Gospel Interpretation and the Q-Hypothesis 9780567670045, 9780567670069, 9780567670052

Citation preview

INTERNATIONAL STUDIES ON CHRISTIAN ORIGINS Editor Michael Labahn Editorial Board Gabriella Gelardini, Tom Holmén, Martin Meiser, Peter Oakes, Bert Jan Lietaert Peerbolte, Beth Sheppard, Loren T. Stuckenbruck, Tom T. Thatcher Published under

Library of New Testament Studies

573 formerly the Journal for 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, Juan Hernández, John S. Kloppenborg, Michael Labahn, Love L. Sechrest, Robert Wall, Steve Walton, Catrin H. Williams

GOSPEL INTERPRETATION AND THE Q-HYPOTHESIS

Edited by Mogens Müller and Heike Omerzu

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 in 2018 Paperback edition first published 2020 Copyright © Mogens Müller and Heike Omerzu, 2018 Mogens Müller and Heike Omerzu have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identified as Editors 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. A catalog record for this book is available from the Library of Congress. ISBN: HB: 978-0-5676-7004-5 PB: 978-0-5676-9248-1 ePDF: 978-0-5676-7005-2 eBook: 978-0-5676-8322-9 Series: International Studies on Christian Origins within Library of New Testament Studies, volume 573 Typeset by Newgen KnowledgeWorks Pvt. Ltd., Chennai, India To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

CONTENTS Preface List of Contributors

vii ix

Chapter 1 INTRODUCTION: WHAT IS AT STAKE BY ADVOCATING OR DISPUTING THE TWO-SOURCE THEORY? Heike Omerzu

1

Part I HISTORY AND THEORY Chapter 2 CONCEPTUAL STAKES IN THE SYNOPTIC PROBLEM John S. Kloppenborg Chapter 3 SAD SOURCES: OBSERVATIONS FROM THE HISTORY OF THEOLOGY ON THE ORIGINS AND CONTOURS OF THE SYNOPTIC PROBLEM Stefan Alkier Chapter 4 WERE THE GOSPEL AUTHORS REALLY ‘SIMPLE CHRISTIANS WITHOUT LITERARY GIFT’ (ALBERT SCHWEITZER)?: ARGUMENTS FOR THE QUEST FOR SOURCES BEHIND THE GOSPELS Mogens Müller Chapter 5 Q AND THE LOGIA: ON THE DISCOVERY AND MARGINALIZING OF P.OXY.1 Francis Watson Chapter 6 WATSON, Q AND ‘L/M’ Christopher M. Tuckett Chapter 7 SEVEN THESES ON THE SYNOPTIC PROBLEM, IN DISAGREEMENT WITH CHRISTOPHER TUCKETT Francis Watson

13

43

79

97

115

139

vi

Contents

Part II TEXTUAL STUDIES Chapter 8 MARK WITH AND AGAINST Q: THE EARLIEST GOSPEL NARRATIVE AS A COUNTER-MODEL Eve-Marie Becker Chapter 9 REFUSING TO ACKNOWLEDGE THE IMMERSER (Q 7.31–35) Clare K. Rothschild Chapter 10 COHERENCE AND DISTINCTNESS: EXPLORING THE SOCIAL MATRIX OF THE DOUBLE TRADITION Hildegard Scherer Chapter 11 TAKING OUR LEAVE OF MARK–Q OVERLAPS: MAJOR AGREEMENTS AND THE FARRER THEORY Mark Goodacre Chapter 12 THE GOSPEL OF LUKE AS NARRATOLOGICAL IMPROVEMENT OF SYNOPTIC PRE-TEXTS THE NARRATIVE INTRODUCTION TO THE JESUS STORY (MARK 1.1–8 PARR.) Werner Kahl Chapter 13 DOES DATING LUKE-ACTS INTO THE SECOND CENTURY AFFECT THE Q HYPOTHESIS? Shelly Matthews Chapter 14 MARCION’S GOSPEL AND THE SYNOPTIC PROBLEM IN RECENT SCHOLARSHIP Dieter T. Roth Author Index Subject Index Index of References

151

165

185

201

223

245

267

283 288 292

PREFACE The essays in this volume derive from a conference on Gospel Interpretation and the Q Hypothesis held in Roskilde, Denmark, in June 2015. The idea to take up this topic was conceived during a conference held at the same place the previous year that was committed to Luke’s Literary Creativity. This theme naturally raised the question of the role of the Two-Source hypothesis in Gospel research. Luckily, it turned out that a number of the most prominent international specialists in this area were willing to contribute with a paper. We are grateful not only to them but also to all the other participants in the conference for lectures, offered short papers, responses and fruitful discussions. Like the 2014 conference, it was funded by the Velux Foundation as part of a larger research project on Narrative Transformation as Theological Interpretation of the Gospels at the Department of Biblical Exegesis, Faculty of Theology, University of Copenhagen. We have all reason to express our great gratitude to the Velux Foundation and to its senior adviser Henrik Tronier for the financial support, making also this follow-up conference possible. Michael Gisinger has corrected and proofread the English language in all contributions written by non-native speakers. We want to thank him for his efforts. A selection of papers from the 2014 conference was published in 2016 in the series Library of New Testament Studies (volume 550), and we are grateful to the editor of the International Studies in Christian Origins, Michael Labahn, for also including this new volume in this series. We also want to thank the staff at Bloomsbury Publishing, especially Sarah Blake, for their support in preparing the manuscript for publication. Mogens Müller and Heike Omerzu Copenhagen

LIST OF CONTRIBUTORS Stefan Alkier is Professor of New Testament and History of the Early Church at Frankfurt University, Germany. He is the author and editor of various books including The Reality of Resurrection: The New Testament Witness (2013). He is also Editor of the Zeitschrift für Neues Testament and of the Internet-Dictionary www.wibilex.de. Eve-Marie Becker is Professor of New Testament Exegesis at Aarhus University, Denmark. She is the author of Das Markus-Evangelium im Rahmen der antiken Historiographie (2006), The Birth of Christian History (2017) and Der früheste Evangelist. Studien zum Markusevangelium (2017). Mark Goodacre is a Professor in the Department of Religious Studies at Duke University, North Carolina. He is the author of several books including The Case Against Q: Studies in Markan Priority and the Synoptic Problem (2002) and Thomas and the Gospels (2012). Werner Kahl is Associate Professor of New Testament at Frankfurt University, Germany, and Head of Studies at the Academy of Mission at Hamburg University, Germany. He is the author of various books including New Testament Miracle Stories in their Religious-Historical Setting (1994) and Jesus als Lebensretter. Afrikanische Bibelinterpretationen und ihre Relevanz für die neutestamentliche Wissenschaft (2007). John S. Kloppenborg is Professor and Chair of the Department for the Study of Religion at the University of Toronto, Canada. His publications include The Tenants in the Vineyard (2006), Greco-Roman Associations: Texts, Translations, and Commentary I (with R. Ascough 2011), Associations in the Greco-Roman World: A Sourcebook (with P. Harland and R. Ascough, 2012) and Luke on Jesus, Paul, and Earliest Christianity (with J. Verheyden, 2017). Shelly Matthews is Professor of New Testament at the Brite Divinity School, Texas Christian University. She is the author of First Converts: Rich Pagan Women and the Rhetoric of Mission in Early Judaism and Christianity (2001), Perfect Martyr: The Stoning of Stephen and the Construction of Christian Identity (2010) and The Acts of the Apostles: Taming the Tongues of Fire (2013). Mogens Müller is Emeritus Professor of New Testament, University of Copenhagen, Denmark. Besides several books in Danish and German he is the author of

x

List of Contributors

The First Bible of the Church: A Plea for the Septuagint (1996) and The Expression ‘Son of Man’ and the Development of Christology (2008). Heike Omerzu is Professor of New Testament Studies, University of Copenhagen, Denmark. Her publications include Der Prozess des Apostels Paulus nach der Apostelgeschichte (2002), Paulus und Petrus: Geschichte – Theologie – Rezeption (with E. D. Schmidt, 2016) and Rewriting and Reception in and of the Bible (with J. Høgenhaven and J. T. Nielsen, forthcoming). Dieter T. Roth is Privatdozent at Mainz University, Germany, and author of The Text of Marcion’s Gospel (2015) and The Parables in Q (2018). Clare K. Rothschild is Professor of Scripture Studies in the Department of Theology at Lewis University, Illinois. She is the author of various books including Luke-Acts and the Rhetoric of History (2004), Baptist Traditions and Q (2005), Hebrews as Pseudepigraphon (2009) and Paul in Athens (2014). Hildegard Scherer is Dr. theol. habil. and currently Interim Professor of New Testament at Theologische Hochschule Chur, Switzerland. She is the author of Geistreiche Argumente. Das Pneuma-Konzept des Paulus im Kontext seiner Briefe (2011) and Königsvolk und Gotteskinder. Der Entwurf der sozialen Welt im Material der Traditio duplex (2016). Christopher M. Tuckett is Emeritus Professor of New Testament Studies, University of Oxford, United Kingdom. He is the author of various books and essays including Q and the History of Early Christianity (1996), The Gospel of Mary (2007), 2 Clement (2012), From the Sayings to the Gospels (2014). Francis Watson holds a Chair in Biblical Interpretation at Durham University, United Kingdom, and is a Professorial Fellow at Australian Catholic University, Melbourne. He is the author of various books including Paul and the Hermeneutics of Faith (2004, 2nd ed. 2016), Gospel Writing: A Canonical Perspective (2013) and The Fourfold Gospel (2016). With Sarah Parkhouse he has co-edited Connecting Gospels: Beyond the Canonical/Non-canonical Divide (forthcoming).

Chapter 1 I N T R O DU C T IO N :   W HAT I S AT S TA K E B Y A DVO C AT I N G O R D I SP U T I N G T H E T WO - S O U R C E T H E O RY ? Heike Omerzu

The Two-Source Theory emerged in the course of the nineteenth century as the traditional assumption that the Gospels were written by eyewitnesses was increasingly disputed in the late eighteenth century (G. E. Lessing, F. D. E. Schleiermacher, J. G. Eichhorn et al.). This insight subsequently gave rise to what became labelled the Synoptic Problem (hereafter, SP), that is, the study of similarities and differences between the Gospels of Mark, Matthew and Luke. The most convincing supposition appeared to be that there must have existed some kind of direct literary dependence between the Synoptic Gospels, that is, at least one of them had served as a written source for the others (‘utilization hypotheses’).1 Already in 1789, Johann Jakob Griesbach had proposed a hypothesis that until today bears his name, namely that Matthew is the oldest Gospel and that it was first used by Luke and later, alongside with Luke, also by Mark.2 The Griesbach theory, together with its underlying assumption of Matthean priority, was revived in the twentieth century and promoted by William R. Farmer

1.  Other principal suppositions apart from a direct literary dependence of one Gospel on another are either ‘proto-Gospel’ or ‘oral tradition’ hypotheses. See, for the history of this taxonomy, the references in John S. Kloppenborg Verbin, Excavating Q: The History and Setting of the Sayings Gospel (Minneapolis, MN: Fortress Press, 2000), 282 n. 19. 2. Johann Jakob Griesbach, Commentatio qua Marci evangelium totum e Matthaei et Lucae commentariis decerptum esse monstratur (Jena 1789/1790; reprint in J. J. Griesbach: Synoptic and Text-Critical Studies 1776–1976, ed. Bernard Orchard and Thomas R. W. Longstaff, Society for New Testament Studies Monograph Series (SNTSMS) 34 [Cambridge: Cambridge University Press, 1978]), 103–35. It is, however, disputed if Griesbach was actually the first to propose this solution or if he had forerunners such as Henry Owen, Observations on the Four Gospels (London: T. Payne, 1764); see on this Mark Goodacre, The Synoptic Problem: A Way Through the Maze (London: Continuum, 2001), 20 n. 6.

2

Heike Omerzu

as the Two Gospel hypothesis (also known as Neo-Griesbach hypothesis).3 Karl Lachmann, however, concluded in 1835 that neither Matthew nor Luke but Mark represented the most primitive order of the synoptic tradition.4 The assumption of Markan priority subsequently became the essential foundation of the Two-Source Theory first advanced by Christian Hermann Weiße and later elaborated by Heinrich Julius Holtzmann and others.5 The theory’s main constituents are that Matthew and Luke, independently of each other, made use of two sources, Mark and an unattested Sayings Source that later got to bear the name Q from German ‘Quelle’6; thus the Sayings Source is often also called the Q hypothesis. The third, major solution7 to the SP, apart from the Griesbach theory and Two-Source Theory, was originally established by Austin Farrer in 1955 and was later further developed by especially Michael D. Goulder.8 This theory also reckons with the assumption that Mark is the oldest gospel that served as a source for Matthew and Luke but instead of a lost Sayings Source, the so-called

3. William R. Farmer, The Synoptic Problem: A Critical Analysis (New York: Macmillan, 1964). See, for a discussion of the adequate denomination of the ‘Griesbach theory’, Christopher M. Tuckett, ‘The Current State of the Synoptic Problem’, in From the Sayings to the Gospels, ed. Christopher M. Tuckett, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 328 (Tübingen: Mohr Siebeck, 2014), 77–116, 81 n. 11. 4. Karl Lachmann, ‘De ordine narrationum in evangeliis synopticis’, Theologische Studien und Kritiken (ThStKr) 8 (1835): 570–90. 5.  Christian Hermann Weiße, Die evangelische Geschichte kritisch und philosophisch betrachtet (2 vols, Leipzig:  Breitkopf und Härtel, 1838); Heinrich Julius Holtzmann, Die synoptischen Evangelien, ihr Ursprung und geschichtlicher Charakter (Leipzig:  Engelmann, 1863); Holtzmann, Die Synoptiker, Hand-Commentar zum Neuen Testament (HCNT) I/1 (Tübingen: Mohr, 3rd edn, 1901). 6. This use most likely goes back to Eduard Simons, Hat der dritte Evangelist den kanonischen Matthäus benutzt (Bonn: Universitäts-Buchdruckerei Carl Georgi, 1880) and was then adopted by others. See on this, Frans Neirynck, ‘The Symbol Q (= Quelle)’, Ephemerides theologicae lovanienses (ETL) 54 (1978): 119–25; Frans Neirynck, ‘Once More: The Symbol Q’, ETL 55 (1979): 382–3; Frans Neirynck, ‘Note on the Siglum Q’, in Evangelica vol. II, Bibliotheca ephemeridum theologicarum lovaniensium (BETL) 99 (Leuven: Leuven University Press, 1991), 474. 7. See for an overview, over other suggestions, Tuckett, ‘Current State’. 8. See Austin Farrer, ‘On Dispensing with Q’, in Studies in the Gospels: Essays in Memory of R. H. Lightfoot, ed. D. E. Nineham (Oxford: Basil Blackwell, 1955), 55–88; Michael D. Goulder, Midrash and Lection in Matthew (London: SPCK, 1974); Michael D. Goulder, Luke: A New Paradigm, Journal for the Study of the New Testament Supplement Series (JSNTSup) 20 (Sheffield: Sheffield Academic Press, 1989); see also Mark Goodacre, Synoptic Problem; Mark Goodacre, The Case against Q: Studies in Markan Priority and the Synoptic Problem (Harrisburg, PA: Trinity Press International, 2002); Mark Goodacre and Nicholas Perrin (eds), Questioning Q. A Multidimensional Critique (Downers Grove: Intervarsity, 2004).

1. Introduction

3

Farrer–Goulder (also Mark-without-Q) theory9 claims that Luke knew and used both, Mark and Matthew. The Two-Source Theory is until today the most widely held and accepted solution to the SP10 but its character as a ‘hypothesis’ is quite differently conceived of in Anglophone scholarship on the one hand, and German-speaking scholarship on the other. In 1993, David Catchpole lamented that ‘Q is experiencing a mid-life crisis, with wounding attacks being made on its identity, indeed its existence, the intention being to kill the hypothesis stone dead.’11 Catchpole’s ‘concern’ addressed especially the challenge to the Sayings Source by advocates of the Farrer–Goulder theory in Great Britain, while the (Neo-) Griesbach theory questioning Markan priority has predominantly influenced debates of the Two-Source Theory in the Unites States.12 Christopher Tuckett, a firm advocate of the Two-Source Theory, concludes that ‘the recent resurgence of interest, especially in the last 30–40 years . . ., has focused the minds of many on issues of methodology as well as raising acutely the issue of certainty with which any solution to the Synoptic Problem can lay claim’.13 Thus, to the majority of biblical scholars in Great Britain and the United States the Two-Source Theory still appears to be the most convincing solution to the SP but its character as a hypothesis is openly discussed.14 In the German-speaking context, however, the somewhat dated bon mot passed on by Erwin R. Goodenough that ‘Q was Luke’s German source’15 seems still valid insofar as the existence of the Sayings Source is barely fundamentally questioned

9. On the designation, see Tuckett, ‘Current State’, 81–2 n. 12; Francis Watson, ‘Q as Hypothesis: A Study in Methodology’, New Testament Studies (NTS) 55 (2009): 397–415, 398–9. 10. Thus, for instance, Goodacre, Synoptic Problem, 21 and 23; Tuckett, ‘Current State’, 77: ‘In one sense, one could say that where we are now in relation to the Synoptic Problem is not very different from where we were 100 years ago. Then, as now, some form of the Two Document Hypothesis (2DH) ruled/rules the roost.’ John C. Poirier, ‘Introduction: Why the Farrer Hypothesis? Why Now?’, in Marcan Priority Without Q: Explorations in the Farrer Hypothesis, ed. John C. Poirier and Jeffrey Peterson, The Library of New Testament Studies (LNTS) 455 (London: Bloomsbury T&T Clark, 2015), 1–15, 1. 11.  David Catchpole, ‘Did Q exist?’, in The Quest for Q (Edinburgh:  T&T Clark, 1993), 1–59, 1. 12. See Goodacre, Synoptic Problem, 21–2. 13. Tuckett, ‘Current State’, 87. 14. But see also the distinction made by Goodacre, Synoptic Problem, 23–4 who holds that ‘[a]mong those who might be called experts on the Synoptic Problem, there is a variety of opinion . . . among those who write books on the Gospels not dealing directly with the Synoptic Problem, there tends to be a kind of blithe confidence, almost a complacency over the correctness of the Two-Source Theory’. 15. Erwin R. Goodenough, ‘The Inspiration of New Testament Research’, Journal of Biblical Literature (JBL) 71 (1952): 1–9, 2 (cited after Poirier, ‘Introduction’, 1).

4

Heike Omerzu

within German scholarship16 while competing theories are at the same time marginalized and underrepresented in standard New Testament introductions or text books that shape exegetical students’ approach to Gospel studies.17 Hence, New Testament introductions typically devote a whole chapter to the Sayings Source but only a few lines to the Neo-Griesbach or Farrer–Goulder theory, often without a proper discussion of their arguments while downplaying evidence against the Q hypothesis.18 At the same time, critical voices disputing the Two-Source Theory are

16. Studies on the Sayings Source might disagree regarding its content, age or character, but do not usually question its existence as such. See, for instance, Andreas Lindemann, ‘Die Logienquelle Q. Fragen an eine gut begründete Hypothese’, in The Saying Source Q and the Historical Jesus, ed. Andreas Lindemann, BETL 158 (Leuven: Leuven University Press, 2001), 3–26, or Jens Schröter, ‘Die Bedeutung der Q-Überlieferungen für die Interpretationen der frühen Jesustradition’, Zeitschrift für die neutestamentliche Wissenschaft (ZNW) 94 (2003): 38–67, who proposes an open variant of the Q hypothesis: ‘Mit “Q” werden also die über Mk hinausgehenden gemeinsamen Texte von Mt und Lk bezeichnet, ohne dass damit bereits weitergehende Annahmen über Herkunft und Zusammengehörigkeit dieser Texte verbunden wären.’ ('By "Q" we mean the common texts of Mt and Lk which extend beyond Mk, without any further assumptions concerning the origin and the relation of these texts.) 17. See, for instance, the examples provided by Werner Kahl, ‘Vom Ende der Zweiquellentheorie oder: Zur Klärung des synoptischen Problems’, Transparent-Extra 75 (2004): 1–36, 3–6. Kahl is one of the few (explicit) German-speaking opponents of the Q hypothesis (see below). 18. See, for instance, Udo Schnelle, Einleitung in das Neue Testament (3rd edn; Göttingen: Vandenhoeck & Ruprecht, 1999; trans. The History and Theology of the New Testament Writing [Minneapolis, MN: Fortress, 1998]), chap. 3.2 on the Synoptic Problem (177–94 [161–79], thereof 190–4 [175–9] devoted to other theories), chap. 3.3 on the Sayings Source (194–214 [179–97]); Martin Ebner and Stefan Schreiber (eds), Einleitung in das Neue Testament (Stuttgart: Kohlhammer, 2008), where ‘alternative models’ are presented on three pages out of eighteen in total devoted to the SP (67–84, Farmer: six lines!), which is followed by 27 pages on the Sayings Source (85–111). In the introductory paragraph of this latter chapter, Ebner refers to the recent publication of James M. Robinson, Paul Hoffmann and John S. Kloppenborg (eds), The Critical Edition of Q, Hermeneia (Leuven: Peeters, 2000) and remarks: ‘Kein Wunder, dass sich im Vorfeld dieses Editionsereignisses und in Reaktion darauf Vertreter all derjenigen Theorien lauthals zu Wort gemeldet haben, die meinen, das synoptische Problem ohne eine hypothetische Quelle lösen zu können, und deshalb provokativ (sic!) auf Q verzichten’ (It is no wonder that during the preparation of this publication and in response to it representatives of all those theories, who claim to be able to solve the Synoptic Problem without a hypothetical source and therefore provocatively (sic!) dispense with Q, vociferiously piped up.)(85). He does not explain how and why the attempt to abandon a hypothetical source should be conceived as a ‘provocation’. See, for a comparison regarding both content and scope, the somewhat more balanced account in Raymond E. Brown, An Introduction to the New Testament (New York: Doubleday, 1997), 111–16 (The Synoptic Problem), 116–22 (The Existence of ‘Q’).

1. Introduction

5

rare. Two more recent exceptions19 are Werner Kahl,20 who argues along the lines of the Farrer–Goulder theory, and Matthias Klinghardt who – on the background of a detailed reconstruction of Marcion’s gospel – holds that an earlier version of this gospel (Mcn) served as source for both Marcion and all the New Testament Gospels.21 A critical history of the SP that not only describes the different solutions but also explores their interrelatedness with theological, ideological and other more or less – consciously or unconsciously – hidden agendas is still a desideratum. This demands a more in-depth analysis than what either this introduction or the current volume as a whole can provide.22 As already indicated, the birth of the Two-Source Theory can be traced back to the beginnings of historical critical scholarship. It is intrinsically tied to the quest for the historical Jesus, both as regards the striving for reliable sources and as it eventually gave rise to liberal ‘Lives of Jesus’ in reaction to first Hermann Samuel Reimarus (1694–1768) and the Fragmentenstreit23 and later to David Friedrich Strauß’s Das Leben Jesu.24 While Mark was the oldest extant Gospel, the lost Sayings Source could possibly be dated even earlier and, in addition, it was supposed mainly to represent logia that could be traced back to Jesus.25 But even as

19. See Kloppenborg, Excavating Q, 271 n. 1, for critical remarks on the older works of Hans-Herbert Stoldt, Geschichte und Kritik der Markushypothese (Göttingen: Vandenhoeck & Ruprecht 1977, trans. History and Criticism of the Marcan Hypothesis [Edinburgh: T&T Clark, 1980]) and Eta Linnemann, Gibt es ein synoptisches Problem? (4th edn, reprint Nürnberg: Verlag für Theologie und Religionswissenschaft 1999; trans. Is There a Synoptic Problem? Rethinking the Dependence of the First Three Gospels [Grand Rapids, MI: Baker, 1992]). On the lack of reception of Stoldt’s work in Germany, see David B. Peabody, with Lamar Cope and Allan James McNicol (eds), One Gospel From Two: Mark’s Use of Matthew and Luke (Harrisburg, PA: Trinity Press, 1983), 9–10. 20.  See Kahl, ‘Vom Ende der Zweiquellentheorie’; Kahl ‘Inclusive and Exclusive Arguments – Towards a Neutral Comparison of the Synoptic Gospels, or: Minor Agreements as Misleading Category’, in Luke’s Literary Creativity, ed. Mogens Müller and Jesper Tang Nielsen, LNTS 550 (London: Bloomsbury T&T Clark, 2016), 44–78; see also Kahl’s contribution to this volume. 21.  Matthias Klinghardt, Das älteste Evangelium und die Entstehung der kanonischen Evangelien, 2 vols, Texte und Arbeiten zum neutestamentlichen Zeitalter (TANZ) 60 (Tübingen: Francke, 2015). 22. See most comprehensively, Kloppenborg, Excavating Q, esp. 271–328 (‘The Jesus of History and the History of Dogma’). 23.  See on this, e.g. Francis Watson, Gospel Writing:  A Canonical Perspective (Grand Rapids, MI: Eerdmans, 2013), 62–113. 24. Tübingen: C. F. Osiander, 1835–36. See on this, Kloppenborg, Excavating Q, 309ff. 25.  Farmer’s claim that the uncritical approval of the Markan priority among German Protestants had an antipapal stance lacks proper corroboration; see William R. Farmer, ‘State Interesse and Marcan Primacy: 1870–1914’, in The Four Gospels 1992, FS Frans Neirynck, ed. F. van Segbroeck et al., BETL 100, 3 vols (Leuven: Leuven University Press, 1993), 3.2477–98; see also the critique by Kloppenborg, Excavating Q, 312–21 and Tuckett, ‘Current State’, 92–3.

6

Heike Omerzu

the quest for the historical Jesus was further developed methodologically, among other things in the wake of form criticism and the attention to extra-canonical material such as the Gospel of Thomas, the Two-Source Theory and the Q hypothesis maintained their dominant status both within the study of the historical Jesus26 and in biblical studies in general. That the Two-Source Theory was developed and established in the context of German biblical studies and theology is most likely an important factor for the reluctance of German-speaking scholarship to discuss (and embrace) alternative solutions to the SP. It has simply become an idea held dear for such a long time that it seems almost unthinkable to let go. This is, on the one hand, facilitated by the rather conservative industry of New Testament Introductions and, on the other, by the somewhat delayed and/or reluctant reception of trends and developments in Anglo-American scholarship. An open discussion of the Two-Source hypothesis might also demand some other impetus. Accordingly, John C. Poirier considers possible reasons for the increased interest of Anglophone studies in alternatives to the Two-Source Theory, and especially the Farrer–Goulder hypothesis, during the past decades and suggests that it ‘is at least partly due to a number of more general developments in NT scholarship, developments that sometimes betray a hollowness in some of the older complaints lodged against the idea of Luke’s dependence on Matthew’.27 Among these developments is the acknowledgement of the evangelists’, and consequently also Luke’s, literary ability28 and interaction with sources as well as ‘a wider acceptance of a late date for Luke’.29 Poirier rightly remarks that the recognition of alternative theories (to him esp. the Farrer theory) depends to a large extent on ‘able defenders’.30 With respect to the Farrer theory, Poirier refers to Goulder and more recently to Mark Goodacre ‘who has transformed the FH [Farrer hypothesis; H. O.] from a (mostly) British phenomenon into a transatlantic phenomenon’.31 Another influential advocate of a solution of the SP without the Q hypothesis is Francis Watson, who assumes Markan priority and Luke’s dependence on Matthew, thereby dispensing with Q.32 It has to be expected for the future that an open and fruitful discussion can also be transferred to German scholarship where the developments mentioned by Poirier are increasingly becoming more established.

26. See, for instance, Lindemann, Saying Source Q. This applies in spite of Kloppenborg’s justified qualification that ‘most of the recent specialists in Q suppose that the contents of Q exhaust what can responsibly be said of Jesus, nor more important, that Q’s particular construal of Jesus represents the historical Jesus’ (Excavating Q, 351). 27. Poirier, ‘Introduction’, 2. 28. See, on this, Müller and Nielsen, Luke’s Literary Creativity. 29. Poirier, ‘Introduction’, 2. 30. Poirier, ‘Introduction’, 2. 31. Poirier, ‘Introduction’, 2 n. 3. 32. See Watson, Gospel Writing.

1. Introduction

7

The present volume might be an important step towards this direction. It represents the results of a conference held in Roskilde, from 21 to 24 June 2015, which had the intention to bring together both proponents and opponents of the Two-Source Theory in order to explore the limitations and strengths as well as the conceptual stakes of different solutions to the SP, especially those operating with Markan priority. Experts in gospel studies as well as scholars with other primary research interests, representing both Anglophone and German-speaking scholarship, should be brought together. John S. Kloppenborg opens the volume’s first section on ‘History and Theory’ with considerations about the ‘Conceptual Stakes in the Synoptic Problem’ in which he both raises methodological issues and sketches the requirements for a critical history of the SP that takes into account the theological and historical agenda underlying the establishment of the Two-Source hypothesis as well as the interrelatedness of the SP with the quest for the historical Jesus and more generally the conceptualization of the Jesus movement. He invokes the complexity of the available data and calls for a critical reflection of the ideological entailment of any solution to the SP. In his contribution, ‘Sad Sources: Observations from the History of Theology on the Origins and Contours of the Synoptic Problem’, Stefan Alkier explores the perception of and reaction to the SP throughout Church history, beginning with the early Church, over the church fathers and Luther to Enlightenment theologians and the emergence of historical critical scholarship. He demonstrates that the history of the study of the SP does not reveal a linear progression and that it is worthwhile, even necessary, to revisit earlier solutions. This is especially true in light of the fact that no solution is found yet. Mogens Müller’s approach in ‘Were the Gospel Authors Really “Simple Christians without Literary Gift” (Albert Schweitzer)? Arguments for the Quest for Sources behind the Gospels’ is also one from the point of view of history of research. He describes as the driving force behind the early historical critical solutions of the SP the urge to establish reliable sources for the quest for the historical Jesus. This interest is still current today but in the wake of both redaction and narrative criticism the literary creativity of the gospel authors became increasingly more important for other scholars. The appreciation of the gospel authors’ literary achievements in combination with a dating of Luke in the second century makes Müller suggest that the Two-Source Theory is superfluous. Francis Watson demonstrates in his essay ‘Q and the Logia: On the Discovery and Marginalizing of P.Oxy.1’ that the existence of an early Sayings Collection genre (which Watson also argues for in his study, Gospel Writing)33 was already discussed in late nineteenth-century Oxford when P.Oxy.1, later identified as a part of the non-canonical Gospel of Thomas, was published by Grenfell and Hunt. The new Logia fragment could possibly have provided a quite different solution to the SP as compared to the Two-Document hypothesis but Watson argues that it

33. See note 32.

8

Heike Omerzu

was especially William Sanday’s rhetorical efforts that established Q as ‘orthodox’ within modern critical scholarship – on the background that the Q hypothesis operates intra-canonically as opposed to the new Logia. The essay ‘Watson, Q and “L/M” ’ by Christopher M.  Tuckett is a critical response to Watson’s claim for a direct dependence of Luke on Matthew without recourse to Q which is one of the main ideas developed in his above-mentioned study Gospel Writing. Tuckett raises, first, more general, methodological issues such as how to define and identify ‘redactional procedures’. Next, he discusses the coincidences between Matthew and Luke as allegedly being an argument against their independence before he eventually addresses (and questions) Watson’s arguments in support of the L/M hypothesis. Francis Watson presents an amended version of his oral response to Tuckett’s critique in ‘Seven Theses on the Synoptic Problem, in Disagreement with Christopher Tuckett’. Herein Watson addresses some misrepresentations by Tuckett and elaborates some of his own arguments. Eve-Marie Becker opens the second section of the volume, ‘Textual Studies’, with the essay ‘Mark With and Against Q:  The Earliest Gospel Narrative as a Counter-Model’. Her point of departure is the reflection of a ‘Markan scholar’ of the state of the art of Q studies, including the relationship between Mark and Q. In a next step, Becker argues that the existence and literary shape of the Gospel of Mark is much better explained on the Q hypothesis than on the Farrer–Goulder theory. Q is subsequently presented to be Mark’s literary forerunner that he tried to exceed by the production of his own Gospel. In her essay, ‘Refusing to Acknowledge the Immerser (Q 7.31–35)’, Clare K.  Rothschild proposes that the expression ὁ υἱὸς τοῦ ἀνθρώπου in Q 7.34 is understood as a self-referential designation of John the Baptist and counters possible objections. Such a reading reveals that the main purpose of the passage within its original Q context is the admonishment of those not acknowledging John’s baptism. With respect to the SP, this reading reminds us that some of the traditions encapsulated in the canonical Gospels might have their origin in Baptist circles and not necessarily in the Jesus movement. In a similar vein, Hildegard Scherer develops the argument in her essay ‘Coherence and Distinctness: Exploring the Social Matrix of the Double Tradition’ on the grounds of the Q hypothesis. Acknowledging the hypothetical nature of the Q document she wants to provide circumstantial evidence to the likelihood of its existence by focusing on the social categories related to household roles in the double tradition texts over against Markan and Sondergut traditions. Scherer concludes that the example of family roles reveals a coherent social texture in the double tradition that is beyond mere coincidence but is the result of conscious reflections. The following two contributions argue from the point of view of the Farrer– Goulder theory. Mark Goodacre addresses in his essay, ‘Taking our Leave of Mark–Q Overlaps: Major Agreements and the Farrer Theory’, a phenomenon that notoriously poses difficulties for the Two-Source Theory, namely agreements in the triple tradition between Matthew and Luke against Mark. After analysing several key passages Goodacre holds that the so-called Mark–Q overlaps are better

1. Introduction

9

explained on the Farrer theory than on the Two-Source Theory. If this assumption is correct, it not only impacts the SP but also affects the criterion of multiple attestation in historical Jesus research. In his essay, ‘The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts: The Narrative Introduction to the Jesus Story (Mark 1.1–8 parr.)’, Werner Kahl exemplifies his conviction that the Farrer–Goulder theory offers a better solution to the SP than the Two-Source Theory by drawing on Greimas’s Narrative Schema in his analysis of Mark 1.1–8 parr. Kahl comes to the conclusion that the differences between the three parallel passages can be explained without recourse to Q because Luke shares with Matthew both the rearrangement of the Markan text and most of the additional material. Differences can be explained as linguistic or narratological changes made by Luke. The final two contributions discuss how recent scholarship on Marcion’s Gospel can cast light on the SP. Shelly Matthews addresses in ‘Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis?’ the relationship between canonical Luke and Marcion’s Gospel. She holds that the canonical Luke is a second-century redaction of an earlier ‘core Luke’ which still seems best explained on the grounds of the Two-Source Theory. Canonical Luke, however, which is characterized by additions primarily at the beginning and end of ‘core Luke’, may have known and drawn on Matthew. Marcionite circles are less likely to have used canonical but rather ‘core Luke’ and expanded it according to their teaching. Dieter T.  Roth provides an instructive and critical overview over ‘Marcion’s Gospel and the Synoptic Problem in Recent Scholarship’. Under the given limitations of space he briefly presents and critically interacts with the recent publications on the topic by Markus Vinzent, Matthias Klinghardt, Jason BeDuhn and Judith Lieu. In sum, Roth welcomes the renewed interest in Marcion’s Gospel and the debates concerning its relation to the canonical Gospels and thus the SP; yet, he cautions that any hypothesis based on Marcion’s Gospel has to take its character as a – potentially hypothetical – reconstruction into account. In sum, the conference from which these chapters originally emanated and the debates continuing in this volume show that after roughly two hundred years of exploring the compositional relationships between the canonical Gospels the dust has still not settled. The contribution of these chapters to the ongoing debate go beyond the details of text analysis and text comparison, essential as these might be because, after all, the formulation of the SP and the resulting source theories exist by virtue of very detailed analysis and comparison of the minutiae of the text of the three Synoptic Gospels and some non-canonical texts. Nevertheless, revisioning the very basis on which the SP is constructed, the contributions in this volume highlight an issue that is of even more fundamental significance. The spotlight is shown here also on historical and geographical trajectories of scholarship, as well as the occasionally incidental and extraneous factors that lead to the construction and use of theory. Given the deep roots of historical critical study of early Christianity since the beginning of the nineteenth century in a countercultural stance vis-à-vis the dominant ecclesial culture in Germany at the time, it is clear (as I have also argued

10

Heike Omerzu

above) that the complex of source hypotheses, the SP, but especially its main solution – the Sayings source Q – has gained its traction as hegemonic theory in service of a search to discover the historical Jesus. The Jesus that emerges from Q is exactly that countercultural figure that stands in some opposition to the high church theology of a resurrected, risen Christ. Certainly, since the 1960s and in the more recent reception of the Jesus-Sayings in studies of Christian origins, it has become clear that Q (and the attendant pillars that continue to support its existence) has become the battleground on which more than just analyses of textual details and comparisons are fought. The debates on the SP illuminate that all scholarship is inflected by invested interests. Scholarship that does not also investigate the role of these underlying interests is likely to continue to re-encode the ideological interests that animated earlier contexts of scholarship. As indicated above in the overview of the history of the SP, the Two-Source Theory has become so hegemonic that, for many in the guild, it now goes without saying and, furthermore, it has become so normalized and naturalized that it achieved a status of reality. In scholarship, our ‘industry’ is not only concerned with the objects we study but also with the process by which these knowledge-objects are approached. This is the desideratum expressed in the ongoing debate contained in this volume.

Part I H ISTORY AND  T HEORY

Chapter 2 C O N C E P T UA L S TA K E S I N T H E S Y N O P T IC PROBLEM John S. Kloppenborg

Ever since the publication of William Farmer’s The Synoptic Problem in 1964 solutions to the synoptic problem other than the Two-Source or Two-Document hypothesis (2DH) have won many advocates (and opponents).1 Of course prior to Farmer’s work there were dissenters from the 2DH: one thinks of B. C. Butler, Austin Farrer, Léon Vagany and others. But it is fair to say that in the wake of Paul Wernle and B. H. Streeter’s monographs and William Sanday’s edited volume on the Synoptic Problem,2 these voices were drowned out by the large majority of those advocating (or simply presupposing) the 2DH. In this chapter, I  will not rehearse the case in favour of the 2DH nor offer arguments against the Farrer hypothesis (FH) or the 2GH. I am not of the view that definitive arguments for or against any of these hypotheses or variations on these are in fact possible. At best, we can propose compositional scenarios that account for most of the data, most of the time, by what we think of as the ‘most probable’ explanations. I  wish instead to think about what might be at stake methodologically and conceptually in various solutions to the Synoptic Problem and why, beyond addressing the relatively narrow problem of the literary genealogy of early Christian gospel writing, the Synoptic Problem (hereafter, SP) might still be of theoretical interest. Before turning to those conceptual stakes, I would like to suggest what should be some common methodological ground in SP discussions.

1. William R. Farmer, The Synoptic Problem: A Critical Analysis (New York: Macmillan, 1964). A bibliography up to 1987 or 1988 was published by Thomas R. W. Longstaff and Page A. Thomas (eds), The Synoptic Problem:  A Bibliography 1716–1988, New Gospel Studies 4 (Macon, GA: Mercer University, 1988). 2.  Paul Wernle, Die synoptische Frage (Leipzig, Freiburg i.Br. and Tübingen:  J. C.  B. Mohr [Paul Siebeck], 1899); (Oxford) Studies in the Synoptic Problem, ed. William Sanday (Oxford: Clarendon, 1911); B. H. Streeter, The Four Gospels: A Study of Origins, Treating of the Manuscript Tradition, Sources, Authorship, and Dates (London: Macmillan, 1924).

14

John S. Kloppenborg

2.1 Methodological Issues All would agree, I think, that the SP is key to the historical understanding of the composition of the gospels. To take the SP seriously is to recognize that the inscription of any of the gospels was not an innocent or naive act, but inevitably involved an engagement and negotiation with the ‘already said’. This engagement might have been elaborative, complementary, corrective, or combative with respect to any particular datum of Jesus material.3 It is also an act that is socially located, that is, it reflects certain social values and postures  – a point to which I  will return later. However the act is construed, inscription was an argument, insofar as it represents a rhetorical selection, disposition and deployment of the Jesus material, and it had a particular social and rhetorical location. The result is a layered text in which prior construals are engaged and submerged, eirenically or otherwise, in a new configuration of textual elements. While purely synchronic approaches to reading have great value, it is hardly incidental to the understanding of a text to note the points at which predecessor materials are manipulated, suppressed, enhanced, or qualified. As I read the literature, two methodological issues have now become important in the framing of the task of SP research. These are, first, the scope of data that ought to be included in SP research; and second, achieving greater clarity on what counts as an argument that has compelling force in this research. 2.1.1 The Primary Data for the Synoptic Problem Perhaps one of the interesting and potentially fruitful recent developments in SP research is Francis Watson’s observation that in the past the parameters of the discussion have artificially been set at the end of the first century, which has had the effect of ruling out of the discussion documents that might have been edited in the second century but which arguably provide access to the history of the inscription of the synoptic gospels.4 Watson himself suggests both that the Gospel of Thomas preserves some materials, and especially a literary genre, that are independent of the Synoptics and that a ‘Thomas-like sayings collection’ may stand behind Matthew.5 3.  Except when reporting the views of others, I deliberately avoid the term ‘Jesus tradition’ as implying too formal and controlled a model of information transfer. ‘Tradition’ conveys an aura of sacrality to the process, which might have included something like controlled or semi-controlled recitation of stories and sayings, but likely also included rumours, gossip, disinformation, exaggeration, speculation and sheer fantasy. For ethnographic examples, see Christine Shojaei Kawan, ‘Legend and Life: Examples from the Biographies of ‘Ā‘ishah Bint Abī Bakr, Mary Carleton, and Friedrich Salomo Krauss’, Folklore 116, no. 2 (2005):  140–54 and on the use of gossip and rumour in historiography, Israel Shatzman, ‘Tacitean Rumours’, Latomus 33, no. 3 (1974): 549–78. 4. Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI: Eerdmans, 2013), 5 and passim. 5. Watson, Gospel Writing, 251–85.

2. Conceptual Stakes in the Synoptic Problem

15

This approach is not new, though it is new in the form that Watson has promoted it. As most will know, Lessing invoked the Gospel of the Nazarenes and the Gospel of the Hebrews, known only from patristic testimony from the early third century onwards, as an extensive Aramaic proto-gospel, existing in several different versions, from which the canonical Gospels were translated.6 Schleiermacher’s Λόγια is structurally even closer to Watson’s model insofar as Schleiermacher began with the testimony of Papias,7 and detected within Matthew two sets of material, a narrative gospel and a sayings collection, although of course Schleiermacher held to the Griesbach hypothesis which placed Mark last as a conflation of Matthew and Luke. Appeals to Urgospels continued throughout the nineteenth century, but with less and less attention paid to the patristic texts originally used to posit an Urgospel. The value of patristic testimony about the composition of the gospels had been soundly rejected by Griesbach and, although it was occasionally invoked later on, by the beginning of the twentieth century the SP was conducted largely without much attention to anything beyond the first three gospels, as Watson rightly notes. Yet beginning at least in the early 1970s, texts that, in Lührmann’s words, were to become ‘apocryphal’,8 had been incorporated into models of Christian origins, at least in North America. Instrumental in this development was James Robinson and Helmut Koester’s Trajectories through Early Christianity9 which programmatically abandoned canonical boundaries and attempted to plot ‘trajectories’ of development and cartographies of theologies that included what, anachronistically, were called ‘canonical’ and ‘extracanonical’ texts. Hence, without detracting from the truth of Watson’s insistence that ‘the canonical/noncanonical distinction is not given with the texts themselves but arises out of their reception’ and that the end of the first century should not set a limit on texts considered pertinent to the Synoptic Problem,10 it is a sad testimony to the theological intransigence in the guild that it has taken almost half a century to appreciate the distinction between empirical and theological histories. In fact the ‘Synoptic Problem’ has been undergoing an expansion of its purview for quite some time:  textual materials are examined that display

6.  Gotthold Ephraim Lessing, ‘Neue Hypothese über die Evangelisten als bloss menschliche Geschichtsschreiber betrachtet’, in Theologischer Nachlass (Berlin:  Voss, 1784), 45–72 [1st edn 1778]. 7.  Friedrich Schleiermacher, ‘Über die Zeugnisse des Papias von unsern beiden ersten Evangelien’, Theologische Studien und Kritiken (TSK) 5 (1832): 735—68. 8.  Dieter Lührmann, Fragmente apokryph gewordener Evangelien in griechischer und lateinischer Sprache, Marburger Theologische Studien 59 (Marburg:  N. G.  Elwert Verlag, 2000). 9.  James M. Robinson and Helmut Koester, Trajectories Through Early Christianity (Philadelphia: Fortress, 1971). 10. Watson, Gospel Writing, 5.

16

John S. Kloppenborg

significant resemblance to the Synoptic Gospels such as Didache 1.3b–2.1; 16.3–8;11 1 Clement 13; the Gospel of Thomas;12 the Gospel of Peter;13 the Dialogue 11. e.g. Bentley Layton, ‘The Sources, Date and Transmission of Didache 1.3b–2.1’, Harvard Theological Review (HTR) 61 (1968):  343–83; John S. Kloppenborg, ‘Didache 16,6–8 and Special Matthaean Tradition’, Zeitschrift für die neutestamentliche Wissenschaft (ZNW) 70 (1979): 54–67; Jonathan A. Draper, ‘The Jesus Tradition in the Didache’, in Gospel Perspectives. Vol. 5: The Jesus Tradition Outside the Gospels, ed. David Wenham (Sheffield: JSOT Press, 1985), 269–87; Christopher M. Tuckett, ‘Synoptic Tradition in the Didache’, in The New Testament in Early Christianity: La réception des écrits néotestamentaires dans le christianisme primitive, ed. J. M. Sevrin, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 86 (Leuven: Peeters and Leuven University, 1989), 197–230; Alan J. P. Garrow, The Gospel of Matthew’s Dependence on the Didache, Journal for the Study of the New Testament Supplement Series (JSNTSup) 254 (Sheffield: Sheffield Academic Press, 2004); John S. Kloppenborg, ‘The Use of the Synoptics or Q in Did. 1.3b–2.1’, in The Didache and Matthew: Two Documents from the Same Jewish-Christian Milieu?, ed. Huub van de Sandt (Assen: Van Gorcum; Minneapolis: Fortress, 2005), 105–29. 12. e.g. Wolfgang Schrage, ‘Evangelienzitate in den Oxyrhynchus-Logien und im koptischen Thomas-Evangelium’, in Apophoreta: FS Ernst Haenchen zu seinem siebzigsten Geburtstag, ed. Walther Eltester and F. H. Kettler, Beihefte zur ZNW (BZNW) 30 (Berlin:  Alfred Töpelmann, 1964), 251–68; John H. Sieber, ‘A Redactional Analysis of the Synoptic Gospels with Regard to the Question of the Sources of the Gospel According to Thomas’, PhD diss. (Claremont:  Claremont Graduate School, 1966); Christopher M. Tuckett, ‘Thomas and the Synoptics’, Novum Testamentum (NovT) 30, no. 2 (1988): 132–57; Thomas Zöckler, Jesu Lehren im Thomasevangelium, Nag Hammadi and Manichaean Studies (NHMS) 47 (Leiden:  Brill, 1999); Nicholas Perrin, Thomas and Tatian: The Relationship Between the Gospel of Thomas and the Diatessaron (Atlanta: Society of Biblical Literature; Leiden: Brill, 2002); John S. Kloppenborg, The Tenants in the Vineyard:  Ideology, Economics, and Agrarian Conflict in Jewish Palestine, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 195 (Tübingen:  Mohr Siebeck, 2006); John Horman, A Common Written Greek Source for Mark and Thomas, Studies in Christianity and Judaism (ESCJ) 20 (Waterloo:  Wilfrid Laurier University, 2011); Stephen J. Patterson, ‘The Gospel of (Judas) Thomas and the Synoptic Problem’, in New Studies in the Synoptic Problem, ed. Paul Foster et  al., BETL 239 (Leuven:  Peeters, 2011), 783–808; Simon Gathercole, The Composition of the Gospel of Thomas:  Original Language and Influences, Society for New Testament Studies Monograph Series (SNTSMS) 151 (Cambridge: Cambridge University Press, 2012); Mark S. Goodacre, Thomas and the Gospels: The Case for Thomas’s Familiarity with the Synoptics (Grand Rapids, MI:  Eerdmans, 2012); John S. Kloppenborg, ‘A New Synoptic Problem: Goodacre and Gathercole on Thomas’, JSNT 36, no. 3 (2014): 199–239. 13. e.g. Nikolaus Walter, ‘Eine vormatthäische Schilderung der Auferstehung Jesus’, New Testament Studies (NTS) 19 (1972–73): 415–29; Joel B. Green, ‘The Gospel of Peter: Source for a Pre-Canonical Passion Narrative?’ ZNW 78 (1987): 293–301; John Dominic Crossan, The Cross That Spoke: The Origins of the Passion Narrative (San Francisco:  Harper & Row, 1988); Alan Kirk, ‘Examining Priorities: Another Look at the Gospel of Peter’s Relationship to the New Testament Gospels’, NTS 40 (1994):  572–95; Martha Stillman, ‘The Gospel of Peter: A Case for Oral-Only Dependence?’ Ephemerides Theologicae Lovanienses (ETL) 73 (1997):  114–20; T. Nicholas Schonhoffer, ‘The Relationship of the Gospel of Peter to the Canonical Gospels: A Composition Critical Argument’, ETL 87 (2011): 229–49.

2. Conceptual Stakes in the Synoptic Problem

17

of the Saviour;14 the Longer (Secret) Gospel of Mark;15 P.Egerton II,16 and several other documents, with the goal of producing a ‘map’ on which to place these various documents. What distinguishes Watson’s view is that unlike most of the analyses of these non-synoptic texts which were content to focus narrowly on the question of whether their contents were dependent in whole or in part on the Synoptics, Watson also proposes a stemma which has direct bearing on the composition of Matthew. This is quite different even from earlier attempts to argue that Q and Thomas had some genetic connection; few attempted to argue that there was a strong or direct connection, and no one, to my knowledge, tried to make Thomas a source of Q, or vice versa. Needless to say, the devil is in the

14.  e.g. Helmut Koester and Elaine Pagels, ‘Report on the Dialogue of the Savior’, in Nag Hammadi and Gnosis:  Papers Read at the First International Congress of Coptology, (Cairo, December 1976), ed. Robert McLachlan Wilson; Nag Hammadi Studies (NHS) 14 (Leiden:  Brill, 1978), 66–74; Christopher M. Tuckett, Nag Hammadi and the Gospel Tradition: Synoptic Tradition in the Nag Hammadi Library (Edinburgh: T& T Clark, 1986); Julian V. Hills, ‘The Three “Matthean” Aphorisms in the Dialogue of the Savior 53’, HTR 84 (1991): 43–58. 15. e.g. Raymond E. Brown, ‘The Relation of “The Secret Gospel of Mark” to the Fourth Gospel’, Catholic Biblical Quarterly (CBQ) 36 (1974): 466–85; Helmut Koester, ‘History and Development of Mark’s Gospel (From Mark to Secret Mark and “Canonical Mark”)’, in Colloquy on New Testament Studies: A Time for Reappraisal and Fresh Approaches, ed. Bruce Corley (Macon:  Mercer University, 1983), 35–57; Scott G. Brown, ‘On the Composition History of the Longer (“Secret”) Gospel of Mark’, Journal of Biblical Literature (JBL) 122 (2003):  89–110; Marvin Meyer, Secret Gospels: Essays on Thomas and the Secret Gospel of Mark (Harrisburg, PA:  Trinity Press International, 2003); Scott G. Brown, Mark’s Other Gospel: Rethinking Morton Smith’s Controversial Discovery, ESCJ 15 (Waterloo, ON: Wilfrid Laurier University, 2005); Stephen Carlson, The Gospel Hoax:  Morton Smith’s Invention of Secret Mark (Waco, TX:  Baylor University, 2005); Peter Jeffery, The Secret Gospel of Mark Unveiled:  Imagined Rituals of Sex, Death, and Madness in a Biblical Forgery (New Haven: Yale University Press, 2007); Francis Watson, ‘Beyond Suspicion: On the Authorship of the Mar Saba Letter and the Secret Gospel of Mark’, Journal of Theological Studies (JTS) 61 (2010): 128–70; Tony Burke, ed., Ancient Gospel or Modern Forgery? The Secret Gospel of Mark in Debate: Proceedings from the 2011 York University Christian Apocrypha Symposium (Eugene, OR: Cascade Books, 2013). The York symposium should have put to rest the suspicion that both the Clementine letter and the Markan extract are modern forgeries by Morton Smith. 16. e.g. Frans Neirynck, ‘The Apocryphal Gospels and the Gospel of Mark’, in The New Testament in Early Christianity: La réception des écrits néotestamentaires dans le christianisme primitive, ed. Jean-Marie Sevrin, BETL 86 (Leuven: Peeters, 1989), 123–75; Jon B. Daniels, ‘The Egerton Gospel:  Its Place in Early Christianity’, PhD diss. (Claremont:  Claremont Graduate School, 1990); Robert L. Webb, ‘Jesus Heals a Leper: Mark 1.40–45 and Egerton Gospel 35–47’, Journal for the Study of the Historical Jesus (JSHJ) 4, no. 2 (2006):  177–202; Tobias Nicklas, ‘Papyrus Egerton 2 – the “Unknown Gospel”’, Expository Times (ExpT) 118 (2007): 261–6.

18

John S. Kloppenborg

details. It is one thing to argue that some of the sayings in Thomas are early and show no dependence on the Synoptics; it is quite another matter to integrate Thomas into a stemma of Synoptic relationship and to posit an early version of Thomas as one of the sources of the Synoptics. 2.1.2 Data, Inference and Editorial Scenarios Second, all would agree that the SP is to be addressed in the first place at the level of literary data and the inferences that these data permit. At least since the publication of Farmer’s The Synoptic Problem there has been a spirited if not sometimes heated discussion of the respective logics and viabilities of solutions to the SP. Most of the discussion to date has focussed, rightly, on three aspects of the problem: (a) the appropriate way(s) to describe the data of the synoptic gospels  – the various patterns of agreements and disagreements in wording and order that exist among the Synoptics; (b) the variety of source-critical configurations that are permitted logically by these data; and (c) the editorial procedures that must be imagined on the part of the evangelists, and the models of the generation and transmission of early Christian documents that must be adopted, in order to render plausible each of the proposed solutions. There are theoretical problems associated with each of these aspects and I will take them in order. 2.1.2.1 Describing the Data First, it is not possible to achieve a completely neutral or objective description of the ‘synoptic data’, both because of text-critical problems and because there are several possible alignments of synoptic texts in parallel, each implying a particular configuration of agreements and disagreements in relation to the overall sequence of pericopae and, hence, each implying a different set of editorial manipulations on the part of successor gospels. From the perspective of text-criticism, we do not possess a stable text of the Synoptics (or any other text containing Jesus-material) from which to argue. And we absolutely do not have texts that imaginably replicate their putative autographs. With each new edition of Nestle–Aland (NA), minor agreements appear and disappear, as do other textual materials that complicate or alleviate problems for one or more solutions to the SP. Huck–Greeven’s policy to avoid harmonistic variants, for example, both created some minor agreements and eliminated others and thus produced a text that was measurably different from the NA texts.17 Because there are many points where, for example, the text of Mark is not unambiguously

17. Frans Neirynck, ‘Greeven’s Text of the Synoptic Gospels’, ETL 58 (1982): 123–4. For an examination of the Greek texts used by synopses of Aland, Huck–Greeven and Orchard, see J. K. Elliott, ‘An Examination of the Text and Apparatus of the Three Recent Greek Synopses’, NTS 32 (1986):  557–82. Elliott does not detect systematic bias in favour of one or other source-critical solution, but notes that individual readings and the way they are displayed favour one or other synoptic solution.

2. Conceptual Stakes in the Synoptic Problem

19

known, it is impossible to be sure whether Matthew and Luke are altering Mark or preserving it. The disappearing and reappearing ἐν at Mark 1.8 is a small case in point. NA25 has simply πνεύματι ἁγίῳ, following B L 892 b t vg, which would suggest that Matthew added ἐν in order to create a better parallel with ἐν ὕδατι in 3.11. Luke’s addition of ἐν would be slightly more difficult to explain, although Neirynck notes a few points at which Luke substitutes ἐν for another preposition.18 Or one might also posit a Q text with ἐν, as the International Q Project (IQP) has done.19 By contrast, NA28 and Huck–Greeven (despite the stated preference for non-harmonistic variants) add ἐν, in which case Matthew and Luke have simply replicated Mark’s phrase. From the perspective of synopsis-construction, the complexity of synoptic data makes it impossible to arrive at a neutral display of synoptic data. Each of the synopses we use inevitably chooses certain parallels as ‘primary’ and makes decisions about which pericopae are ‘in order’ in all three gospels and, by implication, those pericopae that are ‘out of sequence’ in relation to the lead gospel and therefore have been moved by the secondary gospel. For example, the arrangement of pericopae proposed by Aland, who followed Tischendorf ’s Synopsis Evangelica in locating the Sermon on the Mount at Mark 3.19, implies, on the theory of Markan priority, that Matthew delayed Mark 1.29–31, 32–34, 40–45; 2.1–12, 13–17, 18–22, 23–28; 3.1–6, 7–12, 13–19 – ten pericopae in all – to positions after the Sermon on the Mount.20 By contrast the Huck–Greeven arrangement, which places the Sermon on the Mount at Mark 1.39, implies that Matthew moved only Mark 1.29–31, 32–34 to a position after the Sermon. The synopses of Poppi, DenauxVervenne and Crook,21 which put the Sermon on the Mount at Mark 1.21, imply that Matthew delayed only Mark 1.40–45, the cleansing of the leper, to a point after the Sermon. My point is that explanations of Matthew’s editorial procedures 18.  Frans Neirynck, The Minor Agreements of Matthew and Luke against Mark: With a Cumulative List, BETL 37 (Leuven: Leuven University, 1974), 181. 19.  The Critical Edition of Q:  A Synopsis, Including the Gospels of Matthew and Luke, Mark and Thomas, with English, German and French Translations of Q and Thomas, ed. James M.  Robinson, Paul Hoffmann and John S.  Kloppenborg, Hermeneia Supplements (Leuven: Peeters; Minneapolis: Fortress, 2000), 14, hereafter CEQ. 20.  Constantin von Tischendorf, Synopsis evangelica:  ex quattuor evangeliis ordine chronologico concinnavit, brevi commentario illustravit, ad antiquos testes denuo recensuit (Leipzig:  Avenarius & Mendelssohn, 1851 [6. Aufl. 1891; 7. Aufl. 1898]); Kurt Aland, Synopsis Quattuor Evangeliorum (Stuttgart: Württembergische Bibelanstalt, 1963 [2nd edn, 1964; 3rd edn, 1966]) and its successors. 21.  Angelico Poppi, Sinossi diacronica dei quattro vangeli:  italiano (Padova:  Edizioni Messaggero, 2007); Adelbert Denaux and Marc Vervenne, Synopsis van de eerste drie evangeliën (Leuven; Turnhout: Vlaame Bijbelstichtung; Brepols, 1986); Zeba A. Crook, Parallel Gospels: A Synopsis of Early Christian Writing (Oxford and New York:  Oxford University, 2011). This is also the choice of David L. Peabody, Lamar Cope and Allan J. McNicol, One Gospel from Two:  Mark’s Use of Matthew and Luke (Harrisburg, PA:  Trinity Press International, 2002), 85.

20

John S. Kloppenborg

differ, depending on which synopsis is used.22 Did Matthew relocate ten pericopae, or one? What difference does it make to our imaginations of Matthew as an editor and of Matthew’s reasons for editing? As synoptic critics we will have to imagine several ‘Matthews’ – the ‘Aland Matthew’ who is aggressive in his reordering of Mark; a ‘Greeven Matthew’ who is less aggressive; and the ‘Denaux Matthew’ who moves only a single pericope around the Sermon (and in fact a different pericope than the ‘Aland Matthew’) – and think about which ‘Matthew’ is the more plausible. Because there is no neutral way to describe Matthew (or Luke’s) treatment of Mark, an explanation of Matthew’s reordering of Mark on the Aland synopsis will not serve as an explanation of Matthew’s editorial procedures on the Denaux or Greeven arrangement. 2.1.2.2 Logical Inferences Second, the key datum upon which most modern SP hypotheses are built is the non-agreement of Matthew and Luke against Mark in the relative order of the Matt–Mark–Luke material.23 This datum in fact permits any solution in which Mark is medial. It is only with respect to this datum that one can speak of logically possible and impossible solutions. But since the data admit of several logically possible configurations, one cannot ever speak of ‘proof ’ in respect to any particular hypothesis, only disproof in the case of solutions in which Mark is not medial. All of the three most commonly discussed synoptic solutions have Mark as medial: the 2DH, the FH and the 2GH,24 although Mark is medial in these solutions in different ways. It is sometimes suggested that Popperian deductive testing can be invoked to narrow the possibilities.25 Thus Michael Goulder argued that the existence of the so-called minor agreements and the presence of ‘Matthaean’ vocabulary in 22. See John S. Kloppenborg, ‘Synopses and the Synoptic Problem’, in New Studies in the Synoptic Problem, ed. Paul Foster et al., BETL 239 (Leuven: Peeters, 2011), 51–85. 23.  Notwithstanding the claims of Ed P. Sanders, ‘The Argument from Order and the Relationship Between Matthew and Luke’, NTS 15 (1968–69):  249–61, which have been answered in detail by Frans Neirynck, ‘The Argument from Order and St. Luke’s Transpositions’, ETL 49 (1973):  784–815; repr., with additions in Neirynck, Minor Agreements, 291–322. 24. I use the abbreviation 2DH, FH and 2GH for respectively the Two-Document hypothesis, the Farrer (Goulder–Goodacre) hypothesis and the Two Gospel (neo-Griesbach) hypothesis. This usage reflects the collective decision of the participants at the 2008 Oxford conference on the Synoptic Problem, who preferred ‘FH’ for the Farrer (Goulder) hypothesis to other sigla, an abbreviation also used in Marcan Priority Without Q: Explorations in the Farrer Hypothesis, ed. John. C. Poirier and Jeffrey Peterson, Library of New Testament Studies (LNTS) 455 (London and New  York:  Bloomsbury T&T Clark, 2015). The 2015 Roskilde conference preferred L/M (Luke–Matthew), mainly to avoid identifying a hypothesis with a specific individual and to avoid branding the hypothesis with the particular idiosyncrasies of Farrer. 25.  See Karl Popper, Conjectures and Refutations:  The Growth of Scientific Knowledge (London: Routledge and Kegan Paul, 1963), 256.

2. Conceptual Stakes in the Synoptic Problem

21

Luke should be sufficient to refute the hypothesis of the independence of Matthew and Luke.26 There are, however, logical and statistical problems with identifying Goulder’s ‘Matthaean’ vocabulary, and as Goodacre has pointed out, Matthew also contains both ‘Lukan’ and ‘semi-Lukan vocabulary’, which on the logic of Goulder’s argument ought to show that Matthew knew and used Luke.27 Goulder’s argument on the minor agreements would have force only if we were to possess the autographs of the three gospels and if we knew that Matthew and Luke both used the autograph of Mark and only the autograph of Mark. This is hardly the case. For, not only are all of our copies of the Synoptics at least a century younger than their putative autographs but we can hardly suppose that Matthew and Luke used the same copy of Mark. Between MarkMatt and MarkLuke there were likely at least minor variations and perhaps more major disagreements. The many uncertainties that exist in relation to our primary evidence make Popperian deductive testing impossible.28 I  have suggested elsewhere that although Goulder invokes Thomas Kuhn’s model of paradigms, he tries to use (inappropriate) Popperian deductive testing on synoptic hypotheses.29 Kuhn’s model of paradigms in fact allows for the existence of anomalous data without that paradigm collapsing.30 This means that it is best to avoid such terms like ‘disproved’ and ‘discredited’ in

26.  Michael D. Goulder, Luke:  A New Paradigm, JSNTSup 20 (Sheffield:  JSOT Press, 1989), 1–11; ‘Is Q a Juggernaut?’ JBL 115, no. 4 (1996): 667–81. See the response by Robert A. Derrenbacker and John S. Kloppenborg, ‘Self-Contradiction in the IQP? A  Reply to Michael Goulder’, JBL 120, no. 1 (2001): 57–76. 27.  Mark S. Goodacre, Goulder and the Gospels: An Examination of a New Paradigm, JSNTSup: 133 (Sheffield: JSOT Press, 1996), 84–5. 28.  Mark Goodacre (Thomas and the Gospels, 45) has attempted to use deductive testing of the hypothesis of Thomas’s independence of the synoptic gospels, by showing that some sayings in Thomas arguably show influence of the Synoptics, and by appealing to what he called the plagiarist’s charter:  ‘If twenty percent of a student’s essay shows clear signs of plagiarism, it would be no counterargument for the student to complain that the remaining eighty percent of the essay was his or her own work’ (p. 45). I have replied that this argument works only if we knew precisely what we do not know about Thomas and the Synoptics. See ‘New Synoptic Problem’, 199–239, esp. 202–3. 29.  John S. Kloppenborg, ‘Is There a New Paradigm?’ in Christology, Controversy, and Community. FS David Catchpole, ed. David G. Horrell and Christopher M. Tuckett, NovTSup 99 (Leiden:  Brill, 2000), 23–47; repr. with additions as pp.  39–61 in Synoptic Problems: Collected Essays, WUNT 329 (Tübingen: Mohr Siebeck, 2014). 30.  See Thomas S. Kuhn, The Structure of Scientific Revolutions (2nd edn, enlarged; 1962; repr., International Encyclopaedia of Unified Science. Foundations of the Unity of Science 2/2; Chicago: University of Chicago, 1970) and Margaret Masterman, ‘The Nature of a Paradigm’, in Criticism and the Growth of Knowledge: International Colloquium in the Philosophy of Science, London, 1965, ed. Imre Lakatos and Alan Musgrave, Proceedings of the International Colloquium in the Philosophy of Science 4 (Cambridge:  Cambridge University Press, 1970), 59–89.

22

John S. Kloppenborg

relation to any synoptic hypothesis. Such terms are at best rhetorical, attempting to manufacture unearned credibility for one’s own hypothesis, and at worst simply fallacious. Given the nature of synoptic data, we are not in the realm where ‘proofs’ and ‘disproofs’ are possible. General plausibility is all that can be expected. 2.1.2.3 Editorial Scenarios Third, the ‘editorial rationalization’ of any of the possible solutions from 3(b) above is normally conducted by arguing that gospel A is likely the source of gospel B because element y in B can be explained as a plausible borrowing or transformation of x in A, where the inverse relationship is not as plausible:  A(x) → B(y) ^ ¬ B(y) → A(x). It is abundantly clear, however, that such ‘plausibility arguments’ have been adduced for several mutually contradictory directions of borrowing and just as clear that the canons of plausibility differ from critic to critic. This is the level at which most of the SP argument in fact goes on; yet it is also one of the most subjective and reversible parts of the entire enterprise. One of the challenges in respect to assessing ‘plausibility arguments’ is to find ways to discipline one’s own ingenuity in generating possible editorial scenarios. Three kinds of editorial rationalization are typical, which I will call ‘renaming the problem’, ‘coherence arguments’ and ‘externally buttressed arguments’. (i.) Renaming the Problem. In accounting for the differences between one gospel, A (the putative source) that had a certain feature, x, and its successor, B, it is all too common simply to state that B changed A because he ‘liked’ or ‘did not like’ x or had some preference in respect to x. For example, Goulder attempts to account for Luke’s dispersal of material collected in Matthew by appealing to two ‘policies’:  (a) his ‘block policy’  – using Mark as a primary source for one or two chapters and then moving to his other source (Matthew), and (b) his ‘policy’ never to agree ‘with Matthew in the context of a Q saying’.31 This, however, is not an explanation; it only renames the problem. Luke’s disagreement with Matthew is converted into a set of aesthetic preferences of Luke with the help of Goulder’s assumption of Luke’s dependence on Matthew. There is no explanation of why Luke should have such a ‘policy’ of never agreeing with Matthew after Matt 4.11. Francis Watson, in order to account for the fact that Luke’s Sermon on the Plain is much shorter than Matthew’s Sermon on the Mount, posits a Lukan practice of extracting thirteen passages from Matthew’s sermon, writing them in a notebook so that they can be re-incorporated by him later, in any order.32 Although Watson offers this proposal as a response to the oft-made claim that it is impossible to imagine how Luke dismantled Matthew’s sermon, this is also an example of renaming the problem: converting the differences between Matthew and Luke into an aesthetic and scribal procedure of Luke with the help of a source-critical hypothesis. It is hard to see how this is any better an ‘explanation’ (or an explanation at all) of the fact of disagreements in the juxtaposition and sequence of the material in the two sermons than is Vincent Taylor’s suggestion, made long ago 31. Goulder, Luke, 39–41. 32. Watson, Gospel Writing, 170.

2. Conceptual Stakes in the Synoptic Problem

23

(and adapted by Kloppenborg), that Matthew had scanned Q several times (supposing it to be represented by Luke’s order of sayings), extracting from Q sayings that he thought appropriate to the Sermon on the Mount.33 Both ‘proposals’ simply convert disagreements into aesthetic preferences and compositional scenarios of Matthew or Luke. As such, they are not explanations, but simply redescriptions of the problem, parading as explanations. Arguments of this sort are bound to come into conflict with one another. For example, Watson plausibly suggests that on the FH, Luke has extracted Matt 6.9– 13 and 7.7–11 from Matthew’s Sermon, shortened the prayer, composed his own story of the friend at Midnight (11.5–8) in order to create a small cluster of sayings to encourage persistence and patience in prayer.34 Although Watson does not comment on Luke’s relocation of the two Matthaean sayings from the Sermon context, one might also suggest that Jesus’ own prayer in Luke 10.21–22 and his commendation of the disciples provided Luke with the occasion for the insertion of these sayings. But at the same time, one might observe that while the Twelve Thrones saying in Matthew has been rather seamlessly integrated into Mark’s pericope on rewards (Mark 10.23–31 // Matt 19.23–30), Luke detached it from its Matthaean context in order to relocate it to his last supper, Luke 22.24–30, where Luke combines the Twelve Thrones saying with a different Markan text, Mark 10.41–45.35 Aesthetic and ad hoc compositional arguments, alas, support mutually contradictory solutions. 33.  Vincent Taylor, ‘The Order of Q’, JTS New Series 4 (1953):  27–31; ‘The Original Order of Q’, in New Testament Essays: Studies in Memory of T.W. Manson, ed. A. J. B. Higgins (Manchester: Manchester University, 1959), 246–69; John S. Kloppenborg, The Formation of Q:  Trajectories in Ancient Wisdom Collections, Studies in Antiquity and Christianity (Philadelphia: Fortress, 1987), 78–9. 34. Watson, Gospel Writing, 171–3. 35. Watson (Gospel Writing, 175 n. 33) claims that my insistence that Luke’s choice to pass over Matthaean additions to Markan pericopae requires some account ‘seems to assume that the evangelist will incorporate all available source material unless there are specifiable reasons for omitting it’. This is not my assumption. Rather, my point (‘On Dispensing with Q? Goodacre on the Relation of Luke to Matthew’, NTS 49 [2003]: 228–9) is that while one might suppose that Luke did not wish to preserve all of Matthew’s materials, it nonetheless requires some explanation why Luke effected certain relocations. ‘Goodacre himself concedes that it is difficult to know why Luke moved Matt 6.22–3 to its Lukan location (11.34–6). And one wonders why, on Goodacre’s view, Luke did not maintain the connection between Matt 6.24 (serving God or Mammon) and 6.25–34 (an admonition to rely on, and to seek, God rather than possessions) instead of moving the former to 16.13 where it contradicts the point of Luke 16.8, 9–12’ (pp. 228–9). I have no investment in the view that Luke is required to keep the structure of his predecessor materials, but it is incumbent on those who think that Luke abbreviated Matthew to provide some credible reason for his doing so. It is a cop-out to argue that ‘it seemed good to him . . .’ (Watson, Gospel Writing, 175 n. 32): that simply renames the problem and posits an aesthetic preference in Luke’s mind: he did so, because he wanted to. That is not an argument.

24

John S. Kloppenborg

Lest the 2GH be exempted from such criticism, it can be observed that proponents of the 2GH have appealed to similar ‘explanations’. David Peabody and his colleagues account for Mark’s omission of both Matthew and Luke’s Sermons with the conjecture that faced with ‘two very different versions of Jesus’ inaugural sermon at the same point in the common narrative order of Matthew and Luke, Mark chose simply to note the fact that Jesus engaged in teaching at that point’.36 This only renames the absence of the sermon in Mark as his aesthetic ‘choice’. It is doubtful that solutions to the SP can avoid ‘arguments’ that merely redescribe the problem. But it should be recognized that such ‘arguments’ are simply paraphrases of what the ancient author must have done to his source(s), not an explanation of why he did so nor does it make that solution more plausible than alternate solutions. An explanation becomes persuasive if not compelling when one can point to a coherent set of data of which the textual datum under examination is a member, which behave by the same ‘rules’; or that the datum is consistent with ‘rules’ that can be seen to be operative in other texts where one knows the relationship between those texts and their sources. The first kind of tests I will call coherence arguments; the second, externally buttressed arguments. (ii.) Coherence arguments. A more plausible argument can be mounted when a datum to be explained (B’s transformation of A) can be related to a series of analogous transformations in the same documents. This still amounts to positing an aesthetic preference of the editor, but at least that aesthetic preference can be related to a network of supposed transformations evidenced elsewhere. Watson, for example, suggests that differences between Matthew’s ‘spiritualized’ beatitudes, and Luke’s version that focuses on the poor might be explained by pointing to a series of other textual features in Luke where God’s care for the poor is stressed (1.53; 4.18; 16.19).37 One could add that the differences between Matt 6.19–21 and Luke 12.33–34, where Luke rephrases Matthew’s μὴ θησαυρίζετε ὑμῖν θησαυροὺς ἐπὶ τῆς γῆς and substitutes πωλήσατε τὰ ὑπάρχοντα ὑμῶν καὶ δότε ἐλεημοσύνην, is likewise consistent with Luke’s interest in redistributive practices.38 Matthew, however, has an equally pronounced investment in ethical perfectionism and attention to δικαιοσύνη and so his moralizing (rather than ‘spiritualizing’39) of the Q/Lukan beatitudes and his omission of Q 6.24–26 can also be framed as a coherence argument in support of the 2DH.

36. Peabody, Cope and McNicol, One Gospel from Two, 85. 37. Watson, Gospel Writing, 161. 38. See John S. Kloppenborg, ‘Riches, the Rich, and God’s Judgment in 1 Enoch 92–105 and the Gospel According to Luke’, in George W. E. Nickelsburg in Perspective: An On-Going Dialogue of Learning, ed. Jacob Neusner and Alan J. Avery-Peck, Journal for the Study of Judaism Supplement (JSJSup) (Leiden: Brill, 2003), 572–85. 39. The term ‘spiritualized’ is at best misleading and at worst meaningless; ‘moralized’ is a better term, insofar as Matthew has consistently used vocabulary in 5.3–12 that stresses ethical dispositions.

2. Conceptual Stakes in the Synoptic Problem

25

In some cases coherence arguments can be drawn upon to problematize certain spurious assertions. Goulder posits an aesthetic preference in order to ‘explain’ Luke’s version of the Sermon on the Plain. He suggests that Luke did not ‘like’ long speeches and so reduced Matthew’s Sermon on the Mount.40 But this only converts the fact that Luke’s version is shorter than Matthew’s into an aesthetic preference of Luke, with the help of Goulder’s source-critical assumption. His assertion, moreover, runs afoul of the fact that Stephen’s speech in Acts is fifty-two verses, and that Luke has sixty verses of continuous speech material in Luke 12.1–13.9, interrupted only by a few interjections from the crowd or the disciples. Likewise, Luke 15.1–17.10 comprises seventy-three verses of speech material, loosely connected by occasional interjections. Plainly, Luke can and does tolerate long speeches.41 Again I wish to underscore that coherence arguments can be invoked in support of mutually contradictory theories. The observation that Matthew’s wonder accounts are typically much shorter than Mark’s and focus on Jesus’ speech42  – might suggest that Matthew has a consistent practice of streamlining Mark’s stories and eliminating extraneous and redundant detail; But this argument can be reversed, as it has been by proponents of the 2GH, to the effect that Mark consistently expands Matthaean wonder stories to make the accounts more lively.43 In the end, it comes down to which direction of editing one deems to be ‘more plausible’. (iii.) Externally buttressed arguments. Thus far, there have been few efforts to control our fertile imaginations by appealing to the kinds of editorial transformations actually attested in other corpora of literature of a type comparable with the synoptic gospels. Efforts in this directions were made by Thomas Longstaff, who addressed the issue of the incidence of conflation of two exemplars, a fundamental part of the 2GH.44 He adduced mediaeval examples of known conflation and argued that similar literary phenomena are visible in Mark.45 Burton Throckmorton replied, however, that unlike what 2GH Mark must have 40. Goulder, Luke, 39–40, 346. 41. Derrenbacker and Kloppenborg, ‘Self-Contradiction’, 65 n. 21. 42. Heinz Joachim Held, ‘Matthew as an Interpreter of the Miracle Stories’, in Tradition and Interpretation in Matthew (Philadelphia:  Westminster Press; London:  SCM, 1963), 165–300. 43.  For example, Peabody et  al. (One Gospel from Two, 140)  argue that at Mark 5:1– 20 Mark adds numerous details ‘which heighten the story’s dramatic vividness’. Similarly, they note Mark’s expansion of Jairus’ daughter/the woman with the haemorrhage (Mark 5:21–43). 44.  Thomas R.  W. Longstaff, Evidence of Conflation in Mark? A  Study in the Synoptic Problem, Society of Biblical Literature Dissertation Series (SBLDS) 28 (Missoula, MT: Scholars Press for the Society of Biblical Literature, 1977). 45.  Longstaff examined briefly the Greek fragments of Tatian’s Diatessaron and more extensively the accounts of Thomas Beckett’s return to England: John of Salisbury’s Life of Becket and Passio Sancti Thomae, combined by Benedict of Peterborough, and Roger of Hoveden’s conflation of Benedict and the Passio Sancti Thomae.

26

John S. Kloppenborg

done  – microconflation of two parallel exemplars at the level of phrase, clause and sentence – Roger of Hovedon simply alternated blocks of his sources (which is what 2DH and FH Luke does).46 Derrenbacker’s 2005 monograph on compositional practices likewise showed that microconflation was exceedingly rare as a compositional practice.47 F. Gerald Downing has examined the compositional practices of ancient authors in order to offer a critique of Goulder’s suggestions that Luke worked through Matthew’s gospel backwards, and that he had visual access to the whole of the Sermon on the Mount as he moved from Matt 5.42 (Luke 6.30) to 7.12 (Luke 6.31) and then back to 5.46–48 (Luke 6.32–34, 36), and then on again to 7.2 (Luke 6.37– 38),48 deciding what of the Sermon of the Mount to include and what to delay.49 Downing’s point is that the physical procedure that Goulder appears to assume – Luke having visual access to the 9500 characters of Matthew’s Sermon – is unlikely, given the fact that the sermon would represent nineteen average columns of text and, even if one assumed that Matthew’s text was in the largest known columns (3200 characters per column),50 no copy stand (even if they existed) would allow visual access to the entire Sermon.

46. Burton H. Throckmorton, ‘Mark and Roger of Hoveden’, CBQ 39 (1977): 103–6. 47. Robert A. Derrenbacker, Ancient Compositional Practices and the Synoptic Problem. BETL 186 (Leuven: Peeters, 2005). Hence, Luke’s putative procedure on the 2DH or FH is consistent with what is otherwise known of ancient compositional practices. Derrenbacker acknowledges that 2DH Matthew’s conflation of Mark and Q is more problematic, but makes the interesting observation that the more Matthew deviates from the putative (Lukan) order of Q sayings, the lower the verbal agreement with Luke. ‘This trend seems to suggest that as Matthew deviates from Q’s order, he is less inclined to follow the wording of Q. A logical conclusion would be that a lower [agreement] would indicate that Matthew does not have visual contact with that particular Q pericope’ (i.e. that he is citing Q from memory) (p. 238).James Barker challenged Derrenbacker’s findings in a paper entitled ‘Ancient Compositional Practices and the Gospels: A Reassessment’, JBL 135 (2016), 109– 21. He argues that microconflation occurs in 4QDeutn and 8ḤevXIIgr and that conflation required that the scribe have visual access to two exemplars. However the conflation of Deut 5.15 with Exod 22.11 hardly requires visual access, since the two parallel justifications for the sabbath are widely attested, in conflated form in 4QPhyl G (4Q134), and 8QPhyl (8Q3). One hardly needs to posit visual access on the part of the scribe to the phrase, … ‫כי שׂשׂת־ימים עשׂה יהוה את־השׁמים ואת ארץ ואת הים וכול אשׁר בם‬, which is highly formulaic and accessed simply through memory. In the case of 8ḤevXIIgr, it is not a matter of lengthy microconflation at all, but of the scribe translating ‫ יהוה צבאות‬as the palaeo-Hebrew of the tetragrammaton plus τῶν δυνάμεων rather than the LXX’s κύριος παντοκράτωρ. This hardly requires visual access to a Hebrew text of the 12 Prophets. 48. Goulder, Luke, 363–6. 49.  F. Gerald Downing, ‘A Paradigm Perplex: Luke, Matthew and Mark’, NTS 38, no. 1 (1992): 15–36. 50.  Eric G. Turner, Greek Manuscripts of the Ancient World (Princeton:  Princeton University Press, 1971), 108.

2. Conceptual Stakes in the Synoptic Problem

27

Of course, Downing’s rejoinder to Goulder has no effect on other compositional scenarios that might be imagined by proponents of the FH. Other proponents of the FH have been critical of several aspects of Goulder’s imagination of Luke’s procedures. My point here is that attention to the mechanical and physical constraints of composition ought to affect the ways in which we try to solve the SP, at least to rule out procedures that are either otherwise entirely unattested, or that require access to technologies that did not yet exist. One of the more promising ways to ‘normalize’ our imaginations about the kinds of editorial operations that are plausible and those that are not begins with an examination of the ‘canons’ of persuasive speech articulated in the Prosgymnasmata and other rhetorical manuals. As Alexander Damm has shown, the two most commonly recommended rhetorical virtues are clarity (σαφήνεια/ perspecuitas) and propriety (τὸ πρέπον/aptum). Clarity entails both freedom from the risk of obscurity and that the sentence conveys essential information in a way that is not unreasonably delayed. Propriety involves both the skill of inventing and arranging materials to serve the speaker’s purpose and matching the ‘way of speaking’ to the content of the argument.51 Naturally, rhetorical virtues are a function of the purpose of speech; hence, one must always understand the display of rhetorical virtues in the context of the overall purpose of speech. If Luke’s purpose in a pericope differs from Matthew’s, then different rhetorical virtues will apply. Damm examines Luke’s transformation of Matt 12.22–37 through the virtues of clarity and propriety, and finds that Luke 11.14–36 clarifies and shortens Matthew, although the degree of alteration is quite minor. More significant is the issue of propriety, for Luke passes over elements of Matthew’s version that would have enhanced Luke’s argument: ‘Luke describes the demoniac as merely mute (κωφόν). Now it would help Luke to show Jesus’ power if Jesus could cure muteness and blindness – in other words, could make the demoniac “see” (βλέπειν) – especially since ‘blindness’ and ‘sight’ are important for Luke later in the controversy.’52 Damm also notes that Luke’s omission of πόλις from Matt 12.26 seems counterintuitive: [Luke] not only omits ready-made αὔξησις in the form of congeries; he also omits an image germane to his Greek milieu. We cannot explain Luke’s omission of πόλις on grounds that he does not like the term, for he uses it more often (39 times in his gospel) than does Matthew. And once again, retaining πόλις could help Luke weave more tightly his frame of parallels around 10.38–42 . . .53

51.  Alexander Damm, Ancient Rhetoric and the Synoptic Problem:  Clarifying Markan Priority, BETL 252 (Leuven: Peeters, 2013), 69–80. 52. Damm, Ancient Rhetoric, 271. 53. Damm, Ancient Rhetoric, 270.

28

John S. Kloppenborg

Although in Damm’s judgement, Luke misses some opportunities to strengthen his argument, his transformation of Matthew on the FH is ‘on the whole plausible’.54 On the 2DH Luke has done little to the body of Q’s argument: by advancing αὐτός in 11.17 and ὑμῶν in 11.19, he intensifies Q’s emotional proofs. A  more important transformation of Q comes in 11.34–36 where the change of an indicative (εἰ οὖν τὸ φῶς τὸ ἐν σοὶ σκότος ἐστίν becomes an exhortation: σκόπει οὖν μὴ τὸ φῶς τὸ ἐν σοὶ σκότος ἐστίν, something that is typical of an elaborated chria. Second, by changing Q’s εἰ οὖν τὸ φῶς τὸ ἐν σοὶ σκότος ἐστίν, τὸ σκότος πόσον into a positive statement (11.36:  εἰ οὖν τὸ σῶμά σου ὅλον φωτεινόν, . . . ἔσται φωτεινὸν ὅλον ὡς ὅταν ὁ λύχνος τῇ ἀστραπῇ φωτίζῃ σε), ‘Luke makes the classic rhetorical move in conclusions of fostering hope’.55 On Damm’s showing, Luke does not overlook any essential proof material in Q nor does he weaken Q’s treatment of stasis. Of course any talk of Q presupposes that we have independent access to Q, which of course we do not. Yet Damm notes that the analysis of 2DH Luke suggests that Luke does not overlook essential proofs from Mark and so, by inference, he likely did not overlook essential proofs from Q.56 It might be noted that none of Damm’s argument requires us to suppose that Luke or Matthew had rhetorical training; only that as competent writers, each had some grasp of the elements of persuasive speech and would not knowingly overlook the opportunity to craft a persuasive account of Jesus’ engagement with opponents. This kind of approach to assessing the competing models of synoptic relationships elevates the argument beyond the contention that ‘Luke omitted x because it seemed good to him to do so’, insofar as it tries to relate each alteration (or retention) of predecessor materials to the canons of persuasive speech that we know to be current in the Hellenistic world. In this way argument is freed from the subjectivity of what we might think by modern aesthetic standards is a better argument, and grounds judgement in what ancient persons thought was a better and more convincing argument.57 I do not wish to leave the impression that I think that everything is up for grabs and that a solution to the SP is completely open, or to dismiss the hard labour of more than two centuries with the cavalier assertion, ‘it is all, after all, hypothetical’. Still less would I suggest that because of the uncertainties involved, the SP is

54. Damm, Ancient Rhetoric, 274. 55. Damm, Ancient Rhetoric, 276. 56. Damm, Ancient Rhetoric, 279. 57.  A project analogous to Damm’s analysis of synoptic chriae that examines the rhetorical transformation of wonder narratives in Josephus, Plutarch and the synoptics is now published:  Duncan Reid, Miracle Tradition, Rhetoric and the Synoptic Problem, Biblical Texts and Studies 25 (Leuven:  Peeters, 2016). See also Reid, ‘Miracle Stories and the Synoptic Problem’, in New Studies in the Synoptic Problem, ed. Paul Foster et al., BETL 239 (Leuven: Peeters, 2011), 299–319.

2. Conceptual Stakes in the Synoptic Problem

29

a dispensable part of gospel research. In current discussion there are in fact only three serious competitors that can claim substantial bodies of literature (2DH, 2GH and FH),58 and a handful of other hypotheses that are promoted in very restricted circles (the ‘Jerusalem school’)59 or by individuals (e.g. Boismard; Rolland).60 The fundamental points upon which adjudication of these disparate hypotheses turns are relatively few: (a) whether a direct literary relationship among the synoptics can be posited (2DH; 2GH; FH) or not (‘Jerusalem’; Boismard; Rolland); (b)  whether Matthew depends on Mark (2DH; FH [Jerusalem]) or vice versa (2GH); (c) whether Luke used Matthew (2GH; FH) or is independent of Matthew (2DH); and (d) whether Mark can be understood as a conflation of Matthew and Luke (2GH) or not (2DH; FH). While I do not think that we will ever reach a consensus on all of these decision points, a careful and sustained examination will show that the viable explanatory options are finite.

2.2 Towards a Critical History of the Synoptic Problem In addition to the methodological issues outlined above, it is important that we think about conceptual issues connected to SP research. By ‘conceptual issues’ I  have in mind the larger frameworks in the context of which SP research is

58.  There are several variations of the 2DH which include an Ur–Markus or deutero– Markus, or which make Mark dependent upon Q, or which permit secondary influence of Matthew upon Luke. Similarly, within the general scope of the 2GH some solutions posit Lukan access to pre-Matthaean versions of sayings in order e.g. to account for Luke’s apocalyptic sayings in Luke 17. For the purposes of this chapter I will consider these variations to be subtypes of the 2DH and 2GH respectively. 59. On the so-called Jerusalem school, see Robert L. Lindsey, ‘A Modified Two-Document Theory of the Synoptic Dependence and Interdependence’, NovT 6 (1963): 239–63; Malcolm Lowe and David Flusser, ‘Evidence Corroborating a Modified Proto-Matthean Synoptic Theory’, NTS 29 (1983):  25–47; R. Steven Notley, ‘Anti-Jewish Tendencies in the Synoptic Gospels’, Jerusalem Perspective 51 (1996):  20–35, 38. Modern support for the so-called Augustinian hypothesis is negligible. For older literature, see B. C. Butler, The Originality of St. Matthew:  A Critique of the Two-Document Hypothesis (Cambridge:  Cambridge University Press, 1951). For a modified Augustinian hypothesis (also invoking independent use of an oral Urgospel), see John W. Wenham, Redating Matthew, Mark and Luke: A Fresh Assault on the Synoptic Problem (London: Hodder & Stoughton, 1991). 60.  On the ‘multistage hypothesis’, see Marie-Emile Boismard, ‘The Multiple Stage Hypothesis’, in The Interrelations of the Gospels: A Symposium Led by M.-E. Boismard – W. R. Farmer – F. Neirynck. Jerusalem 1984, ed. David L. Dungan, BETL 95 (Leuven: Peeters and Leuven University, 1990), 231–88; Evangile de Marc: sa préhistoire, Etudes bibliques, nouvelle série 26 (Paris: Les Éditions du Cerf, 1994). Another multistage hypothesis is proposed by Philippe Rolland, Les premiers évangiles: un nouveau regard sur le problème synoptique, Lectio Divina 116 (Paris: Les Éditions du Cerf, 1984).

30

John S. Kloppenborg

conducted and the various ways in which those conceptual frameworks have functioned to commend certain solutions and to foreclose others as ‘implausible’. This is a matter of reflecting on the sociology of knowledge – that is, the ways that certain assumptions, conceptual apparatus, vocabulary and explanatory scripts gain salience in certain social, historical and intellectual contexts and, by implication, the ways in which these contexts obscure or render invisible (or ‘improbable’) other assumptions, concepts and scripts.61 It is abundantly clear from an examination of the last two centuries of SP literature that solutions to the SP have commended themselves not simply (sometimes, not even primarily) because of the ways in which they rendered intelligible literary data, but because they were perceived to solve theological or historical problems. It is of course fallacious to argue that principles of advantage or utility ought to justify certain SP solutions, but it is nonetheless a fact that this has been done in the past. We are still in need of a critical history of the SP, one that not only provides a chronicle of the ways in which the synoptic data have been understood and the arguments made on their basis, but the ways in which solutions to the SP have commended themselves because of perceived conceptual utility, sometimes supplementing, sometimes substituting for a genuine analysis of the literary data. Two illustrations will suffice. 1. The birth of the SP was in fact attended by a theological midwife. Lessing’s positing of an Urgospel to account for the synoptics was a direct response to Reimarus’s attack on the credibility of the gospels. For Reimarus, the fraudulent intentions of the evangelists were matched only by their incompetence in perpetrating the fraud, a fraud that he sought to expose by a careful reading of one account against the others. Reimarus was able to use a long list of contradictions among the gospels to drive wedges between Jesus and the evangelists, on the one hand, and between the gospels and theological dogma that was seen to derive from them, on the other.62 Lessing’s answer, which posited an Aramaic Urgospel, had little to do with solving literary problems; indeed he did not discuss specific textual items at all and offered no reconstruction of his hypothetical Urgospel. Instead, Lessing was out to rehabilitate the gospels’ descriptions of Jesus and the work of the evangelists and to preserve the truth of the Christian religion. The latter he did by insisting that ‘the accidental truths of history can never become the proof of necessary truths of reason’.63 The truth of Christian doctrine did not ultimately rest on contingent 61.  A basic work on the sociology of knowledge is Robert K. Merton, ‘Insiders and Outsiders: A Chapter in the Sociology of Knowledge’, American Journal of Sociology (AJS) 78, no. 1 (1972): 9–47. 62. Hermann Samuel Reimarus, ‘Von dem Zwecke Jesu und seiner Jünger’, in Fragmente des Wolfenbüttelschen Ungenannten, ed. Gotthold Ephraim Lessing (Braunschweig:  C. A. Schwetschke & Sohn, 1778), 3–174. 63.  Gotthold Ephraim Lessing, Theological Writings:  Selections in Translation, ed. Henry Chadwick (London:  A. & C.  Black, 1956), 53; see Henry E. Allison, Lessing and the Enlightenment: His Philosophy of Religion and Its Relation to Eighteenth-Century (Ann Arbor, MI: University of Michigan, 1966), 96.

2. Conceptual Stakes in the Synoptic Problem

31

formulations in the gospels. The former purpose he accomplished by an act of sheer speculation:  the gospels derived from an early source supposedly known to Epiphanius and Jerome and purportedly produced by ‘honest people who had had personal converse with Christ’.64 The idea of an early source solved no particular compositional problem but it gave Lessing leverage against Reimarus’s frontal attack on the gospels and the theological assertions that in the eighteenth century were seen to derive from them. As long as Lessing’s ‘solution’ was not expected to provide a compelling resolution of specific literary problems and as long as the dubious patristic ‘evidence’ was not probed too closely, his solution could have survived. But the tide of nineteenth-century scholarship increasingly demanded explanation of fine literary detail and was increasingly dubious about patristic statements. Nevertheless, for a considerable period after Lessing, pre-gospel sources (Schleiermacher’s Λόγια, Q, the Gospel of the Hebrews, UrMarkus, etc.), all left vaguely defined, were invoked by synoptic scholars in order to imagine a connection between Jesus and the later inscription of the gospels. More often than not, such sources did as much to exorcize the ghost of Reimarus as they did to solve specific textual problems. 2. Holtzmann’s 1863 treatment of the SP was qualitatively different from Lessing’s insofar as it put no stake in patristic testimony and attempted a resolution of many actual literary problems. As a full demonstration of a solution to the SP, however, Holtzmann’s treatment was seriously flawed and encumbered by many unnecessary assumptions. This is not to say that Holtzmann’s solution was without merit; he did provide a plausible accounting for the final shape of Matthew and Luke given the assumption of their common use of UrMarkus and a ‘Logia-source’. What was problematic in his solution was not what he did, but what he failed to do. Along with Bernard Weiss, Holtzmann responded directly to the Griesbach hypothesis,65 and specifically engaged Hilgenfeld’s version of the Augustinian hypothesis; he did not, however, systematically examine other options (e.g. Luke’s direct use of Matthew), but proceeded immediately to his own solution.66 In spite of the lack of rigour in Holtzmann’s treatment, it was generally hailed as a ‘demonstration’ of (Ur)Markan priority and of the existence of a sayings source. 64.  Lessing, ‘Neue Hypothese’; English Translation:  ‘New Hypothesis Concerning the Evangelists Regarded as Merely Human Historians’, in Lessing’s Theological Writings, trans. Henry Chadwick (London: A. & C. Black, 1956), 79. 65.  Bernhard Weiss, ‘Zur Entstehungeschichte der drei synoptischen Evangelien’, TSK 34 (1861):  29–100, 646–713; Heinrich Julius Holtzmann, Die synoptischen Evangelien: Ihr Ursprung und geschichtlicher Charakter (Leipzig: Wilhelm Engelmann, 1863), 344–54. 66.  Only later did Holtzmann abandon the idea of an UrMarkus, which left him to explain the Matthew–Luke agreements against Mark. Thus he posited the sporadic influence of Matthew upon Luke:  Heinrich Julius Holtzmann, Lehrbuch der historischkritischen Einleitung in das Neue Testament (2. edn. rev. [and] enl.; Freiburg i.Br.:  J. C. B. Mohr [Paul Siebeck], 1886), 363–65; Lehrbuch der historisch-kritischen Einleitung in das Neue Testament (3. verbesserte und vermehrte Aufl.; Freiburg i.Br.:  J. C.  B. Mohr [Paul Siebeck], 1892), 35.

32

John S. Kloppenborg

Part of this assessment may have had to do with the general collapse of support for the Griesbach hypothesis after the death of F. C. Baur in 1860. But probably a more significant factor in the widespread endorsement of Holtzmann’s solution was the way in which it provided documentation for the historical Jesus of liberal theology. Holtzmann’s UrMarkus (A) lacked the ‘dogmatic’ features so evident in Matthew and Luke and Holtzmann himself produced a sketch of the life of Jesus in which he used Markan material to circumscribe the development of Jesus’ consciousness in seven identifiable stages. Holtzmann in fact declared this life of Jesus as the ‘most valuable result of our investigations’.67 For the next four decades, a long string of ‘lives of Jesus’ would be published, all capitalizing on Holtzmann’s view of Mark. Typical of these ‘lives’ was the interpretation of the kingdom of God as a spiritual kingdom of repentance and the conviction, based on Holtzmann’s reading of Mark, that Jesus’ messianic consciousness developed, precipitated by a ‘Galilean crisis’ in which Jesus came to face the failure of his mission. Albert Schweitzer, commenting on the dominance of Markan priority during this period, sagely observed that ‘[t]he victory . . . belonged, not to the Marcan hypothesis pure and simple, but to the Marcan hypothesis as psychologically interpreted by a liberal theology’.68 Indeed, the salience of Holtzmann’s solution, while in part deriving from the elegance of its solution to various literary problems, probably had as much to do with the way it was presumed to cohere with the dominant Romantic cultural construct that stressed the ‘personality’ of the ‘great man’ in history, a notion that can be traced to Goethe, and which was variously elaborated by Schleiermacher and Rénan.69 In Holtzmann’s ‘A’ source, Jesus emerges as just such a figure – possessing ‘wonderful energy as has not been attested empirically elsewhere’, ‘genuine humanity,’70 developing through various crises into a character of moral excellence and certainly worthy of imitation.71

67.  Holtzmann, Die synoptischen Evangelien, 468 See John S. Kloppenborg, ‘H. J. Holtzmann’s Life of Jesus According to the ‘A’ Source. Part 1’, JSHJ 4, no. 1 (2006): 75–108; ‘H. J. Holtzmann’s Life of Jesus According to the ‘A’ Source. Part 2’, JSHJ 4, no. 2 (2006): 203–23. 68.  Albert Schweitzer, The Quest of the Historical Jesus: A Critical Study of Its Progress from Reimarus to Wrede, trans. William Montgomery, with a preface by F.  C. Burkitt (New York: Macmillan, 1910), 204 (emphasis added). 69.  David Friedrich Strauss, Das Leben Jesu: kritisch bearbeitet, Theologische StudienTexte 15 (Tübingen: C. F. Osiander, 1835–36); Ernest Renan, Vie de Jésus, Histoire des origines du Christianisme 1 (Paris:  Michel Lévy Frères, 1863). On the ‘great man’ see Dieter Georgi, ‘The Interest in Life of Jesus Theology as a Paradigm for the Social History of Biblical Criticism’, HTR 85, no. 1 (1992): 51–83, esp. 77–80; Kloppenborg, ‘H. J. Holtzmann’s Life of Jesus’, esp. 86–93. See also Franz Rosenzweig, ‘Atheistic Theology (1914)’, in Philosophical and Theological Writings, ed. Paul W. Franks and Michael L. Morgan (Indianapolis, IN: Hackett Pub., 2000), 10–24. 70. Holtzmann, Die synoptischen Evangelien, 496. 71. Curiously, even though Holtzmann’s ‘A’ source contained Mark 13, Holtzmann’s ‘life’ for the most part passed over in silence the mini-apocalypse, noting parts of Mark 13 only in connection with the use of the Son of Man title and Jesus’s embrace of futuristic expectations.

2. Conceptual Stakes in the Synoptic Problem

33

Many other such examples could be cited of solutions to the SP being commended not (only) because they provided a satisfying account of the textual details of the Synoptics but because they (also) cohered with larger cultural agenda – F. C. Baur’s use of the Griesbach hypothesis and the liberal appeal to Markan priority are conspicuous cases in point. It should not be concluded that literary and theological arguments were always confused; Lagrange, for example, who recognized the utility of Markan priority for the Catholic idea of the development of doctrine, nonetheless insisted that the SP should be solved quite independently of theological considerations and did not allow such considerations to intrude on his literary analysis. Nor should it be concluded that the 2DH or the 2GH depend on essentially theological assumptions or that these solutions to the SP are simply disguised theological apologetics. That would be both naïve and would trivialize the serious and substantial literary critical observations that were adduced by such literary critics as Lachmann, de Wette, Weiss and Holtzmann and later Lagrange, who saw it as their principal task to address literary relationships. While the theological advantage afforded by a particular solution to the SP should not be a treated as an argument in its favour,72 it is nonetheless true that SP research is conducted in a context where much is at stake conceptually. In the current discussion, the SP is connected to history of Jesus scholarship, and to the issue of the conceptual landscape of the early Jesus movement. 2.2.1 The Synoptic Problem and the Historical Jesus The Synoptic Problem has of course been implicated in the quest of the historical Jesus at various stages. As I have noted above, Holtzmann’s isolation of the ‘A’ source was the basis for a ‘life of Jesus’ that was influential for the next generation of liberal scholarship. Forty years later, and in the wake of Wrede’s analysis of Mark and the collapse of confidence in Mark as a source for Jesus’ messianic consciousness,73 Harnack substituted a reconstruction of Q. Q, he declared, was uncontaminated by dogma, bias and the exaggerated apocalypticism of Mark and provided direct access to Jesus’ self-understanding.74 72.  e.g. R.  Steven Notley (‘Anti-Jewish Tendencies’) argues that the Luke → Mark → Matthew order of the Jerusalem school is to be preferred because it corresponds to a trajectory of increasingly anti-Jewish sentiments. William R. Farmer (The Gospel of Jesus: The Pastoral Relevance of the Synoptic Problem [Louisville, KY: Westminster John Knox, 1994], 198–9) comes perilously close to arguing for the 2GH from the supposed benefits it confers, and against the 2DH for the theological and historical disadvantages he perceives in it. 73.  William Wrede, Das Messiasgeheimnis in den Evangelien (Göttingen:  Vandenhoeck & Ruprecht, 1901). 74.  Adolf von Harnack, The Sayings of Jesus:  The Second Source of St. Matthew and St. Luke, trans. John Richard Wilkinson, New Testament Studies 2 (London:  Williams & Norgate; New York:  G. P. Putnam’s Sons, 1908), 171: ‘Q is a compilation of discourses and sayings of our Lord, the arrangement of which has no reference to the Passion, with an horizon which is as good as absolutely bounded by Galilee, without any clearly discernible bias, whether apologetic, didactic, ecclesiastical, national or anti-national’. See also ibid., 173.

