The Genre and Development of the Didache: A Text-Linguistic Analysis 978-3161483981

Table of contents :
Preface ..............................................................................................................V
List of Abbreviations......................................................................................
Introduction .......................................................................................................1
Chapter 1: The Didache and the Question of Genre .........................................5
1.1 Introduction.............................................................................................5
1.2 Status Quaestionis...................................................................................8
1.3 The Didache as Church Order – Early Research .................................31
1.3.1 Origin of the Term Church Order and its Application to the
Didache.........................................................................................31
1.3.2 Discussion of Genre Prior to the Discovery of the Jerusalem
Manuscript ....................................................................................32
1.3.3 Initial Research on the Didache of H54........................................37
1.3.4 Research at the Turn of the Nineteenth Century...........................43
1.3.5 The Doctrina Apostolorum and the Two Ways ............................46
1.3.6 The Didache and the Dead Sea Scrolls.........................................49
1.3.7 The Genre of the Two Ways .........................................................50
1.4 Studies on Composition, Communication, and Genre..........................52
Chapter 2: A Text-Linguistic Analysis of the Didache...................................65
2.1 Text-Linguistic Methodology ...............................................................65
2.1.1 Introduction...................................................................................65
2.1.2 Text Linguistics and Early Christian Literature............................68
2.1.3 Delimitating a Text .......................................................................69
2.2 The Text of the Didache .......................................................................74
2.3 Structurally Delimiting the Didache of H54.........................................80
2.3.1 Introduction...................................................................................80
2.3.2 Text-Linguistic Abbreviations ......................................................83
2.3.3 Presentation of Delimitation .........................................................84
2.3.4 Structural Overview Based on the Text-Linguistic Analysis........96
2.3.5 Communication Levels .................................................................97
2.3.6 Dramatis Personae .......................................................................99
2.4 The Delimitation of the Didache: Analysis in Detail .........................101
2.4.1 Text-Levels 00ST/0ST – The Titles of the Didache...................101
2.4.1.1 The Titles of the Didache and Their Implications
for Genre..............................................................................105
2.4.1.2 Text-External Functional Aspects of the Incipit .................123
2.4.2 Text-Level 1................................................................................125
2.4.3 Text-Level 2................................................................................133
2.4.4 Text-Level 3................................................................................138
2.4.5 Text-Level 4................................................................................139
Chapter 3: The Generic Development and Compositional History of the
Didache........................................................................................141
3.1 The Genre Didache.............................................................................141
3.2 The Structure of the Didache..............................................................155
3.2.1 Text Level 1 ................................................................................155
3.2.2 Text Level 2 ................................................................................156
3.2.3 Text Level 3 ................................................................................160
3.2.4 The Two Ways Section ...............................................................162
3.3 A Reconstruction of the Development of the Didache.......................184
3.3.1 Stage 1.........................................................................................184
3.3.2 Stage 2.........................................................................................184
3.3.3 Stage 3.........................................................................................185
3.3.4 Stage 4.........................................................................................186
3.4 The Further Development of the Genre Didache ...............................186
Chapter 4: Summary and Conclusion............................................................189
Bibliography of Works Cited ........................................................................193
Reference Index.............................................................................................211
Modern Author Index ....................................................................................225
Subject Index .................................................................................................230

Citation preview

Wissenschaftliche Untersuchungen zum Neuen Testament · 2. Reihe Herausgeber / Editor Jörg Frey (Zürich) Mitherausgeber / Associate Editors Friedrich Avemarie † (Marburg) Markus Bockmuehl (Oxford) James A. Kelhoffer (Uppsala) Hans-Josef Klauck (Chicago, IL)

339

Nancy Pardee

The Genre and Development of the Didache A Text-Linguistic Analysis

Mohr Siebeck

Nancy Pardee, born 1956; 1985 MA, 2002 Ph.D. from the University of Chicago; adjunct instructor at Saint Xavier University in Chicago, 2003–11 member of the steering committee for the Didache in Context Section of the Society of Biblical Literature.

e-ISBN PDF 978-3-16-152475-2 ISBN 978-3-16-148398-1 ISSN 0340-9570 (Wissenschaftliche Untersuchungen zum Neuen Testament, 2. Reihe) Die Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2012 by Mohr Siebeck, Tübingen, Germany. www.mohr.de This book may not be reproduced, in whole or in part, in any form (beyond that permitted by copyright law) without the publisher’s written permission. This applies particularly to reproductions, translations, microfilms and storage and processing in electronic systems. The book was printed by Laupp & Göbel in Nehren on non-aging paper and bound by Buchbinderei Nädele in Nehren. Printed in Germany.

 

Preface The following study is a revision of my doctoral dissertation, “The Genre of the Didache: A Text-Linguistic Analysis,” accepted by the Department of New Testament and Early Christian Literature of the University of Chicago in 2002. Early in my doctoral studies I had the exceptionally good fortune of sitting in on lectures given by Prof. David Hellholm while he was a visiting professor at the Divinity School of the University of Chicago. Prof. Hellholm’s lectures focused on the method of text linguistics and its usefulness for the study of the New Testament and other texts of the early Church. Though having little training in linguistics, I was nonetheless fascinated by his presentation and demonstration of the text-linguistic method on the Shepherd of Hermas and other apocalyptic works. Through the application of this method, Prof. Hellholm was able to show distinct generic characteristics that worked together to differentiate an apocalypse from other genres of ancient literature. The success of his method on these texts was especially intriguing to me since I was concurrently struggling with another of the texts of the Apostolic Fathers, the Teaching of the Twelve Apostles, commonly referred to as the Didache. Vastly different in content and character, the two texts nonetheless shared a common problem: an imprecise understanding of their ancient genres appeared to be impeding a clear understanding of their meaning and function in the early Church. What, I wondered, could text linguistics tell us about the Didache? With the approval of my dissertation committee and the very generous mentoring of Prof. Hellholm, I undertook the task of acquiring a basic knowledge of text-linguistic methodology and applying it to the text of the Didache in the hope of using the results as a component of my dissertation on the development of the text. The results, I believe, have proven fruitful and, combined with data derived from other methods and sources, have resulted in better insight into the genre, developmental stages, and ultimately the meaning and purpose of this important and very ancient Christian text. Due to a variety of circumstances, the completion of the dissertation itself and of this revised version has been inordinately long and it is only thanks to the support of many people that I have been able to finish this project. First of all I must extend my deepest gratitude to two extraordinary scholars: my longtime professor, Robert M. Grant, who, despite my initial deficiencies, none-

 

VI  

Preface

theless saw potential in me, admitted me to the university (twice), and continuously encouraged my progress; and, equally, my dissertation advisor, Adela Yarbro Collins (now of Yale Divinity School), who, upon the retirement of Prof. Grant, graciously agreed to take over the direction of my dissertation at an early stage, a task that, in the end, necessitated a great time commitment on her part. Without the crucial suggestions and corrections, as well as the patience and encouragement, of Prof. Yarbro Collins this work could never have been completed and it is she, as well, who must be thanked for recommending it for publication. My sincere thanks must also be expressed to the other members of my committee: first Prof. Hans Dieter Betz, whose challenging New Testament classes resulted in immeasurable intellectual growth for me and whose important questions and suggestions indeed made my dissertation more sound; and to Prof. Arthur J. Droge, whose guidance and continued encouragement have also been much appreciated. And of course I am deeply grateful to Prof. Hellholm for his incredible willingness not only to instruct me, but also to dialogue with me, on my text-linguistic analysis of the Didache – his many suggestions and criticisms, his sharing of his own publications and of references to other helpful resources, and his continual support of my work have, indeed, been crucial. Secondly I must extend my sincere appreciation to Mohr Siebeck for the opportunity to publish my work: first and foremost to Dr. Henning Ziebritzki, Editorial Director, Theology and Jewish Studies, for his patience and everkind encouragement; to Dr. Jörg Frey, Universität Zürich, for approving my work for publication and providing many important suggestions for revision and supplementation; and to Ms. Tanja Idler, production manager at Mohr Siebeck, for her acute eye for detail and continual support. Of course I give my most heartfelt thanks to all of my family and friends for their unwavering support, but especially to my parents, Victoria, Arthur† and John†, to my son Timothy, and most of all to my husband, Dennis, for his patience and understanding, not to mention his many helpful suggestions. Finally, while acknowledging the help I have received from others, I take full responsibility for the content of the study here presented – I have truly endeavored to represent fairly the views of all whose work I have used and apologize in advance for any omissions or misrepresentations that may be found – they are certainly unintentional. Nancy Pardee, October 2012

 

 

Table of Contents Preface ..............................................................................................................V List of Abbreviations ...................................................................................... IX Introduction .......................................................................................................1 Chapter 1: The Didache and the Question of Genre .........................................5 1.1 Introduction.............................................................................................5 1.2 Status Quaestionis...................................................................................8 1.3 The Didache as Church Order – Early Research .................................31 1.3.1 Origin of the Term Church Order and its Application to the Didache.........................................................................................31 1.3.2 Discussion of Genre Prior to the Discovery of the Jerusalem Manuscript ....................................................................................32 1.3.3 Initial Research on the Didache of H54........................................37 1.3.4 Research at the Turn of the Nineteenth Century...........................43 1.3.5 The Doctrina Apostolorum and the Two Ways ............................46 1.3.6 The Didache and the Dead Sea Scrolls .........................................49 1.3.7 The Genre of the Two Ways .........................................................50 1.4 Studies on Composition, Communication, and Genre..........................52 Chapter 2: A Text-Linguistic Analysis of the Didache...................................65 2.1 Text-Linguistic Methodology ...............................................................65 2.1.1 Introduction...................................................................................65 2.1.2 Text Linguistics and Early Christian Literature............................68 2.1.3 Delimitating a Text .......................................................................69 2.2 The Text of the Didache .......................................................................74 2.3 Structurally Delimiting the Didache of H54.........................................80 2.3.1 Introduction...................................................................................80 2.3.2 Text-Linguistic Abbreviations ......................................................83 2.3.3 Presentation of Delimitation .........................................................84 2.3.4 Structural Overview Based on the Text-Linguistic Analysis........96 2.3.5 Communication Levels .................................................................97 2.3.6 Dramatis Personae .......................................................................99

 

Table of Contents

2.4 The Delimitation of the Didache: Analysis in Detail .........................101 2.4.1 Text-Levels 00ST/0ST – The Titles of the Didache...................101 2.4.1.1 The Titles of the Didache and Their Implications for Genre..............................................................................105 2.4.1.2 Text-External Functional Aspects of the Incipit .................123 2.4.2 Text-Level 1 ................................................................................125 2.4.3 Text-Level 2 ................................................................................133 2.4.4 Text-Level 3 ................................................................................138 2.4.5 Text-Level 4 ................................................................................139 Chapter 3: The Generic Development and Compositional History of the Didache........................................................................................141 3.1 The Genre Didache.............................................................................141 3.2 The Structure of the Didache..............................................................155 3.2.1 Text Level 1 ................................................................................155 3.2.2 Text Level 2 ................................................................................156 3.2.3 Text Level 3 ................................................................................160 3.2.4 The Two Ways Section ...............................................................162 3.3 A Reconstruction of the Development of the Didache.......................184 3.3.1 Stage 1.........................................................................................184 3.3.2 Stage 2.........................................................................................184 3.3.3 Stage 3.........................................................................................185 3.3.4 Stage 4.........................................................................................186 3.4 The Further Development of the Genre Didache ...............................186 Chapter 4: Summary and Conclusion............................................................189 Bibliography of Works Cited ........................................................................193 Reference Index.............................................................................................211 Modern Author Index ....................................................................................225 Subject Index .................................................................................................230

 

Abbreviations Modern Publications The following abbreviations have been appropriated from the SBL Handbook of Style (ed. Patrick H. Alexander, John Kutsko, et al.; Peabody, Mass.: Hendrickson, 1999) or from Siegfried M. Schwertner, Internationales Abkürzungsverzeichnis für Theologie und Grenzbebiete (2d ed.; Berlin; New York: de Gruyter, 1992). The abbreviations of ancient texts can be found in the Index of References. ACW ANRW

AThR AUSS AUU BDAG

BDF BETL BrSR BSac BZNW CCSL ConBNT CRINT CSCO CSEL DJD DLZ EKK FC FRLANT GCS GKC HM

Ancient Christian Writers. 1946– Aufstieg und Niedergang der römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung. Edited by H. Temporini and W. Haase. Berlin, 1972– Anglican Theological Review Andrews University Seminary Studies Acta Universitatis Upsaliensis Bauer, W., F. W. Danker, W. F. Arndt, and F. W. Gingrich. Greek-English Lexicon of the New Testament and Other Early Christian Literature. 3d ed. Chicago, 1999 Blass, F., A. Debrunner, and R. W. Funk. A Greek Grammar of the New Testament and Other Early Christian Literature. Chicago, 1961 Bibliotheca ephemeridum theologicarum lovaniensium Brown Studies in Religion Bibliotheca sacra Beihefte zur Zeitschrift für die neutestamentliche Wissenschaft Corpus Christianorum: Series latina. Turnhout, 1953– Coniectanea neotestamentica or Coniectanea biblica: New Testament Series Compendia rerum iudaicarum ad Novum Testamentum Corpus scriptorum christianorum orientalium. Edited by I. B. Chabot et al. Paris, 1903– Corpus scriptorum ecclesiasticorum latinorum Discoveries in the Judaean Desert Deutsche Literaturzeitung Evangelisch-Katholischer Kommentar zum Neuen Testament Fontes christiani Forschungen zur Religion und Literatur des Alten und Neuen Testaments Die griechischen christlichen Schriftsteller der ersten (drei) Jahrhunderte Gesenius’ Hebrew Grammar. Edited by E. Kautzsch. Translated by A. E. Cowley. 2d. ed. Oxford, 1910 Hallische Monographien

X HNT HSem HTKNT HTR ICC IG Irén JAC JAC.E JBL JECS JR JSJSup JSNTSup JSPSup JTS KEK KlT KuR LCC LCL LCO LSJ LTK MM NewIDB NHS NovT NTS OCD OrChr OrChrAn PG PGL PVTG RB RevScRel RHR RQ RSR RTP SBLSymS SBLTT

Abbreviations Handbuch zum Neuen Testament Horae semiticae. 9 vols. London, 1908–1912 Herders theologischer Kommentar zum Neuen Testament Harvard Theological Review International Critical Commentary Inscriptiones Graecae. Editio minor. Berlin, 1913 [v. 1, 1924]–1940. Irénikon Jahrbuch für Antike und Christentum Jahrbuch für Antike und Christentum Ergänzungsband Journal of Biblical Literature Journal of Early Christian Studies Journal of Religion Supplements to the Journal for the Study of Judaism Journal for the Study of the New Testament: Supplement Series Journal for the Study of the Pseudepigrapha: Supplement Series Journal of Theological Studies Kritisch-exegetischer Kommentar über das Neue Testament (Meyer-Kommentar) Kleine Texte Kirche und Recht Library of Christian Classics. Philadelphia, 1953– Loeb Classical Library Letture cristiane delle origini Liddell, H. G., R. Scott, H. S. Jones, A Greek-English Lexicon. 9th ed. with supplement. Oxford, 1983 Lexicon für Theologie und Kirche Moulton, J. H., and G. Milligan. The Vocabulary of the Greek Testament. London, 1930. New Interpreter’s Dictionary of the Bible. Edited by Katherine Doob Sakenfeld. 5 vols. Nashville, 2006–2009. Nag Hammadi Studies Novum Testamentum New Testament Studies Oxford Classical Dictionary. Edited by N. G. L. Hammond and H. H. Scullard. 2d ed. Oxford, 1970. Oriens christianus Orientalia christiana analecta Patrologia graeca [= Patrologiae cursus completus: Series graeca]. Edited by J.-P. Migne. 162 vols. Paris, 1857–1886 Patristic Greek Lexicon. Edited by G. W. H. Lampe. Oxford, 1968 Pseudepigrapha Veteris Testamenti Graece Revue Biblique Revue des sciences religieuses Revue de l’histoire des religions Römische Quartalschrift für christliche Altertumskunde und Kirchengeschichte Recherches de science religieuse Revue de théologie et de philosophie Society of Biblical Literature Symposium Series Society of Biblical Literature Texts and Translations

Abbreviations SC SCJud SIG SNTSMS STDJ StPatr SVigChr SVTP TDNT

ThH TLZ TQ TRE TU VC WSt WUNT ZKT ZNW ZWT

Sources chrétiennes. Paris: Cerf, 1943– Studies in Christianity and Judaism; Études sur la christianisme et judaisme Sylloge Inscriptionum Graecarum. Edited by W. Dittenberger. 4 vols. 3rd ed. Leipzig, 1915–24. Reprint [4th ed.], Hildesheim, 1960 Society for New Testament Studies Monograph Series Studies on the Texts of the Desert of Judah Studia patristica Supplements to Vigiliae christianae Studia in Veteris Testamenti pseudepigraphica Theological Dictionary of the New Testament. Edited by G. Kittel and G. Friedrich. Translated by G. W. Bromiley. 10 vols. Grand Rapids, 1964– 1976 Théologie historique Theologische Literaturzeitung Theologische Quartalschrift Theologische Realenzyklopädie. Edited by G. Krause and G. Müller. Berlin, 1977– Texte und Untersuchungen Vigiliae christianae Wiener Studien Wissenschaftliche Untersuchungen zum Neuen Testament Zeitschrift für katholische Theologie Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der älteren Kirche Zeitschrift für wissenschaftliche Theologie

Biblical Texts and Versions NA27 UBS4 NAB NEB NRSV

XI

Novum Testamentum Graece, Nestle-Aland, 27th ed. The Greek New Testament, United Bible Societies, 4th ed. New American Bible New English Bible New Revised Standard Version

 

Introduction The Teaching (Didache) of the Twelve Apostles was (re-)discovered in 1873 among the contents of the eleventh-century codex Hierosolymitanus 54 (H54), a manuscript that included other works of the Apostolic Fathers – specifically the Epistle of Barnabas, 1, 2 Clement, and the longer corpus of the Ignatian epistles – as well as the Synopsis Veteris et Novi Testamenti (attributed to Chrysostom) and a list of Hebrew/Aramaic titles of the books of the Old Testament with their Greek equivalents entitled simply Δ∆Onovmata tw'n biblivwn parΔ∆ ÔEbraivoi".1 Today the Didache is accepted by virtually all scholars as a text of the latter half of the first or the early decades of the second century. While not canonical, its importance and authority in the early Church can be seen by attestations to the Teaching(s) of the Apostles, (Didach;[ai;] tw'n ajpostovlwn; Doctrina[e] apostolorum) in the writings of some of the early Fathers, texts clearly related to that of H54.2 Such an early date and stature by themselves would make the Didache an important witness alongside the New Testament to the development of the early Church, but the additional fact that the text is of a more utilitarian nature means that it does not merely supplement the biblical texts, but complements them. In its instruction on baptism and Eucharist, its statements on the acceptance and/or support of apostles, prophets, teachers, overseers/bishops and deacons, its struggle with the Christian ideal of unconditional giving versus the reality of charlatans, the Didache shows another side to the nascent Christian community, to its self-understanding, and to the problems of forming a community of believers in the (post-) apostolic period. Indeed, the last two decades has shown considerable interest in the Didache. Noteworthy among several new commentaries is one of nearly 1000 pages by Aaron Milavec that appeared in 2003;3 this study joined revised

1 In addition, after the colophon, the scribe has filled the rest of the page with a portion of what J. RENDEL HARRIS described as “the patristic explanation of the divergent genealogies in Matthew and Luke” (Three Pages of the Bryennios Manuscript [Baltimore: Johns Hopkins University, 1885], p. 1 of “Explanation of the Photographs”). 2 On this relationship, see below, 105–11. 3 AARON MILAVEC, The Didache: Faith, Hope, & Life of the Earliest Christian Communities, 50–70 C.E. (New York: Newman Press, 2003).

2

Introduction

editions of commentaries by Willy Rordorf/André Tuilier (1998)4 and by Kurt Niederwimmer (1993), the latter also translated into English for the Hermeneia series (1998).5 Collections of essays have also appeared: one presenting new studies (1995, ed. Clayton N. Jefford);6 one gathering together (and translating into English) many important studies of previous years (1996, ed. Jonathan A. Draper);7 and two presenting papers from related international conferences on the Didache and Matthew (2003, ed. Huub van de Sandt)8 and on the Didache, Matthew, and James (2007, eds. Huub van de Sandt and Jürgen K. Zangenberg).9 Of course numerous books, articles, and dissertations centering on the Didache or at least featuring it in a prominent role have also appeared. The modern reader, unfortunately, has been inhibited from a complete understanding of this early work. On the one hand, the preservation of the Didache in writing and the attestation of that text in the early Church means, de facto, that it was a text that communicated desired information to its early audiences. The fact that there is no introduction to the overall purpose of the text and no acknowledgment of its precise author or audience, means, however, that there must have existed underlying presuppositions and concepts held in common by the original author and audience(s) that at one point enabled clear communication but that are no longer present within the modern context. On the most basic level, the author and audience must have shared a common understanding of the genre of the text. Yet scholars have not yet determined what genre the Didache represents: some anachronistically call it the first “church order” while others simply consider it to be a text of “mixed genres”; a few see a clue in the term didache itself, though many dismiss the titles to the text of H54 as later additions. On the other hand, the problems with the Didache are not simply that modern scholars have lost the original, shared presuppositions of its communication. The text exhibits contradictions and abrupt changes in subject, style, and audience, that have left its compositional integrity in question, and schol4 WILLY RORDORF and ANDRÉ TUILIER, La Doctrine des douze apôtres (Didachè) (2d ed.; SC 248; Paris: Cerf, 1998). 5 KURT NIEDERWIMMER, The Didache (ed. Harold W. Attridge; trans. Linda M. Maloney; Hermeneia; Minneapolis: Fortress, 1998); trans. of Die Didache (2d ed.; Kommentar zu den Apostolischen Vätern 1; Göttingen: Vandenhoeck & Ruprecht, 1993). 6 CLAYTON N. JEFFORD, The Didache in Context: Essays on Its Text, History, & Transmission (Leiden: Brill, 1995). 7 JONATHAN A. DRAPER, ed., The Didache in Modern Research (Leiden: Brill, 1996). 8 HUUB VAN DE SANDT, ed., Matthew and the Didache: Two Documents from the Same Jewish-Christian Milieu? (Assen, Netherlands: Royal Van Gorcum; Minneapolis: Fortress, 2005). 9 HUUB VAN DE SANDT and JÜRGEN K. ZANGENBERG, eds., Matthew, James, and Didache: Three Related Documents in Their Jewish and Christian Settings (SBLSymS 45; Atlanta: Society of Biblical Literature, 2008).

Introduction

3

ars still vigorously debate whether there is more than one redactional stage represented. Some important recent studies have come full circle to the original belief of the very first students of the Didache, namely that the text is the work of one author, albeit one who has combined source material with original work. Yet such studies have not satisfactorily explained the presence of contradictions and the overall lack of homogeneity, both of which would be unexpected if the text were the product of a single author or editor. Thus the challenge of the Didache is simultaneously to understand it in the form in which it has been preserved, as a text valued and utilized by the early Christian community, and to uncover what historical developments in that community might be represented by the internal problems of the text. Toward this end, the following study attempts to recover both the generic identity and composition history of the Didache by means of a text-linguistic method first applied to early Christian works by David Hellholm in his study of the Shepherd of Hermas. This specific method focuses on the surface structure of a text with the premise that this structure is generically specific, i.e., that it either fulfills the common expectations of an explicitly stated genre or communicates the genre to the audience indirectly. Such a method is able to treat the Didache from a synchronic view, that is, as a text that was comprehensible to an audience in its present form, but, in its delimitation of the surface structure, has the advantage of being able to point to, or make more explicit, disruptions in that surface structure, signs that may indicate that there has been diachronic development.10

10

A note on the translations provided in this study: for the biblical texts cited I have decided to use a widely accepted translation and have chosen that of the New Revised Standard Version (NRSV). Translations of other texts are either my own or are taken from others as noted.

 

Chapter 1

The Didache and the Question of Genre 1.1 Introduction The Didach; tw'n dwvdeka ajpostovlwn, the Teaching of the Twelve Apostles, is generally acknowledged to be one of the earliest of the extra-canonical writings of the Church and is considered to be a part of the corpus known as the Apostolic Fathers, i.e., it is a text of the sub-apostolic age. A relatively recent discovery (1873)1, its publication (1883)2 caused great excitement among scholars studying early Christianity because it appeared to provide new information about the beginnings of the Church and, simultaneously, new support for various doctrinal positions debated among Protestant denominations, particularly in the U.S. Indeed the Didache seemed to have originated at some point during the first two centuries of the Christian era but, in comparison with other texts of the period, it had a much more practical character. The text is essentially a collection of prescriptions for the life of the Christian community. Structurally it is divided in the tradition into sixteen chapters.3 The first six chapters form a pre-baptismal catechism comprised primarily of a Two Ways text, i.e., a tractate on personal ethics based on the image of a choice

1 The discovery of the Didache was first announced by PHILOTHEOS BRYENNIOS, then the Metropolitan of Serres (Serrae), in his publication, Tou' ejn aJgivoi" Patro;" hjmw'n Klhvmento" ejpiskovpou ÔRwvmh" aiJ duvo pro;" Korinqivou" ejpistolaivv (Constantinople, 1875), h', as being one of the other texts in the Jerusalem manuscript (Hierosolymitanus 54). 2 BRYENNIOS, Didach; tw'n dwdevka ajpostovlwn (Constantinople, 1883). 3 In H54 the text is presently divided into chapters by means of marginal notation. On the one hand, BRYENNIOS had stated that a division into chapters or paragraphs was completely unknown in the manuscript and that the transcription was scriptio continua. (Didachv, le'). Yet in the margins of the Didache (as well as other texts in the manuscript) there are numbers given in Greek letters which correspond to the chapter divisions as given in BRYENNIOS (Didachv). Whether these notations occurred after the discovery of the manuscript by BRYENNIOS is unclear, but the earliest studies of the Didache attributed the chapter delimitation to him (see ADOLF VON HARNACK, Die Lehre der zwölf Apostel nebst Untersuchungen zur ältesten Geschichte der Kirchenverfassung und des Kirchenrechts [TU 2/1–2; Leipzig: Hinrichs, 1884], 2; PHILIP SCHAFF, The Oldest Church Manual Called the Teaching of the Twelve Apostles [2d rev. ed.; New York: Funk & Wagnalls, 1886], 161). The generally accepted versification, however, was the work of HARNACK.

6

Chapter 1: The Didache and the Question of Genre

between the ways of good and evil.4 Chs. 7–10 provide methods and rules for baptism, fasting, prayer, and some type of communal meal. These are followed in chs. 11–15 by instructions concerning itinerant and resident ministers to the community and certain aspects of intra-community relationships. Finally, ch. 16 is an eschatological passage predicting the events of the endtime using apocalyptic imagery.5 Given such topics, it is clear that the Didache must be consulted when reconstructing the earliest history of the Christian Church. Unfortunately the details contained in the Didache are greatly obscured by its literary style. The smallest units of material are, for the most part, understandable, but their relationship to each other and the overall portrait presented of the community are blurred and uncertain and, in some places, even contradictory. Much of the information is clouded by the elliptical nature of the text; apparently certain background knowledge on the part of the ancient reader was assumed, knowledge lost, however, to the modern audience with the result that the text appears laconic. In addition, the larger sections of the Didache often follow one another with little or no transition between them. Finally, the situation is further complicated by an apparent hesitancy to discard outdated material and the inclination instead to edit it or add to it, often with little attempt at overall harmonization. Thus, for example, community meals are described, but their exact function is unclear; traditional church officers – apostles, prophets, teachers, overseers/bishops and deacons – are mentioned, but their roles are not well defined and seem to overlap; perfection is important, though human fallibility is accepted. Such tendencies continued on a larger scale in the early Church with the appropriation of all or parts of the Didache into later texts of this type. 4 That the Two Ways is an independent text is seen by its attestation in works such as the Doctrina apostolorum, the Epistle of Barnabas and the Apostolic Church Order. The function of the Two Ways in these texts, however, is uncertain. See below, pp. 76–78. 5 Though sometimes referred to as the “Little Apocalypse,” Did. 16 lacks the narrative framework and the figure of the heavenly revealer necessary for a formal classification as an apocalypse. It does, however, exhibit certain characteristics of apocalypses, e.g., eschatological crisis in the form of persecution and various social manifestations of an increase of evil in the world, including the appearance of a “world-deceiver.” The processes of eschatological judgment and salvation are also described. The imagery used exemplifies the emphasis in later Jewish as well as Christian apocalypticism on the individual and interpersonal aspects of the eschaton, rather than the cosmic and/or political. While personal eschatology is an important component of virtually all apocalyptic literature, some later Jewish texts, e.g., Testament of Abraham and Apocalypse of Zephaniah, both likely to be products of the first century C.E., describe the rewards or punishment of the individual in the afterlife but lack any reference to a broader transformation of the earth or its political entities such as is common in many earlier Jewish apocalypses, e.g., Dan 7–12 and 1 En. 83–90 (“Animal Apocalypse”) and 93 (“Apocalypse of Weeks”). See JOHN COLLINS, “Introduction: Towards the Morphology of a Genre,” and “The Jewish Apocalypses,” Semeia 14 (1979): 13–18; 22–27.

1.1 Introduction

7

One of the most vexing problems is that the modern reader lacks an understanding of the context of the Didache. Who wrote the text, and for whom? What authority did it presume and how much was it accorded? Why were the topics selected? What information did it presuppose? Certainly an important aspect of the context of a literary work is its generic identity; indeed the recognition of the genre, whether conscious or subconscious, is essential to the comprehensibility of a text. On the one hand there are similarities in terms of form, content, and/or function between the Didache and other earlier or contemporary texts of the Judeo-Christian milieu, in particular writings among the Dead Sea Scrolls such as The Rule of the Community (1QS), the Rule of the Congregation (1QSa) and the Damascus Document (CD); the first-century Sermon on the Mount (Matt 5–7);6 and perhaps other “teachings” alluded to in the New Testament.7 In addition, certain characteristics of the Didache also resemble several ancient, non-JudeoChristian religious texts, for example, the ordinances for a religious association of Philadelphia8 and the regulations for the performance of the Andanian mysteries,9 both dating from the first century B.C.E., as well as the by-laws of an Athenian Society of Iobacchi of the second century C.E.10 Moreover, in the Graeco-Roman period, when rhetoric was a prominent discipline, the correct composition of texts was especially important and consequently the question must be raised as to whether the Didache was composed in accordance with a recognized generic structure, or whether its writing was much less deliberate and more the product of innate, human organizational tendencies.

6 See HANS DIETER BETZ, “A Jewish-Christian Cultic Didache in Matt. 6:1–18: Reflections and Questions on the Problem of the Historical Jesus,” in Essays on the Sermon on the Mount (trans. L. L. Welborn; Philadelphia: Fortress, 1985), 55–69; trans. of “Eine judenchristliche Kult-Didache in Matthäus 6, 1–18,” in Jesus Christus in Historie und Theologie, Neutestamentliche Festschrift für Hans Conzelmann zum 60. Geburtstag (ed. Georg Strecker; Tübingen: J. C. B. Mohr [Paul Siebeck], 1975), 445–57. 7 E.g., 1 Thess 4:1; Rom 6:17; 1 Tim 1:10; Titus 1:9. 8 SIG 985. For a discussion of this inscription that includes the Greek text and an English translation, see S. C. BARTON and G. H. R. HORSLEY, “A Hellenistic Cult Group and the New Testament Churches,” JAC 24 (1981), 7–41. 9 SIG 736. A discussion, selected bibliography, and English translation can be found in MARVIN W. MEYER, The Ancient Mysteries: A Sourcebook (San Francisco: Harper & Row, 1987), 51–59. 10 SIG 1109 (= IG 2/3.1.1368). An English translation, commentary, and bibliography appear in F. W. DANKER, Benefactor: Epigraphic Study of a Graeco-Roman and New Testament Semantic Field (St. Louis: Clayton Publishing House, 1982), 156–66.

8

Chapter 1: The Didache and the Question of Genre

1.2 Status Quaestionis A review of the research from the last fifty years shows that a specific genre for the Didache has yet to be determined. Given the topics addressed by the text, many have called it the earliest example of a church order, i.e., a manual regulating various aspects of Christian community life, particularly worship and rituals on the one hand, and church offices and administration on the other. Texts of this type began to appear especially in the third and fourth centuries C.E., for example, the Apostolic Tradition of Hippolytus,11 the Didascalia Apostolorum, the Apostolic Church Order, and the Apostolic Constitutions. In 1975 Philipp Vielhauer stated the modern consensus: Über die literarische Gattung der Did[ache] herrscht im übrigen Einigkeit: die Did[ache] ist eine Kirchenordnung, and zwar die älteste ihrer Art.12

Vielhauer believed this classification to be most apparent in the middle sections of the text, i.e., the liturgical instruction of chs. 7–10 and the teaching concerning community organization and life in chs. 11–15. Yet, though not quite as obvious, the Two Ways section, in his opinion, was also appropriate to a church order since parenetic material was so abundant in other documents of this type. The fact that the Didache appears to be composed out of other texts of various kinds and origins was also, according to Vielhauer, characteristic of its genre, as could be seen in its Christian counterparts as well as in the analogous texts from among the Dead Sea Scrolls.13 Following Vielhauer, Klaus Wengst wrote in 1984, Die blockweise Zusammenstellung von Vorschriften für verschiedene Lebensbereiche der Christen und der Gemeinde erweist die Didache der Form nach als eine Kirchenordnung.14

Clearly, however, these observations on the genre of the Didache have been greatly influenced by a pre-conceived notion of the characteristics appropriate to a church order and a certain retrojection of the traits of later texts onto the Didache is evident. Yet there have also been other voices in the contemporary discussion. In the introduction to his 1965 commentary on the Didache and the Epistle of Barnabas, Robert Kraft classified both texts under the heading “evolved lit11 In their 2002 commentary, however, PAUL F. BRADSHAW, MAXWELL E. JOHNSON, and L. EDWARD PHILLIPS situate the core of this community text as early as the mid-second century C.E. (The Apostolic Tradition [ed. Harold W. Attridge; Hermeneia; Minneapolis: Augsburg Fortress, 2002], 14). 12 PHILIPP VIELHAUER, Geschichte der urchristlichen Literatur (rev. ed.; Berlin: de Gruyter, 1978), 725. 13 Ibid., 725–26. See below, pp. 161, 166–67 14 KLAUS WENGST, Didache (Apostellehre), Barnabasbrief, Zweiter Klemensbrief, Schrift an Diognet (Schriften des Urchristentums; Munich: Kösel, 1984), 17–18.

