Editing the Septuagint: The Unfinished Task: Papers presented at the 50th anniversary of the International Organization for Septuagint and Cognate Studies, Denver 2018 [1 ed.] 9783666560637, 9783110615463, 9783525560631

Citation preview

De Septuaginta Investigationes (DSI) Edited by Anneli Aejmelaeus, Felix Albrecht, Kristin De Troyer, Wolfgang Kraus, Michael Segal In Co-operation with Kai Brodersen (Erfurt, Germany), Cécile Dogniez (Paris, France), Peter Gentry (Louisville, USA), Anna Kharanauli (Tbilisi, Georgia), Armin Lange (Wien, Austria), Alison Salvesen (Oxford, UK), David Andrew Teeter (Cambridge, USA), Julio Trebolle (Madrid, Spain), Florian Wilk (Göttingen, Germany) Volume 16

Felix Albrecht / Frank Feder (eds.)

Editing the Septuagint: The Unfinished Task Papers presented at the 50th anniversary of the International Organization for Septuagint and Cognate Studies, Denver 2018

Vandenhoeck & Ruprecht

Bibliographic information published by the Deutsche Nationalbibliothek: The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data available online: https://dnb.de. © 2022 by Vandenhoeck & Ruprecht, Theaterstraße 13, 37073 Göttingen, Germany, an imprint of the Brill-Group (Koninklijke Brill NV, Leiden, The Netherlands; Brill USA Inc., Boston MA, USA; Brill Asia Pte Ltd, Singapore; Brill Deutschland GmbH, Paderborn, Germany; Brill Österreich GmbH, Vienna, Austria) Koninklijke Brill NV incorporates the imprints Brill, Brill Nijhoff, Brill Hotei, Brill Schöningh, Brill Fink, Brill mentis, Vandenhoeck & Ruprecht, Böhlau, V&R unipress. All rights reserved. No part of this work may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without prior written permission from the publisher. Typesetting: Datagrafix Vandenhoeck & Ruprecht Verlage | www.vandenhoeck-ruprecht-verlage.com ISSN 2197-0912 ISBN 978-3-666-56063-7

Preface The Greek translation of the Hebrew (and Aramaic) Old Testament, commonly known as Septuagint, has its origins in Ptolemaic Egypt. According to the “Letter of Aristeas”, the Pentateuch was translated in Alexandria during the third century BCE. Since then, Egypt developed into a strongly bilingual country and the Christian mission in Egypt certainly was based on Greek in its beginnings. In the fourth century CE, when Christianity was on firmer ground in Egypt, the Septuagint was also translated into the native Egyptian language, today known as Coptic.1 The Coptic daughter version of the Septuagint, next to the Old Latin, represents the oldest and text historically the most important of all daughter versions. The intertwined and prolific relation between the Greek and the Coptic Old Testament is now aptly reflected also in the joint ventures of the Göttingen Academy of Sciences and Humanities. Göttingen was and is the center of Septuagint research. This research is deeply rooted in the 19th century and linked to the notorious Paul Anton de Lagarde (1827–1891).2 His disciple Alfred Rahlfs (1865–1935) continued Lagarde’s work and founded in 1908 a research institution (until 2015) under the name “Septuaginta-Unternehmen”.3 In 2015, a second major long-term project joined the Göttingen Academy, which deals with the translation of the Septuagint into CopticSahidic and aims at preparing a complete digital edition of the Coptic-Sahidic Old Testament (http://coptot.manuscriptroom.com). Finally, in 2020, the “Editio critica maior des griechischen Psalters” started as a new long-term project at the Göttingen Academy (https://septuaginta.uni-goettingen.de). Our two projects – the edition of the Coptic-Sahidic Old Testament and the edition of the Greek Psalter – work closely together, and the present volume is one of the results of our fruitful collaboration. We are pleased and grateful that this volume appears in the series Investigationes “De Septuaginta”. We wish to thank the editors of the series as well as the anonymous peer-reviewers for their feedback and support. The contributions in this volume go back to a joint panel that we organized as part of the “Society of Biblical Literature” meeting in Denver, Colorado, in November 2018. At this very 1  Cf. Frank Feder, “1.1.6 The Coptic Canon,” in Textual History of the Bible Vol. 2A The Deuterocanonical Scriptures, eds. Frank Feder and Matthias Henze (Leiden: Brill, 2020), 213–39. 2 On Lagarde see Heike Behlmer, Thomas L. Gertzen and Orell Witthuhn, eds. Der Nachlass Paul de Lagarde. Orientalische Netzwerke und antisemitische Verflechtungen. Europäisch-jüdische Studien – Beiträge 46, Berlin, Boston: De Gruyter Oldenbourg, 2020. Open Access: https://doi.org/10.1515/9783110615463. 3  Reinhard Gregor Kratz and Bernhard Neuschäfer, eds. Die Göttinger Septuaginta. Ein editorisches Jahrhundertprojekt. Abhandlungen der Akademie der Wissenschaften zu Göttingen. Neue Folge 22. MSU 30. (Berlin u.a.: De Gruyter, 2013).

6

Preface

meeting, the “International Organization for Septuagint and Cognate Studies” celebrated its 50th Anniversary. Most of the articles have been held as papers at our panel “Göttingen Septuagint: Greek and Coptic”, others have been added at a later stage. To commemorate the special event of the panel held in Denver and this particular anniversary we took the opportunity to publish a selection of papers that deal with (1) the Göttingen Editions, and (2) the Hexapla and Recensions of the Septuagint. Since the Corona pandemic caused a certain delay in the editorial process, we are extremely grateful to the authors for their patience: Domitrix rerum patientia! The editors.

Content Preface�����������������������������������������������������������������������������������������������������������������������������5 I.  Göttingen Editions Anneli Aejmelaeus Challenges in Preparing the Critical Edition of 1 Samuel�������������������������������������11 Paavo Huotari Discovering Old Greek Readings in the Lucianic Text of 2 Samuel by comparison with 4QSama, c–Wishful Thinking?������������������������������������������������27 Pablo A. Torijano/Julio Trebolle The Edition of III–IV Kingdoms. The Critical Reconstruction of the Old Greek Text and the Construction of the Critical Apparatus��������������������������53 Robert J. V. Hiebert A Synopsis of the Textual History of 4 Maccabees�������������������������������������������������81 Peter J. Gentry The Göttingen Edition of Ecclesiastes����������������������������������������������������������������������95 Felix Albrecht The Göttingen Edition of the Psalms of Solomon������������������������������������������������113 Frank Feder A New Textual Witness of the Sahidic Version of Jeremiah and Its Text Historical Assessment�������������������������������������������������������������������������123 II.  Hexapla and Recensions John D. Meade The Dream for a ‘New Field’ Comes True. A Description and Defense of the New Critical Edition of Job 22–42���������������������������������������������������������������131 Benjamin Kantor Discovering the Secunda. Insights from Preparing a New Critical Edition of the Second Column of Origen’s Hexapla�����������������������������������������������������������155 Daniel Olariu Recensional Additions. Insights from Theodotion Daniel����������������������������������193 Indices�������������������������������������������������������������������������������������������������������������������������225

I.  Göttingen Editions

Anneli Aejmelaeus Challenges in Preparing the Critical Edition of 1 Samuel Working on a critical edition – in my case, the critical edition of the Septuagint translation of 1 Samuel – means doing textual criticism on that text, on each and every word and passage of it. The results of this text-critical work will be seen in the critical text, the main text of the edition, which is intended to represent the closest possible approach to the original wording of the book. In the case of the Septuagint, there are fairly many Greek manuscripts that are used for the reconstruction of the textual history of the book.1 All these manuscripts as well as the daughter versions and quotations by early Jewish and Christian writers, which are also used as textual witnesses, will be documented in the apparatus of the edition. The critical text cannot however be established solely on the basis of a survey of the textual witnesses, as problematic cases can rarely be solved by building a stemma of the manuscripts. The reason for this is that the manuscripts seldom represent pure copying of a model manuscript, but often several manuscripts were used when preparing a new manuscript. This was already recognized by Paul de Lagarde, who formulated one of his principles like this:2 (1) Since the manuscripts of the Septuagint are all directly or indirectly the result of an eclectic process, any attempt to restore the original text must also proceed on eclectic principles; and the critic must chiefly depend upon (a) his [or her] acquaintance with the style of the several translators and (b) his [or her] faculty of referring readings to a Semitic original or, when they are not of Semitic origin, recognizing them as corruptions of the Greek archetype.

Lagarde mentions two different sets of criteria: the translation character of the translation in question and the influences from the Hebrew text. These two demand special attention when tackling the problems caused by textual contamination and corruption. Especially the relationship with the Hebrew text plays 1  The manuscripts available for the First Book of Samuel, with their tentative groupings, are the following (with those only partially preserved in parentheses): B A V (M) (842) (845) (846) (867); O = 247-376; L = 19-82-93-108-127; CI = 98-(243)-379-731; CII = 4652-236-242-313-328-530; a = 119-527-799; b = 121-509; d = 44-68-74-106-107-120-122-125134-(370)-610; f = 56-246; s = 64-92-130-314-381-488-489-(762); 29 55 71 158 244 245 318 (342) 460 554 707. As for the group sigla, O stands for the Hexaplaric, L for the Lucianic, C for Catena manuscripts. 2 Lagarde, Anmerkungen zur griechischen Übersetzung der Proverbien, 3; translation according to Swete, An Introduction to the Old Testament in Greek, 484–6.

12

Anneli Aejmelaeus

a special role in the textual criticism of the Septuagint. Behind the translation there is of course the Hebrew Vorlage that at times diverged from the MT, but in addition, the Greek text has been repeatedly approximated to the Hebrew in its different stages, and this makes the textual criticism of the Septuagint different from the textual criticism of any other text. Thus, text-critical problems cannot be solved by external criteria only, but internal criteria always play an essential role. Nevertheless, it is not an either-or situation. Which one of the existing readings is to be considered most original cannot be decided alone by the suitability of the readings in their context either. The most important question in textual criticism is “what happened to the text?” When answering this question, we actually need to combine, on the one hand, what we know about the textual history of the text (the external criteria), and on the other, the evaluation of the kinds of variants there are and the consideration of the reasons for their emergence (the internal criteria). During my work on the Greek 1 Samuel, I have made some observations and discoveries concerning the different factors that have been at work in the textual history of this text. This has led me to an understanding of this textual history that differs to some extent from the understanding Rahlfs had of it, and consequently, the critical text of my edition will be different from Rahlfs’ edition. I shall mention four factors or phenomena of textual history that I have discovered to be decisive in establishing the critical text in problematic cases – the translator, the Vorlage, Jewish revisional activity, and doublets – and then introduce just a few examples of cases demonstrating how these factors come into play in the text-critical procedure. I have chosen examples in which the forthcoming edition will differ from Rahlfs’ edition.3

1.  The Translator The first factor that needs to be taken into account is naturally the translator.4 We need to know the characteristic features of his translation style in order to be able to reconstruct his wordings.5 Various studies have shown that this translator had a fairly strong word-for-word approach to his source text. However, literalism was for him not so much a principle as “an easy technique,” as James Barr put it.6

3  The readings found in Rahlfs’ manual edition are designated by Ra. 4  For a characterization of the translation, see Wirth, Die Septuaginta der Samuelbücher; Aejmelaeus, “The Septuagint of 1 Samuel”, 109–129. 5  For the sake of simplicity, I refer to this translator by the masculine pronoun, but of course, this might be false. We simply do not know enough about the persons who translated the various books of the Septuagint. 6 Barr, The Typology of Literalism in ancient biblical translations, 26, 50.

Challenges in Preparing the Critical Edition of 1 Samuel

13

Unfortunately, this translator was not quite up to his task. He had great difficulties with some Hebrew words and in many cases simply tried to guess on the basis of the context what the text means (e.g. 1 Sam 30:10, 21 below). If he did not know a certain Hebrew root, he often tried to connect it with another one that has at least some similarity (e.g. 1 Sam 14:32 ‫ וַ ַּי ַ֤עט‬and it rushed upon’ from ‫ עיט‬was translated by καὶ ἐκλίθη ‘and it turned,’ as if from ‫)נטה‬. Several words that have been translated correctly in the Pentateuch were not recognized by this translator (e.g. 1 Sam 15:3, 8 ‫ חרם‬hiph. ‘to devote to the ban’), so he was obviously not a learned person who had studied the Pentateuch in both Hebrew and Greek. Time and again he uses transliterations (e.g. 1 Sam 30:8, 15, 23 γεδδούρ for ‫ּגְ דּוד‬ ‘raiding party’; 1 Sam 2:18, 28; 14:3, 18 etc. ἐφούδ for ‫‘ ֵאפֹוד‬priestly garment’),7 which is a clear sign of problems, but also when he uses Greek words, we need to take into consideration that he might have produced a false translation. On the other hand, this translator did very nice work with verbal forms. Frequent use of the historical present is characteristic of him. He was also able to recognize the past iterative forms in Hebrew. In fact, I learned about the past iterative through this translator: in Hebrew, repeated actions are expressed by the alternation of the perfect consecutive and the imperfect, and this is translated by the Greek imperfect. However, this translator alternates both historical presents and (iterative or durative) imperfects with the aorist to make the discourse livelier (e.g. 1 Sam 2:14 below).8

2. The Vorlage The second factor is the Vorlage, which was oftentimes different from the MT, even more so than generally thought. One fundamental principle in my methodology is that the Greek text must be studied in relation with the Hebrew text. In the course of my editorial work, I have not been able to avoid the conclusion that the MT has been deliberately edited at a fairly late stage of the textual history, in any case later than the translation, which I would date to the second half of the 2nd century BCE.9 In many cases it is possible to show the theological

7  Please, note that the former example of transliteration is based on confusion between daleth and resh, which suggests that the false transliteration originated with the translator. 8  For a more detailed discussion, see Wirth, “Das Praesens Historicum in den griechischen Samuelbüchern,” and Wirth “Dealing with Tenses in the Kaige Section of Samuel.” 9  Certain linguistic features in the translation of 1 Samuel suggest that it was made clearly later than the translation of the Torah, which can be dated to the 3rd century BCE (see Lee, A Lexical Study of the Septuagint Version of the Pentateuch, 140–144). As for an ante quem date, the Greek text of Samuel was possibly known to the translator of Sirach, see Aejmelaeus, “When Did the Books of Samuel Become Scripture?,” 263–81 (esp. 268).

14

Anneli Aejmelaeus

or ideological motivation behind the changes of the Hebrew text.10 The Septuagint thus often witnesses to a more ancient form of the text of 1 Samuel. That the Vorlage often differed from the MT was taken into consideration already by Paul de Lagarde, who formulated another one of his principles accordingly:11 (2) Where the critic has to make a choice between two readings, he [or she] will do well to prefer (a) a free translation to one which is slavishly exact, and (b) a translation based upon another Hebrew text to one which represents the MT.

It is important to take into account that the Septuagint may represent another Hebrew text – not only because one or the other Hebrew text was corrupted, but because the MT was later on changed – and that the revisers often corrected the Greek text in cases like these.

3.  Jewish revisional activity The third factor to be mentioned concerns revisions of the Greek text. It is well known that the textual history of the Septuagint includes two Christian recensions: the Hexaplaric and the Lucianic.12 This was already known to Lagarde and Rahlfs, but what they did not know, is that there had been Jewish revisional activity on this Greek text early on.13 In certain problematic cases Rahlfs might have suspected that there had been pre-hexaplaric revision, but he had no proof of it. In this respect, our generation is in a better position to solve textual problems in the Books of Samuel – and elsewhere. The discovery of the Naḥ al Ḥ ever Minor Prophets scroll and Dominique Barthélemy’s interpretation of it have made it clear that the Greek text of the Septuagint was revised by Jewish scholars and that traces of this so-called kaige revision are found in certain sections of Samuel-Kings in Codex Vaticanus (B) and the majority of other witnesses as well as in the B text of Judges – to mention just the most important for my inquiry.14

10  For editorial activity on the Hebrew text, see Aejmelaeus, “Was Samuel Meant to Be a Nazirite?” 11  According to the translation of Swete, Introduction, 189–90. See above note 2. 12  “Recension” is a traditional term used for a new, systematically revised edition of the Septuagint text. The term “revision” is used when referring to more sporadic revision or approximation to the Hebrew text. 13 Aejmelaeus, “What Rahlfs Could not Know: 1 Sam 14,4–5 in the Old Greek.” 14 Barthélemy, Les devanciers d’Aquila,. For the final edition of the revised Minor Prophets Scroll, see Tov, The Greek Minor Prophets Scroll from Naḥ al Ḥ ever.

Challenges in Preparing the Critical Edition of 1 Samuel

15

My discovery is that there are traces of Jewish revisional activity also in 1 Samuel.15 Revision that clearly conforms to the kaige translation philosophy was recognized by Barthélemy in the second half of 2 Samuel.16 Similar variants are, however, also sporadically found in 1 Samuel, showing that there probably existed a kaige revision for 1 Samuel as well and that this has had an influence on the manuscript tradition of the Old Greek, especially on the B text (Codex Vaticanus and the accompanying minuscules 121–509 as well as Aeth). Several examples below illustrate the effects of this phenomenon on the textual history of 1 Samuel.

4. Doublets One more text-historical factor remains to be presented. The earliest layer of corrections in 1 Samuel – earlier than the kaige-type corrections – resulted in numerous doublets that consist of the Old Greek translation of a word, a phrase or a short passage and its correction. The secondary part of the doublet, the formulation considered to be more accurate, must have been first added to the margin of a manuscript from where it slipped into the text, sometimes before the Old Greek counterpart, sometimes after it, and sometimes at a different location. These doublets mark the earliest phase of the textual history attested in the manuscripts, as they are present in practically every manuscript. These early corrections do not always show translation features that would connect with the kaige translation style but there always seems to have been some detail that called for correction.17

5. Examples I shall begin by introducing a case with early Jewish revisional readings, an example that I have discussed on many occasions, but one that is worth repeating because it shows the connection with other exemplars of kaige.18

15 Aejmelaeus, “Kaige Readings in a Non-Kaige Section in 1 Samuel.” The phenomenon was initially discussed in Aejmelaeus, “David’s Return to Ziklag: A Problem of Textual History in 1 Sam 30:1.” 16  Barthélemy accepted the definition of this section as 2 Sam 11:2 – 1 Kings 2:11 by Thackeray, “The Greek Translation of the Four Books of Kings.” The beginning was shown to lie more probably at 2 Sam 10:6 by Wirth, “Dealing with Tenses in the Kaige Section of Samuel.” 17  See also Aejmelaeus, “Where Do Doublets Come from? A Problem of the Septuagint of 1 Samuel.” 18  See Aejmelaeus, “A Kingdom at Stake,” 362–4; Aejmelaeus, “Does God Regret? A Theological Problem that Concerned the Kaige Revisors.”

16

Anneli Aejmelaeus

(1)  1 Sam 15:11 – early revisional readings ‫ת־ׁשאּול ְל ֶמ ֶלְך‬ ָ ‫י־ה ְמ ַל ְכ ִּתי ֶא‬ ִ ‫נִ ַח ְמ ִּתי ִּכ‬

Μεταμεμέλημαι ὅτι ἔχρισα τὸν Σαοὺλ εἰς βασιλέα μεταμεμέλημαι] μεταμέλημαι V 46*-313 55* 71 460; μεταμέλομαι d 554; παρακέκλημαι B A 247 93mg-108mg 121*(vid) Ra: cf MT; παρακέκληκέ με 376(-καί με) | om ὅτι ἔχρισα A | ἔχρισα] ἐβασίλευσα B O L b 244 460 Ra = MT | θ′ παρακέκλημαι σ′ μετεμελήθην 243-731(s nom)

This is the message that Samuel receives concerning Saul: God has rejected Saul saying, “I regret having anointed Saul to be king.” There are two different kinds of corrections in this example. In the first one, the Hebrew verb ‫ נחם‬niph. ‘to regret,’ has been correctly translated in the Old Greek by μεταμέλομαι but this was changed to παρακαλοῦμαι which corresponds to the meaning of the Hebrew verb in pi. ‘to comfort,’ and more precisely, to its passive in niph. ‘to be comforted.’ The change produces a concordant translation of the Hebrew verb with the same Greek verb in all its different forms, although the context demands another equivalent. The resulting text is hardly comprehensible. The same change is found in the kaige section of 2 Samuel (24:16) as well as in the Minor Prophets scroll from Naḥ al Ḥ ever (Jonah 3:9 and 10). There is, however, more at stake here, and I shall come back to this example. The second case is a simpler one. There was a difference in the Hebrew text: the Vorlage of the Old Greek contained the verb ‫‘ משׁח‬to anoint,’ whereas the MT has been changed to ‘to appoint to be king’ (cf. the parallel 15:35). In these two cases as well as elsewhere in 1 Samuel, the main witness for the pre-hexaplaric corrections is Codex Vaticanus, accompanied by the other representatives of the B text (b [= 121-509]) and a few other manuscripts, and so these secondary readings have ended up in all editions of the Greek text, including Rahlfs, but will be corrected in the forthcoming critical edition. The following example reveals one of those early doublets that occur in practically all the manuscripts. In a way, they too represent early revisional activity.

(2)  1 Sam 4:14–16 – an early doublet ‫אמר ֶ ֛מה ֥קֹול ֶה ָה ֖מֹון ַה ֶּז֑ה וְ ָה ִ ֣איׁש ִמ ַ֔הר וַ ּ֖יָב ֹא וַ ּיַ ֵּג֥ד ְל ֵע ִ ֽלי׃‬ ֶ ֹ ‫ת־קֹול ַה ְּצ ָע ָ ֔קה וַ ּ֕י‬ ֣ ‫וַ ּיִ ְׁש ַ ֤מע ֵע ִ ֙לי ֶא‬ 16 ‫ל־ע ֗ ִלי‬ ֵ ‫אמר ָה ִ֜איׁש ֶא‬ ֶ ֹ ‫… וַ ּ֙י‬ ‫ּוׁשמ ֶֹנ֖ה ָׁש ָנ֑ה וְ ֵע ָינ֣יו ָ ֔ק ָמה וְ ֥ל ֹא יָ ֖כֹול ִל ְר ֽאֹות׃‬ ְ ‫ן־ּת ְׁש ִ ֥עים‬ ִ ‫ וְ ֵע ֕ ִלי ֶּב‬15

14καὶ

ἤκουσεν Ηλι τὴν φωνὴν τῆς βοῆς καὶ εἶπεν Τίς ἡ βοὴ τῆς φωνῆς ταύτης; καὶ ὁ ἄνθρωπος σπεύσας εἰσῆλθεν καὶ ἀπήγγειλεν τῷ Ηλι. 15καὶ Ηλι υἱὸς ἐνενήκοντα ἐτῶν, καὶ οἱ ὀφθαλμοὶ αὐτοῦ ἐπανέστησαν, καὶ οὐκ ἔβλεπεν· καὶ εἶπεν Ηλι τοῖς ἀνδράσιν τοῖς περιεστηκόσιν αὐτῷ Τίς ἡ φωνὴ τοῦ ἤχους τούτου; 16καὶ ὁ ἀνὴρ σπεύσας προσῆλθεν πρὸς Ηλι καὶ εἶπεν αὐτῷ… (Rahlfs)

Challenges in Preparing the Critical Edition of 1 Samuel

17

Reading the Greek text according to Rahlfs’ edition, we can see that the parts of text underlined (either dashed or waved) form a doublet, but their formulation is not identical. If we take out the part with waved underlining, we get a text that corresponds to the MT. If we take out the part with dashed underlining, we get a text that has about the same content but differs from the MT in the order of the text and in its wording. The latter alternative obviously represents the Old Greek, which is also based on a slightly different Vorlage. The part with dashed underlining must have been added to the margin as a corrective to approximate the Greek text to the Hebrew proto-MT. When the text was copied, the marginal reading was inserted into the text by a scribe, who obviously understood that it belonged there. Since the doublets are secondary elements in the Greek text – even if they are witnessed by practically all manuscripts – they will not appear in the main text of the critical edition.19 The critical text and apparatus will thus look like this: καὶ ἤκουσεν Ἠλὶ τὴν φωνὴν τῆς βοῆς. † – 15 καὶ Ἠλὶ υἱὸς ἐνενήκοντα ἐτῶν, καὶ οἱ ὀφθαλμοὶ αὐτοῦ ἐπανέστησαν, καὶ οὐκ ἔβλεπεν. – καὶ εἶπεν Ἠλὶ τοῖς ἀνδράσιν τοῖς περιεστηκόσιν αὐτῷ Τίς ἡ φωνὴ τοῦ ἤχου τούτου; 16 καὶ ὁ ἀνὴρ σπεύσας προσῆλθεν πρὸς Ἠλὶ καὶ εἶπεν αὐτῷ … 14 fin] add † καὶ εἶπεν Τίς ἡ φωνὴ τῆς βοῆς ταύτης; καὶ ὁ ἄνθρωπος σπεύσας εἰσῆλθεν καὶ ἀπήγγειλεν τῷ ᾿Ηλί omn codd La115 Sa Aeth Arm Ra = MT: dupl ex 15–16 ⟦ om καί 1 – βοῆς V: homoiotel | ἡ φωνὴ τῆς βοῆς] φωνὴ τῆς ἀκοῆς 82; ἡ βοὴ τῆς φωνῆς B O 799 509 106-107-125-610 55 244 La115 Ra | ταύτης] αὕτη 19 | σπεύσας] ἔσπευσεν καί 29 | εἰσῆλθεν La115] ἦλθε 106 381 | καὶ ἀπήγγειλεν] καὶ ἀνήγγειλε(ν) V 731 74; om καί 158 | om τῷ 509 | Ἠλί] Ἠλεί B A 247 L–93 98 46-242 b f 48 29 244 460 554 707⟧ 14

This is one of the types of cases in which my critical text goes behind the archetype from which all our manuscripts derive. Another kind of case reaching beyond the manuscript witness can be observed in the following example.