34

John S. Kloppenborg

A few more recent authors have argued for a strong correlation between solutions to the SP and the reconstruction of the historical Jesus. Harnack’s enthusiasm for Q has found a recent advocate in James Robinson, who argued: ‘Since the reconstructed Sayings Gospel Q is the best source that exists today to get back to what Jesus actually had to say, it is also the best source for understanding what he thought he was doing, what he was up to in what he did.’75 Relying on the literary and stratigraphical analysis of Q proposed by Kloppenborg, Robinson went further than Harnack in arguing that it was in fact the putative earliest stratum of Q, not Q as a whole, that provides access to the historical Jesus. This conclusion, in my view, confuses compositional history with tradition history and illegitimately turns Q’s editorial choices into evidence of what is authentic or inauthentic.76 Although Michael Goulder did not write a historical Jesus monograph, his synoptic theory is equally strong, but led to very different consequences. At the beginning of the synoptic gospels is what he calls the Peter–James–John material that formed the basis of Mark. He argues strenuously against the easy assumption that Matthew had other ‘traditions’; as he puts it, ‘there is no Troy beneath the Matthaean field except where it overlaps the Marcan’.77 As the title of his first book suggests, Midrash and Lection in Matthew, Matthew was not a collector of ‘traditions’; ‘his editing is what other people call composing’.78 And if Luke is a transformation of Mark and Matthew, with Luke composing what he did not receive from Mark and Matthew, then the volume of material upon which to construct a portrait of the historical Jesus reduces to almost nothing, especially when we take into account the constructed nature of Mark’s gospel. A final example of a strong correlation of the SP and reconstructions of the historical Jesus is found in William Farmer’s work. Farmer sagely observed that some of the more conservative treatments of the historical Jesus that supposedly adopted the 2DH – he mentions Ben Meyer’s The Aims of Jesus and Otto Betz’s Jesus, der Messias Israels79  – ‘employ a sophisticated hermeneutic, according to which the raw consequences of the Two-Source hermeneutic are hidden or obscured by appeal to the theory that later documents can preserve earlier forms of material’.80 For Farmer, this represented a betrayal of principles. Farmer’s own The Gospel of Jesus took the 2GH with utmost seriousness. He based his portrait of

75.  James M. Robinson, The Gospel of Jesus: In Search of the Original Good News (San Francisco: HarperSanFrancisco, 2005), 9. 76. See Kloppenborg, Formation, 245: ‘Tradition-history is not convertible with literary history’ (emphasis original). 77. Michael D. Goulder, Midrash and Lection in Matthew (London: SPCK, 1974), 137. 78. Goulder, Midrash and Lection, 152. 79. Ben F. Meyer, The Aims of Jesus (London: SCM, 1979); Otto Betz, Jesus, der Messias Israels:  Aufsätze zur biblischen Theologie, WUNT II/42 (Tübingen:  J. C.  B. Mohr [Paul Siebeck], 1987). 80. Farmer, Gospel of Jesus, 198.

2. Conceptual Stakes in the Synoptic Problem

35

Jesus (almost) exclusively on Matthew.81 This also meant that Farmer consistently argued that Luke did not have access to non/pre-Matthaean traditions. The Lord’s Prayer, for example, in Luke 11.2–4 was Luke’s abbreviation of an original poetic (Matthaean) prayer, representing the prayer that Jesus taught his disciples.82 The ‘strong’ position has some important consequences beyond the Lord’s Prayer. The first Lukan beatitude, μακάριοι οἱ πτωχοί, ὅτι ὑμετέρα ἐστὶν ἡ βασιλεία τοῦ θεοῦ, commonly thought to reflect a core aspect of Jesus’ interests, would disappear under both the FH and 2GH. This should have been the case for Farmer, but he mounted a rather tortured argument in order to retain this for the historical Jesus. He insisted on the one hand that Jesus was committed to those who were economically disadvantaged, in spite of the plainly moralizing language of Matt 5.3–10 and, on the other, claimed (bizarrely) that the 2DH gives less place to concern for the poor than Matthaean priority. The forced nature of his argument seems a symptom of the problem that a strong approach to the 2GH created for him.83 Goulder, by contrast, is more consistent with his theoretical principles, arguing that Matthew has created the beatitudes from Psalm 119 and Isaiah 61, and that Luke has abbreviated the Matthaean beatitude, in part because of his prioritizing of the poor.84 Neither Jesus’ disciples nor his subsequent followers were poor. On the contrary they had ‘respectable middle-class backgrounds . . . So all the theories of an original addressed by Jesus to the ʿanāwîm, poor, hungry, tearful people in the simple Lukan form, are really only romantic speculations, sustained by the happy coincidence that they do not mention any virtues, and so appeal to believers in justification by faith’.85 81.  See the excellent analysis of William R. Farmer’s Jesus and the Gospel:  Tradition, Scripture, and Canon (Philadelphia, PA: Fortress, 1982) by William E. Arnal, ‘The Synoptic Problem and the Historical Jesus’, in New Studies in the Synoptic Problem: Oxford Conference, April 2008. Essays in Honour of Christopher M.  Tuckett, ed. Paul Foster et  al., BETL 239 (Leuven: Peeters, 2011), 383–95. 82. Farmer, Gospel of Jesus, 43–8. Likewise, Allan J. McNicol, Beyond the Q Impasse – Luke’s Use of Matthew: A Demonstration by the Research Team of the International Institute for Gospel Studies, in collaboration with David L. Dungan and David B. Peabody (Valley Forge, PA: Trinity Press International, 1996), 173–4. Similarly, from the standpoint of the FH, see Ken Olson, ‘Luke 11.2–4: The Lord’s Prayer (Abridged Edition)’, in Marcan Priority Without Q, 101–18. 83. Farmer, Gospel of Jesus, 103. Farmer claims that advocates of the 2DH ‘question the authenticity of [Q 6.20b]’ (ibid.). This claim is plainly false. 84.  Goulder, Midrash and Lection, 186 n.  85, 254; Luke, 346–58. Mark Goodacre (The Case Against Q:  Studies in Markan Priority and the Synoptic Problem [Harrisburg, PA: Trinity Press International, 2002], 133–51) rejects Goulder’s view that Matthew created the beatitudes, but accepts that Luke abbreviated them. 85.  Goulder, Luke, 357. Matthew 5.3 did not figure at all in Ed Parish Sanders, Jesus and Judaism  (London:  SCM, 1985), but in The Historical Figure of Jesus (Harmondsworth, Middlesex: Penguin Books, 1993), 203, Sanders elides the difference between Matt 5.3 and Luke 6.20b with the claim, ‘bless the downtrodden, the poor and the meek, as well as those who hunger and thirst for righteousness, the merciful, the pure in heart, and the peacemakers. These sayings imply demands, but the clearest note is sympathy and promise for those who needed them most’.

36

John S. Kloppenborg

Farmer’s complaint about a betrayal of principles has been put somewhat differently by Dieter Lührmann, who in 2001 observed that despite the hegemony of the 2DH in German New Testament scholarship, it has been Matthew’s portrait of Jesus that has set the agenda for historical Jesus scholarship: A critical examination of literature on Jesus, even the most recent, arouses the suspicion that the Two Document hypothesis may have not been taken seriously enough in its methodological consequences. Generally speaking it seems to me that it is Matthew’s image of Jesus that has left an impression. It is only on this basis that one can explain the fact that the question of Jesus’ understanding of the Law is the apparently unavoidable issue for historical Jesus research, while in Mark the word νόμος does not even occur, and in Q is only in apparently late texts, Q 16.16–17.86

Ten years later, William Arnal came to much the same conclusion. one searches in vain for robust or consistent correlations between the portraits of the historical Jesus and the solutions to the synoptic problem those portraits begin with. This lack of strong and reliable correspondence should in fact be no more surprising than the opposite conclusion that some correlation does exist: the synoptic problem is a literary problem, and the issue of the historical Jesus is, by contrast, an historical one.87

In Arnal’s view, however, the correlation of the SP with historical Jesus research cannot and should not be absolute, for several reasons. First, a strong correlation confuses literary history with tradition history; that is, Matthew or Mark or Luke or Q’s editorial choices to frame their works in a certain way do not have a necessary bearing on the age or authenticity of the materials they chose to include (or exclude). In addition there are simply too many unknowns in the SP: We do not know what form of Mark was used by Matthew and Luke (since it is hardly likely that they used the same copy). Nor can we know whether the changes of Mark that are visible in Matthew and Luke are the result of tendentious editing, or knowledge of other performances of materials parallel to Mark. Nor do we have a real grasp on the degree to which other materials were available to the evangelists. Nor is it clear in what sense, for example, Matthew used Mark – as a slavish copyist/editor, or accessing Mark through memory, or accessing Mark through oral performances of Mark.88 As much as they are decried by Farmer, weaker correlations of the SP and reconstructions of the historical Jesus are probably the norm and are completely justified. Goodacre allows for Luke’s occasional access to orally transmitted material that might be more primitive than Matthew’s versions, in particular in connection

86. Dieter Lührmann, ‘Die Logienquelle und die Leben-Jesu-Forschung’, in The Sayings Source Q and the Historical Jesus, ed. Andreas Lindemann, BETL 158 (Leuven: Peeters and Leuven University Press, 2001), 196 (my translation). 87. Arnal, ‘Synoptic Problem and the Historical Jesus’, 375. 88. Arnal, ‘Synoptic Problem and the Historical Jesus’, 379–80.

2. Conceptual Stakes in the Synoptic Problem

37

with the Lord’s Prayer.89 Departing from Goulder’s strict view of Matthew’s creativity, Francis Watson invokes as a saying collection (perhaps more than one) to which Matthew and Mark had access.90 To conclude that weak correlations are justified is not to argue that there should be no correlation and this is, I take it, partly behind Lührmann’s complaint that material which on the 2DH is largely due to Matthaean redaction has nonetheless been allowed to control certain reconstructions. On the FH it would be a surrender of principles if one were consistently to harmonize Lukan with Matthaean forms of sayings when the Lukan form ‘seemed’ more primitive, just as it would on the 2GH to adduce Mark against the other two when it was convenient to do so.91 Source theories have entailments, even if we are not in a position to map SP solutions neatly onto reconstructions of the historical Jesus. A final point on the historical Jesus. Watson’s admonition not to assume canonical boundaries when thinking about the SP applies as well to the historical Jesus. Since D. F. Strauss, it has been the rather plain-speaking Synoptic Jesus who has controlled the landscape of historical Jesus studies, even if there are many differences among these studies. The gravitational pull of the Synoptic Jesus is extremely strong and when it is coupled with the fact that this Jesus is still the object of worship, prayer and pious sermonizing, the result is that various bits of data are typically left aside in reconstructions of the historical Jesus. For example, the Jesus who praised those who castrated themselves (Matt 19.12), perhaps unsurprisingly, figures in none of the recent historical Jesus books even though on the 2GH it perhaps ought to have been. The agraphon, ‘become approved money-changers’,92 and Thomas’s parable of the assassin (GThom 98) are neglected entirely.93 89. Goodacre, Case Against Q, 64. 90. Watson, Gospel Writing, 249–85. 91.  C. S. Mann, Mark: A New Translation with Introduction and Commentary, AB 27 (Garden City, NY: Doubleday, 1986), 240–3 occasionally (and implausibly) asserts that Mark has access to eyewitness detail not present in the parallel accounts of Matthew and Luke. 92.  Alfred Resch, Agrapha:  Aussercanonische Schriftfragmente, Texte und Untersuchungen (TU) NF 15, 3–4 (2nd edn; Leipzig:  J. C.  Hinrichs, 1906 [repr. Darmstadt:  Wissenschaftliche Buchgesellschaft, 1974]), no. 87. Resch lists more than sixty citations and allusions to this agraphon. 93.  I have suggested elsewhere that the anxiety about the so-called Cynic hypothesis (either in relation to Q or Jesus [often confused with Q]) is a symptom of explanatory habits that reflect assumptions about what Jesus must have been (and therefore what he cannot have been) – assumptions that are not grounded in sound historiography but in theological convictions. See John S. Kloppenborg, ‘A Dog among the Pigeons: The ‘Cynic Hypothesis’ as a Theological Problem’, in From Quest to Q. FS James M. Robinson, ed. Jon Ma. Asgeirsson, Kristin de Troyer and Marvin W. Meyer, BETL 146 (Leuven: Peeters, 1999), 73–117. Recently Stephen Patterson has argued that material transmitted only in the Gospel of Thomas on androgyny might be considered once the monopoly of the Synoptic Gospels on historical Jesus scholarship is rethought:  Stephen J. Patterson, ‘The Apocryphal Gospels and North American Historical Jesus Research’, in Forbidden Texts on the Western Frontier: The Christian Apocrypha in North American Perspectives. Proceedings from the 2013 York University Christian Apocrypha Symposium, ed. Tony Burke (Eugene, OR: Cascade Books, 2015), 173–85.

38

John S. Kloppenborg

2.2.2 The Conceptual Landscape of the Early Jesus Movement Perhaps a more consequential entailment of the SP research has to do with mapping the conceptual terrain of the early Jesus movement. One critical difference between the FH and 2GH on the one hand, and the 2DH on the other, is the latter’s positing of a literary object, and hence the product of some sort of intentional editing, that did not feature a salvific representation of Jesus’ death and perhaps imagined the vindication of Jesus not as a resurrection (bodily or otherwise), but as an assumption.94 However one understands the literary organization and conceptual framing of Q, it is plain that Q does not inscribe the death and vindication of Jesus in the way that Mark (and his successors) did. Even if one resorts to the expedient of positing Q’s knowledge of the passion kerygma – an expedient that is not especially credible95 –, it remains that Q never makes any use of or allusion to that kerygma. The 2DH thus posits as one of the earliest literary stages of the Jesus movement a document (Q) that is far less invested in christological apologetics and in advancing an explanation of Jesus’ death (which flows from christological apologetics) than any of the intracanonical gospels. Although the relationship between Q and the Gospel of Thomas is not especially close, on the matter of a lack of attention to a salvific interpretation of Jesus’ death, they coincide. The absence of a salvific interpretation of Jesus’ death in Q also makes it difficult to imagine how the Jesus movement that Q represented could have adopted eucharistic practices such as that represented in Mark or in the Pauline churches. Hence, to take the 2DH seriously implies imagining at least some degree of significant pluralism in the theology and practice of the early Jesus movement. The 2DH does more than this. I  have suggested elsewhere that Q, unlike Mark, Matthew and Luke, represents the product of a rural Jesus movement and its social imaginary does not range much outside of Jewish Palestine. This is in contrast to Mark’s privileging of non-Jews  – dramatically, a Roman who is the first to grasp the identity of Jesus as son of God – and his attention to the gentile cities of Decapolis, Tyre and Sidon and Caesarea Philippi. Matthew too assumes a world in which travelling magi come to visit a king (and are surprised that the king is in fact born in a small town). Matthew’s Jerusalem is called ‘the holy πόλις’ (4.5) and the ‘πόλις of the Great King’ (5.35). It is a world where kings make gigantic loans or entrust retainers with huge sums of gold (18.23–35; 25.14–30). In Matthew’s spatial imaginary, the city is at the centre. In his parable of the Wedding Banquet (22.1–10) Matthew’s king destroys one πόλις and 94.  On the representation of Jesus’ death, see John S. Kloppenborg, Excavating Q: The History and Setting of the Sayings Gospel (Minneapolis:  Fortress Press; Edinburgh:  T&T Clark, 2000), 363–74; on the vindication of Jesus as an assumption, see Daniel A. Smith, The Post-Mortem Vindication of Jesus in the Sayings Gospel Q, LNTS 338 (London and New York: T&T Clark International, 2007). 95. See John S. Kloppenborg, ‘ “Easter Faith” and the Sayings Gospel Q’, in The Apocryphal Jesus and Christian Origins, ed. Ron Cameron, Semeia 49 (Atlanta: Scholars, 1990), 71–99.

2. Conceptual Stakes in the Synoptic Problem

39

sends out his servants even ἐπὶ τὰς διεξόδους τῶν ὁδῶν of the city in order to invite guests to a feast (22.9). It is where kings sit in judgement over subjects (25.31–46). These are all urban images. Luke even more than Matthew lives in a world of cities. His Gospel is centred on Jerusalem and its temple: this is where his Gospel begins, with an angel appearing to Zechariah, and where it ends, with the disciples remaining in Jerusalem and daily visiting its temple. He repeatedly treats sites that are properly identified as villages as πόλεις. And in Acts, Paul moves from city to city in Asia and Greece, establishing churches. Luke pays no attention at all to the countryside.96 For both Matthew and Luke, the issues precipitated by the expansion of the Jesus movement into non-Jewish areas are critical in the construction of those gospels:  kashruth and sabbath observance for both, and circumcision for Luke. In Q, by contrast, the countryside and private spaces – villages, their houses, and the wasteland  – are the points on Q’s ‘map’ privileged with belief and the embrace of the kingdom of God. Recently Giovanni Bazzana has shown how Q’s social imaginary of the ‘kingdom’ reflects the attitudes and values that can be documented in the low-level village scribes that served as administrators in Ptolemaic and early Roman Egypt.97 The way in which Q imagines the places where the kingdom is embraced is very different from the ways in which Matthew and Luke, in their world of cities, imagine the presence of the kingdom. For Q cities are places where one can expect rejection, arrogance, or indifference; it is in villages and towns – not the cities of Sepphoris or Tiberias which are ignored – that one can expect the ‘children of peace’. If we had only Matthew and Luke, we would not see this face of the Jesus movement, the face of its earliest embodiment in the villages and towns of Galilee. Whereas Luke’s world is the world of the Gentile cities of the eastern Mediterranean, and eventually, Rome, Q’s world is centred in the Galilee, on the towns of Capernaum, Bethsaida, Khorazin and Nazara, the latter two so small and insignificant that they did not merit mention in the Hebrew Bible or in Josephus’s writings. Q’s world is a Jewish or ‘Judean’ world where Gentiles are at the periphery  – to the north in Tyre and Sidon (Q 10.13–14) or the extreme north in Nineveh (Q 11.32). Gentiles make appearances in Q but they are exceptional: a centurion (probably imagined to be a Roman) meets Jesus and expresses a surprising confidence in Jesus (Q 7.1–10). Q tells this story not to construct a model of faithfulness or to promote a gentile mission, but precisely because his confidence

96.  See John S. Kloppenborg, ‘Luke’s Geography: Knowledge, Ignorance, Sources, and Spatial Conception’, in Luke on Jesus, Paul, and Earliest Christianity: What Did He Really Know?, ed. Joseph Verheyden and John S. Kloppenborg, Biblical Tools and Studies (BTS) 29 (Leuven: Peeters, 2016), 101–43. 97.  Giovanni Battista Bazzana, Kingdom of Bureaucracy:  The Political Theology of Galilean Village Scribes, BETL 274 (Leuven:  Peeters, 2015). See earlier, Giovanni Battista Bazzana, ‘Basileia and Debt Relief:  The Forgiveness of Debts in the Lord’s Prayer in the Light of Documentary Papyri’, CBQ 73, no. 3 (2011): 511–25.

40

John S. Kloppenborg

is so exceptional and because it puts to shame those who in Q’s view ought to have confidence in Jesus.98 Finally, in contrast to Mark and Mark’s successors, Q apparently lacked controversies over Sabbath violations (compare Mark 2.23–38; 3.1–6)99 and disputes about kashrut (food laws; Mark 7.15–23).100 In contrast to Luke, there is no anticipation that Christ-followers would give up the observance of Judaean table practices, as Acts 10 seems to enjoin. Nor does Q question the practice of circumcision, a key topic of Paul’s letter to the Galatians and of Acts 15. This is probably because Q presupposed an exclusively Israelite environment where people naturally circumcised their sons, kept kashrut, and observed the Sabbath. There was indeed no reason not to observe these commandments. Q’s complaint with other Jewish groups is not that they observed the Torah and the Q people did not. Rather, Q’s complaint against the Pharisees – no doubt, a caricature – is that they insist on one set of commandments and neglect others. I raise the issue of Q’s rural character and its Judaean complexion not as an argument for its existence – after all that would be a complex version of renaming the problem – but instead to point out that the 2DH has as one of its possible implications that we get a glimpse of a form of the Jesus movement that has, in the editing of Mark and his successors, been mainly obscured. This was a form of the Jesus movement that was largely rural, and oriented to practices connected to a conception of the kingdom of God. Because the framing of Mark’s gospel is driven by christological apologetics – apologetics that Mark’s successors could hardly help but be affected by – the FH and 2GH create a different impression: that christology was at the heart of the Jesus movement from its very beginning. That might (historically) be true. But we arrive at a rather different historical scenario if we wager with the 2DH. Christology, to be sure, is part of Q’s interest, but it is subordinated to the more dominant notion of practices and ethics related to Q’s view of the βασιλεία.

98.  In Formation I argued that Q was implicated in the ‘gentile mission’ and Harry T. Fleddermann, Q: A Reconstruction and Commentary, BTS 1 (Leuven: Peeters, 2005) argues that Q is fully involved with Gentiles. Since Formation I have changed my view. See John S. Kloppenborg, ‘A “Parting of the Ways” in Q?’ in Q in Context I: The Separation Between the Just and the Unjust in Early Judaism and in the Sayings-Source, ed. Markus Tiwald, Bonner biblische Beiträge (BBB) 172 (Göttingen:  V&R Unipress and Bonn University, 2015), 123–43. 99. Some scholars have suggested that at least Luke 14.5 belongs to Q: Frans Neirynck, ‘Luke 14,1–6:  Lukan Composition and Q Saying’, in Der Treue Gottes trauen:  Beiträge zum Werk des Lukas. FS Gerhard Schneider, ed. Claus Bussmann and Walter Radl (Freiburg i.Br.:  Herder, 1991), 243–63; Christopher M. Tuckett, Q and the History of Early Christianity: Studies on Q (Edinburgh: T&T Clark; Peabody : Hendrickson, 1996), 414–16. 100.  CEQ, 170 included in Q Luke 10.7, ἐσθίοντες καὶ πίνοντες τὰ παρ’ αὐτῶν, which might be interpreted as a rejection of dietary rules, but only if it is read in the context of social practices that engaged non-Jews. In any case, it is certainly not a direct rejection of kashruth.

2. Conceptual Stakes in the Synoptic Problem

41

2.2.3 Why Might This Matter? Solutions to the SP and their historiographical entailments might well be considered to be of theoretical interest only to a handful of scholars of the gospels and to none outside of gospel studies. But I suggest that there are larger but largely unexplored issues that lurk behind SP scholarship. Irrespective of what the historical Jesus was  – whether a prophet, wisdom teacher, apocalypticist, benighted rustic, thaumaturge, or religious fraud  –, the SP provides empirical models for how the ‘story’ or representation of Jesus was received, manipulated and massaged into something bigger and more compelling. That in turn raises the critical question, also seldom explored, of cui bono, that is, whose interests are reflected in the literary representations that are the product of the SP? Early gospel writings are the products of situated intellectual activity, and their reception and promotion or neglect is at least in part due to the resonances that they achieved or failed to achieve. There are some important gestures in the direction of what I  have in mind. Bazzana’s work on the intellectual, cultural and political world of low-level scribes in Egypt provides useful models for thinking about the kind of intellectual production that is seen in Q (if such a thing ever existed).101 Likewise, the examination of school practices, in particular, the anthologizing of wise sayings and their use in philosophical inquiry, have considerable potential for thinking about the intellectual setting and goals of the Gospel of Thomas and perhaps other sayings collections.102 Alex Damm’s work on rhetorical practices and gospel editing at least points towards a larger project of situating the production of early gospel material in social-historical contexts that might be set beside those of Plutarch, Philo, Josephus, Virgil and other intellectual producers, in order to inquire into the interests that are reflected in certain representations of Jesus, and why they succeeded (or failed). 101. Bazzana, Kingdom of Bureaucracy. 102.  On anthological practices in philosophical schools, see John Whittaker, ‘The Value of Indirect Tradition in the Establishing of Greek Philosophical Texts, or the Act of Misquotation’, in Editing Greek and Latin Texts: Papers Given at the Twenty-Third Annual Conference on Editorial Problems, University of Toronto, 6–7 November, 1987, ed. John N. Grant (New York: AMS, 1989), 63–95. Attention to the relation between collections of philosophical sayings and the practice of schools has already been employed for thinking about the social practices reflected in the Gospel of Thomas:  Arthur J. Dewey, ‘ “Keep Speaking Until You Find . . .”:  Thomas and the School of Oral Mimesis’, in Redescribing Christian Origins, ed. Ron Cameron and Merrill P. Miller. Symposium Series (Atlanta, GA:  Society of Biblical Literature, 2004), 109–32; John S. Kloppenborg, ‘Beyond Tinkering and Apologetics’, Annali di Storia dell’Esegesi 25, no. 2 (2008):  23–35, esp. 30–3; Kloppenborg, ‘New Synoptic Problem’, 226–32. It should be mobilized by Francis Watson, both to think about the intellectual production involved in his early ‘sayings collection’ and in his positing of Luke’s collection of Matthew’s sayings, both conceived as collections for the purpose of interpretation.

42

John S. Kloppenborg

For the project of understanding what values are reflected in certain representations of Jesus and why they gained attraction in antiquity, the SP is a critical component. In this enterprise, it really does matter whether Matthew was conflating two predecessor texts, or whether Luke was using Mark and rearranging Matthew or whether Mark was compressing two predecessor gospels. It also matters whether Thomas was merely picking odd bits of sayings from Matthew and Luke, or compiling from various sources an anthology of quasi-philosophical wisdom for use in a ‘school’. Thus, there are some interesting conceptual entailments in the SP in relation to our reconstructions of the terrain of the early Jesus movement. The 2DH, and perhaps Watson’s L/M theory, provided that his ‘sayings collection’ does not simply remain a ‘grab bag’ (to use Meier’s infelicitious term)103 with which Matthew can supplement Mark but has an editorial and therefore conceptual profile, leads to a more variegated picture of the earliest stages of the Jesus movement, and leaves the Gospel of Thomas less of a generic and theological anomaly. Farmer’s presentation of the 2GH, by contrast, asserts a high degree of continuity between the historical Jesus and the practices and dogmatic views of the early Jesus movement and a high degree of uniformity among various streams of primitive Christianity thereafter, but leaves Thomas, and other sayings gospels as anomalies. The point of this chapter is on the one hand to inject a degree of humility into our musing about the SP and to caution against the hybris that announces that certain hypotheses have been ‘discredited’ when in fact the particular complexion of data that is available hardly admits of the language of deductive testing and disproof. Nor does it authorize a free-for-all of groundless speculation. On the other, it is a call to reflect on the conceptual entailments of various SP hypotheses and to recognize that they are not innocent of ideological commitments. The most comfortable of those hypotheses are likely the least innocent.

103.  John P. Meier, A Marginal Jew: Rethinking the Historical Jesus. Volume II: Mentor, Message, and Miracles, Anchor Bible Reference Library (New  York, London and Toronto: Doubleday, 1994), 118.

Chapter 3 S A D S OU R C E S :   O B SE RVAT IO N S F R OM T H E H I ST O RY O F T H E O L O G Y O N T H E O R IG I N S A N D C O N T OU R S O F T H E S Y N O P T IC P R O B L E M Stefan Alkier

3.1 The Witness of the Gospels: Difference and Harmony from Luke to Augustine 3.1.1 From the Gospel to the Gospels The images of human, lion, ox and eagle are not a bad characterization of the four Gospels that made their way into the Christian canon as it emerged in late ancient Christianity amidst conflict and diversity. While the canon did not look exactly the same at all times and places, these four Gospels were a stable element in most manifestations and exerted considerable influence on Christian understanding of Jesus and his message of the Kingdom of God; his suffering, death and resurrection;1 and his exaltation and reign as ‘Lord’ of his people. They substantially shaped Christian narrative communities and imagery, as well as the festive calendar of a diverse range of Christian churches. The first thing one needs to know about evangelists in the early Christian movement, however, is that there were more than four of them. Many other evangelists were preaching the gospel of Jesus Christ at the same time that the four Gospels were composed, and evangelists continued to work afterwards, some of whom produced texts that would eventually be labelled ‘Gospel’. The Gospels that later became canonical were grouped together at an early date, and they formed the basis of Tatian’s Diatessaron, but this did not bring the production of Gospels to an end. By the third century CE, several dozen Gospels were circulating, and production did not stop even after widespread stabilization

1.  On the importance of the resurrection for the canonical Gospels, see Stefan Alkier, Die Realität der Auferweckung in, nach und mit den Schriften des Neuen Testaments, Neutestamentliche Entwürfe zur Theologie (NET) 12 (Tübingen: A. Francke-Verlag, 2010); English edition:  The Reality of the Resurrection. The New Testament Witness, Foreword by Richard B. Hays, trans. Leroy A. Huizenga (Waco, TX: Baylor University Press, 2013).

44

Stefan Alkier

of the contours of the catholic canon in the fourth and fifth centuries CE. Gospels continued to be written in the High Middle Ages, some of which influenced the development of Christian piety, practices and art significantly. These include the Protoevangelium of James, which began circulating no later than the third century and influenced the development of Mariology, and the so-called Gospel of PseudoMatthew from the eighth or ninth century, which contains the earliest known depiction of the baby Jesus in the manger. ‘The animals, the ox and the ass, with him in their midst, incessantly adored him.’2 Yet none of these other Gospels were able to push aside or replace the Human (Matthew), Ox (Luke), Lion (Mark) and Eagle (John). Nor was the term ‘evangelist’ reserved for the authors of texts. At the most basic level, an evangelist (εὐαγγελιστής) was simply a ‘bringer of good tidings’, someone who brought ‘good news’ – that is, a ‘gospel’ (Greek εὐαγγέλιον). Good news might be the announcement that a battle had been won, that a peace treaty had been signed, or that an eagerly anticipated child had been born. In the book of Isaiah, the return from Exile – understood as God’s work in bringing salvation – is celebrated as ‘good news’ (see Isa 40.1–2; 52.7; 61.1). The messengers who brought this ‘good news’ of freedom and salvation were considered prophets. A few centuries later, the ‘good news’ consisted of the announcement that the Roman emperor Augustus was the ‘savior’ (σωτήρ) who had brought peace after a century of bloody civil war.3 Narratives about Jesus of Nazareth that would later be called ‘Gospels’ joined this tradition of reporting ‘good news’. Their ‘good news’ included not only the resurrection of Jesus but also his preaching activity – his message was portrayed as consisting of the ‘good news’ that God’s reign was breaking in (see Mark 1.14–15). The Gospels thus depict Jesus’ preaching role as being in continuity with that of earlier prophets. The apostle Paul, who wrote before the production of extant Gospel texts, did not use the term ‘gospel’ to refer to the preaching activity of the human Jesus of Nazareth, but to his resurrection, through which God had established justice and salvation. Paul’s core letters (Rom, 1&2 Cor, Gal, 1 Thess, Phil, Phlm), which were written in the 50s CE, are probably the oldest extant texts of the emerging Christian movement, and understand the ‘gospel’ to consist of the good news that the crucified Jesus has been resurrected. This good news had the power to save those who were captivated by the spirit of the message (see Rom 1.16–17).

2.  Gospel of Pseudo-Matthew 14, translated in J. K. Elliott, The Apocryphal New Testament: A Collection of Apocryphal Christian Literature in an English Translation Based on M.R. James (Oxford: Clarendon Press, 1993), 94. 3. See Stefan Alkier, ‘Leben in qualifizierter Zeit: Die präsentische Eschatologie des Evangeliums vom römischen Novum Saeculum und die apokalyptische Eschatologie des Evangeliums vom auferweckten Gekreuzigten’, Zeitschrift für Neues Testament (ZNT) 22 (2008): 20–34.

3. Sad Sources

45

The term ‘evangelist’ is also applied to Philip in Acts 21.8 and to Timothy in 2 Tim 4.5, and ‘evangelists’ are listed alongside apostles, prophets, pastors and teachers in Ephesians 4.11. The text does not tell us what their specific task was, but one imagines, based on references to ‘good news’ in other contexts, that they communicated the good news about Jesus’ resurrection in narrative or imagistic forms (cf. Gal 3.1b). Their diverse voices consolidated and kept alive this amazing news. The term ‘evangelist’ continued to be used in this more general sense even after the authors of the four most influential Gospels began to be referred to as ‘evangelists’. Eusebius of Caesarea (before 265–339/340) wrote the following in his Ecclesiastical History, in a passage concerning the second century: Very many of the disciples of the time, their hearts smitten by the Word of God with an ardent passion for the true philosophy, first fulfilled the Saviour’s command by distributing their possessions among the needy; then, leaving their homes behind, they carried out the work of evangelists, ambitious to preach to those who had never yet heard the message of the faith and to give them the inspired gospels in writing.4

The term ‘evangelist’ could thus apply to anyone who either read out or recounted the good news of the resurrection of the crucified Jesus of Nazareth. The proclamation of the gospel thus began as both oral and written with a polyphony of evangelists. 3.1.2 Legends about the Four Evangelists in the Early Church It was in the second century CE that the term ‘evangelist’ began to be used to refer to the writers of the four texts that would later become canonical. The men whose names were eventually attached to these texts had not been referred to that way in any of the other texts that would become part of the New Testament. They received the designation in the second century as a shorthand and tribute for their work in putting together the Gospels. The Evangelists were thus created by the Gospels rather than the other way around. This probably explains why there was so little interest in writing down their biographies in the early decades of the Christian movement  – and this lack of interest unfortunately means that we now know almost nothing for certain about their lives. Remarks about the Evangelists in the earliest sources are universally short and give the impression that, by the second century, no one really knew who they had been or when and where they had lived, worked and died. Legends and adventurous conclusions addressed this lack of knowledge, but there was no independent interest in the lives of the Evangelists in the early

4.  Eusebius of Caesarea, H.E. III.37.2. Translated in Eusebius, The History of the Church from Christ to Constantine, trans. G. A. Williamson, rev. ed. Andrew Louth (London: Penguin, 1989), 100.

46

Stefan Alkier

Church. No historical biographies were written about Matthew, John, Mark or Luke, but only fictional legends like the diverse acts of some apostles, even though biography was a popular type of book in the ancient world.5 The Gospels seem to have spoken for themselves, and the poetic attribution of symbols from the book of Revelation – the lion, eagle, ox and human – may have been more meaningful than biographical information about their authors would have been (see Rev 4.7, which draws intertextually on Ezek 1.10 and Isa 6.2). The symbols were probably first attributed to the Evangelists by Irenaeus,6 and the conventional allocations are first attested in the writings of Jerome.7 The earliest legends about the authors of the Gospels also show an interest in portraying them as authoritative witnesses as a means of emphasizing the historical reliability of their narratives. This is well illustrated by a passage from Eusebius’ Ecclesiastical History where he reports and comments on some remarks made by Papias, bishop of Hierapolis, around 130 CE: This, too, the presbyter used to say. ‘Mark, who had been Peter’s interpreter, wrote down carefully, but not in order, all that he remembered of the Lord’s sayings and doings. For he had not heard the Lord or been one of his followers, but later, as I said, one of Peter’s. Peter used to adapt his teachings to the occasion, without making a systematic arrangement of the Lord’s sayings, so that Mark was quite justified in writing down some things just as he remembered them. For he had one purpose only – to leave out nothing that he had heard, and to make no misstatement about it.8

Nothing is said in this passage about Mark’s life – Papias is only interested in connecting the Gospel of Mark with the authority of the apostle Peter. Since Mark had not been an eyewitness of the events about which he wrote, Peter becomes a guarantor of the Gospel’s reliability. Eusebius’ commentary suggests that Papias was the first to make such a connection. ‘Papias also makes use of evidence drawn from 1 John and 1 Peter.’9 Papias may have been thinking of 1 Peter 5.13: ‘Your sister church in Babylon, chosen together with you, sends you greetings; and so does my son Mark’ (NRSV). Papias may have assumed that this ‘Mark’ of 1 Peter – who is otherwise unknown and probably an invented character – was the Evangelist. Needless to say, this ice is far too thin to carry any historical weight.

5.  Cf. Dirk Frickenschmidt, Evangelium als Biographie. Die vier Evangelien im Rahmen antiker Erzählkunst, Texte und Arbeiten zum neutestamentlichen Zeitalter (TANZ) 22 (Tübingen and Basel, 1997), esp. 153–90. 6.  Irenaeus, Adv. Haer. III 11.8. Irenaeus allocates the symbols as follows:  human (Matthew), eagle (Mark), ox (Luke), lion (John). 7.  Jerome, Commentarii in Mattheum, praef. 3.  He allocates the symbols as follows: human (Matthew), lion (Mark), ox (Luke), eagle (John). 8. Eusebius of Caesarea, H.E. III. 39.15. 9. Ibid., III. 39.17.

3. Sad Sources

47

Another interesting aspect of Papias’ remarks is his evident desire to protect the Gospel of Mark from the criticism that he had not recorded events in their proper order, an accusation that had already been lobbed by Luke against his predecessors in the prologue to his Gospel (Luke 1.1–4). Papias acknowledged the validity of this characterization of the Gospel of Mark, but emphasizes that what Mark had written is nevertheless true. Papias’ comments are thus apologetic in nature and therefore not a reliable source of historical information from the perspective of contemporary historical method and hermeneutics. On the other hand, they clearly show that the Gospels were not yet understood as ‘Holy Scripture’ in the second century. Instead, they were ‘memory’, and were authoritative to the extent that they preserved eyewitness testimony either directly (Matthew and John) or indirectly (Mark and Luke)  – a claim that was apparently still controversial in some cases. (Admittedly, some contemporary exegetes think that Papias did not actually have a copy of the Gospel of Mark in front of him, but either an earlier form of the tradition or one of Mark’s sources.) Papias says even less about Matthew:  ‘Matthew compiled the Sayings in the Aramaic language, and everyone translated them as well as he could.’10 This remark refers to a collection of sayings – like the Gospel of Thomas – rather than a narrative like the Gospel of Matthew. It is even less clear than in the case of Mark that Papias is referring to the Gospel of Matthew as we know it. In any case, nothing is said about Matthew’s life. Luke and his works are not even mentioned in extant fragments of Papias, nor is much attention paid to the Gospel of John. Eusebius’ report on Papias only mentions that the author of John’s Gospel was not the same as the author of Revelation, and that both had lived in Ephesus and were buried there.11 Papias, who lived in Hierapolis in Asia Minor, claimed to have gotten his information from a living source, the Ephesian presbyter John. All other known references to the Evangelists are more recent than Papias, and just as biased. Irenaeus’ comments about Mark and Matthew are based on Papias’ account, and he writes about Luke, ‘Luke, the companion of Paul, recorded the Gospel that he [i.e., Paul] preached in a book.’12 Irenaeus presupposes the chronological order Matthew – Mark – Luke, and associates Luke with Paul, just as Mark was associated with Peter. This matchmaking is clearly designed to give apostolic value to works whose authors were not considered apostles themselves. The unknown author of the Muratorian Fragment, which dates to the early third century, claims to know still more about Luke: The third book of the Gospel is that according to Luke. Luke, the well-known physician, after the ascension of Christ, when Paul had taken with him as one zealous for the law, composed it in his own name, according to [the general] 10. Ibid. III. 39.16. 11. Ibid. III. 39, 5f. 12. Irenaeus, Adv. Haer. III 1.1 (translation J. Snyder). Origen takes up this tradition in Eusebius, H.E. VI. 25.3–6.

48

Stefan Alkier belief. Yet he himself had not seen the Lord in the flesh; and therefore, as he was able to ascertain events, so indeed he begins to tell the story from the birth of John.13

Once again, the limitations faced by human authors – ‘as he was able to ascertain’ – are simply accepted. The same fragment contains a legend about the composition of the Gospel of John: The fourth of the Gospels is that of John, [one] of the disciples. To his fellow disciples and bishops, who had been urging him [to write], he said, ‘Fast with me from today for three days, and what will be revealed to each one let us tell it to one another.’ In the same night it was revealed to Andrew, [one] of the apostles, that John should write down all things in his own name while all of them should review it. And so, though various elements may be taught in the individual books of the Gospels, nevertheless this makes no difference to the faith of believers, since by the one sovereign Spirit all things have been declared in all [the Gospels].

All that we learn about the author of the fourth Gospel in this passage is his name and that he was a disciple of Jesus, but the legend of how the Gospel came into being is nevertheless interesting. John did not come up with the plan to write it on his own; he was simply acceding to the request of his fellow disciples and the bishops and the explicit call of the Spirit. The Gospel is thus not depicted as the product of a single human author, but as the work of a collective inspired by the Spirit. The work was carried out by John, but was not really the product of an ‘author’. This passage once again supports the overall conclusion that we do not know anything for certain about the lives of the Evangelists. At the same time, it illustrates an interesting dynamic in early Christian thinking about the Gospels: early Christians invoked the eyewitness status of Matthew and John and relationships between Mark and Peter, Luke and Paul as a means of emphasizing the reliability of the Gospels, without advocating a hermeneutic of authorship. One did not need to know anything about the lives of the Evangelists to understand the Gospels. This hermeneutical perspective was probably the primary factor that led to lack of interest in the Evangelists: ignorance of the authors’ biographies did not diminish the Gospels’ power. With regard to the question of the Gospels’ origins, there is no hint in the authors surveyed above that the four Gospels are dependent on one another literarily. They are understood to be four separate Gospels composed independently. Similarities between them are attributed to the fact that they reflect eyewitness testimony about events that had actually taken place. Differences are traced to the

13.  Translation taken from Bruce M. Metzger, The Canon of the New Testament:  Its Origin, Development, and Significance (Oxford: Clarendon, 1987), 305–7.

3. Sad Sources

49

differing perspectives of the human authors. The four Gospels do not contradict one another, but complement each other. The Human, Lion, Ox and Eagle are powerful witnesses ‘of the events that have been fulfilled among us’ (Luke 1.1). 3.1.3 Strategies to Counteract Confusion That Might be Caused by the Polyphony of the Gospels 3.1.3.1 Luke, Marcion and Tatian In the early Church, the question of the emergence of the Gospels was driven more by apologetic than hermeneutical concern. The main question was whether differences between the Gospels represented contradictions that called their trustworthiness into question, or whether their polyphony could be considered a sort of harmonious diversity.14 The first voice suggesting that differences between gospel narratives might cause confusion was that of the historian Luke, who wrote in the style of ancient historiography in the prologue to his Gospel: Since many have undertaken to set down an orderly account of the events that have been fulfilled among us, just as they were handed on to us by those who from the beginning were eyewitnesses and servants of the word, I too decided, after investigating everything carefully from the very first, to write an orderly account for you, most excellent Theophilus, so that you may know the truth concerning the things about which you have been instructed. (Luke 1.1–4)

Luke’s strategy to ward off potential confusion that might arise from the existence of a variety of narratives about events was not to look for harmony among these voices, but to employ them critically in order to produce his own trustworthy account. Luke was trying to make a singular out of a plural. He did not succeed, however, as one can see from the fact that even more gospels were produced after him than before. It is impossible to know exactly how Luke worked with his sources – whether he had internalized their contents and wrote his own Gospel without referring back to them, or whether he sometimes copied bits of text or even borrowed the outline of his Gospel from a preexisting source. There is no indisputable proof that Luke copied material directly from written documents, as has often been suggested by those trying to explain similarities between Luke and Mark and between Luke and Matthew (or Q, sometimes known as the ‘sayings source’). The most prevalent current hypotheses about the composition of the Gospels, one of which involves only the three Synoptics (Mt, Mk, Lk) and the other of which posits at least one additional source (Q), are at the end of the day nothing more than hypothetical

14.  See Stefan Alkier, ‘Unerhörte Stimmen  – Bachtins Konzept der Dialogizität als Interpretationsmodell biblischer Polyphonie’, in Kleine Schriften des Fachbereichs Evangelische Theologie vol. 3  – Wahrheit und Positionalität, ed. Melanie Köhlmoos and Markus Wriedt (Leipzig: Evangelische Verlagsanstalt, 2012), 45–70.

50

Stefan Alkier

models. Although most contemporary exegetes find the idea of Markan priority plausible, it is impossible to prove either this hypothesis or the existence of Q. These are not ‘facts’, but model-driven hypotheses. Before the advent of the ‘Griesbach Hypothesis’, Anton Friedrich Büsching had suggested that Mark drew on Matthew and Luke, and that the Gospel of Luke was written first  – a difference from Griesbach.15 If one adopts Büsching’s model of Lukan priority, it is impossible to verify whether Luke was actually citing written texts at any particular point. It can still be said with certainty based on Luke 1:1–4, however, that Luke knew of narratives that differed from one another, and that this fact bothered him. He was impelled to step in and create certainty by composing his own consistent, carefully researched account. Marcion adopted a different approach. He selected one narrative from among the available options – the Gospel of Luke – and treated it as the only valid Gospel narrative. In a certain sense, he and Luke shared the same aspiration: to tell the single, tested, trustworthy story. Marcion’s decision-making was not limited to the selection of a single Gospel, however. He also altered Luke’s Gospel to suit his own theological agenda. A third strategy was employed by the Syrian Tatian (110–185 CE), who created a ‘Gospel harmony’ out of several preexisting narratives. He drew primarily on Matthew, Luke, Mark and John, although episodes from other Gospels also appear in his Diatessaron, which was to become the primary Gospel text used in the Syrian church.16 These three strategies  – production of a new text, selection among preexisting texts and harmonization – all address the issue of the polyphony of Gospel narratives in the early centuries, which was evidently considered problematic in some quarters. None of them resulted in widespread adoption of a single, definitive Gospel narrative universally acknowledged as the ultimate locus of truth, however. Luke may have been able to convince Marcion that his narrative was good enough, but he was not able to stop the production of additional Gospels in the future. Marcion’s choice of a single Gospel, which he attached to a number of Pauline letters, led to the establishment of an idiosyncratic form of Christian community that ended up being rejected by the majority of catholic congregations. And while Tatian’s Diatessaron was a great success in Syria, it contributed to a distancing between Syrian congregations and other parts of the Church. Thus a quest for clarity and certainty, motivated by a desire to unify Christians, ended up producing an even more diverse range of Christian communities, many of which looked on one another with suspicion. Such a movement was easy prey for those who considered Christian teaching to be a pernicious superstition. 15.  See Anton Friedrich Büsching, Die vier Evangelisten, mit ihren eigenen Worten zusammengesetzt, auch mit hinlänglichen Erklärungen versehen (Hamburg: Ritter, 1766). 16.  See A. Augustus Hobson, The Diatessaron of Tatian and the Synoptic Problem (Chicago: Chicago University Press, 1904); Helmut Koester, Ancient Christian Gospels: Their History and Development (Philadelphia:  Trinity Press International; London:  SCM Press, 1990), 422–8.

3. Sad Sources

51

3.1.3.2 Origen vs. Celsus As Christ-following communities began to spring up around the Roman Empire in the first century, increasing in number in the second and third centuries, a number of texts were produced that polemicized against Christian ideas, dismissing them as incoherent superstition. As Pliny’s famous letter to Emperor Trajan demonstrates, not all early opponents of this ‘superstition’ made an effort to study the texts produced by its adherents, often preferring to gather their information orally or through torture.17 As early as the second century, however, a more scholarly – although still polemically motivated – engagement with Christian texts began. The extant writings of Justin show that intellectuals were debating the pros and cons of Christian teaching, partly with reference to the nature of Christian texts. Justin countered accusations that the Gospels of Matthew, Luke and Mark are contradictory by employing a strategy of harmonization. His ideas inspired Tatian to compose his Diatessaron; the latter probably worked independently and employed more Gospels than Justin would have liked. One particular work written to demonstrate the irrational nature of Christian teaching was Celsus’ True Doctrine. He highlighted contradictions between Christian texts and pointed out that some of them circulated in multiple different versions. His text critical and literary critical arguments reveal intense study of Christian texts, and represented an important challenge for Christian intellectuals. It was Origen who found the intellectual energy to compose a detailed refutation of Celsus’ arguments.18 In the process, he set forth a workable hermeneutic for the interpretation of texts that continues to be employed in Orthodox churches today, and which contributed significantly to the idea of the ‘multiple senses of Scripture’ that is still taught in the Roman Catholic Church. Origen countered Celsus’ polemical arguments not by denying that there were differences and even contradictions between Christian texts, but by seeking to reduce complexity through emphasis on considerations of a methodological, hermeneutical and normative character. Rather than forcing the plural to become a singular, Origen argued in favour of a skilled plural, rendered manageable through a hermeneutic of the multiple senses of Scripture. Origen employed text criticism, compared texts, reflected hermeneutically and thought about the theology of the canon. He attributed normative status to just four of the Gospels that were circulating at the time. At the beginning of his first Homily on the Gospel of Luke, he comments, ‘The ekklesia has four gospels, the heretics many’.19 His explicit argument for these four Gospels in particular is based on tradition concerning their composition and

17. Pliny, Epistulae X, 96.9. 18. Cf. David Laird Dungan, A History of the Synoptic Problem. The Canon, the Text, the Composition, and the Interpretation of the Gospels, The Anchor Yale Bible Reference Library (New Haven and London: Yale University Press, 2009), 65–88. 19. Origenes Werke 9. Die Homilien zu Lukas, ed. M. Rauer, GCS (Leipzig: Hinrichs, 1930), 4 (my translation).

52

Stefan Alkier

use: ‘I have learned by tradition that the first was written by Matthew, the second was (written) by Mark who composed it according to the instructions of Peter, . . . and the third (was written) by Luke (and is) the Gospel commended by Paul, (and) last of all (is) that (one) by John.’20 Since these four Gospels were circulating in different versions, Origen also employed text critical methods to reduce uncertainty about the wording of the text.21 Origen’s dedicated study of the textual details of the Gospels did not lead him to try to harmonize their referential content, but to posit univocality at the level of authorship: the guarantor of the Gospels’ truth was the Holy Spirit, who had himself introduced contradictions between the Gospels on the level of historical reference and could use them for didactic purposes  – that is, for the education of the attentive reader. Thus, rather than employing a harmonizing strategy that would paper over differences between the Gospels, Origen argued that they should be acknowledged and respected. His main point was that anything that resulted in an obvious contradiction on the level of historical referentiality should be interpreted symbolically: ‘Whoever thinks the writing of these four (Gospels) is (just literal) history, or a representation of real things through historical images, and who supposes God to (literally) be within certain limits in space, and to be unable to present to several persons in different places several versions of Himself at the same time . . . (will find) it impossible (to maintain) that our four writers are writing the truth.’22 3.1.3.3 Augustine vs. Porphyry Although Celsus’ polemic against Christian teaching, texts and practices is well known, he was not the only writer to set the writings of Christ-followers in his sights. Fragments of a tractate Against the Christians by Porphyry, now readily accessible in an excellent new translation and commentary by Matthias Becker, show that Porphyry was well-versed in texts from the Old and New Testaments. He must have studied them carefully. As Dungan remarks, ‘Out of this painstaking study emerged page after page of embarrassing questions pinpointing hundreds of inconsistencies, absurdities, and morally objectionable episodes. He drew heavily on earlier critiques of the Bible, both Christian (especially Marcionite and Gnostic) and non-Christian (Jewish, Neoplatonic [Celsus] and Manichean), thus distilling the concentrated venom directed against the Bible over the previous two hundred years. His attack was especially dangerous in that his Neoplatonic orientation was similar to that of many Christian theologians of his day; his objections were often one they secretly shared.’23

20.  Origen is quoted in Eusebius, H.E. VI.25.4–6. Translation from Dungan, A History of the Synoptic Problem, 70. 21. Cf. Origenes Werke, Matthäuserklärung, ed. Erich Klostermann, GCS (Leipzig: Hinrichs, 1935), 10:387f. 22. Origen, Commentary on John 10.4, ANF 10:383. Translation from Dungan, A History of the Synoptic Problem, 85. 23. Dungan, A History of the Synoptic Problem, 91.

3. Sad Sources

53

Eusebius of Caesarea responded to Porphyry first in ‘The Preparation for the Gospel’, and then in more detail in his ‘Demonstration of the Gospel’. Even his well-known Ecclesiastical History, one purpose of which was to demonstrate the historical reliability of the Christian tradition, was partly directed against Porphyry. Apologetically motivated, Eusebius also drew on his own comparative study of the four Gospels to provide an overview of correspondences between them, which probably served as a tool for Augustine in composing his Gospel harmony.24 Augustine followed in the footsteps of Origen, Papias, Irenaeus and others, acknowledging the plurality of the four Gospels and accepting them as texts in their own right that were rendered authoritative by the Holy Spirit’s divine oversight of their composition and their recourse to eyewitness testimony. Augustine took a different approach than Origen, however. Rather than simply accepting differences between the Gospels, he sought to resolve them by differentiating between ‘narrative’ and ‘history’, distinguishing between the ‘history’ that the Evangelists reported and the manner in which they combined various episodes in their ‘narratives’. According to Augustine, the Evangelists had not attempted to follow the historical order of events when composing their accounts: ‘These . . . had one order determined among them with regard to the matters of their personal knowledge and their preaching [of the gospel] but a different order in reference to the task of giving the written narrative.’25 Augustine attributed differences in the way events were recounted in the various Gospels to differences between divine and human language. His reflections in the areas of semiotics and philosophy of language had led him to the insight that the semiotic requirements of human language also applied to the Bible, and he was convinced that the text of the Bible should not be confused with the Word of God. Augustine’s comparative study of the Gospels also led him to an astonishingly innovative explanation for word-for-word overlap between them, which broke from the tradition of Papias, Irenaeus, Origen and Eusebius. Augustine was the first to posit literary dependence among the Gospels; he was thus the exegetical founder of the sort of thinking that underpins contemporary approaches to their origins. At first, he merely suggested that the extensive overlap in wording between Mark and Matthew showed that Mark must have used Matthew as a Vorlage, which he had shortened.26 The hypothesis that Mark epitomized the Gospel of Matthew is still traced to Augustine, but it is often overlooked that Augustine’s own thinking on the topic continued to develop – he later decided

24. Eusebius’ letter is included in Nestle–Aland, Novum Testamentum Graece, 28th edn, 47–52. 25. St. Augustine: The Harmony of the Gospels, translated by Rev. S. D. F. Salmond, D. D., ed., with Notes and Introduction by Rev. M. B. Riddle, D. D., Nicene and Post-Nicene Fathers Series I, vol. 6 (Edinburgh: T&T Clark, 1887), 1.3 (132). 26. See Harmony of the Gospels, 1.4. (132).

54

Stefan Alkier

that Mark probably had copies of both Matthew and Luke in front of him as he worked. Augustine comments: John remains, between whom and others there is left no comparison to be instituted. For, however the evangelists may each have reported some matters which are not recorded by the others, it will be hard to prove that any question involving real discrepancy arises out of these. Thus, too, it is a clearly admitted position that the first three – namely, Matthew, Mark, and Luke – have occupied themselves with the humanity of the Lord Jesus Christ, according to which He is both king and priest. And in this way, Mark, who seems to answer to the figure of the man in the well-known mystical symbol of the four living creatures, either appears to be preferentially the companion of Matthew, as he narrates a larger number of matters in unison with him than with the rest, and therein acts in due harmony with the idea of the kingly character whose wont it is, as I have stated in the first book, to be not unaccompanied by attendants; or else, in accordance with the more probable account of the matter, he holds a course in conjunction with both [Matthew and Luke]. For although he is at one with Matthew in the larger numbers of passages, he is nevertheless at one rather with Luke in some others. And this very fact shows him to stand related at once to the lion and to the steer, that is to say, to the kingly office which Matthew emphasizes, and to the sacerdotal which Luke introduces, wherein also Christ appears distinctively as man, as the figure of which Mark sustains stands related to both these.27

The hypothesis later articulated by Henry Owen and Johann Jakob Griesbach – that Mark had used the Gospel of Matthew as a Vorlage, adding material from the Gospel of Luke – had thus already been formulated by Augustine. Augustine has not been remembered for his modified position, however, but for the first version of his hypothesis, in which Mark was only dependent on Matthew. This is especially surprising when one considers the incredible success of Augustine’s Gospel harmony, which was so influential on this genre of exegetical apology – despite a proliferation of works of this type28 – that Martin Luther was still drawing on 27.  Harmony of the Gospels, 4.10.11 (400f.). Emphasis Alkier. The passage continues, “On the other hand, Christ’s divinity, in virtue of which He is equal to the Father, in accordance with which He is the Word, and God with God, and the Word that was made flesh in order to dwell among us, in accordance with which also He and the Father are one, has been taken specially in hand by John with a view to its recommendation to our minds. Like an eagle, he abides among Christ’s sayings of the sublime order, and in no way descends to earth but on rare occasions.” 28.  See Christoph Burger, August Den Hollander, Ulrich Schmid (eds), Evangelienharmonien des Mittelalters, Studies in Theology and Religion (STAR) 9 (Assen: Van Gorcum 2004); Dietrich Wünsch, Evangelienharmonien im Reformationszeitalter. Ein Beitrag zur Geschichte der Leben-Jesu-Darstellungen, Arbeiten zur Kirchengeschichte (AzK) 52 (Berlin, New  York:  de Gruyter, 1983); Beate Köster, ‘Evangelienharmonien im frühen Pietismus’, Zeitschrift für Kirchengeschichte (ZKG) 103 (1992):  195–225; Büsching, Die vier Evangelisten; Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI and Cambridge, UK: Eerdmans, 2013), 13–61.

3. Sad Sources

55

Augustine when he discounted any need to worry about the relationship between the Gospels: ‘The evangelists do not all observe the same chronological order. The one may place an event at an earlier, the other at a later time . . . Be that as it may, whether it happened sooner or later, whether it happened once or twice, this will not prejudice our faith.’29

3.2. From the Perspicuity (Klarheit) of Scripture to the Synoptic Problem 3.2.1 From Gospel Harmonies to the Perspicuity of Scripture Differences between the Gospels were not generally considered to be a problem during the Reformation.30 The traditional symbolism of human, lion, ox and eagle was a persuasive and poetic image that sufficiently accounted for the fact that John is a different sort of text than Matthew, Mark and Luke, and that the latter resemble one another. The canon was so well established that there was little interest in the polyphony of the many Gospels that circulated in early Christianity; generally, people only listened to the talented voices of the canonical Four. Yet it was also no problem to place works such as the Protoevangelium of James, which played a role in the cult of the saints, alongside the Quartet in order to expand knowledge about traditional figures. The Church Fathers’ disagreements about the origins of the Gospels had already stopped causing anxiety by the Middle Ages, a time when Gospel 29. Martin Luther, Sermons on the Gospel of St. John Chapters 1–4, in Luther’s Works, ed. Jaroslav Pelikan, vol. 22 (Saint Louis: Concordia, 1957), 218. 30. Debates still went on about the authorship of particular texts and even of individual pericopes, in the tradition of debates about the canon in the Middle Ages, which had primarily been concerned with providing formal justification for the authority of texts that were or should be considered canonical. Without these debates, conversations about the canon during the Reformation and Luther’s establishing of an entirely new design for the Biblical canon would have been impossible. Karlstadt had already argued that the ending of the Gospel of Mark (Mark 16:9–20) is secondary, and had declared it not to be canonical. See Bernhard Lohse, ‘Die Entscheidung der lutherischen Reformation über den Umfang des alttestamentlichen Kanons’, in Verbindliches Zeugnis I. Kanon – Schrift – Tradition, ed. W. Pannenberg and T. Schneider, Dialog der Kirchen 7 (Freiburg i. Br. and Göttingen: Vandenhoeck & Ruprecht, 1992), 182. There was general agreement, however, that everything that was considered canonical was a harmonious source of information about God’s revelation. This is why Luther did not want to include texts that he thought destroyed this harmony (Hebrews, James, Jude and Revelation) in the canon. See Stefan Alkier, ‘Der christliche Kanon als Quelle der Offenbarung. Theologiegeschichtliche Anmerkungen zu einem aktuellen Thema’, in Relationen – Studien zum Übergang vom Spätmittelalter zur Reformation. Festschrift Karl Heinz zur Mühlen, ed. Athina Lexutt and Wolfgang Matz, Arbeiten zur Historischen und Systematischen Theologie (AHSTh) 1 (Münster: LIT Verlag, 2000), 115–38.

56

Stefan Alkier

harmonies proliferated.31 As already remarked, the early line of tradition represented by Papias, Irenaeus, Origen and Eusebius understood the four Gospels to be literarily independent of one another. Matthew and John had been direct eyewitnesses of events, and Luke and Mark mediated the eyewitness testimony of others. Commonalities between the Gospels were attributed to the fact that they all referred back to historical events and either directly or indirectly to the teaching of Jesus; the Holy Spirit had also worked with the authors in the composition process. Differences were attributed to the authors’ individual perspectives and human limitations. While a new approach was initiated by Augustine, who was the first to formulate a hypothesis about the use of one Gospel by the author of another, the earlier approach continued to influence later thinking. One of the reasons Augustine – and Griesbach – thought Matthew had written his Gospel before Mark was that Matthew was considered to be an apostle who had witnessed events himself, whereas Mark had only indirectly transmitted eyewitness testimony. Augustine’s conviction that John had written without knowledge of the other texts also allowed Matthew and John to serve as independent apostolic witnesses of the events recorded in the Gospels, thus undergirding their historical reliability. Both approaches also reflect the common conviction that differences between the four Gospels do not call their truth into question. At the end of the day, the four Gospels speak with the same voice, that of the Son, the Holy Spirit and the Father. Luther’s ideas about the perspicuity of Scripture, a fundamental building block of his radical sola scriptura, would not have been possible without this long apologetic tradition that sought to discern a single voice in the polyphony of the Gospels. 3.2.2 Ad Fontes It is not possible within the confines of the current chapter to recount the entire history of Scriptural interpretation in the Reformation, but a few episodes will now be mentioned that are important to keep in mind, and which have had a significant indirect influence on interpretation of the Gospels.32 31.  Admittedly, detailed studies of Gospel harmonies  – a desideratum for future research – might change this understanding, since they incorporate many interesting exegetical ideas. See the works cited in n. 28 above. An ironic side effect of Historical Criticism has been a widespread ignorance of the history of scholarship. Scholarship from the period before the advent of Historical Criticism is hardly read today. There are also distressing parallels in the field of Philosophy, where philosophers before Kant seem to attract little interest. I hope that these guilds will soon realize what rich treasures await them in the exegetical and philosophical works of the Middle Ages and Early Modernity. 32. See, e.g. Albrecht Beutel, Protestantische Konkretionen. Studien zur Kirchengeschichte (Tübingen: Mohr Siebeck, 1998); Lexutt and Matz (eds), Relationen – Studien zum Übergang vom Spätmittelalter zur Reformation; Volker Leppin (ed.), Reformatorische Theologie und Autoritäten. Studien zur Genese des Schriftprinzips beim jungen Luther (Tübingen:  Mohr Siebeck, 2015).

3. Sad Sources

57

In his book Der Primat der Heiligen Schrift als theologisches und kanonisches Problem im Spätmittelalter, Hermann Schüssler argues persuasively that while the authority of Scripture was not questioned in the Late Middle Ages, the relationship between the authority of the Church and that of Scripture was hotly debated.33 Even those who gave precedence to Scripture in the question of authority did not advocate a radical sola scriptura, but merely a relative primacy of Scripture, reflecting the fact that at least since the time of Augustine, Scripture had been understood as not always being clear and understandable. For the interpretation of difficult passages, Augustine recommended invoking the tradition and authority of the Church. Influenced by this principle of ‘Scripture and tradition’ and the proliferation of Gospel harmonies, many people assumed that there was an essential harmony between Biblical texts and the writings of the Church Fathers. Sometimes the writings of Fathers – those that were acknowledged to be authorities – were even subsumed under the title ‘scriptura’.34 Luther’s slogan, sola scriptura, was initially directed against this harmonizing intermixing of Bible and interpretation, the origins of which he attributed to the idea that the Bible was – at least partially – unclear. In Assertio omnium articulorum,35 which appeared in 1520, Luther argued vehemently for the absolute primacy of the Holy Scriptures, positing a fundamental difference between Scripture and interpretation and thus radically rejecting the practice of extrapolating from Gospel harmonies to the relationship between the Bible and Church tradition. Luther argued that the Bible should be thought of as clear and reliable; its meaning should – and could – be understood by anyone who engaged in serious philological and intertextual studies. He writes, ‘Oportet enim scriptura iudice hic sententiam ferre, quod fieri non potest, nisi scripturae dederimus principem locum, in omnibus quae tribuuntur patribus, hoc est, ut sit ipsa per sese certissima, facillima, apertissima, sui ipsius interpres, omnium omnia probans, iudicans et illumans . . .’36 Whoever even thought about calling the Scriptures ‘dark’ opened the floodgates to human whim. Luther did not mean that the Bible was always easy to understand. On the contrary, its ‘literal’ meaning could only be understood after concentrated philological and intertextual studies. Luther’s sola scriptura was thus a call to methodological instruction in Biblical interpretation, which began with the text and always took Biblical intertexts into account.37

33.  Hermann Schüssler, Der Primat der Heiligen Schrift als Theologisches und kanonistisches Problem im Spätmittelalter, Veröffentlichungen des Instituts für Europäische Geschichte Mainz 86 (Wiesbaden: Phillip von Zabern, 1977). 34.  Schüssler, Der Primat der Heiligen Schrift, 3. 35.  Martin Luther, Assertio omnium articulorum Martini Lutheri per bullam Leonis X. novissimam damnatorum, Weimar edition (hereafter WA) 7, 94–151. 36. Luther, Assertio, WA 7, 97. 37.  Luther, Assertio, WA 7, 97:  Primum sepositis omnium hominum scriptis, tanto magis et pertinacius insudandum erat solis sacris, quo praesentius periculum est, ne quis proprio spiritu eas intelligat, ut usus assidui studii victo periculo eiusmodi, tandem certum nobis faceret spiritum scripturae, qui nisi in scriptura prorsus non invenitur.

58

Stefan Alkier

Luther not only considered the Bible to be clear and understandable but also thought that all rational people could interpret the Bible correctly if they learned to understand its grammar and to read individual sentences in their intra- and intertextual contexts. This is what the interpretative principle ‘scriptura sui ipsius interpres’ actually entails. In order to understand the importance of the Reformers’ doctrine of Scripture within the history of Biblical interpretation, one has to go back to the Renaissance, which began in Italy in the fifteenth century. The multifaceted movement involved not just a turn to ancient Greek and Roman literature and art, but also a new understanding of human abilities, and it sparked a Humanism that fostered joie de vivre, creativity and intellectual abilities. Luther’s conviction about the perspicuity of Scripture is not the only Reformation development that traces to this movement. A desire to read the Biblical Scriptures in their original languages rather than in translation – that is, in the Vulgate – also traces its roots to the Renaissance. A Greek edition of the New Testament produced by Erasmus of Rotterdam (1466?–1536) and the founding of Hebrew Studies by Reuchlin (1455–1522) gave Luther’s understanding of Scripture momentum to spread quickly far beyond Wittenberg and to exercise a defining influence on the subsequent history of Biblical interpretation. Luther’s radical differentiation between Scripture and interpretative tradition was not designed as a radical break with all tradition – after all, he thought about the relationship between the canonical Gospels and Augustine’s Gospel harmony. Luther did not challenge interpretative tradition per se, only the way it had been given precedence over Scripture.38 One could even say that Luther’s semiotichermeneutic insights included not only the idea that signs are chronologically prior to their interpretation but also the conviction that scripture is always richer than any individual interpretation of it. Luther applauded Erasmus’ Greek New Testament and was undisturbed by its considerable philological weaknesses. The Protestant paradox that eventually led to historical-critical exegesis – which has more in common with Celsus and Porphyry than with Origen and Augustine – is that the doctrine of the perspicuity of Scripture, which originated out of Humanism, gave rise to a sort of intense philological and historical study of Biblical texts that almost destroyed its own original idea.

3.3 The Significance of the Wars of Religion for the Development of Christian Biblical Criticism 3.3.1 From the Perspicuity of Scripture to Clarity of Analysis Martin Luther’s doctrine of Scripture was based on the conviction that no interpretation of Scripture is as clear as Scripture itself. The hermeneutical, methodological and theological program encapsulated in the slogan sola scriptura, which 38. Cf. Luther, Assertio, WA 7, 100: Non tamen per haec, sanctis patribus volo detractam auctoritatem.

3. Sad Sources

59

was essentially a call for individual Christians to interpret Christian sources for themselves, combined with the idea that the content of Scripture is clear, led to intensified philological concern about the exact wording of the canonical sources. Roman Catholic intellectuals reacted to Luther’s interpretative method and the theological hermeneutic that grew out of it by employing source criticism to argue that the canonical texts were not as clear as Luther had proclaimed, and that it was necessary to turn to Church tradition to clarify their meaning. The response to Reformation teaching that was formulated at the Council of Trent (1545–63) shows that the new thinking represented a significant challenge, and had already led to a strengthening of Scriptural authority even among Roman Catholics. The primacy granted to ‘original sources’ by Humanism was clearly denied, however, and the authoritative status of the Vulgata – the Latin translation of Scripture – that became a dogma at the Council of Basel-Ferrara-Florenz (1431–45), was confirmed. Furthermore, although Roman Catholics employed source criticism in their arguments, they did not exploit the potential benefits of text criticism or literary criticism, practices that could have helped them challenge the growing dogmatic power of sola scriptura. It was not until the seventeenth century that Richard Simon (1638–1712) would employ text criticism and literary criticism to argue against the Reformation understanding of Scripture, becoming a pioneer of historical-critical exegesis in the process.39 Lutheran orthodoxy countered Roman Catholic objections to sola scriptura by systemizing their doctrine of Scripture. Luther’s dynamic, reception-oriented approach to interpretation became a content-based doctrine about the perspicuity of Scripture. Scripture was described as being free of contradictions, a claim that was bolstered by a doctrine of verbal inspiration.40

39.  Johann Salomo Semler, the most influential early practitioner of historical-critical exegesis in the eighteenth century, was heavily indebted to the work of Richard Simon. See Johann Gottfried Eichhorn, ‘Johann Salomo Semler’, in Allgemeine Bibliothek der biblischen Litteratur, 5. vol. 1.  Stück (Leipzig, 1793), 1–183, here 86–7: Semler ‘trug [. . .] wiederholt die historischen Momente vor, welche der Kritik des A.T.  zur Unterlage dienen müssen [.  . .] Doch hatte er über alle diese Punkte wenig Eigenthümliches, sondern vertraute sich bey ihrer Erörterung ganz der Leitung des vortrefflichen, über sein Jahrhundert weit erhabenem Richard Simon, mit dessen Hülfe er durch die Schranken brach, mit welchen die bis dahin die christlichen Rabbinen das A. T. umschlossen hielten. Gieng er gleich über seinen Führer nicht hinaus, so half er doch durch seinen Beytritt seiner lang genug verschmähten guten Sache fort.’ 40. To date, research on Lutheran orthodoxy has generally been in the vein of systematic theology, with less interest in the history of theology and scholarship. Our understanding of Lutheran orthodoxy may change in the years to come as more detailed historical analyses are conducted. See, e.g. the unpublished dissertation of Daniel Bohnert, Friedrich Balduin (1575–1627) als Vertreter der Wittenberger Universitätstheologie im frühen 17. Jahrhundert (PhD diss., Goethe-University, Frankfurt am Main, FB Evangelische Theologie, 2016).

60

Stefan Alkier

By this point, the conflict between rival Christian groups was no longer being fought merely with intellectual arguments and debates. Actual weapons were being employed in a conflict where theological and political interests were inextricably entangled. Lies and murders were perpetrated in the name of ‘the true faith’ by members of all Christian groups involved, with significant consequences. The Thirty Years’ War (1618–48), which was explicitly proclaimed to be a battle for Christian truth, not only shook up the political order but also resulted in the rejection of authoritarian approaches to questions of truth. The bitter religious conflicts and civil wars of sixteenth- and seventeenth-century Britain had similar results. During the Enlightenment, the field of Mathematics became prominent because intellectuals like Descartes, Spinoza, Leibniz and Christian Wolf were horrified by the havoc that thirty years of civil war among Christians in Europe had wrought and insisted on ways of thinking that produced rules and order and did not facilitate lies. The Wars of Religion thus led to an interest in clear arguments based on verifiable rules and repeatable results. Since various parties in the Thirty Years’ War had based their authoritarian claims on the premise that their own religious group was the only one to represent true Christianity – often bolstered by political rhetoric about how God had worked miracles on their behalf – critique of miracles in the Enlightenment also became a means of criticizing arbitrary claims to power. This politically motivated critique of miracles in the Enlightenment led to changes in how God was viewed. God’s miracles were no longer considered signs of his sovereignty and power, but evidence of his tyrannical and self-contradictory nature. The critique that Enlightenment intellectuals like Spinoza aimed at miracles did not typically lead to atheism, however, but to naturalistic or deistic conceptions of God, where God was seen as a facilitator and preserver of order. The persuasive arguments of Enlightenment thinkers also brought the miracles of Scripture into disrepute, and contributed to new doubts about whether Scripture itself was actually clear, as a lengthy article on ‘Miracles’ in Zedler’s Universallexikon41 demonstrates. Supernaturalism arose in various forms as a rejoinder to a naturalistic approach to the Bible.42 Against this background, it is easy to understand the sensation caused by John Toland’s (1670–1722) publications on apocryphal texts, which featured a rationalistic approach to the Bible inspired by Baruch de Spinoza and John Locke. His

41. Cf. Article ‘Wunder, Wunderwercke’, in Zedlers grosses vollständiges Universallexikon aller Wissenschaften und Künste, welche bißhero durch menschlichen Verstand und Witz erfunden und verbessert worden, Bd. 59 (Halle and Leipzig:  Zedler-Verlag, 1752, The 64 Volumes of Zedler’s Universallexikon appeared in 1733–54. It does not name the authors of the articles. 42.  Cf. Stefan Alkier, ‘Wunderglaube als Tor zum Atheismus. Theologiegeschichtliche Anmerkungen zur Wunderkritik Ferdinand Christian Baurs’, in Ferdinand Christian Baur und die Geschichte des frühen Christentums, ed. Martin Bauspiess, Christof Landmesser and David Lincicum, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 333 (Tübingen: Mohr Siebeck, 2014), 285–312.

3. Sad Sources

61

book, Nazarenus; or Jewish, Gentile, and Mahometan Christianity (1718), suggested that the Gospel of Barnabas that he had discovered in Amsterdam was the original Gospel and inspired others to conduct historical investigations of the canonical Gospels in order to respond to the challenge to the authority of the canonical Gospels that Toland’s work represented. Although Toland’s arguments lack refinement, they should nevertheless be considered a prototype of the hypothesis that there was a ‘proto-Gospel’ that predated the Synoptics. ‘It was he who inaugurated a frontal assault on the legitimacy of the entire biblical canon.’43 The publication of Toland’s book thus treated the question of the Gospels’ historical origins as a ‘problem’, which was all the more fraught because the issue had the potential to cause doubt among believers. This potential for doubt is reflected in the subtitle of a careful study on the literary relationships between the canonical Gospels by Henry Owen, Rector of St. Olave in Hart-street and Fellow of the Royal Society. His Observations on the Four Gospels; Tending Chiefly, to Ascertain the Times of Their Publication; and to Illustrate the Form and Manner of Their Composition, which appeared in 1764 in London, anticipated both of the hypotheses that would later be associated with Johann Jakob Griesbach. According to Owen, the Evangelists wrote for particular audiences. He followed early Church tradition in considering the Gospel of Matthew to have been produced first, and also hypothesized that Luke had used and modified Matthew.44 The Gospel of Mark had been written at the request of the church in Rome, which had needed a ‘history for their use and instruction’. I say such a History. For the Gospel he wrote at their request is evidently a simple and compendious narrative, divested of almost all peculiarities, and accommodated to general use. In compiling this narrative, he had but little more to do, it seems, than to abridge the Gospels which lay before him – varying some expressions, and inserting some additions, as occasion required. That St. Mark followed this plan, no one can doubt, who compares his Gospel with those of the two former Evangelists. He copies largely from both: and takes either the one almost perpetually for his guide. The order indeed is his own, and is very close and well connected. In his account of facts he is also clear, exact, and critical; and the more so perhaps, as he wrote it for the perusal of a learned and critical people.45

Owen supported his arguments with tables showing correspondences between the Synoptic Gospels, which he provided as ‘examples’ of the literary relationships between them.46 He was proud of his comparative method, which he considered

43. Dungan, A History of the Synoptic Problem, 289. 44.  Cf. Henry Owen, Observations on the Four Gospels; Tending Chiefly, to Ascertain the Times of Their Publication, and to Illustrate the Form and Manner of Their Composition (London: T. Payne, 1764), 27. 45. Owen, Observations, 51–2. Emphasis original. 46. Cf. Owen, Observations, 53–72.

62

Stefan Alkier

to be an innovative, objective procedure. He saw himself as following in the footsteps of Biblical authors like Mark, who was notable for his clarity and critical faculties, and did not consider himself indebted to more recent Biblical critics like John Toland, who wanted to destroy the canon. For Owen, there was no inherent conflict between the perspicuity of Scripture and the employment of critical methods of interpreting the Bible: ‘There is a new field of Criticism opened, where the learned may usefully employ their abilities in comparing the several Gospels together and raising observations from that comparative way.’47 3.3.2 Disharmonization Strategies: Semler and Reimarus While Owen was looking for harmony between Bible, tradition and history of interpretation, Johann Salomo Semler (1725–91) was promoting a radical program of disharmonization. Semler latched onto the polyphony that had characterized Christianity from its earliest days and taught that diversity and differences were valuable. As his autobiography shows, he promoted appreciation for difference and respect for different ways of worshipping God as a response to the catastrophic Wars of Religion. He gives poignant descriptions of taking walks around the region where he lived, where one could still see the scars of the Thirty Years’ War.48 These scars served as a continual warning about the dangers of unilateral and intolerant truth claims. The paradigm shift that Semler’s work occasioned – replacing the earlier ideal of a univocal Christianity with a new tolerance for difference and diversity – had significant influence on Gotthold Ephraim Lessing (1729–81)49 and Johann Jakob Griesbach (1745–1812), and thus also on the formulation of the Synoptic Problem and the development of historical-critical hermeneutics and exegesis.50 Semler accepted the Reformation motto ‘ad fontes’, but focused his attention on the individual texts that made up the canon rather than on the canon as such. The theological differences that he observed between Biblical texts led him to reject the idea of the canon and to criticize the doctrine of verbal inspiration. As a corollary, extracanonical texts also became more interesting as legitimate expressions of individual Christian experiences. Semler’s disillusionment with univocality led him not only to reject Gospel harmonies as a shoddy type of apologetics but also to declare that the very idea of the Biblical canon had been a terrible mistake: ‘It was a useless effort to transform all of the books – of the whole Bible, and of the New Testament – into a homogeneous whole, melting away the differences between them in the process.’51 Semler

47. Cf. Owen, Observations, vi. Emphasis original. 48. Johann Salomo Semler, Lebensbeschreibung von ihm selbst abgefaßt I (Halle, 1781). 49.  See Stefan Alkier, Urchristentum Zur Geschichte und Theologie einer exegetischen Disziplin, Beiträge zur historischen Theologie (BHTh) 83 (Tübingen: Mohr Siebeck, 1993). 50. Cf. Eichhorn, ‘Johann Salomo Semler’, 1–183, esp. 1ff. 51. ‘Es war eine unnütze Demonstration, dass man alle Bücher, der ganzen Bibel, und des Neuen Testaments, in ein homogenes Ganze verwandelte; und den daseienden localen Unterschied also wieder wegschmelzte.’ Semler, Lebensbeschreibung von ihm selbst abgefaßt I, 282.

3. Sad Sources

63

attributed the very real differences between texts to the variety of circumstances in which they had been composed: ‘The Holy Scriptures consist of a number of books that were not all written at the same time, by the same authors, or in the same circumstances, and which are not always even discovered in the same circumstances.’52 Semler’s new appreciation for differences between Biblical books led him to reject not only the canon but also any attempt to idealize early Christianity.53 He proclaimed that the commonly held idea that the canon had been determined by the Evangelist John at the end of the first century was a dogmatic construct that did not stand up to historical-critical research. He suggested that the canon had developed gradually, a hypothesis that is still considered compelling today. According to Semler, the diversity of Christian expression traced back to the teaching of Jesus, who had adjusted his speeches for different audiences. Semler also invoked the idea of audience accommodation in connection with the teaching of the apostles and the composition of Biblical books. The diverse backgrounds of producers and recipients of Biblical books had influenced the ways in which ‘Christian religion’ was expressed: ‘Christians in Alexandria, Syria, Arabia, Phoenicia, Italy, the islands, etc., had their own particular forms of Christian religion, which reflected a multitude of different ideas, and Christian religion differed even when they came together because of disparities in the gifts of various teachers.’54 Semler did not think that ‘Christian religion’ had become any less diverse after the texts that would later become part of the New Testament were produced. In fact, plurality had been a necessary condition for the spread of Christianity. The first generations of Christ-followers were characterized by uncontrolled diversity, and there were no canonical texts or significant formal networks of congregations. Semler understood the origins of the Gospels to mirror the polyphony of ‘Christian religion’ in its local expressions: ‘All of the four extant Gospels strive to prove (as John says explicitly about his own work) that Jesus is the expected Messiah, based on the external history about him . . . They were Gospels for Jews . . . not Gospels for gentile Christians.’55

52. ‘Die heilige Schrift besteht aus mehrern Büchern, welche nicht zu einer und derselben Zeit, also auch nicht von einerley Verfassern, unter einerley Umständen, geschrieben sind, auch selbst nicht zu allen Zeiten unter einen und denselben Umständen gefunden werden.’ Semler, Versuch einer nähern Anleitung zu nützlichem Fleiße in der ganzen Gottesgelehrsamkeit (Halle, 1757), 175. 53. See Alkier, Urchristentum, 21–45. 54. ‘Zu gleicher Zeit haben also die Christen in Alexandrien, Syrien, in Arabien, Phönice, Italien, in den Inseln etc. die christliche Religion so wol als einzelne Privati, mit ihren tausendfachen verschiedenen Ideen; als auch in ihren Zusammenkünften nach der Ungleichheit der Gaben der Lehrer.’ Johann Salomo Semler, Theologische Briefe III (Leipzig: Weygandsche Buchhandlung, 1782), 204. 55. ‘Alle die vier noch vorhandenen Evangelien gehen (wie Johannes von dem seinigen ausdrücklich sagt) darauf aus, mittelst der äußeren Geschichte Jesus zu beweisen, daß er der erwartete Meßias sey […] Es waren Evangelien für Juden […] keine Evangelien für Heiden-Christen.’ Eichhorn, ‘Johann Salomo Semler’, 66f.

64

Stefan Alkier

Based on his model of the polyphony of early Christianity, Semler developed his ‘fragment hypothesis’, which Johann Gottfried Eichhorn describes: ‘Isolated pieces of written material about the life of Christ existed before our Gospels; the latter were later simply compiled from the former.’56 ‘John was the only writer to compose his Gospel in a free and independent manner. He wrote before the others, for foreign Christians of a less fearful Jewish character, who were more interested in Christ’s teaching than in his story.’57 ‘With such differences of origin and purpose, of perspectives taken and regions in which they appeared, it was impossible for the Gospels to find the same approval and acceptance in all communities at the time when they were first composed.’58 Semler thought that individual congregations had first begun to band together into regional networks – in a process that involved conflict and did not bring an end to diversity – in the second century, as texts began to be put into collections. Thus, as Semler reconstructed the history of the canon, he was also sketching the origins of the catholic Church, whose history he understood to be one of intolerance and power politics aimed at eliminating diversity among Christians. He saw apologetic interests in the writings of Papias and other early Christian writers who had mentioned the authors of the Gospels, all of whom sought to harmonize them and resolve contradictions between them,59 and his distrust of early Christian tradition also meant that one could no longer assume that Matthew had been composed before Mark and Luke. His analysis of Papias thus laid the foundation for the hypothesis of Markan priority, although he himself never actually articulated it. He was content to argue that the Gospels had been composed independently of one another in different places, at different times, for different local audiences. Meanwhile, his favourite Gospel was clearly John, which he dated before the Synoptics.60 He described the Synoptics as having been compiled from preexisting materials without reference to John’s Gospel. In any case, the Gospels represented not harmony but polyphony for Semler: a variety of voices speaking to a variety of local communities. In Semler’s revolutionary understanding of the canon, the Old and New Testaments were not considered to have been given by God, but to have emerged 56.  ‘Einzelne schriftliche Materialien zu Lebensbeschreibungen Christi waren früher vorhanden, als unsere Evangelien; und letztere wurden späterhin aus den erstern blos zusammengesetzt.’ Eichhorn, ‘Johann Salomo Semler’, 68. 57.  ‘Nur Johannes schrieb sein Evangelium als ein freyer unabhängiger Schriftsteller und früher als die übrigen für ausländische Christen von minder ängstlichem Juden-Geist, denen es mehr um die Lehre, als um die Geschichte Christi zu thun war.’ Eichhorn, ‘Johann Salomo Semler’, 69. 58.  ‘Bey solcher Verschiedenheit ihres Ursprungs und ihres Zwecks der bey ihnen genommenen Gesichtspunkte und der Gegenden ihrer Erscheinung konnten die Evangelien in der Zeit ihres Ursprungs unmöglich bey allen Gemeinen gleichen Beyfall und Eingang finden.’ Eichhorn, ‘Johann Salomo Semler’, 66f. 59. Semler, Theologische Briefe III, 204. 60. Eichhorn, ‘Johann Salomo Semler’, 69.

3. Sad Sources

65

through historical processes. He attributed the individual texts that made up the New Testament to particular Christian groups, and suggested that different versions of ‘Christian religion’ – some narrower and some more open – had competed with each other in the first century, just as in his own day. He considered this sort of antagonism to be a fundamental structural feature of Christianity, despite occasional attempts at compromise. For Semler, the Gospels represented the narrower perspective that he ascribed to ‘Jewish Christianity’, while the Pauline letters illustrated a more open attitude, which he labelled ‘Gnostic-Christian’. The Catholic Epistles belonged in a ‘compromise’ category.61 While Semler’s anti-Jewish comments might at first seem to reflect problems with the scholarly concept of ‘Jewish Christianity’ such as those identified by Helle Lemke,62 it is important to note that he did not operate with a model of progress. Semler believed that narrow, open and comprising stances could also be found among Christians in his own day, and had a right to exist. Disturbed by the Wars of Religion, he was advocating toleration and acceptance of different beliefs and ways of thinking; Historical Criticism served to demonstrate that the polyphony of Christian voices was theologically acceptable. He thus employed his study of the diversity of early Christianity for normative purposes, advocating freedom of religion and tolerance for all forms of Christian faith, as long as this did not present a danger to public order. For Semler, the future of Christianity lay in adopting the freedom and diversity of early Christianity as an ordering principle and allowing freedom of religion on an individual level while maintaining political order. God’s Word, the revelation of eternal truth, was not limited by any particular form of expression, whether of the Holy Scriptures or of early Christian creeds. God’s Word was not constrained by words on the page, but sought to influence the consciences of individuals in various ways. Jesus’ ‘only goal was to convince people that God could not be properly worshiped and loved unless one applied the strength of one’s soul and submitted inwardly to him and his identifiable purposes, with many concrete deeds, carefully performed’.63

61. Cf. Eichhorn, ‘Johann Salomo Semler’, 66. 62.  Helle Lemke, Judenchristentum. Zwischen Ausgrenzung und Integration. Zur Geschichte eines exegetischen Begriffes, Hamburger Theologische Studien 25 (Münster: LITVerlag, 2001), 50:  ‘Immer wieder erscheint die Aussage, daß das Judenchristentum “naturgemäß”, aufgrund seiner ‘Eigenbrödelei und Schrulligkeit’ oder seiner “Verhaftung” mit dem Judentum verkümmern und absterben mußte. Das Judenchristentum sei eine einmalige und beschränkte Größe geblieben, da es durch den ‘gottgewollten’ Übergang der Verkündigung zu den Heiden aufgehoben worden sei.’ This does not at all represent Semler’s position. 63. Jesu ‘einziger Hauptzwek war, die Menschen zu überzeugen, daß Gott ohne Anwendung der Seelenkräfte, ohne innere Ergebenheit gegen ihn und seine kentlichen Absichten, mit noch so vielen eignen äusserlichen Handlungen und noch so ernsthafter Genauigkeit darin, gar nicht gehörig verehret und geliebet heissen könne.’ Johann Salomo Semler, Vorbereitung zur theologischen Hermeneutik (Halle: Carl Hermann Hemmerde, 1760), 82–3.

66

Stefan Alkier

Semler’s arguments, set forth in numerous publications, were so persuasive that the picture he painted of a polyphonic and diverse Christianity in the first century was widely accepted. It was in this context that Gotthold Ephraim Lessing (1729– 81) published, between 1774 and 1778, some fragments of work by Hermann Samuel Reimarus (1694–1768), and  – drawing on Semler’s ‘fragment hypothesis’ – developed his own theory that the ‘Gospel of the Nazarenes’64 had been the ‘Ur-Gospel’, composed in Hebrew. The Reimarus Fragments opened a new chapter in debates about the Holy Scriptures – with a sharper polemical edge.65 Now there began to be talk of ‘fraud’ and ‘lies’, a development that transformed the diversity that Semler had seen in early Christianity into a ‘problem’ and turned the hermeneutic of suspicion into a trademark of historical-critical hermeneutics. Semler may have reacted so vehemently against Reimarus because they agreed at so many essential points.66 They were both convinced that the canonical Gospels reflected Jewish Messianic expectations, and they both acknowledged that there were irreconcilable differences between the texts. While Semler advocated simply accepting these differences, however, on the grounds that they reflected the difference between ‘God’s Word’ and ‘Holy Scripture’, Reimarus spoke of ‘fraud’. One of Lessing’s motivations in publishing the Reimarus Fragments may have been a desire to push for a public clarification of differences between the positions of Semler and Reimarus. In any case, by making the question of the origins of the Gospels a political issue, his publication contributed significantly to the continued development of various hypotheses. It had suddenly become clear that Semler’s version of history, which had achieved almost universal consensus among those with theological education, might be problematic. According to this model, the beginnings of Christianity had been riddled with diversity, and the canonical texts were not the Word of God, but a human expression of what it meant to worship God as a follower of Jesus Christ. It was not even possible to reconcile the Gospels with one another. The canon had not been verbally inspired, but was an instrument of power that had been employed by the catholic Church as it sought to become an official religion of the Roman Empire. What could one still believe? Where could one place one’s trust? Was the whole thing a lie?

64.  Gotthold Ephraim Lessing, Neue Hypothese über die Evangelisten als blos menschliche Geschichtsschreiber betrachtet, in Gotthold Ephraim Lessing, Sämtliche Schriften 16, ed. Karl Lachmann, 3.  auf ’s neue durchg. u.  verm. Aufl. besorgt durch Franz Muncker (Leipzig: Göschne’sche Verlagshandlung, 1902), 370–91(originally Wolfenbüttel, 1778). 65. Cf. Alkier, Urchristentum, 81–9. 66.  Johann Salomo Semler, Beantwortung der Fragmente eines Ungenannten insbesondere vom Zweck Jesu und seiner Jünger, neu hg. von und mit einer Einleitung versehen von Dirk Fleischer, Wissen und Kritik 25 (Waltrop: Hartmut Spenner, 2003 [originally 1781]).

3. Sad Sources

67

3.3.3 Strategies of Reharmonization: The Griesbach Hypothesis and Eichhorn’s ‘Ur-Gospel’ Hypothesis 3.3.3.1 The Griesbach Hypothesis Semler was hardly the only scholar to inquire into the beginnings of Christianity and to study the earliest sources. Philosophers and historians like Voltaire (1694–1778), David Hume (1711–76) and Edward Gibbon (1737–94) devoted themselves to writing ‘secular’ histories of Christian origins, intentionally distancing their work from histories written from a theological or ecclesiastical perspective. The mere fact that accounts of the history of Christianity were included in ‘secular’ histories such as Gibbon’s Decline and Fall of the Roman Empire, which appeared between 1776 and 1787 and was translated into German in 1788,67 was rightly seen as a provocation and challenge to the Church and to Christian doctrine. Despite some philosophical differences between them, all of the secular historians agreed that their own values were not reflected at the ‘beginning’ of any particular historical development, but at the ‘end’. This philosophical conviction led to the idea that history was progressing, and that historians could therefore investigate ‘beginnings’ fairly objectively, because they did not expect to discover ‘truth’ there. They could thus expose differences and contradictions unhindered. Like Semler, the secular historians left a challenging legacy to the nineteenth century: they painted a picture of early Christianity that characterized it as full of division and conflict, but did not think about how the various Christian groups, individuals and texts might be connected with one another.68 Johann Jakob Griesbach (1745–1812), an industrious and conservative student of Semler, tried to bring order to this chaos. He began with text criticism. Erasmus’ edition of the Bible had long been rendered out-of-date by new discoveries, but it was still being used. Griesbach was the first scholar who dared to print a new and, in part, different text that reflected the research on text criticism of more than 200 years.69 His teacher, Semler, was very active in the area of text criticism, and had collected more textual variants than Johann Jakob Wettstein (1693–1754), but he was not really a systematic thinker, so it was up to Griesbach to incorporate Semler’s findings into a new edition of the Bible, which appeared in 1774. Later, Griesbach wrote his Commentatio, qua Marci Evangelium totum e Matthaei et Lucae commentariis decerptum esse monstratur, which appeared in Jena in 1789–90 and offered a less controversial alternative to Reimarus’ ideas about ‘fraud’ in early Christianity, as well as a philological argument against the romanticized idea of an ‘Ur-Gospel’. Griesbach built on Henry Owen’s Observations on the Four Gospels, although without citing it and perhaps without realizing that he was doing so. He owned Owen’s book, however, and may even have met him on a trip to England.

67.  Edward Gibbon, The History of the Decline and Fall of the Roman Empire, ed. with Introduction, Notes and Appendices by J. B. Bury, 7 vols (London: Methuen, 1909–14). 68. On the connections, see Alkier, Urchristentum, 51–112. 69. Cf. Dungan, A History of the Synoptic Problem, 185–97.

68

Stefan Alkier

Griesbach developed his hypothesis in several stages. First, he developed the innovative idea of producing a Synopsis of the Gospels as an alternative to traditional Gospel harmonies. This new format, published by Griesbach for the first time in 1776,70 demonstrated his creativity and erudition, and his interest in examining both similarities and differences between the Gospels. The organization of his Synopsis  – like all synopses  – was by no means ‘objective’, however, but betrayed the influence of his thinking about literary relationships between the Gospels. It represented a milestone in interpretation of the Gospels – indeed, the term ‘Synoptic’ that is now used to refer to the first three Gospels traces its origins to this new way of presenting them. In 1783, Griesbach published a discussion of contradictions between the resurrection narratives.71 This work showed that Griesbach did not always follow in the footsteps of his teacher, Semler. He was not content simply to point out diversity and difference, but wanted to prove that the resurrection narratives were historically reliable. This apologetic desire to demonstrate the historical reliability of the Gospels was the real motivation for his work, in contrast to Semler. The first explicit formulation of Griesbach’s famous Hypothesis appeared in his Commentatio in 1789–9072 and responded to Augustine’s idea that Mark had worked with a copy of the Gospel of Matthew. Griesbach seemed to have been unaware of Augustine’s later suggestion that Mark had used both Matthew and Luke, and no reference is made to Henry Owen. Griesbach’s own aspirations are clear, however. He believed that whoever followed his line of argumentation would become convinced – based on the evidence of the texts themselves – that Matthew had written first, followed by Luke, and that Mark had composed a new Gospel on the basis of those two sources.73 Griesbach not only agreed with early Christian tradition that Matthew had written first but also considered Matthew an eyewitness and accepted as a historical fact the implication of the book of Acts that Mark had been a companion of Paul and Peter (see Acts 12.12).74 He considered his work to have rehabilitated the harmony that had once existed between the twin convictions that Scripture is clear and early Christian tradition is reliable.75 In other words, Griesbach believed that his Synopsis had re-established a harmonious concord between Scripture and tradition.

70. Synopsis Evangeliorum Matthaei, Marci et Lucae una cum iis Joannis pericopis, quae omnino cum caeterorum Evangelistarum narrationibus, conferendae sunt. Textum recensuit [. . .] Johann Jacob Griesbach (Halle, 1776). 71.  Johann Jacob Griesbach, Fontes unde evangelistae suas de resurrection Domini narrations hauserint:  Paschatos solemnia (Jena, 1783); reprinted in Johann Jacob Griesbach, Opuscula academica, ed. Johann Philipp Gabler, vol. 2 (Jena: Frommann, 1825), 241–56. 72.  Johann Jakob Griesbach, Commentatio qua Marci evangelium totum e Matthaei et Lucae commentariis decerptum esse monstratur (Jena, 1789–90). 73. See Griesbach, Commentatio, 12. 74. Griesbach, Commentatio, 12. 75. Griesbach, Commentatio, 11.

3. Sad Sources

69

Displeased with the ‘Ur-Gospel’ hypothesis of his student Johann Gottfried Eichhorn, which he found far too speculative, Griesbach continued writing on the Gospels in subsequent years. In 1794, he would argue in no uncertain terms against the idea that ‘shorter is more original’,76 which had been invoked to support the ‘Ur-Gospel’ hypothesis and the hypothesis of Markan priority. The latter had been developed by J. B. Koppe,77 and by Gottlob Christian Storr as part of an argument in favour of Gospel harmonies.78 Griesbach challenged Koppe and Eichhorn with the question, ‘Need the shorter Gospel be earlier than the longer?’ His own answer was No:  ‘It depends entirely on the intention of the author whether it is preferable to add to, or to subtract from, what others wrote before him.’79 With the help of his Synopsis, Griesbach also expanded philologically on the work of Henry Owen. Like Owen, he considered Mark to have been an intelligent theologian who had not just arbitrarily shortened the Gospels of Matthew and Luke, but who had carefully produced a new Gospel for a particular intended audience. He saw Mark as writing for a gentile-Christian audience somewhere outside of Palestine, although not necessarily for an educated audience in Rome, as Owen had suggested. He also differed from Owen in one more crucial point: according to Griesbach, Mark had not dreamed up a new order of events for the story, but had followed the plot of the Gospel of Matthew, deftly incorporating material from Luke along the way. Griesbach and Owen agreed at other points, however. Neither believed that Mark had carelessly epitomized his sources, and both thought that the literary excellence of the Gospel of Mark showed the author to be an educated and mature theologian who had known exactly what he wanted to accomplish and how he could achieve it. Many intelligent people were convinced by the rigor of Griesbach’s argument and the fact that his hypothesis required so few additional secondary hypotheses. At first, his hypothesis and the ‘Ur-Gospel’ hypothesis competed for 76.  Io. Iac. Griesbachii Theol. D.  et Prof Primar in academia Jenensi commentatio qua Marci Evangelium totum e Matthaei et Lucae commentariis decerptum esse monstratur, scripta nomine Academiae Jenesis, (1798.1790) iam recognita multisque augmentis locupletata (1794), printed in J. J.  Griesbach:  Synoptic and Text-Critical Studies 1776–1976, ed. Bernard Orchard and Thomas R.  W. Longstaff, Society for New Testament Studies Monograph Series (SNTSMS) 34 (Cambridge: Cambridge University Press, 1976), 74–102; English translation on pp. 103–35. 77. Johann Benjamin Koppe, Marcus non epitomator Matthaei (1782; new edition 1798); cf. Heinrich Greeven, ‘The Gospel Synopsis from 1776 to the present day’, in Orchard and Longstaff (eds), Griesbach: Synoptic and Text-Critical Studies, 22–49. 78. Gottlob Christian Storr, Ueber den Zweck der evangelischen Geschichte und der Briefe Johannis (Tübingen:  Jakob Friedrich Heerbrandt, 1786); Gottlob Christian Storr, De fontibus Evangeliorum Matthaei et Lucae, 1794; cf. Walter Schmithals, Einleitung in die drei ersten Evangelien (Berlin/New York: de Gruyter, 1985), 163–6. 79.  Griesbach, in Orchard and Longstaff (eds), Griesbach:  Synoptic and Text-Critical Studies, 119.

70

Stefan Alkier

precedence – until the romanticized idea that the Gospel of Mark was the earliest – because it was the shortest – began to spread. 3.3.3.2 The ‘Ur-Gospel’ Hypothesis of Johann Gottfried Eichhorn In response to the pluralistic view of the rise of Christianity a romanticized interest in ‘origins’ emerged as a way to explain diversity and render it tolerable. According to the principle of ‘origins’, – Ursprung – historical phenomena derive their identity from a temporal starting point – the Ur-Sprung – that bears in full the essence of that phenomenon. This sort of thinking underlies Johann Wolfgang Goethe’s idea that there was an ‘Ur-animal’ and an ‘Ur-plant’, an idea that he did not discard until after turning from Johann Gottfried Herder to Friedrich Schiller. While Goethe moved in other directions, Herder stuck with his interest in ‘origins’ and criticized the histories of Lessing and Kant on related grounds. The idea that origins and essences could be identified was so fascinating that it influenced almost all fields of intellectual endeavour,80 including the work of Johann Gottfried Eichhorn (1752–1827). Eichhorn believed that his ‘Ur-Gospel’ hypothesis had successfully countered Semler’s picture of a radically diverse early Christianity without actually dismissing Semler’s historical work, which Eichhorn valued. Eichhorn summarized his ‘Ur-Gospel’ hypothesis as follows: The lost Gospels are all part of a single tree, which split into two branches, each of which put forth its own twigs. I. Belonging to the main branch, from which the catholic Matthew sprouted, are 1) the Gospel of the Hebrews, 2) Cerinthus’ Gospel, 3) Justin’s Memoirs of the Apostles, and 4) Tatian’s memorandum about his Gospel harmony. II. Belonging to the other branch, from which the catholic Luke emerged, are 1) Marcion’s Gospel, and 2) one of the memoranda from Tatian’s Diatessaron. The tree out of which these two branches sprouted (that is, the common foundation of all of these lost Gospels) is a very ancient, short life of Jesus – an Ur-Gospel.81

80. See Alkier, Urchristentum, 113–72. 81.  ‘Die untergegangenen Evangelien sind alle Theile eines Stammes, der sich in 2 Aeste theilte, von denen jeder wieder seine eigenen Zweige getrieben hat. I.  Zu dem einen Hauptast, aus dem der katholische Matthäus entsprossen ist, gehörte 1)  das Evangelium der Hebräer 2)  Cerinth’s Evangelium 3)  Justin’s Denkwürdigkeiten der Apostel 4)  und Tatian’s eine Denkschrift seiner evang. Harmonie. II.  zu dem andern aus dem der katholische Lukas hervorgegangen ist, gehörte 1)  Marcion’s Evangelium und 2) eine der Denkschriften von Tatian’s Diatessaron. Der Stamm, aus dem diese zwei Aeste entsprossen sind, (oder die gemeinschaftliche Grundlage aller dieser untergegangenen Evangelien) ist ein uraltes kurzes Leben Jesu, ein Urevangelium.’ Johann Gottfried Eichhorn, Einleitung in das Neue Testament, 5  vols (Leipzig:  Weidmann, 1804–27), here I, XII. He first formulated his hypothesis in 1794. See Johann Gottfried Eichhorn, ‘Über die drey ersten Evangelien’, Allgemeine Bibliothek der biblischen Litteratur 5 (1794), 759–62.