1.2 Status Quaestionis

9

erature,” by which he meant that they “show clear evidence of being products of a developing process.”15 Within that broader category Kraft further identified the Didache as a “community manual” based on the fact that it “transmits community instructions for proper conduct and worship.”16 In 1978, Willy Rordorf and André Tuilier openly rejected the idea that the Didache was the first church order, reserving that designation for the later, pseudepigraphic collections such as the Didascalia and the Constitutions.17 Describing the Didache as a catechetical, liturgical, and disciplinary manual, unique with respect to genre, they rejected an overall literary unity to the text and identified four distinct parts, each having its own particular structure and presupposing a different origin.18 While they accepted the unity of the Two Ways section and said that it had a parenetic character that corresponded “tres exactement” to the literary genre didache in primitive Christianity, the other sections showed a lack of literary unity within.19 A more significant study was published in 1980 by Hermann-Adolf Stempel. In it Stempel explored what he believed to be the central role of the teacher in the Didache and called for a “differenzierte Beurteilung der literarischen Gattung der Didache,” describing the work as being “nach dem Selbstverständnis des Verfassers ein Handbuch zur christlichen Unterweisung in Fragen der lebensentscheidenden Lehre, der Liturgie und der Gemeindeordnung.”20 While he agreed with the idea that the Didache contains the oldest church order, at least from ch. 7 on, he maintained that the Sitz im Leben of the text was the instruction of the community, not the establishment of church law.21 In his opinion, the titles, the didactic terminology and references throughout, and the didactic methodology used in the composition of the text, show not only the work as a whole to be a “teaching” (Lehre) but also some

15 ROBERT KRAFT, Barnabas and the Didache (vol. 3 of The Apostolic Fathers: A New Translation; ed. Robert M. Grant; New York: Thomas Nelson & Sons, 1965), 1. 16 KRAFT contrasts this with Barnabas, which, he maintains, differs from the Didache both in form (Barnabas is an epistle) and content (its instruction deals with “ ‘school’ interests,” e.g., exegesis and commentary) (Barnabas and the Didache, 3). 17 RORDORF and TUILIER, La Doctrine, 21, n. 2. 18 Ibid., 11, 17–18, 99. 19 On this see La Doctrine, 17–18. For the meaning of didache as a genre the authors cite in general two works by C. H. DODD (The Apostolic Preaching and Its Developments [Chicago: Willett, Clark & Company, 1937] and Gospel and Law [New York: Columbia University Press, 1951]; see below, p. 49, n. 247; 144–45) as well as the TDNT article by K. H. RENGSTORF (“didavskw, didavskalo", ktl.,” TDNT 2:135–65; see below, 142–54). In the opinion of RORDORF and TUILIER, Didach; tw'n ajpostovlwn was likely the early, though still not original, title of the Two Ways (La Doctrine, 15–16). 20 HERMANN-ADOLF STEMPEL, “Der Lehrer in der Lehre der zwölf Apostel,” VC 34 (1980): 215. 21 Ibid.

10

Chapter 1: The Didache and the Question of Genre

of its individual component sections.22 Here Stempel correctly acknowledged the importance of the term didachv for the text although he did not discuss the concept of didache as an independent, recognizable form or genre: didache is simply whatever is “taught” by the teacher/author of the Didache. While he pointed to some of the structural indicators in the text, he provided neither an overall generic view of its composition nor explanations of its internal inconsistencies and rough transitions with respect to genre, primarily because of his presupposition that the Didache is a single editorial redaction of pre-existent materials by an authoritative teacher within the community. Similarly in 1980 in a study on the ministerial roles represented in the Didache, André de Halleux also supported the original unity of the text and claimed, almost in passing, that the text was, indeed, a didachv, as evidenced by its (later) titles.23 This is only briefly mentioned, however, amidst his attempt to distinguish the tradition-oriented, non-charismatic teacher-author of the Didache from the (sometimes apostolic) teacher-prophet, i.e., a (single) type of charismatic figure whom he believed to be the subject of Did. 10.7–11.12; 13; 15.1–2;24 he did not discuss further the question of genre. From another angle, others have maintained that the designation church order is not really appropriate for the Didache (nor perhaps for other texts in this group) because the work is not comprehensive in scope, either in terms of its range of topics or within each topic itself. The problem first came into focus in 1986 with an article by Georg Schöllgen in which he challenged Vielhauer and the virtually universal classification of the Didache as the earliest church order. According to Schöllgen, such a classification seems correct at first glance since the topics discussed in the text – catechesis, liturgy, and church offices – are typical of early church orders. Yet this appraisal is, in his opinion, erroneous, and leads to certain false presuppositions about the Didache, resulting in the misunderstanding and misuse of the text as a complete guide to the life of the early Christian community it represented. In Schöllgen’s opinion,

22 Ibid., 210–12; 216, n. 10. Here STEMPEL rightly mentions the sub-text defined by Did. 1.3; 2.1; and 6.1, each of which uses the term didachv. More questionable, however, is his assertion that the sections introduced by peri; dev in 7.1; 9.1; and 11.3 should also be considered individual units of teaching (“Lehrinhalte”) – in these instances the term didachv is not present. One should note, moreover, that the last occurrence of the noun didachv comes in Did. 11.2, a datum that may have significance for the text’s redaction history (see below, p. 123). 23 ANDRÉ DE HALLEUX, “Ministers in the Didache,” in The Didache in Modern Research (ed. Jonathan A. Draper; Leiden: Brill, 1996), 319; trans. by J. A. Draper of “Les ministères dans la Didachè,” Irén 53 (1980): 5–29. 24 Ibid., 306–7; 314–15; 319–20. Similarly, MILAVEC, Didache: Faith, Hope, & Life, 438– 41.

1.2 Status Quaestionis

11

If one follows the understanding of the Didache as a comprehensive church order, then the text – as the only source from the post apostolic period – is in the position to give an untruncated picture of the whole external life of an early Christian community. In other words, the criterion of comprehensiveness allows one not only to utilize the positive provisions which the text gives its addressees, but also allows the interpreters to go beyond this to the rare opportunity of a methodologically legitimate use of the otherwise rightly rejected argument from silence: if the Didache is a comprehensive church order, then one can conclude – at least in the areas with which it deals more closely (eg. catechesis, liturgy and church offices) – that whatever it does not deal with was not practiced at the time of its composition. The reverse conclusion from the argument from silence raises the value of the text as a source considerably.25

Schöllgen instead described the Didache as a “selective church order,” i.e., a text written to address specific controversies within the community and to provide direction for the handling of new circumstances that had come about.26 Neither the Didache, nor the other early church orders, were intended to be comprehensive manuals.27 In a later study he further concluded that the Didache, Didascalia, and Apostolic Tradition all considered “Scripture,” i.e., some form of the biblical writings, to be the real church order for their communities.28 The Didache, for example, commands its audience to perform all things wJ" e[cete ejn tw'/ eujaggelivw/ tou' kurivou hJmw'n (as you have [it] in the gospel of our Lord, 15.4). New situations and problems had arisen within the churches, however, and it was deemed necessary that some supra-community rule address these specific issues. On the one hand, according to Schöllgen, the Didache and Didascalia often present their instruction with the support of 25

GEORG SCHÖLLGEN, “The Didache as a Church Order: An Examination of the Purpose for the Composition of the Didache and Its Consequences for Interpretation,” in The Didache in Modern Research (ed. Jonathan A. Draper; Leiden: Brill, 1996), 44; trans. by J. A. Draper of “Die Didache als Kirchenordnung. Zur Frage des Abfassungszweckes und seinen Konsequenzen für die Interpretation,” JAC 29 (1986): 5–26. 26 Ibid., 63. In a more recent study, “Der Abfassungszweck der frühchristlichen Kirchenordnungen: Anmerkungen zu den Thesen Bruno Steimers,” JAC 40 (1997), 61, n. 30, SCHÖLLGEN mentions the use of the alternative term Gemeindeordnung, but finds Kirchenordnung still useful because ejkklhsiva itself can refer to an individual Christian community as well as to the Church at large. 27 SCHÖLLGEN, Zwölf-Apostel-Lehre, Apostolische Überlieferung (FC 1; Freiburg: Herder, 1991), 14–15. SCHÖLLGEN believes that this also explains the fragmentary presentation of the Didache. Since the text is intended to deal only with individual problems, nothing more than a general schema is necessary. Such an understanding of the purpose of the Didache presents as well an alternative to the widespread tendency to dissect the text into smaller, sometimes minuscule, units, attributing each to a different redactional stage. While he does not deny that certain sections of the text were likely pre-existent units, he believes there is no proof of more than one editor at work and one compositional stage. See his “Didache as a Church Order,” 64–67. 28 “Die Didache will die Probleme des Gemeindelebens durch adaptierende Auslegung des Evangeliums lösen,” SCHÖLLGEN, “Pseudapostolizität und Schriftgebrauch in den ersten Kirchenordnungen,” JAC.E 23 (1996), 119.

12

Chapter 1: The Didache and the Question of Genre

Scripture; the Apostolic Tradition, on the other hand, relies more on an orally transmitted tradition of the apostles that is understood as existing alongside Scripture.29 All three additionally support their teaching by claiming apostolic authority. Ultimately, in Schöllgen’s estimation, this claim was a dead end: once the rulings were written down, it was increasingly difficult to update them without casting suspicion on their origins and validity. Thus the church orders were finally replaced by the canons of the early Christian synods.30 Schöllgen’s warnings about the danger of viewing the Didache as a comprehensive manual are well taken. Certainly one must always be wary when making an argument e silentio. Nevertheless his premise that the Didache is “selective” is as much an unverifiable presupposition as the general belief that the text is the first, presumably complete, church order. Both views begin from a pre-conception of church order derived from later orders to which the Didache is compared.31 What the Didache lacks is then pronounced either not (yet) existent by the one group, or pre-supposed and non-controversial by Schöllgen. In his articles, Schöllgen himself seems to have in mind what a comprehensive church order should contain, but rather than conclude that unmentioned topics are non-existent, he believes them to be unimportant; in both cases speculation is involved. His own presupposition is evidenced in several places. Regarding the directions concerning the baptismal water (Did. 7.1–3), he concluded that the purpose of the passage was to address the situation that the preferred form of baptism could not be performed in some communities and that concessions had to be made, in particular for those lacking in water resources. He noted that certain “expected” topics are not represented: The reader learns nothing about the actual rite of baptism, about the baptismal confession, about the person of the baptizer, about the conditions of admission of candidates (age, occupation etc.) or about the baptismal day.32

While admitting that it cannot be assumed that all of these matters were already established in a detailed way in the era of the Didache, he nevertheless believes that, aside from the triadic baptismal formula, the issues which are found in the text are of a secondary importance. Similarly, with respect to the comments on fasting in the text, Schöllgen stated,

29

Ibid., 107–109; 119–20. Ibid., 120–21. See below, pp. 31–32; 43–46. 32 SCHÖLLGEN, “Didache as a Church Order,” 47. 30 31

1.2 Status Quaestionis

13

Again only one aspect of the practice of fasting in the ancient church is regulated. The reader learns nothing e.g. about the length, intensity or even the degree of compulsion of the stationary fast, the extent of its applicability to people (with exceptions for sick, aged and children) and the Paschal fast. All this is assumed as known.33

But are these issues “assumed” or non-existent? The contemporary Jewish literature extols the virtue of the individual who fasts, but it is a voluntary act with only general rules prohibiting fasting on the Sabbath and other days of religious observance.34 Similarly Schöllgen commented on the fact that, among other things, the duties of the various church ministries are not given and the components of the worship service are not stated in the Didache.35 Ultimately he tries to distinguish between a complete church order, with detailed instructions presuming no or only minimal previous knowledge on the part of the reader, and a selective church order, based on certain presuppositions, but this explanation seems anachronistic.36 In 1989 Kurt Niederwimmer took the discussion of the genre of the Didache one step further. Niederwimmer denied that the Didache exhibited a specific genre and stated that, “The Didachist possessed no direct literary model from which to construct this work.”37 He maintained that the work was not easily classified and described the Didache as “a rule for ecclesiastical praxis, a handbook of church morals, ritual, and discipline,”38 that the author composed using various sources from a variety of genres.39 The creation of such a text, as well as its form, came about from the innate character of a cultic community, presumably from the need to perform certain rites precisely. He maintained that the author of the Didache combined a pre-existent literary foundation with other instructions thought necessary for such a rulebook. Once the plan of producing a book of rules had been conceived, the very nature of the enterprise suggested that instructions for baptism and the Eucharist, as well as questions regarding order within the community, should be added.40

Niederwimmer reserved the term church order for chs. 11–15. He speculated that the need for the Didache perhaps came about as a result of a specific situation in the community, but that the nature of the circumstances could not 33

Ibid., 49. E.g., Jdt 8:6–7; T. Sim. 3.4; T. Jud. 15.4; T. Jos. 3; Luke 2:37. 35 SCHÖLLGEN, “Didache as a Church Order,” 59–60. 36 Recently DRAPER has agreed with SCHÖLLGEN that the Didache was “not intended to provide a universal ‘Church Order’.” Stating that there are no examples of the genre church order earlier than the third century, he prefers to describe the text as a community rule, formally comparable to 1QS (“Ritual Process and Ritual Symbol in Didache 7–10,” VC 54 [2000]: 123 and n. 6). See also below, pp. 24–25; 30. 37 NIEDERWIMMER, Didache, 2. 38 Ibid. 39 See also RORDORF/TUILIER, La Doctrine, 17–18. 40 NIEDERWIMMER, Didache, 3. 34

14

Chapter 1: The Didache and the Question of Genre

be derived from the text itself.41 Niederwimmer rightly brought out the practical nature of the Didache and, like Schöllgen, correctly warned against trying to derive information beyond what is stated explicitly. Nonetheless, in his description of chs. 11–15 as a church order, it is clear that he too began from a preconception of what that means. His discussion also raises the question of how genres are “born.” In a 1992 study entitled, Vertex Traditionis: Die Gattung der altchristlichen Kirchenordnungen, Bruno Steimer met the challenge head-on. In this work, Steimer attempted to define the genre of the early Christian church order starting from the texts generally classified as such from the first or second through the fifth centuries. Observing that this designation had usually been based on the contents of the texts,42 Steimer hoped to define the genre more precisely using a form-critical approach. Acknowledging that such an endeavor requires an analysis to be performed on literarily independent works, he focused his efforts on the Didache, the Apostolic Tradition and the Didascalia (the “Basistexte”) from among these writings.43 Steimer actually began his investigation with the Pastoral Epistles (specifically 1 Timothy and Titus) because of the suggestions of some scholars that these works either attest an underlying church order or, alternatively, represent the creation of a church order based on the Haustafeln.44 Ultimately, however, he concluded that neither of these theories was convincing. On the one hand, the disparity of the church order material among the Pastorals was too great and the instruction too general to support the existence of a shared Vorlage; on the other hand, there was insufficient commonality between the Pastorals and other examples of Haustafeln to show a material connection between those texts.45 What really characterized the Pastorals in Steimer’s opinion, and what gave them the sense of “church order,” was a predominance of parenesis and parenesis-related forms (Gliedgattungen) directed toward various aspects of church order and organization and the support of this instruction through the use of theological images, through references to norms already recognized by the community, and through the authority of the author. The purpose of the Pastorals, he concluded, had to do largely (though not 41

NIEDERWIMMER agrees with SCHÖLLGEN that the Didache is “selective” (Didache, 3, n. 15) and similarly rejects a broad reconstruction of the ideas of the author either from what is mentioned or implied in the text (2, n. 8). 42 BRUNO STEIMER, Vertex Traditionis: Die Gattung der altchristlichen Kirchenordnungen (BZNW 63; Berlin: de Gruyter, 1992), 152. 43 The observation that church orders contained within them a number of other genres (“Gliedgattungen”) led STEIMER to classify church order, as it is generally applied to these texts, as a “Rahmengattung,” though it could still also designate a “Gliedgattung,” i.e., a component part of a larger work (Vertex, 156–57). 44 Ibid., 160–61; 190. 45 Ibid., 171.

1.2 Status Quaestionis

15

solely) with a strengthening of church structure in the face of heresy from outside.46 Given these findings, Steimer then turned to his three base texts to investigate form critically the role and character of their parenesis and, in particular, the relationship between formally parenetic (appellative) and juridical (normative/prescriptive) elements within them.47 With respect to the Didache, Steimer agreed with the consensus that sees the text as falling into four parts: the Two Ways teaching (chs. 1–6), the liturgical instructions (chs. 7–10), the church order “im engeren Sinn” (chs. 11–15),48 and the eschatological conclusion (ch. 16). He noted that the first three of these were redactionally joined to one another via transitional phraseology49 and that the last, although syntactically independent, was semantically, clearly, final in nature.50 Transitional phrases could also function as rubrics within the text to introduce new themes on various levels.51 Materials within sections were loosely held together, primarily through word/concept associations.52 Addressing the specific questions concerning parenesis and parenetic vs. juridical forms, Steimer pointed out that all of chs. 7–15 (i.e., both parts 2 and 3) shared a largely prescriptive diction, made up of both apodictic and casuistic forms. In contrast the sections before and after these chapters, i.e., the Two Ways in chs. 1–6 and the eschatological conclusion in ch. 16, contained primarily appellative material that gave a parenetic character not only to those individual sections but, because of their position, to the text overall.53 Yet a certain mixing of the two styles could also be seen, in particular in the Two Ways section.54 A comparison of the results from the study of each of the base texts led Steimer to the conclusion that church orders are comprised of different genres (“Gliedgattungen”) and that they present their material in both appellative and prescriptive styles, i.e., both parenesis (in various “Formen und Gattungen” [“Gattung Paränese”]), and law (“Gattung Recht”).55 The material within sections also exhibited great variation in its forms of address (for example, the

46

Ibid., 175–76; 189. Ibid., 190. 48 Ibid., 203. Here as a “Gliedgattung.” 49 Ibid., 210. 50 Ibid., 194–95. 51 Ibid., 201. 52 Ibid., 210. 53 Ibid. 54 Ibid., 197–98. 55 Ibid., 242–44. On the one hand, the appellative material is essentially found in two “Grundformen,” catalogs and admonition, while the prescriptive texts can be either casuistic or apodictic. “Gattung Recht” is defined as several “Rechtssätze in einem Text aufeinander folgen” (242–45). 47

16

Chapter 1: The Didache and the Question of Genre

variation between singular and plural forms in the Didache).56 Moreover, alongside these shared form-critical characteristics, church orders also exhibited shared interests in cult, discipline, and church organization.57 An element of authority – often, but not always, provided through pseudepigraphy – was also present to support a claim to universality.58 Though, in a nod to Schöllgen, Steimer agreed that these texts are not “complete,” he nonetheless considered the texts to be “comprehensive” in the sense that they addressed, at least minimally, all the important aspects of church life.59 Inappropriately characterized, according to Steimer, as “Streitschriften” by Schöllgen, they instead were concerned with creating a stronger unity and uniformity for Christian communities.60 Schöllgen has taken Steimer to task on several points. In his opinion, a major problem lies in Steimer’s differentiation between “completeness” and “comprehensiveness”: Schöllgen disputes both its linguistic justification and the conclusions to which it leads. In his mind, the minimal overlap of subject matter between the texts in the corpus and their individual concentration on specific issues support his belief that the texts were intended to address problem areas within their own communities and speak against Steimer’s view that the authors are attempting to fix in writing practices already commonly used.61 Moreover, Schöllgen believes that Steimer’s claim of comprehensiveness for the church orders facilitates the very type of conclusions about which he has warned. He cites as an example Steimer’s conclusion from the minimal information found in the Didache that bishops and deacons had no difference in status, that they had little importance overall in the community, and that the members of the community made decisions for the church collectively.62 Schöllgen also disagrees with Steimer and others who maintain that the titles of the Didache are secondary: if the titles, and thus the reference to 56

Ibid., 245. Ibid., 247. 58 Ibid., 268. According to STEIMER, the absence of pseudepigraphy in the Didache (since he rejects the originality of the titles – see below pp. 18; 114; 187, n.146) means that these practices were already largely accepted and that they were put into writing by someone who was already considered authoritative by the community and who was, in that sense, one with the community (321, 339). This did not preclude for him the addition of new instruction within the texts as well (279–80). 59 Ibid., 269–72. 60 Ibid., 316–17. Regarding the Didache, STEIMER cites the possibility that the community might not have a residential prophet (implied in Did. 13.4) as indicating that a plurality of churches is being addressed. As further support for this idea he cites the predominance of the plural form of address beginning in Did. 7 and specifically its occurrence in the instruction to choose bishops and deacons (261–62), though these latter arguments are not as strong since the address could conceivably be to multiple persons within a single community. 61 SCHÖLLGEN, “Abfassungszweck,” 59–60. 62 Ibid., 72. 57

1.2 Status Quaestionis

17

the apostles, are not original, what provided the text with its necessary authority?63 To these criticisms of Schöllgen one can add that, while Steimer has indeed provided useful form-critical analyses of the four major text-parts of the Didache, he does not deal adequately with how these parts work together to make the Didache a cohesive whole. The way in which the parts are syntactically connected to one another is briefly recounted64 but the question as to whether each of them is generically essential to a church order and the relationship of these parts to one another are not explored. Is there, for example, a genre-specific order and/or level of importance to their presentation in the text? The fact of the matter is that Steimer views the Didache as a “Kompilat,” based on what he perceives to be the virtual integrity of the individual sections,65 and he consistently favors the view that this compilation is the work of one author. Too often, however, this perspective leads him to propose not entirely convincing explanations for semantic and syntactic disruptions in the text. For example, while Steimer cites the idea that the abrupt changes in address between singular and plural in Did. 1.3; 5.2u; 7.2–4; and 13.3, 5–7 should be attributed to the use of sources by the author/editor, a widely held opinion, he also suggests the possibility that such a phenomenon (outside of the Two Ways teaching) might have served the purpose of highlighting the instruction for the audience or of inspiring a new hearing.66 Moreover the fact that overseers/bishops and deacons make their appearance only near the end of the Didache is attributed by Steimer to the fact that the section on itinerant ministers (part 3) is the appropriate place for bishops and deacons to be mentioned.67 In addition to this is the problem that Steimer does not explore the 63 Ibid., 70–71; also Zwölf-Apostel-Lehre, 26. SCHÖLLGEN is followed by GUNNAR GARLEFF who has concluded that there is nothing to prevent one or both of the titles from being considered original. He sees them as providing a “Kontinuitätssignal” between the ethical teaching of the Two Ways and Jesus and the early community (Urchristliche Identität in Matthaüsevangelium, Didache, und Jakobusbrief [Beiträge zum Verstehen der Bibel 9; Münster: LIT Verlag, 2004], 111–17). Earlier, STANISLAS DOCKX had suggested that the first title of the Didache, without the reference to “twelve,” was tied to the Jewish Two Ways text while the original version of the longer title, “Didach; tw'n ajpostovlwn toi'" e[qnesin,” was attached to the earliest Christian source of the Didache, a text found in 7.1–10.7; 14.1–15.2 (“Date et origine de la Doctrine des apôtres aux Gentils [Did. 7,1–10,7; 14,1–15,2],” in Chronologies néotestamentaires et Vie de l’Eglise primitive [2d ed.; Leuven: Peeters, 1984], 364). 64 STEIMER, Vertex, 193–95. 65 Ibid., 193. For example, the phrase peri; de, correctly seen by STEIMER as an important structural marker in the text, is interpreted by him as underscoring the compilatory character of the text (201) which in Greek is not necessarily the case. See below pp. 135–36. 66 Ibid., 298, n. 217. 67 STEIMER believes that the inclusion of the instruction regarding bishops and deacons, residential offices of the community, is appropriate in the section dealing with itinerant vs. residential members and following the reference to the Eucharist in ch. 14 (Vertex, 19).

18

Chapter 1: The Didache and the Question of Genre

question as to why the Two Ways begins without introduction and only later is found to be intended for use in baptism,68 or why some teachings from the Jesus tradition are specifically legitimated by reference to “the gospel” or to “the Lord” while most stand alone without support.69 To virtually ignore the possibilities of intrusions into/adjustments of the text at a later time runs the risk of missing clues to the generic character and development of the Didache and related writings. Moreover, his form-critical analysis/es do not yield a satisfactory generic description of church order. For example, while he acknowledges that the third section of the Didache, chs. 11–15, is often characterized as a church order, he concluded that there was nothing from a formcritical perspective that distinguished it from the previous section. In actuality, what distinguished the two was their content. Thus he concluded that the identification of chs. 11–15 as a church order was “nicht verifizierbar.”70 When all is said and done, while Steimer does provide a thorough inventory of the shared characteristics of the ancient texts included under the heading church order, one still would not know exactly how to go about creating one. Moreover, Steimer did not explore the question of the ancient genre of these texts, i.e., while the commonalities among the texts were a logical starting point, he did not address the problem of anachronistically applying to them a later generic term.71 Moreover, because Steimer dismisses the authenticity of the titles to the Didache, he misses the importance of the term didachv as it is used throughout the earliest of those works. Steimer made progress, but still did not resolve the problem. More recently, a 2007 study by Joseph Mueller brought the findings of current research on the relationship between the patristic church order literature and the Jewish Mishnah/Talmud to bear on the question of genre. Mueller contends that the attempt to determine a literary genre for the “church orders” has been problematic, in particular because it requires “putting in the same genre texts compiled between the first century (Didache) and the sixth century of our era (Testamentum Domini) when the life situations of their authors and of their authors’ churches varied so widely.”72 Attempts such as those of Alexandre Faivre73 and Marcel Metzger74 that identify the various shared 68

STEIMER remarks only that the Two Ways text could not have been recited verbatim at baptism (as seems to be indicated by proeipovnte" in Did. 7.1) because the beginning of the text lacks any indication that such was the case (Vertex, 198, n. 40) 69 Compare Did. 8.2; 9.5; 11.3; and 15.3 with the synoptic-like material of 1.3b–5 and ch. 16. See discussion below, pp. 168–73. 70 STEIMER, Vertex, 208. 71 See below pp. 20, 31. 72 JOSEPH G. MUELLER, “The Ancient Church Order Literature: Genre or Tradition?” JECS 15 (2007): 338; 378. 73 See, for example, ALEXANDRE FAIVRE, “La documentation canonico-liturgique de l’Église ancienne,” RevScRel 54 (1980): 204–19; 273–97. There he describes such literature

1.2 Status Quaestionis

19

characteristics among the texts of this corpus in terms of content, as well as that of Steimer, which distinguishes between the formal elements of law and parenesis found within them,75 simply highlight, in Mueller’s opinion, the problem that the topics presented and the forms utilized in the “church orders”

as containing “à la fois des prescriptions d’ordre morale et disciplinaire et des prescriptions ou des descriptions liturgiques. En outre, tous ces documents sont placés directement ou indirectement sous le patronage des apôtres” (287). In fact, he sees the contents of these “church orders” as a more important characteristic than their literary structure, although he does not discount the value of structural features. For example, FAIVRE does not include the Didache among “canonico-liturgical” texts because, although the types of materials (“genres”) it contains match those of the later ancient orders, they are presented by the Didachist as still very distinct from one another. Moreover he sees the Didache as lacking the focus on church office found among the later texts; what instruction on church office does exist comes only after the moral and liturgical materials, a stark contrast, in his opinion, to the general tendency of canonico-liturgical works to direct their instruction first to those of highest status in the community and afterwards to others in an order of decreasing importance (288–89). 74 The development of METZGER’s thought on the genre question is discussed by MUELLER in L’Ancien Testament dans l’ecclésiologie des Pères. Une lecture des Constitutions apostoliques (Instrumenta patristica et mediaevalia 41; Turnhout: Brepols, 2004), 58– 60, and in “Ancient Church Order Literature,” 344–46. METZGER’s earliest position, found in his edition of, and commentary on, the Constitutions, gave three generic characteristics of church orders: 1) the presence of a variety of instructions regarding ecclesiastical institutions; 2) the compilation of earlier texts and traditions; and 3) a pseudepigraphic style (Les Constitutions apostoliques [3 vols.; SC 320, 329, 336; Paris: Cerf, 1985–1987], 1:13). According to METZGER, the Constitutions is the “apogee” of this type of literature (1:49). In a later publication METZGER listed five characteristics: 1) the directing of the text toward Christian communities and their leadership; 2) the treatment of a variety of aspects of ecclesiastical life; 3) the compilation and juxtaposition of earlier traditions; 4) pseudepigraphy; and 5) direct literary links and thus the presence of common materials among the various ancient church orders (“Presentation d’ouvrages: Les Constitutions apostoliques,” RevScRel 62 (1988): 308–9). As MUELLER notes, by 1992 METZGER had pared the list down to two: 1) the presence of regulatory materials relevant to church life; and 2) the claim of apostolicity, in many cases in the form of pseudepigraphy (METZGER, “A propos des règlements ecclésiastiques et de la prétendue Tradition apostolique,” RevScRel 66 (1992): 250–51). In MUELLER’s estimation, however, “this definition could apply … to all conciliar legislation, which is not usually considered to be church order literature” (“Ancient Church Order Literature,” 346). 75 MUELLER, L’Ancien Testament, 61–63. Against the criticism of SCHÖLLGEN, however, MUELLER supports STEIMER’s claim of comprehensiveness for these ancient church order texts. Beyond STEIMER’s idea that all necessary parts of community life are touched upon within them, a view challenged by SCHÖLLGEN, MUELLER maintains that one must also consider STEIMER’s points about the sense of universal validity (and the intentional openendedness) found in much of the material and the fact that the orders provide their readers with outside authorities to cover anything lacking in the order itself (L’Ancien Testament, 64– 67; on the Didache see STEIMER, Vertex, 265–66; 270–72). MUELLER is especially interested in references to the Scripture and in the fact that the early orders, when taken over by later works, were respected and rendered largely intact (L’Ancien Testament, 66–67).

20

Chapter 1: The Didache and the Question of Genre

can be found in what are clearly other genres.76 Thus he concludes that such research has demonstrated “the dubious usefulness of constructing for these ancient texts a literary genre defined by characteristics of content, form, or style.”77 Moreover, citing Eva M. Synek’s work on the Apostolic Constitutions, he brings up the important point that the label church order itself was only first applied to ancient texts by scholars of the modern period because of their similarity to the post-Reformation Kirchenordnungen,78 a point I also raised independently in my dissertation on the Didache. He rightly maintains that such an identification is inadequate, commenting, giving the name Kirchenordnung to the patristic texts opens the possibility of misconceptions about their genre. A new label for these texts could avoid this difficulty.79

Thus, drawing on the work of Charlotte Elisheva Fonrobert on the Didascalia,80 that of Synek on the Constitutions, as well as his own work also on 76 MUELLER, “Ancient Church Order Literature,” 345–48. “Descriptions that boil down to the content areas of the prescriptions in this literature have in the end shown themselves unable to account for the specificity of its genre” (346). 77 Ibid., 337. 78 Ibid., 340–41, referring to EVA M. SYNEK, “Dieses Gesetz ist gut, heilig, es zwingt nicht …”: Zum Gesetzesbegriff der Apostolischen Konstitutionen (KuR 21; Vienna: Plochl, 1997), 12–18. Unaware of SYNEK’s work on the Constitutions, I also explored this issue in my dissertation, “The Genre of the Didache: A Text-Linguistic Analysis” (Ph.D. diss., University of Chicago, 2002), 20–22. See below pp. 31–32; 39–46. 79 MUELLER, “Ancient Church Order Literature,” 343. (As MUELLER points out, STEIMER, in the article “Kirchenordnungen,” LTK3 6:33–38 [1997], also acknowledges SYNEK on this point [“Ancient Church Order Literature,” 341, n. 7]). Like MUELLER, SYNEK too warns about drawing implications for the genre of these ancient texts from the post-Reformation orders stating, “Jedenfalls ist im Auge zu behalten, daß bei der Verwendung des Terminus ‘Kirchenordnung’ zur Bezeichnung einer patristischen Quellengruppe ein spezifisch neuzeitlich geprägter Begriff herangezogen wurde” (“Dieses Gesetz ist gut,” 16). Yet both SYNEK and MUELLER take the time to detail explicit differences between the Reformation texts and the ancient orders. In SYNEK’s opinion, the primary differences between the two are the emphasis on continuity with the past and the derivation of authority from the laws of the Bible and the apostles on the part of the patristic orders (“Dieses Gesetz ist gut,” 18). MUELLER, on the other hand, sees both the ability of the Reformation communities to adapt their church orders as well as their localization of authority in the geo-political leader as important generic differences with the patristic texts (“Ancient Church Order Literature,” 341–43). In the end, SYNEK sees the transference of the Reformation-era generic classification to the ancient texts as “plausibel” (“Dieses Gesetz ist gut,” 17) in that both corpora address similar topics and justify new instruction by recourse to previously accepted tradition. Interestingly, although both scholars note the importance of the terms didachv and didaskaliva for the ancient texts (see below, pp. 23–24), they consider this to be indicative only of a common tradition of Torah exegesis rather than of a specific literary genre and proceed no further along that path of inquiry. 80 CHARLOTTE ELISHEVA FONROBERT, “The Didascalia apostolorum: A Mishnah for the Disciples of Jesus,” JECS 9 (2001): 483–509.