(3)  1 Sam 15:28 – early corruption and reconstruction of a more original text ‫ת־מ ְמ ְל ֧כּות יִ ְׂש ָר ֵ ֛אל ֵמ ָע ֶל֖יָך ַהּי֑ ֹום‬ ַ ‫הוה ֶ ֽא‬ ֜ ָ ְ‫ָק ַ ֙רע י‬

Διέρρηξεν Κύριος τὴν βασιλείαν σου ἐπὶ ᾿Ισραὴλ ἐκ χειρός σου σήμερον σου ἐπί scripsi] σου ἀπό B A O L b d f 55 554 Aeth Luc Par 4 Reg 2; ἀπό CII 64 s 244 460 Aug CD 17,7,9; > V CI a 381 29 71 158 245 318 707 Aug Leg 1,42 Isid I Reg 17,8 Tert Marc 2,24,7 Ra = MT | om Ἰσραήλ V 799 19  In the case of doublets that include a transliteration by the translator and a correction that explains the meaning of the word in question, for the sake of readability of the text, the secondary part may appear in square brackets.

18

Anneli Aejmelaeus

Vorlage: ‫ממלכתך על ישראל מידך‬ Cf. 1 Sam 28:17 καὶ διαρρήξει Κύριος τὴν βασιλείαν σου ἐκ χειρός σου and 1 Sam 13:13 τὴν βασιλείαν σου ἕως αἰῶνος ἐπὶ Ἰσραήλ

In this case the majority of the manuscripts (among others, the B text as well as the Hexaplaric and the Lucianic texts) have the preposition ἀπό which is an inner-Greek corruption that actually makes no sense (“your kingdom from Israel out of your hand”). Rahlfs follows those manuscripts that have an omission in accordance with the MT (“the kingdom of Israel from you”). The Vorlage must have had something different – “your kingdom over Israel” – and the MT has been changed here as well as partially also in the parallel cases 13:13 and 28:17.20 The following is a case that puzzled me from the very beginning of my work on 1 Samuel. Its solution requires familiarity with the translator’s style as well as with the manuscript groups.

(4)  1 Sam 2:14 – an appropriate translation and subsequent revisions21 … ‫ו ְִה ֙ ָּכה ַב ִּכּי֜ ֹור‬14 ‫ׁש־ה ִּׁש ַּנ�֖יִם ְּביָ ֽדֹו׃‬ ַ ‫… וְ ַה ַּמזְ ֵל֛ג ְׁש ֹ֥ל‬ ‫ּובא ַנ ַ֤ער‬ ָ֙ καὶ ἤρχετο τὸ παιδάριον… καὶ κρεάγρα τριόδους ἐν τῇ χειρὶ αὐτοῦ· 14καὶ καθῆκεν αὐτὴν εἰς τὸν λέβητα τὸν μέγαν… καὶ καθῆκεν] pr καὶ ἐπάταξεν 509; καὶ καθίει L; καὶ καθῆκαν 158; κεκράτηκεν d–68′(mend); καὶ ἐπάταξεν Β Α f Ra: cf MT; καὶ ἐπάταξαν 121 68′; καὶ ἐπάτασσεν Ο = MT; et iecit Aeth; et mittebat La115

This is part of the description of the misconduct of the sons of Eli, something that occurred repeatedly in connection with sacrificial meals, and for this reason the verbal forms are mainly those used for iterative past action: in Hebrew the perfect consecutive and imperfect; in Greek the imperfect. In this case, the translator has done a good job choosing a contextually fitting verb καθίημι ‘to send down,’ ‘to let down,’ (rarely also) ‘to strike down.’ The aorist – among the chain of imperfects – was used to express abrupt movement. Among the alternative readings, there are different forms (aorist and imperfect) from two different verbs πατάσσω and καθίημι of which the former is the standard equivalent for ‫ נכה‬hiph. The translator certainly knew this verb and often used the standard rendering for it, but he chose to use here a verb that is

20  The MT reads at 1 Sam 13:13 ‫ד־עֹולם‬ ָֽ ‫ת־מ ְמ ַל ְכ ְּתָך֛ ֶאל־יִ ְׂש ָר ֵ ֖אל ַע‬ ֽ ַ ‫ ֶא‬and at 28:17 ‫ת־ה ַּמ ְמ ָל ָכ ֙ה ִמּיָ ֶ ֔דָך‬ ַ ‫א‬. ֶ For a more thorough argument, see Aejmelaeus, “A Kingdom at Stake,” 359–61. 21  This example has been previously discussed in Aejmelaeus, “The Septuagint of 1 Samuel,” 138, and Aejmelaeus, “Kaige Readings in a Non-Kaige Section in 1 Samuel,” 170–72.

Challenges in Preparing the Critical Edition of 1 Samuel

19

more appropriate in the context.22 The changes of the verb and of the tense that we have among the variants are clearly intentional, whereas the alternation of singular and plural and the spelling error κεκράτηκεν (καί → κε-; καθῆκεν → -κράτηκεν) were most probably unintentional. The emergence of the different readings can only be explained, if καθῆκεν is taken as the original reading. It is represented by the majority of manuscripts, and in a sense, the Lucianic text also participates in this reading, but out of stylistic reasons, that one aorist among several imperfects was changed by the Lucianic reviser to the imperfect καθίει. It would be impossible to reconstruct the development the other way around, presupposing that καθίει was original. There has been plenty of discussion on the position of the Lucianic text in the textual history of the Books of Samuel. It is true that the Lucianic recension is based on a good old text, but it is also clear that this text has recensional features, in particular, changes meant to improve the Greek language and style.23 But where did the reading ἐπάταξεν come from? It is the standard rendering of the Hebrew verb, and thus, it clearly represents an approximation to the Hebrew text, correcting a free rendering. It is typical that this reading is found in Codex Vaticanus (B) and just a few other manuscripts, the accompanying minuscules 121 and 509 showing minor variations. Other manuscripts following the B text in this reading are A and O: A having ἐπάταξεν and the O group (= 247-376) an analogous change to the imperfect as the Lucianic text). A and O were recognized by Rahlfs to be Hexaplaric.24 In this case, Rahlfs did not however see any Hexaplaric influence but considered ἐπάταξεν to be part of the original translation. From my viewpoint, any form of πατάσσω is here secondary, but neither do I consider it to be Hexaplaric. Why not? The Hexaplaric recension is known for its approximations to the Hebrew text – not, however, for approximations of this kind. Origen’s main interest was in the plusses and minuses between the Greek and the Hebrew texts and not in translation equivalents. It is also worth noting that the B text does not contain any of the characteristically Hexaplaric plusses. How did it happen then that the B text and the Hexaplaric witnesses agree in a secondary reading against the great majority of witnesses? The only possible explanation is that Origen knew and used a manuscript that represented the B text.25 Thus, the B text must have been in existence around 200 CE. It is my theory that 22  In the case of ‫ נכה‬hiph. the most common rendering is naturally πατάσσω, but there are – in addition to καθίημι – several alternative renderings (πλήσσω 4:2; 5:12; τύπτω 11:11; 17:36; 27:9; 31:2; παίω 13:4; ἐκζέω 5:6; θανατόω 17:35; 20:33; ἀποκτείνω 17:46). 23  More examples can be found in Aejmelaeus, “Textual History of the Septuagint and the Principles of Critical Editing”, 160–79 (esp. 167–71). 24  Rahlfs in fact includes A in the group siglum O. 25  This was maintained already by Swete, An Introduction to the Old Testament in Greek, 487. Rahlfs, Septuaginta-Studien I–III, 101, suggests that B was “cum grano salis die Vorlage des Origenes.”

20

Anneli Aejmelaeus

the B text is the first Christian edition of the Biblical text from Alexandria – an edition characterized by its adoption of early Jewish corrections.26 Origen obviously knew different manuscripts of the Septuagint, and comparing them, he, of course, preferred readings that were closer to the Hebrew text, believing to recover in this way the genuine Septuagint. Readings that Origen picked up from different manuscripts for his fifth column were by definition pre-Hexaplaric, and he naturally did not mark them in any way. If we now compare our first example 1 Sam 15:11 with this one, we can see that the Hexaplaric recension is following the B text in both cases there too: (1) in the first case, the corrective παρακέκλημαι is found in A 247 and also in 376 in an erroneous form, and (2) in the second case, the corrective ἐβασίλευσα is found in the O group whereas A has an omission. Outside of these two groups, the B text and the Hexaplaric text, the distribution of these readings is not very wide. It is also interesting to see how the Lucianic text behaves in these cases. (1) In the first case, the Lucianic text witnesses the original reading μεταμεμέλημαι. (2) In the second case, however, the approximation to the Hebrew ἐβασίλευσα is also found in the Lucianic text. The pre-Hexaplaric corrections must have been known to the Lucianic reviser, but he did not very often decide for these literal sometimes even Hebraistic renderings. Of the three cases we have seen, the Lucianic text once had the original Old Greek, once a stylistic improvement, and once the early approximation to the Hebrew text. The following two examples will reveal how the Lucianic text sometimes prefers the original longer reading, sometimes decides against such a reading – probably in the interest of the readability of the text.

(5)  1 Sam 1:13 – omission of a genuine Septuagint reading resulting from omission in the MT ‫קֹולּ֖ה ֣ל ֹא יִ ָּׁש ֵ ֑מ ַע‬ ָ ְ‫ו‬

καὶ φωνὴ αὐτῆς οὐκ ἠκούετο· καὶ εἰσήκουσεν αὐτῆς Κύριος. καί 2° – Κύριος] > Β A Ο b f 55 245 707txt Aeth Sa Compl Ra = MT Vorlage: ‫וישמע אליה יהוה‬ Cf. Gen 30:22 ‫ֹלהים‬ ִ֔ ‫יה ֱא‬ ָ֙ ‫ וַ ּיִ ְׁש ַ ֤מע ֵא ֶ ֙ל‬καὶ ἐπήκουσεν αὐτῆς ὁ θεός

26  The B text does not seem to have had a direct contact with the Hebrew, from which follows that the kaige-type readings were most probably excerpted from an actual kaige manuscript. Moreover, the B text used by Origen had to be a Christian text, otherwise he would not have used it to “heal” the text of the Church. For a more detailed discussion, see Aejmelaeus, “Kaige Readings in a Non-Kaige Section in 1 Samuel,” 182–84; Aejmelaeus, “Textual History of the Septuagint and the Principles of Critical Editing,” 172–75; and Aejmelaeus, “Hexaplaric Recension and Hexaplaric Readings in 1 Samuel.”

Challenges in Preparing the Critical Edition of 1 Samuel

21

Hannah was praying silently at the sanctuary in Shilo. “Her voice was not heard, but” – according to the Greek text – “the Lord heard her.” This sentence was obviously inspired by the story of Rachel, who was also suffering from childlessness. The borrowing must however have happened in Hebrew, because the formulation in Greek is different from Gen 30:22. The longer text must have been present in the Vorlage – probably representing the original wording of the Hebrew text. It was removed from the MT by an editor who made several changes in the birth-story of Samuel, obviously with the aim to make Hannah’s vow appear illegitimate.27 Comparison of the Greek text with the shorter Hebrew text led to the omission of the sentence from those Greek manuscripts that witness the early Jewish corrections – that is, the B text followed by a few other manuscripts – and this reading was picked up by Origen for the Hexapla, because it was in harmony with his Hebrew text. Again, omitting parts of the traditional text was not a feature of the Hexaplaric recension, but a typical feature of the kaige revision. The longer text no doubt belongs to the critical text.

(6)  1 Sam 15:29 – a partial omission of a Septuagint reading according to a shorter Hebrew text ‫לֹא יְ ַׁש ֵּקר וְ לֹא יִ ּנָ ֵחם ִּכי לֹא ָא ָדם הּוא ְל ִהּנָ ֵחם‬ καὶ οὐκ ἀποστρέψει (4QSama ‫ )ישוב‬οὐδὲ μετανοήσει,

ὅτι οὐχ ὡς ἄνθρωπός ἐστιν τοῦ μετανοῆσαι· αὐτὸς ἀπειλήσει καὶ οὐκ ἐμμενεῖ;28 αὐτὸς – ἐμμενεῖ cf. Aug CD 17,7 (ipse minatur et non permanet)] pr αὐτός 554: dittmg togr; αὐτός B A O b–121 d–44 Ra; ⲛⲧⲟϥ ⲉϥⲛⲁϭⲱⲛⲧ (≈ αὐτὸς ἀπειλήσει) Sa; quia ipse iratus est Aeth; om αὐτός V 245 707; > L 44 LaM = MT Vorlage? ‫ההוא אמר ולא יקימנה‬ Cf. Num 23:19 ‫ימּנָ ה‬ ֽ ֶ ‫ַה ֤הּוא ָא ַמ ֙ר וְ ֣ל ֹא יַ ֲע ֶׂ֔שה וְ ִד ֶ ּ֖בר וְ ֥ל ֹא יְ ִק‬ αὐτὸς εἴπας οὐχὶ ποιήσει; λαλήσει καὶ οὐχὶ ἐμμενεῖ;

This example reveals a categorical statement that God does not regret or change his mind like human beings do. In a negative statement, the verb μετανοέω seems to pose no problem. But there is a problem with the following sentence, beginning with αὐτός. In fact, in Rahlfs’ edition, the verse ends with the word αὐτός. As usual, Rahlfs follows in this reading Codex Vaticanus B which is accompanied by 121-509 (= b; except for 121mg that has the longer text) as well as A and O and the d group (with the exception of 44). This group of manuscripts – main27  This thesis is discussed in Aejmelaeus, “Was Samuel Meant to Be a Nazirite?” 28  For a more thorough discussion, see Aejmelaeus, “A Kingdom at Stake.” 362–4. See also Aejmelaeus, “Does God Regret? A Theological Problem that Concerned the Kaige Revisers,” 41–53.

22

Anneli Aejmelaeus

ly familiar from the previous examples – makes a partial omission, leaving the pronoun αὐτός in its place – as proof of the omission! The Lucianic text (with manuscript 44 coinciding) leaves out the whole sentence including αὐτός.29 It is possible that the Lucianic reviser made the omission, because he found the sentence hardly comprehensible to those listening to the reading in his Christian community. No doubt, the longer text is original in the Septuagint – and thus part of the critical text. It is present in the majority of the Greek witnesses and is supported by several daughter versions (Aeth, Sa, La through Aug CD). For some reason, it is not represented in the MT, although it must have been in the Vorlage of the Septuagint. The formulation is somewhat puzzling. No one would add a sentence like this, but omitting it is understandable. Essential for its interpretation is that it is a question, which underlines the unchangeableness of God’s decisions, in this case the decision to reject Saul: “Should he threaten and not keep it?”30 The fact that αὐτός was left in its place in the group of manuscripts that witness the kaige-type corrections shows indisputably that the longer text is primary. Comparing the longer Greek text with the shorter Hebrew text, it was perhaps not so easy to see which words should be omitted. Whether the error originated with the kaige reviser, or with the scribe behind the B text who decided to adopt the shorter reading while copying the text, is impossible to know for sure – however, I find the latter more plausible. Since I mentioned that the translator was not quite up to his task, I have to present a case where he made an error and this error was corrected by those early Jewish revisors.

(7)  1 Sam 30:21 – an erroneous translation and its correction ‫ר־ּפּגְ ֣רּו׀ ִמ ֶּל ֶ֣כת׀ ַא ֲח ֵ ֣רי ָד ִ ֗וד‬ ִ ‫אתיִם ָה ֲאנָ ִׁ֜שים ֲא ֶ ֽׁש‬ ַ֙ ‫ל־מ‬ ָ ‫וַ ּ֣יָב ֹא ָד ִ ֗וד ֶא‬

Καὶ παραγίνεται Δαυὶδ πρὸς τοὺς διακοσίους ἄνδρας τοὺς ὑπολειφθέντας τοῦ πορευθῆναι ὀπίσω Δαυίδ ὑπολειφθέντας] ἀπολειφθέντας L–93; ἐκλυθέντας B A O b Sa Ra = MT | πορευθῆναι] πορεύεσθαι B A a b 64 342 460

When chasing after the Amalekites, David had divided his troops and left 200 men behind. When he comes back to them, the Hebrew text refers to these men as those “who were too exhausted to go after David.” The translator made a contextual guess: “who were left behind to go after David,” giving the impression 29  Manuscript 44 probably follows the shorter text of the d group but leaves out the otiose pronoun, thus coinciding with the Lucianic reading. 30  If not understood as a question, it must be connected with the human being: “He threatens and does not keep.”

Challenges in Preparing the Critical Edition of 1 Samuel

23

that they were to follow David later. The Hebrew verb used here, ‫ פגר‬pi. ‘to be exhausted,’ is a rare word, occurring only here and earlier in v. 10 of the same chapter. The translator obviously did not know the word and made a different contextual guess in each case. At the first occurrence in v. 10, the translator rendered ‫ת־נ ַ֥חל‬ ַ ‫ׁשר ִּפּגְ ֔רּו ֵמ ֲע ֖בֹר ֶא‬ ֣ ֶ ‫ֲא‬ ‫ ַה ְּב ֽׂשֹור‬as οἵτινες ἐκάθισαν πέραν τοῦ χειμάρρου τοῦ Βοσόρ. Instead of “too weary to cross the brook,” the Greek text says, “remained on the other side of the brook,” which does not change the story too much, but is in fact a false rendering. For v. 10, no correction has been transmitted,31 but in v. 21 we find ἐκλυθέντας (from ἐκλύω pass. ‘to become weary’) in a group of manuscripts already familiar to us: B and A and the groups O and b. It seems that this alteration caused another change in the following infinitive, at least in B A b: the verb “to become weary” seems to function better with the infinitive in the present πορεύεσθαι, but not all manuscripts follow the same pattern, due to the eclectic nature of most of the manuscripts. Rahlfs’ edition follows the revised text in the participle but not in the infinitive. The critical text is ὑπολειφθέντας, and ἐκλυθέντας reflects early kaige-type approximation to the Hebrew text. The readings of the Three that have been transmitted for these cases confirm that the correction is pre-Hexaplaric.32

6. Conclusion With these few examples from the text of 1 Samuel I have tried to give a glance at the kinds of text-historical problems that an editor of the critical text of a book like the Book of Samuel needs to deal with. I have concentrated on cases that exemplify the new text-historical insights and the methodology used to unravel problem cases that are largely related to the idiosyncrasies of the translator, to early revisional history, and/or to differences of the Vorlage in comparison to the later proto-MT.

7. Bibliography Aejmelaeus, Anneli, “The Septuagint of 1 Samuel,” in VIII Congress of the International Organization for Septuagint and Cognate Studies, Paris 1992, eds. L. Greenspoon and O. Munnich. Septuagint and Cognate Studies 41 (Atlanta, Georgia: Scholars Press, 1995) 109–129. Also published in On the Trail of the Septuagint Translator: Collected Essays. Revised and 31  Nevertheless, the readings of the Three have been transmitted for 1 Sam 30:10: α′ οἳ ἠτόνησαν σ′ ἠδυνάτησαν θ′ ἀπενάρκησαν 243-731(s nom). 32  The readings of the Three for 1 Sam 30:21 are slightly different from those for the parallel verse: α′ οἳ ἐπτωματίσθησαν σ′ ἀτονήσαντας θ′ ἀποναρκήσαντας 243-731(s nom).

24

Anneli Aejmelaeus

Expanded Edition. Contributions to Biblical Exegesis and Theology 50 (Leuven: Peeters, 2007), 123–141. —, “David’s Return to Ziklag: A Problem of Textual History in 1 Sam 30:1,” in XII Congress of the International Organization for Septuagint and Cognate Studies, Leiden 2004, ed. M. Peters. SCS 54 (Leiden/Boston: Brill, 2006), 95–104. —, “A Kingdom at Stake: Reconstructing the Old Greek – Deconstructing the Textus Receptus,” in Scripture in Transition: Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo, eds. A. Voitila and J. Jokiranta. JSJS 126 (Leiden: Brill, 2008), 353–366. —, “What Rahlfs Could not Know: 1 Sam 14,4–5 in the Old Greek,” in After Qumran: Old and Modern Editions of the Biblical Texts – The Historical Books, eds. H. Ausloos, B. Lemmelijn, and J. Trebolle Barrera. Bibliotheca Ephemeridum Theologicarum Lovaniensium 246 (Leuven – Paris – Walpole, MA: Peeters 2012), 81–93. —, “When Did the Books of Samuel Become Scripture?” in From Author to Copyist: Essays on the Composition, Redaction, and Transmission of the Hebrew Bible in Honor of Zipi Talshir, ed. Cana Werman (Winona Lake, IN: Eisenbrauns, 2015), 263–281. —, “Where Do Doublets Come from? A Problem of the Septuagint of 1 Samuel,” in Biblical Greek in Context: Essays in Honour of John A.L. Lee, eds. J.K. Aitken and T.V. Evans. Biblical Tools and Studies 22 (Leuven/Paris/Bristol, CT: Peeters, 2015), 9–19. —, “Textual History of the Septuagint and the Principles of Critical Editing,” in The Text of the Hebrew Bible and Its Editions: Studies in Celebration of the Fifth Centennial of the Complutensian Polyglot, eds. A. Piquer Otero and P. Torijano Morales. Supplements to the Textual History of the Bible 1 (Leiden/Boston: Brill, 2016), 160–179. —, “Does God Regret? A Theological Problem that Concerned the Kaige Revisers,” in The Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila, eds. A. Aejmelaeus and T. Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017), 41–53. —, “Kaige Readings in a Non-Kaige Section in 1 Samuel,” in The Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila, eds. A. Aejmelaeus and T. Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017), 169–184. —, “Was Samuel Meant to Be a Nazirite? The First Chapter of Samuel and the Paradigm Shift in Textual Study of the Hebrew Bible,” in Textus 28 (2019): 1–20. —, “Hexaplaric Recension and Hexaplaric Readings in 1 Samuel,” in On Hexaplaric and Lucianic Readings and Recensions, eds. Dionisio Candido, Joshua Alfaro and Kristin De Troyer. DSI (Göttingen: Vandenhoeck & Ruprecht, forthcoming). Barr, James, The Typology of Literalism in ancient biblical translations. MSU XV (Göttingen: Vandenhoeck & Ruprecht, 1979). Barthélemy, Dominique, Les devanciers d’Aquila. VTSup 10 (Leiden: Brill, 1963). Lagarde, Paul Anton de, Anmerkungen zur griechischen Übersetzung der Proverbien (Leipzig: Brockhaus, 1863). Lee, John A. L., A Lexical Study of the Septuagint Version of the Pentateuch. SCS 14 (Chico, Ca.: Scholars Press, 1983).

