Texts and Artefacts 9780567677716, 9780567677723, 9780567677709

Citation preview

LIBRARY OF NEW TESTAMENT STUDIES

584 Formerly Journal of the Study of the New Testament Supplement Series

Editor Chris Keith

Editorial Board Dale C. Allison, John M.G. Barclay, Lynn H. Cohick, R. Alan Culpepper, Craig A. Evans, Robert Fowler, Simon J. Gathercole, John S. Kloppenborg, Michael Labahn, Love L. Sechrest, Robert Wall, Steve Walton, Catrin H. Williams

TEXTS AND ARTEFACTS

Selected Essays on Textual Criticism and Early Christian Manuscripts

Larry W. Hurtado

T&T CLARK Bloomsbury Publishing Plc 50 Bedford Square, London, WC1B 3DP, UK 1385 Broadway, New York, NY 10018, USA BLOOMSBURY, T&T CLARK and the T&T Clark logo are trademarks of Bloomsbury Publishing Plc First published 2018 Paperback edition first published in 2019 Copyright © Larry W. Hurtado, 2018 Larry W. Hurtado has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as Author of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Hurtado, Larry W., 1943- author. Title: Texts and artefacts : selected essays on textual criticism and early Christian manuscripts / Larry W. Hurtado. Description: New York : Bloomsbury T&T Clark, 2017. |Series: Library of New Testament studies ; volume 584 | Includes bibliographical references and index. Identifiers: LCCN 2017032160 (print) | LCCN 2018000830 (ebook) | ISBN 9780567677709 (epdf) | ISBN 9780567677716 (hb : alk. paper) Subjects: LCSH: Bible. New Testament–Criticism, Textual.| Christian literature, Early–Criticism, Textual. Classification: LCC BS2325 (ebook) | LCC BS2325 .H87 2017(print) | DDC 225.4/86–dc23 LC record available at https://lccn.loc.gov/2017032160 ISBN: HB: 978-0-5676-7771-6 PB: 978-0-5676-8882-8 ePDF: 978-0-5676-7770-9 Series: Library of New Testament Studies, volume 584 Typeset by Forthcoming Publications (www.forthpub.com) To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

to Eldon Jay Epp, teacher and friend

C on t en t s

Reprint Permissions ix Preface xi Introduction xiii Part I Text-Critical and Text-Historical Studies Chapter 1 The New Testament in the Second Century: Text, Collections and Canon

3

Chapter 2 The Early New Testament Papyri: A Survey of Their Significance

28

Chapter 3 New Testament Scholarship and the Dating of New Testament Papyri

48

Chapter 4 God or Jesus? Textual Ambiguity and Textual Variants in Acts of the Apostles

64

Part II Manuscripts as Artefacts Chapter 5 The ‘Meta-Data’ of Earliest Christian Manuscripts

83

Chapter 6 Manuscripts and the Sociology of Early Christian Reading

99

Chapter 7 The Origin of the Nomina Sacra: A Proposal

115

viii Contents

Chapter 8 The Staurogram in Early Christian Manuscripts: The Earliest Visual Reference to the Crucified Jesus?

136

Chapter 9 A Fresh Analysis of Papyrus Oxyrhynchus 1228 (P22) as Artefact

155

Chapter 10 The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654, and Papyrus Oxyrhynchus 655

167

Chapter 11 Who Read Early Christian Apocrypha?

182

Chapter 12 P45 as an Early Christian Artefact: What it Reflects about Early Christianity

200

Index of References Index of Authors

220 228

R ep r i n t P er m i ssi ons

With the exception of the third one, the chapters appearing in this volume were all previously published as journal articles and essays. The author and Bloomsbury T&T Clark wish to thank the respective publishers for their generosity in allowing the studies to be republished in the present volume. The chapters were first printed in the following publications: 1.

‘The New Testament in the Second Century: Text, Collections and Canon’. Pages 3-27 in Transmission and Reception: New Testament Text-Critical and Exegetical Studies. Edited by J. W. Childers and D. C. Parker. Texts and Studies. Piscataway, NJ: Gorgias, 2006.

2.

‘The Early New Testament Papyri: A Survey of Their Significance’. Pages 1–18 in Papyrologie und Exegese: Die Auslegung des Neuen Testaments im Licht der Papyri. Edited by Jens Herzer. Tübingen: Mohr Siebeck, 2012.

4.

‘God or Jesus? Textual Ambiguity and Textual Variants in Acts of the Apostles’. Pages 239–54 in Texts and Traditions: Essays in Honour of J. Keith Elliott. Edited by Peter Doble and Jeffrey Kloha. Leiden: Brill, 2014.

5.

‘The “Meta-Data” of Earliest Christian Manuscripts’. Pages 149–63 in Identity and Interaction in the Ancient Mediterranean: Jews, Christians and Others – Essays in Honour of Stephen G. Wilson. Zeba A. Crook and Philip A. Harland. Sheffield: Sheffield Phoenix, 2007.

6.

‘Manuscripts and the Sociology of Early Christian Reading’. Pages 49–62 in The Early Text of the New Testament. Edited by Charles E. Hill and Michael J. Kruger. Oxford: Oxford University Press, 2012.

7.

‘The Origin of the Nomina Sacra: A Proposal’. JBL 117 (1998): 655–73.

8.

‘The Staurogram in Early Christian Manuscripts: The Earliest Visual Reference to the Crucified Jesus? Pages 207–26 in New Testament Manuscripts: Their Text and Their World. Edited by Thomas J. Kraus and Tobias Nicklas. Texts and Editions for New Testament Study 2. Leiden: Brill, 2006.

x

9.

Reprint Permissions

‘A Fresh Analysis of P.Oxyrhynchus 1228 (P22) as Artefact’. Pages 206–16 in Studies on the Text of the New Testament and Early Christianity: Essays in Honor of Michael W. Holmes on the Occasion of His 65th Birthday. Daniel M. Gurtner, Juan Hernández Jr. and Paul Foster. Leiden: Brill, 2015.

10. ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’. Pages 19–32 in Das Thomasevangelium: Entstehung – Rezeption – Theologie. Edited by Jörg Frey, Enno Edzard Popkes and Jens Schröter. BZNW 157. Berlin: de Gruyter, 2008. 11. ‘Who Read Early Christian Apocrypha?’ Pages 153–66 in Oxford Handbook of Early Christian Apocrypha. Edited by Christopher Tuckett and Andrew Gregory. Oxford: Oxford University Press, 2015. 12. ‘P45 as Early Christian Artefact: What it Reflects about Early Christianity’. Teologisk Tidsskrift 4 (2016): 291–307.

P r efa ce

It is an honour to have been invited by Chris Keith, Editor of the LNTS series, to put together this volume, and I am grateful for this opportunity to bring together a number of my essays originally published in various settings. I am also grateful to Nathan Shedd, Managing Editor for the LNTS series, who generously took on the task of obtaining the many permissions from previous publishers to re-publish these essays. Duncan Burns gave them an excellent copy-editing and prepared the indexes. Indeed, I extend my thanks to all the team at Bloomsbury T&T Clark for their various contributions to the publication of this volume, which included converting essays from several different formats into a cohesive volume. I dedicate this book to my highly esteemed teacher, PhD supervisor (from many years ago), and valued colleague and friend, Eldon Jay Epp. It was from him that I learned the practice of New Testament textual criticism, a discipline which also contributed greatly to my larger practice of historical inquiry into the origins of Christianity. He took me on as a PhD student, recommended me for fellowship-funding for my studies, was both demanding and encouraging through my doctoral work, recommended me effectively for my first academic positions, and has been a friend over the years. He taught me to see textual variants, not as textual ‘corruptions’, but as evidence of how the New Testament writings were read and transmitted, including efforts to make them meaningful in various early settings. I first learned of the nomina sacra from his pointing to the phenomenon as indicative of a shared scribal practice across the otherwise varied Christianity of the earliest centuries. He has also repeatedly urged greater attention to the early New Testament papyri, and his own studies of these items were part of what inspired me to give them close attention. His work in New Testament textual criticism made him internationally respected early on, his status as one of the leading figures in the field confirmed thereafter over the years. It is my great honour to have been one of his PhD students. I cannot be sure that these essays all meet his exacting standards, but I hope, in any case, that this volume will serve as my tribute to him and to his many scholarly contributions.

Abbreviations

ABD AJP ANF ANRW

ANTF APF BASPSup BETL BIOSCS BJRL BZNW CQR CRBR CR:BS EC EncJud ETL FRLANT GCS HTR HUCA JBL JQR JSNT JSNTSup JTS LDAB LSTS

Anchor Bible Dictionary. Edited by David Noel Freedman. 6 vols. New York: Doubleday, 1992 American Journal of Philology Ante-Nicene Fathers Aufstieg und Niedergang der römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung. Part 2, Principat. Edited by Hildegard Temporini and Wolfgang Haase. Berlin: de Gruyter, 1972– Arbeiten zur neutestamentlichen Textforschung Archiv für Papyrusforschung Bulletin of the American Society of Papyrologists Supplements Bibliotheca Ephemeridum Theologicarum Lovaniensium Bulletin of the International Organization for Septuagint and Cognate Studies Bulletin of the John Rylands University Library of Manchester Beihefte zur Zeitschrift für die neutestamentliche Wissenschaft Church Quarterly Review Critical Review of Books in Religion Currents in Research: Biblical Studies Early Christianity Encyclopedia Judaica. Edited by Fred Skolnik and Michael Berenbaum. 2nd ed. 22 vols. Detroit: Macmillan Reference USA, 2007 Ephemerides Theologicae Lovanienses Forschungen zur Religion und Literatur des Alten und Neuen Testaments Die griechischen christlichen Schriftsteller der ersten [drei] Jahrhunderte Harvard Theological Review Hebrew Union College Annual Journal of Biblical Literature Jewish Quarterly Review Journal for the Study of the New Testament Journal for the Study of the New Testament Supplement Series Journal of Theological Studies Leuven Database of Ancient Books The Library of Second Temple Studies

xiv Abbreviations NHS NovTSup NTOA NTS NTTS NTTSD OTP PAAJR RB RHR SCJ SD SNTSMS SNTSU SP SPap SPAW.PH ST STAR STDJ TC TDNT TS TU TynBul VC WUNT ZAC ZPE

Nag Hammadi Studies Supplements to Novum Testamentum Novum Testamentum et Orbis Antiquus New Testament Studies New Testament Tools and Studies New Testament Tools, Studies, and Documents Old Testament Pseudepigrapha. Edited by James H. Charlesworth. 2 vols. New York: Doubleday, 1983, 1985 Proceedings of the American Academy of Jewish Research Revue biblique Revue de l’histoire des religions Studies in Christianity and Judaism Studies and Documents Society for New Testament Studies Monograph Series Studien zum Neuen Testament und seiner Umwelt Sacra Pagina Studia Papyrologica Sitzungsberichte der preussischen Akademie der Wissenschaften. Philosophisch-Historische Studia Theologica Studies in Theology and Religion Studies on the Texts of the Desert of Judah TC: A Journal of Textual Criticism Theological Dictionary of the New Testament. Edited by Gerhard Kittel and Gerhard Friedrich. Translated by Geoffrey W. Bromiley. 10 vols. Grand Rapids: Eerdmans, 1964–76 Theological Studies Texte und Untersuchungen Tyndale Bulletin Vigiliae Christianae Wissenschaftliche Untersuchungen zum Neuen Testament Zeitschrift für Antikes Christentum/Journal of Ancient Christianity Zeitschrift für Papyrologie und Epigraphik

I n t rod uct i on

The essays included in this volume, nearly all originally published in various journals and multi-author works, reflect my conviction that the analysis of earliest Christian manuscripts should be a central feature of the study of early Christianity. Of course, establishing a critical edition of any text, and tracing its textual history as well, requires us to take account of the manuscript witnesses. But, as I show in these essays, the relevance of early Christian manuscripts extends well beyond the familiar tasks of textual criticism. For these manuscripts are not only copies of texts. Their physical and visual properties are also important data, and the Christian manuscripts of the second and third centuries are among the earliest (and, unfortunately, often overlooked) Christian artefacts.1 Though often fragmentary, these manuscripts provide us with direct and fascinating access to various features of the young Christian movement. Early Christianity was characterized by a remarkable investment of efforts and resources in the production, distribution, and use of texts, and these early manuscripts are the physical remains of all that activity.2 They are the artefacts of the copying, reading and study of texts. They directly illustrate the variety of early Christian texts, those that became part of the familiar scriptural canons, and also the many other texts of various kinds that circulated in the first few centuries. The continuing growth in the number of manuscripts now available to view and examine ‘online’ wonderfully facilitates taking account of them in forming our views of early Christianity. Likewise, online resources such as the Leuven Database of Ancient Books now enable us readily

1.  For a fuller presentation of this view, see L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). 2.  Harry Y. Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven: Yale University Press, 1995), remains essential. Note also my discussion of early Christianity as a distinctively ‘bookish religion’ in L. W. Hurtado, Destroyer of the Gods: Early Christian Distinctiveness in the Roman World (Waco: Baylor University Press, 2016), 105–41.

xvi Introduction

to obtain information on manuscripts and conduct various analyses of such matters as the comparative number of copies of given texts, the comparative number of Christian and non-Christian manuscripts, and the comparative number of bookrolls and codices, century by century.3 Students and scholars do not all have to become papyrologists and palaeog­raphers, but anyone committed to the study of early Christianity should feel required to take account of the various data that manuscripts have to offer to the task. The studies included here were written across almost twenty years, but I trust that even the earliest selected for inclusion remain relevant and worth consulting. The essays are not placed in chronological order, but are organized in two parts, reflecting broadly the two emphases of the body of work that they comprise. I have lightly edited a number of them and have updated footnotes to take account of a number of works published subsequent to the original appearance of these essays. Although each essay has its own focus, they do reflect a shared emphasis. So, there is, unavoidably, some repetition of some matters. Text-Critical and Text-Historical Studies The first four essays address questions about the textual transmission of NT writings. Textual criticism was an early focus for me, and I continue to have a strong interest in this subject. In ‘The New Testament in the Second Century: Text, Collections and Canon’, the question is how the writings that came to form part of the NT were transmitted, particularly in the crucial period of the second century.4 I highlight factors that likely 3.  The Leuven Database of Ancient Books: http://www.trismegistos.org/ldab/. There are other online resources that tend to be devoted specifically to NT manuscripts: e.g., the Institut für Neutestamentliche Textforschung in the University of Muenster: http://egora.uni-muenster.de/intf/index_en.shtml, and the Center for the Study of New Testament Manuscripts: http://www.csntm.org/. 4.  See also Michael W. Holmes, ‘Text and Transmission in the Second Century’, in The Reliability of the New Testament: Bart Ehrman and Daniel Wallace in Dialogue, ed. Robert B. Stewart (Minneapolis: Fortress, 2011), 47–65. My former doctoral student, Lonnie D. Bell, produced a detailed study of the many extant portions of the Gospel of John from the second and third centuries, concluding that they reflect a relatively stable text of this writing: ‘Textual Stability and Fluidity Exhibited in the Earliest Manuscripts of John: An Analysis of the Second/ThirdCentury Fragments with Attention also to the More Extensive Papyri (P45, P66, P75)’ (PhD diss., University of Edinburgh, 2015). A published version is to appear in the NTTSD series (Leiden: Brill).

Introduction

xvii

served to promote a certain textual stability. These include the reading of certain writings in corporate worship, which, I contend, would have helped to stabilize their texts.5 ‘The Early New Testament Papyri: A Survey of Their Significance’ is a review of the NT papyri dated to the second and third centuries. In between the initial publication of this essay and the present, further NT papyri have been identified, and so in this republished form I have updated figures to represent the current number of relevant manuscripts. ‘New Testament Scholarship and the Dating of New Testament Papyri’ originated as an invited address at a symposium on early Christian manuscripts, and appears here for the first time. The dating of early papyri is both important and a matter of judgment, and so there are continuing differences of opinion on the matter. NT scholars have sometimes been accused of pushing the dates of NT papyri too early. In this essay I assess this charge. ‘God or Jesus? Textual Ambiguity and Textual Variants in Acts of the Apostles’ arose from my earlier study of Christology in Acts.6 In the course of that study I noticed the curious variations among manuscripts at numerous points in Acts over whether to read Κύριος, Θεός, or Ἰησοῦς. So, the invitation to contribute to a volume honouring Keith Elliott allowed me the opportunity to investigate the matter further. I propose that the variants mainly reflect the efforts of early readers to ‘dis-ambiguate’ the text of Acts by specifying who the ‘Κύριος’ is at points where the referent of the term is not entirely clear. Manuscripts as Artefacts The remaining essays in this volume reflect my emphasis on early Christian manuscripts as physical and visual artefacts. In ‘The “MetaData” of Earliest Christian Manuscripts’ I underscore the point that early manuscripts provide us with much more than copies of texts, reviewing briefly the kinds of additional data that they offer and urging colleagues in the study of early Christianity to take greater account of them.7 5.  For a fuller presentation of the data on this practice, see Brian J. Wright, Communal Reading in the Time of Jesus (Minneapolis: Fortress, 2017). 6.  Larry W. Hurtado, ‘Christology in Acts: Jesus in Early Christian Belief and Practice’, in Issues in Luke–Acts: Selected Essays, ed. Sean A. Adams and Michael Pahl (Piscataway, NJ: Gorgias, 2012), 217–37. 7.  I made a similar plea in an earlier essay: ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra, and the

xviii Introduction

In ‘Manuscripts and the Sociology of Early Christian Reading’, I note the contrast between the elegance and severity of high-quality manuscripts of pagan literary texts and the various features typical of early Christian manuscripts of biblical texts that seem to have served as readers’ aids. I propose that these manuscripts offer us a direct reflection of the social and educational diversity of earliest Christian circles.8 One of my earliest studies of the scribal features of early Christian manuscripts is the essay, ‘The Origin of the Nomina Sacra: A Proposal’. One of the most puzzling features of the ‘nomina sacra’ is the horizontal stroke that is regularly written over the various abbreviated forms, which is not a feature typical of Greek or Latin abbreviations of the time. I propose that this stroke may actually give us a clue to the origin of these identifiably Christian scribal devices, specifically that the practice may have begun with a particular way of writing Jesus’ name. In the nature of the case, any proposal can only be a hypothesis, for the ancient Christians left us no explanation for the nomina sacra. But the advantage of my hypothesis is that it accounts for this curious horizontal stroke, which otherwise is simply a puzzle. Another distinctive feature of some early Christian manuscripts is the device now referred to typically as the ‘staurogram’, a ligature in which the Greek letter Rho is superimposed on the Greek letter Tau. In ‘The Staurogram in Early Christian Manuscripts: The Earliest Visual Reference to the Crucified Jesus?’ I contend that this device was originally appropriated by Christians to serve as a pictographic reference to the crucified Jesus, and so gives us our earliest visual depiction of this, well over a century earlier than the depictions typically cited in histories of early Christian iconography.9

Staurogram’, in Text and Artifact in the Religions of Mediterranean: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo: Wilfrid Laurier University Press, 2000), 271–88. 8.  This essay was prompted by reading the fascinating study by William A. Johnson, ‘Towards a Sociology of Reading in Classical Antiquity’, American Journal of Philology 121 (2000): 593–627, who cogently contends that the severe features of high-quality manuscripts of ‘pagan’ texts intentionally reflect the elite social circles in which they were read. See also my essay, ‘Early Christian Manuscripts as Artifacts’, in Jewish and Christian Scripture as Artifact and Canon, ed. Craig A. Evans and Daniel Zacharias, LSTS 70 (London: T&T Clark International, 2009), 66–80. 9.  But see the acknowledgment of the staurogram as a reference to the crucified Jesus in Robin Margaret Jensen, Understanding Early Christian Art (London: Routledge, 2000), 138.

Introduction

xix

In ‘A Fresh Analysis of Papyrus Oxyrhynchus 1228 (P22) as Artefact’, I provide what I believe is the first detailed analysis of this papyrus fragment of the Gospel of John since the initial publication of it in 1915. P22 is particularly noteworthy as one of the few NT manuscripts that was not a codex. On the basis of an analysis of the physical and visual features of this third-century papyrus, I propose that it likely originated as someone’s personal copy of the Gospel of John, perhaps intended for study or for devotional use. Originally written for a symposium held in Germany on the Gospel of Thomas, in my essay, ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’, I offer an analysis of the physical and visual features of these earliest witnesses to the text, proposing that they reflect a noteworthy interest in it among early Christians. But I also propose that these witnesses show that these copies of the Gospel of Thomas were likely read more for individual edification, and not liturgically as scripture.10 My essay, ‘Who Read Early Christian Apocrypha?’, was written in response to the request to address this question for The Oxford Handbook of Early Christian Apocrypha. The question is not entirely novel, but my own treatment of it is, I believe, a fresh contribution in drawing attention to the physical features of the manuscripts of these texts as possible clues to their usage. Finally, in ‘P45 as an Early Christian Artefact: What it Reflects about Early Christianity’, I highlight features of this manuscript that make it particularly relevant to several historical questions about early Christianity. My purpose in this essay is to illustrate how the physical and visual features of such early manuscripts are relevant to scholarly interests well beyond the circles of papyrologists, including questions about the formation of the fourfold Gospel, the NT canon, and other matters. Summarizing Remarks I hope that the republication of these essays will make them more readily available to students and scholars. But I also hope that thereby these essays will encourage others, particularly younger colleagues, to take account 10.  Cf. AnneMarie Luijendijk, ‘Reading the Gospel of Thomas in the Third Century: Three Oxyrhynchus Papyri and Origen’s Homilies’, in Reading New Testament Papyri in Context / Lire les papyrus du Nouveau Testament dans leur contexte, ed. Claire Clivaz and Jean Zumstein (Leuven: Peeters, 2011), 241–67, who responds to my essay.

xx Introduction

of manuscripts as witnesses to the texts that they bear, and as invaluable artefacts of early Christianity. It is an all too common mistake, even among NT scholars, to regard NT textual criticism and the study of manuscripts as some arcane area to which only specialists in these matters need give much attention. But NT textual criticism and early Christian manuscripts are much too important to be left to text critics and papyrologists!

Part I T e x t - C r i t i ca l a n d T ex t -H i stor i cal S t udi e s

Chapter 1 T he N e w T es ta m en t i n t h e S e cond C e nt ury : T ex t , C ol l ec t i on s a nd C anon *

It has been clear for some time that the second century was a (indeed, perhaps the) crucial period in the development of the NT. The individual writings that comprise the NT were all, or nearly all, written across the latter half of the first century; but the second century is the period in which most of them came to hold a special significance, at least for the great majority of Christian circles. Facilitating and reflecting this, the second century was the time when these writings were copied and disseminated widely. It is all the more frustrating, however, that the extant evidence from the period is almost in inverse relation to its importance. Nevertheless, this importance demands and justifies our efforts to take stock of what we can say, and with what confidence, about the NT writings in the second century. In what follows, I make a modest effort toward this end by underscoring three crucial processes in this period, which also constitute three major areas of scholarly inquiry and controversy: (1) the textual transmission of the NT writings, (2) the phenomenon of early collections of writings (especially the Gospels and Pauline epistles); and (3) certain writings coming to enjoy a special status, authority, and usage, which seems to be the crucial earlier stage of a process that led eventually to a fixed, closed canon of the NT. I shall survey these matters in the light

*  An earlier version of this essay was given as an invited presentation in the New Testament Textual Criticism program unit of the Annual Meeting of the Society of Biblical Literature, 17–20 November 2001, Denver, USA. I particularly acknowledge critical interaction with William L. Petersen then and subsequently in clarifying the discussion of Patristic citation of NT writings, though I must take sole responsibility for the views I offer here. James Kelhoffer also kindly read an earlier draft and gave helpful suggestions.

4

Texts and Artefacts

of current scholarly debates and recently available evidence (e.g., recent Oxyrhynchus volumes). My aim here is not to provide some definitive treatment of any of the data or the issues, but rather to emphasize the importance of these three processes or dynamics, thereby to help to focus further thinking about them. Textual Transmission of the New Testament Writings First, what can we say about the transmission of the text of the NT in the second century? Some scholars emphasize and allege great fluidity and freedom in this period, even alleging that the extant manuscripts (the earliest from the late second and early third century CE) are seriously unreliable as witnesses to the earliest transmission of NT writings and for reconstructing the ‘original’ of these writings. Others, however, contend that the manuscript evidence shows sufficient usefulness to encourage this text-critical effort.1 Some see the undeniable textual fluidity as indicative that the writings held something considerably less than scriptural significance in the earliest centuries, whereas others argue that it shows the opposite. As is reasonably well known, the two main types of evidence that have been used in forming our views of the transmission of NT writings in the second century are, first, the extant manuscripts from that time and early centuries thereafter, and, second, the citations/quotations of NT writings by second-century Christian authors.2 Let us look briefly at recent developments in the study of these bodies of evidence.

1.  Helmut Koester, ‘The Text of the Synoptic Gospels in the Second Century’, in Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission, ed. William L. Petersen (Notre Dame: University of Notre Dame Press, 1989), 19–37, contended that the second century was completely a period of wild variation until sometime near 200 CE when he proposes that a textual recension was undertaken, from which our extant NT manuscripts all derive. Note also the recent problematizing of the task of reconstructing an ‘original’ text of the NT writings: Eldon Jay Epp, ‘The Multivalence of the Term “Original Text” in New Testament Textual Criticism’, HTR 92 (1999): 245–81. 2.  See now especially Barbara Aland, ‘Die Rezeption des neutestamentlichen Textes in den ersten Jahrhunderten’, in The New Testament in Early Christianity, ed. Jean-Marie Sevrin, BETL 86 (Leuven: Peeters, 1989), 1–38. Stuart R. Pickering has complained about inadequate attention given to the potential importance of quotations and allusions to passages in NT writings in early papyri: ‘The Significance of Non-Continuous New Testament Materials in Papyri’, in Studies in the Early Text

1. The New Testament in the Second Century

5

As for the manuscripts of NT writings, there is both bad news and good news. The well-known bad news is that the extant manuscripts that can plausibly be dated to the second century are lamentably few in number, and none of them gives us a complete text of any NT writing.3 In fact, the extant second-century manuscript evidence consists largely in a few incomplete single leaves, though they collectively derive from several manuscripts. Even if we accept Skeat’s argument that P64, P67 and P4 all represent the same multi-gospel manuscript from the late second century, the amount of text preserved in the total body of second-century manuscript material is still frustratingly small.4 The earliest manuscripts that give us substantial portions of texts are dated palaeographically to the early third century or thereabouts. P45 (Gospels and Acts) and P46 (Pauline epistles) date from ca. 200–250 CE, Gospels codices P66 and P75 from ca. 200 CE, P47 (Revelation) ca. 250–300 CE, and P72 (Jude and 1–2 Peter) third to fourth century CE.5

of the Gospels and Acts: The Papers of the First Birmingham Colloquium on the Textual Criticism of the New Testament, ed. D. G. K. Taylor (Birmingham: University of Birmingham Press; Atlanta: Society of Biblical Literature, 1999), 121–41. 3.  E.g., the 28th edition of the Nestle-Aland Novum Testamentum Graece assigns a second-century date only to P52, P90, and P104, and hesitantly to P98. For a recent review of the dating of NT papyri, see now Pasquale Orsini and Willy Clarysse, ‘Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography’, ETL 88 (2012): 443–74. 4.  T. C. Skeat, ‘The Oldest Manuscript of the Four Gospels?’, NTS 43 (1997): 1–34, reprinted in J. K. Elliott, ed., The Collected Biblical Writings of T. C. Skeat, NovTSup 113 (Leiden: Brill, 2004), 158–92; also G. N. Stanton, ‘The Fourfold Gospel’, NTS 43 (1997): 317–46, reprinted (with light revision) in G. N. Stanton, Jesus and Gospel (Cambridge: Cambridge University Press, 2004), 63–91. But cf. Peter M. Head, ‘Is P4, P64 and P67 the Oldest Manuscript of the Four Gospels? A Response to T. C. Skeat’, NTS 51 (2005): 450–57; and Scott D. Charlesworth, ‘T. C. Skeat, P64 + 67 and P4, and the Problem of Fibre Orientation in Codicological Reconstruction’, NTS 53 (2007): 582–604, who argue against Skeat’s proposal; and Charles E. Hill, ‘Intersections of Jewish and Christian Scribal Culture: The Original Codex Containing P4, P64 and P67, and Its Implications’, in Among Jews, Gentiles, and Christians in Antiquity and the Middle Ages, ed. Reidar Hvalvik and J. Kaufman (Trondheim: Tapir Academic, 2011), 75–91, who argues in support of it. 5.  Cf. now Brent Nongbri, ‘The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P. Bodmer II (P66)’, Museum Helveticum 71 (2014): 1–35, who contends that this papyrus could be dated anytime from the late second to the fourth century CE.

6

Texts and Artefacts

Newly Published Manuscripts On the other hand, the good news is that the small fund of second-century and/or early third-century manuscript witnesses has been enriched with the publication of three recent volumes of the Oxyrhynchus papyri.6 Volumes 64–66 give us previously unknown NT papyrus materials that comprise leaves from seven manuscripts of Matthew, four of John, two of Revelation, and one each of Luke, Acts, Romans, Hebrews, and James, the dates of these manuscripts ranging from the second century to the fifth or sixth century CE.7 The earliest are leaves of three manuscripts of Matthew dated to the second or early third century: P.Oxy. 4405 (a new portion of P77, containing Matt 23:30–34, 35–39, second/third century), P.Oxy. 4403 (P103, Matt 13:55–56; 14:3–5, second/third century), P.Oxy. 4404 (P104, Matt 21:34–37, 43, 45, late second century). Prior to the publication of these fragments, per Nestle-Aland 27, the only second-century manuscripts available were the famous P52 (P.Ryl. 457, John 18:31–33, 37–38), P90 (P.Oxy. 3523, John 18:36–19:1; 19:2–7), and (possibly) 6.  Of course, since the original publication of this essay in 2006, several further volumes of Oxyrhynchus Papyri have appeared, some containing additional early NT papyri. But, if anything, these only serve to strengthen the basis for the arguments that I present here. In a later essay, I focus on the NT papyri dated to the second and third centuries: Larry W. Hurtado, ‘The Early New Testament Papyri: A Survey of Their Significance’, in Papyrologie und Exegese: Die Auslegung des Neuen Testaments im Licht der Papyri, ed. Jens Herzer (Tübingen: Mohr Siebeck, 2012), 1–18 (republished as Chapter 2 of the present volume). 7.  The Oxyrhynchus Papyri (London: British Academy for the Egypt Exploration Society): Volume LXIV, ed. E. W. Handley et al., 1997, has P.Oxy. 4401–6 (pp. 1–13, ed. J. David Thomas); Volume LXV, ed. M. W. Haslam et al., 1998, has P.Oxy. 4445–48 (pp. 10–20, ed. W. E. H. Cockle), and P.Oxy. 4449 (pp. 20–25, ed. R. Hübner); Volume LXVI, ed. N. Gonis et al., 1999, has P.Oxy. 4494–95 (pp. 1–5) and P.Oxy. 4497–98 (pp. 7–10, ed. W. E. H. Cockle), P.Oxy. 4496 (pp. 7–10, ed. Tim Finney), P.Oxy. 4499 (pp. 10–35, ed. J. Chapa), and P.Oxy. 5000 (ed. W. E. H. Cockle). These comprise a new portion of P77 (P.Oxy. 4405 part of the same codex as P.Oxy. 2683), and several newly identified manuscripts assigned NT papyri numbers P100–115. Basic information and images are available on the Oxyrhynchus web site: http://www.papyrology.ox.ac.uk/. See especially the valuable discussion by Peter M. Head, ‘Some Recently Published NT Papyri from Oxyrhynchus: An Overview and Preliminary Assessment’, TynBul 51 (2000): 1–16, which includes ample citation of other relevant publications; and J. K. Elliott, ‘Five New Papyri of the New Testament’, NovT 41(1999): 209–13, reviews the NT fragments in Volume 65 of the Oxyrhynchus series, and focuses almost entirely on what readings they contain.

1. The New Testament in the Second Century

7

P98 (P.IFAO 237b, Rev 1:13–20). Even if we add in the recently posited manuscript combination of P4-P64-P67 mentioned already, and grant the (debatable) proposal that the manuscript dates from the late second century, and if we also add NT papyri usually dated ca. 200 CE, such as P66, P75, P46, it is still clear that the very recent Oxyrhynchus fragments add significantly to a very limited body of manuscript material for the second century. Moreover it is further good news that, although comprising a small amount of the text of NT writings, these fragments are actually rich with data. From my own consultation of the relevant Oxyrhynchus volumes and from Peter Head’s valuable survey of these fragments, I mention a few illustrative matters. First, though we have in each case only a small sample of the manuscript from which they come, in general these fragments ‘confirm the text of the great uncials which forms the basis of the modern critical editions’.8 In the main, they provide us with earlier attestation of variants that we already knew of from later witnesses. In some cases, however, these are variants previously attested only in the versions, which warns us that in a good many other cases as well readings presently supported only in the versions may well reflect very early readings that simply happen not to have survived in the extant Greek witnesses.9 But the larger point is that these very early fragments further encourage us to think that the more substantial witnesses from the third century and later are (contra Koester) probably not the results of some supposed major recension of NT writings initiated toward the end of the second century.10 Indeed, the Oxyrhynchus fragments further justify the view that the more substantial early third-century papyri are broadly reliable witnesses of the text of the writings that they contain, as these writings had been transmitted across the second century. Second, these fragments also reinforce the impression given by the NT papyri from 200 CE and a bit later that there were somewhat varying scribal tendencies operative in the textual transmission of the NT in the

8.  Head, ‘Some Recent Published NT Papyri from Oxyrhynchus’, 16. 9.  Perhaps the most significant variant is in P.Oxy. 4445 (P106) at John 1:34, ὁ ἐκλεκτός, where most witnesses, including P66 and P75 read ὁ υἱός. P106 here gives early support for a reading found also in Sinaiticus (original hand), later minuscules (77, 218), Old Latin manuscripts (b, e, ff2*), and the Old Syriac (syrs, c). See Head, ‘Some Recently Published NT Papyri’, 11. 10.  Helmut Koester, ‘The Text of the Synoptic Gospels in the Second Century’, in Petersen, ed., Gospel Traditions in the Second Century, 19–37.

8

Texts and Artefacts

second century.11 That is, the recently published evidence is consistent with the view that the second century was a time of somewhat diverse textual dynamics. To quote from Head’s survey, the fragments ‘illustrate various points along the spectrum from more controlled texts (with corrections, literary features, etc.) to comparatively more free or careless copying’.12 We are thereby further warned against over-simplifications about the textual transmission of NT writings in the second century. Instead, with enhanced confidence we may take up Epp’s proposal that the early NT papyri can be placed in several early ‘clusters’ or ‘textual groups’, and that these represent different ‘textual complexions’ already operative in the second century. Some of the newly published fragments reflect a concern for ‘a high degree of accuracy’, and others indicate a freer readiness to adapt the text, exhibited especially in stylistic changes, harmonizations, higher numbers of accidental changes, and even occasional changes motivated by doctrinal concerns.13 To avoid misunderstanding in this controversial matter, I emphasize my point. I do not deny at all that there was (perhaps considerable) variety of practice in the transmission of the NT writings in the second century.14 I simply stress that along with a readiness of some (perhaps even most)

11.  See esp. James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTS 36 (Leiden: Brill, 2007); and his earlier brief discussion, ‘Scribal Tendencies in the Transmission of the Text of the New Testament’, in The Text of the New Testament in Contemporary Research, ed. Bart D. Ehrman and Michael W. Holmes, SD 46 (Grand Rapids: Eerdmans, 1995), 239–52. 12.  Head, ‘Some Recently Published NT Papyri’, 10. 13.  E. J. Epp, ‘The Significance of the Papyri for Determining the Nature of the New Testament Text in the Second Century: A Dynamic View of Textual Transmission’, in E. J. Epp and Gordon D. Fee, Studies in the Theory and Method of New Testament Textual Criticism, SD 45 (Grand Rapids: Eerdmans, 1993), 274–97. See the somewhat similar categories proposed by Aland, ‘Die Rezeption des neutestamentliches Textes’, 26–27. 14.  The so-called Western Text is perhaps the most striking expression of the fluidity in textual transmission in this early period. But it appears that the ‘Western Text’ is probably more a body of readings produced by somewhat similar scribal tendencies, rather than a cohesive recension. I would now also add that there is no early manuscript evidence of the more substantial variants that distinguish Codex Bezae (fifth century CE). For discussions of the ‘Western Text’, see, e.g., J. N. Birdsall, ‘The Western Text in the Second Century’, in Petersen, ed., Gospel Traditions in the Second Century, 3–17; Joël Delobel, ‘The Text of Luke–Acts: A Confrontation of Recent Theories’, in The Unity of Luke–Acts, ed. J. Verheyden (Leuven: Leuven University Press, 1999), 83–107; and cf. Christopher Tuckett, ‘How Early Is the “Western”

1. The New Testament in the Second Century

9

scribes to introduce variants intended to harmonize the Gospels, remove ambiguities, affirm doctrinal concerns, and even introduce new material intended as edifying, in at least some circles there also appears to have been a somewhat more conservative copying attitude. In addition to their readings, however, the small but fascinating body of early papyri gives us other valuable evidence that should not be overlooked. New Testament scholars, including text critics, have tended to comb early manuscripts for readings; but we also must learn to harvest the fuller and more diverse data that lie in these valuable artefacts.15 For example, the corrections in P.Oxy. 4403 (P103) and P.Oxy. 4405 (P77) are noteworthy. The quality of the hands suggests to me that these manuscripts were not produced by professional calligraphers such as those who produced expensive copies of literary texts.16 Nevertheless, along with some other features, these corrections reflect the sort of mentality (though not the fully developed scribal skills) that we associate later with a scriptorium. In particular, the corrections show a concern for what those correcting the copies regarded as accurate copying. Of course, we must be careful to avoid anachronism in positing too confidently formal scriptoria early in the second century, at least in the sense of the sort of physical settings in which multiple copies of Christian writings were prepared

Text of Acts?’, in The Book of Acts as Church History / Apostelgeschichte als Kirchengeschichte, ed. Tobias Nicklas and Michael Tilly (Berlin: de Gruyter, 2003), 69–86; and W. A. Strange, The Problem of the Text of Acts, SNTSMS 71 (Cambridge: Cambridge University Press, 1992). 15.  Günther Zuntz, The Text of the Epistles: A Disquisition Upon the Corpus Paulinum (London: Oxford University Press for the British Academy, 1953), is a classic study of P46 that demonstrates a fuller use of the data available in the early papyri. Among studies of the major codices, D. C. Parker, Codex Bezae: An Early Christian Manuscript and Its Text (Cambridge: Cambridge University Press, 1992), is a model. I have subsequently emphasized the physical and visual features of early manuscripts in my book, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), and in several essays republished in this volume. 16.  But see now Alan Mugridge, Copying Early Christian Texts, WUNT 2/362 (Tübingen: Mohr-Siebeck, 2016), who contends that many early Christian manuscripts show the services of ‘trained scribes’, and often refers to them as ‘professional’ scribes, i.e., individuals who received payment for their services. I think, however, that he too readily takes evidence of competence in copying as reflecting such ‘professional’ copyists. Cf., also Kim Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature (Oxford: Oxford University Press, 2000).

10

Texts and Artefacts

in later centuries.17 Nevertheless, there are various indications that the copying of early Christian texts in the second century involved emergent scribal conventions that quickly obtained impressive influence, and, at least in some cases and settings, that there was a concern for careful and accurate copying.18 To cite one particular matter, where these new fragments preserve the words known to us as nomina sacra, particularly the four words Θεός, Κύριος, Χριστός, and Ιησούς, they are rather consistently written in the sorts of abbreviated forms that we already know from other ancient Christian manuscripts.19 The minor variations in the precise spelling of the abbreviations do not rightly count against the conclusion that there was a widely attested convention among Christian scribes that certain religiously ‘loaded’ words were to be written in a distinctive manner.20 In so far as earliest Christian manuscripts were not copied by ‘professional’ scribes (or at least often do not exhibit the kind of calligraphy more characteristic of professionally produced literary manuscripts of the period), such widespread and distinctive scribal conventions are all the more notable. They indicate an emergent Christian scribal practice that spread widely and quickly. It is also significant that all of these fragments come from codices. Thus, they collectively reinforce the conclusions drawn from previously known evidence that, by sometime early in the second century 17.  Cf., e.g., Harry Y. Gamble, Books and Readers in the Early Church (New Haven: Yale University Press, 1995), 120–23; Haines-Eitzen, Guardians of Letters, 83–91. Part of the difficulty in the issue is that scholars are not always clear as to what they mean by ‘scriptorium’. 18.  See, e.g., Zuntz, The Text of the Epistles, esp. 263–83. 19.  Unfortunately, the new fragments dated to the second century (P.Oxy. 4405/ P77; P.Oxy. 4403/P103; and P.Oxy. 4404/P104) do not preserve portions of text where the words in question would have occurred. But newly published third-century fragments do: P.Oxy. 4449 (P100), P.Oxy. 4401 (P101), P.Oxy. 4445 (P106), P.Oxy. 4447 (P108), P.Oxy. 4495 (P111), P.Oxy. 4497 (P113), P.Oxy. 4498 (P114), P.Oxy. 4499 (P115). See the table of features in Head, ‘Some Recently Published NT Papyri’, 5. On the nomina sacra, see esp. Colin H. Roberts, Manuscript, Society and Belief in Early Christian Egypt, The Schweich Lectures 1977 (London: Oxford University Press for the British Academy, 1979), 26–48; Larry W. Hurtado, ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998): 655–73 (republished as Chapter 7 of the present volume); and idem, The Earliest Christian Artifacts, 95–134, for a full discussion and bibliographic references. As well, I discuss the matter briefly in other essays in the present volume. 20.  Thus, I do not find Haines-Eitzen’s effort to minimize the significance of the nomina sacra persuasive; cf. Guardians of Letters, 91–96.

1. The New Testament in the Second Century

11

at the latest, Christians overwhelmingly had come to prefer the codex, especially, it appears, for their inherited scriptures (Old Testament) and the Christian writings that were coming to be treated widely as scripture.21 The only extant examples of Christian texts written on unused bookrolls (as distinguished from re-used rolls, ‘opisthographs’) are theological tractates (e.g., Irenaeus’ Against Heresies), and writings that may have been regarded as edifying in some circles but did not gain acceptance as part of the emergent NT canon.22 These new Oxyrhynchus fragments of NT writings also exhibit various aids to reading, such as rough-breathing marks, punctuation, and, a matter of particular significance, occasional spacing at the ends of sentences and perhaps paragraphs.23 These readers-aids are very unusual for literary texts of the period, but there are some similarities to pre-Christian Jewish manuscripts of the OT writings (e.g., P.Ryl. 458). The most cogent inference is that the Christian manuscripts with these various scribal devices were prepared for ease of public reading in churches. That is, these small fragments probably give us further important artefactual evidence confirming second-century reports (e.g., Justin Martyr) of the liturgical practice of reading these NT writings.24 Though inadequately 21.  Cf. P.Oxy. 4443, a fragment of Greek Esther dated to the late first or early second century CE from a bookroll or ‘scroll’ (‘a luxurious copy’), and with at least one occurrence of an uncontracted θεός (plus two other cases proposed for lacunae), which rightly led the editors to assign the manuscript to a Jewish provenance (The Oxyrhynchus Papyri, Volume LXV, ed. M. W. Haslam, 4). For further discussion of the early Christian preference for the codex, see esp. Hurtado, The Earliest Christian Artifacts, 43–93. 22.  E.g., P.Oxy. 3405 (van Haelst 671) is a late second or early third-century roll of Irenaeus, Against Heresies, as is the fourth-century copy, van Haelst 672. The Fayoum fragment (P.Vindob. G. 2325; van Haelst 589) is a roll. P.Oxy. 654 (van Haelst 593) is an opisthograph, but P.Oxy. 655 (van Haelst 595) is another copy of Gospel of Thomas written on a fresh roll. P.Mich. 130 (van Haelst 657) is a late second-century opisthograph of Shepherd of Hermas, whereas P.Berlin inv. 5513 (van Haelst 662, third-century) is another copy of Hermas on an unused roll. 23.  Again, I refer to the table in Head, ‘Some Recently Published NT Papyri’, 5. Further details are given in the relevant Oxyrhynchus volumes. See also Eldon Jay Epp, ‘The New Testament Papyri at Oxyrhynchus in their Social and Intellectual Context’, in Sayings of Jesus: Canonical and Non-Canonical: Essays in Honour of Tjitze Baarda, ed. William L. Petersen, Johan S. Vos, and Henk J. de Jonge, NovTSup 89 (Leiden: Brill, 1997), 47–68 (and Epp’s ‘Appended Note’, which appears in offprints of his essay and takes account of Oxyrhynchus volumes that appeared after the essay was written). 24.  See, e.g., Gamble, Books and Readers, 205–8, 211–31.

12

Texts and Artefacts

noticed, such evidence was already provided in previously available fragments such as the famous fragment of the Gospel of John, P52 (P.Ryl. 457), which exhibits the diaresis (over an initial iota) and curious spaces that seem to register clauses, that is, places where the ‘public’ readers were probably intended to make small pauses.25 It is an unfortunate weakness in Kim Haines-Eitzen’s recent (and in a number of other matters, very helpful) study of early Christian scribal practice that she rather too simply assumes the general literary practice of making private copies for personal usage as the operative setting and model for the production of all early Christian manuscripts. I think that she gives inadequate attention to strong indications that a good many Christian manuscripts were prepared for groups and for reading out as liturgical texts.26 In the following paragraphs, I cite briefly the important matters. Already by the date of our earliest extant evidence, Christians had come to prefer a distinguishable book-form (the codex) over against the wider preference of their culture for the bookroll. As noted earlier, the Christian preference for the codex seems to have been especially strong in copying their most revered writings, those that were regarded as scripture and/ or were coming to be so regarded. Moreover, Christians also developed distinctive scribal practices, among which the nomina sacra are the most striking, but including also the use of punctuation and spacing. They read texts, not simply privately or in the sort of reading circles of the cultured elite, but also, very importantly and characteristically, as a regular part of their liturgical practice and thus as a feature of their gathered worship. In this, they differed from the literary and religious practices of the larger culture. Reading texts did not typically feature in Roman-era cultic practices/settings. The only obvious precedent and analogy for the early Christian religious usage of texts was in the reading of scripture as part of Jewish synagogue practice. All of this cumulatively signals what must be seen as the emergence of an identifiably and somewhat distinctive Christian literary ethos. Indeed, in an essay in the Peter Richardson 25.  Diaresis in recto lines 1 and 2, and verso line 2; spacing recto lines 2 and 3, and verso line 2. C. H. Roberts commented on the spacing he found in P.Ryl. 458 (Greek Deuteronomy, second century BCE) at the ends of sentences or clauses and groups of words, noting how unusual such spacing was, and also that a similar system might be identified in P52 (‘Two Biblical Papyri in the John Rylands Library Manchester’, BJRL 20 [1936]: 219–36, esp. 226–27). 26.  Haines-Eitzen, Guardians of Letters. For a broader and more fully nuanced treatment of the production and use of early Christian texts, see esp. Gamble, Books and Readers in the Early Church.

1. The New Testament in the Second Century

13

Festschrift I have proposed that the early Christian manuscripts offer us our earliest indications of an emergent Christian ‘material culture’.27 The reason I underscore these matters is that there is increasing recognition that the repeated liturgical reading of NT writings is an important factor in their textual transmission. It certainly helps to account for the obviously frequent copying and wide dissemination of these writings, which goes far beyond anything else in antiquity.28 Furthermore, liturgical usage is one of the factors that would have helped to prompt the sort of small stylistic ‘improvements’ intended to make texts clearer and easier to understand that are so well known in Christian manuscripts. The regular liturgical reading of the four canonical Gospels also helps to account for the abundance of harmonizing variants, which are especially frequent in manuscripts of Mark. But repeated public reading of NT writings would also have set real limits on how much a writing could be changed, at least in a given circle, without people noticing (and probably objecting), as anyone familiar with what happens when liturgical changes are introduced can attest. It is, thus, likely not a coincidence that Mark, which appears to have been the least widely and frequently used in liturgical reading, also exhibits the largest number and the most salient variations (especially, of course, the several endings). By contrast, the most widely used Gospel, Matthew, has probably the most stable and fixed text. That is, the practice of repeated liturgical reading of NT writings is yet another factor that ought to lead us to hesitate to characterize the second century as basically a period of ‘wild’ textual tendencies. Along with the surprisingly well-attested preference for the codex and the ubiquitous scribal treatment of the nomina sacra, the practice of liturgical reading of writings provides us with indications of a conventionalization of practice with regard to these writings at a chronological stage of early Christianity to which we are otherwise accustomed to attributing great diversity. It is certainly the case, however, that we have not adequately mined all that is provided to us in the early papyri, whether those that have been known for some time or those newly published. There are some valuable 27.  Larry W. Hurtado, ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo, ON: Wilfrid Laurier University Press, 2000), 271–88. 28.  See now my discussion of the ‘bookish’ nature of early Christianity in Destroyer of the Gods: Early Christian Distinctiveness in the Roman World (Waco: Baylor University Press, 2016), 105–41.

14

Texts and Artefacts

research projects here. For example, Harry Sanders noticed long ago that Codex W (the four Gospels) exhibits a system of spacing to signal sense-units that corresponded with this device in versional evidence, and he proposed that this might reflect ‘an ancient system of phrasing, used in reading the Scriptures in church service’, whose origin ‘must have been as early as the second century’.29 The spacing found in early papyri that have appeared subsequently seems to support Sanders’ suggestion. Already in the second century (although perhaps not everywhere) there appears to have been an embryonic system of subdivision of the texts of the Gospels that probably reflected and supported the practice of Gospel readings as part of Christian worship gatherings.30 But it remains for us to mine the relevant material on this and other intriguing matters. Prospective doctoral students, take note!31 Second-Century Citations The other major body of data that has often been taken as giving evidence about the transmission of the NT writings in the second century are the citations and quotations in second-century Christian writers.32 In very 29.  H. A. Sanders, The New Testament Manuscripts in the Freer Collection. Part I, The Washington Manuscript of the Four Gospels (New York: Macmillan, 1912), 14. 30.  See further Victor Martin’s discussion of spacing signalling subdivisions in P66: Papyrus Bodmer II, Evangile de Jean chap. 1–14 (Cologny-Genève: Bibliothèque Bodmer, 1956), 18–21. 31.  Note the series, ‘Pericope: Scripture as Written and Read in Antiquity’, established in 2000, and published by Van Gorcum (Assen, Netherlands) under the editorship of Marjo Korpel (Utrecht) et al., which focuses on scribal ‘unit delimitation’ in biblical manuscripts (Hebrew, Greek, Syriac, Latin). 32.  Studies include Donald A. Hagner, ‘The Sayings of Jesus in the Apostolic Fathers and Justin Martyr’, in Gospel Perspectives: The Jesus Tradition Outside the Gospels, ed. David Wenham (Sheffield: JSOT, 1984), 233–68; A Committee of the Oxford Society of Historical Theology, The New Testament in the Apostolic Fathers (Oxford: Clarendon, 1905); Helmut Köster, Synoptische Überlieferung bei den apostolischen Vätern, TU 65 (Berlin: Akademie-Verlag, 1957); Edouard Massaux, The Influence of the Gospel of St. Matthew on Christian Literature Before St. Irenaeus. Part 1, The First Ecclesiastical Writers (Macon, GA: Mercer University Press, 1990); Arthur J. Bellinzoni, ‘The Gospel of Matthew in the Second Century’, SecCent 9 (1992): 197–258; idem, The Sayings of Jesus in the Writings of Justin Martyr, NovTSup 17 (Leiden: Brill, 1967). See also the contributions in the two volumes edited by Andrew Gregory and Christopher Tuckett: The Reception of the New Testament in the Apostolic Fathers (Oxford: Oxford University Press, 2005), and Trajectories through the New Testament and the Apostolic Fathers (Oxford: Oxford University Press, 2005).

1. The New Testament in the Second Century

15

basic terms, examination of second-century Christian writers indicates few explicit citations of NT writings; and, where it is clear or highly likely that a NT writing is quoted, the quotation often exhibits curious differences from the text of the writing that is dominant in the extant manuscripts. To be a bit more precise, these things tend to be comparatively more characteristic of Christian writers of the first half of the second century. Noting this, Barbara Aland has proposed that in the latter half of the second century we see the emergence of a ‘text-consciousness’ that is reflected in a more frequently explicit (named) and more exact citation of NT writings. This increased ‘text-consciousness’, she proposes further, may have developed as a result of two major processes: (1) secondcentury controversies over Christian faith that involved questions about the wording of texts (e.g., Marcion), and (2) the prolonged effect of repeated liturgical reading of certain texts.33 As noted earlier, Koester, and William Petersen also, in particular have argued that the loose and fluid wording of the quotations of NT writings in early second-century authors means that the text of NT writings was then considerably more fluid than is reflected in the extant manuscripts.34 Indeed, they have contended that the evidence of the early citations should be preferred over the extant manuscripts of the NT writings in characterizing their textual transmission in the second century. But, as Barbara Aland complained in her 1989 essay, the analysis of second-century Christian citations has tended too much to proceed with insufficient attention to the wider literary practices of the time.35 That is, the import of the citation practices reflected in early second-century Christian authors may not be as obvious or as decisive as has sometimes been assumed. We still do not have the thorough-going ‘history of citation in antiquity’ that Eduard Norden urged long ago.36 But we do have Christopher Stanley’s valuable study of the citations of OT writings in Paul, in which Stanley 33.  Aland, ‘Die Rezeption des neutestamentlichen Textes’, 5–21. 34.  Koester, ‘The Text of the Gospels in the Second Century’; William L. Petersen, ‘What Text Can New Testament Textual Criticism Ultimately Reach?’, in New Testament Textual Criticism, Exegesis and Church History: A Discussion of Methods, ed. Barbara Aland and J. Delobel (Kampen: Kok-Pharos, 1994), 136–52; idem, ‘The Genesis of the Gospels’, in New Testament Textual Criticism and Exegesis: Festschrift J. Delobel, ed. A. Denaux (Louvain: Peeters, 2002). 35.  Aland, ‘Die Rezeption des neutestamentlichen Textes’, 2–3. 36.  Eduard Norden, Die antike Kunstprosa vom vi Jahrhundert vor Christus bis in die Zeit der Renaissance, 2 vols. (Leipzig: Teubner, 1898; reprint, Darmstadt: Wissenschaftliche Buchgesellschaft, 1958), 1:88–90.

16

Texts and Artefacts

includes a comparative analysis of the citation of sources in selected early Roman-era pagan writers and in Jewish writings of the period.37 Stanley shows that in Jewish, pagan and Christian writers of the time, the citation of known written sources is impressively free and adaptive. Writers omit words, phrases and whole lines that they deem superfluous or problematic for their own rhetorical aims; and they also add or substitute words and phrases to serve as ‘interpretive renditions’ of the material cited, making the material fit more closely with the context in their own writing in which the citation is appropriated. Likewise, authors frequently combine and conflate material from different contexts of a cited work and/or from different works. So, in general, the citation practices and techniques that we can observe in the early second-century Christian writers are not very different from the flexible treatment of written sources in the NT and in the broader literary culture of the time. That the wording of these citations is often not attested in any of the extant copies of the cited works suggests that authors exercised a certain freedom in amending what they cited. The differences between citations and the texts of the sources cited often seem to be, not simply the products of imprecise memory, but instead deliberate, sometimes artful adaptations. Moreover, the confidence with which authors made these adaptations of widely known sources suggests that they wrote for readers who accepted such freedom as a legitimate convention in the literary culture of the time. That is, readers familiar with the sources being cited would likely have recognized the adaptations. These readers would have objected to these adaptations only if they dissented from the point being made by the author doing the citation. Moreover, Stanley cogently observed that it is over-simplified to imagine that ancient authors only either cited a work open before them or else worked from memory or ‘oral tradition’. There is impressive evidence, particularly the use of combined and conflate citations, to suggest that authors often worked from written compilations of excerpts from one or more sources, these compilations probably arranged in noteform topically. Furthermore, Stanley proposes that the alterations that we see in the cited material may often have been made at the point of making the excerpt. That is, authors likely combed relevant sources looking for excerpts on some topic and/or in search of support for some point that 37.  Christopher D. Stanley, Paul and the Language of Scripture: Citation Technique in the Pauline Epistles and Contemporary Literature, SNTSMS 74 (Cambridge: Cambridge University Press, 1992). See esp. ‘Citation Technique in Greco-Roman Literature’ (267–91), and ‘Citation Technique in Early Judaism’ (292–337).

1. The New Testament in the Second Century

17

they wished to make. And, given that the literary culture of the day fully permitted adaptation of cited material, authors would likely have adapted what they excerpted, at the point of compiling excerpts, much in the way researchers today often combine cited material and their own observations in their notes. But the key question is whether there was a fully commensurate freedom in the copying of these writings. To be sure, the extant evidence for the writings cited, whether Greek classics, OT, or NT writings, indicates sometimes impressive fluidity, especially in early stages of the textual transmission of these writings.38 But, with a few notable exceptions, the fluidity evidenced in extant manuscripts does not really match the greater extent of the variations that we see in the citations of these works in the authors of the early Roman period. This all means that we should probably think of the copying of texts and the citation of them as somewhat distinguishable processes with different sets of conventions. It follows, thus, that it is dubious to take the form of citations as direct evidence of the state of the texts being cited. To come to the point relevant here, I suggest that it is almost certainly dubious to play off and privilege citations over against our early manuscript evidence for the NT writings. Though they are frustratingly fragmentary, our earliest manuscripts come from within decades of the dates of the early second-century Patristic writers (e.g., Justin), and our more substantial manuscripts are roughly contemporary with, or even earlier than, Patristic writers of the third century and later. We must reckon with all the relevant evidence in characterizing the transmission of texts in second-century Christianity, among which, I contend, early manuscript evidence is the most crucial. A similar cautionary note was sounded several decades ago by Bruce Metzger, who advised that in dealing with Patristic citations of the NT that differ from the textual readings in extant manuscripts, ‘the textual critic must consider whether it was the Father or the scribe of an early copy of

38.  Stanley cites Stephanie West, Ptolemaic Papyri of Homer, Papyrologica Coloniensis 3 (Cologne: Westdeutscher, 1967), on the ‘highly fluid’ state of Homeric epics till they were standardized in mid-second century BCE (West, 5–14). Also Qumran studies, e.g., Shemaryahu Talmon, ‘The Textual Study of the Bible: A New Outlook’, in Qumran and the History of the Biblical Text, ed. F. M. Cross and S. Talmon (Cambridge, MA: Harvard University Press, 1975), 321–400; Emanuel Tov, Textual Criticism of the Hebrew Bible (Assen: Van Gorcum; Minneapolis: Fortress, 1992), esp. 155–98, show an interesting diversity in the text of the OT writings in the pre-Mishnaic period.

18

Texts and Artefacts

the New Testament who was more likely to alter the text’.39 Subsequently, Gordon Fee also demonstrated problems in the use of Patristic citations for recovering the NT text of their times and locales. Fee showed that Patristic authors can cite the NT rather freely, especially in sermons and related writings, whereas in commentaries they adhere more to the wording of the manuscripts available to them.40 If Patristic writers so freely adapted the text of NT writings in citations, well after the NT writings had acquired unquestionably scriptural status and textual stability, we are warned about taking citations in the writings of second-century authors as direct evidence of the state of the text in their time. On the other hand, we should certainly not ignore Patristic citations. In a number of cases they appear to preserve variants otherwise attested only lightly in Greek manuscripts or versions. Therefore, in other cases also where the extant evidence does not allow us to verify matters, we may suspect that this might be so. Clearly, there was fluidity, sometimes what may seem considerable, in the way that the text of NT writings was handled in the second century in some instances. All the same, we should not disregard the other indications that among the scribal tendencies of the time there was also, in some other cases, a recognizable concern to copy relatively carefully and faithfully.41 But there is a good deal more to be said about what the citations of NT writings in second-century authors tell us. My comments here are not intended to pronounce with finality, only to give a reason to recognize that previous analyses are not adequate, and to underscore some important questions and issues for further research.42 39.  Bruce M. Metzger, ‘Patristic Evidence and the Textual Criticism of the New Testament’, in New Testament Studies: Philological, Versional, and Patristic (Leiden: Brill, 1980), 167–88 (quotation from p. 183). 40.  Gordon D. Fee, ‘The Text of John in The Jerusalem Bible: A Critique of the Use of Patristic Citations in New Testament Textual Criticism’, in Epp and Fee, Studies in the Theory and Method, 335–43 [originally published in JBL 90 (1971): 163–73]. 41.  Note, e.g., the judgment by J. Neville Birdsall in his study of the text of Luke in P75 and P66, ‘Rational Eclecticism and the Oldest Manuscripts: A Comparative Study of the Bodmer and Chester Beatty Papyri of the Gospel of Luke’, in Studies in New Testament Language and Text. Essays in Honour of George D. Kilpatrick on the Occasion of his Sixty-fifth Birthday, NovTSup 44 (Leiden: Brill, 1976), 39–51. 42.  See also James A. Kelhoffer, Miracle and Mission: The Authentication of Missionaries and Their Message in the Longer Ending of Mark, WUNT 2/112 (Tübingen: Mohr Siebeck, 2000), esp. 123–30, in critique of Koester’s methods in assessing use/influence of written sources in second-century writers.

1. The New Testament in the Second Century

19

Collections It is a well-known feature of second-century Christianity that collections of writings that came to be part of the NT were formed and circulated. We know that at some point the four canonical Gospels came to be thought of as complementary renditions of the gospel story of Jesus, and came to form a closed set enjoying distinctive regard in many Christian circles. We know also that collections of Pauline epistles were circulating, probably from the late first century, and were likewise treated as scripture in at least some circles.43 These phenomena are regularly and rightly noted in histories of the formation of the NT canon. But I propose that these collections constitute a curious and possibly more significant phenomenon than is reflected in the attention usually given to it in scholarly studies. To be sure, these collections contributed to the larger collection that we know as the NT canon. But, well before there was a ‘New Testament’ canon or the debates about what it comprised, what was the motivation for establishing such collections of writings, and what function(s) were these collections intended to serve? It is remarkable how early these collections of writings appear. Several recent studies agree in pushing back the likely origin of a fourfold Gospels collection to the earliest years of the second century. This scholarly agreement is all the more interesting in that these studies pursue different approaches and questions. Theo Heckel argued that the crucial factor was the production of the familiar form of the Gospel of John in Johannine circles, and Heckel places a fourfold Gospel collection sometime around 120 CE.44 In an astonishingly detailed study of the long ending of Mark, James Kelhoffer argues (persuasively to my mind) that these verses were composed sometime in the first half of the second century, ‘with confidence’, he judges, ca. 120–150 CE, and that they presuppose a fourfold Gospel collection that had been circulating and given ‘high respect’ for some time previously.45 In a study proposing identification of a further fragment of Papias’s comments about the Gospels preserved in Eusebius, Charles Hill contends that Papias knew the four canonical Gospels as a collection sometime ca. 125–135 CE.46 In a recent book 43.  See, e.g., Gamble, Books and Readers, 99–101. 44.  Theo K. Heckel, Vom Evangelium des Markus zum viergestaltigen Evangelium, WUNT 120 (Tübingen: Mohr-Siebeck, 1999). See the review by David C. Parker, JTS 52 (2001): 297–301. Cf. the review by Kari Syreeni in Review of Biblical Literature (http://www.bookreviews.org/Reviews/3161471997.html). 45.  Kelhoffer, Miracle and Mission, esp. 175, and 154–56, 158 (n. 4). 46.  Charles E. Hill, ‘What Papias Said About John (and Luke): A “New” Papian Fragment’, JTS 49 (1998): 582–629, esp. 616–17.

20

Texts and Artefacts

Martin Hengel has weighed in strongly in support of an early fourfold Gospel collection as well.47 But perhaps the most programmatic sketch of the case for an early fourfold Gospel collection has been offered by Graham Stanton in his 1996 SNTS Presidential Address.48 Working chronologically backwards from Irenaeus, Stanton concludes that a fourfold Gospel collection was being promoted from sometime shortly after 100 CE, although this closed collection may have required time to win its well-known supremacy. We can also note slightly earlier studies, in particular Bellinzoni’s careful analysis of Justin’s use of Jesus’ sayings.49 He showed that Justin likely used compilations of Jesus’ sayings drawn from the written Gospels, almost certainly the three Synoptic Gospels. But Justin’s reference to the liturgical reading of memoirs of ‘apostles and those who followed them’ (Dial. 103.8) suggests strongly that he knew of at least two Gospels attributed to apostles and at least two attributed to others. The most likely conclusion is that Justin refers to our four canonical Gospels as regularly read in worship.50 As for a collection of Pauline epistles, the evidence points back at least as early. Indeed, David Trobisch has proposed that Paul himself may have compiled the first collection of his own epistles.51 The reference to ‘all’ of Paul’s epistles in 2 Pet 3:16 probably takes us back to sometime ca. 100 CE or earlier, although it is impossible to say what the ‘all’ comprised. For, as Gamble notes, it appears that the second-century Pauline collections were of varying dimensions, comprising ten, thirteen, or fourteen letters.52 Marcion’s exclusive claims for his ten-letter Pauline collection sometime 47.  Martin Hengel, The Four Gospels and the One Gospel of Jesus Christ (London: SCM; Harrisburg, PA: Trinity Press International, 2000). 48.  Stanton, ‘The Fourfold Gospel’. 49.  Arthur J. Bellinzoni, The Sayings of Jesus in the Writings of Justin Martyr, NovTSup 17 (Leiden: Brill, 1967). See also Helmut Koester, Ancient Christian Gospels (London: SCM; Philadelphia: Trinity Press International, 1990), 360–402. 50.  See now Oskar Skarsaune, ‘Justin and His Bible’, in Justin Martyr and His Worlds, ed. Sara Parvis and Paul Foster (Minneapolis: Fortress, 2007), 53–76. 51.  David Trobisch, Die Entstehung der Paulusbriefsammlung: Studien zu den Anfängen christlicher Publizistik, NTOA 10 (Göttingen: Vandenhoeck & Ruprecht, 1989). 52.  Gamble, Books and Readers, 59–63, 100. On the impact of Paul, see esp. Andreas Lindemann, ‘Der Apostel Paulus im 2. Jahrhundert’, in The New Testament in Early Christianity, ed. Jean-Marie Sevrin, BETL 86 (Leuven: Peeters, 1989), 39–67; idem, Paulus im ältesten Christentum. Das Bild des Apostels und die Rezeption der paulinischen Theologie in der frühchristlichen Literatur bis Marcion (Tübingen: Mohr-Siebeck, 1979).

1. The New Testament in the Second Century

21

around 140 CE probably presuppose a widespread circulation of Pauline letter-collections already by that point. By ca. 200 CE, however, there was an ‘apostolikon’ category of Christian scriptures, comprising a Pauline collection plus letters attributed to other apostolic figures (esp. 1–2 John, 1 Peter, James), as seems reflected in the Muratorian Fragment (which I take to be from ca. 200–250 CE).53 In a sense, then, the NT is a collection of prior collections. David Trobisch has proposed vigorously that the NT as we know it was compiled as a single editorial project sometime in the mid-second century.54 His argument is most intriguing in pointing to passages in various NT writings that could be seen as intended to cross-reference to, and accredit, other NT writings. I am not persuaded, however, that a full NT collection such as later came to be preferred was compiled and circulating as early as Trobisch contends. But it may well be that the compilation of early collections of texts, such as a fourfold Gospel and a Pauline letter-collection, did stimulate the composition of other texts and helped to shape their contents, including the embedding of the sorts of intriguing references that Trobisch highlights. I emphasize here that these collections also probably had an effect upon the transmission of the text of the component writings. The most dramatic demonstration is, of course, Tatian’s Diatessaron (ca. 172 CE), a thorough adaptation and expansion of earlier harmonizing texts (such as may have been used by Justin).55 It appears that one of Tatian’s added features was a full use of John. But I am not persuaded that the few bits of material not paralleled in the extant texts of the canonical Gospels are sufficient for the claim that Marcion used any additional, fifth gospel writing. 53.  See, e.g., Joseph Verheyden, ‘The Canon Muratori: A Matter of Dispute’, in The Biblical Canons, ed. J.-M. Auwers and H. J. De Jonge (Leuven: Leuven University Press/Peeters, 2003), 487–566. 54.  David Trobisch, Die Endredaktion des Neuen Testaments. Eine Untersuchung zur Entstehung der christlichen Bibel (Freiburg: Universitätsverlag Freiburg; Göttingen: Vandenhoeck & Ruprecht, 1996); English translation: The First Edition of the New Testament (New York: Oxford University Press, 2000). 55.  See William Petersen, ‘Tatian’s Diatessaron’, in Koester, Ancient Christian Gospels, 403–30. But I am not so persuaded as Petersen that the Diatessaron incorporated written sources other than the four canonical Gospels. Cf., e.g., Hengel, The Four Gospels and the One Gospel of Jesus Christ, 24–26. The oft-cited Dura fragment may not be so clearly a piece of the Diatessaron as scholars have tended to think. Cf. D. C. Parker, D. G. K. Taylor and M. S. Goodacre, ‘The Dura-Europos Gospel Harmony’, in Taylor, ed., Studies in the Early Text of the Gospels and Acts, 192–228; Jan Joosten, ‘The Dura Parchment and the Diatessaron’, VC 57 (2003): 159–73.

22

Texts and Artefacts

The ‘long ending of Mark’, as Kelhoffer has powerfully argued, is another striking textual phenomenon reflecting the fourfold Gospel collection. This early addition to Mark appears to draw upon the four canonical Gospels, and no other gospel writing. It shows, too, that the four canonical Gospels were not only collected but also compared with one another, which explains best why someone thought that Mark’s ending was deficient and needed to be augmented along the lines of the other Gospels.56 In numerous smaller variants as well, we can probably see the effects of a fourfold Gospels collection, especially the many harmonizations of one Gospel to another that I have already mentioned. Clearly, the special recognition given to the four Gospels did not necessarily involve a reluctance to make such ‘improvements’ to their texts. Of course, it is often thought that the Pauline letters exhibit significant evidence of the effects of circulating in/as a collection. If, as it widely thought, 2 Corinthians is a composite writing, might this composition have taken place in some connection with an early Pauline collection? But a collection could dispose some toward shortening texts too. Gamble has argued that a fifteen-chapter version of Romans (and perhaps a fourteen-chapter version as well) was prepared with a view toward wide ecclesiastical circulation, and a Pauline letter collection is the most likely vehicle for this.57 He has also noted evidence suggesting ‘an early, certainly first-century, effort to overcome the problem [of the particularity of Paul’s letter-destinations] by deleting or generalizing the addresses of some of the letters and sometimes by omitting other locally specific matter as well…’58 The larger question, however, is why such collections emerged at all and became so successfully used. Other letter-collections are a known literary phenomenon in Roman-era antiquity, both in Christian (e.g., Ignatius of Antioch) and pagan circles. A Pauline letter-collection is, thus, not without precedent, though it is still a remarkable development, and perhaps a precedent-setting phenomenon for Christians of the early centuries.59 56.  Kelhoffer, Miracle and Mission, 154–55. 57.  Harry Y. Gamble, The Textual History of the Letter to the Romans, SD 42 (Grand Rapids: Eerdmans, 1977). 58.  Gamble, Books and Readers, 60. 59.  David E. Aune, The New Testament in Its Literary Environment (Philadelphia: Westminster, 1987), 170–72, refers to collections of letters of Aristotle, Isocrates, Demosthenes, Apollonius of Tyana, Libanius, Cicero, Pliny the Younger, and the senior Pliny.