1.2 Status Quaestionis

21

the Constitutions, Mueller “brackets” the discussion of genre81 and instead proposes viewing the ancient orders as “representatives of an exegetical tradition that extracts doctrine on church life from the Old Testament.”82 Fonrobert’s study, strictly speaking, does not deal with the issue of genre but looks, rather, to the Didascalia as an example of the fluidity that still existed between Christianity and Judaism in the third and fourth centuries in Syria.83 Within this study, she cites the fact that the Didascalia recognizes and uses the Jewish Scriptures as authoritative, that it applies rabbinic methods of interpretation, and that it seemingly participates with the Mishnah in a common religious milieu, as evidence that the Didascalia shares a “discursive space” with rabbinic and other forms of Judaism.84 Its appearance at approximately the same time as the codification of the Mishnah in written form suggests to her that the Didascalia is a reaction to that development and that it functions both as a “counter-Mishnah” in its warnings against the “second legislation” of the Torah (i.e., the stipulations given after the incident with the Golden Calf in Exod 32)85 and as a “Christian ‘Mishnah’ ” in providing “its own set of rules, secondary to and based on scriptural law.”86 Synek’s study, a portion of which deals explicitly with genre, explores the concept of law in the Constitutions. She presents what she believes to be parallels between the Constitutions and the Talmud and, more broadly, between the patristic church order literature and the rabbinic interpretation of Torah. She notes that the Constitutions and the Talmud, in addition to being roughly contemporaneous texts, share a material foundation in the Torah as well as formal parallels, specifically a compilatory character, a guarantee of authority through reference to a (fictional) meeting of primary figures (the meeting of the apostles in Jerusalem and the “synods” of Jamnia and Usha respectively), and a support of instruction, at times secondarily, via explicit references to/ excerpts from the biblical writings.87 In Judaism, while the Law per se was the 81

MUELLER, “Ancient Church Order Literature,” 377. Ibid., 337. MUELLER acknowledges that SCHÖLLGEN reached a similar conclusion about the Didache and the Didascalia (ibid., 360, n. 67). It should be noted, however, that SCHÖLLGEN included under “Scripture” the use of New Testament tradition as well by these texts (see above, pp. 11–12). 83 FONROBERT, “Didascalia apostolorum,” 508. 84 Ibid., 486, quoting PETER SCHÄFER, ed., The Talmud Yerushalmi and Graeco-Roman Culture (Tübingen: Mohr Siebeck, 1998), 14. Whereas the Didascalia has traditionally been identified as “Jewish-Christian,” FONROBERT sees that as a modern characterization that obscures the Jewish voices, both authorial and opposing, that are found within (“Didascalia apostolorum,” 483). 85 Ibid., 496. 86 Ibid., 498. 87 SYNEK, “Dieses Gesetz ist gut,” 8–10; 80–81. SYNEK also points out that, while many of the supporting citations in the Constitutions are carried over from its source texts, a significant number come from the final compiler as well (see esp. 26–28 with notes; 33–34). 82

22

Chapter 1: The Didache and the Question of Genre

Torah given through Moses, Synek states that the interpretation (Auslegung), the updating (Aktualisierung), of the written Torah in the Prophets, the Writings, and the Mishnah also became “Torah” in that it too became authoritative.88 In a similar fashion, the Constitutions afforded Jesus’ interpretation of the Law authoritative status as well.89 The passing on and interpretation of Torah (in this broader sense) in the traditions of the apostles, and the subsequent acknowledgment of that body of tradition as authoritative, are, according to Synek, also parallel to the way in which the teachings of the Mishnah both interpreted the Law and at the same time were considered to be part of the Law that derived from the time of Moses.90 Finally, just as the traditions of the apostles found in the Didascalia, Didache, and other source texts used by the Constitutions have been collected and further commented upon to create yet another authoritative actualization of Torah, (in the Constitutions’ own estimation, as part of the canon!91) so too the Talmud contains the authoritative commentary upon the Mishnah.92 Thus, more than simply a component of the Constitutions, this emphasis on the interpretation of Torah (again, in this broader definition) is, in her mind, foundational and absolutely necessary to any discussion of the genre of the ancient church order texts,93 a discussion that heretofore has been inconclusive.94 While the initial comparison with the Reformation texts was understandable, Synek suggests that it is the concept of law as the interpretation or updating of Torah for the Christian community that truly connects the Constitutions with both the rabbinic tradition and the other patristic texts traditionally recognized as church orders. Regarding the Constitutions she writes,

88

Ibid., 82. Ibid. 90 Ibid., 9. 91 Ibid., 83. 92 Ibid., 80, n. 303. 93 Ibid., 77. 94 Ibid., 10–12; cf. 77. SYNEK presents as an example the contrast between NORBERT BROX’s rather “tautologisch” (in her opinion) description of the early church order texts as “die ältesten erhaltenen Aufzeichnungen von Richtlinien fur Recht, Verfassung, Disziplin, Liturgie und Sitte in der Kirche,” (BROX, “Altkirchliche Formen des Anspruchs auf apostolische Kirchenverfassung,” Kairós 12 [1970]: 117) and STEIMER’s problematic “Reduction” to “Kult, Disziplin und Verfassung” (Vertex, 246–47). 89

1.2 Status Quaestionis

23

vom Selbstverständnis des Kompilators ist sein Versuch, unter Verwendung von kirchenordnungsbezogenem Traditionsgut aktuelle Probleme zu regeln, Toraauslegung im Namen der Apostel. Tora (im Sinn der schriftlichen Tora) wie Toraauslegung (respektive aktualisierende Torafortschreibung) zielte im Sinn eines Kontinuität zwischen “altem” and “neuem” Israel bekräftigenden Kirchenbegriffes immer schon auf “Kirchenordnung.” Die Genusbestimmung “Kirchenordnung” kann somit bekräftigt werden. Als Toraauslegung sind die CA “Kirchenordnung.” Nicht zu übersehen ist aber, … daß sie als Kirchenordnung als Toraauslegung verstanden werden wollen. Insofern unterscheiden sie sich also nicht wesentlich von der im Talmud zusammengefasste[n] rabbinische[n] Halacha.95

Thus Synek would retain the genre designation church order but with the understanding that the Constitutions aims at establishing church order via Torah interpretation. Synek finds additional support for a connection between the Constitutions and the rabbinic tradition in the semantic relationship between the selfdesignations of the Greek church order texts, didache and didascalia (teaching), and Hebrew talmud (teaching) from lamad/limmed (learn/teach).96 The remark of Stemberger that “Talmud” can “die gesamte traditionelle ‘Lehre’ bedeuten, v.a. auch die durch Auslegung von M[ischna] gewonnene Lehre der Amoräer, die Bibel und M[ischna] gegenübergestellt wird”97 suggests to her that the Constitutions might be better understood as a “christlichen Talmud.”98 In Synek’s view, the compiler of the Constitutions took over from the author of the Didascalia the concept of “holy teaching,” “im matthäischen Sinn als Auslegung respective Aktualisierung der Tora durch die Apostel.”99 She finds such a concept also in the Didache, likely originating, in her opinion, in the Two Ways tradition, and suggests that the fact that many identify the Two Ways as a proselyte catechism points to the idea that it “etwas mit Toraauslegung zu tun haben.”100 Here Synek adds her voice to those of Rordorf/Tuilier, Stempel, and Halleux mentioned above, all of whom correctly regard such terminology as important for an understanding of the Didache. Her connection of such terminology explicitly with the Two Ways section, however, glosses over the fact that such a characterization of “teaching” as Torah interpretation/updating is too narrow to fit the contents of the Didache overall.101

95

“Dieses Gesetz ist gut,” 83–84. This connection was also noted by J. RENDEL HARRIS in 1887 (see below, p. 41). 97 Dieses Gesetz ist gut, 79, quoting GÜNTER STEMBERGER, Einleitung in Talmud und Midrasch (8th ed.; Munich: C. H. Beck, 1992), 167–68. (Eng. ed., STEMBERGER [H. L. STRACK] Introduction to the Talmud and Midrash [2d ed.; trans. and ed. Markus Bockmuehl; Minneapolis: Fortress, 1996], 165.) 98 “Dieses Gesetz ist gut,” 79. 99 Ibid., 77–78. 100 Ibid., 78, n. 298. 101 The connection of this terminology to the Didache and the other church orders will be more thoroughly discussed below, see pp. 141–55; 186–88. 96

24

Chapter 1: The Didache and the Question of Genre

Mueller, however, agrees with Synek that these terms signal the existence of a common link between the church order literature and the Mishnah/ Talmud. The redactors of the church orders … tended to use a single concept to sum up their compositional objective when speaking of their respective works as a whole. They thought of exhortation and legal prescription as coming under the single heading of doctrine, didachv or didaskaliva, or that of a tradition that is taught or passed on. … We need, then, a different conception of what the redactors intended to do in composing the ancient texts often called church orders, a better account of what they meant by this traditional doctrine. … These three studies [Fonrobert, Synek, Mueller] all point away from the Reformation Kirchenordnungen and toward one or another sacred and authorized work of Old Testament interpretation as the prime analogate for characterizing the texts usually grouped under the label “ancient church orders.” These three studies emphasize that the redactors of these two early church works [Didascalia, Constitutions] intended to function as interpreters of the Old Testament, especially of the law of Moses.102

Specifically Mueller believes that the independent base texts, the Didache, Apostolic Tradition, and Didascalia, “present themselves as doctrine or didascalia, in the sense of a mix of halakhah and aggadah that founds community order on an interpretation of the Old Testament.”103 In addition, the church order texts share with their Jewish counterparts Jewish hermeneutical methods and a claim to an ancient origin.104 Looking specifically at the Didache, Mueller notes the presence of halakhic material in the Two Ways teaching of chs. 1–6, in the instruction on baptism, fasting, and prayer in chs. 7–8, in the discussion surrounding the support of resident ministers in ch. 13, and in the instruction concerning weekly community meetings in ch. 14.105 He finds haggadic material in the ritual prayers of Did. 9–10 as well as in the eschatological teaching of ch. 16.106 He sees Deut 6, the Golden Rule, and the Deca– logue as “essential parts of the bedrock for the prescriptions of the whole Didache,” although the Old Testament must also be interpreted in the light of Jesus and the Messianic Age.107 Structurally, Mueller, citing Draper, notes that the use of peri; de to begin a new topic in the Didache is reflected in the

102 103

MUELLER, “Ancient Church Order Literature,” 348–49. Ibid., 360; see also 365. 104 Ibid., 349–50. 105 Ibid, 356–57. The Old Testament connection is clear in Did. 13, which deals with the community’s obligation to support resident ministers using the term ajparchv (firstfruits), and in ch. 14, where Mal 1:11, 14 is cited. 106 Ibid., 357. MUELLER states that Did. 16 “contains a fragmentary aggadic discourse” but this is questionable. Indeed, what Old Testament passage(s) would he cite as its overall basis? 107 Ibid., 361.

1.2 Status Quaestionis

25

Dead Sea Scrolls and rabbinic writings.108 He sees also a “hint” of typology in the reference to the “vine of David” (Did. 9.2) and in the equation of spiritual food and drink with the physical food and drink provided at creation (Did. 10.3), both found in the eucharistic prayers.109 Finally he maintains that the tracing of its origins back to the very beginning of Christianity via the use of the Jesus tradition and other early materials by the Didache parallels the references to the Great Synod or to the meetings of Jamnia and Uscha seen in the rabbinic materials.110 In Mueller’s opinion, it is thus not a common genre, but the mutual founding of community life on an interpretation of the Old Testament and the sharing of these interpretive and literary methods that connect the Didache with the later, ancient orders as well as with the Mishnah/Talmud.111 Regarding the ancient orders he writes, the exegetical doctrine of community life in these texts represents more a tradition of ecclesiological Old Testament exegesis than a literary genre that has eluded efforts to define it through characteristics of content, form, style, or resemblance with Reformation-era Kirchenordnungen.112

Such a broad definition, however, casts a much wider net, as Mueller himself admits. Indeed, in his opinion, an ancient text can be placed under this heading if it demonstrates just two characteristics: (1) it must be given over in large part to transmitting a self-consciously apostolic tradition of doctrine composed of a mix of halakah and aggadah that founds Christian community order on the interpretation of the Old Testament; (2) it must transmit that doctrine with the aid of exegetical techniques shared with Judaism.113

On this basis Mueller would include texts such as 1 Clement, the De sacramentis of Ambrose, and the Epistula apostolorum with the church order texts.114 It is obviously not genre, he maintains, that connects all of these texts, but rather it is a shared tradition of ecclesiological exegesis of the Old Testament that gives these texts a sense of continuity, despite the broad range of time and circumstances in which they were produced. Mueller writes,

108

Ibid., 357, citing the reference to an analagous usage of the preposition l[ in JONADRAPER, “Weber, Theissen, and ‘Wandering Charismatics’ in the Didache,” JECS 6 (1998): 563, n. 109. 109 MUELLER, “Ancient Church Order Literature,” 361. 110 Ibid., 363–64. 111 Ibid., 353, 355–56, 365. 112 Ibid., 365. 113 Ibid., 370. 114 Ibid., 370–73. THAN

26

Chapter 1: The Didache and the Question of Genre

We need only claim continuity in the evolution of some fundamental elements of an identitypreserving community organization: the techniques of interpretation, the halakhic and aggadic approach to the Old Testament as foundation for community order, and the apostolic reference. This study has tried to demonstrate this continuity instead of a problematic continuity of literary genre.115

Mueller’s study makes an important contribution in showing where the attempts at locating a genre for the church order literature have failed and why. That being said, however, one must ask whether the characterization of these texts as “ecclesiological exegesis of the Old Testament,” is helpful, or even accurate. With respect to the Didache, it is, of course, universally acknowledged that the text reflects a Jewish background of which the Hebrew Scriptures play a foundational role. But can the material presented by the Didache really be described as an exegesis of the Scriptures and is that even the best description of the Mishnah, to which it is being compared? Granted that the Mishnah assumes the written Mosaic Law, it nonetheless contains much that developed independently of it. According to Herbert Danby, although the peculiar literary methods of the Mishnah sometimes give an impression of deducing an ideal and theoretical practice out of academic interpretation and application of Scripture, it may often be that the reverse is the truth: what was sought was Scriptural grounds on which to justify long-established usage not expressly ordained or permitted by Scripture.116

Furthermore, Stemberger, in his discussion of various modern views regarding the relationship of the oral halakhot to the Bible states, The simple evidence of the M[ishnah] text itself shows that there are essentially three groups of halakhot: 1) those derived from Scripture; 2) halakhot independent of Scripture; 3) halakhot which arose independently of Scripture but were later connected with it.117

Regarding the third category, H. L. Strack himself noted that the lack of scriptural support for certain halakot in the Oral Tradition was problematic for the rabbis and that such support in the Mishnah was in many cases secondary.118 In addition, when a biblical reference is provided in the Mishnah, the function is generally one of support for an instruction, not its source.119 While Synek cites such a use of Scripture as yet another formal parallel between the Consti-

115

Ibid., 377–78. HERBERT DANBY, The Mishnah (Oxford: Oxford University Press, 1933), xv. 117 STEMBERGER, Introduction to the Talmud, 128. 118 HERMANN L. STRACK, Introduction to the Talmud and Midrash (Jewish Publication Society of America,1931; repr., New York: Atheneum, 1969), 10. 119 STEMBERGER, Introduction to the Talmud, 127–28, citing both J. N. EPSTEIN, Introduction to Tannaitic Literature: Mishna, Tosephta and Halakhic Midrashim (Hebr.) (ed. E. A. Melamed; Jerusalem: 1957), 503ff.; and CH. ALBECK, Einführung in die Mischna (Berlin and New York, 1971), 56–93. 116

1.2 Status Quaestionis

27

tutions and Jewish halakah,120 the problem is that this is no longer Torah interpretation or exegesis – the Constitutions may support instruction using the Hebrew Scriptures, but they do not derive their instruction from it. This is already present in the Didascalia where an entire series of Old Testament texts can be given to reinforce a teaching, an indicator that the purpose is supportive rather than exegetical.121 The purpose of these two texts is thus not to explain the Torah but to give instruction for the life of a community, backing up this instruction, when possible, using the Old Testament, a text considered authoritative by its members. With the Didache the difference is even more obvious for here the Old Testament is cited only a handful of times and always, only, in a supportive role.122 Thus, while both the Didache and the Mishnah are instructional in nature and share a foundation in Judaism, in neither does the overall characterization as “exegesis” really fit. Moreover there are also important differences in content, form, and purpose between the Mishnah and the Didache and this is where the determination of genre is vital. On the most basic level, the two differ in their presuppositions: while the Mishnah supplements and complements the Torah and assumes a familiarity with the Scriptures on the part of its audience, the Didache assumes nothing of the sort. This important distinction is illustrated by the very first instruction of the Mishnah, m. Ber 1:1, which treats of various situations having to do with the evening recitation of the Shema. Here neither the words of the Shema nor a tradition regarding its regular recitation are provided – these are assumed to be fully known to the reader. Rather it is the case that specific questions regarding the evening recitation are being addressed.123 In contrast, the Didache opens with a statement on the Two Ways (1.1) followed by a summary statement regarding the Way of Life that is comprised of the Double Love Commandment and the Golden Rule (1.2). It is this summary statement that is the subject of the following “didache” (1.3a), which, as has been argued, is a more detailed explanation of how to carry out the commands of 1.2.124 While the instructions of 1.2, a juxtaposition of Deut 6:4 and Lev 19:18b, have an obvious foundation in the Hebrew Scriptures, the exact form exhibited by the Double Love Commandment in the 120 121

SYNEK, “Dieses Gesetz ist gut,” 80–81. See, for example, Didasc. 1.6 (end)–7 where a warning to men regarding the temptations resulting from bathing with women is subsequently supported by Prov 7:1–27, followed immediately by Prov 5:1–14. Similarly in Didasc. 1.8 a comparable instruction to women is supported by Prov 31:10–31; 12:4. 122 Explicit, certain quotations of Old Testament material are introduced only in Did. 14.3 (Mal 1:11, 14) and 16.7 (Zech 14:5). What appears to be intended as a quotation from Scripture is also seen in 1.6, which is often traced to Sir 12:1, but this is uncertain; see the discussion in NIEDERWIMMER, Didache, 83–86. 123 On this general phenomenon see also STRACK, Introduction to the Talmud, 24. 124 As early as HARNACK (see below p. 42).

28

Chapter 1: The Didache and the Question of Genre

Didache is not found in the Torah or even in other Jewish texts.125 Moreover, although versions of the Golden Rule can be found in contemporary Jewish literature,126 Clayton Jefford has suggested that its inclusion adjacent to the Double Love Commandment in the Didache (albeit in a negative form) may be related to the characterization of each as equivalent to “the Law and the Prophets” found only in the Gospel of Matthew (7:12; 22:40).127 Thus Synek appears to be incorrect when she states concerning the Two Ways that, “Der gesamte Text, der über weite Teile eine Dekalogexegese ist, knüpft an Dtn 6,4 an.”128 Rather the Two Ways is an interpretation/expansion/explanation, of the first two statements positioned at the head of the Didache. Under the heading Way of Life, I would maintain, the Didachist provides not an exegesis or actualization of the Torah, but a version of the Torah for gentiles. Indeed the readers are provided with some of the most basic commands of the Decalogue: ouj foneuvsei", ou moiceuvsei", ouj klevyei" – these are not an exegesis of Torah, but laws of the Torah itself presented as specifics of the command found in Did. 1.2. While the Didache clearly presupposes the authority of the Old Testament and on a few occasions uses it to support its own instruction, the main focus, I would propose, is to present teaching under the authority of Jesus, something that is shown by the addition of Q-related material to the original Two Ways in 1.3b–6. Indeed, outside of the Two Ways the characterization of the instruction of the Didache as Old Testament exegesis is even more problematic. As alluded to above, Mueller places the blessings of Did. 9.2, 3–4 and 10.2–5 in the rather open category of haggadah, which he defines as any Old Testament interpretation that is not prescriptive in character. The instructions on baptism, fasting, and prayer (including the text of the Lord’s Prayer) in chs. 7–8, on the other hand, he deems halakhic in nature.129 Is it appropriate, however, to understand any of this as Old Testament exegesis? Two points speak against this characterization. On the one hand, while traditional Jewish imagery certainly appears in the eucharistic prayers, what is presented is not an interpretation of 125

See, for example, KRAFT, Barnabas and the Didache, 137–39; NIEDERWIMMER, Didache, 64–66. 126 See NIEDERWIMMER, Didache, 66, n. 14 for examples. 127 CLAYTON JEFFORD, The Sayings of Jesus in the Teaching of the Twelve Apostles, (SVigChr 11; Leiden: Brill, 1989), 31–38. On the idea that the Two Ways source of the Didache was already “Christian” see HELMUT KOESTER, Synoptische Überlieferung bei den Apostolischen Vätern (TU 65; Berlin: Akademie Verlag, 1957), 172; NIEDERWIMMER, Didache, 64–66; and DAVID HELLHOLM, “Från judisk tvåvägslära till kristen dopkatekes. En inblick i tillkomsten av en första kyrkoordning,” in Ad Acta: Studier til Apostlenes gjerninger og urkristendommens historie (ed. Reidar Hvalvik and Hans Kvalbein; Oslo: Verbum, 1995), 115– 16. 128 SYNEK, “Dieses Gesetz ist gut,” 78, n. 298. 129 MUELLER, “Ancient Church Order Literature,” 356.

1.2 Status Quaestionis

29

Scripture – surely not every statement containing Jewish imagery should be considered haggadic. Secondly, in Did. 8–10 it is the very basic prayers of Christian worship that are being set down for community recitation. While the commands associated with these prayers may be considered halakhic in character, it does not necessarily follow that they are exegetical, as was seen above in Stemberger’s description of the halakhot found in the Mishnah.130 Yet another difference between the Didache and the Mishnah is the fact that the latter text records varying opinions on particular issues. Such is never the case in the Didache and in the rare instance where “backpedaling” occurs, scholars often suspect a redactional hand. One example can be seen in the caveat of Did. 6.2: eij me;n ga;r duvnasai bastavsai o{lon to;n zugo;n tou' kurivou, tevleio" e{sh/: eij dΔ∆ ouj duvnasai, o} duvnh/, tou'to poivei. For if, on the one hand, you are able to bear the entire yoke of the Lord, you will be perfect; but if you are unable, do what you can.

Given the references to perfection found elsewhere in the Didache (1.4; 10.5; 16.2) and the fact that the Two Ways text is being used here as a prebaptismal catechesis (7.1), many believe this caveat to be either the redaction of source material by the Didachist or an addition to the text at an even later stage.131 In either case, it is a technique that differs significantly from that of the Mishnah. While the observations of Synek, Fonrobert, and Mueller regarding shared formal and material characteristics between the church orders and the Mishnah/Talmud certainly highlight the Jewish character of the source communities of the Didache and other church orders, they do not result in a better understanding of the form, content, and/or purpose of the Didache and thus are not a substitute for the information that would be gleaned from an accurate identification of its genre. Mueller himself admits that what many understand as a gentile catechetical function of the Didache distinguishes it from later church orders and states that, “we might find ourselves before a difference of literary genre among the texts commonly grouped under the category of church orders.”132 What Mueller does not consider is the possibility that the difference among the texts might reflect a path of generic development where changes and variations could offer important insight into the circumstances of the communities that produced each of the works. Finally, in other important, recent studies of the Didache, it is clear that a precise generic identity for the text has still not been discerned. Huub van de Sandt, for example, refers to the Didache as a “Christian manual” and “a 130 131

See above p. 26. For a discussion see NIEDERWIMMER, Didache, 120–23. 132 MUELLER, “Ancient Church Order Literature,” 377.

30

Chapter 1: The Didache and the Question of Genre

handbook of church morals, ritual and discipline,”133 while Jonathan Draper calls it a “community rule.”134 In his recent commentary, Aaron Milavec describes the Didache as a “pastoral manual,” “pastoral guide,” or “practical guide for spiritual mentors”135 that “represents the preserved oral tradition detailing the step-by-step training of gentile converts being prepared for full, active participation in the house churches committed to the Way of Life.”136 Although he considers the titles to be later additions to the text,137 he nonetheless believes that the term didache, which he narrowly translates as “the systematic training that a mentor (or a master craftsman) would give to an understudy (apprentice),” reflects well the content and purpose of the work.138 Unfortunately Milavec does not explore whether the text is in fact representative of a specific literary genre. Kari Syreeni, who addresses the genre question more directly, notes that one’s impression of the genre of the text tends to depend on whether one understands the text “progressively,” i.e., from beginning to end, a view that accentuates its catechetical character, or whether, taking a “centric” view, one sees the primary force of the text in the ritual and administrative materials of chs. 7–15, relegating what precedes and follows as a “framing” device. On the one hand Syreeni believes the Didache to be “a purposeful blend of genres” with both “catechetic and … church order elements” given equal weight, yet also refers to the text as “a paraenetic manual of instruction” and a “church order.”139 Also recently, William Varner has labeled the Didache a “Christian handbook” and described its “basic function” as a “catechetical handbook designed for those who were joining the Christian family from among the Gentiles.”140 The consensus is perhaps best illustrated by the comment of Bart Ehrman in the new Loeb edition of the 133 HUUB VAN DE SANDT and DAVID FLUSSER, The Didache: Its Jewish Sources and its Place in Early Judaism and Christianity (CRINT III.5; Assen, Netherlands: Royal Van Gorcum; Minneapolis: Fortress, 2002), xv. 134 See, for example, JONATHAN DRAPER, “Do the Didache and Mattthew Reflect an ‘Irrevocable Parting of the Ways’ with Judaism?” in Matthew and the Didache: Two Documents from the Same Jewish-Christian Milieu? (ed. Huub van de Sandt; Assen, Netherlands: Royal Van Gorcum; Minneapolis: Fortress, 2005), 217., and above pp. 13, n. 36; 24–25. 135 MILAVEC, Didache: Faith, Hope, & Life, e.g., vii, x, xi, 845. 136 Ibid., vii. 137 Ibid., vii; 55–58. 138 Ibid., ix; 71. MILAVEC is followed in this characterization of the text by THOMAS O’LOUGHLIN, The Didache: A Window on the Earliest Christians (London: Society for Promoting Christian Knowledge; Grand Rapids, Mich.: Baker Academic, 2010), 10–13. 139 KARI SYREENI, “The Sermon on the Mount and the Two Ways Teaching of the Didache,” in Matthew and the Didache: Two Documents from the Same Jewish-Christian Milieu? (ed. Huub van de Sandt; Assen, Netherlands: Royal Van Gorcum; Minneapolis: Fortress, 2005), 88, 98, 102. 140 WILLIAM VARNER, The Way of the Didache: The First Christian Handbook (Lanham, Md.: University Press of America, 2007), 3.

1.3 The Didache as Church Order – Early Research

31

Apostolic Fathers. There Ehrman states that, “the book as a whole is usually considered a ‘church manual’ or a ‘church order’, the first of its kind to survive.”141

1.3 The Didache as Church Order – Early Research 1.3.1 Origin of the Term Church Order and its Application to the Didache As was noted above, church order is a translation of the German Kirchenordnung, a term first attested in the sixteenth century,142 signifying a number of church/government-issued constitutions written as a result of the Reformation and the subsequent ecclesiastical movements that took the place of canonical law.143 The various changes had created widespread disorganization and a disturbing lack of uniformity among the churches. Since, however, the Church and the State governments were thoroughly inter-twined, the territories adopting the new Lutheranism had to reorganize both their ecclesiastical communities as well as the interrelationship of these with the secular political powers. These state constitutions typically included a definition of correct Church doctrine and the content of Christian catechism, and also instructions on liturgy, prayer, church offices, synods, church discipline, marriage and care for the poor, as well as the establishment of government-appointed ecclesiastical courts to deal with matters of heresy, etc., thereby establishing an official liaison between the state government and the Church. In some cases, revised editions of these constitutions were valid until the early nineteenth century. The first application of church order to the early Christian texts appears to have been made by Johann Wilhelm Bickell with his publication of the work he called the Apostolic Church Order, a version of the same Two Ways text found in the Didache, the Doctrina, and Barnabas.144 Bickell created the title 141 BART

EHRMAN, Apostolic Fathers (2 vols.; LCL 24, 25; Cambridge: Harvard University Press, 2003), 1.406. Note that EHRMAN also applies the term church order specifically to chs. 7–15 (ibid., 1.409). See as well JOSEPH VERHEYDEN, “Eschatology in the Didache and the Gospel of Matthew,” in Matthew and the Didache: Two Documents from the Same Jewish-Christian Milieu? (ed. Huub van de Sandt; Assen, Netherlands: Royal Van Gorcum; Minneapolis: Fortress, 2005), 214, who also acknowledges the Didache to be a “church order.” 142 JACOB and WILHELM GRIMM, “Kirchenordnung,” in Deutsches Wörterbuch, (1873) cite witnesses only as far back as that period; the New Schaff-Herzog Encyclopedia of Religious Knowledge (1909) knows the term only in reference to the sixteenth-century constitutions. 143 The earliest use of the term is perhaps in the Artikel der ceremonien und anderer kirchen ordnung, by Bishops Georg of Samland and Erhard of Pomesanien in 1525. See ANNELIESE SPRENGLER-RUPPENTHAL, “Kirchenordnungen, II. Evangelische,” TRE 18.676. 144 JOHANN WILHELM BICKELL, Geschichte des Kirchenrechts (1 vol., 2 fasc.; Giessen: Hener, 1843–1849), 1.1.107–32. STEIMER states that the concept (“Begriff”) was applied

32

Chapter 1: The Didache and the Question of Genre

Apostolische Kirchenordnung from what he considered to be the “fitting” title of the Ethiopic Ecclesiastical Canons, a text that begins with the Apostolic Church Order. He translated the Ethiopic title, Dies sind die Canones der Apostel, welche sie zur Ordnung der christlichen Kirche festgesetzt haben.145 The association was certainly understandable as the text dealt with many of the same issues as the constitutions of the sixteenth century, namely, doctrine, sacraments, ministerial offices, and church discipline, though he admitted that the initial moral teaching was more of an introduction to the actual church order.146 In terms of general content and function, however, the ancient and modern texts were very similar, and the genre church order was easily, though anachronistically, applied to the former. The term was later applied to the Didache soon after its publication. But in contemporary times, when the understanding of its genre has become an important tool in interpreting a text and gaining insight into the individuals and communities behind it, the question must be asked whether the application of the generic classification church order is appropriate for the Didache and other similar texts, or whether the use of the term creates improper expectations for these early writings while at the same time impeding our understanding of them by ignoring the issue of their actual, ancient generic identification/development. As has been shown, criticisms of scholars such as Schöllgen, Niederwimmer, and Mueller have brought into question not only the traditional understanding of the Didache as a church order but also the assumption that the work represents a specific literary genre at all. On the other hand, others have noted points of contact between the Didache and ancient notions of and terms for teaching, but have not pursued a more precise generic identification/description of the text based on these observations. All have failed to offer a satisfactory analysis of the Didache from the perspective of ancient literature. 1.3.2 Discussion of Genre Prior to the Discovery of the Jerusalem Manuscript Long before its discovery the Didache was known both through direct attestation in several of the Church Fathers and indirectly through the larger works in which it had been included. As was inevitable, the ideas and discussions surrounding these cognate works were subsequently transferred to the Didache upon the discovery of its text and, for better or worse, set the agenda for its study. In this early period of historical-critical methodology, however, the question of genre was only indirectly addressed. Indeed, the first mentions of genre involving the Didache are found in studies on the Constitutions, a beginning in the seventeenth century but does not cite specific references (“Kirchenordnungen,” 34). 145 Ibid., 1.1.83, n. 2. In reality the title of the Apostolic Church Order varies among the witnesses to the text. 146 Ibid., 1.1.89.

1.3 The Didache as Church Order – Early Research

33

text initially published in the sixteenth century and valued for its antiquity and possible apostolic origin and authority. In pursuit of such authority, some sought to identify the Constitutions (Diatagai; [Diatavxei"] tw'n aJgivwn ajpostovlwn dia; Klhvmento")147 with the Didach; [ai;] ajpostovlwn mentioned as authoritative in the Ecclesiastical History of Eusebius (325 C.E.), in the 39th Festal Letter of Athanasius (367 C.E.), in Against Dice-Throwing falsely attributed to Cyprian (300 C.E.), and in the Stichometry of Nicephorus (850 C.E.);148 and/or with the Diavtaxi" [Diavtaxei"] tw'n ajpostovlwn cited by Epiphanius alongside excerpts generally close to the Constitutions.149 The attempt to connect these texts was natural, given the similarity of the titles and the fact that the Constitutions presented itself as having originated with the apostles and other leaders of the early Church as transmitted through Clement of Rome (Const. ap. 6.18.11). William Whiston, for example, identified the Didach; [ai;] ajpostovlwn as the text found in two Arabic manuscripts, a text that seemed to him to contain extracts of the Constitutions. Since Whiston believed the Constitutions truly to have originated in the apostolic period, however, this did not preclude a relatively early date for the extracts (=Didache) as well. John Grabe, however, contested this identification, stating that the Arabic texts in question were not extracts of the Constitutions but essentially its first six books, though taken from a damaged exemplar.150 He 147

1.100.

For the attested variations of the titles see METZGER, Les Constitutions apostoliques,

148Eusebius, tw'n ajpostovlwn aiJ legovmenai Didacaiv, Hist. eccl. 3.25.4 (text in GCS 9.1.252); Athanasius, Didach; kaloumevnh tw'n ajjpostovlwn, Ep. fest. 39 (text in THEODOR ZAHN, Geschichte des neutestamentlichen Kanons [2 vols.; Erlangen, Germany: Deichert, 1890], 2.1.212); Ps.-Cyprian, in Doctrinis apostolorum, Adv. aleatores 4 (text in WILHELM S. HARTEL, Thasci Caecili Cypriani opera omnia [3 vols.; CSEL 3; Vienna: C. Geroldi filium, 1868–1871; repr., New York: Johnson, 1965], 3.3.96); Nicephorus, Didach; ajjpostovlwn, Stich. 68 (text in ZAHN, Geschichte des neutestamentlichen Kanons, 2.1.301). 149 Epiphanius, Pan. 45.4; 70.10, 11, 12; 75.6 (all with singular form); 80.7 (with plural form). These quotations were often identified as coming from the Constitutions until the publication of the Didascalia by PAUL DE LAGARDE in 1854 showed them to be closer in form to the latter text. Still the extant Didascalia does not always coincide perfectly with the text found in Epiphanius. Thus some today posit the existence of a middle recension of the Didascalia used by Epiphanius and in the Constitutions (see JAMES J. C. COX, “Prolegomena to a Study of the Dominical Logi as Cited in the Didascalia apostolorum, Part 1: Introductory Matters [cont.],” AUSS 13 (1975): 249–50, who follows JAMES VERNON BARTLET, ChurchLife and Church-Order [ed. Cecil John Cadoux; Oxford: Blackwell, 1943]). More recently, in the first edition of the first volume of his translation of the Panarion, FRANK WILLIAMS contended that Epiphanius knew both the Didascalia (at least in some form) and the Constitutions, but in the revised edition he appears to have limited Epiphanius’ knowledge to the Constitutions alone (The Panarion of Epiphanius of Salamis [2 vols.; vol. 1, 1st ed., NHS 35; vol. 1, 2d ed., NHS 63; vol. 2, NHS 36; Leiden: Brill, 1987–2009], compare page xx in the first edition of vol. 1 with page xxv of the second edition). 150 JOHN GRABE, An Essay upon Two Arabick Manuscripts (London, 1711), 20–23.

34

Chapter 1: The Didache and the Question of Genre

pointed out, moreover, that the Arabic text matched neither the catechumenal character of the Didache as described by Athanasius,151 nor the short length of the work (200 lines) as noted by Nicephorus. Grabe maintained that it was more likely that the Arabic manuscripts, one of which was entitled the Book of Dascalia [for Didascalia] which is the Doctrine delivered by the Fathers, the twelve Apostles, and Paul the Apostle, and James Brother of the Lord, Bishop of Jerusalem, consisting of thirty-nine Chapters, the other simply The Doctrine, (both according to his translation) should be identified with the Didaskaliva Klhvmento", a text listed in the seventh-century Catalogue of the 60 Canonical Books separately from the Didacai; tw'n ajpostovlwn.152 In Grabe’s view, this Didaskaliva Klhvmento" was an expanded version of the text entitled The Didascalia or Catholick Doctrine of the twelve Apostles and holy Disciples of our Savior (a Syriac version of the Didascalia Apostolorum), portions of which had been published by Abraham Ecchellensis in 1661. This expanded version represented an intermediate stage of the Constitutions prior to the addition of Books 7 and 8 and it was this that was the foundation of the Arabic texts.153 The Didache, in Grabe’s opinion, was to be identified instead with the Diavtaxi" cited by Epiphanius or, less likely, with the Syriac Didascalia of Ecchellensis.154 Yet the difference in genre nomenclature seen in the various titles was an issue and in 1672 John Pearson briefly discussed the phrases didacaiv, didaskalivai, and diatavxei" tw'n ajpostovlwn in relation to the Constitutions. Finding the terms to be common in the New Testament (the last as a verb) and occurring also in later, pseudonymous texts, he believed that didacaiv and didaskalivai denoted “teachings” while diatavxei" were “commands.” He speculated that the Constitutions were created out of these smaller, older, diverse texts, including the Didache.155 In the decades closer to the discovery of the ancient manuscript of the Didache, discussions involving genre came from two directions. On the one hand, general studies were published on ecclesiastical law, again with the 151 Athanasius stated that, though it was not considered to be Scripture, the Didache, among other books, was prescribed by the Church fathers to be read to catechumens (Bibliva … tetupwmevna de; para; tw'n patevrwn ajnaginwvskesqai toi'" a]rti prosercomevnoi" kai; Boulomevnoi" kathcei'sqai to;n th''" eujsebeiva" lovgon, Ep. fest. 39 [ZAHN, Geschichte des neutestamentlichen Kanons, 2.1.212]). This characterization of the Didache had already been pointed out by JAMES USSHER in his 1644 publication, Polycarpi et Ignatii epistolae (Oxford: Lichfield), where he disputed the identification of the Didache with the Constitutions. 152 GRABE, Arabick Manuscripts, 30. The text for the Catalogue of the 60 Canonical Books can be found in ZAHN, Geschichte des neutestamentlichen Kanons, 2.1.297–301. 153 GRABE, Arabick Manuscripts, 40, citing Ecchellensis, De Alexandrinae Ecclesiae originibus. 154 Ibid., 39–40, 50. 155 JOHN PEARSON, Vindiciae epistolarum S. Ignatii (Cambridge, 1672), 60–62.

