Text-Critical and Hermeneutical Studies in The Septuagint PDF

Text-Critical and Hermeneutical Studies in the Septuagint
Supplements
to
Vetus Testamentum
Editor in Chief
Christl M. Maier
Editorial Board
r.p. gordon – j. joosten – g.n. knoppers – a. van der kooij – a. lemaire –
s.l. mckenzie – c.a. newsom – h. spieckermann – j. trebolle barrera –
n. wazana – s.d. weeks – h.g.m. williamson
VOLUME 157
The titles published in this series are listed at brill.com/vts

Text-Critical and Hermeneutical
Studies in the Septuagint
Edited by
Johann Cook
Hermann-Josef Stipp
LEIDEN • BOSTON
2012
Library of Congress Cataloging-in-Publication Data
Text-critical and hermeneutical studies in the Septuagint / edited by Johann Cook, Hermann-
Josef Stipp.
p. cm. — (Supplements to Vetus Testamentum v. 157)
Text-critical and Hermeneutical Studies in the Septuagint is the title of a bilateral research
project conducted from 2009 to 2011 by scholars from the universities of Munich (Germany)
and Stellenbosch (South Africa). The joint research enterprise was rounded off by a
conference that took place from 31st of August–2nd of September 2011 in Stellenbosch.
It was held in cooperation with the Association for the Study of the Septuagint in South
Africa (LXXSA). Scholars from Germany, the Netherlands, Ireland, Belgium, France, Canada
and the USA, as well as South Africa, delivered papers focusing on the history of the LXX;
translation technique and text history; textual criticism, and the reception of the Septuagint.
Includes bibliographical references and index.
ISBN 978-90-04-24078-0 (hardback : alk. paper)—ISBN 978-90-04-24173-2 (e-book : alk.
paper)
1. Bible. O.T. Greek—Versions—Septuagint—Congresses. 2. Bible. O.T.—Criticism,
interpretation, etc.—Congresses. 3. Bible. O.T.—Hermeneutics—Congresses. I. Cook, Johann.
II. Stipp, Hermann-Josef, 1954–
BS744.T488 2013
221.4’86—dc23
2012035185
This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters
covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the
humanities. For more information, please see www.brill.com/brill-typeface.
ISSN 0083–5889
ISBN 978-90-04-24078-0 (hardback)
ISBN 978-90-04-24173-2 (e-book)
Copyright 2012 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing,
IDC Publishers and Martinus Nijhoff Publishers.
All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV
provided that the appropriate fees are paid directly to The Copyright Clearance Center,
222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.
Fees are subject to change.
This book is printed on acid-free paper.

CONTENTS
Preface ................................................................................................................. ix
Abbreviations .................................................................................................... xi
List of Contributors ......................................................................................... xvii
Part One
History of the Septuagint in General
The Pentateuch in Greek and the Authorities of the Jews ................. 3

Arie van der Kooij
The Biblical Canon and Beyond: Theological and Historical

Context of the Codices of Alexandria .................................................. 21
Heinz-Josef Fabry
Part Two
Translation Technique and Text History
Revisiting the Rock: Tsur as a Translation of Elohim in

Deuteronomy and Beyond ...................................................................... 37
Melvin K. H. Peters
Judges 3:12–30: An Analysis of the Greek Rendering of Hebrew

Wordplay ....................................................................................................... 53
Hans Ausloos
B or not B? The Place of Codex Vaticanus in Textual History

and in Septuagint Research .................................................................... 69
Siegfried Kreuzer
Übersetzungstechnik und Textkritik in den Esdrasbüchern:

Hendiadyoin, Doppelübersetzungen und
Wiederholungsvariationen in 1 Esdr ..................................................... 97
Dieter Böhler SJ
vi contents
In Search of the Old Greek Text of 4 Maccabees .................................. 127

Robert J. V. Hiebert
The Relationship between the LXX Versions of Proverbs and Job ..... 145
Johann Cook
An Analysis of the Use of Hebel as a Metaphorical and Symbolic

Device as Interpreted in LXX Ecclesiastes .......................................... 157
Lawrence Lincoln
The θεός and κύριος Terms in the Isaiah Text and their Impact on
the New Testament: Some Observations ............................................ 173
Peter Nagel
Revisiting the Original Greek of Ezekiel 18 .............................................. 193

Harry F. van Rooy
Theology after the Crisis: The Septuagint Version of Daniel 8–12 ... 207
Martin Rösel
Part Three
Textual Criticism
The Value of the Septuagint for Textual Criticism of the Hebrew

Bible as Illustrated by the Oxford Hebrew Bible Edition of
1 Kings ............................................................................................................. 223
Jan Joosten
Gottesbildfragen in den Lesartendifferenzen zwischen dem

masoretischen und dem alexandrinischen Text des
Jeremiabuches .............................................................................................. 237
Hermann-Josef Stipp
Two Difficult Passages in the Hebrew Texts of Lamentations 5:

Text-Critical Analyses of the Greek Translation ............................... 275
Gideon R. Kotzé
Amos 5:26—Überlegungen zur Textkritik, Textgeschichte und

Übersetzung eines schwierigen Bibelverses ....................................... 297
Eberhard Bons
contents vii
Part four
Reception
Die Schriftzitate im ersten Christentum und die Textgeschichte

der Septuaginta: Ein Wuppertaler Forschungsprojekt .................... 311
Martin Karrer und Johannes de Vries
Ιουδιθ und Iudith: Überlegungen zum Verhältnis der

Judit-Erzählung in der LXX und der Vulgata ..................................... 359
Barbara Schmitz
PsalmsLXX and the Christian Definition of Space: Examples

Based on Inscriptions from Central Asia Minor ............................... 381
Cilliers Breytenbach
The Testament of Job as an Adaptation of LXX Job ............................... 395

Jessie Rogers
On Kingship in Philo and the Wisdom of Solomon ............................. 409

Jonathan More
The Text Form of the Isaiah Quotations in the Sondergut

Matthäus Compared to the Dead Sea Scrolls, Masoretic Text
and Septuagint ............................................................................................. 427
Gert J. Steyn
Die Rezeption von Jer 38:31–34 (LXX) in Hebräer 8–10 und dessen
Funktion in der Argumentation des Hebräerbriefes ....................... 447
Wolfgang Kraus
Index of Ancient Sources .............................................................................. 463

Index of Subjects .............................................................................................. 489
Preface
The heading of this volume is owed to the fact that it arose from a bilat-
eral research programme titled “Text-critical and Hermeneutical Studies
in the Septuagint”, which was conducted from 2009 to 2011 by Johann
Cook, from the Department of Ancient Studies at the University of
Stellenbosch, and Hermann-Josef Stipp, from the Department of Old Tes-
tament at the Faculty of Catholic Theology of the University of Munich.
The project was funded jointly by the National Research Foundation of
South Africa and the Internationales Büro des Bundesministeriums für
Bildung und Forschung (International Bureau of the Federal Ministry of
Education and Research). The shared research enterprise was rounded off
by a conference that took place from 31st of August–2nd of September 2011
at the Wallenberg Research Centre, Stellenbosch Institute for Advanced
Studies (STIAS), and was held in cooperation with the Association for the
Study of the Septuagint in South Africa (LXXSA). Scholars from Germany,
the Netherlands, Ireland, Belgium, France, Canada and the USA, as well as
South Africa, kindly accepted the invitation to deliver papers. Most of the
lectures read on this occasion have been incorporated in the collection.
Moreover, Eberhard Bons contributed an essay although he was unable to
participate in the conference.
As editors of this volume we would like to express our gratitude
towards various persons and institutions. The National Research Founda-
tion of South Africa and the Internationales Büro des Bundesministeriums
für Bildung und Forschung contributed substantially towards the costs of
the research project and the conference. The Universities of Munich and
Stellenbosch offered an environment highly conducive to fruitful research.
The Division of Research Development at the University of Stellenbosch
has been extremely helpful in obtaining financial and other assistance.
A word of thanks to Brigitte Cyster and Tonya Hall from the Department
of Ancient Studies at the US is in order for their unstinting efforts in mak-
ing the arrangements for the congress.
Further, Gert Steyn and Pierre Jordaan in their capacity as executive
committee members of the LXXSA were very supportive throughout this
whole endeavour. We are especially indebted to the participants of the
conference and those who made their contributions available for publi-
cation. We also thank Prof. Christl M. Maier, the editor-in-chief of the
x preface
Supplements to Vetus Testamentum, for accepting the collection for publi-

cation in this series. The publishers Brill, Inc. should also receive a word
of gratitude for publishing this volume. A final word of appreciation goes
to Gideon Kotzé, research assistant of the Department of Ancient Stud-
ies at the US, who took upon himself all the hard work of preparing the
manuscript for publication. Without his professional dedication this vol-
ume would not have been possible.
The Editors, Johann Cook, Dept of Ancient Studies, University of

Stellenbosch, and Hermann-Josef Stipp, Faculty of Catholic Theology,
University of Munich, June 2012.
Abbreviations
AASF Annales Academiae scientiarum fennicae

AB Anchor Bible
ABRL Anchor Bible Reference Library
AGBL Aus der Geschichte der lateinischen Bibel (= Vetus Latina: Die
Reste der altlateinischen Bibel: Aus der Geschichte der lateini-
schen Bibel)
AGJU Arbeiten zur Geschichte des antiken Judentums und des
Urchristentums
AnBib Analecta biblica
ANRW Aufstieg und Niedergang der römischen Welt: Geschichte und
Kultur Roms im Spiegel der neueren Forschung (eds. H. Temporini
and W. Haase; Berlin: De Gruyter, 1972–)
AnSt Anatolian Studies
ANTF Arbeiten zur neutestamentlichen Textforschung
ASP American Studies in Papyrology
ATSAT Arbeiten zu Text und Sprache im Alten Testament
BA Biblical Archaeologist
BASOR Bulletin of the American Schools of Oriental Research
BBB Bonner biblische Beiträge
BCH Bulletin de correspondance hellénique
BdA La Bible d’Alexandrie
BDAG W. Bauer, F. W. Danker, W. F. Arndt and F. W. Gingrich. Greek-
English Lexicon of the New Testament and Other Early Christian
Literature. 3rd ed. Chicago: University of Chicago Press, 1999
BDF F. Blass, A. Debrunner and R. W. Funk. A Greek Grammar of the
New Testament and Other Early Christian Literature. Chicago:
University of Chicago Press, 1961
BEATAJ Beiträge zur Erforschung des Alten Testaments und des anti-
ken Judentum
BETL Bibliotheca ephemeridum theologicarum lovaniensium
BHK Biblia Hebraica (ed. R. Kittel; Stuttgart: Württembergische
Bibelgesellschaft, 31937)
BHQ Biblia Hebraica Quinta (eds. A. Schenker et al.; Stuttgart:
Deutsche Bibelgesellschaft, 2004–)
BHS Biblia Hebraica Stuttgartensia (eds. K. Elliger and W. Rudolph;
Stuttgart: Deutsche Bibelgesellschaft, 1977)
xii abbreviations
BHT Beiträge zur historischen Theologie

Bib Biblica
BiOr Bibliotheca Orientalis
BIOSCS Bulletin of the International Organization for Septuagint and
Cognate Studies
BiPa.Supp Biblia Patristica Supplément
BKAT Biblischer Kommentar, Altes Testament
BN Biblische Notizen
BRev Bible Review
BWANT Beiträge zur Wissenschaft vom Alten und Neuen Testament
ByzZ Byzantinische Zeitschrift
BZ Biblische Zeitschrift
BZAR Beihefte zur Zeitschrift für Altorientalische und Biblische
Rechtsgeschichte
BZAW Beihefte zur Zeitschrift für die alttestamentliche Wissenschaft
CBET Contributions to Biblical Exegesis and Theology
CBQ Catholic Biblical Quarterly
CBQMS Catholic Biblical Quarterly Monograph Series
CCSL Corpus Christianorum: Series latina
CIG Corpus inscriptionum graecarum
CIIP Corpus Inscriptionum Iudaeae/Palastinae
CJZC Corpus jüdischer Zeugnisse aus der Cyrenaika
ConBOT Coniectanea biblica: Old Testament Series
COT Commentaar op het Oude Testament
CRAI Comptes rendus de l’Académie des inscriptions et belles-
lettres
CSCO Corpus scriptorum christianorum orientalium
CSEL Corpus scriptorum ecclesiasticorum latinorum
DCH Dictionary of Classical Hebrew (ed. D. J. A. Clines; Sheffield:
Sheffield Academic Press, 1993–)
DJD Discoveries in the Judaean Desert
DOP Dumbarton Oaks Papers
DSS Dead Sea Scrolls
EstBib Estudios bíblicos
ETAM Ergänzungsbände zu den Tituli Asiae Minoris
FAT Forschungen zum Alten Testament
FRLANT Forschungen zur Religion und Literatur des Alten und
Neuen Testaments
abbreviations xiii
Ges18 W. Gesenius and H. Donner. Hebräisches und Aramäisches

Handwörter-buch über das Alte Testament, 18th ed. Berlin /
Heidelberg / New York: Springer-Verlag, 2007
GKC W. Gesenius, E. Kautzsch and A. E. Cowley. Gesenius’ Hebrew
Grammar. 2nd ed. Oxford: Clarendon Press, 1910
HAL L. Koehler, W. Baumgartner and J. J. Stamm. Hebräisches
und aramäisches Lexicon zum Alten Testament. Leiden: Brill,
1967–1995
HBS Herders biblische Studien
HKAT Handkommentar zum Alten Testament
HNT Handbuch zum Neuen Testament
HR E. Hatch, H. A. Redpath, Concordance to the Septuagint and
Other Greek Versions of the Old Testament. 2 vols. Oxford:
Clarendon Press, 1897
HSM Harvard Semitic Monographs
HThKAT Herders theologisher Kommentar zum Alten Testament
HUCA Hebrew Union College Annual
IC Inscriptiones Creticae
ICC International Critical Commentary
IG Inscriptiones Graecae
IJO Inscriptiones Judaicae Orientis
IK Inschriften griechischer Städte aus Kleinasien
IOSCS International Organization for Septuagint and Cognate Studies
JBL Journal of Biblical Literature
JBS Jerusalem Biblical Studies
JIWE Jewish Inscriptions of Western Europe
JJS Journal of Jewish Studies
JNSL Journal of Northwest Semitic Languages
JQR Jewish Quarterly Review
JSem Journal of Semitics
JSHRZ Jüdische Schriften aus hellenistisch-römischer Zeit
JSJ Journal for the Study of Judaism in the Persian, Hellenistic, and
Roman Periods
JSJSup Supplements to the Journal for the Study of Judaism
JSOT Journal for the Study of the Old Testament
JSOTSup Journal for the Study of the Old Testament: Supplement
Series
JSPSup Journal for the Study of the Pseudepigrapha: Supplement
Series
xiv abbreviations
JSQ Jewish Studies Quarterly

JSS Journal of Semitic Studies
JTS Journal of Theological Studies
KAT Kommentar zum Alten Testament
KEK Kritisch-exegetischer Kommentar über das Neue Testament
KV Korte Verklaring van de Heilige Schrift
LA Letter of Aristeas
LBS Library of Hebrew Bible/Old Testament Studies
LEH J. Lust, E. Eynikel and K. Hauspie. Greek-English Lexicon of
the Septuagint. rev. ed. Stuttgart: Deutsche Bibelgesellschaft,
2003
LSJ H. G. Liddell, R. Scott and H. S. Jones. A Greek-English Lexicon.
9th ed. with revised supplement. Oxford: Oxford University
Press, 1996
LXX Septuagint
LXX.D Septuaginta Deutsch
LUÅ Lunds universitets årsskrift
MAMA Monumenta Asiae Minoris Antiqua
MSU Mitteilungen des Septuaginta-Unternehmens
MT Masoretic Text
NA27 E. Nestle. Novum Testamentum Graece. 27th ed. (eds. B. Aland
et al.; Stuttgart: Deutsche Bibelgesellschaft, 1993)
NETS New English Translation of the Septuagint
NICOT New International Commentary on the Old Testament
NIGTC New International Greek Testament Commentary
NovTSup Novum Testamentum Supplements
NRSV New Revised Standard Version
NT New Testament
NTA Neutestamentliche Abhandlungen
NTS New Testament Studies
OBO Orbis biblicus et orientalis
OHB Oxford Hebrew Bible
OLSJ Online Liddell-Scott-Jones Greek-English Lexicon
OT Old Testament
ÖTK Ökumenischer Taschenbuch-Kommentar
OTE Old Testament Essays
OTL Old Testament Library
OTS Old Testament Studies
OtSt Oudtestamentische Studiën
PAAJR Proceedings of the American Academy of Jewish Research
abbreviations xv
PG Patrologia graeca (ed. J.-P. Migne; Paris, 1857–1886)

PL Patrologia latina (ed. J.-P. Migne; Paris, 1844–1864)
POT De Prediking van het Oude Testament
PVTG Pseudepigrapha Veteris Testamenti Graece
QFRG Quellen und Forschungen zur Reformationsgeschichte
RB Revue biblique
RECAM Regional Epigraphic Catalogues of Asia Minor
RevQ Revue de Qumran
RHR Revue de l’histoire des religions
RNT Regensburger Neues Testament
RQ Römische Quartalschrift für christliche Altertumskunde und
Kirchengeschichte
RSV Revised Standard Version
SBFA Studium Biblicum Franciscanum: Analecta
SBFLA Studii biblici Franciscani liber annus
SBL Society of Biblical Literature
SBLDS Society of Biblical Literature Dissertation Series
SBLSCS Society of Biblical Literature Septuagint and Cognate Studies
SBS Stuttgarter Bibelstudien
SC Sources chrétiennes
SCS Septuagint and Cognate Studies
SEG Supplementum epigraphicum graecum
SemClass Semitica et Classica
SNTSMS Society for New Testament Studies Monograph Series
SOTSMS Society for Old Testament Studies Monograph Series
SPap Studia papyrologica
STDJ Studies on the Texts of the Desert of Judah
StPB Studia post-biblica
SubBi Subsidia biblica
SUC Schriften des Urchristentums
SVF Stoicorum veterum fragmenta
TAM Tituli Asiae Minoris
TANZ Texte und Arbeiten zum neuentestamentlichen Zeitalter
TDOT Theological Dictionary of the Old Testament (eds. G. J. Botterweck
and H. Ringgren; Grand Rapids: Eerdmans, 1974–)
TECC Textos y estudios “Cardenal Cisneros”
TENT Texts and Editions for New Testament Study
ThHK Theologischer Handkommentar zum Neuen Testament
ThQ Theologische Quartalschrift
ThT Theologisch tijdschrift
xvi abbreviations
ThWQ Theologisches Wörterbuch zu den Qumrantexten

TLG Thesaurus linguae graecae
TLZ Theologische Literaturzeitung
TSAJ Texte und Studien zum antiken Judentum
TU Texte und Untersuchungen
TUGAL Texte und Untersuchungen zur Geschichte der altchristlichen
Literatur
TWNT Theologische Wörterbuch zum Neuen Testament (eds. G. Kittel
and G. Friedrich; Stuttgart: Kohlhammer, 1932–1979)
VC Vigiliae christianae
VT Vetus Testamentum
VTSup Vetus Testamentum Supplements
WBC Word Biblical Commentary
WMANT Wissenschaftliche Monographien zum Alten und Neuen
Testament
WSt Wiener Studien
WTJ Westminster Theological Journal
WUNT Wissenschaftliche Untersuchungen zum Neuen Testament
ZAW Zeitschrift für die alttestamentliche Wissenschaft
ZDPV Zeitschrift des deutschen Palästina-Vereins
ZNW Zeitschrift für die neutestamentliche Wissenschaft und die
Kunde der älteren Kirche
list of contributors
Hans Ausloos Université catholique de Louvain, University

of the Free State
Eberhard Bons Université de Strasbourg
Dieter Böhler SJ Philosophisch-Theologische Hochschule
St. Georgen, Frankfurt am Main
Cilliers Breytenbach Humboldt University, Berlin
University of Stellenbosch
Johann Cook University of Stellenbosch
Johannes De Vries Kirchliche Hochschule, Wuppertal
Heinz-Josef Fabry University of Bonn
Robert J. V. Hiebert Trinity Western University, Langley
Jan Joosten University of Strasbourg
Martin Karrer Kirchliche Hochschule, Wuppertal
Gideon R. Kotzé University of Stellenbosch
Wolfgang Kraus Universität des Saarlandes, Saarbrücken
University of Pretoria
Siegfried Kreuzer Protestant University of Wuppertal
Lawrence Lincoln University of Stellenbosch
Jonathan More University of Stellenbosch
Peter Nagel University of Pretoria
Melvin K. H. Peters Duke University, North Carolina
Jessie Rogers Mary Immaculate College (University of
Limerick), Ireland
Martin Rösel University of Rostock
Barbara Schmitz Universität Würzburg
Gert J. Steyn University of Pretoria
Hermann-Josef Stipp Ludwig-Maximilians Universität, München
University of Stellenbosch
Harry F. Van Rooy North-West University, Potchefstroom
Arie Van Der Kooij Leiden University
Part One
History of the Septuagint in General

The Pentateuch in Greek and the Authorities of the Jews
Arie van der Kooij
It is generally held that the Septuagint version of the Pentateuch was

produced in Alexandria, in the first half of the third century B.C.E., but
questions such as its original setting and the reasons why this translation
was made are disputed. An intriguing question, hardly dealt with as far as
I know, concerns the issue on whose authority the Pentateuch was trans-
lated into Greek. In this contribution, I would like to address the question
of whether, as suggested by the Letter of Aristeas, the authorities of the
Jews in Jerusalem were involved in the translation project, or not.1
The Letter of Aristeas (hereafter LA) offers the most ancient answer to
the question about the original setting of the translation of “the Law of
the Jews”. According to this source,2 dating to the second half of the sec-
ond century B.C.E., the translation of the books of the Pentateuch was
part of the policy of the king Ptolemy II Philadelphus (282–246 B.C.E.) to
collect, if possible, all the books of the world. The royal librarian, Dem-
etrius of Phalerum, was commissioned to do so. Demetrius proposed to
include “the books of the Law of the Jews” (par. 30), for which a transla-
tion would be necessary. The king then sent a letter to the high priest of
the Jews, announcing his plan and requesting assistance. Eleazar, the high
priest, agreed to help, and sent a total of seventy-two translators, six men
of each tribe—men of good behaviour, experts in Hebrew and in Greek,
1 This contribution is based on the assumption that the Pentateuch was translated as
an entity, presumably so by a team of five (or six) translators which would account for a
variety of translation styles, and to some extent a lexical variety as well, on the one hand,
and for the fact that the five books are marked by a shared vocabulary as far as significant
terms are concerned, on the other. For another view, see M. Rösel, Übersetzung als Vollen-
dung der Auslegung: Studien zur Genesis-Septuaginta (BZAW 223; Berlin: De Gruyter, 1994),
257–258; C. G. den Hertog, “Erwägungen zur relativen Chronologie der Bücher Levitikus
und Deuteronomium innerhalb der Pentateuchübersetzung,” in Im Brennpunkt: Die Sep-
tuaginta. Studien zur Entstehung und Bedeutung der griechischen Bibel, Bd. 2 (BWANT 161;
eds. S. Kreuzer and J. P. Lesch; Suttgart: Kohlhammer, 2004), 216–228.
2 For an edition of the Letter of Aristeas, see e.g., A. Pelletier, Lettre d’Aristée à Philo
crate: Introduction, texte critique, traduction et notes, index complet des mots grecs (SC 89;
Paris: Cerf, 1962).
4 arie van der kooij
and learned in the Law—to Alexandria to prepare the translation. The

work was done on the isle of Pharos, under the direction of Demetrius of
Phalerum. The new version was read to the leaders of the Jewish commu-
nity in Alexandria, as well as to the Ptolemaic king. It was received most
favourably by both parties.
Because LA clearly bears the marks of an apologetic document, the
question of its historical reliability has been disputed. For a long time,
modern scholarship has been sceptical about any historical clue that LA
might contain, except for the idea that the translation was produced in
Alexandria. In keeping with this tendency, Thackeray developed the view
that the translation originally was made “for synagogue use”, arguing that
the Jews in Alexandria were in need of a Greek version of the Pentateuch
that could be used for reading purposes in their synagogue.3 This the-
ory has found wide acceptance and remains popular up to the present
day. Other scholars, like Brock and Perrot, have suggested that, although
the liturgical matter may have been part of the picture, the translation
arose out of the educational needs of the Jewish community.4 A more
recent theory brought to the fore by Pietersma—the so-called “interlinear
paradigm”—fits in with this proposal.5
These and other suggestions6 made are based on the assumption that
the translation project of the Jewish Law should be seen an inner-Jewish
phenomenon. Interestingly, recent research is marked by a tendency to
regard the promulgation of the Greek Pentateuch as being due to a Ptole-
maic initiative or encouragement, reflecting in this way a less sceptical
attitude towards LA. A good example of this tendency is the theory pro-
posed by a number of scholars, like Bickerman and Barthélemy, that the
initiative of the whole project was taken by the Ptolemaic court.7 The idea
3 H. St. J. Thackeray, The Septuagint and Jewish Worship: A Study in Origins (London:
Milford, 1923), 9.
4 S. P. Brock, “The Phenomenon of the Septuagint,” OtSt 17 (1972): 16; C. Perrot, La lec-
ture de la Bible dans la synagogue: Les anciennes lectures palestiniennes du shabbat et des
fêtes (Hildesheim: Gerstenberg, 1973), 143.
5 A. Pietersma, “A New Paradigm for Addressing Old Questions: the Relevance of the
Interlinear Model for the Study of the Septuagint,” in Bible and Computer: The Stellenbosch
AIBI-6 Conference. Proceedings of the Association Internationale Bible et Informatique “From
Alpha to Byte”. University of Stellenbosch 17–21 July, 2000 (ed. J. Cook; Leiden: Brill, 2002),
337–364.
6 See, e.g., J. Joosten, “Le milieu producteur du Pentateuque grec,” Revue des Études
Juives 165 (2006): 349–361.
7 E. Bickerman, Studies in Jewish and Christian History: Part 1 (AGJU 9; Leiden: Brill,
1976), 171–175; D. Barthélemy, “Pourquoi la Torah a-t-elle été traduite en Grec?” in Études
d’histoire du texte de l’Ancien Testament (OBO 21; Fribourg: Universitaires Fribourg, 1978),
the pentateuch in greek and the authorities of the jews 5
is that the king wanted to have a copy of the books containing the laws
and customs according to which the Jews in Alexandria and in Egypt were
allowed to live. Likewise, Modrzejewski has argued that “royal judges and
officials” needed access to the text of the Jewish Law in order to apply it
effectively.8 Other scholars are of the opinion that the translation proj-
ect rather was elicited by the cultural policy of the early Ptolemies. So
Rajak thinks of a translation made under Ptolemaic patronage,9 whereas
Fernández Marcos has argued that the scholarly milieu in Alexandria is
“the most fitting explanation for the origin of the translation”.10 These and
other proposals made are in line with LA insofar this document is marked
by the idea of a strong interaction between the cultural context in Alexan-
dria, on the one hand, and the religion of the Jews, on the other.11
A stimulating contribution in this regard represents the study by
Honigman, entitled The Septuagint and Homeric Scholarship in Alexandria
(2003). Her work offers a detailed and innovative analysis of LA, and it is
on this basis that she then develops her own thesis regarding the original
setting of the Greek version of the Pentateuch. She argues that the way
the origins of this translation are described in LA was strongly influenced
by the practice and ideology of Homeric scholarship in Alexandria. This
feature, she believes, opens up a window on the original setting of the
translation which was made for the sake of prestige, and not for reasons
of any religious need.
I am not going to discuss in this paper the theory advanced by Honig-
man—I have done so elsewhere12—but I agree with Honigman, and other
scholars referred to above, that in dealing with the complex issue of the
322–340; M. Harl, G. Dorival and O. Munnich, La Bible grecque des Septante: Du Judaïsme
hellénistique au Christianisme ancien (Paris: Cerf, 1987), 78 (“Il faut l’initiative officielle”
[Dorival]).
8 J. Mélèze Modrzejewski, The Jews in Egypt: From Rameses II to Emperor Hadrian
(trans. R. Corman; Princeton: Princeton University Press, 1997), 104–111.
9 T. Rajak, Translation and Survival: The Greek Bible of the Ancient Jewish Diaspora
(Oxford: Oxford University Press, 2009).
10 N. Fernández Marcos, “The Greek Pentateuch and the Scholarly Milieu of Alexan-
dria,” SemClass 2 (2009): 88.
11 Cf. also M. Rösel, Übersetzung als Vollendung der Auslegung; S. Kreuzer, “Entstehung
und Publikation der Septuaginta im Horizont frühptolemäischer Bildungs- und Kultur-
politik”, in Im Brennpunkt: Die Septuaginta. Studien zur Entstehung und Bedeutung der
Griechischen Bibel, Bd. 2 (BWANT 161; eds. S. Kreuzer and J. P. Lesch; Stuttgart: Kohlham-
mer, 2004), 61–75; A. Schenker, “Was führte zur Übersetzung der Tora ins Griechische?
Dtn 4,2–8 und Platon (Brief VII,326a-b),” in Die Septuaginta—Texte, Theologien, Einflüsse
(WUNT 252; eds. W. Kraus and M. Karrer; Tübingen: Mohr Siebeck, 2010), 23–35.
12 A. van der Kooij, “The Septuagint and Alexandrian Scholarship,” BiOr 68 (2011): 492–509.
origins of LXX Pentateuch the picture of LA in its main lines should not
be dismissed too easily. Of course, one could argue that the translation
itself is the only source available for the search after its origins. However,
the Greek version of the Law is only helpful if internal evidence can be
assessed in the light of external data regarding the period in which the
version supposedly was made (first half third century B.C.E.). As a mat-
ter of fact, LA is the only document from Antiquity that tells us a story
about the origins of the Greek Pentateuch. Honigman has shown that it is
marked by an antiquarian intent and that it also contains elements being
part of an oral tradition shared by members of the Jewish elite in Alex-
andria (“collective memory”). However, as she has also made clear, even
in the case of traditions used by the author, it remains difficult to show
whether they testify, in one way or another, to the original setting of the
translation.
In my view, the best way to proceed is to find out whether there are
specific data, external to LA, which may add to the plausibility of any
hypothesis based on LA.13 Any proposal made on the basis of LA can only
serve as a window on the original setting of the translation if there is some
evidence from sources other than LA which is in favour of the suggestion
made. In addition, it is important to see whether specific interpretive ele-
ments in the Greek version itself may fit in, or may even strengthen a
given proposal.
It is from this perspective that I would like to discuss the issue whether
the authorities of the Jews in Jerusalem, in particular the high priest,
played a role in the translation project. According to LA, the high priest
was the one who granted permission to translate the Law, who selected
the translators, and who is also presented as the prime interpreter of the
Law (par. 128–169). What to make of this? Is all this made up for propa-
gandistic reasons only, or does LA at the same time present a picture that
would make sense in the first half of the third century B.C.E.?
The Pentateuch, the Law of the Jews, was a most important part of
the literary heritage of the Jews, and also constituted, together with the
temple and the holy city, the identity of their religion.14 In view of its
13 Cf. Fernández Marcos, “The Greek Pentateuch,” 86.

14 There are good reasons to assume that the Pentateuch did already exist at that time
(ca. 300 B.C.E.). Chronicles, dating to the late fourth century B.C.E. (cf. G. N. Knoppers,
I Chronicles 1–9 [AB 12; New York: Doubleday, 2004], 116), clearly reflects a comprehensive
notion of the term Torah, encompassing not only Deuteronomy (as in Kings) but also
priestly laws of the Pentateuch. It likely presupposes the Pentateuch as an authoritative
writing (cf. e.g. M. Fishbane, Biblical Interpretation in Ancient Israel [Oxford: Clarendon
significance the question of whether the Greek version of it was made on

the authority of the high priest and his colleagues in Jerusalem suggests
itself.
II
The high priest is depicted in LA as the political leader of Judea, and,

as far as we know, this is in line with the historical reality of the early
Hellenistic period, the time when the Law of Moses was translated. This
picture is attested by several sources, such as the work of Ben Sira (ch. 50)
as well as by a source dating to the first decades of Ptolemaic rule: the
description of the Jewish nation by Hecataeus of Abdera, a Greek scholar
who lived in Alexandria, around 300 B.C.E. It has come down to us in the
work of Diodorus Siculus (Bibliotheca Historica 40.3).15 Its focus is on the
laws and customs of the Jewish people. The reader is told that Moses,
“outstanding both for his wisdom and for his courage,” took possession of
the land and founded cities, such as Jerusalem. Hecateaus then continues,
“he established the temple that they hold in chief veneration, instituted
their forms of worship and ritual, drew up their laws, and ordered their
political institutions.” Thus, also in this statement the laws of Moses are
clearly related to the issue of the polity (πολιτεία) of the Jewish nation.
Interestingly, as to the form of government Hecataeus tells his readers
that the priests were selected (by Moses) “to head the entire nation.” They
should not only occupy themselves with the temple and the sacrificial
cult, but were also “appointed to be judges in all major disputes.” It was
to them that Moses entrusted “the guardianship of the laws and customs”.
Hecataeus then states:
For this reason the Jews never have a king, and authority over the people is
regularly vested in whichever priest is regarded as superior to his colleagues
in wisdom and virtue. They call this man the high priest.16
Press, 1985], 264). The evidence from Qumran also points to a date of its final stage before
the third century B.C.E.; see S. White Crawford, “The Use of the Pentateuch in the Temple
Scroll and the Damascus Document in the Second Century B.C.E.” in The Pentateuch as
Torah: New Models for Understanding Its Promulgation and Acceptance (eds. G. N. Knop-
pers and B. M. Levinson; Winona Lake: Eisenbrauns, 2007), 301–317.
15 For the text, see M. Stern, ed., Greek and Latin Authors on Jews and Judaism. Vol. 1:
From Herodotus to Plutarch ( Jerusalem: Israel Academy of Sciences and Humanities, 1974),
26–29.
16 Ibid., 28. On this passage, see e.g. J. C. VanderKam, From Joshua to Caiaphas. High
Priests after the Exile (Minneapolis: Fortress Press, 2004), 118–122; and M. Brutti, The
The picture as presented by Hecataeus is that the Jewish nation was ruled
by “priests”, under the supreme direction of a high priest, a form of gov-
ernment which is regarded to be in accord with the laws of Moses.
The same view was expressed at another occasion of a much later date,
this time by leading Jews. I have in mind an incident that took place in
the year 63 B.C.E. when members of the Hasmonean house came before
Pompey, the Roman general, in Damascus with their dispute over the
kingship. As we are told by Diodorus Siculus:
Likewise the leading men, more than two hundred in number, gathered to
address the (Roman) general and explain that their forefathers [. . .] had sent
an embassy to the senate, and received from them the leadership of the Jews,
who were, moreover to be free and autonomous, their ruler being called
High Priest, not King. Now, however, these men were lording it over them,
having overthrown the ancient laws (τοὺς πατρίους νόμους) and enslaved the
citizens in defiance of all justice17
According to this passage, another party (“the leading men”) appeared
before Pompey, claiming that the ruler of the Jewish nation should not be
a king, but a high priest, because this type of leadership was considered
by them in agreement with the ancient, “ancestral” laws. Two forms of
government are at stake here—that of kingship, and that of aristocracy.
It is clear that the leading men were in favour of the latter, taken in the
sense of priestly rule, while rejecting the former as being a constitution
that brings about the enslavement of the citizens.
Thus, the sources just mentioned not only testify to the fact that the
Jewish nation of the time was ruled by a high priest, but also claim that
(high-)priestly rule was in line with the ancient laws, the Law of Moses. It
may be noted that whereas “the leading men” of Diodorus are focussing
on the figure of the high priest as leader—in contrast to the king—, the
passage of Hecataeus is more nuanced saying that it was the high priest
together with “the priests” who were ruling the nation. The latter ones
are obviously to be seen as leading priests, which is also in line with the
statement, by Hecataeus, that they were the judges “in all major disputes”.
This raises the question which priests might have been envisaged here.
I will come back to this question below.
evelopment of the High Priesthood during the pre-Hasmonean Period: History, Ideology,
D
Theology (Leiden: Brill, 2006), 138–141.
17 Diodorus Siculus, Bibliotheca Historica, 40.2. For this translation, see Stern, Greek and
Latin Authors 1, 185–186.
As noted above, in LA the high priest is also presented as the (prime)

interpreter of the Law. This feature too is in line with the evidence we
have. In the description of the Jews by Hecataeus of Abdera, referred to
above, the following note on the high priest is found:
They call this man the high priest, and believe that he acts as a messenger
(ἄγγελον) to them of God’s commandments. It is he, we are told, who in
their assemblies and other gatherings announces what is ordained, and the
Jews are so docile in such matters that straight way they fall to the ground
and do reverence to the high priest when he expounds the commandments
to them.18
The high priest is depicted here as a “messenger” of God’s command-
ments who in assemblies and gatherings announces and expounds what
is ordained. As we know from other sources of the time, the priests were
considered experts of the law and hence the ones who should teach the
people (see e.g. Sir 45:17; Jub 31:15).19 This sheds light on the statement of
Hecataeus regarding the leading priests being appointed as “judges” which
of course presupposes their expertise of the laws. As these priests were
ruling the nation under the supreme authority of the high priest, the latter
was considered the prime authority as far as the law and its interpretation
is concerned.
This picture of the high priest is also implied in the Wisdom of Ben Sira.
As has been observed by scholars, the passage on Simon, the high priest,
in Sir 50 is strongly linked to the hymnic passage on wisdom and the law
in ch. 24. Both wisdom (Law) and the high priest are said to “minister”
before God (24:10; 50:14). The specific link between the two passages indi-
cates that the high priest was seen as the main authority as far as wisdom
and the Mosaic Law were concerned.20 All this reminds one of the figure
designated, in documents of Qumran, as “the Interpreter of the Law”, the
more so since it is very likely that he is to be regarded a high-priestly
leader.21
18 Ibid., 28.

19 Cf. Deut 17:9; 33:10. For evidence from Qumran, see F. García Martínez, “Priestly
Functions in a Community without Temple,” in Gemeinde ohne Tempel/Community with-
out Temple. Zur Substituierung und Transformation des Jerusalemer Tempels und seines
Kults im Alten Testament, antiken Judentum und frühen Christentum (WUNT 118; eds. B. Ego,
A. Lange, and P. Pilhofer; Tübingen: Mohr Siebeck, 1999), 311–312 (“Priests as Judges”).
20 Cf. A. van der Kooij, “Authoritative Scriptures and Scribal Culture,” in Authoritative
Scriptures in Ancient Judaism ( JSJSup 141; ed. M. Popovic; Leiden: Brill, 2010), 69.
21 Cf. J. J. Collins, The Scepter and the Star. The Messiahs of the Dead Sea Scrolls and
Other Ancient Literature (New York: Doubleday, 1995), 144.
In the light of these (external) data it may be concluded that the pic-
ture of the high priest as leader and prime interpreter, as presented in
LA, concurs with the historical reality at the time when the translation of
the Law was made. Hence, in view of his role and position, one can imag-
ine that the Greek translation of the Law was made on the authority of
the high priest. This is even more plausible if one takes into account that
the Law was an official document and that, as has been argued convinc-
ingly by scholars, the translation should be seen as the result of an official
project.22
III
However, the question arises whether the translation itself contains any
piece of evidence that might shed light on the issue at stake. In this sec-
tion of my contribution, I will discuss a few passages in LXX Pentateuch
that seem to be of interest for our topic. The passages are Exod 19:6, and
23:21–23.
Exodus 19:6, part of which reads thus,23

LXX: “You shall be to me a royal priesthood and a holy nation”
MT: “You shall be to me a kingdom of priests and a holy nation”
The Hebrew expression “a kingdom of priests” (‫ )ממלכת כהנים‬has been
rendered in Greek as βασίλειον ἱεράτευμα. This is not the only rendering
possible; an alternative, and more literal, translation is found in the ver-
sion of Aquila: βασιλεἰα ἱερέων. As to the phrase in LXX it has been sug-
gested to take both words as two substantives (“kingdom”, “priesthood”),24
but scholars like Wevers and others have convincingly argued that Greek
βασίλειον is best understood here as an adjective (“a royal priesthood”).25
22 See A. van der Kooij, “The Septuagint of the Pentateuch and Ptolemaic Rule,” in The
Pentateuch as Torah. New Models for Understanding Its Promulgation and Acceptance (eds.
G. N. Knoppers and B. M. Levinson; Winona Lake: Eisenbrauns, 2007), 292–293 (with lit.).
23 For a discussion of this passage, see also A. van der Kooij, “A kingdom of priests:
Comment on Exodus 19:6,” in The Interpretation of Exodus. Studies in Honour of Cornelis
Houtman (CBET 44; eds. R. Roukema et al.; Leuven: Peeters, 2006), 173–175.
24 So J. H. Elliot, The Elect and the Holy (NovTSup 12; Leiden: Brill, 1966), 70–73. For this
reading/interpretation, see also below.
25 J. W. Wevers, Notes on the Greek Text of Exodus (SCS 30; Atlanta: Scholars Press, 1990),
295. Cf. G. Schrenk, “ἱεράτευμα,” in TWNT III, 249; J. Lust, E. Eynikel and K. Hauspie, Greek-
English Lexicon of the Septuagint, rev. ed. (Stuttgart: Deutsche Bibelgesellschaft, 2003);
A. Le Boulluec and P. Sandevoir, L’Exode (BdA 2; Paris: Cerf, 1989), 200; C. G. den Hertog,
The combination of adjective + noun, not that usual though, is attested at

other places in LXX Pentateuch as a type of rendering of a Hebrew expres-
sion consisting of stat. cstr. followed by a noun.26
The other term employed, ἱεράτευμα, occurs only three times in the
LXX—beside our text in LXX Exod 23:22 (in a large plus) and in 2 Macc
2:17. The word seems to be of late origin; it is only found in the LXX and
related Jewish or Christian writings (cf. Elliot). In distinction to the Greek
ἱερατεία, which occurs more often (e.g. Exod 29:9; 35:19) and denotes the
priesthood and the priestly office in general, the lexeme ἱεράτευμα is a
word which, just like στράτευμα and τεχνίτευμα, refers to a particular
group of people, a body, or college of priests.27
It is to be asked how the expression “a royal priesthood” is related to
the other phrase in the same verse, “a holy nation”. In the early Chris-
tian tradition, particularly so in 1 Peter 2:9, both expressions have been
taken as parallel phrases (1 Peter 2:9: “But you are a chosen race, a royal
priesthood, a holy nation”). This interpretation is also typical of the way
scholars have interpreted our text.28 For example, Wevers remarks that
“Exod’s ‘royal priesthood’ represents an ideal in which the priesthood is of
kingly stock, and in which all Israel constitutes such an ideal”.29 However,
from a Jewish perspective it is far from certain that “royal priesthood” in
Exod 19:6 should be understood as a designation of all Israel, parallel to
“Die griechische Übersetzung von Exodus 19:6 als Selbstzeugnis des frühhellenistischen
Judentums,” in The Interpretation of Exodus. Studies in Honour of Cornelis Houtman (CBET
44; eds. R. Roukema et al.; Leuven: Peeters, 2006), 184.
26 Cf. I. Soisalon-Soininen, Studien zur Septuaginta-Syntax (Helsinki: Suomalainen Tie-
deakatemia, 1987), 66.
27 Cf. Schrenk, “ἱεράτευμα,” 249 (“die Priesterschaft als Korporation”); T. Muraoka,
A Greek-English Lexicon of the Septuagint (Louvain / Paris / Walpole: Peeters, 2009), 338.
28 This also applies to the exegesis of MT Exod 19:6. This interpretation, however, has
been criticized in recent contributions; see in particular A. Graupner, “ ‘Ihr sollt mir ein
Königreich von Priestern und ein heiliges Volk sein’. Erwägungen zur Funktion von Ex
19,3b-8 innerhalb der Sinaiperikope,” in Moses in Biblical and Extra-Biblical Traditions
(BZAW 372; eds. A. Graupner and M. Wolter; Berlin: De Gruyter, 2007), 33–50 (“ein König-
reich, das von Priestern regiert wird” [43]).
29 Wevers, Notes on Exodus, 295. See also E. Schüssler Fiorenza, Priester für Gott: Stu-
dien zum Herrschafts- und Priestermotiv im der Apokalypse (Münster: Aschendorf, 1972), 83;
D. Munoz León, “Un reino de sacerdotes y una nación santa (Ex 19,6),” EstBib 37 (1978): 166
(he has the Jews of the diaspora in mind); O. Camponovo, Königtum, Königsherrschaft und
Reich Gottes in den Frühjüdischen Schriften (OBO 58; Freiburg / Göttingen: Vandenhoeck &
Ruprecht, 1984), 386; Le Boulluec and Sandevoir, L’Exode, 200; A. Schenker, “Drei Mosaik-
steinchen,” in Studies in the Book of Exodus: Redaction—Reception—Interpretation (BETL
126; ed. M. Vervenne; Leuven: Peeters, 1996), 374; W. Horbury, “Monarchy and Messianism
in the Greek Pentateuch,” in The Septuagint and Messianism (BETL 195; ed. M. A. Knibb;
Leuven / Paris / Dudley: Peeters, 2006), 91.
“a holy nation”.30 It seems to be more plausible to interpret our text from

the perspective the text was read in Jewish circles in Antiquity, such as
those behind the Targumim. The latter attest an interpretation according
to which the Hebrew phrase, “a kingdom of priests”, was understood as
referring to the leaders of the people, and not to the people as a whole:
Tg. Onq.: “kings, priests”
Tg. Neof. 1: “kings and priests”
Tg. Ps.-J.: “kings adorned with the crown, ministering priests”
These sources reflect a reading of the Hebrew expression by taking both
terms each on its own, as two substantives.31 It is to be noted that this type
of interpretation is also attested by a source dating to the late second cen-
tury B.C.E.—2 Maccabees. In 2 Macc 2:17 the Greek version of Exod 19:6
is quoted as follows: τὸ βασίλειον καὶ τὸ ἰεράτευμα. Here too both terms do
not refer to the people as a whole, but to the leadership, in this instance
to “the kingship” and “the priesthood” as two distinct institutions.32 Read-
ing from this perspective, the phrase “a royal priesthood” makes perfect
sense because a body, or college, of priests having a “royal” status strongly
evokes the notion of a body of leading priests.33 Interestingly, this is also
in line with other texts which testify to royal aspects of the priesthood,
such as 1Q21 (“kingdom of priesthood”) and the Aramaic Levi Document.34
Thus, our text seems to reflect the polity of the Jewish nation as ruled by a
particular body of priests, under the supreme direction of a high priest.35
30 The ancient Jewish literature does not offer an interpretation of Exod 19:6 in line
with that of the New Testament; see D. R. Schwartz, Studies in the Jewish Background of
Christianity (WUNT 60; Tübingen: Mohr Siebeck, 1992), 57–66.
31 Cf. Theodotion and Symmachus: regnum sacerdotes.
32 On this text, see A. van der Kooij, “The Use of the Greek Bible in II Maccabees,” JNSL
25 (1999): 129–130; T. Rajak, The Jewish Dialogue with Greece and Rome: Studies in Cultural
and Social Interaction (AGJU 48: Leiden: Brill, 2001), 52.
33 For a different interpretation of the phrase ‘royal priesthood’, see Den Hertog, “Die
griechische Übersetzung,” 186 (in his view, the adjective ‘royal’ indicates that Israel “als
eine einem nicht näher definierten externen βασιλεύς zugeordnete Priesterschaft bestimmt
wird”).
34 See M. E. Stone and J. C. Greenfield, “Aramaic Levi Document,” in Qumran Cave
4 XVII: Parabiblical Texts, Part 3 (DJD XXII; G. J. Brooke et al.; Oxford: Clarendon Press,
1996), 17.
35 Horbury (“Monarchy and Messianism,” 92) subscribes to this view, but is also of the
opinion that the phrase ‘royal priesthood’ first of all fits “the view of Israel as priestly
nation”.
Exodus 23:21–23
The passage of LXX Exod 23:20–23 reads in translation,36
20 And look, I am sending my angel in front of you in order to guard you
on the way in order to bring you into the land that I prepared for you.
21 Mind yourself, and listen to him, and do not disobey him. For he shall
not hold you in undue awe, for my name is upon him.
22 If by paying attention you listen to my voice and do all that I tell you,
I will be an enemy to your enemies and will resist those who resist
you.
23 For my angel will go, leading you, and will bring you in to the Amorrite
and Chettite and Pherezite and Chananite and Gergesite and Heuite
and Iebousite, and I will destroy them.37
This translation is based on the edition of Wevers, but, interestingly,
important MSS attest a text which includes a large plus prior to the first
word of v. 22:
If by paying attention you listen to my voice and do all what I command
you, and keep my covenant, you shall be for me a people special above all
nations. For all the earth is mine. And you shall be for me a royal priesthood
and a holy nation. These words you shall say to the sons of Israel.38
In LXX Exod 23:20–23 the figure of “my angel” dominates the scene. He is
designated in those terms both in v. 20 (MT without suffix, but see Sam-
Pent) and in v. 23 (cf. MT). God will send him in front of Israel, and he will
guard the people on the way to the land in order that they arrive in the
land safely. The underlying Hebrew text is about a heavenly messenger,
an angel,39 and it is commonly assumed that this also applies to the Greek
version. However, this version displays elements which seem to point in
another direction. V. 21 is of particular interest in this regard. It reads thus:
“Mind yourself, and listen to him, and do not disobey him. For he shall not
hold you in undue awe, for my name is upon him.”
36 The following discussion of this passage is a shortened version of A. van der Kooij,
“LXX Exodus 23 and the Figure of the High Priest,” in On Stone and Scroll. Essays in Honour
of Graham Ivor Davies (BZAW 420; eds. J. K. Aitken, K. J. Dell and B. A. Mastin; Berlin: De
Gruyter, 2011), 537–549.
37 NETS.
38 In the Septuagint edition of Rahlfs this plus is part of the Greek text. For a transla-
tion of the passage including the plus, see Le Boulluec and Sandevoir, L’Exode, 239. For a
discussion of the plus, see Van der Kooij, “LXX Exodus 23,” 539–541.
39 On the figure of the angel in the Hebrew text, see e.g., H. Ausloos, “The ‘Angel of
YHWH’ in Exod xxiii 20–33 and Judg. ii 1–5. A Clue to the Deuteronomistic Puzzle?”
VT 58 (2008): 1–12.
LXX, “Do not disobey him (μὴ ἀπείθει αὐτῶ)” for MT “do not show bit-
terness towards him”: The Greek version presupposes an interpretation
of the Hebrew different from that of the Masoretes. The latter reflects the
verb ‫( מרר‬cf. e.g. Sym), whereas the Greek phrase is based on the verb
‫( מרה‬cf. Deut 1:26; 9:7, 23, 24 [all cases Hiphil]; Isa 50:5; 63:10 [both cases
Qal]).
LXX “He shall not hold you in undue awe”, for MT “he will not pardon
your transgressions”: The Greek verb used here (ὑποστέλλομαι) means “to
draw, shrink back”. The idea is that “he will not refrain from judging you”.40
The Greek text alludes to the role of a judge, as is clear in the light of Deut
1:16–17 and of Job 13:8 as well.41
Deut 1:16–17 (LXX)

And I commanded your judges at that time, saying: “Give a full hearing
among your brothers, and judge rightly between a man and between his
brother and between his guest. You shall not recognize the person when
judging: like the small so you shall judge the great; you shall not shrink from
the face of a person, for the judgment is God’s. And the matter, if it be too
hard for you, you will bring it to me, and I will hear it.”
Job 13:8 (LXX)
Or will you shrink (from his face, i.e. God)? Nay, you should be judges
yourselves!42
(MT: Will you lift up his face? Will you contend for God?)
As a judge, the messenger will not be overawed by you. The Greek render-
ing presupposes a reading of the underlying Hebrew text different from
MT, namely: “he will not lift up (sc. your face [cf. ‫ )]נשא פנים‬as to your
transgressions”. The issue at stake here is the sensitive matter of showing
partiality in court, pronouncing someone guilty who is not, or the other
way around. Or to put it with the words of Prov 18:5: “It is not good to lift
up the face of the guilty”.
The final clause of the verse reads “My name is upon him (ἐπ᾽ αὐτῶ)” for
MT “my name is in him (‫”)בקרבו‬. According to Wevers, the Greek text is
marked here by an interpretation that is meant to avoid an identification
40 Wevers, Notes on Exodus, 370.

41 Compare also Wis 6:7.
42 NETS: “Really? Will you prevaricate? Go ahead; become judges yourselves!” I prefer
a translation in line with Deut 1, because of the “judges” in the second part of the Greek
text.
of the angel with the Lord. His name is not within him because “he is not
himself the Lord,” and he cannot forgive sins because only God can do
so.43 Ausloos considers this difference being one of the elements in the
Greek version that point to a weakening of the role of the angel.44 How-
ever, both scholars do not address the question of what the phrase, “my
name is upon him,” may mean. It is said of the angel, or messenger, that
the name of the Lord is upon him. It is to be asked to which figure this
idea might apply. In the light of the available data the most likely answer
to this question is: the high priest.
As we know from Exod 28:36 and 39:30, the high priest is the one who
carries the name of the Lord as part of an inscription (“Holy to YHWH”)
being engraved on a plate of pure gold which was fastened on the turban.
In LA when describing the apparel of the high priest (par. 96–99), the
relevant passage reads thus:
Upon his head he has what is called the ‘tiara’, and upon this the inimitable
‘mitre’, the hallowed diadem having in relief on the front in the middle in
holy letters on a golden leaf the name of God (par. 98)45
In this description the focus is on the name of God as being inscribed, no
reference being made to the term “holiness” (ἁγίασμα) as the first word
is rendered in the LXX (Exod 28:32; 36:38). The same idea is attested by
Josephus, B.J. 5.235; Ant. 3.178, the latter of which reads as follows: “the
forehead . . . had a plate of gold, bearing graven in sacred characters the
name of God”.46 The name was inscribed in “holy” (LA), or “sacred”
( Josephus) letters, that is to say, it was not put in square characters, but
in the Old Hebrew script. At another place Josephus provides us with a
striking illustration of the significance attached to the fact that the name
of God is on the high priest. It is a story about Alexander the Great who,
having taken Tyre after a long siege, and also Gaza, went up to the city
of Jerusalem (Ant. 11.326). Alexander’s meeting with the Jewish leaders, at
Mount Scopus, is described as follows:
43 Wevers, Notes on Exodus, 370. For the idea that the (unusual) rendering in Greek is
to be seen as an interpretation by the translator, see also R. Sollamo, Renderings of Hebrew
Semiprepositions in the Septuagint (AASF 19; Helsinki: Suomalainen Tiedeakatemia, 1979),
241.
44 H. Ausloos, “The Septuagint version of Exod 23:20–33. A ‘Deuteronomist’ at work?”
JSNL 22 (1996): 102.
45 Translation of R. J. H. Shutt, “Letter of Aristeas,” in The Old Testament Pseudepigrapha,
vol. 2 (ed. J. H. Charlesworth; New York: Doubleday, 1985), 19.
46 See also Origen, PG 12:1104; Tg. Ps.-J. Exod 29:6.
When Alexander while still far off saw the multitude in white garments
the priests at their head clothed in linen, and the high priest in a robe of
hyacinth-blue and gold, wearing on his head the mitre with the golden plate
on it on which was inscribed the name of God, he approached alone and
prostrated himself before the Name and first greeted the high priest [. . .] the
kings of Syria and the others were struck with amazement at his action and
supposed that the king’s mind was deranged. And Parmenion alone went up
to him and asked why indeed, when all men prostrated themselves before
him, he had prostrated himself before the high priest of the Jews, where-
upon he replied, “It was not before him that I prostrated myself but the God
of whom he has the honour to be high priest . . .” (Ant. 11.331–333).
According to this story, which is considered to be legendary, Alexander
prostrated before the high priest of the Jews because of the Name inscribed
on the head of the latter. The Macedonian king therefore greeted the high
priest first, before being greeted by him as one would expect. To be more
precise, Alexander did not prostrate himself before the high priest, but
before God, as he himself explains. This story which obviously reflects a
Jewish view, offers clear proof of the significance of the name of God on
the high priest, underlining the close relationship between him and God.
In the light of these data the phrase, “my name is upon him,” in the Greek
text of Exod 23:21 makes perfect sense if understood as referring to the
figure of the high priest.
It is to be asked however whether this also would apply to the term
ἄγγελος in v. 20 and v. 23. As such, this Greek word needs not to be taken
in the sense of a heavenly messenger, an angel, since it can also denote a
human messenger (see e.g. Gen 32:3). More importantly, there is evidence
that a priest could be designated that way. In Mal 2:7 the priest is called a
“messenger” of the Lord: “For the lips of a priest should guard knowledge,
and people should seek instruction from his mouth, for he is the messen-
ger of the Lord of Hosts”. Furthermore, as has been noted above, accord-
ing to Hecataeus of Abdera the Jews believed that the high priest acted “as
a messenger (ἄγγελον) to them of God’s commandments”.
All this fits the Greek text of Exod 23 well as it also sheds light on other
elements (modifications) in the passage. In v. 21, the people is urged to
listen and not to disobey the messenger of God of whom it is said that “he
shall not hold you in undue awe,” which, as we have seen, alludes to his
role as judge. It is true that in v. 22 it is not the messenger, but God and
his voice one should listen to, but, just as in the story about Alexander the
Great, our passage seems to be marked by the idea of a very close rela-
tionship between God and the high priest, particularly so because of the
Name being on the latter. It is interesting to note that this is fully in line
with the following statement of Josephus: “Any who disobey him—i.e. the
high priest—will pay the penalty as for impiety towards God himself ”
(C. Ap. 2.194).
Finally, according to v. 20, the angel/messenger will go in front of Israel
“in order to guard you on the way in order to bring you into the land that
I prepared for you” (cf. v. 23). If read as referring to the high priest, this
verse too makes good sense because the priestly leader—Eleazar, the son
of Aaron—was the one who entered the land together with the people (cf.
Num 27:18–23; Josh 22:13; 24:33).
IV
Both instances in LXX Exodus are very interesting as they testify to the
idea that the (leading) priests and the high priest are seen as governing
the Jewish people. They thus reflect a view of the constitution of the Jews
which is in line with the picture provided by Hecataeus of Abdera. For
the sake of clarity it is to be asked however which group of priests might
be envisaged as making up the body of leading priests within the Jewish
polity of the time.
A large number of priests, and Levites, were engaged in the temple ser-
vice, but given the hierarchy involved (see below) only a particular group
is to be regarded as representing the leading ones. It stands to reason to
think here of the priests who are often mentioned as accompanying the
high priest at official occasions. See e.g.:
(To our lord Bigvai, the governor of Judaea,) your servants Yedoniah and his
colleagues, the priests who are in Yeb the fortress (Cowley 30,1)47
(we sent a letter to your lordship and) to Johanan the high priest and his
colleagues, the priests who are in Jerusalem, and to Ostanes the brother of
Anani, and the nobles of the Jews (Cowley 30,18)48
(When he—the high priest—learned that Alexander was not far from the
city,) he went out with the priests and the body of citizens (Ant. 11.329)
With his colleagues [συνιερέων] he (i.e. the high priest, AvdK) will sacrifice
to God, safeguard the laws, adjudicate in cases of dispute, punish those con-
victed of crime (C. Ap. 2.194)
47 A. Cowley, Aramaic Papyri of the Fifth Century B.C. (Oxford: Clarendon Press, 1923
[repr. Osnabrück: Otto Zeller, 1967]), 113.
48 Ibid., 114.
From these examples we learn that the high priest together with “the
priests” is part of the leadership of the Jewish nation, the other part being
the nobles, the body of citizens. The latter college is comprised of the rep-
resentatives of the lay people, elsewhere referred to as “the elders”.49
Who were the priests that together with the high priest were head-
ing the nation? In my view, one has to think here of the priests who
are designated, both in the New Testament and by Josephus, ἀρχιερεῖς,
“chief priests.” They were the ones who constituted the highest level of
the priests acting in the temple, as we know from Josephus and rabbinic
sources.50 To quote Jeremias, the “chief priests permanently employed at
the Temple formed a definite body who had jurisdiction over the priest-
hood and whose members had seats and votes on the council (i.e. the San-
hedrin, AvdK)”.51 A writing from Qumran, 1QM, contains a passage which
is illuminating in this regard. 1QM 2:1–3 provides the following picture of
the priestly hierarchy in the temple:
– The chiefs of the priests behind the High Priest and of his second (in
rank), twelve priests to serve continually before God;
– The twenty-six chiefs of the divisions;
– The chiefs of the Levites to serve continually, twelve;
– The chiefs of their divisions.
The “chiefs of the priests”, representing the highest rank, are to be equated
with the chief priests.52 Therefore, it is likely that they were the ones who
together with the high priest constituted the priestly rule of the Jewish
nation. In the light of these data, it is reasonable to assume that the phrase
“a royal priesthood” refers to this body of leading priests.
These priests were also the ones who acted as judges because the priests
of the highest rank were members of the High Court in Jerusalem. This
is not only clear from the statement by Josephus quoted above, but also
from what Hecataeus tells us about the leading priests as being the ones
who were also appointed “judges in all major disputes”. The reference to
49 See, e.g., 1 Macc 7:33 and 1 Macc 14:28.

50 See J. Jeremias, Jerusalem in the Time of Jesus. An Investigation into Economic and
Social Conditions during the New Testament Period (London: SCM Press, 1976), 147–180.
51 Ibid., 180.
52 See A. van der Kooij, Die alten Textzeugen des Jesajabuches. Ein Beitrag zur Textge-
schichte des Alten Testaments (OBO 35; Freiburg / Göttingen: Universitätsverlag / Vanden-
hoeck & Ruprecht, 1981), 201.
“the major disputes” points to the High Court as described in Deut 17:8–9
(MT; cf. LXX). Both verses read thus,
If any case arises requiring decision between one kind of homicide and
another, one kind of legal right and another, or one kind of assault and
another, any case within your towns which is too difficult for you, then you
shall arise and go up to the place which the Lord your God will choose, and
coming to the Levitical priests, and to the judge who is in office in those
days, you shall consult them, and they shall declare to you the decision.
According to this passage, “the priests” and “the judge”, the latter being
a lay person,53 are the ones who as members of the High Court will deal
with the “difficult” cases. All this also sheds light on the presentation of
the high priest as judge in LXX Exod 23:21.
LXX Exodus contains evidence testifying to a view of the constitution

of the Jews which is in line with the description given by Hecataeus of
Abdera—a body of chief priests, under the supreme direction of the high
priest, heading the nation. As this view obviously serves the interests of
the leading priestly authorities in Jerusalem, it seems likely that the inter-
pretation of the relevant passages actually reflects their own understand-
ing of these texts.
As noted above, in view of his role as leader and prime interpreter of
the Law it seems only natural that the Greek version was made on the
authority of the high priest. If not, that is to say, if the version was made
on the authority of leaders of the Jewish community in Alexandria, it
would have been considered, by the leadership in Jerusalem, a rival one.54
However, it is hardly likely that local leaders in Egypt were in a position to
authorise a translation of the Law, the more so since the Jewish commu-
nity in Alexandria, in its early days, was not significant at all.55 Moreover,
at that time Judea and Jerusalem were part of the Ptolemaic empire. So,
even if leaders of the Jewish community in Alexandria played any role, the
53 Compare Deut 19:17 (“the priests and the judges”).

54 As was the case with the SamPent.
55 Cf. P. M. Fraser, Ptolemaic Alexandria I, Text (Oxford: Clarendon Press, 1972), 55;
S. Honigman, Septuagint and Homeric Scholarship in Alexandria: A Study in the Narrative
of the Letter of Aristeas (London: Routledge, 2003), 98–101.
two passages in LXX Exodus do not support the idea of a rival version. On
the contrary, they rather indicate that the translation of the Pentateuch
was made on the authority of the leading priests in Jerusalem, and of the
high priest in particular.56
56 Compare the role played by king Jehoshaphat who according to 2 Chron 17 sent
nobles, Levites and priests to teach ‘the book of the law of the Lord’ in the cities of Judah
(vv. 7–9).
The Biblical Canon and Beyond:
Theological and Historical Context of
the Codices of Alexandria
Heinz-Josef Fabry
Presented to Emanuel Tov on his 70th Birthday
1. The Question
After the problematisation of the MT induced by the Dead Sea Scrolls

(DSS) and their early manuscripts, we cannot “longer posit MT at the cen-
ter of our textual thinking”;1 rather, we have to devote more attention to
the Hebrew “Vorlage” of the LXX. “We should not (any longer) presume
that the OG of any particular book was translated from the Masoretic
form of that book”.2 The fact is that the text form of the Vorlage is much
older than that of MT.
The well known “textual variety” in the biblical DSS testifies to the
extended textual variability of the text of the Hebrew Bible and the
Christian Old Testament during ancient times. The textual history shows
a gradual phase of stabilisation. This stabilisation may be caused both by
text-internal reasons (such as corrections of obvious mistakes) and, more
so, by text-external reasons (such as the need for a unique text for the
purpose of theological dialogue and defence against heresies). The stabi-
lisation of the Greek text is visible in the context of quotations of the Old
Testament (OT) in the New Testament (NT).3 The Hebrew text became
more solid much later for inner-Jewish theological reasons and because of
1 E. Tov, “Hebrew Biblical Manuscripts from the Judaean Desert: Their Contribution to
Textual Criticism,” JJS 39 (1988): 5–37, esp. 7.
2 E. Ulrich, “The Relevance of the Dead Sea Scrolls for Hexaplaric Studies,” in Origen’s
Hexapla and Fragments (TSAJ 58; ed. A. Salvesen; Tübingen: J. C. B. Mohr [Paul Siebeck],
1998), 401–407, esp. 403.
3 LXX in it earliest form as “Old Greek” has—undoubtedly—preserved a very early stage
of the text in the mode of translation. But E. Tov has instructed us about the splitting of text
traditions, as obviously extant in the Greek manuscripts in the DSS (OG: 7Q-papLXXEx,
4QLXXLeva, 4QpapLXXLevb and 4QLXXNum) and 8ḤevXIIgr (Proto-Theodotion; kaige).
22 heinz-josef fabry
its dissociation from the Christian Bible in the time of the Masoretes from
the 6th until the 9th centuries C.E.4
Against this background, the heuristic hypothesis of the present paper
is as follows: After the Greek text became fixed in the first centuries C.E.
and in the time the MT became fixed, the intentional interventions into
the text of the Bible came to an end. After that time the only opportunity
to introduce Bible-hermeneutical ideas into the text consisted of changing
the order of the biblical books (over-all composition). We will start the
survey by pointing out that the big codices of Alexandria in the 4th and
5th centuries C.E. are examples of interpretation by composition.
Since the LXX text-line shows “textual variety”, E. Ulrich rightly asks: “It
is instructive, if perhaps unfair, to ask the question again this way: which
witness presents an earlier or more original text, the second- and first-
century B.C.E. Qumran manuscripts, or the fourth- and fifth-century C.E.
Vaticanus and Alexandrinus, or the medieval minuscules?”5 The codices
A, B, and S are actually not congruent at all, so that an exhaustive com-
parison of the texts would be necessary.
This is in fact our actual issue. It is striking that in Alexandria in the
4th century C.E. three Greek Bibles (OT and NT) were prepared in a very
short timeframe. The texts of these Bibles are very close, since they origi-
nate possibly from the same Greek textual source. Scholars have certainly
worked off the variants of the codices from a text-critical perspective. This
is not our aim today. Our task, instead, is trying to understand whether
and to what extent these three Greek Bibles may differ in the arrange-
ment of the books which are included in the codices. It is also evident,
furthermore, that such compositional differences also extend to the NT. In
this regard, I intend to demonstrate that the arrangement of the biblical
books is not arbitrary; it rather reflects a specific system, which is what
I am going to focus on here.
To my surprise, this question seems to be new and very unusual, since
there is no hint of it in the history of research in this field. The three codi-
ces are normally not issues of exegetical or of patristic research. Therefore,
I think that the time has come to study the diverging arrangement of the
biblical books.
4 H.-J. Fabry, “Das ‘Alte Testament’,” in “What is Bible?” Part II.: Between Torah and Bible
(eds. K. Finsterbusch and A. Lange; Gütersloh; Gütersloher Verlagshaus, 2012) (in press).
5 Ulrich, “The Relevance,” 406.
the biblical canon and beyond 23
2. Stages of the Early History of the Christian Bible Canon
2.1. Guidelines from Early Judaism

It is difficult to get reliable information on the Jewish canon—apart from
the anachronistic terminology—in the 1st century C.E. We know a few
things about several discussions and disputations about canonical items,
such as the “Synod of Jamnia”. This “Synod”6 discussed purely Jewish
issues; there was not any controversy with or about early Christianity.
And also we know nothing about any discussions or decisions about the
arrangement of the biblical books.
2.1.1. The first available pieces of evidence on the arrangement of the

biblical books in the early Jewish tradition come from Flavius Josephus
(C. Ap. 1.7f.).7 His principle of arrangement follows the precise chrono-
logical sequence (ἀκριβὴς διαδοχή) from the creation until Artaxerxes,
which extends over all three parts of the text: 5 Books of Moses, 13 pro-
phetical books ( Job-Esther) and 4 poetical books (Ps/Prov/Qoh/Song). All
these books are about the revelation of God and were written in the time
between Moses and Artaxerxes. Later books are without prophetical suc-
cession. The hermeneutical principle of the list of Josephus and later Tal-
mud is historically orientated: the world and history are creations of God,
who leads mankind through history until the present time. The world of
today should be understood as being the result of God’s history and His
Magnalia Dei.
2.1.2. A Baraita in b.Baba Batra 14b (5th/6th century C.E.) provides an

arrangement of the 22 biblical books that goes back to older traditions.8
The three parts of the biblical books are the same as in Josephus. But
now we discover an obvious regrouping of the books, a reduction in the
6 To the problems of a “Synod of Jamnia” cf. A. Hahn, Canon Hebraeorum—Canon

Ecclesiae. Zur deuterokanonischen Frage im Rahmen der Begründung alttestamentlicher
Schriftkanonizizät in neuerer römisch-katholischer Dogmatik (Studien zu Theologie und
Bibel 2; Berlin: LIT-Verlag, 2009), esp. 182–193.
7 Cf. W. Fell, “Der Bibelkanon des Flavius Josephus,” BZ 7 (1909): 1–16; R. Beckwith, The
Old Testament Canon of the New Testament Church and its Background in Early Judaism
(Grand Rapids: Eerdmans, 1985), 78ff.
8 Beckwith, Old Testament Canon, 154–162. Beckwith suggest that this list has its origin
in Judas Makkabäus (164 B.C.E.), but see the refutation given by P. Brandt, Endgestalten
des Kanons. Das Arrangement der Schriften Israels in der jüdischen und christlichen Bibel
(BBB 131; Berlin: Philo Verlagsgesellschaft, 2001), 64.
quantity of prophetical books, because the books of Job, Dan, 1 & 2 Chron,
Ezra/Neh and Esther are now allocated to the Hagiographa. The historical
arrangement is dropped. But over all we find a strong stability in the basic
arrangement: the Moses books (not yet classified as torah), prophets and
writings. That corresponds to the Jewish canon.
2.2. The Early Christian Discussion

We would like to have pieces of evidence about the Jewish Bible in the
time of Jesus and even about Jesus’ conception of canonical books. His
vivid disputation against the Pharisees (Matt 23:35) may be of some rele
vance, as Jesus speaks about the innocent blood of Abel (Gen 4) and of
Sacharja, son of Barachiasi (2 Chron 24:20–22); in fact, Jesus seems to
delineate a canonical outline from Genesis to 2 Chronicles. This single
piece of evidence, however, is not significant enough.9
It may be that there are a few lists of biblical books from early Christian
times (especially against the heretic position of Marcion), but the most
important list until now is the one of Melito, where we find the Chris-
tian idiosyncrasy of arranging the books in (a) historical, (b) wisdom and
(c) prophetical books. E. Zenger has told us about the main hermeneutical
differences in his “Introduction”.10 Later Church Fathers changed the pro-
phetical corpus in presenting first the Minor Prophets and then the Great
Prophets. In the following arrangements, one may detect some confusion,
but a closer investigation will bring the main arguments of the Fathers
to light.
First: The Hebrew Bible lost its primary role bit by bit, first in the West-
ern and later in the Eastern Church.11 The deuterocanonical books gradu-
ally became canonical.
Secondly: The middle position of the Prophets in the Jewish canons
was shifted very early to the final position by the Eastern Church Fathers
(Athanasius of Alexandria, Cyril of Jerusalem, Synod of Laodicea, Codex
B), with the aim to present the OT as the promise of Jesus Christ. Origen
9 Cf. L. M. McDonald, The Formation of the Christian Biblical Canon (Peabody: Hen-
drickson, 1995), 46f.
10 E. Zenger, Einleitung in das Alte Testament (ed. C. Frevel; Stuttgart: Kohlhammer,
82012), 22–36.
11 R. Hennings, Der Briefwechsel zwischen Augustinus und Hieronymus und ihr Streit um
den Kanon des Alten Testaments und die Auslegung von Gal. 2,11–14 (Leiden: Brill, 1993), 189:
“Im Osten wird zwar die Autorität der Septuaginta nicht bestritten, aber zumindest in der
Kanontheorie tritt sie dort hinter der hebräischen Bibel zurück.”
and Epiphanius steered a middle course. The Western Church (Synod of

Hippo, Synod of Carthago) held partly to the Jewish arrangement (S und A;
Hippo; Carthago against Codex B and Augustine).
Thirdly: The composition of the New Testament created a new situa-
tion, since now we have to respect mutual parallelisms and influences.
3. The Codices and Their Peculiarities
The Jewish arrangement of the biblical books remained nearly inflexible;

the Christian ones, on the contrary, remained modifiable until present
times. This is precisely the situation of the codices. An analysis of the
arrangements given by the Church Fathers clearly indicates that these
arrangements were not arbitrary.
3.1. H. B. Swete12 suggested that the arrangement of books according to

their chronology, literary nature, contents and authors reflects a typical
idiosyncrasy of the Alexandrian scholars. This explanation seems doubtful
to me. Therefore, we have to inquire about the reasons for the variations
in the arrangements. They are, as I argue, not arbitrary.
3.2. Our second issue is the quest for intentions. The canon lists until the
4th century were mostly homogeneous, but the arrangements in the codi-
ces departed from the established ways. Even if they did not reject their
Jewish roots, they revealed their peculiarity (a) in cancelling the tripartite
structure; (b) in regrouping the books of the prophets; (c) in accepting
the deuterocanonical writings; and (d) in compilating the books of the
Tanakh—now called the “Old Testament”—with the books of the early
Christianity—now called the “New Testament.” These compilations were
new Bibles with clear Christian intentions. They dismissed the historical
arrangement and depreciated the Pentateuch as Torah. It is amazing that
this new Christian hermeneutic was the starting point of a long develop-
ment; it had to grow and to try out several possibilities, and it needed
time to become consistent. Now we have to start our survey of the three
codices to find out how they disagree—despite their nearly concomitant
compilation of the books—and how they combine their Jewish and Chris-
tian hermeneutics.
12 H. B. Swete, An Introduction to the Old Testament in Greek (Cambridge: Cambridge

University Press, 1902 [repr. 1968]), Pt 2 Ch. 1.
4. Codex Vaticanus (B)
4.1. Arrangement of Books and Structure

Codex B is the oldest one; it reflects the tradition of the 39th Easter Letter
of Athanasius (367) and possibly served as the Bible for Athanasius, the
patriarch of Alexandria. B was written in Alexandria (or in Caesarea). The
Pentateuch, which is here understood as a collection of historical books,
is extended to an octateuch (Gen-Ruth), which is followed by a second
octateuch (1 & 2 Sam, 1 & 2 Kgs, 1 & 2 Chron, 1 Esd, Ezra/Neh). Subse-
quently the wisdom literature and the Trilogy of the Persian period follow
against the Jewish arrangement of the books. Tobit—whom Ambrosius
sees as a prophet13—prepares the transition to the corpus propheticum
(3rd Octateuch): first the Minor Prophets, after that the Great Prophets
with Baruch, Threni and EpJer. Daniel represents the apocalyptic conclu-
sion in the same way as in the lists of Cyril and Athanasius.
Athanasius designated Wisdom, Sirach, Esther, Judith and Tobit as
“ἕτερα βιβλία”, Cyril of Jerusalem as “ἀπόκρυφα”, but codex B now accepted
these books as fully canonical wisdom literature and gave them a place
between Job and the Minor Prophets. One may argue that this codex
reflects the biblical theologies of Athanasius and Cyril: moreover, the
place of wisdom literature before the prophetical books may reflect, on
the one hand, a conscious distance from the Jewish arrangement and, on
the other hand, a required dialogue with the philosophy in Alexandria.
4.2. The New Testament Books

The arrangement of the Old Testament books is to some extent parallel to
that of New Testament books, with Gospels, Acts (par. 2 historical octa-
teuchs), Letters of the Apostles (par. wisdom-octateuch) and the Book of
Revelation (par. prophetical octateuch). This is the arrangement given by
Cyril (sequence: Catholic letters-Pauline letters), which does not corre-
spond to the order of Athanasius. What is the ruling hermeneutic idea?
4.3. The Hermeneutical Concept

We have to be aware of the parallelism of both parts of the Bible. Appar-
ently, the book of Daniel stands in the final position (as in the list of the
13 Ambrosius, De Tobia (CSEL 32/II, Wien: Tempsky, 1897) 519.1 et seq.; 520.20 et seq.
Greek Church Fathers Athanasius, Cyril of Jerusalem, Synod of Laodicea)

and forms a possible parallel to the book of Revelation.14 Here one may
possibly find the clue to the pan-biblical hermeneutic. This book deals
with the eschaton, the end of secular powers, the expectation of chang-
ing of all. The book of Daniel amounts to its final sentence: “But go your
way till the end. And you shall rest and shall stand in your allotted place
at the end of the days” (Dan 12:13). Such a final conclusion of the first part
of the Bible is an impressive bridge to the New Testament. The motifs of
the eschatological καιρός and the resurrection are central points of the
New Testament as well; moreover, the concept of the Christians as real
heirs of the Old Testament (see Letter of Barnabas and Irenaeus, Bishop
of Lugdunum in Gaul [Lyon]) lies behind this allocation!
4.4. The Church-Historical Background

The determination of the church-historical background depends on the
precise dating of the codex. If we accept the first part of the 4th century,
then we are close to the time of the Council of Nicaea (325), convened
by Constantine the Great. The main issue addressed by the Council was
the heresy of Arius and the struggle against him. His chief opponent
was Alexander, Bishop of Alexandria. The codex is probably written by
order of Constantine himself.15 Perhaps it was a gift of Athanasius for the
Roman Bishop,16 when he became bishop of Alexandria. The main issue
for the Council was the formula ὁμοούσιος, the consubstantiality of Jesus
Christ with his Father. In this regard, the Fathers of the Council attached
importance to the decisions of the Egyptian theologians, because of their
old pharaonic traditions, where the Pharaoh as Horus was consubstantial
with the Sun-God Re-Harachte.
The next issue at the Council was the dispute about the correct date
for the Easter feast. According to the calendar indications of the OT, the
date of the feast corresponded to the Sunday after the 14th of Nisan. This
issue may also be a reason to finalise the arrangement of OT books with
Daniel and the Easter hope of resurrection (Dan 12:13). And finally, we
can assume that the delight about the restoration of unity in Christian
belief under Alexandrian guidance could have been an excellent reason
14 The book of Revelation took long time getting canonical acceptance.

15 T. C. Skeat, “The Codex Sinaiticus, the Codex Vaticanus and Constantine,” JTS 50
(1999): 583–625.
16 As often suggested by A. Rahlfs.
for preparing the wonderful and precious copy of “the unique Bible”, the
Codex Alexandrinus.
5. Codex Sinaiticus (‫א‬/S)

Codex S dates back to the first half of the 4th century. The text belongs
to the Alexandrian type, but displays a few Western readings. Since the
Codex is not complete, it is difficult to draw significant conclusions. Its
arrangement is close to the canon list of Athanasius, accepted later by the
Council of Laodicea. Here, however, one finds restrictions in accepting the
deuterocanonical writings included 1–4 Maccabees. Those texts follow, in
accurate historical sequence, the books Chron and Ezra; that means they
are historical books.
Since the main part of the historical books is lost, it is difficult to recon-
struct the initial structures of the Codex. It might be possible that the
historical section of the Codex (Gen-1 & 2 Ezra) was identical or similar
to that of Codex B and A. The trilogy from Persian times (Esther, Tobit,
Judith) and the (4?) Maccabees follow the historical part closely. After
that, however, the Codex unexpectedly goes its own way. Next in line is
not the wisdom literature—as in the other Christian codices—but, accord-
ing to the Jewish tradition, the prophets—first the Great Prophets, then
the Minor Prophets. Like the Jewish Bibles, the wisdom literature (not all
the ketûbîm though) forms the final books of the arrangement. The
Psalms—since the time of Ambrosius17 known as the Davidic prophecy
and the prophetic announcement of Jesus Christ—follow immediately
after the Minor Prophets.
Thus the main structure is close to that of the Talmud, where the wis-
dom literature also takes the final position. This alignment with the old
Jewish structure is surprising and, at the moment, enigmatic. It harks back
to the same tradition that is taken up again by Jerome, but now the book
of Job is at the end of the compilation. This is the unique characteristic of
the Codex Sinaiticus and how to interpret this unusual position is unclear.
Either Job is only a negligible annex—this is the assumption of Theodoret,
Bishop of Mopsuestia—or its position is of absolutely prime value. Since
17 Ambrosius, Explanatio Psalmorum I.8 (ed. M. Petschenig, Explanatio Psalmorum,

CSEL 64, Wien: Tempsky, 1999).
Theodoret’s impeachment against Job had no consequences in the later

literature, I am convinced that the final position of Job corresponds to a
very deliberately intended theological concept.
5.2. The Books of the New Testament

The arrangement of the books of the NT is very peculiar in this Codex:
at the beginning are the Gospels, the Pauline letters, the Letter to the
Hebrews and the Pastoral letters, next(!) the Book of Acts, the Catholic
letters, and finally the book of Revelation. Perhaps the New Testament
canon was still under dispute, because we find in final position the Letter
of Barnabas and the “Shepherd of Hermas”.

If we look at the OT part of this Codex, we see the development of pro-
phetical ethics on top of the historical foundation, which acts as their
basis. The arrangement proceeds from history through the prophets to
wisdom. Wisdom is no longer placed before the prophets: rather, it is
understood as the result of learning from history and from prophets. Wis-
dom is pointed out as the goal of the OT, the perspective to the scriptures
about Jesus Christ. Mankind needs a sound knowledge of the Magnalia
Dei to organise life successfully, with wisdom on the base of prophetical
ethics. By placing the book of Job (including the notice of the resurrec-
tion in Job 42:17aLXX) in the final position, the Codex intensifies the idea
of resurrection and states more precisely that only human beings like Job,
who made a great effort to achieve wisdom and a successful life, will have
a share in resurrection after death. In any case, mankind will find substan-
tial wisdom in Jesus Christ, the first of the resurrected ones.

The church-historical background of this Codex may be not substantially
different from Codex Vaticanus. S was written possibly in the histori-
cal context of the struggle of the Arians against the highly controversial
Athanasius, Bishop of Alexandria. His opponents were successful in expel-
ling him from Alexandria to go and live with the monks of the Egyptian
desert. The dispute about the Nicaean formula ὁμοούσιος replaced by the
formula ἀνόμοιος, proposed by Aëtius of Antiochia, found many support-
ers in the Synod of Sirmium. Basileios of Ankyra proposed the alterna-
tive draft ὁμοιούσιος. The oppressive Synod of Seleukia/Rimini (359) gave
the formula ὁμοῖος τῷ πατρί. That was the end of the Nicaean formula for
the moment. “The world groaned and was astonished to find itself Arian.”18
After the proclamation of Julian the Apostate as emperor in 360, violent
pogroms broke out against the Christians as “godless Galilaeans,”19 espe-
cially in Alexandria. Christian doctrine was consequently excluded from
the programmes of education and Christian teachers were expelled. This
historical context is a good place for the Sinaiticus.
5.4.1. First: the structure of this Bible will demonstrate where to look for
true education. The Sinaiticus understood “re-Judaization” as the chosen
way against the persecution of the Christian as “godless Galilaeans.”
5.4.2. Secondly: the Sinaiticus essentially intended to give back the Tal-

mudic Bible to the Christians, in order to enable them to find again the
true roots of the doctrine of God in the time of christological turmoil. In
view of the old thesis of Hellenistic philosophers “presbyteron kreitton,”20
the Christians had to know more and more about the God-Father in order
to know the God-Son. The only way to achieve this knowledge is to read
the Bible preferably in its original form with the revelation of the Father.
5.4.3. Thirdly: the persecution of Christians in the time of Julian the Apos-

tate produced some martyrs, who were very soon greatly respected and
worshipped. Like in the veneration of saints, the martyrs were put side by
side with the great figures of the OT: Abraham, Moses and David; perse-
cuted prophets, Eleazar and the Maccabees were also added. Their vio-
lent death was an anticipated martyrdom and testimony for Jesus Christ.21
This is the reason for the acceptance of the books of the Maccabees
(1–4 Macc) in the Codex.
5.4.4. Fourthly: if we remember Didymus the Blind of Alexandria, we have

to acknowledge that he called Job a “visionary seer” of the incarnation of
God, on the one side, but, like the Sadducees in NT, he had great difficul-
ties in thinking about resurrection, on the other side. The idiosyncrasy
18 “Ingemiscens Orbis terrarum se arianum esse miratus est,” Jerome, Adv. Lucif. 29.
19 Julian, Ep. 84.
20 To this criterion cf. P. Pilhofer, Presbyteron Kreitton. Der Altersbeweis der jüdischen
und christlichen Apologeten und seine Vorgeschichte (WUNT II/39; Tübingen: J. C. B. Mohr
[Paul Siebeck], 1990), 17–75, 143–220.
21 K. Baus, Handbuch der Kirchengeschichte II/1 (Freiburg: Herder, 1973), 338.
of the Sinaiticus is now to declare itself as the orthodox response to the

problematic of martyrs and to denounce Didymus in the course of the
struggle against the scholars of Origen.
6. The Codex Alexandrinus (A)

Codex A was written in the first half of the 5th century C.E. after Athana-
sius, and the tractates of Eusebius and Athanasius are in fact included in
the text of the psalms. Codex A is the youngest of the three.22 The text is
divided in several octateuches: (1) Gen-Ruth; (2) 1 Kgs-2 Chron (Ezra/Neh
are absent); (3) XII-Dan; (4) Esther-3 & 4 Macc; (5) Ps-Sir (incl. Odes). The
books of the prophets begin with the Minor Prophets (as in B and in con-
trast to the MT) and close with Daniel. The trilogy of the Persian period
follows, enlarged by Ezra and Maccabees.23
It is noticeable that the Corpus Propheticum is completely framed by
historical books. Perhaps the prophets are intended to be seen through the
lenses of the historians. As in Sinaiticus, the wisdom literature is arranged
next. The sequence Ps (David)/Odes/Job/Prov (Solomon) is noteworthy,
because it claims Job as the contemporary of Solomon and of the Queen
of Sheba.24 The final position of Sirach is the most enigmatic peculiarity
of this Codex.
6.2. The Books of the NT

The NT books are arranged in a way which is very similar to that of Codex
B: Gospels, Acts, Catholic letters, Pauline letters, Letter to the Hebrews,
Pastoral letters and the book of Revelation. The NT is enlarged by 1 & 2
Clemens and Pss Sol. The Pss Sol seem to be in a marginal position.
22 The Patriarch Cyril noted that the book was written according to the transmission
of Thecla, a noble Egyptian woman, shortly after the Synod of Nicaea, and that her name
could be found at the end of the Codex (now fragmented and lost). Even if such a dating
is surely too early, the relevance of this tradition consists in the reference to Egypt.
23 The Trilogy Ezra-Dan-Esther occurs here for the first time. The trilogy is well attested
from the 3rd century C.E. onwards (Chester Beatty IX und X).
24 Cf. R. Nathan, 2nd century C.E.

The final emphatic position of the wisdom literature points to the same
theological pedagogics as in Codex Sinaiticus. The interpretation of the
books of the Prophets as historical books seems very strange and inexpli-
cable. The search for any hermeneutical concept has to look at the final
position of Sirach, where we read: “Do your work before the appointed
time and at the appointed time (ἐν καιρῷ) he will give you your reward”
(Sir 51:30). The “appointed time”, καιρός, is the very distinctive concept of
NT scriptures. The NT books are closely bound to the OT books through
the significant καιρός terminology. The eschatological perspective of
Sir 51 is a close bridge to the eschatologically imminent expectation in
the Gospels.

The determination of the church-historical background depends here also
on the precise dating of the Codex, which can be determined only approx-
imately. One may refer to three possible historical contexts.
6.4.1. The turn of the century was characterised by the argument of the

doctrine of the trinity. The emperor Theodosius promulgated the Edict
“Cunctos populos” (380) in order to oblige Christians to adhere to the
formula of Nicaea and to the belief in the godly trinity. The Trinitarian
formula of the Council of Constantinople (381), however, missed the
ὁμοούσιος quality of the Holy Spirit. That was the reason for a long con-
flict between Theodoret of Kyrrhos and Cyril of Alexandria (375–444); the
right interpretation of the ὁμοούσιος of the son was disputed again. When
Nestorius became bishop (428) in Constantinople, however, a new front
against Antiochia was built up. That may be the actual context of Codex
Alexandrinus.
6.4.2. A second setting was shaped by Jerome and Augustine. Jerome was
Abbot of a monastery in Bethlehem and he was a zealous heir of Ori-
gen. But when Epiphanios of Salamis tried to condemn Origen, Jerome
changed his mind—to the astonishment of all. John, Bishop of Jerusa-
lem, and Rufin, translator of the works of Origen, became his opponents.
Jerome was encouraged by Theophil of Alexandria, who took up the
struggle against John. In Rome, Rufin wrote the voluminous work Apolo-
gia contra Hieronymum, to which Jerome responded with a multi-volume
polemic paper against Rufin. Theophil succeeded in forbidding the read-

ing of the books of Origen.
6.4.3. In Hippo, Augustine became bishop in 395. Together with Aurelius

of Carthage, he suggested the annual Synod of North African bishops. The
Synod of Carthage (398) adopted a canonical decision: Wisdom, Pss Sol,
Tobit, Judith, and 1 & 2 Macc were decided as fully canonical. Beside this
decision, Augustine himself developed his own list of canonical books.
He subdivided the OT into historical (Gen-2 Chron) and prophetic books
(Ps-2 Ezra). He allocated the wisdom literature to the prophets and Job to
the historical books.
It is quite possible that the Codex Alexandrinus has its place in the
context of these controversies. On the one side, the Alexandrinus differs
distinctly from Origen in regrouping the canon list of Origen. On the other
side, the Codex does not accept the list of Jerome, because of its arrange-
ment, which was very similar to the Jewish one. Finally, the Codex does
not agree with Augustine and the Synod of Hippo. In this way, Alexandria
proved its neutral and impartial stance—for the last time. A few years
later the patriarch Dioscuros was banished by the Council of Chalcedon
(451). After that, the theology of Alexandria was declared monophysitic
and heretical.
7. Perspective
7.1. This survey has hopefully shown that the different arrangements of the
biblical books in the codices are not only well reasoned, but also reflect
clear theological intentions. Despite their imprecise dating, the codices
seem to point to specific purposes and intentions, which plausibly reflect
the prevailing historical controversies and originate from them. Writers
and compilers were actually theologians, who discussed the christological
and trinitarian questions asked by dogmatically informed theologians, in
a prudent and wise manner.
7.2. Writers and compilers did not actively modify these texts; they only
regrouped them in different ways. In this regard, my survey has extended
the possibilities of composition criticism.
7.3. Apart from these analyses of the early Christian Bibles, we have to

keep in view the parallel development(s) of the Hebrew text until its
final standardisation as MT at the end of the 1st millennium. This text

remained relatively unmodified until the time of the Alexandrian codices.
At least since the 6th century a small but important regrouping should
be noticed. The book of Lamentations was separated from Jeremiah, thus
forming an independent book. Ruth was separated from Judges and its
historical context, and regrouped to the ketûbîm. That was a crucial point
on hermeneutical thinking. The obvious early arrangement of the histori-
cal books Gen-Ruth as octateuch was accepted by all Christian Bibles and
understood as the basic document of Israel’s self-consciousness and iden-
tity. In Christian thinking, there was a close bridge between that octa-
teuch and the Tree of Jesse, the family-tree of Jesus in Matt 1, where Ruth
has a prominent place. In this sense, the OT was nothing other than a
salvation-historical introduction to Jesus Christ. The midrashic Jewish tra-
dition transferred the book of Ruth somewhere in the ketûbîm and broke
down this bridge. The octateuch was reduced to the Pentateuch.
In order to be able to understand those processes, we have to go back
to the interaction between the Jewish and Christian Bibles; we have to
analyse such interaction very attentively and understand it as a vivid and
fruitful dialogue between the religions. Once again: there is still a lot of
work to be done.
Part Two
Translation Technique and Text History

Revisiting the Rock: Tsur as a Translation of Elohim in
Deuteronomy and Beyond
Melvin K. H. Peters
The title of this paper telegraphs its objective. Whereas “revisiting”

acknowledges that the subject has been much discussed,1 the subtitle
signals my intention to demonstrate that the representation of ‫ אלהים‬by
the epithet ‫( צוּר‬or ‫ )הצוּר‬throughout the Hebrew Bible was the result
of deliberate, theologically motivated, editorial or redactional activity by
Hebrew tradents whose work was later adopted in the early centuries C.E.
as the received text of the Hebrew Bible.2 Therefore, such Hebrew read-
ings are subsequent to, or distinct from, those reflected in the Hebrew
parent text of the Septuagint, which consistently reflects Elohim in every
instance.
In a recent paper3 presented at a conference celebrating the publica-
tion of the New English Translation of the Septuagint,4 I suggested that in
Deut 32 the Septuagint Vorlage was prior to the received text, and that
modifications preserved in MT were in some instances due to theological
sensitivities. Discussion of ‫ צוּר‬in that paper was limited to some cryptic
1 See, for instance, the entries on ‫ צוּר‬by Thiel and Fabry in TDOT and the extensive
literature cited there.
2 This claim can obviously not be demonstrated empirically. However, it rests on care-
ful observation of the patterns of deviation between the Leningrad Codex and the Hebrew
behind the Septuagint throughout the biblical corpus. It is based, furthermore, on the
demonstrable premise that the Hebrew text of the MT was not the only one in circulation
in the pre-Christian centuries and therefore was not always or necessarily the basis for
Septuagint readings. Textual pluriformity has now been well established, especially since
the discovery of the Dead Sea scrolls. On this, see for example, the articles in A. Schenker,
ed., The Earliest Text of the Hebrew Bible: The Relationship between the Masoretic Text and
the Hebrew Base of the Septuagint Reconsidered (SBLSCS 52; Atlanta: Society of Biblical
Literature, 2003) especially E. Tov’s “The Nature of the Large-Scale Differences between the
LXX and MT S T V, Compared with Similar Evidence in Other Sources,” 125–143.
3 M. K. H. Peters, “Translating a Translation: Some Final Reflections on the Produc-
tion of the New English Translation of Greek Deuteronomy,” in Translation is Required:
The Septuagint in Retrospect and Prospect (ed. R. J. V. Hiebert; Atlanta: Society of Biblical
Literature, 2010), 119–134.
4 A. Pietersma and B. G. Wright, eds., A New English Translation of the Septuagint and
the Other Writings Traditionally Included under That Title (Oxford / New York: Oxford Uni-
versity Press, 2007).
38 melvin k. h. peters
comments, which are now amplified and supplemented with a more care-
ful examination of the entire phenomenon in Deuteronomy and beyond.
But first, some statistics.
The root ‫ר‬-‫ו‬-‫ צ‬appears 217 times in the Hebrew Bible in 19 different
forms. The largest group consists of the simple noun ‫“ צוּר‬rock,” occurring
73 times, followed by various verbal, adjectival or other nominal forms
such as place or personal names. Within the main cluster of nouns are
thirty-six5 where the basic meaning “rock” seems inappropriate, appear-
ing instead to be a substitute for the divine name. Exactly half of those
instances are in the Psalter; the others are scattered between Deut (7), 1&2
Sam (6), Isa (4) and Hab (1). All but one (Isa 30:29) are situated in “poetic”
passages. The first occurrence of ‫ צוּר‬6 is in Exod 17:6 where the god of
Israel is portrayed as standing ‫“ על הצור בחרב‬on the rock at Horeb.”
This may well be the basis for the identification of ‫ אלהים‬with ‫ צוּר‬in later
traditions. The first instance of the specialized use of the term occurs in
Deut 32:4, which in the LXX is rendered θεός; in the received tradition as
‫“ הצוּר‬the Rock.”
The main question of this paper then, is this: Did the translator of Greek
Deuteronomy have before him a text identical to the Leningrad Codex
and intentionally avoided transferring ‫ צוּר‬into Greek (or at least trans-
literating it) or does the translation θεός indicate the existence of a text
containing ‫אלהים‬, which the LXX translator faithfully rendered? Histori-
cally, the overwhelming response to that question in critical scholarship
has been that the Greek translator indeed had the received text before him
and was interpreting or even commenting upon it. The standard narrative
runs roughly as follows. Hellenistic Septuagint translators recoiled at the
presence of such “crass metaphors” for the divine, preferring instead to
use the name itself. Here for example is the critical commentary on this
verse offered by Carmel McCarthy in her recently published edition of
Biblia Hebraica Quinta of Deuteronomy.
“Rock.” When used figuratively of God, was always a problem for the early
translators. G and the other versions in varying typical ways, reinterpret this
5 Deut 32:4, 15, 18, 30, 31(2x), 37; 1 Sam 2:2; 2 Sam 22:3, 32, 47(2x); 23:3; Isa 17:10; 26:4;
30:29; 44:8; Hab 1:12; Ps 18(17):3, 32, 47; 19(18):15; 28(27):1; 31(30):3; 49(48):15; 62(61):3, 7, 8;
71(70):3; 73(72):26; 78(77):35; 89(88):27; 92(91):16; 94(93):22; 95(94):1; 144(143):1.
6 This presumes that the Pentateuch was the first unit collected in the Hebrew Bible,
and that the current order of the books reflects the order of their composition.
revisiting the rock 39
metaphor according to its sense here and at vv. 15, 18 and 30–31 below. See
also v. 37 regarding its use in relation to pagan gods.7
Quite apart from McCarthy’s unqualified use of the word “God,” at least
two striking things are implicit in her comments. First, is the clear notion
that the L text, the so-called MT,8 was the basis for the Septuagint transla-
tion, and second that the translators of “versions” presumably “consulted”
each other to determine their responses to textual problems within it.
Neither of those presuppositions is necessary or compelling.
The Septuagint translators being first had no other version with which
to consult. Translators of subsequent versions could have been aware of
the Greek and influenced by it, but the reverse was not possible. LXX
translators were flying solo; their only points of reference would have
been Hebrew. Indeed, the choice of the word “versions” in so much of
contemporary text critical discourse continuously reinforces the mislead-
ing idea of a single “original” from which copies were made. So, when
modern scholars group together readings found in disparate so-called ver-
sions and draw textual conclusions based on apparent shared patterns,
they risk missing important clues in specific forms of the text, especially
the Septuagint.
But McCarthy is not alone; several, perhaps most, scholars share such
views. Here for instance is John Wevers some years earlier in his Notes on
the Greek Text of Deuteronomy on this verse:
The term ‫ צוּר‬applying to God occurs not only in this verse, but also at
vv. 15, 18, 30, 31, and 37. The word means “rock,” but it is never translated
thus in LXX, which substitutes θεός throughout, except for v. 37 which see.
Presumably the term was used as a poetic name for God to designate the
deity as the one who was solid, firm, unmoveable, but the translator consis-
tently avoided a direct translation, thereby precluding any possible misun-
derstanding of the metaphor.9
Like McCarthy, Wevers uses the word God uncritically, adding theological
characteristics, which he presumes ancients held about divinity. He states
7 C. McCarthy, Biblia Hebraica Quinta editione cum apparatu critico novis curis elabo-
rato, Fascicle 5, Deuteronomy (Stuttgart: Deutsche Bibelgesellschaft, 2007), 139*–140*.
8 Hereafter MT is used only as a convenient equivalent for the Leningrad Codex, and is
not intended in any way to suggest either that there was a single Masoretic text or that the
text of L was (or should be) considered normative at the time the Septuagint Pentateuch
was translated.
9 J. W. Wevers, Notes on the Greek Text of Deuteronomy (SCS 39; Atlanta: Scholars Press,
1995), 510.
that ‫ צוּר‬is substituted by θεός throughout the LXX, a claim that we will
examine presently. Quite clearly Wevers had no doubt that it was the Sep-
tuagint translator who avoided a direct translation in order to preclude
any possible misunderstanding of the metaphor, and that a text identical
with the MT was before him.
The name most often identified with the subject of this paper is Staffan
Olofsson whose dissertation book,10 God is My Rock: A Study of Translation
Technique and Theological Exegesis in the Septuagint, was published in
1990. Olofsson was concerned with theology in the Psalter primarily and
surveyed in the course of his work the use of certain epithets including
‫ צוּר‬in other parts of the Hebrew Bible. His book, as its full title indicates,
is representative of a wide range of scholarship that makes a priori judg-
ments about theological tendencies in the Septuagint and assigns char-
acteristic patterns to its translators as a group.11 Embedded in that type
of scholarship are presuppositions about the activities and motivations
of the translators of the Septuagint. For instance, that there was theo-
logical exegesis within the Septuagint constitutes a starting point for this
kind of research. Conclusions are thus arrived at deductively, not induc-
tively. Septuagint translators are sometimes described as incompetent, at
other times as wildly radical in their responses to the text before them.
Of course that text is always the Hebrew of the Leningrad Codex. Only
rarely (and grudgingly) do such scholars grant the possibility of a differ-
ent Vorlage; every effort is made to conform the Greek translation to the
reading of the so-called MT.
Increasingly, I have grown suspicious of subjective explanations by
modern textual critics for the differences observed in ancient witnesses
to the text of the Hebrew Bible. These explanations are frequently too pre-
dictable and at times so speculative as to be of little value. I have suggested
in another place12 a covert reason for the traditional privileging of Hebrew
over Greek when they diverge. Especially after Qumran, where it has been
clear for some time that multiple textual traditions were in circulation
in the pre-Christian centuries, the persistence of “MT fundamentalism”
10 S. Olofsson, God is My Rock: A Study of Translation Technique and Theological Exegesis
in the Septuagint (ConBOT 31; Stockholm: Almqvist & Wiksell International, 1990). Also
frequently cited is A. Passoni Dell’Aqua “La metafora biblica di Dio come rocca e la sua
soppressione nelle antiche versioni,” Ephemerides Liturgicae 91 (1977): 417–453.
11 Cf. the literature cited in Olofsson, God is My Rock, 167–184 for an overview of the
state of scholarship at that time.
12 See Peters, “Translating a Translation,” 132–134.
seems particularly puzzling and out of step with our contemporary state
of knowledge. The standard argument in defense of current text critical
practice, as I have heard it, is simply this: unless and until we can unearth
a Hebrew text that corresponds precisely to the readings in the Septuagint
in every instance, we will suspect and blame the Septuagint translator
whenever there is variance between the Greek and the Hebrew of Codex
Leningrad.
Now let me be clear. I am not a blind defender of Septuagint readings;
I am fully aware of recensional and revisionistic activity within the his-
tory of the transmission of the Septuagint—activity that invites caution in
regard to uncritical acceptance in every instance of the Greek text as we
now have it. I am also aware of Christian interpolations and interpreta-
tions in parts of the Old Greek text. It also has been shown that Hellenistic
influences are clearly present in parts of the Septuagint, particularly with
respect to personal and place names and the like. But those facts do not
place the unified reading of all existing Greek witnesses under suspicion
at all times in my view. I respect the work of the LXX translators and trust
that what they said is indeed what they meant. They did not create alter-
nate readings in order to be malicious, or to annoy and frustrate research-
ers two thousand years later; they were engaged in a serious enterprise.
The fiction of the “dishonest”13 translator flourishes only as long as we
continue to maintain the idea of the infallible and immutable MT.
In this first case, Deut 32:4, the noun ‫ הצוּר‬is not at all difficult; it did
not pose a problem in this or any other context for the translator of Deu-
teronomy, as McCarthy claims. Nor was the translator “consistently avoid-
ing a direct translation, thereby precluding any possible misunderstanding
of the metaphor,” as Wevers claims. The translator knew the noun and
had rendered it accurately in other places. One need look no further than
v. 13 to find a clear example of ‫ צוּר‬in parallel with ‫ סלע‬both appropriately
translated as πέτρα. Furthermore, a closer look reveals that while three
divine names are represented in the Greek—θεός twice and κύριος once—
only one, ‫ ֵאל‬, is in Hebrew. Furthermore, ὅσιος is not the most adequate
translation of ‫שׁר‬ ֖ ָ ָ‫ י‬either.14 So it is not at all certain that MT was the
13 By “dishonest” I state more bluntly and put in sharper relief what others have termed
“taking liberties,” “interpreting freely,” “harmonizing” and so on. All such modern judg-
ments presuppose the existence of MT as the LXX Vorlage. See also the comments of Orlin-
sky later in this paper.
14 The word ὅσιος usually renders ‫חסד‬, while ‫שׁר‬ ֖ ָ ָ‫ י‬is generally translated εὐθύς (εὐθής)
“right” “upright.”
orlage of the Septuagint in this verse. And even if it were, it would seem
V
counter intuitive for a translator to engage in such an irreverent practice
as explicitly using the divine name at a time and in a community in which
avoidance of calling the name had become normative.15 Translators were,
after all, Jews translating sacred scripture.
Rather, to state the obvious, it would make just as much (or better)
sense to presume that the Vorlage of the Septuagint indeed had ‫ אלהים‬at
the beginning of the verse and ‫ יהוה‬at the end—reflecting a stage in the
Hebrew textual tradition before the selective substitution of the metaphor
‫ צוּר‬had become common practice—as it would be to postulate, without
basis other than prejudice in favor of the received text, that a simple noun
like ‫ הצוּר‬presented a problem for the translator. The noun θεός is not
the usual rendering for ‫ צוּר‬as McCarthy claims16 but overwhelmingly it
is for ‫אלהים‬. Thus, on its face, θεός in the LXX should signal to an unbi-
ased observer the presence of ‫ אלהים‬in the Vorlage, just as κύριος should
signal ‫יהוה‬, as it does at the end of the verse. And Wevers, despite his
unquestioned erudition, exhaustive study of Septuagint materials and
monumental body of work, had no greater insight into the mind of an
ancient translator than any of us. His presumptions are as good as ours.
Why would representation of the divine as an inanimate object, or trans-
literating or translating the word ‫ צוּר‬into Greek be more objectionable
than writing the divine name itself anyway?
But how might one test this hypothesis that the LXX Vorlage had ‫אלהים‬
and not ‫ ?צוּר‬The answer: with careful attention to the treatment of this
expression in the remainder of this chapter and, in time, throughout the
whole Hebrew Bible. That then will be the focus of the remainder of this
paper. Prior treatments of the problem have been either “theological,”
that is to say, they were concerned with Hellenistic views of the divine,
or they have been “text critical,” that is, they were focused on matters of
approaches to translation. What I am proposing is that this be considered
15 Indeed, the observable pattern of movement in the Hebrew Bible is progressively

away from vocalizing the divine name, whether moving from the specific Yahweh to
the more generic Elohim, or from Elohim to any number of substitutes—The Holy One
Blessed be He, One Ancient of Days, etc.,—to ultimately avoiding calling the name alto-
gether in later periods. Analogously, the word choice in the LXX for instance of ἀλλόφυλοι
for Hebrew Philistine appears to be but one of many examples illustrating a tendency to
move away from specific forms to more generic and interpretative ones. Thus the move-
ment from ‫ צור‬to ‫אלהים‬, if the former were in the original, would appear to be out of
character.
16 McCarthy, Deuteronomy, 151*.
a simple “literary-critical” project akin to other well-known examples of

variant forms of the text of the Hebrew Bible.
One question that could guide an inquiry of this kind would be this:
Were the adjustments based on theological/ideological grounds or were
they part of a mechanical process of “find and replace?” If they were made
for theological reasons, one would expect to find ‫ צוּר‬limited exclusively
to the god of Israel, who had come to be viewed in this way. Furthermore,
one might expect to find it replacing YHWH in places, since both forms of
the divine name were regularly interchanged. If on the other hand there
was a global process to selectively “find and replace” ‫ אלהים‬with ‫צוּר‬,
one might expect to discover evidence of inappropriate replacement, for
instance, in relation to ‫ אלהים‬as a plural form or in bound structures in
which the referent is a foreign divinity. That is precisely what one discov-
ers in the rest of Deut 32.
The next occurrence of ‫ צוּר‬in this specialized sense is in v. 15, which
provides insights into the translator’s method and his Vorlage. The Greek
makes explicit reference to Jacob in parallel with ֨‫ יְ ֻשׁרוּן‬Jeshurun. The for-
mer is transliterated as Ιακωβ, but the latter, ֨‫יְ ֻשׁרוּן‬, is interpreted as ὁ
ἠγαπημένος, “the beloved one.” This indicates minimally that the transla-
tor was not averse to rendering or interpreting metaphors, and certainly
points to a Vorlage different from MT, which here has no reference what-
soever to Jacob.
McCarthy explains the omission in MT as due to haplography,17 thus
granting the possibility that the LXX reading might be prior. Wevers by
contrast flatly states: “LXX made two lines out of the first hemistich, with
Ιακωβ as subject of line one and ὁ ἠγαπημένος, ‘the beloved’ freely translat-
ing ֨‫ יְ ֻשׁרוּן‬on the basis of its root ‫ ישׁר‬for the second line.”18 Both explana-
tions, insightful and creative as they seem, are totally subjective. What is
certain is that the Greek mentions Jacob by name; the Hebrew does not.
The texts are different and need not be reconciled.
The next lines ‫“ וַ יִּ טּ ֹ֙שׁ ֱא ֣ל ַוֹה ָע ָ֔שׂהוּ‬He abandoned Eloah who made him”:
‫“ ;וַ יְ נַ ֵ ֖בּל ֥צוּר יְ ֻשׁ ָע ֽתוֹ‬he spurned a rock that saved him” give the first clear
hint that ‫ ֥צוּר‬was likely a late insertion into an otherwise smooth line.
Here, unlike v. 4, ‫ ֥צוּר‬is anarthrous so it could theoretically point to any
rock. Literally translated, the stich would be rendered “he treated with
contempt a rock that saved him.” Whether the reviser forgot the article
17 Ibid., 144*.
18 Wevers, Notes, 518.
or simply presumed that subsequent readers would understand and sup-

ply it in context cannot of course be determined. But it seems significant
that of the 36 instances, only the first, 32:4, carries the definite article. This
might be a clue that the revision indeed began in Deut 32 and proceeded
forward.
Additionally, ‫ צוּר‬seems to disrupt the metrical flow of the line. The
hemistich ‫ וַ יִּ טּ ֹ֙שׁ ֱא ֣ל ַוֹה‬stands in parallel with ‫וַ יְ נַ ֵ ֖בּל אלהים‬, which would
be faithfully reflected if one reconstructs the Vorlage of the Septuagint as
it stands. The LXX reading here is again θεός, the standard gloss for ‫אלהים‬.
The translator does not distinguish between ‫ הצוּר‬and ‫צוּר‬, potential evi-
dence that the difference was not in his text.
In addition to its lack of reference to Jacob, the Hebrew of MT in v. 15
indicates its fractured character in one final way—it fluctuates between
second and third person pronouns: “You grew fat . . . He forsook the God
who made him . . .” The more expansive text reflected in the LXX is con-
sistent, beginning with the explicit naming of Jacob and appropriately
utilizing third person singular pronouns throughout. Indeed, without the
Septuagint, a reader might not have been able to determine the meaning
of the enigmatic ֨‫ יְ ֻשׁרוּן‬or for that matter the much-discussed ‫ ֥צוּר‬.
Verse 18 begins as noted with an anarthrous ‫ ֥צוּר‬. More revealing for
our purposes however are the next two instances, vv. 30 and 31, where a
confusion of ‫אלהי‬, the construct/bound form of the divine name, with
‫אלהים‬, the absolute/free form, seems to have taken place in the received
tradition. Verse 30 further shows that the LXX translator was not follow-
ing the MT tradition. Here is the context: “How could one have routed a
thousand/or two put ten thousand to flight/unless their Rock (‫ צורם‬MT),
God (θεός LXX) had sold them/the Lord had given them up?”
We begin with the obvious. If the LXX translator had ‫ צוּרם‬before
him and recognized ‫ ֥צוּר‬as a cipher for ‫אלהים‬, it would have been quite
straightforward to add αὐτῶν to ὁ θεός.19 But clearly that was not the case.
One of three things therefore seems to have occurred in the received
(MT) tradition. Either the Hebrew original from which modifications to
the divine name were made had ‫אלהיהם‬, or the final ‫ ם‬of an original
‫—אלהים‬the text reflected in the LXX—was read by the reviser as a third
person plural suffix with ‫ אלהי‬being glossed as ‫ ֥צוּר‬. Even more intriguing
and likely is the third prospect: that the original Hebrew, from which both
MT and LXX derived, contained the archaic form of the third plural suffix,
19 See comments on Ps 78:5 in n. 21 below.

‫אלהימו‬, as in v. 37. The LXX Vorlage would then be a variant upon, or mis-
reading of, this archaic form. Another way of framing the situation in the
parent text of the LXX would be to suggest haplography of (or parablepsis
on) the two ‫ה‬s in an original ‫אלהיהם‬, or the apocopation of a final waw
in an original ‫אלהימו‬.
Whatever the case, ‫ צוּרם‬creates ambiguity here. Their “rock,” ‫צוּרם‬, if
accepted as a reference to divinity, could now refer either to the god of
Israel through whose power the one or two persons were able to put thou-
sands to flight, or to the god of the myriads being put to flight, who for
whatever reason became angry and sold them to the one or two Israelite
pursuers. No such ambiguity is reflected in the Vorlage of the LXX. There,
it is the god of Israel who makes the improbable routs possible because
he had sold the many enemies into the hands of the few Israelites. The
equivalent of ‫ צוּרם‬is ‫( אלהים‬θεός) in parallel with Yahweh (κύριος). What
can be inferred here, then, is that the substitution of ‫ צוּר‬for ‫ אלהים‬or
‫ אלהי‬was a mechanical process that did not take into account contextual
or even grammatical concerns in some cases.
This situation is not limited to Deuteronomy or the Pentateuch. Paral-
lels can be seen in the so-called Elohistic Psalter (Pss 42–83) where ‫אלהים‬,
the more generic form of the divine name, was regularly substituted for
‫יהוה‬, the specific ineffable name of Israel’s god. Deuterographs like Ps
14 and 53 illustrate this quite clearly.20 Several other instances of appar-
ently uncritical or unthinking insertion of ‫ אלהים‬in the Elohistic Psalter
have been preserved, in MT. For instance, the ungrammatical expression
‫ יהוה אלהים צבאת‬in Ps 59:6; 80:5, 8, 15, 20 as well as 84:9, is probably
best explained as a substitution of ‫ אלהים‬for ‫אלהי‬. By 89:9, the normal
‫ יהוה אלהי צבאת‬resurfaces. Again, the presence of ‫ אלהים‬in bound
expressions like Elohim Elohai, Elohim Eloheka, Elohim Elohenu, Elohim
Elohe Israel, Elohim Elohe Teshuahti and so on, in the Elohistic Psalter
(43:4, 44:8, 48:15, 50:7, 51:16, 67:7, 68:9), almost surely indicates a substitu-
tion for an original Yahweh in every instance.
Even outside of the Psalter we find the expression Bet Elohim, Elohe
Israel “The house of God, god of Israel” (Ezra 6:22) and Elohim Elohe Avote-
hem “God, God of their fathers” (2 Chron 34:32). In this latter passage the
LXX reflects (the earlier) “Yahweh God of their fathers.” This pattern of
20 Note especially that Yahweh in Ps 14:2, 4 and 7 is represented/replaced by Elohim in

the virtual deuterograph Ps 53. One could argue that these are alternate forms of the same
Psalm but the case for 53 as subsequent to 14 in light of the attitude to the nomina sacra
seems convincing given its location in the collection.
deliberate substitution of the implicit for the explicit is thus not unusual
by any means. Something analogous seems to have occurred to produce
the confusion around ‫ צוּרם‬in Deut 32:30. Another parallel might be the
-boshet/-baal substitutions in Hebrew theophoric names involving Baal. I
could not imagine anyone seriously arguing that Ishboshet was an earlier
version of the name Ishbaal.
The following verse, 31, makes it abundantly clear that the substitution
of ‫ ֥צוּר‬for ‫ אלהים‬in the received tradition was a later mechanical process,
which could not and did not affect the Hebrew behind the Septuagint. It
has led to an undesirable theological situation in the text as it now stands.
The noun ‫ ֥צוּר‬could not have been intended as a metaphor for the god
of Israel exclusively because it is used here and in v. 37 in reference to a
foreign divinity. This indicates that the substitution was primarily at the
lexical level and not ideologically or theologically driven.
As in v. 30, ‫ צוּרם‬seems to represent an original ‫אלהים‬, ‫ אלהיהם‬or
‫( אלהימו‬mis)read, interpreted or (mis)understood in a similar way as
it was there. But the reviser seemed insensitive to the fact that in this
context ‫אלהי‬/‫ אלהים‬was plural since it refers to the gods of the nations,
and so he dutifully plugged in his default substitution for ‫אלהים‬, i.e., ‫ ֥צוּר‬.
Both the unpointed Vorlage of the LXX and the Hebrew version prior to
the revision later received by the Masoretes appear to have read: ‫כי לא‬
‫אלהימו‬/‫כאלהינו אלהיהם‬. The LXX accurately reflects that; the received
tradition retains a corruption that is both ideologically and lexically more
clumsy.
The final verse that makes any reference to ‫ צוּר‬in this specialized way—
v. 37—confirms and brings closure to the notion that the LXX Vorlage was
different from the received text and was uninfluenced by it. Note again,
the anarthrous ‫צוּר‬. Both the LXX and the received text recognize and
transmit the first part of the verse, though the Masoretes point the verb
‫ אמר‬as a waw consecutive with the perfect without a subject, i.e., ‫וְ ָא ַ ֖מר‬
“he will say,” and the LXX reflects a waw consecutive with the imperfect
with the subject being Yahweh, ‫“ ויאמר יהוה‬The Lord said.” I see no need
to harmonize or prioritize these readings. What is more interesting is that
the LXX reflects no evidence of ‫ צוּר‬in its parent text. Rather, the ἐφ᾿ οἷς
seems to suggest the presence of a relative particle, perhaps ‫אשׁר‬, instead
of ‫ צוּר‬in the Vorlage to which the final ἐπ᾿ αὐτοῖς points back.
So, what might we conclude from the evidence in Deut 32 regarding
‫ ?צוּר‬First, it cannot be shown definitively that the Vorlage of the Septua-
gint read the equivalent of MT in any place in this chapter. Indeed the
evidence seems just as, or even more, compelling that all the references
to θεός indicate that it is the normal representation of ‫אלהים‬. Second,

the Vorlage of the LXX demonstrates in some instances a more context
sensitive understanding of the text. Rather, it is the reading of the MT that
displays randomness, ambiguity and inelegance.
How would these findings hold in other sections of the Hebrew Bible
where ‫ צוּר‬seems to have been used metaphorically in MT? I am well
aware that individual translators might not have consulted each other,
and that patterns in one book may not be repeated in other books. But
it could be instructive to observe how the expression is treated in other
parts of the Septuagint. Such a survey might answer questions about the
Vorlage of each book and the habits of different translators. The books of
Samuel (1 & 2 Kingdoms) contain a few instances that might cast light on
the subject at hand. To those we turn now briefly.
In 1 Sam 2:2, the Septuagint and MT are not aligned completely. Only
the first clause, ‫יהו֖ה‬ ָ ‫ין־ק ֥דוֹשׁ ַכּ‬
ָ ‫“ ֵא‬there is no one holy like the Lord” is
shared; the LXX Vorlage also reflects an initial ‫ ִ ֣כּי‬. The second clause in
Hebrew, ‫“ ִ ֣כּי ֵ ֣אין ִבּ ְל ֶ ֑תָּך‬Truly there is no one beside you” finds no exact
parallel in Greek. If there is one, it is the final Greek clause, οὐκ ἔστιν
ἅγιος πλὴν σοῦ “There is no one holy besides you,” but that is a sim-
ple restatement of the first clause. The final clause that includes ‫צוּר‬,
‫אֹלהינוּ׃‬
ֽ ֵ ‫“ וְ ֵ ֥אן ֖צוּר ֵכּ‬there is no rock like our God” is not represented. Greek
has the equivalent of “there is none righteous like our God” as its second
clause. Some might wish to argue, and have suggested, that ‫ צוּר‬is repre-
sented by δίκαιος but that would be a stretch. What this example shows is
that the Hebrew and Greek of the verse, though similar in theme, are not
identical, and further that ‫צוּר‬, if it is accounted for at all, is not glossed
by ‫אלהים‬.
Five passages in 2 Sam (22:3, 22:32, 22:47 twice, 23:3) indicate that the
LXX Vorlage did contain ‫ צוּר‬which, in all but one instance, is glossed
with Greek φύλαξ “keeper/guard” in keeping with the verbal sense of
‫צוּר‬, “to enclose/besiege.” The variant case is 22:32 where ‫ צוּר‬is repre-
sented by κτίστης “creator,” obviously being read as related to the root ‫יצר‬
“to form or create.” But nowhere in the books of Samuel is there evidence
for ‫ אלהים‬as a translation of ‫צוּר‬. This further supports our conclusions
about the Vorlage in Deut 32 being indeed ‫אלהים‬.
The next cases are within the book of Isaiah. The first example (17:10)
is important because it is the only instance in which ‫ צוּר‬in MT corre-
sponds with ‫ יהוה‬in the Vorlage of the LXX. This confirms expectations
mentioned earlier, namely, that ‫ צוּר‬might replace Yahweh on occasion if
substitutions were ideologically driven. It also unfortunately disconfirms
Wevers’ unequivocal statement that the LXX substitutes θεός for ‫צוּר‬
throughout. But more important it shows that the LXX remains faithful
to its parent text, which obviously had Yahweh here and was thus not a
translation of ‫ צוּר‬in this case.
The end of Isa 26:4 ‫עוֹל ִ ֽמים‬
ָ ‫הוה ֖צוּר‬ ֔ ָ ְ‫ ִ ֚כּי ְבּ ָי֣הּ י‬is irrelevant to our discus-
sion because the Greek reflects a different Vorlage and makes no repre-
sentation of ‫צוּר‬. The example in 30:29, ‫ אל צוּר ישׂראל‬takes us back to
the pattern in Deuteronomy where ‫ צוּר‬apparently was not in the Vor-
lage of the Greek but rather ‫אלהים‬. The Greek, πρὸς τὸν θεὸν τοῦ Ισραηλ
seems to represent ‫אל אלהי ישׂראל‬. This would also confirm the earlier
observation that ‫ צוּר‬occasionally replaced the bound form ‫ אלהי‬of the
divine name. In 44:8 it would appear that ‫ צוּר‬was not in the parent text of
the Greek. Its presence in the received text suggests the independence of
the textual traditions.
Hab 1:12, like Isa 17:10, is a key passage in relation to this discussion.
Clearly, ‫ צוּר‬was in the received text and also in the Septuagint Vorlage
preceded in both instances by a ‫ ו‬i.e., ‫וצוּר‬. The LXX translator read his
text as a verb coming from ‫יצר‬, “to form or fashion,” and it appears that
verb was suffixed with the first person singular. Only the LXX recognizes
the waw because for ‫ וצוּר‬it reads: καὶ ἔπλασέν με = “and he has fashioned
me.” Its rendering of the rest of that verse ‫הוֹכ ַיח יְ ַס ְדתּוֹ‬ ֥ ִ ‫“ ְל‬to examine his
chastening” seems an accurate account of what is in the received text.
Modern translators of MT routinely ignore the waw, make ‫ צוּר‬a voca-
tive, and read the third singular suffixes as third plurals, something like:
“O Rock, you have made ‘them’ a cause for punishment/ complaint.”
So it is quite clear when ‫ צוּר‬was in the LXX parent text and how transla-
tors treated it. It is also clear when it was not. We only need to respect and
accept the information before us and not question the motives, actions or
integrity of the messenger.
We have observed thus far instances where ‫צוּר‬: (a) is translated rou-
tinely as πέτρα, (b) it is translated as a nominal relating to the verbal form,
hence φύλαξ “stronghold, keeper” etc., and (c) it is related to the verb ‫יצר‬
“to form/fashion/create.” There is thus no indisputable evidence that ‫צוּר‬
was ever translated as θεός or κύριος in the Septuagint. When those words
appear, they should be taken in their usual sense, namely to indicate the
presence of ‫ אלהים‬or ‫ יהוה‬in the parent text.
The remaining occurrences of ‫ צוּר‬as metaphor in MT occur in the Psal-
ter, and to these we now turn finally. The situation in the LXX Psalter
comports perfectly with our observations thus far. Two distinct patterns
emerge. First, ‫ צוּר‬is unquestionably present in the parent text of the LXX
in one third of the total cases, and in each case it is translated as related
to help or support: i.e., βοηθός—Ps 18:3, 19:15, 49:15, 94:22, and in 89:27,
ἀντιλήμπτωρ “supporter.” In 78:35, ‫ צוּרם‬is clearly present in the Vorlage,
and rendered as βοηθὸς αὐτῶν “their helper.” This supports our judgment
that the collocation ‫ צורם‬was not in the Vorlage of Deut 32: 30 and 31, for
here we see what its translation would look like.21
On the other hand, ‫ אלהים‬not ‫ צוּר‬appears to have been in the par-
ent text in all remaining instances—18:32, 18:47, 28:1, 31:3, 62:3, 62:7, 62:8,
92:16; 95:1; 144:1—in the Psalter. In 71:3 and 73:26 the LXX Vorlage probably
read ‫ לאלהי‬and ‫ אלהי‬respectively as the translations εἰς θεόν and ὁ θεός,
suggest.
Those who would wish to argue otherwise bear the burden to show
credible reasons for the alternate translations of ‫ צוּר‬mentioned above.
Furthermore, they must also answer this basic question: “Why would a
translator follow his text so closely in the main and then in certain places
deviate dramatically from the text before him?” If his deviations are unre-
liable so also should be his agreements.
What then does all this mean? The most significant implication of these
findings is (something I have been emphasizing increasingly in my recent
work, namely,) that the well-established tradition of using the late text of
the Leningrad Codex (MT) as the starting point, the lemma, for textual
analysis is passé and needs to be rethought. Despite its wide acceptance
in religious circles and its rich history, it is just a manuscript like any other
and a rather late one at that. A corollary of this is that the suspicions
surrounding the Septuagint translators who worked centuries before the
Leningrad Codex was written, especially of their methods and motives,
must be lifted if we wish to understand the meaning of the evidence they
provide. Constant efforts to harmonize or explain away obvious differ-
ences in extant forms of the biblical text do violence to each of the forms
and do us no permanent good. The LXX is often the bearer of “inconve-
nient truths” but killing or doubting the messenger only serves to confirm
preexisting beliefs and does not assist in determining the true state of
affairs in the ancient world.
This is not just my view. Three other corroborating voices from dif-
ferent generations, among many other possible candidates, must suffice.
Already in 1974, Harry M. Orlinsky in his Grinfield lectures on the Septua-
gint at Oxford said the following:
21 Cf. n. 19 earlier.
I would put it more bluntly: the modern scholar has no right to assume a
priori that the Septuagint translator manipulated his Hebrew text; in point
of fact, it is the opposite assumption that would be valid viz., that unless
and until a case can be made for regarding the Septuagint translation of
any Book as an unreliable witness to the Hebrew text, the translator must
be regarded as a serious, forthright reproducer of that text.22
And later, speaking of variance between the received text and the Septua-
gint, in Deut 31 no less, Orlinsky says:
What is important is that scholars have not yet sufficiently recognized that
the translator did not edit or emend a text that contained our preserved
reading, for he was not an editor-harmonizer; he was a translator, and he
translated what his Hebrew text offered him. A pity indeed that so many
modern scholars have created anthropomorphic-theological-contextual-
textual problems for the Septuagint translators that they never had, and have
then proceeded to create “solutions” of the same ephemeral substance—
usually at the “expense” of the integrity and scholarship of the translators.23
Thirteen years later Anneli Aejmelaeus, in her frequently quoted 1987
ZAW article “What Can We Know about the Hebrew Vorlage of the Sep-
tuagint?” concludes:
In principle, the MT plus corresponding to the Septuagintal minus and the
Septuagintal plus ought to be treated on equal terms.
All in all, the scholar who wishes to attribute deliberate changes, harmo-
nizations, completion of details and new accents to the translator is under
obligation to prove his thesis with weighty arguments and also to show why
the divergences cannot have originated with the Vorlage. That the translator
may have manipulated his original does not mean that he necessarily did
so. All that is known of the translation techniques of the Septuagint points
firmly enough in the opposite direction.24
Finally, T. M. Law, a recent Oxford graduate, published an impressive and
well-argued defense of an independent pre-MT Vorlage for 3 Kingdoms,
concluding:
Septuagint scholars must not go backwards to a pre-Qumran view of the
biblical text and its associated explanations of the Greek textual history, but
must instead approach the state of the question in the light of what is known
now about textual plurality and in the light of the most current research on
22 H. M. Orlinsky, “The Septuagint as Holy Writ and the Philosophy of the Translators,”
HUCA 46 (1975): 110.
23 Ibid., 112.
24 A. Aejmelaeus, “What Can We know about the Hebrew Vorlage of the Septuagint?”
reprinted in On the Trail of the Septuagint Translators (Kampen: Kok Pharos, 1993), 92.
the OG . . . Hebrew Bible scholars, however, must go backwards; for it was

in the 19th century that Hebrew Bible commentators on Kings . . . evaluated
3–4 Reigns as a different Hebrew Vorlage to MT.25
Instances abound where retroversion of the Hebrew behind the Septu-
agint is difficult and challenging to the modern scholar. We do well to
confine our intellectual energies to resolving those puzzles rather than
creating problems where they do not exist and need not be. There is noth-
ing problematic about the meaning of the divine names in either Hebrew
or Greek and we are well aware of the abundant evidence throughout
the LXX corpus that the Vorlage of the Septuagint, though similar in the
main with the received text, was often quite radically at variance with it.
I suggest that in this matter we have yet another such example. I hope to
have demonstrated in this brief survey, of a well-established but in my
view totally invalid notion, the risks of uncritically trusting late printed
Hebrew over early, easily retrovertable translation Greek.
25 T. M. Law, “How Not to Use 3 Reigns: A Plea to Scholars of the Books of Kings,”
VT 61 (2011): 294.
Judges 3:12–30: An Analysis of the Greek Rendering of
Hebrew Wordplay
Hans Ausloos
1. Wordplay and the translation technique of the Septuagint
If one would like to understand the full meaning of a biblical text, it is

not sufficient to have a good knowledge of Hebrew grammar and syntax.
From the first chapters of the Old Testament on, it becomes clear that
wordplay—whatever this may be—seems to be constitutive for bibli-
cal literature.1 If the reader does not succeed in tracing present forms of
wordplay within a pericope, its full meaning undoubtedly gets lost.
The present contribution does not, by any means, aim to clarify the
problem of Hebrew wordplay on a theoretical level.2 In this paper, I would
like to study if and how the Septuagint (LXX) translator has dealt with the
different types of wordplay with which he was confronted in his Hebrew
Vorlage. Evidently, it is one of the most difficult problems for a translator
to adequately render wordplay from a source language into a target lan-
guage, as becomes evident from the first pages of the Old Testament.
One of the first examples of wordplay in the Hebrew Bible can be
found in the creation narrative in Gen 2–3, where a clear link is made
between ‫ ָה ָא ָדם‬who is formed ‫ ִמן ָה ֲא ָד ָמה‬. Making use of similar-sounding
words, the Hebrew author clearly indicates that, in his view, there is a
close relationship between human beings and the earth (this does not
automatically imply that there is also an etymological link between
1 Despite the importance of wordplay for the understanding of the Hebrew Bible, one
had to wait until the end of the 19th century for the first systematic studies on this topic:
I. M. Casanowicz, Paronomasia in the Old Testament (Ph.D. diss., John Hopkins Univer-
sity, 1894); see also idem, “Paronomasia in the Old Testament,” JBL 12 (1893): 105–167.
Also in more recent times, systematic approaches to the field remain rather rare. See in
this respect, T. Cherry, Paronomasia and Proper Names in the Old Testament. Rhetorical
Function and Literary Effect (Ph.D. diss., Southern Baptist Theological Seminary, 1988);
S. B. Noegel, ed., Puns and Pundits. Wordplay in the Hebrew Bible and Ancient Near Eastern
Literature (Bethesda: CDL Press, 2000).
2 For some remarks with respect to the use of wordplay in biblical literature and an
attempt of classification of different categories, see V. Kabergs and H. Ausloos, “Parono-
masia or Wordplay? A Babel-Like Confusion. Towards A Definition of Hebrew Wordplay,”
Bib 93 (2012): 1–20.
54 hans ausloos
the two lexemes). In cases where one can hardly doubt that the Hebrew
author intended to play with Hebrew words, as is the case in Gen 2, a good
translator should at least try to render his Vorlage as adequately as pos-
sible. Taking for granted that the translator actually did notice the Hebrew
wordplay, he has three possible options. First, he could add a footnote
in order to clarify the wordplay that is present in the source language.
Doing so, he would indicate that he has noticed the Hebrew wordplay,
but at the same time he would admit that he was not able to find an
adequate translation equivalent, and therefore failed in his translation.
Second, the translator could transliterate those Hebrew words that are
constitutive for the wordplay. As such, he equally clearly indicates that
he has noticed the Hebrew wordplay and makes clear to his readers that
there is a link between the two words. At the same time, however, he
does not succeed as a translator in this case either, since he is not giving
a real translation. Third, he could translate the Hebrew words and search
for good alternatives in the target language. However, in practice, it is an
almost impossible task to find two or more terms in the target language
that do not only correspond in meaning, but have a similar connotation
or sound as well. Taking a look at Gen 2:7, one can conclude that the LXX
translator in fact translated ‫ ָה ָא ָדם‬and ‫ ִמן ָה ֲא ָד ָמה‬as καὶ ἔπλασεν ὁ θεὸς
τὸν ἄνθρωπον χοῦν ἀπὸ τῆς γῆς (NETS: “And God formed man, dust from
the earth”).3 In doing so, however, the link between both terms has been
completely lost for the reader of the LXX. Nevertheless, one should be
wary of drawing generalizing conclusions too easily. For example, within
the same context of the Eden narrative (Gen 3:20), the LXX translator suc-
ceeds very well in rendering the Hebrew wordplay into idiomatic Greek.
“Adam” calls his wife Eve (‫ ) ַחוָּ ה‬because she is “mother of all living ones”
(‫) ִכּי ִהיא ָהיְ ָתה ֵאם ׇכּל ָחי‬. The LXX translates it as καὶ ἐκάλεσεν Αδὰμ τὸ
ὄνομα τῆς γυναικὸς αὐτου Ζωη ὅτι αὕτη μήτηρ πάντων τῶν ζώντων (NETS:
“And Adam called the name of his wife Life, because she is the mother
of all the living”).4 In this case, the LXX translator did effectively translate
the proper name ‫ ַחוָּ ה‬with the Greek Ζωή and thus, he has chosen not to
transliterate it. Since ζωή in Greek means “life”, the implicit meaning of
the proper name ‫ ַחוָּ ה‬is preserved in Greek. Moreover, the sound similar-
ity with τῶν ζώντων is present as well.
3 R. J. V. Hiebert, “Genesis,” in A New English Translation of the Septuagint and the
Other Greek Translations Traditionally Included under That Title (eds. A. Pietersma and
B. G. Wright; New York / Oxford: Oxford University Press, 2007), 7.
4 Ibid., 8.
judges 3:12–30 55
Keeping these three major possibilities of rendering Hebrew wordplay

in mind, I am convinced that the study of the specific way in which Greek
translators render the Hebrew wordplay can be helpful in the attempt
to characterise the so-called translation technique of the LXX translators
as adequately as possible. The Greek rendering of Hebrew wordplay—
aetiologies can also be considered as such5—can therefore function as
a so-called content-related criterion6 in addition to the analysis of the
rendering of so-called Hebrew hapax legomena7 and Hebrew jargon-
defined vocabulary.8 These content-related criteria have to be considered
as complementing the “grammatically qualitative” research—as found in
the analyses of the Helsinki school, for example—from the specific angle
of analysing the procedure of the LXX translator when he was confronted
with difficult, highly specific and identifiable questions related to the con-
tent of his Vorlage.
5 H. Ausloos, “LXX’s Rendering of Hebrew Proper Names and the Characterization of
the Translation Technique of the Book of Judges,” in Essays on Septuagint, Hebrew Bible,
and Dead Sea Scrolls (FS R. Sollamo; JSJSup 126; eds. A. Voitila and J. Jokiranta; Leiden:
Brill, 2008), 53–71; H. Ausloos, “The Septuagint’s Rendering of Hebrew Toponyms as an
Indication for the Translation Technique of the Book of Numbers,” in Textual Criticism and
Dead Sea Scrolls Studies. Florilegium Complutense. (FS J. Trebolle Barrera; JSJSup 158; eds.
A. Piquer Otero and P. A. Torijano Morales; Leiden / Boston: Brill; 2012), 35–50.
6 A first methodological presentation of content-related criteria, named as such, has
been systematically presented and illustrated at the LXX.D-conference in Wuppertal, 2008.
See in this respect, H. Ausloos and B. Lemmelijn, “Content-Related Criteria in Characterising
the LXX Translation Technique,” in Die Septuaginta: Texte, Theologien und Einflüsse (WUNT
252; eds. W. Kraus, M. Karrer and M. Meiser; Tübingen: Mohr Siebeck, 2010), 357–376.
A refinement of the initial approach, and more particularly with respect to the Greek
rendering of Hebrew wordplay as a content-related criterion, has been presented at the
LXX.D-conference in Wuppertal, 2010 and has been published as H. Ausloos, B. Lemmelijn
and V. Kabergs, “The Study of Aetiological Wordplay as a Content-Related Criterion in the
Characterisation of LXX Translation Technique,” in Die Septuaginta: Entstehung, Sprache,
Geschichte (WUNT 286; eds. S. Kreuzer, M. Meiser and M. Sigismund; Tübingen: Mohr
Siebeck, 2012), 273–294.
7 H. Ausloos and B. Lemmelijn, “Rendering Love. Hapax Legomena and the Characteri-
sation of the Translation Technique of Song of Songs,” in Translating a Translation. The
LXX and its Modern Translations in the Context of Early Judaism (BETL 213; eds. H. Ausloos
et al.; Louvain / Paris / Dudley: Peeters, 2008), 43–61; H. Ausloos, “The Septuagint’s Render-
ing of Hebrew Hapax Legomena and the Characterization of its ‘Translation Technique’.
The Case of Exodus,” Acta Patristica et Byzantina 20 (2009): 360‑376. See also E. Verbeke,
“The Use of Hebrew Hapax Legomena in Septuagint Studies. Preliminary Remarks on
Methodology,” in Florilegium Lovaniense. Studies in Septuagint and Textual Criticism (FS
F. García Martínez; BETL 224; eds. H. Ausloos, B. Lemmelijn and M. Vervenne; Louvain /
Paris / Dudley: Peeters, 2008), 507–521.
8 B. Lemmelijn, “Flora in Cantico Canticorum. Towards a More Precise Characterisation
of Translation Technique in the LXX of Song of Songs,” in Scripture in Transition. Essays on
Septuagint, Hebrew Bible and Dead Sea Scrolls (FS R. Sollamo; JSJSup 126; eds. A. Voitila and
J. Jokiranta; Leiden, Brill, 2008), 27–51.
56 hans ausloos
The analysis of the Greek rendering of Hebrew wordplay therefore

functions in the context of a kind of artificial laboratory situation that
we create in order to observe the translator’s activity in a particular situ-
ation in which he has been forced to make a specific decision. Of course,
as is the case with hapax legomena (we cannot be entirely certain that
what we consider a hapax was equally a hapax for the translator), in some
cases there can be discussion on whether or not the Hebrew author has
intentionally played with words. As a result, it is theoretically possible
that what we consider as wordplay has not been perceived as such or even
remarked by the translator.9 Nevertheless, after receiving the results of
such a concrete, content-related analysis, the characterisation of a specific
LXX book can be evaluated more precisely. First, by comparing the differ-
ent LXX books ad intra on the basis of one and the same content-related
criterion (e.g. wordplay). Second, ad extra in a double way: comparing
the results on the basis of a particular content-related criterion to the
results of the study of another content-related criterion in different books
(e.g. wordplay compared to hapax legomena). Finally, by comparing the
content-related characterisation to the more traditional quantitative and
grammatically qualitative characterisation of the translation technique of
the respective LXX books.
Against this concise theoretical background, this paper will present a
concrete, elaborated example of an analysis of this kind in studying the
Greek rendering of Hebrew wordplay in Judg 3:12‑30. It will thereby demon-
strate the extent to which the combination and/or comparison of several
distinctive content-related criteria—in particular the analysis of the way
in which Hebrew hapax legomena are rendered and the way in which the
Greek rendering reflects the Hebrew wordplay—can give relevant results
in relation to a more sound characterisation of the LXX translation.
2. The Ehud and Eglon narrative in Judg 3:12–30
The stereotypical narrative pattern of the book of Judges is sufficiently

known. In reaction to Israel’s apostasy, God makes use of Israel’s enemies
9 Cf., however, the valuable remark of J. J. M. Roberts, “Double Entendre in First Isaiah,”
CBQ 54 (1992): 39‑48, esp. 40, with regard to the phenomenon of double entendre: “While
one must remain aware of the danger of overreading, however, it is far more likely that
our lack of familiarity with the wider connotations of classical Hebrew words and phrases
will result in underreading, of missing intentional double entendres”.
judges 3:12–30 57
in order to convert the Israelites, and, through the work of a “judge”,

brings them back on the right path. In the case of Judg 3:12‑30, it is the
corpulent Eglon, the Moabite king, to whose power God delivers the mis-
behaving Israelites. Together with the Ammonites and Amalekites, Eglon
fights Israel and conquers the “city of Palms”. After eighteen years of siege,
the Israelites beg God for help, after which the latter sends them a sav-
iour, Ehud the “Benjaminite”. Ehud and Eglon meet each other for the first
time at the moment that Ehud offers Eglon tribute from the Israelites. In
preparation of his visit, Ehud makes a “double-edged” sword one gomed
long, which he fixes at his right hip under his clothes. When tribute is
paid, Ehud gives permission to the carriers of the tribute to return home.
However, upon reaching the “sculptured stones” of Gilgal, Ehud returns to
Eglon’s palace and tells him that he still has a secret. Eglon immediately
orders all present persons to keep silent, which seems to be understood
as an order to leave the room.10 Ehud enters the throne room and, being
alone with Eglon, tells him that he has a divine message. In a gesture of
respect for the word of God which Ehud will speak out, Eglon rises from
his throne. With his left hand, Ehud takes the sword from his right thigh
and thrusts it in Eglon’s belly, so deep that even the hilt goes in after the
blade. Because Ehud does not withdraw the sword, the fat of Eglon’s belly
closes over the blade.
After the murder, v. 23 describes the way in which Ehud leaves the locum
delicti. He locks the doors of the roof chamber. Thereafter, i.e. after Ehud
has left, the servants come back and observe that the doors of the “roof
chamber” are locked. They speak cryptic words to each other: “Surely he
covers his feet in the cool chamber” (NRSV: “He is only relieving himself in
the closet of the cool chamber”). They wait a while and finally unlock the
doors to find their lord lying dead on the floor. In the meantime, Ehud has
been able to escape “beyond the sculptured stones” to Seira (as opposed
to v. 19, Gilgal is not mentioned in v. 26). When he reaches Ephraim, he
sounds the trumpet, gathers the Israelites and marches with them against
10 Judg 3:19 LXX (A) renders the Hebrew construction ‫אמר ָהס‬ ֶ ֺ ‫ וַ יּ‬as καὶ εἶπεν Εγλωμ
πᾶσιν ἐκ μέσου (NETS: “And Eglon said to everyone, ‘Away!’ ”). With respect to the double
text of Judges, see inter alia, W. R. Bodine, The Greek Text of Judges. Recensional Develop-
ments (HSM 23; Chico: Scholars Press, 1980); O. Munnich, “Le texte de la Septante,” in La
Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien (Initiations
au christianisme ancien; eds. M. Harl, G. Dorival and O. Munnich; Paris: Cerf, 1994), 175;
N. Fernández Marcos, The Septuagint in Context. Introduction to the Greek Versions of the
Bible (Boston / Leiden: Brill, 2001), 94–95.
58 hans ausloos
Moab. The pericope concludes with the mentioning of the killing of ten
thousand strong and fat Moabites, after which rest returns to Israel.
Judg 3:12‑30 represents a real pearl of ancient Hebrew narrative art that
has been passed down to us. Nevertheless, it is a difficult text, not least
because of several hapax legomena in it. In fact, there are three: the length
of Ehud’s sword as one “gomed” (‫ )גּ ֶֺמד‬long (v. 16), the noun ‫ַה ַפּ ְר ְשׁד ֺנָ ה‬
(v. 22) and the word ‫( ַה ִמּ ְס ְדּרוֹנָ ה‬v. 23). Having dealt with these three
hapax legomena in another contribution,11 the present paper will concen-
trate on the wordplay in this passage, where the usage of Hebrew lan-
guage is very much well-thought-out and often ambiguous. The author
constantly plays with words and their meanings, and it seems as if no
single element is accidental, although it is not always easy to define the
exact implications thereof.
3. Ehud as left-hander
The presentation of the Israelite protagonist of the narrative, Ehud, con-

fronts the reader of the Hebrew text with the first play on words. He
is introduced in Judg 3:15. When the Israelites cried unto God, YHWH
“raised up for them a deliverer, Ehud son of Gera, the Benjaminite, a left-
handed man” (NRSV) (‫מֹוׁש ַיע ֶאת ֵאהּוד ֶּבן ּגֵ ָרא ֶּבן ַהיְ ִמינִ י‬
ִ ‫וַ ּיָ ֶקם יְ הוָ ה ָל ֶהם‬
‫) ִאיׁש ִא ֵּטר יַ ד יְ ִמינֹו‬. Ehud is presented as son of Gera (‫) ֶּבן ּגֵ ָרא‬, who,
in turn, is said to be a “son of the Jemini” (‫) ֶּבן ַהיְ ִמינִ י‬.12 There can be
no doubt that the author is playing with words in his double use of the
forms ‫יְ ִמינֹו‬/‫ ַהיְ ִמינִ י‬. The text is not unambiguous. It does not become
clear whether the author of the Hebrew text links Ehud with the tribe
of Benjamin, or whether he would like to characterise Ehud as “son of
my (God’s?) right side”. Usually, when making reference to the person or
the tribe of Benjamin, the biblical authors make use of the proper name
‫ ִבנְ יָ ִמין‬.13 However, the fact that the author links Ehud with Gera seems
11 H. Ausloos and B. Lemmelijn, “Characterizing the LXX Translation of Judges on the
Basis of Content-Related Criteria. The Greek Rendering of Hebrew Absolute Hapax Legom-
ena in Judg 3,12–30,” in After Qumran: Old and Modern Editions of the Biblical Texts—The
Historical Books (BETL 246; eds. H. Ausloos, B. Lemmelijn and J. Trebolle Barrera; Louvain /
Paris / Dudley: Peeters; 2012), 171–192.
12 The combination ‫ ֶּבן ַהיְ ִמינִ י‬also occurs in 2 Sam 16:11; 19:17; 1 Kgs 2:8.
13 Cf., e.g., Gen 35:18; see also Judg 5:14; 10:9; 19:14; 20 (passim); 21 (passim). When Ben-
jamin is born after hard labour, his mother Rachel gives him the name Ben-Oni, which
means “son of my sorrow”. However, his father Jacob names his youngest son Ben-Jamin,
“son of (the) right hand”, which probably can be interpreted as “son of happiness”—cf.
judges 3:12–30 59
to be an argument in favour of Ehud’s identification as a Benjaminite,

because in Gen 46:21, Gera is considered to be a son of Benjamin. In
any case, the lexeme ‫ יָ ִמין‬plays a crucial role in the narrative. The same
v. 15, which presents Ehud as ‫ ֶּבן ַהיְ ִמינִ י‬, mentions that his right hand
(‫ )יְ ִמינֹו יַ ד‬is ‫ ִא ֵּטר‬. The precise meaning of this expression (‫) ִא ֵּטר יַ ד יְ ִמינֹו‬
is not clear. According to some exegetes, the author of Judg 3:15 aimed to
indicate that Ehud was handicapped at his right hand, which would imply
that he was left-handed.14 This is further used as the explanation for the
fact that Ehud fixes the sword at his right hip (v. 16). However, Judg 20:16,
the only other verse were the word ‫ ִא ֵּטר‬occurs, can be used against this
interpretation of a handicapped Ehud. In this verse, the same terminology
is used in characterising seven hundred soldiers of the Benjaminites—
here the term ‫ ְבנֵ י ִבנְ יָ ִמין‬is used ( Judg 20:15)—who are “picked men who
were left-handed; every one could sling a stone at a hair, and not miss”
(NRSV). In this respect, it is not plausible that all seven hundred warriors
are left-handed because of a handicap at their right hand. This leads schol-
ars to the supposition that, according to the biblical tradition, the Benja-
minites would have been specifically trained to fight with their left hands.15
So the adjective ‫ ִא ֵּטר‬has been related to a practice in which the right
hand was systematically bound in order to train soldiers in using their
left hand. In combat, in which attack and defence as well as the construc-
tion of armoury was mostly based on the use of the right hand, it could
be an advantage to be left-handed. It could have an effect of surprise to
the adversary, who was used to fighting against right-handed opponents.
In any case, when the author had the intention of saying that Ehud was
left-handed, it is remarkable that he uses the terminology ‫ִא ֵּטר יַ ד יְ ִמינֹו‬
and not ‫“( יַ ד ְשׂמֺאלוֹ‬his left hand”), as he does in Judg 3:21. Therefore, it
can be concluded that the term “left” is a keyword within the Hebrew
Ehud pericope. It is used again in v. 16 (Ehud fixes the dagger upon his
right thigh—‫ ) ַעל יֶ ֶרְך יְ ִמינֹו‬and finally in v. 21, where it is told that Ehud
C. Westermann, Genesis 12–36 (BKAT 1/2/2; Neukirchen-Vluyn: Neukirchener Verlag, 1981),

675. See also Y. Amit, The Book of Judges. The Art of Editing (Biblical Interpretation Series
38; Leiden / Boston / Köln: Brill, 1999), 171–198, esp. 179–180; U. Hübner, “Mord auf dem
Abort? Überlegungen zu Humor, Gewaltdarstellung und Realienkunde in Ri 3:12–30,”
BN 40 (1987): 130–140, esp. 133 n. 14.
14 Cf. NRSV: “Ehud son of Gera, the Benjaminite, a left-handed man”. Cf. F. Dexinger,
“Ein Plädoyer für die Linkshändler im Richterbuch,” ZAW 89 (1977): 268–269; E. Jugel and
H.‑D. Neef, “Ehud als Linkshänder. Exegetische und medizinische Anmerkungen zu Ri
3,15,” BN 97 (1999): 45–54, esp. 46–47.
15 See, e.g., B. Halpern, “The Assassination of Eglon—The First Locked-Room Murder
Mystery,” BRev 4/6 (1988): 32–41, esp. 35.
60 hans ausloos
takes his dagger with his left hand from his right thigh (‫) ֵמ ַעל יֶ ֶרְך יְ ִמינֹו‬
and thrusts it into Eglon’s belly.
Against this background, the question is raised as to which way the
LXX translator deals with this Leitmotiv of his Hebrew Vorlage. On the one
hand, his translation is rather disappointing. On the other hand, it will
become clear that he is a rather creative translator after all.
Let’s start with the disappointing element. The genealogical notion
‫( ֶאת ֵאהּוד ֶּבן ּגֵ ָרא ֶּבן ַהיְ ִמינִ י‬v. 15) is partially translated, partially very
literally transliterated: τὸν Αωδ υἱὸν Γηρα υἱοῦ τοῦ Ιεμενι ( Judg A and
B—NETS:16 “Aod son of Gera son of Iemeni”). The translator clearly dis-
tinguishes between the usual name of the tribe and the formula in Judg
3:15. In all instances in Judges where the usual Hebrew term ‫ ִבנְ יָ ִמין‬is used
to designate the tribe of Benjamin, the LXX reads Βενιαμιν. In Judg 3:15,
however, ‫ ֶּבן ַהיְ ִמינִ י‬is understood as a (singular) proper personal name.
The other instances in Judg 3, where it is said that Ehud uses his right or
left hand, are translated adequately. According to Judg 3:16, Ehud fastens
the dagger on his right thigh (ἐπὶ τὸν μηρὸν τὸν δεξιὸν αὐτοῦ), which is a lit-
eral translation of ‫ ַעל יֶ ֶרְך יְ ִמינֹו‬. When in v. 21 the murder is described, the
LXX likewise translates literally and correctly: with his left hand (A and
B: τὴν χεῖρα τὴν ἀριστερὰν αὐτοῦ), he takes the dagger from his right thigh
(A: ἀπὸ τοῦ μηροῦ τοῦ δεξιοῦ αὐτοῦ; B: ἐπάνωθεν τοῦ μηροῦ αὐτοῦ τοῦ
δεξιοῦ).
In sum, although the LXX renders the Hebrew Vorlage correctly, the
way in which the Hebrew author plays with the word pair “left” and “right”
in vv. 15, 16, 21, especially with regard to Ehud as son of Jemini, has been
completely lost for the Greek reader.
There is one final element, however, which has to be taken into con-
sideration too, namely Ehud’s characterisation as ‫ ִא ֵּטר יַ ד יְ ִמינֹו‬. Here, we
meet a translator who is looking for a good translation equivalent that
makes sense in the context of the pericope. He renders this enigmatic for-
mula with the term ἀμφοτεροδέξιον (NETS: “an ambidextrous man”). This
lexeme is composed of two common Greek words: the adjective ἀμφοτερος
(“both of two”) and the noun δέξιά (“right hand”) or the adjective δέξιος
(“on the right hand/side”). As a result, ‫ יַ ד יָ ִמין‬has an equivalent in the ele-
ment δέξιον.17 The term ἀμφοτεροδέξιον does not seem to be attested prior
16 P. E. Satterthwaite, “Judges,” in A New English Translation of the Septuagint and the
B. G. Wright; New York / Oxford: Oxford University Press, 2007), 204.
17 Compare, e.g., with Gen 48:17, where ‫ יַ ד יְ ִמינֹו‬is translated as τὴν δεξιὰν αὐτοῦ.
judges 3:12–30 61
to the LXX. Because of that, the lexicon LEH considers it a neologism,

meaning “ambidextrous”.18 Much later, the term can be found in the works
of Galenus Medicus (2nd century C.E.) and Aeristaenetus (5–6th century
C.E.).19 In the Greek papyri, the term does not seem to be present.20 In
any case, probably not knowing the precise meaning of the term ‫ ִא ֵּטר‬, the
translator seems to have been looking for an adequate translation equiva-
lent. Keeping very close to the Hebrew original (the element “right hand”
is conserved), he presents Ehud as able to fight with his left hand as well
as with his right hand. It seems that one can observe here a trace of the
translator’s contextual exegesis, taking care of the context of the narrative,
in which Ehud will use his left hand to kill the fat Eglon. It is with this
person that we will deal in the next part of this paper.
4. Eglon the calf
Besides the presentation of Ehud as left-hander, the Hebrew text seems

to give a specific appreciation of his adversary as well, who is called Eglon
(‫) ֶעגְ לוֹן‬. In the light of what happens in the story, this name could hardly
have been chosen accidentally. Whether the two last consonants (‫ון‬-)
function as a diminutive or not,21 the proper name ‫ ֶעגְ לוֹן‬is undoubtedly
related to the noun ‫ ֵעגֶ ל‬, which means “male calf”, “young bull”.22 The
proper name ‫ ֶעגְ לוֹן‬therefore seems to be an example of a double entendre.23
In Judg 3:12, 14, 17, 19, 20 and 24, ‫ ֶעגְ לוֹן‬is the name of the Moabite king.
However, due to its similar sound with the noun ‫ ֵעגֶ ל‬, the author seems
to give an appreciation of the king’s person.24 Even if the reader does not
18 J. Lust, E. Eynikel and K. Hauspie, Greek-English Lexicon of the Septuagint: Revised
Edition (Stuttgart: Deutsche Bibelgesellschaft, 2003), 34. T. Muraoka, A Greek-English Lexi-
con of the Septuagint (Louvain / Paris / Walpole: Peeters, 2009), 33 translates the term as
“able to use left and right hands equally well”. H. N. Rösel, “Zur Ehud-Erzählung,” ZAW 89
(1977): 270–272 interprets the Hebrew formula equally as indicating “Beidhändigkeit”.
19 H. G. Liddell and R. Scott, A Greek-English Lexicon. With a Revised Supplement
(Oxford: Clarendon Press, 1996), 95.
20 Cf. D. Hagedorn, Wörterlisten aus den Registern von Publikationen griechischer und
lateinischer dokumentarischer Papyri und Ostraka (s.l.: s.n., 132010; [cited 26 May 2011].
Online: http://www.zaw.uni-heidelberg.de/hps/pap/WL/WL.pdf ).
21 Cf. on this topic G. Mobley, The Empty Men. The Heroic Tradition of Ancient Israel
(ABRL; New York: Doubleday, 2005), 77–78; H.‑D. Neef, “Eglon als ‘Kälbermann’? Exeget-
ische Beobachtungen zu Jdc 3:12‑30,” VT 59 (2009): 284–294, esp. 288–289.
22 Cf. D. J. A. Clines, “‫ ֵעגֶ ל‬,” DCH 6, 248–249.
23 With respect to this terminus technicus, see Cherry, Paronomasia, 32–33.
24 Contrary to J. M. Sasson, “Ethically Cultured Interpretations. The Case of Eglon’s
Murder ( Judges 3),” in Homeland and Exile. Biblical and Ancient Near Eastern Studies
62 hans ausloos
immediately understand the wordplay—which seems rather implausible

because of the continuous hostility between Israel and Moab—it becomes
clear in the following verses. In a parenthesis in v. 17, the Moabite king
is presented as ‫ ִאיׁש ָּב ִריא ְמאֹד‬. The lexeme ‫ ָּב ִריא‬is often interpreted
as referring to Eglon as a very fat or corpulent man, not least because of
v. 22, in which it is said that Eglon’s fat closed upon the blade of the dag-
ger. However, human corpulence is normally referred to by making use of
the root ‫ ָשׁ ֵמן‬, as is the case in Judg 3:29.25 The term ‫ ָּב ִריא‬on the contrary,
is mostly applied to animals, meat or plants in order to indicate that they
are good or ripe for consumption.26 Using this term to characterise Eglon
seems therefore to have been done in order to present him as an object
that is ready to be slaughtered.27 In this respect, it is remarkable that the
author of the Ehud-pericope has used some more terms with a double
connotation.
The term ‫ ִמנְ ָחה‬, that occurs four times in Judg 3:12‑30 (vv. 15, 17, 18[bis])
does not only mean “tribute” or “gift”,28 but often also has a cultic and
sacrificial meaning, as becomes clear from several instances in the book
of Leviticus and Numbers,29 where it is sometimes linked with the notion
of “fat”.30 Besides the noun ‫ ִמנְ ָחה‬, the verbal form ‫ וַ יַּ ְק ֵרב‬in Judg 3:17
(‫ ָק ַרב‬: “to present”) also seems to make a link with the offer cult (cf.
‫ וַ יַּ ְק ֵרב ֶאת ַה ִמּנְ ָחה‬in Lev 9:17: “he presented the offering”). The irony of
the Hebrew text is thus clear: while Ehud is bringing a tribute to Eglon
( Judg 3:15), it is in fact king Eglon who will be slaughtered as an offering
(to God?) by Ehud.31
It is beyond doubt that it is very difficult for a translator to transpose
these elements of double entendre or implicit polysemy, which, moreover,
(FS B. Oded; VTSup 130; eds. G. Galil, M. Geller and A. Millard; Leiden / Boston: Brill, 2009),
571–595, esp. 573.
25 Cf. also J. M. Sasson, “Ethically Cultured Interpretations,” 575–576 n. 13.
26 Cf. Gen 41:2, 4, 18, 20 (cows); 41:5, 7 (ears of grain); 1 Kgs 5:3 (oxen); Ezek 34:20
(sheep); Hab 1:16 (food), Zech 11:16 (flesh); see also as object in Ezek 34:3 (the fat animals).
In Ps 73:4, the term refers to the body of the wicked; in Dan 1:15 it refers to the flesh of
Daniel, Hananiah, Mishael, and Azariah.
27 Contrary to J. M. Sasson, “Ethically Cultured Interpretations,” 575, who considers the
term as indicating that Eglon was “imposing (. . .)—a notice that explains why he would
lack guards, a crucial element in the unfolding plot”.
28 See, e.g., Gen 32:14, 19, 21, 22; 33:10; 43:11, 15, 25, 26.
29 Cf., e.g., Lev 2:1, 4, 5, 6, 15; 6:13; Num 15:6.
30 See, e.g., Lev 2:5–6, 15; 14:10.
31 Cf. A. Wénin, “Le ‘point de vue raconté’, une catégorie utile pour étudier les récits
bibliques? L’exemple du meurtre d’Églôn par Éhud ( Jdc 3,15–26a),” ZAW 120 (2008): 14–27,
esp. 17–18.
judges 3:12–30 63
have a clear function in the narrative, to the target language. In what fol-
lows, we will have a look at the manner in which the LXX translator has
dealt with these elements of Hebrew wordplay.
First of all, the translator opts to transliterate Eglon’s name as Εγλωμ.
Doing so, the connotation of the proper name with an animal that will
be slaughtered completely disappears. However, he could have given the
king another name, such as Μόσχον, the noun which in Exod 32:4 is used
as the translation equivalent for the noun ‫ ֵעגֶ ל‬. Secondly, Eglon’s charac-
terisation as ‫ ָּב ִריא‬has been rendered in a very divergent way by the Greek
translator of Judges. In the LXX (A and B), this term is represented by the
Greek adjective ἀστεῖος as a translation equivalent: Εγλωμ ἀνὴρ ἀστεῖος
σφόδρα (NETS: “Eglom was a very handsome man”). This word is rather
rare in the LXX. As an equivalent of a Hebrew word, it occurs in Exod 2:2
(Moses’ mother saw that her baby was ‫—טוֹב‬ἰδόντες δὲ αὐτὸ ἀστεῖον) and
in Num 22:32 (the angel refutes Balaam with the words: οὐκ ἀστεία (‫ )יָ ַרט‬ἡ
ὁδός σου ἐναντίον μου; NETS: “your way was not pretty before me”).32 Only
in Judg 3:17, is ἀστεῖος the translation equivalent of the Hebrew ‫ ָּב ִריא‬.33
LEH translates ἀστεῖος in Judg 3:17 as “handsome”,34 a translation which
is followed by NETS. Muraoka renders it here as “charming”.35 In La Bible
d’Alexandrie, the Greek term is translated as “plantureux” (which in Eng-
lish can be translated as “ample”). The French translation therefore keeps
the middle ground between the meaning in classical Greek (“nice, ele-
gant”) and the meaning which is requested by the context.36 LXX.D argues
that ἀστεῖος was probably meant as a euphemistic or even ironic designa-
tion of Eglon, which would imply that the translator was well aware of the
tenure of Judg 3:12‑30.37 LXX.D translates: “Eglom war ein recht ansehnli-
cher Mann”.
Thirdly, it has been argued that in Judg 3 the term ‫ ִמנְ ָחה‬is used in a
double sense: as a tribute from Ehud to Eglon, but equally as the offering
32 See also Jdt 11:23; Sus (OG) 1:7; 2 Macc 6:23: ὁ δὲ λογισμὸν ἀστεῖον ἀναλαβὼν (NETS:
“making a high resolve”).
33 Elsewhere in the Old Testament, ‫ ָּב ִריא‬has ἐκλεκτός (Gen 41:2, 4, 5, 7, 18, 20; 1 Kgs 5:3;
Hab 1:16; Zech 11:16), ἰσχυρός (Ezek 34:20) or παχύς (Ezek 34:3) as translation equivalent.
34 Lust, Eynikel and Hauspie, Greek-English Lexicon, 90.
35 Muraoka, A Greek-English Lexicon, 98. In Num 22:32, Muraoka translates ἀστεῖος as
“proper, appropriate”.
36 P. Harlé, Les Juges (BdA 7; Paris: Cerf, 1999), 97: “Notre traduction tient la balance
entre le sens du grec classique, ‘de belle apparence, élégant, joli’, et le sens requis par le
contexte”.
37 W. Kraus and M. Karrer, eds., Septuaginta Deutsch: Das griechische Alte Testament in
deutscher Übersetzung (Stuttgart: Deutsche Bibelgesellschaft, 2009), 250.
64 hans ausloos
of Eglon by Ehud. Leaving aside the transliterations,38 ‫ ִמנְ ָחה‬has in the