1. The New Testament in the Second Century

23

But why did a four-Gospel collection emerge so early and manage to have such success? For it is obvious that there were concerns about a plural and somewhat divergent testimony about Jesus.60 Marcion is the most blatant illustration. But the harmonies, especially Tatian’s Diatessaron, also indicate a certain discomfort with four discrete accounts and in some circles a preference for a more cohesive rendition of Jesus. Why, in particular, did Mark obtain a continuing place in the Gospels collection, when in the eyes of many Christians in the second century Matthew seemed to have superseded it so adequately? Collections of Pauline letters circulating by the late first century might be cited as a precedent and stimulus. But a Pauline letter collection represents one apostolic voice, whereas a collection of Gospels, even the recognizably similar canonical four, embodies a diversity of voices and contents. Trobisch is probably right to see in the fourfold Gospel an early and deliberately ‘ecumenical’ move. Likewise, the inclusion of letters attributed to various apostles in what became the NT apostolikon represents a deliberate effort to express and affirm a certain diversity or breadth in what is treated as authoritative.61 Canon The final phenomenon to consider briefly is the emergence of a NT canon, which likewise appears to be well on its way by the end of the second century.62 We can, thus, look to the second century as the period of the key impetus and ‘proto-canonical’ developments. In the interests of limited space, I shall simply mention some key matters. Above all, we again must reckon with the practice of liturgical reading. As we know, those writings that contended for acceptance in the canonical decision-making that is attested in writings from ca. 180 CE onward 60.  Oscar Cullmann, ‘The Plurality of the Gospels as a Theological Problem in Antiquity’, in Oscar Cullmann, The Early Church: Studies in Early Christian History and Theology, ed. A. J. B. Higgins (London: SCM; Philadelphia: Westminster, 1956), 39–54; Helmut Merkel, Die Pluralität der Evangelien als theologisches und exegetisches Problem in der alten Kirche, Traditio Christiana 3 (Bern: Lang, 1978); idem, Die Widersprüche zwischen den Evangelien. Ihre polemische und apologetische Behandlung in der Alten Kirche bis zu Augustin, WUNT 13 (Tübingen: J. C. B. Mohr, 1971). 61.  Trobisch, Die Endredaktion des Neuen Testaments, 158–60. 62.  See, e.g., the various contributions on the formation of the NT canon in The Canon Debate, ed. Lee Martin McDonald and James A. Sanders (Peabody, MA: Hendrickson, 2002).

24

Texts and Artefacts

were those that had already enjoyed widespread inclusion among the texts read in Christian worship gatherings. Those that ‘made it’ the most quickly were those that had the widest usage from the earliest years, and the remaining questions about the rest were settled largely on the basis of whether they had been accepted for liturgical reading sufficiently widely. The earliest precedent and impetus disposing Christian circles to include writings of their own for such liturgical usage was almost certainly the letters of Paul. They were composed as liturgical texts, to be read out in the gathered assemblies to whom they were originally sent, and were ‘kitted out’ with liturgical expressions to make them fit this setting more readily, especially the well-known letter opening and closing expressions. Moreover, if Colossians is a deutero-Pauline composition, it nevertheless shows that the exchange of Paul’s letters among churches began within the years following his execution (4:16); and, if it is an authentic letter of Paul’s, then the practice was earlier still. As I have indicated already, it appears also that the Gospels that became canonical circulated impressively widely and early. That Mark was used so thoroughly as source and model by the authors of Matthew and Luke shows that Mark circulated influentially in various Christian circles within a short period after its composition. Thereafter, at least to judge from the comparatively greater number of early copies extant, it appears that Matthew outstripped all the others in breadth of usage and frequency of copying. But John too appears to have enjoyed impressive success very early.63 Heracleon’s commentary, written sometime ca. 150–175 CE, suggests that John had for some time enjoyed scriptural significance in at least some circles. (It is noteworthy that we have no such commentary on any Christian writing that did not come to form part of the NT.) The various early public-readers’ aids mentioned earlier as characterizing copies of NT writings already in the early second century reflect their usage as liturgical texts. Again, the closest pre-Christian precedents and analogies for these scribal features are found in Jewish copies of OT scriptural writings that came to be included in the closed canon of Judaism.64 63.  I consider the comparative number of extant early copies of various Christian texts in The Earliest Christian Artifacts, 15–41. On the use of the Gospel of John in particular, see now Charles E. Hill, The Johannine Corpus in the Early Church (Oxford: Oxford University Press, 2004). 64.  See, e.g., Tov, Textual Criticism of the Hebrew Bible, 208–17, on scribal practices that can be traced back to pre-Christian manuscripts.

1. The New Testament in the Second Century

25

All this early interest in the public reading of certain writings as part of the liturgical life of Christian groups suggests that we might need to re-think the view that it was only in the later decades of the second century that a ‘text consciousness’ came to be influential. We have, perhaps, somewhat romantically regarded the earliest Christian circles as so given to oral tradition that their writings took a distant second place in their values. I submit that from the earliest observable years Christianity was a profoundly textual movement.65 To cite an early indication, although Paul was an intrepid itinerant preacher, and characterized himself primarily as such (Rom 15:17–21), even in his own lifetime his critics referred to the effects of his letters (2 Cor 10:9–11). The production of the deutero-Pauline letters and, indeed, the larger production of pseudonymous letters as well, attest that writings were early an influential mode of Christian discourse, persuasion, and promotion of religious ideas. The reference to ‘the books and above all the parchments’ in 2 Tim 4:13 shows how much Paul was associated with texts in the subsequent circles that revered him. John of Patmos conveyed his colourful visions and words in a text, which he clothed in prophetic authority and for which he demanded respectful reading and copying (Rev 21:18–19). The production of multiple written renditions of Jesus in the first century and onward shows also that texts were an early and favoured mode for transmitting traditions about him. Even earlier than the canonical Gospels, the Q sayings-source illustrates this as well.66 The continuing proliferation of ‘gospels’ beyond the four that became canonical was apparently already well under way in the second century, and further shows how given to texts early Christian were in circulating their traditions about Jesus. We have, perhaps, read too much into the oft-quoted words of Papias about preferring the reports of ‘living and surviving’ voices over books.67 For Papias’ profession simply echoes the sort of claims that ancient historians regularly made for their works, claims that they either were eyewitnesses themselves or had learned of the events they narrate from 65.  Larry W. Hurtado, ‘Oral Fixation and New Testament Studies? “Orality”, “Performance” and Reading Texts in Early Christianity’, NTS 60 (2014): 321–40; and Destroyer of the Gods, 105–41. 66.  Although the Q sayings source has come in for a good deal of doubt recently, I continue to entertain it as a viable hypothesis. 67.  As quoted in Eusebius, HE 3.39.4. But see L. C. A. Alexander, ‘The Living Voice: Skepticism Towards the Written Word in Early Christian and in Graeco-Roman Texts’, in The Bible in Three Dimensions, ed. D. J. A. Clines (Sheffield: Sheffield Academic, 1990), 221–47.

26

Texts and Artefacts

witnesses.68 That is, Papias’ words do not really represent a preference for oral tradition over against books, but instead he reflects the literary conventions of his time, in which one sought authority for one’s written reports through claiming that they rested on authentic witnesses. And need I remind us that this Papias who supposedly disdained books is himself reported to have composed a five-volume written exposition of the sayings of Jesus? It is true that Christian writers of the decades prior to ca. 150 CE do not characteristically cite texts explicitly in the way that it is done much more frequently in subsequent times. But, I repeat, is the practice of the post-150 CE period indicative of an emergent ‘text consciousness’, or is it more correct to see an emergent author-consciousness? That is, I suggest that what changes in the post-150 CE period is a greater tendency to see texts as the works of authors, and so to cite them as such, rather than simply appropriating the contents of texts. And I further suggest that a major reason for a greater emphasis on texts as products of particular authors is the swirling controversies of the second century over heresies. This, I propose, led Christians to place greater emphasis on authorship of writings as a way of certifying and/or promoting them. So, for example, whereas the canonical Gospels were composed without the authors identifying themselves, across the second century we see an increasing tendency to attribute and emphasize authorship of writings, including a greater tendency to attribute authorship to writings for which authorship was not an explicit feature of the text (e.g., the canonical Gospels, Hebrews). Conclusion Given the breadth of phenomena and issues involved in the three processes that I have addressed here, it has been necessary to limit the extent of my discussion of any of them. Where I have taken a position on controversial matters, the unavoidable brevity means that I cannot hope to have persuaded anyone firmly holding another viewpoint. But I have aspired here, not only to review the relevant phenomena and issues, but also to underscore the importance of the second century for the writings that came to comprise our NT. I hope also to have helped to dispose scholars of the NT, and scholars of the text of the NT in particular, to make a full harvest of the materials available for researching how NT writings were treated in the second century. Recent studies, and recently available 68.  See, e.g., the discussion of Hellenistic historiography by Aune, The New Testament in Its Literary Environment, 80–83.

1. The New Testament in the Second Century

27

manuscripts and their data as well, provide us with some potentially exciting prospects for further knowledge and insight about this crucial period. More than ever, it is in the interests of any particular question or line of inquiry into the second century that we try to take as much account as we can of the spectrum of questions, issues, and available evidence.

Chapter 2 T he E a r ly N ew T es ta me nt P apy r i : A S urv ey of T h ei r S i gni fi cance

The Manuscripts There are now 134 NT papyri assigned numbers in the Gregory-Aland list, actually comprising 128 manuscripts, which represents a massive increase accrued over the course of the twentieth century.1 Before 1900, only a handful of NT papyri were known, none of them early enough to have any perceived value above the major textual witnesses of the 1.  I have updated these numbers from the original publication. As P64 and P67 are now commonly taken as portions of the same codex, and P33 and P58 likewise parts of another, as are P11 and P14, there are then actually 131 papyrus manuscripts currently identified. T. C. Skeat proposed that P4, P64 and P67 all were from the same codex: ‘The Oldest Manuscript of the Four Gospels’, NTS 43 (1997): 1–34, defending a suggestion made by others earlier, but cf. Peter Head, ‘Is P4, P64 and P67 the Oldest Manuscript of the Four Gospels? A Response to T. C. Skeat’, NTS 51 (2005): 450–57; and Scott D. Charlesworth, ‘T. C. Skeat, P64 + 67 and P4, and the Problem of Fibre Orientation in Codicological Reconstruction’, NTS 53 (2007): 582–604. More recently, see Tommy Wasserman, ‘A Comparative Textual Analysis of P4 and P64+67’, TC 15 (2010): 1–26. The most up to date and reliable list of NT papyri is provided online by the Münster Institut für textkritische Textforschung, based on the Kurtzgefasste Liste maintained there: http://intf.uni-muenster.de/vmr/ NTVMR/ListeHandschriften.php. I note, however, that, at the point of accessing the Münster online list in updating this essay (27 October 2016), it goes only to P131, although Gregory-Aland numbers have been assigned to 134 papyri. These include the ‘Willoughby papyrus’, P134, a few verses of Mark on one side of a papyrus fragment, dated third–fourth century. As well, P132 (a fragment of Ephesians) and P133 (it is unclear whether it provides portions of 1 Timothy or 2 Timothy) are to be published in the Oxyrhynchus Papyri series, and at this point I have no information on the date to which they are assigned. Another surprisingly up to date list appears in the Wikipedia entry, List of New Testament Papyri: http://en.wikipedia.org/wiki/List_of_ New_Testament_papyri. The most recent Nestle-Aland Novum Testamentum Graece,

2. The Early New Testament Papyri

29

fourth century. Indeed, NT papyri in significant numbers and of special antiquity appeared on the scene only well into the twentieth century. So, e.g., in 1912, when Henry Sanders published the photographic facsimile of Codex Washingtonianus (the four Gospels), which he dated to the late fourth or early fifth century, this manuscript was then one of the very earliest witnesses to the text of any of the four Gospels, surpassed in date only by Codex Vaticanus and Codex Sinaiticus.2 Moreover, the number of NT papyri available has continued to grow. For example, in a survey of NT papyri published in 1995, Eldon Epp reported the total number of NT papyri as 96, and in a later analysis published in 2007 noted 115 in the official list, the present total of 134 (as of October 2016) thus comprising a 39% increase in twenty-one years and a 16% increase over the total in Epp’s 2007 essay.3 These copies of NT writings form part of a larger body of early Christian copies of literary texts, which include copies of OT writings and various Christian texts such as writings now regarded as Christian apocrypha (e.g., Gospel of Thomas), other Christian religious writings and treatises (e.g., Shepherd of Hermas, Ireneaus, Melito, and a number of unidentified texts), liturgical texts, homilies, and also exorcistic and magical texts. I focus here on the

28th ed. (2012), included 127 papyri in its appended list of witnesses, pp. 792–819. In addition, though not papyrus manuscripts, Gregory–Aland 0189 (a single vellum leaf containing Acts 5:3–21) dated ca. 200 CE, and 0220 (also vellum, containing Rom. 4:23–5:3, 8–13) dated third century are included in the present discussion. 2.  Henry A. Sanders, Facsimile of the Washington Manuscript of the Four Gospels in the Freer Collection (Ann Arbor: University of Michigan Press, 1912); idem, New Testament Manuscripts in the Freer Collection, Part 1, the Washington Manuscript of the Four Gospels (New York: Macmillan, 1912). But cf. the recent critical assessment of the basis for the conventional dating of Codex W: Ulrich Schmid, ‘Reassessing the Palaeography and Codicology of the Freer Gospel Manuscript’, in The Freer Biblical Manuscripts: Fresh Studies of an American Treasure Trove, ed. Larry W. Hurtado (Atlanta: SBL, 2006), 227–49. 3.  Eldon Jay Epp, ‘The Papyrus Manuscripts of the New Testament’, in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael A. Holmes (Grand Rapids: Eerdmans, 1995), 3–21; idem, ‘Are Early New Testament Manuscripts Truly Abundant?’, in Israel’s God and Rebecca’s Children: Christology and Community in Early Judaism and Christianity, ed. David B. Capes et al. (Waco: Baylor University Press, 2007), 77–117. See now his more recent discussion 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), 1–39, in which he counts 127 NT papyri.

30

Texts and Artefacts

earliest NT manuscripts, drawing also upon features of this larger body of early Christian manuscripts, and the studies of the still larger body of manuscripts of the period (Jewish and pagan).4 The primary value of the NT papyri is, of course, not their writing material but their age. Actually, however, a number of these NT papyri are dated to the same centuries from which our well-known and more fully preserved principal witnesses come – fourth to eighth century CE – and so, at least for text-critical purposes, these papyri have not been particularly crucial.5 But there are also a goodly number of papyri (and a few parchment manuscripts as well) that are dated considerably earlier, comprising our very earliest witnesses to NT writings, and so these have a unique historical significance.6 In this discussion, therefore, I focus on the fifty NT papyri and two parchment manuscripts (0189, 0220) palaeographically dated to the second or third century CE, giving some key information about them, and highlighting the principal historical issues on which they uniquely shed light.7 4.  I provided a list of all identifiably Christian copies of all literary texts (including OT, NT and extra-canonical ones) in L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), Appendix 1, 209–29, an updated version of which appears on my personal blog-site: http://larryhurtado.files.wordpress.com/2010/07/christian-lit-texts–2nd–3rd-centuries.pdf. The standard printed (and now increasingly dated) catalogues are Joseph van Haelst, Catalogue des papyrus littéraires juif et chrétiens (Paris: Publications de la Sorbonne, 1976); Kurt Aland, Repertorium der griechischen christlichen Papyri, I: Biblische Papyri, Altes Testament, Neues Testament, Varia, Apokryphen, Berlin: de Gruyter, 1976); idem, Repertorium der griechischen christliche Papyri, II: Kirchenväter-Papyri, Teil 1: Beschreibungen (Berlin: de Gruyter, 1995). Since 1997, Cornelia Römer has produced an annual review of publications on ‘Christian Papyri’ in Archiv für Papyrusforschung, taking up the work of the late Kurt Treu, who produced these annual reviews from 1969 to 1991. 5.  Some are comparatively quite late, e.g., P41 (Acts, eighth century CE), P42 (Luke, seventh/eighth century), P61 (Paulines, ca. 700 CE), P73 (Matthew, seventh century), P74 (Acts, seventh century). 6.  The early parchment manuscripts in question (with dates as given in the Münster online list) are these: 0162 (John 2:11–22, third/fourth century), 0171 (Matt 10:17–23, 25–32; Luke 22:44–56, 61–64, ca. 300 CE), 0189 (Acts 5:3–21, second/ third century), and 0220 (Rom 4:23–5:3, 8–13, third century). 7.  As several papyri given Gregory-Aland numbers are yet to be published (P132–134), I cannot consider them here. Also, taking P64 and P67 as portions of one codex, and omitting from consideration here another 13 NT papyri (as well as 0162 and 0171), which are dated ‘third/fourth’ century CE, we have 52 NT papyri dated to the second or third century. These manuscripts will suffice to illustrate the matters discussed. For earlier discussion of NT papyri, see, e.g., Kurt Aland, ‘The

2. The Early New Testament Papyri

31

The value of these manuscripts is also inverse to the amount of text that they typically preserve. Overwhelmingly, these early items, including most of the 52 earliest considered here, are small remnants of the manuscripts from which they derive. Indeed, in a disappointing number of cases we have only fragments of individual leaves of the codex in question, and in a few other instances we have portions of a handful of leaves of a codex. Of the 52 manuscripts that we consider in this discussion, only five provide us with much more than such small portions of text. Nevertheless, these manuscripts comprise our earliest copies of NT writings and so all of them are invaluable as witnesses to the history of the text of these writings, and for a number of other historical questions as well that I will highlight here. Before we consider their importance with reference to these questions, however, a few further introductory comments are in order. Principal New Testament Papyri Among NT papyri, those included in the Chester Beatty collection hold a major importance. Indeed, the publication of the Chester Beatty biblical papyri between 1933 and 1937, eleven codices (originally thought to be twelve) comprising very early copies of a number of OT, NT and extracanonical texts, decisively presented scholars, especially in NT and LXX studies, with a veritable goldmine.8 Most of these codices are dated to Significance of the Papyri for Progress in New Testament Research’, in The Bible in Modern Scholarship, ed. J. Philip Hyatt (Nashville: Abingdon, 1965), 334–37; Eldon Jay Epp, ‘The New Testament Papyrus Manuscripts in Historical Perspective’, in To Touch the Text: Biblical and Related Studies in Honor of Joseph A. Fitzmyer, S.J., ed. Maurya P. Horgan and Paul J. Kobelski (New York: Crossroad, 1989), 261–88; and idem, ‘The Papyrus Manuscripts of the New Testament’ (2013); 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, STAR 8 (Assen: Royal van Gorcum, 2003), 19–38; idem, ‘Das Zeugnis der frühen Papyri für den Text der Evangelien: Diskutiert am Matthäusevangelium’, in The Four Gospels 1992, ed. F. Van Segbroeck, C. M. Tuckett, G. Van Belle, and J. Verheyden, BETL 100 (Leuven: Leuven University Press/Peeters, 1992), 325–35. As well, in a number of articles Kurt Aland reviewed early NT papyri: ‘Neue neutestamentliche Papyri’, NTS 3 (1956–57): 261–86; ‘Neue neutestamentliche Papyri II’, NTS 9 (1962–63): 303–16; NTS 10 (1963–64): 62–79; NTS 11 (1964–65): 1–21; NTS 12 (1965–66): 193–210; ‘Neue neutestamentliche Papyri III’, NTS 20 (1973–74): 357–81; NTS 22 (1975–76): 375–96. 8.  Frederic G. Kenyon, The Chester Beatty Biblical Papyri, Descriptions and Texts of Twelve Manuscripts of Papyrus of the Greek Bible (London: Emery Walker,

32

Texts and Artefacts

the third century and at least one OT codex (Chester Beatty Papyrus VI, portions of Numbers and Deuteronomy) dated to the mid/late second century CE. Moreover, although there had been fascinating fragments of early copies of NT texts previously unearthed from Oxyrhynchus (e.g., P1 [P.Oxy. 2], a fragment of an early third-century codex of Matthew), the Chester Beatty biblical papyri provided much more substantial portions of remarkably early copies of several biblical texts. For our purposes, three of the Chester Beatty codices are particularly important. P45 (Chester Beatty I) comprises 30 of the original 112 leaves of a codex, preserving portions of all four Gospels (in the ‘Western’ order: Matthew, John, Luke, Mark) and Acts, and is dated to the early/ mid-third century CE.9 In P46 (Chester Beattty II, dated ca. 200 CE), some 86 leaves of an original 102, a codex of Pauline epistles, we have substantial portions of Romans, Hebrews (!), 1 Corinthians, 2 Corinthians, Ephesians, Galatians, Philippians, Colossians, and 1 Thessalonians. P47 (Chester Beatty III, third century) comprises portions of some ten leaves of a codex containing Revelation. The other somewhat substantially preserved, earliest NT manuscripts are part of the Bodmer papyri collection.10 Two papyri in particular are important for this discussion. P66 (Bodmer II) is dated ca. 200 CE and 1933–37). The earlier view that there were twelve codices was later revised, and it is now accepted that we have remains of eleven. 9.  See esp. T. C. Skeat, ‘A Codicological Analysis of the Chester Beatty Papyrus Codex of the Gospels and Acts (P45)’, Hermathena 155 (1993): 27–43, republished in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup, 113 (Leiden: Brill, 2004), 141–57, and more recently note the contributions in The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels – The Contribution of the Chester Beatty Gospel Codex P45, ed. Charles Horton (London: T&T Clark International, 2004). Among Greek manuscripts, the ‘Western’ order of the Gospels is also found in Codex Bezae (Codex D) and Codex Washingtonianus (W). Note also my essay, ‘P45 as an Early Christian Artefact’, which is republished as Chapter 12 of present volume. 10.  Twenty-two manuscripts discovered in 1952 near Dishna, Egypt, they were acquired by the Swiss, Martin Bodmer. Publication of them began in 1954. See now James M. Robinson, The Story of the Bodmer Papyri: From the First Monastery’s Library in Upper Egypt to Geneva and Dublin (Eugene, OR: Cascade, 2011). Developing further a suggestion by C. H. Roberts (‘Books in the Graeco-Roman World and in the New Testament’, in Cambridge History of the Bible. Vol. 1, From the Beginnings to Jerome, ed. P. R. Ackroyd and C. F. Evans [Cambridge: Cambridge University Press, 1970], 48–66 [56]), Robinson contends that the Bodmer and Chester Beatty papyri originated from the same Pachomian monastery library. Thus far, however, this view is still under debate.

2. The Early New Testament Papyri

33

preserves much of the Gospel of John.11 Bodmer Papyrus XIV–XV (P75) which is typically dated to ca. 200 CE as well, preserves substantial portions of Luke and John (102 of an estimated 144 pages survive, along with further fragments, of which eleven were identified subsequently to the publication of the editio princeps).12 The Other Earliest Witnesses As mentioned already, the remaining earliest NT papyri and the early two parchment manuscripts considered here are all fragmentary, often portions from only one page of a codex. But a few of these are dated (palaeographically) even a bit earlier than the Chester Beatty and the Bodmer papyri noted here. Though preserving only small portions of text, therefore, they are of great importance. Among these, the Rylands fragment of John, P52, will be most widely known, which has often been dated ca. 150 CE, but now may have to be placed a bit later toward the end of the second century.13 A few other papyri as well are dated by their

11.  P66 preserves nearly all of John 1:1–14:26, except for pp. 35–38 containing John 6:11–35, and less well 14:29–21:9. Victor Martin and J. W. B. Barns, eds., Papyrus Bodmer II, Supplement: Nouvelle edition augmentee et corrigee (ColognyGenève: Bibliotheca Bodmeriana, 1962). Subsequently, an additional bi-folium containing part of John 19 was published as P.Köln inv. 4274/4298: Michael Gronewald, ‘Johannesevangelium Kap. 19.8–11; 13–15; 18–20; 23–24’, in Kölner Papyri (P.Köln) 5, ed. M. Gronewald, Klaus Maresch, and Wolfgang Schäfer (Opladen: Westdeutscher, 1985), 73–76 (+ plate 7). Apart from this bi-folium and one additional folio (pp. 139–40, John 19:25–28, 31–32) in the Chester Beatty Library, P66 is kept in the Bibliotheca Bodmeriana in Cologny-Genève. I omit here P72 (Bodmer Papyrus VII/ VIII), the earliest copy of 1–2 Peter and Jude, as it is dated somewhat later (‘third/ fourth” century). It now is held in the Vatican Library. 12.  Victor Martin and Rudolf Kasser, Papyrus Bodmer XIV–XV (ColognyGenève: Bibliotheca Bodmeriana, 1961); Kurt Aland, ‘Neue neutestamentliche Papyri III’, NTS 22 (1975–76): 375–96. See James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill; Atlanta: SBL, 2008), 615 (n. 1) for references to various proposals for the date of P75, which vary from ca. 175 to ca. 275 CE. After being put on the market for sale in 2006, P75 was purchased and donated to the Vatican Library where it is now housed. For a news-story on the acquisition by the Vatican Library, see http://dsc.discovery.com/ news/2007/03/05/gospel_arc.html. 13.  See now Brent Nongbri, ‘The Use and Abuse of P52: Papyrological Pitfalls in the Dating of the Fourth Gospel’, HTR 98 (2005): 23–48, who argues that the time-frame for P52 ‘must include dates in the later second and early third centuries’ (46). P52 was dated by comparison with P.Egerton 2 (a fragment of an unknown

34

Texts and Artefacts

editors to the late second century: P90 (John 18:36–40; 19:1–7), P104 (Matt 21:34–37, 43, 45), and P98 (Rev 1:13–20). Several more are dated just a bit later than these, very early in the early third century: P32 (Titus 1:11–15; 2:3–8), P64/P67 (portions of the same codex, Matt 3:9, 15; 5:20–22, 25–28; 26:7–8, 10, 14–15, 22–23, 31–33); and P77 (Matt 23:30–39), P103 (Matt 13:55–56; 14:3–5), and 0189 (Acts 5:3–21) are dated late second and/or early third century. The remaining 37 NT papyri considered here, along with 0220 (a parchment manuscript), are all dated to sometime in the third century.14 Amount of New Testament Text As noted already, the fragmentary nature of most of the earliest NT manuscripts means that collectively they preserve only limited amounts of the text of NT writings. The amounts vary considerably, however. For example, the 18 second- and third-century copies of John together preserve 823 of the 867 verses, about 95% of John.15 By contrast, the ten copies of Matthew from the same period comprise 142 of the 1070 verses, or about 13%. The one copy of Mark (P45) preserves 157 of 666 verses, about 23.5%, and the single copy of Philemon (P87) preserves five of the 25 verses, 20% of that text.16

gospel text), the date of which has also now been moved later. See also L. W. Hurtado, ‘P52 (P.Ryl. Gk. 457) and the Nomina Sacra: Method and Probability’, TynBul 54 (2003): 1–14. 14.  The remaining 37 papyri not already mentioned are these: P4 (Luke), P5 (John), P9 (1 John), P12 (Heb), P15 (1 Cor), P20 (James), P22 (John), P23 (James), P27 (Rom), P28 (John), P29 (Acts), P30 (1–2 Thess), P39 (John), P40 (Rom), P48 (Acts), P49 (Eph), P53 (Matt/Acts), P65 (1 Thess), P69 (Luke), P70 (Matt), P80 (John), P87 (Philemon), P91 (Acts), P95 (John), P101 (Matt), P106 (John), P107 (John), P108 (John), P109 (John), P111 (Luke), P113 (Rom), P114 (Heb), P118 (Rom), P119 (John), P121 (John), P129 (1 Cor), and P131 (Rom). 15.  The verses of John in these 18 papyri are 1:1–51; 2:1–25; 3:1–36; 4:1–54; 5:1–47; 6:1–71; 7:1–52; 8:12–59; 9:1–41; 10:1–42; 11:1–57; 12:1–50; 13:1–38; 14:1–31; 15:1–27; 16:1–4, 6–7, 10–33; 17:1–26; 18:1–40; 19:1–42; 20:1–20, 22–31; 21:1–9, 18–20, 23–25. The total of 867 verses that comprise John involves omitting 8:1–11, a pericope that does not appear in any of the earliest manuscripts. 16.  The verses extant collectively in the ten manuscripts of Matthew are 1:1–9, 12–13, 14–20, 23; 2:13–16, 22–3:1; 3:9–15; 3:16–4:3; 5:20–22, 25–28; 12:24–26, 32–33; 13:55–56; 14:3–5; 20:24–32; 21:13–19, 34–37, 43–45; 23:30–39; 24:3–6, 12–15; 25:41–26:40. I omit from my count of Markan material the Willoughby papyrus (P134), as it has not been published yet.

2. The Early New Testament Papyri

35

Nevertheless, obviously we can only be grateful that we have these early remnants, however limited they are. Collectively, the 52 manuscripts dated to the second or third centuries give us copies of 20 of the 27 NT writings. Moreover, arguably, the very random nature of the portions of text that they preserve actually enhances their value as witnesses to the NT writings. In effect, they provide us random samples of the text of the writings in question, on the basis of which we can make wider (but cautious) inferences about the nature of the text as a whole in the respective manuscripts from which the fragments derive.17 Provenance It is also well known that the earliest NT manuscripts all were found in Egypt, and so it is appropriate to consider how representative they may be of the wider circulation of NT writings in the period of these manuscripts. There are, however, several reasons for thinking that these early manuscripts are likely reflective of the status and transmission of NT writings more widely in the period from which they derive.18 First, these earliest manuscripts reflect a spectrum of transmission practices and policies, ranging from a rather strict/careful reproduction to a somewhat freer handling of the text, and with varying degrees of copyist skill as well. I submit that this variety of copying practices and textual tendencies in this body of manuscripts from Egypt works against any ‘local text’ theory, which would require a more homogenous body of manuscripts in a given geographical locality. Second, there is what Eldon Epp has referred to as ‘a brisk “intellectual commerce” and dynamic interchanges of people, literature, books, and letters between Egypt and the vast Mediterranean region’.19 17.  This is essentially also the view advocated by Kurt Aland and Barbara Aland, The Text of the New Testament, trans. E. F. Rhodes (Grand Rapids: Eerdmans, 1987), 58. See also Barbara Aland, ‘Kriterien zur Beurteilung kleinerer Papyrusfragmente des Neuen Testaments’, in New Testament Textual Criticism and Exegesis: Festschrift J. Delobel, ed. A. Denaux, BETL 161 (Leuven: Peeters, 2002), 1–13. 18.  I echo here the position taken by Eldon J. Epp, ‘The Significance of the Papyri for Determining the Nature of the New Testament Text in the Second Century: A Dynamic View of Textual Transmission’, in Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission, ed. William L. Petersen (Notre Dame: University of Notre Dame Press, 1989), 71–103, citing 90. 19.  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. Birgar A. Pearson (Minneapolis: Fortress, 1991), 35–56, citing 55.

36

Texts and Artefacts

That is, we have evidence of an impressive frequency of contacts between Egypt and other parts of the Mediterranean basin in the Roman era. For example, Epp has shown how commonly and frequently letters were sent and received across considerable distances and with impressive speed of delivery.20 Additionally, as I noted in a previous publication, the diversity of Christian literary texts found in Oxyrhynchus further confirms a vigorous ‘networking’ by Christians in particular trans-locally.21 For example, we have portions of three copies of Shepherd of Hermas (composed in Rome), a copy of Irenaeus’s Against Heresies (composed in Gaul), all these dated to the late second and/or early third century, and works of Melito dated third/fourth century.22 So, if we find this sort of evidence in a provincial centre such as Oxyrhynchus (ca. 200 km south of Cairo), it is fairly certain that in major cities such as Alexandria the exchange and circulation of texts would have been even greater. To use a medical analogy, if my blood circulatory system is healthy, a physician can take a blood sample from any part of my body, even my toe, and be confident that the sample will be reliably indicative. In sum, it is safe to agree with Epp’s judgment that the Egyptian papyri likely ‘represent an extensive if not the full textual spectrum of earliest Christianity’.23 Text-Critical Significance I turn now to survey briefly the significance of these earliest NT manuscripts for NT textual criticism. In simplest terms, their major contribution is that they take us back a hundred or more years earlier than the fourthcentury evidence on which all NT textual criticism had rested prior to their availability. As already noted, all of the 52 manuscripts that form the 20.  Ibid. 21.  Hurtado, The Earliest Christian Artifacts, 26–27. 22.  We have remnants of Melito’s Pascal Homily (P.Bodmer 13), and perhaps portions of another work, On Prophecy (P.Oxy. 5). P.Bodmer 13 was formerly thought to be a portion of some work by Melito as well, but see now Thomas Scott Caulley, ‘A Fragment of an Early Christian Hymn (Papyrus Bodmer 12): Some Observations’, ZAC 13 (2009): 403–14, who argues against Melito’s authorship. 23.  Epp, ‘The Papyrus Manuscripts of the New Testament’, 9. Note also that Guglielmo Cavallo and Herwig Maehler, eds., Hellenistic Bookhands (Berlin: de Gruyter, 2008), 16–17, show that the kind of scripts used in copies of classical literary texts ‘developed along very similar lines’ in Egypt and Italy, suggesting a ‘koine’ of Greek literary scripts in the Mediterranean world. This is consistent with (and provides a larger context for) the indications of a trans-local sharing of Christian texts and copying conventions.

2. The Early New Testament Papyri

37

focus here are dated to the third century or earlier, at least eight of them dated to ca. 200 or soon thereafter, and as many as four (P52, P90, P98, P104) to the (late) second century. Even if we accept Roger Bagnall’s recent argument for moving the dates of these second-century witnesses somewhat later (and in my view he gives no compelling reason for doing so), we would still have a body of evidence that gives us a direct view of the transmission of the NT writings in the early third century, and that perhaps even allows us to peer farther back into the second century.24 Some of the effects of this body of evidence from these early manuscripts have been evident for some time now. I illustrate this first with reference to key individual papyri. For example, as shown several decades ago, the striking agreement of P75 with the text of Codex Vaticanus in Luke and John refuted earlier proposals that Vaticanus was the result of a third-century or fourth-century recension of an earlier and ‘rougher’ kind of NT text.25 It appears instead that the so-called Neutral/ Alexandrian text-type represents and derives from a rather careful and competent transmission practice that goes back at least into the late second century and perhaps earlier still.26 This is of potentially profound significance for any theory and history of the earliest text of NT writings (a matter to which I return later). To cite another example, although Lietzmann flatly stated soon after the publication of the Chester Beatty papyri that P45 and P46 had no great significance for knowledge of the transmission of the NT text, it is now clear that he was simply wrong.27 In his magisterial study, 24.  Roger S. Bagnall, Early Christian Books in Egypt (Princeton: Princeton University Press, 2009), esp. Chapter 1. He makes some valid points, but unfortunately appears to take the publications of Carsten Thiede as representative of the approach to dating NT papyri followed by NT scholars more broadly. Cf. my review in Review of Biblical Literature: http://www.bookreviews.org/pdf/7289_7933.pdf. Bagnall’s reasons for questioning the second-century dating of Christian manuscripts seem rather simplistic to me. 25.  See esp. Gordon D. Fee, ‘P75, P66, and Origen: The Myth of Early Textual Recension in Alexandria’, in New Dimensions in New Testament Study, ed. Richard N. Longenecker and Merrill C. Tenney (Grand Rapids: Eerdmans, 1974), 19–45, republished in Eldon J. Epp and Gordon D. Fee, Studies in the Theory and Method of New Testament Textual Criticism, SD 45 (Grand Rapids: Eerdmans, 1993), 247–73. 26.  Royse’s analysis of P75 confirms these judgments (Scribal Habits, 615–704, esp. 615–18). 27.  Hans Lietzmann, ‘Zur Würdigung des Chester-Beatty Papyrus der Paulusbriefe’, SPAW.PH 25 (1934): 775, republished in his collected essays, Kleine Schriften. Vol. 2, Studien zum Neuen Testament, ed. Kurt Aland, TU 68 (Berlin:

38

Texts and Artefacts

The Text of the Epistles, Günther Zuntz showed the crucial importance of P46 as a basis for understanding the early transmission of the Pauline epistles.28 Likewise, although soon after its publication P45 was enlisted as a supporting member of the so-called Caesarean Text in Mark, subsequent analysis has disproved this. Instead, both in its textual character and physical/visual qualities P45 continues to offer fascinating evidence that requires adjustment of previous views about the early transmission of the Gospels and Acts.29 Studies by Colwell and Royse show that P45 has an unusually large number of ‘significant singular’ readings that likely represent a particular effort to produce a readable and edifying text, ‘improving’ it by many stylistic changes, harmonizations, simplifications, and even pruning.30 This likely explains why P45 does not agree closely with any of the key witnesses to known, major text types. It also shows the kind of editorial freedom exercised by some copyists and readers, which contrasts with the copying stance exhibited particularly in P75. In short, P45 and P75 show that in the earliest period from which there is extant evidence there was a certain variety in copying practice and aims, including both a more strict and a comparatively freer practice (and perhaps a certain spectrum in the freedom exercised). It is interesting to me that the great palaeographer, Eric Turner, identified two broad tendencies in ancient papyri of classical literary texts, one exhibiting greater freedom in adding lines or leaving out lines and with ‘substantial variant phrases or formulas’ (which Turner associates with a Platonic attitude toward books), and the other reflecting a greater respect for the wording of the text and exhibiting a lower ‘coefficient of error’ (which Akademie-Verlag, 1958), 171; idem, ‘Die Chester-Beatty-Papyri des Neuen Testament’, Antike 11 (1935): 147, = Kleine Schriften, 2:168, as cited by Epp, ‘The Papyrus Manuscripts of the New Testament’, 12. 28.  Günther Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum, Schweich Lectures 1946 (London: Oxford University Press, 1953; reprint, Eugene, OR: Wipf & Stock, 2007). 29.  See now L. W. Hurtado, ‘P45 and the Textual History of the Gospel of Mark’, in Horton, ed., The Earliest Gospels, 132–48. As I showed in an earlier publication, P45 and Codex W have a significant level of agreement in Mark, but neither of them has any such significant agreement with the putative key Caesarean witnesses (Θ and 565): Larry W. Hurtado, Text-Critical Methodology and the Pre-Caesarean Text: Codex W in the Gospel of Mark, SD 43 (Grand Rapids: Eerdmans, 1981). 30.  Royse, Scribal Habits, 103–97, esp. 123 and the summary on 197; E. C. Colwell, ‘Method in Evaluating Scribal Habits: A Study of P45, P66, P75’, in Colwell, Studies in Methodology in Textual Criticism of the New Testament, NTTS 9 (Grand Rapids: Eerdmans, 1969), 106–24, esp. 117.

2. The Early New Testament Papyri

39

Turner links with Aristotle).31 It seems that the earliest NT manuscripts show a somewhat comparable variety of transmission practice. The Bodmer papyrus of John, P66, has also had a significant impact. Though initially judged simply a ‘mixed’ text, i.e., not a ‘pure’ witness to any of the major text-types, P66 is now typically linked with the P75-Codex B type of text (albeit, a somewhat looser member of this type, with a number of readings supported also by ‘Western’ and ‘Byzantine’ witnesses).32 As Royse has stated, however, ‘The most striking feature of P66 is the quantity of corrections’, some 465 corrections in the extant 75 leaves.33 This unusually large body of corrections has received a good deal of scholarly attention, most recently and extensively by Royse.34 They reveal much about the copyist of this manuscript, including his many initial failures in copying accurately and his subsequent efforts to make things right. In P66, as perhaps in no other early manuscript, we have a fascinating glimpse into one copyist’s determined efforts to produce an accurate copy of his exemplar, despite his evident difficulties in doing so, and also additional evidence of early variant readings to be considered in establishing the text of John. In addition to the significance of particular key papyri, collectively the 52 early manuscripts considered here comprise a valuable body of data for NT textual criticism.35 Of course, as noted already, their early date makes them all especially important in assessing variants, and also for seeing how copyists did their work. Indeed, recent studies of these manuscripts have required modifications of some traditional principles of textual criticism. For example, James Royse’s massive analysis of copyists’ habits in all the earliest substantially preserved NT papyri shows persuasively that copyists in fact more often produced shorter, not longer, readings, and so the traditional principle of preferring the shorter reading does not carry the force it once did. Likewise, Royse has shown that

31.  Eric G. Turner, Greek Papyri: An Introduction, 2nd ed. (Oxford: Oxford University Press, 1980), 106–9. 32.  Fee, ‘P75, P66, and Origen’, 30–31; idem, Papyrus Bodmer II (P66): Its Textual Relationships and Scribal Characteristics, SD 34 (Salt Lake City: University of Utah Press, 1968), 35. 33.  Royse, Scribal Habits, 409. 34.  See Royse’s detailed classification and discussion of the corrections in Scribal Habits, 409–70. Among earlier studies, Fee, Papyrus Bodmer II (P66), is particularly important. 35.  Epp, ‘The Papyrus Manuscripts of the New Testament’, 13–18, sets out a number of matters for which the early manuscripts are crucial.

40

Texts and Artefacts

harmonization to the immediate context was common, a datum that has obvious implications for assessing variants on the basis of similarity to the wording/style of the text.36 But these manuscripts are also crucial for the larger task (not yet adequately addressed) of constructing a theory and history of the earliest stages of textual transmission of the NT writings. Eldon Epp has complained about what he regards as a surprising under-utilization of the early papyri and also has attempted to develop a picture of earliest transmission of the NT from these manuscripts.37 Of course, beginning with the twenty-sixth edition of the Nestle-Aland NT all published papyri have been cited in the textual apparatus. Also, beginning in 1986, the appearance of successive volumes of Das Neue Testament auf Papyrus is another indication of scholarly interest, as is the IGNTP volume on the papyri of John.38 Nevertheless, Epp has a valid point. So I turn to a brief consideration of Epp’s effort to characterize the earliest period of NT textual transmission on the basis of the early papyri. Essentially, Epp attempts to use early NT papyri to construct a diachronic map of the textual transmission of the NT by assessing how particular papyri line up with the major witnesses of later centuries that have been the basis of the well-known text-types.39 He notes that the strong P75-Codex B connection (with P66 a somewhat weaker member of this ‘textual cluster’) shows a textual ‘trajectory’ of this kind of text going back to ca. 200 (the common dating of P75), and Epp accepts the arguments for tracing this trajectory earlier still, well back into the second century at least. He also posits a looser but real connection of certain other papyri (P29, P48, P38, and also 0171) to the kind of text later found in Codex Bezae (D), at least in Acts. Finally, noting that P45 does not seem to fit readily with either of these kinds of texts, and also noting the similarities of P45 and Codex W in Mark, he proposes a third trajectory in which these are key witnesses. On the basis of this analysis, Epp 36.  Royse, Scribal Habits, e.g., 704–36 (a whole chapter on ‘The Shorter Reading’ criterion), and his concluding remarks, 737–42. Peter Head reached similar results: ‘Observations on Early Papyri of the Synoptic Gospels, Especially on the “Scribal Habits” ’, Biblica 71 (1990): 240–47. 37.  See esp. Epp, ‘The New Testament Manuscripts in Historical Perspective’, 261–88; idem, ‘The Significance of the Papyri’, 71–103. 38.  W. Grunewald et al., eds., Das Neue Testament auf Papyrus (Berlin: de Gruyter, 1986–); W. J. Elliott and David C. Parker, The Gospel According to St. John. Part 1, The Papyri (Leiden: Brill, 1995). 39.  Epp, ‘The Significance of the Papyri’, 100, gives fuller lists of NT papyri for each of his proposed ‘textual clusters’. I cite here the early, major papyri for each one.

2. The Early New Testament Papyri

41

concludes that ‘the claim that at least three distinct “text-types” existed in the dynamic Christianity of the second century can be made with considerable confidence’.40 This is not the occasion for a full assessment of Epp’s proposals, but I will allow myself one critical observation. It is valid to consider whether these early witnesses reflect the text-types associated with later key manuscripts, and so whether these text-types can be traced back into the very period of the earliest papyri.41 But I think that we should also try to analyze the early papyri among themselves and in comparison with one another. Indeed, rather than (or at least in addition to) characterizing the early papyri on the basis of their relationships to later witnesses, I propose that it would be still more heuristically useful simply to characterize the respective ‘textual complexions’ of the earliest manuscripts more inductively, in terms of the patterns of readings that each manuscript supports.42 This latter kind of analysis might give us a better basis for judging what kinds of tendencies and attitudes shaped the textual handling of the NT writings in the earliest period.43 From this, in turn, we might be able to develop a theory and history of the very earliest textual transmission of these writings.44 40.  Ibid., 103. 41.  Some take the term ‘text-type’ to imply a ‘local text’ or a ‘recension’, and so reject it outright. I cannot engage this matter adequately here. By ‘text-types’, I simply mean distinguishable transmission-tendencies of the sort I describe briefly above. 42.  I also have some reservations about Barbara Aland’s characterization of early papyri on the basis of how well their readings accord with the ‘Ausgangtext’ (i.e., the Nestle-Aland text). This produces some interesting observations, but, again, seems to me to import an external standard into the assessment; cf. B. Aland, ‘Der textkritische und textgeschichtliche Nutzen früher Papyri’; idem, ‘Das Zeugnis der frühen Papyri für den Text der Evangelien: Diskutiert am Matthäusevangelium’. Kyoung Shik Min, Die früheste Überlieferung des Matthäusevangeliums (bis zum 3./4. Jh.). Edition und Untersuchung, ANTF 34 (Berlin: de Gruyter, 2006), offers a full application of her approach to the early papyri of Matthew. 43.  One of my recent PhD students conducted such an analysis of early papyri of the Gospel of John: Lonnie D. Bell, ‘Textual Stability and Fluidity Exhibited in the Earliest Greek Manuscripts of John: An Analysis of the Second/Third-century Fragments with Attention also to the More Extensive Papyri (P45, P66, P75)’ (PhD diss., University of Edinburgh, 2015). The published version of Bell’s study it to appear in the NTTSD series (Leiden: Brill). 44.  Michael Holmes proposed that Zuntz’s classic study focused on P46 provides an instructive model for developing a more soundly based theory of the early NT text: ‘The Text of the Epistles Sixty Years After: An Assessment of Günther Zuntz’s Contribution to Text-Critical Methodology and History’, in Transmission and

42

Texts and Artefacts

One further observation about the effects of early NT papyri: With a few notable exceptions, the variants in them are those that we already knew from later witnesses (most often in later Greek witnesses, sometimes only in the Latin or Coptic versions). But we do not find in the early papyri the larger variants that reflect a major change in the text, e.g., the pericope of the adulterous woman, the long ending of Mark, or the major additions in the Codex Bezae text of Acts.45 I highlight two net effects of these data. First, they confirm the earlier view that the great majority of textual variants emerged very early, likely in the second century. But, second, it appears that the earliest state of the text of NT writings was no more diverse than what we have in later witnesses (the versions and the major Greek codices of the fourth to sixth centuries), and that perhaps significant textual variation continued (or achieved success in the manuscript tradition) well beyond the earliest period.46 So, it is now increasingly dubious to cling to the notion sometimes asserted in the past that the second century was a particular period of ‘wild’ textual variation, far greater than what we see in the fourth century and thereafter.47 I reiterate the observation that the early papyri certainly attest varying levels of fluidity in the NT text, and a readiness among some Christians to ‘improve’ the text in various ways (e.g., stylistic changes, harmonizations, etc.); but these manuscripts do not on the whole reflect a careless or ‘wild’ transmission-attitude and process.48 Reception: New Testament Text-Critical and Exegetical Studies, ed. J. W. Childers and D. C. Parker (Piscataway, NJ: Gorgias, 2006), 89–113. In the same volume, I survey major factors that shaped the early transmission of the NT: ‘The New Testament in the Second Century: Text, Collections and Canon’, 3–27 (republished as Chapter 1 of the present volume). I include a brief discussion of recently published papyri, proposing that the early manuscripts exhibit a variety of copying practices and purposes. 45.  The earliest witness with the pericope of the adulterous woman (placed at John 7:53–8:11) is Codex Bezae (fifth century). The earliest Greek witnesses with Mark 16:9–20 are Codex A & D (fifth century). There are indications in Patristic writers that both passages were known earlier, but these are the earliest extant copies of NT writings to include them. 46.  I briefly explore this suggestion in ‘The Pericope Adulterae: Where from Here?’, in The Pericope of the Adulteress in Contemporary Research, ed. David Alan Black and Jacob N. Cerone (London: Bloomsbury T&T Clark, 2016), 147–58. 47.  As Epp observed, the early papyri ‘are not conspicuous for furnishing a mass of new, meaningful variant readings’, but instead typically attest variants already known from later manuscripts: Epp, ‘Are Early New Testament Manuscripts Truly Abundant?’, 106. 48.  For further discussion of these issues and for references to other scholarly literature, see Hurtado, ‘The New Testament in the Second Century’, esp. 6–19, and in this volume 3–27.

2. The Early New Testament Papyri

43

Other Historical Issues In addition to their great importance in NT textual criticism, these early NT manuscripts cast invaluable light on other important historical issues as well. These include various questions about the circulation of particular texts and the role and usage of texts more generally in early Christianity. For example, it is interesting to note the comparative number of copies of various NT writings in the extant earliest manuscripts (second and third centuries).49 Eighteen of these 52 manuscripts are copies of John, exactly one-third of the total, and considerably more than the next most frequently found NT writing, Matthew (9 copies). Thereafter come Acts and Romans (6 copies each), Luke (5), 1 Thessalonians and Hebrews (3 each), 1 Corinthians, Ephesians, James, Philemon and Revelation (2 each), and one copy of each of the remaining texts (Mark, 2 Corinthians, Galatians, Colossians, Philippians, 2 Thessalonians, and 1 John).50 It is also interesting to note that the most frequently found non-canonical text is Shepherd of Hermas (9 copies among manuscripts of this same period). If the comparative number of surviving copies can be taken as indicative of the comparative number of copies circulating in these early centuries, it is very interesting which texts were favoured. These earliest manuscripts also confirm that the ancient Christian preference for the codex book-form (especially, it appears, for those texts that Christians treated as scripture) goes back earlier than all of our extant evidence, into the second century and possibly earlier.51 This must be viewed in the context of an overwhelming preference for the roll in the larger literary and cultural environment of the second and third centuries. About 5% of all second-century copies of catalogued literary manuscripts (pagan, Jewish and Christian) are codices, and about 21% of all third-century manuscripts. By contrast, about 75% of all Christian manuscripts (i.e., of all literary texts, canonical and extra-canonical) dated to the second century, and about 67% of those dated to the third century are codices.

49.  I discuss the comparative number of all texts found in early Christian manuscripts (intra-canonical and extra-canonical) in The Earliest Christian Artifacts, 16–24. In that discussion I include manuscripts dated ‘third/fourth’ century, and so the figures are slightly different. But the broad results are the same. 50.  Again, as it is not yet published, I omit the ‘Willoughby Papyrus’ (P134), which preserves a few verses of Mark. 51.  For further discussion, see Hurtado, The Earliest Christian Artifacts, 43–93, which I draw upon here.

44

Texts and Artefacts

Moreover, if we confine ourselves to copies of texts that Christians treated as scriptures, the preference for the codex is exhibited even more strongly, and in fact is almost total. It is illustrative of the strength of this preference that we do not (as yet) have a single clear example of any NT text copied on an unused roll.52 We have a few cases of NT texts copied on re-used rolls (‘opisthographs’), but otherwise all early NT manuscripts are codices.53 There are a few OT texts on rolls, and these may be Christian (or Jewish) copies, but well over 90% of indisputably Christian copies of OT texts are codices. By contrast, about one third of the early copies of extra-canonical Christian texts (e.g., apocryphal texts, theological treatises, etc.) are on rolls.54 In light of the general view of the time that the roll was the more appropriate form for a valued copy of a literary text, the early Christians’ preference for the codex, and especially for their most highly valued texts, can only represent a deliberate counter-cultural choice.55 52.  The ‘Willoughby Papyrus’ (P134) is reported to have some verses of Mark on the horizontal-fibre side of a piece of papyrus, with remnants of another text on the other side. If the fragment derives from a bookroll, thus, this may be the single instance of a NT writing copied on a fresh bookroll, which then was re-used for the other text. But we must await publication for adequate opportunity to make analysis of the item. 53.  P12 is a citation of Heb 1:1 in a letter on the recto of a roll, with portions of Genesis on the verso, incorrectly included by Epp among the continuous-text copies of NT texts (‘The Papyrus Manuscripts of the New Testament’, 5). P13 (portions of Hebrews) and P18 (Revelation) are opisthographs, the NT texts copied on the outer side of a re-used roll, the inner side containing another text (for which the roll was originally prepared). P22 is two papyrus fragments with portions of John written on the verso side, the recto (horizontal fibres) of the fragments blank. I think it likely it too is an opisthograph. See my essay on this papyrus: ‘A Fresh Analysis of P.Oxyrhynchus 1228 (P22) as Artefact’, in Studies on the Text of the New Testament and Early Christianity, ed. Daniel M. Gurtner, Juan Hernandez Jr. and Paul Foster (Leiden: Brill, 2015), 206–16 (republished as Chapter 9 of the present volume). 54.  E.g., two of the nine earliest copies of Shepherd of Hermas are rolls (P.Oxy. 4706 and P.Berl. inv. 5513), as are both early fragments of Irenaeus (P.Oxy. 405 and P.Jena inv. 18+21), the Dura Gospel harmony fragment (P.Dura 10), one of the copies of Gospel of Thomas (P.Oxy. 655), and the one copy of Gospel of Mary (P.Oxy. 3525). 55.  Bagnall, Early Christian Books in Egypt, proposed that Christians adopted the codex from Roman use of the bookform, but he admits that the strength of the Christian preference for the codex, and especially for scripture texts, is remarkable and an innovation. On the wider preference for, and the characteristics of, literary bookroll, see esp. William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004).

2. The Early New Testament Papyri

45

Texts copied on re-used rolls were typically made for personal reading/ study, and so the examples of such copies of Christian texts among our earliest papyri, which include some NT texts (John, Hebrews, Revelation) and also extra-canonical texts (e.g., Shepherd of Hermas, Gospel of Thomas), can be taken as artifacts of Christians wanting such personal copies. That is, these re-used rolls are likely direct evidence of these texts being read for personal edification. It is likely, however, that most of the earliest NT manuscripts were copied for reading in churches. A variety of features that distinguish these particular manuscripts from high-quality copies of (pagan) classical texts seem intended to facilitate this usage, such as the generous-sized lettering and spacing between the lines, and the use of spaces and elementary punctuation to signal sense-units.56 That is, these and other features seem to be what we may call ‘readers’ aids’, which may have been particularly helpful to those with less than elite schooling (where one would be equipped to handle the more severe format of high-quality classical manuscripts). In another essay I have noted the contrast between the format of earliest Christian manuscripts and contemporary manuscripts of classical texts prepared for elite (pagan) social circles, proposing that the typical layout of Christian copies of scriptural texts reflects directly the more socially diverse nature of early Christian readers.57 Another distinguishing feature of Christian manuscripts, including our earliest papyri, is the practice of writing certain words in a distinctive fashion, the so-called nomina sacra.58 The Greek words in question, 56.  I provide further discussion of a number of these features of earliest Christian manuscripts in The Earliest Christian Artifacts, 155–89. Scot 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, LSTS 70 (London: T&T Clark International, 2009), 148–75, proposed a classification of earliest gospel manuscripts as intended either for public/liturgical or private reading. Some examples are more difficult to classify, but he correctly observes that we have manuscripts indicative of both reader-settings. 57.  L. 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), 49–62 (republished as Chapter 6 of the present volume). William A. Johnson, ‘Toward a Sociology of Reading in Classical Antiquity’, AJP 121 (2000): 593–627, discusses the formatting of manuscripts of classical texts prepared for reading in elite social circles; and now see idem, Readers and Reading Culture in the High Roman Empire: A Study of Elite Communities (Oxford: Oxford University Press, 2010). 58.  I give a fuller discussion in The Earliest Christian Artifacts, 95–134. I must acknowledge, however, the erroneous statements on p. 129 about the treatment

46

Texts and Artefacts

among which Θεός, Κύριος, Ἰησοῦς, and Χριστός are the most consistently treated in this manner, are all written in an abbreviated form (typically first and final letters of the inflected forms, e.g., θς, κς, Ις [but sometimes Ιη], and Χς) with a distinctive horizontal stroke placed over the abbreviation. The presence of the nomina sacra in our earliest NT papyri confirms that this scribal practice is so early that its origin pre-dates all our extant manuscripts, i.e., early second century at the latest. So, the preference for the codex and the nomina sacra taken together reflect an astonishingly early emergence of an identifiably Christian book-practice, or, as I have elsewhere described it, the earliest evidence of an early Christian ‘material and visual culture’.59 A few scholars have contended that the nomina sacra, and perhaps also the preference for the codex, derive from Jewish scribal practices (esp. Kurt Treu and Robert Kraft), but this is very much a minority position.60 It remains the case that in the body of Jewish pre-Christian manuscripts (from Judaea) there is no instance of a literary text on a codex and no instance of any of the nomina sacra forms.61 It is, however, entirely plausible to posit some kind of similarity (and influence) of a reverential attitude or motive behind the practice of the nomina sacra in the various forms that Jewish scribal treatment of YHWH took in ancient biblical manuscripts (e.g., a series of dots in place of the name, or writing YHWH in archaic Hebrew characters, or writing it in Hebrew of the name Ἰησοῦς in P46. Contrary to my statements there, in P46 the name is abbreviated in all the references cited. I am unable to account for this embarrassing error. 59.  Larry W. Hurtado, ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo, ON: Wilfrid Laurier University Press, 2000), 271–88; and also in my essay, ‘The “MetaData” of Earliest Christian Manuscripts’, in Identity and Interaction in the Ancient Mediterranean: Jews, Christians and Others – Essays in Honour of Stephen G. Wilson, ed. Zeba A. Crook and Philip A. Harland (Sheffield: Sheffield Phoenix, 2007), 149–63 (republished as Chapter 5 of the present volume). 60.  Kurt Treu, ‘Die Bedeutung des Griechischen für die Juden im römischen Reich’ Kairos NF 15, Hft. 1/2 (1973): 123–44; Robert A. 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 A. O’Sullivan (London: British Library, 2003), 51–72. 61.  Emanuel Tov, Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert, STDJ 54 (Leiden: Brill, 2004), is now the major resource for these manuscripts.

2. The Early New Testament Papyri

47

characters in Greek copies of OT texts). Likewise, the presence of the ‘readers’ aids’ mentioned earlier in Christian copies of biblical texts may very well reflect Jewish scribal practices.62 Yet the specific scribal device of the nomina sacra seems to be a Christian innovation. Conclusion Within the limits of this discussion I have been able to address only briefly some of the major ways in which the earliest NT manuscripts provide valuable resources for NT scholars. I have also tried to illustrate the usefulness of approaching these manuscripts from the perspective of the study of the larger body of Christian and non-Christian manuscripts of the same period. Most directly, of course, these particular early NT manuscripts are central to the questions at the heart of NT textual criticism. As I have indicated, these precious early copies of NT texts have already re-shaped (and will continue to influence) our views of the earliest stages of the textual transmission of the NT writings, requiring the abandonment of some earlier, confidently held positions. But these manuscripts also open up further lines of investigation and analysis that involve questions wider than traditional NT textual criticism. It is too much to ask NT scholars to become papyrologists, or, for that matter, to expect that all NT scholars should be text critics. But it is not too much to ask that NT scholars develop an awareness of the importance and relevance of the early NT manuscripts for the history of the transmission of the NT writings and, more broadly, for the investigation of Christian origins.

62.  Ibid., 131–236, discusses writing practices in the early Judean manuscripts.

Chapter 3 N ew T es ta m en t S cholar shi p a nd t he D at i n g of N ew T estame nt P apy r i *

I address the topic of the dating of NT papyri as a scholar in NT and Christian Origins with limited expertise (albeit much interest) in palaeography and papyrology. I cannot claim an independent competence for myself in dating early papyri, but have tried to acquire sufficient familiarity with matters to be able to follow the discussions and reasoning of those with such competence.1 My purposes in this essay are to lay out some reasons why competent dating of NT papyri is important to scholars in my own field, and also to engage briefly with how NT scholars have been involved in assigning and communicating dates of NT papyri. The Importance of Papyri Dates for Scholars in New Testament/Christian Origins The oldest and most obvious reason that the dating of NT manuscripts is important to scholars in NT/Christian Origins is the concern to establish the best possible critical edition of the text of NT writings, which can then serve as the basis for all subsequent work of exegesis and analysis of these writings. This is, of course, the traditional aim of NT textual criticism, often stated as the reconstruction of the/an ‘original text’ of NT writings. The term ‘original text’ has come in for a good deal of discussion in recent decades, however, and I do not have the opportunity here to survey the matter adequately. Suffice it to say that nowadays, more and more, NT textual critics refer to aiming to reconstruct the ‘initial text’ (from the *  I thank Willy Clarysse and Brent Nongbri for comments and suggestions on an earlier draft of this essay. 1.  I have urged fellow NT scholars (and aspiring ones) to take greater account of early Christian manuscripts and to acquire some acquaintance with their physical and visual features in Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006).

3. New Testament Scholarship and the Dating of New Testament Papyri

49

German term, Ausgangstext), or ‘earliest accessible text’, that is, the form of the text of a given NT writing from which all extant copies ultimately derived, the form of the text that shaped all subsequent copying of that writing.2 Toward that goal of establishing the best possible critical edition of the NT, from the emergence of the discipline (in the eighteenth century) onward there has been a fervent concern to identify and use the earliest possible witnesses to the text of NT writings. This concern fuelled the quests of nineteenth-century figures such as Tischendorf, Tregelles, and the Smith sisters, who searched libraries in various countries for ancient manuscripts and gave access to witnesses that remain central in the NT text-critical task, e.g., Codex Sinaiticus, Codex Vaticanus, and witnesses to early ‘versions’ (translations) such as the Syriac Gospels text discovered by the ‘sisters of Sinai’.3 It was from such efforts that Westcott and Hort benefited in their 1881 landmark edition of the Greek NT.4 This concern to draw upon earliest textual witnesses continued to drive the great interest shown in the appearance in the twentieth century of remarkable caches of NT papyri, particularly the Chester Beatty Biblical Papyri (in the 1930s) and the Bodmer NT Papyri (1960s).5 But even fragments of individual manuscripts dated early have drawn great excitement and scholarly interest, among which the famous Rylands Library fragment of 2.  Eldon Jay Epp, ‘The Multivalence of the Term “Original Text” in New Testament Textual Criticism’, HTR 92 (1999): 245–81; and now the extensive discussion (with copious bibliography) by Michael W. Holmes, ‘From “Original Text” to “Initial Text”: The Traditional Goal of New Testament Textual Criticism in Contemporary Discussion’, in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes, 2nd ed. (Leiden: Brill, 2013), 637–88. To simplify greatly, the issue arises in considering that there was already some copying (and perhaps some editing) of writings that became part of the NT between the ‘autograph’ (the copy written or released by the author) and any earliest copy from which subsequent copies derive. 3.  I allude to Janet Soskice’s fascinating book about the two Scottish sisters, Agnes and Margaret Smith, who identified the Sinaitic Palimpsest manuscript (fourth century) in St. Catherine’s Monastery, which contained the four Gospels in Syriac (typically referred to in critical apparatuses as ‘syrs’): The Sisters of Sinai: How Two Lady Adventurers Discovered the Hidden Gospels (New York: Knopf, 2009). For a readable review of the development of NT textual criticism, see Robert F. Hull Jr., The Story of the New Testament Text: Movers, Materials, Motives, Methods, and Models (Atlanta: Society of Biblical Literature, 2010). 4.  Brooke Foss Westcott and Fenton John Anthony Hort, The New Testament in the Original Greek (Cambridge: Macmillan & Co., 1881). 5.  Brief descriptions in Hull, The Story of the New Testament Text, 109–19.

50

Texts and Artefacts

the Gospel of John is perhaps the most well-known example (about which more in due course).6 NT textual critics continue to keep a watch out for each new volume of Oxyrhynchus papyri for more early manuscripts (typically only fragments) of NT writings, and over the last couple of decades this interest has been rewarded.7 Unquestionably, for at least most NT textual critics, the early NT papyri (those dated to the second and third centuries) have an inherent and distinctive significance as witnesses to the text of NT writings.8 Even those NT text critics who prefer what is called ‘thoroughgoing eclecticism’ (judgments about variant readings solely on the basis of ‘internal criteria’ and without regard for the dates, quality or numbers of manuscripts supporting variants) express an interest in these early manuscripts.9 6.  C. H. Roberts, An Unpublished Fragment of the Fourth Gospel in the John Rylands Library (Manchester: Manchester University Press, 1935). 7.  See, e.g., Peter M. Head, ‘Some Recently Published NT Papyri from Oxyrhynchus: An Overview and Preliminary Assessment’, TynBul 51 (2000): 1–16; J. K. Elliott, ‘Recently Discovered New Testament Papyri and Their Significance for Textual Criticism’, in Reading New Testament Papyri in Context / Lire les papyrus du Nouveau Testament dans leur contexte, ed. Claire Clivaz and J. Zumstein (Leuven: Peeters, 2011), 89–108. 8.  See the recent review of the use of early papyri in NT textual criticism by Eldon J. Epp, ‘The Papyrus Manuscripts of the New Testament’, in Ehrman and Holmes, eds., The Text of the New Testament in Contemporary Research, 1–39; and earlier discussions of the importance of early NT papyri by Epp, e.g., ‘New Testament Papyri and the Transmission of the New Testament’, in Oxyrhynchus: A City and Its Texts, ed. A. K. Bowman et al. (London: Egypt Exploration Society, 2007), 315–31; and ‘Are Early New Testament Manuscripts Truly Abundant?’, in Israel’s God and Rebecca’s Children: Christology and Community in Early Judaism and Christianity, ed. David B. Capes et al. (Waco: Baylor University Press, 2007), 77–117. In another publication I have offered a more modest survey: Larry W. Hurtado, ‘The Early New Testament Papyri: A Survey of Their Significance’, in Papyrologie und Exegese: Die Auslegung des Neuen Testaments im Licht der Papyri, ed. Jens Herzer (Tübingen: Mohr Siebeck, 2012), 1–18 (republished as Chapter 2 of the present volume). 9.  See, e.g., the comments by perhaps the leading advocate of ‘thoroughgoing eclecticism’, J. K. Elliott, ‘Recently Discovered New Testament Papyri and Their Significance for Textual Criticism’, in Clivaz and Zumstein, eds., Reading New Testament Papyri in Context, 89–108. On the methodological issue, cf. J. Keith Elliott, ‘Thoroughgoing Eclecticism in New Testament Textual Criticism’, in Ehrman and Holmes, eds., The Text of the New Testament in Contemporary Research, 745–70; Michael W. Holmes, ‘Reasoned Eclecticism in New Testament Textual Criticism’, in Ehrman and Holmes, eds., The Text of the New Testament in Contemporary Research, 771–802.

3. New Testament Scholarship and the Dating of New Testament Papyri

51

In addition to the task of trying to restore the ‘initial’ or ‘earliest accessible’ text (wording) of NT writings, there is what is usually regarded as the allied concern to obtain as clear a picture as we can of the earliest textual transmission of NT writings, the early history of their text. For this task, obviously, the oldest manuscripts are usually regarded as supremely important and so, for this reason also, the most reliable dating of NT papyri is essential. Consider this example of the impact of early NT papyri on scholarly views of the history of the textual transmission of the NT. Prior to the availability of key early NT papyri it had been asserted often that sometime in the early fourth century there was an ecclesiastically sponsored recension of the text of the Gospels that is reflected in the so-called Alexandrian Text, for which the fourth-century Codex Vaticanus and Codex Sinaiticus are the classic witnesses. But the publication of early Gospel papyri in the 1960s, particularly P75 (P.Bodmer XIV–XV), typically dated ca. 175–225 CE and exhibiting a text of Luke and John that is virtually identical to Codex Vaticanus, pulled the rug from under that claim, requiring a serious re-think of matters.10 As I have noted elsewhere, earlier still, the publication of the Chester Beatty Gospels–Acts codex, P45, in the 1930s likewise had a dramatic effect, at once giving a snapshot of the text of these writings a century earlier than what had been their earliest Greek witnesses.11 As yet another example, in a magisterial work insufficiently engaged with today, Günther Zuntz explored creatively and cogently the implications of the Chester Beatty Pauline codex (P46) for tracing the earliest textual history of the Pauline Corpus.12 10.  See, e.g., Gordon D. Fee, ‘P75, P66, and Origen: The Myth of Early Textual Recension in Alexandria’, in New Dimensions in New Testament Study, ed. Richard N. Longenecker and Merrill C. Tenney (Grand Rapids: Zondervan, 1974), 19–45, republished in E. J. Epp and G. D. Fee, Studies in the Theory and Method of New Testament Textual Criticism (Grand Rapids: Eerdmans, 1993), 247–73. But the early third-century dating of P66 (P.Bodmer II) has been challenged recently by Brent Nongbri, ‘The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P.Bodmer II (P66)’, Museum Helveticum 71 (2014): 1–35, and he promises a forthcoming article in which he will query the dating of P75 as well. 11.  Larry W. Hurtado, ‘P45 and the Textual History of the Gospel of Mark’, in The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels: The Contribution of the Chester Beatty Gospel Codex P45, ed. Charles Horton (London: T&T Clark International, 2004), 132–48. 12.  Günther Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum, The Schweich Lectures, 1946 (London: Oxford University Press for the British Academy, 1953; reprint edition: Eugene, OR: Wipf & Stock, 2013).

52

Texts and Artefacts

Although ideas of an early fourth-century ecclesiastical recension have been laid to rest, it is still often asserted (assumed?) that in the earliest period of the textual transmission of NT writings both accidental and deliberate changes were much more frequent, and that especially the latter type of change was more freely made, than in later centuries. Indeed, some portray the second century in particular as a time of ‘wild’ and very ‘free’ or ‘loose’ copying, such that it is nigh-impossible to say with confidence what the ‘initial’ wording of NT writings may have been. The extant early papyri do not quite reflect this, showing instead a degree of textual variation no greater than (and perhaps even less than) what we see in manuscripts of later centuries. The curious response of a few scholars, however, has been to dismiss these NT papyri as products of a supposed late second-century recension (a proposal that, with most NT textual critics, I find more desperate than persuasive)!13 In any case, it is understandably important to identify witnesses that can be dated sufficiently early to provide evidence (or at least hints) of the state and transmission of NT writings in this crucial period.14 Among those who have addressed the matter are Barbara Aland and her student Kyoung Shik Min. Their approach was to compare respectively early papyri of Matthew and John with the wording of the Nestle-Aland Greek NT, classifying the papyri as to how closely they reflected this modern scholarly effort to represent the wording of the ‘initial text’ of these writings.15 They posited a certain spectrum of copying practice evidenced 13.  E.g., Helmut Koester, ‘The Text of the Synoptic Gospels in the Second Century’, in Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission, ed. William L. Petersen (Notre Dame/London: University of Notre Dame Press, 1989), 19–37; and for a similar stance, William L. Petersen, ‘The Genesis of the Gospels’, in New Testament Textual Criticism and Exegesis: Festschrift J. Delobel, ed. Adelbert Denaux (Leuven: Peeters, 2002), 33–65. Cf., however, Larry W. Hurtado, ‘The New Testament in the Second Century: Text, Collections and Canon’, in Transmission and Reception: New Testament Text-Critical and Exegetical Studies, ed. J. W. Childers and D. C. Parker (Piscataway, NJ: Gorgias, 2006), 3–27 (republished as Chapter 1 of the present volume). 14.  E.g., J. K. Elliott, ‘The New Testament Text in the Second Century: A Challenge for the Twenty-First Century’, New Testament Textual Research Update 8 (2000): 1–14. 15.  Barbara Aland, ‘Das Zeugnis der frühen Papyri für den Text der Evangelien: Diskutiert am Matthäusevangelium,’, in The Four Gospels 1992, ed. F. Van Segbroeck et al. (Leuven: Leuven University Press, 1992), 325–35; idem, ‘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, STAR 8 (Assen:

3. New Testament Scholarship and the Dating of New Testament Papyri

53

in the early papyri, some showing a relatively strict copying, and others exhibiting varying degrees of a freer attitude and a less exact concern. This is a judgment that roughly corresponds with that reached by some other scholars as well, who have taken other approaches to the early papyri.16 Although not reflecting a ‘wild’ state of the text of NT writings, this diversity in the textual ‘complexion’ of early papyri points against the assertion of a late second-century ecclesiastical recension to which early papyri conform. Instead, the textual diversity in NT papyri suggests a certain variety in copying tendencies. On the one hand, we have what may be termed ‘popular’ copying tendencies that produced manuscripts with comparatively more accidental and/or deliberate changes. On the other hand, we also see other copying efforts across the same early period that seem to have involved a greater concern for using superior-quality archetypes and were aimed at comparatively more exacting transmission of the texts. The copying tendencies toward a ‘freer’ transmission may well have been more common, but the latter, more exacting, kind of efforts, Royal van Gorcum, 2003), 19–38; Kyoung Shik Min, Die früheste Überlieferung des Matthäusevangeliums (bis zum 3./4. Jh): Edition und Untersuchung, ANTF 34 (Berlin: de Gruyter, 2005). For a description of the approach taken, see Barbara Aland, ‘Kriterien zur Beurteilung kleinerer Papyrusfragmente des Neuen Testaments’, in New Testament Textual Criticism and Exegesis: Festschrift J. Delobel, ed. Adelbert Denaux (Leuven: Peeters, 2002), 1–13. Personally, I regard this approach as open to criticism as appearing to involve a certain circular-logic. But, in any case, I prefer the approach being taken by Lonnie D. Bell in his doctoral dissertation (‘Textual Stability and Fluidity Exhibited in the Earliest Manuscripts of John: An Analysis of the Second/Third-century Fragments with Attention also to the More Extensive Papyri [P45, P66, P75]’ [University of Edinburgh, 2015]), in which the variants in papyri in the Gospel of John are analysed in comparison with one another, not in comparison with an external standard such as the Nestle-Aland edition. A published version is to appear in the NTTSD series (Leiden: Brill). 16.  E.g., Eldon Jay Epp, ‘The Significance of the Papyri for Determining the Nature of the New Testament Text in the Second Century: A Dynamic View of Textual Transmission’, in Petersen, ed., Gospel Traditions in the Second Century, 71–103, republished in Eldon Jay Epp and Gordon D. Fee, Studies in the Theory and Method of New Testament Textual Criticism, SD 45 (Grand Rapids: Eerdmans, 1993), 274–97; Hurtado, ‘The New Testament in the Second Century’; Michael W. Holmes, ‘Text and Transmission in the Second Century’, in The Reliability of the New Testament: Bart Ehrman and Daniel Wallace in Dialogue, ed. Robert B. Stewart (Minneapolis: Fortress, 2011), 47–65. See the analysis offered by Bell, ‘Textual Stability and Fluidity’. There, Bell examines the nature and extent of textual variation in secondand third–century papyri of the Gospel of John, and tends toward a similar view as well.