3. Sad Sources

71

Eichhorn understood the Ur-Gospel to go back ultimately to Jesus and to be a replacement for oral accounts of the life and teaching of Jesus, written up by apostles who had been commissioned by Jesus himself.82 In contrast to Semler, Eichhorn denied that Jesus had varied his teaching for different audiences, and suggested that the Ur-Gospel had been just as uniform. Jesus had transmitted his teaching to the original twelve disciples in an unadulterated form, ensuring that the divine origins of Christian truth were not immediately polluted. Pollution arose when sub-apostles developed derivative narratives based on the pure narratives of the original Twelve. The original apostles then put the Ur-Gospel into writing to try to stop this dilution of the message. The term ‘Ur-Gospel’ thus denoted not just the chronological priority of that Gospel, but also its character: it preserved the divine teaching of Jesus in a pure and unalloyed form. Unfortunately, this Ur-Gospel had soon become adulterated. Even the canonical Gospels were adulterated versions, but the Ur-Gospel could be reconstructed from them by applying literary critical methods. In order to reconstruct the origins of Christianity, one therefore had to restore the Ur-Gospel by cleaning the dirt off the Gospels, thus bringing the life and teaching of Jesus to light. Ultimately, Christian truth was found in him. After reconstructing the Ur-Gospel, which came first ontologically and chronologically, it was theologically superfluous to reconstruct other, later versions of Christianity that were derivative and further removed from the original. All that was needed was Christianity’s charter document, the Ur-Gospel. Other Christian texts could go ahead and be diverse and differ from one another; they did not contribute anything significant to Christian truth, anyway. Johann Gottfried Eichhorn’s version of the Ur-Gospel hypothesis was based in its entirety on Semler’s reconstruction of a radically diverse early Christianity, but he interpreted that diversity differently than Semler. For Eichhorn, truth was expressed in a single voice and was threatened by polyphony. To reconstruct the single original voice, one had to reconstruct the pure source. The canonical Gospels had been tainted by polyphony, but could be cleaned up through the application of historical-critical methods. Eichhorn’s hypothesis, which was designed to counter the ‘fraud’ theory of Reimarus, became widely accepted as a model for explaining the emergence of the Gospels, even though he had not argued directly against Griesbach. Eichhorn’s ideas won the day because he had told a nicer story. 3.4 Afraid of the Myth: How the Two-Source Hypothesis Came into Being As we have seen, Semler’s description of early Christianity as a diverse movement was quite persuasive, as was his characterization of harmonizing projects as anxious, apologetically motivated falsifications of history. While Semler viewed polyphony among Christians – in the past and in his own day – as a welcome feature

82. See Alkier, Urchristentum, 140–4.

72

Stefan Alkier

of Christianity, however, some of his students did not share that conviction. For them, the diversity of the Gospels raised questions about their ‘truth’. An earlier approach to taming that diversity, the production of Gospel harmonies, was rejected by Lessing, whose Fragments by an Unknown Author dealt the death blow to the whole genre. Lessing had trouble simply accepting polyphony, however, and sought to bring order to chaos in another way: by positing the existence of a Gospel of the Nazarenes that predated the canonical Gospels. According to his ‘New Hypothesis on the Evangelists as Merely Human Historians’, the publication of which was arranged by his brother Karl Lessing in 1784, ‘The Gospel of the Nazarenes was not some misconceived production substituted at a later date, [and] it was also older than all our four gospels, the first of which was written at least thirty years after Christ’s death.’83 According to Lessing, this first Gospel had been translated from Hebrew into a number of other languages and had been used as a source by the authors of later Gospels. Matthew had been the first to employ it in the production of a Gospel in Greek. Lessing’s hypothesis expresses his desire to find a natural, organic way of connecting the Gospels with one another. Indeed, this was the goal of all theories of literary dependence:  to create a coherent and compelling harmony out of what might otherwise seem to be an incoherent, uncontrolled plurality. The ideal solution was to find a beginning point that accounted for the diversity of the later phenomena and rendered it less threatening. The divergent voices of the Gospels thus led to the Synoptic Problem – an urgent quest to identify the origins of the Gospels, even if that meant troubling the extant texts in the process. It was thought that by tracing genetic relationships between texts, one could restore the cohesion between the Gospels that Semler had destroyed and refute Reimarus’ accusations of ‘fraud’. Historical-critical observations about diversity were thus transformed into an anxious quest for historical certainty about the ‘Gospel history’ of Jesus Christ, which would characterize both Literary Criticism and ‘life of Jesus’ research in the nineteenth century. The genre entitled ‘Life of Jesus’ is the new Harmony of the Gospels that destructs its polyphonic sources for the sake of the one and only ipsissima vox that bears and guarantees the one truth on the literal level. ‘Truth’ was seen as a matter of historical reliability and nothing more – a development that Origen had already warned against. While Johann Gottfried Herder’s ‘Rules for harmonizing our Gospels from Their Origin and Order’ (1797) sought to counteract this reductionist equating of ‘truth’ with historical reliability, he also articulated the idea that plurality is always preceded by singularity. According to Herder, simpler forms of texts, ideas, and the like predate more complex ones. This romantic ideology underlies the hypothesis of Markan priority, which had already been formulated by Koppe and which Herder now sought to make plausible by providing organological arguments. He 83. Gotthold Ephraim Lessing, ‘Neue Hypothese über die Evangelisten als blos menschliche Geschichtschreiber’, in Gotthold Ephraim Lessing, Philosophical and Theological Writings, trans. and ed. H. B. Nisbet, Cambridge Texts in the History of Philosophy (CTHP) (Cambridge: Cambridge University Press, 2005), 148–71, here 155.

3. Sad Sources

73

applied his theory that diversity can always be traced to a common source to the Gospel of Mark,84 countering Griesbach’s philological arguments with aesthetic ones: ‘Is this not the more natural position to take? Is not the shorter form, the unadorned one, generally the earlier, which is later embellished to add explanation, fullness, and roundness? This is clearly the case with Mark, Matthew, and Luke, even if we do not know anything about the authors.’85 Herder also considered oral communication to be earlier and ‘truer’ than written communication, and he thus attributed a sort of double originality to the Gospel of Mark. According to his ‘Diegesentheorie’,86 the Gospel tradition had begun with the oral circulation of simple stories. The Gospel of Mark began to circulate in oral form not long after Jesus’ death, and was first written down by Mark in the 60s. It was the ‘original’ Gospel  – the ‘archetype’.87 The Gospel of Matthew, which was ‘written using the Hebrew alphabet in the vernacular language of Palestine’, drew on the Gospel of Mark and was effectively a ‘commentary’.88 Luke had employed Mark, Matthew and other Gospels. Herder rejected the idea of a lost work like Eichhorn’s Ur-Gospel in no uncertain terms: ‘The attempt to trace our three Gospels to one lost Ur-Text is an adventurous, aimless pursuit. No such unwritten Ur-Text . . . can be reconstructed from the phrases and expressions of a free account with any certainty. Indeed, this audacious endeavor destroys any natural way of looking at our Gospels themselves by heaping problems on problems.’89 Herder was not primarily interested in demonstrating the historical reliability of the Gospels when he formulated his hypothesis about oral tradition and literary dependence. Rather, he wanted to ensure that each text was appreciated on its own poetical and theological merits, regardless of its similarity to other texts. The goal was aesthetic analysis aimed at capturing the uniqueness of each author.90 Two works that appeared independently of one another in 1838 – and which led to the definitive breakthrough of the hypothesis of Markan priority  – were similarly influenced by theological and philosophical premises and anxieties. Although Griesbach’s hypothesis had by no means been forced off the stage by the Ur-Gospel hypothesis and was still being championed by Ferdinand Christian

84. See Johann Gottfried Herder, Ideen zu Geschichte der Philosophie der Menschheit, in J. G. Herder, Sämtliche Werke XIII–XIV, ed. Bernhard Suphan (Berlin: Weidmannsche Buchhandlung, 1887/1909, reprint Hildesheim: Georg Olms, 1967). 85.  Johann Gottfried Herder, ‘Regel der Zusammenstimmung unsrer Evangelien, aus ihrer Entstehung und Ordnung’, in ders., Christliche Schriften. Dritte Sammlung (Berlin:  Weidmann, 1797), 391. Cf. Koppe, Marcus non Epitomator Matthaei ; Gottlob Christian Storr, De fonte evangeliorum Matthaei et Lucae (1794). 86. See Herder, Christliche Schriften, 390. 87. Ibid., 394. 88. Ibid., 409. 89. Ibid., 417. 90. Ibid., 414.

74

Stefan Alkier

Baur (1792–1860) and his students, the authors of the works that I  will now describe saw no need to engage with his philological arguments. Christian Hermann Weisse (1801–66) took exception to the ‘myth theory’ of David Friedrich Strauss (1808–74), one of Baur’s students. Strauss had accepted the Griesbach Hypothesis as an explanation for the written phase of Gospel production, but had given more attention to earlier stages in the Gospel tradition, when material had circulated orally – here one sees the influence of Herder and Eichhorn on his work. Strauss had described the Gospels as consisting of ideas – which had been preserved in the words of Jesus, especially the parables – clothed in mythical garb. When Strauss’ Life of Jesus appeared in 1835, therefore, the historical reliability of the narrative material of the Gospels had been called into question. While Weisse disagreed with Strauss’ ‘myth theory’ of the Gospels, he did not criticize Strauss’ appeal to philosophical theory per se. As the title of Gospel History, Critically and Philosophically Examined (1838) demonstrates, Weisse was convinced – as were all historical-critical exegetes at the time – that exegesis in their generation owed a great debt to philosophy. The first sentence of his book also shows that he was one of the many scholars at the time who did not consider the Scriptures to be ‘perspicuous’: The idea of examining Gospel history pre-dates the appearance of Strauß’ famous work, but I would not dare to claim boldly that it would have been carried out without that work. My inclination, as will be evident, is not negative or critical, but essentially positive: to produce a historical picture of Jesus out of the unclear shell in which tradition early on – and later Church doctrine –encompassed it, a conviction I share with the majority of educated people in our day.91

What had once been seen as ‘perspicuity’ was thus now thought of as an unclear sleeve. Anyone who reads through Weisse’s two long volumes hoping for a historicalcritical analysis of the sources will be disappointed. Weisse was neither a philologist nor an exegete. He was Professor of Philosophy in Leipzig, and had a special interest in philosophy of religion.92 The bulk of his Gospel History consists of a Gospel harmony resembling that of Augustine. In contrast to Augustine, however, Weisse was not interested in the texts themselves, which were simply a means to an end. Nor was the ‘end’ a reconstruction of ‘Gospel history’ as the Gospels portrayed it, but of ‘Gospel history’ at an earlier stage. The once illustrious Human, Lion, Ox and Eagle had become troubled, murky waters that only yielded ‘truth’ if one laboured long and hard.

91. Christian Hermann Weisse, Die evangelische Geschichte kritisch und philosophisch bearbeitet, 2 vols (Leipzig: Breitkopf & Härtel, 1838), here vol 1. 92.  His most important work is Philosophische Dogmatik oder Philosophie des Christenthums, 3 vols (Leipzig: Breitkopf & Härtel, 1855–62).

3. Sad Sources

75

Weisse told his ‘true’, ‘historically reliable’ story of Jesus with apologetic fervour, claiming to be ‘objective’ and seeking to allay fears that the truth of Christianity might be in danger. One might not be able to trust that the sources were clear, but one could trust the ‘objectivity’ of hypotheses about them and the clarity of contemporary reconstructions. Hypotheses about sources were thus now themselves an object of belief. Weisse, the philosopher of religion, and Christian Gottlob Wilke (1786–1854), a Lutheran pastor and preacher, were both so convinced of Markan priority that they did not engage with the potential counterarguments offered by the great philologian Johann Jakob Griesbach. They were more concerned with defending the historical truth of ‘Gospel history’ – and they were also enemies who had competing interests, ideologies and hypotheses. While Weisse saw the Gospel of Mark as drawing on the witness of Peter, albeit second-hand, Wilke considered it to be the ‘Ur-Gospel’, of high literary quality and guarantor for the reliability of the gospel of Jesus Christ. He agreed with early Christian tradition that the order of the episodes in the Gospel did not reflect the historical order of events: Mark is the first evangelist. His work forms the basis for the other Gospels of Matthew and Luke. This work is not a copy of an oral Ur-Gospel, but an artistic composition.93

Wilke’s reflections led him to the conclusion that Scripture on its own was an insufficient basis for faith. Sola scriptura was simply not an option any more. Wilke and his wife converted to Roman Catholicism in 1846. In contrast to the philosopher Weisse, the pastor Wilke would not have accepted the Two-Source Hypothesis. He believed that Luke had used Mark, and that Matthew had used Luke. Weisse, in contrast, posited relationships of literary dependency that involved both the Synoptics and extra-Synoptic sources, laying the foundation for the so-called Two-Source Hypothesis in the process. Weisse’s theory – that Luke and Matthew had used the Gospel of Mark and a collection of sayings put together by the apostle Matthew, as well as additional oral or written sources  – required a whole array of assumptions. The label that would be later applied to it – the ‘Two-Source Hypothesis’ – is therefore misleading. Rather than a simple model involving two and only two sources, this theory combines a complex range of hypotheses about literary dependence, Vorlagen, traditions and fragments, and requires a fair amount of historical fantasy. Heinrich Julius Holtzmann’s book, The Synoptic Gospels, which appeared in 1863,94 helped the ‘Two-Source Hypothesis’ to achieve its definitive breakthrough. 93.  Christian Gottlob Wilke, Der Urevangelist, oder exegetisch kritische Untersuchung über das Verwandtschaftsverhältniss der drei ersten Evangelien (Dresden und Leipzig: G. Fleischer, 1838), 684. 94.  Heinrich Julius Holtzmann, Die synoptischen Evangelien (Leipzig: Wilhelm Engelmann, 1863).

76

Stefan Alkier

Holtzmann initially ‘solved’ problems associated with ‘minor agreements’95  – as they are now known, and which had already been noted by Weisse – by means of an additional hypothesis that posited the existence of a ‘Proto-Mark’. He later abandoned that idea, however,96 and returned to theories of literary dependency, suggesting that Luke had used both the Gospel of Mark and the Gospel of Matthew. The unscholarly confidence with which so many protagonists, in the quest for a ‘solution’ to the Synoptic Problem, laid out their own theories and dismissed others should serve as a warning to all who aspire to scholarly excellence. Hypotheses should not be presented as ‘facts’ or ‘data’, and one should not emulate Wilke, who writes near the end of Der Urevangelist oder exegetisch-kritische Untersuchung über das Verwandtschaftsverhältnis der drei ersten Evangelien, ‘We guarantee for all eternity, that our result is the correct one’.97 3.5 Challenges and Insights from Reflection on the Historical Development of the Synoptic Problem: 10 Theses 1. The ‘Synoptic Problem’ is not a datum or a fact about the canonical Gospels. Rather, it is the result of interpreting the data in constellations of identifiable scientific and political-historical impulses within the larger endeavour of the production of knowledge in Europe being brought to bear on the study of the Bible. 2. Pressure to ‘solve’ the Synoptic Problem does not come from the Gospels themselves, but from the philosophical, hermeneutical and theological presuppositions of exegetes and other scholars who are engaged in the question of the truth of the gospels. 3. The history of scholarship is not a history of steady progress, but of the ways in which discourses have evolved. While some insights have been gained, others have been forgotten while still others have been transformed. 4. It is valuable to revisit the solutions to the Synoptic Problem that were suggested throughout the centuries as a way of adding nuance to contemporary discourse about the Gospels. 5. No single solution to the Synoptic Problem that has been proposed to date accounts so convincingly for all of the data that it can fully satisfy the challenges raised against it. 6. One should ask which approaches are the most plausible. Which approach accounts for the data in the most reasonable manner? Which requires the fewest secondary hypotheses?

95.  Cf. Georg Strecker (ed.), Minor Agreements, Symposium Göttingen 1991 (Göttingen: Vandenhoeck & Ruprecht, 1998). 96. Heinrich Julius Holtzmann, Lehrbuch der historisch-kritischen Einleitung in das Neue Testament (Tübingen: Mohr (Siebeck),1885). 97.  ‘Wir geben für alle Ewigkeit Brief und Siegel, daß unser Resultat das richtige sei.’ Wilke, Urevangelist, 684.

3. Sad Sources

77

7. The assumptions that underlie various approaches to the Synoptic Problem – for example, ‘Mark is the oldest Gospel’, ‘Q existed’98 – should not be treated as ‘data’. The only ‘data’ we have are the text(s) of the Gospels themselves. 8. One of the most important results of the debate about the Synoptic Problem has been the realization that the numerous verbal and syntactical agreements between Matthew, Mark and Luke cannot be explained on the basis of oral tradition alone. Some sort of literary dependence was a factor in their production. 9. It follows from Thesis 8 that the only plausible hypotheses are those in which (at least some of) the authors of the Synoptic Gospels made use of written sources. Fulfilling this criterion are hypotheses that posit the use of sources other than the Synoptic Gospels (e.g. Eichhorn’s hypothesis that there was an ‘Ur-Gospel’), those that suggest literary relationships among the Synoptics alone (e.g. the Griesbach Hypothesis) and those that favour a combination of these factors (e.g. the Two-Source Hypothesis). 10. Current scholarly discourse cannot honestly report on a solution to the Synoptic Problem. Contemporary study demonstrates instead a number of different approaches to a solution. Thomas R.W. Longstaff ’s summary of the findings from the 1976 Muenster Griesbach conference continues to hold and should be foundational for all further discussion of the Synoptic Problem. He writes, ‘Perhaps one of the most important results of the Colloquium was the clear articulation (. . .) of the view that New Testament scholarship has entered a new period of pluralism with regard to the Synoptic Problem. Although the strength of the two-document hypothesis was frequently stressed, nearly everyone present agreed that the solution was not without difficulties and that it could no longer be considered to be established beyond reasonable doubt. In fact, most participants acknowledged that New Testament research would profit from a climate in which a plurality of hypotheses could be accepted as legitimate starting points for exegetical studies.’99

98.  Cf. Bo Reicke, ‘Griesbach’s answer to the Synoptic Question’, in Griesbach: Synoptic and Text-Critical Studies, 67: ‘These two assertions, the priority of Mark and the existence of a Logia source, soon became popular in Protestant Germany, and even took on the character of an article of faith that later spread to other countries and churches. Many scholars accept them as if they were axioms. But such idola theatri should always be called into question.’ 99. Thomas R. W. Longstaff, ‘At the Colloquium’s Conclusion’, in Griesbach: Synoptic and Text-Critical Studies, 173–4.

Chapter 4 W E R E T H E G O SP E L A U T HO R S R E A L LY ‘ S I M P L E C H R I ST IA N S W I T HOU T L I T E R A RY G I F T ’ ( A L B E RT S C H W E I T Z E R ) ? :   A R G UM E N T S F O R T H E Q U E ST F O R S OU R C E S B E H I N D T H E G O SP E L S Mogens Müller

4.1 Introduction We can observe two main tendencies in today’s discussions of the mutual relations between the synoptic gospels.1 One is, to the greatest possible extent, to look for sources behind the three writings, thereby attributing to their authors the least possible independent creativity. The other goes in the opposite direction by operating primarily with the actual texts and, accordingly, viewing the later synoptic gospels as results of the continuous work of otherwise independent and creative authors. Here the hypothesis claiming mutual literary dependence carried the day. By its arrival, the first tendency rapidly called forth hypotheses about later lost sources. In this context, not least the so-called Two-Source hypothesis became influential. Practically from the very beginning it was marked by the fundamental concern of the Enlightenment to reach back to the historical ‘bedrock’ of the traditions. With regard to the gospels, it meant to detect the historical Jesus. According to Enlightenment thinking, true Christianity should be the religion of Jesus. This way a division was established between, on the one hand, the historical Jesus and, on the other, the apostles and their preaching. This divided interpretation from its subject in such a way that a necessary connection was not even supposed. The Jesus For generous help in improving my English I warmly want to thank Revd. Jim West, ThD. 1. For excellent information, see Christopher Tuckett, ‘The Current State of the Synoptic Problem’, in New Studies in the Synoptic Problem. FS Christopher M. Tuckett, ed. Paul Foster, Andrew Gregory, John S. Kloppenborg and Joseph Verheyden, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 239 (Leuven:  Peeters, 2011), 9–50; repr. in Christopher M. Tuckett, From the Sayings to the Gospels, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 328 (Tübingen:  Mohr Siebeck, 2014), 77–116. This overview, however, also makes it clear that inclusion of new aspects, in an essential way, may change the picture Tuckett offers.

80

Mogens Müller

of history and the Christ-proclamation of the apostles and earliest Christianity were seen as something between which one could choose. Seen from this perspective, the New Testament gospels foremost became interesting as, generally speaking, the only sources for the historical Jesus, his life and his fate. Relatively early the Gospel of John was separated from the three others as of no use as a source; the historical here got lost in an already advanced interpretation and mythification process. Thereby it became possible to evade the dilemma with which the obvious difference between the Gospel of John and the Synoptic Gospels had confronted the interpreters from the Early Church and onwards. The time was gone when merely its place in the canon made the Gospel of John a trustworthy source for the Jesus of history. Viewed in another perspective, however, it can be ascertained that what happened was that the harmonization effort, which until then had gone far to make the four gospels’ stories parts of one consistent history, was now replaced by interest in a reliable historical point of departure. In this context, the authors behind the Synoptic Gospels were perceived as vehicles of tradition, and the less they were assumed to have intervened in the material, the better. Any independent creativity on their behalf could only be a misleading factor with regard to the historical point of departure. That is what Albert Schweitzer expressed clearly in the (mostly beseeching) characterizing of the gospel authors as ‘simple Christians without literary gift (literarisch unbegabten Christen)’, which I have used as the title for this chapter. Because of this, it can be interesting to follow up on a tendency in a great part of scholarship, to more and more evade or minimize the individual traits in these writings. It needed a change, in both the view of the theological importance of the historical Jesus and in the perception of what foremost the gospels are sources to, to reach an evaluation of them as what they at least also are; namely, testimonies about stations in a necessary process of interpretation. Form-history had perceived it as its task to define the forces which beforehand had formed the traditions, and which were to be included in the available writings. Their ‘authors’ first really came to sight with the notion of the systematic interest of Redaction criticism in the distinctive, individual stamps of the four. Concurrently, the theological interest in the historical Jesus receded into the background, and a transition from Redaction history to Redaction criticism or Reception criticism came about. The result was a growing ascertaining in scholarship of the relative independence and creativity of these authors. The gospels were now perceived foremost as sources to the arrangement, or rewriting, of the Jesus story according to the theology and preaching of the individual gospel author.

4.2 From Urevangelium Hypothesis to the Mark Hypothesis When critical scholarship, at the end of the eighteenth century, began to question the correctness of the names of the authors attached to the four gospels by early church tradition, it also, at the same time, challenged their historical trustworthiness. Until then a harmonizing reading had taken its point of departure in the

4. Without Literary Gifts

81

acceptance of the four gospels as inspired accounts and therefore fundamentally historically reliable. The detection of their pseudonymity let their anonymous creators appear to be independent, creative authors. While credibility in this period unambiguously becomes a question of historical trustworthiness, these authors turn out to be a problem. They could not any longer be considered trustworthy mediators of an original Jesus tradition, but only as possible sources to it. Said in another way: The true gospel tradition was assumed to precede the available writings.2 The father of the printed gospel synopsis, Johann Jakob Griesbach (1745–1812), actually created his work to problematize a harmonizing reading of the Synoptic Gospels. Through his synoptic arrangement, first to appear in print in 1776, he thus shows that they generally concur in their rendering of the order of the events, but that it is also impossible to tell who is offering the right one when they differ. This is the more remarkable if one assumes a literary dependency between the three. In contrast to the common assumption that it is the case that the Gospel of Matthew underlies the Gospel of Mark, and that both of them were later on used by the author of the Gospel of Luke, Griesbach put forth the hypothesis that the Gospel of Mark was dependent on both of the other two Synoptic Gospels thus being the youngest of them. This so-called Griesbach hypothesis has its adherents to this day. This solution, however, was by far not the only one proposed. Thus the founder of the so-called older Tübingen school, Gottlob Christian Storr (1746–1805), representing a biblical supernaturalism, in an investigation from 1786,3 on the basis of a purely historical argument, reached the conclusion that it was the Gospels of Matthew and Luke which were dependent on the Gospel of Mark. Otherwise, it would be inexplicable that the Gospel of Mark lacks so much of the material included in the two greater gospels. Thus, Storr is an early representative of the Markan hypothesis. In contrast, Gotthold Ephraim Lessing (1729–81), in some posthumously published smaller works,4 makes himself a spokesperson for the view that there was no literary dependence between the three first gospels. Instead, each of them

2.  Below, for the older literature, I  depend mostly on Werner Georg Kümmel, Das Neue Testament. Geschichte der Erforschung seiner Probleme, Orbis Academicus III/3 (München:  Karl Alber, 2., überarb. und erg. Auflage, 1970), thereby also being able to use the translations of the old German texts in the English edition, The New Testament. The History of the Investigation of Its Problems, New Testament Library (London: SCM, 1973). It should, however, be remarked that Kümmel’s presentation is certainly not totally unaffected by his being a devout adherer of the two-document hypothesis. Also, it has to be taken into account that Kümmel writes his history of interpretation as a history of progress in what further narrows the perspective. 3. Ueber den Zweck der evangelischen Geschichte und der Briefe Johannis (1786). 4. Gotthold Ephraim Lessing, Thesen aus der Kirchengeschichte (1776); Neue Hypothese über die Evangelisten als bloß menschliche Geschichtsschreiber betrachtet (1778). Both of these writings were published posthumously.

82

Mogens Müller

build upon an older Aramaic or Hebrew primitive gospel. If the Gospel of Mark is shorter than the others, it is because its author had access only to a shorter edition of this earlier writing. It was also the need, which gradually came about, to have this primitive gospel in translations to the languages of the surrounding world which resulted in the creation – according to Lessing’s assumption – first of the Gospel of Matthew, and the two other gospels later on. Put in another way:  According to Lessing, the creators of the three first gospels were not real authors. This, however, is the case with the author of the Gospel of John. Although he knew the Hebrew primitive writing, in relation to this his gospel constitutes a class of itself. As Lessing argues: If therefore Christianity was not to fall asleep and to disappear among the Jews again as a mere Jewish sect, and if it was to endure among the Gentiles as a separate, independent religion, John must come forward and write his Gospel. It was only his Gospel which gave the Christian religion its true consistency. We have only his Gospel to thank if the Christian religion, despite all attacks, [continues] in its consistency and will probably survive as long as there are men who think they need a mediator between themselves and the Deity: that is, for ever.5

Thus, Lessing estimates the author of the Gospel of John to be a creative theologian who transforms the Jesus story according to his intention in writing it down, while he perceived the other three only as translators of an Aramaic/Hebrew primitive writing. Johann Gottfried Eichhorn (1752–1827) takes over the conception of an Aramaic/Hebrew primitive gospel as the source of the three first gospels.6 Thus, he also deems it necessary to get behind them if we want to detect those parts of the life of Jesus, ‘which the first teachers of Christendom regarded as alone essential for the establishment of the Christian faith among their Jewish contemporaries’.7 This primitive gospel has not, as later on the Gospels of Matthew and Luke would, related stories of the conception, birth and childhood of Jesus, traits surely to be characterized as later, legendary embellishment of tradition, which it must be a legitimate theological task to separate. He would be actually a traitor to religion who in his zeal for matters of tradition would wish to oppose the attempt to reduce the Gospels solely to their apostolic content and to free them again from the additions and embellishments that were

5.  Lessing, Neue Hypothese; cited after Kümmel, Das Neue Testament, 91–2/The New Testament, 77. 6.  See his article: Johann Gottfried Eichhorn, ‘Über die drey ersten Evangelien’ (1794). Kümmel, Das Neue Testament, 93/The New Testament, 77. 7.  Johann Gottfried Eichhorn, Einleitung in das Neue Testament (1804). Kümmel, Das Neue Testament, 93/The New Testament, 78.

4. Without Literary Gifts

83

made at a later time in the course of their reworking. He would oppose the only means by which not only apparent, but also very well-grounded objections to the life and deeds of Jesus can be removed and their credibility be rescued from attacks of intelligent doubters . . . By this separation of the apostolic from the nonapostolic which higher criticism – if only its gift be not spurned – recommends for the most important of reasons, the means are found to establish the credibility and truth of the gospel story on unshakable foundations.8

Again, trustworthiness depends on the eventual historical truth – here defined as genuinely apostolic – that may lie behind the existing gospels, which are conceived with an unapostolic tendency. Contemporary Johann Gottfried Herder (1744–1803), like Lessing, claims that the Gospel of John belongs to its own class. At the same time Herder also acknowledges the distinctive stamps of the three other gospels as independent developments of a primitive gospel. This alone, he postulates, was orally transmitted, and was an oral sermon. ‘Therefore, before any of our Gospels were written, the Gospel was there in the proclamation of Christ and of the apostles.’9 The apostle Peter thus had a ‘full Christian Gospel (Acts 2:22–39)’ at his disposal at the first Pentecost. Thus it all did not begin by writing down a gospel but with the preaching of past and future things (kerygma and revelation), with exposition, teaching, consolation, admonition: with sermon. In general the Gospel of John best demonstrates the idea . . . that they (the Gospels) are not in any sense biographies, but were intended to be historical documentation of the Christian confession of faith that Jesus was the Messiah, and after what fashion he was made the Messiah. John’s Gospel as the latest pursues this purpose in most definite outline, and in the course of so doing an actual biography is wholly lost to view, though it is also true that one ought not to think of biography as the main concern even in the older Gospels. They are what their names implies . . . When therefore under such circumstances a Gospel was written, it could be recorded in no other sense than this. There was neither purpose nor motive to report anecdotes out of the private life of Jesus, for those who composed it [the Gospel] and those for whom it was composed were not the public interested in writing and reading that is characteristic of our times.10

8.  Eichhorn, Einleitung in das Neue Testament. Kümmel, Das Neue Testament, 93/The New Testament, 79. 9.  Johann Gottfried Herder, Vom Erlöser der Menschen. Nach unseren drei ersten Evangelien (1796). Von Gottes Sohn, der Welt Heiland. Nach Johannes Evangelium (1797). Kümmel, Das Neue Testament, 95/The New Testament, 80–1. Regrettably, it is difficult to see in the reference system in Kümmel which writing exactly it is, that is quoted. Thus I only give the page numbers of Kümmel, Das Neue Testament, 91–2/The New Testament, 80. 10. Kümmel, Das Neue Testament, 95–6/The New Testament, 80–1.

84

Mogens Müller

Thus we have to imagine that the evangelists, foremost, wrote down the oral tradition, and to acknowledge that they did so to give rise to faith in Jesus as Christ and Son of God. It was simply impossible that Christianity at that time could be spread among the peoples without written records. But it was also impossible with regard to this period and situation ‘to think of an apostolic Gospel chancellery in Jerusalem that sent tracts with each teacher to each congregation and equipped him with written Gospels’11 The whole idea of our evangelists as scribes (grammateis, scribae) assembling, enlarging, improving, collating, and comparing tracts is strange to, and remote from, that of all ancient writings that speak of their activities, and even more foreign to conclusions drawn from observing themselves, and most of all to their situation, their motivation, and the purpose of their Gospels.12

Herder also forms an idea about these oral evangelists as ‘Evangelischen Rhapsoden’, who – as the first three gospels have it – kept to specified parables, miracles and speeches. For even in the oral narrative not all is equally free. Pregnant sayings, principal statements and parables soon find the same expression as less important parts. Thus, according to Herder: The common Gospel consisted of individual units, narratives, parables, sayings, pericopes. This is evident from the very appearance of the Gospels and from the different order of this or that parable or saga . . . The fact that it consists of such parts vouches for the truth of the Gospel, for people such as most of the apostles were, more easily recall a saying, a parable, an apothegm that they found striking than connected discourses.13

This is also to be observed in the Gospel of Mark, the primacy and distinction of which Herder really has a clear perception of, although without making its author a creative person. According to Herder, it is a mere misconception of this evangelist if he ‘is regarded as a frugal epitomizer of Matthew or as a comparably cautious compiler of our Matthew and our Luke’,14 and if it is – as usual – read after the Gospel of Matthew, its worth nearly disappears. If the Gospel of Mark was to stand by itself (as of course it did when it was first written), it would occupy a high place by reason of the simple principle: ‘Mark’s Gospel is not an abbreviation, but a Gospel in its own right. What others have in a more expanded form and differently has been added by them, not omitted by Mark. Furthermore, Mark is witness to an original, briefer version, to which

11. Kümmel, Das Neue Testament, 96/The New Testament, 81. 12. Kümmel, Das Neue Testament, 96/The New Testament, 81. 13. Kümmel, Das Neue Testament, 97/The New Testament, 82. 14. Kümmel, Das Neue Testament, 97/The New Testament, 82.

4. Without Literary Gifts

85

what the others include over and above what is in it is to be regarded as an addition.’ Is not this the natural point of view? Is not the briefer, the unadorned, usually the more primitive, to which, then, other occasions later add explanation, embellishment, rounding out?15

As Werner Georg Kümmel remarks,16 Herder here exposes the problem, which form-historical scholars later related, just as he, besides the fundament of the Mark hypothesis, also opens up for the Two-Source hypothesis, although he does not count any literary interdependency between the three writings. His basic intention is to find the primitive gospel underlying the three and thereby the historical Jesus, the originator of it. Consequently, what we must read in the Gospels is the Gospel itself, which concerns the teaching, the character, and the work of Jesus, i.e., the provision he wished to make for the highest good of men. Since all these three belong together, we will treat them in relation to one another. The teaching of Jesus was simple and comprehensible to all:  God is your Father and you are all brothers one of another . . .17

This is what finds its expression in the two titles, Son of God and Son of Man. And this character was unmistakably demonstrated in his fulfilment of his work, for work it was, not just teaching. To bring a kingdom of God to the nations, in other words, a real order and disposition that would be worthy of both God and men, that was his calling; that was his purpose.18

Here Herder distinguished between ‘the so-called religion about Jesus (die sogenannte Religion an Jesum)’ and ‘a religion of Jesus (eine Religion Jesu)’ in the same manner as Lessing does between the religion of Jesus and the Christian religion,19 although Herder is not as unappreciative of the latter as Lessing is. However, he thinks that a development was indispensible. Therefore the so-called religion about Jesus must necessarily change with the passage of time into a religion of Jesus, and do so imperceptibly and irresistibly. His God, our God; his Father, our Father. Whoever contributes to bringing back the religion of Jesus from a meretricious slavery and from painfully pious Lord-Lording to that genuine Gospel of friendship and brotherliness, of convinced, spontaneous, free, glad participation

15. Kümmel, Das Neue Testament, 97/The New Testament, 82. 16. Kümmel, Das Neue Testament, 98/The New Testament, 82. 17. Kümmel, Das Neue Testament, 98/The New Testament, 83. 18. Kümmel, Das Neue Testament, 98/The New Testament, 83. 19. In the posthumously edited fragment ‘Die Religion Christi’ (1780).

86

Mogens Müller in the work and intent of Jesus as they are clearly set forth in the Gospels – he himself has taken part in Christ’s work and advanced it.20

Even if Herder is capable of seeing that each of the three synoptic gospels has its own profile, just as the Gospel of John has, in his optic, it alone witnesses that they represent three written versions of the original, oral gospel independent of each other. The material in the Gospels of Matthew and Luke, which they have besides what is in the Gospel of Mark, Herder is only able to consider as mainly embellishing additions to the oral gospel, which accordingly is rendered in its purest form in the shortest of the three. Put another way: To Herder, the most valuable in the gospels is what has received the least influence from those who gave the ‘gospel’ its written form; it is the oral gospel behind that offers access to the historical Jesus. Thus, he also lets the Gospel of John constitute the key to the existing gospels as an expression of the necessary interpretation of the historical memory in the light of the resurrection faith, even if in the first three gospels it happens on a lesser scale than in the fourth.

4.3 The Markan Hypothesis Is Supplemented by the Two-Source Hypothesis The next step was the stabilization of the Markan hypothesis in the 1830s. The most discussed work inside New Testament exegesis in this decennium indisputably was Das Leben Jesu, kritisch bearbeitet (1835–36) by David Friedrich Strauss (1808–74) which indeed eliminated the Gospel of John as a useful source for the life of Jesus. However, with regard to this author’s attitude to the other gospels Albert Schweitzer could ascertain that ‘of the Synoptic question he does not, strictly speaking, take any account’.21 In contrast, the philologist Karl Lachmann (1793–1851), in an article from 1835,22 laid the foundation for the assumption not only of Markan priority but also to the idea that the two other synoptic gospels draw on one further source. Independently of Lachmann, Christian Gottlob Wilke (1786–1854), in the magisterial work Der Urevangelist oder Eine exegetischkritische Untersuchung des Verwandtschaftsverhältniß der drei ersten Evangelien (694 pp.), comes to a similar conclusion:

20. Kümmel, Das Neue Testament, 99/The New Testament, 83. 21.  Albert Schweitzer, Geschichte der Leben-Jesu-Forschung, 2nd edn (Tübingen:  Mohr, 1913), 90. Quoted according to The Quest of the Historical Jesus (trans. W. Montgomery; Newburyport: Dover Publications, 2012), 151. Here it is also noticed that Strauss in this question is eclectic and in reality combines Griesbach’s conception of the Gospel of Mark as secondary with Schleiermacher’s diegesis theory. 22. ‘De ordine narrationum in evangeliis synopticis’, Theologische Studien und Kritiken 8 (1835), 570–90. See Kümmel, Das Neue Testament, 181–2. N. H. Palmer, ‘Lachmann’s argument’, New Testament Studies (NTS) 13 (1966–67): 368–78, contains an English translation of a part of this important article.

4. Without Literary Gifts

87

Mark is the original evangelist. It is his work that forms the basis of both the other Gospels of Matthew and Luke. This work is not a copy of an oral primal gospel, but an artistic composition. Its originator was not one of the immediate followers of Jesus, and this explains why, despite the fact that it has assumed the appearance of a historical narrative, its composition is conditioned less by historical connection than by premeditated principles.23

Thus, even if the Gospel of Mark, according to Wilke, contains the oldest accessible Jesus traditions, it is true that its structure does not so much reflect historical memory as the concepts and principles of its author. In the same year appeared a two-volume work of the professor in philosophy Christian Hermann Weisse (1801–66), Die evangelische Geschichte kritisch und philosophisch bearbeitet. Continuing in the track of Lachmann and with a critical attitude to the inconclusiveness of Strauss – whom he otherwise admired – Weisse attempted to elucidate the mutual relation of the three Synoptic Gospels. Weisse accepted the Markan hypothesis, but found in addition that the material common to the two later gospels, which did not derive from the Gospel of Mark, reflected one more common source. Thereby the two-source hypothesis saw the light of day, and to Weisse it was a most welcome means in his efforts to draw a picture of the historical Jesus. The primacy of the Gospel of Mark is clear, partly because its author ‘in his Greek Style . . . is the most Hebraizing of the evangelists’, partly because it ‘conveys the impression of a fresh naturalness and an unpretentious spontaneity’.24 According to Weisse, the other source was especially visible in the Gospel of Matthew, wherefore he also – a bit misleadingly – designated it ‘The sayings source of Matthew (die Spruchsammlung des Matthäus)’. The two later gospel authors, however, incorporate this source each in their own way. This leads us to reflect briefly on the mutual relationship of the two other Synoptics to one another in those places where they are not both dependent on Mark. We have already noted that we regard this relationship as an independent one, independent, that is to say, in the use of the common sources by each of the two, but not in the sense that each of them, throughout or for the most part, had used sources that the other had not used. It is our most certain conviction that not only Mark but also Matthew’s collection of sayings is a source common to both.25

However, 25 years were to go before the two-source hypothesis seriously gained a foothold. Contributory to this delayed breakthrough was probably the tendency criticism of Ferdinand Christian Baur and his Tübingen school. Here it was difficult

23. Kümmel, Das Neue Testament, 182/The New Testament, 148. 24. Kümmel, Das Neue Testament, 183/The New Testament, 149–50. 25. Kümmel, Das Neue Testament, 185/The New Testament, 151.

88

Mogens Müller

to give up the priority of Matthew. Thus the two-source hypothesis first became effective with the appearance in 1863 of the work Die synoptischen Evangelien. Ihr Ursprung und geschichtliche Character by Heinrich Julius Holtzmann (1832–1910). In this book came the elaboration that made the hypothesis the foundation of the Life-of-Jesus scholarship dominating in New Testament exegesis at that time. One should not be blind to the fact that it was the very need of a second reliable source in the attempt to reconstruct a trustworthy historical Jesus which to a high degree was contributing to make ‘the best’ out of the available sources. As Holtzmann was unable to connect the historical Jesus figure with an expectation of the near fall of the world, he was among the many to interpret Jesus’ preaching of the kingdom of God in a spiritualizing way giving it an ideal meaning. Thus, in the investigation from 1863, Holtzmann defends the thesis about a source older than the Gospel of Mark, that also contained a lot of saying material which, however, this author left out – this hypothesis about an ‘Urmarkus’ was later abandoned by Holtzmann – and this concept allows for letting it be the author of the Gospel of Mark who in his redaction ‘up-eschatologized’ the picture of Jesus. It is undeniable that in ‘A’ [Urmarkus] and in Mark, respectively, we are noticeably closer to the person of the Lord than in Matthew and Luke. The historically conditioned, the humanly individual, retreats least before the general and divine. On the contrary, so much of more finely applied detail, painted with earthy colors whose texture is determined by temporal and local, even individual conditions, is offered the eye of the research scholar that we can say: Nowhere does what the man Jesus was stand out so clearly as in ‘A’ and in the Gospel of Mark, respectively.26

In any case, with its assumption of an independent written source mainly consisting of sayings material, the Two-Source hypothesis made it easier to balance the Markan hypothesis in a way which made it possible to let an idealizing and psychologizing interpretation dominate the overall picture. As Albert Schweitzer could assert, ‘the victory, therefore, belonged, not to the Markan hypothesis pure and simple, but to the Markan hypothesis as psychologically interpreted by a liberal theology’.27 Already here we see the beginning of the later use of a reconstructed source as basis for the reconstruction of a Jesus alternative to the one ascertainable in the Synoptic Gospels. To Holtzmann, however, the Gospel of Mark still is the historically most reliable. Albert Schweitzer (1875–1965) fundamentally was confident in the Synoptic Gospels as trustworthy sources for the life of Jesus. In contrast to, for instance, Socrates, who was portrayed by authors who themselves were creative, ‘Jesus

26. Holtzmann, Die synoptischen Evangelien, 475. Kümmel, Das Neue Testament, 188–9/ The New Testament, 153. Holtzmann does think that the two later Synoptic gospels came into existence independently of each other. See Die synoptischen Evangelien, 163–5. 27. Cf. The Quest of the Historical Jesus, 322–3 (German edn, p. 204).

4. Without Literary Gifts

89

stands much more immediately before us, because he was depicted by simple Christians without literary gifts (literarisch unbegabten Christen)’.28 True enough, these sources suffer from a lack of inner coherence. ‘While the Synoptics are only collections of anecdotes (in the best, historical sense of the word), the Gospel of John – as it stands on record in its closing words – only professes to give a selection of the events and discourses.’29 Nevertheless, the gaps thereby arising can be filled ‘at the worst with phrases, at the best with historical imagination. There is really no other means of arriving at the order and inner connexion of the facts of the life of Jesus than the making and testing of hypotheses’.30 Even if Schweitzer does not contest the Markan hypothesis, which he thinks Holtzmann has made almost a fact, he nevertheless just as well builds on the Gospel of Matthew, then being the second-oldest gospel. For, as Schweitzer asserted many years later, ‘The accounts of both of the oldest gospels are, according to their kind, of equal worth. Matthew’s, however, is – as the more complete – the most valuable’.31 Evidently, Schweitzer here does not need any two-source hypothesis to stabilize the historical trustworthiness of the Gospel of Matthew.

4.4 Alternative Voices As it became clear with Wilke and Weisse as well as with Holtzmann, the general intention was to create a scholarly, reliable fundament under a reconstruction of a historical Jesus figure. This means that the main emphasis is on the gospels, and it also means that the Synoptic Gospels as witnesses to the sources they build upon, incidentally, may lead the interpreter back to the historical point of departure. Even if it is acknowledged that the authors behind the Synoptic Gospels also could be independently creative, this is not what primarily interests us. In this period, however, there are also more critical voices that point to the difficulties in using the Synoptic Gospels as sources for anything other than their authors’ theological interpretation of the Jesus figure. Thus, for instance Bruno Bauer (1809–82), in his Kritik der evangelischen Geschichte des Johannes (1840), had fully accepted the demonstration of Strauss that the Gospel of John is of no use as a source for the historical Jesus. On the contrary, this gospel is a piece of art – Christian art. But in his Kritik der evangelischen Geschichte der Synoptiker I–III (1841–42), where Bauer accepts the Markan hypothesis, he follows it up by

28.  Geschichte der Leben-Jesu-Forschung, 6.  The Quest of the Historical Jesus, 19. Before (14), Schweitzer has expressed gratitude ‘to the early Christians that, . . . they have handed down to us, not biographies of Jesus but only Gospels, and that therefore we possess the Idea and the Person with the minimum of historical and contemporary limitations’. 29. The Quest of the Historical Jesus, 20. 30. The Quest of the Historical Jesus, 20. 31.  ‘Vorrede zur sechsten Auflage’ of Geschichte der Leben-Jesu-Forschung (Tübingen: Mohr Siebeck, 1950), XII. My translation.

90

Mogens Müller

asking if it is also not true in the case of the Gospel of Mark that it is the product of an artist. This way there is no common, reliable tradition, but three authors, where the latter two were without originality, primarily building upon the Gospel of Mark. According to Bauer, it is even so that the author of the Gospel of Matthew is also dependent on the Gospel of Luke. In his description of Bauer’s position, Schweitzer puts it as a question when he writes: But, if so, who is to assure us that this Gospel history, with its assertion of the Messiahship of Jesus, was already a matter of common knowledge before it was fixed in writing, and did not first become known in literary form? In the latter case, one man could have created out of general ideas the definite historical tradition in which these ideas are embodied.32

Thirty years later, Gustav Volkmar (1809–93), in a commentary on the Gospel of Mark, claimed that this gospel essentially consisted of a Pauline theology in the shape of emblems (‘Sinnbilder’) and had a ‘parabolic’ nature. Volkmar’s critics were the first who introduced the designation ‘allegory’ in describing his position.33 Also, the epoch-making investigation by William Wrede (1859–1906), Das Messiasgeheimnis in den Evangelien from 1901, should be mentioned in this connection.34 In a section with the heading ‘Mark as an Author’,35 Wrede contests the idea that the author of the Gospel of Mark offers anything like a trustworthy

32.  The Quest of the Historical Jesus, 227. Schweitzer offers a whole chapter (21 pp.) on Bauer and concludes (256): ‘For us the great men are not those who solved the problems, but those who discovered them. Bauer’s “Criticism of the Gospel History” is worth a good dozen Lives of Jesus, because his work, as we are only now coming to recognize after half a century, is the ablest and most complete collection of the difficulties of the Life of Jesus which is anywhere to be found.’ About fifty years later, this fascination is not affecting Kümmel, who in his history of interpretation banishes Bauer to a single footnote (Das Neue Testament, 557 n. 351/The New Testament, 447 n. 367), because to him criticism here goes too far. 33.  For details, see Anne Vig Skoven, ‘Mark as Allegorical Rewriting of Paul:  Gustav Volkmar’s Understanding of the Gospel of Mark’, in Mark and Paul. Comparative Essays Part II. For and against Pauline Influence on Mark, ed. Eve-Marie Becker, Troels Engberg-Pedersen and Mogens Müller, Beihefte zur Zeitschrift für die neutestamentliche Wissenschaft (BZNW) 199 (Berlin: de Gruyter, 2014), 1–27. 34.  With the subtitle:  Zugleich ein Beitrag zum Verständnis des Markusevangeliums (Göttingen:  Vandenhoeck & Ruprecht, 1901). Quoted according to The Messianic Secret (trans. J.  C. G.  Greig; Cambridge/London:  James Clarke, 1971). As predecessors Wrede (280–6), besides Bruno Bauer and the Dutchman Hoekstra, especially mentions Volkmar (284): ‘Without a doubt Volkmar’s book is the most perceptive and shrewd, and to my mind altogether the most important, that we possess on Mark.’ 35. Messiasgeheimnis, 129–45.

4. Without Literary Gifts

91

picture of the historical Jesus. ‘It must frankly be said that Mark no longer has a real view of the historical life of Jesus’: It therefore remains true to say that as a whole the Gospel no longer offers a historical view of the real life of Jesus. Only pale residues of such a view have passed over into what is a suprahistorical view of faith. In this sense the Gospel of Mark belongs to the history of dogma.36

That this does not mean that Wrede estimates the creator of this gospel especially as an author is seen from the rest of the chapter. However, he has managed to establish a new and ‘dogmatic’ frame around the remnants of ‘history’, which his gospel represents.

4.5 Continuations of the Two-Source Hypothesis The next innovation inside gospel scholarship, the introduction of Form History, meant – although resting upon other premises – a renewed interest in the sources of the Synoptic Gospels, namely the anonymous, oral formation of tradition leading up to it, but which also in a high degree had stamped the material included. In this connection the Two-Source hypothesis could be a part of the argument for assuming that the authors of the three first gospels were not creative in any original way, but by and large were only capable of working with existing traditions. To the three ‘fathers’ of the form-historical approach inside gospel scholarship it thus was a fact that the Synoptic Gospels consisted of popular literature (‘Kleinliteratur’) without any literary ambitions,37 their authors being essentially compilers. Thus, Martin Dibelius (1883–1947), in his Die Formgeschichte des Evangeliums (1919), declares outright: Without doubt these [the Synoptic Gospels] are unliterary writings. They should not and cannot be compared with ‘literary’ works. Nevertheless they are certainly not private notes but are designed for a definite publicity even if it be only humble. But the literary character of these books shows certain marks which differentiate them from other primitive Christian writings. The literary understanding of the synoptics begins with the recognition that they are collections of material. The composers are only to the smallest extent authors. They are principally collectors, vehicles of tradition, editors. Before all else their labour consists in handing down, grouping, and working over the material which has come to

36. Messiasgeheimnis, 129, 131. In the translation, the page numbers are the same. 37. Both with regard to Herder and to the representatives of Form Criticism, one should ascertain that the label ‘popular literature’ (Kleinliteratur) and the like is not deprecatory. It only means that it is not artificial and determined by certain literary conventions. Thus, it is more in accordance with its subject.

92

Mogens Müller them. Therein also their religious presentation of the material comes to essential expression. Thus they make their influence felt with much less independence than, say, the composer of the Book of Acts. The latter is indeed himself an evangelist, but he is much more bound by his material in the Gospel of St. Luke than in the Acts of the Apostles. Here he acts as an author, but in the Gospel rather as a collector and editor. For this reason St. Luke more than the other synoptics shows the strongest literary character. Thereby it can be estimated in how lowly a degree after all St. Mark and St. Matthew may pass as authors.38

Practically the same view is held by Karl Ludwig Schmidt (1891–1956) in an article, ‘Die Stellung der Evangelien in der allgemeine Literaturgeschichte’, where he writes:39 The Gospel is not at all high literature (Hochliteratur), but popular literature (Kleinliteratur), not the outcome of an individual author, but popular books, not biography, but cultic books. In spite of his apparent qualities as an author, Luke neither could nor would create a biography, and the Fourth Gospel that, to a certain degree, is a personal confession, has more received tradition behind it than we are able to discern and is, beyond the personal, carried by a confession community. The Gospels are not to be found in a certain branch of history of literature. (My translation)40

The third great name of Form History, Rudolf Bultmann (1884–1976) takes an important further step in his Geschichte der synoptischen Tradition of 1921. Working with the history of the Synoptic tradition, he is also aware of the redactional work of each of the authors. To begin with – Bultmann says – it is the sociological, not the aesthetic factors that are decisive. ‘But in the literature of primitive Christianity, which is essentially “popular” [“Kleinliteratur”] (Dibelius) in kind, 38.  Martin Dibelius, Die Formgeschichte des Evangeliums, 2nd edn (Tübingen: Mohr Siebeck, 1919, 1933), 2–3. Quoted according to From Tradition to Gospel (trans. Lee Woolf; Cambridge: James Clarke, 1971), 2–3. 39.  The article was published in ΕΥΧΑΡΙΣΤΗΡΙΟΝ. Studien zur Religion und Literatur des Alten und neuen Testaments. FS Hermann Gunkel, ed. Hans Schmidt, Forschungen zur Religion und Literatur des Alten und Neuen Testaments (FRLANT) 36/2 (Göttingen: Vandenhoeck & Ruprecht, 1923), 50–134. 40.  The German reads:  ‘Das Evangelium ist von Haus aus nicht Hochliteratur, sondern Kleinliteratur, nicht individuelle Schriftstellerleistung, sondern Volksbuch, nicht Biographie, sondern Kultlegende. Lukas hat trotz seiner wohl vorhandenen schriftstellerischen Fähigkeiten eine Biographie nicht schaffen können oder wollen, und das vierte Evangelium, das in gewissem Sinne eine persönliche Bekennerschrift ist, hat mehr übernommene Überlieferung hinter sich, als wir feststellen können, und ist über das Persönliche hinaus von einer Bekennergemeinschaft getragen. Auf einem bestimmten Strang der Literaturgeschichte sind die Evangelien nicht zu finden.’

4. Without Literary Gifts

93

this development had not yet taken place, and it is only possible to understand its forms and categories in connection with their “life situation”, i.e. the influences at work in the community.’41 But along the way Bultmann also developed the ‘aesthetic’ understanding of each of the authors as creative, just as Wrede and Dibelius had also done. Thus in conclusion of his investigation he offers relatively substantial characteristics of ‘Mark’, ‘Matthew’ and ‘Luke’, respectively, whereby he points forward to the coming redaction criticism and reception criticism.42 It was not Bultmann’s main interest to create a basis for reconstructing the historical Jesus – as a representative of dialectical theology he did not think that theology could be based on ‘historical’ truths. Other scholars, however, continued the efforts in detecting historically reliable sources behind the gospels.43 A contribution here, which gained great influence especially in the Anglo-Saxon world, was The Four Gospels from 1924 by the Oxford professor Burnett Hillman Streeter (1874–1937). In this book, Streeter proposed his ‘four-sources hypothesis’, claiming that the Gospel of Luke had a predecessor, Proto-Luke. To be sure, Q is the oldest ‘source’, a source that Streeter also accepts as older than the Gospel of Mark. But besides the Gospel of Mark and independent of it existed a Proto-Luke, and in some instances it contained traditions nearer to the historical truth than this Gospel. ‘Neither Mark nor Proto-Luke is infallible; but as historical authorities they should probably be regarded as on the whole of approximately equal value. But, if so, this means that far more weight will have to be given by the historian in the future to the Third Gospel, and in particular to those portions of it which are particular to itself.’44 According to Streeter the conclusion therefore must be the following: So far as historical detail is concerned, Mark and Luke are more to be relied on than Matthew; and where Mark and Luke conflict, Mark is often more to

41.  Geschichte der synoptischen Tradition, FRLANT 29 (Göttingen:  Vandenhoeck & Ruprecht, 1931 [2nd edn]  =  1964 [6th edn]), 5.  Quoted according to The History of the Synoptic Tradition (trans. John Marsh; Peabody, MA: Hendricks, rev. edn 1963), 4. 42. See Geschichte der synoptischen Tradition, 362–92; History, 337–67. 43.  Thus, my conservative predecessor, Frederik Torm (1870–1953), in Hermeneutik des Neuen Testaments (Göttingen:  Vandenhoeck & Ruprecht, 1930), 155, reminds of Schweitzer’s characteristic of the gospels authors, when he, summarizing the results of form history, writes (the words are spaced by Torm): ‘Von der Form aus erkennen wir . . ., daß die Verfasser unliterarische Männer gewesen sind, die nur das einzige Interesse hatten, die übernommene Überlieferung genau und sorgfältig weiterzugeben.’ Cf. 172, where the anonymity with regard to the Synoptic Gospels, according to Torm, is easy to explain:  ‘Die Verfasser wollten die Tatsache unterstreichen, daß sie nur Referenten waren; sie wünschten selber vergessen zu werden; auf den Inhalt der überlieferten Berichte kam alles an.’ 44.  The Four Gospels. A Study of Origins. Treating of the Manuscript Tradition, Sources, Authorship, & Dates (London: Macmillan 1924, 1926 [2nd edn]), 222.

94

Mogens Müller be followed. But as regards the teaching of Christ, much that occurs in a single Gospel is as likely to be genuine as what occurs in two or in all three.45

4.6 Conclusion As mentioned at the beginning we can today observe two main tendencies in gospel scholarship with regard to the value of the Synoptic Gospels as sources. On the one hand, Michael Goulder (1927–2010) could claim that the author of the Gospel of Matthew had mostly himself created all the material in his writing, which he had not taken over from the Gospel of Mark.46 On the other, Ulrich Luz can argue that although this author could make use of creative rewriting of existing traditions, he hardly can be supposed to have created new traditions – thus, according to Luz, the stories about Jesus’ birth and childhood consisted of material which he took over.47 Besides, after the linguistic turn in the 1970s, we have narrative criticism looking at the gospels as writings in their own right and without seeking to detect their eventual sources all the time. Even where scholars were not willing to abandon the historical question about the referentiality of the material which their authors reworked, this approach certainly has strengthened the apprehension of gospels as coherent compositions. Generally today not only the Gospel of John but also the Synoptic Gospels are valued as literary products. The apprehension of their originators as vehicles of tradition and redactors is practically given up. Instead, we have an understanding of their authors as creative interpreters of tradition. The creativity and independence that we observe in the writers behind the Gospel of Matthew and the Gospel of Luke in their use of the Gospel of Mark, we accordingly have to assume with regard to the author of this the earliest gospel. Thus, there is good reason to consider the author of the Gospel of Mark the able creator of the gospel genre.48

45. The Four Gospels, 270. Folker Siegert and Vadim Wittkowsky, Von der Zwei- zur VierQuellen-Hypothese. Vorschlag für ein vollständiges Stemma der Evangelienüberlieferungen, Bibelstudien 15 (Münster:  LIT Verlag, 2014), appearing 90 years after Streeter’s investigation, consists according to its authors (p. 15 n. 2) of ‘ein stark verändertes Echo’ of Streeter’s book. However, here it is not a Proto-Luke, but a Proto-Matthew, which is postulated. 46.  See Michael Goulder, Midrash and Lection in Matthew (London:  SPCK, 1974), 4, where he describes the basic assumption in his investigation also in this way: ‘Or, to put it in other words, I shall consider the grounds for thinking that Matthew was writing a midrashic expansion of Mark: that the new teaching is his teaching, that the new poetry is his poetry, that the new parables are his parables.’ 47. See, for instance, Ulrich Luz, ‘Fiktivität und Traditionstreue im Matthäusevangelium im Lichte griechischer Literatur’, ZNW 84 (1993):  153–77. Cf. Mogens Müller, ‘The New Testament gospels as Biblical rewritings’, Studia Theologica 68 (2014): 21–40, esp. 22. 48.  Today it is not difficult to find scholars underscoring the originality of the author behind the Gospel of Mark.

4. Without Literary Gifts

95

One of the results of Form History is that the Synoptic Gospels in large part consist of traditions which only seemingly are combined to something like a continuous story.49 Already this, however, is a literary achievement. The author of the Gospel of Mark here acts like an anonymous ‘biblical’ author, who in his writing extends the holy story, which is even clearer with regard to the other two synoptics. The question arises, then, whether the only large component, namely the passion story, already existed for this author as a tradition, or if he himself essentially is its creator. A good many arguments speak for the latter assumption.50 This also confirms the impression of this author as independent and creative and invites us to follow Volkmar’s idea about the Gospel of Mark as a narrative and allegorizing edition of Paul’s theology and preaching.51 If the characteristic of the gospel genre as ‘biblical’ is correct, it also means that these writings are meant for use in the congregation. They address all Christbelievers. Because of this, it is also obvious to look for the model in the biblical story and here especially the legends in 1–2 Kings about Elijah and Elisha. They do not pretend to be historical writing, but biblical story where, at the end of the day, it was the intention of the preaching which decided or even created the story. This pertains, for instance, to the stories about the birth and childhood of Jesus in the Gospels of Matthew and Luke, where the two authors seemingly each freely have created his history, the latter maybe even as a conscious alternative to the former. That the holy writings of Judaism, and of course in the shape of the Septuagint, did serve as the literary model becomes especially clear in the Lukan writings.52 Apart from this ‘Luke’, who really seems to have had literary ambitions as a 49.  Fundamental here was Karl Ludwig Schmidt, Der Rahmen der Geschichte Jesu. Literarkritische Untersuchungen zur ältesten Jesusüberlieferung (Berlin: Trowitzsch & Sohn, 1919; repr. Darmstadt: Wissenschaftliche Buchgesellschaft, 1969). See esp. the conclusion on p. 317. 50. This is the main thesis in Niels Willert’s doctoral dissertation, Den korsfæstede konge. Kristologi og discipelbillede i Markusevangeliets passionsfortælling, Bibel og historie 20–1 (Århus: Aarhus Universitetsforlag, 1997). 51.  Cf. William R. Telford’s ‘prophetic’ conclusion on his treatment of the problem ‘Mark and Paul’ in The Theology of the Gospel of Mark, New Testament Theology (Cambridge:  Cambridge University Press, 1999), 164–9:  ‘With the development . . . of narrative-critical tools and an increasing sensitivity on the part of scholars to the nuances of narrative theology, Volkmar’s original suggestion that Mark’s Gospel is an allegorical presentation of Pauline teaching in the form of a narrative may be due, therefore, for a comeback.’ Cf. further Henrik Tronier, ‘Philonic Allegory in Mark’, in Philosophy at the Roots of Christianity, ed. Troels Engberg-Pedersen and Henrik Tronier (Copenhagen: The Faculty of Theology Biblical Studies Section, 2006), 9–48, and the contributions, especially Anne Vig Skoven’s, in the volume Mark and Paul mentioned above in note 32. 52.  See Mogens Müller, ‘Die Lukasschriften und die Septuaginta’, in Die Septuaginta – Entstehung, Sprache, Geschichte, ed. Siegfried Kreuzer, Martin Meiser and Marcus Sigismund, WUNT 286 (Tübingen:  Mohr Siebeck, 2012), 465–79, esp. 468–71 about the ‘pioneer’ with regard to the Septuagint and the Lukan writings, the Swedish scholar Albert Wifstrand (1901–64).

96

Mogens Müller

Christian ‘Josephus’, the model of these authors has not been the biographies from the surrounding world, even if the gospels also show several points of resemblance with this genre. If one, instead, imagines that they have worked inside the genre or interpretation strategy, which has been labelled ‘rewritten Scripture’, this much better explains the freedom they creatively have exercised in rewriting, adding and omitting. If this, finally, is connected to a late dating of the Lukan writings to sometime between 120 and 140 AD,53 it removes the essential presupposition for the twosource hypothesis. If the Gospels of Matthew and Luke were written within an interval of three or four decades, is it not likely that the later of the two would also have known and used this predecessor? Obviously, however, such a late dating builds on a cumulative probability. In contrast to the dating of Paul’s letters to the beginning of the fifties, which should be safe, we first have secure external evidence for the gospels from about the middle of the second century. Thus the question in the actual situation of scholarship is whether the acknowledgement of the originality of the single gospel author is not contributing to making probable that the Two-Source hypothesis is unnecessary, and that no relevant source has been lost. Thus, we should perceive the four canonical gospels as stations in a process, where among the ‘many’ mentioned in Luke 1.1 presumably are the three others, or at least two of them.

53.  Thus, for instance, Niels Hyldahl, ‘Über die Abfassungszeit des lukanischen Doppelwerks’, in Frühes Christentum und Religionsgeschichtliche Schule. FS Gerd Lüdemann, ed. Martina Janssen, F. Stanley Jones and Jürgen Wehnert, Novum Testamentum et Orbis Antiquus/Studien zur Umwelt des Neuen Testaments (NTOA/ SUNT) 95 (Göttingen:  Vandenhoeck & Ruprecht, 2010), 75–82. Cf. Mogens Müller, ‘Acts as Biblical Rewriting of the Gospels and Paul’s Letters’, in Luke’s Literary Creativity, ed. Mogens Müller and Jesper Tang Nielsen, Library of New Testament Studies (LNTS) 550 (London: Bloomsbury T&T Clark, 2016), 96–117.

Chapter 5 Q A N D T H E L O G IA : O N T H E D I S C OV E RY A N D M A R G I NA L I Z I N G O F P . O X Y. 1 Francis Watson

First published in 1955, Austin Farrer’s famous article ‘On Dispensing with Q’ was directed against an orthodoxy associated especially with B. H. Streeter, an older Oxford colleague who had died in 1937 but whose work on gospel origins had already attained virtual canonical status.1 In his opening pages, Farrer makes deft use of Ockham’s Razor against those he calls ‘Streeterians’.2 Entia non sunt multiplicanda praeter necessitatem:  in explaining the origin of the ‘double tradition’ there is no need to postulate an additional entity (Q) if Luke could have derived this material from Matthew. If Luke’s use of Matthew can be shown to be plausible, Q disappears automatically at least in its dominant modern form. I shall suggest here that this claim – that Luke’s use of Matthew entails the elimination of Q – remains true, yet needs to be refined. Farrer overlooks the fact that Q has more than one function. It not only accounts for the non-Markan overlap between Matthew and Luke but also helps to bridge the gap between sayings of Jesus uttered in around 30 CE and their incorporation into one or more written gospels half a century later. Fifty years is a long time to preserve an utterance by purely oral means, especially in the absence of any obvious institutional mechanism for controlling transmission. In dispensing with Q, must we either place our faith in oral tradition or assume with Michael Goulder that many of the sayings and parables attributed to Jesus were actually composed by Matthew?3 If, however,

1. Austin Farrer, ‘On Dispensing with Q’, in Studies in the Gospels in Memory of R. H. Lightfoot, ed. D. E. Nineham (Oxford: Blackwell, 1955), 55–88; Burnett Hillman Streeter, The Four Gospels: A Study of Origins (London: Macmillan, 1924 [1st edn], 1930 [4th edn]). 2. ‘On Dispensing with Q’, 55–7. 3.  Michael Goulder, Midrash and Lection in Matthew (London:  SPCK, 1974); Micheal Goulder, Luke:  A New Paradigm, 2 vols, Journal for the Study of the New Testament Supplement (JSNTSup) 20 (Sheffield: Sheffield Academic Press, 1989).

98

Francis Watson

Farrer’s hypothesis is correct and Q in its conventional form is eliminated, it remains entirely possible for Matthew to have drawn on written collections of Jesus’ sayings which, as such, would in a certain sense be Q-like although they would not be Q. Elsewhere I have argued that there is considerable evidence for the existence of a Sayings Collection genre, of which the Gospel of Thomas is a late but fully extant example.4 Here, I shall show that this possibility was already discussed in late nineteenth-century Oxford, during the lifetime of the long-running seminar on the synoptic problem that was to inspire the work of Streeter. The discussion was occasioned by the publication of an important fragment of a noncanonical gospel, and I will suggest that the hypothetical Q document was preferred on the conservative grounds that it remains safely within the limits of the New Testament canon. This long-forgotten episode in the history of scholarship has an interest of its own, but it is also instructive. It brings to light an often unrecognized factor in the construction of the so-called Synoptic Problem, the assumption that a canonical boundary created towards the end of the second century corresponds to historical realities towards the end of the first. The Synoptic Problem seeks a purely intracanonical solution to the historical problem of the origin and development of early gospel literature. Projecting the canonical boundary back into the last decades of the first century, it assumes that texts later differentiated by the names of Matthew, Mark and Luke already constituted a self-contained collection as soon as they came into being. These texts are, of course, interrelated. It is entirely appropriate to ask how that interrelatedness is to be explained, seeking a literary solution entailing priority and dependence rather than appealing to the vagaries of oral tradition. What is problematic is the assumption that the three interrelated texts are significantly related only to each other, and that they can be considered without reference to other early gospel literature – as though the Gospel of Thomas, the Gospel of Peter, the Egerton gospel, or the Marcionite gospel were, from the very beginning, different in kind from the texts that later became canonical. In this field as in others, canonical or noncanonical status is determined by a text’s reception

4.  Francis Watson, Gospel Writing:  A Canonical Perspective (Grand Rapids, MI:  Eerdmans, 2013), 217–85. In my view, Thomas in its present form shows unmistakeable signs of synoptic influence  – as argued persuasively by Simon Gathercole, The Composition of the Gospel of Thomas: Original Language and Influences, Society for New Testament Studies Monograph Series (SNTSMS) 151 (Cambridge:  Cambridge University Press, 2012), 127–224; Mark Goodacre, Thomas and the Gospels:  The Case for Thomas’s Familiarity with the Synoptics (Grand Rapids, MI: Eerdmans, 2012). Contra Gathercole and Goodacre, however, I consider the format of Thomas – juxtaposed sayings with a simple introductory formula – to have primitive roots which make this relatively late work relevant for synoptic origins.

5. Q and the Logia

99

by user-communities and not by inherent characteristics or conformity to preestablished criteria.5 Within this context of early gospel production Q is ambiguous and plays a variety of roles. It can be presented as an essentially noncanonical text that opposes the dominant emphasis on death and resurrection, miracle and apocalypse, by presenting Jesus as a teacher of wisdom.6 Analysed into various successive layers, it can provide an account of written Jesus tradition prior to Mark.7 Yet Q is also a canonical text. There is nothing in Q that is not already in Matthew and Luke. Unlike genuinely extra-canonical gospels, Q contains nothing that is not already familiar. A  radical theological stance can be assigned to it only on the basis of an argument from silence, and such arguments can always be resisted.8 There is, then, a radical Q and a conservative Q, and it is this ability to appeal to all shades of opinion that explains Q’s extraordinarily successful twentieth-century career. In addition to its role in resolving the Synoptic Problem, Q also generates the concept of ‘alternating primitivity’ which seems so indispensable for conventional 5.  This account of gospel origins is in some respects similar to the one developed most influentially by Helmut Koester, in his Ancient Christian Gospels: Their History and Development (Philadelphia:  TPI, 1990) and From Jesus to the Gospels: Interpreting the New Testament in its Context (Minneapolis: Fortress, 2007). Koester writes: ‘In the vast treasure of non-canonical gospel literature, there are at least some writings that have not found their rightful place in the history of this literary genre’, but which are ‘perhaps at least as old and as valuable as the canonical gospels as sources for the earliest developments of the traditions about Jesus’ (From Jesus to the Gospels, 23). Koester also rightly locates the Gospel of Thomas within a broader Sayings Collection genre, building on his early work on Synoptische Überlieferungen bei den apostolischen Vätern (Berlin: Akademie Verlag, 1957). The major difference between Koester’s position and the one assumed here (and developed elsewhere) is that he presupposes the existence of Q whereas I consider that later evidence allows us only to postulate a primitive Sayings Collection genre, but not to reconstruct any individual instance of it. In addition, a noncanonical gospel does not have to be ‘at least as old . . . as the canonical gospels’ in order to be valuable for our understanding of gospel origins. The canonical gospels may themselves span a period of three or four decades, and a noncanonical gospel composed some years later indicates only that gospel literature continued to proliferate, not that it can be consigned to ‘apocryphal’ status. 6.  This is already the case in Adolf von Harnack, Sprüche und Reden Jesu: Die zweite Quelle des Matthäus und Lukas (Leipzig: J. C. Hinrichs, 1907). For updated versions of this use of Q, see John Dominic Crossan, The Birth of Christianity (New York:  Harper Collins, 1998), and John S. Kloppenborg, Q The Earliest Gospel:  An Introduction to the Original Stories and Sayings of Jesus (Louisville, KY: Westminster John Knox, 2008). 7.  John Kloppenborg, The Formation of Q: Trajectories in Ancient Wisdom Traditions, 2nd edn (Harrisburg, PA: Trinity Press International, 1999). 8. For example in Philip Jenkins’ polemic against the Jesus Seminar, Hidden Gospels: How the Search for Jesus Lost its Way (New York: Oxford University Press, 2001), 54–81.

100

Francis Watson

historical Jesus scholarship.9 Given that, as Farrer rightly emphasized, the very existence of Q hangs on the slender thread of Matthew and Luke’s presumed independence of each other, Q’s 20th century status as critical orthodoxy is a remarkable achievement. At least in its British context, it was a conservative, intracanonical Q that established itself from the 1890s onwards. The Q hypothesis was used to declare the evidence of newly recovered noncanonical texts irrelevant to questions of gospel origins. Thus Streeter’s magnum opus of 1924, subtitled A Study in Origins, limits its horizons to The Four Gospels with a decisiveness that Irenaeus himself would surely have applauded.

5.1 Discovering the Logia In the summer of 1897 a newly discovered Greek fragment was presented to the world under the provocative title, ΛΟΓΙΑ ΙΗΣΟΥ: Sayings of our Lord.10 It became known as P.Oxy.1, and belongs to a text later recovered in a complete Coptic translation where it is entitled peuaggelion pkata ;wmac, The Gospel according to

9.  For a critique of this concept see Mark Goodacre, The Case against Q (Harrisburg, PA: Trinity Press International, 2002), 61–6. Alternating primitivity is classically articulated by B.  H. Streeter:  ‘Sometimes it is Matthew, sometimes it is Luke, who gives a saying in what is clearly the more original form. This is explicable if both are drawing from the same source, each making slight modifications of his own; it is not so if either is dependent on the other’ (Four Gospels, 183). Streeter here establishes the convention that ‘alternating primitivity’ can be regarded as given and does not need to be demonstrated. His argument is reproduced almost verbatim by W. D. Davies and Dale Allison: ‘ [S]ometimes it is the First Evangelist who seems to preserve the more original form of a saying appearing in the double tradition, at other times it is Luke. This is inexplicable if one evangelist is following the other’ (A Critical and Exegetical Commentary on the Gospel according to Saint Matthew, 3 vols [Edinburgh: T&T Clark, 1988–97]), 1.116. For comprehensive evidence of alternating primitivity at work, see The Critical Edition of Q, ed. James M. Robinson, Paul Hoffmann, and John S.  Kloppenborg, Hermeneia (Minneapolis:  Augsburg Fortress, 2000). Here the format (αβγ) represents Matthean agreement with Q against Luke, while [αβγ] represents Lukan agreement with Q against Matthew. 10.  Bernard Grenfell and Arthur S. Hunt, ΛΟΓΙΑ ΙΗΣΟΥ:  Sayings of our Lord (London:  Henry Frowde, 1897) and The Oxyrhynchus Papyri, vol. 1 (London:  Egypt Exploration Fund, 1898), 1–3. For modern editions co-ordinated with the Coptic Gospel of Thomas, see Harold W. Attridge, ‘Appendix: The Greek Fragments’, in Nag Hammadi Codex II, 2–7, ed. Bentley Layton, vol. 1 (Leiden: Brill, 1989), 95–128;118–21; Andrew Bernhard, Other Early Christian Gospels:  A Critical Edition of the Surviving Greek Manuscripts (London:  T&T Clark, 2007), 34–41. I  here follow the enumeration of the sayings in the editio princeps.

5. Q and the Logia

101

Thomas.11 The new discovery was published with an Introduction and Commentary that assigned it an indirect role in canonical gospel origins. Scholarly rejection of any such role was intimately bound up with the development of the Q hypothesis, according to which the most important lost text is to be extracted not from the sands of Egypt but from within the canonical gospels themselves. In more recent scholarship, P.Oxy.1 has long been subsumed into the study of the Coptic Gospel of Thomas.12 It will be instructive to consider it as it appeared to its first modern readers at the end of the nineteenth century. The Logia fragment consists in a single piece of papyrus measuring around 15 x 10 cms, although broken off at the bottom.13 Its editors date it to the early third century. A page number is given at the top right of the verso (ΙΑ = 11), suggesting a text of similar or identical extent to the Coptic Gospel of Thomas and to P.Oxy.654 (if it continued in step with its Coptic counterpart). Text on both verso and recto confirms that this book – unlike P.Oxy.654 – was in codex rather than roll format. The papyrus contains eight sayings of Jesus, although only faint traces of the fourth and eighth of these remain. Logia 2, 6, 7 are intact, however, as are substantial parts of Logia 1, 3 and 5. A standardized introductory formula, λέγει Ἰ(ησοῦ)ς, serves both to isolate the sayings from each other and to place an unusual emphasis on the person of the speaker. The formula is extant in four sayings (2, 3, 6, 7) and may be presumed for the others. Here and elsewhere, conventional Christian scribal abbreviations are employed: Ἰ(ησοῦ)ς (lines 5, 11, 30, 36), θ(εο)ῦ (line 8: ‘kingdom of God’), π(ατέ)ρα (line 11: ‘you will see the Father’), ἀν(θρώπ)ων (line 19: ‘the sons of men’) and, less usually, π(ατ)ρίδι (line 32: ‘his own country’).14 11.  The convention of locating the title at the end of the work but not at the beginning no doubt derives from the translator’s Greek exemplar. P.Oxy.654 (= GThos 1–7) similarly appears to have lacked a superscription, to judge from the blank papyrus fragments above its opening line. 12. For a discussion of the relationship of the Greek fragments P.Oxy.1, 654, 655 to the Coptic GThos, see Simon Gathercole, The Gospel of Thomas: Introduction and Commentary (Leiden: Brill, 2014), 14–24, emphasizing the relative closeness of the Greek and Coptic versions. Of the three Greek manuscripts, P.Oxy.1 is closest to the Coptic text. 13. Cf. Grenfell and Hunt, Sayings, 6–7. 14. On the so-called nomina sacra, see Ludwig Traube, Nomina Sacra: Versuch einer Geschichte der christlichen Kürzung (Munich: Beck, 1907); C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University Press, 1979), 26– 48; Philip Comfort, Encountering the Manuscripts: An Introduction to New Testament Paleography and Textual Criticism (Nashville, TN: Broadman & Holman, 2005), 199–253; Larry Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids, MI: Eerdmans, 2006), 95–134. The appropriateness of Traube’s expression nomina sacra is questioned by Christopher Tuckett, ‘ “Nomina Sacra”: Yes and No?’, in The Biblical Canons, ed. Jean-Marie Auwers and Henk Jan de Jonge (Leuven: Peeters, 2003), 431–58, and defended by Jane Heath, ‘Nomina Sacra and Sacra Memoria before the Monastic Age’, Journal of Theological Studies (JTS) 61 (2010): 516–49. Writing prior to Traube, Grenfell and Hunt refer simply to ‘the regular contractions used in biblical MSS’ (Sayings, 7).

102

Francis Watson

The surviving part of Logion 1 corresponds closely to the end of Luke 6.42: ‘. . . and then you will see clearly to cast out the speck that is in your brother’s eye.’ Logia 2 and 3 are entirely new.15 Jesus says: Unless you fast to the world, you will not find the kingdom of God.16 Unless you observe the sabbath, you will not see the Father. Jesus says: I stood in the midst of the world and I appeared to them in flesh, and I found them all drunk and no one did I find thirsty among them, and my soul grieves over the sons of men because they are blind in their hearts . . .

Owing to the loss of the lower part of the page, only the final word of Logion 4 (πτωχεία) survives at the top of the recto side. The first few lines of the recto are themselves seriously damaged, and Logion 5 exists only in tantalizing fragments until its remarkable conclusion: ‘Raise the stone and there you will find me, split the wood and I am there.’ Logia 6 and 7 have Lukan and Matthean affinities respectively, although their wording is distinctive: Jesus says: A prophet is not acceptable in his own country, nor does a doctor work cures for those who know him. Jesus says: A city built on top of a high mountain and established can neither fall not be hid.

In sum, there are at least three new sayings (2 fasting/sabbath, 3 drunken world, 5 stone/wood) to complement three familiar ones (1 removing the speck, 6 prophet/ doctor, 7 city on hilltop), of which the last two contain significant novel elements. The Logia fragment was published under the auspices of the Egypt Exploration Fund, founded in 1882 ‘for the purpose of conducting excavations in the Delta, which up to this time has been very rarely visited by travellers’. (The year 1882 also marked the beginning of British rule in Egypt.) In 1895 the Fund decided to include the Graeco-Roman period in its scope, and three young Oxford-based archaeologists  – David Hogarth, Bernard Grenfell and Arthur Hunt  – were sent to prospect for promising sites in the Fayum region.17 Grenfell and Hunt returned in the winter of 1896–97 to excavate at Behneseh, site of the ancient city of Oxyrhynchus, around 200 miles south of Alexandria. After a slow start, their 15.  For analysis of these sayings in their Greek and Coptic forms (GThos 26–28), see Gathercole, Thomas, 320–34. 16. The Coptic refers only to ‘finding the kingdom’ (GThos 27.1), and ‘of God’ is uncertain in the Greek, where it is unclear which letter or letters preceded the upsilon with the supralinear stroke that indicates a conventional contraction. There appears to be no satisfactory alternative to ‘kingdom of God’. 17.  See D. W. Rathbone, ‘Grenfell and Hunt at Oxyrhynchus and in the Fayum’, in The Egypt Exploration Society:  The Early Years, occasional papers ed. Patricia Spencer (London: Egypt Exploration Society, 2007), 195–229.

5. Q and the Logia

103

efforts met with spectacular success. In their own words, written after their return to Oxford in the spring of 1897: The ancient cemetery, to which for various reasons the first three weeks’ work was devoted, proved on the whole unproductive; but in the rubbish-heaps of the town were found large quantities of papyri, chiefly Greek, ranging in date from the first to the eighth century, and embracing every variety of subject.18

The first edited collection of these texts was published the following year, in 1898,19 but one papyrus fragment seemed important enough to edit and publish immediately: The document in question is a leaf from a papyrus book containing a collection of Logia or Sayings of our Lord, of which some, though presenting several novel features, are familiar, while othes are wholly new. It was found at the very beginning of our work upon the town, in a mound which produced a great number of papyri belonging to the first three centuries of our era, those in the immediate vicinity of our fragment belonging to the second and third century.20

To this laconic account of the discovery Grenfell later added vivid circumstantial detail in the pages of the October 1897 issue of McClure’s Magazine, an American illustrated periodical, founded in 1893, which offered its readers a miscellany of fiction, history and current affairs.21 Grenfell recounts how, the previous 18. Grenfell and Hunt, Sayings, 5. 19. Oxyrhynchus Papyri Part I (London: Egypt Exploration Fund, 1898). 20. Grenfell and Hunt, Sayings, 6. 21. Bernard P. Grenfell, ‘The Oldest Record of Christ’s Life: The First Complete Account of the Recent Finding of the “Sayings of our Lord” ’, McClure’s Magazine, vol. IX, no. 6 (Oct. 1897), 1022–30. Grenfell’s article follows an account of ‘An Elephant Round-up in Siam’ and a short story entitled ‘The Turf Cutters’, and it is followed by a Civil War-related piece on ‘The Making of a Regiment: What a Service of Seven Months did for a Troop of Raw Volunteers’. Grenfell’s contribution is highlighted on the magazine’s front cover, above an image of a young woman carrying an outsize apple under her arm to signify the arrival of Fall. The title, ‘The Oldest Record of Christ’s Life’ relates primarily to the Introduction to Grenfell’s article provided by Frederick Kenyon (Keeper of Manuscripts in the British Museum, later its Director), who presents the Logia as ‘the earliest, and far the earliest, record of the words spoken by our Lord Jesus Christ upon earth’ (1022). ‘Some of these sayings are certainly authentic, since they are preserved in the inspired Gospels. Some of them are not found in the Gospels; but who shall say whether they are or are not authentic?’ (1023). Kenyon had also contributed to the previous issue of the magazine, where he argued that the recent publication of the Arabic Diatessaron refuted Ferdinand Christian Baur’s late datings for the Gospels and confirmed their historical veracity (‘When were the Gospels written? Discoveries of the Last Twenty Years and what they have done toward answering the Question’, McClure’s, vol. IX, no. 5 [Sept. 1897], 1000–4). In contrast, Grenfell does not engage with historicity issues and confines himself to an account of ‘How We Found the “Logia”’.

104

Francis Watson

autumn, ‘leave was obtained by the Egypt Exploration Fund for Professor Flinders Petrie and myself to excavate anywhere in the strip of desert between the Fayûm and Minya, ninety miles long, in which Oxyrhynchus is situated.’22 Oxyrhynchus was selected for three reasons. First, I had for some time felt that one of the most promising sites left was the city of Oxyrhynchus, on the edge of the western desert, 120 miles from Cairo. Being the capital of one of the districts into which Egypt was anciently divided, it must have been the abode of many rich people who could afford to possess a library of literary texts.23

Second, Oxyrhynchus had been left intact by local antiquities-hunters, in contrast to other sites: Though the ruins of the old town were known to be fairly extensive, and the site still continued partly to be inhabited up to the present day, no papyri appeared to have come from it, a fact which, though it might mean that there were no papyri to be found, made it probable that the place had not been much plundered for antiquities in recent times.24

Third, and above all, Oxyrhynchus seemed to be a site where fragments of Christian literature might be expected of an earlier date than the fourth century, to which our oldest manuscripts of the New Testament belong: for the place was renowned in the fourth and fifth centuries on account of the number of its churches and monasteries, and the rapid spread of Christianity about Oxyrhynchus, as soon as the new religion was officially recognized, implied that it had already taken strong hold during the preceding centuries of persecution.25

Having set the scene in this way, Grenfell proceeds to describe how, after abandoning work on the unproductive cemetery, the excavators switched their attention to the town itself: On January 11th we sallied forth with some seventy workmen and boys, and set them to dig trenches through a mound . . . The choice proved a very fortunate one, for papyrus scraps at once began to come to light in considerable quantities, varied by occasional complete or nearly complete private or official documents containing letters, contracts, accounts, and so on; and there were also a number 22. Grenfell, ‘Oldest Record’, 1025. In view of the exceptional value of Grenfell’s article, I quote from it at length. 23. Grenfell, ‘Oldest Record’, 1024. 24. Grenfell, ‘Oldest Record’, 1024. 25. Grenfell, ‘Oldest Record’, 1024–5.

5. Q and the Logia

105

of fragments written in uncials, or rounded capital letters, the form of writing used in copying classical or theological manuscripts.26

The discovery of the Logia papyrus followed not long after: Later in the week Mr. Hunt, in sorting the papyri found on the second day, noticed on a crumpled uncial fragment written on both sides the Greek word ΚΑΡΦΟΣ (‘mote’), which at once suggested to him the verse in the Gospels concerning the mote and the beam. A further examination showed that the passage in the papyrus really was the conclusion of the verse, ‘Thou hypocrite, cast out first the beam out of thine own eye, and then shalt thou see clearly to pull out the mote that is in thy brother’s eye;’ but that the rest of the papyrus differed considerably from the Gospels, and was, in fact, a leaf of a book containing a collection of sayings of Christ, some of which, apparently, were new. More than that could not be determined until we came back to England.27

There was more to come: The following day Mr. Hunt identified another fragment as containing most of the first chapter of St. Matthew’s Gospel. The evidence both of the handwriting and of the dated papyri with which they were found makes it certain that both the ‘Logia’ and the St. Matthew fragment were written not later than the third century, and they are, therefore, a century older than the earliest manuscripts of the New Testament. It is not improbable that they were the sole remains of a library belonging to some Christian who perished in the persecution during Diocletian’s reign, and whose books were thrown away.28

The Matthew fragment is P.Oxy.2 = P1, and the implication of Grenfell’s account is that it was found in close proximity to the Logia papyrus, presumably in one of the baskets in which the second day’s findings had been gathered.29

5.2 Interpreting the Logia Grenfell’s engaging and illuminating article complements the slim 20-page booklet, published three to four months earlier, in which he and Hunt presented their transcription, reconstruction and interpretation of the Logia papyrus. They consider

26. Grenfell, ‘Oldest Record’, 1027. 27. Grenfell, ‘Oldest Record’, 1027. 28. Grenfell, ‘Oldest Record’, 1027–8. 29. Grenfell’s account continues to narrate how, ‘finding that the rubbish mounds were so fruitful, I proceeded to increase the number of workmen and boys to 110, and the flow of papyri rapidly became a torrent which it was difficult to cope with’ (‘Oldest Record’, 1028).

106

Francis Watson

but reject the possibility that the fragment is derived from an apocryphal text such as the Gospel according to the Egyptians or the Gospel according to the Hebrews, known from patristic references. Whether canonical or apocryphal, a gospel gives a connected narrative of events and discourses, not a series of disjointed sayings . . . A more satisfactory view, though not free from difficulties, is that this fragment is what it professes to be: a collection of some of our Lord’s sayings. These, judging from their archaic tone and framework, were put together not later than the end of the first or the beginning of the second century: and it is quite possible that they embody a tradition independent of those which have taken shape in our Canonical Gospels. . . In any case we may have got for the first time a concrete example of what was meant by the Logia which Papias tells us were compiled by St. Matthew, and the λόγια κυριακά upon which Papias himself wrote a commentary.30

The crucial point here is that the Logia fragment is independent of canonical and noncanonical gospels at least as regards its genre. Gospels are connected narratives, whereas the Logia fragment is a Sayings Collection. We may have here a tradition independent of the synoptic one: that, say Grenfell and Hunt, is ‘quite possible’. What is not just possible but certain is that we have here a ‘concrete example’ of written Jesus tradition being transmitted in non-narrative form. Even if several of the sayings have been influenced by their Matthean or Lukan counterparts, the framework in which they are set would still be independent of all narrative gospels. That framework is unlikely to be unique to the present Logia fragment: ‘[P]robably many such collections were made’.31 The new discovery confirms the existence of a genre. The editors received an early and enthusiastic endorsement of their hypothesis from James Rendel Harris, for whom the new text demonstrated ‘the influence of extra-canonical texts upon the readings of the New Testament.’32 In the new Logia, We find not only that we are behind the Gospels, but that there was more in the sources of the Gospels than is conserved in the Gospels themselves.33

30. Grenfell and Hunt, Sayings, 17, 18. 31. Grenfell and Hunt, Sayings, 18. 32. J. Rendel Harris, ‘The “Logia” and the Gospels’, Contemporary Review 72 (1897): 341– 8; 341. Four years earlier Rendel Harris had written, romantically yet presciently: ‘[I]t is to Egypt that we must more especially look in the coming days, for in the ruins of her cities and amongst her tombs there must yet lie a wealth of buried treasure in literature, which would make the world astonished. Especially should search be made and excavations carried on amongst the remains of cities belonging to the Christian era; for these, although not furnishing material to the student of Egyptology, are likely to contain many Christian and Greek documents’ (A Popular Account of the Newly-Recovered Gospel of Peter [London: Hodder & Stoughton, 1893], 11–12). 33. Rendel Harris, ‘The “Logia” and the Gospels’, 346; italics removed.

5. Q and the Logia

107

In assuming that the new text is itself part of a gospel source, Rendel Harris goes well beyond Grenfell and Hunt’s more cautious and defensible emphasis on the distinctiveness of the Logia genre. Rendel Harris does, however, provide further evidence for the existence of a Sayings Collection genre by citing passages from Acts and the Apostolic Fathers that attribute sayings to Jesus that do not occur in the canonical gospels. The Paul of Acts instructs the Ephesian elders to remember the words of the Lord Jesus, that he said [ὅτι αὐτὸς εἶπεν], ‘It is more blessed to give than to receive.’34

As Rendel Harris points out, the introductory formula, ‘that he said’, is superfluous in this context but recalls the repeated ‘Jesus said’ of the Logia fragment. A similar introduction is provided for an elaborate extra-canonical saying attributed to Jesus by 1 Clement, whose readers are to remember the words of the Lord Jesus which he spoke teaching gentleness and patience. For he said this [οὕτως γὰρ εἶπεν]: ‘Be merciful, that you may receive mercy. Forgive, that it may be forgiven you. As you do, so shall it be done to you. As you give, so shall it be given to you. As you judge, so shall you be judged. As you are kind, so shall you receive kindness. With what measure you measure, with it shall it be measured to you.’35

This carefully constructed saying is related to canonical material but does not derive directly from a canonical gospel. If extra-canonical sayings are introduced with a standardized introductory formula in the new Logia sayings collection, other unfamiliar sayings and their introductions may stem from similar collections. Early citations of Jesus’ sayings regularly diverge from the canonical gospels, suggesting that their sources lie elsewhere – in sayings collections exemplified by the Logia text although not to be identified with it. Rendel Harris’s evidence strengthens the case for a Sayings Collection genre relatively independent of the canonical narrative gospels.36

5.3 Competing Logia Theories If this Sayings Collection hypothesis has any plausibility, it would have significant implications for the Synoptic Problem. In particular, the ‘two-source hypothesis’ popularized by H. J. Holtzmann would face some new questions. The first of the two sources, the Gospel of Mark, would remain in place; the second, also used independently by Matthew and Luke, would encounter potential difficulties. This

34. Acts 20.35. 35. 1 Clem 13.2. 36. On this, see my Gospel Writing, 249–85.

108

Francis Watson

hypothetical second source would shortly come to be known as ‘Q’, but in the 1890s it was still known as the Logia – precisely the term that Grenfell and Hunt chose for their new discovery.37 At this point, then, there are two quite different Logia-hypotheses in play:  one generating a single Logia text from the study of double-tradition passages in Matthew and Luke, the other proposing that a newly discovered Logia text exemplifies an early Sayings Collection genre. The potential problem the new hypothesis poses for Q is that Q does not conform to the generic conventions that may be extrapolated from the new text. Q is a collection of connected discourses, not of individual sayings separated by an introductory formula. For that reason, German scholars sometimes preferred to speak of Q as a Redensammlung, a collection of speeches, rather than as a Spruchsammlung, a sayings collection.38 Q also contains narrative accounts of Jesus’ temptations and the healing of the centurion’s servant, distancing it still further from the Sayings Collection genre suggested by Grenfell and Hunt’s discovery.39 Advocates of the Q hypothesis could not be expected to welcome the rival Logia text. The fundamental questions posed by Grenfell and Hunt’s Logia Iēsou are as follows. Is the newly discovered material so closely related to the canonical texts that it must be integrated into our account of their origin and composition? Or do its affinities lie with later, second century, apocryphal gospel literature, dependent on and inferior to the authentic gospels of the apostolic age? More succinctly: does the new text change anything? Grenfell and Hunt published their Logia papyrus in June 1897, as the nation celebrated the aged Queen Victoria’s Diamond Jubilee. The new text rapidly evoked a flurry of reviews, letters to editors, public lectures and journal articles, in outlets ranging from the Expository Times, the Theologische Literaturzeitung and the Revue Biblique to the Spectator, the Guardian, Leisure Hour and Sunday at Home. One of the more substantial responses was the joint production of two senior Oxford colleagues, Walter Lock, Dean Ireland’s Professor, and William Sanday, Lady Margaret Professor, the foremost English protagonist of the Q hypothesis, described by Grenfell and Hunt as ‘amongst the most conservative of our critics’.40 Lock and 37. In the 2nd edition of his Horae Synopticae (Oxford: Clarendon Press, 1909 [1899]), Sir John Hawkins notes that he has replaced his earlier references to ‘the Logia of Matthew’ with the designation ‘Q’, following the lead of German scholars (107). The terminological shift in English-language scholarship may have been influenced by the publication in 1908 of Adolf Harnack’s book on Q in English translation (The Sayings of Jesus: The Second Source of St. Matthew and St. Luke (Eng. tr. London: Williams & Norgate, 1908). 38. Heinrich Julius Holtzmann, Lehrbuch der historisch-kritischen Einleitung in das Neue Testament, 2nd edn (Freiburg i. Br.: J. C. B. Mohr [Paul Siebeck], 1886), 371. 39.  For criticism of the assumptions that Q and the Gospel of Thomas are works in the same genre, and that the existence of Thomas strengthens the case for Q, see my Gospel Writing, 217–21 (in debate with Helmut Koester and John M. Robinson). 40.  Bernard Grenfell and Arthur S. Hunt, New Sayings of Jesus and Fragment of a Lost Gospel from Oxyrhynchus (London: Henry Frowde, 1904), 27. The title refers to the discovery of P.Oxy.654 and P.Oxy.655 respectively.

5. Q and the Logia

109

Sanday delivered public lectures on 23 October which appeared in print with supplementary material before the end of the year.41 Lock, more sympathetic to Grenfell and Hunt than Sanday, pays tribute to their editorial work: ‘[T]he careful decipherment of the text and the cautious wisdom of the notes have been recognized universally both in England and on the Continent.’42 In his assessment, the new text is ‘a copy of some pre-canonical collection of our Lord’s discourses’.43 If so, however, they would not constitute that new gospel which the Spectator dreads and which the Daily Chronicle welcomes; they would not seriously alter the conditions of the Synoptic problem as Mr. Rendel Harris imagines, for the prologue of St. Luke shows that there were pre-canonical documents out of which our Gospels were framed . . .44

If the Logia fragment is really a copy of a pre-canonical Logia collection, it is hard to see how this would have no bearing on the Synoptic problem. Sanday is more incisive than his colleague. The editors view the fragment as the remnant of an old sayings collection and suggest that ‘probably many such collections were made’. Sanday thinks we should be deeply sceptical of their claim: They are perhaps justified in saying this; but if so, their own discovery is the chief ground for holding the opinion. It is a tenable hypothesis that the new Logia are a specimen of a class, but whether they are or not will need further testing. The chief direction in which this testing can be applied would be through the analysis of our existing Gospels; and this, as inquiry stands at the present moment, can hardly be said to be favourable.45

Sanday does not deny that the fragment represents a sayings collection but he seeks to persuade his audience to view it as a unique literary phenomenon rather than as instantiating a Sayings Collection genre. The editors’ hypothesis is ‘perhaps justified’, it is ‘tenable’, it merits ‘further testing’  – and yet these lukewarm concessions amount to little, for Sanday’s real point is that current trends in gospels scholarship make the Sayings Collection hypothesis extremely unlikely. If we wish to investigate early transmission of the Jesus tradition, we have far more to learn from ‘the analysis of our existing Gospels’ than from bits of old papyrus dug up by junior members of the University with next to no theological background.46 41.  Walter Lock and William Sanday, Two Lectures on the ‘Sayings of Jesus’ Recently Discovered at Oxyrhynchus (Oxford: Clarendon Press, 1897). 42.  Two Lectures, 15. For a ‘continental’ tribute to the editors, see Adolf Harnack, Die jüngst entdeckten Sprüche Jesu (Freiburg i. Br.: J. C. B. Mohr [Paul Siebeck], 1897), 5. Harnack argues that the papyrus contains extracts from the Gospel according to the Egyptians (27–34). 43. Two Lectures, 27. 44. Two Lectures, 27. 45. Two Lectures, 31. 46. At the time of Sanday’s lecture Grenfell was 27, Hunt 26 and Sanday 54.

110

Francis Watson

The issue that Sanday rightly grasps is that the new noncanonical Logia text poses a challenge to the old intracanonical one  – that is, to Q.  But Q will not be so easily dislodged from its privileged position as the earliest record of Jesus’ teaching: It is a widely held opinion that behind the common portions of our First and Third Gospels . . . there lies a yet earlier source which might be described by the name ‘Logia’. But the tendency is to think of these Logia as something more than pure sayings, strung together in no apparent order, and with no connexion beyond the repeated λέγει Ἰησοῦς of the Fragment. If we take a narrative like the Healing of the Centurion’s Servant, which is common to the two Gospels, and is not found in St. Mark, we see there a complete story, and not an isolated saying or sayings, and partly cast into the form of dialogue. In this it is unlike the Fragment. And even those portions of the common matter of the two Gospels which are more strictly made up of sayings yet in one or both of the Gospels usually have a few words of introduction assigning them to some particular occasion.47

While Sanday’s Logia would seem to be a text of greater literary sophistication than the Oxyrhynchus Logia text, we are certainly not meant to conclude that the new text might predate Q. For all its veneer of modernity, Q is a canonical text. Literary and theological degeneration from the early and authentic to the late and apocryphal is only to be expected. In the late nineteenth and early twentieth centuries, Q was establishing itself as a benchmark of critical orthodoxy in the English-speaking world, and Sanday was a key figure in that process.48 It is worth reflecting further on the factors that led Sanday and subsequent generations of scholars to prefer to work with a hypothetical yet intracanonical text rather than engaging seriously with rediscovered extra-canonical ones.

5.4 The Politics of Q Q has its natural habitat within what Sanday calls the ‘Two-Document hypothesis’.49 The two documents are Mark and Q understood as sources used independently of each other by Matthew and Luke. Foundational to the whole construct is the modern assumption of literary interdependence which replaced the traditional 47.  Two Lectures, 31–2. ‘Yet’ in the last sentence seems to mean ‘nevertheless’. Sanday acknowledges that the introductions may well be redactional, but allows his point to stand. 48.  Sanday’s significance is acknowledged by B.  H. Streeter in the dedication of his source-critical classic (The Four Gospels, v): ‘In Memoriam Gulielmi Sanday, S.T.P., insignissimi apud Oxonienses horum studiorum Fautoris’ (‘. . . the most eminent patron of these studies among Oxonians’). 49. Two Lectures, 31 n.

