Escolar Documentos
Profissional Documentos
Cultura Documentos
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
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)
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.
Preface ................................................................................................................. ix
Abbreviations .................................................................................................... xi
List of Contributors ......................................................................................... xvii
Part One
Part Two
The Relationship between the LXX Versions of Proverbs and Job ..... 145
Johann Cook
The θεός and κύριος Terms in the Isaiah Text and their Impact on
the New Testament: Some Observations ............................................ 173
Peter Nagel
Theology after the Crisis: The Septuagint Version of Daniel 8–12 ... 207
Martin Rösel
Part Three
Textual Criticism
Part four
Reception
Die Rezeption von Jer 38:31–34 (LXX) in Hebräer 8–10 und dessen
Funktion in der Argumentation des Hebräerbriefes ....................... 447
Wolfgang Kraus
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
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
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.
6 arie van der kooij
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
II
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
8 arie van der kooij
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.
the pentateuch in greek and the authorities of the jews 9
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.
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 pentateuch in greek and the authorities of the jews 11
“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.
12 arie van der kooij
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”.
the pentateuch in greek and the authorities of the jews 13
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.
14 arie van der kooij
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
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.
16 arie van der kooij
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
the pentateuch in greek and the authorities of the jews 17
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.
18 arie van der kooij
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
“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.
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
1. The Question
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
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.
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.”
the biblical canon and beyond 25
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.
13 Ambrosius, De Tobia (CSEL 32/II, Wien: Tempsky, 1897) 519.1 et seq.; 520.20 et seq.
the biblical canon and beyond 27
for preparing the wonderful and precious copy of “the unique Bible”, the
Codex Alexandrinus.
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.”
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.
the biblical canon and beyond 31
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.
32 heinz-josef fabry
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
the biblical canon and beyond 33
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.
Melvin K. H. Peters
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.
40 melvin k. h. peters
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.
revisiting the rock 41
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.”
42 melvin k. h. peters
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
17 Ibid., 144*.
18 Wevers, Notes, 518.
44 melvin k. h. peters
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
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
revisiting the rock 47
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
revisiting the rock 49
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.
50 melvin k. h. peters
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.
revisiting the rock 51
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 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
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
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
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
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
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
Other Greek Translations Traditionally Included under That Title (eds. A. Pietersma and
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
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
(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
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
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
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.
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
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.
72 siegfried kreuzer
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.
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
74 siegfried kreuzer
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
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).
76 siegfried kreuzer
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
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.
78 siegfried kreuzer
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
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).
80 siegfried kreuzer
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
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.
3.1.1.
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
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.
84 siegfried kreuzer
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).
86 siegfried kreuzer
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
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.
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.
90 siegfried kreuzer
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:
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.
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
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.
49 Cf. Tov, “Lucian and Proto-Lucian,” 101–113: The Antiochene text contains “the LXX”
or “a Septuagint.”
94 siegfried kreuzer
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
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
96 siegfried kreuzer
Dieter Böhler SJ
1. Einführung
3. Hendiadyoin
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.
übersetzungstechnik und textkritik in den esdrasbüchern 101
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).
102 dieter böhler sj
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.
übersetzungstechnik und textkritik in den esdrasbüchern 103
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.“
104 dieter böhler sj
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/
übersetzungstechnik und textkritik in den esdrasbüchern 105
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).
106 dieter böhler sj
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
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).
108 dieter böhler sj
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.
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.
übersetzungstechnik und textkritik in den esdrasbüchern 109
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).
110 dieter böhler sj
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“:
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.
112 dieter böhler sj
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
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.
116 dieter böhler sj
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.
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.“
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).
118 dieter böhler sj
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.
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.
übersetzungstechnik und textkritik in den esdrasbüchern 119
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.
39 Talshir, I Esdras. A Text Critical Commentary, 10.
120 dieter böhler sj
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
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.
übersetzungstechnik und textkritik in den esdrasbüchern 123
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.
124 dieter böhler sj
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.
übersetzungstechnik und textkritik in den esdrasbüchern 125
verdient wie der Jerusalemer Tempel des wahren Gottes. Exakt dieselbe
Differenzierung hatte 1 Esdr in 2:7 auch vorgenommen55.
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
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
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.
in search of the old greek text of 4 maccabees 131
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).
132 robert j. v. hiebert
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
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.
in search of the old greek text of 4 maccabees 133
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.
134 robert j. v. hiebert
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.
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)
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
136 robert j. v. hiebert
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.
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.
in search of the old greek text of 4 maccabees 137
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?
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
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).
138 robert j. v. hiebert
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.
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
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.
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
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).
140 robert j. v. hiebert
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.
in search of the old greek text of 4 maccabees 141
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).
142 robert j. v. hiebert
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
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
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.
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
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).
the relationship between the lxx versions 149
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.
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-
the relationship between the lxx versions 151
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.