LXX the nouns δῶρον, θυσία, θυσίασμα, ξένιον, ὁλοκαύτωμα and προσφορά
as translation equivalents. While some of these terms—such as θυσία,
θυσίασμα and ὁλοκαύτωμα—have an almost exclusively cultic or sacrifi-
cial usage, the translator of Judg 3 consistently translates ‫ ִמנְ ָחה‬as δῶρον.
Parallel to its Hebrew equivalent, this Greek word also has a double usage
in the LXX. As translation equivalent of the noun ‫ ָק ְר ָבּן‬, the term is used
in Leviticus and Numbers as indicating cultic and sacrificial offerings. In
Lev 1:2, for example, the expression ‫ ָק ְר ָבּן ַליהוָ ה‬is rendered as δῶρα τῷ
κυρίῳ. However, the term δῶρον can also be used in the more neutral way
of “gift”, as, for example, in Gen 32:14, 19, 21, 22 ( Jacob sends presents to
his brother Esau). Therefore, although it is difficult to determine whether
the translation equivalent in the LXX of Judg 3 with its double entendre has
been chosen deliberately or not, for the reader of the LXX, the term δῶρον
could have implied a similar double entendre as well.
Finally, the same conclusion can be drawn with respect to the transla-
tion of the verbal form ‫( וַ יַּ ְק ֵרב‬Judg 3:17). Here, LXX uses a form of the verb
προσφέρω (προσήνεγκεν). As in the Hebrew, this verb can have two con-
notations. Although προσφέρω, particularly in the books of Leviticus and
Numbers, is very often used in a cultic context as the translation equiva-
lent of the verb ‫“( ָק ַרב‬to approach, to bring near”), often in combination
with the noun θυσία (e.g. in Gen 43:26), the term has a profane connota-
tion as the translation equivalent of the Hiphil of ‫וַ יָּ בֺא‬: εἰσῆλθεν δὲ Ιωσηφ
εἰς τὴν οἰκίαν καὶ προσήνεγκαν αὐτῷ τὰ δῶρα (NETS: “And Ioseph came into
the dwelling, and they brought to him the presents . . .”).
In sum, whether or not the Greek double entendre was intentional
(but opinions diverge in this respect for the Hebrew as well), with the
exception of the rendering of Eglon’s name, the terminology in Greek can
be interpreted as referring to the profane offering of gifts by Ehud, but
equally to the cultic offering of the king by Ehud.39
38 Cf. the lexemes μαανα, μανα, μαναα(ν), μανααμ, μαναχ, μαννα, μαναειμ. See E. Hatch
and H. A. Redpath, A Concordance to the Septuagint and the Other Greek Versions of the Old
Testament (Including the Apocryphical Books) (2nd ed.; Grand Rapids: Baker, 1998), 292.
39 With regard to the question of how the reader of the LXX interpreted the meaning
of some Greek lexemes, see M. Harl, “La langue de la Septante,” in La bible grecque des
Septante. Du judaïsme hellénistique au christianisma ancien (Initiations au christianisme
ancien; eds. M. Harl, G. Dorival and O. Munnich; Paris: Cerf, 1994), 223–266, esp. 251–253.
judges 3:12–30 65
5. Further examples of wordplay
Besides the occurrences of wordplay within the characterisation of Ehud

as “left-hander” and Eglon as the young bull who will be sacrificed, there
are some minor elements in Judg 3:12‑20 in which the Hebrew author
seems to play with words. Again, the phenomenon of double entendre
seems to be prominent. The first one can be found in v. 16, where it is
said that Ehud made a dagger. The Hebrew text reads as follows: ‫וַ ּיַ ַעׂש לֹו‬
‫ ֵאהּוד ֶח ֶרב‬. According to the grammar of Joüon and Muraoka, this verse
functions as an example of the Hebrew use of the reflexive pronoun.40
Interpreting ‫ לֹו‬as such, the author would indicate that Ehud made a dag-
ger for himself. However, even if one accepts that Judg 3:16 is an example
of the reflexive usage of the personal pronoun, the particle preposition -‫ל‬
with the personal suffix could have another function as well in the pres-
ent context. In order to see so, let us compare it to a number of other
examples. Gen 19:3, for example, narrates how Lot prepares a feast “for
them”, i.e. the two messengers who visit him: ‫וַ יַּ ַעשׂ ָל ֶהם ִמ ְשׁ ֶתּה‬. In Exod
37:2, it is said that Bezaleel makes a crown of gold “for him”, i.e. the ark
(‫)וַ יַּ ַעשׂ לוֹ זֵ ר זָ ָהב‬. So it is at least possible that the author of Judg 3:16 uses
the personal pronoun with the particle preposition in a double sense: the
dagger that Ehud manufactures “for himself ” is actually made “for him”,
i.e. Eglon, who will be killed by the weapon. To a translator, it is not easy
to render this double entendre of the term. The LXX (A and B) has unam-
biguously translated one single meaning by rendering ‫ לוֹ‬as a reflexive pro-
noun (“to/for himself ”): καὶ ἐποίησεν ἑαυτῷ Αωδ μάχαιραν δίστομον (NETS:
“And Aod made for himself a double-bladed dagger”).
A final element of double entendre in the Hebrew text can be found
in the use of the term ‫ ָדּ ָבר‬in Judg 3:19–20. When Ehud returns to king
Eglon, he addresses him twice with parallelly structured words. In v. 19,
Ehud says: ‫“( ְּד ַבר ֵס ֶתר ִלי ֵא ֶליָך ַה ֶּמ ֶלְך‬I have a secret message for you, o
King”). When the servants have left the room, Ehud (v. 20) addresses the
king for the second time: ‫ֹלהים ִלי ֵא ֶליָך‬ ִ ‫“( ְּד ַבר ֱא‬I have a word of God for
you”). There can be no doubt that once more a term is interpreted differ-
ently by Eglon and the attentive reader of the text. For Eglon, the noun
‫ ָדּ ָבר‬refers to a secret message, which, moreover, is a message of God,41
40 P. Joüon and T. Muraoka, A Grammar of Biblical Hebrew (SubBi 14; Rome: Pontificio
Istituto Biblico, 1993), §146k.
41 It is interesting to note that within the Ehud pericope only in v. 20 the term ‫ֱאל ִֺהים‬
is used. Compare with the tetragram in Judg 3:12(ter), 15(bis), 28. J. M. Sasson, “Ethically
66 hans ausloos
and which lets him rise up from his throne (because of respect?). The
reader, on the contrary, can/will have interpreted ‫ ָדּ ָבר‬as a “thing”, refer-
ring to the dagger, which he interprets, moreover, as a divine weapon in
the hands of Ehud.42 In fact, it is not Ehud himself who will rescue Israel
by killing Eglon, but God who has sent Ehud as a deliverer (v. 15).43 It is
well possible that the author of Judg 3 has this double intention in mind,
although it is difficult to give a definitive answer in this regard.
The question at stake here pertains to the way in which the LXX trans-
lator has rendered this double entendre of the noun ‫( ָדּ ָבר‬which com-
monly occurs in Hebrew). In both verses (19 and 20), the LXX (A and B)
translates ‫ ָדּ ָבר‬by the Greek noun λόγος (θεοῦ). Next to the translation
equivalent λόγος, ‫ ָדּ ָבר‬has two other equivalents in the book of Judges:
besides πρᾶγμα (which clearly interprets ‫ ָדּ ָבר‬as “thing”), the term ῥῆμα
is also used. In Judg 11:10, for example, this term is an adequate ren-
dering of ‫ ָדּ ָבר‬in its meaning “word”. The elders of Gilead say to Jeph-
tah: “We will surely do conform to your word (‫”) ִכ ְד ָב ְרָך‬. However, in
Judg 8:1, where ‫ ָדּ ָבר‬clearly is used as “thing”, LXX equally reads ῥῆμα
(‫ית ָלּנוּ‬ ָ ‫— ָמה ַה ָדּ ָבר ַהזֶּ ה ָע ִשׂ‬τί τὸ ῥῆμα τοῦτο ἐποίησας ἡμῖν: “What is the
thing that you have done to us?)”. The same double meaning seems to
be true for the usage of λόγος, which, for example in Judg 11:28, is the
adequate rendering of ‫ ָדּ ָבר‬, meaning “word”: ‫וְ לֺא ָשׁ ַמע ֶמ ֶלְך ְבּנֵ י ַעמּוֹן‬
‫( ֶאל ִדּ ְב ֵרי‬NRSV: The king of the Ammonites did not heed the message
that Jephthah sent him”). The LXX here reads τῶν λόγων Ιεφθαε (A and
B). On the contrary, in Judg 21:11, where ‫ ָדּ ָבר‬seems to mean “thing”
(‫“—וְ זֶ ה ַה ָדּ ָבר ֲא ֶשׁר ַתּ ֲעשׂוּ‬This is the thing that you shall do”), LXX (A)
equally uses here the noun λόγος (καὶ οὗτος ὁ λόγος ὃν ποιήσετε).44 So, for
the LXX of Judg 3:19 and 20, it seems theoretically possible that the trans-
lator, as a sort of Hebraism, equally uses λόγος in a double sense: Ehud is
Cultured Interpretations,” 576 also mentions the possibility that the noun ‫ ֱאל ִֺהים‬func-
tions to denote a superlative: “the gravest message”.
42 In the book of Judges, the noun ‫ ָדּ ָבר‬occurs in Judg 2:4; 3:19, 20; 6:29(bis); 8:1, 3; 9:3,
30; 11:10, 11, 28, 37; 13:12, 17; 16:16; 18:7(bis), 10, 26; 19:19, 24; 20:7, 9; 21:11.
43 Moreover, texts such as 1 Sam 9:27–10:1 seem to mix both meanings of ‫ ָדּ ָבר‬. Samuel
announces to Saul that he has a ‫ ְדּ ַבר ֱאל ִֺהים‬for him. However, instead of saying some-
thing to him, Samuel takes a vial of oil and pours it on Saul’s head.
44 Cf., however, NETS’s translation: “And this is the word that you shall carry out” (καὶ
οὗτος ὁ λόγος ὃν ποιήσετε). The B text reads καὶ τοῦτο ποιήσετε (NETS: “And this is what you
will do”). Compare with LXX.D: “Das ist es, was ihr tun sollt” (A); “Dies sollt ihr tun (B)”.
judges 3:12–30 67
not only the messenger of a (divine) word, but the dagger in his hands is
a (divine) thing used to kill Eglon.45
6. Conclusion
There can be no doubt that the Hebrew text of Judg 3:12‑20 is character-
ised by several types of wordplay, of which the so-called double enten-
dre seems to be the most prominent one. As for the LXX translator, in
some instances—such as when he transliterates Eglon’s proper name—he
is not able to render the double meaning, although we cannot be sure
that he has noticed it. In other instances, however, the translator is look-
ing for some good solutions, such as, for example, with regard to Ehud’s
left-handedness. In translating the problematic Hebrew construction
‫ ִא ֵּטר יַ ד יְ ִמינֹו‬with the neologism ἀμφοτεροδέξιον, the translator clearly
indicates that he has understood the problem and the specificity of the
Hebrew text. Finally, it is sometimes not clear whether the translator him-
self also interpreted his Greek term as implying a double entendre, as is the
case with his rendering of ‫ ָדּ ָבר‬as λόγος. In any case, even if the translator
was using λόγος in a double sense, it is questionable whether the reader
of the LXX would have understood this usage.
When we compare these results with the creative way the translator
deals with the three Hebrew hapax legomena in this pericope,46 one has
to conclude that the LXX translator of Judges surely had the intention of
offering a qualitatively adequate rendering of his Hebrew Vorlage. In this
respect, and at least with respect to this pericope, his labelling as a weak
translator,47 needs to be nuanced. Indeed, on the one hand, the compari-
son of two different content-related criteria, i.c. the Greek rendering of
45 With regard to the possible theological connotation of λόγος as translation equiva-

lent of ‫ ָדּ ָבר‬, see C. Doignez and M. Harl, Le Deutéronome (BdA 5; Paris: Cerf, 2007), 40–43.
See, e.g., also ibid., 222 with regard to Deut 17:1, where ‫ ָדּ ָבר‬has been rendered by ῥῆμα:
“le mot rhema est un décalque de l’hébreu dabar, ‘parole’ ou ‘chose’ ”.
46 H. Ausloos and B. Lemmelijn, “Characterizing the LXX Translation of Judges,” 189:
“Summarizing the conclusions, one observes that the Greek translator of Judges has striven
to offer a meaningful rendering of the Hebrew hapax legomena in his Vorlage. Both with
regard to the mentioning of the weapon, which the translator deliberately calls a μάχαιρα
(dagger) with the length of a σπιθαμή in Judg 3:16, as well as with respect to the enigmatic
verses in Judg 3:22–23 in which he equally attempts to render an understandable Greek
text, the LXX translator demonstrates creativity and originality, each time taking serious
consideration of the particular context of the Ehud narrative”.
47 Cf. I. Soisalon-Soininen, Die Textformen der Septuaginta-Übersetzung des Richter-
buches (AASF 72; Helsinki: Suomalainen Tiedeakatemia, 1951), 60: “Man geht kaum fehl,
68 hans ausloos
Hebrew hapax legomena and the Greek rendering of wordplay, confirms

the rather creative attempts of the translator of Judges. On the other hand,
the confrontation ad extra with the more traditional evaluation of LXX
Judges as very literal seems to be questioned. Knowing that the charac-
terisation of a specific LXX translation must be based on as many differ-
ent criteria as possible, this preliminary evaluation is, of course, open to
debate and should surely be complemented by other research. Neverthe-
less, it opens up a surprising perspective.
wenn man Jdc. für die seinem Sprachgebrauch nach schwächste Übersetzung der ganzen
LXX hält”.
B or not B? The Place of Codex Vaticanus in textual
history and in Septuagint research
Siegfried Kreuzer
1. Introduction
One of the most basic facts in Septuagint studies is the primary text used
for analysis or comparison. This statement will hardly be challenged.
Yet it is a fact which needs to be reflected upon. For most studies on
the Septuagint, but also in studies on the biblical quotations in the New
Testament or even in comparisons with the daughter translations, the
Göttingen edition or Rahlfs’s so-called Handausgabe is often simply taken
as the starting point.1
Now, there is a reason for this approach: in his critical edition Rahlfs
wanted to reconstruct the Old Greek and the Göttingen edition maintains
that same goal. Yet it is and remains a goal only, and how this goal is pur-
sued and perhaps achieved depends on the methods and presuppositions
of the editor; and that’s exactly where Codex Vaticanus comes into play.
The critical editions use different manuscripts for their reconstructions, yet
Codex Vaticanus takes pride of place and dominates the editions. Rahlfs
states in his Handausgabe that he mostly relies on the Majuscules B, S and
A; among them—with a few exceptions as in the book of Judges—he basi-
cally followed Manuscript B, the Codex Vaticanus. The same holds true for
most of the volumes in the Göttingen edition.
1.1. The Importance of B for the Editions

Since the 16th century Codex Vaticanus has remained the most impor-
tant, i.e. the most highly esteemed, manuscript for Septuagint editions
and research. With the exception of the Aldina from 1518 and the Com-
plutensian Polyglot from 1514–1717, and starting with the Sixtina (1587),
1 Septuaginta: Vetus Testamentum Graecum. Auctoritate Academiae Scientiarum Gottin-

gensis editum (Göttingen: Vandenhoeck & Ruprecht, 1931ff.); A. Rahlfs and R. Hanhart, eds.,
Septuaginta: Id est Vetus Testamentum iuxta LXX interpretes (Stuttgart: Deutsche Bibelge-
sellschaft, 2006).
70 siegfried kreuzer
almost all editions of the Septuagint were more or less based on Codex
Vaticanus.
The first printed version of the Septuagint appeared in Venice in 1518
from the printer Aldine and is therefore called the Aldina. It used manu-
scripts which were available in the city at that time. At about the same
time the famous Complutensian Polyglot was prepared at the Alcala Uni-
versity in Complutum, near Madrid. For this edition manuscripts from
Rome and other cities were collected or borrowed.2 After about two gen-
erations the later Pope Sixtus V initiated and published a new edition;
manuscripts were sought and Codex Vaticanus became the basis of the
new edition, not only because it was available in Rome, but also because
of its excellent state of preservation and legibility.
Basically all Septuagint editions from the 17th through to the 19th
centuries,3 and even into the first half of the 20th century, have been based
on the Codex Vaticanus. More or less the only exception was Johannes
Ernestus Grabe’s edition of 1709–1720, which was based on Codex Alex-
andrinus. Practically all other editions are diplomatic editions of Codex
Vaticanus with an ever expanding critical apparatus, i.e. with the readings
of other manuscripts as they became available. This holds true also for the
editions of Holmes-Parsons,4 of Swete,5 and of Brooke-McLean.6 Even the
critical edition by Rahlfs and, to a large extent, the Göttingen edition rely
heavily on Codex “B.”
2 Cf. the basic study by F. Delitzsch, Studien zur Entstehungsgeschichte der Polyglot-
tenbibel des Cardinals Ximenes (Leipzig: Alexander Edelmann, 1871); and, more recently:
S. O’Connell, From Most Ancient Sources: The Nature and Text-Critical Use of the Greek Old
Testament Text of the Complutensian Polyglot Bible (OBO 215, Fribourg / Göttingen: Aca-
demic Press Fribourg / Vandenhoeck & Ruprecht, 2006).
3 H. B. Swete (An Introduction to the Old Testament in Greek [Cambridge: Cambridge
University Press, 1900], 182) counts at least 21 editions of that kind, deriving directly or
indirectly from MS B. Swete himself based his edition on a new collation of MS B.
4 R. Holmes and J. Parsons, eds., Vetus Testamentum Graecum Cum Variis Lectionibus
(Oxford: Clarendon Press, 1798–1820).
5 H. B. Swete, The Old Testament in Greek according to the Septuagint (3 vols.; Cam-
bridge: Cambridge University Press, 1887–1894).
6 A. E. Brooke and N. McLean, eds., The Old Testament in Greek according to the Text
of Codex Vaticanus supplemented from Other Uncial Manuscripts with a Critical Apparatus
Containing the Variants of the Chief Ancient Authorities for the Text of the LXX (Cambridge:
Cambridge University Press, 1906–1940).
b or not b? 71
1.2. Reasons for the Importance of Codex Vaticanus

There are good reasons for the importance of Codex Vaticanus. First, it
was and still is the oldest MS attesting almost all of the Old and the New
Testaments. There is a consensus that it was written in the 4th century,
but there is some debate about whether it was produced in Caesarea in
Palestine or in Egypt. The reason for Caesarea as a suggestion is its rela-
tion to Codex Sinaiticus; the reason for suggesting Egypt is the agreement
of the order of the books with the canon list of Athanasius and because of
some relation to the Egyptian text, especially in the Psalms. Yet both argu-
ments suppose that these traditions (the specific order of the canon list,
and the Greek Vorlage of the Egyptian text) had been available in Egypt
only. So what seems certain thus far is that Codex B originated in the
Eastern Mediterranean in the 4th century. It was written by at least two
scribes, and—as we will see—the Codex or its Vorlage was put together
from scrolls of different textual traditions.
As the ink had faded over the centuries, it was re-inked in the Middle
Ages, some time after the 10th century. This is evident because the so-
called enclitic nu has been re-inked according to the late Byzantine school
rule. The Codex probably came into the Vatican Library via southern Italy.
Unfortunately, the beginning of the Codex, i.e. most of the book of Gen-
esis, is missing.
There are only a few other large codices from late antiquity. There is
Codex Sinaiticus. The complicated history of its discovery, starting with
Tischendorf ’s journey to the St. Catherine Monastery in Sinai in 1844, does
not need to be repeated here.7 There are indications that this Codex origi-
nated in Caesarea, and one of its three or four scribes was possibly identi-
cal with a scribe of Vaticanus. The Codex has been dated between 340 and
360 C.E. It was probably bound only in the 6th century and brought to
St. Catherine’s soon afterwards. Most important is the fact that the Codex
underwent several corrections. Corrector A (Ca) seems to have worked in
the 6th century; he added readings from other traditions. Unfortunately
large parts of the Codex are missing: besides some newly published chap-
ters of Joshua and Judges, the older historical books and also large parts
7 It will be of interest that the famous debate about the legitimacy of its being taken
away from the St. Catherine Monastery has now been resolved, as the documents dealing
with the donation of the Codex to the Tsar by the monastery have been found in Moscow;
cf. C. Böttrich, “Neue Dokumente zur Geschichte des ‘Codex Sinaiticus’,” Early Christianity 1
(2010): 605–613.
of the Pentateuch are missing. With exceptions in a few books, the text,
at least in the Old Testament, is very close to that of Codex Vaticanus.
The remarks of Corrector A present an older textual tradition: some
of the corrections seem to presuppose a Hexaplaric text, others are close
to the Lucianic/Antiochene text. The fact that these corrections have been
added only shows that this tradition was important for the corrector, but
not how old the sources for these corrections are. The Codex with all of its
parts is now accessible in an excellent edition on the internet.8
The next codex is Codex Alexandrinus, generally dated to the 5th cen-
tury. There is practically no information about the origin of the Codex.
Around 1300 it came to Alexandria and in 1627 it was presented to King
Charles I of England. Its text is considered less consistent than Vaticanus,
but this is also a question of the standard for comparison, as Codex Vat-
icanus is also mixed, at least between kaige and non-kaige sections. In
Leviticus, Isaiah and Jeremiah, Alexandrinus seems to be close to Vati-
canus; in other parts it shows Hexaplaric influence; but it is also close
to the Antiochene text; and it has many idiosyncratic readings. As men-
tioned above, it was used for the edition by Grabe 1709–1720. Grabe also
wrote an essay about Alexandrinus attesting the oldest text in the book
of Judges, which was evidently the reason for its specific presentation of
the book of Judges in the editions of Brooke-McLean and Rahlfs. Different
from Vaticanus and Sinaiticus, the beginning of the codex, i.e. the whole
book of Genesis, is preserved in Codex Alexandrinus.
Regarding the older historical books, I would like to mention two more
codices. Codex M, which is Codex Coislianus from the 7th century, is
now in Paris. It comprises the Octateuch and the historical books. It has
many marginal readings with Hexaplaric material. In the historical books
its text is a mixture between the kaige text and the Antiochene text. Its
close counterpart is Codex N, also called Codex V like Codex Venetus.
The reason for the different designations is that N and V are two parts of
one codex. One part is in Rome, while the other part is in Venice. Because
of its close relation to M, the Roman part has been named N in Brooke-
McLean. It was later identified as the first part of a codex whose second
part had become known as Codex Venetus. Rahlfs therefore uses one and
the same siglum for both parts. Codex V dates from the 8th century and
8 http://www.codexsinaiticus.org.
b or not b? 73
in both parts it comprised the entire Old Testament.9 Both codices repre-
sent a text which is partly close to the Antiochene or Old Greek texts and
partly to kaige texts, just like Codex Vaticanus.
Looking at these codices with their later origin and their idiosyncrasies,
it is understandable that Codex Vaticanus became and still is considered
the most important single witness to the Septuagint. Yet there are other
and earlier documents as well.
1.3. Other Early Manuscripts and Witnesses of the Septuagint

The picture opens out if we take into account other manuscripts and the
witnesses of authors writing in Greek and the daughter translations. Besides
some manuscripts from pre-Christian times such as Papyrus Fouad 266,
several fragments from Qumran, and especially the Greek Minor Prophets
scroll from Naḥal Ḥever, today papyri from the 1st to the 4th century C.E.
from different parts of the Old Testament have been identified. For most
books of the Septuagint Codex Vaticanus is currently just one of many
witnesses that are relatively contemporary, but some witnesses are several
centuries older. I cannot go into the details here, but only refer to the
impressive and interesting list “Das handschriftliche Material für die ein-
zelnen Bücher des Alten Testaments” in the Verzeichnis der Griechischen
Handschriften edited by Detlef Fraenkel.10
Interestingly, amidst this wealth of manuscripts there is an excep-
tion for the older historical books, i.e. 1–4 Reigns, 1–2 Chronicles and 1–2
Esdras. There, with the exception of 2 Chron, Codex Vaticanus is still the
oldest witness of the Greek text in the form of a codex.
Yet also in these parts of the Septuagint there are other important
witnesses to the Greek text as well. First, there is the Jewish writer Jose-
phus, who in his Jewish Antiquities referred to the historical books and
quoted them; and even in his allusions it is sometimes possible to identify
the specific form of the text that he is referring to. Adam Mez made an
explicit comparison already in 1895.11 Later Henry St. J. Thackeray did the
9 It probably was also originally a codex of the whole Bible, with the NT portion now
lost; cf. A. Rahlfs and D. Fraenkel, Die Überlieferung bis zum VIII. Jahrhundert (Vol. 1.1 of
Verzeichnis der griechischen Handschriften des Alten Testaments; Göttingen: Vandenhoeck
& Ruprecht, 2004).
10 Ibid., 472–497.
11 A. Mez, Die Bibel des Josephus, untersucht für die Bücher V–VII der Archäologie (Basel:
Jaeger & Kober, 1895). Mez’s results have been pushed aside by Rahlfs in A. Rahlfs, Lucians
Rezension der Königsbücher (Septuaginta-Studien III; Göttingen: Vandenhoeck & Ruprecht,
1911 [repr. 1965]), but they were confirmed by H. St. J. Thackeray, Josephus: The Man and the
same in connection with his studies on Josephus.12 The relevant passages

can now be found in the critical apparatus of the Madrid edition of the
Antiochene text.
But there are also Christian writers who quoted the Septuagint, and at
least some of them quoted not only Genesis, Psalms and the Prophets,
but also the historical books. Unfortunately, the older fathers like Justin,
Irenaeus or Clement of Alexandria do not offer much material from these
books. But in the 4th century there were Theodoret of Cyrus with his com-
mentaries, Chrysostom with his sermons, and other, although less prolific,
authors such as Asterius Sophista. With some caution their works can be
seen as witnesses of the texts they used. Theodoret in his commentaries
is an especially excellent witness to the biblical text he used, which is the
so-called Antiochene or Lucianic text.13
Considering this environment, one must say that the large old codices
have their special importance, but they are not the only textual witnesses.
Beyond that, looking at the whole field, one must admit that Vaticanus
(and Sinaiticus, where extant) represent a minority position within the
field, and that Codex Vaticanus too is of a mixed character as can be seen
especially in the change between kaige and non-kaige sections.
To appreciate the whole picture, a look at the daughter translations,
especially the Old Latin, is necessary. The Old Latin version originated
in the 2nd century, probably at more than one place, probably in more
than one attempt, and probably with even some Jewish origins. But this
does not further impact on our discussion. With its origin in the 2nd cen-
tury the Old Latin is older than Origen’s Hexapla and existed long before
Lucian. It therefore witnesses to a pre-Hexaplaric and a pre-Lucianic text.
As the Old Latin was produced for and among Latin-speaking people in
the western part of the Roman Empire, it must have been based on a
Greek Vorlage used in those western areas.14
Historian (New York: Jewish Institute of Religion, 1929), and E. Ulrich, The Qumran Text of
Samuel and Josephus (Chico: Scholars Press, 1978).
12 Thackeray, Josephus.
13 Cf. the editions: N. Fernández Marcos and A. Sáenz-Badillos, Theodoreti Cyrensis
quaestiones in octateuchum (TECC 17, Madrid: CSIC, 1979); N. Fernández Marcos and
A. Sáenz-Badillos, Theodoreti Cyrensis quaestiones in reges et paralipomena (Madrid, 1984);
Theodoret of Cyrus, The Questions on the Octateuch, Vol. 1. On Genesis and Exodus; Vol. 2.
On Leviticus, Numbers, Deuteronomy, Joshua, Judges, and Ruth (ed. J. F. Petruccione. Trans.
R. C. Hill, Washington: Catholic University of America Press, 2007).
14 Unfortunately the state and tradition of the Old Latin text is complicated. There are
only a few codices and fragments from palimpsests and, for some books, marginal notes in
biblical manuscripts. In spite of these problems and the fragmentary character of the Vetus
b or not b? 75
Something similar can be said about the earlier Egyptian translation,

the so-called Sahidic version from Upper Egypt. Different from the younger
Bohairic translation of Lower Egypt, the Sahidic version was produced in
the 3rd century for people who no longer spoke Greek or who never had
before. Through its Vorlage this Egyptian translation attests a Greek text
from the 3rd century, which was probably older than the Hexapla and also
certainly predates the floruit of Lucian.15
Taking these things together, there is an obvious conclusion: if there
is an agreement between the Antiochene text and the Old Latin and the
Sahidic translation, then we have a textual tradition which was wide-
spread from Syria to Egypt and to the West, and which most probably is
very old, if not identical with the original Septuagint or the so-called Old
Greek. From this wider picture we return to B, the Codex Vaticanus.
2. Codex Vaticanus and the other forms of the text
2.1. Some Peculiarities of Codex Vaticanus

In this section we will concentrate on the historical books. There is a spe-
cial problem in the book of Judges. Besides the textual tradition in Vati-
canus, there is definitely a different textual form in Codex Alexandrinus.
Brooke-McLean highlighted this text by using a different font.16 Rahlfs
went two steps further in his edition: (1) He juxtaposed both traditions;
and (2) he used the text of Alexandrinus as the basis for a critical recon-
struction. In doing so, Rahlfs presented what—at least in his opinion—
Latina, the Old Latin is an important witness to an old stage of the Septuagint. For the Old
Latin texts from Samuel, Kings and Chronicles cf. the apparatus in N. Fernandez Marcos
and J. R. Busto Saiz, El Texto Antioqueno (TECC 53/56/60; Madrid: CSIC, 1989/1992/1996).
15 For the Sahidic Version, cf. S. P. Brock, “Bibelübersetzungen I.5 Bibelüber-setzungen
ins Koptische 2. Altes Testament,” TRE 6: 199–200. Especially for the older Coptic Ver-
sions there is still much to be done. In a large project Karlheinz Schüssler has started
to collect and catalogue the manuscripts: K. Schüssler, ed., Das sahidische Alte und Neue
Testament: Biblia Coptica: Die koptischen Bibeltexte (Forschungsinstitut für Ägyptenkunde
und Koptologie der Universität Salzburg; Wiesbaden: Harrassowitz, 1995ff.).
For the books of Samuel there is the important edition by James Drescher: J. Drescher,
ed., The Coptic (Sahidic) Version of Kingdoms I, II (Samuel I, II) (CSCO 313–314; Scriptores
Coptici 35–36; Louvain: Secretariat du CSCO, 1970).
16 One reason for this decision was most probably the old treatise by Johannes Ernestus
Grabe on the priority of Codex Alexandrinus in the book of Judges: J. E. Grabe, Epistola Ad
Clarissimum Virum, Dn. Joannem Millium . . . Quâ ostenditur, Libri Judicum Genuinam LXX.
Interpretum Versionem eam esse, quam Codex Alexandrinum exhibet (Oxford, 1705).
was the Old Greek, which he called text A, and presented below that the
text of Codex B, as the evidently secondary text.17
Yet in spite of the evident secondary status of B in Judges, in the subse-
quent books Rahlfs basically followed the Codex Vaticanus. The basis for
this procedure was his investigation of the Lucianic text in Kings, pub-
lished in 1911.18 To put it briefly: in the 1860s some MSS had been iden-
tified as presenting the Lucianic or Antiochene text. Julius Wellhausen
in his book on the text of Samuel19 found that often those manuscripts
presented the oldest readings or even confirmed his conjectures. Accord-
ingly, Paul de Lagarde started his search for the Old Greek text by edit-
ing the Lucianic textual tradition. This line of research was supported by
Adam Mez, who—as mentioned above—had compared the Antiquities of
Josephus and had found that Josephus’ biblical quotations basically agree
with the Lucianic text. This meant that the Lucianic text is not the result
of a late reworking by Lucian around 300, but basically already existed in
the 1st century. Besides that, there was the Old Latin text from the 2nd
century, which also agreed with the Lucianic text, most of the time against
Vaticanus.
Contrary to this, Rahlfs basically defended the opposite view in his
investigation. For him Codex Vaticanus without a doubt represented the
oldest text and everything had to be compared with it. Whereas the Luci-
anic text was only a text type connected with the authority of that mar-
tyr (as the famous remark of Hieronymus can be understood as saying),20
Codex Vaticanus was equated with the Old Greek and all the differences
against Vaticanus were interpreted as the result of the comprehensive
redactional activity of Lucian.
For this purpose Rahlfs minimised all evidence that indicated the con-
trary: the agreements with Josephus were explained away; he accepted
only some forms of names as original. Also the evidence of the Old Latin
17 It should be noted that text “A” is not identical with Codex A, while text “B” is basi-
cally identical with Codex B (in the footnotes to B, Rahlfs only mentions corrections from
within the Codex).
18 Rahlfs, Lucians Rezension.
19 J. Wellhausen, Der Text der Bücher Samuelis (Göttingen: Vandenhoeck & Ruprecht,
1871).
20 “Alexandria et Aegyptus in Septuaginta suis Hesychium laudat auctorem, Constan-
tinopolis usque Antiochiam Luciani martyris exemplaria probat, mediae inter has pro-
vinciae palestinos codices legunt, quos ab Origene ela boratos Eusebius et Pamphilius
vulgaverunt, totusque orbis hac inter se trifaria varietate conpugnat.” Jerome, Preface to
Chronicles. Cf. R. Weber and R. Gryson, Biblia Sacra iuxta Vulgatam Versionem (Stuttgart:
Deutsche Bibelgesellschaft, 2007).
b or not b? 77
was explained away as being secondarily influenced by the Lucianic tra-

dition. In the same way, those quotations in the New Testament which
agreed with the Lucianic text were explained as a secondary influence
from the New Testament on the Lucianic manuscripts.
By this procedure—which he also had applied in his investigation on
the Psalms in 190721—Rahlfs came to his understanding of the Lucianic/
Antiochene text, which became most influential for the following decades
and in Septuagint research in general. It became one of the basic prin-
ciples for Rahlfs’s own edition and also for most editors in the Göttingen
edition, e.g. Ziegler and others.
In a description of the editorial work for the Septuagint volumes, Udo
Quast explained in 2000 that at the beginning of the editorial work one
knows little about the manuscripts and the recensional activities. “Lediglich
von dem Vorkommen der zwei großen christlichen Rezensionen des Ori-
genes und Lukian kann von vornherein—oder wenigstens in den meisten
Büchern—ausgegangen werden. Für sie stehen die Rezensionsmerkmale
außerdem weitestgehend fest.”22
2.2. The Problem of the Lucianic Text

What are these “well-known” characteristics of the Lucianic redaction?
Rahlfs mentioned three main traits: the first is the addition of articles;
the second is the addition of explanatory words such as the name of a
person speaking or acting. Yet those traits are irregular. Lucian not only
added words and articles, but he sometimes apparently deleted them.23
For Rahlfs this was not a question of his analysis; rather he made this
irregularity into a further characteristic of Lucian’s work and stated: “Der
Hauptcharakterzug dieser Rezension ist das Fehlen eines klaren Prinzips”.24
21 A. Rahlfs, Der Text des Septuagintapsalters (Septuaginta-Studien II; Göttingen: Van-
denhoeck & Ruprecht, 1907 [repr. 1965]).
22 “Only the occurrence of the two extensive Christian recensions, those of Origen and
of Lucian, can be assumed, at least for most of the books. Beyond that, the characteris-
tics of these recensions are well known and practically certain”. U. Quast, “Einführung in
die Editionsarbeit,” in Der Septuaginta-Psalter und seine Tochterübersetzungen (MSU 24,
eds. A. Aejmelaeus and U. Quast; Göttingen: Vandenhoeck & Ruprecht, 2000), 387–399
(394–395).
23 See in the synopsis below. Articles added in vv. 2, 6, 10; deleted in v. 10. Explanatory
words are added in: v. 2; deleted in v. 10.
24 “The main characteristic of this recension is that it has no clear principle”. Rahlfs,
Lucians Rezension, 293.
Yet one may say that the intention of Lucian was to improve the “Greek-
ness” of the text.25
This basic idea was taken over by Ziegler, among others. In his edition
of Jeremiah he mentioned the same characteristics as Rahlfs, and he also
wrote about Lucian’s irregularity. “Konsequenz war nicht seine Stärke”.26
Things changed with the discoveries from Qumran, especially the bib-
lical texts from Qumran and other places in the Judean desert. The first
scroll of the book of Samuel (4QSama) presented a text that was very close
to the Lucianic text, or rather its Hebrew Vorlage. This means that a sub-
stantial portion of the idiosyncrasies of the Lucianic text is not Lucianic,
but rather old, if not original. This new evidence from Qumran could not
be pushed aside as Rahlfs had done with the evidence from Josephus, from
the New Testament or from the Old Latin. The text from Qumran could
not have been influenced later by the Lucianic tradition. Qumran schol-
ars such as Frank Moore Cross, Eugene Ulrich or Emanuel Tov therefore
accepted the importance and the great age of the Lucianic text, or of the
Antiochene text as more accurate because it is more neutral. Yet they also
25 Rahlfs also mentions Lucian’s Atticising tendency. For Rahlfs this trait was less
important than it became in later scholarship: he mentions Atticising improvements (“atti-
sierende Verbesserungen”), but there are not many and they also are done irregularly.
The word Attic (“attisch”) is mentioned rather late in the study under the heading “other
changes” (“sonstige Änderungen”) (ibid., 176), where he first notices that the Aorist ειπα
was already in use in Attic, that it became common in the Hellenistic period, and that it
is found in both MSS A and B (ibid., 176–177); therefore it is not a sure sign. Indeclinable
δυο is used by the “Attiker”, and is found 4 times in L, while δυσιν, which was especially
favoured in Hellenistic times, is kept by Lucian in 2 Kgs 5:23 (“indeklinables δυο [kommt]
schon bei den Attikern vor (I 2,32; 22,31; II 21,5; 23,12)”), whereas δυσιν represents a “jüngere
Bildung, die in hellenistischer Zeit besonders beliebt ist, in II 5,23 [die] von Lukian beibe-
halten wurde”; (ibid., 259). For the change from ιερεις to ιερεας Rahlfs mentions 5 occur-
rences, but also 7 occurrences where Lucian did not make a change (ibid., 263). Altogether,
Rahlfs mentions seven forms or grammatical phenomena (ibid., 176, 204, 259, 260, 262, 263,
279) with 1 to 7 attestations (except ειπα, for which he does not mention any attestations),
and in almost every case also exceptions. Rahlfs concludes: “But Lucian is not a rigorous
Atticist because, in that case, he would have changed much more. And there are also cases
where Lucian uses a non-Attic form instead of the Attic form, e.g., τριτον instead of τρις,
also . . . I 6,7 πελυξ instead of πελεκυς and II 18,27 τον ουρον instead of το ουρον” (ibid., 281;
“Aber Lucian ist keineswegs strenger Attizist, er hätte sonst sehr viel mehr ändern müssen,
als er getan hat. Auch kommen Fälle vor, wo gerade L eine nichtattische statt der attischen
Form hat, wie τριτον statt τρις, ferner (in Abs. 1 nicht aufgeführt) I 6,7 πελυξ st. πελεκυς und
II 18,27 τον ουρον statt το ουρον”).
26 “Consistency was not his strength”. J. Ziegler, Beiträge zur Ieremias-Septuaginta
(MSU 6; Göttingen: Vandenhoeck & Ruprecht, 1958), 163.
b or not b? 79
embraced the old view about the Lucianic text and therefore developed
some compromise models.27
2.3. Naḥal Ḥever, Kaige and “Lucian Redivivus”

Another text has also become very important: the Greek Minor Prophets
scroll from Naḥal Ḥever. As is well known, this text has become the basis
for the identification of the kaige recension.28 This kaige recension, with
its close and formalistic adaptation of the Old Greek to the Hebrew text, is
not restricted to the Minor Prophets, but can be identified in other books
as well.29 What is most important for the current discussion is that the
kaige recension can also be found in Samuel and Kings. Barthélemy took
up an old observation by Thackeray, who had discerned four different
sections in Reigns, which he named with the Greek letters of the respec-
tive books.30 The sections βγ (2 Kgdms 10–3 Kgdms 2) and the section γδ
(3 Kgdms 22–4 Kgdms 25) are the kaige sections. These sections show the
same traits as the Naḥal Ḥever scroll. The other sections of Samuel and
Kings do not share these characteristics. Therefore they may be called the
non-kaige sections. The Greek text of these passages is of a different char-
acter and closer to the Old Greek.
It needs to be mentioned that this division refers to Codex Vaticanus
alone. While the kaige recension is a widespread phenomenon and can
be found in most books beyond the Pentateuch, the division within 1–4
Kgdms is a unique feature of Codex Vaticanus. It must have arisen from
the combination of different scrolls with different text types.
But what about the Antiochene text in these books? After identifying the
kaige sections, Barthélemy posed a logical question: if Codex Vaticanus is
secondary in the kaige sections, where is the Old Greek to be found? Can
it be identified in some other manuscripts or has it been lost? Barthélemy
compared the kaige text (in his book still called the Palestinian text) with
the Antiochene text and discovered that it must have been the base text
for the kaige recension. The questions and the answers can well be seen
27 E.g. E. Tov, “Lucian and Proto-Lucian,” RB 79 (1972): 101–113, for whom the Lucianic
text contains either the ancient Septuagint or an ancient Septuagint.
28 D. Barthélemy, Les Devanciers d’Aquila (VTSup 10; Leiden: Brill, 1963).
29 Ibid., 89: “Études sur quelques membres déjà connus du groupe καιγε.”
30 H. St. J. Thackeray, “The Greek translators in the four books of Kings,” JTS 8 (1907):
262–266, and idem, The Septuagint and Jewish Worship. A Study in Origins (Schweich Lec-
tures, London: Oxford University Press, 1921).
in the headings of the relevant chapters in his book. At first, by comparing

the two texts, he comes to the conclusion that there is a basic uniformity
between the two text forms: “identité de base entre la forme antiochienne
et la forme palestinienne du text grec”.31 This basic unity between the Pal-
estinian text and the Antiochene text matches clear differences that can
be explained in one way only: “la forme antiochienne ne peut être issue
de la forme palestinienne par abâtardissement”,32 i.e. the Antiochene text
cannot have originated from the kaige text. The dependence is the other
way round: the Antiochene text is the older text, probably very close to
the Old Greek or even identical to it (although not without corruptions in
the course of its transmission).33
This means that the discovery of the kaige recension and the new evalu-
ation of the Antiochene text are two sides of the same coin. This is not the
result of mere speculation or trepidation about the total absence of the
Old Greek for these texts; it is simply because the Antiochene text indeed
represents the text which was used and revised by the kaige revisers.
At about the same time as Barthélemy published his discoveries (1963),
a young man at Oxford had almost finished his dissertation. Sebastian
P. Brock worked on 1 Samuel in his dissertation, which was accepted in
1966.34 He did so in the rather traditional way and, just as many others
had done, followed Rahlfs. He did not use the few papers on the Qumran
scrolls that F. M. Cross had published by that time,35 but he took great
care with the Hexaplaric material and with his analysis of the Lucianic
recension. Evidently, he was not happy with Barthélemy’s book in which
the Lucianic recension had disappeared and the Antiochene text had
become the Old Greek, although with corruptions over time.
In 1965 Brock had the opportunity to deliver a paper on Barthélemy’s
book. Basically he defended the old ideas about Lucian. He gave it the fit-
31 Cf. Barthélemy, Devanciers, 92–102.

32 Cf. ibid., 110–113.
33 Cf. the famous statement: the Antiochene text is “la vielle septante, plus ou moins
corrompue et abatardie” (ibid., 127).
34 Printed much later: S. P. Brock, The Recensions of the Septuagint Version of 1 Samuel
(1966) (Torino: Silvio Zamorani, 1996).
35 F. M. Cross, “A New Qumran Biblical Fragment Related to the Original Hebrew
Underlying the LXX,” BASOR 132 (1953): 15–26; and idem, “The Oldest Manuscripts from
Qumran,” JBL 74 (1955): 147–72. See also W. F. Albright, “New Light on Early Recensions of
the Hebrew Bible,” BASOR 140 (1955): 27–33.
b or not b? 81
ting title “Lucian redivivus”.36 In this paper of only 6 pages he picked out
a few variant readings. His main argument was that the Lucianic text has
the same traits in the kaige and in the non-kaige section. He combined
this correct statement with the assumption that Codex Vaticanus in the
non-kaige sections represents the Old Greek and, therefore, the Lucianic
text is secondary. And he concluded: therefore, the Lucianic or Antiochene
text must also be secondary in the kaige sections.
At first this sounds convincing, but a closer look shows that Brock
jumps from a difference within Codex Vaticanus—i.e. the difference sig-
nalled by the terms kaige and non-kaige—to dating the Antiochene text.
This is simply false reasoning. The difference within the text of Codex
Vaticanus leads to a different relation to the Old Greek. But that is a prob-
lem within Vaticanus and does not say anything about the Lucianic text.
The variant characters and ages within the text of Vaticanus do not affect
the character of the Antiochene text. What changes is not the Antiochene
text, but its relation to the different parts of Vaticanus because of their
different characteristics.
Interestingly, Brock’s small paper was never seriously examined; at
least there is no publication.37 Yet this paper became most influential. Its
consequence was that Barthélemy’s discovery of the kaige recension was
widely accepted, while the other side of the coin, his new evaluation of
the Antiochene text, was pushed aside; the original combination of the
two facts remains practically unknown. Representative of this situation
is the statement in Fernandez Marcos’s “Introduction to the Septuagint”,
where he summarises the importance of Qumran/Naḥal Ḥever and Bar-
thélemy’s discovery only in regard to the kaige recension:
With the obligatory refinements in matters of detail, Barthélemy’s funda-
mental thesis, according to which these fragments belong to a consistent
revision of the LXX to bring it close to a Hebrew text very similar to, but not
identical with, the proto-Masoretic text, has been firmly accepted. Some of
the particular features of this revision which Barthélemy noted, and others
identified in later studies, can be debated . . . However, there is absolutely no
36 S. P. Brock, “Lucian redivivus. Some Reflections on Barthélemy’s Les Devanciers

d’Aquila,” in Studia Evangelica, Vol. V, Papers presented to the Third International Congress
on New Testament Studies held at Christ Church, Oxford 1965 (TUGAL 103; ed. F.L. [sic!]
Cross; Berlin: Akademie Verlag, 1968), 176–181.
37 For a first analysis and discussion see S. Kreuzer, “Lucian Redivivus or Barthélemy
and Beyond?” in IOSCS Congress Volume Helsinki 2010 (ed. M. K. H. Peters; Atlanta,
forthcoming).
doubt that these fragments belong to the LXX, which we knew through more
reliable ancient witnesses, but it was revised to adapt it with greater literal-
ism to the current Hebrew text . . . The finds from Naḥal Ḥever, together with
its general interpretation within the framework of the early history of the
LXX provided by Barthélemy, became an obligatory reference point for all
later studies.38
Lucian as a reviser was indeed revived by Brock and many Septuagint
scholars still take it for granted that there was a general Lucianic recension39
with its specific characteristics.40 For many, Codex Vaticanus is still the
text most close to, or even more or less identical with, the Old Greek, at
least in the non-kaige sections. This problem is not only relevant in Sam-
uel and Kings and in the other historical books such as Judges, Chronicles
and 2 Esdras, but also in other books, i.e. in the prophetic and in some
poetic books.
3. Evaluation and a new approach
3.1. The Relation of Kaige to the Antiochene Text

The typical phenomena can be seen in practically any kaige text in the
historical books. As an example, I use a passage from where my discover-
ies began, i.e. 2 Sam 15.
3.1.1.
The following synopsis41 of 2 Sam 15:2b, 5–6, 10 shows all the typical char-
acteristics as Rahlfs has proposed, and it also shows that Rahlfs even in
the kaige section almost exclusively understood the text of Codex Vati-
canus as the oldest text (except v. 2b with the introduction of ὁ ἀνήρ).
38 N. Fernández Marcos, The Septuagint in Context: Introduction to the Greek Version of
the Bible (Leiden: Brill, 2000 = Atlanta: Society of Biblical Literature, 2009), 72.
39 The question is not about Lucian as a person of the Syrian church, living around
300 C.E. (although his role sometimes has been questioned), nor that there are remarks in
ancient manuscripts which relate specific texts to (someone named) Lucian, but about the
assumed extensive recensional activity of Lucian (or a person of that time).
40 Representative again is Fernández Marcos, Septuagint in Context, 229: “However, no-
one has doubted the peculiar nature of the Lucianic or Antiochene text in the historical
books (Samuel‑Kings‑Chronicles).”
41 In the following tables, I use the vocalised MT as a matter of convenience and also
to indicate that the “text” never consisted of consonants only, but was always a “vocalised”
text with a reading tradition.
b or not b? 83
The typical addition of the article can be seen in v. 2b line 7 (2x!), v. 5

line 2; v. 6 line 8; v. 10 line 3 (2x). The addition of an explanatory word is
found in v. 2b line 6 (in this case even a whole sentence καὶ ἀπεκρίνατο ὁ
ἀνήρ) and v. 6 line 8 (παντῶν). There is also a change in words: σάλπιγγος
instead of κερατίνης (v. 10 line 7) is not exactly the same thing, but it better
represents the function.
Also the opposite can be identified: in v. 2b line 2 there is no πρóς; in
v. 10 line 7 both articles are deleted, and in line 9 the word βασιλεὺς is
missing. Lucian indeed worked in irregular and even contradictory ways.
Why would he add two articles in v. 10, when he deletes the articles a few
words later? Why would he add some words in v. 2b and delete a word
in v. 10?
On the other hand, taking up Barthélemy’s identification of the kaige
recension, the translation is indeed closer to the Hebrew, not only with
regard to πρóς in v. 2b line 2, the missing sentence in line 5, and the miss-
ing counterpart for παντῶν in v. 6 line 8, but also with the word κερατίνη
for ‫ ׁשפר‬in v. 10 line 7.
Yet the representation of the Hebrew in the kaige text is also question-
able. It follows the rules for the article only partially, as in v. 2b line 8,
where the determination of ‫ עבדך‬is expressed by ὁ δοῦλός σου, or in v. 10
line 7, where the determination is mirrored in τὴν φωνὴν τῆς κερατίνης.
But there are also other cases: the determination of ‫מאחד ׁשבטי־יׂשראל‬
has no article as a counterpart, nor does ‫ אנׁשי יׂשראל‬in v. 6 line 8 or the
‫ ׁשבטי יׂשראל‬in v. 10 line 3.
This seeming inconsistency can be explained by the underlying herme-
neutics: for the kaige revisers not only were the grammar and meaning
important but also, and even more so, the surface of the text. The result
was an isomorphic translation (or better, adaptation) of the text that mir-
rored its Vorlage. This explains the seeming irregularity: ‫ׁשבטי יׂשראל‬
(v. 2b line 7 and v. 10 line 3) and ‫( אנׁשי יׂשראל‬v. 6 line 8) are determined
genitive constructions, but there is no visible article. Also ‫ איׁש‬in v. 5 line 2
has no article; therefore there is no article with ἄνδρα either. On the other
hand, where there is an article or a similar visible element42 in Hebrew,
there is also an article in the kaige text.
42 This refers especially to the nota accusativi ‫את‬. As ‫ את‬is used before a determined
object, it has basically the same function as the article.
2 Sam/2 Kgdms 15:2b, 5–6, 10

MT KR (Rahlfs/B) Ant (Madrid Edition)
‫ וַ ּיִ ְק ָ ֙רא‬2b καὶ ἐβόησεν καὶ ἐκάλει
‫ַא ְב ָׁש ֤לֹום ֵא ָל ֙יו‬ πρὸς αὐτὸν Αβεσσαλωμ αὐτὸν Αβεσσαλωμ
‫אמר‬ ֶ ֹ ‫וַ ּ֗י‬ καὶ ἔλεγεν αὐτῷ καὶ ἔλεγεν αὐτῷ
‫יר ַ֔א ָּתה‬ ֙ ‫י־מ ֶּז֥ה ִע‬
ִ ‫ֵא‬ ἐκ ποίας πόλεως σὺ εἶ ἐκ ποίας πόλεως εἶ σύ
καὶ ἀπεκρίνατο ὁ ἀνήρ
‫אמר‬ ֶ ֹ ‫וַ ּ֕י‬ καὶ εἶπεν [ὁ ἀνήρ > B] καὶ ἔλεγεν
‫ֵמ ַא ַ ֥חד ִׁש ְב ֵ ֽטי־יִ ְׂש ָר ֵ ֖אל‬ ἐκ μιᾶς φυλῶν Ισραηλ ἐκ μιᾶς τῶν φυλῶν τοῦ
‫ַע ְב ֶ ּֽדָך׃‬ ὁ δοῦλός σου Ισραηλ
ὁ δοῦλός σου
‫ וְ ָהיָ ֙ה‬5 καὶ ἐγένετο καὶ ἐγίνετο
‫ב־איׁש‬
ִ֔ ‫ִּב ְק ָר‬ ἐν τῷ ἐγγίζειν ἄνδρα ἐπὶ τῶ προσάγειν τὸν
ἄνδρα
‫ְל ִה ְׁש ַּת ֲחֹו֖ ת ֑לֹו‬ τοῦ προσκυνῆσαι αὐτῷ τοῦ προσκυνεῖν αὐτῷ
‫וְ ָׁש ַל֧ח ֶאת־יָ ֛דֹו‬ καὶ ἐξέτεινεν τὴν χεῖρα καὶ ἐξέτεινε τὴν χεῖρα
‫וְ ֶה ֱח ִ ֥זיק ֖לֹו‬ αὐτοῦ καὶ ἐπελαμβάνετο αὐτοῦ
αὐτοῦ καὶ ἐπελαμβάνετο αὐτοῦ
‫וְ ָנ ַׁ֥שק ֽלֹו׃‬ καὶ κατεφίλησεν αὐτόν καὶ κατεφίλει αὐτόν
‫ וַ ּיַ֙ ַעׂש ַא ְב ָׁש ֜לֹום‬6 καὶ ἐποίησεν Αβεσσαλωμ καὶ ἐποίει Αβεσσαλωμ
‫ַּכ ָּד ָ ֤בר ַהּזֶ ֙ה‬ κατὰ τὸ ῥῆμα τοῦτο κατὰ τὸ ῥῆμα τοῦτο
‫ְל ָכל־יִ ְׂש ָר ֵ֔אל‬ παντὶ Ισραηλ παντὶ Ισραηλ
‫ֲא ֶׁשר־יָ ֥בֹאּו ַל ִּמ ְׁש ָ ּ֖פט‬ τοῖς παραγινομένοις εἰς τοῖς παραγινομένοις εἰς
‫ל־ה ֶ ּ֑מ ֶלְך‬
ַ ‫ֶא‬ κρίσιν πρὸς τὸν ὸν βασιλέα κρίσιν πρὸς τὸν βασιλέα
‫וַ יְ גַ ּנֵ ֙ב ַא ְב ָׁש ֔לֹום‬ καὶ ἰδιοποιεῖτο Αβεσσαλωμ καὶ ἰδιοποιεῖτο Αβεσσαλωμ
‫ת־ל֖ב‬ ֵ ‫ֶא‬ τὴν καρδίαν τὰς καρδίας
‫ַאנְ ֵ ׁ֥שי יִ ְׂש ָר ֵ ֽאל׃‬ ἀνδρῶν Ισραηλ παντῶν τῶν ἀνδρῶν τοῦ
Ισραηλ
‫לֹום‬
֙ ‫ וַ ּיִ ְׁש ַל֤ח ַא ְב ָׁש‬10 καὶ ἀπέστειλεν καὶ ἀπέστειλεν
‫ְמ ַרּגְ ֔ ִלים‬ Αβεσσαλωμ κατασκόπους Αβεσσαλωμ
κατασκόπους
‫ל־ׁש ְב ֵ ֥טי יִ ְׂש ָר ֵ ֖אל‬
ִ ‫ְּב ָכ‬ ἐν πάσαις φυλαῖς Ισραηλ εἰς πάσας τὰς φυλὰς τοῦ
Ισραηλ
‫מר‬ ֹ ֑ ‫ֵלא‬ λέγων λέγων
‫ְּכ ָׁש ְמ ֲע ֶכ ֙ם‬ ἐν τῷ ἀκοῦσαι ὑμᾶς ἐν τῷ ἀκοῦσαι ὑμᾶς
‫ת־קֹול ַהּׁש ֔ ָֹפר‬ ֣ ‫ֶא‬ τὴν φωνὴν τῆς κερατίνης φωνὴν σάλπιγγος
‫וַ ֲא ַמ ְר ֶּ֕תם‬ καὶ ἐρεῖτε καὶ ἐρεῖτε
‫ָמ ַ ֥לְך‬ βεβασίλευκεν βασιλεὺς βεβασίλευκεν
‫ַא ְב ָׁש ֖לֹום‬ Αβεσσαλωμ Αβεσσαλωμ
‫ּב ֶח ְב ֽרֹון׃‬. ְ ἐν Χεβρων ἐν Χεβρων
b or not b? 85
This surprising observation fits the fact that in early Judaism not only
the meaning but also the surface of the text had become important. This
can be seen in the strange phenomenon of the kaige recension that the
short form of the Hebrew personal pronoun ‫ אני‬is rendered with ἐγώ and
the long form ‫ אנכי‬with ἐγώ εἰμί, even if a finite verb follows and this
combination in Greek is not only strange but simply wrong. This differ-
ence simply indicates the form of the Hebrew pronoun, although there
is no difference in meaning. The basic idea is that every detail and every
element in the holy text has some importance, even more so if it seems
superfluous, just because it is there.43
From this we can move on to the Antiochene text. If we—for a test—
leave aside the old assumptions about Lucianic redaction and try to see
things the other way around, we arrive at two surprising conclusions
which lead to a new solution. (1) Considering the Antiochene text as a
whole and not only looking at the supposed redactional differences, this
text is a faithful rendering of the Hebrew parent text and at the same
time a text demonstrating rather good Greek. (2) The seemingly irregular
and even contradictory changes by the supposed Lucianic redaction turn
into a consistent explanation as a redaction aiming at isomorphic equiva-
lence. The changes of the article can be explained consistently: as demon-
strated above, the articles in the Antiochene text are in accordance with
the grammar of the Hebrew text, while the kaige recension has added
or deleted the articles according to its isomorphic principle. The same
is the case with the so-called explanatory words. Evidently, the transla-
tor of the Antiochene text added explanatory words or—more probably,
in view of the Qumran texts—there was a Hebrew Vorlage which was
slightly different from the MT and which already in Hebrew contained
such explanatory words. The kaige recension again adapted the Greek to
its Hebrew reference text, in this case a text more or less identical with
MT. According to the principles of the kaige recension explained above,
this again led to additions and to omissions (see the sentence with ἀνήρ in
v. 2b line 6 and παντῶν in v. 6 line 8), according to the same clear principle
as for the article. This close adaptation also leads to changes in number
(e.g. v. 6 line 7: singular τὴν καρδίαν according to ‫ )את־לב‬and addition or
43 On Early Jewish hermeneutics see e.g. C. Dohmen and G. Stemberger, Hermeneutik
der Jüdischen Bibel und des Alten Testaments (Stuttgart: Kohlhammer, 1996); and D. Patte,
Early Jewish Hermeneutic in Palestine (Missoula: Scholars Press, 1975).
adaptation of prepositions (v. 2b line 2: πρός according to ‫ ;אליו‬v. 10 line 3:

ἐν according to ‫ ב‬instead of εἰς.
Also the semantic change from σάλπιγξ to κερατίνη can be explained in
terms of this close formal adaption. While the Antiochene text/Old Greek
aimed at functional equivalence (“trumpet”), the kaige recension wants
material equivalence (“horn”). This and other things like the substitution
of the historical past have been discussed by Thackeray and Barthélemy
and go beyond the redactional principle presented here.
The close adherence to the Hebrew reference text also allows a conclu-
sion regarding its Vorlage. In v. 2 line 3, the Hebrew text evidently had ‫לא‬
(cf. αὐτῷ), and in v. 10 line 9, there must have been a second ‫מלך‬, read
as ‫ ֶמ ֶלְך‬. The reference text of the kaige recension was therefore close to,
but not fully identical with, MT. The Hebrew reference text of the Anti-
ochene text/Old Greek was only slightly different. The plus in v. 2b lines
5–6 presupposes the typical Hebrew combination ‫ויען—ויאמר‬, which was
therefore most probably in the Hebrew Vorlage.
The emphasis (the heart of ) all (men of Israel) may be an emphasis
added by the translator, but may well go back to the Hebrew text.44 On
the other hand, v. 10 line 9 βασιλεύς is missing. This also may be a change
by the translator or—in view of the general closeness to the Hebrew
text—represent a minus in the Vorlage.
3.1.2.
Basically the same situation can be seen in the kaige text of 2 Kgs
(4 Kgdms). Unfortunately, the Qumran evidence for this book is very
meagre. But this does not matter, because the argument developed above
relies on the inner relation of the texts. On the other hand, there is some
interesting evidence from the Old Latin, as the following brief synopsis
shows.
44 The closeness to the Hebrew reference text can be seen by the fact that in v. 2 the
assumed additional ‫ ויען האיש‬is confirmed by ‫ וענה האיש‬in both, 4QSama and 4QSamc
(DJD XVII, 260). The difference between ‫ ויען‬and ‫ וענה‬is regularly found in the Qumran
texts. Both forms express past tense. 4QSama, although in the reconstructed text, has ‫כול‬
as equivalent for παντῶν in v. 6 line 8. Unfortunately, 4QSama is not extant beyond the
beginning of v. 7.
b or not b? 87
2 Kgs/4 Kgdms 6:8–9

MT kaige/B (Rahlfs) Ant (Madrid Edition)
‫ּומ ֶלְך ֲא ָ ֔רם‬
֣ ֶ 8 καὶ βασιλεὺς Συρίας καὶ βασιλεὺς Συρίας
‫ָה ָי֥ה נִ ְל ָ ֖חם ְּביִ ְׂש ָר ֵ ֑אל‬ ἦν πολεμῶν ἐν Ισραηλ ἦν πολεμῶν τὸν Ἰσραήλ,
‫ל־ע ָב ָ ֣דיו‬
ֲ ‫וַ ּיִ ּוָ ַע ֙ץ ֶא‬ καὶ ἐβουλεύσατο πρὸς καὶ συνἐβουλεύσατο
‫מר‬ֹ ֗ ‫ֵלא‬ τοὺς παῖδας αὐτοῦ λέγων τοῖς παισὶν αὐτοῦ λέγων
‫ֹלני‬
֥ ִ ‫ל־מ ֛קֹום ְּפ‬ ְ ‫ֶא‬ εἰς τὸν τόπον τόνδε τινὰ Εἰς τὸν τόπον τὸν φελμουνεὶ
‫ַא ְלמ ִֹנ֖י ַּת ֲחנ ִ ֹֽתי׃‬ ελμωνι παρεμβαλῶ ποιήσωμεν ἔνεδρον,
και ἐποίησαν.
Et consilium habuit cum pueris suis: dicens: In
locum Phelminiim insidia faciamus L115 in locum
Phelmunim obsessionem faciamus L91–95
‫ וַ ּיִ ְׁש ֞ ַלח‬9 καὶ ἀπέστειλεν καὶ ἀπέστειλεν
‫ֹלהים‬ ִ֗ ‫ִ ֣איׁש ָה ֱא‬ Ελισαιε ὁ ἄνθρωπος τοῦ θεοῦ
֙‫ל־מ ֶלְך יִ ְׂש ָר ֵאל‬
֤ ֶ ‫ֶא‬ πρὸς [τὸν >B] βασιλέα πρὸς τὸν βασιλέα Ἰσραὴλ
‫מר ִה ָּׁ֕ש ֶמר ֵמ ֲע ֖בֹר‬ ֹ ֔ ‫ֵלא‬ Ισραηλ λέγων φύλαξαι μὴ λέγων Πρόσεχε τοῦ μὴ
‫ַה ָּמ ֣קֹום ַה ֶּז֑ה‬ παρελθεῖν διελθεῖν
‫י־ׁשם ֲא ָ ֥רם נְ ִח ִ ּֽתים׃‬ ֖ ָ ‫ִ ּֽכ‬ ἐν τῷ τόπῳ τούτῳ τὸν τόπον τούτον,
ὅτι ἐκεῖ Συρία κέκρυπται ὅτι ἐκεῖ Σύροι ἐνεδρεύουσιν.
Et mandavit homo dei L115
We find similar phenomena as above. In v. 8 line 2, the article gets replaced

by the preposition ἐν (Israel) according to the ‫ב‬. In v. 8 line 3, πρóς is
introduced according to the ‫ ֶאל‬in the Hebrew text. In v. 8 line 6, the
intentionally correct but free rendering ποιήσωμεν ἔνεδρον, “let us make an
ambush,” is replaced by the exact rendering παρεμβαλῶ, for ‫תחנתי‬.
In v. 9 line 3, the article is deleted because in ‫ מלך יׂשראל‬there is
no visible article. Verse 9 line 5 seems to presuppose ‫ במקום‬instead of
‫המקום‬, while the accusative and the article in the Antiochene text pre-
suppose the article of the MT. In v. 9 line 6, Σύροι is replaced by Συρία as
the exact rendering of ‫ארם‬.
In this passage the Old Latin is very interesting. The Latin expression
“com pueris suis” may represent either version. But “in locum Phelm-
iniim/Phelmunim” and “insidia/obsessionem faciamus” clearly represent
the Antiochene text. It is interesting that the expression ‫ פלני עלמני‬is
not translated but transcribed in both versions and represented with one
expression only, yet both “Phelminiim” and “insidia/ obsessionem” clearly
represent the Antiochene text. In v. 9 line 2 there is the change between
the name Ελισαιε and the title ἄνθρωπος τοῦ θεοῦ; both expressions can be
found in the context. Interestingly, the Antiochene text again agrees with
MT and it is confirmed by the “homo dei” in the Old Latin.
There are two conclusions:
(1) The Old Latin confirms that the Antiochene text is pre-Hexaplaric and
pre-Lucianic. The comparison again shows that the text is deduced
from the Antiochene/Old Greek text;
(2) Differently from the situation in 2 Sam, in 2 Kgs the Antiochene text
seems closer to the MT, while the kaige text had a slightly different
reference text. Yet the only difference which presupposes a difference
in the Hebrew text is the change between the name Elisaias and the
title “man of God” in v. 9 line 2. With regard to the Hebrew text form,
this is a minor difference because of the repetition and the inter-
change of both elements in the context. But in regard of the age of
the text, the agreement with the Old Latin is proof of the antiquity of
the Antiochene text.
In sum, it can be said that, among other aspects, the two texts show that
in the kaige sections of both 2 Sam and 2 Kgs, the kaige text is a revision
and not the original text of the Septuagint. This proves that the text of
Codex Vaticanus in these sections is secondary.
3.2. The Relation of Codex Vaticanus to the Antiochene Text in

the Non-Kaige Sections
The situation in the non-kaige sections is more difficult. In these sections
of the historical books, especially 1–4 Kgdms, the text of Codex Vaticanus is
much closer to the Old Greek. Most authors hold it to be the witness closest
to the Old Greek, and many practically equate it with the Old Greek.
On the other hand, the Antiochene text has also proven to be very close
the Old Greek. The change between the kaige sections and the non-kaige
sections is a feature of Codex Vaticanus only. The Antiochene text shows
no such difference. Therefore, it can be assumed that also in the non-kaige
sections not only are the character but also the age and the relation to the
Old Greek about the same. So there are two textual traditions which are
seemingly very close to the Old Greek. If both B and the Antiochene text
are close to the Old Greek, B and the Antiochene text must also be closer
together. Yet there are many differences also in the non-kaige sections.
What is the relationship between these two text forms and how do they
relate to the Old Greek? We will consider some examples.
b or not b? 89
3.2.1. ἡ Βααλ—The Seemingly Female Baal

Starting with Judg 2:13 until 4 Kgdms 21:3, and 2 Chron 17:2 there is an
interesting feature: the name of the Canaanite god Baal is combined with
a female article: ἡ Βααλ, τῇ Βααλ or also ταῖς Βααλιμ. The meaning of this
strange feature is most probably a kind of Ketib-Qere in the Greek: the
female article indicates that the name of this God should not be pro-
nounced, but that it should be read as ἡ αἰσχύνη etc.45 This is confirmed by
the well-known change of the name of Ishbaal/Mephibaal to Ishboschet/
Mephiboshet and by the reading αἰσχύνη in 3 Kgdms 18:19, 25. This read-
ing practice can also be found in other books, especially in the book of
Jeremiah. Yet this reading practice was given up later. The kaige recension
in the historical books has returned to the male article with Baal, as can
be seen in Judg 2:13; 3:7; 10:6, 10.
τῇ Βααλ in 3 Kgdms 19:18 and Rom 11,4

Rom 11:4 1 Kgs/3 Kgdms 19:18 1 Kgs/3 Kgdms 19:18
Ant (Madrid) (Rahlfs)
4 ἀλλὰ τί λέγει αὐτῷ ὁ
χρηματισμός; κατέλιπον 18 καὶ καταλείψω ἐξ Ισραηλ 18 καὶ καταλείψεις ἐν
ἐμαυτῷ ἑπτακισχιλίους ἑπτὰ χιλιάδας ἀνδρῶν πάντα Ισραηλ ἑπτὰ χιλιάδας
ἄνδρας, οἵτινες οὐκ τὰ γόνατα ἃ οὐκ ἔκαμψαν ἀνδρῶν πάντα γόνατα
ἔκαμψαν γόνυ γόνυ τῇ Βααλ, καὶ πᾶν ἃ οὐκ ὤκλασαν γόνυ τῷ
τῇ Βάαλ. στόμα ὃ οὐ προσεκύνησεν Βααλ καὶ πᾶν στόμα ὃ οὐ
αὐτῷ προσεκύνησεν αὐτῷ
αὐτῷ] αὐτῇ 127
There is an interesting case in 3 Kgdms 19:18. This passage is quoted in

the New Testament in Rom 11:4. Rahlfs, in his analyses of the Lucianic
text, held the assumption that agreements between the Lucianic text
and the New Testament are not witnesses to an old text, but that the
New Testament has influenced the Lucianic tradition. In this way he also
explained the identical words in the quotation in Rom 11:4 and in the Luci-
anic text of 3 Kgdms 19:18. Yet Rom 11:4 has τῇ Βααλ. This word cannot be
explained as having influenced the Lucianic text, because Rom 11:4 is the
45 A. Dillman, Über Baal mit dem weiblichen Artikel (Monatsberichte der Königlichen
preussischen Akademie der Wissenschaften, Berlin, 1881), 601–620, had already proffered
this postulate and it remains the most probable explanation.
only occurrence of Baal in the New Testament and it would be impossible

to assume that this quotation would have produced all the occurrences
of ἡ Βααλ in the Septuagint. Rather, τῇ Βααλ in Rom 11:4 proves that not
only is this expression old, but the whole quotation uses the old textual
form. This form is the Antiochene text, which in this case again represents
the Old Greek while, on the other hand, the differences in the text of
Vaticanus must go back to a revision that not only changed the reading
of the name of Baal but other words as well. Unfortunately, while Rahlfs
discussed the other differences in this verse,46 he did not mention ἡ Βααλ,
and it is also missing at 3 Kgdms 19:18 in the apparatus of the Handaus-
gabe of 1935. Anyway, the expression τῇ Βααλ in the Antiochene text of 3
Kgdms 19:18 shows that at this place the Antiochene text represents the
old text, i.e. Old Greek, while B has the revised text.47
3.2.2. The Rendering of ‫הטוב בעיניך‬

As mentioned above (in 2.3.), soon after Barthélemy’s Les Devanciers
d’Aquila with its discovery of the kaige recension and its new evaluation
of the Antiochene text, Sebastin P. Brock delivered a paper with the tell-
ing title “Lucian redivivus”.48 In this paper he accepted the kaige recen-
sion, but defended the traditional view of an extensive Lucianic revision.
One of his (rather few) examples was the rendering of the expression
‫ הטוב בעיניך‬in the kaige and in the non-kaige sections. Brock observed
that in the Lucianic text the expression was translated with τὸ ἀρεστὸν
ἐνώπιόν σου, while in Codex Vaticanus there are two renderings of it: in
the kaige section (“Palestinian text”) it is τὸ ἀγαθὸν ἐν ὀφθαλμοῖς σου, and
in the non-kaige section one finds ἀγαθόν ἐνώπιόν σου. Brock rightly said
that the character of the Lucianic text is the same in the kaige and in
the non-kaige sections. But in a strange turn of the argument he did not
discuss the differences in the Vaticanus, but postulated that Vaticanus is
the Old Greek in the non-kaige section and that therefore—and because
of its identical character—the Antiochene text must be secondary, i.e.
the result of Lucian’s revision, in the kaige section as well. Expressed as a
table, his idea is the following:
46 Rahlfs, Lucians Rezension, 251.

47 Interestingly, there is also a case where the opposite is true. In 4 Kgdms 21:3, Codex
Vaticanus has kept ἡ Βααλ, while the Antiochene text has the masculine form. This shows
that Codex Vaticanus sometimes may have kept an older expression, even within a kaige
section.
48 Brock, “Lucian redivivus”.
b or not b? 91
1 Kgdms 2 Kgdms (Pal.)

‫הטוב בעיניך‬ ‫הטוב בעיניך‬
B τὸ ἀγαθόν ἐνώπιόν σου τὸ ἀγαθὸν ἐν ὀφθαλμοῖς σου
Ant τὸ ἀρεστὸν ἐνώπιόν σου τὸ ἀρεστὸν ἐνώπιόν σου
Yet the real problem is the difference within Vaticanus. The difference
shows that there must be two levels. Barthélemy’s identification of the
kaige recension showed that the text in the kaige section belongs to
the revision and that it is the younger text, while the Antiochene text is
the older base text of that revision. If, as Brock had rightly maintained, the
character and the age of the Lucianic text is the same in both sections, it
is old in the non-kaige section as well. Barthélemy’s insights and analyses
lead to the following situation:
1 Kgdms 2 Kgdms 19:38(39)

and Pal. Throughout
Hebrew ‫הטוב בעיניך‬ ‫הטוב בעיניך‬
Ant τὸ ἀρεστὸν ἐνώπιόν σου τὸ ἀρεστὸν ἐνώπιόν σου
B (non-kaige) τὸ ἀγαθόν ἐνώπιόν . . .
B (Pal./kaige) τὸ ἀγαθὸν ἐν ὀφθαλμοῖς σου
The only question is: Is the text of Vaticanus in the non-kaige section
even older (and closer to the Old Greek) than the Antiochene text, or is it
younger, i.e. between the Antiochene tradition and the kaige recension?
As a look at the texts clearly shows, the expression τὸ ἀγαθόν ἐνώπιόν σου is
halfway between the Antiochene text and kaige. It has maintained ἐνώπιόν
σου but it has changed τὸ ἀρεστὸν into τὸ ἀγαθόν, which is formally closer
to the Hebrew. In other words, Brock’s example confirms Barthélemy’s
view: the Antiochene text preserves the oldest text and is close to the
Old Greek, while the kaige text is a later and formalistic revision of the
Greek text towards the Hebrew reference text. And, interestingly, even
the non-kaige sections in Codex Vaticanus show a—somewhat milder—
Hebraising revision.
3.2.3. Two Examples of Whole Verses

For further illustration we look at some verses in 2 Sam 4, comparing the text
of Codex Vaticanus, the Antiochene text, the MT and the text of 4QSama.
2 Sam 4:2
MT Line B Anted
‫ּוׁש ֵנ֣י ֲאנָ ִ ׁ֣שים‬ ְ 1 καὶ δύο ἄνδρες καὶ δύο ἄνδρες
‫דּודים‬֣ ִ ְ‫ָׂש ֵ ֽרי־ג‬ 2 ἡγούμενοι συστρεμμάτων ἡγούμενοι συστρεμμάτων
‫ָהי֪ ּו‬ 3 τῷ Μεμφιβόσθε τῷ Μεμφιβόσθε
‫ן־ׁש ֟אּול‬ ָ ‫ֶב‬ 4 υἱῷ Σαούλ· υἱῷ Σαούλ·
‫ֵׁש ֩ם ָה ֶא ָ֙חד ַ ּֽב ֲע ָ֜נה‬ 5 ὄνομα τῷ ἑνὶ Βαανά, ὄνομα τῷ ἑνὶ Βαναία,
‫וְ ֵ ׁ֧שם ַה ֵּׁש ִנ֣י‬ 6 καὶ ὄνομα τῷ δευτέρῳ καὶ τῷ δευτέρῳ ὄνομα
‫ֵר ָ֗כב‬ 7 Ῥηχάβ, Ῥηχάβ,
‫ְּב ֵנ֛י ִר ּ֥מֹון‬ 8 υἱοὶ Ῥεμμὼν υἱοὶ Ῥεμμὼν
‫ַה ְּב ֶ ֽאר ִ ֹ֖תי‬ 9 τοῦ Βηρωθαίου τοῦ Βηρωθαίου
‫ִמ ְּב ֵנ֣י ִבנְ יָ ִ ֑מן‬ 10 ἐκ τῶν υἱῶν Βενιαμείν, ἐκ τῶν υἱῶν Βενιαμίνd,
‫ם־ּב ֵא ֔רֹות ֵּת ָח ֵ ׁ֖שב‬ ְ ַ‫ִ ּ֚כי ּג‬ 11 ὅτι Βηρὼθ ὅτι καὶ Βηρὼθ
‫ל־ּבנְ יָ ִ ֽמן׃‬
ִ ‫ַע‬ 12 ἐλογίζετο τοῖς υἱοῖς ἐλογίζετο τοῖς υἱοῖς
13 Βενιαμείν. Βενιαμίν.
Qumran: 4QSama
‫ושני אנשים שרי גדודי]ם למפיבשת ֯בן֯ שאול ֗ש[ם הא]חד‬2[
‫בנימ[ין] ֯כי֯ גם‬
֯ ‫הבארתי מבני‬
֯ ‫[בענב ושם השני רכב] בני רמון‬
‫גת[י]ם [ויה]י֯ ו֯ שם‬
֯ ‫ו]י֯ ברחו הברתים‬3 ‫[בארות תחשב] ֯ע ֯ל בנימי[ן‬
As in all these chapters (of the non-kaige section), this verse clearly shows
the closeness of the texts of B and the Antiochene text. This demonstrates
that there were not two different translations (or “Septuagints”), but one
translation (i.e. the original Septuagint, the so-called Old Greek), which
was revised. Besides the general agreements, this is shown by the fact that
there are even agreements of B and the Antiochene text against the MT
such as Μεμφιβόσθε in line 3 and τοῖς υἱοῖς in line 12. Interestingly, the
first agreement Μεμφιβόσθε has its counterpart in the Qumran text, which
shows that it belongs to the Vorlage and not only to the translation. The
second agreement is harder to judge. Most probably it also goes back to
the Hebrew, because it can be explained more easily in Hebrew than in
Greek, as either an addition or an omission because of “homoioarkton”/
‫בני בנימין‬. Beyond that ‫ ֯ע ֯ל‬and the words before it in 4QSama are evi-
dently reconstructed according to MT.
Yet there are also small but significant differences. In line 5 the name
Βαανά is closer to MT than the Antiochene text with Βαναία, and in line 6
the word order in B agrees with MT. In both cases there is no real reason
that the Antiochene text should have changed the text represented by B.
In both cases the B text has no problem and the Antiochene text is not
really better Greek. On the other hand, both differences can be explained
b or not b? 93
as adaptations to the Hebrew text. Therefore, it is highly probable that

the text in B reflects an adaptation, which means that it has undergone
a—mild—revision towards a Hebrew reference text.
2 Kgdms 4:5
MT Line B Anted
‫ֽי־ר ּ֤מֹון‬
ִ ‫וַ ּיֵ֙ ְל ֜כּו ְּב ֵנ‬ 1 Καὶ ἐπορεύθησαν υἱοὶ Καὶ ἐπορεύθησαν οἱ υἱοὶ
‫ֹתי‬
֙ ִ ‫ַה ְּב ֵ ֽאר‬ 2 Ῥεμμὼν τοῦ Βηρωθαίου Ῥεμμὼν τοῦ Βηρωθαίου,
‫ּוב ֲע ָ֔נה‬ַ ‫ֵר ָכ֣ב‬ 3 Ῥεκχὰ καὶ Βαὰμ Ῥηχὰβ καὶ Βαναία
4 ὁ ἀδελφὸς αὐτοῦ,
‫אּו‬֙ ֹ ‫וַ ּיָ ֙ב‬ 5 καὶ εἰσῆλθον καὶ εἰσῆλθον
‫ְּכ ֣חֹם ַהּי֔ ֹום‬ 6 ἐν τῷ καύματι τῆς ἡμέρας ἐν τῷ καύματι τῆς ἡμέρας
‫ל־ּבית‬֖ ֵ ‫ֶא‬ 7 εἰς οἶκον εἰς τὸν οἶκον
‫ִ ֣איׁש ּ֑בֹ ֶׁשת‬ 8 Μεμφιβόσθε, Μεμφιβόσθε,
‫וְ ֣הּוא ׁש ֵֹ֔כב‬ 9 καὶ αὐτὸς ἐκάθευδεν καὶ αὐτὸς ἐκάθευδε
‫ֵ ֖את ִמ ְׁש ַ ּ֥כב‬ 10 ἐν τῇ κοίτῃ
‫ַ ֽה ָּצ ֳה ָ ֽריִ ם׃‬ 11 τῆς μεσημβρίας. τὸ μεσημβρινόν.
Qumran: Not extant.

In this verse, one can again observe that both Greek text forms gener-
ally agree and therefore go back to one single translation that was revised;
there are not enough differences to understand the two versions as two
independent translations.49 There are fewer differences than in the kaige
sections, yet they are clear and significant. There is the difference regard-
ing the article in lines 1 and 7, and in lines 4 and 11 there are differences
regarding addition and omission. According to the old theories about a
Lucianic redaction, Lucian would have added the articles and he would
have added the explanatory words in line 4. On the other hand, he would
have deleted the words in line 11. This was explained by Rahlfs and many
others as a trait of Lucian, i.e. that he worked irregularly and even con-
tradictorily. But this is only an assumption and does not explain why he
should have done this. Again, as in the kaige section, one should test the
other possibility. If one allows the Antiochene text to present the older
text, one arrives at a consistent explanation. Each of the differences can
be explained as an adaptation to the Hebrew reference text, which in
this case almost exactly matched the proto-MT text. The identification of
Benaia as “his brother” (line 4) is missing in the Hebrew text as well.
49 Cf. Tov, “Lucian and Proto-Lucian,” 101–113: The Antiochene text contains “the LXX”
or “a Septuagint.”
The place of Memphiboste’s rest (line 11) is an exact rendering of the

MT. The Hebrew Vorlage of the Antiochene text may have been shorter,
but it is probably only a freer rendering of the same Hebrew words by
naming the circumstance (“at noon”) of his sleeping, while B makes a
word-for-word translation. In any case, B makes an isomorphic adaptation
towards its reference text. The deletion of the articles in lines 1 and 7 fits
exactly into this picture and again demonstrates the isomorphic character
of the revision.50
The strange form of the names in line 3, both Ρεκχα and Βααμ, is a spe-
cific phenomenon of Codex Vaticanus, which in this case is only followed
by a few manuscripts (and for Βααμ also by the Ethiopian version, which
confirms a rather late date). The manuscripts show some variety and inse-
curity with these names.51 Ρεκχα could be explained by a Hebrew form
‫ רככ‬instead of ‫רכב‬. Basically, it looks like a scribal mistake, especially
since in v. 2 line 6 there is the correct form in Codex Vaticanus. More
amazing than the mistake is that it has not been corrected.
We could continue to analyse many more verses from the non-kaige
sections with similar results,52 but for the sake of space this must suffice.
The examples showed that with the text of Codex Vaticanus and the Anti-
ochene text there are two text forms which inseparably belong together
and are very close to the Old Greek. At the same time, there are many
differences, differences that are not scribal mistakes or other uninten-
tional corruptions, but which must be explained as intentional reworking
according to specific rules. The general trend of this reworking is a closer
adaptation of the Greek text towards the Hebrew reference text, mainly
in a formalistic, isomorphic way. This intention is the same as represented
in the kaige revision, i.e. bringing the text closer to its Hebrew reference
text and making it reflect the holy Hebrew text not only in its content,
but also in its form.
50 Interestingly, both text forms read the name Memphiboste and not Isboste, which
confirms that this is the Old Greek reading. The Hebrew equivalent is found in 4QSama,
cf. above v. 2, which confirms that the Vorlage of the Old Greek was not proto-MT but a
slightly different text form. Yet one should notice that in 4:4 MT also reads Mephiboshet/
Memphiboste (see also v. 7), while the Antiochene text has Memphibaal, and that in 4:8
MT again reads Ishboshet. But these are questions of the plurality of the Hebrew text
forms.
51 See the apparatus in Brooke-McLean.
52 This analysis is being undertaken in a research project at Wuppertal sponsored by
the Deutsche Forschungsgemeinschaft.
b or not b? 95
In comparing the text forms and testing the possible directions of

the change, it turns out that the text represented by Codex Vaticanus is
the one which has been revised, while the Antiochene text is closest
to the Old Greek.
As just stated, the formalistic Hebraising revision is not as strong as
in the kaige sections. One may ask if it represents an earlier stage in the
development. Considering the general development of the Septuagint
from a more liberal to a more literal translation, this seems logical. Yet
one must consider that we are dealing with a revision, not with the origi-
nal translation. The idea of greater closeness to the Hebrew in terms of
both content and structure may as well be the result of a cross-influence
from the kaige recension. Therefore, as long as we don’t have indications
of the chronology, it seems better to avoid a term such as pre-kaige and
refer to the revision by its characteristic: the text of Codex Vaticanus at
least in the non-kaige sections of the historical books represents a semi-
kaige text.
4. Conclusions: B or not B?
(1) Codex Vaticanus (B) has become the most important single manu-
script of the Septuagint as a result of historical circumstances, as it
was the basic manuscript for the diplomatic editions from around
1600 until the 20th century, and as it remains the dominant manu-
script in the critical editions. It is probably still the most important
single manuscript, yet it has its worth no longer unto itself, but only in
the context of the other manuscripts and the other textual traditions
(including the Hebrew texts—especially from Qumran—on the one
hand, and the daughter translations—especially the Old Latin, but
also the Sahidic and the Syriac translation—on the other hand). Also
the quotations, both by Josephus and in the New Testament and by
the early Christian writers, have proven to be more important than
had been accepted in the first half of the 20th century.
(2) Most important is the fact that B consists of different text types. At
least the kaige sections and the non-kaige sections as they were iden-
tified by Thackeray and confirmed and interpreted by Barthélemy
give evidence of different text types. This clearly shows that B is a
mixed codex, i.e. a codex with different types of text. Since Barthé-
lemy’s study it is clear that the kaige sections represent a revised
text that is not the Old Greek. This result has been confirmed by the
a forementioned aspects, which demonstrate the early Jewish under-

standing of scripture and the related hermeneutics as the background
of the kaige recension. Yet the non-kaige sections of Codex Vaticanus
also show evidence of a Hebraising revision and therefore are not the
Old Greek, although much closer to it. The Hebraising recension of
the non-kaige sections—although milder—has a similar hermeneuti-
cal background to the kaige sections, i.e. it intends a formalistic adap-
tation towards the authoritative Hebrew reference text. Because it is a
more moderate revision, it can be called the semi-kaige recension. This
means that in both, the kaige sections and the non-kaige sections—at
least of the historical books, but probably also other books, especially
in the prophetic books—Codex Vaticanus represents revised texts,
although revised to different degrees.
(3) The kaige sections in B demonstrate that there had been a Hebraising
revision not only in the Minor Prophets (cf. Naḥal Ḥever scroll), but
also in large parts of the Septuagint—if not in all of it. Codex Vati-
canus is therefore important not only where it is relatively close to
the Old Greek, but also—and probably even more so—where its text
is secondary, as this shows an important and highly influential phase
of the transmission of the Septuagint.
(4) B as a manuscript from the 4th century is older than the other extant
codices and older than the manuscripts of the Antiochene texts, yet
one has to keep in mind that it is not older than the text of the Anti-
ochene fathers and that it is younger than Josephus, the Old Latin, the
Sahidic translation and (although they are in Hebrew) the Qumran
texts. This is the field where the place of Codex Vaticanus has to be
defined today.
(5) So, B or not B? The answer to this question is a clear “yes” and a clear
“no”: yes, in that Codex Vaticanus is still one of the most important
manuscripts of the Septuagint; no, in that it should not be understood
in the old way, as the most important witness of the original Sep-
tuagint, the so-called Old Greek, but rather in the way that it clearly
shows different stages of the transmission of the Septuagint and that
it allows a glimpse at the factors which accompanied this transmis-
sion and the forces which shaped it.
Übersetzungstechnik und Textkritik in den
Esdrasbüchern: Hendiadyoin, Doppelübersetzungen
und Wiederholungsvariationen in 1 Esdr
Dieter Böhler SJ
1. Einführung
In seinem Sendbrief vom Dolmetschen schreibt Martin Luther: „Denn wer

dolmetzschen wil, mus grosse vorrath von worten haben, das er die wol
koenne, wenn eins an allen orten nicht lauten will“1. Was Luther anstrebt,
ist eine durch und durch zielsprachenorientierte Übersetzung. Äquivalen-
tenkonstanz ist hier kein Ideal, vielmehr soll der je im deutschen Kon-
text passende Ausdruck gefunden werden, um einen flüssigen deutschen
Text zu erreichen, der jedes Hebraisieren vermeidet. Das Gegenmodell
zu dieser Übersetzungsphilosophie verwirklichen Buber und Rosenzweig,
die nicht mit Luther wie „die můtter ym haus und der gemeine man“2
reden wollen, sondern etwa durch Äquivalentenkonstanz ausdrücklich
das hebräische Kolorit übertragen und wiedergeben wollen. Beides hat je
nach Verwendungszweck der Übersetzung seine Berechtigung.
In der Septuaginta findet sich die ganze Bandbreite zwischen den
beiden Extremen. Ja, es sind auch die Extreme selbst verwirklicht. So gibt
es bekanntermaßen von Daniel eine alte recht elegante griechische Über-
setzung, die meist als „Septuaginta“ oder „Old Greek“ bezeichnet wird. Ihr
wurde später eine Wort-für-Wort-Übersetzung an die Seite gestellt, die
in der Tradition Theodotion zugeschrieben wurde. Dasselbe gilt für die
beiden Esdrasbücher der LXX: Der älteren um 130 v.Chr. in bestem
Griechisch angefertigten Übersetzung 1 Esdr wurde später, im 1. nach-
christlichen Jahrhundert, die pedantische Interlinearübersetzung 2 Esdr
entgegengestellt. Obwohl im Fall der Esdrasbücher die literarischen Dif-
ferenzen zwischen den beiden Bucheditionen größer sind als im Falle der
Danielfassungen, handelt es sich doch weitgehend noch um zwei Über-
setzungen ein und desselben Buches. Ja es sieht geradezu so aus, als ob
1 M. Luther, „Sendbrief vom Dolmetschen,“ in D. Martin Luthers Werke (kritische

Gesamtausgabe 30/2; Weimar: Hermann Boehlaus Nachfolger, 1909), 627–646; 639.
2 Ibid., 637.
98 dieter böhler sj
jeweils ein Späterer die völlig zielsprachenorientierte Übersetzungsphilo-

sophie des alten Daniel und des 1 Esdr korrigieren oder ergänzen wollte
durch eine nun dezidiert ausgangssprachenorientierte Übertragung.
Die Durchführung zweier Extrempositionen von Übersetzungs-
philosophie am selben Text machen 1 und 2 Esdr zu einem sehr interes-
santen Studienobjekt. Die Übersetzungstechnik von 2 Esdr, die noch nicht
aquilagleich, aber dahin unterwegs ist, macht ihn für die Textkritik sehr
leicht handhabbar. Man muss nur das Griechische Wort für Wort ins Hebrä-
ische und Aramäische rückübertragen. Der späte Textzeuge zeigt sich dann
freilich als weitgehend MT-gleich und damit wenig interessant.
Das Gegenteil gilt für die ältere Version 1 Esdr. In den Textpartien,
wo diese Fassung sich mit Esr-Neh MT (und damit 2 Esdr) überlappt3,
bezeugt 1 Esdr zahllose größere und kleinere Varianten. Doch ist diese
zielsprachenorientierte Übersetzung textkritisch oft schwer auswertbar,
da eine mechanische Rückübersetzung unmöglich ist. Vielmehr muss in
jedem einzelnen Fall eingeschätzt werden, ob eine gut griechische Äqui-
valentenwahl einfach nur der Zielsprachenorientierung des Übersetzers
geschuldet ist, oder eben doch einen von MT abweichenden Vorlagetext
bezeugt.
In vielen Fällen sind sich Textkritiker nicht darüber einig, welchen Text
1 Esdr bezeugt. Es ist bekannt, dass 1 Esdr, außer im Falle bestimmter
technischer Termini, keine Äquivalentenkonstanz pflegt. In der FS für
H.-J. Fabry habe ich etwa gezeigt, dass 1 Esdr ‫ ַהּכ ֵֹהן‬als „der Priester“ aber
je nach Kontext auch als „der Hohepriester“ übersetzen kann4. Das ist bei
ihm eine Frage der Übersetzungstechnik, nicht der Vorlagenvariante.
Es bedarf aber noch vieler Einzeluntersuchungen, um zu sehen, wie
sich der Übersetzer von 1 Esdr in bestimmten Situationen verhält, um ent-
scheiden zu können, ob eine bestimmte Äquivalentenwahl eine Variante
bezeugt oder nicht. Eine solche Einzeluntersuchung will der vorliegende
Versuch bieten.
Es ist in der Fachliteratur bekannt, dass 1 Esdr bisweilen einen einfa-
chen hebräischen Ausdruck durch ein griechisches Hendiadyoin wieder-
gibt, wenn er meint, damit die Bedeutungsbreite des hebräischen Wortes
besser abzudecken. Textkritische Folgerungen darf man daraus nicht
3 Esr 110 + Neh 8:1–13a.

4 D. Böhler, „Literarischer Machtkampf. Drei Ausgaben des Esrabuches im Streit um
das wahre Israel und um die Legitimation von Herrschaft,“ in Juda und Jerusalem in der
Seleukidenzeit (FS Fabry; BBB 159; Hg. U. Dahmen und J. Schnocks; Göttingen: Vanden-
hoeck & Ruprecht, 2010), 125–145; 130–132.
übersetzungstechnik und textkritik in den esdrasbüchern 99
z iehen. Wesentlich verschieden, aber doch benachbart, ist das Phänomen

der Doppelübersetzung. Beim Hendiadyoin geht es um die Kombina-
tion von zwei mehr oder weniger synonymen Ausdrücken. Mit Doppel-
übersetzung löst unser Übersetzer bisweilen das Problem, dass er dem
semitischen Vorlagenwort zwei semantisch völlig verschiedene griechi-
sche Äquivalente zuordnen kann. Manchmal wählt er beide. Eine solche
doppelte Wiedergabe bezeugt dennoch nur ein einfaches Vorlagewort.
Hendiadyoin und Doppelübersetzung werden wir uns gleich zuwenden.
Näher untersuchen möchte ich aber das Verhalten des Übersetzers von
1 Esdr bei Wortwiederholungen. Wenn im hebräischen Text zweimal hin-
tereinander oder mit geringem Abstand dasselbe Wort steht, wird 2 Esdr
das genau so wiederholen. 1 Esdr tut das meist nicht. Wann haben wir nur
Variation in der Äquivalentenwahl, und wann liegt die Bezeugung einer
Variante vor? Das ist die Frage, der diese Untersuchung nachgeht.
2. Die Übersetzungsphilosophien von 1 und 2 Esdr im Vergleich
Zu Beginn will ich ganz knapp an einem konkreten Beispielsatz in die

beiden Übersetzungsphilosophien von 1 und 2 Esdr einführen und die
jeweilige textkritische Auswertbarkeit darstellen.
1 Esdr 5:46 2 Esdr 3:1 Esr MT 3:1

ἐνστάντος δὲ τοῦ ἑβδόμου καὶ ἔφθασεν ὁ μὴν ὁ ‫יעי‬
ִ ‫וַ ּיִ ּגַ ע ַהח ֶֹדׁש ַה ְּׁש ִב‬
μηνὸς καὶ ὄντων τῶν ἕβδομος καὶ οἱ υἱοὶ Ισραηλ ‫ּובנֵ י יִ ְׂש ָר ֵאל ֶּב ָע ִרים ס‬
ְ
υἱῶν Ισραηλ ἑκάστου ἐν ἐν πόλεσιν αὐτῶν καὶ ‫וַ ּיֵ ָא ְספּו ָה ָעם ְּכ ִאיׁש ֶא ָחד‬
τοῖς ἰδίοις συνήχθησαν συνήχθη ὁ λαὸς ὡς ἀνὴρ
ὁμοθυμαδὸν εἷς
Der hebräische Text stellt drei Elemente in Parataxe zusammen: „und

genaht hatte sich,“ „und die Kinder Israel waren“ „und sie versammelten
sich“. 2 Esdr imitiert das genau so: 1. καὶ ἔφθασεν, 2. καὶ οἱ υἱοί, 3. καὶ συνήχθη.
1 Esdr dagegen löst die Dreierparataxe auf, macht aus den ersten beiden
Elementen je einen Genitivus absolutus und ordnet sie dem letzten Ele-
ment als dem Hauptsatz unter. Dadurch verschwindet das letzte Waw, das
mittlere Waw wird zu einem καί und das erste zu einem nachgestellten δέ.
Jedes Waw ist bezeugt, jedes auf eigene Weise. Das ist wichtig: 1 Esdr ist
eine sehr genaue Übersetzung, alles andere als frei oder paraphrastisch.
Aber sie ist nicht mechanisch wie 2 Esdr.
Die im Hebräischen zwingende Anordnung Substantiv-Adjektiv ‫ַהח ֶֹדׁש‬

‫יעי‬
ִ ‫ ַה ְּׁש ִב‬wird von 2 Esdr imitiert: ὁ μὴν ὁ ἕβδομος. 1 Esdr hat öfter einmal
die im Griechischen wohlklingende Wortfolge Adjektiv-Substantiv5: τοῦ
ἑβδόμου μηνὸς.
Das hebräische ‫ ְּכ ִאיׁש ֶא ָחד‬gibt 2 Esdr, wenig überraschend, mit ὡς
ἀνὴρ εἷς wieder. Typisch für 1 Esdr ist die Wahl eines gut griechischen idio-
matischen Ausdrucks wie ὁμοθυμαδόν. Die Vorlage weicht von MT nicht
ab, aber wir hätten sie aus 1 Esdr nicht rekonstruieren können. Bei 2 Esdr
könnten wir das. Für das hebräische ‫ ֶּב ָע ִרים‬bietet 2 Esdr ἐν πόλεσιν αὐτῶν
und bezeugt damit die Variante ‫יהם‬ ֶ ‫ ְּב ָע ֵר‬, wie sie an der Parallelstelle
Neh 7:72 auch in MT steht. Diese Lesart ‫יהם‬ ֶ ‫ ְּב ָע ֵר‬liegt wohl auch dem
idiomatischen Ausdruck ἐν τοῖς ἰδίοις zu Grunde, der einfach „zu Hause“
bedeutet6 und die hebräische Vorlage höchst treffend wiedergibt.
Das ‫ ָה ָעם‬der letzten Zeile, das auch 2 Esdr mit ὁ λαός bezeugt, hat
1 Esdr aber nicht gelesen. Spurlos würde er es nicht verschwinden lassen.
Die „Kinder Israel“ bleiben Subjekt.
Das kurze Beispiel von 1 Esdr 5:46 zeigt, dass 1 Esdr eine sehr genaue
aber eben zielsprachenorientierte Übersetzung ist, deren Auswertung für
die Textkritik deutlich schwerer ist als die der Interlinearübersetzung
2 Esdr. Nach dieser allgemeinen Einführung nun zu der speziellen Frage von
Hendiadyoin, Doppelübersetzung und Variation bei Wortwiederholungen!
3. Hendiadyoin
Hendiadyoin ist die Wiedergabe eines einfachen Ausdrucks der Vorlage

durch zwei in etwa synonyme Äquivalente, die die Reichweite des Vorlage-
ausdrucks gemeinsam besser abdecken, als jeder für sich allein es könnte7.
Ein Beispiel findet sich in 1 Esdr 9:88. Esra fordert die in illegitime
5 Z. Talshir, 1 Esdras. From Origin to Translation (SCS 47; Atlanta: Scholars Press, 1999),
208; D. Böhler, „ ‚Treu und schön‛ oder nur ‚treu‛? Sprachästhetik in den Esrabüchern, in
Im Brennpunkt: Die Septuaginta, Bd. 3: Studien zur Theologie, Anthropologie, Ekklesiologie,
Eschatologie und Liturgie der Griechischen Bibel (BWANT 174; Hg. H.-J. Fabry und D. Böhler;
Stuttgart: Kohlhammer, 2007), 97–105; 101–102.
6 Vgl. Est 5:10; 6:12; Lk 18:28; Joh 19:27; Apg 21:6.
7 R. W. Klein, Studies in the Greek Texts of the Chronicler (unpubl. Diss., Harvard Uni-
versity, Cambridge, 1966), 273: „Numerous double translations and the use of hendiadys,
especially in 1 Esdras, show the strenuous efforts . . . by the translator to present the full
meaning of his Vorlage.“
8 Talshir, 1 Esdras. From Origin to Translation, 241; idem, I Esdras. A Text Critical Com-
mentary (SCS 50; Atlanta: Society of Biblical Literature, 2001), 464–465.
Mischehen verstrickten Judäer auf, Gott die Ehre zu geben und ein
Bekenntnis abzulegen. Im MT heißt das:
Esr 10:11: ‫יכם‬
ֶ ‫י־אב ֵֹת‬
ֲ ‫ֹלה‬
ֵ ‫תֹודה ַליהוָ ה ֱא‬
ָ ‫וְ ַע ָּתה ְּתנּו‬
‫תֹודה‬
ָ bedeutet als Substantiv zu ydh hif. soviel wie „loben“, „preisen“.
Daher übersetzt 2 Esdr:
καὶ νῦν δότε αἴνεσιν κυρίῳ τῷ θεῷ τῶν πατέρων ὑμῶν.
‫תֹודה‬
ָ fungiert aber auch als Substantiv zu ydh hitp. „bekennen“ wie in
Jos 7:19. Daher wählt 1 Esdr 9:8 ein Hendiadyoin als Äquivalent für den
einfachen hebräischen Begriff:
καὶ νῦν δότε ὁμολογίαν δόξαν τῷ κυρίῳ θεῷ τῶν πατέρων ἡμῶν
Mit „Bekenntnis“ und „Verherrlichung“ hat er beide Gebrauchsweisen
von ‫תֹודה‬
ָ eingefangen. Wenn wir nicht aus vielen anderen Beispielen
wüssten, dass Hendiadyoin eine Technik unseres Übersetzers ist, hätten
wir hier wegen der asyndetischen Zusammenstellung den Verdacht einer
Glosse von zweiter Hand9.
4. Doppelübersetzung
Während die Stilfigur des Hendiadyoin einen Begriff mit einem Doppel-
ausdruck aus zwei annähernd synonymen Ausdrücken wiedergibt, um ein
einziges semantisches Feld besser abzudecken, verstehe ich unter einer
Doppelübersetzung die Kombination von zwei Möglichkeiten, die seman-
tisch miteinander nichts zu tun haben. Wenn ein deutscher Übersetzer
sich beim englischen Wort „ear“ nicht entscheiden kann, ob es um das
Hörorgan geht oder eine Ähre („ear of corn“) und deswegen einfach „die
Ähre und das Ohr“ schreibt, dann ist das eine Doppelübersetzung, d.h.
zwei vollkommen verschiedene Wiedergaben eines einzigen Wortes.
In Esr 4:8, 9, 17 schreiben Rechum und Schimschai dem Perserkö-
nig einen Beschwerdebrief gegen die heimgekehrten Judäer. Rechum
trägt den Titel ‫ל־ט ֵעם‬
ְ ‫„ ְּב ֵע‬Berichterstatter,“ Schimschai den Titel ‫ָס ְפ ָרא‬
„Schreiber“ oder „Sekretär“.
9 Syndetisch und asyndetisch: ‫( ֵה ִביא‬2 Chr 36:7) λαβὼν . . . καὶ ἀπενέγκας (1 Esdr 1:39);
‫( ֵה ִביא‬2 Chr 36:18) ἀναλαβόντες ἀπήνεγκαν (1 Esdr 1:51); ‫( ְל ַה ֵּלל‬Esr 3:10) ὑμνοῦντες . . . καὶ
εὐλογοῦντες (1 Esdr 5:57); ‫( וְ ִׁש ְב ַעת יָ ֲעט ִֹהי‬Esr 7:14) καὶ τοῖς ἑπτὰ φίλοις συμβουλευταῖς (1 Esdr
8:11).
Esr 4:17: ‫ל־ט ֵעם וְ ִׁש ְמ ַׁשי ָס ְפ ָרא‬

ְ ‫ל־רחּום ְּב ֵע‬
ְ ‫ִּפ ְתגָ ָמא ְׁש ַלח ַמ ְל ָּכא ַע‬
2 Esdr scheint den Titel ‫ל־ט ֵעם‬
ְ ‫„ ְּב ֵע‬Berichterstatter“ an allen drei Stellen
als Eigennamen verstanden zu haben und transkribiert ihn daher einfach
nur. Dabei bleibt offen, ob Baaltam ein Zweitname Rechums ist oder eine
andere Person neben ihm.
καὶ ἀπέστειλεν ὁ βασιλεὺς πρὸς Ραουμ Bααλταμ καὶ Σαμσαι γραμματέα
1 Esdr dagegen bietet eine doppelte Wiedergabe ein und desselben ara-
mäischen Ausdrucks:
τότε ἀντέγραψεν ὁ βασιλεὺς Ραούμῳ τῷ γράφοντι τὰ προσπίπτοντα καὶ Βεελτεέμῳ
καὶ Σαμσαίῳ γραμματεῖ (1 Esdr 2:21)
Indem der Übersetzer zwischen den übersetzten Titel und den transkri-
bierten Eigennamen auch noch ein καί setzte, hat er aus ein und dem-
selben aramäischen Ausdruck zuerst einen Titel Rechums und dann
zusätzlich eine Person neben Rechum gemacht.
Der aramäische Ausdruck ‫ל־ט ֵעם‬ ְ ‫„ ְּב ֵע‬Berichterstatter“ fand sich bereits
in Esr 4:8 und 4:9. Der Übersetzer von 1 Esdr hatte sich im ersten Fall
für die Interpretation als Eigennamen entschieden und (καί) Βεελτέεμος
transkribiert (1 Esdr 2:15). Im zweiten Fall entschied er sich für die Über-
setzung „Berichterstatter“ und schrieb ὁ τὰ προσπίπτοντα (2:16). Im dritten
Fall Esr 4:17 kombiniert er die beiden Möglichkeiten zu einer Doppelüber-
setzung: τῷ γράφοντι τὰ προσπίπτοντα καὶ Βεελτεέμῳ (1 Esdr 2:21).10
Man würde die Doppelübersetzung an dieser Stelle einem Glossator
zuschreiben, wenn nicht der Übersetzer von 1 Esdr selbst zuvor schon
beide Möglichkeiten, freilich zunächst je für sich, eingesetzt hätte.
Noch deutlicher ist der folgende Fall. Auch hier setzt der Übersetzer
mal eine einfache Wiedergabe, mal eine Doppelübersetzung.
In Esr 5:5 werden die Tempelbauarbeiten der Juden durch einen per-
sischen Beamten auf ihre Rechtmäßigkeit hin überprüft. Der Erzähler
vermerkt, Gottes wohlmeinendes Auge habe auf den Ältesten der Juden
gelegen, so dass sie nicht behelligt wurden:
Esr 5:5: ‫הּודיֵ א‬
ָ ְ‫ל־ׂש ֵבי י‬
ָ ‫וְ ֵעין ֱא ָל ֲההֹם ֲהוָ ת ַע‬
10 Klein, Studies, 235f; Talshir, I Esdras. A Text Critical Commentary, 107 meint: „The
complex rendering of the title suits the nature of our translator.“ Sie lässt offen, ob ein
späterer Revisor in 2:16 die Transkription als Eigenname ersetzte und in 2:21 addierte oder
eine Doppelübersetzung des ursprünglichen Übersetzers vorliegt.
2 Esdr versteht unter šby/sby hier nicht aramäisch „Älteste“ (‫) ָׂש ֵבי‬, sondern
‫„ ְׁש ִבי‬Gefangenschaft“, „Exulantenschaft“ und schreibt: τὴν αἰχμαλωσίαν
Ιουδα.
Kurz darauf schreibt der persische Beamte in Esr 5:9 dem König und
berichtet, er habe jene Ältesten befragt:
‫ֱא ַדיִ ן ְׁש ֵא ְלנָ א ְל ָׂש ַבּיָ א ִא ֵּלְך‬
In diesem Fall hat 2 Esdr sich nun für die andere Alternative, die Ältesten,
entschieden
2 Esdr 5:9: τοὺς πρεσβυτέρους ἐκείνους
Der Übersetzer von 1 Esdr war in derselben Lage, was seine aramäische
Vorlage angeht. Im Fall von Esr 5:9 entscheidet auch er sich für die
Ältesten:
1 Esdr 6:10: τότε ἐπυνθανόμεθα τῶν πρεσβυτέρων τούτων
Im Fall von Esr 5:5 allerdings wählt er die Doppelübersetzung:
1 Esdr 6:5: καὶ ἔσχοσαν χάριν ἐπισκοπῆς γενομένης ἐπὶ τὴν αἰχμαλωσίαν παρὰ τοῦ
κυρίου οἱ πρεσβύτεροι τῶν Ιουδαίων
Nur nebenbei sei bemerkt, dass für hebräisches „Auge“ ein Hendiadyoin
gewählt wird: zuerst χάριν, weil eben zielsprachenorientiert „Wohlwollen“
gemeint ist, dann aber auch noch ἐπισκοπῆς, um das hebräische Bild des
Auges nicht untergehen zu lassen.
Die Verteilung der beiden alternativen Äquivalente „Exulantenschaft“
und „Älteste“ auf den komplexen Satz 1 Esdr 6:5 schließt die Zweithand
eines bloß addierenden Glossators praktisch aus.11
Talshir kommentiert ganz zutreffend: „Unlike standard double transla-
tions, the two solutions do not simply adjoin each other; rather they are
interwoven to form one complex sentence“. Auch hier setzt der Übersetzer
wie in 1 Esdr 2:21 bei Beelteem die beiden Konzepte, die er dem doppel-
deutigen aramäischen Ausdruck entnimmt, als zwei verschiedene Größen
klar nebeneinander. Oder anders ausgedrückt: Die Doppelübersetzung
„Berichterstatter und Beelteem“ in 1 Esdr 2:21 konnte schon dort kaum
einem griechischen Glossator zugewiesen werden, weil die beiden
11 Die „Ältesten“ sind verwoben mit dem Plural des Verbs („sie hatten“) und dem
Genitiv „der Juden“. Die „Exulantenschaft“ hängt an dem „auf/über“ der Vorlage. Talshir,
I Esdras. A Text Critical Commentary, 330: „I Esd presents an ingeniously combined double
translation.“
einzelnen Möglichkeiten zuvor in 2:15 und 2:16 jeweils einmal schon

herangezogen worden waren. Die sicher dem Übersetzer von 1 Esdr 6:5
zuzuschreibende Doppelübersetzung „die Exulantenschaft“ und „die
Ältesten“ verstärkt die Sicherheit, dass ein solches Vorgehen, dem Über-
setzer selbst zuzuordnen ist.
5. Doppelausdrücke
Eine Gruppe von Ausdrücken steht zwischen einfachen Begriffen, die der
Übersetzer als Hendiadyoin wiedergibt, und Wortwiederholungen, die der
Übersetzer durch eine Variation von Äquivalenten überträgt. Ich meine
hebräische (oder aramäische) Doppelausdrücke, die doch nur einen einzi-
gen Begriff meinen, wie etwa „Tag für Tag“, ‫יֹום ְּביֹום‬, im Sinne von „jeden
Tag“, „täglich“. In allen vier (oder fünf) Fällen in denen MT und die Vorlage
von 1 Esdr einen solchen Doppelausdruck verwenden, gibt 2 Esdr (ebenso
wie 2 Par) ihn durch einen entsprechenden griechischen Doppelausdruck
wieder.
Bei Joschijas Pascha in 2 Chr 35 (1 Esdr 1) waren alle Priester und Levi-
ten, die Sänger und Torwächter an ihren Plätzen. In 2 Chr 35:15 heißt es
von den Torwächtern:
2 Chr 35:15: ‫וְ ַהּׁש ֲֹע ִרים ְל ַׁש ַער וָ ָׁש ַער‬
Wort für Wort überträgt der Übersetzer von 2 Par nicht nur den Dop-
pelausdruck „an Tor und Tor“, sondern auch noch das wurzelgleiche
„Torwächter“:
2 Par 35:15: καὶ οἱ πυλωροὶ πύλης καὶ πύλης
Ganz anders der Übersetzer von 1 Esdr! Charakteristisch ist für ihn die
Wiedergabe des hebräischen Doppelausdrucks mit einem echt griechi-
schen Äquivalent, das die hebräische Wortverdoppelung vermeidet:
1 Esdr 1:15: καὶ οἱ θυρωροὶ ἐφ᾽ ἑκάστου πυλῶνος
Die hebräischen ‫ ּׁש ֲֹע ִרים‬heißen bei 1 Esdr immer θυρωροί, in 1–2 Par und
2 Esdr ausnahmslos πυλωροί. Die Torwächter haben also als Fachtermi-
nus auch in 1 Esdr ein Standardäquivalent. Und doch zeigt sich schon an
diesem Beispiel, dass die Wahl der variierenden Äquivalente oft nicht aus-
tauschbar wäre. 1 Esdr hätte nicht umgekehrt formulieren können: „Und
die Torwächter an jeder Tür“, weil man zwar die Wächter an den Türen/
Torflügeln eines Tors als Türwächter bezeichnen kann, nicht aber Tore
als Türen.
Ein ähnlicher Doppelausdruck („Tag für Tag“) findet sich in dem ara-
mäischen Text Esr 6:9:
Esr 6:9: ‫ֶל ֱהוֵ א ִמ ְתיְ ֵהב ְלהֹם יֹום ְּביֹום‬
Wenig überraschend lautet die Wort-für-Wort-Übersetzung von 2 Esdr:
ἔστω διδόμενον αὐτοῖς ἡμέραν ἐν ἡμέρᾳ
Gewählt und elegant dagegen drückt 1 Esdr denselben Gedanken aus:
1 Esdr 6:29: ἀναλίσκεσθαι καθ᾽ ἡμέραν („dass es täglich aufgewandt werde“).
Derselbe Doppelausdruck, nun aber hebräisch, steht in Esr 3:4—und zwar
gleich zweimal:
Esr 3:4: ‫ת־חג ַה ֻּסּכֹות ַּכ ָּכתּוב וְ ע ַֹלת יֹום ְּביֹום ְּב ִמ ְס ָּפר ְּכ ִמ ְׁש ַּפט‬ַ ‫וַ ּיַ ֲעׂשּו ֶא‬
‫ְּד ַבר־יֹום ְּביֹומֹו׃‬
Die griechische Übersetzung von 2 Esdr folgt wie gewohnt Wort für
Wort:
2 Esdr 3:4: καὶ ἐποίησαν τὴν ἑορτὴν τῶν σκηνῶν κατὰ τὸ γεγραμμένον καὶ
ὁλοκαυτώσεις ἡμέραν ἐν ἡμέρᾳ ἐν ἀριθμῷ ὡς ἡ κρίσις λόγον ἡμέρας ἐν ἡμέρᾳ
αὐτοῦ.
Der Übersetzer von 1 Esdr wählt, wie im vorigen Beispiel die gut griechi-
sche Wendung καθ᾽ ἡμέραν.
1 Esdr 5:50: καὶ ἠγάγοσαν τὴν τῆς σκηνοπηγίας ἑορτήν ὡς ἐπιτέτακται ἐν τῷ
νόμῳ καὶ θυσίας καθ᾽ ἡμέραν ὡς προσῆκον ἦν.
1 Esdr bezeugt mit ὡς ἐπιτέτακται ἐν τῷ νόμῳ die hebräische Wendung
‫ּתֹורה‬
ָ ‫ ַּכ ָּכתּוב ַּב‬12.
Schwerer ist aber zunächst zu sagen, ob die Tatsache, dass 1 Esdr die
Wendung „Tag für Tag“ nur einmal hat und keine Spur ihrer Wiederho-
lung verrät, bedeutet, dass seine Vorlage in diesem Punkt abwich, oder
ob der Übersetzer die neuerliche Wiederholung einfach durch Weglassen
vermieden hat. Einfaches Weglassen ist im allgemeinen nicht die Art des
12 Die Wendung kommt in 2 Chr 25:4 und Neh 10:35, 37 vor. In Verbindung mit „Gesetz“
kann ‫ כתב‬mit ἐπιτάσσειν wiedergegeben werden, so wie bei königlichen Befehlen ‫אמר‬
mit ἐπιτάσσειν übersetzt werden kann (Esr 5:15 // 1 Esdr 6:18; ebenso DanLXX 1:18; 2:2, 46;
3:19, 20).
Übersetzers von 1 Esdr. Und da er auch für ‫ ְּב ִמ ְס ָּפר‬kein Äquivalent bietet,
hat er es ziemlich sicher nicht gelesen. Denn ὡς προσῆκον ἦν gibt auf jeden
Fall (wie ὡς καθήκει in 1 Esdr 1:13 und Lev 5:10 und 9:16) ‫ ְּכ ִמ ְׁש ַּפט‬wieder.
Dann dürfte aber auch das Fehlen von ‫ ְּד ַבר־יֹום ְּביֹומֹו‬eher ein Minus in
der Vorlage widerspiegeln. Talshir kommentiert vermutlich richtig: „I Esd
reflects only ‫ ְּכ ִמ ְׁש ַּפט‬. . . ; the rest, it would seem, was not part of his Vor-
lage. It is pleonastic to a certain extent, especially ‫ ְּד ַבר־יֹום ְּביֹומֹו‬immedi-
ately following ‫“יֹום ְּביֹומֹו‬13.
Der letzte Fall eines derartigen Doppelausdrucks findet sich in Esr 10:14.
Die Mischehenangelegenheit soll durch eine Versammlung von Rich-
tern und Ältesten aus jeder Stadt beraten werden. Der hebräische Text
nennt sie:
Esr 10:14: ‫י־עיר וָ ִעיר‬
ִ ֵ‫זִ ְקנ‬
Die Nachahmung in 2 Esdr ist gewohnt sklavisch:
2 Esdr 10:14: πρεσβύτεροι πόλεως καὶ πόλεως
Einen besseren griechischen Ausdruck findet 1 Esdr 9:13:
1 Esdr 9:13: ἑκάστου δὲ τόπου τοὺς πρεσβυτέρους
Der spezielle Fall der eben behandelten hebräischen und aramäischen
Doppelausdrücke zeigt zunächst nur, dass 2 Esdr auch hier die sklavisch
stereotype Wiedergabe verfolgt, wo 1 Esdr keine Äquivalentenkonstanz
einhält. Und es deutet sich an, dass 1 Esdr die Wiederholung desselben
Worts, die die Vorlage ihm vorgibt, lieber vermeidet14. Manchmal tut er
dies nur aus stilistischen Gründen. Oft aber kann man auch sachliche, ja
theologische Überlegungen hinter seiner Wahl erkennen.
6. Wiederholungsvariationen
Mit der Zerstörung Jerusalems durch den babylonischen König Nebukad-

nezzar erfüllen sich nach 2 Chr 36:21 (1 Esdr 1:54f. Vorlage) die Drohungen
des Propheten Jeremia. Das verwüstete Land sollte Brache halten, bis 70
Jahre voll wären. Dieses doppelte Erfüllen der Prophezeiung Jeremias und
13 Talshir, I Esdras. A Text Critical Commentary, 300.

14 Er unterliegt hier aber keinem Zwang. Vgl. Esr 9:8f: ‫ת־לנּו ִמ ְחיָ ה‬
ָ ‫ ָל ֶת‬. . . ‫ּול ִת ֵּתנּו ִמ ְחיָ ה‬
ְ ;
δοῦναι ἡμῖν τροφὴν . . . δοῦναι ἡμῖν τροφὴν (1 Esdr 8:76, 78).
der 70 Brachjahre heißt im hebräischen Vorlagetext von 1 Esdr ebenso wie

in 2 Chr 36:21 zweimal ‫ ְל ַמּלֹאות‬:
2 Chr 36:21: ‫ ְ ל ַמּלֹאות ִׁש ְב ִעים ָׁשנָ ה‬. . . ‫ְל ַמּלֹאות ְּד ַבר־יְ הוָ ה ְּב ִפי יִ ְר ְמיָ הוּ‬
Der griechische Übersetzer von 1 Esdr achtet darauf, dass er die Wortwie-
derholung durch Variation in der Äquivalentenwahl wiedergibt:
1 Esdr 1:54–55: εἰς ἀναπλήρωσιν τοῦ ῥήματος τοῦ κυρίου ἐν στόματι Ιερεμιου
55 . . . εἰς συμπλήρωσιν ἐτῶν ἑβδομήκοντα
In diesem Fall wären beide Äquivalente sowohl für übertragenes wie
numerisches Vollmachen geeignet gewesen. Wichtig war dem Übersetzer
die Variation.
Da die griechische Fassung von 2 Chr 36:21 (2 Par) dieselbe Variation
vornimmt wie 1 Esdr15, haben wir hier einen ersten Hinweis darauf, dass
der Übersetzer von 2 Par und der von 2 Esdr nicht derselbe sind, da sich
in diesem Punkt ihre Übersetzungstechnik merklich unterscheidet.
Unmittelbar nach ihrer Rückkehr aus dem Babylonischen Exil machen
sich die Heimkehrer unter Serubbabel daran, den Altar in Jerusalem wie-
derzuerrichten, um auf ihm Brandopfer darzubringen (Esr 3:2: ‫ ;עֹלֹות‬2 Esdr
3:2 und 1 Esdr 5:48: ὁλοκαυτώσεις). 1 Esdr kann hebräisches ‫ ע ָֹלה‬sowohl
mit ὁλοκαύτωσις (ὁλοκαύτωμα) als auch mit θυσία wiedergeben. In Esr 3:3,
wo zwei Arten von Brandopfern aufgezählt werden, wählt er nacheinan-
der die beiden Äquivalente. Er setzt sie aber wohlüberlegt:
Esr 3:3 ‫עֹלֹות ַליהוָ ה עֹלֹות ַלּב ֶֹקר וְ ָל ָע ֶרב׃‬
1 Esdr 5:49: θυσίας κατὰ τὸν καιρὸν καὶ ὁλοκαυτώματα τῷ κυρίῳ τὸ πρωινὸν καὶ
τὸ δειλινόν
MT spricht ebenso wie die Vorlage von 1 Esdr zweimal16 von ‫עֹלֹות‬, zuerst
generell von Brandopfern für den Herrn (MT) oder Brandopfern zu den
verschiedenen besonderen Terminen und Gelegenheiten (Sabbaten,
15 2 Par 36:21: τοῦ πληρωθῆναι λόγον κυρίου διὰ στόματος Ιερεμιου ἕως τοῦ προσδέξασθαι
τὴν γῆν τὰ σάββατα αὐτῆς σαββατίσαι πάσας τὰς ἡμέρας τῆς ἐρημώσεως αὐτῆς ἐσαββάτισεν
εἰς συμπλήρωσιν ἐτῶν ἑβδομήκοντα.
16 2 Esdr und die Vg bezeugen einen Text, in dem nur einmal vom Brandopfer die Rede
war und zwar im Singular: καὶ ἀνέβη ἐπ᾽ αὐτὸ ὁλοκαύτωσις τῷ κυρίῳ τὸ πρωὶ καὶ εἰς ἑσπέραν;
holocaustum Domino mane et vespere. Demgegenüber bezeugen 1 Esdr und Esr MT bei
allen kleineren Differenzen, die auch zwischen ihnen bestehen, doch gemeinsam einen
Text, der zweimal von Brandopfern sprach—und zwar je im Plural. Kleins Auffassung,
1 Esdr weise hier eine Doppelübersetzung für nur einmaliges „Brandopfer“ seiner Vor-
lage auf, scheitert am Text von MT (Klein, Studies, 36 und 233; dagegen zu Recht Talshir,
I Esdras. A Text Critical Commentary, 299).
Neumonden, Festen, private Opfer)17 und dann von den täglichen Brand-
opfern, dem morgendlichen und dem abendlichen. MT nennt sie beide
‫עֹלֹות‬. 1 Esdr will seine beiden Äquivalente ὁλοκαύτωμα und θυσία zum
Einsatz bringen, verteilt sie aber so, dass er nur die täglichen Brandopfer
ὁλοκαυτώματα nennt, da sie wirklich immer Brandopfer sind, während er
die ersteren, die—etwa als private Opfer—auch ein Schlachtopfer oder
eine Mincha sein könnten, als θυσίας bezeichnet. Die beiden ihm sonst
gleich gültigen Äquivalente werden hier offenbar gezielt eingesetzt und
wären nicht einfach austauschbar.
Die Fortsetzung in Esr 3:4f. nennt die Opfer des Laubhüttenfestes, das
zweimal tägliche Tamid-Opfer und die anderen Festopfer unterschiedslos
‫( עֹלֹות‬2 Esdr: ὁλοκαυτώσεις). 1 Esdr differenziert in diesem Falle—offen-
bar mit leicht abweichender Vorlage („Sabbatopfer“)—so, dass die Opfer
des Laubhüttenfestes (nach Num 29:12–38 Brand- mit Speiseopfer, sowie
Sündopfer) mit dem allgemeinen Begriff θυσία, die täglichen Tamid-Opfer
diesmal προσφορά und die Festopfer wiederum mit θυσία wiedergegeben
werden18.
Esr 3:4–5 1 Esdr 5:50–51 2 Esdr 3:4–5

‫ת־חג‬
ַ ‫ וַ ּיַ ֲעׂשּו ֶא‬4 50 καὶ ἠγάγοσαν τὴν τῆς καὶ ἐποίησαν τὴν ἑορτὴν τῶν
‫ַה ֻּסּכֹות ַּכ ָּכתּוב וְ ע ַֹלת‬ σκηνοπηγίας ἑορτήν ὡς σκηνῶν κατὰ τὸ γεγραμμένον
‫יֹום ְּביֹום ְּב ִמ ְס ָּפר‬ ἐπιτέτακται ἐν τῷ νόμῳ καὶ ὁλοκαυτώσεις ἡμέραν ἐν
‫ְּכ ִמ ְׁש ַּפט ְּד ַבר־יֹום‬ καὶ θυσίας καθ᾽ ἡμέραν ὡς ἡμέρᾳ ἐν ἀριθμῷ ὡς ἡ κρίσις
‫ְּביֹומֹו׃‬ προσῆκον ἦν λόγον ἡμέρας ἐν ἡμέρᾳ αὐτοῦ
‫יכן ע ַֹלת ָּת ִמיד‬ ֵ ‫ וְ ַא ֲח ֵר‬5 51 καὶ μετὰ ταῦτα προσφορὰς 5 καὶ μετὰ τοῦτο ὁλοκαυτώσεις
‫וְ ֶל ֳח ָד ִׁשים‬ ἐνδελεχισμοῦ ἐνδελεχισμοῦ
‫ל־מֹוע ֵדי יְ הוָ ה‬
ֲ ‫ּול ָכ‬
ְ καὶ θυσίας σαββάτων καὶ εἰς τὰς νουμηνίας καὶ εἰς
‫ַה ְמ ֻק ָּד ִׁשים‬ καὶ νουμηνιῶν πάσας ἑορτὰς τὰς ἡγιασμένας
καὶ ἑορτῶν πασῶν ἡγιασμένων
17 Ähnliche Aufzählungen sind in der Chronik nicht selten: 1 Chr 16:40: ‫ְל ַה ֲעלֹות עֹלֹות‬
‫ל־מזְ ַּבח ָהע ָֹלה ָּת ִמיד ַלּב ֶֹקר וְ ָל ָע ֶרב‬ ִ ‫ ; ַליהוָ ה ַע‬2 Chr 2:3: ‫וְ עֹלֹות ַלּב ֶֹקר וְ ָל ֶע ֶרב ַל ַּׁש ָּבתֹות‬
‫מֹוע ֵדי יְ הוָ ה‬
ֲ ‫ּול‬ְ ‫ ;וְ ֶל ֳח ָד ִׁשים‬1 Chr 23:31: ‫ ;עֹלֹות ַליהוָ ה ַל ַּׁש ָּבתֹות ֶל ֳח ָד ִׁשים וְ ַלּמ ֲֹע ִדים‬2 Chr 31:3:
‫ ָלעֹלֹות ְלעֹלֹות ַהּב ֶֹקר וְ ָה ֶע ֶרב וְ ָהעֹלֹות ַל ַּׁש ָּבתֹות וְ ֶל ֳח ָד ִׁשים וְ ַלּמ ֲֹע ִדים‬. Talshir, I Esdras. A
Text Critical Commentary, 299, vermutet mit Blick auf diese Reihungen der Chronik, 1 Esdr
5:49 habe für καιρός ‫ ַלּמ ֲֹע ִדים‬gelesen. In 1:17 steht καιρός für ‫( ֵעת‬2 Chr 35:17). In 1 Esdr 8:76
hat die Vorlage wohl ebenfalls so gelautet (‫ ;) ָּב ֵעת‬MT: ‫ ; ְמ ַעט‬vgl. Talshir, I Esdras. A Text
Critical Commentary, 449).
18 Die Vokabel προσφορά kommt in 1 Esdr nur hier vor. Häufiger steht sie in Sirach (14:11;
34:18, 19; 35:1, 5; 38:11; 46:16; 50:13, 14), da aber nie für eine spezifische Opferart wie Brand-,
Schlacht-, Speiseopfer.
Nach der vom Statthalter Tattenai veranlassten Untersuchung über die

Legitimität der Jerusalemer Tempelbauarbeiten unter Serubbabel schreibt
König Darius in seiner Antwort, die Arbeiten seien zu fördern und wer
dem Reskript zuwiderhandle, dem „reiße man einen Balken aus seinem
Haus und pfähle ihn auf diesem Balken; sein Haus soll wegen seines Ver-
gehens zu einem Trümmerhaufen gemacht werden.“
Esr 6:11: ‫ּוביְ ֵתּה נְ וָ לּו יִ ְת ֲע ֵבד‬
ַ ‫ֹלהי‬
ִ ‫ן־ּביְ ֵתּה ּוזְ ִקיף יִ ְת ְמ ֵחא ֲע‬
ַ ‫יִ ְתנְ ַסח ָאע ִמ‬
‫ל־ּדנָ ה‬
ְ ‫ַע‬
1 Esdr 6:31 λημφθῆναι ξύλον ἐκ τῶν ἰδίων αὐτοῦ καὶ ἐπὶ τούτου κρεμασθῆναι καὶ
τὰ ὑπάρχοντα αὐτοῦ εἶναι βασιλικά
Für das zweimalige „sein Haus“ des aramäischen Textes schwingt sich
selbst 2 Esdr hier zu einer Variation auf, die für ihn an die Grenze des
Möglichen geht. Er schreibt zuerst οἰκία und dann οἶκος. Interessanter ist
1 Esdr. Er wählt zuerst, wo es buchstäblich um das Eigenheim des Täters
geht, τὰ ἴδια αὐτοῦ (so auch Est 5:10 und 6:12)19, während gleich darauf
„sein Haus“ den Besitz im allgemeinen meint und daher mit τὰ ὑπάρχοντα
αὐτοῦ wiedergegeben wird20. Beide Äquivalente sind exakt für ihren Kon-
text gewählt. Hätte man sich möglicherweise für das Eigentum im all-
gemeinen τὰ ἴδια vorstellen können, so doch nicht τὰ ὑπάρχοντα für das
Eigenheim.
In Esr 6:8–9 befiehlt Darius den Behörden, mit den Juden beim Jeru-
salemer Tempelbau zusammenzuarbeiten, sie bei den Kosten für die
Bauarbeiten und den anschließenden Unterhalt des Kultbetriebs aus der
königlichen Kasse zu unterstützen.
19 W. Bauer, Wörterbuch zum Neuen Testament (Hg. K. Aland und B. Aland; Berlin /
New York: De Gruyter 61988), 752.
20 Eine bemerkenswerte Berührung zwischen 1 Esdr (hier auch 2 Esdr) und DanLXX
zeigt sich bei der Übersetzung für das aramäische Wort ‫נְ וָ לּו‬/‫נְ וָ ִלי‬. 1 Esdr 6:31 und 2 Esdr
6:11 interpretieren die zweite Maßnahme gegen das „Haus“ des Übertreters als „Verstaatli-
chung“ (2 Esdr 6:11: καὶ ὁ οἶκος αὐτοῦ τὸ κατ᾽ ἐμὲ ποιηθήσεται). In den beiden Danielparalle-
len (Dan 2:5: ‫ּוב ֵּתיכֹון נְ וָ ִלי יִ ְּת ָׂשמּון‬
ָ und Dan 3:29: ‫ּוביְ ֵתּה נְ וָ ִלי יִ ְׁש ַּתּוֵ ה‬
ַ ) interpretiert DanLXX
ebenfalls „Verstaatlichen“ (2:5: καὶ ἀναληφθήσεται ὑμῶν τὰ ὑπάρχοντα εἰς τὸ βασιλικόν. 3:96
[MT: 3:29]: καὶ ἡ οἰκία αὐτοῦ δημευθήσεται), DanTh dagegen—wie die heutigen Überset-
zungen—„Zerstörung“ (2:5: καὶ οἱ οἶκοι ὑμῶν διαρπαγήσονται. 3:96: καὶ οἱ οἶκοι αὐτῶν εἰς
διαρπαγήν).
Esr 6:8–9 1 Esdr 6:28–29 2 Esdr 6:8–9

‫ּומּנִ ְכ ֵסי ַמ ְל ָּכא ִּדי ִמ ַּדת‬
ִ 8 28 καὶ ἀπὸ τῆς φορολογίας 8 καὶ ἀπὸ ὑπαρχόντων
‫ֲע ַבר נַ ֲה ָרה ָא ְס ַּפ ְרנָ א‬ Κοίλης Συρίας καὶ βασιλέως τῶν φόρων πέραν
‫נִ ְפ ְק ָתא ֶּת ֱהוֵ א ִמ ְתיַ ֲה ָבא‬ Φοινίκης ἐπιμελῶς τοῦ ποταμοῦ ἐπιμελῶς
‫י־לא‬ָ ‫ְלגֻ ְב ַרּיָ א ִא ֵּלְך ִּד‬ σύνταξιν δίδοσθαι τούτοις δαπάνη ἔστω διδομένη τοῖς
‫ְל ַב ָּט ָלא׃‬ τοῖς ἀνθρώποις εἰς θυσίας ἀνδράσιν ἐκείνοις τὸ μὴ
τῷ κυρίῳ Ζοροβαβελ καταργηθῆναι
‫תֹורין‬
ִ ‫ּובנֵ י‬ ְ ‫ּומה ַח ְׁש ָחן‬ ָ 9 ἐπάρχῳ εἰς ταύρους καὶ
‫וְ ִד ְכ ִרין וְ ִא ְּמ ִרין ַל ֲע ָלוָ ן‬ κριοὺς καὶ ἄρνας 29 ὁμοίως 9 καὶ ὃ ἂν ὑστέρημα καὶ
‫ֶל ֱא ָלּה ְׁש ַמּיָ א ִחנְ ִטין‬ δὲ καὶ πυρὸν καὶ ἅλα καὶ υἱοὺς βοῶν καὶ κριῶν καὶ
‫ּומ ַׁשח‬
ְ ‫ְמ ַלח ֲח ַמר‬ οἶνον καὶ ἔλαιον ἐνδελεχῶς ἀμνοὺς εἰς ὁλοκαυτώσεις
‫אמר ָּכ ֲהנַ ּיָ א ִדי־‬ ַ ‫ְּכ ֵמ‬ κατ᾽ ἐνιαυτόν καθὼς ἂν τῷ θεῷ τοῦ οὐρανοῦ πυρούς
‫ירּוׁש ֶלם ֶל ֱהוֵ א ִמ ְתיְ ֵהב‬ ְ ‫ִב‬ οἱ ἱερεῖς οἱ ἐν Ιερουσαλημ ἅλας οἶνον ἔλαιον κατὰ τὸ
‫י־לא‬ ָ ‫ְלהֹם יֹום ְּביֹום ִּד‬ ὑπαγορεύσωσιν ῥῆμα τῶν ἱερέων τῶν ἐν
‫ָׁשלּו׃‬ ἀναλίσκεσθαι καθ᾽ ἡμέραν Ιερουσαλημ ἔστω διδόμενον
ἀναμφισβητήτως αὐτοῖς ἡμέραν ἐν ἡμέρᾳ ὃ
ἐὰν αἰτήσωσιν
Am Ende von 1 Esdr 6:28 findet sich mit εἰς θυσίας τῷ κυρίῳ Ζοροβαβελ
ἐπάρχῳ ein textkritisch, ja eigentlich literarkritisch relevanter Unterschied
zu Esr MT/2 Esdr, der mit dem Einschub der Pagenerzählung 1 Esdr 3–4,
einer Serubbabellegende, zusammenhängt. Diese echte Textdifferenz soll
uns hier nicht interessieren. Auch ἐνδελεχῶς κατ᾽ ἐνιαυτόν (1 Esdr 6:29)
spiegelt eine echte Vorlagedifferenz wider21. Uns beschäftigt hier nur
das Verhalten des Übersetzers von 1 Esdr. Das Partizip hitpe’el von yhb
in Esr 6:8 und 9 wird von 2 Esdr zweimal gleich mit ἔστω διδομένη/ἔστω
διδόμενον wiedergegeben. 1 Esdr wählt im ersten Falle δίδοσθαι, im zweiten
aber ἀναλίσκεσθαι22. In MT und 2 Esdr bezieht sich das erste Geben auf
einen Beitrag zum Tempelbau, das zweite auf eine Unterstützung zum
anschließenden Kultunterhalt. In der Vorlage von 1 Esdr bezieht sich das
erste Geben auf eine Unterstützung des Tempelbetriebs, das zweite aber
auf „Ausgaben“, die zu tätigen sind. Völlig passend und mit Überlegung
ausgewählt sind die beiden Äquivalente δίδοσθαι und ἀναλίσκεσθαι („auf-
wenden“). Sie könnten nicht ausgetauscht werden.
Diesem verwandt ist der Fall von Esr 7:20. Hier steht in einem ebenfalls
aramäischen Dekret des Perserkönigs zweimal ntn für „ausgeben“:
21 Talshir, I Esdras. A Text Critical Commentary, 361 und 363.

22 Ibid., 361 setzt in beiden Fällen ‫ יְ ַהב‬hitp. wie in MT voraus.
Esr 7:20: ‫ן־ּבית ּגִ נְ זֵ י‬

ֵ ‫ל־לְך ְל ִמנְ ַּתן ִּתנְ ֵּתן ִמ‬
ָ ‫ּוׁש ָאר ַח ְׁשחּות ֵּבית ֱא ָל ָהְך ִּדי יִ ֶּפ‬ ְ
‫ַמ ְל ָּכא׃‬
2 Esdr schreibt dafür zweifaches δίδωμι:
2 Esdr 7:20: καὶ κατάλοιπον χρείας οἴκου θεοῦ σου ὃ ἂν φανῇ σοι δοῦναι δώσεις
ἀπὸ οἴκων γάζης βασιλέως
Hier nun scheint der Übersetzer von 1 Esdr den Infinitiv ‫ ְל ִמנְ ַּתן‬nicht gelesen
zu haben, denn er schreibt nur ὅσα ἂν ὑποπίπτῃ σοι (für ‫ל־לְך‬ ָ ‫) ִּדי יִ ֶּפ‬. Wahr-
scheinlicher aber ist, dass er das wiederholte „geben“ vermeidet, indem er
es in einem anderen Ausdruck aufgehen lässt23:
1 Esdr 8:18: καὶ τὰ λοιπά ὅσα ἂν ὑποπίπτῃ σοι εἰς τὴν χρείαν τοῦ ἱεροῦ τοῦ θεοῦ
σου δώσεις ἐκ τοῦ βασιλικοῦ γαζοφυλακίου
Talshir kommentiert ganz richtig: „The clause is reformulated in I Esd:
(1) the word order changes; (2) ‫ ְל ִמנְ ַּתן‬is not represented directly, probably
because of the following ‫ ִּתנְ ֵּתן‬.“ Wenn 1 Esdr hier einen MT-gleichen Vor-
lagetext hatte24, zeigt die Stelle, dass unser Übersetzer Wiederholungen
zwar nicht durch bloßes Weglassen, wohl aber durch Aufgehenlassen in
einem umfassenderen Ausdruck vermeiden kann. Hier hätte er dann ὅσα
ἂν ὑποπίπτῃ σοι deswegen vor εἰς τὴν χρείαν gestellt (anders als in MT),
weil er in letzterem den (Ausgabe-) Bedarf sah, womit in seinen Augen
das erste „ausgeben“ untergebracht wäre. Das erklärt dann auch die
Textumstellung.
In seinem aramäischen Firman verfügt Artaxerxes, Esra und die Juden
dürften etwaige Summen, die nach den Ausgaben für den Kultbetrieb
noch übrigblieben, nach dem Gutdünken ihres Gottes verwenden. In Esr
7:18 heißt es:
‫יטב ִּב ְׁש ָאר ַּכ ְס ָּפא וְ ַד ֲה ָבה ְל ֶמ ְע ַּבד ִּכ ְרעּות ֱא ָל ֲהכֹם‬
ַ ִ‫ל־א ָחיְך י‬
ֶ ‫ּומה ִדי ֲע ָליְך וְ ַע‬ ָ
‫ַּת ַע ְבדּון׃‬
Dem entspricht die buchstäbliche Übersetzung von 2 Esdr:
2 Esdr 7:18: καὶ εἴ τι ἐπὶ σὲ καὶ τοὺς ἀδελφούς σου ἀγαθυνθῇ ἐν καταλοίπῳ τοῦ
ἀργυρίου καὶ τοῦ χρυσίου ποιῆσαι ὡς ἀρεστὸν τῷ θεῷ ὑμῶν ποιήσατε
23 Ibid., 404.
24 Wenn 1 Esdr einen anderen Vorlagetext hatte, zeigt die Stelle in Esr 7:20 nur, dass
aramäisch ntn und yhb (Esr 6:8f ) gleichermaßen für (Geld) „ausgeben“ stehen können.
1 Esdr vermeidet die Wiederholung desselben Äquivalents und wählt zwei

verschiedene:
1 Esdr 8:16: καὶ πάντα ὅσα ἂν βούλῃ μετὰ τῶν ἀδελφῶν σου ποιῆσαι χρυσίῳ καὶ
ἀργυρίῳ ἐπιτέλει κατὰ τὸ θέλημα τοῦ θεοῦ σου
Der Übersetzer hat sich dabei genau überlegt, wann er das allgemeine
„machen“ und wann das spezifischere „ausführen“ setzt. Die Wahl der
beiden Äquivalente für aramäisches ‘bd wäre nicht austauschbar.
In Esr 7:6 wird mitgeteilt, dass der Priester und Schriftgelehrte Esra
einerseits kundig gewesen sei im Gesetz, das der Herr Israel gegeben hat,
dass andererseits der Perserkönig dem Priester alles gab, was er wünschte.
Beides, das Geben des Gesetzes durch Gott und die königlichen Gaben,
werden im hebräischen Text durch ntn ausgedrückt.
Esr 7:6: ‫ֹלהי יִ ְׂש ָר ֵאל וַ ּיִ ֶּתן־לֹו ַה ֶּמ ֶלְך ְּכיַ ד־‬
ֵ ‫תֹורת מ ֶֹׁשה ֲא ֶׁשר־נָ ַתן יְ הוָ ה ֱא‬
ַ ‫ְּב‬
‫ֹלהיו ָע ָליו ּכֹל ַּב ָּק ָׁשתֹו‬
ָ ‫יְ הוָ ה ֱא‬
2 Esdr schreibt zweimal δίδωμι und bildet den hebräischen Text Wort für
Wort ab.
2 Esdr 7:6: ἐν νόμῳ Μωυσῆ ὃν ἔδωκεν κύριος ὁ θεὸς Ισραηλ καὶ ἔδωκεν αὐτῷ ὁ
βασιλεύς ὅτι χεὶρ κυρίου θεοῦ αὐτοῦ ἐπ᾽ αὐτὸν ἐν πᾶσιν οἷς ἐζήτει αὐτός
1 Esdr muss einen leicht veränderten Vorlagetext gehabt haben, da bei
ihm die göttliche durch die königliche Gunst ersetzt ist25. Der Punkt aber,
auf den es uns hier ankommt, ist: Während 1 Esdr das königliche Geben,
das ja nur einem Manne gilt, ebenfalls mit schlichtem δίδωμι wiedergibt,
wählt er für das göttliche Geben der Tora, die ein Buch für die Öffent-
lichkeit ist, das ebenso simple wie ausnehmend passende Kompositum
ἐκδίδωμι („herausgeben“, „publizieren“):
1 Esdr 8:3–4: ἐν τῷ Μωυσέως νόμῳ τῷ ἐκδεδομένῳ ὑπὸ τοῦ θεοῦ τοῦ Ισραηλ
4καὶ ἔδωκεν αὐτῷ ὁ βασιλεὺς δόξαν εὑρόντος χάριν ἐναντίον αὐτοῦ ἐπὶ πάντα τὰ
ἀξιώματα αὐτοῦ
Der Firman des Perserkönigs Artaxerxes, mit dem Esra seinen Auftrag
erhält, endet in Esr 7:26 mit einer Strafandrohung an alle, die sich an das
25 Talshir (I Esdras. A Text Critical Commentary, 389f.) nimmt keine Vorlagedifferenz

an, sondern schreibt dem Übersetzer eine für ihn untypische Abweichung zu. Sie verweist
auf 1 Esdr 6:9 (Esr 5:8) für den richtigen Hinweis, dass δόξα für ‫ יָ ד‬zu stehen pflegt. Das
erklärt aber noch nicht das Fehlen von „des Herrn, seines Gottes“.
Gesetz Gottes und das Gesetz des Königs nicht halten wollten. Der aramä-
ische Ausdruck für das Gesetz ist in beiden Fällen ‫ ָּדת‬.
Esr 7:26: ‫י־א ָל ָהְך וְ ָד ָתא ִּדי ַמ ְל ָּכא‬
ֱ ‫ָּד ָתא ִד‬
2 Esdr übersetzt mechanisch zweimal mit νόμος.
2 Esdr 7:26: νόμον τοῦ θεοῦ καὶ νόμον τοῦ βασιλέως
1 Esdr gibt das zweimalige aramäische ‫ ָּדת‬nicht mit zweimal demselben
griechischen Ausdruck wieder, sondern schreibt:
1 Esdr 8:24: τὸν νόμον τοῦ θεοῦ σου καὶ τὸν βασιλικόν
Klein notiert die Differenz unter „stylistic omissions“26. Talshir scheint
mir die Wahl des griechischen Übersetzers an dieser Stelle treffender zu
beschreiben, wenn sie anmerkt: „he makes a fine distinction between the
divine law . . . and the royal one for which he chooses to condense the
Aramaic phrase“27. Denn es ist völlig klar, dass er zweimal ‫ ָּדת‬gelesen
hat. Auch 1 Esdr redet nicht einfach von einem Gesetz Gottes und des
Königs (‫י־א ָל ָהְך וְ ִּדי ַמ ְל ָּכא‬
ֱ ‫) ָּד ָתא ִד‬, sondern von zweien, nennt aber nur
das erstere νόμος, das letztere τὸν βασιλικόν, ohne noch einmal dasselbe
Wort zu wiederholen, das das göttliche Gesetz bezeichnet hatte. Er inte-
griert das zweite ‫ ָּד ָתא‬in einen neu formulierten Ausdruck und lässt es
nicht einfach weg. Hätte 2 Esdr das zweite νόμος unterdrückt, müssten wir
uns fragen, ob seine aramäische Vorlage nur einmal ‫ ָּדת‬gelesen hat. Wir
hätten ein textkritisches Problem. Bei 1 Esdr aber zeigt sich kein textkriti-
sches Problem, sondern eine Übersetzungsphilosophie.
Ebenfalls im Firman des Artaxerxes findet sich die Vorschrift, dass die
Aufwendungen für den Jerusalemer Tempelgottesdienst aus der Kasse des
Königs beglichen werden. „Haus deines Gottes“ und „Haus der Schätze des
Königs“ stehen im aramäischen Text dicht bei einander:
Esr 7:20: ‫ן־ּבית ּגִ נְ זֵ י ַמ ְל ָּכא׃‬
ֵ ‫ל־לְך ְל ִמנְ ַּתן ִּתנְ ֵּתן ִמ‬
ָ ‫ֵּבית ֱא ָל ָהְך ִּדי יִ ֶּפ‬
2 Esdr setzt für ‫ ֵּבית‬zweimal unbeirrt οἶκος. In 2 Esdr 7:20 heißt es:
οἴκου θεοῦ σου ὃ ἂν φανῇ σοι δοῦναι δώσεις ἀπὸ οἴκων γάζης βασιλέως
Die Pluralform „Schatzhäuser“ muss textkritisch nichts bedeuten. Sie kann
ein Versuch sein, das aramäische „Schätzehaus“ wiederzugeben. 1 Esdr
26 Klein, Studies, 164.

27 Talshir, I Esdras. A Text Critical Commentary, 408f.
wiederholt nicht οἶκος, ja wählt überhaupt zwei völlig verschiedene Äqui-

valente, „Heiligtum Gottes“ für „Gotteshaus“ und „königliche Schatzkam-
mer“ für „Haus der Schätze des Königs“:
1 Esdr 8:18: τοῦ ἱεροῦ τοῦ θεοῦ σου δώσεις ἐκ τοῦ βασιλικοῦ γαζοφυλακίου
1 Esdr hat auch zweimal ‫ ֵּבית‬gelesen, aber der Übersetzer vermeidet die
Wiederholung desselben Ausdrucks und wählt stattdessen zwei je an ihrer
Stelle exakt passende Äquivalente des Hebräischen.
In Neh 8:5–6 findet sich auf engstem Raum viermal hintereinander der
Ausdruck „das ganze Volk“.
‫ּוכ ִפ ְתחֹו ָע ְמדּו‬
ְ ‫ל־ה ָעם ָהיָ ה‬ ָ ‫י־מ ַעל ָּכ‬ ֵ ‫ל־ה ָעם ִּכ‬
ָ ‫וַ ּיִ ְפ ַּתח ֶעזְ ָרא ַה ֵּס ֶפר ְל ֵעינֵ י ָכ‬5
‫ל־ה ָעם ָא ֵמן ָא ֵמן‬
ָ ‫ֹלהים ַהּגָ דֹול וַ ּיַ ֲענּו ָכ‬
ִ ‫וַ יְ ָב ֶרְך ֶעזְ ָרא ֶאת־יְ הוָ ה ָה ֱא‬6 ‫ל־ה ָעם׃‬
ָ ‫ָכ‬
2 Esdr übersetzt ‫ל־ה ָעם‬
ָ ‫ ָכ‬erwartungsgemäß mit πᾶς ὁ λαός:
2 Esdr 18:5–6: καὶ ἤνοιξεν Εσδρας τὸ βιβλίον ἐνώπιον παντὸς τοῦ λαοῦ ὅτι αὐτὸς ἦν
ἐπάνω τοῦ λαοῦ καὶ ἐγένετο ἡνίκα ἤνοιξεν αὐτό ἔστη πᾶς ὁ λαός . . . καὶ ἀπεκρίθη
πᾶς ὁ λαὸς καὶ εἶπαν αμην
Die beiden Textdifferenzen zwischen MT und 2 Esdr, die diese Stelle auf-
weist, werfen textkritische Fragen auf. So hat die Vorlage von 2 Esdr offen-
bar nicht ‫ל־ה ָעם‬
ָ ‫ ֵמ ַעל ָּכ‬gelesen, sondern nur ‫ ֵמ ַעל ָה ָעם‬. Zwar bezeugt der
lukianische Text + πάντος, aber nach allen Regeln der Textkritik muss das
als spätere Angleichung an den Kontext oder den heutigen MT gelten.
Außerdem hat 2 Esdr nur einmal „Amen“ gelesen—ebenso wie 1 Esdr.
1 Esdr gibt das hebräische vierfache ‫ל־ה ָעם‬ ָ ‫ ָכ‬nicht mit dem viermal
gleichen griechischen Ausdruck wieder.
1 Esdr 9:45–47: καὶ ἀναλαβὼν Εσδρας τὸ βιβλίον τοῦ νόμου ἐνώπιον τοῦ πλήθους
προεκάθητο γὰρ ἐπιδόξως ἐνώπιον πάντων 46καὶ ἐν τῷ λῦσαι τὸν νόμον πάντες
ὀρθοὶ ἔστησαν . . . καὶ ἐπεφώνησεν πᾶν τὸ πλῆθος αμην.
Der Übersetzer splittet den Doppelausdruck „alles Volk“ in „alle“ und
„Volk“ und verteilt die beiden Ausdrücke πάντες und τὸ πλῆθος über die
vier Stellen: Menge—alle—alle—ganze Menge. Die Vorlage von 1 Esdr hat
vermutlich wie MT (2 Esdr L) ‫ל־ה ָעם‬ ָ ‫ ֵמ ַעל ָּכ‬gelesen und deswegen hier
πάντων für ein passendes Äquivalent gehalten. Wie 2 Esdr hat auch 1 Esdr
in seiner Vorlage offenbar nur einfaches „Amen“ gelesen. Hätte er nämlich
zweifaches hebräisches „Amen“ gelesen, aber die buchstäbliche Wieder-
holung in der septuagintaüblichen Formel γένοιτο γένοιτο (Num 5:22; Jdt
13:20; Pss 41:14; 72:19; 89:53; 106:48) vermeiden wollen, hätte er viele andere
Wege gehabt, die Doppelung ohne Wiederholung wiederzugeben.
Da neben der Übersetzung γένοιτο (neben den genannten Stellen

z.B. auch durchweg in Dtn 27) auch die einfache Transkription αμην
vorkommt (1 Chr 16:36, Neh 5:13) aber auch die Übersetzung ἀληθῶς
( Jer 28[35]:6) hätte man sich auch irgendeine Kombination wie αμην
γένοιτο oder ἀληθῶς γένοιτο vorstellen können. Das einfache „Amen“ von
1 Esdr 9:47 ist keine Wiederholungsvermeidung, sondern wie in 2 Esdr
abweichende Vorlage. Die Transkription des hebräischen Wortes, wie sie
im NT normal ist (bei den Synoptikern über 50 Mal, immer einfach; bei
Joh immer doppelt: 25 mal „Amen, amen“), spricht dafür, dass der Über-
setzer den Ausruf als liturgischen Ruf kennt und in der Liturgie 1 Esdr 9/
Neh 8 entsprechend setzt.
7. Differenzierte Übersetzung, wo auch MT differenziert
In Esr 7:7–8 beginnt die Erzählung von Esras Heimkehr. Der Erzähler
datiert sehr genau, wann Esra und seine Karawane nach Jerusalem auf-
gebrochen sind:
7 Mit ihm zog im siebten Jahr des Königs Artaxerxes auch eine Anzahl von
Israeliten sowie von Priestern, Leviten, Sängern, Torwächtern und Tempel-
dienern nach Jerusalem. 8 Im fünften Monat dieses siebten Regierungsjahrs
des Königs kam Esra in Jerusalem an.
Der hebräische Text spricht einmal vom „Jahr sieben“ und einmal vom
„siebten Jahr“. 2 Esdr gibt alle Jahreszahlangaben wie die griechische Spra-
che das verlangt, mit Ordinalzahlen wieder28. Bei ihm heißt es also zwei-
mal ἔτος ἕβδομον. Dagegen differenziert 1 Esdr 8:5—freilich nicht beim
Zahlwort, sondern beim Äquivalent für „Jahr“: einmal schreibt er ἔτος ein-
mal ἐνιαυτὸς. Hier kann neben der Neigung von 1 Esdr bei zwei dicht auf-
einander folgenden gleichen hebräischen Ausdrücken in der Übersetzung
öfter einmal zu variieren29, der Wunsch mitgespielt haben, die leichte
Differenz im hebräischen Ausdruck, auch im Griechischen durchschei-
nen zu lassen. Freilich differenziert er gerade nicht bei der Zahlenangabe,
sondern beim Ausdruck für Jahr.
28 2 Esdr 1:1; 3:8; 4:24; 5:13; 6:3, 15; 7:7–8. Nur in 3:8 und 7:8 hat der hebräische Text Ordi-
nalzahlen, alle anderen Stellen sind hebräische und aramäische Kardinalzahlen.
29 Die Neigung ist kein Zwang, wie das zweimalige „fünfter Monat“ in 1 Esdr 8:5–6
zeigt.
Esr 7:7–8 MT 2 Esdr 7:7–8 1 Esdr 8:5

‫ת־ׁש ַבע‬
ֶ ַ‫רּוׁש ָלםִ ִּב ְׁשנ‬ ָ ְ‫ ֶאל־י‬7 εἰς Ιερουσαλημ ἐν ἔτει ἑβδόμῳ εἰς Ιεροσόλυμα ἔτους
‫ְל ַא ְר ַּת ְח ַׁש ְס ְּתא ַה ֶּמ ֶלְך׃‬ τῷ Αρθασασθα τῷ βασιλεῖ ἑβδόμου βασιλεύοντος
‫רּוׁש ַלםִ ַּבח ֶֹדׁש‬ָ ְ‫וַ ּיָ בֹא י‬8 8καὶ ἤλθοσαν εἰς Ιερουσαλημ Ἀρταξέρξου ἐν τῷ πέμπτῳ
‫יׁשי ִהיא ְׁשנַ ת‬ ִ ‫ַה ֲח ִמ‬ τῷ μηνὶ τῷ πέμπτῳ τοῦτο μηνί οὗτος ἐνιαυτὸς ἕβδομος
‫יעית ַל ֶּמ ֶלְך׃‬ ִ ‫ַה ְּׁש ִב‬ ἔτος ἕβδομον τῷ βασιλεῖ τῷ βασιλεῖ
In Esr 6:14 vermeldet der Erzähler die Vollendung des Tempelbaus und
erwähnt, sie sei durch Dekret Gottes und Dekret der Perserkönige erfolgt.
Der aramäische Text spricht zweimal von ‫ ְט ֵעם‬, gebraucht aber die beiden
leicht verschiedenen (aber beide gebräuchlichen) Konstruktusformen:
zuerst ‫( ַט ַעם‬wie z. B. Esr 7:23), dann aber ‫( ְט ֵעם‬wie z. B. Dan 5:2). Auch
die Präposition ‫ ִמן‬wird einmal assimiliert und einmal nicht assimiliert
gesetzt:
Esr 6:14: ‫ּכֹורׁש וְ ָד ְריָ וֶ ׁש‬
ֶ ‫ּומ ְּט ֵעם‬
ִ ‫ן־ט ַעם ֱא ָלּה יִ ְׂש ָר ֵאל‬
ַ ‫וְ ַׁש ְכ ִללּו ִמ‬
‫וְ ַא ְר ַּת ְח ַׁש ְׂש ְּתא‬
2 Esdr drückt die unterschiedliche Vokalisierung, die er vielleicht gar nicht
kennt, nicht aus. Bei ihm heißt es zweimal ununterschieden ἀπὸ γνώμης:
καὶ κατηρτίσαντο ἀπὸ γνώμης θεοῦ Ισραηλ καὶ ἀπὸ γνώμης Κύρου καὶ Δαρείου
καὶ Αρθασασθα.
1 Esdr differenziert hier, wie zu erwarten. Und es ist hier nicht allein der
übliche Wille zur Variation, sondern mit Sicherheit auch die Überzeu-
gung, göttliches und königliches Dekret seien nicht einfach von derselben
Kategorie.
1 Esdr 7:4–5: καὶ συνετέλεσαν ταῦτα διὰ προστάγματος τοῦ κυρίου θεοῦ Ισραηλ
5καὶ μετὰ τῆς γνώμης Κύρου καὶ Δαρείου καὶ Ἀρταξέρξου.
Ganz richtig schreibt Talshir zu unserer Stelle: „The translator may have
chosen different equivalents in order to distinguish between God and the
kings“30. Den Willen, das menschliche Edikt vom göttlichen Ratschluss
zu unterscheiden zeigt ja später auch die masoretische Vokalisierung
(ebenso wie die verschiedenen Formen der Präposition mn). Es ist gut
möglich, dass eine solche Lesetradition schon zu Zeiten des Übersetzers
von 1 Esdr im Umlauf war, obwohl 2 Esdr davon keine Spur zeigt.

8. Übersetzungsphilosophie und Textkritik
Die eingangs behandelte Praxis des Übersetzers von 1 Esdr, bisweilen mit
Doppelübersetzungen zu arbeiten, erlaubt an anderen Stellen ein siche-
reres textkritisches Urteil. Bei der Toraverlesung in Neh 8 heißt es in 8:6:
„Und Esra pries den Herrn, den großen Gott.“ 2 Esdr bezeugt exakt den-
selben Text. Dagegen heißt es in 1 Esdr 9:46: „Und Esdras pries den Herrn,
den höchsten Gott, den allherrschenden Gott Sabaoth.“
Neh 8:6 1 Esdr 9:46 2 Esdr 18:6

6 ‫ וַ יְ ָב ֶרְך ֶעזְ ָרא ֶאת־יְ הוָ ה‬καὶ εὐλόγησεν Εσδρας τῷ καὶ ηὐλόγησεν Εσδρας
‫ֹלהים ַהּגָ דֹול‬
ִ ‫ ָה ֱא‬κυρίῳ θεῷ ὑψίστῳ κύριον τὸν θεὸν τὸν μέγαν
θεῷ σαβαωθ παντοκράτορι
Hanhart schreibt die Erweiterung in 1 Esdr allen Ernstes dem Übersetzer

zu: „Es scheint, dass die vorgegebene Bezeichnung Gottes als ‫ ַהּגָ דֹול‬für
den Übersetzer der Anlaß war, die ihm in den Überlieferungen vorliegen-
den Gottesepitheta zu vereinigen“31. Die Übersetzungsphilosophie unse-
res griechischen Übersetzers schließt dergleichen kategorisch aus.
Richtiger liegt Talshir, die in σαβαωθ παντοκράτορι eine „double
representation“32, „double translation“33 von ‫ ְצ ָבאֹות‬erkennt. Einmal gibt
1 Esdr den wohl auch liturgisch gebräuchlichen hebräischen Ausdruck
in Transkription, dann aber auch noch zusätzlich in griechischer Über-
setzung wieder. Die beiden Wiedergaben σαβαωθ und παντοκράτωρ sind
gängig in der LXX und im NT34. Die hier vorliegende Doppelübersetzung
beweist, dass der Übersetzer in seiner hebräischen Vorlage ‫ֹלהי ְצ ָבאֹות‬
ֵ ‫ֱא‬
tatsächlich vor Augen hatte. Seine Vorlage weicht hier von MT ab. Daraus
31 R. Hanhart, Text und Textgeschichte des 1. Esrabuches (MSU XII, AAWG.PH 91;
Göttingen: Vandenhoeck & Ruprecht, 1974), 69.
32 Talshir, 1 Esdras. From Origin to Translation, 241.
33 Talshir, I Esdras. A Text Critical Commentary, 491. Talshir erwägt hier, anders als
idem, 1 Esdras. From Origin to Translation, 241, noch andere, weit weniger wahrscheinli-
che Möglichkeiten, etwa „der allherrschende Gott Sabaoth“ als Doppel-übersetzung von
„der große Gott.“
34 Das Gottesepitheton ‫ ְצ ָבאֹות‬wird in der LXX bei Jesaja transkribiert (σαβαωθ), im
restlichen AT übersetzt, und zwar in Sam-Kön, Pss und Zeph als Genitiv (κύριος τῶν
δυνάμεων, θεὸς τῶν δυνάμεων), in Ijob, Dodekapropheton (ohne Zeph), Jer mit der Apposi-
tion παντοκράτωρ. Im NT steht zweimal κύριος σαβαωθ (Röm 9:29; Jak 5:4) und einmal κύριος
παντοκράτωρ (2 Kor 6:18).
folgt natürlich noch nicht, dass 1 Esdr hier den älteren Text bietet. Das ist
sogar eher unwahrscheinlich, denn ‫ ְצ ָבאֹות‬kommt zwar in den Chroniken
(1 Chr 11:9; 17:7, 24) als Gottesepitheton vor, nie aber in Esr–Neh. So ist
es in der hebräischen Vorlage von 1 Esdr wohl eine spätere Erweiterung,
die aber (gegen Hanhart) auf jeden Fall schon in der hebräischen Vorlage
stand und keinesfalls dem Übersetzer zuzutrauen ist.
In Esr 5:5—wir haben diese Stelle oben schon besprochen—fanden die
griechischen Übersetzer den aramäischen Ausdruck ‫הּודיֵ א‬ ָ ְ‫„ ָׂש ֵבי י‬die Älte-
sten der Juden“. 2 Esdr entscheidet sich für τὴν αἰχμαλωσίαν Ιουδα, also „die
Exulanten, die Ältesten der Juden“ (1 Esdr 6:5: ἐπὶ τὴν αἰχμαλωσίαν παρὰ τοῦ
κυρίου οἱ πρεσβύτεροι τῶν Ιουδαίων). Es war für den Übersetzer offenbar nicht
entscheidbar, ob er ‫„ ָׂש ֵבי‬Älteste“ oder ‫„ ְׁש ִבי‬Gefangenschaft“ lesen sollte.
Diese Technik der Doppelübersetzung verwendet er sogleich in v. 8
erneut. Hier nun hilft die Kenntnis seiner Übersetzungstechnik bei der
Entscheidung einer textkritischen Frage. 1 Esdr hat in 6:8 gegenüber Esr
MT 5:8 einen merklichen Textüberschuss. Er erwähnt die Ältesten der
Exulantenschaft.
Esr 5:8 MT 2 Esdr 5:8 1 Esdr 6:8

‫יְ ִד ַיע ֶל ֱהוֵ א ְל ַמ ְל ָּכא‬8 8 γνωστὸν ἔστω τῷ βασιλεῖ 8 πάντα γνωστὰ ἔστω τῷ κυρίῳ
‫י־אזַ ְלנָ א ִליהּוד‬ ֲ ‫ ִּד‬ὅτι ἐπορεύθημεν εἰς τὴν ἡμῶν τῷ βασιλεῖ ὅτι παραγενόμενοι
‫ ְמ ִדינְ ָּתא‬Ιουδαίαν χώραν εἰς τὴν χώραν τῆς Ιουδαίας
καὶ ἐλθόντες εἰς Ιερουσαλημ τὴν
πόλιν κατελάβομεν τῆς αἰχμαλωσίας
τοὺς πρεσβυτέρους τῶν Ιουδαίων ἐν
‫ְל ֵבית ֱא ָל ָהא ַר ָּבא‬ εἰς οἶκον τοῦ θεοῦ τοῦ Ιερουσαλημ τῇ πόλει
‫וְ הּוא ִמ ְת ְּבנֵ א ֶא ֶבן‬ μεγάλου καὶ αὐτὸς οἰκοδομοῦντας οἶκον τῷ κυρίῳ
‫ּגְ ָלל וְ ָאע ִמ ְּת ָׂשם‬ οἰκοδομεῖται λίθοις ἐκλεκτοῖς μέγαν καινὸν διὰ λίθων ξυστῶν
‫ְּב ֻכ ְת ַלּיָ א‬ καὶ ξύλα ἐντίθεται ἐν τοῖς πολυτελῶν ξύλων τιθεμένων ἐν τοῖς
τοίχοις τοίχοις
Es ist ziemlich klar und weitgehend anerkannt,35 dass der Rückverweis

auf „diese Ältesten“ im folgenden Vers Esr 5:9 beweist, dass sie auch in
5:8 einst erwähnt gewesen sein müssen, mithin das Plus von 1 Esdr 6:8
35 H. G. M. Williamson, Ezra, Nehemiah (WBC 16; Waco: Word Books, 1985), 70;
J. Blenkinsopp, Ezra-Nehemiah. A Commentary (OTL; London: SCM Press, 1989), 119;
D. Böhler, Die heilige Stadt in Esdra α und Esra-Nehemia. Zwei Konzeptionen der Wieder-
herstellung Israels (OBO 158; Freiburg / Göttingen: Universitätsverlag / Vandenhoeck &
Ruprecht, 1997), 154–158.
den gegenüber Esr 5:8 ursprünglicheren Text bietet. Gunneweg jedoch

hält das Minus von MT für ursprünglich und das Plus von 1 Esdr für eine
erzählerische Auffüllung durch den Verfasser von 1 Esdr36. Da Gunneweg
1 Esdr für „ein Zeugnis des frühhellenistischen Judentums“37 hält, scheint
er an einen griechischen Verfasser oder Kompilator als Auffüller zu den-
ken. Die Kenntnis der Übersetzungspraxis des Übersetzers von 1 Esdr
erlaubt es uns jedoch, auszuschließen, dass das Plus in 1 Esdr 6:8 auf grie-
chischer Ebene entstanden ist.
Die Tatsache, dass er mit τῆς αἰχμαλωσίας τοὺς πρεσβυτέρους wieder eine
Doppelübersetzung für ‫ ָׂש ֵבי‬bietet, beweist, dass er einen aramäischen
Text vor Augen hatte und sich wie schon in 6:5 (Esr 5:5) nicht entscheiden
konnte, ob er ‫„ ָׂש ֵבי‬Älteste“ oder ‫„ ְׁש ִבי‬Gefangenschaft“ lesen sollte38. Das
Plus in 1 Esdr 6:8 ist sicher aramäischen Ursprungs und, wie die meisten
Ausleger mit dem textkritischen Apparat der BHS urteilen, obendrein der
gegenüber MT ursprünglichere Text.
Etwas anders gelagert und doch benachbart ist der Fall von 1 Esdr 1:4
(2 Chr 35:3). Nach MT weist König Joschija bei der Feier des Paschafestes
die Leviten an, Gott und seinem Volk zu dienen:
2 Chr 35:3: ‫יכם וְ ֵאת ַעּמֹו יִ ְׂש ָר ֵאל׃‬
ֶ ‫ֹלה‬
ֵ ‫ַע ָּתה ִע ְבדּו ֶאת־יְ הוָ ה ֱא‬
Der hebräische Text hat ein einziges Verb mit zwei Objekten. Auch 2 Par
35:3 bezeugt eben diesen Text: νῦν οὖν λειτουργήσατε τῷ κυρίῳ θεῷ ὑμῶν
καὶ τῷ λαῷ αὐτοῦ Ισραηλ.
Dagegen bietet 1 Esdr zwei verschiedene Verben mit je einem Objekt:
1 Esdr 1:4: 4καὶ νῦν λατρεύετε τῷ κυρίῳ θεῷ ὑμῶν καὶ θεραπεύετε τὸ ἔθνος αὐτοῦ
Ισραηλ
Die Frage ist nun, ob wir hier ein textkritisches Problem haben, wie
Talshir meint, oder nur eine Frage der Übersetzungstechnik, wie Klein
urteilt. Talshir39 geht davon aus, dass λατρεύετε für ‫ ִע ְבדּו‬steht und den
kultischen Dienst vor Gott meint. Für den Dienst am Volk und damit
das andere Verb θεραπεύετε nimmt sie, nicht ohne Zögern, an, dass ein
anderes Verb im hebräischen Text gestanden haben müsse, näherhin
‫ ָׁש ְרתּו‬. Dagegen glaubt Klein, die Wahl des ersten griechischen Äquiva-
lents habe den Übersetzer gezwungen für das zweite Objekt ein eigenes
36 A. H. J. Gunneweg, Esra (KAT XIX 1; Gütersloh: Gerd Mohn, 1985), 99f.
37 Ibid., 24.
38 Ebenso Klein, Studies, 221 und Talshir, I Esdras. A Text Critical Commentary, 334.
Verbenäquivalent zu suchen: „Since the translator chose λατρεύω (wor-

ship) to translate ‫ ִע ְבדּו‬, he had to use a more neutral word to indicate the
care for the people of Israel“40.
Die unterdessen wohlbekannte Bereitschaft unseres Übersetzers, einen
hebräischen Ausdruck durch eine griechische Doppelübersetzung oder
Hendiadyoin wiederzugeben oder aber ein und dasselbe wiederholte
hebräische Wort mit variierenden griechischen Vokabeln zu überset-
zen, macht es meines Erachtens völlig unnötig, hier ein zweites Wort für
„dienen“ in der hebräischen Vorlage zu postulieren. Allenfalls könnte man
doppeltes ‫ ִע ְבדּו‬vermuten, denn auch das würde der Übersetzer von 1 Esdr
mit λατρεύετε und θεραπεύετε wiedergeben. Es ist ihm aber nach allem,
was wir von ihm wissen, ohne weiteres zuzutrauen, dass er auch ein nur
einmal dastehendes hebräisches Wort mit einer Doppelübersetzung wie-
dergibt, wenn es ihm, wie hier, unpassend erscheint, dieselbe Art des Die-
nens auf Gott und Menschen anzuwenden.
Auf den ersten Blick etwas verwickelt liegt der Fall von Esr 9:12. Esra
richtet nach der Aufdeckung des Mischehenskandals ein langes Bußge-
bet an Gott (Esr 9:5–15). Dabei kommt er in 9:11–12 auf die Vorschriften
der Tora zu sprechen, wie die einwandernden Israeliten mit den im Land
befindlichen Nichtisraeliten umgehen sollen. Dabei zitiert er neben Lev
18 auch Dtn 23:7:
Ihr sollt nicht Friedens- noch Freundschaftsvertrag mit ihnen suchen in
Ewigkeit, damit ihr stark bleibt und das Gut des Landes essen und euren
Kindern vererben könnt in Ewigkeit.
Das hebräische Gebet sagt zweimal ‫ד־עֹולם‬
ָ ‫„ ַע‬in Ewigkeit“. 2 Esdr über-
setzt stereotyp zweimal mit ἕως αἰῶνος. 1 Esdr gibt aber das erste ‫ד־עֹולם‬
ָ ‫ַע‬
mit τὸν ἅπαντα χρόνον wieder, das zweite mit ἕως αἰῶνος.
Esr 9:12 MT 2 Esdr 9:12 1 Esdr 8:82

‫ֹלמם‬ ָ ‫א־ת ְד ְרׁשּו ְׁש‬ ִ ֹ ‫וְ ל‬ οὐκ ἐκζητήσετε εἰρήνην καὶ οὐ ζητήσετε εἰρηνεῦσαι
‫ד־עֹולם‬
ָ ‫טֹוב ָתם ַע‬ָ ְ‫ו‬ αὐτῶν καὶ ἀγαθὸν αὐτῶν τὰ πρὸς αὐτοὺς τὸν ἅπαντα
‫ְל ַמ ַען ֶּת ֶחזְ קּו וַ ֲא ַכ ְל ֶּתם‬ ἕως αἰῶνος ὅπως ἐνισχύσητε χρόνον ἵνα ἰσχύσαντες
‫ֶאת־טּוב ָה ָא ֶרץ‬ καὶ φάγητε τὰ ἀγαθὰ τῆς φάγητε τὰ ἀγαθὰ τῆς γῆς
‫יכם‬ ֶ ֵ‫הֹור ְׁש ֶּתם ִל ְבנ‬ַ ְ‫ו‬ γῆς καὶ κληροδοτήσητε τοῖς καὶ κατακληρονομήσητε τοῖς
‫ד־עֹולם׃‬
ָ ‫ַע‬ υἱοῖς ὑμῶν ἕως αἰῶνος υἱοῖς ὑμῶν ἕως αἰῶνος
40 Klein, Studies, 248.

Klein41 erkennt das Zitat aus Dtn 23:7 in der ersten Vershälfte Esr 9:12a
und 1 Esdr 8:82a:
Dtn 23:7: ‫עֹולם‬
ָ ‫ֹלמם וְ ט ָֹב ָתם ָּכל־יָ ֶמיָך ְל‬
ָ ‫א־ת ְדר ֹׁש ְׁש‬
ִ ֹ‫ל‬
Er vergleicht daher nur die ersten Vershälften mit Dtn 23:7 und sieht,
dass der deuteronomischen doppelten Zeitbestimmung ‫עֹולם‬ ָ ‫ָּכל־יָ ֶמיָך ְל‬
in 2 Esdr das einfache ἕως αἰῶνος, in 1 Esdr 8:82 das ebenfalls einfache τὸν
ἅπαντα χρόνον entspricht. Er schließt nun daraus, dass das ‫ד־עֹולם‬ ָ ‫ ַע‬bzw.
ἕως αἰῶνος von MT bzw 2 Esdr einerseits und das τὸν ἅπαντα χρόνον von
1 Esdr andererseits im MT des Dtn zu einer Konflation verbunden worden
wären. Es scheint mir jedoch von vornherein sehr unwahrscheinlich, dass
der hebräische Text der Tora durch Varianten der eher späten Esrabücher
beeinflusst worden sein sollte. Klein berück-sichtigt vor allem überhaupt
nicht, dass in Esras Gebet gleich zweimal hintereinander ‫ד־עֹולם‬ ָ ‫ ַע‬steht,
wovon nur das erste mit τὸν ἅπαντα χρόνον, das zweite aber mit der für
Klein alternativen Variante ἕως αἰῶνος übersetzt wird.
Talshir42 neigt eher zu einer textkritischen Auswertung der Über-
setzung von 1 Esdr. Sie schließt zwar nicht aus, dass der Übersetzer von
1 Esdr, ebenso wie der von 2 Esdr, das masoretische ‫ד־עֹולם‬ ָ ‫ ַע‬gelesen
habe. Sie vermutet aber eher, 1 Esdr habe in seiner hebräischen Esravor-
lage das deuteronomische ‫ ָּכל־יָ ֶמיָך‬vorgefunden. Dann würde 1 Esdr einen
hebräischen Vorlagetext bezeugen, der das erste Mal ‫ ָּכל־יָ ֶמיָך‬das zweite
Mal aber ‫ד־עֹולם‬ָ ‫ ַע‬gelesen hätte. 2 Esdr bezeugte einfach den jetzigen MT
mit zweifachem ‫ד־עֹולם‬ ָ ‫ ַע‬.
Tatsächlich wäre τὸν ἅπαντα χρόνον eine für den Übersetzer von 1 Esdr
untypisch freie Wiedergabe von ‫ד־עֹולם‬ ָ ‫ ַע‬, ist doch griechisch χρόνος in
fünf von 10 Fällen die Wiedergabe von hebr. ‫ יָ ִמים‬43, dreimal aber auch
für ‫ זְ ָמן‬44. Insofern ist Talshirs Vermutung, er habe hier ‫ל־הּיָ ִמים‬ ַ ‫ ָּכ‬gele-
sen, durchaus begründet. Die Variante ist wohl von Dtn 23:7 MT her in
den hebräischen Vorlagetext von 1 Esdr gekommen. Sie hat sogar einige
Chancen, den ursprünglicheren Esratext wiederzugeben, da MT eine spä-
tere Assimilation darstellen könnte45. Hier spricht der Dtn-Text für eine
41 Ibid., 79.