5. Q and the Logia

111

notion of independent apostolic or post-apostolic authorship. The theory of literary interdependence was limited and sometimes challenged by appeals to an unknowable oral tradition, but it was widely agreed that oral tradition alone was unlikely to have produced gospels in which common material so often occurs in common sequence. The question was which of several competing accounts of literary interdependence was to be preferred. An enduring consensus gradually formed around Markan priority as the best explanation for the ‘triple tradition’ material shared by all three synoptists. The decisive criterion here is the lectio difficilior principle borrowed from textual criticism: in a preponderance of cases it is easier to believe that Matthew or Luke has emended Mark than that Mark has emended Matthew or Luke (as Griesbach had claimed). Yet Matthew and Luke do not normally agree with each other as they rewrite Mark; they must therefore have done so independently. It is at this point that the second document of the ‘TwoDocument hypothesis’ comes into view. If Matthew and Luke are independent of one another in the ‘triple tradition’ material derived from Mark, it is plausible to suppose that they are equally independent of one another in the ‘double tradition’ material they share only with each other. In that case, this material must be drawn from a common source: hence, Q. The Two-Document hypothesis is based on a supposed analogy between the editorial processes that created the triple and the double traditions respectively. If Matthew and Luke appear to engage independently with an earlier source in the one case, that is likely to be true of the other. One reaches Q by way of an analogy assumed a priori and then tested against the textual evidence. Q stands or falls by its ability to make this analogy credible and defend it against counterarguments.50 This is essentially the position that a 29-year-old Sanday had already reached in 1872, when he announced his provisional acceptance of ‘the documentary origin of the Synoptic Gospels, the priority of St. Mark, the existence of two main documents, and the independent use of them by the Evangelists’ (i.e. Matthew and Luke).51 This account is associated especially with H. J. Holtzmann but is also maintained by ‘a majority of the best critics during the last ten or fifteen years’.52 This 1872 appeal to a majority opinion recurs in the 1897 lecture, where the

50.  ‘Half a century of critical investigation has made it clear that the parallel matter in Matthew and Luke falls into two parts – a larger part, which is convincingly explained by their use of a common source still substantially preserved in our Second Gospel, and a lesser part, which it seems natural to explain on the same analogy as due to their use of a second common source now lost – which hypothetical source has been conveniently designated by the symbol “Q” ’ (B. H. Streeter, ‘On the Original Order of Q’, Oxford Studies in the Synoptic Problem, by Members of the University of Oxford, ed. William Sanday (Oxford: Clarendon Press, 1911), 139–64, 141; italics added). 51.  William Sanday, The Authorship and Historical Character of the Fourth Gospel, Considered in Reference to the Contents of the Gospel Itself:  A Critical Essay (London: Macmillan, 1872), ix. 52. Authorship and Historical Character, ix.

112

Francis Watson

two-document hypothesis is described not only as ‘a widely held opinion’ but also as ‘[t]he dominant theory’.53 The concern here is essentially political: to promote a consensus around Markan priority and Q by representing it as an already existing orthodoxy from which it would be foolhardy to dissent. Sanday’s political agenda comes to light especially in the collection entitled, Oxford Studies in the Synoptic Problem, which he edited in 1911 towards the end of his long career. This work, he tells us, has had an unusually long gestation: For some considerable time past (since 1894) a class has been in the habit of meeting in the ‘lodgings’ of the Lady Margaret Professor which we have called a ‘Seminar’, though it has not been quite like the gatherings known by that name on the Continent. The subject of study has been the Synoptic Problem . . . The Seminar has lived through some four or five generations of Oxford life, and it has been attended for the most part by graduates and special students; so that a certain proportion of the members have naturally been ‘birds of passage’, who have stayed for a year or so and have then gone. But it has been the peculiar happiness of this Seminar that it has had a permanent nucleus of members who have been faithful to it from the very first.54

Playing down his own role, Sanday informs us that the essay collection is the work of three of those original members and three younger scholars who have joined the seminar more recently. First and foremost among the senior group is Sir John C. Hawkins, author of a work entitled Horae Synopticae whose two editions present painstaking analyses of the distinct vocabularies of the synoptic evangelists and their sources. Pre-eminent among the second, younger group is B.  H. Streeter. Markan priority is generally accepted, but only Hawkins and Streeter share Sanday’s understanding of Q; two of the founder members remain unpersuaded and develop alternative source-critical hypotheses of their own.55 Embarrassed by this lack of consensus, even after seventeen years of working together, Sanday separates the sheep from the goats. The positions of Hawkins and Streeter ‘fall well within the limits of what may be called the view generally current among scholars’.56 Their essays therefore open the main body of the volume: Sanday thinks it ‘a distinct advantage that the Two-Document hypothesis should have a full statement first’.57 Then there are the views of ‘the two dissentients’ (W. C. Allen and J. V. 53. Two Lectures, 31 n. 54. Oxford Studies, vii. 55.  W. C. Allen reverts to a Schleiermacher-like identification of Q with the Matthean discourses (‘The Book of Sayings Used by the Editor of the First Gospel’, Oxford Studies, 235–86; on Schleiermacher’s Logia hypothesis see my Gospel Writing, 105–7). J. V. Bartlet identifies Q as the source of all non-Markan material in Luke (‘The Sources of St. Luke’s Gospel’, Oxford Studies, 313–63). The other two contributors (W. E. Addis and N. Williams) do not directly engage with Q. 56. Oxford Studies, xv. 57. Oxford Studies, xi.

5. Q and the Logia

113

Bartlet), whose theories ‘fall a little (though in substantial result not very much) outside’ the generally current view represented by Hawkins and Streeter.58 By casting his own contributors as orthodox or dissenters, and by making clear his own decided preference for orthodoxy, Sanday is converting the Synoptic Problem into a Synoptic Solution with Q at its heart. Already canonical by virtue of its origin within Matthew and Luke, Q becomes canonical a second time as an assured result of modern critical scholarship. Debate will continue, no doubt, but questioning the orthodox Q hypothesis will now be redefined as dissent. Sanday’s contribution to the Synoptic problem lies not in any reasoned scholarly arguments but in his rhetoric. Some such political agenda already underlies Sanday’s response to Grenfell and Hunt in 1897. The new Logia come into potential conflict with the intracanonical Q, and they are therefore banished into apocryphal obscurity. In an only apparent concession to the new discovery, Sanday states that ‘[i]n any future investigation of the Synoptic Question, the new Fragment is not likely to be lost sight of ’.59 The statement is disingenuous. Sanday has no intention of integrating the fragment into debate about canonical gospel origins. On the contrary, he is determined to exclude it. In the Oxford Studies it is referred to just once, in a parenthetical reference to the ‘so-called Logia found at Oxyrhynchus’.60 There is only one authentic Logia text, and it is the canonical Q. There is nothing in the hypothesized Q that was not first in Matthew and Luke. If Q’s discourses are coherent, it is a Matthean or Lukan coherence. The wording of a reconstructed Q saying will always be drawn from the Matthean or Lukan version of the same saying, or from both. The order of Q will correspond to the order of Q elements in Matthew and/or Luke. At no point is Q independent of its canonical containers; it has its place firmly within the canonical four gospel collection, distributed and duplicated across two texts. Q is not apocryphal or extra-canonical or even pre-canonical. Although radical conclusions are sometimes drawn from it, it is essentially a conservative hypothesis which leaves the canonical/noncanonical boundary intact. Already in 1897 the publication of P.Oxy.1 brought to light the more challenging possibility that noncanonical gospel material might have a part to play in the continuing investigation of canonical gospel origins.

58. Oxford Studies, xv. 59. Two Lectures, 33. 60.  Oxford Studies, 123 (Hawkins). The index lists a second reference to Oxyrhynchus, which appears to be incorrect.

Chapter 6 W AT S O N , Q A N D   ‘ L / M ’ Christopher M. Tuckett

6.1 Introduction: General Issues In his recent major study of ‘Gospel Writing’, Francis Watson addresses in considerable detail the issue of the relationships between the canonical (and other) gospels.1 He accepts the theory of Markan priority, but he questions radically the other prong of the Two-Source theory (2ST), namely the existence of a Q source; in its place, he mounts a powerful, detailed defence of what he calls the ‘L/M hypothesis’, arguing for the direct dependence of Luke on Matthew.2 Watson deals with a range of many issues in his book. However, this chapter focuses primarily on Watson’s theories about the existence (or otherwise) of Q.

1. Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI: Eerdmans, 2013). See also his earlier essay ‘Q as Hypothesis: A Study in Methodology’, New Testament Studies (NTS) 55 (2009): 397–415. Despite a few small differences in some details, there appears to be no change in his overall views between the two publications. (In informal conversation at the Roskilde conference, Watson suggested that it might be inappropriate to use the earlier article and that he might have ‘moved on’ from this in writing his later book. However, in e.g. Gospel Writing, 163 n. 13, he seems to imply that he regards the methodological arguments in the earlier article as still valid and there is no hint of a retraction in the book of any parts of the earlier article. I therefore draw on both published works here.) Page references in the text here are to Watson’s book unless otherwise stated. 2. The hypothesis is not new, though this name for it is. The theory has been advocated in recent times, notably by Austin Farrer and, more recently, by Michael Goulder: see e.g. his Luke: A New Paradigm, 2 vols, Journal for the Study of the New Testament Supplement (JSNTSup) 20 (Sheffield: JSOT, 1989), and Mark Goodacre: see his The Case against Q: Studies in Markan Priority and the Synoptic Problem (Harrisburg, PA: Trinity Press International, 2002). Also Eric Franklin, Luke: Interpreter of Paul, Critic of Matthew, JSNTSup 92 (Sheffield: JSOT, 1994), and essays in Mark Goodacre and Nicholas Perrin (eds), Questioning Q. A Multidimensional Critique (Downers Grove: InterVarsity, 2004); John C. Poirier and Jeff Peterson (eds), Marcan Priority without Q, The Library of New Testament Studies (LNTS) 455 (London: Bloomsbury

116

Christopher M. Tuckett

Watson assumes that, among hypotheses that seek to explain the Matthew-Luke relationship, the main competitors in the field are the Q and L/M hypotheses.3 He suggests that the L/M hypothesis has never really had a fair hearing (or that it has ‘rarely if ever been advocated in a form in which it deserved to succeed’ (p. 119)) and that the Q hypothesis has been allowed to become the default theory. By contrast, therefore, Watson claims that the L/M theory should be taken more seriously. Further, the method he proposes is that one should ‘compare the redactional procedures entailed by both the Q and the L/M hypotheses’ since, he claims, ‘it is the alleged implausibility of the latter that creates the need for the former’ (‘Q as Hypothesis’, 399). Such an overall method would probably command universal consent today.4 On the other hand, the way in which Watson applies this may be questionable, and this in at least two respects in relation to general methodology. First, there is the issue of what is included in the ‘redactional procedures’ involved. When considering this in more detail, Watson explicitly refuses to discuss a number of aspects that many would regard as ‘redactional’ activity requiring some explanation. For example, he almost always refuses to provide explanations for omissions, whether of individual words or of larger units of tradition, by a secondary evangelist.5 He also generally eschews discussions about the detailed wording at the level of individual words or phrases.6 T&T Clark, 2015). The hypothesis has been variously named (as e.g. the ‘Farrer hypothesis’ or the ‘Farrer-Goulder hypothesis’): see on this, Christopher M. Tuckett, ‘The Current State of the Synoptic Problem’, in From the Sayings to the Gospels, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 328 (Tübingen: Mohr Siebeck, 2014), 77–116, 81–2. Watson prefers the label ‘L/M hypothesis’, partly because it then is not explicitly linked to one particular advocate of the theory (and his/her associated ideas), partly too because he believes that the most persuasive advocate of the theory today is neither Farrer nor Goulder but Goodacre. See Watson, ‘Q as Hypothesis’, 398–9; Gospel Writing, 119. 3.  He notes in passing the Griesbach, or Two Gospel, hypothesis, but does not engage with it in any detail: see ‘Q as Hypothesis’, 399; Gospel Writing, 119–20 n. 5. 4.  For the general approach, comparing the redactional activity by the secondary evangelist(s) on competing hypotheses, see (from many years ago) Christopher M. Tuckett, The Revival of the Griesbach Hypothesis, Society for New Testament Studies Monograph Series (SNTSMS) 44 (Cambridge: Cambridge University Press, 1983), 12–15; more recently, Tuckett, ‘Current State’, 89. A similar approach is adopted by many of the detailed studies of e.g. John Kloppenborg in arguments for the existence of Q. 5.  See e.g. Gospel Writing, 159 on Luke’s non-inclusion of some of the extra beatitudes in Matthew: ‘Luke’s “rejection” [of Matt 5.8] . . . does not mean that he disliked it, and no special explanation is required as to why that might have been the case’ (my stress); also ‘Q as Hypothesis’, 403: ‘little if anything should be read into an omission.’ Similar statements in ‘Q as Hypothesis’, 405–6; Gospel Writing, 144, 172, 175 n. 32. 6.  ‘Each of the competing hypotheses can produce plausible examples of apparent “primitivity” or “secondariness” as required . . . The synoptic problem will not be solved by ad hoc appeals to favoured proof texts.’ (Gospel Writing, 162–3; see also ‘Q as Hypothesis’, 403 n. 17; Gospel Writing, 191.)

6. Watson, Q and ‘L/M’

117

This is however questionable. How one defines ‘redactional’ precisely might vary. However, most would take the word as including changes which a secondary evangelist has made to his (supposed) earlier source.7 Such changes occur at many levels: they can involve omissions and/or changes (and replacements) of individual words and smaller phrases, as well as larger-scale changes involving the choice and ordering of whole sayings or pericopes. To omit one part of this whole network (omissions, and changes in detailed wording) and to focus solely on another part would seem to be unjustified. Some changes made by a secondary writer may well be unimportant, and one should beware of over-interpreting what may be tiny and/ or genuinely insignificant changes. However, some changes may be significant, and one cannot rule these out in deciding what to include in the ‘redactional procedures’ to be investigated. Watson focuses almost all attention on the larger-scale changes in order which Luke must have made to Matthew; and indeed by the end of his chapter advocating the L/M hypothesis (Gospel Writing, ch. 4: ‘Luke the Interpreter’), he has shifted terminology from speaking not of Luke’s ‘redactional’ procedure, but of his ‘compositional’ procedure.8 Nevertheless, the changes which Luke must have made at the smaller level, of changes to detailed wording and also decisions to omit material, may be just as significant and they should not be dismissed prematurely.9 Second, there is the issue about where any ‘burden of proof ’ might lie. Watson states that, methodologically, one should ‘compare’ the redactional procedures presupposed by competing hypotheses. This might suggest that the Q hypothesis and his own L/M hypothesis are to be treated as equals: hence, if Luke’s procedure on the L/M hypothesis can be shown to be reasonable, plausible and coherent, then the hypothesis must be taken seriously and should be seen as on a par with the Q hypothesis. This is unexceptional. However, Watson sometimes implies that, if Luke’s redactional procedure on the L/M hypothesis can be shown to be reasonable and intelligible, then that is enough to establish the validity of the hypothesis over against the Q theory.10 7.  Whether including material unchanged should also be regarded as ‘redaction’ is debated:  it is certainly arguable that this may be just as significant as actively changing the source (and this has led to criticism of some older-style ‘redaction-critical’ approaches which focused solely on changes made by a secondary writer). 8.  Cf. e.g. the summary on p. 215: ‘Luke follows a simple compositional procedure . . .’ (my italics). For some, ‘compositional’ changes relate to the larger changes (involving the overall ‘composition’ of the work), sometimes contrasted with ‘redactional’ changes at the level of detailed wording. 9. In discussion at the Roskilde meeting, Watson said that it was perhaps unreasonable to expect advocates of a hypothesis to explain everything. This may be slightly disingenuous: in any case, critics of the hypothesis have often focused on the omissions which Luke must have made, and some of the changes in detailed wording, as key difficulties for the hypothesis. One cannot therefore simply ignore such evidence in defending the hypothesis anew in the current debate. 10. Cf. ‘Q as Hypothesis’, 407: ‘if the L/M redactional procedure is judged to be no less plausible than the alternative [as implied by the 2ST]’, this ‘undermines the Q hypothesis’ (his italics).

118

Christopher M. Tuckett

The implied logic is however questionable. The greater plausibility of the L/M theory over against the Q hypothesis is only established if the redactional procedure implied by the first is shown to be ‘reasonable and intelligible’ and the procedure implied by the second is shown to be less reasonable or intelligible (or unreasonable and/or unintelligible). Watson’s argument at this point is not quite clear (at least to me). In part he may be appealing to the past history of the Q hypothesis, and the fact that the Q theory only arose because of the difficulty of explaining Luke’s redactional activity on the L/M theory. Thus ‘Q is premised on the unlikelihood that either of the later evangelists is dependent on the other in addition to Mark’.11 It would seem though that, for Watson, if the redactional activity implied by the L/M hypothesis can be shown to be plausible, then this will by itself render the Q hypothesis redundant and/or unnecessary. It is probably true to say that the theory of the existence of Q arose because scholars at the time started from theories of direct dependence, and only moved to a Q theory subsequently. However, as Watson himself notes, the Q hypothesis has now built up an impressive arsenal of arguments and support (cf. p. 119); in particular it has produced a whole raft of explanations about the redactional activity (by Matthew and Luke) that has seemed to many plausible and coherent. Whether these are convincing is of course debatable. However, one cannot now necessarily appeal to the past to justify dismissing the Q hypothesis simply because an alternative theory can be made ‘plausible’ or ‘intelligible’. Two hypotheses generate two sets of redactional procedures by (a) secondary evangelist(s), and it is the task of any student of the synoptic problem to compare these patterns, fairly and as wholes. We cannot, and should not, prejudge the issue by implying that one ‘plausible’ explanation of the overall redactional procedure ipso facto rules out alternative hypotheses. To be fair, Watson does develop an argument against Q (see below): this concerns the claim that the Q hypothesis entails a series of ‘coincidences’ which he claims are too many to be credible. However, in the main part of his analysis of ‘Luke the Interpreter’, Watson focuses his discussion on what Luke has done assuming the L/M hypothesis. Rarely if ever does he consider the alternative scenario and ask how this redactional procedure compares with that presupposed by the Q hypothesis. There is thus very little comparison of redactional (or compositional) procedures in Watson’s book, despite the clear methodological statement in his NTS article. One further point needs to be made about the ‘redactional procedure’ which Watson discerns for the L/M hypothesis. This concerns the extent to which Watson seeks to explain the changes which have occurred. Kloppenborg has drawn

11. ‘Q as Hypothesis’, 398 (similarly on p. 399). Also Gospel Writing, 181 on potentially equally possible explanations (of order) by both the L/M and Q hypotheses: ‘But there is no need for [the Q hypothesis].’

6. Watson, Q and ‘L/M’

119

attention to the need to distinguish ‘description’ and ‘explanation’ in this context.12 To say that, assuming one particular hypothesis, Luke has made this or that set of changes simply describes the situation. It only provides an explanation if one provides reasons why Luke might have changed his source(s) in the way he has. Further claims that a set of changes are plausible (or implausible) relate to proposed explanations of the changes made. What makes an explanation ‘plausible’ is clearly difficult to determine in advance or in the abstract. But an explanation for Luke’s use of a claimed source might be seen to be more (or less) ‘plausible’ insofar as it coheres with, for example, Luke’s use of sources elsewhere, the ways in which other writers of the period used sources, or Luke’s ideas and interests as evidenced elsewhere in Luke-Acts. As such then, an explanation seeks to show that the changes made by Luke can be seen as part of a coherent overall activity in using his sources and in his writing generally.13 Watson seems at times almost deliberately to avoid such attempts at explanation. As noted earlier, Watson tends to avoid all discussion of the detailed wording of individual texts (as inconclusive: cf. above). He thus focuses on the changes in order of larger units of material that Luke has made to Matthew. But even here, the discussion mostly stays at the level of what Luke has done. Why he might have done what he is purported to have done is often studiously avoided. Indeed Watson is occasionally specific about this. For example, in discussing the changes in order Luke must have made to Matthew, Watson says: If it be asked why he [Luke] did so, the best answer is one he himself gives: it seemed good to him to compose his gospel in his own way, so as to address the problems of order bequeathed by his predecessors (cf. Lk. 1.1–4). Beyond that we cannot go (p. 169, my stress).14

However, this then seems to involve a process that is virtually just describing what Luke has done, coupled with an all but explicit refusal to engage with an attempt at explaining (in any detail) what he has done (cf. Watson’s claim above ‘Beyond that we cannot go’).15 I noted earlier that Watson appears to feel under no compulsion at all to explain omissions by Luke; he devotes an enormous amount of attention to considering the ‘compositional procedure’ (at the macro-level) of Luke’s use of 12. Cf. John S. Kloppenborg, ‘On Dispensing with Q? Goodacre on the Relation of Luke to Matthew’, in Kloppenborg, Synoptic Problems, WUNT 329 (Tübingen:  Mohr Siebeck), 2014), 73 (on Farrer’s claim that Luke retains what is ‘Luke-pleasing’). 13. On plausibility, see further §6.3.2. 14.  Cf. too p. 175 n. 32 and a similar appeal to Luke 1.3 (‘it seemed good to me’) and Watson’s comment: ‘that should be good enough for us too.’ Also ‘Q as Hypothesis’, 405, in relation to why Luke might have been ‘motivated’ to break up Matthew’s Sermon on the Mount: any such enquiry about ‘motivation’ is ‘premature’. 15. See below for the issue of ‘order’ and whether this will adequately account for Luke’s redaction of Matthew.

120

Christopher M. Tuckett

Matthean material, but appears to shun any attempt to explain this in any detail beyond referring to the fact that Luke might be concerned about the order and has changed things. That he has changed Matthew’s order (on the L/M hypothesis) is manifest; what is still lacking is a detailed explanation of why he did so in the ways he did.

6.2 Coincidences in Q? A significant part of Watson’s negative case against Q is that the hypothesis of Matthew’s and Luke’s independence from each other involves a whole series of coincidences between the two gospels, coincidences going far beyond the wellknown ‘minor agreements’ between Matthew and Luke (Gospel Writing, ch. 3). Thus in very general terms, on the 2ST, Matthew and Luke both know Mark, both have access to Q and independently choose to supplement Mark by Q material. Further, both add birth narratives and a genealogy to Mark; both also add appearance stories at the end of Mark. Watson also discusses some details within Luke 3–6 and parallels, for further evidence of coincidences. All these are, by implication, too numerous and too significant to be explicable by chance; rather, the agreements between Matthew and Luke call into question radically their alleged independence from each other and hence the whole Q hypothesis. In response, one may raise first some general questions about ‘agreements’. What is an ‘agreement’ in this context? How far can one say that Matthew and Luke ‘agree’ with each other in the aspects Watson highlights? We have to be aware that, in relation to the Synoptic Problem as with history more generally, the way in which we describe the evidence may already impose an interpretation on the data that is, at the very least, debatable. Thus the contents of the gospels might be described in one way that suggests an ‘agreement’ between Matthew and Luke, whereas other descriptions might suggest that there is disagreement.16 This applies especially to the birth narratives and the genealogies in Matthew and Luke. 6.2.1 Birth Narratives Watson claims that Matthew and Luke coincide in adding birth narratives to Mark (and also in adding appearance stories at the end of Mark).17 This is indeed true. However, the birth narratives in question are very different from each other (as are the appearance stories). Is this then a case of Matthew and Luke agreeing (both

16.  Similar problems arise in relation to the ‘minor agreements’:  see my ‘The Minor Agreements and Textual Criticism’, in Minor Agreements, ed. Georg Strecker, Göttinger Theologische Arbeiten (GTA) 50 (Göttingen: Vandenhoeck & Ruprecht, 1993), 119–41, 121–5. 17. The resurrection appearances are not discussed in any detail at all; for the birth narratives, see Gospel Writing, 131–6.

6. Watson, Q and ‘L/M’

121

add birth stories), or disagreeing (they add very different birth stories)? If it is an ‘agreement’ or ‘coincidence’, it is only so at a level of some generality. Watson agrees in part that the stories are very different but claims that there are several common features (Mary as Jesus’ mother, Joseph as Jesus’ father, Mary conceiving through the agency of the Holy Spirit, angelic announcements, a birth in Bethlehem with Jesus’ home in Nazareth). Some of these could be explained by traditions about Jesus’ birth attested elsewhere in the New Testament.18 However, the annunciation is rather different: here we have two very similar stories with angelic appearances (albeit to different people) and some verbal similarities.19 To account for the differences, Watson says that Luke’s annunciation story ‘was composed in reaction to Matthew’s’ (p. 134). However, no reasons are given for a process of such radical rewriting of the Matthean account (if Matthew is Luke’s source). Such rewriting goes far beyond simply reordering the material (a concern which Watson says is dominant for Luke: on this see §6.3.1). Luke here must have radically rewritten Matthew’s account so that, for example, it is Mary not Joseph who receives the visit from the angel; the flight into Egypt, the massacre of the innocents and the coming of the magi are omitted;20 instead Luke has the story of the shepherds as well as the presentation of the child in the temple. Simply to say that Luke’s account is ‘composed in reaction to Matthew’, scarcely scratches the surface of the problem if it is not stated what Luke might be reacting against (and why),21 and/or why he found it appropriate to rewrite Matthew in such a radical way. Further, there is scarcely anything comparable to such radical rewriting of Matthew posited elsewhere.22 Luke’s proposed redaction of Matthew scarcely 18. As Watson himself says (Gospel Writing, 133): e.g. for the names of Jesus’ parents, cf. Mark 6.3; John 1.45; 6.42; for Davidic descent, cf. Rom 1.3; for a possible virginal conception, cf. Gal 4.4, etc. 19.  Cf. p.  135:  Matt 1.21 agrees with Luke 1.31 in having the angel say ‘she/you will bear a son, and you shall call his name Jesus’; see too Goodacre, Case, 56–7 for the same agreement noted. 20.  Watson, Gospel Writing 135 n.  43, states that there is a measure of ‘coincidence’ between Matthew’s story of the magi and Luke’s story of the shepherds: both witness to a ‘ “visitation” motif ’, with both mentioning ‘great joy’ (Matt 2.10; Luke 2.10). But again there is ‘coincidence’, or ‘agreement’, only at a very high level of generality. 21. The possibility that Luke is highly critical of Matthew, and writes to displace and correct Matthew, is developed by Franklin, Luke; see too Poirier, ‘Introduction’, in Poirier and Peterson, Marcan Priority without Q, 1–15, 6–11; in relation to Matt 1–2, Poirier suggests that Luke might have objected to the ‘scandalous suggestion that Jesus’ birth was astrologically determined’ (p. 10: cf. Matt 2.1–12). But this would not explain the other wholesale changes Luke has made. Watson does not seem to go down this general line: he notes (correctly!) that, unlike e.g. Josephus, Luke makes ‘no overt criticism’ of his predecessors in gospel writing (123). 22.  Though see also §6.2.2 on the genealogies. The appearance stories in Matt 28/Luke 24 might provide a comparable situation if Luke is assumed to have used Matthew. The stories are mentioned in passing by Watson as another possible ‘coincidence’ in that Matthew and Luke both expand Mark here: but they expand Mark here in very different ways!

122

Christopher M. Tuckett

‘coheres’ with his use of Matthew elsewhere and this raises doubts about how ‘coherent’ Watson’s explanation of Luke’s redactional procedure here is. On the 2ST, Matthew and Luke presumably had independent access to different (and largely independent) birth narratives. To posit two such traditions is not unreasonable, and some overlap (in motifs, and occasionally odd words or phrases) is not inconceivable.23 Further, the small overlaps are mostly explicable on the basis of common traditions associated with the birth of Jesus attested elsewhere.24 Further, the fact that two later writers, both using Mark, should have added birth narratives is scarcely surprising. For many earlier critics (discussing the genre of Mark), one peculiar feature of Mark was its lack of any birth stories:  this, it was said, rendered Mark’s gospel quite unlike other ‘biographies’ of the time and hence generically very unusual within Greco-Roman literature.25 Whether this claim of itself is justifiable cannot be discussed here; but the addition of birth stories to preface an account of Jesus’ life is very understandable by two later writers working independently, given the possible expectations which readers of Mark’s gospel might have in the ancient world.26 6.2.2 Genealogies Similar considerations apply to the genealogies in Matthew and Luke. For Watson, these constitute another of the ‘coincidences’ which the Q hypothesis must explain (pp.  141–3). Yet once again, this creates an alleged ‘coincidence’, or agreement, by describing the situation at a relatively broad level of generality (Matthew and Luke ‘incorporate genealogies’), whereas others might see here huge differences between the two gospels: the two genealogies are very different in content, and are added at different places in the two gospels (Matthew before the birth narratives themselves, Luke after the baptism of Jesus). Is this then a case of agreement or disagreement? The fact that both agree in showing Jesus’ Davidic descent may also be not surprising as this may have been part of a tradition more widely attested within early Christianity (cf. Rom 1.3). 23. For the common language (‘agreement’) of Matt 1.21 // Luke 1.31 as traditional, see Kloppenborg, ‘Dispensing with Q?’, 76, (against Goodacre):  Kloppenborg calls it ‘purely formulaic’ and compares Gen 16.11; Isa 7.14; 8.3; Hos 1.1–4; 1.6 etc. 24.  Pace Watson, Gospel Writing, 132–3 (referring to Raymond E. Brown, The Birth of the Messiah [New York: Doubleday, 1979]) this is not necessarily related in any way to issues of historicity: it is simply a matter of tradition which may predate Matthew and Luke. 25.  Similarly Richard J. Bauckham, ‘Gospels before Normativization:  A Critique of Francis Watson’s Gospel Writing’, JSNT 37 (2014): 185–200, 188. 26. Similarly the addition of appearance stories at the end of Mark’s gospel is hardly surprising. Mark’s ending (at 16.8) has puzzled modern readers continuously and clearly also caused problems for ancient readers (cf. the endings added to Mark in the textual tradition). Thus the possibility that Matthew and Luke ‘coincided’ in thinking that Mark’s gospel was deficient and needed supplementing with accounts of the appearance(s) of the risen Jesus is hardly surprising.

6. Watson, Q and ‘L/M’

123

The difference in content between the two genealogies relates especially to the line of descent from David to Jesus:  Matthew traces the line of descent via Solomon to Jechoniah, Luke via Nathan, Solomon’s brother (cf. 1 Chr 3.5). Watson says that Matthew’s line of descent is the ‘obvious’ one (p. 142), and hence Luke’s is an ‘alternative’. He then claims: Luke seeks to dissociate Jesus from the Solomonic line of descent, . . . deliberately rejected by Luke even as he draws material for his own genealogy from the same scriptural context. If Luke’s genealogy rejects the Solomonic lineage, and if that lineage is fundamental for Matthew, then the most economical explanation is that Luke presupposes and reacts against the Matthean genealogy. (pp. 142–3).

Luke then ‘deliberately reject[s]’ any claim that Jesus was descended from Solomon. But is this either plausible or coherent as a Lukan strategy? Is there any other evidence (apart from the presumed redactional procedure here) that would make such a move by Luke plausible? Elsewhere in Luke-Acts, Luke is positive about David as a/the ancestor of the Messiah who is Jesus (cf. the speeches in Acts 2, 4, 13). Nothing is said one way or the other about Solomon or other subsequent kings of Israel.27 Statements that are slightly less than fulsome about Solomon appear elsewhere in the gospel tradition (cf. Luke/Q 11.31; 12.27), though both sayings also occur in Matthew who evidently saw no contradiction between these and the claim that Jesus was the Son of David via Solomon.28 Why then Luke should have ‘rejected’, or ‘sought to dissociate’ himself from, the idea that Jesus’ Davidic sonship came via Solomon is not clear. Luke may simply not know it, and the tradition available to him may provide a different line of descent which he incorporates into his gospel unchanged. The claim that Luke ‘reacts against’ Matthew’s genealogy thus lacks any other evidence (from elsewhere in Luke-Acts) to support it and with indeed some counter-evidence the other way. Further, as with the birth narratives, Luke’s alleged process of ‘reacting against’ Matthew must have led him to a process of radical rewriting and alteration which is on a totally different scale from that which he must have undertaken elsewhere in relation to the other material he shares with Matthew. There is then little or no evidence that Luke’s alleged rewriting of Matthew here coheres with his procedure elsewhere.

27.  Watson (p.  142) refers to Fitzmyer’s suggestion that Luke may have ‘avoided the royal line from Solomon to Jeconiah either because of OT strictures on the reigns of some of these kings, or, more likely, because of the oracles in Jer 22:28–30 and 36:30–31 about the coming extinction of the Davidic dynasty’ (Joseph A. Fitzmyer, The Gospel According to Luke [New York:  Doubleday, 1981–85], 1:501). But the point of the Lukan account is that the Davidic dynasty has not become extinct but now has its rightful occupant in the person of Jesus. 28.  Both sayings are not overtly critical of Solomon, but simply claim that the present time far exceeds the time of Solomon (in 11.31) or that nature itself is far more splendid than even the proverbial splendour of Solomon (12.27).

124

Christopher M. Tuckett

There is indeed a ‘coincidence’ between Matthew and Luke at a very general level in that both incorporate genealogies into their stories. But they incorporate very different genealogies, and at different places. The evidence is just as well, if not better, explained if one presumes that Luke did not know Matthew, that he knew an independent genealogy tracing Jesus’ Davidic descent via another route, and he added that into his gospel at the place he saw fit. 6.2.3 Luke 3.1–20 Watson finds a series of further coincidences implied by the Q hypothesis in the accounts of the preaching of John the Baptist and the baptism of Jesus (Luke 3.1– 20 pars.).29 Matthew and Luke both record the preaching of John the Baptist (‘you generation of vipers . . .’ Luke 3.7–9 par.). On the Q hypothesis, this presumably came to them as part of Q; yet both choose to add this material at exactly the same point in Mark, namely, after Mark 1.1–6. Further, the saying about a future baptism in Mark 1.7–8 is treated in exactly the same way in both Matthew and Luke: (i) both separate the two halves of the double baptism saying (Mark 1.8) to enclose the saying about the coming one (Mark 1.7); (ii) both add that the coming one’s baptism will be not only in/with Holy Spirit, but also with ‘fire’; (iii) both add the saying about the gathering of wheat and the fiery destruction of chaff (Luke 3.17 par.). Further, (iv) both agree in adding a version of a ‘Temptation’ account, where again there is extensive verbal agreement between Matthew and Luke against Mark, in the immediate sequel. How far this all tells against a Q theory is however debatable. On the 2ST, this section is a classic example of a so-called Mark-Q overlap, where Mark and Q both shared some material. In general terms, an overlap is scarcely surprising: one would hardly expect two sources containing things purportedly said or done by Jesus (and/or related figures such as John the Baptist) to have absolutely no overlap at all. Here the preaching of John the Baptist (Luke/Q 3.7–9) is a classic example of Q material on the 2ST (displaying almost verbatim agreement between the two gospels over c.  60 Greek words). So too the saying about the chaff and the fire (Luke/Q 3.17) is also Q material (on the 2ST). The agreements between Matthew and Luke in their versions of the double baptism saying (Mark 1.7–8 pars.) can be explained by the existence of a Q version of the saying alongside Mark’s. That Matthew and Luke might have incorporated this material independently at the same point in the Markan outline is scarcely surprising. Both traditions are about John’s own preaching and message. The almost unanimous assumption in the gospels is that John was the forerunner of Jesus, and hence his ministry comes earlier than Jesus’.30 It is thus entirely natural that material about John’s own preaching

29. Gospel Writing, 137–41; also ‘Q as Hypothesis’, 400–3. 30.  So e.g. Matt 17.10–13; John 1; Acts 1.22; 10.37. The vestige of a tradition that Jesus himself baptized people in John 4.2, possibly implying a simultaneous ministry in tandem by John and Jesus together, is anomalous.

6. Watson, Q and ‘L/M’

125

should come at the start of the narrative, before the account of Jesus’ ministry begins. If then Q had such material, and Matthew and Luke both had access to this and to Mark, it is scarcely surprising that they combined the two together at the start of their accounts just before the narrative of Jesus’ ministry.31 In more detail, the tradition of John’s actual preaching (Luke/Q 3.7–9) follows very naturally as the sequel to (and hence filling out in more detail) the summary statement about John’s preaching in Mark 1.4.32 The double baptism saying, if it existed in Q as well as in Mark, is most obviously placed at the same place as Mark has it. So too, if the expansion of the saying about the coming baptism (as with ‘fire’ and spelling out the consequences of that) were part of Q, this too appears here in a completely expected place. Further, the existence of a temptation story in Q would have made it very natural for Matthew and Luke independently to place this at the point in Mark where there is a (much smaller) temptation story (Mark 1.12–13). That there is some agreement here between Matthew and Luke in incorporating Q material into Mark at similar points is undeniable; but it is scarcely unexpected, given the nature of the material concerned and any ‘coincidence’ here scarcely calls into question the assumption of the independence of Matthew and Luke (if both had access to Q).33 For the L/M hypothesis, some of the close verbal agreements between Matthew and Luke are obviously explicable by Luke using Matthew closely.34 There are though other changes, which Watson tends to gloss over. Luke adds his own material on John’s exchanges with crowds, tax-collectors and soldiers (Luke 3.10– 14),35 but he must also have omitted Matthew’s expansion of Mark in the exchange between John and Jesus recorded in Matt 3.14–15. Watson simply says that ‘little if anything should be read into an omission’.36 But, as noted above, discussion of ‘redactional’ changes should perhaps include omissions as well as larger-scale ‘compositional’ changes. 6.2.4 The Position of the Sermon on the Mount/Plain A final ‘coincidence’ appealed to by Watson, and in his view the most important,37 concerns the placing of the Sermon on the Mount/Plain in Matthew and Luke in

31. The situation is different with the material in Luke/Q 11.2–6: this is a report of John’s message to Jesus sent from prison. It must then logically come from a later time. 32. Matthew follows Mark in the note about John’s clothing and diet (Mark 1.5–6); Luke does not. 33. Similarly (for this material) John S. Kloppenborg Verbin, Excavating Q. The History and Setting of the Sayings Gospel (Edinburgh: T&T Clark, 2000), 32. 34. Though why Luke should follow Matthew so closely and faithfully here, having been so free with Matthew in the preceding birth narratives, is less clear. 35. This is Luke’s own added material for the L/M theory and the 2ST alike. 36. ‘Q as Hypothesis’, 403. 37. He refers to it as the ‘most striking’ of all the coincidences (pp. 148, 154).

126

Christopher M. Tuckett

relation to Mark (pp. 148–55). The key point for Watson is that both Matthew and Luke preface their Sermons with the note about the great crowds flocking to Jesus from various places in Mark 3.7–8. Watson presumes that Matthew’s Sermon is to be placed after Mark 1.20 in the Markan order (pp. 148, 150).38 But Matthew prefaces his Sermon with the note about the great crowds from various places (Matt 4.25 // Mark 3.7–8). Luke places his Sermon later in relation to Mark. He follows Mark closely from Mark 1.21 // Luke 4.31 to Mark 3.6 // Luke 6.11. The next Markan verses, Mark 3.7–8/10, apparently leads him to add his version of the Sermon, and he changes the order of the next two Markan pericopes (Mark 3.7–12 // Luke 6.17– 19; Mark 3.13–19 // Luke 6.12–16), perhaps to create an audience for his Sermon (on the Plain).39 Thus the note about the crowds flocking to Jesus (Mark 3.7–8 // Luke 6.17) provides the introduction to Luke’s Sermon. Watson thus argues that The same Markan passage [Mark 3.7–8] provides the setting both for Matthew’s Sermon on the Mount and for Luke’s Sermon on the Plain. There is no reference to teaching in Mark 3.7–19 . . . Yet both the later evangelists have selected this Markan context for their parallel Sermons. (p. 154)

The phenomenon may appear striking at first sight. However, any ‘agreement’ here is quite small when examined more closely. First, one needs to clarify terminology. Watson speaks of the same Markan ‘setting’ or ‘context’ for the two sermons, although conceding that Matthew’s use of this ‘setting’ is ‘out of its original Markan sequence’ (p. 155). Yet this ‘setting’ involves only the use of the note about the great crowds of Mark 3.7–8. Looking at the evidence from a wider perspective, the two versions of the Sermon occur in quite different Markan ‘contexts’ (or ‘settings’): Matthew (probably) places his Sermon on the Mount after Mark 1.20 (or 1.21),40 Luke places his Sermon on the Plain after Mark 3.7–19.41 If ‘setting’ implies simply the use of Mark 3.7–8 alone, 38.  He shows no awareness at all that this is contested! See the discussions in e.g. Frans Neirynck, ‘The Sermon on the Mount in the Gospel Synopsis’, Ephemerides Theologicae Lovanienses (ETL) 52 (1976): 350–7, and above all John S. Kloppenborg, ‘Synopses and the Synoptic Problem’, Synoptic Problems, 120–53, 135–44, who shows how the position of the Sermon in different synopses can affect one’s assessment of the relationships between the gospels. The ‘standard’ positions for the Sermon are after Mark 3.19 (so e.g. Aland), after Mark 1.39 (e.g. Huck), or after Mark 1.21 (e.g. Neirynck). Watson evidently agrees with the latter (though in this case the Sermon is usually proposed as following Mark 1.21 rather than 1.20: the note about ‘teaching in their synagogues’ in Matt 4.23 is generally regarded as taken from Mark 1.21). 39. See e.g. Werner Georg Kümmel, Introduction to the New Testament (London: SCM, 1975), 59. 40. So Watson: I would agree with him in this, though others would not (see n. 38). 41.  Ironically, Watson’s case might be stronger at one level if he had followed Aland’s positioning of the Sermon on the Mount! However, such a positioning does then imply that Matthew’s changesof Mark’s order were more extensive; there might also be other problems for explaining Luke’s use of Matthew here: see Kloppenborg, ‘Synopses’, 137.

6. Watson, Q and ‘L/M’

127

then Watson’s case may stand; but this is not the ‘setting’ or ‘context’ that many would understand by either term generally.42 It is only the very immediate ‘setting’ that is involved here; the broader ‘settings’ or ‘contexts’ in Mark for the placement of the two Sermons are quite different. Second, any link between Matt 4.25(// Mark 3.7–8) and Matthew’s Sermon is very tenuous so that to talk about the verse as providing the ‘setting’ for the Sermon may be misleading. For although the verse immediately precedes the Sermon (and hence constitutes its ‘setting’ in that Matt 5.1 follows Matt 4.25), the substance of the verse does not provide the setting for the Sermon in the sense that the crowds mentioned provide the audience for the Sermon. In fact precisely the opposite is the case: the Sermon is on the ‘mountain’, and Jesus goes up there to escape the crowds, he calls the disciples to him and (initially) addresses the Sermon to them.43 Thus the sequence in Matthew puts the ‘crowds’ and the Sermon side by side, but disassociates one from the other. In fact, Matthew and Luke have quite different compositional aims in the way they use Mark 3.7–8 and the Sermon. Both introduce the Sermon relatively early in their gospels. Luke, using his sources in alternating large ‘blocks’, has used Mark in Luke 4.31–6.19. For his Sermon he wants to create a large audience and hence inverts the order of Mark 3.7–12 + 13–19 for this so that his version of Mark 3.7–8 leads into the Sermon with the crowds mentioned there then providing part of the audience for the Sermon.44 Within Matthew, Matt 4.25 // Mark 3.7–8 functions quite differently. Matthew creates a large ‘double block’ of material, showing Jesus teaching (5–7) and then performing various miracles (8–9). He highlights this by forming an inclusio with two summary statements about Jesus’ twin activity of teaching and healing in 4.23–25 and 9.35.45 In 4.23–25 he uses various Markan summary statements, including Mark 1.39 and 3.7–8.46 But, unlike Luke, the purpose of these statements is not to create a context for the Sermon by providing an appropriate audience; it

42. Many would say that the ‘setting’ or ‘context’ of one element in a narrative is its place within the broader, overall sequence. Hence to claim that Matthew has the same Markan ‘setting’/’context’ as Luke for his Sermon but in a different Markan ‘sequence’ (cf. above) seems to be almost self-contradictory. 43.  There is the well-known ambiguity that, at the end of the Sermon, Jesus appears to have been addressing the crowds all along (Matt 7.28). However, the start of the Sermon seems to presuppose that the disciples, and not the crowds, are the prime addressees (cf. too the second person address in 5.11–12, 13–14). 44.  More clearly than Matthew’s, Luke’s Sermon presupposes an audience of both followers and others (cf. the second person address of the beatitudes and the woes in Luke 6.20–26). 45. The verbal agreement between 4.23 and 9.35, and the resulting inclusio, is regularly noted by commentators. 46.  As these are both out of the Markan sequence in Matthew’s ordering, he may have done so from memory, which might explain the somewhat inexact agreement with Mark.

128

Christopher M. Tuckett

is rather to highlight the wide success enjoyed by Jesus in his general ministry of teaching and healing. For the Sermon itself, Matthew places Jesus on a ‘mountain’ (perhaps to highlight parallels with the Moses tradition), away from the crowds with the disciples (at least initially) the sole audience.47 The conjunction of 4.25 (= Mark 3.7–8) and the Sermon is thus superficial and/or coincidental. If then Luke shows ‘agreement’ with Matthew here, it is only in a somewhat superficial way: he agrees with Matthew in having the note about the crowds from Mark 3.7–8 immediately preceding his Sermon on the Plain; but, as noted earlier, this is in a quite different broader Markan context (after Mark 3.19 not after Mark 1.20), and he uses the link quite differently by making the crowds into the audience for the Sermon, which Matthew explicitly does not. In their use of Mark 3.7–8 and their versions of the Sermon, Matthew and Luke thus have different agendas, and have added the Sermon at different points in relation to Mark. The only agreement between them is that the parallels to Mark 3.7–8 occur just before the Sermon in each gospel. In one way this is a ‘coincidence’. But if Matthew and Luke have independently rearranged the Markan and Q materials in their gospels, it is not surprising if just once they might coincide in a superficial way (as here). The probability that two writers, in rearranging materials over a wide range of subject matter, should never coincide is unlikely. A very small measure of agreement certainly does not tell against the 2ST.48 I turn now to Watson’s arguments for the L/M hypothesis, and his discussion of Luke’s redactional activity on that hypothesis, rather than his negative arguments against Q.

6.3 Watson, Luke and the L/M Hypothesis 6.3.1 Order Watson claims that a primary concern of Luke was the phenomenon of order, in particular the disagreements in order between Mark and Matthew (pp. 121– 31). Papias (in Eusebius, H.E. 3.39) may have been aware of the differences in order between Mark and Matthew (perhaps regarding Matthew as having rectified Mark). A similar problem may have arisen for Luke. Thus in his prologue outlining his reasons for writing, Luke says that he has decided to set down his version ‘in an ordered manner’ (Luke 1.3:  καθεξῆς), implying that he is seeking to give a ‘better’ order in his account than his predecessors have. Luke then may be seeking to correct the order of the two gospels he probably knows, namely, Mark and Matthew. 47. See n. 43 above for the issue of whether the whole Sermon in Matthew is addressed to Jesus’ followers alone. 48.  See also my discussion in my ‘Arguments from Order: Definition and Evaluation’, Sayings to the Gospels, 3–22, 12, in relation to possible ‘coincidences’ in reordering Markan pericopes.

6. Watson, Q and ‘L/M’

129

Whether Luke’s use of καθεξῆς refers precisely to the sequential order of the events recounted may be debatable. But more important here is whether this claim about Luke’s general aims fits the data which the L/M hypothesis seeks to explain. In fact this seems doubtful. First, concerns about the order of gospel accounts arise primarily when there are two or more accounts which disagree with each other. Papias’ worries about Mark’s order are usually taken as arising because of the existence of Matthew with a differing order. Similarly later church writers focus on the differing order of events in the different gospels.49 If Luke’s primary concern is about the differences in order between Mark and Matthew, then (unlike Papias) he seems to have assumed that the Markan order is basically correct:  hence he generally follows the Markan order closely and ignores Matthew’s differences. On its own, this is unexceptional. However, as is well known, Matthew has a great deal of other, non-Markan material. Here Luke must have changed the order very considerably. But why should this material need reordering? Ex hypothesi, Luke has no other source alongside Matthew covering the same material and presented in a different order. Why then does the order of Matthew’s non-Markan material need to be changed?50 Second, the changes which Luke makes on the L/M hypothesis are not simply at the level of order. As noted earlier, Luke must have engaged at times in a process of radical rewriting and/or replacement of equivalent material (e.g. the birth narratives, genealogies, resurrection stories). This would seem to reflect not just a concern about order: Luke here gives a radically new version of the material, at times involving quite different content. Further, other revisions made by omitting substantial parts of Matthew, adding totally new material, as well as rewording parts of Matthew which are retained, all constitute far more than (simply) reordering a text (or two texts) whose order is problematic. It would seem that the order is only a small part of the ‘problem’ which Matthew’s gospel must have posed for Luke! Thus an appeal to the possible problem posed by the differing orders of Matthew and Mark as providing Luke’s primary motive and aim in writing a new account of the events of Jesus’ life does not really fit the evidence that requires explanation.

49. See e.g. Helmut Merkel, Die Widersprüche zwischen den Evangelien (Tübingen: Mohr Siebeck, 1971). 50. Poirier, ‘Introduction’, 11, suggests that Luke may have thought that Matthew’s structuring of non-Markan material was ‘artificial’; elsewhere, he suggests that Luke may have changed Matthew for the sake of it, to show that his account is not just a ‘regional recension’ of Matthew: see his ‘Delbert Burkett’s Defence of Q’, in Poirier and Peterson, Marcan Priority without Q, 191–205, 208–9. Both suggestions seem rather weak and lack support elsewhere. Watson offers no such overarching explanation, beyond referring to the differences between Mark and Matthew (p. 129): but this does not affect the non-Markan material in Matthew.

130

Christopher M. Tuckett

6.3.2 ‘Plausibility’ Watson’s overall claim is that Luke’s redactional activity on the L/M hypothesis can be seen to be plausible and coherent. However, determining what is ‘plausible’ and/or ‘coherent’ here is not easy. As noted earlier, one could compare Luke’s redactional activity on the L/M hypothesis with similar situations. In this respect one could make internal or external comparisons. Internally, one can compare Luke’s redaction of Mark with his redaction of Matthew. Here the two are wildly different, as has often been noted: thus Luke must have treated Mark relatively conservatively, preserving much of the material and, above all, most of the order of his Markan source; on the other hand he must have treated Matthew with considerable freedom, omitting large parts, radically rewriting other parts, and changing the order very considerably. This has often been taken as a strong argument against the plausibility of the overall hypothesis: why would Luke treat his sources in such radically different ways? Would not one expect more consistency by Luke in relation to his two sources?51 One answer would be simply to say that Luke has used two sources, and he treats them in different ways, perhaps because he regards them in a different light, for example, having more respect for Mark (and/or perhaps having known Mark for longer).52 But if this is left in too general terms, it runs the risk of simply describing rather than explaining (cf. above): if Luke used Mark and Matthew, he must have used them in different ways, but why did he do so? To say that he had more respect for Mark than for Matthew does little more than restate the problem in slightly different terms.53 An alternative way of assessing the plausibility of Luke’s procedure might be to compare his activity with that of other contemporary writers using sources. Such work has been undertaken by a number of people in modern debates about the Synoptic Problem.54 Thus both Downing and Derrenbacker have argued that 51. Cf. Christopher M. Tuckett, Q and the History of Early Christianity (Edinburgh: T&T Clark, 1996), 9. 52. See e.g. Goodacre, Case, 88–90; cf. also Poirier, ‘Introduction’, 10–11. 53. In discussion at the Roskilde meeting, Watson asked why one should expect consistency in the use of sources at all, appealing to the way in which many later ‘gospel’ writers used earlier ‘sources’ in very different ways (e.g. the author of the Protevangelium of James). This may however be a false comparison: such later ‘gospel’ writers were not using earlier ‘sources’ in the way that synoptic evangelists used each other and/or (possibly) ‘Q’. The ‘dependence’ of authors of texts such as the Protevangelium, or the Gospel of Mary or the Gospel of Thomas, on e.g. the synoptic gospels is not one of direct literary dependence as is envisaged for, e.g. Matthew’s/Luke’s use of Mark. 54.  See e.g. essays in F. Gerald Downing, Doing Things with Words in the First Century, JSNTSup 200 (Sheffield: Sheffield Academic Press, 2000); Robert Allen Derrenbacker, Ancient Compositional Practices and the Synoptic Problem, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 186 (Leuven:  Peeters, 2005); John S. Kloppenborg, ‘Variation in the Reproduction of Double Tradition and an Oral Q?’, in Kloppenborg, Synoptic Problems, 91– 119. None of this work is referred to in Watson’s work or mentioned in his bibliography.

6. Watson, Q and ‘L/M’

131

other writers of the period, when using more than one source, adopted a relatively simple and straightforward procedure of using one source at a time, generally not engaging in a complex process of conflation of sources, and also not ‘unpicking’ one source from another. Certainly Luke’s radical reordering of Matthew’s material seems quite unlike anything comparable in writers engaged in a similar process of using a source for their writing. Luke’s use of sources on the L/M hypothesis thus does not easily cohere with itself, nor with the use of sources by other contemporary writers. Such a lack of internal and external ‘coherence’ must cast some doubt on the ‘plausibility’ of the underlying hypothesis.55 6.3.3 Luke’s Choice of Material A further issue relates to Luke’s choice of material, including the question of the ‘mechanics’ of how Luke extracted the materials he chose to include. Luke’s use of Mark is relatively simple and not hard to envisage (and indeed is effectively the same for the 2ST or the L/M theory): he uses Mark in large blocks and follows the order of Mark fairly closely, perhaps using Mark as a source available to him visually as he writes. His use of Matthean material on the L/M hypothesis is however harder to envisage. Certainly by the time Luke gets to the section in Luke 9.51–18.14, he is accessing Matthean material from a wide range of scattered Matthean contexts. Watson seeks to show that this is not random, in that Luke’s arrangement has its own (Lukan) logic and sense.56 In one way this is unexceptional (Luke’s order makes Lukan sense); yet the process by which Luke might have decided which material he was going to include, and then how he accessed it, is only discussed briefly. At one point Watson does explicitly address this question. Thus in considering how Luke used the material in Matthew’s Sermon on the Mount, Watson says that Luke used some of this material for his own Sermon, but reserved other bits for later, perhaps ‘copying into a notebook the passages he wishes to reserve for future use’ (p. 170; ‘Q as Hypothesis’, 406), referring to Loveday Alexander’s essay on ‘Ancient Book Production’ and the possible use of wax tablets for preparatory work.57 This, it is implied, gives a plausible picture for how Luke worked. Further, Watson claims, and evidently sees it as very significant, that all the material Luke uses from Matthew occurs later in Luke relative to Mark than in Matthew (pp.  160 n.  5, 175):  the implication is that the general process  – Luke using parts of Matthew’s material (usually in Matthew’s discourses) in Matthew’s

55.  Unless it is deemed ‘plausible’ to be inconsistent and/or out of line with what others do! 56. This is the main thrust of his ch. 4 on ‘Luke the Interpreter’. Similarly Goodacre, Case against Q, ch. 5. 57. Loveday Alexander, ‘Ancient Book Production’, in The Gospels for All Christians, ed. Richard J. Bauckham (Grand Rapids, MI: Eerdmans, 1998), 71–105.

132

Christopher M. Tuckett

context, but then noting down other parts to use later – provides a good model. This is clearer in his earlier essay: The [L/M] hypothesis requires Luke to be able to identify non-Markan material in Matthew, but to do so only after he has given his own rendering of the Markan context in which Matthew has located it, on the basis of Mark’s narrative alone. Luke first rewrites a Markan sequence, and then incorporates selected supplementary material from Matthew.58

The trouble comes when one considers more detail here! First, the extent of the rearranged Matthean material means that the ‘notebook’ which Luke might have used must have been very extensive indeed. Certainly it would have had to contain far more than could be included on a ‘wax tablet’ (often no more than one side and probably not capable of much detail). The notebook would have had to contain almost half of the material Luke takes from Matthew (since almost half is rearranged). Second, and more important, Watson’s model fails to match all the data to be explained. Watson devotes the greater part of his discussion of Luke’s use of Matthew to the material in the Sermon on the Mount, and in slightly less detail to the Mission Discourse. His analysis of the material taken from elsewhere in Matthew in Luke 9.51–18.14 is much shorter, but it is here that problems for Watson’s overall explanation arise. For in this material, Luke must very often have looked ahead in Mark/Matthew, at times by some considerable way, and then decided to include Matthean non-Markan materials earlier than the Markan context in which they appear in Matthew. Watson gives a list of Matthean passages taken over by Luke on p. 187, and well over half of these must have involved Luke’s looking ahead to material he had not yet reached in Mark/Matthew. This includes all the woes of Matt 23, much of the eschatological material in Matt 24–25, as well as a number of smaller units. Watson’s theory thus has to presuppose Luke reading ahead, or reading the whole of Matthew in advance before he has got to the equivalent point in the Markan sequence (and hence not ‘only after he has given his rendering of the Markan context’: cf. above), noting down in his ‘notebook’ (or on his wax tablet) all the material he would use, not only later (for Matt 5–7) but also earlier (for Matt 24–25). In fact Watson’s strong focus on Luke’s use of the Sermon material may have skewed the analysis: the fact that the Sermon on the Mount is Matthew’s first teaching discourse, and used by Luke in his first block of teaching, inevitably means that any Matthean material not used in Luke’s Sermon will come later in Luke. However, with Matthew’s last (eschatological) discourse (Matt 24–25), which is broadly parallel to Luke’s final (eschatological) discourse (Luke 21 // Mark 13), the Matthean material which Luke decides to use, but not in his ch. 21 (and this is not inconsiderable in scope and extent), inevitably comes earlier in Luke. Thus one cannot say that Luke’s general procedure is to identify the non-Markan material in 58. ‘Q as Hypothesis’, 413 (his italics).

6. Watson, Q and ‘L/M’

133

Matthew in Markan contexts and then include that in his own account ‘only after’ he has given his version of Mark (alone). Moreover, he must have been able to identify in this process all the bits in Matthew which do not come from Mark and regarded these as ripe for reordering.59 Further, the list of passages which Luke is said to have taken from Matthew can not always be described as ‘non-Markan material in Matthew’. In Gospel Writing, p. 187, Watson gives a long list of such passages said to comprise ‘individual nonMarkan sayings embedded in Markan contexts’ (p. 186). Among these are Matt 24.17–18 (// Luke 17.31), though this is Matthew’s version of Mark 13.15–16: it is not a ‘non-Markan saying’ at all. Similarly Matt 24.23–25 (// Luke 17.20–21), is Matthew’s rendering of Mark 13.21–23. So too Matt 13.31–33 is parallel to Luke 13.18–21; but Matt 13.31–32 (the mustard seed) is also in Mark so why is this regarded as ‘non-Markan material in Matthew’?60 Other ‘parallels’ cited here are at least questionable. For example, Matt 21.28–32 is said to become Luke 15.11– 32: both are parables involving two sons, but the two parables display far more dissimilarity than similarity. And Matt 25.1–13 can only be said to be the source of Luke 13.25 on the basis of Luke’s omitting the main part of the whole parable here. In sum, Luke thus appears to be engaged in a source-critical dissection of Matthew worthy of any modern student of synoptic problem studies! It would seem that before he ever starts writing, Luke must have read through the whole text of Matthew, identifying some parts as non-Markan and (?hence) ripe for reordering and/or rewriting (though also with a number of exceptions:  see above). On the assumption that it would be very difficult to read two scrolls simultaneously side by side, we perhaps then have to assume that Luke must have known Mark’s text extremely well  – certainly well enough to be able to identify those bits in Matthew which are allegedly derived from Mark and distinguishing these parts from the ‘non-Markan’ material. He must then have marked down (somewhere, somehow) which material he would include (though some of the material he ‘includes’ is also radically rewritten at the same time:  cf. e.g. Matt 21.28–32 which becomes a totally different parable in Luke 15.11–32). Further, it is not clear why Luke would have engaged in this reordering of the Matthean material which is not in Mark. Especially this applies in the case of

59.  In Gospel Writing, 171 n. 27, Watson claims that his own suggestion – that ‘Luke “reserved” or preselected Matthean passages for later inclusion’  – ‘greatly simplifies the compositional procedure’ presupposed on the L/M hypothesis; and in this his own suggestion is superior to that of Goulder who had to postulate a ‘complex to-and-fro movement within the Matthean text’. But Watson’s implied procedure is no different when examined in detail. He too has to postulate a complex ‘to-and-fro’ procedure when it comes to the later material in Luke and Matthew; and indeed this is hardly surprising since Watson and Goulder are effectively trying to deal with (to identify and then explain) exactly the same differences (in the choice and order of material) between Luke and Matthew and hence exactly the same redactional changes which Luke must have made to Matthew! 60. See further §6.3.4.2 below.

134

Christopher M. Tuckett

the eschatological material. He is evidently content to use a substantial block of eschatological teaching in his ch. 21, using the Mark 13 material as his source. But all the non-Markan material in Matt 24–25 (or at least those sections he decides to include) is picked out and placed earlier. Why Luke then has such a strong objection to the Matthean ordering, but not the substance, of non-Markan material is not at all clear (given that it is unlikely that Luke has another source with which Matthew is here disagreeing).61 This then is not a ‘simple’ process of an author using one source sequentially, and delaying some material to be used later in his own work; it is of an author working through the whole of the text of Matthew, apparently in advance, somehow identifying all the bits of Matthew that are not from Mark (a text which he presumably is not accessing while working through Matthew),62 carefully noting down all these parts, some of them very extensive, in a ‘notebook’ before coming to write his own work and then presenting them often in a totally different order. This is scarcely simple, nor is it entirely plausible (at least in the sense of having any comparable parallel in contemporary literature). One may also note here that Watson’s ‘explanations’ for Luke’s decisions about what to include are somewhat sparse. In discussing the material in the Sermon on the Mount used by Luke in a later context (i.e. not in his Sermon on the Plain), Watson sets out the material, grouped in small blocks (mostly only pairs of sayings, but in two cases a single saying) and allegedly showing a ‘pattern’ of pairings of units, where sometimes a ‘thematic link’ can be seen and sometimes ‘traces of the original Matthean order are preserved’.63 The analysis may show something of Luke’s intention in arranging the material in this way. But it does not really show that this arrangement has been derived from Matthew. 6.3.4 ‘Mark-Q Overlaps’ The texts usually assigned to the category of ‘Mark-Q overlaps’ on the 2ST may be particularly significant in a study of the Synoptic Problem. These texts are often thought to create problems for the 2ST, having to postulate the overlapping of two independent sources. However, these texts cause just as great, if not more, problems for theories other than the 2ST.64 In particular, this may be the case for Watson’s L/M hypothesis. I consider here two classic examples, the Beelzebub controversy and the parable of the mustard seed.

61.  Unless it is for the sake of it, to show that he is writing something different from Matthew (cf. n. 50 above). 62. The mechanical difficulties of having more than one text visually available to a scribe while writing (cf. Derrenbacker, Ancient Compositional Practices, 37–9) make this very likely. 63. Gospel Writing, 170–1, modifying slightly (and correcting) the presentation in ‘Q as Hypothesis’, 407. With the single sayings, there is inevitably no pairing, and hence no thematic link, at all. 64. Tuckett, ‘Current State’, 99–100, 109–11.

6. Watson, Q and ‘L/M’

135

6.3.4.1 The Beelzebub Controversy Watson analyses this briefly in his discussion of how Luke chooses material after editing Mark in a long section (Mark 1.29–3.19 // Luke 4.38–6.19). Starting with the Great Sermon, Luke ‘begins to incorporate the supplementary Matthean material he has identified within the Markan material he is editing’ (p. 179). Finding nine such items (including e.g. the inaugural sermon, the centurion’s servant etc.), Luke also ‘takes over from M/Mk1 two episodes he finds in Matthew 12’, namely, the Beelzebub controversy and the saying on Jesus’ true family. But why has Luke then taken this material from Matthew? And why does Luke then omit the Beelzebub pericope from the relevant section of Mark when he gets to this later? It is not ‘supplementary Matthean material’: it is rather material which Matthew has taken from Mark. Matthew has indeed supplemented Mark at times in the pericope (e.g. in the sayings in Matt 12.27–28), but he has also clearly used Mark elsewhere (e.g. in the strong man saying in Matt 12.29 // Mark 3.27). On almost any showing, this is a Markan pericope in Matthew. This should be, on the basis of Watson’s account of Luke’s overall procedure, a pericope that Luke retains in its Markan sequence, using Mark, and not one which he takes from Matthew (as Matthean material ‘supplementary’ to Mark). Thus when Watson says later that the three forms of the Beelzebub controversy are ‘characterised by simplicity and plausibility: Mk →M/Mk → L/M/Mk’ (p. 184), one can hardly say that this is ‘plausible’. ‘Plausibility’ involves giving reasons for the activities identified: and Luke’s compositional activity here does not fit the overall, general explanation which Watson provides. One would expect Luke to give a version in his ‘L/Mk’ sequence, not extracting it to use as ‘L/M/Mk’ material in an ‘L/ M’ sequence (using Watson’s terminology). Hence Luke’s procedure here has not been shown to be plausible.65 Nor is it ‘simple’. He may have identified some Matthean ‘supplements’ to Mark in this pericope, and reproduced them. But in doing so he has carefully avoided, and shunned, all the points where Matthew is very close to Mark (e.g. the strong man saying: Luke’s version in Luke 11.21–22 is quite different). He has thus succeeded in ‘unpicking’ Matthew’s additions to Mark from the Markan material in Matthew, retaining these additions and discarding all the Markan elements. Such a process would be hard to do in any case, and also seems quite unprecedented among contemporary writers using sources.66 Further, Luke’s aversion to this

65. Rather more honest is Watson’s admission in ‘Q as Hypothesis’, 412, that Luke’s treatment of the Beelzebub controversy is ‘anomalous’. 66.  See long ago F. Gerald Downing, ‘Towards the Rehabilitation of Q’, NTS 11 (1965): 169–81. For responses to Downing from the side of the Farrer/‘L/M’ hypothesis, see Ken Olson, ‘Unpicking on the Farrer Theory’, in Goodacre and Perrin (eds), Questioning Q, 127–50; Eric Eve, ‘The Devil in the Detail: Exorcising Q from the Beelzebul Controversy’, in Poirier and Peterson, Marcan Priority without Q, 16–43. Both posit Luke’s using the detailed wording of Matt 9.32–34 as well as Matt 12 here (e.g. for Olson, the parallel between Luke 11.15 and Matt 9.34 means that Luke has not avoided all the close parallels between Matthew and Mark), though this is not easy to envisage in practice.

136

Christopher M. Tuckett

Markan material in Matthew also covers his use of the pericope in Mark, since he must have omitted the story completely from Mark when dealing with Mark’s own sequence (which he normally takes care to preserve and reproduce). Luke’s procedure here is thus not explained by Watson’s analysis: he describes what Luke apparently did, but provides no clear explanation of why he did this.67 By contrast, the Q hypothesis may fare better. This is one of the classic ‘MarkQ overlaps’, with a version of the story in both Mark and Q. Thus Matthew, on the Q hypothesis, combines the two versions together (as he generally does in such instances); Luke prefers the Q version (which again he often does) and hence omits the Markan version. Thus he has no parallel to Mark 3.23–30 when he reproduces the Markan sequence; and he includes the story in his ‘Q+L’ section in Luke 11. Further, the fact that he gives the Q version (without any Markan influence) means that he ends up with parallels to Matthew’s non-Markan material (= Q), but without having to postulate a complex ‘unpicking’ process. The redaction posited for Luke is thus consistent with his practice elsewhere, and also fits well with the ways in which other contemporary writers seem to have used sources. As such it is plausible and reasonably simple. 6.3.4.2 Mustard Seed Very similar arguments apply in the case of Watson’s treatment of the parable of the mustard seed in Matt 13.31–32 // Luke 13.18–19, which also occurs in Mark 4.31–32. I have noted earlier the inconsistency in Watson’s attempt to explain Luke’s procedure in using ‘Matt 13.31–33’ as part of the ‘nonMarkan’ materials extracted from their Markan contexts and used in a later Lukan context, namely, Luke 13.18–21. Matt 13.31–33 includes the parable of the leaven (which is indeed non-Markan), but also the mustard seed which does appear in Mark. It is thus not an example of ‘non-Markan’ material in Matthew at all. Further, as with the Beelzebub controversy, Luke has not only used this parable (from Matthew or Q) in Luke 13 but has also omitted it in Luke 8 when reproducing Mark 4. Thus Luke’s treatment of the parable also differs from Luke’s general policy as proposed by Watson, namely, including all of Mark in its Markan sequence and then identifying non-Markan passages in Matthew which can be easily detached from Mark for inclusion elsewhere. Such a policy does not fit with what Luke must have done with the parable of the mustard seed. Further, in using Matthew here, Luke must have succeeded in cutting out precisely those bits of the parable in Matthew where Matthew is very close in wording 67.  Olson, ‘Unpicking’, 142, suggests that Luke was attracted to Matthew’s version by the extra material in Matthew, and so used Matthew rather than Mark here. But this is not far off saying that Matthew’s extra material was ‘Luke-pleasing’, yet this hardly constitutes an explanation of why Luke thought this: see n. 12 above. Further, Luke’s alleged procedure here (effectively abandoning Mark and using Matthew alone) is still quite unlike what he appears to have done most of the time elsewhere. E.g. in Mark 13 // Matt 24–25, he has evidently wanted to use the extra (non-Markan) material in Matthew, but he has not simply used Matthew rather than Mark – he has rather separated out the extra Matthean material and used it much earlier (in Luke 12, 17) and then used primarily Mark in Luke 21.

6. Watson, Q and ‘L/M’

137

to Mark (the note about the mustard seed being smaller than all other seeds, and the grown mustard seed being larger than other shrubs); but he does keep (almost verbatim) other parts of Matthew where Matthew’s wording diverges from Mark’s (e.g. the man ‘taking’ the seed and sowing it, the seed becoming a ‘tree’ where the ‘birds of heaven dwell in its branches’ [with intertextual echoes of e.g. Dan 4.12, 21]). As with the Beelzebub controversy, we seem to have to envisage Luke carefully ‘unpicking’ Matthew, discarding all the parts where Matthew is close to Mark and keeping almost verbatim those parts where Matthew differs significantly from Mark.68 Such a procedure seems hard to envisage in practical terms, as well as being unlike any other contemporary writer’s way of using sources. It is also quite unlike Luke’s alleged respect for Mark elsewhere and his concern to include all of Mark’s material. It is then both complex and implausible (or at least unexplained). Watson does deal in more detail with the parable’s wording in a long footnote later (p.  198 n.  63). He claims that this is ascribed to a Mark-Q overlap on the 2ST ‘to protect the Q hypothesis at points where so-called “minor agreements” . . . threaten to become major’ (and he lists the extensive verbal agreements here between Matthew and Luke). He cites Sanders and Davies as saying that the parable ‘exists in basically only one form’, and then claims that ‘the second, Q version of the parable is demanded not by the textual evidence but by the refusal to contemplate Luke’s knowledge of Matthew’. Many would disagree. There do seem to be two versions of the parable, one in Mark and one in Luke. The two can be seen to coexist in Matthew, existing side by side in slightly uneasy tension with each other. (Hence, for example, the story starts in Matthew as a story about an event in the past [‘a man sowed . . .’] but later shifts to a generalized statement about how mustard seeds grow; the seed becomes first ‘greater than all shrubs’, apparently keeping with the bounds of the natural realm where mustard seeds do produce shrubs, but then becomes a ‘tree’, which rather goes beyond the natural!)69 Again, a relatively straightforward explanation can be given on the 2ST, positing a Markan and a Q version of the parable. Luke, as before, prefers the Q version and so omits Mark’s version in Mark 4 // Luke 8, and gives the Q version (perhaps largely unchanged) in Luke 13. Matthew seeks to conflate the two in his version, with not entirely happy results.70 As with the Beelzebub controversy, the 2ST provides a reasonably simple explanation of the facts, and also explains the versions reasonably easily and plausibly (within the parameters of the overall hypothesis). In both these instances the 2ST seems to fare considerably better than the L/M hypothesis in explaining the textual evidence.

68. The example was not discussed by Downing and is not analysed by Olson. 69.  The existence of two versions of the parable can thus be seen from the ‘textual evidence’ of Matthew’s version and is not simply postulated to rescue the 2ST from the embarrassment of too many minor agreements. 70. But the tensions exist in Matthew’s version however one accounts for them!

138

Christopher M. Tuckett

6.4 Conclusion Watson’s book is to be welcomed as a significant contribution to the debates about synoptic origins and the synoptic problem. There has not been time or space here to engage with other parts of his overall theories about synoptic origins, including his discussion of non-synoptic sources (such as the Gospel of Thomas or Papyrus Egerton 2), and his theories about a possible ‘sayings collection’ underlying the synoptic gospels (in Gospel Writing, ch. 5).71 This chapter has attempted to provide a response (only) to his claims that Luke’s gospel can be explained on the basis of Luke using Matthew directly without recourse to Q. This response to Watson’s work has inevitably focused on points of disagreement; but it is to be hoped that such a response will be seen as a reflection of respect: it is precisely because of the important issues which Watson’s book raises, and the detail and depth with which it engages with those issues, that a detailed response is (hopefully) appropriate to further the debate and to generate ongoing discussion.

71.  Some might see the latter theory as Q, having been banished in the previous two chapters, creeping in again by the back door.

Chapter 7 S EV E N T H E SE S O N T H E S Y N O P T IC P R O B L E M , I N D I S AG R E E M E N T W I T H C H R I ST O P H E R T U C K E T T Francis Watson

7.1 Thesis One: The ‘Q’ hypothesis should not be regarded as a default position, to be assumed true unless and until proved false. It is understandable that scholars who have long advocated a specific solution to the ‘synoptic problem’, in the classroom and in print, should hold to it tenaciously in the face of challenges to rethink their position. In general, conversions from one position to another are not to be expected; that is as true of the few who accept in broad outline the hypotheses of Griesbach or Farrer as it is of the many disciples of Holtzmann and Streeter. As an advocate of the ‘Q’ hypothesis, however, one has the rhetorical advantage of a long-standing consensus so sure of itself that it prefers to drop the term ‘hypothesis’ and to speak of ‘Q’ as an established fact.1 Thus Tuckett claims that the Q hypothesis has ‘produced a whole raft of explanations about the redactional activity (by Matthew and Luke) that has seemed to many plausible and coherent’ (118). It is, therefore, not dependent on the alleged implausibility of Luke’s knowledge and use of Matthew (L/M). Even if one could show L/M to be ‘reasonable and intelligible’, that would (according to Tuckett) still leave Q intact unless it can be shown to produce a lower degree of reasonableness and intelligibility than L/M. I believe that this can indeed be demonstrated, but the point here is to draw attention to the implications of Tuckett’s position: one might successfully demonstrate that L/M can offer a relatively plausible and coherent

1. See my article, ‘Q as Hypothesis: A Study in Methodology’, New Testament Studies (NTS) 55 (2009): 397–415. Tuckett misleadingly claims, on the basis of a garbled recollection of a conversation (115 n), that I wish to distance myself from my own article. This is entirely untrue. On the contrary, I wish Tuckett had engaged with it more fully – for example, in its demonstration of the problems the ‘Beelzebub controversy’ causes for the Q hypothesis (408–14), which he passes over. The difference between my article and my book (n. 4 below) is that the article positions itself on neutral terrain in order to identify the methodological flaws that make it possible to elide the Q hypothesis with an actual Q document.

140

Francis Watson

account of Luke’s redactional procedures but still leave Q in possession of the field, a document and not just a hypothesis. In another context, however, Tuckett restates and endorses ‘traditional arguments for Q’ which, he acknowledges, ‘are mostly of a negative form, claiming that Luke’s use of Matthew seems very improbable’.2 It follows that Q’s credibility will diminish if L/M’s credibility increases. That is exactly my understanding too of the genesis of Q, which Tuckett now dismisses as an appeal to ‘the past history of the Q hypothesis’ (118, italics added). Q’s status as a default position is evident in its inquisitorial rhetoric.3 On no less than twenty-one occasions in his paper Tuckett poses the question why . . .? Why did Luke rewrite Matthew as he is alleged to have done? Why did he do one thing in one instance and another in another? Why did he reorder his material so drastically? These are ostensibly questions put to me, but in reality their function is to expose – rhetorically at least – L/M’s total lack of explanatory power. These reiterated demands for ever more explanation also express global dissatisfaction with the analysis that has actually been presented; they reflect an unjustified claim to an invulnerable default position which it would be rash indeed to challenge. It is hard to think of any other issue within New Testament scholarship where this kind of inquisitorial harangue is deemed appropriate. Some might see this as a symptom of repressed insecurity.

7.2 Thesis Two: In analysing evangelists’ use of their sources, we may expect to find general tendencies but not rigid uniformity in their editorial practice. In Luke 3.16–17, the evangelist’s version of John the Baptist’s ‘double-baptism’ saying and its sequel corresponds so closely to Matt 3.11–12 that – given Q – this must be seen as ‘a classic example of a so-called “Mark-Q overlap” ’ (124). Coincidentally, both evangelists have conflated Mark and Q in the same way. In general, however, Tuckett supposes Luke to have followed one source at a time, Mark at some points, Q in others. The conflation is uncharacteristic and shows, precisely on the Q hypothesis, that Luke does not always treat his sources in exactly the same way. To take another example, Luke 4.31–44 closely follows the narrative sequence of Mark 1.21–39, from the demoniac in the Capernaum synagogue to the summary statement about mission. Yet, in the passages that precede and follow, Luke diverges radically from his Markan source. Mark’s account of Jesus’ return visit to Nazareth (Mark 6.1–6) is transposed to the beginning of Jesus ministry, where it is elaborated and given programmatic status (Luke 4.16–30). The call of Simon Peter is accompanied by a dramatic miracle (Luke 5.1–11; contrast Mark 1.16–20). Luke does not treat his sources in a rigidly uniform manner. 2.  Christopher M. Tuckett, Q and the Early History of Christianity (Edinburgh:  T&T Clark, 1996), 7. 3.  This has been analysed by Mark Goodacre, The Case against Q (Harrisburg, PA:  Trinity Press International, 2002), 77–80. I must again acknowledge indebtedness to Goodacre’s work.

7. Seven Theses on the Synoptic Problem

141

Obvious though that might seem, Tuckett regards the lack of absolute uniformity in my own account of Luke’s redactional procedures as a major weakness. In the case of the infancy narratives of Luke 1–2, I argue that the evangelist is both influenced by and reacting against the Matthean presentation of the same events. According to Tuckett, Luke’s proposed redaction of Matthew scarcely ‘coheres’ with his use of Matthew elsewhere and this raises doubts about how ‘coherent’ Watson’s explanation of Luke’s redactional procedure here is. (121–2)

Given L/M, the Lukan evangelist also treats Mark differently to the way he treats Matthew. This too is deemed problematic: Luke must have treated Mark relatively conservatively, preserving much of the material and, above all, most of the order of his Markan source; on the other hand he must have treated Matthew with considerable freedom, omitting large parts, radically rewriting other parts, and changing the order very considerably . . . Why would Luke treat his sources in such radically different ways? Would not one expect more consistency by Luke in relation to his two sources? (130)

Tuckett considers himself on particularly strong ground when it comes to the Beelzebub controversy and the Parable of the Mustard seed, absent from Luke’s rendering of their Markan context and located instead in a section of largely double tradition material (Luke 11.14–23; 13.18–21). Given L/M, Luke’s alleged procedure here (effectively abandoning Mark and using Matthew alone) is still quite unlike what he appears to have done most of the time elsewhere. (136n)

Tuckett needs to explain why an evangelist flexible in his treatment of sources on the basis of Q and Markan priority is not allowed similar room for manoeuvre on the basis of L/M. The answer to the question, ‘Would we not expect more consistency . . .?’ is simply: No, we would not.

7.3 Thesis Three: Evangelists may simultaneously both ‘agree’ and ‘disagree’ with their sources. It is not only possible but commonplace for two parties to agree about one aspect of an object of shared concern but to disagree about another. In Tuckett’s mind, however, that seems highly problematic. Agreement and disagreement are mutually exclusive: Thus the contents of the gospels might be described in one way that suggests an ‘agreement’ between Matthew and Luke, whereas other descriptions might suggest that there is disagreement. This applies especially to the birth narratives and the genealogies in Matthew and Luke. (120)

142

Francis Watson

Although Tuckett pursues this point doggedly through a number of pages, there is no need to follow him as he does so. My argument traces a number of points at which parallels emerge between the infancy narratives and genealogies of Luke and Matthew, and in each case there is both agreement and disagreement. Luke agrees with Matthew that Jesus was conceived miraculously through the agency of the Holy Spirit, but diverges from him in making Mary rather than Joseph the protagonist of the pre-birth part of the narrative (Luke 1.26–56; Matt 1.18–25). Luke agrees with Matthew that Jesus was descended from David, but provides a different list of ancestors to connect the two (Luke 3.23–31; Matt 1.6–16). Tuckett demands that I explain why Luke does not have Jesus stemming from the Solomonic line. The answer is that he is an ambitious author who believes his account to be superior to his predecessor’s (cf. 1.1–4) and who asserts his independence at the outset of his work by presenting the same themes differently.

7.4 Thesis Four: It would be an unlikely coincidence for two evangelists to make a series of parallel modifications to the same sources independently of one another. My chapter on ‘The Coincidences of Q’ identifies a whole series of points in which (given Q) Matthew and Luke respond independently yet in parallel to the task of producing new, enlarged and improved gospel books from the same two primary sources, Mark and Q. Unbeknownst to one another, two would-be evangelists decide to combine Mark and Q, to preface the combination with narratives about the Messiah’s birth, and to conclude it by relating his post-resurrection appearances to his disciples. Broad structural parallels are accompanied by parallels in detail: in the case of the already-mentioned double baptism saying, for example, or the choice of the same Markan passage (Mark 3.7–8, 13) to provide the setting for the inaugural Sermon (Matt 4.25–5.2; Luke 6.17–18). It is true that one version of the sermon is preached on a mountain-top (Matt 5.1), the other on the plain (Luke 6.17). But this difference may suggest one author countering the claim of another, rather than independent composition. Tuckett deals with such coincidences piecemeal, although any argument for the implausibility of a series of alleged coincidences is necessarily cumulative. Each coincidence is systematically minimized. In the case of the inaugural sermon, for example, It is only the very immediate ‘setting’ that is involved here; the broader ‘settings’ or ‘contexts’ in Mark for the placement of the two Sermons are quite different. (127)

Of course they are! Does Tuckett suppose I failed to notice that? Wrongly assuming that the link I make is between Matt 4.25 and Mark 3.7–8 only, Tuckett proceeds to query my use of the word ‘setting’, which he finds ‘very tenuous’ (127). If the depiction of crowds gathered at the bottom of the mountain and disciples coming

7. Seven Theses on the Synoptic Problem

143

to Jesus at the top is not the ‘setting’ of the Sermon on the Mount, it is hard to know how else to describe it (Matt 4.25–5.1; cf. Mark 3.7–8, 13).4 In spite of all qualifications, Tuckett does concede that there is here a coincidence of sorts. Yet, though it ‘may appear striking at first sight’ (126), the coincidence is just ‘superficial’ (128). All along, the false distinction between agreement and disagreement is hard at work to ensure that any non-Q-derived agreement between Matthew and Luke is superficial and every disagreement profound.

7.5 Thesis Five: Source-critical hypotheses should focus on compositional process and not just on changes of wording. ‘Compositional process’ refers to the successive operations an evangelist must undertake in creating a new gospel book out of earlier sources, including not only changes to the wording but also structural decisions together with the material means for carrying them out. Tuckett finds my account incredible, requiring the evangelist to work with Mark and Matthew at the same time rather than simply alternating between Mark and Q – although ‘it would be very difficult to read two scrolls simultaneously side by side’ (133). Here Tuckett appeals for support to Downing and Derrenbacker, whom he chides me for failing to cite. These scholars have shown that other writers of the period, when using more than one source, adopted a relatively simple and straightforward procedure of using one source at a time, generally not engaging in a complex process of conflation of sources, and also not ‘unpicking’ one source from another. Certainly Luke’s radical reordering of Matthew’s material seems quite unlike anything comparable in writers engaged in a similar process of using a source for their writing. (119)

As a matter of fact, L/M generally requires the Lukan evangelist to work with one source at a time rather than attempting the complicated conflationary manoeuvres that, a few decades later, Tatian would show to be entirely feasible. That leaves the question of ‘unpicking’, a term favoured by Q advocates because of its negative connotations of intricate and laborious operations on a very small scale. What ‘unpicking’ actually refers to is Luke’s identification of non-Markan elements within Matthew, some of which he selects for his own use. On L/M, the first nonMarkan units that Luke ‘unpicks’ or rather selects from Matthew are the summary of John the Baptist’s preaching (Matt 3.7–12, expanded in Luke 3.7–18), the temptation narrative (Matt 4.1–11; Luke 4.1–13) and parts of the Sermon on the Mount (Matt 5.2–7.27). Tuckett needs to explain why he finds such a simple procedure ‘hard to envisage in practical terms’ (137). As for the other ‘unpicked’

4. Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI: Eerdmans, 2013), 148–55, esp. 152–3.

144

Francis Watson

or selected non-Markan passages from Matthew, many of them are taken from large blocks of Matthean material easily identifiable as such by anyone familiar with Mark – as Luke certainly was. Of the seventy such passages from Matt 5–25 listed in my book, fifty-eight are taken from more extensive Matthean contexts and require no ‘unpicking’ from Mark.5 Of the twelve Matthean passages that have genuinely been ‘unpicked’ from a Markan context, seven are easily recognizable additions to their Markan context.6 The remaining five are passages where Luke has selected a Matthean version of a Markan passage in preference or in addition to a Markan version.7 These include the three passages that Tuckett wrongly claims that I misattribute to Matthew (133), overlooking the justification for this decision I give in each case.8 In my reconstruction of Luke’s compositional procedure, the one additional resource he needs as he selects his seventy passages from the easily identifiable Matthean additions is a notebook, in which to set them aside for future incorporation into his own text. Predictably enough, Tuckett dismisses this proposal as unrealistic: [T]he extent of the rearranged Matthean material means that the ‘notebook’ which Luke might have used must have been very extensive indeed. Certainly it would have had to contain far more than could be included on a ‘wax tablet’ (often no more than one side and probably not capable of much detail). (132, italics original)

These statements are unsubstantiated and misleading. Notebooks in codex format appear to have been widely used in educational contexts, or for notes or drafts that the writer does not wish to preserve. Quintilian has this to say about such books: Certain minor matters – since nothing is immaterial to study – should not be passed over. It is best to write on wax tablets [ceris], making deletion easier . . . 5.  Details in Gospel Writing, 186–7. The larger Matthean contexts include Matt 8.5–13 (Matt 8.11–12 // Luke 13.28–30); Matt 10.16–39 (10.19–20 // Luke 12.11–12, and four other passages); Matt 18.6–21 (18.6 // Luke 17.2, and four other passages); Matt 24.37–25.30 (24.37–39 // Luke 17.26–27, and four other passages). 6. Matt 12.11 // Luke 14.5 (sabbath emergency); Matt 13.16–17 // Luke 10.23–24 (blessed the eyes); Matt 16.2–3 // Luke 12.54–56 (signs of the times); Matt 17.19–20 // Luke 17.5–6 (little faith); Matt 19.28 // Luke 22.29 (twelve thrones); Matt 24.26–27 // Luke 17.23–24 (lightning); Matt 24.28 // Luke 17.37 (body, eagles). 7. Matthean rendering preferred: Matt 12.31–32 // Luke 12.10 (blasphemy against Spirit); Matt 13.31–32 // Luke 13.18–19 (mustard seed); Matt 16.6 // Luke 12.1 (leaven of Pharisees); Matt 24.23–25 // Luke 17.20–21 (kingdom in midst). Luke 17.31 + 21.21 (no return) is a doublet corresponding to Matt 24.17–18 and Mark 13.15 respectively, as their contexts show (cf. Gospel Writing, 187, 214). 8.  Matt 24.17–18 // Luke 17.31 (no return); Matt 24.23–25 // Luke 17.20–21 (kingdom in midst); Matt 13.31–32 // Luke 13.18–19 (mustard seed). See Gospel Writing, 198, 212–13, 214–15.

7. Seven Theses on the Synoptic Problem

145

[S]ome pages [tabellae] should be left blank so that further material can be freely added, for lack of space makes one reluctant to make corrections and will certainly produce confusion in the old when new material is inserted . . . Space should also be left available to note down whatever occurs to one as one writes that is not directly relevant to the matter in hand. Not infrequently our best inspirations come to us at a point where they cannot be inserted; yet it would be a mistake to delay committing them to writing, because they would either slip our minds or distract us from further composition. They are best stored up for later.9

These cerae are thin rectangles of wood with a margin enclosing a wax-filled hollow in which one writes with a stylus. Letters become visible as the darker wooden background is exposed beneath the lighter wax. Quintilian here envisages a set of wooden tablets, of variable size and number according to requirements, tied together to form a codex – the book-format that the early church was so quick to make its own. One important function of such a codex was to store up material for future use. Luke’s use of a codex for just this purpose is entirely routine, and his copying into it of selected passages from Matthew does not require the implausible intellectual gymnastics that Tuckett supposes.

7.6 Thesis Six: An evangelist’s modification to a source is best ‘explained’ by showing it to be an interpretative act. Again and again Tuckett complains that I merely ‘describe’ Luke’s alleged changes to his Matthean source but fail to ‘explain’ them. To say that, assuming one particular hypothesis, Luke has made this or that set of changes simply describes the situation. It only provides an explanation if one provides reasons why Luke might have changed his source(s) in the way he has . . . [T]he discussion mostly stays at the level of what Luke has done. Why he might have done what he is purported to have done is often studiously avoided. (119, italics original)

To this, just two brief remarks will suffice. First, we note again the insistent why, the constant demand for more, the insinuation of a glaring lacuna and a fundamental defect. What I  attempt to do, within the constraints of a single chapter, is not just to ‘describe’ how Luke has presented his Matthean selections in new

9.  Inst. or. x.3.31–33 (Quintilian [Marcus Fabius Quintilianus], The Orator’s Education, ed. D. A. Russell, Loeb Classics, 5 vols [Cambridge, MA: Harvard University Press, 2003]). The translation is my own. On writing tablets and notebooks, see C. H. Roberts and T. C. Skeat, The Birth of the Codex (London: British Academy, 1987), 11–23; Graham N. Stanton, Jesus and the Gospel (Cambridge: Cambridge University Press, 2004), 173–8.

146

Francis Watson

formulations and juxtapositions but also and above all to discern the creative interpretative activity expressed in the evangelist’s editorial work. That is why the title of ch. 4 of Gospel Writing is ‘Luke the Interpreter’. Elucidating the interpretation is integral to the description. Second, Tuckett’s insatiable demand for explanation also covers omissions. Why did the L/M Luke omit Matthean beatitudes like ‘Blessed are the pure in heart’ and ‘Blessed are the peacemakers’? Would we not expect that they would appeal to him? Luke has selected four Matthean beatitudes, reworded them to some degree and paired them with corresponding woes (Luke 6.20–26) – but why these, and why not others? It is seemingly not enough to cite compositional factors or the profound interpretative logic that may underlie them, and that establishes the Lukan evangelist as a major New Testament theologian. To say (as I am severely criticized for doing) that it seemed good to him to do what he did is a Lukan way to suggest that, like definitions, explanations must come to an end somewhere.

7.7 Thesis Seven: There are more than three synoptic gospels for source-critical theories to take into account. Tuckett discusses at some length my chapters on ‘The Coincidences of Q’ and ‘Luke the Interpreter’ but does not consider it worthwhile to ‘allow time or space’ to the chapter that follows, the title of which, ‘Thomas versus Q’, suggests that it may have some relevance to my critique of the Q hypothesis. In this chapter, I argued that the Gospel of Thomas is a relatively late (early second century) exemplar of a primitive Christian genre, that of the collection of Jesus’ sayings, whose formatting conventions it may preserve in, for example, the citation formula ‘Jesus said . . .’ which introduces each saying or sayings-cluster. I suggested that the canonical gospels may themselves have drawn on informal, pre-literary Sayings Collections, although we have (limited) access only to the genre and cannot reconstruct them except in the broadest outline. In one sense, this Sayings Collection hypothesis fulfils the Q-like function of explaining how utterances of Jesus (or those that were early put into his mouth) could have been transmitted over the decades between the 30s CE and the 70s and beyond. Like Q advocates, I do not believe that Jesus tradition remained almost entirely unwritten for four long decades. The only comment that Tuckett makes on this is that ‘[s]ome might see the latter theory as Q, having been banished in the previous two chapters, creeping in again by the back door’ (138 n). The verdict of Tuckett’s anonymous and hypothetical ‘some’ is left unchallenged, as though yet another incoherence in my position could be exposed if only time and space allowed. Yet the Sayings Collection hypothesis has little in common with Q as Tuckett himself understands it. Q is tied to the double tradition of Matthew and Luke, primitive Sayings Collections are not. Q is a modern hypothesis, while the existence of the Gospel of Thomas and analogous extra-canonical sayings traditions is beyond doubt. Above all, Q is intracanonical. There is nothing in it that is not already in the Bible. Tuckett

7. Seven Theses on the Synoptic Problem

147

follows an old Oxford Q tradition which – as I show in my chapter (Chapter 5) in this volume – is deeply conservative in its conviction that there is no place for noncanonical material in discussion of canonical gospel origins. In spite of disagreement, I am nevertheless grateful to Christopher Tuckett for devoting time and trouble to engaging with my work, and also to the editors for providing space to pursue the debate.

Part II T EXTUAL S TUDIES

Chapter 8 M A R K W I T H A N D A G A I N ST Q : T H E E A R L I E ST G O SP E L N A R R AT I V E A S A C O U N T E R- M O D E L Eve-Marie Becker

8.1 Mapping Q and Mark in First Century CE Literary Activity The so-called synoptic problem tends to remain an open question.1 Since Lessing and Griesbach’s time, various hypotheses of relating Mark, Matthew and Luke to each other,2 and especially looking for possible models of literary dependence have been proved out.3 Among those models, the Two-Source or Two-Document hypothesis (2DH) is still the most plausible one. However, since no certainty can be reached, the plausibility of the 2DH is and will – rightly – be under further discussion. Competitive models of solving the synoptic quest constantly challenge the Q hypothesis. In current debate the Q hypothesis is in strong competition with the Farrer4 or Farrer-Goulder (F-G), namely, the L/M hypothesis, as Francis Watson calls it.5 Here, it is argued that Luke used Matthew and Mark as Vorlage. Consequently, 1. Cf. on an overview: Udo Schnelle, ‘Die synoptische Frage in der Geschichte der neueren protestantischen Theologie’, in New Studies in the Synoptic Problem: Oxford Conference, April 2008. FS Christopher  M. Tuckett, ed. P. Foster et  al., Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 239 (Leuven:  Peeters, 2011), 809–830. The organizers of the Roskilde conference in June 2015 – Mogens Müller and Heike Omerzu – have to be thanked for reopening the discussion. 2.  Cf. Christopher  M. Tuckett, ‘The Current State of the Synoptic Problem’, in New Studies, ed. Foster, 9–50. 3.  Cf. Andrew Gregory, ‘What is Literary Dependence?’, in New Studies, ed. Foster, 87–114. 4. Cf. lately: John C. Poirier, ‘Introduction: Why the Farrer Hypothesis? Why Now?’, in Marcan Priority Without Q: Explorations in the Farrer Hypothesis, ed. John C. Poirier and Jeffrey Peterson, The Library of New Testament Studies (LNTS) 455 (London: Bloomsbury T&T Clark, 2015), 1–15. 5.  Cf. Francis Watson, ‘Q as Hypothesis:  A Study in Methodology’, New Testament Studies (NTS) 55 (2009): 397–415.

152

Eve-Marie Becker

this model slightly moves the methodological focus from source criticism and Literarkritik to redaction criticism and hereby enlarges the theoretical model of the so-called Benutzungshypothesen.6 In contrast to, for instance, the Two-Gospel hypothesis (2GH), the L/M hypothesis does not question the assumption of Markan priority as such.7 According to both theories – Q or 2DH and L/M – the Gospel of Mark is considered to be the initiator of the literary history of gospel writing in a narrative sense.8 Also according to both hypotheses, the literary emergence of Mark widely appears to be an isolated phenomenon: in the L/M hypothesis there is no Q at all with which the literary shape of Mark would be in analogy; and according to the Two-Source/2DH theory it is per definitionem presupposed that both written documents – Mark and Q – were produced independently from each other. The consequences for Markan studies are rather similar: In both cases, Markan scholars are seemingly restrained from participating in further speculations about Q and a sayings source. In this contribution we shall see that Markan studies will soon reach a blind end whenever Markan scholars escape from the field of Q studies: Indeed, the quest for the literary prehistory of Mark will always overlap with Q studies.9 In what follows I shall show how much the Q hypothesis in fact can fruitfully influence Markan studies – yet I shall go one step further and suggest that the Q hypothesis in a theoretical sense can stimulate Markan research. As I shall argue, Markan studies will,

6. Cf. Udo Schnelle, Einleitung in das Neue Testament, Uni-Taschenbücher (UTB) 1830 (Göttingen: Vandenhoeck & Ruprecht, 2007 [6th edn]), 188ff., 216f. 7.  On Markan Priority, cf., e.g. Marcan Priority, ed. Poirier and Peterson; Alex Damm, Ancient Rhetoric and the Synoptic Problem: Clarifying Markan Priority, BETL 202 (Leuven: Peeters, 2013). 8.  When John S. Kloppenborg and Christoph Heil argue that Q can already be seen as the actual initiator of the literary history of gospel writing, they operate from the basis of a broad definition of ‘gospel genre’ which is not defined – as I suggest – by the literary concept of narrative and storyline, which relates to the macro-genre of Hellenistic-Roman historiography. Cf. John S. Kloppenborg, Excavating Q: The History and Setting of the Sayings Gospel (Minneapolis, MN:  Fortress, 2000), 398ff.; John S.  Kloppenborg, Q, the Earliest Gospel: An Introduction to the Original Stories and Sayings of Jesus (London:  Westminster John Knox, 2008), 62ff.; and Christoph Heil, ‘Evangelium als Gattung. Erzähl- und Spruchevangelium’, in Christoph Heil, Das Spruchevangelium Q und der historische Jesus, Stuttgarter Biblische  Aufsatzbände (SBAB) 58 (Stuttgart:  Katholisches Bibelwerk, 2014), 41–73. Differently, and here more general: Eve-Marie Becker, The Birth of Christian History. Memory and Time from Mark to Luke-Acts, Anchor Yale Bible Reference Library (New Haven: Yale University Press, 2017). 9.  Wolfgang Schenk, ‘Der Einfluß der Logienquelle auf das Markusevangelium’, Zeitschrift für die Neutestamentliche Wissenschaft (ZNW) 70 (1979): 141–65, 161, e.g. points out that in the case of double traditions we have ‘die einzigen bekannten vormarkinischen Quellenstücke und damit . . . die beste Grundlage, um die markinische Redaktionsarbeit einigermaßen kontrollierbar sicher zu bestimmen’.

8. Mark With and Against Q

153

instead of either ignoring Q studies or simply favouring the Farrer hypothesis,10 directly profit from the Q hypothesis. For a Markan scholar, the theoretical profit of the Q hypothesis lies in the perspective of literary history rather than in source theory.11 In this chapter I shall, thus, tentatively develop some of the heuristic presumptions of 2DH.

8.2 Q and Mark: Some Reflections about the State of the Art Q studies usually circle around a well-established set of historical and philological questions: Does the (sayings) material which Matthew and Luke have in common necessarily stem from a mutual source (Q) that had later not been transmitted as such any longer? By asking such a question, our focus of interpretation will rapidly move along the lines of Literarkritik, redaction criticism and reception history – indeed, from Mark and Q to Matthew and Luke, or even further on to the Gospel of Thomas or the Didache.12 In the narrow field of synoptic studies the evaluation of minor13 and major14 agreements is still a matter of debate. At first sight, the significance for Markan studies is – if at all – only an also-ran, while the scholarly focus is primarily on ‘Q’ as such: As soon as the Q hypothesis is accepted, scholars try to reconstruct its precise wording (J. M. Robinson et al.),15 the different stages of Q’s emergence and its socio-religious setting (J. Kloppenborg et al.)16 in the first century CE Near East world. The specific religious profile of Q might point to a particular Jewish milieu of Jesus followers and seems to allow for a more specific socio-religious diversification

10.  However, similar arguments which I will use in order to develop the Q hypothesis were recently used in order to advocate the Farrer hypothesis, e.g. Poirier, ‘Introduction’, 2ff., especially: ‘increased awareness of the evangelists as writers and reshapers of tradition than as strict . . . compilers’ (2). 11. This was widely the approach of, e.g. Schenk, ‘Einfluß’. – Cf. the overview on different methodological approaches to source theory: David B. Peabody, ‘Reading Mark from the Perspectives of Different Source Hypotheses: Historical, Redactional and Theological Implications’, in New Studies, ed. Foster, 159–85. 12.  On the direct reception of Q in the Didache, cf., e.g. Jonathan A. Draper, ‘Die Didache’, in Die Apostolischen Väter. Eine Einleitung, ed. Wilhelm A. Pratscher, UTB 3272 (Göttingen: Vandenhoeck & Ruprecht, 2009), 17–38, here: 19–21. 13.  Cf. Mark 2.23–28 vs. Matt 12.1–8 // Luke 6.1–5; Mark 1.29–31 parr.; 2.1–12 parr.; 3.22–27 parr.; 4.10–12 parr.; 4.30–32 parr.; 4.35–41 parr.; 9.2–10 parr.; Mark 14.65; 5.27; 9.19; 15.44f. – Cf. Schnelle, Einleitung, 193f. (Lit.!). 14. Cf., e.g. Matt 5.3–12; Luke 6.20–23. 15.  Cf. The Critical Edition of Q: Synopsis Including the Gospels of Matthew and Luke, Mark and Thomas with English, German, and French Translations of Q and Thomas, ed. John M. Robinson et al. (Leuven: Peeters, 2000). 16. Cf., e.g. Kloppenborg, Q, the Earliest Gospel.