27 Cox, “Tying it All Together,” 42.
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.
the relationship between the lxx versions 153
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
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
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
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.
160 lawrence lincoln
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).
an analysis of the use of hebel as a metaphorical 161
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
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).
164 lawrence lincoln
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.
an analysis of the use of hebel as a metaphorical 165
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
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.
an analysis of the use of hebel as a metaphorical 169
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.
170 lawrence lincoln
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
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
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
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
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
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
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.
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
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.
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.
the θεος and κυριος terms in the isaiah text 181
(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).
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.
the θεος and κυριος terms in the isaiah text 183
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.
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.
the θεος and κυριος terms in the isaiah text 185
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.
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).
the θεος and κυριος terms in the isaiah text 187
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
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.
the θεος and κυριος terms in the isaiah text 189
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
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.
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
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.
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.
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”.
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.
198 harry f. van rooy
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.
revisiting the original greek of ezekiel 18 199
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
ἀδικίαν τοῦ πατρὸς
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-
revisiting the original greek of ezekiel 18 201
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.
202 harry f. van rooy
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-
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.
204 harry f. van rooy
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
revisiting the original greek of ezekiel 18 205
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.
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.
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;
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.
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
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.
theology after the crisis 213
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
“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).
theology after the crisis 217
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
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.
theology after the crisis 219
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.
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
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
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 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
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.
the value of the septuagint for textual criticism 229
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:
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]).
the value of the septuagint for textual criticism 231
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.
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:
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.
the value of the septuagint for textual criticism 233
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.
the value of the septuagint for textual criticism 235
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
Hermann-Josef Stipp
1. Das Problem
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
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
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.
240 hermann-josef stipp
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.
gottesbildfragen in den lesartendifferenzen 241
21 Vgl. unten Anm. 30, 34, 43, 46, 49, 63, 74, 83, 99.
22 Vonach, „Jeremias“ (Anm. 11), 2725.
23 Ibid.
24 Ibid.
242 hermann-josef stipp
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).
gottesbildfragen in den lesartendifferenzen 243
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:
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.
gottesbildfragen in den lesartendifferenzen 245
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
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.
gottesbildfragen in den lesartendifferenzen 247
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.
Hand die beiden Verse einer Orthodoxie unterworfen, der an der Domi-
nanz Jhwhs im Verhältnis zu seinem Volk gelegen war36.
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.
gottesbildfragen in den lesartendifferenzen 249
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).
250 hermann-josef stipp
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.).
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.
252 hermann-josef stipp
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.
gottesbildfragen in den lesartendifferenzen 253
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.
254 hermann-josef stipp
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.
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.
258 hermann-josef stipp
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.
gottesbildfragen in den lesartendifferenzen 259
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.
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.
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.
264 hermann-josef stipp
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.
gottesbildfragen in den lesartendifferenzen 265
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.
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.
gottesbildfragen in den lesartendifferenzen 267
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.
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
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.
gottesbildfragen in den lesartendifferenzen 269
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.
270 hermann-josef stipp
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.
gottesbildfragen in den lesartendifferenzen 271
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.
gottesbildfragen in den lesartendifferenzen 273
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.
274 hermann-josef stipp
Gideon R. Kotzé
1. Introduction
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.
278 gideon r. kotzé
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.
two difficult passages in the hebrew texts 279
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
280 gideon r. kotzé
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.
two difficult passages in the hebrew texts 281
3. Method of Analysis
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.
282 gideon r. kotzé
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.
4.1. 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.
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.
two difficult passages in the hebrew texts 283
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.
284 gideon r. kotzé
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
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.
286 gideon r. kotzé
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.
two difficult passages in the hebrew texts 287
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.
two difficult passages in the hebrew texts 289
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.
290 gideon r. kotzé
4.2. Lamentations 5:13
4.2.1. The Hebrew Wordings
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
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.
292 gideon r. kotzé
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
two difficult passages in the hebrew texts 293
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.
294 gideon r. kotzé
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
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
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
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
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.
amos 5:26—überlegungen zur textkritik 301
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.
(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.
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.
18 Hieronymus, Commentarii in prophetas minores, 296.
amos 5:26—überlegungen zur textkritik 303
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
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.
amos 5:26—überlegungen zur textkritik 305
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.
24 Hieronymus, Commentarii in prophetas minores, 297.
25 Ibid., 297.
amos 5:26—überlegungen zur textkritik 307
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
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.
Reception
Die Schriftzitate im ersten Christentum und die
Textgeschichte der Septuaginta:
Ein Wuppertaler Forschungsprojekt
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
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
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.
314 martin karrer und johannes de vries
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)
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.
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.
318 martin karrer und johannes de vries
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.
die schriftzitate im ersten christentum 319
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:
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
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,
die schriftzitate im ersten christentum 321
3.1. Die Quellen
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).
322 martin karrer und johannes de vries
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.
die schriftzitate im ersten christentum 323
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.