42 Talshir, I Esdras. A Text Critical Commentary, 452f.
43 1 Esdr 1:18, 55; 2:15; 5:70; 8:73 // 2 Chr 35:18; 36:21; Esr 4:7; 4:5; 9:7.
44 In 1 Esdr 1:22 ist die Vorlage gänzlich verschieden, in 1:40 möglicherweise auch. In 6:3
gibt χρόνος aramäisch ‫ זְ ָמן‬wieder und in 9:12 hebräisch ‫זְ ָמן‬/‫ ֵעת‬.
45 Die LXX-Fassung der zitierten Dtn-Stelle zeigt, dass weder 1 Esdr noch 2 Esdr auf
diese griechische Dtn-Fassung zurückgegriffen haben. Dtn 23:7 LXX: οὐ προσαγορεύσεις
εἰρηνικὰ αὐτοῖς καὶ συμφέροντα αὐτοῖς πάσας τὰς ἡμέρας σου εἰς τὸν αἰῶνα.
Variation der Zeitbestimmung in der Vorlage von 1 Esdr. Der Variations-

wille des Übersetzers war hier vermutlich gar nicht gefordert.
9. Weitere Schlussfolgerungen aus den Beobachtungen
Eine Passage in 1 Esdr, die in Esr–Neh und damit auch in 2 Esdr keine
Entsprechung hat, ist die sogenannte Pagenerzählung 1 Esdr 3:1–5:6, eine
Serubbabellegende. Seit Torreys Ezra Studies aus dem Jahr 191046 ist weit-
hin anerkannt, dass sie im Original aramäisch war.
Pohlmann hatte in seinen Studien zu 1 Esdr die Auffassung vertre-
ten, die Pagenerzählung sei von einem anderen Übersetzer ins Griechi-
sche übertragen worden als der Rest von 1 Esdr47. Er verweist auf die
verschiedene Übersetzung vermutlich gleicher Vorlageausdrücke in den
beiden Teilen. Demnach wäre die Pagenerzählung erst auf griechischer
Sprachebene interpoliert worden. Talshir dagegen schreibt die griechi-
sche Übersetzung des ganzen Buches ein und derselben Hand zu48. Das
bedeutet, dass die Interpolation der aramäischen Erzählung schon in die
hebräisch-aramäische Tempelbauerzählung erfolgte.
Pohlmanns Argumente beweisen wegen der fehlenden Äquivalenten-
konstanz unseres Übersetzers wenig. Dagegen spricht die Technik der
doppelten Wiedergabe ein und desselben Vorlageausdrucks auch in 1 Esdr
3–4 für die Hand desselben Übersetzers wie im restlichen Buch.
In 1 Esdr 3:4 werden die drei Pagen des Königs zweimal als Leib-wächter
bezeichnet:
1 Esdr 3:4: οἱ τρεῖς νεανίσκοι οἱ σωματοφύλακες οἱ φυλάσσοντες τὸ σῶμα τοῦ
βασιλέως
Das ist, wie Talshir wohl zu Recht vermutet, die doppelte Wiedergabe
eines aramäischen Ausdruckes: „Its first part offers a technical term,
σωματοφύλακες, well attested in the second century B.C.E. (papyri, Polybius).
46 C. Torrey, Ezra Studies (Chicago: University of Chicago Press, 1910 [repr. New York:
KTAV, 1970]), 23–25; K.-F. Pohlmann, Studien zum dritten Esra: Ein Beitrag zur Frage nach
dem ursprünglichen Schluß des chronistischen Geschichtswerks (FRLANT 104; Göttingen:
Vandenhoeck & Ruprecht, 1970), 48.
47 Pohlmann, Studien, 150f.
48 Talshir, 1 Esdras. From Origin to Translation, 103 und 106.
The second part, οἱ φυλάσσοντες τὸ σῶμα τοῦ βασιλέως, presumably repre-

sents the wording of the Vorlage more closely“49.
Das spricht dafür, dass ein und derselbe Übersetzer das ganze Buch
übersetzt hat, mithin die aramäische Pagenerzählung schon in die hebrä-
isch-aramäische Vorlage integriert war.
Nach 1 Esdr 3:6 soll der Sieger des Pagenwettstreits „aus güldenen (Gefä-
ßen) trinken, auf einem Bett aus Gold schlafen und einen goldgezäumten
Wagen (fahren)“:
1 Esdr 3:6: ἐν χρυσώμασιν πίνειν καὶ ἐπὶ χρυσῷ καθεύδειν καὶ ἅρμα χρυσοχάλινον
Im Aramäischen stand hier zweifellos dreimal ‫ ְד ַהב‬. Der Übersetzer der
Pagenerzählung findet drei Variationen.
Nach Talshir50 liegt auch dem griechischen κυριεύει und δεσπόζει in 1
Esdr 4:3 doppeltes aramäisches šlt zu Grunde, ebenso wie dann erneut in
4:14. Der Übersetzer hätte variiert.
Hinter κατ᾽ ἐνιαυτὸν in 1 Esdr 4:51 und 4:52 steht aramäisch šnh bšnh51.
All diese Phänomene, die wir im übrigen 1 Esdr beobachtet haben,
zeigen sich anscheinend auch in der Pagenerzählung, für die wir keine
aramäische Vorlage besitzen. Das stützt die These ihrer Einfügung auf
semitischer Sprachebene52.
10. Ein Blick auf Daniel (MT, LXX, Th)
Der uns aus Esra bekannte Doppelausdruck ‫ יֹום ְּביֹומֹו‬findet sich auch in
Dan 1:5. DanTh schreibt dafür τῆς ἡμέρας καθ᾽ ἡμέραν. DanLXX dagegen καθ᾽
ἑκάστην ἡμέραν. Dieser geht vor wie 1 Esdr, DanTh wie 2 Esdr. Aber beide
hatten die gleiche Vorlage.
In Dan 2:47 spricht Nebukadnezzar von Gott als dem, der Geheimnisse
enthüllen kann, weil Daniel das Geheimnis des Königs enthüllen konnte.
Der aramäische Text spricht zweimal von glh.
Dan 2:47: ‫וְ גָ ֵלה ָרזִ ין ִּדי יְ ֵכ ְל ָּת ְל ִמגְ ֵלא ָרזָ ה ְדנָ ה‬. DanTh gibt das ohne jede
Variation wieder mit καὶ ἀποκαλύπτων μυστήρια ὅτι ἠδυνήθης ἀποκαλύψαι
τὸ μυστήριον τοῦτο. DanLXX variiert das Verb: ὁ ἐκφαίνων μυστήρια κρυπτὰ
49 Talshir, I Esdras. A Text Critical Commentary, 138. Sie denkt an die Vorlage dy ntryn
mlk’.
50 Ibid., 174 und 189f.
51 Vgl. ibid., 233f.
52 Wo der Übersetzer aramäische Wortwiederholungen nachahmt, wie in 4:7–9 und
4:37, ist die rhetorische Absicht der Wiederholungen offensichtlich.
μόνος ὅτι ἐδυνάσθης δηλῶσαι τὸ μυστήριον τοῦτο. Neben dem allgemeinen

Variationswillen den der Übersetzer mit 1 Esdr teilt, mag auch die Überle-
gung mitgespielt haben, dass er göttliches und menschliches Offenbaren
nicht mit demselben Verbum beschreiben wollte. Eine textkitische Frage
stellt sich hier—entgegen dem ersten Anschein—nicht.
Nach Dan 1:8 war Daniel entschlossen, sich durch die königliche Ernäh-
rung nicht zu verunreinigen und bittet den Chefbeamten, sich nicht ver-
unreinigen zu müssen. Im hebräischen Text steht zweimal g’l hitp.
Dan 1:8: ‫ּוביֵ ין‬
ְ ‫ל־לּבֹו ֲא ֶׁשר לֹא־יִ ְתּגָ ַאל ְּב ַפ ְת ַּבג ַה ֶּמ ֶלְך‬
ִ ‫וַ ּיָ ֶׂשם ָּדנִ ּיֵ אל ַע‬
‫יסים ֲא ֶׁשר לֹא יִ ְתּגָ ָאל׃‬ ִ ‫ִמ ְׁש ָּתיו וַ יְ ַב ֵּקׁש ִמ ַּׂשר ַה ָּס ִר‬
DanTh übersetzt: καὶ ἔθετο Δανιηλ ἐπὶ τὴν καρδίαν αὐτοῦ ὡς οὐ μὴ ἀλισγηθῇ
ἐν τῇ τραπέζῃ τοῦ βασιλέως καὶ ἐν τῷ οἴνῳ τοῦ πότου αὐτοῦ καὶ ἠξίωσε τὸν
ἀρχιευνοῦχον ὡς οὐ μὴ ἀλισγηθῇ.
DanLXX variiert das Verb: καὶ ἐνεθυμήθη Δανιηλ ἐν τῇ καρδίᾳ ὅπως
μὴ ἀλισγηθῇ ἐν τῷ δείπνῳ τοῦ βασιλέως καὶ ἐν ᾧ πίνει οἴνῳ καὶ ἠξίωσε τὸν
ἀρχιευνοῦχον ἵνα μὴ συμμολυνθῇ. Die Kenntnis seiner Übersetzungsweise
erübrigt textkritische Bedenken.
In Dan 1:2 steht dreimal „Gotteshaus.“ Zunächst bezeichnet ‫ֹלהים‬ ִ ‫ית־ה ֱא‬
ָ ‫ֵב‬
den Jerusalemer Tempel. Dann ist ‫ֹלהיו‬ ָ ‫ ֵּבית ֱא‬ein babylonischer Götzen-
tempel. Und schließlich meint ‫ֹלהיו‬ ָ ‫אֹוצר ֱא‬
ַ ‫ ֵּבית‬die Schatzkammer in
diesem Götzentempel. DanTh schreibt dreimal οἶκος τοῦ θεοῦ und bezeugt
exakt den MT. DanLXX macht aus dem ersten „Gotteshaus“ ein ἱερόν, setzt
aber nicht einen Genitiv („Geräte des Tempels“), sondern ein Adjektiv
(„Tempelgeräte“) wie es 1 Esdr unzählige Male tut53. Birte Braasch ver-
kennt in ihrer Dissertation über DanLXX die Übersetzungstechnik, wenn
sie meint, die alte Danielübersetzung betone hier die „Heiligkeit“ der
Geräte, stelle „jedoch keinen expliziten Bezug zum Tempel her“54.
An der zweiten Stelle hat die Vorlage von DanLXX offensichtlich das
hebräische ‫ֹלהיו‬ ָ ‫ ֵּבית ֱא‬nicht gelesen. An der dritten Stelle aber hat er
‫ֹלהיו‬
ָ ‫ ֵּבית ֱא‬oder ‫ֹלהיו‬
ָ ‫אֹוצר ֱא‬
ַ ‫ ֵּבית‬gelesen, gibt es aber mit τῷ εἰδωλίῳ
αὐτοῦ wieder, weil ein Götzentempel für ihn nicht dieselbe Vokabel
53 Talshir, 1 Esdras. From Origin to Translation, 211; Böhler, „ ‚Treu und schön‛,“ 102.
54 B. Braasch, Die LXX-Übersetzung des Danielbuchs—eine Orientierungshilfe für das
religiöse und politisch-gesellschaftliche Leben in der ptolemäischen Diaspora (Diss. Ham-
burg, 2004. Veröffentlicht unter URN: urn:nbn:de:gbv:18–21588; URL: http://www.sub
.uni-hamburg.de/opus/volltexte/2004/2158/), 37f.
verdient wie der Jerusalemer Tempel des wahren Gottes. Exakt dieselbe
Differenzierung hatte 1 Esdr in 2:7 auch vorgenommen55.
Dan 1:2 LXX Th

‫ֹלהים‬
ִ ‫ית־ה ֱא‬
ָ ‫ּומ ְק ָצת ְּכ ֵלי ֵב‬ִ καὶ μέρος τι τῶν ἱερῶν καὶ ἀπὸ μέρους τῶν
‫ץ־ׁשנְ ָער ֵּבית‬
ִ ‫יאם ֶא ֶר‬ ֵ ‫וַ יְ ִב‬ σκευῶν τοῦ κυρίου καὶ σκευῶν οἴκου τοῦ θεοῦ
‫ֹלהיו‬
ָ ‫ֱא‬ ἀπήνεγκεν αὐτὰ εἰς καὶ ἤνεγκεν αὐτὰ εἰς
‫ת־ה ֵּכ ִלים ֵה ִביא ֵּבית‬
ַ ‫וְ ֶא‬ Βαβυλῶνα καὶ ἀπηρείσατο γῆν Σεννααρ οἶκον τοῦ
‫ֹלהיו׃‬
ָ ‫אֹוצר ֱא‬ַ αὐτὰ ἐν τῷ εἰδωλίῳ αὐτοῦ θεοῦ αὐτοῦ καὶ τὰ σκεύη
εἰσήνεγκεν εἰς τὸν οἶκον
θησαυροῦ τοῦ θεοῦ αὐτοῦ
11. Schluss
Wer dolmetschen will, hatte Luther gesagt, muss einen großen Vorrat an
Wörtern haben, damit er das je passende Äquivalent setzen kann. Das
gilt für einen dezidiert zielsprachenorientierten Übersetzer wie den von
1 Esdr (und DanLXX). Sie übersetzen exakt und keineswegs frei, variieren
aber in der Äquivalentenwahl—und oft sehr überlegt. Wer diese Technik
kennt, kann besser unterscheiden zwischen der Bezeugung textkritischer
Varianten und bloßer Variation in einer entschieden zielsprachenorien-
tierten Übersetzung. 2 Esdr und DanTh ersparen uns solche Überlegungen,
aber sie geben textkritisch ohnehin wenig her und „grosse vorrath von
worten“, wie Luther meinte, brauchten die auch nicht.
55 1 Esdr 2:7: καὶ ὁ βασιλεὺς Κῦρος ἐξήνεγκεν τὰ ἱερὰ σκεύη τοῦ κυρίου ἃ μετήγαγεν
Ναβουχοδονοσορ ἐξ Ιερουσαλημ καὶ ἀπηρείσατο αὐτὰ ἐν τῷ ἑαυτοῦ εἰδωλίῳ.
In Search of the Old Greek Text of 4 Maccabees
Robert J. V. Hiebert
1. Introduction
The story of the martyrdoms, during the reign of Antiochus IV Epipha-

nes, of an elderly Jewish scribe (γραμματεύς—2 Macc 6:18) or priest and
lawyer (ἱερεύς, νομικός—4 Macc 5:4) named Eleazar and of seven broth-
ers and their mother is told in 2 Macc 6:18–7:42, and in a considerably
expanded version, in 4 Macc 5–18. Antiochus and his officials are deter-
mined to make the Jews “forsake their ancestral laws” (2 Macc 6:1), to
force them “to renounce Judaism by tasting defiling foods” (4 Macc 4:26),
to compel them “to taste pork and food sacrificed to idols” (4 Macc 5:2),
and thereby to signal their willingness “to change over to Greek customs”
(2 Macc 6:9). Antiochus orders that the most barbarous forms of torture
should be inflicted upon the eight males, beginning with Eleazar, in his
attempt to make them yield. When they do not, but instead they defy him
boldly, he has them put to death. According to the 4 Maccabees version,
the mother of the seven brothers is the last one of the group still alive,
and she throws herself into the fire “so that no one might touch her body”
(4 Macc 17:1).1 In both versions, the martyrs’ courage and faithful obedi-
ence “to the holy God-given law” (2 Macc 6:23) are celebrated. In 4 Mac-
cabees, their story is recounted in detail to support the author’s thesis in
his philosophical treatise, i.e., that αὐτοδέσποτός ἐστιν τῶν παθῶν ὁ εὐσεβὴς
λογισμός “pious reason is absolute master of the passions” (4 Macc 1:1; cf.
vv. 7, 13, 30 passim).
Greek 4 Maccabees is the only one of the four books that traditionally
include the name Maccabees in the title for which to date there is no criti-
cal edition in the Göttingen Septuaginta series. I am currently at work to
remedy that lacuna, and a first draft of the critically-reconstructed text,
1 2 Macc 7:41 reports simply that “[l]ast of all, the mother died, after her sons,” without
offering any details as to the circumstances of her death. Throughout this paper, English
translations of Septuagint texts are, unless otherwise indicated, taken from A. Pietersma
and B. G. Wright, eds., A New English Translation of the Septuagint (New York / Oxford:
Oxford University Press, 2007), commonly referred to as NETS. My own translations of
texts are written in italics.
128 robert j. v. hiebert
based on the manuscript evidence of more than 70 extant Greek man-

uscripts along with the Syriac daughter version, is nearing completion.
This is not to say that there have been no preliminary efforts to recover
the wording of the original author. The first such undertaking was that of
Otto F. Fritzsche, whose 1871 edition is based mainly on the great codi-
ces Sinaiticus (S) and Alexandrinus (A)—dated to the fourth and fifth
centuries, respectively—though variant readings from close to a dozen
other manuscripts are recorded in the apparatus.2 A little more than two
decades later, the first edition of Henry B. Swete’s The Old Testament in
Greek according to the Septuagint appeared. Swete produced an essentially
diplomatic edition of 4 Maccabees in which he relied primarily on A, but
whose carelessly written text he corrected on the basis of S and Codex
Venetus (V), an eighth or ninth century uncial.3 In Alfred Rahlfs’ edition
of the Septuagint published in 1935, the text of 4 Maccabees is an eclectic
one that is based on S and A, with preference frequently being given to
S when the two diverge, and with substantive variants from V included
in the apparatus as well.4 The German translation of 4 Maccabees by
H.-J. Klauck, published in 1989, represents an advance in our understand-
ing of the text of this book as it left the hands of the original author.5
He renders a Greek text that has benefitted from text-critical proposals
communicated to him by Robert Hanhart, proposals that are based on
collations that have been carried out at the Septuaginta-Unternehmen in
Göttingen.6
The following list of Greek manuscripts and manuscript groups is based
upon the work of Klauck and Hanhart, but it has also been modified in
the light of my research and in consultation with Detlef Fraenkel of the
Septuaginta-Unternehmen:
Uncials: A S V
A 542 (11:5 – fin libri)
L: 236 491 534 728
q: 71 74 120 370 380 452 731 3002
2 O. F. Fritzsche, ed., ΜΑΚΚΑΒΑΙΩΝ ΤΕΤΑΡΤΟΣ, Libri apocryphi Veteris Testamenti

graece (Leipzig: Brockhaus, 1871), 351–386.
3 H. B. Swete, ed., The Old Testament in Greek according to the Septuagint, vol. 3:
Hosea—4 Maccabees, Psalms of Solomon, Enoch, The Odes (3rd ed.; Cambridge: Cambridge
University Press, 1894; 1905), vi.
4 A. Rahlfs, ed., Septuaginta: Id est Vetus Testamentum graece iuxta LXX interpretes
(Stuttgart: Württembergische Bibelanstalt, 1935).
5 H.-J. Klauck, 4. Makkabäerbuch ( JSHRZ 3.6; Gütersloh: Mohn, 1989), 641–763.
6 Ibid., 678–679.
in search of the old greek text of 4 maccabees 129
q1: 44 107 610

q2: 55 747
q´’ = q + q1 + q2 q´ = q + q1 q’ = q + q2 etc.
m: 316 317 322 325 391 397 446 457 467 472 473 586 591 592 594 595 596 597
607 617 639 640 641 656 677 682 683 686 695 699 713 714 774 778 782 789
m1: 455 585
m2: 587 738
m3 (init libri – 11:4): 62 542 747c/mg
m´’` = m + m1 + m2 + m3 m´ = m + m1 m’ = m + m2 m` = m + m3 etc.
11:5 – fin libri: m´’ 62 747c/mg
Codices mixti: 46 (52 332) 58 340 577 668 690 741 771 773 930
A few explanatory comments regarding these manuscript groupings are
in order.
1. Manuscript groups are delineated on the basis of regular patterns of

agreement among witnesses with regard to divergences from what the
editor of a Septuaginta edition reconstructs as the original text, based
on a careful assessment of the extant manuscript evidence.
2. The uncials, which are the most consistent witnesses to the original
text of 4 Maccabees, do not really constitute a manuscript group inas-
much as they often part company with one another when they do not
bear witness to the original text.
3. The four m groups (m´’`) consist of manuscripts that are either compo-
nents of Greek menologia (collections of lives of the saints arranged by
feast days) or that belong to generally the same textual tradition.
4. From the beginning of 4 Maccabees until 11:4, there exists a group that
I have called m3 and that consists of manuscripts 62 542 and 747c/mg.
Thereafter, 542 becomes aligned with A while 62 and 747c/mg are affili-
ated with m + m1 + m2 (m´’).
5. The original text of manuscript 747 belongs, with that of manuscript 55,
to the q2 group. However, 747 contains a host of corrections or alter-
native readings that are located either between the lines of the text
or in the margins. The majority of these readings are attributable to
a single scribe who did his work after the original scribe penned the
manuscript. The interlinear readings are designated 747c while mar-
ginal readings are designated 747mg, though, as indicated above, both
sets of readings are to be situated within the larger m tradition.
6. The manuscripts and groups that most frequently attest the original
text of 4 Maccabees are A S V L q q1 q2 46 (52 332) 58 340 577 668 690
741 771 773 930. The four m groups, however, are characterized by per-
vasive recensional activity.
2. Daughter versions of Septuagint 4 Maccabees
Text-critically relevant translations of 4 Maccabees are few in number. In

fact, the Syriac version (Sy) is the only one that is fully extant. An eclec-
tic edition based on nine manuscripts was published by R. L. Bensly and
W. E. Barnes.7 A preliminary probe of the degree of affiliation between Sy
and the Greek manuscripts of the book conducted by means of my Web
Application for Textual and Exegetical Research (WATER) indicates that
the highest percentages of agreement are only in the range of about 7.5%.
Thus, for example, Sy agrees with manuscript 741 on 7.55% of the 93 read-
ings attested by one or both of Sy and 741, and Sy agrees with manuscript
668 on 7.54% of the 96 readings attested by one or both of Sy and 668.
Barnes reports that Sy generally agrees with S but seldom with A.8 The
statistics generated by WATER are as follows: S-Sy (89 readings—6.50%);
A-Sy (81 readings—6.01%); V-Sy (45 readings—5.46%). This indicates that
Sy is not very closely affiliated with any other witnesses to the text of this
book. While the Syriac version of 4 Maccabees is, to use Barnes’ terminol-
ogy, a “generally . . . faithful” translation of the Greek,9 it is not the sort of
hyper-literal version that Paul of Tella produced in his seventh century
Syriac translation of Origen’s Hexapla. Instead, it moves “back and forth
along a continuum from the reasonably literal to the freely interpretative.”10
What Hanhart says about the Syriac version of 3 Maccabees in his edition
of the Septuagint of that book is applicable to 4 Maccabees: “In Sy ist der
Text oft frei ausgestaltet, so daß die Textform der Vorlage nur noch schwer
erkennbar ist.”11 Relevant textual evidence from Syriac 4 Maccabees will,
however, be incorporated in the critical edition of this book.
As for other versions, Klauck says: “Nur Hinweise gibt es bislang auf
eine slavische und eine koptische Version.”12 E. Lucchesi reports that only
7 R. L. Bensly and W. E. Barnes, eds., The Fourth Book of Maccabees and Kindred Docu-
ments in Syriac (Cambridge: Cambridge University Press, 1895).
8 Ibid., xiv.
9 Ibid.
10 R. J. V. Hiebert, “Preparing a Critical Edition of IV Maccabees: The Syriac Translation
and Passio Sanctorum Machabaeorum as Witnesses to the Original Greek,” in Interpreting
Translation: Studies on the LXX and Ezekiel in Honour of Johan Lust (BETL 192; eds. F. García
Martínez and M. Vervenne; Leuven: Peeters, 2005), 193–216 (202).
11 R. Hanhart, ed., Septuaginta: Vetus Testamentum graecum, IX, 3: Maccabaeorum liber
III (2nd ed.; Göttingen: Vandenhoeck & Ruprecht, 1980), 9.
12 Klauck, 4. Makkabäerbuch, 679–680.
fragments of a Coptic version have survived,13 and so it will be of very

limited text-critical usefulness. The Slavic version will not be used at all
for the Göttingen edition.
Finally, there is also a fourth century Latin composition known as Pas-
sio Sanctorum Machabaeorum that is based on 4 Maccabees.14 Because it
is a free adaptation rather than a bona fide translation, however, it does
not have much text-critical value.
3. Selected examples involving text-critical decision-making
As is well known, the work of a textual critic involves judicious evalua-

tion of both external and internal evidence in extant textual witnesses. In
other words, in any given context one must make a decision with regard
to which reading is original and which one(s) is/are secondary, based both
on the age and reputation of the available manuscripts and on the assess-
ment of the probabilities with respect to the priority of one reading over
another/others. In the following discussion, I will focus on certain pas-
sages in which this process may be illustrated. I will begin in each case
by citing enough of the texts of Rahlfs’ edition and the NETS version to
provide context for the particular text-critical problem to be discussed.
1. ἀνδραγαθίας vs. καλοκαγαθιας (1:8)
Rahlfs 1:7–9
7 πολλαχόθεν μὲν οὖν καὶ ἀλλαχόθεν ἔχοιμ᾿ ἂν ὑμῖν ἐπιδεῖξαι ὅτι αὐτοκράτωρ ἐστὶν τῶν
παθῶν ὁ λογισμός, 8 πολὺ δὲ πλέον τοῦτο ἀποδείξαιμι ἀπὸ τῆς ἀνδραγαθίας τῶν ὑπὲρ
ἀρετῆς ἀποθανόντων, Ελεαζαρου τε καὶ τῶν ἑπτὰ ἀδελφῶν καὶ τῆς τούτων μητρός. 9
ἅπαντες γὰρ οὗτοι τοὺς ἕως θανάτου πόνους ὑπεριδόντες ἐπεδείξαντο ὅτι περικρατεῖ
τῶν παθῶν ὁ λογισμός.
NETS 1:7–9
7 On the basis of many and diverse considerations I could show you that reason
is absolute ruler of the passions, 8 but I can demonstrate it much better from the
bravery (ἀνδραγαθίας) of those who died for the sake of virtue (ἀρετή): Eleazaros,
13 E. Lucchesi, “Découverte d’une traduction copte du Quatrième livre des Maccabées
(BHG 1006),” Analecta Bollandiana 99 (1981): 302; idem, “Encore trois feuillets coptes du
Quatrième livre des Maccabées,” Écritures et traditions dans la littérature copte (Cahiers de
la Bibliothèque Copte 1; Leuven: Peeters, 1983), 21–22.
14 H. Dörrie, Passio SS. Machabaeorum (Abhandlungen der Gesellschaft der Wissen-
schaften zu Göttingen, Philologisch-historische Klasse 3/22; Göttingen: Vandenhoeck &
Ruprecht, 1938).
the seven brothers and their mother. 9 All of these, in despising sufferings to the
point of death, showed that reason has full control over the passions.
1:8 ἀνδραγαθίας (-θειας V*mg Fritzsche Swete) A L q-q1-q2 m-m3 46 52 58 332 340
577 668 771 Sy Rahlfs Klauck] ανδρειας V* 690; καλοκαγαθιας S 74115
The term ἀνδραγαθία occurs nowhere else in 4 Maccabees, though it does

five times in 1 Maccabees—always in conjunction with πόλεμος (5:56; 8:2;
9:22; 10:15; 16:23)—and once in 2 Maccabees in association with ἀγών.16 In
4 Macc 1:10 the author does state that the men and their mother “died
for the sake of nobility of character” (καλοκαγαθία) and that it is there-
fore appropriate for him to praise them for their virtues (ἀρετῶν). The
term καλοκαγαθία occurs four other times in 4 Maccabees (3:18; 11:22; 13:25;
15:9) but nowhere else in the Septuagint. We therefore have a situation in
4 Macc 1:8 in which the great majority of witnesses attest ἀνδραγαθίας, a
term that does not occur elsewhere in the book but that suits the con-
text well—what with the talk of dying “for the sake of virtue” (v. 8) and
“despising sufferings to the point of death” (v. 9)—while just S and the
eleventh/twelfth century manuscript 741 attest καλοκαγαθιας, a term that
is found in the Septuagint only in 4 Maccabees and that might have come
to replace the original reading ἀνδραγαθίας because of its usage elsewhere
in the book and/or because of a tendency in the transmission history to
enhance the noble character of the martyrs even more. I would, therefore,
go with the majority text and the various listed editions (and against S) in
opting for ἀνδραγαθίας as original.
2. καὶ περὶ τὸ σῶμα vs. ø (1:20)
Rahlfs 1:20
παθῶν δὲ φύσεις εἰσὶν αἱ περιεκτικώταται δύο ἡδονή τε καὶ πόνος· τούτων δὲ ἑκάτερον
καὶ περὶ τὸ σῶμα καὶ περὶ τὴν ψυχὴν πέφυκεν.
NETS 1:20
Of the passions, the two most comprehensive types are pleasure and pain, and
each of these pertains by nature both to the body and to the soul.
1:20 καὶ περὶ τὸ σῶμα V L 370c m-m1-m2-m3-747c 577 690 Rahlfs Klauck] > A S
q(370*)-q1-q2(747*) 46 52 58 332 340 668 741 Fritzsche Swete Breitenstein
15 As is the case in the critical editions of the Göttingen Septuaginta series, throughout
this paper lemma readings include breathing marks and accents, variant readings do not.
16 Apart from the preceding references, ἀνδραγαθία is found in the Septuagint only in
Esth 10:2.
Urs Breitenstein comes to the following conclusion regarding this pas-

sage: “Die von Rahlfs verworfene Lesart der wichtigsten Handschriften
Sinaiticus und Alexandrinus . . . würde ich als lectio difficilior vorziehen.
Sie wäre als Spitze gegen Epikur (wie manches in der Philosophie des
Ps-Ios [Pseudo-Iosephos]) zu verstehen, da nach Epikur alle ἡδοναί,
ἐπιθυμίαι, πόνοι (primär) körperbezogen sind.”17 David deSilva, who trans-
lates the text of S for his commentary in the Septuagint Commentary
Series published by Brill, renders this verse with the shorter reading: “The
two most comprehensive species of passions are pleasure and pain, and
both of these have grown up even around the soul” (τούτων δὲ ἑκάτερος*
[ἑκάτερονc] καὶ περὶ τὴν ψυχὴν πέφυκεν).18 Hanhart, in a communication
to Klauck, however, comments: “Aber es handelt sich doch eher um einen
Abschreibfehler aufgrund eines Homoioteleuton zwischen καί und καί
(Hanhart).”19 Hanhart is likely correct. It should be noted that in 4 Macc
1:28, pleasure (ἡδονή) and pain (πόνος) are, as in the longer reading of v. 20,
associated with both body (σῶμα) and soul (ψυχή): καθάπερ οὖν δυεῖν τοῦ
σώματος καὶ τῆς ψυχῆς φυτῶν ὄντων ἡδονῆς τε καὶ πόνου πολλαὶ τούτων τῶν
φυτῶν εἰσιν παραφυάδες “Just as pleasure and pain are two plants growing
from the body and the soul, so there are many offshoots of these plants.”
In that light, Breitenstein’s suggestion that the shorter reading in 4 Macc
1:20 is to be accepted as the original reading, and that it is to be regarded
as a challenge to Epicurus’s idea that all pleasures, desires, and suffer-
ing are primarily associated with the body, appears to be untenable. Thus
I would favour the longer reading attested by the majority of manuscripts
and included in Rahlfs’ and Klauck’s editions, and go against A S etc.
whose shorter reading is preferred by Fritzsche, Swete, and Breitenstein.
3. ἔργων vs. ετερων (2:9)
Rahlfs 2:8–9
8 αὐτίκα γοῦν τῷ νόμῳ πολιτευόμενος, κἂν φιλάργυρός τις ᾖ, βιάζεται τὸν αὑτοῦ
τρόπον τοῖς δεομένοις δανείζων χωρὶς τόκων καὶ τὸ δάνειον τῶν ἑβδομάδων ἐνστασῶν
χρεοκοπούμενος· 9 κἂν φειδωλός τις ᾖ, ὑπὸ τοῦ νόμου κρατεῖται διὰ τὸν λογισμὸν μήτε
17 U. Breitenstein, Beobachtungen zu Sprache, Stil und Gedankengut des Vierten Makka-
bäerbuchs (Basel / Stuttgart: Schwabe & Co. Verlag, 1976), 136 n. 1.
18 D. A. deSilva, 4 Maccabees: Introduction and Commentary on the Greek Text in Codex
Sinaiticus (Septuagint Commentary Series; eds. S. E. Porter, R. S. Hess, and J. Jarick; Leiden /
Boston: Brill, 2006), 4–5.
19 Klauck, 4. Makkabäerbuch, 692 n. 20c.
ἐπικαρπολογούμενος τοὺς ἀμητοὺς μήτε ἐπιρρωγολογούμενος τοὺς ἀμπελῶνας. ¶ Καὶ

ἐπὶ τῶν ἑτέρων δὲ ἔστιν ἐπιγνῶναι τοῦτο, ὅτι τῶν παθῶν ἐστιν ὁ λογισμὸς κρατῶν·
NETS 2:8–9
8 As soon, indeed, as one adopts a way of life in accordance with the law, even
though a lover of money, one overpowers one’s own bent, lending without inter-
est to the needy and canceling the debt when the seventh year arrives. 9 Even if
one is stingy, one is overcome by the law through reason, neither gleaning the
harvest nor gathering the last of the vineyards’ grapes. ¶ And in all other matters
it is possible to recognize this principle, that reason overcomes the passions.
2:9 ἑτέρων A q-q1-610 46 52 58 340 668 741 771 Fritzsche Swete Rahlfs Klauck] εργων
S V L 610-q2 m-640*-m1-m2-m3 577 690 Sy; > 640*
David deSilva, who translates S, renders the relevant sentence in the lat-
ter part of v. 9: “And on the evidence of deeds, we recognize this—that
reason restrains the passions.”20 The external evidence of the textual wit-
nesses for both εργων and ἑτέρων is strong, so the original reading must
be reconstructed on the basis of internal evidence. It seems likely that
εργων represents a secondary development in the textual history in this
context, whereby the general reference to other (ἑτέρων) manifestations
of living “in accordance with the law” (v. 8) would have been replaced
by the more specific terminology of deeds (εργων) that embody that law.
Furthermore, given the orthographic similarity of the two words (espe-
cially their first and last parts), it is not hard to envision a scenario early
in the textual history in which textual corruption could have given rise
to the alternate reading. So in this case I would side with the editions of
Fritzsche, Swete, Rahlfs, and Klauck, which feature the minority reading
of A etc., and against the majority reading of S V etc.
4. εἴποιεν ἄν τινες vs. ειποι αν τις (1:5) / εἴποι τις ἄν vs. αν τις ειποι (2:24)
Rahlfs 1:5–6
5 πῶς οὖν, ἴσως εἴποιεν ἄν τινες, εἰ τῶν παθῶν ὁ λογισμὸς κρατεῖ, λήθης καὶ ἀγνοίας οὐ
δεσπόζει; γελοῖον ἐπιχειροῦντες λέγειν. 6 οὐ γὰρ τῶν αὑτοῦ παθῶν ὁ λογισμὸς κρατεῖ,
ἀλλὰ τῶν τῆς δικαιοσύνης καὶ ἀνδρείας καὶ σωφροσύνης ἐναντίων, καὶ τούτων οὐχ ὥστε
αὐτὰ καταλῦσαι, ἀλλ᾿ ὥστε αὐτοῖς μὴ εἶξαι.
NETS 1:5–6
5 Some might perhaps ask, “How then, if reason overcomes the passions, does it
not master forgetfulness and ignorance?” Their attempt at argument is ridiculous.
20 deSilva, 4 Maccabees, 6–7.

6 For reason does not overcome its own passions but those opposed to justice,
courage and self-control, and it overcomes these not so that it destroys them but
so that one does not give way to them.
1:5 εἴποιεν ἄν τινες A q2 m-m1-m2-m3 58 332 340 577 690 Fritzsche Swete Rahlfs
Klauck] ειποι (ειπει 491; ειποιεν S* 534 340) αν τις S V L q-q1 46 52 340 668 741
771 Sy
Rahlfs 2:24–3:1
2:24 Πῶς οὖν, εἴποι τις ἄν, εἰ τῶν παθῶν δεσπότης ἐστὶν ὁ λογισμός, λήθης καὶ ἀγνοίας
οὐ κρατεῖ; 3:1 ἔστιν δὲ κομιδῇ γελοῖος ὁ λόγος· οὐ γὰρ τῶν ἑαυτοῦ παθῶν ὁ λογισμὸς
ἐπικρατεῖν φαίνεται, ἀλλὰ τῶν σωματικῶν.
NETS 2:24–3:1
2:24 How is it, then, someone may ask, if reason is master of the passions, that it
does not overcome forgetfulness and ignorance? 3:1 But this argument is entirely
ridiculous, for it is apparent that reason prevails not over its own passions but
over those of the body.
2:24 εἴποι (ειπη 370) τις ἄν A Sc V q-q1-q2 46 52 340 668 741 771 Fritzsche Swete
Rahlfs Klauck] ειποι αν τις L; ειποιτε S*; αν τις ειποι m-m1-m2-m3 58 577 690;
αν ειποι τις 316-473; αν τινες ειποιεν 397-467-686 577; ‫ܐܝܬ ܠܟܘܢ ܠܡܐܡܪ‬
Sy(c var)
This example involves two passages—4 Macc 1:5–6 and 2:24–3:1—that

express similar sentiments and wording. In 1:5, where the manuscript
support for both the singular and the plural readings is quite strong, the
concluding clause in 1:5—γελοῖον ἐπιχειροῦντες λέγειν “Their attempt at
argument is ridiculous”—with its plural participle is the determining fac-
tor in favour of the plural reading at the beginning of the verse attested by
A etc. and adopted in the editions of Fritzsche, Swete, Rahlfs, and Klauck.
In 2:24, however, a singular reading has overwhelming manuscript sup-
port, and the word sequence εἴποι τις ἄν attested by A Sc V etc. and fea-
tured in the same four editions is to be preferred.
5. πρός vs. ως (4:2)
Rahlfs 4:2
ὅθεν ἥκων πρὸς Ἀπολλώνιον τὸν Συρίας τε καὶ Φοινίκης καὶ Κιλικίας στρατηγὸν
ἔλεγεν . . .
NETS 4:2
So he came to Apollonius, governor of Syria, Phoenicia and Cilicia, and said . . .
4:2 πρός A L 610-q2 m-m1 46 52 58 340 577 668 690 741 773 Sy Swete Rahlfs Klauck]
ως S V q-q1-610 m3-747c Fritzsche
David deSilva translates the S reading as follows: “For this reason, he went
when Apollonius [was] the governor of Syria, Phoenicia, and Cilica (sic),
and said. . . .”21 He comments that ως requires an implied verb “was,” as in
“when Apollonius [was] the governor.”22 But πρός would appear to make
better sense in this context. I wonder if a corruption developed early as
a result of attraction to the ending of the preceding word ἥκων that was
then mistakenly followed by ως instead of πρός.
6. ἱερέων vs. γεραιων (4:9)
Rahlfs 4:9–10
9 τῶν δὲ ἱερέων μετὰ γυναικῶν καὶ παιδίων ἐν τῷ ἱερῷ ἱκετευσάντων τὸν θεὸν
ὑπερασπίσαι τοῦ ἱεροῦ καταφρονουμένου τόπου 10 ἀνιόντος τε μετὰ καθωπλισμένης
τῆς στρατιᾶς τοῦ Ἀπολλωνίου πρὸς τὴν τῶν χρημάτων ἁρπαγὴν οὐρανόθεν ἔφιπποι
προυφάνησαν ἄγγελοι περιαστράπτοντες τοῖς ὅπλοις καὶ πολὺν αὐτοῖς φόβον τε καὶ
τρόμον ἐνιέντες.
NETS 4:9–10
9 While the priests, with women and children, were imploring God in the temple
to shield the holy place that was being treated with contempt 10 and while Apol-
lonius was going up with his armed forces to seize the funds, angels on horseback
appeared from heaven with lightning flashing from their weapons, instilling in
them great fear and trembling.
4:9 ἱερέων A q-q1-q2 m 46 52 58 340 577 668 771 Fritzsche Swete Rahlfs Klauck]
γεραιων (γερεων 491 728) S L 542-747mg 741; γηραιων (γιρ. V*; -ρεων m1) V m1-62
690; ‫ ̈ܣܒܐ‬Sy
This case presents an interesting situation. On the one hand, the textual
support for a reference to elders γεραιων/γηραιων/‫ ̈ܣܒܐ‬is quite strong. Fur-
thermore, it might be argued that such a reading would constitute a lectio
difficilior, given that priests might be expected to be in the temple “implor-
ing God” rather than elders. On the other hand, the difference between
γεραιων—or more specifically the itacized version of it (γερεων)—and
ἱερέων is one letter, and in uncial characters the two terms could appear to
be even more similar than in minuscules: ΙΕΡΕΩΝ/ΓΕΡΕΩΝ. So Fritzsche,
Swete, Rahlfs, and Klauck would presumably have concluded that an early
scribal error resulted in a change from priests in the temple to elders. That
palaeographical factor would, therefore, appear to be determinative. How
the reference to elders would have been interpreted once γεραιων came to
21 Ibid., 12–13.
22 Ibid., 13 n. 3.
be part of the textual history, one can only speculate. Could it have been
that this reading was received favourably by some who had differences
with the priestly hierarchy, especially with those who, like Jason in the
time of Antiochus IV (4 Macc 4:15–20), were seen to be complicit with
Gentile authorities in exercising authority over the populace?
7. σφονδυλ- vs. σπονδυλ- (10:8 and 11:18)
Rahlfs 10:8
καὶ εὐθέως ἦγον ἐπὶ τὸν τροχόν, περὶ ὃν ἐκ σπονδύλων ἐκμελιζόμενος
NETS 10:8
Then they immediately brought him [the third brother] to the wheel. On it he
was disjointed at his backbone.
10:8 σφονδύλων A 728 q-452-q1-610-q2 317-542-682* 340 577 668 771 Swete] σφονδυλου
773; σπονδυλων S L-728 452-610 m-317 682*-m1-m2-62 46 52 58 690 741 Fritzsche
Rahlfs
Rahlfs 11:18
ἐφ᾿ οὗ κατατεινόμενος ἐπιμελῶς καὶ ἐκσπονδυλιζόμενος
NETS 11:18
He [the sixth brother] was stretched tight upon it with great diligence; his back
was broken
11:18 ἐκσφονδυλιζόμενος (εκσφενδ. 457 592* 699; εκφονδ. 597*; -διλ. 71) A-542 491
q-452-q1-q2 m-316 397 467 473 586 640 686 789-m1 340 668c Fritzsche Swete] εκσπονδ.
(εκπονδ. 62) S 452 62-316-397-467-473-586-640-686-789-m2 46 52 58 668* 690
741 Rahlfs; σφονδ. L-491 577; μηεφονδυλιμενον (sic) 771
There is relatively strong manuscript support for both σφονδυλ- and

σπονδυλ- in 10:8 and for εκσφονδυλ- and εκσπονδυλ- in 11:18. A search in
Thesaurus linguae graecae reveals that the σφονδυλ- root occurs some
774 times in that database, while σπονδυλ- occurs 936 times.23 As Henry
St. J. Thackeray points out, Attic Greek attests forms in σφ-, whereas Ionic
and some κοινή writers use σπ-.24 In the only other place in the Septuagint
23 A Thesaurus linguae graecae (TLG) search yields two occurrences of the εκσφονδυλ-
root, and three of εκσπονδυλ-.
24 H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Sep-
tuagint (Cambridge, Cambridge University Press, 1909), 106 (§7, 17). The LSJ entry for this
term is σφόνδυλος, but it is followed by the note “Ion. and later Greek σπόνδυλος” (The
Online Liddell-Scott-Jones Greek-English Lexicon [OLSJ], σφόνδυλος).
in which any form of the σφονδυλ-/σπονδυλ- root is found (Lev 5:8), the
editions of Wevers, Swete, and Rahlfs all have σφονδυλ-:
σφονδύλου 802 Wevers Swete Rahlfs] σπονδυλου (-δηλου 528 125 646) 58-72-
707(mg)(vid) C´’ 118´-537 125 730 74-76 527 318 628´ 18 55c 59 424 646´
In 4 Maccabees, therefore, it would seem to be easier to account for the
more frequently occurring and later form σπονδυλ- as a secondary replace-
ment of the original σφονδυλ- than vice versa. In 4 Maccabees, it appears
that Rahlfs relied too much on the testimony of Sinaiticus when he opted
for the σπονδυλ- form.
8. βασανισμόν vs. + παραιτεισθαι (11:2)
Rahlfs 11:2
Οὐ μέλλω, τύραννε, πρὸς τὸν ὑπὲρ τῆς ἀρετῆς βασανισμὸν παραιτεῖσθαι
NETS 11:2
Tyrant, I am not about to beg to be excused from torture for the sake of virtue.
Klauck 11:2
“Ich will mich, Tyrann, zu den Foltern um der Tugund willen nicht erst auffordern
lassen.”25
Scarpat 11:2
“Non mi rifiuterò, o tiranno, all tortura inflitta a causa della virtù.”26
deSilva 11:2
Οὐ μέλλω, τύραννε, πρὸς τὸν ὑπὲρ τῆς ἀρετῆς βασανισμόν
“I do not hold back, tyrant, in regard to torture for the sake of moral
excellence.”27
11:2 βασανισμόν (βανισμον 457) S L q-q1 m-607-m2-m3-747c 46 52 58 773 Fritzsche]
θανατον 340 668 771; + παραιτεισθαι A 741 Swete Rahlfs Klauck; + δειλιαν (διλ.
340) q2(747*) 607 340 577 668 690 771; + ‫“( ܠܡܐܬܐ‬to come”) Sy
The presence of the verb παραιτεισθαι seems, on the basis of both external
and internal evidence, to be secondary. With regard to external evidence,
only A 741 attest it. As for internal evidence, it would appear that the
remaining manuscripts bear witness to the conviction that μέλλω requires
an infinitive complement. But that conviction seems to be predicated on
25 Klauck, 4. Makkabäerbuch, 731.

26 G. Scarpat, ed. and trans., Quarto libro dei Maccabei (Biblica testi e studi 9; Brescia:
Paideia, 2006), 305.
27 deSilva, 4 Maccabees, 36–37.
the notion that in this context μέλλω means “be about to.” I would argue,
however, that here it denotes “delay, put off,”28 and that the verse means:
Tyrant, I do not put off torture for the sake of virtue.
9. ὀροφοφοιτοῦντα vs. οροφοιτουντα and οροφοκοιτουντα (14:15)
Rahlfs 14:13–16
13 θεωρεῖτε δὲ πῶς πολύπλοκός ἐστιν ἡ τῆς φιλοτεκνίας στοργὴ ἕλκουσα πάντα
πρὸς τὴν τῶν σπλάγχνων συμπάθειαν, 14 ὅπου γε καὶ τὰ ἄλογα ζῷα ὁμοίαν τὴν εἰς
τὰ ἐξ αὐτῶν γεννώμενα συμπάθειαν καὶ στοργὴν ἔχει τοῖς ἀνθρώποις. 15 καὶ γὰρ τῶν
πετεινῶν τὰ μὲν ἥμερα κατὰ τὰς οἰκίας ὀροφοιτοῦντα προασπίζει τῶν νεοττῶν, 16 τὰ
δὲ κατὰ κορυφὰς ὀρέων καὶ φαράγγων ἀπορρῶγας καὶ δένδρων ὀπὰς καὶ τὰς τούτων
ἄκρας ἐννοσσοποιησάμενα ἀποτίκτει καὶ τὸν προσιόντα κωλύει·
NETS 14:13–16
13 Consider how complex is the affection of a mother’s love for her children,
channeling all her feeling into a sympathy rooted deep within. 14 Even unreason-
ing animals show sympathy and affection for their offspring like that of human
beings. 15 For among birds, the tame ones that roam the mountains protect their
young on housetops, 16 and others, building their nests on mountain peaks, in
clefts of ravines, in holes of trees or on their tops, give birth to their young and
fend off the intruder.
14:15 ὀροφοφοιτοῦντα 741(|)] ‫ܥܐܠܐ‬ ̇ ‫ ܬܚܝܬ ܡܛܠܠܐ‬Sy; οροφοκοιτουντα (conjecture)
Bekker, Fritzsche, Deißmann, Klauck; οροφοιτουντα (-φητ. 71; -φυτ. S 457-472-586-
591-592-594-595-607-617-639-640-656-683-699-713-778-789 340 771; -φυτουντο
682) A S L q-q2 m-m1-m3 46 52 58 340 668 771 Swete Rahlfs; οροφοιτωντα V
577(-φυτ.); φοιτωντα m2
There are a number of problems in interpreting v. 15, specifically with

respect to the dependent clause and subsequent main clause τὰ μὲν ἥμερα
κατὰ τὰς οἰκίας ὀροφοιτοῦντα προασπίζει τῶν νεοττῶν (Rahlfs) “the tame ones
that roam the mountains protect their young on housetops” (NETS). The
association of tame birds with houses is logical enough, but what kind of a
connection is intended in the phrase that is introduced by the preposition
κατά? Furthermore, how does roaming the mountains (οροφοιτουντα) fit
with this picture? This brings us to v. 16, where other birds are described
building nests “on mountain peaks” (κατὰ κορυφὰς ὀρέων) as well as “in
clefts of ravines, in holes of trees or on their tops” (φαράγγων ἀπορρῶγας
καὶ δένδρων ὀπὰς καὶ τὰς τούτων ἄκρας). Although it is not stated explicitly,
28 Included in the OLSJ entry for μέλλω are the words “delay, put off . . . . inf. is freq. omit-
ted.” Among the cited examples are μακρὰ μέλλεται it is a long delay (Sophocles, Oed. col.
219), and μὴ μέλλωμεν ἔτι let us delay no longer (Plato, Leg. 712b).
the birds described in v. 16 would appear to be wild ones in contrast to

the specifically-designated tame ones (ἥμερα) in v. 15. As such, their situ-
ation out in the wilds would also appropriately contrast with the location
among houses of the tame birds. In that light, again one asks, how do
mountains suit their situation?
It is undoubtedly such questions that have given rise to the variety of
alternatives for the participle in v. 15.
ὀροφοφοιτοῦντα 741(|)29
‫ܥܐܠܐ‬̇ ‫ ܬܚܝܬ ܡܛܠܠܐ‬Sy
οροφοκοιτουντα (conjecture) Bekker, Fritzsche, Deißmann, Klauck
οροφοιτουντα (-φητ. 71; -φυτ. S 457-472-586-591-592-594-595-607-617-639-640-656-
683-699-713-778-789 340 771; -φυτουντο 682) A S L q-q2 m-m1-m3 46 52 58 340
668 771 Swete Rahlfs
οροφοιτωντα V 577(-φυτ.)
φοιτωντα m2
Leaving aside for a moment the reading of 741 that I have chosen as my
lemma, let us consider first the conjecture, οροφοκοιτουντα, which has
been adopted by a number of textual critics over the years.30 This is a
compound consisting presumably of ὄροφος “roof ” and κοιτέω, which is
not attested but would likely be a verbal cognate of κοίτη “nest” (of a
bird).31 Klauck translates 14:15 with the conjecture (οροφοκοιτουντα) as fol-
lows: “Nehmen wir nur die Vögel. Die zahmen beschirmen ihre Jungen,
indem sie unter den Dächern der Haüser nisten.”32 Commenting on this
verse, Klauck suggests that this reading appears to be supported by the
wording of the Syriac version ‫ܥܐܠܐ‬ ̇ ‫ ܬܚܝܬ ܡܛܠܠܐ‬enter under the roof and
also by manscript 741 ὀροφοφοιτοῦντα roam the roofs. He does, however,
make the following acknowledgement: “Allerdings ist die Vokabel ander-
wärts nicht belegt. Rahlfs liest mit den Handschriften ὀροφοιτοῦντα (nach
Liddell-Scott s. v. ein Beleg bei Hesych.), was ‘die Berge durchstreifend’
bedeutet. Will man es beibehalten, müßte man so erklären: Die zahmen
Vögel in V. 15 benutzen die Haüser als ‘Berge’, wie die wilden Vögel in
29 Note that the word break occurs after iota in manuscript 741: οροφοφοι|τουντα.
30 I. Bekker, Flavii Josephi Opera Omnia, vol. 6 (Leipzig: Teubner, 1856), 295; Fritzsche,
ΜΑΚΚΑΒΑΙΩΝ ΤΕΤΑΡΤΟΣ, 378; A. Deißmann, “Das vierte Makkabäerbuch,” in Die Apokry-
phen und Pseudepigraphen des Alten Testaments, vol. 2 (ed. E. Kautzsch; Hildesheim: Georg
Olms, 1900), 169; Klauck, 4. Makkabäerbuch, 742.
31 OLSJ, ὄροφος, κοίτη.
32 Klauck, 4. Makkabäerbuch, 742.
V. 16 die echten Berge (Hanhart).”33 As Klauck points out, the term attested
by Rahlfs and the majority of manuscript witnesses—ὀροφοιτέω—is not
found elsewhere in extant Greek literature. It is undoubtedly cognate to
ὀρειφοιτέω “roam the mountains,”34 which appears in a citation from the
first century B.C.E. grammarian Sostratus in the commentary on Homer’s
Odyssey by Eustathius, the twelfth century bishop of Thessalonica.35 The
form οροφοιτωντα is presumably based on the root ὀροφοιτάω which, in
LSJ/OLSJ, is listed as an equivalent to ὀρειφοιτέω. The form φοιτωντα is
based on the frequently attested simplex verb φοιτάω “go to and fro,” “on
the wing” (of birds), “roam wildly about.”36 The latter two readings in this
verse—οροφοιτωντα and φοιτωντα—can quite readily be accounted for as
variants of the majority reading, οροφοιτουντα.
This brings us back to my lemma reading, ὀροφοφοιτοῦντα, which is
attested in only the eleventh/twelfth century manuscript 741, one of my
codices mixti. This term occurs nowhere else in extant Greek literature, but
its occurrence in manuscript 741 in 4 Macc 14:15 means that there is more
textual evidence for it than there is for the conjecture, οροφοκοιτουντα,
that is favoured by Bekker, Fritzsche, Deißmann, and Klauck. Like that
conjecture, the first element in this compound word is ὄροφος “roof.” The
second element, however, is φοιτέω, an alternate spelling for φοιτάω.37 In
the entry in OLSJ for ὀροφοιτάω, one reads the following: “= ὀρειφοιτέω,
LXX 4 Ma. 14.15 (v.l. [varia lectio] -οῦντα, but perh. ὀροφοφοιτ- shd. be
read).”38 In his Italian edition of 4 Maccabees, Giuseppe Scarpat com-
ments that οροφοιτουντα “è risultato di un’aplologia *ὀροφο-φοιτέω ‘fre-
quento i tetti’.”39 The suggestion that the original reading ὀροφοφοιτοῦντα
could have given rise to the majority reading οροφοιτουντα is quite plausi-
ble.40 That possibility seems to me to represent a more logical progression
than the reverse one since, as noted above, tame birds are the subject of
the participle in question, and their association with the roofs of houses
makes more sense than the prospect of them roaming the mountains. At
33 Ibid., n. 14a.
34 OLSJ, ὀρειφοιτέω.
35 Eustathius, Commentarii ad Homeri Odysseam 1.390.9.
36 OLSJ, φοιτάω.
37 Ibid.
38 OLSJ, ὀροφοιτάω.
39 Scarpat, Quarto libro dei Maccabei, 353.
40 Oddly enough, despite this comment, Scarpat retains the majority reading
οροφοιτουντα “roam the mountains” in his Greek text of 14:15, and he translates it as
“s’annidano sotto il tetto” nest under the roof as though the Greek text were the conjecture
οροφοκοιτουντα nest in/under the roofs (Scarpat, Quarto libro dei Maccabei, 346–347).
subsequent stages in the textual history, οροφοιτωντα and φοιτωντα could

have been derived from the majority reading. In this kind of a situation, it
would appear to be more prudent to settle on an attested reading than to
posit one that has no manuscript support, as is the case with the conjec-
̇ ‫ܬܚܝܬ ܡܛܠܠܐ‬
ture. Finally, in response to Klauck’s contention that ‫ܥܐܠܐ‬
in the Syriac version seems to support the conjecture οροφοκοιτουντα,
I would argue that one can make at least as good a case for associating
the Syriac reading with ὀροφοφοιτοῦντα of manuscript 741 as one can for
linking it with οροφοκοιτουντα.
10. ἐπιλογίσασθαι vs. επιλογισασθε (16:5)
Rahlfs 16:5
Καὶ γὰρ τοῦτο ἐπιλογίσασθε
NETS 16:5
Consider this also
16:5 ἐπιλογίσασθαι A S 534 q2(747*) 340 577 741 771 Swete] pr εστιν q-q1; επιλογεισθαι
542; επιλογισασθε L-534 46 52 668 Fritzsche Rahlfs; διαλογισασθαι 682 713; δει
λογισασθαι m1; διαλογισασθε m-682 713-m2-62-747c; + εστιν 577; + χρη 741
The complicating factor in evaluating various options in this case is the

fact that the development of alternative readings could well have been
the result of etacism (αι > ε as a termination or vice versa). I maintain,
however, that my lemma reading ἐπιλογίσασθαι is to be preferred as origi-
nal over Rahlfs’ lemma επιλογισασθε. The manuscript support for the for-
mer is stronger, and when one factors in the variant that involves the
addition of εστιν prior to my lemma (q-q1) and possibly also the readings
διαλογισασθαι (682 713) and δει λογισασθαι (m1), the case for an infinitive
rather than an imperative construction here is even more compelling.
One can more readily account for the progression from the infinitive
ἐπιλογίσασθαι and secondarily the longer alternative εστιν επιλογισασθαι
to the imperative επιλογισασθε than vice versa. My lemma could, in fact,
be regarded as the lectio difficilior. In the only other place in 4 Maccabees
where the verb ἐπιλογίζομαι occurs, the construction is Ἔστιν . . . τοῦτο . . .
ἐπιλογίσασθαι “this can be explained” (3:6). So in 16:5, I would suggest that
the clause Καὶ γὰρ τοῦτο ἐπιλογίσασθαι is to be translated, For there is also
this to consider.
11. ὡς οὐδέ vs. ουδ ως (16:12)
Rahlfs 16:12
Ἀλλὰ τούτῳ τῷ θρήνῳ οὐδένα ὠλοφύρετο ἡ ἱερὰ καὶ θεοσεβὴς μήτηρ οὐδ᾿ ἵνα μὴ
ἀποθάνωσιν ἀπέτρεπεν αὐτῶν τινα οὐδ᾿ ὡς ἀποθνῃσκόντων ἐλυπήθη
NETS 16:12
Yet the holy, God-fearing mother bewailed none of them with this lament and
neither attempted to dissuade any of them from dying, nor, as they died, did she
grieve.
16:12 ὡς οὐδέ (ουδ q2 577; ουδεν 71) S V L q-q1-q2 m2 46 52 340 577 668 771] ως γαρ
ουδε m-m1-62; ουδ ως A-542 741 Fritzsche Swete Rahlfs
I think I understand why Fritzsche and Rahlfs chose ουδ ως as their lemma
(Swete typically follows A as a matter of course). I suspect that they opted
for ουδ ως because it maintains stylistic/syntactical symmetry with the
preceding clause that begins οὐδ᾽ ἵνα. Nevertheless, given the strength of
the support for ὡς οὐδέ, I question why the earliest extant witness (S) and
all the manuscript groups would choose to break that proposed original
symmetry by introducing an expression that is attested nowhere else in
the Septuagint (though there are plenty of examples of ὡς οὐδέ outside
the Septuagint). This is why I conclude that it was the other way round:
ὡς οὐδέ was original and ουδ ως secondary. My proposal for the original
text thus creates a sense of progression in the statement about the mother
by the author of 4 Maccabees, inasmuch as οὐδέ is construed to be more
adverbial than conjunctival: οὐδ᾿ ἵνα μὴ ἀποθάνωσιν ἀπέτρεπεν αὐτῶν τινα
ὡς οὐδὲ ἀποθνῃσκόντων ἐλυπήθη and she did not attempt to dissuade any of
them from dying, just as she did not even grieve while they were dying.41
4. Conclusion
The preceding discussion has dealt with examples of kinds of problems

that I encounter in assessing the manuscript evidence while attempting to
reconstruct the original Septuagint text of 4 Maccabees. Clearly each case
must be decided on its own merits. And, as indicated above, evaluating
both the available external and internal evidence is required.
41 My thanks to Albert Pietersma for his suggestion in a private communication (28 May
2011) that the sense of progression seems to be underscored when οὐδέ is understood to
be adverbial.
The Relationship between the LXX Versions
of Proverbs and Job
Johann Cook
1. Introduction
1.1. The Problem
On the issue of the relationship between the Septuagint versions of Prov-
erbs and Job scholars have deviating opinions. Gerleman expressed the
view that the same translator was responsible for both LXX Proverbs and
OG Job.1 Gammie2 tested Gerleman’s thesis and reached the opposite con-
clusion. Other scholars also dealt with aspects of the relationship between
these two translators. Heater3 studied the LXX of Job and found many
examples of intra-textual readings, what he called anaphoric translation
technique, in LXX Job taken from other parts of the Septuagint. However,
he did not directly address the issue at stake. In his series of articles,
Orlinsky 4 also analysed various characteristics of the Greek translation
of Job, without explicitly dealing with the relationship between the men-
tioned translators. Cox in passing refers to these translators and suggests
that they might come from the same circle of translators.5 In a paper deliv-
ered at IOSCS in New Orleans 2009 based on content analysis of these two
units, I came to the conclusion that these versions originate from different
translators and contexts.6 It is the intention of this article to contribute
towards this discussion.
1 G. Gerleman, Studies in the Septuagint I, Book of Job (LUÅ Bd 43. Nr 2; Lund: Gleerup,
1946), 14–17.
2 J. G. Gammie, “The Septuagint of Job: Its Poetic Style and Relationship to the Septu-
agint of Proverbs,” CBQ 49/1 (1987): 14–31.
3 H. Heater, A Septuagint Translation Technique in the Book of Job (CBQMS 11; Washing-
ton DC: The Catholic Biblical Association of America, 1982).
4 See especially H. M. Orlinsky, “The Character of the Septuagint Translation of the
Book of Job,” HUCA 39 (1958): 229–271.
5 C. E. Cox, “The Historical, Social, and Literary Context of Old Greek Job,” in XII Con-
gress of the International Organization for Septuagint and Cognate Studies, Leiden 2004
(SBLSCS 54; ed. M. K. H. Peters; Atlanta: Society of Biblical Literature, 2006), 116.
6 J. Cook, “Were the Septuagint Versions of Job and Proverbs Translated by the Same
Person?” Hebrew Studies 51 (2010): 129–156.
146 johann cook
1.2. Applicable Criteria
Appropriate criteria need to be formulated in order to address this issue.
As is well-known, there are two sets of criteria: firstly, linguistic ones and
more specifically on the micro level, lexically based criteria, as well as, on
the macro level, syntactical and stylistic issues; secondly, arguments from
content analysis. One would naturally expect the same translator to show
correspondence on these two levels. In this paper, I will deal with the first
criterion; since I have addressed the second criterion at the mentioned
meeting of IOSCS.
Some work has been done in this regard on the OG of Job by Cox.7 I
will take a cue from this research and in this paper I will concentrate on
a small number of linguistic criteria with a focus on lexical items. These
criteria will be chosen randomly. Firstly, I will research the way particles
have been applied in the two translated units; secondly, I will deal with
the manner in which certain lexical items are utilized in these two units.
Here I will endeavour to select those items that I deem typical of each
of the translators and try to determine how such items function in these
units.
1.3. Textual Basis
Since the Old Greek of LXX Proverbs has not yet been determined sys-
tematically, the scholarly edition of Rahlfs8 must be utilised. Proverbs has
been allocated to Peter Gentry to prepare in the Göttingen edition. The
research into Job is based on the critical edition by Joseph Ziegler.9 In this
regard, see also Pietersma10 in his review of Ziegler and Gentry11 on the
asterisked materials.
7 C. E. Cox, “Tying it All Together: The Use of Particles in Old Greek Job,” BIOSCS 38
(2005): 41–54.
8 A. Rahlfs, Septuaginta. Id est Vetus Testamentum graeca iuxta LXX interpretes (Stutt-
gart: Deutsche Bibelgesellschaft, 1979).
9 J. Ziegler, Septuaginta: Vetus Testamentum Graecum. Auctoritate Scientiarum Gottin-
gensis editum XI,4: Job (Göttingen: Vandenhoeck & Ruprecht, 1982).
10 A. Pietersma, Review of Job. Septuaginta. Vetus Testamentum Graecum, II/4 ed. J. Zie-
gler, JBL 104 (1985): 305–311.
11 P. J. Gentry, The Asterisked Materials in the Greek Job (SCS 38; Atlanta: Scholars Press,
1995).
the relationship between the lxx versions 147
1.4. Conclusion
There is a consensus that both LXX Proverbs and Job are less faithfully
translated units.12 However, there are also differences between these trans-
lations. A pertinent one is that, whereas LXX Proverbs is fundamentally an
expansive text, the OG of Job, on the contrary, is a shortened, abbreviated
text. However, this does not mean that the former contains no minuses at
all, or that the OG of Job has no pluses. As a matter of fact, LXX Job con-
tains two (three?) important additions, one in ch. 2 vv. 9a–f, the diatribe
on the wife of Job and the second, ch. 42:17a–e. Another conspicuous dif-
ference is the change in the order of chapters13 in LXX Proverbs from ch.
24 onwards, which I have ascribed to the Greek translator.14 These units
are thus unique in the corpus of Septuagint translations.15
In one crucial respect these two units agree, namely in their size. The
number of words are comparable. LXX Job has 13561 words, which include
the Theodotionic text, and Proverbs totals 11164 words. Although there is
a subjective element in this counting it may be used as a basis for com-
parison.
2. The Micro Level (Lexical Items)
There can be no doubt that both translators of Proverbs and Job were
excellently educated in the Jewish and Greek cultures. They both were, as
Cox16 said about the OG Job, a work of good literary quality. As far as LXX
Proverbs is concerned, I identified the following significant pattern. This
12 E. Tov and B. G. Wright, “Computer-assisted Study of the Criteria for Assessing the
Literalness of Translation Units in the LXX,” Textus 12 (1985): 149–183. See also the mono-
graph by Tov, The Text-critical Use of the Septuagint in Biblical Research (JBS 3; Jerusalem:
Simor, 1981–1998).
13 See E. Tov, “Recensional Differences Between the Masoretic Text and the Septuagint
of Proverbs,” in Of Scribes and Scrolls, Studies on the Hebrew Bible, Intertestamental Juda-
ism, and Christian Origins presented to John Strugnell (eds. H. W. Attridge et al.; Lanham:
University Press of America, 1990), 43–56.
14 J. Cook, “The Greek of Proverbs—Evidence of a Recensionally Deviating Hebrew
Text?” in Emanuel: Studies in Hebrew Bible, Septuagint, and Dead Sea Scrolls in Honor of
Emanuel Tov (eds. S. M. Paul, R. A. Kraft, L. H. Schiffman and W. W. Fields; Leiden: Brill,
2003), 605–618.
15 C. E. Cox (“Job,” in A New English Translation of the Septuagint. A New Translation
of the Septuagint and the Other Greek Translations traditionally Included Under That Title
[eds. A. Pietersma and B. G. Wright; Oxford / London: Oxford University Press, 2007], 667)
thinks “OG Job is one of a kind in the Septuagint corpus.”
16 Ibid.
148 johann cook
translator clearly was a creative stylist with an exceptional knowledge of

Jewish and Greek culture which is observed on various niveaus. As far as
his creative approach is concerned, he made ample use of hapax legom-
ena which he borrowed from the Greek world. There are also a number
of neologisms;17 he moreover, applied a rather large number of lexemes
unique to Proverbs.18 As to be expected, he interprets extensively in some
instances. I have not made the same extensive analyses of LXX Job. But
the research by Cox points in the same general direction.
2.1. Individual Lexemes
A word on the databases that I used is in order. I utilized the Thesaurus Lin-
gua Graecae and combined it with Hatch & Redpath (HR). In connection
with LXX Job, I double checked these sources against Ziegler’s edition. I
also partially made use of Wordsmith which is a corpus concordance tool
developed by Mike Scott (www.lexically.net). It compares word frequen-
cies from different corpora in order to determine which words are most
different between the two. It calculates a keyness score, which measures
the degree of difference in the two corpora, based on the assumption that
words with identical reference frequencies would have a keyness of 1. The
higher the keyness, the bigger the chance that an observed difference is
significant. The analyst selects the level of chance that a trivial finding is
mistakenly regarded as significant, which is by default one in a million in
Wordsmith.19
2.1.1. Particles
Cox20 has dealt with this issue exhaustively. He deems these particles
important: “The use of such ‘little words’ provides nuance, continuity,
change of direction, qualification, color, and emotion, to what we say or
write.” However, the intention of this paper is different from Cox’s who
was interested in the way the translator used particles to tie the OG of
17 J. Cook, “The translator(s) of LXX Proverbs,” TC—a Journal of Biblical Textual Criti-
cism 7 (2002): 1–50.
18 J. Cook, The Septuagint of Proverbs Jewish and/or Hellenistic Proverbs. Concerning the
Hellenistic Colouring of LXX Proverbs (VTSup 69; Leiden: Brill, 1997), 335–342.
19 I was introduced to Wordsmith by a corpus linguist, Bertus van Rooy, from the Uni-
versity of North-West, Vaaltriangle campus. He provided this description and some of the
data utilised below.
20 Cox, “Tying it All Together,” 41–54. See also J. D. Denniston, The Greek Particles
(Oxford: Clarendon Press, 1934).
Job together. As stated above, the intention is to compare LXX Proverbs

and Job.
There are prominent correspondences as well as differences between
the way these items were applied by the individual translators. I will deal
with differences first of all.
2.1.1.1. Differences
The first category is where a specific particle was used by either the trans-
lator of Proverbs and Job and not by both.
The particle ἄρα, as an inferential particle, namely “thus, accordingly,”21
is used 40 times in LXX, not at all in Proverbs and 11 times in Job. Another
inferential particle, μέντοι, “to be sure,” that is used 4 times in LXX Prov-
erbs, in Prov 5:4; 16:25; 22:9a and 26:12, does not appear in Job at all.
The coordinating conjunction ἀτάρ is used uniquely in the LXX. It
occurs only twice, namely in Job 6:21 (“But yet you too trod on me, without
mercy; so now that you have seen my hurt, be afraid!”)22 and 7:11 (“That
said, however, I will not be sparing with my mouth; I will speak, though I
am in anguish; I will open up the bitterness of my soul, though I am in dire
straits”). In both cases it occurs together with other particles in what Cox23
calls “the compounding of particles” (οὐ μὴν δὲ ἀλλά, “nonetheless,”—Job
2:5a; 5:8a; 13:3a; 17:10a and 33:1a). In Job 6:21, as “ἀτὰρ δὲ καί” and in 7:11
as “ἀτὰρ οὖν οὐδέ.” This compounding of particles, according to Cox, is
a characteristic of LXX Job. I made various searches in LXX Proverbs in
order to analyse this issue: There are 3 examples of δὲ καί, namely 3:16a;
6:3 and 14:22. However, this could hardly be seen as a compounding since
καί is used paratactically. I could find no examples of this compounding
feature in connection with the conjunction ἀλλά (Cox24 quotes ἀλλὰ μὴν
οὐδἕ—LXX Job 32:21b), nor with ὅτι (ὅτι μέν γάρ—LXX Job 9:19) in LXX
Proverbs. The closest is the combination of εἴ and γάρ as γὰρ εἴ in LXX
Proverbs 23:7 (however, this is a conditional clause); εἴ γάρ in 2:20 and εἰ δέ
in 23:3. But again this is not compounding in the order of those discussed
by Cox. Only one conclusion can be drawn on account of this data, that
the translator of LXX Proverbs did not compound particles, contrary to
LXX Job.
21 Cox, “Tying it All Together,” 42.

22 Translations from NETS.
24 Ibid.
150 johann cook
The subordinating conjunction ὁπόταν is a hapax legomenon in Job

29:22. According to HR, a related form, (ὁπότε), appears 10 times in the
LXX and in Job 26:14, but not in Proverbs.
Another category, adverbs and adverbials point in the same direction.
The adverb ἄλλως appears twice in Job, namely 11:12 and 40:8, but not in
LXX Proverbs. The adverb πάλιν is not used in Proverbs but appears 9
times in Job. The adverb καθώς occurs in Job 6:17; 10:4 (Theodotion) and
42:9 but not at all in Proverbs.
The emphatic particle δή, “indeed,” occurs 15 times in Job but not in
Proverbs. According to Van Rooy, it has a keyness of 19.3. The coordinat-
ing particle, διό, “therefore,” is used in Job 9:22; 32:6 and 10; 34:10 and 34;
37:24 and 42:6 but not in Proverbs. The subordinating conjunction διότι,
“because,” appears in Job 32:2 and 3; 33:9 and 36:12. It is not used in Prov-
erbs. According to HR, the coordinating conjunction τοίνυν, “for,” occurs
21 times in LXX and is used in Job 8:13 (together with οὕτως) and 36:14,
but not in Proverbs. The emphatic particle γε, “yet, even,” is not used in
Proverbs but it appears in Job 13:2, 9; 15:10 (2x); 21:25 and 30:2, 24. The
subordinating conjunction ἐπεί occurs in Job 13:15 and 35:7, and not in
Proverbs. According to HR, the negative combination μὴ οὐ/μὴ οὐχί occurs
18 times in LXX, not at all in Proverbs but once in Job 22:12.
It is rather difficult to interpret this data. For one thing many of these
particles are used in limited cases only. Nevertheless, at the least this cat-
egory is an indication that different translators were at work. The lack of
the compounded list of particles, οὐ μὴν δὲ ἀλλἄ, that appears in Job and
not in LXX Proverbs is the most significant piece of evidence in this regard.
One would surely expect the same translator to have used this list.
2.1.1.2. Correspondences
According to Wordsmith, the particle δἕ is found 4852 times in the LXX.
Of these 735 occurrences are found in Job and 556 in Proverbs. Van Rooy
is of the opinion that this particle is the most conspicuous in LXX Job
compared to the rest of the LXX. It has a relative frequency of 4.3% of
all the words in LXX Job, compared to 0.7% in all the other books. This
unique characteristic applies to LXX Proverbs as well, which has a relative
frequency of 4.3%. This is an indication of large correspondence between
these corpora.
The adverb εἶτα occurs in Prov 6:11 and 7:13 and 12 times in Job. The
subordinating conjunction ἵνα occurs 44 times in Proverbs and 23 times in
Job. This is perhaps surprising since Job is probably more of an argumen-
tative text than Proverbs. However, the individual translators could have
made use of other particles.
The subordinating conjunction ἐπειδή occurs in Job 9:29 and 25:7 as
well as in Prov 1:24. The coordinating conjunction ἤ, “or,” occurs 29 times
in Proverbs and 76 times in Job. The negative particle μηδέ occurs in Job 7
times and 23 times in Proverbs. According to Van Rooy, this is a significant
difference with a keyness of 15. The inferential particle τοιγαροῦν occurs
12 times in the LXX, thrice in Job, namely in chs. 7:11; 22:10 and 24:22 and
twice in Proverbs, namely in 1:26 and 31. Of the conjunction καί there are
472 instances in LXX Proverbs. According to TLG, there are 594 cases in
LXX Job. However, this number must be relativized. A substantial por-
tion, 187 cases come from Theodotion, and moreover, some examples are
indicated to be καὶ γε in TLG, whereas Ziegler defined these as the particle
καίγε. Examples occur in Job 15:10 (2x), 30:2 from Theodotion. Another fac-
tor is the question as to what text should be deemed the OG. I recently
argued that the addition to ch. 2 v. 9 should be deemed the OG text.25
There are 5 instances of καί in these additions. On the contrary, the addi-
tions in ch. 42 v. 17 are most probably the result of at least two later hands.
One should distinguish between addition a and b–e in this regard.26 How-
ever, for the sake of this exercise, I will only discard the Theodotionic ver-
sions. Hence there are 407 instances of καί in LXX Job. This statistic is also
difficult to interpret, since each example could be understood differently
either as copulative, adversative, etc.27 Be that as it may, it is clear that
both translators used this conjunctive extensively. The coordinating con-
junction γάρ is used extensively in both units. There are 171 occurrences
in LXX Job of which only 4 come from Theodotion, thus 167 instances in
total. This particle is used 102 times in LXX Proverbs.
The combination of the interrogative τί and the coordinating conjunc-
tion γάρ occurs in Job, firstly the accusative neuter sg. form τί γάρ in 3:11
and 20; 4:17, 6:5 and 22; 7:17; 10:19; 15:7 and 9; 16:3; 18:4; 21:4; 22:3; 25:2; 31:14.
Also, the nominative m. sg. τίς γάρ occurs in Job 6:11; 13:19; 14:4; 15:14; 19:23;
27:8; 36:22; 41:2 and 42:3.
According to HR, the negative particle μήτε is used 18 times in the LXX
and not in Job and Proverbs. The particle καίτοι is a hapax legomenon in
25 See J. Cook, “Are the Additions in LXX Job 2,9a–e to be deemed as the Old Greek
Text?” Bib 19/2 (2010): 275–284.
26 See A. Y. Reed, “Job as Jobab: The Interpretation of Job in LXX Job 42:17b–e,” JBL 120/1
(2001): 31. See also Gentry, The Asterisked Materials in the Greek Job, 586.
152 johann cook
4 Macc 2:6. The coordinating conjunction οὖν is used 16 times in LXX Job
and 4 times in Proverbs.
The negative particle μηδέ is used in Job 3:4, 6, 7 and 9; 16:18 and 20:17
and 40:3. In Prov 1:10; 3:11 and 31; 4:5 and 6 (2x), 14 and 27; 5:20; 6:4, 7 (2x),
25 (2x); 9:18a; 23:6, 7 and 20; 24:1, 15, 19, 28 and 25:6.
According to TLG, the coordinating conjunction τε is used 18 times in
Job. There are, however, discrepancies between TLG and HR. In the lat-
ter the following instances do not occur in TLG, 1:5; 6:18; 12:7; 14:21; 26:13;
30:4; 37:13 and 39:11. These instances do not occur in Ziegler’s edition. This
brings the number of occurrences to 10. Τε is used 13 times in Proverbs. It is
striking that it is applied 5 times in the first six verses of ch. 1 in Proverbs.
This is done in order to tie together the introduction as to what is needed
to be wise, or to become wiser (v. 5).28 As can be seen there is some cor-
respondence between the way this particle is used in both units.
The interrogative ποτε is used 3 times in Job and 5 times in Proverbs.
The interrogative μήποτε occurs in Job 1:5 and in Prov 22:25; 23:9; 25:16
and 17 and 30:10. According to HR, μή τις/μή τι occurs 11 times in LXX and
in Job 6:22 and 25:3 (the coordinating conjunction γάρ is added in this
instance) and in Prov 3:30. The negative particle μή occurs frequently in
both units. The negative conjunction οὐδέ occurs approximately twice as
much in Job as in Proverbs.
The adverb ἔτι appears only once in Proverbs, namely in 31:7 and 28
times in Job. The adverb νῦν is used extensively in the LXX, but only 16
times in Job. In addition, 5 cases are of the form νυνί (6:28; 7:21; 30:1 and
9 and 42:5) but none appears in LXX Proverbs. There are three cases of
νῦν in Proverbs (5:7; 7:24 and 8:32). The adverb οὕτως is used frequently in
both Job and Proverbs. The interjection ἰδού occurs once in Proverbs (1:23)
and 26 times in Job. The subordinating conjunction ὅτε occurs twice in
Proverbs and 12 times in Job.
The data under this paragraph could be an indication that the same
translator was responsible for both units. However, there are no striking
characteristics in this regard.
28 J. Cook, “Inter-textual relations between the Septuagint versions of the Psalms and
Proverbs,” in The Old Greek Psalter. Studies in honour of Albert Pietersma (eds. R. J. V. Hie-
bert, C. E. Cox and P. J. Gentry; Sheffield: Sheffield University Press, 2001), 228.
2.2. Lexical Items Used Uniquely by the Translators

2.2.1. LXX Proverbs
Being wisdom literature one can naturally expect LXX Proverbs to have
religiously orientated lexemes. A prominent lexeme is found in ch. 8:22.
The translator intentionally opted for the verb κτίζω as equivalent for ‫ ָקנָ ה‬.
The Greek verb appears 63 times in the LXX, but only once in Proverbs
and not at all in Job. The Hebrew lexeme ‫ ָקנָ ה‬has various connotations
in its semantic fields. It is used in Prov 1:5; 4:5 and 7; 8:8; 15:32; 16:16; 17:16;
18:15; 19:8; 20:14 and 23:23. According to Lisowski,29 Prov 8:8 is the sole
example where the nuance of “to create” obtains. It therefore seems as
if the translator intentionally applied κτίζω, in order to avoid possible
misunderstanding.30
LXX Proverbs also contains readings that appear in the political and
administrative contexts and that are not used in Job at all. The hapax
ἀβασίλευτόν (“without a king”) occurs in Prov 30:27. It appears in classi-
cal Greek literature in Thucydides Histories 2.80.5, 6; Xenophon Historia
Graeca 5.2.17 and Plutarch Theseus 24.2. It has a bearing on the issue of
kingship in Proverbs.31 The noun ἀντίδικος is used 8 times in the LXX and
once in Proverbs, namely in 18:17, but not in Job at all. Here it depicts the
adversary/opponent. The hapax δωρολήμπτης is found in 15:27 where it
describes “a receiver of bribes.” The verb παρεδρεύω occurs only in Prov
1:21 and 8:3 in the LXX in connection with the activity of the figure of
wisdom.32 The lexeme συκοφάντης occurs only twice in the LXX, in Ps 71
(72):4 and Prov 28:16. In the latter, it is used to describe a king who is a
swindler. Aitken33 relates it with Ptolemaic papyri.
Other significant lexemes have a bearing on political and religious insti-
tutions. The word συνέδριον appears 8 times in Proverbs with an interest-
ing distribution: 11:13 (‫ ;)סוד‬15:22 (‫ ;)סוד‬22:10 (-)(‫ ;)דין‬24:8 (-); 26:26 (‫;)כהל‬
27:22 (✝) and 31:23 (-). It is used 12 times in the LXX and not at all in Job.
29 G. Lisowsky, Konkordanz zum Hebräischen Alten Testament, Zweite Auflage

(Württembergische Bibelanstallt: Stuttgart, 1958).
30 Cook, The Septuagint of Proverbs Jewish and/or Hellenistic Proverbs, 21.
31 Cf. J. Aitken, “Poet and Critic: Royal ideology and the Greek translator of Proverbs,” in
Jewish Perspectives on Hellenistic rulers (eds. T. Rajak et al.; Berkely / Los Angeles / London:
University of California Press, 2007), 195 on the significance of the poetic lexical choices
this translator made.
32 Cf. ibid., 196. Aitken finds correspondences with the function of judges during the
Hellenistic period.
33 Cf. ibid.
154 johann cook
The following words appear only in LXX Proverbs and not in LXX Job:
The noun ἀγγελία Prov 12:25 (‫) ָדּ ַבר‬, 25:25 (‫מוּעה‬ָ ‫ ) ְשׁ‬and 26:16 (‫) ָט ַעם‬. This
noun is used 14 times in the LXX. Here it represents 3 Hebrew words in
LXX Proverbs. The lexeme αἴσθησις appears 25 times in LXX of which 21
times in Proverbs, consistently as equivalent for ‫ ַדּ ַעת‬. The noun ἀπαίδευτος
is found 16 times in LXX and 7 times in Proverbs. It represents 5 Hebrew
words in LXX Proverbs, namely ‫מוּסר‬ ָ , ‫ ְכּ ִסיל‬, ‫ ֵלץ‬, ‫ ֵאין נָ ָבל‬and ‫ ֱאוִ יל‬. The
noun γνῶσις is used frequently in the LXX and in Proverbs 15 times, mostly
for ‫ ַדּ ַעת‬. There are no equivalents in Job. The verb ἐγκωμιάζω appears
only in Proverbs, namely 5 times. Of these 4 times it represents ‫ הלל‬and
once ‫רבה‬.
As stated already, I have identified numerous hapax legomena in LXX
Proverbs.34
2.2.2. LXX Job
As far as OG Job is concerned, I did not make similar exhaustive lexical
analyses as in the case of LXX Proverbs. I identified a number of hapax
legomena in chs. 1, 2, 14, 19, 28 and 42. A few examples must suffice.
LXX Job ch. 2 contains 4 hapax legomena. The verb ἐκτίνω in v. 5 ren-
ders ‫יִ ֵ ּ֖תן‬. Three more examples come from the set of additions, namely the
verb διανυκτερεύω in 9c and πλανῆτις35 and λάτρις in 9d.
There are lexemes that can help to determine the provenance of OG
Job. Cox36 agrees with Gerleman that the Greek word φορολόγος, “tax
gatherer,” (3:18 and 39:7) is a term that reflects an Alexandrian (Egyptian)
context. This word appears only in 2 Esdr 4:7 and 18; 5:5; Job 3:18 and 39:7,
as well as 1 Macc 3:29 and not in Proverbs. Three other terms fall into the
same category. The noun πάπυρος (“papyrus”) appears only in Job 8:11 and
40:21 (in A) as well as in Isa 19:6. The noun κάλαμος (“reed”) is used 25
times in the LXX, but only once in Job, viz. 40:21. The third term, βούτομον
(“sedge”) also appears in Job 40:21 and in 8:11. Strikingly, none of these
terms occur in LXX Proverbs. This applies also to the term, τετράδραχμον,
which is a hapax legomenon, used as the standard denomination in the
monetary system of the Ptolemies.37 These lexemes seem to point to an
34 Cook, The Septuagint of Proverbs Jewish and/or Hellenistic Proverbs, 335–342.

35 This word occurs only in some MSS.
36 Cox, “The Historical, Social, and Literary context of Old Greek Job,” 108.
37 G. Gerleman, Studies in the Septuagint I, Book of Job (LUÅ Bd 43. Nr 2; Lund: Gleerup,
1946), 36.
Alexandrian provenance of LXX Job. I have argued that the Greek version
of Proverbs in fact came into being in Palestine.38
A number of lexemes occur only in LXX Job and not at all in LXX Prov-
erbs. The noun βροτός occurs only in Job namely 17 times as equivalents
for ‫ ֱאנוֹשׁ‬, ‫ ָא ָדם‬and ‫ ָבּ ָשׂר‬. The adjective γεννητός is used only in Job, 5
times, consistently as equivalent of ‫ילד‬. The verb δείδω is used 8 times in
Job and once in Isaiah. In LXX Job, it represents ‫ ָח ְפ ִשׁי‬, ‫ יגר‬and ‫שׁחח‬. HR
5 times uses the siglum ✝. Διάβολος is a word used 19 times in the LXX and
12 times in Job always for ‫ ָשׂ ָטן‬. Δίαιτα appears 11 times in Job and is used
for ‫נָ וֶ ה‬, ‫ א ֶֹהל‬and ‫ ַבּיִ ת‬. HR uses the siglum ✝ 3 times and the noun occurs
once in Judg 12:15. The lexeme ἐξαίσιος occurs only in Job (9x) for ‫ִמ ָקּרוֹב‬
‫ ֵשׁ ֶמץ‬, ‫פּלא‬, ‫ ֶצ ַלע‬, and ‫ ִפּ ְתאֹם‬respectively. The noun ἐπιστήμη is used 13
times in Job with different Vorlagen (‫ ִבּינָ ה‬, ‫ ְתּבוּנָ ה ֵדּ ַע‬, ‫ ַדּ ַעט‬and ‫ ) ָשׂכל‬and
not in Proverbs. The verb ἐφοράω occurs 4 times in Job and as equivalent
for 3 Hebrew words. In Job 21:16, for ‫ ;רחק‬22:12 for ‫ ;ראה‬28:24 for ‫נבט‬.
One example (34:24), according to HR, has no Semitic Vorlage. The noun
σαπρία (“refuse”) is used 9 times in the LXX and primarily in Job, namely
in 2:9; 7:5; 8:16; 17:14; 21:26 and 25:6, mostly in connection with ‫ ִר ָמּה‬.
As I stated earlier, the choice of lexemes has been made at random.
This exercise naturally needs to be extended, but can be deemed repre-
sentative. It is also evident that the Wordsmith option needs to be worked
out further.
3. Conclusion
I am forced by the evidence to reiterate the conclusion I reached in New

Orleans. I find it difficult to accept that two translators who exhibit such
divergent attitudes to their parent texts and to specific lexical items, could
have belonged to the same circle of translators, nor could be deemed as
one and the same translator.39
38 Cook, “Theological/ideological Tendenz in the Septuagint—LXX Proverbs a case

study,” in Interpreting Translation. Studies on the LXX and Ezekiel in honour of Johan Lust.
(eds. F. García Martínez and M. Vervenne; Leuven / Paris / Dudley: Leuven University
Press / Peeters, 2005), 79.
39 Cf. also Gammie, “The Septuagint of Job,” 29. Jan Joosten reached the same conclusion
in respect of the way another category, elaborate similes, is rendered differently in LXX
Proverbs and Job (“Elaborate Similes—Hebrew and Greek. A Study in Septuagint Transla-
tion Technique,” Bib 77 [1996]: 236).
An Analysis of the Use of Hebel as a Metaphorical and
Symbolic Device as Interpreted in LXX Ecclesiastes
Lawrence Lincoln
1. Introduction
This paper intends to examine how the Greek translation of Qohelet rep-
resented the lexeme ‫( הבל‬hebel) and how it incorporated the thematic
nature and concepts of the meanings of this lexeme into a Greek text
that would have been accessible and understandable to Jewish and Greek
reading audiences.
Qohelet is in many ways a controversial and unusual book when com-
pared to the rest of the Hebrew canon. As one of the five books com-
prising the group known as the Megillot (or Ketubim), its controversial
themes makes it stand more or less on its own in the collection of inspired
biblical works. Its enigmatic nature speaks not only of the world of the
third century B.C.E. when it was most probably composed,1 but also of a
universalistic humanism that is readily recognizable even today. Although
by tradition ascribed to the hand of King Solomon, most modern schol-
ars have rejected the book’s identification with Solomon as its author.2
1 The date of the composition of Qohelet has been extensively researched, not only
because of the controversial message of the book, but because of its attribution of Solomon
as its author. With few and somewhat vague historical allusions to go by, the extensive
scholarship has proposed a range of possible dates for its composition range from the
time of Solomon himself to the Persian period, the Greek age down to 100 C.E., a period
spanning six hundred years. Attempts to date the work by analysing the linguistic charac-
ter of the Hebrew have also fallen far short of expectations. Choon-Leong Seow probably
echoes the view of most modern scholars on Qohelet that it was most likely composed
shortly before the conquest of Palestine in 333 B.C.E. C. L. Seow, Ecclesiastes: A New Trans-
lation with Introduction and Commentary (AB 18; New York: Doubleday, 1997). In a PhD
dissertation L. R. Tyler analysed 66 individual lexical items to ascertain the lateness of
the book’s composition or, its possible earlier authorship attributed to King Solomon. In
all there were only six truly late lexical items that proved a later authorship, and as noted
by Tyler in his closing word: “The linguistic evidence, then, does not disprove Solomonic
authorship, nor does it prove that Solomon wrote Ecclesiastes.” L. R. Tyler, The Language
of Ecclesiastes as a Criterion for Dating (PhD diss., Texas, 1988). An earlier yet still relevant
study is that of E. Bickerman, Four Strange Books of The Bible (New York: Schocken Books,
1967), who also argues for a third century date of composition.
2 The prescript to the book states that it was written by the son of King David, although
the Hebrew version does not specifically refer to Solomon as such. Only the Aramaic
158 lawrence lincoln
Although Solomon is not mentioned specifically by name, the attribution

to Solomon by later rabbinical hands gave the text the cachet of legiti-
macy. Midrash Qohelet (known as Haggdat Qohelet in the Arukh) but only
called the Qohelet Rabbah since the Venice Edition of 1545,3 is a much later
work, probably eighth century, which fixes Solomon clearly as the spiri-
tual inspiration behind the creation of Qohelet: “The words of Koheleth,
the son of David, King in Jerusalem. That is what scripture declares by the
Holy Spirit through Solomon, King of Israel (1.1).”4
Hebrew Qohelet does not contain any stirring or dramatic narrative,
nor is it populated by sages or memorable characters (except obliquely
when referring to the putative authorship of King Solomon), nor does it
deal with sacred, legal or Hallachic matters. Singularly, as a Jewish crea-
tion, it contains no doctrinal imperatives; no messianic or end-of-days
grand plan; nor is mention made of the covenant and the term “Israel”
does not occur at all. The effect of all this is to give the work a philosophi-
cal tone rather than one of any direct religiosity.5 Compared to the other
books of the Hebrew Bible which deal with the historical experiences of
the Israelite nation and the development of a theological foundation for
the Jewish faith, the wisdom books, on the other hand, present a pan-Near
Eastern secular basis for Jewish wisdom thought.6 There is little there-
fore that identifies Qohelet, Job and Proverbs as particularly Israelite in
character. In fact, Job himself for instance, appeared to have dwelt in the
land of Uz, probably in southern Palestine in Edom or northern Arabia,7
but illustrates the wider context in which wisdom-type thought devel-
Targum adds “Solomon, the Son of King David in Jerusalem” as does the Midrash in
Qohelet Rabbah. The LXX also does not mention Solomon by name, although the para-
phrase by Gregory of Thaumaturgos goes further by stating in the opening sentence to the
book, “Solomon (the son of the king and prophet David) . . .” (Τάδε λέγει Σαλομών, ὁ τού
Δαβίδ βασιλέως καί προφήτου). J. Jarick Gregory Thaumaturgos’ Paraphrase of Ecclesiastes
(Atlanta: Scholars Press, 1990); likewise the Vulgate also only mentions the son of David:
filii David Regis Hierusalem.
3 See H. L. Strack and G. Stemberger, Introduction to the Talmud and Midrash (Edin-
burgh: T & T Clark, 1991), 342.
4 A. Cohen, Midrash Rabbah (vol. 8; London: Soncino Press, 1971).
5 Seow, Ecclesiastes, 54.
6 L. H. Schiffman, From Text to Tradition: a History of Second Temple & Rabbinic Juda-
ism (Hoboken: Ktav Publishing House, 1991), 32. As noted by James Crenshaw, “Wisdom
expresses itself with remarkable thematic coherence. Wise men and women address com-
mon problems, whether the dangers of adultery, the perils of the tongue, the hazards of
strong drink, the enigma of undeserved suffering, the inequities of life, or the finality of
death” (cf. J. L. Crenshaw, Old Testament Wisdom: An Introduction [London: SCM Press,
1981], 27–41).
7 See also Gen 36:28 and Lam 4:21.
an analysis of the use of hebel as a metaphorical 159
oped. This could partly explain why these three wisdom works were not
concerned with Israelite revelation and national redemption at all.8 In
Qohelet, God is mentioned 40 times (‫)אלוהים‬9 as such; Qohelet never
questions the existence of God, but is rather obsessed with the theme of
injustice and the capriciousness of human existence and a God who does
not appear to have a direct or determinative role in the fate of mankind.
Job too rails against the world that is filled with injustices in a universe
governed by a God who is supposed to be a God of justice. Qohelet’s treat-
ment of God (‫ )אלוהים‬is peculiar and confuses the reader; ‫ אלוהים‬is first
mentioned in 1:13 in a somewhat confusing context: “I gave my heart to
search and explore with wisdom all that is done under the heavens and
it is a hard thing that God has given men to preoccupy themselves with”
(1:13).10 God is therefore a problematic concept in Qohelet and, as H. Fisch
has observed, the God of Israel is “half-concealed and half-revealed” and is
only discerned through dialectic of presence and absence.”11
In Qohelet the cosmos and God are a mystery. God is undeniably in
control, but his actions and that of the cosmos are mysterious to men,
who have to deal with events over which they have little or no control,
while enduring a life that is unpredictable and harsh, and one in which
they are ultimately faced with death, like any of the animals of the earth.
Qohelet’s single most important feature therefore is its ontological
quest to finding out what being human is all about. Martin Heidegger, in
his monumental work, Being and Time, asked the same question but from
a different perspective: “if we are capable of dying, how can we possibly
8 See R. Alter, The Wisdom Books: Job, Proverbs and Ecclesiastes (New York: Norton,
2010), 3–10. The commentaries on wisdom literature are numerous and space does not
permit mentioning them in detail.
9 The term ‫אדוני‬, the commonly accepted personal appellation of God, does not occur
at all. Qohelet appears to speak to and of a “generic” God in keeping with the philosophic
tone of the book.
10 I have translated ‫ בו לענות‬as “preoccupy themselves with”; however, the Targum
has used the term “afflicted” ‫לאיסתגפא‬. The LXX uses περισπᾶσθαι. The Targum, however,
acknowledged the problematic nature of this passage and reinterpreted it entirely as fol-
lows: “And I set my mind to seek instruction from the Lord at the time when he revealed
himself to me at Gibeon to test me and to ask me what I wanted from him. And I asked
of Him only the wisdom to know the difference between good and evil understanding of
everything that happened under the sun in this world. I saw all the deeds of sinful people
were an evil matter which the Lord gave to the people so that they should be afflicted
by it”. I have used the translation here by P. Knobel, The Aramaic Bible (Collegeville: The
Liturgical Press, 1991) 24.
11 H. Fisch, Poetry with a Purpose: Biblical Poetics and Interpretation (Bloomington: Indi-
ana Press, 1988), 165.
understand death?”12 In other words, if death is so fundamental to being

human, how then does one begin to understand what being human really
means? This kind of enquiry is by its very nature inherently melancholy,
a feature often ascribed to Qohelet. This is a universalistic theme and one
that has its roots in ancient philosophical enquiry, which is possibly why
it resonates so strongly and appears curiously familiar at first reading,
according to R. Whybray.13
It is therefore no surprise that this difficult book should have aroused
controversy before it was finally accepted as part of the Hebrew Bible
tradition. It has been traditionally accepted that the Council of Yavneh
played the determinative role in fixing the Hebrew collection of inspired
books at the end of the first century C.E. and thereby implying the time
when Qohelet’s reception was apparently finalised. This kind of linear
approach must be handled cautiously, as Schiffman has noted; there is no
evidence for supposing that the council at Yavneh was responsible for for-
mally approving and accepting the books of the Ketubim into the canon,
or that the canon as a whole had reached its final stage of acceptance.14
Qohelet continued to be a controversial book and generated internal
debate amongst the rabbis at least until as late as the fourth century C.E.,
according to Jerome.15 “The Hebrews say that . . . this book should be oblit-
erated, because it asserts that all the creatures of God are vain, and regards
the whole as nothing, and prefers eating and drinking and transient pleas-
ures before all things.” The Mishnah provides a succinct example of the
type of debate that Qohelet generated in m. Yad 3.5. This section names
the rabbis involved and includes the following observation:
All the Holy Scriptures renders the hands unclean. The Song of Songs and
Ecclesiastes render the hands unclean. R. Judah says: The Song of Songs
renders the hands unclean, but about Ecclesiastes there is dissension. R. Jose
says: Ecclesiastes does not render the hands unclean, and about the Song of
Songs there is dissension. R. Simeon says: Ecclesiastes is one of the things
12 M. Heidegger, Being and Time (trans. J. Macquarie and E. Robinson; Oxford: Black-
well, 1995), 41.
13 R. Whybray, “Qoheleth, Preacher of Joy,” JSOT 23 (1982): 87–98.
14 Schiffman, From Text to Tradition, 56–58.
15 Jerome, Commentarius Ecclessiasten in Patrologia Cursus Completus: Series Latina 23
(ed. J. P. Migne; Paris), 1172. Another point of reference is Josephus, who wrote in the last
decade of the first century: “For we have not an innumerable multitude of books among
us, disagreeing from and contradicting one another (as the Greeks have), but only twenty-
two books, which contain the records of all the past times” (Contra Apionem 1.38); cf.
W. Whiston, The Works of Josephus (Peabody: Hendrickson Publishers, 1987).
about which the School of Shammai adopted the more lenient, and the
school of Hillel the more stringent ruling.16
Despite Talmudic references to the various debates around its acceptabil-
ity, the rabbinical pericopes are scholarly in nature, indicating that the
book had been in use already in some form or another, and that the decree
for its acceptance into the Hebrew canon was probably a fait accompli
but was still being debated and discussed.17 Arie van der Kooij explains
that
the ancient books and scriptures would not have been seen as carrying
authority if their teachings had not been brought into force and if they had
not been studied by the appropriate authorities—the scholar scribes. Inter-
pretation of books that were considered authoritative required authoritative
and authorised persons to bring the ideas into effect.18
A further clue as to the book’s earlier existence and influence before
the first century lies in the possibility that it was used by the author of
Ecclesiasticus, or the Wisdom of Ben Sirach, which was composed about
180–175 B.C.E. It is also likely that Qohelet was known to the author of the
Wisdom of Solomon, which was most likely compiled in the first century
B.C.E.19
2. Hebel
Qohelet begins the introductory passage with the following startling state-
ment: “Everything is futile, said Qohelet, it’s futile, it’s all completely futile”
(1:2) (my translation). The concept of hebel, most commonly rendered in
English as “vanity,” has been one of the most studied and disputed con-
cepts/lexemes in the Qohelet text and an extensive number of commen-
taries have been devoted to this one aspect of the Qohelet book alone.
The term hebel appeared in the King James Version as “vanity” and hebel
habalim as “vanity of vanities,” according to Robert Alter, most probably
16 English translation by H. Danby, The Mishnah (London: Oxford University Press,

1933), 781.
17 See Talmud b. Sab. 30b; Lev. Rab. 28.1; and Abot de R. Nat. 1.4.
18 A. van der Kooij, “Authoritative Scriptures and Scribal Culture,” in Authoritative
Scriptures in Ancient Judaism (JSJSup 131; ed. M. Popovic; Leiden: Brill, 2010), 55–71.
19 The author of the Wisdom of Solomon attempted to rectify certain of Qohelet’s say-
ings he deemed to be sinful. See G. Barton, A Critical and Exegetical Commentary on the
Book of Ecclesiastes (ICC; Edinburgh: T & T Clark, 1908), 53–58.
as a result of the Vulgate influence from the Latin vanitas.20 In my reading

of the concept of hebel as used by Qohelet, I do not think that the modern
English equivalent of “vanity” has quite the correct sense of the meaning
of the Hebrew term hebel. In my opinion, Qohelet was principally trying
to convey a particular message, that in spite of all the advantages of fear-
ing God, the advantages to be gained in studying wisdom and pursuing
wealth and other material pleasures of life, in spite of all these, they are
all just pointless as nothing lasts in the cycles of life. In other words, the
lot of mankind is essentially pointless; as in the beasts of the earth, death
and rotting in a grave are the lot of mankind and it is futile to expect more
than that. Michael Fox has translated hebel as “absurd” and “absurdity of
absurdities.” He bases part of his argument on the work of Albert Camus
and his work The Myth of Sisyphus, and demonstrates that both Camus
and Qohelet are similar in that they excelled in portraying the irratio-
nalities of human existence, “The two thinkers are aligned also in their
unflinching determination to strip away illusions and to face life’s harsh
irrationality.”21 Fox, however, realizes that the “absurd” is not applicable
in all cases and advocates related synonyms, such as ephemeral, as in the
case of the comparison between animals and man when living beings are
referred to as ha-kol in 3:18–19. As Fox has noted, the death of an animal
cannot be regarded as “absurd” and equated with that of humans and
proposes in this case the term ephemeral when dealing with life forms
in general.22 The point is that life and the very notion of being human
and the way we live as it is desires rational outcomes and hope, but the
reality is in fact the opposite, which is that it is absurd and therefore futile
and pointless to be striving for any outcome (my emphasis). In a recent
book by Robert Alter, hebel and its variants have been translated as “mere
breath” and “merest breath,” which although it acknowledges the literal
and material nature of hebel, sounds contrived and also does not convey
the context, purpose and intent we can presume of the author or authors,
nor do these terms capture the extreme irony and pathos of human exis-
tence, according to Qohelet’s worldview.
The point is that hebel as a metaphorical and abstract concept is prob-
lematic from a translational point of view as well as when striving for
20 Alter, The Wisdom Books: Job, Proverbs and Ecclesiastes, 339.

21 M. V. Fox, A Time to Tear Down and a Time to Build Up: A Rereading of Ecclesiastes
(Grand Rapids: Eerdmans, 1999), 10.
22 Ibid., 39.
a complete reading of the book. As Douglas Miller has noted: “Qohelet

employs hebel as a ‘symbol’, an image which holds together a set of mean-
ings, or ‘referents’, that can neither be exhausted nor adequately expressed
by any single meaning.”23 Significantly, Miller also states correctly that “no
single metaphor has been found adequate for the range of hebel’s diverse
usage” (my emphasis).24
On a practical level the lexeme hbl appears 38 times in the Hebrew ver-
sion of Qohelet, if one includes hebel habalim in 1:2.25 Hebel is a noun and
in its various uses in the MT it is used metaphorically to describe specific
conditions of human experience and in a few cases as a verb in the MT.26
It is also used more than one hundred times in the Babylonian Talmud
and in other biblical texts.27 The original Hebrew meaning is to describe
an exhalation of breath, a mist, light wind or vapour. In short, it describes
some form of breath/vapour/mist that can be seen, but has no physical sub-
stance beyond that and just disappears as quickly as it had been observed.28
Trying to take hold of it is thus physically impossible—thus it is futile.
The term thus acquired its metaphorical sense to describe the condi-
tions of living beings (humans and animals) that are observable, ironic,
and contradictory. Typical synonyms used are most notably “ephemeral,”
insubstantial, incomprehensible, enigmatic and inconsistent.29 In French
it is rendered as “fumeé,” in German “Vergänglichkeit,” “Nichtigkeit” and
“Eitelkeit,” in Dutch “ijdel” and in Afrikaans “tevergeefs.”
Fox has noted that in the absence of any clear history of the develop-
ment of the word in the Hebrew language, its metaphorical development
23 D. B. Miller, Symbol and Rhetoric in Ecclesiastes: The Place of Hebel in Qohelet’s Work
(Atlanta: Society of Biblical Literature, 2002), 15. Also, P. Wheelwright, Metaphor and Real-
ity (Bloomington: Indiana University Press, 1962), 92.
24 Miller, Symbol and Rhetoric, 14.
25 Some scholars believe that ‫ הכל‬in 9:2 should read as ‫ הבל‬and that at 9:9 the second
‫ הבל‬should be omitted because of scribal errors, as has been done in the LXX.
26 Job 27:12; Ps 62:11; Jer 23:16; 2 Kgs 17:15; Jer 2:5.
27 Miller, Symbol and Rhetoric, 187–194.
28 Seow, Ecclesiastes, 47; all the other studies on Qohelet are in broad agreement with
this explanation.
29 Also L. Koehler and W. Baumgartner, Lexicon in Veteris Testamenti Libros; (Leiden:
Brill, 1958); B. Davidson, The Analytical Hebrew and Chaldee Lexicon (London: Samuel Bag-
ster, 1959).
is hinted at from its other uses in OT.30 Some of the other MT references
in which hebel and its variants appear are noted below.31
1. Ephemeral connotations
Ps 39:6: “Surely every man walks in a vain show; surely they are disqui-
eted in vain (‫)אך הבל יהמיון‬.”
Prov 21:6: “The obtaining of treasures by a lying tongue is a vanity
(‫ )הבל‬tossed to and fro of them who seek death.”
Ps 144:4: “Man is like to vanity; his days are as a shadow (‫אדם להבל‬
‫ )דמה‬that passes away.”
Job 7:16: “I loathe it; I would not live forever—leave me alone, for my
days are vanity (‫)כי הבל ימי‬.”
2. Vain
Isa 49:4: “I have laboured in vain (‫)הבל‬.”
Isa 30:7: “Egypt will help only in vain (‫ )הבל‬and emptily (‫)רק‬.”
See also Lam. 4:17; Job 9:29; and Ps 94:11 for similar meanings of hebel
as meaning “vanity.”
3. Deceit
Hebel is also synonymous with “lies” and “deceit.” Whenever šeqer,
‘awen, ma‘al or kazab appears in the following verses, it includes hebel
within the same verse to highlight the vanity of deceitful deeds and
thoughts as in Prov 31:30; Job 21:34; Ps 62:10 and Zech 10:2.
The theme of deceit is taken further to also mean the uselessness of
pursuing false gods as in Jer 16:19; Deut 32:21; Jon 2:9; Zech 10:2; 2 Kgs
17:15; Jer 2:5, 8:19, 14:22.
But it is a moot point whether Qohelet was influenced in any way by
the passages noted above.32
30 Cf. Fox, A Time to Tear Down and a Time to Build Up, 27–30. See C. D. Ginsberg,
Coheleth, Commenly Called the Book of Ecclesiastes (London: Longman, Green, Longman
& Roberts, 1861), 13; and O. Loretz, Qohelet und der alte Orient: Untersuchungen zu Stil und
theologischer Thematik des Buches Qohelet (Freiburg: Herder, 1964), 223, who both describe
hebel as denoting “nothing,” but Fox (A Time to Tear Down and a Time to Build Up, 28)
counters with the argument that hebel is never used in Qohelet to mean “nothing.”
31 Adapted from a summary by Fox, A Time to Tear Down and a Time to Build Up,
28–29.
32 G. S. Ogden, “The Meaning of the Term Hebel,” in Reflecting with Solomon: Selected
Studies in the Book of Ecclesiastes (ed. B. Zuck; Grand Rapids: Baker Books, 1994), 227–232.
Hebel’s use is in fact quite extensive in numerous biblical and post-biblical

texts. A summary of these various usages and their loci is provided by
Douglas B. Miller (Appendix II), which shows how widespread the lexeme
was and the variety of metaphorical and non-metaphorical usages it had
in biblical and post-biblical literature. In short, it appears in hebel form
in Hebrew or as habla in its Aramaic form 96 times in noun forms and
9 times in verb forms in the MT. Hebel can also be found over 100 times
in the Babylonian Talmud.33
The term hebel and the metaphorical sense it connoted were there-
fore rooted in the Semitic consciousness and had a number of different
contextual meanings, all of which bear negative values and connotations
which would have been commonly understood as such by their intended
target audience. Different negative inferences and contexts included, diffi-
cult human conditions (Job 7:21), errors of speech (Job 35:16), futile labour
(Isa 49:4, Job 9:29), acquiring wealth dishonestly (Prov 21:6) and numerous
other contexts in which hebel was used symbolically and metaphorically.34
The debate on the intellectual influences on Qohelet has been widely
studied and the consensus amongst most scholars appears to be that it
is impossible to verify that there were any foreign, i.e. non-Jewish, influ-
ences on the author’s thought. Besides numerous Aramaisms, there are no
other convincing or discernible traces of language influences to be found
in his work.35
3. The Greek Version of Qohelet
There is no evidence of an Old Greek equivalent of Qohelet in the Alexan-

drian translation tradition as there are for the other biblical books of the
Hebrew canon. The Pentateuchal books were most likely translated in the
third century B.C.E. and the Ketubim after that, probably at various later
dates. The Greek texts are preserved in six uncial MSS with the fourth-
century Codex Vaticanus considered to be the most reliable.36
33 Miller, Symbol and Rhetoric, 187–194.

34 Ibid., 190–194.
35 Seow concludes that although there are some loan words from other languages,
these in themselves do not provide sufficient evidence to prove that the work was indeed
Phoenician, as proposed by M. Dahood, “The Language of Qoheleth,” CBQ 14 (1952): 227–
232; idem, “Canaanite Words in Qoheleth,” Bib 46 (1965): 210–212; idem, “The Phoenician
Background of Qoheleth,” Bib 47 (1966): 264–282.
36 Hamburg Papyrus Codex (possibly end of third century), written in both Greek and
Coptic and close to the Codex Vaticanus, confirms the existence of variant readings with
little or no other significance. A complete list of the Greek manuscripts is:
Heinrich Graetz put forward the suggestion that LXX Qohelet was in
fact the work of Aquila, based on a number of similarities between the
LXX and the so-called Aquilan style.37 This generated vigorous and on-
going debate on whether the extant versions came from an Aquilan origi-
nal. Although this hypothesis has been widely accepted by some, some
scholars have argued against it, most notably Hyvärinen,38 John Jarick39
and Seow.40 The search for the correct answers probably lies somewhere
in between the opposing hypotheses, and Seow is most probably correct
when he concludes that the LXX is not the work of Aquila, but a version
that, like Aquila, is motivated by the desire to facilitate certain kinds of
exegesis promoted by the rabbis. Thus the style of the LXX may be bet-
ter explained as “proto-Aquilan” but not strictly Aquilan—“not Aquila
himself.”41 Thus the LXX version is “Aquilan” in nature, while the third
column of the Hexapla is in fact the work of Aquila as noted by Origen,
a view that is also supported by Fox.42 It is not known if there had been
an earlier version than Aquila’s first edition. According to Barton, “The
translation we have in the LXX was at all events made from a text which
differed a good deal from our present Hebrew, and was therefore made
from a text that Aquiba had not revised.”43 Barton goes further by add-
ing that if this is correct, then the LXX Qohelet version we have probably
dates from the second quarter of the second century.
The character of the translation of LXX Qohelet can be described as
being extremely literal and very close to the Hebrew original in most
respects. Fox calls it a mimetic approach and “consistent in word-
correspondences,” thus making it easier to identify and possibly explain
• Hamburg Papyrus Codex—third century

• Codex Vaticanus—fourth century
• Codex Sinaiticus—fourth century
• Codex Alexandrinus—fifth century
• Codex Ephraemi—fifth century
• Codex Venetus—eight/ninth century.
37 Graetz’s argument is based largely on the fact that the lexeme “et is translated as
σύν.” H. Graetz, Kohelet (Leipzig: Winter, 1871), 171–189; A. H. MacNeile, An Introduction to
Ecclesiastes (Cambridge: Cambridge University Press, 1904), 115–168.
38 D. K. Hyvärinen, Die Űbersetzung von Aquila (ConBOT 10; Lund: Gleerup, 1977),
88–99.
39 J. Jarick, “Aquila’s Koheleth,” Textus 15 (1990): 131–139.
40 Seow, Ecclesiastes, 7–8.
41 Ibid., 8.
42 M. V. Fox, Qohelet and his Contradictions (JSOTSup 71; Sheffield: Almond, 1987), 164.
43 Barton, Ecclesiastes, 8–11.
the variances.44 As a result of such a rigid translation formula, in which

each lexeme attempts to match the Hebrew equivalent, the LXX Qohe-
let ended up being probably somewhat stilted and alien-sounding to the
readers of Hellenistic Greek of the time—a point made by Peter Gentry
in the introduction to the NETS translation of Qohelet.45 The translator
tried to emulate some of the Semitic stylistic forms, thereby coming up
with non-conventional syntactical elements.
Several examples to illustrate this point and demonstrate the trans-
lator’s consistency and uniformity in finding rigid and consistent Greek
equivalents to match the Semitic equivalent.
he Hebrew ‫ גם‬is consistently rendered as καίγε (1:17; 2:1, 7, 8, 14, 15bis,

• T
19, 21, 23, 24, 26; 3:11; 4:4, 8tris, 11, 14, 16bis; 5:9, 16, 18; 6:5, 9; 7:14, 21, 22bis;
8:10, 12, 14, 16; 9:1bis, 6bis, 12, 13; 10:20; 12:5. The Hebrew lexeme ‫ וגם‬is
also consistently translated with καίγε, as in 1:11; 3:13; 5:15; 6:3, 7; 7:6, 18;
8:17; 9:3, 11bis; 10:3, 11:2.
• To illustrate the pedantic exactness of the translator even further, the
Hebrew lexeme ‫את‬, where it is used as a marker denoting the accusa-
tive, were in all cases rendered consistently in the LXX as σύν (1:14; 2:12,
17, 18; 3:10, 11bis, 17bis; 4:1, 2, 3, 4bis, 15; 5:3, 6; 7:14, 15, 18, 26, 29; 8:8, 9, 15,
17bis; 9:1bis, 15; 10:19, 20; 11:5, 7; 12:9, 14).
• There are a number of other less frequent translation idiosyncrasies. In
2:1 ‫ בתוב הראה‬is rendered in Greek as καί ἰδέ ἐν ἀγαθώ, thereby chang-
ing the Hebrew idiomatic phrase from “enjoy with pleasure,” to the lit-
eral “I see in good”, which contextually does not make any sense.
In terms of the Greek equivalent Qohelet hebel is translated by the lex-

eme ματαιότης (“vanity”) in all cases, except later in Origen’s third column,
where it is translated as ἀτμίς (“breath”).46 Ματαιότης therefore ensured
that in Greek there could be no ambivalence about the meaning of hebel in
its Greek forms and it may have been meant to consistently convey a clear
ethical message, with a range of abstract meanings—empty, frivolous, vain,
44 Fox, A Time to Tear Down and a Time to Build Up, 156.

45 cat.sas.upenn.edu/nets/edition.
46 The exceptions are Codex Vaticanus of the Septuagint and Hamburg Papyrus, where
alternatives are used at 9:9 where ἀτμος occurs. This can also be seen in Aquila, Symma-
chus and Theodosius at 9:9 and elsewhere occasionally. This has been noted by Miller,
Symbol and Rhetoric, 59.
idle and trifling.47 One can conclude that the problems of translating the
concrete term hebel was not lost on Jerome, as he also used an abstract
term in Latin, vanitas and the corresponding adjective vanus.
Gregory Thaumaturgos recognised the problems inherent in the LXX
version based on the Hebrew and reinterpreted the work considerably,
especially when dealing with the contextual situations concerning hebel/
ματαιότης. He did so by using a range of alternative terms for hebel not
only as a means to clarify and interpret, but probably also to reduce the
repetition of ματαιότης.48
In 1:14, hebel is accompanied by the phrase, ‫רעות רוח‬, (“chasing the
wind”), which is unique to the MT and anyway would not have made
much sense if literally translated as such into Greek. This phrase appears
seven times with hebel in Qohelet: 1:14; 2:11, 17, 26; 4:4, 6, 16. It therefore
appears as προαίρεσις πνεύματος, which in Greek means “preference/choice
of spirit” (NETS), thus losing the metaphorical association of the wind, in
that ‫ רוח‬in the Hebrew also referred to spirit/ soul.49 ‫ רעות רוח‬therefore
referred to the useless/pointless chasing of something that one cannot
catch and thereby also extended the metaphorical sense in which hebel
was employed.50 In Greek, however, the intended meaning of προαίρεσις
πνεύματος was to point out the dangers and the harm that can come to a
man who seeks things that are unattainable, especially if left to chance.51
Qohelet deals with youth and with being young and how young people
should enjoy this early stage of their lives without forgetting that in the
end there will ultimately be God’s judgment awaiting them, presumably
47 H. G. Liddell and R. Scott, Greek-English Lexicon (Oxford: Clarendon Press, 1975), 489;
S. Holm-Nielsen (“On the Interpretation of Qoheleth in Early Christianity,” VT 24 [1974]:
168–177) also argues that ματαιότης gives hebel a more ethical slant. Robin Salters noted
that ματαιότης meaning “purposelessness” is preferable contextually because the alterna-
tive, ἀτμίς, did not possess any figurative connotations. R. B. Salters, The Book of Ecclesi-
astes: Studies in the Versions and the History of Exegesis (PhD diss., St. Andrews 1973).
48 Gregory used alternative words such as κενός, ἀνόνητος, ἀτοπία, ἄχρηστος, δολερός,
ἄθλιος, πονηρία, ἄνοια and πλάνη.
49 In Hebrew/Jewish eschatology man possesses a body as well as a life spirit (ruaḥ).
At the time of death the body and the spirit separate and God takes back the gift of life
he granted. See, for example, Ps 104:29: “you hide your face and they are troubled; you
take away their breath, they die, and return to their dust.” Refer also to Job 34:14–15 and
Gen 2:7.
50 The use of “wind” can also be found in other sources of wisdom literature, such as,
Job 15:2; Prov 27:16; Prov 11:29; Prov 30:4 and Job 16:3. See Miller, Symbol and Rhetoric, 94.
51 In the Vulgate this is translated as universa vanitas et adflicto spiritus (“all is vanity
and distress of purpose”). This is probably based on an assumption that ‫ רעות‬is derived
from the Aramaic root ‫רעע‬. See Miller, Symbol and Rhetoric, 93.
with approaching old age (11:9–10). In Qohelet’s thought, the state of youth
is hebel, a mere transient phase (part of a cycle) of being human. The
Hebrew uses a metaphor for this description, ‫כי הילדות והשחרות הבל‬.
It is likely that the LXX translator may have struggled somewhat to find a
suitable metaphorical phrase to match ‫ ;ילדות והשחרות‬šahªrŭt, meaning
the prime of life (and most probably derived from ‫שחר‬, “dawn”), prob-
ably did not translate well into Greek and anyway would have sounded
repetitive.52 The Greek is therefore translated as ὅτι ἡ νεότης καὶ ἡ ἄνοια
ματαιότης—“Put away anger from your heart, and divert pain from your
body, for youth and lack of understanding are vanity.” In my view, this is
another example of how subtle changes were made to the Hebrew text
in order to translate Semitic metaphors and symbols, ambiguous mean-
ings, words and repetitiveness, and at the same time to include an ethi-
cal dimension without detracting from the overall literal approach in the
translation technique.
That the LXX translator took great care word-for-word to make sure
that the translation would make sense to a Greek-speaking person can be
seen, for example, in 6:7. The Hebrew Bible version reads: “All the labour
of man is for his mouth and yet the appetite (‫ )נפש‬is not filled.” In the
LXX, nephesh is replaced with ψυχή, thereby conveying a more theological
connotation than a literal form of nephesh would have and also avoid-
ing a Hebrew metaphor which would not have translated well into the
equivalent Greek.53
An additional example comes from 1:11 and, although it is not directly
connected with hebel, it is worth quoting in order to illustrate how care-
ful the translator was to preserve the original, but at the same time avoid
textual and contextual confusion. The verse in Hebrew has an apparent
repetition which may have been confusing in a translation setting, but
the interesting point is that the Greek replaces the last two words with
εἰς τὴν ἐσχάτην, “who will be born at the last.” This is a seemingly eschato-
logical viewpoint which may have been due to later rabbinical influences.
52 Koehler-Baumgartner has “prime of life” (Lexicon, 962), while Davidson in his Ana-
lytical Hebrew and Chaldee Lexicon attributes a meaning of “dawn of life” as a metaphor
for youth (709). There is also a semantic connection with “dark hair,” presumably as a
metaphor to distinguish between a young child and an adolescent. See also W. L. Holladay,
Hebrew and Aramaic Lexicon of the Old Testament (Grand Rapids: Eerdmans, 1988), 366;
similarly in F. Brown, S. R. Driver and C. A. Briggs, Hebrew and English Lexicon of the Old
Testament (Oxford: Clarendon Press, 1906), 1007.
53 According to Koehler and Baumgartner (Lexicon, 627), nephesh is translated in the
OT 755 times by ψυχή.
However, the end of day’s connotation is clear in the Vulgate and may
have been what the LXX translator was also trying to imply.
4. Conclusion
The consistent use of some idiosyncratic words and morphological con-

structions such as the use of καίγε and σύν indicates that the translator, or
translators, went to great lengths to produce a relatively literal translation
based on the original Hebrew source text without compromising on the
overall theme or message of Qohelet. Where adaptations to the meaning
of a passage were required to make them understandable in Greek, the
translator was able to do so effectively. LXX Qohelet does not detract in
any way from the essential Jewishness of its content, but it does appear
to soften its tone slightly to convey a more ethical slant. This approach is
in fact even more evident in the Peshitta. Although the Syriac version was
apparently translated from a Hebrew original, there are some similarities
to the LXX and it goes even further to spiritualise the text.54 The Aramaic
version, a much later production, on the other hand, is wholly interpreta-
tive in nature and, while maintaining the major themes of Qohelet, it is a
combination of literal translation and midrashic paraphrases in line with
the Babylonian rabbinic world view of circa seventh century.55
The value of the LXX Qohelet is really to be found in its adoption by the
Christian church when it became part of the canon of Holy Scriptures for
Christians in the Hellenistic world. Jerome recorded that he had made two
translations of Ecclesiastes from the Hebrew and that he had made exten-
sive use of the LXX as well as other Greek translations from the second
century.56 Gregory Thaumaturgos (ca. 213–270) produced the first Chris-
tian version of Ecclesiastes apparently largely based on the LXX. This work
is known as a paraphrase because of the effort made by Gregory, inspired
by his beloved mentor, Origen, to follow in his tradition of reinterpreting
and clarifying original and obscure sacred texts.57
54 Cf. Holm-Nielsen “Interpretation of Qoheleth,” 168–177.

55 Knobel, The Aramaic Bible, 9–15; E. Levine, The Aramaic version of Qohelet (New York:
Sepher-Hermon Press, 1978), 5–13.
56 Jerome, Commentarius, 1009.
57 Cf. Jarick, Gregory Thaumaturgos’ Paraphrase of Ecclesiastes.
James Barr said that words have meaning in specific contexts and in
relation to those who use them.58 Thus the translators were faced with
having to find suitable equivalents for hebel and other Semitic words
and phrases and select appropriate Greek equivalents without detracting
from their source texts. The translators of the LXX were aware of the com-
plexities associated with the lexeme hebel and were able to distinguish
its multiple uses.59 They recognised too that hebel was not restricted to
its elemental meaning of “vapour,” “wind” or “breath,” but was meant as
a symbolic and metaphorical device to produce a wider range of nuances
and meanings. This ultimately resulted in the necessity of making appro-
priate adaptations and additions that would have made sense to Greek-
speaking Jews.
The strength of the LXX is that it does not dilute the overall Judaic
nature of the book, albeit it lacking in direct historical references to Isra-
elite tradition. The LXX retains the integrity of Qohelet’s message as well
as, presumably, that of the Vorlage at its disposal, and it made the Greek
version accessible to “Jews who spoke Greek but did not read Hebrew
or Aramaic with a translation that would reflect the Hebrew as much as
possible.”60 The Greek translation recognised the central role played by
hebel throughout the book and used the Greek form consistently so as not
to detract from Qohelet’s central thesis. Hebel provided the means to link
the opening statement in 1:2 with the concluding one in 12:8 by using it
in appropriate contexts throughout the text. Hebel therefore became the
frame of reference by which Qohelet could apply his thought and belief
system to all of human experience.
58 J. Barr, The Semantics of Biblical Language (London: Oxford University Press, 1961),
171.
59 For the multiple uses of hebel and a detailed exposition of its metaphorical and sym-
bolic place in Qohelet; see Miller, Symbol and Rhetoric.
60 Seow, Ecclesiastes, 7.
The Θεος and Κυριος Terms in the Isaiah Text and their
Impact on the New Testament: Some Observations
Peter Nagel
1. Introduction
According to the MT, the Hebrew deity is not only referred to using the
terms ‫אלוה‬, ‫ אלהים‬and ‫אל‬,1 but also named ‫ יהוה‬while being called ‫אדני‬.2
In addition, one should also consider the so-called “biblical” manuscripts
found in and around the Judean desert, in which the paleo-Hebrew terms,
such as ‫)אלהים( אלהימ‬, ‫)אל( אל‬, ‫ )יהוה( יהוה‬and ‫( אדוי‬Adonai),
among others, were used to represent the Hebrew deity.3 It goes without
saying that these terms had to be translated into the Greek frame of
conceptual reference; hence the Hebrew deity “became” θεός, was ‘named’
κύριος, while being ‘called’ δεσποτής. The latter is not to say that the
Hebrew deity was called “δεσποτής” as such, but according to the “rule
of thumb” accepted by scholars in general, these terms are considered to
1 Noteworthy is the philological and literary approach of A. Murtonen, A Philological

and literature treatise on the Old Testament divine names [El, Eloha, Elohim] and [Yahweh]
(Helsinki: Societas Orientalis Fennica, 1952); see also Gericke’s philosophical approach
towards El-ness in J. W. Gericke, “What is an ‫ ?אל‬A Philosophical Analysis of the Concept
of Generic Godhood in the Hebrew Bible,” OTE 22/1 (2008): 21–46.
2 The in-depth investigation of M. Rösel, Adonaj—warum Gott ‘Herr’ gennant wird (Göt-
tingen: Mohr Siebeck, 2000), is considered to be foundational regarding the “naming” and
“calling upon” the Hebrew deity using Adonaj; see also Rösel’s thorough yet condensed
contribution “‫‍אדו‌‌ן‬,” in Theologisches Wörterbuch zu den Qumrantexten ([ThWQ]; Bd. 1;
eds. H.-J. Fabry and U. Dahmen; Stuttgart: Kohlhammer, 2011); cf. M. Delcor, “Des diverses
manières d’écrire le tétragramma sacré dans les anciens documents hébraïques,” RHR 147
(1955): 145–173, as well as S. T. Byington, “‫אדני יהוה‬,” JBL 76/1 (1957): 58–59, who treats of
the attestation of ‫ יהוה‬and ‫ אדני‬in the Hebrew tradition.
3 Cf. J. P. Siegel, “The Employment of Paleo-Hebrew Characters for the Divine Names
at Qumran in the Light of Tannaitic Sources,” HUCA 42 (1971): 159–172; see also, what is
considered to be a “new” substitute for the Tetragrammaton, J. M. Baumgarten, “A New
Qumran Substitute for the Divine Name and Mishnah Sukkah 4.5,” JQR 83/1–2 (1992): 1–5.
Another valuable study in this regard is D. W. Parry, “4QSama and the Tetragrammaton,”
in Current Research and Technological Developments on the Dead Sea Scrolls: Conference on
the Texts from the Judean Desert, Jerusalem (eds. D. W. Parry and S. D. Ricks; Leiden: Brill,
1996), 106–125; cf. idem, “Notes on Divine Name avoidance in Scriptural Units of the Legal
texts of Qumran,” in Legal Texts and Legal Issues: Proceedings of the Second Meeting of the
International Organization for Qumran Studies, Published in Honour of Joseph M. Baumgar-
ten (eds. M. Bernstein, F. García Martinez and J. Kampen; Leiden: Brill, 1997), 437–439.
174 peter nagel
be the most suitable Greek equivalents for their corresponding Hebrew

counterparts. Hence, θεός = ‫אלהים‬, κύριος = ‫ יהוה‬and δεσποτής = ‫אדני‬.
The problem with such a “rule of thumb” presupposition is that one has a
tendency to overlook the discrepancies, even though they might seem to
occur infrequently. The question, is how valid and useful such a “one-to-
one” literary and conceptual transmission is? The intent with this study is
not to answer the latter question, but to make some preliminary remarks
based on observations on the use of the θεός, κύριος and δεσποτής terms
in the Isaiah text against its Hebrew backdrop.4 It would be beyond the
scope of this study to discuss the use of θεός, κύριος and δεσποτής in the
whole of the Isaiah corpus; the study would therefore limit itself to those
“out-of-the-ordinary” occurrences. Stated differently, those occurrences
that deviate from the so-called “rule of thumb” will be the focal point of
the investigation. The theory that the term θεός could have been the “first
suitable” Greek term used as equivalent for the Tetragram in particular
and for the Hebrew deity in general sprung from this; a theory introduced
by Traube;5 summarised, together with other theories, by Hurtado6; and
later re-interpreted by De Troyer.7
The first step in this investigation would be to briefly consider the
Vorgeschichte of these terms as per the Isaiah text; stated differently, one
is compelled to consider the state of a possible Hebrew Vorlage in this
regard.8 This section of the investigation will be followed by an overview
4 Numerous studies pertinent to the issue of the term κύριος as a reproduction of the
Tetragram have seen the light of day; the monumental work of W. W. Baudissin, Κυριος
als Gottesname im Judentum und seine Stelle in der Religionsgeschichte (ed. O. Eissfeldt;
Giessen: A. Töpelmann, 1926) as a standard reference work on the term κύριος deserves to
be named in particular; for references made to the Hebrew deity in the Greek Psalter, see
H. U. Steymans, “Die Gottesbezeichnung kyrios im Psalter der Septuaginta,” L’Ecrit et
l’Esprit: Etudes d’histoire du texte et de théologie biblique en hommage à Adrian Shenker
(OBO 214; eds. D. Böhler, I. Himbaza and P. Hugo; Fribourg / Gottingen: Academic Press /
Vandenhoeck & Ruprecht, 2005), 325–338.
5 L. Traube, Nomina Sacra: Versuch einer Geschichte der christlichen Kürzung (München:
C. H. Beck, 1907).
6 L. W. Hurtado, “The Origin of the Nomina Sacra: A Proposal,” JBL 117/14 (1998): 655–
673; see the summary on pp. 664–671.
7 K. de Troyer, “The Pronunciation of the Names of God,” in Gott Nennen (eds. I. U.
Dalferth and P. Stoellger; Tübingen: Mohr Siebeck, 2008), 145–172.
8 For a thorough and comprehensive investigation into the Textgeschichte of the Isaiah
text, see A. van der Kooij, Die Alten Textzeugen des Jesajabuches: Ein Beitrag zur Textge-
schichte des Alten Testaments (Göttingen: Vanderhoeck & Ruprecht, 1981); see also idem,
“Isaiah in the Septuaginta,” in Writing and Reading the Scroll of Isaiah (eds. C. C. Broyles
and C. E. Evans; Leiden / New York / Köln: Brill, 2008), 513–529 (517); cf. the work of Rösel,
Adonaj, esp. 78–124 with regard to Isaiah; J. Lust in turn investigated “The Divine Titles
‫ האדון‬and ‫ אדני‬in Proto-Isaiah and Ezekiel,” in Isaiah in Context: Studies in Honour of
the θεος and κυριος terms in the isaiah text 175
of possible tendencies where “disregarding” the “rule of thumb” would

demonstrate to be most prolific.9 The third and final step will reflect on
the impact these discrepancies might have had on the New Testament
authors. The comparison of the “Old” Greek (hereafter OG) Isaiah with
the MT10 as well as witnesses from the Judean desert11 will be done on a
textual level.
2. Vorgeschichte—The Hebrew Vorlage
There are principally three terms used in the Hebrew dialect when and
if reference is made to an omnipotent, transcendental, all-powerful being
(translated in virtually all English translations as “God” with a capital “G”);
they are ‫אל‬, ‫ אלהים‬and ‫אלוה‬.12 The Isaiah text is no exception; the term
‫ אלהים‬is used in 94 instances consisting of 84 phrase structures of which
most occur in Isa 21–66; the term ‫ אל‬is attested in more than 10 verses;13
while ‫ אלוה‬is deployed only once, in Isa 44:8. The term ‫ אלהים‬is used
Arie van der Kooij on the Occasion of his Sixty-Fifth Birthday (eds. M. N. van der Meer
et al.; Leiden: Brill, 2010), 131–149. For an investigation into the translation of the divine
name limited to the Pentateuch see M. Rösel, “The Reading and Translation of the Divine
Name in the Masoretic Tradition and the Greek Pentateuch,” JSOT 31/4 (2007): 411–428.
For “names” of deities in the Ancient Near East with specialized focus on Mesopotamia
as a region, with Sumerian and Akkadian as relevant dialects, see C. Uehlinger, “Arbeit
an altorientalischen Gottesnamen—Theonomastik im Spannungsfeld von Sprache, Schrift
und Textpragmatik,” in Gott Nennen (eds. I. U. Dalferth and P. Stoellger; Tübingen: Mohr
Siebeck, 2008), 23–72. In turn, F. Hartenstein, “Die Geschichte JHWHs im Spiegel seiner
Namen,” in Gott Nennen (eds. I. U. Dalferth and P. Stoellger; Tübingen: Mohr Siebeck, 2008),
73–96, focused his attention on the history of YHWH as “name” of the Hebrew deity (cf.
W. H. Brownlee, “The Ineffable Name of God,” BASOR 226 [1977]: 39–46) counterbalanced
by the treatment of Elohim as the so-called “name” for the Hebrew deity. He does this by
probing semantics and grammatical characteristics, among others. For “Gottesnamekri-
terium” see E. Blum, “Der vermeintliche Gottesname ‘Elohim’,” in Gott Nennen (eds. I. U.
Dalferth and P. Stoellger; Tübingen: Mohr Siebeck, 2008), 97–119. For an investigation of
the “substitutes” of the Tetragram see J. Z. Lauterbach, “Substitutes for the Tetragrammaton,”
PAAJR 2 (1930–1931): 39–46.
9 Cf. De Troyer, “Pronunciation,” 144–150, who offers a brief but thorough overview
if and to what extent the Tetragram was pronounced or not and how the oldest Hebrew
manuscripts assist in determining this.
10 The MT as represented by the Biblia Hebraica Stuttgartensia: SESB Version (Stuttgart:
German Bible Society, 2003).
11 As confirmed by A. Aejmelaeus, “What can we know about the Hebrew ‘Vorlage’
of the Septuagint,” ZAW 99/1 (1987): 58–89 (58), the manuscripts found in and around
the Judean desert assist us in getting closer to how the OG’s Hebrew Vorlage might have
looked.
12 Gericke, “What is an ‫ ”?אל‬21.
13 Cf. Isa 8:5; 10:21; 12:2; 40:18; 42:10; 44:10, 15; 45:14, 15, 20, 21; 46:6, 9.
176 peter nagel
Reference 1QIsaa MT Notes LXX

3:15a ‫יהוה‬ ‫אדני‬ Superscript 1 ‫ אדני‬QIsa a
–
3:15b ‫יהוה‬ –
3:16 ‫יהוה‬ ‫יהוה‬ κύριος
3:17a ‫אדני‬14 ‫אדני‬ Superscript ‫ יהוה‬in 1QIsaa ὁ θεός
3:17b ‫אדני‬ ‫יהוה‬ κύριος
3:18 ‫יהוה‬ ‫אדני‬ Superscript ‫ אדני‬in 1QIsaa κύριος
in correlation with ‫ יהוה‬in more than twenty instances,15 with the term
‫ אדני‬deployed in 48 verses in comparison to ‫ יהוה‬utilised in 450 verses,
occurring 394 times.16 In addition to the use of the term ‫ אדני‬in Isaiah,
the author(s) also made use of the term ‫אדון‬.17
If one compares the MT with 1QIsaa and other related manuscripts
found in and around the Judean desert, the Hebrew text tradition appears
for the most part intact. There are, however, some small variations which
require reflection. The variants found in Isa 3:15–18 are a classic example
of Hebrew variants used to represent the Hebrew deity. The table below
presents the Hebrew variants in comparison to their Greek equivalents.
The data suggest that the scribes responsible for 1QIsaa as well as the
Masoretes appear inconsistent in applying the terms used when referring
to the Hebrew deity. The underlying issue at hand is one definable as
the Ketib-Qere problem, which requires some clarification. The standard
explanation is represented by Rösel; the Masoretes vocalised ‫ יהוה‬with
the vowels assigned to ‫אדני‬. The latter “forced” the reader to pronounce
(Qere—what ought to be read) against what was written (Ketib—what
ought to be written).18 The exception would be that if and when ‫אדני‬
‫ יהוה‬is written in combination, the term ‫ יהוה‬would be vocalised to read
‫אלהים‬, this would counter the duplicate reading of Adonai Adonai.19 An
opposing stance on this matter, of which De Troyer is a representative, is
that the most “usual” form of the Tetragram in Codex Leningrad as well
as in Codex Aleppo testifies to ‫( יְ הוָ ה‬shema—what ought to be read) and
14 4QIsab (4Q56) reads ‫אדני‬.

15 Cf. Isa 7:11; 17:6; 21:10, 17; 24:15; 25:1; 26:13; 35:2; 36:7; 37:42, 16, 20, 21; 40:28; 41:13; 48:1;
17; 49:4, 5; 51:15, 22; 55:5; 60:9.
16 The use of ‫ אדון‬and ‫ אדני‬in relation to ‫ יהוה‬and ‫ צבאות‬in Isaiah has been covered
for the most part by Rösel, Adonaj, 78–124; see also Lust, “The Divine Titles”, 131–149.
17 Cf. Rösel, Adonaj, 79.
18 Cf. Rösel, Adonaj, 2.
19 Ibid., 3; cf. De Troyer, “Pronunciation,” 144–145.
not ‫( יְ הו ֺה‬Αdonai), implying that the vowels adopted from the Aramaic
‫ ְׁש ָמא‬indicated what ought to be read;20 even though there are exceptions
to the rule.21
Returning to Isa 3:15–18, it is thus reasonable to assume that a redac-
tor of 1QIsaa wanted to make sure that the Tetragram in Isa 3:15a is pro-
nounced Adonai while the Masoretes “wrote” what they in all probability
heard, but wrote ‫ יהוה‬as an indication of what was implied by what was
read.22 Isa 3:17a in turn seems to indicate that the Masoretes copied what
ought to be read, while Isa 3:17b testifies to the fact that they interpreted
the term ‫( אדני‬1QIsaa) as an indication of what ought to be read, but wrote
what should be written. The reverse is again visible in Isa 3:18, with 1QIsaa
bearing witness to the Ketib form ‫יהוה‬, while the redactor of 1QIsaa indi-
cated what ought to be read, ‫אדני‬. In Isa 3:18 the Masoretes thus “inserted”
‫ אדני‬into the main body of the text either based on the superscript or
they wanted the Ketib form to be representative of the Qere form;23 the
latter explanation could be rejected as mere speculation. It could also be
argued that a Hebrew Vorlage was available to the Masoretes from where
they copied the text verbatim. A similar attested issue is found in Isa 28:16,
where the MT reads ‫ אדני יהוה‬while 1QIsaa attests to ‫ יהוה‬with a super-
script ‫אדני‬.24 The ‫ יהוה־אדני‬alternating readings continue in Isa 28:22
with 1QIsaa reading ‫ יהוה‬in comparison to the MT reading ‫אדני‬. Further-
more, the MT appears to have “ignored” ‫ יהוה‬in Isa 30:19 while 1QIsaa does
indeed read the Tetragram.25 These ‫ יהוה־אדני‬alternating variants, partic-
ularly attested to in Isa 3:15–18 (1QIsaa), confirm and reinforce Rösel’s posi-
tion that the vocalisation of Adonai testifies what ought to be read when
20 De Troyer, “Pronunciation,” 145.

21 E.g. Exod 3:2 (ΜΤ).
22 Cf. De Troyer, “Pronunciation,” 144.
23 Cf. D. Trobisch, Die Endredaktion des Neuen Testaments: Eine Untersuchung zur Ent-
stehung der christlichen Bibel (Göttingen: Vanderhoeck & Ruprecht, 1996), 21 n. 19. Note-
worthy is the fact that the Greek counterpart of Isa 3:18, as represented by the LXXGött, does
not attest to any equivalent of these terms, while some Greek manuscripts read κύριος
κύριος (cf. [※ 22-48-763-96] φησι[ν] κυριος κυριος [> oII 233]) others read πιπι πιπι (cf. ※
φησι κυριος κυριος [adn. πιπι πιπι]). Similar cases of Ketib-Qere “confusion” are detectable
in Isa 6:11; 7:14; 8:7 (‫ אדני‬as superscript in 1QIsaa with a probable reading of ‫ ;)יהוה‬Isa 9:7;
21:16; 28:2, 16; 30:15; 49:7; 61:1.
24 1QIsab however, appears to be closer to the MT with the second ‫ יהוה‬reading, with
an uncertainty of what term is to be read in the first instance. Isa 30:15 attests to a similar
issue; 1QIsaa again reads ‫ יהוה‬with ‫ אדני‬superscript, compared to 4QIsac most probably
reading ‫יהוה יהוה‬.
25 Additional discrepancies are found in Isa 9:7, where 1QIsaa reads ‫ יהוה‬as opposed to
‫ אדני‬attested in the MT. A similar case is found in Isa 28:2.
178 peter nagel
‫ יהוה‬was written.26 However, such alternating readings are very limited

and should thus not be taken as the “standard” practice of the time. The
Ketib-Qere problem surrounding the “naming” of the Hebrew deity might
not have been a case of reading aloud ‫ ְׁש ָמא‬, ‫ אלהים‬or ‫ אדני‬for ‫ ;יהוה‬but
it is indeed plausible that both practices could have been employed simul-
taneously by different scribes or scribal groups. It is nevertheless clear
that “naming” or making reference to the Hebrew deity was a complex
matter, at least from the 3rd century B.C.E. onwards.
Variant readings revolving around the term ‫ אלהים‬also occur; 1QIsaa (Isa
37:20) has ‫ יהוה אלהים‬while the MT only reads ‫יהוה‬.27 Furthermore, Isa
49:14 (1QIsaa) reads ‫ יהוה ואדוני‬with a superscript ‫ ואלוהי‬directly above
‫ ואדוני‬presumably implying that Elohim is to be read which would sup-
port the argument that if and when ‫ יהוה‬and ‫ אדני‬is consecutively read,
‫ יהוה‬should be pronounced ‫ אלהים‬to avoid the repetition of Adonai.28 A
slightly different but related issue is the MT reading in Isa 50:5 attesting
to both ‫ אדני‬and ‫ יהוה‬compared to 1QIsaa reading ‫ ;אדני אלהים‬the latter
seems to indicate that the Masoretes wrote what they considered to be
an indication of what ought to be read with the term ‫ אלהים‬in 1QIsaa.29
Finally, Isa 61:1 and Isa 61:11 testify to interesting variants, presented in the
table below.
Ref 1QIsaa 1QIsab 4QIsam MT

Isa 61:1 ‫יהוה אלהים‬ ‫יהוה אלהים‬ ‫אד‬ ‫אדני יהוה‬
Isa 61:11 ‫יהוה אלהים‬ – – ‫אדני יהוה‬
Thus, both 1QIsaa as well as the MT were consistent in applying the same
terms in Isa 61:1 and Isa 61:11. The only plausible assumption one could
draw from the data is that 1QIsaa and 1QIsab present a text tradition,
opposing the text tradition offered by 4QIsam and the MT, if a ‫אדני יהוה‬
reconstruction for 4QIsam is accepted. The Greek text traditions might
shed some light on the matter, and will consequently be considered in the
section to follow.
26 Cf. Rösel, Adonaj, 2ff.

27 The Greek and Syriac equivalents in turn only account for the term ‫אלהים‬.
28 Both 1QIsab and 4QIsab (4Q56) do not attest to any superscript.
29 Cf. Isa 54:6, which differs with regard to ‫( יהוה אלהים‬1QIsaa) and ‫( אלהים‬MT).
3. The κύριος and θεός terms: an overview
The Greek manuscripts, dated to the 2nd century B.C.E. and 2nd century
C.E., testifies to the following terms as “suitable” Greek equivalents for rep-
resenting the Hebrew deity:30
– ΙΑΩ (4QpapLXXLevb)31—Tetragram representation

– θεος P.Fouad 266b—Elohim representation
– blank space P.Fouad 266b—most probably Tetragram representation
– ‫ יהוה‬θεος P.Fouad 266b—Tetragram and Elohim representation
respectively
– ‫ יהוה‬8ḤevXIIgr—Tetragram representation
– θεος κυριος P.Oxy 656—Elohim and Adonai representation
The data snippet above presents an array of Greek possibilities in repre-

senting the Hebrew deity; the variations and discrepancies found in the
Greek version(s) of Isaiah also confirm the complexity in deciding on a
“suitable” Greek equivalent for a Hebrew term used when representing
the Hebrew deity. Sensitivity for this complexity intensifies with the reali-
sation that even the Hebrew text tradition struggled with finding ways
and means to refer to or “name” the Hebrew deity. The data from OG
Isaiah, or rather, the data inferred from a reasonable critical construction
of the OG Isaiah,32 reveals that the term δεσποτής is used as Greek equiva-
lent for ‫ אדון‬in Isa 1:24; 3:1 and 10:33,33 with the term κύριος utilised as a
representation of ‫ אדני‬in a relatively large number of cases.34 Moreover,
the term θεός is used, not without exceptions, as an equivalent for ‫אלהים‬,35
as well as for ‫אל‬36 while the term κύριος is a rendition of ‫יהוה‬. Below is a
list of some supplementary characteristics and peculiarities:
30 Cf. De Troyer, “Pronunciation,” 160–161.

31 Cf. Rösel, Adonaj, 4–5 and De Troyer, “Pronunciation,” 153.
32 J. Ziegler, Septuaginta: Vetus Testamentum Graecum, XIV: Isaias (Göttingen: Vanden-
hoeck & Ruprecht, 1983).
33 The term δεσποτής is also used, given the presumptions, as equivalent for ‫ אדני‬in
Gen 15:2, 8; Jer 1:6; 4:10; as well as ‫ אלהים‬in Job 5:8b and for ‫ יהוה‬in Jer 15:11.
34 Cf. Isa 4:4, 6:1, 8, 11; 7:14, 20; 8:7; 9:8(7); 10:12; 11:11; 21:6, 8, 16; 24:2; 26:13b; 28:2; 29:13;
30:20; 36:12; 38:14, 16; 49:14.
35 Cf. Isa 2:3; 5:16 (‫ ;)אל‬7:11; 10:21 (‫ ;)אל‬12:2 (‫ ;)אל‬13:19; 14:13 (‫ ;)אל‬17:6, 10; 21:10, 17; 25:1, 9.
36 Cf. Isa 5:16; 8:8; 10:21.
180 peter nagel
a.) Isaiah OG (Isa 4:5; 5:13; 16:14; 25:9b; 27:3; 28:21; 30:32; 41:16; 45:24; 53:1;
55:5; 58:5; 59:13) ≠ The MT reads ‫;יהוה‬37
b.) Κύριος ὁ θεός in OG Isaiah = ‫ יהוה‬MT (Isa 17:6; 24:15; 25:1; 26:13; 30:18;
37:4, 20, 21; 38:5; 41:21; 42:5, 6, 8, 13, 21; 43:1, 10, 11, 12, 14, 15; 44:2; 45:1, 3,
5, 6, 7, 11);38
c.) Isaiah OG reads the term θεός ≠ MT (Isa 9:11, 17; 13:6; 23:16; 26:9, 12;
28:13b; 41:17c; 45:23, 25; 49:7;39 the opposite is in turn attested in Isa
21:9; 53:4;
d.) The θεούς = ‫( אליל‬MT);
e.) Κύριος ὁ θεός ≠ in MT (Isa 24:16; 27:4);40
f.) Κύριος = ‫( צור‬Isa 17:10);
g.) The term κύριος in Isa 40:18 = ‫ אל‬in the MT;41
h.) The 1st κύριος term in Isa 54:5 = ‫ בעל‬in the MT;
i.) The OG Isaiah ≠ The MT, the latter testifies to the term ‫ אלהים‬in Isa
55:7; 61:2.
3.1. The Term κύριος

The term κύριος dominates the OG Isaiah, with 422 occurrences in 365
verses, 5 times of which κύριος is used in the vocative case in relation to
the term θεός with its accompanying definite article;42 in 30 instances it
is used in the nominative case in combination with the term θεός.43 Cau-
tion is duly noted when concordance related statistics are given such as
being offered here. The statistics should not be considered as necessar-
ily “argumentatively sound data”, but merely introductory in nature. The
first noteworthy issue brought to the fore by the term κύριος is found in
Isa 3:15–18; it appears as if the Greek scribes “ignored” ‫אדני יהוה צבאות‬
37 In Isa 12:2; 37:14 the MT testifies to two ‫ יהוה‬terms, compared to one κύριος term each
in both the references. The opposite is true in Isa 37:6 offering two κύριος terms weighed
against the term ‫ יהוה‬by the MT; and in Isa 57:15 two κύριος terms are read in the LXX
compared to no reference made in the MT or any other Hebrew text tradition for that
matter; cf. Isa 58:6; 61:1 (2nd κύριος term); 63:7 (4th κύριος term); 64:2.
38 This “equivalent” was consistently used, at least in Isa 41–45.
39 In Isa 37:19 the OG Isaiah reads εἴδωλα as equivalent for ‫אלהים‬, while the Greek does
not contest the term ‫ אלהים‬in Isa 60:9.
40 Another variant is attested in Isa 41:21, where the MT reads ‫ יהוה‬with κύριος ὁ θεός as
the Greek equivalent; cf. Isa 42:5, 6, 8, 13, 21; 43:1, 10, 11, 12, 14, 15; 44:2; 45:1, 3, 5, 6, 7, 11.
41 Cf. Isa 7:13 representing the term κύριος as equivalent for ‫אלהים‬.
42 Cf. Isa 24:16; 25:1; 26:12, 13 and Isa 37:20.
43 Cf. Isa 8:10; 17:6; 27:4; 30:18; 37:4, 21; 38:5; 41:17, 21; 42:5, 6, 8, 13, 21; 43:1, 3, 10, 12, 14, 15;
44:2; 45:1, 3, 5, 6, 7, 11; 51:22; 52:12 and 57:21.
(Isa 3:15).44 Greek Isaiah continues opposing the “rule of thumb”; in Isa
3:17a with ὁ θεός seemingly representing ‫( אדני‬MT, 4QIsab) or ‫ אדני‬with
superscript ‫( יהוה‬1QIsaa). One could infer from the evidence that, as the
term θεός is used to render ‫ יהוה‬in an overwhelming amount of cases, ὁ
θεός probably represents the 1QIsaa reading (‫ אדני‬with superscript ‫)יהוה‬
in Isa 3:17a. The term κυριός in Isa 3:17b appears to be the equivalent for
either ‫( אדני‬1QIsaa) or ‫( יהוה‬MT), with the term κύριος again utilised to
represent either ‫( אדני‬MT) or ‫ יהוה‬with superscript ‫( אדני‬1QIsaa) in Isa
3:18. The reason for this “unusual” and inconsistent representation could
be argued from both a theological as well as a literary stance.
There appears to be no obvious theological cue that supports a theo-
logically determined reason why ὁ θεός was chosen as an equivalent for
the terms in question (cf. Isa 3:17a). A literary induced reason, in combina-
tion with a theological-conceptual motivation is that the superscript ‫יהוה‬
is best rendered using ὁ θεός. The latter term would call the ‘‫אל‬-ness’ of
the Hebrew deity to mind, a term that would encapsulate an overarching
monotheistic deity, best represented by ‫אל‬, ‫ אלהים‬or ‫אדני יהוה צבאות‬,
none of which are present in the Hebrew of Isa 3:17a. The Hebrew text
does however attest to ‫ אדני יהוה צבאות‬in Isa 3:15, offering the transla-
tor ample opportunity to use ὁ θεός; instead, these terms are ironically
ignored. It is, also possible that the translator completely ignored his Vor-
lage and followed his own conceptual “creativity”.
The avoidance or ignorance of the term ‫ אדני‬introduced in Isa 3:15–18,
whereby a single κύριος term, or any other related term for that matter, is
used as a Greek equivalent for ‫אדני יהוה‬, is not confined to Isa 3:15–17.45
Another intriguing Hebrew text tradition that requires due consideration
in its Greek form is Isa 61:1 and Isa 61:11 (the table below summarises the
various readings).
44 Some Greek manuscripts read an additional κυριος κυριος (※ 22-48-763-96) φησι(ν)

κυριος κυριος (> oII 233) στρατιων V-oI’ L′’`-46–233 C 403′ 770 Chr.Tht. = 𝔐).
45 Cf. Isa 7:7; 22:5, 12 (B-Qmg-109 403′ 538 read κυριος κυριος which is considered to be
“closer” to the MT; minuscule 91 in turn reads κυριος ο θεος); 28:16, 22 (+※ κυριου QSyh
86); 40:10 (The hexaplaric tradition, together with Eusebius, accounts for two κυριος terms;
whereas Hieronymus “adds” deus equivalent for the term θεός); 49:22; 50:4, 5, 7, 9 (The BHS
text-critical apparatus suggests that the LXX, Ethiopian and Arabic versions do not read
an equivalent term for ‫אדני‬. The textual evidence from the Judean desert (1Q8 Isaiah and
1QIsaa) supports the MT text reading); 52:4; 56:8; 61:1, 11; 65:13 and Isa 65:15; Cf. Isa 36:9
where the Hebrew tradition reads ‫ אדני‬compared to the absolute silence of the Greek
tradition.
182 peter nagel
Ref 1QIsaa 1QIsab 4QIsam MT LXXGött Qmg ʘ

Isa 61:1 ‫אד יהוה אלהים יהוה אלהים‬ ‫ אדני יהוה‬κύριος κύριος
κύριος
Isa 61:11 ‫– יהוה אלהים‬ – ‫ אדני יהוה‬κύριος
If the “rule of thumb” presupposition is accepted as “reasonable”, then

one could assume that in this case the term κύριος represents the read-
ing of 1QIsaa best. When dealing with the Isaiah text, however, the more
argumentatively sound inference is that the only “rule of thumb” to be
adopted in general is that there are not any.46 The latter is confirmed
when one considers Isa 61:11 in relation to Isa 61:1a; the former testifies to a
single κύριος term as opposed to ‫( אדני יהוה‬MT), while 1QIsaa reads ‫יהוה‬
‫אלהים‬. Thus, in Isa 61:1a the term κύριος best represents either the MT
(used once to avoid repetition of the term ‫)אדני‬, or 1QIsaa. If the scribes
are considered to be consistent, the term κύριος best represents the MT in
Isa 61:11. One thing that could be deduced from the evidence is that LXXGött
seems to have simplified the matter by consistently reading a single term
κύριος, while manuscripts belonging to the hexaplaric tradition in general
follow the MT, as expected.
Another captivating issue is the “omission” of both the ‫ יהוה‬terms,
including ‫ צבאות‬in Isa 22:14a and ‫ צבאות‬. . . ‫ אדני‬in Isa 22:14b.47 The lat-
ter two occurrences do indeed strengthen the idea that the term ‫ אדני‬is to
a large extent “downplayed” by the translators of OG Isaiah. Isa 49:14 (‫יהוה‬
‫ )אדני‬offers another twist to the ‫—אדני יהוה‬κύριος dilemma; the LXXGött
renders κύριος ὁ κύριος as equivalent with an array of alternatives offered
by manuscripts, varying between the term θεός in the first instance,48 while
one is expected to read the term θεός in the second instance.49 The avoid-
ance of the term ‫אדני‬, especially when used in combination with ‫יהוה‬, is
wonderfully illustrated in the catena of occurrences found in Isa 50:1–10
(Isa 50:4, 5, 7 and 9); all the ‫( אדני יהוה‬MT) occurrences are rendered
46 Van der Kooij (“Isaiah in the Septuaginta,” 518) confirms that it is generally accepted
that the LXX Isaiah follows a more “free” approach towards the original, while mentioning
that Ziegler and Seeligmann are of the opinion, an opinion he supports, that the Vorlage
of the LXX Isaiah does not differ much from the MT. These statements might appear to
be confusing; why would a solid Hebrew text tradition cause such a “loose” or “free” Greek
equivalent, especially with theologically significant terms such as discussed here?
47 The “omission” of ‫ אדני‬in Isa 22:14b is also attested in the Syriac as well as in
Theodotion.
48 κύριος1°] ο θεος 538.
49 κύριος2°] θεος A′ 88–oII L′’`-86c-233 564 403′ 534.
with a single term κύριος. These instances, together with others, seem
to indicate a “discomfort” with the term ‫ אדני‬on the part of the Greek
translators, the reason of which is currently unknown and thus demands
further investigation.
2.2. The Term θεός

The term θεός occurs 195 times in 174 verses; in almost all cases it is accom-
panied with a definite article and is a term that implies the Hebrew deity
“proper”, except for its nominative plural use in Isa 37:19; 41:23; 42:17 and
the accusative plural use in Isa 44:15, the latter which refers to deities in
general, not the Hebrew deity in particular. The theory that the term θεός
was considered to be the most suitable Greek equivalent to represent ‫יהוה‬
appears to be a plausible one, at least in OG Isaiah. The data produced
from a “surface” comparison between the Hebrew and Greek text tradi-
tions of Isaiah, indicates that the term θεός is used, in a large number of
cases, as an equivalent for the Tetragram.50 Because the “general rule of
thumb” states that θεός is the Greek equivalent for ‫אלהים‬, the term is
conceptually devaluated as a reasonable candidate to represent ‫יהוה‬. A
channel of conceptual possibilities is opened up when θεός is seen as a
significant Greek equivalent for the Tetragram from a very early stage in
the Greek text tradition; a theory that would require a broader investiga-
tion into how the Tetragram was rendered elsewhere. For now the focus
will be confined to the Isaiah text.
Ref 1QIsaa MT LXXGött Sc 90–456 S* 36 O’-Qmg oII Rom

377–cII 9:28
393 403′
Isa ‫אדני‬ ‫ אדני‬ὁ θεος κύριος ὁ κύριος κύριος
10:23 ‫יהוה‬ ‫יהוה‬
Isa ‫אדני‬ ‫ אדני‬κύριος κύριος ὁ θεός
10:24 ‫יהוה‬ ‫יהוה‬ κύριος
Isa ‫יהוה‬ ‫ יהוה‬ὁ θεος
10:26
50 Cf. Isa 1:10; 2:2 4:2; 6:12; 7:17; 8:17, 18; 9:11(10); 10:20, 23, 26; 11:2, 3; 13:6; 14:2, 3, 5, 27;
24:21; 25:10; 26:4b; 27:1; 30:9, 18, 29; 31:1; 33:5, 22; 36:15, 18, 20b; 37:20b, 22; 38:7, 20b, 22; 40:7,
27, 28, 31; 41:4, 14; 42:12, 19, 24; 44:5, 6, 23; 48:17b; 51:13; 54:5, 13; 55:6; 58:8, 9, 11, 13; 61:2; 65:23.
The reverse is attested in Isa 7:13, where the term κύριος appears to be an equivalent for
‫ ;אלהים‬cf. Isa 61:10; 62:5.
184 peter nagel
In Isa 10:23 the LXXGött reads ὁ θεός as an equivalent for ‫ אדני יהוה‬supported
by the first hand of codices A and S, while codex B reads the term κύριος.
Other Greek manuscripts alternate between two possibilities; κύριος sup-
ported by some manuscripts belonging to the hexaplaric tradition,51 while
others include the definite article.52 Isa 10:24 in turn opted for the term
κύριος as an equivalent for ‫אדני יהוה‬. This reading is opposed by the first
hand of codex S which reads κυριος ο θεος, supported by some manuscripts
belonging to the hexaplaric tradition.53 A corrector of S offers another
possibility, reading κυριος κυριος.54 In Isa 10:26, the translators again opted
to use ὁ θεός to render ‫יהוה‬. In all these cases, the Hebrew text tradition
is supported by 1QIsaa. The “out-of-the-ordinary” cluster of attested terms
found in ch. 14 (Isa 14:2, 3, 5 and 27) all read the term θεός compared to the
MT which only attest to the term ‫יהוה‬. There seems to be no evidence to
suggest the contrary, showing that the Hebrew text tradition is intact. Isa
26:12 testifies to κύριος ὁ θεός as equivalent for ‫יהוה‬, the only occurrence
of its kind in the LXXGött.55 In 26:13, however, the MT reads ‫יהוה אלהים‬
with the expected Greek counterpart κύριος ὁ θεός. The reading in Isa 26:12
only makes sense if a “being consistent” argument is adopted to explain
the unusual Greek equivalent for the Tetragram. The reading κυρίου τοῦ
θεοῦ in Isa 28:13 is repeated only in Isa 51:20 in the whole of the LXXGött,56
while some manuscripts (Is 28:13) in the hexaplaric tradition together
with the Lucianic tradition calls for an “omission” of τοῦ θεοῦ,57 with the
catena manuscript group supporting the “omission” of κυρίου.58 A further
noteworthy case is the θεός reading in Isa 37:20b; the MT reads ‫ יהוה‬with
1QIsaa testifying to ‫יהוה אלהים‬, the latter which opens the possibility that
the 1QIsaa reading’s ‫ אלהים‬could present a possible alternative Hebrew
Vorlage for the LXXGött reading if the theory is not accepted that the term
θεός was considered to be the most “suitable” Greek equivalent for ‫יהוה‬.59
51 O’-Qmg, a reading that is confirmed by the Syro-Palestinian translation and the church
fathers Eusebius, Basilius and Tertullian.
52 (ο κυρ. oII).
53 + ο θεος S* 36 377–cII 393 403′; A 2nd hand of codex S reads κς ο κς while a 3rd hand
opted for κς κς. Some manuscripts belonging to the Lucianic tradition “omits” the definite
article.
54 + (※) κυριος Sc 90–456 = 𝔐 ↓.
55 Cf. Isa 17:6; 24:15; 25:1; 26:13; 30:18; 37:4, 20, 21; 38:5; 42:6, 8, 13, 21.
56 Cf. Isa 7:11; 48:1.
57 om. τοῦ θεοῦ O′’ L′’`-233-456 301 403′ 449′ 534 Eus.Tht.Hi. = 𝔐; om. τοῦ 393.
58 om. κυρίου C′’-566.
59 Contra Van der Kooij, “Isaiah in the Septuaginta,” 518–518. In Isa 41:13 one finds a
similar case, where the MT reads ‫ יהוה אלהים‬while the LXX only reads θεος; cf. Isa 51:15.
One would, however, have to justify why the translator would “discard”
‫יהוה‬. It is, probable that the translator did not consider it necessary to
reproduce an equivalent for both the Hebrew terms, as it was prohibited
to pronounce the former. The occurrence in Isa 38:11 seems to confirm the
theory that the term θεός is considered the most suitable equivalent for
‫ ;יהוה‬the MT reads ‫ יה יה‬with two Hebrew manuscripts testifying to ‫יהוה‬
supported by both Symmachus as well as the Syriac translation with the
LXXGött again reading θεός.
In Isa 41:17 κύριος ὁ θεός appears to represent ‫( יהוה‬MT) with no text
witnesses suggesting an alternative Hebrew reading,60 while κύριος ὁ θεός
reproduces ‫ האל יהוה‬in Isa 42:5; in the latter case an alternative Hebrew
text reading is found in 1QIsaa reading ‫אל אלהים‬.61 It is suggested that the
κύριος ὁ θεός reading in Isa 41:17 as well as Isa 42:5, among others, should
not be explained on account of its Hebrew Vorlage. One should rather
consider the κύριος ὁ θεός construction as theologically determined, espe-
cially in Isa 41–45. In these chapters, the κύριος ὁ θεός construction appears
to be governed by the first personal pronoun ἐγώ together with λέγει,62
which could imply that for the translators, if and when the Hebrew deity
declares or utters something in reference to the godly self, the existential
theological I, as logion, the theological maximus is required. If this is a
reasonable assumption, it fuels the proposal that neither κύριος nor θεός
are “suitable” Greek equivalents to render the Hebrew deity proper. This
theologically determined understanding of κύριος ὁ θεός is supported by
Isa 57:21 reading κύριος ὁ θεός in comparison to the MT which reads ‫אלהי‬
only.63
Another variant reading is attested in Isa 51:15; the term ὁ θεός with
a definite article appears to be the Greek equivalent for ‫יהוה אלהים‬64
where one would have expected a Greek equivalent κύριος ὁ θεός.65 In Isa
43:1 the Greek text reads κύριος ὁ θεός compared to the MT that only has
60 Some Greek manuscripts opted for ο αγιος (Q ʘ), while others “omitted” κυριος (198
538 Sa); cf. Isa 41:20.
61 Ms 51 “omits” κύριος while some manuscripts from the hexaplaric tradition, including
early Church fathers (oII 407 410 Ir. [hab. Ir.lat] Cyr.[lem]) in turn “omit” ὁ θεός.
62 A noteworthy exception is found in Isa 44:24, reading Οὕτως λέγει κύριος ὁ λυτρούμενός
σε καὶ πλάσσων σε ἐκ κοιλίας Ἐγὼ κύριος ὁ συντελῶν πάντα ἐξέτεινα τὸν οὐρανὸν μόνος καὶ
ἐστερέωσα τὴν γῆν; this phrase seems to suggest an opposing stance to what is offered as
an explanation for the frequent use of κύριος ὁ θεός in Isa 41–45.
63 1QIsaa, 1QIsab and 4QIsad confirm the Hebrew text reading.
64 Cf. Isa 41:13; 48:17; 55:5.
65 The Hebrew text tradition does not offer an alternative that would “justify” the cur-
rent Greek equivalent.
186 peter nagel
the term ‫יהוה‬. Why the inconsistency? In Isa 51:15 it is probably a case
of influence from the Vorlage; such a possibility is introduced by 4QIsac
which does not testify to the term ‫אלהים‬, but only to ‫יהוה‬, a term fre-
quently represented by the term θεός, as shown above. In the case of Isa
43:1, the reason behind the “out-of-the-ordinary” reading could be theo-
logical consideration, as was pointed out above. Finally, in Isa 60:9 “nam-
ing” the Hebrew deity is brought to the fore; according to OG Isaiah it is
the name κύριος that is deemed holy, while for the Hebrew text it is ‫יהוה‬
‫אלהים‬. Before moving on to how the variants, alternatives and peculiari-
ties could have impacted the New Testament authors, it might be worth-
while to examine how the OG Isaiah dealt with ‫ ִע ָ ּ֥מנּו ֵ ֽאל‬which came to
be known in the New Testament as Ἐμμανουήλ (Matt 1:23). The concept
‫ ִע ָ ּ֥מנּו ֵ ֽאל‬occurs only three times in the MT, all of which are confined to
Isaiah. In 7:14 the Hebrew terms are rendered with Ἐμμανουήλ. In Isa 8:8
however, the Hebrew concept is translated with μεθ᾿ ἡμῶν ὁ θεός. Finally,
in Isa 8:10 the phrase in turn is represented using μεθ᾿ ἡμῶν κύριος ὁ θεός;
three distinct Greek equivalents, one of which found its way into the New
Testament (Isa 7:14 cited in Matt 1:23a) and one cited in Matt 1:23b (Isa 8:8).66
The effect of this translation will be discussed in the next section of
this paper.
4. The New Testament impact
Reflecting upon the Isaiah text, its history within Christianity and its use
in the New Testament is by no means a new venture. A general overview
of the history of such influence was undertaken by John Sawyer in 1996,67
followed by a compilation of essays edited by Steve Moyise and Maarten
Menken on those New Testament books that contain Isaiah citations.68
Studies with a more focused and specialised approach are works such as
66 NA27 notes that the phrase μεθ᾿ ἡμῶν ὁ θεός is sourced from either Isa 8:8 or Isa 8:10,
which is technically incorrect; the reading in Isa 8:10 is clearly testifying to κύριος ὁ θεός
and not ὁ θεός.
67 J. F. A. Sawyer, The fifth gospel: Isaiah in the history of Christianity (Cambridge: Cam-
bridge University Press, 1996).
68 S. Moyise and M. Menken, eds., Isaiah in the New Testament: The New Testament and
Scriptures of Israel (London / New York: T & T Clark, 2005).
that of Koch,69 Stanley,70 Wilk,71 Shiu-Lun Shum72 and Wagner;73 all of

whom investigated in some way or form the use of Isaiah as scripture in
the Pauline literature. Peter Mallen considered the reading and transfor-
mation of Isaiah in Luke–Acts.74 The objective with this section of the
investigation is to consider the impact the Isaianic use of the θεός and
κύριος terms might have had on the New Testament authors, using Matt
1:23 (Isa 7:14 and Isa 8:8) and Rom 9:27–29 as test cases; the latter which
testifies to two explicit Isaiah citations, Isa 10:22–23 (Rom 9:27–28) and
Isa 1:9 (Rom 9:29).
4.1. Isa 7:14; 8:8 (Matt 1:23)

The impact of Isa 7:14 and 8:8 in the text of Matt 1:23 and the significance
thereof should not be underestimated. The introductory formula launch-
ing the citation reads: ἵνα πληρωθῆ τὸ ῥηθὲν ὑπὸ κυρίου διὰ τοῦ προφήτου
λέγοντος (“for the fulfilment of what kyrios said through the prophet”).
Kyrios is thus the primary acting agent, speaking through the prophet Isa-
iah, in this case, saying: ἰδοὺ ἡ παρθένος ἐν γαστρὶ ἕξει καὶ τέξεται υἱόν, καὶ
καλέσουσιν τὸ ὄνομα αὐτοῦ Ἐμμανουήλ (“behold, the virgin will have in her
womb and she will give birth to a son, and they will call him Emman-
uel”). The author continues by offering a translation of the Aramaic term
with the words: ὅ ἐστιν μεθερμηνευόμενον μεθ᾿ ἡμῶν ὁ θεός (“which is trans-
lated: Theos with us”). Interpreting these citations within their immedi-
ate thought-structure (Matt 1:18–25), the import of the κύριος-θεός citation
becomes evident. Logically inferred, it is Ἰησοῦ Χριστοῦ (Matt 1:18) who
should be named Ἰησοῦν (Matt 1:21) and Ἐμμανουήλ (Matt 1:23a); the lat-
ter holding the semantic potential of μεθ᾿ ἡμῶν ὁ θεός (Matt 1:23b). The
citation in Matt 1:23a causes both author and reader to draw a direct con-
ceptual link between Ἰησοῦν and Ἐμμανουήλ. The latter would supposedly
69 D.-A. Koch, Die Schrift als Zeuge des Evangeliums (Tübingen: Mohr Siebeck, 1986).
70 C. D. Stanley, Paul and the language of Scripture: Citation technique in the Pauline
Epistles and contemporary literature (Cambridge / New York: Cambridge University Press,
1992).
71 F. Wilk, Die Bedeutung des Jesajabuches für Paulus (FRLANT 179; Göttingen: Vanden-
hoeck & Ruprecht, 1998).
72 Shiu-Lun Shum, Paul’s use of Isaiah in Romans: a comparative study of Paul’s Letter to
the Romans and the Sibylline and Qumran sectarian texts (Tübingen: Mohr Siebeck, 2002).
73 J. R. Wagner, Heralds of the Good News: Isaiah and Paul in Concert in the Letter to the
Romans (Leiden: Brill, 2003).
74 P. Mallen, Reading and Transformation of Isaiah in Luke-Acts (London / New York:
T & T Clark, 2007).
188 peter nagel
not have been an obvious link, if the Aramaic ‫ׁשּוע‬

ַ ֵ‫י‬, semantically related
to ‫ׁשּועה‬
ָ ְ‫י‬, was considered and understood as a literary-conceptual fore-
runner of Ἰησοῦς with the potential meaning “helper, saviour”. If such an
assumption is accepted as plausible, then the citation taken from Isaiah
expands the possible meanings of the term Ἰησοῦν from the semantic pros-
pect of “saviour” to Emmanuel to the semantic value of Theos with us. This
expansion of the semantic field results in the conclusion: Ἰησοῦς is Theos
among us. It is doubtful that the author of the Matthean gospel could have
“creatively” conceptualised that Ἰησοῦς is Ἐμμανουήλ and therefore μεθ᾿
ἡμῶν ὁ θεός without the support of Isa 7:14 and Isa 8:8. It remains an open-
ended question, if and to what extent the Matthean author considered the
theological implication of his citation.
4.2. Isa 10:22c–23 (Rom 9:27–28)

The author creates a Χριστός—θεός theological frame of reference in Rom
9:1–25 with its theological thrust fixated in and around Rom 9:5.75 The fol-
lowing theos-kyrios concept (Rom 9:26–33) is made possible by the three
cited texts: Hos 2:1b–c (Rom 9:26), Isa 10:22c–23 (Rom 9:27–28) and Isa
1:9 (Rom 9:29).76 Although the attention of this study is on the impact
the Isaiah citations might have had on Pauline thought, it is worth noting
that the textual integrity of both Hos 2:1b–c and Rom 9:26 in terms of θεός
and κύριος appears to be intact.77 Literarily and theological-conceptually
speaking the term θεός in Hos 2:1b–c could not have referred to any other
entity than ‫אל‬, the “wisdom” deity of ancient Israel. The same, though,
cannot be said for Isa 10:22c–23 in Rom 9:27–28. A variety of possibilities
are presented by the Greek witnesses for both Old Testament (hereafter
OT) and New Testament (hereafter NT) texts.78 All the NT text witnesses
75 Rom 9:5 sets the literary context in which the intriguing thrust of the inter-relatedness
of the Χριστός and θεός is put to the fore.
76 According to H. Schlier, Der Römerbrief: Kommentar (Freiburg: Herder, 1977), 303,
Paul wants to appeal that the Gentiles belongs to the people of God with the Hos 2:1
citation; in turn, the Isa 10:22–23 confirms the true Israel as the Ekklesia. These citations
also confirm the sovereign action of God (ibid., 304). For Schlier (ibid., 304), it is evident
that ἐκεῖ κληθήσονται υἱοὶ θεοῦ ζῶντος (Rom 9:26) indicates Paul’s apposing stance over and
against the Jewish-Apocalyptic and Rabbinic position. Koch (Schrift, 146), in turn suggests
that the redaction of the Israelites in Isa 10:22 is portrait as a legal act of Yahweh.
77 The Hebrew text tradition reads ‫ אל‬with the Greek text tradition offering θεός as
equivalent.
78 The Hebrew text tradition seems to be intact for the most part, except for a text
critical note in the BHS on Isa 10:23 that states that two Hebrew manuscripts “omits” ‫יהוה‬
‫צבאות‬.
read κύριος, while the OT text witnesses vary between ὁ θεός, κύριος κύριος
and κύριος. From the textual evidence it can be deduced that the ‫ְצ ָבֹאות‬
term was either “ignored” from a very early stage of transmission or it had
been “omitted” by the Greek translators. The fact remains that ‫אדני יהוה‬,
when used in combination, often appears to have been contracted into
one single Greek equivalent. Paul’s “consistent” use of the term κύριος in
Rom 9:28 and Rom 9:29 as sourced from Isa 10:23 and Isa 1:9 respectively,
together with the fact that he had ample theologically determined reason
to “alter” the citation, points to the fact that he stringently followed his
Vorlage. If the term κύριος belongs to a Vorlage, the reading in Rom 9:28
would seem to be in line with some manuscripts from the hexapla. The
theological thrust of ‫“ אדוני ְצ ָבֹאות יהוה‬Lord God of Hosts” was already
watered down by the Greek translators from a very early stage because
they struggled to find a “suitable” Greek equivalent for these Hebrew
terms in combination. This struggle spilled over into the New Testament.
It is thus more than possible that “other” equivalents for these terms were
available, one of which might have primarily impacted Paul. This would
imply, taking Rom 9:26 which attests to the term θεός into consideration,
that in Rom 9:28 and Rom 9:29 Paul regarded the referent of both the
terms θεός and κύριος as one and the same entity: the Hebrew deity. Is
this really the case?
What then is the implication of these κύριος citations? Is Paul calling
Jesus to mind as the κύριος or as ‫ ?יהוה‬In Rom 9:8 Paul speaks of τοῦτʼ
ἔστιν, οὐ τὰ τέκνα τῆς σαρκὸς ταῦτα τέκνα τοῦ θεοῦ (“that is, not the children
of flesh are these children of Theos”),79 while he asserts in Rom 9:29 that
εἰ μὴ κύριος σαβαὼθ ἐγκατέλιπεν ἡμῖν σπέρμα (“if Kyrios, lord of hosts, did
not leave a remnant behind for us”). Both κύριος and θεός thus reserve the
right to decide which nation or clan to accept or deny, to include or to
exclude. Moreover, Paul also commands that ὁ λόγος τοῦ θεοῦ (Rom 9:6)
should not be considered invalid, for it will come to fulfilment once and
for all on earth (λόγον γὰρ συντελῶν καὶ συντέμνων ποιήσει κύριος ἐπὶ τῆς
γῆς [Rom 9:28]). The Hos 2:1b–c citation in Rom 9:26 reinforces the θεός
concept introduced in Rom 9:1–25; a concept that called the “supreme”
Hebrew deity to mind. The two κύριος citations (Isa 10:23 and Isa 1:9) seem
out of place. However, with these two citations, due to their overlapping
theme, Paul appears to regard the term θεός (Rom 9:6–8) and the term
κύριος (Rom 9:28) as having the same referent; the same could be said
79 Cf. Rom 9:7.