154

Eve-Marie Becker

of transmission processes within and beyond the Jesus movement. During the so-called Late Second Temple period – more precisely, the Herodian or apostolic age (cf. W. Horbury)17 – there are at least multiple processes of identity formation going on which are mirrored in Pauline letter-writing, synoptic and Johannine ‘Christianity’ and, eventually, the Q group.18 Simon J.  Joseph has presented one of the latest attempts of localizing the Q movement in its socio-religious context, indeed by placing it in close relationship to the Essenes and the Qumran library.19 We will come back to this shortly. The current state of the art in Q studies is certainly manifold and pluralistic. How will Markan scholars then be best involved in Q studies? It is one of the most fascinating tasks, when teaching the gospel writings in class,20 to discuss with students the historical and literary plausibility of the Q hypothesis,21 the impact it made (and still makes) on gospel studies in twentieth and twenty-first century’s scholarship (last but not least on Jesus research also), and the consequences it has for our reading and understanding of the Synoptic Gospels  – including Mark  – as much as John.22 Besides, Markan studies constantly have to take into account the broader cultural and intellectual context

17.  Cf. William Horbury, ‘Beginnings of Christianity in the Holy Land’, in Christians and Christianity in the Holy Land: From the Origins to the Latin Kingdoms, ed. Ora Limor and Guy G. Stroumsa, Cultural Encounters in Late Antiquity and the Middle Ages, vol. 5 (Turnhout: Brepols, 2006), 7–89; Eve-Marie Becker, ‘Jesus and Capernaum in the Apostolic Age:  Balancing Sources and Their Evidence’, in The Mission of Jesus:  Second Nordic Symposium on the Historical Jesus, Lund, 7–10 October 2012, ed. Samuel Byrskog and Tobias Hägerland, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) II/391 (Tübingen: Mohr Siebeck, 2015), 113–39, 131. 18.  Christoph Heil, ‘Einleitung’, in Die Spruchquelle Q.  Studienausgabe Griechisch und Deutsch, ed. Paul Hoffmann and Christoph Heil (Darmstadt:  Wissenschaftliche Buchgesellschaft, 2002), 12–30, 27. 19.  Cf. Simon J. Joseph, Jesus, Q, and the Dead Sea Scrolls:  A Judaic Approach to Q, WUNT II/333 (Tübingen: Mohr Siebeck, 2012), esp. 22ff. 20.  Cf. on this, various contributions in Communication, Pedagogy, and the Gospel of Mark, ed. Elizabeth Evans Shively and Geert Van Oyen, Resources for Biblical Study 83 (Atlanta: Society of Biblical Literature Press, 2016), herein, esp. Becker, ‘Mark in the Frame of Ancient History Writing’, 13–27. 21.  Cf., e.g. Christoph Heil, ‘Die Q-Rekonstruktion des Internationalen QProjekts:  Einführung in Methodik und Resultate’, Novum Testamentum (NT) 43 (2001): 128–43. 22. As a matter of fact, the separation of the Fourth Gospel from the Synoptic Gospels – which is problematic in itself – is even further enforced by the Q hypothesis. On the relation between John and the Synoptics, see valuable contributions such as Hans Windisch, Johannes und die Synoptiker. Wollte der Vierte Evangelist die älteren Evangelien ergänzen oder ersetzen?, Untersuchungen zum Neuen Testament (UNT) 12 (Leipzig:  Hinrichs’sche Buchhandlung, 1926); John and the Synoptics, ed. Adelbert Denaux, BETL 101(Leuven: Peeters, 1992).

8. Mark With and Against Q

155

in which the literary emergence of the genre of ‘gospel’ writing took place. Here the Q collection can play a significant role. But how can Markan scholars best discuss the implications the Q hypothesis has for interpreting the earliest gospel narrative? Studying the interrelations of Mark and Q is not a totally new endeavour  – there have already been attempts where various scholars have looked for possible (literary) interdependencies between Q and Mark. However, since these attempts were mostly based on the assumption of (direct) dependency, they soon became schematic: It was either discussed how intensively Mark was literarily dependent on Q,23 or how Q could have drawn from pre-Markan material.24 There was hardly any convincing result obtained from these discussions. For many scholars who are in favour of the Q hypothesis, the result has been to let Q and Markan studies rather move along separate paths. But even though we would agree to separate Q and Markan studies, the relation of Q and Mark still is a complex one: How do we finally deal with the general theoretical claim that there is no Q hypothesis if Q was not part of one coin where Mark pictures the other side? To what degree does the hypothesis of the Markan priority, which is crucial for Markan studies, constitutively interfere with the Q hypothesis – even though in terms of the history of research, the latter results from the former, and not vice versa?25 Can the Markan priority, as it is currently argued, really be imagined without necessarily hypothesizing Q? The L/M hypothesis precisely departs from this insight by offering one possible alternative to the Two-Source theory/2DH: Markan priority without Q. Does the Q hypothesis in consequence become superfluous – as Marc Goodacre and others suggest?26 We have to question this conclusion, indeed, for two reasons:  First, the Q hypothesis is still significant for reading Mark in terms of religious history and theology. On the one hand, the historical validity of Jesus traditions in the sense of ‘multiple attestations’ can be stated when we have to deal with the so-called double (or triple) traditions.27 But again, it is mostly in its strict separation from Q then that Mark is credited for its ‘historical accuracy’. On the other hand, as soon as scholars tend to see in Matthew a ‘special relationship’ to Q in terms of content

23.  Cf., e.g. Bernard Weiß; Wilhem Bousset; Wolfgang Schenk; Walter Schmithals; Jan Lambrecht; David R.  Catchpole; Harry T.  Fledderman  – references in Schnelle, Einleitung, 233f. 24. Schenk, ‘Einfluß‘, discusses and votes for both options. 25. The hypothesis of the Markan priority is prior to the Q hypothesis; the latter results from the presumption of the Markan priority (Karl Lachmann et al.). 26.  Cf. Mark Goodacre, ‘A World without Q’, in Questioning Q:  A Multidimensional Critique, ed. Mark Goodacre and Nicholas Perrin (London: SPCK, 2004), 174–9. 27.  Cf. Mark 1.2, 7f., 12f.; 3.22–26, 27–29; 4.21, 22, 24, 25; 4.30–32; 6.7–13; 8.11f.; 8.34–35, 38; 9.37, 40, 42, 50; 10.10f., 31; 11.22f.; 12.37b–40; 13.9, 11, 33–37. – Cf. Schnelle, Einleitung, 233f. However, here it is presumed that Mark and Q had independent access to those traditions.

156

Eve-Marie Becker

and theology,28 Mark’s obvious distance to Q is judged as a certain type of theology that places itself beyond the pure rendering of Jesus traditions – it possibly rather aims to join Pauline thinking in a more or less recognizable way.29 Second and much more important are the implications the Q hypothesis has for gospel studies in light of literary history: When we aim to understand the literary emergence of Mark in the overall frame of a commencing early ‘Christian’ literary history which generally has to be seen against the background of transition processes from orality to literacy,30 we will best understand the written gospel concept as initiated by Mark by contrasting it with a conceptual counter-model (Q), which must soon afterwards have disappeared from the early Christian book market.31 While Mark intends to narrativize and interpret the storyline, the Q concept reflects a more traditional claim of documenting and collecting authoritative words being possibly close to the claim of conserving the ‘viva vox’ to which Papias of Hierapolis one to two generations later will still commit himself (cf., esp. Eusebius, Historia ecclesiastica 3.39.4; 3.39.15; Jerome, De viris illustribus 18). In terms of its literary genre Q is thus best described as a ‘Spruch- bzw. Redensammlung’ (Christoph Heil).32 So far we have tried to reach a fair description of the current state of the art in Q studies, seen with the eyes of a Markan scholar who in general tends to sympathize with the Q hypothesis. This description might also help us to better appreciate the scholarly progress visible in Q studies as compared to a rather ‘dualistic thinking’ which we met in this field in the 1960s.33 However, when reflecting about the theoretical need of the Q hypothesis, there is continuously a crucial desideratum visible among scholars from various ‘camps’ of synoptic studies. Synoptic scholarship still tends to ignore how Mark’s composition as gospel writing within early ‘Christian’

28. Cf. Linden Youngquist, ‘Matthew, Mark and Q’, in Mark and Matthew I, Comparative Readings: Understanding the Earliest Gospels in their First Century Setting, ed. Eve-Marie Becker and Anders Runesson, WUNT 271 (Tübingen:  Mohr Siebeck, 2011), 233–61, 239. Even among the representatives of the Farrer hypothesis, it is stated that Matthew could have used a source which is not Q, but which is somehow similar to Q. 29.  Cf. in general Paul and Mark, Comparative Essays Part II: For and Against Pauline Influence on Mark, ed. Eve-Marie Becker, Troels Engberg-Pedersen and Mogens Müller, Beihefte zur ZNW (BZNW) 199 (Berlin: de Gruyter, 2014/2017). 30.  However, a general scepticism about the schematic orality-literacy divide in antiquity was recently formulated again by Gregory Woolf, ‘Ancient Illiteracy?’, Bulletin of the Institute of Classical Studies 58 (2015): 31–42. 31.  If we choose a literary and conceptual model of interpreting this fact, was it the lack of narratives and a storyline that discloses itself in historiographical and logical terms which made Q lose its place in early Christian literary history? Or should we rather try to upgrade Q in regard to its narrative outline and its literary claim? 32. Heil, ‘Evangelium als Gattung’, 60. 33.  Cf., e.g. Theodore E. Rosché, ‘The Words of Jesus and the Future of the “Q” Hypothesis’, Journal of Biblical Literature (JBL) 79 (1960): 210–20, esp. 220.

8. Mark With and Against Q

157

literary history must conceptually have been affected by Q as a (possible) forerunner document. Up to now, Q is not seen in the overall light of the emerging literary culture among Christ believers. This restrictive view is typical of the advocates of the Q hypothesis who separate Q from Mark and – if at all – look for material overlaps only and the representatives of the L/M hypothesis who finally leave it open to explain how the Markan composition of a pre-historiographical narrative can have emerged, supposing there was no literary activity at all among Jesus followers and Christ believers except Pauline letter writing. Even in their exegetical work, both ‘camps’ are finally restricted to the theoretical frame of literary dependence or independence.34 In what follows, I will argue for an alternative theoretical frame: Mark’s response to Q as a literary counter-model. Before reaching this, I will further elaborate on the question of how Mark – in terms of its literary genre – came into being as (pre) historiographical Hellenistic literature.35 I shall initially work with both presumptions and ask how Mark came into being, presuming both that there was no Q and that there was an existent piece of Q literature.36

8.3 Mark without Q: Resuming Desiderata To a Markan scholar to whom – being a fellow of F-G or L/M – the Q hypothesis is easily dispensable, Mark’s Gospel would appear to be like a creatio ex nihilo: Mark would be the first writer who has transformed a diversity of (oral) Jesus traditions (logia, miracle, passion narrative) into a written account. This account seemingly was preliminary only: It would first be the historian Luke who – by having the choice of studying various gospel accounts (possibly Matthew, in any case Mark) – would have presented himself as a selective, evaluative and interpretive historiographical author. According to the L/M hypothesis, the literary history of gospel writing consequently would follow a kind of a ‘teleological model’ where Luke represents the final stage. Thinking Mark wrote without Q – while simultaneously maintaining the assumption of Markan priority – would mean to read the Markan Gospel as a first draft of a gospel story which lacks the literary skills of narrative selection, evaluation and interpretation which is characteristic of ancient literary authors, especially historians. Is this picture of Mark plausible, fair and appropriate? 34. Cf., e.g. Zeba A. Crook, ‘The Synoptic Parables of the Mustard Seed and the Leaven: A Test-Case for the Two-Document, Two-Gospel, and Farrer-Goulder Hypotheses’, Journal for the Study of the New Testament (JSNT) 78 (2000): 23–48. 35.  Cf. Eve-Marie Becker, Das Markus-Evangelium im Rahmen antiker Historiographie, WUNT 194 (Tübingen: Mohr Siebeck, 2006). 36.  Only after having settled on the title of this chapter, I remembered that Ed Parish Sanders once used a similar description: ‘Mark without Q’, cf. Studying the Synoptic Gospels, ed. Ed Parish Sanders and Margaret Davies (London:  SCM, 1989), 63  – also quoted in Poirier, ‘Introduction’, 14.

158

Eve-Marie Becker

Considering the high literary quality of Mark’s writing (storyline, logical consistency, language etc.) which nowadays (in contrast to F. Overbeck et al. and the representatives of the Formgeschichte) finds its constant  – or rather increased  – appraisal among classical (e.g. H.  Cancik et  al.) as well as modern (e.g. H.  M. Gauger et al.) philologists, we can hardly consider the Markan Gospel to be a preliminary ‘draft’ only.37 By saying this we have to argue against all scholars who, for different reasons, downplay Mark’s literary status and quality. The group of scholars arguing for the Two-Source theory/2DH considers Mark to have source, namely, documentary, status; likewise, for the advocates of the so-called Farrer or L/M hypothesis (or various types of Benutzungshypothesen), Mark finally is no more than a literary source or Vorlage for later, more distinguished gospel writings (esp. Luke). This perspective on Markan studies, however, evidently falls too short. In the end, it refers back to the state of synoptic scholarship in the late nineteenth and early twentieth century.38 Synoptic studies have still not really recovered from the deficient valuation of the Markan Gospel as suggested once by source theorists. It seems to me that we should move on a different path because whether or not we assume the emergence of Mark to happen in temporal proximity to the (final redaction of the) Q document, we have to explain most plausibly how the Markan Gospel as a written narrative of comparably high literary standards originated. The question must be: Can the Q hypothesis help us to better explain the literary shape of the Markan Gospel anyhow? My tentative answer in this chapter is: yes, it can – as soon as we develop some of its premises. Instead of thinking along schematic and restrictive lines of source theory and literary dependence or independence, we should open up to the model of a coexistent literary quality and more extensive literary overlaps. The literary shape of the Markan Gospel is more plausible when we place it in a productive coexistence to a more expanded Q writing rather than in a strict separation from it.

8.4 Mark with Q: Developing the Q Hypothesis Let us instead in a final step suppose that Q did exist, gradually being transformed from an oral collection of Jesus sayings (and deeds) to a literary document which in its final stage also includes a certain amount of narrative sequences (esp. Q 3–4). The ‘Q-Endredaktion’39 might eventually have taken place around the time the Markan Gospel came into being. The question to be answered will thus be: Was the

37. More comprehensively on this, cf. Eve-Marie Becker, ‘Literarische Innovation in hellenistisch-römischer Zeit: Die Evangelien im Lichte des genre-Diskurses’, in Eve- Marie Becker, Der früheste Evangelist: Studien zum Markus-Evangelium, WUNT 380 (Tübingen: Mohr Siebeck), 17–33. 38.  This is, interestingly enough, just around the time when the Q hypothesis was worked out extensively (Adolf von Harnack et al.). 39. Heil, Einleitung, 22.

8. Mark With and Against Q

159

author of the earliest gospel narrative – when conceptualizing his story about the ἀρχὴ τοῦ εὐαγγελίου – aware of the Q document? Suppose he was. How does he relate his literary product to an already existing (written) collection of Jesus traditions which memorize and interpret Jesus’ ministry and teaching? 8.4.1 Mark’s Knowledge of Q From what we know in general about ancient literary activity, including the circulation of literary products, it is hard to believe that Mark would have not known anything about a document like Q as a contemporary, slightly forerunning piece of literature.40 Even though scribal or, more generally speaking, literary activities – such as the Oxyrhynchus example documents41 – are much more widespread than we tend to think, they are not to be seen as a self-evident and common task in ancient times either: Literary activity indeed is an exceptional venture – this can be said without either necessarily presuming an elitist culture behind all scribal activities in antiquity,42 or presupposing only a high-rank concept of ancient ‘literature’.43 Written texts which exceeded daily-life communication were considered literature and made known as such, that is, sent around (e.g. Col 4.16) – especially when we assume that there were vital networks of interaction existent within and between various groupings of Jesus followers and Christ believers and when we also describe the transmitters of Q as migrants who showed much and constant mobility. If Mark got to know about Q, does this scenario imply that Mark drew on Jesus traditions that are transmitted by Q, that is, that he drew on mutual Palestinian Jesus material? We can indeed envisage that possibility (without necessarily arguing for direct literary dependence). However, instead of using Q as a source in a material sense – a source Mark would have been dependent on in a literary sense – Mark might have rather ‘used’ Q as a ‘source of inspiration’ and/or a counter-model when composing his book. Such a – rather indistinct – recourse to Q would also

40. Mark himself was aware of collections of sayings (e.g. Mark 4; 6.7ff.; 13). Eventually, he even got to know Q, as John might have had knowledge of Mark and Q without necessarily being directly dependent on the Synoptic gospels in literary terms. See above (note 22). 41.  Cf. Lincoln H. Blumell, Lettered Christians:  Christians, Letters, and Late Antique Oxyrhynchus (Leiden: Brill, 2012), 163ff. 42. Cf. in general on the diversity of social milieus behind either scribal or literary activities in the ‘pagan’ world:  William A. Johnson, Readers and Reading Culture in the High Roman Empire: A Study of Elite Communities (Oxford:  Oxford University Press, 2010); cf. on the Q movement: Giovanni B. Bazzana, Kingdom of Bureaucracy: The Political Theology of Village Scribes in the Sayings Gospel Q, BETL 274 (Leuven: Peeters, 2015). 43.  We should instead highlight the fact that the variety and diversity of what can be called ‘literature’ in Hellenistic times is huge, cf. Ernst Vogt, ‘Die griechische Literatur’, in Griechische Literatur, ed. Ernst Vogt, Neues Handbuch der Literaturwissenschaft (NHL) 2 (Wiesbaden: Athenaion, 1981), 1–18, 1–2.

160

Eve-Marie Becker

explain remarkable divergences whenever Mark and Q present mutual material. This is, for instance, the case when Mark and Q both refer to Capernaum – a place that must have played a central role in Jesus’ Galilean ministry (cf. Mark 2 and Q 7.1; 10.15).44 Mark and Q obviously had different literary and theological intentions when mentioning eminent places around the Galilean Sea. Let us go one further step in our presumed scenario: If Mark cannot and did not ignore the existence of Q – why does he not at all refer to it? Does he intend to suppress Q by providing a more – or even ‘better’, that is, a more comprehensive – literary writing (Mark 1–16)? It seems as if we rapidly move into the field of pure speculation here. However, if we look for analogies between how first and second century CE (historiographical) writers in general have dealt with contemporary or (slightly) forerunning socio(-religious) groupings and (competing) writers we might shed some important light on Mark’s tentative dealing with Q. We will first investigate Mark’s dealing with Q’s Trägerkreis as a group of forerunning foreigners that he would have had to relate to (socio-religious perspective), and then consider his possible strategy of dealing with Q’s text-producers as literary competitors (intellectual perspective). We shall reach in the end a surprisingly consistent picture which can explain Mark’s critical relation to Q. 8.4.2 Mark’s Silence about the Q Movement and His Depiction of ‘Others’ As far as I see, the Markan Gospel does not expose any particular group of Jesus followers which would reflect the precise image of how we reconstruct the Q movement/network. At the same time, Mark takes more or less explicit notice of many other, mostly differing groupings which were around in first century CE Palestine – Eyal Regev has listed those again lately.45 Mark in fact mentions Pharisees (Mark 2.18 etc.), Sadducees (Mark 12.18) and eventually even Zealots and apocalyptic movements (cf. Mark 13). According to Regev’s listing, Mark does not refer to Essenes and Sicarii. However, in addition to Regev’s mapping of Judean society, Mark also names γραμματεῖς (e.g. Mark 1.22) and a group of disciples around John the Baptist (Mark 2.18ff.). Especially Mark’s interest in the mention of γραμματεῖς is hardly contingent since the Markan gospel narrative constantly emphasizes Jesus’ success as teacher and interpreter of scripture in contradistinction from his opponents (esp. Mark 2; 12).46

44. Cf. Becker, ‘Jesus and Capernaum’. 45.  Cf. Eyal Regev, ‘Flourishing before the Crisis: Mapping Judean Society in the First Century CE’, in Jews and Christians in the First and Second Centuries: How to Write Their History, ed. Peter J. Tomson and Joshua Schwartz, Compendia rerum iudaicarum ad novum testamentum (CRINT) 13 (Leiden: Brill, 2014), 52–69. 46.  Cf. Lorenzo Scornaienchi, ‘Jesus als Polemiker:  Wie polemisch darf Jesus sein? Historische und normative Aspekte’, in Polemik in der frühchristlichen Literatur. Texte und Kontexte, ed. Oda Wischmeyer and Lorenzo Scornaienchi, BZNW 170 (Berlin:  Walter de Gruyter, 2011), 381–413, 403ff.

8. Mark With and Against Q

161

Mark’s portrait of the Palestinian socio-religious setting in the Late Temple period is not far from how historians of his time in general make reference to diverse socio-religious groupings. This applies first of all but not exclusively to Josephus. Philo and Josephus both, in their portrayal of the Essenes, for instance, make Judaism appear as a ‘sort of ideal “philosophy”, able to compete with Greek philosophical schools and with Hellenistic-Roman religions’ as Per Bilde has pointed out.47 But even when investigating how ‘Jews’ and ‘Christians’ – indeed as distinct groups48 – are perceived by Romans, that is, outsider authors, around 100 CE, John Barclay reaches similar results and reveals a certain ‘rule of thumb’: The more Roman authors move into the literary field of history-writing stricto sensu, the more comprehensive is their interest in and picture of socio-religious (as well as ethnic) diversification. It does not seem accidental then that, in the early Roman Empire, Suetonius and Tacitus offer a more thorough picture of socioreligious groupings than Pliny the Elder and Pliny the Younger, Martial, Seneca, or Juvenal do. Can this observation about historians who provide a quite comprehensive, though intentionally coloured picture of socio-religious stratigraphy match how Mark is seemingly silent about the Q movement? Does Mark ignore Q since the document and its transmitter group is in close relationship to the Essenes and the Qumran library (see above S. J. Joseph), a grouping which is – perhaps not accidentally  – also left out in Mark’s portrait of Herodian Palestine? Or does Mark conceal the Q movement programmatically because he rejects it or considers himself (and his community, or better yet, audience) to be in religious and/or literary competition with it? In some parts of Pauline letter-writing we find at least a remarkable and different strategy of not concealing ‘others’ who are opponents and competitors: In 1 Cor (partly 2 Cor 10–13) and Gal 1–2, Paul makes reference to competing missionaries – he even names them and points out their influence on various parties explicitly (Peter, Apollo, James),49 in order to reveal their inferiority. However, while Paul in his letter-writing can directly interact with contemporary opponents, the challenge for Mark is different. The Q movement is not a natural part of Mark’s world of storytelling (‘erzählte Welt’) since it is posterior to the time frame narrated (‘erzählte Zeit’). Mark can thus hardly mention the representatives of Q explicitly as he talks about the disciples of John the Baptist in their distinction from the Jesus group (Mark 2.18). Mark’s way of interfering with Q can thus only be implicit and subcutaneous. Mark corrects, modifies,

47. Per Bilde, ‘The Essenes in Philo and Josephus’, in Per Bilde, Collected Studies on Philo and Josephus, ed. Eve-Marie Becker et  al., Studia Aarhusiana Neotestamentica (SANt) 7 (Göttingen: Vandenhoeck & Ruprecht, 2016), 121–50, 146. 48. Cf. John M. G. Barclay, ‘ “Jews” and “Christians” in the Eyes of Roman Authors c. 100 CE’, in Jews and Christians, ed. Tomson and Schwartz, 313–26, esp. 320f. 49.  Does the fact that Paul, on other occasions, does conceal competitors point to the possibility that they are eventually also active as (letter) writers?

162

Eve-Marie Becker

re-conceptualizes Q’s outline, for example, when identifying John the Baptist (cf. Mark 1.2 in difference to Q 7.27),50 presenting parables (cf. Mark 4.30–32 and Q 13.18f.),51 arranging prophetic and/or apocalyptic speech (cf. Mark 13.4ff. as differing from Q 17–19), or predicting the temple’s fortune (cf. Mark 13.2; 14.58; 15.29 as differing from Q 13.35). 8.4.3 Mark’s Dealing with Q as a Literary Forerunner and/or Competitor Against this background we move into the broad field of literary strategy and intellectual discourse.52 In a final step we will develop Barclay’s above-mentioned ‘rule of thumb’ when observing how long in fact historiographical writers provide a comprehensive picture of socio-religious milieus: My impression is that historians in the early Roman Imperial time are eager to describe socio-religious groupings as long as those groupings do not interfere with the historian’s own literary and intellectual ambitions. In other words, as long as historians consider themselves to be narrators and interpreters of (contemporary) history, they collect all available information even in order to strengthen their narrative position as writers who are more than well informed. However, as soon as they would, in particular, be forced to mention literary activities within ‘other’, often adverse milieus, they would rather be silent or imprecise about them. Since literacy was a dominant weapon in cultural and intellectual battles at that time, historians tend to suppress competing authors. There might be only one exception to this:  when historians choose autobiographical forms of writing, it is an elementary part of their literary strategy to expose competitive historians  – as we see it in Josephus’ ‘Vita’ when fighting against Justus of Tiberias (§ 34ff.). Even more, explicitly addressing the literary competitor strengthens the author’s – in this case Josephus’ – individual profile as a literate and intellectual (cf. also 2 Pet 3). Being silent about competitors or making only imprecise references to forerunning historiographical accounts, in contrast, is a literary motif that we mostly find in the prologues of history-writing (Luke 1:1ff.; Josephus, BJ 1:1ff.; ant 1:1ff.).53 This motif, again, appears to be a literary strategy which strengthens once more the narrator’s role – especially when 50. Cf. Becker, Markus-Evangelium, 243ff. 51.  Can we even better explain the so-called minor agreements in Matthew and Luke when we make the assumption that Q would have contained more extensively literary material we also find in Mark? 52.  Cf., in general, Kendra Eshleman, The Social World of Intellectuals in the Roman Empire:  Sophists, Philosophers, and Christians (Greek Culture in the Roman World) (Cambridge: Cambridge University Press, 2012). 53.  Since Mark 1.1 or 1.1–3 cannot be seen as a prologue as such, this option is missing to the author of the Markan Gospel. On the implicit rejection of Q 7.27 in Mark 1.2, see above. I would like to thank Christoph Heil (Graz) for his stimulating response to my paper during the conference and further helpful remarks when preparing the paper for publication.

8. Mark With and Against Q

163

the audience recognizes that the historian has evidently drawn upon forerunners by deconstructing or surpassing their narrative account. The Markan Gospel in its suppressive interaction with Q can indeed be read in light of such a literary strategy. From here, we can reach the following conclusion: Mark has dealt with Q as a source of inspiration in order to surpass and counteract the existing piece of literature. The fact that Mark has concealed Q does not necessarily mean that he has not shared or even drawn on its material or even literary basis. Rather, to the contrary, the best way for Mark to argue against Q would be to make use of it and to surpass it (implicit aemulatio). Mark as well as Q are thus to be seen within the broader field of early Christian literary activity. Herein, we can shed new light on the relationship of Mark and Q: Mark was stimulated and challenged by Q’s outline; he drew upon the Q document without incorporating it like a source. Mark’s literary ambition was to exceed the forerunner text.

Chapter 9 R E F U SI N G T O A C K N OW L E D G E T H E I M M E R SE R (Q 7.31–35) Clare K. Rothschild

9.1 Introduction This chapter examines ‘the children who refuse to play’ parable in Q 7.31–35, hypothesizing a new reading, for what it may tell us about the Synoptic Problem. It explores the three parts of the episode, asking to what extent the interpretation (vv. 33–34) follows from the parable (vv. 31–32), and how the concluding wisdom saying (v. 35) is related to each.1

1. I wish to thank R. Matthew Calhoun for critical comments on a prior version of this chapter, as well as Mark Goodacre, John Kloppenborg, Christopher Tuckett and Francis Watson for excellent suggestions at a presentation of the paper in Roskilde, and Mogens Müller and Heike Omerzu for the opportunity to present the paper at that conference (20– 24 June, 2015). In the paper, I assume the existence of Q and that John the Baptist was the original subject of its sayings. I cannot litigate this position in the present chapter but have attempted to make the case in my book, Baptist Traditions and Q, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 190 (Tübingen: Mohr Siebeck, 2005). In this chapter, I focus on the International Q Project’s (IQP) version of the passage, which favours the Lukan version, since Luke is the Gospel thought to preserve the more original ‘Q form’ of this pericope. See Joseph Fitzmyer, The Gospel according to Luke: Introduction, Translation, and Notes, AB 28–28A (Garden City, NY: Doubleday, 1981–85), 1:679. Pace Mark Goodacre (citing Christopher Tuckett, ‘The Existence of Q’, in The Gospel Behind the Gospels: Current Studies on Q, ed. Ronald A. Piper, Novum Testamentum Supplements (NovTSup) 75 [Leiden: Brill, 1994], 19–47), The Case Against Q (Harrisburg, PA: Trinity Press International, 2002), 88 n. 24.

166

Clare K. Rothschild

9.2 History of Interpretation Form-critically speaking, most scholars agree as to the composite nature of this textual unit.2 Joseph A. Fitzmyer describes the three elements as: (1) parable (or simile, vv. 31–32); (2) explanation of the parable (vv. 33–34);3 and (3) wisdom saying (v. 35).4 Not all agree that the three parts originally comprised a single comprehensive unit, however. Rudolf Bultmann cites the parable as an example of a similitude whose original meaning is irrecoverable.5 Some specify incompatibility among the segments. Dieter Zeller argues that the parable clearly identifies ‘this generation’ with the children in the marketplace, identifying the crux interpretum as the parable’s relationship to the interpretation.6 Christl Maier and Jens Herzer view the parable and its interpretation as contradictory, assigning only the parable and wisdom logion to the earliest stratum of tradition.7 Not only is the relationship between the three sections difficult to parse, but each section alone poses interpretive challenges. Ulrich Luz categorizes scholarship on the parable into three groups:  (1) those who view this generation as the children addressed by the parable; (2) those who view this generation as the children who are calling out in the parable; and (3) those who view this generation as all of the children (i.e. those calling out and those called) (cf. v. 35).8 The three positions bring to light yet another question of the text, namely, how many groups of children are implied? Most interpreters argue that Luke implies only one

2. Wendy J. Cotter, in an extremely helpful article, describes the passage (i.e. Q 7.31–35) as ‘clearly composite’: ‘The Parable of the Children in the Market-Place, Q (Lk) 7:31–35: An Examination of the Parable’s Image and Significance’, NovT 29 (1987):  289–304 at 293. Cf. Dieter Zeller, ‘Die Bildlogik des Gleichnisses Mt 11,16f. / Lk 7,31f.’, Zeitschrift für die Neutestamentliche Wissenschaft (ZNW) 68 (1977): 252–7; Stephan Witetschek, ‘The Stigma of a Glutton and Drunkard: Q 7,34 in Historical and Sociological Perspective’, Ephemerides Theologicae Lovanienses (ETL) 83 (2007): 135–54 at 141–3. 3.  γάρ connects the interpretation to the parable. What is more, with Ulrich Luz, the parable alone all but demands interpretation (Matthew:  A Commentary. Hermeneia [Minneapolis: Augsburg Fortress, 1989–2007], 2:146; cf. 2:148). 4. Fitzmyer, Gospel according to Luke, 1:677. Cf. Walter Grundmann, ‘Weisheit im Horizont des Reiches Gottes; Eine Studie zur Verkündigung Jesu nach der Spruchüberlieferung Q’, in Die Kirche des Anfangs: Festschrift für Heinz Schürmann zum 65. Geburtstag, ed. Rudolf Schnackenburg (Leipzig:  St. Benno, 1977), 175–99; M. J. Suggs, Wisdom, Christology, and Law in Matthew’s Gospel (Cambridge, MA: Harvard University, 1970), 33–61. 5.  The History of the Synoptic Tradition, trans. J.  Marsh (rev. edn; Oxford:  Oxford University Press, 1963; repr., Peabody, MA, 1995), 199. 6. Zeller, ‘Die Bildlogik’, 255–7. 7.  ‘Die spielenden Kinder der Weisheit (Lk 7,31–35 par. Mt 11,16–19): Beobachtungen zu einem Gleichnis Jesu und seiner Rezeption’, in Exegese vor Ort. FS Peter Welten (Leipzig: Evangelische Verlagsanstalt, 2001), 277–300, esp. 282–5. 8. Luz, Matthew, 2:146–48.

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

167

group with some calling to ‘each other’ (ἀλλήλοις), whereas Matthew implies two groups, one calling ‘the others’ (τοῖς ἑτέροις).9 The number of groups is important insofar as two groups may correspond better to the interpretation (i.e. John the Baptist and the Son of Man). Shifting focus to the interpretation (vv. 33–34), is there any correspondence between what the children say and subsequent comments concerning the eating practices of John and Jesus? Does Jesus invite people to dance, but John invite people to mourn?10 Are piping and dancing diametrically opposed to chanting dirges and lamenting? Do John and Jesus both reject this generation? Or are they included in this generation, perhaps calling out and being ignored?11 Or have they been rejected by this generation because they did not respond when it called out to them? If John and Jesus are rejected, what exactly did they fail to do? Did they fail to live in accordance with the Jewish liturgical calendar?12 James A. Kelhoffer has recently proposed that John and Jesus are rejected in vv. 33–34, but their rejection has nothing to do with their prophetic messages; rather they are rejected for not participating in normal communal meals.13 Is this correct? Are the diets in this passage purely literal? With regard to the interpretation section, it is further unclear why Jesus comes off slightly worse, as φάγος καὶ οἰνοπότης (‘glutton and drunkard’),14 than John, who merely fasts, assisted by a δαίμων (δαιμόνιον ἔχει). To be sure, a δαίμων is lucky if it allows an ancient prophet, seer, or sage to exercise miraculous control of his dietary intake for an extended period of time (Luke: μὴ ἐσθίων ἄρτον μήτε

9. e.g. I. Howard Marshall, The Gospel of Luke, The New International Greek Testament Commentary (NIGTC) (Grand Rapids, MI: Eerdmans, 1978), 300. 10.  However, flute playing occurs first in the text parallel to John, mourning is parallel to Jesus (Luz, Matthew, 2:146 n. 9). 11. See Alfred Plummer, The Gospel according to St. Luke, ICC (Edinburgh: T&T Clark, 1922 [5th edn], repr. 1953), 206. 12. See Olof Linton, ‘The Parable of the Children’s Game: Baptist and Son of Man (Matt. xi. 16–19 = Luke vii. 31–5): A Synoptic Text-Critical, Structural and Exegetical Investigation’, New Testament Studies (NTS) 22 (1975–76): 159–79, here 175. 13. ‘John the Baptist as an Abstainer from Table Fellowship and Jesus as a “Glutton” (Luke 7:31–35||Matt 11:16–19): (Non-)Participation in Meals as a Means for Gaining Social Capital That Can Confirm or Jeopardize a Person’s Standing in Society’, in The Eucharist – Its Origins and Contexts: Sacred Meal, Communal Meal, Table Fellowship in Late Antiquity, Early Judaism, and Early Christianity, ed. David Hellholm and Dieter Sänger, WUNT 376; 3 vols (Tübingen: Mohr Siebeck, 2017), 313–29. 14. Eating with tax collectors and sinners does not necessarily imply failing to keep kashrut, but rather denotes an ardent approach to the eschaton. Luz comments, ‘While many people probably praised John’s asceticism, “glutton” and “drunkard” traditionally are negatively loaded words’ (Matthew, 2:149). Cf. also François Bovon, Luke, Hermeneia; (Minneapolis: Fortress, 2002), 1:286 on the eschatological scenario. Fitzmyer (Gospel according to Luke, 1:681) notes that the line may echo Deut 21.20.

168

Clare K. Rothschild

πίνων οἶνον/Matt:  μήτε ἐσθίων μήτε πίνων).15 It is further unclear why the interpretation focuses on the diets of two individuals as opposed to other aspects of their ministries.16 Is there a correspondence between minstrel music or dance and dietary intake? Are they connected as outward signs on an eschatological horizon? The concluding wisdom saying (v. 35)  is also enigmatic. Daniel Harrington describes it as ‘obscure’.17 Some scholars doubt whether it belongs to the same stratum of tradition as the parable and interpretation (the καί of the saying makes only a loose connection with what comes before).18 It would seem that the perplexing parable and its enigmatic interpretation demand some kind of conclusion, if for no other reason than to mitigate Jesus’ apparent failure.19 Does the saying exonerate Jesus? Who are Wisdom’s children? Are they John and Jesus?20 Or are they those who have heeded John’s, Jesus’, or both of their messages? Is Jesus to be identified with Wisdom? Was the saying in v. 35 originally independent, merely linked to the parable and interpretation through the catchword concept ‘child’ (παιδίον, τέκνον) or the catchword ‘to vindicate’ (ἐδικαίωσαν, ἐδικαιώθη)?21 Perhaps commentators are pressing connections between the three sections of the pericope too hard? And as if these questions were not sufficiently daunting, I. Howard Marshall points out

15. The two Greek words are rare. Fitzmyer notes about John’s diet: ‘John’s asceticism is regarded as unreasonable’ (Gospel according to Luke, 1:681). While both caricatures seem negative, the diet itself is not the point. 16.  Cotter observes: ‘The issues by which they [John and Jesus] are condemned do not relate to the differences in the preaching of John and Jesus, but rather to the differences in eating and drinking habits, i.e. externals’ (‘Children in the Market-Place’, 302, emphasis original). 17. Harrington writes: ‘It should probably be taken as a general comment on the results of the teaching of John and Jesus, something like “you will know them by their fruits” (Matt 7:20)’ (The Gospel of Matthew, Sacra Pagina [SP] 1 [Collegeville, MN:  Liturgical Press, 1991], 158). 18. e.g. Fitzmyer, Gospel according to Luke, 1:679: ‘The added saying of Jesus in v. 35 not only allegorizes the parable further in identifying the “children” as children of Wisdom, but even recalls the “vindication” of God by all the people and the toll-collectors of v. 29 above.’ So also Bovon, Luke, 1:287 (on 7.35); Luz, Matthew, 2:146. 19. W. F. Albright and C. S. Mann, Matthew, AB 1 (Garden City : Doubleday, 1971), 139. Failure would have been equally embarrassing in Matthean and Lukan traditions. 20. Bovon, Luke, 1:288. 21.  Matthew has ἔργων for τέκνων. Cf. also Josephus, Ant. 18.116–19, a passage that both opens and closes with references to John’s punishment as God’s vindication: ‘But to some of the Jews the destruction of Herod’s army seemed to be divine vengeance, and certainly a just vengeance, for his treatment of John, surnamed the Baptist . . . Though John, because of Herod’s suspicions, was brought in chains to Machaerus, the stronghold that we have previously mentioned, and there put to death, yet the verdict of the Jews was that the destruction visited upon Herod’s army was a vindication of John, since God saw fit to inflict such a blow on Herod’ (Eng. Trans.: Louis Feldman, Loeb Classical Library [LCL]).

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

169

that Lukan redaction may be at issue in the passage: ‘It appears that Luke has had to alter a saying, originally addressed to the Pharisees, into a statement about them in order to accommodate the material to its present context (7.24).’22 Hopefully, it is clear that the five verses of Luke 7.31–35 pose a discouraging number of issues to the interpreter. For the sake of argument, they may be grouped into two general categories: (1) number of groups of children and the relationship of the children to each other; and (2) compatibility of the parable, interpretation, and wisdom saying, especially relationship of parable to interpretation, that is, the relationship of ‘this generation’/‘children in the marketplace’ to John the Baptist and Jesus. The present essay cannot hope to solve all the issues. It merely proposes a hypothesis that no one, to my knowledge, has yet explored. The hypothesis solves some of the hotly contested issues with one simple interpretive shift, namely, understanding the Son of Man in v. 34 as originally a self-referential expression of John the Baptist.23 The proposal is not entirely new since it involves reducing the subjects of the interpretation from two to one, a theory that has been argued with regard to Jesus (i.e. the people liked neither Jesus’ summons to joy nor his call to repentance, vv. 33–34).24 However, interpretation of Jesus as the Son of Man encounters challenges. Ulrich Luz explains: The origin of the entire text is still an open question. The Son of Man title is what makes it difficult to attribute it to Jesus. ‘Son of Man’ is to be understood here as a title. If we may assume that in Aramaic ‘son of man’ used alone is not a simple equivalent for ‘I,’ but means ‘I’ with a generalizing connotation (‘I’ as human; a human being such as I), then a nontitular ‘son of man’ simply does not fit here where Jesus speaks only of John and himself.25

Interpreting the Son of Man expression as John in v.  33 avoids this difficulty. Without the Son of Man reference, the entire section (vv. 18–35) contains no

22.  Luke, 297. Marshall aptly calls attention to the fact that the parable begins with no indication that direct speech is being introduced (Luke, 299). 23. The ‘Son of Man’ expression is so problematic, it has been argued that it was not original in the present context (Marshall, Luke, 302). 24.  Heinz Schürmann:  ‘Perhaps the metaphor [children] was originally connected to the invitation and the call to repentance of Jesus alone’ (‘The Son of Man Title in the Sayings Source’, in The Shape of Q: Signal Essays on the Sayings Gospel, ed. J. S. Kloppenborg [Minneapolis:  Fortress, 1994], 81). See also Heinz Schürmann, Die Lukasevangelium, Herders Theologischer Kommentar zum Neuen Testament (HTKNT) 3/1 (Freiburg i.Br.: Herder, 1969), 425–6. Cf. Marshall, Luke, 297–8. 25.  Luz, Matthew, 2:148, emphasis added. Heinz Schürmann writes: ‘The titular use of ‘Son of Man’ for the earthly Jesus who is coming again at the parousia and judgment is certainly secondary, though perhaps Palestinian’ (‘Son of Man Title in the Sayings Source’, 81); Die Lukasevangelium, 428–9.

170

Clare K. Rothschild

explicit reference to Jesus.26 Therefore, in this essay, I postulate that John is the sole subject and speaker of Luke 7.31–35. Preliminarily recommending the hypothesis is that the two prior sections in both Luke and Matthew treat John (Luke 7.18–30). In what follows, I will explore whether the entire section (7.24– 35) is originally traceable to John, their incorporation into this Lukan secondary or tertiary context accounting for much of the confusion surrounding the passage.27 First, I will examine vv. 31–35 in terms of the issues raised above, after which I will consider vv. 24–28 for coherence with vv. 31–35 and other Baptist traditions.

9.3 Analysis 9.3.1 Luke 7.31–35 We begin by considering the two general interpretive conundrums noted above: (1) number of groups of children and the relationship of some children to others; and (2) compatibility of the three parts of this section (i.e. parable, interpretation and wisdom saying), especially the relationship of the parable to the interpretation, that is, of ‘this generation’ to John the Baptist and the Son of Man. 9.3.1.1 Groups of Children and Intergroup Relationships If, with most commentators, we interpret Luke 7.31–35 as part of a larger literary context beginning in v. 18,28 then these five verses perpetuate discussion of John the Baptist’s identity as prophet of God. This final segment takes the form of a parable introduced by a

26. All the arguments that Son of Man is Jesus’ self-referential expression may apply to John the Baptist. See Carsten  Colpe, ‘ὁ υἱος τοῦ ἀνθρώπου’, Theological Dictionary of the New Testament  (TDNT) 8:429–61, esp. 430–3. 27.  With characteristic perspicacity, after denying that Elijah is the precursor of the Messiah in pre-Christian Judaism, Fitzmyer outlines the problem: ‘In Stage I of the gospel tradition, John seems to have thought of Jesus as Elias redivivus, the “One who is to come” in the role of the fiery reformer. Jesus arrives on the scene and plays out his role as the bringer of the bounties promised by Isaiah. John hesitates and doubts whether he is really the ‘One who is to come’ (7:19). Then Jesus reverses the roles: John is Elias redivivus (7:27). Then, because Jesus is eventually recognized in the gospel tradition (Stages II–III) as the “Messiah”, and because John has been identified by Jesus himself as this precursor and as Elijah (implicitly, in Luke), Elijah becomes the precursor of the Messiah. Something like this:  John:  Jesus:  Elijah:  Messiah’ (Gospel according to Luke, 1:672–3). One need not agree with this model to appreciate the attempt to clarify redactional layers. 28.  e.g. Schürmann (‘Son of Man Title in the Sayings Source’, 81): ‘But the Son of Man logion – along with Q 7:35 – does not merely comment on the foregoing metaphor. At the same time it concludes the entire section on the Baptist in Q 7:18–35.’

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

171

rhetorical question. The present ‘generation’ (a pejorative term)29 is compared to children sitting in the marketplace making music. The children comprise either one or two groups. Three scenarios are possible. In the first scenario, one group of children may call out the entire rhyme, ‘When we play the flute . . .’ and ‘when we play a dirge . . .’ you (i.e. unspecified other children) refuse to act. In the second scenario, one group is divided in half. Half plays the flute expecting others to dance but receives no response. The non-respondent half, then, plays or sings a dirge expecting wailing but likewise receives no response.30 The third scenario envisages two groups. Children of Group #1 want their comrades to dance, whereas unrelated children of Group #2 want their comrades to cry. The difference between two halves of one group and two separate groups lies in the origin of the second group, that is, whether they are a part of the one large corps or possess status as outsiders vis-à-vis the original corps. Without further information it would be impossible to decide how many groups are implied; however, study of the passage’s parallelism offers a clue. Virtually all commentators interpret the rhyme as reflecting antithetical parallelism. ‘Piping’ or ‘playing a flute’ (ηὐλήσαμεν < αὐλέω) and ‘chanting’ or ‘singing’ ‘a dirge’ (ἐθρηνήσαμεν < θρηνέω) are understood as opposites on the analogy of comedy and tragedy. Yet the αὐλός (‘flute’), an instrument associated with Athena and possibly the most common Greek instrument, was played for a wide variety of different social occasions, including festivals, processions of births and deaths, athletic games, private drinking parties and theatrical performances of tragedy. Furthermore, the Greeks had a strong tradition of dance. Ritual dance was an integral part of ancient Greek culture.31 It was considered divinely inspired and

29.  Cf. 9.41; 11.29–32, 50–51; 17.25; Acts 2.40. In the OT:  Jer 2.31; 7.29; Deut 32.5; Ps 78.8. Cf. other negative uses of γένεα:  Matt 12.39–42; 16.4; 17.17; 23.36; 24.34. See Bovon: ‘In the background, then, is the deuteronomistic motif of obduracy’ (Luke, 1:284). Marshall too offers the background: ‘Behind it [γένεα] lies the usage in Dt. 32:5, 20; Jdg. 2:10; Pss. 78:8; 95:10; Je. 7:29, which suggests the faithlessness of Israel and its subjection to the wrath of God’ (Luke, 299). Cotter points out that the verb προσφωνέω (‘to address’) is consistently used in a ‘formal context’, addressing the people or a king (e.g. 2 Macc 15.15; Acts 21.40; 1 Esd 2.21) (‘Children in the Market-Place’, 296–7). She also notes that the use of κάθημαι and ἀγορά together suggests judges (302): ‘[T]he parable does not draw us into the world of children, but into the world of adults, and in particular the centres of civic justice, the courts’ (302). 30.  Franz Mussner evaluates all the ways in which the game might be construed, concluding that the game involves two groups, each speaking one half of the rhyme (‘Der nicht erkannte Kairos [Mt 11,16–19 = Lk 7,31–35]’, Biblica 40 [1959]: 599–612, here 599–601). See also Zeller, ‘Die Bildlogik’, 252–7. Plummer comments (St. Luke, 207): ‘There are two parties of children. This is more clearly marked by τοῖς ἑτέροις in Mt. than by ἀλλήλοις here.’ Pace Bovon, Luke, 1:286 and n. 67. 31. On Greek dance, see W. O. E. Oesterley, Sacred Dance in the Ancient World (Mineola, NY:  Dover, 2002); Lilian B. Lawler, The Dance in Ancient Greece (Connecticut:  Wesleyan University Press, repr. 1978 [1st edn, 1962]); Steven H. Lonsdale, Dance and Ritual Play in Greek Religion (Baltimore, MD:  Johns Hopkins University Press, 1993). On Roman

172

Clare K. Rothschild

inseparable from music; dancers often sang. Its civilizing effect made it more crucial to tragic than other festive occasions. It was thought to communicate wisdom and truth just like music and poetry. All present at holidays and other special occasions would be expected to dance, including funerals. The κομμός was a powerful dirge to which dancers beat their breasts. Thus, Luke 7.33, ‘We played the flute for you, and you did not dance’ does not necessarily imply a ‘wedding’ or even a joyous occasion.32 In fact, it is possible that – prompted by the purported presence of two individuals in the interpretation section (i.e. John and the Son of Man, vv. 33–34) – scholars have falsely assumed that the rhyme featured two different types of social occasions when it is more likely that the two lines are synonymously parallel. The sense of the parable would then be that some children prompted others to dance by playing a traditional funeral melody on the flute, but the others disregarded their social obligation. Then the same children prompted the others to dance by playing a funeral dirge on the flute, but again the others ignored their obligation. In other words, the rhyme features only one circumstance, a funeral, and only one ritual requirement, the obligation to dance. The complaint is that some children exercise defiant refusal. Support for this reading is found in the widely cited parallel in Herodotus 1.141 (cf. Aesop, Fable 27b)33 that features only the circumstance of death and only the ritual of dance: dance, see Richard C. Beacham, Spectacle Entertainments of Early Imperial Rome (New Haven:  Yale University Press, 2011); Richard C.  Beacham, The Roman Theatre and Its Audience (Cambridge, MA:  Harvard University Press, reissue edn 1996); and Günther Wille, Musica Romana:  Die Bedeutung der Musik im Leben der Römer (Amsterdam:  P. Schippers, 1967). On Jewish wedding dance, see Michael L.  Satlow, Jewish Marriage in Antiquity (Princeton: Princeton University Press, 2001). I wish to thank Charles Cosgrove for his very generous assistance on this topic. 32. Greek dance was of two primary types: gymnastic and mimetic. Mimetic dancing might represent the life and adventures of a god (e.g. Dionysia). The dance known as the βακχική (Lucian, Salt. 79) was a Satyric dance practiced in Ionia and Pontus in which illustrious statesmen took part. Often spectators became so captivated they sat for the whole day watching. See William Smith, A Dictionary of Greek and Roman Antiquities (London: John Murray, 1875). 33. The entire Greek text is provided because the degree of semantic conformity may compel some readers to do their own lexical comparisons: Ἁλιεύς τις αὐλοὺς εἶχε καὶ σοφῶς ηὔλει· καὶ δή ποτ’ ὄψον ἐλπίσας ἀμοχθήτως πολὺ πρὸς αὐλῶν ἡδυφωνίην ἥξειν, τὸ δίκτυον θεὶς ἐτερέτιζεν εὐμούσως. ἐπεὶ δὲ φυσῶν ἔκαμε καὶ μάτην ηὔλει, βαλὼν σαγήνην εἷλκεν ἰχθύων πλήρη. ἐπὶ γῆς δ’ ἰδων σπαίροντας ἄλλον ἀλλοίως, τοιαῦτ’ ἐκερτόμησε τὸν βόλον πλύνων· ‘ἄναυλα νῦν ὀρχεῖσθε. κρεῖσσον ἦν ὕμας πάλαι χορεύειν, ἡνίκ’ εἰς χοροὺς ηὔλουν.’ [Οὐκ ἔστιν ἀπόνως οὐδ’ ἀλύοντα κερδαίνειν. ὅταν καμὼν δὲ τοῦθ’ ἕλῃς ὅπερ βούλει, τὸ κερτομεῖν σοι καιρός ἐστι καὶ παίζειν.] ‘There was once a fisherman who saw some fish in the sea and played on his pipe, expecting them to come out onto the land. When his hopes proved false, he took a net and used it instead, and in this way he was able to haul in a huge catch of fish. As the fish were all leaping about, the fisherman remarked, “I say, enough of your dancing, since you refused to dance when I played my pipe for you before!” ’ (Babrius #9, Perry #11 in Aesop’s Fables. Trans. Laura Gibbs, World’s Classics [Oxford: Oxford University Press, 2002]).

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

173

Once, he said, there was a flute-player who saw fish in the sea and played upon his flute, thinking that they would come out on to the land. Disappointed of his hope, he cast a net and gathered it in and took out a great multitude of fish; and seeing them leaping, he said, ‘You had best stop your dancing now; you would not come out and dance before, when I played to you.’34

Herodotus records that Cyrus employed this parable in a statement to an Ionian and Aeolian messenger bearing news that his people, although unwilling to revolt from Croesus prior to Cyrus’s victory, now offered to be Cyrus’s subjects on the terms they had under Croesus. Naturally Cyrus refuses. The point is not that by obeying Cyrus’s order the Ionians and Aeolians might have avoided domination, but that by their disobedience they had burned a bridge, so to speak  – later opportunistically hoping that their egregious offence would be overlooked. On the terms of the parable, the sense is not that some fish might have saved their lives. All fish will be caught either way. The sense is also not when all fish will be caught, since the difference in timing is negligible. Rather, the parable emphasizes obedience – namely, dancing when the music is played – and foolishness – that is, imagining that after a flagrant act of disobedience some fish might alter their fate. The very moment that the flute-player appears, the fate of every Ionian and Aeolian fish is sealed. Luke 7.32b appears to reduce this brief fable to a two-line aphorism featuring the most prominent characteristic of Hebrew poetry, parallelism. In the epitomization process, the context is lost.35 In all likelihood, the context from which Q adopted the saying featured God’s prophets as the flute-players. Musical instruments often served as vehicles of a prophetic message (e.g. 1 Sam 10.5; 1 Chron 25.1). When the prophets arrive with God’s message of judgement (the music begins), it is incumbent upon the people not simply to listen, but to dance. Once God casts his net, however, it is too late. By then, they will have either obeyed or disobeyed; their fate will be sealed. The pipers sit because, according to ancient custom, musicians may ‘sit’ (καθημένοις, v. 32) when they are playing their instruments (whether it is necessary or not). When not playing their instruments, however, even the musicians are expected to dance. In short, vv. 33 and 34 do not reflect antithetical parallelism comparable to comedy and tragedy or wedding and funeral. The event is a funeral. Each verse alone contains an antithesis (God’s prophets [pipers] vs. disobedient others [dancers who refuse to dance]), but the two verses together reflect synonymous parallelism. This parable characterizes the generation of its speaker as comprised of two groups of children: (1) Group #1, God’s prophets, represented by John the Baptist in the present generation, piping a dirge to warn the people of the imminent eschaton; and (2) Group #2, the Pharisees and lawyers – disobedient Jews refusing to 34. Engl. Trans.: A. D. Godley (LCL, with minor modifications). 35. Demetrius, Eloc. 89 notes the importance of brevity when transforming a metaphor into a simile.

174

Clare K. Rothschild

acknowledge God’s message through John. Since John is referred to as κάλαμος in v. 24, ἄνθρωπος in v. 25, and προφήτης in v. 26, when Luke inserts the accusative subject, τοὺς ἀνθρώπους in the initial question in v. 31 (τίνι οὖν ὁμοιώσω τοὺς ἀνθρώπους τῆς γενεᾶς ταύτης καὶ τίνι εἰσὶν ὅμοιοι, cf. Matt 11.16, τίνι δὲ ὁμοιώσω τὴν γενεὰν ταύτην;), it is clear that John is included in this generation.36 In. v. 34, the Son of Man is likewise referred to as ἄνθρωπος – φίλος of tax collectors and sinners.37 In the Gospel of Luke only those who misunderstand him refer to Jesus an ἄνθρωπος (e.g. 23.4, 6, 14, 18). The parable projects them attempting too late to ‘dance’ on Judgement Day.38 Similar to Luke 13.28 (κλαυθμός), the parable concludes with ‘weeping’.39 The environs of the Jordan River and Galilee (with its lake) make it easy to imagine how such a non-biblical saying became integrated among its earliest traditions.40 9.3.1.2 Compatibility between the Parable and Its Interpretation The main impediment to compatibility between the three elements of this section is applicability of the interpretation to the parable. Having discussed the parable, it remains to discuss its relationship to the interpretation and final wisdom saying. Utilizing a Baptist hermeneutic, our initial observation is that, if ὁ υἱὸς τοῦ ἀνθρώπου represents John, then together with the main verbs in each verse (ἐλήλυθεν), and the

36.  Although affirmation is given only to the final question in v. 26, John surely merits all three distinctions (i.e. reed, man and prophet), albeit with qualifications. He is a man dressed in soft clothing, fit for a kingdom (Matt 11.8, ἰδοὺ οἱ τὰ μαλακὰ φοροῦντες ἐν τοῖς οἴκοις τῶν βασιλέων εἰσίν), insofar as v. 28 places John as least in the kingdom. 37.  Cf. also Luke 20.4 in which the author implies that as an ἄνθρωπος John’s baptism was from heaven. 38.  W. C. Van Unnik traces Jesus’ dance in the Acts of John 94–97 to the Passover sermon by Bishop Melito of Sardis in which eating and drinking gives way to praise, in turn, giving way to dancing. ‘A Note on the Dance of Jesus in the Acts of John’, Vigiliae Christianae (VC) 18 (1964): 1–5. Text: Campbell Bonner, The Homily on the Passion by Melito, Bishop of Sardis, in Mélanges Franz Cumont, Annuaire de l’Institut de philologie et d’histoire orientales et slaves 4 (1936), 107–19. 39. Interesting also is Luke 13.29, which states that after the weeping and gnashing of teeth (ὁ κλαυθμὸς καὶ ὁ βρυγμὸς τῶν ὀδόντων, v. 28), people will come from east and west, from north and south, and will recline for dinner in the kingdom of god (καὶ ἥξουσιν ἀπὸ ἀνατολῶν καὶ δυσμῶν καὶ ἀπὸ βορρᾶ καὶ νότου καὶ ἀνακλιθήσονται ἐν τῇ βασιλείᾳ τοῦ θεοῦ). Fitzmyer comments about ‘reclining at table in the kingdom’: ‘I.e. and share in the eschatological banquet to be provided by God for his chosen ones. See the OT motif and its development in Isa 25:6–8; 55:1–2; 65:13–14; 1 Enoch 62:14; 2 Bar. 29:4; Pirqe ʼAboth 3:20. Luke alludes to it again in 14:15; 22:16, 29–30; cf. Rev 3:20; 19:9’ (Gospel according to Luke, 2:1026). 40. See n. 31 (above).

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

175

references to ‘eating’ and ‘drinking’ (first denied, then affirmed), John and ὁ υἱὸς τοῦ ἀνθρώπου are parallel matching the parallelism of the parable: V. 33, ἐλήλυθεν γὰρ Ἰωάννης ὁ βαπτιστὴς μὴ ἐσθίων ἄρτον μήτε πίνων οἶνον, καὶ λέγετε· Δαιμόνιον ἔχει· V. 34, ἐλήλυθεν ὁ υἱὸς τοῦ ἀνθρώπου ἐσθίων καὶ πίνων, καὶ λέγετε· Ἰδοὺ ἄνθρωπος φάγος καὶ οἰνοπότης, φίλος τελωνῶν καὶ ἁμαρτωλῶν.41

One prophet, John arrives with a message from God for his generation. Sometimes he fasts, other times he resumes eating  – unremarkable behaviour for any Jew, let  alone a prophet.42 Yet rather than listen to and, more importantly, obey his message, the self-appointed interpreters du jour preoccupy themselves with his unexceptional dietary practices. The synonymous parallelism of the parable and the synonymous parallelism of most of the interpretation strongly suggest that John the Baptist and the Son of man are to be identified as one and the same individual. It is typical of similes and parables to draw their comparison with the aim of censure.43 9.3.1.3 Excursus: The Dietary Habits of John the Baptist Although most commentators prefer to cite punishment of the rebellious children in Deut 21.18–21 as a parallel to the accusation of ‘glutton and drunkard’ against the Son of Man in Q 7.34, a few stipulating factors recommend a comparison with 1 Kgs 19.8 instead. First, the rare substantives (i.e. φάγος, οἰνοπότης) of Q 7.34 share the same Greek roots as the two verbs in 1 Kings (i.e. φαγεῖν, πίνειν). Although the law concerning children charged with disreputable conduct in defiance of parental authority in Deut 21.20 in Hebrew (MT) specifies eating and drinking as evidence of disobedience,44 the Septuagint’s version of this passage does not (i.e. ἀπειθεῖν, ‘to be disobedient’ and ἐρεθίζειν, ‘to rouse to fight’): Q 7.34:  ἐλήλυθεν ὁ υἱὸς τοῦ ἀνθρώπου ἐσθίων καὶ πίνων, καὶ λέγετε· ἰδοὺ ἄνθρωπος φάγος καὶ οἰνοπότης, φίλος τελωνῶν καὶ ἁμαρτωλῶν. 41. Jesus’ disciples eat and drink with tax collectors and sinners at Luke 5.30. 42. One might ask whether John ate. In fact, it would be incompatible with the portrait of John in Luke (especially) and Matthew (but not Mark) to claim that John only fasted. John’s diet is missing in Luke; pace Joel B. Green who claims ‘John is known in the Gospel of Luke for his ascetic practices (cf. 1:15; 5:33)’, The Gospel of Luke, New International Commentary on the New Testament (NICNT) (Grand Rapids, MI:  Eerdmans, 1997), 303. See James A. Kelhoffer, The Diet of John the Baptist: ‘Locusts and Wild Honey’ in Synoptic and Patristic Interpretation, WUNT 176 (Tübingen: Mohr Siebeck, 2005), 4–12. See ‘Excursus’ below. 43. Rhet. Her. 4.62. 44. ‫סוֹרר וּמ ֶֹרה ֵאינֶ נּוּ שׁ ֵֹמﬠַ ְבּקֹלֵ נוּ זוֹלֵ ל וְ ס ֵֹבא‬ ֵ ‫וְ ָא ְמרוּ ֶאל־זִ ְקנֵ י ﬠִ ירוֹ ְבּנֵ נוּ זֶ ה‬:

176

Clare K. Rothschild

1 Kings 19.8 (LXX):  καὶ ἀνέστη καὶ ἔφαγεν καὶ ἔπιεν· καὶ ἐπορεύθη ἐν τῇ ἰσχύι τῆς βρώσεως ἐκείνης τεσσαράκοντα ἡμέρας καὶ τεσσαράκοντα νύκτας ἕως ὄρους Χωρηβ. Deut 21.20 (LXX):  καὶ ἐροῦσιν τοῖς ἀνδράσιν τῆς πόλεως αὐτῶν ῾Ο υἱὸς ἡμῶν οὗτος ἀπειθεῖ καὶ ἐρεθίζει, οὐχ ὑπακούει τῆς φωνῆς ἡμῶν, συμβολοκοπῶν οἰνοφλυγεῖ.

Second, the wider context of Q 7.18–35 resembles the story line in 1 Kgs 19.8 in which Elijah is said to have fasted until the Lord deemed food necessary for his journey thereby providing it for him: ‘And he [Elijah] arose and ate and drank, and he went in the strength of that food forty days and forty nights as far as Mount Horeb.’ Elijah fasted until God provided food; then he ate and fasted again.45 Just like this passage, fasting precedes eating in Q 7.33–34. It goes without saying that the correspondence between John and Elijah in Q also recommends 1 Kings over Deuteronomy. Third, Matthew does not have an equivalent to Luke 7.29–30. Instead, Matthew 17 contains four different verses (vv. 12–15) immediately following the ‘among those born of women’ Q statement about John. The first two verses of Matthew’s insertion comment on violence against the kingdom of God beginning in the days of John the Baptist. Then, v. 14 specifies that John the Baptist is Elijah, a verse echoed later in Jesus’ conversation with his disciples after the transfiguration. Coming down the mountain, Jesus discusses the appearance of ‘Elijah with Moses’ (καὶ ἰδοὺ ὤφθη αὐτοῖς Μωϋσῆς καὶ Ἠλίας, Matt 17.3) on the mountain after which Matthew clarifies that John the Baptist is Elijah: He replied, ‘Elijah is indeed coming and will restore all things; but I  tell you that Elijah has already come, and they did not recognize him, but they did to him whatever they pleased. So also the Son of Man is about to suffer at their hands.’ Then the disciples understood that he was speaking to them about John the Baptist (Matt 17.11–13, NRSV).46

Similar to Q 7.33–34, in the above passage a statement about John is followed by another purportedly about Jesus, but referred to as ὁ υἱὸς τοῦ ἀνθρώπου. Finally, while difficult to demonstrate that the expression ‘Son of Man’ would have been understood in the same way in Greek (ὁ υἱὸς τοῦ ἀνθρώπου) as it was in Hebrew (‫ )בן־אדם‬or Aramaic (‫)ברנש‬, namely, as another expression for ‘man’, this problem is

45.  The δαίμων that enables John to fast might also compel him to dance, that is, connecting him to the reference to dancing in v. 32. 46. On the resurrection of John the Baptist in Mark, see Baptist Traditions and Q, 129– 72. In John 3.13, Elijah is described as ὁ υἱὸς τοῦ ἀνθρώπου: ‘No one has ascended into heaven [Elijah] except the one who descended from heaven, the Son of Man.’ Cf. Jas 5.17 (1 Kgs 17.1; 18.41–45): ‘Elijah was a human being like us (Ἠλίας ἄνθρωπος ἦν ὁμοιοπαθὴς ἡμῖν), and he prayed fervently that it might not rain, and for three years and six months it did not rain on the earth.’

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

177

not unique to John.47 Moreover, with John’s qualification, ἐν γεννητοῖς γυναικῶν (‘among those born of women’) in v. 28 may emphasize the same point – namely, the distance between a divine God and a human being. In sum, if ὁ υἱὸς τοῦ ἀνθρώπου refers to John not Jesus in v. 34, then the long debated question concerning Jesus’s self-description as ὁ υἱὸς τοῦ ἀνθρώπου would have a new explanation:  namely, tradition (i.e., Jesus learned it from his teacher, John).48 Reducing the individuals in the interpretation to one, namely, John the Baptist, solves a few additional problems that are well documented in the history of the interpretation of this passage. First, understanding ‘the Son of Man’ as John not Jesus in v. 34 eliminates the problem of the order of elements in the interpretation, namely, that John’s fasting followed by Jesus’ feasting correlates incorrectly with ‘piping’/’dancing’ followed by ‘mourning’/’weeping’. Second, this solution avoids the apparent equality (or even the priority of John – because possessing a demon that facilitates extended prophetic fasting is superior to the designation of glutton and drunkard)49 of Jesus and John contradicting a clear bias for Jesus elsewhere in Luke and Matthew, yet corresponding to Q’s predilection for John. Finally, if ὁ υἱὸς τοῦ ἀνθρώπου represents John not Jesus, then he, not Jesus, is the ‘friend of tax collectors and sinners’ (cf. tax-collectors and soldiers, Luke 3.12–14). This eliminates the anomalous reference of Jesus as friend.50 John is referred to as ὁ φίλος τοῦ νυμφίου in John 3.29,51 yet in the entire NT and corpus of ECL, Jesus is never referred to as φίλος. Summarizing findings thus far, if the Son of Man in Q 7.34 refers to John, the interpretation implies that the Pharisees and the lawyers who refused John’s baptism (v. 29)  did so out of disobedience to God.52 They obsessively monitored John’s eating habits for a sign of his authenticity, scrutinizing both his

47. Colpe, ‘ὁ υἱὸς τοῦ ἀνθρώπου’, esp. 402–5. 48.  Cf. Paul’s similar reference to Jesus as son of God, ‘born of a woman’ in Gal 4.4: ἐξαπέστειλεν ὁ θεὸς τὸν υἱὸν αὐτοῦ, γενόμενον ἐκ γυναικός. 49. See n. 14 above. 50. See above. 51.  Cf. Luke 5.34, 35 in which Jesus refers to himself as the bridegroom:  ὁ δὲ Ἰησοῦς εἶπεν πρὸς αὐτούς· μὴ δύνασθε τοὺς υἱοὺς τοῦ νυμφῶνος ἐν ᾧ ὁ νυμφίος μετ’ αὐτῶν ἐστιν ποιῆσαι νηστεῦσαι; ἐλεύσονται δὲ ἡμέραι, καὶ ὅταν ἀπαρθῇ ἀπ’ αὐτῶν ὁ νυμφίος, τότε νηστεύσουσιν ἐν ἐκείναις ταῖς ἡμέραις. 52.  Green notes the connection between the parenthetical insertion and what follows: ‘The degree to which Luke’s parenthetical aside in vv. 29–30 is integrated into the logic of Jesus’ address to the crowds is indicated by the use of a resultative “then” (or “therefore”) at the beginning of v. 31 and by the parallelism of vv. 29–30 and vv. 31–35. At first glance, it appears that Jesus is drawing out the implications of Luke’s commentary on the polarity of responses to John’s message of repentance-baptism: Since the Pharisees and scribes rejected God’s purpose of themselves, to what shall I compare those of this generation?’ Green also

178

Clare K. Rothschild

fasting (sheer luck)53 and his eating (glutton, drunkard).54 Rather than embrace his message that the kingdom of God will feature God’s justice (reward for the repentant whose works exemplify righteousness vs. punishment for the unrepentant), they debated, prevaricated and vacillated, all the while insisting on their own exonerated status as children of Abraham. In doing so, they missed the decisive hour.55 9.3.1.4 Compatibility between the Parable, Interpretation and Wisdom Saying How then might we understand the relationship of the parable and its interpretation to the final Wisdom saying of v. 35? Although Ragnar Leivestad argues that the appearance of σοφία is ironic56 and Stephan Witetschek argues that in the final saying Jesus ‘turns the blame on them: By rejecting (John and) Jesus they themselves prove that they are not on the side of Wisdom’,57 a Baptist hermeneutic views the final wisdom saying as a challenge to its audience. At the level of design, it poses a riddle or αἴνιγμα with the implication that it is not to be decoded in terms of allegory. It makes its point like an ἔλεγχος, by demanding that the audience (within and outside of the narrative) decide for themselves. Luke 7.31–34 addresses rejection of God’s prophets. The summation is posed: Wisdom’s children will prove her correct, that is, the prophets will be vindicated.58 As in Q 11.49–51, this generation persecuted and killed some of its prophets and retribution will be demanded.59 Just like the disobedient fish in the parable cited by Herodotus, once

acknowledges, however, that vv. 31–35 also flow well from v. 28 (Luke, 302). He also points out parallelism between vv. 29 and 35, arguing for a chiasm: vv. 29, 30, 31–34, 35 (304). Marshall observes: Matthew has no parallel to v. 29 (but cf. 21.31) (Luke, 296). The passage raises the question of tests of authenticity in early Christian sources regarding prophets and also apostles. 53. Cf. the refutation in Gos. Ebion. 2 which replaces ‘locusts’ with ‘cake’. 54. Differently, Mark portrays John’s diet as solely consisting of locusts and wild honey (and on which one could survive only a matter of weeks), Matthew more plausibly records that John occasionally ate locusts and wild honey. See Kelhoffer, Diet of John the Baptist, 4– 12 and the literature cited, esp. 11 n. 21. Cf. also descriptions of John the Baptist in Josephus, Ant. 18.5.2 [§§116–19]). 55. Fitzmyer, Gospel according to Luke, 1:680. 56.  Leivestad, ‘An Interpretation of Matt 11:19’, Journal of Biblical Literature (JBL) 71 (1952): 179–81, here 180. 57. Witetschek, ‘Stigma of a Glutton and Drunkard’, 151, emphasis added. 58. Johan Huizinga, Homo Ludens: A Study of the Play Element in Culture (Boston: Beacon Press, 1950), 109. 59.  ‘Therefore also the Wisdom of God said, “I will send them prophets and apostles, some of whom they will kill and persecute,” so that this generation may be charged with the blood of all the prophets shed since the foundation of the world, from the blood of Abel to the blood of Zechariah, who perished between the altar and the sanctuary. Yes, I tell you, it will be charged against this generation.’ (NRSV)

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

179

the net is cast, it is too late to dance. Although all past prophets ignored by Israel will one day vindicate Wisdom, in this saying concerning the present generation, John the Baptist is the specific child denoted.60 He is the child playing the pipe who will one day vindicate Wisdom. Such a solution does not, however, deny that the saying contains a conundrum. Like an aporia parlour game, it remains for the reader to figure out who Wisdom’s children are. The riddle of the Sphinx may lurk in the background particularly since, in both cases, the penalty for an incorrect answer is death. Such a riddle was designed to invite reflection (by both the narrative and real audiences) but defeat comprehension. The question was meant to ‘catch’, expose, or refute (ἐλέγχειν) smug confidence. Baptist traditions demonstrate a predilection for this strategy. One need only think of his tantalizingly puzzling description of the future: I baptize you with water; but one who is more powerful than I is coming; I am not worthy to untie the thong of his sandals. He will baptize you with the Holy Spirit and fire. His winnowing fork is in his hand to clear his threshing floor and to gather the wheat into his granary; but the chaff he will burn with unquenchable fire.

This saying has powerful parallels to the one about the children of wisdom under consideration. While the topic cannot be treated at length, it constitutes another riddle. The Fourth Gospel offers two additional, enigmatic Baptist sayings: ‘Behold, the Lamb of God’ and ‘he must increase and I must decrease’. 9.3.1.5 Self-referentiality: Luke 7.31–35 as a Saying of John Before closing, I would like to entertain the possible self-referentiality of Q 7.31–35. As noted above, the parable beginning in Q 7.31 bears a resemblance to John’s address to the people in Q 3.7. Q 3.7 addresses ‘generation of vipers’ (γεννήματα ἐχιδνῶν); Q 7.31 addresses ‘this generation’ (τίνι οὖν ὁμοιώσω τοὺς ἀνθρώπους τῆς γενεᾶς ταύτης, καὶ τίνι εἰσὶν ὅμοιοι;). If we incorporate Q 7.24–28 into this comparison we see that in v. 24 John’s disciples depart after which the text states that ‘he’ begins to speak to the crowds about ‘him’. As with v. 21, all translations and commentaries understand Jesus as the implied subject, John the implied object. Given the level of obvious redaction in this Gospel, however, a limited measure of doubt is in order.61 If we extend the Baptist hermeneutic to vv. 24–28 and imagine momentarily that John is the speaker, we note a similar emphasis on interrogatives: Question #1a: What did you go out into the wilderness to look at? Question #1b: A reed shaken by the wind?62 Question #2a: What then did you go out to see? 60. John S. Kloppenborg (‘The Formation of Q and Antique Instruction Genres’, in The Shape of Q, 155) writes: ‘Q 7:35 represents Jesus and John as Sophia’s children.’ 61. Matthew supplies Jesus in these cases. 62. Cf. Gos. Thom. 78. These are sarcastic responses reminiscent of the ἔλεγχος.

180

Clare K. Rothschild

Question #2b: A person dressed in soft robes? Look, those who put on fine clothing and living in luxury are in royal palaces. Question #3a: What then did you go out to see? Question #3b: A prophet? Answer: This is the one about whom it is written, ‘See, I am sending my messenger ahead of you, who will prepare your way before you.’ I tell you, among those born of women no one is greater than [I]; yet the least in the kingdom of God is greater than [I].63 We note that once the answer, incorporating a citation of Mal 3.1 (cf. 4.5–6), is finally provided, it concerns John’s identity as messenger preparing the way of the Lord  – John’s distinctive claim in Mark 1.2–3 (attributing it to Isa 40.3). In Luke 1.76, Zechariah alludes to this passage (i.e. Mal 3.1) in his benediction on the birth of John the Baptist (1.76). In John’s programmatic statement in Q 3.7 (τίς ὑπέδειξεν ὑμῖν φυγεῖν ἀπὸ τῆς μελλούσης ὀργῆς;), interrogation is also the preferred mode of instruction.64 Returning to Q 7.31, the passage begins with a double interrogative formulation: ‘To what shall I compare this generation? And what are they like?’ The echo in v. 31 of vv. 24–28 would have been immediately recognized in an oral presentation of this material. In Q 7.24–28, we also note an emphasis on empirical evidence. Three times the crowd is asked what they went out to the wilderness to see: ‘What did you go out in the wilderness to look at? . . . What then did you go out to see? . . . What then did you go out to see?’ First, the verb ἐξέρχομαι (vv. 24, 25, 26) implies that both speaker and audience are in the wilderness – a setting favouring John over Jesus. Next, these three parallel rhetorical questions pick up on John’s parallel question in v. 19 (repeated in v. 20): ‘Are you the one who is to come? Or are we to look for another?’ Jesus answers that John’s messengers ought to tell John about the miracles they have seen (v. 21), but, again, the pronoun in v. 21 (cf. v. 24) makes the subject ambiguous. Furthermore, it is somewhat mystifying why, in Q 7.31–35, Jesus makes children the subject of his parable. Nowhere else does Jesus use children as a metaphor; the only children he speaks about are Jews (as the literal children of God) and actual children.65 In Q 3.7–9, John refers to rocks as potential children of God. Furthermore, in both this passage and 7.31–35, this generation is compared to children. Q 3.8b: Πατέρα ἔχομεν τὸν Ἀβραάμ, λέγω γὰρ ὑμῖν ὅτι δύναται ὁ θεὸς ἐκ τῶν λίθων τούτων ἐγεῖραι τέκνα τῷ Ἀβραάμ; Q 7.31–32:  Τίνι οὖν ὁμοιώσω τοὺς ἀνθρώπους τῆς γενεᾶς ταύτης καὶ τίνι εἰσὶν ὅμοιοι; ὅμοιοί εἰσιν παιδίοις τοῖς ἐν ἀγορᾷ καθημένοις καὶ προσφωνοῦσιν ἀλλήλοις. In both passages, some people 63. Cf. Gos. Thom. 46. 64. Cf. Acts 13.25 in which Luke 3.16 is modified to begin with a question: ‘What do you consider me to be? I am not he. No, but one is coming after me; I am not worthy to untie the thong of the sandals on his feet.’ 65. In Matt 13.38, ‘seed’ is a metaphor for children.

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

181

who consider themselves ‘children’ do not measure up.66 In the first passage, they claim to be children but cannot be since they bear no resemblance to their purported father; and in the second passage John agrees that some are children, but of a disobedient type soon to suffer punishment by their father. Q 3.7 and 7.31– 35 also share the eschatological requirement of ethical action.67 Q 3.7 demands καρποὺς ἀξίους τῆς μετανοίας (‘fruit worthy of repentance’), threatening those who fall short with punishment by fire. Q 7.31–35 castigates intransigence, threatening any who fall short with Wisdom’s vindication.68 Naturally, the question arises, if Q 7.31–35 is a saying of John the Baptist (and the ‘Son of Man’, a self-referential expression of John’s), must we also agree that John refers to himself by his own name and title, Baptizer, in v. 33? It is possible he did as the Wisdom teacher Qoheleth seems also to have done,69 although it is equally likely that the proper nouns reflect the redactional habit of either evangelist. With regard to Marcion’s Gospel (Mcg), neither Tertullian nor Epiphanius makes any reference to vv. 7.29–35. According to S. Baring-Gould: That verses 29–32 should have been purposely excluded, it is impossible to suppose, as they favored Marcion’s tenets. It has been argued that the rest of the verses, 33–35, were cut out by Marcion because in verse 34 it is said, ‘The Son of Man is come eating and drinking; and ye say. Behold a gluttonous man and a winebibber.’ But the ‘Gospel of the Lord’ contained Luke v. 33: ‘Why do the disciples of John fast often, and make long prayers, and likewise the disciples of the Pharisees; but thine eat and drink;’ and the example of Christ going to the feast prepared by Levi is retained (v. 29).70

Since Tertullian refers to vv. 7.18–28 and Epiphanius notes vv. 7.23, 27, it is possible that if vv. 7.29–35 were missing, one or the other would have pointed this out,

66.  In Luke 7.31–35 familial language plays a role. Children exchange taunts in the marketplace, but the parent-like Wisdom issues the verdict and riddle. 67.  So also Josephus, Ant. 18.116–19: ‘For Herod had put him to death, though he was a good man and had exhorted the Jews to lead righteous lives, to practice justice towards their fellows and piety toward God, and so doing to join in baptism. In his view this was a necessary preliminary if baptism was to be acceptable to God. They must not employ it to gain pardon of whatever sins they committed, but as a consecration of the body implying that the soul was already thoroughly cleansed by right behavior’ (Engl. Trans.: Louis Feldman, LCL). 68. Finally, we note that Luke 11 also shares significant similarities with Luke 7, including demon possession (Luke 11.15), ὁ υἱὸς τοῦ ἀνθρώπου as a sign to this generation (11.30), the Wisdom of God (11.49), and appreciation of the prophets by the present generation. I wish to thank R. Matthew Calhoun for his assistance on this aspect of my argument. 69. See Peter Machinist, ‘The Voice of the Historian in the Ancient Near Eastern and the Mediterranean World’, Interpretation (2003): 117–37, here 133–4. 70. The Lost and Hostile Gospels (London: Williams and Norgate, 1874), 264.

182

Clare K. Rothschild

unless they did not see these verses in their version of Luke. Since these verses are present in 𝔓45 (𝔓 75 is fragmentary here) and Matt 11.1–19 is parallel, Mcg probably did contain these verses – Tertullian and Epiphanius declining to mention them because they found no variation between Mcg and Luke.71 Dieter T. Roth argues for the methodological necessity of approaching a reconstruction with no use of an argument e silentio either for or against the presence of a passage in Marcion. He, therefore, states only that 7.31–35 is unattested in Mcg. Zahn and Harnack were quite confident that it was missing. Tsutsui and Beduhn state that the passage is unattested, although noting that Harnack thought it was absent. According to Roth, quite more analysis of the attested elements (both attested as present and attested as not present) is necessary before guesses should be hazarded as to the status of unattested passages. What is more, John the Baptist appears in Mcg at Luke 5.33–35 (first mention of John the Baptist in Mcg); 7.18– 20, 22–23, 26, 27, 28; 9.7–8; 11.1.72

9.4 Conclusion Q 7.31–35 poses an array of interpretive difficulties. No simple solution exists. This essay proposes that the passage makes sense if ὁ υἱὸς τοῦ ἀνθρώπου expression in v. 34 is understood as a self-referential designation of John the Baptist. At least two objections to this reading include that in Luke 5.33 the dietary habits of John the Baptist and Jesus are contrasted in a way that may seem comparable at first glance to the description in 7.33–34; namely, John’s disciples are said to fast and pray frequently, whereas Jesus’ disciples are said to eat and drink. We note, however, that the reference occurs in Luke and not Q, it denotes a behaviour of the disciples not John or Jesus, and Jesus’ disciples just eat and are not gluttonous drunks. Second, although John exhorts and baptizes tax collectors and other sinners in Luke 3.12, Jesus’ disciples are accused of eating with tax collectors and sinners in 5.30 and 7.29,73 and in 15.1 the tax collectors and others flock to hear him. Here again, the references occur in Luke not Q, they denote behaviour of disciples not teachers, and in no passage is either Jesus or his disciple referred to as φίλος of such tax collectors and sinners. Luke may have added all three of these passages and others (e.g. 18.10–13; 19.2–7) as a means of smoothing out cooptation of Q’s Baptist sayings for Jesus.

71. https://sites.google.com/site/inglisonmarcion/Home/marcion/marcion-s-gospelcompared-verse-by-verse-with-luke/lk-7. 72.  See Dieter T. Roth, The Text of Marcion’s Gospel, New Testament Tools, Studies and Documents (NTTSD) 49 (Leiden: Brill, 2015), 113–15, 137, 193, 211, 294–5, 365, 400, 403 and 416. 73. We note, however, that Luke 5.29 modifies Mark 2.15, replacing καὶ πολλοὶ τελῶναι καὶ ἁμαρτωλοὶ with καὶ ἦν ὄχλος πολὺς τελωνῶν καὶ ἄλλων.

9. Refusing to Acknowledge the Immerser (Q 7.31–35)

183

The advantages of this reading are as follows: (1) the section coheres with what comes before it (Luke 7.18–30), namely, a discussion of John the Baptist’s identity and significance, including both the Lukan and Matthean insertions (Luke 7.29–30; Matt 11.12–15); (2) the number of groups of children is clarified: Group #1 are God’s prophets and Group #2 are the Pharisees and lawyers; (3) the three components of the section make sense as a unit; (4) the section accurately reflects the fable from which it was derived; (5) the section coheres with Elijah traditions coherently developing the discussion of John’s identity in Q (and Luke); (6) the section encounters no anomalous equality between John the Baptist and Jesus; (7) the interpretation encounters no problem in the order of its elements vis-à-vis the parable; and (8) the section adheres in both form and content to other sayings of John. The primary purpose of the passage in Q is to reprimand those failing to acknowledge John’s baptism. What it reminds us about the Synoptic Problem is precisely what we learn from the Lukan Sondergut of chs 1–3, namely, to be wary of Baptist traditions lurking among the prized traditions of the Jesus.

Chapter 10 C O H E R E N C E A N D D I ST I N C T N E S S : E XPLORING THE S O CIAL M ATR I X OF T H E D OU BL E T R A DI T ION Hildegard Scherer

The synoptic question is far from resolved. The existence of any ‘Q’ document as literary source for the double-tradition materials shared by Matthew and Luke beyond Mark is especially under scrutiny.1 As no material remains of this hypothetical document have ever come to our knowledge, our trial is based on circumstantial evidence – to which I would like to contribute by analysing the social texture of double-tradition materials over against the Markan Jesus traditions as well as the so-called Sondergut.2

10.1 Methodological Considerations This approach recommends itself as it starts out from existing textual material: It is a fact that Matthew and Luke share textual material beyond Mark, and, taking Markan priority for granted, a substantial part of these texts contain traditions 1. The most prominent effort to rethink the Gospel sources is launched from the proponents of a hypothesis labelled as ‘Mark without Q’/ ‘Farrer-Goulder’/‘L/M’ hypothesis. They argue for – or at least wish to have seriously considered – Lukan dependence on Matthew, whereas both of them used and favoured Mark. As recent proponents of this solution, Mark Goodacre and Francis Watson contribute to our volume. 2.  In this chapter, I  summarize findings of my Habilitationsschrift (Bonn 2014; in print: Königsvolk und Gotteskinder: Der Entwurf der sozialen Welt im Material der Traditio duplex, Bonner Biblische Beiträge [BBB] 180 [Göttingen: V & R unipress, 2016]), esp. 365– 439 and 533–47, with full reference to secondary literature, of which I here only quote the most important examples. Whereas in this chapter, I  confine myself to social categories related to household and family, in my book I also treat Israel and related social categories, i. e. genealogy and ethnos construction, prophets, cities and their institutions, the opposing group in Q 11.39–52 and the societal order proposed in the programmatic speech Q 6.20– 49, and moreover the categories related to Basileia, all of them on their socio-historical background.

186

Hildegard Scherer

not found in Mark. What interests me now is the texture of this double-tradition material: Taking into account the patchwork of synoptic tradition, is there some distinctive structure in the double-tradition fabric? Are there characteristic threads which are not, or not fully, interwoven with their surroundings? And how does this texture relate to Markan texts? To extract some texture from textual materials, we are in need of parameters running through as much of the material as possible. As an example, I have focused on social categories, by which I understand any categories grouping human beings. Social categories seem to me an apt parameter, as Early Christian texts deal with a group of human beings. They address the Jesus group more or less overtly and profile it against other culturally established groups. In their narrative worlds, Early Christian texts thus construct social worlds. The social categories operating in the worlds of the text offer modes of social identification to their audience, applicable to ‘oneself ’ as well as ‘the others’. By means of attribution, they shape the perception and evaluation of the groups.3 To understand in full the social categories thus involved in the texts, they are to be examined on the basis of their cultural background, which provides the necessary inferences – and at the end of the day, patterns may appear which may be distinctive for some textual corpus. Working with double-tradition material, we have to consider further methodological premises: In contrast with Mark, we do not know with certainty the wording and the extent of any would-be source of the double tradition. As a consequence, we have to base our work on a ‘critical minimum’4: To circumvent hazardous reconstructions, the arguments should be based on categories and text fragments which are identical in the Matthean and Lukan versions.5 As most of the stylistic features or word order are not relevant for our purpose of analysing social categories, the number of problems with reconstruction shrinks significantly. This said, for the most part, my own relevant reconstructions converge with the results of the Critical Edition of Q.6 Thus, for reasons of space, I will refrain

3.  A helpful overview on group processes, and among them social categorization, is provided by Rupert Brown, Group Processes:  Dynamics Within and Between Groups (Oxford: Blackwell, 2nd edn, 2000). 4. Nicholas Perrin, ‘The Limits of a Reconstructed Q’, in Questioning Q: A Multidimensional Critique, ed. Mark S. Goodacre and Nicholas Perrin (London:  SPCK, 2004), 71–88, 79. The problem is also brought to the surface e. g. by Michael Wolter, ‘Reconstructing Q?’, Expository Times (ExpTim) 115 (2003): 115–19, 115–17. 5.  Checking the representation of the social categories of Markan Jesus speech in Matthew  and Luke, I  could not find any significant minor agreements (cf. Scherer, Königsvolk, 102). In analogy, I take it as probable that Matthew and Luke have hardly any minor agreements against a would-be double-tradition source. 6.  The Critical Edition of Q: Synopsis Including the Gospels of Matthew and Luke, Mark and Thomas with English, German and French Translations of Q and Thomas, ed. James M.  Robinson, Paul Hoffmann and John S.  Kloppenborg, Hermeneia (Minneapolis, MN: Fortress Press, and Leuven: Peeters, 2000).

10. Coherence and Distinctness

187

from presenting the texts and commenting on the reconstruction, as the critical minimum can easily be looked up in any synopsis.7 One further point: Most probably our double tradition is only a fragment of what may have been its source. In any case, despite being incomplete, a fragment can be telling as to its texture. If in this selection structures can be found, they function like nodal points of a matrix. If, as it is most probable for any ancient text, the text was coherent, the full text does not turn this matrix into its opposite. Rather, the full text would have followed the same matrix. Anyway, the insights gained from the fragment are in some way representative of the complete text – as can be seen from the Dead Sea Scrolls of which we use fragments to reconstruct social or theological structures of whole documents.

10.2 Exemplary Analysis: Household Roles After these preliminaries, I would like to introduce one exemplary analysis of the double tradition’s social matrix, which deals with the social categories related to the ancient household. The roles played in ancient households are well known. In patriarchal societies, Greco-Roman and Jewish alike, we find one single male figure at its head, the pater familias. As far as the basic structure is concerned,8 he is husband to his wife, father to his children and master to his slaves,9 the latter institution being some rather undisputed part of ancient households. Formally, the

7. Whenever I mention the would-be double-tradition Vorlage, I make use of the abbreviation Q plus Lukan verse numbers only for convention’s sake. 8. What is missing in this model is the category of siblings: According to the ideal, any married male leads a household of his own – siblings only appear in the role of children under the power of the pater familias. While they are young, they are of no particular interest to the familial discourse. As adults, siblings are expected to show solidarity towards each other. Of course, reality may be different: The oldest son is expected to inherit his father’s house – and to live under his tutelage even when married, cf. Q 14.26. Moreover, divorced women are expected to return to their father’s house which, after his death, is the brother’s house, cf. Lev 22.13; Judg 19.2 9.  e. g. Aristotle, Politics (Pol.) 1253b; Seneca, Epistulae (Ep.) 94.1; Philo, Hypothetica (Hypoth.) 7.14; Josephus, Contra  Apionem (C. Ap.) II 190–210; Pseudo-Phocylides (Ps.-Phoc.) 175–227; moreover the economics literature and New Testament household codes operate on that model, cf. Martin Ebner, Die Stadt als Lebensraum der ersten Christen:  Das Urchristentum in seiner Umwelt I, Grundrisse zum Neuen Testament (GNT) 1.1 (Göttingen:  Vandenhoeck & Ruprecht, 2012), 173–5; Marlis Gielen, Tradition und Theologie neutestamentlicher Haustafelethik:  Ein Beitrag zur Frage einer christlichen Auseinandersetzung mit gesellschaftlichen Normen, Athenäums Monografien. Theologie/ Bonner Biblische Beiträge 75 (Frankfurt a. M.: Hain, 1990), 55–62, 66–7. From Palestinian wisdom tradition, the family models represented in Sirach or 4QInstructions might be added; and prominently: Deut 5.14/Exod 20.10.

188

Hildegard Scherer

pater familias has power over the household, but also responsibilities. For example, he is the owner of all goods but, at the same time, it is his duty to provide for the livelihood of all of his subordinates.10 In ancient times, and especially in agricultural societies, belonging to a household is existential. Households provide the houses – as spaces necessary for the protection of life and goods and as spaces marking out social belonging. Playing a household role grants access to land or further possessions and moreover grants access to food – which is to be produced by the household’s division of labour.11 10.2.1 Rupture Between Generations – and a New House of the ‘Father’ In the double tradition, we find some discourse on household roles. In three texts of the double tradition relevant to our topic, existential ruptures between the younger and the older generations of a household appear.12 In Q 9.59–60, the son is pulled away from familial norms to the authority of Jesus. His (implied) wish to follow Jesus collides with his obligation to be present at home for his father’s burial. To me, his request seems to make most sense if his father was still alive and if he asked to stay with him until his death.13 Jesus’ answer is open to some interpretation: Is he qualifying all immobile family members as ‘spiritually dead’– or is he ironically relativizing the importance of any action

10.  e. g. Isa 4.1; Ps.-Phoc. 153–174; Sir 33.20–22, 24; 25.22; T. Job 21–25; P.Yadin 10 l. 9– 10, 14; P.Yadin 19 l. 15–16.; P.Yadin 37 l. 9. 11.  The staple food of the day is bread, cf. e. g. Gildas Hamel, Poverty and Charity in Roman Palestine:  First Three Centuries C.E., University of California Publications:  Near Eastern Studies 23 (Berkeley, CA:  University of California Press, 1990), 22. It is a cultural product which is cooperatively produced in agricultural household units: In his analysis of baking facilities in Massada, Jerusalem and Jericho, no traces of central food production appear according to Ronny Reich, ‘Baking and Cooking at Masada’, Zeitschrift des Deutschen Palästina-Vereins (ZDPV) 119 (2003): 140–58, 152, 155. That provision of bread is precarious outside households or on journeys can be demonstrated from e. g. 1 Kgs 13.15; Gen 28.20–21.; Exod 16.35. 12.  Of course, the compilation of the three texts is far from innovative: cf. e. g. Christopher Tuckett, ‘Q and Family Ties’, in Metaphor, Narrative, and Parables in Q, ed. Dieter T. Roth, Ruben Zimmermann and Michael Labahn, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 315 (Tübingen: Mohr Siebeck, 2014), 57–71; Arland D. Jacobson, ‘Jesus against the Family: The Dissolution of Family Ties in the Gospel Tradition’, in From Quest to Q. FS James M. Robinson, ed. Jon Ma. Asgeirsson, Kristin de Troyer and Marvin W. Meyer, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 146 (Leuven:  Leuven University Press and Peeters, 2000), 189–218, 190–202. 13.  Cf. R. T. France, The Gospel of Matthew, The New International Greek Testament Commentary (NIGTC) (Grand Rapids, MI: Eerdmanns, 2007), 329.

10. Coherence and Distinctness

189

towards those who belong to the realm of the dead?14 In any case, he strips them of any power: Dead persons can no longer act. Apart from that, Jesus’ pragmatic message is unambiguous: there is no way back home. ‘Following’ Jesus means moving in the opposite direction. Moreover, it means sharing his lifestyle, as described in v. 57–58: homelessness with no safe place for one’s belongings and with no ties to any familial household whatsoever. It means taking leave. And it is the young who actively break their familial loyalties. Q 12.51 and 53, with its ἦλθον formula, summarizes Jesus’ task – and tries to legitimize the harshness experienced by his followers,15 which is not naturally expected from a mandatary of the God of Israel. The metaphorical sword of v. 51 cuts off the younger from the elder generation or, to be more precise, a young married couple (‘son’ and ‘bride’) from the parents of the husband, and moreover a single daughter from her mother, indicating the gendered spheres of the households. The subtext of v. 53 is found in Mic 7.6 LXX. The tradition fits into situations where houses had been woefully split because of the Jesus movement. Those concerned – the former sons, daughters, daughters-in-law and perhaps parents – try to make sense of their situation. It is explained to them as a necessity, in accord with biblical prophecy.16 In spite of all differences in wording, there is some common ground between the sayings of Matthew and Luke in Q 14.26: It contains preconditions for calling oneself related to Jesus. One of these conditions – and maybe the practical realization of ‘carrying one’s cross’ (v. 27)17 – means hating (one’s) parents and children. In addition to the two aforementioned texts, we find some initiative shown by parents here. In its biblical sense, ‘love’ and ‘hate’ are not firsthand emotional moves: ‘to love’ means to practice loyalty towards somebody; its opposite, ‘to hate’, means to give up loyal behaviour,18 which will automatically shake the foundations of any household community. The rupture is at hand. But apart from these three examples of household dissolutions, we find another thread in the texture of the double tradition which seems to provide an answer to the existential loss of household protection: In the double tradition, God appears

14.  Whereas the first option is held by many commentators, some think of secondary burial, e. g. Byron M. McCane, Roll Back the Stone: Death and Burial in the World of Jesus (Harrisburg, PA: Trinity Press International, 2003), 73–7. 15. As for the legitimatory character, cf. e.g. Jacobson, ‘Jesus against the Family’, 193; Tuckett, ‘Q and Family Ties’, 63–7, interprets Q 12,51.53, but also Q 9.57–60 and 14.26 as legitimizing retrospectives rather than appeals for the present. 16.  Mic 7.6 is adapted to this situation. It is the only verse in the Micah passage which refers to the future. Stripped of its context, it can be understood as prophecy. 17. Cf. W. D. Davies and Dale C. Allison, Jr., The Gospel according to Saint Matthew, vol. II: Matthew VIII-XVIII, International Critical Commentary (ICC) (London:  T&T Clark, 1991), 221; Tuckett, ‘Q and Family Ties’, 67, is reluctant to full identification. 18.  Cf. Joel B. Green, The Gospel of Luke, New International Commentary on the New Testament (NICNT) (Grand Rapids, MI:  Eerdmans, 1997), 565; Tuckett, ‘Q and Family Ties’, 67.

190

Hildegard Scherer

as the heavenly ‘father’ integrating his ‘sons’ into a new virtual household.19 Astonishingly, from among the many rights and duties of an ancient pater familias, the double tradition focuses four times on the father providing subsistence for his children – who in real life, when deciding for Jesus, have lost their base of living when splitting up from their native pater familias. The motive of the ‘foster father’ God seems to answer the problem of the socially outcast Jesus followers.20 In the Lord’s Prayer, Q 11.2–4, both Matthew and Luke contain an appeal to the heavenly father: ‘Give us this day our daily bread.’21 This urgent wish refers to more than just the goods of creation: It is not about sun and rain (cf. Q 6.35–36), but about the bread ready at hand. This is the outcome of an agricultural household’s production chain. The focus on household substitution is reinforced by another double tradition close to the Lord’s Prayer, Q 11.11–13.22 This Qal-wa-homer sets the heavenly father in an analogy to earthly fathers23 as ‘givers’ of good food or 19. An often made observation, cf. e. g. Halvor Moxnes, Putting Jesus in His Place. A Radical Vision of Household and Kingdom (Louisville, KY: Westminster John Knox, 2003), 114–21, with reference to the problem of family dissolution and food provision. Yet, Tuckett, ‘Q and Family Ties’, 69 n. 36, sees the problem of the family tensions and of reintegrating into a new social formation, but thinks the father motive to be too general. I think it becomes very specific when the role of the foster father is highlighted. 20.  Cf. also 1QH xvii 34–36 and Jos. Asen. 11–12 for family outcasts who pray to God as father to provide for them. As for the background of God as father in early Judaism, cf. Angelika Strotmann, ‘Mein Vater bist du!’ (Sir 51,10): Zur Bedeutung der Vaterschaft Gottes in kanonischen und nichtkanonischen frühjüdischen Schriften, Frankfurter Theologische Studien 39 (Frankfurt a. M.: Knecht, 1991), who collects and comments on all relevant texts. 21. The most apt interpretation of ἐπιούσιον seems to me ‘for the following day’, said in the morning and relating to the day which has just started, cf. France, Matthew, 248. Morning is the time proper for preparing bread in the households. For a short survey of the different options cf. e. g. Michael Wolter, Das Lukasevangelium, Handbuch zum Neuen Testament (HNT) 5 (Tübingen: Mohr Siebeck, 2008), 407–8, who also opts for a realistic interpretation as daily food portion. Social realties confirm this interpretation: Flour of hand mills had to be consumed within a short time, ‘as it contained the oily wheat germ, which tends to go rancid’, Dalit Thon, ‘From Seed to Loaf ’, in Bread: Daily and Divine, ed. No’am Ben-Yossef (Jerusalem: The Israel Museum, 2006), 13–31, 24; as the production of 1 kilogram flour in a hand mill takes one hour (cf. Shukri ‘Arraf, ‘From the Field to the Table: Bread in Palestinian Culture and Tradition’, in Bread: Daily and Divine, ed. No‘am Ben-Yossef [Jerusalem: The Israel Museum, 2006], 191–215, 202), daily bread production gets the more plausible . 22.  In between, we find the appeal to search, ask  – and knock. While for the pair of searching and asking there are biblical traditions, as far as I can see we do not find any traditional precedent to ‘knock’ at God’s door. To me, this special move seems to have grown out of a situation of people out of houses. 23. The ‘you’ group need not refer specifically to Jesus followers who were addressed as heads of households – with v. 13a in mind, the ‘you’ addresses human beings in general as opposite to God.

10. Coherence and Distinctness

191

goods on a more abstract level. Earthly fathers play out the role of heads of households who are in charge of the household goods.24 Concerning the pragmatics of this little sequence,25 some development seems to have taken place as compared to the request for bread in Q 11.3. The text states in the abstract that everything the father gives on request is something ‘good’ – as fathers naturally give ‘good’ things to their children. This persuasion seems necessary wherever prayers do not lead to any expected outcome. It appeals to God’s foresight, in the sense that he gives what in the long run will turn out to be a ‘good’ option. Significantly, this reflection originates from the role of the father who provides nurture for his children. Furthermore, in the double-tradition temptation story, Satan in Q 4.3f. first appeals to Jesus as the son who gains bread by the power of his father. As v. 4 indicates, this is not a coincidental miracle – it is about survival. In my opinion, the temptations present Jesus as a paradigmatic ‘son of God’ in its inclusive sense, as are the ‘sons of God’ in Q 6.35–36. The strongest evidence is given by Satan’s address to an anarthrous son: If a title were meant, there would have been an article, but no copula between ‘son’ and ‘of God’ (εἰ υἱὸς εἶ τοῦ θεοῦ), disturbing the word order of any title. There is another tradition which brings together desert, children of God, food and the good father God: Wis 16.20f., which builds on Exod 16 and Ps 77.24–2526 LXX. It makes the devil’s request fully plausible – and fits Jesus’ role as an exemplary Israelite. But apart from abstraction and biblical implications, we find the basic motive once more in the double-tradition material in Q 12.29–31. The text has its own history of tradition.27 It seems to have been reworked already on the level of any double-tradition Vorlage. But it still reflects problems with fulfilling roles in household work division. The promise of v. 31b – food and clothing will be provided – is once more based on the notion that there is a superior ‘father’ (v. 30b) who is at the same time king and who feels responsible for his children.28

24.  In its shortness, v. 11 draws an ideal picture to fit most properly with the heavenly analogy: In real life, there most probably would be some mother responsible for bread at least (cf. Gen 27.17; 1 Kgs 17.11; Prov 31.15); but there are male bread givers (cf. Gen 25.34). They appear when bread is given outside the household (Judg 8.4–6; Tob 1.17; Luke 11.5–8). 25. Cf. e. g. France, Matthew, 279; Sarah E. Rollens, Framing Social Criticism in the Jesus Movement: The Ideological Project in the Sayings Gospel Q, WUNT II/374 (Tübingen: Mohr Siebeck, 2014), 151. 26.  Cf. W. D. Davies and Dale C. Allison, Jr., The Gospel according to Saint Matthew, vol.  I:  Introduction and Commentary on Matthew I-VII, ICC (London:  T&T Clark, 1988), 352. 27.  Cf. Martin Ebner, Jesus  – ein Weisheitslehrer? Synoptische Weisheitslogien im Traditionsprozess, Herders Biblische Studien (HBS) 15 (Freiburg i.  Br.:  Herder, 1998), 250–75. 28. The relation to paternal obligation is also noted in Wolter, Luke, 456.