1.3 The Didache as Church Order – Early Research

35

overarching purpose of determining what constituted an authoritative text for the Christian Church. For example, Bickell treated briefly the development of the term kanwvn in the first three centuries of the Church, from a word used more generally for everything that could be regarded as a guide for the Christian life, to a denotation of the tradition (oral or written) safeguarded by the Church (or a specific teaching thereof), to a term referring to various synodal decrees.156 A short but more comprehensive discussion was found in an 1856 study of Christian legal writings in the pre-Constantinian era by Paul de Lagarde. Here Lagarde distinguished between the terms novmoi (legal rulings of the Pentateuch), ejntolaiv (mandates of Christ which were encouraged but not required), diatavxei"/constitutiones (rulings ascribed, however doubtfully, to the apostles), kanovne" (rulings fixed by Church councils), o{roi, yh'foi, and dovgmata (decrees of synods but not recognized by the entire Church), and epistolae, (whose authority depended on the worthiness of the author).157 In this context he also mentioned the Didaskaliva tw'n ajpostovlwn in conjunction with the Constitutions and distinguished this text from the Didaca;" tw'n ajpostovlwn, which he identified with a Syriac text by that name.158 On the other hand, some aspects of the genre question can be found in studies generated by the publication of the Apostolic Church Order by Bickell.159 With a text in hand that was obviously related to both Constitutions Book 7 as well as Barnabas, Bickell thought perhaps to have procured the ancient Didache. In support of this idea he cited, among other evidence, the fact that the Apostolic Church Order used the term didachv, though admitting that its contents were more like “commands” than “teachings.”160 Thus ques156

BICKELL, Kirchenrechts, 1.1.3–8. PAUL A. DE LAGARDE, Reliquae juris ecclesiastici antiquissimae (graece edidit) (Leipzig, 1856), iii–vii. 158 Ibid., vi. This text has no direct relation to that represented by H54. With an opening based on the scenes found in Acts 1–2, it lists twenty-seven “apostolic” ordinances dealing with worship and church office, guarantees these through apostolic succession, and concludes with an account of the geographical division of the mission of the Church among the various apostles. An account of the publication and the history of research of this text can be found in WITOLD WITAKOWSKI, “The Origin of the Teaching of the Apostles,” in IV Symposium Syriacum 1984: Literary Genres in Syriac Literature (eds. H. J. Drijvers, R. Lavenant, S. J., C. Molenberg, and G. J. Reinink; OrChrAn 229; Rome: Pontifical Institutum Studiorum Orientalium, 1987), 161–71. 159 With respect to the discovery of the Apostolic Church Order, it should be noted that GRABE himself had seen the text in the Greek Vindobonensis manuscript and had speculated that it was the libellus ab apostolis mentioned by Origen (Arabick Manuscripts, 36). 160 BICKELL cited four reasons that made the identification of the Didache and the Apostolic Church Order plausible to him: 1) the term didachv occurs in Apos. C.O. 5; 2) Eusebius mentions the Didache immediately before Barnabas, a text that, in its Two Ways passage, is related to the Apostolic Church Order; 3) the size of the Didache as attested by Nicephorus is too short to be the Constitutions; and 4) in an Oxford manuscript of the Catalogue of the 60 157

36

Chapter 1: The Didache and the Question of Genre

tions regarding the form, size, and subject matter of a didachv were already being raised. Subsequent discussions concentrated, however, on the possible relationship of the Didache with the Apostolic Church Order and/or the Duae Viae mentioned by Rufinus,161 as well as the theory that there was some other common source.162 Thus the announcement of the discovery of the Didache in 1875 (among the texts of codex H54) and its subsequent publication in 1883 found an academic world well prepared to receive it. The most closely related works, the Epistle of Barnabas, the Apostolic Church Order and the Constitutions, had been published and their potential interrelationships discussed; the possibility of identifying the Apostolic Church Order with the Didache and/or with the Two Ways/Judicium Petri had been raised alongside the theory that a lost, common source, perhaps with these titles and either pre- or postdating Barnabas, had been used. Finally, with Philotheos Bryennios’s publication of the Didache, there was at last a text that bore the title DIDACH TWN DWDEKA APOSTOLWN and whose primitive characteristics placed it chronologically prior to the Apostolic Church Order and the Constitutions.

Canonical Books, the Didache is listed as a text distinct from the Didascalia of Clement, a text which can be identified with the first six books of the Constitutions but which also circulated independently and which includes, in some eastern manuscripts, the Apostolic Church Order. On the other hand, in addition to the seeming discord between the title of the Didache and the contents of the Apostolic Church Order, Bickell acknowledged the fact that the material attributed to the Didache by the author of Adversus aleatoribus was lacking in the Apostolic Church Order (though not, as Bickell believed, in the Constitutions) (Kirchenrechts, 1.1.96–97, n. 14). 161 “In nouo uero Testamento libellus qui dicitur Pastoris siue Hermae, et is qui appellature Duae uiae, uel Iudicium secundum Petrum,” Rufinus, Symb. 36 (38) (CCSL 20.171). An April 1854 article in the Christian Remembrancer (“Article 1,” n.s. 27 [April, 1854]: 251– 309) suggested the identification of the Apostolic Church Order published by BICKELL with the Doctrina/Didache found in Athanasius and Eusebius as well as the Two Ways/Judgment of Peter mentioned by Rufinus (277, 289–97). In 1866 ADOLF HILGENFELD further pushed the identification with the Didache (Evangeliorum secundum Hebraeos, secundum Petrum, secundum Aegyptios, Matthiae traditionum, Petri et Pauli praedicationis et actuum, Petri apocalypseos, Didascaliae apostolorum antiquioris quae supersunt [pt. 4 of Novum Testamentum extra canonem receptum; 1st ed.; Leipzig: Weigel; 1866], 93–106; [2d rev. ed.; 1884], 110–21). 162 The idea of a common source behind the texts of the Two Ways in Barnabas, the Apostolic Church Order and the Constitutions was raised by BICKELL (Kirchenrechts, 1.1.90–92) and further discussed by OSCAR VON GEBHARDT and ADOLF VON HARNACK in “Barnabae epistula,” Patrum apostolicorum opera (eds. Oscar von Gebhardt, Adolf Harnack and Theodor Zahn; fasc. 1, pt. 2; 2d ed.; Leipzig, 1878). A theoretical reproduction of the source was published by A. KRAWUTZCKY, “Über das altkirchliche Unterrichtsbuch: ‘Die Zwei Wege oder die Entscheidung des Petrus’,” TQ 64/3 (1882): 359–445.

1.3 The Didache as Church Order – Early Research

37

1.3.3 Initial Research on the Didache of H54 In the editio princeps of the text and subsequent journal articles, Bryennios presented his analysis of the historical character of the Didache and its connection to other early, ecclesiastical texts clearly related to it. Bryennios believed the Didache to be the work of a simple, pious, Christian layperson, writing to an early Christian community in order to provide them with the practical requirements of the faith and to warn them against those preaching false doctrine.163 In his view, the text, while not limited to catechumens, nonetheless had been designed with gentile converts in mind, at a time when there was not yet a very clear distinction between clergy and laypeople.164 While not addressing the topic of genre directly, Bryennios did discuss several interrelated aspects of the Didache that have a bearing on its genre, namely its titles, structure, and overall compositional integrity. The Didache bears two titles, the first, Didach; tw'n dwvdeka ajpostovlwn (Teaching of the Twelve Apostles) followed by the longer title Didach; Kurivou dia; tw'n dwvdeka ajpostovlwn toi'" e[qnesin (Teaching of the Lord Through the Twelve Apostles to the Gentiles). Both of the titles call the text a didache and this raises the question as to the relationship between that term and the form and contents of the text. In addition, the outside witnesses to the Two Ways and the obvious transition at Did. 7.1 led many to regard the first six chapters as having come from an earlier, independent work. Subsequently questions were raised regarding the use of sources in the composition of the Didache, the possibility of redactional stages in the text, and the relationship of the titles to the Two Ways source and to the text as a whole. With respect to the term didache, Bryennios pointed to its use in Acts and 2 John: Acts 2:42 h\san de; proskarterou'nte" th'/ didach'/ tw'n ajjpostovlwn. … They devoted themselves to the apostles’ teaching. … 2 John 9 pa'" oJ proavgwn kai; mh; mevnwn ejn th'/ didach'/ tou' Cristou' qeo;n oujk e[cei: Everyone who does not abide in the teaching of Christ, but goes beyond it, does not have God.165

He believed the shorter title of the Didache in H54 to be a later abridgment of the longer, original one, which itself had referred only to the Two Ways section.166 While the entire book was composed for a community of believers, chs. 1–6 contained the teaching of the Lord per se, teaching whose special

163

BRYENNIOS, Didachv, d', e'. Ibid., q'–i'. 165 Ibid., b', n. 1. 166 Ibid., g'. 164

38

Chapter 1: The Didache and the Question of Genre

use, according to 7.1, was for the instruction of new converts.167 That the long title only applied to chs. 1–6 allowed, in his opinion, for the subsequent transition to instruction on ritual and community organization, instruction that could not have originated with the Lord in any way and must be from the author. The phrase in 7.1 tied the Two Ways to the remainder of the text and “saved,” according to Bryennios, its compositional integrity.168 Yet, despite the title, Bryennios noted that the one “speaking” in the text could be neither the Lord nor the apostles, and must remain anonymous.169 Still, unlike the Constitutions or the Apostolic Church Order, the text did not attempt to derive authority by falsely claiming the authorship of a prestigious figure.170 Bryennios envisioned the overall structure of these sixteen chapters as consisting of four sections: a) the Two Ways (1–6); b) instructions for worship, i.e., baptism, fasting, prayer, Eucharist and meeting on the Day of the Lord (7–10, 14); c) instructions concerning prophets, teachers, traveling Christians, and overseers/bishops and deacons (11–13, 15); and d) instruction on the Parousia (16).171 Despite the evident lapses in continuity, he believed the text to be the unified work of one individual. He ultimately dated the Didache to 120– 160 C.E. and identified it with the text mentioned by Clement of Alexandria,172 Eusebius, Athanasius, and Nicephorus. The character of the numerous scholarly reactions to the publication of the Didache was largely divided along national lines. In Germany, where the majority of the earlier scholarship on Didache-related materials had taken place, research continued along the lines of historical criticism. On the other hand, in Great Britain and the U.S., where there was more diversity in Protestantism and where the historical-critical method had been perceived by many as disruptive, if not destructive, to Christian faith, the questions asked of the Didache gave a high priority to the issues of ritual practice and ecclesiastical organization. Despite the difference in emphases, however, most agreed that

167 PHILOTHEOS BRYENNIOS, “Peri; th'" Didach'" tw'n dwvdeka ajpostovlwn,” Δ∆Ekklhsiastikh; ajlhvqeia (11/10 [22]/1884): 52a–b. 168 Ibid., 54b. 169 BRYENNIOS, Didachv, g'–d'. EHRMAN also considers the Didache to be anonymous stating, “neither title claims that the book was actually written by the apostles, simply that it conveys their teachings; it is, therefore, anonymous rather than pseudonymous,” (Apostolic Fathers, 1.406). 170 BRYENNIOS, Didachv, d'. 171 Ibid., g' 172 BRYENNIOS believed that Strom. 1.20.100.4, where Clement reproduces a passage very close to Did. 3.5 and cites it as “scripture,” is a reference to the Didache (Didachv, 16, n. 15). See below, p. 106.

1.3 The Didache as Church Order – Early Research

39

the Didache was an authentic first- or second-century Christian text, and therefore a valid source for historical information.173 As with Bryennios, the subject of the genre of the Didache was generally only indirectly discussed by early commentators. In 1884 various anonymous articles in The Independent, for example, called the text “the oldest book of church order,” “a directory of worship,” and “the first church manual ever written,” using descriptive rather than technical terminology.174 Theodor Zahn believed the Didache to be an early, second-century Christian “praktisches Hilfsbüchlein” intended to stand alongside other sources of Christian teaching and edification,175 while Roswell Hitchcock and Francis Brown considered the “treatise” to contain “the substance of that which Christians need concern themselves about.”176 Philip Schaff called the Didache a “church manual or brief directory of apostolic teaching, worship and discipline,” that “serves its purpose admirably: it is theoretical and practical, short and comprehensive.”177 L. Massebieau described the Didache as “une sorte de livre d’eglise, manuel très court, rédigé certainement de manière à pouvoir être appris par coeur et où se trouve le nécessaire pour la constitution et le fonctionnement d’une communauté chrétienne. On peut dire, en employant ici des mots trop techniques, qu’il contient ce qui paraissait indispensable en fait de catéchuménat, de liturgie et de discipline.”178 Paul Sabatier also described the Didache as “un manuel de religion chrétienne,” and likewise Charles Taylor called it a “church manual.”179 G. Bonet-Maury believed the text to be the lost “Discipline ecclésiastique” attested by the early Fathers (though he thought it originally to be only a Two Ways text),180 and H. Holtzmann called it the

173 Among the dates suggested by scholars for the Didache were, for example, SABATIER, 50 C.E.; FUNK, MASSEBIEAU, DE ROMESTIN, SPENCE, 75–100 C.E.; LIGHTFOOT, 80–110 C.E.; ZAHN, 80–120 C.E.; SCHAFF, 90–100 C.E.; FARRAR, HITCHCOCK/BROWN, BICKELL, 100 C.E.; HARNACK, 120–160 C.E.; HILGENFELD, BONET-MAURY, 160–190 C.E.; KRAWUTZCKY, latter half of the second century C.E. 174 The Independent (New York), 28 February, 6 March, 3 April, 24 April 1884. 175 ZAHN, Geschichte des neutestamentlichen Kanons, 2.1.289. 176 ROSWELL D. HITCHCOCK and FRANCIS BROWN, Didach; tw'n dwvdeka ajpostovlwn: Teaching of the Twelve Apostles (2d ed.; New York: Charles Scribner’s Sons, 1885), lxxxv, n. 2. 177 SCHAFF, Teaching of the Twelve Apostles, 16. 178 L. MASSEBIEAU, “L’enseignement des douze apôtres,” RHR 10 (September–October 1884): 130. 179 PAUL SABATIER, Didach; tw'n ie' ajpostovlwn: La Didache ou l’Enseignement des douze apôtres (Paris: Fischbacher, 1885), 68; CHARLES TAYLOR, The Teaching of the Twelve Apostles with Illustrations from the Talmud (Cambridge: Deighton Bell and Co., 1886), v. 180 G. BONET-MAURY, La doctrine des douze apôtres (Paris: Fischbacher, 1884), 14–15, 35–36.

40

Chapter 1: The Didache and the Question of Genre

“älteste Kirchenordnung, die wir haben.”181 Still others characterized the Didache as “a compendium of apostolic teaching – a ‘Summa’ accepted by Christians in A.D. 140,”182 and “the oldest uninspired book of Church Order known to this generation.”183 Most scholars, whether they believed the text to be in its original form or to exhibit redactional stages, recognized a clear break in chs. 6–7, and it is at that point that Karl München believed the actual “church order” to begin, describing the initial chapters as a “catechism.”184 Similarly F. X. Funk described the first six chapters as “eine einfache Unterweisung in der christlichen Sittenlehre” and the last ten chapters as a “Rituale” or “Kirchenordnung.”185 The phenomenon of the double title was certainly addressed in the more thorough of the early studies. Of primary importance was the debate over which of the two titles might be the original. Since neither of them agrees precisely with the external ancient attestations (nor do the attestations agree among themselves) this evidence was not determinative. Following Bryennios, most scholars, including Hitchcock and Brown, Schaff, and Aug. Wünsche believed the long title to be original and the short to be an abbreviation, though only a few like Wünsche and Funk agreed completely with Bryennios that the long title applied solely to the Two Ways.186 On the other hand Adolf Hilgenfeld thought the longer title was secondary while Benjamin Warfield, deeming the Two Ways to have been originally an independent work, thought the short title, omitting dwvdeka and possibly tw'n, was the original title of that text.187 Similarly Bonet-Maury thought the Two Ways to be the earliest part of the Didache and the original title to be the first, shorter title,188 while Taylor also theorized that the Two Ways (apart from a few interpolations) was originally an independent Jewish manual of ethics, possibly pre-Christian.189 181 H. HOLTZMANN, review of Adolf von Harnack, Die Lehre der zwölf Apostel nebst Untersuchungen zur ältesten Geschichte der Kirchenverfassung und des Kirchenrechts, DLZ 5, no. 40 (4 October 1884): 1452. 182 W. E. ADDIS, in Notices of Books, Dublin Review, 3d ser., 12 (July–October 1884): 444–45. 183 J. Z. TYLER, “The Oldest Book of Church Order,” Disciple of Christ (1 April 1884): 199. 184 KARL MÜNCHEN, “Die Lehre der zwölf Apostel, eine Schrift des ersten Jahrhunderts,” ZKT 10 (1886): 662. 185 FRANZ XAVER FUNK, “Die Doctrina apostolorum,” TQ 66 (1884): 394, 396. 186 AUG. WÜNSCHE, Lehre der zwölf Apostel (2d ed.; Leipzig: Schulze, 1884), 9, n. 1; FRANZ XAVER FUNK, “Zur Apostellehre und Apostol. Kirchenordnung,” TQ 69 (1887): 2–3. Contrary to BRYENNIOS, however, WÜNSCHE believed the short title to have been added with the addition of chs. 6–16. 187 BENJAMIN WARFIELD, “Textual Criticism of the Two Ways,” Expositor, 3d ser., 3 (February 1886): 158. 188 BONET-MAURY, La doctrine, 14–15, 35–36. 189 TAYLOR, Teaching of the Twelve Apostles, 22, 49, 113.

1.3 The Didache as Church Order – Early Research

41

A. Chiappelli, believing the text to be a joining of two originally independent, late-first-century works, namely a Two Ways text and a ritual text, reconstructed the title of the first as “duvo oJdoiv” or “didach; tw'n duvo oJdw'n,” a title that became didach; tw'n ajpostovlwn after the addition of the second text.190 Generally, the points addressed concerned the originality of one or the other title, the supposedly gentile audience, and the relationship between the text and actual words of the Lord. Only a few writers dealt with the term didache itself. Schaff, who upheld the originality of the long title, agreed with Bryennios that the title was “borrowed” from Acts 2:42, though he added that this was to be understood “of the contents, not of the form” of the Didache.191 E. de Muralt, on the other hand, cited the same passage, but maintained the originality of the short title.192 Taylor believed the text of the Didache to be a “skeleton” of the complete, oral tradition referred to in the New Testament as “the Teaching.” To the evidence from Acts and 2 John, he added Titus 1:9 and also pointed to the occurrences of the expression in Justin Dial. 35 and Barn. 18.1.193 Funk also connected didache with evidence from the New Testament and other early literature, citing also Acts 13:12; Titus 2:10, Justin 1 Apol. 53 and Mart. Ascen. Isa. 3.194 J. Rendel Harris, however, believed that the title “Didache” was “Hebraistic” and an equivalent to Hebrew “Talmud.” He saw especially the beginning of the text as a presentation of “Thorah” (Did. 1.1–2) followed by commentary, i.e., Targum or Talmud (Did. 1.3ff.).195 Beyond this, the idea of didache as representing a genre was not explored by scholars. In this early period of research on the Didache, among the most important publications were those of Adolf von Harnack, which began to appear almost immediately in 1884.196 Harnack identified the text with the Didache attested 190

ALESSANDRO CHIAPPELLI, “Le relazioni fra la Dottrina dei dodici Apostoli, il Pastore d’ Erma e la Lettera di Barnaba,” in Studii di antica letteratura cristiana (Turin: 1887; reprint, Bologna: Arnaldo Forni Editore, 1977), 142–43. He maintained, against BRYENNIOS and HARNACK, that, given the differences between the two parts, the long title could not have originally applied to the entire work (127). 191 SCHAFF, Teaching of the Twelve Apostles, 14–15. 192 E. DE MURALT, “L’enseignement des douze apôtres,” RTP 17 (May 1884): 281. 193 TAYLOR, Teaching of the Twelve Apostles, v–vi, 112–13. 194 FRANZ XAVER FUNK, Doctrina duodecim apostolorum (Tübingen: Henrici Laupp., 1887), 3–4. 195 J. RENDEL HARRIS, The Teaching of the Apostles (Baltimore: Johns Hopkins University; London: C. J. Clay and Sons, 1887), 79–80. 196 In addition to HARNACK’s 1884 book, Die Lehre der zwölf Apostel nebst Untersuchungen zur ältesten Geschichte der Kirchenverfassung und des Kirchenrechts, his publications included: a review of BRYENNIOS, Didachv tw'n ajpostovlwn, TLZ 9 (9 February 1884): 49–55; a review of HILGENFELD, Evangeliorum secundum Hebraeos, secundum Petrum, secundum Aegyptios, Matthiae traditionum, Petri et Pauli praedicationis et actuum, Petri apocalypseos, Didascaliae apostolorum antiquioris quae supersunt, TLZ 9 (1884): 337–43;

42

Chapter 1: The Didache and the Question of Genre

by the early Fathers and estimated its date at 135–165 C.E., primarily because he believed that the author used both Barnabas and Hermas. Harnack maintained the integrity of the text while logically dividing it into three subsections: I) chs. 1–10; 2) chs. 11–15; and 3) ch. 16.197 He held that the long title of the text was the original,198 and, against Bryennios, that it referred to the entirety of the text and not just the Two Ways.199 He felt that the title truly expressed the intention of the author for the Didache to be an explanation to gentile Christians of the Lord’s teaching regarding the life of the individual Christian as well as the Christian community, teaching that had been passed on by the apostles.200 The work contained instruction based on, and in the form of, words of the Lord, (Did. 1.3) sometimes from the “gospel,” (Did 8.2; 11.3; 15.3, 4), sometimes from the ejntolaiv (1.5; 4.13; 13.5, 7).201 The term didache seemed to him to have originally designated ‘moral teaching’ and within this text referred to the subsequent exegesis of the opening directive to love God and love neighbor.202 Yet Harnack observed that, until the middle of the second century, it was apparently allowable for moral teaching and ecclesiastical instruction, derived through the medium of the apostles from the Lord, to be designated “didach; kurivou dia; tw'n ibV ajpostovlwn.”203 Thus the title was appropriate for the whole text and applied to the entire congregation.204 That the Didache emphasized the practical over the theological Harnack found to be a common trait among early Christian texts, including Barnabas, 1 Clement, 2 Clement, Hermas, and the works of Justin, and independently confirmed by the witness of Pliny.205 According to Harnack, theological material was not at first transmitted in didactic forms, but in cultic and hymnic formulae; it was only in the battle against Gnosticism that theology Die Apostellehre und die jüdischen beiden Wege (Leipzig: Hinrichs, 1886); and “Zur Lehre der zwölf Apostel – III Artikel,” TLZ 11 (12 June 1886): 271–73; (24 July 1886): 344–47; 12 (1887): 32–34. 197 HARNACK, Lehre der zwölf Apostel nebst Untersuchungen, 37–40. 198 HARNACK compared the long title of the Didache to Pol. Phil. 6.2; 2 Pet 3:2; and especially Matt 28:19–20, for which the title of the Didache seemed to him to be a gloss (Lehre der zwölf Apostel nebst Untersuchungen, Proleg. 25, n. 2). To show that the abbreviation found in the short title was not exceptional, he cited as an example the work, Diatagai; tw'n ajpostovlwn dia; Klhvmento" (seemingly the Apostolic Constitutions), which he stated was found abbreviated as Diatagai; Klhvmento" (Proleg. 30–31). 199 Ibid., Proleg. 29–30. 200 Ibid., Proleg. 29–30. 201 Ibid., Proleg. 24–25. 202 Ibid., Proleg. 25. HARNACK points to Did. 1.3–4, and 2.1. 203 Ibid., Proleg. 239. 204 Ibid., Proleg. 30. 205 Ibid., Proleg. 32–34. In his description of Christians, Pliny notes as part of their meetings a pledge, among other things, not to commit robbery and adultery (Pliny, Ep. 10.96.7 [Radice, LCL 59]).

1.3 The Didache as Church Order – Early Research

43

was codified, a phenomenon that could be observed in the writings of Irenaeus.206 Harnack’s comments on the title and structure of the Didache laid the foundation for a study on its genre. Simultaneously, however, he also referred to the work as the oldest “church order”207 and regarded it as the basis and model of all the subsequent constitutional literature of the eastern Church, applying the term Kirchenordnung to these as well, perhaps encouraged by Bickell’s title of the Apostolic Church Order.208 While Harnack did not use only this term to describe the Didache209 and certainly would have been aware of its original meaning, it was a convenient way to communicate the character of the text, whose original genre remained undetermined or at least without example. Thus it seems likely that the title of the Apostolic Church Order as well as Harnack’s appropriation of that term to the Didache encouraged a still broader use among scholars. A. Krawutzcky, expanding upon his earlier study of the Two Ways text, also contributed an important study on the sources and redaction of the Didache. According to Krawutzcky, the “new teacher” used: 1) a Two Ways document comprised of parts of chs. 1–6; and 2) an anti-Ebionite “Verordnung” that prescribed rules for baptism and the Eucharist, included instruction on fasting, prayer, Sunday celebration, and the selection and importance of ministers, and concluded with a word on watchfulness (7–11.2; 14–15; 16.1– 2). To these texts, the (pro-Ebionite) author added other source material as well as his own comments.210 1.3.4 Research at the Turn of the Nineteenth Century Subsequent to the initial flood of publications on the Didache, the underlying question of its genre developed in several directions. First of all, the appearance of the Didache encouraged the study of texts related to it in content, as well as those similar to it in character, and the convenient designation of such works individually and collectively as ”church orders” subsequently increased. On the one hand, work continued on the Constitutions,211 the Apos206

HARNACK, Lehre der zwölf Apostel nebst Untersuchungen, Proleg. 34–35. HARNACK, Review of BRYENNIOS, cols. 50, 55; see also Die Apostellehre und die jüdischen beiden Wege, 2. 208 HARNACK, Lehre der zwölf Apostel nebst Untersuchungen, Proleg. 223. 209 HARNACK also called it a Leitfaden in which the editor compiled “in übersichtlicher, leicht fasslicher und leicht behaltlicher Form die wichtigsten Regeln für das christliche Leben, die didavgmata tou' kurivou” (Lehre der zwölf Apostel nebst Untersuchungen, Proleg. 32). 210 A. KRAWUTZCKY, “Ueber die sog. Zwölfapostellehre, ihre hauptsächlichsten Quellen und ihre erste Aufnahme,” TQ 66/4 (1884): 579, n. 1; 573–75. 211 FRANZ XAVER FUNK, Die Apostolischen Konstitutionen (Rottenburg: Bader, 1891); idem, Didascalia et Constitutiones apostolorum (2 vols.; Paderborn: Schöningh, 1905). 207

44

Chapter 1: The Didache and the Question of Genre

tolic Church Order,212 and the Didascalia,213 the latter which, along with the publication of the Didache, enabled the tracing of the sources of the Constitutions. Also studied was another family of texts whose origins could all be traced to the Apostolic Tradition of Hippolytus: the Egyptian (and related Ethiopic) Church Order,214 the Canons of Hippolytus,215 the Testamentum Domini,216 Const. ap. 8.3–46, and the Constitutions through Hippolytus (also called the Epitome of Const. ap. 8.3–46). As with the Constitutions, many of the shorter texts were found in the larger collections of the Syriac Octateuch (Testamentum Domini, Apostolic Church Order, Epitome and Apostolic Canons217) and the Clementine or Egyptian Heptateuch, also called the Ecclesiastical Canons (Apostolic Church Order, Egyptian Church Order, Epitome and, in some versions, Apostolic Canons). One of the most important works was the publication of the Latin Verona fragments in 1900 by Edmund Hauler attesting parts of the Didascalia, Apostolic Church Order and Egyptian Church Order.218 In 1904, G. Horner published the Ethiopic, Arabic, and Sahidic (fragmentary) versions of the Ecclesiastical Canons, and in 1905 Funk 212 Most important are the publications of ADOLF VON HARNACK, Die Quellen der sogenannten Apostolischen Kirchenordnung (TU 2/5; Leipzig: Hinrichs, 1886); and J. P. ARENDZEN, “An Entire Syriac Text of the Apostolic Church Order,” JTS 3 (1901–1902): 59– 80. 213 First published in its entirety by PAUL A. DE LAGARDE, Didascalia apostolorum syriace (Leipzig: 1854); later by MARGARET DUNLOP GIBSON, The Didascalia apostolorum in Syriac and The Didascalia apostolorum in English (HSem 1, 2; London: 1903); FUNK, Didascalia et Constitutiones; and J. M. HARDEN, The Ethiopic Didascalia (Translations of Christian Literature, ser. IV, Oriental Texts; London: SPCK, 1920). 214 First published in Coptic with English translation by HENRY TATTAM, The Apostolic Constitutions or Canons of the Apostles in Coptic (London: Oriental Translation Fund, 1848); then by LAGARDE in 1853 (publ. 1854), “Constitutiones apostolicae graece … accedunt Constitutiones ecclesiae aegyptiacae,” in Reliquiae canonicae (vol. 2 of Analecta ante-nicaena, by Christian Karl Josias von Bunsen; London, 1854; reprint, Darmstadt: Scientia Verlag Aalen, 1968), 449–77. It was LAGARDE who gave it this title. Post-Didache it was published by G. HORNER, The Statutes of the Apostles or Canones ecclesiastici (London: Williams & Norgate, 1904). 215 Editio princeps in Arabic with Latin translation by D. B. DE HANEBERG, Canones S.Hippolyti (Munich: 1870); also published by HANS ACHELIS, Die ältesten Quellen des orientalischen Kirchenrechtes. Erstes Buch: Die Canones Hippolyti (TU 6/4; Leipzig: Hinrichs, 1891). 216 The Syriac version with a Latin translation was published by I. E. RAHMANI, Testamentum Domini nostri Jesu Christi (Mainz: Kirchheim, 1899). Later appeared an English translation by JAMES COOPER and ARTHUR JOHN MACLEAN, The Testament of Our Lord (Edinburgh: T. & T. Clark, 1902); and a study by F. X. FUNK, Das Testament unseres Herrn und die verwandten Schriften (Mainz: Kirchheim, 1901). 217 Apostolic Canons is the title given to the 85 canons found at the end of some manuscripts of the Constitutions Book 8. 218 EDMUND HAULER, Didascaliae apostolorum fragmenta ueronensia latina (Leipzig: Tuebner, 1900).