190 peter nagel
for Rom 9:29. It remains unclear if Paul adopted the “underlying” concept
of the term κύριος to be a rendering of the Tetragram, although it seems
quite plausible. Paul is thus not calling to mind Jesus as the κύριος, but the
authoritative nature of the Hebrew deity. These two Isaiah citations sug-
gest that the κύριος term within the explicit citations refers to a different
entity than the κύριος term which clearly refers to Jesus as the Christ. The
deployment of the two Isaiah citations would ultimately have an impact
on how one interprets the conceptual “relationship” between the terms
Χριστός and θεός in Rom 9:5. How do the quotations considered above and
the conclusions drawn about their impact reflect on the interpretation
and understanding of Rom 9:5?
The term Χριστός in Rom 9:5 should first and foremost be understood
in relation to the Χριστός terms in Rom 9:1 and Rom 9:3. In both these
instances the term Χριστός is presented within a prepositional clause:
Ἀλήθειαν λέγω ἐν Χριστῷ and ἐγὼ ἀπὸ τοῦ Χριστοῦ respectively. It should
further be noted that in both cases the first person singular pronoun in
the relevant clause, which refers to Paul, is used. In Rom 9:1 Paul’s truth is
considered to be justified ἐν Christ (Rom 9:1), while ἀπό refers to a second-
ary position over and against someone who is ἐν Christ (Rom 9:3).80 One
could thus, with a reasonable amount of certainty, conclude that the con-
cept underlying the term Χριστός in these verses are one of Christ being
a mediator. On the other hand, it would be difficult to deny that through
this mediating role, Χριστός, in the mind of Paul, should be praised as θεός,
who is θεός over all.81 This concept is strengthened when one considers
the idea that ἐγὼ ἀπὸ τοῦ Χριστοῦ ὑπὲρ τῶν ἀδελφῶν μου τῶν συγγενῶν μου
κατὰ σάρκα seamlessly fits into the concept of θεός’ free will to make or
appoint nations, clans or any group as “his sons” or “his children” as is evi-
dently assigned to both θεός and κύριος in ch. 9. Not only is θεός and κύριος
conceptually considered terms referring to the Hebrew deity, but in this
case the Χριστός term also belongs to this concept. The impact of the term
κύριος in both Isa 1:9 and Isa 10:23 leads one to deduce: (1) Paul stringently
follows his Vorlage. (2) He considered both the terms θεός and κύριος as
terms that “translate” or “represent” the Hebrew deity, even though this
is not always the case.
80 Cf. E. Käsemann, Commentary on Romans (London: SCM Press, 1980), 259.

81 To quote H.-C. Kammler, “Die Prädikation Jesu Christi als ‘Gott’ und die paulini-
sche Christologie: Erwägungen zur Exegese von Rom 9,5b,” ZNW 94/3–4 (2003): 171: “Der
Begriff θεός wäre dabei nicht im Sinne eines Nomen proprium verwendet, sondern als
Wesensbezeichnung.”
5. Conclusion
Rendering those Hebrew terms used when reference was made to the
Hebrew deity was as complex as the transmission of the Hebrew text
itself. The term ‫ אדני‬was frequently “misrepresented” if used in correla-
tion with the term ‫יהוה‬. Moreover, the Ketib-Qere perplexity had a major
influence on how the terms referring to the Hebrew deity was rendered;
this in turn spilled over and impacted the Greek translation of the Hebrew
text. The sensitivity, and hence, the prohibition in uttering the “name” of
the Hebrew deity became a major factor in the rendering of the terms
referring to the Hebrew deity in its Greek conceptual frame of reference.
The inference drawn from the investigated data of Isaiah is that the term
‫ אדון‬is consistently rendered using the term δεσποτής as a Greek equiva-
lent, while the term κύριος was primarily employed as the Greek equiva-
lent for the term ‫ אדני‬and ‫יהוה‬. However, the Greek translation of ‫יהוה‬
by κύριος is not without exception; in fact, the evidence appears to suggest
that the term θεός could be considered as the most suitable term not only
as a Greek equivalent for ‫יהוה‬, but as a term that represents the Hebrew
deity in general. The data suggests that the term θεός was considered, from
the 2nd century B.C.E. onwards, as a suitable Greek equivalent not only as
a rendering of the terms referring to the Hebrew deity proper, but also for
the personal Hebrew deity “named” ‫יהוה‬. Therefore, the “rule of thumb”
or any other systematised rendering of the Hebrew deity in OG Isaiah, at
least, should be discarded. Finally, due to the Greek variants available to
the New Testament authors, the theological and conceptual shift made by
the Greek translators had an impact not so much on the structure of the
New Testament text, but undoubtedly on the theo-logy and kyrio-logie of
the New Testament.
Revisiting the Original Greek of Ezekiel 18
Harry F. van Rooy
1. Introduction
When Ziegler1 prepared the critical text of Ezekiel for the Göttingen edi-
tion of the Septuagint (first published in 1952), he gave pride of place to
the B text, with Codex Vaticanus (B) and Papyrus 967 as the most impor-
tant witnesses. However, when he compiled his text, he had only a part of
this papyrus at his disposal. The publication of the parts of the papyrus in
Cologne2 and Madrid3 has made it possible to revisit the original Greek.
In some instances, Ziegler did not accept the reading of B as his origi-
nal Greek. The question is whether readings of Papyrus 967 may change
this choice in some instances. In many instances the papyrus supports
the choice of the reading of B as the original Greek. However, in some
instances where Ziegler did not accept the reading of B, this papyrus sup-
ports B, whereas in others it disagrees with B and supports the reading of
Ziegler. This paper will examine the readings of Ziegler in Ezekiel 18 in
the light of the readings of Papyrus 967. Many examples of Papyrus 967
agreeing with B and the critical text occur, but examples where Papyrus
967 and B agree with a reading not accepted by Ziegler (18:4, 9), or where
Papyrus 967 agrees with the critical text against B (18:10), are important.
These readings will be evaluated, with suggestions for revising the original
Greek of Ezekiel 18.
2. Papyrus 967 in Recent Discussions
As regards the value of Papyrus 967, two main views can be distinguished.
Some scholars regard the papyrus as very important for the history of the
1 J. Ziegler, Sepuaginta: Vetus Testamentum Graecum, XVI, 1: Ezechiel (Göttingen: Van-
denhoeck & Ruprecht, 1977).
2 L. G. Jahn, Der griechische Text des Buches Ezechiel nach dem Kölner Teil des Papyrus
967 (Papyrologische Texte und Abhandlungen 15; Bonn: Habelt, 1972.)
3 M. Fernández-Galliano. “Nuevas páginas del códice 967 del A.T. Grieco (Ez 28,19–
43,9),” SPap 10 (1971): 5–79.
194 harry f. van rooy
text of Ezekiel in the Hebrew transmission of the book, with the papyrus
representing an older version of the Hebrew. On the other hand, there are
scholars who restrict the importance of this papyrus to the transmission
of the Greek Ezekiel. It is impossible to discuss these questions in detail
in this paper. Important surveys of the problems and proposed conclu-
sions can be found in the works of Schwagmeier,4 Flanagan,5 Olley6 and
O’Hare.7
Schwagmeier, Flanagan and Olley support the idea that Papyrus 967 is
very important for the history of the Hebrew Ezekiel as well, while Flana-
gan argues in favour of its importance being restricted to the Greek trans-
mission of the book. In both these views, the fact that Ezekiel 36:23–30 is
omitted by the papyrus and the rearrangement of chapters 36–39 play an
important role. In the first view the insertion of 36:23–38 is regarded as
a late feature in the Hebrew tradition, while the rearrangement of 36–39
must be regarded as original. In the second view, these two factors are
ascribed to the transmission of Ezekiel in Greek. These issues will prob-
ably not find a consensus very easily. The aim of this paper is, however,
not related to that problem, but rather to the importance of Papyrus 967
for determining the original Greek. The manuscript frequently has read-
ings agreeing with the text as reconstructed by Ziegler, frequently agree-
ing with B. There are many instances, however, where the papyrus does
not agree with the text of Ziegler, agreeing or disagreeing with B. It will
be impossible to give attention to all the variants in Ezekiel 18, but a
representative sample will be discussed to throw light on the use of Papy-
rus 967 in determining the reading of the original Greek in Ezekiel 18.
As Papyrus 967 is an important pre-hexaplaric witness to the Greek
text of Ezekiel, attention will first be given to passages that are important
for the hexaplaric tradition of Ezekiel, then to passages important for the
relationship between B and Papyrus 967, and finally to some other inter-
esting variants in Papyrus 967.
4 P. Schwagmeier, Untersuchungen zu Textgeschichte und Entstehung des Ezechielbuches

in masoretischer und griechischen Überlieferung (D.Th. diss., University of Zurich, 2004),
180–187.
5 J. Flanagan, “Papyrus 967 and the Text of Ezekiel: Parablepsis or an Original Text?” in
Studies in Scripture in Early Judaism and Christianity (Library of Second Temple Studies 70;
ed. C. A. Evans; London: T & T Clark, 2009, 105–116).
6 J. W. Olley, Ezekiel. A Commentary Based on Iezekiel in Codex Vaticanus (Septuagint
Commentary Series; Leiden / Boston: Brill, 2009), 8–9.
7 D. M. O’Hare, “Have You Seen, Son of Man?” A Study in the Translation and Vorlage of
LXX Ezekiel 40–48 (SCS 57; Atlanta: Society of Biblical Literature, 2010), 7–19.
revisiting the original greek of ezekiel 18 195
3. Papyrus 967 and the Hexaplaric Tradition
3.1. Readings Marked with an Asterisk Not in Papyrus 967

Papyrus 967 was published over a period of time, with sections of the
papyrus in different localities. Ezekiel 18 is in that part of the papyrus in
Cologne, published by Jahn in 1972.8 In his discussion of the readings of
Papyrus 967, Jahn looks at all the instances where comparisons with hexa-
plaric readings can be made. In the first place he discusses the instances
where hexaplaric readings are indicated with an asterisk in hexaplaric
witnesses and where these readings do not occur in Papyrus 967, as is
to be expected. He discusses five such examples in Ezekiel 18. The first
of them is in Ezekiel 18:16. The MT has ‫ה־בגֶ ד‬ ָ ‫ וְ ֵעֹרום ִּכ ָּס‬at the end of
the verse. The last word is not translated in the original Greek accord-
ing to Ziegler. This omission is supported by Papyrus 967, A, B and some
other witnesses, while the hexaplaric reading adds ιματιον. Here Papyrus
967 and B have the reading of the original Greek, as is to be expected in
instances such as this. The same omission of the Hebrew occurs in 18:7,
supported by Papyrus 967, B and quite a number of witnesses. In this
instance an asterisk is not indicated in any hexaplaric witness, but the
reading of this word is clearly hexaplaric. Similar examples occur in 18:18,
22, 28 and 32.9 In all these instances Papyrus 967 agrees with the original
Greek as reconstructed by Ziegler.
3.2. Readings Marked with an Obelus Occurring in Papyrus 967

There are only two examples of this in Ezekiel 18, namely in vv. 2 and 8.
In 18:2, Ziegler has Υἱὲ ἀνθρώπου at the beginning of the verse. This reading
does not occur in the MT and is marked with an obelus in hexaplaric wit-
nesses. The fact that it occurs in Papyrus 967, B and many other witnesses
demonstrates that this reading is pre-hexaplaric and part of the original
Greek. It can probably be regarded as a pre-hexaplaric addition to bring
this address to the prophet in line with 12:22.10
In v. 8, Papyrus 967 omits the καί at the beginning of the verse, agree-
ing with the MT. It is interesting to note that this omission occurs in a
number of witnesses, including the Coptic in the Bohairic version.11 This
8 Jahn, Der griechische Text.

9 Ibid., 131–132.
10 Cf. W. Zimmerli, Ezekiel 1 (Hermeneia; Philadelphia: Fortress Press, 1979), 369.
11 The others are V-449 Armp Clem. Or.XIV 224 Lo. Tht. Tert.III 473.
raises the question whether the reading without καί must be regarded as
the original Greek, although there are many examples in Ezekiel 18 of the
Septuagint reading καί at places where the MT does not have “and”.
3.3. Readings Marked with an Asterisk Occurring in Papyrus 967

There is only one example of this in Ezekiel 18, namely in v. 24. This is,
however, a very interesting example. Ziegler has the following reading:
ἐν δὲ τῷ ἀποστρέψαι δίκαιον ἐκ τῆς δικαιοσύνης αὐτοῦ καὶ ποιήσῃ ἀδικίαν
κατὰ πάσας τὰς ἀνομίας, ἃς ἐποίησεν ὁ ἄνομος, πᾶσαι αἱ δικαιοσύναι αὐτοῦ, ἃς
ἐποίησεν, οὐ μὴ μνησθῶσιν· ἐν τῷ παραπτώματι αὐτοῦ, ᾧ παρέπεσε, καὶ ἐν ταῖς
ἁμαρτίαις αὐτοῦ, αἷς ἥμαρτεν, ἐν αὐταῖς ἀποθανεῖται.
When this is compared to the MT, two differences can be observed. Zie-
gler does not have anything for the Hebrew ‫ יַ ֲע ֶׂשה וָ ָחי‬after ὁ ἄνομος. Zim-
merli regards these two words as a gloss in the MT and wants to delete it.12
Allen argues, however, that it should be regarded as original, as linking up
with the same verb earlier in the verse.13 Jahn has noted that some wit-
nesses have added something to correct this deficiency.14 The hexaplaric
witnesses read και ποιηση και ζησεται, or something similar, equivalent to
the MT. The Lucianic witnesses have εαν ποιηση ου ζησεται, or something
similar. The majority of witnesses do not have this addition, making it
probable that the omission of the phrase of the MT was already part of
the original Greek. Papyrus 967 has an addition, but only εαν ποιηση. Jahn
thinks that these different readings support Ziegler’s reading of the origi-
nal Greek, and that the difference between the Lucianic reading and the
reading of Papyrus 967 could be attributed to Papyrus 967 having omitted
the part appearing in the Lucianic reading, or that the Lucianic reading
has added it.15 As these witnesses all attest to attempts to bring the Greek
in line with the Hebrew, an omission by Papyrus 967 is more likely. In
this instance B has the original Greek and Papyrus 967 has an attempt,
pre-hexaplaric, to bring the Greek in line with the Hebrew.
The second difference in v. 24, the plural noun ἁμαρτίαις in the Greek
for the Hebrew singular, is one of the instances where all the Greek wit-
nesses agree against the Hebrew, with only two exceptions (Sa and Con-
12 Zimmerli, Ezekiel 1, 373.

13 L. C. Allen, Ezekiel 1–19 (WBC 28; Dallas: Word), 266. Cf. also D. I. Block, The Book of
Ezekiel 1–24 (NICOT; Grand Rapids: Eerdmans), 581.
14 Jahn, Der griechische Text, 55 and 145.
15 Ibid., 145.
stit). However, in v. 21 the same Hebrew consonantal text appears, but

vocalised by the Masoretes as a plural (‫אתו‬
ָ ֹ ‫) ַחּט‬, while in v. 24 it is vocal-
ised as a singular. In both instances the Greek reads a plural, probably
indicating that the Greek used the same Vorlage, but interpreted the two
nouns as being both plurals, while the Masoretes took the one as plural
and the other as singular.
4. Readings with B and Papyrus 967 in Agreement
Many examples of Papyrus 967 agreeing with B and the critical text occur,
but important are those examples where Papyrus 967 and B agree with a
reading not accepted by Ziegler (18:4, 9) or where Papyrus 967 agrees with
the critical text against B (18:10). A typical example of Papyrus 967 agree-
ing with the critical text and B occurs in 18:2, as discussed above, where
the Greek has “son of man” as a plus at the beginning.
In 18:2 another very interesting example occurs. This example is noted
in his apparatus by Jahn,16 but not discussed in detail. The MT reads as
follows:
‫ל־א ְד ַמת יִ ְׂש ָר ֵאל ֵלאמֹר ָאבֹות‬
ַ ‫ת־ה ָּמ ָׁשל ַהּזֶ ה ַע‬
ַ ‫מ ְׁש ִלים ֶא‬ ֹ ֽ ‫ה־ּל ֶכם ַא ֶּתם‬ ָ ‫ַמ‬
‫אכלּו ב ֶֹסר וְ ִׁשּנֵ י ַה ָּבנִ ים ִת ְק ֶהינָ ה‬
ְ ֹ‫י‬
All the witnesses of the Septuagint do not have an equivalent for the
Hebrew phrase ‫מ ְׁש ִלים ֶאת־‬ ֹ ֽ ‫ ַא ֶּתם‬. Zimmerli regards the reading of the
Septuagint as a smoothing of the text in agreement with 12:22.17 Block
thinks it may be due to haplography.18 These proposals take the Hebrew
as the original text, with the Septuagint shortening it. It would not be so
easy to explain the reading of the MT as an expansion, disagreeing with
12:22.
The more important variant, referred to by Jahn, is related to the phrase
‫ל־א ְד ַמת יִ ְׂש ָר ֵאל‬
ַ ‫ ַע‬. Ziegler has the verse as follows: Υἱὲ ἀνθρώπου, τί ὑμῖν ἡ
παραβολὴ αὕτη ἐν τοῖς υἱοῖς Ισραηλ λέγοντες Οἱ πατέρες ἔφαγον ὄμφακα, καὶ
οἱ ὀδόντες τῶν τέκνων ἐγομφίασαν. For the relevant phrase he reads ἐν τοῖς
υἱοῖς Ισραηλ. However, quite a number of variants appear in the different
manuscripts of the Septuagint. The MT says that the proverb is about the
land of Israel, while the Septuagint says that the proverb is current among
16 Ibid., 51.
17 Zimmerli, Ezekiel 1, 369. Cf. also Allen, Ezekiel 1–19, 265.
18 Block, Ezekiel 1–24, 557.
the children of Israel. As the proverb deals with a saying in which the
children blame their fathers for the evil they are suffering, the phrase of
the Septuagint fits easier into the context. The proverb is not about the
land of Israel as such. Some of the variants in manuscripts of the Septu-
agint agree with the MT and can be regarded as editorial work to bring
the Greek closer to the Hebrew. Ziegler lists the following in his appara-
tus: ἐν τοῖς υἱοῖς] επι της γης Syhmg L´’ (449txt) Tht. Constit. (om. της) Aug.
Iul.op.imp. 3,38 = 𝔐; pr. επι της γης A´ ’-410; εν τω Cyr.IV 420 VI949: ex 3 |
Ισρ.] pr. του L´-V-449txt Tht.; > 46. The last two variants are not important
for this discussion. The other two are, however, important. If one looks at
the witnesses listed, B and related texts as well as hexaplaric texts support
Ziegler’s reading. A reading agreeing with the MT appears in the Lucianic
group, agreeing to some extent with the reference to Israel in the next
verse. The group of A has a combination of the two readings: επι της γης
εν τοις υιοις Ισραηλ (“about the land among the children of Israel”). It is
interesting to note that the MT has the same Hebrew expression in 12:22,
there rendered with επι της γης του Ισραηλ. This makes it quite possible
that the original Greek was translated from a different Vorlage in 18:2.19
Papyrus 967 has a unique reading: εν τῳ Ισραηλ επι της γης. It has the
reference to the land at the end of the phrase and refers to Israel, not
the children of Israel. This reference to Israel appears in v. 3 as well, and
it is also found in Cyrillus Alexandrinus. Papyrus 967 has a longer read-
ing than the MT and Ziegler in this instance, and this goes against the
general trend of the Septuagint having the shorter reading. The original
Greek probably had the reading as accepted by Ziegler and the reading of
Papyrus 967 can then be regarded as a pre-hexaplaric attempt to bring the
reading of the Greek closer to the Hebrew.
In Ezekiel 18:19 an interesting set of variants related to word order
occurs. Marquis made a very interesting study of word order variants in
Ezekiel.20 According to him, the Greek of Ezekiel 1–39 agree for 90% with
the word order of the MT, comparable to Jeremiah, and not much lower
than the figure for Psalm 1–78, 1 Samuel and 2 Kings.21 He regards Ezek-
19 Block, ibid., 557 n. 15, thinks that a different Vorlage may be possible, but also har-
monisation with other texts, such as 12:24. Zimmerli, Ezekiel 1–24, 369–370 regards the
reading of the Septuagint as secondary, pointing to other instances where the Septuagint
introduces the children of Israel.
20 G. Marquis, “Word Order as a Criterion for the Evaluation of Translation Technique
in the LXX and the Evaluation of Word-order Variants as Exemplified in LXX-Ezekiel,”
Textus 13 (1986): 50–84.
21 Ibid., 64–65.
iel as quite literal in the rendering of word order. He discusses a large

number of examples from Ezekiel, but unfortunately not the example to
be discussed here. Ziegler has the following reading in 18:19: οὐκ ἔλαβε
τὴν ἀδικίαν ὁ υἱὸς τοῦ πατρὸς. The MT reads: ‫לֹא־נָ ָׂשא ַה ֵּבן ַּב ֲעֹון‮ ָה ָאב‬. The
reading of Ziegler does not follow the Hebrew word order, by putting the
subject of the verb between the two words of the Hebrew genitive con-
struction, and does not have an equivalent for the Hebrew preposition.
Ziegler lists the different variants: τὴν ἀδ. ὁ υἱὸς τοῦ π. B Q-62] την αδ.
του π. ο υιος 147´ Las Arm Didasc.; αδ. υιος π. αυτου A; ο υιος την αδ. του
π. αυτου 88 (om. αυτου) L´’-403´ (om. την) Tht.Constit. = 𝔐. Papyrus 967
agrees exactly with 88, a manuscript from the Alexandrian group, while
many manuscripts from the Lucianic tradition has the same word order,
with the addition of the genitive pronoun to father. The reading with the
pronoun occurs in two Alexandrinian manuscripts as well (403 and 416).
The reading of Papyrus 967 may attest to the original Greek, or it could
be an early attempt to bring the Greek closer to the Hebrew. The fact that
the reading of Papyrus 967 does not occur in hexaplaric witnesses, and
is close to readings of Alexandrian and Lucianic witnesses, may support
the possibility of Papyrus 967 having the reading of the original Greek,
with the other witnesses reflecting either a different Vorlage or stylistic
adaptations.
In vv. 10–20 the words for father and son occur a number of times in
the MT. In some instances the words occur with the article and in others
without the article, and in some instances it occurs with the pronomi-
nal suffix for the third person singular. The following list presents all the
instances in the order they appear in the MT, with the verse in the first
column, the Hebrew in the second, an indication of the context in the
third and Ziegler’s rendering in the fourth.
Verse Hebrew Context Ziegler

10 ‫ֵּבן‬ ‫ן־ּפ ִריץ‬
ָ ‫ֹהוליד ֵּב‬
ִ ְ‫ו‬ καὶ ἐὰν γεννήσῃ υἱὸν λοιμόν
14:1 ‫ֵּבן‬ ‫ֹהוליד ֵּבן‬
ִ ‫וְ ִהּנֵ ה‬ ἐὰν δὲ γεννήσῃ υἱόν
14:2 ‫ָא ִביו‬ ‫ל־חּטֹאת ָא ִביו‬
ַ ‫ת־ּכ‬
ָ ‫וַ ּיַ ְרא ֶא‬ καὶ ἴδῃ πάσας τὰς ἁμαρτίας
τοῦ πατρὸς αὐτοῦ
17 ‫ָא ִביו‬ ‫הּוא ל֥ א יָמּות ַּב ֲעֹון ָא ִביו‬ οὐ τελευτήσει ἐν ἀδικίαις
πατρὸς αὐτοῦ
18 ‫ָא ִביו‬ ‫י־ע ַׁשק ע ֶֹׁשק‬
ָ ‫ָא ִביו ִּכ‬ ὁ δὲ πατὴρ αὐτοῦ ἐὰν
θλίψει θλίψῃ
Table (cont.)
Verse Hebrew Context Ziegler
19:1 ‫ַה ֵּבן‬ ‫ַמ ֻּד ַע לֹא־נָ ָׂשא ַה ֵּבן ַּב ֲעֹון ָה ָאב‬ Τί ὅτι οὐκ ἔλαβε τὴν
ἀδικίαν ὁ υἱὸς τοῦ πατρὸς
19:2 ‫ָה ָאב‬ ‫ַמ ֻּד ַע לֹא־נָ ָׂשא ַה ֵּבן ַּב ֲעֹון ָה ָאב‬ Τί ὅτι οὐκ ἔλαβε τὴν
ἀδικίαν ὁ υἱὸς τοῦ πατρὸς
19:3 ‫ַה ֵּבן‬ ‫ּוצ ָד ָקה ָע ָׂשה‬
ְ ‫וְ ַה ֵּבן ִמ ְׁש ָּפט‬ ὅτι ὁ υἱὸς δικαιοσύνην καὶ
ἔλεος ἐποίησε
20:1 ‫ֵּבן‬ ‫ֵּבן לֹא־יִ ָּׂשא׀ ַּב ֲעֹון ָה ָאב‬ ὁ δὲ υἱὸς οὐ λήμψεται τὴν
ἀδικίαν τοῦ πατρὸς
20:2 ‫ָה ָאב‬ ‫ֵּבן לֹא־יִ ָּׂשא׀ ַּב ֲעֹון ָה ָאב‬ ὁ δὲ υἱὸς οὐ λήμψεται τὴν
20:3 ‫ָאב‬ ‫וְ ָאב לֹא יִ ָּׂשא ַּב ֲעֹון ַה ֵּבן‬
‬ οὐδὲ ὁ πατὴρ λήμψεται τὴν
ἀδικίαν τοῦ υἱοῦ
20:4 ‫ַה ֵּבן‬ ‬
‫וְ ָאב לֹא יִ ָּׂשא ַּב ֲעֹון ַה ֵּבן‬ οὐδὲ ὁ πατὴρ λήμψεται τὴν
ἀδικίαν τοῦ υἱοῦ
The words for father and son appear in three forms each in the Hebrew:
without the article (10, 14 and 20), with the article (19 and 20) and with
the pronominal suffix for the third person masculine singular (14, 17 and
18). In vv. 14 and 18, the Greek uses the article before the noun as well, but
not in v. 17. The form without the article is rendered without the article in
Ziegler’s text of vv. 10 (a son) and 14 (a son), but with the article in v. 20
(the son, the father). The forms with the article in Hebrew are rendered
with the article in the Greek as well (vv. 19 and 20). For the forms with
the suffix the Greek uses the genitive of the personal pronoun in vv. 14, 17
and 18 (his father).
The two nouns appear twelve times in this section, with only two exam-
ples in v. 20 not following the Hebrew closely in Ziegler’s text. What is also
interesting to note is the rendering with the genitive pronoun but without
the article in vv. 17 and 18. In three instances no variants are listed by Zie-
gler (v. 10, 14:1 and 18) and in these instances Papyrus 967 has no variants
either. In another four instances the reading of Papyrus 967 agrees with
the text of Ziegler (17, 19:3, 20:1 and 20:2).
In v. 17, a number of witnesses insert the article (62’ L’’ 91 Tht). In 19:3
only one witness omits the article. The article is omitted by A and a few
other witnesses, while a number add the genitive pronoun to 20:2. The
fact that B and Papyrus 967 agree, supports Ziegler’s choice for the origi-
nal Greek in these instances. Wevers notes that the addition of the geni-
tive pronoun appears in both hexaplaric and Lucianic witnesses, although

they are not asterisked.22
This leaves five of the instances to be considered: 14:2, 19:1 and 2 and
20:3 and 4. In 14:2 Papyrus 967 is the only witness to omit the genitive pro-
noun, and this is probably an error in transmission, as there is no reason
to suspect a different Vorlage here.
The two examples in 19:1 and 19:2 have been touched upon in the dis-
cussion of the word order of this verse above. Here the rendering of the
words for father and son is the issue under discussion. The Hebrew has:
‫ ַמ ֻּד ַע לֹא־נָ ָׂשא ַה ֵּבן ַּב ֲעֹון ָה ָאב‬. Ziegler reads: Τί ὅτι οὐκ ἔλαβε τὴν ἀδικίαν
ὁ υἱὸς τοῦ πατρός. The Greek follows the Hebrew by having both words
with the article and without the genitive form of the pronoun. The four
possible variants are the omission of the two articles and the addition of
the genitive pronoun. In this instance it is argued above that Papyrus 967
probably has the word order of the original Greek. It can further be noted
that Papyrus 967 agrees with the Hebrew with regard to having the article
before both nouns and not adding the genitive pronoun. This strengthens
the possibility that Papyrus 967 has the original Greek for the whole sen-
tence under discussion.
As far as 20:2 and 3 are concerned, Ziegler’s text follows the Hebrew
closely, with articles before both nouns and without the genitive pronoun.
However, Papyrus 967 omits the article before the second father and adds
the genitive pronoun after the second son. The omission of the article
before father occurs in a number of witnesses (A´ O C´’-86´). These wit-
nesses represent the Alexandrinian text (A and 26), the main hexaplaric
witnesses, the catenae group as well as 86 and 710. The spread of witnesses
can be regarded as supporting the reading of Papyrus 967 as representing
the original Greek. There can be no other reason for these witnesses to
omit the article, whereas the addition of the article to bring the reading in
line with the Hebrew and with the previous sentence makes sense.
The addition of the genitive pronoun to son also has widespread support.
It does not occur in B and in some hexaplaric witnesses.23 The reading of
22 J. W. Wevers, “The L Text of Ezekiel,” in Studies in the Text Histories of Deuteronomy
and Ezekiel (MSU XXVI; J. W. Wevers and D. Fraenkel; Göttingen: Vandenhoeck & Rupre-
cht, 2003), 78.
23 Ibid.
Papyrus 967 can again be regarded as the original Greek, on account of

the same arguments as the previous example.
5. Other Interesting Readings in Papyrus 967
Jahn is of the opinion that Papyrus 967 is based in the original Greek, but
that it was corrected to bring it closer to the Hebrew in many instances.24
These are pre-hexaplaric corrections. In these instances the possibility of
Papyrus 967 representing the original Greek must be kept in mind.
A very interesting set of variants occur in Ezekiel 18, related to the
rendering of the Hebrew ‫ּוצ ָד ָקה‬ ְ ‫ ִמ ְׁש ָּפט‬that occurs in the MT in vv. 5,
19, 21 and 27. These two words occur as a pair in Ezekiel 33:14, 16 and
19 and 45:9 as well. Jahn discusses the examples in Ezekiel 18 and 45,25
but not the examples in Ezekiel 33 (probably because that chapter of the
book is not in the Cologne manuscript of Papyrus 967). These two words
were not treated in the same way in all the places where they occur. In
all the eight instances in the Hebrew of Ezekiel, the two words are the
object of the verb ‫ ָע ָׂשה‬. The verb in the perfect precedes the two words
in Ezekiel 18:5 and 21, 33:14 and 19, and follows them in 18:19 and 33:16. In
18:27 it precedes the verb in an imperfect with waw consecutive. In 45:9
the verb is an imperative and follows the two words. The Hebrew verb
is translated into Greek with forms of the verb ποιέω: in 18:5 a participle,
in 18, 19 and 27, 33:16 an aorist active indicative, in 18:21 and 33:14 and 19
an aorist active subjunctive and in 45:9 an aorist active imperative. In all
these instances the Septuagint follows the Hebrew word order. There are
only a few variants with regard to the form of the verb. The most impor-
tant one is in 18:19, where B and some hexaplaric (O-407) witnesses read a
perfect (πεποιηκε[ν]) for ἐποίησε. In this instance Papyrus 967 agrees with
the majority of the witnesses against B. This indicates that Papyrus 967
does not always agree with B.
However, there are many differences with regard to the rendering of
the two Hebrew words. According to Jahn, the normal rendering in the
Greek would be κρίμα καὶ δικαιοσύνη.26 This rendering occurs in most wit-
nesses in 18:5 and 33:14 and in all witnesses in 18:27, 33:16 and 19 and 45:9.
It is accepted by Ziegler as the reading of the original Greek in 18:5. How-
24 Jahn, Der griechische Text, 156.

25 Ibid., 162.
26 Ibid.
ever, the fact that κρίμα καί is omitted by B, Papyrus 967 and LaCS in 18:5
makes it more probable that this was the reading of the original Greek.
This is supported by Jahn as well.27 Zimmerli regards the reading of B as
a summarising of the Hebrew.28
In 18:19 Ziegler has δικαιοσύνην καὶ ἔλεος ἐποίησε. In this instance there
is a variety of variants. Papyrus 967 is the only witness with the reading
δικαιοσύνην ἐποίησε καὶ ἔλεον. Lucianic witnesses have κρίμα καὶ δικαιοσύνην
καὶ ἔλεον ἐποίησε. Hexaplaric witnesses have ἔλεος καὶ δικαιοσύνην πεποίηκεν
and B δικαιοσύνην καὶ ἔλεος πεποίηκεν. It is interesting to note that manu-
script 26, a witness of the A group, omits καὶ ἔλεος.
In 18:21 Ziegler has ποιήσῃ δικαιοσύνην καὶ ἔλεος, with the verb at the
beginning, but the same rendering for the two Hebrew words as in 18:19.
The Lucianic witnesses have the same three words for the two Hebrew
words as in 18:19. Papyrus 967 (and 62) has δικαιοσύνην καὶ ἔλεον. A and
related witnesses have τα δικαιωματα μου.
In 33:14 Papyrus 967 is the only witness to omit καὶ δικαιοσύνην, while all
the others have the reading regarded by Jahn as the standard rendering:
κρίμα καὶ δικαιοσύνην. This standard rendering, however, occurs in 33:16
and 19 in Papyrus 967, as in all the other witnesses. If one keeps the prob-
ability in mind that in 18:5 the original Greek probably only had one word,
and not two, it may be possible that Papyrus 967 has the original reading
in 33:14.
Jahn discusses many examples where Papyrus 967 has unique readings.
Most of them are not really important, but some warrant discussion. He
refers to two variants in 18:10.29 The text of Ziegler is as follows: καὶ ἐὰν
γεννήσῃ υἱὸν λοιμόν ἐκχέοντα αἷμα καὶ ποιοῦντα ἁμαρτήματα. The variants
are related to the words αἷμα and ἁμαρτήματα. The MT reads:
‫ן־ּפ ִריץ ׁש ֵֹפְך ָּדם וְ ָע ָׂשה ָאח ֵמ ַא ַחד ֵמ ֵא ֶּלה‬
ָ ‫ֹהוליד ֵּב‬
ִ ְ‫ו‬
The ‫ ָאח‬is probably the result of dittography. The rendering of the Sep-
tuagint can be regarded as an attempt to make sense of the difficult
Hebrew at the end. The words for “blood” and “sins” have variants in the
Greek. The Hebrew word for blood in the singular is usually rendered as
a singular in the Greek as well. The word for “sins” is in the plural in the
Septuagint, with no direct Hebrew equivalent. Ziegler has the singular
27 Ibid.
28 Zimmerli, Ezekiel 1, 370.
29 Jahn, Der griechische Text, 52.
αἷμα in the first instance and the plural ἁμαρτήματα in the second. In his
apparatus he lists 130 with the plural of the first word and B, 88, 106 and
LaCS with the singular in the second instance. Papyrus 967 has the plural
in both instances, in the first instance in agreement with 130 and in the
second instance in agreement with many witnesses, excluding B, 88, 106
and LaCS. In this instance Papyrus 967 supports the plural in the second
instance. However, the plural in the first instance in Papyrus 967 requires
more attention.
The combination of ‫ שפך דם‬occurs eight times in Ezekiel, in 16:38,
18:10 (as discussed above), 22:3, 6, 9, 12, 27 and 23:45.
In 16:38 Ziegler does not include the translation of this phrase in
his reconstructed text, but the phrase is added by many witnesses (και
εκχεουσων αιμα O-407 L-311-V-46 Tht.Hi. = 𝔐; pr. και εκχεουσης αιμα rel).
In all instances where it is added, the singular of the noun αιμα is used.
Unfortunately, this section is missing in Papyrus 967.
In 22:3 Ziegler has the plural in his text, with Papyrus 967 and a number
of witnesses having the singular (967 o L´’−36 Las* Co Aeth Arm Tht. = 𝔐).
In 22:6, 12 and 27 all the witnesses have the singular. In 22:9 all the wit-
nesses do not have the same word order, but they all have the noun in
the singular. In 23:45 the noun is also in the singular, but in the genitive,
in Ziegler’s text, without the verb for shedding (blood). In agreement with
the Hebrew, that verb is added in a number of witnesses, with only 233
having the noun in the plural. Although the singular is the normal render-
ing of the Hebrew noun, there are a number of instances where the plural
occurs in the majority of the manuscripts in 22:3 and some witnesses in
18:10 and 23:45. The plural would be the normal rendering of the Hebrew
noun in the plural, one would expect.
The Hebrew noun in the plural occurs in the MT in 9:9, 16:6 (twice), 9,
18:13, 22:2 and 24:6 and 9. In 9:9 the Hebrew says that the land will be full
of blood (plural). This phrase does not occur in the Septuagint. In 18:13 it
is stated that the blood (plural in MT) of the unrighteous person will be
on his head. The Hebrew plural is rendered by the singular in the Septu-
agint. In 16:6 (twice) and 9 the noun is used in the plural in the MT with
reference to blood at the birth of a baby. These are all rendered in the
singular in the Septuagint. The expression “city of blood” (plural) occurs
in the MT in 22:2 and 24:6 and 9. The phrase does not appear in 24:9. It
is rendered by a plural by the Septuagint in 22:2 and 24:6. Even in the
case of the Hebrew plural noun, the majority of instances are rendered
by the singular, except when it occurs in the expression “city of blood”.
Taking all these instances into account, it would be easier for a plural in
the Septuagint to be changed to a singular than vice versa. This makes the
possibility of Papyrus 967 having the original Greek in 18:10 quite strong,
and this should probably be the reading.
6. Conclusion
In the light of the discussion above, some proposals are made for different
readings for the original Greek as proposed by Ziegler.
Verse Ziegler Proposal

5 κρίμα καὶ δικαιοσύνην δικαιοσύνην
10 αἷμα αἷματα
19 οὐκ ἔλαβε τὴν ἀδικίαν ὁ υἱὸς τοῦ πατρὸς οὐκ ἔλαβε ὁ υἱὸς τὴν
20 οὐδὲ ὁ πατὴρ λήμψεται τὴν ἀδικίαν τοῦ υἱοῦ οὐδὲ πατὴρ λήμψεται τὴν
ἀδικίαν τοῦ υἱοῦ αὐτοῦ
There is also one instance in 18:8 where a different reading has some sup-
port for replacing Ziegler’s reading. There is a related example in 33:14, but
these two examples warrant further study in the light of a discussion of
related variants elsewhere in Ezekiel.
Verse Ziegler Proposal

8 καὶ τὸ ἀργύριον τὸ ἀργύριον
33:14 καὶ δικαιοσύνην Omit
What this discussion demonstrates is that one must be careful to equate

Papyrus 967 too easily with the original Greek. This is often done by
referring to the important variants in Ezekiel 7, the omission of 36:23c–38
and the different order of chapters 36–39 in Papyrus 967. However, the
detailed analysis of the variants in Papyrus 967 in Ezekiel 18 indicates
that there are many variants in this manuscript that cannot be equated
with the original Greek. There are instances where the papyrus has the
original Greek, often in agreement with B, but this cannot be taken as a
general rule.
Theology After the Crisis:
The Septuagint Version of Daniel 8–12
Martin Rösel
The Book of Daniel is one of the most fascinating writings in the Bible—
but also one of the most difficult. I would like to remind of some of the
highly disputed questions concerning this book. First of all there is the
notorious problem of the two language switches from Hebrew to Aramaic
and back in Dan 2:4 and after chapter 7. It is generally agreed that the
Aramaic parts on the whole belong to the older stratum of the literary
history of the book, stemming from late Persian or early Hellenistic times.1
The Hebrew chapters 8–12 are obviously younger. Here the situation has
changed. The issue is no longer life in the Babylonian or Persian diaspora,
but the events in Israel itself, when the temple cult in Jerusalem and the
religious life of the Israelites were threatened by a wave of Hellenisation,
fuelled by the Seleucid king Antiochus IV Epiphanes.2
The events in the early 2nd century B.C.E. under the reigns of Anti-
ochus III and IV, and the violent Maccabean response to the desecration
of the temple led to the formulation of an apocalyptic concept of history.
The different contributors to the Book of Daniel held the conviction that
the obvious decline of the worldly kingdoms attests to God’s plans to end
this period of history and to introduce a new, just and eternal government.3
Thus the figure of Daniel, a diaspora Jew full of “enlightenment, under-
standing, and wisdom like the wisdom of the gods” (Dan 5:11), was trans-
formed into an apocalyptic seer who saw strange visions and received
insights and understanding from heavenly messengers.
1 Some fundamentalist circles still hold the traditional view that the book comes from
an exilic author; cf. as a recent example W. Vogel, The Cultic Motif in the Book of Daniel
(New York: Peter Lang, 2010).
2 This short characterisation is based on inter alia: K. Koch et al., Das Buch Daniel (EdF
144; Darmstadt: Wissenschaftliche Buchgesellschaft, 1980); J. J. Collins, Daniel. A Commen
tary on the Book of Daniel (Minneapolis: Fortress Press, 1993), 24–64.
3 For a sound theological interpretation of the visions, cf. A. C. Merrill Willis, Disso
nance and the Drama of Divine Sovereignty in the Book of Daniel (OTS 520; London-New
York: Continuum, 2010). For a detailed concept of the literary development of the visions,
cf. R. G. Kratz, “The Visions of Daniel,” in The Book of Daniel. Composition and Reception
(Vol. 1; VTSup 83/1; eds. J. J. Collins and P. W. Flint; Leiden: Brill, 2001), 91–113 (although the
argument presented there is not convincing).
208 martin rösel
In chapter 11 the historical events from the last Persian kings until the
beginning of the final days is revealed to Daniel by an angel. This difficult
chapter is largely a vaticinium ex eventu, retelling elements of the history
of the Seleucid and Ptolemaic wars in the 3rd and 2nd centuries. But since
the verifiable data end in the year 165 B.C.E. and the reconsecration of the
temple in Jerusalem is not mentioned, scholars generally agree that the
Hebrew/Aramaic version of the book was finished in exactly this year.4
But the literary development of the book was not finished. It is here
that the Septuagint comes into play,5 because the Greek versions attest to
several independent additions to the book, namely the Susanna story to
introduce the figure of the wise and just Daniel; the prayers in chapter 3;6
and the stories of Bel and the dragon, which highlight the strictly mono-
theistic theology of the book. Moreover, the Greek version also offers a
glimpse into the earlier history of the Daniel traditions, because in chap-
ters 4–6 the text of the OG differs to such an extent from the parent MT
that the assumption of a largely deviating Semitic Vorlage is unavoidable.7
To add another field of discussion, the textual tradition of the Greek
translation is in itself difficult, because very early a second translation
known under the name of Theodotion was made, which was much closer
to the Hebrew text than the Old Greek.8 Since there are some citations
in the NT from this proto-Theodotionic version, it must come from pre-
Christian times.9 Eventually the Old Greek version was no longer used;
4 Collins, Daniel, 38, 388–390.

5 O. Munnich, “Texte massorétique et Septante dans le livre de Daniel,” in The Earli
est Text of the Hebrew Bible. The Relationship between the Masoretic Text and the Hebrew
Base of the Septuagint reconsidered (SBLSCS 52; ed. A. Schenker; Leiden-Boston: Brill, 2003),
93–120.
6 Cf. J. Joosten, “The Prayer of Azariah (DanLXX 3). Sources and Origin,” in Septuagint
and Reception. Essays prepared for the Association for the Study of the Septuagint in South
Africa (VTSup 127; ed. J. Cook; Leiden: Brill, 2009), 5–16, who opts for an original Greek
composition, but see also K. Koch, Daniel 1–4 (BKAT 22/1, Neukirchen-Vluyn: Neukirchener
Verlag, 2005), 314–375 with arguments in favour of an Aramaic original.
7 R. Albertz, Der Gott des Daniel. Untersuchungen zu Daniel 4–6 in der Septuaginta
fassung sowie zu Komposition und Theologie des aramäischen Danielbuches (SBS 131; Stutt-
gart: Katholisches Bibelwerk, 1988), who holds that the Vorlage of LXXDan 4–6 is older
than the Aramaic text of Dan 4–6.; see also Collins, Daniel, 5f.; Koch, Daniel, 377–380,
387–401 (on Dan 4 only, but with a theological exposition of the Greek text.)
8 A. A. di Lella, “The Textual History of Septuaginta-Daniel and Theodotion-Daniel,”
in The Book of Daniel. Composition and Reception (Vol. 2; VTSup 83/2; eds. J. J. Collins and
P. W. Flint; Leiden: Brill, 2001), 586–607.
9 A. Schmitt, “Stammt der sogenannte ‘θ’-Text bei Daniel wirklich von Theodotion?”
MSU 9 (1966): 279–392; idem, “Die griechischen Danieltexte (‘θ’ und o’) und das Theodo-
tionproblem,” BZ 36 (1992): 1–29.
theology after the crisis 209
we have the famous saying from the prologue of Jerome’s translation of

Daniel that he does not know why this has happened, but that he can
confirm that the text of the Seventy differs from the veritas (hebraica).10
The OG therefore survived in one pre-hexaplaric Ms (Pap. 967) and late
hexaplaric witnesses only (88, SyH).11 Because of the weak textual attesta-
tion, it is often difficult to decide about the oldest stratum of the Greek
translation.
Scholarly research on the Old Greek of Daniel has mainly concentrated
on chapters 1–7 and the problem of the deviating Vorlage in 4–6.12 Sev-
eral authors have focused on the translation of the “son of man” in 7:13,
because of its implications on the New Testament and the question of
the messianic titles of Jesus.13 Only a few studies are concerned with the
Hebrew chapters 8–12 and their translation into Greek.14 This is regret-
table, because the translation must have been carried out shortly after
the events of the Maccabean struggle. Most scholars opt for a dating in
the second half of the 2nd century B.C.E.15 Comparable to Ben Sira, the
10 Danihelem prophetam iuxta Septuaginta interpretes Domini Salvatoris ecclesiae non

legunt, utentes Theodotionis editione, et hoc cur accederit nescio . . . hoc unum adfirmare pos
sum, quod multum a veritate discordet et recto iudico repudiatus sit.
11 Cf. the new “Einleitung” by O. Munnich in the second edition of the Göttingen edi-
tion of Susanna, Daniel, and Bel et Draco (J. Ziegler and O. Munnich, Septuaginta: Vetus
Testamentum Graecum, XVI, 2: Susanna, Daniel, Bel et Draco [Göttingen: Vandenhoeck &
Ruprecht, 1999], 9–121).
12 B. Braasch, Die LXX-Übersetzung des Danielbuches—eine Orientierungshilfe für das
religiöse und politisch-gesellschaftliche Leben in der ptolemäischen Diaspora. Eine rezeptions
geschichtliche Untersuchung von Dan 1–7 (Diss., Hamburg 2003); Albertz, Gott; T. J. Mea-
dowcraft; Aramaic Daniel and Greek Daniel. A Literary Comparison (JSOTSup 198, Sheffield:
Sheffield Academic Press, 1995), with a short outlook on Dan 8–12, 245–261; T. McLay, The
OG and Th Versions of Daniel (SBLSCS 43; Atlanta: Scholars Press, 1996) with chapters on
Dan 8:1–10 and 12:1–13.
13 Cf. the recent discussion between O. Hofius and K. Koch: O. Hofius, “Der Septua-
ginta-Text von Daniel 7,13–14. Erwägungen zu seiner Gestalt und seiner Aussage,” ZAW
117 (2005): 73–90; K. Koch, “Der ‘Menschensohn’ in Daniel,” ZAW 119 (2007): 369–385. See
also M. Müller, The Expression ‘Son of Man’ and the Development of Christology. A History
of Interpretation (London: Equinox, 2008), esp. 326–374.
14 Worth mentioning is the dissertation of S. Pace Jeansonne, The Old Greek Translation
of Daniel 7–12 (CBQMS 19, Washington: Catholic Biblical Association, 1988); A. McCrystall,
Studies in the Old Greek Translation of Daniel (Diss., Oxford, 1980).
15 The main argument for this dating is the citation of βδέλυγμα ἐρημώσεως from Dan
11:31 in 1 Macc 1:54: Collins, Daniel, 8f. Cf. Meadowcraft, Aramaic Daniel and Greek Daniel,
275: “the stories were very soon translated into Greek”. See also the summary of Braasch,
Die LXX-Übersetzung, 291: “wenig wahrscheinlich, dass der LXX-Übersetzer lange nach 164
v.Chr. die Übersetzungstätigkeit aufnahm”. McCrystall (Studies, 386) dates between 161
and 152 B.C.E.
210 martin rösel
book was translated not more than about 50 years after its final composi-
tion in Hebrew.
This temporal proximity raises the question of whether some reflection
of the events around the Maccabean crisis can be traced in the Greek
version. From 165 B.C.E. onwards the course of history was quite different
from what Daniel had predicted. Therefore we can ask: are there indica-
tions that the translator has slightly adjusted the message of the book
so that it fits the new historical situation? Exploring this question is not
mere academic guesswork, because we know that the Book of Daniel did
not lose its theological and political importance. In Josephus and other
sources such as the Revelation of John, the fourth kingdom of Dan 7 was
identified with the Romans16 after the end of the world and the beginning
of the eternal kingdom did not come about during the time of one of the
Hellenistic rulers. This identification with the Romans seemed to be plau-
sible, since the depiction of the “fourth beast, terrifying and dreadful and
exceedingly strong . . . (with) great iron teeth” (Dan 7:7) seemed to fit the
overwhelming power of the Romans much better than a regional Syrian
kingdom. So the question of this paper is whether there are elements in
the Greek translation of the book of Daniel which kept the promises alive
rather than presenting a closed chapter of Judaean history.
The Old Greek of Daniel has been described as a dynamic translation,17
which shows several attempts to explain the difficult text of its Vorlage
and to state things more precisely than a mere literal translation could do.
16 See M. Tilly, “Die Rezeption des Danielbuches im hellenistischen Judentum,” in Die

Geschichte der Daniel-Auslegung in Judentum, Christentum und Islam. Studien zur Kommen
tierung des Danielbuches in Literatur und Kunst (BZAW 271; eds. K. Bracht and D. S. du Toit;
Berlin-New York: de Gruyter, 2007), 31–54; K. Koch, “Spätisraelitisch-jüdische und urchrist-
liche Danielrezeption vor und nach der Zerstörung des zweiten Tempels,” in Rezeption
und Auslegung im Alten Testament und in seinem Umfeld. Ein Symposion aus Anlass des 60.
Geburtstags von Odil Hannes Steck (OBO 153; ed. R. G. Kratz; Fribourg-Göttingen: Univer-
sitätsverlag-Vandenhoeck & Ruprecht, 1997), 93–123.
17 Cf. R. T. McLay “Daniel,” in A New English Translation of the Septuagint and the
B. G. Wright; New York-Oxford: Oxford University Press, 2007), 992; cf. H. D. Neef, “Daniel /
Das Buch Daniel,” in Septuaginta Deutsch. Erläuterungen und Kommentare zum griechi
schen Alten Testament. Band II: Psalmen bis Daniel (eds. M. Karrer and W. Kraus; Stuttgart:
Deutsche Bibelgesellschaft, 2011), 3016; Tilly, “Rezeption des Danielbuches,” 34, with a list
of characteristics. Pace Jeansonne (Old Greek Translation, 131f.) has come to the result that
the Old Greek is “reasonably accurate and faithful”, but calls for caution when it comes to
assessing a theological Tendenz in the work of the translator. Cf. also her earlier article:
idea., “The Stratigraphy of the Text of Daniel and the Question of Theological Tendenz in
the Old Greek,” BIOSCS 17 (1984): 15–35 (on Dan 7:13). In the course of this paper it will
become clear that Pace Jeansonne’s approach is too narrow.
This becomes particularly obvious when comparing the instances in ques-

tion with the later, more literal version of Theodotion. In the first chapters
of the book there are several hints that the translation must originate from
Egypt. The figures of Daniel and the king are reshaped so that the Greek
story of Daniel enhances the idea that competent Jews can be promoted
by the king to get important positions in the hierarchy of the royal court.18
This fits nicely the situation of the Jewish community in Alexandria.
But it is not completely clear whether the LXX version of Daniel can
be seen as a unit; there are several differences between chapters 1–7, on
the one hand, and 8–12, on the other. Therefore caution is advised so that
one does not draw conclusions too quickly from one part of the book to
the other.
In chapters 8–12 one can see several attempts to clarify difficult histori-
cal or geographical details. Thus in every instance in 11:5–40 the Hebrew
‫ ֶמ ֶלְך ַהּנֶ גֶ ב‬has been clarified by βασιλεὺς Αἰγύπτου (9 times) while the
“King of the North” ‫ ֶמ ֶלְך ַה ָּצפֹון‬was translated literally as βασιλεὺς βορρᾶ
(7 times in 11:6–40); Theodotion has the more literal ὁ βασιλεὺς τοῦ νότου
throughout.19 For later readers the unspecific “king of the north” in oppo-
sition to “Egypt” must have opened the possibility to relate this wicked
kingdom to the empire of the Romans. This interpretation could then be
supported by 11:30, because here the “ships of the Kittim” (‫) ִצּיִ ים ִּכ ִּתים‬
were correctly identified with the Ῥωμαῖοι.20 In 11:43 the ‫ֻל ִבים וְ ֻכ ִׁשים‬
were identified with the Λίβυες καὶ Αἰθίοπες; in 10:4 we can read Τίγρης
where MT has ‫( ִח ָּד ֶקל‬cf. Gen 2:24; Theodotion gives the transcription
Εδδεκελ) and in 10:20 the “prince of Greece” (‫ ) ַׂשר־יָ וָ ן‬has been translated
as στρατηγὸς Ἑλλήνων.
18 This is one of the important findings of the dissertation by Braasch, Die LXX-Über
setzung, 288–302.
19 Cf. McCrystall, Studies, 321f; F. F. Bruce, “The Earliest Old Testament Interpretation,”
OtSt 17 (1972): 37–52 (41f.). Unfortunately Pace Jeansonne (Old Greek Translation, 29) does
not comment on this striking actualisation, but only mentions it when presenting the
article by Bruce.
20 Neef, “Daniel,” 3046: “Die LXX ‘entziffert’ den MT ‫‘ )ציים( כתים‬kittäische (= zyprische
= mittelmeerische’ (Schiffe)”; J. A. Montgomery, A Critical and Exegetical Commentary on
the Book of Daniel (ICC; Edinburgh: T & T Clark, 1927 [repr. 1979]), 454: “a correct historical
midrash”. F. F. Bruce (“Prophetic Interpretation in the Septuagint,” BIOSCS 12 (1979): 17–26
[23]) hints at an intertextual connection with Num 24:24 (‫ )וְ ִצים ִמּיַ ד ִּכ ִּתים‬and assumes
that the translator has seen this prophecy fulfilled in the events of 168 B.C.E. G. Jahn (Das
Buch Daniel nach der Septuaginta hergestellt [Leipzig: Pfeiffer, 1904]) has seen the Vorlage
of the Septuagint as the older and better version than the MT, therefore he has offered a
large number of retranslations and conjectures, e.g. p. 111 on 11:30: The Vorlage had ‫ ָר ִמים‬,
which the translator read as ‫ר ִֹמים‬, MT later introduced the reference to Num 24:24.
212 martin rösel
Moreover, the actions of the Romans were characterised by καὶ

ἐξώσουσιν αὐτὸν καὶ ἐμβριμήσονται αὐτω “they will expel him and rebuke
him” (NETS), so that according to Robert Hanhart, “one believes to hear
the growling of the hard and curt man Gaius Pompilius”.21 Hanhart and
also Arie van der Kooij have convincingly shown that in 11:29f. and 11:1422
detailed knowledge about the wars between the Seleucid kings of the
North and the Ptolemaic kings of the south lies behind the translation.23
So one can conclude from these and other observations that the translator
was willing to interpret his text and to convey its meaning to his readers.
Although especially Sharon Pace Jeansonne and to a smaller extent also
Tim McLay24 have argued that those translations can all be explained as
technical solutions for the problems of the Vorlage, I have the impression
that it is possible to go—in the footsteps of Robert Hanhart and Arie van
der Kooij—one step further.25 I will concentrate on verses concerning the
events around the Maccabean crisis and the future.
In chapter 8 Daniel has a vision of a ram and a goat. Horns are growing
out of the goat; one of them “grew exceedingly great toward the south,
toward the east, and toward the glorious land” (8:9, MT ‫ל־ה ֶּצ ִבי‬
ַ ‫ ;וְ ֶא‬LXX
different: καὶ ἐπὶ βορρᾶν).26 The Greek version of vv. 10–12 is extremely
difficult to explain,27 but it is obvious that the cosmic dimension of the
threat against the temple is amplified: the horn was raised unto the stars
21 R. Hanhart, “Die Übersetzungstechnik der Septuaginta als Interpretation. Daniel 11,29
und die Ägyptenzüge des Antiochus Epiphanes,” in idem, Studien zur Septuaginta und zum
hellenistischen Judentum (FAT 24; ed. R. G. Kratz; Tübingen: Mohr Siebeck, 1999), 85, using
a characterisation by Mommsen; my translation into English.
22 A. van der Kooij, “A Case of Reinterpretation in the Old Greek of Daniel 11,” in Tradi
tion and Re-interpretation in Jewish and Early Christian Literature (FS J.-C. H. Lebram; StPB
36; ed. J. van Henten et al.; Leiden: Brill, 1986), 72–80.
23 This is also the conclusion reached by McCrystall, Studies, 384.
24 See McLay, OG and Th Versions, 191 n. 9: “we have given considerable reason to
doubt McCrystall’s view that the OG translator actually intentionally introduced signifi-
cant changes to MT for theological purposes”. Although I am also not convinced that the
Greek translation has its own chronological system, as McCrystall argues, he has collected
several instances which show that the translator has had his own theological ideas.
25 See also Meadowcraft, Aramaic Daniel and Greek Daniel, 250, who states a “mixture
of interpretation and misreading” as being characteristic of Dan 8–12 and gives some
examples for theological differences between MT and LXX.
26 Obviously the translator did not understand the concept of the “glorious / beautiful
land” (‫)צבי‬, because it is never translated adequately; the allusion to the north seems to
be taken from 8:4 (Meadowcraft, Aramaic Daniel and Greek Daniel, 251). Pace Jeansonne
(Old Greek Translation, 108) rejects the possibility of an intentional change and assumes a
“faulty Vorlage” or a misreading: ‫ ָצפֹון‬.
27 Cf. e.g. Montgomery, Daniel, 335–340; Pace Jeansonne, Old Greek Translation, 87;
both were assuming that glosses were moved into the text of the Old Greek.
of heaven (ἀστέρων for ‫)צבא‬, some of the stars will fall (or: he will be cast
down from the stars; cf. NETS), the mountains which existed from eternity
were overthrown. It is interesting to notice that, while most verbs are in
the Aorist, the statements about the deliverance of the captives (8:11 ἕως
ὁ ἀρχιστράτηγος ῥύσεται τὴν αἰχμαλωσίαν for ‫ר־ה ָּצ ָבא ִהגְ ִּדיל‬ ַ ‫ )וְ ַעד ַׂש‬and
the desolation of the sanctuary (καὶ τὸ ἅγιον ἐρημωθήσεται 8:11; καὶ τὰ ἅγια
ἐρημωθήσεται εἰς καταπάτημα 8:13) are in the future tense, although the
Hebrew text has no yiqtol forms. In v. 14 the idea is introduced that the
sanctuary has to be purified after 2 300 days (only here καθαρισθήσεται for
‫וְ נִ ְצ ַּדק‬, is used). Although this translation is in itself not very significant, it
has to be seen together with 11:35 and 12:6. Here the idea of a final purifica-
tion has been introduced over against the Hebrew text; in 11:35 the intel-
ligent (‫ילים‬ ִ ‫ ) ַה ַּמ ְׂש ִּכ‬are preparing for their purification; in 12:6 two angels
are talking about the consummation of the wonders and the purification.
One may suspect that the translator held the position that the temple
still has to be cleaned—although the Maccabeans have reconquered and
reconsecrated it.28
Another interesting characteristic of the Greek Daniel is introduced a
little later in 8:16. Here the anthropos angel says that “the vision is for this
ordinance” (NETS; ἐπὶ τὸ πρόσταγμα ἐκεῖνο ἡ ὅρασις, without counterpart
in MT). From this verse on πρόσταγμα is used several times for prophe-
cies and visions,29 e.g. in 9:2, when it is about the “ordinance of the Lord
to Jeremiah” or in 9:23 when an ordinance went out from the Lord to
answer Daniel’s prayer and Daniel is asked to understand this ordinance.
In 9:25 he is then told that he will “discover ordinances to respond, and
you will build Jerusalem as a city for the Lord” (καὶ εὑρήσεις προστάγματα
ἀποκριθῆναι καὶ οἰκοδομήσεις Ιερουσαλημ πόλιν κυρίῳ; the Hebrew text has
‫ד־מ ִ ׁ֣ש ַיח נָ ֔ ִגיד‬
ָ ‫רּוׁש ַלםִ ַע‬ ָ ְ‫) ִמן־מ ָֹצא ָד ָבר ְל ָה ִׁשיב וְ ִל ְבנֹות י‬30. The role of Daniel
has been significantly changed, for now he is responsible for the future
restoration of Jerusalem.31 According to 12:4 and 12:9, Daniel is ordered to
28 Similarly McCrystall, Studies, 386: “The Temple needs the true line of Priesthood to
return”. He assumes that the translator of Daniel is opposed to pro-Seleucid members of
the Jewish community.
29 In the Aramaic parts πρόσταγμα is only used for orders of the kings (for ‫ ִמ ָּלה‬cf. 2:15;
3:22 [28]; 5:7; 6:13).
30 It is not clear whether the translator had exactly this text in his Vorlage; the transla-
tion of ‫ ָמ ִ ׁ֣ש ַיח נָ גיד‬by κύριος is unusual (Montgomery, Daniel, 378); a misreading of ‫ עד‬to
‫ > עיר‬πόλις is not impossible, cf. Pace Jeansonne, Old Greek Translation, 126f.
31 Thus already A. Bludau, Die alexandrinische Uebersetzung des Buches Daniel und ihr
Verhaeltniss zum massorethischen Text (Freiburg: Herder, 1897), 120.
214 martin rösel
seal and hide these προστάγματα until the time of the end. Again, one can
ask whether the translator had a different expectation of the final events
to the authors of the Hebrew book of Daniel. It is evident that the transla-
tion πρόσταγμα for ‫ ָּד ָבר‬has changed the meaning considerably, because
Daniel not only sees visions and hears the word of God, but he is now
receiving orders and is involved in the events of the last days.
Chapter 9 is the meditation about the 70 years of Jeremiah 29:10 / 25:11.
In v. 2 the problem is stated: ἐγὼ Δανιηλ διενοήθην ἐν ταῖς βίβλοις τὸν ἀριθμὸν
τῶν ἡμερῶν ὅτε ἐγένετο πρόσταγμα κυρίου32 ἐπὶ Ιερεμιαν τὸν προφήτην
ἐγεῖραι εἰς ἀναπλήρωσιν ὀνειδισμοῦ Ιερουσαλημ ἑβδομήκοντα ἔτη: “I, Daniel,
pondered in the books the numbers of days when the ordinance of the
Lord came to the prophet Jeremiah to be awake (or: to rise) until the
fulfilment of the disgrace of Jerusalem—seventy years.” We have already
noticed the use of πρόσταγμα for “word”. It is interesting to see that the
Greek text has introduced the contents of this order to Jeremiah: to be
awake to see when the disgrace of Jerusalem is fulfilled. So this chapter is
not about the destruction of the temple, but about its defilement; the text
is adapted to reflect the situation in the 2nd century. In v. 4 Daniel begins
his prayer to the Lord God who is maintaining his covenant to those who
keep his ordinances. Here πρόσταγμα is used in a usual way for ‫ ִמ ְצוָ ה‬.
Obviously the prophecies to Jeremiah and the visions of Daniel are seen
on the same level as God’s commandments (cf. also v. 12: καὶ ἔστησεν ἡμῖν
τὰ προστάγματα αὐτοῦ).
The translation goes even one step further when it states in v. 10 that
the Lord has given his law to Moses (not in MT) and to us through his
servants the prophets (τῷ νόμῳ σου ᾧ ἔδωκας ἐνώπιον Μωσῆ καὶ ἡμῶν διὰ
τῶν παίδων σου τῶν προφητῶν). Now Moses and his law are of secondary
importance when compared to the prophets.33 Daniel, on the other hand,
who has insights into the mysteries of the prophets, will receive ordi-
nances and rebuild the city of the Lord. For me it is obvious that another
concept of prophecy lies behind these translations. This assumption fits
the observation that, according to 9:6, the prophets were sent not only to
32 This verse is of special interest because in MS 88 obviously a previous Hebrew tet-

ragram ‫יהוה‬, which served to replace original κύριος (now attested by Pap. 967), was no
longer understood by a later scribe, who corrected it to the similar looking τῇ γῇ; cf. the
apparatus of the Göttingen edition. For the problem of the replacement of the divine name
see M. Rösel, “The Reading and Translation of the Divine Name in the Masoretic Tradition
and the Greek Pentateuch,” JSOT 31 (2007): 411–428.
33 Unfortunately, Pace Jeansonne did not comment on this interesting addition, which
is clearly intentional; cf. Neef, “Daniel,” 3034: “ergänzt präzisierend”.
“all the people of the land” (‫ל־עם ָה ָא ֶרץ‬ַ ‫) ֶאל ָּכ‬, but to “every nation on
earth” (παντὶ ἔθνει ἐπὶ τῆς γῆς).34
Chapter 9 shows some more interesting deviations. A minor point is
that here in this chapter we have the only instance in the LXX where ‫ּתֹורה‬ ָ
is translated by διαθήκη: 9:13: κατὰ τὰ γεγραμμένα ἐν διαθήκῃ Μωσῆ “what
is written in the covenant of Moses”. This is clearly a contextual transla-
tion, because vv. 13f. are not dealing with prescriptions of the torah, but
with the evils that have come to Israel because they have not obeyed to
the law. Moreover, Israel has not only transgressed the law, but has com-
pletely forsaken it; ἐγκαταλείπω in 9:11 is a highly unusual translation for
‫( עבר‬only Judg 2:20).
Special attention is obviously devoted to the problem of understanding
and comprehension. In chapters 8–12 the verb διανοέομαι and its cognates
are used for several Hebrew words like ‫ ִּבינָ ה‬, ‫בין‬, ‫חשב‬, ‫ׂשכל‬. This unifor-
mity is not in line with the translator’s usual habit of stylistic variation.
Some examples: the justice of the Lord has to be understood (9:13); Daniel
has to understand God’s prostagmata (9:23+25); the angels give insight /
understanding (9:22). Moreover, Israel’s sin against God can be described
as lack of understanding, cf. 9:15: ἡμάρτομεν ἠγνοήκαμεν for ‫ ָח ָטאנּו ָר ָׁש ְענּו‬.
Therefore the 70 weeks which were disclosed to Israel (9:24) are meant to
bring the people to end the sin and to understand the vision (διανοηθῆναι
τὸ ὅραμα, without parallel in MT).35 It is obvious that emphasis is laid on
the intellectual side of the Israelite religion. This is in line with the Hebrew
version of Daniel, because here ‫ ַה ַּמ ְׂש ִּכ ִלים‬are obviously the group from
which the final form of the text comes.
Daniel’s request to understand the 70 years of Jeremiah is answered
by Gabriel in vv. 24–27. We have already seen that the expectation of the
Greek version differs considerably from the Hebrew text when Daniel is
presented as the one who will rebuild Jerusalem (v. 25). The second part
34 Cf. F. F. Bruce, “The Oldest Greek Version of Daniel,” OtSt 20 (1977): 22–40 (24). Pace
Jeansonne (Old Greek Translation, 114–116) has tried to show that this interpretation of
greater prophetic universalism is not valid, because there are more instances where ἔθνος
has been used to translate ‫ ָעם‬and referring to Israel. But the references she gives are
not convincing. In 11:14 she misses the “decidedly pro-Ptolemaic interpretation” (Collins,
Daniel, 380) of this verse; ἔθνος here refers to one of the peoples in the Hellenistic empire.
Moreover, the verse refers back to Am 9:11; cf. the explanation of McCrystall, Studies, 144f.,
333–335. In 11:33 the context is negative, which may lay behind the choice of ἔθνος for ‫ ָעם‬.
Pace Jeansonne’s own conclusion is hesitant (Old Greek Translation, 118): “the OG does not
necessarily give us any consistent indication of greater universalism”.
35 Cf. also 10:1: τὸ πλῆθος τὸ ἰσχυρὸν διανοηθήσεται τὸ πρόσταγμα; “the forceful multitude
will understand the decree” (NETS).
216 martin rösel
of v. 25 on the Vorlage of the LXX must have differed from what we have
today in the MT, because there is another sequence of the events revealed
to Daniel. Moreover, there are important differences in meaning. In v. 26
the angel is not speaking about 62 weeks of Jerusalem being rebuilt but
of 77 + 62 (weeks), after which an anointed (χρῖσμα) will be removed.36
After that a king of nations will demolish the city and the sanctuary with
its anointed. Then the end of this king will come with wrath (v. 26b) and
in wars. The covenant will prevail, will return and will be rebuilt (v. 27a).
This passage reads like a summary of the events of the Maccabean wars.
But the text goes on; the interval until the final righteousness is much
longer in the Greek text. At the end of times sacrifice and libation will
cease and the abomination of desolations will be given until the end of
time (v. 27b).
Admittedly, the textual situation of this verse is extremely difficult, as
one can see in the Göttingen edition, where Olivier Munnich has given
some parts of the text in brackets only. But two important elements are
obvious: the Greek version abandons the scheme of 70 weeks or year-
weeks. Either it has no longer understood the way the Hebrew text was
calculating,37 or it wanted to enhance the time span until the events of the
last days.38 Moreover, as the text reads now, the events of the Maccabean
crisis are only a kind of prelude. The end of days has been postponed,
so that it was possible to actualise the text in the 1st century C.E.; cf. the
explicit reference to the βδέλυγμα τῆς ἐρημώσεως in Mark 13:14.39
Again, this interpretation of the end of chapter 9 is supported by obser-
vations from chapters 11 and 12, the long final vision of the end of days. I
have already mentioned that in some verses the translator has shown his
knowledge of historical matters. When it comes to questions of the final
days, differences between Vorlage and translation become obvious. In
36 For the purpose of this paper it is not necessary to engage in the discussion about
the textual condition of 9:24–25 and possible reconstructions of a deviating Vorlage; cf. e.g.
Montgomery, Daniel, 401f.; McCrystall, Studies, 237–260, although his theory of a consistent
chronology in the Greek Daniel is not convincing. The same is true for the calculation of
Bruce (“Prophetic Interpretation,” 25) according to whom the translator switches to year
139 of the Seleucid chronology, beginning in 311 B.C.E.
37 See already Bludau, Die alexandrinische Uebersetzung, 116f. and Neef, “Daniel,” 3036f.
for possible misreadings of the Hebrew text to explain the Greek numbers.
38 Thus Bruce, “Prophetic Interpretation,” 25f., but he favours an explanation that the
translator wanted to bring the dating of the events in line with the historical facts about
the Maccabean crisis; cf. already Bludau, Die alexandrinische Uebersetzung, 117.
39 C. Evans, “Daniel in the New Testament,” in The Book of Daniel. Composition and
Reception (Vol. 2, VTSup 83/2; eds. J. J. Collins and P. W. Flint; Leiden: Brill, 2001), 490–527
(519–523).
11:33 the Hebrew text introduces the ‫ילי ָעם‬ ֵ ‫ ַמ ְׂש ִּכ‬, a group which, accord-
ing to 12:3, has to its credit “leading many to righteousness”. In the Hebrew
text the maskilim are to be distinguished from the “little help” (11:34), who
support the wise. It is widely accepted that ‫ ֵעזֶ ר ְמ ָעט‬refers to the Mac-
cabees, while the maskilim are the apocalyptic group behind the book of
Daniel.40
The Greek text displays a different idea of the sociology of the groups
of resistance. According to 11:33 LXX, there is a group of thoughtful ones
of the nation (NETS) or of people who care about the nation (LXX.D: “die
Nation im Sinn haben”), which will stumble by the sword and will be soiled
by pillaging. This group will then gather a little strength (συνάξουσιν ἰσχὺν
βραχεῖαν)41 and many will be gathered to that group. Then, according to
v. 35, some of this group of understanding people42 will be able to clean
themselves (cf. v. 33, they were soiled), and then they will be chosen and
purified. While the Hebrew text was speaking of two groups, the Greek
has only one. The Maccabees are expunged from history.
The Hebrew text then goes on to retell the last days of the king. The
final battle between the king of the north and the king of the south will
take place in the beautiful land—obviously Israel (11:41). In the LXX this
verse is entirely missing;43 according to v. 40, the king of the north will
come to Egypt to plunder the whole land (v. 43).
Only after the war against Egypt will the king move to a place “between
the seas and the mountain of the will of the holy one” (v. 45; NETS) where
he finds his end.44 Again, the LXX deviates in an important detail from
40 Cf. Montgomery, Daniel, 458f.; Collins, Daniel, 386, doubts that ‫ ֵעזֶ ר ְמ ָעט‬refers to a
group but is only stating that the maskilim received only little help from any group.
41 According to Collins (Daniel, 367), the translator had ‫ עז‬not ‫ עזר‬in his Vorlage; Jahn
(Das Buch Daniel, 113) assumes a reading ‫חיל‬.
42 It is interesting to notice that—perhaps for stylistic reasons—in 11:35 συνιέντων is
used for ‫ילים‬ ִ ‫ ; ַה ַּמ ְׂש ִּכ‬while in v. 33 ἐννοούμενοι was used. One could ask why the translator
did not use a cognate of διανοέομαι, which is one of the key words in Dan 8–12. Instead
the used διανοηθήσονται to render the following verb ‫( יכׁשלו‬perhaps reading ‫יׂשכלו‬, Neef,
“Daniel,” 3047). Anyway, it is obvious that he does not see ‫ ַמ ְׂש ִּכיל‬as a technical term, as
it is attested in texts from Qumran.
43 Rahlfs has: καὶ ἐπελεύσεται εἰς τὴν χώραν μου from MS 88 and the Syh; Collins (Daniel,
368) thinks of haplography in 967 and an attempt at reconstruction in 88 and Syh; already
Montgomery (Daniel, 468) was assuming a homoioteleuton. Also McCrystall (Studies, 381)
argues that a mistake must have occurred; A. Geissen (Der Septuaginta-Text des Buches
Daniel. Kap. 5–12, zus. mit Susanna, Bel et Draco sowie Esther Kap. 1, 1a-2,15 [Bonn: Rudolf
Habelt Verlag, 1968], 261) also thinks of a deviating Vorlage.
44 McCrystall (Studies, 383–385) suspects that v. 45 refers to the battles between Judas
Makkabaios and Antiochus IV at Emmaus and (because of the plural θαλασσῶν) near Bet
Zur, but this is not convincing.
218 martin rösel
its Vorlage because of the expectation of a final battle between northern

and Ptolemaic troops in Egypt. It is difficult to say whether or not the
translator had specific historical events in mind. After the campaign of
Antiochus IV against Egypt in 168 B.C.E., Demetrios II was in 129 B.C.E. the
next Seleucid king who tried to attack Egypt, but he did not get further
than the border at Pelusium.45 One can only assume that the battles of
Antiochus formed the matrix for the events of the final days which the
translator was expecting.
Chapter 12 then reveals what will happen in the last days. Again the
Greek version draws a considerably different picture. The first, minor
deviation is that the whole people of Israel will be exalted; LXX has πᾶς ὁ
λαός, while MT has ‫ ַע ְּמָך‬only (12:1). In v. 2 the Hebrew text speaks about
two groups who then will rise from the dust: some to everlasting life, and
some to shame and everlasting contempt: ‫עֹולם וְ ֵא ֶּלה ַל ֲח ָרפֹות‬ ָ ‫ֵא ֶּלה ְל ַחּיֵ י‬
‫עֹולם‬
ָ ‫ ְל ִד ְראֹון‬. In the Greek we have three groups: “some of everlasting life,
but others to shame and others to everlasting dispersion” (οἱ μὲν εἰς ζωὴν
αἰώνιον οἱ δὲ εἰς ὀνειδισμόν οἱ δὲ εἰς διασπορὰν αἰώνιον, translation: cf. NETS).46
One of the three groups will have a positive afterlife, v. 3, “those who are
intelligent will light up like the luminaries of heaven (καὶ οἱ συνιέντες
φανοῦσιν ὡς φωστῆρες τοῦ οὐρανοῦ), which is in accordance with the MT.
In the second part of this verse the credit due to this group is described
differently: they did not lead many to righteousness (‫יקי ָה ַר ִּבים‬ ֵ ‫ּומ ְצ ִּד‬
ַ ), but
they did strengthen God’s words (καὶ οἱ κατισχύοντες τοὺς λόγους μου); the
educational aspect of the work of the maskilim is no longer present.47 This
is in line with earlier observations concerning the ‫ ; ַמ ְׂש ִּכיל‬cf. footnote 42.
Then Daniel is ordered to hide these prostagmata and “to seal the book
until the time of consummation, until the many rage violently and the
earth is filled with injustice” (ἕως ἂν ἀπομανῶσιν οἱ πολλοὶ καὶ πλησθῇ ἡ γῆ
ἀδικίας, cf. NETS). The Hebrew text is quite different: many shall be running
back and forth, and knowledge48 shall increase (for ‫יְ ׁש ְֹטטּו ַר ִּבים וְ ִת ְר ֶּבה‬
‫) ַה ָּד ַעת‬. As observed at the end of chapter 9, the Greek version has a differ-
45 G. Hölbl, Geschichte des Ptolemäerreiches. Politik, Ideologie und religiöse Kultur von
Alexander dem Grossen bis zur römischen Eroberung (Darmstadt: Wissenschaftliche Buch-
gesellschaft, 2004), 178.
46 Bruce (“Oldest Greek Version,” 26) asks whether the third group is a result of a vari-
ant translation of one clause only.
47 Even if the Vorlage of the translator was different (cf. Neef, “Daniel,” 3049f.: perhaps
the translator has read ‫ )?ומחזיקי דברי‬it is interesting to notice that ‫ דבר‬was not trans-
lated with the usual prostagma.
48 NRSV and Collins, Daniel, 369: evil, reading ‫ הרעה‬for ‫הדעת‬, based on LXX: ἀδικίας.
ent expectation concerning the end of days: the worst is yet to come. And
as could be seen earlier, there is another important link between chapters
9 and 12 in the Greek version, because also 12:6 speaks—without reference
in the Hebrew text—about the need for a final purification: (πότε οὖν . . . ὁ
καθαρισμὸς τούτων).
Finally, in v. 9, another small but important deviation points in the
same direction. In the Hebrew text the angel orders Daniel: “Go your way,
Daniel, for the words are to remain secret and sealed until the time of the
end” (‫ד־עת ֵקץ‬ֵ ‫) ַע‬. In the Greek version only ‫ ַעד‬is translated and con-
nected with the following verse: ἕως ἂν πειρασθῶσι καὶ ἁγιασθῶσι πολλοί καὶ
ἁμάρτωσιν οἱ ἁμαρτωλοι . . . Again, the LXX gives the impression that in the
last days there will be a final threat. The same idea seems to lie behind
the expansion in v. 13 that there are still days and hours until the end. But
at least for Daniel the end will not be too bad, because here, in the last
sentence of the book, he receives the promise that he will rest and then
rise upon his glory (καὶ ἀναστήσῃ ἐπὶ τὴν δόξαν σου; the MT has ‫) ְלג ָֹר ְלָך‬.49
A short conclusion might be in order. It has become obvious that the
first Greek version differs from the Hebrew text, especially in respect to
the question of the events of the final days. In the eyes of the translator
the events of the Maccabean crises have only been a kind of prefigura-
tion of what will come in the future.50 God has revealed these coming
threats by way of his ordinances to the prophets and to the apocalyptic
seer Daniel. Those of the Israelites who think about these revelations can
gain eternal life and will light up like stars.
Admittedly, the textual situation of the Greek Daniel is so problematic
that not all of the collected observations are equally convincing and it is
not always clear whether a variant comes from the Vorlage or from the
translator. But seen in the larger framework of the history of reception,
it is understandable why the expectations of the book of Daniel could
be applied to the Roman empire, although the predictions of this book
about the Hellenistic rulers failed. The Greek book of Daniel can easily
be inscribed in the history of the apocalyptic movement, but this is a task
for another paper.
49 Only here δόξα is used for ‫ּגֹורל‬

ָ ; this translation is not mentioned in T. Muraoka,
Index, s.v.
50 Others like Bludau (Die alexandrinische Uebersetzung, 126f.) or McCrystall (Studies,
385f.) concluded that the Old Greek of Daniel wanted to comment on the Maccabean
events; they did not see that the translation opens the book to a new interpretation of
coming events.
Part Three
Textual Criticism
The Value of the Septuagint for Textual Criticism of the
Hebrew Bible as Illustrated by the Oxford Hebrew Bible
Edition of 1 Kings
Jan Joosten
1. Introduction
From the late nineteen eighties onward, Septuagint studies have been
going from strength to strength. Major projects aimed at producing trans-
lations into modern languages, with more or less extensive notes—La
Bible d’Alexandrie in France, NETS in North America, Septuaginta Deutsch
in Germany—have attracted lots of young scholars. There are schools of
Septuagint studies in Finland and Spain. The international meetings of
the International Organization for Septuagint and Cognate Studies (IOSCS)
have grown from intimate gatherings of a handful of experts to large-scale
conferences with parallel sessions—70 papers were programmed at the
latest IOSCS meeting in 2010 in Helsinki. And the boom is not coming to
an end: the Hexapla Project, the projected seven-volume Handbuch zur
Septuaginta, and the recently launched Historical and Theological Lexicon
of the Septuagint witness to the continuing vitality of the field. In South
Africa, too, the study of the Greek version of the Bible has grown to the
point where the creation of a special organization, the Association for the
Study of the Septuagint in South Africa became a necessity.1
The growth of Septuagint studies has gone hand in hand with the rise
of a relatively new approach to the old Greek version. In earlier times,
starting with Louis Cappel in the seventeenth century, the Septuagint was
mostly consulted as a tool, not as a scientific object in its own right.2 The
single most important context in which the Septuagint was exploited was
textual criticism of the Hebrew Bible. Old Testament scholars used the
Greek version as a quarry for readings apt to improve the Hebrew text
where it was felt to be faulty. Emanuel Tov’s celebrated introduction to
the Septuagint, first published in 1981, is called The Text-Critical Use of
1 A first volume of essays prepared for this new Association has recently been pub-
lished: Septuagint and Reception (VTSup 127; ed. J. Cook; Leiden: Brill, 2009).
2 L. Cappel, Critica sacra (Paris, 1650).
224 jan joosten
the Septuagint in Biblical Research.3 Over the last thirty years, however,
the angle of approach has shifted. More and more scholars declare their
interest in the Septuagint as a writing, or a collection of writings, in its
own right.4 Research has focused on questions such as how the Septua-
gint translators interpreted their Vorlage, and how the Greek text itself
became a source of meaning and interpretation. The historical origins of
the version, its original and subsequent status among Jews and non-Jews,
the inner coherence and continuity of the books making up the Greek
“canon” (if that is the right word) have been at the center of scholarly
interest. Questions of textual history are treated from the point of view
of Greek text forms such as the kaige recension or the Lucianic/ proto-
Lucianic amalgam.5
It would be an exaggeration to say that the new approaches have super-
seded the earlier focus on textual criticism of the Hebrew Bible. There
have been some vivacious exchanges, of course: I remember a vigorous
discussion between Martin Rösel and Ron Hendel on a number of Septua-
gint readings in Genesis—explained by Rösel as reflective of the transla-
tors’ exegesis and claimed by Hendel as witnesses to a different Hebrew
source text.6 But mostly what can be said is that the new spirit in Sep-
tuagint research is complementary to the earlier text-critical enterprise:
the Septuagint should be understood on its own terms before it can be
compared to Hebrew text forms; “retroverting” the Greek into Hebrew can
only be done when the version has thoroughly been analyzed as a transla-
tion. And in fact, the traditional text-critical exploitation of the Septuagint
has never gone away but has continued to be practiced by knowledge-
3 E. Tov, The Text-Critical Use of the Septuagint in Biblical Research (Jerusalem Biblical
Studies 8; Jerusalem: Simor, 1997).
4 See, e.g., M. Harl, “La ‘Bible d’Alexandrie’ et les études sur la Septante. Réflexions sur
une première expérience,” VC 47 (1993): 313–340; W. Kraus, “Hebräische Wahrheit und
Griechische Übersetzung. Überlegungen zum übersetzungsprojekt Septuaginta-deutsch
(LXX.D),” TLZ 129 (2004): 989–1007.
5 A representative cross-section of Septuagint studies is provided by the proceedings
of IOSCS conferences edited in the SBLSCS series (editor M. K. H. Peters). The proceed-
ings of Septuaginta-deutsch conferences held in Wuppertal also document the state of the
art well: Die Septuaginta—Texte, Kontexte, Lebenswelten (WUNT 219; eds. M. Karrer and
W. Kraus; Tübingen: Mohr Siebeck, 2008); Die Septuaginta—Texte, Theologien, Einflüsse
(WUNT 252; eds. W. Kraus, M. Karrer and M. Meiser; Tübingen: Mohr Siebeck, 2010); a
third volume is forthcoming.
6 See M. Rösel, “The Text-Critical Value of Septuagint-Genesis,” BIOSCS 31 (1998): 62–70;
R. S. Hendel, “On the Text-Critical Value of Septuagint Genesis: A Reply to Rösel,” BIOSCS
32 (1999): 31–34.
the value of the septuagint for textual criticism 225
able scholars in different places.7 An important niche for the text-critical

use of the Septuagint in our own time is the Oxford Hebrew Bible proj-
ect launched in the early 2000s by the aforementioned Ron Hendel. The
Oxford Hebrew Bible will be a new edition of the Hebrew Bible featuring a
critical text, an apparatus and a textual commentary.8
2. The OHB of 1 Kings
Producing a critical text of the Hebrew Bible is a daunting enterprise, for

many different reasons.9 On the theoretical level, an important question is
which textual form one should aim to reconstruct.10 Biblical literature was
shaped by tradition, and books may have known different Hebrew edi-
tions through time. For some books, such as Jeremiah and Joshua, there is
textual evidence of such divergent editions. Should textual critics aim for
the earliest form attainable, or for some later form, and if so, which one?
The answer to these questions varies a bit from book to book.11 In practice,
what will be done for most books is nothing more adventuresome than to
emend the more egregious errors of the received MT.12
Another thorny issue is what one should do with the Masoretic vocaliza-
tion and accents: one might argue they have no place in a reconstruction of
7 See, e.g., B. Albrektson, Text, Translation, Theology. Selected Essays on the Hebrew
Bible (SOTSMS; Farnham: Ashgate, 2010); O. Munnich, “Retouches rédactionnelles au
texte proto-massorétique: l’apport des versions grecques de Daniel,” in Congress Volume
Ljubljana 2007 (VTSup 133; ed. A. Lemaire; Leiden: Brill, 2010), 71–86; M. Richelle, Le Tes-
tament d’Élisée. Texte Massorétique et Septante en 2 Rois 13.10–14.16 (Cahiers de la Revue
Biblique 76; Paris: Gabalda, 2010).
8 See S. White Crawford, J. Joosten, E. Ulrich, “Sample Editions of the Oxford Hebrew
Bible: Deuteronomy 32:1–9, 1 Kings 11:1–8, and Jeremiah 27:1–10 (34 G),” VT 58 (2008): 352–
366.
9 R. Hendel, “The Oxford Hebrew Bible: Prologue to a New Critical Edition,” VT 58
(2008): 324–351; H. G. M. Williamson, “Do We Need A New Bible? Reflections on the Pro-
posed Oxford Hebrew Bible,” Bib 90 (2009): 153–175; E. Tigchelaar. “Editing the Hebrew
Bible: An Overview of Some Problems,” in Editing the Bible (ed. J. Kloppenborg and J. New-
man; Toronto: University of Toronto Press, forthcoming).
10 See B. K. Waltke, “Aims of OT Textual Criticism,” WTJ 51 (1989): 93–108.
11 As can be seen in the sample published in the article referred to in n. 8, Jeremiah
will be edited in two columns, one representing the MT, the other the Vorlage of the Sep-
tuagint (see also the remarks on 1 Kgs 2:1 below).
12 Emanuel Tov has criticized the use of the term “emendation” in reference to the
adoption of preferred readings from other witnesses than the MT: the goal of textual criti-
cism is to restore the original text, not to emend the MT, see E. Tov, Textual Criticism of
the Hebrew Bible (Minneapolis / Assen: Fortress Press / Van Gorcum, 1992), 351; similarly
Hendel, “Oxford Hebrew Bible,” 331. Theoretically Tov is right, but it is also necessary to
recognize the limits of what is possible.
226 jan joosten
the “oldest text attainable”, because at the time when this text would have
circulated, no vowel pointing existed;13 but one might argue, too, that the
pointing encodes information going back to earlier times: when the text
was written in consonantal form only, people would still know how to
pronounce it.14 In the meantime, OHB policy is to keep the vocalization
and accents wherever the MT is adopted, and to vocalize—but not to
accentuate—emended passages.15
On the practical level, the main problem is the dearth—or rather, the
patchy attestation—of Hebrew textual material. A few passages in the
Pentateuch are attested in the MT, the Samaritan Pentateuch, and one
or more texts from Qumran. In Isaiah, or in parts of the books of Samuel,
the Qumran material is sufficiently extensive to start working. In many
other books, however, the only Hebrew witness available is the MT. The
MT is not entirely without variants, of course, but it is surprisingly unified
nonetheless. There would be scant justification for a critical edition of the
Hebrew Bible based solely on medieval Hebrew manuscripts. In passages
where no non-Masoretic Hebrew material is available, the lion’s share of
the text-critical enterprise will rest on a comparison of the MT with the
Greek version, and on conjecture. Without putting too fine a point on it,
this is the situation in regard to the first book of Kings.
Initially, the OHB of 1 Kings was attributed to Steve McKenzie. McKen-
zie did produce a critical edition of 1 Kgs 1:1–7—including an apparatus
but no textual commentary. But soon afterwards he abandoned the proj-
ect. In 2004, the book was reallocated to the present author. The proj-
ect of editing 1 Kings started in earnest in 2009, when funding became
available allowing close collaboration between Jan Joosten, of Strasbourg,
and Jean Koulagna, professor of OT at the Lutheran School of Theology
in Meiganga, Cameroon.16 Since 2009, two chapters of 1 Kings have been
analyzed and provisional versions of the critical text, the apparatus, and
the textual commentary have been created.
Although the critical text is based on a confrontation of all available
direct witnesses, and although the apparatus includes extensive references
13 See Williamson, “Do We Need A New Bible?”

14 For the antiquity of the oral tradition issuing in the Masoretic vocalization see, e.g.,
F. Buhl, Canon and Text of the Old Testament (trans. J. MacPherson; Edinburgh: T & T Clark,
1892), 236–239; J. Barr, Comparative Philology and the Text of the Old Testament. With Addi-
tions and Corrections (Winona Lake: Eisenbrauns, 1987), 194–207.
15 This will be done on the “copy-text principle,” see Hendel, “Oxford Hebrew Bible,”
343–345.
16 The funding comes mainly from the Institut Universitaire de France.
to readings in the Peshitta, the Targums, the Vulgate, and Flavius Josephus’
Antiquities, really meaningful divergences from the received Hebrew text
are almost wholly confined to the Greek versions. In practically all places
where a non-Masoretic reading has been adopted, our authority is the
Septuagint. While the aforementioned Aramaic and Latin versions hew
closely to the Masoretic tradition, diverging from it only rarely and usu-
ally in small details, the Greek tradition clearly reflects a different state of
the Hebrew text of 1 Kings.17 Unfortunately, the Greek textual tradition is
itself notoriously complicated in the books of Kingdoms/Samuel-Kings.
3. The Greek Text of 3 Kingdoms
3 Kingdoms is one of the books of the Septuagint for which the Göttingen
edition is still in preparation.18 Perusal of the old Brooke-McLean edition,
known also as the “Cambridge edition”, will quickly show the polymor-
phous nature of manuscript evidence.19 Successive stages in the history of
investigation have permitted, nevertheless, to perceive some measure of
order in the chaos. The most important discoveries were made by Thack-
eray, at the beginning of the twentieth century, and by Barthélemy in the
nineteen sixties.
On the basis of his grammatical analyses, Henry Thackeray realized that
the bulk of Greek manuscripts of 1–4 Kingdoms combine two distinct ver-
sions: a literal version in 1 Kgds 1–2 Kgdms 11:2 and in 3 Kgdms 2:12–21:43,
and an ultra-literal one, closely resembling Theodotion, in 2 Kgdms 11–3
Kgdms 2:11 and in 3 Kgdms 22–4 Kgdms 25.20 He correctly judged the less
literal parts to be older than the more literal ones:
Earlier portions: 1 Sam 1–2 Sam 11:1 1 Kgdms 1–2 Kgdms 11:1
1 Kgs 2:12–21:29 3 Kgdms 2:12–21:43
Later portions: 2 Sam 11:2–1 Kgs 2:11 2 Kgdms 11:2–3 Kgds 2:11
1 Kgs 22–2 Kgs 25 3 Kgdms 22–4 Kgdms 25
17 The textual nature of Josephus’ paraphrase in the Antiquities is a more complicated

matter. Josephus appears to know both the MT and the Old Greek, mixing elements as he
sees fit and adding others on his own account.
18 The Göttingen edition is being prepared in Madrid by Julio C. Trebolle Barrera and
his team, notably Pablo Torijano Morales and Andrés Piquer Otero.
19 The Old Testament in Greek. Vol. II, Part II, I and II Kings (eds. A. E. Brooke, N. McLean
and H. St. J. Thackeray; Cambridge: Cambridge University Press, 1930).
20 See H. St. J. Thackeray, A Grammar of the Old Testament in Greek (Cambridge: Cam-
bridge University Press, 1909), 10–11.
228 jan joosten
Thackeray explained the presence of earlier and later portions in Kingdoms

as reflecting a translation in two stages: at first, only parts of Samuel–Kings
were translated, leaving aside the “more disastrous” bits (the gradual
downfall of David in 2 Sam and 1 Kgs 1–2, and the end of the monarchy in
2 Kgs).21 When, in a second stage, it was decided to add the missing parts,
this was done according to an ultra-literal technique that had become
fashionable in the meantime. Later research would disprove this expla-
nation. Moreover the exact demarcation of “earlier” and “later” portions
would be tweaked a bit.22 The identification and relative dating of the
different portions, however, has become an inalterable component of Sep-
tuagint research.
Half a century later, Dominique Barthélemy was able to show that
Thackeray’s “later portions” did not represent a translation, but rather the
recension of an older version.23 The basic hypothesis was suggested to
Barthélemy by the Greek Minor Prophets scroll discovered in Nahal Hever
a few years before. This scroll turned out to contain a recension of the
Septuagint text of the Dodecapropheton following the same principles of
literal translation as could be observed in parts of Kingdoms.
An important element in Barthélemy’s demonstration was the obser-
vation that an “Old Greek” version of Thackeray’s later portions was
preserved in the so-called Antiochene or Lucianic text, represented in
Kingdoms by five minuscules.24 While the majority text of 2 Kgdms 11:2–3
Kgdms 2:11 shows much similarity to Theodotion—and to the Minor
Prophets scroll of Nahal Hever—, the translation technique of the Anti-
ochene text coheres with that of Thackeray’s “earlier portions”. Thus it
turns out that the books of Samuel-Kings were from the start translated
completely. Later, however, the text was revised according to the prin-
ciples of a Palestinian school aiming at greater conformity between the
Greek version and the emerging MT. On the basis of one striking feature,
Barthélemy proposed to call this revision the kaige recension. Like Thack-
eray’s distinction of different portions, Barthélemy’s explanation of the
“later portions” as recensional has become a foundation stone of Septuagint
21 Ibid., 11.
22 See J. D. Shenkel, Chronology and Recensional Development in the Greek Text of Kings
(HSM 1; Cambridge: Harvard, 1968), 117–120.
23 D. Barthélemy, Les Devanciers d’Aquila (VTSup 10; Leiden: Brill, 1963), 47.
24 Referred to by the sigla b o c2 e2 in the Cambridge edition and by the numbers 19, 82,
93, 108 and 127 in the Göttingen edition (b = 19 + 108, the two manuscripts tending to go
together). For a description of the manuscripts, see N. Fernández Marcos and J. R. Busto
Saiz, El Texto Antioqueno de la Biblia Griega. I 1–2 Samuel (Madrid: CSIC, 1989), XV–XXIII.
research. Thackeray’s “later portions” are nowadays usually referred to as

“kaige sections”.
In Barthélemy’s view, the mixture of Old Greek and kaige in the major-
ity text of Kingdoms reflected the partial nature of the revision in a first
stage. The manuscripts used by Origen for the Septuagint column in his
Hexapla were mixed because the kaige recension had been applied only
to 2 Kgdms 11:2–3 Kgdms 2:11 and 3 Kgdms 22–4 Kgdms 25.25 This remains
a possibility. It is difficult to explain, however, why a reviser would tar-
get only those two blocks, leaving alone the first 41 chapters as well as
some 20 chapters between 3 Kgdms 2:12 and 21:43.26 Later scholars have
tended, therefore, to ascribe the peculiar mixture of Old Greek and kaige
in the main manuscript tradition of 1–4 Kgdms to chance. In this view, the
recension was applied to the whole work. Later however, a scribe inadver-
tently copied parts of the work from a scroll containing the Old Greek and
parts from another scroll containing the kaige recension.27 By accident,
this mixed manuscript became the root of the later textual tradition—
excepting the Antiochene tradition, which preserved the original transla-
tion throughout.
In the post-Barthélemy period, two further adjustments were made.
Firstly, it was pointed out by several scholars that the Antiochene text
cannot simply be qualified as a representative of the Old Greek, not even
in the kaige sections (Thackeray’s “later portions”, where the main tradi-
tion reflects the kaige recension). Although the Antiochene text remained
largely untouched by the kaige recension, it underwent many other
changes, as can be proven easily from the non-kaige section (Thackeray’s
“earlier portions,” where the main tradition reflects the Old Greek). It is
more correct, therefore, to say that in the kaige sections the Old Greek can
be partially retrieved from the Lucianic tradition.
Secondly, detailed study has shown that all available manuscripts of 1–4
Kgdms are contaminated to some extent by different textual traditions.
While older studies habitually oppose the “B text” of Codex Vaticanus
25 Barthélemy, Devanciers, 140–141.

26 Barthélemy tried to justify the partial nature of the revision for the section 2 Kgdms–
3 Kgdms 2:11, but his explanations are not convincing, see ibid., 141–143.
27 The first scholar who proposed this hypothesis seems to be Emanuel Tov, see E. Tov,
“The Methodology of Textual Criticism in Jewish Greek Scriptures, with Special Attention
to the Problems in Samuel–Kings: The State of the Question: Problems and Proposed Solu-
tions,” in 1972 Proceedings: International Organization for Septuagint and Cognate Studies
and the SBL Pseudepigrapha Seminar (SCS 2; ed. R. Kraft; SBL, 1972), 3–15, reprinted in
E. Tov, The Greek and Hebrew Bible (Leiden: Brill, 1999), 489–499.
230 jan joosten
to the “Antiochene text” of minuscules 19, 82, 93, 108 and 127, and the
“Hexaplaric text” of Codex Alexandrinus, recent investigations indicate
that this terminology is at best approximative:
– In the non-kaige sections, Codex Vaticanus is expected to reflect the

Old Greek, but it also contains a number of recensional readings.
– Codex Alexandrinus and its congeners usually give a secondary text, but
here and there they alone appear to have preserved the most ancient
version available.
– The text of minuscules 19, 82, 93, 108, 127 is supposed to be imperme-
able to the kaige recension throughout 1–4 Kgdms, but it is not: several
kaige readings are found in it.
The textual situation of the Greek tradition in Kingdoms is very compli-

cated and the history of investigation has seen the emergence of many
hypotheses that later turned out to be unhelpful.28 The ideas of Thackeray
and Barthélemy, widely accepted as they are (although with slight modi-
fications), provide no easy protocol for deciding which Greek reading if
any reflects the Old Greek. They do however afford a certain perspective
when it comes to comparing the Greek data with the MT and other tex-
tual witnesses.
4. Case studies illustrating the Use of the LXX in Textual

Criticism of the Hebrew Text
As was stated above, the Greek text of 1 Kgs 1–2, in both the Old Greek and
the kaige recension, is rather faithful to its Hebrew Vorlage (or Vorlagen).
Nevertheless, there are a surprising number of divergences between the
received Hebrew text and the Greek tradition, particularly the Antiochene
text. No doubt some of these divergences go back to the Greek translator.
Some of them, however, appear to reflect a divergent Hebrew text. Among
28 To give only one notable example, Frank Cross has developed the idea of a “proto-
Lucianic” recension, postulating that the Lucianic/Antiochene tradition contains two dis-
tinct recensional strata: the “Lucianic” revision of the 4th Century, and an earlier revision
on the basis of a Hebrew manuscript textually akin to the 4QSama fragments. This theory
has proved almost impossible to root out, even although only Cross’ students ever really
supported it, against the unanimous agnosticism of other competent scholars. As has been
shown by Tuukka Kauhanen of Helsinki University in his recent doctoral dissertation, this
theory rests on no certain foundation. See T. Kauhanen, The Proto-Lucianic Problem in
1 Samuel (Ph.D. diss., Helsinki, 2011 [to be published by Vandenhoeck & Ruprecht, Göt-
tingen]).
these are several variants that must be preferred to the reading of the
MT on internal grounds. We will explore two passages where the Greek
tradition may be held to offer a superior reading and one where, although
probably reflecting a Hebrew text, it is clearly secondary.
4.1. 1 Kgs 1:8
‫שׁר‬
֣ ֶ ‫בּוֹרים ֲא‬
֖ ִ ִ‫יא וְ ִשׁ ְמ ִ ֣עי וְ ֵר ֔ ִעי וְ ַהגּ‬
֙ ‫וּבנָ יָ֙ הוּ ֶבן־יְ הוֹיָ ָ ֜דע וְ נָ ָ ֤תן ַהנָּ ִב‬
ְ ‫וְ ָצ ֣דוֹק ַהכּ ֵֺהן‬ BHS
‫ם־אד ֺנִ ָיּֽהוּ‬
ֲ ‫ְל ָדִו֑ד ל֥ ֺא ָהי֖ וּ ִע‬
“But Zadok the priest, and Benaiah the son of Jehoiada, and Nathan the prophet,
and Shimi and Rei, and the Heroes of David, were not with Adonijah.”
‫שׁר‬
֣ ֶ ‫בּוֹרים ֲא‬
ִ ִ‫יא וְ ַשׁ ָמּא וְ ֵר ָעיו ַהגּ‬ ְ ‫ וְ ָצ ֣דוֹק ַהכּ ֵֺהן‬OHB
֙ ‫וּבנָ יָ֙ הוּ ֶבן־יְ הוֹיָ ָ ֜דע וְ נָ ָ ֤תן ַהנָּ ִב‬
‫ם־אד ֺנִ ָיּֽהוּ‬
ֲ ‫ְל ָדִו֑ד ל֥ ֺא ָהי֖ וּ ִע‬
“But Zadok the priest, and Benaiah the son of Jehoiada, and Nathan the prophet,
and Shammah and his comrades, the Heroes of David, were not with Adonijah.”
OHB Apparatus
‫ *ושמא ורעיו הגבורים‬GL (Σαμαιας καὶ οἱ ἑταῖροι αὐτοῦ οἱ ὄντες δυνατοί) ] ‫ושמעי ורעי‬
‫ והגבורים‬M
The lemma is taken from the eclectic text as edited in the OHB. The asterisk
shows that the Hebrew text is not attested as such, but retroverted from the Luci-
anic Greek (GL), quoted in parentheses. The non-preferred variant is quoted from
the MT (M). No other textual witnesses are indicated because they agree with
MT in the expected way: this is true for the Syriac, the Targum, and the Vulgate,
but also for the non-Lucianic Greek (καὶ Σεμεϊ καὶ Ρηι καὶ οἱ δυνατοί) which here
reflects the kaige recension.29
In MT, the proper names “Shimei and Rei” fall completely from the blue
sky. Occurring as they do between the well-known Zadok, Benaiah and
Nathan on the one hand and the famous group of “David’s heroes” (cf.
2 Sam 23:8) on the other, they are hard to interpret. The MT is hardly
correct, even although it is confirmed by the B text of the LXX (heavily
influenced by the kaige recension in this passage), by the Peshitta, the
Targum and the Vulgate.
The Antiochene text has several important variants. Where MT reads
the proper noun Rei, the Antiochene text has a common noun with a pos-
sessive pronoun: “his comrades.” The Greek reflects a different reading
29 Flavius Josephus reads differently from both the MT and the Antiochene Text: “And
Shimei, David’s friend, and all the bravest (warriors)” (Ant. 7.346).
232 jan joosten
of the consonants attested in MT: the waw was not attached to the word
‫ הגבורים‬as in MT, but to the letters ‫ ורעי‬to form the word ‫( ורעיו‬cf. rēϲāyw,
“his friends” Job 32:3).30 In addition, the first proper name seems to have
been read differently: in the LXX, Σαμαια(ς) corresponds most often to
‫ שמעיה‬and is found also for ‫( ישמעיה‬1 Chron 12:4) ‫( שמע‬Neh 8:4) and
‫( שמא‬2 Sam 23:11)—but nowhere, excepting this passage, for ‫שמעי‬.
Without being overly transparent, the Antiochene text makes good
sense in the context. The “heroes” (‫הגבורים‬, οἱ δυνατοί) of David are a
well-defined group of warriors accompanying David throughout his mili-
tary career. In 2 Sam 23:8–39, they are listed by name and divided into
two categories: the three, and the thirty.31 The third among “the three” is
‫—שמא‬Σαμαια(ς) in Greek (v. 11).32 Apparently, then, Shammah is named
in our verse as the leader of the famous “thirty” heroes of David at the
time of Adonijah’s coup. Why the third name in the list of 2 Sam 23 was
singled out and not the first (or the second) is impossible to say. The text
suggests that, between the glorious period described in 2 Sam 23:8–39
and the end of David’s life, something happened that made Shammah
the principal figure instead of Ishbaal.
In weighing the two alternative versions, no importance should be given
to the rule lectio difficilior potior. If MT is corrupt, as many critics admit,
its difficulty gives no indication of originality. The decision between the
two versions hinges exclusively on the possibility to explain one of them
as secondary to the other. Two scenarios seem possible:
– The Antiochene version is a facilitating interpretation of the corrupted

text of MT: the translator, or the Hebrew scribe producing the imme-
diate Vorlage of the Antiochene text, creatively rearranged the conso-
nants of proto-MT in order to arrive at a meaningful text. Casting about
for a likely identification of the proper nouns in his Vorlage, he hit upon
Shammah in the list of heroes in 2 Sam 23 and decided this was good
enough for his purposes.
30 Cf. BHS (where however there is no indication as to whether or not the Antiochene
reading is preferred).
31 The text of the list is much deteriorated in the MT and needs to be partly restored
on the basis of parallels in Chronicles and of the Greek versions. See S. R. Driver, Notes on
the Hebrew Text and the Topography of the Books of Samuel (2nd ed.; Oxford: Clarendon,
1913), 362–363.
32 In the Antiochene text of 2 Sam 23:11 it is clear that Samaias comes after Iesbaal
and Eleazar as the third of the three. The precise form of the proper name in the Hebrew
Vorlage of the Antiochene text can of course only be guessed.
– Alternatively, the Hebrew reading reflected in the Antiochene is the

original from which the proto-MT corrupted. A possible reason for
the corruption would be the failure of an ancient scribe to identify the
Shammah to which his Vorlage referred. In view of the deplorable tex-
tual state of the list of heroes in 2 Sam 23, this seems to be a reasonable
hypothesis. Instead of Shammah, the scribe introduced the much more
frequent proper name Shimei. Once Shammah became Shimei, the link
between the proper name and the list of David’s heroes was broken,
leading to the further corruption of ‫ ורעיו‬into ‫ ורעי ו‬with ‫ רעי‬being
taken as a second proper name.
Since the whole principle of “tendentious palaeographical exegesis” is

moot,33 it seemed to us that the second scenario is more likely than the
first.34
4.2. 1 Kgs 2:3
֨‫מר ֻחקּ ָ ֺ֤תיו ִמ ְצו ָֺתיו‬

ֺ ֨ ‫הו֣ה ֱאל ֶֺ֗היָך ָל ֶל ֶ֤כת ִבּ ְד ָר ָכיו֨ ִל ְשׁ‬
ָ ְ‫ת־מ ְשׁ ֶ ֣מ ֶרת ׀ י‬
ִ ‫וְ ֽ ָשׁ ַמ ְר ָ֞תּ ֶא‬ BHS
‫וֹתיו‬ ָ֔ ‫וּמ ְשׁ ָפּ ָ ֣טיו וְ ֵע ְד‬
ִ
“And keep the charge of the LORD your God, to walk in his ways, to keep his
statutes, his commandments, and his judgments, and his testimonies.”
֨‫מר ֻחקּ ָ ֺ֤תיו ִמ ְצו ָֺתיו‬

ֺ ֨ ‫הו֣ה ֱאל ֶֺ֗היָך ָל ֶל ֶ֤כת ִבּ ְד ָר ָכיו֨ ִל ְשׁ‬ ִ ‫ וְ ֽ ָשׁ ַמ ְר ָ֞תּ ֶא‬OHB
ָ ְ‫ת־מ ְשׁ ֶ ֣מ ֶרת ׀ י‬
‫וּמ ְשׁ ָפּ ָ ֣טיו‬
ִ
OHB Apparatus
‫ ] משפטיו‬+ ‫ ועדותיו‬M GL, > G(-L)
The apparatus indicates that after the lemma there is a plus in MT (M) and in the
Antiochene text (GL). This element is lacking in the Greek tradition excepting the
Antiochene text (G(-L)).
What is striking in this example is that the non-Masoretic variant is

attested in the main witnesses of the Septuagint, but not in the minus-
cules representing the Antiochene text. This is contrary to expectation,
33 See Tov, Text-critical Use, 100–103.

34 The reading has been adopted by several other textual critics: See, e.g., H. Oort,
Textus hebraici emendationes, quibus in Vetere Testamento neerlandice vertendo usi sunt
(Leiden: Brill, 1900), 45. It is rejected in B. Stade and F. Schwally, The Sacred Books of the
Old Testament. A Critical Edition Of The Hebrew Text, Part 9: The Books of Kings (Leipzig:
Hinrichs, 1904), 62: “ingenious but merely conjectural.”
234 jan joosten
but it is not a unique constellation. The non-Antiochene Septuagint may

reflect the Old Greek here, or, perhaps more probably, it may reflect the
kaige recension made on the basis of a Hebrew Vorlage that didn’t yet
have the addition.
The secondary nature of the phrase ‫ ועדותיו‬in MT (followed by the
Antiochene [καὶ τὰ μαρτύρια αὐτοῦ], Peshitta, Targum, and Vulgate) is
shown by:
– the lack of a counterpart to this element in the non-Antiochene

Septuagint;
– the unusual distribution of the connective waw in MT (“his statutes, his
commandments, and his judgments, and his testimonies); one would
expect either: A, B, C, and D, or: A, and B, and C, and D.
– the fact that ‫ עדות‬nowhere else co-occurs with the three other terms
of the list.
The phrase ‫ ועדותיו‬may have been added at a late stage through assimila-
tion to other texts where ‫ עדות‬figured in lists of this type (although not
with the exact same elements).
*
A common objection to text-critical operations on the Hebrew Bible is
that one should not take elements from one textual tradition to mend
another one.35 It is certainly true that MT and Septuagint represent differ-
ent streams of tradition in 1 Kings. As these two examples show, however,
textual emendations are not based solely on a divergence between wit-
nesses. Their foremost justification is a problem in the received Hebrew
text. In the two passages discussed, the MT is indeed anomalous to a cer-
tain extent. The comparison with other textual witnesses helps to identify
what is problematic in the MT, and it shows a way to a textual solution.
The emendation, however, is in a way called forth by the Masoretic tradi-
tion itself.36
35 See, e.g., D. Barthélemy, Études d’histoire du texte de l’Ancien Testament (OBO 21; Fri-
bourg / Göttingen: Editions Universitaires / Vandenhoeck & Ruprecht, 1978), 369.
36 I owe this insight to my former doctoral student, René Pfertzel.
4.3. 1 Kgs 2:1
‫מר‬
ֺ ֽ ‫מה ְבנ֖ וֹ ֵלא‬
ֺ ֥ ֺ ‫ת־שׁל‬
ְ ‫י־דִ ֖וד ָל ֑מוּת וַ יְ ַצ֛ו ֶא‬
ָ ‫וַ יִּ ְק ְר ֥בוּ יְ ֵ ֽמ‬ BHS
“Now the days of David drew nigh that he should die; and he charged Solomon
his son, saying”
A B37 OHB
‫י־דִ ֖וד ָל ֑מוּת‬ ָ ‫וַ יִּ ְק ְר ֥בוּ יְ ֵ ֽמ‬ ‫ויהי אחרי כן וימת דוד וישכב עם אבותיו‬
‫מר‬
ֺ ֽ ‫מה ְבנ֖ וֹ ֵלא‬
ֺ ֥ ֺ ‫ת־שׁל‬
ְ ‫וַ יְ ַצ֛ו ֶא‬ ‫ויצו את בנו שלמה לפני מותו לאמר‬
OHB Apparatus
A = M, B = GL (καὶ ἐγένετο μετὰ ταῦτα καὶ ἀπέθανε Δαυιδ καὶ ἐκοιμήθη μετὰ τῶν
πατέρων αὐτοῦ καὶ ἐνετείλατο τῷ υἱῷ αὐτοῦ Σολομῶντι ἔμπροσθεν τοῦ θανάτου αὐτοῦ
λέγων)
Where textual variants line up in such a way as to indicate the presence of two
Hebrew recensions, the OHB format permits to print two parallel columns.38 We
have taken advantage of that possibility for the present verse. The A column, the
earliest text form according to us, contains the MT (M), the B column a Hebrew
text reconstructed on the basis of the Antiochene Greek (GL).
The narrative verse introducing David’s last will is transmitted in two dif-
ferent textual forms, the one represented by the MT, the main witnesses
of the Septuagint, Peshitta, Targum, and Vulgate, the other by the Antio-
chene text.
The Antiochene Greek, καὶ ἐγένετο μετὰ ταῦτα καὶ ἀπέθανε Δαυιδ καὶ
ἐκοιμήθη μετὰ τῶν πατέρων αὐτοῦ καὶ ἐνετείλατο τῷ υἱῷ αὐτοῦ Σολομῶντι
ἔμπροσθεν τοῦ θανάτου αὐτοῦ λέγων, can easily be translated into Hebrew
and probably goes back to a Hebrew Vorlage:
– καὶ ἐγένετο μετὰ ταῦτα = ‫( ויהי אחרי כן‬see Judg 16:4 etc.)
– καὶ ἐκοιμήθη μετὰ τῶν πατέρων αὐτοῦ = ‫( וישכב עם אבותיו‬see 1 Kgs 2:10
etc.)
– ἔμπροσθεν τοῦ θανάτου αὐτοῦ = ‫( לפני מותו‬cf. 1 Chron 22:5 [LXX ἔμπροσθεν
τῆς τελευτῆς αὐτοῦ]).
37 “And it happened after this that David died and slept with his fathers; but before his
death, he commanded his son Solomon, saying . . .”
38 In 1 Kgs, this is a felicitous possibility. In many passages, the textual facts do indeed
indicate the presence of two (sometimes more) recensions, yet, it is not always possible to
indicate which recension is earlier. The two-column format permits the presentation of all
the textual evidence without needing to decide on the question of priority.
236 jan joosten
But this divergent Hebrew text is almost certainly secondary. In the books
of Kings, the expression ‫ וישכב עם אבותיו‬is always used as a euphemism
for “he died”.39 In Kings, it never combines with the verb ‫“ מות‬to die”.40
The combination of ‫ וישכב עם אבותיו‬with ‫ וימת‬cannot therefore repre-
sent the earliest stage of the text. The alternative version represented by
GL appears to be a later rewriting, drawing on 1 Kgs 2:10 (‫וַ יִּ ְשׁ ַ ֥כּב ָדִּ ֖וד ִעם־‬
‫ ) ֲאב ָ ֺ֑תיו‬and perhaps on a reminiscence of 1 Chron 22:5.
5. Conclusions
In 1 Kings (as in Samuel–Kings in general), the Old Greek is not readily

available, but needs to be recovered from the jungle of attested Greek
text forms. In kaige sections—such as 3 Kgdms 1–2:11—it must usually
be looked for in the Antiochene tradition, although this too represents a
mixture of primitive and secondary elements. Where the Old Greek can
be attained, it is seen to reflect a very good and old Hebrew text that is in
places superior to the MT. In the first two chapters of 1 Kings, we emended
the received Hebrew text twelve times on the basis of the Septuagint, of
which at least five times in a way that crucially alters the meaning of the
passage (as in 1 Kgs 1:8 discussed above).41 Some of our emendations will
perhaps not pass muster, but the large number gives an indication of the
important potential of the Greek tradition in textual studies on 1 Kings.
As the last example indicates, 3 Kgdms represents not only a differ-
ent Hebrew text, but also a different edition of the book. In the first two
chapters, this aspect of the relation between the MT and the Septuagint
is not prominent. But starting in chapter 3, it will become an important
issue. 1 Kings is one of the books (like Jeremiah and Daniel) where tex-
tual criticism bleeds into redaction-historical criticism. Fortunately, this
dimension needs not be dealt with in the present paper.
39 See Stade and Schwally, Books of Kings, 66.

40 It does so combine however in the late biblical Hebrew of Chronicles, see 2 Chron
16:13.
41 The verses concerned are (an exclamation mark indicates important cases): 1 Kgs 1:2!,
8!, 9, 10 (2x), 11, 19, 35 ; 2:3, 4, 5 (2x)!, 22!
Gottesbildfragen in den Lesartendifferenzen zwischen
dem masoretischen und dem alexandrinischen
Text des Jeremiabuches
Hermann-Josef Stipp
1. Das Problem
Die ursprüngliche antike Übersetzung des Jeremiabuches ins Griechische

(JerG*) ist eine besonders interessante Zeugin für die literarische Genese
des Buches, weil sie eine Fassung bewahrt hat, die etwa um ein Siebtel
kürzer ist als die masoretische Ausgabe und in der Makrostruktur sowie
bisweilen auch der Mikrostruktur von jener abweicht1. Dazu kommen
qualitative Varianten, die an Bedeutung jedoch weit hinter den quantitati-
ven zurückbleiben. Die textkritische Auswertung von JerG* wird erleichtert
durch ihre typische Übersetzungssprache, die hochgradig standardisiert
ist und wenig Zugeständnisse an die Stilistik und Idiomatik des zeitge
nössischen Griechischen macht. Diese Eigenart gestattet, die hebräische
Vorlage, wo sie von MT abweicht, zumeist mit befriedigender Sicherheit
wiederherzustellen. Seitdem Fragmente des Jeremiabuches aus Qumran
veröffentlicht wurden, die der rekonstruierten JerG*-Vorlage nahestehen
(4Q71 bzw. 4QJerb und 4Q72a bzw. 4QJerd)2, setzt sich zunehmend die
Überzeugung durch, dass die durch JerG* repräsentierte Textform—sie
sei hier die alexandrinische genannt (AlT)—der masoretischen Ausgabe
(MT) an Alter vorausliegt. Ich habe mich diesem Urteil andernorts in dem
Sinne angeschlossen, dass JerAlT gegenüber JerMT eine globale Priorität
innehat; d. h. JerAlT ist kein Vorfahr von JerMT in direkter Linie, hat
sich aber seit der Gabelung der Textüberlieferungsstränge in erheblich
1 Vgl. z. B. E. Tov, The Text-critical Use of the Septuagint in Biblical Research, Second
Edition, Revised and Enlarged (JBS 8; Jerusalem: Simor 1997), 243f.—Idem, Der Text der
Hebräischen Bibel. Handbuch der Textkritik (Stuttgart: Kohlhammer 1997), 265, bemisst die
Unterschüsse von JerG* sogar auf ein Sechstel des Buches.
2 Eine vorläufige Edition wurde erstellt von J. G. Janzen, Studies in the Text of Jeremiah
(HSM 6; Cambridge: Harvard University Press, 1973), 173–184; die editio princeps findet sich
bei E. Tov, „4QJera–e,“ in Qumran Cave 4. X. The Prophets (DJD XV; Hg. E. Ulrich; Oxford:
Clarendon Press, 1997), 145–207.
238 hermann-josef stipp
geringerem Maß weiterentwickelt als JerMT 3. Hier sei lediglich nochmals

jene Datenklasse hervorgehoben, der m. E. die größte Beweiskraft zugun-
sten dieser Verhältnisbestimmung zukommt: Die Sonderlesarten von
JerMT besitzen eine starke sprachliche Eigenprägung durch den präma
soretischen Idiolekt, ein Repertoire von sprachlichen Merkmalen, die im
masoretischen Sondergut jeweils mindestens doppelt bezeugt sind, aber
in dem mit JerAlT geteilten Bestand des Buches oder sogar im gesamten
Rest des AT fehlen. Derzeit (Mai 2012) zähle ich 82 solche Phänomene
mit zusammen über 250 Belegen; davon sind 38 Merkmale mit zusam
men 92 Fällen ausschließlich in den masoretischen Sonderlesarten des
Jeremiabuches bezeugt4. Zusätzlich hat Jan Joosten auf Züge des nach-
klassischen Hebräischen im masoretischen Sondergut hingewiesen5. Da
ein derartiges Korpus nicht aus Kürzungen herrühren kann, ist der Befund
nur erklärbar, wenn das masoretische Sondergut weit überwiegend aus
Nachträgen besteht, die auf eine kleine Zahl von Ergänzern zurückgehen.
Weil indes die globale Priorität von JerAlT definitionsgemäß ein geringes
Maß an Eigenentwicklung einschließt, bleibt die Einzelprüfung der Vari-
anten unverzichtbar.
Damit steht die generelle Verhältnisbestimmung der beiden wichtigs
ten antiken Textformen des Jeremiabuches fest. Allerdings haben in den
vergangenen Jahren Georg Fischer und sein Schüler Andreas Vonach
in einer Serie von Publikationen zäh die gegenteilige These verfochten.
Für Fischer, der sogar dem prämasoretischen Idiolekt jede Beweiskraft
abspricht6, hat dem Schöpfer von JerG* nichts anderes als MT zur Vorlage
gedient; entsprechend gingen sämtliche Differenzen auf den Übersetzer
zurück, der ein beträchtliches Maß an Interpretation in sein Werk ein
fließen ließ, wie zumal der Umstand zeige, „dass Jer G dazu neigt, Gott
3 Vgl. H.-J. Stipp, Das masoretische und das alexandrinische Sondergut des Jeremia
buches. Textgeschichtlicher Rang, Eigenarten Triebkräfte (OBO 136; Freiburg / Göttingen:
Universitätsverlag / Vandenhoeck & Ruprecht, 1994), sowie die Aktualisierungen in idem,
„Zur aktuellen Diskussion um das Verhältnis der Textformen des Jeremiabuches,“ in Die
Septuaginta—Texte, Kontexte, Lebenswelten (WUNT 219; Hg. M. Karrer und W. Kraus;
Tübingen: Mohr Siebeck, 2008), 630–653.
4 Vgl. vorläufig H.-J. Stipp, „Linguistic Peculiarities of the Masoretic Edition of the Book
of Jeremiah. An Updated Index,“ JNSL 23/1 (1997): 181–202; sowie den in Anm. 3 genannten
Aufsatz. Ein vervollständigtes Inventar befindet sich in Vorbereitung.
5 J. Joosten, „L’excédent massorétique du livre de Jérémie et l’hébreu post-classique,“ in
Conservatism and Innovation in the Hebrew Language of the Hellenistic Period. Proceedings
of a Fourth International Symposium on the Hebrew of the Dead Sea Scrolls & Ben Sira (STDJ
73; Hg. J. Joosten und J.-S. Rey; Leiden: Brill, 2008), 93–108.
6 G. Fischer, „Die Diskussion um den Jeremiatext,“ in idem, Der Prophet wie Mose. Stu-
dien zum Jeremiabuch (BZAR 15; Wiesbaden: Harrassowitz, 2011), 73–89 (87–89).
gottesbildfragen in den lesartendifferenzen 239
milder, weniger menschlich [d. h. weniger anthropomorph, HJS] oder

gewalttätig zu schildern“7. Vonach erklärt in der Einleitung zur Wieder
gabe des griechischen Jeremiabuches in der „Septuaginta Deutsch“
(LXX.D): „Inhaltlich-theologisch bemüht sich die Septuaginta um ein mil
deres Sprechen Gottes selbst, aber auch über ihn . . . Sie formuliert generell
vorsichtiger und pietätvoller und betont mehr die Souveränität Gottes,
indem sie z. B. allzu anthropomorphe Regungen vermeidet.“8 Fischer und
Vonach widersprechen damit diametral dem Urteil von Bernard Zloto-
witz, JerG* lasse keine Neigung zur Reduktion von Anthropomorphismen
erkennen9. Mehr noch: Die Modifikationen am Gottesbild bildeten sogar
bloß einen exemplarischen Ausschnitt aus vielfältigen Freiheiten, die
sich der Übersetzer gestattete, um seinem Zielpublikum den Zugang zum
Jeremiabuch zu ebnen. In den Worten Fischers: „Jer-LXX bezeugt einen
Übersetzer, der als frommer Jude die Botschaft dieses Buches seinen
Zeitgenossen in ihrem hellenistisch-ägyptischen Kulturkreis nahebringen
wollte.“10 Vonach hat diese Überzeugung nunmehr zur Grundlage seiner
Erläuterung des griechischen Jeremiabuchs im Kommentar zur LXX.D
erhoben11.
Wie an anderer Stelle nochmals zu bekräftigen ist, liegt diese zusam
menfassende Charakterisierung von JerG* fernab der Wirklichkeit12. Das
griechische Jeremiabuch repräsentiert vielmehr eine „imitative technique
of translation“13, wie sie typisch ist für eine „ausgangssprachliche“ Wieder-
gabe, die „wörtlich, aber grammatisch und/oder inhaltlich nicht korrekt“
7 Ibid., 85.
8 A. Vonach, „Jeremias—Das Buch Jeremia. Einleitung,“ in Septuaginta Deutsch. Das
griechische Alte Testament in deutscher Übersetzung (Hg. W. Kraus und M. Karrer; 2., ver-
besserte Aufl. Stuttgart: Deutsche Bibelgesellschaft, 2010), 1288f. (1289).
9 B. M. Zlotowitz, The Septuagint Translation of the Hebrew Terms in Relation to God in
the Book of Jeremiah (New York: KTAV, 1981), mit dem Fazit S. 183: „The data throughout
this thesis have profusely demonstrated that the Septuagint translator(s ?) of the book of
Jeremiah reproduced literally and correctly the Hebrew terms related to God.“
10 G. Fischer, Jeremia. Der Stand der theologischen Diskussion (Darmstadt: Wiss. Buch-
gesellschaft, 2007), 52.
11 A. Vonach, „Jeremias,“ in Septuaginta Deutsch. Erläuterungen und Kommentare zum
griechischen Alten Testament, Bd. II: Psalmen bis Daniel (Hg. M. Karrer und W. Kraus;
Stuttgart: Deutsche Bibelgesellschaft, 2011), 2696–2814.
12 H.-J. Stipp, „Die Jeremia-Septuaginta als theologische Programmschrift. Zur Kommen-
tierung des griechischen Jeremiabuches im Rahmen der ‚Septuaginta Deutsch‘ (LXX.D),“
erscheint in: BZ.
13 J. Barr, The Typology of Literalism in Ancient Biblical Translations (MSU 15 = NAWG.
PH 11; Göttingen: Vandenhoeck & Ruprecht, 1979), 292.
verfährt14. „The most basic norm of the translator . . . was that of isomor-

phism. That is to say, typically each morpheme of the source text is rep-
resented in the target text.“15 Das Ergebnis muss seinerzeit einen überaus
fremdartigen, schwerfälligen Klang besessen haben, denn es strotzt von
Hebraismen und mutet seinen Benutzern nicht selten kaum verständli-
che Passagen zu. Der Wille, den Text für die Leserschaft aufzubereiten,
war in Wahrheit zwar nicht völlig inexistent, aber eng begrenzt. JerG*
enthält ein geringes Maß interpretierender Elemente, die eindeutig nicht
der Vorlage, sondern dem Übersetzer zuzurechnen sind. Um einige beson
ders klare, da mehrfach bezeugte Beispiele zu nennen: Der Klageruf ‫א ָההּ‬,ֲ
in Jer nur in Reden des Propheten an Jhwh belegt, wird auf die Wurzel
‫ היה‬bezogen und als Gottesanrede ὁ ὤν wiedergegeben16, eine offenkundig
von ExG* 3:14 inspirierte Deutung17. Das Substantiv ‫ נָ ִביא‬wird in Fällen,
die von Falschpropheten reden, häufig mit ψευδοπροφήτης übersetzt18. Für
das Verb ‫ קרא‬tritt, wenn in Aufforderungen Jhwhs an Jeremia gerichtet,
immer ἀναγινώσκω vorlesen ein, also auch dann, wenn der Kontext vom
Vortrag einer schriftlichen Fassung nichts weiß19.
Da ein beschränktes Ausmaß solcher Einflüsse nicht zu leugnen ist, soll
im Folgenden ein Kernpunkt der Position Fischers und Vonachs geprüft
werden, nämlich die These, der Übersetzer habe bevorzugt die Gottes
bilder seiner Vorlage entschärft20. Ein Vergleich der antiken Ausgaben
des Jeremiabuches unter dieser Rücksicht erscheint lohnend, weil JerG*
zwar nicht entfernt einen solch ausgeprägten Hang zu interpretierenden
Wiedergaben erkennen lässt, wie Fischer und Vonach behaupten, aber
in der Tat mehrfach an solchen Stellen von MT abweicht, wo Aspekte
der Gottesvorstellung zur Sprache kommen. Vorweg ist zu betonen, dass
14 F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die
Septuaginta (Münsteraner Judaistische Studien 9, Münster: Lit, 2001), 133.
15 A. Pietersma und M. Saunders, „To the Reader of Ieremias,“ in A New English Trans-
lation of the Septuagint and the Other Greek Translations Traditionally Included under that
Title (Hg. A. Pietersma und B. G. Wright; New York: Oxford University Press, 2007), 876–881
(876).
16 1:6; 4:10; 14:13; 32:17.
17 Vgl. Siegert, Einführung (Anm. 14), 253–255; E. G. Dafni, „Οἱ οὐκ ὄντες θεοί in der Sep-
tuaginta des Jeremiabuches und in der Epistel Jeremias. Ein Beitrag zur Frage nach dem
Werdegang des sogenannten alexandrinischen Kanons,“ in The Biblical Canons (BETL 163;
Hg. J. M. Auwers und H. J. de Jonge; Leuven: University Press, 2003), 235–245 (241f.).
18 6:13; 26:7, 8, 11, 16; 27:9/34:7; 28:1; 29:1, 8; ferner Sach 13:2.
19 3:12; 11:6; 19:2. Bei auch in MT vorausgesetzten Akten des Vorlesens: 29:29; 36:6, 8, 10,
13, 14, 15, ‌21, ‌23; 51:61, 63.
20 Vgl. schon die Bemerkungen bei Stipp, „Diskussion“ (Anm. 3), 651f., doch soll das Pro-
blem hier nochmals auf verbreiterter Beobachtungsbasis wiederaufgenommen werden.
kein objektivierbares oder gar automatisierbares Verfahren zur Erschlie-

ßung des einschlägigen Materials existiert. Die Fall für Fall vorgehende
Beurteilung, welche Devianzen die transportierten Gottesbilder beein-
flussen, ist nicht zu umgehen. Die unten besprochenen Stellen wurden
über längere Zeit hinweg entlang der Arbeit am Text des Jeremiabuches
gesammelt. Damit wird zwar keine Vollständigkeit, aber hinreichende
Repräsentativität beansprucht. Eingeschlossen sind die recht wenigen
diskussionswürdigen Beispiele, die Fischer und Vonach zur Stützung ihrer
These anführen21.
Selbstredend können hier unmöglich alle Argumente Fischers und
Vonachs zurückgewiesen werden. Dies geht schon aus Raumgründen
nicht, weil hier ihr Verfahren, bei der Rechenschaft über die relevanten
Fakten recht sparsam vorzugehen, nicht fortgesetzt werden kann. Über
gangen werden namentlich solche Fälle, die unerhebliche Differenzen
aufbauschen, offenkundige Fehler enthalten oder gravierende Tatsachen
ignorieren. So liest man etwa bei Vonach zu 2:20: „Die Vermeidung harter
Aussagen über oder von Gott wird oft lediglich durch eine einfache Ände-
rung der Kommunikationsstruktur oder des Subjekts erzielt. So spricht in
2,20 das Wort über die Hurerei in der LXX durch Weiterführung der direk-
ten Rede das Volk selbst, in MT hingegen Gott.“22 Vonach scheint nicht
zu wissen, dass die Lesung διαχυθήσομαι auf den Verlust des Spatiums in
‫ ַא ְתּ צ ָֹעה‬zurückgeht, das grammatisch inkorrekt als ‫ ֶא ְת ָצ ֶעה‬gelesen wurde
(richtig: ‫)א ְצ ָט ֶעה‬.ֶ Die Verschiebung des Sprecherwechsels resultiert daher
nicht aus einem interpretierenden Eingriff des Übersetzers, sondern aus
einer fehlerhaften Vorlage.—Ebenso deutet Vonach das Nebeneinander
von ‫ יָ ק ְֹשׁ ִתּי ָלְך‬und ἐπιθήσονταί σοι in 50:24 als „Subjektänderung“ des Über-
setzers zur Milderung des Gottesbildes23. Dabei deutet das Verb auf die
Vorlage ‫יָ ִשׁתוּ‬, die durch Ausfall des ‫ ק‬und Anpassung der Endung aus der
masoretischen Lesart hervorging.—Weiterhin wird erklärt: „Viele solche
‚Korrekturen‘ des Gottesbildes werden in der LXX außerdem dadurch
erreicht, dass in 1. Sg. formulierte Aussagen Gottes oder solche in 3. Sg.
über ihn in ein unpersonales Passiv oder eine indifferente Nominalkon-
struktion umformuliert werden.“24 Das soll beispielsweise für 6:2 gelten.
Dabei fehlt der unerlässliche Verweis auf die Äquivalenz ‫יתי‬ ִ ‫ ≙ ָדּ ִמ‬τὸ ὕψος
σου, die zeigt, dass in der Vorlage das (offenbar defektiv geschriebene)
21 Vgl. unten Anm. 30, 34, 43, 46, 49, 63, 74, 83, 99.
22 Vonach, „Jeremias“ (Anm. 11), 2725.
23 Ibid.
24 Ibid.
Prädikat durch eine gängige ‫ר‬/‫ד‬-Verwechslung zu der nominalen Form

‫*ר ָמ ֵתְך‬
ָ verstümmelt und daher verlorengegangen war, was den Übersetzer
nötigte, aus dem Kontext ein neues Prädikat zu erraten. Dabei fiel seine
Wahl auf ‫וְ ַה ְמּ ֻענָּ גָ ה‬, eine Bildung, die seine Hebräischkenntnisse ohnehin
vor eine hohe Hürde gestellt haben dürfte. Das Vorurteil, die meisten
Devianzen gingen auf den Übersetzer zurück, hindert Vonach jedoch
daran, eine abweichende Vorlage überhaupt in Betracht zu ziehen. Die
gegebenen Umstände bilden freilich eine enorme methodische Erschwer-
nis, will man aus JerG* konzeptionelle Differenzen zu MT erheben.—Zu
49:37 (AlT 25:17) bemerkt Vonach: „In 25,17 richtet in MT Gott selbst Elam
aktiv zugrunde (‫‚ עד כלותי אותם‬bis ich sie zugrunde gerichtet habe‘),
während das zugrunde richtende Subjekt in der LXX durch die Wieder-
gabe im Infinitiv das Schwert des vorangehenden Satzgliedes ist“25. Dabei
wird übersehen, dass Jhwh in JerG* dieses Schwert ausdrücklich als τὴν
μάχαιράν μου mein Schwert reklamiert. Das auf ihn zurückverweisende
Pronomen hat lediglich die Position gewechselt, und zwar derart, dass
an der göttlichen Urheberschaft kein Zweifel bleibt.—Argumentationen,
wie sie diese Beispiele repräsentieren, sind bei Fischer und Vonach leider
keine Seltenheit.
Die Lesartenunterschiede sind zu befragen, ob sie dazu beitragen,
JerMT und/oder JerAlT mit distinkten konzeptionellen Profilen zu ver-
sehen. Denn selbstverständlich können Differenzen nicht a priori dem
Übersetzer angelastet werden. Vielmehr kommen stets mindestens drei
verschiedene Ursachen in Betracht, zwischen denen zu entscheiden ist, was
indes nicht immer gelingt: (1) MT hat sich von AlT wegentwickelt; (2) die
hebräische Fassung von AlT hat sich von MT entfernt; (3) der Schöpfer
von JerG* ist von seiner Vorlage abgewichen26. Es gilt jeweils zu klären,
welche dieser Quellen die größte Wahrscheinlichkeit besitzt. Nur wenn
Ort und Richtung einer Modifikation bestimmt sind, lassen sich auch ihre
Motive eingrenzen. Konvergenzen können Hinweise liefern auf Trends
oder gar spezifische Theologien, die in dem betroffenen Texttyp bzw. in
der Übersetzung wirksam sind.
Die nachstehende Erörterung listet Lesartendifferenzen auf, die Anlass
zur Frage geben, ob bei ihrer Entstehung die Reflexion von Gottes
25 Ibid.
26 Die drei Alternativen vermeiden das methodische Manko der Arbeiten Fischers, der
aufgrund seiner Vorentscheidung, Lesartdifferenzen seien immer den Übersetzern zuzu
schreiben, niemals die Möglichkeit einer nichtmasoretischen Vorlage von JerG* in Betracht
zieht. In abgemilderter Form vertritt dieses Apriori auch Vonach, „Jeremias“ (Anm. 11).
bildproblemen eine Rolle gespielt hat. Die Analyse wird in einen Versuch
münden, die Belege in Klassen zu ordnen, doch um nichts zu präjudi
zieren, wird das Material nach dem masoretischen Bucharrangement
dargeboten. Zitate aus dem Hebräischen folgen den Konventionen einer
textkritischen Synopse zum Jeremiabuch, die sich in Vorbereitung befin
det27. Rückübersetzte alexandrinische Lesarten werden mit tiberischer
Vokalisation versehen, die als Symbolsystem zur Repräsentanz einer
bestimmten Interpretation des Konsonantentextes dient. Eine Reihe
diakritischer Symbole hebt alexandrinische Abweichungen hervor, die auf
die JerG*-Vorlage zurückgeführt werden:
[. . .] Eckige Klammern markieren masoretische Überschüsse.

<. . .> Spitze Klammern bezeichnen Überhänge der alexandrinischen Aus-
gabe.
. . . \ . . . Schrägstriche geben qualitative Varianten an: zuerst in Leserichtung die
masoretische Lesart, danach die alexandrinische.
* Drehachse: Die durch das Drehachsensymbol getrennten Passagen ste-
hen im alexandrinischen Text in umgekehrter Reihenfolge.
#. . .§ Bei Bedarf stecken Grenzzeichen die Reichweite einer Notation ab.
(. . .) Runde Klammern fügen erforderlichenfalls weitere Informationen
hinzu.
| Ein senkrechter Strich bezeichnet Satzgrenzen in AlT.
≙ entspricht
Gram! ungrammatisch
Abkürzungen:
TT Tiberischer Text (d. h. MT in tiberischer Vokalisation).
JE12 Die Büchergruppe Jer, Ez und Dodekapropheton, die wahrscheinlich
von demselben Übersetzer(kreis) ins Griechische übertragen wurde28.
Nötigenfalls werden die Verse in Sätze zerlegt (a, b, c, . . .). Wo es geboten

erscheint, wird auch der griechische Text beigegeben. Dagegen werden
für den vorliegenden Zusammenhang unerhebliche Varianten ausgespart.
Die andersartige Kapitel- und Verszählung von AlT in Jer 26–51 wird nur
bei Bedarf nach Schrägstrich beigegeben.
27 H.-J. Stipp, Textkritische Synopse zum Jeremiabuch (Manuskript, 7. korrigierte interne

Auflage, München, 2011).
28 Vgl. Stipp, Sondergut (Anm. 3), 17–19.
2. Das Material29
1:1 ‫ יִ ְר ְמיָ הוּ‬# ‫( | ֲא ֶשׁר ָהיָ ה ֶאל‬τοῦ θεοῦ) ‫§ ִדּ ְב ֵרי \ ְדּ ַבר־יְ הוָ ה‬
τὸ ῥῆμα τοῦ θεοῦ ὅ ἐγένετο ἐπὶ Ιερεμίαν
MT deklariert den Inhalt des Buches als Worte Jeremias, während AlT
das Folgende überschreibt als Wort Jhwhs, das an Jeremia erging. JerG*
verwendet typisches Übersetzungsgriechisch (ῥῆμα . . . ἐγένετο ἐπί) und
untermauert so, dass die Variante eine hebräische Vorlage spiegelt. Diese
Fassung verdoppelt die Wortereignisformel V. 2 und unterstreicht die
Kontrolle Jhwhs über das Offenbarungsgeschehen, was für den sekundä
ren Einfluss theologischer Reflexion spricht. Somit wurde in der alexan
drinischen Texttradition die Souveränität Jhwhs gesteigert, freilich schon
in der hebräischen Phase, nicht durch den griechischen Übersetzer30. Ein
analoger Eingriff ist in 51:59 AlT zu beobachten (s. u.).
1:17 ‫יהם‬
ֶ ֵ‫ ִל ְפנ‬# ‫ל־תּ ַחת‬ֵ ‫ן־א ִח ְתָּך \ וְ ַא‬ֲ ‫ § ֶפּ‬f ‫יהם‬
ֶ ֵ‫ל־תּ ַחת \ ִתּ ָירא ִמ ְפּנ‬
ֵ ‫ ַא‬e
<‫ילָך נְ ֻאם־יְ הוָ ה‬
ֶ ‫י־א ְתָּך ֲאנִ י ְל ַה ִצּ‬
ִ ‫> ִכּ‬
μὴ φοβηθῇς ἀπὸ προσώπου αὐτῶν μηδὲ πτοηθῇς ἐναντίον αὐτῶν ὅτι μετὰ
σοῦ ἐγώ εἰμι τοῦ ἐξαιρεῖσθαί σε λέγει κύριος
In MT trägt Jhwh eine scharfe Drohung vor: Wenn Jeremia sich von sei
nen Gegnern einschüchtern lässt, wird Jhwh diesen Effekt noch zusätzlich
in die Höhe treiben. AlT dagegen bietet die mit ‫ ירא‬und ‫ חתת‬gebildete
erweiterte Beruhigungsformel im üblichen Vetitiv: ‫ל־תּ ַחת‬ ֵ ‫ל־תּ ָירא וְ ַא‬
ִ ‫ ַא‬31.
Es folgt eine weitere Zusage des Mitseins nebst Gottesspruchformel mit
wörtlichen Parallelen in 8b—dort nach der einfachen Beruhigungsformel—
und 19c. Das Übersetzungsgriechische (ἐγώ εἰμι für ‫ ; ֲאנִ י‬τοῦ + Infinitiv für
‫ ל‬+ Infinitivus constructus) bestätigt die Abkunft von einer hebräischen
Vorlage. Der Überhang widerspricht dem früher beliebten Standpunkt,
die alexandrinische Tradition habe Dubletten getilgt; eher ist JerAlT
hier umgekehrt konflationär angereichert worden. Laut dieser Textform
29 Der Text von JerG* ist entnommen aus der Edition von J. Ziegler, Ieremias, Baruch,
Threni, Epistula Ieremiae (Septuaginta. Vetus Testamentum Graecum Auctoritate Acade-
miae Scientiarum Gottingensis editum 15, Göttingen: Vandenhoeck & Ruprecht ²1976).
Übersetzungen von JerG* sind angelehnt an die Wiedergabe von G. Fischer und A. Vonach
in LXX.D: Septuaginta Deutsch. Das griechische Alte Testament in deutscher Übersetzung
(Hg. W. Kraus und M. Karrer; Stuttgart: Deutsche Bibelgesellschaft, 2009), 1288–1342.
30 So aber Vonach, „Jeremias“ (Anm. 11), 2712.
31 Jer 30:10 MT || 46:27; Dtn 1:21; Jos 8:1; 10:25; Jes 51:7; Ez 2:6; 3:9; 1 Chr 22:13; 28:20; 2 Chr
20:15, 17; 32:7; als Prohibitiv: Dtn 31:8; vgl. Narrativ 1 Sam 17:11.
gewährt Jhwh dem Propheten uneingeschränkten, in klischierte Ter-

minologie gefassten Zuspruch. Der Überschuss, der nur mit der alex-
andrinischen Fassung von 17ef harmoniert, und der konventionellere
Sprachgebrauch deuten auf eine sekundäre Milderung des Gottesbildes,
indem man die Drohung durch eine unbedingte Beistandszusage ersetzte,
sodass der schroffe Umgang Jhwhs mit Jeremia entfiel. Der Eingriff wurde
allerdings nicht vom Übersetzer vorgenommen, sondern fand bereits in
der vorausliegenden hebräischen Tradition statt.
6:11 ‫אתי‬
ִ ‫ ָמ ֵל‬# ‫ וְ ֵאת § ֲח ַמת יְ הוָ ה \ ֲח ָמ ִתי‬a
(συνετέλεσα √ ‫יתי ָה ִכיל )כלה‬ ִ ‫ נִ ְל ֵא‬b
‫חוּרים יַ ְח ָדּו‬
ִ ‫( וְ ַעל סוֹד ַבּ‬ἔξωθεν) ‫ל־עוֹלל ַבּחוּץ‬ ָ ‫ > ֶא< ְשׁפְֹך ַע‬c
on der Zornesglut Jhwhs bin ich voll; ich bin überdrüssig, (sie) zurück-
V
zuhalten.
Gieß (sie) aus über den Säugling auf der Gasse und über die Runde der
Burschen zugleich!
καὶ τὸν θυμόν μου ἔπλησα καὶ ἐπέσχον καὶ οὐ συνετέλεσα αὐτούς ἐκχεῶ
ἐπὶ νήπια ἔξωθεν καὶ ἐπὶ συναγωγὴν νεανίσκων ἅμα
nd ich füllte meinen Grimm, aber ich hielt ihn zurück und bereitete
U
ihnen nicht das Ende. Ich werde (ihn) ausgießen über unmündige Kin-
der draußen und über eine Versammlung von Jugendlichen zugleich.
Laut MT bekennt in 11ab ein menschlicher Redner—laut Buchkontext
Jeremia—, von nicht mehr beherrschbarer Zornesglut Jhwhs erfüllt zu
sein. In 11c reagiert nach einem unmarkierten Sprecherwechsel die frag-
los göttliche Stimme mit dem Befehl (Imperativ), die Zornesglut über die
Judäer zu ergießen. Diese Rederollen setzen einen weiteren unmarkierten
Sprecherwechsel im Vortext voraus. Denn die Botenformel in V. 9 dekla-
riert das Folgende als Gotteswort, aber wenn in 11a Jeremia zu vernehmen
ist, hat er ohne explizites sprachliches Signal das Wort ergriffen, was am
ehesten mit Beginn von V. 10 geschieht. In AlT hingegen bildet V. 9–15 eine
geschlossene Gottesrede (die Selbstreferenz Jhwhs wie auf einen Dritten
nach Art von 10f ist gängig32 und daher kein Einwand). Folglich ist es dort
Jhwh, der laut 11ab meine Zornesglut (‫ ) ֲח ָמ ִתי‬nicht mehr zügeln will und
32 H.-J. Stipp, Elischa—Propheten—Gottesmänner. Die Kompositionsgeschichte des

Elischazyklus und verwandter Texte, rekonstruiert auf der Basis von Text- und Literarkritik
zu 1 Kön 20.22 und 2 Kön 2–7 (ATSAT 24; St. Ottilien: Eos, 1987), 120–130.
in 11c ihre Entfesselung (‫ ) ֶא ְשׁפְֹך‬ankündigt33. Die unbefriedigend ange

zeigten Sprecherwechsel in MT könnten die alexandrinischen Tradenten
veranlasst haben, den Passus zu vereinfachen, indem sie ihn durchgehend
der Stimme Jhwhs unterstellten. Dann müssten sie zu diesem Zweck das
Gottesbild allerdings deutlich getrübt haben, und man fragt sich, warum
sie hierzu die Anrede an den Propheten in 11c aufgaben, die ohnehin nur
aus dem Mund Jhwhs kommen kann. Daher überzeugt es mehr, die Modi-
fikation in der masoretischen Überlieferung zu suchen, wo man einen Teil
des Passus in den Mund Jeremias verlegte, um den Eindruck zu vermei-
den, Jhwh habe seine eigenen Emotionen nicht im Griff. Der Imperativ
‫ ְשׁפְֹך‬mit (hier kontextgetilgtem) Objekt ‫ ֵח ָמה‬kann zusätzlich durch die
Parallele in 10:25 angeregt worden sein.
Demnach erwuchsen die Devianzen zwar aus der theologischen Arbeit
am Gottesbild, jedoch im masoretischen Strang, und sie sollten auch
nicht Jhwh sanfter zeichnen, sondern seine Selbstkontrolle dem Zweifel
entziehen. Sollte dagegen der Wandel tatsächlich auf alexandrinischer
Seite eingetreten sein, hätte man dort zur Behebung eines Problems der
Textkohärenz den Charakter Jhwhs erheblich verdunkelt. Dies gälte in
geringerem Maß für den Übersetzer, gingen die Abweichungen auf ihn
zurück. Indem er die Verbform ‫ ָה ִכיל‬falsch analysierte, bewahrte er Jhwh
vor dem Verdacht mangelnder Herrschaft über sich selbst, doch hätte er
durch ἐκχεῶ den Vollzug des Unheils in Jhwhs eigene Hände verlegt.
9:9 ‫ל־ה ָה ִרים ֶא ָשּׂא \ ְשׂאוּ ְב ִכי ]וָ נֶ ִהי[ וְ ַעל־נְ אוֹת ִמ ְד ָבּר ִקינָ ה‬
ֶ ‫ַע‬
ἐπὶ τὰ ὄρη λάβετε κοπετὸν καὶ ἐπὶ τὰς τρίβους τῆς ἐρήμου θρῆνον
Der Satz ist Gottesrede, vorweg markiert durch die Botenformel V. 6, die
Gottesspruchformel V. 8 und ein Ich, das sich Handlungen zuschreibt,
die nur Gott offenstehen (V. 6, 8). Daher erhebt Jhwh in MT ein Weinen
und ein Klagelied, während er in AlT eine sonst im näheren Kontext nicht
genannte Hörerschaft zu diesen Riten aufruft. Dies kann den Anschein
erwecken, ein alexandrinischer Tradent oder der Übersetzer habe den
anthropomorphen Zug der offen zur Schau getragenen Trauer Gottes
beseitigen wollen34. Allerdings zitiert auch AlT Jhwh kurz zuvor in 8:23/
33 Die griechische Wiedergabe von 11b gibt kein Recht, eine abweichende Vorlage zu
erschließen, sondern erklärt sich aus dem irrigen Bezug von ‫ ָה ִכיל‬auf ‫כלה‬. Außerdem
bereitete dem Übersetzer das Verb ‫ לאה‬immer Schwierigkeiten; vgl. sonst 9:4; 12:5; 15:6;
20:9.
34 So für den Schöpfer von JerG* Fischer, „Diskussion“ (Anm. 6), 84.
9:1 mit Worten, in denen er sich offen zu seinen Tränen bekennt, und
in 8:21 erweitert sogar ein alexandrinischer Überhang die Gottesklage um
die massiv anthropomorphe Vorstellung, ‫יּוֹל ָדה‬ ֵ ‫( ִחיל ַכּ‬ὠδῖνες ὡς τικτού
σης) hätten Jhwh erfasst. Vom Weinen Gottes weiß AlT ferner in 48/
31:32.—Eine zu 9:9 analoge Variantenkonstellation begegnet in 14:17 (s.
u.). Diese Befunde sind mit 13:17 zu vergleichen, wo laut MT Jeremia das
Schicksal der Judäer beweint, während AlT die Klage ebenfalls den Judä-
ern zuschreibt: ‫י\כם ִמ ְפּנֵ י‬ ֶ ‫י<כם> ִדּ ְמ ָעה ְבּ ִמ ְס ָתּ ִרים ִתּ ְב ֶכּה־נַ ְפ ִשׁ‬
ֶ ִ‫וְ ֵת ַרד ֵעינ‬
]. . .[ ‫גֵ וָ ה‬. Wäre es den alexandrinischen Texttradenten darum gegangen,
Anthropomorphismen zu vermeiden, hätten sie in die Klage Jeremias
nicht einzugreifen brauchen. Nach 13:17 zu schließen, war AlT indes pri-
mär bestrebt, die Trauer über die Not der Judäer generell den Betroffenen
aufzubürden, während die Reduktion von Anthropomorphismen allen
falls ein nachrangiges Motiv lieferte. Vor allem sollten die Judäer ihre aus
eigener Schuld erwachsene Misere selbst beweinen. Die alexandrinischen
Lesarten folgen somit einheitlich dem Muster von 9:17, wo der Appell
‫ וְ ִת ֶשּׂנָ ה ָע ֵלינוּ נֶ ִהי‬an Menschen (Klagefrauen V. 16) ergeht35 und auch den
Judäern die Tränen fließen: ‫וְ ֵת ַר ְדנָ ה ֵעינֵ ינוּ ִדּ ְמ ָעה‬. Im gegebenen Fall sind
keine Merkmale zu erkennen, die ein Urteil erlauben, ob der Wandel im
Zuge der Übersetzung oder vielmehr schon in der hebräischen Phase der
alexandrinischen Textüberlieferung eingetreten ist.
10:16 ‫ וְ יִ ְשׂ ָר ֵאל ֵשׁ ֶבט[ נַ ֲח ָלתוֹ‬c] ‫י־יוֹצר ַהכֹּל הוּא‬

ֵ ‫ ִכּ‬b ‫א־כ ֵא ֶלּה ֵח ֶלק יַ ֲעקֹב‬
ְ ֹ‫ל‬a
51:19 ‫ וְ ֵשׁ ֶבט[ נַ ֲח ָלתוֹ‬c] ‫י־יוֹצר ַהכֹּל הוּא‬ֵ ‫ ִכּ‬b ‫א־כ ֵא ֶלּה ֵח ֶלק יַ ֲעקֹב‬
ְ ֹ‫ל‬a
Die Dublette kontrastiert den „Anteil Jakobs“ mit den nichtigen Götzen
bildern (10:14–16 || 51:17–19). Gegenüber den drei Sätzen der masoretischen
Ausgabe kennen die alexandrinischen Fassungen nur zwei. Dabei fährt
der zweite Satz jeweils folgerichtig fort, weil ‫ נַ ֲח ָלתוֹ‬dort parallel mit ‫ֵח ֶלק‬
‫ יַ ֲעקֹב‬gefügt ist: ‫ נַ ֲח ָלה‬präzisiert ‫ ֵח ֶלק‬, und das Enklitikon vertritt ‫יַ ֲעקֹב‬,
sodass in beiden Sätzen Jhwh als Israels An- bzw. Erbteil proklamiert
wird. In 10:16c MT hingegen ist umgekehrt Israel der Stamm von Jhwhs
Erbteil, wozu 51:19c MT eine verstümmelte Variante bietet. MT vertritt die
gängigere Konzeptionalisierung des Verhältnisses von Jhwh und Israel,
wie die näheren Parallelen in Jes 63:17 und Ps 74:2 sowie die entfernte-
ren in Dtn 32:9 und Ps 78:71 belegen. Zudem kommt ‫ ֵשׁ ֶבט‬in Jer nur hier
vor und hat damit idiolektalen Status. Folglich hat eine prämasoretische
35 Von dort dürfte MT seinerseits ‫ וָ נֶ ִהי‬entlehnt haben.

Hand die beiden Verse einer Orthodoxie unterworfen, der an der Domi-
nanz Jhwhs im Verhältnis zu seinem Volk gelegen war36.
10:18 ‫ת־יוֹשׁ ֵבי ָה ָא ֶרץ ] ַבּ ַפּ ַעם[ ַהזֹּאת‬

ְ ‫( ֶא‬Gram!) ‫קוֹל ַע‬
ֵ \ ‫צוֹל ַע‬
ֵ ‫ ִהנְ נִ י‬b
# ‫רוֹתי ָל ֶהם \ ַבּ ָצּ ָרה‬
ִ ‫ § וַ ֲה ֵצ‬c
ἰδοὺ ἐγὼ σκελίζω τοὺς κατοικοῦντας τὴν γῆν ταύτην ἐν θλίψει
Siehe ich, ich lasse die Bewohner dieses Landes stolpern37 in Drangsal.
Für 18c liest AlT nur in Drangsal, während MT Jhwh nochmals direkt
involviert mit und ich werde sie bedrängen. Vorweg in 18b kündigt Jhwh
ein massiv anthropomorphes Handeln an, das der Übersetzer unver
wässert wiedergab (wobei das Partizip in der Vorlage von ‫קוֹל ַע‬ֵ zu ‫צוֹל ַע‬
ֵ
verschrieben war, dem er wider die Grammatik eine transitive Bedeutung
unterschob). Daher ist nicht einzusehen, weswegen allein 18c entschärft
worden sein sollte38. Mehr Wahrscheinlichkeit hat für sich, dass in MT die
Beteiligung Jhwhs nachträglich gesteigert worden ist.
11:1; 18:1; 21:1; 32:1; 40:1 # ‫ַה ָדּ ָבר ֲא ֶשׁר ָהיָ ה § ֶאל־יִ ְר ְמיָ הוּ * ֵמ ֵאת יְ הוָ ה‬
Dies ist die konstante Abfolge der Relativsatz-Variante der Wortereig
nisformel in MT 39, während AlT in den genannten Fällen die beiden Prä
positionalverbindungen in umgekehrter Reihenfolge bietet. Der Befund
lässt zwei Deutungen zu, zwischen denen nicht leicht zu entscheiden ist:
Entweder hat man in AlT den Wortlaut mehrfach revidiert, um den Vor
rang Jhwhs vor dem Propheten auch sprachlich zu repräsentieren40, oder
sämtliche Belege wurden in MT nach diesem Muster vereinheitlicht.
36 Vonach, „Jeremias“ (Anm. 11), 2724, 2756, vertritt den umgekehrten Entwicklungsweg,
lässt aber alle genannten Gesichtspunkte außer Acht. Der hebräische Wortlaut von 10:16
bietet überdies keinerlei „komplexe und nur schwer verständliche Formulierung“ (2756),
die den Übersetzer zu Vereinfachungen hätte veranlassen können.
37 Fischer und Vonach (Anm. 29) übersetzen σκελίζω mit wegschleudern. Der Fehler ist
korrigiert im Kommentar z. St. bei Vonach, „Jeremias“ (Anm. 11), 2756, entsprechend H. G.
Liddell und R. Scott, A Greek-English Lexicon, Revised and Augmented throughout by H. St.
Jones (Oxford: Clarendon Press, 1940), s. v.: trip up one’s heels, upset; J. Lust, E. Eynikel und
K. Hauspie, A Greek-English Lexicon of the Septuagint, Revised Edition (Stuttgart: Deutsche
Bibelgesellschaft, 2003), s. v.: to overthrow, to upset.
38 Gegen Vonach, „Jeremias“ (Anm. 11), 2726.
39 Ferner 7:1 > AlT; 30:1; 34:1, 8; 35:1.
40 So noch entschieden Stipp, Sondergut (Anm. 33), 152f.
11:18 ‫יתי‬
ִ ‫יתנִ י \ ָר ִא‬
ַ ‫ ָאז ִה ְר ִא‬c ‫ וָ ֵא ָד ָעה \ וְ ֵא ָד ָעה‬b ‫י\ענִ י‬
ֵ ִ‫יענ‬
ַ ‫הוֹד‬
ִ ‫ [וַ ]יהוָ ה‬a
‫יהם‬
ֶ ‫ַמ ַע ְל ֵל‬
κύριε γνώρισόν μοι καὶ γνώσομαι τότε εἶδον τὰ ἐπιτηδεύματα αὐτῶν
Laut MT berichtet Jeremia in 18ab von einem Erkenntnisprozess, den
Jhwh bei ihm bewirkte (18a ‫יענִ י‬ ַ ‫הוֹד‬
ִ er ließ mich erkennen), wobei aus
drücklich der Erfolg konstatiert wird (18b ‫ וָ ֵא ָד ָעה‬und ich erkannte: Nar
rativ der 1. Person mit Endung -a41). Der Gegenstand der Einsicht wird erst
in 18c offengelegt, wo sich der Prophet direkt an Jhwh wendet: Da hast
du mir ihre Untaten gezeigt. Redet somit in MT der dritte Satz Jhwh an,
ist es in AlT der erste: 18a, der dort einen Appell an Jhwh richtet, indem
die Verbalform als Imperativ interpretiert wird (‫יענִ י‬ ֵ ‫הוֹד‬
ִ lass mich erken-
nen) und der asyndetische Gottesname als Anrede fungiert. 18b benennt
als Finalsatz mit syndetischem Kohortativ (wֺ=’iqtul‑a: ‫ וְ ֵא ָד ָעה‬damit ich
erkenne)42 die erhoffte Folge der erbetenen Erkenntnis, während 18c
die Erfüllung wie MT in die Vergangenheit verlegt, aber keine Mitwir-
kung Jhwhs thematisiert: Da sah ich ihre Untaten. Die präteritale Deixis
steht im Einklang mit der Fortsetzung V. 19–20, der zufolge Jeremia die
Machenschaften seiner Gegner ebenfalls schon kennt. In beiden Lesarten
ist das Verhältnis von 18ab zu c problematisch: in MT wegen des Wech
sels der Rederichtung, in AlT wegen der temporalen Inkonsistenz. Einen
glatten Text ergibt nur eine Mischung beider Fassungen: ‫יענִ י‬ ַ ‫הוֹד‬
ִ ‫יְ הוָ ה‬
‫יהם‬ֶ ‫יתי ַמ ַע ְל ֵל‬
ִ ‫ וָ ֵא ָד ָעה ָאז ָר ִא‬Jhwh ließ mich erkennen und ich erkannte;
da sah ich ihre Untaten. Dies dürfte den originalen Wortlaut darstellen,
den die alexandrinische Tradition bzw. der Übersetzer in 18ab über eine
Revokalisierung umdeutete, wahrscheinlich angestoßen durch die relativ
seltene Formation wa=’iqtul‑a in 18b und ermöglicht durch die Asyndese
des Gottesnamens in 18a. Die Modifikation auf alexandrinischer Seite
war also anscheinend vor allem grammatischer Natur, indem man mehr-
deutige Verbalformen nach einem geläufigeren Muster reinterpretierte.
Dagegen griff die masoretische Überlieferung tiefer in den Textbestand
ein: Sie hob in 18c die Urheberschaft Jhwhs hervor, indem sie dem Verb
eine kausative Form verlieh, es dem Subjekt Jhwh zuordnete und so das
41 Vgl. GKC 134 (§ 49e); P. Joüon und T. Muraoka, A Grammar of Biblical Hebrew. Second
Reprint of the Second Edition, with Corrections (SubBi 27, Roma: Gregorian & Biblical Press,
2009), 129 (§ 47d).
42 Vgl. E. Kuhr, Die Ausdrucksmittel der konjunktionslosen Hypotaxe in der ältesten
hebräischen Prosa. Ein Beitrag zur historischen Syntax des Hebräischen (Beiträge zur semi-
tischen Philologie und Linguistik 7; Leipzig: Hinrichs, 1929) 49–52; Joüon und Muraoka,
Grammar (Anm. 41) 596 (§ 168b).
Sehen Jeremias in das Zeigen Jhwhs verwandelte. Auch hier fand somit
Arbeit am Gotteskonzept statt, allerdings im masoretischen Strang, wo
man die Souveränität Gottes über das Geschehen erweiterte. Ein gleich
artiger Vorgang mit Wechsel von G- zu H-Stamm und von menschlichem
zu göttlichem Subjekt hat in 18:2 und vielleicht auch 19:9 stattgefunden.
Anders zu bewerten ist wohl 15:17 (s. u.).
13:12 ‫ ָכּל־נֵ ֶבל יִ ָמּ ֵלא יָ יִ ן‬c [‫ֹלהי יִ ְשׂ ָר ֵאל‬

ֵ ‫ֹה־א ַמר יְ הוָ ה ֱא‬
ָ ‫ ]כּ‬b
Die Botenformel 12b fehlt in AlT. Die Meinung, sie sei dort wegen des
folgenden Trinkspruchs 12c gestrichen worden43, ist allerdings an der Tat-
sache zu messen, dass man die Phrase häufig in MT ergänzt hat: 11:22;
18:11; 22:30; 27:21; 29:25; 31:37; 35:19, wobei in 27:21; 29:25 und 35:19 überdies
idiolektale Gottesepitheta den sekundären Charakter der Belege unter
mauern44. Danach erfolgte der Eingriff eher auf masoretischer Seite, wo er
jedoch keinen spezifischen konzeptionellen Intentionen entsprang, son
dern zur routinemäßigen Expansion geprägter Elemente gehörte, wie sie
für diese Textform kennzeichnend ist.
14:17 ‫( ֵעינַ י> ֶכם< ִדּ ְמ ָעה‬2. Pl f, Gram!) ‫ ֵתּ ַר ְדנָ ה‬b

καταγάγετε ἐπ᾿ ὀφθαλμοὺς ὑμῶν δάκρυα
Der Passus eröffnet laut dem Redebefehl 17a ‫ת־ה ָדּ ָבר ַהזֶּ ה‬ ַ ‫יהם ֶא‬ ֶ ‫וְ ָא ַמ ְר ָתּ ֲא ֵל‬
und dem Enklitikon in ‫ת־ע ִמּי‬ ַ ‫ ַבּ‬17d eine Gottesrede. Folglich spricht MT
wie in 9:9 (s. o.) die Vorstellung aus, dass Jhwh weint: Meine Augen fließen
über von Tränen. G* dagegen bildet das Verb hier wie dort als Imperativ
Plural. Die Form καταγάγετε deutet die Präfixkonjugation als 2. Ps Pl f in
Funktion Aufforderung, und zwar irrigerweise45, da 17a auf die Adressaten
im maskulinen Plural verweist (‫יהם‬ ֶ ‫ ) ֲא ֵל‬und somit ‫יכם‬
ֶ ֵ‫ ֵעינ‬mit maskulinem
Pronomen zu rekonstruieren ist; 17d wiederum imaginiert dieselbe Größe
singularisch als ‫ת־ע ִמּי‬
ַ ‫ ַבּ‬. Überdies bekräftigt bei der parallelen Verbalform
‫ ִתּ ְד ֶמינָ ה‬17c auch G* die Interpretation als 3. Person. Demnach gebührt
dem Bezug auf die 3. Ps Pl f die Priorität. Hier wie in 9:9 scheint sich der
Schluss nahezulegen, man habe in AlT die Vorstellung vom Weinen Got
tes umgehen wollen46. Doch wie dort bereits dargelegt, gab wahrschein
43 So Fischer, „Diskussion“ (Anm. 6), 84.