Challenges in Preparing the Critical Edition of 1 Samuel

25

Rahlfs, Alfred, Septuaginta-Studien I–III. 2. Auflage. Vermehrt um einen unveröffentlichten Aufsatz und eine Bibliographie mit einem Nachruf von Walter Bauer (Göttingen: Vandenhoeck & Ruprecht, 1965 [original publication I 1904; II 1907; III 1911]). Swete, Henry Barclay, An Introduction to the Old Testament in Greek. Second Edition, Revised by R. R. Ottley (Cambridge: University Press, 1914 [original publication 1900]). Thackeray, Henry St. J., “The Greek Translation of the Four Books of Kings,” JTS 8 (1907): 262–278. Tov, Emanuel, The Greek Minor Prophets Scroll from Naḥ al Ḥ ever (8Ḥ evXIIgr). Discoveries in the Judean Desert VIII (Oxford: Clarendon Press, 1990). Wirth, Raimund, Die Septuaginta der Samuelbücher: Untersucht unter Einbeziehung ihrer Rezensionen. DSI 7 (Göttingen: Vandenhoeck & Ruprecht, 2016). —, “Das Praesens Historicum in den griechischen Samuelbüchern,” in In the Footsteps of Sherlock Holmes: Studies in the Biblical Texts in Honour of Anneli Aejmelaeus, eds. Kristin De Troyer, T. Michael Law and Marketta Liljeström. CBET 72 (Leuven: Peeters, 2014), 117–132. —, “Dealing with Tenses in the Kaige Section of Samuel,” in The Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila, eds. A. Aejmelaeus and T. Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017), 185–197.

Paavo Huotari Discovering Old Greek Readings in the Lucianic Text of 2 Samuel by comparison with 4QSama, c–Wishful Thinking? 1. Introduction At least since the Qumran discoveries it has become clear that the Septuagint (LXX) is an important textual witness of the Hebrew Bible. It offers a glimpse of a Hebrew text which is often older than the tradition of the Masoretic text (MT).1 Several agreements between the Qumran scrolls and the LXX against the MT prove that the translators of the LXX, in general, translated their contemporary Hebrew text literally. The literal translation is evident particularly in the Historical books, which show traces of a literal translation technique. The translator(s) of Samuel were able to translate small details in Hebrew with good Greek but did not always recognize the larger syntactical structures of the Hebrew text.2 Shortly after the Qumran findings, another discovery was made at Naḥ al Ḥ ever, close to Qumran, in 1952: the scroll 8Ḥ evXIIgr, which contains a Greek translation of the Minor Prophets.3 Dominique Barthélemy identified the text as a Greek text that was revised according to a Hebrew text similar to the proto-MT.4 Similarly, he identified this revised Greek text or the so-called Hebraizing revision (kaige) in the Historical books, particularly in two sections (2 Sam 11:2–1 Kgs 2:11; 1 Kgs 22 – 2 Kgs), which earlier were thought to represent different translation techniques. Against the Greek Majority Text (Maj.) which was now influenced by kaige, it is the Lucianic text (the L-text) that preserves the Old Greek (the OG) in the kaige section.5 The relationship of 4QSama, c to the L-text is an intriguing subject: It is an early Hebrew version(s), which may correspond to the OG attested in the L-text against the MT and the tradition of Maj. This relationship raises several questions regarding what has happened to the text. Three observations must be taken into account before drawing any conclusions. First, it cannot be argued that the L-text preserves the OG just because it offers unique readings in the kaige section in comparison to other Greek witnesses. It may preserve the OG because Maj. or some witnesses are influenced by kaige. Since the influence of kaige is stronger in 1  See, for example, De Troyer, Rewriting the Sacred Text; Aejmelaeus, “What Can We Know about the Hebrew Vorlage of the Septuagint?,” 71–106; Tov, The Text-Critical Use of the Septuagint in Biblical Research, 201–38. 2 Aejmelaeus, “The Septuagint of 1 Samuel,” 140–1. 3 Tov, The Greek Minor Prophets Scroll from Naḥ al Ḥ ever. 4  Barthélemy, Devanciers. 5 Barthélemy, Devanciers, 34–41.

28

Paavo Huotari

the kaige section, the instances in which the L-text preserves the OG occur more there than in the non-kaige section. Second, the L-text is also a recensional text that is not related to kaige at all. Thus, the L-text attests several recensional features that occur in both sections. Third, the nature of the L-text is revealed only when it is compared to Maj. It is not only the Codex Vaticanus, although a good witness to the kaige revision, but Maj. which is influenced by kaige. In this study, all agreements between the L-text of 2 Samuel and 4QSama, c are studied by taking into account all the Greek witnesses. Without these three premises, one may come to the simplistic solution that the L-text is the OG in the kaige section and the Codex Vaticanus and/or Maj. is the OG in the non-kaige section.

2. Starting points 2.1  The Qumran Texts of 2 Samuel

The Hebrew text of 2 Samuel is preserved in two Qumran manuscripts 4QSama and 4QSamc. 4QSama is dated around 50–25 BCE. The text is elegant and written in late Hasmonean or early Herodian script, probably the work of a skilled scribe. The scroll itself, although extensive, is relatively fragmentary with hundreds of fragments from all over 1–2 Samuel.6 4QSamc is paleographically dated around 100–75 BCE. It contains one long passage of 2 Samuel (14:7–15:15) presented in three columns and one fragment of 1 Samuel (25:30–32). It is a product of the same scribe who copied the Rule of the Community (1QS) and the Testimonia (4Q175).7

2.2 The Kaige revision

The beginning of the kaige section in 2 Samuel is debated. In this study, it is assumed that the kaige revision starts with 2 Sam 10:6.8 The identification of a kaige reading is extremely important when reconstructing the OG, particularly in the kaige section. Furthermore, the kaige revision has sporadically influenced witnesses in the non-kaige section as well.9 In general, a kaige reading corresponds with the MT text. A kaige reading characteristically 1) repeats the same number of elements as the Hebrew text, 2) renders a given Hebrew lexeme with the same Greek equivalent and 3) produces the basic mean6  Frank Moore Cross et al., Qumran Cave 4.XII:1–2 Samuel, DJD XVII (Oxford: Clarendon, 2015), 2–28; See also Seppänen, “The Hebrew Text of Samuel: Differences in 1 Sam 1 – 2 Sam 9 between the Masoretic Text, the Septuagint, and the Qumran Scrolls,” 3. 7  DJD XVII, 247–54. 8  See Wirth, “Dealing with Tenses in the Kaige section of Samuel,”, 197: “Lexical and syntactical observations in 2 Samuel 10 prove that […] detailed analysis can show that the line between non-kaige and kaige goes between 2 Sam 10:5 and 10:6.” 9 Aejmelaeus, “Kaige readings in a Non-Kaige Section in 1 Samuel,” 169–84.

Discovering Old Greek Readings in the Lucianic Text

29

ing of the Hebrew word. A kaige reading is also 4) found more often in the kaige section than outside it, and 5) in the translations of Aquila and Theodotion.10 In Samuel, the kaige revision is best attested in the following witnesses: B, b (121509), 68′ (68-122), 64′ (64-381), 55, 244, 245, 460, 707. Yet, in the kaige section the previous witnesses are often joined by Maj. (see below) when preserving the kaige readings.11

2.3  The Lucianic text – The Two Layers

The Lucianic text (the L-text) of Samuel-Kings is attested in the following five Greek witnesses 19 82 93 108 127.12 In the Göttingen editions, they constitute the L-group, which is divided into two sub-groups: 19′ (19 108) and L-19′ (82, 93 and 127). The L-text is often followed by the Greek witnesses Mmg, V, 158, 245s, 318, 460 and 554, which share some Lucianic readings to a widely varying degree.13 The L-text contains two textual layers.14 The earlier layer is the Vorlage and the unrevised Greek text of the Lucianic reviser. This earlier layer is sometimes called the proto-Lucianic text but to avoid any confusion with the proto-Lucianic recension (see below), I will call it the base text of the Lucianic reviser. The base text attests the Greek text that has, against B (+) and Maj., escaped the influence of the kaige revision.15 The later layer complicates the identification of the base text. The later layer is the Lucianic revision produced by the reviser at the beginning of the 4th c. CE. The Lucianic revision can still be divided into 1) the revisional elements and 2) the Hexaplaric approximations (see below). In order to investigate the base text of the Lucianic reviser (≈OG), the Lucianic revision needs to be identified and removed. The identification requires knowledge of the revisional elements and the Hexaplaric text. A Lucianic reading aims to produce a logically and stylistically better text that reflects the rules of Attic Greek.16

10  See the lexicographical and external criteria for a kaige reading introduced by Kauhanen, “Lucifer of Cagliari and the Kaige Revision,” 147. 11  Tuukka Kauhanen, “The Best Greek Witnesses,” 561–73. 12  The Cambridge edition of The Old Testament in Greek (Brooke-McLean) uses the following signs for the Lucianic MSS: b’ (19) b (108) o (82) e2 (93) c2 (127). 13 Kauhanen, “The Best Greek Witnesses.” 14 Metzger, Chapters, 24–27; Fernández Marcos, Septuagint, 232–36. 15 Barthélemy, Devanciers, 126–27. See also Kauhanen, Best Greek Witnesses, 564: “Among the notable features of this textual tradition is that it attests only very few kaige readings. According to the standard theory, the base text of the revision, the proto-Lucianic text, was a good, old text-type that for a large part escaped the kaige revision”. 16 Brock, Recensions, 224–99. See also Metzger, Chapters, 30–35; Fernández Marcos, Septuagint, 230–32.

30

Paavo Huotari

2.4  The Proto-Lucianic recension?

The discovery of the Qumran scrolls and of the kaige revision and the new awareness of the L-text’s nature were a trigger for new research. One particular interest was possible agreements between the witnesses which precede Lucian of Antioch, such as 4QSama, c, and their relationship with the L-text. On the basis of several assumed agreements, Frank M. Cross suggested that the L-text shows recensional elements other than the Lucianic revision. Cross attempted to interpret these early revisional elements as traces of the proto-Lucianic recension, which represents an even earlier Hebraizing revision than the kaige revision. Instead of the proto-MT behind the kaige revision, Cross argued that the proto-Lucianic recension was revised toward the Hebrew text attested in 4QSama.17 The theory of the proto-Lucianic recension has since been rejected.18 From this perspective, the L-text would not contain traces of an early Hebraizing revision but traces of the OG itself.19 The rejection of the proto-Lucianic recension, therefore, is not a denial of the agreements between 4QSama, c and the L-text. Instead of a proto-Lucianic recension, it would be preferable to speak of agreement between individual readings of the OG attested in the L-text and the Qumran scrolls 4QSama, c.

2.5  The Hexaplaric text

The Hexaplaric text attests several additions and a few omissions according to the Fifth Column of Origen’s Hexapla. Origen’s contemporary Hebrew text was similar to the MT. His aim of “healing” the Septuagint, therefore, did produce a Greek text different from the OG.20 The Hexaplaric text of Samuel is witnessed by A (the Codex Alexandrinus) and O (247-376). Also, the L-text is often influenced by the Hexapla, either by the Fifth Column or the Jewish translations, particularly Symmachus.21 The Hexaplaric approximations in the L-text also explain some agreements between the L-text and the Hebrew text. Some agreements between the L-text and 4QSam, therefore, may be only coincidental as a result of the Hexaplaric influence on the L-text.

17  F. M. Cross, “The history of the Biblical text in the Light of Discoveries in the Judaean Desert,” HTR 57 (1964): 281–99. 18  See especially Kauhanen, The Proto-Lucianic Problem in 1 Samuel. 19 Aejmelaeus, “The Septuagint of 1 Samuel,” 127: “The proto-Lucianic recension is a hypothesis created to fit another hypothesis, the neat pattern of the theory of local texts, with little practical significance.” See also Tov, “Lucian and Proto-Lucian: Toward a New Solution of the Problem,” 103: “The second layer is the historical Lucian, and I suggest that its substratum contained either the Old Greek translation or any Old Greek translation.” 20  Ep. Afr. (5–9); Comm. Matt (15.14) 21 Brock, Recensions, 170–173; Albrecht, “Von der hebraica veritas zur vera graecitas: ­Origenes – Hesych – Lukian,” 105–41 (126–27).

Discovering Old Greek Readings in the Lucianic Text

31

2.6  The Majority Text and its relation to other textual traditions

The Majority text (Maj.) is a loose text tradition that is preserved in the Greek witnesses that do not share distinctive traces of other traditions, such as the Lucianic revision or the Hexaplaric text. Maj. is particularly witnessed by M, CI (98243-379-731), 74′ (74-106-120-134-370), s−64′ (92-130-314-488-489-762), 29, 71. In the non-kaige section, Maj. does not follow B (+) when the latter shows traces of kaige. In the kaige section, however, Maj. often follows the kaige witnesses. Therefore, the kaige revision cannot be reduced to B (+) only or one would underestimate the phenomenon. In the kaige section, it is Maj., in addition to B (+), that is influenced by kaige. It is essential to understand and remember that the L-text has gone through a more extended Lucianic revision in both sections than B (+) and Maj. are influenced by kaige. Thus Maj. is a quantitatively better witness of the OG in the kaige section as well.22 However, when B (+) and Maj. are influenced by kaige – and only then – the L-text is the better witness of the OG. Because the influence of kaige is more evident in the kaige section, the number of OG readings in the L-text against B (+) and Maj. is greater in the kaige section than the non-kaige section.23 Because of this difference in number, the value of the L-text has sometimes been overestimated.24 In the kaige section, the OG may occasionally be lost under the Lucianic revision of the L-text and the influence of kaige in B (+) and Maj.

3. Data The focus of this study is 2 Samuel, of which the kaige section and the non-kaige section provide an excellent reference point.25 I have included all the Lucianic readings in 2 Samuel that agree or at least seem to agree with the Qumran scrolls 4QSama and 4QSamc, but with the following criteria: 1) The reading in 4QSama and 4QSamc is not totally reconstructed, but visible, 2) the reading in 4QSama and 4QSamc is against the MT and 3) the reading in the L-text is against Maj. Using this definition, I found 25 agreements in the L-text with 4QSama and seven agreements with 4QSamc. What makes this significant is that, of the 32 agreements, 30 are found in the kaige section. This could highlight the nature of the L-text in the kaige section. However, if the agreements do not have a common origin, they are not really agreements at all. The difference in the numbers could also be caused by the fragmentary nature of 4QSama and 4QSamc. 22 Kauhanen, “Best Greek Witnesses,” 565, 573. 23 Ibid. 24  See, for example, Barthélemy, Devanciers, 35. 25  A similar study of 1 Samuel has already been done. See Kauhanen, The Proto-Lucianic Problem in 1 Samuel, 165–88.

32

Paavo Huotari

In their article “A Statistical analysis of the Textual Character of 4QSamuela” from 2006, Cross and Saley have suggested that the manuscript 4QSama preserves 41 readings26 that agree with the L-text.27 The difference of 16 agreements between their study and this presentation is due to the application of different criteria. First, Cross and Saley have included agreements in which the L-text is against the Codex Vaticanus but not Maj. There are eight agreements in which the L-text agrees with 4QSama, but is also supported by Maj. against the MT and B: 12:16

13:16 13:39

17:25

18:11 21:4 24:171 24:172

‫ בשק ארצה‬4QSama; ‫ ָא ְר ָצה‬MT

ἐν σάκκῳ ἐπὶ τὴν γῆν V M 376 L CI a−527 f 64′ 29 55 71 158 244 245 318 342 372 460] καὶ ἐκοιμήθη ἐπὶ τῆς γῆς A; ἐπὶ τῆς γῆς B b 68′ 707; pr καὶ ἐκοιμήθη rel ]‫ [אל] ֯א ֯ח[י‬4QSama; ‫ ַאל־אֹוד ֹת‬MT μή ἄδελφε L Maj.] περὶ B A O 509 71 ‫ [רו]ח המלך‬4QSama; ‫ ָּדוִ ד ַה ֶּמ ֶלְך‬MT τὸ πνεῦμα τοῦ βασιλέως Δαυιδ L; τὸ πνεῦμα τοῦ βασιλέως Maj.] ὁ βασιλεὺς δαδ B O 527 509 55; δαδ ὁ βασιλεὺς A a−527 488 245 ׄ‫[י]שי‬ ֯ 4QSama; ‫ נָ ָחׁש‬MT ιεσσαι L Maj.; ιεσαι 46 74; ιεσσαι (litt ιε sup ras 707); eesse La115] ναας B A 247 509 55* 244c 318; αας 460; ασα 700 ‫חמ]שים‬ ֯ [ 4QSama; ‫ ֲע ָׂש ָרה ֶכ ֶסף‬MT πεντήκοντα σίκλους ἀργυρίου L Maj.] δέκα ἀργυρίου (-ους O) B A O ]‫ ֯מ[כול ישראל‬4Qsama; ‫ ְּביִ ְׂש ָר ֵאל‬MT ἐκ παντὸς Ισραηλ L Maj.] ἐν ισραηλ B A 247 509 ‫הרעה‬ ׄ 4QSama; > MT ὁ ποιμὴν L Maj.] > B 55 ‫הרעתי‬ ׄ 4QSama; ‫יתי‬ ִ ֵ‫ ֶה ֱעו‬MT ἐκακοποίησα L Maj.] ὁ κακοποιήσας CI 489 29 244; καὶ ἠδίκησα 509; > B 55

These agreements do not contribute to the question addressed in this study. When studying the uniqueness of the L-text, the vast majority of the Greek witnesses should be against the L-text. However, all agreements probably attest the OG reading (the L-text adds Δαυιδ in 13:39, which is a minor Hexaplaric addition [ὁ βασιλεὺς δαδ B O 527 509 55; δαδ ὁ βασιλεὺς A]).

26  I need to emphasize that I am not certain that my 25 (+16 rejected) agreements are the same as those suggested by Cross and Saley. 27  Cross and Saley, “A Statistical Analysis of the Textual Character of 4QSamuela (4Q51).”

Discovering Old Greek Readings in the Lucianic Text

33

Second, another eight readings, although only partially reconstructed in 4QSama, are based on such fragmentary evidence that a visual agreement is already uncertain. 2:7

6:9

11:6 11:8 12:1 13:37

18:9 23:4

]‫ליהם ל[מלך‬ ׄ ‫ ׄע‬4QSama; ‫יהם‬ ֶ ‫ ְל ֶמ ֶלְך ֲע ֵל‬MT

ἐφ᾽ ἑαυτοῖς εἰς βασιλέα L] ἐφ᾽ ἑαυτοὺς εἰς βασιλέα Maj.; ἐφ᾽ ἑαυτόν εἰς βασιλέα B 376 b ]‫]ארון יהוׄ [ה‬ ֯ ‫ [הארון ויבוא‬4QSama; ‫ יְ הוָ ה ֲאֹרון‬MT ἡ κιβωτὸς τοῦ θεοῦ καὶ ἦλθεν ἡ κιβωτὸς τοῦ κυρίου L] ἡ κιβωτὸς κυρίου Maj. (=B) ]‫ [א]ל[יו‬4QSama; ‫ל־ּדוִ ד‬ ָ ‫ ֶא‬MT πρὸς αὐτόν L] πρὸς Δαυιδ Maj. (=B) ]‫ מ[פני המלך‬4QSama; ‫ ִמ ֵּבית ַה ֶּמ ֶלְך‬MT ἐκ προσώπου τοῦ βασιλέως L] ἐξ οἴκου τοῦ βασιλέως Maj. (=B) ‫ נ[תן‬4QSama; > MT Ναθαν L] > Maj. (=B) ]‫ בא[רץ חי]ל[ם‬4QSama; > MT εἰς γῆν Χαλααμα L] εἰς γῆν Χαμααχαδ (≈) Maj.; εἰς γῆν Μαχαδ B A O 55* 460 ‫[באל]ה‬ ֯ ‫יתל‬ ֯ ֯‫ ו‬4QSama; ‫ וַ ּיֻ ַּתן‬MT καὶ ἀνεκρεμάσθη ἐν τῷ δένδρῳ L] καὶ ἐκρεμάσθη Maj. (=B) ]‫כד[שא‬ ֯ ‫[ממ]טר‬ ֯ 4QSama; ‫ ִמ ָּמ ָטר ֶּד ֶׁשא‬MT ὡς ὑετός ὡς βοτάνη L] καὶ ὡς ἐξ ὑετοῦ χλόης Maj. (=B)

Regarding 4QSamc, Ulrich suggests nine agreements between the L-text and 4QSamc.28 The difference of two readings between his study and this one is also the result of fragmentary evidence that makes the agreement visually uncertain although only partially reconstructed.29 14:9

‫ממלכ]תו֯ [ נ]קי‬ ֯ ‫‏[וכסא‬4QSamc; ‫ וְ ִכ ְסאֹו נָ ִקי‬MT

15:12

‫ וה]יׄ ֯ה‬4QSamc; ‫ וַ יְ ִהי‬MT

καὶ ὁ θρόνος τοῦ βασιλέως ἀθῷος L] καὶ ὁ θρόνος αὐτοῦ ἀθῷος Maj. (=B) καὶ ἦν L] καὶ ἐγένετο Maj. (=B)

28  DJD XVII, 254. 29  Similarly, I need to emphasize that I am not certain that my seven (+ 2 rejected) agreements are the same as those suggested by Ulrich.

34

Paavo Huotari

4. The Criteria and the Evaluation of the Agreements In this study, I have compared the readings of the L-text with 4QSama, c in order to find real agreements that are against other Greek witnesses (Maj.) and the MT. Next, I have aimed to determine that these agreements are either traces of the proto-Lucianic recension (Cross), or evidence of the OG in the L-text as in 1 Samuel (Kauhanen). Consequently, when the L-text is in real agreement with the 4QSama, c reading against both Maj. and the MT, there are two mutually exclusive explanations: – L-reading = Proto-Lucianic recension (Cross) If so, the reading in the L-text should show secondary elements that precede Lucian of Antioch. At the same time, 2) the competing reading in Maj. or in other witnesses should attest the OG, or, in a rare case, it represents the kaige and the OG reading is lost. If the L-text represents the proto-Lucianic recension, agreements between the L-text only and the Qumran MSS are expected in both sections. – L-reading = OG If so, the reading in the L-text should not show recensional elements such as Lucianic or Hexaplaric. At the same time, the competing reading in Maj. and perhaps in other witnesses should show secondary elements such as kaige. Because Maj. is influenced by kaige particularly in the kaige section, the agreements between the L-text and 4QSama, c are therefore expected particularly in this section. Based on the criteria presented above, all agreements between the L-text and 4QSama and 4QSamc are classified into the following categories:30 – Real agreement: The agreement of two readings goes back to a common origin. – Probable agreement: The agreement of two readings probably goes back to a common origin. Probable agreement is often based on agreement in a minor detail, which reduces the likelihood of a common origin since the agreement could also have come about polygenetically. – Coincidental agreement: The agreement goes back to a common origin, but 4QSama, c need not be the Hebrew base text of the L-text. For example, the L-text may attest a Hebrew text similar to Qumran but it may be a Hexaplaric approximation in the L-text. – Pseudo-agreement: The agreement is only apparent due to revisional activity of a later scribe/reviser.

30  This classification is inspired by, although slightly different from, the terminology in Kauhanen, The Proto-Lucianic Problem in 1 Samuel, 28–29.

35

Discovering Old Greek Readings in the Lucianic Text

– Uncertain agreement: The notion of agreement is based on an observation which may signal an real or only pseudo-agreement. Often this concerns inner-Greek or inner-Hebrew variation, so-called polygenetic changes or errors, which may happen in any witness anytime. In this study, the polygenetic changes particularly concern conjunctions and tenses.