1.3 The Didache as Church Order – Early Research

45

published two volumes, the first of which presented the Didascalia and Didache in parallel with the Constitutions and the second which gave texts related to these works and other witnesses, including the Constitutions through Hippolytus and the Egyptian Church Order.219 The nature of the material and the previous use of the term church order in the titles of several of the texts seem to have made its use as a general designation natural. In his 1891 publication on the Constitutions, Funk used “church order” in his description of the Constitutions as well as the Didascalia, and Hans Achelis used it in his study of the Canons of Hippolytus that same year.220 In 1901, John Wordsworth published a collection of studies in early Christianity, including a section entitled “Church Orders,” where he discussed the individual texts as well as the larger collections and other related materials.221 Horner also used the term when he mentioned the theory of a lost church order behind the texts related to the Apostolic Tradition.222 In their work on the Testamentum Domini, James Cooper and Arthur John Maclean used “church order” (in quotation marks!) to describe the Testament, its related works, and other similar works, including the Didache, which they called an “elementary ‘Church Order’.”223 Then in 1910, Maclean provided a handbook “to give a description of certain early Christian manuals of instruction and worship which are conveniently called Church Orders.”224 Among the general characteristics of this literature, he included pseudonymous form, antiquarianisms (attempting to make the texts appear older), the experimental nature of the liturgical forms, and the tendency toward compilations.225 Again the term was used by R. H. Connolly in 1916 when he explored the interrelationship of the “closely related family of Church Orders in which the eighth book of the Constitutions and the Canons of Hippolytus are the most conspicuous members.”226 In 1924 James Vernon Bartlet delivered lectures on “Church-Life and Church-Order During the First Four Centuries,” and included in these the development of the “church orders.” Within this development, the Didache was the earliest written form of what had become known as the “didache” in the apostolic age, i.e., a standard collection of the most 219

FUNK, Didascalia et Constitutiones. FUNK, Konstitutionen, 356, 360; ACHELIS, Canones Hippolyti, forward. 221 JOHN WORDSWORTH, The Ministry of Grace (London: Longmans, Green, and Co., 1901), 13–63. 222 HORNER, Statutes, viii. 223 COOPER and MACLEAN, Testament, 11. 224 ARTHUR JOHN MACLEAN, The Ancient Church Orders (Cambridge: Cambridge University Press, 1910), 1. 225 Ibid., 3–9. 226 RICHARD HUGH CONNOLLY, The So-Called Egyptian Church Order and Derived Documents (Cambridge: Cambridge University Press, 1916), vii. For other examples, see HARDEN, Ethiopic Didascalia, ix–xi. 220

46

Chapter 1: The Didache and the Question of Genre

important traditions on church order.227 The Didache, according to Bartlet, had been composed in three stages between 65–100 C.E. and initially (65– 75 C.E.) was comprised of the Two Ways (prior to the addition of 1.3b–2.1) and most of the present text on baptism, on the Eucharist, and in the concluding eschatological chapter, all under the title “Teaching of the Lord.”228 Through all of these examples it is clear that the term church order had become a de facto genre designation for this type of early Christian literature. 1.3.5 The Doctrina apostolorum and the Two Ways Another development in the study of the Didache that is important in determining its genre came with the publication in 1900 by Joseph Schlecht of a complete edition of the Doctrina Apostolorum, a third-century Latin text nearly identical to Did. 1.1–6.1, though it lacks 1.3–2.1 and contains some Barnabean elements.229 A fragment of this text had previously been published by Oscar von Gebhardt in 1884 and Funk in 1886,230 but the appearance of a complete text prioritized the question of whether there was a pre-existent Two Ways source used by the Didache and/or Barnabas. That there existed such a source, whether Jewish or Christian, used in the Didache, had earlier been accepted, for example, by Warfield, Bonet-Maury, Taylor, and also Harnack in his second book on the text, Die Apostellehre und die jüdischen beiden Wege. The publication of the complete Doctrina, however, added to the plausibility of this scenario and fed the discussion of the interrelationships of the Didache, Barnabas, Doctrina, and Hermas. In 1912, J. Armitage Robinson attempted to show that the author of the Didache used a Two Ways text and other early writings for the purpose of presenting to his own, mid-secondcentury community a “manual of Church discipline,” a manual which the 227

BARTLET, Church-Life and Church-Order, 34. Ibid., 53–55. The second stage (80–90 C.E.) saw the expansion of the title to Teaching of the Lord through the Twelve Apostles to the Gentiles and the addition of chs. 8, 11.3–12, 14, and various other smaller units, especially material related to the Gospels. The final redaction (about 100 C.E.) updated the materials with a few instructions on church organization. Unfortunately BARTLET, prior to his death, nowhere gave a complete delineation of the text development (see editor’s note, p. 55). BARTLET, however, designated the Apostolic Tradition as the “earliest formal Church-Order,” (67). 229 JOSEPH SCHLECHT, Doctrina duodecim apostolorum (Freiburg: 1900); idem, Doctrina xii apostolorum: Die Apostellehre in der Liturgie der katholischen Kirche (Freiburg: Herder, 1901). 230 GEBHARDT discovered the fragment in the work of BERNHARD PEZ, Thesaurus anecdotorum novissimus (6 vols.; 1721–1729T), 4.2.5–8, and published it under the title, “Ein übersehenes Fragment der Didachv in alter lateinischer Übersetzung,” in HARNACK’s Lehre der zwölf Apostel nebst Untersuchungen zur ältesten Geschichte der Kirchenverfassung und des Kirchenrechts, 275–86. It was later published by FUNK, “Zur alten lateinischen Übersetzung der Doctrina apostolorum,” TQ 68 (1886): 650–55. 228

1.3 The Didache as Church Order – Early Research

47

author believed to contain the original teachings of the Twelve Apostles to the gentiles.231 According to Robinson, the long title of the Didache, created by the author from Matt 28:19–20, revealed this purpose.232 In his opinion, the Didache had been intentionally made to appear ancient by its author and much of the picture it portrayed was the author’s imaginative reconstruction of the early Church and could not be taken as historically accurate. In 1929 Robinson expanded his theory to include the idea that the Two Ways text was not an independent work but an original and integral part of Barnabas.233 That same year James Muilenburg published his own study also supporting the priority and originality of the Two Ways in Barnabas and concluding that the Didache was a “compilation of sources designed to supply the functions of a Church manual.”234 F. E. Vokes agreed with this assessment and reintroduced the idea that the author of the Didache was an early Montanist,235 dating the text to the latter part of the second or early part of the third century.236 Connolly also supported the priority of Barnabas,237 as did W. Telfer, who believed the Didache to be the creation of a second-century, Syrian Christian trying to portray an apostolic model of Church life based on the suggestion in Matt. 28:19–20 that the Twelve Apostles “were authors of a Catholic Church Order” and on the account of the apostolic council in Acts 15 “which gave its charter to the primitive Syrian church.”238 In it the Didachist had used Barnabas as well as Matthew. On the other side, however, B. H. Streeter239 as well as Edgar Goodspeed defended the theory of a common source of the Two Ways, Goodspeed proposing that the Doctrina was a Latin translation of the “original Teaching of the Apostles,”240 a text that was later “expanded into the church manual we know as the Didache.”241 The idea of an independent Two Ways text encouraged study into the possible existence of a more or less standard early Christian catechism. In several 231

J. ARMITAGE ROBINSON, “The Problem of the Didache,” JTS 13 (1912): 341–42. Ibid., 340–41. 233 J. ARMITAGE ROBINSON, Barnabas, Hermas and the Didache (London: SPCK; New York: Macmillan, 1920). 234 JAMES MUILENBURG, “The Literary Relations of the Epistle of Barnabas and the Teaching of the Twelve Apostles” (Ph.D. diss., University of Marburg, 1929), 72. 235 The suggestion that the Didache was a Montanist reworking of pre-existent material had been proposed earlier by HILGENFELD. See, for example, “Die Lehre der zwölf Apostel,” ZWT 28 (1885): 73–102. 236 F. E. VOKES, The Riddle of the Didache (London: SPCK, 1938), 209–10. 237 R. H. CONNOLLY, “The Didache in Relation to the Epistle of Barnabas,” JTS 33 (1932): 237–53. 238 W. TELFER, “The ‘Plot’ of the Didache,” JTS 45 (1944): 146. 239 B. H. STREETER, “The Much-Belaboured Didache,” JTS 37 (1936): 369–74. 240 EDGAR J. GOODSPEED, “The Didache, Barnabas and the Doctrina,” AThR 27 (1945): 237. 241 Ibid., 235. 232

48

Chapter 1: The Didache and the Question of Genre

publications beginning in 1903, D. Alfred Seeberg attempted to prove the existence of a universal, pre-baptismal, Christian catechism that had been adapted from an originally Jewish, oral, proselyte catechism.242 He believed that this teaching, called “didache,” consisted of a dogmatic section, an ethical section called the “Ways,” and an eschatological section, and was already attested in the New Testament. Originally circulating in oral form among the Christian communities, it had been adapted and reformulated in various ways.243 The Didache itself was based on this teaching but lacked the dogmatic material while containing additional instruction on the Lord’s Prayer and Eucharist.244 In 1909 G. Klein, in an effort to retrace the steps from Jewish proselytism to the early Christian catechism, postulated that part of the oral instruction used in Judaism for converts included Two Ways material joined with an eschatological conclusion.245 This schema had been taken over into Christianity, as was clearly seen, he believed, in the Didache as well as in other texts.246 From the studies of Seeberg and Klein one could conclude that the genre of the Didache, possibly known as “didache,” was directly descended from an oral genre of catechetical teaching, a genre that perhaps was still visible in the Two Ways section, possibly originally concluding with ch. 16. At that time, however, the issue of genre was not pursued further.247 242 D. ALFRED SEEBERG, Der Katechismus der Urchristenheit (Leipzig: Deichert, 1903); Das Evangelium Christi, (Leipzig: Deichert, 1905); Die beiden Wege und das Aposteldekret (Leipzig: Deichert, 1906); Die Didache des Judentums und der Urchristenheit (Leipzig: Deichert, 1908). 243 Idem, Der Katechismus der Urchristenheit, 39–41. 244 Idem, Die Didache des Judentums, 85. 245 G. KLEIN, Der älteste christliche Katechismus und die jüdische Propaganda-Literatur (Berlin: Reimer, 1909). 246 Ibid., 244, 249. 247 In later works PHILIP CARRINGTON, (The Primitive Christian Catechism [Cambridge: Cambridge University Press, 1940]), and EDWARD GORDON SELWYN (“On the Inter-relation of 1 Peter and other N.T. Epistles,” in his book The First Epistle of St. Peter [London: Macmillan & Co. Ltd., 1964], 363–466) attempted to reconstruct a primitive baptismal catechism(s) based on recurring patterns and phraseology within certain New Testament texts and to isolate its sources. CARRINGTON’s model began with an introduction that included an exhortation to no longer live (“walk”) as the Gentiles and to endure temptation and persecution, a reference to the word of God, and the image of new birth/creation. This was followed by a baptismal parenesis based on the pattern: 1) deponentes (putting off vices) and worship; 2) subiecti (submission to God and leaders); 3) vigilate (vigilance with respect to temptation and persecution; and 4) resistite (resisting the devil) and state (keeping morally upright) (58; for a more developed version see p. 90). In his revision of CARRINGTON’s model, SELWYN postulated the existence of two different versions of this primitive catechism and separated from it the Vigilate/State section, which he believed to reflect a tradition independent from the catechism (“Inter-relation of 1 Peter,” 439–61). The reconstructions of the catechism for the New Testament texts by these scholars is problematic enough; though shared patterns and vocabulary are evident, variations are also abundant. When one compares these models to the

1.3 The Didache as Church Order – Early Research

49

1.3.6 The Didache and the Dead Sea Scrolls A move toward a broader perspective was made with the discovery of the Dead Sea Scrolls, in particular the Rule of the Community (1QS). The first scholar to point to a connection between the Didache and 1QS was Jean-Paul Audet in 1952.248 Clearly the two texts shared terminology, concepts, and, to some extent, structure, and these observations led Audet to the conclusion that the Didache represented a Christian rendition of a Jewish Two Ways text. According to Audet, the Didache had been composed in three stages: Stage 1) 1.1–11.2; Stage 2) 11.3–16.8, which was later but by the same author; and Stage 3) several passages interpolated at a later date by another editor, 1.3b– 2.1, 6.2–3, 7.2–4, 13.3, 5–7.249 With respect to genre, Audet pointed out the difficulty in recognizing a genre from a point in time in which it is no longer used.250 He saw the issue of the genre of the Didache as intimately connected with the problem of the double title in H54.251 In his reconstruction of the history of the title, Audet believed that the author of the Didache had begun with a Jewish Two Ways document entitled “Didach; kurivou toi'" e[qnesin,” to which was added instruction thought appropriate for the Christian communities. This new text was entitled “Didacaiv (pl.!) tw'n ajpostovlwn,” the plural being only natural since it was followed by the original title and text of the Two Ways. Thus the first title, in his opinion, pertained to the entire document while the second covered only the Two Ways material.252 In the early

Didache, even if only to chs. 1–6, the differences are much more extensive and even CARRINGTON had to admit that the Two Ways – of which the Way of Life was, in his opinion, originally an oral Jewish catechism of the Greek synagogue (Primitive Christian Catechism, 5, 95) – does not exhibit the same pattern (89). While the instruction included in the Two Ways is broadly the same as in the New Testament texts, i.e., traditional Jewish parenesis, the specific content, vocabulary, and organization are different. Along with this research on the catechism came C. H. DODD’s work on the relationship between this catechism and the primitive “kerygma” in The Apostolic Preaching and Its Developments (1937) and Gospel and Law (1951). DODD sought to make a clear distinction between kerygma and didache: kerygma, i.e., preaching, involved only the proclamation of the basic points of the Christian message, a proclamation directed to potential converts; didache, on the other hand, was community oriented and dealt mainly with ethical instruction. The two were not to be confused and the presence of didache always presupposed the existence of the kerygma. 248 JEAN-PAUL AUDET, “Affinités littéraires et doctrinales du Manuel de discipline,” RB 59 (1952): 219–38; La Didachè: Instructions des apôtres (Paris: Librairie Lecoffre, 1958). 249 La Didachè, 104–20. 250 Ibid., 91. 251 “Une erreur sur l’un est déjà le commencement d’une erreur sur l’autre,” La Didachè, 92. 252 According to AUDET, at a later time the meaning of the plural in the first title was no longer understood and the change to the singular was made. In addition the “mistaken” notion that both titles covered the entire work led to a desire for clarification and dia; tw'n dwvdeka

50

Chapter 1: The Didache and the Question of Genre

Church didachv had taken on a special meaning of ‘the instruction given by apostles as part of their function of establishing churches’. At some point, one of the itinerant apostles, i.e., the apostles mentioned in Did. 11, drew up a collection of these instructions to take along, a text that consisted of Did. 1.1– 11.2 (apart from 1.3b–2.1; 6.2–3; 7.2–4). Later this same apostle added Did. 11.3–16.8; finally, an interpolator inserted 1.3b–2.1; 6.2–3; 7.2–4; and 13.3, 5–7. Thus didachv became a “genre littéraire,” of which Audet cited as a New Testament example the instructions of 1 Corinthians 7–14, a “suite de didacaiv sur des sujets divers.”253 It is clear by this statement, however, that Audet was using genre in the sense of form and as a component rather than the entire text of the extant Didache. For Audet, a didachv is akin to an ejntolhv and its object is to prompt its audience to choose the correct way of life.254 Thus its subject matter has to do with peripatei'n, poiei'n, pra'xi".255 “Les Didacai; tw'n ajpostovlwn sont un recueil d’‘instructions’ de ce genre.”256 1.3.7 The Genre of the Two Ways In the 1960s and 1970s other important studies appeared that dealt, albeit indirectly, with the generic identity of the Didache. Klaus Baltzer, as part of his 1964 study entitled, The Covenant Formulary, attempted to show that the Two Ways of the Didache (as well as the entirety of Barnabas) followed the covenant formulary (Bundesformular) found in the Old Testament. By the era of early Christianity the formulary had undergone certain developments, but could, he maintained, still be recognized in various contexts. Thus the Didache and 1QS III, 13–IV, 26 showed the formulary within a community rule; 1QS I, 18–II, 18 illustrated a liturgical use of the formulary; and in Barnabas the form had been appropriated for preaching.257 The Bundesformular, however, did not explain the entire form of the Didache, nor, as later critics pointed out, did it really match the structure of the Two Ways texts. In a study on the virtue and vice lists of early Christianity, including the Two Ways texts, Ehrhard Kamlah disputed Baltzer’s conclusions on the Two Ways and traced the form instead to a dualistic Persian myth that he reconstructed

ajpostovlwn was added to the second title (La Didachè, 101–2). For further details on AUDET’s position, see below pp. 109, 115. 253 Ibid., 249. He also added Col 4:9–12; Eph 4:1–6.17; and 2 John 4–11. 254 Ibid., 250. 255 Ibid., 250–51. 256 Ibid., 251. 257 KLAUS BALTZER, The Covenant Formulary (trans. David E. Green; Oxford: Blackwell, 1971), 177.

1.3 The Didache as Church Order – Early Research

51

from several other attestations.258 In 1972 M. Jack Suggs expanded Kamlah’s study and concluded that the Two Ways texts showed a development away from the original mythological motifs.259 Of the three Christian witnesses to the Two Ways, the text of Barnabas was closest to the original myth while the Didache attested the latest redaction of the form.260 In addition, beginning from Kamlah’s observation that the New Testament catalogs were often associated with baptismal language, Suggs showed that the Two Ways form or genre had its original (though not exclusive) Sitz im Leben in initiatory settings.261 Rordorf reached conclusions similar to Suggs but attributed the differences in the Two Ways witnesses to the influence of Iranian religion on one segment of the ethical tradition found in the Bundesformular.262 This resulted in both dualistic and non-dualistic Two Ways traditions existing simultaneously, both finding their way into Christianity. Then in 1981, Mary Margaret McKenna published a dissertation that surveyed the occurrences of Two Ways imagery in texts of the Graeco-Roman period and addressed the question of whether the Two Ways could be classified as a Gattung, which she defined as an oral form, or as a genre, i.e., a literary form.263 Through her extensive analysis of these texts she concluded that the Two Ways functioned as a Gattung in its early existence and later became a literary genre. Though it was related to the covenant formulary, its main function was a call to repentance and thus could be appropriated into other contexts. Like Kamlah, Suggs, and Rordorf, she discerned two “trajectories” in Two Ways texts, which she called “Persian” and “Greek.” The former type, according to McKenna, was seen primarily in the reactionary texts related to Qumran; the latter, more common form found continuity within the wisdom/scribal tradition.264 She believed that the Sitz im Leben of the common source of the Two Ways in Didache, Barnabas and Doctrina was an ethical teaching or rule used for initiates, possibly gentile proselytes, into a distinct Jewish religious community 258 EHRHARD KAMLAH, Die Form der katalogischen Paränese im Neuen Testament (WUNT 7; Tübingen: Mohr [Siebeck], 1964). KAMLAH reconstructed the myth from 1QS III, 13–IV, 26; Philo QE 1.23; and Plutarch Is. Os. 46–47. 259 M. JACK SUGGS, “The Christian Two Ways Tradition: Its Antiquity, Form and Function,” in Studies in the New Testament and Early Christian Literature: Essays in Honor of Allen P. Wikgren (ed. David Aune; Leiden: Brill, 1972), 60–74. 260 Ibid., 70–71. 261 Ibid., 72. SUGGS uses “form” and “genre” interchangeably. 262 WILLY RORDORF, “An Aspect of the Judeo-Christian Ethic: The Two Ways,” in The Didache in Modern Research (ed. Jonathan A. Draper; Leiden: Brill, 1996), 148–64; trans. by J. A. Draper of “Une chapitre d’éthique judéo-chrétienne: les deux voies,” RSR 60 (1972): 109–28. 263 MARY MARGARET MCKENNA, “ ‘The Two Ways’ in Jewish and Christian Writings of the Graeco-Roman Period: A Study of the Form of Repentance Parenesis” (Ph.D. diss., University of Pennsylvania, 1981). 264 Ibid., 386–87.

52

Chapter 1: The Didache and the Question of Genre

that included gentiles and that seemed to have had a connection with the Essenes.265

1.4 Studies on Composition, Communication, and Genre This survey of research on the Didache and its component Two Ways text shows that the genre of the work is indeed still in question. Its overall similarities with secular as well as both Jewish and Christian texts, along with its use of earlier sources and its later function as a source itself, all warrant a deeper investigation into the genre issue. Such a study must be carried out both on a synchronic level, in order to understand the way in which the parts fit together within the context of the work as a whole, and on a diachronic level, in order to see its later development, to see what kinds of information came to be considered “didache” for the Church. One method that has been fruitful for uncovering the genre in texts of this period is the text-linguistic method demonstrated by David Hellholm on apocalyptic writings. Although it is a method that, like others in the study of human language, is somewhat weighted down with technical terminology, Hellholm's text-linguistic approach is, in actuality, simply the application of linguistic concepts to the analysis of the literary structure of a text. Such a study can provide both a more detailed look at the mechanics of an individual text and a means of comparison with other, similar texts. Of course some type of structural presentation of the Didache, in more or less detail, has virtually always been included in the commentaries. From a communications standpoint, the recognition of a text’s structure is essential for the understanding of the information that the text is seeking to impart to its audience – it’s an important aspect of successful communication. And, as with most texts, the Didache does have some easily recognizable structural indicators on its surface level. The issue of exactly how the Didache works syn265

Ibid., 454–55. In her study MCKENNA posited and employed a broader concept of Essenes that included the community represented by the sectarian texts of the Dead Sea Scrolls as well as other sub-groups such as those described by Josephus and Philo and that was connected with texts such as the Testaments of the Twelve Patriarchs. In her opinion, this could account for the clear relationship, yet significant differences, between the Dead Sea texts, the Testaments, and the Two Ways source of Didache, Doctrina, and Barnabas. See, for example, “Two Ways,” 123–27; 384–85; 458–60. Note that VAN DE SANDT, in his extensive work on the Two Ways, also suggests a literary connection between the “Greek Two Ways,” (i.e., his reconstruction of the common text underlying the Didache, Barnabas, and the Doctrina) and the Two Ways docrine of IQS as well as a correspondence of both with the Testaments of the Twelve Patriarchs and the early Derekh Eretz materials. He locates the source of these traditions “on the border where the fringes of the Essene movement met the ‘left wing’ of the Pharisaic circles” (Didache, 180–81).

1.4 Studies on Composition, Communication, and Genre

53

chronically, however, is especially relevant today, given a current trend to reject the idea of compositional stages and to see the text instead as the work of one author, created with or without the use of outside sources. An important example of such a trend is found in the commentary by Milavec.266 On the one hand, Milavec begins from the same premise as a text-linguistic study, namely that the Didache must be examined synchronically in order to discover how its components function together as a whole. This is of special interest for him because it is his opinion that the text represents the written version of an originally oral training program for gentiles267 emanating from the middle of the first century C.E.268 As a training manual, it was not, in its oral form, the creation of one author, but, rather, was the tradition of a community. Only later and for unknown reasons was it put into writing.269 Thus, parallel to the Greek text, which he bases primarily on H54, he presents in English an “analytical translation,” i.e., he identifies the “units of thought, repetitious patterns, and key transitions” in order to show the text’s “internal logic.”270 He rightly notes that the authors of the text “made ample use of topic sentences (or phrases) in order to signal the beginning of new sections” and “used summary statements (4:14b; 13:1f.; 15:4) and final cautions (4:12–14a; 6:1f; 266 Along with SCHÖLLGEN (see above pp. 10–11 and n. 27), NIEDERWIMMER also sees only one stage of composition (Didache, esp. 42–52), though he does acknowledge the existence of later textual “glosses,” (43; see also “Textprobleme der Didache,” WSt 95 [1982]: 114–30, where he discusses specifically Did.1.4a; 2.5b; 10.5; 10.7; and 13.4). VAN DE SANDT too considers the Didache to be the creation of one editor, albeit using much older sources (Didache, 28–48). Additional supporters of a single author/editor include UMBERTO MATTIOLI, Didachè: Dottrina dei dodici apostoli (3d rev. ed.; LCO 5; Rome: Figlie di San Paolo, 1980), esp. 21–29; MICHELLE SLEE, The Church in Antioch in the First Century C.E.: Communion and Conflict (JSNTSup 244; London: T&T Clark International, 2003), esp. 160; WILLIAM VARNER, Way of the Didache, esp. 58; and GARLEFF, Urchristliche Identität, 92. 267 MILAVEC claims that various “clues” within the Didache indicate that it was originally an oral composition. See MILAVEC, Didache: Faith, Hope, & Life, esp., xxxii–xxxiii; 57–58; 718–19. The evidence he provides there, however, is not persuasive. Among other things, he does not take into account: 1) that the fact that written texts of this period were read aloud, even by a solitary reader, does not mean that the text was originally oral; 2) that word can refer as well to written communication; 3) that a text may recount spoken rites without itself having been originally oral (against his citing of proeipovnte" in Did. 7.1); and 4) that verbs originally oral in meaning can sometimes be used with reference to written texts (against his citing of proeirhmevna in 11.1; see “proei'pon,” BDAG 867–68, 2b). 268 MILAVEC’s dating of the written text is unclear to me. On the one hand he states that the Didache was “written in the generation following the death of Jesus, when the message of Jesus was not yet encapsulated in stories about Jesus,” (Didache: Faith, Hope, & Life, vii); yet he also believes that the text was originally an oral composition, only written down at a later time. It is unclear whether his mid-first century dating applies to the oral form only or to the written text as well. 269 Ibid., vii; 57. 270 Ibid., 9.

54

Chapter 1: The Didache and the Question of Genre

11:1f.) in order to bring closure to blocks of material.”271 From these “linguistic clues” he concludes that the text essentially is composed of five parts that exhibit a deliberate progression: I. II. III. IV. V.

Training Program in the Way of Life (Did. 1.1–6.2) Regulations for Eating, Baptizing, Fasting, Praying (Did. 6.3–11.2) Regulations for Hospitality/Testing Various Classes of Visitors (Did. 11.3–13.2) Regulations for First Fruits and for Offering a Pure Sacrifice (Did. 13.3–15.4) Closing Apocalyptic Forewarnings and Hope (Did. 16.1–8).272

In his commitment to its compositional integrity, however, Milavec does not, in my opinion, adequately note or explain major structural incongruities in the text. One example of this is his understanding of the passage Did. 6.3–7.1: 6.3 Peri; de; th'" brwvsew": o} duvnasai bavstason: ajpo; de; tou' eijdwlolquvtou livan provsece: latreiva gavr ejsti qew'n nekrw'n: 7.1 Peri; de; tou' baptivsmato": ou{tw baptivsate: tau'ta pavnta proeipovnte": baptivsate eij" to; o[noma tou' patro;" kai; tou' uiJou' kai; tou' aJgivou pneuvmato" ejn u{dati zw'nti: Concerning food, bear what you are able but be especially on guard against meat sacrificed to idols for it is the worship of dead gods. Concerning baptism, baptize in this way: After first stating all these things, baptize in the name of the Father and of the Son and of the Holy Spirit in living water.

That this passage is transitional in character is widely acknowledged. On the one hand, the introductory phrase peri; dev in both 6.3 and 7.1 signals to the reader that a new topic will be addressed.273 Yet with the instruction presented in 7.1 there begins also a change from the second person singular verb forms of 6.1–3 (predominant in the Two Ways in general) to second person plural forms. Moreover there is an unexpected reference in 7.1 seemingly back to the entirety of the preceding material (tau'ta pavnta proeipovnte").274 Thus most scholars view 7.1 as beginning, however awkwardly, a new section of the text and not merely a new topic. On the other hand, 6.1–2 seems to be a conclusion to the Two Ways: 271

Ibid, 58. Ibid., 58–59. This is a revision of his earlier work in “The Pastoral Genius of the Didache: An Analytical Translation and Commentary,” in Religious Writings and Religious Systems: Systemic Analysis of Holy Books in Christianity, Islam, Buddhism, Greco-Roman Religions, Ancient Israel, and Judaism, Volume 2, Christianity (eds. Jacob Neusner, Ernest S. Frerichs and A. J. Levine; BrSR 2; Atlanta: Scholars Press, 1989), 101. 273 See below, pp. 135–36. 274 In text-linguistic terms, this happens through the combination of a verb of communication proeipovnte" with the demonstrative pronominal phrase tau'ta pavnta, which clearly stands for all of the previous material. Because this statement is not part of the previous material but refers to it as an independent entity, the level of communication has changed from that of a teacher instructing catechumens (in 1.1–6.3) to that of the author instructing the community. See below pp. 84, 138. 272

1.4 Studies on Composition, Communication, and Genre

55

6.1 {Ora mhv ti" se planhvsh/ ajpo; tauvth" th'" oJdou' th'" didach'": ejpei; parekto;" qeou' se didavskei: 6.2 eij me;n ga;r duvnasai bastavsai o{lon to;n zugo;n tou' kurivou tevleio" e[sh/: eij dΔ∆ ouj duvnasai o} duvnh/ tou'to poivei: Watch that no one cause you to stray from this way of teaching for such a person teaches you apart from God. For if you are able to bear the entire yoke of the Lord, you will be perfect; but if you cannot, do what you can.

Here the phrases tauvth" th'" oJdou' th'" didach'" and, in my opinion, o{lon to;n zugo;n tou' kurivou clearly refer back to the preceding Two Ways text and bring a closure to the instruction.275 Thus one is faced with the problem of explaining the awkward position of 6.3, a statement that is simultaneously part and not part of the Two Ways, an incongruity that is, however, only minimally represented in Milavec’s analytical translation via the placing of umlauts over the primary vowels of the plural verbs beginning in 7.1.276 While many scholars believe 6.3 to be a secondary addition to the original catechism represented by the Two Ways, Milavec is convinced that it is part of the “deliberate progression” of the text and that the instruction on food is found “in just that place where it would be expected to be introduced to the candidates for baptism – not earlier and not later.”277 Thus, acknowledging that 6.1–2 concludes the Two Ways, Milavec sees the next major section of the text as beginning with 6.3 rather than with 7.1. In his reconstruction of the authors’ intentions, he believes that the prohibition on foods offered to idols was put off until this point in the training, i.e., just prior to baptism, because the home environment of many gentile converts would have required participation in meals in some way connected to pagan sacrifice278 and to prohibit them from doing so “would have created an unnecessary hardship.”279 Following baptism, in his reconstruction, the catechumens would, however, have been allowed to join in the community meals, thus resolving the problem. The position of this instruction also coincides, in Milavec’s view, with the pre275 See below pp. 70; 84; 139; 184–85. This is in contrast to the view that “yoke” refers to the entirety of the Law of Moses, for which see especially the argument of VAN DE SANDT, Didache, 238–70. In my opinion VAN DE SANDT’s misunderstanding of the term forces him to see the criticism of the “hypocrites” in Did. 8.1, 2 as evidence of a radical change in the community toward “an attitude of animosity to Jews and Judaism” (296). 276 Thus, for example, “bäptize” (Didache: Faith, Hope, & Life, 29). There is no other indication of the change in communication levels that has taken place. 277 Ibid., 237. 278 MILAVEC states, “Of necessity … most candidates would have been constrained to take part in family meals wherein, either regularly or periodically, some offering was made to the household gods as part of the meal or some portion of the meats served had been previously offered at a public altar,” (Didache: Faith, Hope, & Life, 100) and further that a convert might have to “tolerate his father’s pious habit of dedicating every principle meal to the household gods,” (237); see also pp. 249–253. 279 Ibid., 237. On the idea that the Didache was addressing the hostile environment in which catechumens resided, see also pp. 112–16; 741–68.

56

Chapter 1: The Didache and the Question of Genre

baptismal fast commanded in Did. 7.4, an act that he maintains served to rid the body of the catechumen of prohibited food.280 The change to a plural verb form in 7.1 is explained, according to Milavec, by the fact that the true audience of the text has finally been revealed – the intended audience of the Two Ways was never really the catechumens, as one might have assumed, but rather the mentors of the catechumens. He states: The training in the Way of Life is offered in the second person singular because the trainers need to hear the specific rules in the form in which they will use them to train their novices.281

In other words, the Two Ways of the Didache provided teachers with the verbatim instruction they were to teach the initiates. After the directive of 6.3 the verb form changes to plural, in Milavec’s opinion, because the instruction regarding the performance of the baptismal rite was, of course, intended only for the teachers who would actually carry out the rite. Milavec’s attempt to rescue the integrity of the text at this point is unconvincing. Leaving aside the question as to whether one can assume with him that the majority of gentile catechumens were living in circumstances where the only food available was connected with idolatry,282 one is struck at how 280

Ibid., 100, 237. Ibid., 245. 282 Briefly, there are two problems with MILAVEC’s portrayal of that situation. First of all, in the Roman tradition at least, it was not the entire meal but only a portion that was offered to the household gods, this being thrown into the hearth (see, for example, LESLEY ADKINS and ROY A. ADKINS, Handbook to Life in Ancient Rome, updated edition [New York: Facts On File, Inc., 2004], 307–8; FRITZ M. HEICHELHEIM, CEDRIC A. YEO, and ALLEN M. WARD, A History of the Roman People, [2d ed.; Englewood Cliffs, N.J.: Prentice-Hall, 1984], 42). In this scenario one would not be eating the food sacrificed to the household gods. MILAVEC, however, provides no data on Gentile meal customs to support his statements. Secondly, while the late Hellenistic Jewish/early Christian term eijdwlovquton literally means “something offered to a cultic image/idol, food sacrificed to idols,” it is nonetheless often understood as specifically referring to meat, namely the meat left over from pagan temple sacrifices and made available for purchase in the public market (BDAG 280; see also FRIEDRICH BÜCHSEL, “eijdwlovquton,” TDNT 2:378–79, for additional information and references). Indeed “meat” seems clearly to be the meaning in 1 Cor 8 where Paul, in his discussion of whether it is appropriate for Christians to consume eijdwlovquton, uses kreva" (meat) as a synonym (8:13). This is supported also by the reference to the mavkellon in 1 Cor 10:25, which, according to C. K. BARRETT, “was mainly concerned with the sale of meat and fish” (see C. K. BARRETT, “Things Sacrificed to Idols,” NTS 11 [1965]: 144–46 for examples from ancient literature; for an opposing opinion see, however, PETER D. GOOCH, Dangerous Food: I Corinthians 8–10 in Its Context [SCJud 5; Waterloo, Ontario: Wilfrid Laurier University Press, 1993], 53–55). If “meat” is the correct translation in Did. 6.3, rather than MILAVEC’s more general “food” (Didache: Faith, Hope, & Life, 27; 237) it further speaks against the hardship he imagines since meat was expensive and consumed only rarely by most people in the Graeco-Roman world (see, for example, WAYNE MEEKS, The First Urban Christians 281

1.4 Studies on Composition, Communication, and Genre

57

contrary such a positioning of 6.3 would be to an effective communication of the message he envisions. At the very least, if the instruction in 6.3 is vital to the catechumens, so much so that prohibited foods must be evacuated from their bodies prior to baptism, why would it not be placed prior to 6.1–2, where it would benefit from the inclusion and protection of that passage while still preserving a final position? For such instruction to be intentionally set outside of the preceding oJdo" th'" didach'" (6.1) makes no sense. In addition, Milavec’s explanation of the movement from singular to plural verb forms in 7.1 does not account for the same phenomenon in 1.3 where plural verb forms occur within what is clearly intended to be instruction for the catechumens. Ultimately his explanation of 6.1–3 seems convoluted. A similar problem occurs with Milavec’s understanding of Did. 1.3b–6, also generally considered an interpolation. Not only do the commands change from singular in 1.2, to plural in 1.3, to primarily singular again beginning in 1.4 and continuing throughout the remainder of the Two Ways (with the exception of 4.11 and 5.2), but there is also a somewhat overlapping change from the occurrence of the future indicative and present imperative in 1.2 to strictly the use of imperatival forms in 1.3b–6, to the predominant use of the future indicative throughout Did. 2; 3.9–10; and most of ch. 4. Milavec does not deal with the change to plural in 1.3 and believes the alternation in mood to represent deliberate distinctions between commands that candidates were expected to follow immediately (i.e., those presented in the imperative [1.3b– 6]), commands that applied to all people (presented in the future indicative [2.1–3.6]), and commands that applied again to the candidates but only after full entrance into the Christian community (also in the future indicative [3.7– 4.14]).283 But this distinction between the function of the future indicative in the latter two sections is only made in order to resolve the problems that instructions concerning giving are presented twice, both in 1.4–6 and 4.5–8, and that the teaching in 1.6, cautioning against giving indiscriminately, seemingly contradicts the notion that one should give freely in 1.5 and 4.5–8. Milavec, however, sees the future indicative of 4.5–8 not as a type of imperative but simply signifying future expectations that are only to be implemented upon entry into the community. Yet nothing in the text indicates that these future verb forms are to be understood in any other way than in the same imperatival sense that he too admits is the meaning of the text from 2.1–3.6.284 Moreover [New Haven: Yale University Press, 1983], 98; JULIANA CLAASSENS, “Food,” NewIDB 2:473). 283 MILAVEC, Didache: Faith, Hope, & Life, 102. 284 Note that MILAVEC understands this future construction as only a “mild imperative” imitating the language of the Decalogue, (Didache: Faith, Hope, & Life, 102). In actuality the Hebrew aúl + imperfect of the Decalogue had the “strongest expectation of obedience,” GKC 107o. See also below, pp. 73–74; 158.

58

Chapter 1: The Didache and the Question of Genre

Milavec ignores the fact that while Did. 4.5–8 is attested in the Two Ways witnesses of Doctrina and Barnabas, Did. 1.3b–6 is missing, and that where it is found, namely in Matthew and Luke, it is the imperative that is used, as is the case in the Didache. Surely this indicates that the change in mood here is connected with the characteristics of the source rather than any deliberate authorial intent. Milavec believes in the textual integrity of the Didache because of the semantic connections he sees between its various text-parts285 but he does not give adequate value and consideration to structural markers that speak against integrity. This leads, in my opinion, to an inadequate portrayal of the structure of the text, an erroneous conclusion about its composition, and an unsound basis for the further task of determining a community setting for its creation. The problems of his study underscore the benefits of applying an already established text-linguistic method to the Didache, thereby allowing for a more detailed, consistent, and comprehensive description of the structure of the text while providing a basis for comparison with other texts and for scholarly critique and refinement. A more detailed study that also utilized the concept of orality to explain the composition of the Didache was published in 1992 by Ian Henderson with a subsequent article appearing in 1995.286 Henderson himself describes his first publication as arguing “that the symbolic, formulaic, and argumentative unity of the Didache can fruitfully be reconciled with its evident episodic and stylistic disunity if we recognize and understand the deliberate, oral quality of the text.”287 In Henderson’s view, the seeming repetitions and contradictions within the Didache as well as the abruptness of the text at certain points are not due to subsequent editorial redactions, as is often argued, but are characteristics of its “oral sensitivity.” Recurring symbols of “verbal, spoken authority,” in his opinion, create a unifying structure and communicate the purpose of the Didache, i.e., to “provide an interpretative framework within which to harmonize conflicting norms”288 and to “frustrate any interrogation about its specific tradition-historical loyalties and connections.”289 Henderson believes that the concept didachv, a term with definite oral connotations, is central for the text. In his view, “the literary relationship of didavskalo" (and terminological echoes) to implied author, to the notion of the text as didachv, and especially to the reader(s) as prospective mediator(s) of didachv, is a key to what 285

MILAVEC, Didache: Faith, Hope, & Life, 59. 286 IAN H. HENDERSON, “Didache and Orality in

Synoptic Comparison,” JBL 111 (1992): 283–306; “Style-Switching in the Didache: Fingerprint or Argument?” in The Didache in Context: Essays on Its Text, History & Transmission (ed. Clayton N. Jefford; Leiden: Brill, 1995), 177–209. 287 Idem, “Style-Switching,” 177. 288 Idem, “Didache and Orality,” 302. 289 Idem, “Style-Switching,” 178.