54

Texts and Artefacts

though perhaps less common in the second century, appear to go back just as early.17 A key factor in understanding the early transmission-history of NT writings is an accurate grasp of how the copyists of that period actually did their work. Taking early NT papyri as our best evidence, James Royse made a detailed and extended study of the ‘scribal habits’ that they exhibit, his work correcting some earlier generalizations, and now enhancing considerably the basis on which to proceed in refining use of what is called ‘internal criteria’ for evaluating variants.18 Without engaging here with the specifics of Royse’s study, my point is simply that the early dates ascribed to these NT papyri make them particularly crucial for establishing a view of how early copyists operated. To cast the net more broadly beyond NT textual criticism, the production of critical editions of the NT provides the basis upon which virtually all other scholarly work in the NT rests. So, if early papyri are directly crucial for the text-critical task, they are indirectly crucial for practically all of the wider exegetical and historical tasks with which the majority of NT scholars are engaged. Whether, for example, it be commentaries, analyses of the Gospel-traditions (whether source criticism, form criticism, redaction criticism, or literary criticism), discussions of NT theology, social-description of early Christian circles, or any other matter for which the NT writing are important, the possibility and exactness of this work all rest upon efforts to establish the earliest-available text of these writings. For this task the secure dating of earliest witnesses is crucial. To cite other examples of wider historical questions about earliest Christianity for which early papyri are central, these witnesses may allow us to gauge which writings were circulating among Christians in the earliest period, and perhaps even to estimate their comparative popularity. Based on counts of copies of texts in the extant Christian papyri from the second and third centuries, it is interesting, for example, that the single most frequently attested writing is Psalms (18 copies), and that among Christian writings the Gospel of John is most frequent (18 17.  This also roughly reflects the judgments of Zuntz, The Text of the Epistles, who proposed an influence of concerns for exact copying stemming from the copying of classical literary texts in Alexandria in particular. See also my discussion in ‘The New Testament in the Second Century’. Moreover, E. G. Turner observed a somewhat similar duality in copying tendencies in classical texts, a freer practice in Homeric papyri of the Ptolemaic period and a comparatively ‘steady respect for the authority of the text’ (that he attributed largely to scholars in Alexandria): Greek Papyri: An Introduction (Oxford: Clarendon, 1980), 107–110. 18.  James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTS 36 (Leiden: Brill, 2007).

3. New Testament Scholarship and the Dating of New Testament Papyri

55

copies), followed by the Gospel of Matthew (12 copies) and Shepherd of Hermas (11 copies), whereas we have only one early copy of Mark from the centuries before Constantine.19 These data also show that in this early period the Gospels circulated mainly as individual writings (manuscripts such as P75 and P45 apparently early exceptions), that some were more frequently copied and read than others, and that some (ultimately) non-canonical writings (esp. Hermas) were more popular than many that became canonical.20 Moreover, from the physical and visual features of the earliest papyri, we can form inferences about the people who copied them (e.g., in the main, copyists whose intention and/or abilities lay more in producing a readable text than an elegantly formatted one).21 We can also make inferences about readers and the intended uses of writings. For example, it is clear that some copies of Christian writings were made for private reading/study, and other copies for reading in circles of believers gathered for worship, providing us direct evidence of the ‘reading culture’ of earliest Christianity (i.e., the situations in which writings were read).22 As one indication of how physical format suggests intended usage, it appears that early Christians favoured the codex, especially (but not exclusively) 19.  I cite here figures from the discussion of these data in my book, The Earliest Christian Artifacts, 15–41. These figures include papyri dated third/fourth century. 20.  Accepting the date commonly assigned to it, P46 shows that manuscripts containing a collection of Paul’s letters also appeared early, certainly by ca. 200 CE, and perhaps much earlier. On the dating question, see S. R. Pickering, ‘The Dating of the Chester Beatty-Michigan Codex of the Pauline Epistles (P46)’, in Ancient History in a Modern University: Essays in Honor of Edwin A. Judge, ed. T. W. Hillard et al., 2 vols. (Grand Rapids: Eerdmans, 1998), 2:216–27. 21.  Kim Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Trans­ mitters of Early Christian Literature (Oxford: Oxford University Press, 2000), argues that most manuscripts of Christian texts were made informally, copies passed from one believer to another for subsequent ‘in house’ copying. 22.  See, e.g., Scott D. Charlesworth, ‘Public and Private: Second- and ThirdCentury Gospel Manuscripts’, in Jewish and Christian Scripture as Artifact and Canon, ed. Craig A. Evans and H. Daniel Zacharias, LSTS 70 (London: T&T Clark International, 2009), 148–75; and now idem, Early Christian Gospels: Their Production and Transmission, Papyrologica Florentina (Florence: Gonnelli, 2014); and my own discussions: Larry W. Hurtado, ‘What Do the Earliest Christian Manuscripts Tell Us About Their Readers?’, in The World of Jesus and the Early Church: Identity and Interpretation in Early Communities of Faith, ed. Craig A. Evans (Peabody, MA: Hendrickson, 2011), 179–92; idem, ‘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), 49–62 (republished as Chapter 6 of the present volume).

56

Texts and Artefacts

for those writings that they used/regarded as scripture (i.e., read in corporate worship), whereas Christian texts copied on rolls (and, even more so, copies on re-used rolls) seem to have been intended for personal/private usage.23 Likewise, miniature copies of writings were almost certainly for personal reading (e.g., the miniature roll containing The Gospel of Thomas, P.Oxy. 655).24 So, in the features of the earliest-dated papyri it is particularly interesting that we can see evidence of the use of certain texts as scripture (i.e., read in corporate worship). The early papyri also provide evidence of the emergence of what I have called a Christian ‘material and visual culture’. In addition to the Christian preference for the codex already mentioned, I draw attention to the ‘nomina sacra’ and the monogram-like device called the ‘staurogram’ (the tau-rho compendium), which reflect an apparently distinctive early Christian scribal tradition. The ‘staurogram’, which appears in NT papyri dated ca. 175–225 CE (P66, P75), is particularly significant as a visual expression of early Christian faith. In these earliest uses, the device seems to function as a visual reference to the crucified Jesus, and some 150 years earlier than what has commonly been cited by historians of early Christian art as the earliest depictions of Jesus on the cross.25 The nomina sacra (distinctive abbreviated forms of key words, especially Θεός, Κύριος, Χριστός, and Ἰησοῦς) strongly suggest that Christian texts were copied by Christians. Moreover, of course, readers had to acquaint themselves with this scribal convention in order to read Christian texts. 23.  E.g., the Greek fragments of the Gospel of Thomas all seem to be from three manuscripts likely prepared for personal reading/study. See my analysis: Larry W. Hurtado, ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’, in Das Thomasevangelium: Entstehung – Rezeption – Theologie, ed. Jörg Frey, Enno Edzard Popkes and Jens Schröter, BZNW 157 (Berlin: de Gruyter, 2008), 19–32 (republished as Chapter 10 of the present study). 24.  Thomas Kraus, ‘Die Welt der Miniaturbücher in der Antike und Spätantike. Prolegomena und erste methodische Annäherungen für eine Datensammlung’, SNTSU 35 (2010): 79–110. 25.  On the ‘nomina sacra’ and the ‘staurogram’, see Larry W. Hurtado, ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo, ON: Wilfrid Laurier University Press, 2000), 271–88; idem, Earliest Christian Artifacts, 95–134 (on the nomina sacra), and 135–54 (on the ‘staurogram’). Commendably, Robin Margaret Jensen has recognized the significance of the early instances of the ‘staurogram’ as visual references to the crucified Jesus: Understanding Early Christian Art (London: Routledge, 2000), 138.

3. New Testament Scholarship and the Dating of New Testament Papyri

57

New Testament Scholars and Dating Manuscripts I turn now to consider how NT scholars have dealt with the question of the dates of early Christian manuscripts, particularly NT papyri. As discussed here already, NT scholars are certainly keen to have access to the earliest Christian manuscripts available. But have NT scholars conducted themselves with due circumspection in their handling of questions about the dates of early papyri? I consider this question in light of recent criticism alleging a tendency among (some?) NT scholars to posit or prefer unduly early dates. Alleging an ‘excessively self-enclosed character and absence of selfawareness’ of much scholarship that has led some to conclusions that are ‘profoundly at odds with fundamental social realities of the ancient world and with basic probability’, Roger Bagnall proposed that the number of Christian papyri assigned to the second century is too great, and that nearly all should probably be assigned dates no earlier than the third century.26 Precisely who he included in the ‘self-enclosed’ scholarship he criticized, he did not specify.27 In any case, instead of considering the particulars of specific papyri that may have been dated too early, Bagnall proceeds to contend, on the basis of what I regard as ‘guestimates’ of the possible number and percentage of Christians in the general population of second-century Egypt, that the number of Christian papyri should be ‘proportionate to the Christians’ share of the population at a given moment’.28 So, he concludes, instead of the eight or so copies of Christian texts widely dated to the second or early third century, we should have far fewer, perhaps only one from the second century in particular. To speak for myself, for several reasons I find this approach simplistic and unpersuasive.29 For example, although he notes that Christians inherited ‘a writing-centered culture’, he fails to consider whether Christians might 26.  Roger Bagnall, Early Christian Books in Egypt (Princeton: Princeton University Press, 2009), citing 1. 27.  From Bagnall’s comments in a panel discussion of his book in a session held as part of the Annual Meeting of the Society of Biblical Literature, however, NT scholars (or at least some of them) seem to have been the object of his scorn. 28.  Bagnall, Early Christian Books, 17–24. I refer to his ‘guestimates’ as the best term to describe the dubious basis for his calculations. 29.  I echo briefly here criticisms that I lodged in my review of Bagnall’s book in the Review of Biblical Literature: http://www.bookreviews.org/pdf/7755_9195.pdf, my criticisms cited with approval by Don Barker, ‘The Dating of New Testament Papyri’, NTS 57 (2011): 582 (571–82).

58

Texts and Artefacts

well therefore have had greater reasons than the general population for making copies of their writings. Thus they may have generated a number of copies disproportionate to the percentage of Christians in the general population.30 Moreover, it seems a very dubious procedure to posit numbers of Christians in Egypt as a basis for arguing the matter, when this requires one to posit something that itself requires better justification. His only sustained discussion of dates of NT papyri, is in a rather caustic critique of the claim of the late Carsten Peter Thiede that P64 (P.Mag. Greek 17) is a fragment of a first-century copy of the Gospel of Matthew.31 It is noteworthy that in the course of this discussion, Bagnall draws upon earlier critiques of Thiede, especially those by several NT scholars.32 Actually, there is (to my knowledge) scant indication that 30.  Bagnall, Early Christian Books, 2. His argument that the price of books produced by professional copyists would have made copies of texts prohibitive for most Christians (‘The Economics of Book Production’, 50–69) is likewise simplistic, reflecting a failure to consider the indications that most early Christian manuscripts were copied informally from texts passed from hand to hand, as contended by HainesEitzen (Guardians of Letters). On many questions about the use and copying of books among Christians, Harry Y. Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven: Yale University Press, 1995) remains important. 31.  Bagnall, Early Christian Books, 25–40. Cf. Carsten P. Thiede, ‘Papyrus Magdalen Greek 17 (Gregory-Aland P64): A Reappraisal’, ZPE 105 (1995): 13–20. The Magdalene College fragments were initially catalogued as Magdalene College Gr 18, but were subsequently re-catalogued as P. Mag. Greek 17. 32.  E.g., Klaus Wachtel, ‘P64/67: Fragmente des Matthäusevangeliums aus dem 1. Jahrhundert?’, ZPE 107 (1995): 73–80. Note also E. Puech, ‘Des fragments grecs de la Grotte 7 et le Nouveau Testament? 7Q4 et 7Q5, et le papyrus Magdalen grec 17 = P64’, RB 102 (1995): 570–84; Peter M. Head, ‘The Date of the Magdalen Papyrus of Matthew (P.Magd.Gr. 17 = P64): A Response to C. P.Thiede’, TynBul 46 (1995): 251–85; Harald Vocke, ‘Magdalen 17 ‒ Weitere Argumente gegen die Frühdatierung des angeblichen Jesus-Papyrus’, ZPE 113 (1996): 153–57; and Sigrid Peterson’s review (initially published on the ‘Ioudaios’ discussion-list): http://ccat. sas.upenn.edu/~petersig/thiede.txt.final.reply. It is now widely accepted that P64 and P67 are portions of the same codex. Whether P4 is another portion of the same codex remains under debate. Cf. Philip W. Comfort, ‘Exploring the Common Identification of Three New Testament Manuscripts: P4, P64 and P67’, TynBul 46 (1995): 43–54; T. C. Skeat, ‘The Oldest Manuscript of the Four Gospels?’, NTS 43 (1997): 1–34 (republished in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott [Leiden: Brill, 2004], 158–92); Peter M. Head, ‘Is P4, P64 and P67 the Oldest Manuscript of the Four Gospels? A Response to T. C. Skeat’, NTS 51 (2005): 450–57; Tommy Wasserman, ‘A Comparative Textual Analysis of P4 and P64+67’, TC 15 (2010): 1–26; Scott D. Charlesworth, ‘T. C. Skeat, P64 + 67 and P4, and the Problem

3. New Testament Scholarship and the Dating of New Testament Papyri

59

many NT scholars have been swayed by the allure of Thiede’s claims, and so any possible implication that NT scholarship as a whole has been ‘self-enclosed’ and insufficiently critical in this matter would seem unfounded. More recently, however, Brent Nongbri published a critique of early dates assigned to the famous Rylands fragment of the Gospel of John, complaining about ‘the way scholars of the New Testament have used and abused papyrological evidence’ with regard to this important papyrus.33 Granted, Nongbri does provide citations of a number of NT scholars that show a kind of drift toward (or even beyond) the earlier end of the time-frame posited by palaeographers. Nongbri’s accusation is that ‘Kurt Aland appears to be the one guilty of popularizing this earlier dating’ of 125 CE as ‘the later limit’, and Nongbri cites others who have echoed a similarly early dating.34 Indeed, although the table of manuscripts in the Nestle-Aland edition simply assigns the papyrus to the second century, the online Kurzgefasste Liste of the Münster Institute for New Testament Text-Critical Research (established by Aland) does date P52 more narrowly to 100–125 CE.35 But Nongbri also acknowledges other NT scholars who have expressed greater caution in the matter.36 I regard as justifiable his complaints about some influential NT scholars positing an unduly precise, and over-confidently early, dating of this papyrus; and he rightly urges that palaeographical dating does not allow much more than a time-frame of ca. 50–75 years at best.37 He is also of Fibre Orientation in Codicological Reconstruction’, NTS 53 (2007): 582–604; Charles E. Hill, ‘Intersections of Jewish and Christian Scribal Culture: The Original Codex Containing P4, P64 and P67, and Its Implications’, in Among Jews, Gentiles, and Christians in Antiquity and the Middle Ages, ed. Reidar Hvalvik and J. Kaufman (Trondheim: Tapir Academic, 2011), 75–91. 33.  Brent Nongbri, ‘The Use and Abuse of P52: Papyrological Pitfalls in the Dating of the Fourth Gospel’, HTR 98 (2005): 46 (23–48). 34.  Nongbri, ‘Use and Abuse’, 30 n. 22. 35.  http://ntvmr.uni-muenster.de/liste (consulted January 2014). 36.  Among them are Bart D. Ehrman, most recently in ‘The Text as Window: New Testament Manuscripts and the Social History of Early Christianity’, in Ehrman and Holmes, eds., The Text of the New Testament in Contemporary Research, 803–30 (819 n. 54); and my own suggestion that P52 may be more safely dated to the late second century: ‘P52 (P. Rylands Gk. 457) and the Nomina Sacra: Method and Probability’, TynBul 54 (2003): 1–14 (7 n. 20). 37.  On the limits of palaeographical dating, see, e.g., E. G. Turner, Greek Manuscripts of the Ancient World, 2nd ed. (London: Institute of Classical Studies, 1987), 20–24.

60

Texts and Artefacts

correct in observing that one reason for the tendency of some NT scholars to push the date of P52 extremely early has been the desire to invoke the papyrus to defend an early date of the composition of the Gospel of John.38 But, as he judges, it is unwise to make so much rest upon an extremely early and overly precise second-century date of P52. Don Barker, for example, has likewise urged caution, judging it difficult to be more precise in dating this papyrus than sometime in the second or third century.39 As for the time-frame for the composition of the Gospel of John, there are other data that are probably as important as the date of P52 (and which seem to me to make a late first-century time most likely).40 More recently still, there is an important study of the dating of early NT manuscripts by Pasquale Orsini and Willy Clarysse, whose sub-title, ‘A Critique of Theological Palaeography’, signals their impetus.41 Early in the article they complain about ‘some New Testament scholars’ positing 38.  See Nongbri, ‘Use and Abuse’, 23–25, for examples. 39.  Barker, ‘The Dating of New Testament Papyri’, 575. Barker also urged that P64+67+4 cannot be dated more precisely than mid-second to mid-fourth century (578), and that P46 should be assigned 150–250 CE (581). But cf. Stanley E. Porter, ‘Recent Efforts to Reconstruct Early Christianity on the Basis of its Papyrological Evidence’, in Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament, ed. Stanley E. Porter and Andrew Pitts (Leiden: Brill, 2010), 71–84, who mounts a case for dating both P52 and P.Egerton 2 to ca. 100–150 CE (82; cf. 84, ‘somewhere in the middle second century, perhaps tending toward the early part of it’). 40.  See recent studies of the evidence of use of the Gospel of John in the second century: Titus Nagel, Die Rezeption des Johannesevangeliums im 2. Johrhundert: Studien zur vorirenäischen Aneigung und Auslegung des vierten Evangeliums in christlicher und christlich-gnostischeer Literatur (Leipzig: Evangelische Verlagsanstalt, 2000); Charles E. Hill, The Johannine Corpus in the Early Church (Oxford: Oxford University Press, 2004); Tuomas Rasimus, ed., The Legacy of John: Second-Century Reception of the Fourth Gospel, NovTSup 132 (Leiden: Brill, 2010). In his contribution to the last volume, István Czachesz pushes for a second-century date of the Gospel of John, citing Nongbri’s article as allowing for this: ‘The Gospel of the Acts of John: Its Relation to the Fourth Gospel’, in Rasimus, ed., The Legacy of John, 49–72 (69 n. 75); but cf. Anne Pasquier’s essay in the same volume, ‘Influence and Interpretation of the Gospel of John in Ancient Christianity: The Example of Eugnostos (NHC III,3 and V,1)’, 210–31, who cites P52 as ‘[t]he most ancient fragment’ of John, dating it ‘between 94–127 CE’ (229 n. 44)! It is curious that in this multi-author volume, there is no contribution on the early papyri, even though John is the most-frequently represented Christian text in early papyri. 41.  Pasquale Orsini and Willy Clarysse, ‘Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography’, ETL 88 (2012): 443–74.

3. New Testament Scholarship and the Dating of New Testament Papyri

61

earlier dates for some NT papyri than those commonly affirmed in ‘an uneasy consensus’ that they see as commendable and represented in the dates given in the table in the Nestle-Aland edition.42 Their critique of offending ‘New Testament scholars’ commences with a brief mention of Carsten Thiede (again!) and José O’Callaghan (particularly claims by these two figures that portions of some NT writings could be identified among the fragments from Qumran). But Orsini and Clarysse focus more on Philip Comfort and Karl Jaroš, and their early dating of certain NT papyri.43 The critique of Comfort and Jaroš seems well founded and justified. But I want to register some mild dissatisfaction over their characterization of the contours of scholarship more broadly. First, it is not evident that all those cited are recognized as contributing scholars in NT/Christian Origins.44 Moreover, by any measure of scholarly opinion, their very early dates of some NT papyri are (to my knowledge) commonly regarded as ‘maverick’ views that have not been taken up favourably by scholars generally, whether in NT studies or in papyrology. Among palaeographers as well as NT scholars there have been (and remain) what may be 42.  Orsini and Clarysse, ‘Early New Testament Manuscripts’, 444–45. 43.  See esp. Philip W. Comfort and David P. Barrett, The Text of the Earliest Greek New Testament Manuscripts: A Corrected, Enlarged Edition of the Complete Text of the Earliest New Testament Manuscripts (Wheaton, IL: Tyndale House, 2001); and now Philip Comfort, Encountering the Manuscripts: An Introduction to New Testament Paleography and Textual Criticism (Nashville: Broadman & Holman, 2005), in which Comfort seems to have adjusted a few dates a little later. It is a bit puzzling that I could find only a couple of reviews of the book, none by a recognized figure in NT textual criticism or in papyrology. But Orsini and Clarysse do engage it and Comfort’s earlier work at numerous points (‘Early New Testament Manuscripts’). Karl Jaroš, Das Neue Testament nach den älteseten griechischen Handschriften: Die handschriftliche Überlieferung des Neuen Testaments vor Codex Sinaiticus und Codex Vaticanus (CD-ROM) (Ruhpolding: Franz Philipp Rutzen; Würzburg: Echter, 2006), seems to have had little notice among NT scholars. 44.  Thiede was essentially an autodidact (and relentless self-publicist), who came to popular attention through his dubious claims about the Qumran fragments and the Magdalen College fragment of Matthew (Magdalen Greek 17; P64). For a brief (and positive) biographical entry, see http://en.wikipedia.org/wiki/Carsten_Peter_Thiede. O’Callaghan established an institute for papyrology, but his claims seem to have convinced virtually no one but Thiede. Karl Jaroš (University of Vienna) is an OT scholar with no record of expertise in NT or in papyrology: http://homepage.univie. ac.at/karl.jaros/. Philip Comfort has published a number of books dealing with early manuscripts, but is mainly known for his assigning earlier dates to certain NT papyri than commonly accepted in the field.

62

Texts and Artefacts

described as clusters or ‘camps’ of scholars advocating comparatively earlier or later time-frames for manuscripts.45 But there is no significant camp of ‘theological palaeography’ among NT scholars that is supposedly influenced by the very early dates of NT papyri advocated by Comfort and Jaroš.46 Second, although Orsini and Clarysse appear to set the judgments of papyrologists over against ‘theological papyrology’ (which implicitly seems identified by them with NT scholars), in fact (as reflected in the footnotes in their article) the most robust refutations of the extremely early dates proposed by Thiede and others have come from NT scholars, not from papyrologists.47 I make one final relevant observation about the results of the Orsini/ Clarysse article. At the end, they provide a table of 91 early NT manuscripts, with columns giving for each one the various dates assigned by Comfort and Barrett, Jaroš, and Nestle-Aland, and their own judgments.48 In comparison with the dates in the Nestle-Aland column (which, again, represents what they call the ‘uneasy consensus’ among most NT scholars), in the overwhelming number of cases (by my count, 73 of the 91 manuscripts listed) Orsini and Clarysse agree. They do propose somewhat later dates for 13 (in general, from a few decades to roughly a century later), but also, notably, somewhat earlier dates than in NestleAland for another five.49 As a result (contra Bagnall’s proposal) they date 45.  Bagnall (Early Christian Books in Egypt, 15) posits one camp of palaeographers tending toward earlier dates for Christian papyri (Bell, Roberts, and Skeat), and another tending toward somewhat later dates (Turner, Thomas, Grenfell and Hunt). Porter (‘Recent Efforts’, 73–74) posits two additional clusters or camps of opinion, citing Comfort as an advocate of still earlier dates (but also citing Adolf Deissmann, Ulrich Wilken and Wilhelm Schubart as similarly having supported early dates of certain manuscripts), and a fourth group who prefer dates later than all the others (Michael Gronewald and Nongbri). 46.  In private communication, however, Clarysse reiterated a concern that the views of these ‘outriders’ might be having (or might come to have) an influence upon NT scholars who are unable to judge matters well, and so may find these early views dates very attractive. 47.  See, e.g., the works cited in Orsini and Clarysse, ‘Early New Testament Manuscripts’, 444 n. 12. Their references to reviews of the Comfort/Barrett book are a bit misleading (445). The reviewers commend the book as a handy quick-reference tool for accessing the texts, but do not endorse the early dates. 48.  Ibid., 469–72. 49.  P30 (175–225), P64+67+4 (175–200), 0171 (175–225), 0212 (175–225), 0308 (250–300). It is very interesting that their dates for P64+67+4 and 0171 are

3. New Testament Scholarship and the Dating of New Testament Papyri

63

seven NT manuscripts to sometime in the second century (or late second/ early third century).50 Conclusion Unquestionably, the dates of early Christian papyri, and NT papyri in particular, are important for NT scholarship. Moreover, as NT scholars more broadly come to recognize the significance of the ‘para-textual’ data as well as the textual data provided by early papyri, the accurate/secure dating of these important artifacts will be all the more important. Very few NT scholars have the competence to make authoritative judgments about the matter, and so it will continue to be crucial to draw on the expertise of papyrologists and palaeographers (even if they do not always agree among themselves).51 By and large, I contend, NT scholars have been judicious in their invoking of the dates of early papyri (with the notable exception of what Nongbri has termed the ‘use and abuse’ of dates for P52). Contra the anxieties of Orsini and Clarysse, among NT scholars there is to my knowledge no salient or growing movement of ‘theological palaeography’, no phalanx attempting a push for unduly early dates. Instead, in general, the dates of NT papyri in the list included in the Nestle-Aland Greek NT, representing the ‘uneasy consensus’ generally approved by Clarysse and Orsini, are those widely accepted among NT scholars. So, as a final exhortation to colleagues in my field, it would be well for NT scholars to continue to be guided by this ‘uneasy consensus’ and treat with due caution attempts to push for earlier dates of NT papyri that do not enjoy the support of papyrologists and palaeographers.

closer to those proposed for these manuscripts by Comfort/Barrett, and the date for 0212 a bit closer to Jaroš’s, than to the dates given for these manuscripts in Nestle-Aland! 50.  P30 (175–225), P52 (125–175), P64+67+4 (175–200), P90 (150–200), P104 (100–200), 0171 (175–225), 0212 (175–225). 51.  As an example of palaeographers differing, Clarysse and Orsini (‘Early New Testament Manuscripts’, 472) agreed with the third-century date for P118 (LDAB 10081) given also in Nestle-Aland. By contrast, Cornelia Römer dated the manuscript fifth–sixth century: ‘Christliche Texte VII’, APF 50 (2004): 277 (275–83). In email exchanges Clarysse informed me that he and Orsini think that Römer’s dating is too late, but also that their third-century dating may be too early. So, they now propose a dating of 375–425 CE, and the entry in the LDAB has been modified to reflect this.

Chapter 4 G od or J es us ? T e xt u a l A m b i g ui t y a n d T extual V ar i ant s i n A c t s of t h e A post le s

One of the most interesting features of the Acts of the Apostles is what looks like an intentional correlation of references to God and Jesus in a number of expressions.1 To be sure, Acts also distinguishes God and Jesus. So, for example, some 160 times the God of biblical tradition is designated ὁ θεός, and is consistently the referent of this construction in Acts, whereas Jesus is unambiguously referred to almost 70 times by name, (ὁ) Ιησούς.2 But, to reiterate, at various points the author also seems to have intended to correlate God and Jesus, conspicuously linking them in discourse and references to religious practices. For example, in Acts we have several textually secure references to the grace of God (11:23; 13:43; 14:26; 20:24), but 15:11 refers to ‘the grace of the Lord Jesus’ (cf. also 14:3; 15:40), which seems to have created some differences among ancient readers about whose grace is referred to in 20:32, as reflected in the variants there (to which I return later in this essay). It is interesting that this sense of ambiguity in the text at certain points is not ours alone, but, as we shall see, is reflected in the manuscript tradition. That is, it seems that at a number of places in Acts we have 1.  Studies of Christology in Acts are too numerous to list them here. In another recent publication I have attempted my own characterization of some major features: L. W. Hurtado, ‘Christology in Acts’, in Issues in Luke–Acts, ed. S. A. Adams and M. W. Pahl (Piscataway, NJ: Gorgias, 2012), 217–37. C. K. Rowe argues that there is a similar intentional ambiguity, especially in the use of the term κύριος in Luke: Early Narrative Christology: The Lord in the Gospel of Luke (Berlin: de Gruyter, 2006). 2.  I use approximate numbers because at a number of places there are textual variants, and approximate numbers will serve to make my points. Anarthrous forms of θεός with the biblical deity as referent appear in 5:29, 39; 7:55. In 12:22, I take the anarthrous form in the crowd’s acclamation as ‘the voice of a god’.

4. God or Jesus?

65

textual artefacts of the efforts of ancient readers to clarify for themselves the referents in some statements, and so to disambiguate them.3 Certainly, one of the factors contributing to this referential ambiguity in a number of places in Acts is the pattern of the author’s use of ὁ κύριος and κύριος.4 In the majority of their 70 (or so) uses in Acts, the arthrous-singular forms of κύριος are applied unambiguously to Jesus, either along with his name (e.g., 1:21; 4:33; 8:16; 11:17, 20; 15:11, 26; 16:31; 19:5, 13, 17; 20:21, 24, 35; 21:13), or an arthrous form on its own but the referent clear contextually (e.g., 9:1, 11, 15, 17, 27, 28; 11:16; 13:12; 14:3; 18:8; 22:10; 23:11).5 In a few other instances, however, God is rather obviously the referent of the arthrous-singular of κύριος (e.g., 3:20; 4:26; 7:33; 13:47).6 On the other hand, the typical referent of the 3.  I have been persuaded that we should view most intentional changes to the text as more likely made by readers, not copyists (‘scribes’). See especially M. W. Holmes, ‘Codex Bezae as a Recension of the Gospels’, in Codex Bezae: Studies From the Lunel Colloquium, June 1994, ed. D. C. Parker and C.-B. Amphoux (Leiden: Brill, 1996), 123–60; and U. Schmid, ‘Scribes and Variants: Sociology and Typology’, in Textual Variation: Theological and Social Tendencies?, ed. H. A. G. Houghton and D. C. Parker (Piscataway, NJ: Gorgias, 2008), 1­–23. 4.  See, esp. J. D. G. Dunn, ‘ΚΥΡΙΟΣ in Acts’, in Jesus Christus als die Mitte der Schrift, ed. C. Landmesser et al. (Berlin: de Gruyter, 1997), 363­–78; G. Schneider, ‘Gott und Christus als ΚΥΡΙΟΣ nach der Apostelgeschichte’, in Lukas, Theologe der Heilsgeschichte: Aufsätze zum lukanischen Doppelwerk, ed. G. Schneider (Bonn: Hanstein, 1985), 213­–26; and, e.g., G. D. Kilpatrick, ‘ΚΥΡΙΟΣ Again’, in The Principles and Practice of New Testament Criticism: Collected Essays of G. D. Kilpatrick, ed. J. K. Elliott (Leuven: Leuven University Press, 1990), 216–22, for observations about anarthrous and arthrous forms of κύριος in the LXX and the Gospels. I think we see the influence of the LXX pattern of usage in Acts. Unfortunately, neither Dunn nor Schneider gives much attention to textual variants. 5.  I have listed only those places where the text is secure and where the referent is unambiguously clear. There are a number of other instances where Jesus is in my view likely the referent as well (e.g., 9:31, 35, 42; 11:21; but cf. Dunn, ‘ΚΥΡΙΟΣ in Acts’, 369–72), but to argue the cases would distract us unnecessarily from the focus of this essay. 6.  In 7:33 Bezae has καὶ ἐγένετο φωνή, and in v. 31 instead of this phrase has ὁ κύριος εἶπεν. I find unpersuasive Read-Heimerdinger’s claim that the speeches of characters such as Stephen and the apostles do not reflect the theology of the author of Acts, and so I must dissent from her further claim that the author did not use arthrous forms of κύριος with reference to God. Cf. J. Read-Heimerdinger, The Bezan Text of Acts: A Contribution of Discourse Analysis to Textual Criticism, JSNTSup 236 (London: Sheffield Academic, 2002), e.g., 280–81. Cf. also her view of the referent of ‘ὁ κύριος’ in 13:47 (ibid., 284–85).

66

Texts and Artefacts

anarthrous-singular forms of κύριος is God (e.g., 2:25, 39; 3:22; 7:31, 49; 15:17; 17:24, similarly to the frequent use of anarthrous singular forms of κύριος in the LXX as the translation-equivalent for ‫)יהוה‬, although in a few other instances in Acts the referent is equally clearly Jesus (e.g., 2:36; 10:36!).7 This leaves a goodly number of places in Acts where it is not entirely clear whether κύριος/ὁ κύριος refers to God or to Jesus (e.g., 5:9, 19; 8:22, 24, 26, 39; 11:21; 12:7, 11, 17; 13:2, 10–11). For example, in his study of the uses of κύριος in Acts, Dunn lists 33 (of 110) as ‘ambiguous’.8 As an initial example, in the episode about Simon Magus (8:14–24), Peter reprimands him for thinking he could purchase ‘the gift of God’ (τὴν δωρεὰν τοῦ θεοῦ, v. 20), and warns him that his heart is not right before God (τοῦ θεοῦ, v. 21). So, does Peter’s exhortation to petition ‘the Lord’ (τοῦ κυρίου, v. 22) refer to God or to Jesus? As noted already, Jesus is more typically the referent of the arthroussingular forms of κύριος, but the immediate context suggests that God could be intended as the recipient of the petition for forgiveness. It should 7.  I omit from this discussion uses of the vocative form, κύριε, given that this address can have a range of connotations. Note that in 10:36 the christological claim is that Jesus is ‘Lord of all’ (πάντων κύριος), which may not, thus, be a real instance of the simple anarthrous form of κύριος applied to Jesus. Also, in 2:36 and 10:36 the anarthrous forms are predicates in copula constructions, which distinguish them from the constructions in which God is the referent. Cf. the discussion of anarthrous κύριος in Bezae by Read-Heimerdinger, Bezan Text of Acts, 294–97. By my count, of the twenty or so uses of anarthrous forms of κύριος in Acts in at least ten instances the referent is clearly God. Of the remaining instances, in several (11:21; 12:7, 23; 13:11) it is not entirely clear what the ‘hand/angel of the Lord’ represents, but in each case it is at least plausible that God is intended. In any event, there is a clear general pattern of distinction between referents for the arthrous and anarthrous forms, and so I find Read-Heimerdinger’s summary of the matter insufficiently precise (Bezan Text of Acts, 293–94). 8.  Dunn, ‘ΚΥΡΙΟΣ in Acts’, 369­–72. Among these 13:2, ‘λειτουργούντων δὲ αὐτῶν τῷ κυριῷ’ is particularly intriguing. In light of the general pattern of usage of the arthrous singular forms of κύριος in Acts, Read-Heimerdinger’s confidence that Jesus is the likely referent has some basis (Bezan Text of Acts, 281). The contextual variants ‘word of the Lord/word of God’ in 12:24 and 13:5 show efforts to clarify the referent in these statements, but there seems to be no variation in the phrasing in question in 13:2. Curiously, some commentators devote space to the possible connotation of the verb but do not consider who ‘the Lord’ is: e.g., C. K. Barrett, The Acts of the Apostles, ICC (Edinburgh: T & T. Clark, 1994), 1:604–5. In the LXX, the verb λειτουργέω frequently designates priestly service to God (e.g., Deut 10:8; 17:2; 1 Sam 2:11, 18; 3:1).

4. God or Jesus?

67

not be surprising, therefore, that the manuscript evidence reflects different efforts to judge the matter. So, in v. 24, whereas most witnesses have the Magus ask Peter to petition τὸν κύριον, a number of other witnesses prefer τὸν θεόν (D 614. 1505 and others).9 It seems to me that it is most likely that the latter variant reflects an effort by some ancient readers of Acts to clarify the intended recipient of the petition here, in this case replacing τὸν κύριον with τὸν θεόν. Similarly, in v. 25, whereas many witnesses have τὸν λόγον τοῦ κυρίου, a number of others (P74 A Ψ et al.) have τὸν λόγον τοῦ θεοῦ, the latter variant also removing any ambiguity in the alternative phrasing, and so most likely a secondary variant. Granted, in ancient manuscripts the difference between the nomina sacra forms of κύριος and θεός typically involved only the initial letter of each word. But I propose that these variants most likely reflect intentional changes, not accidental ones. I think that it is considerably less likely that they arose simply through copyists confusing a theta and a kappa. So now let us examine other instances where it appears that ancient readers sought to resolve this sort of ambiguity. I give particular attention to Codex Bezae’s readings at these points. Because its text of Acts is so distinctive, it will be especially interesting to see how the textual transmission recorded in this manuscript handled this matter.10 One of 9.  There are some other interesting variants in Codex D in v. 24 that do not concern us here. See, e.g., B. M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (Stuttgart: United Bible Societies, 1994), 314. 10.  There is, of course, a copious body of scholarly publications on Codex Bezae, as reflected in the valuable reference work by J. K. Elliott, A Bibliography of Greek New Testament Manuscripts, 2nd ed., SNTSMS 109 (Cambridge: Cambridge University Press, 2000), 49­–53. In particular, note the essays in D. C. Parker and C.-B. Amphoux, eds., Codex Bezae: Studies from the Lunel Colloquium June 1994 (Leiden: Brill, 1996), including a valuable contribution by J. K. Elliott, ‘Codex Bezae and the Earliest Greek Papyri’, 161–82. Relevant to the present discussion, cf. also especially Read-Heimerdinger, Bezan Text of Acts, esp. 275–310. E. J. Epp, The Theological Tendency of Codex Bezae Cantabrigiensis in Acts, SNTSMS 3 (Cambridge: Cambridge University Press, 1966), is also important (e.g., 61­–64). There are discussions of a few variants in studies such as G. D. Kilpatrick, ‘An Eclectic Study of the Text of Acts’, in Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. J. N. Birdsall and R. W. Thomson (Freiburg: Herder, 1963), 64­–77; and J. Dupont, ‘Notes sur les Acts des Apôtres’, RB 62 (1955): 47–49. Note also now the large project by J. Rius-Camps and J. Read-Heimerdinger, The Message of Acts in Codex Bezae: A Comparison with the Alexandrian Tradition, 4 vols. (London: T&T Clark International, 2004–2009). I do not engage here the long-standing questions about the ‘Western’ Text, on which see, e.g., W. A. Strange, The Problem of the Text of Acts, SNTSMS 71 (Cambridge: Cambridge University

68

Texts and Artefacts

the questions will be whether we perceive any pattern or ‘drift’ to the preferred readings in Bezae at these points, that is, any tendency to prefer God or Jesus. The Data I list here 32 instances in Acts where textual witnesses diverge in seeking to identify the referent in a sentence/phrase, in 22 instances the witnesses simply preferring either κύριος or θεός (2:17; 6:7; 8:24, 25; 10:33; 12:11, 24; 13:5, 44, 48; 15:17, 40; 16:10, 15, 32; 17:27; 18:26; 19:20; 20:28, 32; 21:14, 20). In the remaining ten variation-units we see other efforts to make explicit the referents in certain expressions, e.g., by added identifying words (14:25; 16:6; 20:25), or other measures (2:21, 34; 5:9; 13:10; 16:7; 18:9; 18:25). In each variation-unit listed below I first record the reading preferred in the Nestle-Aland 27th edition of the Novum Testamentum Graece (hereafter NA27), followed by the main variants that seem to reflect an effort to judge whether the referent in the sentence is God or Jesus, with a few major witnesses for these variants from NA27 indicated in round brackets.11 2:17 2:21 2:34 5:9 6:7

λέγει ὁ θεός: λέγει κύριος (D E et al.) #τὸ ὄνομα κυρίου: τὸ ὄνομα τοῦ κυρίου (D) ὁ κύριος τῷ κυρίῳ: κύριος τῷ κυρίῳ (‫ *א‬B* D) #τὸ πνεῦμα κυρίου: τὸ πνεῦμα τοῦ κυρίου (D) | τὸ πνεῦμα τὸ ἅγιον (P74 et al.) ὁ λόγος τοῦ θεοῦ: ὁ λόγος τοῦ κυρίου (D E Ψ 614 et al.)

Press, 1992); B. Aland, ‘Entstehung, Charakter und Herkunft des sog. westlichen Textes untersucht an der Apostelgeschichte’, ETL 62 (1986): 5–65; and, more recently, C. Tuckett, ‘How Early Is the “Western” Text of Acts?’, in The Book of Acts as Church History / Apostelgeschichte als Kirchengeschichte, ed. T. Nicklas and M. Tilly (Berlin: de Gruyter, 2003), 69­–86. 11.  For this exercise, I have drawn upon the textual apparatus of NA27 to identify points of variation in the text and for the key witnesses for variants, and have doublechecked readings of Codex Bezae by consulting the highly regarded transcription by F. H. Scrivener, Bezae Codex Cantabrigiensis (Cambridge: Deighton, Bell & Co., 1864), and photographs available online: http://www.lib.cam.ac.uk/exhibitions/KJV/ codex.php?id=1. This produced a couple of additional instances of variation not noted in NA27, which I mark with a # in the list. In instances where NA27 prints words in square brackets, I simply record here the NA27 reading without the brackets. I do not attempt a complete list of supporting witnesses, but only major ones in most cases. I use standard sigla, including ‘M’ for the ‘Majority text’, reflected in the mass of medieval manuscripts. Variants are separated by a vertical mark ( | ).

4. God or Jesus? 8:24 πρὸς τὸν κύριον: πρὸς τὸν θεόν (D 614, 1505 et al.) 8:25 τὸν λόγον τοῦ κυρίου: τὸν λόγον τοῦ θεοῦ (P74 A Ψ 316 et al.) 10:33 ὑπὸ τοῦ κυρίου: ἀπὸ τοῦ θεοῦ (P74 D M et al.) 12:11 ἐξαπέστειλεν ὁ κύριος: ἐξαπ. κύριος (‫ א‬A D E 33 M) | ἐξαπ. ὁ θεός (36 323 453 1739 et al.) | ἐξαπ. κύριος ὁ θεός (1241) 12:24 ὁ δὲ λόγος τοῦ θεοῦ: ὁ δὲ λόγος τοῦ κυρίου (B et al.) τὸν λόγον τοῦ θεοῦ: τὸν λόγον τοῦ κυρίου (D et al.) 13:5 13:10 τὰς ὁδοὺς τοῦ κυρίου: τὰς ὁδοὺς κυρίου (P74 ‫א‬2 A C D E Ψ M) 13:44 τὸν λόγον τοῦ κυρίου: τὸν λόγον τοῦ θεοῦ (B* C E Ψ M et al.) | Παυλοῦ πολύν τε λόγον ποιησαμένου περὶ τοῦ κυρίου (D) 13:48 ἐδόξαζον τὸν λόγον τοῦ κυρίου: ἐδόξαζον τὸν λόγον τοῦ θεοῦ (B E et al.) | ἐδέξαντο τὸν λόγον τοῦ θεοῦ (D) | ἐδόξαζον τὸν θεόν καὶ ἐπίστευσαν τῷ λόγῳ τοῦ κυρίου (614 et al.) 14:25 τὸν λόγον: + τοῦ κυρίου (‫ א‬A C Ψ et al.) | + τοῦ θεοῦ (P74 E et al.) 15:17 #τὸν κύριον: τὸν θεόν (D) 15:40 τῇ χάριτι τοῦ κυρίου: τ. χ. κυρίου (D) | τ. χ. τοῦ θεοῦ (P45 C E Ψ et al.) 16:6 τὸν λόγον: + τοῦ θεοῦ (D) 16:7 τὸ πνεῦμα Ἰησοῦ: τὸ πνεῦμα κυρίου (C* et al.) | τὸ πνεῦμα (M) | τὸ πνεῦμα τὸ ἅγιον (armmss Epiphanius) 16:10 ὁ θεός: ὁ κύριος (D M et al.) 16:15 πίστιν τῷ κυρίῳ: πίστην τῷ θεῷ (D) 16:32 τὸν λόγον τοῦ κυρίου: τὸν λόγον κυρίου (D) | τὸν λόγον τοῦ θεοῦ (‫ *א‬B et al.) 17:27 ζητεῖν τὸν θεόν: μάλιστα ζητεῖν τὸ θεῖον ἒστιν (D) | ζητεῖν τὸ θεῖον (Cl) | ζητεῖν τὸν κύριον (E M) 18:9 #εῖπεν δὲ ὁ κύριος: εῖπεν δὲ κύριος (D) 18:25 τὴν ὁδὸν τοῦ κυρίου: τὸν λόγον τοῦ κυρίου (D) | τὴν ὁδ. κυρίου (P41 B et al.) 18:26 τὴν ὁδὸν τοῦ θεοῦ: τὴν ὁδὸν (D) | τὸν λόγον τ. κυρίου (323. 945. 1739. et al.) | τὴν τοῦ θεοῦ ὁδὸν (Ψ M) 19:20 κατὰ κράτος τοῦ κυρίου ὁ λόγος ηὒξανεν καὶ ἲσχυεν: κατὰ κράτος ἐνίσχυσεν καὶ ἡ πίστις τοῦ θεοῦ ηὒξανεν καὶ ἐπληθύνε (D) | κατὰ κράτος ὁ λόγος τοῦ κυρίου ηὒξανεν καὶ ἲσχυσεν (P74 ‫א‬2 M et al.) 20:25 τὴν βασιλείαν: τ. β. τοῦ Ιησοῦ (D) | τ. β. τ. θεοῦ (E M et al.) | τὸ ἐυαγγέλιον τοῦ θεοῦ (323. 1739. 1891 et al.) 20:28 τὴν ἐκκλησίαν τοῦ θεοῦ: τὴν ἐκκλησίαν τοῦ κυρίου (D P74 A C E Ψ et al.) | τὴν ἐκκλησίαν τοῦ κυρίου καὶ τοῦ θεοῦ (C3 M) 20:32 τῷ θεῷ: τῷ κυρίῳ (B 326 et al.) 21:14 #τοῦ κυρίου τὸ θέλημα: τὸ θέλημα τοῦ θεοῦ (D) | τὸ θέλημα τοῦ κυρίου (Ψ M) 21:20 ἐδόξαζον τὸν θεὸν: ἐδόξαζον τὸν κύριον (D Ψ M et al.)

69

70

Texts and Artefacts

Analysis We can begin analysis of these data by noting the number of times that Codex Bezae prefers κύριος or θεός in the 22 instances in Acts cited here where the textual witnesses diverge over these words. In eleven of these variation-points Bezae prefers κύριος (sometimes arthrous and sometimes anarthrous): 2:17; 6:7; 8:25; 12:11; 13:5, 44; 15:40; 16:10, 32; 20:28; 21:20.12 In nine other instances, however, Bezae prefers ὁ θεός: 8:24; 10:33; 12:24; 13:48; 15:17; 16:15; 19:20; 20:32; 21:14.13 This immediately suggests that there is no obvious, consistent preference, at least no programmatic effort to insert references to God or to Jesus.14 Instead, the impression one has is that, at least in many instances, the readers of Acts who left their mark in the text witnessed by Bezae essentially attempted to judge the referents on a case-by-case basis, likely attempting to decide the matter in light of the immediate context of the statements or phrases in question. The same seems to be true for other manuscripts too. The variants in 12:11 illustrate this, with some witnesses (e.g., B Ψ 614) supporting ὁ κύριος, an expression that typically designates Jesus, whereas other witnesses (‫ א‬D et al.) have the anarthrous κύριος, which more often designates God in Acts. Still other witnesses (e.g., 36 and 323) have ὁ θεός, or κύριος ὁ θεός (1241) here, both of which variants

12.  In 2:17; 12:11 and 15:40, Bezae has the anarthrous form, for which I suspect God was the intended referent, as I shall suggest in the following discussion. Bezae’s preference for κύριος in 16:10, 32 and 20:28 does not fit the claim that Bezae ‘reserves κύριος for situations involving people of Jewish origin or for reporting their words…creating a distinction between gentiles and Jews’ (cf. Read-Heimerdinger, Bezan Text of Acts, 286). 13.  In addition, note that in 16:6 Bezae (with some others) specifies that τὸν λόγον is τ. λ. τοῦ θεοῦ. Also, in 19:20 Bezae supports a distinctive reading that refers to ἡ πίστις τοῦ θεοῦ growing powerfully (κατὰ κράτος), whereas other witnesses refer to the powerful growth of τοῦ κυρίου ὁ λόγος or ὁ λόγος τοῦ κυρίου; in 17:27 Bezae refers to humankind seeking τὸ θεῖον, where other witnesses have either ‘God’ or ‘the Lord’; and in 18:26 Bezae has simply τὴν ὁδόν, whereas other witnesses have ‘the way of God’ or ‘the word of the Lord’. 14.  I must register dissent, therefore, from Read-Heimerdinger’s view that Vaticanus reflects a ‘higher’ Christology and that Bezae reflects an earlier christological stance (Bezan Text of Acts, 292–93). She fails to take adequate account of other evidence (e.g., that already in Paul’s letters Jesus is linked with God in belief and religious practice), and in my view misunderstands the textual evidence that she considers.

4. God or Jesus?

71

transparently reflect the view that God is the referent who sent the angel to rescue Peter from jail.15 Among the remaining variation-units other than those involving simply a choice between κύριος or θεός, note 14:25, where Bezae (with Vaticanus and others) supports the reading that in Perga Paul and Barnabas spoke τὸν λόγον, whereas some other witnesses have either τὸν λόγον τοῦ θεοῦ or τὸν λόγον τοῦ κυρίου. By contrast, in 16:6, where again we have the reading τὸν λόγον supported by many witnesses, Bezae has τὸν λόγον τοῦ θεοῦ. All of these variants reflect what appear to be authentic phrases used in Acts. So in each of these instances, whichever variant is judged to be ‘original’, I suggest that the others all represent efforts to clarify the text slightly by using terminology native to Acts.16 In 16:7, although a number of important witnesses (including Bezae and Vaticanus) have τὸ πνεῦμα Ιησοῦ preventing Paul from going into Bithynia (a unique occurrence of this expression in the NT), a few have τὸ πνεῦμα κυρίου or τὸ πνεῦμα τὸ ἃγιον, and others (as reflected in the ‘Majority’ text) have simply τὸ πνεῦμα. Each of these latter variants is a more familiar expression, and ‘the Spirit of the Lord’ could have been intended either as a more reverential way of referring to Jesus as ‘Lord’, or as an identification of the Spirit as God’s (the probable import of the

15.  If ancient readers perused the context to help them judge the referent in 12:11, they still had to choose what to make of the data. E.g., in 12:5, the church prays to God (πρὸς τὸν θεὸν) for Peter’s release, and so the ἂγγελος κυρίου in 12:7 might readily be taken as from God, which might have prompted a preference for this variant in 12:11. On the other hand, in 12:17 Acts securely credits ὁ κύριος with rescuing Peter, which might have led some to prefer ὁ κύριος in v. 11. Cf. Read-Heimerdinger (Bezan Text of Acts, 283), who rightly suggests that ἂγγελος κυρίου (cf. her incorrect ὁ ἂγγελος κυρίου) is somewhat ambiguous, ‘the duality perhaps being intentional’. But she seems not to consider adequately the different pattern of usage of the anarthrous and arthrous forms of κύριος in Acts (and in the LXX), and so over-confidently claims that the frequent reference to Jesus as ὁ κύριος is ‘a deciding factor’ for thinking that ‘Luke intends the angel of the Lord to be understood as that of Jesus’. Contrast the instances in Acts where the arthrous κύριος is used for Jesus, even in formulaic expressions: 8:16; 9:28; 11:17; 15:26; 19:5, 17; 21:13 (but cf. Bezae’s curious anarthrous variants in 19:5, 17!). A similar pattern is reflected in other NT writings, e.g., 1 Cor 1:2; 5:4; 6:11. 16.  Examples of what seem to be secure instances of τὸν λόγον are in 4:4; 6:4; 10:44; 11:19; 14:12; 17:11; 20:7; τὸν λόγον τοῦ θεοῦ in 4:31; 6:2; 8:14; 11:1; 13:7, 46; 17:13; 18:11; and τὸν λόγον τοῦ κυρίου in 15:35, 36; 19:10. See, e.g., the comments about which variants are likely original and which secondary in Metzger, Textual Commentary, 375, 390.

72

Texts and Artefacts

variant ‘the Holy Spirit’). Moreover, these variants might have been intended to avoid any thought that ‘the spirit of Jesus’ was some ghostly apparition. We have a very different set of variants in 17:27, but I suggest that they too reflect efforts to understand and ‘clarify’ the statement in which they appear. Was it God’s plan for all nations ‘to seek God’ or ‘to seek the Divine/Deity’ (Bezae’s reading) or ‘to seek the Lord’? As noted by Metzger, τὸ θεῖον may have been placed here under the influence of this term in v. 29, and ‘since θεός is the subject of the sentence (cf. v. 24), there was an added incentive for scribes to alter θεόν to either θεῖον or κύριον’, to avoid repetition.17 Also, however, some early Christian readers may have wanted to avoid any idea that people can seek the true God apart from God’s revelation in the gospel, and so may have preferred a reference to seeking τὸ θεῖον, ‘the divine’, perhaps connoting a less well informed religiousness. In any case, once again, I suggest that we have variants as products of serious efforts to engage the text and to render it less difficult. The import of the variants in 18:9 (not noted in NA27) is a bit more difficult to judge. It is possible that the absence of the definite article before κύριος in Bezae here is an accidental omission. Moreover, in 18:8 after καὶ ἐβαπτίζοντο, Bezae has a distinctive additional clause, πιστεύοντες τῷ θεῷ διὰ τοῦ ὀνόματος τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ, which may have served to identity who the κύριος is who appears to Paul in v. 9.18 On the other hand, this latter, distinctive reading clearly shows a ‘dyadic’ combination and distinction of two figures, God and Jesus. So, it is also possible that Bezae’s anarthrous form in v. 9 represents an effort to specify the referent as God, reflecting the dominant pattern of usage of this form of κύριος in Acts.

17.  Ibid., 405. This statement seems in conflict, however, with the initial sentence in Metzger’s comment on this variation-unit, in which he refers to ‘the careless substitution by a scribe of ΚΝ for ΘΝ…’ I think it more likely that the change was intentional. Also, as indicated earlier, I find it more plausible that any such intentional changes were made by readers taking the time to study the text and puzzle over its difficulties, rather than by copyists ‘on the fly’. 18.  Cf. Epp’s discussion of 18:8 and other instances in Acts (in readings supported both by Vaticanus and Bezae), who judged that there seems to be ‘a definite pattern in the usage of πιστεύειν τῷ θεῷ/κυριῷ (or πιστεύειν ἐπὶ/εἰς τὸν θεόν/κύριον), gentiles referred to as believing in God, and Jews as believing in the Lord (Theological Tendency, 88­–90).

4. God or Jesus?

73

This latter is plausible if we take time to consider briefly some other instances where Bezae uses the arthrous and anarthrous forms of κύριος with some apparent concern to distinguish God and Jesus. For example, whereas in 2:21 there is a certain ambiguity in the anarthrous form of κύριος in the NA27 reading, as to whether it refers to calling upon God or Jesus, the Bezae variant with the article, τοῦ κυρίου, tilts the probability strongly toward the latter.19 Note also that in 2:34 Bezae (and ‫ *א‬B*) has εἶπεν κύριος τῷ κυριῷ μου, the anarthrous form, κύριος, referring to God and the form with the article designating ‘the Lord’ Jesus.20 In 13:10, Bezae (with numerous other witnesses) has Elymas opposing τὰς ὁδοὺς κυρίου (cf. τὰς ὁδοὺς τοῦ κυρίου in ‫ *א‬B et al.), and in v. 11 Bezae has him struck by ἡ χεὶρ κυρίου.21 In light of Bezae’s distinctively explicit statement in v. 12 that the proconsul believed τῷ θεῷ and was astonished by τῇ διδαχῇ τοῦ κυρίου, I think it plausible that the anarthrous forms of κύριος in vv. 10–11 were intended to refer to God, with Jesus designated later in v. 12 by the arthrous form.22 On the other hand, I find it more difficult to judge confidently whether in 11:21 χεὶρ κυρίου refers to Jesus, who is designated with the arthrous forms in the immediate context (τὸν κύριον in vv. 20–21, and τῷ κυριῷ in v. 24), or to God, mentioned explicitly in v. 23 (τὴν χάριν τοῦ θεοῦ) and in the preceding context too (vv. 17–18). Clearly, in Bezae as in other NT manuscripts, both God and Jesus factor in religious discourse and practice.23 The variants in 5:9 suggest several efforts to identify the Spirit here, apparently with reference to God or Jesus, a couple of the variants perhaps further illustrations of the distinguishable connotations typical in the anarthrous and arthrous forms of κύριος. The reading supported by

19.  Noted also by Read-Heimerdinger, Bezan Text of Acts, 279, although I am less confident than she about her claim that the Bezae variant in 2:17 was intended to make Jesus the author of the words quoted from Joel. Acts has several references to ‘calling upon’ Jesus’ name (7:59; 9:14, 21; 22:16), confirming that the author sees this action as corresponding to the exhortation in the Joel quotation. 20.  Cf. the reading ειπεν ὁ κύριος τῷ κυριῷ μου favoured in other witnesses (e.g., P74 ‫א‬2 A B2). 21.  I do not consider significant the omission of the article, ἡ, in Vaticanus and some other witnesses. 22.  Cf. Read-Heimerdinger (Bezan Text of Acts, 284). 23.  Cf. Read-Heimerdinger (ibid., 280­–81), who confidently takes χεὶρ κυρίου as referring to Jesus. But in all comparable expressions elsewhere in Acts, the ‘hand’ is God’s: 4:28, 30; 7:25 (in addition to 11:21 and 13:11, which are under discussion here).

74

Texts and Artefacts

Vaticanus and most other witnesses, τὸ πνεῦμα κυρίου (anarthrous κυρίου), may indicate that God is the intended referent, whereas the Bezae reading, τὸ πνεῦμα τοῦ κυρίου (arthrous form), may have served to identify the Spirit with reference to Jesus.24 The variant, τὸ πνεῦμα τὸ ἃγιον (P74 et al.), certainly seems to identify the Spirit as God’s by means of a common early Christian expression. In 19:20, we have what appears to be a revised word-order in the variant reading supported by P74 and other witnesses, and a distinctive reading in Bezae that involves reference to ‘the faith of God’ growing instead of ‘the word of the Lord’, giving us the only occurrence of ἡ πίστις τοῦ θεοῦ in Luke–Acts. As others have proposed, Bezae’s reading seems to reflect some sort of conflation, perhaps through a reader noticing the similarity of this verse to 6:7 and 12:24, where ‘the word of God/the Lord’ ηὒξανεν καὶ ἐπληθύνετο.25 The variants in 20:25 show efforts to clarify whether Paul refers here to preaching ‘the kingdom’, ‘the kingdom of God’, ‘the kingdom of Jesus’, or ‘the gospel of God’. I am particularly interested in the variants that exhibit a preference for identifying the kingdom either with reference to God or, as in Bezae, with reference to Jesus. This might allow us to add this variation-unit to those eleven other places where Bezae prefers κύριος to θεός. But, even so, as indicated earlier, the data do not suggest to me a uni-directional preference, but instead an effort to puzzle out each passage in its own light. I return now to the 22 variation-units listed in which the major variants involve identifying the referent either as κύριος or θεός. In the first of these, 2:17, however, it is not entirely clear that the variants reflect different referents. Certainly, the referent of ὁ θεός (supported in most witnesses) is clear enough, but what are we to make of Bezae’s preference here for the anarthrous κύριος? Given that this form is often used in Acts as the Greek substitute for YHWH (e.g., 2:39; 3:22, as is also the dominant translationchoice in the LXX), in Bezae here it might have been simply a preferred way of introducing the Joel quotation as words by the OT deity, using

24.  Cf. comments by Read-Heimerdinger (ibid., 282), who says that Bezae’s arthrous form ‘undoes’ the ‘set phrase’ from the LXX, and ‘reactivates it so that what once referred to Yahweh can now also apply to Jesus as Lord’. It seems to me, instead, that all the variants reflect honest struggles by various ancient readers to make sense of the statement in 5:9. 25.  E.g., J. H. Ropes, The Beginnings of Christianity. Vol. 3, The Text of Acts (London: Macmillan & Co., 1926), 185; Metzger, Textual Commentary, 418–19, although neither mentions the similarity of wording in 12:24.

4. God or Jesus?

75

an expression familiar in the LXX (e.g., Amos 1:6, 11, 13; 2:1).26 On the other hand, in light of statements later in the context (2:34) where Jesus is identified as the one who has ‘poured out [ἐξέχεεν]’ the Spirit-phenomena recounted in the narrative here, and has now been made both κύριον…καὶ χριστόν (2:36), it is also possible (though I think less likely) that Bezae’s reading in 2:17 served to ascribe the prophetic words to Jesus, or perhaps it served to project the sort of ambiguity of referent that we have noted is a characteristic of numerous uses of κύριος in Luke–Acts.27 In at least many of the remaining instances, however, it seems to me likely that the variants more clearly reflect different judgments about the referent. So, for example, in several instances (6:7; 8:25; 12:24; 13:5, 44, 48; 16:32), the main variants (‘word of God’ or ‘word of the Lord’) may be taken as reflecting preferences about how to identify the message. Bezae shows a frequent, though not consistent, preference in these variation-units for ‘word of the Lord’ (6:7; 8:25; 13:5; 16:32), which I take as identifying the message with reference to Jesus, whereas, interestingly, Vaticanus frequently shows a contrasting preference for ‘word of God’ (6:7; 13:5, 44, 48; 16:32).28 Both expressions seem to be 26.  See, e.g., Ropes (The Text of Acts, 16–17) for a discussion of the variant forms of the Joel quotation in Acts witnesses. He judged λέγει ὁ θεός as ‘the undoubtedly original words’ here (17). 27.  Cf. Read-Heimerdinger (Bezan Text of Acts, 289), who claims that the Bezae reading here makes ‘the Lord known to Joel (Yahweh) to be one and the same as the Lord acting in the present times (Jesus)’. I think that she fails to consider here the potential significance of the anarthrous form of κύριος, for which in Acts the referent is more often God. At a number of other points as well, Read-Heimerdinger’s views seem to me somewhat curious: e.g., on the variants in Acts 16:10 (ibid., 291), or her claim that Bezae presents Paul as resisting the Spirit’s directions in 20:32; 19:1 and 20:3 (ibid., 288). Rius-Camps and Read-Heimerdinger, Message of Acts, 1:169, judge that Bezae’s κύριος in 2:17 is ‘potentially ambiguous’, but that 2:33 ‘will make it clear that Jesus is intended’. The alternate reading, ὀ θεός, clearly designates God, and they claim that ‘it is typical of B03 [Vaticanus] to avoid identification of Jesus with the Lord who spoke to Israel in the past’. I cannot here engage the rather sweeping characterizations of Bezae and Vaticanus offered by these scholars, so I will simply note that I think they are based on a selection of data from the fuller evidence. E.g., at numerous points, Bezae adds honorific titles for Jesus (e.g., 1:21; 2:38; 4:33; 15:11), which hardly suggests a more primitive christological stance than is reflected in Vaticanus. Also, as Epp noted (Theological Tendency, 63), Bezae has several more references to ‘the name of our Lord Jesus Christ’ (6:8; 14:10; 18:4, 8). 28.  We might also take account of Bezae’s variant in 13:44, which refers to a λόγον…περὶ τοῦ κυρίου, and the variant in 16:6, τὸν λόγον τοῦ θεοῦ. Including these, Bezae prefers ‘word of the Lord’ in five instances and ‘word of God’ in three others

76

Texts and Artefacts

authentic to Acts, each with a number of secure occurrences where there is no variation among witnesses: ‘word of God’ in 4:31; 6:2; 8:14; 11:1; 13:7, 46; 17:13; 18:11; and ‘word of the Lord’ in 15:35, 36; 19:10.29 So, to repeat my thesis, it appears that ancient readers often chose the one or the other expression, and on a case by case basis, perhaps reflecting their phrasing preferences, but also perhaps their exegetical judgments about which expression best fitted the statements in each context. That a given major witness varies in preference from one instance to another seems to me to support this proposal. For example, given that Bezae reflects a general preference for ‘word of the Lord’, the choice for ‘word of God’ at some other points is all the more interesting. Bezae’s preference for ‘the word of God’ in 12:24 (cf. ‘word of the Lord’ in Vaticanus) is heavily supported by other witnesses too, and so might simply be the prior reading, which transmitters of the text reflected in Bezae saw no reason to alter.30 It is also interesting to note, however, that in the immediately preceding narrative we have a disapproving reference to Herod being acclaimed in the cry θεοῦ φωνὴ καὶ οὐκ ἀνθρώπου (v. 22), and then being struck a mortal blow by ἂγγελος κυρίου because Herod did not give glory τῷ θεῷ (v. 23). So, most readers may have found ‘the word of God’ in v. 24 a more suitable expression to express a contrast with the blasphemous activities of Herod. Bezae’s preference for ‘the word of God’ in 13:48 (also supported by B E et al.) is also very interesting, as there are several references to the ‘word of God/the Lord’ in the immediate context (13:44, 46, 48, 49), with variants in each instance, except for 13:46, where we have an apparently secure reading, ‘the word of God’.31 Bezae’s reading in 13:44 is distinctive, but basically supports a reference to ‘the Lord’. Thereafter, Bezae has ‘the word of God’ in vv. 46 and 48, and then ‘the word of the Lord’ again in v. 49. Is this a case of a deliberate linkage of the two expressions in close succession by framing two uses of the one with two uses of the other?32 (12:24; 13:48; 16:6). Interestingly, in the variation-units where the options are ‘word of the Lord’ and ‘word of God’, the only instance where Vaticanus and Bezae agree is 13:48. 29.  The expression ‘the word of God’ appears four times in Luke (5:1; 8:11, 21; 11:28), along with instances of ‘the word’ in reference to the Christian message (1:2; and probably 8:12, 13, 15). ‘The word of the Lord’, however, does not appear in Luke. 30.  Note that this reading is adopted in NA27, reversing the decision made in previous editions. 31.  In v. 49, P45 and a few other witnesses have simply ὁ λόγος, omitting τοῦ κυρίου. 32.  Cf. Vaticanus’ preference for ‘word of God’ in 13:44, 46, and 48, and ‘word of the Lord’ in v. 49.

4. God or Jesus?

77

I noted earlier that in 16:6, Bezae (with some other witnesses) has Paul and Silas hindered by the Holy Spirit from speaking τὸν λόγον τοῦ θεοῦ in Asia, whereas most other witnesses have τὸν λόγον. I do not see an obvious reason for Bezae’s reading, and it may simply reflect a preference here for the fuller expression.33 As I noted earlier, in 8:24 Bezae (with some others) has the Magus ask Peter to pray for him πρὸς τὸν θεόν, and so this may represent a desire to clarify who ‘the Lord’ mentioned in v. 22 is, and who is the recipient of the prayer here. In the first sentence of 15:17, most witnesses have people seeking τὸν κύριον, but Bezae (and a few Latin witnesses) have τὸν θεόν (a variationunit not noted in NA27). Bezae’s reading here may reflect an effort to clarify the referent in light of the immediate context, in which we have a secure statement in v. 14 that God (ὁ θεός) chose to look favourably upon gentiles, and also the statement in v. 19 that the gentiles were turning ‘to God’ (ἐπὶ τὸν θεόν).34 The difference among witnesses over whether in 15:40 Paul and Silas were committed by believers to ‘the grace of God’ (P45 et al.) or ‘the grace of the Lord’ (D B et al.) is another instance where the variants may reflect a desire to clarify the referent. Granted, ‘the grace of God’ is the more securely well-attested expression in Acts (11:23; 13:43; 14:26; 20:24; cf. also 20:32), whereas ‘the grace of the Lord Jesus (Christ)’ appears explicitly only in 15:11 (but cf. also 14:3), so ‘the grace of God’ in 15:40 may only reflect a preference for the more familiar expression.35 But it also avoids any ambiguity over who ‘the Lord’ is in the alternate expression supported by D and others. In 16:10, the choice is whether ὁ κύριος (Bezae) or ὁ θεός called Paul’s missionary ensemble to evangelize Macedonia, both variants with ample support. One might take ὁ θεός here as less ambiguous than the alternative, and so a clarifying move. But it is also possible that ὁ κύριος seemed more appropriate to some readers in light of the preceding context in which ‘the spirit of Jesus’ prohibits entry into Bithynia (v. 7).36 33.  Cf., e.g., the same preference in E and several other witnesses in 6:4. 34.  Likewise, λέγει κύριος in v. 17 was probably taken as referring to God, in keeping with the dominant sense of the anarthrous κύριος. Cf. Read-Heimerdinger’s claim that Bezae avoids using ὁ κύριος in referring to gentiles (Bezan Text of Acts, 286). 35.  ‫ א‬A B and most other witnesses read τῆς χάριτος τοῦ κυρίου Ἰησοῦ, whereas C D Ψ and a number of other witnesses read τῆς χάριτος τοῦ κυρίου Ἰησοῦ Χριστοῦ. 36.  I do not find persuasive the claim that Bezae presents Jesus as the Macedonian man who appeared to Paul in a vision. Cf. Rius-Camps and Read-Heimerdinger, The Message of Acts in Codex Bezae, 3:255.

78

Texts and Artefacts

The story of Lydia (16:11–15) seems to have presented some ambiguities as well. In 16:14, she is referred to as reverencing τὸν θεόν, but also as having her heart opened to the gospel message by ὁ κύριος.37 So, it is interesting that in v. 15, although most witnesses have her as faithful τῷ κυρίῳ Bezae prefers τῷ θεῷ, the latter likely a clarifying variant that also aligns more with the description of her in v. 14. In the account involving the Philippian jailor (16:25–34), similarly, we have references to God and Jesus intertwined, which probably generated the textual variants in v. 32, where witnesses divide over whether Paul spoke ‘the word of the Lord’ (probably referring to Jesus) or ‘the word of God’. Note that in v. 31 Paul urges the jailor to ‘believe on the Lord Jesus’, but in v. 34 witnesses agree that he believed in ‘God’.38 This may have led ancient readers to ponder how best to characterize Paul’s message in v. 32. In Bezae’s reading here (τὸν λόγον κυρίου), the anarthrous κυρίου may reflect the LXX usage of this form for YHWH, and if so might be an alternate attempt to refer to Paul’s message as stemming from God. But this reading might also have seemed a bit ambiguous, and so could have generated the other two, as attempts to specify more clearly the referent in the phrase. In 20:28 the key and connected questions for ancient readers were whether God or Jesus ‘obtained’ (περιεποιήσατο) the church, and how to understand the reference to blood in the final words of the verse. Most scholars judge that the readings favoured in NA27 are more likely original.39 But, if so, they comprise a statement that posed for many ancient readers some ambiguity and potential misunderstanding in referring to ‘the church of God’ which he obtained διὰ τοῦ αἳματος τοῦ ἰδίου, especially if this expression is understood as ‘through his [God’s] own blood’.40 The variant reading, ἐκκλησίαν τοῦ κυρίου supported by Bezae and some other 37.  Scrivener judged that the original hand of Bezae had Lydia as σεβομένη τὸν κν, the kappa of the final word changed to a theta (Bezae Codex Cantabrigiensis, 444, col. 3 n. 481). I was unable to determine the matter from the online photos. If correct, however, I would take the change as instancing the efforts of ancient readers to clarify referents in the text. 38.  For our purposes, the stylistic difference between the NA27 reading and Bezae’s variant in v. 34 is inconsequential. 39.  E.g., Metzger, Textual Commentary, 425–27. 40.  As now widely thought, however, this expression should likely be understood as ‘through the blood of his own (son)’. See, e.g., Metzger, Textual Commentary, 426–27. Cf. also B. D. Ehrman, The Orthodox Corruption of Scripture (New York: Oxford University Press, 1993), 87–88, 264, who reads the variants in the context of ‘Patripassianist’ controversies.