5. Discussion Agreement Real

Probable

Coincidental

Pseudo

Uncertain

Total

4QSama = L 3 (1)31 4QSamc = L 2 Total 5

4 2 6

3 – 3

7 (1) – 7

8 3 11

25 7 32

Of the 32 agreements (see full appendices at the end), five agreements are real and six are probable (= eleven). In all eleven agreements, the L-text probably preserves the OG, whereas Maj. attests a secondary reading, either a kaige reading or some other early correction. This confirms the previous observations about the proto-Lucianic recension. 4QSama and 4QSamc do not provide any evidence of agreements in secondary readings to prove the existence of an entire revision toward the Hebrew text attested in 4QSama and 4QSamc. Four real agreements are in the kaige section and one in the non-kaige section, which is what could be expected. However, as only two of the suggested agreements are in the non-kaige section, one cannot evaluate whether there is a statistically significant difference between the sections. Furthermore, the Qumran scrolls’ fragmentary nature also impacts what material is available to us. Nevertheless, five real and six probable agreements between 4QSama and 4QSamc and the L-text show individual instances (ten in the kaige section) in which the L-text preserves the OG (4QSam) against the MT and Maj. In the following, all five real agreements are studied case by case with other agreements, if in the same passage. At the end, I provide full appendixes of all 32 agreements with a classification.

31  A number in parentheses is the number of readings in the non-kaige section.

36

Paavo Huotari

2 Sam 5:11 4QSama L-textSam

MTSam Maj.Sam

AO CI−243txt

La115 Sam

MTChr Maj.Chr

καὶ ׄ‫[ו]ח ֯רשי‬ ֯ τέκτονας ‫ עץ‬ξύλων

καὶ ‫ וְ ָח ָר ֵׁשי‬τέκτονας ‫ ֵעץ‬ξύλων

καὶ τέκτονας ξύλων

et fabros lignarios

καὶ ‫ וְ ָח ָר ֵׁשי‬οἰκοδόμους

καὶ ‫ וחרשי‬τέκτονας

καὶ ‫ וְ ָח ָר ֵׁשי‬τέκτονας

καὶ τέκτονας

et structores

καὶ ‫ וְ ָח ָר ֵׁשי‬τέκτονας

λίθων τοίχου

parietum

‫קיר‬ ֯ τοίχου (+λίθων 19′)

‫ ֶא ֶבן‬λίθων ‫ִקיר‬

‫ ִקיר‬τοίχων

‫ ֵע ִצים‬ξύλων

L-textChr

καὶ τέκτονας τοίχου καὶ τέκτονας ξύλων

As the table above shows, the text has multiple variations in both the Hebrew and Greek witnesses in 2 Sam 5:11. DJD (Cross, Parry and Saley) suggests one agreement here.32 4QSama reads “and craftsmen of wall” whereas MTSam reads “and craftsmen of wall stone.” L-textSam and La115 agree totally with 4QSama whereas Maj.Sam reads “craftsmen of stone.” The Greek witnesses 247, 376 (= O) and the Codex Alexandrinus (A) preserve a longer reading agreeing totally with MTSam. The different witnesses of Chronicles provide a parallel text, similar to the text in Samuel but with a different word order: “the craftsmen of wood” precedes “the craftsmen of wall.” Except for the word order, all Chronicles witnesses in the table agree with 4QSama against MTSam and Maj.Sam. Only Maj.Chr reads οἰκοδόμους instead of τέκτονας. This variation of readings provides multiple alternative options for what has happened to the text. The first immediate observation is that the agreement between 4QSama and L-textSam probably is real regarding translation equivalents. In the LXX, ‫ קיר‬is generally translated with τοῖχος but never with λίθος. On the other hand, λίθος is the general translation equivalent for ‫אבן‬, which is never translated with τοῖχος. The second immediate observation is that MTSam does not agree with Maj.Sam. Of the three options (4QSama: ‫ ;קיר‬Maj.Sam: λίθων = ‫ ;אבן‬MTSam: ‫)אבן קיר‬, MTSam preserves the longest reading, which is probably a conflated reading.33 This conflated reading agrees with the Hexaplaric witnesses A O and CI−243txt, which prob32  DJD XVII, 122. 33 McCarter, II Samuel, 144.

Discovering Old Greek Readings in the Lucianic Text

37

ably preserve the Hexaplaric addition according to the Hebrew text attested in MTSam. The Greek witness 243mg contains an asterisk on the word λίθων which suggests that the LXX text as known to Origen did not read that word. Origen added λίθων according to the Hebrew text attested in MTSam, thus, repeating the conflated reading already existing in the Hebrew text. It must be noted that the Lucianic witnesses also have a variation. The witness group 19′ agrees with A and O regarding the conflated reading but the witness group L−19′ (i.e. MSS. 82 93 127), as the sole Greek witness, preserves only τοίχου. In general, MSS. 82 93 127 should be taken as better witnesses of the L-text if there are no internal reasons to argue otherwise.34 Thus, the earlier reading in the Hebrew text is either ‫( אבן‬Maj.Sam) or ‫קיר‬ (4QSama and L-textSam). One of them is probably a variant reading that has replaced the other. In the end, a variant reading has been chosen and produced a conflated reading that is preserved in MTSam. It is most likely that the OG attests an earlier Hebrew reading. L-reading = Proto-Lucianic recension: If so, ‫ קיר‬has replaced ‫אבן‬. Ulrich argues that the OG is in Maj.Sam on the basis of Lagarde’s principle “furthest from the MT.”35 The L-text, however, is as far away from the MT as from Maj. Ulrich’s choice is based on the number of the word that is plural in Maj.Sam against the singular in MTSam and L-textSam. L-reading = OG: If so, ‫ אבן‬has replaced ‫קיר‬.36 ‫ אבן‬creates a fine parallel with ‫עץ‬, which is even more visible in the plurals of Maj.Sam “the craftsmen of woods and the craftsmen of stones.” On the level of Hebrew, this parallelism may be secondary, and instead of “wall,” “stone” fits slightly better with the narrative of building a palace for David. Externally, L-textSam is supported by the OL witness La115, Chronicles, and the asterisk in 243mg. The Maj.Sam reading λίθων may preserve a kaige-revision in the non-kaige section towards a now lost Hebrew text.37 The plural λίθων may be a slight stylistic choice due to parallelism with ξύλων. Internal reasons and the attestation of witnesses lends more support to the conclusion “The L-text attests the OG.”

34  Busto Saiz, “On the Lucianic manuscripts in 1–2 Kings,” 305–8. 35 Ulrich, The Qumran Text, 100. 36 McCarter, II Samuel, 144; Against Ulrich, Qumran Text, 99; Anderson, 2 Samuel , 80: ‫קיר‬ has replaced the ‫אבן‬. 37  Cross also agrees that Maj. preserves a later reading. See DJD XVII, 122.

38

Paavo Huotari

2 Sam 14:32 4QSamc L-text

‫^ועתה‬ ׄ καὶ νῦν ‎‫[א]^ה‬ ׄ ‫אר‬ ֯ ὄψομαι ] ‫‏נא פנ[י‬δὴ τὸ πρόσωπον ‫ המלך‬τοῦ βασιλέως

MT Maj.

‫ וְ ַע ָּתה ֶא ְר ֶאה‬καὶ νῦν ἰδοὺ ‫ ְּפנֵ י‬τὸ πρόσωπον ‫ ַה ֶּמ ֶלְך‬τοῦ βασιλέως οὐκ εἶδον· 

A O 328 καὶ νῦν ἰδού] ὅτι O | om ἰδού 328 | τὸ – βασιλέως] post εἶδον tr A O

In 2 Sam 14:32, DJD (Ulrich) suggests one agreement: the use of the particle ‫ נא‬in 4QSamc and δή in the L-text, and their absence in the MT and Maj.38 In general, such agreements concerning minor details should be regarded as uncertain. However, the particles δὴ and ‫ נא‬are not very frequent compared to conjunctions. Their agreement, therefore, cannot be disregarded on the basis of inner-language variation. Furthermore, δή is such a common equivalent of ‫ נא‬in Samuel-Kings that the agreement cannot be only pseudo. On the other hand, the L-text is close to both Hebrew texts in not reading ἰδού. The future ὄψομαι and the absence of οὐκ εἶδον in the L-text also agree with the Hebrew hortative ‫אראה‬. Maj. provides the identical meaning but through the past tense, which results in the negative expression “and now behold, the face of the king I have not seen (yet)”. L-reading = Proto-Lucianic recension: If so, δή in the L-text is secondary, agreeing with 4QSamc (‫ )נא‬against MT,39 whereas the absence of it in Maj. would be the OG. However, δή is attested 120 times in Samuel-Kings (69 times in Samuel) with no significant difference between the non-kaige and the kaige sections (52 vs. 68). Thus, the usage of δή fits well with the translator’s toolbox. Furthermore, in the Hebrew text, the addition of ‫ נא‬is in no way evidential. A similar reading ‫ ראה נא‬is found twice in MT2Sam (7:2, 13:28), as is the reading ‫ נא‬+ hortative (14:15; 15:7; 16:9; 17:1 18:19. 22; 24:14). Thus, the use of the particle ‫ נא‬with similar readings is not unique to 4QSama, but is also typical of the Hebrew text attested in the MT. The Hexaplaric witness O lacks ἰδού, and both A and O have a word order agreement with the L-text. Even though the L-text is closer to either Hebrew texts, it is unlikely influenced by Hexaplaric tradition since A O do not feature the verb ὄψομαι and the particle δή.

38  DJD XVII, 262. 39  DJD XVII, 262. See also Auld, I & II Samuel, 498; Ulrich, “4QSamc,” 14.

Discovering Old Greek Readings in the Lucianic Text

39

L-reading = OG: If so, δὴ in the L-text is the OG agreeing with 4QSamc (‫)נא‬, whereas the absence of it in Maj. would be secondary. In this instance, ‫ נא‬is probably an earlier reading, which has dropped away. This may be either an omission of a redundant element or perhaps a minor corruption of the Hebrew text. Maj. probably reflects the Hebrew text ‫הנה‬, which may be a variant reading for the imperative ‫( ראה‬cf. ‫ אראה‬in the MT and 4QSamc). MT2Sam reads ‫ ִהּנֵ ה־נָ א‬twice in a close context (13:24, 14:21); thus, it may be a possible variant here too. If ‫הנה‬ is attested at the beginning of the reading instead of the verb ‫ראה‬/‫אראה‬, the end requires a verb after ‫ =( המלך‬Maj.). Such transposition does not have a direct impact on the particle ‫נא‬, but it may have dropped out at the same time. The transposition, nevertheless, suggests that Maj. is secondary. Internal reasons support the conclusion “The L-text is the OG.” If one argues that the L-text is Hexaplaric and Maj. is the OG, the agreement between the Ltext and 4QSamc would be only coincidental. Either way, this agreement does not provide evidence of the proto-Lucianic recension.

2 Samuel 15:2 4QSama ]‫וה[יה‬ ‫[כ]ול איש‬ ‫‏[אשר יהיה‬ ]‫לו] [ריב‬ ]‫[לבוא‬ ]‫[אל‬ ‫א]ל‬ ֯ ‫[המלך‬ ‫המשפט‬ ‫וקרא‬ ‫לו‬ ‫אבשל[ו]ם‬ ֯ ]‫‏[ואמר‬ ]‫[אי מזה‬ ]‫[עיר אתה‬ ]‫[וענה‬ ]‫[האיש‬ ‫ואמר‬ ‫[ח]ד‬ ׄ ‫מא‬ ׄ ׄ‫שבטי‬ ‫ישראל‬ ]‫[עבדך‬

4QSamc L-text

MT Maj.

]‫ [והיה‬καὶ ἦν ‫ וַ ִיְהי‬καὶ ἐγένετο ]‫ [כול איש‬πᾶς ἀνήρ ‫ל־ה ִאיׁש‬ ָ ‫ ָּכ‬πᾶς ἀνήρ, ]‫ [אשר] [לוא‬ᾧ ἐγίνετο ‫ר־יִהיֶה־ֹּלו־‬ ְ ‫ ֲא ֶׁש‬ᾧ ἐγίνετο ‫ ֯ריב‬κρίσις ‫ ִריב‬κρίσις, ‫לב[ו]א‬ ׄ καὶ ἤρχετο ‫ ָלֹבוא‬ἦλθεν ‫ אל‬πρὸς τὸν ‫ ֶאל־‬πρὸς τὸν ‫ המלך‬βασιλέα ‫ ַה ֶּמ ֶלְך‬βασιλέα ‫ משפט‬εἰς κρίμα ‫ ַל ִּמ ְׁש ָּפט‬εἰς κρίσιν, ]‫ [וקרא‬καὶ ἐκάλει ‫ּיִק ָרא‬ ְ ַ‫ ו‬καὶ ἐβόησεν ]‫ [לוא‬αὐτὸν ‫ ַא ְב ָׁשֹלום‬πρὸς αὐτὸν ]‫ [אבשלום‬Αβεσσαλωμ ‫ ֵא ָליו‬Αβεσσαλωμ ‫ [וא]מר‬καὶ ἔλεγεν ‫אמר‬ ֶ ֹ ‫ וַ ּי‬καὶ εἶπεν ‫ אימזה‬αὐτῷ ἐκ ποίας ‫י־מּזֶ ה‬ ִ ‫ ֵא‬αὐτῷ ᾿Εκ ποίας ‫ עיר אתה‬πόλεως εἶ σὺ ‫ ִעיר ַא ָּתה‬πόλεως σὺ εἶ; ‫ וענה‬καὶ ἀπεκρίνατο ὁ ‫אמר‬ ֶ ֹ ‫ וַ ּי‬καὶ εἶπεν ]‫ [האיש‬ἀνήρ ]‫ [ואמר‬καὶ ἔλεγεν ]‫ [מאחד‬ἐκ μιᾶς ‫᾿ ֵמ ַא ַחד‬Εκ μιᾶς ‫ שבטי‬τῶν φυλῶν ‫ ִׁש ְב ֵטי־‬φυλῶν ]‫ [ישראל‬τοῦ Ισραηλ ‫ יִ ְׂש ָר ֵאל‬Ισραηλ ‫ ‏[עבד]כה‬ὁ δοῦλός σου ‫ ַע ְב ֶּדָך‬ὁ δοῦλός σου

Variants to Maj. ἐγένετο] ἐγίνετο a−527 554 | ἐγίνετο] ἐγένετο B A O 93 CI 509 64′ 244 245 372 460 707 Ra | ἐβόησεν] ἐβόα O 318; ἐκάλει 554mg; ἔλεγεν 372 | αὐτόν] τὸν 82* εἶπεν1] ἔλεγεν B A O a 509 488 55 244 318 460 Ra | εἶπεν2] ἔλεγεν O 318; > 158; + αὐτῷ CII−46′ 313 ἐκ2] pr ὁ ἀνήρ A O CII a 509 d−370 s 55 71 244 245 342 460 554; + πρὸς αὐτόν 707

40

Paavo Huotari

The passage 2 Sam 15:1–12 tells about Absalom’s conspiracy against David. Absalom is standing by the side of the road leading to the city. If someone came with a complaint for David, the king, Absalom would call him aside and administer justice. Nevertheless, he was not able to act officially for he had not been appointed a judge. His example and conduct, however, engage the attention of the people. In this particular verse four agreements have been suggested between the 4QSama/4QSamc and the L-text. This is a good example of agreements, of which only one, however, is real and three are uncertain. DJD (Cross, Parry and Saley) suggests three agreements based on the imperfect forms of the verbs καὶ ἦν, καὶ ἐκάλει, and καὶ ἔλεγεν in the L-text, which would agree with the Hebrew consecutive perfects ]‫וה[יה‬, ‫וקרא‬, and ‫ ואמר‬in 4QSama. Maj. reads those verbs in the aorist καὶ ἐγένετο, καὶ ἐβόησεν, and καὶ εἶπεν whereas the MT has the consecutive imperfects ‫יְהי‬ ִ ַ‫ו‬, ‫וַ ּיִ ְק ָרא‬, and ‫אמר‬ ֶ ֹ ‫וַ ּי‬. There is surely a correlation between the tenses but that is not yet a sign of causality.40 It is known that the OG translator often used the aorist, but sometimes the historical present, when translating the Hebrew consecutive imperfect. The kaige reviser occasionally changed historical presents to aorists to better fit the Hebrew consecutive imperfect.41 The OG translator also used, every now and then, the imperfect when translating the Hebrew consecutive perfect, whereas the kaige reviser also changed some of those to aorist. However, the OG translator may also translate the Hebrew imperfect consecutive with the imperfect. Thus, the L-text probably preserves OG readings in these three verbs in the imperfect, whereas Maj. preserves kaige readings in the aorist. However, the OG translator’s and the kaige reviser’s Hebrew Vorlage may have contained either imperfect consecutives as attested in the MT or perfect consecutives as attested in 4QSama, but the outcome is, nevertheless, the same: that is, three imperfects in the OG (the L-text) and three aorists in the kaige revision (Maj.). Therefore, there is no certainty that 4QSama and the L-text agree against the MT and Maj. The fourth agreement is quantitative in nature.42 Both 4QSama and 4QSamc agree with the L-text “and the man answers and says.” This longer reading is against Maj. “and the man said” and the MT “he said.” The L-text has the longest reading with two verbs and a separate subject. Maj. has one verb and the separate subject. The MT has the shortest reading for it has only one verb without an explicit subject such as ‫האיׁש‬. The subject ὁ ἀνήρ is also omitted in B (+). It must be noted that only one verb is visible in both Qumran manuscripts, ‫ וענה‬in 4QSamc – at line 23 on 40 Ulrich, “4QSamc,” 15; DJD XVII, 154: “L. 1–9 (15:1–6) In these verses, the principal verb forms are modal, habitual, or customary, and especially perfect with waw conversive. […] L generally uses Greek imperfects reflecting the verbal forms of 4QSama. B, generally using aorists, no longer expresses the modal character of the original text, for the most part following MT.” See also Auld, I & II Samuel, 500; Anderson, 2 Samuel, 193; McCarter, II Samuel, 354. Cf. Wellhausen, Der Text der Bücher Samuelis, 194. 41 Voitila, “The Use of Tenses in the L- and B-texts in the Kaige-Section of 2 Reigns,” 213–37. 42 Ulrich, “4QSamc,” 15; DJD XVII, 263.

41

Discovering Old Greek Readings in the Lucianic Text

fragment 5 in column 2 – and ‫ ואמר‬in 4QSama – at line 4 on fragment 113. Considerations of space do not require reconstructing both verbs and ‫ האיש‬in either manuscript: ‫( וענה‬visible in 4QSamc) and ‫( ואמר‬4QSama) could thus simply be variants. L-reading = Proto-Lucianic recension: If so, καὶ ἀπεκρίνατο in the L-text is secondary, agreeing with 4QSamc (‫)וענה‬, and its absence in Maj. would be the OG. This, however, is unlikely (see below). L-reading = OG: If so, καὶ ἀπεκρίνατο in the L-text is the OG agreeing with 4QSamc (‫)וענה‬, and the absence of it in Maj. would be secondary. Ulrich, while arguing that the L-text has preserved the OG, holds that the longer Hebrew reading is an expansion.43 However, it is more likely to be the original Hebrew reading: ‫ ענה‬is often accompanied by ‫ אמר‬in Samuel if the speaker is a human being, whereas alone ‫ ענה‬is reserved only for God.44 The longer reading explains how the MT reading came about: a parablepsis due to homoioarkton of the waw consecutive ‫וענה האיש ואמר‬, which led to the omission of both ‫ וענה‬and ‫האיש‬. In addition to the quantitative agreement (the L-text: καὶ ἀπεκρίνατο and 4QSamc: ‫ )וענה‬there are also qualitative reasons to prefer the L-text as the OG reading. The OG translator probably recognized and translated the iterative consecutive perfect (‫ )ואמר‬with the imperfect (ἔλεγεν) as usual. The reading in B (+), again, is probably the kaige reading agreeing with the MT, whereas Maj. perhaps represents an early revision of the text in which καὶ ἀπεκρίνατο had been dropped as a redundant element. Thus, I agree with Ulrich about the real agreement, and the L-text is the OG. The longer reading in the Hebrew text, however, is not an expansion in the Hebrew text attested in 4QSamc but an earlier text, which is seen in the underlying Hebrew of the OG, preserved in the L-text.

2 Sam 22:46. 49 4QSama L-text 22:46

22:49

‫ לא יחגרו‬ἐλυτρώθησαν ‫ ממסרותם‬ἐκ δεσμῶν

MT Maj.

‫ וְ ְיַחּגְ רּו‬καὶ

MTPs (18) LXXPs (17)

‫ וְ יַ ְח ְרגּו‬46

σφαλοῦσιν αὐτῶν ‫ֹרותם‬ ָ ְ‫ ִמ ִּמ ְסּג‬ἐκ τῶν ‫יהם‬ ֶ ‫ֹרות‬ ֵ ְ‫ִמ ִּמ ְסּג‬ συγκλεισμῶν αὐτῶν. ‫ ֯תצרני‬διετήρησάς με ‫ ַּת ִּצ ֵילנִ י‬ῥύσῃ με ‫ּצילנִ י‬ ֵ ‫ ַּת‬49

καὶ ἐχώλαναν ἀπὸ τῶν τρίβων αὐτῶν 49 ῥύσῃ με 46

43  DJD XVII, 263: For Ulrich, the only secondary element in the MT is the imperfect consecutive ‫ ויאמר‬for an original perfect consecutive ‫ואמר‬. 44  See, for example, 1 Sam 7:19, 8:18, 9:17.

42

Paavo Huotari

Second Samuel 22 is David’s song of praise, which is also attested in Psalm 17 LXX (18 MT). DJD (Cross, Parry and Saley) suggests two agreements here. In both cases we are probably dealing with scribal errors in the Hebrew text.45 The error may go in either direction. In both cases the L-text does not correspond to Ps 17 LXX (18 MT) and agrees with 4QSama. In 22:46 ‫“ ממסרותם‬bond/band” in 4QSama is against ‫“ ִמ ְסּגֶ ֶרת‬fortress; stronghold” in the MT and MTPs. The L-text reads δεσμός “bond, chain” which fully agrees with 4QSama, whereas Maj. reads συγκλεισμός “prison, a closed place” mostly agreeing with the MT. ‫( ִמ ְסּגֶ ֶרת‬MT) is also translated by συγκλεισμός in Mic 7:17. The agreement between 4QSama and the L-text is thus real. The MT reads ‫רֹותם‬ ָ ְ‫“ וְ יַ ְחּגְ רּו ִמ ִּמ ְסּג‬and they will gird/tremble out of their strongholds” making no sense. The Hebrew lexeme ‫ ִמ ְסּגֶ ֶרת‬in the MT, nevertheless, corresponds to the MTPs that indicates that the MT is probably earlier than 4QSama in this case. If so, 4QSama preserves the corrupt reading, which is probably due to a scribal error (minus ‫)ג‬.46 However, the underlying Hebrew Vorlage of the OG can still be the corrupt reading in 4QSama. It seems unlikely that the L-text preserves a Hexaplaric reading, or its agreement with 4QSama to be coincidental. Because the Lucianic reviser probably did not have Hebrew competence, the agreement between the L-text and 4QSama is probably prior to him, in other words, due to the translator. Thus, the L-text preserves the probable OG reading, of which the underlying Hebrew is attested in 4QSama, and Maj. preserves the kaige-reading according to the MT. In 22:49 ‫“ ֯תצרני‬you watch me” in 4QSama is against ‫“ ַּת ִּצ ֵילנִ י‬you deliver me” in the MT. The L-text reads διετήρησάς με “you watch me closely” agreeing with 4QSama, whereas Maj. reads ῥύσῃ με “you rescue me” mostly agreeing with the MT. The difference between the Hebrew texts is only reš and yod/lamed. Contrary to the previous case, both readings in the Hebrew texts are understandable. Because Maj. and the MT agree with the LXXPs and the MTPs, the Hebrew ‫ַּת ִּצ ֵילנִ י‬ is probably earlier than the competing reading in 4QSama, which is probably a scribal error from ‫ יל‬to ‫ר‬. Nevertheless, the underlying Hebrew Vorlage of the OG can yet be 4QSama as was in the case mentioned above. As an equivalent of ‫נצל‬, ῥύομαι is attested nine times in Samuel-Kings (5 times in Samuel), and always in kaige sections. On the other hand, διατηρέω is the equivalent of ‫ נצר‬three times in the LXX (Deut 33:9, Ps 12:8, Prov 22:12). Therefore, ῥύομαι is probably the kaige reading, and διετήρησας the OG.47 45  DJD XVII, 186–87. Cf. McCarter, II Samuel, 462–63; Auld, I & II Samuel, 589. 46  Contrary to Jason K. Driesbach, 4QSamuela and the Text of Samuel, VTS 171 (Leiden: Brill, 2016), 90: “MT (=GB) is odd” but still “this reading was corrupted via metathesis of gimel and reš in MT (=GB). 4Q’s ‫ מסרותם‬resulted from an error of lectio facilior for ‫מסגרותם‬. Finally, the reading of 4Q was adapted to the context through the addition of a negation.” See McCarter, II Samuel, 462, 472. 47  Tim McLay, “Kaige and Septuagint Research,” Textus 19 (1998): 131–34, see n. 81. I mostly agree (see points 1 and 2) with Ulrich, Qumran Text, 112: “Thus we can say: (1) that G reflects M,

Discovering Old Greek Readings in the Lucianic Text

43

To sum up, both readings (22:46 and 49) in the L-text (δεσμός, διετήρησάς με) are most likely OG readings that agree with 4QSama, whereas Maj. preserves kaige corrections (συγκλεισμός, ῥύσῃ με) according to the earlier Hebrew reading. In general, it is likely that kaige is revised according to the secondary Hebrew text attested in the MT against 4QSama. However, in these two cases the issue is scribal error in the Hebrew text, not intentional editing. Consequently, it is probable as argued above that here the corruption is already found in the Hebrew text behind the Greek translation, and kaige is revised according to the earlier Hebrew text attested in the MT (i.e. proto-MT) against 4QSama.