1.4 Studies on Composition, Communication, and Genre

59

the text is about.”290 More specifically on the subject of genre, Henderson agrees with others in classifying the second half of the Didache as a church order, which, in retrospect, permits one to describe the entire work as a church order in terms of its “Grossgattung.”291 In actuality, however, he believes that the text “emphasizes a more specific hybridization of older genres,”292 which is evidence of an “intermediate creative sensibility, combining positive characteristics of orality and literacy.”293 On the one hand, Henderson maintains that an emphasis on orality can be seen in the Didache in its conscious preference for verbs of speaking (levgw, lalevw, didavskw, keleuvw, ejlevgcw, eujcaristevw) over those of writing, this in spite of the fact that it makes use of written sources.294 For example, the formally indicated quotations in the Didache (according to Henderson, 1.6; 9.5; 14.3; 16.7) are “strongly marked by their exclusive use of levgw-citation formulas.”295 On the other hand he states that “the fluidity of the normative and text-linguistic boundaries between different authorities is an important sign not only of oral sensibility in Didache, but also of its argumentative tenor.”296 Thus, of the three Old Testament quotations in the Didache, he notes that one is “marginal in its scriptural authority” (1.6) and another “diminishes the distinction between divine and dominical speech, adding to the citation formula Didache’s ambiguous theo-/Christological symbol (14.3, uJpo; kurivou)” – together these signal that “while the OT’s distinctness is recognized, its distinctiveness as a text and as a source of normative authority is significantly qualified.”297 An example of such traits can also be found, in his view, in the introduction to the Lord’s Prayer in Did. 8.2, which states that one should pray wj" ejkevleusen oJ kuvrio" ejn tw'/ eujaggelivw (as the Lord commanded in the gospel) – in addition to the verb of speaking (ejkevleusen) there is once again a reference to kuvrio". Moreover, Henderson notes that here the “gospel” is both cited and reproduced in the text which follows, clearly intended to “emphasize normative complementarity – and to gloss over normative dissonance – between the Gospel and Didache itself, between both of these and the prayer itself as text, and between these texts and the voice(s) of the Lord.”298 He concludes from this that, 290

Idem, “Didache and Orality,” 287. 290, though he later states that he doubts that the Didache was meant to be read as a church order (“Style-Switching,” 189). 292 Idem, “Didache and Orality,” 290. 293 Ibid., 289. 294 Ibid., 295. 295 Idem, “Style-Switching,” 206. 296 Idem, “Didache and Orality,” 295. 297 Ibid., 296. 298 Ibid. This sense of harmonization and of a “blurring” of sources in order to give a sense of “normative complementarity” is also seen by HENDERSON in 11.3 (kata; to; dovgma 291 Ibid.,

60

Chapter 1: The Didache and the Question of Genre

the distinctiveness of literacy events is effaced so consistently and unexpectedly as to constitute a deliberate policy; sources, some of them certainly written, are formally acknowledged, but their normative and even their textual distinctiveness is denied. Scripture, Gospel, and even Didache itself (4.1; 7.1; 11.1–2) are together the semantically indivisible voice of the Lord, definitively heard within Didache’s implied readership.299

Indeed, according to Henderson, the use of the present tense and verb of possession in Did. 15.3, 4 (wJ" e[cete ejn tw'/ eujaggelivw/) shows that it is the readers who are given the administrative and interpretative authority within the community.300 Thus, “by implicating the reader in its own oral sensibility, Didache suggests the possibility of creating or re-creating an atmosphere in which the variety of normative language can be simultaneously and cumulatively, not hierarchically, maintained.”301 In Henderson’s view, understanding the oral character of the Didache allows one to appreciate its unified literary structure created by the use of “conversational techniques of accumulation, juxtaposition and climax.”302 Such techniques allow the simultaneous presence of opposing norms, implying “their ultimate prescriptive harmony”303 and explain, for example, the problematic phrase in Did. 1.4, oujde; ga;r duvnasai: in characteristically oral style, Didache resolves the tension between possibilism and perfectionism not by extended argumentation, but by repeatedly juxtaposing the two, here and in closely parallel contexts, to imply their ultimate prescriptive harmony (1.3b–6; 6.2, … tevleio" e[sh//: eij d'Δ∆ ouj duvnasai, [you will be perfect/but if you are unable]; … 13.5–7, … wJ" a[n soi dovxh/, do;" kata; th;n ejntolhvn [as it seems best to you/give according to the commandment]).304

Also important for the structure of the text, in his opinion, is its formulaic composition, involving the repetition of certain words/word groups and the use of structural formulae and parallelism. Such characteristics allow the

tou' eujaggelivou) and 15.3, 4 (wJ" e[cete ejn tw'/ eujaggelivw/ [tou' kurivou hJmw'n]). In the latter case, “arguably the work’s climax,” HENDERSON makes the interesting observation that “although Didache emphatically distinguishes ‘the gospel’ from itself, no help is offered toward distinguishing the implied author’s contribution normatively or textually from that of the source/model: the acknowledged debt is to the gospel’s authority, not to its text,” (297). 299 Ibid., 297–98. 300 Ibid., 297. On this see also STEIMER, Vertex, 297–99. 301 HENDERSON, “Didache and Orality,” 305. 302 Ibid. 303 Ibid., 301. As an example, HENDERSON cites here Did. 1.3b–2.1 where “pragmatic grounding (oujc e[xete ejcqrovn [you will not have an an enemy], 1.3b; … oujde; ga;r duvnasai [for you are unable], 1.4) is argumentatively unconnected with the transcendent motive (e[sh/ tevleio" [you will be perfect], 1.4a; cf. 6.2), except implicitly in the next, climactic, cumulative step, to giving kata; th;n ejntolhvn [according to the commandment] (1.5, 6), where pragmatic considerations and eschatological imperative at last coincide (esp. 1.6).” 304 Ibid.

1.4 Studies on Composition, Communication, and Genre

61

Didache to be “understood nonsequentially.” For example, deutevra in Did. 2.1 does not denote a logical sequence, but a “symbolic, rhetorical” priority of ch. 1 over what follows; tau'ta pavnta in 7.1 also implies the tau'ta pavnta of 11.1, as well as the ejntola;" kurivou of 4.13, 14, and the pavsa" ta;" pravxei" … wJ" e[cete ejn tw'/ eujaggelivw/ tou' kurivou hJmw'n of 15.4;305 and the repetition of words in the field didask- “describes the whole relationship between the Teacher/Speaker and audience, so that 11.2 recalls formally the anticipatory note in 4.1 at the same time that it introduces the series of personal/functional definitions in 11.7–11.”306 Ultimately Henderson concludes, Didache argues for the complementarity of diverse authorities by receiving them conversationally and pragmatically rather than ideologically, as didachv/ejntolhv/dovgmav; the symbols chosen for this task are therefore preferentially and not only accidentally or subliminally those of speech and action rather than those of a more literary logic.307

Though similarly oriented, Henderson’s studies do not provide a thorough text-linguistic analysis of the structure of the Didache and the way that it communicates to a reader. In his brief exploration into the genre he rightly points to the importance of the term didachv, but given his working hypothesis that the text is a unified creation, he does not explore what he himself describes as a “progressive” defining of the term,308 and the genre question remains largely unanswered. Ultimately his resolution of the literary difficulties of the Didache by recourse to the presence of a conscious, oral sensibility is unconvincing. The preponderance of oral over literary terminology in the text can largely be attributed to factors other than a “normative hermeneutical strategy,”309 including the intent to provide specific prayers and ritual texts along with instruction for their use; the quotation of sayings of Jesus, who produced no literary texts; and ancient literary convention.310 Further, one 305 Ibid., 303. HENDERSON also states that, “tau'ta pavnta become explicit and actual only in the whole imagined conversation in which tradition, text, and readership all have decisive voices (4.1–2; 6.1–7.1; 11.1–3; 14.1–16.2),” (299). 306 Ibid., 301–4. 307 Ibid., 304. 308 I.e., in Did. 1.3; 2.1; 6.1; and 11.2, (“Didache and Orality,” 304). 309 Ibid., 305. 310 The “levgw-citation formula,” for example, is a common method of citing Old Testament passages and is found regularly in the past tense (as in the Didache) as well as in the present (see “ei\pon,” BDAG, 1e). (For instances of a parallel phenomenon outside Jewish and Christian writings, for example in Epictetus, see “levgw,” BDAG 588–90, 1bh). Contra HENDERSON (“Style-Switching,” 206) there seems to be no special strategy on the part of the Didachist in the use of this phrase, indeed HENDERSON himself notes that the phenomena is “an extraordinary convention among early Christian documents.” Moreover, in his second publication, HENDERSON removed Did. 7.1 and 11.1 as examples of this formula, preferring to call them “marked self-allusions” (loc. cit.). Note, however, that in 7.1 proeipovnte" is

62

Chapter 1: The Didache and the Question of Genre

questions whether tension between conflicting norms can be alleviated simply by juxtaposing them (even within or alongside an already authoritative text like the Two Ways311) and citing a variety of authorities in order to diffuse the situation and give the sense that the issues remain “open.” In actuality the text warns against other teachings and sham authority figures (6.1; 11.1); provides a verbatim baptismal text and eucharistic prayers, in the case of the latter allowing an exception only for prophets (10.7); and calls for “perfection” in the end time (16.2). If its intent is to provide a dialogue on problematic issues, this does not seem clearly communicated within the text. In addition, Henderson’s view that each part of the Didache presupposes the text as a whole would imply that the audience must know all of the Didache before they can understand any part of it. Indeed Henderson states that “there is nothing in this didachv that the reader does not already know.”312 This seems to presume far too much in an effort to preserve the integrity of the text as a literary creation of one author. Although Henderson at times mentions many of the same “markers” that will be presented in the following text-linguistic study, he does not use these markers to show the overall structure of the text or the progression (and interruption) of communication from beginning to end. In comparison to the work of these two scholars, a thorough and more systematic text-linguistic study does, I believe, yield a better understanding of the structure of the Didache in several ways. First of all it brings to the text a system of linguistic analysis, including concepts and terminology, developed outside of Christian studies, and previously shown to be useful for the comprehension of texts, both Christian and secular. Although, admittedly, using any system of analysis to some extent predetermines one’s results, such a study will nonetheless provide a new set of data towards the resolution of the genre problem. These data are more detailed in their explanation of how the Didache communicates its message to its intended audience than was the case in earlier studies. In addition, they provide information regarding another dimension of the text in that they delineate its structure not only from beginning to end but also from the surface level to deeper levels of communication. The result of all this is a clearer, more in-depth picture of the Didache as an indisimply a circumstantial participle referring literally to the recitation of tau'ta pavnta “all these things,” i.e., the text of chs. 1–6, by community members performing a baptism, while ta; proeirhmevna in 11.1 is an attributive perfect passive participle used as a substantive, a form commonly found in ancient literature to refer to something previously stated in a written text (see “proei'pon,” BDAG 867–68, 2b) – one wonders how much deliberate authorial strategy should be seen in what appears to be conventional usage. 311 Regarding the Two Ways in the Didache, HENDERSON states that “its implied authorship and its imperatival and protreptic address are purposefully borrowed to bias reception of the public wisdom of chs. 7–15,” (“Didache and Orality,” 291). 312 Ibid., 292.

1.4 Studies on Composition, Communication, and Genre

63

vidual text and a better ability to compare its structure with other, similar works. Indeed, his interest in the topic after discussions early in my dissertation work led Hellholm himself to publish a brief text-linguistic study of the Didache in which he explores the composition of the text and includes a structural outline of it in an appendix.313 Hellholm makes a strong case for such a study in that, while others have suggested various structural overviews of the Didache, they do not deal with either the relationship of the text-parts to one another, nor with the rationale for their organization within the text as a whole, data which would provide additional and crucial information regarding the early Christian author and intended readers of the text. To truly understand the text, according to Hellholm, one must consider not only the pre-existing sources, the authorial redaction of these sources, and original authorial contributions (aspects, one must add, that are often presented by studies on the Didache), but also the data provided by the various extant witnesses to the text and the text’s own internal structural markers.314 Thus, based on information gleaned from each of these areas, Hellholm presents the Didache as a product of an author/editor who collected, redacted, and supplemented various pre-existing sources to form a “church order” (kyrkoordning) for a particular community. According to Hellholm, the sources included an originally Jewish (but already Christianized315) Two Ways text (represented in chs. 1–6); instructions for performing the rites of baptism and Eucharist (chs. 7–10); and a “church order,” in the narrow sense of the word (at least 11.4–12316). By rendering the sources in this new, unified form, Hellholm sees the author as having transformed the Two Ways source into a Christian baptismal catechism (made clear from the reference to the Two Ways material in the instruction on baptism in Did. 7.1 and emphasized by the inclusion of gospel material in 1.3b–2.1)317 and as having extended the narrower meaning of church order (originally applicable to 11.4–12) to include also the baptismal catechism and the instructions for the performance of baptism and Eucharist.318 The author, however, was not simply a compiler – in some places there is evidence of redactional activity, in others original material has been added by the author himself. In agreement with Schöllgen, Hellholm sees this new “church order” as created specifically for the needs of a particular community: neither the 313 HELLHOLM, “Kristen dopkatekes,” 109–39. While the structural outline presented in his article had been shared with me in an earlier exchange of outlines, I was unaware of his final publication until my own analysis was completed. 314 Ibid., 111–12. 315 Ibid., 115–16. See above, p. 28, n. 127; and below, p. 156, n. 61. 316 Ibid., 128, following NIEDERWIMMER, et al. 317 Ibid., 123–27. 318 Ibid., 130.

64

Chapter 1: The Didache and the Question of Genre

catechism of chs. 1–6 nor the text as a whole should be considered as comprehensive in scope.319 In this article, Hellholm is certainly successful in his attempt to illustrate the importance of including an investigation of the markers by which a text is structured as part of its historical study.320 Indeed, at many points Hellholm’s structural analysis concurs with my own (specific agreements and disagreements will be discussed in detail below). Yet it is also the case that the two analyses differ in significant, and thus important, ways for, as Hellholm’s study has shown, the discernment of the structural markers within a text is a crucial element in reaching an understanding of its historical origins and function. On the one hand, Hellholm’s analysis sees only one stage of composition for the Didache – my analysis will challenge that view and make the case that the text has undergone several periods of redaction. Secondly, his study is subject to the same criticism as that of Schöllgen, namely that it is hindered by an anachronistic application of the genre church order to the Didache.321 This presupposition leads him to envision the joining of Did. 1–10 with the source found in 11.4–12 as a development of an earlier and more narrow concept of church order represented by the latter. Yet the instruction on the discernment of apostles and prophets found in 11.4–12 and the material of the remainder of chs. 12–15 are clearly different in content, function, and even in form from chs. 1–10, whose primary purpose seems to be to provide instructions and authorized texts for the performance of church rituals. Indeed the connection between the two larger sections is only made via the warning against false teaching in 11.2. To see the preceding chapters as dependent on the genre of 11.4–12 is, I believe, unwarranted and in my analysis it will be argued that the genre of the Didache finds its origin instead in the Two Ways text of chs. 1–6 and the subsequent development of the concept of didache. Indeed, Hellholm does not explore the ancient genre of the text: he refers to the Didache only as a “church order” and a generic understanding of the term didache is implied only in the appended structural outline where the individual sources of the text are referred to as didacaiv.322 While Hellholm makes an important first attempt at understanding the form and function of the Didache, the need for a more complete study is certainly warranted.

319

Ibid., 125, 127, 130, 132. Ibid., 132. 321 For this see also DAVID HELLHOLM and VEMUND BLOMKVIST, “Parainesis as an Ancient Genre-Designation: The Case of the ‘Euthalian Apparatus’ and the ‘Affiliated Argumenta’,” in Early Christian Paraenesis in Context (eds. Troels Engberg-Pedersen and James M. Starr; BZNW 125; Berlin, New York: de Gruyter, 2004), 469. 322 HELLHOLM, “Kristen dopkatekes.” This designation is only found in the appended hierarchical text analysis beginning on p. 132. 320

Chapter 2

A Text-Linguistic Analysis of the Didache 2.1 Text-Linguistic Methodology 2.1.1. Introduction The field of text linguistics (and its American relation, discourse analysis) essentially refers to the application of linguistic concepts and methodologies to the text as an act of communication. Originally, linguistics dealt with the smaller units of language and communication using descriptive or structural approaches. On the one hand, language was described in terms of sounds (phonemes), individual signs (graphemes), meaningful groups of letters (morphemes) and words (lexemes). Later, structural methodology dealt with the relationship, the syntax, of words on the level of the phrase, clause, and sentence; the various “slots” or components of such expressions; and the transformation from one structure to another. Neither of these methods, however, reached beyond the level of the sentence and it was only with work beginning in the 1950s and expanding in the 1960s–1970s that the focus of linguistic research turned also to the text, i.e., a more extensive communicative entity. Text linguistics actually encompasses a variety of methods that have as their common basis the study of the text as the primary unit. Within a text, the linguistic aspects mentioned above, i.e., micro-structural components, are still important, but also necessary are such aspects as overall theme, societal context, and function. Most importantly, a text-linguistic analysis is text-centered, i.e., it starts with a specific text to see how that individual text communicates to its audience. Moreover, beyond rhetorical methodology, which evaluates a text against a given set of rules, text-linguistic methods allow for the individuality of an author and presuppose that a given text is comprehensible, that it communicates. A text-centered analysis “is necessary in order to be able to judge the extent to which conventional influence has controlled individual expressions or to which the individual author has bent and shaped the conventional into something new and appropriate to the situation in hand.”1 To better understand text-linguistic methodology, it is useful to start with the definition of text as developed by Robert de Beaugrande and Wolfgang 1 BRUCE C. JOHANSON, To All the Brethren: A Text-Linguistic and Rhetorical Approach to 1 Thessalonians (ConBNT 16; Stockholm: Almqvist & Wiksell International, 1987), 6.

66

Chapter 2: A Text-Linguistic Analysis of the Didache

Dressler in their 1981 introduction to the methodology.2 A text is a communicative occurrence that meets seven aspects and standards of textuality. From the perspective of the text (the text-centered realm): 1) Cohesion – the ways in which the components of the surface text are mutually connected in a sequence, including and especially grammatical dependencies. Cohesion is basically concerned with the syntax of the surface text. 2) Coherence – the ways in which the components of the textual world, that is, the configuration of concepts and relations which underlie the surface text, are mutually accessible and relevant. Such relationships include, for example, those of a causal, temporal or final character.3 From the perspective of the author/audience (the user-centered realm), the text must show: 3) Intentionality – the purpose of a text in that it is produced to achieve a certain goal. 4) Acceptability – the assumption on the part of the receiver that a text is cohesive, coherent and purposeful. 5) Informativity – the need that a text offer some new information. 6) Situationality – the factors which make a text relevant to a situation of occurrence. 7) Intertextuality – the dependency of a text on the receiver’s knowledge of other texts for successful communication.

Understanding these as “constitutive principles,” de Beaugrande and Dressler added three more concepts which they considered “regulative principles”: efficiency of communication on the part of the participants, effectiveness of the text on its receivers, and appropriateness of the characteristics of the text in terms of its setting.4 In the early period of text linguistics, the emphasis was on the text as a unit just beyond the sentence and on the structural description and classification of such units as occurred. But as de Beaugrande and Dressler point out, “the descriptive method eventually breaks down in the face of complexity … and open systems.”5 The fact that language is dynamic and entails virtually unlimited possibilities prohibits complete description and prediction. Even when transformational grammar was applied to the text, demonstrating the interdependence of sentences in communicating meaning, the results were still largely descriptive rather than explanatory.

2 ROBERT DE BEAUGRANDE and WOLFGANG DRESSLER, Introduction to Text Linguistics (London: Longman, 1981), 3–11. 3 HELLHOLM defines cohesion and coherence somewhat more narrowly. Looking at it from the point of view of individual text-parts, he understands cohesion as having to do with the semantic content of a text-part while coherence deals with its argumentative function visà-vis the text. These, along with connectivity, i.e., matters of text-syntax, are the aspects that create the continuity between a text-part and what precedes and follows (“Die Argumentative Funktion von Römer 7.1–6,” NTS 43 [1997]: 385). 4 DE BEAUGRANDE/DRESSLER, Introduction, 11. 5 Ibid., 23.

2.1 Text-Linguistic Methodology

67

An important development was Roland Harweg’s concept of substitution as an element that makes a text cohesive.6 Substitution is simply the bringing to mind of another, usually previous, expression using a different form, for example, a pronoun, synonym, or relationship such as class/instance, cause/ effect, part/whole. Because such a substitution refers to something either forward (cataphoric) or backward (anaphoric) in the text, it helps to hold the text together structurally and semantically. Important progress in the understanding of how a text actually produces meaning came with the work of Teun van Dijk for whom the generating of a text began with a main idea and involved the expression of this idea in ways in which it could be understood by its audience.7 By means of the overall structure, the macro-structure, of a text, certain operations such as deletion, generalization, and construction create communication by allowing the meaning to be extracted by the reader. Van Dijk’s model of text is thus processoriented; instead of being only descriptive, it analyzes how the communication actually takes place. In collaboration with Walter Kintsch, Van Dijk also learned through experimentation that readers of texts bring to their reading prior knowledge of macro-structure based on their real-world experiences, another aspect of how a text communicates. On this foundation, de Beaugrande and Dressler advocated what they termed the “procedural approach,” the study of “the operations which manipulate units and patterns during the utilization of language systems in application.” In their mind, “a text cannot be explained as a configuration of morphemes or sentences,” rather “morphemes and sentences function as operational units and patterns for signaling meanings and purposes during communication.”8 The concept of “rules” dictating minimum requirements for a text is replaced by them with a listing of characteristics whose measurement will give an indication of how well an expression fulfills its intent, how efficient, effective, and appropriate it is. A text is “a document of decision, selection and combination,” where the “occurrences are significant by virtue of the other alternatives which might have occurred instead.”9 The model of de Beaugrande and Dressler involves the performance of roughly sequential (though certainly overlapping) phases by the text producer: planning (setting a goal), ideation (collecting of ideas to express the goal), development (to expand, modify, and connect the ideas), expression, and parsing (the actual committing of the expressions into grammatical dependencies in a linear for6 ROLAND HARWEG, Pronomina und Textkonstitution (Munich: Fink, 1968). Discussed in DE BEAUGRANDE/DRESSLER, Introduction, 22. 7 For a discussion and reference to the pertinent articles of van Dijk see DE BEAUGRANDE/ DRESSLER, Introduction, 26–27. 8 DE BEAUGRANDE/DRESSLER, Introduction, 33. 9 Ibid., 35

68

Chapter 2: A Text-Linguistic Analysis of the Didache

mat for the surface text). The recipients of the text then perform a reversal of this process to retrieve the main idea behind the text. It is this “reversal” process that is described and analyzed by text linguistics in order to better understand how the text actually communicates to an audience. 2.1.2 Text Linguistics and Early Christian Literature The application of text-linguistic methodology was first performed on an early Christian work by David Hellholm. In the first part of a two-part study on Visions 1–4 of the Shepherd of Hermas, Hellholm performed a textlinguistic analysis in an attempt to determine the constitutive elements of the genre apocalypse. Research into this genre by the Apocalypse Group of the SBL Genres Project had earlier explored the problem of distinguishing among apocalypses, apocalyptic texts as component parts of other texts, and apocalyptic imagery. In a major step forward, the group provided a “masterparadigm” of the “prominent recurring features” found in apocalyptic texts, based on a survey of all texts that had been, or could be, classified as apocalyptic from the period of 250 B.C.E.–250 C.E from the area of the eastern Mediterranean.10 With this paradigm the group was then able to determine the features that were shared among all apocalypses and, conversely, what features separated this genre from other, similar writings. Hellholm contended, however, that there were two major flaws in the master-paradigm. First, although the group had presented its work as a first step towards an understanding of the genre apocalypse and stated that it did not intend to take up the issues of social setting or function,11 Hellholm believes function to be an essential component of any genre and consequently that it is necessary to include function in any such paradigm. A genre, he maintains, like any other human artifact, must have not only content and form, but function as well. This, he states, is not simply a text-linguistic concept, but is corroborated by the fact that literary critics and New Testament scholars have also come to similar conclusions independently.12 The choice by an author of a specific genre in which to present his/her message in and of itself has a meaning, an overall function, the specifics of which are then related in the text. Secondly, while Hellholm supported the group’s attempt at a hierarchization of the features included in the paradigm, he believed that further refinement was necessary in order to indicate the relationship of the features in terms of their level of abstraction: the fact that certain features could incorporate other, secondary features was not accurately represented in the para10

J. COLLINS, “Introduction,” 5. Ibid., 4. 12 DAVID HELLHOLM, “The Problem of Apocalyptic Genre and the Apocalypse of John,” Semeia 36 (1986): 20–21. 11

2.1 Text-Linguistic Methodology

69

digm.13 Finally, in addition to a paradigm of component features, Hellholm believes that a text-linguistic analysis is necessary to determine the generic structure of a text since it is the structure that helps to communicate the genre to an audience.14 Indeed, even in the case of the Didache where a general consensus on the structure of the text already exists, a text-linguistic analysis is necessary to show the relationship of the text-parts to one another and to the text as a whole.15 2.1.3. Delimitating a Text The text analysis presented here follows the method first outlined by David Hellholm in his study on Hermas, a system that was developed earlier by Elisabeth Gülich and Wolfgang Raible for narrative texts and then further modified and refined by Hellholm.16 Their model assumes an author (encoder), an audience (decoder), a shared universe, a text as an act of communication, and a more or less shared language among the parties, including shared concepts, such as the conventions of a generic code.17 In fact, within a text, the surface structure is a vital element of its communication to the audience in that it is generically specific, i.e., it reflects the intended genre of the text, either by more or less fulfilling the expectations of a stated genre, or, when such a statement is lacking, by indirectly informing the audience of the genre and arousing these expectations. For example, a letter is easily recognized as such without a title, but something entitled “epistle” that was missing a salutation would be somewhat puzzling to a reader. The overall structure of a text must be clearly recognizable for efficient communication. In an attempt to determine and describe this surface structure, Gülich/ Raible and Hellholm posit various text delimitation markers, i.e., signs within a text that serve to set off its component parts from one another while yet connecting these parts in an overarching structure. The markers are classified according to a three-level system, the pragmatic, semantic, and syntactic dimensions. These dimensions are hierarchical in that the most important signs for delimiting a text, those on the pragmatic level, govern those of the semantic and syntactic levels; likewise the semantic governs also the syntactic. 13

Ibid., 26–28 Ibid., 31. 15 See above pp. 62–64. 16 DAVID HELLHOLM, Das Visionenbuch des Hermas als Apokalypse: Formgeschichtliche und texttheoretische Studien zu einer literarischen Gattung, vol. 1, Methodologische Vorüberlegungen und makrostrukturelle Textanalyse (ConBNT 13/1; Lund: CWK Gleerup, 1980), especially 77–95. According to JOHANSON, the model of Gülich and Raible is subscribed to as one of the most useful for describing the most important, basic, textual processes and features of verbal communication (To All the Brethren, 7). 17 JOHANSON, To All the Brethren, 7. 14

70

Chapter 2: A Text-Linguistic Analysis of the Didache

Taken together, these markers will demonstrate the cohesion and coherence aspects of the text pointed to by de Beaugrande and Dressler. Thus on the pragmatic plane, one finds markers on the meta-level, i.e., words or clauses indicating that a new act of communication, a new communication situation (thus the term “pragmatic”18), is taking place within the text (and so they are text-internal markers). Such markers delimit the most essential parts of the surface structure of a text. First in importance are metacommunicative sentences, sentences that are comprised of a communicatorsubject and a verb of communication (speaking, hearing, writing, etc.). These sentences most often occur at the start of the text so as to introduce it, though sometimes they are found at the end of the text, indicating that it is concluded.19 One example from the Didache would be the clause ou{tw" proseuvcesqe which occurs at the beginning and end of the text of the Lord’s Prayer (Did. 8.2–3). Here the level of communication changes in that, while it is the author of the Didache who is introducing the text of the prayer, it is the audience of the Didache who are commanded to recite the prayer to God and who are thus envisioned as the speakers. The other type of delimitation marker on this level is the substitution on the meta-level, i.e., the occurrence of a word or phrase that stands for a specific text-internal act of communication, for example, “chapter,” “section,” “commandment,” etc. When one reads in Did. 6.1 the phrase tauvth" th'" oJdou' th'" didach'", it is clear that this stands for the instructions just delivered in the first five chapters.20 Again, such markers are said to be on the meta-level because they are communicating at one level above the actual text that they are introducing or concluding. Similar to this marker is a substitution on the abstraction level, which also stands for a portion of the text. An example of this would be the use of the pronoun au{th in Did. 1.2a. to denote the text that follows. Note that, as a pronoun, au{th has a wider range of possible meanings than simply the object for which it stands in any individual text situation, i.e., it is more abstract. This also means that, unlike the substitution on the meta-level, a substitution on the abstraction level does not stand completely outside of the text it represents. Thus according to Raible and followed by Hellholm, such substitutions are really at a position in between markers on the meta-level and the semantic level.21 Next in order are markers that signal a new text-part but remain on the same communication level with that text-part, not on a meta-level. Such 18

Ibid., 26. meta-communicative sentence is one which “funktioniert als Signal für den Beginn bzw. – allerdings seltener – das Ende eines Kommunikationsaktes,” HELLHOLM, Visionenbuch, 80. According to JOHANSON, however, they can also be reiterated within the text (To All the Brethren, 26). 20 See also HELLHOLM, “Kristen dopkatekes,” 116, 120. 21 HELLHOLM, Visionenbuch, 86–87. 19 A

2.1 Text-Linguistic Methodology

71

markers are semantic in nature and delimit a new part by referring to some change of circumstances in the “outside world” of the text (thus they are textexternal references). Examples of such markers would be changes in time or place (episode markers) or in the characters present within the text (dramatis personae). When these appear, they function within one sub-text to distinguish it semantically from other sub-texts. Again in the Didache, the phrase meta; de; to; ejmplhsqh'nai in 10.1 signals a change in time and helps to delimit this as a new text-part. Finally, the markers on the lowest level of delimitation are those that function to connect various text-parts syntactically. Unlike the semantic markers, they do not refer to some reality outside of the text, at least not directly, but simply connect or disconnect text-parts by their grammatical function. The first of these are instances of renominalization, i.e., markers that bring back into the text characters (or entities) previously introduced. This can be signaled through a variety of mechanisms, for example, through a change from indefinite to definite articles, through the use of pronouns, or through the creation of some other equivalent identifying expression – the means vary in different languages.22 An example from the Didache comes in the very opening of the text where miva th'" zwh'" in 1.1 becomes hJ oJdo;" th'" zwh'" in 1.2 and helps to delimit the beginning of a sub-section. Because an instance of renominalization depends for its meaning ultimately on the original introduction of the person (or entity) in view, it can have (only) an indirect relation to the text-external world. In contrast, adverbs and conjunctions are also syntactic markers that connect the various sentences and sections of a text but they function solely on the text-internal level, i.e., without any relation to the world outside of the text in which they are found.23 An example in the Didache would be the particle ou\n that occurs in 11.1 and 15.1 and functions to end the previous text-part while beginning a new one. The delimitation markers used by Hellholm can thus be summarized as follows: I. Pragmatic markers a. meta-communicative sentences b. substitutions on the meta-level II Semantic markers a. change in world structure b. episode markers (temporal, local) c. change in dramatis personae III. Syntactic markers a. renominalization b. sentence and text connectors24 22

Ibid., 94–95. Ibid., 95. 24 Ibid., 80–95. 23

72

Chapter 2: A Text-Linguistic Analysis of the Didache

As might be obvious, parts of a text can themselves be comprised of even smaller text-parts and consequently the communication of a text can occur on a number of levels within that text. In applying Hellholm’s text-linguistic method to the Didache, the question may be raised as to the appropriateness of using a method developed for narrative texts on a non-narrative work. Bruce Johanson dealt with this problem in his text-linguistic study on 1 Thessalonians. As Johanson points out, the method of Gülich and Raible was never limited by them to narrative texts and in fact their original study had been informed by non-narrative alongside narrative texts. Yet Johanson found it necessary to modify the set of text delimitation markers to incorporate data he found in 1 Thessalonians, due both to its text type and to individual characteristics of its author.25 Similarly, important differences from the markers as described by Gülich/Raible and Hellholm were also found in applying this method to the Didache, another non-narrative text. The markers on the meta-level, i.e., meta-communicative sentences and substitutions on the meta-level, do appear in the Didache, though the first are much less frequent than the latter. Important differences, however, show up on the semantic and syntactic levels. On the semantic level, the first marker, change in world structure (e.g., earth/heaven/hell), is not at all applicable to the Didache; with respect to episode markers, there are no local markers and few temporal ones in the text; and, finally, there are only subtle changes in dramatis personae, essentially changes in number (singular vs. plural forms), perhaps where different sources have been joined.26 On the syntactic level, the text has a modest number of examples of renominalization as this has been defined (though it does exhibit a type of repetition akin to this – see below). Of course it also employs both adverbs and conjunctions as connectors. An additional feature, however, that seems to delimit certain sections of the Didache is repetition. Hellholm states that repetition (“Rekurenz”) is a phenomenon that can strengthen a particular marker,27 but in the Didache two types of repetition seem to function as delimitation markers themselves. On the one hand, a term mentioned peripherally in one text-part is repeated in an adjoining text-part in order to facilitate a change in topic, a phenomenon commonly referred to as word association. Examples in the Didache are the 25 JOHANSON, To All the Brethren, 29–31, 32. 26 See discussion below, pp. 99–100. Note also

that the address to the various groups within the household found in Did. 4.9–11 is not really a change in dramatis personae, but rather a specification of sub-groups within the audience, though this still helps to create the structure of the text. The introduction of the various church officers are also not actual changes in dramatis personae, since the audience remains the same, but are better understood as individual topics of instruction on the same level as the instruction on baptism and the Eucharist. 27 HELLHOLM, Visionenbuch, 61.