324 martin karrer und johannes de vries
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).
die schriftzitate im ersten christentum 325
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).
326 martin karrer und johannes de vries
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.
Die Datenbank ist seit Ende 2011 elektronisch nutzbar. Der Zugang ist frei.
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).
328 martin karrer und johannes de vries
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
die schriftzitate im ersten christentum 329
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.
330 martin karrer und johannes de vries
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,
332 martin karrer und johannes de vries
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.).
die schriftzitate im ersten christentum 333
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.
334 martin karrer und johannes de vries
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.
die schriftzitate im ersten christentum 335
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.
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.
338 martin karrer und johannes de vries
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
die schriftzitate im ersten christentum 339
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
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).
340 martin karrer und johannes de vries
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
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.
342 martin karrer und johannes de vries
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.
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.
die schriftzitate im ersten christentum 343
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.
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)
(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
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.
die schriftzitate im ersten christentum 347
• Num 24:17: ἀνατελεῖ ἄστρον ἐξ Ἰακώβ, der Vers liest weiter: καὶ ἀναστήσεται
ἄνθρωπος ἐξ Ἰσραήλ.
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).
348 martin karrer und johannes de vries
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,
die schriftzitate im ersten christentum 349
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.
350 martin karrer und johannes de vries
6. Ergebnis
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
352 martin karrer und johannes de vries
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.
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 schriftzitate im ersten christentum 353
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 rekon- struierte 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.
die schriftzitate im ersten christentum 355
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:
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:
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
die schriftzitate im ersten christentum 357
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
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
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
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.
ιουδιθ und iudith 363
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.
ιουδιθ und iudith 365
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
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.
ιουδιθ und iudith 367
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
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.
ιουδιθ und iudith 369
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
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]).
ιουδιθ und iudith 371
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
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.
ιουδιθ und iudith 373
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.
ιουδιθ und iudith 375
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.
ιουδιθ und iudith 377
3. Fazit
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).
ιουδιθ und iudith 379
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
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
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
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).
384 cilliers breytenbach
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
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.
psalmslxx and the christian definition of space 385
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.
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.
386 cilliers breytenbach
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
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.
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.
(ornament)
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
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
“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.
392 cilliers breytenbach
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.
psalmslxx and the christian definition of space 393
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).
394 cilliers breytenbach
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
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.
the testament of job as an adaptation of lxx job 399
• 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
16 A. Y. Reed, “Job as Jobab: The interpretation of Job in LXX Job 42:17b–e,” JBL 120/1
(2001): 31.
the testament of job as an adaptation of lxx job 401
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
the testament of job as an adaptation of lxx job 403
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
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).
the testament of job as an adaptation of lxx job 407
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).
Jonathan More
1. Introduction
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
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
2. God as King
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”.
on kingship in philo and the wisdom of solomon 413
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.
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).
on kingship in philo and the wisdom of solomon 415
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
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.
on kingship in philo and the wisdom of solomon 417
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.
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.
on kingship in philo and the wisdom of solomon 419
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).
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).
on kingship in philo and the wisdom of solomon 421
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.
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.
on kingship in philo and the wisdom of solomon 423
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
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.
on kingship in philo and the wisdom of solomon 425
Gert J. Steyn
1. Introduction
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
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.
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.
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
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,
the text form of the isaiah quotations 433
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
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:
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.
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 text form of the isaiah quotations 437
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.
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. (cont.)
(1Q Isaa) 4Q56 MT Isa 42: LXX Isa 42:1–4 Matt 12:18–21
(4QIsab) 1–4
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
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.
the text form of the isaiah quotations 441
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. (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).
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
the text form of the isaiah quotations 445
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
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
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
2.1. 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.
die rezeption von jer 38:31–34 (lxx) in hebräer 8–10 451
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
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.
23 Ibid., 263.
24 Groß, “Erneuerter oder neuer Bund,” 60.
25 Ibid., 62.
26 Zusammenstellung in Anlehnung an M. Meiser, Die Rezeption von Jer 38 (31) in
Hebr 8, Manuskript 1f.
454 wolfgang kraus
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.
die rezeption von jer 38:31–34 (lxx) in hebräer 8–10 455
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.
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.
die rezeption von jer 38:31–34 (lxx) in hebräer 8–10 457
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.
39 Ibid., 103f.
458 wolfgang kraus
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
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.
die rezeption von jer 38:31–34 (lxx) in hebräer 8–10 461
5. Fazit
Die Einführung der neuen (καινή) bzw. besseren (κρεῖττον, 7:22) bzw. fri-
schen (νεά, 12:24) διαθήκη dient
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
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
466 index of ancient sources
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
index of ancient sources 467
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
482 index of ancient sources
1QM CD A
2:1–3 18 VII, 14–15 299, 307
VII, 17 299, 307
1QS
3:18 422
Jewish Authors
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
484 index of ancient sources
Rabbinic Writings
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
index of ancient sources 485
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
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
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