4. God or Jesus?

79

witnesses removes this difficulty, ‘the Lord’ here referring to Jesus as the one who obtained the church through his own blood.41 The difference over whether Paul commended his hearers to ‘God’ or ‘the Lord’ in 20:32 was likely prompted by the following reference to ‘the word of his grace’. On the one hand, there are other references to the grace of God in Acts (11:23; 13:43; 14:26; 20:24; cf. ‘the grace of the Lord Jesus’ in 15:11). On the other hand, in 14:3, ‘the word of his grace’ clearly refers to ‘the Lord (Jesus)’. Also, in 20:35 there is a secure reference to ‘the words of the Lord Jesus’, which may have further prompted some readers to prefer ‘the Lord’ in v. 32. In 21:14 (a variation-unit not noted in NA27), most witnesses refer to ‘the will of the Lord’ (with a variation in word-order among witnesses), but Bezae has τὸ θέλημα τοῦ θεοῦ. ‘The will of the Lord’ is an unusual expression, whereas references to the will of God abound in the NT (e.g., Acts 22:14; Rom 1:10; 2:18; 12:2; 15:32), and this may have been one factor in a preference for the latter phrasing in Bezae.42 It might also have served to remove any ambiguity as to the referent. Finally, in 21:20 we have variants as to whether, upon hearing Paul recount his ministry among the gentiles, the Jerusalem church glorified ‘God’ or ‘the Lord’. The latter reading probably refers to Jesus, and is supported by Bezae and numerous other witnesses including the mass of medieval manuscripts. But it is an unusual expression in Luke–Acts, the closest we have to an analogy in the reference in 13:48, ἐδόξαζον τὸν λόγον τοῦ κυρίου (with variants discussed earlier).43 Glorifying ‘God’, however, is reflected numerous times in Luke–Acts (Luke 2:20; 5:25–26; 7:16; 13:13; 17:15; 18:43; 23:47; Acts 4:21; 11:18). So τὸν κύριον could have been replaced by some readers with this more familiar expression, perhaps also echoing the somewhat similar scene in 11:18. Or (if the judgment of NA27 is followed) was τὸν κύριον preferred here by many readers, perhaps to ‘Christianize’ the statement? 41.  Here again, the arthrous form is important to note, designating, as it usually does in Acts, Jesus. Rius-Camps and Read-Heimerdinger (Message of Acts in Codex Bezae, 4:119) err in stating that ‘τῆν ἐκκληςίαν τοῦ κυρίου is found in the LXX’, citing Deut 23:2–4, 9; 1 Chr 28:8; Mic 2:5. In these and other LXX instances, the anarthrous κύριος appears, and the arthrous form in the Bezae reading reflects the early Christian adaptation of the term as a christological title. Bezae also has περιεποιήσατο ἑαυτῷ, the latter word further emphasizing that ‘the Lord’ who acquired the church is Jesus. The variant supported by M is an obvious conflation of the other two. 42.  In Eph 5:17 many witnesses have τὸ θέλημα τοῦ κυρίου (adopted in NA27), others have τὸ θέλημα τοῦ θεοῦ, and P46 has τὸ θέλημα τοῦ χριστοῦ. 43.  In 13:48 Bezae says the people ‘received the word of God’.

80

Texts and Artefacts

Conclusion In the hope that the foregoing discussion has adequately given a plausible basis for my main contentions, I shall conclude by simply re-stating them. At a number of points in the text of Acts, ancient readers were presented with ambiguities, especially whether the referent was God or Jesus, and the variants at these points reflect readers’ efforts to judge the matter. It appears that they may often have done this by reference to the immediate context, and so the textual variants are artefacts of their exegetical efforts. We do not see a programmatic effort to insert Jesus or God, but instead it seems that readers engaged each of these variation-units on a case-by-case basis, simply seeking to grasp what they thought the text meant. In some cases (e.g., 20:28), doctrinal issues in the early church may have been a factor disposing readers to one variant or another, but these appear to be few.

Part II M a n u s cr i p t s a s A rt e fact s

Chapter 5 T he ‘ M e ta -D ata ’ of E a rli e st C hr i st i an M a n u s cr i p ts *

The earliest extant physical artefacts of Christianity are manuscripts, and scholars concerned with the origins of Christianity should feel more obliged to familiarize themselves with these artefacts as a matter of some priority. This obligation is not, however, sufficiently recognized in the field, largely because many scholars do not realize what these items have to offer. So, the aim in this short discussion is to illustrate what sorts of data early Christian manuscripts present to students of Christian origins. Here I offer some selective illustrations regarding features of early Christian manuscripts that may be of significance for larger historical issues, including Jewish and Christian relations in the first three centuries. I have presented an extended discussion of these matters in the form of a modest-sized book.1

*  An earlier draft of this essay was given as an invited presentation in November 2004 to a special joint session in the Annual Meeting of the Society of Biblical Literature, sponsored by two programme units: the New Testament Textual Criticism Section and the Early Jewish and Christian Relations Section. I offer this discussion here in honour of Stephen Wilson, a long-time friend and colleague. 1.  The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). In that book and in this essay, I develop further an emphasis expressed in a previous publication: ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo, ON: Wilfrid Laurier University Press, 2000), 271–88. Among other relevant publications are two particular useful volumes: Harry Y. Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven: Yale University Press, 1995); Kim Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature (Oxford: Oxford University Press, 2000).

84

Texts and Artefacts

It will help if I clarify a few introductory matters before we look at actual cases. First, by ‘meta-data’ I refer to the data given in manuscripts beyond the readings of the texts that they convey. More commonly, insofar as early Christian manuscripts are noted at all, they tend to be seen almost solely as early copies of texts. So, for example, textual critics take account of them in tracing textual variation and the textual history of these texts. But in this essay, I point to other phenomena, particularly the physical and visual features of manuscripts, such as how they are laid out and the various scribal practices involved. Second, by ‘earliest Christian manuscripts’ I mean those manuscripts, whatever the text they contain, that can be judged with some confidence to have been produced for Christian usage, and probably by Christian scribes. NT scholars, for example, have tended to focus heavily on our earliest manuscripts of NT texts, and mainly with a concern for the readings that they attest, as is proper if the main aims are tracing the history of these texts and establishing a modern critical edition of them. But for other, wider questions about Christian origins, it is important to take account of all texts of Christian provenance. I shall support this by giving some illustrations. There are, however, disagreements among the experts over how confidently we can ascribe a Christian provenance to some manuscripts. Of course, in the case of copies of unquestionably Christian texts, such as those that came to make up our NT or other Christian writings, it is most likely that we are dealing with manuscripts produced for, and probably by, Christians. But how do we tell if particular copies of other texts that were not originally composed by Christians, such as writings now in the Hebrew Bible, are to be regarded as Christian artefacts? In general, three factors are widely (but not universally or without quibble) accepted as bases for deciding: use of a codex format, use of nomina sacra, and if a given manuscript was found among others that are more obviously Christian productions. I shall have more to say about the first two of these matters later. Third, by ‘earliest’ Christian manuscripts, I mean here items that are dated to the second or early third centuries CE. As to precise dating, here too there are disagreements among the experts. A manuscript dated to the early or middle second century by one palaeographer may be dated to the late second or even early third century by another expert. Unlike ancient manuscripts containing some ‘documentary’ texts, such as marriage documents or some letters, all our early Christian copies of literary texts are undated. So, the dating of all our putatively earliest Christian manuscripts is entirely a palaeographical judgment.

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

85

To be able to offer a judgment worthy of the attention of others requires a considerable amount of time spent in developing a wide and detailed familiarity with many, many specimens of Greek writing across at least a few centuries. In short, it is a demanding specialization of its own, upon which the rest of us have to depend, and for which we ought to be particularly grateful. I can follow intelligently the observations of palaeog­raphers, and sometimes have my own, somewhat hesitant and cautious views of this or that case, but in the main I shall depend here upon the judgments of experts in Greek palaeography. Where the experts differ, I shall note that and try to qualify my use of the evidence accordingly. With these introductory comments sufficing, I now turn to specifics. The Codex Format The first matter to note is the Christian preference for the codex over the roll, a phenomenon evident already in our earliest identifiably Christian manuscripts.2 This preference is all the more striking in comparison to the wider general preference for the roll-format in the second and third centuries CE, particularly for ‘literary’ texts, that is, writings of literary, philosophical, or religious significance. Outside of Christian circles, this wider preference for the roll only began to shift to a preponderance of codex manuscripts in the fourth century CE and later.3 We may use some data helpfully compiled in the Leuven Database of Ancient Books (LDAB) to illustrate this.4 Taking into account the catalogued ‘literary’ manuscripts dated from the third century BCE through the eighth century CE, identifiably Christian rolls amount to 2.7% of the total number of rolls (3,033), whereas Christian codices amount to 73% of all the total number of codices (3,188). Codices (of all provenances) amount to about 5% of second-century manuscripts and 2.  The key study on the topic is Colin H. Roberts and T. C. Skeat, The Birth of the Codex (London: Oxford University Press, 1983). In the following paragraphs, I draw upon the fuller discussion in The Earliest Christian Artifacts, Chapter 2. 3.  Roberts and Skeat, The Birth of the Codex, 37, provide figures of rolls and codices to show ‘it is clear that the codex scarcely counted for Greek literature before about A.D. 200. It was not until about A.D. 300 that it achieved parity with the roll.’ 4.  This valuable catalogue is accessible online: http://ldab.arts.kuleuven.ac.be/. In the following discussion, I use figures and percentages at the time of my consultation of the LDAB. As manuscripts are continually being published and catalogued, precise figures and percentages will change across time, of course, but the figures given here suffice to give the general picture, and this general picture is not changed by such further accretions of catalogued items.

86

Texts and Artefacts

about 15% of third-century manuscripts. But when we turn to manuscripts of Christian provenance, the codex is clearly the favourite book-form. For example, in the Leuven database overall, at least 91.6% of copies of NT writings are identified as codex form, and only 1.1% are rolls (and of the latter, it seems likely that all, or nearly all, are actually opisthographs, re-used rolls, the copies likely prepared for personal study).5 Among all second- and third-century NT manuscripts (our earliest), the percentage of codices is at least as high. By contrast, of all manuscripts of Homer (third century BCE through seventh century CE), 62.8% are rolls, and only 18.5% identified as codices. For manuscripts of Euripides (third century BCE through eighth century CE), 65.9% are rolls and 17.9% are codices.6 If we were to confine our attention to copies of these texts dated no later than the third century CE, the preponderance of rolls over codices would be even greater. All the data support the commonly accepted conclusion held among scholars acquainted with ancient book-production that the roll was overwhelmingly the preferred format for any text considered of literary, philosophical, or religious significance, the codex generally reserved for ‘documentary’ texts (e.g., account-books, notebooks). It is, therefore, all the more noteworthy that early Christian circles particularly preferred the codex for those writings that they regarded the most highly and treated as ‘scripture’. Though all pre-Christian (i.e., unquestionably Jewish) copies of ‘Old Testament’ writings are rolls, the copies that are of uncontested Christian provenance are all (or nearly all) codices.7 So, this suggests that the equally strong Christian preference of the codex for writings that became part of the NT certainly does not 5.  About 5.6% of MSS of NT writings are classified in the Leuven database as ‘sheet’ (i.e., a single page, the editor unable or unwilling to classify as from a roll or codex), and another 1.8% as ‘fragment’ (which usually means pieces of a MS so small that it is very difficult to say what form the parent MS may have been). The Leuven database simply records the expressed judgments of editors of MSS, and the introduction to the database indicates that some editors seem somewhat reluctant to identify single sheets as to the original MS form. This means that the percentages for codex and roll forms (particularly the former) are modest calculations. Moreover, the database does not distinguish between texts copied on the ‘recto’ side of rolls and ‘opisthographs’ (re-used rolls). 6.  The codices of these authors tend to come from the later centuries covered, when, for whatever reason, the codex was gaining ground in general book production. If we restrict our coverage to copies dated no later than the third century CE, there are very few if any codex copies. 7.  The Leuven database classifies 71.4% (434) of catalogued copies of OT writings dated from the second century BCE through the eighth century CE) as

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

87

indicate a lack of esteem for these texts. Instead, this preference, evident particularly in the case of writings that came to make up the Christian Bible, may well reflect their emerging scriptural status, with a strong preference for the codex format for these writings that matches the equally strong Christian preference for the codex for OT writings, which unquestionably functioned as scriptures for at least the main body of Christians of the time. This Christian preference for the codex was exercised particularly in copying their scriptural texts, but was by no means restricted to such texts. That is, the codex format does not by itself indicate that the text copied was treated by its user(s) as scripture. But it does seem that the codex was strongly preferred by early Christians for their scriptures. So, copies of Christian texts (scripture or extra-canonical) in a roll format are noteworthy (and the later the manuscript, the more noteworthy), for it probably means that the text put on a roll (or at least that particular copy) was not regarded (at least by the copyist who produced the manuscripts, or the party for whom the manuscript was copied) as having scriptural significance. Early Christians certainly did use the roll format for some texts, such as religious treatises, and some liturgical and magical texts, and this provides very interesting data in support of my proposals. In their indispensable study of the origins of the codex, Roberts and Skeat cited 118 Christian copies of writings other than OT texts and those that became the NT, 83 of which were codices and 35 rolls (three of these opisthographs).8 For instance, the sole two manuscripts of Irenaeus from the second to fourth centuries are rolls. Of the earliest catalogued manuscripts of Clement of Alexandria, one is a roll, one a codex, and one a fragment of unidentified book-form. Of the LDAB catalogued manuscripts for Shepherd of Hermas, 22 are codices (mainly third to sixth centuries CE) and four (among the earliest copies) are rolls. Of the three Oxyrhynchus fragmentary copies of Gospel of Thomas (early/mid-third century CE), one is a codex and two are rolls (one a re-used roll or opisthograph).9 By contrast, it is notable that we do not have a single instance of a NT writing copied on an unused roll (as distinguished from a re-used roll), and there are, at most, perhaps codices, 5.1% (31) as rolls, 21.5% (131) as ‘sheet’, and 2% (12) as ‘fragment’. The overwhelming mass of catalogued Greek OT manuscripts are of Christian provenance and from fourth century CE and later. 8.  Roberts and Skeat, The Birth of the Codex, 43. 9.  The technical term is ‘opisthograph’, which designates a roll originally used for one text, written on the inner surface, and then turned over and re-used for another text, written on the outer surface.

88

Texts and Artefacts

two to five instances of OT writings on rolls that may be of Christian provenance. By my own counting of Christian manuscripts dated earlier than 300 CE, about 35% of Christian extra-canonical literary texts are on unused rolls. This contrasts markedly with the figures for Christian copies of OT writings and writings that came to form the NT. Of these, about 95% are codices.10 Christians certainly did not invent the codex, and, to be sure, we even have a few examples of the codex used for ‘pagan’ literary works in the same period (but these ‘pagan’ works are parchment codices, often of small size, whereas Christians appear to have preferred papyrus codices). Yet Christians do seem to have been particularly active in experimenting with, and developing, this book-form. Among earliest Christian codices, we have examples of the single-gathering (or single-quire) book (all the sheets arranged in a single stack and then all folded in half, e.g., the Chester Beatty Pauline codex, P46, originally comprising 52 papyrus sheets, and the Bodmer Gospels codex P75, originally comprising 36 folded sheets), and multiple-gathering constructions with quires of various numbers of sheets (e.g., P45, made up entirely of folded single-sheet quires, or P66, made up of quires of varying numbers of sheets). This suggests to me that in this period (late second and early third centuries CE) Christians were themselves pioneering in the more serious use of the codex book-form. Their experimentation with these various modes of codex construction would not have been necessary had the codex already been well developed for book copying. It is easier, however, to demonstrate that early Christians preferred the codex-format than it is to provide a convincing explanation for how and why they came to do so. Proposals tend to fall into one of two types, which we may label as ‘pragmatic’ or ‘semiotic’. I am not persuaded, however, that any of the several proposals about the supposed practical advantage of the codex is successful in accounting for the wholesale Christian preference for this format.11 ‘Pragmatic’ proposals include suggestions that Christians may have been attracted to the codex because it allowed use of both sides of the writing material, thereby saving on the cost of copies. But careful attempts to calculate costs of copying texts suggest that any actual savings that might have been gained by use of codex format were not significant.12 Furthermore, to anticipate other data 10.  For a fuller presentation of the evidence on the matter, see Hurtado, The Earliest Christian Artifacts, esp. 53–61. 11.  For further discussion, see now ibid., 63–82. 12.  T. C. Skeat, ‘The Length of the Standard Papyrus Roll and the Cost-advantage of the Codex’, in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott,

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

89

discussed later in this essay, the rather wide margins and line-spacing in many earliest Christian manuscripts suggest that the copyists were not particularly concerned to save the amount of papyrus used. The Australian scholar G. H. R. Horsley has proposed that Christians preferred the codex because they were from lower-educated circles more accustomed to dealing with documentary texts than literary ones, the codex thus seeming to them a more familiar book-form.13 But this seems to me unconvincing. To cite one reason, I fear that it presupposes a somewhat over-simplified view of the socio-economic level of early Christianity in the first two centuries CE. What will seem initially a more plausible suggestion is that the codex was favoured because it may have been more easily transportable (perhaps carried in a pouch across one’s shoulder), something perhaps attractive to a religious movement that obviously devoted a lot of effort and resources to networking trans-locally among various Christian circles. But was Christianity the only movement in which trans-local use of texts was a feature? Other suggestions of an ‘obvious’ superiority of the codex have been offered. For instance, we are so accustomed to the bound book that many will imagine the codex to be obviously superior for such things as flipping back and forth between references or marking/finding one’s place in a text. But we should be a bit suspicious about what seems so ‘obvious’ today. For one thing, it appears that writers tended to make excerpts of texts for use in their own compositions, and did not customarily need to thumb to particular texts from their full copies. Moreover, once again, were Christians the only ones who used particular texts as authoritative, proof-texting their claims from them? Surely, Jewish use of scripture was comparable, and yet it is clear that Jews overwhelmingly continued in this period to prefer the roll-format, particularly for their scriptures. In any case, were Christians somehow uniquely able to perceive such advantages of the codex that seem so obvious to some moderns but somehow escaped others in the second and third centuries CE? Anything is possible, but I confess that this seems to me counter-intuitive.

NovTSup 113 (Leiden: Brill, 2004), 65–70; the original essay appeared in ZPE 45 (1982): 169–76. Skeat judged perhaps a 26% savings on the cost of the papyrus, but when we note the additional time, labour and skills involved in constructing a codex, the putative cost-advantage is narrowed further considerably. 13.  G. H. R. Horsley, ‘Classical Manuscripts in Australia and New Zealand, and the Early History of the Codex’, Antichthon 27 (1995): 60–85, esp. 76–83.

90

Texts and Artefacts

What I have called the ‘semiotic’ proposals include the suggestion that the Christian preference for the codex represented a move to identify and distinguish Christian copies of texts. That is, the preference for the codex may exhibit part of what we may think of as an emergent Christian ‘material culture’ in the second century CE.14 I recognize that this goes against some current views that it is inappropriate to distinguish ‘Christianity’ in that early period, but those who have argued such views seem to me not to have taken account of the evidence to which I point here.15 Certainly, the preference for the codex seems to be a general convention among Christians already in the second century, one of a few phenomena (others of which I discuss shortly) pointing toward emergent Christian conventionalizations at a time when we generally assume nothing but diversity. My purpose here, however, is not to engage fully this larger historical issue, but merely to point out that the sort of data that I highlight here need to be considered in dealing with the larger question of when and how ‘Christianity’ may have emerged as an identifiable entity, and how it may have expressed itself (verbally, visually, and physically) in the earliest period. Nomina Sacra A second feature of earliest Christian manuscripts that is well known among papyrologists and palaeographers, but insufficiently explored by scholars in Christian origins, is the curious scribal practice referred to as the ‘nomina sacra’.16 Essentially, a number of key words in early Christian religious discourse are characteristically written in special abbreviated forms (commonly, first and last letters, in some cases with a medial letter too) with a distinctive supralinear horizontal stroke placed 14.  For the proposal that we can identify an emergent Christian material and visual culture in the second century, I refer readers to my earlier essay, ‘The Earliest Evidence’. 15.  Cf., e.g., Judith M. Lieu, Neither Jew Nor Greek? Constructing Early Christianity (London: T&T Clark International, 2002), 171, whose judgment that for the first two centuries CE ‘material remains are not available as markers of Christian Identify, or/and, if available, they would not be or perhaps are not distinguishable’ appears to neglect the artifactual import of earliest Christian manuscripts. 16.  The key discussion of the matter is in Colin H. Roberts, Manuscript, Society and Belief in Christian Egypt (London: Oxford University Press, 1979), 26–48. I have attempted to contribute further in an article published several years ago: Larry W. Hurtado, ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998): 655–73 (republished as Chapter 7 of the present volume). Also see now my discussion in The Earliest Christian Artifacts, Chapter 3.

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

91

over the abbreviated form. The words most consistently treated in this manner in the earliest extant evidence are the four terms Θεός, Κύριος, Ἰησοῦς, and Χριστός. That is, the most characteristic examples, and probably the words that first were given this scribal treatment, are key designations of God and Jesus, which Schuyler Brown termed ‘nomina divina’.17 But already in manuscripts that are dated to the late second or early third century CE (e.g., the Egerton ‘Unknown Gospel’, P.Egerton 2), we see other words treated as nomina sacra (e.g., ‘Son’, ‘Father’, ‘Spirit’, and also ‘David’, ‘Moses’, ‘heaven’, and ‘Jerusalem’). So, clearly, from whatever beginning point, the practice spread to include additional terms. There are controversies over the origin of this scribal practice and also over its function or significance. Roberts concluded that the particular scribal device in question was a novel Christian development, though likely motivated by religious concerns similar to those reflected in the treatment of the Tetragrammaton in pre-Christian Jewish manuscripts of biblical writings (a view to which I have lent my support).18 But Kurt Treu and Robert Kraft have contended that Christians appropriated a scribal practice that had already developed in Jewish circles, and that the only significant Christian innovation was in extending the practice to certain terms directly expressive of early Christian faith, especially, of course, Ἰησοῦς.19 Admittedly, the extant evidence is frustratingly spotty, and so educated guesses are unavoidably all that we can offer. Also, it is not feasible here to set out an adequate argument on the issue, which I have attempted in another publication.20 I content myself here with emphasizing that the issue is not merely a palaeographical one. If the scribal practice of nomina sacra was adopted from prior Jewish scribal practice, then this further indicates a remarkably direct indebtedness of earliest Christian circles to their Jewish matrix, and also shows another feature of shared ‘material culture’ between ‘Judaism’ and ‘Christianity’ in the second century. If, on the other hand, nomina sacra represent in some real sense a distinctively Christian innovation, then, as appears to be the case with the preference 17.  Schuyler Brown, ‘Concerning the Origin of the Nomina Sacra’, Studia Papyrologica 9 (1970): 7–19. 18.  Esp. Roberts, Manuscript, 26–48. 19.  Kurt Treu, ‘Die Bedeutung des Griechischen für die Juden im römischen Reich’, Kairos NF 15, no. 1/2 (1973): 123–44; Robert A. 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 A. O’Sullivan (London: British Library, 2003), 51–72. 20.  Hurtado, The Earliest Christian Artifacts, esp. 113–30.

92

Texts and Artefacts

for the codex, this scribal practice may be another emergent convention by which Christians marked their manuscripts. And this would be a notable finding relevant to larger questions about how early ‘Christianity’ began to emerge as an identifiable social phenomenon. That is, there are larger issues for which the specific question of the origin of nomina sacra is very relevant. It will be important, however, not to allow preferred dispositions (or declared positions) on these larger issues (e.g., whether and when we can speak of ‘Judaism’ and ‘Christianity’ in the early second century) to skew our leanings on what must finally be a matter to be judged on the basis of the available manuscript evidence. My plea is that the sort of evidence that I point to here should be taken into account. The other issue about nomina sacra currently (and more recently) debated is what the scribal practice really signifies. Probably, most scholars who have considered the phenomenon have judged that nomina sacra reflect early Christian piety (whatever the historical relationship of this scribal practices to Jewish scribal practice), the words in question given this special treatment to set them off from the surrounding text out of reverence for what the words represent or designate. On this view, nomina sacra are prime evidence of earliest Christian faith and religious devotion. Indeed, I have emphasized that the four earliest and most consistently treated words (‘God’, ‘Lord’, ‘Jesus’, and ‘Christ’) vividly reflect the ‘binitarian’ shape of earliest Christian piety, particularly as the key words for Jesus are given the same scribal treatment as key words for God.21 Moreover, it is important to note that the scribal phenomenon in question was purely a visual one. Although it seems likely that ancient Jewish scribal practices with reference to the Tetragrammaton were intended to signal readers to pronounce a reverential substitute-word (typically, ‫ ֲאד ֹנָ י‬in Hebrew, or Κύριος in Greek), there is no indication that the use of nomina sacra in Christian manuscripts functioned to signal any equivalent action by readers. So far as we can tell, lectors of these early Christian manuscripts pronounced fully and explicitly the words written as nomina sacra, including the key words for God and Jesus. This is why, therefore, I emphasize that the nomina sacra registered solely as visual phenomena and could be experienced solely by those who read (or otherwise viewed?) the manuscripts in which they were written.

21.  Hurtado, ‘The Origin of the Nomina Sacra’; idem, Lord Jesus Christ: Devotion to Jesus in Earliest Christianity (Grand Rapids: Eerdmans, 2003), 625–27.

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

93

Accordingly, echoing an observation by Erich Dinkler several decades ago, I emphasize that nomina sacra are to be taken as textual/scribal phenomena with an ‘iconographic’ significance.22 As such, they form perhaps our earliest extant expressions of an emergent Christian ‘visual culture’, and should be taken account of more in efforts to chart the emergence of early Christian efforts to express their faith visually. Indeed, perhaps the nomina sacra should be included in our efforts to explore the origins of Christian art and iconography. Furthermore, given that the practice in question may have originated as early as the late first century, and certainly not very long thereafter, it is all the more important to take account of it in our larger analysis of Christian origins. However, Christopher Tuckett has contended recently that the scribal device functioned simply as abbreviation of certain frequently used words in early Christian vocabulary, and that nomina sacra are not particularly expressive of early Christian devotion.23 In my recent book, I have engaged Tuckett’s view at some length, and there is neither space nor necessity to do so here.24 Suffice it to say that I find his argument flawed and his conclusion highly implausible. For instance, Tuckett’s proposal that the nomina sacra may have been intended simply as readers’ aids is readily refuted. We know very well the various devices that ancient copyists used to aid readers, and such abbreviations are not among them. But my main point here is that the issue of what to make of the nomina sacra is sufficiently important to deserve the attention of all scholars concerned with the origins of Christianity. For, whether nomina sacra represent an interesting expression of early Christian faith or merely a Christian scribal convenience-device, in either case, if they were a Christian innovation, they comprise a noteworthy phenomenon. The ‘Staurogram’ There is another curious, and even more widely overlooked, item that makes its first appearance in a few early Christian manuscripts: the monogram-like combination of the letters tau and rho that sometimes 22.  Erich Dinkler, ‘Älteste christliche Denkmäler: Bestand und Chronologie’, in Art, Archaeology and Architecture of Early Christianity, ed. Paul Corby Finney (New York: Garland, 1993), 22–66 (originally published in Dinker’s essay-collection, Signum Crucis [Tübingen: Mohr, 1967], 134–78). 23.  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), 431–58. 24.  Hurtado, The Earliest Christian Artifacts, esp. 122–32.

94

Texts and Artefacts

forms part of an abbreviated nomina sacra treatment of the Greek words σταυροω and σταυρος. It is one of several monogram devices referring to Jesus, including the more well known chi-rho, the iota-chi, and the iota-eta.25 Although they all were adapted from pre/non-Christian usage, in Christian usage these scribal devices reflect Christological convictions. The chi-rho is, of course, the first two letters of Χριστός, the iota-chi the initial letters of Ἰησοῦς Χριστός, and the iota-eta the initial two letters of Ἰησοῦς. But it is important to note that the tau-rho makes no such reference to Christological terms or titles. As already indicated, the earliest Christian use, in NT manuscripts dated variously from the late second to early or mid-third centuries (P75, P66, and P45), is entirely in direct or indirect references to Jesus’ cross or crucifixion. If the other monogramdevices can be termed ‘christograms’ (references broadly to Jesus and his significance), the tau-rho is more correctly designated the early ‘staurogram’.26 Moreover, the tau-rho device seems more truly and purely a visual phenomenon, perhaps intended, as several scholars have suggested, as an early pictographic representation of the crucified Jesus. From other early texts, particularly the Epistle of Barnabas (12:8), we know that the Greek letter tau was used by Christians as a visual reference to Jesus’ cross. The appropriation of the tau-rho, involving the superimposition of the rho upon a tau, thus, may have been intended (in Christian usage) as a simple, stylized depiction of the head of a figure on a cross. If this is correct, then it means that historians of early Christian art must revise commonly held notions of when we can date earliest visual references to the crucified Jesus. This usage of the tau-rho is some two hundred years or earlier than what is otherwise usually regarded as the earliest extant Christian depiction. I am pleased to note that in her valuable recent survey of early Christian art, Robin Margaret Jensen acknowledges this.27 25.  For a brief discussion with illustrations, see Jack Finegan, The Archeology of the New Testament: The Life of Jesus and the Beginning of the Early Church, rev. ed. (Princeton: Princeton University Press, 1992), 352–55. I offer a fuller discussion in The Earliest Christian Artifacts, Chapter 4. 26.  Matthew Black, ‘The Chi-Rho Sign – Christogram and/or Staurogram?’, in Apostolic History and the Gospel: Essays Presented to F. F. Bruce, ed. W. Ward Gasque and Ralph P. Martin (Grand Rapids: Eerdmans, 1970), 319–27; Kurt Aland, ‘Bemerkungen zum Alter und zur Entstehung des Christogramms anhand von Beobachtungen bei P66 und P75’, in Studien zur Überlieferung des Neuen Testaments und seines Textes (Berlin: de Gruyter, 1967), 173–79. 27.  Robin Margaret Jensen, Understanding Early Christian Art (London: Routledge, 2000), esp. 138. See also my essay-length study of the ‘staurogram’: Larry

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

95

Other Features of Christian Manuscripts Other features of the earliest Christian manuscripts are worth mentioning here, all of them potentially relevant for wider historical issues. Most of the scribal ‘hands’ in earliest Christian manuscripts, including copies of the NT writings, represent what is often called a ‘semi-documentary’ type, noticeably less elegant than the more calligraphic hands typical of formal copies of Greek literary texts of the time. Also, the Christian scribal hands seem less elegant than those that are often found in early Jewish copies of biblical texts in Greek. Any suggestion that the less sophisticated hands in Christian copies of NT writings reflects a lack of high esteem for these texts, however, is undermined by the fact that Christian copies of OT writings (whose scriptural status for most early Christians is clear) display equivalent scribal hands. Moreover, though the scribal hands in earliest Christian manuscripts are often noticeably less elegant than high literary book-products of the time, it is usually clear that the scribes in question were aiming to produce as good a product as they could, and that they seem often to have been scribes well-practiced in copying, if not in formal book-trade calligraphy. In short, the lack of elegance probably does not connote a lack of esteem for the texts copied. So, instead, might these Christian ‘semi-documentary’ scribal hands reflect the socio-economic levels of earliest Christianity? That is, were there simply very few Christian scribes with the specific calligraphic training to produce the more elegant writing preferred in the formal book trade of the time? This may be worth considering. But we also have to be careful about over-generalizing. For, already in the Christian manuscript fragments of the second and third centuries CE we also have a few examples of somewhat more sophisticated scribal hands.28 That is, although they may have been a minority among copyists of Christian texts, by the late second century (and perhaps earlier) there were some copyists with a certain calligraphic ability (and/or some Christians financially able to pay for copyists with calligraphic ability).

W. Hurtado, ‘The Staurogram in Early Christian Manuscripts: The Earliest Visual Reference to the Crucified Jesus?’, in New Testament Manuscripts: Their Texts and Their World, ed. Thomas J. Kraus and Tobias Nicklas (Leiden: Brill, 2006), 207–26 (republished as Chapter 8 of the present volume). 28.  E.g., P.Oxy. 4403 (P103, a single sheet of a copy of Matthew dated to the late second or early third century CE), the scribal hand of which is described as ‘quite elegant’ by the editors. E. W. Handley, U. Wartenberg, R. A. Coles, et al., The Oxyrhynchus Papyri, Volume LXIV (London: Egypt Exploration Society, 1997), 6.

96

Texts and Artefacts

To turn now to other data, it is worth noting that Christian manuscripts often have very wide margins, generous line spacing, and also writing that, though often less than elegant, is clear, carefully formed, and with generous-size letters. As noted by the master of Greek papyrology, Eric Turner, Christian manuscripts (especially copies of biblical texts) often have fewer lines per page than typical of Greek literary works with equivalent column height.29 To repeat a point made briefly earlier in this essay, it is rather clear that copyists of earliest Christian texts were not usually concerned about saving writing material. Instead, these features probably reflect a concern to produce copies that were easy to read. More specifically, it seems likely that we have early manuscripts prepared for public or liturgical reading, the manuscripts serving as actual artefacts of this use of the texts in question. This conclusion is strengthened by other phenomena that appear to represent an emerging constellation of devices that may be thought of as readers’ aids intended to facilitate the reading of the texts in question. Already in our earliest fragmentary materials (e.g., P52, the Rylands fragment of John dated sometime in the second century), we have emergent use of the dieresis over an initial iota of a word following a word that ends in another vowel. As well, from about the same period we begin to see use of some punctuation and occasional breathing marks. Another striking feature is the use of enlarged spaces within lines, violating the more strict scripta continuo characteristic of Greek literary copies, and apparently intended to mark sense-units that may be early paragraphdivisions (e.g. the second-century Yale Genesis fragment, P.Yale 1). Likewise, page-numbering appears as early as the Chester Beatty codex of Numbers–Deuteronomy (late second or early third century CE). These scribal features all amount to an interesting pattern of bookproduction. Formal copies of Greek literary works do not usually have these sorts of phenomena. But it is important to note that early Jewish copies of biblical writings have a number of scribal features that are similar and that seem to reflect a corresponding concern for ease of reading.30 So, whatever may be the origin of the nomina sacra or the Christian preference 29.  E. G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977), 84–87. 30.  Emanuel Tov, ‘Scribal Features of Early Witnesses of Greek Scripture’, in The Old Greek Psalter: Studies in Honour of Albert Pietersma, ed. Robert J. V. Hiebert, Claude E. Cox and Peter J. Gentry (Sheffield: Sheffield Academic, 2001), 125–48. Many years ago, Colin Roberts noted such sense-unit spacing in the pre-Christian roll of Deuteronomy (P.Ryl. 458): C. H. Roberts, Two Biblical Papyri in the John Rylands Library Manchester (Manchester: Manchester University Press, 1936).

5. The ‘Meta-Data’ of Earliest Christian Manuscripts

97

for the codex, it seems rather clear that at least some features of early Christian scribal practice reflect the influence and appropriation of Jewish scribal practice. More precisely, I emphasize that these features in Jewish and early Christian manuscripts seem to indicate the preparation of copies intended for public, liturgical reading. One question for further investigation is whether these scribal features are in fact more characteristic of Christian texts generally or are more specifically characteristic of copies of certain texts prepared for public reading. If the latter is the case, then in the presence or absence of these scribal phenomena we might have indications of how individual Christian texts were regarded in the second and third centuries by the scribal devices deployed in the copies of them. One final feature has a significance beyond its apparently pedestrian character. Our earliest Christian manuscripts also include corrections. In some cases, the corrections seem to have been made by the copyist, but in other cases by another hand. Both kinds of corrections are worth noting. The first and most obvious thing to observe is that corrections probably indicate an interesting concern for care in what the text says, and may well even reflect a concern for care in copying. In the case of corrections by the original copyist or by another contemporary hand, we may presume that often they were made by comparing the new copy against an exemplarcopy. This makes an interesting and somewhat contrasting phenomenon in comparison with other undeniable indications that Christians could also engage in a certain freedom and flexibility in citing and in copying their texts. In short, in the second and third centuries CE Christian textual transmission may have been characterized by more than one trend or type of practice: interesting flexibility and an equally interesting concern for care in copying. Moreover, corrections may also reflect a setting that may amount to a scriptorium. Corrections in another contemporary scribal hand may be especially significant as evidence of this. That is, the latter type of corrections seems to indicate one copyist preparing a manuscript, and another person, perhaps someone in a kind of supervisory role, reviewing the work. These corrections certainly seem to comprise an interesting suggestion of a concern for ‘quality control’. I repeat that these corrections are evident already in Christian copies of biblical texts as early as the late second century, and the extant examples are unlikely to be the first manuscripts that received this kind of care. So, although it will go against the contentions of some scholars, perhaps the idea of second-century Christian scriptoria, or at least some scriptoria-like care in copying texts in some circles, is not so implausible after all.

98

Texts and Artefacts

Conclusion Although brief and introductory, I hope that my remarks have been sufficient to highlight the importance of early Christian manuscripts as historical artefacts crucial for historical analysis of the origins of Christianity. The kinds of issues on which these artefacts may cast some useful light include the formation of a Christian canon, especially the formation of what came to be the NT, the emergent group-identity of Christianity in the second century (and perhaps even earlier), and the historical relationship and indebtedness of emergent Christianity to Jewish traditions. I hope that a growing number of specialists in Christian origins and ancient Judaism will come to take account of manuscripts as artefacts in terms of their value for understanding broader historical questions.

Chapter 6 Manuscripts and the Sociology of Early Christian Reading

I In this essay, I take a cue from a programmatic and path-breaking study by William A. Johnson, ‘Toward a Sociology of Reading in Classical Antiquity’, in which he cogently argued that previous attempts to portray reading in the Roman era were too generalized, and that The more proper goal…is to understand the particular reading cultures that obtained in antiquity, rather than to try to answer decontextualized questions that assume in ‘reading’ a clarity and simplicity it manifestly does not have.1

In the concluding lines of his study, Johnson urges that ‘we urgently need…to frame our discussions of reading, whether ancient or modern, within highly specific sociocultural contexts’.2 Johnson chose to focus on ‘the reading of Greek literary prose texts by the educated elite during the early empire (first and second centuries A.D.)’, freely acknowledging, however, that it was one of a number of specific ‘reading cultures’ of the time.3 In the much more modest and exploratory study that follows, I focus on the particular ‘reading culture’ comprised of Christians in the first three centuries. I shall argue that there is a distinguishable Christian reading-culture, another ‘specific sociocultural context’, and that early Christian manuscripts are direct artefacts of it. The

1.  William A. Johnson, ‘Toward a Sociology of Reading in Classical Antiquity’, AJP 121 (2000): 593–627 (quotation from 606). 2.  Ibid., 625. Evidently, for Johnson, elite social circles considered collectively comprised a ‘highly specific’ reading context. So, I trust that it will be equally appropriate to treat early Christian circles broadly as well as another specific reading context. 3.  Ibid., 606.

100

Texts and Artefacts

comparison is appropriate, for early Christians were particularly given to the reading of certain ‘literary texts’ (especially those that functioned for them as scriptures) in their gatherings, as we know from a variety of Christian references of the time (e.g., Col 4:16; 1 Tim 4:13). Indeed, as Harry Gamble showed in his magisterial study of early Christian books and readers, the production, copying, circulation and reading of texts was a remarkably prominent part of early Christian activities.4 Perhaps the most intriguing contribution of Johnson’s essay was his argument that the typical format of high-quality Greek prose-text manuscripts was intended to reflect and validate the elite nature of the reading groups in which they were to be read. Taking his analysis of this matter as sound, I wish to consider here the formatting of the earliest Christian literary-text manuscripts with a view to considering how they, too, likely reflect the nature of the circles in which they were to be read. We shall see that there are striking differences in format between the manuscripts that Johnson considered and those addressed here. I shall argue that these differences were deliberate, and that the format-features of earliest Christian manuscripts reflect and affirm the very different socio-cultural character of the Christian circles of the time. Before we turn to the Christian manuscripts and their features, however, it will be helpful to take further notice of Johnson’s analysis of pagan prose literary manuscripts and what they tell us about the ancient readers for whom they were intended. It is well known that high-quality Greek prose-text manuscripts of the Roman era were formatted in what seems to us (and rightly so) to have made huge demands on readers. The most obvious feature, ‘scriptio continua’, required readers to form words in an uninterrupted flow of Greek alphabetic characters, with no word-division or sense-unit demarcation, and typically no punctuation. This format is especially noteworthy, given that copies of ancient school exercises often have word-division, and ‘elaborate visual structural markers’ appear often in documentary texts and inscriptions. Moreover, Hebrew manuscripts of the time, such as the Qumran texts, have word-separation and spaces to indicate senseunits. Indeed, in copies of Roman literary texts prior to the period under consideration here, ‘word separation is the norm, in fact universal so far 4.  Harry Y. Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven: Yale University Press, 1995). I have referred to early Christianity as a particularly ‘bookish’ religion in its time: Larry W. Hurtado, Destroyer of the Gods: Early Christian Distinctiveness in the Roman World (Waco: Baylor University Press, 2016), 105–41.

6. Manuscripts and the Sociology of Early Christian Reading

101

as we know’.5 This latter datum makes it all the more interesting that in the centuries we are concerned with here the Romans departed from wordseparation and adopted scriptio continua. As Johnson observed, this is ‘a choice they would hardly have made if it interfered fundamentally with the Roman reading system’.6 So, assuming that Romans were not stupid, the deliberate move to this rather demanding format must have served something in the ancient reading culture. In short, prose literary texts were not formatted in scriptio continua because the ancients were incapable of thinking of a less demanding way of presenting texts for reading. Instead, Johnson cogently contends, this format was intended to reflect and serve the specific elite cultural settings in which the texts were to be read. This will become still clearer if we take further note of the visual features of these manuscripts.7 They were typically not codices but handwritten rolls, held horizontally between the hands, the texts written in vertical columns ranging from 4.5 to 7.0 centimetres width, about 15–25 letters per line, left and right justification, and about 15–25 in height, with about 1.5–2.5 centimetres spacing between columns. The letters were carefully written, calligraphic in better quality manuscripts, but with no spacing between words, little or no punctuation, and no demarcation of larger sense-units. The strict right-hand justification was achieved by ‘wrapping’ lines (to use a computer term), ending each line either with a given word or a syllable, and continuing with the next word or syllable on the next line, the column ‘organized as a tight phalanx of clear, distinct letters, each marching one after the other to form an impression of continuous flow, the letters forming a solid, narrow rectangle of written text, alternating with narrower bands of white space’.8 As Johnson observed, The product seems, to the modern eye, something almost more akin to an art object than a book; and, with its lack of word spaces and punctuation, the ancient bookroll is, to the modern perception, spectacularly, even bewilderingly, impractical and inefficient as a reading tool. But that the ancient reading and writing systems interacted without strain is indisputable: so 5.  Johnson, ‘Toward a Sociology of Reading’, 608. Johnson cites E. Otha Wingo, Latin Punctuation in the Classical Age (The Hague: Mouton, 1972). 6.  Johnson, ‘Toward a Sociology of Reading’, 609. 7.  I draw directly here on Johnson’s concise summary, ‘Toward a Sociology of Reading’, 609–10. For a fuller discussion, see Eric G. Turner, Greek Manuscripts of the Ancient World, 2nd ed., rev. Peter J. Parsons (London: Institute of Classical Studies, 1987); and now especially the detailed study by William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004). 8.  Johnson, ‘Toward a Sociology of Reading’, 609.

102

Texts and Artefacts stable was this idea of the literary book, that with only small variations it prevailed for at least seven hundred years in the Greek tradition. The economical hypothesis is that the reading culture was likewise stable, and that readers were so thoroughly comfortable with the peculiarities of the writing system that adjustments…proved unnecessary over a great deal of time.9

Johnson probed further the visual qualities of literary bookrolls, pointing to ‘the beauty of the letter shapes, and the elegant precision of placement for the columns’, which reflect ‘[t]he elaborate care taken by scribes in the production of a literary prose text’, achieving ‘an elegant harmony that speaks loudly to aesthetic sensibilities’. In his words, That the physical literary roll not only contained high culture, but was itself an expression of high culture, does not need to be argued at length. The product itself makes it fairly obvious.10

Johnson then shows that in the cultured circles in which such manuscripts were read, ‘the use of literary texts…is deeply rooted in that sense of refined aesthetic enjoyment so formative in the interior construction of a cultural elite’. He makes an interesting comparison with the way that opera functions in contemporary elite culture, ‘the very difficulty serv[ing] to validate the activity as one exclusive to the educated and cultured’.11 Having noted that the reading of such prose texts was a favoured feature of social gatherings of the cultural elite, he emphasizes that in these settings ‘it was the reader’s job [emphasis his] to bring the text alive, to insert the prosodic features and illocutionary force lacking in the writing system’. He goes on to observe that ‘[t]he reader played the role of performer, in effect, and the sort of direction for pause and tone given by the author’s para-linguistic markup in our texts (commas, quotes, italics, indentation, etc.) was left to the reader’s interpretation of the lines’. In short, ‘A surprising amount of the burden to interpret the text was shifted from author to reader’.12 The ability to rise to this challenge is part of what marked off very skilled readers from others. 9.  Ibid., 609–10. As Johnson shows, this means that modern proposals that texts were not read out but were memorized and declaimed orally are very much dubious. 10.  Ibid., 612. 11.  Ibid., 615. 12.  Phrasings cited in this paragraph are all from ibid., 620. Johnson notes (n. 39) the extant manuscripts that have punctuation and other readers’ aids added by their users, evidence of the preparatory work that readers did to deliver the text appropriately.

6. Manuscripts and the Sociology of Early Christian Reading

103

It will be important for what follows also to take account of Johnson’s observations that in the ancient Roman setting reading is ‘tightly bound up in the construction of the community’, and that ‘[r]eading of literary prose, often difficult and inaccessible to the less educated, is part of that which fences off the elite group from the rest of society’.13 II Taking Johnson’s study as the basis, in what follows I offer a complimentary (but much more modest) pilot-study of the reading of literary texts in worship gatherings of Christian circles of the first three centuries. As in Johnson’s study, I wish to draw particular attention to the visual features of earliest Christian manuscripts as reflective of the social character of the early Christian circles that used them, and perhaps also as indicative of a deliberate effort to format Christian texts in a manner that contrasts with the sort of elite-oriented copies that were the Johnson’s focus. The Social Setting of Early Christian Reading As noted already, we know from various sources that the reading of certain texts formed a frequent part of the worship gatherings of Christian believers. I trust that it will be sufficient to provide some limited illustration of this, especially given the impressive treatment of the matter by Gamble.14 In addition to the references to the reading of texts in Christian worship gatherings already cited, note also Paul’s demand that his letter to the Thessalonian church be read ‘to all the brothers’ (1 Thess 5:27), and Justin’s description of Christian assemblies in the mid-second century, in which he states that ‘the memoirs of the apostles or the writings of the prophets are read for as long as time permits’, followed by an address by ‘the president’ (1 Apol. 67). Indeed, Edwin Judge showed that the place of preaching/teaching in Christian gatherings (activity often connected to the reading of sacred texts) led to some pagan observers likening these more to philosophical circles than religious ones.15 We also know something of the social character of Christian circles of the first few centuries, although certainly not as much as we would wish. Starting with evidence of first-century churches, it seems that they 13.  Ibid., 623. 14.  See, e.g., Gamble, Books and Readers in the Early Church, 205–31. 15.  Edwin A. Judge, ‘The Early Christians as a Scholastic Community’, Journal of Religious History 1 (1961): 4–15, 125–37.

104

Texts and Artefacts

typically involved a diversity of people: free(d) and slaves, males and females, older and younger people, and a variety of socio-economic levels from workers to small/medium property owners and owners of businesses and, occasionally, individuals of some wealth and modest social status.16 As Meeks observed in his study of Pauline churches, ‘The extreme top and bottom of the Greco-Roman social scale are missing from the picture’, with ‘no landed aristocrats, no senators, equites, nor (unless Erastus might qualify) decurions’. Likewise, ‘There may well have been members of the Pauline communities who lived at the subsistence level, but we hear nothing of them’. Social strata in between the extremes are, however, ‘well represented’, including slaves. But in the Pauline letters, ‘The “typical” Christian…the one who most often signals his presence in the letters by one or another small clue, is a free artisan or small trader’. Overall, however, there was ‘a mixture of social levels in each congregation’, reflecting ‘a fair cross-section of urban society’.17 Across the second and third centuries, there appears to have been an increasing, but still small, number of higher-status converts. In Justin Martyr we even have a Christian seeking to be taken seriously as a philosopher (but we cannot be confident whether he succeeded in this aim beyond Christian circles).18 Granted, by the third century CE, Christian converts included some from upper echelons of Roman society. Nevertheless, the overall picture of the social makeup of Christian groups does not radically change from the first-century urban groups reflected in Paul’s letters. This is broadly in line with Lampe’s detailed analysis of evidence for Christians in Rome in the first few centuries.19 Trebilco’s study of early Christianity in Ephesus does not deal explicitly with social strata, but he does include a discussion of references to material 16.  The pioneering study is Edwin A. Judge, The Social Pattern of Christian Groups in the First Century (London: Tyndale, 1960), reprinted in Edwin A. Judge, Social Distinctives of the Christians in the First Century: Pivotal Essays by E. A. Judge, ed. David Scholer (Peabody, MA: Hendrickson, 2008), 1–56. Note also Abraham J. Malherbe, Social Aspects of Early Christianity (Baton Rouge: Louisiana State University Press, 1977), and Wayne A. Meeks, The First Urban Christians: The Social World of the Apostle Paul (New Haven: Yale University Press, 1983), esp. 51–73, on ‘The Social Level of Pauline Christians’. 17.  The phrasings cited are all from Meeks, The First Urban Christians, 73. 18.  See the multi-author volume from the recent conference on Justin held in the University of Edinburgh: Sara Parvis and Paul Foster, eds., Justin Martyr and His Worlds (Minneapolis: Fortress, 2007). 19.  Peter Lampe, From Paul to Valentinus: Christians at Rome in the First Two Centuries (Minneapolis: Fortress, 2003 [German 1989]), esp. 138–50, 351–55.

6. Manuscripts and the Sociology of Early Christian Reading

105

possessions in the Christian texts that may reflect Ephesian Christianity, noting indications of some moderately wealthy believers among larger bodies of Christians generally of more modest means.20 As far as levels of education and abilities in reading/writing in particular, we have indications of some Christians with what looks like a ‘grammar school’ education, but it is hard to find anyone from the elite cultural levels, such as Celsus, the second-century critic of Christianity (who progressed on through the highest level of the three levels of classical education).21 The main point, however, is that typical Christian circles of the first three centuries were socially mixed, with most adherents from lower social strata, a minority from middle levels, and a very few from upper levels. To be sure, studies by Judge, Malherbe and others have forged a majority-view different from early twentieth-century pictures of early Christianity as a movement wholly made up of the dispossessed.22 We should recognize that earliest Christian circles comprised typically a variety of social levels, and that people with some property and level of education, from the first, seem often to have exercised particularly influential roles. But this very mixture of social levels immediately distinguishes Christian groups from the more homogenous elite circles that formed the focus of Johnson’ study. That is, the diversity of social levels typical of early Christian circles is what gives them a specific, perhaps distinctive, sociocultural identity. Judge remarked that early Christianity ‘broke through social barriers and encompassed people of every level of community life in a way that had never been the case with any movement of ideas of an organized kind’.23

20.  Paul Trebilco, The Early Christians in Ephesus from Paul to Ignatius, WUNT 166 (Tübingen: Mohr Siebeck, 2004), 404–45, and esp. his conclusion to this chapter, 443–45. 21.  Rafaella Cribiore, Gymnastics of the Mind: Greek Education in Hellenistic Egypt (Princeton: Princeton University Press, 2001), esp. Chapters 6–8, on the three levels of education. For Celsus, see Henry Chadwick, trans., Origen: Contra Celsum (Cambridge: Cambridge University Press, 1965); Jeffrey W. Hargis, Against the Christians: The Rise of Early Anti-Christian Polemic (New York: Lang, 1999). 22.  Abraham J. Malherbe, Social Aspects of Early Christianity (Baton Rouge: Louisiana State University Press, 1977); and Edwin Judge’s review of scholarship, ‘The Social Identity of the First Christians: A Question of Method in Religious History’, Journal of Religious History 11 (1980): 210–17, reprinted in Scholer, ed., Social Distinctives, 117–35. I cite this reprint of the essay here. 23.  Judge, ‘Social Identity’, 134.

106

Texts and Artefacts

Early Christian Manuscripts With this very brief and general sketch of the typical social settings comprised in early Christian gatherings, let us now turn to the visual features typical of earliest Christian manuscripts for what they tell us about the readers and settings for which they were copied.24 In what follows, I shall argue that earliest Christian manuscripts present us with a striking contrast to the sort of artefacts to which Johnson drew attention.25 I propose that the format typical of early Christian manuscripts suggests that they were prepared for a certain spectrum of mainly non-elite reader-competence. The first and most obvious feature to note about early Christian manuscripts is that the great majority of them are codices, reflecting a curious preference by early Christians for the codex over the bookroll for their literary texts. This preference is comparatively well known among scholars, but I am not confident that the full pattern of data has been engaged in some attempts to account for this preference. More specifically, the extant manuscript evidence suggests that early Christians generally preferred the codex for their literary texts, but especially for those literary texts that they most highly prized, those that functioned as scripture. Christians were somewhat more ready to use the bookroll for 24.  In the following discussion, I draw heavily on my fuller discussion of all these matters in L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). In other publications as well I have drawn attention to the physical and visual features discussed here, e.g., Larry W. Hurtado, ‘Early Christian Manuscripts as Artifacts’, in Jewish and Christian Scripture as Artifact and Canon, ed. Craig A. Evans and Daniel Zacharias, LSTS 70 (London: T&T Clark International, 2009), 66–80. 25.  In a recent stimulating essay, Scott D. Charlesworth discusses the physical/ visual features of early Christian Gospels manuscripts with a view to determining whether a given copy was intended for liturgical/public or private reading: ‘Public and Private: Second- and Third-century Gospel Manuscripts’, in Evans and Zacharias, eds., Jewish and Christian Scripture, 148–75. His essay is rich in details and refer­ ences, and his question and basic approach seem to me cogent. In particular, I agree that manuscripts with readers’ aids provided by the original copyist likely were intended for ease of public/liturgical reading. But I question his claim that the various readers’ aids he discusses are more common in second-century Gospel manuscripts than in third-century copies (148). Likewise, I hesitate over his insistence that the Christian copyist-conventions such as nomina sacra must indicate ‘controlled settings’ and early Christian copying centres (e.g., 149, 171–74). There may well have been some such settings, but the spread of these Christian copyist conventions do not seem to me to require these settings or necessarily reflect them.

6. Manuscripts and the Sociology of Early Christian Reading

107

other texts, such as theological treatises, liturgical texts, and other texts that may have functioned more for personal edification or study, although even in these genres the codex dominated. It may be helpful to give some figures, which are derived from a recent consultation of the Leuven Database of Ancient Books (LDAB).26 Of 41 Christian manuscripts dated to the second century in the LDAB catalogue, 24% are rolls, 76% codices; and of about 190 third-century CE Christian manuscripts 23% are rolls, and 77% codices. Moreover, although Christian items make up about 2% of the total of second-century manuscripts, about 27% of the total of second-century codices are identifiably Christian books. Christian items make up about 12% of the total of third-century manuscripts, but 38% of third-century codices.27 Across the first four centuries, over 70% of all codices are Christian manuscripts. Although the general preference for the codex among early Christians is clear and striking, I contend that it is even more interesting that use of the codex is nearly total for copies of the texts that came to form the Christian canon. Among about 75 copies of OT texts dated to the second and third centuries CE, probably no more than 7% of those that we may confidently take as Christian manuscripts are rolls. So far as NT writings are concerned, we do not have a single extant copy written on an unused roll (excluding, thus, the few examples of re-used rolls, ‘opisthographs’, such as P22). By contrast, of 58 second/third-century copies of extracanonical Christian texts, some 34% are bookrolls. So, as noted already, there appears to have been a somewhat greater Christian readiness to retain the bookroll for these sorts of texts. In light of the overwhelming general preference for the bookroll in the period considered here, especially for copies of literary texts, the Christian preference for the codex book-form suggests some further observations. The first of these is that this preference for the codex must have been conscious and deliberate. Early Christians cannot have been unconscious that their preferred book-form was out of step with the larger book culture 26.  The LDAB is a valuable online database which as of the date of writing this essay (March 2010) has over 15,000 manuscripts logged: http://www.trismegistos. org/ldab/. 27.  Roger Bagnall has recently questioned the second-century dating of some Christian manuscripts, contending that many/most should be dated a bit later, to the third century CE: Early Christian Books in Egypt (Princeton: Princeton University Press, 2009). This would make little difference to my points here. Bagnall’s discussion of the Christian preference for the codex seems to me beset with some problems, as I note in my review of his book in Review of Biblical Literature: http:// www.bookreviews.org/pdf/7289_7933.pdf.

108

Texts and Artefacts

of the time. Indeed, the evidence suggests a particularly deliberate effort to move away from the bookroll for copies of texts that were intended to function in their assemblies as scripture, as part of their ritual culture, as texts that were associated closely with their gathered worship settings. This is all the more likely when we recognize what was involved. Preparation of a codex required skills and judgments distinguishable from and additional to those required of copyists on bookrolls. For the latter, the basic steps were to acquire a sufficient length of writing material (sold in shops by length and quality), and then copy the text in neat columns. But using a codex requires, for example, the ability to estimate how many sheets will be needed, decisions about how to construct the codex (e.g., one gathering or multiple gatherings, and if the latter the number of sheets in each gathering), and use of a different layout (wider lines). So, the choice of whether to use a bookroll or a codex was certainly not a casual one. I suggest that it was unlikely also that professional copyists of literary texts, who would have been given to the use of the bookroll, were ready to use the one or the other book-form with equanimity. A small constellation of other formatting features strengthens the suspicion that we are dealing with evidence of a specific reading-culture very different, even deliberately different, from the elite circles to which Johnson pointed.28 The space available in this essay requires me to present here only a brief overview of these matters.29 Let us begin with noting the nature of the copyists’ hands typical in Christian manuscripts. The hands are usually clear, competent and readable, but not calligraphic in visual appearance.30 The letters often include rounded forms and are less 28.  See now Alan Mugridge’s analysis of the quality and nature of the copyist hands of over 500 Christian manuscripts of the first four centuries: Copying Early Christian Texts: A Study of Scribal Practice, WUNT 362 (Tübingen: Mohr Siebeck, 2016). He concludes that the overwhelming majority reflect skilled (‘professional’) copyists, but fewer exhibit a calligraphic hand. 29.  I summarize briefly here manuscript features discussed more fully in Hurtado, The Earliest Christian Artifacts, 155–89. 30.  Among NT manuscripts of this early period, the hand(s) of P4/P64/P67, noticeably toward a calligraphic appearance, represents an exception (these three all possibly from the same codex). T. C. Skeat argued that these were remnants of a four-Gospel codex: ‘The Oldest Manuscript of the Four Gospels’, NTS 43 (1997): 1–34, reprinted in J. K. Elliott, ed., The Collected Biblical Writings of T. C. Skeat, NovTSup 113 (Leiden: Brill, 2004), 158–92. But see the criticism of Skeat’s case by Peter M. Head, ‘Is P4, P64 and P67 the Oldest Manuscript of the Four Gospels? A Response to T. C. Skeat’, NTS 51 (2005): 450–557. Recently, Don Barker has proposed that P39, 0206, 0232, and P88 may all be early third-century or even late

6. Manuscripts and the Sociology of Early Christian Reading

109

regular in size, but are carefully written separately and without ligatures. In a number of cases, the lettering is somewhat larger than characteristic of literary bookrolls, and the spacing between lines somewhat greater, resulting in fewer lines per column than in high-quality pagan literary manuscripts with comparable column-heights. Even in comparison with contemporary pagan codices containing literary texts, Christian codices of equivalent size tend to have fewer lines per page and few letters per line, as noted by Eric Turner in his invaluable study of ancient codices.31 In addition, Christian codices, especially those containing scripture texts, exhibit punctuation marking sense-units and probably intended to signal where a reader should pause slightly (e.g., often a ‘middle-stop’, a dot placed vertically mid-way in a line). As examples, among earliest Christian manuscripts we find punctuation in P75 (Gospels of Luke and John), and P66 (Gospel of John), both manuscripts dated ca. 200 CE. These and other early Christian manuscripts also show the use of slightly enlarged spaces in lines to mark sense-units (roughly corresponding to our sentences), which would similarly signal where to make a pause in reading the text. One also finds the use of a diaeresis, a set of two dots resembling the German umlaut, written above the initial vowel of a word that follows immediately another word ending in a vowel. This mark signals to the reader that the vowel above which it is written is not part of a diphthong but the first letter of a new word. In some cases, the diaeresis may also signal a vowel to be aspirated (i.e., a ‘rough-breathing’). We have an early examples of the diaeresis in P52 (the Rylands fragment of the Gospel of John), where it appears over the initial iota of ιουδαιοι (recto, line 1, John 18:31) and also over the initial iota of ινα (recto, line 2, John 18:32; verso, line 2), another manuscript which takes us back at least to the late second century CE.32 second-century manuscripts, all of which he also describes as ‘deluxe’ editions with a calligraphic hand and careful bilinearity: ‘How Long and Old Is the Codex of which P.Oxy. 1353 Is a Leaf?’, in Evans and Zacharias, eds., Jewish and Christian Scripture, 192–202. 31.  Eric G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977), esp. 85–87. In Turner’s list of 23 codices from the second/ third centuries with fifty or more lines per page (which he judged to be the upper end of the spectrum of lines per page), P.Chester Beattty IX–X (a copy of Daniel, Esther, et al.) is the only Christian codex (which has 45–57 lines per page). 32.  In light of the later dating of the ‘Egerton Gospel’ (P.Lond.Christ. 1 + P. Köln 6.255) towards the late second century CE, which formed the basis for Roberts’ dating of P52, the latter might well also have to be dated later than previously assumed. Cf.

110

Texts and Artefacts

I emphasize that these features are typically provided by the copyist, and not added by readers. In terms from the world of the automobile, they are not owner customizations but ‘factory equipment’. Collectively, they amount to a conspicuous effort to produce copies of texts with an emphasis on legibility, ease of reading, and even built-in guidance on how the text should be read. That is, the Christian manuscripts with these features reflect an effort to place somewhat less of a burden on the reader to decide how to deliver the text orally.33 I submit further that this constitutes a very different purpose in com­ parison to the manuscripts studied by Johnson. In the Christian manu­scripts, we have a greater concern for the content, the text itself, and what looks like comparatively less concern to produce a copy with strong aesthetic qualities. Unlike the manuscripts prepared for elite pagan circles, copies that are as much the product of high craftsmanship, almost objets d’art, the Christian manuscripts to which I refer seem to have been prepared to serve fully and simply the texts that they contain, and those who will read them. But it would be simplistic to conclude that early Christian manuscripts are merely utilitarian. If Johnson is correct that the format of the pagan literary rolls was intended to reflect and affirm the exclusivity of the elite social circles in which they were to be read, then Christian manuscripts (especially those that appear to have been prepared for public reading) typically seem to reflect a very different social setting, perhaps deliberately so. I propose that they reflect a concern to make the texts accessible to a wider range of reader-competence, with fewer demands made on readers to engage and deliver them. In turn, this probably reflects the more socially diverse and inclusive nature of typical early Christian groups. That is, I submit that these early Christian manuscripts are direct evidence and confirmation of the greater social breadth and diversity represented in early Christian circles, in comparison to the elite social circles in which pagan literary texts were more typically read. Can we go further and surmise that these manuscripts evidence a conscious turn away from the elitist format of high-quality literary manuscripts? If so, then the formatting of earliest Christian literary codices would represent the artefacts of a deliberate effort to reflect, affirm, and C. H. Roberts, An Unpublished Fragment of the Fourth Gospel in the John Rylands Library (Manchester: Manchester University Press, 1935), and the recent critique by Brent Nongbri, ‘The Use and Abuse of P52: Papyrological Pitfalls in the Dating of the Fourth Gospel’, HTR 98 (2005): 23–48. 33.  A similar proposal was offered by Turner, The Typology of the Early Codex, 85–86.

6. Manuscripts and the Sociology of Early Christian Reading

111

facilitate a distinguishable Christian book/text culture, one characterized by social inclusiveness and diversity. If high-quality pagan bookrolls were intended to be daunting to anyone but the elite, these Christian manuscripts appear to be intended to enable a greater range of Christians to serve in the public reading of texts in Christian gatherings. Indeed, we could also say that the social effects of this (and perhaps one of the intentions) included ‘enfranchising’ a wider social diversity of people in Christian circles in the public reading and discussion of literary texts, activities that were otherwise dominantly associated with elite social strata. The prominent place of the public reading and discussion of literary texts (esp. scriptures) in churches meant that these experiences (which, again, were more associated with cultured elite social strata) were extended to a wider diversity of people, including many for whom these sorts of experiences would otherwise not be likely. That is, early Christian manuscripts are probably also artefacts of this very interesting further social consequence of the centrality of the reading of literary texts in Christian gatherings. Also, there are noteworthy instances of what we may think of as Christian watermarks, or at least visual references to Christian faith. Indeed, I have referred to these as comprising the earliest extant evidence of an emerging ‘visual culture’ in ancient Christianity.34 I refer specifically to the ‘nomina sacra’ and the ‘staurogram’. I have discussed these phenomena rather fully elsewhere, and so here I shall simply present a summary of relevant matters.35 The so-called nomina sacra have been noted often, and in recent years have been the object of renewed interest.36 Although some scholars have proposed that they originated simply as readers-aids, intended to provide navigation points on the page, the majority view is that these curious abbreviations of key words in early Christian discourse were 34.  Larry W. Hurtado, ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity: Essays in Honour of Peter Richardson, ed. Stephen G. Wilson and Michel Desjardins (Waterloo, ON: Wilfrid Laurier University Press, 2000), 271–88. 35.  Hurtado, The Earliest Christian Artifacts, 95–134 (the nomina sacra), 135–54 (the ‘staurogram’). 36.  Colin H. Roberts, Manuscript, Society and Belief in Christian Egypt (London: Oxford University Press, 1979), 26–48, was the stimulus for renewed interest in the nomina sacra. My own article later drew further attention to the matter: L. W. Hurtado, ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998): 655–73 (republished as Chapter 7 of the present volume).

112

Texts and Artefacts

visual expressions of Christian piety.37 Moreover, they are distinctively Christian. Neither the specific words themselves (the earliest and most consistently treated are Θεός, Κύριος, Ἰησοῦς, and Χριστός) nor the forms in which they are written have a direct analogy or precedent beyond early Christian manuscripts. So widely is their distinctiveness recognized among experts in ancient manuscripts that the presence of any one of them on an otherwise unidentifiable fragment is typically sufficient for a palaeographer to judge it likely part of some Christian manuscript. Although there is a certain similarity in the reverential attitude behind the ways that the Tetragrammaton was treated in ancient Jewish manuscripts (e.g., written in a distinctive script, or replaced with a series of dots, or sometimes with ‘Elohim’), there are also crucial differences. Allowing for their variations, the nomina sacra are much more standardized. They are abbreviated forms of the words in question, with a horizontal stroke placed over the abbreviation. Moreover, it appears that the Jewish special treatment given to the Tetragrammaton was intended particularly to signal to readers not to pronounce YHWH but to use a reverential substitute, namely, ‫ ֲאד ֹנָ י‬in Hebrew or Κύριος in Greek. But there is no indication that in the public reading of early Christian texts reverential substitutes were used for the words represented by the nomina sacra. Instead, it appears that the lector pronounced the words normally, the abbreviated forms making no difference.38 That is, the nomina sacra seem to have been purely a visual phenomenon, a written/visual expression of reverence for the referents of the words in question. Readers of Christian texts would encounter and have to deal with them, but auditors would not.39 The so-called staurogram is a device that likewise seems to have been deployed in early Christian manuscripts as an expression of Christian faith.40 The device involves the superimposing of the majuscule letter 37.  Cf., e.g., C. M. Tuckett, ‘ “Nomina Sacra”: Yes and No?’, in The Biblical Canons, ed. J.-M. Auwers and H. J. de Jonge, BETL 98 (Leuven: Peeters, 2003), 431–58; and my discussion of the significance of the devices in Hurtado, The Earliest Christian Artifacts, 120–33. 38.  Granted, this is an argument from silence, as I am not aware of any reference to the way that Christian lectors handled the nomina sacra. 39.  But given the high regard for scripture texts in many early Christian circles, is it wildly imaginative to suppose that ordinary believers, even illiterate ones, might have asked to view the copy of a sacred text, out of admiration and reverence? 40.  In addition to my discussion in The Earliest Christian Artifacts, 135–54, see also my essay, ‘The Staurogram in Early Christian Manuscripts: The Earliest Visual Reference to the Crucified Jesus?’, in New Testament Manuscripts: Their Text and Their World, ed. Thomas J. Kraus and Tobias Nicklas (Leiden: Brill, 2006), 207–26 (republished as Chapter 8 of the present volume).

6. Manuscripts and the Sociology of Early Christian Reading

113

rho on the majuscule tau. The bare device itself can be traced in various non/pre-Christian utilitarian uses (e.g., as a symbol for ‘3’ or ‘30’), but it was adopted in some early Christian manuscripts and deployed in a uniquely Christian manner and with a distinctively Christian meaning. Specifically, the earliest Christian uses of the device are as part of the way that the words σταυρός (‘cross’) and σταυρόω (‘crucify’) are written in some early manuscripts containing NT texts. In these cases, the words are abbreviated, i.e., treated as nomina sacra (so with a horizontal line over the abbreviated form), the abbreviation including the first and final letter(s), and including also the tau and the rho of these words combined to form the ‘staurogram’ device. That the earliest extant Christian use of the device is in these manuscripts and solely as part of the words ‘cross’ and ‘crucify’ has led a number of scholars to judge that the Christian purpose was to allude visually to the crucified Jesus, the loop of the rho intended as a pictographic reference to the head of a crucified figure on a cross (represented by the tau).41 If this is correct (and I think it is), these instances of the ‘staurogram’ comprise our earliest visual references to the crucified Jesus, earlier by some 150 years than what are usually taken by historians of Christian art as the initial examples.42 In any case, the nomina sacra and the staurogram represent efforts to mark early Christian manuscripts visually as Christian. These scribal devices were not utilitarian in purpose. The nomina sacra were not intended really as abbreviations in the ordinary sense of that word; they did not function to save space. Nor did they have some pedestrian function, such as orientation points for readers on a codex page.43 They originated and developed as visual expressions of Christian piety, especially in the case of the four earliest words so treated, which have been referred to more 41.  E.g., Kurt Aland, ‘Bemerkungen zum Alter und Entstehung des Christogramms anhand von Beobachtungen bei P66 und P75’, in Studien zur Überlieferung des Neuen Testaments und seines Textes, ANTF 2 (Berlin: de Gruyter, 1967), 173–79; Erika Dinkler-von Schubert, ‘ΣΤΑΥΡΟΣ: Vom “Wort vom Kreuz” (1 Kor. 1,18) zum Kreuz-Symbol’, in Byzantine East, Latin West: Art-Historical Studies in Honor of Kurt Weitzmann, ed. Doula Mouriki et al. (Princeton: Department of Art and Archaeology, 1995), 29–39. 42.  Two Christian gems dated to the fourth century CE and a fifth-century seal in the Metropolitan Museum of Art in New York City are frequently cited as earliest visual depictions of the crucified Jesus. But cf. now Robin Margaret Jensen, Understanding Early Christian Art (London: Routledge, 2000), 131–41, who has recognized the ‘staurogram’ as likely an earlier pictographic reference to the crucified figure of Jesus. 43.  Contra Tuckett, ‘ “Nomina Sacra”: Yes and No?’

114

Texts and Artefacts

specifically as ‘nomina divina’.44 The earliest Christian use of the tau-rho device, the ‘staurogram’, is even more obviously a visual expression of Christian devotion. Earliest Christian manuscripts are not often calligraphic or luxurious, and as we have noted reflect an impressively conscious turn from the literary bookroll toward the codex, which in the general culture of the time was regarded as less elegant or appropriate for literary texts. But the nomina sacra and the staurogram in particular show a concern for imprinting a distinctive semiotic quality on early Christian manuscripts, identifying them specifically as Christian items. Conclusion Freely acknowledging the limited dimensions of the preceding essay, nevertheless, I hope to have shown that the earliest Christian manuscripts hold a significance well beyond the technical interests of palaeographers and papyrologists, that these ancient items are artefacts of historically noteworthy social developments comprised by earliest Christian circles. In particular, I contend that these manuscripts reflect and promoted a specifically Christian reading-culture that in its historical setting was innovative and remarkable. It comprised a social phenomenon very different from the elite reading culture studied by Johnson, and involved the enfranchising and affirmation of a diversity of social strata in the public reading and discussion of literary texts, specifically texts that formed the charter documents of their religious life.