6. Conclusion The Qumran manuscripts 4QSama and 4QSamc in 2 Samuel do not provide any evidence that would support the proto-Lucianic recension (Cross). This would have required several agreements between 4QSama, c and the L-text in secondary readings. This study did not find even a single case. There are five real agreements (+ six probable agreements) between 4QSama, c and the L-text that are simultaneously against the MT and Maj. In these five cases, nevertheless, the agreements evidence the OG in the L-text. This strengthens the view of the previous study by Kauhanen regarding the relationship of the OG in the L-text (1 Samuel), and 4QSama, b. This demonstrates well that the L-text offers an insight into the earlier Greek text. At the same time competing readings in Maj. probably preserve kaige-revisions that agree with the Hebrew tradition that later on became the MT. In 22:46, 49, exceptionally, the readings in the MT (=proto-MT) show traces of earlier text than found in competing and secondary readings of 4QSama. Three coincidental and seven pseudo-agreements, however, strongly remind us that any default assumption may lead us astray when evaluating the agreements. The similar appearance of two readings in two different witnesses is not yet evidence of a common origin. This is particularly evident in the L-text, which through Hexaplaric approximations or the Lucianic revision may attest a Hebrew text similar to Qumran scrolls. The largest category is uncertain agreement with eleven cases, in which no conclusion in either direction can be made. Changes in minor details may happen in any witness at any time and produce polygenetic agreements.

while L reflects 4Q; (2) that G = KR; (3) “L = pL, because after the mid first century AD there was no possible basis for διετήρησάς; (4) that it is impossible to determine the OG.” Ulrich’s last point (4) probably results from his third point, “L = pL.” If the L-text represents the proto-Lucianic recension, it cannot be the OG. In this instance, however, the most unambiguous conclusion is L = OG.

44

Paavo Huotari

7. Appendix 4QSama ≈ L In this appendix I have included all 25 agreements between 4QSama and the L-text against the MT and Maj. 4QSama L-text48

3:28

5:11

11:4

12:15

‫ ודם‬αἷμα

‫ וחרשי‬καὶ τέκτονας ‫קיר‬ ֯ τοίχου

καὶ

]‫ותב[וא‬ ׄ ἀπῆλθεν ‫ אלוהים‬ὁ θς

MT Maj.

‫ ִמ ְּד ֵמי‬ἀπὸ τῶν

αἱμάτων

‫ וְ ָח ָר ֵׁשי‬καὶ τέκτονας ‫ ֶא ֶבן‬λίθων ‫ִקיר‬

καὶ

‫ וַ ָּת ָׁשב‬ἀπέστρεψεν ‫יְהוה‬ ֗ ָ κύριος

Selected variants49

Classification

ἀπὸ τῶν αἱμάτων] ἀπὸ τοῦ αἵματος 245 | αἱμάτων] bis scr V*

Pseudo: The Lucianic revision (ἀπὸ τοῦ νῦν καὶ ἕως τοῦ αἰῶνος αἷμα Αβεννηρ) Real

λίθων] pr τοίχου 19′; τοίχου L−19′; + τείχου 376; + τοίχου A 247 CI−243txt; sub ※ 243mg; La115; cf. 1 Par 14:1 ἀπέστρεψεν] υπ. 328; et intravit La115 κύριος] ὁ θς 158

Probable

Pseudo: The Lucianic reviser may have harmonized the word towards τὸν θεόν in 12:16

48  The text with shading marks the agreement between Qumran and the L-text 49  The Lucianic witnesses are also provided in this section if there is variation among them. A lemma attests the reading in Maj.

Discovering Old Greek Readings in the Lucianic Text

4QSama L-text48

12:16

‫ויב[ו]א‬ ֯ εἰσελθὼν ‫ וישכב‬ἐκάθευδεν ‫ בשק‬ἐν σάκκῳ

12:17

‫ ‏‬‎ καὶ ֯‫ [זקנ]י‬προσῆλθον οἱ ‫ויקו֯ ֯מ[ו]‏‬‎ πρεσβύτεροι ‫ ביתו‬τοῦ οἴκου αὐτοῦ

‫ אליו‬πρὸς αὐτὸν 13:3

13:24

‫ [י]הונתן‬Ιωναθαν

‫ אל‬πρὸς ‫ עבדך‬τὸν δοῦλον αὐτοῦ

13:32

‫ [כי על‬ὅτι ἐν ὀργῇ ]‫ אף‬ἦν αὐτῷ ]‫ [אבשלום‬Αβεσσαλωμ ‫היה‬ (om) (om)

MT Maj.

Selected variants49

‫ּובא‬ ָ καὶ εἰσῆλθεν ‫ וְ ָלן‬καὶ ηὐλίσθη ‫וְ ָׁש ַכב‬

ἐν σάκκῳ V M 376 L CI a−527 f 64′ 29 55 71 158 244 ἐν σάκκῳ 245 318 342 372 460] καὶ ἐκοιμήθη  A; > B b 68′ 707; pr καὶ ἐκοιμήθη rel καὶ ἀνέστησαν ἐπ᾽ αὐτόν] ‫ וַ ֻּיָקמּו‬ἐπ᾽ αὐτὸν οἱ > 379txt; post ‫ זִ ְקנֵי‬πρεσβύτεροι αὐτοῦ tr A O L ‫ ֵביֹתו‬τοῦ οἴκου CII−242 527 αὐτοῦ d−68′370 ‫ָע ָליו‬ s−64′ 554 | ἐπ᾽] πρὸς 799 f

‫ֹיונָדב‬ ָ Ιωναδαβ

‫ ִעם־‬μετὰ ‫ ַע ְב ֶּדָך‬τοῦ δούλου σου

‫י־על־‬ ַ ‫ִּכ‬ ‫ִּפי‬ ‫ַא ְב ָׁשֹלום‬ ‫ָה ָיְתה‬ ‫ׂשּומה‬ ָ

ὅτι ἐπὶ στόματος Αβεσσαλωμ ἦν κείμενος

Ιωναδαβ] ιωναδαμ B* 530 509 488 342; ιωναθαν L−127 μετὰ − σου] πρὸς τὸν δοῦλον αὐτοῦ 554mg; μ. τῶν δούλων αὐτοῦ εἰς τὸ δοῦλο | σου 245; at seruum suum La115 ἐπὶ – κείμενος] ἐν ὀργῇ ἦν α. τῷ αμνων 372; ἐν ὀργῇ ἦν αὐτῷ αβεσσαλωμ (‑εσα. L−93 127; > M) M L; in ira enim est at abessalon | La115

45 Classification

Uncertain: The reading in the L-text may also render the reading in the MT.

Pseudo: The Lucianic reviser may have changed a preposition according to the prefix of προσῆλθον Probable

Pseudo: The Lucianic reviser may have improved the syntax

Coincidental: Hex. influence in the L-text ὅτι – κείμενος]  α′ σ′ ὅτι ἐν ὀργῇ ἦν αὐτῷ Ἀβεσσαλώμ 554

46

Paavo Huotari

4QSama L-text48

15:2

15:2

15:2

18:9

]‫ וה[יה‬καὶ ἦν

‫ וקרא‬καὶ ἐκάλει

MT Maj.

Selected variants49

Classification

‫ וַ ִיְהי‬καὶ ἐγένετο

ἐγένετο] ἐγίνετο a−527 554

Uncertain: The reading in the L-text may also render the reading in the MT. Uncertain: The reading in the L-text may also render the reading in the MT. Uncertain: The reading in the L-text may also render the reading in the MT

‫ּיִק ָרא‬ ְ ַ‫ ו‬καὶ ἐβόησεν

καὶ

]‫ [וענה‬ἀπεκρίνατο ὁ ]‫ [האיש‬ἀνήρ ‫ ואמר‬καὶ ἔλεγεν

καὶ

‫והוא‬ ֯ αὐτὸς

καὶ εἶπεν

‫אמר‬ ֶ ֹ ‫ וַ ּי‬ὁ ἀνήρ

καὶ

‫ וְ ַא ְב ָׁשֹלום‬Αβεσσαλωμ

20:10

‫ על‬ἐπὶ ‫החמ ֯ש‬ ֯ τὴν λαγόνα

‫ ֶאל־‬εἰς ‫ ַהח ֶֹמׁש‬τὴν ψόαν

20:10

‫ רדפו‬κατεδίωκον

‫ ָר ַדף‬ἐδίωξαν

20:13

‫ ֯כ[ו]ל‬πᾶς ‫ העם‬ὁ λαὸς

‫ ָּכל־‬πᾶς ‫ ִאיׁש‬ἀνὴρ Ισραηλ

ἐβόησεν] ἐβόα O 318; ἐκάλει 554mg; ἔλεγεν 372

εἶπεν] ἔλεγεν O 318; > 158; + αὐτῷ  CII−46′ 313 | ὁ ἀνήρ] πρὸς αὐτόν 707; > B V M CI f 29 158 318 372; om ὁ 530 125 Αβεσσαλωμ] αβεσλω 158*; αὐτός 554c εἰς] ἐπὶ A O 527 460;  τοῦ ἐπὶ τὴν γῆν 242 ἐδίωξαν] κατεδίωκον L; -‍ξεν B M 121 d−125 246 71 158 372 554 Ra

Probable

Coincidental: Hex. Influence in the L-text Uncertain: The reading in the L-text may also render the reading in the MT Pseudo: The Lucianic reviser may have harmonized the text according to πᾶς ὁ λαός in 20:12, 15.

Discovering Old Greek Readings in the Lucianic Text

4QSama L-text48

21:16

]‫ [והוא‬καὶ οὗτος ]‫ [חגור‬περιεζωσμένος ‫חורה‬ ֯ παραζώνην

MT Maj.

Selected variants49

47 Classification

‫ וְ הּוא‬καὶ αὐτὸς περιεζωσμένος] + Coincidental: ‫ ָחגּור‬περιεζωσμένος ξίφος Hex. influence ‫ ֲח ָד ָׁשה‬κορύνην 707 | κορύνην] in the L-text κορυφὴν 707; κορύνην] α′ παραζώνην Mmg; κενήν 243; σ′

+ παραζώνην 342 μάχαιραν 243379(s nom); θ′ παραζώνην CI−98 (379s noms nom) 22:41 τοὺς μισοῦντάς τοὺς − αὐτούς] Uncertain: ]‫ [משנאי‬μισούντων με ‫ ְמ ַׂשנְ ַאי‬με, καὶ μισούντων με The absence )om( (om) ἐθανάτωσας κατεπάτησα  of a ]‫ ׄאצמית[ם‬κατεπάτησα ‫ וָ ַא ְצ ִמ ֵיתם‬αὐτούς (‑σαν 93) L; om conjunction καί V 247 488′ 460 | με] + ἐξωλόθρευσας 530 318 22:43 ]‫ [כעפר על‬ὡς χνοῦν ‫ ַּכ ֲע ַפר־‬ὡς χοῦν γῆς] ἐπὶ πρόσωπον Pseudo: ‫ ‏פני‬ἐπὶ πρόσωπον ἀνέμου The Lucianic ‫ ארח‬ἀνέμου ‫ ָא ֶרץ‬γῆς (‑μων 19′) reviser ‫[כ]טיט‬ ֯ ὡς πηλὸν ‫ ְּכ ִטיט־‬ὡς πηλὸν  L−93 127;  may have ‫חוצ[ו]ת‬ ֯ τῶν ἐξόδων ‫ חּוצֹות‬ἐξόδων ἐπὶ (απο 93)  harmonized to ‫ ארקעם‬λεανῶ ‫ ֲא ִד ֵּקם‬ἐλέπτυνα πρόσωπον ἀνέμου Ps 18:43 αὐτούς ‫ ֶא ְר ָק ֵעם‬αὐτούς 93-127 | ἐξόδων] pr  τῶν L−82 108;  pr τὸν 108; ἐξ ᾅδου 247 | αὐτούς] + στερεωματησω αὐτούς 247 22:46 ‫ לא יחגרו‬ἐλυτρώθησαν ‫ וְ יַ ְחּגְ רּו‬καὶ σφαλοῦσιν συγκλεισμῶν]  Real ‫ ממסרותם‬ἐκ ‫ֹרותם‬ ָ ְ‫ ִמ ִּמ ְסּג‬ἐκ τῶν ‑ματων 247 460; δεσμῶν αὐτῶν συγκλεισμῶν δεσμῶν L−127; αὐτῶν. δυσμῶν 127 22:49 ‫ ֯תצרני‬διετήρησάς με ‫ ַּת ִּצ ֵילנִי‬ῥύσῃ με ῥύσῃ με] ρυσητε Real 130(|) | ῥύσῃ] ῥύσεις 246; ῥῦσαί V M 328 509 488 158 245 460; διετήρησε 19′; διετήρησας L−19′

48

Paavo Huotari

4QSama L-text48

22:51

‫ ישועת‬τὴν σωτηρίαν ‫ מלכו‬τοῦ βασιλέως αὐτοῦ

23:1

23:3

‫הקי֯ ׄם‬ ‫אל‬ ‫משיח‬ ‫[אלו]הי‬ ]‫֯יע[קב‬

]‫ [משל‬ἄρξον ‫ׄבאדם‬ ‫[צדי]ק‬ ׄ ‫משל‬ ‫[ביראת‬ ]‫אלוהים‬

24:18

ἀνέστησεν ὁ θεὸς χριστὸν θεὸς Ιακωβ

ἐν ἀνθρώποις δικαίως ἄρχε φόβῳ θεοῦ

‫ ויאמר‬καὶ εἶπεν

MT Maj.

‫ יְ ׁשּוֹעות‬τάς σωτηρίας ‫ ַמ ְלֹּכו‬βασιλέως

‫ֻה ַקם‬ ‫ָעל‬ ‫ֹלהי‬ ֵ ‫ְמ ִׁש ַיח ֱא‬ ‫יַעקֹב‬ ֲ

‫ֹמוׁשל‬ ֵ ‫ָּב ָא ָדם‬ ‫ַצ ִּדיק‬ ‫ֹמוׁשל‬ ֵ ‫ְיִר ַאת‬ ‫ֹלהים‬ ִ ‫ֱא‬

‫אמר ֹלו‬ ֶ ֹ ‫וַ ּי‬

Selected variants49 τάς σωτηρίας] τὴν σωτηρίαν L; om τάς B A 247 (98vid) a b 56 64′ 29 245 707

Classification

Uncertain: The reading in αὐτοῦ the L-text may also render the reading in the MT ἀνέστησεν ὁ κύριος] ὁ θς Pseudo: ὁ κύριος L; om ὁ B A V The Lucianic ἐπὶ χριστὸν 247 CI 328 b 68′ f reviser θεοῦ Ιακωβ 64′-489 29 55 71 may have 158 244 245 318 harmonized 460 707 the text according to following θεὸς Παραβολὴν πῶς κραταιώσητε Uncertain: εἰπόν φόβον] δικαίως The reading in ᾿Εν ἀνθρώπῳ ἄρχε (ἀρχαὶ 19′) the L-text may πῶς φόβῳ L; om πῶς also render κραταιώσητε 247 | κραταιώσητε the reading φόβον B V 121 68′ 64′ in the MT θεοῦ 29 318 460c] κραταιωσηται M a−527 509 f 55 71 158 342 460* 707; ‑σατε A 247; κραταιωσεται rel καὶ εἶπεν αὐτῷ αὐτῷ ] πρὸς αὐτόν Probable 460;  > 245 707; + γαδ CI 29 244

49

Discovering Old Greek Readings in the Lucianic Text

8. Appendix 4QSamc ≈ L In this appendix I have included all seven agreements between 4QSamc and the L-text against the MT and Maj. 4QSamc L-text

καὶ σβεσθήσεται ὁ σπινθὴρ ὁ

14:7

(Top ὑπολελειμμένος margin) μοι ‫[ש]ארתי‬ ֯ ‫֯ה‬ 14:11 ‫ משער]ות‬ἀπὸ τῶν τριχῶν

14:20

14:21

‫[לדע]ת‬ ׄ )om( ‫֯אשר‬ ‫בארצ‬ ‫ולך והשב‬

τοῦ γνῶναι (om) τὰ ἐπὶ τῆς γῆς πορεύου καὶ ἐπίστρεφε

14:27

‫ והיאה‬καὶ αὕτη ]‫היתה[ אשה‬ ׄ ἦν γυνὴ

14:32

‫ ‏נא‬ὄψομαι ‎^‫[א]^ה‬ ׄ ‫אר‬ ֯ δὴ

15:2

‫ וענה‬καὶ ἀπεκρίνατο ]‫ [האיש‬ὁ ἀνήρ ]‫ [ואמר‬καὶ ἔλεγεν

MT Maj.

καὶ

‫ וְ ִכּבּו‬σβέσουσιν ‫ ֶאת־ּגַ ַח ְל ִּתי‬τὸν ἄνθρακά

Some noteworthy variants

Classification

om μου 527

Probable

μου

‫ ֲא ֶׁשר‬τὸν κατα‫נִׁש ָ֔א ָרה‬ ְ λειφθέντα ‫ ִמ ַּׂש ֲע ַרת‬ἀπὸ τῆς τριχὸς

‫ָל ַד ַעת‬ ‫ת־ּכל־‬ ָ ‫ֶא‬ ‫ֲא ֶׁשר‬ ‫ָּב ָא ֶרץ‬ ‫ה ֵׁשב‬ ‫ְך‬ ָ ‫וְ ֵל‬

‫ִהיא‬ ‫א ָּׁשה‬ ‫ה‬ ִ ‫ָה ָיְת‬ ‫ֶא ְר ֶאה‬ ‫אמר‬ ֶ ֹ ‫וַ ּי‬

Uncertain: The innerlanguage variation Probable

τοῦ γνῶναι πάντα τὰ ἐν τῇ γῇ. πορεύου πορεύου] καὶ Uncertain: ἐπίστρεψον πορεύθητι καὶ The plus A O; > B 509 55 of a conjunction αὕτη Uncertain: ἦν γυνὴ The plus of a conjunction ἰδοὺ καὶ νῦν ἰδού] Real ὅτι O καὶ εἶπεν ]  Real ἔλεγεν O 318;  εἶπεν > 158; + αὐτῷ ὁ ἀνήρ CII−46′ 313 |  ὁ ἀνήρ] πρὸς αὐτόν 707; > B V M CI f 29 158 318 372;  om ὁ 530 125 

50

Paavo Huotari

9. Bibliography Aejmelaeus, Anneli, “What Can We Know about the Hebrew Vorlage of the Septuagint?” in On the Trail of the Septuagint Translators, ed. A. Aejmelaeus (Leuven: Peeters, 2007), 71–106. —, “The Septuagint of 1 Samuel.” in On the Trail of the Septuagint Translators, ed. A. Aejmelaeus (Leuven: Peeters, 2007), 123–141. —, “Kaige readings in a Non-Kaige Section in 1 Samuel,” in The Legacy of Barthélemy. Edited by Anneli Aejmelaeus and Tuukka Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017), 169–184. Albrecht, Felix, “Von der hebraica veritas zur vera graecitas: Origenes – Hesych – Lukian,” BN 184 (2020): 105–141. Anderson, Arnold A, 2 Samuel. WBC 11 (Dallas: Word Books, 1989). Auld, A. Graeme, I & II Samuel. A Commentary (Louisville: Westminster John Knox, 2012). Barthélemy, Dominique, Les Devanciers d’Aquila: Première publication intégrale du texte des fragments du Dodécaprophéton. VTSup 10 (Leiden: Brill, 1963). Brock, Sebastian, The Recensions of the Septuaginta version of 1 Samuel (Torino: Silvio Zamorani editore, 1996). Busto Saiz, J. R, “On the Lucianic manuscripts in 1–2 Kings,” in Sixth Congress of the International Organization for Septuagint and Cognate Studies: Jerusalem, 1986, ed. C. E. Cox. SBLSCS 23 (Atlanta: Scholars Press, 1987), 305–311. Cross, Frank Moore, “The History of the Biblical text in The Light of Discoveries in the Judean Desert.” HTR 57 (1964): 281–299. Cross, Frank Moore and Richard J. Saley, “A Statistical Analysis of the Textual Character of 4QSamuela (4Q51),” DSD 13,1 (2006): 46–54. Cross, Frank Moore et al., Qumran Cave 4.XII:1–2 Samuel. DJD XVII (Oxford: Clarendon, 2015). De Troyer, Kristin, Rewriting the Sacred Text. What the Old Greek Texts Tell Us about the Literary Growth of the Bible (Leiden: Brill, 2003). Fernández Marcos, Natalio. The Septuagint in Context (Leiden: Brill, 2000). Kauhanen, Tuukka, The Proto-Lucianic Problem in 1 Samuel. DSI 3 (Göttingen: Vandenhoeck & Ruprecht, 2012). —, “Lucifer of Cagliari and the Kaige Revision,” in The Legacy of Barthélemy, ed. Anneli Aejmelaeus and Tuukka Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017), 146–168. —, “The Best Greek Witnesses for 2 Samuel,” in Die Septuaginta – Themen, Manuskripte, Wirkungen, ed. Eberhard Bons, Michaela Geiger, Frank Ueberschaer, Marcus Sigismund and Martin Meiser, WUNT 444 (Tübingen: Mohr Siebeck, 2020), 561–573. McCarter, P. Kyle, II Samuel. A New Translation with Introduction, Notes and Commentary. AB 9 (New York: Doubleday, 1984). McLay, Tim. “Kaige and Septuagint Research,” Textus 19 (1998): 127–139. Metzger, Bruce M, Chapters in the History of New Testament Textual Criticism (Leiden: Brill, 1963).

Discovering Old Greek Readings in the Lucianic Text

51

Seppänen, Christian, The Hebrew Text of Samuel: Differences in 1 Sam 1 – 2 Sam 9 between the Masoretic Text, the Septuagint, and the Qumran Scrolls (PhD diss., The University of Helsinki, 2018). Tov, Emanuel, “Lucian and Proto-Lucian: Toward a New Solution of the Problem,” RB 79 (1972): 101–13. —, The Greek Minor Prophets Scroll from Naḥ al Ḥ ever (8HevXIIgr). DJD VIII (Oxford: Clarendon, 1990). —, The Text-Critical Use of the Septuagint in Biblical Research (Winona Lake: Eisenbrauns, 2015). Ulrich, Eugene C, The Qumran Text of Samuel and Josephus. HSM 19 (Atlanta: Scholars Press, 1978). —, “4QSamc: A Fragmentary Manuscript of 2 Samuel 14–15 from the Scribe of the Serek Hayyaḥ ad (1QS).” BASOR 235 (1979): 1–25. Voitila, Anssi, “The Use of Tenses in the L- and B-texts in the Kaige-Section of 2 Reigns,” in Die Septuaginta, Entstehung, Sprache, Geschichte, ed. Siegfried Kreuzer, Martin Meiser and Marcus Sigismund. Tübingen: Mohr Siebeck, 2012), 213–237. Wellhausen, Julius, Der Text der Bücher Samuelis (Göttingen: Vandenhoeck & Ruprecht, 1871). Wirth, Raimund, “Dealing with Tenses in the Kaige section of Samuel,” in The Legacy of Barthélemy, ed. Anneli Aejmelaeus and Tuukka Kauhanen. DSI 9 (Göttingen: Vandenhoeck & Ruprecht, 2017) 185–197.