44 Vgl. ferner 17:5 sowie als Bestandteil größerer Einschübe 29:16, 17; 33:17, 20, 25.
45 Außerdem wird ‫‑ירד‬G ein kausativer Sinn beigelegt. Zur Äquivalenz ‫‑ירד‬G ≙ κατ
άγω vgl. ferner 9:17c; 13:17d.
46 Vgl. Anm. 34; Vonach, „Jeremias“ (Anm. 11), 2762.
lich weniger die Aversion gegen Anthropomorphismen den Ausschlag als

vielmehr der Wunsch, die Judäer ihre Notlage selbst beweinen zu lassen
(s. zu 9:9). Und wie dort bleibt unentscheidbar, ob die theologischen Kor-
rekturen schon in der hebräischen Phase der alexandrinischen Texttradi
tion eintraten oder erst vom Übersetzer vorgenommen wurden.
15:7 (ἠτεκνώθην) ‫ ִשׁ ַכּ ְל ִתּי \ ֻשׁ ַכּ ְל ִתּי‬b

Der Satz fällt innerhalb einer Gottesrede, in der Jhwh auf vergangenes
Strafhandeln an seinem Volk zurückschaut (6c–7c). Die Retrospektive
steht in TT durchgehend im Aktiv, während G* in 7b eine passivische
Vokalisation voraussetzt. Diese Deutung des Konsonantentextes ist in
einem einzelnen Punkt weniger hart, doch kann angesichts der Aussagen
im Kontext von einer Milderung des Gottesbildes keine Rede sein. Zugleich
ist die G*-Lesart sogar anthropomorpher47. Wie immer sie sich zu TT ver-
halten mag, belegt sie jedenfalls keine Scheu von JerAlT oder JerG*, die
Gottesvorstellung mit anthropomorphen Metaphern zu artikulieren.
15:17 ‫אתי‬ִ ‫אתנִ י \ ָמ ֵל‬

ָ ‫ִכּי־זַ ַעם ִמ ֵלּ‬
Innerhalb der Konfession 15:10–21 schildert Jeremia mit 15:17d seinen
bedauernswerten Zustand. Dabei erhebt er in MT einen scharfen Vorwurf
an die Adresse Jhwhs, denn er macht ihn für seine Gemütsverfassung
verantwortlich, während AlT diese nur konstatiert. MT kann damit die
göttliche Lenkung des Geschehens herausstellen, aber AlT kann auch
die Anklage eliminiert haben, die Jhwh einen harten Umgang mit dem
Propheten vorhält. Angesichts der geringfügigen Differenz im Schrift
bild kommt überdies ein bloßes Schreiberversehen—in beiden Rich
tungen—in Betracht. Der analoge Befund in 1:17 spricht allerdings für eine
Modifikation in AlT. Vgl. auch den folgenden Fall 15:18.
15:18 ‫( ִלי ְכּמוֹ ַא ְכזָ ב ַמיִ ם‬Aor 3. f ) ‫ ָהיוֹ ִת ְהיֶ ה‬. . . ‫נוּשׁה‬
ָ ‫וּמ ָכּ ִתי ֲא‬
ַ
In dem Klagegebet Jeremias ist die Verbalgruppe ‫ ָהיוֹ ִת ְהיֶ ה‬auf der sprach-
lichen Oberfläche doppeldeutig, da sie einerseits für die 2. Ps Sg m eintre-
ten kann und somit Jhwh in einer besonders scharfen Anklage mit einem
Trugbach vergleicht. Andererseits kann die Verbalgruppe die 3. Ps Sg f
bezeichnen; so würde sie allerdings auf ‫ ַמ ָכּ ִתי‬als ihr Subjekt verweisen
47 Die Variante ἠτεκνώθησαν hat Ziegler, Jeremias (Anm. 29), dagegen zurecht als Glät-
tung in den Apparat verwiesen; gegen Vonach, „Jeremias“ (Anm. 11), 2763.
und meine Wunde höchst implausibel mit einen Trugbach gleichsetzen48.

G* wählt trotzdem die zweite Möglichkeit, sodass der direkt an Jhwh
gerichtete Vorwurf entfällt (vgl. auch oben zu 1:17; 15:17)49. Unentscheid-
bar bleibt, ob der Übersetzer selbst damit den Text entschärfte oder bloß
eine Leseweise applizierte, die in jenen jüdischen Kreisen üblich war, in
denen er die Traditionen seines Glaubens pflegte. Denn schließlich sind
die Väter der Septuaginta ihrer Vorlage kaum erst im Zuge der Überset
zung begegnet; weit mehr Wahrscheinlichkeit hat es für sich, dass sie ihre
heiligen Schriften schon vorweg in ihrer heimischen Gemeinde studiert
und herkömmliche Leseweisen des mehrdeutigen Konsonantentextes
kennen gelernt hatten50.
16:13 ‫יִתּנוּ‬
ְ \ ‫א־א ֵתּן‬
ֶ ֹ ‫[יוֹמם וָ ַל ָיְלה] ֲא ֶשׁר ל‬
ָ ‫ֹלהים ֲא ֵח ִרים‬
ִ ‫ת־א‬
ֱ ‫ם־שׁם ֶא‬
ָ ‫וַ ֲע ַב ְד ֶתּ‬
‫ָל ֶכם ֲחנִ ינָ ה‬
Weil dem ‫ ֲא ֶשׁר‬-Satz in MT ein Beziehungswort fehlt, gehört die Kon
junktion dort zu den wenigen Fällen, wo sie eine kausale Funktion aus
übt51. AlT dagegen liest ein verbales Prädikat in der 3. Ps Pl und somit
einen Relativsatz, der auf ‫ֹלהים ֲא ֵח ִרים‬
ִ ‫ ֱא‬bezogen ist. Der geringe Auf
wand, mit dem sich der ‫ ֲא ֶשׁר‬-Satz der üblichen relativen Funktion zuord-
nen lässt, spricht gegen eine sekundäre Glättung. Vielmehr dürfte AlT
den originalen Stand repräsentieren, während man in der masoretischen
Tradition der Vorstellung entgegentrat, dass Fremdgötter den judäischen
Exilanten Erbarmen gewähren oder verweigern könnten bzw. überhaupt
aktionsfähig wären52. Folglich erwuchs die masoretische Lesart aus einer
theologischen Korrektur mit dem Ziel, die Souveränität Jhwhs zu sichern,
selbst um den Preis eines grausameren Gottesbildes.
48 Der zitierte Passus enthält weitere Divergenzen, die sich aber nur auf Formulie
rungsvarianten beziehen und daher hier ausgeklammert werden.
49 Vgl. Vonach, „Jeremias“ (Anm. 11), 2725, 2764.
50 Vgl. z. B. Siegert, Zwischen Hebräischer Bibel und Altem Testament (Anm. 17), 123–125;
H.-J. Stipp, „Bemerkungen zum griechischen Michabuch aus Anlass des deutschen LXX-
Übersetzungsprojekts,“ JNSL 29 (2003): 103–132 (120–122); G. D. Martin, Multiple Originals.
New Approaches to Hebrew Bible Textual Criticism (SBL Text-Critical Studies 7; Leiden: Brill,
2011), 82f.
51 HAL 95b; vgl. Ges18 112a.
52 So z. B. W. L. Holladay, Jeremiah 1. A Commentary on the Book of the Prophet Jeremiah
Chapters 1–25 (Hermeneia; Philadelphia: Fortress Press, 1986), 474. Nach Vonach, „Jere-
mias“ (Anm. 11), 2765, hingegen „nimmt die LXX der Darstellung des Gottes Israels einiges
an Schärfe.“ Dann müsste der Übersetzer die polytheistischen Implikationen seiner Wie-
dergabe freiwillig in Kauf genommen haben.
16:15 ‫וּמכֹּל ָה ֲא ָרצוֹת‬

ִ ‫ת־בּנֵ י יִ ְשׂ ָר ֵאל ֵמ ֶא ֶרץ ָצפוֹן‬
ְ ‫ַחי־יְ הוָ ה ֲא ֶשׁר ֶה ֱע ָלה ֶא‬
‫( ָשׁ ָמּה‬ἐξώσθησαν) ‫יחם‬ ָ ‫ֲא ֶשׁר ִה ִדּ‬
Laut MT wird Jhwh die Israeliten aus aller Welt versammeln, wohin er
sie zuvor versprengt hat. Der kausativen Verbalform ‫יחם‬ ָ ‫ ִה ִדּ‬korrespon
diert die passivische Wiedergabe ἐξώσθησαν, die den Akteur verschweigt.
Weil JerG* die Genera verbi großzügig behandelt, ist die Entscheidung
erschwert, ob das Passiv schon der Vorlage eignete (‫ )נִ ְדּחוּ‬oder auf den
Übersetzer zurückgeht, der Jhwh absichtlich oder bloß faktisch von der
Verantwortung für die Diaspora entband. Daher sind die Konsequenzen
mehrerer Alternativen zu bedenken.
Wann immer das Passiv in den Text geriet, ist es jedenfalls kaum aus
einer theologisch motivierten Glättung erwachsen. 16:15 gehört in eine
Reihe mit 12 weiteren geprägten Wendungen in Jer, die ein topisches Por-
trät der Diasporasituation entwerfen, indem sie das Leben in der Fremde
durch die Verben ‫ נדח‬und ‫ פוץ‬auf Zerstreuung zurückführen und deren
Reichweite mittels der Stichworte ‫ ֲא ָרצוֹת‬, ‫ גּוֹיִ ם‬und ‫ ְמקֹמוֹת‬weltweit ent-
grenzen53. Diese Passagen folgen weit überwiegend dem Muster von 16:15
MT, insofern sie ihre verbalen Prädikate als H-Stamm formen und Jhwh
die Subjektsrolle zuweisen. Dies gilt für 7 Fälle in dem Bestand, den MT
mit AlT gemeinsam hat; dort bekennt sich auch JerG* zur Urheberschaft
Jhwhs und gibt somit keinen Anlass zur Frage nach abweichenden Vorla-
gen oder Eingriffen des Übersetzers54. Hinzu kommen 3 weitere Belege in
masoretischen Überschüssen, die, wie idiolektale Komponenten erweisen,
nicht in AlT ausgeschieden wurden, sondern erst nach der Gabelung der
Texttraditionen in MT eingedrungen sind55. Daneben existieren nur zwei
passivische Fälle, die ebenfalls dem masoretischen Sondergut angehören,
eine Art der Bezeugung, die sie selbst zu einem Merkmal des prämaso-
retischen Idiolekts erhebt (40:12; 43:5)56. Akteursblinde Varianten der
fraglichen geprägten Wendung treten also erst in den jüngsten Zutaten
zur masoretischen Ausgabe des Buches auf, während die alexandrinische
Edition gerade die passivischen Formulierungen vermissen lässt, die theo-
logisch die geringsten Probleme bereitet haben müssten.
53 Vgl. H.-J. Stipp, Deuterojeremianische Konkordanz (ATSAT 63; St. Ottilien: Eos, 1998),
88f.
54 8:3; 9:15; 23:3, 8; 24:9; 32:37; 46:28 (|| 30:11 MT). In 23:8 MT wird Jhwh sekundär in 1.
Ps repräsentiert.
55 29:14, 18; 30:11 (|| 46:28). Die idiolektalen Züge sind verzeichnet bei Stipp, „Updated
Index“ (Anm. 4), 192f.
56 Vgl. Stipp, „Updated Index“ (Anm. 4), 193f. Sollte in 16:15 das Passiv in der Vorlage
wurzeln, heben sich 40:12; 43:5 immerhin durch die Form ‫ ָשׁם‬statt ‫ ָשׁ ָמּה‬ab.
Das Bild wird bekräftigt durch nicht-formelhaftes Vergleichsmaterial,

das geeignete Verben mit dem Subjekt Jhwh und dem Objekt Israel bzw.
Juda verbindet. G* gibt auch solche Passagen ungeschönt wieder: Vgl. mit
‫‑פוץ‬H 13:24; 18:17; mit ‫ זרה‬15:7; 31:10; mit ‫‑גלה‬H 22:12 (wo sogar allein AlT
die Deportation in die Hände Jhwhs verlegt, weswegen der Passus weiter
unten zu erörtern ist); 29:4, 7. Der Beleg mit ‫‑גלה‬H in 29:14 fällt in einen
masoretischen Überschuss, den idiolektale Elemente als späte Interpola-
tion markieren57. Näherer Analyse bedarf lediglich die Sachlage bei ‫‑נדח‬H
in 27:15, wo AlT einzig den pronominalen Bezug auf Jhwh übergeht; siehe
dazu unten.
So ergibt sich: Bei 16:15 sind keine Indizien zu erkennen, die ein Urteil
gestatten, ob das Passiv bereits in der JerG*-Vorlage stand oder erst vom
Übersetzer geschaffen wurde. Jedenfalls zeigt sich die alexandrinische
Tradition weder in ihrer hebräischen noch in ihrer griechischen Pha
se geneigt, Jhwhs Regie hinter der Entwurzelung der Israeliten zu ver
schleiern. Der masoretische Strang hat erst auf seiner jüngsten Stufe zwei
akteursblinde Formulierungen hervorgebracht, allerdings neben weiteren
Belegen herkömmlichen Typs, die wie gehabt Jhwh als Verursacher iden-
tifizieren. Sofern sich die diskutierte Lesartendifferenz in 16:15 überhaupt
mit in den Textformen wirksamen Trends verbinden lässt, ergeben sich
zwei Möglichkeiten, je nach dem, wo die passive Variante erstmals auftrat.
War sie schon in der JerG*-Vorlage enthalten, bildete sie dort kaum einen
sekundären Zug, sondern darf eher die Priorität beanspruchen, während
der masoretische Kausativ nachträglich aus der Adaption an einen domi-
nanten Sprachgebrauch erwuchs und/oder dem wiederholt beobachtba-
ren Bedürfnis nachgab, die Steuerung der Geschichte ausdrücklich in die
Hände Jhwhs zu verlegen. Hat erst der Übersetzer das Passiv eingeführt,
wollte er ebensowenig Jhwh ein freundlicheres Gesicht verleihen, son
dern er pendelte spontan zwischen den Genera verbi, wie er es häufig tat,
ohne dass er dazu spezieller Gründe bedurfte. Welche Alternative auch
immer den Vorzug verdient, gibt der Befund jedenfalls kein Recht, der
alexandrinischen Texttradition bzw. dem Übersetzer einen Hang zur Mil-
derung der Gottesbilder zu bescheinigen. Dafür gibt es zu viele Gegenbei-
spiele58. Für den Übersetzer ließe sich dieses Fazit durch weitere Beispiele
aus JE12 erhärten59.
57 S. Anm. 55.

58 Gegen Vonach, „Jeremias“ (Anm. 11), 2765.
59 Vgl. namentlich die bei Stipp, Konkordanz (Anm. 53), 88f., gesammelten Belege aus
Ez.
18:2 ‫ת־דּ ָב ָרי‬
ְ ‫יעָך \ ִת ְשׁ ַמע ֶא‬
ֲ ‫יּוֹצר וְ ָשׁ ָמּה ַא ְשׁ ִמ‬
ֵ ‫קוּם וְ יָ ַר ְד ָתּ ֵבּית ַה‬
Anstelle des Kausativstamms in MT setzt AlT den Grundstamm voraus
(ἀκούσῃ). Plausibel ist allein der Schritt von der alexandrinischen zur
masoretischen Lesart, die das Bestreben verrät, die Kontrolle Jhwhs über
das Geschehen herauszustreichen. Nach gleichem Muster und mit iden
tischer Absicht hat man in MT 11:18 sowie vielleicht auch 19:9 Stämme und
Subjekte ausgetauscht.
19:9 ‫יהם‬ ֶ ‫יהם וְ ֵאת ְבּ ַשׂר ְבּנ ֵֹת‬ ֶ ֵ‫ת־בּ ַשׂר ְבּנ‬
ְ ‫ וְ ַה ֲא ַכ ְל ִתּים \ וְ ָא ְכלוּ ֶא‬a
‫יהם‬
ֶ ‫ ֲא ֶשׁר ִיָציקוּ ָל ֶהם אֹיְ ֵב‬c ‫וּב ָמצוֹק‬ְ ‫אכלוּ ְבּ ָמצוֹר‬ֵ ֹ ‫ר־ר ֵעהוּ י‬ ֵ ‫ וְ ִאישׁ ְבּ ַשׂ‬b
Dies ist eine typische Konstellation, die die Frage nach sekundärer Auf-
hellung des Gottesbildes in der alexandrinischen Überlieferung anregt:
Hat man dort die grauenvolle Vorstellung ausgelöscht, dass Jhwh wün-
schen könne, die Angehörigen seines Volkes zum Kannibalismus an ihren
eigenen Kindern zu nötigen (‫אכל‬-H)60? Allerdings bezeugen in 9b beide
Textformen übereinstimmend den Grundstamm von ‫אכל‬, und die alex
andrinische Variante von 9a entspricht der Parallele Dtn 28:53 ‫וְ ָא ַכ ְל ָּת‬
‫ּוב ָמֹצוק‬ְ ‫ֹלהיָך ְּב ָמֹצור‬ ֶ ‫ן־לָך יְ הוָ ה ֱא‬
ְ ‫ּובנ ֶֹתיָך ֲא ֶׁשר נָ ַת‬
ְ ‫י־ב ְטנְ ָך ְּב ַׂשר ָּבנֶ יָך‬
ִ ‫ְפ ִר‬
‫ ; ֲא ֶׁשר־יָ ִציק ְלָך אֹיְ ֶבָך‬vgl. ferner die ebenfalls nicht-kausativen Belege Lev
26:29 und Bar* (hebräisch) 2:361. Der Befund lässt daher entgegengesetzte
Deutungen zu: Entweder hat man 19a in AlT theologisch geglättet und/
oder an 19b samt der Parallelen angeglichen, oder der deuteronomistische
Autor von 19:1–1362 benutzte von vornherein die geprägte Wendung mit
‫‑אכל‬G, die in der masoretischen Tradition im Interesse eines Machter
weises Jhwhs kausativ abgewandelt wurde. Weil MT gehäuft im Interesse
der Souveränität Jhwhs modifiziert und dazu wiederholt der G-Stamm mit
menschlichem Subjekt gegen den H-Stamm mit göttlichem Verursacher
ausgewechselt wurde (11:18; 18:2), hat der Verdacht der alexandrinischen
Priorität gute Gründe auf seiner Seite, ohne dass die wünschenswerte
Sicherheit zu gewinnen wäre. Jedenfalls ist der Befund nicht ohne die
genannten Parallelen sachgerecht zu beurteilen63.
60 So G. Fischer und A. Vonach, „Tendencies in the LXX Version of Jeremiah,“ in

G. Fischer, Der Prophet wie Mose (Anm. 6), 64–72 (69).
61 Diese und entferntere Parallelen sind zusammengestellt bei Stipp, Konkordanz
(Anm. 53), 16.
62 Vgl. zur Verfasserfrage H.-J. Stipp, „Jeremia und der Priester Paschhur ben Immer.
Eine redaktionsgeschichtliche Studie,“ in Kulte, Priester, Rituale—Beiträge zu Kult und
Kultkritik im Alten Testament und Alten Orient (FS Th. Seidl; Hg. St. Ernst und M. Häusl;
ATSAT 89; St. Ottilien: Eos, 2010), 375–401.
63 Gegen Fischer und Vonach (Anm. 60).
20:16 <‫ר־ה ַפְך יְ הוָ ה > ְבּ ַאף‬

ָ ‫וְ ָהיָ ה \ יְ ִהי ָה ִאישׁ ַההוּא ֶכּ ָע ִרים ֲא ֶשׁ‬
Die wohl von Dtn 29:22 (weniger wahrscheinlich: Ijob 9:5) inspirierte
Glosse belegt, dass man auch in der alexandrinischen Tradition bereit
war, Jhwh zusätzlich mit negativen Leidenschaften zu assoziieren.
21:7 ‫ת־ע ָב ָדיו וְ ֶאת־‬

ֲ ‫הוּדה וְ ֶא‬
ָ ְ‫ת־צ ְד ִקיָּ הוּ ֶמ ֶלְך־י‬
ִ ‫י־כן נְ ֻאם־יְ הוָ ה ֶא ֵתּן ֶא‬
ֵ ‫ וְ ַא ֲח ֵר‬a
. . . ‫ָה ָעם‬
‫וּביַ ד[ ְמ ַב ְק ֵשׁי נַ ְפ ָשׁם‬
ְ ] ‫יהם‬ֶ ‫ְך־בּ ֶבל וּ[ ְביַ ד אֹיְ ֵב‬
ָ ‫אצּר ֶמ ֶל‬ ַ ‫בוּכ ְד ֶר‬
ַ ְ‫] ְבּיַ ד נ‬
‫יהם‬ ֶ ‫ לֹא־יָ חוּס \ ָאחוּס ֲע ֵל‬c ‫י־ח ֶרב‬ ֶ ‫ וְ ִה ָכּם \ וְ ִה ֻכּם ְל ִפ‬b
‫ וְ לֹא יְ ַר ֵחם \ ֲא ַר ֵחם‬e [‫ ]וְ לֹא יַ ְחמֹל‬d
In AlT 7a kündigt Jhwh an, König Zidkija und seine Untertanen ihren
Feinden auszuliefern, die zuvor in V. 2, 4 als die babylonischen Belagerer
Jerusalems identifiziert wurden. Anschließend sollen die Sieger laut 7b ein
Blutbad anrichten, das Jhwh in 7ce als Ausdruck seiner Gnadenlosigkeit
deutet (1. Ps Sg). In MT sind die Gegner in einer Reihe von Nominal
gruppen ausdifferenziert, angeführt vom babylonischen König, der zwar
in beiden Fassungen der Einheit 21:1–10 erwähnt wird (V. 2), aber nur in
der masoretischen Ausgabe einen Namen trägt (ferner V. 2 MT)64. Das
Massaker gilt in 7b TT als Tat Nebukadnezzars und in 7c–e MT als Folge
von dessen Mangel an Erbarmen (3. Ps Sg). Dabei bezeugt MT mit 7d die
zusätzliche Aussage ‫וְ לֹא יַ ְחמֹל‬, im Einklang mit der Parallele 13:14, wo die
Trias kein Mitleid haben—nicht schonen—sich nicht erbarmen allerdings
wie in 21:7 AlT Eigenschaften Jhwhs aufreiht. Der Wandel von MT zu AlT,
mithin die Umwidmung der Passage von Nebukadnezzar auf die Baby-
lonier (7b) und Jhwh (7ce) ist unwahrscheinlich, denn dann wäre die
Streichung von Namen und Titel aus 7a kaum zu motivieren (die nicht
jünger sein können als die 3. Person in 7b–e, der sonst die Bezugsgröße
fehlt). Folglich wurde 21:7b–d umgekehrt auf Nebukadnezzar umdirigiert,
möglicherweise gefördert durch die Defektivschreibung von ‫וְ ִה ֻכּם‬65. Als
Konsequenz ging die Mitleidlosigkeit von Jhwh auf das menschliche
Strafwerkzeug über. Diese Annahme passt zu 13:14 und dem Umstand,
dass die masoretischen Überschüsse an Namen und Titeln den babyloni
schen Großkönig ohnehin besonders freizügig bedenken66, darunter nicht
selten an Stellen, die sich aus anderen Gründen als sekundär erweisen.
64 In V. 4 erwähnt ihn MT zusätzlich mit seinem Titel.

65 Theoretisch kann der Vorgang mehrstufig verlaufen sein: Erst wurde 7b singularisch
gelesen, dann hat man 7c(d)e auf die 3. Ps umgestellt.
66 Vgl. die Zusammenstellung bei Janzen, Studies in the Text of Jeremiah (Anm. 2),
139–141.
Demnach wurde hier zwar das Gotteskonzept abgemildert, aber erstens

nicht im alexandrinischen, sondern im masoretischen Strang. Zweitens
hat der Eingriff allem Anschein nach gar nicht die Aufhellung des Gottes
bilds bezweckt, denn dafür hätte die Umwidmung auf die babylonischen
Widersacher ausgereicht, indem man 7c–e in die 3. Person Plural ver-
setzte. Vielmehr ging es dem Revisor vor allem um die Einführung Nebu-
kadnezzars, mit dem sanfteren Profil Jhwhs als Nebeneffekt. Wer dagegen
MT die Priorität zuspricht, muss stattdessen eine Verdunkelung des Got-
tesbildes in AlT annehmen67.
22:12 ‫ר־הגְ לוּ \ ִהגְ ֵל ִתי אֹתוֹ ָשׁם יָ מוּת‬

ִ ‫<אם> ִבּ ְמקוֹם ֲא ֶשׁ‬
ִ ‫ִכּי‬
Laut MT wurde Joahas bzw. Schallum von Menschen unwiderruflich
deportiert, während dies nach AlT ein Werk Jhwhs gewesen ist. Die
Prioritätsfrage ist schwer entscheidbar, doch in Frage kommen nur eine
Aufhellung des Gottesbildes in der masoretischen Tradition oder eine
Verschärfung im alexandrinischen Strang, um hinter dem Vorgang die
Fügung Gottes aufzuweisen. Auch ein schlichter Schreibfehler ist nicht
auszuschließen, der in beiden Richtungen eingetreten sein kann: entwe
der durch Ausfall des ‫ ת‬mit anschließender Korrektur ‫ ו ⇒ י‬in MT oder
umgekehrt durch Verschreibung ‫ י ⇒ ו‬und Ergänzung des ‫ת‬. Die Hypo
these von Glättungstendenzen in JerG* findet jedenfalls erneut keine
Handhabe.
23:23 ‫ֹלהי ֵמ ָרחֹק‬
ֵ ‫ֹלהי ִמ ָקּר ֹב ָאנִ י נְ ֻאם־יְ הוָ ה וְ לֹא ֱא‬
ֵ ‫[א‬ֱ ‫]ה‬ ַ
In MT ist V. 23 eine rhetorische Frage, in AlT hingegen eine Feststellung;
infolgedessen deklariert MT Jhwh als einen „Gott von ferne,“ während
AlT ihn als „Gott aus der Nähe“ beschreibt. Die Konsequenz ist laut V.
24 in beiden Texttypen gleich: ‫א־א ְר ֶאנּוּ‬ ֶ ֹ ‫ ִאם־יִ ָסּ ֵתר ִאישׁ ַבּ ִמּ ְס ָתּ ִרים וַ ֲאנִ י ל‬.
24b ist aufgrund der negierten Apodosis nur als rhetorische Frage sinn-
voll, die einschärft: Kein Versteck vermag vor Jhwhs Blick zu schützen.
Der fragende Charakter von 24b spricht dafür, dass diese Eigenart auch
vorweg formal markiert wurde, und damit zugunsten der masoretischen
Fragepartikel, die eine disjunktive Alternativfrage . . . ‫ ִאם‬. . . ‫ ֲה‬erzeugt68.
Zudem ist die Streichung der Fragepartikel in AlT gut begründbar. In MT
verbürgt gerade die Ferne und damit der geweitete Blickwinkel Gottes seine
67 Diese größere Härte des Gottesbildes in JerG* wird von Vonach, „Jeremias“ (Anm.
11), 2771, folgenlos notiert.
68 Vgl. HAL 59a; Ges18 70b.
Allwissenheit. Diese Logik hat man in der alexandrinischen Texttradition

offenbar nicht mehr verstanden; deshalb wandelte man die Frage in eine
Feststellung um, sodass nun umgekehrt die Nähe seine Unentrinnbarkeit
garantiert. Damit ändert sich nicht der maßgebliche Zug des Gottesbildes,
sondern nur der gedankliche Anweg dorthin. Mittels einer anderen Logik
hat AlT den Sinn des Passus gewahrt69.
25:6 ‫ ָל ֶכם‬# ‫ § וְ לֹא ָא ַרע \ ְל ָה ַרע‬c ‫יכם‬

ֶ ‫אוֹתי ִבּ ַמ ֲע ֵשׂה יְ ֵד‬
ִ ‫א־ת ְכ ִעיסוּ‬
ַ ֹ ‫ וְ ל‬b
MT bietet zwei Sätze: In 6b warnt Jhwh davor, ihn zu kränken; für den
Fall, dass sein Ruf Gehör findet, verspricht er den Judäern in 6c, ihnen
nichts Böses zuzufügen. Anstelle von 6c kennt AlT nur einen Infinitiv
ausdruck, der die schlimmen Konsequenzen als selbst verursachte Frucht
der Beleidigung Jhwhs stilisiert. Keine Fassung schreibt Jhwh schmerz
hafte Taten an Juda zu; der masoretische Satz 6c mit dem Subjekt Jhwh
dient ja dem Zweck, für den Fall des Gehorsams seinen Verzicht auf Straf-
handlungen zu erklären. Die alexandrinische Lesart kann also nicht dem
Wunsch entspringen, das Gottesbild aufzuhellen. Für die Bewertung ist
zu beachten: Die masoretische Variante ‫ ָא ַרע‬ist in Jer singulär, während
der alexandrinische Infinitiv ‫ ְל ָה ַרע‬mehrere Parallelen besitzt (25:29; 4:22;
31:28), ein ambivalenter Tatbestand, da er sowohl einen älteren Wortlaut
als auch Adaption an eine gebräuchlichere Formulierung anzeigen kann.
Sollte die masoretische Version sekundär sein, dürfte sie aus dem Motiv
herrühren, das Geschehen der Lenkung durch Jhwh zu unterstellen.
25:30 ‫ ָ שׁאֹג יִ ְשׁ ַאג ַעל־נָ וֵ הוּ ֵה ָידד ְכּד ְֹר ִכים יַ ֲענֶ ה‬. . . ‫יְ הוָ ה ִמ ָמּרוֹם יִ ְשׁ ָאג‬
JerG* 32:16 κύριος ἀφ᾿ ὑψηλοῦ χρηματιεῖ . . . λόγον χρηματιεῖ ἐπὶ τοῦ
τόπου αὐτοῦ καὶ αιδεδ ὥσπερ τρυγῶντες ἀποκριθήσονται
Der Vers fällt auf durch die ungewöhnliche, doppelte Gleichung ‫≙ שׁאג‬
χρηματίζω. Das Standardäquivalent zu ‫ שׁאג‬brüllen lautet ἐρεύγομαι brül-
len; JE12 kennt darüber hinaus noch ὠρύομαι mit derselben Bedeutung70.
Alle diese Verben bezeichnen vorwiegend Rufe von Raubtieren. Einzig
im zitierten Passus wird ‫ שׁאג‬mit χρηματίζω wiedergegeben. G benutzt
69 Nach der Erklärung von Vonach, „Jeremias“ (Anm. 11), 2775, wird in den beiden
Fassungen „inhaltlich das direkte Gegenteil ausgesagt . . . Drückt der Satz im Hebräischen
nämlich v.a. die Ferne Gottes aus, so wird in der LXX gerade seine Nähe betont.“
70 Vgl. E. Hatch und H. A. Redpath, A Concordance to the Septuagint and the Other Greek
Versions of the Old Testament, 2. Aufl. (Grand Rapids: Baker, 1998), s. v.; E. C. dos Santos, An
Expanded Hebrew Index for the Hatch-Redpath Concordance to the Septuagint (Jerusalem:
Dugith, o. J. [1973]), s. v.
das Verb in seiner Grundbedeutung handeln, beschäftigt sein in 1 Kön/

3 Bas 18:27 und Ijob 40:8; daneben kommt es nur und gehäuft in JerG*
32–43 (MT 25–36) vor. Es bildet also ein Spezifikum der zweiten Hälfte
des griechischen Jeremiabuches, die sich durch eine Anzahl von Überset
zungskonventionen von der ersten Hälfte abhebt71. Dabei repräsentiert
χρηματίζω, von 25:30/32:16 abgesehen, sonst immer ‫דבר‬-D, und zwar stets
bezogen auf Reden Jhwhs (30:2; 36:2, ‌4) oder eines Propheten (26:2cd),
der auch ein Falschprophet sein kann (29:23). In JerG* hat χρηματίζω
daher die Spezialbedeutung (prophetisch) reden angenommen72, die im
NT fortwirkt73.
In 25:30/32:16 charakterisiert das Verb ebenfalls Kundgaben Jhwhs.
Der Vers besitzt zwei Parallelen mit ‫ שׁאג‬und Subjekt Jhwh, wobei G*
gleichermaßen die Standardäquivalente vermeidet: In Joël 4:16 wird ‫שׁאג‬
mit ἀνακράζω aufschreien und in Am 1:2 mit φθέγγομαι äußern, reden
wiedergegeben. Die Fälle stehen in JE12, deren griechische Ausgaben ‫שׁאג‬
in andersartigen Kontexten durchaus wörtlich repräsentieren (ἐρεύγομαι
Hos 11:10; Am 3:4, 8; ὠρύομαι Jer 2:15; Ez 22:25; Hos 11:10; Zef 3:3). Diese Pra-
xis legt den Schluss nahe, dass der JE12–Übersetzer ‫ שׁאג‬in unverfängli-
chen Zusammenhängen wörtlich wiedergab, aber das Subjekt Jhwh nicht
mit Verben verbinden wollte, die sonst für das Gebrüll von Tieren eintre
ten. Aus ähnlichen Pietätsgründen dürfte er für ‫ יַ ֲענֶ ה‬den Plural gewählt
und so den ‫ ֵה ָידר‬-Ruf auf (ungenannte) menschliche Akteure verschoben
haben74.
25:31 ‫ ָה ְר ָשׁ ִעים נְ ָתנָ ם \ נִ ְתּנוּ‬c ‫ל־בּ ָשׂר‬

ָ ‫ נִ ְשׁ ָפּט הוּא ְל ָכ‬b ‫ ִכּי ִריב ַליהוָ ה ַבּגּוֹיִם‬a
‫ַל ֶח ֶרב‬
Laut MT liefert Jhwh eigenhändig die Frevler dem Schwert aus; AlT 32:17
dagegen gebraucht das Passiv, wobei indes der Vortext keinen Zweifel
zulässt, wer die Fäden in der Hand hält. Da Jhwhs Gerichtshandeln den
71 Vgl. E. Tov, The Septuagint Translation of Jeremiah and Baruch. A Discussion of an

Early Revision of the LXX of Jeremiah 29–52 and Baruch 1:1–3:8 (HSM 8; Missoula: Scholars
Press, 1976); speziell zu χρηματίζω S. 71. Zu den Konsequenzen der Zweiteilung vgl. zuletzt
T. S. L. Michael, „Bisectioning of Greek Jeremiah: A Problem to Be Revisited?“ BIOSCS 39
(2006): 93–104; A. Pietersma, “An Excursus on Bisectioning Ieremias” (2007): http://ccat
.sas.upenn.edu/nets/edition/ieremias-excursus.pdf.
72 Lust, Eynikel und Hauspie, Lexicon (Anm. 37), s. v.
73 W. Bauer, Griechisch-deutsches Wörterbuch zu den Schriften des Neuen Testaments
und der frühchristlichen Literatur, 6., völlig neu bearb. Aufl. hg. von K. Aland und B. Aland
(Berlin: De Gruyter, 1988), s. v.
74 In diesem Fall ist daher der Meinung von Fischer, „Diskussion“ (Anm. 6), 85, und
Vonach, „Jeremias“ (Anm. 11), 2791, beizupflichten.
Übeltätern gilt und obendrein primär die Fremdvölker heimsucht, lud es

nicht dazu ein, dem Porträt Jhwhs freundlichere Züge zu verleihen. MT
weist mit der Pendenskonstruktion . . . ‫ ָה ְר ָשׁ ִעים נְ ָתנָ ם‬die syntaktisch kom
plexere Struktur auf, doch zeigt die alexandrinische Tradition generell
keine Neigung, dieses Bildemuster durch einfachere zu ersetzen. Daher
hat wahrscheinlich AlT den originalen Wortlaut bewahrt, während MT
wieder die göttliche Regie ins Licht rückt.
26:3 und öfter: ‫נחם‬-N mit göttlichem Subjekt ≙ (ἀνα)παύομαι

In der ersten Hälfte von JerG* 75 wird ‫נחם‬-N bereuen überwiegend mit
μετανοέω bereuen wiedergegeben, gleichgültig ob Menschen (8:6) oder
Jhwh (4:28; 18:8, 10) als Subjekt fungieren. Daneben wird in 20:16 bei
göttlichem Subjekt μεταμελέομαι bereuen, bedauern eingesetzt; ferner
bietet 15:6 in Gottesrede die Äquivalenz ‫יתי ִהנָּ ֵחם‬ ִ ‫לא‬
ֵ ִ‫ נ‬ich bin es müde,
Mitleid zu haben ≙ καὶ οὐκέτι ἀνήσω αὐτούς und ich werde sie nicht mehr
freilassen, wo das Schriftbild des Infinitivs anscheinend von ‫נוח‬-H abge-
leitet wird (‫) ַהנִּ ָחם‬. Diese Praxis bewegt sich in den Bahnen der übrigen
Bücher aus JE12, die für ‫נחם‬-N mit Subjekt Jhwh konstant μετανοέω wäh-
len76. Die zweite Hälfte von JerG* hingegegen gebraucht bei göttlichem
Subjekt (ἀνα)παύομαι aufhören, was ebenfalls an eine Ableitung von ‫נוח‬
denken lässt (26/33:3, 13, 19; 42/49:10), während für die Reue Efraims in
31/38:19 weiterhin μετανοέω eintritt. Die Gleichung ‫נחם‬-N ≙ παύομαι ist
nochmals in Jes 1:24 belegt, wiederum bei göttlichem Akteur. Allerdings
findet sie sich auch bei Rahel in 31/38:15 ‫נחם‬-N getröstet werden, sich
trösten lassen. Obendrein wird παύομαι für ‫דכא‬-D pass (Subjekt: Judäer)
44/51:10 herangezogen. Die Basis ‫ דכא‬nebst ihren Varianten ‫דכה‬, ‫דכך‬
und ‫ דוך‬ist sonst in JE12 nicht belegt und könnte dem Übersetzer unbe-
kannt gewesen sein. Der Befund legt den Schluss nahe, dass man in JerG*
29–52 eine theologisch motivierte Wiedergabe vorgezogen hat, die den
Anthropomorphismus vermeiden sollte, Gott könne Reue empfinden. Das
Äquivalent (ἀνα)παύομαι dürfte von der graphischen Ähnlichkeit ‫נחם‬/‫נוח‬
angeregt sein und wurde noch in weiteren Fällen eingesetzt, wo es dem
Übersetzer wünschenswert erschien.
75 Vgl. oben die Ausführungen zu 25:30 und speziell Anm. 71.

76 Jhwh/Gott: μετανοέω: Joël 2:13, 14; Am 7:3, 6; Jon 3:9, 10; 4:2; Sach 8:14; Menschen:
μεταμελέομαι: Ez 14:22.
27:8 . . . # ‫יהם‬ ֶ ‫]וּב ֶדּ ֶבר[ ֶא ְפקֹד § ַעל־הגּוֹי ַההוּא \ ֲע ֵל‬

ַ ‫וּב ָר ָעב‬
ָ ‫ַבּ ֶח ֶרב‬
‫( \ ֻתּ ָמּם ]א ָֹתם[ ְבּיָ דוֹ‬Gram!) ‫ד־תּ ִמּי‬
ֻ ‫ַע‬
Der Satz kündigt Völkern, die sich der göttlich verfügten Weltherrschaft
Nebukadnezzars (V. 6) verweigern, den Untergang von der Hand des
babylonischen Großkönigs an. MT betont eigens die Urheberschaft Jhwhs
in diesem Geschehen: ‫ד־תּ ִמּי א ָֹתם ְבּיָ דוֹ‬ ֻ ‫ ַע‬mit ungrammatischer kausativer
Funktion des Grundstamms von ‫תמם‬. AlT bietet eine regelhafte Variante,
allerdings ohne Verweis auf das göttliche Zutun: ‫ד־תּ ָמּם ְבּיָ דוֹ‬ ֻ ‫ ַע‬. Weil die
Drohung nicht nur Juda, sondern allen Völkern gilt, ist so wenig wie bei
25:31 ersichtlich, warum der Passus den Drang zur Aufhellung des Got
tesbildes geweckt haben sollte. Allenfalls könnte man in AlT den Wort
laut an 24:10 ‫ד־תּ ָמּם‬ ֻ ‫ ַע‬angeglichen haben. Die mangelnde Grammatizität
spricht indes für eine späte Korrektur auf masoretischer Seite, als die
Sicherheit im Umgang mit der hebräischen Sprache bereits im Schwin-
den begriffen war. Offenbar hat man erneut dem Bedürfnis genügt, die
Autonomie und Majestät Jhwhs zu steigern77. In ähnlicher Weise wurde
auch in AlT 42:12 und MT 43:10, ‌12 eine Initiative vom König von Babel auf
Jhwh verlagert. Wenn der Vorgang in 21:7 umgekehrt verlief, so deshalb,
weil dort Nebukadnezzar erstmals in den Vers eingeführt wurde.
27:10 ‫ ִכּי ֶשׁ ֶקר ֵהם נִ ְבּ ִאים ָל ֶכם ְל ַמ ַען ַה ְר ִחיק ֶא ְת ֶכם ֵמ ַעל ַא ְד ַמ ְת ֶכם‬a
[‫ וַ ֲא ַב ְד ֶתּם‬c ‫ ]וְ ִה ַדּ ְח ִתּי ֶא ְת ֶכם‬b
Laut MT wird die Zerstreuung der von den Falschpropheten verleiteten
Fremdvölker von Jhwh selbst vollstreckt (10b), während AlT 10bc über
springt. Wie schon im vorigen Beispiel ist kein Grund zu erkennen,
warum ein alexandrinischer Tradent gewünscht haben sollte, diesen Zug
aus dem Charakterprofil Jhwhs zu tilgen. Ohnehin hätte es dafür gereicht,
10b auszuscheiden oder in einen Infinitiv umzuformen, der wie ‫ַה ְר ִחיק‬
10a der Präposition ‫ ְל ַמ ַען‬untergeordnet wäre und die Zerstreuung als Ziel
bzw. Effekt der Falschprophetie hinstellen würde, wie es auch V. 15 AlT
tut (dazu sogleich). Da dies nicht geschehen ist, hat wahrscheinlich nicht
AlT das Gottesbild gemildert, sondern MT hat es zugespitzt, offenbar im
Interesse der Geschichtssouveränität Jhwhs.
27:15 ‫תכם‬ֶ ‫יח]י[ ֶא‬

ִ ‫וְ ֵהם נִ ְבּ ִאים ִבּ ְשׁ ִמי ַל ָשּׁ ֵקר ְל ַמ ַען ַה ִדּ‬
Nach MT hat die judäische Falschprophetie zur Folge, Jhwh zur Zer
streuung Judas zu treiben, während AlT diese Beteiligung Jhwhs nicht
77 Vgl. Anm. 102.

kennt. Für sich betrachtet, lässt die Variante die Erklärung zu, die alex
andrinische Tradition habe das Gottesbild um ein unerwünschtes Detail
bereinigt, doch im Lichte des Parallelfalls in V. 10 (s. o.) hat die Stei
gerung der göttlichen Verfügungsmacht in der masoretischen Tradition
die höhere Plausibilität für sich.
28:13 ‫יהן מֹטוֹת ַבּ ְרזֶ ל‬

ֶ ‫\תי ַת ְח ֵתּ‬
ִ ‫ית‬
ָ ‫מוֹטֹת ֵעץ ָשׁ ָב ְר ָתּ וְ ָע ִשׂ‬
Während laut MT Hananja jene Verendgültigung der babylonischen Welt-
herrschaft vollzieht, die metaphorisch als die Ersetzung des hölzernen
Jochs durch ein eisernes vergegenwärtigt wird, ist dies für AlT ein Werk
Jhwhs. Plausibel ist nur die masoretische Priorität: Offenkundig gab AlT
dem Drang nach, eine Maßnahme von solcher Reichweite in die Verfü-
gung Gottes zu verlegen, wie dies auch V. 14 tut, wo es Jhwh ist, der den
Nationen das eiserne Joch aufbürdet: ‫ל־הגּוֹיִ ם‬ ַ ‫ל־צוַּ אר ָכּ‬
ַ ‫עֹל ַבּ ְרזֶ ל נָ ַת ִתּי ַע‬
]‫[ה ֵא ֶלּה‬
ָ . In Dtn 28:48 handelt Jhwh ebenso an Israel. Das härtere Auftre-
ten Gottes hat man dafür in Kauf genommen.
29:14 ‫( ָל ֶכם‬καὶ ἐπιφανοῦμαι = ‫יתי‬ ִ ‫אתי \ וְ נִ ְר ֵא ִתי )? וְ נִ גְ ֵל‬

ִ ‫וְ נִ ְמ ֵצ‬
Die Rückübersetzung von G* καὶ ἐπιφανοῦμαι ist unsicher; die sonstigen
verwertbaren Belege von ἐπιφαίνομαι in JE12 stehen für ‫‑גלה‬N (Ez 39:28
AlT) und ‫‑ראה‬N (Zef 2:11)78. Ez 39:28 AlT propagiert ähnlich wie hier die
Idee, dass Jhwh sich Judäern in der Fremde offenbare (‫יהם‬ ֶ ‫לוֹתי ֲא ֵל‬
ִ ָ‫ְבּ ִהגּ‬
‫ ָבּגּוֹיִ ם‬anstelle von MT ‫ל־הגּוֹיִ ם‬ ַ ‫לוֹתי א ָֹתם ֶא‬ ִ ְ‫ ְבּ ַהג‬.), weswegen damit zu
rechnen ist, dass ‫יתי‬ ִ ‫ וְ נִ גְ ֵל‬eine übergreifende Tendenz von AlT wider
spiegelt. Andererseits können ‫אתי‬ ִ ‫ וְ נִ ְמ ֵצ‬und ‫ וְ נִ ְר ֵא ִתי‬durch Verschreibung
auseinander hervorgegangen sein. Ferner kann ‫אתי‬ ִ ‫ וְ נִ ְמ ֵצ‬dem Bestreben
entspringen, das sinnlichere Konzept der Gottesnähe, wie von ‫‑ראה‬N
verkörpert, über den Austausch gegen ‫‑מצא‬N zu spiritualisieren bzw.
zu transzendentalisieren. Zugleich lässt sich MT als Versuch deuten, den
Kontext V. 13–14 geläufigeren Ausdrucksweisen anzunähern: Die engste
Parallele zu V. 13–14 MT ist Jes 65:1 mit drei gemeinsamen Verben (‫בקשׁ‬-D,
‫דרשׁ‬, ‫מצא‬-N); vgl. ferner die Beschreibungen von Gotteserfahrungen mit
den Wortpaaren ‫דרשׁ‬/‫מצא‬-N Jes 55:6; 1 Chr 28:9; 2 Chr 15:2; ‫בקשׁ‬/‫מצא‬-N
2 Chr 15:4, 15. Bei dieser unklaren Argumentationslage wird man besser
auf ein Prioritätsurteil verzichten. Jedenfalls vertritt G* mit ἐπιφανοῦμαι
eine Lesart, die eher anthropomorpher ist als MT.
78 Vgl. ferner Ez 17:6 für ‫פנה‬.

30:12 [‫ ָאנוּשׁ ] ְל[ ִשׁ ְב ֵר]ְך‬b ≙ ἀνέστησα σύντριμμα

Das Adjektiv ‫ ָאנוּשׁ‬unheilbar war dem JerG*-Übersetzer unbekannt (vgl.
15:18; 17:9, 16)79. In 30:12 leitete er das Schriftbild behelfsmäßig von der
Wurzel ‫ נשׂא‬ab mit dem Ergebnis Ich (Jhwh) richtete Zusammenbruch
auf. Infolgedessen “the Lord was presented . . . as the one who brought to
Israel the grief mentioned in the unit Jer. 30:12–17.”80 Das Missverständnis
bedingte, dass sich Jhwh in JerG* über die Vorlage hinaus eine grausame
Tat zuschrieb.
30:23 ‫ ≙ ַס ֲע ַרת יְ הוָ ה‬ὀργὴ κυρίου; ‫גּוֹרר‬

ֵ ‫ ≙ ַס ַער ִמ ְת‬ὀργὴ στρεφομένη81
‫ ַס ֲע ָרה‬wird in JerG* sonst in der Parallele 23:19 durch σεισμός repräsentiert;
für ‫ ַס ַער‬steht dort συσσεισμός und in 25:32 λαῖλαψ. Die interpretierende
Wiedergabe von ‫ ַס ֲע ַרת יְ הוָ ה‬und ‫ ַס ַער‬durch ὀργή (κυρίου) zeigt, dass der
Übersetzer keineswegs zögerte, Gott sogar über die Vorlage hinaus mit
Unmutsausbrüchen zu assoziieren. Die Bedeutung der beiden Substantive
wird regelrecht auf intensive negative Emotionen Jhwhs umgelenkt, pas-
send zur Fortsetzung V. 24. Vielleicht hat sich der Übersetzer dabei von
Anklängen an das Substantiv ‫ זַ ַעם‬leiten lassen, das in JE12, wenn es einen
Affekt Jhwhs beschreibt, überwiegend durch ὀργή repräsentiert wird (Jer
50:25; Ez 21:36; 22:24; Nah 1:6; vgl. Ez 22:31; Hab 3:12). Sollte hingegen schon
die Vorlage ‫ זַ ַעם‬gelesen haben, ist die Verschärfung des Gottesbildes für
die hebräische Phase der alexandrinischen Tradition anzunehmen. In die-
selbe Kategorie gehört die Äquivalenz ‫( ְמזִ ָמּה‬bezogen auf Jhwh) ≙ ὀργή
51:11; vgl. ferner 36:7 (s. u.).
31:20 ‫ לוֹ‬# (ἔσπευσα) ‫ל־כּן § ָהמוּ ֵמ ַעי‬

ֵ ‫ַע‬
In Gottesrede wird ‫ ָהמוּ ֵמ ַעי‬meine Eingeweide gerieten in Wallung durch
ἔσπευσα ich eilte repräsentiert. Dies ist ein glaubwürdiges Beispiel für
interpretierende Wiedergabe in JerG*. Offenbar wollte der Übersetzer die
Vorstellung vom Bauch (κοιλία; so JE12 für ‫ ֵמ ֶעה‬82) Jhwhs vermeiden83.
79 Die funktionslose Präposition in [‫ ] ְל[ ִשׁ ְב ֵר]ְך‬ist aus einer Konflation mit Nah 3:19
‫ ְל ִשׁ ְב ֶרָך נַ ְח ָלה ַמ ָכּ ֶתָך‬. . . entstanden. Das enklitische Personalpronomen hingegen ist kaum
zu entbehren und dürfte daher in der noch paläohebräisch geschriebenen alexandrini-
schen Tradition durch Haplographie mit dem folgenden ‫ נַ ְח ָלה‬entfallen sein: ‫ן → ך‬.
80 B. Becking, Between Fear and Freedom. Essays on the Interpretation of Jeremiah 30–31
(OTS 51; Leiden: Brill, 2004), 27.
81 Anstelle von ‫גּוֹרר‬ ֵ ‫ ִמ ְת‬las die hebräische Fassung von AlT wahrscheinlich ‫חוֹלל‬ֵ ‫ִמ ְת‬
(vgl. Jer 23:19).
82 Jer 4:19; Ez 3:3; 7:19; Jon 2:1, 2.
83 Dies ist weithin anerkannt. Vgl. z. B. Becking, Fear (Anm. 80), 37; Fischer und Vonach,
„Tendencies“ (Anm. 60), 69; Vonach, „Jeremias“ (Anm. 11), 2799.
Er setzte daher an die Stelle emotionaler Aufgewühltheit ein gesteuertes,

wenngleich vieldeutiges Verhalten. Hinreichender Anlass zur Frage nach
einer devianten Vorlage besteht nicht84.
32:28 ‫וּל ָכ ָדהּ‬

ְ ‫ְך־בּ ֶבל‬
ָ ‫[ ֶמ ֶל‬. . .] ‫ת־ה ִעיר ַהזֹּאת ְבּיַ ד‬
ָ ‫ ֶא‬# ‫§ ִהנְ נִ י נ ֵֹתן \ ִהנָּ תֹן ִתּנָּ ֵתן‬
34:2 <‫וּל ָכ ָדהּ‬ ְ > ‫ְך־בּ ֶבל‬
ָ ‫ת־ה ִעיר ַהזֹּאת ְבּיַ ד ֶמ ֶל‬
ָ ‫ ֶא‬# ‫§ ִהנְ נִ י נ ֵֹתן \ ִהנָּ תֹן ִתּנָּ ֵתן‬
39:17 ‫יהם‬ ֶ ֵ‫ר־א ָתּה יָ גוֹר ִמ ְפּנ‬
ַ ‫וְ לֹא ִתנָּ ֵתן \ ֶא ֶתּנְ ָך ְבּיַ ד ָה ֲאנָ ִשׁים ֲא ֶשׁ‬
Die drei Übereignungsformeln sind gemeinsam zu behandeln. Die ersten
beiden stimmen weitgehend wörtlich überein und teilen überdies dieselbe
Lesartendifferenz: MT bietet eine aktivische Fassung im Futurum instans,
während AlT eine passivische Figura etymologica bezeugt. Infolgedessen
macht jeweils nur MT explizit Jhwh als Akteur der Auslieferung Jerusalems
an die Babylonier kenntlich. Zu vergleichen sind hierzu die Parallelen in
32:3, wo MT und AlT übereinstimmend die aktivische Fassung vertreten,
und 38:3, wo beide Textformen passivisch formulieren. Auf Quereinflüsse
zwischen diesen Passagen weist die konflationäre Entlehnung von ‫וּל ָכ ָדהּ‬ ְ
aus 32:28 bzw. 38:3 in 34:2 AlT. Bei der negierten Übereignungsformel
in 39:17, die Ebed-Melech die Bewahrung vor anonymen Gegnern ver
heißt, sind die Genera verbi umgekehrt verteilt: MT gebraucht das Pas
siv, während AlT durch das Aktiv Jhwh als Handelnden identifiziert. Die
Zusammenstellung zeigt, dass kein Texttyp bei den betroffenen Übereig-
nungsformeln konsequent ein bestimmtes Genus verbi bevorzugt.
Wie sind die Abweichungen zu bewerten? Die Varianten in 32:28 und
34:2 können zunächst einfach auf die spontane Adaption an 32:3 oder
38:2 zurückgehen. Sollten die Formeln in der alexandrinischen Tradition
entschärft worden sein85, ist dies, wie 32:3 demonstriert, jedenfalls nicht
konsequent geschehen. Ferner ist hervorzuheben, dass JerMT sich von AlT
durch zusätzliche Belege des Deiktikons ‫ ִהנֵּ ה‬unterscheidet, die sich oben-
drein gerade in Kap. 32 häufen86. Daher ist eher den alexandrinischen
Lesarten von 32:28 und 34:2 die Priorität zuzuerkennen, während der
masoretische Wortlaut sich der Vorliebe für ‫ ִהנֵּ ה‬verdanken dürfte, der
ein Tradent unter konflationärer Anlehnung an Vorbilder nachgab. Die
These der Aufhellung des Gottesbildes findet hier jedenfalls keine brauch-
bare Stütze.
84 σπεύδω repräsentiert in JE12 sonst ‫עוז‬-H (Jer 4:6), I ‫( נהר‬Mi 4:1, in der G*-Vorlage
verschrieben zu ‫)?מהר‬, ‫מהר‬-D (Nah 2:6); vgl. Ez 30:9.
85 So Vonach, „Jeremias“ (Anm. 11), 2801, 2803, für den Übersetzer.
86 Vgl. 24:1; 32:24, 27, 28; 49:12, 35.
39:17 bot, da Heilszusage, keinen Anreiz zur theologischen Glättung. Die

sekundäre Wahl des Passivs (MT) ist daher kaum zu motivieren. Dagegen
kann die alexandrinische Fassung an die Parallele 38:16 ‫ם־א ֶתּנְ ָך‬ ֶ ‫ַחי־יְ הוָ ה ִא‬
‫ ְבּיַ ד ָה ֲאנָ ִשׁים ָה ֵא ֶלּה‬. . . angepasst sein; vielleicht hat auch der Wunsch mit-
gespielt, den Vorgang deutlicher der Regie Jhwhs zu unterstellen. Folglich
hat man allenfalls in 39:17 AlT die göttliche Kontrolle über den Vorgang
zu erhöhen gesucht, doch eine Milderung des Gottesbildes lässt sich dem
Material nicht entnehmen.
36:7 ‫ל־ה ָעם ַהזֶּ ה‬

ָ ‫ר־דּ ֶבּר יְ הוָ ה ֶא‬
ִ ‫ִכּי־גָ דוֹל ָה ַאף וְ ַה ֵח ָמה ֲא ֶשׁ‬
ὅτι μέγας ὁ θυμὸς καὶ ἡ ὀργὴ κυρίου ἥν ἐλάλησεν ἐπὶ τὸν λαόν τοῦτον
Der Übersetzer hat die Wiedergabe des Gottesnamens aus dem Relativsatz
in den Matrixsatz verlagert. Die Abfolge kann nicht in der Vorlage gestan-
den haben, denn sie ergäbe eine Konstruktuskette mit mehreren Nomina
regentia, eine in hebräischer Prosa überaus seltene Wortverbindung87.
Demzufolge ist es der Übersetzer gewesen, der mit dem Ausdruck der
Grimm und der Zorn des Herrn die Assoziation Jhwhs mit heftigen nega-
tiven Emotionen nachgerade verstärkt hat. Eine regelrechte Neigung, den
Zorn Jhwhs hervorzuheben, zeigt die alexandrinische Tradition auch in
30:23 und 51:11.
36:26 ‫] § וַ יַּ ְס ִתּ ֵרם‬. . .[ ‫] וְ ֵאת יִ ְר ְמיָ הוּ‬. . .[ ‫ת־בּרוְּך‬

ָ ‫ ָ ל ַק ַחת ֶא‬. . . ‫וַ יְ ַצוֶּ ה ַה ֶמּ ֶלְך‬
# ‫יְ הוָ ה \ וַ יִ ָסּ ֵתרוּ‬
Baruch und Jeremia entgehen dem Zugriff von Jojakims Häschern, indem
sie ein Versteck aufsuchen. Laut MT ist dies ein Werk Jhwhs, wovon AlT
nichts weiß. Glaubhaft ist nur der Weg von AlT zu MT, mithin der Nach-
druck auf der aktiven Intervention Jhwhs in MT88.
40:3 ‫ל־ה ָמּקוֹם ַהזֶּ ה׃‬

ַ ‫ת־ה ָר ָעה ַהזֹּאת ֶא‬ָ ‫ֹלהיָך ִדּ ֶבּר ֶא‬ ֶ ‫ יְ הוָ ה ֱא‬2
‫אתם ַליהוָ ה‬
ֶ ‫י־ח ָט‬
ֲ ‫ ִכּ‬d [‫ ] ַכּ ֲא ֶשׁר ִדּ ֵבּר‬c ‫ וַ יַּ ַעשׂ יְ הוָ ה‬b [‫ ]וַ יָּ ֵבא‬a 3
Die Rede Nebusaradans an Jeremia in 40:2–5 beginnt mit einer kurzge
fassten theologischen Deutung der Katastrophe Judas. Über AlT hinaus
betont MT, dass Jhwh das angekündigte Unheil (V. 2) auch herbeigeführt
87 Vgl. B. K. Waltke und M. O’Connor, An Introduction to Biblical Hebrew Syntax (Winona

Lake: Eisenbrauns, 1990), 139 (§ 9.3b).
88 Vgl. Anm. 102.
habe (‫ וַ יָּ ֵבא‬3a)89. Für eine Tilgung ist kein Grund ersichtlich. Ferner sind
Aussagen, die ‫בוא‬-H mit dem Subjekt Jhwh verbinden, in Jer derart häu
fig90, dass der überwiegende Teil der Belege von ‫בוא‬-H dieses göttliche
Subjekt trägt. Eine Ergänzung wie 40:3a entspricht daher den typischen
Mustern des masoretischen Sonderguts. Die Position des Gottesnamens
beim zweiten, von AlT bestätigten verbalen Prädikat unterstützt die
Annahme einer Zutat, die erneut dem Ziel gedient haben dürfte, die gött-
liche Geschichtslenkung zu exponieren.
42:12 ‫ וְ ֶא ֵתּן ָל ֶכם ַר ֲח ִמים‬a

‫ל־א ְד ַמ ְת ֶכם‬
ַ ‫ וְ ֵה ִשׁיב \ וַ ֲה ִשׁב ִֹתי ֶא ְת ֶכם ֶא‬c ‫ וְ ִר ַחם \ וְ ִר ַח ְמ ִתּי ֶא ְת ֶכם‬b
Der Vers aus dem Orakel Jeremias an die angehenden Ägyptenauswan
derer ist seinem Kontext nach dazu bestimmt, den Judäern jene Heils
gaben zu verheißen, die sie erwarten dürfen, wenn sie auf die gewünschte
Emigration verzichten. Faktisch wendet sich der Passus jedoch in einem
Fiktionsbruch an die Adressaten des Erzählers im Exil, um ihnen die
Heimkehr in Aussicht zu stellen (12c)91. Die Rückführung ist nach MT
eine Tat des Königs von Babylon, laut AlT hingegen wird sie von Jhwh
bewirkt. Die letztere Variante wird aus doppeltem Grund als sekundär
erwiesen: Erstens lässt sich 12c AlT aus dem historischen Wissen ableiten,
dass es nicht die Babylonier waren, die die Erlaubnis zum Aufbruch nach
Juda erteilten; schon deshalb hätte man nicht nachträglich diesen Akt in
ihre Hände verlegt. Zweitens hat die Revision in 12b AlT eine Dublette zu
12a hervorgebracht. Infolgedessen hat hier die alexandrinische Tradition
bestimmte Handlungen explizit aus menschlicher in göttliche Regie über
tragen, doch nicht aus innertheologischen Gründen, sondern um retro-
spektiv den Einklang mit dem Geschichtsverlauf herzustellen.
89 Überdies ist ‫ ַכּ ֲא ֶשׁר ִדּ ֵבּר‬3c mit Sicherheit prämasoretischer Zusatz, wie das idiolek
tale Gepräge zeigt; vgl. Stipp, „Updated Index“ (Anm. 4), 194; idem, Konkordanz (Anm. 53),
35. Zu objektlosem ‫ עשׂה‬im Sinne von handeln, es ausführen vgl. von Jhwh 1 Sam 14:6; Ps
119:126; mit anderen Akteuren z.B. Gen 41:34; Jer 12:5; Hag 2:4; Spr 13:16; 21:25; 31:13; Esr 10:4;
1 Chr 28:10; 2 Chr 19:7.
90 2:7; 3:14; 4:6; 5:15; 6:19; 11:8MT, 11, 23; 15:8 u. v. a.
91 Diese Interpretation wird näher gerechtfertigt bei H.-J. Stipp, Jeremia im Parteien
streit. Studien zur Textentwicklung von Jer 26, 36–43 und 45 als Beitrag zur Geschichte Jere-
mias, seines Buches und judäischer Parteien im 6. Jahrhundert (BBB 82; Frankfurt am Main:
Hain, 1992), 188f., 254–256; ferner idem, „Das judäische und das babylonische Jeremiabuch,“
in Congress Volume Ljubljana 2007 (VTSup 133; Hg. J. Krašovec; Leiden: Brill, 2010), 239–264
(262 Anm. 106). Dort werden auch weitere Details der Textbezeugung von 42:12 erörtert.
43:10 [‫ְך־בּ ֶבל ] ַע ְב ִדּי‬

ָ ‫אצּר ֶמ ֶל‬ ַ ‫בוּכ ְד ֶר‬
ַ ְ‫ וְ ָל ַק ְח ִתּי ֶאת־נ‬d ‫ ִהנְ נִ י שׁ ֵֹל ַח‬c
‫ וְ ַשׂ ְמ] ִתּי[ ִכ ְסאוֹ ִמ ַמּ ַעל ָל ֲא ָבנִ ים ָה ֵא ֶלּה‬e
Laut 10e MT kündigt Jhwh an, für Nebukadnezzar in Ägypten einen Thron
zu errichten, wohingegen nach AlT der Babylonier dies selbst tun wird.
Hier hat anerkanntermaßen die masoretische Tradition Souveränität
vom König von Babel auf Jhwh verlagert92. Eine gleichartige Intervention
bezeugt sogleich das nächste Beispiel.
43:12 ‫ֹלהי ִמ ְצ ַריִ ם‬
ֵ ‫וְ ִה ַצּ ִתּי \ וְ ִה ִצּית ֵאשׁ ְבּ ָב ֵתּי ֱא‬
In MT legt Jhwh selbst Feuer an die ägyptischen Tempel, was nach AlT
Nebukadnezzar obliegt. In der Fortsetzung 12b–df, 13a weisen dann beide
Textformen dem Großkönig die Subjektsrolle zu. Dies und der Befund in
V. 10 (s. o.) bezeugen, dass erneut die masoretische Tradition die Macht-
fülle Jhwhs auf Kosten Nebukadnezzars gesteigert hat93.
46:10 <‫וְ ָא ְכ ָלה ֶח ֶרב > ַליהוָ ה‬

Der Präpositionalausdruck wurde in AlT konflationär aus dem paralle-
len Satz in 12:12 interpoliert, mit der Konsequenz, dass die Grausamkeit
Jhwhs gegenüber Ägypten deutlicher hervortrat.
49:35 ‫ילם‬ ָ ‫־ק ֶשׁת ֵע‬

ֶ [‫] ִהנְ נִ י[ שׁ ֵֹבר \ יִ ָּשׁ ֵבר ] ֶאת‬
Das Deiktikon ‫ ִהנֵּ ה‬ist in MT mehrfach nachgetragen worden94, was dafür
spricht, dass dies auch hier geschah. Indem man zudem das Prädikat ins
Aktiv versetzte95 und dem göttlichen Subjekt unterstellte, betonte man
Jhwhs weltweites Walten. Weil ein Fremdvolk betroffen ist, bestand
wenig Anreiz zu abmilderndem Eingreifen.
50:20 ‫ִכּי ֶא ְס ַלח ַל ֲא ֶשׁר ַא ְשׁ ִאיר \ נִ ְשׁ ֲאר‬

Die in das Babylonorakel eingeflochtenen Heilsworte für Israel und Juda
V. 17–20 sagen an ihrem Ende dem überlebenden Rest Vergebung zu. MT
rechnet das Überdauern durch ‫שׁאר‬-H ausdrücklich dem Eingreifen Jhwhs
an, während AlT durch das Passiv (‫שׁאר‬-N) keinen Urheber thematisiert.
92 Vgl. z. B. W. L. Holladay, Jeremiah 2. A Commentary on the Book of the Prophet Jer-

emiah Chapters 26–52 (Hermeneia; Philadelphia: Fortress Press, 1989), 277; anders Vonach
(Anm. 102).
93 Anders wieder Vonach (Anm. 102).
94 Ferner 24:1; 32:1, 24, 27, 28; 34:2; 49:12; 50:12; vgl. 49:15.
95 Dabei ist wohl ebenfalls ‫ ֵאת‬ergänzt worden, das in passivischen Sätzen indes auch
das Subjekt markieren kann.
Weil sich nur der Übergang von AlT zu MT einsichtig machen lässt, ist
abermals eine Steigerung der göttlichen Geschichtssouveränität in MT zu
konstatieren96.
51:11 ‫יתהּ‬
ָ ‫( ְל ַה ְשׁ ִח‬ἡ ὀργὴ αὐτοῦ) ‫ל־בּ ֶבל ְמזִ ָמּתוֹ‬
ָ ‫י־ע‬
ַ ‫ת־רוּח ַמ ְל ֵכי ָמ ַדי ִכּ‬
ַ ‫ֵה ִעיר יְהוָ ה ֶא‬
Die Äquivalenz ‫ ≙ ְמזִ ָמּה‬ὀργή ist in G singulär. JerG* wählt für das Sub
stantiv sonst βδέλυγμα Abscheulichkeit (11:15) und ἐγχείρημα Vorhaben, Ver-
such (23:20 || 30:24). Für das Verb ‫ זמם‬tritt ἐγχειρέω vorhaben, trachten ein
(so im folgenden Vers 51:12)97, für ‫ זִ ָמּה‬ἀπαλλοτρίωσις Entfremdung (13:27).
Wie der mehrmalige Gebrauch von ἐγχειρ- erweist, war dem Übersetzer
die Bedeutung der Wurzel ‫ זמם‬vertraut, doch gab er sie uneinheitlich wie-
der. War seine Vorlage mit MT identisch, hat er hier mit ὀργή ein beson-
ders scharfes Äquivalent eingesetzt. Ähnlich wie in 30:23 stellt sich die
Frage, ob ihm der Anklang an die Wurzel ‫ זעם‬als Inspiration diente, und
wie dort ist nicht völlig auszuschließen, dass bereits die Vorlage entspre-
chend zugespitzt worden war. Der Vergleichsfall 36:7 lässt den Ursprung
der Variante indes eher beim Übersetzer suchen, der folglich keine Scheu
kannte, von Jhwhs Zorn zu reden und dabei sogar über seine Vorlage hin-
auszugehen.
51:59 ‫ר־צוָּ ה >יְ הוָ ה ֵאת< יִ ְר ְמיָ הוּ ַהנָּ ִביא > ֵלאמֹר< ֶאת־ \ ֶאל ְשׂ ָריָ ה‬
ִ ‫ַה ָדּ ָבר ֲא ֶשׁ‬
MT berichtet von dem Wort, das der Prophet Jeremia dem Seraja auftrug;
AlT hingegen spricht von dem Wort, das Jhwh dem Propheten Jeremia auf-
trug, (es) dem Seraja zu sagen. Der Gebrauch von εἰπεῖν für ‫ ֵלאמֹר‬98 deu-
tet auf eine ungeschickt redigierte hebräische Vorlage. Demnach hat AlT
ähnlich wie in 1:1 (s. o.) die Kontrolle über einen Offenbarungsvorgang
ausdrücklich von Jeremia auf Jhwh übertragen99.
3. Auswertung
Ein sachgerechtes Urteil über das vorgelegte Material hat zunächst zu

berücksichtigen, dass die Belegsammlung nur Unterschiede zwischen dem
tiberischen Text und JerG* auflistet. Die behandelten Fälle sind jedoch
96 Laut Vonach, „Jeremias“ (Anm. 11), 2781, wird sogar hier „Gott in der LXX eher milder
dargestellt.“
97 Vgl. auch 4:28: ὁρμάω?
98 Sonst 27:4/34:3; 30/37:1; 37/44:17; vgl. 44/51:26 (MT ‫)א ֵֹמר‬.
99 Vgl. Vonach, „Jeremias“ (Anm. 11), 2785, für den Übersetzer.
eingebettet in eine Fülle von Passagen, in denen die beiden Zeugen

übereinstimmend Jhwh eine Härte im Denken und Handeln zuschreiben,
die bei der Vorliebe für mild getönte Gottesbilder, wie sie in unserem
Kulturkreis verbreitet ist, erfahrungsgemäß erheblichen Anstoß erregen
kann. Schon dies nötigt zur—immerhin banalen—Einsicht, dass antike
jahwistische Textproduzenten und -rezipienten andere Maßstäbe an
ihre autoritative Literatur anlegten als wir heute; was Leser des 21. Jahr
hunderts verstört, braucht ein antikes Publikum keineswegs befremdet
zu haben. Wie die Lektüre des gesamten Buches bestätigt, hat JerG* die
schroffen Züge Jhwhs normalerweise ungefiltert an ihre Leser weiterge
reicht. Dies gilt uneingeschränkt auch für Jhwhs Umgang mit Israel und
Juda100. Dasselbe lässt sich—wie hier nicht mehr näher ausgeführt werden
kann—für die zahlreichen Anthropomorphismen und -pathismen feststel-
len101. Eine zusammenfassende Würdigung muss also in Rechnung stellen,
dass die untersuchten Lesartendifferenzen keine Regelfälle repräsentieren,
sondern ein recht kleines Korpus von Ausnahmen bilden.
Die Analyse hat nicht immer eindeutige Prioritätsurteile geliefert. Sie
hat freilich ergeben, dass in der Tat vereinzelt textliche Modifikationen
plausibel auf das Anliegen rückführbar sind, unerwünschte Details aus
dem Porträt Jhwhs zu tilgen, also Gottesbilder an vorgängige Erwartun
gen anzugleichen. Diese Eingriffe sind aber gering an Zahl und haben
den Status begrenzter Anomalien, zeigen also keineswegs vorherrschende
Tendenzen an, schon gar nicht in JerG*. Insgesamt fördern die hinter den
Devianzen aufscheinenden Motive nicht den Eindruck, dass die Tradenten
der masoretischen und der alexandrinischen Ausgabe des Jeremiabuches
mitsamt dem griechischen Übersetzer von der Sorge über grausame
Implikate der Gotteskonzepte des Werkes umgetrieben wurden. Entgegen
anders lautenden Thesen war ihr Drang, solche Eigenarten zu dämpfen,
wenig ausgeprägt.
Sofern sich ein Trend ausmachen lässt, der eine nennenswerte Anzahl
von Retuschen hervorgebracht hat, betrifft dieser die Neigung des maso
100 Vgl. z. B. 2:30; 4:4, 8–10, 26; 5:3; 5:9 (|| 5:29; 9:8); 5:14–17; 6 passim; 7:14–16, 20, 33f.;
8:1–3, 10, 17f.; 9:10/11, ‌14f./15f., 20f./21f.; 11:11, 14, 17, 22f.; 12:7–13; 13:13f., 17, ‌24–27; 14:11f., 15–18;
15:1–9, 13f.; 16:3–5, 9, 13, 18; 17:27; 18:17; 19:3, 7–13, 15; 20:4–6; 21:4–14; 22:5–7, 22; 23:19f., 39f.;
24:8–10; 25:9–11 u. v. a.
101 So ist in JerG* die Rede vom Grimm (ὀργή, θυμός, παροξυσμός) Jhwhs ebenso gängig
wie in MT und wird, wie oben gezeigt, in 30:23; 36:7; 51:11 sogar noch ausgeweitet. Vgl.
ferner z. B. 1:9; 2:21, 30; 4:19; 8:19–9:1/2; 9:14/15; 10:12f. (|| 51/28:15f.), 17f.; 11:17; 12:2, 7f.; 13:11, 13;
16:17; 17:27; 19:7, 11; 21:4–7, 14; 22:24–26; 23:15, 19f., 39; 24:6; 25:3; 46/26:15; 48/31:36; 49:10/29:11;
49:38/25:18; 50/27:25; 51/28:25, 39, 57 u. ö.
retischen Strangs, die souveräne göttliche Herrschaft über alles Geschehen

zu betonen. Diese Tendenz charakterisiert also gerade nicht JerG*, son-
dern im Gegenteil die masoretische Textüberlieferung. Mehrfach wurden
Handlungen aus menschlichen Händen in die Initiative Gottes verlegt.
Während Jeremia nach dem älteren Wortlaut in 11:18 die Untaten seiner
Widersacher sieht (‫ראה‬-G) und in 18:2 das Gotteswort hört (‫שׁמע‬-G), ver-
schiebt MT die Verben in den H-Stamm und stellt so die Wahrnehmungen
des Propheten als von Gott verursacht hin. Diese Fälle stützen besonders
nachdrücklich die alexandrinische Priorität, weil dort keinerlei Anlass
bestand, die Urheberschaft Jhwhs zwecks Aufhellung seines Porträts oder
Vermeidung von Anthropomorphismen herunterzuspielen. Dies bekräf-
tigt das Urteil über die folgenden Beispiele: In 10:18 hat man Jhwhs Regie
durch den Austausch von ‫ ַבּ ָצּ ָרה‬gegen ‫רוֹתי ָל ֶהם‬ ִ ‫ וַ ֲה ֵצ‬zusätzlich unter
strichen. Dasselbe geschah in 49:35 MT, wo man das akteursblinde Passiv
‫ יִ ָּשׁ ֵבר‬gegen das auf Jhwh bezogene Aktiv ‫ ִהנְ נִ י שׁ ֵֹבר‬vertauschte; ebenso
hat man in 50:20 MT ‫ נִ ְשׁ ֲארוּ‬durch den Kausativ ‫ ַא ְשׁ ִאיר‬ersetzt. Wenn
ferner in 36:26 AlT Baruch und Jeremia sich vor den Nachstellungen Joja-
kims verbergen (‫)וַ יִ ָסּ ֵתרוּ‬, macht MT eine Tat Jhwhs daraus (‫)וַ יַּ ְס ִתּ ֵרם יְ הוָ ה‬.
Nach 27:8 AlT gehen die rebellischen Völker von der Hand Nebukadnez
zars zugrunde (‫ד־תּ ָמּם ְבּיָ דוֹ‬ ֻ ‫) ַע‬, doch laut MT vollzieht dies Jhwh, wäh-
rend der Großkönig zum ausführenden Organ herabsinkt (‫ד־תּ ִמּי א ָֹתם‬ ֻ ‫ַע‬
‫) ְבּיָ דוֹ‬. Wo AlT zufolge Nebukadnezzar in Ägypten seinen Thron errichtet
(43:10) und die Tempel einäschert (43:12), erkennt MT göttliche Akte102.
In 25:6 dürfte älteres ‫ ְל ָה ַרע‬dem negierten Kausativ ‫ וְ לֹא ָא ַרע‬gewichen
sein. Außerdem hat man in 6:11 den Verdacht mangelnder Selbstkontrolle
Jhwhs unterbunden. Und während Jhwh in 10:16 || 51:19 AlT als Erbteil
Jakobs deklariert wird, kehrte man in MT mit Rücksicht auf die göttliche
Majestät die Beschreibungsrichtung um, sodass nun Israel als der Stamm
von Jhwhs Erbteil erscheint.
Wiederholt hat die masoretische Tradition im Interesse der Souverä
nität Jhwhs sogar zusätzliche Kanten in seinem Profil in Kauf genom
men. Um in 16:13 dem Eindruck der Aktionsfähigkeit von Fremdgöttern
einen Riegel vorzuschieben, zog man es vor, deren Erbarmungslosigkeit
102 Die Meinung von Vonach, „Jeremias“ (Anm. 11), 2725, in 27:8; 36:26; 43:10, 12 habe
umgekehrt der Übersetzer zwecks Vermeidung von Anthropomorphismen die göttlichen
Akte in menschliche Hände verlegt, ist an den in Anm. 101 genannten Passagen zu messen.
Nach Vonach müsste sich der Übersetzer auf wenige Beispiele gängigen Redens von Gott
beschränkt und gerade die massiven Anthropomorphismen und ‑pathismen unangetastet
gelassen haben.
auf Jhwh zu übertragen (‫א־א ֵתּן ָל ֶכם ֲחנִ ינָ ה‬

ֶ ֹ ‫)ל‬. In 27:15 wurde die Zerstreu
ung Judas ausdrücklich als sein Werk deklariert (‫יח]י[ ֶא ְת ֶכם‬ ִ ‫) ְל ַמ ַען ַה ִדּ‬
und in 40:3 seine Urheberschaft der Katastrophe deutlicher hervorgekehrt
([‫)]וַ יָּ ֵבא‬. Daher würde es zum Gesamtbild passen, wenn—was allerdings
nicht mit Sicherheit zu klären ist—in 19:9 der Kannibalismus der Judäer
an ihren Kindern (AlT: ‫אכל‬-G) nachträglich explizit auf das Wirken Jhwhs
zurückgeführt worden wäre (MT: ‫אכל‬-H mit göttlichem Subjekt).
Dass diese Modifikationen tatsächlich im masoretischen Strang ein
traten mit dem Ziel, die göttliche Geschichtslenkung selbst um den Preis
zusätzlicher Härten zu exponieren, und nicht etwa umgekehrt auf alex
andrinischer Seite die Züge Jhwhs freundlicher gestaltet wurden, belegen
insbesondere jene masoretischen Sonderlesarten, die Jhwh Grausam
keiten an Frevlern und Fremdvölkern zuschreiben und daher kaum Anlass
zur Abschwächung unerwünschter Charakteristika bieten konnten (25:31;
27:8, 10; 49:35).
Geringfügig hat der Nachdruck auf Jhwhs Regie hinter dem Geschehen
auch in AlT seine Spuren hinterlassen. So wandelten sich in 1:1 die Worte
Jeremias in das Wort Jhwhs, das an Jeremia erging, und in 51:59 wurde der
Auftrag Jeremias an Seraja ersetzt durch das Geheiß Jhwhs an Jeremia
zur Beauftragung Serajas. In 28:13 ging der weltgeschichtliche Akt, der
metaphorisch aufscheint als Herstellung der eisernen Jochstangen, von
Hananja auf Jhwh über. Ferner ist damit zu rechnen, dass man aus Respekt
vor der Hoheit Jhwhs bei mehreren Belegen der Wortereignisformel die
Nennung des Gottesnamens sekundär vor das Glied mit dem Adressaten
Jeremia gerückt hat (11:1; 18:1; 21:1; 32:1; 40:1). Dagegen erscheint Zurück-
haltung geboten bei der Frage, ob die alexandrinische Tradition in 22:12
die Deportation des Joahas nachträglich in göttliche Hände gelegt hat;
gleiches gilt für die aktivische Fassung der Zusage an Ebed-Melech in
39:17, ihn nicht seinen Feinden auszuliefern. Auszuklammern ist jeden
falls 42:12, wo AlT aus der Rückschau die Erlaubnis zur Heimkehr aus
dem Exil vom König von Babylon auf Jhwh verlagert hat, denn hier gaben
keine theologischen Motive (im engeren Sinne) den Ausschlag, sondern
die Kenntnis des tatsächlichen Geschichtsverlaufs. Ist somit eine für MT
charakteristische Tendenz stark vermindert auch in AlT anzutreffen, ent-
spricht dies dem generellen Panorama, wonach in beiden Überlieferungs
strängen ähnliche Praktiken der Textpflege wirksam waren, obgleich in
sehr unterschiedlichem Ausmaß103.
103 Vgl. Stipp, Sondergut (Anm. 3), 165.

Leicht an Grausamkeit gewann das Gottesbild in AlT 20:16 durch den

Zusatz, wohl aus Dtn 29:22 (vgl. Ijob 9:5) entlehnt, Jhwh habe sein Zer
störungswerk <‫ > ְבּ ַאף‬verrichtet; ferner wurde 46:10 an 12:12 adaptiert mit
dem Ergebnis <‫וְ ָא ְכ ָלה ֶח ֶרב > ַליהוָ ה‬. Der Nachdruck auf Jhwhs Zorn in 36:7
ist dagegen mit Sicherheit dem Übersetzer anzurechnen, der sich folglich
mitnichten scheute, den Gott Israels mit negativen Gefühlsausbrüchen
zu verbinden. Dasselbe gilt mit hoher Wahrscheinlichkeit auch für 30:23
und 51:11; andernfalls wurzeln diese heftigen Zuspitzungen schon in der
hebräischen Phase der alexandrinischen Texttradition. Bedingt durch ein
Missverständnis der Vorlage, hat der Übersetzer in 30:12 mit ἀνέστησα
σύντριμμα Jhwh noch eine weitere Unheilstat an Israel zugeschrieben.
Anthropomorphismen, die sogar über MT hinausgehen, bietet AlT bzw.
G* in 15:7 ‫ ֻשׁ ַכּ ְל ִתּי‬/ἠτεκνώθην und 29:14 ‫וְ נִ ְר ֵא ִתי‬/καὶ ἐπιφανοῦμαι.
Von einer Entschärfung des Gottesbildes kann dagegen nur in weni
gen Ausnahmefällen die Rede sein. Einigermaßen sicher ist lediglich,
dass AlT in 1:17 und wohl auch 15:17, 18 den barschen Umgang Jhwhs
mit seinem Propheten abschwächte104. Der Schöpfer von JerG* beschnitt
Anthropomorphismen, als er die Vorstellungen vom Brüllen (25:30) und
vom Bauch Jhwhs (31:20) seinem Publikum vorenthielt; weiterhin scheint
er aus demselben Grund in der zweiten Hälfte des Buches die Rede von
der Reue Gottes vermieden zu haben, indem er für ‫נחם‬-N mit göttlichem
Subjekt (ἀνα)παύομαι aufhören wählte. Wenn jedoch ein alexandrinischer
Tradent oder der Übersetzer das Weinen Jhwhs in 9:9; 14:17 auf das Volk
umleitete, so zeigt die gleichartige Umlenkung des Weinens Jeremias 13:17,
dass nicht Vorbehalte gegen Anthropopathismen die Feder führten, son-
dern das Aussageziel, die Schuldigen sollten ihre Notlage selbst betrauern.
Eine Milderung des Gottesbildes könnte die masoretische Tradition in
22:12 beabsichtigt haben, wenn, was offen bleiben muss, die Deportation
des Joahas nicht in AlT auf Jhwh übertragen (‫) ִהגְ ֵל ִתי‬, sondern umgekehrt
in MT auf Menschen verschoben worden wäre (‫) ִהגְ לוּ‬. AlT hätte das Got-
tesbild in 19:9 aufgehellt (‫‑אכל‬G), sofern nicht im Gegenteil die maso-
retische Überlieferung Jhwh sekundär zum Anstifter des Kannibalismus
erhoben hat (‫‑אכל‬H MT), was insgesamt plausibler ist.
Keinen signifikanten Einfluss auf das Gottesbild hat dagegen die ale
xandrinische Transformation der Frage 23:23 in eine Feststellung, sodass
104 Vielleicht meldet sich hier dieselbe Sorge um die Reputation der authentischen Pro-
pheten zu Wort, die in 26:21 zur Streichung der Notiz von der Furcht und Flucht des Urija
ben Schemaja geführt hat; vgl. Stipp, Sondergut (Anm. 3), 153.
Jhwh von einem ‫ֹלהי ֵמ ָרחֹק‬ֵ ‫ ֱא‬in einen ‫ֹלהי ִמ ָקּר ֹב‬
ֵ ‫ ֱא‬verwandelt wird, denn
die Neufassung leitet dasselbe Theologumenon—die Unentrinnbarkeit
Gottes—mit einer anderen Logik her, nachdem das Verständnis für die
ältere Begründung geschwunden war. Kaum zum Thema gehören die prä-
masoretischen Eingriffe in 21:7, die nicht die Entlastung Jhwhs bezweck
ten, sondern Nebukadnezzar in den Kontext einführen sollten. Einem
Urteil entzieht sich die Passivkonstruktion ἐξώσθησαν beim Rückblick auf
die Zerstreuung Israels in JerG* 16:15, denn ihr textgeschichtlicher Ort
und die Triebkräfte hinter ihrer Entstehung sind nicht befriedigend zu
erhellen. Ein negatives Ergebnis lässt sich immerhin konstatieren: Aus
sekundärem Bemühen um das Gottesbild ist das Passiv wahrscheinlich
nicht erwachsen.
Die verbleibenden Differenzen verdanken sich wohl ganz andersarti
gen Beweggründen. Die nur von MT bezeugte prophetische Botenformel
in 13:12 rührt aus der routinemäßigen Vermehrung klischierter Phrasen in
diesem Texttyp her, und die ehemals passiven Übereignungsformeln in
32:28 || 34:2 nahmen aktive Form an, weil ein Rezensor in der masoreti
schen Tradition das Deiktikon ‫ ִהנֵּ ה‬besonders schätzte.
Die Behauptung, die alexandrinische Tradition bzw. der Übersetzer
hätten anstößige Züge der Gottesvorstellung abgeschliffen, trifft nach all-
dem nur vereinzelt zu. Die wenigen gültigen Belege werden jedoch durch
gegenteilig orientierte Beispiele aufgewogen, und ohnehin ändern sie
nichts am extremen Übergewicht der Unheilsansagen, die eine Unmenge
massiver Härten in den Gottesbildern mit sich bringen, sodass selbst dann,
wenn sich weitere Milderungen nachweisen ließen, das Porträt Jhwhs nur
marginal nuanciert würde.
So lässt sich resümieren: Zu den Motiven, die nach der Gabelung der
Überlieferungsstränge des Jeremiabuches die Textentwicklung antrieben,
gehörten auch Impulse, die man als Arbeit am Gottesbild klassifizieren
kann. Sie waren jedoch erstens insgesamt nur recht schwach entwickelt.
Zweitens wirkten sie bloß minimal in der alexandrinischen Tradition—
vom griechischen Übersetzer ganz zu schweigen—, während die meisten
einschlägigen Retuschen aus prämasoretischen Händen hervorgingen,
wie es das generelle Verhältnis der beiden Arme der Textüberlieferung
ohnehin erwarten lässt. Drittens ging es den Tradenten nur in seltenen
Ausnahmefällen darum, Jhwh mildere Züge zu verleihen, sondern sie
wollten vor allem seine Lenkung allen Geschehens propagieren, und sei
es sogar um den Preis gesteigerter Härte seines Profils. Allem Anschein
nach galt ihr Interesse angesichts ihrer eigenen politischen Machtlosig
keit viel mehr der Geschichtssouveränität Jhwhs als seiner Gnädigkeit.
Was theologische Programme der Septuaginta-Übersetzer angeht, bleibt

der Ertrag bescheiden: Der Schöpfer des griechischen Jeremiabuches
setzte alles daran, seinem Publikum durch getreuliche Wiedergabe des
Muttertextes ein zuverlässiges Bild vom verehrten hebräischen Original
zu vermitteln—so, wie er es verstand.
Two Difficult Passages in the Hebrew Texts of
Lamentations 5: Text-Critical Analyses of the
Greek Translation
Gideon R. Kotzé
1. Introduction
The Greek translation of Lamentations counts amongst the more interest-

ing translation units in the corpus of Greek Jewish scriptures traditionally
known as the Septuagint (LXX). LXX Lamentations is generally charac-
terised as a “literal”1 or “quantitative formal equivalent”2 translation of a
Hebrew Vorlage that was close to the consonantal base of the MT.3 It is
also considered as a member of the kaige group of translations and revi-
sions.4 Concerning the “literal” nature of LXX Lamentations, Albrektson5
is of the opinion that the translator rendered his Vorlage in a mechanical
way and often word for word. By disregarding the context of passages, he
1 According to H. Ausloos and B. Lemmelijn (“Content-Related Criteria in Characteris-

ing the LXX Translation Technique,” in Die Septuaginta—Texte, Theologien, Einflüsse. 2.
Internationale Fachtagung veranstaltet von Septuaginta Deutsch [LXX.D], Wuppertal 23.—
27.7.2008 [eds. W. Kraus and M. Karrer; Tübingen: Mohr Siebeck, 2010], 358), “a transla-
tion is labelled as ‘literal’ when it renders its Vorlage very accurately, translating it in an
almost mechanical way and often word for word”. In contrast to such literal translations,
“a translation is characterized as ‘free’ when it does not proceed in this manner, in other
words when it handles its Vorlage, both linguistically and exegetically, more freely”. They
point out, however, that, in the past, the “distinctive features of ‘literalness’ have not been
precisely defined and there has been very little interest in the ‘freedom’ of a translation”.
Ausloos and Lemmelijn (“Content-Related Criteria,” 359) insist that one must determine
in which ways a translation unit is either literal or free. In this regard, see J. Barr, The
Typology of Literalism in Ancient Biblical Translations (MSU XV; Göttingen: Vandenhoeck
& Ruprecht, 1979).
2 K. J. Youngblood, “The Character and Significance of LXX Lamentations,” in Great
Is Thy Faithfulness? Reading Lamentations as Sacred Scripture (eds. R. A. Parry and H. A.
Thomas; Eugene: Pickwick, 2011), 65.
3 See G. R. Kotzé, “The Greek Translation of Lamentations: Towards a More Nuanced
View of its ‘Literal’ Character,” in Septuagint and Reception (VTSup 127; ed. J. Cook; Leiden:
Brill, 2009), 78.
4 See D. Barthélemy, Les Devanciers d’Aquila (VTSup 10; Leiden: Brill, 1963), 158–160;
and I. Assan-Dhôte and J. Moatti-Fine, Baruch, Lamentations, Lettre de Jérémie (BdA 25.2;
Paris: Cerf, 2005), 155–157.
5 B. Albrektson, Studies in the Text and Theology of the Book of Lamentations with a Criti-
cal Edition of the Peshitta Text (Lund: CWK Gleerup, 1963), 208–210.
276 gideon r. kotzé
fails, at times, to capture the meanings of the Hebrew sentences. Simi-

larly, Hirsch-Luipold and Maier6 depict the approach of the translator as
follows:
Der Septuaginta-Übersetzer versucht, die hebräische Vorlage möglichst dec-
kungsgleich nicht primär im Blick auf den Sinn, sondern auch im Blick auf
die Satzstruktur, die Grammatik sowie die Bedeutung und Form der Einzel-
wörter ins Griechische zu übernehmen. Dabei entstehen äußerst gewagte
Satzkonstruktionen und eine fremdartige, bisweilen poetische Sprache,
aber auch Sätze, die im Griechischen rätselhaft bleiben. Der Sinn der grie-
chischen Übersetzung weicht gerade wegen dieser Übersetzungsmaxime an
manchen Stellen erheblich von dem des hebräischen Textes ab.
With regard to the membership of LXX Lamentation to the kaige group,
only five of the nine key features of the group discussed by Barthélemy are
present in this translation unit.7 First, all the occurrences of ‫ גם‬in Lam-
entations are rendered by καί γε (Lam 1:8, 2:9, 3:8, 4:3, 4:15 and 4:21). Sec-
ondly, in LXX Lam 3:53, ἀνήρ serves as the translation equivalent of ‫איש‬,
but the latter is not used in a distributive sense in this case. Thirdly, forms
of ‫ נצב‬appear in Lam 2:4 and Lam 3:12. In the first passage, ‫ נצב ימינו‬is
translated as ἐστερέωσε δεξιὰν αὐτοῦ. The rendering does not reflect the
kaige group’s characteristic translation equivalent for ‫נצב‬, στηλόω. In Lam
3:12, however, ‫ ויציבני‬is rendered by a form of στηλόω. Fourthly, the Greek
present is only found once in LXX Lam 4:19 (= MT Lam 4:18). Gentry there-
fore remarks that “it can safely be said that the translator was not in the
habit of using historical presents”.8 Finally, ‫ אין‬is translated six times as
οὐκ ἔστιν (Lam 1:9, 1:17, 1:21, 2:9, 4:4 and 5:8), three times as οὐχ ὑπάρχει/
ὑπάρχουσιν (Lam 1:2, 5:3 and 5:7) and once as οὐκ ἦν (Lam 1:7). Gentry9
concludes from this that “only three of the five patterns provide strong
support that the Greek Lamentations belongs to the καίγε tradition . . . It
6 R. Hirsch-Luipold and C. M. Maier, “Threnoi/Die Klagelieder,” in Septuaginta Deutsch.

Das griechische Alte Testament in deutscher Übersetzung (eds. W. Kraus and M. Karrer;
Stuttgart: Deutsche Bibelgesellschaft, 2009), 1349. Cf. also idem, “Threnoi/ Threni Seu
Lamentationes/Die Klagelieder,” in Septuaginta Deutsch. Erläuterungen und Kommentare
zum griechischen Alten Testament. Band II: Psalmen bis Daniel (eds. M. Karrer and W. Kraus;
Stuttgart: Deutsche Bibelgesellschaft, 2011), 2829–2830.
7 Cf. Barthélemy, Les Devanciers d’Aquila, 48–80. In the wake of Barthélemy’s initial
study, scholars have suggested a large number of other features that supposedly also char-
acterise the kaige group. R. T. McLay (“Kaige and Septuagint Research,” Textus 19 [1998]:
127–139) shows that these features are not consistently shared by the members of the
group nor are the differences among these texts sufficiently recognised.
8 P. J. Gentry, “Lamentations,” in A New English Translation of the Septuagint (eds.
A. Pietersma and B. G. Wright; New York / Oxford: Oxford University Press, 2007), 934.
9 Ibid.
two difficult passages in the hebrew texts 277
is clear that some relationship exists between Greek Lamentations and

other texts in the καίγε tradition, but it is not systematic.”
Another interesting issue, relevant to the discussion on the kaige-
membership of LXX Lamentations, is the possibility that it is the work of
Theodotion. This possibility is raised, on the one hand, by the fact that
Origen indicated that the versions of Aquila and Theodotion did not exist
for Lamentations.10 On the other hand, a few readings are attributed to
Aquila in the critical apparatus of Ziegler’s Göttingen edition of Lamen-
tations.11 Gentry12 refers to a study of one of his students, Kevin Young-
blood, who investigated this possibility by comparing and contrasting LXX
Lamentations with the Greek Minor Prophets scroll from Naḥal Ḥever, the
Greek Psalter, Aquila and the materials in the book of Job that are attrib-
uted to Theodotion. He reports that the “comparison between Lamenta-
tions and Theodotion Job was interesting. The evidence showed not only
strong similarities, but also enough differences to make it fairly certain
that the same person did not produce both translations”.13 As a result,
Gentry concludes: “Theodotion Job and OG Lamentations have the same
type of approach to translation, but the Greek Lamentations is probably
not Theodotion”.14
Although the “literal” character of LXX Lamentations seems well estab-
lished, it is necessary and possible to produce a more nuanced profile of
this translation unit.15 One way to achieve this is to study how the Greek
translator rendered passages in the Hebrew text that can be considered
“difficult”.16
10 Cf. P. J. Gentry, “Old Greek and Later Revisors: Can We Always Distinguish Them,” in
Scripture in Transition. Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour
of Raija Sollamo (eds. A. Voitila and J. Jokiranta; Leiden / Boston: Brill, 2008), 326; and
Youngblood, “LXX Lamentations,” 65–66.
11 J. Ziegler, Septuaginta. Vetus Testamentum Graecum Auctoritate Academiae Scien-
tiarum Gottingensis editum XV: Jeremias, Baruch, Threni, Epistula Jeremiae (Göttingen:
Vandenhoeck & Ruprecht, 1976).
12 Gentry, “Old Greek and Later Revisors,” 326.
13 Ibid., 327.
14 Ibid.; L. J. Greenspoon (“The Kaige Recension: The Life, Death, and Postmortem
Existence of a Modern- and Ancient Phenomenon,” in XII Congress of the International
Organization for Septuagint and Cognate Studies, Leiden 2004 [ed. M. K. H. Peters; Atlanta:
Society of Biblical Literature, 2006], 5–16) also argues that the members of the kaige group
cannot always be linked to the version of Theodotion.
15 Kotzé, “The Greek Translation of Lamentations,” 93.
16 This suggestion is based on the conclusions of a previous study: G. R. Kotzé, “Lamen-
tations 4:7 and 4:14. Reflections on the Greek Renderings of the Difficult Hebrew Wordings
of These Verses,” JSem 20/1 (2011): 250–270.
2. Difficult Readings and Textual Criticism
James Barr defines a “difficult” reading as follows:17 A reader

feels that [the reading] ‘does not make sense’. The grammar is “wrong”, i.e.
does not fit with usual patterns of usage. The use of words is anomalous.
Or perhaps the text contradicts what is said elsewhere in the same literary
work, so that it seems to “spoil the effect” of the whole; or it may contradict
something well known from altogether other sources.
Thus, a textual difficulty can pertain to the semantics of words, the gram-
mar and syntax of a passage and/or the content of the passage in relation
to its larger literary contexts.
From the perspective of Old Testament textual criticism, a reader can
deal with such textual difficulties in different ways. Traditionally, the dis-
cipline of Old Testament textual criticism studies the biblical texts, their
development during the processes of transmission (copying and transla-
tion) and evaluates different readings in order to determine which read-
ing is the original one.18 With regard to a textual difficulty, the reader who
follows this approach to textual criticism will, in the first instance, want
to determine whether there are readings in the available textual repre-
sentatives that differ from the difficult reading. If a collation of the extant
Hebrew manuscripts and/or an analysis of the ancient translations, which
were based on Hebrew Vorlagen, reveal the existence of diverging Hebrew
readings, the reader must account for these different readings and how the
difficult reading came into being. An examination of the different readings
may lead the reader to assume that the difficult reading was the result of
some scribal error. Alternatively, the reader can come to the conclusion
that the other readings developed from the difficult reading and therefore,
that this reading, despite its difficulties, represents the original wording of
17 J. Barr, Comparative Philology and the Text of the Old Testament (Oxford: Clarendon
Press, 1968), 3.
18 See E. Tov, Textual Criticism of the Hebrew Bible. Second Revised Edition (Assen / Min-
neapolis: Van Gorcum / Fortress Press, 2001), 289–290. Text-critics formulate their findings
in connection with the evaluation of diverging readings in different ways (Tov, Textual
Criticism of the Hebrew Bible, 310): “Some speak in terms of preferable readings, others refer
to better or (more) original readings, and again others try to identify the reading from
which the other ones presumably derived”. See, for example, A. van der Kooij, “Textual
Criticism of the Hebrew Bible. Its Aim and Method,” in Emanuel. Studies in the Hebrew
Bible, Septuagint, and Dead Sea Scrolls in Honor of Emanuel Tov (eds. S. M. Paul, R. A. Kraft,
L. H. Schiffman and W. W. Fields; Leiden / Boston: Brill, 2003), 729–739; and B. Lemmelijn,
“What Are We Looking For in Doing Old Testament Text-Critical Research?” JNSL 23/2
(1997): 69–80.
the passage.19 If the reader is satisfied that none of the readings preserved
in the textual representatives constitute the original text, the wording can
be emended so as to restore its presumed original form. This is referred
to as conjectural emendation, given that the suggested original reading is
conjectured from the actual readings in the textual representatives.
In cases where the Hebrew manuscripts and the Vorlagen of the ancient
translations (as far as their details can be restored with certainty) present
a uniform text, a textual difficulty in the Hebrew wording can be seen as
a copyist error. The reader can then solve the difficulty by means of con-
jectural emendation.20 However, if the reader does not regard the textual
difficulty as an error, attempts can be made to make sense of the difficult
reading with the help of comparative philology.
In the approach to textual criticism, where the available textual rep-
resentatives are treated as witnesses to the original Hebrew texts, the
ancient translations are relevant to the discussion of textual difficul-
ties, insofar as they were based on Hebrew wordings that differ from
the Hebrew text containing the difficulty. The text-critical “significance”
of the variant readings in the ancient translations will, in this approach,
depend on whether they are considered to be more original than the dif-
ficult readings in the Hebrew text. In the philological treatment of tex-
tual difficulties, the ancient translations are used primarily to confirm an
understanding of the Hebrew text that is based on evidence culled from
the cognate languages.21
Methodologically speaking,22 these ways of dealing with the ancient
translations are appropriate to text-critical efforts at editing or elucidating
19 One of the (internal) criteria for evaluating diverging readings is that the more dif-
ficult reading is the preferable one (lectio difficilior praeferanda est). According to R. W.
Klein (Textual Criticism of the Old Testament. From the Septuagint to Qumran [Philadelphia:
Fortress Press, 1974], 75), “Grammatical, historical, theological, and lexical difficulties often
were eliminated or modified by the scribes as they copied the manuscripts. The scribes
would not knowingly insert a more difficult form for a common one or an archaic or rare
word instead of one in everyday usage.” P. K. McCarter (Textual Criticism. Recovering the
Text of the Hebrew Bible [Philadelphia: Fortress Press, 1986], 21) notes that this text-critical
rule of thumb is another way of posing the question: Which reading would have been
more likely to give rise to the others (utrum in alterum abiturum erat)?
20 Barr (Comparative Philology, 3) correctly indicates that conjectural emendations pro-
ceed from the supposition that textual difficulties constitute errors in the wording of a
passage.
21 See Barr, Comparative Philology, 238–272, concerning the use of the ancient versions,
especially the LXX, in philological treatments of textual difficulties.
22 On the appropriate methodologies for using the ancient translations in text-critical
research, see, for example, E. Tov, The Text-Critical Use of the Septuagint in Biblical
the Hebrew text of Old Testament writings, especially in cases of prob-

lematic passages. However, text-critical research can also be relevant to
the interpretation of Old Testament writings in other ways than solving
difficulties in the Hebrew wordings and providing exegetes with prob-
lem-free Hebrew texts to base their exegesis on. If the putative original
Hebrew text and the textus receptus are not regarded as the most impor-
tant wordings of an Old Testament writing and therefore, as (practically)
the only legitimate representatives of the writing’s content, the wordings
of the ancient translations gain importance as witnesses to the content
of a biblical book.23 When the ancient translations are studied as repre-
sentatives of an Old Testament writing’s content, the reader must have a
good understanding of why the wordings of the translations look the way
they do.24 Seeing as text-critics investigate the ways in which readings in
the textual representatives were created during the processes of copying
and translation, textual criticism can contribute to such a better under-
standing of the wording of ancient translations and how they present the
content of passages in an Old Testament writing.
Accordingly, this study, which focuses on LXX Lamentations, takes as
its point of the departure the assumption that text-critical analyses of the
Greek renderings of difficult passages in the Hebrew wordings of Lam-
entations can open new vistas on the way in which the Greek translator
translated his Vorlage. This, in turn, can make a small but significant con-
tribution to a better understanding of the wording of LXX Lamentations,
as well as a more nuanced profile of this translation unit.
Research. Revised and Enlarged Second Edition (Jerusalem: Simor, 1997); and J. A. Adair, “A
Methodology for Using the Versions in the Textual Criticism of the Old Testament,” JNSL
20/2 (1994): 111–142.
23 The content of an Old Testament writing refers to what a reader takes the wording
of the writing to convey. In other words, it is the reader’s perception of what the wording
of the writing means.
24 This is one of the lessons learned from rhetorical criticism as practiced by, for instance,
Phyllis Trible. Trible (God and the Rhetoric of Sexuality [Philadelphia: Fortress Press, 1978])
stresses the organic unity of the form or wording of a text and its content: “Form and
content are inseparable. On the one hand, the text is not a container from which ideas or
substance can be abstracted to live an independent life. On the other hand, the text is not
a subject matter from which stylistic and structural wrappings can be removed to exist
autonomously. How the text speaks and what it says belong together in the discovery of
what it is. To convey content is to employ form; to convey form is to employ content”. See
also P. K. Tull, “Rhetorical Criticism and Intertextuality,” in To Each Its Own Meaning. An
Introduction to Biblical Criticisms and Their Applications. Revised and Expanded (eds. S. L.
McKenzie and S. R. Haynes; Westminster: John Knox Press, 1999), 159.
3. Method of Analysis
Bearing this assumption in mind, two verses of Lamentations 5, which

contain textual difficulties will be subjected to text-critical analyses. The
problematic passages are Lam 5:10 and 5:13. In each case, the problematic
aspect of the Hebrew wording(s)25 is first identified and discussed before
the Greek text26 is subjected to analysis. The analyses will treat the Greek
texts of Lam 5:10 and 5:13 as both translations and as witnesses to the con-
tent of the verses in their own right.27 The analyses therefore entail exami-
nations of the readings in the Greek text28 that attempt to determine the
most plausible explanations of how these readings were created during
the processes of translation. The analyses also include comparisons of
the Greek translation with other ancient translations. Such comparisons
serve three purposes. First, they highlight unique renderings of difficult
Hebrew words, phrases or clauses in the Greek translation. Secondly, they
reveal instances where ancient translators arrived, independently from
each other, at the same interpretation of the Hebrew text, as well as cases
where the Greek text might have influenced later translations. Thirdly,
similar readings in the Greek text and another ancient translation raise
the possibility that these translations were based on Hebrew Vorlagen
that differ from the extant Hebrew text(s) used in the text-critical analy-
sis. For the purposes of this study, comparisons with the Peshitta29 and
25 For the wording of the MT, I primarily make use of the BHQ fascicle edition:
R. Schäfer, “Lamentations,” in Biblia Hebraica quinta editione cum apparatu critico novis curis
elaborato. General Introduction and Megilloth (eds. A. Schenker et al.; Stuttgart: Deutsche
Bibelgesellschaft, 2004), 54–72, 113*–136*. I also consult the critical apparatuses of BHK and
BHS. Codex Leningradensis is used as the base text in all these editions. With regard to the
Qumran manuscripts, the wordings of some verses of Lamentations 5 were preserved on a
fragmentary manuscript found in cave 5 (5QLama). I make use of the DJD edition prepared
by J. T. Milik (“Lamentations [Premier Exemplaire],” in Les “Petites Grottes” de Qumran:
Exploration de la Falaise, Les Grottes 2Q, 3Q, 5Q, 6Q, 7Q à 10Q, Le Rouleau de Cuivre [DJD III;
M. Baillet, J. T. Milik and R. de Vaux; Oxford: Clarendon Press, 1962], 174–177).
26 The Göttingen edition of Lamentations was prepared by J. Ziegler. I make critical use
of this edition, as well as of A. Rahlfs’s edition of the LXX, which was edited by R. Hanhart:
Septuaginta. Id est Vetus Testamentum graece iuxta LXX interpretes (Stuttgart: Deutsche
Bibelgesellschaft, 2006).
27 Cf. J. Cook, “Exegesis in the Septuagint,” JNSL 3/1 (2004): 2.
28 Given that this study has the work of the original Greek translator of Lamentations in
view, the Old Greek text of LXX Lamentations must be used for the analysis. The Old Greek
text refers to the earliest form of the translation’s wording that can be (re)constructed on
the basis of the available manuscripts and conjectural emendation.
29 B. Albrektson’s critical edition of the Peshitta text of Lamentations serves as the
source for readings of the Syriac translation.
the Vulgate30 will suffice. After the most plausible explanations for the
(Old) Greek readings are identified, the analyses subsequently establish
the effect that the particular readings in the wording of the Greek text
have on the content of the passages in question.
4. Analysis of the Hebrew and Greek Wordings of

Lamentations 5:10 and Lamentations 5:13
4.1. Lamentations 5:10
4.1.1. The Hebrew Wordings
MT ‫ֹעור ֙נּו ְּכ ַתּנ֣ ּור נִ ְכ ָ֔מרּו ִמ ְּפ ֵנ֖י זַ ְל ֲע ֹ֥פות ָר ָ ֽעב׃‬
ֵ֙
Our skin is wrinkled like an oven, because of the rages of hunger.
5QLama ‫‍עורי֯ [נ]ו֯ ֯כתנור נכמרו ֯מ[פ]ני ז֯ לפות ֯רעב‬

[Our] skins are wrinkled like an oven, be[cau]se of the rages31 of hunger.
The wordings of Lam 5:10 are almost identical in 5QLama and the MT.
In the Masoretic manuscript, there is incongruence between the verb
‫נִ ְכ ָמרּו‬, which is plural, and its subject, ‫עֹורנּו‬
ֵ , which is singular. This dis-
crepancy in number was corrected by the scribe who copied 5QLama. He
initially copied the opening word of the verse as ‫עורנו‬, but then (presum-
ably after copying ‫ )נכמרו‬inserted a yôd in the interlinear space after the
rêš of ‫עורנו‬. This scribal correction changes the suffix from one denoting
a singular noun to one indicating a plural noun;32 therefore, the word
can be restored as ‫עורינו‬. The form ‫ עורינו‬is also found in a number of
30 For the Vulgate, I use the fifth edition of R. Weber edited by R. Gryson (Biblia Sacra
iuxta Vulgatam Versionem [Stuttgart: Deutsche Bibelgesellschaft, 2007]).
31 E. Y. Kutcher (A History of the Hebrew Language [Jerusalem: Magness Press, 1982],
96) notes that gutturals weakened in Qumran Hebrew. These consonants were sometimes
confused with one another or elided altogether. See also M. G. Abegg, “The Linguistic
Analysis of the Dead Sea Scrolls: More Than (Initially) Meets the Eye,” in Rediscovering
the Dead Sea Scrolls. An Assessment of Old and New Approaches and Methods (ed. M. L.
Grossman; Grand Rapids: Eerdmans, 2010), 59; and E. Qimron, The Hebrew of the Dead Sea
Scrolls (Atlanta: Scholars Press, 1986), 25–26. The reading ‫ ז֯ לפות‬in 5QLama is an example
of the elision of the pharyngeal ‘áyĭn.
32 See Schäfer, “Lamentations,” 135* and Milik, “Lamentations (Premier Exemplaire),”
175.
medieval Masoretic manuscripts.33 Ehrlich34 argues that ‫ עורנו‬should be

emended into ‫עורינו‬, in order to eradicate the incongruence between the
subject and verb in the passage. However, Rudolph35 maintains that this
will not mitigate the problem, because the normal plural form of ‫ עור‬is
‫עורות‬.36 He is of the opinion that the wāw of ‫ נכמרו‬must be removed so
that the verbal form will agree in number with its subject. Scholars such as
Westermann, Kraus and Dyserinck share this view.37 In contrast, Dahood38
proposes that ‫עֹורנּו‬ֵ in the MT should be taken as scriptio defectiva for
‫עורינו‬, whilst Renkema39 thinks that one can understand ‫עֹורנּו‬
ֵ in a collec-
tive sense. These two suggestions have the benefit of upholding the lectio
difficilior, ‫ עורנו‬and removing the need to emend the verb ‫נכמרו‬.
The main difficulty in the opening clause of the verse is, however, not
the form of ‫נכמרו‬, but its meaning. The verb is in the Niph‘al stem forma-
tion and according to the standard lexica, it means “to grow warm/ hot”
or “to grow excited”.40 Apart from Lam 5:10, it appears only three more
times in the Hebrew Bible: Gen 43:30, 1 Kgs 3:26 and Hos 11:8. In these pas-
sages, the subject of ‫ נכמרו‬is either ‫ רחמים‬or ‫ נחומים‬and the verb refers
to compassion that is “stirred up”. Yet, in Lam 5:10, the precise meaning
of ‫ נכמרו‬is unclear, given that it marks the point of comparison between
human skin and an oven. Furthermore, the compound preposition ‫מ‍פני‬
probably has a causal function in the verse. This implies that the “rages of
hunger” (‫)זלעפות רעב‬, that is, the effects of starvation, are the cause for
33 See J. B. de Rossi, Variae lectiones Veteris Testamenti librorum. Volume III (Parma:
Bodoni, 1788), 246.
34 A. B. Ehrlich, Randglossen zur hebräischen Bibel. Textkritisches, Sprachliches und
Sachliches (Leipzig: J. C. Hinrichs, 1914), 53.
35 W. Rudolph, “Der Text der Klagelieder,” ZAW 56 (1938): 121.
36 Albrektson, Studies in the Text and Theology of the Book of Lamentations, 199–200,
also regards the form ‫ עורינו‬with suspicion on the grounds that the usual plural form of
‫ עור‬is ‫עורות‬. Nevertheless, P. Wernberg-Møller, Review of B. Albrektson, Studies in the
Text and Theology of the Book of Lamentations, JSS 10 (1965): 109–110, indicates that ‫עורינו‬
is “no more remarkable” than the form ‫בנפשינו‬, a variant reading for ‫ בנפשנו‬in Lam 5:9
that is found in some Masoretic manuscripts. He also points out that he unusual plural
form ‫ נפשים‬appears in Ezek 13:20 and suggests that it might only be by accident that the
form ‫ עורים‬is otherwise unattested.
37 C. Westermann, Lamentations. Issues and Interpretation (Minneapolis: Fortress Press,
1994), 210; H.-J. Kraus, Klagelieder (Threni) (BKAT; Neukirchen-Vluyn: Neukirchen Verlag,
1983), 85; J. Dyserinck, “De Klaagliederen uit het Hebreeuwsch opnieuw vertaald,” ThT 26
(1892): 380.
38 M. J. Dahood, “New Readings in Lamentations,” Bib 59 (1978): 194.
39 J. Renkema, Klaagliederen (COT; Kampen: Kok, 1993), 436.
40 F. Brown, S. R. Driver and C. A. Briggs, Hebrew and English Lexicon of the Old Testa-
ment (Oxford: Clarendon Press, 1906), 485; L. Koehler and W. Baumgartner, Lexicon in
Veteris Testamenti Libros; (Leiden: Brill, 1958), 442.
the comparison between human skin and an oven expressed by ‫נכמרו‬.

To judge from the various translations of ‫ נכמרו‬in modern commentaries
and other studies on the book, scholars think that the Hebrew texts of
Lam 5:10 compare the skin of starved human beings with the blackening
of an oven,41 with the heat it emits,42 or with the cracked and flaked tex-
ture of the clay of an oven, caused by the heat of the fire.43
4.1.2. The Greek Translation

LXX τὸ δέρμα ἡμῶν ὡς κλίβανος ἐπελιώθη ἀπὸ προσώπου καταιγίδων λιμοῦ
Our skin has become black like an oven, because of the squalls of hunger.
The Greek manuscripts seem to exhibit a doublet at Lam 5:10. They contain
two equivalents for the verb ‫נכמרו‬, ἐπελιώθη and συνεσπάσθησαν (“they
were shrivelled”).44 In his Handausgabe of the LXX, Rahlfs45 retains both
verbs and punctuates the text of the verse in such a way that ἐπελιώθη
belongs to the first clause with τὸ δέρμα ἡμῶν as its subject, whilst
συνεσπάσθησαν constitutes the verb of a second clause:46 “Our skin has
41 Cf. D. R. Hillers, Lamentations. A New Translation with Introduction and Commentary.

Second, Revised Edition (AB 17; New York: Doubleday, 1992), 158; R. Gordis, The Song of
Songs and Lamentations. A Study, Modern Translation and Commentary (New York: KTAV,
1974), 151.
42 Cf. A. Berlin, Lamentations. A Commentary (OTL; Louisville: Westminster John Knox
Press, 2004), 115; Westermann, Lamentations, 209; Kraus, Klagelieder (Threni), 85; G. C. Aal-
ders, Klaagliederen. Opnieuw uit den Grondtekst Vertaald en Verklaard (KV; Kampen: Kok,
1952), 107; and M. R. H. Löhr, Die Klagelieder des Jeremia (HKAT; Göttingen: Vandenhoeck
& Ruprecht, 1893), 25.
43 Cf. R. B. Salters, Lamentations (ICC; London / New York: T & T Clark, 2010), 355;
P. R. House, Lamentations (WBC; Nashville: Thomas Nelson Publishers), 454; U. Berges,
Klagelieder (HThKAT; Freiburg / Basel / Wien: Herder, 2002), 271; Renkema, Klaaglied-
eren, 437; Albrektson, Studies in the Text and Theology of the Book of Lamentations, 200;
W. Rudolph, Das Buch Ruth. Das Hohe Lied. Die Klagelieder (KAT; Gerd Mohn: Gütersloher
Verlaghaus, 1962), 258; and G. R. Driver, “Hebrew Notes on ‘Song of Songs’ and ‘Lamentati-
ons’,” in Festschrift Alfred Bertholet zum 80. Geburtstag (eds. W. Baumgartner et al.; Tübin-
gen: J. C. B Mohr, 1950), 143.
44 In the Syrohexapla, there is an obelus before ‫ܩܦܕܘ‬, the equivalent of συνεσπάσθησαν,
to indicate that there is no Hebrew equivalent for this word. See A. M. Ceriani (ed.), Codex
Syro-Hexaplaris Ambrosianus photolithographice editus (Milan: Bibliothecae Ambrosianae,
1874), 142 (recto). Regarding the variants for ἐπελιώθη in the minuscules, see the critical
apparatus in Ziegler’s Göttingen edition (Septuaginta, 492) and F. Field, Origenis hexa-
plorum quae supersunt sive veterum interpretum graecorum in totum vetus testamentum
fragmenta (Vol. 2; Oxford: Clarendon Press, 1875), 761.
45 Rahlfs, Septuaginta, 765.
46 In F. Feder’s edition of the Sahidic Coptic version, Biblia Sahidica. Ieremias, Lamenta-
tiones (Threni), Epistula Ieremiae et Baruch (Berlin: Walter de Gruyter, 2002), 216, the word-
ing of this daughter translation reflects the same sentence structure: ⲁⲡⲉⲛϣⲁⲁⲣ ⲕⲙⲟⲙ
become black like an oven; they were shrivelled, because of the squalls of
hunger”.47 Ziegler, however, characterizes the reading in the Greek manu-
scripts as a clear case of double translation and omits συνεσπάσθησαν from
his reconstructed Old Greek text.48
The lectio duplex in the Greek manuscripts can be accounted for in
different ways.49 First, it is possible that the original Greek translator
decided to render ‫ נכמרו‬with two alternative equivalents. The choice for
two translations of ‫נכמרו‬, one with a singular form to match the number
of the preceding subject and the other with a plural form to be faithful
to the form of the Hebrew verb, might have been a way to deal with the
incongruence between the subject and verb in the Hebrew text. This
explanation assumes that the translator’s Hebrew Vorlage was identical
to the consonantal base of Codex Leningradensis. Secondly, there might
have been a doublet in the Hebrew Vorlage, which the Greek translator
rendered faithfully. Robinson50 explains the plus in the Greek text along
these lines by suggesting that συνεσπάσθησαν might have been a transla-
tion of ‫סמרו‬, a doublet of ‫נכמרו‬.51 Without any manuscript evidence to
support such a suggestion, Robinson’s explanation remains moot. In my
opinion, the third explanation for the doublet in the Greek manuscripts
is the most likely one. Ἐπελιώθη, the aorist indicative passive form of
the verb πελιόομαι, “to become livid/pale”,52 is a rare word that appears
only once in the LXX. Driver53 thinks that it means “blackened,” that
is, “scorched,” in LXX Lam 5:10. To explain the rendering of ‫ נכמרו‬with
ἐπελιώθη, he argues that the Greek translator related the Hebrew verb to
the same root as the Syriac words ‫“( ܟܡܝܪܐ‬gloomy”/“dark”/“black”) and
‫“( ܐܬܟܡܪ‬to become sad/darkened”). It might be suggested that a later
ⲛ̅ⲑⲉ ⲛ̅ⲟⲩⲧⲣⲓⲣ ⲁⲩⲥⲟⲕⲟⲩ ⲉⲡⲉⲥⲏⲧ ⲹⲓ ⲟⲩⲥⲟⲡ ⲙ̅ⲡⲉⲙ̅ⲧⲟ ⲉⲃⲟⲗ ⲛ̅ⲛ̅ⲹⲁⲧⲏⲩ ⲙ̅ⲫⲉⲃⲱⲱⲛ (“Our skin
became black like an oven; they brought us down together before the whirlwinds of hun-
ger”).
47 Cf. also the translations of L. C. L. Brenton, The Septuagint with the Apocrypha: Greek
and English (London: Bagster & Sons, 1851), 979.
48 J. Ziegler, Beiträge zur Ieremias-Septuaginta (MSU VI; Göttingen: Vandenhoeck &
Ruprecht, 1958), 99.
49 See Tov, The Text-Critical Use of the Septuagint in Biblical Research, 129.
50 T. H. Robinson, “Notes on the Text of Lamentations,” ZAW 51 (1933): 259.
51 ‫ סמרו‬means “to bristle up” (Brown, Driver and Briggs, Lexicon, 702). In Ps 119:120 it
seems to be used in the sense of flesh that “creeps.”
52 LEH, 364; T. Muraoka, A Greek-English Lexicon of the Septuagint (Louvain / Paris /
Walpole: Peeters, 2009), 543. Assan-Dhôte indicates that πελιόομαι is used by Hippocrates
and Aristotle in the sense “to become livid” (Assan-Dhôte and Moatti-Fine, Baruch, Lam-
entations, Lettre de Jérémie, 280).
53 G. R. Driver, “Notes on the Text of Lamentations,” ZAW 52 (1934): 308.
scribe was unfamiliar with the interpretation of ‫ נכמרו‬exhibited by the

rendering ἐπελιώθη and understood the Hebrew verb in the sense of “to
shrink” or “to be wrinkled,” which the root ‫ כמר‬can have in Aramaic. This
scribe created the reading συνεσπάσθησαν, from the verb συσπάω (“to draw
together”/“to contract”),54 either as a correction of ἐπελιώθη, seeing as it is
closer to the form and perceived meaning of the Hebrew verb, or merely
as an alternative reading. Thus, συνεσπάσθησαν can be explained as a sec-
ondary expansion based on a different understanding of the Hebrew verb
‫ נכמרו‬than the one reflected in the original reading, ἐπελιώθη. Moreover,
the fact that the Vetus Latina only has an equivalent for ἐπελιώθη, livida
facta est (“it was made livid/bluish”) and not one for συνεσπάσθησαν lends
support to the view that the latter was not part of the original Greek trans-
lation.55 In view of the evidence from the Old Latin translation and the
explanation of συνεσπάσθησαν as a secondary reading that found its way
into the wording of the original translation during the process of transmis-
sion, Ziegler’s reconstructed Old Greek text can be used to examine how
the Greek translator dealt with the difficult Hebrew wording of Lam 5:10.
The Greek translation is quite literal when compared to the MT. It fol-
lows the word order of the Hebrew closely and contains a Hebraism, the
rendering of the compound preposition ‫ מפני‬with ἀπὸ προσώπου. This is
a “Hebraistic circumlocution” of the prepositions ἀπό and παρά with geni-
tive.56 Since the preposition ἀπό + genitive can have a causal function and
‫ מפני‬is probably used in this way in the Hebrew wording of the verse, it is
reasonable to assign ἀπὸ προσώπου a causal meaning here.
The idiomatic expression ‫ זלעפות רעב‬is translated as καταιγίδων λιμοῦ.
‫ זלעפה‬is a rare word in the Hebrew Bible. It appears only in Ps 11:6, Ps 119:53
and Lam 5:10. According to the lexicon of Brown, Driver and Briggs,57 its
basic meaning is “raging heat.” The connotation with heat is not reflected
in any of the Greek translations of the Hebrew phrases containing forms
54 LSJ, 1733. In the manuscripts of LXX Lam 5:10, συνεσπάσθησαν probably has the mean-
ing “they were shrivelled.” Cf. G. B. Caird, “Towards a Lexicon of the Septuagint,” in Septu-
agint Lexicography (SCS 1; ed. R. A. Kraft; Missoula: Scholars Press, 1975), 149.
55 The Old Latin translation reads as follows: pellis nostra, ut clibanus livida facta est a
facie plagarum famis (“Our skin was made livid like an oven from the blows of hunger”).
The text is quoted from P. Sabatier’s edition, Bibliorum sacrorum latinae versiones antiq-
uae seu Vetus Italica et caeterae quaecunque in codicibus manuscriptis et antiquorum libris
reperiri potuerunt: quae cum Vulgata Latina, & cum Textu Graeco comparantur (Vol. 2; Remis:
Reginaldum Florentain, 1743), 732.
56 BDF §217. See also R. Sollamo, Renderings of Hebrew Semiprepositions in the Septu-
agint (Helsinki: Suomalainen Tiedeakatemia, 1979), 85–86.
57 Brown, Driver and Briggs, Lexicon, 273.
of ‫זלעפה‬. The Greek equivalent of ‫ זלעפה‬in Ps 118 (119):53 is ἀθυμία

(“despondency”),58 whilst forms of καταιγίς (“squall”/“hurricane”/“storm”)59
serve as renderings for ‫ זלעפות‬in Ps 10 (11):660 and Lam 5:10. With regard
to the latter, the phrase καταιγίδων λιμοῦ conveys the pangs of hunger in
terms of the gusts of a violent wind-storm.
Turning to the Greek translation of the problematic first colon in the
Hebrew text, the incongruence in number between the subject and verb
in the MT is not imitated in the Old Greek text. Both the subject (τὸ δέρμα
ἡμῶν) and the verb (ἐπελιώθη) are singular in the translation. This does
not necessarily presuppose that the Old Greek text was based on a differ-
ent Hebrew Vorlage, seeing as the change of number from the source text
to the target text can readily be explained as a translational adjustment.61
The use of ἐπελιώθη as a translation equivalent for ‫ נכמרו‬implies that the
Greek translator made discolouration the point of comparison between an
oven and the skin of humans who suffer from hunger. Κλίβανος is a dome-
shaped oven made of clay that was used for baking bread.62 Wood and dry
grass were burned in these ovens and their regular use in the preparation
of bread would have caused them to become blackened from the soot of
the smoky fires.63 Concerning the blackening of the skin, the translator
might have had the discolouration of bruised skin in mind, seeing as the
verb πελιόομαι is derived from the adjective πελιός, which describes parts
of the body that are discoloured as a result of extravasated blood.64 It
is noteworthy that a darkened complexion is also brought in connection
with hunger in Lam 4:8. In this passage, both the MT and the Old Greek
text state that the appearance of the Nazirites has become “blacker than
soot” (‫חשך משחור‬/ ἐσκότασεν ὑπὲρ ἀσβόλην). The darkened and emaciated
58 LEH, 10. Cf. A. Rahlfs, Septuaginta. Vetus Testamentum Graecum. Auctoritate

Academiae Scientiarum Gottingensis editum X: Psalmi cum Odis (Göttingen: Vandenhoeck
& Ruprecht, 1979), 292.
59 LEH, 234.
60 Cf. Ralhfs, Psalmi cum Odis, 93.
61 Schäfer, “Lamentations,” 136*.
62 Cf. J. P. Louw and E. A. Nida, Greek-English Lexicon of the New Testament Based on
Semantic Domains (Vol. 1; Cape Town: Bible Society of South Africa, 1993), 92.
63 Cf. F. E. Deist, The Material Culture of the Bible. An Introduction (Sheffield: Sheffield
Academic Press, 2000), 194–195. See also P. J. King and L. E. Stager (Life in Biblical Israel
[Louisville / London: Westminster John Knox Press, 2001], 67) concerning the shape and
function of the clay oven (‫ )תנור‬in Israel.
64 LSJ, 1357. The expression “black and blue” is an English equivalent.
appearance of the Nazirites65 count amongst the dire effects of famine and
starvation, a theme that is developed in vv. 4, 5, 9 and 10 of Lam 4.66
4.1.3. Other Translations
Like the Greek manuscripts, there are two equivalents for the verb ‫ נכמרו‬in
the Peshitta text of Lam 5:10.67 The fact that ‫ ܐܬܩܦܕܘ‬and ‫ ܩܢܐܘ‬have mean-
ings that are similar to συνεσπάσθησαν and ἐπελιώθη raises the question
whether the Syriac text was influenced by the Greek manuscripts. Weitz-
man68 discusses three potential explanations for instances where readings
in the Peshitta agree with those in the LXX over against the correspond-
ing readings in the MT: First, the Peshitta and the Greek translation were
based on Vorlagen that differ from the MT; secondly, the Syriac translator
interpreted the passage in the same way as the Greek translator did (poly-
genesis) or consulted the LXX during the translation process; and finally,
later copyists revised the text of the Peshitta in order to bring in line with
the LXX. With regard to Lam 5:10, Albrektson69 maintains that there is a
clear connection between the Peshitta text and the reading in the Greek
manuscripts. The argument that συνεσπάσθησαν in the Greek manuscripts
is a secondary reading that did not form part of the original Greek transla-
tion rules out the possibility that the Greek and Syriac translations share a
Hebrew Vorlage that contained a doublet. Furthermore, the Peshitta does
not simply conform to the Greek.70 The sentence structures in the Syriac
text and the wording of the Greek manuscripts are different. First, both
65 The Greek translation of the second bicolon of Lam 4:8 indicates that the skin of the
Nazirites cleaves to their bones (ἐπάγη δέρμα αὐτῶν ἐπὶ τὰ ὀστέα αὐτῶν) and that they are
withered and have become like wood (ἐξηράνθησαν ἐγενήθησαν ὥσπερ ξύλον). In the MT, it
is the skin of the ‫ נזירים‬that become dry as wood (‫ָ‍עץ‬ ֵ ‫)יָ ֵבׁש ָהיָ ה כ‬, but in the Old Greek
text, the verbs ἐξηράνθησαν and ἐγενήθησαν are plural and, therefore, probably have ναζιραῖοι
as their subject and not δέρμα αὐτῶν, which is singular. See, in this regard, the comments
of Albrektson, Studies in the Text and Theology of the Book of Lamentations, 183.
66 Cf. Salters, Lamentations, 304; House, Lamentations, 441; Berlin, Lamentations, 108;
Hillers, Lamentations, 158.
67 ‫ܡܫܟܝܢ ܐܝܟ ܕܡܢ ܬܢܘܪܐ ܐܬܩܦܕܘ ܘܩܢܐܘ ܡܢ ܩܕܡ ܛܘܪܦܐ ܕܟܦܢܐ‬ ̈ (“Our skins are shrivelled
as from an oven and they became pale from the exhaustion from hunger”).
68 M. P. Weitzman, The Syriac Version of the Old Testament. An Introduction (Cam-
bridge: Cambridge University Press, 1999), 68–86.
69 Albrektson, Studies in the Text and Theology of the Book of Lamentations, 201.
70 Albrektson (Studies in the Text and Theology of the Book of Lamentations, 200–201)
therefore rightly rejects Abelesz’s view that this passage proves that the Peshitta transla-
tion of Lamentations was revised to conform to the Greek translation.
Syriac verbs are plural and have ‫ܡܫܟܝܢ‬̈ as their subject.71 Conversely, one
of the Greek verbs is singular, whilst the other one is plural. This means
that only the first verb, ἐπελιώθη, can have τὸ δέρμα ἡμῶν as its subject. It
is unclear what the intended subject of συνεσπάσθησαν is. Secondly, the
corresponding verbs are in reversed order in the Syriac translation and the
Greek manuscripts. The first verb in the Peshitta, ‫ܐܬܩܦܕܘ‬, corresponds
to the second verb in die Greek manuscripts, συνεσπάσθησαν, whilst the
second verb in the Syriac translation, ‫ܩܢܐܘ‬, corresponds to ἐπελιώθη, the
initial verb in the text preserved by the Greek manuscripts. This would
suggest that the Syriac translator’s first instinct was to understand the
Hebrew verb as referring to a shrinking or shrivelling action.72 He (or,
maybe, a later copyist) then incorporated the interpretation of ‫ נכמרו‬as
denoting discolouration into the wording of the Syriac translation, pre-
sumably after consulting a Greek manuscript.
The Vulgate translation of the opening colon of Lam 5:10 also differs
from the Old Greek text: pellis nostra quasi clibanus exusta est (“Our skin
is inflamed like an oven”). The rendering of ‫ נכמרו‬with exusta est, the per-
fect indicative passive form of exurere, implies that Jerome saw the hot-
ness of an oven as the point of comparison with the skin of the speakers,
who suffer from the attacks (literally: the storms) of hunger (tempestatum
famis).73
Although it remains uncertain what the precise connotation between
an oven and human skin is that the author of the Hebrew text had in
mind, it is nevertheless clear from the comparison with the Peshitta and
the Vulgate that the Old Greek text of Lam 5:10 presents a novel interpre-
tation of the verse’s difficult Hebrew wording.
71 The plural form of ‫ܡܫܟܝܢ‬̈ might be an attempt by the Syriac translator to eliminate the
discrepancy in number between the subject and verb in his Hebrew Vorlage. Alternatively,
the translator’s source text might have contained the reading ‫עורינו‬.
72 The Syriac translator also employed a form of the verbal root ‫ܩܦܕ‬, “to contract”/“make
to shrink” (J. Payne Smith, ed., A Compendious Syriac Dictionary [Oxford: Oxford University
Press, 1902], 512), to render the Hebrew verb in the clause ‫“( צפד עורם על עצמם‬Their skin
has shriveled on their bones”) in Lam 4:8.
73 Cf. the midrash in Lamentations Rabbah 5:10 §1: “Two teachers comment. One said:
Like a heated mass of grapes. The other said: Like an oven insufficiently heated”. The
translation is quoted from A. Cohen, “Lamentations,” in Midrash Rabbah. Volume 7 (eds
H. Freedman and M. Simon; London: Soncino Press, 1961), 240.
4.2. Lamentations 5:13
4.2.1. The Hebrew Wordings
MT ‫ים ְט ֹ֣חון נָ ָׂ֔שאּו ּונְ ָע ִ ֖רים ָּב ֵ ֥עץ ָּכ ָ ֽׁשלּו׃‬

֙ ‫חּור‬
ִ ‫ַּב‬
Young men took up/bore the mill and youths stumbled because of the
wood.
5QLama ‫נשאו ] [ כשלו‬

֗ ‫[טחון‬
֯ ‫בחורי]ם‬
Young me[n ] took up/bore the mill [ ] stumbled.
The chief difficulty in the Hebrew wordings of Lam 5:13 pertains to the
hapax legomenon ‫טחון‬. Together with the verb ‫נשאו‬, ‫ טחון‬is open to
more than one interpretation. The vocalisation of the word in the MT
gives it a qǐṭâl form. In nouns of this type, the ĭ vowel changes to a vocal
šewâ and the long â vowel in the second syllable is sometimes replaced by
ô.74 Driver75 points out that qǐṭâl-forms can denote concrete objects and
since ‫ טחון‬is related to the verbal root ‫טחן‬, “to grind,” its meaning in the
MT is probably “mill(stone).” Driver also argues that the meaning of the
verse is that young men bare or endure the labour of grinding.76 On this
interpretation, “the mill(stone)” refers to the menial labour of processing
grain, a task usually performed by maidens or slaves.77 Some commenta-
tors render ‫ נשאו טחון‬as “they carry/take up the hand-mill/millstone” and
claim that the work of grinding grain at the mill would be demeaning and
humiliating to the young men.78 Rudolph79 proposes a third solution to
the textual difficulty. He suggests that ‫ טחון‬should be taken as an infini-
tive construct with ‫ בחורים‬as the object of the verb ‫“( נשאו‬They took
young men to grind”). This solution seems unlikely in view of the paral-
lelism exhibited by the two cola of the verse.80 In the second colon, ‫נערים‬
is the subject of the verb ‫ כשלו‬and the same sentence structure applies
74 Cf. GKC §84an.

75 Driver, “Hebrew Notes on ‘Song of Songs’ and ‘Lamentations’,” 143.
76 Ibid., 143–144.
77 See King and Stager, Life in Biblical Israel, 94. By picturing the young men as per-
forming the work of young women or servants, the verse describes the circumstances in
Jerusalem during the Babylonian occupation with the metaphor of a mundus inversus. Cf.
Deist, The Material Culture of the Bible, 262.
78 Cf. Salters, Lamentations, 340, 360; Renkema, Klaagliederen, 439–440; Kraus,
Klagelieder (Threni), 90; Gordis, The Song of Songs and Lamentations, 196.
79 Rudolph, “Der Text der Klagelieder,” 122.
80 Cf. Albrektson, Studies in the Text and Theology of the Book of Lamentations, 202.
to the first colon. Nevertheless, the interpretation of ‫ טחון‬as an infinitive

construct is reflected in the version of Symmachus.81
4.2.2. The Greek Translation
LXX ἐκλεκτοὶ κλαυθμὸν ἀνέλαβον καὶ νεανίσκοι ἐν ξύλῳ ἠσθένησαν

Choice men took up weeping and young men grew weak under wood.
The reading κλαυθμόν in the Greek manuscripts is often branded as a cor-

ruption of a more original translation equivalent.82 Robinson wonders
whether κλαυθμόν is an error for ἀληθόμενον, a participle form of the verb
ἀλήθω (“to grind”).83 However, Rudolph objects that ἀληθόμενον is “ungrie-
chisch” and says that “man wird an κλασμα μυλου (oder μυλικον λιθον?),
den Mühlstein, denken müssen.”84 Concerning these proposals, Ziegler is
of the opinion that κλάσμα μύλου “paßt nicht gut, da es zwei Wörter sind,
und da es nicht als Objekt zu ἀνέλαβον geeignet ist.”85 He also quotes, with
approval, Katz’s evaluation of μυλικὸν λίθον as an “unzeitige Reminizenz
aus dem N.T.”86 Ziegler thinks it is best to adopt Katz’s proposed reading,
μύλον (“hand-mill”), as the original translation of ‫טחון‬, although he notes that
μύλον is not used elsewhere as an equivalent for ‫טחן‬.87 Albrektson rejects
all of these emendations on the grounds that they are graphically too far
removed from κλαυθμόν, the reading found in the Greek manuscripts.88 See-
ing as in the LXX, forms of ‫ טחן‬are rendered by ἀλήθω (Num 11:8, Judg 16:21,
Qoh 12:3), ἀλέω (Isa 47:2) and καταλέω (Exod 32:20, Deut 9:21), Albrektson
81 In the edition of Ceriani, Codex Syro-Hexaplaris, 142 (recto), the reading of Symma-
chus quoted in the margin of the Syrohexapla reads ‫ܝܢܣܩܐ ܠܡܛܚܢ ܢܣܒܘ‬. ̈ According to Field
(Origenis hexaplorum, 761) the Greek version is νεανίσκους εἰς τὸ ἀλέσαι ἔλαβον.
82 Cf. Ziegler, Beiträge, 36; Rudolph, “Der Text der Klagelieder,” 122; Robinson, “Notes
on the Text of Lamentations,” 259. Salters (Lamentations, 359) refers to the suggestion of
Kelso that the Vorlage of the Greek translator might have contained the reading ‫בכי‬. It
is, however, difficult to see how ‫ טחון‬could have developed, through scribal errors, into
‫ בכי‬during the transmission of the Hebrew text or how the Greek translator could have
misread ‫ טחון‬as ‫בכי‬. If ‫ בכי‬was present in the Hebrew Vorlage or only in the mind of the
translator, it was not the result of any kind of scribal or reading error.
83 Robinson, “Notes on the Text of Lamentations,” 259.
84 Rudolph, “Der Text der Klagelieder,” 122.
85 Ziegler, Beiträge, 36.
86 Ibid. Cf. also P. Katz, Review of BH3, Heft 13: Megilloth. TLZ 63 (1938): 34.
87 Ziegler, Beiträge, 37. Accordingly, the Greek wording of the verse’s first clause in the
Göttingen edition is ἐκλεκτοὶ μύλον ἀνέλαβον (Ziegler, Septuaginta, 493). Rahlfs (Septuag-
inta, 765) does not emend the wording and prints the reading found in Codex Vaticanus
and Codex Alexandrinus: ἐκλεκτοὶ κλαυθμὸν ἀνέλαβον.
88 Albrektson, Studies in the Text and Theology of the Book of Lamentations, 202–203.
considers the possibility that a related Greek word served as the original
translation equivalent for ‫ טחון‬in Lam 5:13 and suggests that the Greek
translator might have used ἀλεσμός (“grinding”).89 This is an important
suggestion, because it puts forward a reading that is not only close to the
Greek equivalents of ‫ טחן‬in other passages of the LXX, but also allows for
the creation of κλαυθμόν through scribal errors. In this regard, Albrektson
draws attention to the fact that the interchange of the letters ΑΛ → ΛΑ
and the confusion of С and Θ are common mistakes,90 but he does not
explain why the kappa of κλαυθμόν was added and how the diphthong αυ
was created. A scribe could have confused some of the letters of ἀλεσμόν
during the copying of the Greek text, which resulted in a reading that
did not make sense. In an attempt to rectify the damage, another scribe
changed the corrupt reading into κλαυθμόν so that the passage would have
a new meaning.
Another possible explanation of the reading κλαυθμόν in the Greek
manuscripts is to disregard the assumption that it is a corruption and to
attribute it to the original Greek translator.91 He might have been puz-
zled by the hapax legomenon ‫ טחון‬in his Hebrew Vorlage and produced
what he thought was a meaningful rendering in the context of the verse.
In the next verse, the choice men92 are said to cease from their songs
(ἐκλεκτοὶ ἐκ ψαλμῶν αὐτῶν κατέπαυσαν)93 and the translator might have
89 Ibid., 203. According to the critical apparatus of the Göttingen edition, minuscules
26 and 538 contain additions which include the reading ἀλεσμόν.
90 Albrektson, Studies in the Text and Theology of the Book of Lamentations, 203. See also
N. Fernández Marcos, The Septuagint in Context: Introduction to the Greek Versions of the
Bible (Leiden / Boston / Köln: Brill, 2000), 199.
91 Unfortunately, the Old Latin and Sahidic Coptic daughter versions shed little light on
the original Greek translation of Lam 5:13. In Sabatier’s edition (Bibliorum sacrorum latinae
versiones antiquae, 732) the Vetus Latina reads potentes molis molebant, which is similar
to the Lucianic version: ἐκλεκτοὶ ἐν μυλοις ἤλεσαν (“Choice men grinded in mills”). There
is a conflation of images in the Coptic version, which presents the choice men as sitting,
weeping and grinding at the mill (Feder, Biblia Sahidica, 216): ⲛⲥⲱⲧⲡ̅ ⲁⲩⲹⲙⲟⲟⲥ ⲁⲩⲣⲓⲙⲉ
ⲁⲩⲛⲟⲩⲧ ⲹⲙ̅ ⲡⲙⲁ ⲛ̅ⲛⲟⲩⲧ.
92 It is interesting that the translator translated ‫ בחורים‬in both Lam 5:13 and Lam 5:14
as though it is the passive participle of ‫“( בחר‬to elect”/“to choose”). In Lam 1:15, ‫“( בחורי‬my
young men”) is also rendered by ἐκλεκτούς μου (“my choice men”). However, at Lam 1:18
and Lam 2:21, ‫ בחורי‬is translated as νεανίσκοι μου (“my young men”).
93 This is the wording of the second clause of LXX Lam 5:14 in Rahlfs’ edition. Ziegler
(Septuaginta, 493) deletes κατέπαυσαν from the wording of the Greek text. He thinks that
the first word of v. 15 in the original translation, κατέπαυσε, intruded the wording of v. 14
(Ziegler, Beiträge, 98). In the Greek manuscripts, the opening word of v. 15 is κατέλυσεν
(“it made an end”), but καταλύω is only used for Hiph‘il forms of ‫ שבת‬and not for Qal
forms. Accordingly, Ziegler argues that the first word of the Old Greek text of Lam 5:15 was
κατέπαυσε, which was corrupted into κατέλυσε. The appearance of κατέπαυσαν at the end
considered it appropriate that these men wept instead of making music.

If κλαυθμόν is retained as the reading of the Greek translation, it does not
reproduce the parallelism of the verse in the Hebrew text; rather, it states
that choice men “took up” weeping and young men fainted under (the
weight of) wood.94
4.2.3. Other Translations
The Peshitta text of Lam 5:13 has a free translation of the first colon. It
states that young men grind at the mill: ‫ܓܕܘܕܐ ܛܚܢܘ ܪܚܝܐ‬.̈ The Vulgate
also presents a free translation, but it is quite obscene compared to the
renderings in the Peshitta and the Old Greek. According to this Latin
translation, the enemy “unchastely misused” young men (adulescentibus
inpudice abusi sunt). This implies that ‫טחון‬, in the sense of grinding, was
understood as a euphemism for sexual intercourse. This line of interpre-
tation is also found in Lamentations Rabbah 5:13 §1,95 which refers to a
similar use of ‫ טחן‬in Judg 16:21 where it is said that Samson grinded at
the mill.96
Neither the Peshitta nor the Vulgate show signs of being influenced by
the Greek translation. Moreover, there is no reason to assume that the
Hebrew Vorlagen of these ancient translations differed much from the
of v. 14 was probably the result of dittography of κατέπαυσε in v. 15. The change in number
from singular to plural occurred under the influence of κατέπαυσαν earlier in v. 14. Ziegler
therefore rejects the proposal of Rudolph (“Der Text der Klagelieder,” 122) that the second
κατέπαυσαν in v. 14 is either an unnecessary repetition of the first κατέπαυσαν in the verse
or a second translation of ‫ שבת‬in v. 15, apart from κατέλυσεν. Cf. also Katz, Review, 34; and
idem, “ΚΑΤΑΠΑΥΣΑΙ as a Corruption of ΚΑΤΑΛΥΣΑΙ in the LXX,” JBL 65/3 (1946): 320.
94 The Greek translator of Lamentations also used ἀσθένεω to translate the Hiph‘il form
of ‫ כשל‬in Lam 1:14. He rendered the clause ‫“( הכשיל כחי‬he caused my strength to fail”)
as ἠσθένησεν ἡ ἰσχύς μου (“my strength became weak”). Whereas ‫ כחי‬is the object of the
verb in the Hebrew wording, the translator interpreted it as the subject. In LXX Lam 2:8,
ἀσθένεω also serves as the translation equivalent of the Pu‘lal verb ‫אמללו‬. The bicolon, in
which this verb appears, reads as follows: ‫“( אמללו ויאבל חל וחומה יחדו‬And he caused
rampart and wall to mourn; they languish together”). To judge from the Greek translation
(καὶ ἐπένθησε τὸ προτείχισμα καὶ τεῖχος ὁμοθυμαδὸν ἠσθένησεν), the translator took ‫ חל‬as
the subject of ‫ ויאבל‬and ‫ חומה‬as the subject of ‫אמללו‬. To avoid incongruence in number
between subject and verb, the translation equivalent of ‫אמללו‬, ἠσθένησεν, is singular. Cf.
Schäfer, “Lamentations,” 121*.
95 Cf. A. Cohen, “Lamentations,” 241.
96 Tractate Soṭah 10a, which forms part of Seder Nashim in the Babylonian Talmud,
attributes the same meaning to ‫ טחן‬in connection with Judg 16:21: “R. Joḥanan said: ‘Grind’
means nothing else than [sexual] transgression; and thus it is stated, Then let my wife grind
unto another”. This quotation is from Job 31:10. Cf. A. Cohen, “Soṭah: Translated into English
with Notes, Glossary and Indices,” in The Babylonian Talmud. Seder Nashim (Vol. 3; ed. I.
Epstein; London: The Soncino Press, 1936), 45.
available Hebrew manuscripts. The preceding analysis of the wordings

of the Old Greek translation, the Peshitta and the Vulgate suggests that
each translator gave his own unique interpretation of the difficult opening
clause in the Hebrew text.
5. Conclusions
The goal of the preceding text-critical analyses of LXX Lam 5:10 and 5:13
was to determine how the readings in the Greek wordings of these verses
were created during the process of translation and therefore, to establish
whether the assumption that such analyses can lead to a better under-
standing of the Old Greek text as a translation is justified.
The analyses of the reconstructed Old Greek texts of Lam 5:10 and 5:13
showed that the textual difficulties in the available Hebrew manuscripts
are not reproduced in the Greek translation. Rather, there are differences
between the Hebrew and Greek wordings of Lam 5:10 and 5:13. Concern-
ing the causes of the differences, the analyses indicated that they were
not the result of the Greek translator’s mechanical approach to transla-
tion; rather the differences can be attributed to the Greek translator’s own
unique interpretations of the difficult Hebrew passages. The uniqueness
of the Greek translator’s interpretations of Lam 5:10 and 5:13 is borne out
by the comparisons with the Peshitta and the Vulgate translations. There-
fore, although the translator rendered his Vorlage “literally” (with regard
to factors such as word order, quantitative representation, segmentation
and internal consistency), he also made an attempt to give intelligible
interpretations of the difficulties in the wordings of Lam 5:10 and 5:13.
It should also be pointed out that the Greek translator did not treat
all the problematic passages in his Hebrew Vorlage of Lamentations 5 in
the same way. For example, scholars have found the clause ‫על צוארנו‬
‫“( נרדפנו‬on our necks we are pursued”) in Lam 5:5 and the phrase ‫מפני‬
‫“( חרב המדבר‬because of the sword of the desert”) in Lam 5:9 quite taxing
and many emendations and clarifications have been proposed. Instead
of interpreting these passages, as he did with the textual difficulties in
Lam 5:10 and 5:13, the Greek translator was content with producing the
literal renderings ἐπὶ τὸν τράχηλον ἡμῶν ἐδιώχθημεν97 and ἀπὸ προσώπου
97 ‫ על צוארנו נרדפנו‬is probably an idiomatic expression and J. Joosten (“Translating

the Untranslatable: Septuagint Renderings of Hebrew Idioms,” in “Translation is Required”.
The Septuagint in Retrospect and Prospect [ed. R. J. V. Hiebert; Atlanta: Society of Bibli-
ῥομφαίας τῆς ἐρήμου98 as equivalents for ‫ על צוארנו נרדפנו‬and ‫מפני חרב‬

‫המדבר‬, respectively. The fact that the Greek translator did not render all
the difficult passages in Lamentations 5 in the same manner underscores
the view that he did not translate his Vorlage in a way that is rigid and
mechanical.
In conclusion, this study demonstrates that text-critical analyses of the
Greek renderings of difficult passages in the Hebrew wording of Lamenta-
tions 5 which focus on how the readings in the Old Greek text were cre-
ated during the translation process, on the one hand, and which compare
these readings with their counterparts in other ancient translations, on
the other hand, can indeed shed light on the “translation technique” of
LXX Lamentations. The assumption that formed the point of departure
of this study is, therefore, valid—at least for LXX Lam 5:10 and 5:13. The
data provided by these analyses not only lead to a better understanding of
the Greek wording of Lam 5:10 and 5:13, but can also contribute to a more
nuanced translation profile of LXX Lamentations.
cal Literature, 2010], 59–70) shows that, in the LXX, translators dealt with the difficulties
involved in rendering idiomatic expressions in different ways. In some cases, the transla-
tors rendered the Hebrew idioms literally, thereby sacrificing the clarity of the translation
in favour of fidelity to the form of the source text. In other instances, the translators cap-
tured the meaning of the idiomatic expressions with free translation. Joosten (“Translating
the Untranslatable,” 63) also identifies a third approach followed by the Greek translators
in translating Hebrew idioms; sometimes they combined a free rendering of the idiom’s
meaning with literal rendering of its form. Ἐπὶ τὸν τράχηλον ἡμῶν ἐδιώχθημεν in LXX Lam
5:5 is an example of a literal rendering of a Hebrew idiomatic expression. It is uncertain to
what extent the meaning of the expression would have been clear to a Greek reader, but
it might have been partly comprehensible from its context.
98 Many commentators agree with Kraus’s assessment that the phrase ‫ חרב המדבר‬is
a “verkürzte Redeweise” that has the meaning “das Schwert der Wüstenbewohner” (Kraus,
Klagelieder [Threni], 89). Ἀπὸ προσώπου ῥομφαίας τῆς ἐρήμου in LXX Lam 5:9 can also be
interpreted along these lines. Sollamo (Renderings of Hebrew Semiprepositions, 85) notes
that the referent of ἀπὸ προσώπου is usually a person or a concrete object. In connec-
tion with the prepositional phrase in LXX Lam 5:9, she claims that ῥομφαία refers to both
the weapon itself (concrete object) and its wielder (person). The case of τῆς ἐρήμου can
be taken as a genitive of origin (the sword, and by extension its user, comes from the
desert) or a genitive of place/space (the sword, and its user, is confronted in the desert).
If τῆς ἐρήμου is understood as a genitive of place/space, its meaning would be similar to
the renderings of ‫ חרב המדבר‬in the Peshitta and the Vulgate. In the Syriac translation
(‫)ܚܪܒܐ ܕܒܡܕܒܪܐ‬, the relative pronoun ‫ ܕ‬+ the preposition ‫ܒ‬, followed by the noun ‫ܡܕܒܪܐ‬
(“the sword that is in the desert”), serve to reproduce the Hebrew postconstructus, while
Jerome decided on a prepositional phrase for his Latin translation: gladii in deserto (“the
sword in the desert”).
Amos 5:26—Überlegungen zur Textkritik, Textgeschichte
und Übersetzung eines schwierigen Bibelverses
Eberhard Bons
1. Einleitung
Amos, pastor et rusticus ruborum mora destringens, paucis uerbis explicari

non potest. Dieses Zitat findet sich in einem Brief, den der Bibelüberset-
zer und -kommentator Hieronymus gegen Ende des 4. Jahrhunderts an
den zum Christentum konvertierten und für seine literarische Bildung
bekannten Paulinus von Nola adressiert1. Die Aussagen des Propheten
Amos, auf dessen landwirtschaftliche Tätigkeit Hieronymus hier anspielt
(vgl. Am 7:14), lassen sich—so der Kommentator—nicht mit wenigen
Worten erklären. Dies erläutert er anhand von einigen Beispielen, die sich
offenbar nur mit Mühe deuten lassen: die drei oder vier Verbrechen, derer
insgesamt sieben Völker in den Kapiteln 1–2 beschuldigt werden, die fet-
ten Kühe (Am 4:1), zu denen der Prophet spricht, weiterhin Themen wie
der Schöpfer der Heuschrecke (Am 7:1), der Herr auf der stählernen Mauer
(Am 7:7), schließlich der Obsthaken (Am 8:1 [Vulgata-Text]). Vielleicht
denkt Hieronymus hier auch an eine Stelle wie Am 5:26. Denn wie ein
langer Abschnitt in seinem Kommentar zum Zwölfprophetenbuch zeigt,
ist er sich der Schwierigkeit auch dieses Verses bewusst2. Aus heutiger
Sicht kennt der Masoretentext (= MT) von Am 5:26 keine seltenen oder
gar rätselhaften, ja unübersetzbar erscheinenden Wörter oder Ausdrücke.
Und selbst in den beiden Eigennamen ‫ ִסּכּות‬und ‫ ִּכּיּון‬, die anderswo in
der Hebräischen Bibel nicht mehr vorkommen, erkennt man inzwischen
Bezeichnungen für fremde Götter: Der Name ‫ ִסּכּות‬weist zwar nicht
auf den babylonischen Gott „Sakkut“ (richtig „Sag-kud“) hin, wie lange
angenommen wurde, sondern möglicherweise auf eine andere vorder
orientalische Astralgottheit. Hinter ‫ ִּכּיּון‬wiederum verbirgt sich anschei
nend der nach dem Muster von ‫„ ִׁשּקּוץ‬abscheuliches Kultbild“ vokalisierte
1 Hieronymus, Brief 53, zitiert nach: Hieronymus, Epistulae (CSEL 54; Hg. I. Hilberg;
Wien: Österreichische Akademie der Wissenschaften, 21996), 457.
2 Hieronymus, Commentarii in prophetas minores (CCSL 76; Hg. M. Adriaen; Turnhout:
Brepols, 1969), 296–298.
298 eberhard bons
Gott „Kaiwan,“ der den Planeten Saturn bezeichnet3. Dennoch scheint die
Übersetzung des Verses Am 5:26 nicht unumstritten zu sein. Das bestätigt
auch ein Blick in Bibelübersetzungen neueren Datums. Denn diese geben
nicht einheitlich den MT wieder, vielmehr weichen einige leicht vom MT
ab. Worin scheinen aber die Probleme von Am 5:26 zu bestehen? Und
wie lassen sich auf dem heutigen Kenntnisstand von Textgeschichte und
Übersetzungstechnik diese Probleme beschreiben?
Im Rahmen dieses kurzen Beitrags soll darauf verzichtet werden, die
verschiedenen textkritischen Emendationen sowie einige sehr hypotheti-
sche Übersetzungen zu resümieren und zu kommentieren4. Ebensowenig
wird noch einmal die Frage aufgegriffen, ob der gesamte Vers Am 5:26
späteren Datums ist5 oder ob er mindestens teilweise dem Propheten
Amos selbst zugeschrieben werden kann6. Vielmehr besteht nach wie
vor eine der entscheidenden Fragen darin, welche Gemeinsamkeiten und
Unterschiede die verschiedenen antiken Textzeugen von Am 5:26 aufwei-
sen und wie sich vor allem die Unterschiede zwischen ihnen erklären las-
sen. Von der Beantwortung dieser Frage wird dann abhängen, wie man
den MT von Am 5:26 bewerten kann. Im folgenden Abschnitt (2) wer-
den zunächst die hebräischen Textzeugen des Verses vorgestellt. Sodann
werden die griechischen und lateinischen Übersetzungen von Am 5:26
untersucht (3). Abschließend sollen die Ergebnisse zusammengefasst und
einige Schlussfolgerungen gezogen werden (4).
3 Für beide Götternamen vgl. die Artikel von M. Stol, „Kaiwan“ und „Sakkuth,“ in Dic-
tionary of Deities and Demons in the Bible (Hg. B. Becking et al.; Leiden: Brill, 21999), 478,
722f.; außerdem H. M. Barstad, The Religious Polemics of Amos. Studies in the Preaching of
Am 2,7B-8; 4,1–13; 5,1–27; 6,4–7; 8,14 (VTSup 34; Leiden: Brill, 1984), 122–126.
4 Beispiele bei W. R. Harper, Amos and Hosea (ICC; Edinburgh: T & T Clark, 1905 [repr.
1994]), 139–141; S. M. Paul, A Commentary on the Book of Amos (Hermeneia; Minneapolis:
Fortress Press 1991), 194–196; D. Barthélemy, Critique textuelle de l’Ancien Testament. Tome 3:
Ézéchiel, Daniel et les 12 Prophètes (OBO 51/3; Fribourg / Göttingen : Editions Universitaires /
Vandenhoeck & Ruprecht, 1992), 665.
5 Diese Zweifel äußerte schon J. Wellhausen, Die Kleinen Propheten übersetzt und erklärt
(Berlin: De Gruyter, 41963), 84, aufgrund von inhaltlichen Bedenken. So bemerkt er u.a.,
dass Amos den Israeliten die Verehrung fremder Gottheiten anderswo nicht zum Vorwurf
mache. In den letzten Jahrzehnten erkennen verschiedene Autoren in Am 5:25–26 einen
deuteronomistischen Zusatz, vgl. W. H. Schmidt, „Die deuteronomistische Redaktion des
Amosbuches,“ ZAW 77 (1965): 168–193, bes. 189f., sowie H. W. Wolff, Dodekapropheton 1: Joel
und Amos (BKAT XIV 2; Neukirchen-Vluyn: Neukirchener Verlag, 31985), 137, 310. Anders
Paul (A Commentary on the Book of Amos, 197–198) der Am 5:26 durchaus in die Zeit und
in die Prophetie des Amos einordnen kann.
6 So W. Rudolph, Joel-Amos-Obadja-Jona (KAT XIII 2; Gütersloh: Gerd Mohn, 1971), 208.
amos 5:26—überlegungen zur textkritik 299
2. Der Hebräische Text von Amos 5:26
Was den hebräischen Textbefund angeht, war über Jahrhunderte lediglich

der MT von Am 5:26 bekannt. Dies änderte sich erst, als mit der Entdec-
kung und Entzifferung der Damaskusschrift der wohl bis heute älteste
Textzeuge des Verses in das Blickfeld der Forschung trat. Zwar wird in
der Damaskusschrift der Vers Am 5:26 nicht komplett zitiert, und der
vorhergehende und nachfolgende Text weichen weit vom MT ab. Den-
noch ist bemerkenswert, dass die Reihenfolge der Wörter im MT und in
der Damaskusschrift übereinstimmt und der Konsonantentext in beiden
identisch ist. Dies zeigt sich auch noch an einem Detail: Die einzelnen
Elemente des Bibelzitats werden in der Damaskusschrift erklärt, und in
diesem Zusammenhang wird der Ausdruck ‫ כיון צלמיכם‬leicht verändert
aufgegriffen: ‫כיון הצלמים‬. Außer diesen beiden Zitaten in der Damas-
kusschrift haben sich keine anderen hebräischen Fragmente des Verses
erhalten, auch nicht in der Vielzahl der Handschriften vom Toten Meer7.
Aus der folgenden Tabelle, die nur die in Frage kommenden Wörter
berücksichtigt, werden die Gemeinsamkeiten der Texte ersichtlich:
MT CD A VII, 14–15 CD A VII, 17

‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬
‫ֵ ֚את ִס ּ֣כּות ַמ ְל ְּכ ֶ֔כם‬ ‫את סכות מלככם‬ ‫וכיון הצלמים‬
‫וְ ֵ ֖את ִּכּי֣ ּון ַצ ְל ֵמ ֶיכ֑ם‬ ‫ואת כיון צלמיכם‬
‫יכם‬
ֶ֔ ‫ּכֹוכ ֙ב ֱא ֹ֣ל ֵה‬
ַ
‫יתם ָל ֶ ֽכם׃‬ ֖ ֶ ‫ֲא ֶ ׁ֥שר ֲע ִׂש‬
Aufschlussreich ist noch die Peschitta. Deren Text von Am 5:26 stimmt
im wesentlichen mit dem MT überein, mit Ausnahme davon, dass die
Peschitta ‫ ִסּכּות‬nicht als Eigennamen interpretiert, sondern—ähnlich
wie die Septuaginta (= LXX) und die Vulgata (s.u.)—als Substantiv, näm-
lich als Constructus-Form von ‫ ֻס ָּכה‬. Das folgende hebräische Substantiv,
im MT ‫ ַמ ְל ְּכ ֶכם‬, wird dann als Eigenname „Milkom“ verstanden, so dass
sich der Ausdruck ‫ ܡܫܟܢܗ ܕܡܠܟܘܡ‬ergibt: „Zelt Milkoms.“ Außerdem liest
die Peschitta singularisches ‫„ ܨܠܡܟܘܢ‬dein Götterbild,“ was wohl besser
7 Vgl. B. Ego et al., Hg., Biblia Qumranica. Volume 3b: Minor Prophets (Leiden: Brill,
2005), 58f.; A. Lange, Handbuch der Textfunde vom Toten Meer. Band 1: Die Handschriften
biblischer Bücher von Qumran und den Fundorten (Tübingen: J. C. B. Mohr [Paul Siebeck],
2009), 334.
300 eberhard bons
zum unmittelbar vorausgehenden Götternamen im Singular passt8. Beide

Dokumente—Damaskusschrift und Peschitta—überliefern insofern einen
Text von Am 5:26, der vom MT nicht völlig abweicht. Die Variante „Zelt
Milkoms,“ die die Peschitta bietet, ist lediglich auf eine andere Interpre
tation des Konsonantentextes zurückzuführen und veranlasst zu keiner
Korrektur des MT. Und dennoch scheint der MT von Am 5:26 Kommen-
tatoren und Übersetzer nicht vollständig zufriedenzustellen. Zwar wird
in den letzten Jahrzehnten meist darauf verzichtet, dem Text durch mehr
oder weniger weitreichende Emendationen eine völlig neue Gestalt zu
geben, aber dennoch werden immer wieder Zweifel an der richtigen
Wortfolge laut. Einer der Eingriffe, der den Text nur geringfügig verändert,
orientiert sich an der LXX und besteht in einer leichten Umstellung der
Wörter der zweiten Vershälfte9:
MT Änderungsvorschlag
‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬ ‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬
‫ֵ ֚את ִס ּ֣כּות ַמ ְל ְּכ ֶ֔כם‬ ‫ֵ ֚את ִס ּ֣כּות ַמ ְל ְּכ ֶ֔כם‬
‫וְ ֵ ֖את ִּכּי֣ ּון‬ ‫וְ ֵ ֖את ִּכּי֣ ּון‬
‫ַצ ְל ֵמ ֶיכ֑ם‬ ‫יכם‬ ֶ֔ ‫ּכֹוכ ֙ב ֱאֹל֣ ֵה‬
ַ
‫יכם‬ ֶ֔ ‫ּכֹוכ ֙ב ֱאֹל֣ ֵה‬ַ ‫ַצ ְל ֵמ ֶיכ֑ם‬
‫יתם ָל ֶ ֽכם׃‬
֖ ֶ ‫ֲא ֶ ׁ֥שר ֲע ִׂש‬ ‫יתם ָל ֶ ֽכם׃‬
֖ ֶ ‫ֲא ֶ ׁ֥שר ֲע ִׂש‬
Eine derartige Änderung hätte ohne Zweifel den Vorteil, dass das Wort
‫יכם‬
ֶ ‫ ַצ ְל ֵמ‬nicht auf die zuvor genannten Namen der Götter bezogen würde,
sondern das Bezugswort für den folgenden Relativsatz wäre: Die Israeli-
ten stellen nicht ihre Götter her, sondern die Götterbilder der Gottheiten,
die sie verehren10. Eine solche geringfügige Textänderung hat in verschie
denen neueren Bibelübersetzungen Spuren hinterlassen. So liest etwa
die revidierte Luther-Bibel von 1984: „Ihr trugt den Sakkut, euren König,
und Kewan, den Stern eures Gottes, eure Bilder, welche ihr euch selbst
8 Vgl. A. Gelston, The Twelve Minor Prophets (BHQ 13; Stuttgart: Deutsche Bibelgesell-
schaft, 2010), 84*.
9 Vgl. etwa I. Willi-Plein, Vorformen der Schriftexegese innerhalb des Alten Testaments.
Untersuchungen zum literarischen Werden der auf Amos, Hosea und Micha zurückgehen-
den Bücher im hebräischen Zwölfprophetenbuch (BZAW 123; Berlin / New York: De Gruyter,
1971), 39; C. van Leeuwen, Amos (POT; Nijkerk: Callenbach, 1985), 230.
10 Vergleichbar auch die Lesart des Targums, das im wesentlichen dem MT entspricht,
aber statt ‫יכם‬ֶ ‫ֹלה‬
ֵ ‫ּכֹוכב ֱא‬
ַ die Wörter ‫ּכֹוכב ַצ ְל ֵמיכֹון‬
ַ hat.
gemacht habt.“ Analog die New Revised Standard Version: “You shall take
up Sakkuth your king, and Kaiwan your star-god, your images, which you
made for yourselves”11.
Man wird keineswegs behaupten können, dass die beschriebene Textän-
derung zu beträchtlichen Sinnverschiebungen führe. Dennoch stellt sich
die Frage, ob ein solch punktueller Eingriff zulässig ist, selbst wenn man
ihn mit dem Text der LXX begründet. Denn bevor man einen Textbau-
stein gleichsam aus der LXX herauslöst und ihn dann in den hebräischen
Text einsetzt, muss eine gründliche Untersuchung der betreffenden LXX-
Stelle erfolgen12. Bei genauerer Betrachtung stellt sich nämlich heraus,
dass die LXX dem zitierten Änderungsvorschlag nicht genau entspricht,
sondern sie weicht in mehrfacher Hinsicht vom MT ab. Dieser Frage ist
im nächsten Abschnitt nachzugehen. Dort ist ebenso zu untersuchen, wie
sich die lateinische Textüberlieferung zur griechischen und zur hebräi
schen verhält.
3. Die Griechischen und Lateinischen Texte von Amos 5:26
(a) Der LXX-Text
Auf den ersten Blick fallen die Unterschiede zwischen dem hebräischen
und dem griechischen Text von Am 5:26 auf, vor allem dort, wo die Wort-
folge abweicht. Zur Illustration sei noch die lateinische Übersetzung der
LXX angeführt, die Hieronymus in seinem Kommentar zum Amos-Buch
liefert13. Sie entspricht im wesentlichen dem griechischen Text.
MT LXX Hieronymus, Amos-

Kommentar
‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬ καὶ ἀνελάβετε et assumpsistis
‫ֵ ֚את ִס ּ֣כּות ַמ ְל ְּכ ֶ֔כם‬ τὴν σκηνὴν τοῦ Μολοχ tabernaculum Melchom
‫וְ ֵ ֖את ִּכּי֣ ּון ַצ ְל ֵמ ֶיכ֑ם‬ καὶ τὸ ἄστρον τοῦ θεοῦ ὑμῶν et sidus Dei vestri
‫יכם‬ ֶ֔ ‫ּכֹוכ ֙ב ֱא ֹ֣ל ֵה‬
ַ Ραιφαν, τοὺς τύπους αὐτῶν14 Repham, figuras eorum
‫יתם ָל ֶ ֽכם׃‬ ֖ ֶ ‫ֲא ֶ ׁ֥שר ֲע ִׂש‬ οὓς ἐποιήσατε αὐτοῖς. quas fecistis vobis.
11 Ähnlich auch die „English Standard Version“ von 2007.