192

Hildegard Scherer

The idea of solitary outcasts being integrated into and nurtured by a superior household of the heavenly father sounds appealing – but what about reality? How does this integration work in practice? The passive προστεθήσεται in 12.31b is revealing:  It can be read as a passivum sociale.29 As some connection to earthly households is essential for survival, there must have been earthly households providing shelter for those leaving their former homes for the Gospel’s sake. There must have been households of sympathizers or even group members who were saved from the aforementioned existential ruptures.30 The double traditions mirror these household followers too, the most telling example being the slave of Q 12.42–46. This slave parable has some specific accent: Why does it highlight the slave who ‘gives food in time’?31 Food distribution seems to be the pragmatic background of the whole scenario (cf. v. 45d). In real life, we know that the pater familias as the owner of the household goods is responsible for distribution. By this parable, the real-life master of the household is written into the role of a slave in God’s superior household. He is no longer master of the goods, but obedient agent, he is slave among his co-slaves, which means everybody in God’s service, familial household member or not. If the person addressed fulfils his role as slave among slaves, he will stand his eschatological home trial – but if he behaves as patres familias usually do – exercising their power and dining with friends32 – he will not. If the heads of working households act out their responsibility for their fellow group members, the problem of supply in the God’s new household is successfully solved.33 So far, exemplary nodal points of the double tradition’s social matrix have appeared: By starting out in several places from the problem of family ruptures, there seems to be a corresponding notion of a superior divine father who substitutes the earthly one, especially in his role of supplying goods for living. The role 29. That the promise of provision is realized on a social level is also seen by e. g. Richard Valantasis, The New Q:  A Fresh Translation with Commentary (London:  T&T Clark, 2005), 165. 30.  Cf. Q 16.18: The implicit obligation to keep marriages going fits with the need for Christian households. 31.  The text only presupposes very basic knowledge of the institution of slavery. The specific task to provide food seems over-narrated, as actually, the responsible slave has more things to do than that; the specific action of food provision is no suitable summary of his task. 32.  Cf. Christine Gerber, ‘Es ist stets höchste Zeit (Vom treuen und untreuen Haushalter):  Q 12.42–46 (Mt 24.45–51 // Lk 12.42–46)’, in Kompendium der Gleichnisse Jesu, ed. Ruben Zimmermann (Gütersloh:  Gütersloher Verlagshaus, 2007), 161–70, 163: assumptions of status privileges of the master. 33.  Cf. a similar conclusion by e.  g. Harry T. Fleddermann, Q:  A Reconstruction and Commentary, Biblical Tools and Studies (BTS) 1 (Leuven:  Peeters, 2005), 635; Melanie Johnson-DeBaufre, Jesus among Her Children:  Q, Eschatology, and the Construction of Christian Origins, Harvard Theological Studies (HTS) 55 (Cambridge, MA:  Harvard University Press, 2005), 103.

10. Coherence and Distinctness

193

of the divine pater familias can be exploited for variation: Being not only the father of children, but also the master of slaves, he has the power to establish norms in his household. In spite of all the modifications and sophistications the texts might have undergone, they seem to be variations of the same motif. The motif appears several times, and the texts interact with some logic. This seems to be more than coincidence. But what about these motifs – family rupture, heavenly breadwinner – in the corpora of Mark and Matthew/Luke Sondergut? As far as I can see, despite some continuity, there are characteristic differences in Mark’s treatment of family rupture:34 Mark 3.21 reconstructs the conflict between Jesus and his family. Indeed, it is Jesus who has gone away and is now to be taken home by his family. And indeed, Jesus distances himself and talks about his new family headed by God. At first glance, we find our motifs in a nutshell – but there are three significant differences: The story is about Jesus’ family rupture, not about his followers; on a pragmatic level, it deals with gaining the status of intimately belonging to Jesus, not with supplying what is essential for one’s existence; and finally, the heavenly father is not fixed in his role as a provider of food, but as norm-giving leader (3.35). Mark 10.29–30 varies on the motif of leaving one’s household and gaining substitutes in this world, but the father is mentioned only implicitly (i.e. not at all). Finally, Mark does not depict family ruptures as brutal as the double tradition: In Mark 10.29, the house is simply ‘left’. Jesus’ call of his first followers in Mark is irenic: They follow him ‘at once’, but Simon and Andrew are not cut off from their families:  Afterwards, Jesus visits the house of Peter and his mother-in-law who ‘serves’ them (Mark 1.18, 20, 29–31). More significantly, it is Jesus himself who in Mark twice appeals to the fourth commandment, explicitly impressing upon his audience the duty of caring for one’s parents in their old age (Mark 7.10–12; 10.19). Of course there is eschatological family rupture in Mark 13.12, but this conflict is not initiated by the followers of Jesus, but by the non-believing family members; moreover, it leads straight into capital trial, not into being homeless. Beyond Mark, Matt 10.34–39 has read Q 12.51, 53 and Q 14.26 under the headline of Mark 13.12/Matt 10.21, as illustrating the eschatological familial tensions leading into trials. By implication, Matthew’s sword turns into a sword of capital punishment. By contrast, Luke seems to have kept to the more original interpretation in which the Micah quotation makes full sense:  The saying is not about cutting the heads of family members, but cutting the ties between them. If Luke had only known Matthew, would he have isolated the saying and would he have interpreted it the way he does? The cross saying in Matt 10.37 seems more apt, but whereas before, the hostile family members cancel their loyalty towards the addressees, Matt 10.37 is about the loyalty of addressees. In Luke, this notion is clearly fleshed out when the saying

34. A pioneer in the comparison of materials: Jacobson, ‘Jesus and the Family’, 190–210, unfortunately only compares the family ties topic. Cf. also Jacobson, ‘The Literary Unity of Q’, Journal of Biblical Literature (JBL) 101 (1982): 365–89.

194

Hildegard Scherer

appears in a parenetic sequence culminating in giving up one’s possessions. If the little sequence had flowed from the pen of Matthew, would it bear all those tensions? In my opinion, it is best explained by preceding traditions, which Matthew has subsumed under ‘eschatological trials’. As Luke treats the material rather differently and more to the sense of the words, I think it rather probable that he has had independent access to it. As regards family rupture, Matthew shows some coincidentia oppositorum: He fits the apophthegms concerning the Jesus followers rather clumsily into the miracle chapter Matt 8 – and creates some tension between the son of man having no shelter and Jesus’ lodgings in Capernaum (Matt 4.13). In the same way, Matt names Jesus’ harsh saying about the dead burying their dead – but nevertheless has his Jesus speak the Markan critique of neglecting the fourth commandment (Matt 15.3–6). Matthew seems to tone down the family rupture/homeless followers topic, but he does not manage to cover his tracks which appear more clearly when isolating the double tradition. As concerns the new household and the ‘father-master’, both Matthew and Luke pen their versions of the slave parable Q 12.42–46. For Matthew, it is part of a sequence of eschatological parables; for Luke, it is about vigilance, curtailed to Peter. But neither Matthew nor Luke are concerned with the singular phenomenon of a food-providing slave. Of course, with some metaphorical moves, one can make sense of this narrative element even in their context. But would they have created a parable like this? In Matthew, the tensions are especially strong: His eschatological parables address all Jesus followers alike – but this parable is about some extraordinary slave responsible for others. Luke seems to have been aware of this and refers it to Peter as the ‘functionary’. But would he have done so if he only had known Matthew – and not the sequence of three slave parables in the double tradition? Without a doubt, as concerns the motives of family rupture and God’s new household, we could argue about the evaluation of coherences in spite of all the differences. What, by contrast, seems to me much more distinctive is the motif of the bread-giving, food-providing father role. Turning up four times in the double tradition,35 it is found neither in Mark nor in Matthew and Luke beyond the double tradition.36 10.2.2 Slaves in the Household of God: Models for (Particular) Followers Let me turn back to the aforementioned slave parable: It has three more points of connection in double-tradition material. The metaphorical model underlying Q 35. Not to mention the father–son relationship in Q 6.35–36, which has similar undertones: It appeals to distribution of goods. Anyway, it puts the father in the role of creator, not of pater familias. 36.  Despite the introduction in Matt 6.8, the Lord’s Prayer of Matthew is focussed on forgiving sins (Matt 6.14–15 par Mark 11.25) – the food element is superfluous to him.

10. Coherence and Distinctness

195

12.42–46 and the other slave parables of the double traditions manifests itself in Q 16.13: Master is the Lord, the addressees are his slaves, and the behaviour expected of the addressees is to serve their master loyally. In Q 19.12–26, the so-called parable of the talents, we find another slave parable which shows affinities with Q 12.46–47.37 There is also a master leaving his household and giving the slaves responsibilities; there are also ‘faithful’ slaves fulfilling their duties; as a reward they are also ‘set above’ something. At first glance, Matthew and Luke do not have very much in common apart from the basics of the plot. But in reading Matthew 25.14–30 closely, we find logical tensions in the text which indicate that he must have altered some kind of Vorlage:  (1) In Matthew, the slaves are given several talents, ‘the belongings’ of their master. But when the master comes back, he qualifies the sums of money as ‘little’.38 (2) In Matthew, the faithful slaves receive two different rewards:39 They are ‘set above many things’ (which means more work) and at the same time invited into the ‘joy’ of their master (which means leisure). Whereas this second reward meets the ends of Matthew’s eschatological spacing, the unfaithful slave being removed outside, the first reward is superfluous for Matthew. But at the same time, it links this text to the second slave parable of Q 12.42–46, where the faithful slave is also rewarded by being ‘set above many things’. (3)  There is something wrong with Matthew’s notion of distributing the talents ‘according to one’s power’: Indeed, the slaves are trusted with different sums of money, but the faithful slaves gain the same profit: both of them return 200 per cent of the sum invested. Relative to each other, they show the same amount of ‘power’.40 But why, then, does the master order the talent of the third slave to be given to slave no. 1? What is he actually rewarded for? (4) v. 29 misses the topic of the parable. The parable revolves around the problem of gaining or not gaining while the sentence in v. 29 talks about having or not having.41 And by the way, (5) where do the addressees of v. 28 and v. 30 come from?

37.  The link between the two parables is often noted, e.g. Gerber, ‘Es ist stets höchste Zeit’, 167. 38.  Cf. Adelbert Denaux, ‘The Parable of the Talents/Pounds (Q 19,12– 27): A Reconstruction of the Q Text’, in The Sayings Source Q and the Historical Jesus, ed. Andreas Lindemann, BETL 158 (Leuven:  Leuven University Press and Peeters, 2001), 429–60, 437. 39.  The problem is also seen by e.g. Denaux, ‘The Parable of the Talents’, 448; Fleddermann, Q, 848. 40. Cf. Denaux, ‘The Parable of the Talents’, 437–8. 41.  Cf. Christoph Heil, ‘Was erzählt die Parabel vom anvertrauten Geld? Soziohistorische und theologische Aspekte von Q 19,12–26’, in Metaphor, Narrative, and Parables in Q, ed. Roth, Zimmermann and Labahn, 339–70, 355.

196

Hildegard Scherer

As regards these observations, the Lukan text, far from lacking problems,42 is more consistent: The slaves are given one mina (often rendered pound) – a sum that in our context can really be called ‘little’ – and they gain tenfold or fivefold profit. The slave with the 10 (or: 11) mina is rightly being rewarded, as he was in fact the most productive. How come does Luke have more logic in his plot? Has he ameliorated the redactional mistakes of Matthew – or has he simply had independent access to the Vorlage of the text which was not yet flawed by Matthew’s inconsistencies?43 Of course, one single example of the so-called alternating primitivity

42.  It seems hyperbolic to be granted the power over ten or five cities for a gain of 9 or 4 mina (v. 16.18); the number of ten slaves at the beginning seems to be redactional, as only three slaves give their accounts (v. 13); the Lukan slave of v. 24 should have eleven, not ten mina (cf. e. g. Fleddermann, Q, 840f.); the answer of the slave in v. 20f. bears tensions: Why does he hide the money despite being given an explicit order to deal and despite the fact he was afraid of his master? But there are breaks already on the level of any assumed Vorlage of the double tradition. v. 25 is doubly attested, but does not fit anyway, as noted above. As for Lukan redaction, many agree that Luke introduced the topic of gaining the title of a king (v. 12a.14.27). 43.  Moreover, Luke develops the text into a Beispielgeschichte, where the narration and its application stem from the same sphere – in our case, the social sphere. One paradigmatic incident is narrated to illustrate the ways of the (social) world – one paradigmatic role model is given for behaviour in this same social sphere. For Luke, the text demonstrates how worldly kings or masters abuse their power and are oriented towards profit. Whereas two slaves play their parts in this game, the Lukan hero is the third slave who peacefully resists the system. The addressees are meant to follow him. To my mind – and against Ruben Zimmermann, ‘Parabeln – sonst nichts! Gattungsbestimmung jenseits der Klassifikation in “Bildwort”, “Gleichnis”, “Parabel” und “Beispielerzählung” ’, in Hermeneutik der Gleichnisse Jesu:  Methodische Neuansätze zum Verstehen urchristlicher Parabeltexte, ed. Ruben Zimmermann, WUNT 231 (Tübingen:  Mohr Siebeck, 2008), 383–419  – parables in their proper sense function differently: Narration and application stem from different spheres, the one being theological, the other social. The social incident provides analogy for theological realities – as it is the case in Matthew’s eschatological version (and already in our assumed Vorlage): When the Lord is God, the first two slaves are meant to be the positive examples, the third one is the bad guy. Matthew has fleshed out this idea of eschatological reward and punishment and has put his ‘parable’ in a row with other texts of this kind. I wonder whether Luke would have been able to detect the Beispielgeschichte potential of the text if he had only known Matthew’s version. To me, it seems more probable that he knew a shorter and still more ambiguous version, as it is represented by the common core of Matthew and Luke. As concerns the double tradition Vorlage, I agree with Heil, ‘Was erzählt die Parabel’, 362–70 (cf. there for secondary literature on any positions); but as concerns tradition history, it seems to me more plausible to categorize the stage before Q as well as Luke as Beispielgeschichte. Pro Beispielgeschichte, cf. e. g. Martin Ebner, ‘Face to FaceWiderstand im Sinn der Gottesherrschaft. Jesu Wahrnehmung seines sozialen Umfeldes im Spiegel seiner Beispielgeschichten’, Early Christianity 1 (2010):  406–40, 418–23; some would-be metaphorical elements of the text are extremely difficult (421–2).

10. Coherence and Distinctness

197

cannot prove anything. Later on, we will have to consider this phenomenon in cumulo. But let us now come back to the pragmatics of this second slave parable, as we reconstruct its common core in the double tradition: It appeals to would-be ‘slaves’ in order to put their ‘money’ in circulation and to gain profit from it. That sounds like missionary work: What God has entrusted to those obedient to him is to be given to others, and, even if the image does not fit 100 per cent, the gain will be new group members. As not all slaves in the household of God are ‘financial experts’, it seems that a particular group of active missionaries is at the centre of this parable.44 I will make only some short remarks about one last slave parable of the double tradition, Q 14.16–23. This parable is about legitimizing unexpected invitations to God’s dinner. I do not dare reconstruct the excuses of the persons originally invited  – there is too little common ground and too much redactional interest to say something valid. Anyway, the dinner seems to take place at an unusual point in time: As ordinary dinners start in the evening, after the work is done,45 this dinner collides with the daily duties of the would-be guests. For this reason, the slave has to invite others who have no established relationship with the host. Unconventional as it is, this kind of invitation needs legitimization:46 Those who collect the guests are written into the role of the faithful slave who is nothing but obedient to his master. Any responsibility for the guest list lies with him. Four slave parables, three of them explicitly parenetic, two of them structural siblings, three of them role models for particular positions in Christian groups – do the other corpora show similar nodal points? The Markan tradition knows only one little remnant of parenetic slave parables in Mark 13.34–36. As far as I can see, it does not carry any particular weight for Mark. There are some more slaves in the vineyard parable Mark 12.1–11, but without any parenetic impetus. The slaves are the prophets of old, the story is culminating in the death of the ‘son’  – any would-be followers and their functions are completely out of sight of this parable. Whereas in the double-tradition material, the slave–master metaphor is much more fleshed out, Mark transposes the parenetic impulse of ‘being a slave’:  The Markan Jesus requires his followers to act as slaves or – more frequently – servants towards each other (Mark 9.35; 10.42–45). As regards Matthew’s Sondergut, we find two more parables structurally cohering with our double-tradition material. In Matt 13.24–30, there are slavefunctionaries who wish to eliminate the weeds from their master’s field; in Matt 18.23–34, we find a slave in a position to cancel the debts of his co-slave. The

44.  Cf. also Jean-Marc Babut, Un tout autre christianisme: Traduction nouvelle et commentaire de la Source Q, (Paris: Desclée de Brouwer, 2010), 261. 45.  Cf. e.g. Ebner, Stadt, 180; Graeco-Roman practices are plausible at least for Palestinian elites. 46. As for the legitimatory aspect, cf. Jacques Dupont, ‘La parabole des invités au festin dans le ministère de Jésus’, in Jacques Dupont, Études sur les évangiles synoptiques, vol. 2, BETL 70b (Leuven: Leuven University Press and Peeters, 1985), 667–705, 690–1, 704.

198

Hildegard Scherer

redactional interests behind these parables are obvious  – did the double tradition inspire Matthew to create them? If there are no clear hints as for a Matthean reworking of tradition, were they part of a double-tradition Vorlage? At the moment, I do not feel ready to prove the one or the other option.47 And here the limitations of my method turn up: By analysing the social matrix, we will never be able to enclose any precise extent of a double-tradition Vorlage. There are Sondergut texts which might fit; at the same time there are double-tradition texts lacking any social nodal points. But one last glance at the slaves in Luke’s Sondergut: Beyond the double tradition, he has shown little interest in slaves as role models:48 His ‘householder’ in Luke 16.1–7 is clearly free, as he can be fired. As concerns the double tradition, Luke tries to portray his slaves in a better light, writing about a householder in Q 12.42 or putting them in a Beispielgeschichte. In Luke 17.7–10, there is one more slave parable about submissiveness in the service of the Lord. Its introduction is as revealing as it is awkward in the synoptic tradition: The metaphor is addressed to real slave owners who in real life are served by their slaves. This fits the ways of the world – the critical force of Q 12.42–46, which has the real pater familias behave like a slave among co-slaves, has vanished.

10.3 Conclusion 10.3.1 Coherence and Distinctness By the example of familial roles, I have tried to show that the social matrix of the double tradition makes some sense of its own. When isolating the double-tradition texts, we find some interplay of social concepts which is more than coincidence and which shows reflections on urgent problems of a group of Jesus followers. I have focused here on familial roles, but in the social texture of the double tradition we can find further threads. For example, we find an intensive discourse on prophets: There is a double-tradition Jesus who grapples for acceptance (e.g. Q 7.34; 11.14–22) – and who is to be legitimized as a messenger of God on the one hand by positive comparison to John the established prophet and on the other hand by the penetrating motive of the killing of the prophets. Or we can find some discourse on genealogy, on who is son of the patriarchs or of God (e.g. Q 3.8; 13.28; 6.35–36). These discourses are clearly marked off from the Markan traditions, which I have tried to show by the example of household roles. As for the other examples, Mark neither cares about patriarchs nor is he consistent about prophet killings – he can even state Mark 6.4 as if the only problem of a prophet was his hometown.

47.  At this point, there is little impetus behind the question why Luke should have left these texts out if he knew them. There might be texts where this question is more pressing. 48. Luke 12.37, 47–48 seem to be knitted out of Luke 12.42–46.

10. Coherence and Distinctness

199

Moreover, I  think there is ample evidence that Matthew, when writing his Gospel, has had access to foregoing traditions: traditions which would have looked differently if Matthew had composed them and which still bear the marks of their own tradition histories. If these traditions show some common matrix and coherence, there is reason for exegetical interest: The more they cohere, the more they attest to some stream of tradition parallel to Mark. 10.3.2 Luke’s Access to the Traditions There is one question left, which is in my opinion to be treated separately from the question of traditions foregoing Matthew: the question of Luke’s access to traditions apart from Mark. Two options are hotly debated: Has he had access via Matthew  – or independently of him? From the perspective of double-tradition investigations following Q, this question is not too relevant for large parts of the material, for in many cases Matthew’s wording seems to be closer to the Vorlage anyway. And even without comparing the material with Luke, the profile of his foregoing tradition has not completely vanished despite his redactional moves. But in the course of analysis, we stumble again and again across the phenomenon of alternating primitivity.49 As regards sequence and wording, Luke sometimes seems to be more authentic to his tradition. It is more difficult to explain how Luke had derived Luke 12.51, 53; 14.26 or Luke 19,12–27 out of Matthew than to explain how Matthew had derived those out of Luke’s wording or Vorlage. Of course, the sparse examples which have turned up in my analysis are not enough for substantial evidence: We can make the point only cumulatively, taking into regard, e.g. Matt 23 as compared to Luke 11.39–52, especially Luke 11.51; Luke 13.18–19; the painstaking cluster Luke 16.16–18; the sign of Jona Q 11.30 where Luke deteriorates Matthew by skipping a plausible explanation for an enigmatic saying, or the awkward second eschatological discourse Luke 17.22–37.50 10.3.3 Prospects I have concentrated on the social matrix of the double-tradition material  – of course, my reflections on Mark, Matthew and Luke are far from complete. I have singled out what I  regarded as helpful comparisons, but of course, much more could be said about their redactional moves. Moreover, I am sure that social categories are not the only parameters useful for probing the double tradition. To me

49.  Alternating primitivity has become a staple argument pro Q, cf. e.g. John S. Kloppenborg, ‘On Dispensing with Q? Goodacre on the Relation of Luke and Matthew’, New Testament Studies (NTS) 49 (2003): 210–36, 224–5. 50.  When checking the arguments of reconstruction, alternating primitivity recommends itself in many cases. The examples listed are by no means exhaustive, but only chosen to be illustrative.

200

Hildegard Scherer

it seems also promising to compare, for example, the modes of referring to biblical texts: In a recent contribution, Marco Frenschkowski has proposed a model for the double-tradition texts as based on narrated knowledge.51 In contrast to that, at first sight, it is obvious that the Markan Jesus quotes the Scriptures more often and more explicitly. Perhaps new parameters and comparisons will help to substantiate our theories on the synoptic patchwork in a fresh way.

51.  Marco Frenschkowski, ‘Nicht wie die Schriftgelehrten:  Nichtschriftgelehrte Rezeption alttestamentlich-jüdischer Traditionen in der Logienquelle und im Koran’, in Metaphor, Narrative, and Parables in Q, ed. Roth, Zimmermann and Labahn, 221–52.

Chapter 11 T A K I N G OU R L E AV E O F M A R K – Q O V E R L A P S : M AJOR A GR E E M E NT S AND THE F A RRE R  T H E ORY Mark Goodacre

11.1 What Are the Mark-Q Overlaps? For students of the Synoptic Problem (hereafter SP), the ‘Mark–Q overlap’ passages represent a particularly fascinating set of data. These passages have a special place in the Two-Source Theory, occupying the space where the two major, independent sources of Matthew and Luke appear to ‘overlap’. In this chapter, I would like to argue that the Mark-Q overlaps draw attention to the weakness of the Two-Source Theory. These passages contradict one of the major arguments for the existence of Q and they compromise the elegant architecture of the Two-Source Theory. When they are taken seriously, the model loses its appeal. The Mark–Q overlap passages, and the difficulties they pose for the Two-Source Theory, are not widely understood.1 It will be helpful, therefore, to lay out which passages are involved, why the theory of ‘Mark–Q overlap’ is invoked, and why it is problematic. The basic pattern for much, but not all, of the triple-tradition 1. The focus of this chapter is the Two-Source Theory and the difficulties that the MarkQ overlaps pose for that theory. Other theories explain the Mark-Q overlaps in other ways, e.g. recently Delbert Burkett, Rethinking the Gospel Sources, Volume 2: The Unity and Plurality of Q, SBL Early Christianity and Its Literature (SBLECL) 1 (Atlanta:  Society of Biblical Literature, 2009), 6–7, and see the response in John C. Poirier, ‘Delbert Burkett’s Defense of Q’, in Marcan Priority Without Q:  Explorations in the Farrer Hypothesis, ed. John C. Poirier and Jeffrey Peterson, The Library of New Testament Studies (LNTS) 455 (London: Bloomsbury T&T Clark, 2015), 191–225 (195–9). For the Griesbach hypothesis, see especially William R. Farmer, The Synoptic Problem: A Critical Review of the Problem of Literary Relationships Between Matthew, Mark and Luke (New York:  Macmillan, 1964; 2nd edn, Dillsboro, NC:  Western North Carolina Press, 1976), esp. 142, and see the critique in Christopher M. Tuckett, The Revival of the Griesbach Hypothesis:  An Analysis and Appraisal, Society for New Testament Studies Monograph Series (SNTSMS) 44 (Cambridge:  Cambridge University Press, 1983), 76–93. For Matthean posteriority theories, see below.

202

Mark Goodacre

material is that there is a lot of agreement between Matthew and Mark alone, a lot of agreement between Mark and Luke alone, a lot of agreement between all three, but less agreement between Matthew and Luke. In other words, Mark is usually the middle term.2 Proponents of the Two-Source Theory and the Farrer Theory, both of which postulate the priority of Mark, explain this phenomenon by suggesting that Matthew and Luke relied on Mark for the triple tradition. So far the Two Source Theory and the Farrer Theory, both Marcan Priority theories, are united. However, there are several places where it is less clear that Mark is the middle term. These are the passages under discussion here, the so-called Mark–Q overlap passages, where there is substantial agreement between Matthew and Luke against Mark. For the Farrer theory, according to which Luke is familiar with Matthew as well as Mark, these passages present no difficulty.3 They are places where Luke is primarily dependent on Matthew, where he turns to Matthew’s account rather than Mark’s, generating much higher levels of agreement with Matthew than with Mark. For the Two-Source Theory, according to which Matthew and Luke are independent of one another,4 these passages appear problematic. How can Matthew and Luke be agreeing so extensively against Mark when they are dependent on Mark for their triple-tradition material? The Two-Source Theory solves the problem that these passages present by suggesting that the two sources of Matthew and Luke sometimes overlapped. On such occasions, Matthew and Luke turned to the hypothetical source Q instead of or as well as Mark, thereby generating major agreements between one another and

2. See Goodacre, Synoptic Problem, ch. 2, esp. 35–9, for an introduction to this material. This description is itself a massive oversimplification of the evidence. See further below. 3.  The theory is named after Austin Farrer, ‘On Dispensing with Q’, Studies in the Gospels: Essays in Memory of R. H. Lightfoot, ed. D. E. Nineham (Oxford: Blackwell, 1955), 55–88, but owes most to Michael Goulder. See especially Luke: A New Paradigm, Journal for the Study of the New Testament Supplement Series (JSNTSup) 20 (Sheffield:  Sheffield Academic Press, 1989). See also Mark Goodacre, The Case Against Q: Studies in Markan Priority and the Synoptic Problem (Philadelphia:  Trinity Press International, 2002); Mark Goodacre and Nicholas Perrin (eds), Questioning Q:  A Multidimensional Critique (London:  SPCK, 2004); Poirier and Peterson Marcan Priority, and Francis Watson, Gospel Writing:  A Canonical Perspective (Grand Rapids, MI:  Eerdmans, 2013), esp. 117–216. Watson prefers the term ‘L/M theory’ (Gospel Writing, 118–19), but ‘Farrer theory’ is now well established in the literature, and ‘L/M theory’ could equally well characterize the very different Griesbach (Two-Gospel) hypothesis, which also postulates Luke’s use of Matthew, but without Marcan Priority. On the hard-fought battle to distinguish the Farrer Theory from the Griesbach theory, and to stress the role played by Marcan Priority, see my Case Against Q, 9–14. 4.  For useful introductions to the Two-Source Theory, see Robert H. Stein, Studying the Synoptic Gospels: Origin and Interpretation (2nd edn; Grand Rapids, MI:  Baker, 2001) and John S. Kloppenborg, Q, the Earliest Gospel: An Introduction to the Original Stories and Sayings of Jesus (Louisville, KY: Westminster John Knox, 2008).

11. Taking our Leave of Mark–Q Overlaps

203

against Mark. Thus, the passages for which ‘Mark–Q overlap’ is invoked are tripletradition passages that feature substantial agreement between Matthew and Luke against Mark. B. H. Streeter explained the phenomenon in this way: There are several places where Matthew, Mark, and Luke are all three substantially parallel, but where the variations in detail and additions in which Matthew and Luke agree against Mark are so striking that it is clear they must have derived their versions in part, if not wholly, from some other source than Mark.5

Although there are many passages that have been assigned to Mark–Q overlap,6 Streeter himself limited the category to six key passages: (1) John the Baptist (Matt 3.1–12 // Mark 1.1–8 // Luke 3.1–18) (2) Baptism of Jesus (Matt 3.13–17 // Mark 1.9–11 // Luke 3.21–22)7 (3) Temptation (Matt 4.1–11 // Mark 1.12–13 // Luke 4.1–13)

5.  B. H. Streeter, ‘St. Mark’s Knowledge and Use of Q’, in Studies in the Synoptic Problem, ed. William Sanday (Oxford:  Clarendon, 1911), 165–83 (166). In this article, Streeter attributes the overlap passages to Mark’s use of Q. In The Four Gospels: A Study of Origins, Treating of the Manuscript Tradition, Sources, Authorship and Date (London:  Macmillan, 1924), Streeter abandoned this explanation. On this, see further Ed Parish Sanders, ‘Mark-Q Overlaps and the Synoptic Problem’, New Testament Studies (NTS) 19 (1973): 453–65. 6. See, e.g. the expansive list of 25 overlap texts in Rudolf Laufen Die Doppelüberlieferungen der Logienquelle und des Markusevangeliums (Königstein: Peter Hanstein, 1980), 91–2. But Laufen counts elements within pericopae separately, e.g. Mark 1.2 is treated separately from Mark 1.7–8, and the Beelzebub Controversy is divided into three separate texts, parallel to Mark 3.22–26, 3.27 and 3.28–9. See similarly Harry T. Fleddermann, Mark and Q. A Study of the Overlap Texts, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 122 (Leuven:  Leuven University Press, 1995), ix-xi and Q: A Reconstruction and Commentary, Biblical Tools and Studies (BTS)1 (Leuven: Peeters, 2005), 75–7, for a list of 29 overlap texts, and this still excludes parallels to Mark 1.1–6, 1.9–11, 1.12–13 and 12.28– 24. For F. C. Burkitt, see below, n. 46. 7.  Several scholars do not include the Baptism in Q, e.g. John S. Kloppenborg, The Formation of Q:  Trajectories in Ancient Wisdom Collections, Studies in Antiquity and Christianity (Philadelphia: Fortress, 2000), 84–5 and Excavating Q: The History and Setting of the Sayings Gospel (Minneapolis: Fortress, 2000), 93; Fledderman, Mark and Q, 16–18. The question of whether or not a given passage should be assigned to Q inevitably comes up repeatedly in a piece like this so my policy will be to focus on the six key passages isolated by Streeter in 1924. All six of these are treated as Q passages by the International Q Project, for which see especially James M. Robinson, Paul Hoffmann and John S. Kloppenborg Verbin; managing editor Milton Moreland, Critical Edition of Q in a Synopsis, Including the Gospels of Matthew and Luke, Mark and Thomas, with English, German and French Translations of Q and Thomas, Hermeneia (Philadelphia: Fortress, 2000).

204

Mark Goodacre

(4) Mission Discourse (Matt 10.5–15 // Mark 6.6b–13 // Luke 9.1–6 // Luke 10.1–12) (5) Beelzebub Controversy (Matt 12.25–32 // Mark 3.23–30 // Luke 11.17–23, 12.10) (6) Mustard Seed (Matt 13.31–32 // Mark 4.30–32 // Luke 13.18–19)8 In each case, the agreements between Matthew and Luke against Mark are substantial. There are agreements in order, where Matthew and Luke agree in placing the same double-tradition material in the same triple-tradition context, something that is not supposed to happen on the Two-Source Theory,9 and there are major agreements in wording. Of the six passages, the first (John the Baptist) and third (Temptation) are especially striking in that there is so much additional material in Matthew and Luke. Mark lacks parallels to Matt 3.7–10 // Luke 3.7–9 (‘Brood of vipers . . . ’), Matt 3.12 // Luke 3.17 (‘whose winnowing fork . . .’) and Matt 4.3–10 // Luke 4.3–12 (‘If you are the Son of God . . .’). Moreover, it is not simply a question of wedges of additional double-tradition material getting spliced into the triple-tradition material. Matthew and Luke sometimes feature the same alternative means of structuring material. This is particularly clear in the Preaching of John the Baptist, and no more so in the following section, where they have the same

8.  Streeter, Four Gospels, 305–6. His list here is more minimal than his earlier list (‘St. Mark’s knowledge’, 167–75), where he has these six passages minus the Baptism, but also has Mark 4.21–25 and parallels (171–2), Mark 9.42–50 and parallels (175–6) and Mark 12.38–40 and parallels (176). But see his qualifiers in Four Gospels, 306. Ed Parish Sanders and Margaret Davies (Studying the Synoptic Gospels [London:  SCM, 1989], 81) talk about ‘the principal five passages’ as the six above minus the Baptism. 9.  The agreement is so precise in the first three cases, John the Baptist, Jesus’ baptism and the Temptation story, that the standard articulation of the Two-Source Theory has to build these pericopae in as exceptions to the rule in a manner reminiscent of ‘What have the Romans done for us?’, e.g. John S. Kloppenborg, Q, the Earliest Gospel, 32, ‘Apart from three Q pericopae concerning John the Baptist and the temptation that Matthew and Luke have connected with Mark in the same way . . ., Matthew and Luke always place the Mark-Q combination differently.’ The fourth, fifth and sixth examples above also have similarities in ordering. Luke 9.1–6 is in the Marcan order, while Luke 10.1–12 comes shortly afterwards at the outset of Luke’s central section setting out Jesus’ journey as ‘the Way of the Lord’. Both Matthew and Luke add the parable of the Leaven to the parable of the Mustard Seed (Matt 13.33 // Luke 13.20) on which see further Sanders and Davies, Studying, 79 and Ed Parish Sanders, ‘The Argument from Order and the Relationship between Matthew and Luke’, NTS 15 (1969): 249–61 (260–1). Luke features a chreia about Jesus’ true family shortly after the Beelzebub Controversy (Luke 11.27–28, cf. Matt 12.46–50 // Mark 3.31–35 // Luke 8.19–20), on which see Michael Goulder, Luke, 509–11, and Eric Eve, ‘The Devil in the Detail: Exorcising Q from the Beelzebul Controversy’, in Marcan Priority, 16–43, here 32–3.

11. Taking our Leave of Mark–Q Overlaps

205

fresh structure alongside more than a verse of new material not found in Mark, at exactly the same point in mid-sentence:10

Matthew 3.11–12

11

Mark 1.7–8

Luke 3.15–17

7

15

καὶ ἐκήρυσσεν λέγων·

Ἐγὼ μὲν ὑμᾶς βαπτίζω ἐν ὕδατι εἰς μετάνοιαν· ὁ δὲ ὀπίσω μου ἐρχόμενος Ἔρχεται ὁ ἰσχυρότερός μου ὀπίσω ἰσχυρότερός μού ἐστιν, μου, οὗ οὐκ εἰμὶ ἱκανὸς οὗ οὐκ εἰμὶ ἱκανὸς τὰ κύψας λῦσαι τὸν ἱμάντα τῶν ὑποδημάτων ὑποδήματα βαστάσαι· αὐτοῦ· 8 ἐγὼ ἐβάπτισα ὑμᾶς ὕδατι, αὐτὸς ὑμᾶς βαπτίσει αὐτὸς δὲ βαπτίσει ὑμᾶς ἐν πνεύματι ἁγίῳ καὶ ἐν πνεύματι ἁγίῳ. πυρί· 12 οὗ τὸ πτύον ἐν τῇ χειρὶ αὐτοῦ, καὶ διακαθαριεῖ τὴν ἅλωνα αὐτοῦ καὶ συνάξει τὸν σῖτον αὐτοῦ εἰς τὴν ἀποθήκην, τὸ δὲ ἄχυρον κατακαύσει πυρὶ ἀσβέστῳ.

Προσδοκῶντος δὲ τοῦ λαοῦ καὶ διαλογιζομένων πάντων ἐν ταῖς καρδίαις αὐτῶν περὶ τοῦ Ἰωάννου, μήποτε αὐτὸς εἴη ὁ χριστός,16 ἀπεκρίνατο λέγων πᾶσιν ὁ Ἰωάννης· Ἐγὼ μὲν ὕδατι βαπτίζω ὑμᾶς· ἔρχεται δὲ ὁ ἰσχυρότερός μου, οὗ οὐκ εἰμὶ ἱκανὸς λῦσαι τὸν ἱμάντα τῶν ὑποδημάτων αὐτοῦ· αὐτὸς ὑμᾶς βαπτίσει ἐν πνεύματι ἁγίῳ καὶ πυρί· 17 οὗ τὸ πτύον ἐν τῇ χειρὶ αὐτοῦ διακαθᾶραι τὴν ἅλωνα αὐτοῦ καὶ συναγαγεῖν τὸν σῖτον εἰς τὴν ἀποθήκην αὐτοῦ, τὸ δὲ ἄχυρον κατακαύσει πυρὶ ἀσβέστῳ.

10.  Matthew’s and Luke’s parallel reworking of Mark’s composite quotation from Exod 23.20, Mal 3.1 and Isa 40.1 earlier in the pericope (Matt 3.3 // Mark 1.2–3 // Luke 3.4–6) is also striking. They agree (1) in extracting Isa 40.1, (2) saving Exod 23.20 and Mal 3.1 for the same point later in the narrative (Matt 11.10 // Luke 7.27), and (3) doing so using the same wording found in Mark 1.2–3. This presents difficulties for the Two-Source Theory. See Mark Goodacre, ‘The Evangelists’ Use of the Old Testament and the Synoptic Problem’, in New Studies in the Synoptic Problem: Oxford Conference, 2008. FS Christopher M. Tuckett, ed. Paul Foster, Andrew Gregory, John S. Kloppenborg and Joseph Verheyden, BETL 239 (Leuven: Peeters, 2011), 281–98, here 284–9, and literature cited there.

206

Mark Goodacre

Table (Continued) Matthew 3.11–12

Mark 1.7–8

Luke 3.15–17

11 ‘I, on the one hand, am baptizing you with water for repentance, but he who is coming after me is mightier than I, and I am not fit to remove his sandals; he will baptize you with the Holy Spirit and fire. 12 His winnowing fork is in his hand, and he will clear his threshing floor; and he will gather his wheat into the barn, but he will burn up the chaff with unquenchable fire.’

7 And he was preaching, 15 Now while the people and saying, were in a state of expectation and all were wondering in their hearts about John, as to whether he might be the Christ, 16 John answered and said to them all, ‘I, on the one hand, am baptizing you with water; ‘After me one but one is coming who is is coming who is mightier than I, and I am mightier than I, and I am not fit to stoop down and not fit to untie the thong of his untie the thong of his sandals. 8 I baptized you sandals; with water; but he will he will baptize you with the Holy baptize you with the Holy Spirit.’ Spirit and fire. 17. His winnowing fork is in his hand to clear his threshing floor, and to gather the wheat into His barn; but he will burn up the chaff with unquenchable fire.’

The agreement between Matthew and Luke is clear and substantial. Mark’s John first speaks about the coming of ‘the one who is mightier than me’( Ἔρχεται ὁ ἰσχυρότερός μου, Mark 1.7), and subsequently says, ‘I baptized you with water . . .’ (ἐγὼ ἐβάπτισα ὑμᾶς ὕδατι . . ., Mark 1.8). Both Matthew and Luke, in parallel, structure the saying differently, with a μὲν . . . δέ sentence (‘On the one hand . . . on the other hand’) in which John first announces that he is baptizing in water (now present tense, Ἐγὼ μὲν ὑμᾶς βαπτίζω ἐν ὕδατι . . .) before he talks about the stronger one who is coming (Matt 3.11 // Luke 3.16). They then insert almost exactly the same wording at exactly the same place in the Marcan narrative, even to the extent of running on at the same point with a conjunction not found in Mark. ‘He will baptize you with the holy spirit’, Mark’s John prophesies (Mark 1.8). ‘And fire!’ continue Matthew’s and Luke’s John in parallel (Matt 3.11 // Luke 3.16)

11. Taking our Leave of Mark–Q Overlaps

207

before running on with an entire verse describing harvest, separation and fiery judgement (Matt 3.12 // Luke 3.17).11 It is, of course, far too much to attribute this kind of agreement to Matthew’s and Luke’s independent redaction of Mark, and no two-source theorist takes this route. In order to retain the notion of Luke’s independence from Matthew, two-source theorists appeal to Mark–Q overlap. According to this theory, Mark and Q here overlap in content and wording, and Matthew and Luke independently make the same use of the overlapping content and wording in Mark and Q.12

11.2 What Is the Problem? It is important to notice what is happening here. In order to explain substantive agreement between Matthew and Luke against Mark, the Two-Source Theory appeals to an ‘overlapping’ of its two sources, but in making this appeal, its advocates are conceding one of the fundamental arguments that is made for the existence of Q in the first place, that Luke never features Matthew’s additions to Mark in triple-tradition material. It is worth remembering just how fundamental an argument for the existence of Q this is. Influential advocates of the existence of Q press the point. ‘Is it conceivable’, Kümmel asks, ‘that Luke would have taken over none of the Matthean additions to the Markan text?’13 Christopher Tuckett similarly asks, ‘If Luke knew Matthew, why does he never show any knowledge of Matthew’s redaction of Mark?’14 So too Fitzmyer, ‘Luke is never seen to reproduce the typically Matthean additions within the Triple Tradition.’15 In each case, these scholars use the word ‘never’.16 Luke supposedly never reproduces Matthew’s 11. Cf. Watson, Gospel Writing, 140, ‘Matthew’s addition of “and fire” . . . also connects this with the typically Matthean judgement-saying that now follows, concluding as it does with a reference to “unquenchable fire” (cf. Mt. 13.30; 25.41).’ On the Matthean apocalyptic scenario here, see further below. 12.  Watson, Gospel Writing, 139–41, notes the extraordinary coincidences that are involved if Luke and Matthew are here redacting Mark and Q independently of one another. 13.  Werner Georg Kümmel, Introduction to the New Testament (Engl. Trans., London: SCM, 1966; rev. ed., 1975), 50. 14.  Christopher M.  Tuckett, ‘The Existence of Q’, in The Gospel Behind the Gospels, ed. Ronald A.  Piper, 19–47 (25); Q and the History of Early Christianity:  Studies on Q (Edinburgh: T&T Clark, 1996), 8. 15.  Joseph A. Fitzmyer, The Gospel According to Luke I-IX, The Anchor Bible (AB) 28 (New York: Doubleday, 1981), 73–4. 16.  Stein, Studying the Synoptic Gospels, 99, appears to be conscious of the difficulty and so places ‘never’ in inverted commas, ‘One of the strongest arguments against the use of Matthew by Luke is the fact that when Matthew has additional material in the triple tradition (‘Matthean additions to the narrative’), it is “never” found in Luke.’ Here ‘never’ appears to mean ‘occasionally’.

208

Mark Goodacre

additions to Mark, a claim that is contradicted by the phenomenon of the ‘Mark– Q overlap’ passages in which Luke agrees with Matthew’s substantive additions to Mark. The scholars in question generally qualify the argument by listing tripletradition passages where Luke does not feature Matthew’s additions to Mark,17 but this does not detract from the misleading nature of an argument that uses the term ‘never’ when ‘sometimes’ is meant.18 The Q sceptic’s objection to the Mark–Q overlap theory is not, therefore, an objection to the idea of overlapping sources. It is a question of what the concession implies about arguments for the very existence of Q. These are passages that should not be present given the way that the model is constructed. The foundation of the Two-Source Theory is the notion that Matthew and Luke edited Mark independently of one another, yet here they are manifestly editing Mark in identical ways, with substantial agreements in wording and order. Proponents of the Q hypothesis frequently frame the difficulty posed by the Mark–Q overlaps as an abstract objection to the idea of overlapping sources. Kloppenborg, for example, says: Although a few critics of 2DH [the Two-Document Hypothesis] make much of the Mark–Q overlaps as a significant problem, these really pose little difficulty. 17.  Even here, the evidence is sometimes overstated, e.g. Fitzmyer and Tuckett suggest that it would be surprising for Luke to have omitted Matthew’s additions to an entire pericope that Luke omits from Mark (Matt 14.28–31, Peter’s Walking on the Water), on which see my Case Against Q, 49–51. This point is conceded by Christopher Tuckett in ‘The Current State of the Synoptic Problem’, in New Studies, ed. Paul Foster et al., 9–50 (41 n. 82). Similarly, the frequent expressions of surprise at Luke’s omission of Matt 16.17–19 (Jesus’ commendation of Peter) generally ignore Luke’s omission of Mark 8.33 (Jesus’ rebuke of Peter), on which see Mark Goodacre, ‘The Rock on Rocky Ground: Matthew, Mark and Peter as Skandalon’, in What Is It That the Scripture Says?:  Essays in Biblical Interpretation, Translation, And Reception. FS Henry Wansbrough Osb, ed. Philip McCosker, LNTS 316 (London: Continuum, 2006), 61–73 (68–9). John Kloppenborg, ‘On Dispensing with Q: Goodacre on the Relation of Luke to Matthew’, NTS 49 (2003):  210–36, here 221–2 and Tuckett, ‘Current State’, 40–1 find Luke’s omission of Matt 16.17–19 surprising in the light of Luke’s interest in Peter, but my point (Case Against Q, 51) is not about redactional tendency but about narrative development. Peter indeed has ‘a key role’ in Acts (Kloppenborg, ‘On Dispensing’, 222) but what Matthew’s Jesus prophesies is not that Peter is important but that he is foundational. 18. Compare Sanders’s more accurate summary, ‘The two-document hypothesis is dependent on the statement that Matthew and Luke do not agree together against Mark, and holders of that hypothesis are forced to explain away the existence of the actual agreements. The principal weakness of the two-source theory is that it is constructed to explain the synoptic interrelationships on the assumption that there are no agreements between Matthew and Luke against Mark, rather than on the assumption that there are almost always some, but usually not many such agreements. Since the latter statement is the true one, the two-document hypothesis must be considered inadequate, at least as a hypothesis which gives an account of all the evidence’ (‘Mark-Q Overlaps’, 453). See further my discussion in Case Against Q, 49–54.

11. Taking our Leave of Mark–Q Overlaps

209

In the first place, it is hardly surprising that two independent tellings of the Jesus tradition (Mark and Q) should sometimes narrate the same events or sayings. And they do.19

The difficulty with the Mark–Q overlaps has nothing to do with the plausibility or otherwise of overlapping sources. The point is, in the words of E. P. Sanders, as follows: The proposed Mark–Q overlaps show the weakness of the two-document hypothesis. They are mostly points at which there are too many agreements between Matthew and Luke against Mark – both in words and in order – to be overlooked.’20

The category effectively achieves a kind of argumentative coup for the TwoSource Theory by allowing its advocates to insist on the counter-factual claim that Luke ‘never’ agrees with Matthew in substantive ways in triple-tradition material. Moreover, there is an element of misdirection here. Two-Source theorists do discuss the minor agreements between Matthew and Luke against Mark but by placing these in a separate category of their own, as if fundamentally different in nature from the Mark–Q overlaps, they are able to imply that Matthew and Luke only agree in ‘minor’ ways in the triple-tradition material. The issue here is partly a question of how we choose to describe the data. The difficulty with the term ‘Mark–Q Overlaps’ is that it describes a particular set of data, passages featuring major agreements between Matthew and Luke against Mark, using the terminology of the Two-Source Theory.21 This is far from ideal 19. Kloppenborg, Q: The Earliest Gospel, 34 (emphasis original). Kloppenborg compares the overlapping of John and the Synoptics, and Thomas and the Synoptics (ibid.). The analogy will not be especially helpful to those who attribute the overlapping to John’s familiarity with the Synoptics and Thomas’s familiarity with the Synoptics. See also Christopher Tuckett, ‘Current State’, 46, ‘In general terms, the possibility that sources overlapped seems not at all implausible or impossible. Hence a ‘Mark-Q overlap’ in principle is surely not an inherent problem’. See similarly Brice Jones, Matthean and Lukan Special Material: A Brief Introduction with Texts in Greek and English (Eugene: Wipf and Stock, 2011), 7, ‘Most advocates of the Two-Source hypothesis, however, do not think that the Mark-Q overlaps pose any real threat, since two independent yet similar traditions are bound to have existed prior to the composition of the Gospels as we know them’ which likewise responds to an argument that no one is actually making. See also Stein, Studying the Synoptic Gospels, 120–1, ‘Yet, on a purely theoretical basis, it would be most unusual if two sources concerning Jesus, such as Mark and Q, did not overlap in some way. After all, they do deal with the same person, with incidents in his life and sayings that he uttered, so that some overlap would be expected.’ 20. Sanders, ‘Mark-Q Overlaps’, 463. 21.  In German speaking scholarship, the term often used for the Mark-Q overlaps is Doppelüberlieferung, e.g. especially Laufen, Die Doppelüberlieferungen.

210

Mark Goodacre

if there is to be a level playing field for studying the phenomenon. Adherents of the Farrer theory do not label minor agreements between Matthew and Luke as ‘Luke’s minor borrowing from Matthew in triple tradition’. If they did, they would be rightly criticized for allowing their description of the set of data to be influenced by their explanation of the data.22 In order to encourage an open and fair discussion, therefore, there should be ‘mandatory retirement’23 for using ‘Mark–Q overlaps’ as a label for a set of data. Instead, the phenomenon should be described more neutrally as ‘major agreements between Matthew and Luke against Mark’ or more simply ‘the major agreements’.24

11.3 Major and Minor Agreements The development of these discrete categories, ‘Mark–Q overlaps’ on the one hand and ‘minor agreements’ on the other, is not just a failure to describe the data as neutrally and objectively as possible. By assigning different passages to different categories, the student can fail to see that the agreement between Matthew and Luke is on a continuum, from lesser to greater degrees of agreement between Matthew and Luke. The continuum ranges from pure triple tradition to pure double tradition, with varying degrees of agreement along the way, from relatively minor to quite major agreement between Matthew and Luke against Mark. It is a pattern that makes good sense on the assumption that Luke uses Mark, but supplements his use of Mark with his use of Matthew, sometimes in minor ways, sometimes in major ways. The Mark-Q overlaps are not a different category of data to be

22.  In Mark Goodacre, The Synoptic Problem:  A Way through the Maze (London: Continuum, 2001), ch. 2, I attempted to describe the synoptic data without prejudice to any Synoptic theory. This was in conscious contrast to the usual approach that refracts the data through the particular solution that is being proposed. My suggestion is that introductory students are often prejudiced in favour of a particular solution because of the way that the data is described. 23.  For the term, see Paula Fredriksen, ‘Mandatory Retirement: Ideas in the Study of Christian Origins Whose Time Has Come to Go’, Studies in Religion / Sciences Religieuses 35 (2006): 231–46. 24.  See further my Case Against Q, ch. 8, ‘Major and Minor Agreements’ for the terminology and the argument. Occasionally, the Mark-Q overlaps are described as ‘major agreements’ by Q advocates, e.g. Kloppenborg, Q:  The Earliest Gospel, 35, ‘If the MarkQ overlaps could be called the major agreements of Matthew and Luke against Mark, there are also a number of smaller agreements of Matthew and Luke against Mark.’ Compare also Streeter’s language, ‘In nearly all the passages we have examined, the verbal agreements between Matthew and Luke against Mark are very substantial’ (‘St. Mark’s Knowledge’, 178).

11. Taking our Leave of Mark–Q Overlaps

211

separated off and explained in a unique way. They are, rather, points that one can plot on a graph, points on a continuum.25 Christopher Tuckett responds to this point in the following way: Many would however argue that there is no continuous spectrum of agreements in the tradition:  rather, they virtually all seem to cluster into two groups: one of ‘minor’ agreements made up almost exclusively of tiny details, mostly totally inconsequential in relation to meaning, and one of ‘major’ agreements, where many would argue that the agreements are so extensive that they seem to provide evidence for the existence of a non-Markan version of the tradition. And there is not much in between these poles. Thus to talk about a ‘continuum’ of agreements may not be fully persuasive. That there are different categories of agreement is undeniable  – but they seem to be more discrete than continuous.26

The idea that there is a spectrum of agreement is, however, straightforward to illustrate. The two ends of the spectrum are clear. There are triple-tradition passages with just a few minor agreements on one end of the spectrum,27 and there are

25.  I attempted to develop this point in Case Against Q, 163–4. The point was anticipated, though I  had not realized it, by Sanders in ‘Mark-Q Overlaps’, 454–5, ‘Virtually every single pericope in the triple tradition has some such agreements. One could construct a chart enumerating the agreements of Matthew and Luke against Mark by pericope, and the chart would show numbers ranging from 0 to 34, with many of the numbers in between being listed. (This is counting only strong positive agreements – the presence of the same word in Matthew and Luke which is not in Mark – and excepting the proposed overlaps).’ 26.  Tuckett, ‘Current State’, 43, emphasis original (but note one minor concession at 43 n.  87 in relation to Mark 12.28–34); cf. Christopher Tuckett, ‘Review of Mark Goodacre, Case Against Q’, Novum Testamentum (NovT) 46 (2004):  401–3, here  402, ‘Goodacre contents himself with the general point about Luke’s using Matthew’s additions to Mark, and/or referring to different levels of ‘agreements’ against Mark here; he talks about a ‘broad spectrum’ and a ‘sliding scale’ (161) or a ‘continuum’ (163) of the level of Matthew-Luke agreements against Mark. However, he never analyses any of these ‘overlap’ passages in any detail. And in terms of any ‘broad spectrum’, the trouble is that there is not much by way of a ‘continuum’: there are examples at both ends of the spectrum but not much in between.’ 27.  Note, however, Sanders’s important point that there are very few passages that feature no minor agreements between Matthew and Luke against Mark, ‘One can best observe the extent of these agreements, both positive agreements and agreements in omission, by trying to find a triple-tradition pericope which does not have any agreements between Matthew and Luke against Mark; there are only one or two’ (‘Mark-Q Overlaps’, 453 n. 4, emphasis original); cf. Sanders and Davies, Studying the Synoptic Gospels, 67.

212

Mark Goodacre

double-tradition passages where Mark is not present at all at the other end of the spectrum. The material in between these two poles is what is here in focus. These are the figures for five of the six major Mark–Q overlap passages:28 (1) Matt 3.1–12 // Mark 1.1–8 // Luke 3.1–18 (John the Baptist): 94 Matt-Luke agreements against Mark out of 345 words in Luke (27%). (2) Matt 3.13–17 // Mark 1.9–11 // Luke 3.21–2 (Baptism of Jesus): 3 Matt-Luke agreements against Mark out of 43 words in Luke (7%). (3) Matt 4.1–11 // Mark 1.12–13 // Luke 4.1–13 (Temptation): 114 Matt-Luke agreements against Mark out of 203 words in Luke (56%). (4) Matt 12.25–32 // Mark 3.23–30 // Luke 11.17–23, 12.10 (Beelzebub): 65 Matt-Luke agreements against Mark out of 146 words in Luke (45%). (5) Matt 13.31–2 // Mark 4.30–2 // Luke 13.18–19 (Mustard Seed): 11 MattLuke agreements against Mark out of 40 words in Luke (28%). In this selection of key Mark–Q overlap passages, there are Matt-Luke agreements ranging from 7 per cent of Luke’s words in a given pericope to 56 per cent in a given pericope, with agreement levels well spaced in that range, at 27, 28 and 45 per cent. If one were to sample particular sub-pericopae within the larger pericopae, there can be remarkably high degrees of agreement. In Matt 3.12 // Luke 3.17, for example, 88 per cent of Luke’s words agree with Matthew; in Matt 3.7–10 // Luke 3.7–9, 85 per cent of Luke’s words agree with Matthew’s. To place this in context, this kind of agreement is as high as one sometimes sees in the pure doubletradition material.29 The question, though, is how this level of agreement compares with levels of agreement among triple-tradition passages that feature minor agreements. There are plenty of pericopae with just a handful of minor agreements, producing a percentage of agreement lower than the 7 per cent we see in the Baptism, for example: (6) Matt 8.1–4 // Mark 1.40–45 // Luke 5.12–16 (Leper): 5 Matt-Luke agreements against Mark out of 98 words in Luke (5%). (7) Matt 9.1–8 // Mark 2.1–12 // Luke 5.17–26 (Paralytic): 12 Matt-Luke agreements against Mark out of 212 words in Luke (6%). (8) Matt 9.9–13 // Mark 2.13–17 // Luke 5.27–32 (Levi): 5 Matt-Luke agreements against Mark out of 94 words in Luke (5%).

28. The figures are from Sanders, ‘Mark-Q Overlaps’, 457–8. He does not provide figures for the Mission Discourse which is more complex because the overlap is across two Lucan contexts, one with more agreements with Mark (Luke 9.1–6) and one with more agreements with Matthew (Luke 10.1–12). 29.  On the question of high verbatim agreement in the double tradition, see Mark Goodacre, ‘Too Good to Be Q:  High Verbatim Agreement in the Double Tradition’, in Marcan Priority without Q, ed. Poirier and Peterson, 82–100.

11. Taking our Leave of Mark–Q Overlaps

213

The lower end of agreement in a Mark–Q overlap passage (7%) is just above the level of agreement in these triple-tradition passages that feature minor agreements (5% and 6%). It is worth asking the additional question whether there are any (non Mark–Q overlap) triple-tradition pericopae in which the number of Matthew-Luke agreements against Mark are as great as 7 per cent of the total Lucan words in the pericope. If so, it will be clear that the continuum is one in which the supposedly different categories in fact overlap with one another. There are indeed such pericopae: (9) Matt 14.13–21 // Mark 6.30–44 // Luke 9.10–17 (Five Thousand): 15 MattLuke agreements against Mark out of 164 words in Luke (9%). (10) Matt 21.1–9 // Mark 11.1-10 // Luke 19.28–38 (Entry into Jerusalem): 12 Matt-Luke agreements against Mark out of 167 words in Luke (7%). (11) Matt 21.23–27 // Mark 11.27–33 // Luke 20.1–8 (Question About Authority): 10 Matt-Luke agreements against Mark out of 118 words in Luke (8%). (12) Matt 22.34–40 // Matt. 12.28–34 // Luke 10.25–28 (Great Commandment): 18 Matt-Luke agreements against Mark out of 73 words in Luke (25%).30 (13) Matt 26.57–75 // Mark 14.53–72 // Luke 22.54–71 (Trial and Peter’s Denial): 25 Matt-Mark agreements against Mark out of 263 words in Luke (9.5%).31 These examples show that the degree of agreement between Matthew and Luke against Mark is sometimes higher in passages with so-called ‘minor agreements’ than it is in passages that are labelled ‘Mark–Q overlap’. In other words, it is clear that there is not only a continuum with different degrees of agreement from low (triple-tradition passages with few Minor Agreements) to high (Mark–Q overlap passages with many Major Agreements), but there is also some overlapping between the degree of agreement between Matthew and Luke in passages normally designated Mark–Q overlap and passages normally designated triple tradition.32

30.  This passage has occasionally been included in Q and so would also be a Mark-Q overlap. In order to avoid subjectivity, I  have included as Mark-Q overlap passages only those included in Q by the International Q Project. (See also above, n. 7.) 31. The figures for the triple-tradition passages are extrapolated from A. M. Honore, ‘A Statistical Study of the Synoptic Problem’, NovT 10 (1968): 95–147. 32. Moreover, there is one Matthew-Luke agreement assigned to Mark-Q overlap by the International Q Project that amounts to just one word, Ναζαρά in Matt 4.13 // Luke 4.16, on which see Shawn Carruth and James M.  Robinson, Q 4:1–13, 16. The Temptations of Jesus – Nazara, volume editor: Christoph Heil (Leuven: Peeters, 1996), 392–462 and Frans Neirynck, ‘NAZARA in Q: Pro and Con’, in From Quest to Q. FS James M. Robinson, ed. Jon Ma. Asgeirsson, Kristin de Troyer and Marvin W. Meyer, BETL 146 (Leuven:  Leuven University Press, 2000), 159–69 and Michael Goulder, ‘Two Significant Minor Agreements (Mat. 4:13 Par.; Mat. 26:67–68 Par.)’, NovT 45 (2003): 365–73.

214

Mark Goodacre

11.4 The Degree of Mark in Luke’s Mark–Q Overlap Passages There is a further, related point about the nature of the Mark–Q overlap passages. It is commonly claimed that Matthew’s and Luke’s versions of the Mark–Q overlap passages look different. Matthew tends to blend Mark and Q in the overlap passages whereas Luke tends to reproduce Q alone. Drawing on Gerald Downing’s work,33 Christopher Tuckett underlines this point: In these passages, one can indeed refer to Luke’s use of Matthew’s additions to Mark, and/or to extensive non-trivial Matthew-Luke agreements. However, any non-Q theory has to explain Luke’s apparently almost pathological refusal in some of these texts to use any Markan material at all (e.g. the Beelzebul controversy, or the Mustard Seed). As Gerald Downing argued many years ago, Luke’s procedure on the Farrer-Goulder-Goodacre model appears totally at odds with his procedure elsewhere (where, according to Goodacre and others, Luke knows Mark far better than Matthew and uses Mark in preference to Matthew). In these passages, Luke must have studiously avoided all the points where Matthew and Mark agree and reproduced only Matthew’s additions to Mark.34

It is one of the curiosities of scholarship on the SP that a point with so slim an evidentiary basis is so often repeated. It is simply not the case that Luke lacks Marcan material in the Mark–Q overlap passages. In John the Baptist’s preaching (Matt 3.1–12 // Mark 1.1–8 // Luke 3.1–18), for example, there are extensive triple agreements as well as agreements between Mark and Luke alone. In the section surveyed above (Matt 3.11–12 // Mark 1.7–8 // Luke 3.15–17), there are clear Mark-Luke agreements like the claim that John is unworthy ‘to loose the thong on his sandals’ (λῦσαι τὸν ἱμάντα τῶν ὑποδημάτων αὐτοῦ, Mark 1.7 // Luke 3.16; contrast Matt 3.11, τὰ ὑποδήματα βαστάσαι).35 Similarly, the baptism story features many triple agreements (represented by double underline) as well as a couple of minor Mark-Luke agreements (represented by single underline):

33.  The key article is F. Gerald Downing, ‘Towards the Rehabilitation of Q’, NTS 11 (1964):  169–81 (but see also more recently ‘Disagreements of Each Evangelist with the Minor Close Agreements of the Other Two’, Ephemerides Theologicae Lovanienses (ETL) 80 [2004]: 445–69). Downing’s 1964 article is persuasively answered in Ken Olson, ‘Unpicking on the Farrer Theory’, in Questioning Q: A Multidimensional Critique, ed. Mark Goodacre and Nicholas Perrin (London: SPCK, 2004), 127–50. Tuckett does not mention Olson’s discussion. See now also Eric Eve, ‘The Devil in the Detail’, 16–43, which provides a persuasive exegesis of the Beelzebub Controversy on the basis of the Farrer theory. 34. Tuckett, ‘Review of Case Against Q’, 402. The point is made again in Tuckett, ‘Current State’, 44. 35. See the synopsis of the passages above 205-6 which clearly shows the triple agreements as well as the Mark-Luke only agreements.

11. Taking our Leave of Mark–Q Overlaps

Matt 3.13, 16–1736 1

13

Luke 3.21–2

9

21

Καὶ ἐγένετο ἐν ἐκείναις ταῖς ἡμέραις ὁ Ἰησοῦς ἀπὸ ἦλθεν Ἰησοῦς ἀπὸ τῆς Γαλιλαίας Ναζαρὲτ τῆς Γαλιλαίας ἐπὶ τὸν Ἰορδάνην πρὸς τὸν Ἰωάννην τοῦ καὶ ἐβαπτίσθη εἰς τὸν βαπτισθῆναι ὑπ’ αὐτοῦ . Ἰορδάνην ὑπὸ Ἰωάννου. ... 16 βαπτισθεὶς δὲ ὁ Ἰησοῦς 10 καὶ εὐθὺς εὐθὺς ἀνέβη ἀπὸ τοῦ ἀναβαίνων ἐκ τοῦ ὕδατος· καὶ ἰδοὺ ὕδατος εἶδεν ἠνεῴχθησαν οἱ σχιζομένους τοὺς οὐρανοί, καὶ εἶδεν οὐρανοὺς καὶ πνεῦμα θεοῦ τὸ πνεῦμα καταβαῖνον ὡσεὶ περιστερὰν ὡς περιστερὰν ἐρχόμενον ἐπ’ αὐτόν· καταβαῖνον εἰς αὐτόν· 17 καὶ ἰδοὺ φωνὴ ἐκ 11 καὶ φωνὴ ἐγένετο  ἐκ τῶν οὐρανῶν λέγουσα· τῶν οὐρανῶν· Οὗτός ἐστιν ὁ υἱός μου ὁ Σὺ εἶ ὁ υἱός μου ὁ ἀγαπητός, ἐν ᾧ ἀγαπητός, ἐν σοὶ εὐδόκησα. εὐδόκησα. 13

Τότε παραγίνεται

Mark 1.9–10

Then Jesus came from Galilee to John at the Jordan, to be baptized by him … . 16 And when Jesus had been baptized, just as he came up from the water, suddenly the heavens were opened to him and he saw the Spirit of God descending like a dove and alighting on him. 17 And a voice from heaven said, ‘This is my Son, the Beloved, with whom I am well pleased.’

215

Ἐγένετο δὲ ἐν τῷ βαπτισθῆναι ἅπαντα τὸν λαὸν καὶ Ἰησοῦ

βαπτισθέντος καὶ προσευχομένου ἀνεῳχθῆναι τὸν οὐρανὸν 22 καὶ καταβῆναι τὸ πνεῦμα τὸ ἅγιον σωματικῷ εἴδει ὡς περιστερὰν ἐπ’ αὐτόν, καὶ φωνὴν ἐξ οὐρανοῦ γενέσθαι· Σὺ εἶ ὁ υἱός μου ὁ ἀγαπητός, ἐν σοὶ εὐδόκησα.

9

In those days Jesus came 21 Now when all the from Nazareth of Galilee people were baptized, and was baptized by and when Jesus also had John in the Jordan. been baptized and was praying, 10 And just as he was coming up out of the the water, he saw the heaven was opened, heavens torn apart and 22 the Spirit and the Holy Spirit descending like a descended upon him in dove into him. bodily form like a dove. 11 And a voice came from And a voice came from heaven, ‘You are my heaven, ‘You are my Son, the Beloved; with Son, the Beloved; with you I am well you I am well pleased.’ pleased.’

36.  I have omitted Matt 3.14–15 from the synopsis because there are no parallels in Mark and Luke so it is not relevant to the discussion of Mark // Luke agreements here. Kloppenborg (‘On Dispensing’, 219–20) argues that Luke would not have omitted these

216

Mark Goodacre

There is more triple agreement in this pericope than there is Matthew-Luke agreement. It is clear that Luke is not avoiding Mark’s wording in passages like this.37 The specific examples mentioned by Tuckett are the Beelzebub Controversy (Matt 12.25–32 // Mark 3.23–30 // Luke 11.17–23, 12.10) and the Mustard seed (Matt 13.31–32 // Mark 4.30–32 // Luke 13.18–19). These passages include several agreements between Mark and Luke. There are triple agreements and there are Mark-Luke agreements. Luke’s versions are clearly not Mark-free zones. On Sanders’s count, the Beelzebub Controversy features 31 triple agreements, 35 Matthew-Mark agreements, 5 Mark-Luke agreements and 65 Matthew-Luke agreements. Similarly, the Mustard Seed, on Sanders’s count, features 14 triple agreements, 11 Matthew-Mark agreements, 6 Mark-Luke agreements and 11 Matthew-Luke agreements.38 These are straightforward to see in synopsis: Matt 13.31–32

Mark 4.30–32

Ἄλλην παραβολὴν Καὶ ἔλεγεν· Πῶς παρέθηκεν αὐτοῖς λέγων· ὁμοιώσωμεν τὴν βασιλείαν τοῦ θεοῦ, ἢ ἐν τίνι αὐτὴν παραβολῇ Ὁμοία ἐστὶν θῶμεν; 31 ὡς ἡ βασιλεία τῶν οὐρανῶν κόκκῳ σινάπεως, ὃν κόκκῳ σινάπεως, ὃς λαβὼν ἄνθρωπος ὅταν ἔσπειρεν ἐν τῷ ἀγρῷ σπαρῇ ἐπὶ τῆς γῆς, αὐτοῦ· 32 ὃ μικρότερον μικρότερον μέν ἐστιν πάντων τῶν ὂν πάντων τῶν σπερμάτων, σπερμάτων τῶν ἐπὶ τῆς ὅταν δὲ αὐξηθῇ γῆς – 32 καὶ ὅταν σπαρῇ, ἀναβαίνει καὶ γίνεται μεῖζον τῶν μεῖζον πάντων τῶν λαχάνων ἐστὶν καὶ λαχάνων καὶ γίνεται δένδρον, ποιεῖ κλάδους ὥστε ἐλθεῖν μεγάλους, ὥστε δύνασθαι ὑπὸ τὴν σκιὰν

Luke 13.18–19 Ἔλεγεν οὖν· Τίνι ὁμοία ἐστὶν ἡ βασιλεία τοῦ θεοῦ, καὶ τίνι ὁμοιώσω αὐτήν; 19 ὁμοία ἐστὶν κόκκῳ σινάπεως, ὃν λαβὼν ἄνθρωπος ἔβαλεν εἰς κῆπον ἑαυτοῦ,

καὶ ηὔξησεν καὶ

ἐγένετο εἰς δένδρον, καὶ

verses if he had known them. However, Luke has completed his John narrative in the previous passage with John’s arrest, which moves him off stage (Luke 3.18–20), so he is hardly likely to reintroduce the character at this point. 37.  The Mission Discourse (the fourth example above) is also rich in Marcan material but is more complicated because unlike the other examples, there are two versions in Luke, the first in Luke 9.1–6 and the second in Luke 10.1–12. But even Streeter acknowledged the presence of Marcan material in Luke 9.1–6, ‘Matthew as usual conflates Mark and Q, and so for once to some small extent does Luke in 9.1–5, but Luke also has a version in 10.1–12, much, if not all, of which is Q’ (‘St. Mark’s knowledge’, 173). 38. Sanders, ‘Mark-Q Overlaps’, 458.

11. Taking our Leave of Mark–Q Overlaps

217

Table (Continued) Matt 13.31–32

Mark 4.30–32

Luke 13.18–19

τὰ πετεινὰ τοῦ αὐτοῦ τὰ πετεινὰ τοῦ οὐρανοῦ καὶ οὐρανοῦ κατασκηνοῦν ἐν τοῖς κατασκηνοῦν. κλάδοις αὐτοῦ.

τὰ πετεινὰ τοῦ οὐρανοῦ κατεσκήνωσεν ἐν τοῖς κλάδοις αὐτοῦ.

He put before them another parable: ‘The kingdom of heaven

He said therefore, ‘What is the kingdom of God like? And to what should I compare it? 19 It is like a mustard seed that someone took and sowed in the garden;

He also said, ‘With what can we compare the kingdom of God, or what parable will we is like a use for it? 31 It is like a mustard seed that mustard seed, which, someone took and sowed when sown in his field; 32 it is the upon the ground, is the smallest of all the seeds, smallest of all the seeds on earth; 32 yet when it is but when it has grown it sown it grows up and is the greatest of becomes the greatest of shrubs and becomes a all shrubs, and puts forth tree, so that large branches, so that the birds of the air come the birds of the air can and make nests in its make nests in its branches. shade.

it grew and became a tree, and the birds of the air made nests in its branches.

There are plenty of triple agreements here (represented by double underline). Matthew has the blander opening, simply stating ‘He put before them another parable . . .’ but Mark and Luke agree with one another in formulating the same kind of double-question about comparison with the kingdom of God, an element that is particularly striking given its presence in the introduction to the parable. Even here, in one of Tuckett’s choice examples, there is hardly an ‘almost pathological refusal’ to include any Marcan material. It is true, of course, that there is a substantial degree of agreement between Matthew and Luke against Mark in these passages, but that is why Q theorists characterize them as ‘Mark–Q overlap’ rather than as triple-tradition pericopae with minor agreements. In cases like these, the complaint is effectively that Luke agrees with Matthew too much. But if Luke is using Matthew as well as Mark, it is not surprising that on some occasions he will agree with Matthew far more than he does with Mark. It is one end of a spectrum where there is a greater degree of agreement between Matthew and Luke than there is between Mark and Luke. It is a difference in degree and not a difference in kind. On the Farrer theory, the explanation is straightforward. On occasions like this, where Matthew is the middle term among the Synoptics, Luke is working with Matthew as his primary source and not Mark. The usual triple-tradition situation, where there are major agreements with Mark and minor

218

Mark Goodacre

agreements with Matthew, is reversed and, instead, there are major agreements with Matthew and minor agreements with Mark. If Luke is working with both Matthew and Mark, it is not surprising that on occasions Luke turns to Matthew as his primary source, even in triple-tradition material. Indeed it is worth noting how often this happens where Matthew has a fuller account than Mark, in the John the Baptist material, the Temptations, Beelzebub and the Mission discourse.

11.5 The Direction of Dependence If it is true that the Mark–Q overlap passages pose serious difficulties for the TwoSource Theory, is it clear that the Farrer theory should take the spoils? The idea of standing the Farrer theory on its head and reversing the direction of dependence is occasionally aired.39 One way of approaching this question is to ask whether Luke appears to betray his knowledge of Matthew through docile reproduction while engaged in the complex business of working from his two sources.40 E. P. Sanders points to such an example in Luke’s reworking of the Mission discourse.41 Matt 10.11–15

Mark 6.10–11

11

10 καὶ ἔλεγεν αὐτοῖς· Ὅπου ἐὰν εἰσέλθητε εἰς καὶ εἰς ἣν ἂν οἰκίαν, οἰκίαν εἰσέλθητε,

εἰς ἣν δ’ ἂν πόλιν ἢ κώμην εἰσέλθητε, ἐξετάσατε τίς ἐν αὐτῇ ἄξιός ἐστιν· κἀκεῖ μείνατε ἕως ἂν ἐξέλθητε. 12 εἰσερχόμενοι δὲ εἰς τὴν οἰκίαν ἀσπάσασθε αὐτήν· 13 καὶ ἐὰν μὲν ᾖ ἡ οἰκία ἀξία, ἐλθάτω ἡ εἰρήνη ὑμῶν ἐπ’ αὐτήν· ἐὰν δὲ μὴ ᾖ ἀξία, ἡ εἰρήνη ὑμῶν

ἐκεῖ μένετε ἕως ἂν ἐξέλθητε ἐκεῖθεν.

Luke 9.4–5

ἐκεῖ μένετε καὶ ἐκεῖθεν ἐξέρχεσθε.

39. See Robert K. MacEwen, Matthean Posteriority: An Exploration of Matthew’s Use of Luke and Mark as a Solution to the Synoptic Problem, LNTS 501 (London: Bloomsbury T&T Clark, 2015). See too Ronald V. Huggins, ‘Matthean Posteriority: A Preliminary Proposal’, NovT 34 (1992): 1–21. Alan Garrow, The Gospel of Matthew’s Dependence on the Didache, JSNTSup 254 (London: T&T Clark, 2004) is also a Matthean posteriority theory, but it retains a Q source. 40.  Cf. Mark Goodacre, ‘Fatigue in the Synoptics’, NTS 44 (1998):  45–58 (54–8). In discussions of the Farrer theory over against the Two-Source theory, fatigue can only be suggestive given the hypothetical nature of Q, with which Luke may have been fatigued. However, in the discussion of the direction of dependence between Luke and Matthew, the category is more useful. 41.  Sanders and Davies, Studying the Synoptic Gospels, 95. In this synopsis, I have not added the Lucan doublet of the Mission Discourse to the Seventy (Luke 10.1–12), but it is

11. Taking our Leave of Mark–Q Overlaps

219

Table (Continued) Matt 10.11–15 πρὸς ὑμᾶς ἐπιστραφήτω. 14 καὶ ὃς ἂν μὴ δέξηται ὑμᾶς μηδὲ ἀκούσῃ τοὺς λόγους ὑμῶν, ἐξερχόμενοι ἔξω τῆς οἰκίας ἢ τῆς πόλεως ἐκείνης ἐκτινάξατε τὸν κονιορτὸν τῶν ποδῶν ὑμῶν.

11

Whatever town or village you enter, find out who in it is worthy, and stay there until you leave. 12 As you enter the house, greet it. 13 If the house is worthy, let your peace come upon it; but if it is not worthy, let your peace return to you. 14 If anyone will not welcome you or listen to your words, as you leave that house or that town, shake off the dust from your feet.

Mark 6.10–11

Luke 9.4–5

11

 καὶ ὃς ἂν τόπος μὴ δέξηται ὑμᾶς μηδὲ ἀκούσωσιν ὑμῶν,

5

καὶ ὅσοι ἂν μὴ δέχωνται ὑμᾶς,

ἐκπορευόμενοι ἐκεῖθεν ἐξερχόμενοι ἀπὸ τῆς πόλεως ἐκείνης ἐκτινάξατε τὸν τὸν χοῦν τὸν ὑποκάτω τῶν κονιορτὸν ἀπὸ τῶν ποδῶν ὑμῶν ποδῶν ὑμῶν εἰς ἀποτινάσσετε εἰς μαρτύριον αὐτοῖς. μαρτύριον ἐπ’ αὐτούς. 10 He said to them, 4 ‘Wherever Whatever you enter a house, house you enter, stay there until you leave stay there, and leave the place. from there.

11

 If any place will not welcome you and they refuse to hear you, as you leave, shake off the dust that is on your feet as a testimony against them.’

5

Wherever they do not welcome you, as you are leaving that town shake the dust off your feet as a testimony against them.

In Luke 9.5, τῆς πόλεως ἐκείνης (‘that town’) has no antecedent. Only οἰκίαν (‘house’) has been mentioned. It appears to have come over from Matt 10.14, where reference to ‘that town’ is coherent since entry to a town is mentioned earlier in the passage (Matt 10.11). On the Matthean posteriority hypothesis, Luke is working here from Mark and his reference to ‘that town’ is incomprehensible, while Matthew presciently fixes the problem ahead of time by supplying the antecedent. Luke’s use of Matthew is the stronger hypothesis.42 worth noting that there are some triple and many double (Matthew-Luke) agreements here in Luke 10.5–11. 42.  On Matthean posteriority, Matthew is here conflating Mark’s Mission Discourse (Mark 6.6b–13) with Luke’s Mission Discourse (Luke 9.1–6) and Luke’s Mission to the

220

Mark Goodacre

Luke’s use of Matthew also provides a stronger account of the distinctively Matthean nature of the material they share in the first major Mark–Q overlap passage, the Preaching of John the Baptist. The language, imagery and rhythm of the new material is Matthean through and through. The first line of John’s preaching provides a telling example: Matt 3.7 // Luke 3.7:  γεννήματα ἐχιδνῶν, τίς ὑπέδειξεν ὑμῖν φυγεῖν ἀπὸ τῆς μελλούσης ὀργῆς ‘Brood of vipers! Who warned you to flee from the coming wrath?’

Michael Goulder labels this rhythm an ‘echidnic’. The offensive vocative + rhetorical question43 occurs twice again in remarkably similar forms: 3.7: γεννήματα ἐχιδνῶν, τίς ὑπέδειξεν ὑμῖν φυγεῖν ἀπὸ τῆς μελλούσης ὀργῆς; 12.34: γεννήματα ἐχιδνῶν, πῶς δύνασθε ἀγαθὰ λαλεῖν πονηροὶ ὄντες; 23.33, ὄφεις, γεννήματα ἐχιδνῶν, πῶς φύγητε ἀπὸ τῆς κρίσεως τῆς γεέννης; 3.7: ‘Brood of vipers! Who warned you to flee from the coming wrath?’ 12.34: ‘Brood of vipers! How can you speak good things when you are evil?’ 23.33:  ‘Snakes, brood of vipers! How can you flee from the judgement of gehenna?’ These links should not be played down. These are not everyday phrases. ‘Brood of vipers’ occurs in only these places in the gospel tradition, and the notion of fleeing from wrath coheres with the broader Matthean stress on a fiery judgement expressed using harvest imagery. This becomes clear in Matt 3.10 // Luke 3.9: 3.10: πᾶν οὖν δένδρον μὴ ποιοῦν καρπὸν καλὸν ἐκκόπτεται καὶ εἰς πῦρ βάλλεται ‘Therefore every tree not producing good fruit will be cut down and thrown into the fire.’

Virtually the identical sentence occurs again in Matt 7.19. Once again it is not just the language but also the imagery and thought that is Matthean. It is Matthew among the evangelists who exploits harvest imagery to tell the story of judgement and hell-fire. The Matthean apocalyptic scenario, here appearing for the first time in the Gospel, will be repeated at regular intervals. The pattern in this scenario is clear: Seventy (Luke 10.1–12). On this scenario, Matthew must have seen ‘that town’ in Luke 9.5 and realized that he could make it work in context by drawing in the references to the town in Luke 10.8–11. 43. Michael D. Goulder, Midrash and Lection in Matthew (London: SPCK, 1974), 79; but see my critique in Goulder and the Gospels: An Examination of a New Paradigm, JSNTSup 133 (Sheffield:  Sheffield Academic Press, 1996), 48–9 and 83. Goulder over-stressed distinctive Matthean vocabulary elements, but he is on stronger ground in relation to rhythm, imagery and thought.

11. Taking our Leave of Mark–Q Overlaps

221

(1) There is a stark contrast. In every case, there is a black-and-white contrast, in which the good are distinguished from the wicked, usually using harvest imagery. Here in John’s speech, there is ‘wheat’ and ‘chaff ’ (Matt 3.12 // Luke 3.17); later, there are ‘good trees’ and ‘bad trees’ (Matt 7.15–20 // Luke 6.43–45); there are ‘wheat’ and ‘tares’ (Matt 13.24–30; 36–43) and fish that are either good or bad (Matt 13.48), virgins who are either ‘wise’ or ‘foolish’ (Matt 24.1-13) and ‘sheep’ and ‘goats’ (Matt 25.31–46). (2) There is a demand for good fruits. Here (Matt 3.8–10 // Luke 3.8–9), the challenge is to ‘bear fruits worthy of repentance’ (3.8), leading to the metaphor of people as trees (3.9) who produce good or bad fruit. The metaphor is developed with good trees and bad trees subsequently in Matt 7.15–20 // Luke 6.43–5 and Matt 12.33. (3) There is a separation at the harvest. ‘The angels’ (Matt 13.34, 41, 49, cf. 25.31) separate ‘the wicked from among the righteous’ (Matt 13.48–9) at the end of the age, symbolized as a harvest when wheat is separated from chaff (Matt 3.9 // Luke 3.9), or wheat from tares (Matt 13.24–30; 36–43), or good fish from bad fish (13.49), or sheep from goats (Matt 25.32). (4) There is an eternal fire where the evil are burned. The chaff, the tares, the bad fish, the bad trees are all thrown into the eternal, unquenchable fire (Matt 3.10, 12; 7.19, 13.30, 13.42, 13.50). It is a view of hell that coheres with Matthew’s characteristic stress on ‘weeping and gnashing of teeth’ (Matt 8.12 // Luke 13.28, Matt 13.42, 50; 22.13; 24.51; 25.30). Several of these motifs do, of course, also appear in Luke,44 but it is Matthew who regularly clusters them together, as when in Matt 12.34, the ‘brood of vipers’ invective is again combined with teaching about trees and their fruits (Matt 12.33) alongside the day of judgement (Matt 12.36). The apocalyptic scenario, with its clusters of related themes, with regular use of harvest imagery and fiery judgement are clearly characteristic of Matthew, and it suggests that the direction of dependence in the Mark–Q overlap material in Matt 3 and Luke 3 goes from Matthew to Luke.

11.6 Conclusion The SP is often perceived as unimportant to the wider task of discussing the history and theology of early Christian texts. As a complex and obscure aspect of the SP, discussion of the Mark–Q overlaps can seem particularly abstruse. The contributors to this volume do not, of course, share such a perception, but it is worth thinking about how we draw attention to the importance of the topic to the wider field. One area where Mark–Q overlaps have major importance is in historical

44.  Luke’s one major exploration of the afterlife (Luke 16.19–31) also features a fiery Hades, but it is worth noting that Luke’s eschatological reversal is here typically not so much about the evil and the righteous as the rich and the poor (cf. Case Against Q, 136–8).

222

Mark Goodacre

Jesus research. The role played by the criterion of multiple attestation in the work of many historical Jesus scholars,45 alongside an acceptance of the Two-Source Theory, means that traditions apparently attested in both Mark and Q become especially important and, as a result, passages attributed to Mark–Q overlap attain a significance that may be out of proportion to their historical value.46 One of the reasons commonly given for treating Jesus’ baptism as historical is that it occurs in both Mark and Q. But if Luke knew Matthew as well as Mark, those agreements are simply the result of Luke’s taking over several of Matthew’s modifications of Mark, and not evidence of an independent attestation in another source, a conclusion that will be disappointing to many historical Jesus scholars. And if Luke knew Matthew, then it becomes less clear that John’s apocalyptic preaching derives from a parallel source that is independent of Mark. It may simply be Matthew’s characteristic expansion of his Marcan source, drawing on the same stock of imagery that he uses also for Jesus’ preaching, with the brood of vipers, the wheat and the chaff and the eschatological separation and fiery judgement, a conclusion that will be a disappointment to those who have invested a great deal in reconstructing the historical John the Baptist. The inconvenience of a possible conclusion should not, however, deter us from asking the difficult questions. Reflection on the role played by Mark–Q overlaps may encourage interesting discussion on a range of topics in scholarship on Christian origins, not least the exploration of the way one literary work appropriates, expands and reworks another. The more limited purpose of this chapter, though, has been to draw attention to the difficulties that they pose for the Two-Source Theory and to suggest that the Farrer theory explains the data better. These passages are problematic not because overlapping sources are implausible but because they contradict one of the primary reasons for accepting the Q hypothesis in the first place. It is said that Matthew and Luke ‘never’ agree with one another in substantial ways against Mark, but these passages illustrate that this is not true. Along with the socalled minor agreements, they are part of a broader spectrum of evidence according to which Luke sometimes agrees with Matthew in minor ways and sometimes in major ways. They are what we would expect to find if indeed Luke knew Matthew as well as Mark. Perhaps it is time to take our leave of Mark–Q overlaps, to acknowledge these major agreements, and to think again about dispensing with Q. 45.  For a critique of this criterion, including the role Mark-Q overlaps play in its practice, see Mark Goodacre, ‘Criticizing the Criterion of Multiple Attestation: The Historical Jesus and the Question of Sources’, in Jesus, Criteria, and the Demise of Authenticity, ed. Chris Keith and Anthony Le Donne (London: T&T Clark, 2012), 152–69. 46.  Cf. Benedict T. Viviano, ‘The Historical Jesus in the Doubly Attested Sayings: An Experiment’, RB 103/3 (1996):  367–410, ‘If Q and Mark are indeed independent of each other, the cases where they echo the same saying or phenomenon take on high significance as the most solidly grounded early sources available’ (367). The appeal to the double attestation of Mark and Q goes back to F. Crawford Burkitt, The Gospel History and Its Transmission (Edinburgh:  T&T Clark, 1906); see especially 147. Burkitt did not use the term Mark-Q overlaps but his list is far larger than Streeter’s, e.g. it includes the Parable of the Sower (152).

Chapter 12 T H E G O SP E L O F L U K E A S NA R R AT O L O G IC A L I M P R OV E M E N T O F S Y N O P T IC P R E - T E X T S THE N A R R AT I V E I NTRODUCTION TO T H E J E SU S S TORY ( M AR K 1.1– 8   P A RR . ) Werner Kahl

12.1 Introduction In its 180 years of fictional existence, the career of the Logienquelle Q has undergone remarkable transformations:  In the first half of the nineteenth century it was described first as a source of Jesus’ utterances, not of his deeds. As such it was a welcome ‘discovery’ to critically react to the challenges posed by Hermann Samuel Reimarus1 and later by David Friedrich Strauss,2 according to whom the Gospels presented a mixture of unreliable historical data, lies, myth and true words of Jesus. Of course, and unfortunately, Q was not excavated like many other manuscripts in the Near East during the course of the nineteenth century.3 It was rather uncovered – due to a particular, and problematic, interpretation of the reference to τὰ λόγια with respect to Matthew found in a literary fragment of Papias’ writings as quoted in Eusebios’ church history4  – from beneath the

1.  Gotthold Ephraim Lessing (ed.), Fragmente des Wolfenbüttelschen Ungenannten (4th edn; Berlin: Sandersche Buchhandlung, 1835 [orig.: 1774–78]). 2.  David Friedrich Strauss, Das Leben Jesu, kritisch bearbeitet (2 vols; Tübingen: C. F. Osiander, 1835/1836). Cf. also David Friedrich Strauss, Hermann Samuel Reimarus und seine Schutzschrift für die vernünftigen Verehrer Gottes (Leipzig: Brockhaus, 1862). 3. Cf. John S. Kloppenborg, Excavating Q: The History and Setting of the Sayings Gospel (Philadelphia: Fortress Press, 2000). 4. It was Friedrich Schleiermacher who suggested in 1832 that τὰ λόγια in Eusebios’ quotation of Papias in the former’s History of the Church (III 39) should refer to a collection of ‘sayings and speeches’ (of Jesus), cf. his ‘Ueber die Zeugnisse des Papias von unsern beiden ersten Evangelien’, in Theologische Studien und Kritiken (ThStKr) 4 (1832): 735–68. This suggestion was accepted before Christian Hermann Weisse by Karl August Credner in his Einleitung in das Neue Testament (Halle: Verlag der Buchhandlung

224

Werner Kahl

existing Synoptic Gospels. Christian Hermann Weisse’s ‘solution’ to the synoptic problem in the form of the so-called Two-Source Theory (2ST) in 1838 was not the result of an objective investigation. It came as a direct response to Strauss’ Leben Jesu, which he critically appreciated.5 The assumed Sayings Source provided a recourse to what increasingly was considered the essentials of the historical Jesus, that is, his ethical and wisdom teachings – and not his miracles.6 The quest for a source behind received texts was, of course, part of the general program in classical philology of the nineteenth century to recover originals in order to proceed to unmutilated truth. The emphasis on the teachings of Jesus at the expense of narratives about events that were rather embarrassing to the modern mind proved to be extremely persuasive for liberal theologians at the turn of the century. Especially after William Wrede had made it likely that the Gospel of Mark was a historically untrustworthy document,7 Adolf von Harnack

des Waisenhauses, 1836), 91. Based on Schleiermacher’s interpretation of the Papias fragment, according to which both our Gospel of Mark and our Gospel of Matthew presupposed more original versions (Proto-Mark and Matthew’s λόγια), Credner developed a rather complex scenario of the development of the synoptic gospels that could serve as a basis for the 2ST as well as for the Synoptic Improvement Model (cf. below 12.3.1): Luke used, like Matthew, the λόγια (in ‘Hebrew’) and Greek Proto-Mark; in addition, Luke also knew other sources and traditions which probably also included both the Gospel of Matthew and the Gospel of Mark, cf. Einleitung, 203–6. In research on the history of the synoptic problem, the significance of Credner’s contribution – especially as a link between the proposals of the late eighteenth and the nineteenth century – has generally gone unnoticed. The interpretation of τὰ λόγια as ‘sayings’ or ‘speeches’ is, however, due to a forced argument. Internal and external evidence rather suggests that τὰ λόγια in the Papias fragment refers to general Jesus traditions, cf. Gerhard Kittel, λόγιον im Gebrauch der Kirche, in Theologisches Wörterbuch zum Neuen Testament (ThWNT) 2 (1942): 143–5. 5.  Christian Hermann Weisse, Die evangelische Geschichte kritisch und philosophisch bearbeitet (Leipzig:  Breitkopf und Härtel, 1838), vol. I, iv and 8. See the overview of the development of the 2ST in Werner Kahl, ‘Vom Ende der Zweiquellentheorie oder:  Zur Klärung des synoptischen Problems’, in Kontexte der Schrift. Band II:  Kultur, Politik, Religion, Sprache. FS Wolfgang Stegemann, ed. Christian Strecker (Stuttgart:  Kohlhammer 2005), 404–42. 6.  Reimarus had already drawn this distinction between the ‘practical teachings’ of Jesus, which he highly valued, and the miracle stories which he despised, cf. Hermann Samuel Reimarus, Apologie oder Schutzschrift für die vernünftigen Verehrer Gottes, vol. 2, ed. Gerhard Alexander (Frankfurt: Insel Verlag, 1972), 540–3. To Reimarus, the Sermon on the Mount is the epitome of Jesus’ teachings, cf. ibid., 30–6. 7.  William Wrede, Das Messiasgeheimnis in den Evangelien. Zugleich ein Beitrag zum Verständnis des Markusevangeliums (Göttingen: Vandenhoeck & Ruprecht, 1901).

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

225

could claim that the Sayings Source Q provided the most reliable access to the true teachings of Jesus.8 In the twentieth century New Testament scholars began, however, also to question the reliability of Q as directly reflecting the utterances of the historical Jesus, and eventually form and redaction criticism were also applied to this ‘source’. This procedure resulted in various scenarios of a diachronic development of Q until its final redaction.9 In contemporary scholarship on Q, the ‘Sayings Collection’ has mutated into a ‘Sayings Gospel’ or a ‘Semi-Gospel’ running structurally  – and at times with respect to content – parallel to the Gospel of Mark (esp. at the beginning) while not including a passion narrative.10 This parallelism is, in my understanding, not surprising but, on the contrary, to be expected due to the method of constructing Q from material shared by Matthew and Luke over against Mark, since all three Gospels follow more or less the same order. If Matthew and Luke were structurally as different from each other and compared to Mark as, for instance, John and the Gos. Thom., one would also not find a parallel structure between Mark and Q. Not only narrative sections in Q – including miracle traditions11 – have recently received considerable attention, but even the whole of Q has been described as a narrative.12 In addition, it has been claimed that the ‘gaps’ of this narrative – and there are plenty of gaps in Q – were intentional in order to involve the interpretive creativity of the recipients to fill those gaps.13 8.  Adolf von Harnack, Sprüche und Reden Jesu. Die zweite Quelle des Matthäus und Lukas (Leipzig: J. C. Hinrichs’sche Buchhandlung, 1907), 173: ‘Wer ist wertvoller? (. . .) Die Spruchsammlung und Markus müssen in Kraft bleiben, aber jene steht voran. Vor allem wird die Übertreibung des apokalyptisch-eschatologischen Elements in der Verkündigung Jesu und die Zurückstellung der rein religiösen und moralischen Momente hinter jenes immer wieder ihre Widerlegung durch die Spruchsammlung finden. Sie bietet die Gewähr für das, was in der Verkündigung Jesu die Hauptsache gewesen ist: die Gotteserkenntnis und die Moral zu Buße und Glauben, zum Verzicht auf die Welt und zum Gewinn des Himmels – nichts anderes.’ 9. See among many others, Kloppenborg, Excavating Q. 10.  Cf. Christoph Heil, ‘Einleitung’, in Die Spruchquelle Q.  Studienausgabe Griechisch und Deutsch, ed. Paul Hoffmann and Christoph Heil (Darmstadt:  Wissenschaftliche Buchgesellschaft, 2002), 11–28, 17–19. 11.  Cf. Martin Hüneburg, Jesus als Wundertäter in der Logienquelle. Ein Beitrag zur Christologie von Q. Arbeiten zur Bibel und ihrer Geschichte 4 (Leipzig:  Evangelische Verlagsanstalt, 2001). 12.  Harry T. Fleddermann, ‘The Narrative of Jesus as the Narrative of God in Q’, in Dieter T. Roth, Ruben Zimmermann and Michael Labahn (eds), Metaphor, Narrative, and Parables in Q, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 315 (Tübingen: Mohr Siebeck, 2014), 141–62, 142. 13.  Michael Labahn, ‘Was “Lücken” berichten: Exemplarische Beobachtungen zu narrativen “gaps” in Q’, in Roth, Zimmermann and Labahn, Metaphor, 163–88, 165.

226

Werner Kahl

This latest twist in the study of Q is informed by the application of insights and methods in literary studies, especially narratology and textual semiotics. As laudable as the reception of literary studies is for tackling the synoptic problem, the concrete procedures of application to the discourse evolving around the 2ST are far from satisfying. This holds true in at least three respects: 1. Eclecticism: Q scholars tend to connect to only those insights in literary studies that seem to avail themselves to Q studies. At the same time, insights contradicting the Q project in, at times, fundamental ways are glossed over. For example and foremost, a basic axiom in Editionswissenschaft – the discipline of critically establishing texts whose originals are lost – has been ignored by the International Q Project (IQP): A critical edition and reconstruction of a text for which there is no direct manuscript evidence (Hauptüberlieferung) but which would have to rely only on quotations or paraphrases (Nebenüberlieferung) is academically untenable.14 2. Application: In order to avail the Q text to methods developed in literary studies with the intention to demonstrate the narrativity of Q, standards of literary criticism tend to be ignored. For example, with respect to Michael Labahn’s suggestion of the intentionality of gaps in Q, Wolfgang Iser’s Reception Aesthetics in general and his convincing explanation of the involvement of readers by means of narratological gaps in particular, have been developed and tested with ‘real’ textual compositions.15 Fictional Q is marked less by narratological gaps than by broad textual gaps. Another example is Harry Fleddermann’s claim that ‘Q is a full, complete narrative’:16 In his attempt to make Q plausible as a narrative, one of Fleddermann’s arguments concerns the beginning of Q.  It runs as follows:  Since ‘a narrative must have an initial situation followed by a sequence of events that leads up to a reversal or change’ and since ‘John’s Preaching serves as the initial situation’17 we have in John’s preaching the beginning of the Q narrative. From the perspective of narratology and structuralism this is a surprising argument since preaching does not constitute a ‘situation’. According to narratological theory, preaching is clearly a performance aiming at a transfer of savoire-faire, to put it in terms of Algirdas Greimas’ universal narrative schema.18 Such a rather unwary application 14.  Werner Kahl, ‘Erhebliche matthäisch-lukanische Übereinstimmungen gegen das Markusevangelium in der Triple-Tradition  – ein Beitrag zur Klärung der synoptischen Abhängigkeitsverhältnisse’, Zeitschrift für die neutestamentliche Wissenschaft (ZNW) 103 (2012): 20–46, 21–4. 15.  Cf. Wolfgang Iser, Der implizite Leser. Kommunikationsformen von Bunyan bis Beckett (München: Wilhelm Fink, 1972); Wolfgang Iser, Der Akt des Lesens. Theorie ästhetischer Wirkung (München: Wilhelm Fink, 1976). 16. Fleddermann, ‘Narrative of Jesus’, 142. 17. Fleddermann, ‘Narrative of Jesus’, 142. 18. See below 12.3.2.

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

227

of narratological insights to the field of Q studies does not help the cause of Q but might rather be another indication for the fundamental flaw of the Q hypothesis. 3. High degree of academic uncertainty: The textual basis for the application of complex narratological and semiotic approaches within Q studies is too unreliable. This uncertain textual foundation of Q, combined with the complexity of the applied methods informed by literary criticism and with unsound procedures of application, generates highly uncertain and speculative results which can hardly be verified or falsified or made probable by any plausible academic standards.19

12.2 Q’s Parallel Beginning to Mark’s Gospel Mark’s ‘Good news about Jesus Christ’ begins with a narrative about John the Baptist. Both Matthew and Luke include a prelude about Jesus’ conception, birth and early life, before they get to the narrative about the Baptist’s proclamation. Matthew in ch. 2 puts the focus on (the divine preservation of the life of) Jesus alone without mentioning the Baptist. Luke in chs 1 and 2 interweaves the birth narratives of Jesus and John by means of which the relationship between the two figures and the function of John within the wider program of God’s plan with humankind through Jesus are established. By means of this procedure the following scenario, as narrated in Mark 1.1–8, appears to be more sufficiently prepared in the Gospel of Luke than in the Gospel of Matthew. Luke’s version of the infancy story also seems to correspond well with his claim to write down all events καθεξῆς, that is, in the correct sequence or in chronological order (Luke 1.3). According to the so-called Critical Edition of Q (CEQ), Q, like Mark, also began with a reference to John  – however, not necessarily with a narrative but with utterances in the form of a judgement proclamation. Andreas Lindemann is not the first to find it ‘at least astonishing’ that a collection of Jesus’ sayings ‘would begin with a longer discourse of another speaker, who has not even been introduced to the readers’.20 Already to the founder of the 2ST, Weisse, this constellation 19.  Even though the works of Labahn and Fleddermann may not be considered representative of Q studies, there is a recent trend that is less concerned with the fact of – the lack of – a precise reconstruction of the wording of Q and more with the exploration of meta-structures which in turn are regarded as supporting the ‘factuality of Q’; see Ruben Zimmermann, ‘Metaphorology and Narratology in Q Exegesis: Literary Methodology as an Aid to Understanding the Q Text’, in Roth, Zimmermann and Labahn, Metaphor, 3–30, 5; see my review of this collection of new perspectives on Q, in Theologische Literaturzeitung (ThLZ) 7 (2016): 776–8. 20. Andreas Lindemann, ‘Die Logienquelle Q. Fragen an eine gut begründete Hypothese’, in The Sayings Source Q and the Historical Jesus, ed. Andreas Lindemann, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 158 (Leuven: Peeters, 2001), 3–26, 5. Cf. the Coptic Gospel of Thomas which, after a short introduction, begins with sayings of Jesus.