Pablo A. Torijano/Julio Trebolle The Edition of III–IV Kingdoms The Critical Reconstruction of the Old Greek Text and the Construction of the Critical Apparatus The history of the text of 3–4 Kingdoms is clearly different from the textual history of the Pentateuch and other books edited in the Göttingen Series.1 The textual history of these books is reduced to the transmission of the text of the original version (OG) over several centuries until its revision in the Hexaplaric and Lucianic recensions.2 On the contrary, in 3–4 Kingdoms the Hexaplaric recension and the previous recensions of Aquila, Symmachus and Theodotion were preceded by a recension done around the change of era that was the first attempt to bring the Old Greek text (OG) in line with an early form of the Hebrew Masoretic tradition. This kaige-Theodotionic recension, identified by the study of the MS of the Minor Prophets from Naḥ al Ḥ ever, affected the text of two sections of 3–4 Kingdoms established by Thackeray: 3 Kgdms 1:1–2:11 (βγ) and 3 Kgdms 22:1–4 Kgdms 25:30 (γδ).3 In these kaige sections the majority text represented by the B group (manuscripts B 121 509) transmits the text of the kaige-Th. recension. The fact that 4QSama contains Hebrew readings agreeing with variants of the “Lucianic” manuscripts 19-82-93-108-127 came to prove that the Antiochene text represents an old textual form close to the OG,4 so that in those kaige sections this Antiochene pre-Lucianic text is the only one to preserve a text “substantially identical” to the Old Greek.5 The fragments of Kings in Qumran MSS are scarce, so it is particularly meaningful that the OG ἄρσιν αὐτῶν attested 1 When referring to the Greek text, we will use the denomination 1–4 Kingdoms (3–4 Kdgms); if the discussion deals with the Hebrew text, the denomination will be either 1–2 Samuel / 1–2 Kings. 2  Christian Schäfer, Benutzerhandbuch zur Göttinger Septuaginta. Band 1: Die Edition des Pentateuch von John William Wevers (Göttingen: Vandenhoeck & Ruprecht, 2012), 93–112. 3  H. St. J. Thackeray, The Septuagint in Jewish Worship (London: Oxford University Press, 1921), 114. 4  É. Puech proposes new readings of the Qumran manuscripts of Samuel and Kings with the result that the number of agreements of the Qumran manuscript with the Antiochene text or the OG is greater than previously assumed, cf. É. Puech, “Les manuscrits des Livres de Samuel dans les grottes 1 et 4 de Qumrân 1Q17, 4Q51–52–53, l’Apocryphe de Samuel – 4Q160 et l’Apocryphe de Samuel–Rois – 6Q9,” RB 126 (2019): 5–51, esp. 31–18; É. Puech “Les Livres des Rois à Qumrân: 4Q54, 4Q54a(?), 5Q2 et 6Q4,” (forthcoming, author’s information). 5 Barthélemy, Devanciers; D. Barthélemy, “Les problèmes textuels de 2 Sam 11,2 – 1 Rois 2,11 reconsidérés à la lumière de certaines critiques des Devanciers d’Aquila,” Bull. IOSCS (1972): 16–89.

54

Pablo A. Torijano/Julio Trebolle

by the Antiochene text (19-93-108-127) agrees with the reading of 6Q4 frg. 11 2 Kgs 7:8 ‫משאם‬, masā’ām and stands against the kaige reading ἐκεῖθεν of LXXB, Syr, Vulg and Targ (MT ‫)משם‬. Rahlfs acknowledges that the L reading is “Vorlucianisches Gut.”6 The Antiochene pre-Lucianic text is often supported by readings of Chronicles and Paralipomena, Josephus, the Old Latin, the pre-Hexaplaric stratum of the Armenian version and the Coptic and Georgian versions. The Septuagint edition of Rahlfs represents the OG in the non-kaige sections (α, ββ and γγ), but follows the kaige-Th text in sections βγ and γδ. In practice, Rahlfs considers that the Old Greek text of the kaige sections has been lost, as Tov affirms as well: “that text [the OG] has been lost in two sections in 1–4 Kingdoms that have been replaced with the kaige-Th. revision.”7 However this first revision was not as systematic as to substitute the OG entirely. Nor was it as superficial as it may seem given the small number of identified kaige characteristics. The OG text can be considered sometimes lost or substituted by the kaige text, but in many other occasions the Antiochene pre-Lucianic text preserves a text “substantially identical” to the Old Greek. Consequently, the problem of the edition of 3–4 Kingdoms lies basically in identifying and differentiating kaige and pre-Lucianic textual levels in order to restore the pre-recensional Old Greek text and ultimately its Hebrew Vorlage. As stated by Barthélemy, “les livres des Règnes sont peut-être des tous les livres de la Bible ceux pour lesquels on peut attendre d’une récupération de la Septante ancienne les lumières les plus précieuses pour la restitution du texte hébraïque original.”8 According to Wevers the original of LXX Pentateuch “was in the main much like the consonantal text of MT; in other words, the extant textual tradition must be taken seriously. The age of rampant retroversion and wild emendations is over.”9 Conversely the Septuagint of Kings translates a Hebrew text that differs significantly from the (proto-) Masoretic. Therefore, the challenge of the edition of 3–4 Kingdoms lies in taking seriously the fact that LXX represents a different edition of these books, as it is also the case in sections of the historical books as shown by the Qumran manuscripts 4QJosa, 4QJudga and 4QSama.10 Thus, it is possible to state that the age of attributing the variants of 3–4 Kingdoms to interpretations or tendencies of the translator following the Jewish exegesis of the period is, to a large degree, over.

6 Rahlfs, Lucians Rezension, 288. 7 Tov, Textual Criticism of the Hebrew Bible, 247. 8 Barthélemy, Devanciers, 140. 9  J.W. Wevers, Notes on the Greek Text of Genesis (Atlanta GA: Scholars Press, 1993), p. xiii. 10  E. Ulrich, The Dead Sea Scrolls and the Developmental Composition of the Bible (Leiden: Brill, 2015), 47–108.

The Edition of III–IV Kingdoms

55

Wevers devoted his first publications to the study of the translator’s exegetical or theological principles in 1–4 Kingdoms.11 Thus, for example, according to Wevers the use of ἐν χειρὶ (κυρίου) in 2 Kgdms 24:14 ἐμπεσοῦμαι δὴ ἐν χειρὶ κυρίου… εἰς δὲ χεῖρας ἀνθρώπου οὐ μὴ ἐμπέσω (“I shall fall now in the hand of the Lord…, but into human hands I will not fall”) “implies a tendency to remove God as far as possible from Mankind.” As Raija Sollamo notes Wevers operated “without knowledge of the translator’s translation technique.” Sollamo shows that the version ἐν χειρί is only “an attempt at as literal a rendering as possible,” so that “there was no theology here at all, only philology.”12 The analysis of the translation techniques must be preceded by criticism of the external evidence. The evidence provided by the Greek manuscripts and the secondary versions helps to identify the textual filiation of the two readings ἐν χειρί/εἰς χεῖρας in 2 Kgdms 24:14 (in the kaige section γδ). The reading ἐν χειρί in 2 Kgdms 24:14 is witnessed only by the MSS B 509 (Aeth), as well as by the MSS that represent the Hexaplaric text (A 247). The rest of the manuscript tradition attests the reading εἰς (τὰς 19-93-108-127 120 Josuid) χεῖρας: rel. Arm OL Jos(uid) Or-lat Chr Thdt. The distribution of the translation εἰς χεῖρας or ἐν χειρί in the no kaige and kaige sections of 1–4 Kgdms shows that the version εἰς χεῖρας is characteristic of the original LXX version, whereas the literal rendering ἐν χειρί (MT ‫)ביַ ד‬ ְ is characteristic of the kaige recension in the βγ and γδ sections. Thus, in the non-kaige section the OG reading (παραδίδωμι/ἀποδίδωμι/δίδωμι) εἰς χεῖρας is found in 1 Kgdms 12:9; 14:10; 14:37; 17:47; 21:3; 23:4; 23:12 (Hexaplaric); 23:14; 24:5; 24:11 (χεῖρα B Aeth] χεῖρας A 509 56 554 707; τὰς χεῖρας rel); 2 Kgdms 5:19 (2x); 3 Kgdms 15:18; 18:9 (χεῖρα B Aeth Syr] χεῖρας AV om. Arm); 21:13; 21:28 (χεῖρα] χεῖρας L Arm). In the kaige sections the OG reading (παραδίδωμι) εἰς χεῖρας appears in 3 Kgdms 22:6.12 and 4 Kgdms 3:13; 19:10; 21:14, witnessed by the whole manuscript tradition. The kaige reading ἐν χειρί is found in 2 Kgdms 16:8; 21:9; 4 Kgdms 3:10.18; 13:3 (2x); 18:30, but the Antiochene text, followed by the Armenian version in its pre-Hexaplaric stratum, preserves the old version εἰς 11  J.W. Wevers, “A Study in the Hebrew Variants in the Books of Kings,” ZAW 61 (1945–48): 43–76; J.W. Wevers, “Exegetical Principles underlying the Septuagint Text of I Kings ii 12 – xxi 43,” OTS 8 (Leiden, 1950): 300–322; J.W. Wevers, “Principles of Interpretation Guiding the Fourth Translator of the Book of the Kingdoms (3 K. 22:1 – 4 K. 25:30),” CBQ 14 (1952): 40–56; J.W. Wevers, “A Study in the Exegetical Principles Underlying the Greek Text of 2 Sam 11:2– 1 Kings 2:11,” CBQ 15 (1953): 30–45. 12  R. Sollamo, “The Study of Translation Technique,” in Die Sprache der Septuaginta. The Language of the Septuagint, ed. Eberhard Bons and Jan Joosten (Gütersloh: Gütersloher Verlaghaus, 2016), 161–171, esp. 169. As Karen J. Jobes and Moisés Silva note, “some elements of the LXX previously attributed to translation technique or recensional activity are now known to represent a Hebrew Vorlage different from the MT,” in Invitation to the Septuagint (Grand Rapids MI: Baker Academic, 2000), 281.

56

Pablo A. Torijano/Julio Trebolle

χεῖρας. In 3 Kgdms 22:15 the Antiochene text presents the recensional form ἐν χειρί, but the rest of the manuscripts transmits the old reading εἰς χεῖρα. Finally, in 4 Kgdms 17:20 only the MS 55 preserves, together with the Armenian version, the OG reading εἰς χεῖρας. The critical edition of 3–4 Kgdms will have in the lemma in every occasion the reading εἰς χεῖρας against ἐν χειρί adopted in Rahlfs’ edition. The LXX secondary versions contribute to establish “certain rules of thumb,” “certain combinations of A B S La Aeth and Co, i.e. those commonly accepted as pre-recensional texts,” […] “often taken as almost automatically embodying the old Greek.”13 In 3–4 Kingdoms two basic combinations or groupings of manuscripts and versions are distinguished. The first, corresponding to the non-kaige γγ section, is formed by B L Cro OL Arm Cop Aeth Jos, compared to the Hexaplaric group (A 247). Texts B and L basically and the versions that follow them transmit the same OG text, although L can hold OG readings more frequently than previously thought. The second grouping corresponds to the kaige βγ and γδ sections and is formed by L Cro OL Arm (Cop Aeth) Jos that transmit an OG text, compared to B Aeth that represent the kaige text. In these sections the pre-Lucian level of the Antiochene text is the sole representative, supported above all by the OL, of the oldest recoverable text from the ancient Septuagint. In this way, the analysis of the critical value of each manuscript or group of manuscripts together with that of the translation characteristics is essential to distinguish the kaige and the pre-Lucianic or OG text, as well as, sometimes, to identify the Hebrew Vorlage used by the Septuagint translator.14

1. The Old Greek Text of 3 Kgdms 12:24a–z The text of 3 Kgdms 12:24a–z is a good testing ground to begin the textual analysis of 3–4 Kingdoms. This passage is a “fossil” that has survived virtually unchanged throughout its textual transmission. As it had no correspondence in MT, it did no undergo the influence of the kaige and Hexaplaric recensions. In this passage the two groups of manuscripts that represent the OG text, B 509 and 19-82-93-107-127 (L), show the translation features of the original version, as the use of the historical present (κοιμᾶται 24a; θάπτεται 24a; πορεύεται 24k) and ἐνώπιον (24a.r) against the aorist and the expression ἐν ὀφθαλμοῖς (MT ‫)בעני‬ characteristics of the kaige recension.

13 Wevers, “The Göttingen Pentateuch,” 51–60, 55. 14  A. Aejmelaeus, “What Can We Know about the Hebrew Vorlage of the Septuagint,” in On the Trail of Septuagint Translators: Collected Essays (Kampen: Kok Pharos, 1993), 71–106.

The Edition of III–IV Kingdoms

57

1.1  The Questionable Text-Critical Value of the Unique Readings of Codex Vaticanus (B) and the Group B 509 External criticism does not attribute as much weight today as in the past to the antiquity of the manuscripts or to their uncial script. Nor does it grant so much authority to the pandect codices as Vaticanus (B)15 or Alexandrinus (A), on which Rahlfs’s edition is based to a great extent. The isolated readings of B against the rest of the textual tradition are generally haplographic and lack critical value,16 as in the following cases in which the secondary versions provide important testimony: 12:24b πόρνη B] pr γυνη rel Aeth Lat Luc; 24e αὕτη B] pr και rel Aeth Lat SyrHmg; 24e ἡ B] ην rel Aeth Lat SyrHmg; 24g ἀρρωστίᾳ κραταιᾷ B] αρρωστιαν κραταιαν omnia SyrHmg; 24k ἐλθούσης B] εισελθουσης rel Aeth SyrHmg Luc; 24k ἐπαποστέλλω B CI 55 244] επαποστελω rel Lat;17 24m κόψεται B 372] κοψονται rel Aeth; 24o δώδεκα 2º B 71 318 707] δεκα rel Aeth.18 Rahlfs follows the reading of the whole manuscript tradition against the isolated reading of B, except for 24e αὕτη B. However, in this case the Aethiopic, the Old Latin and the marginal reading of the Syrohexaplaric text (SyrHmg) support the reading καὶ αὕτη (ἦν) of the rest of the Greek textual tradition. The text of the group B 509 is also defective in 24a Αμμων Β 509 318 460] pr υιων rel Aeth Thdt Rahlfs; υιων αμμος V 19 | 24g πορεύου B 509] pr και Aeth SyrHmg Rahlfs rel. In 12:24q the reading of B 509 246 460 followed by Rahlfs οὕτως ἐλάλησεν πρὸς σὲ ὁ λαός is defective against the reading ὡς ἐλάλησεν πρὸς σὲ ὁ λαὸς λαλήσεις πρὸς τὸν λαόν (CI o x 55 244 245 318 342 372 627 707 Aeth). The reading of B 509 246 460 οὕτως ἐλάλησεν… corresponds to the first part of the complete reading, whereas the reading of the rest of the tradition (V rel) οὕτως λαλήσεις πρὸς τὸν λαόν corresponds to the second part. The OL et sic dices ad populum translates and confirms the critical value of the Antiochene reading and of the rest of the tradition. The group B 509 (plus CI 244 318 372 460) presents the version εἰς ἀπάντην in 24k (plus CI 244 318 372 460) and 24n, against the version εἰς ἀπάντησιν of the rest of the manuscript tradition (and SyrHmg in 24n). Barthélemy classifies the 15  According to Barthélemy, Thackeray is “mesmerized” by the textual type of Codex Vaticanus, Barthélemy, Devanciers, 63. 16  Bo Johnson, Die hexaplarische Rezension des 1. Samuelbuches der Septuaginta (Lund: Gleerup, 1963), 42. 17  According to Kauhanen, “in light of the usage of the future in the passage (see 24l, m) it is best to accept Rahlfs’s solution that the future ἐπαποστελῶ is the original reading. The B reading with the present tense [ἐπαποστέλλω] is probably a corruption from it. The latter reading is probably the one that both La115 and Lucifer follow…,” Kauhanen, Lucifer of Cagliari, 57. The reading inmittam (Legionensis) reflects the future ἐπαποστελῶ. 18  On the grounds of B λάβε σεαυτῶ δώδεκα (δέκα rel Aeth), H. Seebass supposes that Jeroboam was designed king of the “twelfe” tribes of Israel and of the whole of Solomon’s kingdom, but this proposal is based on an isolated reading of Codex Vaticanus, cf., H. Seebass, “Zur Königserhebung Jeroboams I,” VT 17 (1967): 325–333, esp. p. 328.

58

Pablo A. Torijano/Julio Trebolle

version ‫ = לקראת‬εἰς ἀπάντην among the features of the kaige recension versus the OG version εἰς ἀπάντησιν. As a matter of principle, a kaige characteristic is not to be found in 12:24a–z in non-kaige section. In 12:24l all manuscripts read εἰς συνάντησιν. In the same way, in 3 Kgdms 2:35n, in a passage that belongs to the “supplement” of LXX 2:35a–o, the group B 509 has εἰς ἀπάντην against εἰς συνάντησιν of the rest of the manuscripts. The translation of ‫ לקראת‬appears a total of 27 times in kaige section in 1–4 Kingdoms. In 7 cases the reading of only B εἰς ἀπάντην (5x) or εἰς συνάντην (2x) is opposed to the rest of the manuscript tradition εἰς ἀπάντησιν.19 In 9 cases the reading ἀπάντην of B plus some other MSS is opposed also to the reading ἀπάντησιν of the rest of the manuscripts.20 In 5 cases most of the MSS have the form ἀπάντησιν.21 In 4 cases Codex Vaticanus has συνάντησιν (3x) / ἀπάντησιν (1x).22 In non-kaige section ‫ לקראת‬appears translated in 3 Kgdms 2:19 with ἀπάντην (B 379 CII–242’ d s t z) / ἀπάντησιν (rel); 2:35n ἀπάντην (B 509 56 460) / ἀπάντησιν (rel); 20:18 ἀπάντην (V CI x 55 71 158 245 318 372 707) / ἀπάντησιν (A V L CII 247 121 d f o s t z 244 342 460 554); 4 Kgdms 21:27 ἀπάντην (B 370 460) / ἀπάντησιν (rel). Ιn 3 Kgdms 18:16 Codex Vaticanus reads συνάντην versus συνάντησιν in the rest of the manuscripts. Therefore, εἰς ἀπάντησιν seems to be the OG translation of ‫ לקראת‬while εἰς ἀπάντην (or εἰς συνάντην) is not a characteristic of the kaige recension but rather a peculiar feature of B or of B and 509. Once again, the study of the translation technique must be accompanied by the analysis of the manuscript attestation, although frequently it is not easy to combine internal criticism with external evidence.

1.2  Textual Variants between B 509 and L (19-82-93-108-127)

The readings of the two groups of manuscripts representing the OG text, B 509 and L, agree almost completely in 3 Kgdms 12:24a–z. The only significant variant between B 509 and L is found in 3 Kgdms 12:24r: καὶ ἐλάλησεν αὐτοῖς τὰ αὐτά καὶ ταῦτα ἀπέστειλεν πρός με λέγων ὁ λαός (B 598) – καὶ ἐλάλησεν αὐτοῖς τὰ αὐτὰ (MS 19 ταῦτα) λέγων κατὰ ταῦτα ἀπέστειλεν πρός με ὁ λαός (LXXL). The L reading could be considered an example of stylistic and syntactic variatio,23 but the fossilized OL text allows us to recognize the original text of LXX among the 19 2 Kgdms 16:1; 4 Kgdms 2:15; 4:26.31; 5:26; 8:9; 16:10. 20  2 Kgdms 10:5 (B 372); 15:32 (B 119 55 700); 19:16 (B 55 509); 4 Kgdms 5:21 (B CI f 244); 8:8 (B CI f 74 244); 9:18 (B CI 56 244 418); 9:21 (B 56 x−527 318 418); 10:15 (B f); 23:29 (B CI 244). 21  2 Kgdms 19:17 (A V 247 93-127 52 56 370 554 707); 19:21 (V 247 121 19’ 56 372 554); 19:25 (A V L 52 106 92-130 134 55 372 554 707); 3 Kgdms 2:8 (A M V L CI–379 242-530 121 106 488 x 71 158 244 245 342 372); 20:18 (A V L CII 247 121 d f o s t z 244 342 460 554). 22  3 Kgdms 19:26; 4 Kgdms 1:3.6.7. 23  G. Toloni, Jéroboam et la division du royaume. Étude historico-philologique de 1 Rois 11,26 – 12,33 (Turnhout: Brepols, 2017), 173.

The Edition of III–IV Kingdoms

59

different Greek variants. The OL reading et locutus est eis dicens haec et haec mandauit ad me populus translates the Greek καὶ ἐλάλησεν αὐτοῖς λέγων Ταῦτα καὶ ταῦτα ἀπέστειλεν πρός με ὁ λαός. Rahlfs follows the unique reading of the B group: τὰ αὐτά B 509] ταῦτα λέγων 19’; λέγων 246; > Aeth: + λέγων L–19’; ταῦτα rel.24 The reading of the Antiochene text τὰ αὐτὰ λέγων is conditioned by the wrong position of λέγων, as it is also wrong the position of λέγων in B 509 (πρός με λέγων ὁ λαός). The B group and L differ in 3 Kgdms 12:24a.b in the translation of the Hebrew ‫וְ ֵׁשם ִאּמֹו‬: καὶ ὄνομα τῆς μητρὸς (B) / τῇ μητρὶ (L) αὐτοῦ. The distribution of the variants between the non-kaige and kaige sections throughout 3–4 Kingdoms shows that the genitive τῆς μητρός is the OG version, whereas the dative τῇ μητρί is a kaige characteristic. In non-kaige section the version τῆς μητρός is found in six cases.25 In kaige section we have τῇ μητρί in ten cases.26 Despite the weak attestation of τῆς μητρός in kaige sections, the edition should surely give preference to this reading throughout the text of 3–4 Kingdoms. The dative τῇ μητρί of the Antiochene text in 3 Kingdoms 12:12a.b cannot be explained by the influence of the kaige recension, but rather by contamination in the process of textual transmission. The B group and L differ also in the reading πύλην B / πόλιν L (12:24l): πύλην B CI 328 x 55 71 158 244 318 342 372 Aeth Lat Aeth] pr εἰς SyrHmg; εἰς τὴν πόλιν rel. Both texts show this difference throughout 1–4 Kingdoms: 2 Kgdms 10:8; 11:23: τῇ θύρᾳ τῆς πύλης B Rahlfs / πόλεως L (OG);27 4 Kgdms 10:8 τὴν θύραν τῆς πύλης B Rahlfs / πόλεως L (OG); 23:8 ἄρχοντος τῆς πύλης B Rahlfs / πόλεως L (OG). Commenting on 12:24l Kauhanen affirms that “since there is no Greek evidence for a combined reading containing both the words πύλη and πόλις, it is best to accept Lucifer [portam ciuitatis] as a witness to the B reading τὴν πύλην, which is likely the original reading; the reading τὴν πόλιν results from

24 Talshir, The Alternative Story, 122–123, follows Rahlfs’ τὰ αὐτὰ καὶ ταῦτα (‫)וכזאת כזאת‬. She acknowledges that 19-82-93-108-127 present “a reasonable text”, but it “should not be taken as the original text.” 25  12:24a (τῇ μητρί L d 245 Thdt); 12:24b (τῇ μητρί L d−125 246); 14:21; 15:2 (τῇ μητρί 246 55 71); 15:10; 16:28a. The translation τῆς μητρός appears also in two Hexaplaric additions: 3 Kgdms 11:26 A 247 Arm SyrH (* α′ SyrH) and 14:31 A 247 CII f d-125 s t z 554 Arm SyrH (* α′ SyrH). 26 3 Kgdms 22:42; 4 Kgdms 15:2 (μητρός V L 700 328 460 Aeth Syrh Cyr); 18:2 (μητρός L 460 Aeth Syrh); 21:1 (μητρός 19’ Aeth); 21:19 (μητρός 247 82 121 o 488 x 318 342 372 460 Aeth); 22:1 (μητρός CI 328 244 342 Aeth); 23:31; 23:36 (μητρός 19’ Aeth); 24:8; 24:18 (μητρός 55 Aeth). 27  “LXXL has preserved the OG, regardless of whether its Vorlage had ‫ ָּׁש ַער‬or ‫עיר‬, ִ ” Leonardo Pessoa da Silva Pinto, Different Literary Editions in 2 Samuel 10–12. A Comparative Study of the Hebrew and Greek Textual Traditions (Madrid: CSIC, 2019), 93.