44.  Schuyler Brown, ‘Concerning the Origin of the Nomina Sacra’, SP 9 (1970): 7–19.

Chapter 7 The Origin of the Nomina Sacra: A Proposal*

The nomina sacra are a collection of words (ultimately, fifteen became common) written in special abbreviated forms in Christian sources to indicate their sacred character (see Fig. 1).1 The words given this special treatment fall into three groups (1) the four earliest attested and most consistently rendered words, Ἰησοῦς, Χριστός, Κύριος, Θεός; (2) three additional terms, which appear to be slightly later and less uniformly

*  Earlier versions of this essay were given at the 1996 Annual Meeting of the Society of Biblical Literature and to postgradu­ate seminars at the universities of St Andrews and Aberdeen, and I thank all discussants for their comments. I particularly thank Τ. C. Skeat and John Gager for comments. I am also grateful to William Parkinson, one of my PhD students, for discussing my ideas with me and for help in verifying Greek paleographical practices as discussed later in this essay. 1.  Of these fifteen words, some are nouns that can have a variety of references, of course, even in the case of Θεός and Κύριος. In such cases, it is clear that the scribal aim (though not always consistently observed) was to write the words in the form of nomina sacra (i.e., in abbreviated forms) when they had sacred referents. The nomina sacra are briefly described in Β. M. Metzger, Manuscripts of the Greek Bible (Oxford: Oxford University Press, 1981), 36–37; C. H. Roberts, ‘Books in the Graeco-Roman World and in the New Testament’, in Cambridge History of the Bible. Vol. 1, From the Beginnings to Jerome, ed. P. R. Ackroyd and C. F. Evans (Cambridge: Cambridge University Press, 1970), 48–66; H. Y. Gamble, Books and Readers in the Early Church (New Haven: Yale University Press, 1995), 74–78, and now David Trobisch, Die Endredaktion des Neuen Testaments, ΝΤΟΑ 31 (Freiburg: Universitatsverlag Freiburg; Göttingen: Vandenhoeck & Ruprecht, 1996), 16–31. Probably the most influential discussion in English is C. Η. Roberts, ‘Nomina Sacra Origins and Significance’, in Manuscript, Society and Belief in Early Christian Egypt: The Schweich Lectures 1977 (London: Oxford University Press, 1979), 26–48. See now also my discussion in The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 95–134

116

Texts and Artefacts

treated: πνεύμα, άνθρωπος, σταυρός;2 and (3) the remaining eight, πατήρ, υιός, σωτήρ, μήτηρ, ουρανός, Ισραήλ, Δαυειδ, Ιερουσαλήμ, which are abbreviated less consistently and appear to have joined the list of sacred terms latest.3 Ludwig Traube’s 1907 book is usually credited with having drawn the attention of scholarly circles to this subject, and it is he from whom the label ‘nomina sacra’ derives.4 Among more recent scholars, C. H. Roberts in particular has emphasized the importance of the nomina sacra as a distinguishing feature of early Christianity.5 The major questions connected with the nomina sacra are (1) whether this scribal practice originated in pre-Christian circles or was a Christian innovation, and (2) what the practice represents and what religious impetus lies behind it. In this essay I wish to review the issues involved and offer a proposal as to how and why the phenomenon may have begun in Christian circles. I The nomina sacra appear in Greek, Latin, Coptic, Slavonic, and Armenian sources, including manuscripts, inscriptions, amulets, and icons as well, 2.  See M. Black, ‘The Chi-Rho Sign – Christogram and/or Staurogram?’, in Apostolic History and the Gospel Biblical and Historical Essays Presented to F. F. Bruce on His 60th Birthday, ed. W. W. Gasque and R. Ρ. Martin (Grand Rapids: Eerdmans, 1970), 319–27, for an interesting discussion of the special scribal treatment of σταυρός and verbal cognates as well. For briefer discussion, but with good illustrations of examples from P66, P75, and Gospel of Truth (NH I,3), see Jack Finegan, The Archaeology of the New Testament: The Life of Jesus and the Beginning of the Early Church, rev. ed. (Princeton: Princeton University Press, 1992), 381–82. See also Kurt Aland, ‘Bemerkungen zur Alter und zur Entstehung des Christogramms anhand von Beobachtungen bei P66 und P75’, in Studien zur Überlieferung des Neuen Testaments und Seines Textes, ANTF 2 (Berlin: de Gruyter, 1967), 173–79, and my discussion in The Earliest Christian Artifacts, 135–54. 3.  In addition to the more well-known fifteen, there are a few other terms written similarly but only in a few Christian sources (e.g., P.Bodmer VII and VIII has Μιχαήλ, Νώε, Σαρρα, Αβραάμ written with a stroke over them, and P.Bodmer XIII has the stroke over Αδάμ, δνιν, δυιν, Αβρμ. Only the last three are abbreviations, but the stroke written above them appears to signal that they are being treated as sacred terms by the scribe. The Egerton Gospel has Μω (Μωυσης), Ησας (Ησαΐας), Προφας (προφητας), and one instance of βαλευσι (βασιλευσι). The Nag Hammadi Cop­tic document Gospel of the Egyptians (NH 3 2) has ιχθύς with a stroke over it, which likely makes it a nomen sacrum there. 4.  Ludwig Traube, Nomina Sacra Versuch einer Geschichte der christlichen Kürzung (Munich: Beck, 1907). 5.  See esp. Roberts, Manuscript, 26–48.

7. The Origin of the Nomina Sacra

117

down to the Middle Ages.6 It is, however, particularly significant for the investigation of the origins of Christianity that nomina sacra (esp. the earliest four mentioned above) are found even in the very early scraps of Christian manuscripts, which take us back perhaps to the late or middle second century.7 6.  Traube discussed Greek and Latin evidence A H R E Paap (Nomina Sacra in the Greek Papyri of the First Five Centuries [Leiden: Brill, 1959]), updated Traube’s discussion and gave reference to evidence that had appeared subsequent to Traube’s study. Jose O’Callaghan (Nomina Sacra in Papyrus Graecis Saecuh HI Neotestamentarus [Rome: Biblical Institute, 1970]) supplements Paap. See also O’Callaghan’s list of Greek manuscripts of the fourth to eighth centuries in SPap 10 (1971): 99–122. For treatment of the nomina sacra in LXX manuscripts, see F. Bedodi, ‘ “Nomina sacra” nei papiri veterotestamentari pre-christiani’, SPap 13 (1974): 89–103, and Stanislaw Jankowski, ‘I “nomina sacra” nei papiri dei LXX (secoli II e III d C)’, SPap 16 (1977): 81–116. For examples of the nomina sacra on amulets, see Campbell Bonner, Studies in Magical Amulets, Chiefly Graeco-Roman (Ann Arbor: University of Michigan Press; London: Geoffrey Cumberlege/Oxford University Press, 1950), 172, 183–85. On Slavonic evidence, see Ute Sill, ‘Nomina Sacra’ im Altkirchenslavischen bis zum 11 Jahrhundert, Forum Slavicum 80 (Munich: Wilhelm Fink, 1972). D. C. Parker has recently compared the nomina sacra in the Greek and Latin columns of Codex Bezae (D), with interesting results (Codex Bezae: An Early Christian Manuscript and Its Text [Cambridge: Cambridge University Press, 1992], 97–106). David Taylor (University of Birmingham) informs me that the nomina sacra appear in Christian Armenian manuscripts of the NT. 7.  See, e.g., H. I. Bell and T. C. Skeat, Fragments of an Unknown Gospel and Other Early Christian Papyri (London: British Museum, 1935), who discuss Egerton Papyrus 2 and date it in the mid-second century. In personal correspondence (letter dated 21 June 1997), Skeat points to a small fragment of the Egerton Papyrus that ‘turned up in Cologne some years ago’ (P.Köln 6.255), which exhibits ‘the hook-shaped mark between consonants which is common in the 3rd cent, but decidedly rare in the 2nd’, and notes that ‘this may necessitate revising the dating of the MS’. For discussion of P.Egerton 2 and the Cologne fragment, see now J. Jeremias and W. Schneemelcher, ‘Papyrus Egerton 2’, in W. Schneemelcher, New Testament Apocrypha, trans. R. M. Wilson, rev. ed. (Louisville: Westminster John Knox; Cambridge: James Clarke, 1991), 196–99, who see the Cologne fragment as requiring that we must ‘be much more cautious with an early date than hitherto’, and accept suggestions that 200 CE might be a more secure dating. Cf. H. Koester, Ancient Christian Gospels: Their History and Development (London: SCM; Philadelphia: Trinity Press International, 1990), 205–16, who now consents to the later dating of the manuscript, but contends (contra Jeremias, Schneemelcher, et al.) that it is not dependent on the canonical Gospels. On the Bodmer papyri, see F. G. Kenyon, ‘Nomina Sacra in the Chester Beatty Papyri’, Aegyptus 13 (1933): 5–10 (approx. 200–250 CE). More recently, Τ. C. Skeat has proposed that P4, P64, and P67 are all

118

Texts and Artefacts

1. Four Earliest Nomina Sacra (from second-century manuscripts onward) ΙNT

Contracted form = ��, ��, etc.; suspended form = ��; conflate form = ���

Χρίστος

Contracted forms = ��, ��, etc.

Θεός

Contracted forms = ��, ��, etc.

Κύριος

Contracted forms = ��, ��, etc.

2. Other Nomina Sacra Πνευμα

���, ���, etc.

Υιος

��, ��, etc.

Δαυειδ

���

Μητηρ

���, etc.

Πατηρ

���, etc.

Ισραηλ

���

Σωτηρ

���

Ανθρωπος

����

Ιερουσαλημ

����

ουρανος

�����

σταυρος

���, and variation, e.g., �����, ����, ����, ���

(Also, the ‘staurogram’ [attested in some early manuscripts, e.g., P66, P75 D (ca. 200 CE)] = σ οσ ) Figure 1. Illustrations of the Nomina Sacra

These specially written forms appear in Christian documentary and literary manuscripts, the only apparent exceptions being some private texts8 (e.g., letters, prayers, magical texts) or errors from a careless scribe. They are found in Christian biblical manuscripts, noncanonical religious texts (e.g., the Egerton Gospel fragment), and in ‘orthodox’ and unorthodox Christian writings (e.g., the Coptic Gospel of Thomas, Acts of Peter, Acts of John).9 All this indicates a remarkable instance of portions of the same four-Gospel codex, and that the manuscript of which these are portions should be dated in the late second century, which would make it among the earliest Christian manuscripts (‘The Oldest Manuscript of the Four Gospels?’, NTS 43 [1997]: 1–34). Skeat mentions the nomina sacra in these fragments on p. 6. 8.  E.g., Ρ.Oxy. 3 407, a third-century CE Christian prayer text that does not abbreviate the nomina sacra. 9.  E.g., Ρ.Oxy. 6 849 (Acts of Peter), Ρ.Oxy. 6 850 (Acts of John), Ρ.Oxy. 6 851 (unknown apocryphal acts).

7. The Origin of the Nomina Sacra

119

standard­ization that contrasts with the wide diversity we have come to associate with the earliest centuries of Christianity.10 These abbreviated words are distinctive in form, subject matter, and function from other scribal phenomena, so much so that it is widely (but not universally, as indicated below) accepted that the presence of any of them in a manuscript is itself a good indication of its Christian provenance. Most familiarly, the nomina sacra are abbreviated in a contracted form, usually first and last letters of the particular inflected form of the word, with a horizontal stroke over the letters of the abbreviated form (Fig. 1). This distinguishes them from the kinds of abbreviations in non-Christian Greek manuscripts, ostraca, and inscriptions, which are usually abbreviation by ‘suspension’, the first letter or two written and the rest omitted, with varying marks to indicate an abbreviated word.11 As to subject matter, the four earliest nomina sacra are divine names or titles designating the two central figures in early Christian religious devotion (God and Jesus), and the remaining ones are religious terms important in early Christian circles representing other features of belief and devotion. This restricted range of the words, all of them significant terms in early Christian religious vocabulary, sets the nomina sacra apart from pagan abbreviations, which tend to be common words (e.g., καί) and other terms that over long usage had acquired sufficient recognition to permit abbreviated forms to suffice (e.g., in inscriptions the names of emperors and titles of officials – in British usage ‘HRH’ would be an analogy), or other easily recognizable words that are abbreviated if they are written at the end of a line.

10.  This observation is made by E. J. Epp, ‘The Significance of the Papyri for Determining the Nature of the New Testament Text in the Second Century: A Dynamic View of Textual Transmission’, in E. J. Epp and G. D. Fee, Studies in the Theory and Method of New Testament Textual Criticism, SD 45 (Grand Rapids: Eerdmans, 1993), 288. 11.  Kathleen McNamee, Abbreviations in Greek Literary Papyri and Ostraca, BASPSup 3 (Chico, CA: Scholars Press, 1981); Alain Blanchard, Sigles et abbréviations dans les papyrus documentaires grecs: Recherches de paléographie, Bulletin of the University of London Institute of Classical Studies Supplement 30 (London: Institute of Classical Studies, 1974); Michael Avi-Yonah, Abbreviations in Greek Inscriptions [The Near East, 200 ΒC–AD 1100] (London: Humphrey Milford, 1940), reprinted in A. Ν. Oikonomides, Abbreviations in Greek Inscriptions Papyri Manuscripts and Early Printed Books [Chicago: Ares, 1974]). Blanchard writes, ‘…dans le domaine papyrologique, la contraction n’est qu’un mythe’ (p. 2).

120

Texts and Artefacts

This brings us to the distinction in function between the nomina sacra and non-Christian Greek abbreviations. The simple purpose of Greek abbreviation is to save space (especially in inscriptions) and/or time and labour (especially in texts and ostraca). Abbreviations appear frequently in documentary papyri but scarcely in literary texts except in marginal notes and in copies prepared for private use of scholars.12 Frequently, the words abbreviated are simply those at random that occur at the end of a line. There is no standard system of abbreviation, and scribes appear to have followed their own preferences as to when to abbreviate and how to do so. In contrast, the nomina sacra do not really serve as abbreviations at all. They are not intended to conserve space or labour. They appear more frequently in Christian manuscripts prepared for formal usage, such as public reading, the Christian equivalent of ‘literary’ texts.13 As already mentioned, the words involved are a relatively fixed set of terms, all of which have fairly obvious religious meaning. The aim is clearly to express religious reverence, to set apart these words visually in the way they are written. In the nomina sacra, we encounter a fascinating manifestation of ancient Christian devotion, and these scribal symbols are perhaps the earliest surviving artifacts of an emerging Christian material culture.14 In fact, the origin of nomina sacra appears to take us back beyond the second-century manuscripts, in all likelihood well back into the first century. By the second century, the four divine epithets (Ἰησοῦς, Θεός, Χριστός, Κύριος) are consistently written as nomina sacra, and allowing even minimal time for the practice to gain sufficient recognition and standardization this would require an origin no later than the late first century. Epigraphical evidence of pagan abbreviations indicates that there is a time lag of at least several decades, and often longer, for terms to become sufficiently familiar to be recognized in abbreviated form by the intended readership.

12.  McNamee, Abbreviations, xi. 13.  On early Christian use and production of books, see now Gamble, Books and Readers in the Early Church. Cf. E. J. Epp, ‘The Codex and Literacy in Early Christianity and at Oxyrhynchus: Issues Raised by Harry Y. Gamble’s Books and Readers in the Early Church’, CRBR 10 (1997): 15–37. 14.  The other physical phenomenon peculiar to early Christianity often mentioned as indicative of an emerging ‘material culture’ is the Christian preference for the codex. See, e.g., Gamble, Books and Readers, 42–66.

7. The Origin of the Nomina Sacra

121

II Given that the nomina sacra are apparently both distinctively Christian and amazingly early, what relation do they have to the religious and cultural background of the early church, and what influences might have prompted and shaped them? As indicated already, pagan abbreviation practice does not seem to provide either a true analogy or the impetus. In individual scribal features, however, the Christian practice seems to show adaptation of some non-Christian scribal techniques. The horizontal stroke over the abbreviated term, which is standard in the Christian nomina sacra, may be compared with the variety of abbreviation marks found in non-Christian inscriptions and texts, among which we often find a stroke sometimes placed over the final letter or two and extending slightly beyond the abbreviation. But the particular form of the stroke used with the nomina sacra, placed squarely over the abbreviated form rather than at the end of the form, is different from abbreviation marks in non-Christian Greek materials and is more exactly mirrored in the stroke used above letters when they are intended to be read as numbers, a matter to which I return below. Though the Christian preference for contraction is distinctive, occasional contraction is found in some non-Christian sources, and so the basic device may not have been a Christian invention but may have been appropriated and adapted. Moreover, Christian nomina sacra abbreviated by suspension, though considerably less frequent in Christian texts, likewise show the appropriation of the mechanics of abbreviation from profane Greek practice.15 If Greek abbreviation practices provided a quarry of scribal devices to adapt, it is probably in Jewish tradition that we find the closest analogies for understanding the religious meaning and aim involved in the nomina sacra. It is well known that by the first Christian century devout Jews were very particular about the oral and written treatment of the divine

15.  Cf. Alan Millard (‘Ancient Abbreviations and the Nomina Sacra’, in The Unbroken Reed: Studies in the Culture and Heritage of Ancient Egypt in Honour of A. F. Shore, ed. C. Eyre, A. Leahy, and L. M. Leahy [London: Egypt Exploration Society, 1994], 221–26), who has suggested possible relevance of scribal practices found in ancient Phoenician sources. But one problem with his suggestion is that his examples are separated by a few centuries or more from the date of the Christian phenomena. Moreover, even if Millard’s suggestion were accepted, we would have only a possible derivation of the scribal device of abbreviation by contraction. The larger historical questions about what the nomina sacra signified for early Christians would remain unanswered.

122

Texts and Artefacts

name, YHWH.16 In extant pre-Christian Jewish biblical manuscripts, the divine name is characteristically written in special ways intended to distinguish it from the surrounding text.17 Emanuel Τον notes that the 16.  See M. Delcor, ‘Des diverses manières d’écrire le tétragramme sacré dans les anciens documents hébraïques’, RHR 147 (1955): 145–73; J. Z. Lauterbach, ‘Substitutes for the Tetragrammaton’, PAAJR (1930–31): 39–67. The ancient Jewish reverence for the name is reflected in the LXX translation of Lev 24:16, which in Hebrew forbids ‘blaspheming the name of Yahweh’, but in the LXX invokes death on one who ‘pronounces the name of the Lord’. See also Philo, Vit. Mos. 2.114, 205; Josephus, Ant. 2 §275. On the significance of names, especially divine names, see H. Bietenhard, ‘ὄνομα’, TDNT 5:242–83. On rabbinic traditions about the divine name, see Ε. E. Urbach, The Sages (Cambridge, MA: Harvard University Press, 1987), 124–34. See also C. H. Dodd, The Interpretation of the Fourth Gospel (Cambridge: Cambridge University Press, 1965), 93–96, for brief discussion of rabbinic notions about God’s name and their possible relevance for Johannine Christology. 17.  See the introductory discussion in Metzger, Manuscripts of the Greek Bible, 33–35. On the Qumran materials, see J. P. Siegel, ‘The Employment of Palaeo-Hebrew Characters for the Divine Names at Qumran in the Light of Tannaitic Sources’, HUCA 42 (1971): 159–72; H. Stegemann, ‘Religionsgeschichtliche Erwägungen zu den Gottesbezeichnungen in den Qumrantexten’, in Qumrân: Sa piété, sa théologie et son milieu, ed. M. Delcor (Leuven: Leuven University Press, 1978), 195–217; and P. W. Skehan, ‘The Divine Name at Qumran, in the Masada Scroll, and in the Septuagint’, BIOSCS 13 (1980): 14–44. For a review of pre-Christian Greek biblical manuscript evidence, see George Howard, ‘The Tetragram and the New Testament’, JBL 96 (1977): 63–68; but cf. Albert Pietersma, ‘Kyrios or Tetragram: A Renewed Quest for the Original LXX’, in De Septuaginta: Studies in Honour of John William Wevers on his Sixty-fifth Birthday, ed. A. Pietersma and Claude Cox (Mississauga, ON: Benben, 1984), 85–101, who argues that the rendering of the divine name in Hebrew characters in some early Greek biblical manuscripts represents an effort at re-hebraizing and that the original practice in Greek biblical manuscripts had been to translate YHWH as Kyrios. Pietersma also contends against Howard that it is most unlikely that this re-hebraizing process was sufficiently successful to allow for Howard’s theory that all pre-Christian, Jewish Greek biblical manuscripts would have had YHWH represented in Hebrew characters. More recently, J. R. Royse has argued (a) that Philo must have read biblical texts with the tetragrammaton written in paleoHebrew or Aramaic (‘square’) characters, not translated by kyrios, and (b) that Philo probably pronounced the tetragrammaton as kyrios in his reading of these biblical manuscripts (‘Philo, ΚΥΡΙΟΣ, and the Tetragrammaton’, in The Studia Philonica Annual: Studies in Hellenistic Judaism, Vol. 3, ed. D. T. Runia [Atlanta: Scholars Press, 1991], 167–83). In some Qumran texts, the Hebrew El (used with reference to God) is sometimes written in paleo-Hebrew characters as well, e.g., 1QH 1.26; 2.34; 7.5; 15.25 (noted by Delcor, ‘Des diverses manières’, 147 n. 2; and Skehan, ‘Divine Name’, 17).

7. The Origin of the Nomina Sacra

123

tetragrammaton is accorded special treatment in a variety of ways in the Qumran Hebrew material: sometimes represented by four or five dots, sometimes preceded by a colon, and sometimes represented in paleoHebrew characters.18 The divine name is also accorded special treatment in some Greek biblical manuscripts of Jewish provenance. In the Qumran Greek scroll of Leviticus (4QLevb) the tetragrammaton is written as ιαω.19 In the Greek Ρ.Fouad 266 manuscript (Rahlfs 848, containing Genesis and Deuteronomy), the tetragrammaton appears in square Hebrew characters, and in a Greek scroll of the minor prophets (8HevXIIgr), ‫ יהוה‬is written in paleo-Hebrew form.20 Ρ.Oxy. 50.3522 (first century CE) likewise has ‫ יהוה‬in archaic Hebrew characters.21 In Ρ.Oxy. 7.1007, a Greek manuscript of Genesis from the third century CE, we have the divine name written as a double yod (the ‫ י‬written in the form of a Z) with a horizontal

Josephus (Ant. 12 §89) refers to Hebrew biblical manuscripts with gold characters, probably reserved for the tetragrammaton, and Aristeas 176 may refer to the same practice. Codex purpureus Petropolitanus is an example from later times. Origen (Psalmos 2.2) refers to the Jewish practice of pronouncing ‫( ֲאד ֹנָ י‬in Hebrew) or Κύριος (in Greek) when reading the scriptures aloud, and he also refers to Jews writing YHWH in archaic Hebrew characters in Hebrew biblical scrolls. 18.  Emanuel Τον, Textual Criticism of the Hebrew Bible (Minneapolis Fortress, 1992), 216, 220. In addition to ‫יהוה‬, ‫אלהים‬, and ‫ אל‬appear in Paleo-Hebrew characters in the Qumran scrolls. Τοv also points to indirect evidence that ‫ יהוה‬may have been abbreviated by the initial ‫ י‬in some (not extant) Hebrew manuscripts, as seems to be reflected in some LXX readings (pp. 256–57). 19.  Ρ. W. Skehan, E. Ulrich, J. E. Sanderson, and Ρ. J. Parsons, Qumran Cave 4 IV: Palaeo-Hebrew and Greek Biblical Manuscripts, DJD 9 (Oxford: Clarendon, 1992), see pp. 168–69 on use of Ιαω for ‫ יהוה‬in 4QLXXLevb (Lev 3:12 [frag. 6], Lev 4:27 [frag. 20]). G. A. Deissmann (‘Greek Transcriptions of the Tetragrammaton’, in Bible Studies [Edinburgh: Τ. & Τ. Clark, 1901], 321–36) gives examples of Ιαω and other renderings of ‫ יהוה‬in various nonbiblical Greek documents, particularly in magical texts, where it is crucial for readers to be given the pronunciation of the name(s) of the deities being invoked. 20.  Emanuel Τοv, The Greek Prophets Scroll from Nahal Hever (8HevXllgr) (The Seiyal Collection), DJD 8 (Oxford: Clarendon, 1990), see p. 12 on the writing of the tetragrammaton in paleo-Hebrew characters (24 samples in Hand A, 4 in Hand B). This practice may reflect a Jewish adaptation of the use of foreign/exotic characters or signs (called characteres) attested in magical amulets (see Bonner, Magical Amulets, 12, 194–95). 21.  Ρ. J. Parsons, J. R. Rea, and E. G. Turner, eds., The Oxyrhynchus Papyri, Vol. 50 (London: British Academy, 1983), 1–3.

124

Texts and Artefacts

line through the two letters, another substitute for the divine name similar to substitutes found in later Jewish manuscripts.22 There is no indisputably Jewish manuscript in which any of the nomina sacra are written as we find them in undeniably Christian manuscripts. But it seems likely that Jewish reverence for the divine name, and particularly the Jewish practice of marking off the divine name reverentially in written forms, probably provides us with the key element in the religious background that early Christians adapted in accordance with their own religious convictions and expressed in the nomina sacra. The four earliest nomina sacra represent Chris­tian reverence of God and Christ expressed in the special way these key terms were written in Christian texts. But in this desire to show reverence for divine names and titles in the way they are written out, we have a continuity in what we may call religious psychology between Jewish tradition and emerging Christian tradition.23 As Schuyler Brown noted, the four earliest words are more correctly nomina divina, terms that in early Christian devotion function somewhat analogously to the divine name in Jewish tradition.24 22.  Ρ.Oxy. 7.1007 also has the contracted form of theos, which is otherwise attested only in Christian manuscripts Roberts (Manuscript, 33–34) suggests that Ρ.Oxy. 7.1007 may come from a Jewish-Christian group, which would account for the apparent mixture of Jewish and Christian writing practices. Kurt Treu argues that Ρ.Oxy. 7.1007 is Jewish and that contraction of theos was practiced among Jews and may have begun among them (‘Die Bedeutung des Griechischen fur die Juden im romischen Reich’, Kairos NF 15 [1973]: 123–44). I am inclined toward Roberts’s view in light of the evidence of undisputed Jewish manuscripts. Whatever the case may be, this manuscript is unique in having both a nomina sacra form of theos and the abbreviation/substitute for ‫יהוה‬. The abbreviation of ‫ יהוה‬as ΖΖ (with a horizontal stroke through the two letters) is also found on some intaglia. For discussion of these data and of other substitutions for the tetragrammaton, see G. R. Horsley, New Documents Illustrating Early Christianity, Vol. 4 (Sydney, NSW: Macquarie University, 1987), 232. 23.  The likelihood of early Christian appropriation and adaptation of Jewish scribal practices is suggested by P. J. Parsons (‘The Scripts and Their Date’, in Τοv, Greek Prophets Scroll from Nahal Hever, 19–26), who points out that the use of enlarged initial letters at line beginnings, phrase beginnings, and new sections, characteristic of Christian books and in contrast to copies of the Greek classics, seems to have been inherited from Greek-speaking Jewish circles (pp. 23–24). Millard, also, suggests that the Christian scribal practices may have been influenced by Jewish and other West Semitic scribal traditions (‘Ancient Abbreviations’, 223–24). Finegan points to marginal scribal marks in 1QIsa that seem to be comparable to marks employed by Christian scribes later (Archeology of the New Testament, 346–48). 24.  S. Brown, ‘Concerning the Origin of the Nomina Sacra’, SPap 9 (1970): 7–19 (19).

7. The Origin of the Nomina Sacra

125

There are differences to be noted. First, Jewish reverence for the divine name extended to a reluctance to pronounce it in Hebrew, and a use of oral substitutes such as ‫ ֲאד ֹנָ י‬in Hebrew and Κύριος in Greek. Their various ways of writing the tetragrammaton were intended both to express reverence and probably to signal to readers that they should use a substitute for the divine name. There is, however, no indication of early Christian reluctance to pronounce the nomina sacra or of the use of substitutes for them in reading aloud. Second, though our evidence is limited, it appears that there were a variety of devices used by Jewish scribes to express reverence for the divine name. By contrast, Christian practice seems to have limited itself to relatively standardized abbreviated forms of key words, usually contracted forms, with the horizontal stroke over the abbreviated word. Third and most significant, of course, the Christian nomina sacra quickly (perhaps originally) included key terms (names/titles) for God and Christ, reflecting the binitarian shape of earliest Christian devotion, which distinguished it from the Jewish matrix of Christianity.25 Whatever may be the origin of the nomina sacra, and whatever the relationship to Jewish scribal practices, the treatment of Ἰησοῦς and Χριστός in the same special way as Κύριος and Θεός is undeniably an early Christian innovation and signifies momentous religious developments. III In our earliest surviving Christian sources, the four terms Θεός, Κύριος, Ἰησοῦς, and Χριστός are already established as nomina sacra, but scholars won­der if earlier still a reverential practice initially focused on one of these was the point of origin. I shall mention a few theories of others and then offer a proposal of my own.26 25.  1 have explored the emergence of this binitarian devotional practice and its relationship to the Jewish matrix in my book One God, One Lord: Early Christian Devotion and Ancient Jewish Monotheism (Philadelphia: Fortress; London: SCM, 1988; 2nd ed., Edinburgh: Τ. & Τ. Clark, 1998; 3rd ed., London: Bloomsbury T&T Clark, 2015). 26.  I omit here reference to the theory of E. Nachmanson (‘Die schriftliche Kontraktion auf den griechischen Inschriften’, Éranos 10 [1910]: 101–44), who proposed that the nomina sacra were more directly influenced by Greek contraction of words as evidenced in Greek ostraca and inscriptions. The evidence Nachmanson cites shows only that the technique of abbreviation by contraction was not a Christian invention. The nomina sacra, however, are not simply Christian abbreviations, and their function is not accounted for in any way by abbreviations in non-Christian

126

Texts and Artefacts

Traube proposed that the initial term treated as a nomen sacrum was Θεός and that the practice of writing it in contracted form began among Greek-speaking Jews who sought thereby to imitate the Hebrew consonantal writing of the divine name by omitting the vowels of Θεός.27 From this initial term, Traube posited, the practice spread among Greekspeaking Jews to include contracted forms of Κύριος, πνεύμα, πατήρ, ουρανός, άνθρωπος, Δαυειδ, Ισραήλ, and Ιερουσαλήμ, and this custom was subsequently adopted by Chris­tians, who then added to the list Ἰησοῦς, Χριστός, υιός, σωτήρ, σταυρός, and μήτηρ. Paap agreed that Θεός was the initial nomen sacrum, that the contracted form was an imitation of the consonantal spelling of ‫יהוה‬, and that the practice began among graecized Jews. But, citing the dates and provenance of the manuscript evidence, much of it not available to Traube, Paap argued that the development of the fuller list of nomina sacra beyond Θεός took place among Christians.28 Schuyler Brown pointed to major logical and factual problems in the theories of Traube and Paap, in particular the preference of Κύριος (not Θεός) among Greek-speaking Jews as the translation and Greek vocalization of ‫יהוה‬, and Brown proposed that Κύριος was the original nomen sacrum, written as a contraction and first used by Christian scribes as a reverential way to render the Greek substitute for ‫ יהוה‬in Christian copies of the Greek OT.29 Then, Brown suggested, from the Christian use of Κύριος to refer both to God and Christ, the practice of reverential contraction was ‘rapidly extended in one direction to θεός and in the other direction to Ἰησοῦς and χριστός’.30 Thereby, Christian scribes gave ‘graphic expression to the theological equation already present in the earliest apostolic preaching, in which Κύριος, the name for the God of Israel, was used as a title for Jesus Christ’.31 More recently, Kurt Treu has registered his preference for the theory that the nomina sacra began among Jews prior to Christian usage and initially included both Θεός and Κύριος, written as contractions with a horizontal stroke placed over them to distinguish them in Greek texts materials. See Paap (Nomina Sacra, 121–23) for criticisms of Nachmanson’s views. See also Gunnar Rudberg, ‘Zur palaographischen Kontraktion’, Éranos 10 (1910): 71–100; idem, Neutestamentlicher Text und Nomina Sacra (Uppsala Humamstiska Vetenskapssam fundet Skrifter, 1915 [1917]). 27.  Traube, Nomina Sacra, 36. 28.  Paap, Nomina Sacra, 119–27. 29.  Schuyler Brown, ‘Concerning the Origin of the Nomina Sacra’, SPap 9 (1970): 7–19. 30.  Ibid., 18. 31.  Ibid., 19.

7. The Origin of the Nomina Sacra

127

where they served as translation equivalents for ‫יהוה‬. Christians took up the idea and quickly broadened it to include ‘the remaining persons of their Trinity’ and then a wider list of religious terms.32 George Howard has argued that Κύριος and Θεός were the initial nomina sacra, ‘first created by non-Jewish Christian scribes who in their copying the LXX text found no traditional reason to preserve the tetragrammaton’, and may have considered the contracted forms of these words ‘analogous to the vowelless Hebrew Divine Name…’.33 All things considered, however, it seems to me that the most promising option is to follow up suggestions by Colin Roberts, as I hope to show in what follows.34 As he argues, the provenance of the manuscript evidence suggests that the nomina sacra are a Christian innovation reflecting the influence of Jewish reverence for the name of God, reverence reshaped under the impact of Christian religious convictions. The early variation in the spellings of some of the nomina sacra (contraction, suspension, combinations of both) also suggests that they were a Christian invention and not an already developed practice taken over from Jewish tradition. Moreover, I think Roberts is probably correct to posit Ἰησοῦς as the first of the nomina sacra, and I wish to explore further the reasons for this view.35 First, the early frequency of both the contracted form of Ἰησοῦς (ις and inflected forms, as in Fig. 1) and the suspended form (ιη), over against the much more regularly contracted forms of the other three early nomina sacra is consistent with the hypothesis that Ἰησοῦς began to be written reverentially before there was a more standardized scribal practice. If the nomina sacra had begun with the convention of contracted 32.  Treu, ‘Die Bedeutung des Griechischen für die Juden im römischen Reich’. 33.  George Howard, ‘Tetragrammaton in the New Testament’, ABD 6:392–93; idem, ‘The Tetragram and the New Testament’. In his assertion of a similarity between the ‘vowelless’ tetragrammaton and the contracted forms of Κύριος and Θεός, Howard is subject to the same sort of criticism that Brown directed against Traube. All Hebrew words are written ‘vowelless’, so in that respect there was nothing special about ‫ יהוה‬that would suggest similarities to contacted forms of Κύριος and Θεός. Also, of course, the contracted form of Κύριος involves not only the omission of vowels but also a consonant! Even more importantly, Howard fails to take account of the other rich evidence that the person and name of Jesus were early and widely accorded divine significance, which included an association of Jesus’ name with the divine name, as noted later in this essay. 34.  Roberts, Manuscript, esp. 35–48. I am more doubtful, however, as to his speculations about the influence of the Jerusalem church leadership and about a Palestinian provenance for the nomina sacra (pp. 42–46). 35.  Ibid., 37.

128

Texts and Artefacts

forms of Θεός and/or Κύριος, how and why would the suspended form of Ἰησοῦς have developed so early and spread so widely? The suspended spelling of Ἰησοῦς is attested in earliest sources but gives way later to the contracted spelling or another form, ιης, which is probably a conflation of suspended and contracted forms.36 Second, it may also be significant that, uniquely among all the nomina sacra, the numerical value of the suspended form of Ἰησοῦς (ιη = 18) is commented on in early Christian sources. Indeed, I suggest that this is an important clue to the origin of the nomina sacra. We know from Barn. 9.7–8 that the numerical value of the suspended form of Jesus’ name was noted by the early second century; and a passage in Clement of Alexandria appears to refer to this as already a venerable tradition in the late second century.37 Both Barnabas and Clement mention the numerical value of ιη in connection with explaining the 318 servants of Abraham in Gen 14:14 as representing the cross (the cross-shaped τ = 300) and Jesus (ιη = 18). That this is a use of gematria in exegesis suggests a Jewish-Christian tradition that is likely earlier still than the sources in which it is now preserved. Clement introduces the exegesis as if it was already venerable by his date.38 I suggest that the likely early date and Jewish-Christian provenance of the suspended form of Jesus’ name points to the possible origin of the nomina sacra in early Christian reverence for Jesus’ name. 36.  E.g., Ρ.Egerton 2 (mid/late second century CE) has the suspended form of Ἰησοῦς with contracted forms of the other earliest nomina sacra. See Η. I. Bell and Τ. C. Skeat, eds., Fragments of an Unknown Gospel and Other Early Christian Papyri (London: British Museum, 1935), 2–4, who note variation in the abbreviated forms of Ἰησοῦς and support the view that the suspended form may be the earlier, probably ‘from the Apostolic age downwards’. Ρ.Oxy. 2.210 has ιη. The conflated form ιης is found in Ρ.Oxy. 2.208, 3.402 (third century CE), 2.209 (early fourth century), 6.850, and 6.847 (fourth century CE) has both the suspended and conflated forms of Jesus’ name. By the third century, the suspended form was clearly giving way to the contracted and conflated forms. 37.  Clement, Strom. 6 278–80 (ANF 2:499), cited in Roberts, Manuscript, 37 n. 2. G. Scholem mentions a rabbinic interpretation of the 318 men of Gen 14:14 as referring to Eliezer, the servant of Abraham, the numerical value of his name in Hebrew characters being 318 (‘Gematria’, EncJud 7:370, citing b. Ned. 32a, Gen. R. 43 2, and see also L. Ginzberg, Legends of the Jews [New York: Jewish Publication Society, 1909–38], 5:224 n. 93, for further references to rabbinic literature). Scholem cogently suggests that the rabbinic interpretation here may have originated as a reply against the Christian interpretation reflected in Barnabas and Clement. 38.  Clement begins the exegesis with ‘phasin oun einai…’, which, as noted by Roberts (Manuscript, 37 n. 2), ‘suggests that ιη was no longer current [as the preferred abbreviation of Jesus’ name] in Clement’s day’.

7. The Origin of the Nomina Sacra

129

It is most unlikely that the suspended form of Ἰησοῦς was derived from this curious exegesis of Gen 14:14. More probably, the Christian gematria of the number 318 in that passage depends on a prior acquaintance with the suspended form of Jesus’ name, ιη, and an appreciation of its numerical value as 18. But what would have been the significance of eighteen? I suggest that originally it may have been intended to signify an association between Jesus and ‘life’, alluding to the numerical value of the letters of the Hebrew word for life, ‫ = ח( חי‬8; ‫ = י‬10). This gematria of ‫ חי‬is well known popularly in later Jewish tradition, but, admittedly, is not explicitly attested in sources early enough to serve as background for earliest Christian groups.39 But it is not necessary that the number 18 was already bandied about among Jews as representing ‘life’ as it is popularly in more recent times. My proposal simply requires either Christian Jews or Gentiles sufficiently acquainted with the Hebrew alphabet to be able to compute the numerical value of the Hebrew letters for ‘life’, who would then have been able to see the numerical equivalence in the Greek letters of the suspended-form abbreviation of Ἰησοῦς, prompted by their beliefs in the risen Jesus as himself powerfully alive and as life-giver.40 This ability to make gematria associations across Hebrew and Greek alphabets is not, I think, too improbable. It is, for example, commonly thought that the 666 of Rev 13:17 and 15:12 requires one to use Hebrew

39.  A check with specialists, including Philip Alexander, Lawrence Schiffman, and Shaye Cohen, failed to produce any known evidence of 18 being cited as religiously significant in any relevant early Jewish sources. But there is evidence that numbers were a subject of speculation among Jews (e.g., Philo, De Plant. 117–25; Opif. Mund. 13–14, 89–110; Quest. Gen. 1.91; 2.5; Quest. Exod. 2.87, reflecting Pythagorean traditions). Note also Sib. Or. 1:141–45; 5:12–51; 11:256, 266; 12:16–271; and Rev 13:17; 15:12 for early examples of gematria. On gematria, see Scholem, ‘Gematria’; Joshua Trachtenberg, Jewish Magic and Superstition (New York: Atheneum, 1982), 262–63; Franz Dornseiff, Das Alphabet in Mystik und Magie (Leipzig: Teubner, 1922), 91–118; and F. W. Farrar, History of Interpretation: Bampton Lectures 1885 (New York: E. P. Dutton, 1886; reprint, Grand Rapids: Baker, 1961), 98–100. For further discussion of early Jewish and Christian evidence of the religious significance of numbers, especially 8, see Reinhart Staats, ‘Ogdoas als ein Symbol für die Auferstehung’, VC 26 (1972): 29–52. 40.  The technical term for what I am suggesting is isopsephy (sometimes ‘isopsephism’), the linking/association of two or more words whose letters add up to the same numerical value (in this case, the suspended spelling of Jesus’ name and ‘life’). For discussion of isopsephy, see Dornseiff, Alphabet, 96–104, and the rabbinic examples given in Farrar, History of Interpretation, 98–100, e.g., the equation/linkage of ‘Menahem’ with the royal-messianic ‘Branch’ of Zech 3:8 (in Hebrew both = 138).

130

Texts and Artefacts

characters to obtain its meaning (Nero Caesar).41 Likewise, it is commonly accepted that the number 14 emphasized in the Matthean genealogy of Jesus (Matt 1:17) presupposes the numerical value of the name ‘David’ in Hebrew characters (‫ = ד‬4 + ‫ = ו‬6 + ‫ = ד‬4).42 The conviction that Jesus had been raised from the dead and made the Lord of life for the elect could have suggested the association between ‘life’ and Jesus’ name. Could this association be reflected in John 20:31, where the (Jewish Christian?) author expresses hope that readers will believe in Jesus the Christ and may thus ‘have life in his name’?43 In further support of this suggestion, I offer an observation about a striking feature of the Christian nomina sacra noticed by everyone who has studied the phenomenon, but not, to my knowledge, ever satisfactorily explained: the curious horizontal stroke characteristically placed over the nomina sacra. As noted already, the placement of this stroke is not really the same as the stroke that is one of the abbreviation marks used in ancient Greek sources. The stroke over the nomina sacra is, however, exactly the same as the horizontal stroke placed over Greek letters when they served as number signs.44 I propose that the familiar horizontal stroke characteristic of the nomina sacra began its career in Christian usage over the suspended form of Ἰησοῦς and was initially intended to signal to 41.  See, e.g., Richard Bauckham, The Climax of Prophecy: Studies on the Book of Revelation (Edinburgh: Τ. & Τ. Clark, 1993), 384–88. 42.  See, e.g., the discussion in R. E. Brown, The Birth of the Messiah (Garden City, NY: Doubleday, 1977), 74–81, esp. 80 n. 38. 43.  From a later period, Eusebius (Praep. 10.5) associates the Hebrew letter ḥet (the first letter of the Hebrew word for life) with ‘life’ (ζωή), an association found also in the Hypomnestikon of Joseph 26.15. For the latter, see R. M. Grant and G. W. Menzies, Joseph’s Bible Notes (Hypomnestikon) (Atlanta: Scholars Press, 1996). 44.  See, e.g., P.Oxy. 1.108, a monthly meal bill from the late second or early third century CE, which contains many examples. Blanchard states, ‘le trait horizontal signale la lettre a valeur de chiffre’ (p. 3), and notes that when thus used with a letter combination functioning as a number the horizontal stroke is placed squarely over the letter combination, whereas when a horizontal stroke is used to indicate an abbreviation (other than the nomina sacra) the stroke does not extend over the whole of the abbreviated word but is fixed over the final letter (p. 21 n. 20). For more extended discussion of abbreviation marks, including the use of the horizontal stroke, see Avi-Yonah, Abbreviations, 33–39. On the origin of the practice of representing numbers by alphabetical letters, see Karl Menninger, Number Words and Number Symbols: A Cultural History of Numbers (Cambridge, MA: MIT, 1968), 262–74. My PhD student William Parkinson kindly checked a large number of Greek papyri to verify the widespread use of this horizontal stroke to designate letters serving as numbers.

7. The Origin of the Nomina Sacra

131

readers to note the numerical value of ιη, which was, I suggest, the initial nomen sacrum. As other words were quickly added to the growing list of sacred terms to be written in special forms as nomina sacra, the original function of the horizontal stroke became irrelevant (the numerical values of these other abbreviations were not particularly significant) and the stroke functioned simply as the Christian scribal device for highlighting the abbreviations as specially sacred words. The initial numerical significance of the suspended spelling of Jesus’ name as I have described it would have resonated only with (1) Christian Jews with some acquaintance with Hebrew and/or (2) with those sufficiently interested in Jewish-Christian traditions to appreciate the numerical significance of ιη. By the time of the extant second-century sources cited earlier (Barnabas, Clement of Alexandria), it appears that there remained a memory that the suspended spelling of Jesus’ name equals 18 and that this could be employed in the exegesis of Gen 14:14, but little more as to the significance of this numerical value.45 In fact, the particular significance and association I propose for the suspended spelling of Jesus’ name may have been lost by the latter part of the first century. If the nomina sacra began with the special abbreviation of Jesus’ name, it appears that very quickly the other early-attested nomina divina (Θεός, Κύριος, Χριστός) were also accorded scribal reverence and written as sacred abbreviations. For reasons about which we can only speculate, the practice of abbreviation by contraction rather than suspension was developed and preferred for this growing list of sacred terms. Perhaps contraction may have been preferred as a standardized technique because there was a growing list of words given reverence and there was a need for some standardized abbreviation system for all these words. The convention of contraction afforded a system that permitted a variety of 45.  Irenaeus refers to Valentinian use of the numerical value of Jesus’ name written as ιη to calculate the number of a set of heavenly aeons (Adv. Haer. 1.3.2; ANF 1:319). I owe this reference to my graduate student William Parkinson. This Valentinian interpretation seems to me to reflect a later gnosticizing reinterpretation of the earlier (Jewish-Christian?) tradition that highlighted the numerical value of ιη. See the discussion of this passage by Dornseiff, Alphabet, 131. In the intricate body of numerological speculation attributed to the heretic Marcus, the numerical value of Ἰησούς fully spelled (888) is featured, as is the value of the initial iota (10) (Irenaeus, Adv. Haer. 1.15.2; cf. 1.16.2). This Marcusian system may likewise represent a much more elaborate development of simpler and more primitive Jewish-Christian gematria involving Jesus’ name. (I thank an anonymous JBL assessor of this essay for reminding me of this reference.)

132

Texts and Artefacts

sacred words (including all the inflected forms of them) to be written in a relatively standardized way that could easily be recognized once this basic abbreviation scheme was known among Christian readers of Greek. The principle was simple: (normally) first and last letter of certain important words in Christian religious vocabulary, with occasional orthographic variations (e.g., sometimes χρς for Χριστός). Roberts suggested that both suspension and contraction may have been appropriated independently of each other and at about the same time, both used initially to render Jesus’ name, the suspended form ‘owing something to number symbolism, the other, perhaps with an allusion to Alpha and Omega, taking the first and last letters’.46 If this suggestion is accepted, then the subsequent preference for contraction would have been a case of one practice being seen to be more serviceable than the other for the growing list of sacred words being written in abbreviated forms. There is, however, in my view a slightly greater chance that the practice of writing Jesus’ name in suspended form as ιη, with the horizontal stroke intended to signal the numerical value of the abbreviation, precedes the practice of sacred abbreviation by contraction. This hypothesis has the advantage of explaining the origin of the horizontal stroke over the nomina sacra, which is otherwise not accounted for in other hypotheses. The hypothesis that the name ‘Jesus’ was the first of the Christian nomina sacra is also consistent with the rich evidence of the enormous religious significance attached to Jesus’ name, and the ritual use of it in early Christian circles.47 New Testament references to Christian baptism as ‘in/into the name of Jesus’ (e.g., Acts 2:38; 19:5) indicate that the rite included the cultic pronunciation/invocation of Jesus’ name.48 In the NT, 46.  Roberts, Manuscript, 37. In his study of P66, Victor Martin suggested that both the contracted and suspended forms of Ἰησοῦς go back to the earliest stage of the practice and that the Christian sacred abbreviations began with Ἰησούς and Χριστός, followed quickly by Θεός and Κύριος (Papyrus Bodmer II: Evangile de Jean chap. 1–14 [Cologny-Genève: Bibliotheca Bodmeriana, 1956], 29). 47.  Wilhelm Heitmuller, ‘Im Namen Jesu’ Eme sprach- und religionsgeschichtliche Unter­suchung zum Neuen Testament, speziell zur altchristhchen Taufe, FRLANT 1/2 (Gottingen: Vandenhoeck & Ruprecht, 1903), esp. 128–265; Jean Daniélou, The Theology of Jewish Christianity (London: Darton, Longman & Todd; Chicago: Regnery, 1964), 147–63. 48.  On the use of Jesus’ name in baptism, see esp. Heitmuller, ‘Im Namen Jesu’, 9–127, 266–336; L. Hartman, ‘Early Baptism – Early Christology’, in The Future of Christology Essays in Honor of Leander E. Keck, ed. A. J. Malherbe and W. A. Meeks (Minneapolis: Fortress, 1993), 191–201; idem, ‘Baptism “Into the Name of Jesus” and Christology: Some Tentative Considerations’, ST 28 (1974): 21–48; idem, ‘ “Into the Name of Jesus” ’, NTS 20 (1974): 432–40.

7. The Origin of the Nomina Sacra

133

Christians can be described simply as those who ‘call upon the name of the Lord’ (e.g., Acts 9:14; 1 Cor 1:2; 2 Tim 2:22), which must refer to the ritual use of Jesus’ name in the worship setting. Both Rom 10:9 and 1 Cor 12:3 refer to this pronouncing of Jesus’ name in worship settings, and in 1 Cor 5:3–4 Paul calls for a disciplinary ritual that involves drawing on the power of Jesus’ name to punish an offending member of the church. Jesus’ name was characteristically used in early Chris­tian healings (e.g., Acts 3:6) and exorcisms (e.g., Acts 16:18; cf. Acts 19:14–17). In Phil 2:9–11, it is ‘at/in [έν] the name of Jesus’ that the universal eschatological submission is to be made; and in Rev 14:1 the elect are marked with the names of the Lamb and God.49 Among the striking indications of how Jesus’ name was regarded, we can also note the statement in the Gospel of Truth (38:6) that ‘the name of the Father is the Son’.50 In the Dialogue with Trypho (75.1–2), Justin claims that Jesus’ name is the name of God disclosed to Moses, the name borne by the angellos sent before Israel in Exod 23:20, whom Justin sees as Joshua (Gr. Ἰησοῦς). This is unlikely to be an original idea with Justin and thus may have much earlier roots, very possibly in JewishChristian circles of the first century. Tertullian bears witness to the same christological exegesis of Exod 23:20, and there seems to be an echo of this tradition also in Barnabas.51 These references all are best accounted for as remnants of a once-thriving tradition that the name ‘Jesus’ was itself of divine significance, as well as being a powerful invocation for various ritual purposes. 49.  Note also Odes Sol. 42:20, where Jesus is pictured as saying, ‘And I placed my name upon their head because they are free and they are mine’ (translation from J. H. Charlesworth, OTP 2:771). 50.  The Gospel of Truth (Nag Hammadi [NH] I,3; XII,2) 37–41 presents one of the most extended and developed treatments of the theme of the Son as God’s Name. Daniélou has described this document as containing ‘a Christology of the Name more explicit and more fully developed than any other’ ancient Christian text (Theology of Jewish Christianity, 157). For other references to Jesus’ name in Gnostic writings, see also Gos. Phil. (NH II,3) 56:3–15 and Gos. Eg. (NH III,2; IV,2) 65:9–66:24 (translations in J. M. Robinson, The Nag Hammadi Library, rev. ed. [Leiden: Brill, 1988]). 51.  Tertullian, An Answer to the Jews 9 (ANF 3:163–64). In Barn. 12.8–10, we have a reference to the Jesus/Joshua theme, with Moses’ bestowing of ‘Joshua/ Ἰησοῦς’ upon Hoshea, the son of Nun (Num 13:16) taken as prefiguring Jesus. Cf. Philo, De Mut. Nom. 121, where the Greek name Ἰησοῦς (Joshua) is referred to in very positive terms. Examples of the abbreviation of Jesus’ name as ιη in early Christian inscriptions are given by Bellarmino Bagatti, The Church from the Circumcision: History and Archaeology of the Judaeo-Christians (Jerusalem: Franciscan Printing Press, 1971), 166–69.

134

Texts and Artefacts

IV In light of the evidence for the ritual use and religious significance of Jesus’ name in early Christianity, it is easier to consider the possibility that the practice we know as the nomina sacra may have begun with scribal expressions of reverence for the name of Jesus, perhaps initially through the abbreviation of the name as ιη with the original intention of highlighting the numerical significance of the abbreviation. The horizontal stroke placed over the letters of this abbreviation originally may have been intended to serve its customary role of signalling to readers to calculate the numerical value of the letters. It is the particular advantage of this hypothesis that it accounts for the very early dating of the suspended form of abbreviation of Jesus’ name and also accounts for the origin of the curious horizontal stroke that came to characterize the nomina sacra, which is otherwise a puzzling paleographic feature. Whatever the point of origin might have been, from the earliest observable stages of development onward the nomina sacra are fascinating evidence of early Christian faith and devotion. The earliest surviving evidence shows already a developed binitarian pattern of devotion reflected in the representation of the terms Ἰησοῦς, Χριστός, Θεός, and Κύριος with terms referring to Jesus and God treated with the same sort of scribal reverence.52 If, as argued here, the nomina sacra may have originated from the early Christian reverence shown to the name of Jesus, then the initial impulse was christological, an expression of what I have termed ‘Christ-devotion’.53 As to the geographical and chronological origin of the practice, I am unable to be specific. For reasons outlined earlier, we probably have to think in terms of a first-century point of origin among Christian Jews able to appreciate the isopsephy I propose, and this seems to suggest (though it does not require) a time prior to 70 CE when we commonly suppose the influence of Christian Jews was greater than in later decades. If a necessary component of a theory of the origin of the nomina sacra is a ‘plausible context’, I suggest that the most important context is the evident devotion to Jesus that characterized early Christian groups. The piety of early Christians gives us the impetus for the scribal practice, and the probable primary reason the practice became so quickly 52.  I have referred to this inclusion of Christ with God in the devotional pattern of early Christianity as an apparently distinctive ‘mutation’ in Jewish monotheistic practice (One God, One Lord, esp. 93–124; idem, ‘What Do We Mean by “FirstCentury Jewish Monotheism”?’, SBLSP 32 [1993]: 348–68). 53.  L. W. Hurtado, ‘Christ-Devotion in the First Two Centuries: Reflections and a Proposal’, Toronto Journal of Theology 12 (1996): 17–33.

7. The Origin of the Nomina Sacra

135

and so widely embraced among various Christian groups. More detailed proposals about dates or geography would be little more than speculative and of little heuristic value.54 In appreciating the nomina sacra as evidence of early Christian devotion, it is interesting to note in passing the words that became nomina sacra and those that did not. In addition to the absence of such terms as Λόγος and Μονογενής, it is particularly curious that there are no specifically Gnostic nomina sacra (e.g., Πλήρωμα or Βυθός) and that even the Gnostic texts treat as nomina sacra words from the same list that we find in other early Christian writings. The words that did become regularly treated as nomina sacra all suggest a provenance among ‘orthodox’ Christian circles and probably with a strong sense of connection to the OT and Jewish traditions.55 The nomina sacra also illustrate powerfully how important it is for NT scholars to familiarize themselves with historical realia such as NT manuscripts and not to confine their studies to printed editions (where one is unlikely to find any hint of such things as the nomina sacra). These scribal devices are our earliest (though insufficiently appreciated) evidence of an emerging Christian ‘material culture’ that is probably to be traced back to the first century. Not until the third century do we have such things as Christian inscriptions and Christian art. But the nomina sacra, though outwardly perhaps less impressive, are a significantly earlier indication of Christian desires to register physically and visually their religious devotion. I suggest 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.56 54.  ‘Plausible context’ is the phrase of an anonymous JBL assessor who expressed frustration at my reluctance to situate the proposed origins of the nomina sacra more precisely geographically and chronologically. I share the frustration but am bound to observe what I perceive as the limits of the surviving evidence. I welcome suggestions from others who may be more perceptive than I. 55.  For a provocative study of now widely accepted views of ‘orthodox’ and ‘heretical’ versions of Christianity, see A. J. Hultgren, The Rise of Normative Christianity (Minneapolis: Fortress, 1994). 56.  In his survey of oldest Christian artifacts, Erich Dinkier chastised scholars in Christian art history and archaeology for sometimes overlooking the nomina sacra (‘Älteste christliche Denkmäler: Bestand und Chronologie’, in Signum Crucis [Tübingen: Mohr, 1967]: 134–78, esp. 176–78 [reprinted in Art, Archaeology and Architecture of Early Christianity, ed. P. C. Finney (New York: Garland, 1993), 22–66, esp. 64–66]).

Chapter 8 T h e S ta ur og ram i n E a r ly C h ri s t i a n M anuscr i pt s : T h e E a r l i es t V i s ua l R e fe r e nce to t h e C r uci f i ed J e sus ?

Among the several monograms used by early Christians to refer to Jesus, the so-called staurogram or cross-monogram, which is comprised of the Greek majuscule forms of the letters tau and rho, the vertical line of the rho superimposed on the vertical stroke of the tau, is of particular historical significance.1 The specific proposal that I shall support in the present essay is that the Christian use of this device in certain early manuscripts represents the earliest extant visual reference to the crucified Jesus, indeed, considerably prior to what is commonly thought to be the time (fourth or fifth century CE) when Christians began to portray the crucifixion of Jesus visually.2 This has significant implications well 1.  For further discussion of the ‘staurogram’, see now Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 135–54. The most important previous studies are by Kurt Aland, ‘Bemerkungen zum Alter und Entstehung des Christogramms anhand von Beobachtungen bei P66 und P75’, in Studien zur Überlieferung des Neuen Testaments und seines Textes (Berlin: de Gruyter, 1967), 173–79; Matthew Black, ‘The Chi-Rho Sign – Christogram and/or Staurogram?’, in Apostolic History and the Gospel: Essays Presented to F. F. Bruce, ed. W. Ward Gasque and Ralph P. Martin (Grand Rapids: Eerdmans, 1970), 319–27; and, more recently, Erika Dinkler-von Schubert, ‘CTAYROC: Vom ‘Wort vom Kreuz’ (1 Kor. 1,18) zum Kreuz-Symbol’, in Byzantine East, Latin West: Art-Historical Studies in Honor of Kurt Weitzmann, ed. Doula Mouriki et al. (Princeton: Department of Art and Archaeology, 1995), 29–39. 2.  I provide here further support for a point made earlier by Erich Dinkler, Signum Crucis (Tübingen: J. C. B. Mohr [Paul Siebeck], 1967), 177–78, about the historical importance of the tau-rho compendium, who in turn was seconding and amplifying observations by Kurt Aland, ‘Neue Neutestamentliche Papyri II’, NTS 10 (1963–64): 62–79 (esp. 75–79); idem, ‘Neue Neutestamentliche Papyri II’, NTS 11 (1964–65): 1–21 (esp. 1–3). I have discussed the matter more briefly in an earlier essay,

8. The Staurogram in Early Christian Manuscripts

137

beyond the area of codicology and palaeography, extending also into questions about early Christian beliefs and expressions of piety.3 Before we examine this specific proposal, however, I address some introductory and background questions and set the staurogram into an appropriate historical context. Several questions obviously present themselves. What is the historical relationship of these various Christian monograms to one another? Were some or all of them created de novo by Christians, or do they represent or include Christian appropriations of ligatures already in use? In any case, what did these devices signify and how did they function in Christian usage, especially in the earliest instances? It is not possible here to deal comprehensively with these questions with reference to all these monograms. Instead, I shall provide some limited discussion of general matters and then focus more specifically on questions about the tau-rho device. Early Christian Monograms In addition to this tau-rho combination, early Christians also made use of several other such devices to refer to Jesus Christ.4 Perhaps most well known is the chi-rho, which certainly obtained the most widespread and most long-lasting usage, down to the present time.5 Less familiar, but also certainly found in ancient Christian usage, were the iota-chi monogram (which with its six points can look like a stylized star) and also the iota-ēta (ΙΗ). In some later (post-Constantinian) instances of Christian L. W. Hurtado, ‘The Earliest Evidence of an Emerging Christian Material and Visual Culture: The Codex, the Nomina Sacra and the Staurogram’, in Text and Artifact in the Religions of Mediterranean Antiquity, ed. Stephen G. Wilson and Michel Desjardins, SCJ 9 (Waterloo, ON: Wilfrid University Press, 2000), 271–88 (esp. 279–82). 3.  Joseph van Haelst, Catalogue des papyrus littéraires Juifs et Chrétiens (Paris: Publications de la Sorbonne, 1976), is an indispensable reference work for the use of papyrological data. 4.  See the discussion of ‘Abbreviations and Monograms’ in Jack Finegan, The Archeology of the New Testament: The Life of Jesus and the Beginning of the Early Church, rev. ed. (Princeton: Princeton University Press, 1992), 352–55; and Hurtado, Earliest Christian Artifacts, 95–134. 5.  E.g., Wolfgang Wischmeyer, ‘Christogramm und Staurogramm in den lateinischen Inscriften altkirchlicher Zeit’, in Theologia Crucis – Signum Crucis: Festschrift für Erich Dinkler zum 70. Geburtstag, ed. Carl Andresen and Günter Klein (Tübingen: J. C. B. Mohr [Paul Siebeck], 1979), 539–50.

138

Texts and Artefacts

usage, we have two or more of these devices used together, as is the case with the Christian inscription from Armant (ancient Hermonthis, Egypt), at the bottom of which there is a tau-rho and an ankh flanked on either side by a chi-rho.6 As to their derivation, with the possible exception of the iota-ēta compendium, these are all pre-Christian devices and were appropriated by early Christians.7 In each case, the Christian innovation was to ascribe new meanings and significance to these devices, so that in Christian usage they referred to Jesus and reflected early Christian piety. A ‘monogram’ is an interweaving or combination of two (or sometimes more) alphabetic letters, the component-letters of the resultant device typically referring to a person’s name or title. But such letter-combinations (called ‘ligatures’ and/or ‘compendia’) can also serve other purposes, particularly as abbreviations of common words. For instance, in pre/non-Christian Greek papyri of the Roman period, the chi-rho is used as an abbreviation for several words (e.g., forms of χρόνος), and in Greek inscriptions this ligature is found as an abbreviation for ἑκατονταρχια, ἑκατονταρχης, ἑκατονταρχους, χιλιαρχης, and a few other terms.8 As well as the more familiar form of the chi-rho device, one of the two letters superimposed over the other, there are also instances where the one component letter is written above the other.9 To cite another early non-Christian 6.  Finegan, The Archeology of the New Testament, 387–88, gives a photograph and discussion. The inscription was originally published in 1892, and is thought to have been made sometime between the fourth and sixth centuries CE. 7.  In the following discussion of pre/non-Christian usage of these devices, I draw upon the following studies: Alain Blanchard, Sigles et abbréviations dans les papyrus grecs: Recherches de paleographie, Institute of Classical Studies Bulletin Supplement 30 (London: Institute of Classical Studies, 1974); Kathleen McNamee, Abbreviations in Greek Literary Papyri and Ostraca, Bulletin of the American Society of Papyrologists Supplements 3 (Chico, CA: Scholars Press, 1981); Michael Avi-Yonah, Abbreviations in Greek Inscriptions (The Near East, 200 B.C.–A.D. 1100) (repr., Chicago: Ares, 1974; originally published as a supplement to Quarterly of the Department of Antiquities in Palestine 9 [1940]). Dinkler-von Schubert (‘CTAYROC’, 33–34) also surveys the pre/non-Christian usage of the tau-rho and the chi-rho devices. The most comprehensive survey of ancient monograms known to me is Victor Garthausen, Das alte Monogramm (Leipzig: Karl W. Hiersemann, 1924), but unfortunately his discussion of earliest Christian monograms (esp. 73–79) is clearly incorrect in light of subsequently discovered evidence such as I discuss here. 8.  See, e.g., Don Pasquale Colella, ‘Les abbreviation ‫ ט‬et (ΧΡ)’, Revue Biblique 80 (1973): 547–58, who comments on the likely import of chi-rho marks on (non-Christian) amphorae. 9.  Examples cited by McNamee, Abbreviations, 118; Blanchard, Sigles et abbreviations, 26 n. 36; and Avi-Yonah, Abbreviations, 112.

8. The Staurogram in Early Christian Manuscripts

139

instance of the familiar form of this particular ligature, Randolph Richards drew attention to a chi-rho in P.Mur. 164a (line 11), a text of Greek tachygraphic writing (an early form of shorthand) on parchment which, with the other manuscripts found in Wadi Murabba’at, is probably to be dated to the Jewish revolt of 132–135 CE.10 In yet another instance, the device also appears in the margin of a hypomnema on Homer’s Iliad, dated to the first century BCE, the chi-rho here a sign for χρηστον (marking passages ‘useful’ for excerpting).11 The tau-rho combination, the focus of this discussion, appears in pre/non-Christian usage as an abbreviation for τρ(οπος), τρ(ιακας), and τρ(οκονδας).12 Among specific noteworthy instances, there is the use of this device on some coins of King Herod (37–4 BCE), the tau-rho intended to identify them with the third year of his reign.13 The iota-chi combination was an archaic form of the Greek letter psi, and was also sometimes used on Roman-era coins (probably as a numerical symbol). Moreover, there is an obvious similarity to six-pointed devices used for decoration ubiquitously in various cultures, and sometimes as stylized stars.14 The uses of the iota-ēta combination, however, in relevant surveys of the data with which I am acquainted are all Christian instances, comprising the first two letters of the name Ἰησοῦς and intended as an obvious reference to him.15 But there are similar ligatures of other letters in non-Christian Greek documentary papyri, such as the combination of 10.  E. Randolph Richards, The Secretary in the Letters of Paul (Tübingen: J. C. B. Mohr [Paul Siebeck], 1991), 40–41. The full description of the manuscript is in P. Benoit et al., Les grottes de Murabba’āt, DJD 2 (Oxford: Oxford University Press, 1961), 275–79. 11.  E. G. Turner, Greek Manuscripts of the Ancient World, 2nd ed. (London: Institute of Classical Studies, 1987), plate 58. 12.  McNamee, Abbreviations, 119; Avi-Yonah, Abbreviations, 105. 13.  Baruch Kanael, ‘The Coins of King Herod of the Third Year’, JQR 62 (1951–52): 261–64; idem, ‘Ancient Jewish Coins and Their Historical Importance’, Biblical Archaeologist 26 (1963): 38–62 (esp. 48). Use of devices involving a tau-rho ligature were also noted on items from Dura Europos, at least some instances likely craftsmen’s marks. See R. N. Frye, J. F. Gillam, H. Inghold, and C. B. Welles, ‘Inscriptions from Dura-Europos’, Yale Classical Studies 14 (1955): 123–213 (esp. 191–94). 14.  For instances and discussion, see Max Sulzberger, ‘Le Symbole de la Croix et les Monogrammes de Jésus chez les premier Chrétiens’, Byzantion 2 (1925): 394–95 (337–448), who also cites Gardthausen, 76–77. 15.  Avi-Yonah, Abbreviations, 72.

140

Texts and Artefacts

mu and epsilon (for μεγας, μερις, μετοχος, and other terms).16 So, even if the specific iota-ēta combination may have been first employed as a monogram by Christians, the Christian use of other ligatures, for example the stylized six-pointed decorative device (iota-chi) to refer to Ἰησοῦς Χριστός, may have helped to suggest this device, and, in any case, the joining of various letters to form a ligature was familiar to readers of the time, especially in documentary texts and inscriptions. As indicated already, in Christian usage, all of the monograms/ compendia in question served in one way or another as references to Jesus. Thus, the Christian appropriation of them all reflects the enormous place of Jesus in early Christian devotion, and these curious devices thereby became themselves expressions of this piety.17 The chi-rho, for example (using the first two letters of Χριστός), was a direct reference to ‘Christ’ and became one of the most familiar and widely used emblems in Christian tradition.18 The iota-chi seems to have functioned mainly as a combination of the initial letters of Ἰησοῦς Χριστός, and likewise served simply as a way of referring to him, as did the iota-ēta, which was formed from the first two letters of Ἰησοῦς.19 Moreover, it is important to note that all of these devices represent visual phenomena, and so, as reverential references to Jesus in early Christian usage, they have a certain iconographic function and significance, which should be recognized. The earliest Christian use of these devices, which takes us back at least to the late second century and quite possibly earlier, represents the emergence of what we may term a Christian ‘visual culture’. I shall return to this point later. But in the case of the Christian use of tau-rho monogram, there are also interesting distinctives that now require further attention. 16.  Blanchard, Sigles et abbreviations, 4. 17.  On early Jesus-devotion, see now, L. W. Hurtado, Lord Jesus Christ: Devotion to Jesus in Earliest Christianity (Grand Rapids: Eerdmans, 2003). On the nomina sacra as expressions of this devotion, see pp. 625–27. 18.  Note, e.g., the use of the chi-rho in the Trisomus inscription in the Catacomb of Priscilla (Rome), a prayer to God, the last line of which reads σοι δοξα εν [χρ]. For full text and discussion, see Finegan, Archeology of the New Testament, 380. For other instances, see M. Burzachechi, ‘Sull’ Uso Pre-Costantiniano del Monogramma Greco di Christo’, Rendiconti della Pontificia Accademia Romana di Archeologia, Series III 28 (1955–56): 197–211. 19.  Finegan, Archeology of the New Testament, 379–80, gives a photo and discussion of a painted sign in the Catacomb of Priscilla that appears to have an iota-ēta compendium, but in this instance the horizontal stroke extends through and beyond the letters, giving the appearance of three connected equilateral crosses.