228

Werner Kahl

seemed ‘enigmatic’.21 Lindemann therefore begins to ask ‘questions with respect to a well founded hypothesis’.22 His strategy to overcome the difficulty is typical of the treatment of problems within the discourse of the 2ST: He suggests the existence of further hypothetical sources. John’s judgement proclamation could have been derived by both Matthew and Luke independently, from a ‘singular tradition’ (Einzelüberlieferung). Also the narrative about the temptation of Jesus in its enlarged form in Matthew and Luke could not have belonged to Q since it is closely intertwined with the narrative of Jesus’ baptism by John which, according to Lindemann, was also missing in Q. The longer temptation narrative is then also attributed to another ‘particular tradition’. Therefore, Q would have begun with the sayings’ material of Jesus now to be found in Matthew’s ‘Sermon on the Mount’ and Luke’s ‘Sermon on the Plain’. This is the traditional conception of the beginning of Q as proposed initially by Schleiermacher in 1832.23 Notwithstanding the increase in hypotheses, Lindemann is content that the 2ST ‘still provides the best and simplest explanation for the synoptic tradition with respect to the whole textual evidence’.24 This claim is not convincing since Lindemann fails to compare the potency of the 2ST with that of other explanatory models with respect to the passage in question and to the whole synoptic tradition. The very beginning of the synoptic tradition especially poses severe challenges to any version of the 2ST.25 The difficulty of squaring the textual evidence of Mark 1.1–8 parr. with any 2ST had already caused changes of position in the case of the two founders of the theory. Both Weisse and Holtzmann, who each favoured different conceptions of the 2ST, in their original monographs on the synoptic

21. Weisse, Die evangelische Geschichte, vol. 2, 5. 22. See the subtitle of Lindemann’s article. 23.  Schleiermacher, Zeugnisse des Papias, 752:  ‘Außer diesem Hauptstück aber, vor Kap.  5 und nach Kap.  25, glaube ich nicht, daß wir Bestandteile unserer apostolischen Sammlung zu erwarten haben.’ According to Schleiermacher, the reason for this suggestion is that the source in question only contained speeches of Christ (752: ‘die nur Reden Christi zusammenstellen wollte’). Cf. also Credner, Einleitung, 204, who suspects that – in addition to the nine chapters of Matthew proposed by Schleiermacher (5–7; 10; 13; 18; 23–25) – Matt 4.23–25 too was taken from the sayings source. 24.  Lindemann, ‘Logienquelle’, 3.  In so doing, Lindemann  – like many proponents of the 2ST regularly do  – repeats a conviction that was first propounded by Heinrich Holtzmann, ‘Umschau auf dem Gebiete der neutestamentlichen Kritik. I. Die Evangelien’, in Jahrbücher für Protestantische Theologie (1875) 1, 583–635, here 586: ‘Thatsache ist (. . .) dass sich jene Lösung der synoptischen Frage wie durch die Stütze des Papias-Zeugnisses, so auch durch ihre innere Einfachheit und Durchsichtigkeit in immer weiteren Kreisen empfohlen hat.’ 25. The explanations of the Matthean and Lukan versions of Mark 1.1–8 within the 2ST discourse are as manifold as they are contradictory, cf. Michael Wolter, Das Lukasevangelium, Handbuch zum Neuen Testament (HNT) 5 (Tübingen: Mohr Siebeck, 2008), 154, and – with respect to the common treatment by Matt and Luke of the quotation in Mark 1.2 – his perplexed concession: ‘Eine eindeutige Erklärung für diesen Sachverhalt gibt es nicht . . .’ (157).

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

229

problem presented distinct interpretations of this passage that they later rejected. According to Weisse, the sayings ascribed by both Matthew and Luke to the Baptist (Matt 3.7–10 / Luke 3.7–9; Matt 3.11b–12 / Luke 3.16b–17) were found in the source of the apostle Matthew (‘Q’). Since he had defined – with Schleiermacher – that source as a collection of sayings of Jesus, he concludes that the utterances in question were originally sayings of Jesus put into the mouth of the Baptist in order to emphasize the inner connection of the two.26 Less than twenty years after this explanation, in 1856, Weisse concedes that the challenges of this passage – together with a few others – had forced him to a single correction of his previous version of the 2ST. Now he attributes the pronouncements of the Baptist to an original version of the Gospel of Mark, now lost – a theory that also goes back to Schleiermacher’s article on Papias. In his earlier work, this part of Schleiermacher’s interpretation of the Papias fragment was rejected by Weisse.27 A few years after the revision of the position of Weisse, Holtzmann, in his major contribution to the subject from 1863, designed a 2ST that presupposes the distinction of our Gospel of Mark from an original Gospel of Mark  – his ‘Apostolische Quelle’ or ‘Urmarcus’. This original – and now lost – source was the basis of the Gospels of Mark, Matthew and Luke. The Matthean-Lukan exclusive agreements with respect to Mark 1.1–8 are attributed to this lost Markan source. Mark simply did not copy the verses in question for his surviving Gospel.28 Holtzmann even gives Mark’s motivation for each of his narrative moves vis à vis the original Gospel. Fifteen years later, Holtzmann began to consider a possibility of explaining the ‘minor agreements’ which he formerly categorically rejected  – namely, that Luke and Matthew did not write completely independently of each other.29 In fact, the passage under consideration here cannot easily be accounted for synoptically on the basis of the generally held presumption that, on the one hand, Matthew and Luke did not use one another’s work, and, on the other, that the ‘minor agreements’ were due to the use of an original Gospel of Mark by Matthew and Luke.30 Eduard Simons then tried in 1880 to

26. Weisse, Die evangelische Geschichte, vol. 2, 8–9. 27. Weisse, Die Evangelienfrage in ihrem gegenwärtigen Stadium (Leipzig: Breitkopf und Härtel, 1856), 156–7. 28. Holtzmann, Die Synoptischen Evangelien, 67–8. 29.  Cf. Holtzmann, ‘Zur synoptischen Frage. II. Die Stellung Jesu zum Gesetz’, in Jahrbücher für Protestantische Theologie 4 (1878), 533–68, here 553–4; Review by G. Meyer, La Question Synoptique, in ThLZ 3 (1878): 553–4, here 553. 30. The fundamental decision of Matthew-Lukan independency was simply – and uncritically – presupposed ‘philosophically’ as a given fact by the founder of the 2ST, Christian Weisse, Die evangelische Geschichte, vol. 2, 8:  ‘Eine Benutzung des einen dieser beiden Werke durch das andere anzunehmen, haben wir uns in unserer allgemeinen Betrachtung über den Charakter der Evangelien nicht bewogen finden können.’ For later theologians, it seemed inconceivable that Luke could have ‘destroyed’ the beautiful speech compositions of Jesus found in the Gospel of Matthew. This, however, is simply an esthetic argument which is historically unfounded, as Holtzmann concedes, cf. his review of Eduard Simons, ‘Hat der dritte Evangelist den kanonischen Matthäus benutzt?’ in ThLZ 8 (1881): 180–3, here 181.

230

Werner Kahl

test the synoptic relationships from this new perspective proposed by Holtzmann, that is, presupposing Luke’s use of the canonical Gospels of Mark and Matthew and of Q.31 In his analysis of Mark 1.1–8 parr., Simons explains the ‘minor agreements’ almost completely with Luke’s knowledge of both the canonical Gospel of Mark and of Matthew, here without any recourse to ‘Q’ – and I think he does so convincingly.32 In his review of Simons’ work, Holtzmann fully agrees with his student’s solution of the synoptic problem.33 And in his Lehrbuch der historischkritischen Einleitung in das Neue Testament from 1885, Holtzmann clearly states his change of mind in this regard.34

12.3 Methodological Considerations 12.3.1 The Synoptic Improvement Model In my assessment, the 2ST does not provide the best explanation for most synoptic problems. It is neither the simplest nor the most convincing attempt to account for the similarities and differences between the Synoptic Gospels. The fewest difficulties of any attempt to make sense of these problems seem to occur in what I call the Synoptic Improvement Model (SIM).35 According to this theory most synoptic similarities and differences can be made plausible without presupposing any hypothetical sources, first by comparing Lukan and Matthean tendencies in extending, rewriting and refocusing the Gospel of Mark, with Luke also using the Gospel of Matthew, and second by taking into account typical narratological and theological features.36 It is the advantage of SIM that it presupposes 31.  Eduard Simons, Hat der dritte Evangelist den kanonischen Matthäus benutzt? (Bonn: Emil Strauss, 1880). 32. Simons, Evangelist, 21–5. 33. ThLZ 8 (1881): 180–3. 34.  Heinrich Holtzmann, Lehrbuch der historisch-kritischen Einleitung in das Neue Testament (Freiburg i.Br.:  Mohr, 1885), 339. In the second edition of this work from 1886, Holtzmann added a lenghty section in which he argues for the plausibility of Simons’ solution (363–5). 35. I propose the name Synoptic Improvement Model for the model presupposed here. It is a critical adaption of the model elaborated by Michael Goulder and identical with the model preferred by Mark Goodacre and Francis Watson, the latter refers to it as L/M theory, cf. below n. 36. 36. Following his teacher, Austen M. Farrer, Michael Goulder has described and applied the theory that Luke used both the Gospel of Mark and the Gospel of Matthew in detail in his commentary Luke. A New Paradigm, Journal for the Study of the New Testament Supplement Series (JSNTSup) 20 (Sheffield: Academic Press, 1989). A more balanced version of the theory has been proposed by Mark S. Goodacre, Goulder and the Gospels. An Examination of a New Paradigm, JSNTSup 133 (Sheffield:  Academic Press, 1996); Mark S. Goodacre, The Synoptic Problem. A Way Through the Maze (London: T&T Clark, 2004); Mark S. Goodacre

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

231

texts that have approximately been established by careful text-critical procedures.37 Proposed explanations of synoptic problems from the perspective of SIM can be checked for plausibility with reference to the interpretation of texts that are available in an identical form – the most recent Nestle–Aland edition – to all exegetes. The 2ST, however, depends on the plausibility of at least one hypothetical text – Q – which has been constructed in academically questionable ways, as indicated above. In practice, however, most proponents of the 2ST introduce additional hypothetical texts whenever a textual feature cannot be easily explained by reference to Q: a form of Deutero-Mark, single traditions, Q in various stages of its assumed history, and so on. Taking recourse to hypothetical sources and increasing their number is the great disadvantage of the 2ST. As a matter of fact, right from its general endorsement in New Testament scholarship that began in the 1860s, there has not been a single 2ST but a whole variety of versions of the 2ST depending on the particular choice of additional hypotheses.38 Notwithstanding this multiplicity of – contradictory! – versions of the 2ST, it has generally gained the status of an exegetical axiom not to be questioned. This holds true especially for the German exegetical tradition. Here the 2ST has often been treated as beyond criticism.39 The unquestioned authority attributed to the 2ST becomes understandable if one considers the theological functions this theory fulfilled: From the very beginning of its inception, solutions of the synoptic problem that were interested in hypothetical sources behind the existing Gospels reacted to fundamental attacks on the historical trustworthiness of the Gospel accounts. With Q it seemed possible to establish a theology grounded in the teachings and the character of Jesus himself, bypassing or at least minimizing the all-too-embarrassing mythological or miraculous

and Norman Perrin (eds), Questioning Q:  A Multidimensional Critique (London:  IVP Academic, 2004); Kahl, ‘Vom Ende’; Werner Kahl, ‘Erhebliche matthäisch-lukanische Übereinstimmungen gegen das Markusevangelium in der Triple-Tradition  – ein Beitrag zur Klärung der synoptischen Abhängigkeitsverhältnisse’, ZNW 103 (2012):  20–46; Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI: Erdmanns, 2013). 37. It should be noted that any synoptic theory has to take into account the relative uncertainty with respect to the exact wording of the synoptic texts. Comparing SIM with the 2ST, this problematic appears much aggravated and potentized in the latter due to the – academically untenable – procedure of establishing the Q text from a comparison of the Gospels of Matthew and Luke, the result of which is by necessity a completely uncertain text. 38. For an overview of various additional hypotheses within the 2ST discourse, cf. Kahl, ‘Übereinstimmungen’, 25–6. 39.  Cf., however, the balanced treatment of alternative explanations of the synoptic problem by Martin Ebner, ‘Die synoptische Frage’, in Einleitung in das Neue Testament, ed. Martin Ebner and Stefan Schreiber, Kohlhammer Studienbücher Theologie (KStTh) 6 (Stuttgart:  Kohlhammer, 2008), 67–84. This might also be an indication that the 2ST is about to lose its monolithic authority in Germany.

232

Werner Kahl

features of the existing Gospels.40 It is no coincidence that the enlightened critique of miracles and the search for a hypothetical sayings source converged historically in the second half of the eighteenth and in the beginning of the nineteenth century. In the past decades, a new and more positive philosophical and theological assessment of both myths and miracles has gained ground.41 At the same time, the focus in literary studies has shifted away from the establishment of sources behind texts in order to gain truths in ‘originals’ to academically controlled readings of texts in their final form. In addition, a new appreciation of variation and difference has emerged in cultural studies due to the influence of postmodern philosophies and postcolonial studies. The attempts at creating ‘Q’ are still an expression of the paradigm of harmonizing the Gospels – not on a synchronic level, as was the rule in theology until around 1800, but diachronically.42 It is in this context that SIM has increasingly become plausible while the 2ST has lost credibility since it does not seem to be able to solve synoptic problems. The 2ST rather creates unnecessary problems.43 Taking recourse to a highly speculative and methodologically problematic procedure in establishing Q and then working with this hypothetical source in explaining synoptic relations should be a last resort only to be engaged if other, less hypothetical and less problematic procedures, which rely on received texts, fail. This scenario, however, is not given due to the explanations of synoptic relationships possible with SIM. Only if these explanations prove to be too speculative and implausible or more forced than those in the 2ST discourse should models which propose hypothetical sources be engaged. John Kloppenborg, in his contribution to this volume, has drawn our attention to the problematic procedure of simply ‘renaming the problem’ when referring to

40. The two monographs of the nineteenth century in which the 2ST was developed each contain a discussion of the miracles of Jesus according to the synoptic tradition: Weisse, Evangelische Geschichte, vol. 1, 335–76; Holtzmann, Die Synoptischen Evangelien, 497–514. 41. Cf. Kurt Hübner, Die Wahrheit des Mythos (München: Beck, 1985); Stefan Alkier and Annette Weissenrieder (eds), Miracles Revisited. New Testament Miracle Stories and their Concepts of Reality, Studies of the Bible and its Reception (SBR) 2 (Berlin: de Gruyter, 2013). 42.  I owe this observation to Stefan Alkier, cf. also his contribution to the present volume, where he points out the high value of tolerance among philosophers and theologians of the eighteenth century. 43. For example, the problem of the evidence of the existence of Q: If Q had been used independently from one another by Matthew and Luke in very different locations, it seems highly unlikely that such an important source of the very words of Jesus, who was considered Lord, was lost and not mentioned in any document of antiquity. By contrast, the Gospel of Mark, which was considered problematic and insufficient at the beginning of the second century (cf. the Papias fragment and possibly Luke 1.1–4) had survived even though it was almost completely absorbed by both the Gospels of Matthew and Luke. Cf. Bartosc Adamczewski, Q or not Q? The So-Called Triple, Double, and Single Traditions in the Synoptic Gospels (Frankfurt a.M.: Peter Lang, 2010), 83–4.

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

233

what he calls ‘aesthetic and ad hoc compositional arguments’.44 And he has also shown that ‘coherence arguments can be invoked in support of mutually contradictory theories.’45 Both arguments are to be taken seriously in research on the synoptic problem; not only with respect to SIM but also – and even more so – in the discourse on the 2ST. On the part of Michael Goulder, for example, it is clearly to the disadvantage of his proposal when, in his commentary on Luke, he tries to explain every single narrative move of Luke over against the Gospel of Matthew. Here Goulder overstretches the point. Often – and by necessity – his explanations are highly speculative and therefore counterproductive. Due to the following circumstances, we should be aware of the limitations of our possible understanding of Luke’s mind:46 1.  Apart from the technical information given to us in Luke’s prologue, we do not know any further details about Luke’s intentions in rewriting his pre-texts. The intentions and strategies of the Gospel writers can only tentatively be inferred, especially if we are able to compare a given Gospel with its source(s). 2.  The Gospel writers conceptualized and organized the world in general with a world-knowledge-system or encyclopaedia widespread in Mediterranean antiquity –while presupposing God’s interaction with the world in and through Christ as a given fact – very differently from intellectuals in the Western world of the twenty-first century. 3.  The Synoptic Gospels and all prevailing early Christian writings represent only a fraction of the written and oral communication between Christ believers in the first century. 4. We do not know anything precise about the Gospel writers: their addressees, their locations, the exact time of their writing and the particular situation in which they wrote. Our knowledge about the Synoptic Gospel writers is extremely sparse, and our understanding of their motives and of their writings is limited. Studying and comparing the different Gospels, however, allows us to identify particular emphases of each Gospel account. These observations can be either verified or falsified or made more or less plausible with reference to texts that we all have in front of us in the most recent Nestle–Aland edition.47 If Luke used the Gospel of Mark – which I presuppose alongside both SIM and the 2ST – then changes that function as narratological or theological improvements regarding his pre-text were to be expected, unless the Lukan prologue had exclusively rhetorical functions, which I doubt. And if the differences between the Gospel of Luke and the Gospel of Matthew could generally also be accounted for in terms of Lukan attempts of an improvement of the Gospel of Matthew – and

44. Both citations Kloppenborg, ‘Conceptual Stakes’, 13–42. 45. Kloppenborg, ‘Conceptual Stakes’, 13–42. 46. This applies even more to Matthew’s mind and most of all to Mark’s. 47. Cf. the apt remark by the ‘father’ of SIM, A. M. Farrer, ‘On Dispensing with Q’, in Studies in the Gospels: Essays in Memory of R. H. Lightfoot, ed. D. E. Nineham (Oxford: Blackwell, 1955), 55–88, here 66: ‘The Q hypothesis is a hypothesis, that is its weakness. (. . .) Now St. Luke’s book is not a hypothetical entity. Here is a copy of it on my desk.’

234

Werner Kahl

not the other way round48 – then SIM stands on sufficiently firm ground. Therefore, this explanation of the synoptic interdependencies would be preferable to the 2ST with an assumed and constructed source Q. Also the 2ST works with arguments of coherence and Gospel tendencies, but it does so to determine the original wording of Q after a critical comparison of Matthean and Lukan writing strategies; once Q is thus established, the redactional activity of both Matthew and Luke is described with respect to Q! In this way, a hermeneutical circle has been completed and the tendencies of either Matthew or Luke as compared to their presumed pre-text Q cannot be verified or falsified. Therefore – as compared to SIM – the problems of the 2ST are of a higher degree, that is, they are more severe and increased. This also applies to the arguments put forth as explanation for the so-called minor agreements (MA), in an attempt to minimize their weight against the 2ST and to weaken SIM: 1. coincidental identical changes of the text of the Gospel of Mark by Matthew and Luke; 2. harmonizations of the Gospel of Matthew and the Gospel of Luke in the textual tradition; 3. harmonizations of the Gospel of Luke and the Gospel of Matthew by a second-century redaction.49 All these possibilities are generally plausible, the third to a much lesser degree due to the difficulties in establishing the text of Marcion’s Gospel. The first two points need to be taken into consideration in SIM and they serve as a reminder that we do not have the autographs in front of us with the Nestle–Aland edition. However, the possibility of later harmonizations also – and even more so – affects the 2ST with its reconstruction of Q. 12.3.2 Narratological Analysis The Universal Narrative Schema as developed by the Hungarian-French Semiotician A.  Greimas in the second half of the twentieth century might help to identify the narratological motivation behind variations in a later Gospel as compared to its pre-text(s). The development of the Narrative Schema was the result of a structuralist attempt to describe a universal deep structure of narratives in terms of the chronological organization of single narrative transformations in a particular narrative. It has been applied mainly in ethnology, in the study of

48.  As has been recently maintained by, among others, Robert K. MacEwen, Matthean Posteriority. An Exploration of Matthew’s Use of Mark and Luke as a Solution to the Synoptic Problem, The Library of New Testament Studies (LNTS) 501 (London:  Bloomsbury T&T Clark, 2015). My comparative analysis of the wording, style and syntax of the Gospels of Mark, Matthew and Luke indicates that Luke used both Mark and Matthew and that he intentionally tried to improve both his ‘Vorlagen’, see Kahl, ‘Übereinstimmungen’, 37–40. 49. See the new perspective on Marcion’s Gospel, according to which it derived – like the canonical Gospel of Luke – from a lost original Gospel of Luke; see for a detailed discussion Dieter T. Roth, The Text of Marcion’s Gospel, New Testament Tools, Studies and Documents (NTTSD) 49 (Leiden: Brill, 2015).

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

235

Table 12.1 The Narrative Schema A. Need

B. Preparedness

C. Performance

D. Sanction

A subject of a circumstance disjoined from a desirable object or conjoined with an undesirable object.

An active subject, willing or obliged (MANIPULATION), and able (having the power) to overcome the need (COMPETENCE), specified in A, by a performance.

The active subject performing the action to transform the circumstance specified in A into the opposite.

Recognition of the success or failure of the performance, or of the achievement of a desired value.

literature – especially fairy tales and myths – and in Biblical studies. It can also be applied to discourse analysis.50 The Narrative Schema promises to be particularly helpful against the background of Luke’s claim to present narrative traditions he received in a good order (Luke 1.1–3) with the didactic purpose to assure his addressee of the reliability of the content of the teachings (about the Gospel of Christ) he had received (Luke 1.4). By implication this means that, from Luke’s perspective, earlier accounts fell short of the narratological standards he set for himself (cf. with respect to the Gospel of Mark, the Papias fragment as it makes transparent that the reliability of this Gospel was obviously questioned until the time of Papias). Reading Mark 1.1–8 – the same applies of course for the whole Jesus Christ narrative in its variations as presented in the Synoptic Gospels – with the different stages of the Narrative Schema in mind helps to recognize chronological gaps, potentially irritating twists and inconsistencies in the Gospel of Mark as they also might have occurred to either Matthew or Luke or both of them. With the help of the Narrative Schema the possible intended functions of additions, omissions, translocations, word changes or other alterations of a later Gospel as compared to an earlier Gospel might become transparent.

50.  I adopted the Narrative Schema derived from Algirdas J.  Greimas’ reorganization (Algirdas J. Greimas and Joseph Courtés, Sémiotique. Dictionnaire raisonné de la théorie du langage [Paris:  Hachette, 1993], 244–7) of Vladimir Propp’s thirty-one narrative functions (Vladimir Propp, Morphology of the Folktale [2nd edn; Austin:  University of Texas Press, 1968]), which have been introduced into the field of New Testament studies by Hendrikus Boers (Hendrikus Boers, ‘Introduction’, in Wilhelm Egger, How to Read the New Testament. An Introduction to Linguistic and Historical-Crititical Methodology [Peabody : Hendrickson, 1996], xxxvii–lxix, lxv; Hendrikus Boers, Neither on This Mountain Nor in Jerusalem. A  Study of John 4 [Atlanta:  Society of Biblical Literature, 1988], 9–14), and I  applied it to the analysis and comparison of approximately 150 narratives of ‘miracle healing stories’ from the ancient Mediterranean (see Werner Kahl, New Testament Miracle Stories in Their Religious-Historical Setting:  A Religionsgeschichtliche Comparison from a Structural Perspective, Forschungen zur Religion und Literatur des Alten und Neuen Testaments [FRLANT] 163 [Göttingen: Vandenhoeck & Ruprecht, 1994], 44–62).

236

Werner Kahl

12.3.3 A Synoptic Analysis of Mark 1.1–8 // Matt 3.1–12 // Luke 3.1–20 In this chapter I will present my synoptic analysis of Mark 1.1–8 parr. As shown above, this passage at the very beginning of the synoptic tradition has proven extremely unwieldy for proponents of the 2ST. And it was a critical synoptic reading of these verses that initially caused me to question the appropriateness of the 2ST in general and the existence of Q in particular. I derived from my analysis the hypothesis that it might be sufficient to understand the synoptic interrelationships if we assume that the Gospel of Luke is the result of a critical integration of mainly two sources:  The Gospel of Matthew, which itself is the result of a rewriting of Mark, and the Gospel of Mark. After reading through the whole Greek Synopsis from this perspective, I arrived at the conclusion that the assumption of Q is an unnecessary complication in the attempt of understanding the synoptic interrelatedness. In addition, the history of the life-of-Jesus research in the nineteenth century indicated that the invention of ‘Q’ was theologically motivated. For a neutral comparison of the synoptic material it is essential to overcome the traditional arbitrary distinction between problematically so-called MA which are not regarded as belonging to Q, and larger agreements between the Gospel of Matthew and the Gospel of Luke in the material in common with the Gospel of Mark which would be attributed to Q.51 Instead I  have introduced the category exclusive agreements (EA) of any two against the third in any material (EAMt-Lk, EAMk-Lk, EAMt-Mk). I have distinguished them from inclusive agreements (IA), that is, agreements between all three Synoptics which cannot give any indication as to the direction of dependency: Luke, or for that matter Mark or Matthew if any of them was the writer of the youngest Gospel, could have derived identical matter from either of the other two Synoptics.52 In order to arrive at a tentative quantification of dependencies I have also suggested a count of words identical or similar in the Synoptics in a given episode. Exact identical words get the count number 1, similar words or phrases the count number 0.5. This count is a rough attempt to get closer to an objectification of the evidence and to indicate degrees of proximity and distance between the Synoptics with respect to a given pericope. In the following table (Table  12.2), EAMt-Lk are marked in bold, EAMk-Lk with single underlining, EAMt-Mk with broken underlining, and IA with double underlining.53 The word count for Mark 1.1–8 yields the following results: EAMt-Lk 120 / EAMk-Lk 20.5 / EAMt-Mk 17.5 / IA 12

51. Cf. Mark Goodacre, The Case Against Q: Studies in Markan Priority and the Synoptic Problem (Harrisburg, PA: Trinity Press International, 2002), 52–4. 52. Kahl, ‘Übereinstimmungen’, 28–36. 53. With respect to variations in word choice or placement that have an identical function in a sentence, I have highlighted the first and last letter of the parallel word, cf. the temporal phrase in Matt 3.1a // Luke 3.1a.

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

237

From the word count we see already that the Gospel of Matthew and the Gospel of Luke by far have most of the material in common, with Luke sharing more with Mark than Matthew with Mark. Compared with the overall picture of synoptic proximities and distances in Mark 1–16 parr. this result is in so far atypical as Matthew in general is closer to Mark than is Luke, and as Luke in the Triple-Tradition is only in about 40 per cent of the pericopes closer to Matthew than he is to Mark.54 However, if the EAMt-Lk outside the TripleTradition in the whole synoptic material are also considered – as they should – it becomes clear that Matthew and Luke share overwhelmingly much material over against Mark. In the following synoptic presentation of Mark 1.1–8 parr. I  have chosen Matthew as the middle term – following a suggestion by Mark Goodacre – indicating the historical sequence of the Synoptic gospels on the one hand and illustrating the – most likely – fact that Luke read the Gospel of Mark critically through the Matthean re-lecture of it on the other hand. The chart makes transparent that in this passage the Gospel of Luke shares with the Gospel of Matthew the rearrangement of the Markan material in all instances and Matthean additions to the Gospel of Mark in most instances. Here, ‘the parallels between Matthew and Luke are more extensive than those between Mark and Luke’.55 Furthermore, Luke adds to these common additions, over against the Gospel of Mark, material unique to and typical of his Gospel (3.1–2, 5–6, 10–16, 18–20). Rearrangement of the Markan material I. Both Matthew and Luke translocate Mark 1.2–3 in an identical manner and to the same location: They present the section after the first reference to John and his proclamation (Matt 3.3 // Luke 3.4). At the same time they isolate Mark 1.2b and move it to an identical location in a different context: Matt 11.10 // Luke 7.27. The reason for this isolation is clearly a correction of the Markan ascription of the verse to Isaiah, while it actually is a particular mixed quotation from Mal 3.1 and Exod 23.20.56 II. Both Matthew and Luke move forward Mark 1.8a in front of John’s proclamation of the stronger one (Matt 3.11a / Luke 3.16b) with both sharing a variant wording over against Mark. By means of this translocation, a better narrative flow is achieved.

54. Cf. Kahl, ‘Übereinstimmungen’, 35–6. 55. Goodacre, Case, 52. 56.  All three Synoptics show here a particular influence of the MT text, while usually they presuppose the LXX. This striking constellation motivated Albrecht Ritschl, who favoured a version of the 2ST already in 1851, to assume a knowledge of Matthew by Luke, cf. Albrecht Ritschl, ‘Ueber den gegenwärtigen Stand der Kritik der symoptischen Evangelien’, in Gesammelte Aufsätze von Albrecht Ritschl (Leipzig: Mohr 1893), 1–51, here 48–9 (original: 1851).

238

Werner Kahl

Table 12.2 Synoptic Presentation of Mark 1.1–8 parr. Mark 1.1-8

Matt 3.1-12

1 Ἀρχὴ τοῦ εὐαγγελίου Ἰησοῦ

Luke 3.1-20

(see 1.1)

(see 1.1-4)

Χριστοῦ [υἱοῦ θεοῦ]. 1 Ἐν δὲ ταῖς ἡμέραις ἐκείναις

1 Ἐν ἔτει δὲ πεντεκαιδεκάτῳ τῆς ἡγεμονίας Τιβερίου Καίσαρος, ἡγεμονεύοντος Ποντίου Πιλάτου τῆς Ἰουδαίας, καὶ τετρααρχοῦντος τῆς Γαλιλαίας Ἡρῴδου, Φιλίππου δὲ τοῦ ἀδελφοῦ αὐτοῦ τετρααρχοῦντος τῆς Ἰτουραίας καὶ Τραχωνίτιδος χώρας, καὶ Λυσανίου τῆς Ἀβιληνῆς τετρααρχοῦντος, 2 ἐπὶ ἀρχιερέως Ἅννα καὶ Καϊάφα,

3a ൻ

4a ൻ

(see 11.10)

(cf. 7.27)

3b ൻ

4b ൻ

2 Καθὼς γέγραπται ἐν τῷ Ἠσαΐᾳ τῷ προφήτῃ, Ἰδοὺ ἀποστέλλω τὸν ἄγγελόν μου πρὸ προσώπου σου, ὃς κατασκευάσει τὴν ὁδόν σου· 3 φωνὴ βοῶντος ἐν τῇ ἐρήμῳ, Ἑτοιμάσατε τὴν ὁδὸν κυρίου, εὐθείας ποιεῖτε τὰς τρίβους αὐτοῦ, 4 ἐγένετο Ἰ ω ά ν ν η ς β α π τ ίζων

παραγίνεται Ἰωάννης ὁ βαπτιστὴς ἐγένετο ῥῆμα θεοῦ ἐπὶ Ἰωάννην

ἐν τῇ ἐρήμῳ

κηρύσσων ἐν τῇ ἐρήμῳ τῆς

τὸν Ζαχαρίου υἱὸν ἐν τῇ ἐρήμῳ. 

Ἰουδαίας  3 καὶ ἦλθεν εἰς πᾶσαν [τὴν] 5b ൻ

περίχωρον τοῦ Ἰορδάνου

καὶ κηρύσσων βάπτισμα μετανοίας 2 [καὶ] λέγων, Μετανοεῖτε, ἤγγικεν κηρύσσων βάπτισμα μετανοίας εἰς εἰς ἄφεσιν ἁμαρτιῶν. 

γὰρ ἡ βασιλεία τῶν οὐρανῶν. 

ἄφεσιν ἁμαρτιῶν, 

3 οὗτος γάρ ἐστιν ὁ ῥηθεὶς διὰ 4 ὡς γέγραπται ἐν βίβλῳ λόγων 2a ↑

Ἠσαΐου τοῦ προφήτου

Ἠσαΐου τοῦ προφήτου,

λέγοντος, 3↑

Φωνὴ βοῶντος ἐν τῇ ἐρήμῳ,

Φωνὴ βοῶντος ἐν τῇ ἐρήμῳ,

Ἑτοιμάσατε τὴν ὁδὸν κυρίου,

Ἑτοιμάσατε τὴν ὁδὸν κυρίου,

εὐθείας ποιεῖτε τὰς τρίβους

εὐθείας ποιεῖτε τὰς τρίβους

αὐτοῦ.

αὐτοῦ·

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

Mark 1.1-8

Matt 3.1-12

239

Luke 3.1-20 5 πᾶσα φάραγξ πληρωθήσεται καὶ πᾶν ὄρος καὶ βουνὸς ταπεινωθήσεται, καὶ ἔσται τὰ σκολιὰ εἰς εὐθείαν καὶ αἱ τραχεῖαι εἰς ὁδοὺς λείας· 6 καὶ ὄψεται πᾶσα σὰρξ τὸ σωτήριον τοῦ θεοῦ.

4 Αὐτὸς δὲ ὁ Ἰ ω ά ν ν η ς εἶχεν τὸ ἔνδυμα αὐτοῦ ἀπὸ τριχῶν καμήλου 6ൻ

καὶ ζώνην δερμ α τ ί ν η ν π ε ρ ὶ τ ὴ ν ὀσφὺν αὐτοῦ, ἡ δὲ τροφὴ ἦν αὐτοῦ ἀ κ ρ ί δες κ α ὶ μ έ λ ι ἄ γ ρ ι ο ν. 

5 καὶ ἐξεπο ρεύ ετ ο πρ ὸς αὐ τ ὸ ν

5 τότε ἐξεπορεύετο πρὸς α ὐ τ ὸ ν

πᾶσα ἡ Ἰουδαία χώρα κ α ὶ οἱ Ἱ ε ρ ο-

Ἱεροσόλυμα κ α ὶ π ᾶσα ἡ Ἰ ο υ δαία

σ ο λ υ μ ῖται πάντες,

καὶ πᾶσα ἡ περίχωρος τοῦ

3a ↑

Ἰορδάνου,  κ α ὶ ἐ β α π τ ί ζο ν τ ο ὑ π’ α ὐ τοῦ ἐν τῷ 6 καὶ ἐβαπτίζοντο ἐν τῷ Ἰορδάνῃ Ἰορδάνῃ π οταμῷ ἐξομολογούμενοι π οταμῷ ὑπ’ α ὐτοῦ ἐ ξ ο μ ο λ οτ ὰ ς ἁ μ α ρ τ ί α ς α ὐ τ ῶ ν. 

γο ύμ ε ν οι τὰ ς ἁμ α ρ τ ί α ς α ὐ τ ῶ ν.

6 καὶ ἦν ὁ Ἰ ω ά ν ν η ς ἐ ν δ ε δ υμένος τ ρίχας κ α μ ή λ ο υ κ α ὶ ζ ώ ν η ν

4↑

δερματίνην περὶ τὴν ὀσφὺν αὐτοῦ, καὶ ἐ σ θίων ἀ κ ρ ί δας κ α ὶ μ έ λ ι ἄ γ ρ ι ο ν. 7 Ἰδὼν δὲ πολλοὺς τῶν Φαρισαίων 7 Ἔλεγεν οὖν τοῖς ἐκπορευομένοις καὶ Σαδδουκαίων ἐρχομένους ἐπὶ ὄχλοις βαπτισθῆναι ὑπ’ τὸ βάπτισμα αὐτοῦ εἶπεν αὐτοῖς, αὐτοῦ, Γεννήματα ἐχιδνῶν, τίς ὑπέ-

Γεννήματα ἐχιδνῶν, τίς ὑπέ-

δειξεν ὑμῖν φυγεῖν ἀπὸ τῆς

δειξεν ὑμῖν φυγεῖν ἀπὸ τῆς

μελλούσης ὀργῆς; 8 ποιήσατε οὖν μελλούσης ὀργῆς; 8 ποιήσατε οὖν καρπὸν ἄξιον τῆς μετανοί-

καρποὺς ἀξίους τῆς μετανοίας·

ας· 9 καὶ μὴ δόξητε λέγειν ἐν ἑαυ- καὶ μὴ ἄρξησθε λέγειν ἐν ἑαυτοῖς, τοῖς, Πατέρα ἔχομεν τὸν Ἀβραάμ, Πατέρα ἔχομεν τὸν Ἀβραάμ, λέγω γὰρ ὑμῖν ὅτι δύναται ὁ θεὸς λέγω γὰρ ὑμῖν ὅτι δύναται ὁ θεὸς ἐκ τῶν λίθων τούτων ἐγεῖραι

ἐκ τῶν λίθων τούτων ἐγεῖραι

τέκνα τῷ Ἀβραάμ.  10 ἤδη δὲ ἡ

τέκνα τῷ Ἀβραάμ.  9 ἤδη δὲ καὶ

ἀξίνη πρὸς τὴν ῥίζαν τῶν

ἡ ἀξίνη πρὸς τὴν ῥίζαν τῶν

δένδρων κεῖται· πᾶν οὖν δένδρον δένδρων κεῖται· πᾶν οὖν δένδρον μὴ ποιοῦν καρπὸν καλὸν ἐκκόπ- μὴ ποιοῦν καρπὸν καλὸν ἐκκόπτεται καὶ εἰς πῦρ βάλλεται.

τεται καὶ εἰς πῦρ βάλλεται. 

240

Werner Kahl

Mark 1.1-8

Matt 3.1-12

Luke 3.1-20 10 Καὶ ἐπηρώτων αὐτὸν οἱ ὄχλοι λέγοντες, Τί οὖν ποιήσωμεν ; 11 ἀποκριθεὶς δὲ ἔλεγεν αὐτοῖς, Ὁ ἔχων δύο χιτῶνας μεταδότω τῷ μὴ ἔχοντι, καὶ ὁ ἔχων βρώματα ὁμοίως ποιείτω.  12 ἦλθον δὲ καὶ τελῶναι βαπτισθῆναι καὶ εἶπαν πρὸς αὐτόν, Διδάσκαλε, τί ποιήσωμεν ;  13 ὁ δὲ εἶπεν πρὸς αὐτούς, Μηδὲν πλέον παρὰ τὸ διατεταγμένον ὑμῖν πράσσετε. 14 ἐπηρώτων δὲ αὐτὸν καὶ στρατευόμενοι λέγοντες, Τί ποιήσωμεν καὶ ἡμεῖς; καὶ εἶπεν αὐτοῖς, Μηδένα διασείσητε μηδὲ συκοφαντήσητε, καὶ ἀρκεῖσθε τοῖς ὀψωνίοις ὑμῶν. 15 Προσδοκῶντος δὲ τοῦ λαοῦ καὶ διαλογιζομένων πάντων ἐν ταῖς καρδίαις αὐτῶν περὶ τοῦ Ἰωάννου, μήποτε αὐτὸς εἴη ὁ Χριστός,

7 καὶ ἐκήρυσσεν λέγων,

16 ἀπεκρίνατο λέγων πᾶσιν ὁ Ἰωάννης,

8a ൻ

11 ἐγὼ μὲν ὑμᾶς βαπτίζω ἐν

Ἐγὼ μὲν ὕδατι βαπτίζω ὑμᾶς·

ὕδατι εἰς μετάνοιαν· Ἔρχεται ὁ ἰσχυρότερός μου ὀ π ί σ ω ὁ δὲ ὀ π ί σ ω μ ο υ ἐρχόμενος μ ο υ, οὗ οὐκ εἰμὶ ἱκανὸς κύψας

ἔρχεται δὲ ὁ ἰσχυρότερός μου,

ἰσχυρότερός μού ἐστιν, οὗ οὐκ εἰμὶ οὗ οὐκ εἰμὶ ἱκανὸς λῦσαι τὸν

λῦσαι τὸν ἱμάντα τῶν ὑποδημάτων ἱκανὸς τὰ ὑποδήματα βαστάσαι· ἱμάντα τῶν ὑποδημάτων αὐτοῦ· αὐτοῦ· 8 ἐγὼ ἐβάπτισα ὑμᾶς ὕδατι,

11a ↑

16b ↑

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

241

Mark 1.1-8

Matt 3.1-12

Luke 3.1-20

αὐτὸς δὲ βαπτίσει ὑμᾶς ἐν

αὐτὸς ὑμᾶς βαπτίσει ἐν πνεύματι αὐτὸς ὑμᾶς βαπτίσει ἐν πνεύματι

πνεύματι ἁγίῳ.

ἁγίῳ καὶ πυρί· 

ἁγίῳ καὶ πυρί· 

12 οὗ τὸ πτύον ἐν τῇ χειρὶ αὐτοῦ, 17 οὗ τὸ πτύον ἐν τῇ χειρὶ αὐτοῦ καὶ διακαθαριεῖ τὴν ἅλωνα

διακαθᾶραι τὴν ἅλωνα αὐτοῦ καὶ

αὐτοῦ, καὶ συνάξει τὸν σῖτον

συναγαγεῖν τὸν σῖτον εἰς τὴν

αὐτοῦ εἰς τὴν ἀποθήκην, τὸ δὲ

ἀποθήκην αὐτοῦ, τὸ δὲ ἄχυρον

ἄχυρον κατακαύσει πυρὶ

κατακαύσει πυρὶ ἀσβέστῳ.

ἀσβέστῳ. 18 Πολλὰ μὲν οὖν καὶ ἕτερα παρακαλῶν εὐηγγελίζετο τὸν λαόν·  19 ὁ δὲ Ἡρῴδης ὁ τετραάρχης, ἐλεγχόμενος ὑπ’ αὐτοῦ περὶ (see 6 . 17 - 1 8)

(see 1 4 .3 - 4)

Ἡρῳδιάδος τῆς γυναικὸς τοῦ ἀδελφοῦ αὐτοῦ καὶ περὶ πάντων ὧν ἐποίησεν πονηρῶν ὁ Ἡρῴδης, 20 προσέθηκεν καὶ τοῦτο ἐπὶ πᾶσιν, κατέκλεισεν τὸν Ἰωάννην ἐν φυλακῇ.

Additions of material in the Gospels of Matthew and Luke: 1. Both Matthew and Luke add a temporal indicator at the very beginning of the discourse, introduced by the same phrase, Ἐν δέ, plus reference to time. Matthew’s introduction, Ἐν δὲ ταῖς ἡμέραις ἐκείναις παραγίνεται . . ., seems to be informed by Mark 1.9: Καὶ ἐγένετο ἐν ἐκείναις ταῖς ἡμέραις . . . As a narrative bridge between Matt 1–2 and the following discourse about John the Baptist, this temporal indicator is narratologically problematic since it seems to indicate, taken at its face value, that John the Baptist appeared historically at about the same time that Jesus’ parents settled with their infant in Nazareth (cf. Matt 2.19–23).57 Luke has here a detailed list of historical references, which is a typical Lukan feature. His temporal introduction does not create any narratological tension with the preceding chapters. Presupposed that Luke also knew the Gospel of Matthew, his account appears to be informed at the beginning by Matthew’s introductory formula Ἐν δέ . . . At the same time, 57. This inconsistency had already been discussed by David Friedrich Strauss, Das Leben Jesu, kritisch bearbeitet, vol. 1 (Tübingen: C. F. Osiander, 1835), 309–10, where he calls it a ‘schlaffe Zeitbestimmung’ (310).

242

Werner Kahl

Luke’s version could be understood as a narratological improvement on the Gospel of Matthew.58 2. In Luke 3.3 an addition of Matthew over against Mark occurs: Matt 3.5b reads καὶ πᾶσα ἡ περίχωρος τοῦ Ἰορδάνου. Luke has a variant form of this phrase: καὶ ἦλθεν εἰς πᾶσαν [τὴν] περίχωρον τοῦ Ἰορδάνου. In Matthew the phrase introduces the people of the region of Jordan just before it is narrated in Matt 3.6 that the people were actually baptized there. Its phrasing is a parallel to the mentioning of Judea: καὶ πᾶσα ἡ Ἰουδαία. In addition, Matthew changes the sequence of named locations as compared to Mark. In my view, the addition prepares the location of the baptism which in Mark comes without introduction. In Luke, however, the phrase features with a different function. Here it is not the subject of an activity but the object of an activity of John, clarifying right from the beginning what Matthew attempted to communicate in a somewhat clumsy way: the location where John’s proclamation and baptism took place. Here we have an improvement on Matthew’s improvement on Mark. According to Mark it seems as if John baptized and proclaimed ‘in the desert’ while in Matthew the reference to his baptizing activity occurs as an appositional attribute of John, the Baptizer, possibly in order to avoid the puzzling Markan statement of a baptism in the desert. Luke however does not have any qualification of John as Baptizer, but calls him in accordance with the information given in his birth narrative, son of Zacharias. This John is qualified – according to the Narrative Schema: in the phase of preparedness involving his competence and probably indicating a divine manipulation – by a Word of God coming to him. This happens in the desert from where John moves to the region around Jordan where he then proclaims – as he does in the Markan version over against the Gospel of Matthew, that is, ‘baptism of repentance for forgiveness of sins’. Matthew here has a typical reference to the nearness of the kingdom of heaven. In the CEQ, the addition to the Markan version in the Gospel of Matthew and the Gospel of Luke – under consideration here – is registered as a torso by deleting all letters particular to either the Matthean or the Lukan versions. The editors could not decide if Matthew or Luke kept the ‘Q original’. As it stands in the CEQ, bare of any grammatical function, the remaining letters are narratologically meaningless.59

58.  The same tendency of a Lukan correction and narratological improvement of a Matthean imprecision and narratological inconsistency caused by the addition of material to the Gospel of Mark can be observed at the beginning of the Sermon on the Mount (Matt 5.1 // Luke 6.17–19), where, according to the sequence of Matthew’s account, the number of disciples gathering around Jesus can only be four. 59. Hoffmann and Heil, Spruchquelle, 32. It should also be noted that in the CEQ, with respect to the passage under consideration here, shorter EAMt-Lk and the rearrangements of Markan material common to Matthew and Luke are not considered.

12. The Gospel of Luke as Narratological Improvement of Synoptic Pre-texts

243

1. Luke glosses over Mark 1.5–6 and Matt 3.4–6, probably because he wants to de-focus John (cf. also 3.19–20 where he moves forward and sharply shortens the reference to John’s imprisonment and death as compared to the accounts in Mark 6 and Matt 14). He shares with Matthew the following condemnation (Matt 3.7–10 / Luke 3.7–9), but with a remarkable change as concerns the addressees of the utterances: While for Matthew they are typically Pharisees and Sadducees, in Luke all present at the baptismal scene are addressed. This is on the one hand typical of Luke’s universalism (see his extension of the preceding quotation: πᾶσα σὰρξ) and on the other hand typical of his general avoidance in blaming Jews and Pharisees in particular. 2. It can be seen in the chart that, four times in this narrative sequence, Luke adds sections to additions of Matthew with respect to Mark: Luke 3.1–2, 5–6, 10–16, 18–20. While Matthew in 3.7 is specific about the Pharisees and Sadducees, Luke in 3.12–14 is specific about tax collectors and Roman soldiers, indicating different groups which were probably of particular concern for either Matthew or Luke. In 3.16 Luke gives a reason for John’s utterance about ‘the stronger one’, which is lacking in both the Gospel of Mark and the Gospel of Matthew, answering the open question of the bystanders whether John might be the Messiah. While preparing the utterance and thus providing a narratological frame for it, this reason is at the same time a correction of Matthew’s version. Since the utterance in Matt 3.11 immediately follows the words of warning against the Pharisees and Sadducees, ὑμᾶς as direct object of John’s and the stronger one’s baptism refers only to them. This is an irritating scenario that Luke corrects: The recipients of these baptisms are ‘all’ (3.16) or the ‘people’ (3.15, 18). 3. In Matt 3.11c–12 // Luke 3.16c–17, a common passage about the baptism with fire is added over against Mark 1.8b. Again, in the Gospel of Matthew it is directed only against the Pharisees and Sadducees while Luke has it as a general statement. This strong statement is balanced in the additional Luke 3.18, which indicates an admonishment qualified as a proclamation of ‘good news’. This further admonishment is analogous to the advice given in the additional section Luke 3.10–14, since these pieces of advice give concrete direction as to how to escape the expected judgement. In the Matthean version such advice is not given.

12.4 Conclusion The analysis has shown with respect to Mark 1.1–8 parr. that it is possible and seems to be most plausible to explain the differences between the Synoptics without recourse to any hypothetical source. It seems sufficient, economical and reasonable to assume that Luke used both the Gospel of Mark and the Gospel of Matthew. While Matthew attempted an improvement and actualization of the Gospel of Mark relevant to his context, Luke critically integrated both sources serving the needs of his account. Luke’s reading of both the Gospel of Mark and the Matthean improvement of this Gospel resulted in lexical, grammatical,

244

Werner Kahl

syntactic and narratological changes.60 These changes, in line with Luke’s intentions as communicated in his prologue, can be considered particularly Lukan attempts at improving his Vorlagen – not only narratologically but also historically and theologically.61 The coherence of his presentation of the story of Jesus Christ as compared to both the Gospel of Mark and the Gospel of Matthew demonstrates that Luke’s attempts were narratologically quite successful.62 This holds true not only for the passage under consideration here, but in general for the whole of the Gospel of Luke and with Acts.63

60. This scenario is supported by the detailed word and style studies of Wilhelm Larfeld, Die neutestamentlichen Evangelien nach ihrer Eigenart und Abhängigkeit (Gütersloh:  C. Bertelsmann, 1925), even though Larfeld also reckoned with the additional use of ‘Q’ by Matthew and Luke. 61.  Cf. the instructive study of Finn Damgaard, Rewriting Peter as an Intertextual Character in the Canonical Gospels. Copenhagen International Seminar (Routledge: New York, 2016). 62.  At times, however, Luke commits narratological errors in his attempt at critically integrating his sources, cf., e.g., the reference to Jesus’ miracles in Capernaum in Luke 4.23, even though at this point of the Lukan narrative Jesus had not been to Capernaum yet. 63.  Cf. the Lukan omission of the discourse on purity (Mark 7.1–23 parr. and also 7.24–30 parr.) in light of the ultimate, revelatory treatment of the subject in Acts 10–15 (cf. Wolter, Lukasevangelium, 30). This appears as a deliberate alteration of Luke, since otherwise the historical opposition of Jesus’ disciples as apostles in Jerusalem to the inclusion of non-Jesus as non-Jews would appear inexplicable. Cf. in this context also his correction of the Markan-Matthean notion of the disciples’ move to Galilee (Mark 16.7 parr. and Luke 24.13–53 // Acts 1.1–8.4).

Chapter 13 D O E S D AT I N G L U K E - A C T S I N T O T H E S E C O N D C E N T U RY A F F E C T T H E Q H Y P O T H E SI S ? Shelly Matthews

The Two-Source hypothesis has commonly been predicated on the assumption that Matthew and Luke, the two gospels using Mark and Q as sources, were both composed in the 80s of the Common Era. The question pursued in this chapter, as indicated by its title, is whether we might still assume the same set of relationships between source and redaction if a persuasive case can be made that the canonical Gospel of Luke did not reach a relatively stable form until the early decades of the second century.1 At first glance, proposing a second-century date for Luke’s final form might be thought to be an argument in favour of some version of a ‘Mark without Q’ hypothesis, since the whole of the double tradition might then be explained by Luke’s ready access to the Gospel of Matthew, a Gospel in circulation for decades before Luke writes.2 The argument here takes a different tack. The focus of my recent research has been on the ‘book ends’ of Luke, where signs of late redaction appear most clearly.3 Adjudicating the rhetorical force of 1.  I use the phrase ‘relatively stable form’, rather than ‘final form’, in recognition of the fact that of the canonical Gospels, Luke was an especially ‘open book’, subject to considerable expansion and reworking in the first three centuries of the Common Era. See D. C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997), 148–74. 2.  As suggested by Mogens Müller at the first Roskilde conference:  ‘Acts as Biblical Rewriting of the Gospels and Paul’s Letters’, in Luke’s Literary Creativity, ed. Mogens Müller and Jesper Tang Nielsen, The Library of New Testament Studies (LNTS) 550 (London: Bloomsbury T&T Clark, 2016), 96–117. 3.  On Luke 1–2, see Shelly Matthews, Perfect Martyr:  The Stoning of Stephen and the Construction of Christian Identity (New  York:  Oxford University Press, 2010), 43–54. On Luke 24, see Shelly Matthews, ‘Elijah, Ezekiel and Romulus: Luke’s Flesh and Bones (Luke 24.39) in Light of Early Jewish and Roman Narratives of Ascent and Resurrection’, in On Prophets, Warriors and Kings: The Former Prophets through the Eyes of their Interpreters, ed. Ariel Feldman and George Brooke, Beihefte zur ZAW (BZAW) 470 (Berlin: de Gruyter, 2016), 161–82; and Shelly Matthews, ‘Fleshly Resurrection, Authority Claims, and the Scriptural Practices of Lukan Christianity’, Journal of Biblical Literature (JBL), 136 (2017): 163–83.

246

Shelly Matthews

this redactional work has required engagement with a number of scholars reviving the theory that this redaction is aimed directly at Marcion.4 This in turn has led to the question of how best to understand the relationship between Marcion’s Gospel and canonical Luke, and what insight into the synoptic problem can be gleaned through bringing Marcion’s Gospel into the debate. While in my view the Two-Source hypothesis remains the best way to account for the bulk of agreements of Matthew and Luke not attributable to Mark, the late dating of the ‘book ends’ of canonical Luke requires some modification of this hypothesis. I suggest here that canonical Luke was preceded by a ‘core Luke’5 which utilized Q and Mark as sources. In a later redactional stage, canonical Luke may have depended on Matthew as a source for its prefatory materials. To build this argument, I first take up the question of dating Lukan materials, beginning with the book of Acts.

13.1 Dating Lukan Materials Acts. I  hold that the Acts of the Apostles was composed in the second century of the Common Era, sometime between 100 and 130CE, likely towards the later end of that date range.6 The persuasive arguments of Richard Pervo,

4.  See, for instance, Joseph Tyson, Marcion and Luke-Acts:  A Defining Struggle (Columbia, SC: University of South Carolina Press, 2006); Matthias Klinghardt, ‘Markion vs. Lukas: Plädoyer für die Wiederaufnahme eines alten Falles’, New Testament Studies (NTS) 52 (2006): 484–513; Matthias Klinghardt, ‘The Marcionite Gospel and the Synoptic Problem: A New Suggestion’, Novum Testamentum (NT) 50 (2008):  1–27; Matthias Klinghardt, Das älteste Evangelium und die Entstehung der kanonischen Evangelien, 2 vols, Texte und Arbeiten zum neutestamentlichen Zeitalter (TANZ) 60 (Tübingen: Francke Verlag, 2015). 5. To be clear, the ‘core Luke’ postulated here is not a revival of the so-called Proto-Luke hypothesis popularized by B. H. Streeter, The Four Gospels: A Study of Origins, Treating of the Manuscript Tradition, Sources, Authorship & Dates (London: MacMillan, 1924), 233–70. 6.  Among recent scholars who either argue specifically for dating Acts to the second century, or assume such a date as they work to interpret the text, see Vernon Robbins, ‘A Socio-Rhetorical Look at the Work of John Knox on Luke-Acts’, in Cadbury, Knox, and Talbert:  American Contributions to the Study of Acts, ed. Mikeal C. Parsons and Joseph B. Tyson (Atlanta:  Scholars Press, 1992), 91–105; Christopher Mount, Pauline Christianity:  Luke-Acts and the Legacy of Paul, Novum Testamentum Supplements (NovTSup) 104 (Leiden: Brill, 2002); Christopher Mount, ‘Constructing Paul as a Christian in the Acts of the Apostles’, in Engaging Early Christian History: Reading Acts in the Second Century, ed. Rubén R. Dupertuis and Todd Penner (Durham:  Acumen, 2013), 141–52; Mary Rose D’Angelo, ‘ The ANHP Question in Luke-Acts: Imperial Masculinity and the Deployment of Women in the Early Second Century’, in A Feminist Companion to Luke, Feminist Companion to the New Testament and Early Christian Writings (FCNTECW) 3 (Sheffield: Sheffield Academic Press, 2002), 44–69; Richard I. Pervo, Dating Acts: Between

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 247

supplemented by several additional scholars in recent years, that the author of the Acts of the Apostles uses portions of Josephus’ Antiquities as a source, and has access to a collection of the letters of Paul, sets the terminus a quo for the writing at 100CE.7 Though recognition of Acts’ dependence on these sources is useful in establishing the earliest possible date of its composition, Acts’ close relationship with other textual materials securely dated to the second century, along with the social world depicted in the narrative, lead me to think it more likely that the text was written after 115 CE. A compositional window between 115 and 130 would situate Acts after Pliny’s correspondence with Trajan,8 and in contrapuntal relation with Hadrian’s establishment of the Panhellenion.9 A date closer to 130 than to 100 also helps to make sense of the fact that its author not only seems aware of the searing criticisms that come to be levelled against

the Evangelists and the Apologists (Sonoma, CA:  Polebridge, 2006); Richard I. Pervo, Acts:  A Commentary, Hermeneia (Minneapolis:  Fortress, 2009); Tyson, Marcion and Luke-Acts; Klinghardt, ‘Markion vs. Lukas’; Klinghardt, Das älteste Evangelium; Matthews, Perfect Martyr; Dennis Smith and Joseph Tyson, eds., Acts and Christian Beginnings: The Acts Seminar Report (Salem, OR: Polebridge, 2013); John Kloppenborg, ‘Literate Media in Early Christ Groups: The Creation of a Christian Book Culture’, Journal of Early Christian Studies (JECS) 22 (2014): 21–59. 7. Acts’ Use of Josephus: Steve Mason, Josephus and the New Testament (Peabody, MA: Hendrickson, 1992), 185–225; Pervo, Dating Acts, 149–60; Matthews, Perfect Martyr, 63–5. See also Andrew Gregory, ‘Acts and Christian Beginnings: A Review Essay’, Journal for the Study of the New Testament (JSNT) 39 (2016): 97–115. Gregory presses the point that it would be possible to use the Antiquities as a source as early as the mid 90s CE, but acknowledges also that this seems unlikely for any author living outside Rome. Acts’ Use of a Pauline Letter Collection: Pervo, Dating Acts, 51–147; Lars Aejmelaeus, ‘ The Pauline Letters as Source Material in Luke-Acts’, in The Early Reception of Paul, ed. Kenneth Liljeström (Helsinki: Finnish Exegetical Society, 2011), 54–75; Heikki Leppä, ‘Luke’s Selective Use of Gal 1 and 2: A Critical Proposal’, in The Early Reception of Paul, ed. Liljeström, 91–124; Ryan Schellenberg, ‘The First Pauline Chronologist? Pauls’ Itinerary in the Letters and in Acts’, Journal of Biblical Literature 134/1 (2015): 193–213. Aejmelaeus (p. 67, with regard to the escape from Damascus); Leppä (p. 101, with regard to Paul’s activity immediately following his conversion); and Schellenberg (pp. 211–13 with regard to Paul’s itinerary) share in common the observation that when details concerning significant events in Paul’s biography are absent from the Pauline epistles, they are also absent from Acts. 8.  Thomas E. Phillips, ‘How did Paul Become a Roman “Citizen?” ’ and Mark G. Bilby, ‘Pliny’s Correspondence and the Acts of the Apostles: An Intertextual Relationship?’, in Luke on Jesus, Paul, and Earliest Christianity: What Did He Really Know?, ed. Joseph Verheyden and John S. Kloppenborg, Biblical Studies and Tools 29 (Leuven: Peeters, 2017), 147–69 and 171–89, respectively. 9. Laura Nasrallah, ‘The Acts of the Apostles, Greek Cities, and Hadrian’s Panhellenion’, JBL 127:3 (2008): 533–66.

248

Shelly Matthews

Christians once they are recognized as a distinct social group by literate outsiders but also works to diffuse them.10 Luke 24. Close study of the final chapter of canonical Luke, and consideration of its programmatic function for the Acts narrative, has led me also to question the view that the date of this composition – at least the ‘final form’ of this final chapter11 – could lie in the mid-80s of the first century, as many scholars assume. While my arguments for dating Acts to the second century above rely heavily on work that has been published for some time now, and thus could be summarized briefly, this argument questioning the dating of the final chapter of canonical Luke requires more detailed elaboration here.12 There are a number of reasons to question the assumption that canonical Luke 24 was composed in the 80s, and to ask instead whether it might have reached its final form in close connection to the composition of the Acts of the Apostles sometime in the second century. One might note, for instance, the repeated observations of John Alsup, in his study of source and redaction of post-resurrection appearance stories, that Luke 24 stands out among canonical gospels for the complexity of the relationships it sketches, and its overall compositional sophistication, and then ask whether it is likely for such compositional sophistication in tying together so many threads of early resurrection tradition to have been manifest already by 85 of the common era.13 One might ask about the obvious Eucharistic 10. Though I make no claim that Acts is written as late as the Octavius of Minucius Felix (c.160–250), it is yet striking that Acts seems to anticipate virtually every one of the disparaging criticisms against Christianity levelled by Minucius’s antagonist Caecilius – from the low status and mixed gender of the group, to the charges of jabbering in corners, spitting at the gods, scorning the purple robes of public office and behaving in ways even more sociopathic than those he dismisses as the ‘wretched tribe’ of the Jews. See Shelly Matthews, Acts of the Apostles: Taming the Tongues of Fire, Phoenix Guides to the New Testament (PGNT) 5 (Sheffield:  Sheffield Phoenix, 2013), 45–57. For a contrasting view of the relationship of Acts to the Apologists, see Andrew Gregory, ‘Among the Apologists? Reading Acts with Justin Martyr’, in Engaging Early Christian History: Reading Acts in the Second Century, ed. Rubén R. Dupertuis and Todd Penner (Durham, UK: Acumen, 2013), 169–86, esp. 184. 11. Again I note that establishing a ‘final form’ is a relative, rather than absolute undertaking, especially given the fluidity of the textual tradition for canonical Luke, and especially Luke 24. On the seven so-called Western non-interpolations contained in Luke 24 (24.3b, 24.6a, 24.12, 24.36b, 24.40, 24.51b; 24.52a), see Michael Wade Martin, ‘Defending the “Western Non-Interpolations”: The Case for an Anti-Separationist Tendenz in the Longer Alexandrian Readings’, JBL 142/2 (2005): 269–94. The fact of the cluster of significant divergent readings in Luke 24 arising in the second century, along with the divergent readings of canonical Luke and Marcion’s Evangelion, supports my general observation that the final chapter of Luke was a matter of intense interest, and considerable reworking in the early to mid-second century. 12. See related arguments in Matthews, ‘Fleshly Resurrection’. 13.  John Alsup, The Post-Resurrection Appearance Stories of The Gospel Tradition:  A History-of-Tradition Analysis with Text-Synopsis (Stuttgart: Calwer Verlag, 1975), esp. 147– 213. Alsup himself does not pursue this question, but assumes a composition shortly after the time of the Gospel of Matthew.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 249

overtone in the Emmaus episode, where eyes are opened and Jesus is recognized after bread is blessed, broken and given (Luke 24.30–31) and ask how much time is required for such a ritual performance to be translated so neatly into this particular narrative form. One might raise questions about relationships between Luke 24 and John 20, and even John 21 (widely accepted as a late appendix to the fourth Gospel), at points where divergences seem best explained by Luke’s dependence upon John, rather than vice versa.14 Though each of these questions merits further study, my own work on Luke 24 up until now has focused on the following two distinctive aspects of Lukan resurrection narratives which suggest a late dating for the redaction of this material: first, how Luke 24 (and Acts) approach the question of scripture interpretation; second, how fleshly resurrection in Luke 24 (in tandem with assertions pertaining to fleshly resurrection in Acts) function to authorize exclusive claims pertaining to the twelve apostles as authoritative witnesses to the resurrection. Consider first the question of how Luke 24 treats the matter of scripture fulfilment and scripture interpretation. The three scenes in Luke 24 – the report of the empty tomb, the appearance to the travellers on the Emmaus road and the appearance to the eleven and those with the eleven – when taken together, function as the most expansive assertion of proof-from-prophecy of any canonical Gospel. The proof-from-prophecy claim here is expansive not just in acknowledging a tripartite division of Hebrew Scriptures – the law, the prophets and the Psalms (Luke 24.44) – but also in asserting that the events in the life of Jesus fulfil the scriptures in their totality.15 Further, the narrative of a post-resurrection appearance to the apostles, in which Jesus teaches the apostles to read scripture in a manner enabling them especially to understand how Jesus’ suffering fulfils scripture, is a tradition Luke shares not with other canonical gospels, but with the apology of Justin Martyr, demonstrating that with respect to this idea Luke is in the orbit of the apologists, rather than the earlier evangelists.16 Finally – and most compellingly 14.  For one argument that Luke uses John as a source, see Barbara Shellard, ‘The Relationship of Luke and John:  A Fresh Look at an Old Problem’, Journal of Theological Studies (JTS) 46 (1995): 70–98. For the specific argument that the western non-interpolations at 24.12, 36b and 40 are best explained as deriving from John 20, see Martin, ‘Western Non-Interpolations’, 282–5. 15. Luke 24.27b: διερμήνευσεν αὐτοῖς ἐν πάσαις ταῖς γραφαῖς τὰ περὶ ἐαθτοῦ; 24.44b: δεῖ πληρωθῆναι πάντα τὰ γεγραμμένα. Cf. Mogens Müller, ‘The Reception of the Old Testament in Matthew and Luke-Acts: From Interpretation to Proof from Scriptures’, Novum Testamentum (NovT) 43 (2001): 315–30; Jacob Jervell, ‘The Center of Scripture in Luke’, in Jacob Jervell, The Unknown Paul: Essays in Luke-Acts and Early Christian History (Minneapolis: Augsburg, 1984), 122–37, 179–83; on the programmatic nature of Luke 24, see especially P. Schubert, ‘The Structure and Significance of Luke 24’, in Neutestamentliche Studien für Rudolf Bultmann zu seinem siebzigsten Geburtsag (Berlin: Töpelmann, 1957), 165–86. 16. Justin, 1 Apol. 50.12; Dial. 53.5; 106.1. Oskar Skarsaune, The Proof from Prophecy: A Study in Justin Martyr’s Proof-Text Tradition:  Text-Type, Provenance, Theological Profile (Leiden: Brill, 1987), 256–9.

250

Shelly Matthews

to my mind – this chapter, along with key passages concerning interpretation of scriptures in Acts, embraces a hermeneutical principle articulated by mid- and late-second century Christian authors that the scriptures have a ‘nous’, a mind, or an underlying sense, that needs to be exposited by a ‘qualified guide’ for those who seek to understand them.17 This hermeneutical principle aligns with the admonition in 2 Peter (perhaps the latest book to be written which has been included in the canon) that its addressees adhere foremost to the knowledge that ‘no prophecy of scripture is a matter of one’s own interpretation πᾶσα προφητεία γραφῆς ἰδίας ἐπιλύσεως οὐ γίνεται’ (2 Peter 1.20). It is illustrated in Acts, for instance, by the Ethiopian Chamberlain reading Isaiah who, when he is asked by the apostle Philip whether he understands what he is reading, responds, ‘How shall I be able to understand unless someone guides me?’ (Acts 8.31). This hermeneutical principle is invoked, at least once in Luke 24, by Jesus expositing the scriptures for those on the Emmaus road (24.32). If one accepts the recent arguments of Matthew Bates for translation of 24.45, then here also Jesus ‘opens the mind of the scripture’ for the Eleven.18 Thus, Luke 24.32, and possibly Luke 24.45, along with Acts 8.31, share a hermeneutical principle with 2 Peter and early Christian authors such as Justin Martyr and Irenaeus, a principle not articulated by earlier generations of Jesus followers, but one adopted by a number of second-century textual communities, as they articulate norms for reading earlier texts.19 Second, consider how Luke 24 and Acts closely connect the appearance of the resurrected Jesus in flesh and bone to the authority of the twelve apostles. In Luke 24 Jesus appears before the 11 (and those with them), saying ‘touch me and see; for a spirit does not have flesh and bones as you see that I have’ (24.39b: ψηλαφήσατέ με καὶ ἴδετε, ὅτι πνεῦμα σάρκα καὶ ὀστέα οὐκ ἔχει καθὼς ἐμὲ θεωρεῖτε ἔχοντα).

17.  Here I  follow closely the arguments of Matthew Bates, ‘Closed-Minded Hermeneutics?: A Proposed Alternative Translation for Luke 24.45’, JBL 129/3 (2010): 537–57. 18. While 24.45: τότε διήνοιξεν αὐτῶν τὸν νοῦν τοῦ συνιέναι τὰς γραφάς is commonly translated, ‘he opened their minds to the scriptures’, Matthew Bates has recently proposed translating it as, ‘Jesus exposited the Scriptures so that the disciples could understand their meaning’, a translation which is syntactically possible, and also conforms to Luke’s use of ‘διήνοιγεν + τὰς γραφάς’ in Luke 24.32 and Acts 17.2–3. Bates himself focuses on the hermeneutical implications of understanding scriptures as having an underlying mind and requiring a suitable guide, a ‘mind’ which he suggests in the end is the ‘core content of the early Christian kerygma’, (‘Closed-Minded Hermeneutics’, 556). He himself does not raise the question of the implications of his argument for dating Luke 24. 19.  Consider Justin, Dial 29.2: ‘For these words have not been fashionably arranged by me, nor embellished by human technique, but rather David sang them, Isaiah preached them, Zechariah heralded them, Moses recorded them. Do you recognize them Trypho? They are stored up in your Scriptures, or rather not in yours but in ours, for we are obedient to them, but when you read them, you do not understand the ‘mind’ in them [ὑμεῖς δὲ ἀναγινώσκοντες οὐ νοεῖτε τὸν ἐν αὐτοῖς νοῦν]’, cited in Bates, ‘Closed-Minded Hermeneutics’, 537.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 251

A first clue signalling Luke’s awareness of an internal and ongoing quarrel concerning the nature of the risen Christ, as Daniel Smith has noted, comes from the fact that Jesus is said to interrupt and clarify a resurrection debate already in progress, even among those within the inner circle. Before he asserts that he stands before them in flesh and bone, Jesus is said to perceive the ‘disputes’ [dialogismoi] that arise in the hearts of the eleven (24.38) concerning the nature of what they are seeing.20 As further proof that he stands before them in flesh, he eats the grilled fish (24.41–43).21 This is the only explicit reference to Jesus’ fleshly resurrection in all of the canonical gospels, already a clue that Luke 24 may be a participant in a debate best placed in the third generation of Jesus followers.22 Unless one grants an early date to the composition of Barnabas, indications of contests concerning the fleshly substance of the resurrected Christ in the apostolic fathers are generally dated from the Trajanic period and beyond.23 A stronger case may be mounted for this later window of time if one recognizes that the assertion of Luke 24 that the resurrected Jesus stands in flesh is closely tied to his assertions pertaining to apostolic authority. As I have argued elsewhere, canonical Luke 24 sets out a programmatic statement on the connection of the appearance of the resurrected Jesus in the flesh to the exclusive authority of the eleven (soon-to-be-twelve) apostles/witnesses. This connection is underscored in the inaugural speeches of both Peter (2.31–32) and Paul (13.31,37) in which assertions that Jesus’ flesh was not corrupted through the process of death and resurrection are closely linked to authorization of the twelve

20.  Daniel A. Smith, ‘Seeing a Pneuma[tic Body]:  The Apologetic Interests of Luke 24.36–43’, Catholic Biblical Quarterly (CBQ) 72 (2010): 752–72, esp. 753. 21. This, in distinction to John 21.9–13, where Jesus distributes, but does not partake of, the grilled fish and bread. 22. To avoid confusion, I clarify here that in the ancient world, bodily resurrection and fleshly resurrection are not identical concepts. Of course, the apostle Paul understands Jesus to have risen bodily but, as is well known, insists that the resurrected body is not a body of flesh, but rather one suited for the heavens (1 Cor 15.35–50). For comparison of resurrection claims in John and Luke, see Matthews, ‘Fleshly Resurrection’. For a careful study of the variety of views on the afterlife in early Judaism, see Outi Lehtipuu, Debates Over the Resurrection of the Dead: Constructing Early Christian Identity, Oxford Early Christian Studies (OECS) (Oxford: Oxford University Press, 2015). 23. For fleshly resurrection in Barnabas, see Barn. 5.6–7. For discussion of the dating of the apostolic fathers, see Clayton N. Jefford, The Apostolic Fathers and the New Testament (Peabody, MA:  Hendricks 2006), 7–37; and the introductions supplied in Bart D. Ehrman (ed.), The Apostolic Fathers, 2 vols, Loeb Classical Library (LCL) (Cambridge, MA: Harvard University Press, 2003). Jefford favours the range 96–100 CE for Barnabas; Ehrman (Apostolic Fathers, 2.7) argues for c. 130 CE. Additional passages pertaining to the fleshly resurrection of Jesus in the apostolic fathers include Ign. Smyrn. 32 and 2 Clem. 9.1–5.

252

Shelly Matthews

as exclusive witnesses to that resurrection.24 This connection is also signalled in the story of the restoration of the twelve in Acts 1.12–26, and most elaborately in Peter’s speech before Cornelius: [God] made him manifest not to all the people, but to those witnesses who were foreordained by God, to us, we who ate and drank with him after he rose from the dead. And he commanded us to preach to the people and to bear witness that this is the one ordained by God as judge of the living and the dead. To him all the prophets bear witness [καὶ ἔδωκεν αὐτὸν ἐμφανῆ γενέσθαι, οὐ παντὶ τῷ λαῷ, ἀλλὰ μάρτυσιν τοῖς προκεχειροτονημένοις ὑπὸ τοῦ θεοῦ, ἡμῖν, οἵτινες συνεφάγομεν καὶ συνεπίομεν αὐτῷ μετὰ τὸ ἀναστῆναι αὐτὸν ἐκ νεκρῶν. Καὶ παρήγγειλεν ἡμῖν κηρύξαι τῷ λαῷ καὶ διαμαρτύρασθαι ὅτι οὗτός ἐστιν ὁ ὡρισμένος ὑπὸ τοῦ θεοῦ κριτὴς ζώντων καὶ νεκρῶν τούτῳ παντές οἱ προφῆται μαρτυροῦσιν]. (10.40b–43a)

Consider the weighty assertions pertaining to early Christian authority and proclamation condensed into this succinct formula. The role of the twelve as witnesses is said to have been foreordained (προκεχειροτονημένοις). In yet another instance of the expansive nature of proof-from-prophecy in Acts, Peter claims authorization to preach a message to which all the prophets testify (10.43). This text employs the (pre)-creedal affirmation that Jesus is judge of the living and the dead (κριτὴς ζώντων καὶ νεκρῶν) a phrase it shares with many turn-of-the-century writings, including 1 Peter, Barnabas, 2 Timothy, 2 Clement, and Polycarp.25 Finally, the exclusive status asserted for the twelve apostles owes to their having witnessed the resurrected Jesus in the flesh, proof of which lies in their having eaten and drunk with him.26 It is my contention that the groundwork supporting these arguments in Acts for the exclusive authority of the twelve to speak for Jesus because they saw Jesus in the flesh (Acts 1.12–26; 2.31–32; 10.40–43; 13.31,37) is laid in canonical Luke 24. Elaine Pagels recognized long ago that Luke’s particular apostolic concept was used from the second century onwards to validate the apostolic succession, and to combat

24.  Peter connects the incorruptibility of Jesus’ flesh with the witness of the twelve at Acts 2.31b-32: οὔτε ἡ σάρξ αὐτοῦ εἶδεν διαφθοράν. Τοῡτον τὸν Ἰησοῦν ἀνέστησεν ὁ θεός, οὗ πάντες ἡμεῖς μάρτυρες). Paul argues similarly, proclaiming Jesus’ incorruptibility at 13.37 (ὅν δὲ ὁ θεὸς ἤγειρεν, οὐκ εἶδεν διαφθοράν) prefaced by the reminder at 13.31 that after he was raised, Jesus appeared to the witnesses/the twelve. For further elaboration of the significance of these speeches, see Matthews, ‘Fleshly Resurrection’. 25.  Compare 2 Tim 4.1:  κρίνειν ζῶντας καὶ νεκρούς; 1 Pet 4.5:  κρῖναι ζῶντας καὶ νεκρούς; Barn. 7.2:  ὢν κύριος καὶ μέλλων κρίνειν ζῶντας καὶ νεκρούς; 2 Clem. 1.1:  περὶ κριτοῦ ζώντων καὶ νεκρῶν; Pol. Phil 2.1:  ὃς ἐρχεται κριτής ζώντων καὶ νεκρῶν; and also Hegesippus (Eusebius, Hist. Eccl. 30.20.4): κρίνεῖ ζῶντας καὶ νεκρούς. 26. For arguments that eating functions here, as in Acts 1.3 and Luke 24.43, as a cipher for fleshly resurrection, see Matthews, ‘Fleshly Resurrection’.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 253

the authority of visionaries.27 My argument here tweaks that of Pagels by suggesting that Luke’s writing serves that purpose so well, not because he anticipated an argument that would take place only among future generations of Christians, but because the final redaction of Lukan materials takes place in the second century itself, when these debates concerning fleshly resurrection and authority arise.28 In sum, the arguments crafted in Luke 24 pertaining to scripture interpretation, fleshly resurrection and exclusive apostolic witness, which are subsequently developed in Acts, lead me to the conclusion that the final chapter of canonical Luke also received its final form in the second century. Though these conclusions have been reached independently from the recent revival of interest in the relationship of canonical Luke to Marcion’s Evangelion I turn now to the question of how Marcion’s Gospel (and/or Marcionitism29) affects my own thinking about Lukan materials.

13.2 Canonical Luke and Marcion’s Evangelion When one takes into account not just canonical Luke but also Marcion’s version of Luke, (referred to from now on as the Evangelion), questions pertaining to how the canonical Gospels reach their final form become more complicated. Though

27. Elaine Pagels, ‘Visions, Appearances and Apostolic Authority: Gnostic and Orthodox Traditions’, in Gnosis. FS Hans Jonas, ed. Barbara Aland (Göttingen: Vandenhoeck & Ruprecht, 1978), 415–30; Elaine Pagels, The Gnostic Gospels (New York: Random House, 1979), 3–32. 28.  Though it should be noted that while Acts anticipates Irenaeus in terms of the link between flesh and apostolic authority, this motif is absent from Justin. Canonical Luke and Justin share the understanding that only the Twelve qualify as apostles; that the mission of the apostles originates in Jerusalem and spreads outwards to the nations (1 Apol. 39.3); and that Jesus appeared to the Apostles to teach them to read scripture in a manner enabling them to understand the fulfilment of scripture in Jesus’ suffering and resurrection (see note 16 above, and 1 Apol. 67.8; Dial. 53.5; 76.6; 106.1). Yet, in contrast to Luke, where each invocation of Jesus’ fleshly resurrection is tied to an assertion of the authority of the twelve, none of the references in Justin to the teachings granted to the Apostles by the resurrected Jesus lay explicit emphasis on the fleshliness of that resurrection appearance (unless one grants that the treatise on resurrection attributed to Justin is authentic rather than pseudonymous. See Pseudo-Justin, Res. 9.5–9). See further, Matthews, ‘Fleshly Resurrection’. 29. Term used to acknowledge ideas that would have been ‘in the air’ at the time Acts is composed, whether or not they originated with Marcion himself, by Walter Schmithals, The Office of Apostle in the Early Church, trans. John E. Steely (Nashville: Abingdon, 1969), 272–86; trans. of Das kirchliche Apostelamt: Eine historische Untersuchung, Forschungen zur Religion und Literatur des Alten und Neuen Testaments (FRLANT) 61 (Göttingen: Vandenhoeck & Ruprecht, 1961). The usefulness of the term is that it allows us to think through some of the issues that consolidate in Marcion’s own teaching, but that may have circulated among Jesus believers before Marcion begins teaching in Rome, c. 150 CE.

254

Shelly Matthews

any reconstruction of the Evangelion remains even more in the realm of the hypothetical than the establishment of the original texts of any of our canonical gospels, portions of the Evangelion can be reasonably reconstructed, on the basis of our sources – especially Tertullian, Epiphanius and Adamantius – in view of the fact that these authors are reading the Evangelion and citing it with some care. This work is facilitated by recent publications containing elaborate reconstructions of Marcion’s text, including those of Jason BeDuhn and Dieter Roth.30 We turn now to consider how the synoptic problem shifts when the Evangelion as well as canonical Luke is brought into the mix. While many scholars accept the assertions of the heresiologists that Marcion knew each of the four canonical gospels, but chose Luke alone from the four as his Evangelion, and then proceeded to mutilate it so that it would conform to his heretical teaching, a number of recent scholarly arguments has called into question this ancient framing.31 Andrew Gregory has observed that while Marcion seems to have subjected the text of the Gospel he had before him to some editing, his redactional tendencies generally bend in the direction of conservation rather than expurgation or mutilation.32 Jason BeDuhn, after voicing suspicion that the charges that Marcion mutilated canonical Luke conform quite neatly to the heresiological trope that orthodoxy precedes heresy, asks further what possible material evidence Irenaeus and Tertullian could have for ascertaining that Marcion mutilated a preexisting text. BeDuhn and Klinghardt also note that one might defend Marcion from the charge of radically redacting or mutilating canonical Luke on grounds that echo a complaint made by Tertullian – namely, if Marcion did indeed attempt to redact canonical Luke so that it could stand in conformity to his ideological agenda, then he was a very incompetent redactor. It is possible to compile a very long list of passages in the Evangelion that seem to be at odds with marcionite ideology in so far as it can be reconstructed, and also passages present in canonical

30.  Jason D. BeDuhn, The First New Testament:  Marcion’s Scriptural Canon (Salem OR:  Polebridge Press, 2013). Dieter T. Roth, The Text of Marcion’s Gospel, New Testament Tools, Studies, and Documents (NTTSD) 49 (Leiden: Brill, 2015). See also Klinghardt, Das älteste Evangelium. 31.  Which is not to suggest that these recent arguments break ground that has never been ploughed. As BeDuhn notes (The First New Testament, 78), the question of the priority of canonical Luke or the Evangelion ‘is one of the founding debates within modern biblical studies’. For his history of scholarship see BeDuhn, First New Testament, 78–92. See also the review of nineteenth-century scholarship on the debate over Marcionite or Lukan priority, which contextualizes it within the agenda of the Tübingen’s school’s concern about how ‘Pauline’ Luke might be understood to be, in Judith Lieu, ‘Marcion and the Synoptic Problem’, in New Studies in the Synoptic Problem: Oxford Conference, April 2008: Essays in Honor of Christopher M. Tuckett (Leuven: Peeters, 2011), 731–51. 32.  Andrew Gregory, The Reception of Luke and Acts in the Period before Irenaeus. Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) II/169 (Tübingen: Mohr Siebeck, 2003), 173–209.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 255

Luke that would have aided Marcion’s ideology, but appear to be missing from the Evangelion. In the words of Matthias Klinghardt, the difficulty in understanding the Evangelion as a redaction of canonical Luke is that Marcion’s version lacks ‘an editorial concept’.33 If it is accepted that Marcion’s text is not the product of his extensive mutilation of canonical Luke, a Gospel he supposedly chose from a preexisting four gospel canon for the purpose of challenging Christian orthodoxy, then it is not necessary to regard the Evangelion as posterior to, and derivative of, canonical Luke.34 How then, to understand the relationship between these two versions of the Gospel? Two possible options remain: either canonical Luke is explained as Luke’s expansion of the Evangelion (recently suggested by Matthias Klinghardt; compare also Joseph Tyson, John Knox);35 or, both canonical Luke and the Evangelion are best explained as Gospels that have arisen from an earlier form of Luke, which was available to both authors. This latter position has been recently argued by Jason BeDuhn, a position which concurs, at least in general outline, with that of Judith Lieu in her new monograph on Marcion.36 As to the question of whether canonical Luke is finalized in direct response to Marcion, Lieu’s observations are tentative and cautious: On the one hand, Marcion did edit the version of the written Gospel that he received, although arguably not to such an extent as his opponents believed, and as might appear from a comparison between a reconstruction of his ‘Gospel’ and

33. Jason BeDuhn, ‘The Myth of Marcion as Redactor: The Evidence of Marcion’s Gospel against an Assumed Marcionite Redaction’, Annali di storia dell’esegesi 29 (2012):  21–48. Klinghardt, ‘Marcionite Gospel and the Synoptic Problem’, 7; Klinghardt, Das älteste Evangelium 1.163–74. Roth stands apart from those who argue for priority of the Evangelion over canonical Luke, arguing that more research is required before passing judgement. Consider his remarks with regard to Luke 4.43 and 16.16 in The Text of Marcion’s Gospel, 437–8. 34. This argument has implications for the question of when the four-gospel collection emerged, since those who accept that Marcion mutilated Luke frequently assume he had a four-gospel canon to repudiate. On the question of when the four gospels began to circulate as a collection, see Harry Y. Gamble, ‘The New Testament Canon: Recent Research and the Status Quaestionis’, in The Canon Debate, ed. Lee Martin McDonald and James A. Sanders (Peabody, MA: Hendrickson, 2002), 267–94, esp. 276–82. 35.  Klinghardt, ‘Marcionite Gospel and the Synoptic Problem’; Klinghardt, Das älteste Evangelium; Tyson, Marcion and Luke-Acts; John Knox, Marcion and the New Testament: An Essay in the Early History of the Canon (Chicago: University of Chicago Press, 1942). 36. BeDuhn, First New Testament. BeDuhn, tracing the history of scholarship on this question understands himself to be reviving the late eighteenth-century thesis of Johann Semler with this argument. Judith Lieu, Marcion and the Making of a Heretic: God and Scripture in the Second Century (Cambridge: Cambridge University Press, 2015), esp. 196–209. Cf. also Gregory, Reception of Luke and Acts, 193.

256

Shelly Matthews

canonical Luke. A consequence . . . must be that canonical Luke is itself the result of redactional development subsequent to the form known by Marcion, both on the level of textual variants and probably also of more extensive passages; however, that any such development was specifically directed against Marcion, although a matter for investigation, would be difficult to demonstrate.37

Conceding Lieu’s assessment that demonstrating this particular redactional tendency is difficult, but also taking to heart her affirmation that such a question is indeed ‘a matter for investigation’, I make the cautious observations that follow. Joseph Tyson’s argument that Luke 1 and 2 have been added to an earlier core Luke for purposes of refuting marcionite teaching merits further consideration. Tyson has noted the thick clustering of vocabulary that is distinctive to canonical Luke in these first two chapters. He notes further that the emphasis on the fleshly incarnation, along with the Jewish markers of the scene – including the circumcision of Jesus, the boy Jesus’ direct words concerning his Father while in the Jerusalem temple and the myriad of Septuagintal echoes – make these chapters suitable for refuting Marcion’s version of Christianity.38 In a previous publication, I have augmented Tyson’s arguments by noting that the emphasis in Luke 1 and 2 on the militant, battle-ready attributes of this God of Israel, intertwined with the announcement of Jesus as savior, serve well to refute any distinction made in marcionite thinking between the High God of Jesus on the one hand, and the warlike god of the Old Testament on the other.39 That is, the issues relating to Luke 1 and 2 seem to constitute a special case. They are clearly not a part of the Evangelion, they share features that set them apart from the remainder of the canonical gospel, and they contain a high condensation of assertions that are antithetical to marcionite claims.40 37. Lieu, Marcion, 209, my emphasis. 38.  See here Tyson, Marcion and Luke-Acts, 90–100. For arguments that the Lukan preface 1.1–4 is interlaced with anti-marcionite elements, see Klinghardt, Das älteste Evangelium, 1.154–159. 39.  Matthews, Perfect Martyr, 43–53. Consider, for instance, the common recognition that the canticles of Luke resemble Maccabean battle hymns in form and content; the echoes of Jael and Judith in Elizabeth’s acclamation of Mary; the heavenly army [stratia] attendant on the annunciation to the shepherds. 40. Consider even F. F. Bruce on the Lukan prologue as anti-Marcionite, ‘Some Thoughts on the Beginning of the New Testament Cannon’, Bulletin of the John Rylands University Library of Manchester 65.2 (1983):  37–60, esp. 54. While the view that canonical Luke chs 1 and 2 are a late redaction in direct response to Marcionite concerns is not widely held among biblical scholars, it should be noted that a wider number have recognized that these two chapters stand somehow apart from the main body of the gospel. Among those who recognize that chs 1–2 may have been composed later than the rest of the third gospel are Joseph Fitzmyer, who argues: ‘the Lucan infancy narrative was composed with the hindsight not only of the gospel tradition prior to Luke, but also of the Lucan Gospel proper”. See The Gospel According to Luke I-IX, Anchor Bible (AB) (New York: Doubleday, 1981), 306.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 257

On the other hand, in my judgement, canonical Luke 24 is not best understood as directly and solely aimed against Marcion.41 This argument owes to my conclusions above on scripture interpretation, Luke’s distinctive attention to fleshly resurrection and the manner in which witnessing fleshly resurrection is linked to the authority of the twelve. Consider again the emphasis in Luke 24 on scripture fulfilment, especially as articulated in verses 27 and 44–45. These passages are unattested, and likely absent from Marcion’s Evangelion. The affirmation that Jesus fulfils ‘everything’ written in Moses and the prophets stands in opposition to Marcion’s disregard for these scriptures, and thus could serve as polemical refutation of Marcionitism.42 But as argued above, more is being affirmed in these programmatic statements than scripture fulfilment alone. That the scriptures prove most particularly that it was necessary for the Christ to suffer is not an obvious plank in an anti-Marcionite agenda since Marcion seems not to have disputed Christ’s suffering. The assertion that scriptures have a ‘mind’ and that the reader requires an expert guide to their meaning do not seem to be a specifically anti-Marcionite trope, but rather serve a more general rhetorical purpose for the author of Luke-Acts, as they do for Justin or Ireneaus. Second, consider again the matter of the resurrected Jesus’ appearance specifically in flesh and bone (24.36–43). A key difficulty in understanding this passage as owing solely to anti-Marcionite concerns (or for that matter, in understanding Marcion’s Evangelion 24 as a direct refutation of canonical Luke 24) is that considerable overlap exists between these resurrection narratives, in so far as Marcion’s text can be reconstructed.43 Granted, the Evangelion at 24.39 seems to have included neither an invitation to touch, nor the reference to flesh. Yet, even in the Evangelion the resurrected Jesus asserts that he retains his bones, a puzzling acknowledgement of solidity, if two communities are here imagined to be duelling over a question of spiritual vs. fleshly resurrection.44 Further, the insistence in canonical Luke that the resurrected Jesus eats (tied so closely in Luke to the assertion of fleshly resurrection) may not be understood as a rhetorical

41. Pace Tyson, Marcion and Luke-Acts, 100–9. 42. Klinghardt, Das älteste Evangelium 1.171. 43.  For discussion on possible reconstructions here, see BeDuhn, First New Testament 196–200; Roth, Text of Marcion’s Gospel, 182–4; 267–9; 345–6; 393–5; Klinghardt, Das älteste Evangelium, 1.166–69. 44.  The certainty of the bones in the reconstruction of the Evangelion is based on Tertullian’s remarks pertaining to this passage at Marc. 4.43.6-7. Here is an instance in which Tertullian charges that while Marcion should have expunged the phrase to fit his theology, he retained it as a means of deceiving his readers into thinking he has not mutilated Luke. See BeDuhn, First New Testament, 127; Roth, 182–4, 345–6; 393–5, 435. Roth even allows it possible that ‘flesh’ might also have been in Marcion’s text.

258

Shelly Matthews

counter to Marcion’s understanding of the resurrection since Marcion’s Jesus seems to eat as well.45 Moreover, when canonical Luke moves from the insistence on flesh in the earthly appearance to the story of the ascent into the heavens, a Marcionite backdrop loses further explanatory value. Marcion’s view of the resurrected body as a heavenly body might stand in some tension with the affirmation of Luke 24.39.46 But, as Turid Karlsen Seim has demonstrated, Acts asserts the fleshliness of Jesus only in his earthly appearances, and not after his ascension. In so far as Luke imagines the ascended Jesus to bear a body suited for the heavens, Lukan and Marcionite views of Jesus’ resurrection are not so disparate.47 In short, while canonical Luke and the Evangelion seem to reflect radical disagreements over the nature of Jesus’ entry into the world, it is more difficult to see radical difference in their stories of Jesus’ ultimate fate, both in terms of the specific substance of the resurrected body and his ascension into the heavens. Thus, it 45. According to Tertullian, Marcion holds that while Jesus was not of the flesh, he, like the angels in Genesis, assumed a state of ‘putative flesh’ and ate as they did (Marc. 3.9). Both BeDuhn (First New Testament, 127) and Klinghardt (Das älteste Evangelium 2.1155–1156) citing Eznik, De Deo 407, include the eating of broiled fish in their reconstructions of the Evangelion 24.41–43. Roth’s reconstruction of the Evangelion acknowledges the fish eating narrative as tentative, but not certain. See The Text of Marcion’s Gospel, 436. 46.  Although it is not even clear whether Marcion would have averred on the question of a fleshly resurrected Jesus per se. As Judith Lieu has noted, Tertullian’s complaints about Marcion’s views of Jesus’ flesh suggest that a central debate between Marcion and opponents was not framed in terms of flesh vs. spirit, but rather involved defining ‘what constituted “flesh”, and specifically the flesh of human beings’. Lieu, Marcion, 372–80. Furthermore, as I argue (‘Fleshly Resurrection’), even for Luke the flesh of Jesus seems not to be the ordinary flesh of humankind. 47. Turid Karlsen Seim, ‘ The Resurrected Body in Luke-Acts: The Significance of Space’, in Metamorphoses: Resurrection, Body and Transformative Practices in Early Christianity, ed. Turid Karlsen Seim and Jorunn Øklund (Berlin: de Gruyter, 2009), 19– 39. Though Luke offers no narrative details on how Jesus’ body is transformed as it moves from earth to heaven, and the substance of that heavenly body is never a point of emphasis or contention, references in Acts to Jesus communicating from the heavenly sphere after the ascension assume that Jesus bears a body appropriate to this sphere. In this twofold staging of resurrection, first in flesh, and then in heavenly body, Luke’s ideas of afterlife may be considered as roughly analogous to those in 4 Ezra 7.26–115, and 2 Baruch 50–51. See Outi Lehtipuu, ‘Biblical Body Language: The Spiritual and the Bodily Resurrection’, in Anthropology in the New Testament and its Ancient Context: Papers from the EABS Meeting in Piliscaba/Budapest, ed. Michael Labahn and Outi Lehtipuu (Leuven: Peeters, 2010), 151–68, esp. 158. Apothesosis narratives often include the detail that mortal flesh melts away as a body is transported from the earthly to heavenly sphere. See, for example, Ovid, Metam. 15.824–28; compare also Philo’s description of Moses’ translation, Mos. 2.228.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 259

is too simple to argue that the Evangelion and canonical Luke represent a quarrel over the binary of ‘spiritual’ vs. ‘fleshly’ resurrection.48 Further, consider how canonical Luke 24, along with linked passages in Acts, foreclose the apostolate to the twelve who saw the resurrected Jesus in the flesh. Granted, this foreclosing requires Luke to strip from Paul his understanding of apostolicity, and thus subordinates him to the Jerusalem apostles in a manner Marcion would not countenance. But Lukan materials linking apostolic authority to fleshly resurrection appearances are not only excluding Marcion’s beloved Apostle Paul from the list of those receiving legitimating appearances but they also exclude James, the five hundred, and ‘all of the apostles’ (1 Cor 15.5–8); Mary and the women (Luke 24.1–11; Matt 28.1,9, Mark 16.1, John 20.11–18); and Cleopas and his companion (Luke 24.13–35). Especially in the assertion that the risen Christ was not made to appear to all the people but only to the foreordained witnesses, these Lukan materials seem to be aimed beyond one particular opponent. They challenge, rather, a notion expressed across a wide range of materials, including materials eventually canonized, that authority to speak on behalf of the resurrected Christ is granted to the visionary (Gal 1.12; 1 Cor 9.1; John 20.14–17; Rev. 1.10–19; cf. Gospel of Mary; Dialogue of the Savior; Pistis Sophia).49 Finally, the privileging of Peter among the apostles as the first to whom the fleshly Jesus appeared, and as spokesman in three of the four passages in Acts in which fleshly resurrection is invoked, does not point in any obvious way to a marcionite backdrop. This feature, taken together with the diminished agency of Mary Magdalene in canonical Luke, calls to mind a number of extra-canonical texts in which quarrels are staged between Peter and Mary Magdalene concerning her vision and speech. That traditions pertaining to Mary and the women, and not Marcion, fuel Peter’s assertion that Jesus was not made manifest to all the people, is further suggested by redactional tendencies in Luke-Acts to modify traditions

48.  Pace Markus Vinzent, ‘Der Schluß des Lukasevangeliums bei Marcion’, in Marcion und seine kirchengeschichtliche Wirkung:  Vorträge der Internationalen Fachkonferenz zu Marcion, gehalten vom 15.-18. August 2001 in Mainz, ed. Gerhard May and Katharina Greschat, Texte und Untersuchungen zur Geschichte der altchristlichen Literatur (TUGAL) 150 (Berlin: de Gruyter, 2002), 79–94, esp. 86. 49.  For arguments that John 20.14–17 parallels Rev. 1.10–19 in form, and the Pauline appearance traditions of 1 Corinthians in function, see Mary Rose D’Angelo, ‘ “I have seen the Lord”. Mary Magdalen as Visionary, Early Christian Prophecy and the Context of John 20.14-18’, in Mariam, the Magdalen, and the Mother, ed. Deirdre Good (Bloomington, IN:  Indiana University Press, 2005), 95–122. For the exalted function of Mary as visionary in Gnostic texts, see Jane Schaberg, The Resurrection of Mary Magdalene:  Legends, Apocrypha and the Christian Testament (New  York:  Continuum, 2004), 146–9, 168–77; Dierdre Good, ‘Pistis Sophia,’ in Searching the Scriptures: A Feminist Commentary Vol. 2, ed. Elisabeth Schüssler Fiorenza (New York: Crossroad, 1994), 678–707.

260

Shelly Matthews

pertaining to leading women in the Jesus movement in the direction of subordination and silence.50 In summary, Luke 1–2 may be readily understood as a challenge to major components of Marcion’s teaching – that Jesus did not gestate in a womb, was not a Jew and is the emissary of the High God who has nothing to do with the Old Testament God and his scriptures. But the issues raised in Luke 24 point to a different array of concerns. Luke 24 and Evangelion 24 share significant overlap and the distinctively Lukan material is best explained by other means than as owing to Marcionite polemic alone. For this reason I gravitate towards the premise of a core Luke, taken over and expanded in different directions in Marcionite circles on the one hand, and in ‘Lukan’ circles on the other. Perhaps one of the concerns of the Lukan textual community is refuting Marcionitism. But marshalling such a refutation is not the sole concern of the final redactor.

13.3 A ‘Marcionite Solution’ to the Synoptic Problem? From the standpoint that canonical Luke is a late redaction, and that both canonical Luke and the Evangelion are reworking an earlier ‘core Luke’, the question may now be asked what insights into the synoptic problem might be gleaned through introducing the Evangelion into the debate – that is, by holding both the Evangelion and canonical Luke alongside of the Gospels of Mark and Matthew when trying to account for overlaps and discrepancies among them. Here, I  engage directly two arguments made by Jason BeDuhn,51 pertaining to how the synoptic problem might be answered differently if the Evangelion is taken into consideration, namely the issue of the minor agreements, and the explanation for the overlapping narrative material in the double tradition concerning John the Baptist and Jesus’ temptations. Consider first the question of the minor agreements. A  number of scholars who challenge the Two-Source hypothesis have held that the minor agreements of Matthew and Luke against Mark in the triple tradition are the shoals on which the hypothesis founders.52 BeDuhn poses a serious challenge to this view in his 50. On Luke’s redactional tendencies concerning women, see Matthews, ‘Fleshly Resurrection’, 169–72; Jane Schaberg and Sharon H. Ringe, ‘Gospel of Luke’, in Women’s Bible Commentary rev. and enl., ed. Carol A Newsom, Sharon H. Ringe and Jacqueline E. Lapsley (Louisville, KY: Westminster John Knox, 2012), 493–511; Barbara Reid, Choosing the Better Part?: Women in the Gospel of Luke (Collegeville, MN: Liturgical Press, 1996); Marianne Sawicki, Seeing the Lord: Resurrection and Early Christian Practice (Minneapolis: Fortress, 1994), 150–67; Elisabeth Schüssler Fiorenza, ‘A Feminist Critical Interpretation for Liberation: Martha and Mary: Luke 10.38–42’, Religion and Intellectual Life 3 (1986): 21–35. 51.  BeDuhn, First New Testament, 92–7; BeDuhn, ‘Marcion’s Evangelion and the Two Source Hypothesis’, forthcoming. 52.  For instance, Klinghardt (‘Synoptic Problem’, 4)  considers the minor agreements ‘fatal to the Q hypothesis’.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 261

observation that the Evangelion appears to share considerably fewer of the minor agreements than are shared between Matthew and the current critical text of Luke; and, conversely, that the Evangelion possesses some harmonizations with Matthew that are not found in canonical Luke.53 From these observations, BeDuhn concludes that all such minor agreements arise as secondary scribal harmonizations, rather than owing to compositional relationships. To arrive at these conclusions, BeDuhn takes special notice of the 52 minor agreements deemed significant by S.  McLoughlin,54 and asks what form of the passage in question might be reconstructed from witnesses to Marcion’s text. After striking 21 instances from McLoughlin’s list of 52 because a reading for the Evangelion is either unattested or too vague to be determined, BeDuhn concludes: ‘Of the remaining 31 cases the Evangelion has the minor agreement in 15 and does not have the agreement in 16.’55 Because reconstructing Marcion’s text is more an art than a science, some might assess the evidence differently and dispute BeDuhn’s specific numerical tally here. This may be so particularly in instances where his judgements made about the ‘absence’ of a minor agreement in the Evangelion might be regarded by others as owing only to the silence of Tertullian or Epiphanius about its presence in the text they read and reproduce. Yet, whether it is held that the number of cases in which the Evangelion reads differently from canonical Luke is precisely 16 cases, or a smaller number, I find BeDuhn’s overarching conclusions with regard to these significant minor agreements to be sound: the Evangelion bears witness to the scribal propensity to harmonize readings, once written gospels are able to be compared as they are copied.56 These conclusions significantly weaken arguments against the two-document hypothesis based upon the minor agreements.

53. The fact that Matthew agrees sometimes with canonical Luke and sometimes with the Evangelion is one ground for BeDuhn’s argument that canonical Luke and the Evangelion are both reworking a core Luke. 54.  S. McLoughlin, ‘Les accords mineurs Mt-Lc contre Mc et le problème synoptique. Vers la thèorie des deux sources’, Ephemerides Theologicae Lovanienses (ETL) 43 (1967): 17–40. 55. BeDuhn, ‘Marcion’s Evangelion’. 56. It is beyond the scope of this paper to reproduce and assess each of the 52 significant minor agreements tested by BeDuhn against his reconstruction of the Evangelion. I cite here two of the more readily reproducible cases, as an illustration of the process by which these assessments are made. (1) Where Mark 9.19 reads ὦ γενεὰ ἄπιστος, Matt 17.17 and Luke 9.41 read ὦ γενεὰ ἄπιστος καὶ διεστραμμένη. Neither Tertullian (Marc. 4.23.1; Marc. 4.23.2) nor Epiphanius (Scholion 19) reproduce διεστραμμένη. (2) Where Mark 5.27 reads ἥψατο τοῦ ἱματίου αὐτοῦ, Matt 9.20 and Luke 8.44 read ἥψατο τοῦ κρασπέδου τοῦ ἱματίου αὐτοῦ. Neither Tertullian (Marc. 4.20.8) nor Epiphanius (Scholion 14) reproduce τοῦ κρασπέδου. On this second reconstruction, Roth is considerably more reluctant than BeDuhn to regard an unattested word as an absent word. Cf. BeDuhn, First New Testament, 147 and Roth, The Text of Marcion’s Gospel, 210.