60

Pablo A. Torijano/Julio Trebolle

a transcriptional error.”28 However, the OL portam ciuitatis seems to attest rather a reading like those of 4 Kgms 7:10 πύλην τῆς πόλεως; 23:8 πύλῃ τῆς πόλεως. The form Σολομων (L) versus Σαλωμων (B 509) is not strictly Antiochene since, as Rahlfs already noted, it is found in Eupolemus, Josephus and the New Testament, so it cannot be said that we are dealing here with a late correction.29 Σολομων is the form transmitted by the majority Greek text and the OL as 12:24b Σαλωμων B V* 509 64 55 245 372] σολομων d−44 106 71 342; σαλομων Vc CI 328 381 x 158 244 318 460 627; σωλομωντος 93; σολομωντος rel; Solomonis Lat Luc; 12:24p Σαλωμων B V* 509 64 55 245 372 707] σαλομων Vc CI 328 x 158 244 318 460 627; σολομων 342; Σολομωντος rel; Solomonis Lat. The same combination of witnesses with slight changes is observed in 12:24a.b.c.d.

1.3  The Witness of the Secondary Versions

There is a tendency to consider that the Lucianic MSS (19-82-93-108-127-700) transmit a text that is isolated from the rest of the tradition. However, this text form was the one translated in the old versions before the diffusion of the Hexaplaric recension, which explains the coincidences between versions that did not have contact between them, as the OL, the Coptic, Ethiopic, Armenian and Georgian versions. Barthélemy speaks about a substantial identification of L with the text used also in Alexandria and that Origen inherited from Clemens of Alexandria, a text form substantially identical with the one preserved in the Antiochene tradition.30 The OL represents a pre-Lucianic Greek text, very close to OG. This textual filiation is the ground for the critical value of the OL text. The passage of 3 Kgdms 12:24k.l is a unique case where the three witnesses of OL have preserved an identical text that, with light intra Latin variants, translates faithfully a pre-Lucianic Greek text: the Vienna Palimpsest (Vindobonensis, L115),31 the text quoted by Lucifer of Cagliari32 and the marginal readings in Leon Bible codices (Beuron 9195).33 The only significant variant readings mihi attulisti (Lucifer, Legionensis) and attulisti mihi (Vindob.) reflect respectively the Greek readings μοι ἐνήνοχας (B CI 509 244 318 460) and ἐνήνοχάς μοι (rel Aeth SyrHmg Thdt). Rahlfs follows 28 Kauhanen, Lucifer of Cagliari, 59. 29 Rahlfs, Lucians Rezension, 184. 30 Barthélemy, Devanciers, 138–139. 31  B. Fischer, “Palimpsestus Vindobonensis: a Revised Edition of L115,” BIOSCS 16 (1983): 13–87; reprinted as: B. Fischer, “Palimpsestus Vindobonensis. II: Manuscript 115 of the Books of Kingdoms,” in Beiträge zur Geschichte der lateinischen Bibeltexte (Freiburg: Herder, 1986), 315–333. 32  G. F. Diercks, “De regibus apostaticis,” in Luciferi Calaritani Opera quae supersunt. Ad fidem dvorum codicum qui adhuc extant necnon adhibitis editionibvs veteribvs, ed. G. F. Diercks (Turnholt; Brepols, 1978), I. 58–60. 33 Moreno, Las glosas marginales de Vetus Latina.

The Edition of III–IV Kingdoms

61

the B 509 text, but the Aethiopian version (Aeth) and the marginal reading of the Syro-hexaplaric (SyrHmg) follow the rest of the MSS, the Lucianic ones included. The L readings attested in Aeth and SyrHmg are significant, both by their quantity and the quality of the Greek text they represent: 24c ἐζήτει Rahlfs] εζητησεν L−82 246 Aeth | 24d ἐγώ Rahlfs] post μου tr. Aeth; > L 328 d−44 125 246 o 245 318 Aeth SyrHmg | 24e πρεσβυτέραν Rahlfs] + αδελφην L 328 106-107 SyrHmg; + και αδελφην 610; + Nuham Lat; 24e ἡ B] ην rel SyrHmg Rahlfs | 24f ἀπελεύσομαι B V CI−328 509 o x 55 71 158 244 245 318 342 372 460 554 627 707 Rahlfs] -σωμαι 158 245; + καὶ ἀπέστειλεν αὐτὸν Σουσακειμ rel SyrHmg | 24f γῆν Rahlfs] pr. την 509 460; την L d−44 125 246 372 Aeth SyrHmg | 24f Εφραιμ 2º B V 82 CI 328 509 106-107 71 244 318 460 Rahlfs] + προς Ιεροβοαμ rel Aeth | 24g ἀρρωστίᾳ κραταιᾳ B] αρρωστιαν κραταιαν omnia SyrHmg Rahlfs | 24g πορεύου B 509 460] pr. και rel Aeth SyrHmg Rahlfs | 24k πορεύεται Rahlfs] επορευθη L−127 SyrHmg LatLuc | 24k ἐλθούσης B] εισελθουσης rel Aeth SyrHmg LatLuc Rahlfs | 24l τὴν πύλην B CI 328 x 55 71 158 244 318 342 372 Aeth Lat (Vindobonensis) Luc] pr. εις SyrHmg; τη πυλη v: εις την πολιν rel; > V | 24m κόψεται B 372] κοψονται rel Aeth Rahlfs | 24n ἀπαντήν B 509 46 Rahlfs] απαντησιν V CI o x 55 71 244 245 318 372 627vid 707; απαντησιν αυτης rel SyrHmg | 24o λόγος κυρίου ἐγένετο Rahlfs] εγενετο λογος κυριου V 328 527 Aeth | 24o ισραηλ] + συ βασιλευσεις 460; + βασιλευσεις L Aeth | 24t εἶπεν] ειπον Aeth | 24u το σκηνωμα] τα σκηνοματα L 509 246 Aeth.

Aeth agrees with rel in 24l, 24m, 24o; with L OL in 24o; with L in 24c, 24l, 24o and L in 24t, 24u. SyrHmg follows omnia / rel in 24f, 24g, 24m, 24n; it agrees with rel Aeth OL in 24k (2º); with L OL in 24k and with L in 24d, 24e, 24f. The marginal readings of the Syro Hexaplaric text are taken from the Greek “Lucianic” but they follow omnia or rel, agreeing with Aeth OL or with L. Therefore, these readings cannot a priori be considered late or “Lucianic”. They may sometimes represent pre-Lucianic readings going back to the OG (cf. infra).

2. The Kaige and the Old Greek Texts of 3–4 Kingdoms The observations made on 3 Kingdoms 12:24a–z are to be extended to the whole text of 3–4 Kingdoms, but the textual situation is now much more complex, especially regarding the γδ kaige section. To the difficulty in identifying and isolating the kaige and pre-Lucianic / OG texts, the difficult of making a critical judgment on textual and literary phenomena that are absent in 12:24a–z is added, such as alternative readings, double readings and transposition of clauses or even larger sections of text. In the opening lecture for the Centenary celebration of the Göttingen Septuaginta Martin L. West points out a basic question in textual criticism of classical texts, also crucial in LXX textual criticism:

62

Pablo A. Torijano/Julio Trebolle

“how far it is legitimate to regularize the linguistic dress in which the poems appear, and to restore older forms than those given by the manuscripts. In certain cases, more archaic forms appear in some passages and their more modern equivalents in others. It is certainly possible that the poet was inconsistent. But as we know that the tradition was subject to creeping modernization, I take the view that we are more likely to hit the truth by generalizing the ancient forms attested in some passages than by ascribing the inconsistency of the tradition to the poet.”34

The critical analysis must be carried out case by case and reading by reading, as it is usually affirmed, but this way of proceeding should not lead to “ascribe inconsistency” to the translator who is rather consistent in using specific translation features, while, on the contrary, the recensors may be inconsistent in their revisions, especially in their first attempts as is the case of the kaige recension. On the other hand, the analysis of external criticism must also be governed by a series of constants of the textual tradition and certain combination of witnesses that generally transmit the oldest text. The necessary internal coherence of the edition of 3–4 Kingdoms entails to “generalize the ancient forms.” This sometimes leads to restore OG forms attested only by late witnesses or by the secondary versions and to reinstate such forms when they have been substituted by “more modern equivalents” or by readings based in the Hebrew text that superseded the Hebrew of the Greek version.

2.1  The Text-Critical Value of the Unique Readings of Codex Vaticanus and 509 The unique readings of Codex Vaticanus, followed sometimes by a few manuscripts, are very frequent throughout 3–4 Kingdoms. They generally lack any critical value as in the following examples taken from only three verses, 3 Kingdoms 21:14–17 (in kaige section).35 21:14 χορῶν B 247 93 d Cop] πολεων A SyrH; χωρων rel Arm Aeth Luc. v. 15 τοὺς ἄρχοντας τὰ παιδαρία B] τους αρχοντας και τα παιδαρια 108; τα παιδαρια τους αρχοντας 318; τους αρχοντας και τα παιδαρια των αρχοντων L−19′ CI 244; om. τα παιδαρια 19; τους παιδας (τα παιδαρια 243; om. τους 236) των αρχοντων rel Arm SyrH. v. 15 τῶν χορῶν B d−610 71] > 247 121 488 z; των χωρων rel Arm SyrH.

34  M.L. West, “Critical Editing,” in Die Göttinger Septuaginta: Ein editorisches Jahrhundertprojekt, ed. R.G. Kratz and B. Neuschäfer (Berlin: de Gruyter, 2013), 13–27, 25. 35  Montgomery expressed “his criticism of uncritical use of B,” cf, James A. Montgomery, A Critical and Exegetical Commentary on the Books of Kings, ed. H. Gehman, ICC (Edinburgh: T&T Clark, 1951) 10.

The Edition of III–IV Kingdoms

63

v. 15 διακόσια B 243-731 o 244 318 342] σ′ 92-214-488-762 74 158; τριακοσιοι A 68 71; Σλβ 247 98′ 46-236-242 107-125-610 246 74 245 554; διοκοσιοι rel Jos. v. 15 ἑξήκοντα B] + (pr επτα CI o x 244 318 372) χιλιαδας rel Arm Copt SyrH. v. 17 ἄρχοντες παιδαρία τῶν χορῶν B] οι αρχοντες και τα παιδαρια των αρχοντων των χωρων L 328 246; παιδαρια (pr. τα CI 44 o) των (> A CI o x 244 318 372) αρχοντων των χωρων rel Arm Aeth SyrH; pueri de ciuitate principum regionum LatLuc..36

Rahlfs follows generally the “long” reading against the “short” haplographic reading of B, even in instances in which it is marked with asterisk as in 3 Kgdms 20:10 παρανόμων B] εξ εναντιας αυτου… ΑΝ omn Arm La Aeth SyrH (* 127 θ′). 4 Kgdms 1:14 τοὺς πρώτους B 460] καὶ ἑκάστου τοὺς πεντήκοντα L; καὶ τοὺς πεντήκοντα αὐτῶν rel Arm SyrH (sub * α′ σ′ θ′) | 2:18 καί 1º B 530 245 460 707 Aeth] καὶ ἐπέστρεψαν αὐτοί 247 121; καὶ ἀνέστρεψαν πρὸς αὐτόν (sub * 56 127) rel Arm SyrH* α′ σ′ θ′ | 3:13 σου B 55 245 460 707 Aeth] + καὶ πρὸς τοὺς προφήτας τῆς μητρός σου rel Arm Syr (sub * α′ σ′ θ′).

In other cases, Rahlfs follows the possibly defective reading of B against the (almost) whole manuscript tradition and the secondary versions that may represent the OG text, already known by Aquila, Symmachus or Theodotion as in 3 Kgdms 20:1: καί Β V 318] + ἐγένετο μετὰ τὰ ῥήματα ταῦτα καί omn Arm Aeth SyrH (sub * α′ SyrH) (= MT) | 18:46 ἐπί B 460] pr ἐγένετο rel Arm Lat (erat) SyrH (sub *), attributed to Symmachus | 4 Kgdms 1:14 τοὺς πρώτους B 460] καὶ ἑκάστου τοὺς πεντήκοντα L; καὶ τοὺς πεντήκοντα αὐτῶν rel Arm SyrH (sub * α′ σ′ θ′).

The same can be said of 3 Kgdms 18:46 ἐγένετο, witnessed by the whole Greek tradition and also the OL (erat, Lucifer) and attributed to Symmachus, against the reading ἐπί of only B and 460, followed by Rahlfs. Equally in kaige section Rahlfs follows the isolated reading τοὺς πρώτους of B and 460 in 4 Kgdms 1:14 against the reading καὶ ἑκάστου τοὺς πεντήκοντα L; καὶ τοὺς πεντήκοντα αὐτῶν rel Arm SyrH (sub * α′ σ′ θ′). This reading agreeing with MT could be OG text known by the “three.” Again, a conflict arises between the manuscript attestation of a reading and the application of a principle of textual criticism, between rec36  In 3 Kgdms 21:14.15.17.19 A. Schenker follows χορῶν attested only in Codex Vaticanus (“boys of the choir leaders”) against χωρῶν (“administrative districts”) of the rest of the manuscript tradition and the versions, particularly the OL (regionum), A. Schenker, “Junge Garden oder akrobatische Tänzer? Das Verhältnis zwischen 1 Kön 20 MT und 3 Regn 21 LXX,” in The Earliest Text of the Hebrew Bible. The Relationship between the Masoretic Text and the Hebrew Base of the Septuagint Reconsidered, ed. A. Schenker (Atlanta: SBL, 2003) 17–34. The reading χορῶν seems to be a mistake of copying by confusion of ω/ο.

64

Pablo A. Torijano/Julio Trebolle

ognizing a haplography in B or considering that the LXX original attests the same reading as MT. As the unique readings of Codex Vaticanus, those of the group B 509 (and in several cases 460) present also a defective text on numerous occasions. These are a few examples taken from the initial chapters of the two books (in 4 Kgdms 509 is not extant): 3 Kgdms 1:17 κύριέ B 509] + μου (>L–82) βασιλεῦ rel Arm Aeth SyrH; 1:46 βασιλείας B 509 460] pr τῆς rel | 1:47 Σαλωμων B 509 460] + τοῦ υἱοῦ σου rel Arm Aeth SyrH SyrJ | 2:19 τοῦ θρόνου B 509 460] τὸν θρόνον αὐτοῦ Z L; αὐτοῦ rel Arm Aeth (uid) SyrH | 2:35c τὸν οἶκον B 509] pr. αὐτοῦ τὸν οἶκον καὶ L; + αὐτοῦ καὶ τὸν οἶκον rel Aeth SyrH | 2:46 αὐτόν B 509 460 Aeth] + καὶ ἀπέθανεν ὁ σεμεει L; + καὶ ἀπέθανεν rel Arm SyrH | 2:46e σαλωμων B 509 Rahlfs] + ἐν ἡμέρα μιᾁ rel Aeth | 3:2 θυμιῶντες B 509 460 Aeth] πλὴν ὁ λαός ἦσαν θυμιῶντες καὶ θύοντες Z L 158 Thdt; πλὴν ὁ λαός ἦσαν θυμιῶντες rel Arm SyrH | 3:24 λάβετέ B 509 460 Aeth] + μοι rel Arm Lat SyrH. 4 Kgdms 1:11 καί 4º B Aeth] pr. καὶ ἀνέβη rel Arm SyrH | 1:13 ἡγούμενον B] πεντηκόνταρχον τρίτον L−19 Arm SyrH; + πεντηκόνταρχον τρίτον rel | 1:13 τούτων B 44 460] τῶν πεντήκοντα L 328; om. Aeth; + τῶν πεντήκοντα rel Arm SyrH | 1:14 μου B 460 Aeth SyrH] τῶν δούλων σου rel Arm | 1:16 ἀπέστειλας B CI 244 460 Aeth] + ἀγγέλους L f o x–119 245 318 242 372 Aeth SyrH; ἐξαπέστειλας ἀγγέλους rel | 2:10 ἔσται B Arm] + σοι rel Aeth SyrH Chr | 2:14 τὸ ὕδωρ BA] τὰ ὕδατα καὶ οὐ διῃρέθη L 106 158 Compl; aquas Arm; + καὶ οὐ διέστη (διῃρέθη 460) rel Aeth SyrH; 3:12 ῥῆμα B 460] + κυρίου rel SyrH | 3:25 πηγήν B 82 460 Aeth] + ὕδατος rel Arm SyrH | 4:2 οὐθέν B] pr. in domo eius Aeth; ἐν τῷ οἴκω SyrH; οὐδὲν ἐν τῷ οἴκω rel; ἐν τῷ οἴκω tr. post ἀλλ᾽ ἤ L.

In all these cases the B 509 reading is clearly defective and the reading of the rest of the manuscript tradition is to be followed. But there are numerous cases in which the reading transmitted by the Greek tradition (except B 509) is marked with an asterisk as if it were a Hexaplaric addition. In such cases the question arises as to whether the reading absent in B 509 is a Hexaplaric addition as indicated by the asterisk or, rather, an OG reading known also by the Jewish recensors but lost in B or in the group B 509. The haplographic character of this group makes problematic to qualify as “Hexaplaric” readings witnessed by the whole Greek tradition and the secondary versions against the reading of B 509. Rahlfs follows the B or B 509 readings in cases where the long reading of the rest of the tradition may be OG as in 3 Kgdms 1:35 και B 56 509 Cop] pr. (sub * V 127; sub * θ′ SyrH) και αναβησεσθε (-σεσθαι A 82 242 460*) οπισω αυτου (+ και αναβησεσθε post αυτου 342) και εισελευσεται Α Μ V O L CI CII 121 d o s t x z mixti–245 707 Arm Aeth Geor SyrH |

The Edition of III–IV Kingdoms

65

2:34 και 1º Β] και ανεβη βαναιας υιος ιωδαε (om υιος Ιωαδ d–106 Geor) rel Arm Aeth Geor (* θ′) | 2:42 αποθανη 121-509] και ειπας μοι αγαθον το ρημα ο ηκουσα rel Arm SyrH (sub * 127 pr * θ′ SyrH) | 4:4 και 1º B 121-509 f z 245 460 707 Aeth] pr. και βαναιας υιος ιωδαε επι της στρατιας rel Arm Lat SyrH (sub * θ′).37

In other instances, Rahlfs prefers the reading of the majority text even it is marked with an asterisk: 3 Kgdms 2:5 πολέμου Β] + εν ειρηνῃ και εδωκεν αιμα αθῳον (innocentium La) O 243 52 121b 106-107 92-130-314-489 120-134 554 Arm Lat SyrH (sub *) | 2:46 αὐτόν B 509] + και απεθανεν ο σεμεει L; και απεθανεν rel Arm SyrΗ | 6:10 (6:5 MT) δαβιρ B V 509 71 245 460 707 Lat] + και εποιησεν πλευρας κυκλοθεν rel Arm Aeth SyrH (πλευρας κυκλοθεν * SyrH) | 6:23 (MT 24) πτερύγιον B 509] το πτερυγιον αυτου το εν και πεντε πηχεων πτερυγιον O Arm SyrΗ; πτερυγιον του χερουβ του ενος και πεντε πηχεων πτερυγιον* rel | 6:23 (MT 24) αὐτοῦ 2º B 509 106 o 74 71 342 460] + και εως μερους πτερυγιου A L SyrH sub *; + εις μερος πτερυγιου αυτου rel | 18:29 θυσίαν Β 82-93] + και ουκ ην φωνη και ουκ ην ακροασις SyrH; + καὶ οὐκ ἦν φωνή A-247 127 rel Arm; 16:15 επτα ετη B 509 Aeth] επτα ημερας rel Arm Lat SyrH | 4 Kgdms 1:14 τοὺς πρώτους Β] και εκαστου τους πεντηκοντα L; και τους πεντηκοντα αυτων rel Arm SyrH (τους πρωτους sub * α′ σ′ θ′) | 2:18 καὶ Β 530 245 460 707 Aeth] pr και επεστρεψαν αυτοι 247 121 488; και ανεστρεψαν προς αυτον (sub * 127) rel Arm SyrH (sub * α′ θ′).

As stated by Barthélemy the group B 509 is the only direct witness to the text used by Origen in his fifth column, from which the text of the Hexaplaric group of manuscripts comes (A 247 = O).38 The critical decision on the “approximations to MT” in OG readings probably lost in B 509 is one of the thorniest issues in the editing work of 3–4 Kingdoms.39 In 4 Kgdms 5:15 ἔστη B 82 245 707] + ενωπιον αυτου (sub * 127) rel Arm Aeth ; + εις προσωπον αυτου A (* α′ σ′) one may wonder if the B reading is OG (Rahlfs) or rather it has lost ἐνώπιον αὐτοῦ, so that this reading is not an assimilation to MT (‫)ל ָפנָ יו‬ ְ but the OG version known to Aquila and Symmachus and reflecting the Hebrew ‫וַ ּיַ ֲעמֹד ְל ָפנָ יו‬. OG regularly translates the Hebrew expression “to stand before X” by ἐνώπιον (τοῦ),40 except in 3 Kgdms 1:2 where the verb παρίστημι is followed by the dative. However, the Hexaplaric 37  The term στρατιάς instead of δυνάμεως (M V L 328 121 f o x z 55-71-158-318-342) may vouch for the OG nature of the “addition.” 38 Barthélemy, Dévanciers, 138–9. 39  T.M. Law, Origenes Orientalis. The Preservation of Origen’s hexapla in the Syrohexapla of 3 Kingdoms (Göttingen: Vandenhoeck & Ruprecht, 2011), 321–361. 40  ιστημι ενωπιον, 3 Kgdms 1:28; 3:16; 19:11; 22:21; 4 Kgdms 4:12; 5:15;8:9; 25:8; ιστημι κατα προσωπον, 3 Kgdms 3:15; 8:22; 4 Kgdms 10:4; παριστημι ενωπιον, 1 Kgdms 6:20; 16:22; 3 Kgdms 1:2; 10:8: 12:6.(8.10 προ προσωπου); 17:1; 18:15: 4 Kgdms 3:14; 5:16.