12 Vgl. zu diesem Postulat E. Tov, The Text-Critical Use of the Septuagint in Biblical
Research (Jerusalem: Simor, 21997) Kapitel VII.
13 Hieronymus, Commentarii in prophetas minores, 296.
14 Das Pronomen findet sich nur im Codex W.
302 eberhard bons
Einige Unterschiede zwischen MT und LXX sind relativ leicht zu erklä-

ren15. Ähnlich wie die Peschitta (s.o.) setzt die LXX anstelle des Eigen-
namens ‫ ִסּכּות‬eine Constructus-Form von ‫ ֻס ָּכה‬voraus und liest statt des
Substantivs ‫ ַמ ְל ְּכ ֶכם‬den Eigennamen „Moloch“; dabei bleibt das enkliti-
sche Personalpronomen der 2. Pers. plur. unübersetzt16. Sodann scheint
die LXX den Text im Vergleich zum MT umzustellen. Dabei entspricht
griechisches τὸ ἄστρον τοῦ θεοῦ ὑμῶν dem Ausdruck ‫יכם‬ ֶ ‫ֹלה‬
ַ des
MT, und τοὺς τύπους αὐτῶν ist anscheinend die Wiedergabe von ‫יכם‬ ֶ ‫ ַצ ְל ֵמ‬,
wiewohl das hebräische enklitische Personalpronomen der 2. Pers. plur.
mit αὐτῶν übersetzt wird. Betont werden muss allerdings, dass diese
Umstellung nur teilweise derjenigen entspricht, die heutige Kommenta-
toren und Übersetzer vorschlagen (s.o.). Zu erklären bleibt noch Ραιφαν:
Diese ungewöhnliche Form ist wohl auf eine fehlerhafte Lesung zurückzu-
führen. Ob das Wort Ραιφαν schon auf das Stadium der Übersetzung aus
dem Hebräischen zurückgeht oder erst im Prozess der Überlieferung des
LXX-Textes entstanden ist, kann letztlich schwer entschieden werden17.
Jedenfalls weiß schon Hieronymus davon, dass die LXX Repham lese, wäh-
rend der hebräische Text das Wort chion habe—ebenso wie Aquila und
Symmachus18. Wie auch immer, dieser Name begegnet auch im Amos-
Zitat in Apg 7:43, und es ist völlig unklar, ob er mit einer bekannten Gott-
heit assoziiert werden konnte.
Zwei Fragen, die eng miteinander verbunden sind, bleiben an dieser
Stelle noch offen: (1) Wie ist die vom MT abweichende Wortfolge des
LXX-Textes zu erklären? (2) Was bedeuten die Worte τοὺς τύπους αὐτῶν in
ihrem unmittelbaren Kontext? Da beide Fragen in der Sekundärliteratur
nicht ausführlich behandelt werden, ist hier eine genauere Untersuchung
notwendig.
15 Vgl. zur LXX des Amosbuches neben den Kommentaren J. A. Arieti, „The Vocabulary
of Septuagint Amos,“ JBL 93 (1974): 338–347; J. De Waard, „Translation Techniques used
by the Greek Translators of Amos,“ Bib 59 (1978): 339–350; J. M. Dines, The Septuagint of
Amos. A Study in Interpretation (Diss., London, 1992); S. P. Carbone und G. Rizzi, Il libro di
Amos. Lettura ebraica, greca e aramaica (Bologna: Edizioni Dehoniane, 1993); A. W. Park,
The Book of Amos as Composed and Read in Antiquity (Studies in Biblical Literature 37; New
York: Peter Lang, 2001); R. Pierri, Parole del Profeta Amos (SBFA 59; Jerusalem: Franciscan
Printing Press, 2002); A. Gelston, “Some Hebrew misreadings in the Septuagint of Amos,”
VT 52 (2002): 493–500; W. E. Glenny, Finding Meaning in the Text. Translation Technique
and Theology in the Septuagint of Amos (VTSup 126; Leiden: Brill, 2009).
16 Vgl. hierzu C. Dogniez, „ ‚Lost in Translation‘: la désignation des chefs dans le Dode-
kapropheton,“ JSJ 39 (2008): 192–210, bes. 200f. Die Form Melchom, die Hieronymus zitiert,
entspricht der Peschitta-Lesart Milkom.
17 Vgl. hierzu Dines, The Septuagint of Amos, 168.
zu (1): Grundsätzlich kann nicht ausgeschlossen werden, dass der

LXX-Text eine analoge hebräische Vorlage übersetzt, die sich nicht mehr
erhalten hat. Die Übersetzungstechnik des Übersetzers des Zwölfprophe-
tenbuches besteht ja darin, die Wortfolge und Syntax des hebräischen
Textes weitgehend im Griechischen nachzubilden. Dies lässt sich im
Amos-Buch an zahlreichen Phänomenen zeigen, z.B. am pleonastischen
Gebrauch des Personalpronomens in einem Relativsatz (Am 2:9; 4:7) oder
an der redundanten Verwendung des Personalpronomens im Genitiv
unmittelbar nach einem Substantiv (z.B. Am 3:7; 4:9). Dennoch scheinen
zwei Überlegungen dafür zu sprechen, dass die im Vergleich zum MT
andere Wortfolge nicht auf eine entsprechende Vorlage zurückzuführen
ist, sondern auf einen Eingriff des Übersetzers19.
(a) Dem Übersetzer schien offenbar—vielleicht durch Kenntnis anderer

Bibeltexte (4 Kgt 23:10 [Rahlfs])—der Name der Gottheit „Moloch“ ver-
traut zu sein. Aber kannte er auch den zweiten Gott, den der hebräische
Text mit „euer Gott“ bezeichnet und den er mit einem Stern assoziiert?
Angesichts des ungewöhnlichen Namens ist dies nicht sicher. Vielleicht lag
es aus diesem Grund nahe, die parallelen Elemente „Sikkut, euer König //
und Kijjon, eure Bilder“ anders anzuordnen, als dies der hebräische Text
tut. Konkret: Der Übersetzer stellte vor den unbekannten Eigennamen die
Worte τὸ ἄστρον τοῦ θεοῦ ὑμῶν. Auf diese Weise machte er deutlich, dass
„Raiphan“—oder wie auch immer die ursprüngliche Form gelautet haben
mag—ein Gott ist. Somit stehen also in der LXX folgende Elemente par-
allel nebeneinander:
τὴν σκηνὴν τοῦ Μολοχ

καὶ τὸ ἄστρον τοῦ θεοῦ ὑμῶν Ραιφαν
An diese Elemente wird dann wie eine Apposition der Ausdruck τοὺς
τύπους αὐτῶν angefügt, und von ihm hängt der abschließende Relativsatz
οὓς ἐποιήσατε ἑαυτοῖς ab.
19 Diese Meinung vertreten u.a. De Waard, „Translation Techniques Used by the Greek
Translators of Amos,“ 350; Barthélemy, Critique textuelle de l’Ancien Testament, III, 667;
Park, The Book of Amos as Composed and Read in Antiquity, 173; Pierri, Parole del Profeta
Amos, 108f.
304 eberhard bons
(b) Vorausgesetzt dass mit dem Substantiv τύπος ein Abbild—in

diesem Kontext ein Götterbild—gemeint ist (s.u.), konnte die vorhin
beschriebene neue Anordnung der Satzteile zu einem Ergebnis führen,
das durchaus intendiert war. Im hebräischen Bibeltext scheint zwar die
Aussage, dass Menschen „Götter“ in irgendeiner Weise verfertigen, an vie
len Stellen theologisch anstößig zu sein, aber dennoch wird kein Zweifel
daran erhoben, dass Menschen „Götter“ handwerklich herstellen können
oder wollen. Dies zeigt schon ein einschlägiger Text wie Ex 32:1, und auch
der MT von Am 5:26 scheint eine solche Deutung zuzulassen: „eure Göt-
ter, die ihr euch gemacht habt“20. Die LXX scheint dagegen an manchen
Stellen—gerade im Bereich der prophetischen Schriften—zu diffe
renzieren. Im MT von Jes 37:19 // 2 Kön 19:18 wird gesagt, „ihre Götter“
(‫יהם‬ֶ ‫ֹלה‬
ֵ ‫— ֱא‬gemeint sind die Götter der Völker) seien ins Feuer geworfen
worden. Und sogleich wird hinzugefügt, dass diese Götter eigentlich gar
keine Götter, sondern Werke von Menschenhand gewesen seien. Die LXX
von Jes 37:19 sucht diesem Unterschied zwischen Göttern und von Men-
schen hergestellten Kultobjekten Rechnung zu tragen, indem sie—anders
als im Fall von 4 Kgt 19:18—nicht das Wort θεοί verwendet, sondern von τὰ
εἴδωλα αὐτῶν spricht. Diese sind von Menschen angefertigt worden—sie
sind aber keine die Götter. Möglicherweise spiegelt sich im LXX-Text von
Am 5:26 eine ähnliche Differenzierung wider: Die Israeliten stellen höch
stens Abbilder von Göttern her, aber nicht die Götter selbst.
zu (2). Im Griechischen ist das Substantiv ὁ τύπος vieldeutig. Wie ein
Blick in die Wörterbücher zeigt, hat das Wort ein breites Bedeutungsspek-
trum und kann je nach Kontext verschiedene Realitäten bezeichnen, z.B.
die Guss- oder Prägeform, das aus ihr hergestellte Objekt, weiterhin den
Umriss, die Skizze, die durch eine mechanische Einwirkung hinterlasse
nen Spuren usw.21. Von daher kann das Substantiv grundsätzlich sowohl
für ein materielles oder immaterielles Urbild als auch für sein Nachbild
verwendet werden. In der LXX ist das Wort äußerst selten. Es begegnet
in verschiedenen Bedeutungen: „Muster“ (Ex 25:40; MT: ‫) ַת ְבנִ ית‬, „Inhalt“
(eines Briefes, so 3 Makk 3:30), „Beispiel“ (an Gottlosigkeit, 4 Makk 6:19).
Im Sinne von „materielles Abbild“ ist τύπος aber auch außerhalb der LXX
äußerst selten und kommt in der jüdischen Literatur in griechischer
Sprache anscheinend nur noch bei Flavius Josephus, Antiquitates, 15.329,
20 Pierri (Parole del Profeta Amos, 108f.) erwägt noch die Deutung, dass das Bezugswort
des Relativsatzes der zuvor genannte Stern sei.
21 Zahlreiche Beispiele bei C. Spicq, Lexique théologique du Nouveau Testament (Fri-
bourg / Paris: Editions Universitaires / Cerf, 1991), 1524–1527.
vor (ὡς ἀγάλματα καὶ τύπους). Derselbe Autor verwendet das Wort in der
Bedeutung „Götterbild“ noch einmal in den Antiquitates, 1.322: Dort ist
von den Götterbildern die Rede, die Rachel beim Abschied von Laban
unter ihrem Sattel verbirgt (τοὺς τύπους, vgl. Gen 31:36LXX: τὰ εἴδωλα).
Dieser spärliche Befund lässt nur eine Folgerung zu: Im Gegensatz zu τὸ
εἴδωλον kann ὁ τύπος kaum als ein terminus technicus gelten, der Götterbil-
der bezeichnet. Kennt die griechische Literatur paganer Herkunft Stellen,
die zu einem besseren Verständnis von τύπος im Kontext von Am 5:26
beitragen können22? Man kann hierzu zwei Belege bei Herodot anführen,
die von der Herstellung von menschenartigen Abbildern handeln. In den
Historien, II, 86, beschreibt Herodot die Bestattungsbräuche der Ägyp-
ter. In diesem Zusammenhang spricht er davon, dass die Mumien in ein
menschenförmiges Gefäß gelegt werden, das man zuvor angefertigt hat
(ποιεῦνται ξύλινον τύπον ἀνθρωποειδέα). Das bedeutet: der τύπος ist buch-
stäblich menschenförmig, und zwar wohl nach den Körpermaßen des
Verstorbenen aus Holz hergestellt. An einer anderen Stelle, in den Histo
rien, III, 88, erwähnt Herodot das Abbild eines Menschen: Der persische
König Kambyses ließ zu seinen Ehren einen τύπος aus Stein aufstellen
(τύπον ποιησάμενος λίθινον ἔστησε), der ihn als Reiter abbilden sollte. An
beiden Stellen wird somit deutlich, dass ein τύπος ein materielles Abbild
eines Menschen ist, das eigens hergestellt wird (Verb ποιέω). Selbstver
ständlich kann man aus diesen Stellen nicht die Schlussfolgerung ziehen,
dass der Übersetzer des Amos-Buches Herodots Historien gelesen hatte.
Es ist aber nicht unwahrscheinlich, dass er mit einem entsprechenden
Sprachgebrauch vertraut war und das Substantiv τύπος im Sinne von
„materielles Abbild“ kannte, das handwerklich herzustellen war. Wie man
sich diese Abbilder im konkreten Kontext von Am 5:26 vorzustellen hat
und in welcher Weise sie ihre göttlichen „Urbilder“ repräsentieren sollten,
bleibt jedoch im LXX-Text offen.
Zurück zu den vorhin gestellten Fragen: Die bisher gesammelten Beob-
achtungen legen die Folgerung nahe, dass der LXX-Text von Am 5:26
wahrscheinlich keine Wort-für-Wort-Übersetzung ist, sondern das Ergeb-
nis einer Interpretation darstellt. Diese besteht darin, dass der Überset-
zer die Elemente des Textes neu anordnet. Außerdem führt er mit dem
Begriff τύπος einen Terminus ein, der in der LXX ohne Parallele ist und
wohl das materielle Abbild bezeichnet. Dies ist um so bemerkenswerter,
22 C. D. Isbell, „Another Look at Amos 5:26,“ JBL 97 (1978): 97–99, bemerkt zwar, dass
das Wort τύπος in Am 5:26 ungewöhnlich ist, belässt es aber bei dieser Feststellung.
306 eberhard bons
als gerade das Zwölfprophetenbuch mit dem Wort τὸ εἴδωλον ein einschlä-
giges Substantiv für „Götterbilder“ verwendet (z.B. Hos 8:4; Mi 1:7; Hab
2:17). Jedenfalls ist eindeutig, dass nach dem LXX-Text die Israeliten keine
Götter herstellen, sondern lediglich deren Abbilder.
(b) Der Vulgata-Text
Zuletzt stellt sich noch die Frage, ob der Vulgata-Text zu weiteren Erkennt-
nissen über die Textgeschichte von Am 5:26 führt. Stellt man die Texte
synoptisch nebeneinander, ergibt sich folgendes Bild:
MT Vulgata LXX
‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬ et portastis καὶ ἀνελάβετε
‫ֵ ֚את ִס ּ֣כּות ַמ ְל ְּכ ֶ֔כם‬ tabernaculum Moloch vestro τὴν σκηνὴν τοῦ Μολοχ
‫וְ ֵ ֖את ִּכּי֣ ּון ַצ ְל ֵמ ֶיכ֑ם‬ et imaginem idolorum vestrorum καὶ τὸ ἄστρον τοῦ θεοῦ ὑμῶν
‫יכם‬ ֶ֔ ‫ּכֹוכ ֙ב ֱא ֹ֣ל ֵה‬
ַ sidus dei vestri Ραιφαν, τοὺς τύπους αὐτῶν
‫יתם ָל ֶ ֽכם׃‬ ֖ ֶ ‫ֲא ֶ ׁ֥שר ֲע ִׂש‬ quae fecistis vobis οὓς ἐποιήσατε ἑαυτοῖς
Zunächst fällt auf, dass Hieronymus den Text—im Gegensatz zur LXX—
nicht anders anordnet. Dies bedeutet: Er hat wahrscheinlich eine hebrä-
ische Vorlage benutzt, die sich von demjenigen hebräischen Text nicht
wesentlich unterscheidet, den der MT und (wenigstens in Teilen) die
Damaskusschrift überliefern. Und trotzdem entfernt auch Hieronymus
sich vom hebräischen Text—trotz seiner grundsätzlichen Präferenz für
die hebraica veritas23. Eine wichtige Autorität ist für ihn dabei das neu-
testamentliche Zitat von Am 5:25–27, das Stephanus in seine Rede ein-
baut (Apg 7:42f.) und das im wesentlichen dem Text der LXX entspricht.
Hieronymus argumentiert nun, dass der erste Märtyrer nicht geirrt haben
könne (nec putandus est primus martyr errasse)24, ja dass die neutestament-
lichen Autoren die alttestamentlichen Schriften nicht nach dem Wortlaut
zitieren, sondern entsprechend ihrem Sinn25, und aus diesen Gründen
23 Vgl. hierzu Hieronymus, Brief 106, 2 zitiert nach: Hieronymus, Epistulae (CSEL 55; Hg.
I. Hilberg; Wien / Leipzig: Tempsky / Freytag, 1912), 239: Sicut autem in nouo testamento, si
quando recurrimus ad fontem Graeci sermonis [...], ita in veteri testamento, si quando inter
Graecos Latinosque diversitas est, ad Hebraicam confugimus veritatem, ut, quicquid de fonte
proficiscitur, hoc quaeramus in rivulis.
25 Ibid., 297.
gibt er in Am 5:26 der LXX den Vorzug. Jedoch ist ihm der hebräische Text
keineswegs unbekannt. Zwar weiß er offenbar nichts von der Gottheit,
die sich hinter dem Namen ‫ ִסּכּות‬verbirgt. Er zitiert aber das folgende
Wort, das im hebräischen Text melchekem laute, und ist auch über dessen
Wiedergabe mit regis vestri durch Theodotion und Aquila informiert26.
Offen bleibt nur, wodurch Hieronymus’ Übersetzung imaginem für ‫ִּכּיּון‬
zustandekommt. Da imago und idolum für ihn gleichbedeutend sind27, ist
imaginem vielleicht als eine gewisse Verlegenheitsübersetzung des schwer
verständlichen ‫ ִּכּיּון‬anzusehen, mit der Hieronymus ein Äquivalent zum
ebenfalls singularischen tabernaculum schaffen wollte.
Somit bleibt als Ergebnis festzuhalten, dass Hieronymus zwar von eini-
gen gewichtigen Unterschieden zwischen der LXX und dem hebräischem
Text weiß. Trotzdem hält er aber wegen der Autorität des neutestament-
lichen Schriftzitates an einem Text fest, der der LXX nahesteht. Jedoch
übernimmt Hieronymus nicht die Wortfolge der LXX, sondern orientiert
sich am hebräischen Text. Daher erlaubt der Vulgata-Text Rückschlüsse
auf seine ihm zugrunde liegende hebräische Vorlage, aber nur in einem
begrenzten Maß.
4. Abschliesende Bemerkungen
Aus dem Vergleich der verschiedenen Textzeugnissen von Am 5:26 kann

man die Folgerung ableiten, dass sich für den Wortlaut, der im MT über-
liefert ist, antike Zeugen finden, und zwar CD A VII,14–15, 17, die Peschitta,
ferner die Vulgata, wenn man von ihren durch Apg 7:43 beeinflussten Ele-
menten absieht. Der MT ist keineswegs unübersetzbar. Wenn ‫יכם‬ ֶ ‫ ַצ ְל ֵמ‬die
beiden vorher genannten Gottheiten qualifiziert, könnte man in diesem
Wort das Bezugswort des abschließenden Relativsatzes erkennen. Solche
Bilder kann man „machen“ (Verb ‫עׂשה‬, vgl. Ez 7:20; 16:17). Der Ausdruck
‫יכם‬ ֶ ‫ֹלה‬
ַ wäre dann wohl ebenfalls vom einleitenden Verb ‫אתם‬ ֶ֗ ‫ּונְ ָׂש‬
abhängig. Eine andere Möglichkeit besteht darin, entweder nur das Wort
‫ּכֹוכב‬
ַ oder aber den gesamten Text von ‫יכם‬ ֶ ‫ֹלה‬
ַ bis zum abschlie
ßenden Relativsatz als eine Parenthese aufzufassen. Diese Rekonstruktion
der Syntax führt dann zu der Übersetzung: „ . . . und Kijjon, eure Bilder
(ein Stern), eure Götter, die ihr euch gemacht habt“28 oder „ . . . und Kijjon,
26 Ibid., 297.
27 Ibid., 296.
28 Vgl. J. A. Soggin, Il profeta Amos (Studi Biblici 61; Brescia: Paideia, 1982), 131.
308 eberhard bons
eure Bilder (den Stern eures Gottes, den ihr euch gemacht habt)“29. Wie
auch immer man sich entscheiden mag: Der MT ist infolge seiner Syntax
uneindeutig und schwer verständlich30. Ob die Wortfolge, die er bietet,
fehlerhaft ist, kann derzeit nicht entschieden werden, da abweichende
hebräische Textzeugnisse nicht verfügbar sind. Aufgrund der vorhin
durchgeführten Untersuchung des LXX-Textes von Am 5:26 erweist sich
jedoch eine Aussage als wahrscheinlich: Der griechische Text besteht in
einer anderen Anordnung der Textelemente, die dem Text eine größere
Eindeutigkeit verleiht31.
Zuletzt sei noch Folgendes bemerkt: Wenn heutige Bibelübersetzungen
sich in Am 5:26 nicht am MT orientieren, sondern den Text umstellen,
wie in Abschnitt 2 gezeigt, dann ist dieses wenigstens in einer Fußnote zu
kennzeichnen. Aus dieser sollte hervorgehen, dass nicht der MT in seiner
überlieferten Gestalt übersetzt wird, sondern dass der Übersetzer die Rei-
henfolge der Wörter geringfügig verändert hat.
29 So etwa Barthélemy, Critique textuelle de l’Ancien Testament, III, 667.

30 Vgl. F. I. Andersen und D. N. Freedman, Amos (AB 24A; New York: Doubleday, 1989),
533.
31 Vgl. Dines, The Septuagint of Amos, 167: „It is widely acknowledged that the order of
LXX is more balanced and logical than that of MT.“
Part Four
Reception
Die Schriftzitate im ersten Christentum und die
Textgeschichte der Septuaginta:
Ein Wuppertaler Forschungsprojekt
Martin Karrer und Johannes de Vries
1. Textforschung und Schriftzitate
Die biblische Textforschung erhielt durch die Quellenfunde und Papyrus

editionen des 20. Jh. wesentliche neue Impulse. Der protomasoretische
Text hatte, wie sich zeigte, in der Zeit vor und um die Entstehung des
ersten Christentums besonderes Gewicht, herrschte aber hebräisch noch
nicht alleine vor. Nicht minder Bewegung kam in die Betrachtung der
griechischen Überlieferung. Die kritische Edition der Septuaginta wurde
durch neue Papyri und eine Aufwertung des sog. antiochenischen Textes,
der lange Zeit gering geachtet wurde, belebt1. Darüber hinaus ließen sich
durch den hebräischen Text angeregte Revisionsprozesse (kaige und Ver
wandtes) schon vor der Zeitenwende nachweisen. Viele Wurzeln für Text
formen, die lange Zeit als jung galten (sog. jüngere Übersetzungen), fallen
nach heutiger Kenntnis noch vor das Ende des Neuen Testaments. Das
von der Forschung vor allem gesuchte Old Greek der Septuaginta tritt des
halb in eine größere Vielfalt der Überlieferung, die nach Würdigung ruft2.
Dieser Wandel wirkt sich auf die Betrachtung der Zitate aus Israels
Schriften aus. Herkömmlich spielten sie in der Textforschung eine geringe
Rolle3. Denn Abweichungen der Zitate vom zitierten Text schienen durch
1 N. Fernández Marcos und J. R. Busto Saiz, Hg., Biblia: El texto antioqueno de la biblia
griega (3 Bde.; Madrid: CSIC, 1989/1992/1996); Weiteres bei S. Kreuzer „B or not B? The
Place of Codex Vaticanus in Textual History and in Septuagint Research,“ in vorliegendem
Band.
2 Für Übersichten sei auf A. Lange, „From Literature to Scripture: The Unity and
Plurality of the Hebrew Scriptures in Light of the Qumran Library,“ in One Scripture or
Many? Canon from Biblical, Theological, and Philosophical Perspectives (Hg. C. Helmer und
C. Landmesser; Oxford: Oxford University Press, 2004), 51–107 und E. Tov, Textual Criticism
of the Hebrew Bible (Minneapolis: Fortress Press, 32011), 127–147 verwiesen.
3 Obwohl der Vorgang des Zitierens lange vor dem frühen Christentum begann: s. die
inneralttestamentlichen Zitate 2Kön (LXX 4Kgt) 14:6 und 2Chr 25:4 nach Dtn 24:16 und Jer
26(LXX 33):18 nach Mi 3:12. Der im Ketef Hinnom gefundene Priestersegen, eine Variante
von Num 6:24–26 (wohl 7./6. Jh.), könnte älter sein als der verschriftlichte Pentateuch
(D. M. Salzer, Die Magie der Anspielung: Form und Funktion der biblischen Anspielungen in
312 martin karrer und johannes de vries
redaktionelle Eigenheiten (Erinnerungslücken, Integration in einen

neuen Kontext, theologische Adaptionen) weithin erklärbar. Als nicht
minder einfach zu erklären galt der Sachverhalt, wenn sich Textformen
von Neuem Testament und Septuaginta in einem Teil der Überlieferung
unerwartet näherten: Die Editoren vermuteten in solchen Fällen nach
trägliche Einflüsse aus dem jeweils anderen Feld.
A. Rahlfs führte das Siglum „ex“ in seine Handausgabe der Septuaginta ein,
um anzugeben, wo eine Textfassung „aus“ neutestamentlichem Einfluss
stamme, „cf.“, wo das nahe liege, und =, wo zumindest eine Übereinstim
mung vorliege4, und sah die Abhängigkeit der Überlieferung vom Neuen
Testament keinesfalls selten gegeben5. Vice versa formulierten neutesta
mentliche Textforscher die Faustregel, Entsprechungen von ntl. Varianten
zur Septuaginta (OG oder spätere Varianten) gingen gerne auf eine nach
trägliche Anpassung an die LXX zurück (wenn auch keineswegs immer)6.
Dieses Bild geriet durch die Breite und Vielfalt der antiken Schriftüber
lieferung ins Wanken. Robert Kraft stellte den alten Konsens 1978 in
Frage, und Robert Hanhart wagte 1981 die Gegenthese gegen die ältere
Forschung, „daß die neutestamentlichen Schriften in ihren alttestamentli
chen Zitaten, auch dort, wo sie vom alten LXX-Text abweichen, auf vorge
gebener jüdischer Überlieferung beruhen“7. Seither steht eine Neuprüfung
des textkritischen Befundes an.
den magischen Texten der Kairoer Geniza Biblia magica [TSAJ 134; Tübingen: Mohr Siebeck,
2010], 87), wäre dann aber kein Schriftzitat.
4 Z.B. bei PsLXX 13:3: s. A. Rahlfs, Der Text des Septuaginta-Psalters, nebst einem Anhang:
Griechische Psalterfragmente aus Oberägypten nach Abschriften von W.E. Crum (SeptSt 2;
Göttingen: Vandenhoeck & Ruprecht, 21965), insb. 227, 325, 327, und idem, Hg., Psalmi cum
Odis (Septuaginta: Vetus Testamentum Graecum 10; Göttingen: Vandenhoeck & Ruprecht,
1931), 30f., 96: „ex Rom. 3,13–18, ubi Paulus haec uerba [. . .] cum Ps. 13,3 iunxit.“
5 Seine bekanntesten Beispiele sind PsLXX 13:3 (auch laut Apparat in der Handausgabe
[A. Rahlfs und R. Hanhart, Hg., Septuaginta: Id Est Vetus Testamentum Graece Iuxta 70
Interpretes (Stuttgart: Deutsche Bibelgesellschaft, 22006), II 11] Langtext „ex Rom. 3,13–18,
ubi Paulus haec uerba [. . .] cum Ps 13,3 iunxit”) und PsLXX 39:7, wo er sich gegen alle grie
chischen Handschriften für ὠτία (nach Ga, dem hebräischen Text entsprechend) entschied,
weil er die griechischen Handschriften zu Hebr 10:5 verglich. Zu diesen Beispielen vgl.
unten Anm. 79.
6 B. Aland und K. Aland, Der Text des Neuen Testaments (Stuttgart: Deutsche Bibel
gesellschaft, 21989a), 285; idem, The Text of the New Testament (Grand Rapids / Leiden:
Eerdmans / Brill, 21989b), 281.
7 R. Hanhart, „Das Neue Testament und die griechische Überlieferung des Judentums,“
in Überlieferungsgeschichtliche Untersuchungen (TU 125; Hg. F. Paschke; Berlin: Akademie-
Verlag, 1981), 293–303, hier 296. Vgl. zuvor R. A. Kraft, „Christian Transmission of Greek
Jewish Scriptures. A Methological Probe,“ in Paganisme, Judaïsme, Christianisme. Influen-
ces et affrontements dans le monde antique (FS Marcel Simon; Hg. A. Bénoit; Paris: Boccard,
1978), 207–226.
die schriftzitate im ersten christentum 313
Allerdings muss diese Neuprüfung ein nicht unwichtiges Hindernis über

winden. Die Textforschung zu Septuaginta und Neuem Testament sowie
den außerbiblischen frühjüdischen und altkirchlichen Quellen erfolgt—
verständlich aufgrund der Fülle des Materials—weitgehend getrennt. Ein
Austausch über die Bezeugung der zitierten Texte/Zitate wurde zwischen
den bekanntesten Textforschungsstellen, dem Septuaginta-Unternehmen
der Akademie der Wissenschaften zu Göttingen und dem Institut für
Neutestamentliche Textforschung Münster, früher nicht institutionali
siert. Die materiale Basis ist deshalb über bisherige Datenbanken hinaus
zu erschließen, um sichere Entscheidungen fällen zu können. Die inzwi
schen vorzüglich zugänglichen Handschrifteneditionen erlauben das.
2. Schriftzitate vor dem frühen Christentum
Unterschiedliche Zugänge können sich ergänzen, um die Zitate genauer

in die Textgeschichte einzuordnen. Zum einen lässt sich die Hebräische
Bibel zum Ausgangspunkt nehmen und das Spektrum der Zitationen aus
ihr suchen8. Das Wiener, von Armin Lange unter Mitarbeit von Matthias
Weigold geleitete Forschungsprojekt „The Meaning of Ancient Jewish Quo
tations and Allusions for the Textual History of the Hebrew Bible,“ dessen
Dokumentation gerade veröffentlicht wurde, setzt hier an. Es sammelt die
Zitate und Anspielungen aus der Hebräischen Bibel in der jüdischen Lite
ratur bis 70 n.Chr., unterstützt durch ein elektronisches Suchprogramm.
Dabei stellt es Zitationsformeln zurück und benützt eine offene Definition:
Alle Referenzen werden in die Sammlung aufgenommen, die eindeutig
8 Seit den 1950er Jahren beobachtet die Forschung Zitate in den Qumrantexten und
der jüdisch-hellenistischen Literatur: vgl. bes. J. Carmignac, „Les citations de l’Ancien Tes
tament dans la Guerre des Fils de Lumière contre les Fils de Ténèbres,“ RB 63 (1956):
234–260, 375–390; und idem, „Les citations de l’Ancient Testament, et spécialement des
Poèmes du Serviteur, dans les Hymnes de Qumrân,“ RevQ 2 (1960): 357–394; S. Delamarter,
A Scripture Index to Charlesworth’s Old Testament Pseudepigrapha (London: Sheffield
Academic Press, 2002); J. A. Fitzmyer, „The Use of Explicit Old Testament Quotations in
Qumran Literature and in the New Testament,“ NTS 7 (1960–61): 297–333. Nachdruck in
J. A. Fitzmyer, Hg., Essays on the Semitic Background of the New Testament (London:
Geoffrey Chapman, 1971), 3–58; A. Lange, Handbuch der Textfunde vom Toten Meer, Band
1: Die Handschriften biblischer Bücher von Qumran und den anderen Fundorten (Tübingen:
Mohr Siebeck, 2009); J. Maier, Die Qumran-Essener: Die Texte vom Toten Meer, Band 3:
Einführung, Zeitrechnung, Register und Bibliographie (München: Reinhart, 1996), 161–182;
F. J. Morrow, The Text of Isaiah at Qumran (Diss., Washington, 1973); J. C. VanderKam und
P. W. Flint, The Meaning of the Dead Sea Scrolls: Their Significance for Understanding the
Bible, Judaism, Jesus, and Christianity (San Francisco: HarperCollins, 2002), 427–433.
erkennbar sind. Das ermöglicht dankenswert, die Rezeptionsgeschichte

der Hebräischen Bibel umfassend zu erhellen.
Für die Textgeschichte im engeren Sinn ist der Ausschnitt der Samm
lung relevant, in dem aus den Referenzen Schriftzeugen werden9. Außer
dem muss sich eine textgeschichtliche Untersuchung für im Wiener
Verzeichnis nicht erfasste Quellen öffnen. Philo etwa schrieb vor 70 n.Chr.
Doch die Textüberlieferung ist weit jünger, und ein gewichtiger Teil der
Forschung10 vermutet für einen Teil der Überlieferung, ihre Zitate11 seien
unter Einfluss des Aquila-Textes spätantik nachträglich redigiert (so
zuletzt R. A. Kraft 2005)12. Die Wiener Dokumentation nimmt die Philo-
Zitate ebenso wie das Neue Testament nicht auf.
Zum anderen kann die Forschung den Gegenstand begrenzen und bei
den Zitaten im engeren Sinn sowie bei einer spezifischen Schriftgestalt
ansetzen, neben dem hebräischen Text besonders der Septuaginta. Das
ist der Schwerpunkt des hier vorzustellenden Wuppertaler Forschungs-
projektes.
Wie verhalten sich Zitat und zitierter Text? Werfen wir dazu einen
Blick auf die Stellen, die bereits in der Hebräischen Bibel und Septuaginta
durch Zitateinleitungen markiert sind13.
9 Die Veröffentlichung—A. Lange und M. Weigold, Biblical Quotations and Allusions

in Second Temple Jewish Literature (Journal of Ancient Judaism. Supplements 5; Göttingen:
Vandenhoeck & Ruprecht, 2011)—erfolgte nach Abschluss des vorliegenden Beitrags zum
Vortrag in Stellenbosch. Bemerkenswert ist die Reflexion von Intertextualitätstheorie und
die Typologie von „Citationality” (ibid., 19ff.). Vgl. auch die Homepage des Wiener Instituts
für Judaistik (www.univie.ac.at/judaistik/, abgerufen am 16/6/2011).
10 Zur älteren Forschungsgeschichte s. P. Katz, Philo’s Bible: The Aberrant Text of Bible
Quotations in Some Philonic Writings and Its Place in the Textual History of the Greek Bible
(Cambridge: Cambridge University Press, 1950) und D. Barthélemy, „Est-ce Hoshaya Rabba
qui censura le ‘Commentaire allégorique’? À partir des retouches faites aux citations bibli
ques, étude sur la tradition textuelle du Commentaire Allégorique de Philon,“ in Études
d’histoire du Texte de l’Ancien Testament (OBO 21; Hg. idem; Göttingen: Vandenhoeck &
Ruprecht, 1978), 140–173, 390–391.
11 Aufgelistet in Philon d’Alexandrie (BiPa.Supp; Hg. J. Allenbach et al.; Paris: Centre
national de la recherche scientifique, 1982), nicht ganz vollständig, doch für unsere Zwecke
zureichend.
12 R. A. Kraft, „Philo’s Bible Revisited: The ‘Aberrant Texts’ and Their Quotations of
Moses,“ in Interpreting Translation: Studies on the LXX and Ezekiel in Honour of J. Lust
(BETL 192; Hg. F. García Martínez und M. Vervenne; Leuven: Peeters, 2005), 237–253.
13 Vgl. zu den Schriftzitaten innerhalb des AT W. Oswald, „Zitat (AT),“ s.v. „Zitat,“ in Das
wissenschaftliche Bibellexikon im Internet (WiBiLex; Hg. M. Bauks und K. Koenen), abgeru
fen am 22/8/2011. Online: www.wibilex.de, insb. 2.2.1.
Tabelle 1 Dtn 24:16 und 2Kön (LXX 4Kgt) 14:6/2Chr 25:4

Hebräischer Text OG nach RaHa Antiochenischer Text
Dtn 24:16 ‫א־יּומתּו ָאבֹות‬
ְ ֹ ‫ ל‬Οὐκ ἀποθανοῦνται
πατέρες
‫ל־ּבנִ ים‬
ָ ‫ ַע‬ὑπὲρ τέκνων,
‫א־יּומתּו‬
ְ ֹ ‫ּובנִ ים ל‬ָ καὶ υἱοὶ οὐκ
ἀποθανοῦνται
‫ל־אבֹות‬
ָ ‫ ַע‬ὑπὲρ πατέρων14
Nicht werden Väter Nicht werden sterben
um der Söhne Väter für Kinder, und
willen getötet, und Söhne werden nicht
Söhne werden nicht sterben für Väter.
getötet um der
Väter willen.
2Kön 14:6 ‫ ַּכ ָּכתּוב‬γέγραπται γέγραπται
(4Kgt 14:6) ]. . .[ [. . .] [. . .]
‫א־יּומתּו ָאבֹות‬
ְ ֹ ‫ ל‬Οὐκ ἀποθανοῦνται Οὐκ ἀποθανοῦνται
πατέρες πατέρες
ָ ‫ ַע‬ὑπὲρ υἱῶν ὑπὲρ τέκνῶν,
‫א־יּומתּו‬
ְ ֹ ‫ּובנִ ים ל‬ ָ καὶ υἱοὶ οὐκ οὐδὲ υἱοὶ
ἀποθανοῦνται ἀποθανοῦνται
‫ל־אבֹות‬ָ ‫ ַע‬ὑπὲρ πατέρων ὑπὲρ πατέρων15
Nicht werden sterben Nicht werden sterben
Väter für Söhne, und Väter für Kinder, auch
Söhne werden nicht werden nicht Söhne
sterben für Väter. sterben für Väter.
2Chr 25:4 ‫ ַכ ָּכתּוב‬γέγραπται γέγραπται
]. . .[ [. . .] [. . .]
‫ לֹא־יָמּותּו ָאבֹות‬Οὐκ ἀποθανοῦνται Οὐκ ἀποθανοῦνται
πατέρες πατέρες
ָ ‫ ַע‬ὑπὲρ τέκνων, ὑπὲρ τέκνων,
‫ּובנִ ים לֹא־יָמּותּו‬ָ καὶ υἱοὶ οὐκ καὶ υἱοὶ οὐκ
ἀποθανοῦνται ἀποθανοῦνται
‫ל־אבֹות‬ָ ‫ ַע‬ὑπὲρ πατέρων ὑπὲρ πατέρων16
Nicht werden Nicht werden sterben Nicht werden sterben
sterben Väter um Väter für Kinder, und Väter für Kinder, und
der Söhne willen, Söhne werden nicht Söhne werden nicht
und Söhne werden sterben für Väter. sterben für Väter.
nicht sterben um
der Väter willen.
14 Gö stimmt damit überein.

15 Fernández Marcos und Busto Saiz, Biblia, 122f. z.St.
16 Wiedergabe nach Fernández Marcos und Busto Saiz, Biblia, 126 z.St.
2Kön 14:6 zitiert nach eigener Angabe, was geschrieben ist im Buch der
Tora/des Gesetzes (Tabelle 1). Der hebräische Text von Vorlage (Dtn 24:16)
und Zitat stimmt völlig überein. Anders ist dies aber beim parallelen Zitat
in 2Chr 25:4; dort steht statt des Hof ’al (MT ‫יּומתּו‬
ְ ) das Qal (MT ‫;)יָ מּותּו‬
schon die hebräische Zitatgeschichte lässt eine stilistische Variation zu.
Die griechische Übersetzung gewährt der Überlieferung einen zusätz
lichen stilistischen Freiraum. Der antiochenische Text (Ant.) von 4Kgt 14:6
variiert die Wiedergabe von ‫( בנים‬τέκνα/υἱοί) und wählt οὐδέ für die Ver
neinung. Die von Rahlfs bevorzugten Handschriften dagegen bieten die
konsistente Wiedergabe von ‫ בנים‬mit υἱοί und die Verneinung καὶ οὐκ.
Beides weicht vom heute rekonstruierten Old Greek des Dtn leicht ab (Ant.
durch οὐδέ, RaHa [die Edition von Rahlfs/Hanhart] durch ὑπὲρ υἱῶν).
Blicken wir auf 2Chr 25:4, stimmen Rahlfs-Text und Ant. dort dagegen
mit LXX Dtn 24:16 überein. Die Übersetzer könnten auf Gleichheit zu dem
in ihrer Zeit maßgeblichen Text des Gesetzes geachtet haben. Das aber
wirkt sich nicht auf 4Kgt 14:6 aus; 2Chr 25:4 und 4Kgt 14:6 werden weder
im Rahlfs-Text noch in Ant. zueinander abgeglichen. Der griechische Text
bleibt in der Ant.- wie der Rahlfs-Fassung in sich flexibel.
Das Ergebnis vertieft sich, wenn wir auch den zweiten Teil des Zitates
betrachten, Dtn 24:16b und 2Kön (LXX 4Kgt) 14:6b/2Chr 25:4b (Tabelle 2):
Selbst die hebräische Tradition (innerhalb des Codex Leningradensis)
Tabelle 2 Dtn 24:16b und 2Kön (LXX 4Kgt) 14:6b/2Chr 25:4b

Hebräischer Text LXX (nach A und Antiochenischer Text
(nach L) B) (nach Fernández
Marcos und Busto Saiz)
Dtn 24:16b ‫ ִאיׁש ְּב ֶח ְטֹאו‬ἕκαστος
‫יּומתּו‬
ָ (B add. ἐν) τῇ
ἑαυτοῦ ἁμαρτίᾳ
ἀποθανεῖται.
2Kön 14:6b ‫ם־איׁש‬
ִ ‫ ִּכי ִא‬ὅτι ἀλλ᾽ ἢ ἕκαστος ἀλλ᾽ ἢ ἕκαστος
(4Kgt 14:6b) ‫( ְּב ֶח ְטֹאו יָמּות‬B om. ἕκαστος)
ἐν ταῖς ἁμαρτίαις ἐν τῇ ἁμαρτίᾳ
αὐτοῦ αὐτοῦ
ἀποθανεῖται. ἀποθανεῖται.
2Chr 25:4b ‫ ִּכי ִאיׁש ְּב ֶח ְטֹאו‬ἀλλ᾽ ἢ ἕκαστος ἀλλ᾽ ἢ ἕκαστος
‫ יָמּותּו‬τῇ ἑαυτοῦ ἁμαρτίᾳ ἐν τῇ ἁμαρτίᾳ αὐτοῦ
ἀποθανοῦνται ἀποθανεῖται.
(A ἀποθανεῖται).
lässt hier große Unterschiede zu; dies betrifft die Verbform (Singular in
2Kön 14:6, Plural in Dtn 24:16 und 2Chr 25:4) und eine Ergänzung gegen
über der Zitatvorlage (‫ ִּכי ִאם־‬in 2Kön 14:6 bzw. ‫ ִּכי‬in 2Chr 25,4).
Unterschiedliche Konjunktionen finden sich daraufhin auch in der grie
chischen Übersetzung. Außerdem wird das hebräische ‫ ְּב ֶח ְטֹאו‬griechisch
an allen drei Stellen unterschiedlich wiedergegeben. Varianten zwischen
A und B erklären sich als Anpassungen an die hebräische Vorlage (B add.
ἐν in Dtn 24:16) oder Vereinfachungen innerhalb des griechischen Textes
(A ἀποθανεῖται in 2Chr 25:4). Aber eine Angleichung zwischen den Stel
len findet nicht statt. Diese Beobachtungen lassen vermuten, dass die im
antiochenischen Text erreichte Konkordanz zwischen 4Kgt 14:6 und 2Chr
25:4 eher auf eine Anpassung an den hebräischen Text bzw. eine Verein
fachung innerhalb des griechischen Satzes zurückgeht anstatt auf eine
Harmonisierung zwischen den beiden Stellen.
Viel spricht dafür, dass Ant. dem Old Greek näher steht, während
der Rahlfs-Text in 4Kgt 14:6 die hebräische Textoberfläche nachträglich
genauer abbildet (καὶ υἱοὶ οὐκ entspricht präzis der hebräischen Wortfolge
‫ּובנִ ים לֹא‬
ָ ). Doch in einem Punkt wirkt sich die Textgeschichte nicht ent
scheidend aus: Weder Ant. noch der Rahlfs-Text suchen eine durchgängige
Konkordanz zum zitierten Text und der Zitate untereinander. Das Zitat
behält in der griechischen Überlieferung einen gewissen Freiraum.
Tabelle 3 Mi 3:12 und Jer 26:18 (JerLXX 33:18)17

Hebräischer Text OG nach RaHa Varianten
Mi 3:12 ‫ ָל ֵכן ִּבגְ ַל ְל ֶכם‬διὰ τοῦτο δι᾽ ὑμᾶς
‫ ִצּיֹון ָׂש ֶדה ֵת ָח ֵרׁש‬Σιων ὡς ἀγρὸς
‫ירּוׁש ַלםִ ִעּיִ ין ִּת ְהיֶ ה‬
ָ ִ‫ ו‬ἀροτριαθήσεται
καὶ Ιερουσαλημ ὡς ὡς BQSVW] εἰς A
ὀπωροφυλάκιον ἔσται ὀπωροφυλάκιον
‫וְ ַהר ַה ַּביִ ת‬ καὶ τὸ ὄρος τοῦ οἴκου ABQSVW
‫ְל ָבמֹות יָ ַער׃‬ ὡς ἄλσος δρυμοῦ 3–10 ]‫ עיין‬masoretische
Hss. haben ‫עיים‬
(vgl. Jer 26:18)18.
17 Eine Einführung in das Zitat und weitere Literatur bei G. Fischer, Jeremia 26–52
(HThKAT 39; Freiburg / Basel / Wien: Herder, 2005), 35–37. Für den Hinweis auf Jer 26:18
danken wir Darius Müller (vgl. auch Anm. 42).
18 So BHS; die Anmerkung mit dem Verweis fehlt in BHQ.
Tabelle 3 (cont.)
Hebräischer Text OG nach RaHa Varianten
Darum, Deswegen, um
euretwegen wird euretwillen wird
Zion als Acker Sion wie ein Acker
gepflügt werden, umgepflügt und
und Jerusalem wird Jerusalem wird sein
zu Steinhaufen wie eine Wachhütte
und der Berg und der Berg des
des Hauses zu Hauses wird zu
Waldeshöhen. einem verwilderten
heiligen Hain.
(LXX.D)
Jer 26:18 ‫יכיה‬ָ ‫ּמֹור ְׁש ִּתי ִת‬ַ ‫ַה‬ Μιχαίας ὁ
(JerLXX ]. . .[ ‫ָהיָ ה נִ ָּ֔בא‬ Μωραθίτης ἦν [. . .]
33:18) ‫] ֵלאמֹר‬. . .[ ‫אמר‬ ֶ ֹ ‫וַ ּי‬ καὶ εἶπεν [. . .]
‫ֹה־א ַמר יְ הוָ ה‬ ָ ‫ּכ‬ οὕτως εἶπεν κύριος
19‫ְצ ָבאֹות‬ Σιων ὡς ἀγρὸς
‫ִצּיֹון ָׂש ֶדה ֵת ָח ֵרׁש‬ ἀροτριαθήσεται
‫ירּוׁש ַליִ ם ִעּיִ ים‬
ָ ִ‫ו‬ καὶ Ιερουσαλημ εἰς
‫ִּת ְהיֶ ה‬ ἄβατον ἔσται ἄβατον B SC]
‫וְ ַהר ַה ַּביִ ת‬ καὶ τὸ ὄρος τοῦ οἴκου ἄβατος S
‫ְל ָבמֹות יָ ַער׃‬ εἰς ἄλσος δρυμοῦ ὀπωροφυλάκιον AQV
Micha der Michaias der εἰς BQSV] ὡς A
Moreschiter Morasthiter lebte
wirkte prophetisch [. . .]
[. . .] und sprach und sagte [. . .]:
[. . .]: „So spricht „So hat der Herr
der Herr der gesprochen: Sion
Heerscharen: Zion wird wie ein Acker
wird als Acker umgepflügt werden
gepflügt werden, und Jerusalem wird
und Jerusalem wird unbesiedelt und
zu Steinhaufen der Berg des Hauses
und der Berg zum Dickicht-Hain
des Hauses zu werden.“ (LXX.D)
Waldeshöhen.“
19 Im Leningradensis fehlt der in BHS nun folgende Zeilenumbruch vor dem Zitat.
Somit hebt BHS das Zitat künstlich hervor.
Das zweite Beispiel, die Aufnahme von Mi 3:12 in Jer 26:18 (JerLXX 33:18),
unterstreicht die Beobachtungen. Auch der hebräische Text geht hier mit
der Vorlage souverän um. Er enthält in den maßgeblichen Handschrif
ten die Sprachvariante ‫עיים‬/‫עיין‬, ergänzt im Zitat die vertraute, aber bei
Micha selbst fehlende Redeformel „so spricht der Herr der Heerscharen“
und passt die Syntax des Zitats dem Kontext an, so dass der Anfang des
Zitats grammatisch von Micha abweicht.
Der griechische Text (rekonstruiertes Old Greek) verkürzt die Rede
formel, vielleicht einer knapperen hebräischen Vorlage folgend (κύριος
statt κύριος παντοκράτωρ)20, und bietet bei Mi und Jer unterschiedliche
Wiedergaben für das seltene hebräische ‫ ִעּיִ ים‬, ὡς ὀπωροφυλάκιον (Mi;
„wie eine Wachhütte“) bzw. εἰς ἄβατον (Jer; „zu einem Unbetretbaren“).
In den Handschriften sind die Unterschiede geringer (AQV haben in Jer
ὀπωροφυλάκιον; BS/Sc unterstützen den kritischen Text ἄβατον). Doch
keine einzige Handschrift gleicht die Präposition εἰς/ὡς an. Dies gilt auch
für A: Zwar liest A in Jer ὡς (übrige Hss. εἰς), dafür liest es in Mi εἰς (übrige
Hss. ὡς), ein deutlicher Hinweis, dass sich die stilistischen Varianten der
Stellen unabhängig voneinander entwickeln. Hebräische und griechische
Überlieferung verraten eine nur geringe Neigung, den zitierten Text mit
dem zitierten Wort in Übereinstimmung zu bringen21.
Ziehen wir aus diesen Beobachtungen die Konsequenzen:
a. Die alte Schriftüberlieferung duldete in der Regel Varianten. Das zeigt

sich nicht nur in den neu gefundenen Quellen aus der Umgebung des
Toten Meeres, sondern schon in den inneralttestamentlichen Zitaten.
b. Die Handschriften für diese alte Überlieferung stammen in der Regel
aus jüngerer Zeit, handle es sich um den MT, die Hauptcodices der
Septuaginta oder die Zeugen des antiochenischen Textes (Ant.). Selbst
diese späteren Zeugen aber zeigen keine wesentliche Neigung, Konkor
danz zwischen Zitat und Zitatvorlage herzustellen.
Falls wir diese Tendenz auf die frühchristlichen Zitate verlängern dürfen,
spricht viel für die von Robert Hanhart eingeleitete Forschungswende:
Das frühe Christentum findet einen variantenreichen Schrifttext vor und
20 Letzteres findet sich z.B. in JerLXX 32:27.

21 Gemäß dem Apparat der BHS entsprechen immerhin 3–10 Mss. von Micha dem Jere
miatext. Falls das eine sekundäre Angleichung ist—wofür einiges spricht—, ist sie jung
und erfolgt gegen die Zitatrichtung, aber in der Buchreihenfolge zum zuerst Geschriebe
nen und grammatisch Einfacheren hin (der Plural mit ‫ ם‬ist die Normalform).
rezipiert diesen. Die Handschriften der Septuaginta und der frühchrist

lichen Literatur bewahren die textliche Vielfalt. Daher bieten die früh
christlichen Zitate an mehr Stellen als früher angenommen Gestalten des
Septuagintatextes, ohne dass redaktionelle Eingriffe der frühchristlichen
Autoren geleugnet werden dürften.
3. Die Zitate des frühen Christentums und die Septuaginta
Mit diesen Beobachtungen ist die Basis für das Wuppertaler Forschungs
projekt gelegt. Es untersucht die frühchristlichen Schriftzitate. Deren
textgeschichtlichen Hintergrund bildet primär die Septuaginta, da das
frühe Christentum griechisch schreibt22. Doch der hebräische Text ist zu
berücksichtigen (wie in den obigen Beispielen), da er die Textgeschichte
der Septuaginta (durch „kaige“-Tendenzen und Verwandtes) sowie die
Schriftkenntnis einzelner frühchristlicher Autoren beeinflusst (zu erwä
gen ist das bes. für Mt). So berührt sich die Forschung mit der erwähnten
Wiener Dokumentation über die Rezeption der Hebräischen Bibel. Aber
Doppelarbeit entfällt, da die Wiener Dokumentation nicht bei der Septua
ginta beginnt und die frühchristlichen Schriften ganz ausklammert.
In der Edition des Neuen Testaments könnte das Gewicht der Septuaginta
Folgen zeitigen: Da das Interesse für die Septuaginta erst in jüngerer Zeit
wuchs, orientieren sich die Randhinweise im Novum Testamentum Graece
bislang am hebräischen Text (abweichende Kapitelzählung bei Jer etc.). Ob
eine zukünftige Auflage das ändert, ist derzeit nicht abzusehen, entsprä
che aber der Forschungsentwicklung. Referenzen zum hebräischen Text
wären darob freilich nicht klein zu schreiben. Z.B. steht Jak 1,19 (ἔστω δὲ
πᾶς ἄνθρωπος ταχὺς εἰς τὸ ἀκοῦσαι, βραδὺς εἰς τὸ λαλῆσαι) dem Text von Sir
5,11 nach der hebräischen Handschrift A fast zitathaft nahe, LXX dagegen
ferner23.
22 Spezialstudien wiesen eine Priorität der Septuaginta ab Paulus nach: bes. D.-A. Koch,
Die Schrift als Zeuge des Evangeliums: Untersuchungen zur Verwendung und zum Verständ-
nis der Schrift bei Paulus (BHTh 69; Tübingen: Mohr Siebeck, 1986); H. Hübner, Vetus Testa-
mentum in Novo, Band 2: Corpus Paulinum (Göttingen: Vandenhoeck & Ruprecht, 1997).
Wichtige thematische Aspekte lassen sich gleichwohl zu Qumran vergleichen: vgl. T. H.
Lim, Holy Scripture in the Qumran Commentaries and Pauline Letters (Oxford: Clarendon
Press, 1997).
23 Vgl. Kraus in W. Kraus und M. Karrer, „Umfang und Text der Septuaginta: Erwägun
gen nach dem Abschluss der deutschen Übersetzung,“ in Die Septuaginta—Texte, Kontexte,
Lebenswelten (WUNT I/219; Hg. M. Karrer, W. Kraus und M. Meiser; Tübingen: Mohr Sie
beck, 2008), 8–63, hier 10. Da eine Zitationsformel fehlt und kleine Abweichungen bleiben,
3.1. Die Quellen
3.1.1. Die Haupthandschriften, Zitatmarkierungen und stilistische

Phänomene
Das materiale Zentrum der Handschriften zu Septuaginta und Neuem
Testament bilden die sogenannten Vollbibeln des 4./5. Jh. Sie vereinen
beide Schriftenkreise mit den damals noch vorhandenen offenen Rändern;
offen ist bei der Septuaginta die Zugehörigkeit von Oden und PsSal, im
frühen Christentum die von Barn, 1/2Clem und Herm. Die Codices werden
überdies lange benützt und mit Korrekturen versehen (bes. umfangreich
der Sinaiticus). Zitierte, zitierende Texte und Korrekturen lassen sich
daher in ein und demselben materialen Bestand prüfen.
Sir MS C Sir MS A Jak 1:19 Lxx Sir 5:11

Ἴστε, ἀδελφοί μου
ἀγαπητοί·
Wisset, meine
geliebten Brüder:
‫היה‬ ‫היה‬ ἔστω Γίνου
δὲ πᾶς ἄνθρωπος
Sei Sei Es sei jeder Mensch Werde
‫נכון‬ ‫ממהר‬ ταχὺς ταχὺς
fest eilend schnell schnell
‫בשמועה טובה‬ ‫להאזין‬ εἰς τὸ ἀκοῦσαι, ἐν ἀκροάσει σου
bei einem guten zum Hören zum Hören bei deinem Hören
Hören
‫ובארך‬ ‫ובארך רוח‬ βραδὺς καὶ ἐν μακροθυμίᾳ
und nach einer Weile und mit Langmut langsam und mit Langmut
‫ענה תענה נכונה‬ ‫השב פתגם‬ εἰς τὸ λαλῆσαι, φθέγγου ἀπόκρισιν
antworte Festes lass ein Wort zum Reden äußere eine
zurückkehren. Antwort.
βραδὺς εἰς ὀργήν·
langsam zum Zorn.
liegt eine freie griechische Übersetzung der Sentenz von MS A oder eine dichte Anspie
lung, kein volles Zitat vor; der sentenzartige Charakter des Textes ginge in letzterem Fall
auf den Autor des Jak zurück (so H. Frankemölle, Der Brief des Jakobus, Bd. 1 [ÖTK 17/1;
Gütersloh: Gütersloher Verlagshaus, 1994], 326f.). Den Vergleich zur hebräischen Überlie
ferung erlaubt obige Übersicht (für die hebräischen Quellen und ihre Bearbeitung danken
wir Friedrich Reiterer, Salzburg, und Matthias Millard, Wuppertal).
Zwei Phänomene verdienen besondere Aufmerksamkeit. Zum einen

schaffen die Zitate einen potentiell offenen Rand des Kanons. An der
Markierung von Zitaten im Neuen Testament durch ein eigenes Zeichen
(die Diplé, auf die wir noch zurückkommen) im Sinaiticus und Alexan
drinus wird das nicht sichtbar. Sie beschränken die Markierung auf Zitate
aus Schriften des Alten Testaments; die kirchliche Kanonbildung zeichnet
sich also im Zitationsverfahren dieser Handschriften ab. Freier dagegen
verfährt das Skriptorium der berühmtesten Handschrift, des Vaticanus:
Es markiert nicht nur die Aufnahme von 1Hen 1:9 in Jud 14f. als Zitat (s.
Tabelle 12), was sich als Nachklang eines weiteren alttestamentlichen
Kanons verstehen ließe (obwohl das Henochbuch nicht in den Bestand
der Septuaginta und nicht in den Septuagintateil des Vaticanus einging)24.
Es markiert vielmehr darüber hinaus auch ein Zitat der Völker, das Wort
τοῦ γὰρ καὶ γένος ἐσμέν aus Arat (Phainomena 5) in Apg 17:2825. Dieses
Wort galt im Judentum schon vor Paulus als eine faszinierende Beschrei
bung des Gottesverhältnisses der Menschen; bereits Aristobul, fr. 4,6
zitierte es26. Allerdings griff Aristobul in den Kontext ein; während Arat
vorschlug, bei Zeus zu beginnen (ἐκ Διός; Phain. 1), schrieb er ἐκ θεοῦ; der
eine Gott verdrängte Zeus (Beginn von fr. 4,6 = Euseb, Praep. Ev. XIII 12,6).
In diesem Sinn dürfte das Wort im Christentum wirksam geworden sein;
d.h. Arat wird einer solennen Schriftmarkierung im Bibelkodex würdig, weil
er unter den Völkern davon spricht, wie das Menschengeschlecht „aus Gott”
ist. In der Sache christianisiert das die Dichtung der Völker, und tatsäch
lich korrigiert B die Zitateinleitung. Das Skriptorium schreibt (vielleicht
nach seiner Vorlage) gegen den Ausgangstext nicht mehr, „eure” (griechi
schen) Dichter hätten das gesagt, sondern „welche der Dichter bei uns”
(ἡμᾶς statt ὑμᾶς in 28b)27. Die Wahrnehmung eines Zitats ist nicht an die
heiligen Schriften Israels gebunden (passend dazu, dass das Markierungs
24 Die Referenz des Jud weicht im Übrigen erheblich von der griechischen Textfassung
des 1Hen (nach der Edition von M. A. Knibb, Hg., The Ethiopic Book of Enoch [2 Bde.; Oxford:
Clarendon Press, 1978]) ab; vgl. H. Paulsen, Der Zweite Petrusbrief und der Judasbrief (KEK
12/2; Göttingen: Vandenhoeck & Ruprecht, 1992), 74–76. Das Skriptorium orientierte sich
an der Zitateinleitung Προεφήτευσεν δὲ καὶ τούτοις ἕβδομος ἀπὸ Ἀδὰμ Ἑνώχ, λέγων, ohne
einen Abgleich der Textform zu suchen.
25 U. Schmid, „Diplés im Codex Vaticanus,“ in Von der Septuaginta zum Neuen Testa-
ment. Textgeschichtliche Erörterungen (ANTF 43; Hg. M. Karrer, S. Kreuzer und M. Sigis
mund; Berlin / New York: De Gruyter, 2010), 99–114, hier 104, 110.
26 Übrigens wörtlich identisch zu Paulus und übereinstimmend mit dem heute herge
stellten kritischen Arat-Text; vgl. J. Martin, Hg., Phénomènes / Aratos (Paris: Belles Lettres,
1998), 1f.
27 ‫ א‬und A dokumentieren dagegen den Ausgangstext (ὑμᾶς) und verzichten auf die
Markierung durch die Diplé.
Tabelle 4 PsLXX 15:8 und Apg 2:25: Stilunterschiede in den Editionen

LXX Ps 15:8 (Gö/ LXX Apg 2:25 NT Haupthandschriften
RaHa) Haupthandschriften (NA27)
8 προωρώμην προορωμην A B S U 25 προορώμην προορωμην ‫ א‬A B* 04 05
τὸν κύριον τὸν κύριον 08 18* 33
[. . .] προωρωμην L´ Bc [. . .] προωρωμην B2 18c 424
1739
zeichen, die Diplé auf die vorchristliche Philologie zurückgeht), bleibt

aber eine solenne Hervorhebung. Neben dem Schriftbewusstsein zeigen
die großen Codices damit an dieser einen Stelle die souveräne Inkultura
tion des frühen Christentums in die Spätantike.
Zum zweiten stehen alle Skriptorien bei ihren Zitaten vor stilistischen
Entscheidungen: Sollen sie den Text in alten, gegebenenfalls attisierenden
Sprachformen wiedergeben oder in der umlaufenden Stilistik ihrer Zeit
(mit der Bandbreite der Koine)? Diese Frage taucht bei modernen Editio
nen wieder auf und sorgt gelegentlich für scheinbare Textunterschiede.
Nennen wir ein Beispiel, PsLXX 15:8 // Apg 2:25 (Tabelle 4). Die kri
tischen Ausgaben des griechischen Psalters (Gö und RaHa) geben die
Aussage „ich sah den Herrn“ mit dem klassischen Präteritum wieder
(προωρώμην), das Novum Testamentum Graece (NA27) wählt dagegen die
Koine-Form (προορώμην; s. Tabelle 4). Auf den ersten Blick unterschei
den sich Zitat und Zitatvorlage. Betrachten wir aber die Handschriften,
schreiben die ersten Hände von A, B und S/‫ א‬in LXX und NT die Koine
form. Die klassische Form der LXX geht auf den lukianischen Text zurück
und wird von einem Korrektor des Vaticanus als besseres Griechisch
erkannt, ist also Stilvorliebe von L´ Bc. Die Edition der Septuaginta folgt
dem besseren Griechisch, die Edition des Neuen Testaments den ersten
Händen der Haupthandschriften. Beides hat editorisch Reiz. Doch könnte
die Editionsentscheidung aufgrund der materialen Textbasis auch anders
erfolgen28. Um stilistische Textabweichungen zu reduzieren, wäre eine
Abstimmung der Editionsvorhaben von Septuaginta und Neuem Testa
ment sinnvoll und wünschenswert (beachte nochmals unter 4.1).
28 Genauer zur Stelle M. Karrer und U. Schmid, „Old Testament Quotations in the New
Testament and the Textual History of the Bible—the Wuppertal Research Project,“ in
Von der Septuaginta zum Neuen Testament: Textgeschichtliche Erörterungen (ANTF 43; Hg.
M. Karrer, S. Kreuzer und M. Sigismund; Berlin / New York: De Gruyter, 2010), 155–196,
hier 173f.
3.1.2. Papyri und weitere Handschriften
Tabelle 5 Gen 15:13 und Apg 7:6: Stilvarianten in der Überlieferung

Apg 7:6 Apg 7:6 Gen 15:13 Gen 15:13 Gen 15:13
(NA27 nach (D) (Gö nach (A) (P. 961; 4. Jh.;
‫א‬AB) jüngeren nach der Edition
Zeugen, da die Pietersma)
älteren [insb.
‫א‬B] z.St. nicht
erhalten sind)
Ἐλάλησεν δὲ
οὕτως ὁ θέος
[. . .]
δουλώσουσιν δουλωσουσιν δουλώσουσιν κακωσουσιν δου]λωσουσι[ν
αὐτὸ29 αυτους αὐτοὺς29 αυτο αυτους
καὶ και καὶ και και] κακωσο[υσιν
κακώσουσιν κακωσουσιν κακώσουσιν δουλωσουσιν και
αὐτοὺς καὶ αυτους και τα]πεινω[σουσιν
ταπεινώσουσιν ταπεινωσουσιν αυτους]
αὐτοὺς αυτους
ἔτη ετη υ30 τετρακόσια ἔτη τετρακοσια ετη ετη υ30
τετρακόσια
Zu den Vollcodices kommen umfangreiche weitere Zeugen – Papyri,

Majuskeln etc. –, deren Erschließung stets fortschreitet und sich auf die
Textherstellung sowie den Textvergleich auswirkt. Nennen wir wieder
ein Beispiel: Papyrus 961 (Sigel nach Rahlfs-Fraenkel; = Chester Beatty
IV) wurde von Al Pietersma 1977 neu ediert31. Folgen wir den vor diesem
Zeitpunkt erstellten kritischen Editionen zu Gen 15:13 (Rahlfs [RaHa] und
Wevers [Gö, 1974]), kürzt das neutestamentliche Zitat in Apg 7:6 nicht
nur den Referenztext der Septuaginta, sondern stellt auch die Zahlangabe
„400 Jahre” in sich um (ἔτη τετρακόσια statt τετρακόσια ἔτη). Betrachten
wir dagegen den ältesten LXX-Zeugen der Gen-Stelle, Papyrus 961 nach
29 Bezug: σπέρμα.
30 Zahlzeichen für τετρακόσια.
31 Wevers konnte sich nur auf die editio princeps von 1934 stützen (F. G. Kenyon, Hg.,
Genesis [CBBP 4.1 (text) / 4.2 (plates); London: Emery Walker, 1934–1935]); Neuedition:
A. Pietersma, Hg., Chester Beatty Biblical Papyri IV and V: Genesis (ASP 16; Toronto: A. M.
Hakkert, 1977).
der Textherstellung Pietersmas, handelt es sich um eine alte Stilvariante

der Septuagintaüberlieferung. Es bestätigt sich, dass wir Stilvarianten text
geschichtlich nicht überbewerten dürfen. Zudem finden wir einen schö
nen Nachweis dafür, dass die neutestamentlichen Zitate als Schriftzeugen
für die Septuagintaüberlieferung heranzuziehen sind, auch wenn das in
den kritischen Editionen noch nicht voll sichtbar wird.
Denn Gen 15:13 ist nur in wenigen alten Handschriften erhalten; für
die Überlieferung der ersten acht Jahrhunderte ergibt sich folgendes (voll
ständiges) Bild:
ἔτη τετρακόσια Apg (1. Jh.); P. 961 (4. Jh.); M (7. Jh.)
τετρακόσια ἔτη A (5. Jh.)32
Berücksichtigt man zudem, dass die Variante τετρακόσια ἔτη dem hebrä
ischen Text (MT) entspricht )‫(א ְר ַּבע ֵמֹאות ָׁשנָ ה‬
ַ , könnte die in Apg
erhaltene Variante ἔτη τετρακόσια (gegen Gö) Old Greek sein, während
der A-Text eine Korrektur nach der Oberfläche des hebräischen Textes
bezeugt.
Bleiben wir noch einen Augenblick bei der Stelle und betrachten die
weitere Überlieferung. In der Septuaginta bietet A einen eigenwilligen
Text mit Umstellung der Satzglieder und einmal dem Singular αὐτό statt
Plural αὐτούς; der Singular αὐτό des neutestamentlichen Zitats könnte
daher gleichfalls in der Septuagintaüberlieferung vorbereitet sein. Umge
kehrt bietet der westliche Text von Apg (D; s. Tabelle 5) den Plural αὐτούς.
Stilvarianten der Überlieferung kennzeichnen Septuaginta- und neutesta
mentlichen Text und machen selbst verschiedene Überlieferungsstränge
zueinander durchlässig.
Ronald van der Bergh (Universität Pretoria) konnte 2010/2011 eine Forschungs
zeit in Münster verbringen und den Codex Bezae (D 05) zu den Evangelien
und Apg in das Wuppertaler Projekt eintragen, ein erfreuliches Beispiel der
Kooperation. Nach seinen Untersuchungen ist nicht auszuschließen, dass
die Septuaginta-Hauptüberlieferung den neutestamentlichen Text von D05
sekundär beeinflusste, auch wenn er eine Annäherung durch unabhängige
Weiterentwicklung des Textes für wahrscheinlicher hält. Ersteres würde das
hohe Gewicht der Septuaginta für die Alte Kirche unterstreichen.
32 D ist in Gö scheinbar für den Obertext verzeichnet, tatsächlich ist Gen 15:13 in D
beschädigt und nur partiell überliefert; die hier diskutierte Stelle ist nicht erhalten (vgl.
K. v. Tischendorf, Hg., Biblia: Fragmenta Evangelii Lucae et Libri Genesis [MSIn.NC 2; Leip
zig, 1857], 105).
3.2. Die Erstellung einer digitalen Datenbank

International ergab sich eine bemerkenswerte Zusammenarbeit des Wup
pertaler Forschungsprojektes mit der Universität Pretoria. Dort konzipiert
Gert Steyn eine Dokumentation der frühchristlichen Zitatgeschichte vor
wiegend auf Papier. Beide Male werden—wie Gert Steyn formuliert—
Manuskripte mit Manuskripten verglichen33. Proprium Wuppertals ist die
Erstellung einer Datenbank, die kurz skizziert sei (vgl. die Übersicht im
Anhang):
3.2.1. Die Datenbank
Die digitale Datenbank erfasst die Schriftzitate des Neuen Testaments
nach den Hauptzeugen der Überlieferung. Neben den griechischen
Haupthandschriften von Neuem Testament und Septuaginta listet sie die
Septuaginta-Papyri bis zum 5. Jh., den MT und die biblischen Handschrif
ten aus Qumran auf. Da die Hauptcodices auch Barn, Herm, 1 und 2Clem
enthalten und damit den Textraum zu den apostolischen Vätern öffnen,
werden deren Parallelen zu den neutestamentlichen Zitaten gleichfalls
geboten. Alte Übersetzungen (Peshitta, Vulgata) geben zusätzliche Infor
mationen. Darüber hinaus versuchen wir je auch den antiochenischen
Text der Septuaginta zu den Zitatvorlagen zu rekonstruieren, da das die
Verbindung zu einem zweiten Wuppertaler Forschungsprojekt herstellt,
der Neubewertung des antiochenischen Textes durch Siegfried Kreuzer.
Die Vernetzung der Datenbank mit den New Testament Transcripts
aus Münster (nttranscripts.uni-muenster.de) ist erfolgt. Sie wird wei
tere neutestamentliche Handschriften erschließen, weshalb der neute
stamentlich-frühchristliche Teil der Datenbank sich mit einer knappen
Auswahl bescheiden kann, aber aufgrund der geplanten Vervollstän
digung der NT Transcripts auf die Dauer besonders aussagekräftig sein
wird. Umgekehrt wird die neutestamentliche Textforschung dank der
Vernetzung zwischen Münster und Wuppertal auch die Septuaginta-
Papyri bis zum 5. Jh. künftig elektronisch nützen können. Eine langjäh
rige Lücke in der Koordination von Septuaginta- und neutestamentlicher
Forschung schließt sich.
33 Vgl. G. J. Steyn, „Comparing Manuscripts with Manuscripts: Thoughts on the Com

pilation of a Synopsis of Textual Variants on the Old Testament Quotations in the New
Testament,“ in Von der Septuaginta zum Neuen Testament: Textgeschichtliche Erörterungen
(ANTF 43; Hg. M. Karrer, S. Kreuzer und M. Sigismund; Berlin / New York: De Gruyter,
2010), 211–226.
Die Datenbank ist seit Ende 2011 elektronisch nutzbar. Der Zugang ist frei.
3.2.2. Die Ausklammerung von Anspielungen

Die Datenbank konzentriert sich auf Zitate. Damit vermag sie Anspie
lungen nicht zu erfassen, obwohl auch diese textgeschichtlich oft reizvoll
sind. Die breiteren Hinweise der Wiener Dokumentation zur Hebräischen
Bibel und des Randes von NA27 für das NT sind ergänzend zu nutzen. Das
Gesamtbild wird sich freilich durch die Anspielungen nicht grundsätz
lich ändern. Beide Akzente, die wir hervorhoben, finden sich auch dort:
LXX- und neutestamentliche Überlieferung beeinflussen sich weniger als
früher erwartet, und stilistische Differenzen dürfen nicht überbewertet
werden.
Tabelle 6 Codex C—mögliche Vergleichsstellen in Septuaginta und

Neuem Testament
In C erhaltener LXX-Text vgl. C im NT34 Hinweis
Hiob (fragmentarisch)
Hi 16:19 Mk 11:10; Mt 21:9 Zitatcharakter fraglich
συνιστωρ μου εν υψιστοις ωσαννα εν τοις υψιστοις (keine Referenzformel
und begrenzte
Textentsprechung)
Hi 41:3 Röm 11:35
[lacuna] αντισθησεται μοι η τις προεδωκεν αυτω και Zitatcharakter fraglich
και υπομενει ανταποδοθησεται αυτω (keine Referenzformel
und begrenzte
Textentsprechung)35
Prov (Teile) Ntl. Zitate aus Prov liegen
stets in Lücken von C LXX
34 C ist in der Regel noch nicht in den NT Transcripts wiedergegeben, doch soll diese
Lücke in absehbarer Zeit geschlossen werden. Ein ernsthaftes Problem stellt der schlechte
Erhaltungszustand von C dar.
35 Vgl. jedoch B. Schaller, der als Vorlage für Röm 11:35 eine „revidierte Septuagintafas
sung des Hiobbuches” annimmt (“Zum Textcharakter der Hiobzitate im paulinischen
Schrifttum,“ ZNW 71 [1980]: 22–26, hier 26).
Tabelle 6 (cont.)
In C erhaltener LXX-Text vgl. C im NT Hinweis
Ekkl 7:20 Röm 3:10 fehlt in C Vergleich nicht möglich
οτι ανος (= ανθρωπος) wegen Lücke zu Röm 3:10
ουκ εστιν δικαιος εν τη in C; kritisch hergestellter
γη Text wäre καθως γεγραπται
οτι ουκ εστιν δικαιος ουδε εις.
Dass auch C diesen Text (οτι
statt οτι ανθρωπος) enthielt,
ist angesichts der stabilen
Überlieferung des Verses zu
vermuten.
Weish (einige Teile) keine Zitate im NT
Sir 4:1 Mk 10:19 Mk 10:19 par. fügt in
μη αποστερησης μη αποστερησης Motive des Dekalogs (Ex
20:12–16; Dtn 5:16–20) das
Verbot ein, jemand zu
benachteiligen. Es steht in
der Form οὐκ ἀποστερήσει[ς]
in Ex 21:10 und in Dtn
24:14 A (konkretisiert auf
besondere Lebensbereiche),
in der knappen Abstraktion
μὴ ἀποστερήσῃς
(„benachteilige nicht“)
nur bei Sir. C belässt (wie
alle Handschriften) den
Mischtext, obwohl er die
sprachliche Form des
Dekalogs ändert, stellt also
keine Zitatfolge her. Das
stilistisch gute μή muss
nicht Sir 4:1 entnommen
sein, sondern kann eine
allgemeine sprachliche
Weiterentwicklung sein;
denn Mk verwendet μή
auch bei den Geboten des
Dekalogs (gegen οὐ mit
Futur von LXX).
Sir 17:26 2Tim 2:19 Sir nicht Leittext für
αποστρεφε απο αδικιας αποστητω απο αδικιας das Zitat und begrenzte
πας ο ονομαζων το Textentsprechung; der
ονομα κυ (= κυριου) weitere Bezugstext Jes 26:13
ist in C LXX verloren
Das verdeutlicht Codex C, die vierte, allerdings so unvollständig erhaltene

Vollbibel des 4.–5. Jh., dass kein Vergleich von Zitaten, sondern lediglich
der von Anspielungen möglich ist.
Ein Vollcodex war neben dem Sinaiticus (S/‫)א‬, Vaticanus (B), Alexandri
nus (A) und Ephraemi rescriptus (C) auch der jüngere Venetus (V; 8. Jh.)36.
Doch ist im Venetus das ganze NT verloren, so dass kein Vergleich zwi
schen zitiertem und zitierendem Text möglich ist. C ist umgekehrt im Alten
Testament zu unvollständig erhalten, um Zitate und Zitatvorlagen zu paral
lelisieren37. Der neutestamentliche Text von C wird in den New Testament
Transcripts sichtbar werden, sobald die Münsteraner Eintragungen bis dort
hin fortgeschritten sind, und dadurch dann mit der Wuppertaler Datenbank
verlinkt.
An keiner der erkennbaren Anspielungen im Codex C (Tabelle 6) bemüht
sich das Skriptorium, sprachliche Differenzen zwischen Septuaginta und
Neuem Testament auszugleichen, seien diese Differenzen stilistisch (wie
der Artikel τοις in Mk 11:10 par. gegenüber Hi 16:19) oder semantisch (wie
bei αποστητω 2Tim 2:19 versus αποστρεφε). Falls das Skriptorium die
Bezüge zwischen Septuaginta-Text und Neuem Testament wie die spä
teren Ausleger wahrnahm, leitete es aus ihnen also keinerlei Druck zur
Angleichung der Texte ab.
4. Zusätzliche Beobachtungen an den Zitaten
Es ist nicht möglich, hier alle Aspekte des Wuppertaler Projektes zu

berichten. Geben wir daher nur einige ausgewählte Hinweise:
36 Vgl. A. Rahlfs und D. Fraenkel, Hg., Verzeichnis der griechischen Handschriften des
Alten Testaments, Band 1,1: Die Überlieferung bis zum VIII. Jahrhundert (Septuaginta: Vetus
Testamentum Graecum. Supplementum; Göttingen: Vandenhoeck & Ruprecht, 2004),
313–315.
37 Vom Bestand des Codex Ephraemi rescriptus blieben neben 145 Blättern zum Neuen
Testament 64 Blätter mit Teilen der Weisheitsschriften erhalten. Allerdings weist der
Codex gerade dort im Neuen Testament Lücken auf, wo die erhaltenen Weisheitsschriften
zitiert werden könnten. Die Anspielungen, die vorhanden sind, sind in der Datenbank zu
übergehen. Stellennachweise zu den libri sapientales bei Rahlfs und Fraenkel, Verzeichnis,
314, zum NT am einfachsten im Anhang des NA27.
4.1. Zitateinleitungen und Zitattext

Einleitungen sind eine wichtige Entscheidungshilfe, um Zitate zu erken
nen und von Anspielungen zu unterscheiden. Deshalb führt die Daten
bank die „quotation marks” auf 38.
Das „Inventory of Structurally Important Literary Features in the Anony
mous and Pseudepigraphic Jewish Literature of Antiquity”39 erstellt derzeit
eine gemeinsame Begriffsbildung für Phänomene der antiken jüdischen
Literatur. Zitate (“quotations”) fallen dort unter die Strukturierung eines
Textes durch einen anderen. Erkennbar sind sie an metasprachlichen Phä
nomenen (“Meta-linguistic structuring of a text by another text”)40. Das
Inventory berücksichtigt die Varianz der Textüberlieferung (“the base text
wording [. . .] can vary from ms to ms”)41, führt aber einleitende Formulie
rungen, die die Zitate eindeutig erkennbar machen, noch nicht auf. Das
sollte dort nachgetragen werden.
Dieses heuristische Kriterium erlaubt, der Datenbank die frühchristlichen
Zitateinleitungen zu entnehmen42. Textgeschichtlich interessant, ist nicht
immer leicht zu entscheiden, ob ein Markierungselement schon zum Zitat
oder noch zur Einleitung gehört. Namentlich gilt das für das sehr häufige
und bedeutungsreiche ὅτι (vgl. schon den Hinweis auf Ekkl 7:20/Röm 3:10
oben in Tabelle 6). Im Einzelfall lässt sich hier eine Korrektur für die kriti
sche Edition des Neuen Testaments vorschlagen, die Zitate kursiviert und
daher eine Überprüfung der kursivierten Texte verlangt:
Tabelle 7 Lk 4:10/PsLXX 90:11

Lk 4:10 (A B) γεγραπται γαρ οτι τοις αγγελοις αυτου εντελειται
(A εντελειτε) περι σου του διαφυλαξαι
(A διαφυλαξε) σε
Lk 4:10 (‫)א‬ γεγραπται γαρ οτι τοις αγγελοις αυτου εντελιται περι σου
του διαφυλαξαι σε
38 Vgl. den Anhang (Anm. 34).

39 Verantwortlich A. Samely, P. Alexander, R. Bernasconi und R. Hayward (Forschungs
projekt der Universität Manchester); vgl. www.llc.manchester.ac.uk/research/projects/
ancientjewishliterature (Zugriff am 19/8/2011), das Inventory kann hier abgerufen werden:
www.llc.manchester.ac.uk/research/projects/ancientjewishliterature/inventory (Zugriff am
19/8/2011).
40 Punkt 6 des genannten Papieres (vgl. Anm. 39).
41 Weiterhin Punkt 6 des genannten Papieres (vgl. Anm. 39).
42 Eine Übersicht entsteht in der Magisterarbeit von Darius Müller „Die Gegenwart der
Schrift im Neuen Testament. Eine sachkritische und theologische Untersuchung zu den
Markierungen der alttestamentlichen Zitate im Neuen Testament” derzeit in Begleitung
des Projektes an der Kirchlichen Hochschule Wuppertal/Bethel.
Tabelle 7 (cont.)
Lk 4:10 (Vulgata) scriptum est enim quod angelis suis mandabit de te ut
conservent te
PsLXX 90:11 (A) oτι τοις αγγελοις αυτου εντελειτε περι σου
του διαφυλαξε σε [. . .]
PsLXX 90:11 (B) oτι τοις αγγελοις αυτου εντελειται περι σου
του διαφυλαξαι σε [. . .]
PsLXX 90:11 (S) oτι τοις αγγελοις αυτου εντελειται περι σου
του διαφυλαξαι σε [. . .]
Ps 90:11 (Vulgata) quia (sic iuxta Hebr.; iuxta LXX: quoniam)
angelis suis mandabit de te ut custodiant
te [. . .]
PsMT 91:11 [. . .]‫ִכּי מַ ְלאָ כָ יו יְ צַ ֶוּה־לָ ְּך ִל ְשׁמָ ְרָך‬
Z.B. schlägt NA27 (S. 163 zSt.) vor, das ὅτι in Lk 4:10 noch zur Zitateinlei
tung mit γέγραπται zu nehmen (es wird nicht kursiviert und durch einen
Absatz vom Zitat abgesetzt). Γέγραπται ὅτι ist eine zweifellos häufige
Zitateinleitung, aber das ὅτι gehört nicht zwingend dazu; das im Evan
gelium unmittelbar vorangehende Zitat 4:8 etwa begnügt sich mit der
Markierung durch γέγραπται ohne ὅτι. Daher ist ebenso möglich, ὅτι in
Lk 4:10 noch zum Zitat zu rechnen. Tatsächlich entspricht das ὅτι dort
wie die gesamte zitierte Phrase (ὅτι τοῖς ἀγγέλοις αὐτοῦ ἐντελεῖται περὶ
σοῦ τοῦ διαφυλάξαι σε) genau der Septuaginta. Zu vermuten ist, dass die
Unterscheidung vom Zitat erst sekundär, unter Einfluss der lateinischen
Überlieferung entstand. Erkennbar ist das in der Datenbank durch die
Wiedergabe der Vulgata. Diese nämlich überträgt das hebräische ‫כי‬/grie
chische ὅτι des Psalms (MT 91:11) mit „quoniam“ (iuxta LXX) bzw. „quia“
(iuxta Hebr.), das neutestamentliche ὅτι dagegen mit „quod.“ Der lateini
sche Text unterscheidet Altes und Neues Testament. Doch ein Grund, das
in die griechische Grundüberlieferung zurück zu übertragen, fehlt. Wir
würden daher vorschlagen, ὅτι in Lk 4:10 zum Zitat zu zählen. In einer
künftigen Auflage des Novum Testamentum Graece (Nestle-Aland) sollte
es in die erste Zeile des Zitats gezogen und kursiviert werden.
Bleiben wir noch einen Augenblick bei der Vulgata. Nicht nur
„quoniam“/„quia“ und „quod“ unterscheiden sich dort, sondern auch die
Wiedergabe von ‫שמר‬/διαφυλάξαι. Beim Psalm wählt die Vulgata (iuxta
Hebr. und LXX) „custodire,“ im Evangelium „conservare.“ Die Übersetzung
des Alten und des Neuen Testaments erfolgen unabhängig voneinander,
eine deutliche Warnung davor, einen hohen Anpassungsdruck bei Zitaten

zu postulieren.
Die griechischen Haupthandschriften unterstreichen das durch eine
stilistische Eigenheit. Sie erliegen nämlich gelegentlich einer phonetisch
bedingten Schreibvariante (‫ *א‬Lk 4:10 εντελιται, A Ps 90:11 εντελειτε und
διαφυλαξε). Indes findet sich diese Vereinfachung nur teilweise überein
stimmend in zitiertem und zitierendem Text (‫ *א‬liest in Ps 90:11 εντελειται,
A in Lk 4:10 εντελειται und διαφυλαξαι). Demnach neigen die Skriptorien
in Stilfragen zur Übereinstimmung von zitiertem und zitierendem Text,
aber nicht durch wechselseitigen Einfluss NT-LXX, sondem wegen der
gemeinsamen Zeitlage der Schreiber. Es zeichnet sich unbeschadet der
obigen Hinweise zum westlichen Text ab, dass wir im Allgemeinen mit
nur geringen wechselseitigen Einflüssen zwischen der Überlieferung der
Septuaginta und der des Neuen Testaments rechnen dürfen.
4.2. Mehrfachzitate in den Codices

Besonders aufschlussreich sind Mehrfachzitate in unseren Leitcodices.
In vielen Fällen bestätigen sie den kritisch rekonstruierten LXX-Text.
Dieser (das Old Greek) hatte also in der neutestamentlichen Zeit noch
eine erhebliche Verbreitung. Die frühchristliche Rezeption, die in der
Regel älter als die Septuaginta-Handschriften ist, unterstützt insofern die
Rekonstruktion der kritischen Septuaginta-Edition. Zugleich bringt sie
manche zusätzliche Erkenntnis. Verdeutlichen wir das an Prov 3:12, weil
der Alexandrinus diese Stelle auch im 1Clem enthält, der sich im neute
stamentlichen Kanon nicht durchsetzte: Prov 3:12 und die Zitate in Hebr
12:6 sowie in 1Clem 56:4 stimmen in A völlig überein, und weitere Hand
schriften unterstützen diese Lesart (im Hebr ‫ א‬und für die Zeile ον γαρ
αγαπα κυριος παιδευει auch p46). Der zitierte Text ist nach der derzeiti
gen Kenntnis der Handschriften—und daher zu Recht in den kritischen
Editionen—Old Greek.
Nun war aber die Übersetzung παιδεύει für das Hebräische ‫ יוכיח‬unge
wöhnlich. Sie findet sich in der Septuaginta (nach RaHa) allein an unserer
Stelle. Weit häufiger war die Übersetzung durch ἐλέγχειν (Gen 21:25 usw.)43.
Deshalb dringt diese Übersetzungsalternative gelegentlich in die Überlie
ferung von Prov 3:12 ein, am gewichtigsten im Vaticanus (s. Tabelle 8).
43 Belege bei E. Hatch und H. A. Redpath, A Concordance to the Septuagint and the
Other Greek Versions of the Old Testament, Bd. 1 (Oxford: Clarendon Press, 1897), 449, s.v.
3a und in den digitalen Ausgaben (BibleWorks u.ä.).
Tabelle 8 Prov 3:1244/Hebr 12:6/Apk 3:19/1Clem 56:4

Prov 3:12 (A) ον γαρ αγαπα κυριος παιδευει μαστιγοι δε παντα υιον
ον παραδεχεται
Prov 3:12 (B) ον γαρ αγαπα κς ελεγχει μαστιγοι δε παντα υιον
Prov 3:12 (S) ον γαρ αγαπα κς παιδευει μαστιγοι δε παντα υιον
Philo, Congr. 17745 ὃν γὰρ ἀγαπᾷ κύριος ἐλέγχει, μαστιγοῖ δὲ πάντα υἱὸν
ὃν παραδέχεται
Prov 3:12 (Vulgata) quem enim diligit Dominus corripit et quasi pater in filio
conplacet sibi
Prov 3:12 (MT) ‫ּוכאָ ב אֶ ת־בֵ ּן יִ ְרצֶ ה‬
ְ ‫יח‬ַ ‫יוֹכ‬
ִ ‫ִכּי אֶ ת אֲשֶ ׁר ֶיאֱהַ ב יְ הוָה‬
Hebr 12:6 (AS) ον γαρ αγαπα κς παιδευει μαστιγοι δε παντα υιον
ον παραδεχεται (A παραδεχετε)
Hebr 12:6 (P46) ον γαρ αγαπα κς παιδευει
Hebr 12:6 (Vulgata) quem enim diligit Dominus castigat flagellat autem omnem
filium quem recipit
Vgl. Apk 3:19 (A) εγω οσους εαν φιλω ελεγχω και παιδευω ζηλευε ουν
και μετανοησον
Apk 3:19 (S) εγω οσους αν φιλω ελεγχω και παιδευω ζηλωσον ουν
και μετανοησον
1Clem 56:4 (A) ον γαρ αγαπα κς παιδευει μαστιγοι δε παντα υιον
Erstmals zitiert Philo (Congr. 177) die Variante46. Eine Anspielung der
Apk (die also zur Datenbank zu ergänzen ist) kombiniert daraufhin beides:
44 Der Text von Congr. folgt der Ausgabe von L. Cohn und P. Wendland, Hg., Philo-
nis Alexandrini opera quae supersunt ediderunt Leopoldus Cohn et Paulus Wendland, Bd. 3
(Berlin: Georg Reimer, 1898); dort sind keine Varianten verzeichnet.
45 ProvLXX 3:12 ist (nach Rahlfs und Fraenkel, Verzeichnis) nur in den großen Codices
ABSV erhalten. Eine besondere antiochenische Textform gibt es zum Proverbienbuch ver
mutlich nicht. Vgl. hierzu die im Wuppertaler Projekt entstandenen Überlegungen unter
www.kiho-wb.de/ISBTF/datenbank/annotationes (Nr. 28; Marcus Sigismund).
46 Da Cohn und Wendland, Philonis Alexandrini opera, z.St. keine Varianten verzeich
nen (vgl. auch Anm. 45), ist davon auszugehen, dass der Text nicht nur von den Hss.,
die einen vermutlich überarbeiteten Text bezeugen (GF), sondern auch von den übrigen
(MAH) gelesen wird. Vgl. Cohn und Wendland, ibid., xvii und Kraft, „Philo’s Bible Revisited,“
237–253, hier 239.
Tabelle 9 Jes 40:3/Mt 3:3/Mk 1:3/Lk 3:4/Joh 1:23/Barn 9:3

Jes 40:3 (ABS) φωνη βοωντος εν τη ερημω
Jes 40:3 (Ant. rekonstruiert)47 φωνὴ βοῶντος ἐν τῇ ἐρήμῳ
Jes 40 :3 (Vul.) vox clamantis in deserto
Mt 3:3 (BS) φωνη βοωντος εν τη ερημω
Mk 1:3 (ABS) φωνη βοωντος εν τη ερημω
Lk 3:4 (ABS) φωνη βοωντος εν τη ερημω
Lk 3:4 (Vulg.) vox clamantis in deserto
Joh 1:23 (ABS) φωνη βοωντος εν τη ερημω
Joh 1:23 (Vul.) vox clamantis in deserto
Barn 9:3 (S) και παλιν λεγει [. . .] και παλιν λεγει [. . .] και παλιν λεγει ακουσατε τεκνα
φωνη βοωντος εν τη ερημω
S nach jüngerer Korrektur (cc) φωνης [βοωντος εν τη ερημω]
Der Autor übernimmt παιδεύειν und ergänzt in Anlehnung an die zeit

genössische Fortschreibung des Textes ἐλέγχειν. Die Überlieferung Phi
los und des ersten Christentums hilft, das Old Greek zu bestätigen, und
erlaubt, den Wechsel zu ἐλέγχειν zu beschreiben. Dieser gehört noch in
vorchristliche Zeit, vermag jedoch bis zum Ende des 1. Jh. παιδεύειν nicht
zu verdrängen.
Im Einzelfall wird sich wieder eine Korrektur der kritischen Edition
ergeben. Wählen wir das Beispiel dafür diesmal aus einer nachneutesta
mentlichen Quelle, dem Barn: Die Zeile φωνὴ βοῶντος ἐν τῇ ἐρήμῳ aus Jes
40:348 wird übereinstimmend so von allen Evangelien (Mk 1:3; Mt; Lk; Joh)
und von Barn 9:3 ‫( *א‬prima manus) zitiert. Die frühchristliche Rezeption
bestätigt das Old Greek der kritischen LXX-Edition (Tabelle 9).
Der Barn enthält dabei eine Besonderheit. Er bildet in seiner Rezeption
eine Zitatkette und leitet in ihr das Jes-Wort durch παλιν λεγει ακουσατε
τεκνα ein. Die prima manus versteht diese Phrase insgesamt als Zitati
onsformel (vgl. προσεχετε τεκνα in 15:4); die Eröffnung des Zitats mit dem
Nominativ φωνη präzis nach LXX ist nur konsequent. Ein jüngerer Korrek
tor des Sinaiticus (in der digitalen Edition [www.sinaiticus.de] cc genannt
und im Rang dem herausragenden Korrektor ca nicht zu vergleichen) löst
47 Die Variante (add. διὰ τῆς ἀβάτου) des rekonstruierten antiochenischen Textes zieht
sich über alle Untergruppen der Ant-Tradition und ist daher sicherlich die ursprünglich
antiochenische Lesart, zumal sie flankierend zur eigentlichen Hs.-Überlieferung auch von
den starken Zeugen Qmg und Theodoret geboten wird. Ant gleicht hier (wie gelegentlich
in Jes zu beobachten) an Proto-MT (‫ )בערבה‬an.
48 J. Ziegler, Isaias (Septuaginta: Vetus Testamentum Graecum 14; Göttingen: Vanden
hoeck & Ruprecht, 1939), 266f.
diesen Bezug auf. Seiner Ansicht nach beginnt mit ακουσατε eine Rede des
Geistes (vgl. 9:2). Er passt die Syntax dem an und korrigiert den Text zu
ακουσατε [. . .] φωνης, „hört die Stimme.“ Eine Korrektur gegen LXX berei
tet ihm kein Beschwer (eines der vielen Indizien dafür, dass die Textüber
lieferung auch in später Zeit unabhängig vom zitierten Text geschah). Der
Hierosolymitanus, die zweite, 1056 geschriebene49 Haupthandschrift des
Barn unterstützt das. Auch wenn sie einen guten Text bietet, kommen wir
mit der Variante nicht vor die byzantinische Zeit.
Die derzeit im deutschen Raum weitest verbreitete kritische Barn-
Edition (Wengst 1998; vgl. älter Bihlmeyer) schließt sich trotzdem der Kor
rektur an (also φωνης). Der Grund ist nicht genannt, aber wahrscheinlich
die Differenz zu LXX. Unsere bisherigen Beobachtungen stützen das nicht.
Sie raten eher dazu, der prima manus des Sinaiticus zu folgen und ein
hohes Zitatbewusstsein beim frühchristlichen Autor anzunehmen. Erst
lange nach ihm, vielleicht erst gegen Ende des 1. Jahrtausends gewinnt
die geänderte Auffassung der Syntax die Oberhand.
Auch dies freilich ist hoch interessant. Denn noch über den neutesta
mentlichen Text hinaus entfernen wir uns von der alten These neute
stamentlicher Textforschung, Entsprechungen zur LXX in Handschriften
frühchristlicher Texte seien sekundär gegenüber Abweichungen von der
Septuaginta.
Runden wir die Beobachtungen an PsLXX 109:1 (PsMT 110:1), dem verbrei
tetsten Psalmzitat der frühchristlichen Literatur ab (Mt 22:44; Mk 12:36;
Lk 20:42f.; Apg 2:34f.; Hebr 1:13; 1Clem 36:5; Barn 12:10 vgl. 1Kor 15:25;
Tabelle 10). Es begegnet in Mk 12:36/Mt 22:44 mit der Lesart ὑποκάτω,
sonst mit ὑποπόδιον. In die erhaltenen Septuagintahandschriften drang
ὑποκάτω nicht ein50. Dort dominiert ὑποπόδιον und wird von Rahlfs als
Ausgangstext rekonstruiert. Nachträgliche Unterstützung erhält Rahlfs’
Rekonstruktion von Papyrus 2110, der um 200 entstanden sein dürfte, aber
für die Edition noch nicht benutzt wurde. Folgen wir dem, dann korrigiert Lk
den Text seiner frühchristlichen Vorlage (Mk) zum Septuaginta-Haupttext;
der Septuaginta-Haupttext bewährt sich als ein interner frühchristlicher
Maßstab.
49 Datierung nach K. Wengst, Hg., Didache (Apostellehre), Barnabasbrief, zweiter

Klemensbrief, Schrift an Diognet (SUC 2; Darmstadt: Wissenschaftliche Buchgesellschaft,
1998), 5f.
50 Soweit sich das der Edition der Psalmen durch Rahlfs, Psalmi cum Odis, entnehmen
lässt, die der Erneuerung bedarf.
Tabelle 10 PsLXX 109:1/Mk 12:36/Mt 22:44/Lk 20:42f.

PsLXX 109:1 (Gö) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποπόδιον τῶν ποδῶν σου
PsLXX 109:1 (Ant. rekonstriert) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποπόδιον τῶν ποδῶν σου
Mk 12:36/Mt 22:44 (NA27) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποκάτω τῶν ποδῶν σου
Lk 20:42f (NA27) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποπόδιον τῶν ποδῶν σου
Apg 2:35 (NA27) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποπόδιον τῶν ποδῶν σου
Hebr 1:13 (NA27) ἕως ἂν θῶ τοὺς ἐχθρούς σου ὑποπόδιον τῶν ποδῶν σου
1Clem 36:5 A εως αν θω τους εχθρους σου υποποδιον των ποδων σου
Barn 12:10 S εως αν θω τους εχθρους σου υποποδιον των ποδων σου
Trotzdem verlangt ein Indiz eine textgeschichtliche Differenzierung:

Ὑποκάτω gibt den Ausgangstext ‫ הדם‬vorzüglich zielsprachlich (nämlich
mit einem präpositional verstehbaren Adverb) wieder, während ὑποπόδιον
das Nomen „Schemel” aus der Ausgangssprache abbildet. Ein spezifisch
christliches Interesse für die Wahl von ὑποκάτω lässt sich nicht ausma
chen51, jedoch die Wendung ὑποκάτω τῶν ποδῶν bei Septuaginta-Über
setzern finden (im kritischen Septuaginta-Text [RaHa] 2Kgt [2Sam] 22:10;
Mal 3:21 und PsLXX 8:7). In PsLXX 8:7 wird es laut Rahlfs52 von allen Haupt
handschriften getragen, dort als Äquivalent zu ‫)תחת‬. Berücksichtigen
wir, dass abbildende Lesarten in der Überlieferung seit dem 1. Jh. v.Chr.
vordrangen (sie sind ein Kennzeichen von Kaige), können wir nicht ganz
ausschließen, dass das frühchristliche ὑποκάτω eine bereits vorchristliche,
dem Old Greek nahe Stilvariante repräsentiert53. Das frühe Christentum
würde uns dann eine in der Septuaginta-Überlieferung verlorene alte Text
form bewahren, die vorzüglich zu PsLXX 8:7 passen würde. Die Daten in
der Wuppertaler Datenbank erlauben keine Lösung der Frage; ihre Stärke
liegt darin, eine solche Frage sichtbar zu machen. Hoffen wir für die Ent
scheidung auf die anstehende Neuedition des Psalters.
51 Ὑποκάτω begegnet bei Mk dreimal, ist also kein ausgeprägter Vorzugswortschatz.

Das Mt bietet es nur an unserer Stelle, ist hier also von Mk abhängig.
52 Rahlfs, Psalmi cum Odis, z.St. S. 89.
53 Vgl. M. Karrer, „Der Septuaginta-Text im frühen Christentum,“ in Handbuch zur
Septuaginta, Band 1: Einleitung in die Septuaginta (Hg. S. Kreuzer; Gütersloh, 2012, im
Erscheinen).
5. Markierungen in den Handschriften
Die Aufnahme der Handschriften für die Wuppertaler Datenbank führte

zu einer aufschlussreichen ergänzenden Beobachtung, die sich in der
Datenbank leider nicht abbilden lässt, weil sie keine Fotos vorsieht: Die
Handschriften fügen in ihren Text Zeichen ein, die uns die Zitationsge
schichte zu verstehen helfen.
5.1. Leseakzente in neutestamentlichen Papyri

Die Anfänge hervorhebender Zeichen sind bislang nur unzureichend
erschlossen. Aber bereits einer der ältesten neutestamentlichen Papyri,
P46 (herkömmlich um 200 datiert), weist solche Zeichen auf (Abb. 1). Ein
Schreiber (erste oder frühe zweite Hand) fügt dort Akzente über dem Text
ein54. Sie signalisieren, soweit sich das gegenwärtig sagen lässt, eine Her
vorhebung oder eine Stelle, wo ein Leser bzw. Vorleser kurz innehalten
könnte; das Blatt mit Hebr 4:6–8 (folio 24r) etwa enthält solche Zeichen
am Anfang von Hebr 4:6 (Z. 3, kein Zitat), am Anfang von Hebr 4:7 (Z. 6,
kein Zitat, Zitat erst später im Vers), innerhalb von Hebr 4:7 (Z. 8, passend
zum Beginn des Zitats aus LXX Ps 94:7f.) und am Übergang von Hebr 4:7
zu 4:8 (Z. 10, Ende des Zitats). Oder nehmen wir Röm 12:19–21 (Tabelle 11):
Die Akzente stehen in v. 19 vor der Zitatformel γέγραπται γάρ und hin
ter ἀνταποδώσω, umgeben somit den in Nestle-Aland27 als Zitat aus Dtn
32:35 markierten Text. Kurz danach steht ein Akzent jenseits eines Zitat
bereichs (in v. 21 hinter κακοῦ).
Tabelle 11 Dtn 32:35/Od 2:35/Prov 25:22/Röm 12:19/Hebr 10:30

Röm 12:19 (NA27) γέγραπται γάρ· ἐμοὶ ἐκδίκησις, ἐγὼ ἀνταποδώσω,
λέγει κύριος
Dtn 32:35 (Gö) ἐν ἡμέρᾳ ἐκδικήσεως ἀνταποδώσω
Dtn 32:35 (σ´ Gö) mihi ultio et retribuam
Od 2:35 (Gö) ἐν ἡμέρᾳ ἐκδικήσεως ἀνταποδώσω
Prov 25:22 ὁ δὲ κύριος ἀνταποδώσει σοι ἀγαθά
Hebr 10:30 (NA27) οἴδαμεν γὰρ τὸν εἰπόντα· ἐμοὶ ἐκδίκησις, ἐγὼ ἀνταποδώσω
(A ‫א‬c: add. λέγει κύριος/λέγε κς)
54 Dieselbe Hand korrigiert, soweit sich sehen lässt, in 12:16 φρονουτες (first hand) zu
φρονουντες (Korrektor). Die Tintenfarbe der Akzente unterscheidet sich vom Text selbst.
Demnach gehören die Akzente zu einem frühen Bearbeitungsgang.
Auszug aus P.Mich.inv. 6238 (NT P46/Chester Beatty II); Recto (3554) zu Röm 12, Abb.
nach http://quod.lib.umich.edu/a/apis/x-3554/6238_31.tif, abgerufen am 21.6.2011 (image
digitally reproduced with the permission of the Papyrology Collection, Graduate Library,
University of Michigan); die von uns nachträglich eingefügten Ringe markieren drei
Akzente (vermutlich Lesezeichen).
Abb. 1 P46—Leseakzente
Demnach sind die Akzente des Papyrus nicht als Zitatmarkierungen ent
standen, sondern als allgemeine Lesehinweise. Gleichwohl sind sie für die
Zitatgeschichte interessant. Denn sie beweisen, dass Zitathinweise (“quo
tation markers“) und manchmal auch das Ende von Zitaten auffielen.
Der im Beispiel zitierte Text in Röm 12:19 entspricht übrigens nicht
der Septuaginta, sondern bis auf die asyndetische Reihung (Verzicht auf
ein καί) Symmachus (Tabelle 11). Die frühen Christen benutzten in ihren
Zitaten also auch Textformen, die die sog. jüngeren Übersetzungen vorbe
reiten. Die Zitate verdeutlichen die Textgeschichte über die Septuaginta
hinaus.
Der Symmachustext von Dtn 32:35 (Oden 2:35), „mihi ultio et retribuam,“
lässt sich über das Syrische rekonstruieren55. Das zweite frühchristliche
Zitat in Hebr 10:30 unterstützt den asyndetischen Wortlaut des Paulus (ohne
καί). Das Neue Testament benützt mithin eine Textform, die wir Proto-Sym
machus nennen können56.
5.2. Die Diplé
Tabelle 12 Die Diplé bei Jud 14

Die häufigste Gestalt der Diplé ist >. Graphisch begegnen auch Varianten.
Vgl. z.B. Jud 14 in Codex Vaticanus:
ΕΠΡΟΦΗΤΕΥΣΕΝΔΕΚΑΙ
ΤΟΥΤΟΙΣΕΒΔΟΜΟΣΑΠΟ
ΑΔΑΜΕΝΩΧΛΕΓΩΝΙ
> ΔΟΥΗΛΘΕΝΚΣΕΝΑΓΙΑΙΣ
> ΜΥΡΙΑΣΙΝΑΥΤΟΥΠΟΙ
> ΗΣΑΙΚΡΙΣΙΝΚΑΤΑΠΑΝ
> ΤΩΝΚΑΙΕΛΕΓΞΑΙΠΑΝ
> ΤΑΣΤΟΥΣΑΣΕΒΕΙΣΠΕ
ΡΙΠΑΝΤΩΝΤΩΝΕΡΓΩΝ
Im Codex Vaticanus ist z.B. Jud 14 mit Diplé ausgezeichnet, der Vers nimmt 1Hen
1:9 auf und ist mit ἐπροφήτευσεν δὲ καὶ τούτοις ἕβδομος ἀπὸ Ἀδὰμ Ἑνὼχ λέγων
eingeleitet57.
55 W. Baars, Hg., New Syro-hexaplaric Texts (Leiden: Brill, 1968), 95, 144, 148.
56 Zur weiteren Diskussion M. Karrer, Der Brief an die Hebräer, Bd. 2 (ÖTK 20/2;
Gütersloh: Gütersloher Verlagshaus, 2008), 213f. z.St.; G. J. Steyn, A Quest for the Assumed
LXX Vorlage of the Explicit Quotations in Hebrews (FRLANT 235; Göttingen: Vandenhoeck
& Ruprecht, 2011).
57 S. das Faksimile 2.St.; Bibliorum Sacrorum Graecorum Codex Vaticanus B (Rom: Isti
tuto Poligrafico e Zecca dello Stato, 1999).
Einen nächsten Schritt könnten Spatia oder Absätze nach zunächst ein
zelnen Zitaten gebildet haben. Beides setzen die großen Handschriften
des 4. Jh. voraus (vgl. z.B. ‫ א‬zu 1Kor 2:9 oder Apg 28:25–28)58. Doch stellen
wir dies zurück. Denn diese Merkmale werden nun durch eine andere
Markierung in den Schatten gestellt:
Die biblischen Vollcodices führen den Randhinweis der Diplé ein, ein
Zeichen, das—wie der Name verrät—aus zwei aufeinander stoßenden
Balken gebildet ist (> und ähnlich). Das philologische Zeichen ist alt59, der
Gebrauch als innerbiblischer Verweis neu. Die Einführung geht unsyste
matisch vor: Die Skriptorien setzen das Zeichen nur in einem von Codex
zu Codex differierenden Teil der neutestamentlichen Schriften, und wo sie
es setzen, folgen sie textinternen Signalen, namentlich Zitationsformeln.
Im Vaticanus begegnet die Diplé daher auch bei der gerade angespro
chenen Stelle 1Kor 2:9 (wegen der Zitationsformel καθὼς γέγραπται), deren
Schriftreferenz sich textextern nicht verifizieren lässt, und im Alexandrinus
bei dem Zitat unbekannter Herkunft in 1Tim 5:18 (nach der Zitationsformel
λέγει γὰρ ἡ γραφή)60. Dieser Befund schließt aus, dass die markierten Texte
regelhaft an alttestamentlichen Bezugstexten geprüft wurden61, obwohl die
biblischen Vollcodices einen solchen Vorgang am selben materialen Objekt
im selben Skriptorium erlauben würden. Immerhin gibt es aber im Vatica
nus an dritter Stelle ein Indiz, dass das Problem bewusst wurde: Der Codex
setzt bei Jak 4:6 eine Diplé (διὸ λέγει, Zitat aus Prov 3:34 LXX), bei dem nicht
identifizierbaren Zitat von Jak 4:5 trotz der Zitationsformel ἡ γραφὴ λέγει
58 Mithilfe der Absätze markiert ‫ א‬Sinnabschnitte, dadurch auch häufiger Abschnitte

wörtlicher Rede. So ist z.B. die Paulusrede Apg 26:25–27 in zwei Absätze aufgeteilt (26:25f.
und 26:27). Ähnlich ist auch der Beginn der Paulusrede in Apg 28:25b–28 durch einen
Absatzeinschnitt markiert. Jedoch umfasst der Absatz nicht die ganze wörtliche Rede, son
dern nur die Einleitung und das Schriftwort aus Jes 6:9f. Die Zitatauszeichnung geschieht
hier also vermutlich anhand textinterner Impulse, doch bereitet sie den Rückverweis auf
die LXX vor.
59 Im ersten christlichen Beleg, dem Irenäus-Papyrus P.Oxy 3.405 (kurz nach 200),
dient es dem Verweis einer altkirchlichen Quelle auf das Neue Testament. Vgl. U. Schmid,
“Die Diplé: Einführung,“ in Von der Septuaginta zum Neuen Testament: Textgeschichtliche
Erörterungen (ANTF 43; Hg. M. Karrer, S. Kreuzer und M. Sigismund; Berlin / New York:
De Gruyter, 2010), 77–81, hier 78.
60 Weiteres in den einschlägigen Beiträgen in Von der Septuaginta zum Neuen Testa-
ment: Textgeschichtliche Erörterungen (ANTF 43; Hg. M. Karrer, S. Kreuzer M. Sigismund;
Berlin / New York: De Gruyter, 2010), zur Schreibung bes. 84–91 (bes. 84), 117–122.
61 Wenn Zitate, die von den neutestamentlichen Autoren nicht durch Zitatmarker ein
geleitet wurden, die Diplé tragen, müssen wir dies deshalb durch ein Allgemeinwissen der
Skriptorien um biblische Texte begründen (s. im Vaticanus 1Kor 10:26 [Zitat aus PsLXX 23:1]
und Eph 6:2 [Zitat aus Ex 20:12], im Alexandrinus Eph 5:31 [Zitat aus Gen 2:24]).
nicht. Vielleicht erkannte der Auszeichner durch seine allgemeine Schrift

kenntnis, dass Jak 4:5 keine LXX-Vorlage besitzt.
Interessanterweise steht die Diplé bei keinem der erwähnten inneralt
testamentlichen Zitate, also nicht bei 2Kön (LXX 4Kgt) 14:6/ 2Chr 25:4
(nach Dtn 24:16) und Jer 26:18 (JerLXX 33:18) (nach Mi 3:12). Es ist im Zitat
bereich ein spezifisches Zeichen für Rückverweise vom Neuen Testament
aus62.
Diese Zeichen stellen eine explizite Referenz der neutestamentlichen
Schriften zum je zitierten Text und damit der Septuaginta her. Die Herstel
lung von Konkordanz im Wortlaut könnte folgen. Doch die Skriptorien ver
zichten auf einen solchen Abgleich. Obwohl sie Vollbibeln herstellen und
deshalb material mit Sicherheit Vorlagen für Septuaginta- und neutesta
mentliche Schriften besitzen, prüfen sie nicht wechselseitig den Wortlaut.
An den Rand von Apg 13:41 hat ein Schreiber ΙΩΗΛ eingefügt. Das Zitat in Apg 13:41
stammt aber tatsächlich aus Hab 1:5. Die Abbildung stammt aus der digitalen Edition
(www.sinaiticus.de).
Abb. 2 Marginalie zu Apg 13:41 im Codex Sinaiticus
Dieser Verzicht auf Prüfung lässt sich durch eine zweite Erscheinung
nachweisen: Der Codex Sinaiticus nennt oft in einer Marginalie neben
der Diplé oder an ihrer Stelle die Schrift, aus der das Zitat nach Ansicht
des Skriptoriums stammt—und irrt sich wiederholt (‫ א‬Mt 2:6 usw.).
62 Im Septuaginta-Bereich ist die Diplé damit frei dafür, als Verweiszeichen in allge
meinem Sinn verwendet zu werden. Das geschieht in späterer Zeit stellenweise: In Codex
19 (12. Jh.) steht sie am Rand eines Teils der poetischen Totenklage Davids in 2Kgt 1:22–27
(fol. 189v; Auszeichnung von 8 Zeilen). Für den Hinweis danken wir Marcus Sigismund.
Solche Irrtümer setzen sich noch lange fort. So fügt eine spätantike Hand63
( jünger als die prima manus und flüchtig schreibend, aber noch Majuskel,
daher 5.–8. Jh.) im Sinaiticus neben Apg 13:41 ein: ΙΩΗΛ. Das Zitat stammt
indessen nicht aus Joel, sondern aus Hab 1:5 (Abb. 2).
Es versteht sich von selbst, dass Zitat und zitierter Text weder im Codex
noch durch den Schreiber der Marginalie aufeinander abgestimmt wer
den; Sinaiticus Hab 1:5 bietet mit der Wendung εργον εγω εργαζομαι (“ein
Werk tue ich“) nach heutigem Wissen das Old Greek (eine Wiedergabe
der etymologischen Figur ‫)פ ַֹעל ּפ ֵֹעל‬. ‫( *א‬prima manus) Apg 13:41 dage
gen schreibt (irrig oder nach einer anderen Auflösung der hebräischen
Phrase) εργον ο εγω εργαζομε εγω („ein Werk, das ich tue, ich“). Der bedeu
tende Korrektor ca löscht das Pronomen (ο), belässt jedoch die Doppe
lung des εγω in Apg. Die Korrektur des neutestamentlichen Textes erfolgt
also unabhängig vom Septuagintatext desselben Codex.
Blicken wir kurz zur Seite: Die von Hab 1:5 abweichende Wortstellung
ἐργάζομαι ἐγώ ist in Apg 13:41 vorzüglich bezeugt (P74 ABD; nur die Doppe
lung des εγω ist ein Proprium des Sinaiticus). Das neutestamentliche Zitat
bietet eine alte Nebenform des Septuagintatextes. Dritte Zeugen ab dem
5. Jh. allerdings korrigieren die Wortstellung doch noch; erstmals C (5. Jh.)
bietet εγω εργαζομαι wie das Old Greek des Dodekapropheton (danach
E08 [6. Jh.] und 097 [7. Jh.] u.a.). Das kann eine innerneutestamentliche
stilistische Korrektur sein. Aber etwas wahrscheinlicher scheint ein Ein
fluss des Septuaginta-Haupttextes.
Tabelle 13 2Kor 9:9 und PsLXX 111:9

LXX Ps 111:9 (Gö nach ἐσκόρπισεν, ἔδωκεν τοῖς πένησιν. ἡ δικαιοσύνη αὐτοῦ μένει
AS und anderen Zeugen; εἰς τὸν αἰῶνα τοῦ αἰῶνος
in B fehlt PsLXX 111:9)
2Kor 9:9 (NA27 nach BS ἐσκόρπισεν, ἔδωκεν τοῖς πένησιν, ἡ δικαιοσύνη αὐτοῦ μένει
und anderen Zeugen; in εἰς τὸν αἰῶνα
A fehlt 2Kor 9:9)
2Kor 9:9 (mit ἐσκόρπισεν, ἔδωκεν τοῖς πένησιν. ἡ δικαιοσύνη αὐτοῦ μένει
Ergänzung in FGK und εἰς τὸν αἰῶνα τοῦ αἰῶνος
anderen Zeugen ab
dem 9. Jh. nach NA27)
63 U. Schmid, „Diplés und Quellenangaben im Codex Sinaiticus,“ in Von der Septuaginta
zum Neuen Testament: Textgeschichtliche Erörterungen (ANTF 43; Hg. M. Karrer, S. Kreuzer
und M. Sigismund; Berlin / New York: De Gruyter, 2010), 83–98, hier 88–90, weist sie dem
Schreiber A zu, die Quellenangabe des Zitates in Apg 13:41 gehöre zur ursprünglichen Pro
duktionseinheit im Skriptorium.
Es würde sich lohnen, die Entwicklung der Handschriften über den Aus
gang der Antike hinaus zu verfolgen. Denn einiges spricht dafür, dass nun
die Bemühung um Konkordanz zunimmt (ohne dass sich der Gebrauch
der Diplé allgemein durchsetzen würde). So ergänzen jüngere Hand
schriften in 2Kor 9:9 τοῦ αἰῶνος nach PsLXX 111:9 (Tabelle 13). Für uns ist
vor allem bemerkenswert: Wo sich Einflüsse feststellen lassen, führt das
Zitationsbewusstsein zu einer Anpassung der neutestamentlichen Über
lieferung an den alttestamentlichen Bezugstext. Die frühere, für Rahlfs’
Septuaginta-Edition wesentliche Annahme, der Septuagintatext sei umge
kehrt dem Neuen Testament angepasst worden, ist schwer zu belegen.
5.3. Markierungen in der Septuaginta

Der Befund der großen Codices zur Septuaginta korrespondiert dem. Die
Skriptorien der Vollcodices führen kein Zeichen für Vorverweise aufs
Neue Testament ein. D.h. sie setzen an keiner einzigen Stelle der Sep
tuaginta, die vom neutestamentlichen Teil dieser Codices aufgegriffen
und wiederholt wird, ein Vorverweiszeichen. Das Zitatbewusstsein der
Alten Kirche besitzt demnach eine entscheidende Richtung: Die jüngeren
frühchristlichen Schriften fußen auf den älteren heiligen Schriften und
weisen auf sie zurück. Das Alte dagegen bedarf zu seiner Wahrnehmung
das Neue nicht in gleicher Notwendigkeit. Dass es weniger Einflüsse des
Neuen Testaments auf die Septuaginta gibt als früher vermutet, gehört in
den Kontext der antiken Hochschätzung des Alten, Vorangehenden.
Jes 42:1–5 in 1QJesa (200 v.Chr.); in der Hs. Dtn 17:5 in Codex Sarravianus (IV/V); der
finden sich insgesamt 11 solcher Kreuze, von Origenes nach Proto-MT hinzugefügte
die vermutlich ausgewählte Passagen Text ist durch Asteriskoi am Anfang der
hervorheben sollen; im Einzelnen ist unklar, Einfügung und in den Marginalien angezeigt;
ob sie sich auf den Text rechts oder links die Funktion des Metobelos am Ende des
beziehen. (Abb. nach der digitalen Edition Einschubs übernimmt der Hochpunkt.
dss.collections. imj.org.il/isaiah) (Abb. nach www.katapi.org.uk/BibleMSS/
G.htm)
Abb. 3 Marginalzeichen in 1QJesa und Codex Sarravianus

Doch gibt es auch in jüdischen Handschriften seit vorchristlicher Zeit

Marginalmarkierungen: Abbildung 3 zeigt ein dem X ähnelndes Zeichen
in 1QJesa, das vermutlich ausgewählte Passagen hervorheben sollte64. In
Septuaginta-Handschriften verbreiten sich hexaplarische Zeichen, die
bei längeren Abschnitten zu jeder Zeile treten (Asteriskos/Obelos und
Metobelos; s. Abb. 3) und darin graphisch den Diplé-Markierungen (§5.2)
ähneln. So stellt sich die Frage, ob nicht in etwas jüngerer Zeit der Bezug
zum Neuen Testament durch Kennzeichnungen hergestellt wird und sich
textgeschichtlich auswirkt.
Betrachten wir dazu zunächst den Codex Ambrosianus aus der 1. Hälfte
des 5. Jh., der besonders reichhaltige Anmerkungen enthält65. Hier finden
sich neben anderen Markierungen verschieden gestaltete Kreuzzeichen.
64 Die Bedeutung und Verwendung der Zeichen im Einzelnen ist unklar—einschließ

lich der Frage, ob sie den Abschnitt in der linken oder rechten Spalte markieren. Doch
dass es sich um „Marks Drawing Attention to Certain Matters in the Text” handelt, dürfte
derzeitiger Forschungskonsens sein (so u.a. O. H. Steck [Die erste Jesajarolle von Qumran
(1QIsa): Schreibweise als Leseanleitung für ein Prophetenbuch (SBS 173/1; Stuttgart: Katho
lisches Bibelwerk, 1998), 33 Anm. 87] und E. Tov [“Scribal Markings in the Texts from
the Judean Desert,“ in Current Research and Technological Developments on the Dead Sea
Scrolls (STDJ 20; Hg. D. W. Parry und S. D. Ricks; Leiden: Brill, 1996), 41–77, hier 65f.], von
dem das Zitat stammt). Vgl. zur Problemlage und für weitere Literatur Steck, Erste Jesaja-
rolle, 33 mit Anm. 87 und E. Tov, „Scribal Notations in the Texts from the Judaean Desert,“
in The Texts from the Judaean Desert: Indices and an Introduction to the Discoveries in the
Judaean Desert Series (DJD 39; Hg. idem; Oxford: Clarendon Press, 2002), 323–349, hier
insb. 339f., und idem, Textual Criticism, 103, 205.
65 Zu nennen sind u.a. Marginalnotizen, Textvarianten, Gliederungshilfen, Zählungen
(bei den Geboten) usw. Vgl. die aktuellen Arbeiten zu Codex Ambrosianus: M. Fincati
(“Per la storia dell’Esateuco Ambrosiano A 147 inf.,“ Aevum 83 [2009]: 299–339 und eadem,
„New Testament Quotations and the Text of the Ambrosian Hexateuch,“ in Die Septuaginta
III. Entstehung, Sprache, Geschichte [WUNT]; Hg. S. Kreuzer, M. Meiser und M. Sigismund;
Tübingen 2012 [im Erscheinen]) und C. Boyd-Taylor (“The Greek Bible Amongst Jews and
Christians in the Middle Ages—The Evidence of Codex Ambrosianus,“ in XIII Congress of
the International Organization for Septuagint and Cognate Studies, Ljubljana 2007 [SBLSCS
55; Hg. M. K. H. Peters; Atlanta: Society of Biblical Literature, 2008], 29–39 [Nachdruck:
Leiden, 2010]), hier auch weitere einleitende Literatur zu Codex F. Boyd-Taylor befasst sich
insb. mit den mittelalterlichen Anmerkungen im Codex (Fb); vgl. hierzu auch die dem
nächst erscheinende Dissertation von Mariachiara Fincati. Zum Text von F vgl. den Bei
trag von J. de Vries (in Textgeschichte und Schriftrezeption im frühen Christentum / Textual
History and the Reception of Scripture in Early Christianity [SBLSCS]; Hg. J. de Vries und
M. Karrer; 2012 [im Erscheinen]).
Die hier vorgestellten Beobachtungen wurden anhand eines Mikrofilms des Göttin
ger Septuaginta-Unternehmens gemacht. Etliche Markierungen der ersten Hand, die
A. M. Ceriani (Pentateuchi et Josue quae ex Prima Scriptura Supersunt in Codice Ambrosiano
Graece Seculi Fere V [MSP 3; Mailand: Bibliotheca Ambrosiana, 1864]) notierte, sind auf
diesem nicht zu erkennen, was die folgenden Beschreibungen eventuell beeinträchtigt.
Gliederungszeichen Gliederungszeichen Markierung am Hervorhebungszei-

vor Lev 20:6; dieses vor Num 12:11; es Beginn des Buches chen neben Num
Zeichen geht steht neben dem Arithmoi (Num 1:1). 24:17. Der Vers
vermutlich auf die Paragraphos; (ἀνατελεῖ ἄστρον
erste Hand zurück. möglicherweise ἐξ Ἰακώβ) hat
wurde es erst im 16. christologische
Jh. ergänzt. Relevanz; vgl. Apk
22:16).
Abb. 4 Kreuzzeichen im Codex Ambrosianus66
Dabei sind aufgrund äußerlicher Gestaltung und vermutlicher Funktion

zwei Fälle zu unterscheiden:
(1) In fast allen Teilen des Codex finden sich in unterschiedlicher Häufig
keit Kreuzzeichen, die die Textgliederung unterstützen67.
So sind einige Paragraphoi nicht als einfacher Querstrich, sondern als
Kreuz ausgeführt; beides geht nach Ceriani (1864), der das Originalmanu
skript studierte, auf die erste Hand zurück68. Hinzu kommen viele Stellen
(insb. in den hinteren Büchern Num und Dtn), an denen der Paragraphos
66 Fotos in Abb. 4 und 5: Johannes de Vries vom Mikrofilm beim Septuagintaunterneh

men Göttingen. Abdruckgenehmigung erteilt vom Rechteinhaber (© Veneranda Biblioteca
Ambrosiana—Milano/De Agostini Picture Library).
67 Hinzu kommen Sonderfälle wie das singuläre Kreuzzeichen, das den Anfang des
Buches Numeri anzeigt. Fincati (“La storia,“ 303) erwägt eine Datierung dieses Zeichens in
die Spätantike oder das Frühmittelalter.
68 Zur Abschnittsmarkierung finden sich in Codex Ambrosianus neben dem klassi
schen Paragraphos (–) noch das Kreuz (+) und die Diplé (>). Vgl. z.B. die Zeichen vor Lev
20:6 (+ Kreuz), 20:9 (– Paragraphos) und 20:12 (> Diplé). Diese markieren je einen neuen
Abschnitt, von Ceriani (Pentateuchi et Josue) werden sie der ersten Hand zugerechnet und
in seinem Transkript mit = (für + und >) bzw. mit—(für—) wiedergegeben.
später um ein nebengeschriebenes Kreuz erweitert wurde69. Kreuzzei

chen als Layoutmarker zur Textgliederung finden in den ntl. Teilen von
Codex Alexandrinus und Codex Sinaiticus eine Parallele70. Doch zeigen
sie in Codex F keine besondere Affinität zu frühchristlich aufgegriffenen
Passagen. Eine Orientierung an Zitatvorlagen für die Schriften des Neuen
Testamentes lässt sich also nicht feststellen.
(2) Neben diesen relativ häufig verwendeten Gliederungskreuzen steht an

vereinzelten Stellen ein (meist etwas größeres) Kreuz, das einzelne Passa
gen hervorhebt und schon aufgrund seiner Form für unsere Fragestellung
besonders relevant ist: Es ist (zumindest mehrfach) als Staurogramm, also
als explizit christologisches Zeichen gestaltet71. Markante Hervorhebun
gen betreffen (vgl. außerdem unten Dtn 18:15f., 18):
• Ex 3:6: ᾿Εγώ εἰμι ὁ θεὸς τοῦ πατρός σου.

• Num 24:7: ἐξελεύσεται ἄνθρωπος ἐκ τοῦ σπέρματος αὐτοῦ. Der Vers liest
weiter: καὶ κυριεύσει ἐθνῶν πολλῶν, καὶ ὑψωθήσεται ἢ Γὼγ βασιλεία αὐτοῦ,
καὶ αὐξηθήσεται ἡ βασιλεία αὐτοῦ.
69 Vgl. die Markierungen (+) neben den Paragraphi zu Num 12:2–14 (bei Ceriani, Pen-
tateuchi et Josue, 226). Diese sind nicht der ersten Hand zuzurechnen. Mariachiara Fincati
war es möglich, diese Zeichen in der Originalhandschrift einzusehen, sie teilte uns dan
kenswerter Weise mit, diese seien in „light brown ink” („rostbraun”) ausgeführt und fänden
sich auch auf den Nachtragsseiten aus dem 11. Jh. Da dieselbe Hand darüber hinaus weitere
Anmerkungen im Codex (insb. Überschriften ab Num) einfügte, lässt sie sich auf das 16.
Jh. datieren (so ebenfalls Fincati).
70 So findet sich in einzelnen Büchern des Alexandrinus ein durch Kreuze angezeigtes
und durch beigefügte Zählung vertieftes Gliederungssystem (Kephalaia). Dabei ist dieses
System teilweise sehr eng (ca. 80 Markierungen in Lk), teilweise weiter (knapp 20 Mar
kierungen im Joh). Im Codex Sinaiticus lässt sich ebenfalls ein durch Kreuze realisiertes
Gliederungssystem beobachten, vgl. z.B. das Mt, wo sich diese Markierungen neben dem
farbig eingetragenen Eusebischen System finden. Die Einschnitte entsprechen ebenfalls in
etwa denen der Kephalaia, sind aber nicht durchgezählt (deutlich zu sehen z.B. vor Mt 3:1;
5:1; 5:11; 5:17 u.ö.; daneben übrigens auch hier dem Diplé ähnliche Gliederungszeichen, vor
5:3; 5:4 u.ö.). Eine ausführliche Erörterung der Kephalaia findet sich bei H. Freiherr von
Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt herge-
stellt aufgrund ihrer Textgeschichte, Band 1,1: Untersuchungen: Die Textzeugen (Göttingen:
Vandenhoeck & Ruprecht, 21911), 402–432.
71 Eventuell sind nicht alle Zeichen gleich ausgeführt, könnten also von verschiedenen
Händen stammen, doch lässt sich dies auf dem Mikrofilm nicht sicher entscheiden. In
zumindest vielen Fällen (vielleicht allen?) ist es als Staurogramm (= Rho mit Querstrich)
gezeichnet, wie in Abb. 18 ersichtlich. Das Staurogramm verbindet das Chi-Rho-Mono
gramm mit dem Kreuzeszeichen, indem es das Chi um seinen Schnittpunkt dreht. Zur
Relevanz des Staurogramms in frühen christlichen Handschriften vgl. L. W. Hurtado, „The
Staurogramm in Early Christian Manuscripts: The Earliest Visual Reference to the Cruci
fied Jesus?” in New Testament Manuscripts: Their Texts and Their World (TENT 2; Hg. T. J.
Kraus und T. Nicklas; Leiden / Boston: Brill, 2006), 207–226.
• Num 24:17: ἀνατελεῖ ἄστρον ἐξ Ἰακώβ, der Vers liest weiter: καὶ ἀναστήσεται
ἄνθρωπος ἐξ Ἰσραήλ.
Neben diesen Stellen sind noch einige weniger eindrückliche markiert

(z.B. Dtn 9:10; 29:17 etc.)72, doch der eklektische Eindruck, den diese Mar
kierungen von möglicherweise verschiedenen Händen vermitteln, bleibt.
Zwar besitzen einige der genannten Stellen christologische Signifikanz
(z.B. Num 24:17; vgl. Apk 22:16), und einzelne werden neutestamentlich
zitiert (Ex 3:6 in Mt 22:32 parr. und Apg 3:13; 7:32). Doch ergeben sich
die Markierungen nicht als Folge des Zitats, sondern betreffen auch nicht
zitierte Stellen (Num 24:7 wird im NT nicht christologisch aufgegriffen).
Mehr noch, selbst an den Stellen, die im NT deutlich aufgegriffen sind,
passt kein Korrektor den Ambrosianustext an den Text des NT an und
ergänzt keine Marginalie einen explizierenden Vorverweis auf die rezipie
rende Schrift des Neuen Testaments.
Eine Kreuzmarkierung (Staurogramm) neben Dtn 18:18 hebt den Vers deutlich hervor. Auf
dem Bild ist auch der Unterschied zu den Gliederungsmarkierungen (kleines Kreuzzei
chen neben Paragraphos) vor Dtn 18:17, 21 zu erkennen. (Vgl. Anm. 65.)
Abb. 5 Codex Ambrosinanus—Markierung bei Dtn 18:18
72 Die Liste ist keinesfalls vollständig. Außerdem ist auf dem Mikrofilm oft nicht ein
deutig zu erkennen, ob die Kreuze zum Gliederungssystem gehören (wodurch freilich
auch verschiedene Verse markiert werden), oder ob einzelne Passagen explizit hervorge
hoben werden sollen. Die hier genannten Beispiele sind deutlich unterscheidbar von den
Kreuzzeichen, die als Gliederungsmarker fungieren. Zur Markierung in Ex 3:6 vgl. Fincati,
„La storia,“ 301 mit Anm. 17. Nicht ganz sicher ist z.B. die Funktion der Kreuze in Ex 3:12;
Ex 24:1–3. Dtn 29:17 ist insofern besonders wahrzunehmen, als der Text von F hier eine
besondere Nähe zu seiner NT-Anspielung in Hebr 12:15 aufweist. Allerdings teilt F diese
Nähe mit A; sie ist vermutlich nicht mit einer Angleichung an den neutestamentlichen
Text zu erklären (vgl. de Vries, s. Anm. 65).
Tabelle 14 Dtn 18:15f., 18 und Apg 3:22; 7:37

Apg 3:22 (NA27) ὅτι προφήτην ὑμῖν ἀναστήσει κύριος ὁ θεὸς ὑμῶν ἐκ τῶν ἀδελφῶν
ὑμῶν ὡς ἐμέ·
αὐτοῦ ἀκούσεσθε κατὰ πάντα ὅσα ἂν λαλήσῃ πρὸς ὑμᾶς.
Apg 7:37 (NA27) προφήτην ὑμῖν ἀναστήσει ὁ θεὸς ἐκ τῶν ἀδελφῶν ὑμῶν ὡς ἐμέ
Dtn 18:15f (Gö = F) προφήτην ἐκ τῶν ἀδελφῶν σου ὡς ἐμὲ73 ἀναστήσει σοι κύριος ὁ
θεός σου,
αὐτοῦ ἀκούσεσθε (16) κατὰ πάντα, ὅσα ᾐτήσω . . .
Dtn 18:18 (Gö + F) προφήτην ἀναστήσω αὐτοῖς ἐκ [F add. μέσου] τῶν ἀδελφῶν
αὐτῶν ὥσπερ σέ
Das auffälligste Beispiel bestätigt die Beobachtung. Neben Dtn 18:15 und
18:18 befinden sich gleich zwei Kreuzmarkierungen (neben 18:18 offen
sichtlich ein Staurogramm [vgl. Abb. 5], bei 18:15 ist dies möglich, aber im
Mikrofilm nicht sicher zu erkennen), die sich deutlich von denen neben
den Paragraphoi unterscheiden (vgl. auf derselben Seite des Folios die
Zeichen vor 18:17, 21; 19:1). Sie spiegeln das frühchristliche Interesse an der
christologischen Bedeutung des Wortes: Christus gilt seit Apg 3:22; 7:37 als
der in der Schrift erwartete „Prophet nach Mose.“ Trotzdem werden die
Texte im Dtn nicht an die Schlüsselstellen in Apg angeglichen. Dtn 18:18
weicht in F sogar stärker vom Text in Apg 3:22; 7:37 ab als der kritisch
rekonstruierte Text: F liest ἐκ μέσου τῶν ἀδελφῶν αὐτῶν anstatt des kürze
ren und auch vom NT bezeugten ἐκ τῶν ἀδελφῶν αὐτῶν74.
Diese Kreuzeszeichen im Ambrosianus sind möglicherweise sehr alt.
Aber Sicherheit ist darüber nicht zu gewinnen.
Ceriani notiert sie nicht mit den übrigen Zeichen der ersten Hand,
diskutiert sie jedoch als Zeichen, die „in diesem Zeitalter, welches wir
den ältesten Codices zuordnen, und danach” weite Verbreitung fanden75.
Mariachiara Fincati schreibt sie der ersten Hand zu76. Doch sind viele
73 Ὡς ἐμέ fehlt im Grundtext, wird aber noch von der ersten Hand nachträglich
eingefügt.
74 Dtn 18:15f ist die für das NT-Zitat insgesamt relevantere Vorlage. Anders als 18:18 ist
18:15f (i) Moserede (wie die Zitate in Apg 3:22; 7:37), so findet sich auch nur in Dtn 18:15f
(ii) das Subjekt (κύριος) ὁ θεός und (iii) das kürzere ὡς ἐμέ. Auch (iv) der zweite Teil des
Zitates in Apg 3:22 (αὐτοῦ ἀκούσεσθε κατὰ πάντα ὅσα) stammt aus Dtn 18:15f.
75 Ceriani, Pentateuchi et Josue, xix–xxi; vgl. xxi: „ea fere aetate, quae vetustissimis
codicibus tribuitur, et deinceps, idem monogramma occurrit in Inscriptionibus Graecis
totius fere veteris orbis Christiani.“
76 So Mariachiara Fincati in einer persönlichen Mitteilung nach Einsicht der relevanten
Stellen im Original. Als terminus ante quem ist aber in jedem Fall das 11. Jh. zu sehen,
Randnotizen des Codex mittelalterlich (vgl. Anm. 66). Die Überlegungen

zu den Markierungen (Interesse an markanten Stellen ohne Textanpas
sung ans NT) gelten also wenigstens für die frühe, vielleicht sogar für die
spätere Zeit (bis spätestens zum 11. Jh., vgl. Anm. 76).
Wenden wir uns deshalb noch einem zweiten, jüngeren Codex zu,
dem Codex Chludov (vermutlich kurz nach 843 entstanden)77. Die beiden
gefundenen Aspekte—das Interesse an markanten, christologisch oder
sonst relevanten Stellen und das Desinteresse an einer Abgleichung des
Wortlauts zum neutestamentlichen Text—vertiefen sich dort. Das Kreuz
zeichen hebt hier ebenfalls—neben anderen Zeichen—einzelne Passagen
hervor, die im Codex Chludov nun zum Teil illustriert werden78. Mehrfach
werden durch die Abbildung Psalmtext und neutestamentliche Motive
verbunden. Doch es bleibt bei der motivischen Verknüpfung. Auf genaue
textliche Konkordanz wird unverändert verzichtet.
Illustrieren wir das anhand von Folio 88r. (Die Wiedergabe eines Fotos
ist leider nicht möglich, so dass wir uns mit der Beschreibung begnügen.)
Auf dieser Seite markiert ein Kreuz PsLXX 88:10 (σὺ δεσπόζεις τοῦ κράτους
τῆς θαλάσσης τὸν δὲ σάλον τῶν κυμάτων αὐτῆς σὺ καταπραΰνεις). Der untere
Teil des Folios ist mit der neutestamentlichen Erzählung von der Sturm
stillung (Tempestas Sedata, Mt 8:23–27 parr.) illustriert. Ein zweites Kreuz
neben dem Kopf Jesu verknüpft das Psalmwort mit der Darstellung, die
um die Beischrift κινδυνεύοντος τοῦ πλοῦ ἐπιτιμῶν ὁ ἸΣ τῷ ἀνέμῷ καὶ τῇ
θαλάσσῃ ergänzt ist. Nicht einmal Beischrift und Septuagintatext werden
aufeinander abgeglichen.
in welchem ein Restaurator zumindest einen Teil dieser Zeichen verstärkte (so bei Dtn
32:43). Vgl. auch Fincati, „La storia,“ 300.
77 Für eine aktuelle Einführung zu Codex Chludov (Moskau, Hist. Mus. Cod. 129,
Chludov-Psalter) siehe R. Stichel, Beiträge zur frühen Geschichte des Psalters und zur
Wirkungsgeschichte der Psalmen (Abhandlungen der Nordrhein-Westfälischen Akademie
der Wissenschaften 116; Paderborn: Schöningh, 2007), 141–150, hier auch weitere Litera
tur; für eine Einordnung der Miniaturen in die Kunstgeschichte vgl. die Arbeit von K. A.
Corrigan, Visual polemics in the ninth-century Byzantine psalters (Cambridge: Cambridge
University Press, 1992) und die Übersicht von S. Dufrenne, Tableaux synoptiques de 15 psau-
tiers médiévaux (Paris: Centre national de la recherche scientifique, 1978); die Miniaturen
sind in der russischen Faksimile-Ausgabe zugänglich (M. V. Ščepkina, Miniatjury Chludov-
skoj psaltyri: grečeskij illjustrirovannyj kodeks IX. veka [Moskau, 1977]).
78 Auch jenseits von Illustrationen hebt das Kreuzzeichen einzelne Stellen hervor, die
die Bearbeiter für bedeutsam erachteten: So sind auf Folio 66v beispielsweise sieben Zei
len von PsLXX 68 ( jeweils die ausgerückten Zeilen Ps 68:14–20) mit Kreuzen vor der Zeile
markiert, ohne damit auf ein Bild zu verweisen. Oft sind diese Markierungen als Verweis
zeichen zu lesen, die ein Psalmwort mit einer Illustration verknüpfen. Dabei entstehen an
vielen Stellen Verbindungen zum Neuen Testament.
Sturm (ὁ ἄνεμος) und Meer (θάλασσα) sind in der Szene personifiziert

dargestellt und ebenfalls beschriftet. Psalmwort und Sturmstillung sind
durch die zentralen Motive des Meeres und der Wogen (κυμάτων Mt 8:24
wie im Ps) verbunden, doch auf den Psalmtext wirkt sich das nicht aus.
Die Marginalie nennt nicht einmal das Parallelmotiv der κύματα aus PsLXX
88:10/Mt 8:24.
Ein zweites Beispiel, Ps 2:7b, ist gar mit einem roten Kreuz und je einem
roten Paragraphos vor und hinter dem Vers versehen (Abb. im vorliegen
den Band ebenfalls nicht abdruckbar); dem Psalmwort wird eine doppelte
Markierung zuteil. Der Text des Psalmworts lautet wie in Gö: κύριος εἶπεν
πρός με υἱός μου εἶ σύ ἐγὼ σήμερον γεγέννηκά σε. Die nebenstehende Illu
stration von der Geburtsszene Jesu zeigt die christologische Rezeption des
Wortes in der Weihnachtsgeschichte. Doch entspricht das der byzantini
schen Frömmigkeit, nicht den frühchristlichen Zitaten.
Denn nach dem Zitat in Apg 13:33 wäre Ps 2:7 auf die Auferstehung zu
beziehen, nach demjenigen in Hebr 5:5 auf Jesu Berufung zum Hoheprie
ster; auch eine Anspielung auf die Taufe wäre möglich (Mt 3:17; Lk 3:22).
Aber all dies übergeht der Chludov-Psalter.
Solche Markierungen in Septuagintahandschriften gehören demnach
mehr in die Frömmigkeits- als in die Textgeschichte. Es liegt in der Kon
sequenz dieses Vorgehens, dass sich im Codex Chludov auch sonst keine
auffälligen Anpassungen an den Text des NT finden.
Runden wir die Beobachtungen durch ein letztes Beispiel ab (auch
dies leider ohne Abbildung): Der Codex interessiert sich für die im
Neuen Testament aufgegriffenen Psalmworte, so PsLXX 68:10, zitiert bei
der Tempelreinigung in Joh 2:17. Ein Schreiber markiert den Psalmvers
(Folio 66r) mit einem schwarzen Kreuz und illustriert ihn mit der neu
testamentlichen Purgatio Templi (Mt 21:12f parr.; Joh 2:14–17). Die Abbil
dung des Codex lässt sich nicht sicher einem der Evangelien zuordnen,
allerdings wird die dargestellte Geißel nur im Joh explizit erwähnt. Das
Joh ist auch das einzige Evangelium, das das Psalmwort im Zusam
menhang mit der Tempelreinigung zitiert (Joh 2:17). So entsteht ein
deutlicher Vorverweis auf Joh 2:17. Doch wird das Psalmwort in seinem
Wortlaut nicht an Joh 2:17 angepasst. Der Psalter Chludov bietet wie Gö
Tabelle 15 PsLXX 68:10 und Joh 2:17

PsLXX 68:10 (Gö = Codex Chludov) ὅτι ὁ ζῆλος τοῦ οἴκου σου κατέφαγέν με
Joh 2:17 (NA27 ohne Varianten) ὁ ζῆλος τοῦ οἴκου σου καταφάγεταί με
den Aorist Aktiv, während die neutestamentliche Überlieferung (übri

gens zusammen mit dem Psalter von B und S) Futur Medium liest (vgl.
Tabelle 15).
Vielleicht wurde trotz der Nähe nicht einmal der besondere Zusam
menhang zum Joh erkannt. Denn der Beitext zur Zeichnung merkt an:
ἐκ τοῦ κατὰ Ματθαῖον. Der Schreiber dieses Beitextes (der Illustrator oder
eine spätere Hand) unterstreicht den Vorverweis aufs Neue Testament,
orientiert sich dabei aber nicht am textlich nächst verwandten Evange
lium, dem Joh, sondern am kirchlich weitest verbreiteten, dem Mt.
Das Fazit ist rasch gezogen: Markierungen in christlichen Septuaginta
handschriften bekunden—bis weit in die byzantinische Zeit hinein—das
Interesse, für das christliche Leben wichtige Stellen hervorzuheben. Doch
dieses Interesse gilt dem Motiv, nicht dem Wortlaut. So beeinflusst es den
Septuagintatext nicht. Seine Entwicklung erfolgt im Wesentlichen unab
hängig von den neutestamentlichen Zitaten.
6. Ergebnis
Die Ergebnisse sind damit eindeutig:
a. Die Geschichte des Zitierens beginnt vor dem frühen Christentum. Das
Zitat kann dabei Nebenformen des zitierten Textes aufgreifen. Eine
Angleichung des Textes erfolgt bei den Zitaten innerhalb der Septua
ginta nicht.
b. Die neutestamentliche Textüberlieferung knüpft daran an. Sie nimmt
wahr, wo aus heiligen Schriften Israels zitiert wird, und kennzeichnet
das allmählich. Aber daraus erwächst kein großer Druck, die zitie
renden Stellen an die Referenzpassagen der Septuaginta anzupassen.
Sekundäre Anpassungen der neutestamentlichen Zitate an den Sep
tuaginta-Haupttext erfolgen nur gelegentlich und meistens spät.
c. Auf diese Weise bleiben uns durch die frühchristlichen Schriften,
namentlich das Neue Testament, Nebenformen der Septuaginta-
Überlieferung und Einflüsse jüngerer Übersetzungen bewahrt, die sonst
verloren gingen; wir betrachteten Beispiele bis zu [Proto-] Symmachus.
d. Die frühchristliche Septuagintaüberlieferung blieb durch die neutesta
mentlichen Zitate bis zur Spätantike fast unbeeinflusst. Die Skripto
rien sahen bis ins 5. Jh. keine Notwendigkeit, von der Septuaginta aufs
Neue Testament vorauszuweisen. Spätere Verweise folgen der Fröm
migkeitsgeschichte und sind an einer textlichen Konkordanz zu den
neutestamentlichen Zitaten nicht interessiert. Ein Einfluss neute

stamentlicher Überlieferung auf die Septuaginta wird dadurch sehr
fraglich.
Damit können wir zur Forschungsdiskussion Stellung nehmen, bei der wir
begannen: Kraft und Hanhart hatten Recht damit, die ältere These eines
wechselseitigen Einflusses zwischen Septuaginta- und neutestamentlicher
Überlieferung zu hinterfragen. Von der These, die Septuaginta habe neu
testamentliche Zeugen sekundär beeinflusst, bleiben immerhin Spuren;
einzelne Handschriften weisen Einflüsse auf, doch diese Einflüsse erfas
sen nicht die Gesamtüberlieferung und sind daher leicht als sekundär zu
erkennen. Gravierender ist der Einschnitt gegenüber der These von Alfred
Rahlfs, Übereinstimmungen zum NT in der LXX gingen auf eine Anpas
sung der alten Überlieferung an das Neue Testament zurück. Alle Stellen,
an denen Rahlfs ein „ex Novo Testamento” o.ä. vermerkte, bedürfen der
Prüfung und möglicherweise einer Korrektur79.
Anhang 1: Die Wuppertaler Datenbank
Die Datenbank ist online verfügbar unter www.kiho-wb.de/lxx_nt. Zwei

Möglichkeiten des Zugriffs werden geboten. Das intuitiv bedienbare Fron
tend „easyview” bietet einen Überblick über die wichtigsten Datensätze.
Daneben ist es möglich, auf alle Sätze der Datenbank mit ausschließlichem
Leserecht zuzugreifen. Dieser Zugang sei kurz vorgestellt. Das Erschei
nungsbild und der skizzierte Datenumfang entsprechen dem Stand bei
Ende des DFG-Förderungszeitraumes (Ende 2011).
Zu den neutestamentlichen Zitaten sind in der Datenbank erfasst (389
Verse):
• der Text von NA27 (als Basistext)

• Zitatmarker im neutestamentlichen Text (einleitende Formulierungen)
und in den Marginalien der großen Codices (Diplés)
79 Wir nannten in Anm. 4 und 5 beispielhaft PsLXX 13:3 Langtext und PsLXX 39:7. Diese
Stellen prüfte das Wuppertaler Forschungsprojekt daher exemplarisch: s. M. Karrer,
U. Schmid und M. Sigismund, „Textgeschichtliche Beobachtungen zu den Zusätzen in
den Septuaginta-Psalmen,“ in Die Septuaginta: Texte, Theologien und Einflüsse (WUNT 252;
Hg. W. Kraus, M. Karrer und M. Meiser; Tübingen: Mohr Siebeck, 2010), 140–161 und
M. Karrer, „Ps 39:7–10 in Hebrews 10:5–7,“ in Psalms and Hebrews: Studies in Reception (LBS
527; Hg. D. J. Human und G. J. Steyn; New York / London: T & T Clark, 2010), 126–146.
• die Varianten der Codices Sinaiticus, Alexandrinus und Vaticanus samt

Tätigkeit der Korrektoren
• die Varianten der griechischen Spalte des Codex Bezae (D05 gr), wie
derum samt Korrekturen
• die Varianten der neutestamentlichen Papyri bis ca. 300
• der byzantinische Text
• lateinische Versionen: Vulgata, Codex Bezae (d lat.) und Codex Usse
rianus
• syrische Versionen: Peshitta und die Vetus Syra Codices Curetonianus
und Sinaiticus, mit englischer und deutscher Übersetzung
• als deutsche Übersetzung: Neue Zürcher Bibel
• als englische Übersetzung: Revised Standard Version (RSV).
• Weitere griechische Handschriften können über die New Testament
Transcripts eingesehen werden.
An den Referenzstellen in der Septuaginta (356 Verse) bzw. der Hebrä

ischen Bibel sind erfasst:
• der Septuagintatext von Rahlfs’ Handausgabe (als Basistext)

• die Abweichungen dazu in der Göttinger Ausgabe
• die Septuagintavarianten der Codices Sinaiticus, Alexandrinus und
Vaticanus samt Tätigkeit der Korrektoren
• die Septuagintavarianten des Codex Ambrosianus (F)
• die Varianten der Septuagintapapyri bis ca. 500 (einschließlich Qumra
nica Graeca)
• ein rekonstruierter antiochenischer Text (mit Varianten)
• der hebräische Text: MT (nach BHS/BHQ; mit den in BHS/BHQ ver
zeichneten Varianten, Ketiv/Qere) und Varianten aus den Funden in
der Judäischen Wüste
• Schriftzitate bei Philo
• Schriftzitate bei Barnabas, 1 und 2 Clemens (Schriften, die auch in S/‫א‬
bzw. A enthalten sind)
• lateinische Versionen: Vulgata, bei den Psalmen iuxta Graecos und
Hebraeos
• syrische Version: Peshitta mit deutscher und englischer Übersetzung
• als deutsche Übersetzungen: Septuaginta Deutsch; Einheitsüberset
zung; Neue Zürcher Bibel (konsequente Übersetzung des hebräischen
Textes)
• als englische Übersetzungen: A New English Translation of the Septua
gint; L. C. L. Brenton, Septuagint; NRSV.
Über ein Pull-Down-Menü lassen sich die verschiedenen biblischen

Bücher auswählen, anschließend aus einer Liste die verzeichneten Verse
aus den jeweiligen Büchern. Die folgende Übersicht zeigt die Auswahl zu
Hebr 1:6, einer repräsentativen Stelle von mittlerem Datenumfang (der
Text links und rechts der Abb. ist in sich fortlaufend zu lesen).
1
Knapp 70 Einträge zu Hebr 1:6. 2
Zum ntl. Leittext werden die atl.
Zitatadressen mit ihren jeweiligen
Leittexten genannt.
Der byzantinische Text sowie Va- 3
rianten aus den großen Codices
(AB[D]S) werden aufgeführt.
Weitere Varianten zu den ntl. Stellen aus den Papyri, sowie die
Texte der Vulgata und Peshitta sind verfügbar.
Der atl. Teil ist zu jeder Stelle nach Sprachen (LXX, MT +
Qumran, Peshitta, Vulgata) unterteilt.
Zu den LXX-Stellen ist der rekonstruierte antiochenische Text
aufgeführt sowie die Varianten der großen Codices (ABS),
außerdem sind für den Pentateuch auch die Varianten des Codex
Ambrosianus verzeichnet.
Varianten aus den Handschriften vom Toten Meer sind verzeich-
net. 4
Sämtliche LXX-Papyri bis ca. 500 sind eingetragen.
Moderne Übersetzungen zu LXX (LXX.D, NETS, Brenton) und
MT sind verfügbar.
Ein Klick auf die verschiedenen Einträge ruft die jeweiligen Übersichten
auf. Vier Beispiele seien genannt:
Ein Klick auf die Leitstelle 1 zeigt den Basistext (Leittext) an:
Neues Testament (3313)

NT Buch Hebr
NT Kapitel 1
NT Begin Vers 6
NT Lesetext ὅταν δὲ πάλιν εἰσαγάγη τὸν πρωτότοκον
εἰς τὴν οἰκονµένην, λέγει· καὶ
προσκυνησάτωσαν αὐτῷ πάντες ἄγγελοι
θεοῦ
NT Basistext (ohne Akzente) οταν δε παλιν εισαγαγη τον πρωτοτοκον
εις την οικουµενην λεγει και
προσκυνησατωσαν αυτω παντες αγγελοι
θεου
Counted Text οταν2 δε4 παλιν6 εισαγαγη8 τον10
πρωτοτοκον12 εις14 την16 οικουµενην18
λεγει20 και22 προσκυνησατωσαν24
αυτω26 παντες28 αγγελοι30 θεου32
Der ntl. Leittext ist durchgezählt (hier 2–32), so dass alle Varianten sicher
zugeordnet werden können, mit ungraden Zahlen werden Additionen
verzeichnet.
Unter der NT-Zitat-Adresse 2 sind u.a. auch die Einleitungstexte ver
zeichnet:
NTZ Einleitungstext λέγει

NTZ Einleitungstext ohne Akzente λεγει
Ein Beispiel 3 : B liest zu Wort 32 (θεου) ein Nomen Sacrum (θυ):
Haupthandschrift B (11997)
B: Sichtvermerk nomen sacrum
B: Zitat ausgezeichnet? ja
B: Art der Zitats-Auszeichnung >
B: Zitatangabe der Hs.
B: Abweichender Text θυ
B: ggf.: orthographisch richtige θεου
Schreibweise der Abweichung
B: Hand der Abweichung B
B: Aktion 1. Hand
B: Abweichung zu Buch Hebr
B: Abweichung zu Kapitel 1
B: Abweichung, Beginn Vers 6
B: Abweichung, Beginn Wort 32
B: Abweichung, Beginn Vers 6
B: Abweichung, Beginn Wort 32
B: Sonderinformationen Die ersten beiden Zeilen
(Auszeichnungen etc.) ausgezeichnet
B: Anmerkungen/Notizen
Hier 4 ein Bsp. für eine Variante (Om[ission]) gegenüber dem LXX-
Leittext von Dtn 32:43; Bei allen Papyri (NT + LXX) wird stets der Volltext
angegeben, so dass der Benutzer sich bei schwierigen Entscheidungen ein
eigenes Bild machen kann, ggf. unterstützt von Sonderinformationen.
RA 848 (19299)
RA 848: Sichtvermerk om.
RA 848: Zitat ausgezeichnet?
RA 848: Art der Zitats-Auszeichnung
RA 848: Zitatangabe der Hs.
RA 848: Abweichender Text om. παντες
RA 848: ggf. orthographisch richtige
Schreibweise
RA 848: Versvolltext [ευφρανθητε ουρανοι αµα αυτωι και
προσκυνησατωσαν αυτωι υιοι θε]ου
[ευφρανθητε εθνη µετα του λαου αυτου
και ενισχυσατωσαν αυτωι παντες
αγγελοι θε]ου
RA 848: Hand der Abweichung scribe
RA 848: Aktion 1. Hand
RA 848: Abweichung zu Buch Dtn
RA 848: Abweichung zu Kapitel 32
RA 848: Abweichung, Beginn Vers 43
RA 848: Abweichung, Beginn Wort 16
RA 848: Abweichung, Ende Vers 43
RA 848: Abweichung, Ende Wort 16
RA 848: Sonderinformationen Die beschädigte Spalte ist zu schmal
für die vom Leittext gelesene Text
zeile, so dass die auch in anderen
Hss. bezeugte Omission von παντες
naheliegt.
RA 848: Anmerkungen/Notizen
Ιουδιθ und Iudith: Überlegungen zum Verhältnis der
Judit-Erzählung in der LXX und der Vulgata
Barbara Schmitz
„Das Buch Judit ist in zwei wesentlich voneinander unterschiedenen Text-

formen überliefert: einerseits in der griechischen Fassung und den sich
daran anschließenden alten Übersetzungen (vetus latina, syr., aeth., sah.,
arm.), zum anderen in der Fassung der lateinischen Vulgata und, mit die-
ser durchweg parallel laufend, in mittelalterlichen hebräischen Hss“1—so
fasst Helmut Engel die Textsituation des Buches Judit prägnant zusam-
men. Damit liegen heute zwei unterschiedliche Textformen der Juditer-
zählung vor: Die griechische Juditerzählung in der Septuaginta und die
lateinische Iuditherzählung in der Vulgata. Beide Textfassungen unter-
scheiden sich deutlich voneinander: Die Vulgata-Fassung ist gegenüber
der LXX-Fassung um 20 Prozent kürzer; weil die Vulgata-Fassung aber
auch umfangreiche Erweiterungen aufweist, teilen LXX und Vulgata nur
ungefähr die Hälfte des Textbestandes.
Lange ist man davon ausgegangen, dass hinter der Septuaginta- und der
Vulgata-Fassung der Juditerzählung ein hebräischer „Ur“-Text gestanden
habe. Diese Annahme, die die Forschung seit dem 19. Jh. prägt, gründet
sich im Wesentlichen auf zwei Argumente:
Erstens sei der griechische Text der Septuaginta so durchgehend von
einem hebräischen Sprachduktus geprägt, dass man hier von einem
hebräischen Originaltext, der recht wörtlich ins Griechische übertragen
worden sei, ausgehen müsse.
Zweitens spreche Hieronymus (ca. 347–420) in seiner praefatio zu sei-
ner lateinischen Übertragung von einem ihm vorliegenden „chaldäischen“
Text, der ihm als Vorlage gedient habe.
Beide Argumente werden im Folgenden diskutiert: Im ersten Teil wird
auf die in den letzten Jahren neu aufgerollte Frage nach der Originalspra-
che der LXX-Fassung der Juditerzählung eingegangen; der zweite Teil ist
der Analyse der praefatio des Hieronymus gewidmet.
1 H. Engel, „Das Buch Judit,“ in Einleitung in das Alte Testament (Hg. E. Zenger; Stuttgart /
Berlin / Köln: Kohlhammer, 72008), 289–301, 298.
360 barbara schmitz
1. Die Frage der Originalsprache der Septuaginta
In der Forschung zum Buch Judit galt es lange als unhinterfragter Kon-
sens, dass der Judit-Text der LXX eine Übersetzung aus dem Hebräischen
oder Aramäischen sei. Die Argumente für diese Hypothese sind u.a. von
Louis Soubigou2 in seinem Kommentar von 1952, von Yehoshua Grintz3 in
der von ihm hypothetisch rekonstruierten hebräischen Fassung der Judi-
terzählung von 1957 und von Erich Zenger4 in seinem 1981 erschienenen
Kommentar zur Juditerzählung in der Reihe Jüdische Schriften aus helle-
nistisch-römischer Zeit zusammengestellt worden. Eine weitere umfang-
reiche Auflistung findet sich im Judit-Kommentar der Anchor Bible von
Carey Moore5 1985 im Anschluss an Robert Pfeiffer6 1949. Außerordentlich
häufige Parataxe, hebraisierende Syntax, „Übersetzungsfehler“ und Hebra-
ismen in der Wortwahl sind in diesen Listen zahlreich aufgeführt. Dies
hat zu der weithin vertretenen Überzeugung geführt, dass die griechische
Juditerzählung die Übersetzung eines verloren gegangenen hebräischen
Originals sei. Diese Überzeugung der Juditforscher fasst Robert Hanhart so
zusammen: „Der griechische Text des Buches Iudith ist ein Übersetzungs-
text. Seine Vorlage war entweder hebräisch oder aramäisch“7—Robert
Hanhart muss allerdings selbst zugeben, dass es für diese angenommene
Vorlage keine handschriftlichen Zeugen gibt: „Aber solche Vorlagen schei-
nen schon früh sowohl unter Juden als auch unter Christen nur noch spo-
radisch bekannt gewesen zu sein“8.
Diese scheinbar geklärte Frage wurde 1974 durch Hans Yohanan Prie-
batsch9, der typisch griechisch-hellenistische Züge der Juditerzählung
hervorhob, neu gestellt. Eine Vielzahl von Beobachtungen, wie etwa die
Thyrsos-Stäbe in Judits Händen (Jdt 15:12), Worte wie ἀκινάκης für den
Säbel (Jdt 13:6) sowie die merkwürdigen geographischen Angaben ließen
2 L. Soubigou, „Judith,“ in La Sainte Bible de L. Pirot et A. Clamer 4 (Paris: Letouzey &
Ane, 1952), 481–575, 483–485.
3 Y. M. Grintz, Sefer Jehudît. A Reconstruction of the Original Hebrew Text with Intro-
duction, Commentary, Appendices and Indices (hebr.) (Jerusalem: Bialik Institute, 1957),
56–61.
4 E. Zenger, Das Buch Judit (JSHRZ I/6; Gütersloh: Gerd Mohn, 1981), 430–431.
5 C. A. Moore, Judith (AB 40; Garden City: Doubleday, 1985), 66–67.
6 R. H. Pfeiffer, „The Book of Judith,“ in History of New Testament Times (New York:
Harpers, 1949), 285–303, 298–299.
7 R. Hanhart, Text und Textgeschichte des Buches Judith (MSU 14; Göttingen: Vanden-
hoeck & Ruprecht, 1979), 9.
8 Ibid.
9 H. Y. Priebatsch, „Das Buch Judit und seine hellenistischen Quellen,“ ZDPV 90 (1974):
50–60.
ιουδιθ und iudith 361
Priebatsch zu dem vorsichtigen Schluss kommen, diese erschwerten „die