8. The Staurogram in Early Christian Manuscripts

141

The Staurogram: Origin The first observation to make is that, whereas all of the other Christian letter-compendia that I have mentioned are true monograms, the component letters in each case directly referring to Jesus by name and/or a christological title, the tau-rho combination did not have any such function. Its component letters neither derive from, nor refer to, Jesus’ name or any of the familiar christological titles. Indeed, in Christian usage, the two component letters in this device do not appear to refer to any words at all. So what suggested the Christian appropriation of this particular lettercompendium?20 Furthermore, although the tau-rho seems later to have had some usage simply as a free-standing reverential cipher for the figure of Jesus, and/or perhaps simply as an emblem intended to signify Christian faith, what was its initial function and significance, and when might it first have appeared in Christian usage? Let us first address the question of origins. Our most important evidence, and certainly the earliest, is provided by the instances of this device in some very early Christian manuscripts.21 We may begin with Papyrus Bodmer II (P66), the extant portion of a codex of the Gospel of John (chs. 1–14 relatively well preserved, the rest of John through ch. 21 in very fragmentary condition), and dated palaeographically to ca. 200 CE.22 In this manuscript the noun σταυρος (‘cross’, three instances) and at least seven uses of forms of the verb σταυροω (‘crucify’) are written in abbreviated forms, and with the tau and rho of these words written as a compendium. In each case, the statement in which the noun or verb appears refers to Jesus’ cross/crucifixion.23 20.  Cf. Dinkler-von Schubert (‘CTAYROC’, 32), who judged the question no closer to an answer. I acknowledge the difficulty involved in being entirely precise, as the following discussion will show. But I do not think that we are entirely without clues and a likely basic association of the device in earliest Christian use. 21.  Kurt Aland has the credit for first drawing attention to this evidence in two important articles (cited above, n. 2) in successive volumes of NTS in the early 1960s. 22.  Victor Martin, Papyrus Bodmer II: Evangile de Jean Chap. 1–14 (ColognyGenève: Bibliotheca Bodmeriana, 1956); idem, Papyrus Bodmer II, Évangile de Jean, Supplément, Chaps. 14–21 (Cologny-Genève: Bibliotheca Bodmeriana, 1958); and Victor Martin and J. W. B. Barns, Papyrus Bodmer II, Supplement, Évangile de Jean chap. 14–21, Nouvelle edition augmentée et corrigée (Cologny-Genève: Bibliotheca Bodmeriana, 1962). 23.  Aland identified instances of σταυρος abbreviated and with the tau-rho in John 19:19, 25, 31, and abbreviated forms of σταυροω with this device in John 19:6 (three), 15 (two), 16, 18, (‘Neue Neutestamentliche Papyri II’ [NTS 10]: 75), and further possible cases in 19:17, 20. Cf. the instances identified by Martin and Barns in the 1962 augmented and corrected edition of chs. 14–21 of P66: forms of σταυρος

142

Texts and Artefacts

Likewise, in P75, dated to about the same time and comprising portions of the Gospel of Luke (Papyrus Bodmer XIV) and the Gospel of John (Papyrus Bodmer XV), there are further instances of the tau-rho compendium used in abbreviated forms of the same two Greek words.24 But the scribal practice in this manuscript was not so consistent. In all three cases where σταυρος appears in the extant portions of Luke (9:23; 14:27; 23:26) the word is written in an abbreviated form, and in two of these cases (9:23; 14:27) the tau-rho compendium is also used.25 In the six extant occurrences of the verb σταυροω, however, the word is abbreviated twice (23:33; 24:7), and in the other four cases is written fully (23:21 [×2], 23; 24:20). Only at Luke 24:7 in P75 is there a verb-form extant with the tau-rho compendium.26 These abbreviations of σταυρος and σταυροω (in each case with a horizontal stroke over the abbreviation) mean that the copyists in question were extending to them the special, and apparently distinctively Christian, abbreviation-practice now commonly referred to as ‘nomina sacra’.27 But, as Aland observed, on the basis of these two early, and roughly contemporary, manuscripts, it appears that the Christian practice of writing σταυρος as a nomen sacrum was somewhat more quickly and more firmly established than was the case for the verb σταυροω.28 We should also note that in the Vienna fragment of P45 (dated ca. 200–250 CE), at Matt 26:2 (the sole place where either the relevant noun or verb appears in the extant portions of the manuscript) the verb-form σταυρωθηναι (‘to be crucified’) is written in a contracted form and with in 19:19, 25, plus another one restored as ‘des plus probables’ in 19:18, and forms of σταυροω in 19:6 (two), 16, 18, plus a proposed restoration of another instance in 19:20. My own examination of the photos published in their 1962 edition enabled me to verify clear instances in abbreviated forms of σταυρος in 19:19, 25, and 31, and in forms of σταυροω in 19:6, 15, 16, and 18. 24.  Victor Martin and Rodolphe Kasser, Papyrus Bodmer XIV, Evangile de Luc, chaps. 3–24 (Cologny-Genève: Bibliotheca Bodmeriana, 1961). 25.  The statements in Luke 9:23 and 14:27 have Jesus demanding his followers to ‘take up daily’ and ‘bear’ their own cross. But in each case, there is a clearly implied reference to his crucifixion. 26.  Martin and Kasser, Papyrus Bodmer XIV, 18; Kurt Aland, ‘Neue Neutestamentliche Papyri II’ (NTS 11): 2. The extant portions of John in P75 (P.Bodmer XV) do not include any uses of σταυρος or σταυροω. 27.  For discussion and references to other key studies, see Larry W. Hurtado, ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998): 655–73 (republished as Chapter 7 of the present volume); and, more fully, Hurtado, Earliest Christian Artifacts, 95–134. 28.  Aland, ‘Neue Neutestamentliche Papyri II’ (NTS 11): 2.

8. The Staurogram in Early Christian Manuscripts

143

the tau-rho compendium.29 That is, we have three early third-century Christian manuscripts with this curious device, in all of which it is used in the same way, as part of a nomina sacra treatment of the Greek words for ‘cross’ and ‘crucify’. It is unlikely that we happen to have the very first Christian usages of the tau-rho. We must suppose that this device had already been in Christian usage for some period of time for it to have been used independently by the copyists of these three manuscripts.30 This obviously means that we should date the initial Christian appropriation of the tau-rho device at least as early as the final decades of the second century, and quite plausibly somewhat earlier. It is a very interesting question as to whether the earliest appropriation of the tau-rho was made by copyists of still earlier Christian manuscripts in references to Jesus’ cross/crucifixion, or whether there was some previous and/or wider Christian usage of this ligature, i.e., beyond its use in Christian manuscripts. Unfortunately, I know of no clear evidence to settle the matter. P45, P66 and P75 offer us the earliest extant Christian uses of the tau-rho device, and in all these cases it is used in references to Jesus’ cross/crucifixion. But we can say with some confidence that these three early manuscripts are not likely the first such uses of the tau-rho. Instead, P45, P66 and P75 offer us evidence of a Christian appropriation of the tau-rho device that (whatever and whenever its origin) was already becoming familiar in Christian circles at the time that these copyists worked. In any case, this important manuscript evidence about the Christian appropriation of the tau-rho device rather clearly means that earlier (and still echoed) views, such as the influential analysis of early Christian 29.  Hans Gerstinger, ‘Ein Fragment des Chester Beatty-Evangelienkodex in der Papyrussammlung der Nationalbibliothek in Wien (Pap. Graec. Vinob. 31974)’, Aegyptus 13 (1936): 67–72 (69). The fragment (Matt 25:41–26:39) forms part of Chester Beatty Papyrus I (van Haelst 371), 30 leaves of a codex originally comprising the four Gospels (in ‘Western’ order) and Acts. See esp. T. C. Skeat, ‘A Codicological Analysis of the Chester Beatty Papyrus Codex of Gospels and Acts (P45)’, Hermathena 155 (1991): 27–43, reprinted in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup 113 (Leiden: Brill, 2004), 141–57. Dieter Roth has recently shown another instance of the staurogram in P45: ‘P45 as Early Christian Artifact: Considering the Staurogram and Punctuation’, in Mark, Manuscripts, and Monotheism: Essays in Honor of Larry W. Hurtado, ed. Chris Keith and Dieter T. Roth (London: Bloomsbury T&T Clark, 2014), 118–29. 30.  Although these three manuscripts are dated to a roughly similar period, the differences in scribal hands and a number of other features indicate that P45, P66 and P75 must derive from three distinguishable settings, which means that the copyists likely worked independently of one another.

144

Texts and Artefacts

Jesus-monograms by Sulzberger, must be judged incorrect on a couple of important matters, and that any history of early Christian symbols must take account of this.31 Most obviously, contra Sulzberger, the Christian tau-rho monogram did not first emerge in the post-Constantinian period, and is not to be understood as a derivation from a prior Christian usage of the chi-rho.32 Instead, the evidence cited from P45, P66 and P75 gives instances of the Christian use of the tau-rho considerably earlier than datable instances of the Christian usage of the chi-rho, and well before Constantine! Indeed, as Aland noted several decades ago, to go by this manuscript evidence, the earliest Jesus-monogram appears to be the tau-rho, not the chi-rho.33 Moreover, and perhaps of equal significance, 31.  Max Sulzberger, ‘Le Symbole de la Croix et les Monogrammes de Jésus chez les premiers Chrétiens’, Byzantion 2 (1925): 337–448. A very similar schema of the evolutionary development of Christian monograms was set out earlier and briefer by Lewis Spence, ‘Cross’, in Encyclopedia of Religion and Ethics, ed. James Hastings (Edinburgh: T. & T. Clark, 1911), 4:324–30. Likewise in need of correction is the analysis by M. Alison Frantz, ‘The Provenance of the Open Rho in the Christian Monograms’, American Journal of Archaeology 33 (1929): 10–26 (esp. 10–11). 32.  Sulzberger also made several other claims that have been influential but are shown to be incorrect by the manuscript evidence: that the earliest Christian symbol for Jesus’ cross was the chi, not the tau (366), that as a general rule ‘on ne trouve ni croix, ni monogrammes de Jésus, ni representations de la Passion avant le quatrième siècle’ (371), that possibly with rare exceptions there are no direct representations of Jesus’ cross before Constantine (386), that the iota-chi is the earliest-attested Jesusmonogram, and neither the chi-rho nor the tau-rho can be dated prior to the fourth century (393). Granted, Sulzberger wrote before the Chester Beatty and Bodmer papyri were available to scholars, and he leaned heavily on inscriptional data. Based on Christian manuscripts then available, he observed that ‘Il est remarquable que, dans les papyrus chrétiens, on ne trouve ni croix ni monogramme avant le Ve siècle’ (446). But he cannot be excused entirely. Even on the basis of evidence available to him, he had reason to question his views. But, instead, he seems to have allowed what seemed to him an elegant theory to determine how to handle evidence, rather than shaping his theory to fit the evidence. To cite an important instance, in considering a Christian inscription from Egypt which ends with a tau-rho flanked by an alpha and an omega, he preferred to assume that these were added ‘après coup’ (376–77). P45, P66 and P75 now clearly confirm that this was a serious mis-judgment. The influence of his weighty article is reflected in writings of many other historians of early Christian art, e.g., C. R. Morey, Early Christian Art, 2nd ed. (Princeton: Princeton University Press, 1953), 128. 33.  I restrict attention here to the use of these ligatures, and cannot engage the wider questions about other early Christian symbols, among which fish are prominent, including the anagram ΙΧΘΥΣ (= Ἰησοῦς Χριστός Θεοῦ Υιός Σωτήρ), which probably goes back to the early third century or even earlier. On the latter,

8. The Staurogram in Early Christian Manuscripts

145

the instances of the tau-rho device in these manuscripts (i.e., in abbreviations of the Greek words for ‘cross’ and ‘crucify’ in NT passages referring to Jesus’ death), the earliest Christian uses extant, show us that this compendium was used in this early period, not simply as a general symbol for Jesus, but more specifically to refer reverentially to Jesus’ death.34 In an article on P66 and P75, Jean Savignac noted that evidence indicating the chronological priority of the tau-rho over the chi-rho rendered Sulzberger’s view of the origin of the Christian use of these two ligatures invalid, but Savignac’s own proposal seems to me no more persuasive. Based on the frequently noted Armant Christian inscription from the fourth century CE (or later) which features a tau-rho and the hieroglyphic ankh-sign flanked by two chi-rhos, he suggested that the appropriation of the tau-rho derived from its visual similarity to the ankh (the hieroglyphic significance of the latter being ‘life’), which, he further proposed, had been adopted previously, perhaps in certain Valentinian circles in Egypt.35 Savignac recognized that, in general, early Christians, especially those whose faith remained more influenced by Jewish monotheistic concerns, may have been loath to adopt a pagan religious symbol such as the ankh. But, claiming the appearance of an ankh on the final page of the copy of the Gospel of Truth in the Jung codex, and taking the widely shared view that this text derives from Valentinian circles, Savignac offered this as a basis for thinking that Valentinians may have been more ready to adopt this ancient Egyptian symbol for ‘life’, interpreting it as referring to the life given through Jesus. There are, however, major problems with Savignac’s proposals. First, his core thesis does not adequately reflect the respective dates of the evidence. The earliest verifiable Christian uses of the ankh symbol are considerably later than the uses of the tau-rho device in P66, P75 and P45.36 It is simply not sound historical method to derive the clearly see, e.g., Graydon F. Snyder, Ante Pacem: Archaeological Evidence of Church Life Before Constantine (Macon, GA: Mercer University Press, 1985), 24–26 (with further references), and esp. Franz J. Dölger, ΙΧΘΥΣ. Das Fisch Symbol in frühchristlichen Zeit (Münster: Aschendorff, 1928). 34.  Aland, ‘Neue Neutestamentliche Papyri II’ (NTS 10): 78. 35.  Jean de Savignac, ‘Les Papyrus Bodmer XIV Et XV’, Scriptorium 17 (1963): 50–55 (51). Much earlier, Gardthausen (78–79) had proposed that the chi-rho was the earliest Christian monogram, and that a subsequent Christian use of the tau-rho derived from the ankh. Both of his proposals are now refuted by the evidence of early Christian manuscripts. 36.  Aland disputed whether an ankh could really be read on the last page of the Jung Codex (‘Neue Neutestamentliche Papyri II’ [NTS 11]: 2–3). But, whatever the valid reading of this particular manuscript, the ankh symbol indisputably appears

146

Texts and Artefacts

attested Christian use of the tau-rho from a supposedly prior Christian use of the ankh, when the evidence for the Christian use of the latter device is much later. It is always a better approach to develop a theory out of the evidence, taking the dates of evidence seriously. If there was any causative relationship between the Christian appropriation of the ankh and the tau-rho, the chronological data actually make it more likely that Savignac’s proposal should be stood on its head: The appropriation of the ankh may well have resulted from its visual resemblance to the tau-rho device. In any case, the sequential relationship between the Christian appropriation of the tau-rho and the ankh is rather clearly the opposite to Savignac’s theory. There is a second problem in Savignac’s proposal, and it is not confined to him. It may be a mistake to presume that the Christian appropriation of the various Jesus-monograms must have involved one initial monogram from which subsequent Christian appropriation of the others then developed.37 It seems to me that this insufficiently examined assumption contributed to the misjudgments of Sulzberger as well as Savignac, leading them to posit their respective developmental schemes, even though the evidence did not actually suggest either one. Why should we suppose that there had to be one initial Jesusmonogram from which the others somehow developed?38 It is at least as reasonable to view the Christian uses of the various Jesus-monograms as reflecting quasi-independent appropriations of at least some of the various pre/non-Christian compendia, each of the appropriations suggested to elsewhere in the Nag Hammadi texts, particularly on the leather cover of Codex 2 and at the end of the text titled ‘The Prayer of the Apostle Paul’. Moreover, other data such as the Armant inscription mentioned above rather clearly indicate Christian appropriation of the ankh by the fourth to sixth centuries CE, and this appropriation seems not to have been particularly connected to Valentinian circles. Although some of the Nag Hammadi texts may well have originated in Greek-speaking ‘gnostic’ circles, the fourth-century Coptic manuscripts of the Nag Hammadi collection were likely prepared by monastic scribes who were certainly strongly ascetic, but not particularly ‘Valentinians’. See, e.g., the discussion by J. M. Robinson, Gen. Ed., The Nag Hammadi Library, rev. ed. (Leiden: Brill, 1988), 10–22. 37.  To be sure, I have given reasons that the nomina sacra practice may have originated with an abbreviated form of Jesus’ name, the first two letters, ΙΗ (Hurtado, ‘Origin of the Nomina Sacra’). But we should not presume that the various christograms derived from one original. 38.  Is the uncritical assumption of such a schema simply indicative of how Darwinian concepts of unilinear evolution have become so much a part of Western intellectual culture that we assume that the ‘historical’ explanation of anything must have proceeded along these lines?

8. The Staurogram in Early Christian Manuscripts

147

Christians by the perceived capability of the respective devices to express Christian faith and piety.39 As we have noted already, all of the devices in question here were in pre/non-Christian use already, and thus were readily available. All that was needed for the appropriation of any one of them was for some Christian to perceive it in a new light, seeing in it a reference to Jesus. Of course, it is in principle possible that an initial Christian appropriation of one of these compendia may have helped to stimulate Christians to seize upon others as well. But this seems to me no more than a possibility. In any case, even such a scenario does not amount to the various Jesus-monograms evolving out of an initial one. In summary of the import of the chronological data, the earliest extant Christian uses of the tau-rho are notably prior to the attested Christian usage of any of the other ligatures. This alone makes it unlikely that the Christian appropriation of the tau-rho was directly influenced by prior Christian use of any of these other devices. Indeed, the chronological data suggest strongly that the tau-rho may have been the earliest of the several ligatures that were appropriated by early Christians to refer to Jesus. Likewise, to repeat the point, the earliest Christian use of the tau-rho was probably not derived from Christian use of the ankh, for this is attested only considerably later. Also, it is significant that, in distinction from the other ligatures, the Christian tau-rho was not functionally a monogram. That is, unlike the other ligatures in question, the tau-rho was not derived from, and did not refer to, the name of Jesus or christological titles. This is a further reason for doubting that the Christian appropriation of the tau-rho ligature was derived from a supposedly prior use of one of the others. In earliest Christian usage, only the tau-rho appears as part of the nomina sacra treatment of certain words (σταυρος and σταυροω), and, in comparison with the other christograms, simply functioned differently as an early Christian symbol. Indeed, an answer to the question of how and why the Christian use of the tau-rho originated is probably connected to its earliest function. So, to this we now give further attention.

39.  By ‘quasi’ independent, I mean that the appropriation of the various devices as Jesus-monograms obviously happened among circles of Christians, who to a greater or lesser extent shared features of faith and piety. Moreover, Christians clearly made efforts to ‘network’ with other Christian circles, both locally and trans-locally. So, if any given ligature was first adopted in some circle of Christians, they may well have known of the appropriation of one or more of the other ligatures among their own or other circles of Christians.

148

Texts and Artefacts

The Staurogram: Earliest Function and Significance The difference in derivation of the staurogram corresponds to a difference in function. To reiterate an important point, unlike the other compendia, the tau-rho did not function as a direct allusion to Jesus by name or title. In the earliest instances of the tau-rho, of course, the two letters are two of those that make up the Greek words for ‘cross’ and ‘crucify’. But this in itself is unlikely to explain either the reason for the Christian appropriation of the ligature or its original Christian symbolic meaning. The earliest manuscript evidence cited earlier shows that the writing of the Greek words in question as nomina sacra did not consistently involve the use of the tau-rho ligature, which strongly suggests that the two phenomena (the nomina sacra practice and the staurogram) arose independently. A more likely approach to the origin and original function/significance of the tau-rho is readily available. We know that the Greek letter tau was invested with symbolic significance by Christians very early, specifically as a visual reference to the cross of Jesus. In the Epistle of Barnabas 9:7–9 (dated sometime 70–130 CE), commenting on the story of Abraham’s rescue of Lot with a force of 318 servants (Gen 14:14), the number represented by the use of the Greek letters ΤΙΗ, the author interprets the two letters iota and ēta (the first two letters of Ἰησοῦς) as referring to Jesus, and letter tau as a reference to (and prediction of) Jesus’ cross.40 We have other evidence confirming that the Greek letter tau was viewed by Christians in the second century CE as a visual symbol of the cross of Jesus. Indeed, Justin Martyr (1 Apol. 55) indicates that 40.  This rendering of the number in Gen. 14:14 is clearly instanced in, e.g., the Chester Beatty Genesis manuscript (Chester Beatty Papyrus IV, Rahlfs 961, fourth century CE), and was almost certainly used also in the early fragment of Genesis, P.Yale 1 (P.Yale inv. 419, van Haelst 12, variously dated from early second to third century CE). Although there is a lacuna in this fragment at this spot, the space is scarcely adequate to have accommodated the number written out in words. The likelihood that the number was written as ΤΙΗ is one of the reasons that most papyrologists take P.Yale 1 to be an early Christian copy of Genesis. On this fragment, see esp. C. H. Roberts, ‘P. Yale 1 and the Early Christian Book’, in Essays in Honor of C. Bradford Welles, ed. A. E. Samuel, American Studies in Papyrology 1 (New Haven: American Society of Papyrologists, 1966), 27–28; and the stimulating reflections by Erich Dinkler, ‘Papyrus Yalensis 1 als ältest bekannter christlicher Genesistext: Zur Frühgeschichte des Kreuz-Symbols’, in Im Zeichen des Kreuzes: Aufsätze von Erich Dinkler, mit Beiträgen von C. Andresen, E. Dinkler-v.Schubert, E. Grässer, G. Klein, ed. Otto Merk and Michael Wolter (Berlin: de Gruyter, 1992), 341–45. The way the number is written out in Greek, τριακοσιους δεκα και οκτω, would have suggested to early Greek-speaking Christians the use of the three Greek letters in question.

8. The Staurogram in Early Christian Manuscripts

149

second-century Christians could see visual allusions to Jesus’ cross in practically any object with even the remote shape of a Τ (e.g., a sailing mast with cross-beam, a plow or other tools with a cross-piece of any kind, the erect human form with arms extended, even the face with the nose extending!).41 In another fascinating passage (1 Apol. 60), Justin cites a statement from Plato’s Timaeus, εχιασεν αυτον εν τῳ παντι (‘He placed him crosswise in the universe’), which Justin appropriates as a reference to Jesus (‘concerning the Son of God’, 1 Apol. 60.1). The verb, εχιασεν, suggests a chi shape (Χ), but Justin claims (1 Apol. 60.2–5) that Plato derived the idea from a misunderstanding of the biblical account where Moses was directed by God to erect a brass object for the healing of the Israelites who had been bitten by serpents (Num 21:8–9). Justin claims that Plato misunderstood the object that Moses made as chi-shaped, when in fact it was in the figure of a cross.42 In light of his earlier comments about cross-shaped objects in 1 Apol. 55, we can say that Justin almost certainly had some Τ-shaped object in mind here as well in claiming that Moses’ brass object was ‘the figure of a cross’.43 Closer to the probable date of the manuscripts in which the tau-rho device appears, there is another significant piece of evidence. Tertullian (Contra Marcionem 3.22), citing the passage in Ezekiel where God directs an angel to mark the foreheads of the elect, takes the ‘mark’ as the Greek letter tau, and then comments as follows: Now the Greek letter Tau and our own [Latin] letter T is the very form of the cross, which He [God] predicted would be the sign on our foreheads in the true Catholic Jerusalem…44 41.  Somewhat later, Minucius Felix (Octavius 29; ANF 4:191) echoes basically the same attitude. On the history and various types of cross-symbols, see, e.g., Erich Dinkler and Erica Dinkler-von Schubert, ‘Kreuz’, in Lexikon der christlichen Ikonographie, ed. Engelbert Kirschbaum, 8 vols. (Rome: Herder, 1968–76), vol. 2, cols. 562–90. 42.  Justin says of Plato’s putative reading of the Numbers account, μήδε νοήσας τύπον εἶναι σταυροῦ ἀλλὰ χίασμα, τὴν μετὰ τὸν πρῶτον θεὸν δύναμιν κεχιάσθαι ἐν τῶ παντὶ ειπε (1 Apol. 60.5). 43.  The LXX has Moses fashion a brass serpent and place it ἐπὶ σημείου. The Hebrew has Moses place a brass serpent on a ‫‘( נס‬pole’). 44.  Contra Marcionem was written 207 CE. I cite here the translation of Tertullian in ANF 3 (pp. 340–41). The LXX of Ezek 9:4, however, has the angel directed to place a σημειον upon the foreheads of the righteous. Tertullian seems to cite the reading that is reported by Origen to have featured in the translations of Theodotion and Aquila (Origen, Selecta in Ezekiel; Migne, 3.802), which is a more literal rendering of the Hebrew (‫)תו‬.

150

Texts and Artefacts

So it seems most reasonable to see the Christian appropriation of the tau-rho ligature as connected to, and likely prompted by, this strong association of the Greek letter tau with Jesus’ cross. This certainly also fits with the fact that the earliest known Christian uses of the tau-rho device are in the special ‘nomina sacra’ writing of the words for ‘cross’ and ‘crucify’. But what is the significance of the superimposed letter rho in the Christian use of the tau-rho compendium? Many years ago, Dölger cited intriguing evidence indicating that the Greek letter rho (the numerical value of which is 100) could represent ‘good fortune’ (by ‘isosephy’ the letters in the expression επ’ αγαθα amount to 100).45 Dölger also cited a statement by the Christian teacher and hymnist Ephraem the Syrian (ca. 306–73 CE) that is of interest. The statement comes in Ephraem’s comments on the meaning of the Christian symbol that apparently comprised a tau-rho with the alpha and omega placed respectively under the left and right horizontal arms of the tau. Ephraem says that in this device we have represented the cross of Jesus (the tau, for which he says that Moses’ outstretched hands are an OT ‘type’ or foreshadowing), the alpha and omega signifying that Jesus (‘the crucified one’) is the beginning and end, and, he continues, ‘The Ρ [Greek letter rho] signifies βοηθια [= “help”], the numerical value of which is 100’.46 Dölger took Ephraem’s statement to mean that he interpreted the tau-rho device by isosephy as signifying ‘Salvation is in the Cross’ or ‘the Cross is our help’.47 This seems to me a persuasive inference. Might this be also the original meaning and function of the tau-rho device? Is this how the scribes who first employed the tau-rho in the nomina sacra forms of σταυρος and σταυροω regarded the device? Ephraem is, of course, considerably later than the time of the manuscripts that we are focusing on here, and so it is a valid question whether his numerical interpretation manifests his own fascination with such things or reflects more broadly early Christian interpretation of the tau-rho. 45.  ε–5, π–80, α–1, γ–3, α–1, θ–9, α–1 = 100. Franz Joseph Dölger, Sol Salutis: Gebet und Gesang im christlichen Altertum, 3rd ed. (Münster: Aschendorff, 1972 [1925]), 73–74, citing Artemidorus of Ephesus (‘Daldianus’, late second century CE), and an inscription from Pergamon from the time of Hadrian. 46.  β–2, ο–70, η–8, θ–9, ι–10, α–1 = 100. I translate the Greek from the citation of Ephraem in Dölger (Sol Salutis, 74 n. 2). On Ephraem, see, e.g., Kathleen McVey, ‘Ephraem the Syrian’, in Encyclopedia of Early Christianity, ed. Everett Ferguson, 2nd ed. (New York: Garland, 1998), 376–77 (with bibliography). McVey describes Ephraem as holding ‘a vision of the world as a vast system of symbols or mysteries’ (376). 47.  Dölger, Sol Salutis, 74.

8. The Staurogram in Early Christian Manuscripts

151

To be sure, we have evidence that at least some Christians in the first and second centuries engaged in isosephy. Most familiar, of course, is the number of ‘the beast’ in Rev 13:17–18, which is ‘the number of his name’.48 We should also recall the interpretation of the 318 servants of Abraham noted previously in Epistle of Barnabas (9:7–9). In an earlier publication, I have offered support for C. H. Roberts’ proposal that the nomina sacra writing of Jesus’ name as ΙΗ may have derived from an association of the numerical value of these two Greek letters (18) with the same numerical value of the Hebrew word for life, ‫( חי‬ḥet-yod).49 But even if this particular proposal is not deemed persuasive to all, it is clear that some Christians from the earliest period were interested in using numerical symbolism to express their faith.50 So, it is in principle a plausible possibility that the numerically based meaning of the rho in the tau-rho device stated by Ephraem might go back much earlier, and might even have been the originating impulse for the Christian appropriation of the device. But there are some reasons to hesitate. Precisely given the evidence of a readiness among Christian in the first few centuries to employ isosephy, it is curious that we have no hint that the tau-rho was interpreted in this way earlier than Ephraem. Moreover, there is to my knowledge no evidence that the number 100 featured in second-century Christian isosephy or that the word βοηθια was particularly prominent in Christian vocabulary of that period. Indeed, Ephraem’s strong interest in finding mystical symbols of his faith everywhere in the world and nature suggests that the numerical interpretation of his tau-rho which he proposes may be his own contribution. Most significantly, Ephraem was commenting on the Christian use of a ‘free-standing’ tau-rho device, that is, the tau-rho used on its own as a Christian symbol, such as we see in the Armant inscription cited previously.51 48.  As is well known, there is some textual variation in manuscripts of Revelation, the best supported number being 666, but some witnesses reading 616 (C and Irenaeus), and even 665 (the minuscule 2344). 49.  Hurtado, ‘Origin’, 655–69. 50.  To cite another example, the number 8 was appropriated by early Christians as a symbol for the resurrection and eschatological hopes. See esp. Franz J. Dölger, ‘Die Achtzahl in der altchristlichen Symbolik’, Antike und Christentum 4 (1934): 153–87; Reinhart Staats, ‘Ogdoas als ein Symbol für die Auferstehung’, Vigiliae Christianae 26 (1972): 29–52. 51.  Finegan, Archeology of the New Testament, 387–88. Granted, the free-standing form of the tau-rho that Ephraem comments on includes the use of the alpha and omega symbols as well, but this is only a more elaborate version of the sort of free-standing use of the tau-rho we have reflected in the Armant inscription.

152

Texts and Artefacts

But I contend that this much later free-standing use of the tau-rho is significantly different from what we have in the earliest evidence of Christian use of the device, in which it appears within texts and as part of an effort to mark off words that refer to Jesus’ cross/crucifixion. That is, in our earliest evidence of its Christian use, the tau-rho consistently appears in a crucial context as part of a text that has to do with Jesus’ death. Used as a free-standing symbol, however, a device such as the tau-rho invites, perhaps requires, some imaginative interpretation such as Ephraem offered. But used in the way that we see the device employed in P66, P75 and P45, the tau-rho takes a Christian meaning and function from the words of which it is a crucial part, and the sentences in which it is deployed. This leads us to another intriguing possibility. The tau-rho device may have been appropriated by Christians originally, not (or not simply) on the basis of numerical symbolism, but because it could function as a visual reference to the crucified Jesus. This is not an original suggestion, but was proposed earlier, notably by Aland and then supported strongly by Dinkler.52 In this proposal, the tau-rho device was appropriated initially because it could serve as a stylized visual reference to (and representation of) Jesus on the cross. The tau is confirmed as an early symbol of the cross, and it is proposed that the loop of the superimposed rho in the tau-rho suggested the head of a crucified figure. This very simple pictogram-reference to the crucifixion of Jesus fits with the simplicity and lack of decorative detail that characterize earliest Christian art. As Robin Margaret Jensen notes in her recent excellent introduction to early Christian art, the simple nature of the visual expressions of faith in the earliest material ‘suggests that communication was valued above artistic quality or refinement and that the emphasis was on the meaning behind the images more than on their presentation’.53 Commendably (and unusually among historians of early Christian art), Jensen takes note of the instances of the tau-rho device in the early papyri to which I draw attention in this essay, characterizing the combined letters as forming ‘a kind of pictogram, the image of a man’s head upon a cross’, and observing that the device ‘seems to be an actual reference to the cross of crucifixion…’54 The wider importance of this view of the tau-rho is considerable. As Dinkler put it in his enthusiastic endorsement of Aland’s study, 52.  Aland, ‘Bemerkingen’; Dinkler, Signum Crucis, 177–78. 53.  Robin Margaret Jensen, Understanding Early Christian Art (London: Routledge, 2000), 24. 54.  Ibid., 138.

8. The Staurogram in Early Christian Manuscripts

153

Mit Recht macht Aland darauf aufmerksam, dass somit das Staurogramm älter ist als das Christogramm [chi-rho]…älter als jedes ‘christliche’ Bild, älter als die christianisierte oder auch schöpferisch-christliche Ikonographie, das Zeichen der Christen für das Heilsereignis, für das Kreuz Christi ist.55

That is, if this proposal is correct (and if the common dating of the papyri in question is correct), the tau-rho represents a visual reference to Jesus’ crucifixion about 150 to 200 years earlier than the late fourthor fifth-century depictions that are usually taken by art historians as the earliest.56 Significance for Scholarship If in earliest Christian use the tau-rho is rightly referred to as a ‘staurogram’, it is a noteworthy phenomenon to be reckoned with in charting the history of earliest Christian iconography.57 As I noted in a previous publication, however, it is unfortunate that a good many historians of early Christian art are not aware of the staurogram (largely because early Christian manuscripts are not usually thought of as offering data for the study of art), and so do not take account of its import.58 But the staurogram is both important and rather unusual. In its earliest extant occurrences, it is a scribal device but entirely with a visual function, and so an iconographic phenomenon, a visual/material expression of early Christian faith/piety. Whether the tau-rho was adopted originally as a

55.  Dinkler, Signum Crucis, 178. 56.  Two Christian intaglio gems usually dated to the fourth century, and a fifth-century seal held in the Metropolitan Museum of Art in New York City are the frequently cited items. For a discussion of these items and other relevant evidence, see now Jensen, Understanding Early Christian Art, 131–41. 57.  A tau-rho written in red ink appears at the beginning of a single papyrus page containing Ps 1:1 (Ralhlfs 2116; van Haelst 84) dated initially (by A. Traversa) to the second century. Writing before the publication of the early manuscript data that I underscore here, and under the influence of Sulzberger’s thesis, C. R. Morey (Early Christian Art [Princeton: Princeton University Press, 1953], 128) rejected this dating because he was confident that the Christian use of the tau-rho did not pre-date Constantine. Morey was right to suspect the second-century date of the manuscript, but his reason was wrong! Cf., e.g., Roberts, ‘P. Yale 1’, 27–28. 58.  Hurtado, ‘The Earliest Evidence’, esp. 281–82. I cite there as an example of otherwise valuable histories of early Christian art that omit any reference to the staurogram, Robert Milburn, Early Christian Art and Architecture (Berkeley: University of California Press, 1988), but this omission is in fact typical of the genre.

154

Texts and Artefacts

pictogram of the crucified Jesus (as I tend to think), or was interpreted more along the lines of Ephraem’s numerical symbolism, either way it was a visual reference to the cross of Jesus. Moreover, this has ramifications far beyond papyrology or the history of early Christian art. On what has been the dominant assumption that visual references to Jesus’ crucifixion do not pre-date the fourth century CE, some scholars have drawn far-reaching conclusions about the nature of Christian faith/piety in the pre-Constantinian period.59 For instance, in a study of the earliest archaeological evidence of Christianity with many other positive features, Graydon Snyder emphatically denied that there was any evidence of a visual reference to Jesus’ crucifixion prior to the fourth century.60 On this basis, he then made the further dubious claim that there was ‘no place in the third century [or earlier] for a crucified Christ, or a symbol of divine death’.61 But Snyder showed no awareness of the staurogram, and so his estimate of cross-symbolism in the pre-Constantinian period is likely wrong, or at least certainly inadequate in the coverage of relevant evidence.62 We can also say, therefore, that his sweeping characterization of pre-Constantinian Christian piety/faith is equally questionable. In the earliest instances of Christian usage, the staurogram (again, whether taken as a pictogram or a numerical symbol) obviously makes reference to the crucifixion/cross of Jesus, and so (along with the abundant textual evidence) reflects an importance given to Jesus’ crucifixion in Christian faith/piety, from at least as early as the late second century. 59.  In an essay written before he became aware of the manuscript evidence of the Christian use of the staurogram, Dinkler (‘Comments on the History of the Symbol of the Cross’, Journal for Theology and Church 1 [1965]: 124–46, German original 1951) once referred to the ‘absolute dogma that the symbol of the cross makes its first appearance in the age of Constantine’ (132), and claimed an absence of archaeological evidence of cross-marks made by Christians from the first two centuries (134), reflecting, of course, the influential judgment by Sulzberger (cited above). 60.  Snyder, Ante Pacem, 26–29 (I have not yet had access to the revised edition of this work which appeared in 2003, but from reviews it appears that it does not rectify the inadequately informed view of the matter expressed in first edition). 61.  Ibid., 29. 62.  I intend no particular condemnation of Snyder, for a failure to take account of the staurogram (and of the phenomena of early Christian manuscripts generally) is, sadly, rather widely demonstrated in contemporary studies of Christian origins. In another publication I aim to help students and scholars recognize the importance of the data offered (L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins [Grand Rapids: Eerdmans, 2006]).

Chapter 9 A F res h A n a lysi s of P a p y r us O x y rh y n chus 1228  ( P22) a s A rt efa ct

P.Oxyrhynchus 1228 (Gregory-Aland P22) designates two papyrus fragments preserving portions of John 15:25–16:2 and 16:21–32. In their editio princeps of 1914, Grenfell and Hunt dated the manuscript palaeographically to the late third century CE, a dating now commonly accepted.1 Although noted in several subsequent publications and cited in textual apparatuses to the Greek NT, it is not clear that the papyrus itself has received any close examination or detailed description since Grenfell and Hunt. Moreover, the papyrus exhibits some interesting features that further warrant the fresh analysis offered here.2 Roberts and Skeat referred to it as ‘an eccentric production’, in particular for being ‘written on the verso of a roll the recto of which is left blank’, a matter to which I return shortly.3 The readings of P.Oxy. 1228 have received some attention, especially as evidence of the state and transmission of the text of John in the early centuries.4 I shall consider some variation-units later. But first, 1.  Bernard P. Grenfell and Arthur S. Hunt, eds., The Oxyrhynchus Papyri, Volume X (London: Egypt Exploration Society, 1914), 14–16. 2.  P.Oxy. 1228 is housed in the Special Collections of the University of Glasgow Library (catalogued as MS Gen 1026/13). I was able to make an autopsy analysis in early July 2014. 3.  Colin H. Roberts and T. C. Skeat, The Birth of the Codex (London: Oxford University Press, 1983), 39. 4.  E.g., Grenfell and Hunt (Oxyrhynchus Papyri, 14) judged it as showing ‘a good and interesting text’, but ‘its affinities are not strongly marked, and it does not agree at all consistently with any one of the chief authorities’ (by which they must have meant the major codices of the fourth century CE and later). P. L. Hedley, ‘The Egyptian Texts of the Gospels and Acts’, Church Quarterly Review 118 (1934): 23–39, 188–230 (205), noted general agreement with the Westcott and Hort text, and so classified it as a supporter of their ‘Alexandrian’ text of the Gospels. More recently,

156

Texts and Artefacts

and in an effort to offer a fresh contribution, I focus on the physical and visual features with a view to characterizing the manuscript as an artefact of early Christianity, and with a particular concern to see what we can infer about its intended usage.5 The Manuscript The dimensions of the two fragments are as follows: The smaller one (hereafter, fragment 1) measures 5 × 3.4 cm, and the larger one (fragment 2) measures 18.5 × 6.4 cm.6 Part of an upper margin survives on both fragments, measuring 8 mm on fragment 1 and 15 mm on fragment 2. Unfortunately, however, nothing of any other margin survives on either fragment, which makes it difficult to make precise calculations about line lengths. As noted already, on the smaller fragment we have parts of eight lines of John 15:25–16:2. On fragment 2 (the larger one) we can detect remains of 29 lines of John 16:21–32. The text is written on only one side of the papyrus, and so (as long postulated) the fragments must derive from a roll, not a codex. Moreover, it appears that fragment 2 is part of a column of text that followed the column from which fragment 1 survives. That the text is written across the fibres, however, that is, on what would normally be the outer surface of a roll, suggests that this is a portion of a re-used roll (often referred to as an ‘opisthograph’).7 There however, it has been described as displaying ‘an independent text’ by Philip Comfort, Encountering the Manuscripts: An Introduction to New Testament Palaeography and Textual Criticism (Nashville: Broadman, 2005), 62, but he offers no basis for this somewhat cryptic judgment. 5.  I have laid out this approach to manuscripts at greater length in L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). 6.  My measurement of the width of fragment 2 differs from that given by Grenfell and Hunt (5 cm) and echoed in subsequent publications. The fragment has a long upper part and a smaller lower part that also extends further to the right of the upper part. I suspect that Grenfell and Hunt measured the widest point of the upper part. I measured the distance between the right edge of the lower part and the left edge of the upper part. 7.  Some scholars reserve the term ‘opisthograph’ to refer to manuscripts in which one text is written on the inner surface and then continued onto the outer surface, and they urge that in the case of rolls originally with one text on the inner surface and another text subsequently written on the outer surface, we should perhaps simply use the term ‘re-used rolls’. But see the recent comments on the issue by George W. Houston, Inside Roman Libraries: Books Collections and Their Management in Antiquity (Chapel Hill, NC: University of North Carolina Press, 2014), 6 n. 24.

9. A Fresh Analysis of Papyrus Oxyrhynchus 1228

157

are a number of other examples of re-used rolls, for example, P.Oxy. 654, a portion of a copy of the Gospel of Thomas in Greek written on the outer side of a roll, the inner surface originally used for what appears to have been a land survey. Among other early examples there is also P. Oxy. 657, which, together with PSI 12.1292, makes up portions of Hebrews (P13), on the outer surface of a roll, the inner surface containing a copy of Livy’s Epitome. P.Oxyrhynchus 1079 (P18) is another example, a copy of Revelation on the outer surface of a roll, the inner surface containing a copy of Exodus.8 The re-use of rolls is commonly taken as reflecting the wish for a basic copy of a text for personal usage, the intended reader either uninterested in a more formal copy on fresh writing material, or unable to cover the cost of preparing one.9 But the oft-noted puzzle about P.Oxy. 1228 is that no text appears on the other side of the extant writing material, the side with horizontal fibres, i.e., the side of a roll normally used as the primary writing surface. Grenfell and Hunt posited that ‘in other parts the roll included sheets which had previously been inscribed’, and, more specifically, Kurt Aland proposed that the extant fragments derive from the end-sheets of a roll that served simply to protect it, and so were not inscribed with text on their inner surface.10 On this proposal, to complete the copying of the Gospel of John additional sheets would have had to be added, or part of another roll would have had to be re-used and pasted onto the initial one from which our extant fragments come. This is still a plausible proposal as to why P.Oxy. 1228 has text only on the side of the papyrus with vertical fibres, and I cannot offer any more compelling suggestion myself.11 8.  For a table of manuscripts of literary texts of Christian provenance commonly dated to the second or third century CE, see Hurtado, Earliest Christian Artifacts, 209–29. There is still no instance of a copy of any NT writing on an unused roll. Instead, all extant copies are either codices or a small number of re-used rolls. 9.  P. L. Hedley’s reference to re-used rolls, however, as texts written ‘on waste paper’ is perhaps a bit severe: ‘The Egyptian Texts of the Gospels and Acts’, CQR 118 (1934): 188–230 (227). 10.  Grenfell and Hunt, Oxyrhynchus Papyri, 14; Kurt Aland, Studien zur Überlieferung des Neuen Testaments und seines Textes (Berlin: de Gruyter, 1967), 114. 11.  A number of ‘magical’ texts and amulets on papyrus were written across the fibres. This may have been a convention of sorts among at least some of those who prepared such items. But there is no indication that P.Oxy. 1228 was written as an amulet. At my presentation of this essay in the 2014 Annual Meeting of the Society of Biblical Literature, Malcolm Choat and AnneMarie Luijendijk also referred me to the study by Jean-Luc Fournet, ‘Esquisse d’une anatomie de la lettre antique tardive d’après les papyrus’, in Correspondances. Documents pour l’histoire de l’Antiquité

158

Texts and Artefacts

Aland estimated that another 39–40 lines would have been required to accommodate the text of John between the last line of fragment 1 (which takes us into John 16:2) and the first line at the top of fragment 2 (which commences in John 16:21). This would have made an original column of 47–48 lines, measuring approximately 25–26 × 8–10 cm. Adding space for the upper and lower margins, he suggested a roll of about 30 cm in height, made up of papyrus sheets of about 25 cm width. Assuming that the copyist was basically consistent in the number of lines per column, Aland calculated that the entire copy of John would have required some 50 columns, the two extant fragments coming from columns 37 and 38. Estimating two columns of writing per sheet of the re-used roll, he judged that another six sheets (perhaps from another roll re-used for this task) would have been needed to complete the copying of John.12 Although he did not indicate it, integral to any such calculation is an estimate of the likely average number of characters per line. I tried to make my own estimates of original line-lengths, the number of lines per column and column height in an effort to check Aland’s calculations. In one step, I attempted to estimate the original average number of characters per line in the 29 lines of the larger fragment, which commences in John 16:21 and extends into John 16:32. The number of characters to take into account in the Nestle-Aland (27th edition) text of John 16:21–32 is about 1,034.13 If we divide this amount by 29 (the number of partially preserved lines on this fragment), this produces a rough average 35–36 characters per line. Trying the same experiment on fragment 1, however, which contains the text of John commencing in 15:25 and extending into 16:2, we have ca. 341 characters to reckon with in the full text of 15:25–16:2 as given in Nestle-Aland. If we divide this number by 8 (the number of partially preserved lines of this fragment), we get an average of 42 characters per line. I suspect that Aland used something close to my estimate of the average number of characters per line on the larger fragment (i.e., 35–36 characters per line) in forming his calculations of the likely number of lines per column and the approximate width of the columns. For if we divide the relevant number of characters in the verses that originally stood between tardive, ed. Roland Delmaire, Janine Desmulliez, and Pierre-Louis Gatier (Lyon: Maison de l’Orient et de la Méditerranée, 2009), 23–66, who notes a shift in the format of letters in/after the fourth century CE, in which they came to be written across the fibres of papyrus sheets. 12.  Aland, Studien, 114. 13.  These counts involved removing punctuation and writing forms of πατήρ, θεός, and Ἰησούς in the nomina sacra forms used in the fragments.

9. A Fresh Analysis of Papyrus Oxyrhynchus 1228

159

the smaller and the larger fragments (i.e., John 16:3–20), approximately 1,365 characters, by 35, this suggests 39 missing lines – very close to Aland’s estimate of 39–40. If, however, we use an estimate of the average number of characters per line needed to accommodate the text that originally made up the eight lines of fragment 1, i.e., 42, this produces a requirement of roughly 33 lines missing between fragment 1 and fragment 2. But it may be wiser to work with the estimated average number of characters per line based on the larger fragment, simply because we have a larger body of text on which to form an estimate.14 Perhaps, however, it is wiser still to conclude that, in the absence of any side-margin, all such calculations can only be rough estimates. As for the original height of the roll from which the fragments derive, I would judge Aland’s estimate of 30 cm as perhaps at the lower end of possibilities. By my calculations, allowing for reasonable top and bottom margins, the height of the roll may have been anything from 30 cm (a minimum) to ca. 35 cm.15 In his authoritative study of Greek literary bookrolls, William Johnson judged that in the Roman period ‘roll heights rarely fell below 25 cm or above 33 cm’, which would allow for a roll of the approximate height as estimated either by Aland or by me.16

14.  Also, at the bottom of fragment 1 we may have the now-indecipherable remains of a ninth line. If we divine the characters in John 15:25–16:2 (341) by nine, we get ca. 38 characters per line, not so far from the average of 35–36 characters per line estimated for fragment 2. 15.  The 29 lines on the larger fragment take up ca. 17 cm, or just under 0.6 mm per line. So, an estimated 47 lines for a complete column would require ca. 27.5 cm. If we then add a top margin of at least 2 cm and a bottom margin of at least 3 cm (bottom margins on rolls were typically larger than top margins), that yields an estimate roll-height of ca. 33 cm, and the top and bottom margins may have been larger. Ellwood Mearle Schofield, ‘The Papyrus Fragments of the Greek New Testament’ (dissertation, Southern Baptist Theological Seminary, 1936), estimated a roll height of ca. 38 cm, which he judged ‘quite impossible for any ancient roll’ (197). But his estimate would require top and bottom margins combining to amount to 10 cm, which seems just a bit excessive to me. (I thank Brent Nongbri for bringing Schofield’s unpublished thesis to my notice, and for commenting on an earlier draft of this essay.) 16.  William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004), 143. Schofield (‘Papyrus Fragments’, 197) briefly noted as the only other options that the fragments might derive from a copy of John very different from that we know, or from some excerpt text or lectionary text, but judged that ‘no definite conclusions can be reached when the evidence is so fragmentary as in this case’. I find the former option unlikely, and the latter likewise less probable than presuming that P.Oxy. 1228 derives from a manuscript copy of John.

160

Texts and Artefacts

The Hand I turn now to a discussion of the hand of P.Oxy. 1228, to my knowledge offering here a more detailed analysis than available previously. For example, Grenfell and Hunt briefly referred to ‘an upright informal hand of medium size’, whereas Schofield saw ‘a rather rough and heavy upright semi-uncial of small size, the work of a poor scribe’.17 Neither these scholars nor others subsequently have given much more detail. The first observation easily made is that the copyist was rather unsuccessful at producing a bi-linear script. To turn to a more detailed description of the letters, they measure 3–3.5 mm for the most part, except for the rho in particular, which measures 3.5–4 mm, and whose vertical stroke descends well below the line of other letters, and also the iotas, which often extend either below or above the other letters. This fits well within the average size of letters on other third-century Christian papyri, and I would not characterize the script as small. Compare, for example, P.Oxy. 1, a codex leaf containing part of the Gospel of Thomas, which has majuscule letters mainly of ca. 2 mm height (initial letters of lines slightly larger, and ‘descenders’ noticeably larger still), or P.Oxy. 654, which is also a portion of a re-used roll and a portion of another copy of the Gospel of Thomas, with majuscule letters varying from ca. 1.6 mm (most letters) to 3 mm (descenders such as upsilon), and the initial letters of lines ca. 3–4 mm.18 The letters of P.Oxy. 1228 are also somewhat irregular in size and formation. Compare, for example, omicrons in fragment 1 line 2 and fragment 2 line 1 with fragment 1 line 3 and fragment 2 lines 20 and 27. For a striking example of variation in letter size, compare the shape and position of the two upsilons in fragment 1, line 5. The theta is written with an exaggerated cross-stroke, and is sometimes rounder and other times more oval. The mid-stroke of the nu often strikes the right-vertical stroke 17.  Grenfell and Hunt, eds., Oxyrhynchus Papyri, 14; Schofield, ‘Papyrus Fragments’, 197. 18.  For further discussion, see L. W. Hurtado, ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’, in Das Thomas­ evangelium: Entstehung – Rezeption – Theologie, ed. Jörg Frey, Enno Edzard Popkes and Jens Schröter, BZNW 157 (Berlin: de Gruyter, 2008), 19–32 (republished as Chapter 10 of the present study); and now also AnneMarie Luijendijk, ‘Reading the Gospel of Thomas in the Third Century: Three Oxyrhynchus Papyri and Origen’s Homilies’, in Reading New Testament Papyri in Context / Lire les papyrus du Nouveau Testament dans leur contexte, ed. Claire Clivaz and Jean Zumstein (Leuven: Peeters, 2011), 241–67.

9. A Fresh Analysis of Papyrus Oxyrhynchus 1228

161

a bit up from the bottom (e.g., frag. 1 line 1, frag. 2 lines 3, 4, 5, 6, 21), but compare the instance in fragment 1 line 3. The alpha was formed of one continuous movement, tending in a cursive direction. The epsilon varies slightly in size, and typically has an elongated mid-stroke that sometimes connects it with the following letter. The rho has a curiously small head, and, as already noted, the vertical stroke descends well below the bottom line (e.g., frag. 1 lines 1 and 3). The iota likewise is typically a bit larger than other letters, often extending above the upper line (e.g., frag. 2 lines 18, 19, 21, 27), but sometimes below the bottom line (frag. 2 line 3). The cross-stroke of the eta intersects the two vertical strokes somewhat above the mid-point, and there are sometimes hooks on these vertical strokes. There are also hooks often on the vertical strokes of the nu and the pi. The upsilon is formed in one pen movement, looking sometimes like a V. The delta is formed in a threestep movement without lifting the pen (frag. 2 lines 2, 7), and has a small hook at the top of the right-downward stroke. The Kappa is often slightly larger than adjacent letters, the lower, right-downward stroke elongated (e.g., frag. 1 lines 4–5, 19, 29). The pi has a hook in the foot of the rightvertical stroke, and the cross-stroke is written with a slight wave shape (e.g., frag. 1 line 4, frag. 2 line 5). Sigmas usually have a rounded top (frag. 1 lines 6–7, frag. 2 lines 3, 6, 27), but sometimes a more flattened and extended top (e.g., frag. 1 line 4, frag. 2 line 15). The top stroke of the tau is often written slightly lower than surrounding letters and the vertical stroke extends below them (e.g., frag. 1 lines 6–7, frag. 2 lines 2–3, 8–9). There are numerous instances of connected letters.19 For example, the top-stroke of a gamma is extended to become the mid-stroke of an epsilon in fragment 1 line 1, and instances of an alpha and a following iota are written without lifting the pen (frag. 1 line 5, frag. 2 line 6). In fragment 1 line 6, we see the top of a sigma extended to become the top-stroke of a tau and then also the mid-stroke of an epsilon (frag. 1 line 6), producing a three-letter connection! In fragment 1 line 7, the right-stroke of a lambda connects with the following iota, and in the same line the cross-stroke of a theta becomes also the cross-stroke of the following eta, and then the top-stroke of a tau becomes the mid-stroke of the following epsilon. In fragment 2 line 1, the letters of ὅταν appear to be connected, forming a cursive-like construction, and in line 3 the mid-stroke of an epsilon continues upward to form the following iota. In line 4 the cross-stroke of a theta becomes the cross-stroke of the following eta. 19.  In personal communication, AnneMarie Luijendijk informed me that in Medieval Studies connected letters are sometimes called ‘kissing letters’. Though that is a more colourful expression, I shall use the more bland ‘connected letters’ here.

162

Texts and Artefacts

The nomina sacra forms are πρς (frag. 1 lines 2–3, frag. 2 line 15), πρα (frag. 2 line 17), ανος (frag. 2 line 4), and ιη[ς] (frag. 2 line 27, the vertical fibre that likely had the sigma torn off after the eta). There is no punctuation, but we have instances of a diaeresis, over the upsilon of ὑμ[εῖς] in fragment 1 line 5, and in fragment 2 over the upsilon of ὑμῶν in line 6, and (likely) over the upsilon of ὑμ[εῖς] in line 18.20 Overall, these numerous instances of connected letters, the informality and even irregularity in letter-formation, the absence of punctuation, and the obvious inability at bi-linearity combine to give the impression of a copyist capable of a certain level in what we may term ‘everyday’ writing, but unable to produce the sort of simple book hand that we see in manuscripts such as P.Bod. XIV–XV or even P.Chester Beatty II (or uninterested in doing so).21 To re-state a judgment explored a bit more fully later in this study, these features of the script, combined with the likelihood that we have here a re-used roll, strongly suggest an informal copy of John quickly prepared for personal reading/study. The Text Given the small amount of text that survives in P.Oxy. 1228, it is perhaps unwise to try to align it with any particular textual cluster.22 Nevertheless, there have been various attempts to characterize its textual affiliation. In his survey of early witnesses to the text of the Gospels and Acts from Egypt, for example, Henry Sanders ascribed to the manuscript a ‘Western’ textual alignment.23 But P. L. Hedley countered that in eight out twelve variation-units in the Grenfell–Hunt transcription, it agrees with

20.  Grenfell and Hunt (Oxyrhynchus Papyri, 15) also saw a diaeresis over the upsilon that formed part of a proposed ὑμεῖν in fragment 2 line 10. They were probably correct, but at present there is at most only a trace of the letter and possible diaeresis, perhaps due to fading since their examination of the papyrus. 21.  In comparing the hands of several other third-century NT papyri with that of P.Oxy. 1171 (P20), Schofield described the hand of P. Oxy. 1228 as ‘decidedly cruder’ (‘Papyrus Fragments’, 190). 22.  Recognizing current differences over whether the term ‘text type’ remains meaningful, I adopt the term ‘textual cluster’ from a major essay by Eldon Jay Epp, ‘Textual Clusters: Their Past and Future in New Testament Textual Criticism’, in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, 2nd ed. (Leiden: Brill, 2013), 519–77. 23.  Henry A. Sanders, ‘The Egyptian Text of the Four Gospels and Acts’, HTR 26 (1933): 77–98 (esp. the table on 90).

9. A Fresh Analysis of Papyrus Oxyrhynchus 1228

163

the Alexandrian textual tradition.24 In their introduction to NT textual criticism, the Alands characterized P22 as ‘at least normal text’, by which phrasing they mean a kind of text with ‘a more or less distinct tendency toward the “strict” text’ (a ‘strict’ text being one that ‘reproduced the text of its exemplar with greater fidelity’).25 In a subsequent study, Barbara Aland characterized the text of the papyrus as ‘on the whole excellent, because it almost always reproduces the hypothetical Ausgangstext’.26 On a scale ranging from papyri reflecting a careful and exacting copying of the Ausgangstext of John (P39) to those papyri reflecting a much freer practice (P95, P80, P45), she placed P22 (P.Oxy. 1228) with a number of others ‘Dazwischen’ (‘in between’).27 Let us now consider variation-units where I found the text of P.Oxy. 1228 to be readable.28 I list them below, in each case giving first the reading preferred in Nestle-Aland (27th edition) with major witnesses that support it, followed by other variant-readings and their principal witnesses. 15:25

ἐν τῷ νόμῷ αὐτῶν γεγραμμένος (P22vid ‫)*(א‬2 B D L Ψ et al.): γεγραμμένος ἐν τῷ νόμῷ αὐτῶν A Θ f13 M ¦ ἐν τῷ νόμῷ γεγραμμένος P66vid.

24.  Hedley, ‘Egyptian Texts’, 205. 25.  Kurt Aland and Barbara Aland, The Text of the New Testament, trans. E. F. Rhodes (Grand Rapids: Eerdmans, 1987), 93, 95, 99. 26.  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, STAR 8 (Assen: Royal van Gorcum, 2003), 19–38 (20), translation mine. ‘Ausgangstext’ = the earliest attainable form of the text from which all extant copies ultimately derive. She considered twelve variation-units where the text of P.Oxy. 1228 can be checked. 27.  Aland, ‘Der textkritische Nutzen’, 37. 28.  The Grenfell and Hunt transcription also posits a number of partially readable letters in fragment 2 lines 15–29 for which I could not detect sufficient remaining traces now. I have, thus, included here only readings of P.Oxy. 1228 that can be verified with reasonable confidence at present either by autopsy inspection or by use of good-quality photos, e.g., from the Center for the Study of New Testament Manuscripts, http://www.csntm.org/Manuscript/View/GA_P22. So, I list ten variation-units, whereas Grenfell and Hunt’s transcription offered previous scholars a basis for twelve. But I doubt that this will make any substantial difference to the characterization of the text.

164

Texts and Artefacts 15:26

ὁταν ἐλθῃ (P22 ‫ א‬B Δ et al.): ὁταν δὲ ἐλθῃ A D K L Γ Θ Ψ M et al.

16:21

Θλίψσεως (P22 ‫ א‬B et al.): λύπης D 579 pc.

16:22

οὖν νῦν μὲν λύπην (P5 P22vid P66 ‫א‬2 B C* D L W Ψ et al.): νῦν μὲν οὖν λύπην ‫ ¦ *א‬οὖν λύπην μὲν νῦν A C3 Θ M ¦ μὲν οὖν λύπην νῦν f13.



ἐχετε (P22 ‫ *א‬B C f13 M et al.): ἐξετε P66 ‫א‬2 A D N W* Θ Ψ et al.



αἰρει (P22 et al.): ἀρεῖ P5 B D* Γ et al. ¦ ἀφαιρεῖ W.

16:23

ἀν τι (P5 Β C L et al.): ὁ τι (οτι?) (ε)αν P22vid A D2 W et al. ¦ ὁτι ὁ (ε)αν ‫ א‬Θ 33 et al. ¦ ὁτι ὁσα (ε)αν f1 f13 M.

16:28

ἐξῆλθον παρὰ τοῦ πατρὸς (P5 P22vid ‫ א‬A C2 Θ f1 f13 M): omit D W et al. ¦ ἐξῆλθον ἐκ τοῦ πατρὸς B C* L Ψ 33 et al.

16:32

ἐλήλυθεν (P22vid P66 ‫א‬2 A B C* D* L W et al.): νυν ἐλήλυθεν C3 D1 Θ Ψ f1 f13 M et al. ¦ ἐλήλυθεν ἡ ὠρα ‫*א‬.



κἀμὲ (‫ א‬B et al.): καὶ ἐμὲ P22vid A C3 D W M et al.

From such a small number of variation-units, it is perilous to posit any strong textual alignment. In eight of these ten variation-units, however, P22 (P.Oxy. 1228) bears the readings preferred also in Nestle-Aland as more likely those of the ‘Ausgangstext’. In six instances, P22 agrees with principal ‘Alexandrian’ witnesses (‫ א‬and B), in two of these with support also from D (whether original hand or correctors) and some other witnesses. In three variation-units P22 agrees with ‫ א‬and B against D and some other witnesses, and in two other cases it agrees with D against ‫א‬ and B. As to the nature of these variation-units, they mainly involve differences in word-order and minor stylistic preferences. It is worth noting that in at least some clear cases, P22 supports what some ancient readers may have judged to be less elegant syntax (e.g., 15:26; 16:32). Likewise, note P22’s support for present-tense verbs in 16:22, in contrast with the futuretense preferences of some other witnesses, the latter likely reflecting readers’ efforts to smooth out the sense of the text. In at least one other case (16:21), we likely see a variant that represents a harmonization with the immediate context, the λύπη in Bezae (D) and some other witnesses

9. A Fresh Analysis of Papyrus Oxyrhynchus 1228

165

perhaps reflecting the use of the same word to describe a woman’s labour pains earlier in the verse. In this case, note that P22 does not support this variant, but instead has what might have seemed to some ancient readers a slightly more unusual term to describe labour pains, θλίψις. In sum, then, with due allowance for the very limited body of text extant for analysis, P.Oxy. 1228 (P22) exhibits a text of John that is by no means aberrant, but, instead, seems on the whole one made up of readings that enjoy support from other important witnesses. Moreover, in several cases where we may see efforts of ancient readers to smooth out minor stylistic matters in some other witnesses, P.Oxy. 1228 supports what we could describe as the more ‘difficult’ readings. So, Barbara Aland’s characterization of the text of this papyrus cited earlier as in general reflecting a concern for faithful copying seems to me valid. Implications Let us now attempt to draw together results of this analysis. From our examination of the physical and visual properties of P.Oxy. 1228, it is rather clear that we have remnants of a copy of John made for someone’s personal reading, not a copy intended for public/liturgical usage. Moreover, that it may be a re-used roll combines with the informal nature of the hand to suggest someone either uninterested in a more elegant copy or unable to acquire or produce one. So, we may surmise that this may have been a reader of limited financial means, but with a strong interest in having his/her own copy of John for personal reading/study. If we entertain this possibility, then P.Oxy. 1228 (along with other papyri of the period) may give us artifactual evidence of a body of sub-elite Christians in the third century who, though of limited financial resources, were keenly interested in reading texts such as the Gospel of John.29 To posit that many Christians in that period were sub-elite is nothing new; but papyri such as P.Oxy. 1228 may comprise one kind of artefact of their participation in, and commitment to, their faith, with a particular interest in having direct access to texts such as John. Given the sub-elite nature of P.Oxy. 1228, it is all the more interesting that it exhibits a copy of John that seems to reflect a concern for faithful transmission of it the text. Granted, there is not much text in P.Oxy. 1228 29.  For another attempt to probe the physical/visual features of early Christian biblical manuscripts for implications about their intended usage, see 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), 49–62 (republished as Chapter 6 of the present volume).

166

Texts and Artefacts

on which to make a judgment, but what we have gives us no indication of a free-wheeling attitude toward the transmission of the text of John. In the variation-units where the papyrus is extant, we have noted that it does not generally support the readings favoured in some other manuscripts that seem to have arisen as stylistic ‘improvements’ or corrections. Instead, we appear to have a product that reflects a concern simply to have a copy of the text for the personal edification of some third-century Christian, apparently someone more concerned with the content of John than with what some other readers may have seen as minor stylistic infelicities. This may be significant. The papyrus seems to have been produced under ‘informal’ conditions, not for liturgical/formal usage, and so not likely under the supervision of any ecclesiastical authority. So, the quality of the text of this privately produced copy of John is all the more interesting in witnessing to the care with which at least some Christians handled the transmission of this text in the early period from which P.Oxy. 1228 derives.

Chapter 10 T h e G r eek F r a g me nts of t he G os p el of T h om a s as A rt e fact s : P a py r ol og i ca l O b s ervat i ons on P a p y rus O x y rh y n chus 1, P a p y rus O x y rh y n chus 654, a n d P a p y r us O x y r h ynchus 655

In the voluminous body of publications on the Gospel of Thomas (hereafter GThomas), and in surveys of this mass of scholarship as well, there has been a concentration on a host of questions of an exegetical and traditioncritical nature. It is the text that has drawn the overwhelming attention of scholars, whether considering the Nag Hammadi Coptic version or the very fragmentary Greek witnesses.1 This is perfectly understandable; but I suggest that in all this activity we have not given sufficient attention to the actual manuscripts themselves as artefacts that may give us valuable information about this intriguing collection of sayings ascribed to Jesus. A ‘text’ is, in one obvious sense, a body of cognitive content in written form, whether the content is intended to inform, inspire, entertain, exhort, or befuddle. But texts really acquire their life when made available to 1.  Key surveys of scholarship on GThomas include the following: Francis T. Fallon and Ron Cameron, ‘The Gospel of Thomas: A Forschungsbericht and Analysis’, ANRW 2.25/6: 4195–4251; G. J. Riley, ‘The Gospel of Thomas in Recent Scholarship’, CR:BS 2 (1994): 227–52; Stephen J. Patterson, ‘The Gospel of Thomas and the Synoptic Tradition: A Forschungsbericht and Critique’, Foundations and Facets Forum 8 (1992): 1–54; and Philip Sellew, ‘The Gospel of Thomas: Prospects for Future Research’, in The Nag Hammadi Library after Fifty Years: Proceedings of the 1995 Society of Biblical Literature Commemoration, ed. John B. Turner and Anne McGuire (Leiden: Brill, 1997), 328–46. Riley (227) noted that already by 1985 there were more than 600 books and articles by scholars on GThomas, and that by the date of his writing ‘that number is greatly increased’. David M. Scholar’s inventory of publications on all the Nag Hammadi texts, Nag Hammadi Bibliography 1948–1969, NHS 1 (Leiden: Brill, 1971), has been updated annually in Novum Testamentum.

168

Texts and Artefacts

readers (copied) and used (by readers and those to whom texts are read). A full historical approach to ancient texts, therefore, should include attention to how they were handled, used and understood, how they functioned in the ancient setting(s). Manuscripts are direct artefacts of the actual transmission and usage of texts, and can give us hints about the real-life ways that texts functioned, at least for those who copied and used the manuscripts that survive. This is the broad emphasis that I have elaborated in another publication, and in this essay I offer some illustration of it with reference to the extant portions of the three Greek manuscripts of GThomas.2 Of course, a number of previous scholars have given attention to these intriguing items. In the main, however, these studies have concentrated on textual concerns, such as the correct restoration of the often lacunose lines, and/or comparing the similarities and differences between the Greek fragments and the Coptic version.3 Indeed, so far as I am aware, the only previous, sustained papyrological analyses of the three Oxyrhynchus manuscripts of GThomas were done by the original editors and a very few others in the early decades after their original publication.4 In this essay, 2.  L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). 3.  Among such studies, I note in particular J. A. Fitzmyer, ‘The Oxyrhynchus Logoi of Jesus and the Coptic Gospel according to Thomas’, TS 20 (1959): 543–51, reprinted in Fitzmyer, Essays on the Semitic Background of the New Testament (Missoula: Scholars Press, 1974), 355–433 (which is cited here); Otfried Hofius, ‘Das Koptische Thomasevangelium und die Oxyrhynchus-Papyri Nr. 1, 654 und 655; II’, Evangelische Theologie 20 (1960): 189–92; Robert A. Kraft, ‘Oxyrhynchus Papyrus 655 Reconsidered’, HTR 54 (1961): 253–62; M. Marcovich, ‘Textual Criticism on the Gospel of Thomas’, JTS 20 (1969): 53–74; H. A Attridge, ‘Appendix: The Greek Fragments’, in Nag Hammadi Codex II,2–7 together with XIII,2*, Brit. Lib. OR. 4926(1), and P. Oxy. 1, 654, 655, ed. Bentley Layton, NHS 20 (Leiden: Brill, 1989), 96–128; and now April DeConick, ‘Corrections to the Critical Reading of the Gospel of Thomas’, VC 60 (2006): 201–8. There is now a useful Internet site offering a critical transcription and translation of the Greek fragments: http://www.gospels.net/ thomas/. 4.  The key descriptions, on which virtually all subsequent work depends, are these: Bernard P. Grenfell and Arthur S. Hunt, ΛΟΓΙΑ ΙΗΣΟΥ: Sayings of Our Lord from an Early Papyrus (London: Egypt Exploration Fund, 1897); idem, New Sayings of Jesus and Fragment of a Lost Gospel from Oxyrhynchus (London: Egypt Exploration Fund, 1904); idem, eds., The Oxyrhynchus Papyri, Volume I (London: Egypt Exploration Fund, 1898); and idem, eds., The Oxyrhynchus Papyri, Volume IV (London: Egypt Exploration Fund, 1904), 1–28. Fitzmyer (‘The Oxyrhynchus Logoi’, 420–33) aimed to give a complete listing of bibliography on the Oxyrhynchus fragments of GThomas

10. The Greek Fragments of the Gospel of Thomas as Artefacts

169

I draw gratefully upon these crucial studies, supplemented by my own opportunities to conduct an autopsy analysis of P.Oxy. 1, and P.Oxy. 654.5 I believe that on some matters I am able to supplement in greater detail previously published descriptions of the manuscripts. But my principal aim here is to point to physical features of the three Greek manuscripts that may be more significant and heuristically useful than has been recognized. I shall consider the three Oxyrhynchus manuscripts individually, in the order of their original identification, and then draw some general conclusions. P.Oxyrhynchus 1 P.Oxyrhynchus 1 is a portion of a single papyrus leaf from a codex.6 The extant portion (ca. 15 × 9.5 cm) comprises 21 lines of incompletely preserved text (each line comprising 15–17 letters) that roughly corresponds to sayings 26–30 (+77b), and 31–33 of the Coptic text. On the assumption that the extent of the Greek text basically corresponded to what we have in the Coptic, Fitzmyer posited that there were originally 37–38 lines of text on each page of the manuscript.7 The single column

published by ca. 1973. In this vast body of work, there is scarcely any focus on papyrological questions of the sort I explore here. Fitzmyer’s otherwise very helpful study is illustrative of my point. He makes only brief comments about the physical features of the Oxyrhynchus manuscripts, and focuses almost entirely on establishing the reading of the sayings that they contain. See also the concise descriptions and bibliographies in Joseph van Haelst, Catalogue des papyrus littéraires juifs et chrétiens (Paris: Publications de la Sorbonne, 1976), #593 (P.Oxy. 654), #598 (P.Oxy. 1), and #595 (P.Oxy. 655). More recently, AnneMarie Luijendijk, ‘Reading the Gospel of Thomas in the Third Century: Three Oxyrhynchus Papyri and Origen’s Homilies’, in Reading New Testament Papyri in Context / Lire les papyrus du Nouveau Testament dans leur contexte, ed. Claire Clivaz and Jean Zumstein (Leuven: Peeters, 2011), 241–67, considers the items and engages the present study as originally published. 5.  In a visit to Oxford in July 2005, I was able to examine a number of early Christian papyri in the Bodleian Library and the Sackler Library, including P.Oxy. 1 and P.Oxy. 654. In the original form of this essay I had to rely on a high-quality photograph of P.Oxy 655. Subsequently, however, I was able to examine the manuscript in the Houghton Library, and so the present discussion is based on direct examination of all three Greek items. I thank the authorities at all these libraries for their cooperation in my research. 6.  The leaf is held in the Bodleian Library (Oxford), catalogued as Bodl. MS. Gr. th. e 7 (p). 7.  Fitzmyer, ‘The Oxyrhynchus Logoi’, 355–56 n. 2.

170

Texts and Artefacts

of writing on each page measures ca. 6.5 cm width on both sides of the leaf, and it seems clear that the extant text ran from the verso side of the extant leaf (i.e., the papyrus fibres perpendicular to the lines of text) to the recto side (the fibres parallel to the writing). On both sides of the leaf, the extant top margin is about 2.8 cm. Based on the height of the extant 21 lines of text (ca. 12.2 cm), an additional 16–17 lines would require ca. another 9.7 cm, making a total column height of about 22 cm. If we also assume a bottom margin of at least 3.0 cm (bottom margins are usually somewhat greater than top margins), we can estimate a total page height of ca. 28 cm. On both sides, the outside margin measures 1.5–1.7 cm and the inside margin 1.0–1.6 cm (the inner margin on the recto side slightly narrower, and the endings of the lines somewhat less regular than on the verso side).8 So, allowing for some loss of material from both margins, we can assume an original page-width of perhaps 10–13+ cm. This sort of tall and narrow page has analogies in some other Christian papyrus manuscripts, and fits easily Turner’s proposed ‘Group 8’ of papyrus codices, one of the more common shapes among codices of the second and third centuries CE.9 On the left edge of the recto side, there is a repair strip about 3.8 cm wide. Grenfell and Hunt noted this strip, and seem to have suggested that it reflects a greater wear that accrued to the outer margins of codex pages through frequent usage. But in this instance, it is clear that the damage requiring the repair strip happened before the text was copied, for the lines of writing (all in the same hand) commence on the repair strip and extend onto the remaining portion of the page.10 It is more likely, therefore, that the particular sheet of papyrus of which the extant leaf is a portion suffered damage to this edge in the process of cutting the bi-folia for this codex. In short, the repair strip tells us nothing about how heavily

8.  The inside margin is also noticeably more ragged, suggesting that the leaf may have been torn from the codex of which it was originally a part. It does not appear simply to have fallen out. 9.  To my knowledge, this is the first attempt to estimate the original page size. See Eric G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977), 20, for a list of other codices (Christian and non-Christian) in this category, which involves a breadth of 12–14 cm and a height of 25–30 cm. Other Christian examples of the same approximate date include P.Chester Beatty II (P46; 13.5–15.2 × 26.5–27 cm) and P.Oxy 208+1781 (P5; ca. 12.5 × 25 cm). In his table of codices surveyed, Turner gives no estimated page-size for P.Oxy. 1 (Typology, 143). 10.  Cf. Grenfell and Hunt, ΛΟΓΙΑ ΙΗΣΟΥ, 7.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

171

the text may have been used.11 There is, however, considerable abrasion to the beginnings of lines on this side, resulting in the loss of one or two letters from some. But this abrasion probably happened in the course of the fragment lying in the ground for many centuries. As for the scribal hand, it is clear and competent, but workaday and certainly not calligraphic, and a dating to the early third century CE is commonly accepted.12 For instance, it is clearly not bilinear, and there is considerable variation and obvious inconsistency in letter sizes (cf., e.g., the different sizes of the epsilons in verso lines 1 and 3, the bēta in lines 2 and 7, and the sigma in lines 1 and 2).13 Other features, such as ligatures, confirm a copyist of very limited aesthetic abilities. The majuscule letters are mainly ca. 2 mm in height, initial letters of lines often slightly larger (ca 3 mm), and ‘descenders’ noticeably larger still (the phi ca. 6 mm, the psi 4.5 mm). In general, however, the size of the writing tends toward the smaller end of the spectrum of hands usual for a page of this size, and this also means a somewhat larger number of lines per page, ca. 37–38, than more typical of Christian literary manuscripts. To cite two roughly contemporary examples of codices of similar page shape and size, P.Oxy. 208+1781 (P5, fragments of John) had 27 lines per page, and P.Chester Beatty II (P46, Pauline epistles) had 25–28 lines per page. If, as is usually assumed, fewer lines per page and somewhat larger letter-sizes reflect a copy prepared for ease of reading (perhaps public reading in particular), then the smaller-size letters and somewhat greater number of lines per page may signal that P.Oxy. 1 was copied more for personal reading/ usage.