2.1 Text-Linguistic Methodology

73

two occurrences of word association in the section on baptism (Did. 7–8). In Did. 8.1, the writer changes the topic from baptism to weekly fasting by picking up on the immediately preceding mention of pre-baptismal fasting in 7.4 and reiterating the term several times. Then in 8.2 the topic again changes, this time to daily prayer, through the repetition of the term uJpokritaiv (hypocrites), people who had been negatively associated with fasting in 8.1 (and Matt 6:16!). On the one hand, neither weekly fasting nor daily prayer is directly tied to the baptismal theme of the overarching text-part.28 It is, of course, possible that the theme “baptism” brought to the minds of the original audience a shared, underlying, conceptual framework that included instruction on fasting and prayer as well. It seems more likely, however, that the repetition of “fasting” and of “hypocrites,” in essence the only elements providing a sense of cohesion and coherence between those text-parts and the ones preceding them, is a weak connecting device on the semantic level employed by a later redactor.29 Thus I have added word association as the lowest marker among those on the semantic level. There is also another kind of repetition in the Didache, this time occurring on the syntactic level. It consists of repeated phraseology, essentially the figure of anaphora in Greek. This phenomenon can function both to furnish connections within a text-part and to delimit that text-part from others. An example of this is found in the tevknon (or “fences”) section, Did. 3.1–6, where a sequence of four introductory phrases is repeated five times.30 Such a pattern not only serves to make this section cohesive, but also distinguishes it from what precedes and what follows. In addition, the existence of this distinctive pattern gives support to the likelihood that this section was originally from a separate source, a possibility already raised by the fact that it is lacking in the Two Ways witness in Barnabas. Moreover the exact and proximate repetition of phraseology in this section may support and reflect the primary use of the Two Ways as stated in Did. 7.1, i.e., in the performance of baptism. Along these lines note also that the recurrent phrase tevknon mou is seemingly a type of renominalization, though it is not substitutionary in character but rather repeats exactly the form of the first occurrence and thus perhaps should be seen as continuing to participate in the text-external character of the original. Another new marker on the syntactic level is the stylistic shift in the expression of commands between the imperative mood (including mhv + aorist 28 In his discussion on thematic markers, JOHANSON, citing van Dijk, states that “a thematic marker may be seen as governing a text-sequence thematically in so far as the propositions of that text-sequence ‘satisfy’ it directly or indirectly,” To All the Brethren, 29. On the idea that the Lord’s Prayer was an original part of the Didache and that chs. 7–10, though showing later redaction, originally continued the initiation process begun with the Two Ways see DRAPER, “Ritual Process.” 29 See below, pp. 127, 157, 185, 190. 30 See below, p. 163.

74

Chapter 2: A Text-Linguistic Analysis of the Didache

subjunctive as the corresponding prohibitive form) and the volitive future. To give some examples, Did. 1.3–6, as noted earlier, uses only imperatival forms while, immediately following it, ch. 2 has only future forms; 3.1–8 employs imperatives again but 3.9–4.4 contains only future forms; chs. 6–16 favor the imperative, but they are interspersed with instances or series of future forms. Such variation is disjunctive in nature and seemingly disrupts the cohesiveness of the text, though to what extent this was sensed by the ancient reader is uncertain. Nonetheless, the phenomenon would appear to indicate something more than a change of nuance and seems likely to signal a shift in sources.31 Although the points at which these stylistic variations occur are beyond the text levels presented in the structural delimitation below, their ramifications for the structure of the text will be seen later in this study. Thus the delimitation markers must be adapted for the Didache as follows: I. Pragmatic markers a. meta-communicative sentences b. substitutions on the meta-level II. Semantic markers a. episode markers (temporal, local) b. change in dramatis personae c. change in topic via word association III. Syntactic markers a. renominalization b. exact repetition of phraseology c. change in syntactical style d. sentence and text connectors

2.2 The Text of the Didache The text used for the text-linguistic study is essentially that of the Jerusalem manuscript (H54). It is preferred because it is the only extant version of the entire Didache in Greek. The other sources for the text of the Didache fall into three categories: 1) fragments or manuscripts of the Didache itself; 2) later “church orders” that incorporate all or parts of the Didache into their texts; and 3) independent 31 In this sense it can be considered a “diachronic signal on the surface level” – my thanks to DAVID HELLHOLM for this observation. The future form as it is found in the legal context in which it is used in the Didache is generally considered to be a Semitism, reflecting the legal language of the Hebrew Bible as it is translated in the LXX. See, for example, BDF §362; A. T. ROBERTSON, A Grammar of the Greek New Testament in the Light of Historical Research (2d ed; New York: Hodder & Stoughton, 1915), 874–75; ERNEST DE WITT BURTON, Syntax of the Moods and Tenses in New Testament Greek (3d ed.; Edinburgh: T. & T. Clark, 1898), 67. See above, pp. 57–58, and below p. 158.

2.2 The Text of the Didache

75

witnesses to the Two Ways section (Two Ways), Did. 1–6. In more detail, these are as follows: 1) Extant witnesses to the Didache Within the first category are two fourth-century fragments – Papyrus Oxyrhynchus 178232 containing Did. 1.3c–4a and 2.7–3.2 in Greek; Coptic Br. Mus. Or. 927133 that includes Did. 10.3b–12.2 – and one complete Georgian manuscript34 that was previously collated against H54 but has since been lost. 2) Related church orders In the second group the complete Didache (in a somewhat reworked form) is found incorporated into the Greek Apostolic Constitutions,35 while a section comprised of most of Did. 11.3–13.7 followed by 8.1–2a is included in the Ethiopic version of the Ecclesiastical Canons,36 both works generally believed to have originated in the fourth century. Traces of the Didache in the third century Syriac Didascalia apostolorum are possible but disputed.37

32 Published in A. S. HUNT, Oxyrhynchus Papyri (vol. 15; London: Oxford University Press, 1922), 12–15. 33 First published with English translation by G. HORNER, “A New Papyrus Fragment of the Didaché in Coptic,” JTS (April, 1924): 225–31; with German translation and a more thorough investigation, CARL SCHMIDT, “Das koptische Didache-Fragment des British Museum,” ZNW 24 (1925): 81–99. Most recently, see the comprehensive study with photos of the manuscript, transcription, and English translation by F. STANLEY JONES and PAUL A. MIRECKI, “Considerations on the Coptic Papyrus of the Didache (British Library Oriental Manuscript 9271),” in The Didache in Context: Essays on Its Text, History & Transmission (ed. Clayton N. Jefford; Leiden: Brill, 1995), 47–87 (photos at end of volume). 34 Description and collation with H54 published by GREGOR PERADSE, “Die Lehre der zwölf Apostel in der georgischen Überlieferung,” ZNW 31 (1932): 111–16. 35 The text of the Constitutions can be found in METZGER, Les Constitutions apostoliques; for a discussion of the date of the Constitutions see Les Constitutions apostoliques, 1.57–60. 36 The text and English translation of the Ethiopic version of the Ecclesiastical Canons can be found in HORNER, Statutes. On the dating of the Ethiopic Ecclesiastical Canons consult the argument of AUDET, La Didachè, 34–45. 37 See especially R. H. CONNOLLY, “The Use of the Didache in the Didascalia,” JTS 24 (1923): 147–57. CONNOLLY concluded not only that the author of the Didascalia knew the Didache, but also that the author knew it in the form of H54. The correspondences given show shared concepts, terminology, and phraseology from throughout the Didache, including the so-called interpolated section, 1.3b–2. This points to a knowledge of the entire Didache and not simply to a Two Ways text. Yet the parallels in the Didascalia are short, inexact, and in different contexts than the corresponding material in the Didache, which leans toward the conclusion that the author knew the Didache but that the use of it was indirect.

76

Chapter 2: A Text-Linguistic Analysis of the Didache

3) Two Ways texts Finally, the third category of texts indicates that the Two Ways had an existence apart from the Didache. The Two Ways is found in an independent third-century Latin version under the title Doctrina apostolorum;38 a form of the Two Ways is included in the final chapters of the second century Epistle of Barnabas;39 the Way of Life, the first part of the Two Ways, forms the opening section of the fourth century Apostolic Church Order;40 and the Two 38

The Doctrina is attested in its entirety in an eleventh-century manuscript (Cod. Lat. Monac. 6264; olim Frising. 64) and as a fragment in one ninth- or tenth-century manuscript (Cod. Mellic. 597, olim 914, olim Q52). Regarding the fragment, see above, p. 46, for its early publication information; it was more recently published by KURT NIEDERWIMMER, “Doctrina apostolorum (Cod. Mellic. 597),” in Theologia Scientia Eminens Practica (ed. Hans-Christoph Schmidt-Lauber; Vienna: 1979), 266–72. As noted previously, the complete manuscript was first published by SCHLECHT, Doctrina xii apostolorum, with a broader study on the entire manuscript the next year, Doctrina xii apostolorum: Die Apostellehre in der Liturgie der katholischen Kirche. In 1913, LEO WOHLEB published a linguistically-oriented study of the Doctrina, including a collation of the Latin texts and a parallel Greek reconstruction, in Die lateinische Übersetzung der Didache kritisch und sprachlich untersucht (Paderborn: Schöningh, 1913). The text of the Doctrina and a short study was also published by RORDORF/TUILIER in an appendix to their commentary on the Didache, La Doctrine, 203–10. 39 A comparison of the texts of the Didache and Barnabas in English can be found in KRAFT, Barnabas and the Didache. 40 The Apostolic Church Order contains a Two Ways text plus an additional legislative section. The Greek text is extant in the twelfth-century manuscript Vindobonensis hist. gr. 7 olim 45, fol. 4–7, which was first published by BICKELL in Kirchenrechts, 1.1.107–32, and later also in: LAGARDE, Reliquiae (graece), 74–79; J. B. PITRA Iuris ecclesiastici Graecorum historia et monumenta (vol. 1; Rome, 1864), 77–88; HILGENFELD, Evangeliorum secundum Hebraeos, secundum Petrum … (2d ed.), 110–21; and THEODOR SCHERMANN, Die allgemeine Kirchenordnung, frühchristliche Liturgien und kirchliche Überlieferung (3 vols.; Paderborn: Schöningh, 1914), 1:12–34. It is also extant in Syriac manuscripts, including a codex of the eleventh- or twelfth-century, Cantabr. O. o. I.2 (complete but parts illegible), and a seventeenth-century Mosul Codex, now called by ARTHUR VÖÖBUS Mar Behnam 1/1, both of which are collated and translated into English in ARENDZEN, “Apostolic Church Order,” 59– 80; an additional eighth-century Syriac witness, Mardin Orth. 309, is discussed in VÖÖBUS, “Die Entdeckung der ältesten Urkunde für die syrische Übersetzung der Apostolischen Kirchenordnung,” OrChr 63 (1979): 37–40, where other, later witnesses are also listed. The Two Ways portion of the Apostolic Church Order, however, also had an independent circulation and can be seen in four Greek manuscripts and one Syriac manuscript. Interestingly, while the Syriac represents a complete text of the Two Ways, the Greek manuscripts attest two different abridgments of the work. One recension is found in the Greek Mosquensis bibl. s. Synodi 125 from the tenth century (see GEBHARDT and HARNACK, Barnabae epistula, Proleg. xxix–xxxi); the other Greek recension, the so-called “X Recension” or “Epitome,” is represented by three manuscripts: Ottob. gr. 408 fol. 88v; Paris gr. 1555A fol. 177; and Napol. II C34 (olim 35) fol. 83; they can be found in THEODOR SCHERMANN, Eine Elfapostelmoral oder die X-Recension der “beiden Wege” (Munich, 1903). The Syriac text is attested in Sangerm. Paris. Orient. syr. 38, published by PAUL A. DE LAGARDE, Reliquiae juris ecclesiastici antiquissimae syriace (Leipzig, 1856).

2.2 The Text of the Didache

77

Ways is found as well in several other Christian texts, though in even more modified forms.41 As already noted, although some early twentieth-century scholars held that there was a literary dependency between the Didache and Barnabas, most contemporary scholars believe instead that there was a common Two Ways source shared by these texts, a conclusion reached based on the co-existence of overall agreement in wording alongside important differences.42 The Two Ways of Barnabas, unlike that of the Didache, occurs at the end of the text, i.e., in chs. 18–20. The transition from the previous chapters to the Two Ways is abrupt: Barn. 18.1 tau'ta me;n ou{tw". Metabw'men de; kai; ejpi; eJtevran gnw'sin kai; didachvn. ÔOdoi; duvo eijsi;n didach'" kai; ejxousiva" … So it is with these things. Now let us move on to another insight and teaching. There are two ways of teaching and power …

The fact that the author calls this section another “gnosis” and “teaching” adds to the sense that this is a new source and not a composition of the author.43 Yet the Two Ways of Barnabas is very different from the Didache. The “ways” are not those of “life/death” but of “light/darkness,” which are presided over by “angels of God/Satan” respectively. A form of the instruction on love of neighbor may be present in Barn. 19.5 (“you shall love your neighbor more than yourself”) but the Golden Rule is lacking in Barnabas as is 1.3b–2.1 and the tevknon section, Did. 3.1–6. The content of the Way of Light in Barnabas is much the same as the Way of Life in the Didache, but is 41

In the Apostolic Church Order, however, there may be evidence that the entire Didache was known to the author (see below, pp. 78–79; 106–7;109–10). Likewise, the fourth century Syntagma didascalias (text in PG 28.835–46 and H. HYVERNAT, “Le Syntagma doctrinae,” in Studia patristica: Études d’ancienne littérature chrétienne [ed. Pierre Batiffol; Paris: Leroux, 1890], 121–28) and the related and contemporaneous Fides cccxviii patrum (text in PG 28.1637–1644) reproduce much of the Two Ways (though apart from the Two Ways framework) but several passages make it possible that the Didache was the source used. (For a summary of the arguments and bibliography see NIEDERWIMMER, Didache, 13–14; 34–35.) The Arabic version of the Life of Shenoute in its introductory section reproduces the Two Ways in a form that testifies to the use of only a Two Ways document (for the Arabic version see ÉMILE AMÉLINEAU, Monuments pour servir a l’histoire de l’Egypte chrétienne aux IVe et Ve siècles [vol. 4 of Mémoires publiés par les membres de la Mission archéologique française au Caire, 1885–1886; ed. E. Grébaut; Paris: Leroux, 1888], xlviii–xciii; 289–478; 480.) For other, later, texts and their bibliographies see NIEDERWIMMER, Didache, 14. 42 For a presentation of the Doctrina, Didache, Barnabas and the Life of Shenoute in parallel, see GOODSPEED, “Didache, Barnabas and the Doctrina,” 238–47. 43 A Latin version of Barnabas that includes only chs. 1–17, dated possibly as early as the third century, once led some to believe that the Two Ways was a later addition, but the fact that Two Ways imagery is seen throughout the text makes it more likely that the Latin version is abbreviated. See KRAFT, Barnabas and the Didache, 5.

78

Chapter 2: A Text-Linguistic Analysis of the Didache

organized very differently. Yet, beginning with the ending of the Way of Light in Barn. 19.11b, the wording of both texts becomes so alike that a belief in the existence of a common source is warranted. In addition, there is no indication that the Two Ways material in Barnabas had a baptismal context, as is the case in the Didache. Even closer to the wording of the Didache is the Two Ways in the Doctrina. It is so close, in fact, that some early scholars called the Doctrina a Latin translation of the Didache.44 In the opening of the Doctrina, the “ways” are identified as life/death, as in the Didache, but also as light/darkness, as in Barnabas. Over these preside two angels, one of righteousness and one of iniquity, again very close to Barnabas. While on the one hand the Doctrina contains both parts of the “Double Love Commandment,” i.e., love of God and love of neighbor, as found in the Didache, it nonetheless lacks 1.3b–2.1, as does Barnabas. The tevknon section is also included in the Doctrina, though without Did. 3.3. Yet overall, the Doctrina appears to be the fullest recension of the Two Ways source. Edgar Goodspeed counted 161 separate items in the Doctrina, of which 145 are present in the Didache and 104 in Barnabas.45 But there are also eight items shared by the Didache and Barnabas not present in the Doctrina46 and, in comparison with the Didache, there is some shifting of lines and some items missing in the Latin. Again, the evidence is quite compelling for the existence of a common Two Ways source among these texts and against their literary dependency. Another witness to the Two Ways is found in the fourth-century Apostolic Church Order (Apos. C.O.). This text circulated in a fuller form and in attenuated forms (only the Two Ways),47 as well as within the larger Ecclesiastical Canons. Roughly speaking, the Apostolic Church Order attests Did. 1.1– 3a and 2.2–4.8 (witness Ottob. gr. 408 also includes 4.9–14), omitting 1.3b– 2.1 and the Way of Death, but including the tevknon section. The Way of Life, however, has been apportioned to eleven apostles48 and, in the fuller edition, this apportionment begins to repeat, serving to maintain the format. The Two Ways text of the Apostolic Church Order is closest to the Doctrina, yet has elements peculiar to the Didache and Barnabas respectively.49 44

E.g., WOHLEB entitled his 1913 work on the Doctrina, Die lateinische Übersetzung der Didache. GOODSPEED also pointed out that as late as Berthold Altaner in 1938 and M. L. W. Laistner in 1943, the Doctrina was still described as a Latin translation of the Didache (“Didache, Barnabas and the Doctrina,” 230, n. 3). 45 GOODSPEED, “Didache, Barnabas and the Doctrina,” 237. 46 Ibid., 234. 47 See above, p. 76, n. 40. 48 It is interesting that the eleven includes both Cephas and Peter, along with Nathaniel, and lacks Thaddeus (Judas) and James the Lesser. 49 Apos. C.O. 4 parallels Barn. 19.2 in including in the first part of the Double Love Commandment, doxavsei" to;n lutrwsavmenovn se ejk qanavtou (glorify him who redeemed

2.2 The Text of the Didache

79

Certainly the existence of multiple witnesses to an ancient text is always welcome. Alongside a text-linguistic analysis of H54, they provide an exceptional (among early Christian texts) opportunity to look at the text diachronically, i.e., to trace its growth and development and to explore the possible causes of such change. They also provide information on the synchronic level to help in the understanding of the function and meaning of the text at each stage of its history. Unfortunately the one hundred years of research already devoted to the Didache in its various forms leaves the student of the text somewhat predisposed to certain redactional theories, making the (always doomed, yet still necessary) attempt at objectivity in the text-linguistic analysis all the more difficult. Circular reasoning notwithstanding, however, the advantage of a text-linguistic study is that it requires of the scholar a complete and detailed structural analysis based on specific data. In the past, scholars have studied the structure of the Didache, noting one inconsistency or another in the text and suggesting that it was indicative of editorial and/or redactional activity. Very often they have been right. In fact the observation of such phenomena illustrates a basic tenet of text linguistics, the necessity for the surface structure of a text to be recognizable to its audience. A text-linguistic analysis points out those places in the Didache where the surface structure has suffered some kind of disruption, problems sometimes mentioned in the past, though without a precise and over-arching way to describe them. The field of text linguistics has the advantage of having developed a technical language with which to deal with such phenomena, applicable to any text. In addition, such an analysis highlights other phenomena that have gone unnoticed. The purpose of doing a text-linguistic analysis of the Didache, then, lies in the hope not only that such a study may make clearer and more precise what past scholarship has already discerned about the state of the text, but that it may by you from death). Also Apos C.O. 14 contains material that is parallel to Barn. 21.2–4, 6; 19.11, including an emphasis on doing good for the needy, especially in light of the imminence of the end-times, and on the necessity of maintaining the overall body of instruction unchanged. Yet, like the Didache, the Apostolic Church Order lacks the light/darkness imagery of the introduction in Barnabas and the Doctrina; Apos. C.O. 6 lists yeudomarturhvsei" in the same position as Did. 2.3, in contrast with its earlier occurrence in the Doctrina; Apos. C.O. 6 includes the phrase ou{" me;n ejlevgxei", peri; de; w}n proseuvxh/ found in Did. 2.7 but lacking in the Doctrina and Barnabas; Apos. C.O. 8 includes the teaching in the tevknon section regarding lust and fornication that is missing from the Doctrina (and Barnabas); Apos. C.O. 13 includes the teaching ouj lhvyh/ provswpon ejlevgxai tina; ejpi; paraptwvmati found in Did. 4.3 but lacking in the Doctrina and Barnabas. Corresponding to the Doctrina, however, is the absence of dia; pantov" (Did. 3.8) and of ajlla; memestwmevno" pravxei (Did. 2.5). Some believe that the Apostolic Church Order was familiar with the entire Didache because of the inclusion of parallels to Did. 10.3b and 13.1–2 in Apos. C.O. 12. Most, however, feel that these are not enough to support that conclusion. KRAFT believes that the author began with a source comparable to the Doctrina to which was added excerpts from at least Barnabas and possibly the Didache (Barnabas and the Didache, 10).

80

Chapter 2: A Text-Linguistic Analysis of the Didache

its very nature reveal new data from which to recover the genre and the redaction history of the text (and thereby a part of Church history), and that it may present these results in a manner which is more easily accessible and amenable to scholarly review, debate, and modification.

2.3 Structurally Delimiting the Didache of H54 2.3.1. Introduction In creating a picture of the structure of the Didache, the text-linguistic analysis must present the text not only from beginning to end but also from the surface level of the text, where the basic structure is communicated to the reader, to its most deeply imbedded sub-texts. Thus the structural outline has both a vertical component to it, representing the sequence of the text-parts as one would read them, as well as a horizontal aspect, showing the depth of a particular text-part and indicating its relationship to other text-parts. The results of the text-linguistic analysis are given below. First a list of the technical abbreviations used in the analysis is provided (2.3.2). This is followed by a detailed outline presented on facing pages (2.3.3): the proposed text-linguistic delimitation is shown on the right-hand (odd-numbered) pages;50 the relevant portions of the text and a brief explanation of the textlinguistic information are given on the left-hand side (even-numbered pages). Following the outline I have condensed the results into a more simplified representation showing the most important structural components of the text (2.3.4) and have also included a discussion of the various communication levels found therein (2.3.5) as well as the dramatis personae appearing in the text (2.3.6). Finally, a more in-depth explanation of the results of the analysis is presented (2.4). As mentioned earlier, the text given here is that of H54 and I have used as a starting point the 1887 transcription of Harris, who rendered into standard Greek the various abbreviations, contractions, tachygraphical signs, itacisms, alternative accentuations, and other paleographical phenomena characteristic of manuscripts of that period in an unobtrusive but clearly documented manner and with minimal emendation. From the vantage point several years after the original publication of the text, Harris was also able to consider and present the readings of Bryennios, Harnack, Hilgenfeld, and others and I have noted those where appropriate.51 I have not, however, followed any one 50 As mentioned above (p. 5, n. 3), the delineation of the chapters of the Didache in H54 was from Bryennios while the versification is attributed to Harnack. I have further divided the verses into sub-units based on the stops in the text. These sub-parts are noted alphabetically. 51 See HARRIS, Teaching, 1–10 for the transcription and 12–14 for initial discussion.

2.3 Structurally Delimiting the Didache of H54

81

scholar in emending the text or in the reading of the punctuation.52 On the latter point, it is, of course, impossible to know what, if any, information the punctuation of this eleventh-century witness can provide about the original structure of the Didache. Still, the marking of sense divisions, however sparse, can be found in manuscripts not too far removed from the date generally accepted for the Didache53 and thus the transmittal of such data by H54 must not be excluded out of hand. It is true that the system of punctuation found in H54 has already moved away from the system of Aristophanes of Byzantium (third-century B.C.E.) toward the modern system adopted today in editions of Greek texts.54 In the earlier system a full stop was represented by a high point (stigmh; teleiva); the pause of our semi-colon was represented by a low point (uJpostigmhv); and the pause of the comma was signaled by a middle point, (stivgmh mevsh) (though, according to Bruce Metzger, the value assigned to these last two marks could vary among scribes). The use of the middle point eventually died out, its place having been taken by the comma starting about the ninth century. In contrast, Bryennios noted that in H54 the stigmh; teleiva/full stop was represented by a low point; the (more common) pause of a semi-colon was signaled by a stigmh; mevsh or a[nw, i.e., a middle or high point; and the modern comma sign had now become the uJpostigmh, i.e., the shortest pause.55 In addition, Bryennios stated that H54 shows one 52 With respect to emendations, I have generally accepted, without note, HARRIS’s standardizations (e.g., in accentuation, in places where itacism has likely occurred, and in his rendering of phrases such as mh; dev into single words) as well as his virtually universal retention of the manuscript’s inconsistent presentation of the phenomena of movable n and ". In terms of the portions of the text presented here, the widely accepted emendations in 9.4, 10.4 and 11.5, with which I concur, have been noted in brackets. In matters where suggested corrections seemed unnecessary with respect to the understanding of the text, namely in the addition of ou}" de; ejlehvsei" after ou}" me;n ejlevgxei" in 2.7 (based on other witnesses), the change to poiei' for ms. poivei in 6.2, the correction of keleuvei" to keleuvsei" in 7.4, the correction of prosexomologhsavmenoi to proexomologhsavmenoi in 14.1, the change of thvn to tivna in 14.2, and the change of eJtevrou to eJtaivrou in 15.3, I have opted for a more conservative approach and reproduced the text of H54 without emendation. 53 The following examples can be found in BRUCE METZGER, Manuscripts of the Greek Bible: An Introduction to Greek Paleography (Oxford and New York: Oxford University Press, 1981): p46, dated to about 200 C.E., uses slight spaces to indicate pauses (64–65, Plate 6); p66, also dated to 200 C.E., appears to have actual punctuation marks (66–67, Plate 7); and Rahlfs 967, dated to the early third century C.E., has marks of punctuation “here and there” (70–71, Plate 10). 54 On the system of Aristophanes see EDWARD MAUNDE THOMPSON, A Handbook of Greek and Latin Palaeography (London: 1901; repr., Chicago: Ares Publishers, 1975), 70; on its varied application see METZGER, Manuscripts, 32. 55 BRYENNIOS, Didachv, z'. Following BRYENNIOS, I am representing the positions of the points in relation to the central axis of the letters rather than the actual ruled lines of the page since the writing of H54 is pendent from the line, as is the case generally with manuscripts of its period according to METZGER, Manuscripts, 26, 28.

82

Chapter 2: A Text-Linguistic Analysis of the Didache

example (in 1 Clem. 16) of the mark of interrogation, the ejrwthmatikovn, represented by the sign used in English today for the semi-colon,56 a phenomenon that began about the eighth or ninth century.57 Could the sense pauses found in H54 have any relation to the original, even perhaps oral, tradition of the Didache? Because it is still remotely possible that the text of the Didache in H54 bears within it some traces of early or even original punctuation (whether oral or literary), and because any other presentation would be even more arbitrary, I have decided to present the punctuation according to my reading of the manuscript.58

56

BRYENNIOS, Didachv, q'. See THOMPSON, Handbook, 70; METZGER, Manuscripts, 32. 58 BRYENNIOS punctuated his edition of the text according to his own interpretation of its meaning. HARRIS, however, attempted to reproduce the actual punctuation of the manuscript and I have done the same. My presentation differs from that of HARRIS in the reading of the full stop, i.e., the low point. Throughout the text as a whole, I find many fewer instances of the full stop as compared with HARRIS’ twenty-four (not including the lozenge of the title represented by him by a period), and several of mine were not recognized as such by HARRIS. Moreover twelve of HARRIS’ full stops were presented either without stop or by a comma in BRYENNIOS. In my opinion, several factors complicate the determination of the qevsei" (positions of the points). The first is the fact that the writing of the scribe sometimes drifts above the ruled line which makes it uncertain as to whether the placing of the point is related to the position of the line or to that of the letters – I have chosen the latter (see, for example, folio 78a. 14, 15). Secondly, there are instances where what appears visually to be a low point is, from the semantics of the text, unlikely to have been considered a full stop (see, for example, 78a. 11). Finally, parallel, proximate, grammatical constructions at times are assigned different punctuations by HARRIS based on his observation of the manuscript (see, for example, 77a. 15 cf. 8, 10, 13, 18). Thus any transcription of the punctuation is open to question and one suspects that perhaps precision was not of the greatest importance to the scribe himself. 57

2.3 Structurally Delimiting the Didache of H54

83

2.3.2 Text-Linguistic Abbreviations Based on the system used by David Hellholm, the notations used in the analysis are as follows: ST: sub-text. In the notation nST n the superscript number on the left indicates how deeply the text-unit is imbedded with respect to the surface level (Text Level 1) of the text; the number or series of numbers to the right indicates the sequential position of the text-unit vis-à-vis the text as a whole. MS: meta-communicative sentence/clause. A meta-communicative sentence/clause serves both to delimit an adjacent text-unit and move the communication to a new, deeper level. In the notation MS n the superscript number to the right indicates the level of communication at which the MS occurs vis- à-vis the initial communication level of the text (Communication Level 1 = between author and readers). SM: substitution on the meta-level. A substitution on the meta-level is a word or phrase that stands in place of a text or sub-text, as is the case, for example, with titles or chapter headings. In the notation SM n the superscript number to the right indicates the level at which the substitution occurs vis-à-vis the surface level of the text. MNS: meta-narrative sentence/clause. When an MS and SM occur together, their entire sentence/clause designates the adjacent text-unit as a text in its own right. In the notation MNS n the superscript number to the right indicates the level at which the MNS occurs vis-àvis the text’s surface communication level. SA: substitution on the abstraction-level. A substitution on the abstraction level is a word or phrase that stands in place of a text or sub-text but which, semantically speaking, is broader than the object that it represents as, for example, with a pronoun. Thus an SA delimits a text-unit but is not quite a text-level above that unit, i.e., it is not at a meta-level. DP:

dramatis personae. A change in agents, their grouping, and/or their status.

EM:

episode markers. Terms indicating local and/or temporal circumstances.

WA: word association. The repetition of a noun/phrase from a previous section to introduce a new theme. RN:

renominalization. The re-entry into the text of an entity previously introduced.

RP: repetition. The repetition of a word or phrase to produce two or more distinct, yet connected, sub-texts. (STL: syntactical style. A change in writing style via the use of alternative syntactical forms, for example a move from the use of the volitive future to that of the imperative.) TC:

text connector. For example, a conjunction.

(surr): surrogate. A word or phrase that is, technically speaking, not the delimitation marker specified, but nonetheless functions as one.

84

Chapter 2: A Text-Linguistic Analysis of the Didache

2.3.3 Presentation of Delimitation didach; tw'n dwvdeka ajpostovlwn:. Didach; kurivou dia; tw'n dwvdeka ajpostovlwn toi'" e[qnesin: 1.1) oJdoi; duvo eijsiv. miva th'" zwh'": kai; miva tou' qanavtou: diafora; de; pollh; metaxu; tw'n duvo oJdw'n: 6.1) o{ra mhv ti" se planhvsh/ ajpo; tauvth" th''" oJdou' th'" didach'": ejpei; parekto;" qeou' se didavskei: 2) eij me;n ga;r duvnasai bastavsai o{lon to;n zugo;n tou' kurivou tevleio" e[sh/: eij dΔ∆ ouj duvnasai o} duvnh/ tou'to poivei (Harris poiei'): 3) peri; de; th'" brwvsew": o} duvnasai bavstason: ajpo; de; tou' eijdwloquvtou livan provsece: latreiva gavr ejstin qew'n nekrw'n: 7.1) peri; de; tou' baptivsmato" ou{tw baptivsate: tau'ta pavnta proeipovnte" baptivsate: eij" to; o[noma tou' patro;" kai; tou' uiJou' kai; tou' aJgivou pneuvmato" ejn u{dati zw'nti: 11.1) o}" a]n ou\n ejlqw;n didavxh/ uJma'" tau'ta pavnta ta; proeirhmevna devxasqe aujtovn: 2) eja;n de; aujto;" oJ didavskwn strafei;" didavskh/ a[llhn didach;n eij" to; katalu'sai. mh; aujtou' ajkouvshte: eij" de; to; prosqei'nai dikaiosuvnhn kai; gnw'sin kurivou. devxasqe aujto;n wJ" kuvrion: 15.4) ta;;" de; eujca;" uJmw'n kai; ta;" ejlehmosuvna" kai; pavsa" ta;" pravxei": ou{tw" poihvsate wJ" e[cete ejn tw'/ eujaggelivw/ tou' kurivou hJmw'n: 16.1) grhgorei'te uJpe;r th'" zwh'" uJmw'n: The second heading Didach; kurivou dia; tw'n dwvdeka ajpostovlwn toi'" e[qnesin: (Teaching of the Lord Through the Twelve Apostles to the Gentiles) stands for the entire following text and is thus a title or substitution (SM) for it. Because the title is not part of its co-text, however, it is said to be on the meta-level, i.e., one level above it. The title is cataphoric in that it occurs before its co-text. That it is also a suppositio materialis means that the text referred to is here present (vs. a suppositio formalis, i.e., a reference to an external text – see below pp. 128 and n. 161; 140; 185). The title also functions as a meta-communicative sentence/clause (MS) in that the co-text corresponding to “teaching” occurs at a deeper communication level: at the juncture between the title and the co-text the communication moves from the level of author/reader to that of Lord (apostles)/Gentiles. That the markers SM + MS occur together here also creates a metanarrative sentence/clause (MNS), marking the co-text as a complete text in its own right. The surface level of the text is divided at 15.4 by the phrase ta;;" de; eujca;" uJmw'n kai; ta;" ejlehmosuvna" kai; pavsa" ta;" pravxei" (your prayers and alms and all your practices) which functions as a summary of the entire preceding discussion. The occurrence of this surrogate SM after its co-text is designated as anaphoric. In the sub-text 1.1–15.3, a break is created in 11.1 by tau'ta pavnta ta; proeirhmevna (all these aforementioned things) which is a reference (SA/MS) to the text in 1.1–10.7; this delimitation is supported by the phrase a[llhn didachvn (another teaching) in 11.2, a SM (surr) which implies that 1.1–10.7 is itself a “teaching” (see below, pp. 137–38; 156–60; 185–85). The section 1.1–10.7 is itself sub-divided at 7.1 by the phrase tau'ta pavnta (all these things), a SA which refers back to the text in 1.1–6.3, and by proeipovnte" (saying beforehand), a MS designating 1.1–6.3 explicitly as an act of communication; together they form a meta-narrative sentence/clause (MNS) and indicate that 1.1–6.3 is itself a separate text. In 6.1–2 tauvth" th''" oJdou' th'" didach'" (this way of teaching) and o{lon to;n zugo;n tou' kurivou (the whole yoke of the Lord) are substitutions that both refer back to the preceding text in 1.1–5.2. Within that sub-text, 1.1 oJdoi; duvo eijsiv: miva th'" zwh'": kai; miva tou' qanavtou: diafora; de; pollh; metaxu; tw'n duvo oJdw'n (there are two ways, one of life and one of death, and there is a great difference between the two ways) is a cataphoric reference (SM) to what follows in 1.2–5.2.