262

Shelly Matthews

Second, consider the question of pericopes pertaining to the preaching of John the Baptist and the temptations of Jesus found in canonical Luke (3.7–9; 3.16b– 17; 4.1–13), and commonly assigned to Q in reconstructions of the source.57 The Evangelion begins with Luke 3.1, ‘in the fifteenth year of Tiberius Caesar, when Pilate was governing Judea’, and then proceeds to narrate the movements and teachings of Jesus in Capernaum and Nazareth, without any reference to John’s inaugural sermon, the baptism, or the temptation. This leads to the question whether these passages were present in the ‘core Luke’ known by Marcion and then deleted by him. If so, then Marcion certainly was a ‘mutilator’ as the church fathers supposed, at least as he struggled to compose a proper introduction to his Gospel. If, however, one is open to the suggestion that Marcion is not working quite this fast and loose with his knife, other options are possible. Here again, I turn to a proposal recently offered up by Jason BeDuhn. If Q were to be reconstructed from double tradition shared by Matthew and the Evangelion (rather than by Matthew and canonical Luke) our sayings source would begin, essentially, at 6.20 with the sermon on the Mount/Plain. Further, in this hypothetical reconstruction, the presence of the narratives concerning John’s preaching and the temptations of Jesus in canonical Luke would be understood as part of a later stage of redaction, when the final redactor has a copy of the Gospel of Matthew in hand. An advantage of imagining the prefatory material in canonical Luke as not owing directly to Q, as BeDuhn notes, is that a sayings source that begins at Luke/Q 6.20 can be understood more fully and truly a sayings source, and the Two-Source hypothesis in this form becomes unencumbered of the need to explain how Q comes to include so much narrative material in its preface.58 Further, a hypothetical reconstruction in which Luke 3 and 4 draw on the Gospel of Matthew would also account for the fact that Luke inserts the material of the Double Tradition into the same Marcan context as Matthew only at 3.7–9, 17; and 4.2b–13, but not in other instances.59 In sum, dating canonical Luke to the second century does not necessarily require us to dispense with the Two-Source hypothesis. If the major signs of second-century editorial activity are concentrated at front and back sections of the canonical

57.  The Critical Edition of Q, ed. James M.  Robinson, Paul Hoffmann and John S. Kloppenborg, Hermeneia (Minneapolis: Fortress, 2000), 2–43. 58. BeDuhn, First New Testament, 94–7. These arguments are significantly expanded in BeDuhn, ‘Marcion’s Evangelion’. Kloppenborg accounts for the narrative relating to Jesus’ temptation as conforming to ancient form, whereby sayings of lesser known sages might be prefaced with materials that demonstrate the character of the sage under trial. See for example John Kloppenborg Verbin, Excavating Q: History and Setting of the Sayings Gospel (Minneapolis: Fortress, 2000), 202. 59.  Significantly, at the Roskilde conference on the Synoptic Problem, three of the papers arguing against the existence of Q, and for Luke’s use of the Gospel of Matthew point to Luke 3 and 4 as instances of Luke’s use of Matthew (see contributions of Mark Goodacre, Francis Watson and Werner Kahl in this volume). Compare also Mark Goodacre, The Case Against Q (Harrisburg, PA: Trinity Press International, 2002), 170–85.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 263

narrative,60 it is possible to postulate a core Luke composed decades earlier, through the incorporation of Mark and the sayings source, without access to the Gospel of Matthew. The fact that a number of minor agreements evident between canonical Luke and Matthew are absent from the Evangelion, while the Evangelion sometimes harmonizes with Matthew and/or Mark against canonical Luke, suggests that both canonical Luke and the Evangelion moved through various, independent stages of harmonization. These independent harmonizations challenge the argument that the minor agreements of Matthew and Luke against Mark in the triple tradition serve to undo the Two-Source hypothesis. Further, BeDuhn’s proposal that we might profitably reconstruct the Q document from the shared materials of the Gospel of Matthew and the Evangelion, rather than canonical Luke, merits further consideration. Such a proposal would dramatically reduce the length of the hypothetical document and also complicate theories of the relationship between Q, Luke and Matthew. Yet, the proposal is advantageous in providing a way of accounting for common agreement without common order (Luke/Q 6.20 and following), and common ordering of the preaching of John the Baptist and the temptations Luke 3.7–9,17; 4.2b–13. To be sure, understanding canonical Luke as ‘core Luke’ plus additions clustered primarily at the beginning and end of the text moves us quite deeply into the realm of the hypothetical, in a way that positing a solution to the synoptic problem on a relatively more secure textual base might not appear to do. This chapter makes no claim that the hypothesis unfurled here for consideration is a particularly elegant solution to the problem. But in defence of this inelegant proposal, I note that a number of hypotheticals have been proposed at the Roskilde conference, perhaps inevitably, given our increasing awareness of (1) the fluidity of the textual tradition;61 (2) the active role of authors, redactors and subsequent scribes in shaping 60.  For an argument, based on the materiality of classical manuscripts, that the endings of ancient texts were ready sites for editorial intrusion, see Stephanie West, ‘Terminal Problems’, in Hesperos; Studies in Ancient Greek Poetry Presented to M. L. West on His Seventieth Birthday (2007), 3–21. 61.  Consider for instance, David Parker, The Living Text of the Gospels, for demonstration that the last three chapters of Luke were not fixed for several centuries after its ‘original’ composition; William L. Petersen, ‘The Genesis of the Gospels’, in New Testament Textual Criticism and Exegesis. FS J. Delobel, ed. A. Denaux, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 169 (Leuven:  Peeters, 2002), 33–65, on the wide gap that must be acknowledged to exist between our critical editions and the text that was current in the second century; Arthur J. Bellinzoni, ‘The Gospel of Luke in the Apostolic Fathers: An Overview’, in Trajectories through the New Testament and the Apostolic Fathers, ed. Andrew F. Gregory and Christopher M. Tuckett (Oxford: Oxford University Press, 2005), 45–66, for challenging the hypothesis that archetypes of textual tradition are identical with autographs, in view of the substantial revisions that must have occurred during the first 100  years of transmission; and Robbins, ‘A Socio-Rhetorical Look’, who, working from his socio-rhetorical method, and engaging with the work of Bakhtin, argues that a late dating of canonical Luke-Acts makes more possible the ‘socio-ideological conversations’ that seem necessary in the formation of the text.

264

Shelly Matthews

materials known to them62 and (3) the fact that ‘the canonical Gospels in their current form do not hold all the answers to a final solution.’63

13.4 Concluding Observations The following concluding reflections may be overly optimistic about the direction in which the field of biblical scholarship might move. Still, in conclusion, hope is expressed that a clearer recognition of the nature of the shifting sands on which we all build hypotheses about the relationships among the canonical gospels might prompt biblical scholars to consider whether it is a reasonable and worthy exercise to continue to frame our hypotheses solely within the bounds of the historicalcritical method. Exclusive embrace of this method, with its rhetoric of scientific rigor and objectivity, positions our field in the past – solely within the bounds of the discourses of Enlightenment modernity, untouched by the postmodern, poststructuralist, (including feminist) recognition that all knowledge is situated, that the historical narratives we construct are contingent and perspectival, and that especially because we deal with something as theologically/ideologically weighted as Christian beginnings, our decisions on how to reconstruct these beginnings have consequences.64 To be sure, embrace of such a poststructuralist frame does not require the abandonment of the historical-critical method. It is possible from ‘the other side of the linguistic turn’, to still employ historical-critical tools, to make judgements about historical plausibility and to offer up one’s own particular reading of an ancient text as most persuasive. Further, recognition of one’s biases does not necessarily mean they are false.65 But the poststructuralist frame – at least as it has been employed by certain feminist, minoritized and postcolonial subjects – calls scholars to a heightened awareness of their own subjectivity. Such framing requires one to articulate, in so far as one’s partial vision allows, how one’s own

62.  Cf. Mogens Müller’s contribution to this volume; also Klinghardt, ‘Marcionite Gospel’, 26–7. 63. Lieu, ‘Marcion and the Synoptic Problem’, 746. 64. On this point, see, among others, Elisabeth Schüssler Fiorenza, Rhetoric and Ethic: The Politics of Biblical Interpretation (Minneapolis: Fortress, 1999); Elisabeth Schüssler Fiorenza, ‘Revisioning Christian Origins: In Memory of Her Revisited’, in Christian Beginnings: Worship, Belief and Society, ed. Kieran O’Mahoney (London: Continuum, 2003), 225–50; Hal Taussig, ‘The End of Christian Origins? Where to Turn at the Intersection of Subjectivity and Historical Craft’, in Review of Biblical Literature (RBL) 13 (2011): 1–46; Shelly Matthews, Perfect Martyr, 20–4; Matthews, Acts of the Apostles, ix–xii. 65. A point made with some eloquence by William E. Arnal, ‘Making and Re-Making the Jesus Sign: Contemporary Markings on the Body of Christ’, in Whose Historical Jesus?, ed. William E. Arnal and Michel Desjardins, Studies in Christianity and Judaism 7 (Waterloo, ON: Wilfrid Laurier University Press, 1997), 308–19.

13. Does Dating Luke-Acts into the Second Century Affect the Q Hypothesis? 265

perspectives and values shape ones readings, and the ethical consequences of the reconstructions one pursues.66 William Arnal, in a recent essay, offers up one model for the direction this reflection on what is at stake in our historical constructions of early Christianity might take, with respect to historical Jesus scholarship.67 He suggests that reconstructions of Jesus as anti-cultural, anti-imperial and as reacting to the antihumane aspects of empire reflect our common contemporary experience: ‘Greater forces than any single person can master or even conceptualize are changing the world, and it is consequently difficult to feel this force of change as anything but dehumanizing.’ Thus he locates in this Jesus scholarship ‘a passionate appeal for a renovated, more humane world . . . filtered through a persuasive and poetic reconstruction of our collective past’.68 Writing more directly about the Q hypothesis, Melanie Johnson-DeBaufre has argued that ‘what is at stake in debates over Q is Christian identity’, and that in spite of recognition that early Christianity was diverse, most Q studies seem interested in ‘countering a singular authoritative view of Christian identity with another singular authoritative version of Christian identity’. Her own study counters this rhetoric by offering up a reading of Q texts that give voice to ‘communal traditions of solidarity’, as a more ethically preferable framing within a context of religious pluralism.69 The arguments laid out here are based on my close reading of primary sources, aided by the scholarship produced by historical critics I have judged to be most persuasive. Yet I still recognize that my arguments for canonical Luke as a fluid text reaching some stability, but not perfect stability, in the second century within the context of early Christian authority debates, along with my arguments that Lukan scriptural materials were shaped not by the singular hand of ‘Luke’ the evangelist, but by many hands in conversation and contest do resonate with the sort of Christianity that appeals to me. Namely, a Christianity that is more fluid, less dogmatic and open to a variety of voices beyond the hegemonic voices of the canon.70 I would invite all readers engaged in the question of the synoptic problem to reflect on how their perspectives and desires inform the scholarship they produce.

66. Though proceeding from a different standpoint than the one articulated here, John S. Kloppenborg makes comparable claims in his ‘Conceptual Stakes in the Synoptic Problem’, in this volume. 67. Arnal, ‘Making and Re-Making the Jesus Sign’. 68. Ibid., 318–19. 69.  Melanie Johnson-DeBaufre, Jesus Among Her Children:  Q, Eschatology and the Construction of Christian Origins, Harvard Theological Studies (HTS) 55 (Cambridge, MA: Harvard University Press, 2005), 41–4. 70.  Consider also Kloppenborg’s observations on the appeal of the Two-Document Hypothesis, as one that ‘leads to a more variegated picture of the earliest stages of the Jesus movement’, while the Two-Gospel Hypothesis ‘asserts a high degree of continuity between the historical Jesus and the practices and dogmatic views of the early Jesus movement and a high degree of uniformity among various streams of primitive Christianity thereafter’.

Chapter 14 M A R C IO N ’ S G O SP E L A N D T H E S Y N O P T IC P R O B L E M I N R E C E N T S C HO L A R SH I P Dieter T. Roth

14.1 Introduction The resurgence in studies of Marcion and Marcion’s Gospel over the past decade has highlighted the continuing relevance of this figure and his Gospel text for numerous questions related to New Testament (NT) studies and the history of early Christianity as well as the ongoing scholarly disagreements concerning nearly every aspect of Marcion and Marcion’s Gospel.1 One relevant issue concerning which the debate is particularly noteworthy is the question of the place and significance of Marcion’s Gospel in the consideration and study of the Synoptic Problem. Though the question of the relationship between Marcion’s Gospel and the Gospel of Luke, along with the significance of Marcion’s Gospel for understanding the relationship between the Synoptic Gospels more generally, has often played a

1. The renewed interest in Marcion, and especially Marcion’s Gospel, is particularly evident in the number of monographs and articles that have been published over the past 10 years. Major monographs include Joseph B. Tyson, Marcion and Luke-Acts: A Defining Struggle (Columbia:  University of South Carolina Press, 2006); Sebastian Moll, The ArchHeretic Marcion, Wissenschaftliche Untersuchungen zum Neuen Testament (WUNT) 250 (Tübingen:  Mohr Siebeck, 2010); Jason BeDuhn, The First New Testament: Marcion’s Scriptural Canon (Salem:  Polebridge, 2013); Markus Vinzent, Marcion and the Dating of the Synoptic Gospels, Studia Patristica Supplement (StPatrS) 2 (Leuven:  Peeters, 2014); Matthias Klinghardt, Das älteste Evangelium und die Entstehung der kanonischen Evangelien, 2 vols, Texte und Arbeiten zum neutestamentlichen Zeitalter (TANZ) 60 (Tübingen, Francke: 2015); Judith M. Lieu, Marcion and the Making of a Heretic: God and Scripture in the Second Century (Cambridge:  Cambridge University Press, 2015); and Dieter T. Roth, The Text of Marcion’s Gospel, New Testament Tools, Studies, and Documents (NTTSD) 40 (Leiden: Brill, 2015).

268

Dieter T. Roth

role in the history of research,2 several recent proposals have attached particular significance to Marcion’s Gospel when studying the canonical Gospels. Though space precludes a full discussion of the range of proposals found in the writings of Markus Vinzent, Matthias Klinghardt, Jason BeDuhn and Judith Lieu, the following discussion seeks to offer a succinct presentation of and brief critical interaction with their recent publications and the views found therein concerning the place of Marcion’s Gospel in the study of the Synoptic Problem.

14.2 Prolegomena A significant challenge confronting any attempt to work on Marcion’s Gospel or to consider its place in discussions concerning the Synoptic Problem is that, in order to do so, one must first reconstruct Marcion’s text. Without a reconstruction, the types of comparisons and arguments required for studies of the Synoptic Problem are not possible. The precise nature of the challenge is actually twofold in that such a reconstruction involves two fundamental questions: (1) Which sources are relevant for reconstructing Marcion’s Gospel? and (2) How are the relevant sources to be read and utilized in the pursuit of a textual reconstruction? In another publication I have provided a more extensive analysis of these issues than can be offered here;3 however, a few, brief comments are in order due to their relevance for the ensuing discussion. First, concerning the relevant sources for Marcion’s Gospel, Lieu has written, ‘The primary resources for any reconstruction of Marcion’s “Gospel” are the efforts of Tertullian and Epiphanius; . . . Other critiques of Marcion’s text are found in more piecemeal fashion, particularly, from Origen onwards, in exegetical contexts.’4 Beduhn rightly notes that the three primary sources are Tertullian, Epiphanius and the Adamantius Dialogue, and also adds that ‘a number of lesser sources provide important confirmation of readings given by the major three, or fill in gaps otherwise left by them’.5 Though I  evaluate some of these lesser sources slightly

2.  For an extensive history of research on Marcion’s Gospel cf. Roth, Text of Marcion’s Gospel, 7–45. Problematic representations of the especially important era of Marcion research in the 1840s and 1850s involving, in particular, Albrecht Ritschl, F. C. Baur, Adolf Hilgenfeld and Gustav Volkmar, are discussed in Dieter T. Roth, ‘Marcion’s Gospel and Luke: The History of Research in Current Debate’, Journal of Biblical Literature (JBL) 127 (2008):  513–27. For the place of these debates within the larger discussion concerning the Synoptic Problem, cf. R. H. Fuller, ‘Baur versus Hilgenfeld:  A Forgotten Chapter in the Debate on the Synoptic Problem’, New Testament Studies (NTS) 24 (1978): 355–70. 3.  Cf. Dieter T.  Roth, ‘The Link between Luke and Marcion’s Gospel:  Prolegomena and Initial Considerations’, ed. John S.  Kloppenborg and Joseph Verheyden, Bibliotheca Ephemeridum Theologicarum Lovaniensium (BETL) 29 (Leuven: Peeters, 2017), 59–80. 4. Lieu, Marcion and the Making of a Heretic, 187. 5. BeDuhn, First New Testament, 34.

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

269

differently than BeDuhn, in particular Pseudo-Ephrem A and 𝔓69,6 my own work on reconstructing Marcion’s Gospel is in basic agreement with these statements by Lieu and Beduhn.7 Methodologically problematic, by contrast, is Vinzent’s comment concerning Marcion’s texts that ‘the Gospel and his Apostolikon (of ten Pauline letters) can be recovered only partially from the glimpses that are given by his opponents, unearthed from their writings (primarily Tertullian, Epiphanius, Adamantius’ [Pseudo-Origen?] Dialogue I–II, and Codex Bezae)’.8 Though one can profitably compare readings attested for Marcion’s Gospel in the polemical writings of the heresiologists with readings found in the NT manuscript tradition, including readings found in Codex Bezae,9 this manuscript is not an opponent of Marcion nor is it a source for Marcion’s Gospel. It is simply a confusion of categories to include this manuscript in the list of heresiologists whose works interact with and thus attest Marcion’s Gospel. Similar problems attend Klinghardt’s attempts to reconstruct Marcion’s text not only upon the basis of the church fathers’ testimony but also on the basis of the manuscript tradition for the Gospel of Luke for passages unattested by the heresiologists.10 Second, when reconstructing Marcion’s scriptural texts, the foundational methodological principle, as already recognized and argued in Urich Schmid’s

6.  Cf. Roth, Text of Marcion’s Gospel, 50 n 18 and Roth, ‘The Link between Luke and Marcion’s Gospel’, 64–68. Lieu also comments that relying upon (Pseudo-)Ephrem as a ‘source for Marcion’s text is questionable’ (Marcion and the Making of a Heretic, 269 n. 112). 7. Cf. Roth, Text of Marcion’s Gospel, 46–9, 83–91, 270–85, 347–58 and 396. 8. Vinzent, Marcion and the Dating of the Synoptic Gospels, 9. 9.  Cf., e.g., Dieter T. Roth, ‘Marcion and the Early Text of the New Testament’, in The Early Text of the New Testament, ed. Charles E. Hill and Michael J. Kruger (Oxford: Oxford University Press, 2012), 307–12 in which I considered several readings that demonstrated both points of contact with, along with clear indications that Marcion’s Gospel was not the same as, the ‘D-text’. 10.  Cf. Klinghardt’s comments concerning the use of both the testimony of the church fathers and the manuscript tradition when reconstructing Marcion’s Gospel in Klinghardt, Das älteste Evangelium, 1:23. The issue in Klinghardt’s work is complicated by his use of the siglum ‘Mcn’ as representative both of Marcion’s Gospel and Klinghardt’s postulated ‘ältestes Evangelium’. For more extensive discussion of Klinghardt’s methodology and use of this siglum, cf. Roth, ‘The Link between Luke and Marcion’s Gospel’, 72–75. A  summary overview of ‘Mcn’ is as follows: Klinghardt first states that he uses this siglum as ‘die Bezeichnung des marcionitischen Evangeliums’ (ibid., 1:24) without in any way assuming that Marcion authored this Gospel. In addition, however, Klinghardt states that though he has avoided the use of certain terms due to the baggage they carry in NT scholarship, terms such as ‘Urevangelium’ or ‘Protevangelium’ are ‘gut denkbar’ for what ‘Mcn’ is and that ‘Protolukas’ or ‘Protomarkus’ would also be ‘angemessen’ (ibid.). What is more, in instances where the attestation of the church fathers differs concerning Marcion’s Gospel, Klinghardt contends that it would be advisable to speak ‘vom vorkanonischen Evangelium anstatt von Mcn [emphasis original]’ (ibid.).

270

Dieter T. Roth

work on Marcion’s Apostolikon, is that if readings found in Marcion’s Gospel are to be gleaned from the ‘citations’11 found in the writings of Marcion’s adversaries ‘müssen wir das Zitierverhalten unserer Quellen möglichst präzise beschreiben, und das geschieht am überzeugendsten, indem man sämtliche Bibelzitate in allen Schriften eines Kirchenvaters untersucht’.12 When an author’s citation habits are not taken into consideration, or at least not adequately taken into consideration, a less than critical reconstruction of Marcion’s texts is the inevitable result. As such, it will be seen in the discussion below that both the issue of a problematic identification of which sources are actually relevant for reconstructing Marcion’s Gospel as well as the problematic use of correctly identified, relevant sources have affected the scholarly discussion of Marcion’s Gospel and the Synoptic Problem.

14.3 Marcion’s Gospel as the Key to Solving the Synoptic Problem In recent works considering Marcion’s Gospel and the Synoptic Problem, Vinzent and Klinghardt have presented quite radical solutions to the Synoptic Problem which, though differing from each other in important respects, have in common the vital role played by Marcion’s Gospel in their proposed solutions. The more daring proposal is that of Vinzent who contends that Marcion was ‘an inspirational source who, in proximity with the authors of the other Gospels, wrote the very first Gospel and, only after having come to know the re-writings of his own text through colleagues, published his (revised?) version of his Gospel, together with the Antitheses and Paul’s letters’.13 Thus, ‘Marcion, who created the new literary genre of the “Gospel” and also gave his work this title, had no historical precedent in the combination of Christ’s sayings and narratives.’14 For Vinzent, Marcion’s Gospel was, ‘through combining prefaces of Acts and Luke, put under the name of Luke. Luke was complemented by Mark, Matthew, and John.’15 Though

11.  Ulrich Schmid makes the important and helpful observation:  ‘Wenn die antimarcionitischen Polemiker versuchen, den Häretiker ex his revinci, quae servavit [Tertullian, Marc. 5.4.2], dann muß dies nicht notwendigerweise bedeuten daß sie seinen Text auch in jedem Fall wörtlich zitieren’ (Marcion und sein Apostolos: Rekonstruktion und historische Einordnung der marcionitischen Paulusbriefausgabe, Arbeiten zur neutestamentlichen Textforschung (ANTF) 25 [Berlin: de Gruyter, 1995], 26). Section numbers for Adversus Marcionem are taken from Contre Marcion, critical text by Claudio Moreschini, trans. René Braun, 5 vols, SC 365, 368, 399, 456, 483 (Paris: Les É ditions du Cerf, 1990–2004). 12. Schmid, Marcion und sein Apostolos, 26. 13. Vinzent, Marcion and the Dating of the Synoptic Gospels, 188. Cf. also Vinzent’s comments on pp. 99–100. 14. Ibid., 277. 15. Ibid., 282.

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

271

Vinzent’s voice is an important one in recent scholarship on Marcion, these proposals have been strongly criticized.16 A particularly glaring problem is that though in the introduction Vinzent indicates that ‘work is progressing on a forthcoming reconstruction (as far as the sources allow) of Marcion’s Gospel together with a Synoptic commentary’, in numerous places in his monograph he provides lengthy citations from Marcion’s Gospel without any indication of the basis for the cited text. In addition, he employs a novel numbering for Marcion’s Gospel that is apparently taken from this still unpublished work.17 In my estimation this is putting the cart before the horse and it is difficult to evaluate Vinzent’s proposal on the basis of a debate concerning the content of Marcion’s Gospel, for Vinzent’s work in this regard has not appeared. In addition, Vinzent largely bases his argument concerning Marcion’s Gospel upon his reading of Tertullian, ‘some elements of which have so far received no attention in scholarship’.18 Three of these elements are:  (1) Tertullian presents

16.  Cf., e.g., Paul Foster, ‘Review of Markus Vinzent, Marcion and the Dating of the Synoptic Gospels’, Journal of Ecclesiastical History (JEH) 66 (2015): 144–5, who bluntly states ‘It is . . . difficult to find his [Vinzent’s] own proposal clearly articulated in the book, or to see how the extensive textual data that he marshals actually supports what he appears to be arguing’ and summarizes, ‘Vinzent’s attempts to redate the Synoptic Gospels to the period after the writing of Marcion’s gospel are entirely unpersuasive. Textual evidence is read in ways that are highly problematic, and the details that are discussed do not lead to the conclusions being drawn. Moreover, the whole argument is disconnected and logically fallacious. As a whole the book is unpersuasive and idiosyncratic throughout.’ Similarly, Clare Rothschild, ‘Review of Markus Vinzent, Marcion and the Dating of the Synoptic Gospels’, Review of Biblical Literature (RBL) (2016), 4 pages, cited 30 April 2016, online:  http://www.bookreviews.org/ pdf/10107_11206.pdf, who comments ‘no argument that Vinzent puts forward effectively overturns the classical arguments of amplification and refinement for Markan priority’ and concludes that the point she is attempting to make ‘is that when evidence is insufficient or worse, when the interpretation of the evidence is verifiable but not falsifiable, then judgments about that evidence are not just unpersuasive or controversial; they are unscientific, that is, they belong to a nonscientific genre such as historical fiction, philosophical speculation, or personal reflection.’ Though Lieu notes that Vinzent’s work ‘appeared too late and is too comprehensive a theory’ to be addressed in her volume, she does let the reader know that she ‘is not persuaded’ (Marcion and the Making of a Heretic, 183 n. 1). 17.  For instance, Vinzent provides a column of Marcion’s text in a table spanning nine pages (Marcion and the Dating of the Synoptic Gospels, 264–72) and it is not until n. 188 on p. 273 that Vinzent informs the reader that the confusing numbering of verses for Marcion’s Gospel used throughout the monograph is taken from his unpublished Marcion’s Gospel: A Synoptic Commentary. 18.  Vinzent, Marcion and the Dating of the Synoptic Gospels, 91. The importance of Tertullian is highlighted on the preceding page when Vinzent writes, ‘The best insight into Marcion’s Gospel is undoubtedly provided by Tertullian’s books, Against Marcion, especially Book IV, . . .’ (ibid., 90).

272

Dieter T. Roth

Marcion as the author/writer (evangelizator) of his Gospel, (2) in contradistinction to presenting Marcion as the author of the Gospel, Tertullian presents Marcion as the redactor of Paul’s epistles, and (3) Tertullian refers to Marcion as a ‘literary innovator in form and content (nova forma sermonis: similitudes and dialogues)’.19 Perhaps, however, there are very good reasons for why a reading along these lines has received ‘no attention in scholarship’.20 First, the interpretation of evangelizator as Gospel writer is less than persuasive. The context in which Tertullian refers to Marcion as an evangelizator clearly has Gal 1.8 as its background (Marc. 4.4.5), where a proclamation is in view. Furthermore, the one other reference to evangelizator (Marc. 5.5.1) and the two references to evangelizatores (Marc. 5.7.11; 5.19.5) are rather unambiguously references to individuals proclaiming, and not writing, the Gospel. Second, though there certainly are differences between Tertullian’s interaction with Marcion’s Gospel and Marcion’s collection of Pauline letters, in Marc. 5.1.9 Tertullian explicitly states that the precedence of Marcion’s treatment of the Gospel leads one to expect the mutilation of the number of Paul’s letters in Marcion’s Apostolikon. It is difficult, therefore, to follow Vinzent in his radical disjunction between Tertullian supposedly presenting Marcion as author of the Gospel but redactor of the Pauline letters. Finally, Vinzent states that ‘Tertullian asserts that with his Gospel Marcion introduced a nova forma sermonis, a literary innovation, that there is in Christ a novel style of discourse, when he sets forth similitudes, when he answers questions’.21 In support of this point, Vinzent refers to Marc. 4.11.12, citing the words forma sermonis in Christo nova, cum similitudines obicit, cum quaestiones refutat.22 This incomplete citation is, in reality, quite misleading for Tertullian actually asserts: Nec forma sermonis in Christo nova. Cum similitudines obicit, cum quaestiones refutat, de septuagesimo23 venit psalmo: Aperiam, inquit, in parabolam os meum, id est similitudinem; eloquar problemata, id est edisseram quaestiones. Even if Vinzent wishes to argue that since Tertullian refutes that there is any novel style of discourse in Christ, Marcion must have argued that there was such a novel style of discourse, it simply is not correct to say that Tertullian asserts that Marcion introduced a nova forma sermonis. Seeking to unravel Marcion’s views from Tertullian’s arguments is a hazardous enterprise and mirror reading a statement by Tertullian most certainly is not the same thing as a statement by Tertullian describing Marcion’s views.24 19. Ibid., 91. 20.  For further discussion of the ensuing critical comments, cf. Dieter T. Roth, ‘Review of Markus Vinzent, Marcion and the Dating of the Synoptic Gospels’, Journal of Theological Studies (JTS) 66 (2015): 800–3 and Roth, ‘The Link between Luke and Marcion’s Gospel’, 68–71. 21. Vinzent, Marcion and the Dating of the Synoptic Gospels, 92. 22. Ibid., 92 n. 352. 23.  Since the citation comes from Psalm 77, many editors corrected the text to septuagesimo septimo, though the Moreschini text in the Source chrétiennes edition (cf. n. 11 above) retains the above reading. 24. Though her comment is made in a different context, Lieu’s observation that ‘distinguishing Tertullian’s rhetoric from actual Marcionite exegesis is often hazardous’ is also à propos here (Marcion and the Making of a Heretic, 258).

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

273

Though there is much more that could be critically discussed in Vinzent’s monograph, the abovementioned issues already reveal the unlikelihood that his approach is a fruitful one for advancing the discussion of the Synoptic Problem. The approach to the Synoptic Problem from the perspective that Marcion, towards the middle of the second century CE, authored the very first Gospel, which was subsequently utilized by the authors of the Synoptic Gospels (and John) is, at least in the presentation of Vinzent, unpersuasive. Klinghardt, by contrast, recently published a massive, two-volume work encompassing both his theory concerning Marcion’s Gospel/the oldest Gospel (Mcn) and his reconstruction of Mcn.25 His discussion of the relationship between Mcn and the canonical gospel encompasses pp. 117–347 in volume 1 of Das älteste Evangelium und die Entstehung der kanonischen Evangelien and cannot even begin to be discussed in detail here. In essence, Klinghardt argues that Mcn was a source, not only for Luke, but also for (pre-canonical) Mark, (pre-canonical) Matthew and, to a lesser but no less real extent, (pre-canonical) John.26 In addition, Luke also used (pre-canonical) Mark, (pre-canonical) Matthew and (precanonical) John as sources; (pre-canonical) Matthew also used (pre-canonical) Mark; and (pre-canonical) John was also familiar with (pre-canonical) Mark and (pre-canonical) Matthew.27 Finally, one single redactor created the canonical forms of the four Gospels through a canonical redaction.28 The absolutely central role played by Marcion’s Gospel in this complex model cannot be overstated. In Klinghardt’s own words: Die hier skizzierte Überlieferungsgeschichte ergibt daher das erstaunliche Bild eines schrittweisen Wachstums der Evangelienüberlieferung, das aus einer 25. On the siglum ‘Mcn’, cf. n. 10 above. 26. Cf. the diagrams in Klinghardt, Das älteste Evangelium, 1:191, 193. It is important to note that this position is different from the one advocated by Klinghardt in an earlier article, in which he presented a diagram with Mcn being dependent upon Mark (cf. Matthias Klinghardt, ‘The Marcionite Gospel and the Synoptic Problem’, Novum Testamentum (NovT) 50 [2008]: 1–27, with the diagram found on p. 21). Lieu is rightly critical of this earlier view, noting that ‘while this hypothesis, unsurprisingly, explains why Luke does not have certain distinctive Matthean additions to their common material, it does demand the additional direct influence of Mark on Luke, for example in Luke 20.9-19 (= Mark 12.1-12), which is omitted by Marcion . . . he does not address Luke’s acquisition of other omissions by Marcion from Mark (e.g. Luke 19.29–46), nor does he seriously consider the number of Matthew’s omissions from Marcion . . . More particularly, he does not undertake a close comparison of the textual relationships, which, as already noted, considerably exacerbate the problem’ (Marcion and the Making of a Heretic, 202–3 n. 49). Though some of these criticisms are no longer relevant for Klinghardt’s revised view, as will be seen in the ensuing discussion, the final point involving the comparison of textual relationships, and especially the basis upon which textual relationships are made, remains a problem. 27. Cf. the diagrams in Klinghardt, Das älteste Evangelium, 1:193, 305. 28. Cf. the diagram in ibid., 1:313.

274

Dieter T. Roth

gemeinsamen Wurzel hervorgeht. Diese gemeinsame Wurzel bleibt von der ersten Rezeption durch Mk bis zum letzten Schritt der lk Redaktion die eine, maßgebliche Bezugsgröße, an der sich alle späteren Rezeptionsschritte mehr (Mk, Mt, Lk) oder weniger (John) orientiert haben . . . Mcn ist nicht nur die gemeinsame Wurzel, aus der alle vier kanonischen Evangelien auf die eine oder andere Art entsprungen sind, sondern auch der Baum, um den herum sie sich ranken.29

It is not surprising that many controversial arguments are found in the several hundred pages to which Klinghardt devotes his attempt to establish this complex theory of Gospel relationships and development, the vast majority of which simply cannot be considered within the confines of this chapter. Since the focus in the present discussion is on Marcion’s Gospel, two examples highlighting the problem of the reconstruction of this text and its use in Klinghardt’s theory will be considered as they are illustrative of the types of difficulties that attend Klinghardt’s discussion throughout. The first example involves the question of the word-level reconstruction of Mcn in 12.51. Luke 12.51 concludes with οὐχί, λέγω ὑμῖν, ἀλλ᾽ ἢ διαμερισμόν, whereas the Matthean parallel in Matt 10.34 reads οὐκ ἦλθον βαλεῖν εἰρήνην ἀλλὰ μάχαιραν. Klinghardt points out this difference and then also observes, ‘In einer einzigen Minuskel (1242*) findet sich diese Wortwahl auch für Lk 12,51 (οὐχί, λέγω ὑμῖν, ἀλλ᾽ ἢ μάχαιραν).’30 At this point, however, Klinghardt’s view becomes massively problematic. He concludes, ‘Allerdings bezeugt Adamantius genau diese Variante für Mcn, die in 1024* [sic] folglich nicht sekundär ist, sondern eine ausnahmsweise nicht getilgte Interferenz zwischen dem vorkanonischen und dem kanonischen Text anzeigt.’31 In other words, Klinghardt views Mcn as having employed the word μάχαιρα, Matthew having followed this reading, and then the Lukan redaction of Mcn (and Matthew) having changed the reading using διαμερισμός.32 Yet, his suggested reading for Mcn here is, in my estimation, indefensible. Though Klinghardt indicates in a footnote that Tertullian attests the Lukan reading in Marc. 4.29.13 (non, dico vobis, sed separationem), Klinghardt is apparently convinced that the Adamantius Dialogue has retained the original

29.  Ibid., 1:310. Klinghardt recognizes numerous fundamental similarities between his view and the view of Vinzent, even as there are also some important differences. For this reason, Klinghardt briefly interacts with Vinzent’s theory, though indicates that a full discussion and evaluation ‘ist eine Aufgabe für die Zukunft’ (Das älteste Evangelium, 1:388). 30. Klinghardt, Das älteste Evangelium, 1:101. The attestation for the reading is presumably taken from the apparatus in The Gospel According to St. Luke, ed. American and British Committees of the International Greek New Testament Project, 2 vols, The New Testament in Greek 3 (Oxford: Clarendon, 1984–87). 31. Klinghardt, Das älteste Evangelium, 1:101. 32. Cf. ibid., 2:804.

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

275

reading in Adam. 68,1 (2.5).33 He does not mention here that Tertullian, more than simply offering the reading, actually accuses Marcion of having altered the reading from μάχαιραν to διαμερισμόν.34 Thus, it is nigh on to certain that Tertullian’s copy of Marcion’s text contained the Lukan word.35 Klinghardt also fails to mention that the Adamantius Dialogue, whether originally or due to the process of transmission, regularly provides Matthean readings in passages ‘cited’ in the debates with the Marcionites. This evident element of the citation habit of the author (or copier) of the Adamantius Dialogue is vitally important to bear in mind when considering the readings attested therein.36 As Kenji Tsutsui observes, ‘Daß die Markioniten häufiger aus dem Matthäusevangelium zitieren als aus dem Lukasevangelium,

33. The reference to the Adamantius Dialogue provides the page and line number from the W. H. van de Sande Bakhuyzen edition of the Greek text in Pseudo-Origen, Der Dialog des Adamantius: ΠΕΡΙ ΤΗΣ ΕΙΣ ΘΕΟΝ ΟΡΘΗΣ ΠΙΣΤΕΩΣ, GCS 4 (Leipzig: J. C. Hinrichs, 1901) followed by the divisions, set in parentheses, in C. P. Caspari’s edition of Rufinus’s Latin translation of the text in Kirchenhistorische Anecdote: Nebst neuen Ausgaben patristischer und kirchlich-mittelaltlicher Schriften:  I. Lateinische Schriften:  Die Texte und die Anmerkungen (Oslo: Malling, 1883), 1–129. 34. Klinghardt does note this in his reconstruction (Das älteste Evangelium, 803). 35.  For further discussion of this passage, cf. Roth, Text of Marcion’s Gospel, 87, 90 and Dieter T. Roth, ‘Matthean Texts and Tertullian’s Accusations in Adversus Marcionem’, JTS 59 (2008):  596–97. In the latter work I  point out that Tertullian nowhere else cites Luke 12.51 but that he twice cites Matt 10.34, with one of those citations coming earlier in Adversus Marcionem (cf. Scorp. 10.17 and Marc. 3.14.5). In both of the above referenced publications I  argue that Tertullian has confused the Matthean and Lukan version and that his accusation therefore stems from a memory slip. In a note just prior to discussing Tertullian’s comments on Luke 12.51, Lieu states, ‘Roth . . . argues that Tertullian is not claiming (mistakenly) that Marcion has eliminated such [Matthean] verses from Luke, but that he has omitted what belongs to the “true” Gospel, but this does not explain the following example’ (Marcion and the Making of a Heretic, 191 n.  18). This comment is rather curious, since concerning Luke 12.51, I explicitly argue that ‘a different explanation is most likely’ (Roth, ‘Matthean Texts’, 596) and reiterate in the conclusion of the article: ‘I would submit that with only one clear exception, where Tertullian’s memory failed him [here referring to Luke 12.51], the accusations, or at least most of them, concerning the deletion of Matthean verses or elements levelled by Tertullian against Marcion are best explained by Tertullian’s theological perspective that the four gospels comprise a single, unitary “Gospel”, and therefore were motivated by Marcion having rejected the Gospel of Matthew’ (ibid., 597). 36.  Cf. also the observation by Judith Lieu, ‘Marcion and the Synoptic Problem’, in New Studies in the Synoptic Problem: Oxford Conference, April 2008: Essays in Honour of Christopher M. Tuckett, ed. Paul Foster, Andrew Gregory, John S. Kloppenborg and Joseph Verheyden, BETL 239 (Leuven: Peeters, 2011), 737: ‘There is a marked tendency in patristic citation for Matthew to influence quotations of Luke (or Mark), and this is more generally evident in both Epiphanius and Adamantius.’

276

Dieter T. Roth

spricht ebenfalls gegen die Zuverlässigkeit des Dialogs’.37 Adolf von Harnack rightly contended that the reading μάχαιραν ‘zu Unrecht vom Dialog ihm [Marcion’s text] zugeschrieben wird’.38 It seems to me that dismissing Tertullian’s attestation and following the reading in the Adamantius Dialogue is rather problematic. Seeking to support this decision by appealing to the original hand of a corrected thirteenthcentury minuscule (!) is text-critically highly questionable to say the least.39 Would Klinghardt really have us believe that in the entirety of the Lukan textual tradition, this minuscule alone has preserved the reading of the pre-canonical source for Luke?40 Thus, it is not the case that Mcn and Matthew shared a reading here and Klinghardt’s model would not simply involve the usually postulated redactional move from ‘sword’ to ‘division’,41 but actually need to explain why Matthew made the surprising move of replacing διαμερισμόν in Mcn with μάχαιραν.

37.  Kenji Tsutsui, Die Auseinandersetzung mit den Markioniten im AdamantiosDialog:  Ein Kommentar zu den Büchern I–II, Patristische Texte und Studien (PTS) 55 (Berlin: de Gruyter, 2004), 92. 38. Adolf von Harnack, Marcion: Das Evangelium vom fremden Gott: Eine Monographie zur Geschichte der Grundlegung der katholischen Kirche, 2nd edn, Texte und Untersuchungen (TU) 45 (Leipzig: J. C. Hinrichs, 1924), 216*. 39.  In addition to the typographical error in the reference to the minuscule in the citation referenced by n. 31 above, Klinghardt commits a serious error in a note found in his reconstruction of 12.51. Klinghardt writes, ‘Eine einzige Minuskel (1241*; s. XII, Sinai) liest hier μαχαιραν anstelle von διαμερισμον. Allerdings ist der Evangelientext der Minuskel 1241 einigermaßen wichtig, weil er eine Fülle von bemerkenswerten Lesarten enthält (vgl. H. von Soden, Die Schriften des Neuen Testaments I/2, Berlin 1907, 989; H. J. Vogels, Beiträge zur Geschichte des Diatessaron im Abendland, Münster 1919, 13; im Lk-Text sind die Lesarten zu 12,58; 13,29; 18,16; 21,11; 23,34; 24,3 wichtig; s. dazu die Liste in Anhang III)’ (Das älteste Evangelium, 803 n. 10). The glaring problem is that it is not 1241*, but rather 1242* that attests the reading. Even if one were to agree with Klinghardt in his evaluation of the significance of 1241 (the references to von Soden and Vogels are also to 1241  =  von Soden δ371 whereas 1242  =  von Soden δ469), the evaluation is completely irrelevant here because Klinghardt is discussing the wrong manuscript. 40. It is not warranted to assume that Tertullian’s text of Luke contained the reading found in Matthew as Klinghardt apparently does: ‘Ironischerweise zeigt Tertullians katholischer Lk-Text hier die Lesart des vorkanonischen Evangeliums’ (Das älteste Evangelium, 804). Cf. n. 35 above. 41.  μάχαιραν is generally viewed as the reading of Q (cf. James M. Robinson, Paul Hoffmann and John S. Kloppenborg, The Critical Edition of Q: Synopsis Including the Gospels of Matthew and Luke, Mark and Thomas with English, German, and French translations of Q and Thomas, Hermeneia [Minneapolis: Fortress, 2000], 380). For the view that Luke changed the Q reading whereas Matthew preserved it, cf., e.g. W. D. Davies and Dale C. Allison, Jr., The Gospel According to Saint Matthew, 3 vols, International Critical Commentary (ICC) (Edinburgh: T&T Clark, 1988–97), 2:218 and François Bovon, Das Evangelium nach

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

277

The second example involves both broader questions of the reconstruction of Marcion’s Gospel and the use of the manuscript tradition in positing readings which form the basis of Klinghardt’s theory of redactional development. When considering the relationship between 14.34–35 in Mcn and Luke 14.34–35, Matt 5.13 and Mark 9.50, Klinghardt admits that 14.25–35 is unattested for Marcion’s Gospel.42 This, however, does not hinder Klinghardt from providing a word for word reconstruction of the text. Such a procedure is, in my estimation, highly dubious, as is the assertion that in Mcn ‘die metaphorik des Salzes . . . sich hier auf die Funktion der Jünger [bezieht], die alles verlassen haben bzw. ihr Kreuz auf sich nehmen (*14,25)’.43 There is no source for Marcion’s Gospel for these verses and a critical reconstruction of Marcion’s text should, at this point, read ‘unattested’.44 What is more, Klinghardt contends that the pre-canonical reading for Mark 9.49 is actually found in the reading of D and several OL manuscripts (πᾶσα γὰρ θυσία ἁλὶ ἁλισθήσεται). Thus, he believes that ‘der primäre Vergleichstext für das mcn Salzlogion *14,34f in *Mk 9,49f (D, it) zu sehen ist’ and concludes that even though there is some semantic uncertainty, ‘die Rezeption des Salzlogions in Mt 5,13 auf die überlieferungsgeschichtliche Priorität von *14,34f vor *Mk 9,49f (D it) hin[deutet]’.45 Thus, if I have understood Klinghardt correctly, the evidence for Mcn’s priority, it having been utilized as a source by Mark, with Matthew having used both Mcn and Mark, is found in an unattested reading for Marcion’s Gospel and the supposed pre-canonical reading of Mark preserved in D.  I, for one, have serious doubts about such an approach and significant difficulties in

Lukas, 4 vols, Evangelisch-katholischer Kommentar zum Neuen Testament (EKKNT) 3 (Neukirchen-Vluyn: Neukirchner, 1989–2009), 2:354. Cf. also the extensive documentation in Albrecht Garsky, Christoph Heil, Thomas Hieke and Josef E. Amon, Q 12:49–59: Children against Parents, Judging the Time, Settling out of Court, ed. Shawn Carruth, Documenta Q: Reconstructions of Q through Two Centuries of Gospel Research Excerpted, Sorted and Evaluated (Leuven:  Peeters, 1997), 98–106, where the vast majority of scholars views the direction of the redaction as being from the Matthean reading to the Lukan reading. 42.  He writes, ‘*14,25–35; unbezeugt, aber vermutlich vorhanden’ (Klinghardt, Das älteste Evangelium, 216). 43. Ibid. 44.  Cf. Roth, Text of Marcion’s Gospel, 425. Stating, ‘Für Mcn nicht bezeugt, aber vielleicht ganz vorhanden und geringfügig durch die lk Redaktion bearbeitet’ (Klinghardt, Das älteste Evangelium, 2:843) immediately raises the question of how this can be known. Even if the passage were to have been present in Marcion’s Gospel, there is no critically controlled manner in which one can move from the various attestations of readings in these verses in the manuscript tradition to the precise wording of Marcion’s Gospel. For an extensive discussion concerning the methodology of reconstructing Marcion’s texts; questions of methodology in the history of research on Marcion’s Gospel; and lists of passages attested as present, passages attested as not present, and unattested passages for Marcion’s Gospel in the sources, cf. Roth, Text of Marcion’s Gospel, 1–82. 45. Klinghardt, Das älteste Evangelium, 1:217.

278

Dieter T. Roth

agreeing with viewpoints arising from it. I would therefore conclude that though there certainly is much in Klinghardt’s work that is worth considering and that must be discussed at greater length, in the light of issues such as those illustrated in the examples above, many will find the challenges to accepting Klinghardt’s methodology, reconstruction and conclusions concerning Marcion’s Gospel and the Synoptic Problem difficult to overcome.

14.4 Marcion’s Gospel as Relevant, but not Determinative, for the Synoptic Problem Rather different are the proposals by BeDuhn and Lieu as to the place of Marcion’s Gospel in scholarly discourse concerning Marcion’s Gospel. BeDuhn argues that behind Marcion’s creation of the first New Testament ‘there is the high likelihood that there stands an earlier significant event: the composition of the Evangelion, by an unknown author, at an uncertain time, in an undetermined location.’46 As such, BeDuhn is of the opinion that ‘Marcion may not have produced a definitive edition of the Evangelion after all, but rather took up a gospel already in circulation in multiple copies’, with the result that ‘Marcion’s Evangelion and Luke are both independent developments of a common proto-gospel’.47 On the basis of this view, BeDuhn highlights several implications that are directly relevant for the discussion of the Synoptic Problem. On the one hand, if the evidence of Marcion’s Evangelion is taken into consideration and it, instead of Luke, is utilized as the point of comparison with Matthew, the scope of Q would be reduced, ‘but far from undermining the Q hypothesis, this reduction in the scope of Q actually strengthens it’.48 BeDuhn argues that if Q opened with the Sermon on the Mount/Plain material, all of the problems for the Two-Document theory found in the introduction of John and the temptation passage ‘evaporate at a stroke’.49 For this reason, ‘The evidence of the Evangelion . . . points to a Q that more consistently takes the familiar form of a sayings collection than does the text of Q reconstructed on the basis of Luke.’50 On the other hand, when considering the so-called minor agreements, which BeDuhn identifies as ‘the single greatest point of criticism of the two-source hypothesis’, BeDuhn contends that ‘The Evangelion has substantially fewer (one third) of the “minor agreements” accepted in the current critical text of Luke [emphasis original].’51 Though this line of argumentation is certainly worth pursuing 46.  BeDuhn, First New Testament, 92. Of course, Beduhn’s view that Marcion actually created the first ‘New Testament’, and especially the idea that Marcion actually used this term, is debated (cf. Lieu, Marcion and the Making of a Heretic, 406–8). 47. BeDuhn, First New Testament, 90, 79. 48. Ibid., 94. 49. Ibid., 95. 50. Ibid., 96. 51. Ibid., 93.

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

279

when considering Marcion’s Gospel and the Synoptic Problem,52 it is an unfortunate reality that BeDuhn does not provide the evidence for or the documentation underlying this assertion. Furthermore, since BeDuhn has provided an English translation of Marcion’s Gospel, the requisite basis for such a contention, namely the Greek text, is also not found in BeDuhn’s volume. At the same time, in several instances BeDuhn does offer comments on specific ‘minor agreements’ in the notes following his English text. When considering these comments, the practical challenges of following this theoretically insightful path become readily apparent. In several instances where BeDuhn makes reference to minor agreements, the issue concerning the reconstruction of Marcion’s text once again becomes relevant. For instance, in the note on 24.4, BeDuhn comments ‘The “minor agreement” between Matthew and Luke in the characterization of their shining cloaks (astrapē/astraptousē) was not present in the Evangelion (lampra is used instead).’53 Though the characterization is not of the cloak in Matthew, there may indeed be a ‘minor agreement’ here in Matthew and Luke.54 It is far from certain, however, that Marcion’s Gospel did not have this reading. Tertullian does not refer to this element in v. 4 and concerning Epiphanius’s testimony, as I noted in my reconstruction, his ‘only reference to v. 4 is the allusion to the men in shining garments at the conclusion of the verse. The rather clearly summary and imprecise nature of Epiphanius’s introduction to scholion 76 makes any reconstruction of precise wording tenuous and speculative’.55 Even Harnack, though deciding to follow Epiphanius, noted the uncertainty by reconstructing ἐν ἐσθῆτι λαμπρᾷ (ἀστραπτούσῃ?).56 In another instance, namely in 5.12, BeDuhn observes that ‘ “Master” (kyrie) is not directly attested for the Evangelion; it is one of the “minor agreements” between Matthew and Luke against Mark in this verse’.57 In my estimation, since this element is unattested for Marcion’s Gospel it is methodologically impermissible to use the verse to support either the inclusion or the exclusion of a ‘minor agreement’.58 In yet a further instance, BeDuhn states concerning 5.18: ‘Here and in the following verses the Evangelion apparently read “cot” (Latin grabatum > Gk krabatton) with Gk ms D, several OL manuscripts, and Mark, instead of “bed” and “little bed,” respectively, in most witnesses to Luke, which appears to be influenced by Matthew and constitutes one of the “minor agreements” absent from the Evangelion.’59 Yet, even

52.  Klinghardt, Das älteste Evangelium, 1:234–45 also discusses the so-called minor agreements. 53. BeDuhn, First New Testament, 194. 54.  Cf. Frans Neirynck, The Minor Agreements of Matthew and Luke against Mark with a Cumulative List, BETL 37 (Leuven: Leuven University Press, 1974), 194 and the comments in Andreas Ennulat, Die ‘Minor Agreements’, WUNT II/62 (Tübingen: Mohr Siebeck, 1994), 411. 55. Roth, Text of Marcion’s Gospel, 342. 56. Harnack, Marcion, 238*. 57. BeDuhn, First New Testament, 132. 58. Cf. n. 44 above. 59. BeDuhn, First New Testament, 133.

280

Dieter T. Roth

BeDuhn only reconstructs the term for 5.2460 and in my estimation the evidence here is so uncertain that I evaluated the reading as only ‘possible’, meaning that it is a reading ‘attested by a source, though ultimately no confidence can be placed in these readings being found in Marcion’s text’.61 In a final instance in 7.27, BeDuhn comments, ‘The final phrase “before you” is absent in D and the OL, as it is in Mark 1.2, and both Epiphanius and Tertullian end their quotes before reaching it (Epiphanius lacks the entire clause “who will prepare your road before you”); but Adam* 2.18 attests it, making this one of the few “minor agreements” between Luke and Matthew against Mark that is also found in the Evangelion.’62 Here, however, basing the reconstruction solely on the Adamantius Dialogue is problematic and ultimately I would argue that Marcion’s reading at this point is uncertain.63 In two other instances, however, BeDuhn may be on firmer footing. In 9.20, BeDuhn contends that the Evangelion ‘had “You are the Christos” without the additional qualification “of God” (Tertullian, Marc. 4.21.6; 4.22.6; 4.34.16; Adamantius), as do also the SSyr and CSyr and OL ms a, in line with the text of Mark 8.29, and so lacking another of the “minor agreements” with Matthew found in most witnesses to Luke.’64 Though BeDuhn is perhaps more certain than warranted, ‘of God’ may indeed have been missing in Marcion’s text.65 In 9.41, BeDuhn observes, ‘Our sources seem unanimous that the additional phrase “mistrustful and twisted” found in v. 41 of Luke and Mt 17.17, was absent from the Evangelion, as it is from OL mss a and e, and from the parallel in Mk 9.19; thus another one of the “minor agreements” between Matthew and Luke against Mark evaporates [emphasis original].’66 In this instance, I would concur.67 In sum, though the types of questions that BeDuhn is asking of Marcion’s Gospel are of significance and importance for the Synoptic Problem, it is also evident that the scholarly discussion and debate concerning both the reconstruction of the text being queried and the precise relationship between Marcion’s Gospel and Luke remains of vital significance. A final view to be discussed here is that of Lieu. In her monograph she concluded that ‘at the end of these investigations [into Marcion’s Gospel] it remains certain that the Gospel that Marcion used as his core text followed the same structure and sequence of textual units as canonical Luke, but that it may have lacked some of the passages and verses now part of the latter’.68 As such, though advocating the view that Marcion’s Gospel represents ‘some form of “pre-canonical Luke” ’, 60. Cf. ibid., 100. 61.  Roth, Text of Marcion’s Gospel, 411. For a discussion of the four levels of certainty employed in my reconstruction of Marcion’s Gospel, cf. ibid., 410–12. 62. BeDuhn, First New Testament, 143. 63. Cf. Roth, Text of Marcion’s Gospel, 114–15, 365–6, and 416. 64. BeDuhn, First New Testament, 149. 65. Cf. Roth, Text of Marcion’s Gospel, 418. 66. Ibid., 151. 67. Cf. Roth, Text of Marcion’s Gospel, 214, 304 and 419. 68. Lieu, Marcion and the Making of a Heretic, 208.

14. Marcion’s Gospel and the Synoptic Problem in Recent Scholarship

281

Lieu contends that attempts to go beyond such a basic statement ‘and to integrate Marcion’s “Gospel” directly into any of the conventional theories of the interrelationship and interdependency of the three Synoptic Gospels have proved unpersuasive  – even at a time when such conventional theories are again becoming contentious’.69 Thus, instead of such a direct integration into, and thus attempting to make Marcion’s Gospel the key to understanding, the Synoptic relationships, Lieu more generally observes that ‘any theory of synoptic relationships cannot avoid considering the implications of the different forms of the Lukan text or the effect of both oral and written traditions on it over a considerable period of time’.70 For this reason, Lieu seeks to understand and locate Marcion’s Gospel, not at the centre of any proposed solution to the Synoptic Problem, but rather as one of numerous second-century Gospel traditions that cannot be ignored. In sum, Lieu helpfully suggests ‘that Marcion’s Gospel should continue to contribute to debates about the Synoptic Problem, just as those debates should contribute to the attempt to understand his [Marcion’s] own procedures and their place in the growth of authoritative texts’.71

14.5 Conclusion In the light of the foregoing discussion, it is evident that recent scholarship on Marcion and Marcion’s Gospel has brought this Gospel to bear on the discussion of various issues in NT scholarship, not least of which is the discussion of the Synoptic Problem. Though the renewed interest in Marcion’s Gospel and recent resurgence of scholarship on it is, in my estimation, a welcome development, it is patently obvious that great care needs to be exercised in the use of this text and in appeals made to it. On the one hand, Lieu is absolutely correct in stating that ‘given the many uncertainties in our knowledge of Marcion’s Gospel, it must be recognised that any hypothesis can claim to be nothing more than a hypothesis’.72 On the other hand, she is equally right in insisting that ‘total scepticism about the textual details of Marcion’s text is not warranted’, and for this reason it is vitally important that ‘any hypothesis [concerning Marcion’s Gospel] must work on the textual level as well as on that of content’.73 It is precisely at this point, however, that there are significant problems in recent work on Marcion’s Gospel and the Synoptic Problem, especially in the works of Vinzent and Klinghardt, and also, to a certain extent, in BeDuhn’s monograph. When grand theories are constructed on the basis of improbable, and at times even indefensible, reconstructions, one only further complicates the discussion of topics such as the Synoptic Problem, a topic

69. Ibid., 202. 70. Lieu, ‘Marcion and the Synoptic Problem’, 739. 71. Ibid., 751. 72. Lieu, ‘Marcion and the Synoptic Problem’, 746. 73. Ibid.

282

Dieter T. Roth

that is complicated enough already! In the opening pages of my monograph on the text of Marcion’s Gospel, I noted that a number of textual critics, NT scholars and patristic scholars had highlighted the need for a new and more rigorous reconstruction of Marcion’s Gospel.74 In the conclusion of my own attempt to address this scholarly desideratum, I wrote that ‘the text reconstructed here is by no means considered to be the “last word” in reconstructions of Marcion’s Gospel; rather, it is an attempt to move scholarly debate and discussion forward on firmer ground in the hopes that it may engender further consideration of the readings of this text’.75 It is precisely such further consideration, debate and discussion concerning the readings of Marcion’s Gospel that is a necessary prerequisite for moving forward in scholarship on Marcion’s Gospel and the Synoptic Problem. Here again, it is my hope that a few of the thoughts, observations and even criticisms offered here may serve as an impetus for future work on Marcion’s Gospel and its implications for issues in the field of NT scholarship, not least of which is the ongoing debate concerning the Synoptic Problem and ‘solutions’ posited for it.

74.  Cf., e.g. the sentiments expressed by Kurt and Barbara Aland, The Text of the New Testament, trans. Erroll F. Rhodes, 2nd edn (Grand Rapids, MI:  Eerdmans, 1989), 172; Karlmann Beyschlag, ‘Marcion von Sinope’, in Alte Kirche I, vol. 1 of Gestalten der Kirchengeschichte, ed. Martin Greschat (Stuttgart: Kohlhammer, 1984), 71; J. K. Elliott, ‘The New Testament Text in the Second Century: A Challenge for the Twenty-First Century’, New Testament Textual Research Update (NTTRU) 8 (2000):  12; Martin Hengel, ‘Aufgaben der neutestamentlichen Wissenschaft’, NTS 40 (1994): 342; and Gerhard May, ‘Markion in seiner Zeit’, in Gerhard May: Markion: Gesammelte Aufsätze, ed. Katharina Greschat and Martin Meiser, Veröffentlichungen des  Instituts für Europäische Geschichte (VIEG) 68 (Mainz: Philipp von Zabern, 2005), 8. 75. Roth, Text of Marcion’s Gospel, 437.

AUTHOR INDEX Adamczewski, Bartosz 232 Aejmelaeus, Lars 247 Aland, Barbara 282 Aland, Kurt 19, 282 Albright, W. F. 168 Alexander, Loveday 131 Alkier, Stefan 43–4, 49, 55, 60, 62–3, 66–7, 70–1, 232 Allen, W. C. 112 Allison, Dale C., Jr. 100, 189, 191, 276 Allison, Henry E. 30 Alsup, John 248 Amon, Josef E. 277 Arnal, William E. 35–6, 264–5 Attridge, Harold W. 100 Babut, Jean-Marc 197 Barclay, John M. G. 161 Baring-Gould, S. 181 Barker, James 26 Bartlet, J. V. 112 Bates, Matthew 250 Bauckham, Richard J. 122 Bauer, Bruno 89–90 Baur, Ferdinand Christian 32–3, 74, 87, 103, 268 Bazzana, Giovanni Battista 39, 41, 159 Beacham, Richard C. 172 Becker, Eve-Marie 154, 157–8 Becker, Matthias 52 BeDuhn, Jason D. 182, 254–5, 257–8, 260–2, 267–9, 278–81 Bellinzoni, Arthur J. 263 Bernhard, Andrew 100 Betz, Otto 34 Beutel, Albrecht 56 Beyschlag, Karlmann 282 Bilby, Mark G. 247 Bilde, Per 161 Blumell, Lincoln H. 159 Bohnert, Daniel 59 Boismard, Marie-Emile 29

Bonner, Campbell 174 Bousset, Wilhelm 155 Bovon, François 167–8, 276 Brown, Raymond E. 17 Brown, Rupert 186 Brown, Scott G. 17 Bruce, F. F. 256 Bultmann, Rudolf 92–3, 166 Burger, Christoph 54 Burke, Tony 17 Burkett, Delbert 201 Burkitt, F. Crawford 203, 222 Büsching, Anton Friedrich 50, 54 Butler, B. C. 13, 29 Calhoun, R. Matthew 165, 181 Cancik, H. 158 Carlson, Stephen 17 Catchpole, David R. 3, 155 Comfort, Philip 101 Cope, Lamar 19, 24 Cosgrove, Charles 172 Cotter, Wendy J. 166, 168, 171 Credner, Karl August 223, 228 Crook, Zeba A. 19, 157 Crossan, John Dominic 16, 99 Damgaard, Finn 244 Damm, Alexander 27–8, 41, 152 D’Angelo, Mary Rose 246, 259 Daniels, Jon B. 17 Davies, Margaret 211, 218 Davies, W. D. 100, 137, 189, 191, 276 Denaux, Adelbert 19, 195 Derrenbacker, Robert Allen 21, 25–6, 130, 143 de Wette, W.M.L. 33 Dewey, Arthur J. 41 Dibelius, Martin 91 Downing, F. Gerald 26–7, 130, 135, 137, 143, 214 Draper, Jonathan A. 16

284

Author Index

Dungan, David Laird 51–2, 61, 67 Dupont, Jacques 197 Ebner, Martin 4, 187, 191, 196–7, 231 Ehrman, Bart D. 251 Eichhorn, Johann Gottfried 1, 59, 62, 64–5, 69, 70–1, 74, 77, 82–3 Elliott, J. K. 18, 44, 282 Ennulat, Andreas 279 Eshleman, Kendra 162 Eve, Eric 135, 204, 214 Farmer, William R. 1–2, 5, 13, 18, 33–5, 201 Farrer, A. M. 2, 13, 97–8, 100, 115, 151, 202, 230, 233 Fiorenza, Elisabeth Schüssler 260, 264 Fitzmyer, Joseph A. 165–8, 170, 174, 178, 207–8, 256 Fleddermann, Harry T. 40, 155, 192, 196, 203, 225–7 Flusser, David 29 Foster, Paul 271 France, R. T. 188, 191 Franklin, Eric 115, 121 Franks, Paul W. 32 Fredriksen, Paula 210 Frenschkowski, Marco 200 Frickenschmidt, Dirk 46 Fuller, R. H. 268 Gamble, Harry Y. 255 Garrow, Alan J. P. 16, 218 Garsky, Albrecht 277 Gathercole, Simon 16, 98, 101 Gauger, H. M. 158 Georgi, Dieter 32 Gerber, Christine 192, 195 Gibbon, Edward 67 Gielen, Marlis 187 Goethe, Johann Wolfgang 70 Good, Dierdre 259 Goodacre, Mark S. 1, 3, 6, 16, 21, 23, 35, 37, 98, 100, 115, 130–1, 140, 155, 165, 185, 202, 205, 208, 210, 212, 218, 221, 230, 236–7 Goodenough, Erwin R. 3 Goulder, Michael D. 2, 20–2, 25, 27, 34–5, 37, 94, 97, 115, 202, 204, 213, 220, 230, 233

Green, Joel B. 16, 175, 189 Greeven, Heinrich 69 Gregory, Andrew 151, 247–8, 254–5 Greimas, Algirdas J. 226, 234–5 Grenfell, Bernard 100–5, 108–9 Griesbach, Johann Jakob 1, 50, 54, 56, 61–2, 67, 75, 81, 116, 151 Grundmann, Walter 166 Hamel, Gildas 188 Harnack, Adolf von 33–4, 99, 108–9, 182, 224–5, 276, 279 Harrington, Daniel 168 Hawkins, John C. 108, 112 Heath, Jane 101 Heil, Christoph 152, 154, 156, 158, 162, 195–6, 225, 243, 277 Held, Heinz Joachim 25 Hengel, Martin 282 Herder, Johann Gottfried 70, 72–4, 83–6 Herzer, Jens 166 Hieke, Thomas 277 Hilgenfeld, Adolf 268 Hills, Julian V. 17 Hobson, A. Augustus 50 Hoffmann, Paul 100, 203, 243, 276 Hogarth, David 102 Hollander, August Den 54 Holtzmann, Heinrich Julius 2, 31–3, 33, 75–6, 88, 107–8, 111, 139, 228–30 Honore, A. M. 213 Horbury, William 154 Horman, John 16 Hübner, Kurt 232 Huggins, Ronald V. 218 Huizinga, Johan 178 Hume, David 67 Hüneburg, Martin 225 Hunt, Arthur S. 100–3, 105, 108–9 Hurtado, Larry 101 Hyldahl, Niels 96 Iser, Wolfgang 226 Jacobson, Arland D. 188, 193 Jeffery, Peter 17 Jefford, Clayton N. 251 Jenkins, Philip 99 Jervell, Jacob 249

Author Index Johnson, William A. 159 Johnson-DeBaufre, Melanie 192, 265 Jonas, Hans 253 Jones, Brice 209 Joseph, Simon J. 154 Kahl, Werner 4–5, 224, 226, 231, 236–7 Kant, Immanuel 56, 70 Karlstadt, Andreas B. 55 Kawan, Christine Shojaei 14 Kelhoffer, James A. 167, 178 Kenyon, Frederick 103 Kirk, Alan 16 Kittel, Gerhard 224 Klinghardt, Matthias 5, 246, 254–8, 260, 264, 267–70, 273–7 Kloppenborg, John S. 1, 5–7, 13, 16, 20–1, 23–5, 32, 34, 37–41, 99–100, 116, 118–19, 122, 130, 152–3, 165, 179, 199, 202–4, 208–10, 223, 225, 232–3, 262, 265, 276 Knox, John 255 Koester, Helmut 15, 17, 50, 99 Koppe, J. B. 69, 72–3 Köster, Beate 54 Kuhn, Thomas S. 21 Kümmel, Werner Georg 81, 83–5, 87, 90, 207 Labahn, Michael 188, 225–7 Lachmann, Karl 2, 33, 86 Lagrange, M.-J. 33 Lambrecht, Jan 155 Larfeld, Wilhelm 244 Laufen, Rudolf 203, 209 Lawler, Lilian B. 171 Layton, Bentley 16 Lehtipuu, Outi 251, 258 Leivestad, Ragnar 178 Lemke, Helle 65 Leppä, Heikki 247 Leppin, Volker 56 Lessing, Gotthold Ephraim 1, 15, 30–1, 62, 66, 70, 72, 81–2, 151, 223 Lexutt, Athina 55–6 Lieu, Judith M. 255–6, 258, 264, 267–9, 272–3, 275, 278, 280–1 Lindemann, Andreas 4, 6, 227–8 Lindsey, Robert L. 29

Linnemann, Eta 5 Linton, Olof 167 Lock, Walter 108–9 Locke, John 60 Lohse, Bernhard 55 Longstaff, Thomas R. W. 25, 77 Lonsdale, Steven H. 171 Lowe, Malcolm 29 Lührmann, Dieter 15, 36–7 Luther, Martin 54–8 Luz, Ulrich 94, 166–7, 169 MacEwen, Robert K. 218, 234 Machinist, Peter 181 Maier, Christl 166 Mann, C. S. 37, 168 Marshall, I. Howard 167–9, 171 Martin, Michael Wade 248 Mason, Steve 247 Masterman, Margaret 21 Matthews, Shelly 245, 247–8, 264 Matz, Wolfgang 55–6 May, Gerhard 282 McCane, Byron M. 189 McLoughlin, S. 261 McNicol, Allan J. 19, 24 Meier, John P. 42 Merkel, Helmut 129 Merton, Robert K. 30 Metzger, Bruce M. 48 Meyer, Ben F. 34 Meyer, G. 229 Meyer, Marvin 17 Moll, Sebastian 267 Morgan, Michael L. 32 Mount, Christopher 246 Moxnes, Halvor 190 Müller, Mogens 94–6, 245, 249, 264 Nasrallah, Laura 247 Neirynck, Frans 2, 17–20, 40, 213, 279 Nicklas, Tobias 17 Notley, R. Steven 33 Oesterley, W. O. E. 171 Olson, Ken 35, 135–6, 214 Overbeck, F. 158 Owen, Henry 1, 54, 61–2, 67–9

285

286

Author Index

Pagels, Elaine 17, 252–3 Palmer, N. H. 86 Parker, David 245, 263 Patterson, Stephen J. 16, 37 Peabody, David 5, 19, 24–5, 153 Perrin, Nicholas 16, 115, 202 Perrin, Norman 231 Pervo, Richard I. 246–7 Peter, Simon 140 Petersen, William L. 263 Peterson, Jeff 115, 202 Petrie, Flinders 104 Phillips, Thomas E. 247 Plummer, Alfred 167 Poirier, John C. 6, 115, 121, 129, 151, 201–2 Popper, Karl 20 Poppi, Angelico 19 Propp, Vladimir 235 Rathbone, D. W. 102 Regev, Eyal 160 Reicke, Bo 77 Reid, Barbara 260 Reid, Duncan 28 Reimarus, Hermann Samuel 5, 30–1, 66, 71, 223–4 Rendel Harris, James 106–7, 109 Resch, Alfred 37 Ringe, Sharon H. 260 Ritschl, Albrecht 237, 268 Robbins, Vernon 246 Roberts, C. H. 101, 145 Robinson, James M. 15, 34, 100, 153, 203, 276 Rolland, Philippe 29 Rollens, Sarah E. 191 Rosché, Theodore E. 156 Rosenzweig, Franz 32 Roth, Dieter T. 182, 188, 195, 234, 254–5, 257, 267–9, 272, 277, 279–80, 282 Rothschild, Clare 271 Sanday, William 13, 108–11, 113 Sanders, Ed Parish 20, 35, 137, 157, 203–4, 208–9, 211, 216, 218 Satlow, Michael L. 172 Sawicki, Marianne 260 Schaberg, Jane 259–60

Schellenberg, Ryan 247 Schenk, Wolfgang 152, 155 Schiller, Friedrich 70 Schleiermacher, Friedrich 1, 15, 86, 223–4, 228–9 Schmid, Ulrich 54, 269–70 Schmidt, Karl Ludwig 92, 95 Schmithals, Walter 155, 253 Schnelle, Udo 4, 151 Schonhoffer, T. Nicholas 16 Schrage, Wolfgang 16 Schreiber, Stefan 4 Schröter, Jens 4 Schubert, P. 249 Schürmann, Heinz 169–70 Schüssler, Hermann 57 Schweitzer, Albert 32, 80, 86, 88–90 Scornaienchi, Lorenzo 160 Seim, Turid Karlsen 258 Semler, Johann Salomo 59, 62–7, 71–2, 255 Shatzman, Israel 14 Shellard, Barbara 249 Sieber, John H. 16 Siegert, Folker 94 Simon, Richard 59 Simons, Eduard 2, 229–30 Skeat, T. C. 145 Skoven, Anne Vig 90, 95 Smith, Daniel A. 38, 251 Smith, William 172 Spinoza, Baruch de 60 Stanton, Graham N. 145 Stein, Robert H. 202, 207 Stillman, Martha 16 Stoldt, Hans-Herbert 5 Storr, Gottlob Christian 69, 81 Strauß, David Friedrich 5, 32, 37, 74, 86, 89, 223–4, 242 Strecker, Georg 76, 120 Streeter, Burnett Hillman 13, 93, 97, 100, 110, 112, 139, 203–4, 210, 216, 246 Strotmann, Angelika 190 Suggs, M. J. 166 Taussig, Hal 264 Taylor, Vincent 22–3 Telford, William R. 95 Thon, Dalit 190 Throckmorton, Burton 25–6

Author Index Tischendorf, Constantin von 19 Toland, John 60–2 Torm, Frederik 93 Traube, Ludwig 101 Tronier, Henrik 95 Tsutsui, Kenji 182, 275–6 Tuckett, Christopher M. 2–3, 16–17, 40, 79, 101, 116, 130, 134, 139–41, 144, 146–7, 151, 165, 189, 201, 207–9, 211, 214, 216 Tyson, Joseph B. 246, 255–6, 267 Vagany, Léon 13 Valantasis, Richard 192 van Unnik, W. C. 174 Vervenne, Marc 19 Vinzent, Markus 259, 267–74 Viviano, Benedict T. 222 Vogt, Ernst 159 Volkmar, Gustav 90, 95, 268 Walter, Nikolaus 16 Watson, Francis 3, 5–6, 14, 17, 22–4, 37, 42, 54, 98, 115–22, 124, 129–32, 141, 143, 151, 165, 185, 202, 207, 230–1 Webb, Robert L. 17

287

Weiß, Bernard 31, 33, 155 Weiße, Christian Hermann 2, 74–5, 87, 89, 224, 227–9 Weissenrieder, Annette 232 Wenham, John W. 29 Wernle, Paul 13 West, Stephanie 263 Wettstein, Johann Jakob 67 Whittaker, John 41 Wifstrand, Albert 95 Wilke, Christian Gottlob 75, 86, 89 Wille, Günther 172 Willert, Niels 95 Windisch, Hans 154 Witetschek, Stephan 166, 178 Wittkowsky, Vadim 94 Wolf, Christian 60 Wolter, Michael 186, 190, 228 Woolf, Gregory 156 Wrede, William 33, 90–1, 224 Wünsch, Dietrich 54 Youngquist, Linden 156 Zeller, Dieter 166, 171 Zimmermann, Ruben 188, 196, 227 Zöckler, Thomas 16

SUBJECT INDEX aesthetics 22, 23, 24–5, 93, 233 agraphon 37 allegory 90, 95, 178 alternating primitivity 99–100, 196, 199 Anglo-Saxons 93 apologetics 33, 38, 40–1, 47, 49, 62, 64, 68, 71, 75 Augustine 52–8, 74 Augustinian hypothesis 29, 31 Augustus, emporer 44 baptism 124–5, 181, 242–3 of Jesus 203, 212, 222 beatitudes 35, 146 Beelzebub controversy 135–6, 137, 203–4, 212, 216 Beispielgeschichte 196, 198 Benutzungshypothesen 152, 158 Bible 39, 52–3, 57–8, 60, 62, 67, 76, 146 birth narratives of Jesus 120–2, 227 canonical texts 15 canons 43, 51, 55, 62, 98 of persuasive speech 27–8 Catholic Epistles 65 Celsus 51, 58 Cerinthus’ Gospel 70 Christian art 89 Christian identity 265 Church doctrine 74 Codex Bezae 269 coherence and distinctness 198–9 coherence arguments 233 compositional practices 26 compositional process 143 Council of Basel-Ferrara-Florenz 59 Council of Trent 59 Critical Edition of Q (CEQ) 227, 243 crux interpretum 166 Cynic hypothesis 37

David 123, 142 Demetrius 173 Descartes 60 dialectical theology 93 diegesis theory 86 disharmonization 62 double traditions 97, 155, 186, 192, 204, 262 eclecticism 226 Editionswissenschaft 226 editorial scenarios 22–9 Enlightenment 60, 79 Epiphanius 31, 181–2, 254, 261, 268–9, 279–80 Erasmus of Rotterdam 58, 67 eschatological material 134, 194 ethical perfectionism 24 Eusebius of Caesarea 45–6, 53, 56, 128 Evangelion, see Marcion’s Gospel Evangelische Rhapsoden 84 evangelist(s) 6, 18, 23, 30, 36, 43–8, 53–5, 61, 75, 84, 87, 92, 100, 111–12, 116–18, 126, 130, 140–3, 145–6, 153, 181, 220, 249 exclusive agreements (EA) 236 exegetical axiom 231 Farrer–Goulder–Goodacre model 214 Farrer–Goulder theory 3–6, 116, 151, 185. See also Mark without Q and L/M theory Farrer hypothesis (FH) or theory 13, 20, 23, 26–9, 35, 38, 40, 116, 139, 153, 156, 158, 202, 214, 217–18 Form Criticism or History 80, 91, 95 Fragmentenstreit 5 fragment hypothesis 64, 66 fraud 30, 71–2

Subject Index genealogies, Matthew and Luke coincidence in 120–2 German Protestants 5 Gospel harmony 50, 53–4, 56–8, 62, 68–70, 74 Gospel Writing 120, 124, 133, 144, 146 Griesbach hypothesis 15, 31–3, 50, 67–70, 73–4, 81, 139, 201 harmonizations 234 Historical Criticism 56, 65 historical Jesus 6–7, 9–10, 32–3, 33–7, 41–2, 74, 79–80, 85–9, 91, 93, 100, 221–2, 224–5, 265 Hochliteratur 92 Holy Scriptures 57, 63, 66 Holy Spirit 48, 52–3, 56, 142 Huck–Greeven arrangement 18–19 Humanism 58–9 idealism 88 inclusive agreements (IA) 236 infancy narratives and genealogies of Luke and Matthew 142, 227 intergroup relationships 170–4 International Q Project (IQP) 19, 165, 226 interpretative tradition 58, 166–70 Irenaeus 46–7, 53, 56, 100, 250, 253–4 Jerome 31, 46 Jesus Christ 5–6, 28, 32, 35, 41, 44–5, 54, 56, 71–2, 75, 79, 85–6, 97, 107, 124, 167, 169–70, 174, 186, 191, 193, 231, 250, 256 Jesus’ sayings 10, 98, 102, 227 Jewish Christianity 65 Johannine ‘Christianity’ 154 John the Baptist 203, 212, 227, 229 Judaism 95, 161, 190 kashruth 39–40, 167 kerygma 38, 83 Kingdom of God 40, 43, 102 Kleinliteratur 91–2 lectio difficilior principle 111 Leibniz 60 Life-of-Jesus scholarship 88, 90

289

L/M hypothesis 8, 42, 115–18, 120, 125, 128–31, 133–4, 137, 139, 141, 143, 146, 151–2, 155, 157–8, 185, 202, 230–1, 234 Logia 31, 113 competing theories 107–10 discovering 100–5 interpretation of 105–7 Longer (Secret) Gospel of Mark 17 Lord’s Prayer 35, 37, 190, 194 Luke access to traditions 199 choice of material 131–4 Mark–Q overlap passages 214–17 major agreements 210–13 Marcan hypothesis 32 Marcion 50 Marcion’s Apostolikon 270 Marcion’s Gospel 70, 98, 234, 253–64,  267–9 as key to Synoptic Problem solving 270–8 as relevant for Synoptic Problem 278–81 Mark dealing with Q as a literary forerunner and/or competitor 162–3 knowledge of Q 159–60 silence about the Q movement and depiction of ‘others’ 160–2 without Q 157–8, 185, 245 Markan hypothesis 81, 85, 86–9 Mark–Q overlaps 124, 134, 140, 142, 155 Beelzebub controversy 135–6 definition of 201–7 direction of dependence 218–21 the problem 207–10 ‘Matthaean’ vocabulary 20–1 Memoirs of the Apostles 70 microconflation 26 minor agreements 20–1, 76, 137, 162, 209, 210–13, 229–30, 234, 236, 261, 278–80 miracles 224, 232, 235 mission discourse 204, 216, 218 multistage hypothesis 29 Muratorian Fragment 47

290

Subject Index

mustard seed, in Matthew and Luke 136–7, 136–7, 204, 212, 216 narratology 226–7, 234–5 Neo-Griesbach hypothesis. See Two Gospel hypothesis Nestle–Aland (NA) 18–19, 53, 231, 233 nomina sacra 101 non-Markan materials 129, 132–4, 136 oral tradition hypothesis 1 Origen 51–2, 56, 58 Papias 15, 46–7, 53, 56, 64, 106, 128–9, 156, 223–4, 228–9, 232, 235 pater familias 187–8, 190, 192–4, 198 Pauline letters 50, 65, 96, 154, 157, 161, 247, 269, 272 Pauline theology 90, 95 Pentecostals 83 Peter–James–John material 34 plausibility arguments 22, 130–1 Pliny 51 polyphony 45, 49–50, 55–6, 62–5, 71–2 Porphyry 52–3, 58 ‘practical teachings’ of Jesus 224 preaching 226 of John the Baptist 262 proof-from-prophecy claim 249 Prosgymnasmata 27 proto-Gospel hypothesis 1, 61 Proto-Luke hypothesis 93, 246 Proto-Mark hypothesis 76 Q hypothesis 2–4, 6, 10, 17, 19, 23–4, 26, 28, 33–4, 36, 38–40, 49, 93, 97–101, 108, 110, 115–16, 118, 130, 139–40, 146, 151, 153–5, 185, 196, 201, 236 coincidences in 120–8 factuality of 227 and Mark 153–7 parallel beginning to Mark’s Gospel 227–30 politics of 110–13 self-referentiality 179 redactional procedures 116–17, 119, 140, 198 redaction criticism 80, 93, 152–3

Redensammlung 108 Reformation 55–6, 58–9, 62 Renaissance 58 Reuchlin 58 rewritten Scripture 96 rhetorical virtues 27 Sabbath 39–40, 102 Sayings Collection hypothesis 98–9, 103, 105–9, 146, 158, 225, 229 Sayings Gospel 225 sayings source 2–5, 10, 49, 224–5, 232. See also Q hypothesis Scriptures 56–9, 65, 68, 250 secondary evangelists 116–18 Semi-Gospel 225 Sermon on the Mount 19, 22–3, 25–6, 131–2, 134, 143, 224, 228, 242 Sermon on the Plain 22, 25, 228 slaves/slavery 192, 194 in the household of God 194–8 Socrates 88 sola scriptura 56–9, 75 Solomon 123, 142 Sondergut 185, 197–8 Son of God 85 Son of Man 32, 85, 169, 174–7 source-critical hypotheses 112, 143 Spruchsammlung 108 structuralism 226, 234 supernaturalism 60, 81 Synopsis 68 synoptic analysis 236–44 Synoptic Gospels 16, 61, 80–1, 87–8, 91, 93–5, 111, 224, 230, 233 Synoptic Problem (SP) 1–3, 5–6, 9–10, 13–15, 62, 72, 76, 98, 120, 130, 134, 151, 165, 201, 221, 230, 233, 260–4 birth of the 30–1 data description 18–20 editorial rationalization 22–9 historical development, reflection on 76–7 and the historical Jesus 33–7 Holtzmann’s treatment of 31–3 importance of 41–2 intracanonical solution for 98 logical inferences 20–2

Subject Index primary data for 14–18 solution without the Q hypothesis 6 synoptic re-lecture model 230–4 Tatian 43, 50, 70 teachings of Jesus 224–5 Tertullian 181–2, 254, 257–8, 261, 268–9, 271–2, 274–6, 279–80 Theophilus 49 Thirty Years’ War 60, 62 tradition 14. See also double traditions; triple traditions triple traditions 155, 210, 237 Tübingen School 81, 87 Two-Document hypothesis (2DH) 13, 20, 24, 26, 28–9, 33–4, 36–8, 40, 42, 265, 278 Two-Gospel hypothesis (2GH) 2–4, 13, 20, 24–5, 29, 33, 35, 37–8, 40, 265

291

Two-Source hypothesis (2SH) 2–6, 10, 71–7, 79, 115, 120, 122, 124, 131, 134, 137, 151, 202, 224, 227–31, 233, 245, 260, 263 continuations of 91–4 universal narrative schema 226,  234–5 Ur-Gospel hypothesis 15, 30, 67, 69, 70–1, 73, 75, 77 UrMarkus 31–2, 88 utilization hypotheses 1 Voltaire 67 Vorlage 158, 191, 195–6, 199 Vulgata 59 Word of God 45, 53, 65–6, 243

INDEX OF REFERENCES Old Testament Genesis 16.11 25.34 27.17 28.20–21

122 191 191 188

Exodus 16 16.35 20.10 22.11 23.20

191 188 187 26 205, 237

Leviticus 22.13 Deuteronomy 5.14 5.15 21.18–21 21.20

187

32.5

187 26 175 167, 175, 176 171

Judges 8.4–6 19.2

191 187

1 Samuel 10.5

173

1 Kings 13.15 17.1 17.11 18.41–45 19.8

188 176 191 176 175, 176

1 Chronicles 3.5 25.1

123 173

Psalms 77 77.24–25 78.8 119

272 191 171 35

Proverbs 31.15

191

Isaiah 4.1 6.2 7.14 8.3 25.6–8 40.1 40.1–2 40.3 52.7 55.1–2 61 61.1 65.13–14

188 46 122 122 174 205 44 180 44 174 35 44 174

Jeremia 2.31 7.29 22.28–30 36.30–31

171 171 123 123

Ezekiel 1.10

46

Daniel 4.12 4.21

137 137

Hosea 1.1–4 1.6

122 122

Micah 7.6

189

Malachi 3.1 4.5–6

New Testament Matthew

1–2 1.1–4 1.6–16 1.18–25 1.21 2.1–12 2.2 2.10 2.19–23 3 3.1–12 3.1a 3.3 3.4–6 3.5b 3.6 3.7 3.7–9 3.7–10 3.7–12 3.8–10 3.9 3.10 3.11

Index of References

293

3.11–12

140, 205, 206, 214 237 229 244 204, 207, 212, 220, 221 215 203, 212 125, 215 243 243 262 243 243 143, 203, 212 262 204 38 22 194, 213 228 142 143 142 132, 228 144 142, 242 143 35 35 24, 153 277 38 26 26 194 23 194 24 23 23 24 23 26 23 26 220, 221

180, 205, 237 180

47, 53–4, 61, 68–9, 73, 81–2, 90 , 94 , 224 , 227 , 229 , 236 –7 , 243 121, 242 142 142 142 121, 122 121 121 121 242 221 203, 212, 214, 238–41 236 205, 237 243 242 242 219, 243 262 204, 212, 229, 243 143 220 221 220, 221 19, 206, 207, 214, 243

3.11a 3.11b–12 3.11c–12 3.12

3.13 3.13–17 3.14–15 3.15 3.16 3.17 3.18 3.19–20 4.1–11 4.2b–13 4.3–10 4.5 4.11 4.13 4.23–25 4.25 4.25–5.1 4.25–5.2 5–7 5–25 5.1 5.2–7.27 5.3 5.3–10 5.3–12 5.13 5.35 5.42 5.46–48 6.8 6.9–13 6.14–15 6.19–21 6.22–3 6.24 6.24–26 6.25–34 7.2 7.7–11 7.12 7.15–20

294 7.19 8 8.1–4 8.5–13 8.11–12 8.12 9.1–8 9.9–13 9.20 9.32–34 9.34 10 10.5–15 10.11 10.11–15 10.14 10.16–39 10.19–20 10.21 10.21–22 10.34 10.34–39 10.37 11.1–19 11.5–8 11.8 11.10 11.12–15 11.16 11.34–36 12 12.1–8 12.11 12.22–37 12.25–32 12.26 12.27–28 12.28–34 12.29 12.31–32 12.33 12.34 12.36 12.39–42 12.46–50 13 13.16–17 13.24–30

Index of References 220, 221 194 212 144 144 221 212 212 261 135 135 228 204 219 218, 219 219 144 144 193 23 274, 275 193 193 182 23 174 205, 237 183 174 23 135 153 144 27 204, 212, 216 27 135 213 135 144 221 221 221 171 204 228 144 197, 220, 221

13.30 13.31–32

13.31–33 13.33 13.34 13.36–43 13.38 13.41 13.42 13.48 13.48–49 13.49 13.50 14 14.13–21 14.28–31 15.3–6 16–17 16.2–3 16.4 16.6 16.13 16.17–19 17 17.3 17.10–13 17.11–13 17.12–15 17.14 17.17 17.19–20 18 18.6 18.6–21 18.23–34 18.23–35 19.12 19.23–30 19.28 21.1–9 21.23–27 21.28–32 22.1–10 22.9 22.13 22.34–40 23

221 133, 136, 144, 204, 212, 216, 217 133, 136 204 221 220, 221 180 221 221 220 221 221 221 243 213 208 194 215 144 171 144 23 208 176 176 124 176 176 176 171, 261 144 228 144 144 197 38 37 23 144 213 213 133 38 39 221 213 132, 199

Index of References 23–25 23.36 24–25 24.1–13 24.17–18 24.23–25 24.26–27 24.28 24.34 24.37–25.30 24.37–39 24.51 25.1–13 25.14–30 25.30 25.31 25.31–46 25.32 26.57–75 28 28.1 28.9 Mark

1–16 1.1 1.1–3 1.1–6 1.1–8

1.2

1.2–3 1.2b 1.4 1.5–6 1.7

228 171 132, 134, 136 220 133, 144 133, 144 144 144 171 144 144 221 133 38, 195 221 221 39, 220 221 213 121 259 259 46–7, 61, 73, 75, 81–2, 84, 86, 91, 93–4, 107, 224, 229, 236, 243 160, 237 162 162 124, 203 203, 212, 214, 227–30, 235–41, 244 155, 162, 203, 228, 280 180, 205, 237 237 125 243 124, 206, 214

1.7–8

1.8 1.8a 1.8b 1.9–10 1.9–11 1.12–13 1.14–15 1.16–20 1.18 1.20 1.21 1.21–39 1.22 1.29 1.29–3.19 1.29–31 1.32–34 1.40–45 2 2.1–12 2.13–17 2.15 2.18 2.18–19 2.18–22 2.23–28 2.23–38 2.32–34 3.1–6 3.7–8 3.7–12 3.13 3.13–19 3.21 3.22–26 3.22–27 3.23–30 3.27 3.27–29 3.28–29 3.31–35 3.35 4 4.10–12 4.21

295 124, 203, 205, 206, 124, 214 19, 124, 206 237 244 215 203, 212 125, 203, 212 44 140 193 193 19 140 160 19 135 19, 153, 193 19 19, 212 160 19, 153, 212 19, 212 182 160, 161 160 19 19, 153 40 19 19, 40 142, 143 19 142, 143 19 193 155, 203 153 136, 204, 212, 216 135, 203 155 203 204 193 136, 159 153 155

296 4.21–25 4.22 4.24 4.25 4.30–32

4.31–32 4.35–41 5.1–20 5.21–43 5.27 6 6.1–6 6.3 6.4 6.6b–13 6.7–8 6.7–13 6.10–11 6.30–44 7–8 7.1–23 7.10–12 7.15–23 7.24–30 8.11–12 8.29 8.33 8.34–35 8.38 9.2–10 9.19 9.35 9.37 9.40 9.42 9.42–50 9.49 9.50 10.10–11 10.19 10.23–31 10.29 10.29–30 10.31 10.41–45 10.42–45 11.1–10

Index of References 204 155 155 155 153, 155, 162, 204, 212, 216, 217 136 153 25 25 153, 261 243 140 121 198 204, 219 159 155 218, 219 213 155 244 193 40 244 155 280 208 155 155 153 153, 261 197 155 155 155 204 277 155, 277 155 193 23 193 193 155 23 197 213

11.22–23 11.25 11.27–33 12 12–13 12.1–11 12.18 12.28–24 12.28–34 12.37b–40 12.38–40 13

13.2 13.4–5 13.9 13.11 13.12 13.15–16 13.21–23 13.33–37 13.34–36 14.53–72 14.58 14.65 15.29 15.44–45 16.1 16.7 16.8 16.9–20 Luke

1 1–2 1.1 1.1–2 1.1–3 1.1–4 1.3 1.4 1.15 1.26–56

155 194 213 160 155 197 160 203 211 155 204 32, 132, 134, 136, 159, 160 162 162 155 155 193 133 133 155 197 213 162 153 162 153 259 244 122 55 50, 54, 81–2, 90, 92, 174, 227, 234, 236–7, 269 256 141, 245, 260 49, 96 162 235 47, 49, 50, 119, 232 119, 128 235 175 142

Index of References 1.31 1.53 1.76 2 2.10 2.31–32 3 3–4 3–6 3.1 3.1–2 3.1–18 3.1–20 3.1a 3.3 3.4 3.4–6 3.5–6 3.7 3.7–9

3.7–18 3.8 3.8–9 3.8b 3.9 3.10–14 3.10–16 3.12 3.12–14 3.15–17 3.16

3.16–17 3.16b 3.16b–17 3.16c–17 3.17

3.18 3.18–20

121, 122 24 180 256 121 251 221, 262 158 120 260 237, 243 203, 212, 214 124–5, 238–41 236 242 237 205 237, 243 179–81, 219 124, 125, 180, 204, 212, 229, 243, 262, 263 143 198 220 180 220, 221 125, 244 237, 243 182 177, 243 205, 206, 214 180, 206, 207, 214, 243 140 237 229, 262 244 124, 204, 207, 212, 220, 263 244 215, 237, 243

3.21–22 3.23–31 4 4.1–13 4.2b–13 4.3–4 4.3–12 4.4 4.16 4.16–30 4.18 4.23 4.31–44 4.38–6.19 4.43 5.1–11 5.12 5.12–16 5.17–26 5.18 5.27–32 5.29 5.30 5.33 5.33–35 5.34 5.35 6.1–5 6.17 6.17–18 6.17–19 6.20–23 6.20–26 6.20–49 6.20b 6.30 6.31 6.32–34 6.35–36 6.36 6.37–38 6.42 6.43–45 7 7.1 7.1–10

297 203, 212, 215 142 262 143, 203, 212, 262 263 191 204 191 213 140 24 244 140 135 255 140 279 212 212 279 212 182 175, 182 175, 182 182 177 177 153 142 142 242 153 146 185, 260, 261 35 26 26 26 190, 191, 194, 198 26 26 102 220, 221 181 160 39

298 7.18–20 7.18–28 7.18–30 7.18–35 7.19 7.20 7.21 7.22–23 7.23 7.24

7.24–28 7.24–35 7.25 7.26 7.27

7.28 7.29

7.29–30 7.29–32 7.29–35 7.30 7.31 7.31–32 7.31–34 7.31–35

7.32 7.32b 7.33

7.33–34 7.33–35 7.34

Index of References 182 181 170, 183 169, 170, 176 170, 180 180 179, 180 182 181 169, 174, 179, 180, 196 179, 180 170 174, 180, 196 174, 180, 182 162, 170, 181, 182, 205, 237 182 168, 177, 178, 181, 182, 195 176, 177, 183 181 181 178 179, 180 165, 166, 180 178 165, 166, 169, 170–82 173, 176 173 169, 172, 173, 175, 181 165–7, 169, 172, 176, 182 181 169, 173–5, 177, 181, 182, 198

7.35

7.73 8 8.19–20 8.44 9.1–6 9.4–5 9.5 9.7–8 9.10–17 9.20 9.41 9.51–18.14 9.57–58 9.57–60 9.59–60 10.1–12

10.5–11 10.7 10.8–11 10.13–14 10.15 10.23–24 10.25–28 11 11.1 11.2–4 11.5–8 11.11 11.11–13 11.13a 11.14–22 11.14–23 11.14–36 11.15 11.17 11.17–23 11.19 11.27–28 11.29–32 11.30 11.31 11.32

165, 166, 168, 170, 178, 179 182 136 204 261 204, 212, 216, 219 218, 219 219 182 213 280 171, 261, 280 131, 132 189 189 188 204, 212, 216, 218, 219 219 40 219 39 160 144 213 136, 181 182 35, 190 191 191 190 190 198 141 27 135, 181 28 204, 212, 216 28 204 171 181, 199 123 39

Index of References 11.34–36 11.36 11.39–52 11.49 11.49–51 11.50–51 11.51 12 12.1 12.1–13.9 12.10 12.11–12 12.20–21 12.27 12.28 12.29–31 12.30 12.30b 12.31 12.31b 12.33–34 12.37 12.42 12.42–46 12.45d 12.46–47 12.47–48 12.51

12.53 12.54–56 13 13.18–19

13.18–21 13.20 13.21–22 13.25 13.28 13.28–30

28 28 185, 199 181 178 171 199 136 144 25 144, 204, 216 144 196 123 174, 177, 178, 195 191 195 191 174, 177, 180 191 24 198 198 192, 194, 195, 198 192 195 198 189, 193, 199, 274, 275 189, 193, 199 144 136, 137 136, 144, 162, 199, 204, 212, 216, 217 133, 136, 141 204 135 133 174, 198, 221 144

13.29 13.31 13.35 13.37 14.5 14.15 14.16–23 14.25–35 14.26 14.27 14.34–35 15.1 15.1–17.10 15.11–32 16.1–7 16.8 16.9–12 16.13 16.16 16.16–17 16.16–18 16.19 16.19–31 17 17–19 17.2 17.5–6 17.7–10 17.20–21 17.22–37 17.23–24 17.25 17.26–27 17.31 17.37 18.10–13 19.2–7 19.12–26 19.12–27 19.28–38 20.1–8 20.4 21 21.31 22.16 22.24–30 22.29 22.29–30 22.54–71

299 174 251 162 251 40, 144 174 197 277 189, 193, 199 189, 196 277 182 25 133 198 23 23 195 255 36 199 24 221 136 162 144 144 198 133, 144 199 144 171 144 133, 144 144 182 182 195 195, 199 213 213 174 132, 136 178 174 23 144 174 213

300 23.4 23.6 23.14 23.18 24

24.1–11 24.3b 24.4 24.6a 24.12 24.13–35 24.13–53 24.27 24.27b 24.30–31 24.32 24.36–43 24.36b 24.38 24.39 24.39b 24.40 24.41–43 24.43 24.44 24.44–45 24.45 24.51b 24.52a John

1 1.45 3.13 3.29 3.51 4.2 6.42 20 20.11–18 20.14–17 21 21.9–13

Index of References 174 174 174 174 121, 245, 248–53, 257, 259, 260 259 248 279 248 248 259 244 257 249 249 250 257 248 251 258 250 248 251 252 249 257 250 248 248 47–8, 80, 82–3, 86, 89, 94 124 121 176 177 177 124 121 249 259 259 249 251

Acts 1.1–8.4 1.3 1.12–26 1.22 2 2.31–32 2.40 4 8.31 10 10–15 10.37 10.40–43 10.40b–43a 10.43 12.12 13 13.25 13.31 13.37 15 17.2–3 20.35 21.8

244 252 252 124 123 252 171 123 250 40 244 124 252 252 252 68 123 180 252 252 40 250 107 45

Romans 1.3 1.16–17

121, 122 44

1 Corinthians 9.1 15.5–8 15.35–50

259 259 251

2 Corinthians 10–13

161

Galatians 1–2 1.8 1.12 3.1b 4.4

161 272 259 45 121, 177

Ephesians 4.11

45

Colossians 4.16

159

Index of References 2 Timothy 4.1 4.5

252 45

James 5.17

176

1 Peter 4.5 5.13

252 46

2 Peter 1.20 3

250 162

Revelation 1.10–19 3.20 4.7 19.9

259 174 46 174

Apocrypha, Old Testament Sirach 25.22 188 33.20–22 188 33.24 188 Tobit 1.17 Wisdom of Solomon 16.20–21 Pseudepigrapha 1 Enoch 62.14 2 Baruch 29.4 50–51 4 Ezra 7.26–115 Joseph and Asenath 11–12 Testament of Job 21–25

191

301

Dead Sea Scrolls  1QH xvii 34–36

190

4QDeut

26

4QPhyl G (4Q134)

26

8H.evXIIgr

26

8QPhyl (8Q3)

26

Ancient Jewish Writers Josephus Jewish Antiquities 247 18.5.2 178 18.116–19 168, 181 Pseudo-Phocylides 153–174

188

Apostolic Fathers 1 Clement 13 13.2

16 107

2 Clement 1.1

252

Barnabas 5.6–7 7.2

251 252

Didache 1.3b–2.1 16.3–8

16 16

Polycarp Philippians 2.1

252

191

174

174 258

Apocrypha, New Testament Dialogue of the Saviour 16, 259 258

190

188

Egerton Gospel

98

Gospel of Barnabas

61

Gospel of the Hebrews

15, 70

302

Index of References

Gospel of Mary

130, 259

Gospel of the Nazarenes

15, 66, 72

Gospel of Peter

16, 98

Gospel of Pseudo-Matthew

44

Gospel of Thomas

6, 14, 16, 37–8, 41–2, 47, 98–9, 108, 130, 146, 153 102 102 37

26–28 27.1 98 Pistis Sophia

259

Early Christian Writings Adamantius Dialogue 254, 268–9, 274–6, 280 Eusebius  Ecclesiastical History 3.39 3.39.4 3.39.15 30.20.4 Eznik of Kolb On God 407 Jerome De viris illustribus 18 Justin 1 Apology 39.3 67.8 Dialogue with Trypho 53.5 76.6 106.1 Pseudo-Justin De resurrectione 9.5–9

128 156 156 252

258

156

253 253

253 253 253

253

Tertullian Against Marcion 3.9 3.14.5 4.4.5 4.11.12 4.20.8 4.21.6 4.22.6 4.23.1 4.23.2 4.29.13 4.34.16 4.43.6–7 5.1.9 5.4.2 5.5.1 5.7.11 5.19.5

258 275 272 272 261 280 280 261 261 274 280 257 272 270 272 272 272

Antidote for the Scorpion’s Sting 10.17 275 Classical Writings Herodotus 1.141

172

Ovid Metamorphoses 15.824–28

258

Papyri P.Egerton II

17

P.Oxy.1

100, 101, 113

P.Oxy.2 = p1

105

P.Oxy.654

101, 108

P.Oxy.655

101, 108

P.Yadin 10 l. 9–10, 14

188

P.Yadin 19 l. 15–16

188

P.Yadin 37 l. 9

188