11.  I thank Dr Nick Gonis of the Sackler Library (Oxford) for discussing the matter with me during my visit to Oxford in July 2005 and confirming this view of the likely occasion for the repair. I assume that the person cutting and preparing the bi-folia for copying cut the horizontal fibres vertically in a straight line, which then permitted an even strip of vertical fibres to be pasted onto the damaged area. 12.  Cf. Attridge, ‘Appendix’, 96–97, who describes the writing as ‘rather flattened capitals of medium height’, and ‘an informal literary hand’ with analogies in other late second and third-century CE papyri. I would say that among papyri of this period and page-size, the letters are somewhat on the small side, and that the hand is unskilled at literary quality, rather than ‘informal’. 13.  ‘Bilinear’ means that the letters in lines all are written evenly in size and placement so that they fit within two imaginary parallel lines. For an excellent entreé into papyrology, see Eric G. Turner, Greek Papyri: An Introduction (Oxford: Clarendon, 1980).

172

Texts and Artefacts

There are several scribal devices worth noting. As observed already, the copyist was clearly concerned to make the right-hand ends of lines on the verso more regular than on the recto side, likely because on the verso side the lines ended on the outer edge of the page, where irregularities would be more noticeable.14 In an effort to achieve this visual regularity of line endings, the scribe used a line-filler mark (lines 3, 9, 17, 18), whereas no such mark appears on the recto side. The nomina sacra forms (all with the usual supralinear stroke) are ΙΣ (verso lines 5, 11; recto lines 23 [restored], 30, 36, and 41), ΘΥ (verso line 8), ΑΝΩΝ (verso line 19), ΠΡΑ (verso line 11), and ΠΡΙΔΙ (recto line 32). A diaeresis (functioning as a rough breathing mark) appears over the initial upsilon of ΥΙΟΙΣ (verso line 19), and a supralinear stroke is used at the end of some lines in place of a final nu (verso lines 10, 16; recto lines 27, 35). There is, however, no punctuation, and I cannot detect any clear use of spacing to signal sense units. There are some itacisms, νηστεύσεται (verso lines 5–6), ευρηται (verso line 7), σαρκει (verso line 13), δειψω(ν) (verso line 16), and γεινωσκοντας (recto line 35), and a clear correction of πτωχια(ν) (recto line 22, an epsilon written above the iota). However, οικοδομημενη is uncorrected in lines 36–37 (recto).15 The number on the upper right-hand corner of the verso side (ΙΑ = 11) is very interesting. Grenfell and Hunt suggested that it was ‘usual to foliate the right-hand pages’ of a codex, which would confirm further that the text on the verso preceded the material on the recto side.16 But, in fact, it is not so easy to be sure what the number represents.17 Numbering can function to order the sheets of papyrus, which were folded to form ‘bi-folia’ (i.e., two leaves, or four pages of writing space), in order to ensure their correct order for assembly to form the codex (particularly in single-gathering codices). But sometimes leaves, not sheets, are numbered, and in other cases numbers identify gatherings (quires), especially, of course, in the 14.  This was first observed by Grenfell and Hunt, ΛΟΓΙΑ ΙΗΣΟΥ, 6–7. 15.  Here, too, I reiterate items mentioned by Grenfell and Hunt, ΛΟΓΙΑ ΙΗΣΟΥ, 7. For very recent discussion of itacism in Koine Greek, see now Chrys C. Caragounis, The Development of Greek and the New Testament: Morphology, Syntax, Phonology, and Textual Transmission, WUNT 167 (Tübingen: Mohr-Siebeck, 2004), esp. 339–96; and Francis T. Gignac, ‘Phonological Phenomena in the Greek Papyri Significant for the Text and Language of the New Testament’, in To Touch the Text: Biblical and Related Studies in Honor of Joseph A. Fitzmyer, S.J., ed. Maurya P. Horgan and Paul J. Kobelski (New York: Crossroad, 1989), 33–46. 16.  Ibid., 6. 17.  For a careful discussion of the functions of numbering in codices, see Turner, Typology, 73–80.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

173

case of multiple-gathering codices. With only this one leaf from the codex extant, it is wise to be cautious. That the number on this leaf appears to be from a hand other than the copyist further justifies caution. Turner cites other examples of numbering added by a second hand, and suggests that in these instances it might have been added by a user of the codex, perhaps to facilitate reference.18 So, the number 11 on this leaf might mean that it was preceded by ten leaves (or twenty pages), which is my own guess. But the number might instead reflect a pagination scheme added by a user, a scheme that might have involved numbering the right-hand pages. Or, just conceivably, the number might mean that this leaf was preceded by ten previous gatherings, each of an uncertain number of papyrus sheets. The beginning of text on the verso of P.Oxy. 1 corresponds to saying 26 of the Coptic text. As noted, if the number reflects a marking of the leaves or the papyrus sheets, then the extant verso was preceded by ten leaves (twenty pages). At about 36 lines per page, 20 pages would accommodate about 720 lines of text, or about 575 letters per page (each line about 16 letters). As Fitzmyer judged, it would not have required 20 pages to accommodate the equivalent in Greek of sayings 1–25 of the Coptic, and so some other text probably preceded GThomas in the codex.19 But, unfortunately, we have no basis for proposing what other text(s) this codex might have included.20 In summary, I suggest that P.Oxy. 1 is a copy prepared by a scribe of modest literary ability and quite possibly for someone of modest financial resources (and so was not able to afford a more elegant copy), or else by/for a reader who wanted only a readable copy of the text. The codex format reflects the general preference for this book-form among early Christians, and the likely page-size is a common one among codices of the period of this one. The absence of the sort of scribal devices that we 18.  Ibid., 75. 19.  Fitzmyer, ‘The Oxyrhynchus Logoi’, 355–56 n. 2. 20.  In her presentation at the Eisenach conference on the GThomas (October 2006) at which the original version of the present essay was given, Jutta LeonhardtBaltzer noted that the Coptic version of GThomas in Nag Hammadi Codex 2 is preceded by a version of the Apocryphon of John, and suggested that this might have been intended to make the latter text a kind of interpretative context for reading GThomas. Her presentation made me realize all the more how very interesting it is that also in P.Oxy. 1 the Greek GThomas was preceded by some other text. It is intriguing to wonder if this text was some version of the Apocryphon of John! For an analysis of the contents and arrangements of the Nag Hammadi codices, see also Michael A. Williams, Rethinking ‘Gnosticism’: An Argument for Dismantling a Dubiou Category (Princeton: Princeton University Press, 1996), 235–66, esp. 253–55 for discussion of Codex 2.

174

Texts and Artefacts

customarily associate with copies prepared for public reading suggests that this may have functioned as a personal/private copy.21 P.Oxyrhynchus 654 In P.Oxy. 654 we have a very different item.22 First, it is a portion of an ‘opisthograph’, a re-used roll, the extant text of GThomas having been copied on the outer side of a roll, whose inner side had been used previously for a land survey-list. Grenfell and Hunt dated the cursive hand of this survey-list to the late second or early third century CE, and assigned the hand of this copy of GThomas to the middle or late third century CE. This dating has obtained general acceptance.23 The extant portion measures 24.4 (h) × 7.8 cm (w). Portions of the top margin (1.5 cm) and the left-hand space between columns (2.6 cm) survive, as well as 42 incomplete lines of text that correspond to sayings 1–7 of the Coptic text of GThomas. Unfortunately, there is an uneven vertical break that cuts through all the extant lines, so that only the left-hand portion of each line survives. Reconstructions of the remainder of the lines based on retranslation from the Coptic suggest an original column of ca. 9 cm width, and ca. 26–36 letters per line.24 The text is written in majuscule letters that vary considerably in height, from 1.6 mm (omicron and omega), to 2 mm (sigma, thēta, and ēta), and 3 mm (descenders such as upsilon), and the initial letters of lines somewhat larger, ca. 3–4 mm. Attridge repeated the description of the hand by Grenfell and Hunt, ‘a common informal literary type of the third century’, but I would suggest that this is a rather generous characterization.25 The complete inability at bilinear writing, the irregularities of letters in size and formation, and other features seem to me to indicate a scribe of very limited skill (or little interest) with regard to the aesthetic properties usually expected in copies of literary texts. Furthermore, the errors in spelling (e.g., γνωσθε, line 20; αποκαλυφησετ[αι], line 29), and the bizarre 21.  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 International, 2009), 148–75, proposes scribal features that may distinguish copies used for liturgical or personal purposes. 22.  P.Oxy. 654 is catalogued as B. L. pap. 1531 and held in the British Library (London). 23.  Grenfell and Hunt, ΛΟΓΙΑ ΙΗΣΟΥ, 9. 24.  The maximum number of extant letters in any line is 19 (line 28). 25.  Attridge, ‘Appendix’, 97.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

175

first line (οιτοι οι οι λογοι οι…, which Grenfell and Hunt described as ‘intolerable, even in third century Greek’) combine to indicate a scribe characterized by a noticeable level of carelessness or limited ability.26 Note also the two cases where the scribe accidentally omitted words and then inserted them above the line (υμεις, line 19; οτι, line 25). The only nomina sacra form is the consistent use of ΙΗΣ (lines 2, 27, 36). Compare this three-letter form of the name with the two-letter spelling used in P.Oxy 1. It is also interesting that neither πατρος (line 19) nor ουρανος (lines 11–12) is treated as a nomen sacrum in this manuscript, whereas in P.Oxy 1 πατρος is written as a nomen sacrum. There is no punctuation, but there are a number of other interesting scribal devices. In particular, note the ‘coronis’ marks (shaped somewhat like an arrow) that usually precede the phrase λεγει Ιη(σου)ς (lines 5, 9, and 36); but note also line 27, where the mark appears mistakenly placed following this phrase. At several points, horizontal lines extending from the left margin and a few letters into the text signal the first full line of a saying (lines 6, 10, 22, 28, 32). These lines, however, appear to have been added by someone other than the copyist. A diaeresis over an initial upsilon or iota appears in lines 13, 14, 15, 19, and 21. To summarize, all indications make it likely that P.Oxy. 654 is a personal copy of GThomas, intended for private study. The carelessness or limited skill of the copyist may suggest either that the intended user was unable to afford a better quality copy, or was simply not sufficiently concerned to have one. P.Oxyrhynchus 655 The remaining item to be considered, P.Oxy. 655, is comprised of several fragments of a papyrus roll.27 Unfortunately, two of the eight fragments originally identified as part of this manuscript are now missing (fragments 26.  Greenfell and Hunt. Eds., Oxyrhynchus Papyri, Volume IV, 2. 27.  P.Oxy. 655 is catalogued as SM 4367 in the Houghton Library (Semitic Museum Collection), Harvard University. Originally, for this discussion I studied a good-quality photograph kindly supplied by the Houghton Library, which I was able to scan into digital form and then enlarge for examination of details. In the republished version here, I draw upon my own examination of the fragments in the Houghton Library, which produced some alterations and additions to the results of earlier studies. I also draw upon the results of Robert A. Kraft’s analysis of the fragments (‘Oxyrhynchus Papyrus 655 Reconsidered’, HTR 54 [1961]: 253–62), which corrects the results of earlier studies, including Grenfell and Hunt’s editio princeps, at several points.

176

Texts and Artefacts

f and h), but these were rather small and their loss probably does not detract much from what we can say about the manuscript. Of the remaining ones, fragment b is the largest, 8.3 cm (w) × 8.2 cm (h). According to Kraft’s reconstruction, which I can confirm, fragment a contains parts of eleven lines from the top of one column, and fragment b has another 17 lines of the same column (most lines incompletely preserved) and the initial few letters of ten lines of a second column. In what follows, I include more detailed observations of the size and nature of the lettering than, to my knowledge, otherwise available.28 When connected, fragments a and b give us a combined height of ca. 12.2 cm. The extant top margin on fragment b is ca. 10 mm. The intercolumn space amounts to ca. 14 mm. Fragment c has the initial one or two letters of ten lines from the top of this same second column. Kraft has also proposed that fragments d and e might be bits of some other column, but not one adjacent to the material in the larger fragments.29 In sum, Kraft proposed that ‘the papyrus once contained columns of about 30 lines each [ca. 5 cm width], with approximately 12–16 letters per line (usually 14–15)’.30 It is worth noting that these are narrow columns in comparison with more typical layout on literary rolls (and cf. the estimated 9 cm width of columns in P.Oxy. 654). These narrow columns may have been designed to compensate for the very small size of the lettering. Most letters are mainly 2 mm in height, although the alpha and upsilon are ca. 3 mm in height, and the phi and psi closer to 4 mm, whereas sigmas and epsilons are sometimes as small as 1.5–1.6 mm. The writing is tightly compressed, with line-spacing only ca. 3 mm. It can be confirmed easily that the eye finds long lines of small print harder to read, and this would have been all the more true in the absence of punctuation and word division. The extant fragments preserve portions of text that correspond roughly to sayings 36–39, and possibly a bit of saying 24 as well, of the Coptic GThomas. There is, however, at least one major difference. Nearly all of lines 4–17 of Kraft’s re-construction of the first extant column (also accepted in Attridge’s edition) have no parallel in the Coptic text of

28.  I hope that the more recent practice of giving measurements of letters (in millimetres) will be more widely followed in descriptions of ancient manuscripts. Characterizations such as ‘small’ or ‘medium sized’ are far too imprecise for scientific analysis and comparison. 29.  See Kraft’s discussion of the possible contents of the smaller fragments in ‘Oxyrhynchus 655’, 260–62. 30.  Ibid., 262.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

177

saying 36.31 This and a number of other instances of variation in content and order of material between the Oxyrhynchus manuscripts and the Coptic GThomas should make us a bit cautious (perhaps more cautious than some scholars have been) about using the Coptic to make conjectural restorations of the Greek manuscripts.32 The small, majuscule characters noted earlier are confidently and skilfully formed, with serifs or small hooks on some letters (e.g., some horizontal strokes of tau, vertical strokes of some letters such as lambda and phi). It is clear that the scribe aimed to produce a very compact manuscript. Attridge estimated that the original height of the roll was approximately 16 cm, which had to accommodate top and bottom margins and some 30 lines of text. For modern comparison purposes, the volumes in the Loeb Classical Library have pages of just over 16 cm height, and a maximum of about 32 lines per page. A comparison with roll-heights of the same approximate period as P.Oxy. 655 confirms that it was very much on the small/compact end of the spectrum. In his invaluable study of Greek literary rolls, William Johnson judged that in the Roman period roll heights were characteristically 25–33 cm, with exceptions on either end of the range.33 So, the small size of P.Oxy. 655 indicates that it was very likely a personal copy. The letters are evenly spaced, and the scribe appears to have aimed basically for bilinear layout, but without complete success. As is often the case in manuscripts of the period other than the highest-quality literary hands, the omicron is significantly smaller than the other letters. Also, although in general the individual letters are formed fairly consistently, there is some small variation in a few (e.g., the various instances of sigma, epsilon, and thēta). There is no observable punctuation, no breathing marks or accents or use of spaces to mark sense-units, and no ekthesis or 31.  See Kraft’s note on this (ibid., 254), in which he weighs whether P.Oxy. 655 or the Nag Hammadi Coptic text better represents ‘the original form’ of GThomas. But it may also be that the text was always subject to additions or deletions. Cf. also Attridge’s restoration of these lines (‘Appendix’, 121–22), which have some relationship to the sayings in Matt 6:27–29 and Luke 12:25–27. 32.  Another famous instance of significant difference between the Oxyrhynchus fragments and Coptic GThomas is, of course, the position of the saying about lifting the stone and splitting ‘the wood’, which is part of a saying corresponding to Coptic saying 30 in P. Oxy. 1, but is attached to saying 77 in the Coptic text. Attridge listed a number of ‘substantial differences’ between the Greek fragments and the Coptic text, ‘Appendix’, 99–101. 33.  William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004), 141–43.

178

Texts and Artefacts

any other device to mark paragraphs. On line 3 of fragment d, however, there is a small >-shaped mark apparently serving as a line filler. There is also one correction in line 2 of this same fragment, a small epsilon written above the iota in what remains of the word φωτινω. For a somewhat analogous hand (perhaps just slightly less elegant), note, for example, P.Oxy. 1016, a third-century CE roll of Plato’s Phaedrus.34 Unfortunately, none of the words characteristically treated as nomina sacra survives in these fragments. Those who have attempted reconstructions of the text (e.g., Attridge, Kraft) have tended to assume, however, that Jesus’ name was written as ΙΣ, although a three-letter form (ΙΗΣ) is also entirely possible. Summarizing Observations and Inferences Let us now try to gather up the data from our analysis of the three Oxyrhynchus manuscripts of GThomas and attempt some observations and inferences. We may begin by noting that the multiple copies of GThomas clearly indicate a certain level of interest in this text among Christians in the early third century. If, however, we compare the number of copies of all literary texts in identifiably Christian manuscripts dated prior to 300 CE, we can put the three copies of GThomas into some perspective.35 Of course, any such comparison rests on the assumption (by no means incontrovertible) that the extant number of copies of a text from a given period is some general reflection of its comparative popularity at that time.36 I list here the literary texts in question in the decreasing order of the numbers of extant copies among Christian manuscripts dated to the second and third centuries CE: Psalms (18), John (17), Matthew (12), Shepherd of Hermas (11), Genesis and Exodus (8 each), Luke and Acts (7 each), Isaiah (6), Revelation (5), Romans and Hebrews (4 each), James, Ephesians, Leviticus, Acts of Paul, Gospel of Thomas (3 each), 1 Corinthians, 34.  See plate 84 in Eric G. Turner, Greek Manuscripts of the Ancient World, 2nd ed., Institute of Classical Studies Bulletin Supplement 46 (London: Institute of Classical Studies, 1987). 35.  For a fuller survey of all literary texts in earliest Christian manuscripts, see Hurtado, Artifacts, 15–41, and Appendix 1 for a complete list as of the date of that publication. The question marks in the list above reflect uncertainties about the identification of the texts in question. 36.  In the following numbers, I omit papyri dated ‘third/fourth century’, and count only those date clearly to either the second or third century. This results in some numbers different from the original version of this essay.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

179

Philippians, 1 Thessalonians, 2 Thessalonians, Jude, Irenaeus, Gospel of Mary, Gospel of Peter (?),37 Deuteronomy, 2 Chronicles, Esther, Proverbs, Ecclesiastes, Sirach, Jeremiah, Ezekiel, Daniel, Minor Prophets, and Tobit (2 each), and a large number of other writings for which we have only one copy from this period (Numbers, Joshua, Judges, Job, Wisdom, Susannah, 2 Maccabees, Mark, 2 Corinthians, Galatians, Colossians, Philemon, Titus, 1 Peter, 2 Peter, 1 John, 2 John, Protevangelium of James, ‘Egerton’ Gospel, ‘Fayum’ Gospel, Correspondence of Paul and Corinth, Apocalypse of Peter, Apocryphon of James and Jambres, Apocryphon of Moses, Melito’s Paschal Homily, Melito’s On Prophecy?, Melito’s Paschal Hymn?, Odes of Solomon, Julius Africanus’ Cesti, Origen’s Gospel Commentaries, a Homily by Origen, Origen’s De Principii, Sibylline Oracles, Diatessaron?, Theonas’ Against Manichaeans?, and a small number of unidentifiable texts. So, the three copies of GThomas suggest a readership interest somewhat greater than for many other texts, but it hardly stands out. In fact, GThomas ties for thirteenth place along with James, Ephesians, Leviticus, and Acts of Paul, the three copies of GThomas suggesting an interest in the second and third centuries perhaps approximate to that given to these other writings. On the other hand, obviously the three copies place it ahead of many texts, including a number of canonical ones (e.g., Mark)! It is also interesting that in the three copies of GThomas we have an example of each of the three book-forms of the ancient period. It is almost certain that ‘opisthographs’ (re-used rolls) represent economical copies of texts made for private reading and study, and this is no doubt what we have in P.Oxy. 654.38 As for P.Oxy. 655, the roll format, compact size, the smallness of the writing and the large number of lines per column, combine to suggest likewise a private copy (as already noted), and perhaps a copy intended for portability.39 But, in comparison with P.Oxy. 654, the greater skill of the copyist of P.Oxy. 655 and the use of 37.  It is now less certain that P.Oxy. 2949 and P.Oxy. 4009 are fragments of the Gospel of Peter, in light of the discussion by Thomas J. Kraus and Tobias Nicklas, eds., Das Petrusevangelium und die Petrusapokalypse: Die Griechischen Fragmente mit deutscher und englischer Übersetzung, GCS 11 (Berlin: de Gruyter, 2004); and Paul Foster, ‘Are There Any Early Fragments of the So-called Gospel of Peter?’, NTS 52 (2006): 1–28. 38.  Other early examples of Christian opisthographs include P.IFAO 2.32 (P98; Revelation), P.Oxy. 1228 (P22; John), PSI 8.921v (Psalms), P.Mich. 130 (Hermas), and a couple of unidentified theological treatises or homilies (P.Gen. 3.125 and P.Mich. 18.763). 39.  There are likewise examples of compact-size codices, Christian and nonChristian, and these are thought to have been prepared for portable usage. See my discussion of codex sizes in Artifacts, 155–65.

180

Texts and Artefacts

a fresh roll also suggest that it was copied for someone able to pay the likely greater cost involved, or perhaps someone able to draw upon the services of a skilled slave to do the copying (which would also indicate a person of some financial means and able to own such a skilled slave). Also, given the strong general preference for the codex among ancient Christians, especially for texts used as scripture, the choice to copy a text in a fresh roll surely further indicates that this text (or at least this copy of the text) was not used as scripture, i.e., not read publicly in worship settings.40 It is, however, more difficult to be certain about the intended usage of P.Oxy. 1. The probable size of this codex is rather typical of papyrus codices of its time, including many Christian copies of biblical texts. But, as noted already, in comparison with some other Christian manuscripts, the somewhat larger number of lines per page and the small size of the letters, plus the good likelihood that the codex included some other text(s) as well as GThomas combine to make one wonder if this manuscript was some sort of compendium, perhaps for personal usage/study. The lack of sense-unit markers or spacing and punctuation is consistent with this, as Christian codices copied for public/liturgical usage tend to have such readers’ aids supplied by the original copyist. So, collectively, these three third-century copies of GThomas likely reflect a use of this text in personal or non-liturgical settings, perhaps for religious edification, reflection, and/or study. With allowance for the limits of our artefactual evidence and the need to take account of any other evidence in forming conclusions, we can say that we have no indication from these manuscripts that GThomas functioned as ‘scripture’ for their intended readers. Or, perhaps a bit more circumspectly, we could say that it is at least unlikely that any of these three copies so functioned.41 It is worth noting that the likely personal usage of these Greek copies of GThomas fits well with the emphasis in this text on the individual and on personal spiritual fulfilment. From the opening words onward, the exhortations are largely to individuals, as reflected, for example, in the singular forms of the Greek verbs in the exhortations in the prologue and sayings 1, 3, 5, 24, 26, and particularly in the very difficult saying 30, with 40.  For further discussion of the Christian preference for the codex, see Artifacts, Chapter 2. It remains interesting that (at least so far) we have no instance of a writing that became part of the NT copied on an unused roll (in contrast to the few examples of opisthographs), and only a very few copies of OT writings on rolls likely from Christian hands. 41.  Cf. the drift of Luijendijk, ‘Reading the Gospel of Thomas’, who queries my conclusion.

10. The Greek Fragments of the Gospel of Thomas as Artefacts

181

its emphatic statement that οπου εις εστιν μονος λεγω εγω ειμι μετ’ αυτου.42 The incidence of such singular exhortation-forms could be multiplied if we take account of the fuller text given in the Coptic version. By contrast, nothing in GThomas (in the extant Greek or the Coptic) seems to me to promote corporate/congregational religious life. So, GThomas may have been intended originally for the private/personal attention of individuals seeking some sort of deeper/higher truth, a further spiritual attainment, or a further revelation of purportedly esoteric truths. Perhaps like-minded souls formed some sort of loose network, sharing texts such as GThomas with one another. But I see little reason to think that demarcated ‘communities’ lie behind this text, or that it functioned as a liturgical text. It is also worthwhile to note that the variation in the quality of the hands, book-forms, and handling of nomina sacra (esp. the variation in the way Jesus’ name is written) all combine to indicate that these three manuscripts were copied under separate circumstances. It is, I think, unlikely that they derive directly from a common archetype. If so, then these several copies of GThomas probably derive from a preceding equivalent number of prior copies. Again, this reflects a certain level of readers’ interest in the text that probably pre-dates the time of the extant manuscripts of GThomas. The use of nomina sacra forms for Ἰησοῦς shows both a well-known Christian scribal convention, and also a certain variation in the specific implementation of that practice among the three copyists of these manuscripts. Unfortunately, I am not aware that we know the precise find-spot(s) for these three copies of GThomas in relation to the other Christian manuscripts from Oxyrhynchus, but there is no particular reason to link these manuscripts with some distinctive circle of Christians. So far as we know, these artefacts were copied and used among the same Christian circles in which the other texts found in Oxyrhynchus and other ancient Egyptian sites functioned. There are no doubt further inferences that might be drawn. But I trust that these will suffice to make the basic point that it is worthwhile to take account of the physical and visual properties of the actual manuscripts of texts in forming judgments about their place, role, function and significance in earliest Christianity. My key concern is that in all our efforts to understand such texts as GThomas, and to imagine what kind(s) of readers and significance the text might have enjoyed, we take full account of the actual artefacts of these phenomena, the earliest Christian manuscripts. 42.  I reproduce here the reconstruction of lines 25–30 of the recto of P.Oxy. 1 by Attridge, ‘Appendix’, 119.

Chapter 11 W ho R e a d E a rly C h ri s t i an A pocry pha ?

The question posed in the title of this essay is difficult to answer for several reasons. First, there is a diversity of texts usually included under the category ‘early Christian apocrypha’, including typically some texts often linked with dissident, ‘heretical’ groups and/or ideas, but also other texts that may have been intended to supplement or expand upon early ‘orthodox’ texts and ideas, and/or simply to promote a version of Christian edification, and perhaps entertainment.1 So an immediate answer to the question of who read ‘early Christian apocrypha’ is that the variety of texts under this heading likely signals a variety of readers; and this is confirmed by other evidence that we will note shortly. Given the diversity and number of texts that comprise ‘early Christian apocrypha’, however, within the limits of this discussion it is not realistic to attempt anything more than an illustrative treatment of the matter. Another problem lies in the available evidence for a number of these apocryphal texts. For some of them (e.g., the gospels often linked specifically with Jewish Christianity), we have only what purport to be brief quotations given by some early Christian writers, mere snippets of what may have been texts of some considerable size. So inferring readers from contents (always difficult) is all the more difficult with regard to these texts. This has not stopped some scholars from speculating about the provenance of these writings, but in my view these speculations carry little probative force.2 1.  I place the words ‘heretical’ and ‘orthodox’ in scare-quotes to signal that I use them simply to reflect how certain texts have been seen traditionally. 2.  For an effort (not always persuasive) to posit the origins and nature of the Jewish-Christian gospels that may lie behind the putative citations of them in early Christian writers, see A. F. J. Klijn, Jewish-Christian Gospel Tradition, VCSup 17 (Leiden: Brill, 1992). Cf., e.g., Andrew Gregory, ‘Jewish-Christian Gospels’, in The Non-Canonical Gospels, ed. Paul Foster (London: T&T Clark International, 2008), 54–67.

11. Who Read Early Christian Apocrypha?

183

Moreover, some apocryphal texts survive only in translation, often from a considerably later time than the probable point of original composition, reflecting a transmission-history that likely involved significant redactional changes, sometimes perhaps of incalculable dimensions. Even when texts survive in the likely original language, it is often clear that they underwent significant re-shaping and adaptation in the course of being copied across the early centuries (as, e.g., we can see in comparing the extant portions of the earlier Greek text of the Gospel of Thomas with the later Coptic version).3 On the one hand, the translation of these texts certainly reflects interest in them and usage of them by readers in various cultural settings. On the other hand, this usage by different kinds of readers across time further complicates any answer to the question of who read the texts in question. Positing Readers from Contents One of the ways that scholars have gone about trying to posit the readers of early Christian texts (both canonical and extra-canonical) is by making inferences based on their contents. Indeed, this kind of ‘mirror reading’ has often generated proposals that this or that kind of readers or ‘community’ is reflected in, and lay behind, a given text. Granted, to some extent early Christian texts do likely reflect interests and situations of authors and intended readers. But defining with any precision the identity and nature of the intended readers on the basis of the contents of unprovenanced texts is more difficult that some may have realized, and involves making a few dubious assumptions. By contrast, in the case of some texts, e.g., such as the undisputed letters of Paul, we may well have explicit identification of the author and addressee(s), as well as reference to the situation (either the author’s or the addressees’) that occasioned the letter. But by ‘unprovenanced’ texts I mean texts that do not give such information, making it more often very difficult to judge these things. 3.  See, e.g., Harold W. Attridge, ‘Appendix: The Greek Fragments’, in Nag Hammadi Codex II, 2–7 Together with XIII, 2*, Brit. Lib. Or. 4926(1), and P. Oxy. 1, 654, 655, Vol. 1: Gospel According to Thomas, Gospel According to Philip, Hypostasis of the Archons, and Indexes, ed. Bentley Layton (Leiden: Brill, 1989), 95–128. One of the few commentators on the Gospel of Thomas who takes serious notice of differences between the extant Greek fragments and the Coptic text is Richard Valantasis, The Gospel of Thomas (London: Routledge, 1997). As another example, in his recent study of the Infancy Gospel of Thomas, Tony Burke discusses its complex recension-history: Tony Burke, De Infantia Iesu Evangelium Thomae – Graece, Corpus Christianorum Series Apocryphorum 17 (Turnhout: Brepols, 2010).

184

Texts and Artefacts

One of the dubious assumptions sometimes made is that the characters, settings and emphases in texts reflect the historical circumstances and characteristics of the intended readers sufficiently to permit us to determine these matters. But in fact literary texts may reflect imaginary settings, may project fictional, perhaps idealized people and circumstances, and may even portray circumstances that contrast with those of the author and intended readers, for example, for the purpose of offering something exotic or interesting to readers. It is actually not clear that texts so readily and directly mirror their intended readers. Any claim that they do must be argued on a case-by-case basis, preferably supported with some other corroborating evidence. A related dubious assumption is that texts such as the apocryphal writings were intended for some specific ‘community’ or particular circle of readers in the first place. Several decades ago, Frederik Wisse criticised the assumption that texts such as those often labelled ‘gnostic’ reflect specific circles or ‘communities’ whose beliefs and practices can be read off the texts in question. He judged it ‘very questionable’ to attribute the special features of a given text to ‘a certain, otherwise unknown, branch of early Christianity’.4 To cite Wisse further, the beliefs and practices advocated in these writings, insofar as they vary from those reflected in other Christian texts, cannot be attributed to a distinct community or sect. Rather, these writings were more likely idiosyncratic in terms of their environment. The ‘teaching’ they contain was not meant to replace other teaching but to supplement. They did not defend the beliefs of a community but rather tried to develop and explore Christian truth in different directions.5

That is, the intended readers of these ‘gnostic’ texts may as likely have been assorted individuals of various types, interested (perhaps for various reasons) in what they may have regarded as the explorative or innovative approach to Christian faith taken in the texts. Moreover, despite assumptions to the contrary, there is no good reason for thinking that these texts functioned as ‘scripture’ for defined Christian groups such as ‘Thomasine’ Christians distinguishable from other Christian circles.6 4.  Frederik Wisse, ‘The Use of Early Christian Literature as Evidence for Inner Diversity and Conflict’, in Nag Hammadi, Gnosticism, and Early Christianity, ed. Charles W. Hedrick and Robert Hodgson (Peabody, MA: Hendrickson, 1986), 177–90 (181). 5.  Ibid., 188. 6.  Cf., e.g., Bentley Layton, The Gnostic Scriptures: A New Translation with Annotations and Introductions (New York: Doubleday, 1987).

11. Who Read Early Christian Apocrypha?

185

An example of the dubious tendency to ascribe specific groups or types of readers to early Christian texts is the claim that some of the apocryphal Acts were intended for women readers.7 The basic reason for such proposals seems to be that the texts in question feature prominent women characters. Here, again, we see the somewhat simplistic notion that texts rather directly mirror their intended readers. This particular view, however, that certain of the apocryphal Acts were particularly directed to women readers, has been challenged recently by Kim Haines-Eitzen. Focusing on the physical form of the early manuscripts containing these texts (a topic to which I return later in this discussion), she judges that ‘there is nothing to suggest a gendered readership’.8 Instead, she concludes, If, indeed, the form of ancient books can tell us something about their readers – a subject we still have much to learn about – then these apocryphal acts were read not by the ‘popular’ masses or necessarily by ‘women’ but rather by those members of the upper echelons who likewise enjoyed poetry, history, and perhaps philosophy.9

In sum, instead of assuming that any specific readership can be inferred directly from the contents of a given text, it is wiser to take account of other data as well. In what follows, I underscore this other evidence, some of which has perhaps not been considered adequately. ‘External’ References to Usage In what follows, I first discuss ‘external’ evidence, by which I mean here explicit references in ancient Christian writings to the use of apocryphal texts. One of the most explicit of such references is the oft-cited passage where Eusebius describes how Serapion (bishop of Antioch, 199–211 CE) reacted to being informed that a Gospel of Peter was read by certain members of the church in Rhossus (Eusebius, HE 6.12.1–6), one of 7.  Examples of proponents include Stevan Davies, The Revolt of the Widows: The Social World of the Apocrypha Acts (Carbondale: Southern Illinois University Press, 1980), esp. 95–96; Virginia Burrus, Chastity as Autonomy: Women in the Stories of the Apocryphal Acts (Lewiston: Edwin Mellen, 1987), esp. 108. 8.  Kim Haines-Eitzen, The Gendered Palimpsest: Women, Writing, and Representation in Early Christianity (New York: Oxford University Press, 2012), 53–64 (citation 57). See also her earlier study, ‘The Apocryphal Acts of the Apostles on Papyrus: Revisiting the Question of Readership and Audience’, in New Testament Manuscripts: Their Texts and Their Worlds, ed. Thomas J. Kraus and Tobias Nicklas (Leiden: Brill, 2006), 293–304. 9.  Ibid., 62–63.

186

Texts and Artefacts

the communities within his episcopal authority. Initially assuming that the text was harmless (though he apparently doubted the attribution to Peter), Serapion did not object to Rhossus Christians continuing to read it. Then, however, after being prompted to read the text himself carefully, he came to the view that, though for the most part it was ‘in accordance with the true teaching of the Saviour’, there were some things ‘added’ (προσδιεσταλμένα), which he apparently regarded as possibly tending in a heretical direction. Although it cannot detain us here, an obvious question concerns the identity and contents of the text that Serapion considered. A majority of scholars have assumed that it is substantially the same as the partially preserved text included in the ‘Akhmîm codex’ (P.Cair. 10759, dated variously seventh–ninth century CE), which, accordingly, has acquired the title ‘Gospel of Peter’.10 In the most thorough examination of matters to date, however, Paul Foster urges that ‘one should exercise caution before too quickly identifying the Akhmîm text with the Gospel of Peter that Serapion declared open to docetic interpretation’.11 But it is not crucial here whether the Akhmîm does or does not preserve substantially the text that Serapion wrote about; instead, we focus on what Eusebius relates about the use of the text that he refers to as the Gospel of Peter. Clearly, this text was read by Christians in Rhossus, and likely in addition to, not instead of, other texts more familiar to us that came to form the emerging Christian canon. Eusebius says that Serapion wrote a refutation of the ‘false statements’ in the Gospel of Peter, because some in Rhossus ‘on the ground of the said writing had turned aside into heterodox teachings’ (HE 6.12.2). But it is by no means clear that these people comprised some separate circle or ‘community’ that identified itself with particular reference to this text. Instead, it appears to have been simply one of the texts read among at least some Christians in Rhossus, with

10.  The partially preserved text in the Akhmîm codex bears no title, but is typically referred to as the Gospel of Peter, largely because of the statement in the final lines, ‘But I, Simon Peter, and Andrew my brother, having taken up our nets went to the sea…’ (Gospel of Peter 14.58). I translate the transcription of the text in Paul Foster, The Gospel of Peter: Introduction, Critical Edition and Commentary (Leiden: Brill, 2010), 205. 11.  Foster, The Gospel of Peter, 90. Note also Thomas J. Kraus and Tobias Nicklas, eds., Das Evangelium nach Petrus: Text, Kontexte, Intertexte (Berlin: de Gruyter, 2007). Reflecting the common assumption that the Akhmîm text is essentially the Gospel of Peter referred to by Eusebius, see, e.g., J. K. Elliott, ed., The Apocryphal New Testament (Oxford: Clarendon, 1993), 150–51.

11. Who Read Early Christian Apocrypha?

187

some of them apparently taking it as justifying some ideas that Serapion considered heterodox. However, it is interesting that, although Serapion became concerned about this, he also judged that ‘for the most part’ the text was acceptable. It is not clear from Eusebius’ report whether, in addition to refuting the bits that he regarded as susceptible to heterodox interpretation, Serapion also actively sought to suppress this Gospel of Peter, or was content to correct what he regarded an inappropriate use of the text. If we return briefly to consider the Akhmîm text regarded by many as a later copy or version of the Gospel of Peter, it is noteworthy that the small, composite codex containing the text (along with portions of the Apocalypse of Peter, 1 Enoch and the Martyrdom of St. Julian) was found in the grave of a man buried in a cemetery that seems to have been associated with a nearby monastery.12 The most natural assumption is that the man was a member of the local Christian community, quite possibly a monk, and was given a Christian burial, this small codex buried with him, perhaps because it was some special possession of his. So, whether the text is or is not related to the Gospel of Peter that Serapion read and refuted, the indications are that this apocryphal text (along with the other extra-canonical texts in this codex) was read (and perhaps treasured) by this Christian man, with no indication that he was particularly heterodox or that he was part of some discrete circle attached to the text. There are also several references in ancient Christian writers to other gospels that they link specifically to Jewish Christians.13 But there are problems that create major uncertainties about these texts.14 The most significant problem is that all we have are a handful of what these writers present as citations of these texts, and no copy of any of them survives.

12.  For information on the discovery of the codex, see U. Bouriant, ‘Fragments du texte grec due livre d’Énoch et de quelques écrits attribués à saint Pierre’, in Mémoires publiés par les membres de la Mission archéologique française au Caire, tome IX, fasc. 1 (Paris: Ernest Leroux, 1892), 93–147. 13.  Klijn, Jewish-Christian Gospel Tradition, gives introduction, text and commentary on the citations of these texts. Cf., however, the more recent review of matters by Gregory, ‘Jewish-Christian Gospels’. On Jewish-Christianity more generally, see the multi-author work, Jewish Believers in Jesus, ed. Oskar Skarsaune and Reidar Hvalvik (Peabody, MA: Hendrickson, 2007), and on the various JewishChristian gospels, Craig A. Evans, ‘The Jewish Christian Gospel Tradition’, 241–77, which includes copious citation of other scholarly studies. 14.  Evans, ‘The Jewish Christian Gospel Tradition’, 241, lists several difficulties in the extant evidence.

188

Texts and Artefacts

A second difficulty is that the references to and descriptions of these texts do not cohere very well. So, for example, it is not entirely clear whether we are dealing with two or three such gospels.15 In scholarly discussion one finds references to a Gospel according to the Hebrews, a Gospel of the Ebionites, and a Gospel of the Nazoraeans, but only the first title appears in the ancient sources, and the others are convenient titles attached by scholars to the putative excerpts of texts associated with Jewish-Christian circles in the ancient writers in question.16 The earliest, Irenaeus (Adv. Haer. 1.26.2; 3.11.7), refers to Ebionites having their own gospel, which he took to be a version of the Gospel of Matthew.17 Later, Epiphanius (Pan. 30) also refers to Ebionites, whom he characterizes as having certain heretical beliefs, and he gives several citations of what he says was their own gospel.18 But scholars are not agreed that Epiphanius actually knew either Ebionites or the gospel that he ascribes to them. Whatever we make of Ephiphanius’ statements, as Gregory noted, the apparent acquaintance with a Gospel according to the Hebrews shown by Clement of Alexandria, Origen and Didymus ‘suggests that it was both widely known and widely acceptable, at least among those for whom they wrote’, and that it was likely ‘a text with wide appeal in the period before the boundaries of the New Testament canon were formally closed’.19 In short, although there were extra-canonical gospel-texts that may well have emerged in Jewish-Christian circles and may have been intended initially for such readers, it appears that they then obtained a wider readership in the early Christian centuries. If these texts functioned as scripture for some, they were also read, and probably with some appreciation, also among those who did not include these texts among their scriptures. 15.  Noting recent studies, Evans (ibid., 246) states, ‘Obviously, the state of the scholarly question has been thrown into the air; there simply is no consensus’. 16.  Clement of Alexandria, Strom. 2.45.5 and Jerome, Comm. Mich. 7.7; Comm. Isa. 40.9 refer to a gospel according to the Hebrews. Gregory, ‘Jewish-Christian Gospels’, 56–59, reviews the difficulties in identifying how many such gospels there may have been and what they may have comprised. 17.  Oskar Skarsaune, ‘The Ebionites’, in Skarsaune and Hvalvik, eds., Jewish Believers in Jesus, 419–62. 18.  See, e.g., Gregory, ‘Jewish-Christian Gospels’, 61–66, for discussion of the citations, which he judges may more likely derive from a gospel harmony that shows dependence on Matthew and Luke, but also some distinctive tendencies, e.g., vegetarianism. 19.  Ibid., 67.

11. Who Read Early Christian Apocrypha?

189

As another category of apocryphal gospel, let us now consider the so-called Infancy Gospels. Once again, we have texts that seem to have enjoyed a wide readership and reception in Christian circles.20 In fact, the most influential of these, known variously as the Protevangelium Jacobi, the Nativity of Mary, and The Infancy Gospel of James, exercised ‘an influence on the faith and piety of the church which rivals that of the canonical gospels’.21 This is particularly evident in Christian art down the centuries after this text was composed, with numerous examples of scenes and figures that are derived from it.22 For example, depictions of Joachim and Anna (Mary’s parents), and type-scenes such as Mary’s birth, all reflect its influence long after the text itself was forgotten in Western Christianity (though it continued to be copied and read in Eastern Christianity). Moreover, if we note that the paintings in question were typically commissioned by ecclesiastical authorities, it becomes clear that this extra-canonical text exercised influence (and perhaps had a readership) in the hierarchy of Christianity, at least in late antiquity and the medieval period, and not only among the laity.23 Another leading example of this type of text is The Infancy Gospel of Thomas (known in earlier centuries as Evangelium Thomae Israelitae or Evangelium Thomae), which is less a connected narrative and more ‘a collection of largely self-contained stories that are only loosely held together by a series of indications of Jesus’ age’ at various points in his childhood.24 Stephen Gero referred to it as ‘the fixation in writing of a 20.  For a handy volume that combines introduction, Greek text, and translation of the two major writings of this type, see Ronald F. Hock, The Infancy Gospels of James and Thomas (Santa Rosa, CA: Polebridge, 1995). 21.  David R. Cartlidge and J. Keith Elliott, Art and the Christian Apocrypha (London: Routledge, 2001), 21. Hock (Infancy Gospels, 4) refers to ‘a bewildering array of titles’ of the text in the many manuscripts of it. 22.  For examples, see Cartlidge and Elliott, Art and the Christian Apocrypha, 21–46. 23.  The text seems to have become virtually unknown in the Western churches at some point, however, and was re-introduced in a Latin translation by Guillaume Postel in 1552 from a Greek manuscript that is now lost. The earliest complete copy extant is P.Bodmer V (fourth century), a small codex whose provenance is probably upper Egypt, on which see Joseph van Haelst, Catalogue des papyrus littéraires juifs et chrétiens (Paris: Publications de la Sorbonne, 1976), no. 599 (212–13). There are also fragments of two other early copies: PSI 1.6 (fragments of five leaves, fourth century) and P.Grenf. 1.8 (parchment, fragments of eight leaves, fifth–sixth century), van Haelst, nos. 600 and 601 (213–14). Van Haelst also referred to a single leaf from another codex (now lost) and dated fourth century (no. 602). 24.  Hock, Infancy Gospels, 85.

190

Texts and Artefacts

cycle of oral tradition, of religious folklore’.25 The extant manuscripts exhibit considerable fluidity in the transmission of the text, with deletion and insertion of some stories, all the easier in such a loose collection of vignettes, at least some of which circulated on their own before being collected in Infancy Thomas. So, although some individual stories (especially the exchange between the child Jesus and a Jewish teacher about the meaning of the Greek letters alpha and beta) are attested in some early writings, it is not always easy to determine whether their authors actually knew and read something like the full text of Infancy Gospel of Thomas or instead simply were acquainted with some of these particular vignettes.26 But the multiple later copies of this infancy gospel, and its translation into a number of ancient languages, surely show that it came to be read in various Christian circles.27 As Gero noted, ‘Apparently, both Greek and Slavonic versions of apocryphal gospels were quite generally used in Orthodox monasteries as devotional reading’.28 Moreover, to judge by the references to Jesus in the Qur’an as ‘he who made clay birds fly’ (Surah 3.49; 5.110; one of the famous stories in Infancy Thomas 2:1–7), it appears that Infancy Thomas (or at least some of its contents) may have been enjoyed (and influential) well beyond early Christian circles. The several works in the category of ‘apocryphal acts’ likewise were widely read and influential for centuries.29 Of the five main works in this category (Acts of John, Acts of Paul, Acts of Peter, Acts of Andrew, Acts of Thomas), only the Acts of Thomas survives completely, but the extant portions of the others, along with references to them in other Christian 25.  Stephen Gero, ‘The Infancy Gospel of Thomas: A Study of the Textual and Literary Problems’, NovT 13 (1971): 46–80 (56). 26.  The story appears in Infancy Thomas 6:13–23. Gero, ‘The Infancy Gospel of Thomas’, 63, cites references to earlier versions of the story in Ep. Apos. 4 (extant only in Ethiopic) and Irenaeus, Adv. Haer. 1.13.1 (1.20.1 in ANF 1:344–45). 27.  Gero, ‘The Infancy Gospel of Thomas’, 48–56, discusses versions in Greek, Latin, Syriac, Arabic, Armenian, Georgian, Ethiopic and Old Church Slavonic. 28.  Ibid., 75. 29.  Hans-Josef Klauck, The Apocryphal Acts of the Apostles, trans. Brian McNeil (Waco: Baylor University Press, 2008), is an excellent recent introduction to these texts. He proposes (p. 3) the following approximate dates: Acts of John (ca. 150–160 CE), Acts of Paul (ca. 170–180 CE), Acts of Peter (ca. 190–200 CE), Acts of Andrew (ca. 200–210 CE), Acts of Thomas (ca. 220–240 CE). A roughly similar order and dating is proposed by Jan Bremmer, ‘The Apocryphal Acts: Authors, Place, Time and Readership’, in The Apocryphal Acts of Thomas, ed. Jan Bremmer (Leuven: Peeters, 2001), 148–70, esp. 153–54.

11. Who Read Early Christian Apocrypha?

191

texts, clearly show a great interest in all these works. As a collection, all five works were known and widely read among Manichaeans in particular.30 But also, despite critical comments about them by figures such as Eusebius (HE 3.3.2; 3.25.4), these works enjoyed varying levels of popularity and usage across wider circles of early Christianity as well. Indeed, in the case of a portion of the Acts of Peter that recounts his martyrdom, we have material taken into the liturgy of some ancient churches.31 Moreover, characters and scenes in this work (e.g., Peter’s crucifixion upside-down) and the other apocryphal acts were also popular subjects in Christian art.32 We have only to think of the enormous place of the figure of Thecla in Christian tradition and art, a figure that seems to have emerged influentially in the Acts of Paul, this account subsequently generating other texts about her (e.g., the Life and Miracles of St. Thecla dated to the fifth century).33 To be sure, the Acts of John exhibits certain emphases (especially in §§87–105) that raised doubts about its orthodoxy (e.g., Eusebius, HE 3.25.6), leading to condemnation of the text at the Second Council of Nicaea in 787 CE. Klauck judged that in this material the text exhibited certain ‘points of contact’ with ‘the Eastern version of Valentinianism’, but predating ‘the elaboration of the great gnostic systems’.34 But it is likely that for at least some Christian readers more generally Acts of John and the other apocryphal Acts were read variously as interesting, provocative, and edifying texts. I return to this point later in this discussion. Some recent studies emphasize certain similarities between the apocryphal Acts and Roman-era ‘pagan’ novels. Jan Bremmer firmly contended, ‘In fact, the intertextuality of the AAA [Apocryphal Acts of the Apostles] with the novel cannot be doubted’, underscoring similar motifs in both bodies of texts.35 Christine Thomas posited strong similarities to what she called ‘historical novels’ in particular, fictional accounts featuring characters regarded by writers and readers as real figures of history, and she 30.  Ibid., 3–5, cites key evidence of Manichaean usage. 31.  R. J. Bauckham, ‘The Martyrdom of Peter in Early Christian Literature’, ANRW 2.26/1 (1992): 539–95; Klauck, Apocryphal Acts, 100–105. 32.  Cartlidge and Elliott, Art and the Christian Apocrypha, 134–235, for numerous examples and extended discussion. 33.  On Thecla in Christian art, see ibid., 148–62. Gilbert Dagron, Vie et miracles de sainte Thècle (Bruxelles: Société des Bollandistes, 1978), gives the Greek text and French translation. 34.  Klauck, Apocryphal Acts, 17–18. Consequently, he regards Acts of John as the earliest of the apocryphal acts, proposing a date of ca. 150–160 CE. 35.  Bremmer, ‘Apocryphal Acts’, 164.

192

Texts and Artefacts

proposed the Alexander romance as providing ‘the best generic parallel among the novelistic products of the Roman Empire’.36 Moreover, she contended that the ‘narrative fluidity’ of these apocryphal texts, ‘their existence in multiple translations, redactions, abridgments, and expansions’, comprises evidence of the popularity of the material that they contain among various ‘audiences’ over time and place.37 I noted earlier that some scholars have taken the prominent place of Christian women in some of the apocryphal Acts as indicating that these texts were intended for, and were particularly read by, Christian women. To cite another scholar who takes such a view, Bremmer confidently contended that women were both ‘the actual readers’ and the intended readers of these texts. He further proposed that the authors of the apocryphal Acts had a ‘missionary’ intention, particularly aiming for the attention, and conversion, of upper-class women, a conclusion he judged ‘inescapable’ in light of the ‘female focus’ of these writings.38 But we have also already noted the critical assessment of this stance by HainesEitzen. Moreover, in likening the apocryphal Acts to the ancient historical novel, a type of writing that likely enjoyed a readership of males and females, Thomas seems implicitly to come down closer to Haines-Eitzen than to Bremmer. To be sure, women may well have been among those who enjoyed the apocryphal Acts (and other early Christian writings), but there does not seem to be a good a reason to make them the predominant readers. Instead, we should probably allow for a diversity of people for whom these texts were meaningful in various ways. The Artefacts of Ancient Reading I turn now to consider evidence that is often overlooked in studies of early Christian apocrypha and their likely readers, the extant physical artefacts of ancient readers, the remains of early manuscripts of these texts. It has not been adequately recognized that the physical features of ancient manuscripts can give us hints of the kinds of readers for whom they were 36.  Christine M. Thomas, The Acts of Peter, Gospel Literature, and the Ancient Novel (New York: Oxford University Press, 2003), esp. 87–105, citing 89. She distinguishes between ancient ‘historical novels’ and ‘erotic novels’ in that the former ‘are all “referential” texts that narrate major historical events, however constitutive novelistic techniques of embellishment may also be for their genre’, whereas the erotic novels ‘all avoid this sort of referentiality, focusing on the private events of obscure characters’ (93). 37.  Ibid., 89. 38.  Bremmer, ‘Apocryphal Acts’, 166–67.

11. Who Read Early Christian Apocrypha?

193

copied, and the settings in which they were read.39 Here also, because of the limitations of space, I shall confine the discussion to some examples. Perhaps the best known Christian apocryphal text today is the Gospel of Thomas, and so the portions of three manuscripts of this text that are dated palaeographically to the third century which the earliest copies extant, provide a good place to start.40 Although one must be cautious in doing so, it is interesting to consider whether the comparative number of extant copies of a given text may reflect its comparative popularity in the ancient setting.41 In the case of the Gospel of Thomas, the portions of three early manuscripts certainly indicate some ancient interest in this text. To be more specific, on the one hand, it is not among the most frequently attested (cf., e.g., the 12 copies of Matthew or 16 copies of John). By my calculation, the remains of three copies of the Gospel of Thomas make the text tie for thirteenth place (with James, Ephesians, Leviticus and Acts of Paul) in the number of extant copies among literary texts in Christian manuscripts the second and third centuries CE. On the other hand, for many Christian texts, both canonical and extra-canonical, we have portions of only one or two copies from this early period.42 39.  I emphasize this point more broadly in an earlier publication: L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). More recently, see L. 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), 49–62 (republished as Chapter 6 of the present volume). The pioneering studies influential for my own work, and very much important still, include Eric G. Turner, Greek Papyri: An Introduction (Oxford: Oxford University Press, 1980), particularly 74–96; and C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University Press, 1979). 40.  According to one tabulation, by 1985 there were already more than 600 scholarly books and articles on the Gospel of Thomas, and there are surely many more now: G. J. Riley, ‘The Gospel of Thomas in Recent Scholarship’, CR:BS 2 (1994): 227–52. Among more recent publications is the multi-author volume, Das Thomasevangnelium: Entstehung – Rezeption – Theologie, ed. Jörg Frey, Enno Edzard Popkes and Jens Schröter (Berlin: de Gruyter, 2008). I draw here on my contribution to that volume, ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’, 19–32 (republished as Chapter 10 of the present study), which gives the details on which I base the present discussion. 41.  I have surveyed the comparative number of all the extant copies of texts of Christian provenance from before 300 CE, tentatively considering what we may infer in Earliest Christian Artifacts, 24–41. 42.  Compare, e.g., figures for each NT writing of which there is evidence from before 300 CE. I indicate in round brackets the number of extant copies: Matthew (12),

194

Texts and Artefacts

But what more might these fragments of early copies tell us about who read the Gospel of Thomas and the circumstances in which it was read? Papyrus Oxyrhynchus 1 is a portion of one leaf of a papyrus codex, the extant text roughly corresponding to sayings 26–30 (+77b) and 31–33 of the Coptic text from Nag Hammadi. The original page-size of the codex (ca. 27 × 10–13 cm) makes it analogous to some other Christian codices of the same period containing literary texts, whether (what became) canonical or extra-canonical ones. The copyist-hand likewise fits within the spectrum of hands that we see in some other early Christian literary manuscripts, in this case a copyist aiming for clarity but of limited calligraphic abilities, as reflected in the inconsistency in letter sizes, the ligatures and an inability to produce truly ‘bilinear’ writing.43 It is, however, perhaps more interesting that, in comparison with some other early Christian manuscripts of equivalent page-size (particularly copies of texts known to have been treated as scripture in early Christian circles), the letters are somewhat smaller and the number of lines per page (ca. 37–38) somewhat larger, producing a more compressed format. A smaller number of lines per page suggests a text intended for greater ease of reading, perhaps public reading, whereas a larger number of lines per page may suggest a copy intended for individual reading/study.44 Moreover, the intriguing page-number (ια = 11) on the verso side of the leaf (apparently added by an early reader) suggests that some other text (which, unfortunately, we cannot identify) preceded the Gospel of Thomas, and so this codex combined two or more texts. The desire to combine multiple texts in one codex would also help explain the somewhat more compressed writing mentioned above. In sum, the physical features of P.Oxy. 1 (e.g., compressed writing, absence of punctuation, etc.) suggest that it more likely was copied (as some sort of compendium of certain texts?) for personal reading and study, rather than for public reading (e.g., in corporate worship).

Mark (1), Luke (7), John (16), Acts (7), Romans (4), 1 Corinthians (2), 2 Corinthians (1), Galatians (1), Ephesians (3), Philippians (2), Colossians (1), 1 Thessalonians (3), 2 Thessalonians (2), Philemon (1), Titus (1), Hebrews (4), James (3), 1 Peter (1), 2 Peter (1), 1 John (1?), 2 John (1), Jude (2), Revelation (5). For discussion of these figures, see Hurtado, Earliest Christian Artifacts, 20–21. 43.  ‘Bilinear’ means letters of fairly even height, the line of letters set evenly within imaginary top and bottom lines, which is one of the key earmarks of a more skilled copyist. 44.  See my discussion of the matter in The Earliest Christian Artifacts, 171–77.

11. Who Read Early Christian Apocrypha?

195

In P.Oxy. 654 we have remnants of a re-used papyrus roll (an ‘opisthograph’), the extant portion of Gospel of Thomas copied on the outer side of a roll, the inner side (with horizontal fibres) containing the remains of a survey-list. As for the copy of Gospel of Thomas, the complete inability at bilinear writing, the irregularities in letter-size and formation, and other features of the copyist-hand (errors in spelling, a bizarre first line) suggest someone of very limited skill. There are horizontal lines extending from the left margin and into the text to signal the first line of a saying of Jesus, these apparently added by a reader rather than by the copyist. Here, too, the physical features of this manuscript suggest that it was prepared for private study. This was the typical purpose of re-used rolls.45 P.Oxyrhynchus 655 comprises fragments of a small papyrus roll, the small majuscule characters confidently and skilfully formed, forming neat, narrow columns of ca. 12–16 characters per line. The original height of the roll was about 16 cm (inclusive of top and bottom margins), yet with some 30 lines of text per column. This was clearly a manuscript toward the small/compact end of the spectrum of sizes of literary rolls.46 As a modern comparison, volumes in the Loeb Classical Library series have pages of just over 16 cm height, and a maximum of ca. 32 lines of text per page. As the case with the Loeb volumes, P.Oxy. 655 was almost certainly intended (and used) for personal reading. Another text that is witnessed in remnants of early manuscripts is the Gospel of Mary (Magdalene).47 Although its text is preserved most 45.  Opisthographs are ‘usually taken as a sign of an economy-minded collector’, according to George W. Houston, ‘Papyrological Evidence for Book Collections and Libraries in the Roman Empire’, in Ancient Literacies: The Culture of Reading in Greece and Rome, ed. William A. Johnson and Holt N. Parker (Oxford: Oxford University Press, 2009), 233–67 (257). He cites Julian Krüger, Oxyrhynchos in der Kaiserzeit: Studien zur Topographie und Literaturrezeption, Europäische Hochschulschriften 441 (Frankfurt am Main: Lang, 1990), 161, as showing that of all recovered Oxyrhynchus papyri 17.9% were opisthographs. 46.  William Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004), 141–43, concluded that heights of rolls of literary texts in this period were typically 25–33 cm. On miniature manuscripts, see Thomas Krauss, ‘Die Welt der Miniaturbücher in der Antike und Spätantike. Prolegomena und erste methodische Annäherungen für eine Datensammlung’, SNTSU 35 (2010): 79–110. 47.  See now, C. M. Tuckett, The Gospel of Mary (Oxford: Oxford University Press, 2007), for full and recent discussion of issues concerning this fascinating text. For a helpful summary-discussion, see C. M. Tuckett, ‘The Gospel of Mary’, in The Non-Canonical Gospels, ed. Paul Foster (London: T&T Clark International, 2008), 43–53.

196

Texts and Artefacts

extensively in a fifth-century Coptic copy, it is the remains of two early Greek copies that will concern us here. The one, P.Oxy. 3525, is a scrap measuring 11.5 × 12 cm, the text written in a cursive hand on the ‘recto’ (the side of papyrus with horizontal fibres), the ‘verso’ side blank, which strongly suggests that it is a portion of a roll.48 The cursive hand, the roll book-form, and the modest size of the roll, all make it rather certain that this is the remnant of someone’s personal copy of the text. The other Greek remnant of the Gospel of Mary is P.Ryl. 463, a papyrus fragment (8.9 × 9.9 cm) with writing on both sides and pagenumbering, which make it the remains of a leaf of a codex.49 In this copy, the writing is majuscule and a hand described by Roberts as ‘clear and upright [but] also ugly and ill-proportioned, and show[ing] considerable cursive influence’.50 Tuckett noted the numerous mistakes made by the copyist, some of them corrected, and some of the mistakes producing Greek text that ‘makes little sense’ such that at these points it can only be understood on the basis of the later Coptic translation.51 Here, again, in view of these data, plus the small size of the codex (ca. 11 × 16 cm), we likely have an informal copy intended for someone’s personal reading of the text.52 The general point I wish to make here is that, based on the nature of the remnants of the early manuscripts of the Gospel of Thomas and the Gospel of Mary, in neither case do we have any reason to link these copies with distinctive circles of Christians. That is, it would be dubious to posit a circle of ‘Thomas’ Christians or ‘Mary’ Christians as connected with these manuscripts. Indeed, I suggest that this should be taken as illustrative more widely of how apocryphal gospels functioned. 48.  Peter Parsons, ‘3525. The Gospel of Mary’, The Oxyrhynchus Papyri, Volume 50 (London: Egypt Exploration Society, 1983), 12–14. There is an excellent photograph of the fragment in Tuckett, The Gospel of Mary, among the photos placed after p. 110. 49.  Tuckett, The Gospel of Mary, provides photographs of both sides, among photos after p. 110. He accurately identifies the recto and verso side, correcting the labels attached to the items by the Rylands Library. 50.  C. H. Roberts, ‘463. The Gospel of Mary’, in Catalogue of Greek Papyri in the John Rylands Library, Volume III (Manchester: Manchester University Press, 1938), 18–23 (20). 51.  Tuckett, The Gospel of Mary, 84. ‘Overall, the scribe appears to have been somewhat careless’, citing a similar judgment by Roberts, ‘The Gospel of Mary’, 20. 52.  Roberts estimated a writing-column of ca. 7.5 × 12 cm (‘The Gospel of Mary’, 20). My estimated page-size allows for margins of at least 1.5 cm left, right and top, and 2 cm bottom. For further discussion of the significance of the size of codices, see Hurtado, The Earliest Christian Artifacts, 155–65.

11. Who Read Early Christian Apocrypha?

197

There were obviously Christians who wrote these and other gospel-texts featuring figures such as Thomas and Mary, and there were obviously other Christians who enjoyed reading these texts, as reflected in the remnants of early copies of them, and the subsequent translations of these and other such texts (e.g., Coptic and other languages). But the features of the extant artefacts of the early reading and readers suggest that these texts were (typically?) copied for, and read by, individuals, the texts likely circulated and copied among those Christians who expressed an interest in them. As to the social connections of these individuals, at the most, we should probably imagine loose networks of sorts, rather than defined circles or sects of Christians.53 I take it to be congruent with this view that the remnants of these texts were found in the same locations (e.g., in the same mounds in Oxyrhynchus) as the remnants of familiar Christian texts such as copies of writings of the OT and the emergent NT, and writings by Irenaeus, Melito, and others, all of which are typically taken as reflective of ‘mainstream’ Christianity of the time. This suggests that the ‘apocryphal’ texts were likely read by at least some in the same Christian circles in which these other texts also functioned. Along similar lines, consider again Wisse’s contention that the so-called gnostic texts likely circulated primarily among individual Christians who were intrigued by the alternative ideas in them.54 Whether or not the Gospel of Thomas or the Gospel of Mary, for example, is taken as ‘gnostic’ by scholars today, the earliest extant artefacts of these texts in themselves give no basis for positing discrete circles of Christians devoted particularly to these texts and their distinctive teachings. Instead, as noted already, these early manuscripts appear to have been prepared for individual usage. Of course, discrete Christian circles of various sorts may well be posited on other grounds (e.g., the claims of writers such as 53.  It takes us into broader issues than can be handled adequately here, but my proposal of loose networks of somewhat like-minded individuals has some resonance with some recent explorations of the circumstances in which so-called gnostic writings were copied, translated and read, in particular Michael A. Williams, Rethinking ‘Gnosticism’: An Argument for Dismantling a Dubious Category (Princeton: Princeton University Press, 1996), 235–62, and Stephen Emmel, ‘The Coptic Gnostic Texts as Witnesses to the Production and Transmission of Gnostic (and Other) Traditions’, in Das Thomasevangelium: Enstehung – Rezeption – Theologie, ed. Jörg Frey, Enno Edzard Popkes and Jens Schröter, BZNW 157 (Berlin: de Gruyter, 2008), 33–49. 54.  Wisse, ‘The Use of Early Christian Literature as Evidence for Inner Diversity and Conflict’.

198

Texts and Artefacts

Irenaeus or Eusebius). But my point here is that, whatever such circles there were, the sorts of texts that we have been considering also likely circulated and were read, perhaps mainly, by interested individuals who otherwise were members of ‘mainstream’ Christian churches of their time and place. This also seems to me broadly in accord with Haines-Eitzen’s view cited earlier about the likely readers of the apocryphal Acts, based on her analysis of the extant manuscripts of these texts from the first six centuries.55 As noted already, she was particularly concerned to test contentions by other scholars that these writings were read especially by women, concluding against this view.56 She also weighed the view that these writings had a ‘popular’ readership, i.e., people of relatively lower social levels, and found this, too, dubious. The 13 manuscripts of apocryphal Acts that she surveyed are all codices (as we would expect generally for Christian literary texts increasingly after the third century).57 Moreover, she judged that they show no particularly distinctive physical features in copyist hands or other matters in comparison with copies of other texts (including scriptural texts), making it ‘highly problematic to continue to argue for the popular readership – or the popular/female readership – of the Apocryphal Acts of the Apostles’.58 Instead, she proposed that the readers of these texts were simply individuals (more likely ‘members of the upper echelons’) who enjoyed the stories of the adventures (and martyrdoms) of heroic Christian figures.59 For the purpose of this discussion, the main point is that the apocryphal Acts, as also the case with the other Christian apocryphal literature, found readers who for various reasons were interested in the contents of these texts. In some cases, this or that apocryphal text might have been of special significance for a given circle of Christians. But, more generally 55.  Haines-Eitzen, The Gendered Palimpsest, 53–64. 56.  Ibid., 57. See also pp. 58–61 for her brief comments on several illustrative examples of manuscripts of apocryphal Acts. 57.  In the earliest period from which we have manuscript remains (second– third centuries CE), Christian preference for the codex seems to have been most pronounced for copies of texts that functioned as scripture. For example, we have no NT writing on an unused roll from this period. Christians did, however, use the roll for other literary texts (e.g., theological treatises), though even for these texts the codex was more often used. For further discussion, see Hurtado, The Earliest Christian Artifacts, 53–61, and the list of texts in early Christian manuscripts, 209–29, which includes information about book-form and other matters. 58.  Haines-Eitzen, The Gendered Palimpsest, 62. 59.  Ibid., 63.

11. Who Read Early Christian Apocrypha?

199

it seems, as a diverse body of writings, early Christian apocrypha likely drew a diversity of readers, most of whom were simply interested individuals who found the texts variously intriguing, entertaining, inspiring, provocative, or edifying, and perhaps other readers who found them dubious or even objectionable. Who read early Christian apocrypha? The best general answer would seem to be a variety of people who took an interest in these diverse texts for a diversity of reasons.

Chapter 12 P 4 5 as a n E a rly C h r i s t i an A rt e fact : W h at i t R ef l ect s a b out E a r ly C hr i st i ani ty *

Among the invaluable collection of biblical papyri in the Chester Beatty Library, for NT scholars the ‘jewel in the crown’ is likely P45 (P.Chester Beatty I).1 It is the earliest extant four-Gospel codex (palaeographically dated mid-third century CE), and also includes our most extensively preserved early copy of Acts. For these reasons, this manuscript is extremely important for questions about the formation of the fourfold Gospel and the NT canon, as well as for the textual history of the five particular writings that it contains.2 But P45 is also highly important as an *  This is a lightly revised version of a presentation given as part of the celebration of Reidar Hvalvik’s sixty-fifth birthday held on 29 September 2016. I reiterate here my respect and appreciation for his contributions to scholarship on the NT and early Christianity. The present form of the essay retains something of the character of that presentation to a mixed audience of scholars and the general public 1.  But some might well argue that P46 (P.Chester Beatty II) is a contender for pride of place, as it may be dated a little earlier, and is likewise a historic artefact, as our earliest example of a Pauline Corpus in one codex. The classic study remains Günther Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum, The Schweich Lectures, 1946 (London: Oxford University Press for the British Academy, 1953; reprint Wipf & Stock, 2007), although his larger theory of Christian recensional activity in early second-century Alexandria seems not to have won a wide number of adherents. 2.  Frederic G. Kenyon, The Chester Beatty Biblical Papyri, Fasciculus II, The Gospels and Acts: Text (London: Emery Walker, 1933); idem, The Chester Beatty Biblical Papyri, Fasciculus II, The Gospels and Acts: Plates (London: Emery Walker, 1934). High-resolution digital photos of the Chester Beatty biblical papyri are now also available online in the Center for the Study of New Testament Manuscripts: http://www.csntm.org/manuscript/View/GA_P45. Other/subsequent fragments have been identified: Hans Gerstinger, ‘Ein Fragment des Chester Beatty-Evangelienkodex in der Papyrussammlung der Nationalbibliothek in Wien’, Aegyptus 13 (1933): 67–72; T. C. Skeat and B. C. McGing, ‘Notes on Chester Beatty Biblical Papyrus I

12. P45 as an Early Christian Artefact

201

early Christian book, a physical artefact, and so in this discussion I wish to consider this remarkable papyrus book from various angles, seeking to discover what we can of its ‘story’.3 In what follows, I draw upon previous scholarly analyses of the extant portions of the original manuscript and judgments about its contents, size, construction, and textual character.4 My purpose here is simply to illustrate the various kinds of data that are generated in such analyses, and their relevance for the wider historical study of early Christianity.

P45 and the Fourfold Gospel Let us begin by considering P45 as our earliest extant four-Gospel codex, which gives us artefactual confirmation that the familiar four NT Gospels had acquired a special significance and comprised a closed set of such writings (at least, in many Christian circles) by the third-century (Gospels and Acts)’, Hermathena 150 (1991): 21–25; Thomas J. Kraus, ‘Ad Fontes ‒ The Benefit of the Consultation of Original Manuscripts as for Instance P.Vindob.G. 31974’, in Ad Fontes. Original Manuscripts and Their Significance for Studying Early Christianity: Selected Essays (Leiden: Brill, 2007), 25–45. See Brent Nongbri, ‘The Acquisition of the University of Michigan’s Portion of the Chester Beatty Biblical Papyri and a New Suggested Provenance’, Archiv für Papyrusforschung 60 (2014): 93–116, for discussion of how portions of the codex got scattered and sold to various purchasers. The extant portions of the texts included in P45 are given in the list of manuscripts in the Nestle-Aland Novum Testamentum Graece, 28th ed., 794. 3.  This reflects the emphasis in my book, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006); and idem, ‘What Do the Earliest Christian Manuscripts Tell Us About Their Readers?’, in The World of Jesus and the Early Church: Identity and Interpretation in Early Communities of Faith, ed. Craig A. Evans (Peabody, MA: Hendrickson, 2011), 179–92. A key previous study of P45 is T. C. Skeat, ‘A Codicological Analysis of the Chester Beatty Papyrus Codex of the Gospels and Acts (P45)’, Hermathena 155 (1993): 27–43, republished in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup 113 (Leiden: Brill, 2004), 141–57, the latter cited hereafter. As well, note Gunther Zuntz, ‘Reconstruction of One Leaf of the Chester Beatty Papyrus of the Gospels and Acts (P45)’, Chronique d’Egypte 26 (1951): 191–211; Skeat and McGing, ‘Notes on Chester Beatty Biblical Papyrus I’. 4.  The precise provenance of P45 remains uncertain, and need not detain us here. For discussion of its acquisition, see Charles Horton, ‘The Chester Beatty Biblical Papyri: A Find of the Greatest Importance’, in The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels ‒ The Contribution of the Chester Beatty Gospel Codex P45, ed. Charles Horton (London: T&T Clark International, 2004), 149–60, esp. 157–58.