Annahme hebräischer Ursprache und vorhellenistischer Entstehung“10.
Während Priebatsch den Blick auf Besonderheiten der Juditerzählung
gerichtet hatte, die sich s.E. leichter aus einem griechisch-hellenistischen
Umfeld erklären lassen, hat Helmut Engel—wenn ich recht sehe—als
Erster11 in seinem 1992 erschienenen Artikel: „ ‚Der HERR ist ein Gott, der
Kriege zerschlägt‘. Zur Frage der griechischen Originalsprache und der
Struktur des Buches Judith“ die Frage nach der Originalsprache der Judi-
terzählung neu aufgerollt.12 Dabei hat er die Sprachgestalt der vorliegen-
den griechischen Erzählung in den Blick genommen und aufgezeigt, dass
die in der griechischen Juditerzählung zitierten oder eingespielten bibli-
schen Texte aus der LXX genommen wurden und nicht den hebräischen
Text wiedergeben. Engel konnte u.a. an Num 23:19 (vgl. Jdt 8:16) und Exod
15:3 (vgl. Jdt 9:7; 16:2) überzeugend aufzeigen, dass die LXX-Fassungen und
eben nicht die sich deutlich unterscheidenden hebräischen Textfassungen
zugrunde liegen. Weitere Argumente sind die Formulierung in Jdt 7:28,
die im Hebräischen unmöglich ist, der Bezug von Jdt 9:2 auf Gen 34:7 nach
der LXX, das die Rede Achiors in Jdt 5 prägende Motiv des „Wohnens“
anhand des Stammes -οικ- und einige andere mehr13. Damit kündigt Engel
den bis dahin verbreiteten Konsens auf: Statt eine Übersetzung eines
verloren gegangenen hebräischen Originals anzunehmen, geht er davon
aus, dass der uns vorliegende griechische Text des Juditbuches eine ori-
ginal auf Griechisch entstandene Erzählung sei, deren Verfasser die LXX
benutzt und sich eines hebraisierenden Sprachstils bedient habe.
Jan Joosten hat 2007 in der Festschrift Dimant einen eigenen Nachweis
für Griechisch als Originalsprache geführt: An einer Reihe von Beispielen
hat er aufgezeigt, dass sich die griechische Erzählung durch gehobenes
Griechisch in Wortwahl und Syntax auszeichne14: So gebe es z.B. nicht nur
für griechische Futur-Infinitiv-Formen und für Lexeme wie ἀπεγνωσμένοι
10 Ibid., 52.
11 Ein Vorläufer dürfte Franz Steinmetzer gewesen sein, der davon ausgeht, dass ein
großer Teil der Juditerzählung auf Griechisch entstanden sei und zwar jener vierte Teil,
den er in seinem literarkritischen Modell der makkabäischen Zeit zuordnet: F. Steinmet-
zer, Neue Untersuchungen über die Geschichtlichkeit der Judith-Erzählung (Leipzig: Haupt,
1907), 110.
12 H. Engel, „ ‚Der HERR ist ein Gott, der Kriege zerschlägt‛. Zur Frage der griechischen
Originalsprache und der Struktur des Buches Judith,“ in Goldene Äpfel in silbernen Schalen
(BEATAJ 20; Hg. K.-D. Schunk und M. Augustin; Frankfurt: Peter Lang, 1992), 155–168.
13 Zu den Argumenten vgl. Engel, Helmut, „ ‚Der HERR ist ein Gott‛,“ 155–168, 157–159.
14 J. Joosten, „The Original Language and Historical Milieu of the Book of Judith,“ in
Meghillot. Studies in the Dead Sea Scrolls (FS D. Dimant; Hg. M. Bar-Asher und E. Tov;
Jerusalem / Haifa: Bialik Institute / Haifa University Press, 2007), *159–176, *160–167.
362 barbara schmitz
und ἀπηλπισμένοι in Jdt 9:11 oder ἀνυπέρβλητος in Jdt 16:13, kein „simple
Hebrew equivalent“, sondern die Juditerzählung zeichne sich darüber
hinaus durch guten, z.T. gehobenen, spezifisch griechischen Satzbau aus
(so etwa in Jdt 5:3 bzw. 5:8, 12, 23; 7:30; 9:2 etc.), so dass man davon ausge-
hen müsse, dass die Juditerzählung auf Griechisch verfasst worden sei.
Die hebräisch klingenden Spracheigentümlichkeiten des griechischen
Judittexts sind sodann von Jeremy Corley15 2008 untersucht worden:
Jeremy Corley hat die Liste von Carey Moore, der über 30 Hinweise für
einen hebräischen Text auflistet, erneut einzeln durchgesehen und ist zu
dem Ergebnis gekommen, „that the proposed instances of Hebraic phra-
seology and style in the Greek text of Judith do not necessarily indicate
a Hebrew origin of the book, since they can easily be evidence of either
mimetic appreciation of Septuagintal style or Semitic interference. [. . .] a
Hebrew Vorlage cannot be presumed, while a Greek origin can be sugge-
sted as very possible“16.
Neben diesen sprachlichen Beobachtungen möchte ich über die von
Hans Yohanan Priebatsch notierten Facetten hinaus weitere Beobach-
tungen benennen, anhand derer ich das hellenistische, vielleicht sogar
römisch-hellenistische Kolorit der Erzählung aufgezeigt habe: Dies sind
erstens die zahlreichen Bezüge der Juditerzählung zu den Historien des
Herodot17, zweitens die nach den Regeln griechischer Rhetorik gestaltete
Rede Achiors in Jdt 5:5–2118, drittens die Bezüge, die sich über das Motiv
des κωνώπιον (Jdt 10:21; 13:9, 15; 16:19) in die lateinische Literatur ergeben19
sowie viertens die offensichtlichen Aufnahmen der Reflexionen über den
Tyrannenmord im Anschluss an die literarisch viel gepriesene Tat von
15 J. Corley, „Septuagintalisms, Semitic interference, and the original language of the
Book of Judith,“ in Studies in the Greek Bible. Essays in honor of Francis T. Gignac (CBQMS
44; Hg. J. Corley; Washington: The Catholic Biblical Association of America, 2008), 65–96.
16 Ibid., 96.
17 B. Schmitz, „Zwischen Achikar und Demaratos—Die Bedeutung Achiors in der Judit-
erzählung,“ BZ 48 (2004): 19–38, 28–37; vgl. auch A. Momigliano, „Biblical Studies and Clas-
sical Studies: Simple Reflections about Historical Method,“ BA 45 (1982): 224–228; M. S.
Caponigro, „Judith, Holding the Tale of Herodotus,“ in „No one spoke ill of her“. Essays on
Judith (SBL Early Judaism and its Literature 2; Hg. J. VanderKam; Atlanta: Scholars Press,
1992), 31–46 = idem, in Women in the Hebrew Bible: A Reader (Hg. A. Bach; New York /
London: Routledge, 1999), 377–386.
18 B. Schmitz, Gedeutete Geschichte. Die Funktion der Reden und Gebete im Buch Judit
(HBS 40; Freiburg: Herder, 2004), 56–70.
19 B. Schmitz, „Holofernes’s Canopy in the Septuagint,“ in The sword of Judith. Judith
Studies across the Disciplines (Hg. K. R. Brine, E. Ciletti und H. Lähnemann; Cambridge:
Open Book Publishers, 2010), 71–80.
Harmodius und Aristogeiton20. Diese Beobachtungen ließen sich wahr-

scheinlich noch erweitern.
Angesichts der jüngsten Diskussionen würde man heute auf die alte
Frage „Welches ist die Originalsprache der Juditerzählung?“ wohl eine
neue Antwort geben:
Erstens: Die in der LXX vorliegende Juditerzählung scheint auf Grie-
chisch entstanden zu sein; sie weist gehobenen griechischen Sprachstil
auf, integriert aber in diesen eine das Hebräische imitierende griechische
Sprache. Dies wirft neue Fragen auf: Geschieht dies bewusst? Wird der
hebraisierende Sprachstil vielleicht sogar als Stilmittel eingesetzt? Sollte
damit der so hellenistisch geprägten Juditerzählung ein hebräisches, sozu-
sagen „biblisches“ Kolorit verliehen werden? Denkbar wäre es. Die These,
dass die vorliegende Fassung der Erzählung auf Griechisch entstanden
sei, beantwortet allerdings nicht die Frage, ob es hebräische oder aramä-
ische Vorformen des griechischen Juditbuchs gegeben habe, sondern lässt
sie offen.
Zweitens weist die Juditerzählung durchgehend deutlich griechisch-
hellenistisches Kolorit auf, so dass davon auszugehen ist, dass die grie-
chische Juditerzählung in den theologischen, politischen und kulturellen
Vorstellungen des Hellenismus beheimatet ist, wie sie sich in der LXX und
in profan-hellenistischen Kontexten zeigen.
Damit hat sich das Koordinatensystem, in das die Juditerzählung der
LXX und damit die eine der beiden großen Textformen, einzuordnen ist,
für die Auslegung grundlegend verändert.
2. Die praefatio des Hieronymus zum Buch Iudith
Hieronymus spricht in seiner praefatio davon, dass ihm ein Text in „chaldä-
ischer Sprache“ (Chaldaeo . . . sermone; Z. 2–3) vorgelegen habe21.
20 B. Schmitz, „War, violence and Tyrannicide in the Book of Judith,“ in Deuterocanoni-
cal and Cognate Literature. Yearbook 2010 (Hg. J. Liesen und P. Beentjes; Berlin / New York:
De Gruyter, 2010), 103–119.
21 Zitiert nach: R. Weber & R. Gryson, Hg., Biblia Sacra. Iuxta Vulgatam Versionem. Edi-
tio Quinta (Stuttgart: Deutsche Bibelgesellschaft, 20075), im Folgenden „Weber/Gryson‟.
Aus dieser Ausgabe stammt auch die Nummerierung der Zeilen.
Die deutsche Übersetzung des Iudith-Buchs und des Prologs stammen von Helmut
Engel.
364 barbara schmitz
1Apud Hebraeos liber Iudith inter Bei den Juden wird das Buch
Agiografa legitur: cuius auctoritas ad Judit den „Agiografa“ zugeordnet,
2roboranda illa quae in contentione deren Autorität zur Stützung eines
veniunt, minus idonea iudicatur. Arguments in Streitfragen als
Chaldaeo 3tamen sermone conscriptus, weniger geeignet betrachtet wird. Es
inter historias conputatur. Sed quia ist jedoch in chaldäischer Sprache
hunc 4librum sinodus nicena in verfasst und wird den geschichtli
numero Sanctarum Scripturarum chen Darstellungen zugerechnet.
legitur conputasse, 5adquievi Aber da, wie es heißt, das Konzil
postulationi vestrae, immo exactioni: von Nikaia dieses Buch der Zahl der
et sepositis occupationibus, quibus heiligen Schriften zugerechnet hat,
6vehementer artabar, huic unam habe ich mich mit Eurem Anliegen,
lucubratiunculam dedi, magis sensum ja Eurer Forderung, einverstanden
e sensu quam ex 7verbo verbum erklärt, mich heftig bedrängende
transferens. Multorum codicum Beschäftigungen beiseite gelegt und
varietatem vitiosissimam amputavi; diesem [Buch] eine einzige kleine
8sola ea, quae intelligentia integra in Nachtschicht gewidmet, wobei ich
verbis chaldaeis invenire potui, latinis es mehr sinngemäß als wortwörtlich
expressi. übertragen habe. Die sehr fehlervolle
Verschiedenheit der vielen Hand
schriften habe ich weggeschnitten:
Nur das, was ich in chaldäischen
Worten voll verständlich finden
konnte, habe ich lateinisch ausge
drückt.
9Accipite Iudith viduam, castitatis Nehmt die Witwe Judit in Empfang,
exemplum, et triumphali laude ein Vorbild der Keuschheit, und
perpetuis eam 10praeconiis declarate. verkündigt sie mit triumphierendem
Hanc enim non solum feminis, Lob in unablässigen Preisungen!
sed et viris imitabilem dedit, 11qui, Diese hat nämlich nicht nur den
castitatis eius remunerator, virtutem Frauen, sondern auch den Männern
ei talem tribuit, ut invictum omnibus derjenige zur Nachahmung gegeben,
12hominibus vinceret, insuperabilem der ihr als Belohner ihrer Keuschheit
superaret. solche Stärke zuteilte, dass sie den
von allen Menschen Unbesiegten
besiegte, den Unüberwindlichen
überwand.
Hieronymus dürfte das Buch Iudith22 am Ende der dritten Phase seiner
Bibelübersetzungen, zwischen 405 und 407, angefertigt haben23.
Die praefatio24 ist in zwei Teile gegliedert: Im ersten Teil (Z. 1–8) ordnet
Hieronymus das Buch Judit in den jüdischen bzw. christlichen Kanon ein
und erläutert die Textsituation und seine Arbeitsweise. Der zweite Teil
(Z. 9–12) setzt mit zwei Aufforderungen im Imperativ Plural neu ein und
nimmt zum Inhalt des Buches Iudith Stellung, wobei er nur die Figur der
Iudith in den Blick nimmt.
Die Analyse der praefatio wird im Folgenden in vier Themengebiete
gegliedert: Zuerst wird die Frage nach den Adressaten der praefatio
gestellt (2.1.), zweitens werden die Informationen zum Text in den Blick
genommen (2.2.), drittens werden die Informationen zur Arbeitsweise des
Hieronymus (2.3.) und schließlich die inhaltlichen Akzente der praefatio
analysiert (2.4.).
2.1. Die Frage nach den Adressaten der Vorrede zum Buch Iudith
Hieronymus hat viele Bücher in seiner Bibelübersetzung mit einer Vor-
rede ausgestattet25. Die Vorreden (praefatio oder prologus) zu den meisten
biblischen Büchern sind an Persönlichkeiten aus Hieronymus’ persönli-
chem Umfeld gerichtet26. Diese sind jedoch nicht mit privaten Widmun-
gen zu verwechseln, wie sie heute Büchern vorangestellt werden, vielmehr
handelt es sich bei den Angeredeten um Förderer und Geldgeber, die die
Projekte des Hieronymus finanziert und so überhaupt ermöglicht haben.
Der Begriff der „amicitiae“ bezeichnet weniger private Freundschaften,
sondern vielmehr ein Netz persönlicher Beziehungen zu wirtschaftlich
22 Zur besseren Orientierung bezieht sich die Schreibung Iudith wie auch die Abkür-
zung „Idt“ immer auf den Vulgata-Text; wo der Text der Vetus Latina gemeint ist, wird dies
durch die Hinzufügung „[VL]“ kenntlich gemacht.
23 Zur Biographie von Hieronymus vgl. J. N. D. Kelly, Jerome. His life, writings, and con-
troversies (London: Duckworth, 1975); G. Grützmacher, Hieronymus. Eine biographische
Studie zur alten Kirchengeschichte, 3 Bde. (Leipzig, 1901; Berlin 1906/1908 [Nachdr.: Aalen:
Scientia-Verlag, 1969/1986]); F. Cavallera, St. Jérôme. Sa vie et son oeuvre, 2 Bde. (Louvain /
Paris: Champion, 1922); vgl. auch A. Kotzé, „Augustine, Jerome and the Septuagint,“ in
Septuagint and Reception (VTSup 127; Hg. J. Cook; Leiden: Brill, 2009), 245–260.
24 Die Übersetzung der praefatio stammt von Helmut Engel.
25 Eine Übersicht findet sich bei A. Fürst, Hieronymus. Askese und Wissenschaft in der
Spätantike (Freiburg / Basel / Wien: Herder, 2003), 86–87.
26 Keine namentlich genannten Adressaten haben die Vorreden zu den Büchern Ijob
(Übersetzung aus dem Hebräischen), Josua, Jeremia, Ezechiel und Iudith.
366 barbara schmitz
potenten Freunden27. Hieronymus darf als ein äußerst erfolgreicher „net-

worker“ gelten, dem es Zeit seines Lebens gelang, Freundschaften zu rei-
chen Christinnen und Christen aufzubauen, zu pflegen und über diese
Netzwerke Gelder für seine Projekte zu akquirieren28. Die in den Vor-
reden Genannten sind in diesem Sinne „Freunde“ des Hieronymus, die
seine wissenschaftliche Arbeit direkt oder indirekt finanziert haben. So
sind die Vorreden zu den Samuel- und Königsbüchern (prologus galeatus)
sowie die Vorreden zu Jesaja, Daniel und zum Zwölfprophetenbuch an die
beiden Frauen Paula und ihre Tochter Eustochium gerichtet. Paula (die
Ältere), die aus einer reichen römischen Adelsfamilie stammte, finanzierte
beispielsweise aus ihrem Vermögen den Bau von drei Frauen- und einem
Männerkloster mit Pilgerherberge in Bethlehem und ermöglichte Hie-
ronymus seine wissenschaftliche Tätigkeit. Dass es sich bei den Bibelüber-
setzungen auch um Auftragsarbeiten handelt, ist aus der ersten Vorrede,
die an Papst Damasus gerichtet ist, zu erkennen. Dieser hatte Hieronymus
im Jahre 382 beauftragt, das Neue Testament zu revidieren. Es handelt
sich um ein Auftragswerk, dem die Vorrede als eine Art „Begleitbrief zur
bestellten Ware“29 beigefügt ist.
Die praefatio der Juditerzählung zeichnet sich durch zwei Besonder-
heiten aus: zum einen wird kein Adressat in der praefatio genannt, zum
anderen wird im Prolog auf den Inhalt des Buches Iudith eingegangen.
Anders als bei den meisten Vorreden, die sich an namentlich genannte
Adressaten wenden, wendet sich in der praefatio zum Buch Iudith ein
„Ich“ an Leser in der zweiten Person Plural (postulationi vestrae, Z. 5; acci-
pite, Z. 9; declarate, Z. 10). Wer könnte diese zweite Person Plural sein?
27 So bestimmt dies auch die Schrift „Über Freunde“ (Περὶ φίλων) des Rhetors Libanus
(314–393), die in der handschriftlichen Überlieferung den sprechenden zweiten Titel „Über
Armut“ (Περὶ πενίας) trägt, vgl. hierzu und zu den Beziehungen des Hieronymus Fürst,
Hieronymus, 72–76.
28 Vgl. hierzu S. Rebenich, Hieronymus und sein Kreis. Prosopographische und sozialge-
schichtliche Untersuchungen (Hist. Einzelschriften 72; Stuttgart: Steiner, 1992).
29 “Damasus hat Hieronymus ausgesucht, nicht umgekehrt. Hieronymus bekommt den
Auftrag, die Evangelien zu revidieren. So schreibt er dann: ‚Novum opus facere me cogis
ex veteri . . .‛ (F. Stummer, Einführung in die lateinische Bibel. Ein Handbuch für Vorlesun-
gen und Selbstunterricht [Paderborn: Schöningh, 1928], 255f.; PL 29,557ff.) Es wird an dem
klaren Fall dieser Vorrede zu den Evangelien etwas exemplarisch deutlich, was—wie mir
scheint—auch für alle anderen Vorreden des Hieronymus gilt: Hieronymus stellt sich
hier als einer vor, dessen Arbeit an der Übersetzung, wenn nicht tatsächlich—wie es am
Anfang nachweislich der Fall ist –, so doch wenigstens seiner Meinung nach auf einen
Auftrag hin geschieht. Die Vorrede bringt den Antwortcharakter der geleisteten Arbeit
zum Ausdruck: wenn sie schon ein Brief ist, so ein Begleitbrief mit der bestellten Ware,
nicht primär ein Werbebrief!“ M. E. Schild, Abendländische Bibelvorreden bis zur Lutherbi-
bel (QFRG 39; Gütersloh: Gerd Mohn, 1970), 13–70, 17, vgl. weitere Belege 17–18.
Zunächst: Wir wissen es nicht. Aber es sprechen gute Gründe dafür,

dass die praefatio zum Buch Iudith an die gleichen Adressaten wie die
zum Buch Tobit gerichtet ist. Diese ist an Chromatius und Heliodorus
gerichtet (Cromatio et Heliodoro episcopis Hieronymus in Domino salutem;
Tobit-Prolog Z. 1; Weber/Gryson 676). Für die enge Beziehung zwischen
den Vorreden zu den Büchern Tobit und Iudith spricht erstens die zeitli-
che Nähe; beide Bücher dürften zwischen 405 und 407 übertragen worden
sein30. Zweitens werden beide Bücher zu den „Hagiografen“ gezählt (inter
Agiografa legitur Z. 1; vgl. Tobit-Prolog Z. 5; Weber/Gryson 676). Drittens
sind beide Bücher aus „chaldäischer Sprache“ übertragen worden ([liber]
Chaldaeo . . . sermone conscriptus Z. 2–3; vgl. librum Chaldaeo sermone con-
scriptum Tobit-Prolog Z. 3 und Chaldaeorum lingua Tobit-Prolog Z. 8–9;
Weber/Gryson 676). Viertens findet sich in beiden Vorreden durchgehend
die Anrede in der zweiten Person Plural (postulationi vestrae, Z. 5; accipite,
Z. 9; declarate, Z. 10; in der Tobit-praefatio passim), die in der Vorrede zum
Buch Iudith nicht aufgelöst wird, sondern offen bleibt, sich in der praefa-
tio zum Buch Tobit aber an die im Anfangsgruß genannten Bischöfe Chro-
matius und Heliodorus richtet. Fünftens geht aus beiden Vorreden hervor,
dass Hieronymus nur ungern die beiden Übersetzungen angefertigt habe.
So betont er, dass er für die Übertragung der Iudith andere Arbeiten wider-
willig beiseite gelegt habe31. Sechstens erläutert Hieronymus in beiden
Vorreden, dass er die Übertragungen durch äußeren Druck erledigt habe:
In der praefatio zum Buch Iudith erläutert er, dass er der Übertragung nur
auf dringenden Wunsch der Angesprochenen, wohl der Bischöfe Chroma-
tius und Heliodorus, nachkomme32 und sich damit einer Entscheidung des
Konzils von Nizäa beuge (sed quia hunc librum sinodus nicena in numero
Sanctarum Scripturarum legitur conputasse, adquievi postulationi vestrae,
Z. 3–4). In der praefatio zum Buch Tobit schreibt Hieronymus, dass er
sich über das Insistieren der beiden Bischöfe wundere, die von ihm die
Übertragung des Buches Tobit ins Lateinische forderten33.
30 Chromatius oder Heliodor starb 407. Daher ist die Abfassungszeit des Tobit-Prologs
vor 407 anzusetzen. Geht man von einer Zusammengehörigkeit der beiden Prologe aus,
liegt mit 407 auch ein terminus ante quem für die Übertragung von Iudith vor.
31 „mich heftig bedrängende Beschäftigungen beiseite gelegt“ (et sepositis occupationi-
bus, quibus 6vehementer artabar [Iudith-Prolog Z. 5–6]).
32 „. . . habe ich Eurem Anliegen, ja Eurer Forderung, nachgegeben“ (5adquievi postula-
tioni vestrae, immo exactioni [Iudith-Prolog Z. 5]).
33 „Ich höre nicht auf, mich über die Inständigkeit Eurer Forderung zu wundern. Ihr
fordert ja von mir, dass ich ein in chaldäischer (= aramäischer) Sprache verfasstes Buch
ins Lateinische übertrage“ (Mirari non desino exactionis vestrae instantiam; exigitis enim ut
librum Chaldaeo sermone conscriptum, ad Latinum stilum traham [Tobit-Prolog Z. 2–3]).
368 barbara schmitz
Die genannten Argumente sprechen dafür, den Iudith-Prolog wie den

Tobit-Prolog als an Chromatius und Heliodorus gerichtet zu sehen34.
Damit würden sich die offen bleibenden Formulierungen in der praefatio
zum Buch Iudith in einfacher Weise klären.
Wer waren Chromatius und Heliodorus? Chromatius war Mitglied der
asketischen Gemeinschaft in Aquileja35 und wurde vor 393, möglicher-
weise um 388, Bischof von Aquileja (bis 407/408). Die Beziehungen zwi-
schen Hieronymus und Chromatius gestalteten sich—v.a. während des
Origenismusstreits—wechselhaft. Heliodorus und Hieronymus kannten
sich aus Studienzeiten und waren seitdem gut befreundet. Nachdem
Heliodorus den Dienst in der kaiserlichen Verwaltung quittiert hatte,
nahm er die asketische Lebensweise auf und unternahm eine Pilgerreise
nach Jerusalem, während der er zwischen 373 und 375 Hieronymus traf.
Heliodorus wurde Bischof in seiner Heimatstadt Altinum in Oberitalien.
Beide stammten aus reichen Familien und gehörten zu jenen finanzstarken
Freunden, die Hieronymus tatkräftig unterstützt haben36. An Heliodorus
und Chromatius zusammen waren die Prologe zu den „Salomo–Büchern“
(Sprichwörter, Kohelet und Hoheslied) und eben auch die praefatio zum
Buch Tobit gerichtet; an Chromatius allein sind darüber hinaus die Kom
mentare zu Habakuk und Jona sowie die Übersetzung der Chronikbücher
gerichtet.
2.2. Informationen zum übersetzten Text in der praefatio

Hieronymus gibt in seiner praefatio im Wesentlichen drei Informationen
darüber, welche Textsituation er vorgefunden habe:
Erstens sei ein liber Iudith bei den Juden bekannt, gelte aber bei ihnen
nicht als biblischer Text, sondern zähle unter die „agiografa“ (Apud
Hebraeos liber Iudith inter agiographa legitur; Z. 1).
34 So auch Fürst, Hieronymus, 87 Anm. 139; ebenso P.-M. Bogaert, Judith. Einleitung
(Vetus Latina 7/2, Faszikel 1; Freiburg: Herder, 2001), 31.
35 Die Informationen zu Chromatius und Heliodor sind entnommen aus Fürst, Hie-
ronymus, 164–165; 182–183. Zu Chromatius siehe B. Studer, „Chromatius d’Aquilée,“ in
Dictionnaire encyclopédique du christianisme, Bd. 1 (Paris: Cerf, 1990), 476; zu Heliodor:
R. Aubert, „Héliodore,“ in Dictionnaire d’Histoire et de Géographie Ecclésiastique, Bd. 23
(Paris: Letouzey & Ane, 1990), 908.
36 Dazu gehört möglicherweise auch die Finanzierung des Sprachkundigen im „Chaldä-
ischen“ und Hebräischen (utriusque linguae peritissimum loquacem repperiens) und die des
„Schnellschreibers“ (notario Z. 11), von denen im Tobit-Prolog die Rede ist.
Über dieses Buch erfahren die Leser zweitens, dass es in „chaldäischer

Sprache“ (Chaldaeo tamen sermone conscriptus; Z. 2–3; vgl. 8) verfasst sei.
Hieronymus ist der einzige, der von einem Buch Judit in „chaldäischer
Sprache“ berichtet. Anders hört sich eine Aussage des Origenes (ca. 185–
254 n. Chr.) an, der schreibt, dass die Juden Iudith und Tobit „nicht im
Gebrauch haben und sie auch nicht unter den Apokryphen in hebräischer
Sprache besitzen“37. Was versteht Hieronymus unter „Chaldäisch“? Dass
Hebräisch nicht gemeint sein kann, ergibt sich daraus, dass Hieronymus
im Prolog zum Buch Tobit schreibt, dass das Hebräische und das Chaldä-
ische nahe verwandte Sprachen seien (quia vicina est Chaldaeorum lingua
sermoni Hebraico; Tobit-Prolog Z. 8–9). Die Schwierigkeit in der Identifi-
kation des „Chaldäischen“ besteht darin, dass Hieronymus damit sowohl
Syrisch als auch Aramäisch bzw. die verschiedenen Ausprägungen des
Aramäischen bezeichnet hat38. Ein wenig Syrisch hatte Hieronymus wohl
während seines ersten Aufenthalts im Osten (373–382) in der „Wüste“
nahe der Stadt Chalkis gelernt (vgl. Ep. 17,2). Ebenso konnte Hieronymus
auch etwas Aramäisch lesen, wie aus dem Prolog zum Buch Daniel hervor-
geht: „Neulich habe ich mich an das Buch Daniel gemacht und . . . (nach
früheren Versuchen) noch einmal begonnen, Chaldäisch zu lernen. Doch
um die Wahrheit zu sagen: Bis heute kann ich die chaldäische Sprache
eher lesen und verstehen als sprechen“ (Weber/Gryson 1341). Dass seine
Kenntnisse im „Chaldäischen“ gering geblieben sind, zeigt auch der Prolog
zum Buch Tobit (vgl. hierzu 2.3.)39.
Drittens berichtet Hieronymus, dass ihm viele Handschriften in „sehr
fehlerhafter Verschiedenheit“ vorgelegen hätten (multorum codicum
varietatem vitiosissimam amputavi; Z. 7). Über die Sprache dieser unter-
schiedlichen Handschriften erfahren wir nichts. Pierre-Maurice Bogaert
schließt sich den Analysen von Philipp Thielmann an, der in seiner Studie
„Beiträge zur Textkritik der Vulgata“ von 1883 davon ausgeht, „1) dass
37 ὅτι Ἑβραῖοι τῷ Τωβίᾳ οὐ χρῶνται, οὐδὲ τῇ Ιουδήθ· οὐδὲ γὰρ ἔχουσιν αὐτὰ ἐν ἀποκρύφοις
ἑβραϊστί, Brief des Origenes an Julius Africanus (um 240 n. Chr.). Auf Judit wird zwar recht
häufig in der christlichen Literatur verwiesen; die ältesten Belege sind Tertullian, De mono-
gamia (TE mon 17,1 [204]) und Klemens von Rom (CLE-R 55 [50,15–20]; 59 [55,2–3]); aber
es wird auffallend selten aus dem Buch zitiert, so dass die Zitate und Verweise bei den
Kirchenvätern kaum Rückschlüsse auf die Textsituation erlauben.
38 Belege hierzu finden sich bei Fürst, Hieronymus, 77.
39 Beide Sprachen werden in der Forschung diskutiert; die meisten gehen von einem
aramäischen Text aus. Da aber das Syrische eine ähnliche Nähe zum Hebräischen auf-
weist wie das Aramäische—und dies ist der einzige Hinweis, den Hieronymus gibt,—hält
Bogaert beide Sprachen für möglich, vgl. Bogaert, Judith, 31–32.
370 barbara schmitz
Hieronymus bei seiner Bearbeitung des Buches Judith in ausgedehn-

tester Weise die altlateinischen Texte verwendet hat, 2) dass er hierbei
insbesondere dem cod. Corb. öfter als den übrigen Handschriften gefolgt
ist“40. Daher resümiert Bogaert: „Il n’est pas dit que ces derniers sont
grecs; et quand Jérôme s’en prend à la vitiosissima varietas, ce n’est pas
aux grecs qu’il pense; il doit s’agir ici de témoins latins [. . .] L’examen de
la version de Jérôme, mené avec la méthode par Philip Thielmann il y a
plus de cent ans, montre que le vocabulaire n’en est guère hiéronymien,
mais franchement vieux-latin“41.
Bemerkenswert ist, dass Hieronymus ein Zeuge für die Pluralität und
Unterschiedlichkeit der im Umlauf befindlichen Judit-Handschriften, aber
zugleich er auch der einzige Zeuge für ein „chaldäisches“ Buch Iudith ist.
2.3. Informationen zur Arbeitsweise des Hieronymus

Der ausführliche Hinweis des Hieronymus auf eine „chaldäische“ Textfas-
sung ist nicht nur textgeschichtlich interessant, sondern auch mit Blick
auf seine Arbeitsweise. Aus dem Tobit-Prolog lernt man, dass Hieronymus
selbst so wenig „Chaldäisch“ konnte, dass er sich diesen Text übersetzen
lassen musste. Dazu beschäftigte er zwei Mitarbeiter: einen Dolmetscher,
der die „chaldäischen“ Worte mündlich ins Hebräische übersetzte, und
einen „Schnellschreiber“, dem Hieronymus dann die mündlich ins Hebrä-
ische übersetzten Worte auf Lateinisch diktierte42.
40 P. Thielmann, „Beiträge zur Textkritik der Vulgata, insbesondere des Buches Judith,“
Programm der Studienanstalt Speyer (1883): 37 vgl. 19–20. Dem schließt sich Bogaert, Judith,
63 an und lehnt die Analysen von E. E. Voigt, The Latin Version of Judith (Leipzig: Drugu-
lin, 1925) ab. Vgl. idem, 63.67: „La Vetus Latina de Judith est une traduction du grec [. . .]
Les minuscules 58 et 583, la Vetus Latina, la Peshitta et la Syro-hexaplaire . . . représentent
indiscutablement une même forme du livre de Judith.“
41 P.-M. Bogaert, „Judith dans la première Bible d’Alcala (Complutensis 1) et dans la
version hiéronymienne (Vulgata),“ in Philologia Sacra. Biblische und Patristische Studien für
Hermann J. Frede und Walter Thiele zu ihrem siebzigsten Geburtstag, Bd. 1: Altes und Neues
Testament (AGBL 24/1; Hg. R. Gryson; Freiburg: Herder, 1993), 116–130, 117; Bogaert, Judith,
32: „ils ne peuvent être que latins en l’abscence de mention du grec“.
42 „Und da die Sprache der Chaldäer dem Hebräischen nahe ist, habe ich einen in bei-
den Sprachen sehr bewanderten Sprecher eilig aufgegriffen und, was er mir in hebräischen
Worten ausdrückte, das habe ich unter Beiziehung eines Schnellschreibers in lateinischen
Wendungen dargestellt.“ (et quia vicina est Chaldaeorum lingua sermoni Hebraico, utriusque
linguae peritissimum loquacem repperiens . . . et quidquid ille mihi Hebraicis verbis expressit,
hoc ego accito notario, sermonibus Latinis exposui [Tobit-Prolog Z. 8–11]).
Wie muss man sich die Arbeit des Hieronymus am Buch Iudith
vorstellen?43 Welche Arbeitsschritte sind anzunehmen? Eine Rekonstruk-
tion des Arbeitsprozesses, wie dieser aus den Vorreden zu den Büchern
Iudith und Tobit zu erkennen ist, ergibt44:
Im ersten Arbeitsschritt ist der „chaldäische“ Judit-Text von einem Dol-
metscher mündlich ins Hebräische übersetzt worden, den Hieronymus
im zweiten Arbeitsschritt ad hoc ins Lateinische übertragen hat. Diese
Übertragung wurde dann im dritten Arbeitsschritt von einem Schreiber
schriftlich festgehalten. Diesen von ihm selbst erstellten lateinischen Text
hat Hieronymus im vierten Arbeitsschritt mit anderen ihm vorliegenden
Textfassungen, wahrscheinlich altlateinischen Handschriften, verglichen.
Den fünften Arbeitsschritt beschreibt Hieronymus folgendermaßen: „Die
sehr fehlervolle Verschiedenheit der vielen Handschriften habe ich wegge
schnitten: Nur das, was ich in chaldäischen Worten voll verständlich fin-
den konnte, habe ich lateinisch ausgedrückt“45. Hieronymus hat somit
eine eigene Auswahl aus dem ihm vorliegenden Textbestand getroffen. Für
diese waren zwei Leitkriterien entscheidend: Das erste Leitkriterium war,
dass Hieronymus dem „chaldäischen“ Text die Priorität gab, d.h. jenem
Text, der über das Hebräische ins Lateinische übertragen worden war. Das
zweite Leitkriterium war, dass von den „chaldäischen“ Worten nur die in
den Iudith-Text des Hieronymus aufgenommen wurden, denen Hierony-
mus „volle Verständlichkeit“ (intelligentia integra) zubilligte. Angesichts
dieser für zahlreiche und vielfältige Verschiebungen, Veränderungen
und auch Fehler offenen Arbeitsweise ist die Beschreibung, die Hierony-
mus selbst davon gibt, wohl zutreffend: Er habe „mehr sinngemäß als
wortwörtlich übertragen“46. Anders beschreibt Hieronymus seine Arbeits-
weise z.B. in der Vorrede zum Buch Ester, wo er erklärt, dass er „ganz
eng Wort für Wort“ übertragen habe (verbum e verbo pressius transtuli,
Z. 1–2; Weber/Gryson 691). Hieronymus betont somit in seiner Vorrede
zum Buch Iudith selbst, dass sich seine Arbeitsweise hier deutlich von der
bei anderen Büchern unterschieden habe. Ob dies mit den wenig geliebten
43 Vgl. für die Septuaginta: T. A. W. van der Louw, „The dictation of the Septuagint
Version,“ JSJ 39 (2008): 211–229.
44 So auch U. Köpf, „Hieronymus als Bibelübersetzer,“ in Eine Bibel—viele Übersetzun-
gen. Not und Notwendigkeit (Hg. S. Meurer; Stuttgart: Deutsche Bibelstiftung, 1978), 71–89
(75).
45 Multorum codicum varietatem vitiosissimam amputavi, 8sola ea, quae intelligentia
integra in verbis chaldaeis invenire potui, latinis expressi (Iudith-Prolog Z. 7–8).
46 . . . magis sensum e sensu quam ex 7verbo verbum transferens (Iudith-Prolog Z. 6–7).
372 barbara schmitz
„Hagiografen“ zusammenhängt?47 Oder vielleicht mit dem Inhalt der

Erzählung und der Figur der Iudith?
Hinzu kommt, dass Hieronymus behauptet, der Arbeit am Buch Iudith
nur „eine einzige kleine Nachtschicht“ (huic unam lucubratiunculam dedi;
Z. 6) gewidmet zu haben. Philipp Thielmann formuliert es vorsichtig so:
„Mag man von der Willenskraft und Gelehrsamkeit des Hieronymus, von
seiner Geschicklichkeit und Erfahrung im Übersetzen noch so hoch den-
ken, es lässt sich von vorneherein annehmen, dass bei solcher Eilfertigkeit
ein Werk zustande kommen musste, das den anderen mit grösster Sorg-
falt übertragenen Partien des alten Testaments in gar manchen Punkten
nachsteht“48. So scheint es sich bei der Aussage von der „einen kleinen
Nachtschicht“ eher um eine rhetorische, denn um eine wörtlich zu neh-
mende Aussage zu handeln, zumal im Prolog zum Buch Tobit Hierony-
mus betont, dass er nur einen einzigen Tag Arbeit darauf verwendet hätte
(unius diei laborem arripui). Einen einzigen Tag für Tobit und eine einzige
Nachtschicht für Iudith—diese Aussagen sind eine schöne rhetorische
Figur, vor allem, wenn man sich vor Augen hält, dass eine Übersetzung
der 14 Kapitel des Buches Tobit und der 16 Kapitel des Buches Iudith mit
der beschriebenen Arbeitsweise an einem Tag bzw. in einer Nacht nicht
realisierbar ist.
Wenn es sich hier also um eine ausschmückende, wenig wörtlich zu
nehmende Aussage handelt, dann stellt sich die Frage, inwieweit dies
auch für die anderen Aussagen der praefatio gelten könnte. Was könnte
dies beispielsweise für die Frage nach dem „chaldäischen“ Text bedeuten,
deren einziger Zeuge Hieronymus ist . . .?
47 So schreibt Hieronymus im Tobit-Prolog: „Ich habe Eurem Wunsch Genüge getan,
nicht jedoch meinem Studium“ ( feci satis desiderio vestro, non tamen meo studio). Dies
vermutet auch Stummer, Einführung in die lateinische Bibel, 97. Vgl. zu hebraica veritas:
C. Markschies, „Hieronymus und die ‚Hebraica Veritas‛. Ein Beitrag zur Archäologie des
protestantischen Schriftverständnisses?“ in Die Septuaginta zwischen Judentum und Chri-
stentum (WUNT 72; Hg. M. Hengel und A. M. Schwemer; Tübingen: Mohr Siebeck, 1994),
131–181.
48 Thielmann, „Beiträge zur Textkritik der Vulgata,“ 18. In Bezug auf die Qualität der
Übertragung urteilt Kaulen: „Am tiefsten stehen unter all seinen Uebertragungen die deu-
terokanonischen Bücher Judith und Tobias“. F. Kaulen, Geschichte der Vulgata (Mainz:
Kirchheim, 1868), 180.
2.4. Inhaltliche Akzentsetzungen der praefatio für die Lektüre des

Buches Iudith
Der zweite Teil der Vorrede (Z. 9–12) besteht aus einer Stellungnahme
des Hieronymus zum Inhalt der folgenden Erzählung und zur Figur der
Iudith. Dies ist die zweite Besonderheit, durch die sich die praefatio zum
Buch Iudith auszeichnet: Normalerweise nimmt Hieronymus in seinen
Vorreden selten Stellung zum Inhalt des übersetzten biblischen Buches—
anders jedoch bei Iudith. Damit fällt die praefatio zum Buch Iudith aus
dem Rahmen—Anlass genug, dies näher in den Blick zu nehmen.
Wen hat Hieronymus im Blick, wenn er den zweiten Teil seiner praefa-
tio mit einer zweifachen Verwendung des Imperativs Plural (Z. 9) begin-
nen lässt?
Folgt man der These der Zusammengehörigkeit der Prologe der Bücher
Tobit und Iudith, dann sind zunächst Heliodorus und Chromatius ange-
sprochen. Ihnen würde dann die Aufforderung „empfangt“ (accipite, Z. 9)
und „verkündet“ (declarate, Z. 10) gelten.
Zugleich dürfte sich Hieronymus, ganz in antiker Tradition der „Wid-
mung“, bewusst gewesen sein, dass seine Vorreden keine privaten Begleit-
schreiben und auch nicht nur Schreiben an seine finanziellen Gönner sind,
sondern sich über die Angesprochenen hinaus an die Öffentlichkeit der
Leserschaft wenden. Das wird beispielsweise im prologus galeatus deut-
lich: Dieser ist zunächst an die beiden Frauen Paula und deren Tochter
Eustochium gerichtet. Im Laufe des Prologs wird aber auch ein „Leser“ in
der 3. Person Singular maskulinum (obsecro te lector; Z. 58; Weber/Gryson
365) angesprochen, an den sich der ganze folgende Abschnitt richtet (Z.
58–73); etwas weiter im Prolog wechselt dann die Anrede wieder, und
Paula und Eustochium werden als „Dienerinnen Christi“ (sed et vos famu-
las Christi rogo . . .; Z. 74; Weber/Gryson 366) wieder in der 2. Person Plur.
femininum angeredet. Damit ist deutlich, dass Hieronymus neben den
direkt Angesprochenen ein breites Publikum im Blick hatte.
Dass Hieronymus in seiner Vorrede zum Buch selbst Stellung nimmt,
macht auf den ersten Blick den Übergang von der Vorrede zur folgenden
Erzählung geschmeidiger. Schaut man allerdings genauer hin, wird sehr
schnell deutlich, dass Hieronymus mit den wenigen Worten zur Figur der
Iudith die Optik der Lesenden so vorprägt, dass die folgende Erzählung
ganz unter dem Eindruck seiner Einleitungsworte steht. Der zweite Teil
der Vorrede ist somit weit mehr als nur ein geschmeidiger Einstieg in die
Erzählung; vielmehr bewirkt er eine starke Lenkung der Aufmerksamkeit
der Leserinnen und Leser, die nun die Erzählung in der Perspektive lesen,
374 barbara schmitz
die er vorgegeben hat. Dies gilt sowohl für die anzunehmenden Adressa-
ten Heliodorus und Chromatius als auch für die gesamte spätere Leser-
schaft. Die folgende Analyse des Inhalts hat dies mit zu bedenken.
Für die Untersuchung des zweiten Teils der Vorrede stellt sich die
Frage, mit welchen Informationen Hieronymus den Leser versorgt sehen
möchte: Was möchte Hieronymus, dass der Leser über Iudith wisse, bevor
er die Iudith-Erzählung liest?
Iudith wird gleich zu Beginn als „Witwe“ (vidua; Z. 9) eingeführt. Als
ihr Hauptcharakteristikum stellt Hieronymus ihre Keuschheit heraus, die
er in der kurzen praefatio gleich zwei Mal erwähnt (castitatis exemplum,
Z. 9; castitatis remunerator, Z. 11). Hieronymus ehrt Iudith einerseits in
hymnischen Worten: „. . . und verkündigt sie mit triumphierendem Lob in
unablässigen Preisungen!“ (et triumphali laude perpetuis eam 10praeconiis
declarate, Z. 9–10), reduziert aber ihre Vorbildfunktion, die auch zwei Mal
erwähnt wird (castitatis exemplum, Z. 9; hanc . . . imitabilem dedit, Z. 10),
einzig auf ihre Keuschheit. Iudiths Keuschheit sei es—so Hieronymus –,
die Frauen wie Männer nachahmen sollten. Aufgrund ihrer Keuschheit
habe Gott ihr „Tugend/Tüchtigkeit/ Stärke“ (virtus, Z. 11) verliehen, die
sie zu ihrer—übrigens auch zweifach—als übermenschlich gepriesenen
Tat befähigt hätte (1. ut invictum omnibus hominibus vinceret; 2. insupera-
bilem superaret, Z. 12). Damit liefert Hieronymus eine Interpretation der
Erzählung, die den Leserinnen und Lesern schon vor ihrer Lektüre klar
machen soll, dass es nicht Iudith, sondern einzig Gott49 ist, der sie zu der
Tat befähigt hat. Iudiths einziger (!) Beitrag ist ihre Keuschheit. Mit dieser
Interpretation unterscheidet sich die Deutung des Hieronymus grundle-
gend von der LXX-Fassung, in der Judit als selbstbewusste, theologisch
gebildete Frau auftritt, deren wesentliche Qualität nicht ihr keusches
Witwendasein, sondern ihre Gottesfurcht ist ( Jdt 8:8)50. Das Motiv der
Witwenschaft findet sich freilich auch in der LXX-Fassung (χήρα Jdt 8:4, 5,
6; 9:4, 9; 10:3; 16:7), hier aber dient es dazu, Judits Freiheit und Unabhän-
gigkeit, die sich in ihrer Witwenschaft und ihrem Vermögen gründen, zu
erklären, die erst ihr eigenständiges Handeln ermöglichen.
49 Gott wird nicht direkt genannt, sondern nur indirekt als der eigentliche, aber ganz
im Hintergrund bleibende Lenker der Geschichte am Ende gezeichnet. Dies ist insofern
interessant, als die hintergründige, aber nie offen zu Tage tretende Aktivität Gottes die
LXX-Fassung prägt, die Vulgata-Fassung diese aber nicht in gleichem Maße widergibt.
Die der Vulgata-Fassung eigene Gottes-Vorstellung ist ebenfalls zu untersuchen: So fehlt
die für die LXX so wichtige Rettungszusage in Jdt 4:13.
50 Vgl. Schmitz, Gedeutete Geschichte, 154–156.
Zu fragen ist, ob und inwiefern diese Interpretation der praefatio der

Darstellung Iudiths in dem von Hieronymus gestalteten Text entspricht.
Mit anderen Worten: Wie wird die Figur Iudith in der folgenden Textfas-
sung des Hieronymus porträtiert?51
Exemplarisch soll dieser Frage im Folgenden nur an dem für die praefa-
tio zentralen Motiv der „Keuschheit“ nachgegangen werden: Das Motiv
der Keuschheit findet sich in der Iudith-Erzählung nur noch in Idt 15:11
und 16:26. Beide Verse gehören zu den Texten, die sich nur bei Hierony-
mus, nicht aber in der LXX finden52.
In Idt 15:11 preist der Hohepriester Joachim Iudith dafür, dass sie wie
ein Mann gehandelt habe und ihr Herz stark sei, weil sie die Keuschheit
geliebt und sich nach dem Tod ihres Mannes nicht erneut verheiratet
habe (quia fecisti viriliter et confortatum est cor tuum eo quod castitatem
amaveris et post virum tuum alterum non scieris).
Am Ende der Erzählung wird in Idt 16:26 betont, dass sich Iudith
durch Tugend und Keuschheit auszeichnet. Damit werden die aus Idt
15:11 bekannten Motive, „Tugend/Tüchtigkeit/Stärke“ (virtus), Keuschheit
(castitas) und Iudiths Enthaltsamkeit nach dem Tod ihres Mannes wieder
miteinander verbunden, um Iudith zu charakterisieren.
Dass diese Trias der Motive in der Iudith-Erzählung ausschließlich an
diesen beiden „Plus“-Stellen auftreten und zudem als zentrales Motiv
die inhaltliche Präsentation der Iudith-Figur in der praefatio prägen, ist
auffallend. Es könnte entweder daran liegen, dass Hieronymus gerade
diese Motive in die Erzählung eingefügt, oder dass er diese in den ihm
vorliegenden Manuskripten bereits vorgefunden hat. Diese Frage ist nicht
mit Gewissheit zu entscheiden, weil wir nicht wissen, welche Textfassun-
gen Hieronymus vorlagen. Aufgrund der Analysen von Thielmann und
Bogaert gehört der Codex Corbeiensis (= Handschrift 15153) zu den älte-
sten altlateinischen Textfassungen und weist die größte Nähe zur Fassung
des Hieronymus auf. Bogaert vermutet, dass Hieronymus die Textform
51 Zu dieser Frage ist gerade eine von mir betreute Dissertation von Lydia Hilt in Arbeit,
die sich der Figur der Iudith in der Vulgata widmet.
52 In den syrischen Versionen finden sich diese Plus-Texte ebenso nicht, zumal die syri-
schen Versionen der Judit-Erzählung den griechischen Text übersetzen; vgl. L. van Rompay,
„No evil word about her. Two Syriac Versions of the book of Judith,“ in Text, Translation,
and Tradition. Studies on the Peshitta and its use in the Syriac Tradition (FS K. D. Jenner; Hg.
W. T. van Peursen und R. B. Ter Haar Romeny; Leiden: Brill, 2006), 205–230.
53 Die Nummerierung der Handschriften folgt dem Projekt des Vetus-Latina-Instituts.
Paris, Bibliothèque nationale de France lat. 11549 (Corbeiensis 7). Diese ist publiziert von
P.-M. Bogaert, „Recensiones de la vielle version latine de Judith I. Aux origines de la vulgate
hiéronymienne. Le ‚Corbeiensis‛,“ Rbén 85 (1975): 7–37.
376 barbara schmitz
gekannt haben könnte, die in der Complutensis 1 (HSS 109) erhalten ist54.
Signifikanterweise finden sich aber in diesen altlateinischen Textfassun-
gen die Keuschheitsaussagen von Idt 15:11 und 16:26 nicht55. Diese finden
sich ausschließlich als Zusätze in der Iudith-Erzählung des Hieronymus.
Nur in der Vulgata wird Iudiths Leistung auf ihre Keuschheit reduziert, die
der einzige Grund dafür sei, dass Gott ihr virtus verliehen hätte (Idt 15:11;
16:26). Weil diese Aussagen in der LXX ebenso wie in den altlateinischen
Textfassungen, die Hieronymus vorgelegen haben könnten, fehlen, kann
man dies als ein deutliches Indiz dafür werten, dass diese Passagen nicht
aus einer Vorlage stammen, sondern in der Tat von Hieronymus selbst
hinzugefügt wurden. Sie tragen seine Handschrift und zeigen die spezifi-
sche Perspektive des Hieronymus auf die Iudith-Erzählung56.
Angesichts dieses Befundes stellt sich mit Blick auf den textexternen
Bereich die Frage, ob und inwiefern hier ein Frauenideal aus der Zeit des
Hieronymus bzw. sein spezifisches Bild einer Frau Eingang in das Porträt
der Iudith-Figur gefunden hat57.
Zum persönlichen Umfeld von Hieronymus gehörten (reiche) Frauen,
die oftmals aus der römischen Oberschicht stammten und die sich nach
dem Tod ihres Ehemannes einem christlichen, gemäßigt asketischen
Lebensideal verpflichteten und mit ihrem Vermögen u.a. die Projekte des
Hieronymus entscheidend mitfinanzierten. Hieronymus selbst hat Zeit
seines Lebens mit unterschiedlichen asketischen Lebensformen experi-
mentiert: ein eher zurückgezogenes Leben in der Großstadt, ein Leben in
ländlicher Umgebung bei Chalkis oder ein gemeinschaftlich strukturiertes
Zusammenleben in Bethlehem. Dabei darf man sich unter „Askese“ an der
Wende vom 4. zum 5. Jh. kein sich selbst kasteiendes Aussteigerleben in
der Wüste jenseits aller gesellschaftlichen Normen vorstellen. Vielmehr
54 P.-M. Bogaert, „Judith dans la première Bible d’Alcala,“ 116–130. Publiziert in idem,
„La version latine du livre de Judith dans la première Bible d’Alcala,“ Rbén 78 (1968): 7–32;
181–212.
55 Ebenso in den weiteren von Bogaert publizierten Fassungen: P.-M. Bogaert, „Recen-
siones de la vielle version latine de Judith I. Aux origines de la vulgate hiéronymienne. Le
‚Corbeiensis‘; II. Le ‚Monacensis‘; III. La tradition allémanique; IV. Trois Manuscrits et deux
recensions; V. La tradition caroligienne,“ Rbén 85 (1975): 7–37; 241–265; ibid., 86 (1976):
7–37.182–217; ibid., 88 (1978): 7–44.
56 Auch Thielmann, „Beiträge zur Textkritik der Vulgata,“ 59, führt Idt 15:11 und 16:26
auf Hieronymus selbst zurück.
57 Vgl. hierzu die Angaben zu Iudiths Wohnumfeld: Sie wohne im oberen Teil des Hau-
ses in einem abgesonderten Raum (et in superioribus domus suae fecit sibi secretum cubi-
culum in quo cum puellis suis clausa morabatur [Idt 8:5]), was eher an ein Frauenkloster
erinnert als an das offene luftige Zelt der LXX-Fassung, in dem sich über Tag aufzuhalten
sehr angenehm ist (Jdt 8:5). Für den Hinweis danke ich Helmut Engel.
handelt es sich um ein selbst gewähltes, bewusst-bescheiden organisier-

tes, aber letztlich elitäres Leben vermögender Menschen aus der Ober-
schicht, für die diese Lebensform keinen vollständigen Bruch mit ihrem
sozialen Umfeld bedeutete, wohl aber die Möglichkeit bot, ein Leben
jenseits traditioneller Normen zu gestalten58. So bestand der Aufenthalt
von Hieronymus in der „Wüste“ Chalkis in einem Leben nahe der Stadt
Chalkis, wo er sich auf einem ausgedehnten Landgut zwei bis drei Jahre
aufhielt und wissenschaftliche Studien (dort gab es eine Bibliothek!) mit
ländlichem otium verband. Gerade für Frauen bot das christliche Ideal der
Askese die Möglichkeit, nach dem Tod ihres Mannes und ihrer Familien-
phase ihr Leben neu auszurichten. In zahlreichen Briefen hat Hieronymus
verwitwete (vermögende) Frauen zu einem asketischen Leben aufgefor-
dert. Hieronymus propagierte bereits 382–385 ein Virginitätsideal, das zu
einer drastischen Abwertung der Familie führte (vgl. Ep. 22 von 384 an
Paulas Tochter Eustochium), was ihm innerhalb der Kirche viele Feinde
machte. Trotz der vermeintlichen Wertschätzung gegenüber Frauen, die
man in Hieronymus’ Ideen erkennen kann,
leidet die Haltung des Hieronymus gegenüber den Frauen, bedingt nicht
zuletzt durch die Geschichte der Askesebewegung, an der Zweigleisigkeit
und Realitätsferne, die das kirchliche Frauenbild der Epoche allgemein
kennzeichnet. Die Männer der Kirche schätzten gebildete, vermögende
Frauen als Lebensgefährtinnen, als Brief- und Diskussionspartnerinnen, als
finanzielle Helferinnen, als Vertreterinnen des eigenen Lebensideals der
Askese. Diese Wertschätzung von Frauen als Asketinnen beruht allerdings
auf einem Frauenbild, das auf „Vermännlichung“ hinauslief. Zudem blieben
alle Theologen, auch der amts- und institutionskritische Hieronymus, auf
ein Kirchenbild festgelegt, das im Gefolge einer emanzipationsfeindlichen
Denklinie der Antike und einer entsprechenden kirchengeschichtlichen Ent-
wicklung Frauen den Zugang zu Amt, Leitung und Lehre verwehrte und nur
noch karitative Aufgaben erlaubte. Die persönlichen Freiräume der Frauen
um Hieronymus und ihre enorme Bedeutung für die frühe Geschichte des
Mönchtums hatten auch aus der Sicht des Hieronymus keine Auswirkun-
gen auf Bild und Stellung der Frau in der Kirche, die von „Entweiblichung‟,
Unterordnung und Gehorsam geprägt blieben59.
58 Vgl. hierzu B. Feichtinger, „Zäsuren, Brüche, Kontinuitäten. Zur aristokratischen

Metamorphose des christlichen Asketenideals am Beispiel des Hieronymus,“ WSt 110
(1997): 187–220.
59 Fürst, Hieronymus, 54–55. Vgl. auch B. Feichtinger, Apostolae apostolorum. Frauenas-
kese als Befreiung und Zwang bei Hieronymus (Studien zu klassischen Philologie 94; Frank-
furt a.M.: Peter Lang, 1995), 165–235; P. Laurence, Le Monarchismue féminin Antique: Idéal
hiéronymien et réalité historique (Spicilegium Sacrum Lovaniense. Études et Documents 52;
378 barbara schmitz
Vor diesem geistes- und theologiegeschichtlichen Hintergrund stellt sich

die Frage, ob sich die Ideen von Hieronymus auch in der Porträtierung der
Iudith-Figur in seiner Übertragung und Gestaltung der Iudith-Erzählung
spiegeln? Hieronymus’ praefatio mit der Reduktion der Iudith-Figur allein
auf das Ideal der Keuschheit scheint dafür zu sprechen.
Doch nicht allein das: Dadurch, dass Männer wie Frauen in der praefa-
tio erwähnt werden, die sich Iudith als Vorbild nehmen sollen, scheint
Iudith für Hieronymus zugleich ein geschlechterübergreifendes Modell für
enthaltsames, monastisches Leben zu sein. Nicht zu vergessen ist, dass die
Iudith-Fassung ja wahrscheinlich an die Bischöfe Heliodorus und Chroma-
tius gerichtet ist. Interessanterweise hatte Hieronymus mit Chromatius
in Aquileja in einer asketischen Gemeinschaft gelebt. Ist es vor diesem
Hintergrund ein Zufall, dass anscheinend gerade auch an ihn der Prolog
zum Buch Iudith gerichtet ist? Es ist gut möglich, dass dieser Aspekt die
Aufmerksamkeit von Hieronymus an der ihn sonst wenig interessieren-
den Iudith-Erzählung geweckt hat und er sich von diesem Aspekt in seiner
Textgestaltung maßgeblich hat leiten lassen, da er ja, wie er selbst sagt,
mehr dem Sinn als dem Wortlaut nach übertragen hat. Ein solches Motiv
ist umso plausibler, als Hieronymus zur Zeit der Übertragung des Iudith-
Buches eine hoch emotionale Schrift Contra Vigilantium (406) verfasste,
in der er sein asketisches Lebensideal gegen den Asketen-Skeptiker Vigi-
lantius verteidigte.
Die praefatio zum Buch Iudith legt somit nahe, dass sich die Unter-
schiede der Judit-Erzählung der LXX zur Vulgata-Fassung nicht allein
auf die Hieronymus vorliegenden Textfassungen zurückführen lassen,
sondern dass man auch mit einem hohen Anteil des Hieronymus an der
Gestaltung der Iudith-Erzählung zu rechnen hat.
3. Fazit
Ausgangspunkt war die Frage nach der Originalsprache der Juditerzäh-

lung. Lange ist man von einem hebräischen Original ausgegangen, das—
wie auch immer die Modelle im Einzelnen aussahen—sowohl hinter der
LXX-, als auch hinter der Vulgata-Fassung angenommen wurde60.
Leuven, 2010); K. Cooper, The Virgin and the Bride. Idealized Womanhood in Late Antiquity
(Cambridge / London: Harvard University Press, 1996), 68–115.
60 Vgl. hierzu A. M. Dubarle, Judith. Formes et sens des diverses tradition. Tome I: Etudes,
Tome II: Textes (AnBib 24,1–2; Rom: Institut Biblique Pontifical, 1966).
Für die LXX war die Annahme eines hebräischen Originaltextes jahr-
zehntelang in der Juditforschung leitend. Diese Auffassung ist aber in der
jüngsten Zeit neu bewertet worden: Das bisherige Minderheitsvotum für
eine auf Griechisch entstandene LXX-Fassung gewinnt in den letzten Jah-
ren immer mehr an Plausibilität.
Das andere wesentliche Argument für die Annahme eines hebräischen
„Ur“-Textes war der Hinweis des Hieronymus in seiner praefatio, dass ihm
ein „chaldäischer“ Text vorgelegen habe. Eine genauere Analyse der praefa-
tio kann erstens deutlich machen, dass die in der Vorrede beschriebene
Arbeitsweise mit den mehrfachen Übertragungen („chaldäisch“–hebrä-
isch–lateinisch) den Schluss von der uns vorliegenden Vulgata-Fassung
auf den von Hieronymus (als Einzigem) behaupteten „chaldäischen“ Text
erschwert.
Am Beispiel des Motivs „Keuschheit“ konnte zweitens gezeigt werden,
dass der praefatio eigene Gestaltungsprinzipien zugrunde liegen, die theo-
logischen und anthropologischen Vorstellungen im Buch Iudith entspre-
chen. Dies spricht dafür, dass Hieronymus mit großer Freiheit den Stoff
seiner Iudith-Erzählung gestaltet hat. Von diesen ersten Analysen her ist
mit einem hohen Eigenanteil des Hieronymus an seiner Textfassung zu
rechnen.
Daher steht es m.E. für die Erforschung der Judit-Tradition an, das spe-
zifische Konzept der Vulgata herauszuarbeiten und in sich zu erforschen.
Damit lassen sich die Fragen an die Vorgeschichte des Textes zwar nicht
klären, aber es könnte das theologische Konzept deutlicher zu Tage treten
und damit neue Perspektiven auf die Unterschiede bzw. die Gemeinsam-
keiten zur LXX werfen. Dies könnte nicht nur zu Erhellung der Gesamtsi-
tuation des Judit-Stoffs beitragen, sondern würde auch eine Erforschung
genau jener Textfassung(en) bedeuten, die für die reiche Rezeption des
Judit-Stoffs in der abendländischen Kunst-, Musik- und Literaturge-
schichte leitend waren.
So war es Martin Luther, der die Vulgata-Fassung des Hieronymus im
Westen mit populär gemacht hat: Obwohl er das Buch Judit—ähnlich wie
Hieronymus es in seiner praefatio betont—als „apokryphes“ Buch behan-
delt, so hat Luther das Buch Judit in seiner Vorrede als ein „fein/gut/heilig/
nützlich Buch“61 bezeichnet—übersetzt aber wurde nicht die griechische,
sondern die lateinische Textfassung des Hieronymus.
61 „Darumb ists ein fein/gut/heilig/nützlich Buch.“ M. Luther, Die Gantze Heilige Schrift
Deutsch (Wittenberg: Hans Lufft, 1545; [Nachdruck 1973, Band II]).
PsalmsLXX and the Christian Definition of Space:
Examples Based on Inscriptions from
Central Asia Minor1
Cilliers Breytenbach
1. Introduction
Jewish influence on some of the Christians from late antiquity can be

detected on material objects like inscriptions. Sometimes it is proved ono-
mastically by the use of originally Jewish names2 like Μαρία3 (for ‫)מרים‬
and Ἰωάννης 4 (for ‫)יוחנן‬, or iconographically by traditional Jewish sym-
bols, like a palm branch incised on a rough block from Çarıklar, 25 km
south-southeast of Iconium.5 Sometimes the language of the inscription
itself documents Jewish influence. On an inscription from Iconium the
initial epithet ὁ θε(ὸς) τῶν φ<υλ>ῶν τοῦ Ἰστρα|ήλ . . . and the final invo-
cation “we beseech you in the name of the almighty God” (ἐνορκιζόμ[ε-]
θ[α] | τὸν παντ[ο]κράτο<ρ>α θ(εὸ)ν ΠΑ | — — —)6 prove Jewish influence.
Warnings like the latter against the unauthorized re-use of graves are not
1 This research was made possible by the DFG Excellence Cluster 264 TOPOI: The For-
mation and Transformation of Space and Knowledge in Ancient Civilizations (www.topoi.
org). For readers unfamiliar with the epigrapher’s habit, abbreviations of epigraphical
corpora have been decoded at their first occurrence. For other abbreviations, cf. P. H.
Alexander et al., eds., The SBL Handbook of Style: For Ancient Near Eastern Studies, Biblical,
and Early Christian Studies (Peabody: Hendrickson, 1999).
2 Cf. T. Ilan, Lexicon of Jewish Names in Late Antiquity (4 vols.; TSAJ; Tübingen: Mohr
Siebeck, 2002sqq.).
3 Cf. W. M. Calder, Monumenta Asiae Minoris Antiqua (MAMA I) (Manchester: Man-
chester University Press, 1928), nos. 168, 169, 268; idem, Monuments from Eastern Phrygia
(MAMA VII) (Manchester: Manchester University Press, 1956), no. 484.
4 E.g. MAMA I nos. 186, 312, 365; MAMA VII nos. 102, 574, 593; G. Laminger-Pascher, Die
kaiserzeitlichen Inschriften Lykaoniens, Vol. 1: Der Süden (Ergänzungsbände zu den Tituli
Asiae Minoris [ETAM] 15; Vienna: Österreichische Akademie der Wissenschaften, 1992),
no. 143.
5 W. M. Calder and M. H. Balance, Monuments from Lycaonia, the Pisido-Phrygian
Borderland, Aphrodisias (MAMA VIII) (Manchester: Manchester University Press, 1962),
no. 304.
6 W. Ameling, Inscriptiones Judaicae Orientis, Vol. 2: Kleinasien (IJO II) (TSAJ 99; Tübin-
gen: Mohr Siebeck, 2004), no. 226.
382 cilliers breytenbach
uncommon.7 Often the Christians took over the Jewish practice to allude
to the legal stipulations of Deuteronomy.8
2. Allusions
Amongst Christian inscriptions there are also allusions from the Psalms.9
From Anazarbus in Cilicia a part of an arch has survived. During the 5th or
6th centuries a benediction εὐλογ[ητὸς ὁ θ]εός was engraved on it.10 This
benediction is well known from the Psalms. In the majority of cases in the
LXX God or the Lord is being praised for delivering the speaker(s) from
a perilous situation.11 Christian reception of what seems to be a liturgical
tradition is already documented by the Pauline letters.12 A marble block
with a cross in the pediment and two leaves from Savatra (Yağlıbayat),
55 kilometres east of ancient Iconium in Lycaonia marked the 5th/6th
centuries grave of a very notable (περίφρων) priest named Παῦλος. Using a
7 Supplementum epigraphicum graecum (SEG) 30 (1980) no. 1060 (Naxos); SEG 44

(1994) no. 765 (Katane).
8 Cf. L. Robert, “Malédictions funéraires grecques,” CRAI (1978): 241–289, 246–249;
P. R. Trebilco, Jewish Communities in Asia Minor (SNTSMS 69; Cambridge: Cambridge Uni-
versity Press, 1991), 60–69; J. H. M. Strubbe, “Curses Against Violation of the Grave in Jewish
Epitaphs from Asia Minor,” in Studies in Early Jewish Epigraphy (ed. J. W. van Henten and
P. W. van der Horst; AGJU 21; Leiden: Brill, 1994), 70–128.
9 This applies to Jewish inscriptions to a lesser extent. But see the line on a bronze seal
from Neapolis (modern Naples) in Campania (‫ )יהי שלום‬where Ps 122:7 is cited; cf. D. Noy,
Jewish Inscriptions of Western Europe, Vol. 1: Italy (excluding the City of Rome), Spain and
Gaul (JIWE I; Cambridge: Cambridge University Press, 1993), no. 25. Ps 128:6 “Peace upon
Israel” is cited in Palestinian synagogues.
10 Cf. M. H. Sayar, Die Inschriften von Anazarbos und Umgebung, Vol. 1: Inschriften aus
dem Stadtgebiet und der nächsten Umgebung der Stadt (Inschriften griechischer Städte aus
Kleinasien [IK] 56/1; Bonn: Habelt, 2000), no. 657.
11 E.g. Gen 14:20—καὶ εὐλογητὸς ὁ θεὸς ὁ ὕψιστος, ὃς παρέδωκεν τοὺς ἐχθρούς σου ὑπο-
χειρίους σοι. καὶ ἔδωκεν αὐτῷ δεκάτην ἀπὸ πάντων; Ps 17:47—ζῇ κύριος, καὶ εὐλογητὸς ὁ θεός
μου, καὶ ὑψωθήτω ὁ θεὸς τῆς σωτηρίας μου; Ps 65:20—εὐλογητὸς ὁ θεός, ὃς οὐκ ἀπέστησεν τὴν
προσευχήν μου καὶ τὸ ἔλεος αὐτοῦ ἀπ᾽ ἐμοῦ; Ps 67:36—θαυμαστὸς ὁ θεὸς ἐν τοῖς ἁγίοις αὐτοῦ∙ ὁ
θεὸς Ισραηλ αὐτὸς δώσει δύναμιν καὶ κραταίωσιν τῷ λαῷ αὐτοῦ. εὐλογητὸς ὁ θεός; Dan θ’ 3:95—
καὶ ἀπεκρίθη Ναβουχοδονοσορ καὶ εἶπεν εὐλογητὸς ὁ θεὸς τοῦ Σεδραχ, Μισαχ, Αβδεναγω, ὃς
ἀπέστειλεν τὸν ἄγγελον αὐτοῦ καὶ ἐξείλατο τοὺς παῖδας αὐτοῦ, ὅτι ἐπεποίθεισαν ἐπ᾽ αὐτῷ καὶ τὸ
ῥῆμα τοῦ βασιλέως ἠλλοίωσαν καὶ παρέδωκαν τὰ σώματα αὐτῶν εἰς πῦρ, ὅπως μὴ λατρεύσωσιν
μηδὲ προσκυνήσωσιν παντὶ θεῷ ἀλλ᾽ ἢ τῷ θεῷ αὐτῶν.
12 E.g. 2 Cor 1:3—Εὐλογητὸς ὁ θεὸς καὶ πατὴρ τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ, ὁ πατὴρ τῶν
οἰκτιρμῶν καὶ θεὸς πάσης παρακλήσεως; 1 Pet 1:3 according to P72 and minuscule 33—Εὐλο-
γητὸς ὁ θεὸς καὶ πατὴρ τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ, ὁ κατὰ πολυέλεος αὐτοῦ ἀναγεννήσας
ἡμᾶς εἰς ἐλπίδα ζῶσαν δι᾽ ἀναστάσεως Ἰησοῦ Χριστοῦ ἐκ νεκρῶν. God is qualified as πατὴρ τοῦ
κυρίου ἡμῶν Ἰησοῦ Χριστοῦ. Motivations for the benediction are still God’s compassion and
abundant mercy. Cf. also Eph 1:3.
psalmslxx and the christian definition of space 383
phrase from the Psalms,13 he invokes God in his sorrow for the loss of his
beloved brother: θε(ὸς) βοήθι . . .14 The lines from the Psalm at the begin-
ning of the inscription should help this Paul to overcome the deep sor-
row at the grave of his brother. The fact that such allusions to the Greek
Psalms were widespread amongst Christians, is counterbalanced by the
observation that there are few direct quotations. In studies on the use of
the Old Greek translations of the Hebrew Bible, this evidence from epig-
raphy is often neglected.
3. Direct Citations
The quotation of passages from the Bible in Hebrew or in Greek trans-

lation in Jewish inscriptions from antiquity is not very common.15 In
Jewish inscriptions there are several quotes from Deuteronomy.16 Prov 10:7
is quoted eleven, 1 Sam 25:29 four times, Zech 5:1–5 twice and Prov 6:23,17
PsLXX 92:5,18 122:7 and 135:25 one time each. In late antiquity Christians
quoted their Greek Old Testament more often, especially specific Psalms.19
13 Cf. PsLXX 53:6 (ἰδοὺ γὰρ ὁ θεὸς βοηθεῖ μοι); 69:6 (ὁ θεός, βοήθησόν μοι); 78:9 (βοήθησον
ἡμῖν, ὁ θεὸς ὁ σωτὴρ ἡμῶν); 93:18 (κύριε, βοηθεῖ μοι); 108:26 (βοήθησόν μοι, κύριε ὁ θεός μου,
σῶσόν με κατὰ τὸ ἔλεός σου); 118:86 (βοήθησόν μοι), 117 (βοήθησόν μοι, καὶ σωθήσομαι).
14 SEG 52 (2002) no. 1459 (Savatra) = I.Konya; cf. B. H. McLean, Greek and Latin Inscrip-
tions in the Konya Archeological Museum (Regional Epigraphic Catalogues of Asia Minor
[RECAM] IV; London: British Institute of Archaeology at Ankara, 2002), no. 206 (with
photo, photo of squeeze and translation).
15 In Jewish inscriptions published in JIWE, in G. Lüderitz, Corpus jüdischer Zeugnisse
aus der Cyrenaika (CJZC) (Beihefte zum Tübinger Atlas des Vorderen Orients, Reihe B, 53;
Wiesbaden: Ludwig Reichert Verlag, 1983), and Corpus Inscriptionum Iudaeae/Palaestinae:
A Multi-Lingual Corpus of the Inscriptions from Alexander to Muhammad, Vol. 1 (CIIP I) (ed.
H. M. Cotton-Paltiel et al.; Berlin: De Gruyter, 2010) there are no quotations. In both IJO II
(no. 153) and CIIP II (no. 1348) single quotations from the Psalms have been found.
16 From Palmyra in southern Syria, Hebrew citations from Deut 6:4–9 (D. Noy and
H. Bloedhorn (ed.), Inscriptiones Judaicae Orientis, Vol. 3: Syria und Cyprus [IJO III] [TSAJ
102; Tübingen: Mohr Siebeck, 2004], no. 44), 7:14 (IJO III no. 45), 7:15 (IJO III no. 46), 28:5
(IJO III no. 47) on doorways are known.
17 Cf. the list by S. Fine and L. V. Rutgers, “New Light on Judaism in Asia Minor During
Late Antiquity: Two Recently Identified Inscribed Menorahs,” JSQ 3 (1996): 1–23, 8; P. W.
van der Horst, Ancient Jewish Epitaphs: An Introductory Survey of a Millennium of Jewish
Funerary Epigraphy (300 BCE–700 CE ) (CBET 2; Kampen: Kok Pharos, 1991), 37–39.
18 Cf. CIIP II no. 1348.
19 Cf. D. Feissel, “Notes d’épigraphie chrétienne VII,” BCH 108/1 (1984): 545–579, and
A. E. Felle, Biblia Epigraphica: La sacra scrittura nella documentazione epigrafica dell’Orbis
Christianvs antiqvvs (III–VIII secolo) (Bari: Edipuglia, 2006).
In the following attention will be given to cases from Asia Minor,20 where
lines from the Psalms are cited.21
3.1. Nicea
We start with the only Jewish example from Asia Minor. In the edition of
Inscriptiones Judaicae Orientis on Asia Minor, Walter Ameling22 registered
one on a marble block from ancient Nicea (Iznik). The menorah above
the inscription of PsLXX 135:25 clearly marks it as originally Jewish.23 It has
been reused twice:
1. IJO II no. 153 διδοὺς ἀρ|[τ]ὸν τῖ πάσι σ|αρκί, ὅτι εἰ[ς] | ἐῶνα ἔλεο[ς] | αὐτοῦ
PsLXX 135:25 ὁ διδοὺς τροφὴν πάσῃ σαρκί,

ὅτι εἰς τὸν αἰῶνα τὸ ἔλεος αὐτοῦ.
He, who gives nourishment to all flesh—
because his mercy is forever. (trans. NETS)24
PsLXX 135:25 uses τροφή (nourishment, food),25 to translate ‫לחם‬, differing

from the common usage in the LXX. This word is a little less specific than
the normal translation equivalent ἄρτος,26 signifying a loaf of baked wheat
bread.27 The inscription follows this latter normal use. The use of ε for αι
in αἰῶνα follows the normal pattern since Roman times28 and documents
20 There are many more quotations on monuments from Syria, Palestine and Egypt.
Cf. L. Jalabert, “Citations bibliques dans l’épigraphie grecque,” in vol. 3/2 of Dictionnaire
d’archéologie chrétienne et de liturgie (ed. F. Chabrol and H. Leclercq; Paris: Letouzey,
1914), 1731–1756, and the more recent review by D. Feissel, “La Bible dans les inscriptions
grecques,” in Le monde grec ancien et la Bible (ed. C. Mondésert; Paris: Beauchesne, 1984),
223–231.
21 The citations from IsaiahLXX will be treated elsewhere; cf. Corpus inscriptionum grae-
carum (CIG) IV no. 8935; W. M. Ramsay, The Cities and Bishoprics of Phrygia: Being an
Essay of the Local History of Phrygia from the Earliest Times to the Turkish Conquest, Vol. 1/2:
West and West-Central Phrygia (Oxford: Clarendon Press, 1897), 740, nos. 674–676. Cf.
Eusebius Hist. eccl. 10.4.48–52.
22 Cf. supra n. 6.
23 Cf. Fine and Rutgers, “New Light,” 1–10.
24 A. Pietersma and B. G. Wright, eds., A New English Translation of the Septuagint
(Oxford: Oxford University Press, 2007).
25 Alexandrinus and Sinaiticus.
26 The same reading occurs in α’ and σ’.
27 Cf. LSJ, s.v.
28 Cf. F. T. Gignac, A Grammar of the Greek Papyri of the Roman and the Byzantine Peri-
ods, Vol. 1: Phonology (Testi e documenti per lo studio dell’antichità 55/1; Milan: Cisalpino-
Goliardica, 1976), 191–193.
the rule of the spoken sound, this might also be the reason why the “food”
became more specifically “bread”. But the text used could have been
Aquila,29 whose translation also reads ἄρτον. It is possible that the ash-
lar stems from a soup kitchen in a synagogue,30 thus illustrating how the
Psalm functions to place the location of the common daily meal under
the mercy of God.
3.2. Antioch in Pisidia
Our first example from Christianity comes from Antioch on the border
of Pisidia. The text was amongst the previously unpublished inscriptions
in the note books of William Ramsay. These inscriptions have now been
edited and published in Inschriften griechischer Städte aus Kleinasien.31
During 1912/13 Ramsay recorded the prayer of the Christian with the
Roman family name Sergius from the Sofular quarter. Sergius, also known
as Φολέας, was Christian, as can be inferred from the cross.
2. IK 67 no. 109 + Εὐχὴ Σεργίου τοῦ λεγομένου Φολέα· δήκασο-

ν κ(ύρι)ε τὺς ἀδηκõντας με· πολεμήσε τοὺς πολε-
μõντα με· ὁ θ(εό)ς τὴν ἔνεσή μου μὴ παρασηοπή-
σης ὅτη στόμα ἁμαρτολοῦ κὲ ἔστομα δο-
λήου ἐπ᾽ ἐμεῦ ὐνύκθη.32
PsLXX 34:1 δίκασον, κύριε, τοὺς ἀδικοῦντάς με,33

πολέμησον τοὺς πολεμοῦντάς με.
Render judgment, O Lord, on those who do me wrong;
fight against those who fight against me! (trans. NETS)
PsLXX 108:1–2 (1) ὁ θεός, τὴν αἴνεσίν μου μὴ παρασιωπήσῃς,

(2) ὅτι στόμα ἁμαρτωλοῦ καὶ στόμα δολίου
ἐπ᾽ ἐμὲ ἠνοίχθη.
(1) O God, do not pass over my praise in silence,
(2) because a sinner’s mouth and a deceiver’s mouth—
it was opened against me. (trans. NETS)
29 The extra article τῖ (for τῇ) leads to Aquila; pace C. Zuckerman, “Psalms 135:25 in
Symmachus’ Translation on a Jewish Inscription from Nicaea (Iznik),” Scripta Classica
Israelica 20 (2001): 105–111.
30 Cf. Fine and Rutgers, “New Light,” 17.
31 M. A. Byrne and G. Labarre, Nouvelles inscriptions d’Antioche de Pisidie: D’après les
Note-books de W. M. Ramsay (IK 67; Bonn: Habelt, 2006).
32 Ibid., 56 (facsimile of Ramsay’s transcription, 144).
33 α’—δίκασαι, κύριε, τοῖς δικαζομένοις μοι.
The inscription intentionally connected two quotations, PsLXX 34:1 and

PsLXX 108:1–2. For the omitted superscription, Ps 34:1 reoccurs complete.
The vowel ι was changed to η, the diphthong ου contracted to υ or ο
(cf. δήκασον κ(υρί)ε34 τὺς δηκõντας; πολεμõντα and τὴν ἔνεσή μου).35 The
weak aorist imperative πολέμησον was changed into a strong form
πολέμησε. The omitted ς of πολεμõντα might be unintentional. All these
changes suggest that the mason chiselled according to the spoken form
of the text.
PsLXX 108:1–2 confirms this hypothesis. First and foremost, the word
denoting “praise in a religious sense and vocally”, αἴνεσις, does not appear
in Greek prior to the LXX.36 Here it is written as it was pronounced: ἔνεση.
The change of ι to η also continues, the diphthong αι contracted to ε and
there is no distinction indicated by a longer ω in relation to a shorter o (cf.
μὴ παρασηοπήσης ὅτη στόμα ἁμαρτολοῦ κὲ ἔστομα δολήου).37 In the case of
ἠνοίχθη, the η is changed and the diphthong contracted to an υ.
Unfortunately there is no information about the stone in Ramsay’s note
book. The only other instance where PsLXX 34 is quoted however, is in a
funerary context.38 The interpretation of Ramsay’s transcription can be
based on the text only. The εὐχή after the cross marks the following lines
as the prayer of the Christian Sergius. The supplicant re-uses the biblical
invocations κύριε and ὁ θ(εό)ς,39 asking the Lord to condemn his wrongdo-
ers, to wage war against his enemies, praying to God not to remain silent40
about Sergius’ praise. The reason is introduced by a ὅτι clause: Sergius was
defamed. The mouth of a sinner, a mouth of deceit was opened against
him, vilifying him. Sergius put up the stone with the inscription in order to
34 The invocation κύριε is not omitted. Instead of reading an adversative κέ (= καί) in

line 2, as Ramsay did, κε should be taken as nomen sacrum; Cf. the comment and transla-
tion of Byrne and Labarre, Nouvelles inscriptions, 57.
35 The change from ι to η is well documented in Roman and Byzantine papyri (cf. Gignac,
Grammar, 237–238). The interchange of ου with ο is frequent, with υ occasional (cf. ibid.,
208, 211–212, 214–215).
36 Cf. T. Muraoka, A Greek-English Lexicon of the Septuagint (Louvain / Paris / Walpole:
Peeters, 2009), s.v.
37 Frequent throughout the Roman and Byzantine periods (cf. Gignac, Grammar, 275–
277).
38 PsLXX 34:4 is quoted on the lid of a 6th century larnax in the Cyprus Museum (note
also the text of PsLXX 24:18 on a broken marble plate in the Museum); cf. T. B. Mitford,
“Some New Inscriptions from Early Christian Cyprus,” Byzantion 20 (1950): 105–175, no. 9
(and 21).
39 The nominative is used as vocative.
40 With reference to Polybius, LSJ, s.v. glosses “to pass over in silence”.
counter the slander. He publicly prayed to God to come to his assistance

and to punish his enemies. In order to achieve this, he combined two
Psalms from the Greek Bible on the slab, creating a locus to influence
common opinion about him.
3.3. Anazarbus
The second Christian inscription to be treated comes from the southern
edge of Anazarbus (near the modern Dilekkaya) in ancient Cilicia. The
city lies north of the bay of Issus on the river Pyramus, which winds its
way through the Cilicia plains. The inscription was hewn high up into the
rock face of the southern side of a cleft splitting the limestone crag on
which the city was built.41 It is unclear whether the earthquakes of A.D.
525 and 561 which levelled the city caused or damaged the cleft.42 Be it as
it may, the cleft was used as eastern access road to the city.43 The text of
the 5th or 6th centuries inscription conforms to the standard reading of
the major LXX manuscripts.44
3. IK 56/1 no. 60 ὁ θεὸς ἡμῶ|ν καταφυγὴ | καὶ δύναμις.44
PsLXX 45:2 ὁ θεὸς ἡμῶν καταφυγὴ καὶ δύναμις

God is our refuge and power (trans. NETS)
Neither the Hebrew text, nor the Greek translation in the LXX manuscripts
should steer the interpretation of the text of this inscription. The current
text and its context should set the limits for our interpretation. Since the
gorge was used as access road, this was a dangerous and scary place to
pass. “The cleft is 250 m. in length and varies between 4.00 m. and 15.00
m. in width. On either side of the path the rock rises to a height of about
50 m.”45 The point is, the Psalm was used to give reassurance in this dan-
gerous situation. It should reassure the travellers that even in the narrow
space at the bottom of the cleft, God is their refuge and power.
41 For more detail see M. Gough, “Anazarbus,” AnSt 2 (1952): 85–150.

42 For more detail see M. H. Sayar, “Anazarbus,” Brill’s New Pauly, online: http://
referenceworks.brillonline.com/entries/brill-s-new-pauly/anazarbus-e120710; accessed
24/2/2012. The inscription is unscathed and could predate the rebuilding under Justin I
and Justianus I.
43 Cf. Gough, “Anazarbus,” 102.
44 Cf. Sayar, Inschriften von Anazarbos, 52 (photo: plate XXIII).
45 Gough, “Anazarbus,” 102.
From Semendere (modern Ovacık in the Tuneeli province) comes a

limestone plate which marked the grave of Theodosius. Konon, an Isaurian
saint, is mentioned in the first line. The line alludes to PsLXX 45:2: + ἅγιε
Κόνον, σὲ κατέφυγα.46 Ἁγεῖα καταφυγή becomes an epitheton for God the
most high God (θέος ὑψίστος).47
An interesting parallel comes from the summit of Kolokytha, an islet
east of the island of Spinalonga (in the Gulf of Elounda in north-eastern
Crete). Here PsLXX 20:3 is cited as part of a petition for help on a 4th cen-
tury inscription hewn into the uneven surface of the rock:
4. Bandy no. 68 + εὐχαριστῶ σε, Κ(ύρι)ε, ὥτι τὴν

ἐπιθυμείαν τῆς κραδί-
ας μου εἴδωκάς μη καὶ τὴν θέλη-
σιν τῶν χειλέων μου οὐκ ἐστέρησάς [με.(?)]
Κ(ύρι)ε, Ἰ(ησο)ῦ, Κ(ύρι)ε, φύλαξον τὸν δοῦλόν σου
[Ἁδριανόν·
Χ(ριστ)έ, διάσωσων Κώνωνα.48
PsLXX 20:3 τὴν ἐπιθυμίαν τῆς ψυχῆς αὐτοῦ ἔδωκας αὐτῷ

καὶ τὴν θέλησιν τῶν χειλέων αὐτοῦ οὐκ ἐστέρησας αὐτόν.
The desire of his heart you gave him,
and of the wish of his lips you did not deprive him.
(trans. NETS)
The introductory thanksgiving εὐχαριστῶ σε, Κ(ύρι)ε, addressed to the

Lord Jesus is motivated by introducing Ps 20:3 with ὥτι. In applying the
Psalm to the situation of the supplicant, the third person is consequently
altered into the first person.49 Again the quote follows the spoken version,
although there are translations supporting κραδίας50 as translation for ‫לב‬
instead of ψυχῆς.51 With the words of Ps 20:3 the supplicant thanks the
Lord for always giving him the desire of his heart and not refusing him
the wish of his lips. Against this backdrop the Lord is addressed again:
46 D. Berges and J. Nollé, Tyana: Archäologisch-historische Untersuchungen zum

südwestlichen Kappadokien, Vol. 1 (IK 55/1; Bonn: Habelt, 2000), no. 17. Cf. also H. Rott,
Kleinasiatische Denkmäler aus Pisidien, Pamphylien, Kappadokien und Lykien: Darstellender
Teil (Leipzig: Dieterich, 1908), 372–373, no. 85.
47 Cf. SEG 19 (1963) no. 852 (Sibidunda in Pisidia).
48 A. C. Bandy, The Greek Christian Inscriptions of Crete, Vol. 10/1: IV–IX A.D. (Athens:
Christian Archaeological Society, 1970), no. 68.
49 Μη instead of μοι.
50 Κραδίας is an epic form of καρδίας; cf. LSJ, s.v.
51 In the Latin and Sahidic translations.
Κ(ύρι)ε, Ἰ(ησο)ῦ. The real plea is formulated using the imperative aorist.
“Lord, guard your servant Hadrian; Christ, let Konon preserve.” Possibly
Hadrian and Konon were ship wrecked and reached the islet, praying for
help. After they were rescued they came back, expressing their gratitude
by inscribing the rock of their anguish with the line from Ps 20:3. Maybe
they were still castaways and a relative inscribed the rock. We would not
know, but we can see how Christians expressed their fear and hope in
the language of the Greek Psalms, redefining the places of their anguish
in spaces of hope.
3.4. Kesmez and Laranda

Back in Asia Minor, we move westward through the Tarsian gates into
east Lycaonia. A 5th century inscription introduced with a forked cross
from the village Kesmez east of Karapınar52 in the region of Thebasa also
cites Ps 117:20 on two tabulae ansatae on a lintel found amongst the spoils
of a church:
5. ETAM 15 no. 158 I + αὕτη ἡ πύλη | εἰσ]ελεύσον
(ornament)
II τοῦ κ(υρίο)υ. δίκέοι | [ται] ἐν [α]ὐτῇ.
PsLXX 117:20 αὕτη ἡ πύλη τοῦ κυρίου,

δίκαιοι εἰσελεύσονται ἐν αὐτῇ.
This is the gate of the Lord;
righteous ones shall enter in it. (trans. NETS)
At Dereköy/Fisandon, 9 km south southeast of Laranda, also in eastern

Lycaonia the first two lines of a lintel above the northern entrance of a
church from the 9th century cites PsLXX 117:20, too; the third line cannot
be related to the Greek Bible. That what is left of the broken block might
be: [κ]ύ|ριε βοωήθι.
6. ETAM 15 no. 58 αὕτη ἡ πύλη τοῦ κυ/ρίου δίκεοι εἰσελ[εύσ]οντε ἐν αὐτῇ.|

κύριε ῥώνο/σον τοὺς [ . . . . . . . . ]
[κ]ύ/ριε β . . ωηθ . . . ς[ . . . . . . . . . . . . . . . . . . . ].
52 Karapınar lies 91 km east-southeast of Konya.

The impressive lintel of the 6th century church of the archangel Michael
in Alakilise in ancient Lycia is also inscribed with PsLXX 117:20.53 The use of
the verse on the lintels above entrances to the narthex was widespread, as
can be seen from Elaioussa-Sebaste in Cilicia,54 from Komana-Hierapolis in
Cappadocia,55 from Myra56 and from Alacahisar in Lycia,57 from Germia
in Galatia,58 from Gortyna on Crete,59 Kerkyra on Corfu60 and Karpathus.61
These are merely some examples from Asia Minor, Greece and the Islands.
East of the Amanus there are many more cases.62
A decorated epistyle reused as lintel in a 4th(?) century church in
Afyonkarahisar in south-east Phrygia has a catena of texts from the Psalms
(PsLXX 31:1; 33:9, 6; 26:1; 96:11) on the flat back side.63 In the Byzantine
period Psalms were increasingly quoted on lintels,64 crosses65 and even
floors66 over a widespread area, transferring the idea that God dwells in
the temple in Jerusalem to the churches.67
53 Cf. Rott, Kleinasiatische Denkmäler, 322–323. Also cited on lintels in Cataonia in

ancient Cappadocia (cf. W. H. Waddington, “Inscriptions grecques et latines de Cataonie,”
BCH 7 [1883]: 125–148, no. 25) and in Sivas in ancient Armenia (cf. H. Grégoire, “Rapport sur
un voyage d’exploration dans le Pont et en Cappadoce,” BCH 33 [1909]: 3–169, no. 17).
54 Cf. G. Dragon and D. Feissel, Inscriptions de Cilicie (Travaux et mémoires du Centre
de recherche d’histoire et civilisation de Byzance 4; Paris: De Boccard, 1987), no. 21.
55 Cf. R. P. Harper, “Tituli Comanorum Cappadociae,” AnSt 18 (1968): 93–147, 135, no. 8.01.
56 Cf. Felle, Biblia Epigraphica, no. 480.
57 Cf. R. M. Harrison, “Churches and Chapels of Central Lydia,” AnSt 13 (1963): 117–151,
136, no. 12.
58 Cf. SEG 36 (1986) no. 1190.
59 Cf. M. Guarducci and F. Halbherr, Tituli Gortynii (Inscriptiones Creticae [IC] IV;
Rome: Libreria dello Stato, 1950), no. 473.
60 Cf. W. Dittenberger, Inscriptiones Phocidis, Locridis, Aetoliae, Acarnaniae, insularum
maris Ionii (Inscriptiones Graecae [IG] IX/1; Berlin: Georg Reimer, 1897), no. 720.
61 Cf. Felle, Biblia Epigraphica, no. 470.
62 Cf. PsLXX 117:20 on lintels in Syria (CIG IV nos. 8930–8934); Ps 33:6 in Attica (IG II/
III2 no. 13302).
63 W. H. Buckler, Monuments and Documents from Phrygia and Caria (MAMA VI) (Man-
chester: Manchester University Press, 1939), no. 385: [μακάρι]οι ὧν ἀφέθησαν αἱ ἀνομίαι καὶ
ὧν ἐπεκαλύφθη|[σαν αἱ] ἁμαρτίαι. γεύσασθε καὶ εἴδεται ὅτι χρηστὸς | [ὁ Κ(ύριο)ς. π]ροσέλθατε
πρὸς αὐτὸν καὶ φωτεισθῆται καὶ τὰ |[πρόσ]ωπα ὑμῶν οὐ μὴ καταισχυνθῇ. Κ(ύριο)ς φωτισμός
||[μου κ]αὶ σωτήρ μου, τίνα φοβηθήσομαι; Κ(ύριο)ς ὑπερασπισ|[τὴς τῆς ζω]ῆς μου, ἀπὸ τίνος
διλειάσω; φῶς ἀνέτειλε |[τῷ δικαίῳ καὶ] τοῖς εὐθέσει τῇ καρδίᾳ εὐφροσύνη. Apart from ι written
as ει in lines 2, 3 and 7, the text matches the standard LXX manuscripts.
64 Cf. Bandy, Greek Christian Inscriptions of Crete, no. 53 (6th–8th centuries in Galia).
65 Cf. the florilegium of citations from the Psalms in ‘Edjāz (M. von Oppenheim and
H. Lucas, “Griechische und lateinische Inschriften aus Syrien, Mesopotamien und Kleina-
sien,” ByzZ 14 [1905]: 1–72, no. 78).
66 In a mosaic floor of a Byzantine church in the Lycian Arycanda PsLXX 5:8 is cited. Cf.
S. Şahin, Die Inschriften von Arykanda (IK 48; Bonn: Habelt, 1994), no. 310.
67 Cf. the full text of PsLXX 14 engraved during the 4th century on a white marble slab
from Lapēthos (modern Karavas) in Cyprus. (The initial publication by P. Perdrizet,
3.5. Savatra
As the Roman crosses show, the final example is an inscription on a Chris-
tian gravestone from Savatra (Yağlıbayat) 55 kilometres east of ancient
Iconium in Lycaonia, now kept in the museum in Konya. Under the
assumption that it is a funerary inscription, the original context of the
marble block should set the limits for our interpretation.68
7. I.Konya no. 207 + Κύριος σκέπη σου ἐπ[ὶ χεῖ]-

ρα δεξίαν σου·+ Κ[ύρι]-
ο<ς> φυλάξι σε ἀπὸ παν#τ[ὸς]
κακοῦ·+ φυλάξι τὴν ψυ-
χήν σου ὁ Κύριος·+ Κύ-
ριος φυλάξαι τὴν εἴσο-
δόν σου καὶ τὴν ἔξοδόν
σου ἀπὸ τοῦ νῦν καὶ ἕως τοῦ
αἰῶνος +69
PsLXX 120:5–8 (5) κύριος φυλάξει σε,

κύριος σκέπη σου ἐπὶ χεῖρα δεξιάν σου·
(6) ἡμέρας ὁ ἥλιος οὐ συγκαύσει70 σε
οὐδὲ ἡ σελήνη τὴν νύκτα.
(7) κύριος φυλάξει σε ἀπὸ παντὸς κακοῦ,
φυλάξει τὴν ψυχήν σου.
(8) κύριος φυλάξει τὴν εἴσοδόν71 σου καὶ τὴν ἔξοδόν σου
ἀπὸ τοῦ νῦν καὶ ἕως τοῦ αἰῶνος.
(5) The Lord will keep you;
the Lord is your shelter at your right hand.
. . .
(7) The Lord will protect you from every evil;
he will guard your soul.
(8) The Lord will guard your entrance in and your exit,
from now on and forevermore.
“Inscriptions de Chypre,” BCH 20 [1896]: 336–363, 349–351 [no. 4, with plate XXIV], has
been corrected by Mitford, “Inscriptions,” 167–168.).
68 This is not the only occurrence of PsLXX 120:7–8. On the wall of a 5th century family
tomb in the Crimmerian Kertsch the Psalm is quoted. Cf. J. Kulakowsky, “Eine altchrist-
liche Grabkammer in Kertsch aus dem Jahre 491,” RQ 8 (1894): 49–87 and 309–327 (repr.
1991).
69 Cf. photo of the squeeze: I.Konya figure 246.
70 α’ and σ’ have πλήξει.
71 σ’ has προέλευσιν.
The first part of the stichos in Ps 120:5 is omitted on the small marble
block. The original second person is kept. The Lord is your shelter72 at
your right hand. The text is thus expressing the trust of the family mem-
bers who erected the monument to address the buried and the passersby.
Normally, when κύριος is used on Christian inscriptions, the title refers
to God and not to the resurrected Christ. The family thus put their trust
in God.
The reoccurrence of φυλάσσω proves that it was the belief in the Lord’s
protection which motivated the selection of these lines from Ps 120. The
omission of v. 6 of Ps 120 proves that the pretext was altered to fit the
new context. The following exposition proposes an interpretation which
is based upon the shortened text of the inscription. Since the deceased are
buried, it would be awkward to write on a tombstone “By day the sun shall
not burn you up, nor the moon during the night”. The focus is shifted from
the protection during day and night. On the marble block, the text from
Ps 120:7 thus directly follows v. 5b. The omission of the epsilon in line 3
(φυλάξι and not φυλάξει) shows that the engraver relied on hearing.73
The word φυλάσσειν however, does not have to have the same mean-
ing in all cases. In the syntagma φυλάσσειν τινὰ ἀπό τινος the verb means
“to protect from someone or something”.74 Φυλάξι is used in this way in
lines 3 and 4: “The Lord will protect you from every evil.” Used transi-
tively, the recurrent verb φυλάσσειν can also mean “to guard someone or
something”.75 Lines 4 and 5 of the funeral inscription express, that the
Lord will guard your ψυχή. The lexical sense of ψυχή can be glossed with
“breath” as the sign of life or the “life” itself, or as in this context with “soul”.76
In lines 4 and 5 the word designates the soul or spirit of the deceased in
opposition to the buried corpse. Lines 4 and 5 thus probably mean that
God will guard the soul of the deceased. In line 6 the inscription again
followed the spoken word, writing φυλάξαι instead of φυλάξει as in the
72 The Greek word σκέπη can, as the Hebrew ‫צל‬, designate the shade afforded by trees
(cf. BDAG, s.v.), but it rather has the meaning shelter, signifying the protective function
of the object; cf. LSJ, s.v.
73 The classical ει was often identified with the shorter simple ι (cf. Gignac, Grammar,
189–190).
74 Cf. LSJ, s.v. 1.B.
75 Cf. LSJ, s.v. 1.A.
76 Cf. LSJ, s.v.
LXX.77 Taking the unmarked meaning of εἴσοδος and ἔξοδος which are the
objects of φυλάσσειν in the transitive sense, it is best to translate “the
Lord will guard your entrance and your exit”. Those buried under the
cross are protected by the Lord from the moment they are interred until
they are resurrected. The final lines express that this protection goes even
beyond resurrection: from now on (ἀπὸ τοῦ νῦν) and until forever (ἕως τοῦ
αἰῶνος).
But is this the only context we can construct around the stone? What
if it were from a house78 or a church?79 Then those who went in and out
of the church have been put under the protection and guardianship of the
Lord. Again it is the words of the Psalm that are used to express his care,
to ensure that there is no difference between his protection inside and
outside, that it continues from the present into evermore.
4. Conclusion
Petitions for help alluding to or quoting from the Psalms were used on rock
faces,80 or gateways of working places,81 lintels of churches’ entrances,82
77 There are occasional instances of this interchange since the 2nd century (cf. Gignac,
Grammar, 260).
78 PsLXX 120:8 is quoted above the entrance of a house on the St. Nicolas island (Gemiler,
near Ölüdeniz, ancient Lycia). Cf. D. Feissel, Chroniques d’épigraphie byzantine: 1987–2004
(Centre de recherche d’histoire et civilisation de Byzance—Monographies 20; Paris: Asso-
ciation des amis du centre d’histoire et civilization de Byzance, 2006), no. 336.
79 Verses 7–8 are also found on lintels of churches in Stratonicea in Caria (cf. M. Ç.
Şahin, Die Inschriften von Stratonikeia, Vol. 2/2: Neue Inschriften und Indices [IK 22/2; Bonn:
Habelt, 1990], no. 88), in Al-Bara in Syria (cf. Kulakowsky, “Grabkammer,” 63–64), on a
stone from Cnidus (cf. W. Blümel, Die Inschriften von Knidos, Vol. 1 [IK 41/1; Bonn: Habelt,
1992], no. 242 [= SEG 12 (1955) no. 458]), on a mosaic in the narthex of a basilica in Eleuth-
erna on Crete (cf. Feissel, Chroniques, no. 138).
80 In Ambar/Anbar at Sidamaria (ETAM 15 no. 143): Κ(ύρι)ε βοήθι Ἰωύννου | καὶ Πέτρου
στράτορος | καὶ Ταρασίου. ΗΨ. . . . αὐτοῖς. Cf. also G. Laminger-Pascher, Beiträge zu den grie-
chischen Inschriften Lykaoniens (ETAM 11; Wien: Österreichische Akademie der Wissen-
schaften, 1984), no. 175.
81 At Kale Köy near Sidamaria (ETAM 15 no. 151): + κ(ύρι)ε βοήθι Θεοδ[ώ]ρου. Cf. ETAM
15 no. 207.
82 E.g. Christian inscriptions on lintels in Syria: Ḳaṣr Nawā (dated A.D. 398, citing
PsLXX 83:11 [von Oppenheim and Lucas, “Inschriften,” no. 22]), Ḳaṣr el Berūdj (one dated
A.D. 550, alluding to PsLXX 30:4a and 60:4 [no. 37], and another dated A.D. 569, citing
PsLXX 30:3 and 120:7–8 [no. 38]). Also on a loose stone at Umm Ṣahrīdje (dated A.D. 602, cit-
ing PsLXX 120:8 [no. 22]), and above a door in Ḳaṣr ibn Wardān (freely rendering PsLXX 117:20
[no. 48]). Amongst the ruins of a church in Ḳaṣr el Andarīn, PsLXX 117:20 and 121:1 are cited
(nos. 58 and 60).
in churches,83 amulets,84 bracelets,85 and funerary inscriptions,86 making

the Psalms the book from the Old and New Testament most used in Chris-
tian inscriptions. The Psalms have a protective function; they document
Christian practice to trust God for help in order to overcome danger and
fear in various locations, even at the grave yard. What is meant is well
expressed by the words of Ps 90:1–3, one of the most quoted87 Psalms—on
lintels, amulets, bracelets88 and gravestones:
(1) He that dwells in the help of the Highest,
shall sojourn under the shelter of the God of heaven.
(2) He shall say to the Lord, Thou art my helper and my refuge:
my God; I will hope in him.
(3) For he shall deliver thee from the snare of the hunters,
from every troublesome matter. (trans. Brenton)
83 Cf. PsLXX 92:5 quoted in various churches: Medeia in Trachea ( J. Robert and L. Robert,
Bulletin Épigraphique, Vol. 7 [1971–1973] [Paris: Société d’éditions “Les belles lettres”,
1974], 462, no. 427a); Selymbria (modern Silivri) on the Propontis (P. Magdalino, “Byz-
antine Churches of Selymbria,” DOP 32 [1978]: 309–318, 316); Nicaea (S. Şahin, Katalog
der antiken Inschriften des Museums von Iznik [Nikaia], Vol. 1 [nr. 1–633]: Stadtgebiet und
die nächste Umgebung der Stadt [IK 9; Bonn: Habelt, 1979], no. 49); Telemessus in Lycia
(E. Kalinka, Tituli Lyciae linguis Graeca et Latina conscripti, Vol. 2/1: Pars Lyciae occidentalis
cum Xantho oppido [Tituli Asiae Minoris (TAM) II; Vienna: Österreichische Akademie der
Wissenschaften, 1920], no. 118,2); Cilicia (SEG 37 [1987] no. 1365); Caesarea Maritima in
Palestine (CIIP II no. 1348).
84 E.g. Ps 135 in Greek on a Christian amulet form Gebel Qarâra in Egypt; cf. F. Bilabel
et al., Griechische, koptische und arabische Texte zur Religion und religiösen Literatur in
Ägyptens Spätzeit, Vol. 1: Textband (Veröffentlichungen aus den badischen Papyrus-
Sammlungen 5/1; Heidelberg: Verlag der Universitätsbibliothek, 1934), 364–366, no. 127.
85 E.g. PsLXX 90:1–3 on bracelets from Cyprus, Syria, Sidon, and Jerusalem; cf. Feissel,
“Notes,” 575–576.
86 Cf. MAMA VII nos. 473 (Bulduk), 567 (Ceşmelisebil), both in the region of Gdanmaa,
and no. 517 (Sülüklü/Vetissos); W. A. Ramsay, “Laodiceia Combusta and Sinethandos,”
Athenische Mitteilungen 13 (1888): 231–272, 250, no. 47 (Kadınhanı in the region of Laodi-
cea Combusta).
87 Cf. the allusion to PsLXX 90:1–2 on a marble slate from Prusias ad Hypium (W. Ameling,
Die Inschriften von Prusias ad Hypium [IK 27; Bonn: Habelt, 1985], no. 119).
88 Cf. M. Piccirillo, “Un braccialetto Christiano della regione de Betlem,” SBFLA 29
(1979): 244–252 (plates XXV–XXVII).
The Testament of Job as an Adaptation of LXX Job1
Jessie Rogers
The book of Job has generated an impressive volume and range of retell-
ings across a wide variety of media and genres.2 This is indicative of the
diverse reading strategies, the precommitments and interpretative choices
that readers bring to this complex text. In this article, I am concerned with
one particular instance of the reception of the book of Job, the pseude-
pigraphic Testament of Job, written somewhere in the period 1st century
B.C.E. to 1st century C.E. Spittler describes it as a “folkloristic elaboration
of the biblical story of Job that may be compared in its method of treat-
ment with the elaboration of incidents in the lives of the patriarchs by the
author of the Testaments of the Twelve Patriarchs.”3 Job, on his deathbed,
recounts incidents in his life to his assembled children, using the tale as
the basis for admonishing and encouraging them to emulate his example
of steadfast endurance (ὑπομονή). He gives his children their inheritance
and then dies and is lamented.
There are four extant copies of the Greek text representing three tex-
tual traditions (P, S and V), and two ancient versions, Old Slavonic and
Coptic.4 The last is the oldest, dating from the 5th century C.E., but it is
incomplete. I have accessed the Greek texts through the Online Critical
Pseudepigrapha.5 This has Brock’s edition6 as the default text and pres-
ents all the surviving Greek MSS as well as the eclectic editions compiled
1 This paper was presented at the International Conference on the Septuagint: Text-
Critical and Hermeneutical Studies in the Septuagint (LXXSA) in Stellenbosch, South Africa
on 31 August 2011.
2 C. L. Seow, “Reflections on the History of Consequences: The Case of Job,” in Method
Matters: Essays on the Interpretation of the Hebrew Bible in Honor of David L. Petersen (eds.
J. M. LeMon and K. H. Richards; Atlanta: Society of Biblical Literature, 2009), 561–586,
gives a good, albeit selective, overview of the reception of the book of Job in various
traditions.
3 R. P. Spittler, “The Testament of Job: A History of Research and Interpretation,” in
Studies on the Testament of Job (eds. M. A. Knibb and P. W. van der Horst; Cambridge:
Cambridge University Press, 1989), 7.
4 These are described by Spittler, ibid., 8–10.
5 Available at http://www.purl.org/net/ocp.
6 S. P. Brock, “Testamentum Iobi,” in Testamentum Iobi, Apocalypsis Baruchi Graece
(PVTG 2; ed. idem; Leiden: Brill, 1967), 19–59. This edition offers a corrected transcription
of P and gives the variants of V and S in the critical apparatus.
396 jessie rogers
by Brock and Kraft,7 following the versification system of Brock which

this paper also uses. The differences between the Greek texts are numer-
ous, but usually of minor relevance and they do not substantially affect
the story line. In this paper, I use P unless otherwise stated. All English
quotations from the Testament of Job are from Spittler’s translation in the
Doubleday Pseudepigrapha.8
The literary category of “adaptation” is very broad and non specific: the
possibilities for retelling a story are endless. An adaptation is not a trans-
lation, summary or amplification of the parent text, but a distinct literary
creation. Testament of Job is not a commentary on the biblical story of Job,
but a replacement story. The blurb on the paperback reissue of Studies on
the Testament of Job9 claims that “the Testament adopts from the biblical
story scarcely more than the framework.” Seow goes even further, ques-
tioning whether Testament of Job is even a reception of the biblical story
and suggesting it may be based on an alternative version.10 I disagree with
these assessments. I hope to show that the relationship between the two
books is more than simply shared characters and a vaguely similar plot. It
is a retelling which can rightly be called an adaptation. The kind of adap-
tation that we find here transposes the biblical story into a different genre,
omits large sections and adds a considerable amount of new material to
produce a very different work. Nevertheless, many of the deviations can
be explained in terms of an attempt to fill in gaps and remove ambiguities
and perceived difficulties in the biblical story to bring it more in line with
conventional piety. In claiming that Testament of Job is an adaptation of
LXX Job, I will be arguing for a familiarity with the actual text of LXX Job,
and not just dependence on a generally circulated memory of the story
of Job.
Biblical Job is itself most likely an adaptation of a traditional folk tale.
We know from Ezek 14:14, 20 that the figure of Job, alongside Noah and
Daniel,11 was a paragon of righteousness familiar to ancient readers. The
7 R. A. Kraft, ed., The Testament of Job According to the SV Text (TT 5; Pseudepigrapha
Series 4; Missoula: Scholars Press, 1974).
8 R. P. Spittler, “Testament of Job,” in Old Testament Pseudepigrapha volume 1 (ed. J. H.
Charlesworth; Garden City: Doubleday, 1983), 829–868. Spittler’s translation also includes
significant variations and their translation in the footnotes.
9 M. A. Knibb and P. W. Van der Horst, eds., Studies on the Testament of Job (Cam-
bridge: Cambridge University Press 1989).
10 Seow, “Reflections on the History of Consequences”, 585.
11 Or perhaps “Danel”—the Hebrew spelling may indicate a person other than the Dan-
iel of the biblical book of the same name. The LXX has Δανιηλ.
the testament of job as an adaptation of lxx job 397
narrative framework of the book of Job, its prologue ( Job 1–2) and epi-
logue ( Job 42:7–17), give us the contours of that original tale. This narra-
tive frame is clearly delineated in the Hebrew because it is in prose, while
the rest of the book is poetry. The distinction between narrative frame
and dialogue is less marked in the Greek translation which tends to col-
lapse the parallelism of the Hebrew poetic dialogue.
At the start of the biblical story in both its Hebrew and Greek incarna-
tions, Job is both exceedingly righteous and very wealthy. The impetus
for his suffering comes from Satan’s challenge to God about whether Job’s
religion is disinterested or ultimately an investment in Job’s own well-
being. Twice God agrees to allow Satan to test Job’s faithfulness. First, he
loses his wealth and his children in a series of calamities on a single day,
and then he is struck with a terrible plague. He is understandably devas-
tated. Nevertheless, his response is conventionally pious and he rejects his
wife’s plea to curse God and die. Job’s friends arrive to comfort him. At
this point the prologue ends and the dialogue begins.
The original folktale may have moved quickly on to Job’s restoration
as a reward for his patient endurance, but in the biblical retelling the tale
has been considerably bulked up by the addition of almost forty chap-
ters of poetic dialogue. Job’s opening monologue ushers in three cycles of
speeches where Eliphaz, Bildad and Zophar speak and Job answers each
in turn. The friends argue the case of conventional wisdom and retributive
justice. Job berates them for being such miserable comforters, gives voice
to his pain and confusion and insists on his integrity. He both accuses God
and appeals to God for help. The third cycle of speeches peters out with
Bildad speaking only briefly and Zophar not at all,12 dramatically high-
lighting the ineffectiveness of their wisdom in answering Job. Then, after
a poem about wisdom in chapter 28 which is probably in the narrator’s
voice and not strictly speaking part of the dialogue, Job concludes with a
soliloquy in which he vehemently defends his innocence and challenges
God to state his case against him. A fourth friend, Elihu, then steps up to
the fray, waxing lyrical about the educative effects of suffering and the
sovereign wisdom of God. His presence in the text is something of an
enigma in that he is not mentioned in the prologue or epilogue and Job
12 There have been many attempts made to reallocate part of Job’s speech to Zophar, as
reflected in the headings in the Jerusalem Bible, for example. But there is no justification
in the Hebrew for this, nor is any such rearrangement reflected in the LXX.
398 jessie rogers
does not reply to him. His speeches do at least provide a pause in the
story between Job’s defiant challenge to God and God’s answer that comes
after Elihu has fallen silent. God’s speeches to Job are a litany of rhetorical
questions with striking images from the natural world. Twice God speaks
and twice Job recants.
In the prose epilogue, there is no mention of the Satan. Yahweh declares
his approval of Job who, unlike his friends, has spoken of him what is
right. After the friends have sacrificed and Job has prayed for them, Job is
restored to health, he regains his great wealth and has ten more children,
including three daughters even more beautiful than the first set. Job then
lives happily ever after or, to put it in biblical terms, he dies “old and full
of years” ( Job 42:17).
One key question that any interpreter must decide concerns the rela-
tionship between the poetic dialogue and the narrative frame. Is the book’s
prose/poetry divide semantically significant? Does the dialogue confirm
or destabilize the meaning of the prose tale? The Greek translators of
Hebrew Job seem to have decided that they are to be read of a piece,
because they did not preserve the marked distinction between prose and
poetry. There are a number of difficulties or “gaps” in the text. Why does
the Satan who plays such a key role in Job’s suffering disappear after the
prologue and not even get a mention in the epilogue? Why is there no
mention of Elihu in the narrative frame? When God proclaims at the end
of the text that his servant Job has spoken of him what is right, which
words of Job are being approved, the words that he speaks in response to
God’s speeches where he capitulates and withdraws his challenge against
God, or all of his words including the ranting and raving in the dialogue
with his friends and his concluding soliloquy? There are also ambiguities
and interpretative challenges of a more religious or theological nature.
What is the Satan doing in heaven, and why is God so quick to entertain
the Satan’s insinuations? What is the point of God’s speeches from the
storm, and how are they to be construed as an answer to Job? Why is Job,
the paragon of piety and disinterested faith in Yahweh, a non-Israelite?
Scholars have used some of these gaps and discontinuities to build up
a credible picture of the textual history of the biblical book of Job. But
my interest here is on the other side of the text, the world in front of the
text, where these gaps and ambiguities exist as problems and possibilities
for the reader. Many of the variations in the story as recounted in the
Testament of Job appear to be an elaborate exercise in ironing out these
difficulties.
Before tackling the Testament of Job we need to consider the Greek

translation of the biblical book of Job13 because, as I will demonstrate,
it is in its Greek form that the book was available to the author of the
Testament. Translation is itself an interpretation, a reading. This is all the
more so where the translation is not a particularly literal one. The Old
Greek text of biblical Job is one sixth shorter than the Hebrew MT, fol-
lowing it closely to begin with, but omitting more and more as the story
proceeds, so that a full 35% of Elihu’s speeches are omitted, with fewer
omissions again toward the end.14 There is not unanimity among scholars
that the MT is actually the Hebrew Vorlage lying behind the OG transla-
tion15 but, for my purposes, it is safe enough to assume that the Greek is a
relatively free translation of a text largely similar to MT. It follows exactly
the same structure, shortening some of the poetic verses by collapsing the
parallelism.
The Greek translation contains four additions, all of which also appear
in the Testament of Job in some form:
• Job’s wife is given a longer speech when she exhorts Job to “curse God
and die” (2:9);
• The friends are identified as “kings” (2:11);
• There is a reference to the certainly of Job’s resurrection after the notice
of his death at the end of the book (42:17);
• The book concludes with an appendix which refers to a “Syrian” source,
locates Job’s land of origin, and equates Job with the Jobab mentioned
13 I work with the unasterisked material. Material added by Origen to harmonise the
text with the MT is marked with an asterisk in the Hexapla. These additions were largely
drawn from the later and more literal translation from the MT attributed to Theodotion.
This Hexaplaric version has strongly influenced subsequent copies of LXX Job.
14 Orlinsky provides the following statistics: “Of the first 11 chapters practically nothing
is absent from the Greek—only about 7 stichoi; of the following 3 chapters (11–14), 4% is
lacking; of the next 7 chapters (15–21), 16% of the Hebrew text is not reproduced in the
Septuagint; of the following 10 chapters (22–31). 21% does not exist in the pre-Origenian
Septuagint; of the next 6 chapters (32–37, the Elihu Speeches), fully 35% is omitted; and
of the last five chapters (mostly the Lord’s Speeches), but 16% is absent.” H. M. Orlinsky,
“Studies in the Septuagint of the Book of Job (chapter 1),” HUCA 28 (1957): 63.
15 N. Fernández Marcos (“The Septuagint Reading of the Book of Job,” in The Book of
Job [BETL 114; ed. W. A. M. Beuken; Leuven: Peeters, 1994], 251–266) argues that the Greek
translator tended toward paraphrase in order to create a meaningful text out of an often
obscure Vorlage whereas J. Cook (“Aspects of Wisdom in the Texts of Job [Chapter 28]:
Vorlage[n] and / or Translator[s]?” OTE 5 (1992): 26–34) suggests a combination of free
translation and a slightly different source text to account for differences.
400 jessie rogers
in the genealogy of Esau in Gen 36:33 before once again listing his
three friends and the lands they ruled. This appendix is not found in
the Peshitta, Qumran Targum or Rabbinic Targum of Job.16
The LXX Job appendix provides an ingenious solution to the problem of

Job’s location outside of Israel. The Jobab in Esau’s genealogy with whom
he is equated is not only fifth from Abraham (through Isaac and Esau),
but is also the grandson of Ishmael’s daughter Basemath. He thus takes his
place firmly within biblical history as a descendant of Abraham. The Tes-
tament of Job will make this connection even stronger by naming Dinah,
Jacob’s daughter, as the mother of his second set of children.
In light of the dramatic changes that the Testament of Job will make to
the role and depiction of Satan, it is interesting to note how the Greek
translator treats this character. All thirteen instances where the Satan is
named in the Hebrew are rendered by the term ὁ διάβολος. The name is
not transliterated as ὁ Σατανάς, a term which appears in the Testaments
of the Twelve Patriarchs. This is consistent with how Satan is translated in
the Septuagint of Zech 3:1–3, the one other place in the Hebrew Scriptures
where we have a heavenly being called the Satan who acts as an accuser
or counsel for the prosecution in the heavenly council. Names aside, the
Greek translation presents his role and function in a way that is entirely
faithful to the Hebrew.
The Testament of Job does not keep to the structure of the biblical text.
It begins with a deathbed scene, recounts the events retrospectively, and
is predominantly in prose, with some poetic sections inserted. Neverthe-
less, it does keep the same cast as the biblical story: Job, his wife, his chil-
dren (particularly his three daughters), his friends Eliphaz, Baldad and
Zophar, Elihu, God and the Satan. The names are the same as in LXX
Job, although only Eliphaz’s place of origin is noted. The one exception to
consistency in nomenclature with LXX Job is the Satan who, consistently
with the Testaments of the Twelve Patriarchs, is called Σατανάς as well as
ὁ διάβολος. The alternate designations of the main protagonist as Job or
Jobab found in the Testament has no precedent in the Hebrew text of Job,
but draws on the same tradition as is reflected in the appendix of LXX Job.
The LXX gives a coherent account of the reason for this identification,
described above, whereas the Testament simply assumes it and along with
16 A. Y. Reed, “Job as Jobab: The interpretation of Job in LXX Job 42:17b–e,” JBL 120/1
(2001): 31.
it Job’s descent from Esau (T.Job 1:6) without referring to the genealogy in
Genesis upon which it is based. Reed suggests that what was a functional
association of Job/Jobab in LXX Job becomes in Testament of Job a name
change modelled upon the Abram/Abraham paradigm.17 This suggests
that the LXX appendix precedes the use of the tradition in the Testament
of Job. The essential outline of the plot is also there: a twofold attack by
Satan on righteous and wealthy Job causes him to lose everything, includ-
ing his health. His friends come to comfort him and end up arguing with
him, and afterwards Job is fully restored. There are also significant altera-
tions: Job’s wife, daughters and the Satan are all given extended roles, the
dialogues are greatly truncated and their content is very different, and the
way in which some characters are portrayed is radically changed.
In this retelling of the story in Job’s voice, there is no third person nar-
ration of a scene in heaven nor a wager between God and Satan. Instead,
the trials that Job endures are presented as Satan’s retaliation against Job
for destroying an idolatrous temple in Job’s realm. An angelic voice fore-
warns Job and promises complete restoration if Job endures. Job’s suffer-
ing is therefore linked to his righteous actions and to his zeal for God. He
is aware of the connection and of the reason behind his suffering which
he can trace directly back to Satan. This is not simply a reinterpretation
of the Prologue in biblical Job; it is a complete rewriting. The question of
the prologue in biblical Job—“Does Job serve God for nothing?”—raised
by the Satan, around which biblical Job turns and which is faithfully trans-
lated in LXX Job, has been completely replaced. The trials are given the
much more concrete and conventional explanation of retaliation by Satan
for a direct attack from Job. The arbitrary nature of Job’s suffering and
God’s complicity in it in biblical Job is rewritten as suffering that Job will-
ingly accepts as part of his righteous opposition to Satan. Here is none of
the existential crisis of biblical Job where the suffering has to be endured
in the absence of any explanation, where there is no promised end to the
suffering, and where Job remains ignorant of Satan’s involvement. In the
words of the angel, Job “will be like a sparring athlete, both enduring pains
and winning the crown” (T.Job 4:8). When Job is promised restoration it is
“so that you may know that the Lord is impartial—rendering good things
to each one who obeys” (T.Job 4:9). This is a reassertion of the very tenets
that are challenged by the dialogue in biblical Job! The Testament of Job is
therefore about patient endurance under trial; it is not an exploration of
17 Ibid., 51.
402 jessie rogers
unexplained suffering. The effect that this has upon the tone of the whole
book is immense. Although Job suffers, there will be no cries of anguished
confusion or pressing countersuits against God. Difficult questions raised
by the Prologue are thus neatly sidestepped and the whole rationale for
telling the story changed at a stroke.
Elements of the biblical Prologue are still retained, however, in that
Satan has to ask the Lord for authority to destroy Job’s goods and family
(T.Job 8:3) and then to strike his body (T.Job 20:1–3). One difference is that
Satan goes out “under the firmament” (ὑπο τὸ στερέωμα) to make his request
(T.Job 8:2) as opposed to presenting himself to God in heaven along with
the angels. This keeps Satan out of heaven, thus “correcting” one poten-
tially problematic aspect of the Prologue. Nevertheless, although God is
not implicated by the narrative in Job’s suffering, the Testament does not
consistently use Satan as a device for distancing God from responsibility
for what happens to Job. When asked by Baldad: “Who destroyed your
goods or inflicted you with these plagues?” (T.Job 37:3), Job unequivocally
answers “God” (T.Job 37:4). Job’s response to the loss of his wealth and
children is word-for-word the same as in LXX Job 1:21b: “The Lord gave,
the Lord took away. As it seemed good to the Lord, so it has happened.
Blessed be the name of the Lord” (T.Job 19:4). This is slightly different from
the Hebrew which lacks any equivalent for the middle piece: ὡς τῷ κυρίῳ
ἔδοξεν οὕτως καὶ ἐγένετο.
The problematic absence of Satan in the Epilogue of biblical Job is a
“gap” which is thoroughly filled in by the Testament of Job. In the Testa-
ment, Satan tries several times to trick Job and he is explicitly seen to be
behind Job’s wife’s “Curse God and die” speech, standing quite literally
just behind her as she delivers it. But Satan is no match for the wisdom,
faithfulness and patient endurance of Job and eventually bursts into tears
and admits defeat, before exiting (T.Job 27:2–6). That is not quite the last
we see of Satan; he returns incognito when he possesses Elihu, as dis-
cussed later. The problematic elements of Satan’s role in biblical Job have
thus been addressed or removed. He is unambiguously evil and not por-
trayed as a member of God’s council. He does not actually enter heaven,
and is clearly shown to get his comeuppance. Job is not ignorant of Satan
and is able to outsmart him. His defeat is explicitly acknowledged, and
we know where he’s gone and why he doesn’t reappear at the end. All of
the additional story elements associated with Satan’s expanded role occur
prior to the arrival of the friends (apart from his association with Elihu),
and Satan’s concession of defeat occurs immediately after the speech of
Sitis and Job’s reply which are modelled upon LXX Job 2:9–10, as discussed
below. So the Testament of Job does still respect the order of the account
in biblical Job where Satan’s role ends prior to the dialogue.
Job’s wealth and piety are both mentioned in the Prologue of bibli-
cal Job. The question of their link is part of the dynamic of the story.
Testament of Job makes that link explicit by placing the account of Job’s
wealth—hagiographical and exaggerated to even more ridiculous
proportions—within an account of his great generosity and almsgiving,
so that the wealth of Job is expressed entirely in terms of his benevo-
lence.18 The Testament uses the list of Job’s livestock in Job 1:3 but multi-
plies the quantity of each, using the smaller total found in the biblical Job
for the subsets which were set aside for the benefit of the poor.19 So, for
instance: “I used to have 130 000 sheep, of them I designated 7000 to be
sheared for the clothing of orphans and widows, the poor and the help-
less” (T.Job 9:2–3). The Testament’s account of Job’s wealth and benevo-
lence draws upon elements from Job’s final soliloquy in biblical Job ( Job
29–31)20 where Job longs for the past and pleads his innocence. Since Job’s
innocence is not in doubt and he is not confused about what is happening
to him in the Testament of Job, the author had no need for such a speech,
but instead has gleaned elements from it to enrich his description of the
piety of Job. It is no doubt from Job 30:1 that the Testament author comes
up with the sheepdogs in the list of Job’s animals (T.Job 9:3).21 Several
details in the Testament appear to be narrativizations of elements from
this soliloquy. For example, Job’s description of past blessings—“when my
roads were flowing with butter and my hills flowed with milk!” (LXX Job
29:6 NETS)—is combined with his protestation of innocence—“And if too
my female attendants often said, ‘Who might grant us to be filled with
his flesh?’ though I was very kind” (LXX Job 31:31 NETS)—to become this
strange little chapter:
Those who milked the cows grew weary, since milk flowed in the mountains.
Butter spread over my roads, and from its abundance my herds bedded down
18 M. Wisse, Scripture between Creativity and Identity. A Hermeneutical Theory Building
upon Four Interpretations of Job (Utrecht: Ars Disputandi, 2003), 37.
19 For example, sheep to be sheared for the clothing of orphans and widows (T.Job 9:3)
and camels used to carry gifts to the needy in the cities and villages (T.Job 9:5).
20 Spittler (Testament of Job, 842 n. 9a) also notes this connection: “Typical Midrashic
embellishment characterises this section magnifying the pious generosities of Job. Job 29
and 31, both LXX (less so Job 30 LXX) clearly inform the author here, supplying numerous
details and actual language and illuminating several textual problems in TJb 9–15.”
21 How could Job disdain to put the fathers of his mockers with his sheepdogs if he
didn’t have any?
404 jessie rogers
in the rocks and mountains because of the births. So the mountains were
washed over with milk and became as congealed butter. And my servants,
who prepared the meals for the widows and the poor, grew tired and would
curse me in contempt, saying, “Who will give us some of his meat cuts to be
satisfied? Nevertheless, I was quite kind. (T.Job 13)
It is probably because of this technique of drawing from the soliloquy that
the description of Job’s benevolence in the Testament of Job carries such
overtones of resentment and opposition from some of the beneficiaries
and has them participating with Satan in Job’s ruin.
If biblical Job’s description of Job’s sacrificing on behalf of his children
is open to an interpretation of either excessive scrupulousness or bad
parenting on Job’s part, Testament of Job’s retelling expands upon and
modifies the account so that there can be no question of Job being an
indulgent parent with spoilt children prone to blasphemy. It becomes a
more straightforward demonstration of Job’s great piety. In T.Job 15, their
shared dinners are no longer parties, but a serving and eating alongside
the slaves and Job’s worry is that his sons may have thought: “We are sons
of this rich man, and these goods are ours. Why then do we also serve?”
Job’s concern, therefore, is not direct blasphemy, but that his sons may
have exhibited pride, which equates to thinking evil things in their heart
toward God. Spittler points out that “TJob follows Job 1:4 LXX in making
definitely daily banquets out of the probably annual celebrations men-
tioned in MT.”22 This once again points to the Testament’s dependence
upon the Greek translation of Job.
The greatly expanded role of women in the Testament has been
explored by a number of different commentators.23 A new talking part is
introduced to the story in the form of a servant girl. She is ordered to give
a burnt loaf to the beggar at the door, whom Job knows to be Satan, and to
say to him: “You shall no longer eat from my loaves at all, for I have been
estranged from you. Yet I have given you this loaf of bread in order that
I may not be accused of providing nothing to a begging enemy.” (T.Job
7:10). This appears to me to be an innovation that does not build upon any
element in the biblical story of Job, but it does illustrate Prov 25:21.24 The
22 Spittler, Testament of Job, 845 n. 15b.

23 M. C. Legaspi, “Job’s Wives in the Testament of Job: A Note on the Synthesis of Two
Traditions,” JBL 127/1 (2008): 71–79; S. R. Garrett, “The ‘Weaker Sex’ in the Testament of
Job,” JBL 112/1 (1993); and P. W. van der Horst, “Images of Women in the Testament of Job,”
in Studies on the Testament of Job (eds. M. A. Knibb and P. W. van der Horst; Cambridge:
Cambridge University Press, 1989), 93–116.
24 ἐὰν πεινᾷ ὁ ἐχθρός σου τρέφε αὐτόν ἐὰν διψᾷ πότιζε αὐτόν
burnt loaf also supplies the image which Satan uses to warn about what
he will do to Job’s body, providing a bridge to the following section.
Job’s wife is given a name, Sitis (which sounds like σῖτος, the word
for grain or bread), and her own story of suffering. Much of that story
is inserted prior to the “Curse God” speech. While the Hebrew text has
Job’s wife’s reaction following on immediately from Satan’s attack on Job,
the LXX inserts χρόνου δὲ πολλοῦ προβεβηκότος (LXX Job 2:9a). It is into
this space that the account of Sitis’ servitude and her deception by Satan
is inserted. Sitis goes to the market and is tricked by Satan, whom she
takes to be a bread merchant, into selling her hair for bread. This allows
Satan to gain some power over her and explains why she utters her “Curse
God and die” speech. This speech, which is interrupted by a lament for
Sitis, contains the LXX addition to Job 2:9 and 2:8b, with the exception
of the second half of v. 9d: καὶ οἰκίαν ἐξ οἰκίας προσδεχομένη τὸν ἥλιον πότε
δύσεται ἵνα ἀναπαύσωμαι τῶν μόχθων καὶ τῶν ὀδυνῶν αἵ με νῦν συνέχουσιν.
The elements of this omitted fragment—going into the houses of others
and looking to lie down to refresh herself—are woven into her story as
recounted in the Testament. After Job is afflicted with plague and worms
and relocates to the dung heap outside the city, he describes his wife
thus: “. . . I saw with my own eyes . . . my first wife carrying water into the
house of a certain nobleman as a maidservant so that she might get bread
and bring it to me” (T.Job 21:1–2). At night, when her work is done, she
returns to her husband to share her meagre food with him. Later, when
Job’s three friends arrive, she pleads with them to retrieve the bones of
her children from the fallen house, because knowing that none of them
are buried is the worst of her afflictions. After being granted a vision of
her children in heaven she says: “Now I know that I have a memorial with
the Lord. So I shall arise and return to the city and nap awhile and then
refresh myself before the duties of my servitude” (T.Job 40:4). That night,
however, she dies in peace. Although the introduction of this extended
story of the suffering of Sitis involves a substantial shift from the script of
biblical Job, several of the elements in the story are a narrativization of
the LXX addition to her speech in Job 2:9 (cf. T.Job 24:1–3) with the story
of her wanderings and servitude. The incident of selling her hair to Satan
also allows the author of the Testament to present Sitis in a good light
despite her “Curse God” speech. He does not omit it, but shows that the
deception of Satan is behind her foolishness, and allows her a (relatively)
happy death. The additional elements in her speech have mainly to do
with the bread she shares with Job and recounting the story of Satan’s
deception. This emphasis on bread is probably a play on her name, or
406 jessie rogers
vice versa. Job’s reply to his wife includes the words εἰ τὰ ἀγαθὰ ἐδεξάμεθα
ἐκ χειρὸς κυρίου τὰ κακὰ οὐχ ὑπομένομεν which is word-for-word the same
as LXX Job 2:10b, with the exception of the final verb where the synonym
ὑπομένω replaces ὑποφέρω.
While the overlap between LXX Job 2:9 and the speech and story
of Sitis is clear, there has been some debate about the direction of the
dependence. Delcor25 in 1968 concluded that the addition in LXX Job 2:9
derives from T.Job, a conclusion which Schaller26 firmly rejected, con-
cluding instead that LXX Job is prior to T.Job. Testament of Job is clearly
dependent upon LXX Job in a number of places, but it is theoretically
possible that the additions to LXX Job 2:7 were added later, under the
influence of the Testament of Job. However, because the speech in LXX Job
is distributed over a wider range of verses, and because the technique of
narrativization of speech or poetry has already been demonstrated to be
one of the compositional techniques employed by the author of the Testa-
ment, I think it more likely that the Testament of Job draws upon a single
speech to construct his story of Sitis, scattering the elements slightly, than
that someone has summarized elements of Sitis’ story and placed them
in an expanded speech in LXX Job. This concurs with the conclusion of
Johann Cook, reached through a different process of argumentation, that
the addition in LXX Job 2:9a–e is the work of the translator of the OG.27
Having Job’s wife die is another departure from the biblical tale, prob-
ably motivated by the desire to harmonise her story with the tradition
attested to in other early Jewish literature28 that Job’s wife is Dinah,
daughter of Jacob. Dinah is taken as a second wife in the period of restored
prosperity and is the father of his children who can now be described as “a
chosen and honoured race from the seed of Jacob” (T.Job 1:5) and are not
contaminated by association with the imperfect, albeit sympathetically
portrayed, Sitis. The impulse to bring Job within the ambit of the biblical
story, evident in the appendix added in LXX Job, is even more explicit
25 M. Delcor, “Le Testament de Job, la priere de Nabonide et les traditions targou-
miques”, in Bibel und Qumran (FS H. Bardtke; ed. S. Wagner; Berlin, Evangelische Haupt-
bibelgesellschaft), 54–74.
26 B. Schaller, “Das Testament Hiobs und die Septuaginta-Übersetzung des Buchs Hiob,”
Bib 61 (1980): 377–406.
27 J. Cook, “Are the Additions in LXX Job 2,9a–e to be deemed as the Old Greek text?”
Bib 9 (2010): 275–284.
28 E.g. Pseudo-Philo (Bib. Ant 8:8), the Targum to Job (2:9), and later rabbinic mid-
rashim (e.g. Gen. Rab. 19:12, 76:9).
here. This, incidentally, has the added spinoff of solving another gap in the
biblical text of Genesis—the fate of Dinah.
Job’s friends are designated as kings, an innovation of LXX Job which
is picked up and developed in the narrative. It follows the order of the
biblical narrative in having them come to Job only after his wife’s speech.
There is extensive overlap in wording between T.Job 28:2–4 and LXX Job
2:11–13 in the description of their arrival, although the author of the Testa-
ment is at pains to point out that their sitting in silence for seven days is
not due to patience (μακρόθυμος) but to shock (T.Job 28:5). Patient endur-
ance is the preserve of Job in this story. They are still miserable comfort-
ers in the long run, but because the whole circumstance of Job’s suffering
has been changed, the author of the Testament has largely ditched the
dialogue in biblical Job and replaced it with a different set of interactions.
This relieves him of the puzzle of trying to make sense of the circuitous
dialogue as he replaces it with a much more coherent and straightforward
conversation which reads as genuine question and answer, statement and
response. Eliphaz29 takes offence at Job’s calm confidence that he has a
throne in heaven, Baldad is concerned to test Job’s sanity, and Zophar
simply offers Job the use of their physicians, an offer which Job refuses.
In all his replies, Job is a model of patient endurance and firm confidence
in God, a very different Job from the one we meet in the poetic section
of biblical Job.
The treatment of the figure of Elihu is a particularly interesting example
of “filling in the gaps.” In biblical Job, Elihu is not mentioned in either the
Prologue or the Epilogue; he disappears as abruptly as he appears. In the
Testament, while it is not clear exactly when he makes his appearance,30
there is no reference to the number of kings when they are introduced
and T.Job 42:3 mentions four kings, so he is probably included with the
others when they arrive. He is somewhat surprisingly demonized by the
author, who has Satan entering him so that he says all sorts of things too
terrible for the author to repeat.31 He is then destroyed by God, explaining
his absence at the end of the story. This somewhat unexpected reinterpre-
tation of the character of Elihu makes sense in the light of the author’s
strategy for rereading God’s speeches in biblical Job.
29 Or perhaps Elihu. P, S and V read Ἐλιούς in T.Job 32:1. In T.Job 33:1, P has Ἐλιοῦ, S has
Ἐλιούς and V has Ἐλιφαζ. Spittler emends this to “Eliphaz” in these chapters.
30 See n. 29 and the uncertainty over when he begins to speak.
31 The text simply refers the reader to the “records left/miscellanies of Elihu.”
408 jessie rogers
It appears that the Testament is drawing upon the opening lines of God’s
speech in biblical Job: “Who is this that darkens counsel with words with-
out knowledge?” ( Job 38:2). Since this would be an utterly inappropriate
thing for God to say to this pious, patient incarnation of Job, the words are
taken to refer to Elihu, hence the “demonizing” strategy. Since these words
in biblical Job come immediately after Elihu has fallen silent, one can see
how this strategy suggested itself. According to the Testament of Job, “the
Lord—having appeared plainly to me through a hurricane and clouds—
spoke and censured Elihu, showing me that the one who spoke in him
was not a human but a beast” (T.Job 42:2). This hunch is strengthened by
the fact that God’s twice repeated exhortation to Job at the beginning of
each of God’s speeches in the biblical text: “Gird up your loins like a man”
( Job 38:3, 40:7) is also retained by the Testament author (T.Job 47:5). The
words are still addressed to Job, but their meaning is completely changed
from being a challenge to Job to an invitation to take up magical girdles
the wearing of which completely heals Job. These girdles are to become
the inheritances that his daughters receive from him (T.Job 46–50). Apart
from this, God’s speeches still emanate from the storm but are referred to
rather than quoted (T.Job 42:1–3; 47:9).
This has been a representative rather than a comprehensive treatment of

the Testament of Job as an adaptation of LXX Job. I have indicated how
the Prologue is taken up in the Testament. The same could be done for the
Epilogue. There are also additional links between the soliloquy in LXX Job
29–31 and the description of Job’s wealth and benevolence to be explored.
The examples I have given are sufficient, however, to counter claims that
the Testament of Job is not a reception of the biblical Job or that it adopts
scarcely more than its framework. The Testament of Job takes up the
actual text of LXX Job in places and it respects the order of events in its
precursor. All the same, it is a very free re-reading. It takes great liberties
with the story, adding material (which may have been created de novo or
drawn from other sources) and changing the sense or recipients of words
in order to present a picture of Job, God and Satan more in line with
conventional piety. It uses the techniques of elaboration, summary, con-
flation and narrativisation of speech and poetry in constructing elements
of the story. The result is a very different tale, but one that can still be
described as an adaptation of LXX Job.
On Kingship in Philo and the Wisdom of Solomon1
Jonathan More
1. Introduction
The Wisdom of Solomon, also known as the Book of Wisdom, is addressed

to an assorted group of kings, rulers, judges and princes. The author of the
work claims to be a king—more specifically, the king of God’s people.2
The absence of named characters makes it difficult for the uninitiated
to know who this people and their king are, yet those who recognise the
literary allusions can identify him as Israel’s King Solomon.3
Numerous texts purportedly written to, and for, kings were produced in
the Graeco-Roman period, and the Wisdom of Solomon would not have
been unusual in this regard. Tracts entitled “περὶ βασιλείας” occur over
a broad chronological range and were produced by various philosophi-
cal schools.4 Not one of these περὶ βασιλείας volumes has been fully pre-
served. However, Stobaeus’ anthology contains a small group of fragments
identified as coming from Pythagorean περὶ βασιλείας tracts attributed to
Ecphantus, Diotogenes and Sthenidas.5 In an influential essay published
1 On the occasion of his seventieth birthday, I would like to dedicate this essay to
Dr Paul Bowers, teacher and friend, who instructed me in the importance of these Jewish
texts.
2 Wis 7:1–6; 8:10–15; 9:7, 12.
3 The writer identifies himself as a king who sought wisdom from God (7:7–12; cf. 1 Kgs
3:4–15; 4:29–34) and was responsible for the building of the Jewish temple (9:8; cf. 1 Kgs
7:13–8:21). The allusions in chapter 10 to characters and events from Genesis and Exodus
confirm that the author is drawing on the Hebrew scriptures.
4 In his work On the Lives of Eminent Philosophers, Diogenes Laertius lists volumes
entitled περὶ βασιλείας by Euphantus (2.110), Aristotle (5.22), Theophrastus (5.42; 5.49),
Strato (5.59), Persaeus (7.36), Cleanthes (7.175), Sphaerus (7.178) and Epicurus (10.28).
Other works with similar titles were written by Xenocrates (Στοιχεῖα πρὸς Ἀλέξανδρον περὶ
βασιλείας; 4.14), Theophrastus (Περὶ παιδείας βασιλέως; 5.42; Πρὸς Κάσανδρον περὶ βασιλείας;
5.47), Antisthenes (Κῦρος ἢ περὶ βασιλείας; 6.16; Ἀρχέλαος ἢ περὶ βασιλείας; 6.18). Plato’s
Statesman is given the second title of περὶ βασιλείας by Thrasylus (3.58). There are other
texts which might draw from a similar conceptual pool, e.g., Theophrastus On Tyranny
(5.45) and Xenophon’s Hieron or Of Tyranny (2.57).
5 A lack of consensus regarding the dating of these fragments (see n. 9) makes it difficult
to know how to refer to them. “Hellenistic Pythagorean” is most appropriate since, even
if they come from a later period, they reflect Hellenistic thought and deliberately echo a
certain Pythagorean tradition. For the sake of brevity, though, I use the term “Pythagorean”
410 jonathan more
in 1928,6 E. R. Goodenough presented a translation and discussion of these

fragments and concluded his article with the following claim about the
Pythagorean texts:
Even though so much has had to be omitted from the present discussion,
this philosophy of royalty will, I think, prove to have been the official politi-
cal philosophy of the Hellenistic age.7
Even if “the official political philosophy of the Hellenistic age” is somewhat
more elusive than Goodenough suggests,8 these fragments provide a use-
ful glimpse of the philosophical discussion of kingship in certain circles
during the Hellenistic period.9
In his study of the Hellenistic influences on the Wisdom of Solomon,
James M. Reese observes that Pseudo-Solomon draws on the same tra-
ditions as these Pythagorean kingship treatises in order to construct
elements of his anthropology and to enrich his description of God.10
The influence of the Pythagorean tracts on Philo has also been noted by
to refer to these tracts without prejudicing any conclusion regarding their date or the exact
nature of the philosophical tradition represented by them.
6 E. R. Goodenough, “The Political Philosophy of Hellenistic Kingship,” Yale Classical
Studies 1 (1928): 53–102. The article also includes a discussion of a fragment from Archytas
entitled περὶ νόμου καὶ δικαιοσύνης which is similar to the other fragments but does not
claim to make a direct contribution to the περὶ βασιλείας topos.
7 Ibid., 102.
8 See, for example, Aalders’ description of the varieties of Hellenistic political thought
which begins with the observation that “Hellenistic political thought is by no means a unity.
It consists of a number of strongly diverging, but also interacting tendencies and currents
of thought” (G. J. D. Aalders H. Wzn., Political Thought in Hellenistic Times [Amsterdam:
Adolf M. Hakkert, 1975], 1).
9 The Pythagorean texts have been dated over a span of six centuries from the Hel-
lenistic period (H. Thesleff, An Introduction to the Pythagorean Writings of the Hellenistic
Period. [Acta Academiae Aboensis. Humaniora. Vol. 24 no. 3; Åbo: Åbo Akademi, 1961],
72) to the third century c.e. (W. Burkert, “Zur geistesgeschichtlichen Einordnung einiger
Pseudopythagorica,” in Pseudepigrapha I [ed. K. von Fritz; Entretiens sur l’Antiquité Clas-
sique XVIII; Vandœuvres / Genève: Fondation Hardt, 1972], 25–55). The discussion has
recently been summarised by F. Calabi (God’s Acting, Man’s Acting: Tradition and Philoso-
phy in Philo of Alexandria [trans. H. C. Tooke; Studies in Philo of Alexandria 4; Leiden /
Boston: Brill, 2008], 185–188) and so will not be rehearsed here. The nature of the works,
whether considered as textbooks (Thesleff, Introduction, 72) or as summaries of the doc-
trines of various schools (O. Murray, Peri Basileias: Studies in the Justification of Monarchic
Power in the Hellenistic Period [D.Phil. diss., Oxford, 1971]) suggests their utility for recon-
structing the content of aspects of Hellenistic kingship theory regardless of their date of
composition.
10 J. M. Reese, Hellenistic Influences on the Book of Wisdom and Its Consequences (AnBib
41; Rome: Biblical Institute Press, 1970), 10, 71–79.
on kingship in philo and the wisdom of solomon 411
various scholars.11 In this essay, I will bring together these three groups of
texts by comparing Philo’s construction of Moses as Israel’s ideal king and
the portrayal of Solomon that is found in the Book of Wisdom with the
Pythagorean περὶ βασιλείας fragments.
Given the enormity of the Philonic corpus and the relative brevity of
this essay, I will focus most of my attention on Philo’s work On the Life
of Moses. Philo indicates explicitly that the first half of this two-volume
work is concerned with Moses’ kingship (Moses 1.334), while the climax
of this volume, the Sinai epiphany, treats Moses’ appointment as king
(1.148–162).
Philo’s text differs from the Wisdom of Solomon in important ways, not
the least of which is genre. In the Life of Moses, Philo engages in a retelling
of the Exodus story in order to make Moses’ story known in Greek circles.12
The Wisdom of Solomon also contains a retelling of the Exodus story,13
but the work as a whole is generally classified as “protreptic” or “didactic
exhortation,”14 and claims to be addressed to rulers by the pseudonymous
author. The primary concern of this essay, however, is not with questions
of genre, but with the portraits of kingship found within these texts.
11 E. Brehier recognised the similarities between the Pythagorean kingship frag-
ments and Philo’s representation of Moses and the ideal king (Les Idées Philosophiques
et Religieuses de Philon d’Alexandrie [Paris: Libraire Alphonse Picard & Fils, 1908], 18–23).
Goodenough’s discussion of Philo’s presentation of “Moses for Gentiles” (By Light, Light:
The Mystic Gospel of Hellenistic Judaism [Amsterdam: Philo Press, 1969 (repr. of 1935)],
181–98) builds on these similarities, as does his examination of Philo’s political thought
(The Politics of Philo Judaeus: Practice and Theory. With a General Bibliography of Philo by
Howard L. Goodhart and Erwin R. Goodenough [Hildesheim: Georg Olms Verlagsbuchhan-
dlung, 1967 (repr. of 1938)], 44–63, 90–100). G. F. Chesnut’s discussion of the ruler as the
embodied law or reason of God (“The Ruler and the Logos in Neopythagorean, Middle
Platonic, and Late Stoic Political Philosophy,” ANRW II 16.2 [1978], 1310–1332) includes
Philo and the Pythagorean περὶ βασιλείας texts. F. Calabi’s recent work on Philo’s theology
contains an appendix comparing certain elements in Philo with Ecphantus’ περὶ βασιλείας
(Tradition and Philosophy, 185–215).
12 Moses 1.1–4.
13 See, e.g., S. Cheon, The Exodus Story in the Wisdom of Solomon: A Study in Biblical
Interpretation (JSPSup 23; Sheffield: Sheffield Academic Press, 1997).
14 So Reese, Hellenistic Influences, 90–121; D. Winston, The Wisdom of Solomon: A New
Translation with Introduction and Commentary (AB 43; New York / London: Doubleday,
1979), 18–20; and J. J. Collins, Jewish Wisdom in the Hellenistic Age (OTL; Louisville: West-
minster John Knox, 1997), 179–182. For arguments against this classification, see M. Gilbert,
“Wisdom Literature,” in Jewish Writings of the Second Temple Period: Apocrypha, Pseude-
pigrapha, Qumran Sectarian Writings, Philo, Josephus (ed. M. E. Stone; Compendia Rerum
Iudaicarum Ad Novum Testamentum. Section Two: The Literature of the Jewish People
in the Period of the Second Temple and the Talmud, Vol. 2.; Assen / Philadelphia: Van
Gorcum / Fortress Press, 1984), 306–309.
412 jonathan more
Both Jewish texts were written in Alexandria at approximately the

same time,15 and the possibility of a Pythagorean revival in Alexandria16
suggests a further link between the Jewish texts and the Pythagorean frag-
ments, although the comparative method being employed in this essay
does not require geographical or chronological proximity.17
This three-way comparison will highlight the similarities and differ-
ences between these writings and thus enable us to see more clearly how
the Jewish Alexandrian authors have adopted and adapted the Hellenistic
περὶ βασιλείας topos.
2. God as King
In certain Stoic writings, the cosmos is compared to a well-ordered city

or society in which the gods are rulers and human beings their subjects.18
This metaphor depicting God as a heavenly king provides the point
of departure in the texts being studied here. Ecphantus calls God “the
higher king” (τῶ ἀνωτέρω βασιλέως; 80.5) and the “heavenly king” (ὠρανὸν
βασιλέως; 81.10) who rules well (81.8).19 He asserts that the orderliness and
beauty of the universe consists in the fact that everything in it falls under
some sort of rule (81.2). Sthenidas can write that “nothing which lacks a
king or ruler is good” (188.10–11). The last section of Diotogenes’ text iden-
tifies Zeus as the ruler of all things (ὁ κρατέων πάντων; 75.9) who holds a
lightning bolt in his hand as symbol of his universal rule and his severity
(75.14–15). Furthermore, both Diotogenes (75.12–13) and Sthenidas (188.5)
15 Philo: J. Morris, “The Jewish Philosopher Philo,” in E. Schürer, The History of the
Jewish People in the Age of Jesus Christ (175 b.c.–a.d. 135) (ed. G. Vermes, F. Millar and
M. Goodman; 3 vols.; Edinburgh: T & T Clark, 1986), 3.2.813–816. Wisdom: Winston,
Wisdom of Solomon, 20–25.
16 See J. M. Dillon, The Middle Platonists: A Study of Platonism 80 b.c. to a.d. 220 (Classical
Life and Letters; London: Duckworth, 1977), 117–121. Conclusions regarding the provenance
of the Pythagorean texts being discussed are tied up with those pertaining to their date
of composition. See n. 9.
17 R. A. Segal, “In Defense of the Comparative Method,” Numen 48/3 (2001): 339–373;
J. Z. Smith, Drudgery Divine: On the Comparison of Early Christianities and the Religions
of Late Antiquity (Jordan Lectures in Comparative Religion 1988; Chicago: University of
Chicago Press, 1994), 36–53.
18 SVF §§ 1.98, 2.528; see also A. A. Long and D. N. Sedley, The Hellenistic Philosophers:
Translations of the Principal Sources, with Philosophical Commentary (Cambridge: Cam-
bridge University Press, 1987), 434.
19 The Pythagorean texts are taken from H. Thesleff’s edition (The Pythagorean Texts of
the Hellenistic Period. [Acta Academiae Aboensis, 30, 1; Åbo: Åbo Akademi, 1965]) and the
English translations from Goodenough, “Hellenistic Kingship”.
call God “the father of gods and men” and Sthenidas also identifies God
as the creator and lawgiver (188.7, 9). These metaphors contribute to the
picture of God as the benevolent ruler of the cosmos.
God’s sovereignty over the universe is axiomatic in the Wisdom of Solo-
mon. Although God is never called βασιλεύς, the author declares that God
will reign (βασιλεύσει) over the righteous who, in turn, judge nations and
rule over peoples (3:8). God manages (διέπω)20 all things rightly (12:15a)
and has “sovereignty over all” (12:16b). God is sovereign, yet judges with
mercy (ἐπιείκεια) and governs with forbearance (12:18). God is kind and
true and longsuffering, ordering all things with mercy (15:1). God acts as
a “pilot” in the universe (14:3), a metaphor used to express God’s provi-
dence, an important element in the author’s conception of God. (14:3;
17:2; 6:7). God is also portrayed as benefactor of the righteous (εὐεργετέω
3:5; 11:5, 13; 16:2; εὐεργεσία 16:11, 24; Wisdom is described as εὐεργετικός in
7:22).21 Like a good king, God punishes the wicked and rewards the righ-
teous.22 Pseudo-Solomon thus portrays God using the language of Hel-
lenistic kingship. Elements of his theological vocabulary and imagery can
also be found in the Pythagorean kingship treatises.23
Philo’s biography of Moses follows the biblical story more closely than
most of his other treatises and contains very little allegorical exposition.
There is relatively little explicit talk about God in this treatise, yet God is
portrayed as the cosmic king throughout Philo’s writings. Although other
images are perhaps more important in Philo’s description of God,24 he fre-
quently uses royal language to speak of God and draws on imagery from
the Persian monarchy to describe God as the Great King and the King of
Kings.25 The first commandment in the Decalogue summarises “the laws
20 Ecphantus uses διέπω to refer to divine (81.16) and human (80.23; 81.1) rule.
21 See Reese, Hellenistic Influences, 78 n. 204. For the pilot metaphor used slightly differ-
ently, see Diotogenes 72.6; for the king as benefactor, see Diotogenes 72.14. The language
of benefaction is found only infrequently in the LXX: Pss 12:6; 56:2; 114:7; 2 Macc 10:38
(εὐεργετέω); Ps 77:11 (εὐεργεσία) but in these cases speaks of God.
22 This conviction forms the heart of the narrative in chapters 1–5.
23 Reese, Hellenistic Influences, 6–7, 10–11, 71–79.
24 “Father” and “Creator,” for example, occur commonly, either alone or in combina-
tion (e.g., Moses 1.158, 272; 2.24, 48, 67, 134, 238). These images are combined when God
is called ὁ γεννητής (Moses 2.205). In Moses 2.88, God is called “Father and Ruler (ἡγεμών)
of All.” I have used the Loeb edition for both the Greek text and English translation of
Philo’s works.
25 “Great King”: Decalogue 61; Agriculture 78; QE 2.44; “King of Kings”: Spec. Laws 1.18;
Cherubim 99. See D. T. Runia, Philo of Alexandria, On the Creation of the Cosmos Accord-
ing to Moses: Introduction, Translation and Commentary (Philo of Alexandria Commentary
Series 1; Leiden: Brill, 2001), 232; Goodenough, Light, 38–41.
414 jonathan more
on God’s monarchical rule” through which God is declared to be, not only
the “First Cause of the World,” but also ἡγεμών and βασιλεύς, ἡνιοχῶν and
κυβερνῶν.26 The royal virtues such as mercy and kindness are ascribed to
God (Moses 1.198), and God is portrayed as a benefactor (Moses 1.199; 2.41,
256, 259).27
In their portrayal of God, the authors of these Jewish texts draw on
language reminiscent of the Pythagorean fragments’ descriptions of both
God and king. God is portrayed as a heavenly king who rules over the cos-
mos while exhibiting virtues that are also identified with the ideal human
king. In isolation, these individual verbal and conceptual items might be
considered commonplace and unremarkable. However, the conceptual
matrix formed by the combination of these items is unmistakable and the
presence of this matrix in these writings suggests that a similar tradition
of heavenly and earthly monarchy influenced them.
3. The King’s Imitation of God
God’s monarchical rule over the cosmos leads to an argument for kingship
based on “naturalness”: in the same way that God rules over the universe,
the sun can be seen to rule over the stars, and different groups of animals
all seem to obey one leader.28 The Pythagoreans observe that “nothing
unruled can be found” (Ecphantus 81.2) and that “nothing which lacks a
king or ruler is good” (Sthenidas 188.10–11). It follows that if human beings
should attempt to conform their lives to nature,29 then monarchy is the
most desirable form of rule since it is that form of rule which best con-
forms to the nature of the cosmos.30
26 Decalogue 155; cf. 12; Spec. Laws 1.12. See also Decalogue 53 where, in addition to
being γεννητής, God is also called στρατάρχης, κυβερνήτης and ἄρχων τῆς μεγαλοπόλεως.
This last description is similar to the Stoic analogy in which God relates to the cosmos as
a ruler to a city.
27 Moses is never called “benefactor” in this biography, but he is praised for being con-
cerned with his subjects (1.151).
28 O. Murray, “Philosophy and Monarchy in the Hellenistic World,” in Jewish Perspec-
tives on Hellenistic Rulers (eds. T. Rajak et al.; Hellenistic Culture and Society 50; Berkeley /
Los Angeles / London: University of California Press, 2007), 23–24. See, e.g., Xenophon,
Cyr. 5.1.24–25; Seneca, Clem. 1.19.2.
29 See, e.g., SVF 3.16; Diogenes Laertius, Lives 7.87–89. The various ways in which a
life κατὰ φύσιν was conceived are discussed in Long and Sedley, Hellenistic Philosophers,
400–401, 364–366.
30 Murray observes that this preference for monarchy can be traced back to Homer:
“The rule of many is not good; let there be one ruler, one king” (Iliad 2.204–205) (“Philoso-
phy and Monarchy,” 24).
The Pythagorean texts also assert that the king stands in the same rela-
tionship to the polis as God does to the cosmos (Diotogenes 72.19–20; cf.
Ecphantus 81.26–82.3). One of the most important characteristics of the
king is established through this analogical reasoning: the king must rule
on earth as God rules in the heavens, or, stated differently, the king must
imitate God. God is king “by nature” but the earthly king is king by imi-
tation (Sthenidas 187.12–13),31 while “royalty is an imitation of divinity”
(Diotogenes 75.15–16). Diotogenes encourages the king to adopt the royal
virtues exhibited by the gods (75.8–9), including majesty and gracious-
ness, with justice, mercy and kindness being subsumed under the latter
(74.20–21).
Perhaps the most striking example of the imitation topos is found in
Ecphantus. Since God used himself as the archetype when creating the
king, the earthly king is a copy (τύπος) of the heavenly king (80.4–5). Simi-
lar ideas are expressed in Philo’s writings.
In his discussion of the conduct of rulers, Philo states that the king’s
sovereignty is “formed in the image of its archetype the kingship (βασιλεία)
of God” (Spec. Laws 4.164). It might be objected that it is kingship which
is formed according to the archetype of the kingship and God, and not
the king himself. However, later in the same treatise Philo interprets the
Mosaic law as teaching that if good rulers wish to experience assimilation
(ἐξομοίωσις) to God, they must imitate (μιμέομαι) God in seeking the good
of their subjects (Spec. Laws 4.186–188).32 It is therefore possible to read
the first statement as pertaining to the king himself and not simply to his
reign.
31 Sthenidas’ asserts that the king is ruler “by birth and imitation” (γενέσει καὶ μιμάσει).
The addition of “by birth” suggests that the author might have had a dynastic structure in
mind. In contrast, Philo’s Moses is praised for refusing to allow his sons to succeed him
(Moses 1.150).
32 Calabi, Tradition and Philosophy, 190–193 notes the similar emphasis that both
Ecphantus and Philo place on the goodness of God.
The concept of assimilation can be traced back to Plato (e.g., Theaetetus 176b; Timaeus
90d; Laws 716c), see D. N. Sedley, “The Ideal of Godlikeness,” in Plato 2: Ethics, Politics,
Religion, and the Soul (ed. G. Fine; Oxford Readings in Philosophy; Oxford: Oxford Uni-
versity Press, 1999), 309–28. J. M. Armstrong (“After the Ascent: Plato on Becoming Like
God,” Oxford Studies in Ancient Philosophy 26 [2004]: 171–83) argues that this theme in
Plato’s writing does not always entail retreat from the sensible world but can include a
commitment to improve it, thus confirming the link made in Philo and the Pythagorean
writings between the king’s positive actions and his assimilation to God. Calabi (Tradition
and Philosophy, 191–92) goes on to point out that assimilation to God/imitation of God is a
common topos in writings about kingship (cf. LA 188–294; Seneca, Clem. 1.7.1; 1.19.8–9).
416 jonathan more
Moses proves to be exemplary in these matters. He always seeks to

benefit those over whom he exercises authority (Moses 1.151), thus show-
ing himself to be a careful imitator of God. Not surprisingly, he is given
the title of “god and king” because of his exceedingly great virtues (Moses
1.154–158), many of which are focused on the well-being of his subjects.33
In this same section, Philo draws on Plato’s concept of the philosopher-
king when explaining Moses’ ascent to God in Exod 20:21.34 Like Plato’s
philosopher-king who must necessarily experience a vision of the good
in order to rule well (Resp. 504d–505b), Moses, having led the Israelites
out of Egypt, sees the “archetypal essence of existing things” (Moses 1.158).
Regardless of whether Moses’ vision here is of the intelligible world or of
τὸ ὄν,35 upon his descent from the darkness he has indeed become “beau-
tiful and godlike” (1.158). His assimilation to, and imitation of, God has
reached its zenith. The nature of this change will be discussed in the fol-
lowing section.
Nothing explicit is said in the Wisdom of Solomon about the king’s imi-
tation of God or God’s sovereignty. Both God and Wisdom are described
in language that is also used of the king, leaving the reader in no doubt
as to the similarity between the earthly and heavenly rulers. There is no
indication, however, that the earthly king’s sovereignty comes through
imitation of the divine king. Instead, sovereign rule comes through rela-
tionship with Wisdom (Wis 6:21; 8:10–18),36 which, in turn, is a gift from
33 Philo tells us that God appointed Moses king over the Israelite refugees because of
his virtues. These are displayed in different ways, including Moses’ willingness to leave
the high position he enjoyed in Egypt (1.149), his desire to benefit his subjects (1.151, remi-
niscent of Aristotle’s second criterion in his six different types of government [Politics
1279a16]) and his indifference to wealth (1.152–153, 155). A list of sixteen virtues illustrates
Moses’ superior nature (1.154).
34 When Moses 2.2 is compared with Resp. 473d it becomes clear that Philo has one eye
on Plato when producing his portrait of Moses.
35 The former is preferred by D. T. Runia, “God and Man in Philo of Alexandria,” JTS
NS 39/1 (1988): 54 n. 26; cf. Posterity 14, Names 7. More recently, S. D. Mackie has argued
for the latter (“Seeing God in Philo of Alexandria: The Logos, the Powers, or the Existent
One?” Studia Philonica Annual 21 [2009]: 25–47).
36 D. Winston sees a similarity between Timaeus 90a and the kinship with Wisdom
that is expressed in Wis 8:17 (Wisdom of Solomon, 197–198; idem, “Wisdom in the Wis-
dom of Solomon,” in In Search of Wisdom: Essays in Memory of John G. Gammie [ed. L. G.
Perdue, B. B. Scott and W. J. Wiseman; Louisville: Westminster John Knox Press, 1993],
149–164; here 154–155). But the emphasis in this passage is on relationship and not being.
This becomes clear in the second half of the verse when the idea is developed in terms of
friendship with Wisdom.
God (Wis 7:7; 8:21; 9:10, 17). The absence of imitation language is striking,37
especially when the allusion to Gen 1 in Wis 9:1–3 would have allowed the
author to discuss this topos in terms of humanity’s creation in the image
of God.38
There is a fundamental difference in the way Moses and Solomon are
construed in these texts. Philo’s description of Moses follows the Platonic-
Pythagorean model in which the king’s sovereignty is an imitation of God’s
cosmic reign. Pseudo-Solomon describes kingship as arising through kin-
ship with Wisdom which comes as a gift from God. This difference is
explained when the nature of the king is examined more carefully.
4. Nature of the King
In the Pythagorean texts, we find that Ecphantus’ king is superior and

unique among the rest of humanity because he was created as the type of
which God is the archetype. The king was created out of the same material
as the rest of humanity, but embodies the best of that which is common
to humanity (80.1–4). Similarly, Diotogenes’ king is the man who exhibits
superlative virtues and leadership skills. He will distinguish himself from
the rest of humanity and become like the gods through his superior vir-
tue (74.10–12). Just as God is the most superior being in nature, so the
king is “the best man” (ὁ ἄριστος) amongst human beings (72.16–19). It
would seem that here the Pythagorean texts elevate the king to the high-
est position within the earthly realm, but stop short of ascribing divinity
to him.39
37 M. McGlynn (“Solomon, Wisdom and the Philosopher-Kings,” in Studies in the Book
of Wisdom [eds. G. G. Xeravits and J. Zsengellér; JSJSup 142; Leiden / Boston: Brill, 2010],
61–81; here 72–77) describes a number of parallels between Plato’s Republic and Wisdom,
arguing that the former “was a formative book” for the latter (72). This relationship makes
the absence of the Platonic imitation topos all the more conspicuous.
38 Wis 2:23 states that God created humanity for immortality “and made them the
image of his own nature” (NETS). It does not seem that this is considered a particularly
royal trait in the Book of Wisdom and this idea cannot be considered to contribute to the
kingly imitation-of-God topos.
39 Similarly, Aristotle’s Politics (1284a3–14) describes one whose virtue excels that of all
those around him and who is thus a “God among men.” There is no law for such a person
who is himself a law. Such a person could never be ruled by someone else, for “that would
be as if mankind should claim to rule over Zeus” (1284b30–31). Plato’s ruler comes from
among his brothers (Resp. 415a), yet he is gold while the rest of the city might be silver or
bronze or iron.
418 jonathan more
Other passages imply that the king is something more than human.
After drawing attention to the similarities between God and the king,
Diotogenes states that the king, with his absolute rulership, and since he
is νόμος ἔμψυχος, has been “metamorphosed into a deity among men” (θεὸς
ἐν ἀνθρώποις παρεσχαμάτισται; 72.22–23). Alone, this text might speak of
the king’s divinity. In the context of this comparison, however, it becomes
clear that the author is making an analogical argument. The king’s abso-
lute sovereignty (ἀρχὰν ἔχων ἀνυπεύθυνον)40 and his existence as embod-
ied law (νόμος ἔμψυχος) mean that when he is compared to other human
beings, he is like a god. This is a form of relative rather than absolute
divinity.41
The distinction between humanity and the gods was not as clear in
ancient times as it is to many in the modern world. Heroes, rulers and
other divine persons occupied a position that seems to straddle these two
categories.42 One might think that Philo’s monotheism would not have
allowed any breach of the human-divine barrier, yet even Philo, at times,
seems ambivalent on this point. Like Ecphantus’ king (81.10–12), Moses is
sent “as a loan to the earthly sphere” (Sacrifices 8–10) and, as has already
been observed, Moses receives the titles “god and king” (Moses 1.158).
Philo’s understanding of Moses’ nature has been investigated by a num-
ber of scholars and, with some exceptions,43 most agree that Philo did not
see Moses as divine in any absolute sense.44 Rather, Philo inherits this
40 In Aristotle (Pol. 1295a9–22), this sort of rule is identified as a third type of tyranny,
the opposite of the perfect monarchy. In this tyranny, the ruler is ἀνυπεύθυνος and rules
all alike, whether equals or betters, with a view to his own advantage. This phrase is not
pejorative in Diotogenes’ text.
41 Contrary to Calabi (Tradition and Philosophy, 197) and Chesnut (“Ruler and Logos,”
1317–1318) who see the king described in the Pythagorean texts as “ontologically superior”
to the rest of humanity.
42 See, e.g., J. D. Mikalson, “Greek Religion: Continuity and Change in the Hellenistic
Period,” in The Cambridge Companion to the Hellenistic World (ed. G. R. Bugh; Cambridge:
Cambridge University Press, 2006), 208–222, esp. 213–216.
43 Goodenough (Light, 223–234) argues that Philo vacillates between the monotheism
of his ancestral Judaism and Platonic/Pythagorean philosophical traditions, on the one
hand, and popular beliefs in which the good and the great were deified, on the other.
B. Mack (“Imitatio Mosis: Patterns of Cosmology and Soteriology in the Hellenistic Syna-
gogue,” Studia Philonica 1 [1972]: 27–55) places Philo’s Moses in the Hellenistic θεῖος ἀνήρ
tradition.
44 C. R. Holladay, Theios Aner in Hellenistic-Judaism: A Critique of the Use of This Cat-
egory in New Testament Christology (SBLDS 40; Missoula: Scholars Press, 1977), 103–198;
Runia, “God and Man”; L. W. Hurtado, One God, One Lord: Early Christian Devotion and
Ancient Jewish Monotheism, 2nd ed. (London / New York: T & T Clark, 1998 [repr. of 1988]),
59–63; I. W. Scott, “Is Philo’s Moses a Divine Man?” Studia Philonica Annual 14 (2002):
87–111.
language from Exod 7:1, where Moses is told by the Lord: “Look, I have
given you as a god to Pharao” (NETS). This is explained as divinity in a
relative sense:
[T]he wise man is said to be a god to the foolish man, but . . . in reality he
is not God . . . when the wise man is compared with Him that is, he will be
found to be a man of God; but when with a foolish man, he will turn out to
be one conceived of as a god, in men’s ideas and imagination, not in view
of truth and actuality. (Worse 162)
The title “god” is an expression of Moses’ human excellence as a sage and
does not imply divinity in any absolute sense. Similar limits are placed
upon Plato’s ideal ruler who has his gaze fixed on “eternal realities” and
“will himself become orderly and divine” but only “in the measure permit-
ted to man” (Resp. 500b–d).45 Moses’ divinity is linked to his identity as the
perfect sage and his portrayal as an ideal king in the Platonic mould.46
Even though Moses is not divine, it is clear that he is unique among,
and superior to, the rest of the human race. He alone is able to contem-
plate “what is hidden from the sight of mortal nature” (Moses 1.158). It is
impossible for the ordinary person to achieve what Moses has achieved,
yet it remains a good thing for humanity to strive after this goal and to
imitate him as far as that is possible (Moses 1.159). For sages take God as
their guide and teacher, but the less perfect follow the sage (Heir 19).
While a cursory reading of Philo might lead some to the conclusion that
he is ambivalent about Moses’ divinity, the Wisdom of Solomon leaves its
reader in no doubt regarding this question. Pseudo-Solomon goes to great
lengths to emphasise Solomon’s complete and unambiguous humanity:
I myself also am mortal like everyone . . . And I myself, when I was born,
drew in the common air and fell upon the kindred earth, with the same first
sound crying like everyone. . . . For no king has had a different beginning of
existence, but there is one entrance into life and the same way out. (7:1–6;
NETS)
Ruler-cult etiology is described and the practice criticised by Pseudo-
Solomon in 14:15–21 as part of his denunciation of idolatry.47 The divinity
of the ruler is not attacked directly but implicit in this critique of idolatry
is the denial of any divinity to that which is not God (Wis 14:20b).
45 Trans. P. Shorey (Loeb).

46 Holladay, Theios Aner, 108–129.
47 Winston, Wisdom of Solomon, 270–271.
420 jonathan more
Divine Wisdom dwells in Solomon as a gift from God, bringing with

it great honour (8:10–12), political success and military victory (8:14–15;
9:10–12). Solomon’s address to the kings and “judges of the ends of the
earth” suggests that the same blessings are available to those rulers who,
like Solomon, would receive Wisdom and her instruction (6:21–25).
In the Hellenistic urge to assign a god-like nature to kings, we find an
attempt to justify monarchy in recognition of the king’s superiority.48 Only
someone who is god-like can, and should be permitted to, wield this sort
of power. Philo’s writings exhibit a similar pattern of thought in attribut-
ing this exalted status to Moses in order to cast him in the mould of a
Hellenistic βασιλεύς. Not only was Moses Israel’s exemplary king, he was
also its lawgiver, high priest and prophet par excellence (Moses 2.1–7). His
occupation of these offices, which was based on his superior virtue, meant
that he was a god-like figure. Philo is able to identify Moses as such with
confidence because he finds a reference to Moses as “a god” in the biblical
text. However, Philo consistently explains this reference as indicating that
Moses is divine only in a relative sense.49
The portrait of God’s king in Wisdom is quite different. Solomon is not
a super-human sage who holds his position because of his superiority
over the rest of humanity. There is no hint of any god-like qualities being
ascribed to Solomon. Instead, his unexceptional humanity is stressed and
any good that he exhibits or has achieved can all be traced back to God’s
gift of Wisdom. The corollary of this line of thought is that Solomon’s
experience is available to all who would follow in his footsteps.
5. Imitation of the King
The relationship between kings and their subjects is the final matter to
be investigated. Two possible approaches exist according to Ecphantus:
subjects might be controlled through force and compulsion, or, where
48 Murray (“Philosophy and Monarchy”) argues that attempts to justify and control
monarchy lay at the heart of the περὶ βασιλείας treatises.
49 This idea is illustrated by a fragment from Thomas Mangey’s 1742 edition of Philo’s
works quoted by Goodenough (Politics of Philo, 99): “In his material substance (οὐσία) the
king is just the same as any man, but in the authority of his rank he is like the God of all.
For there is nothing upon earth more exalted than he. Since he is a mortal, he must not
vaunt himself; since he is a god he must not give way to anger. For if he is honored as
being an image of God, yet he is at the same time fashioned from the dust of the earth,
from which he should learn simplicity to all” (Mangey II, 673).
goodwill (εὐνοία) exists, they will imitate the king (82.28–83.17). There are
thus two paths presented here: the one marked by spontaneity and imi-
tation, the other by persuasion that comes through listening and subse-
quent obedience.50 As the king imitates God, so the king’s subjects would
do well to imitate him. Should they fail to do so, they must be persuaded
and/or compelled to obey. The former, imitation, is seen as superior
and desirable. But with a rare nod to realpolitik, the Pythagorean tracts
acknowledge that subjects might not always respond in this way and that
compulsion might become a necessity.
Diotogenes highlights the effect that the visage of the king should have
on his subjects. The king should, first, be majestic (σεμνός) “in his appear-
ance and utterances, and by his looking the part of a ruler”; second, he
should be gracious and beneficent; and third, inspire fear in his subjects
by dealing justly with the righteous and the wicked (73.23–28). Majesty is
a godlike thing that will result in admiration and honour from the king’s
subjects (73.28–74.1). Diotogenes concludes:
So will he succeed in putting into order those who look upon him, amazed
at his majesty, at his self-control, and his fitness for distinction. For to look
upon the good king ought to affect the souls of those who see him no less
than a flute or harmony. (74.15–19)
Ecphantus commends spontaneous imitation of the king and Diotogenes
would have the ideal king’s appearance inspire his subjects to act com-
mendably. These two approaches to ruling are combined in Philo’s Moses.
Following his ascent, Philo’s Moses is described as one whose life is set
before the Israelites “like some well-wrought picture, a piece of work
beautiful and godlike, a model for those who are willing to copy it” (Moses
1.158).
The use of the term “godlike” here reminds the reader that the king’s
reign imitates the divine: now that Moses has become like God, he is able
to present himself as a paradigm that his subjects might imitate. The fol-
lowing two sections (1.160–161) show that Philo is familiar with the polit-
ical discussion of imitation. He affirms the power that the king wields
because of the fact that his subjects imitate him. Philo also recognises that
this imitation can happen for better or for worse—an implicit acknowl-
edgement of the possibility of bad kingship.
50 See Calabi, Tradition and Philosophy, 205–214.