2.3 Structurally Delimiting the Didache of H54

85

00ST Inscriptio: didach; tw'n dwvdeka ajpostovlwn 0ST Incipit: didach; kurivou dia; tw'n dwvdeka ajpostovlwn toi'" e[qnesin = MS0(surr) SM0: didach; kurivou dia; tw'n dwvdeka ajpostovlwn toi'" e[qnesin cataphoric suppositio materialis = the entire following text SM0 + MS0 = MNS0 DP: kurivou (Jesus) = quoted speaker; tw'n dwvdeka ajpostovlwn = speaker(s); toi'" e[qnesin = listeners 1ST1 A compilation of instruction on Christian life/practice, on church rites and on certain practical aspects of community life: 1.1–15.4. SM1(surr): ta;" de; eujca;" uJmw'n kai; ta;" ejlehmosuvna" kai; pavsa" ta;" pravxei" (15.4) = 1.1–15.3 (anaphoric) 2ST1.1 Instructions on Christian life/practice and church rites: 1.1–11.2 SA (SM2 surr): tau'ta pavnta ta; proeirhmevna (11.1a) = 1.1–10.7 SM2(surr): a[llhn didachvn (11.2a) suppositio materialis = 1.1–10.7 (anaphoric) DP: o}" didavxh/, oJ didavskwn = hypothetical agent referred to but not participating 3ST1.1.1 Instruction on Christian life/practice, recited prior to baptism: 1.1–6.3 SA (SM3 surr)/MS1: tau'ta pavnta proeipovnte" (7.1b) = 1.1–6.3 (anaphoric) SA + MS1 = MNS DP: proeipovnte" (7.1b) readers (= speakers) 4ST1.1.1.1 The intermediate stage of the Two Ways: 1.1–6.2 SA/SM4: tauvth" th'" oJdou' th'" didach'" (6.1) =1.1–5.2 SM4:

(anaphoric) o{lon to;n zugo;n tou' kurivou (6.2) = 1.1–5.2 (anaphoric)

5ST1.1.1.1.1 The body of the Two Ways teaching, a parenesis on righteous living vs. wickedness: 1.1–5.2 SM5: oJdoi; duvo eijsiv: miva th'" zwh'": kai; miva tou' qanavtou: diafora; de; pollh; metaxu; tw'n duvo oJdw'n (1.1) = 1.2–5.2 (cataphoric).

86

Chapter 2: A Text-Linguistic Analysis of the Didache

1.1) oJdoi; duvo eijsiv: miva th'" zwh'": kai; miva tou' qanavtou: diafora; de; pollh; metaxu; tw'n duvo oJdw'n: 2) hJ me;n ou\n oJdo;" th'" zwh'" ejstin (ms. e[stin passim; Harris ejsti;n) au{th: prw'ton: ajgaphvsei" to;n qeo;n to;n poihvsantav se: deuvteron: to;n plhsivon sou wJ" seautovn: pavnta de; o{sa eja;n qelhvsh/" mh; givnesqaiv soi: kai; su; a[llw/ mh; poivei: 3) touvtwn de; tw'n lovgwn hJ didachv ejstin au{th: 2.1) deutevra de; ejntolh; th'" didach'": 2) ouj foneuvsei": ouj moiceuvsei": ouj paidofqorhvsei": … 7) ouj mishvsei" pavnta a[nqrwpon: ajlla; ou}" me;n ejlevgxei": peri; de; w|n proseuvxh/: ou}" de; ajgaphvsei" uJpe;r th;n yuchvn sou: 3.1) tevknon mou feu'ge ajpo; panto;" ponhrou': kai; ajpo; panto;" oJmoivou aujtou': 2) mh; givnou ojrgivlo" (ms., Harris ojrgi'lo"): oJdhgei' ga;r hJ ojrgh; pro;" to;n fovnon: mhde; zhlwthv": mhde; ejristikov": mhde; qumikov": ejk ga;r touvtwn aJpavntwn fovnoi gennw'ntai: 3) tevknon mou. mh; givnou ejpiqumhthv": oJdhgei' ga;r hJ ejpiqumiva pro;" th;n porneivan: mhde; aijscrolovgo": mhde; uJyhlovfqalmo": ejk ga;r touvtwn aJpavntwn moicei'ai gennw'ntai: … 7) i[sqi de; prau?": … 4.1) tevknon mou tou' lalou'ntov" soi to;n lovgon tou' qeou': mnhsqhvsh/ nukto;" kai; hJmevra": 4.12) mishvsei" pa'san uJpovkrisin kai; pa'n o} mh; ajresto;n tw'/ kurivw/: 13) ouj mh; ejgkatalivph/" ejjntola;" kurivou: fulavxei" de; a} parevlabe": mhvte prostiqei;": (per ms.) mhvte ajfairw'n: 14) ejn ejkklhsiva/ ejxomologhvsh/ ta; paraptwvmatav sou: kai; ouj proseleuvsh/ ejpi; proseuchvn sou ejn suneidhvsei ponhra'/: au{th ejsti;n hJ oJdo;" th'" zwh'": 5.1) hJ de; tou' qanavtou oJdov" ejstin au{th: prw'ton pavntwn ponhrav ejsti kai; katavra" mesthv: fovnoi: moicei'ai: … The phrases hJ me;n ou\n oJdo;" th'" zwh'" ejstin au{th (on the one hand, the way of life is this) in 1.2a, and au{th ejsti;n hJ oJdo;" th'" zwh'" (this is the way of life) in 4.14c (both renominalizations [RN] of miva th'" zwh'" in 1.1a) are substitutions (SA/SM) that denote the sub-text 1.2b–4.14b (see also Hellholm, “Kristen dopkatekes,” 119; Hellholm and Blomkvist, “Parainesis as an Ancient Genre-Designation,” 469). That text is in turn sub-divided at 1.3a where touvtwn de; tw'n lovgwn (and of these words) refers back to 1.2 while hJ didachv ejstin au{th (the teaching is this) refers ahead to 1.3b–4.11. That the sub-text ends at 4.11 is determined by the phrase in 4.12, pa'san uJpovkrisin kai; pa'n o} mh; ajresto;n tw'/ kurivw/ (every hypocrisy and everything which is not pleasing to the Lord), a summary (SM[surr]) of the preceding text in 1.3b–4.11. Likewise ejjntola;" kurivou (commands of the Lord) (SM) and a} parevlabe" (everything you have received) (SA) in 4.13 are also anaphoric references to the text in 1.3b–4.11. Note how 4.14a-b, ejn ejkklhsiva/ ejxomologhvsh/ ta; paraptwvmatav sou: kai; ouj proseleuvsh/ ejpi; proseuchvn sou ejn suneidhvsei ponhra'/: (in church confess your sins and do not come to your time of prayer with an evil conscience) falls in between the two anaphoric SMs in 4.12-13 and the SM in 4.14c. The implication is that the instruction in 4.14a-b is part of the Way of Life yet not part of the ejjntola;" kurivou and may indicate redactional activity. Indeed the Two Ways witnesses attest variations here: Barnabas has all of Did. 4.12–14 except for ejn ejkklhsiva/; Doctrina, however, lacks Did. 4.13a, 14a. Deutevra de; ejntolh; th'" didach'" (the second command of the teaching) in 2.1 refers to the sub-text following in 2.2–4.11 (SM); at the same time it also implies that the text in 1.3b– 6 was the “first command of the teaching.” Within the sub-text 2.2–4.11, 3.1 is a summary of the instruction that will follow in 3.2–6 and therefore functions as a substitution (SM [surr]) for it. In addition, the repetitious elements (RP) in 3.1–6 bind the section together structurally and also serve to distinguish it from the surrounding text. While the single element tevknon mou occurs again in 4.1, it is not accompanied by the other recurring elements of section 3.1– 6 and is also separated from this section by the material of 3.7–10. It thus falls outside of that text-part though, along with the renominalization tevkna in 5.2u, it lends cohesion to the larger section.

2.3 Structurally Delimiting the Didache of H54

87

6ST1.1.1.1.1.1 Way of Life: 1.2–4.14 SA/SM6: hJ me;n ou\n oJdo;" th'" zwh'" ejsti;n au{th (1.2a) = 1.2b–4.14b (cataphoric) SA/SM6: au{th ejsti;n hJ oJdo;" th'" zwh'" (4.14c) =1.2b–4.14b (anaphoric) (with 1.2b forms inclusio) RN: hJ oJdo;" th'" zwh'" (1.2a; 4.14c) = miva th'" zwh'" (1.1a) 7ST1.1.1.1.1.1.1 Statement of main principles: 1.2b–e 7ST1.1.1.1.1.1.2 Explication of main principles:1.3–4.11 SA: touvtwn de; tw'n lovgwn (1.3a) = 1.2 (anaphoric) SA/SM7: hJ didachv ejstin au{th (1.3a) = 1.3b–4.11 (cataphoric) SM7(surr): pa'san uJpovkrisin kai; pa'n o} mh; ajresto;n tw/' kurivw/ (4.12) = 1.3b–4.11 (anaphoric) SM7: ejntola;" kurivou (4.13a) = 1.3b–4.11 (anaphoric) SA: a} parevlabe" (4.13b) = 1.3b–4.11 (anaphoric) 8ST1.1.1.1.1.1.2.1 First command: 1.3b–6 (SM8 [implicit] = 1.3b–6) 8ST1.1.1.1.1.1.2.2 Second command: 2.1–4.12 SM8: deutevra de; ejntolh; th'" didach'" (2.1) = 2.2–4.11 (cataphoric) 9ST1.1.1.1.1.1.2.2.1 List of prohibitions: 2.2–7 9ST1.1.1.1.1.1.2.2.2 Tevknon section: 3.1–6 SM9 (surr): feu'ge ajjpo; panto;" ponhrou': kai; ajpo; panto;" oJmoivou aujtou' (3.1) = 3.2–6 (cataphoric) RP: tevknon mou 3.1a, 3a, 4a, 5a, 6a RP: mh; givnou 3.2a, 3b, 4a, 5a, 6a RP: oJdhgei' gavr 3.2b, 3c, 4b, 5b, 6b RP: mhdev … mhdev 3.2c-e, 3d-e, 4c-f, 5c-d, 6c-d RP: ejk ga;r touvtwn aJpavntwn … gennw'ntai 3.2f, 3f, 4g, 5e, 6e 9ST1.1.1.1.1.1.2.2.3 Miscellaneous commands: 3.7–10 9ST1.1.1.1.1.1.2.2.4 Inter-community relations: 4.1–11 RP: tevknon mou 4.1a 7ST1.1.1.1.1.1.3 Summary and conclusion: 4.12–14b

88

Chapter 2: A Text-Linguistic Analysis of the Didache

5.1) hJ de; tou' qanavtou oJdov" ejstin au{th: prw'ton pavntwn ponhrav ejsti kai; katavra" mesthv: fovnoi: moicei'ai: … 2u) rJusqeivhte tevkna ajpo; touvtwn aJpavntwn: 6.1) o{ra mhv ti" se planhvsh/ ajpo; tauvth" th''" oJdou' th'" didach'": ejpei; parekto;" qeou' se didavskei: 2) eij me;n ga;r duvnasai bastavsai o{lon to;n zugo;n tou' kurivou tevleio" e[sh/: eij dΔ∆ ouj duvnasai o} duvnh/ tou'to poivei: 3) peri; de; th'" brwvsew": o} duvnasai bavstason: ajpo; de; tou' eijdwloquvtou livan provsece: latreiva gavr ejstin qew'n nekrw'n: 7.1) peri; de; tou' baptivsmato" ou{tw baptivsate: tau'ta pavnta proeipovnte" baptivsate: eij" to; o[noma tou' patro;" kai; tou' uiJou' kai; tou' aJgivou pneuvmato" ejn u{dati zw'nti: 2) eja;n de; mh; e[ch/" u{dwr zw'n: eij" a[llo u{dwr bavptison: eij dΔ∆ ouj duvnasai ejn yucrw'/: ejn qermw'/. 3) eja;n de; ajmfovtera mh; e[ch/": e[kceon eij" th;n kefalh;n tri;" u{dwr eij" o[noma patro;" kai; uiJou' kai; aJgivou pneuvmato": 4) pro; de; tou' baptivsmato" pronhsteusavtw oJ baptivzwn kai; oJ baptizovmeno": kai; ei[ tine" a[lloi duvnantai: keleuvei" de; nhsteu'sai to;n baptizovmenon pro; mia'" h] duvo: 8.1) aiJ de; nhstei'ai uJmw'n: mh; e[stwsan meta; tw'n uJpokritw'n: nhsteuvousi ga;r deutevra/ sabbavtwn kai; pevmpth/: uJmei'" de; nhsteuvsate tetravda kai; paraskeuhvn: 2) mhde; proseuvcesqe wJ" oiJ uJpokritaiv: ajllΔ∆ wJ" ejkevleusen oJ kuvrio" ejn tw'/ eujaggelivw/ aujtou': ou{tw" proseuvcesqe: pavter hJmw'n oJ ejn tw'/ oujranw'/: … 3) tri;" th'" hJmevra" ou{tw" proseuvcesqe: 9.1) peri; de; th'" eujcaristiva" ou{tw eujcaristhvsate: 2) prw'ton peri; tou' pothrivou: eujcaristou'mevn soi pavter hJmw'n … 3) peri; de; tou' klavsmato": eujcaristou'mevn soi pavter hJmw'n … 5) mhdei;" de; fagevtw mhde; pievtw ajpo; th'" eujcaristiva" uJmw'n: ajllΔ∆ oiJ baptisqevnte" eij" o[noma kurivou: kai; ga;r peri; touvtou ei[rhken oJ kuvrio": mh; dw'te to; a{gion toi'" kusiv: The clause hJ de; tou' qanavtou oJdov" ejstin au{th (but the way of death is this) is both a substitution (SM/SA) for, and an introduction to, the text in 5.1b–5.2. Note that hJ tou' qanavtou oJdov" (5.1a) is a renominalization [RN] of miva th'" qanavtou in 1.1a. Touvtwn aJpavntwn (all these things) in 5.2u, on the other hand, is a summary statement (SA) referring back to this section. The renominalization tevkna in 5.2u brings up again the addressee of chs. 3-4 and adds an additional element of cohesion to the section. An addition to the Two Ways instruction begins with the phrase peri; de; th'" brwvsew" (and concerning food), peri; dev being a typical heading in Greek and therefore a substitution on the meta-level (SM) for the instruction which follows (see below pp. 135; 139–40). Similarly peri; de; tou' baptivsmato" (and concerning baptism) in 7.1 is a heading and a substitution on the meta-level (SM) for the text in 7.1b–8.3 (see also Hellholm and Blomkvist, “Parainesis as an Ancient Genre-Designation,” 469); the clause immediately following, ou{tw baptivsate (thus baptize), with its use of a relative adverb, is a substitution on the abstraction level (SA) for this same text. Note, however, that the topics of ch. 8, fasting and prayer, are related only tangentially with baptism and joined to the text via the low-level semantic marker of word association and basic syntactical connectors. The clauses wJ" ejkevleusen oJ kuvrio" (as the Lord commanded) and ou{tw" proseuvcesqe (thus pray) in 8.2b both employ a relative adverb (a SA for the text which follows) and a verb of communication (an MS) to form a meta-narrative sentence/clause (MNS) with respect to the prayer which follows. Eujaggelivw/ (gospel), an external text, is thus a suppositio formalis. The heading peri; de; th'" eujcaristiva" (concerning the Eucharist) in 9.1a is a substitution (SM) for the text in 9.1b–10.7; the following clause, ou{tw eujcaristhvsate (thus give thanks), with its use of a relative adverb (an SA for the text which follows) and verb of communication (an MS), forms a meta-narrative sentence (MNS) for the text in 9.2–10.7. The phrases peri; tou' pothrivou (concerning the cup) (9.2a) and peri; de; tou' klavsmato" (concerning the bread) (9.3a) serve as headings for their respective instructions and are thus substitutions on the meta-level (SM).

2.3 Structurally Delimiting the Didache of H54

89

6ST1.1.1.1.1.2 Way of Death: 5.1–2 SA/SM6: hJ de; tou' qanavtou oJdov" ejstin au{th (5.1a) = 5.1b–2t (cataphoric) SA: touvtwn aJpavntwn (5.2u) = 5.1b–2t RN: hJ tou' qanavtou oJdov" (5.1a) = miva th'" qanavtou (1.1a) RN: tevkna (5.2u) = tevknon (3.1a, 3a, 4a, 5a, 6a) 5ST1.1.1.1.2 The conclusion of the recitation/warning: 6.1–2 4ST1.1.1.2 Addition to the Two Ways teaching: 6.3 SM4: peri; de; th'" brwvsew" (6.3a) = 6.3b–d (cataphoric) 3ST1.1.2 Instructions on various church rites/activities: 7.1–11.2 4ST1.1.2.1 Instructions regarding baptism with other additional Christian obligations: 7.1–8.3 SA/SM4:

peri; de; tou' baptivsmato" ou{tw baptivsate (7.1a) = 7.1b–8.3 (cataphoric) (SA [SM3 surr]/MS1: tau'ta pavnta proeipovnte" [7.1b] = 1.1–6.3 [anaphoric]) 5ST1.1.2.1.1 Instructions on performing baptism: (7.1–4) 5ST1.1.2.1.2 Instructions on fasting: (8.1) WA: aiJ de; nhstei'ai uJmw'n … nhsteuvsi … nhsteuvsate (8.1) < 7.4 5ST1.1.2.1.3 Instructions on prayer: (8.2–3) SA/MS1: wJ" ejkevleusen oJ kuvrio" ejn tw'/ eujaggelivw/ aujtou' (8.2b) = 8.2d–o (suppositio formalis) SA/MS1: ou{tw" proseuvcesqe (8.2c) = 8.2d–o SA + MS1 = MNS DP: readers = speakers of prayer; God = recipient of prayer WA: wJ" oiJ uJpokritaiv (8.2) < 8.1 4ST1.1.2.2 Instructions on celebrating the Eucharist: 9.1–10.7 SM4: peri; de; th'" eujcaristiva" (9.1a) = 9.1b–10.7 (cataphoric) SA/MS1: ou{tw eujcaristhvsate (9.1b) = 9.2–10.7 (cataphoric) SM4 + MS1 = MNS 5ST1.1.2.2.1 Instructions for the meal: 9.2–4 6ST1.1.2.2.1.1 Prayer for the wine: 9.2a–c SM6: prw'ton peri; tou' pothrivou (9.2a) = 9.2b–d (cataphoric) 6ST1.1.2.2.1.2 Prayer for the bread: 9.3–4 SM6: peri; de; tou' klavsmato" (9.3a) = 9.3b–4 (cataphoric)

90

Chapter 2: A Text-Linguistic Analysis of the Didache

9.1) peri; de; th'" eujcaristiva" ou{tw eujcaristhvsate: 2) prw'ton peri; tou' pothrivou: eujcaristou'mevn soi pavter hJmw'n uJpe;r th'" aJgiva" ajmpevlou Dau>i;d tou' paidov" sou: h|" ejgnwvrisa" hJmi'n dia; jIhsou' tou' paidov" sou: soi; hJ dovxa eij" tou;" aijw'na": 3) peri; de; tou' klavsmato": eujcaristou'mevn soi pavter hJmw'n uJpe;r th'" zwh'" kai; gnwvsew" h|" ejgnwvrisa" hJmi'n dia; jIhsou' tou' paidov" sou: soi; hJ dovxa eij" tou;" aijw'na": 4) w{sper h\n tou'to [to;] (add Gebhardt, et al.) klavsma dieskorpismevnon ejpavnw tw'n ojrevwn kai; sunacqe;n ejgevneto e{n: ou{tw sunacqhvtw sou hJ ejkklhsiva ajpo; tw'n peravtwn th'" gh'" eij" th;n sh;n basileivan: o{ti sou' ejstin hJ dovxa kai; hJ duvnami" dia; jIhsou' Cristou' eij" tou;" aijw'na": 5) mhdei;" de; fagevtw mhde; pievtw ajpo; th'" eujcaristiva" uJmw'n: ajllΔ∆ oiJ baptisqevnte" eij" o[noma kurivou: kai; ga;r peri; touvtou ei[rhken oJ kuvrio": mh; dw'te to; a{gion toi'" kusiv: 10.1) meta; de; to; ejmplhsqh'nai ou{tw" eujcaristhvsate: 2) eucaristou'mevn soi pavter a{gie uJpe;r tou' aJgivou ojnovmatov" sou ou| kateskhvnwsa" ejn tai'" kardivai" hJmw'n [with Bryennios, Harris, et al.; codex uJmw'n]: kai; uJpe;r th'" gnwvsew": kai; pivstew" kai; ajqanasiva": h|" egnwvrisa" hJmi'n dia; jIhsou' tou' paidov" sou[:] (add Bryennios, Harris, et al.) soi; hJ dovxa eij" tou;" aijw'na": 3) suv devspota pantokravtor (pantovkrator Harris) e[ktisa" ta; pavnta e{neken tou' ojnovmatov" sou: trofhvn te kai; poto;n e[dwka" toi'" ajnqrwvpoi" eij" ajpovlausin: i{na soi eujcaristhvswsin: hJmi'n de; ejcarivsw pneumatikh;n trofh;n kai; poto;n kai; zwh;n aijwvnion dia; tou' paidov" sou: 4) pro; pavntwn eujcaristou'mevn soi o{ti dunato;" ei\ su; [: soi;] (add Harnack, et al.) hJ dovxa eij" tou;" aijw'na": 5) mnhvsqhti kuvrie th'" ejkklhsiva" sou tou' rJuvsasqai aujth;n ajpo; panto;" ponhrou': kai; teleiw'sai aujth;n ejn th'/ ajgavph/ sou: kai; suvnaxon aujth;n ajpo; tw'n tessavrwn ajnevmwn: th;n aJgiasqei'san eij" th;n sh;n basileivan: h}n hJtoivmasa" aujth'/: o{ti sou' ejstin hJ duvnami" kai; hJ dovxa eij" tou;" aijw'na": 6) ejlqevtw cavri": kai; parelqevtw oJ kovsmo" ou|to": wJsanna; [codex wJ" ajnna;] tw'/ qew'/ Dau>ivd: ei[ ti" a{giov" ejstin ejrcevsqw: ei[ ti" oujk e[sti metanoeivtw: maranaqav ajmhvn: 7) toi'" de; profhvtai" ejpitrevpete eujcaristei'n o{sa qevlousin: 11.1) o}" a]n ou\n ejlqw;n didavxh/ uJma'" tau'ta pavnta ta; proeirhmevna devxasqe aujtovn: 2) eja;n de; aujto;" oJ didavskwn strafei;" didavskh/ a[llhn didach;n eij" to; katalu'sai. mh; aujtou' ajkouvshte: eij" de; to; prosqei'nai dikaiosuvnhn kai; gnw'sin kurivou. devxasqe aujto;n wJ" kuvrion: 3) peri; de; tw'n ajpostovlwn kai; profhtw'n kata; to; dovgma tou' eujaggelivou: ou{tw" poihvsate: In 9.5 peri; touvtou (concerning this) points both forward and backward: as the marker for a change in topic, periv looks forward, but the demonstrative pronoun touvtou is a substitution on the abstraction level (SA) for the command just given. Nonetheless, ei[rhken, a verb of communication forming a meta-communicative sentence/clause (MS), along with the SA creates a meta-narrative sentence/clause (MNS) preceding the citation of the Lord’s teaching. In 10.1 meta; de; to; ejmplhsqh'nai (after being filled [with food and drink]) is a temporal marker indicating the beginning of a new sub-text within 9.1–10.7. Indeed ou{tw" eujcaristhvsate, like the virtually identical clause in 9.1b, forms a meta-narrative sentence/clause (MNS) with respect to what follows. Since it is a part of the instructions for the Eucharist, however, it is one level deeper into the text than the clause of 9.1b. Again, that 11.1–2 serves as a conclusion to 1.1–10.7 is shown by the phrase tau'ta pavnta ta; proeirhmevna which points back to all the instructions in the text thus far. Indeed 11.2 shows that the text up to this point is generically a “didache.” The heading peri; de; tw'n ajpostovlwn kai' profhtw'n (concerning the apostles and prophets) in 11.3 is a substitution (SM) for the text in 11.4–15.3; ou{tw" poihvsate (thus do), with its use of a relative adverb, may be a substitution (SA) for this same text or may refer to the “gospel.” The introduction of a sub-text on the prophets here may be connected via word association (WA) to the short, abrupt appearance of prophets, articular and without introduction, in 10.7. The extensive treatment of the prophets beginning in ch. 11 raises the possibility of later redaction.

2.3 Structurally Delimiting the Didache of H54 5ST1.1.2.2.2 Prohibition of meal to unbaptized: 9.5 6ST1.1.2.2.2.1 Command excluding the unbaptized: 9.5a–b 6ST1.1.2.2.2.2 Citation of the Lord to support command: 9.5c SA/MS1: peri; touvtou ei[rhken oJ kuvrio" (9.5c) (cataphoric) SA + MS1 = MNS DP: the Lord (Jesus) = speaker 5ST1.1.2.2.3 Prayers following the meal: 10.1–7 SA/MS1: ou{tw" eujcaristhvsate (10.1b) = 10.2–6 (cataphoric) SA + MS1 = MNS EM: meta; de; to; ejmplhsqh'nai (10.1a) 4ST1.1.2.3 Conclusion: 11.1–2 TC: ou|n 5ST1.1.2.3.1 Receive “orthodox”: 11.1 5ST1.1.2.3.2 Re: others: 11.2 2ST1.2 Instruction regarding hospitality to itinerant ministers and to travelers, new settlers in the community, obligations to resident ministers, worship, and selection of ministers within the community: 11.3–15.3 SA/SM2: peri; de; tw'n ajpostovlwn kai; profhtw'n … ou{tw" poihvsate (11.3) = 11.4–15.3 (cataphoric) WA: profhtw'n (11.3) < 10.7 3ST1.2.1 Instructions regarding hospitality to itinerant ministers and to travelers, new settlers in the community and obligations to resident ministers: 11.3–13.7 4ST1.2.1.1 General teaching to act in accordance with gospel: 11.3 RN: tou' eujaggelivou = tw'/ eujaggelivw/ aujtou' (8.2b)

91

92

Chapter 2: A Text-Linguistic Analysis of the Didache

11.4) pa'" de; ajpovstolo" ejrcovmeno" pro;" uJma'" decqhvtw wJ" kuvrio": 5) ouj menei' de; [eij mh;] (add Harris, et al.) hJmevran mivan: eja;n de; h/\ creiva kai; th;n a[llhn: trei'" de; eja;n meivnh/ yeudoprofhvth" ejstivn: 6) ejxercovmeno" de; oJ ajpovstolo" mhde;n lamlbanevtw: eji mh; a[rton e{w" ou| aujlisqh'/: eja;n de; ajrguvrion aijth'/ yeudoprofhvth" ejstiv: 7) kai; pavnta profhvthn lalou'nta ejn pneuvmati, ouj peiravsete. oujde; diakrinei'te: pa'sa ga;r aJmartiva ajfeqhvsetai: au{th de; hJ aJmartiva oujk ajfeqhvsetai: 8) ouj pa'" de; oJ lalw'n ejn pneuvmati profhvth" ejstivn: ajllΔ∆ eja;n e[ch/ tou;" trovpou" kurivou: ajpo; ou\n tw'n trovpwn gnwsqhvsetai oJ yeudoprofhvth" kai; oJ profhvth" 9) kai; pa'" profhvth" oJrivzwn travpezan ejn pneuvmati, ouj favgetai ajpΔ∆ aujth'": eij de; mhvge yeudoprofhvth" ejstiv: 10) pa'" de; profhvth" didavskwn th;n ajlhvqeian eij a} didavskei ouj poiei' yeudoprofhvth" ejstiv: 11) pa'" de; profhvth" dedokimasmevno" ajlhqino;" poiw'n eij" musthvrion kosmiko;n ejkklhsiva": mh; didavskwn de; poiei'n o{sa aujto;" poiei', ouj kriqhvsetai ejfΔ∆ uJmw'n: meta; qeou' ga;r e[cei th;n krivsin: wJsauvtw" ga;r ejpoivhsan kai; oiJ ajrcai'oi profh'tai: 12) o}" dΔ∆ a]n ei[ph/ ejn pneuvmati dov" moi ajrguvria h] e{terav tina oujk ajkouvsesqe aujtou'. eja;n de; peri; a[llwn uJsterouvntwn ei[ph/ dou'nai mhdei;" aujto;n krinevtw: 12.1) pa'" de; oJ ejrcovmeno" ejn ojnovmati kurivou: decqhvtw: … 13.1) pa'" de; profhvth" ajlhqino;" qevlwn kaqh'sqai pro;" uJma'": a[xiov" ejstin th'" trofh'" aujtou' 2) wJsauvtw" didavskalo" ajlhqinov" ejstin a[xio" kai' aujto;" w}sper oJ ejrgavth" th'" trofh'" aujtou': 3) pa'san ou\n ajparch;n gennhmavtwn lhnou' kai; a{lwno": bow'n te kai' probavtwn labwvn: dwvsei" th;n ajparch;n toi'" profhvtai": aujtoi; gavr eijsin oiJ ajrcierei'" uJmw'n: 4) eja;n de; mh; e[chte profhvthn dovte toi'" ptwcoi'": 5) eja;n sitivan poih'/" th;n ajparch;n labw;n do;" kata; th;n ejntolhvn: 6) wJsauvtw" keravmion oi[nou h] ejlaivou ajnoivxa", th;n ajparch;n labw;n do;" toi'" profhvtai": 14.1) kata; kuriakh;n de; kurivou sunacqevnte" klavsate a[rton: kai; eujcaristhvsate prosexomologhsavmenoi ta; paraptwvmata uJmw'n: o{pw" kaqara; hJ qusiva hJmw'n (codex; uJmw'n Georgian, Const) h\/: 2) pa'" de; e[cwn th;n ajmfibolivan meta; tou' eJtaivrou aujtou', mh; sunelqevtw uJmi'n e{w" ou| diallagw'sin: i{na mh; koinwqh'/ hJ qusiva uJmw'n: 3) au{th gavr ejstin hJ rJhqei'sa uJpo; kurivou ejn panti; tovpw/ kai; crovnw/ prosfevrein moi qusivan kaqaravn: o{ti basileu;" mevga" eijmi; levgei kuvrio": kai; to; o[nomav mou qaumasto;n ejn toi'" e[qnesi: 15.1) ceirotonhvsate ou\n eJautoi'" ejpiskovpou" kai; diakovnou" ajxivou" tou' kurivou a[ndra" praei''" kai; ajfilarguvrou" kai; ajlhqei'" kai; dedokimasmevnou": uJmi'n ga;r leitourgou'si kai; aujtoi; th;n leitourgivan tw'n profhtw'n kai; didaskavlwn: 2) mh; ou\n uJperivdhte aujtouv": aujtoi; gavr eijsin oiJ tetimhmevnoi uJmw'n: meta; tw'n profhtw'n kai; didaskavlwn: 3) ejlevgcete de; ajllhvlou". mh; ejn ojrgh'/: ajllΔ∆ ejn eijrhvnh/: wJ" e[cete ejn tw'/ eujaggelivw/: kai; panti; ajstocou'nti kata; tou' eJtevrou mhdei;" laleivtw: mhde; parΔ∆ uJmw'n ajkouevtw e{w" ou| metanohvsh// Starting in 11.4, the markers delimiting the structure of the text are rather weak. Though semantically the sub-text 11.4–15.3 includes material beyond the scope of “apostles and prophets,” especially in chs. 14 and 15, there is no conclusion corresponding to the introduction of 11.3 at any point prior to that of 15.4. In 11.4–13.7 the repetition of pa'" dev (and variants), a syntactic signal, serves to connect the section. Renominalizations of tw'n ajpostovlwn (11.4– 6) and then profhtw'n (11.7–12; 13.1–7; 15.1–2), both introduced in 11.3, also serve to hold the larger sub-section together. The renominalizations in 15.1–2 of the figure of the didavskalo", first mentioned in conjunction with the figure of the prophet in 13.1, a connection continued here, add to the cohesion of the sub-sections though, like the unexpected prophets of 10.7, may be an indication of later redaction. Aside from these, however, markers are virtually non-existent in 14.1–15.3 – the text simply changes topics without warning. An SA/ MS(surr) occurs when the text from Mal 1 is quoted in 14.3. One weak marker, the text connector ou\n (therefore), introduces the sub-text 15.1–2, but its meaning there is uncertain.

2.3 Structurally Delimiting the Didache of H54

93

4ST1.2.1.2 Specific instructions concerning hospitality to itinerant ministers, prophets and travelers; stipulations regarding those who choose to settle in the community: 11.4–13.7 RP: pa'" dev, kai; pavnta, and variants 11.4a, 7a, 9a, 10a, 11a; 12.1; 13.1 5ST1.2.1.2.1 Hospitality toward itinerant apostles: 11.4–6 RN: pa'" de; ajpovstolo" (11.4a); oJ ajpovstolo" (11.6a) = tw'n ajpostovlwn (11.3) 5ST1.2.1.2.2 Discernment of prophets: 11.7–12 RN: kai; pavnta profhvthn (11.7a); oJ profhvth" (11.8c); kai; pa'" profhvth" (11.9a); pa'" de; profhvth" (11.10a, 11a) = profhvtw'n (11.3) 5ST1.2.1.2.3 Hospitality toward travelers and to those who wish to settle in the community: 12.1–5 5ST1.2.1.2.4 Tithing as the way to support prophets and teachers who take up permanent residence in the community: 13.1–7 RN: pa'" de profhvth" (13.1a); toi'" profhvtai" (13.3c); profhvthn (13.4a); toi'" profhvtai" (13.6b) = profhvtw'n (11.3) 3ST1.2.2 A group of miscellaneous instructions concerning worship and community life: 14.1–15.3 4ST1.2.2.1 The command to meet for worship and the necessity of community harmony for worship to be considered correctly performed: 14.1–3 5ST1.2.2.1.1 Meet weekly on the Lord’s Day: 14.1a 5ST1.2.2.1.2 Keep Eucharist pure: 14.1b–3 6ST1.2.2.1.2.1 Confession of sins: 14.1b 6ST1.2.2.1.2.2 Reconciliation of disputes: 14.2 6ST1.2.2.1.2.3 Quotation from the Lord to support command: 14.3 SA/MS1(surr): au{th gavr ejstin hJ rJhqei'sa uJpo; kurivou (14.3) (cataphoric) DP: kurivou (God) = speaker 4ST1.2.2.2 Selection of overseers/bishops and deacons: 15.1–2 RN: tw'n profhtw'n (15.1–2) = profhvtw'n (11.3) RN: didaskavlwn (15.1–2) = didavskalo" (13.2) TC: ou\n 4ST1.2.2.3 Instruction on relations with errant members: 15.3 RN: tw'/ eujaggelivw/ = tw'/ eujaggelivw/ aujtou (8.2b); tou' eujaggelivou(11.3) 2ST1.3 Conclusion: Always act according to the gospel: 15.4 RN: tw'/ eujaggelivw/ tou' kurivou hJmw'n (15.4) = tw'/ eujaggelivw/ aujtou (8.2b); tou' eujaggelivou (11.3); tw'/ eujaggelivw/ (15.3) RP: ou{ poihvsate wJ" e[cete ejn tw'/ eujaggelivw/ tou' kurivou hJmw'n 15.4b