202

Texts and Artefacts

date of this manuscript. Later in this discussion, I consider the physical feat involved in constructing a single papyrus codex adequate for the four Gospels and Acts. At this point, I focus on the relevance of P45 for questions about the formation of a four-Gospel ‘canon’. It is, I think, commonly accepted that at least by the latter half of the second century the familiar four Gospels had acquired (or were acquiring) a wide acceptance as a closed list of accounts of Jesus, and in many Christian circles were treated as scriptures. Of course, other ‘Jesus books’ were circulating and continued to be produced, but the four familiar NT Gospels by then were increasingly treated as forming a distinct and closed group. Irenaeus’ defence of these four Gospels in the late second century CE against the alleged interests of ‘gnostics’ in other gospels is typically cited as key evidence of this (Adv. Haer. 3.11.7–9). But Oskar Skarsaune judged that already in Justin Martyr, a few decades earlier than Irenaeus’ text, we see ‘an incipient canon in the way he refers to the Gospels’.5 Similarly, Darrell Hannah contended that the Epistula Apostolorum, which he dates to the 140s CE, gives evidence of a four-Gospel collection earlier than Irenaeus and perhaps even slightly earlier than Marcion’s single-Gospel canon.6 In further support of an early date for a four-Gospel collection, James Kelhoffer argued that the ‘long ending’ of Mark draws upon the other NT Gospels and was likely circulating by ca. 120–140 CE.7 Others, such as Heckel, Hengel and Stanton, also lodged their support for the view that the notion of a delimited ‘fourfold’ Gospel likely formed sometime in the first half of the second century.8 More recently still, Charles Hill has argued a vigorous case for this position.9 5.  Oskar Skarsaune, ‘Justin and His Bible’, in Justin Martyr and His Worlds, ed. Sara Parvis and Paul Foster (Minneapolis: Fortress, 2007), 53–76 (76). 6.  Darrell D. Hannah, ‘The Four Gospel “Canon” in the Epistula Apostolorum’, JTS 59 (2008): 598–633. 7.  James Kelhoffer, Miracle and Mission: The Authentication of Missionaries and Their Message in the Longer Ending of Mark, WUNT 2/112 (Tübingen: Mohr Siebeck, 2000). 8.  G. N. Stanton, ‘The Fourfold Gospel’, NTS 43 (1997): 317–46; Theo K. Heckel, Vom Evangelium des Markus zum viergestaltigen Evangelium, WUNT 120 (Tübingen: Mohr [Siebeck], 1999); idem, ‘Jesus Traditions and Gospels in Justin Martyr and Irenaeus’, in The Biblical Canons, ed. J.-M. Auwers and H. J. De Jonge, BETL 143 (Leuven: Leuven University Press, 2003), 353–70; Martin Hengel, The Four Gospels and the One Gospel of Jesus Christ (London: SCM, 2000). More recently, see also Charles E. Hill, Who Chose the Gospels: Probing the Great Gospel Conspiracy (Oxford: Oxford University Press, 2010). 9.  Charles E. Hill, ‘A Four-Gospel Canon in the Second Century? Artifact and Arti-Fiction’, EC 4 (2013): 310–34.

12. P45 as an Early Christian Artefact

203

There are, to be sure, remnants of other manuscripts of approximately the same date as P45, or even earlier, that may give us further indications that early Christians sought to combine the ‘fourfold’ Gospel collection in a single book. Papyrus Bodmer XIV–XV (P75) is particularly interesting, preserving generous portions of Luke and John; these two Gospels clearly in this order in a single ‘gathering’ (or ‘quire’) that originally consisted of 72 leaves (36 folded sheets). This manuscript is typically dated to the early third century CE.10 It is possible, of course, that the codex originally contained only Luke and John, but Theodore Skeat proposed, plausibly in my view, that the manuscript originally may have comprised two gatherings, Matthew and Mark contained in another, now-lost, gathering/ portion of the same manuscript.11 Nevertheless, to underscore the point, P45 remains our earliest extant incontrovertible four-Gospel codex.12 It is 10.  Victor Martin and Rodolphe Kasser, Papyrus Bodmer XIV, Évangile de Luc, chap. 3–24 (Cologny-Genève: Bibliotheca Bodmeriana, 1961); Victor Martin and Rodolphe Kasser, Papyrus Bodmer XV, Évangile de Jean chap. 1–15 (ColognyGenève-: Bibliotheca Bodmeriana, 1961). More recently, further fragments of the manuscript have been identified: Marie-Luise Lakmann, ‘Papyrus Bodmer XIV–XV (P75): Neue Fragmente’, Museum Helveticum 64 (2007): 22–41; and James M. Robinson, ‘Fragments from the Cartonnage of P75’, HTR 101 (2008): 231–52. Luke ends and John begins on the same page, making the order of these certain. But cf. the recent challenges to the early dating of these Bodmer papyri by Brent Nongbri: ‘The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P. Bodmer II (P66)’, Museum Helveticum 71 (2014): 1–35; idem, ‘Reconsidering the Place of Papyrus Bodmer XIV–XV (P75) in the Textual Criticism of the New Testament’, JBL 135 (2016): 405–37. 11.  Theodore Skeat, ‘The Origin of the Christian Codex’, ZPE 102 (1994): 263–68, republished in J. K. Elliott, ed., The Collected Writings of T. C. Skeat, NovTSup 113 (Leiden: Brill, 2004), 79–87 (cited here), esp. 80–81. Hengel (The Four Gospels, 41) endorsed this proposal. However, I do not find Skeat’s proposal persuasive that Christians initially adopted the codex to accommodate the four Gospels in the early second century CE, and that the single-Gospel codex (dominant in the second and third centuries) developed subsequently. This seems to be backwards to the data. 12.  Skeat’s proposal that P4 (fragments of Luke) was part of the same codex as P64 and P67 (fragments of Matthew) has been contested and defended; cf. T. C. Skeat, ‘The Oldest Manuscript of the Four Gospels?’, NTS 43 (1997): 1–34, republished in The Collected Writings of T. C. Skeat, 158–92; Peter M. Head, ‘Is P4, P64 and P67 the Oldest Manuscript of the Four Gospels? A Response to T. C. Skeat’, NTS 51 (2005): 450–57; Scott D. Charlesworth, ‘T. C. Skeat, P64 + 67 and P4, and the Problem of Fibre Orientation in Codicological Reconstruction’, NTS 53 (2007): 582–604. Charles E. Hill, ‘Intersections of Jewish and Christian Scribal Culture: The Original Codex Containing P4, P64 and P67, and Its Implications’, in Among Jews, Gentiles, and Christians in Antiquity and the Middle Ages, ed. Reidar Hvalvik and

204

Texts and Artefacts

a priori unlikely, however, that it was the first such codex ever constructed. It would be extraordinary if the first experimental four-Gospel codex just happened to survive! So, we must presume that P45 reflects similar, and quite plausibly, earlier efforts by Christians to put into physical form the closed circle of the four Gospels that became canonical.13 But the somewhat varying textual affiliations of the individual Gospels in P45 (a topic to which I return later in this discussion) reflect the separate prior transmission and textual history of the individual Gospels that preceded the gathering of them into a single codex (whenever the latter first took place).14 Likewise, the apparent order of the Gospels in P45 (Matthew, John, Luke, Mark) bears witness to this same prior J. Kaufman (Trondheim: Tapir Academic Press, 2011), 75–91, defended Skeat’s proposal and judged that his claim that these fragments come from a four-Gospel codex ‘remains viable’ (78). Even more uncertain, however, is H. A. Sanders’ suggestion that P53 (P.Mich. 6652), comprising one leaf of Matthew and one leaf of Acts, may derive from the same codex, which might have included the four Gospels and Acts: H. A. Sanders, ‘A Third Century Papyrus of Matthew and Acts’, in Quantulacumque: Studies Presented to Kirsopp Lake, ed. R. P. Casey, S. Lake and A. K. Lake (London: Christophers, 1937), 151–61. Cf. Joseph van Haelst, Catalogue des papyrus littéraires juifs et chrétiens (Paris: Publications de la Sorbonne, 1976), no. 380. But Hill recently judged it ‘likely, though of course not certain’, that these two leaves are also remnants of a four-Gospel codex: ‘A Four-Gospel Canon in the Second Century?’, 323 n. 46. In any case, in P45 we have fragments of all four Gospels, making it undeniably the earliest four-Gospel codex. 13.  Among the Chester Beatty papyri are other artefacts of early Christian efforts to combine multiple scripture-texts in one codex. In addition to P45, the other most famous one among NT scholars, of course, is P46 (P.Chester Beatty II), which, as mentioned earlier, is our earliest extant codex comprising a collection of Paul’s letters. In addition, there is P.Chester Beatty VI (Rahlfs 963), the remains of Numbers and Deuteronomy (second/third century CE; van Haelst, Catalogue, no. 52), and P.Chester Beatty IX–X (Rahlfs 967), remnants of Ezekiel, Daniel (with Susanna and Bel and the Dragon) and Esther. Portions of this same codex are held in other locations, as described in van Haelst, Catalogue, §315. 14.  See, e.g., recent analyses by James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTS 36 (Leiden: Brill, 2007), 103–6; and comments on P45 by several contributors to The Early Text of the New Testament, ed. Charles E. Hill and Michael J. Kruger (Oxford: Oxford University Press, 2012): Tommy Wasserman, ‘The Early Text of Matthew’, 83–107 (91–94); Peter M. Head, ‘The Early Text of Mark’, 108–20 (114–17); Juan Hernández Jr., ‘The Early Text of Luke’, 121–39 (128–30); Juan Chapa, ‘The Early Text of John’, 140–56 (150–52). The fragmentary nature of the remnants of the Gospels in P45 (especially Matthew and John) makes it difficult to conduct the sort of thorough analysis of textual

12. P45 as an Early Christian Artefact

205

separate transmission of them. The more familiar order found in most manuscripts reflects the sequence in which Irenaeus refers to their composition (Adv. Haer. 3.1.1), with Gospels ascribed to apostles forming the first and last of the group.15 In the so-called Western order of the Gospels in P45 (reflected also in Codex Bezae and Codex W and some other witnesses), however, the Gospels ascribed to apostles (Matthew and John) appear first, in declining length, followed by Luke and Mark, again in declining length.16 So, it appears that the Gospels were initially transmitted individually and then brought together physically in one sequence or another as in P45 and other early manuscripts.17 It is relevant to take account of Trobisch’s thesis that there was a fixed edition of the NT by ca. 150 CE that included an ordering of its writings. But this seems to me to founder on the evidence of our earliest manuscripts, which show, for relationships that we could wish. But in Matthew, John and Luke P45 does not align firmly with any of the major witnesses to the familiar text-types, exhibiting a mixture of readings of ‘Alexandrian’ and ‘Western’ identity, although it has none of the larger and distinctive readings of Codex Bezae. In Acts, P45 is closer to the ‘Alexandrian’ witnesses than to the ‘Western’ ones. I have shown that in Mark P45 has a markedly close relationship to Codex W, but that neither manuscript is related to the so-called Caesarean text-type: Larry W. Hurtado, Text-Critical Methodology and the Pre-Caesarean Text: Codex W in the Gospel of Mark, SD 43 (Grand Rapids: Eerdmans, 1981), confirmed recently by Tommy Wasserman, ‘P45 and Codex W in Mark Revisited’, in Mark, Manuscripts, and Monotheism: Essays in Honor of Larry W. Hurtado, ed. Chris Keith and Dieter T. Roth (London: Bloomsbury T&T Clark, 2014), 130–56. 15.  Bruce M. Metzger, The Canon of the New Testament: Its Origin, Development, and Significance (Oxford: Clarendon, 1987), includes an appendix discussing the ordering of NT writings in ancient manuscripts (295–300). See 296–97 for a list of the various (nine!) orders of the Gospels with supporting witnesses. See also Hengel (The Four Gospels, 38–47) for a vigorous discussion of the data. Is there any significance that the ‘Western’ order of the Gospels seems to reflect Justin’s reference to the Gospels in two groups, those ‘memoirs’ written by apostles and those written ‘by those who followed them’ (1 Apol. 103.8)? 16.  The probability that the Gospels stood in the ‘Western’ order in P45 is based mainly on the observation that Mark and Acts alone have distinctive marks added by a reader, suggesting that Mark was last among the Gospels, immediately followed by Acts. 17.  This accounts for the significant differences in the numbers of extant early copies of each Gospel, as I discussed in The Earliest Christian Artifacts, esp. 20–24. Cf. Skeat, ‘The Origin of the Christian Codex’, who (unconvincingly in my view) proposed that the four Gospels initially circulated together and only thereafter as individual texts.

206

Texts and Artefacts

example, that well into the third century there was no single fixed ordering of the Gospels.18 Moreover, contrary to some assertions, the Christian adoption of the codex bookform did not instigate, enable, or shape the development of a fourfold Gospel or the NT canon.19 The emergence of the fourfold Gospel and the subsequent formation of the complete NT canon comprise one process, and the Christian preference for the codex is a quite distinguishable phenomenon, and it was not essential to the process of canon formation. As illustration of this, in roughly the same period ancient Judaism developed a canon without recourse to the codex!20 It is more accurate to conclude that the early Christian preference for the codex had its own initial impetus and continuing rationale, and that the emergence of a growing body of writings that Christians treated as scripture then promoted their various efforts to construct codices adequate to contain multiple texts, eventuating in the ‘pandects’ of the fourth century and later (e.g., Codex Vaticanus and Codex Sinaiticus) that contained the entire NT or even the entire Christian canon; that is, two factors, the early Christian preference for the codex and the emergence of a body of writings treated as scripture in Christian circles, combined to promote efforts to develop codex technology adequate to accommodate collections of writings in a physical form.

P45 and Acts As noted earlier, in addition to the four NT Gospels, P45 also gives us the earliest substantial amount of the text of Acts; all the other extant pre-Constantinian manuscripts of Acts consisting of a single page or part of a page.21 As to its textual affiliation in Acts, Tuckett recently judged that P45 is ‘very similar in general terms to the Alexandrian text, but also exhibits some freedom in reproducing the sense of the text’, showing ‘rather less concern to reproduce the precise wording down to the tiniest 18.  Cf. David Trobisch, Die Endredaktion des Neuen Testaments. Eine Unter­ suchung zur Entstehung der christlichen Bibel (Freiburg: Universitätsverlag Freiburg; Göttingen: Vandenhoeck & Ruprecht, 1996); English edition: The First Edition of the New Testament (Oxford: Oxford University Press, 2000). 19.  Cf. Martin Wallraff, Kodex und Kanon: Das Buch im frühen Christentum, Hans-Lietzmann-Vorlesungen 12 (Berlin: de Gruyter, 2013). 20.  See now Timothy H. Lim, The Formation of the Jewish Canon (New Haven: Yale University Press, 2013) 21.  For a recent review of the evidence, see Christopher Tuckett, ‘The Early Text of Acts’, in Hill and Kruger, eds., The Early Text of the New Testament, 157–74, which includes references to earlier studies. See 165–67 for his analysis of P45.

12. P45 as an Early Christian Artefact

207

detail’.22 This characterization broadly reflects the results of other recent studies of the overall character of the textual transmission reflected in P45, such as the detailed analysis by Royse.23 With the great majority of the other early papyri of Acts, P45 basically shows that something close to the ‘Alexandrian’ text-type was probably dominant in/by the third century CE, albeit with a certain spectrum in the transmission tendencies exhibited, and the papyrus reflects the absence of the more expansionist type of text of Acts found in Codex Bezae.24 However, the particular datum that I wish to draw attention to here is the inclusion of Acts with the four Gospels in one book. In other (later) manuscripts, Acts is more typically linked with the Catholic Epistles, serving, it seems, as a kind of narrative framework for them.25 But P45 suggests a treatment of Acts as a kind of sequel to the Gospels narratives of Jesus’ ministry. Andrew Gregory showed that in Irenaeus (Against Heresies) and the Muratorian Fragment we have instances of Acts likely read as part of the same work as the Gospel of Luke, and also 22.  Ibid., 167. 23.  Royse, Scribal Habits, 105–97. As with most other scholars, Royse ascribed the textual character of P45 almost entirely to ‘the scribe’ of the manuscript. But we should also probably allow for the influences of readers/users of NT writings, as urged by Ulrich Schmid, ‘Scribes and Variants: Sociology and Typology’, in Textual Variation: Theological and Social Tendencies?, ed. H. A. G. Houghton and D. C. Parker (Piscataway, NJ: Gorgias, 2008), 1–23; and Michael W. Holmes, ‘Codex Bezae as a Recension of the Gospels’, in Codex Bezae: Studies from the Lunel Colloquium, June 1994, ed. D. C. Parker and C.-B. Amphoux (Leiden: Brill, 1996), 123–60. Of course, the copyist could also have been an interested reader, but my point is that intentional changes more likely came from reading rather than copying, whereas as accidental changes likely happened in the copying process. 24.  The early papyrus of Acts that comes closest to Codex Bezae is P38 (P.Mich. inv. 1571). See Tuckett, ‘Early Text of Acts’, 162–65, who granted that P38 is ‘closely related’ to the text found in Bezae but by no means identical to it. Instead, P38 may witness to an earlier stage of ‘the textual tradition of the Western text’ (165). Also useful is the earlier study by Peter Head, ‘Acts and the Problem of Its Texts’, in The Book of Acts in Its First Century Setting. Vol. 1, Ancient Literary Setting, ed. Bruce W. Winter and Andrew D. Clarke (Grand Rapids: Eerdmans; Carlisle: Paternoster, 1993), 415–44. 25.  E.g., in P74, Acts is followed by the Catholic Epistles, and in Codex Sinaiticus, the order of NT writings is the Gospels, Pauline Epistles, Acts, the Catholic Epistles, Revelation, and then Barnabas and Shepherd of Hermas. In Codex Vaticanus, and Codex Alexandrinus, Acts appears after the Gospels and just before the Catholic Epistles, the Pauline Epistles coming afterwards. There are still other placements in the manuscript tradition.

208

Texts and Artefacts

as complementing the four Gospels as a group.26 Kavin Rowe argued, however, that in the early centuries Acts was typically read in connection with the four Gospels collectively, and not simply or primarily as a sequel to Luke. Indeed, Rowe contended that Luke and Acts were written and released separately (albeit by the same author), and that in the earliest centuries the two texts were never circulated or taken as parts of one work.27 It is neither possible nor necessary here to engage these particular issues adequately. Instead, I simply want to complement the recent textual studies of the early reception of Acts by drawing attention to P45 as an important physical artefact of early Christian usage of Acts, and an apparent regard for Acts as scripture, as evidenced in its inclusion with the four Gospels in this manuscript.28 Granted, P45 is dated several decades later than Irenaeus’ Against Heresies, in which he discusses Acts in connection with the Gospels. However, given its Egyptian provenance, P45 appears to show either that Irenaeus reflects a wider trans-local use of Acts in early Christian circles in connection with the Gospels, or, perhaps, that Irenaeus’ use of Acts had a subsequent influence upon other early Christian circles trans-locally. Certainly, there is evidence that Irenaeus’ Against Heresies was read in Egypt not long after its composition, and so his influence there is fully plausible.29 26.  Andrew Gregory, The Reception of Luke and Acts in the Period Before Irenaeus, WUNT 2/169 (Tübingen: Mohr Siebeck, 2003), 38–43. 27.  C. Kavin Rowe, ‘History, Hermeneutics and the Unity of Luke–Acts’, in Rethinking the Unity and Reception of Luke and Acts, ed. Andrew F. Gregory and C. Kavin Rowe (Columbia, SC: University of South Carolina Press, 2010), 43–65 (this essay originally published in JSNT 28 [2005]: 131–57); and also in the same volume, idem, ‘Literary Unity and Reception History: Reading Luke–Acts as Luke and Acts’, 74–81. But cf. Andrew F. Gregory, ‘The Reception of Luke and Acts and the Unity of Luke–Acts’, 82–93, who urges that ancient readers such as Irenaeus could well have read Acts both in connection with the fourfold Gospel and other texts and also as particularly connected to the Gospel of Luke, and that it is just as likely that Luke and Acts were composed and released initially together as parts of one work. Cf. also the arguments that Luke and Acts come from different authors in Mikeal C. Parsons and Richard I. Pervo, eds., Rethinking the Unity of Luke and Acts (Minneapolis: Fortress, 1993). 28.  See also Jens Schröter, ‘Die Apostelgeschichte und die Entstehung des neutestamentlichen Kanons. Beobachtungen zur Kanonisierung der Apostelgeschichte und ihrer Bedeutung als kanonischer Schrift’, in Auwers and De Jonge, eds., The Biblical Canons, 395–429, esp. 398–418, on the early reception of Acts and its emerging place in the formation of the NT canon. 29.  I refer, of course, to P.Oxy. 405, fragments of a bookroll palaeographically dated late second or early third century CE containing the work.

12. P45 as an Early Christian Artefact

209

In any case, the inclusion of Acts with the Gospels in P45 reflects a reading of Acts as complementing the Gospels in some way.30 Irenaeus drew upon Acts to affirm the validity of the apostles collectively, over against Marcion and other sectarian tendencies, and the inclusion of Acts with the Gospels in P45 may well reflect physically such a ‘protoorthodox’ stance, affirming Acts and its more inclusive treatment of the various apostolic figures.31 But, again, the placement of Acts in P45, separated from Luke (in this case by Mark) but following the four Gospels, appears to make Acts more a sequel to the Gospels collectively, and not particularly the second volume of ‘Luke–Acts’. Book-form and Construction I turn now to consider the book-form and construction of P45. As a first observation, P45 obviously reflects the distinctive early Christian preference for the codex, a preference exhibited especially for those texts treated as scriptures in the early centuries. It is well known that this early Christian preference for the codex contrasts with the overwhelming preference for the bookroll (scroll) more generally in ancient Roman contexts, especially for literary texts.32 Illustrative of this, about 95% of extant second-century non-Christian copies of literary texts are bookrolls, whereas at least 75% of all second-century Christian manuscripts of literary texts are codices. There is a similarly marked difference between Christian and non-Christian book-form preferences in the third century CE as well.33 However, this is not the appropriate occasion to explore further the reasons for this Christian preference, but it is worth noting that it was already established by the time of our earliest extant Christian manuscripts. 30.  Cf. Schröter, ‘Die Apostelgeschichte’, 407, who appears to endorse the proposal of Trobisch, Endredaktion, 52–53, that P45 may have also included the Catholic Epistles, Acts serving as their introduction; but Trobisch’s proposal is entirely speculative. 31.  Schröter, ‘Die Apostelgeschichte’, 423–27, concludes that Acts became part of the emerging NT canon precisely as a text that affirmed the authority of the apostles collectively, over against Marcion and gnostic circles. 32.  William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004), gives now the most complete discussion of the Roman-era bookroll. 33.  See my fuller discussion of the early Christian preference for the codex in The Earliest Christian Artifacts, 43–93; and my survey of the matter with updated data in my book, Destroyer of the Gods: Early Christian Distinctiveness in the Roman World (Waco, TX: Baylor University Press, 2016), 133–38.

210

Texts and Artefacts

Moreover, I repeat that it seems to have been expressed with particular force for copies of those literary texts that Christians read as scriptures. Second- and third-century Christian copies of OT writings and the Christian writings that came to form part of the NT are almost entirely codices, whereas among copies of other early Christian texts of the same period (e.g., theological tractates), about one-third are bookrolls.34 So, given this general Christian preference for the codex, if a given text was copied in a codex, that by itself does not mean that the text was regarded as scripture. But we can say that a bookroll copy of a Christian literary text likely means that the text (or at least that copy) did not function as scripture. However, there are also other things to take account of in considering the physical features of P45. One of these is the size of the manuscript. For comparison, among ancient bookrolls and codices some are classified as ‘miniatures’, which must have served as personal copies of texts that could be carried more easily on one’s person and/or taken along on a journey.35 Among early Christian examples, there are remnants of a copy of the Gospel of Thomas (P.Oxy. 655) from a miniature bookroll (originally, ca. 16cm height), both the bookroll format and the miniature size confirming that this manuscript was prepared for personal/private usage.36 Likewise, the early third-century copy of the Gospel of Mary (P.Ryl. 463) was a miniature codex.37 In the case of P45, however, we have 34.  Whereas about 65% of other second/third-century copies of Christian literary texts are codices, at least 95% of writings that came to form the Christian canon are codices. Only a few Christian copies of some OT writings are bookrolls, and the only bookrolls with NT texts are re-used bookrolls, these copies likely made for personal reading. 35.  Thomas Kraus, ‘Die Welt der Miniaturbücher in der Antike und Spätantike. Prolegomena und erste methodische Annäherungen für eine Datensammlung’, SNTSU 35 (2010): 79–110; idem, ‘Miniature Codices in Late Antiquity: Preliminary Remarks and Tendencies About a Specific Book Format’, EC 7 (2016): 134–52. An example is P62 (P.Osl. inv. 1661), fragments of a Greek-Coptic bilingual codex, whose pages originally measured ca. 6.6 × 5.6 cm, as described by Leiv Amundsen, ‘Christian Papyri from the Oslo Collection’, Symbolae Osloensis 24 (1945): 121–40. 36.  For further discussion of this and the other earliest manuscripts of this text, see Larry Hurtado, ‘The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655’, in Das Thomasevangelium: Entstehung – Rezeption – Theologie, ed. Jörg Frey, Enno Edzard Popkes and Jens Schröter, BZNW 157 (Berlin: de Gruyter, 2008), 19–32 (republished as Chapter 10 of the present study). 37.  C. H. Roberts, ‘463. The Gospel of Mary’, in Catalogue of the Greek Papyri in the John Rylands Library, Vol. 3 (Manchester: Manchester University Press, 1938), 18–23.

12. P45 as an Early Christian Artefact

211

‘a substantial volume’ with an original page size of about 25 cm height and 20 cm width, originally comprising 112 folios, or 224 pages, with ‘a thickness of perhaps 5–6 cm. apart from any binding’.38 This reflects a well-represented codex size among third-century manuscripts, a full-size book.39 Accommodating in one codex the considerable amount of text comprised by the four Gospels and Acts required some thought. For one thing, the copyist had to try to get a fair amount of text on each page in order to keep the total number of pages within reason. So, P45 has a larger number of lines per page (36–39) than most other early NT manuscripts, and also a large number of letters per line (averaging ca. 50).40 Nevertheless, the letters are not particularly crowded, and the spacing between the lines (in printer’s terms, the leading) is generous, indicating a desire to facilitate reading the text. As additional reader’s aids, the original copyist deployed punctuation ‘occasionally’, marking off sense-units roughly corresponding to sentences, and a diaeresis more generally over an initial iota or upsilon.41 So, clearly, this book was designed to package a large body of text, but it was also prepared to facilitate usage by readers. However, there are differences of opinion on the precise intended usage of the book. Was it prepared for liturgical usage, to be read from in corporate worship? Or was it prepared for some individual, for personal study and edification?42 The curious marks added by some early reader in 38.  I draw here upon Skeat’s calculations in ‘A Codicological Analysis’, 156. 39.  Its dimensions place P45 in ‘Group 4’ in the classification of codices by Eric G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977; reprint, Eugene, OR: Wipf & Stock, 2010), 16. 40.  Zuntz, ‘Reconstruction’, 194, noted lines varying from 47–48 letters in length (fol. 12r) to as many as 57 (fol. 14v), but judged the average to be about 50 letters per line. Skeat, ‘Codicological Analysis’, 145, calculated about 53 stichoi per page, whereas Kraus, ‘Ad Fontes’, 33, estimated about 39–40 lines per page. If Skeat’s stichoi designate lines, it is not clear why or how Skeat’s calculation differs so markedly. Kraus also noted that P45 has more letters per line than other early NT manuscripts. P75 (P.Bodmer XIV–XV) likewise has about 40 lines per page, but its pages have very different dimensions, ca. 26 × 13 cm, and so narrower lines of ca. 25–30 letters. 41.  See, e.g., Kenyon, Chester Beatty Biblical Papyri, Gospels and Acts: Text, ix, for brief description of visual features of the writing. 42.  Peter M. Head, ‘The Early Text of Mark’, in Hill and Kruger, eds., The Early Text of the New Testament, 108–20, judged that, despite the small and compact nature of the hand, the manuscript may have been used ‘in public reading’ (115). Cf. Kraus, ‘Ad Fontes’, 34, who posited that, although ‘a usage for public purposes…is by no means and generally ruled out…’, the codex ‘was hardly meant and prepared

212

Texts and Artefacts

Mark and Acts (which appear to function as a reader’s aid in identifying sense-units) surely suggest the subsequent preparation of the text (at least those portions) for reading, probably for public reading.43 Also, in this early period, the task of constructing a codex to contain such a large amount of text required some thought.44 The typical (and simplest) form of earliest Christian codices was a single stitched ‘gathering’ (or ‘quire’) of multiple sheets of papyrus, reflecting what seems to have been the earliest and basic technique for making a codex generally at the time. This was adequate for a text of moderate size, such as an individual Gospel, which could be accommodated with a modest number of folded sheets of writing material. Moreover, initially, the single-gathering construction was also used by Christians for larger bodies of text. For example, P46, the Chester Beatty Pauline codex, was originally a single gathering of 52 folded sheets of papyrus.45 But there was a practical upper limit to how many sheets could be accommodated in this method of codex construction.46 For one thing, when such a single-gathering codex was closed, the inner pages protruded and had to be trimmed, which resulted in them being narrower than the outer ones, and the more sheets in the gathering, the more the inner sheets had to be trimmed. There was obviously a point of diminishing returns! So, in the case of P45, the copyist chose another method. The original 56 papyrus sheets were folded individually and then stitched together, each folded sheet a gathering/quire comprising two folios or four pages, amounting to 224 pages in all.47 This technique avoided having pages of for public reading, but probably for private purposes or a usage within a relatively limited circle’. 43.  These marks are in a darker ink and are shaped like a / placed at the top of a line after the final word in a sense-unit (roughly like our modern sentence). They are visible in Kenyon’s facsimile volume and also online: http://www.csntm.org/ manuscript/View/GA_P45. 44.  See especially Turner, Typology of the Early Codex, 55–71, ‘How a Codex Was Made Up’, for further examples of the various forms of codex construction. It is often overlooked that the early Christian preference for the codex required the development of skills additional to those required for copying in a bookroll. 45.  Frederic G. Kenyon, The Chester Beatty Biblical Papyri, Fasciculus III Supplement, Pauline Epistles: Text (London: Emery Walker, 1936), ix. 46.  P.Chester Beatty IX–X, a single-gathering codex that contained Ezekiel, Daniel and Esther, was composed of 59 folded sheets (118 leaves, 236 pages). Turner, Typology of the Early Codex, 58–59, lists single-gathering codices, one of which ran to more than 280 pages (or 70 folded sheets)! 47.  Skeat, ‘Codicological Analysis’, 156.

12. P45 as an Early Christian Artefact

213

different widths. Also, it may have meant that the copyist did not have to calculate in advance exactly how many sheets were needed. He could simply add individual folded sheets as required to complete the task. But, of course, the resultant book was not very strong structurally. If the binding thread broke, the individual folded sheets could readily fall apart. There was, however, yet another method of codex construction experimented with in the third century that proved more successful in the end. Two or more sheets of writing material were folded together to form a gathering, the codex then made up of multiples of these gatherings or quires. The Chester Beatty codex of Genesis (P.Chester Beatty V; Rahlfs 962), for example, seems to have been composed mainly of gatherings/ quires of five folded sheets (ten leaves each).48 In time, of course, this last method of multiple-gatherings/quires won out, and four folded sheets comprising eight leaves or sixteen pages, became the standard ‘quire’ for codices, and for the printed book to this day.49 So, I think that the various experiments with codex-construction reflected in extant third-century Christian papyri show that Christians of that time were at the leading edge of experimentation in codex construction, pushing what originated as a simple device mainly used for sub-literary texts into a bookform for extended bodies of literary texts. For if codex technology had already been settled for serious literary purposes, early Christians would scarcely have spent their efforts in these various experiments in codex construction. Instead, to reiterate an earlier observation, we must conclude that the prior Christian preference for the codex, especially for their sacred texts, combined with the desire (at least in some circles) to have multiple texts in one manuscript, drove these experiments with various forms of codex construction. The point to note here is that the particular method of codex construction used in P45 seems to have been an early experiment that was superseded in time by the multiplegathering method.50 We should also take note of the ‘hand’, which is consistent throughout the whole codex, indicating that it is entirely the product of one copyist. Moreover, this was a skilled copyist, well able to produce a readable text 48.  Frederic G. Kenyon, The Chester Beatty Biblical Papyri, Fasciculus IV, Genesis: Text (London: Emery Walker, 1934), vii. 49.  The term ‘quire’ derives from the Latin quaternio, referring to a set of four sheets of writing material. 50.  Turner, Typology of the Early Codex, 60, designated the codices constructed of individual folded sheets ‘Uniones’, and judged that the early dates of most of the extant examples make this method ‘contend with single-quire codices to be regarded as the earliest form of codex…’

214

Texts and Artefacts

whose letters are ‘amazingly even’ in size, the orthography, punctuation, and use of nomina sacra ‘astonishingly consistent’. The beginnings of lines are arranged ‘with great regularity’, producing ‘an almost faultless, straight line’ down the left-hand margin. The right-hand endings of lines vary a bit, though generally by only one or two letters, and there is an effort at regular line-widths, sometimes involving a filler-mark (>) at the end of a shorter line, and sometimes by starting a word on one line and then completing it on the next.51 In short, although the hand of P45 is not ‘calligraphic’, it obviously exhibits a trained and experienced copyist who was concerned to produce a very readable text and was capable of doing so.52 The copyist’s skill is also reflected in the comparatively small number of nonsense readings and corrections. However, the copyist was not faultless. Royse observed ‘a marked tendency to omit portions of the text, often (as it seems) accidentally, but perhaps also by deliberate pruning’. He also noted frequent harmonization, particularly harmonization to the immediate context, and occasional attempts at stylistic and grammatical ‘improvements’.53 The accidental omissions (usually only a word or two), and probably at least some of the small variations in word-order as well, are probably attributable to the copyist. However, I think that the apparently intentional variants are more often/likely traceable to readers/users of the texts.54 For example, harmonizations to immediate context, especially those involving harmonization to the subsequent context, could require reading ahead, sometimes a few or more lines ahead, noting differences and then considering how best to handle them. This amounted to an early exegesis of the text to determine meaning. In my view, it is not likely that copyists would usually take the time for such matters. It is more likely, instead, that a reader studying the text would, for example, take account of the context and sometimes make changes to harmonize an earlier 51.  The phrases quoted and the general characterization of the hand are drawn from Zuntz, ‘Reconstruction’, 192–93, confirmed by my own analysis of the Kenyon facsimiles. 52.  Skeat, ‘A Codicological Analysis’, 157. 53.  Royse, Scribal Habits, 197. See 103–97 for Royse’s extended and detailed analysis. 54.  Ulrich Schmid, ‘Scribes and Variants: Sociology and Typology’, in Houghton and Parker, eds., Textual Variation, 1–23; and Michael W. Holmes, ‘Codex Bezae as a Recension of the Gospels’, in Parker and Amphoux, eds., Codex Bezae, 123–60, who contend (cogently in my view) that copyists mainly copied, and that readers were mainly responsible for intentional variants.

12. P45 as an Early Christian Artefact

215

statement with a subsequent one. Similarly, intentional stylistic changes seem to me more often plausibly assigned to readers rather than copyists, for these changes, too, required someone to take the time to consider introducing them to ‘improve’ the text. Then, when that manuscript was copied, such intentional changes could be introduced into the subsequent textual tradition, copyists simply copying the text before them in their exemplars.55 Whoever made such changes, however, P45 gives us artefactual evidence that at least some early Christians (whether copyists or readers) both revered texts such as the four Gospels and also felt a certain freedom to make minor alterations in them. However, the changes were from a ‘friendly’ motive. They appear mainly designed to smooth the texts, the phrasing adjusted to make them conform to stylistic preferences and conventions of those who read them, and occasionally removing what seemed ambiguities. Among the intentional changes, only a comparatively small number could seem intended to alter the meaning of the text.56 The Nomina Sacra and the Staurogram As typical of many early Christian manuscripts (especially copies of biblical texts), P45 shows us the early Christian scribal practice now typically referred to as the ‘nomina sacra’, the distinctive abbreviations of certain key words in early Christian discourse, with a distinctive horizontal stroke placed over the abbreviated forms. Four words in particular are written as nomina sacra with special regularity: ΘΕΟΣ, ΚΥΡΙΟΣ, ΙΗΣΟΥΣ, and ΧΡΙΣΤΟΣ, which are typically represented by 55.  In taking this view, I therefore think that Royse’s ascription of all the variants in early papyri such as P45 to ‘scribal habits’ is a bit over-simplified, and that he fails to take adequate account of the role of readers/users of texts. But in email discussion of the matter, Royse indicated that he found it difficult (even dubious) to try to identify ‘intentional’ variants, and remains persuaded that scribes played an active role in generating the several types of variants that he noted. 56.  Contra the impression one might take from the title of Bart Ehrman’s widely noticed book, although there are some variants that reflect doctrinal concerns, there is in fact no evidence of a concerted attempt to alter the NT writings in a ‘proto-orthodox’ direction. Cf. Bart D. Ehrman, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament (New York: Oxford University Press, 1993); and, e.g., Frederick Wisse, ‘The Nature and Purpose of Redactional Changes in Early Christian Texts: The Canonical Gospels’, in Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission, ed. W. L. Petersen (Notre Dame: University of Notre Dame Press, 1989), 39–54.

216

Texts and Artefacts

the first and final letters of their inflected forms.57 Space does not permit me here to discuss the matter at any length, except to note briefly that in P45 Jesus’ name is written in a manner that differs from this pattern, using the first two letters: ΙH. This form of Jesus’ name seems to reflect an early practice that was then succeeded by the first-letter/last-letter convention typically used for the other key words that came to be treated as nomina sacra.58 So, this form of Jesus’ name in P45 reflects its early dating and likely the way Jesus’ name was treated in its exemplar. P45 also exhibits another striking device that is now referred to typically as the ‘staurogram’. This device involves the vertical line of the Greek capital letter rho superimposed on the Greek capital letter tau. In P45, as in a few other early Christian manuscripts (e.g., P66 and P75), this device is used in writing forms of the words σταυρος and/or σταυροω, and the intended effect seems to be to depict visually a crucified figure, the loop of the rho the head on a T-shaped cross.59 This makes these uses of this device in P45 and a few other early manuscripts likely the earliest Christian visual references to the crucified Jesus.60

P45 and the Early Text of the NT Writings This brings me finally to consider briefly the import of P45 for our sense of how the NT writings were transmitted in the earliest period to 57.  The classic and highly influential study is Colin H. Roberts, Manuscript, Society and Belief in Christian Egypt (London: Oxford University Press, 1979), 26–48. Other words were also treated as nomina sacra in early Christian manuscripts, although not with the same regularity. In P45, these include the inflected forms of Πατηρ, Υιος, Σταυρος, Πνευμα, and (apparently) one instance of Χριστιανος in Acts 11:26. As well, the words were often written in three-letter forms, e.g., χρς. 58.  In an earlier publication I proposed that the nomina sacra practice may have derived from this abbreviation of Jesus’ name: Larry W. Hurtado, ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998): 655–73 (republished as Chapter 7 of the present volume). 59.  The device appears in P45 at Matt 26:2 and Luke 14:27, as shown by Dieter T. Roth, ‘P45 as an Early Christian Artifact: Considering the Staurogram and Punctuation’, in Mark, Manuscripts, and Monotheism: Essays in Honor of Larry W. Hurtado, ed. Chris Keith and Dieter T. Roth (London: Bloomsbury T&T Clark, 2014), 118–29. Robin Margaret Jensen, Understanding Early Christian Art (London: Routledge, 2000), 138, refers to the staurogram as ‘a kind of pictogram’ of a crucified figure. More recently, see Bruce W. Longenecker, The Cross Before Constantine: The Early Life of a Christian Symbol (Minneapolis: Fortress, 2015), 106–10. 60.  See, e.g., my discussion of the device in Artifacts, 135–54, and also Chapter 8 of this volume.

12. P45 as an Early Christian Artefact

217

which we have access.61 As noted earlier in this discussion, the textual affiliation/character of the writings included in P45 varies somewhat. I note again that in Acts, P45 is markedly closer to the ‘Alexandrian’ text (‫ א‬B A C) than to the ‘Western’ text (D it), with ‘none of the major variants characteristic of the text of D in Acts’.62 As for the Gospels, in the case of Matthew there is little extant text on which to base any confident judgment about textual affiliation in P45.63 In Mark, P45 aligns clearly with Codex W, but the textual tradition from which they both developed seems to me to have been something like that represented in Sinaiticus and Vaticanus.64 In a recent study, Juan Hernández judged that ‘P45’s close ties to primary Alexandrian witnesses prevail throughout the text of Luke’.65 Juan Chapa has recently described the text of John in P45 as ‘at a mid-way point between the “Alexandrian” and the “Western” MSS’, but with none of the larger variants found in the latter witnesses, and he labelled P45 as exhibiting ‘a “free” text’, characterized by numerous small variants of the kind already noted.66 In his study focused on ‘singular’ readings, Royse characterized P45 similarly.67 Nevertheless, despite the numerous small variants, including the obviously intentional ones, in P45 we recognize easily the NT writings in question. There are no indications of major deletions, insertions, or re-arrangements in their texts. In short, although there is a good deal of ‘microlevel fluidity’ involving the sort of small variants noted, there appears to be ‘macrolevel stability’ in the texts of these writings in P45.68 61.  Larry W. Hurtado, ‘P45 and the Textual History of the Gospel of Mark’, in Horton, ed., The Earliest Gospels, 132–48; idem, ‘The New Testament in the Second Century: Text, Collections and Canon’, in Transmission and Reception: New Testament Text-Critical and Exegetical Studies, ed. J. W. Childers and D. C. Parker (Piscataway, NJ: Gorgias, 2006), 3–27 (republished as Chapter 1 of the present volume). But the major resource now is the multi-author work: Hill and Kruger, eds., The Early Text of the New Testament. 62.  Kenyon, Chester Beatty Biblical Papyri, Gospels and Acts: Text, xviii; and also idem, ‘Some Notes on the Chester Beatty Gospels and Acts’, in Casey and Lake, eds., Quantulacumque, 145–48. Also now, Tuckett, ‘The Early Text of Acts’, 165–67. 63.  So, recently, Wasserman, ‘The Early Text of Matthew’, 93–94. 64.  E.g., Head, ‘The Early Text of Mark’, 114, echoing my own judgment in Text-Critical Methodology, 88–89. 65.  Hernández, ‘The Early Text of Luke’, 129–30. 66.  Chapa, ‘The Early Text of John’, 150–52. 67.  Royse, Scribal Habits, e.g., his summary, 197. 68.  I borrow these terms and the important distinction that they represent from a programmatic essay by Michael W. Holmes, ‘From “Original Text” to “Initial Text”: The Traditional Goal of New Testament Textual Criticism in Contemporary

218

Texts and Artefacts

This forms a striking contrast with the evidence concerning the transmission of some other early Christian texts, such as the Gospel of Thomas and Shepherd of Hermas, which are marked by much more substantial variation.69 Granted, P45 exhibits the exercise of a certain level of textual freedom in the many small variants in word-order, tense, etc., which contrasts with the apparently more reserved and fixed transmissionpractice reflected in, for example, P75. However, P45 does not justify any notion of a ‘wild’ or chaotic state of the text of the Gospels and Acts in the early period it reflects.70 Conclusion I intend this study as more illustrative than exhaustive, or particularly original, for the analysis of P45 as an early Christian artefact with a ‘story’ to tell could be extended further. I hope, however, that the foregoing will at least show that there is more to note about early Christian manuscripts than the wording of their texts. In the case of P45, we have a manuscript specifically prepared to combine the four canonical Gospels and Acts in one codex-volume, giving us material confirmation of the special character and status that these writings had obtained by the time the manuscript was prepared. P45 also gives us physical evidence that, although the fourfold Gospel was by then a closed set of the familiar four texts (at least in a growing number of Christian circles), these texts had been transmitted individually, and there was not yet agreement on how to order them when they were placed in one book. P45 attests the early Christians’ preference for the codex bookform, and particularly for copies of those texts that were to be used as scripture. The method used to construct P45 also shows us that well into the third century Christians were still experimenting with various ways to build codices adequate for such a sizeable body of texts. P45 tells us that the early Christians’ commitment to the codex, especially for their scriptures, Discussion’, in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes, 2nd ed. (Leiden: Brill, 2013), 637–88, esp. 670–77. Kenyon’s basic description was similar: Gospels and Acts: Text, xix–xx. 69.  See, e.g., Hurtado, ‘The Greek Fragments of the Gospel of Thomas as Artefacts’; Carolyn Osiek, Shepherd of Hermas: A Commentary (Minneapolis: Fortress, 1999), esp. 1–4. 70.  To be sure, the very fragmentary nature of the manuscript makes it impossible to judge whether it included either of the two major ‘macrolevel’ variants in the Gospels: the Pericope Adulterae in John and the ‘long ending’ of Mark (16:9–20).

12. P45 as an Early Christian Artefact

219

led them to engage in these ambitious efforts, which placed Christians then at the leading edge of codex development. Likewise, P45 attests the noteworthy scribal devices that distinguish early Christian manuscripts, the ‘nomina sacra’ and the ‘staurogram’. These probably reflect early Christian scribal efforts to express their faith and devotion visually. The nature of the texts in P45 bears witness to the freedom that some early Christians felt to alter their wording, but also the limits of that freedom. In P45, we see the activities of early Christian copyists and readers as well, the manuscript serving as a precious artefact of these ancient believers and their handling of these texts.

I n d ex of R ef er e nce s Hebrew Bible/ Old Testament Genesis 14:14 128, 129, 131, 148

Amos 1:6 75 1:11 75 1:13 75 2:1 75

Exodus 23:20 133

Micah 2:5 79

Leviticus 3:12 123 4:27 123 24:16 122

Zechariah 3:8 129

Numbers 13:16 133 21:8–9 149 Deuteronomy 10:8 66 17:2 66 23:2–4 79 23:9 79 1 Samuel 2:11 66 2:18 66 3:1 66 1 Chronicles 28:8 79 Psalms 1:1 153 Ezekiel 9:4 149

New Testament Matthew 1:1–9 34 1:12–13 34 1:14–20 34 1:17 130 1:23 34 2:13–16 34 2:22 34 3:1 34 3:9–15 34 3:9 34 3:15 34 3:16–4:3 34 5:20–22 34 5:25–28 34 6:27–29 177 10:17–23 30 10:25–32 30 12:24–26 34 12:32–33 34 13:55–56 6, 34 14:3–5 6, 34 20:24–32 34 21:34–37 6, 34 21:43 6, 34

21:45 6, 34 23:30–39 34 23:30–34 6 23:35 6 23:39 6 24:3–6 34 24:12–15 34 25:41–26:40 34 25:41–26:39 143 26:2 142 26:7–8 34 26:10 34 26:14–15 34 26:22–23 34 26:31–33 34 31:13–19 34 31:34–37 34 31:43–45 34 Mark 16:9–20

42, 218

Luke 1:2 76 2:20 79 5:1 76 5:11 76 5:21 76 5:25–26 79 7:16 79 8:12 76 8:13 76 8:15 76 9:23 142 12:25–27 177 13:13 79 14:27 142 17:15 79

221 18:43 79 22:44–56 30 22:61–64 30 23:21 142 23:23 142 23:26 142 23:33 142 23:47 79 24:7 142 24:20 142 John 1–14 141 1:1–14:26 33 1:1–51 34 1:34 7 2:1–25 34 2:11–22 30 3:1–36 34 4:1–54 34 5:1–47 34 6:1–71 34 6:11–35 33 7:1–52 34 7:53–8:11 42 8:1–11 34 8:12–59 34 9:1–41 34 10:1–42 34 11:1–57 34 12:1–50 34 13:1–38 34 14–21 142 14:1–31 34 14:29–21:9 33 15:1–27 34 15:25–16:2 156, 158, 159 15:25 158, 163 15:26 164 16:1–4 34 16:2 158 16:3–20 159 16:6–7 34 16:10–33 34 16:21–32 156, 158 16:21 158, 164

Index of References 16:22 164 16:23 164 16:28 164 16:32 158, 164 17:1–26 34 18:1–40 34 18:31–33 6 18:31 109 18:32 109 18:36–19:1 6 18:36–40 34 18:37–38 6 19 33 19:1–42 34 19:1–7 34 19:2–7 6 19:6 141, 142 19:15 141 19:16 141, 142 19:17 142 19:18 141, 142 141, 142 19:19 19:20 142 19:25–28 33 19:25 141, 142 19:31–32 33 19:31 141 20:1–20 34 20:22–31 34 20:31 130 21 141 21:1–9 34 21:18–20 34 21:23–25 34 Acts 1:21 2:17

65, 75 68, 70, 73–75 2:21 68, 73 2:25 66 2:33 75 68, 73, 75 2:34 2:36 66, 75 2:38 75, 132 2:39 66, 74 3:6 133

3:20 65 3:22 66, 74 4:4 71 4:21 79 4:26 65 4:28 73 4:30 73 71, 76 4:31 4:33 65, 75 5:3–21 29, 30, 34 66, 68, 73, 5:9 74 5:19 66 5:29 64 5:39 64 6:2 71, 76 71, 77 6:4 6:7 68, 70, 74, 75 6:8 75 7:25 73 7:31 65, 66 7:33 65 7:49 66 7:55 64 7:59 73 8:14–24 66 8:14 71, 76 65, 71 8:16 8:20 66 8:21 66 66, 77 8:22 8:24 66–70, 77 67–70, 75 8:25 8:26 66 8:39 66 9:1 65 9:11 65 73, 133 9:14 9:15 65 9:17 65 9:21 73 9:27 65 65, 71 9:28 9:31 65 9:35 65 9:42 65

222 Acts (cont.) 10:33 68–70 10:36 66 10:44 71 71, 76 11:1 11:16 65 11:17–18 73 65, 71 11:17 11:18 79 11:19 71 11:20–21 73 65, 66, 73 11:21 11:23 64, 73, 77, 79 11:24 73 11:26 216 12:7 66 66, 68–71 12:11 12:17 66, 71 64, 76 12:22 12:23 66 66, 68–70, 12:24 74–76 13:2 66 66, 68–70, 13:5 75 13:7 71, 76 13:10–11 66, 73 13:10 68, 69, 73 66, 73 13:11 13:12 65, 73 13:43 64, 77, 79 68–70, 75, 13:44 76 13:46 71, 76 13:47 65 68–70, 75, 13:48 76, 79 13:49 76 14:3 64, 65, 77, 79 14:10 75 14:12 71 14:25 68, 69, 71 14:26 64, 77, 79 15:11 64, 65, 75, 77, 79

Index of References 15:14 77 15:17 66, 68–70, 77 15:19 77 15:26 65, 71 15:35 71, 76 15:36 71, 76 15:40 64, 68–70, 77 16:6 68–71, 75, 76 68, 69, 71, 16:7 77 16:10 68–70, 75, 77 16:11–15 78 16:14 78 16:15 68–70, 78 16:18 133 16:25–34 78 16:31 65, 78 16:32 68–70, 75, 78 16:34 78 17:11 71 17:13 71, 76 66, 72 17:24 17:27 68–70, 72 17:29 72 18:4 75 65, 72, 75 18:8 18:9 68, 69, 72 18:11 71, 76 18:25 69 18:26 68–70 18:28 68 19:1 75 65, 71, 132 19:5 19:10 71, 76 19:13 65 19:14–17 133 19:17 65, 71 19:20 68–70, 74 20 65 20:3 75 20:7 71 20:21 65

20:24

64, 65, 77, 79 68, 69, 74 20:25 20:28 68–70, 78, 80 20:32 64, 68–70, 75, 77, 79 65, 79 20:35 21:13 65, 71 21:14 68–70, 79 68–70, 79 21:20 22:10 65 22:14 79 22:16 73 23:11 65 45 218 75 218 Romans 1:10 79 2:18 79 4:23–5:3 29, 30 5:8–13 29, 30 10:9 133 12:2 79 15:17–21 25 15:32 79 1 Corinthians 1:2 71, 133 5:3–4 133 5:4 71 6:11 71 12:3 133 2 Corinthians 10:9–11 25 Ephesians 5:17 79 Philippians 2:9–11 133 Colossians 24, 100 4:16

223

Index of References

1 Thessalonians 5:27 103

7:5 122 15:25 122

Epistle of the Apostles 4 190

1 Timothy 4:13 100

Babylonian Talmud Nedarim 32a 128

Gospel of Peter 14:58 186

2 Timothy 2:22 133 4:13 25 Titus 1:11–15 34 2:3–8 34 Hebrews 1:1 44 1 Peter 3:16 20 Revelation 7, 34 1:13–20 13:17–18 151 13:17 129 14:1 133 15:12 129 21:18–19 25 Pseudepigrapha Aristeas 176 123 Odes of Solomon 42:20 133 Sibylline Oracles 1:141–45 129 5:12–51 129 11:256 129 11:266 129 12:16–271 129 Dead Sea Scrolls 1QH 1:26 122 2:34 122

Midrash Genesis Rabbah 43:2 128 Philo De mutatione nominum 121 133 De opificio mundi 13–14 129 89–110 129 De plantation 117–25 129 De vita Mosis 2:114 122 2:205 122 Quaestiones et solutiones in Exodum 2:87 129 Quaestiones et solutiones in Genesin 1:91 129 2:5 129 Josephus Antiquities 122 2 §275 12 §89 123 New Testament Apocrypha and Pseudepigrapha Acts of John §§87–105 191

Infancy Gospel of Thomas 2:1–7 190 6:13–23 190 Nag Hammadi Codices Gospel of Truth I 38:6 133 133 I 37–41 Gospel of Thomas – xix, 11, 29, 44, 45, 56, 87, 118, 157, 160, 167, 168, 178, 183, 193–97, 210, 218 1 180 3 180 5 180 24 180 26 180 30 180 Gospel of Philip 56:3–15 133 Gospel of the Egyptians 65:9–66:24 133 Apostolic Fathers Barnabas 148, 151 9:7–9 9:7–8 128 12:8–10 133 12:8 94

224 Classical and Ancient Christian Writings Clement of Alexandria Stromateis 2.45.5 188 6 278–80 128 Epiphanius Panarion 30 188 Eusebius Historia ecclesiastica 3.3.2 191 3.25.4 191 3.25.6 191 3.39.4 25 6.12.1–6 185 6.12.2 186 Praeparatio evangelica 10.5 130 Hypomnesticon of Joseph 26.15 130 Irenaeus Adversus haereses 1.3.2 131 1.13.1 190 1.15.2 131 1.16.2 131 1.26.2 188 3.1.1 205 3.11.7–9 202 3.11.7 188 Jerome Commentariorum in Isaiam 40.9 188 Commentariorum in Michaeum 7.7 188

Index of References Justin Apologia I 38 207 55 148, 149 60 149 60.1 149 60.2–5 149 60.5 149 67 103 74 207 103.8 205 Dialogus cum Tryphone 75.1–2 133 103.8 20 Minucius Felix Octavius 29 149 Origen Psalmos 2.2 123 Tertullian An Answer to the Jews 9 133 Adversus Marcionem 3.22 149 Papyri Þ (Gregory-Aland numbers) 1 32 4 5, 7, 28, 34, 60, 62, 63, 108, 117, 203 5 34, 170, 171 9 34 11 28 34, 44 12 13 44, 157 14 28 15 34

18 44, 157 20 34 22 34, 44, 107, 155, 163–65, 179 23 34 27 34 28 34 29 34, 40 34, 62, 63 30 32 34 33 28 38 40 39 34, 108, 163 40 34 41 30 42 30 5, 32, 34, 37, 45 38, 40, 51, 55, 76, 77, 94, 142–45, 152, 163, 201–17, 219 46 5, 7, 9, 32, 37, 38, 41, 46, 51, 55, 79, 88, 170, 171, 204, 212 47 5, 32 48 34, 40 49 34 5, 6, 12, 33, 52 37, 59, 60, 63, 96, 109 52 recto l. 2 109 l.1 109 verso l. 2 109 53 34, 204 58 28 61 30 62 210

225 64

5, 7, 28, 30, 34, 58, 60–63, 108, 117, 203 65 34 5, 7, 32, 33, 66 39, 40, 56, 88, 94, 109, 116, 118, 132, 141, 143–45, 152, 216 5, 7, 28, 30, 67 34, 58, 60, 62, 63, 108, 117, 203 69 34 70 34 5, 33 72 73 30 30, 67, 68, 74 74 75 5, 7, 33, 37–40, 51, 55, 56, 88, 94, 109, 116, 118, 142–45, 152, 203, 211, 216 77 6, 9, 34 34, 163 80 87 34 88 108 90 5, 6, 34, 37, 63 91 34 34, 163 95 98 5, 7, 34, 37, 179 100–115 6 100 10 10, 34 101 103 6, 9, 10, 34, 95 104 5, 6, 10, 34, 37, 63 106 7, 10, 34

Index of References 107 34 108 10, 34 109 34 111 10, 34 113 10, 34 114 10, 34 115 10 118 34, 63 119 34 121 34 129 34 28, 34 131 132–134 30 132 28 133 28 134 28, 34, 43, 44 P.Berlin inv. 5513

11, 44

P.Bodmer V VII VIII XIII XIV XIV–XV XV

189 116 116 36, 116 142 51, 162, 211 142

P.Cair. 10759 186 P.Chester Beatty I 32, 143 II 32, 162, 170, 171 III 32 213 V VI 32 IV 148 IX–X 109, 204, 212 P.Dura 10 44

P.Egerton 2

91, 117, 128

P.Fouad 266 123 P.Gen. 3.125 179 P.Grenf. 1.8 189 P.IFAO 2.32 179 237b 7 P.Jena inv. 18+21 44 P.Köln 6.255 inv. 4724/ 4298

109, 117 33

P.Lond.Christ 1 109 P.Mag. Greek 17 58, 61 P.Mich. 18.763 179 11, 179 130 6652 204 inv. 1571 207 P.Mur. 164a, l. 11 139 P.Oslo 1661 inv. P.Oxy. 1

210 160, 169–71, 173, 175, 177, 180, 194

226 P.Oxy. 1 (cont.) verso 171 l. 1 l. 3 171. 172 l. 2 171 ll. 5–6 172 l. 5 172 l. 7 171 l. 8 172 172 l. 9 l. 10 172 172 l. 11 l. 13 172 172 l. 16 l. 17 172 172 l. 18 l. 19 172 recto l. 22 172 172 l. 23 ll. 25–30 181 172 l. 27 l. 30 172 172 l. 32 l. 35 172 ll. 36–37 172 l. 36 172 172 l. 41 1.108 130 2 32 2.208 128 2.209 128 118 3 407 3.402 128 6.849 118 6.850 118 6.851 118 6.847 128 6.850 128 7.1007 123, 124 50.3522 123 208+1781 170, 171 405 44, 208 654 11, 157, 160, 169, 174, 176, 179, 195

Index of References l. 2 l. 5 l. 6 l. 9 l. 10 ll. 11–12 l. 13 l. 14 l. 15 l. 19 l. 20 l. 21 l. 22 l. 25 l. 27 l. 28 l. 29 l. 32 l. 36 655

175 175 175 175 175 175 175 175 175 175 174 175 175 175 175 174, 175 174 175 175 11, 44, 56, 169, 175, 177, 179, 195, 210

frag. d l. 2 178 l. 3 178 657 157 1016 178 1079 157 1171 162 155, 157, 1228 159, 160, 162–66, 179 frag. 1 l. 1 161, 161 ll. 2–3 162 l. 2 160 l. 3 160, 161 ll. 4–5 161 l. 4 161 l. 5 160, 162 ll. 6–7 161 l. 6 161 l. 7 161 l. 19 161 l. 29 161

frag. 2 l. 1 160, 161 l. 1 161 ll. 2–3 161 161 l. 3 l. 4 161, 162 l. 5 161 l. 6 161 l. 7 161 ll. 8–9 161 l. 10 162 161, 162 l. 15 l. 17 162 161, 162 l. 18 l. 19 161 160 l. 20 l. 21 161 l. 27 160–62 2683 6 2949 179 3405 11 3523 6 3525 44, 196 4009 179 4401–6 6 4401 10 4403 6, 9, 10, 95 6, 10 4404 4405 6, 9, 10 4443 11 4445–48 6 4445 7, 10 4447 10 4449 6 4494–95 6 4495 10 4496 6 4497–98 6 4497 10 4498 10 6, 10 4499 4706 44 5000 6 P.Ryl. 457 457,

6, 12

Index of References

227 recto l. 1 l. 2 l. 3 l. 2 458 463

12 12 12 verso 12 11, 12, 96 196, 210

P.Vindob. G. 2325 11 P.Yale 1

96, 148

PSI 1.6 189 8.921v 179 12.1292 157

Parchment Manuscripts 0162 30 0171 30, 40, 62, 63 0189 30, 34 0206 108 62, 63 0212 0220 30 0232 108 0308 62 Quran Surah 3.49 190 5.110 190

I n d ex of A ut hor s

Aland, B. 4, 8, 15, 31, 35, 41, 52, 53, 68, 163 Aland, K. 30, 31, 33, 35, 94, 113, 116, 136, 141, 142, 145, 152, 158, 163 Alexander, L. C. A. 25 Amphoux, C.-B. 67 Amundsen, L. 210 Attridge, H. A. 168, 171, 174, 177, 181, 183 Aune, D. E. 22, 26 Avi-Yonah, M. 119, 130, 138, 139 Bagatti, B. 133 Bagnall, R. S. 37, 44, 57, 58, 62, 107 Barker, D. 57, 60, 109 Barns, J. W. B. 33, 141 Barrett, C. K. 66 Barrett, D. P. 61 Bauckham, R. 130 Bedodi, F. 117 Bell, H. I. 117, 128 Bell, L. D. xvi, 41, 53 Bellinzoni, A. J. 14, 20 Benoit, P. 139 Bietenhard, H. 122 Birdsall, J. N. 8, 18 Black, M. 94, 116, 136 Blanchard, A. 119, 138, 140 Bonner, C. 117 Bouriant, U. 187 Bremmer, J. 190–92 Brown, R. E. 130 Brown, S. 91, 114, 124, 126 Burke, T. 183 Burrus, V. 185 Burzachechi, M. 140 Cameron, R. 167 Caragounis, C. C. 172 Cartlidge, D. R. 189, 191 Caulley, T. C. 36 Cavallo, G. 36 Chadwick, H. 105

Chapa, J. 6, 204, 217 Charlesworth, J. H. 133 Charlesworth, S. D. 5, 28, 45, 55, 58, 59, 106, 174, 203 Clarysse, W. 5, 60–63 Cockle, W. E. H. 6 Colella, D. P. 138 Coles, R. A. 95 Colwell, E. C. 38 Comfort, P. W. 58, 61, 156 Cribiore, R. 105 Cullmann, O. 23 Czachesz, I. 60 Dagron, G. 191 Daniélou, J. 132 Davies, S. 185 DeConick, A. 168 Deissmann, G. A. 123 Delcor, M. 122 Delobel, J. 8 Dinkler, E. 93, 135, 136, 148, 149, 152–54 Dinkler-von Schubert, E. 113, 136, 138, 141, 149 Dodd, C. H. 122 Dölger, F. J. 145, 150, 151 Dornseiff, F. 129, 131 Dunn, J. D. G. 65, 66 Dupont, J. 67 Ehrman, B. D. 59, 78, 215 Elliott, J. K. 6, 50, 52, 67, 186, 189, 191 Elliott, W. J. 40 Emmel, S. 197 Epp, E. J. 4, 8, 11, 29, 31, 35–37, 39–42, 44, 49–51, 53, 67, 72, 75, 119, 120, 162 Evans, C. A. 187, 188 Fallon, F. T. 167 Farrar, F. W. 129 Fee, G. D. 18, 37, 39, 51, 53



Index of Authors

Finegan, J. 94, 116, 124, 137, 138, 140, 151 Finney, T. 6 Fitzmyer, J. A. 168, 169, 173 Foster, P. 104, 179, 186 Fournet, J.-L. 157, 158 Frantz, M. A. 144 Frey, J. 193 Frye, R. N. 139 Gamble, H. Y. xv, 10–12, 19, 20, 22, 58, 83, 100, 103, 115, 120 Garthausen, V. 138 Gero, S. 190 Gerstinger, H. 143, 200 Gignac, F. T. 172 Gillam, J. F. 139 Ginzberg, L. 128 Gonis, N. 6 Goodacre, M. S. 21 Grant, R. M. 130 Gregory, A. 14, 182, 187, 188, 208 Grenfell, B. P. 155, 157, 160, 162, 168, 170, 172, 174, 175 Gronewald, M. 33, 40 Haelst, J. van 30, 137, 169, 189, 204 Hagner, D. A. 14 Haines-Eitzen, K. 9, 10, 12, 55, 58, 83, 185, 198 Handley, E. W. 6, 95 Hannah, D. D. 202 Hargis, J. W. 105 Hartman, L. 132 Haslam, M. W. 6, 11 Head, P. M. 5, 7, 10, 11, 28, 40, 50, 58, 108, 203, 204, 207, 211, 217 Heckel, T. K. 19 Hedley, P. L. 155, 157, 163 Heitmuller, W. 132 Hengel, M. 20, 21, 202, 203, 205 Hernández, J., Jr. 204, 217 Hill, C. E. 5, 19, 24, 59, 60, 202–4, 217 Hock, R. F. 189 Hofius, O. 168 Holmes, M. xvi, 41, 42, 49, 50, 53, 65, 207, 214, 217, 218 Horsley, G. H. R. 89, 124 Hort, F. J. A. 49 Horton, C. 32, 201

229

Houston, G. W. 156, 195 Howard, G. 122, 127 Hübner, R. 6 Hull, R. F., Jr. 49 Hultgren, A. J. 135 Hunt, A. S. 155, 157, 160, 162, 168, 170, 172, 174, 175 Hurtado, L. W. xv, xvii, xviii, 6, 10, 11, 13, 24, 25, 30, 32, 34, 36, 38, 42–46, 48, 50–53, 55–57, 59, 64, 83, 88, 90– 95, 100, 106–8, 111, 112, 125, 134, 136, 137, 140, 142, 146, 151, 153, 154, 156, 157, 160, 165, 168, 178, 179, 193, 194, 196, 198, 201, 205, 209, 210, 216–18 Inghold, H. 139 Jankowski, S. 117 Jaro, K. 61 Jensen, R. M. xviii, 56, 94, 113, 152, 153, 216 Jeremias, J. 117 Johnson, W. A. xviii, 44, 45, 99, 101–3, 159, 177, 195, 209 Joosten, J. 21 Judge, E. A. 103–5 Kanael, B. 139 Kasser, R. 33, 142, 203 Kelhoffer, J. A. 18, 19, 22, 202 Kenyon, F. G. 31, 32, 117, 200, 211–13, 217, 218 Kilpatrick, G. D. 65, 67 Klauck, H.-J. 190, 191 Klijn, A. F. J. 182, 187 Koester, H. 4, 7, 14, 15, 20, 52, 117 Korpel, M. 14 Kraft, R. A. 46, 91, 168, 175–77 Kraus, T. 56, 179, 186, 201, 210, 211 Krauss, T. 195 Krüger, J. 195 Kruger, M. J. 204, 217 Lakmann, M.-L. 203 Lampe, P. 104 Lauterbach, J. Z. 122 Layton, B. 184 Lietzmann, H. 37, 38 Lieu, J. M. 90 Lim, T. H. 206

230

Index of Authors

Lindemann, A. 20 Longenecker, B. W. 216 Luijendijk, A. xix, 160, 169, 180 Maehler, H. 36 Malherbe, A. J. 104, 105 Marcovich, M. 168 Martin, V. 14, 33, 132, 141, 142, 203 Massaux, E. 14 McDonald, L. M. 23 McGing, B. C. 200, 201 McNamee, K. 119, 120, 138, 139 McVey, K. 150 Meeks, W. A. 104 Menninger, K. 130 Menzies, G. W. 130 Merkel, H. 23 Metzger, B. M. 18, 67, 71, 72, 74, 78, 115, 122, 205 Milburn, R. 153 Millard, A. 121, 124 Min, K. S. 41, 53 Morey, C. R. 144, 153 Mugridge, A. 9, 108 Nachmanson, E. 125 Nagel, T. 60 Nicklas, T. 179, 186 Nongbri, B. 5, 33, 51, 59, 60, 110, 201, 203 Norden, E. 15 O’Callaghan, J. 117 Oikonomides, A. N. 119 Orsini, P. 5, 60–63 Osiek, C. 218 Paap, A. H. R. E. 117, 126 Parker, D. C. 9, 19, 21, 40, 67, 117 Parsons, M. C. 208 Parsons, P. J. 123, 124, 196 Parvis, S. 104 Pasquier, A. 60 Patterson, S. J. 167 Pervo, R. I. 208 Petersen, W. L. 15, 21, 52 Peterson, S. 58 Pickering, S. R. 4, 5, 55 Pietersma, A. 122 Popkes, E. E. 193

Porter, S. E. 60, 62 Puech, E. 58 Rasimus, T. 60 Rea, J. R. 123 Read-Heimerdinger, J. 65–67, 70, 71, 73–75, 77, 79 Richards, R. 139 Riley, G. J. 167, 193 Rius-Camps, J. 67, 75, 77, 79 Roberts, C. H. 10, 12, 32, 50, 85, 87, 90, 91, 96, 110, 111, 115, 116, 124, 127, 128, 132, 148, 153, 155, 193, 196, 210, 216 Robinson, J. M. 32, 133, 146, 203 Römer, C. 30, 63 Ropes, J. H. 74, 75 Roth, D. T. 143, 216 Rowe, C. K. 64, 208 Royse, J. R. 8, 33, 37–40, 54, 122, 204, 207, 214, 217 Sanders, H. A. 14, 29, 162, 204 Sanders, J. A. 23 Sanderson, J. E. 123 Savignac, J. de 145 Schmid, U. 29, 65, 207, 214 Schneemelcher, W. 117 Schneider, G. 65 Schofield, E. M. 159, 162 Scholar, D. M. 167 Scholem, G. 128, 129 Schröter, J. 193, 208, 209 Scrivener, F. H. 68, 78 Sellew, P. 167 Siegel, J. P. 122 Sill, U. 117 Skarsaune, O. 20, 188, 202 Skeat, T. C. 5, 28, 32, 58, 85, 87–89, 108, 117, 118, 128, 143, 155, 200, 201, 203, 205, 211, 212, 214 Skehan, P. W. 122, 123 Snyder, G. F. 145, 154 Soskice, J. 49 Spence, L. 144 Staats, R. 129, 151 Stanley, C. D. 16 Stanton, G. N. 5, 20, 202 Stegemann, H. 122 Strange, W. A. 9, 67, 68



Index of Authors

Sulzberger, M. 139, 144 Syreeni, K. 19 Talmon, S. 17 Taylor, D. G. K. 21 Thiede, C. P. 58 Thomas, C. M. 192 Thomas, J. D. 6 Tov, E. 17, 24, 46, 47, 96, 123 Trachtenberg, J. 129 Traube, L. 116 Trebilco, P. 105 Treu, K. 46, 91, 124, 127 Trobisch, D. 20, 21, 23, 115, 206, 209 Tuckett, C. 8, 9, 14, 68, 93, 112, 113, 195, 196, 206, 207, 217 Turner, E. G. 39, 54, 59, 96, 101, 109, 110, 123, 139, 170–73, 178, 193, 211–13 Ulrich, E. 123

Valantasis, R. 183 Verheyden, J. 21 Vocke, H. 58 Wachtel, K. 58 Wallraff, M. 206 Wartenberg, U. 95 Wasserman, T. 28, 58, 204, 205, 217 Welles, C. B. 139 West, S. 17 Westcott, B. F. 49 Williams, M. A. 173, 197 Wingo, E. O. 101 Wischmeyer, W. 137 Wisse, F. 184, 197, 215 Wright, B. J. xvii Zuntz, G. 9, 10, 38, 51, 54, 200, 201, 211, 214

231