422 jonathan more
This section concludes with Philo’s recognition of Moses as embodied

and reasonable law (νόμος ἔμψυχός τε καὶ λογικός; Moses 1.162).51 Here,
as in the second volume (2.4–5), this concept connects Moses’ kingship
and his role as Israel’s lawgiver. Before the law was given, Moses and the
other patriarchs mirrored God’s law in their own lives. When Abraham
is described as one of the ἔμψυχοι καὶ λογικοὶ νόμοι, it is because he
lives in accordance with nature and follows the (as yet) unwritten law
(Abraham 3–5).52 In the Hellenistic political thinking illustrated by the
Pythagorean texts,53 the king was νόμος ἔμψυχος. Moses is embodied law
on both accounts: he is portrayed as a Hellenistic king and is also the one
who lives in accordance with God’s law and who is thus supremely quali-
fied to be Israel’s lawgiver. In the passage being considered here, Moses
1.158–162, these elements combine to make Moses worthy of imitation by
his subjects. Even though it is unlikely that they would ever be able to
fulfil this goal, the desire to do so is important (1.159).
The fact that the Wisdom of Solomon is addressed to the king’s fellow
rulers might initially lead one to assume that this text has nothing to say
to the ordinary person. A closer reading of the book indicates that this
is not the case and that the relationship between subject and king has
taken a distinctly democratic turn.54 Pseudo-Solomon has drawn on the
biblical tradition of universal kingship to affirm the royal potential of all
human beings. Alluding to the account of the creation of humanity in
Genesis, the king’s prayer in chapter 955 recognizes that all human beings
were originally created to rule over the other creatures and to manage the
world (9:2–3),56 a goal which, nevertheless, will not be achieved by all of
humanity. The opening chapters of the book introduce the righteous who
will enjoy immortality and eschatological rule over the wicked (3:7–8;
51 For a recent discussion on Philo’s understand of law, see J. W. Martens, One God, One
Law: Philo of Alexandria on the Mosaic and Greco-Roman Law (Studies in Philo of Alexan-
dria and Mediterranean Antiquity 2; Leiden: Brill, 2003), 83–101.
52 Pace Martens, ibid., 94 who argues that, for Philo, only Moses meets the full require-
ments for being considered νόμος ἔμψυχος.
53 Only Diotogenes 71.21–22 uses the phrase νόμος ἔμψυχος, but the concept is adum-
brated in Archytas (33.8–10) and contributes to the overall scheme of kingship presented
by Ecphantus. See Goodenough, “Hellenistic Kingship”; Chesnut, “Ruler and Logos,” 1313–
1320; Martens, One God, One Law, 31–66.
54 See Newman, “Democratization of Kingship.”
55 Cf. 1 Kgs 3:6–9; 2 Chron 1:8–10.
56 So, too, Philo: Moses 2.65; Creation 88 (see Runia, On the Creation of the Cosmos, 257).
See also Ps 8:6–8; Sir 17:2–4; 1QS 3.18; 2 Bar 14:18 and other texts cited by Winston, Wisdom
of Solomon, 201.
5:1–23). The righteous will receive “a glorious crown and a beautiful dia-
dem” (5:16) and will be made divine children of God (5:5). The inclusio
created by the address to rulers in 1:1 and 6:1 further highlights this link
between righteousness, wisdom/Wisdom and sovereignty.57
In the Book of Wisdom, immortality and rule are brought about by God
(9:2; 15:3) through the work of Wisdom (8:13, 17; 10:2). Kingship is thus not
limited to a unique and superior human being, but is available to all who
are righteous and who call upon God in order to receive Wisdom: “the
book affirms the democratization of kingship: ‘Everyman’ can be a king, to
the degree that it is possible for all to gain wisdom.”58 In confirmation of
the universal availability of Wisdom, chapter 10 provides a list of examples
of righteous people from the past, from Adam to Moses, who were rescued
and guided by Wisdom.59
Universal kingship is possible because Wisdom is available to all who
are willing to receive her from God (9:13–18). Solomon’s very real human-
ity (7:1–6) is proof of this. While the language of imitation does not occur
explicitly with regard to these kings, they are encouraged to be instructed
by Solomon’s words (6:11, 25) which, in turn, praise Wisdom and those
who would seek after her as Solomon has.60
The concept of the king as νόμος ἔμψυχος is absent in the Book of
Wisdom. The “incorruptible light of the law . . . given to the world” (18:4)
refers to the law of God, presumably as given through Moses, and yet
there is no mention of specifically Jewish elements like circumcision,
sabbath-keeping or food laws. Furthermore, the fact that the world-rulers
are meant to keep this law (6:4) highlights its universal applicability and
suggests that Pseudo-Solomon sees “the Law of Moses as the embodi-
ment of a universal law.”61 There is a link between the law/instruction
57 N. LaCoste (“Solomon the Exemplary Sage: The Convergence of Hellenistic and Jew-
ish Traditions in the Wisdom of Solomon,” The University of Toronto Journal for Jewish
Thought 1 [2010]: online: http://cjs.utoronto.ca/tjjt/node/18; accessed 1/4/2011) illustrates
Pseudo-Solomon’s blending of Jewish and Hellenistic sage-traditions. The relationship
between wisdom and sovereignty is also common to both of these traditions.
58 Newman, “Democratization of Kingship,” 327.
59 In two of these cases, Wisdom’s gift is closely tied to sovereignty: she gives Adam
strength to rule all things (10:2) and she led Joseph from being imprisoned to a position of
authority over those who had originally ruled over him (10:14).
60 The difficulty of deciding whether the Wisdom of Solomon should be thought of as
protreptic or encomium is illustrated here. For the former, see Winston, Wisdom of Solo-
mon, 18–20; for the latter, Gilbert, “Wisdom Literature,” 306–309.
61 Collins, Jewish Wisdom, 192.
424 jonathan more
and Wisdom (2:12; 3:11; 6:17–20),62 but Solomon is nowhere described as

“having become Wisdom” through God’s gift. The persons of Solomon and
Wisdom remain distinct, as, presumably, do Solomon and the law. The
narrative in chapters 1–5 suggests that Wisdom enables one to follow the
law and, in doing so, to become righteous. There is no sense in which a
righteous person or king defines the law in their own person.
The subjects’ response in the Book of Wisdom is not portrayed in terms
of either spontaneous imitation or persuasion as it is in the Pythagorean
texts and Philo. Nor is the close link between the king and the law as
described in those texts evident here. The call to imitation is not a call
to imitate Solomon’s person but to imitate his desire for Wisdom, and
through this, to follow God’s law. For this reason, the work is concerned
primarily with praising Wisdom and not with praising Solomon. The
ideal king and sage has been relegated to the background as Wisdom has
moved to the fore. In this regard, the Wisdom of Solomon stands in much
closer relationship to the Jewish wisdom tradition than to the Pythago-
rean political texts.
6. Conclusion
Philo and the author of the Book of Wisdom have drawn on Jewish as
well as Graeco-Roman traditions in their respective portrayals of Moses
and Solomon as kings. While it is difficult to adduce sufficient evidence to
prove a direct relationship between these Jewish texts and the Pythago-
rean περὶ βασιλείας writings, their dependence on a common pool of ideas
has been demonstrated. These commonalities serve to highlight the link
in Alexandria between Hellenistic Pythagorean philosophy, on the one
hand, and Jewish thought, on the other.
This is not to say that the two Jewish authors have drawn from the
Pythagorean texts in the same way. The elements examined above have
illustrated some of the differences between the Jewish authors. Philo’s
conceptualization of ideal kingship contains a number of similarities to
those found in the Pythagorean texts. Many of these ideas are absent or
suppressed in Pseudo-Solomon. Some of the differences between the texts
might be explained on the basis of the different audience envisaged by
each author.
62 The relationship between Torah and Wisdom was already established by the time
Pseudo-Solomon wrote; see Collins, Jewish Wisdom, 54–56.
Philo’s biography of Moses seems to have been written for “outsiders;”

it is an exoteric text that serves as an introduction to Moses and his laws.63
Philo’s explicit purpose is to make Moses known to those who have
been ill-served by Greek writers (Moses 1.1–2). He thus portrays Moses as
standing in the tradition of the Platonic philosopher-king which reaches
its apotheosis in the ideas represented by the Pythagorean περὶ βασιλείας
fragments. The authority and superiority of the Mosaic laws, a common
theme in Philo’s writings,64 are reinforced by coming from Moses, the one
who is uniquely βασιλεύς and νόμος ἔμψυχος.
Possibly written for an audience more familiar with the Jewish scrip-
tural traditions,65 the Wisdom of Solomon has less of an apologetic tone
than Philo’s Moses. The focus is not on Solomon as the ideal king, but on
the benefits of acquiring Wisdom. The fundamental idea of imitation of
God’s virtues as presented in the person of the king has been replaced by
the reception of God’s Wisdom.
While the similarities between these two authors have often been
noted,66 the comparisons presented in this essay show how Pseudo-
Solomon and Philo have put to quite different use one topos drawn from
philosophical traditions current in the Alexandria of their time. This
illustrates how the same conceptual matrix can be employed for different
ends. These observations highlight, again, the variegated nature of reli-
gious and philosophical thinking in Alexandrian Judaism.
63 Morris, “Philo,” 855.

64 E.g., Moses 2.12–16; see Morris, “Philo,” 875–876.
65 Winston, Wisdom of Solomon, 63–64; Gilbert, “Wisdom Literature,” 312.
66 E.g. Winston, Wisdom of Solomon, 59–62.
The Text Form of the Isaiah Quotations in the
Sondergut Matthäus Compared to the Dead Sea Scrolls,
Masoretic Text and Septuagint
Gert J. Steyn
1. Introduction
Early Christian hermeneutics stands in the centre of the formation of early

Christian theology. It was during the process of interpreting the very same
Scriptures of the Jews now in terms of the Jesus event that early Christian-
ity developed as another alternative Jewish movement to those that were
in existence during the 1st century C.E.1 The LXX, available in a range of
different versions at this time, was the “Bible” of early Christianity. They
quoted from it, referred to it, alluded to it, paraphrased it and used its
language and terminology when dealing with OT motifs. The Greek trans-
lation of the Hebrew OT itself built an interpretative bridge between the
Jewish Scriptures and the Greek speaking world. But which particular LXX
versions were available? Do we have alternative Hebrew textual traditions
that confirm such LXX versions? These are vitally important questions,
because only once this has been compared with the text form of the NT
reading, can one get an idea of what possible changes might have been
made by the NT author.
Major shifts took place in studies on “the OT in the NT” during the
last decade. These were particularly due to the compilation of the eclec-
tic editions of the LXX by the Septuaginta Unternehmen, to studies on
the language and terminology of the LXX,2 to research on the relation
between the Dead Sea Scrolls (DSS) and the LXX,3 and to the work on the
1 Examples include groups such as those at Qumran, the Therapeutae to whom Philo
of Alexandria referred, and that of John the Baptist.
2 To be included in this area of LXX research are studies on the language and ter-
minology, work that is mainly done by the schools of Louvain (Belgium) and Helsinki
(Finland).
3 Notice should be taken here of work by OT scholars such as Emanuel Tov, Albert
Pietersma on the Psalms (e.g. “The Present State of the Critical Text of the Greek Psalter,”
in Der Septuaginta-Psalter und seine Tochterübersetzungen. Symposium in Göttingen 1997
[MSU XXIV; eds. A. Aejmelaeus and U. Quast; Göttingen: Vandenhoeck & Ruprecht, 2000],
428 gert j. steyn
LXX translations into English,4 German5 and French.6 One such develop-
ment was an awareness that arose for the particular text forms of the LXX
that were in circulation7—an aspect that is of great importance for NT
scholarship. Attention is now focused on the question about which of the
available OT text forms might underlie a particular OT quotation in the
NT. Unless this has not been properly established, no responsible observa-
tions can be made about the NT author’s hermeneutic, and especially his
adaptation and application of his Scriptures for his particular theological
purposes. There is thus today a realization that one cannot just simply
attempt to discuss the function of such OT material unless at least the
range of available Hebrew and Greek textual witnesses were first identi-
fied and assessed in order to determine the (mainly LXX) text form that
was at the disposal of the NT author.8 Only then can one determine in
each instance the places where differences are noted between the recon-
structed texts of the OT and that of the NT, whether a NT author made use
of another OT text version, how the text form of that version might have
looked like and then, after this possibility has been eliminated, to identify
the possible changes that might be due to the NT author’s hand due to
12–32); Adrian Schenker on Jeremiah and Siegfried Kreuzer on the Antiochian Text of par-
ticularly Kingdoms and Paralipomenon.
4 A. Pietersma and B. G. Wright (eds.), A New English Translation of the Septuagint and
the Other Greek Translations Traditionally Included under That Title (Oxford / New York:
Oxford University Press, 2007). See also the recent work of R. J. V. Hiebert (ed.), “Transla-
tion is Required”. The Septuagint in Retrospect and Prospect (SCS 56; Atlanta: Society of
Biblical Literature, 2010).
5 W. Kraus and M. Karrer (eds.), Septuaginta Deutsch. Das griechische Alte Testament in
deutscher Übersetzung (Stuttgart: Deutsche Bibelgesellschaft, 2009).
6 La Bible d’Alexandrie, Tome 1–25 (Paris: Le Cerf, 1986–2011).
7 Cf., for example, D.-A. Koch, Die Schrift als Zeuge des Evangeliums. Schriftverwendung
und Schriftverständnis bei Paulus (Tübingen: Mohr Siebeck, 1985); M. J. J. Menken, Mat-
thew’s Bible. The Old Testament Text of the Evangelist (BETL CLXIII; Leuven: Peeters, 2004);
G. J. Steyn, A Quest for the Assumed LXX Vorlage of the Explicit Quotations in Hebrews (Göt-
tingen: Vandenhoeck & Ruprecht, 2011). An important contribution in this regard is also
the work done on the electronic databank of the Wuppertal-Münster “Institut für Septua-
ginta- und biblische Textforschung” by Martin Karrer, Ulrich Schmid, Marcus Sigismund,
Michael Labahn and others.
8 See G. J. Steyn, “Which ‘LXX’ are we talking about in New Testament scholarship?
Two examples from Hebrews,” in Die Septuaginta—Texte, Kontexte, Lebenswelten (WUNT
219; eds. M. Karrer and W. Kraus; Tübingen: Mohr Siebeck, 2008), 297–307; and particu-
larly idem, “Comparing Manuscripts with Manuscripts. Thoughts on the Compilation of a
Synopsis of Textual Variants in the Old Testament Quotations in the New Testament,” in
Von der Septuaginta zum Neuen Testament. Textgeschichtliche Eröterungen (eds. S. Kreuzer,
M. Karrer and M. Sigismund; Berlin: De Gruyter, 2010), 211–226.
the text form of the isaiah quotations 429
grammatical, and more interesting for biblical scholarship, to theological

reasons.
This brief study is an attempt to investigate a very small section of
the NT in this regard. The focus is on the Isaiah quotations in Matthew’s
gospel—but particularly narrowed down to those Isaiah quotations where
extant witnesses from the DSS are available and only to those that occur
in the Sondergut Matthäus material. In this sense the OT comparative
basis of textual forms is expanded to include the DSS texts. By doing this,
the relation between the DSS, MT and LXX can be accommodated in
the research. Furthermore, by focusing only on the Sondergut Matthäus
material we can establish closer the text form generally used by Matthew
(excluding his material from Mark and Q) for the Isaiah quotations that
he employed. The purpose is then to establish the relation between the
text forms of the Hebrew versions (thus also the DSS) and the available
LXX witnesses with that of Matthew’s text form. The question asked here
is then whether Matthew’s text form in these quotations shows closer
similarities with a LXX version or with the Hebrew witnesses.
Maarten Menken has, in my opinion, extensively dealt with and largely
answered the questions on Matthew’s OT text form in recent years.9 In
that sense, this study is almost superfluous. However, it is hoped that by
firstly unpacking and repacking the same material in a different manner—
focusing on the role of the DSS alongside the LXX—one might discover
even more, or simply confirm those results. Secondly, it is particularly
hoped that one could identify closer similarities to particular text types
and textual trajectories of the OT that might underlie the Isaiah quota-
tions in Matthew’s Sondergut material.
2. Isaiah Quotations with DSS Attestation in the

Sondergut Matthäus
2.1. The Introductory Formulae

There are six quotations from Isaiah in the Sondergut Matthäus of which
we have extant textual witnesses amongst the DSS. All six quotations
belong to the “fulfillment quotations”10 in Matthew and are introduced
9 Cf. Menken, Matthew’s Bible.

10 There are 11 fulfillment quotations in Matthew: six from Isaiah, two from Zechariah,
one from Jeremiah, one from Hosea, one from the Psalms and one from an unknown
source. No fulfillment quotations were taken from the Torah.
430 gert j. steyn
with an introductory formula ἵνα πληρωθῇ τὸ ῥηθὲν (ὑπὸ κυρίου, 1:23) διὰ
(Ἠσαΐου, 4:15–16; 8:17; 12:18–21) τοῦ προφήτου λέγοντος. It should be noted
that only the first of these quotations contains the phrase “by the Lord”
(ὑπὸ κυρίου) and that the next three cases (Matt 4:15–16; 8:17 and 12:18–21),
in its consecutive sequence in Matthew, all have an identical introductory
formula. The last case omits both “by the Lord” as well as “Isaiah.” The
fifth case, which is embedded within the third of the five big discourses
of Jesus in Matthew, differs somewhat in its formulation (καὶ ἀναπληροῦται
αὐτοῖς ἡ προφητεία Ἠσαΐου ἡ λέγουσα). In the last five of these cases the
Isaiah quotation is embedded as Sondergut Matthäus insertions within
Markan material used by Matthew. Despite my own skepticism regarding
the “Testimony Book” hypothesis11 of Harris12 (lately supported by Albl ),13
this might indeed point here in this case to the existence of a florilegium,
or a testimony list, which was at Matthew’s disposal—most probably
compiled by Matthew himself 14 (due to its absence elsewhere in early
Jewish and Christian literature) prior to the writing of his gospel.
Matthew SMt place Isaiah Differences Introductory formula

1 Matt 1:23 SMt Isa 7:14 1 Fulfillment citation
2 Matt 4:15–16 SMt in Mk Isa 8:23–9:1 10+ Fulfillment citation
3 Matt 8:17 SMt in Mk Isa 53:4 5 Fulfillment citation
4 Matt 12:18–21 SMt in Mk Isa 42:1–4 19 Fulfillment citation
5 Matt 13:14–15 SMt in Mk Isa 6:9–10 1 Fulfillment citation
(3rd Mt Jesus-Discourse)
6 Matt 21:5 SMt in Mk Isa 62:11 none in the Fulfillment citation
(+Zech 9:9) Isaiah section
11 Cf. Steyn, Quest.

12 J. R. Harris, Testimonies Part I and Part II (Cambridge: Cambridge University Press,
1916/1920).
13 M. C. Albl, “And Scripture Cannot be Broken”. The Form and Function of the Early
Christian Testimonia Collections (NovTSup 96; Leiden: Brill, 1999).
14 “Scholars are uniform in their opinion that they (formula quotations, GJS) derive
from the final editor of Matthew’s work” (R. Beaton, “Isaiah in Matthew’s Gospel,” in Isa-
iah in the New Testament [eds. S. Moyise and M. J. J. Menken; London / New York: T & T
Clark], 64).
The Isaiah quotations in the SMt sections contain the longest quotations
from Isaiah in comparison to those found in the Markan and Q material.
This is an interesting tendency and can be observed by other NT authors
as well. The unknown author of Hebrews, for instance, quoted long pas-
sages from Jeremiah, Ps 40 and Ps 95 that occur only in his work as quo-
tations and nowhere else prior to his time of writing. He also expanded
on existing quotations (such as that of Ps 8:7)—similar to what Matthew
does with his Isa 6:9–10 quotation.
Case 1. Isa 7:14 in Matt 1:23

4Q55 (4QIsaa) 4Q65 (4QIsal ) MT Isa 7:14 LXX Isa 7:14 Matt 1:23
‫] העלמה‬ ‫]הנה‬
֯ [ ‫ִה ֵּנ֣ה ָה ַע ְל ָ֗מה‬ ἰδοὺ ἡ παρθένος ἰδοὺ ἡ παρθένος
‫הרה‬ ‫ָה ָר ֙ה‬ ἐν γαστρὶ ἕξει ἐν γαστρὶ ἕξει
‫וילדת בן‬ ‫וְ י ֶֹל ֶ֣דת ֵּ֔בן‬ καὶ τέξεται υἱόν, καὶ τέξεται υἱόν,
‫וקר א‬ ‫וְ ָק ָ ֥רא ת‬ καὶ καλέσεις καὶ καλέσουσιν
‫שמו‬ ‫ְׁש ֹ֖מו‬ τὸ ὄνομα αὐτοῦ τὸ ὄνομα αὐτοῦ
‫עמנואל‬ ‫ִע ָ ּ֥מנּו ֵ ֽאל׃‬ Εμμανουηλ· Ἐμμανουήλ,
The reconstructed LXX reading15 and that of the NT seem to be here vir-
tually identical, except for the change in number where LXX reads a 2nd
person singular, whilst Matthew reads a 3rd person plural.16 None of the
Hebrew witnesses, including the DSS, support Matthew’s 3rd person plu-
ral reading.17 The closest seems to be the 3rd person in the Hebrew 1QIsaa
(which is probably an attempt to simplify “a difficult form”18 compared
to the MT) and the Greek LXX Codex Sinaiticus—but in both instances
in the singular19 and not in the plural as in Matthew’s version. (Menken20
15 J. Ziegler, Isaias Vol. XIV (Göttingen: Vandenhoeck & Ruprecht, 1983), 147.
16 According to D. A. Hagner, “This is probably Matthew’s alteration of the text rather
than a variant (Matthew has ‘she’), made in order to avoid the conflict between the com-
mand to Joseph to name the child Jesus and the statement of Isaiah that the child shall be
named Emmanuel” (Matthew 1–13, [WBC 33A; Dallas: Word, 2002]), 20–21.
17 According to G. L. Archer and G. Chirichigno, “The καλέσουσιν is an Aramaic-type
third person plural, equivalent to a passive” (Old Testament Quotations in the New Testa-
ment [Chicago: Moody Press, 1983], 95).
18 H. P. Scanlin, The Dead Sea Scrolls and Modern Translations of the Old Testament
(Wheaton: Tyndale House Publishers, 1993).
19 So correctly, G. L. Archer and G. Chirichigno, OT Quotations, 95.
20 Menken, Matthew’s Bible, 117–131.
432 gert j. steyn
and Blomberg21 mistakenly took the reading in LXX Codex Sinaiticus as

a plural with the possibility that Matthew might have followed this LXX
variant).
One can thus assume with fair certainty that at least the plural of the
3rd person, as found in Matthew’s version, is due to his own hand.
2nd sg = καλέσεις (fut ind) 3rd sg = καλέσει 3rd pl = καλέσουσιν

cf. Matt 1:21; Luke 1:31 (fut ind) (fut ind)
• MT Isa 7:14 [‫]וְ ָק ָ ֥ראת‬ • 1QIsaa [‫( ]וקרא‬Qal perf • Matt 1:23
(Qal perf 2nd fem sg)22 3rd masc)23 • LXX minuscules
• LXX Isa 7:14 (Codex A, • LXX (Codex ‫[ א‬καλέσει] (10th–14th cent.)
B). Aquila, Symmachus, (fut ind); Codex Q 26–106; 90 mg 130; 233;
Theodotion (Marchalianus, 393; 410c; 499′; 456;
• Matt 1:23 (Codex Bezae; 6th cent.) 534; 764c
pc Bomss Origen, • LXX translations: O. Lat • LXX Translations:
Eusebius) (vocabitis), Syriac, Bo Syp
Vulgate (vocabitur) • LXX quotations in
• LXX minuscules Church Fathers:
(12th–14th cent.) 311–46 Origen, Epιhanius,
Chrysostom,
Theodoret
The LXX translates the Hebrew term ‫— ָה ַע ְל ָ֗מה‬used generally for a “girl of
marriageable age,”24 “a young woman who may or may not be a virgin”25—
at some places (Exod 2:8; LXX Ps 67:26) with νεᾶνις. The same applies to the
LXX recensions of Aquila, Symmachus and Theodotion. At other places,
however, ‫ ָה ַע ְל ָ֗מה‬is translated with ἡ παρθένος (e.g., Gen 24:43)—such
as here in Isa 7:14.26 The NT continues to use in this quotation the LXX
21 C. L. Blomberg, “Matthew,” in Commentary on the New Testament Use of the Old Tes-
tament (eds. G. K. Beale and D. A. Carson; Grand Rapids: Baker Academic, 2007), 1–110,
here 4.
22 According to H. P. Scanlin, “. . . apparently second person feminine singular, but per-
haps third person . . .” (Dead Sea Scrolls). Cf. also B. Davidson, The Analytical Hebrew and
Chaldee Lexicon (Lynn: Hendrickson, 1981), 665.
23 So H. P. Scanlin, Dead Sea Scrolls. According to the parsing of Logos, it is taken as
feminine.
24 W. Gesenius and S. P. Tregelles, Gesenius’ Hebrew and Chaldee Lexicon to the Old
Testament Scriptures (Bellingham, WA: Logos Research Systems, Inc, 2003), 634.
25 Hagner, Matthew I, 20–21.
26 According to W. Gesenius and S. P. Tregelles, “The notion of unspotted virginity is
not that which this word conveys, for which the proper word is ‫תּולה‬ ָ ‫( ְּב‬see Cant. 6:8,
term παρθένος without any difficulties in the sense of “a young unmarried

girl” with the “implication of virginity”27 at places such as Matt 1:23,28
Luke 1:27,29 Acts 21:9, etc.
Case 2. Isa 8:23–9:1 in Matt 4:15–16

(1Q Isaa) MT Isa 8:23–9:1 LXX Isa 9:1–2 Matt 4:15–16
‫ארץ זבולון‬ ‫לּון‬

֙ ‫ ַ ֤א ְר ָצ ה זְ ֻב‬χώρα Ζαβουλων, γῆ Ζαβουλὼν
‫ א ְר ָצ ה נַ ְפ ָּת ֔ ִלי וה ארץ נפתלי‬
֣ ַ ְ‫ ו‬ἡ γῆ Νεφθαλιμ καὶ γῆ Νεφθαλίμ,
‫וְ ָה ַא ֲח ֹ֖רון ִה ְכ ִ ּ֑ביד והאחרון הכביד‬
‫דרך הים‬ ‫ ֶ ּ֤ד ֶרְך ַהּיָ ֙ם‬ὁδὸν θαλάσσης ὁδὸν θαλάσσης,
καὶ οἱ λοιποὶ οἱ τὴν
παραλίαν
κατοικοῦντες
‫עבר הירדן‬ ‫ֵע ֶ֣בר ַהּיַ ְר ֵ ּ֔דן‬ καὶ πέραν τοῦ Ιορδάνου, πέραν τοῦ Ἰορδάνου,
‫גליל הגואים‬ ‫ּגְ ִ ֖ליל ַהֹּגויִ ֽם׃‬ Γαλιλαία τῶν ἐθνῶν, Γαλιλαία τῶν ἐθνῶν,
τὰ μέρη τῆς Ιουδαίας.
‫ העם ההולכים‬a‫ָה ָע ֙ם ַהה ְֹל ִ ֣כים‬ ὁ λαὸς ὁ πορευόμενος ὁ λαὸς ὁ καθήμενος
. ‫בחוש‬ ‫ַּב ֔חֹ ֶׁש ְך‬ ἐν σκότει, ἐν σκότει
‫ראו אור גדול‬ ‫ָר ֖אּו ֹ֣אור ּגָ ֹ֑דול‬ ἴδετε φῶς μέγα· φῶς εἶδεν μέγα,
‫יושבי‬ ‫בי‬
֙ ֵ ‫י ְֹׁש‬ οἱ κατοικοῦντες καὶ τοῖς καθημένοις
‫בארץ‬ ‫ְּב ֶ ֣א ֶרץ‬ ἐν χώρᾳ καὶ ἐν χώρᾳ καὶ
‫צלמות‬ ‫ַצ ְל ָ֔מוֶ ת‬ σκιᾷ θανάτου, σκιᾷ θανάτου
‫יהם׃ אור נגה עליהם‬
ֽ ֶ ‫ֹ֖אור נָ ַגּ֥ה ֲע ֵל‬ φῶς λάμψει ἐφʼ ὑμᾶς . φῶς ἀνέτειλεν
αὐτοῖς .
The LXX and NT have numerous differences amongst each other and the
reconstructed LXX reading leaves the impression of a paraphrase or free
and Prov. loc. cit; so that in Isa. loc. cit. the LXX. have incorrectly rendered it παρθένος)”
(Hebrew and Chaldee Lexicon, 634).
27 I. H. Marshall, The Gospel of Luke: A Commentary on the Greek Text (NIGTC; Exeter:
Paternoster Press, 1978), 64.
28 “Matthew, unquestionably delighted with the agreement between the tradition
about Jesus’ birth and the words of Isaiah, not only prefaces the quotation with a formula
of fulfillment but even conforms the wording of the surrounding narrative to that of the
quotation” (Hagner, Matthew I, 20–21).
29 Marshall argues at this point: “In view of 1:34 this implication is undoubtedly present
here, a view which is strengthened by the probable allusions to Is. 7:14 here and in v. 31”
(Gospel of Luke, 64).
434 gert j. steyn
translation30 in comparison with the Hebrew texts at our disposal. Apart

from the fact that both Matthew and the LXX witnesses lack the Hebrew
phrase ‫“( וְ ָה ַא ֲח ֹ֖רון ִה ְכ ִ ּ֑ביד‬but in the latter times he will glorify”), which is
present in the MT and DSS, Matthew seems to be in general closer here to
the Hebrew MT. If Blomberg is correct in assuming that “Matthew can be
understood as following the LXX but improving its translation at several
points to bring it more in line with the Hebrew,”31 then it implies that
(a) Matthew knew the Hebrew version with its differences just as well as
that of the LXX, and that (b) the reconstructed LXX version used for our
comparisons was indeed the Greek OT version that Matthew had at his
disposal. Both these assumptions are problematic.32
The reconstructed LXX text is longer than the Hebrew textual traditions
and includes two expansions33 that are neither present in Matthew, nor
in the Hebrew traditions: (a) καὶ οἱ λοιποὶ οἱ τὴν παραλίαν κατοικοῦντες and
(b) τὰ μέρη τῆς Ιουδαίας. Beaton argues that Matthew’s “omission” of the
latter phrase (τὰ μέρη τῆς Ιουδαίας) implies “that the regional locatedness
of Galilee is important for Matthew.”34 But these phrases are also absent
(apart from the DSS, MT and NT) in the following LXX witnesses:
καὶ οἱ λοιποὶ οἱ τὴν παραλίαν κατοικοῦντες τὰ μέρη τῆς Ιουδαίας

LXX Codex ‫*א‬ LXX minuscules (8th–13th cent)
LXX minuscules (8th–12th cent) O´’ L´’-46–233–456–764c C 239´ 393
O´ L´-311–456–764c C 393 407 410 538
LXX translations Co Syp
LXX quotations in Church Fathers: LXX quotations in Church Fathers:
Eusebius, Basilius, Chrysostomus, Eusebius, Basilius, Theodoret, Cyrillus,
Cyrillus, Tertullian Tertulian, Cyprian, Hieronymus
30 Similarly Blomberg, “Matthew,” 19.

31 Blomberg also keeps the possibility open that it “is just barely possible that Matthew
arrived at the same translation of the Hebrew independently” (“Matthew,” 19).
32 Cf. ibid. Beaton defines it better: “The text-form shares similarities with both the MT
and LXX, but there are also several readings in it that are unique to Matthew” (“Isaiah,”
67).
33 The marginal notes of the 6th century Codex Marchalianus, which originated in
Egypt, also expands οδον θαλασσης slightly to read: τινες ουκ εχ(ουσιν) οδον θαλασσης.
34 Beaton, “Isaiah,” 68.
Despite the fact that the Matthean version of the extant Codex Alexandri-
nus only starts at Matt 25:6, it might be useful to still compare at least one
other uncial, namely Codex Sinaiticus. It is particularly interesting that
the first expansion is absent in Isa 8:23–9:1 in the original of Codex Sinaiti-
cus, but present in the overwritten folio—whereas both versions also lack
the phrase ὁδὸν θαλάσσης35—added by a later hand. In Matt 4:15–16, how-
ever, the phrase is completely absent in this NT part of Codex Sinaiti-
cus. Furthermore, all the LXX textual traditions read χώρα—ἡ γῆ36 at the
beginning of the text quoted by Matthew. This is strange as all the Hebrew
textual traditions have ‫ וְ ַ ֣א ְר ָצה— ַ ֣א ְר ָצה‬and the NT traditions γῆ—καὶ γῆ.
Four further differences, which are mostly substitutions, between the
reconstructed LXX reading and Matt 4:15–16, are the following:
The LXX has ὁ λαὸς ὁ πορευόμενος Matthew has ὁ λαὸς ὁ καθήμενος

(walking) (sitting)
The reconstructed LXX reading is in Amongst the LXX witnesses, also
agreement with the Hebrew traditions Codex Alexandrinus,37 a number of
(MT, DSS) and is supported, amongst minuscules and some Church Fathers
others, by the 3rd century Chester support the Matthean reading. These
Beatty papyrus 965 and by the Church include LXX oII 36c cI’ 301 407 534
Father Cyprian. Origen, Cyril, Tertullian.
The LXX reads ἴδετε φῶς μέγα Matthew reads φῶς εἶδεν μέγα
The LXX reading with the 2nd person Matthew’s reading reflects more
plural and in an imperative form closely the reading of the Hebrew
differs from that of the Hebrew traditions (MT, DSS).38 The reading is
traditions which have the 3rd person also supported by improvements to
plural (‫) ָר ֖אּו‬. the LXX Lucian text. A number of LXX
minuscules and Church Fathers follow
the Matthean reading: oII L ’́ -36–46–456
C 2́ 39 ́ 403 ́ 449 ́ 538 Sa Syp Or Chr Tht
Cyr Hi.
35 The latter is also absent in O ‘́ L ́ -96–311–456–764c C 301 393 538 544 Sa Syp Eus Bas
Chr Tert Hi.
36 Some LXX later witnesses replace the article ἡ with καί—which brings it in line with
the NT reading: 106 olI lI-764c 301 403 ́ 538 Syp Eus Tert Cypr.
37 One should be cautious, however, with Codex Alexandrinus as it is known for its
changes in the OT text of passages quoted in the NT to bring it on a par with the NT
reading.
38 So also Archer and Chirichigno, OT Quotations, 99.
436 gert j. steyn
Not only is the word order different here, but Matthew’s verb is presented
in the aorist indicative active 3rd singular of εἶδον, whereas the LXX verb
is an aorist imperative active 2nd plural of ὁράω.
The LXX reads οἱ κατοικοῦντες Matthew has καὶ τοῖς καθημένοις

(dwelling). (sitting).
There are no NT textual witnesses that There are no Hebrew or LXX witnesses
testify to this LXX reading. that support this reading of Matthew.
The LXX ends with λάμψει (shine) Matthew reads ἀνέτειλεν (rise) αὐτοῖς.
ἐφʼ ὑμᾶς. The 9th century LXX minuscule 301
The Church Father Cyprian supports and the Church Father Tertullian
this LXX reading. follow this NT reading.
Case 3. Isa 53:4 in Matt 8:17

(1Q Isaa) (1Q Isab) MT Isa 53:4 LXX Isa 53:4 Matt 8:17
‫ ָא ֵכ֤ן‬οὗτος τὰς αὐτὸς τὰς

‫חולי‸י‸נו‬ ‫חלינו‬ ‫ ֳח ָל ֵ֨י ֙נּו‬ἁμαρτίας ἀσθενείας
ἡμῶν ἡμῶν
‫הואה נשא‬ ‫הוא נשא‬ ‫֣הּוא נָ ָׂ֔שא‬ φέρει ἔλαβεν
‫ומכאובינו‬ ‫ומכאבינו‬ ‫ּומ ְכא ֵ ֹ֖בינּו‬ַ καὶ περὶ καὶ τὰς
‫סבלם‬ ‫[סבלם‬ ‫ְס ָב ָל֑ם‬ ἡμῶν νόσους
ὀδυνᾶται ἐβάστασεν
Regarding the Hebrew text, the Syriac versions and the Vulgate39 assume
an emphatic ‫ הּוא‬between ‫ּומ ְכא ֵ ֹ֖בינּו‬
ַ and ‫ ְס ָב ָל֑ם‬. There are no variant
readings of Matt 8:17 amongst the NT witnesses. The reading of Matt 8:17
is also closer to the Hebrew versions (DSS, MT)40 and the Greek transla-
tions of Aquila and Symmachus, than to the reconstructed LXX and the
Targums.41 Despite the fact that the reconstructed LXX reading and that of
39 Vere languores nostros ipse tulit, et dolores nostros ipse portavit (Isa 53:4).
40 Cf. also Archer and Chirichigno, OT Quotations, 121; Beaton, “Isaiah,” 69. So also
Blomberg, “Matthew,”, 32. He states that “The MT obviously fits Matthew’s context of
physical healings much better than does the LXX.”
41 Menken, Matthew’s Bible, 35–49; Beaton, “Isaiah,” 69.
the NT show the same format, their contents differ substantially amongst
each other.42 These differences are the following:
The LXX starts with the nominative The NT starts with the 3rd person
masculine singular demonstrative singular personal pronoun αὐτός
pronoun οὗτος
LXX minuscule 88 (the only extant
Greek witness of the 5th column of
Origen’s Hexapla) and the marginal
notes of the Syrohexaplaric translation
have the same reading as the NT. The
next verse (LXX Isa 53:5) starts with
αὐτός. The Hebrew versions also use
the 3rd person personal pronoun.
The LXX reading has ἁμαρτίας The NT reading chooses ἀσθενείας
(accusative feminine plural, (accusative feminine plural, “sickness,
“sin, failure”) disease”)
The Lucian group L’ and a few 11th–12th The 11th–13th century LXX minuscules
century LXX minuscules ([22*]–62–96 22c-93 also read ἀσθενείας as Matthew
449’) as well as some Church Fathers does. The NT reading agrees here with
(Athanasius, Theodoret) chose to read the Hebrew versions (DSS, MT).
μαλακιας here, which aligns the reading
with the MT.
The LXX has the verb φέρει The NT uses ἔλαβεν
(praesens indicative active 3rd singular, (aorist indicative active 3rd singular,
“to bear, to carry”) “to take away, to remove”)
The LXX reading, rather than the NT, is There are no LXX witnesses that
here closer to the Hebrew versions. support this NT reading.
The LXX ends with περὶ ἡμῶν ὀδυνᾶται The NT ends with τὰς νόσους
(praesens indicative medium/passive (accusative feminine plural—“disease,
3rd singular—“to cause pain”, ὀδυνάω) illness”) ἐβάστασεν (aorist indicative
active 3rd singular—“to carry away,
to remove”)
The NT is here again closer to the
Hebrew versions.
42 Cf. Blomberg: “Nothing in Matthew’s translation requires dependence on the LXX”

(“Matthew,” 32). Menken holds a different opinion (Matthew’s Bible, 35–49).
438 gert j. steyn
Scholars tend to think that Matthew might have done the Greek trans-
lation from the Hebrew himself.43 A good case in point is the use of
ἡμῶν after νόσους which is a “completely literal rendering” of the Hebrew
‫ ַמ ְכא ֵ ֹ֖בינּו‬. However, there is insufficient evidence that Matthew also trans-
lated other OT passages directly from the Hebrew into Greek. It would
thus be difficult to explain why he would do so in some instances, but use
existing Greek translations for other passages. The explanation for Mat-
thew’s closeness to the Hebrew textual traditions should probably rather
be searched for in the existence of another LXX text form which was at
Matthew’s disposal.
Case 4. Isa 42:1–4 in Matt 12:18–21

(1Q Isaa) 4Q56 MT Isa 42: LXX Isa 42:1–4 Matt 12:18–21
(4QIsab) 1–4
‫הנה‬ a ‫ ֵ ֤הן‬Ιακωβ ἰδοὺ

‫עבדי‬ ‫ ַע ְב ִ ּ֙די‬ὁ παῖς μου, ὁ παῖς μου
‫אתמוכה בו‬ ‫ְך־ֹּבו‬
֔ ‫ ֶא ְת ָמ‬ἀντιλήμψομαι ὃν ᾑρέτισα ,
αὐτοῦ · Ισραηλ
‫בחירי‬ b‫ ְּב ִח ִ ֖ירי‬ὁ ἐκλεκτός μου , ὁ ἀγαπητός μου
‫רצתה‬ ‫ ָר ְצ ָ ֣תה‬προσεδέξατο εἰς ὃν εὐδόκησεν
αὐτὸν
‫נפשי נתתי‬ ‫ נַ ְפ ִ ׁ֑שי נָ ַ ֤ת ִּתי‬ἡ ψυχή μου· ἡ ψυχή μου·
ἔδωκα θήσω
‫רוחי‬ ‫רּוחי‬
ִ֙ τὸ πνεῦμά μου τὸ πνεῦμά μου
‫עליו‬ ‫ ָע ֔ ָליו‬ἐπʼ αὐτόν, ἐπʼ αὐτόν, καὶ
‫ומשפטי לגואים‬ ‫ ַלֹּגויִ ֥ם‬c ‫ ִמ ְׁש ָ ּ֖פט‬κρίσιν τοῖς κρίσιν τοῖς
‫יוציא‬ ‫ֹיוציא׃‬ ִֽ ἔθνεσιν ἐξοίσει . ἔθνεσιν
ἀπαγγελεῖ .
‫לוא יזעק‬ ‫ ֥ל ֹא יִ ְצ ַ ֖עק‬2 οὐ κεκράξεται οὐκ ἐρίσει
‫ולוא ישא‬ ‫ וְ ֣ל ֹא יִ ָ ּׂ֑שא‬οὐδὲ ἀνήσει , οὐδὲ κραυγάσει ,
‫ולוא ישמיע‬ ‫ וְ ֽל ֹא־יַ ְׁש ִ ֥מ ַיע‬οὐδὲ οὐδὲ ἀκούσει
43 Beaton, “Isaiah,” 69; Blomberg, “Matthew,” 32.

Case 4. (cont.)
(1Q Isaa) 4Q56 MT Isa 42: LXX Isa 42:1–4 Matt 12:18–21
(4QIsab) 1–4
ἀκουσθήσεται τις ἐν ταῖς

‫בחוץ‬ ‫[בחו]ץ‬
ֿ ‫ ַּב ֖חּוץ‬ἔξω πλατείαις
‫קולו‬ ‫קולו‬ ‫ֹקוֹלו׃‬
ֽ ἡ φωνὴ αὐτοῦ. τὴν φωνὴν
αὐτοῦ.
‫קנה‬ ‫קנה‬ ‫ ָק ֶנ֤ה‬3 κάλαμον κάλαμον
‫רצוץ‬ ]‫֯ר[צוץ‬ ‫צּוץ‬
֙ ‫ָר‬ τεθλασμένον συντετριμμένον
‫לוא ישבור‬ ‫֣ל ֹא יִ ְׁש ֹּ֔בור‬ οὐ συντρίψει οὐ κατεάξει
‫ופשתה כהה‬ ‫ּופ ְׁש ָ ּ֥תה ֵכ ָ ֖הה‬
ִ καὶ λίνον καὶ λίνον
καπνιζ όμενον τυφόμενον
‫לוא יכבה‬ ‫ ֣ל ֹא יְ ַכ ֶ ּ֑בּנָ ה‬οὐ σβέσει, οὐ σβέσει,
‫לאמת יוציא‬ ‫ֹיוציא‬
֥ ִ ‫ ֶל ֱא ֶ ֖מת‬ἀλλὰ εἰς
‫משפט‬ ‫[משפ]ט‬
ֿ ‫ ִמ ְׁש ָ ּֽפט׃‬ἀλήθειαν
ἐξοίσει κρίσιν.
‫ולוא יכהה‬ ‫לא יכהה‬ ‫ ֤ל ֹא יִ ְכ ֶה ֙ה‬4 ἀναλάμψει καὶ
‫ולא ירוץ‬ a‫ וְ ֣ל ֹא יָ ֔רּוץ‬οὐ
‫ולוא ירוץ‬
θραυσθήσεται,
‫ ַעד־יָ ִ ׂ֥שים ָּב ָ ֖א ֶרץ ֯ע ֯ד י֯ [שים ] עד ישים בארץ‬ἕως ἂν θῇ ἐπὶ ἕως ἂν ἐκβάλῃ
‫ ] [ משפט‬vacat [ ] ‫ ִמ ְׁש ָ ּ֑פט‬τῆς γῆς κρίσιν· εἰς νῖκος τὴν
κρίσιν.
‫ולתורתיו‬ ‫ֹתור ֹ֖תו‬ ָ ‫ּול‬ ְ καὶ ἐπὶ τῷ νόμῳ καὶ τῷ ὀνόματι
‫איים ינחילו‬ ‫ ִא ִּי֥ים יְ יַ ֵ ֽחילּו׃‬αὐτοῦ ἔθνη αὐτοῦ ἔθνη
ἐλπιοῦσιν. ἐλπιοῦσιν.
There is a large number of differences between the reconstructed LXX

text and the NT.44 They differ substantially45 and this raises again the
44 Archer and Chirichigno calls this “a classic case of the independence of the LXX
on Matthew’s part” (OT Quotations, 113). Blomberg takes a similar position: “. . . Matthew
seems independent of the LXX and provides a more literal rendering of the Hebrew than
does the LXX” (“Matthew,” 43). Beaton rightly describes Matthew’s text form of this quota-
tion as distinct from both the MT and LXX (“Isaiah,” 71).
45 Blomberg aptly states: “Matthew’s Greek shares a couple of key words per line with
the LXX, but in most instances these are the most natural Greek words that anyone would
use to translate the Hebrew . . .” (“Matthew,” 43).
440 gert j. steyn
question about whether this is due to another Textvorlage underlying the

LXX version, or whether this is due to the NT author’s rewriting. The fol-
lowing differences are noted:
The identity and qualification of the child (son or servant, παῖς),
differs in the LXX and NT versions. In the LXX he is identified as Jacob
(Ιακωβ), and qualified as that (a) God “will help him” (ἀντιλήμψομαι
αὐτοῦ),46 that (b) he is “Israel, the elected one” of God (Ισραηλ47 ὁ ἐκλεκτός
μου),48 that (c) God’s very being (ἡ ψυχή) “accepted him” (προσεδέξατο
αὐτόν),49 that (d) God “gave” (ἔδωκα, aorist) his spirit upon him, and that
(e) “he will carry out” (ἐξοίσει,50 future) justice to the nations. The manner
in which this LXX child will operate is as follows: Firstly, his voice (nomi-
native case) will not be cried aloud (οὐ κεκράξεται,51 from κράζω) or
released (οὐδὲ ἀνήσει), nor will it be heard (οὐδὲ ἀκουσθήσεται, future
passive) outside (ἔξω). Secondly, metaphorically speaking, a reed that is
crushed (τεθλασμένον) he will not bruise (οὐ συντρίψει) and a wick that is
smoking (καπνιζόμενον) he will not quench, but in truth he will carry out
justice (ἀλλὰ εἰς ἀλήθειαν ἐξοίσει κρίσιν)—a phrase present in the Hebrew
but absent in Matthew’s version. The reconstructed LXX sets the term for
this action to be “until he will establish justice upon the earth” (θῇ ἐπὶ τῆς
γῆς). The phrase ἀναλάμψει52 καὶ οὐ θραυσθήσεται is part of the section that
is also present in the Hebrew, but lacks in Matthew. The phrase καὶ ἐπὶ
τῷ νόμῳ is closer to the Hebrew versions. The Greek codices and versions,
however, read here ὀνόματι—as Matthew does.
In the NT (as in the Hebrew versions), there is no direct indication in
the quotation itself about the identification of this child. Instead, where
the name of Jacob appears in the LXX, the NT reads the attention marker
ἰδού53 and the Hebrew versions ‫( ֵ ֤הן‬MT) or ‫( הנה‬DSS). Also where the
46 Some important LXX witnesses only read ἀντιλήμψομαι. These include Codices ‫א‬
A–Q* and B*, as well as minuscules 198 393 544 and 965.
47 Supported also by the 6th century Egypt Codex Q (Marchalianus).
48 It is interesting that the Boharic translation lacks the phrase ὁ ἐκλεκτός μου.
49 The Church Father Justin is closer to this reading with προσεδέξεται αὐτόν.
50 This is a more literal rendering of the Hebrew ‫ֹיוציא‬
֥ ִ (cf. Archer and Chirichigno, OT
Quotations, 113).
51 The Alexandrian group, Codex A + minuscule 106, read κράξεται and the minuscule
group 377–564–565 read οὐ κραυγάσει (the latter being closer to the NT reading).
52 The LXX minuscule group 377–564–565 read only ἀναλαμπει; minuscules 88 and 407
(followed by Justin) read ἀναληψει.
53 The LXX minuscule group oII (109–736) reads the same as the NT, whereas minus-
cule 88 adds ἰδου and minuscule 106 adds καὶ ἰδου to Ιακωβ.
name of Israel stands in the LXX,54 there is no such reference in the NT

or the Hebrew55 versions. Furthermore, Matthew’s version qualifies this
child (servant) to be (a) the one “whom (God) has chosen” (ὃν ᾑρέτισα),
that he is (b) “the beloved of God” (ὁ ἀγαπητός μου), (c) that God’s
very being (ἡ ψυχή) is “well pleased with him” (εἰς ὃν εὐδόκησεν),56 that
(d) God “will put” (θήσω, future) his spirit upon him, and that (e) “he
will proclaim” (ἀπαγγελεῖ, future) justice to the nations. The manner in
which the NT child will operate is as follows: Firstly, he will not quarrel
(οὐκ ἐρίσει)57 or shout (οὐδὲ κραυγάσει, from κραυγάζω), nor will anyone
(τις) hear (οὐδὲ ἀκούσει, future active) his voice (accusative case) in the
streets (ἐν ταῖς πλατείαις). Secondly, metaphorically speaking, a reed that
is bruised (συντετριμμένον),58 he will not break (οὐ κατεάξει) and a wick
that is smoldering (τυφόμενον)59 he will not quench—which is further
from the Hebrew than the LXX. Matthew’s version sets the term for this
action to be “until he brings justice to victory” (ἐκβάλῃ εἰς νῖκος)—which
is closer to the Hebrew versions. (See the quotation from Isa 7:14 in Matt
1:23 where the phrase τῷ ὀνόματι60 αὐτοῦ also occurred).
There is another interesting difference between our text versions with
the presence of an additional waw in 1QIsaa and its equivalent καί before
κρίσιν in the NT text—which is also supported by the Boharic translation
of the LXX. Both the MT and the reconstructed LXX, however, lack this
copulative.
Blomberg considers the differences between the Hebrew and the LXX
to be the result of the LXX that “translates, at times more paraphrasti-
cally than literally.”61 Archer and Chirichigno identify several “wrong” and
“accurate” translations of the Hebrew in the LXX, for example:
54 Except for the LXX minuscule group oII (109–736) that also lacks the name as is the
case with the Hebrew and the NT.
55 This was also stated in a Greek note in the marginal notes of both the 6th century
Codex Marchalianus and in those of the Syrohexaplaric translation.
56 This formulation in Matthew is closer to the Hebrew versions than to the recon-
structed LXX and is a good alternative for ‫ ָר ְצ ָ ֣תה‬.
57 This reading is also supported by the Boharic translation of the LXX and minuscule
534.
58 This NT reading is supported only by the 13th–14th century LXX minuscule 46.
59 The NT reading is also supported by the Church Fathers Justin and Cyril.
60 The absence of ἐπί here in the NT is also supported by the 6th century LXX wit-
ness Codex Marchalianus and by minuscule group 311 + 130. On the other hand, however,
the NT witness, Codex Washingtoniensis, minuscule 0233 and a few other witnesses do
include ἐπί here—which brings it closer to the Hebrew and LXX versions.
61 Blomberg, “Matthew,”, 43.
442 gert j. steyn
(i) ἐρίσει (“will contend/argue contentiously”) is a bit paraphrastic for ‫יִ ְצ ַ ֖עק‬
(“will cry aloud”); (ii) κραυγάσει is much more accurate for ‫ּׂשא‬ ֑ ָ ִ‫ י‬. . . ‫ֹקוֹלו‬
ֽ than
LXX’s ἀνήσει (“will desist”), which is definitely wrong; (iii) ἀκούσει τις is not
accurate for ‫יַ ְׁש ִ ֥מ ַיע‬, although it is closer in meaning perhaps than LXX’s
ἀκούσει (which seems to suggest that the messianic Servant is the one who
will not hear).62
The fundamental flaw with this form of argumentation is the assumption
that the reconstructed LXX text is a translation of the Hebrew texts at our
disposal today, that there was only one version of the LXX translation in
circulation and that it was intended to be a literal translation. Due to the
lack of physical evidence of a particular LXX version that might have been
closer to Matthew’s reading (and to the Hebrew) than that of the recon-
structed LXX text, the most appropriate summary of the situation might
be that of Hagner: “Matthew may well have formed the translation from
sources available to him (e.g. Targumim) or produced his original trans-
lation (so Davis and Allison) in order to suit his own purposes.”63 What
remains largely unexplained, though, is the large number of differences
between the Hebrew texts and the reconstructed LXX—especially in the
light of the other LXX passages quoted by Matthew that seem to be much
closer to the Hebrew texts known to us.
Case 5. Isa 6:9–10 in Matt 13:14–15

(1Q Isaa) 4Q60 MT Isa LXX Isa 6:9–10 Matt 13:14–15
(4QIsaf ) 6:9–10
‫שמעו‬ ‫ִׁש ְמ ֤עּו‬ Ἀκοῇ ἀκοῇ
‫שמוע‬ ‫ָׁש ֹ֨מ ַ ֙וע‬ ἀκούσετε ἀκούσετε
‫ועל תבינו‬ ‫ל־ּת ִ֔בינּו‬ ָ ‫וְ ַא‬ καὶ οὐ μὴ συνῆτε καὶ οὐ μὴ συνῆτε,
‫ראו‬ ‫ְּור ֥אּו‬ καὶ βλέποντες καὶ βλέποντες
‫ראו‬ ‫ָר ֹ֖או‬ βλέψετε βλέψετε
‫ועל תדעו‬ ‫ל־ּת ָ ֽדעּו׃‬ ֵ ‫וְ ַא‬ καὶ οὐ μὴ ἴδητε· καὶ οὐ μὴ ἴδητε.
‫ השם‬10 ‫מן‬֙ ֵ ‫ ַה ְׁש‬10 ἐπαχύνθη γὰρ ἐπαχύνθη γὰρ
‫לב‬ ‫ֵלב־‬ ἡ καρδία τοῦ ἡ καρδία τοῦ
‫העם הזה‬ ‫ָה ָע֣ם ַה ֶּ֔זה‬ λαοῦ τούτου, λαοῦ τούτου,
‫ואוזניו‬ ‫[וב]אזנו‬ ‫וְ ָאזְ ָנ֥יו‬ καὶ τοῖς ὠσὶν καὶ τοῖς ὠσὶν
‫הכבד‬ ‫ַה ְכ ֵ ּ֖בד‬ αὐτῶν βαρέως βαρέως
ἤκουσαν ἤκουσαν
62 Archer and Chirichigno, OT Quotations, 113.

63 Hagner, Matthew—quoted by Blomberg, “Matthew”, 43; contra Menken, Matthew’s
Bible, 67–88.
Case 5. (cont.)
(1Q Isaa) 4Q60 MT Isa LXX Isa 6:9–10 Matt 13:14–15
(4QIsaf ) 6:9–10
‫ועיניו‬ ‫ וְ ֵע ָינ֣יו‬καὶ τοὺς καὶ τοὺς
ὀφθαλμοὺς αὐτῶν ὀφθαλμοὺς αὐτῶν
‫השע‬ ‫ָה ַ ׁ֑שע‬ ἐκάμμυσαν, ἐκάμμυσαν,
‫פן יראה‬ ‫ֶּפן־יִ ְר ֶ֨אה‬ μήποτε ἴδωσι μήποτε ἴδωσιν
‫בעיניו‬ ‫ְב ֵע ָ֜יניו‬ τοῖς ὀφθαλμοῖς τοῖς ὀφθαλμοῖς
‫ובאוזניו‬ ‫ּוב ָאזְ ָנ֣יו‬
ְ καὶ τοῖς ὠσὶν καὶ τοῖς ὠσὶν
‫ישמעו‬ ‫ישמע‬ ‫יִ ְׁש ָ֗מע‬ ἀκούσωσι ἀκούσωσιν
‫בלבבו‬ ‫ובלבבו‬ a‫ּול ָב ֹ֥בו‬ ְ καὶ τῇ καρδίᾳ καὶ τῇ καρδίᾳ
‫יבין‬ ‫יבין‬ ‫יָ ִ ֛בין‬ συνῶσι συνῶσιν
‫ושב‬ ] ‫וש[ב‬ b‫ׁשב‬ ֖ ָ ָ‫ו‬ καὶ ἐπιστρέψωσι καὶ ἐπιστρέψωσιν
‫ורפא‬ ‫וְ ָ ֥ר ָפא‬ καὶ ἰάσομαι καὶ ἰάσομαι
‫לו‬ ‫ֹֽלו׃‬ αὐτούς. αὐτούς.
Apart from some differences between the Hebrew and the (reconstructed)
LXX,64 the latter translates the Hebrew more literally in general with Mat-
thew closely following the LXX65 version—if it is assumed that the LXX
translation was made from a Hebrew text similar to that at our disposal.
This passage was already referred to in a brief three line paraphrase in
Mark 4:10–12. The Markan version shows close similarities with the tex-
tual tradition of Targum Isaiah.66 Matthew, however, explicitly quotes
the passage from Isa 6:9–10 in a longer text form virtually identical to
that of the reconstructed LXX, except for the LXX which contains an addi-
tional αὐτῶν before βαρέως. Other LXX witnesses lack this word in Isa 6:10,
amongst which the original of Codex Sinatiticus.67 In the NT text of Matt
13:15, however, some NT witnesses amongst which, Codex Sinaiticus (!),
include it here.68 (Cf. also the version of Acts 28:27 in this regard).
64 Cf., for example, the translation of ‫ל־ּת ָ ֽדעּו‬

ֵ ‫ ַא‬with οὐ μὴ ἴδητε in the LXX (followed by
Matthew) instead of μὴ γνῶτε or εἰδῆτε (Archer and Chirichigno, OT Quotations, 93).
65 Blomberg, “Matthew,” 43. Beaton (“Isaiah,” 72–73) refers to the discussion of W. D.
Davies and D. C. Allison regarding the debate on the authenticity of Matt 13:14–15 (Mat-
thew II [ICC; Edinburgh: T & T Clark, 1991], 393–394).
66 Cf. Blomberg, “Matthew,” 47.
67 Other LXX witnesses that do not have αὐτῶν at this point, include minuscule 538 and
the Church Fathers Athanasius, Basilius, Theodorethus, Tertullian and Cyprian.
68 Others are Codex Ephraimi and minuscules 33. 892. 1241 pc it vgmss sys.c.p.
444 gert j. steyn
Case 6. Isa 62:11 + Zech 9:9 in Matt 21:5

(1Q Isaa) (1QIsab) MT Isa 62:11 LXX Isa 62:11 LXX Zech 9:9 Matt 21:5
‫אמורו‬ ‫̇א ̇מרו‬ ‫רּו‬
֙ ‫ִא ְמ‬ Εἴπατε Χαῖρε σφόδρα, εἴπατε
‫לבת‬ ‫לבת‬ ‫ְל ַבת‬ τῇ θυγατρὶ θύγατερ τῇ θυγατρὶ
‫ציון‬ ‫ציון‬ ‫־צֹּי֔ ון‬
ִ Σιων Σιων· Σιών·
‫הנה‬ ‫הנה‬ ‫ִה ֵּנ֥ה‬ Ἰδού κήρυσσε,
‫ישעך‬ ‫יש ֿע ֯ך‬ ̇ ‫יִ ְׁש ֵ ֖עְך‬ σοι ὁ σωτήρ θύγατερ
‫בא‬ ‫̇ב ̇א‬ ‫ָ ּ֑בא‬ παραγίνεται Ιερουσαλημ·
‫הנה‬ ‫̇הנה‬ ‫ִה ֵּנ֤ה‬ ἔχων τὸν ἰδοὺ ὁ ἰδοὺ ὁ
‫שכרו‬ ‫ֿש[כרו‬ ‫ְׂש ָכ ֹ֙רו‬ ἑαυτοῦ βασιλεύς βασιλεύς
‫אתו‬ ‫א]תו‬̇ ‫ִא ֹּ֔תו‬ μισθὸν σου σου
‫ופעלתיו‬ ‫ופעלתו‬ ‫ּופ ֻע ָּל ֹ֖תו‬ ְ καὶ τὸ ἔργον ἔρχεταί σοι, ἔρχεταί σοι
‫לפניו‬ ‫[לפניו‬ ‫ְל ָפ ָנֽיו׃‬ πρὸ δίκαιος καὶ
προσώπου σῴζων αὐτός,
αὐτοῦ. πραῢς καὶ πραῢς καὶ
ἐπιβεβηκὼς ἐπὶ ἐπιβεβηκὼς
ὑποζύγιον καὶ ἐπὶ ὄνον
πῶλον νέον . καὶ ἐπὶ
πῶλον υἱὸν
ὑποζυγίου .
This is a conflated quotation taken from passages from Isa 62:11 and Zech
9:9. Only the beginning of the quotation in Matt 21:5 shows similarities
with the reconstructed Isaiah LXX text form. The last two thirds do not
belong to Isaiah but to Zechariah. Neither the Isaiah part nor the Zecha-
riah part of the quotation appears in Mark or Luke. John’s gospel, how-
ever, provides a paraphrase of the Zechariah section in John 12:15.
Apart from two possible omissions in the latter part of the quotation,
the NT textual tradition shows no alternative traditions to the text as it
stands. Furthermore, none of the variant readings in the Isaiah LXX textual
tradition supports any part of the last two thirds of the NT quotation.
3. Conclusion
(a) There are hardly any differences regarding the Hebrew text form in
a comparison between the DSS and MT versions. Not only in the six
cases that were investigated above is this the case, but also in the
four cases where Matthew used Isaiah quotations from his synoptic
sources Mark and Q. This confirms again what scholarship noted
before, namely that the (2nd to 1st centuries B.C.E.) Isaiah Scrolls (in
particular 1QIsaa) represent a text that is basically the same as that
of (the 11th century C.E.) Codex Leningradensis. (The situation is dif-
ferent with 1QIsab and others). There are only very few differences
between these Hebrew versions in the cases investigated above.
(b) There are no indications that the differences with “the” reconstructed
LXX in these Matthean Isaiah quotations are necessarily due to the
section from Isaiah from which Matthew quoted in his SMt material.
This can be seen in cases 1 and 5 (traditionally known to be part of
“proto Isaiah”) where only one difference occurred, in contrast with
case 2 where more than ten differences were noted.
(c) There are also hardly any differences amongst the gospels themselves
with regard to these Isaiah quotations where they overlap in occur-
rence between the gospels. In a preliminary survey of those Isaiah
quotations that came from Markan or Q material, it became clear that
Matthew adopted the text form of those quotations as he found it in
his source material. Chances are thus good then that if he stayed close
to the text form of those quotations as found in his synoptic source
material, that he would probably do the same regarding the text form
of the source(s) he used for these Sondergut quotations.
(d) From the ten cases of Isaiah quotations occurring in Matthew’s gospel
(six in the SMt, three from Mark and one from Q), all six of those in
the Sondergut Matthäus, without exception, belong to the group of
“fulfillment quotations.” As these are only to be found in Matthew’s
Gospel, this raises the issue again about the possible use of a Testimo-
nium source for these quotations. In theory, the existence of a list that
contained the passages of the fulfillment quotations cannot be ruled
out completely. Lack of physical evidence and broader application by
Matthew calls, however, for great caution.
(e) There seems to be a tendency that the text form of the Isaiah quo-
tations in the SMt sections moves closer at a number of places to
the Hebrew versions (DSS, MT) and to some of the early transla-
tions of the LXX, such as the Syriac translations (Syro-Palestinian69
and Syro-Hexapla),70 as well as to the Egyptian Coptic translations
69 This is a Syriac, or more particularly, a western Aramaic translation that was made
of the LXX between the 4th and 6th centuries C.E.
70 This is a Syriac translation made by Paul of Tella from Origen’s 5th column of his
Hexapla (which Origen created during the 2nd to 3rd century during his terms in Alexandria
446 gert j. steyn
(especially in the Boharic71 dialect). Furthermore, there are clear traces

of Hexaplaric connections. This particular trajectory of the textual tra-
dition seems to be followed generally by the Isaiah commentaries of
Church Fathers such as Tertullian (from Carthage in the Roman prov-
ince of Africa during the 2nd to 3rd century) and Eusebius (Bishop of
Cesarea during the 4th century),72 in the six cases of this study. There
are also a few interesting cases where Codex Sinaiticus (with connec-
tions to both an Alexandrian and Hexaplaric text type) testified to a
text form closer to that of Matthew’s version. These commonalities
are pointing to another LXX text form73 than the reconstructed LXX
of today that was used by Matthew (and one which is also closer to
the Hebrew of the DSS and Codex Leningradensis). A synoptic over-
view of the differences and similarities between these texts confirms
not only that the Hebrew text form underlying the six quotations in
Matthew was close to that of the DSS and Codex Leningradenis, but
points also to a closer possible connection with the 5th column of
Origen’s Hexapla. This beckons for another investigation.
and Caesarea). The Syro-Hexapla is based on this revised LXX text from the so-called eccle-
siastical LXX text and probably made during the 7th century C.E. What makes this work
remarkable is that it contains the obeluses and asterisks from Origen’s 5th column.
71 A fairly literal translation of the LXX made probably during the 3rd century C.E.
72 Eusebius produced an independent edition of the 5th column of Origen’s Hexapla
with variant readings from other editions in the margin (F. G. Kenyon, Our Bible and the
Ancient Manuscripts [London: Eyre & Spottiswoode, 1895; 19394]. (Prepared for katapi.org
.uk by Paul Ingram, 2003).
73 Menken is thus correct in his judgment that Matthew used a form of the LXX, differ-
ent to the one known to us (Matthew’s Bible).
Die Rezeption von Jer 38:31–34 (LXX) in Hebräer 8–10
und dessen Funktion in der Argumentation des
Hebräerbriefes*
Wolfgang Kraus
1. Einleitung
Hebr 8:1f. beginnt mit einer erstaunlichen Aussage. Nachdem der Autor
bereits (nach unserer Zählung) sieben Kapitel teilweise nicht unkompli-
zierte Ausführungen vorgelegt hat, heißt es jetzt: κεφάλαιον δὲ ἐπὶ τοῖς
λεγομένοις. Er kommt also jetzt zu seiner Hauptsache. Der Leser stellt sich
erstaunt die Frage, ob alles Vorherige nur “Vorgeplänkel” war?
Der Begriff κεφάλαιον stellt einen rhetorischen terminus technicus dar1.
Er signalisiert, in einem auch sonst rhetorisch durchgefeilten Schreiben,
dass der eigentliche “main point” oder “big point” jetzt explizit zur Spra-
che kommt. Georg Gäbel spricht vom “leitenden Gesichtspunkt”2. Wir tun
gut daran, wenn wir uns von diesem rhetorischen Signal bei unserer Inter-
pretation leiten lassen und die Aussage in 8:1f. nicht nur als Vorgriff auf
Kap. 8–10 verstehen3, sondern sie auf den ganzen Brief beziehen. Denn
die Aussage in 8:1f. ist eingebettet in ein Netz von Hinweisen, angefangen
mit Hebr 1:3. Bereits dort heißt es von Jesus, er habe sich ἐν δεξίᾳ gesetzt.
Dahinter steht zweifellos Ps 109(110):1. In Hebr 1:13 wird dann explizit die-
ser Psalmvers zitiert. Auf die Aussage des Sitzens zur Rechten Gottes wird
jedoch mehrfach Bezug genommen: neben 1:3, 13 und 8:1 auch in Hebr
10:12 und 12:2. Es handelt sich somit um ein Leitthema.
* Mit diesem Beitrag grüße ich Prof. Dr. Dietrich-Alex Koch zu seinem 70. Geburtstag.
1 K. Backhaus, Der neue Bund und das Werden der Kirche. Die Diatheke-Deutung des
Hebräerbriefes im Rahmen der frühchristlichen Theologiegeschichte (NTA 29; Münster:
Aschendorff, 1996), 153; vgl. idem, Der Brief an die Hebräer (RNT Regensburg: Pustet, 2010),
z.St.
2 G. Gäbel, Die Kulttheologie des Hebräerbriefes. Eine exegetisch-religionsgeschichtliche
Studie (WUNT II.212; Tübingen: Mohr Siebeck, 2006), 240.
3 So allerdings in einigen Kommentaren zum Hebr: H. Hegermann, Der Brief an die
Hebräer (ThHK 16; Berlin: Evangelische Verlagsanstalt, 1988), 162; H.-F. Weiß, Der Brief an
die Hebräer (KEK 13; Göttingen: Vandenhoeck & Ruprecht, 1991), 428; M. Karrer, Der Brief
an die Hebräer. Kapitel 5,11–13,25 (ÖTK 20/II; Gütersloh: Gütersloher Verlagshaus, 2008), 101;
nicht ganz eindeutig H. Braun, An die Hebräer (HNT 14; Tübingen: Mohr Siebeck, 1984),
227.
448 wolfgang kraus
Worin besteht der leitende Gesichtspunkt? Folgende drei Elemente

haben nach 8:1f. Gewicht:
– die Glaubenden haben einen Hohenpriester

– dieser hat sich zur Rechten des Thrones der Majestät gesetzt
– dieser Hohepriester dient am himmlischen, das heißt am wirklichen,
wahrhaftigen Heiligtum.
Georg Gäbel hat die inhaltliche Bedeutung der Aussage sachgemäß so

zusammengefasst:
Nachdem Hebr in Kap. 7 den Eidschwur Ψ 109:4 ausgelegt hatte, verbindet er
in Kap. 8 die Hohepriesterthematik mit der sessio ad dexteram nach Ψ 109:1.
In 8:1f kommen die beiden Hauptstränge der Christologie des Hebr (und
die grundlegenden Belegstellen dafür aus der Schrift, Ψ 109:1.4) ausdrück-
lich zusammen; Hebr lehrt den zur rechten Hand Gottes inthronisierten
Christus des traditionellen Erniedrigungs- und Erhöhungsschemas als den
himmlischen Hohepriester zu verstehen. Die traditionelle Christologie und
ihre Erhöhungsaussagen werden durch die Hohepriester-Christologie kult-
theologisch interpretiert4.
Sehen wir uns die Struktur von Hebr 8:1–13 etwas näher an:
V.1f. bilden These und Ausgangspunkt.

V.3 führt aus, dass jeder Hohepriester etwas darbringen muss.
V.4 konstatiert: Wären wir auf der Erde, wäre die Stelle des Hohenpriesters
bereits besetzt und Jesus könnte dies nicht sein.
V.5: Das irdische Heiligtum ist allerdings nur ein Abglanz des himm
lischen, ganz so, wie es Mose in Ex 25 befohlen wurde.
V.6: Jesus ist Hoherpriester mit höherer Dignität. Sein Dienst beruht auf
einer besseren διαθήκη, deren Mittler er ist, weil diese mit besseren
Verheißungen verordnet wurde (νενομθέτηται—d.h. gesetzlich ange-
ordnet).
V.7 und V.13 sind Rahmungen des Jer-Zitats.
V.8–12 bieten dann das Zitat von Jer 38(31):31–34.
4 Gäbel, Kulttheologie, 240.

die rezeption von jer 38:31–34 (lxx) in hebräer 8–10 449
In V.6 findet sich die zweite Erwähnung des Stichwortes διαθήκη, das in
Hebr 7:22 überraschenderweise das erste Mal fällt. Es ist dort unerwartet
und überraschend, aber durchaus nicht unvorbereitet.
Mit der Aussage in V.13, dass, nachdem die neue διαθήκη eingesetzt
wurde, die alte διαθήκη folglich dem Verschwinden nahe ist, schließt das
Kap. ab.
In 9:1–10:14 folgen dann Ausführungen zur Art des Hohepriester-
Dienstes Jesu in überbietender Analogie zum Hohepriester-Dienst der
Aaroniden.
10:14 bietet den Zielsatz der Argumentation: Jesus hat durch sein einmali-
ges Opfer die Geheiligten zur Vollendung geführt.
10:15 enthält die Einführung des folgenden erneuten Zitats aus Jer 38 LXX
wobei hier der Hl. Geist als Zeuge für Gottes Rede auftritt.
10:16f. werden nochmals Jer 38:33f. LXX zitiert. Sie haben hier die Funk-
tion eines Schriftbeleges für die vollzogene Vergebung der Sünden und
Gesetzlosigkeiten.
10:18 bietet eine weiterführende Variation des Zielsatzes in 10:14 unter
dem Aspekt, dass dort, wo Sündenvergebung erfolgt ist, es keiner Opfer
mehr bedarf.
Das Zitat aus Jer 38(31) hat somit für den Argumentationsgang Hebr 8–10
fundamentale Bedeutung. Durch die Zitate in 8:8–12 und 10:16f. wird eine
inclusio gebildet, wobei in 10:16f. nur verkürzt zitiert wird: Die neue διαθήκη
bedeutet, dass die Gesetze in Herz und Sinn geschrieben werden und der
Sünden (Hebr ergänzt: und Gesetzlosigkeiten) nicht mehr gedacht wird.
Sebastian Fuhrmann hat die Aufnahme von Jer 38(31) im Hebr mit Recht
unter das Stichwort “Vergeben und Vergessen” gestellt5.
Bevor wir nach der Funktion der διαθήκη-Vorstellung im Hebr fragen,
geht es im folgenden Schritt zunächst um den Text von Jer 31(38) in seiner
hebräischen und griechischen Fassung.
5 S. Fuhrmann, Vergeben und Vergessen. Christologie und Neuer Bund im Hebräerbrief
(WMANT 113; Neukirchen-Vluyn: Neukirchener Verlag, 2007).
450 wolfgang kraus
2. Jer 31 (38):31–34 MT und LXX
2.1. Jer 31:31–34 MT 6
Walter Groß spricht mit Recht von der “Einzigartigkeit” dieses Textes:
Es handelt sich um den einzigen Beleg für das Syntagma “neue Berit” im
Alten Testament7.
Die zeitliche Einordnung ist in der Forschung umstritten. Nachdem
Siegfried Herrmann und Winfried Thiel den Abschnitt als literarisch ein-
heitlich angesehen und einer deuteronomistischen Schicht zugewiesen
hatten, wurde der Text durch Christoph Levin erheblich dekomponiert:
Auf ein kurzes frühexilisches Heilswort (V.31a und 34*) folgt nach Levin
als erste Erweiterung aus dem 5. Jh. die Hinzufügung der “neuen Berit,”
um dann eine zweite Erweiterung in spät-atl. Zeit zu erfahren, wo davon
die Rede sei, dass die Torah auf’s Herz geschrieben werde8. Nach Walter
Groß ist das Wort post-deuteronomistisch, was von Konrad Schmid im
Sinn von anti-deuteronomistisch zugespitzt wurde9.
Wir müssen diese Frage in unserem Zusammenhang nicht entschei-
den. Klar ist, dass es sich nicht um ein Heilswort des Propheten Jeremia
handelt. Nach meiner Meinung haben das Verständnis von Groß als post-
deuteronomistisch und das von Schmid als anti-deuteronomistisch viel
für sich.
6 Zum Folgenden einschlägig sind vor allem W. Groß, “Erneuerter oder neuer Bund,” in
Bund und Tora (WUNT 92; Hg. F. Avemarie und H. Lichtenberger; Tübingen: Mohr-Siebeck,
1996), 41–66; W. Groß, „Der neue Bund“, Theologische Qartalschrift 176 (1996): 259–272;
A. Schenker, Das Neue am neuen Bund und das Alte am alten (FRLANT 212; Göttingen: Van-
denhoeck & Ruprecht, 2006); H.-J. Stipp, “Die Perikope vom ‘Neuen Bund’ (Jer 31:31–34)
im masoretischen und alexandrinischen Jeremiabuch. Zu Adrian Schenkers These von der
‘Theologie der drei Bundesschlüsse’,” JNSL 35 (2009): 1–25; und A. Schenker, “Welche Argu-
mente wiegen schwerer auf der Waagschale? Zwei Weisen, die Textunterschiede in Jer
31:32–34 zu erklären,” JNSL 36 (2010): 113–124.
7 Groß, “Erneuerter oder neuer Bund,” 43.
8 S. Herrmann, Die prophetischen Heilserwartungen im Alten Testament (BWANT 85;
Stuttgart: Kohlhammer, 1965), 179ff.; W. Thiel, Die deuteronomistische Redaktion von Jere-
mia 25–45 (WMANT 52; Neukirchen-Vluyn: Neukirchener Verlag, 1981), 20ff.; C. Levin, Die
Verheißung des neuen Bundes in ihrem theologiegeschichtlichen Zusammenhang ausgelegt
(FRLANT 137; Göttingen: Vandenhoeck & Ruprecht, 1985), 30–32; K. Schmid, Buchgestal-
ten des Jeremiabuches. Untersuchungen zur Redaktions- und Rezeptionsgeschichte von Jer
30–33 im Kontext des Buches (WMANT 72; Neukirchen-Vluyn: Neukirchener Verlag, 1996),
302–304, 348, 372f.
9 Groß, “Erneuerter oder neuer Bund,” 49 Fn. 40; idem, „Der neue Bund“, 262 samt Fn. 15.
Grammatische Probleme liegen vor allem an zwei Stellen vor:
1. Wie ist der ‫אשׁר‬-Satz in V.32 zu verstehen?

2. Welches Verständnis von ‫ נתתי‬liegt in V.33 vor?
Zu 1: Was den ‫אשׁר‬-Satz angeht, so scheint es am wahrscheinlichsten,

dass es sich um einen Relativsatz zu ‫ ברית‬handelt, also um eine Parallele
zum vorausgehenden ‫אשׁר‬-Satz. Damit wäre zu übersetzen: “die sie ‫המה‬
gebrochen haben—meine ‫”ברית‬10.
Zu 2: Was ‫ נתתי‬angeht, so stellt es eine Perfektform dar und heißt nor-

malerweise: ich hatte/habe gegeben. Es wird häufig gedeutet als perfectum
propheticum, um zu der futurischen Form ‫ אכתבנה‬zu passen. Diese Deu-
tung beruht jedoch nicht auf syntaktischen, sondern auf theologischen
Kategorien, weshalb sie problematisch ist11.
Manche Ausleger rechnen—unter Bezug auf Peschitta, Vulgata und
Targum, die eine Zukunftsaussage bieten12—mit Textverderbnis und
gehen von einem ursprünglichen Waw-Perfekt ‫ ונתתי‬aus. Aus strukturel-
len Gründen entscheidet sich auch Groß hierfür13.
Schenker und andere halten am perfektischen Sinn fest und überset-
zen: Ich hatte/habe meine Torah in ihre Mitte gegeben14. Der Anschluss
des Folgesatzes muss dann adversativ erfolgen: Ich hatte (zwar) meine
Torah in ihre Mitte gegeben, aber ich werde sie (künftig) auf ihr Herz
schreiben ‫ועל־לבם‬.
“In ihre Mitte” hätte dann soziologische Bedeutung15. Die Neuerung
bestünde somit darin, dass die Torah künftig auf ’s Herz geschrieben wird,
so dass sie innerlich und nicht mehr nur äußerlich sein wird. Die dadurch
entstehende strukturelle Spannung innerhalb der Verse ist m.E. besser zu
ertragen als die Annahme von Textverderbnis oder die Übersetzung als
perfectum propheticum.
10 Groß, “Erneuerter oder neuer Bund,” 41 Fn. 2.

11 Ibid., 42 Fn. 3.
12 Vgl. Stipp, “Perikope,” 11.
13 Nach Stipp, ibid., ist diese Lesung nur als “Notlösung” zu werten und hat alle Wahr-
scheinlichkeit gegen sich.
14 Schenker, Bund, 30f.; Karin Finsterbusch, “Ich habe meine Tora in ihre Mitte gege-
ben.” Bemerkungen zu Jer 31,33 Biblische Zeitschrift 46 (2005): 86–92; Karrer, ÖTK 20/II,
116; vgl. daneben H. Tita, “ ‘Ich hatte meine Tora in ihre Mitte gegeben’: Das Gewicht einer
nicht berücksichtigten Perfektform in Jer. XXXI 33,” VT 52 (2002): 551–556.
15 Schenker, Bund, 28.
452 wolfgang kraus
Was sagt der Text inhaltlich?
V.31 beginnt mit einer positiven These: Es gibt eine neue ‫ברית‬. V.34
schließt mit einer Begründung, welche erst die “innere Voraussetzung”
dafür angibt, dass es zu einer solchen neuen Berit kommen kann16.
Die Zielaussage des ganzen Abschnitts lautet: “Ich werde ihrer Schuld
nicht mehr gedenken.”
Die Vv.32–34a benennen die Eigenart dieser ‫ברית‬, und zwar teilweise
in Antithese zur früheren, teilweise als Beschreibung der Kennzeichen der
neuen.
V.32 beginnt antithetisch zur früheren ‫ברית‬. Gemeint sind wohl alle Gene-
rationen seit dem Exodus. Damit entsteht ein “Schuldkontinuum von den
Ägypten-Tagen an bis zum Augenblick der Verheißung”17. Wie Walter
Groß richtig herausstellt, kennt das Jeremia-Buch “überhaupt keinen Zeit-
punkt, zu dem der mit den Vätern geschlossene Bund jemals gehalten
worden wäre”18. D.h. die neue ‫ ברית‬ist auch deshalb neu, weil die vorhe-
rige gebrochen wurde.
Die Frage, ob man besser von einer erneuerten oder einer neuen Berit
sprechen soll, lässt sich m.E. so beantworten: Es gibt einerseits Elemente,
die sich durchhalten: (1) die Adressaten, (2) die Torah, (3) das mit der sog.
Bundesformel umschriebene Gottesverhältnis. Insofern ließe sich zurecht
von einer erneuerten Berit sprechen19.
Schenker legt großes Gewicht darauf, dass die ‫ ברית‬von Seiten Gottes
nie gekündigt wurde20. Allerdings fordert die “nominale Formulierung
‚einen neuen Bund schließen’” und die Struktur von Jer 31:31–34, ande-
rerseits doch mit “neue Berit” zu übersetzen21. Demgegenüber stellt der
Exodusbund die bekannte Größe dar22.
Das bedeutet: Von Gott her nie gekündigt, vom Haus Israel und dem
Haus Juda aber gebrochen, bezieht sich die Ankündigung einer neue
Berit auf eine neue Initiative Gottes. Sie gipfelt darin, dass Gott die bisher

17 Groß, „Der neue Bund“, 200.
18 Ibid., 260 Fn. 5.
19 E. Zenger, Das erste Testament. Die jüdische Bibel und die Christen (Düsseldorf: Pat-
mos Verlag, 1991), 115.
20 Schenker, Bund, 37.
21 Groß, „Der neue Bund“, 261.
22 Ibid., 262.
äußerliche Torah ins Herz schreiben wird, so dass sie aus innerem Antrieb
heraus getan werden wird.
Damit ist eine Theologie des Lernens, wie sie uns aus dem Dtn bekannt
ist, nicht mehr vereinbar. Die Torah bleibt sich inhaltlich gleich, aber
ihre Aneignung erfolgt nicht mehr durch Lehren und Lernen, sondern
geschieht spontan, von innen heraus. Jer 31 kommt damit der Verkündi-
gung des neuen Herzens in Ez 36 nahe. Denn es geht hier wie dort um
die gleiche Frage, nämlich das grundsätzliche Problem der “Befähigung
Israels zum YHWH-Gehorsam” überhaupt23.
Voraussetzung zu dieser Befähigung ist die Vergebung der Schuld, d.h.
wir finden hier eine radikale gnadentheologische Antwort. Durch die
Torah im Herzen wird “jeder zukünftige Bundesbruch verhindert.”24 Auf
deuteronomistischem Hintergrund ist das kaum verständlich zu machen.
Die Frage, die im Text Jer 31:31–34 zur Debatte steht, lautet nicht wie
im deuteronomistischen Bereich: “Wie kann Israel nach einer solchen
Geschichte der Sünde Gott wieder dienen?”, sondern sie geht tiefer: “Wie
kann Israel allererst dazu befähigt werden?”25 Die Antwort lautet: Gott
schließt eine neue ‫ברית‬.
2.2. Jer 38:31–34 LXX

Kommen wir von hier aus zu Jer 38 LXX, so fallen sofort fünf relevante
Unterschiede zum MT auf, von denen mindestens zwei sachliches Gewicht
haben26.
1. Statt der idiomatischen Wendung ‫ כרת ברית‬liest die LXX διαθήσομαι

διαθήκην (V.31).
2. An Stelle von ‫( בעלתי בם‬V.32), “ich war ihr Meister,” steht in der LXX
ἠμέλησα αὐτῶν, “ich kümmerte mich nicht mehr um sie.” Als Vorlage
sind verschiedene Vokabeln diskutabel: ‫בעל‬, ‫ בחל‬II, ‫געל‬, aber auch
‫ מאס‬oder ‫ שׁלח‬in der Bedeutung “verwerfen.” Das Verbum ἀμελέω
begegnet insgesamt nur viermal in der LXX: neben Jer 38 noch Jer 4:17;
2 Makk 4:14; SapSal 3:10. Im Neuen Testament neben Hebr 8 noch in
Mt 22:5 und 1 Tim 4:14.
23 Ibid., 263.
25 Ibid., 62.
26 Zusammenstellung in Anlehnung an M. Meiser, Die Rezeption von Jer 38 (31) in
Hebr 8, Manuskript 1f.
454 wolfgang kraus
3. Statt des hebräischen Perfekt ‫ נתתי‬bietet die LXX eine Futurform,

und zwar als Doppelung διδοὺς δώσω. Gegenüber einer einfachen fini-
ten Verbform lässt dies auf eine anderslautende hebräische Vorlage
schließen.
4. Der MT bietet den suffigierten Singular ‫( תורתי‬V.33). Die LXX schreibt
pluralisch νόμους μου. Allerdings könnte dies auf den gleichen Kon-
sonantenbestand zurückgehen, wenn defective Schreibweise (‫תורתי‬
statt ‫ )תורותי‬vorausgesetzt wird. Dies ist auch in anderen Fällen als
Ursache divergierender Übersetzungsmöglichkeiten bekannt.
5. Der MT liest ‫( בקרבם‬V.33), was sowohl “in ihre Mitte” als auch “in ihr
Inneres” bedeuten könnte. Die LXX bietet εἰς τὴν διάνοιαν αὐτῶν. Das
kann nur in einer Richtung verstanden werden.
Von sachlichem Gewicht sind vor allem zwei Differenzen:
1. Nach Jer 38:32 LXX steht die Möglichkeit im Raum, dass nach dem
Bundesbruch durch die Menschen auch Gott selbst an seinem Volk
das Interesse verloren hat. Adrian Schenker spricht von einer Zeit ohne
Bund.27 Mindestens von einer Zeit der faktischen Unwirksamkeit des
Bundes wird man auszugehen haben. Das geht über die Aussage des
MT hinaus.
2. Jer 38:33 LXX rechnet mit einer neuen Willenskundgabe Gottes, bei der
nicht klar ist, ob es sich nur um Inhalte handeln wird, die bereits aus
der Torah vom Sinai her bekannt sind oder ob nicht auch neue Gesetze
dabei sein werden. Der MT hat dies nicht im Blick. Die Neuheit der
neuen ‫ ברית‬bezieht sich dort auf die Art der Vermittlung, nicht auf die
Inhalte.
Die Frage stellt sich, ob alle Änderungen als sekundär zu begreifen sind,
u.a. theologisch motiviert, oder ob sie anders erklärt werden müssen.
Dieser Frage nach der hebr. Vorlage von Jer 38 LXX ist intensiv Adrian
Schenker nachgegangen. Nach seiner Analyse hat der Übersetzer von
Jer 38 (31) eine andere hebr. Vorlage gelesen als MT (bzw. eine Vorstufe
desselben). Wichtiges Indiz dafür ist v.a. die ungriechische Konstruktion
διδοὺς δώσω, die eine hebr. figura etymologica im Hintergrund haben
27 Schenker, Bund, 24—dagegen Fuhrmann, Vergeben, 153f. Fn. 97 und 238 Fn. 37.
dürfte, nämlich finites Verb plus Inf.abs. (‫)נתון אתן‬28. Nimmt man die
grundsätzliche Einsicht hinzu, dass die Jeremia-Übersetzung, wo wir dies
erkennen können, ihre hebr. Vorlage im allgemeinen treu wiedergegeben
hat, müssen wir von einer anderen Textform ausgehen als der uns im MT
erhaltenen.
Über die Frage der zeitlichen Priorität von MT oder der hebr. Vorlage
der LXX ist damit noch nicht entschieden. Die Argumente, die Schenker
vorbringt, wonach der MT die jüngere Texttradition repräsentiert, haben
einige Plausibilität für sich, wenngleich sie durch Stipp Widerspruch
erfuhren29:
1. Die Aussage einer Vernachlässigung Israels durch Gott sucht man eher
zu vermeiden, als dass man sie sekundär einträgt.
2. Ein Hinweis auf die bereits erfolgte Willenskundgabe Gottes am Sinai
wird eher sekundär eingetragen als gewaltsam getilgt.
3. Das Konzept der einen Torah als Kontinuum zwischen proton und
eschaton wurde eher nachträglich eingebracht als später zerstört.
Und Schenker ging sogar noch einen Schritt weiter, und führte die der
LXX vorliegende hebr. Fassung auf den Propheten Jeremia selbst zurück.
Letztere Frage muss in unserem Kontext keinesfalls entschieden werden30.
Sebastian Fuhrmann ist der Argumentation von Schenker weitgehend
gefolgt31. Hermann-Josef Stipp hat Schenker widersprochen, insbeson-
dere seiner “Theologie der drei Bundesschlüsse,” hält allerdings manche
der Rekonstruktionen Schenkers für nachvollziehbar32.
Doch abgesehen von einer möglichen Autorschaft Jeremias, und abge-
sehen von der Frage, ob es sich bei der Fassung des MT um eine Fort-
schreibung einer früheren Form handeln sollte oder nicht33, unbestreitbar
bleibt, dass Jer 38:31–34 LXX eine eigenständige Vorstellung beinhaltet, die
neben dem MT Anspruch hat, Gehör zu finden. Deren Kennzeichen sind
die folgenden:
28 Im Fall von ‫ נתן‬findet sich eine solche figura etymologica siebenmal in der Hebrä-
ischen Bibel; s. Schenker, Bund, 33.
29 Ibid., 45ff., 49ff.; vgl. dagegen Stipp, “Perikope,” und die Replik von Schenker, “Argu-
mente.”
30 Karrer mahnt zur Vorsicht: Das sei “schwer zu verifizieren” (Karrer, ÖTK 20/II, 119).
31 Fuhrmann, Vergeben, 153f.
32 Schenker, Bund, 38; Stipp, “Perikope,” passim, Zusammenfassung 22f.
33 Letzteres scheint mir wahrscheinlicher als dass die LXX die Korrektur einer proto-
masoretischen Textform gewesen sein sollte; vgl. Schenker, Bund, 42 zu Torah im 2. Jh.
456 wolfgang kraus
1. Nicht nur die Väter haben wie im MT den Bund gebrochen, sondern
Gott selbst hat sich nicht mehr um sein früheres Volk gekümmert. Ob
das heißt, dass auch von Gott her kein Bund mehr existierte, lässt sich
nicht allein über die Bedeutung von ἀμελεῖν erschließen34, sondern nur
aus dem Gesamtkontext des Jer-Buches. Und hierbei scheinen die bei
Schenker genannten Stellen (Jer 11:1–14, bes. 2–5; 41:8–22 LXX; 38:35–37
LXX) doch dafür zu sprechen, dass keine διαθήκη mehr bestand. Die
Ankündigung einer neuen διαθήκη bedeutete daher tatsächlich einen
völligen Neuanfang.
2. Diese neue διαθήκη in Jer 38 LXX ist anderes konzipiert als im MT: V.33
heißt es, dass die Gesetze Gottes (nicht die Torah) in den Verstand
gegeben und auf die Herzen geschrieben werden. In welchem Verhält-
nis diese Gesetze zur Sinai-Torah stehen, muss offen bleiben35, aber
klar ist, dass die Formulierung auch für neue Gesetze Raum lässt.
3. Die Schärfe in der Aussage des Bundesbruches ist in der LXX gegen-
über dem MT größer. Die gnadenhafte Zuwendung Gottes über die
Geschichte des Bundesbruches hinweg wird damit aber nicht kleiner,
sondern nur noch größer und radikaler.
3. Jer 38 LXX in Hebr 8 und 10
Das Zitat im Hebr repräsentiert eindeutig eine LXX-Textform36, wenn-

gleich nicht eine solche, die mit der unserer kritischen Ausgaben in jeder
Hinsicht identisch ist. Knut Backhaus siedelt die vom Hebr verwendete
Textform in der “Nähe zur Textüberlieferung des Codex Alexandrinus”
an37. Neben kleineren Differenzen, wie der Verwendung von λέγει statt
φήσιν, haben einige Unterschiede zwischen dem Text in Jer 38 LXX und
dem Zitat im Hebr sachliches Gewicht38.
1. Statt der Formulierung der LXX in V.31 διαθήσομαι τῷ οἰκῷ Ισραηλ καὶ
τῷ οἰκῷ Ιουδα διαθήκην καινήν bietet Hebr 8:8: συντελέσω ἐπί. Also statt
διαθήσομαι mit Dativ haben wir συντελέσω ἐπί mit Akkusativ. (Anders
34 So aber Fuhrmann, Vergeben, 153f. Fn. 97; 238 Fn. 37.
35 Karrer, ÖTK 20/II, 118.
36 Ibid., 103; Fuhrmann, Vergeben, 153.
37 Backhaus, Bund, 171; Zustimmung bei Fuhrmann, Vergeben, 153, vgl. zur genaueren
Verortung der Textvorlage des Hebr in der Textgeschichte noch Karrer, ÖTK 20/II, 104.
38 Für eine genaue Auflistung s. Karrer, ÖTK 20/II, 103.
verhält es sich in Hebr 8:10, wo das διαθήσομαι mit Dativ der LXX sich
auch im Hebr findet.)
2. Statt des Verbums διαθήσομαι mit Dativ in V.32 der LXX bietet Hebr 8:9
ποιέω mit Dativ.
3. Bei der in Jer 38:33 LXX ungewöhnlichen grammatischen Konstruktion
διδοὺς δώσω bietet Hebr 8:10 das δώσω nicht, versteht also die erste
Hälfte als einen abhängigen Partizipialsatz: indem ich meine Gesetze
in ihren Verstand gebe, werde ich sie ihnen auch aufs Herz aufschrei-
ben (eingravieren).
4. Das γράψω in Jer 38 LXX wird zu ἐπιγράψω in Hebr 8.
5. Das αὐτῶν nach μικρού in V.34 wird in Hebr 8:11 nicht geboten.
In Hebr 10, wo Teile des Jer-Textes erneut zitiert werden, fallen zwei Dif-
ferenzen auf, die vermutlich aktualisierende Änderungen des Hebr-Autors
darstellen.
1. Bei der Wendung διαθήσομαι τῷ οἰκῷ Ισραηλ in Jer 38:33 ist der Dativ
τῷ Ισραηλ in Hebr 10:16 ersetzt durch πρὸς αὐτούς.
2. In das Zitat von τῶν ἁμαρτιῶν αὐτῶν οὐ μὴ μνησθήσομαι ist eingefügt:
καὶ τῶν ἀνομιῶν αὐτῶν. Hierdurch wird deutlich, worauf es dem Hebr-
Autor besonders ankommt.
Anders als in Hebr 10, wo es sich vermutlich um Aktualisierungen handelt,

die auf den Hebr-Autor zurückgehen, lassen fast alle Abweichungen im
Zitat in Hebr 8 nicht auf bewusste Änderungen durch den Autor schlie-
ßen, sondern sind durch Varianten in der Vorlage zu erklären.
Umstritten ist lediglich die Verwendung von συντελέσω, und zwar text-
geschichtlich sowie inhaltlich. An anderen Stellen, wo der Hebr-Autor
nicht zitiert, sondern selbständig formuliert (Hebr 2:10; 5:9; 7:19, 28; 9:9;
10:1, 14; 11:40; 12:23) verwendet er (als eines seiner Lieblingswörter) τελειοῦν
als Simplex und nicht als Kompositum. Im Vergleich mit Jer 41:8, 15 LXX;
Jes 61:8 und der Symmachus-Fassung von Jer 38:31 LXX ergibt sich nach
Martin Karrer jedoch zwingend, dass der Autor den Wechsel des Verbums
nicht selbst zu verantworten hat, sondern es aus der ihm vorliegenden
Textüberlieferung übernommen hat39.
39 Ibid., 103f.
458 wolfgang kraus
Inhaltlich bedeutet συντελέσω . . . διαθήκην καινήν: “eine neue Setzung

zum Vollzug bringen”40. Die Übersetzung mit “einen Bund vollendend
eingehen”41 ist m.E. zu stark von der Zweiseitigkeit eines Bundes her
gedacht. Die Übersetzung in Hebr 8:10: “einen Bund eingehen”42 διαθήσομαι
διαθήκην ist von daher ebenfalls nicht ganz problemfrei.
4. Zur Funktion der Vorstellung von der neuen διαθήκη im Hebr
Welche Funktion hat nun die διαθήκη-Vorstellung in der Argumentation

des Hebr? Ist sie, wofür Knut Backhaus eingetreten ist, der Versuch, in
einer “semantischen Sattelzeit” eine neue Selbstidentifikation zu ermög-
lichen? Hat sie also die Funktion, in einer Situation, in der frühere Aus-
einandersetzungen mit dem Judentum überwunden sind, die eigene,
christliche Identität neu zu beschreiben43?
Oder geht es bei der Bundesthematik um die “Artikulation jüdisch-
judenchristlicher Identitätskonkurrenz,” wofür Manuel Vogel eintritt44?
Der Autor wolle den christlichen Glauben als “legitime Neuinterpretation
der biblischen Tradition” verständlich machen und sich gegen “konkurrie-
rende jüdische Auffassungen der biblischen Religion” behaupten45. Vogel
setzt daher den Hebr früher an als Backhaus dies tut.
Oder geht es dem Hebr primär darum, anhand von Jer 38 LXX die
Bedeutung des “heilseffektiven Leidens und Todes Jesu” im Sinn von “Ver-
gebung” und “Amnestie” neu zu verdeutlichen, wie Sebastian Fuhrmann
dies vertreten hat46? “Mit der von Gott gewirkten, von Christus vermit-
telten Einsetzung des neuen Bundes geht ein göttliches Vergessen, eine
Amnestie einher”47.
Um die Frage nach der Funktion der διαθήκη-Vorstellung im Hebr zu
beantworten, müssen wir uns noch einmal den Gedankengang vergegen-
wärtigen. Der Ausgangspunkt lautet: Jesus sitzt zur Rechten Gottes (Hebr
40 Braun, HNT 14, 239.

41 Karrer, ÖTK 20/II, 98, 104.
42 Ibid., 98.
43 K. Backhaus, “Das Bundesmotiv in der frühchristlichen Schwellenzeit,” in Der spre-
chende Gott. Gesammelte Studien zum Hebräerbrief (WUNT 240; idem; Tübingen: Mohr
Siebeck, 2009), 153–173; 157f.
44 M. Vogel, Das Heil des Bundes. Bundestheologie im Frühjudentum und im frühen Chri-
stentum (TANZ 18; Tübingen / Basel: Francke Verlag, 1996), 318.
45 Ibid., 323.
46 Fuhrmann, Vergeben, 250.
47 Ibid., 226.
1:3, 13). Dies gehört zum grundlegenden Bekenntnis. Dahinter steht Ps

109:1. Dieser zur Rechten Gottes erhöhte Jesus, ist Hoherpriester. Es gibt,
wie Martin Karrer erneut herausgestellt hat, für die Vorstellung des Hohe-
priestertums Jesu keine traditionsgeschichtlich überzeugende Ableitung.
Es handelt sich um eine Invention des Hebr-Autors. Erschlossen wird sie
aus der Fortsetzung in Ps 109, nämlich aus V.4.
Das Hohepriestertum Jesu ist nach der Ordnung Melchisedeks einge-
setzt. Es steht damit in einer gewissen Konkurrenz zum aaronidischen—
und daneben jeglichem irdischen—Priestertum, ist diesem aber überlegen.
Das wird in Hebr 7 begründet anhand von Gen 14.
In Hebr 7:11–12 taucht erstmals die Vorstellung auf, dass mit der Einset-
zung Jesu zum Hohenpriester auch die Kultordnung sich ändert.48 Wenn
das Priestertum wechselt, wechselt auch die Ordnung (bzw. das Gesetz
oder die Verordnung) für Priesterschaft und Volk.
Diese Aufhebung der früheren Ordnung war nötig, weil sie sich als zu
schwach erwies (Hebr 7:18). Hebr 7:18 steht sachlich parallel zu 8:7, d.h.
bereits hier hat der Hebr-Autor den Text Jer 38 LXX im Blick, was sich in
V.22 dann auch begrifflich niederschlägt.
Hebr 7:19 spricht von der besseren Hoffnung, die mit der Aufhebung
der vorhergehenden ἐντολή gestiftet wird.
In Hebr 7:22 wird dies—erstmals—als κρεῖττον διαθήκη beschrieben,
deren “Bürge” (ἔγγυος) Jesus sei. Und zwar wird diese Aussage mit κατὰ
τοσοῦτο in direktem Anschluss an das Zitat von Ps 109:4 eingeführt. Jesus
ist “in demselben Maß”49 bzw. “dementsprechend”50 bzw. “in so großem
Mass”51 oder “accordingly”52 Bürge einer besseren διαθήκη geworden.
Daraus folgt: Die Einführung der διαθήκη-Vorsstellung hat zwingend
etwas mit Ps 109:4 zu tun. Wobei dort ja gar nicht “Hoherpriester” steht,
sondern nur “Priester nach der Ordnung Melchisedeks” begegnet. Dass die-
ser Priester Hoherpriester ist, ergibt sich erst aufgrund der gezera schawa
mit Gen 14 und deren Anwendung auf Jesus. Denn auch dort steht nicht
Hoherpriester, sondern Priester Gottes, des Höchsten. Wenn Melchisedek
allerdings von Abraham den Zehnten empfängt, muss er höher stehen als
alle von Abraham abstammenden Priester.
48 Ibid., 124.
49 So die Übersetzung bei Braun, HNT 14, z.St.
50 So die Übersetzung bei Weiß, KEK 13, z.St.
51 So die Übersetzung bei Karrer, ÖTK 20/II, z.St.
52 So die Übersetzung der New English Translation (Dallas: Bible Press, 2003), z.St.
460 wolfgang kraus
Die erstmalige Einführung des διαθήκη-Begriffes, der dann im Verlauf

der Argumentation noch weitere sechszehn Mal begegnet, erfolgt über-
raschend—aber nicht unvorbereitet. Διαθήκη steht hier in Parallele zu
ἐντολή, V.18, und zu νόμος in V.12 und V.19.
Wie ist hier διαθήκη zu verstehen: als “Bund”53, als “Heilsordnung”54, als
“Setzung”55 oder als “Verfügung”56? Wir lassen das noch einen Augenblick
offen.
Der Gedankengang wird in 7:28 zu dem Ziel geführt, dass zwar das
Gesetz eine Vielzahl schwacher Menschen zu Hohepriestern einsetzt, die
eidliche Versicherung jedoch, die zeitlich nach dem Gesetz ergangen ist,
den Sohn, der in Ewigkeit vollendet ist.
In Hebr 8:6 geht es dann um die darauf gegründete “bessere Leiturgia”
Jesu, der “auch” Mittler einer überlegenen διαθήκη geworden ist. Dies wird
in 8:7ff. mit Jer 38 LXX begründet.
Daraus ist zu folgern: Die Einführung des Jer-Zitats steht in unmit-
telbarem Zusammenhang mit dem Verständnis Jesu nach Ps 109:4 und
liefert die Begründung für die durch Jesus vermittelte bessere διαθήκη.
Die Ankündigung einer neuen διαθήκη in Jer 38 LXX bezog sich—so der
Hebr—auf die Einsetzung Jesu nach der Ordnung Melchisedeks. Sie ist
die Grundlage für Jesu überbietendes Handeln. Sie hat ihr Ziel in der voll-
kommenen Sündenvergebung (Hebr 10:18). Jer 38 LXX liefert dafür die
biblische Begründung.
Das Handeln Jesu am himmlischen Heiligtum wird in Hebr 9 in überbie-
tender Analogie zum Jom-Kippur-Ritus (Lev 16), dem Ritus der Roten Kuh
(Num 19) und dem Ritus des Bundesblutes (Ex 24) näher dargestellt.
In Hebr 9:15 wird die Reinigung des Gewissens von den toten Werken
als neue διαθήκη interpretiert, deren Mittler Jesus ist.
Hebr 9:16f. begründen rechtsmetaphorisch, dass Jesus nur aufgrund
seines Todes zu diesem Mittler werden konnte.
Hebr 9:18ff. führen aus, dass die Einweihung der διαθήκη nur durch Blut
erfolgen kann, analog und zugleich überbietend zu Ex 24.
53 Backhaus, Hebr, z.St.; Fuhrmann, Vergeben, 137–150; Karrer, ÖTK 20/II, z.St.
54 Weiß, KEK 13, 411–415; J. Frey, “Die alte und die neue διαθήκη nach dem Hebräer-
brief,” in Bund und Tora (WUNT 92; Hg. F. Avemarie und H. Lichtenberger; Tübingen:
Mohr Siebeck, 1996), 263–310: 267.
55 Braun, HNT 14, 217f.
56 So wird der Begriff in Septuaginta Deutsch im Bereich des Pentateuch wiederge
geben.
Hebr 10 schließlich begründet, warum das Kultgesetz keine Vollendung

bringt: es geht nämlich darum, den Willen Gottes zu tun. In diesem Wil-
len ist die Gemeinde geheiligt, ein für allemal. Und das bezeugt der Hl.
Geist, der genau diese Sündenvergebung mit Jer 38 LXX im Blick hatte.
Der Bogen schließt sich.
5. Fazit
Die Einführung der neuen (καινή) bzw. besseren (κρεῖττον, 7:22) bzw. fri-
schen (νεά, 12:24) διαθήκη dient
1. der Begründung des Hohepriestertums Jesu und

2. der inhaltlichen Erläuterung, dass mit Jesu Wirken vollständige Sün
denvergebung kommt.
Um die Anwendung von Ps 109:4 auf Jesus durch Schriftbeweis zu belegen,

wird Jer 38 LXX eingeführt, und um die soteriologische Qualität des Hohe-
priestertums Jesu zu erläutern fährt er fort: Jesus bringt als Hoherpriester
die angekündigte Sündenvergebung. Die hohepriesterliche Tätigkeit Jesu
wird dann durch weitere AT-Kontexte expliziert. Leitender Gesichtspunkt
in Hebr 8–10 (bzw. 7–10) ist jedoch nicht eine “Bundes”-Theologie, son-
dern Ps 109:4.
Die διαθήκη-Theologie des Hebr hat eine wichtige, aber begrenzte
Funktion. Jer 38 LXX ist zunächst und vor allem Schriftbeweis. Der Text
verheißt Sündenvergebung und begründet eine neue διαθήκη jenseits bzw.
nach der früheren, durch welche die levitische Priesterschaft eingesetzt
wurde. Der Autor des Hebr zieht Jer 38 LXX zunächst heran, um das Hohe-
priestertum Jesu als schriftgemäß zu begründen. Doch es wäre zu wenig,
hier stehen zu bleiben. Im Folgenden argumentiert der Hebr aufgrund des
Inhaltes von Jer 38 LXX, um die konkrete Füllung des Hohepriestertums
Jesu zu erläutern.
Die Argumentation bezüglich der neuen διαθήκη im Hebr verläuft nach
dem gleichen Muster wie die zu κατάπαυσις in Hebr 3:7–4:11: Gott hat
geschworen, sie werden nicht in die Ruhe kommen. Weil aber feststeht,
dass einige eingehen werden (4:6), lässt Gott zu einem späteren Zeitpunkt
durch David sagen: “heute, so ihr seine Stimme hört” (4:7). Wenn durch
Josua der Eingang in die Ruhe erfolgt wäre, würde nicht später von einem
anderen Tag geredet werden müssen (4:8).
462 wolfgang kraus
Der Beleg aus Jer 38 LXX stellt somit für den Autor des Hebr primär
einen Schriftbeweis dar. Eine “Bundes”-Theologie, die christliche Identität
beschreiben soll, liegt m.E. nicht im Interesse des Verfassers. Die Interpre-
tation des heilseffektiven Todes Jesu als vollkommene Sündenvergebung
aufgrund von Jer 38 LXX gehört zur inhaltlichen Konkretion aufgrund des
Schriftbeweises.
Dass die Argumentation des Hebr im weiteren Verlauf der Geschichte
des frühen Christentums dann zur Begründung eines neuen “christlichen”
Selbstverständnisses herangezogen werden konnte, und dass dieses dann
zunehmend in Absetzung vom Judentum formuliert wurde (vgl. Barnabas-
brief, Justin), gehört noch nicht in die Theologie des Auctors ad Hebraeos,
und im Grunde auch nicht in dessen Wirkungsgeschichte, denn der Hebr
war noch von einer anderen Vorstellung geleitet57. Vielmehr haben wir
hier von dessen Rezeptionsgeschichte in veränderter Zeit zu sprechen.
57 Es fällt auf, dass der Hebr zur Beschreibung “christlicher” Identität eine Begriff-
lichkeit verwendet, die hiermit nicht zusammenstimmt—das allerdings ist Thema einer
anderen Ausarbeitung, die ich demnächst vorzulegen beabsichtige. Die Unterstellung
von F. Crüsemann, Das Alte Testament als Wahrheitsraum des Neuen. Die neue Sicht der
christlichen Bibel (Gütersloh: Gütersloher Verlagshaus 2011), wonach die Interpretation von
Jer 31(38) durch A. Schenker u.a. zur “Wiedereinsetzung des traditionellen christlichen
Antijudaismus in sein (wenn auch etwas relativiertes) Recht” führe (174), ist allerdings
zurückzuweisen. Crüsemann redet einerseits die Differenzen zwischen Jer 31 MT und Jer
38 LXX klein (Analoges gilt für seine Darstellung des Hebr insgesamt, ibid. 174–179). Seine
Behauptung beruht andererseits nicht auf exegetischen Argumenten, sondern einer ideo-
logischen petitio principii, die eine andere Interpretation als die eigene unter das Verdikt
des Antijudaismus stellt (vgl. ibid., 150f.). Damit wird jegliche differenzierende Argumen-
tation von vornherein verdächtig, wenn sie nicht politisch korrekt erscheint. Hermeneuti-
scher Überlegungen, wie nach der Schoa mit bestimmten Bibeltexten (nicht nur mit deren
Auslegungsgeschichte) umzugehen sei, bedarf es dann gar nicht mehr.
INDEX of ancient sources
MT/LXX
Genesis 3:6 346–347

1 417 3:12 347
2–3 53 3:14 240
2:7 54, 168 7:1 419
2:24 211, 340 15:3 361
3:20 54 17:6 38
4 24 19:6 10–12
14 459 20:12–16 328
14:20 382 20:12 340
15:2 179 20:21 416
15:8 179 21:10 328
15:13 324–325 23:20 13, 16–17
19:3 65 23:21–23 10, 13
21:25 332 23:22 11, 13, 16
24:43 432 24 460
31:36 305 24:1–3 347
32:3 16 25 448
32:14 62, 64 25:40 304
32:19 62, 64 28:32 15
32:21 62, 64, 164 28:36 15
32:22 62, 64 29:9 11
33:10 62 32:1 304
34:7 361 32:4 63
35:18 58 32:20 291
36:28 158 35:19 11
36:33 400 36:38 15
41:2 62–63 37:2 65
41:4 62–63 39:30 15
41:5 62–63
41:7 62–63 Leviticus
41:18 62–63 1:2 64
41:20 62–63 2:1 62
41:34 266 2:4 62
43:11 62 2:5 62
43:15 62 2:5–6 62
43:25 62 2:6 62
43:26 62, 64 5:8 138
43:30 283 5:10 106
46:21 59 6:13 62
48:17 60 9:16 106
9:17 62
Exodus 14:10 62
2:2 63 16 460
2:8 432 18 120
3:2 177 20:6 345
464 index of ancient sources
20:9 345 29:22 256, 272

20:12 345 31 50
26:29 255 31:8 244
32 37, 46–47
Numbers 32:4 38, 41, 43–44
1:1 345 32:9 247
5:22 114 32:13 41
6:24–26 311 32:15 38–39, 43–44
11:8 291 32:18 38–39, 44
12:2–14 346 32:30–31 39
12:11 345 32:30 38, 44, 46, 49
15:6 62 32:31 38, 44, 46, 49
19 460 32:35 337, 339
22:32 63 32:37 38–39, 45–46
23:19 361 32:43 349
24:7 346–347 32:48 357
24:17 345, 347 33:10 9
24:24 211
27:18–23 17 Joshua
29:12–28 108 7:19 101
8:1 244
Deuteronomy 10:25 244
1:16–17 14 22:13 17
1:21 244 24:33 17
1:26 14
5:16–20 328 Judges
6:4–9 383 2:4 66
7:14 383 2:13 89
7:15 383 2:20 215
9:7 14 3 63–64, 66
9:10 347 3:7 89
9:21 291 3:12–30 56–58, 62–63,
9:23 14 65, 67
9:24 14 3:12 61, 65
17:1 67 3:14 61
17:5 343 3:15 58–60, 62,
17:8–9 19 65–66
17:9 9 3:16 58–60, 65, 67
18:15 346, 348 3:17 61–64
18:17 347–348 3:18 62
18:18 346–348 3:19–20 65
18:21 347–348 3:19 57, 61, 65–66
19:1 348 3:20 61, 65–66
19:7 19 3:21 59–60
23:7 120–121 3:22–23 67
23:21 13–14, 16, 19 3:22 58, 62
23:23 13, 16–17 3:23 57–58
24:14 328 3:24 61
24:16 311, 315–317, 3:26 57
341 3:28 62
27 115 3:29 62
28:5 383 5:14 58
28:48 262 6:29 66
28:53 255 8:1 66
29:17 347 8:3 66
index of ancient sources 465
9:3 66 16:11 58
9:30 66 19:17 58
10:6 89 19:38 (39) 91
10:9 58 22:3 38, 47
10:10 89 22:10 336
11:10 66 22:32 38, 47
11:11 66 22:47 38, 47
11:28 66 23 232
11:37 66 23:3 38, 47
12:15 155 23:8–39 232
13:12 66 23:8 231
13:17 66 23:11 232
16:4 235
16:16 66 1 Kings
16:21 291, 293 (3 Kingdoms)
18:7 66 1–2 228, 230
18:10 66 1–2:11 LXX 236
18:26 66 1:1–7 226
19:14 58 1:2 236
19:19 66 1:8 231, 236
19:24 66 1:9 236
20 58 1:10 236
20:7 66 1:11 236
20:9 66 1:19 236
20:15 59 1:35 236
20:16 59 2:1 235
21 58 2:3 233, 236
21:11 66 2:4 236
2:5 236
1 Samuel (1 Kingdoms) 2:8 58
1–2 Kgdms 11:2 227 2:10 235–236
2:2 38, 47 2:12–21:43 LXX 227
9:27–10:1 66 2:12 LXX 229
14:6 266 2:22 236
17:11 244 2:32 78
25:29 383 3 236
3:4–15 409
2 Samuel (2 Kingdoms) 3:6–9 422
1:22–27 LXX 341 3:26 283
4 91 4:29–34 409
4:2 92, 94 5:3 62–63
4:4 LXX 94 6:7 78
4:5 LXX 93 7:13–8:21 409
4:7 LXX 94 18:19 LXX 89
4:8 LXX 94 18:25 LXX 89
10–3 Kgdms 2 79 18:27 259
11–3 Kgdms 2:11 227 19:18 89–90
11:2–3 Kgdms 2:11 228–229 21:43 LXX 229
15 82 22–4 Kgdms 25 79, 227, 229
15:2 77, 82–86 22:31 78
15:5–6 82, 84, 86
15:5 83–84 2 Kings
15:6 77, 83–86 (4 Kingdoms)
15:7 86 5:23 78
15:10 77, 83–84, 86 6:8–9 87
6:8 87
6:9 87–88 1 Esdras
14:6 311, 315–317, 341 1 104
17:15 163–164 1:4 119
18:27 78 1:13 106
19:18 304 1:15 104
21:3 LXX 89–90 1:17 108
21:5 78 1:18 121
23:10 LXX 303 1:22 121
23:12 78 1:39 101
1:40 121
1 Chronicles 1:51 101
11:9 118 1:54–55 107
12:4 232 1:54 106
16:36 115 1:55 121
16:40 108 2:7 125
17:7 118 2:15 102, 104, 121
17:24 118 2:16 102, 104
22:5 235–236 2:21 102–103
22:13 244 3–4 110, 122
23:31 108 3:1–5:6 122
28:9 262 3:4 122
28:10 266 3:6 123
28:20 244 4:3 123
4:7–9 123
2 Chronicles 4:14 123
1:8–10 422 4:37 123
2:3 108 4:51 123
15:2 262 4:52 123
15:4 262 5:46 99–100
15:15 262 5:48 107
16:13 236 5:49 107–108
17:2 89 5:50–51 108
17:7–9 20 5:50 105
19:7 266 5:57 101
20:15 244 5:70 121
20:17 244 6:3 121
24:20–22 24 6:5 103–104, 118–119
25:4 105, 311, 315–317, 6:8 118–119
341 6:9 112
31:3 108 6:10 103
32:7 244 6:18 105
34:32 45 6:28–29 110
35 104 6:28 110
35:3 119 6:29 105, 110
35:15 104 6:31 109
35:17 108 7:4–5 116
35:18 121 8:3–4 112
36:7 101 8:5–6 115
36:18 101 8:5 115–116
36:21 106–107, 121 8:11 101
8:16 112 7:20 110–111, 113

8:18 111, 114 7:23 LXX 116
8:24 113 7:26 112–113
8:73 121 9:5–15 120
8:76 106, 108 9:7 121
8:78 106 9:8 106
8:82 120–121 9:11–12 120
9 115 9:12 120–121
9:8 100–101 10:4 106, 266
9:12 121 10:11 101
9:13 106
9:45–47 114 Nehemiah (2 Esdras)
9:46 117 5:13 115
9:47 115 7:72 100
8 115, 117
Ezra (2 Esdras) 8:1–13 98
1–10 98 8:4 232
1:1 LXX 115 8:5–6 (18:5–6) 114
3:1 99–100 8:6 (18:6) 117
3:2 107 10:35 105
3:3 107 10:37 105
3:4–5 108
3:4 105, 108 Esther
3:8 LXX 115 5:10 100, 109
3:10 101 6:12 100, 109
4:5 121 10:2 132
4:7 121, 154 (LXX)
4:8 101–102 Judith
4:9 101–102 4:13 374
4:17 101–102 5 361
4:18 LXX 154 5:3 362
4:24 LXX 115 5:5–21 362
5:5 102–103, 118–119, 5:8 362
154 (LXX) 5:12 362
5:8 112, 118–119 5:23 362
5:9 103, 118 7:28 361
5:13 LXX 115 7:30 362
5:15 105 8:4 374
6:3 LXX 115 8:5 374, 376
6:8–9 109–110 8:6 374
6:8 110–111 8:8 374
6:9 105, 110 8:16 361
6:11 109 9:2 361–362
6:14 LXX 116 9:4 374
6:15 LXX 115 9:7 361
6:22 45 9:9 374
7:4 101 9:11 362
7:6 112 10:3 374
7:7–8 115–116 (LXX) 10:21 362
7:8 LXX 115 11:23 63
7:18 111 13:6 360
13:9 362 2:24–3:1 135

13:15 362 2:24 134–135
13:20 114 3:6 142
15:11 375–376 3:18 132
15:12 360 4:2 135
16:2 361 4:9–10 136
16:7 374 4:9 136
16:13 362 4:15–20 137
16:19 362 4:26 127
16:26 375–376 5–18 127
5:2 127
1 Maccabees 5:4 127
1:54 209 6:19 304
3:29 154 10:8 137
5:56 132 11:2 138
7:33 18 11:4 129
8:2 132 11:18 137
9:22 132 11:22 132
10:15 132 13:25 132
14:28 18 14:13–16 139
16:23 132 14:15 139–141
14:16 139–141
2 Maccabees 15:9 132
2:17 11–12 16:5 142
4:14 453 16:12 143
4:17 453 17:1 127
6:1 127
6:9 127 Psalms
6:18–7:42 127 1–78 198
6:18 127 2:7 350
6:23 63, 127 5:8 LXX 390
7:41 127 8:6–8 422
10:38 413 8:7 336 (LXX), 431
10:1 LXX 336
3 Maccabees 11 (10):6 286–287
3:30 304 12:6 LXX 413
13:3 LXX 312, 352
4 Maccabees 14 45, 390 (LXX)
1:1 127 14:2 45
1:5–6 134–135 14:4 45
1:5 134–135 14:7 45
1:7–9 131 15:8 LXX 323
1:7 127 17:47 382
1:8 132 18 (17):3 38, 49
1:9 132 18 (17):32 38, 49
1:10 132 18 (17):47 38, 49
1:13 127 19 (18):15 38, 49
1:20 132–133 20:3 LXX 388–389
1:28 133 23:1 LXX 340
1:30 127 24:18 LXX 386
2:6 152 26:1 LXX 390
2:8–9 133–134 28 (27):1 38, 49
2:8 134 30:3 LXX 393
2:9 133–134 30:4 LXX 393
31:1 LXX 390 89 (88):27 38, 49

31 (30):3 38, 49 89:53 114
33:6 LXX 390 90:1–3 LXX 394
33:9 LXX 390 90:1–2 LXX 394
34:1 LXX 385–386 90:11 LXX 330–332
34:6 LXX 386 91:11 331
39:6 164 92:5 LXX 383, 394
39:7 LXX 312, 352 92 (91):16 38, 49
40 431 93:18 LXX 383
41:14 114 94:7 LXX 337
42–83 45 94:11 164
43:4 45 94 (93):22 38, 49
44:8 45 95 431
45:2 LXX 387–388 95 (94):1 38, 49
48:15 45 96:11 LXX 390
49 (48):15 38, 49 104:29 168
50:7 45 106:48 114
51:16 45 108:1–2 LXX 385–386
53 45 108:26 LXX 383
53:6 LXX 383 109 (110):1 447–448, 459
56:2 LXX 413 109:4 448, 459–461
59:6 45 110 (109):1 335
60:4 LXX 393 111:9 LXX 342–343
62 (61):3 38, 49 114:7 LXX 413
62 (61):7 38, 49 117:20 LXX 389–390, 393
62 (61):8 38, 49 118:86 LXX 383
62:10 164 118:117 LXX 383
62:11 163 119:26 266
65:20 382 119:120 285
67:7 45 119 (118):53 286–287
67:26 LXX 432 120 LXX 392
67:36 382 120:5–8 LXX 391
68:9 45 120:5 LXX 392
68:10 LXX 350 120:6 LXX 392
68:14–20 LXX 349 120:7–8 LXX 391–393
69:6 LXX 383 120:7 LXX 392
71 (70):3 38, 49 120:8 LXX 393
72 (71):4 153 121:1 LXX 393
72:19 114 122:7 LXX 382–383
73:4 62 128:6 382
73 (72):26 38, 49 135 LXX 394
74:2 247 135:25 LXX 383–384
77:11 LXX 413 144 (143):1 38, 44
78:5 44 144:4 164
78:9 LXX 383
78 (77):35 38, 49 Odes
78:71 247 2:35 337, 339
80:5 45
80:15 45 Proverbs
80:20 45 1:5 152–153
83:11 LXX 393 1:10 152
84:9 45 1:21 153
88:10 LXX 349–350 1:23 152
89:9 45 1:24 151
1:26 151 23:23 153

1:31 151 24 147
2:20 149 24:8 153
3:11 152 25:16 152
3:12 332–333 25:17 152
3:16 149 25:21 404
3:30 152 25:22 337
3:31 152 25:25 154
3:34 340 26:12 149
4:5 152–153 26:16 154
4:6 152 26:26 153
4:7 153 27:16 168
4:14 152 27:22 153
4:27 152 28:16 153
5:4 149 30:4 168
5:7 152 30:10 152
5:20 152 30:27 153
6:3 149 31:7 152
6:4 152 31:13 266
6:7 152 31:23 153
6:11 150 31:30 164
6:23 383
6:25 152 Qohelet
7:13 150 1:2 161, 163, 171
7:24 152 1:11 167, 169
8:3 153 1:13 159
8:8 153 1:14 167–168
8:22 153 1:17 167
8:32 152 2:1 167
10:7 383 2:7 167
11:13 153 2:8 167
11:29 168 2:11 168
12:25 154 2:12 167
13:16 266 2:14 167
14:22 149 2:15 167
15:22 153 2:17 167–168
15:27 153 2:18 167
15:32 153 2:19 167
16:16 153 2:21 167
16:25 149 2:23 167
17:16 153 2:24 167
18:5 14 2:26 167–168
18:15 153 3:10 167
18:17 153 3:11 167
19:8 153 3:13 167
20:14 153 3:17 167
21:6 164–165 3:18–19 162
21:25 260 4:1 167
22:9 149 4:2 167
22:10 153 4:3 167
22:25 152 4:4 167–168
23:3 149 4:6 168
23:7 149 4:8 167
23:9 152 4:11 167
4:14 167 Job

4:15 167 1–2 397
4:16 167–168 1 154
5:3 167 1:3 403
5:6 167 1:4 404
5:9 167 1:5 152
5:15 167 1:21 402
5:16 167 2 154
5:18 167 2:5 149, 154
6:3 167 2:7 406
6:5 167 2:8 405
6:7 167, 169 2:9–10 402
6:9 167 2:9 147, 151, 154–155,
7:6 167 399, 405–406
7:14 167 2:10 406
7:15 167 2:11–13 407
7:18 167 2:11 399
7:20 328, 330 3:4 152
7:21 167 3:6 152
7:22 167 3:7 152
7:26 167 3:9 152
7:29 167 3:11 151
8:8 167 3:18 154
8:9 167 3:20 151
8:10 167 4:17 151
8:12 167 5:8 149, 179
8:14 167 6:5 151
8:15 167 6:11 151
8:16 167 6:17 150
8:17 167 6:18 152
9:1 167 6:21 149
9:2 163 6:22 151–152
9:3 167 6:28 152
9:6 167 7:5 155
9:9 163, 167 7:11 149, 151
9:11 167 7:16 164
9:12 167 7:17 151
9:13 167 7:21 152, 165
9:15 167 8:11 154
10:3 167 8:13 150
10:19 167 8:16 155
10:20 167 9:5 256, 272
11:2 167 9:18 152
11:5 167 9:19 149
11:7 167 9:22 150
11:9–10 169 9:29 151, 164–165
12:3 291 10:4 150
12:5 167 10:19 151
12:8 171 11:12 150
12:9 167 11–14 399
12:14 167 12:7 152
13:2 150
Song of Songs (Canticles) 13:3 149
6:8 432 13:9 150
13:15 150 30:4 152

13:18 14 30:9 152
13:19 151 30:24 150
14 154 31 403
14:4 151 31:10 293
14:21 152 31:14 151
15–21 399 31:31 403
15:2 168 32–37 399
15:7 151 32:2 150
15:9 151 32:3 150, 232
15:10 150–151 32:6 150
15:14 151 32:10 150
16:3 151, 168 32:21 149
16:18 152 33:1 149
16:19 327, 329 33:9 150
17:10 149 34:10 150
17:14 155 34:14–15 168
18:4 151 34:24 155
19 154 34:34 150
19:23 151 35:7 150
20:17 152 35:16 165
21:4 151 36:12 150
21:16 155 36:14 150
21:25 150 36:22 151
21:26 155 37:13 152
21:34 164 37:24 150
22–31 399 38:2 408
22:3 151 38:3 408
22:10 151 39:7 154
22:12 150, 155 39:11 152
23:6 152 40:3 152
23:7 152 40:7 408
23:20 152 40:8 150, 259
24:1 152 40:21 154
24:15 152 41:2 151
24:19 152 41:3 327
24:22 151 42 154
24:28 152 42:3 151
25:2 151 42:5 152
25:3 152 42:6 150
25:6 152, 155 42:7–17 397
25:7 151 42:9 150
26:13 152 42:17 29, 147, 151,
26:14 150 398–399
27:8 151
27:12 163 Wisdom of Solomon
28 154, 397 1–5 413, 424
28:24 155 1:1 423
29–31 403, 408 2:12 424
29 403 2:23 417
29:6 403 3:5 413
29:22 150 3:7–8 422
30 403 3:8 413
30:1 152, 403 3:10 453
30:2 150–151 3:11 424
5:1–23 423 17:2–4 422

5:5 423 17:26 328
5:16 423 24:10 9
6:1 423 34:18 108
6:4 423 34:19 108
6:7 14, 413 35:1 108
6:11 423 35:5 108
6:17–20 424 38:11 108
6:21–25 420 45:17 9
6:21 416 46:16 108
6:25 423 50 7, 9
7:1–6 409, 419, 423 50:13 108
7:7–12 409 50:14 9, 108
7:7 417 51:30 32
7:22 413
8:10–18 416 Hosea
8:10–15 409 2:1 188–189
8:10–12 420 8:4 306
8:13 423 11:8 283
8:14–15 420 11:10 259
8:17 416, 423
8:21 417 Amos
9:1–3 417 1–2 297
9:2–3 422 1:2 259
9:2 423 2:9 303
9:7 409 3:4 259
9:8 409 3:7 303
9:10–12 420 3:8 259
9:10 417 4:1 297
9:12 409 4:7 303
9:13–18 423 4:9 303
9:17 417 5:25–27 306
10 409 5:25–26 298
10:2 423 5:26 297–301, 304–308
10:14 423 7:1 297
11:5 413 7:3 260
11:13 413 7:6 260
12:15 413 7:7 297
12:16 413 7:14 297
12:18 413 8:1 297
14:3 413 9:11 215
14:15–21 419
14:20 419 Micha
15:1 413 1:7 306
15:3 423 3:12 311, 317, 319, 341
16:2 413 4:1 264
16:11 413
16:24 413 Joel
17:2 413 2:13 260
18:4 423 2:14 260
4:16 259
Ben Sira
4:1 328 Jonah
5:11 320–321 2:1 263
14:11 108 2:2 263
2:9 164 5:16 179

3:9 260 6:1 179
3:10 260 6:8 179
4:2 260 6:9–10 430–431, 442–443
6:9 340
Nahum 6:10 443
1:6 263 6:11 177, 179
2:6 264 6:12 183
3:19 263 7:7 181
7:11 176, 179
Habakkuk 7:13 180, 183
1:5 341–342 7:14 177, 179, 186–188,
1:12 38, 48 430–433, 441
1:16 62–63 7:17 183
2:17 306 7:20 179
3:12 263 8:5 175
8:7 177, 179
Zephaniah 8:8 179, 186–188
2:11 262 8:10 180, 186
3:3 259 8:17 183
8:18 183
Haggai 8:23–9:1 430, 433, 435
2:4 266 9:1–2 433
9:7 177
Zechariah 9:8 (7) 179
3:1–3 400 9:11 (10) 180, 183
5:1–5 383 9:17 180
8:14 260 10:12 179
9:9 430, 444 10:20 183
10:2 164 10:21 175, 179
11:16 62–63 10:22–23 187–188
13:2 240 10:22 188
10:23 183–184, 189–190
Malachi 10:24 183–184
2:7 16 10:26 183–184
3:21 336 10:33 179
11:2 183
Isaiah 11:3 183
1:9 187–190 11:11 179
1:10 183 12:2 175, 179–180
1:24 179, 260 13:6 180, 183
2:2 183 13:19 179
2:3 179 14:2 183–184
3:1 179 14:3 183–184
3:15–18 176–177, 180–181 14:5 183–184
3:15–17 181 14:13 179
3:15 176–177, 181 14:27 183–184
3:16 176 16:14 180
3:17 176–177, 181 17:6 176, 179–180, 184
3:18 176–177, 181 17:10 38, 47–48,
4:2 183 179–180
4:4 179 17:16 180
4:5 180 19:6 154
5:13 180 21–66 175
21:6 179 37:19 180, 183, 304

21:8 179 37:20 176, 178, 180,
21:9 180 183–184
21:10 176, 179 37:21 176, 180, 184
21:16 177, 179 37:22 183
21:17 176, 179 38:5 180, 184
22:5 181 38:7 183
22:12 181 38:11 185
22:14 182 38:14 179
23:16 180 38:16 179
24:2 179 38:20 183
24:15 176, 180, 184 38:22 183
24:16 180 40:3 334
24:21 183 40:7 183
25:1 176, 179–180, 40:18 175, 180
184 40:27 183
25:9 179–180 40:28 176, 183
25:10 183 40:31 183
26:4 38, 48, 183 41–45 185
26:9 180 41:4 183
26:12 180, 184 41:13 176, 184–185
26:13 176, 179–180, 184, 41:14 183
328 41:16 180
27:1 183 41:17 180, 185
27:3 180 41:21 180
27:4 180 41:23 183
28:2 177, 179 42:1–5 343
28:13 180, 184 42:1–4 430, 438–439
28:16 177, 181 42:5 180, 185
28:21 180 42:6 180, 184
28:22 177, 181 42:8 180, 184
29:13 179 42:10 175
30:7 164 42:12 183
30:9 183 42:13 180, 184
30:15 177 42:17 183
30:18 180, 183–184 42:19 183
30:19 177 42:21 180, 184
30:20 179 42:24 183
30:29 38, 48, 183 43:1 180, 185–186
30:32 180 43:10 180
31:1 183 43:11 180
33:5 183 43:12 180
33:22 183 43:14 180
35:2 176 43:15 180
36:7 176 44:2 180
36:9 181 44:5 183
36:12 179 44:6 183
36:15 183 44:8 38, 48, 175
36:18 183 44:10 175
36:20 183 44:15 175, 183
37:4 176, 180, 184 44:23 183
37:6 180 44:24 185
37:14 180 45:1 180
37:16 176 45:3 180
45:5 180 61:1 177–178, 180–182

45:6 180 61:2 180, 183
45:7 180 61:8 457
45:11 180 61:10 183
45:14 175 61:11 178, 181–182
45:15 175 62:5 183
45:20 175 62:11 430, 444
45:21 175 63:7 180
45:23 180 63:10 14
45:24 180 63:17 247
45:25 180 64:2 180
46:6 175 65:1 262
46:9 175 65:13 181
47:2 291 65:15 181
48:1 176 65:23 183
48:17 176, 183, 185
49:4 164–165, 176 Jeremiah
49:5 176 1:1 244, 268, 271
49:7 177, 180 1:2 244
49:14 178–179, 182 1:6 179, 240
49:22 181 1:8 244
50:1–10 182 1:9 269
50:4 181–182 1:17 244–245,
50:5 14, 178, 181–182 251–252, 272
50:7 181–182 1:19 244
50:9 181–182 2:5 163–164
51:7 244 2:7 266
51:13 183 2:15 259
51:15 176, 184–186 2:20 241
51:20 184 2:21 269
51:22 176, 180 2:30 269
52:4 181 3:12 240
52:12 180 3:14 266
53:1 180 4:4 269
53:4 180, 430, 436 4:6 264, 266
53:5 437 4:8–10 269
54:5 180, 183 4:10 179, 240
54:6 178 4:19 263, 269
54:13 183 4:22 258
55:5 176, 180, 185 4:26 269
55:6 183, 262 4:28 260, 268
55:7 180 5:3 269
56:8 181 5:9 269
57:17 180 5:14–17 269
57:21 180, 185 5:15 266
58:5 180 5:29 269
58:6 180 6 269
58:8 183 6:2 241
58:9 183 6:9–15 245
58:11 183 6:9 245
58:13 183 6:10 245
59:13 180 6:11 245–246, 270
60:9 176, 180, 186 6:13 240
6:19 266 13:13 269

7:1 248 13:14 256
7:14–16 269 13:17 247, 250, 269,
7:20 269 272
7:33 269 13:24–27 269
8:1–3 269 13:24 254
8:3 253 13:27 268
8:6 260 14:11 269
8:10 269 14:13 240
8:17 269 14:15–18 269
8:19–9:1/2 269 14:17 247, 250, 272
8:19 164 14:22 164
8:21 247 15:1–9 269
8:23 246 15:6–7 251
9:1 247 15:6 246, 260
9:4 246 15:7 251, 254, 272
9:6 246 15:8 266
9:8 246, 269 15:10–21 251
9:9 246–247, 15:11 179
250–251, 272 15:13 269
9:10 269 15:17 250–252, 272
9:11 269 15:18 251, 263, 272
9:14 269 16:3–5 269
9:15 253, 269 16:9 269
9:16 245 16:13 252, 269–270
9:17 247, 250 16:15 253–254, 273
9:20 269 16:17 269
9:21 269 16:18 269
10:12 269 16:19 164
10:14–16 247 17:5 250
10:16 247–248, 270 17:9 263
10:18 248, 270 17:16 263
10:25 246 17:27 269
11:1–14 456 18:1 248, 271
11:1 248, 271 18:2 250, 255, 270
11:2–5 456 18:8 260
11:6 240 18:10 260
11:8 266 18:11 250
11:11 266, 269 18:17 254, 269
11:14 269 19:1–13 255
11:15 268 19:2 240
11:17 269 19:3 269
11:18 249, 255, 270 19:7–13 269
11:19–20 249 19:7 269
11:22 250, 269 19:9 250, 255, 271–272
11:23 266 19:11 269
12:2 269 19:15 269
12:5 246, 266 20:4–6 269
12:7–13 269 20:9 246
12:7 269 20:16 256, 260, 271
12:12 267, 272 21:1–10 256
13:11 269 21:1 248, 271
13:12 250, 273 21:2 256
21:4–14 269 28:14 262

21:4–7 269 29–52 260
21:4 256 29:1 240
21:7 256–257, 261, 273 29:4 254
21:14 269 29:7 254
22:5–7 269 29:10 (25:11) 214
22:12 254, 257, 271–272 29:13–14 262
22:22 269 29:14 253–254, 262,
22:24–26 269 272
22:30 250 29:16 250
23:3 253 29:17 250
23:8 253 29:18 253
23:15 269 29:23 259
23:16 163 29:25 250
23:19 263, 269 29:29 240
23:20 268 30 (37):1 248, 268
23:23 257, 272 30:2 259
23:24 257 30:10 244
23:39 269 30:11 253
24:1 264, 267 30:12–17 263
24:6 269 30:12 263, 272
24:8–10 269 30:23 263, 265,
24:9 253 268–269, 272
24:10 261 30:24 263, 268
25–36 (32–43) 259 31 (38) 449, 453–454,
25:3 269 462
25:6 258, 270 31:10 254
25:9–11 269 31 (38):15 260
25:29 258 31 (38):19 260
25:30 258–260, 272 31:20 263, 272
25:31 259, 261, 271 31:28 258
25:32 263 31:31–34 452–453
26–51 243 31:31 450, 452
26:2 259 31:32–34 452
26 (33):3 260 31:32 451–452
26:7 240 31 (38):33 449, 451
26:8 240 31:34 450, 452
26:11 240 31:37 250
26 (33):13 260 32 264
26:16 240 32:1 248, 267, 271
26 (33):18 311, 317–319, 341 32:2 264
26 (33):19 260 32:17 240, 259
26:21 272 32:24 264, 267
27:4 (34:3) 268 32:27 264, 267, 319
27:6 261 (LXX)
27:8 261, 270–271 32:28 264, 267, 273
27:9 (34:7) 240 32:37 253
27:10 261–262, 271 33:17 250
27:15 254, 261, 271 33:20 250
27:21 250 33:25 250
28:1 240 34:1 248
28 (35):6 115 34:2 264, 267, 273
28:13 262, 271 34:8 248
35:1 248 49:35 264, 267, 270–271

35:19 250 49:37 (25:17) 242
36:2 259 49:38 (25:18) 269
36:4 259 50:12 267
36:6 240 50:17–20 267
36:7 263, 265, 50:20 267, 270
268–269, 272 50:24 241
36:8 240 50 (27):25 263, 269
36:10 240 51:11 263, 265,
36:13 240 268–269, 272
36:14 240 51:12 268
36:15 240 51 (28):15 269
36:16 259 51:17–19 247
36:21 240 51 (28):17 269
36:23 240 51:19 247, 270
36:26 265, 270 51 (28):25 269
37 (44):17 268 51 (28):39 269
38 LXX 457–462 51 (28):57 269
38:3 264 51:59 244 (LXX), 268,
38:16 265 271
38 (31):31–34 447–448, 450, 51:61 240
453, 455 51:63 240
38:31 LXX 453, 456–457
38:32 LXX 453–454, 457 Baruch
38:33 LXX 454, 456–457 2:3 255
38:34 LXX 457
38:35–37 LXX 456 Lamentations
39:17 264–265, 271 1:2 276
40:1 248, 271 1:7 276
40:2–5 265 1:8 276
40:2 265 1:9 276
40:3 265–266, 271 1:14 293
40:12 253 1:15 292
41:8–22 LXX 456 1:17 276
41:8 LXX 457 1:18 292
41:15 LXX 457 1:21 276
42 (49):10 260 2:4 276
42:12 261, 266, 271 2:8 293
43:5 253 2:9 276
43:10 261, 267, 270 2:21 292
43:12 261, 267, 270 3:8 276
43:13 267 3:12 276
44 (51):10 260 3:53 276
44 (51):26 268 4:3 276
46:10 267, 272 4:4 276, 288
46 (26):15 269 4:5 288
46:27 244 4:8 287–289
46:28 253 4:9 288
48 (31):32 247 4:10 288
48 (31):36 269 4:15 276
49:10 (29:11) 269 4:17 164
49:12 264, 267 4:19 (18) 276
49:15 267 4:21 158, 276
5 281, 294–295 18:20 200, 205

5:3 276 18:21 197, 202–203
5:5 294 18:22 195
5:7 276 18:24 196–197
5:8 276 18:27 202
5:9 294–295 18:28 195
5:10 281–289, 294–295 18:32 195
5:13 281–282, 290, 19:1 200–201
292–295 19:2 200–201
5:14 292–293 19:3 200
5:15 292–293 20:1 200
20:2 200–201
Ezekiel 20:3 200–201
1–39 198 20:4 200–201
2:6 244 21:36 263
3:3 263 22:2 204
3:9 244 22:3 204
7 205 22:6 204
7:19 263 22:9 204
7:20 307 22:12 204
9:9 204 22:24 263
12:22 195, 197–198 22:25 259
12:24 198 22:27 204
14:1 199–200 22:31 263
14:2 199–200 23:45 204
14:14 396 24:6 204
14:20 396 24:9 204
14:22 260 30:9 264
16:6 204 33 202
16:7 307 33:14 202–203, 205
16:9 204 33:16 202–203
16:38 204 33:19 202
17:6 262 34:3 62–63
18 193–196, 202, 34:20 62–63
205 36–39 194, 205
18:2 195, 197–198 36 453
18:4 193, 197 36:23–38 194, 205
18:5 202–203, 205 36:23–30 194
18:7 195 39:28 262
18:8 195, 205 45 202
18:9 193, 197 45:9 202
18:10–20 199
18:10 193, 197, 199–200, Susanna
203–205 1:7 63
18:13 204
18:14 200 Daniel
18:16 195 1–7 209, 211
18:17 199–200 1:2 124–125
18:18 195, 199–200, 202 1:5 123
18:19 198–200, 202–203, 1:8 124
205 1:15 62
1:18 105 9:11 215

2:2 105 9:12 214
2:4 207 9:13 215
2:5 109 9:23 213, 215
2:15 213 9:24–27 215
2:46 105 9:24–25 216
2:47 123 9:24 215
3 208 9:25 213, 215–216
3:19 105 9:26 216
3:20 105 9:27 216
3:22 (28) 213 10:1 215
3:29 (3:96) 109 10:4 211
3:95 LXX 382 10:20 211
4–6 208–209 11 208, 216
5:2 116 11:5–40 211
5:7 213 11:6–40 211
5:11 207 11:14 212, 215
6:13 213 11:29 212
7 207, 210 11:30 211
7:7 210 11:31 209
7:13 209–210 11:33 215, 217
8–12 207, 209, 211, 215, 11:34 217
217 11:35 213, 217
8 212 11:40 217
8:4 212 11:41 217
8:9 212 11:43 211, 217
8:10–12 212 11:45 217
8:11 213 12 216, 218–219
8:14 213 12:1 218
8:16 213 12:2 218
9 214–216, 218–219 12:3 217–218
9:2 213–214 12:4 213
9:4 214 12:6 213, 219
9:6 214 12:9 213, 219
9:10 214 12:13 27, 219
OT Pseudepigrapha
1 Enoch par. 98 15
1:9 322, 339 par. 128–169 6
par. 188–294 415
2 Baruch
14:18 422 Pseudo-Philo Bib. Ant.
8:8 406
Jubilees
31:15 9 Testament of Job
1:5 406
Letter of Aristeas 1:6 401
par. 30 3 4:8 401
par. 96–99 15 4:9 401
7:10 404 4QIsaa 431

8:2 402
8:3 402 4QIsab 176, 178, 181,
9–15 403 438–439
9:2–3 403
9:3 403 4QIsac 186
9:5 403
13 404 4QIsad 185
15 404
20:1–3 402 4QIsaf 442–443
21:1–2 405
24:1–3 405 4QIsal 431
27:2–6 402
28:2–4 407 4QIsam 178, 182
28:5 407
32:1 407 4QJerb 237
33:1 407
37:3 402 4QJerd 237
37:4 402
40:4 405 4QLXXLeva 21
42:1–3 408
42:2 408 4QLXXLevb 21
42:3 407
46–50 408 4QLXXNum 21
47:5 408
47:9 408 4QSama 78, 86, 91–92,
94
Dead Sea Scrolls
1Q21 12 4QSamc 86
1QIsaa 176–178, 181–185, 5QLama 281–282, 290

343–344, 431–433,
436, 438–439, 7QpapLXXEx 21
441–445
8ḤevXIIgr 21, 179
1QIsab (1Q8) 178, 181–182, 185,
436, 444–445 Aramaic Levi 12
1QM CD A
2:1–3 18 VII, 14–15 299, 307
VII, 17 299, 307
1QS
3:18 422
Jewish Authors
Aristobulus 7.346 231

Fr. 4 322 11.326 15
Fr. 6 322 11.329 17
11.331 16
Josephus
Antiquities (Ant.) Bellum Judaicum
1.322 305 (B. J.)
3.178 15 5.235 15
Contra Apion 2.1–7 420

(C. Ap.) 2.2 416
1.7 23 2.4–5 422
1.38 160 2.12–16 425
2.194 17 2.24 413
2.41 414
Philo 2.48 413
Abraham 2.65 422
3–5 422 2.67 413
2.88 413
Agriculture 2.134 413
78 413 2.205 413
2.238 413
Cherubim 2.256 414
99 413 2.259 414
Congr. Names
177 333 7 416
Creation Posterity
88 422 14 416
Decalogue QE
53 414 2.44 413
61 413
155 414 Sacrifices
8–10 418
Heir
19 419 Spec. Laws
1.12 414
Life of Moses 1.18 413
1.1–4 411 4.164 415
1.1–2 425 4.186–188 415
1.148–162 411
1.149 416 Worse
1.150 415 162 419
1.151 414, 416
1.152–153 416 Targums
1.154–158 416 Tg. Onq.
1.154 416 Exod 19:6 12
1.155 416
1.158–162 422 Tg. Neof.
1.158 413, 416, 418–419, Exod 19:6 12
421
1.159 419, 422 Tg. Ps.-J.
1.160–161 421 Exod 19:6 12
1.162 422 Exod 29:6 15
1.198 414
1.199 414 Tg. Job
1.272 413 2:9 406
1.334 411
Rabbinic Writings
Abot de R. Nat. Lev. Rab.

1.4 161 28:1 161
b. Baba Batra Qoh. Rab.

146 23 1:1 158
b. Sab Lam. Rab.

30b 161 5:10 289
5:13 293
m. Yad
3.5 160 Soṭah
10a 293
Gen. Rab.
19:12 406
76:9 406
New Testament
Matthew Mark
1 34 1:3 334
1:18–25 187 4:10–12 443
1:18 187 10:19 328
1:21 187, 432 11:10 327, 329
1:23 186–187, 430–433, 12:36 335–336
441 13:14 216
2:6 341
3:1 346 Luke
3:3 334 1:31 432–433
3:17 350 1:34 433
4:15–16 430, 433, 435 3:4 334
5:1 346 3:22 350
5:3 346 4:8 331
5:4 346 4:10 330–332
5:11 346 18:28 100
5:17 346 20:42 335–336
8:17 430, 436
8:23–27 349 John
8:24 350 1:23 334
12:18–21 430, 438–439 2:14–17 350
13:14–15 430, 442–443 2:17 350
13:15 443 12:15 444
21:5 430, 444 19:27 100
21:9 327
21:12 350 Acts
22:5 453 2:25 323
22:32 347 2:34 335
22:44 335–336 2:35 336
23:35 24 3:13 347
25:6 435 3:22 348
7:6 324 Ephesians

7:32 347 1:3 382
7:37 348 5:31 340
7:42 306 6:2 340
7:43 302, 307
13:33 350 1 Timothy
13:41 341–342 4:14 453
17:28 322 5:18 340
21:6 100
21:9 433 2 Timothy
26:25–27 340 2:19 328–329
26:25 340
26:27 340 Hebrews
28:25–28 340 1:3 447, 459
28:27 443 1:6 354
1:13 335–336, 447,
Romans 459
3:10 328, 330 2:10 457
3:13–18 312 3:7–4:11 461
9 190 4:6–8 337
9:1–25 188–189 4:6 337, 461
9:1 190 4:7 337, 461
9:3 190 4:8 337, 461
9:5 188, 190 5:5 350
9:6–8 189 5:9 457
9:6 189 7–10 461
9:7 189 7 448, 459
9:8 189 7:11–12 459
9:26–33 188 7:12 460
9:26 188–189 7:18 459–460
9:27–29 187 7:19 457, 459–460
9:27–28 187–188 7:22 449, 459, 461
9:28 183, 189 7:28 457, 460
9:29 117, 187–190 8–10 447, 449, 461
11:4 89–90 8 448, 453, 457
11:35 327 8:1–13 448
12 338 8:1 447–448
12:16 337 8:3 448
12:19–21 337 8:4 448
12:19 337, 339 8:5 448
8:6 448–449, 460
1 Corinthians 8:7 448, 459–460
2:9 340 8:8–12 448–449
10:26 340 8:8 456
15:25 335 8:9 457
8:10 457–458
2 Corinthians 8:11 457
1:3 382 8:13 448–449
6:18 117 9 460
9:9 342–343 9:1–10:14 449
9:9 457 James

9:15 460 1:19 320–321
9:16 460 4:5 340–341
9:18 460 4:6 340
10 457, 461 5:4 117
10:1 457
10:5 312 1 Peter
10:12 447 1:3 382
10:14 449, 457 2:9 11
10:15 449
10:16 449, 457 Jude
10:18 449, 460 14 322, 339
10:30 337, 339
11:40 457 Revelation
12:2 447 3:19 333
12:6 332–333 22:16 345, 347
12:15 347
12:23 457
12:24 461
Early Christian Writings
1 Clement Epistulae
36:5 335–336 17, 2 369
56:4 332–333 22 377
53 297
Ambrosius 106 306
Explanatio Psalmorum
1.8 28 Praef. Chron. 76
Barnabbas Praef. Esth.

9:2 335 1–2 371
9:3 334
12:10 335–336 Praef. Iudit
15:4 334 1–8 365
1 367–368
Eusebius 2–3 363, 367, 369
Praep. Ev. 3–4 367
XIII 12, 6 322 5–6 367
5 366–367
Eustathius 6–7 371
Commentarii ad 6 372
Homeri Odysseam 7–8 371
1.390.9 141 7 369
8 369
Jerome 9–12 365, 373
Adv. Lucif. 9–10 374
29 30 9 366–367, 373–374
10 366–367, 373–374
Comm. Eccl. 160, 170 11 374
12 374
Comm. proph. min. 297, 301–302, 306
Praef. Tobit 8–9 367, 369

1 367 11 368
2–3 367
3 367 Julian
5 367 Epistulae
8–11 370 84 30
Greek and Latin Writings
Aristotle 72.19–20 415

Politics 72.22–23 418
1279a16 416 73.23–28 421
1284a3–14 417 73.28–74.1 421
1284b30–31 417 74.10–12 417
1295a9–22 418 74.15–19 421
74.20–21 415
Aratus 75.8–9 415
Phaenomena 75.9 412
1 322 75.12–13 412
5 322 75.14–15 412
75.15–16 415
Diodorus Siculus 80.1–4 417
Bib. Hist. 80.4–5 415
40.2 8 80.5 412
40.3 7 80.23 413
81.1 413
Diogenes Laertius 81.2 412, 414
Lives 81.6 413
2.110 409 81.8 412
3.58 409 81.10–12 418
4.14 409 81.10 412
5.22 409 81.26–82.3 415
5.42 409 82.28–83.7 421
5.45 409 187.12–13 415
5.47 409 188.5 412
5.49 409 188.7 413
5.59 409 188.9 413
6.16 409 188.10–11 412, 414
6.18 409
7.36 409 Herodotus
7.87–89 414 Histories
7.175 409 II, 86 305
7.178 409 III, 88 305
10.28 409
72.6 413 Homer
72.14 413 Iliad
2.204–205 414
Hellenistic
Pythogorean Texts Plato
33.8–10 422 Laws
71.21–22 422 712b 139
72.16–19 417 716 415
Republic Sophocles
415 417 Oed. col.
473 416 219 139
500 419
504–505 416 Thucydides
Histories
Theaetetus 2.80.5 153
176 415 2.80.6 153
Timaeus Xenophon
90 415–416 Hist. Gr.
5.2.17 153
Plutarch
Theseus Cyr.
24.2 153 5.1.24–25 414
Seneca
Clem.
1.7.1 415
1.19.2 414
1.19.8–9 415
INDEX of subjects
Alexandria(n) 3–7, 19, 22, 24–30, Corbeiensis 375

32–34, 72, 74, 76, Curetonianus 353
154–155, 165, 199, Ephraemi (C) 166, 327, 329, 443
211, 412, 424–425, Hamburg Pap. 165–167
427, 440, 445–446 Leningradensis 37–41, 49, 176,
Antiochus IV 127, 137, 207, 212, 281, 285, 316,
217–218 445–446
Aquila (α’) 10, 90, 98, Marchalianus (Q) 319, 432, 434,
166–167, 277, 302, 440–441
307, 314, 384–385, Sarravianus 343
391, 432, 436 Sinaiticus (ℵ/S) 28–32, 71–72,
Asterisk 146, 151, 195–196, 74, 128, 133, 138,
201, 231, 343–344, 166, 183–184, 319,
399, 446 321–324, 329–336,
340–342, 346,
Biblical quotation(s) 21, 69, 76–77, 351, 353–354, 384,
89–90, 95, 121, 431–432, 434–435,
190, 243, 293, 299, 443, 446
302, 306–307, Usserianus 353
311–314, 316–332, Vaticanus (B) 22, 24–29, 31,
334–335, 337, 69–76, 79, 81–82,
339–344, 346–348, 88, 90–91, 94–96,
350–354, 356–357, 165–167, 184,
369, 383–384, 193–195, 197, 200,
386, 427–432, 434, 202–205, 229–230,
439, 441, 444–446, 291, 317, 319,
448–449, 456–457, 322–323, 329–334,
459–460 339–340, 342, 351,
353–354, 432
Canon 21, 23–29, 33, 71, Venetus (N/V) 72, 128, 166, 319,
157, 160–161, 165, 329, 333
170, 224, 322, Washingtoniensis 301, 441
332, 365, 372 Coptic 75, 131, 165, 195,
Codex 292, 395, 445
Aleppo 176 Bohairic 75, 195, 440–441,
Alexandrinus (A) 22, 28, 31–33, 70, 446
72, 75–76, 128, Sahidic 75, 95–96, 284,
133, 166, 184, 195, 292, 388
200, 230, 291,
317, 319, 322–324, Daughter translation(s) 69, 73–74, 95, 284
329–334, 336, 340, Dead Sea scrolls 21, 37, 299, 354,
342, 346–347, 427, 429, 431,
353–354, 384, 432, 434–437, 440,
435, 440, 456 444–446
Ambrosianus (F) 344–349, 353–354 Double entendre 56, 61–62, 64–67
Bezae (D05) 325, 353, 432 Double translation 97, 99–104, 107,
Chludov 349–350 117–120, 122,
Coislianus (M) 72 284–285, 288
490 index of subjects
God 9, 13–19, 23, 27, Jesus Christ 24, 27–30, 34,

30, 32, 38–40, 189–190, 209,
43–47, 54, 56–58, 348–350, 373,
62, 65–66, 88–89, 388–389, 392,
101–102, 111–114, 427, 430–431, 433,
116–120, 123–125, 447–449, 458–462
127, 136, 143,
159–160, 162, 168, Kaige 21, 72–74, 79–83,
175, 185, 188–189, 85–96, 224,
207, 214–215, 228–231, 234, 236,
218–219, 233, 275–277, 311, 320,
237–242, 244–247, 336
249–252, 254–255, Ketib/Qere 89, 176–178, 191,
257–258, 260–273, 353
297–298, 300–303,
307–308, 322, 361, Lucian 74–83, 90, 93
374, 376, 381–383, Lucianic version 72–82, 85–92,
385–388, 390, (Antiochene text) 94–96, 114, 184,
392, 394, 397–402, 196, 198–199, 201,
404–405, 407–425, 203, 224, 228–236,
440–441, 447–449, 292, 323, 435, 437
452–456, 458–459,
461 Manuscript(s) 21–22, 49, 69–70,
Gods 39, 46, 164, 207, 73–77, 79, 82,
252, 270, 297–298, 94–96, 128–133,
300, 304, 306–307, 135, 137–138,
412–413, 415, 140–143, 165, 173,
417–418 175–177, 179,
181–182, 184–185,
Hapax legomenon 55–56, 58, 67–68, 188–189, 194,
148, 150–151, 197–199, 202–205,
153–154, 290, 292 226–230, 278–279,
Haplography 43, 45, 197, 217, 281–286, 288–289,
263 291–292, 294,
Hendiadys 97–101, 103–104, 299, 312–313, 316,
120 319–326, 328, 332,
Hermeneutic 22–27, 29, 32, 335–337, 340,
34, 83, 85, 96, 343–344, 346,
427–428 350–354, 356,
Hexapla 72, 74–75, 80, 88, 360, 364, 366,
130, 166, 184–185, 369–371, 375–376,
194–196, 198–199, 387, 390
201–202, 209, 223, Masorete(s) 14, 22, 46,
230, 344, 399, 437, 176–178, 197
445–446 Masoretic text (MT) 10, 13–14, 19,
21–22, 31, 34, 37,
Idol 127, 299–300, 39–41, 43–51, 82,
304–307, 401, 419 84–88, 91–94,
98–101, 104,
Jerome 28, 32–33, 76, 160, 107–111, 114–121,
170, 181, 209, 289, 123–124, 163–165,
295, 297, 301–302, 168, 173, 175–178,
306–307, 359, 363, 180–186, 195–199,
365–379, 434 202–204, 208,
index of subjects 491
Masoretic text (cont.) 211–216, 218–219, P2110 335

225–228, 230–238, P.Fouad 266 73, 179
240–253, 255–262, P.Oxy 3.405 340
264–273, 275–276, P.Oxy 656 179
281–283, 286–288, Peshitta 170, 227, 231,
290, 297–304, 234–235, 275,
306–308, 316, 319, 281, 288–289,
325–326, 331, 333, 293–295, 299–300,
335, 353–354, 399, 302, 307, 326,
404, 427, 429, 353–354, 370, 375,
431, 433–445, 450, 400, 451, 466
453–456, 462 Proto-Masoretic text 81, 93–94,
232–233, 238, 247,
Neologism 61, 67, 148 253, 273, 311, 334,
343
Obelus 195, 284, 344, 446 Ptolemaic 4, 5, 7, 19, 153,
Old Greek (OG) 21, 41, 51, 63, 73, 208, 212, 215, 218
75–76, 79–82, 86, Ptolemies 5, 154
88, 90–92, 94–97, Ptolemy II 3
127, 145–148, 151,
154, 165, 175, Qumran 7, 9, 18, 22, 40,
179–180, 182–183, 50, 73, 78, 80–81,
186, 191, 208–210, 85–86, 92–93,
212, 215, 219, 95–96, 217, 226,
223, 227–230, 237, 281–282, 313,
234, 236, 277, 326, 353–354, 400,
281–282, 285–289, 427
292–295, 311–312,
315–319, 325, 332, Received text 37–38, 42, 46, 48,
334, 336, 342, 383, (textus receptus) 50–51, 227, 230,
399, 406 234, 236, 280
Old Latin 74–76, 78, Retroversion 51, 231, 262
(Vetus Latina) 86–88, 95–96,
286, 292, 359, 365, Scribal error 94, 136, 163, 278,
370–371, 375–376 291–292
Origen 15, 24, 31–33, Scribe 71, 127, 129, 161,
74, 76–77, 130, 176, 178, 180, 182,
166–167, 170, 229, 214, 229, 232–233,
277, 368–369, 279, 282, 292, 357
399, 432, 435, 437, Seleucid 207–208, 212–213,
445–446 216, 218
Original text 22, 88, 129, 143, Septuagint (LXX) 3, 5–6, 10–11,
197, 225, 279 13–15, 17, 19–22,
24, 29, 37–51,
Papyrus 61, 122, 153–154, 53–57, 60–61,
311, 324, 326, 337, 63–70, 73–77,
339, 353–354, 357, 79, 81–82, 88, 90,
386 92–93, 95–97, 105,
P46 332–333, 337–338 109, 114, 117, 121,
P62 203 123–125, 127–130,
P74 342 132–133, 137, 141,
P961 324–325 143, 145–155,
P965 435 157–159, 163,
P967 193–205, 209, 214 166–171, 176–177,
492 index of subjects
Septuagint (cont.) 180–185, 193, Translator(s) (cont.) 251–254, 259–260,

196–198, 202–205, 263–265, 268–270,
208–209, 211–212, 272–277, 280–281,
215–219, 223–225, 285–289, 291–295,
227–236, 239, 297, 300, 302–303,
241–242, 244, 252, 305, 308, 316, 336,
258, 274–277, 398–400, 406, 454
279–281, 284–286, Translation(s) 3–6, 10, 13–14,
288, 291–295, 19–21, 37–42,
299–308, 311–337, 47–51, 54, 56, 60,
339–345, 349–354, 63–64, 68, 75, 83,
357, 359–363, 371, 92–95, 97–100,
374–376, 378–379, 102, 105, 109, 111,
381–391, 393–397, 115, 117, 119,
399–408, 413, 121–122, 124–125,
427–429, 431–447, 128, 130–131, 145,
449–450, 453–462 147, 157, 162,
Symmachus (σ’) 12, 167, 185, 291, 165–167, 169–171,
302, 339, 351, 384, 175, 184, 186–187,
391, 432, 436, 457 191, 204, 208–209,
Syriac version 95, 128, 130, 140, 210–216, 218–219,
(see also Peshitta) 142, 170, 178, 182, 223–224, 228–229,
185, 231, 281, 285, 237, 239, 242, 244,
288–289, 295, 247, 252, 259,
339, 353, 369, 375, 275–281, 284–289,
432, 436, 445 291–295, 297–298,
Syrohexapla 284, 291, 370, 437, 300–302, 305,
441, 445–446 307–308, 311,
316–317, 321, 326,
Targum 12, 158–159, 227, 331–332, 339, 351,
231, 234–235, 300, 353–354, 360–361,
400, 406, 436, 365–368, 372, 383,
442–443, 451 385, 387–388,
Tetragram 65, 173–177, 179, 396–397, 399–400,
183–184, 190, 214 404, 410, 427–428,
Textual criticism 97–98, 100, 114, 432, 434, 436,
117, 223–225, 230, 438, 441–443,
236, 278–280, 445–446, 451,
297, 369 454–455, 458
Theodotion (Th/θ’) 12, 21, 97, 109, Translation 54, 60–61, 63–64,
123–125, 147, equivalent(s) 66–67, 276, 287,
150–151, 167, 182, 291–293, 384
208, 211, 227–228, Translation technique 53, 55–56, 97–99,
277, 307, 382, 399, 107, 113, 117–119,
432 124, 146, 169, 228,
Translator(s) 3, 6, 15, 32, 38–44, 295, 298, 303
47–50, 53–56,
60–68, 85–86, Vorlage 21, 37, 40–51,
98–107, 110–125, 53–55, 60, 67,
145–153, 155, 167, 71, 74–75, 78, 83,
169–171, 181–185, 85–86, 92, 94, 98,
189, 191, 209–219, 99–100, 103–108,
224, 230, 232, 110–115, 117–118,
238–246, 248–249, 120–124, 130, 155,
index of subjects 493
Vorlage (cont.) 171, 174–175, 177, Vulgate 158, 162, 168, 170,
181–182, 184–186, 227, 231, 234–235,
189–190, 197–199, 282, 289, 293–295,
201, 208–213, 297, 299, 306–307,
216–219, 224–225, 326, 331, 333,
230, 232–235, 353–354, 359,
237–238, 240–244, 365, 369–370,
246, 248, 252–254, 372, 374–376,
263–265, 268, 378–379, 432,
272, 275–276, 436, 451
278–281, 285,
287–289, 291–295, Wordplay 53–56, 58, 62–63,
303, 306–307, 65, 67–68
316–317, 319,
322–323, 326–327, YHWH 15, 42–43, 45–48,
329, 335, 339, 341, 58, 173, 175, 188,
346, 348, 359–360, 240, 242, 244–273,
362, 375–376, 399, 398, 453
440, 453–457

Text-Critical and Hermeneutical Studies in The Septuagint PDF

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Text-Critical and Hermeneutical Studies in The Septuagint PDF

Enviado por

Direitos autorais:

Formatos disponíveis

Text-Critical and Hermeneutical Studies in the Septuagint

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

Copyright 2012 by Koninklijke Brill NV, Leiden, The Netherlands.

This book is printed on acid-free paper.

History of the Septuagint in General

The Pentateuch in Greek and the Authorities of the Jews ................. 3

The Biblical Canon and Beyond: Theological and Historical

Translation Technique and Text History

Revisiting the Rock: Tsur as a Translation of Elohim in

Judges 3:12–30: An Analysis of the Greek Rendering of Hebrew

B or not B? The Place of Codex Vaticanus in Textual History

Übersetzungstechnik und Textkritik in den Esdrasbüchern:

In Search of the Old Greek Text of 4 Maccabees .................................. 127

An Analysis of the Use of Hebel as a Metaphorical and Symbolic

Revisiting the Original Greek of Ezekiel 18 .............................................. 193

The Value of the Septuagint for Textual Criticism of the Hebrew

Gottesbildfragen in den Lesartendifferenzen zwischen dem

Two Difficult Passages in the Hebrew Texts of Lamentations 5:

Amos 5:26—Überlegungen zur Textkritik, Textgeschichte und

Die Schriftzitate im ersten Christentum und die Textgeschichte

Ιουδιθ und Iudith: Überlegungen zum Verhältnis der

PsalmsLXX and the Christian Definition of Space: Examples

The Testament of Job as an Adaptation of LXX Job ............................... 395

On Kingship in Philo and the Wisdom of Solomon ............................. 409

The Text Form of the Isaiah Quotations in the Sondergut

Index of Ancient Sources .............................................................................. 463

Supplements to Vetus Testamentum, for accepting the collection for publi-

The Editors, Johann Cook, Dept of Ancient Studies, University of

AASF Annales Academiae scientiarum fennicae

BHT Beiträge zur historischen Theologie

Ges18 W. Gesenius and H. Donner. Hebräisches und Aramäisches

JSQ Jewish Studies Quarterly

PG Patrologia graeca (ed. J.-P. Migne; Paris, 1857–1886)

ThWQ Theologisches Wörterbuch zu den Qumrantexten

Hans Ausloos Université catholique de Louvain, University

History of the Septuagint in General

Arie van der Kooij

It is generally held that the Septuagint version of the Pentateuch was

and learned in the Law—to Alexandria to prepare the translation. The

13 Cf. Fernández Marcos, “The Greek Pentateuch,” 86.

significance the question of whether the Greek version of it was made on

The high priest is depicted in LA as the political leader of Judea, and,

As noted above, in LA the high priest is also presented as the (prime)

18 Ibid., 28.

Exodus 19:6, part of which reads thus,23

The combination of adjective + noun, not that usual though, is attested at

“a holy nation”.30 It seems to be more plausible to interpret our text from

Deut 1:16–17 (LXX)

40 Wevers, Notes on Exodus, 370.

49 See, e.g., 1 Macc 7:33 and 1 Macc 14:28.

LXX Exodus contains evidence testifying to a view of the constitution

53 Compare Deut 19:17 (“the priests and the judges”).

Presented to Emanuel Tov on his 70th Birthday

After the problematisation of the MT induced by the Dead Sea Scrolls

2. Stages of the Early History of the Christian Bible Canon

2.1. Guidelines from Early Judaism

2.1.1. The first available pieces of evidence on the arrangement of the

2.1.2. A Baraita in b.Baba Batra 14b (5th/6th century C.E.) provides an

6 To the problems of a “Synod of Jamnia” cf. A. Hahn, Canon Hebraeorum—Canon

2.2. The Early Christian Discussion

and Epiphanius steered a middle course. The Western Church (Synod of

3. The Codices and Their Peculiarities

The Jewish arrangement of the biblical books remained nearly inflexible;

Ges18 W. Gesenius and H. Donner. Hebräisches und Aramäisches

Hans Ausloos Université catholique de Louvain, University