66

Pablo A. Torijano/Julio Trebolle

manuscripts A 247 have ἐνώπιον, L ἔναντι τοῦ βασιλέως and SyrH ἔμπροσθεν αὐτοῦ, a reading marked with an asterisk and attributed to Theodotion.41 External evidence seems to make it clear that this is a Hexaplaric approximation to MT, but internal criticism seems to support the reading that responds to a constant characteristic of the OG version, ἵστημι/παρίστημι ἐνώπιον with genitive. A translator may be inconsistent at times, but for some justifiable reason. On the other hand, recensors and copyists or groups of manuscripts are often inconsistent. In many of the alleged “approximations to MT” or “Korrekturen nach M” (Rahlfs) textual criticism is torn between giving greater weight to internal criticism or to external evidence. The principle according to which the Greek reading closest to MT is to be regarded as secondary42 is not equally applicable to the translator who works on a non-Masoretic Hebrew text and to a recensor or a copyist who works on a Greek text which may not correspond exactly to the Old Greek. Among the corrections of L according to MT noted by Rahlfs is that of 4 Kgdms 25:18 υἱὸν τῆς δευτερώσεως] τον ιερεα τον δευτερον L Arm SyrH(mg). According to Peter Walters the translation of ‫ ּכ ֵֹהן ִמ ְׁשנֶ ה‬as υἱὸν τῆς δευτερώσεως “must be emended into ιερεα τ.δ. as L and Aquila rightly read, cf. 23:4.”43 Moreover δεύτερος is the OG translation of ‫( ִמ ְׁשנֶ ה‬1 Kgdms 8:2; (17:13); 23:17; 2 Kgdms 2:3), while δευτέρωσις is the kaige counterpart (4 Kgdms 23:4; 25:18; cf. Sir 42:1).44 The L reading τὸν ἱερέα τὸν δεύτερον attested by Arm SyrH(mg) is pre-Lucianic and OG. Its agreement with Aquila and MT simply means that LXX and MT transmit in this case and surely in many others the same text. Rahlfs also points out as correction of L according to MT the reading of 4 Kgdms 3:18 καὶ δώσει κύριος τὴν Μωαβ εἰς χεῖρας ὑμῶν compared to the majority text καὶ παραδώσω τὴν Μωαβ ἐν χειρὶ ὑμῶν. According to Rahlfs εἰς χεῖρας harmonizes with v. 13 and δώσει depends on MT ‫וְ נָ ַתן‬. However, εἰς χεῖρας is OG versus kaige ἐν χειρί and δώσει has to be explained in relation to the preceding L reading καὶ κοῦφον τοῦτο ἐνώπιον κυρίου ποιῆσαι αὐτό versus the majority reading καὶ κούφη αὕτη ἐν ὀφθαλμοῖς κυρίου. L ἐνώπιον κ. is OG while 41 Law, Origenes Orientalis, 45 and 256. According to Rahlfs the Syriac was back translated by Field by εμπροσθεν, but the Syriac term may correspond to ενωπιον, see Rahlfs, Lucians Rezension, 163, note 1. 42  In the comparison between the texts of MT and LXX, Lagarde’s criterion, as formulated by S.R. Driver, is applied: “If two readings co-exist, of which one expresses the Masoretic text, while the other can only be explained from a text deviating from it, the latter is to be regarded as the original,” S.R. Driver, Notes on the Hebrew Text of the Books of Samuel (Oxford: Clarendon Press, 1890; 19132), XLIV. 43 Peter Walters (formerly Katz), The Text of the Septuagint: Its Corruptions and their Emendation, ed. D.W. Gooding (London: Cambrige University Press, 1973), 313. 44  J. Trebolle, “Recensional Criticism of 4 Kingdoms 25:18–1,” in Textual and Literary Criticism of the Books of Kings. Collected Essays, ed. Andrés Piquer Otero and Pablo A. Torijano, VTS 185 (Leiden: Brill, 2020) 12–15.

The Edition of III–IV Kingdoms

67

ἐν ὀφθαλμοῖς κ. is kaige. Rahlfs considers that κοῦφον τοῦτο improves the mechanical translation of the majority text κούφη αὕτη (= MT ‫)נָ ַקל זֹאת‬, which for this very reason is a kaige reading.45 The form κούφη αὕτη is irregular, since the usual expression is κοῦφον + infinitive translating Hebrew ‫( קלל‬niphal) + infinitive followed by a verb with the same subject: “And as if it had been a light thing for him to walk…, he took as his wife,” καὶ οὐκ ἦν αὐτῷ ἱκανὸν τοῦ πορεύεσθαι… καὶ ἔλαβεν γυναῖκα =‫( וַ יְ ִהי ֲהנָ ֵקל ֶל ְכּתֹו… וַ ּיִ ַּקח ִא ָּׁשה‬1 Kgs 16:31); “It is normal for the shadow to lengthen ten intervals; rather let the shadow retreat,” κοῦφον τὴν σκιὰν κλῖναι… ἐπιστραφήτω ἡ σκιά =‫( נָ ֵקל ַל ֵּצל ִלנְ טֹות… יָ ׁשּוב ַה ֵּצל‬2 Kgs 20:10).46 The L reading in 4 Kgdms 3:18 seems to go back to the OG καὶ κοῦφον ἐνώπιον κυρίου ποιῆσαι τοῦτο καὶ δώσει τὴν Μωαβ εἰς χεῖρας ὑμῶν and by conjecture to the Hebrew ‫ת־מֹואב ְּביֶ ְד ֶכם‬ ָ ‫וְ נָ ַקל ְּב ֵעינֵ י יְ הוָ ה ַל ֲעשו ֺת זֹאת וְ נָ ַתן ֶא‬. The Septuagint is a work a se stante whose translator fulfills the role of author, so that the edition of a LXX book could be carried out in principle with the only data of the Greek manuscript tradition and its versions. Nevertheless, the edition of 3–4 Kingdoms, unlike the edition of other LXX books, cannot be done independently of the LXX’s Hebrew Vorlage as regards the OG and independently of the (proto-) MT as regards the text of the kaige and Hexaplaric recensions.

2.2  The Critical Value of the Versions According to Wevers, the collation of the daughter versions “does represent the weak link in the preparation of the Göttingen editions;” “the one area where change is both possible and needed is the work on the daughter versions.”47 The secondary versions of the Pentateuch contribute to establish “certain rules of thumb” or “certain combinations of A B S La Aeth and Co, i.e. those commonly accepted as pre-recensional texts, […] often taken as almost automatically embodying the old Greek.”48 In the same way, the combination of L OL Aeth, Co, Arm, Georg, Josephus usually represents the OG in the kaige sections. The secondary versions are frequently decisive to establish the textual history of 3–4 Kingdoms and to reconstruct the oldest stages of its text. Paralipomena is the oldest testimony of 4 Kingdoms as witness of pre-Lucianic/OG readings in

45 Rahlfs, Lucians Rezension, 243–244 and 272. 46 Cf. likewise εἰ κοῦφον ἐν ὀφθαλμοῖς ὑμῶν ἐπιγαμβρεῦσαι = ‫יכם ִה ְת ַח ֵּתן‬ ֶ ֵ‫ַהנְ ַק ָּלה ְב ֵעינ‬ (1 Sam 18:23); μέγα σοί ἐστιν τοῦ κληθῆναί σε παῖδά μου τοῦ στῆσαι = ‫יֹותָך ִלי ֶע ֶבד ְל ָה ִקים‬ ְ ‫נָ ֵקל ִמ ְה‬ (Isa 49:6); μὴ μικρὰ τῷ οἴκῳ Ιουδα τοῦ ποιεῖν = ‫הּודה ֵמ ֲעׂשֹות‬ ָ ְ‫( ֲהנָ ֵקל ְל ֵבית י‬Eze 8:17); κοῦφον ἐν ὀφθαλμοῖς κυρίου … πλουτίσαι (Sir 11:21). 47  J.W. Wevers, “The Future of Septuagintal Textual Studies,” in The Bible as Book. The Transmission of the Greek Text, ed. S. McKendrick and O.A. O’Sullivan (New Castle: Oak Knoll Press, 2003), 209–219, esp. 213. 48 Wevers, “The Göttingen Pentateuch,” 55.

68

Pablo A. Torijano/Julio Trebolle

this kaige section,49 as for example in 2 Par 23:8 Ιωδαε ὁ ἱερεύς … καὶ ἔλαβον ἕκαστος τοὺς ἄνδρας αὐτοῦ, 4 Kgdms 11:9 Ιωδαε ὁ συνετὸς ἱερεύς … καὶ ἔλαβον ἕκαστος τοὺς ἄνδρας αὐτοῦ L, Ιωδαε ὁ συνετός, καὶ ἔλαβεν ἀνὴρ τοὺς ἄνδρας αὐτοῦ B.50 The OL has an extraordinary critical value for recognizing pre-Lucianic readings and establishing the Old Greek text. According to Rahlfs and Dieu the OL Lucianisms would be late revisions based on readings of a Lucianic text common in the West,51 but as stated by Fischer “the more Lucianic readings the OL has, the older its text is.”52 A basic task in the study of the OL consists in discriminating between the original OL readings which translate a pre-Lucianic/OG text and the readings which reflect a kaige Greek text and entered in the OL textual tradition at a late stage.53 In 4 Kgdms 17:2 OL prae omnes qui fuerant ante eum (L 115) translates the pre-Lucianic/OG παρὰ πάντας τοὺς γενομένους, whereas non sicut reges Israel qui fuerant ante ipsum (L 91–95 Al.) reflects οὐχ ὡς οἱ βασιλεῖς Ισραηλ οἳ ἦσαν (kaige, LXXAB). The OL reflects the pre-Lucianic/OG readings against the lexical options of the kaige revision, as in the following instances: 3 Kgdms 49  The fact that only the OL of the second book of Paralipomena has been preserved has induced Hanhart to publish the edition of the second book before the first. This fact contributes to get a clearer picture of the textual history of 3–4 Kingdoms, cf. R. Hanhart, Paralipomenon liber II (Göttingen: Vandehoeck & Ruprecht, 2014), 4. 50  Other examples in 2 Par 23:10 ἕκαστον, 4 Kgdms 11:11 ἕκαστος L, ἀνήρ B; 2 Par 25:4 τέκνων, 4 Kgdms 14:6 τέκνων L, υἱῶν B; 2 Par 25: 4 ἀλλ᾽ ἤ, 4 Kgdms 14:6 ἀλλ᾽ ἤ L, ὅτι ἀλλ᾽ ἤ B; 2 Par 26:4 ἐνώπιον 4 Kgdms 15:3 ἐνώπιον L, ἐν ὀφθαλμοῖς B; 2 Par 33:3 στρατία, 4 Kgdms 21:3 στρατία L, δυνάμει B; 2 Par 33:6 ἐπλήθυνεν τοῦ ποιῆσαι τὸ πονηρὸν ἐναντίον κυρίου τοῦ παροργίσαι αὐτόν, 4 Kgdms 21:6 ἐπλήθυνε τοῦ ποιῆσαι τὸ πονηρὸν ἐνώπιον κυρίου τοῦ παροργίσαι αὐτόν L, ἐπλήθυνεν τοῦ ποιεῖν τὸ πονηρὸν ἐν ὀφθαλμοῖς κυρίου παροργίσαι αὐτόν B; 2 Par 33:8 κατὰ πάντα τὸν νόμον, 4 Kgdms 21:8 κατὰ πάντα τὸν νόμον L, κατὰ πᾶσαν τὴν ἐντολήν B, Hanhart, Paralipomenon liber II, 290, 307, 314, 317, 328, 383–385. 51 Rahlfs, Lucians Rezension, 161; L. Dieu. “Retouches Lucianiques sur quelques textes de la vieille latine (I et II Samuel),” RB 16 (1919): 390–403; according to Brock, OL may witness old readings but these may have originated from later correctors who used L manuscripts, S. Brock. “Lucian redivivus: Some Reflections on Barthélemy’s Les Devanciers d’Aquila,” Studia Evangelica 5 (1968): 176–181; S. Brock, The Recensions of the Septuaginta Version of 1 Samuel (Torino: S. Zamorani, 1996), 217–218. 52  B. Fischer, “Lukian-Lesarten in der Vetus Latina der vier Königsbücher,” Studia Anselmiana 27–28 (1951): 169–177, especially 173, 175, 177. 53  M. Bogaert, “Bulletin de la Bible latine (1955–75),” Bulletin d’ancienne littérature chrétienne latine 5, Revue bénédictine de critique, d’histoire et de littérature religieuses 74–84 (1964– 74): 162; E. Ulrich, “The Old Translation of the LXX and the Hebrew Scrolls from Qumran,” in The Dead Sea Scrolls and the Origins of the Bible, ed. E. Ulrich (Grand Rapids / Leiden: Eerdmans / Brill, 1999), 233–274; J. Trebolle, “From the ‘Old Latin’ through the ‘Old Greek’ to the ‘Old Hebrew’ (2 Kings 10:23–25),” Textus 11 (1984): 17–36; A. Schenker, Älteste Textgeschichte der Königsbücher. Die hebräische Vorlage der ursprünglichen Septuaginta als älteste Textform der Königsbücher (Fribourg: Academic Press; Göttingen: Vandenhoeck & Ruprecht, 2004) 134–167.

The Edition of III–IV Kingdoms

69

1:34 tuba = σάλπιγγι L/OG (κερατίνῃ B kaige); 22:10 unusquisque = ἕκαστος L/ OG (ἀνήρ B kaige); 4 Kgdms 4:29 and 8:10 vade = πορεύου L/OG (δεῦρο B kaige); 7:9 iniquitatem = ἀδικίαν L/OG (ἀνομίαν B kaige).54 The OL attests sometimes Antiochene readings that involve a different Hebrew Vorlage, as in 4 Kgdms 1:2 inquirite = ἐπερωτήσατε L/OG (‫ ≠ )שאל‬ἐπιζητήσατε Β kaige (‫ ;)דרש‬3:9 ascendit = ἀνέβη L/OG (‫ ≠ )ויעל‬ἐπορεύθη Β kaige (‫)וַ ּיֵ ֶלְך‬.55 The Ethiopic version of Kings reproduces faithfully according to Dillman and Rahlfs a B text like that of Codex Vaticanus. Gehman affirms that the text of 3 Kingdoms shows influence of the Antiochene text and Davies is of the same opinion regarding 4 Kingdoms.56 The Aeth agreements with L are more frequent in the kaige sections, which supports the pre-Lucianic/OG character of the Antiochene text against the B kaige text. Taking as an example 4 Kgdms 4 these agreements correspond to Greek variants of various kinds. They attest an OG reading in 4:3 δεῦρο] Abi Arm Aeth SyrHmg (πορεύου); 4:7 δεῦρο] πορεύου L Arm Aeth; 4:29 δεῦρο] πορεύου L 460 Arm Aeth Lat Spec. In 4:28 the omission of ὅτι in Aeth and L Arm represents the OG as accepted by Rahlfs in his edition. Aeth agrees with an Hexaplaric reading of L in 4:2 εἶπεν 1º] + αὐτῇ L Aeth (uid); + πρὸς αὐτήν 247 121 488 z 372 SyrH (sub * α′ θ′ ε′); ei Arm | 4:8 φαγεῖν 2º] + ἄρτον 247 L−82 x 318 342 554 Aeth SyrHmg (pr *) | 4:26 ἐρεῖς] + αὐτῇ L 460 Aeth (uid) SyrH (pr *). Aeth seems not to follow the haplographic reading of its usual model, the Codex Vaticanus, in 4:3 γειτόνων B] + σου L rel Arm Aeth (uid) SyrH. Other agreements simply highlight a certain affinity between the Antiochene and the Alexandrian texts and seem to point to the fact that the LXX text that was widespread in Syria and Egypt in pre-Origen times was relatively homogeneous: 4:2 οικω] + σου L Aeth (uid) | 4:5 αυτοι] pr. και L Aeth | 4:19 παιδαριω] + αυτου L Aeth (uid) | 4:23 και ειπεν] + αυτη 318; + αυτη ο L 460 700; ei Aeth | 4:23 σημερον προς αυτον 93-127 Aeth Thdt | 4:24 μη] pr και L 460 Aeth Arm | 4:26 παιδαριῳ] + σου L 460 44 242 Aeth (uid) Lat | 4:27 ηγγισεν] προσηλθε L 460 Aeth (uid) | 4:27 η ψυχη αυτης κατωδυνος αυτη] κατωδυνος η ψυχη αυτης L 460 Aeth Chr Thdt | 4:29 οτι] καὶ 19 Arm Aeth SyrH Spec | 4:36 εξεβοησεν (hapax)] εκαλεσεν L 460 Aeth | 4:36 προς 1º] τον L−108 460 Aeth | 4:39 εγνωσαν] ηδει L−82 460 Aeth Arm | 4:41 ειπεν 1º] + Ελισαιε L 460 Aeth | 4:41 το παιδαριον] > L Aeth | 4:41 εκει] > L−82 460 Aeth Lat | 4:42 ανηρ διηλθεν] ηλθεν ανηρ L−82 460 Aeth Arm.

54 Moreno, Las glosas marginales de Vetus Latina, 179. 55  Kauhanen notes eighteen instances in which “it is possible that the hypothetical OG text witnessed by Lucifer may occasionally have preserved the reading of the Hebrew Vorlage of the LXX against most or all of the other Greek witnesses,” Kauhanen, Lucifer of Cagliari, 316–318. 56  H.S. Gehman, “The Old Ethiopic Version of 1. Kings and its Affinities,” JBL 50 (1931): 81–114.

70

Pablo A. Torijano/Julio Trebolle

The actual text of the Armenian version supports in most cases the Hexaplaric tradition, but this stage of the text results from the revision of a previous translation made on Greek manuscripts that shared many readings with the Antiochene pre-Lucianic tradition. The not so rare agreements between the secondary versions Aeth, Cop, Arm and OL increase the critical value of their readings.57 We have previously found in 3 Kgdms 12:24a–z a large number of agreements of L Aeth and the marginal readings of the Syrohexaplaric text (SyrHmg). The kaige sections contain many more of such agreements, which Rahlfs considers to be “Lucianic,”58 but in fact witness the pre-Lucianic/OG text, reflected also in the OL, Aeth and Arm versions:59 4 Kgdms 3:4 SyrHmg ην φερων φορον και επιστρεφον τω βασιλει Ισραηλ reproduces the doublet formed by the pre-Lucianic/OG reading ην φερων φορον reflected by OL ferebat tributum and the kaige B reading επιστρεφον; 6:8 SyrHmg εις τον τοπον του φελμουνι ποιησωμεν ενεδρον is the pre-Lucianic/OG reading reflected in OL in locum phalmunum obsessionem faciamus; 7:2 SyrHmg και εαν κυριος ποιηση, pre-Lucianic/OG reading (‫)? ואם‬, attested by the Armenian and Ethiopic versions against the B kaige reading ιδου (‫ הנה‬MT); 23:33 SyrHmg και δεκα ταλαντα (‫)ܠ‬. Rahlfs recognized the early character of the reading δεκα, witnessed by the Armenian text and the Peshitta, against B εκατoν (‫ מאה‬MT).

The Targum, Peshitta and Vulgate agree in not a few cases with OG readings, especially with the pre-Lucianic ones in the kaige sections, as for example in60 4 Kingdoms 3:27 γῆν αὐτῶν L, ‫ ארצם‬Mss, իւրեանց Arm, ‫ܐܠܪܥܗܘܢ‬/‫ ܐܠܬܪܗܘܢ‬S, terram suam V] γην rel (‫ ָל ָא ֶרץ‬MT) | 7:6 πλησίον L, ‫ רעהו‬nonn Mss, ‫ חברה‬Tfmss, ‫ ܠܚܒܪܗ‬S] αδελφον (‫ ָא ִחיו‬MT) rel | 14:13 ἀπὸ τῆς πύλης L 700, ‫ שער‬pc Mss, ի դրանէ Arm, ‫ מתרע‬T, ‫ ܡܢ ܬܪܥܐ‬S, a porta V, ‫ משער‬2 Chr] εν τη πυλη rel (‫ ְּב ַׁש ַער‬MT) | 23:6 57  Cf. A. Piquer, “An Old Greek Reading Attested in the Sahidic and Old Latin Fragments of 1Kgs 1:52. Text-critical Analysis and Relationship with the Hebrew Text,” Hénoch 30/1 (2008): 80–93; A. Suciu and F. Albrecht, “Remarks on a Coptic Sahidic Fragment of 3 Kingdoms, Previously Described as an Apocryphon of Solomon,” JBL 136 (2017): 57–62. 58 Rahlfs, Lucians Rezension, 30–34. 59 Cf. Julio Trebolle, “Pre-Lucianic Readings of 3–4 Reigns in Marginal Notes of the Syro-hexaplaric Version and in the Syriac Text of Jacob of Edessa,” in From Scribal Error to Rewriting. How Ancient Texts Could and Could not be changed, eds. A. Aejmelaeus and D. Longacre (Göttingen: Vandenhoeck & Ruprecht, 2020), 73–98. 60  See J. Trebolle and P. Torijano, “The Behavior of the Hebrew Medieval Manuscripts and the Vulgate, Aramaic and Syriac Versions of 1–2 Kings vis-à-vis the Masoretic Text and the Greek Version in 1–2 Kings,” in The Text of the Hebrew Bible: From the Rabbis to the Masoretes, ed. E. Martín-Contreras and L. Miralles-Maciá (Göttingen: Vandenhoeck & Ruprecht, 2014), 101–136.

The Edition of III–IV Kingdoms

71

̈ τοὺς τάφους L 460, ‫ קברי‬Ms, ‫ לקברי‬T, ‫ܩܒܪܐ‬ S] τον ταφον rel (‫ ֶק ֶבר‬MT) | 23:17 τὸ ἐν βαιθελ L, ‫ בבית אל‬pc Mss, T, ‫ ܒܒܝܬ ܐܝܠ‬S] βαιθελ rel (‫ית־אל‬ ֵ ‫ ֵּב‬MT) | 24:12 πρὸς (τὸν L x 460) βασιλέα) 247 L o x 55 342 460, ‫ אל מלך‬pc Mss, ար արքայ Arm, Aeth, ‫ אל מלכא‬TMs, ‫ ܠܘܬ ܡܠܟܐ‬S, ad V] ἐπὶ (βασιλέα) rel (‫ל־מ ֶלְך‬ ֶ ‫ ַע‬MT).

2.3  Translation Features and Readings of the Kaige and the Pre-Lucianic / OG texts. Rahlfs classified the translation ἕκαστος (MT ‫)איׁש‬ ִ of the Lucianic manuscripts among the “corrections of expression” made by the Lucianic recensor. In the same way, among the “Lucianic” synonym changes (“Umtauschung von Synonymen”) Rahlfs included those of θυσιάζειν by θύειν, ἄρχων τῆς δυνάμεως by ἀρχιστράτηγος and κερατίνη by σάλπιγξ.61 Today, it is well known that ἀνήρ, θυσιάζειν, ἄρχων τῆς δυνάμεως and κερατίνη are kaige characteristics, while ἕκαστος, θύειν, ἀρχιστράτηγος and σάλπιγξ are OG readings preserved by the pre-Lucianic text in the kaige sections. The necessary homogeneity of a critical text leads to introduce in the lemma of the critical edition the pre-Lucianic/OG reading even when it was replaced by the corresponding kaige version and therefore no actual testimony of it has been preserved, as for example in 4 Kgdms 3:25 ἔρριψαν ἕκαστος (ἀνήρ B) τὸν λίθον and 7.6 εἶπεν ἕκαστος πρὸς τὸν πλησίον (L, ἀδελφὸν B). Similarly, the conjectures ἀρχιστράτηγον instead of kaige ἄρχοντα τῆς δυνάμεως (B, 4 Kgdms 22:19) and OG τέκνον instead of kaige υἱός (B, 4 Kgdms 4:4.5; 14:6; 17:31; 21:6) are to be introduced in the lemma. The textual transmission of onomastics and toponymy leads to the multiplication of textual variants, as shown particularly in the lists of governors and districts of 3 Kingdoms 4. But the most significant variants of personal and place names go back to the previous editorial process and make a significant difference between the two editions of LXX and MT.62 The Antiochene text preserves OG Ορνια (‫)ארניה‬, reflected in OL (2:24), in part of the Armenian tradition and clearly in the Georgian version (Ornia/Ornea MS O), against kaige Αδωνιας (MT ‫)אד ֹנִ ּיָ הּו‬. ֲ 63 In non-kaige section both B and L preserve OG Σαδδουκ (2 Kgdms 8:17 and 3 Kgdms 4:4). L has Σαδδουκ also in 2:35; 2:46h and 4:2, to be restored in the edition against B Σαδωκ. In the kaige section βγ L preserves OG Χορρ(ε)ι (3 Kgdms 1:38.44; 4 Kgdms 11:4.19), reflected in OL Chori / Corri (3 Kgdms 1:38), in agreement with the Ketib ‫ ְּכ ִרי‬of 2 Sam 20:23 and against the

61 Rahlfs, Lucians Rezension, 179 and 181. 62  Piquer, Torijano, Trebolle, “Versions of the Septuagint,” 251–282. 63  3 Kgdms 1:8. 9. 11. 13. 24. 41. 42. 43. 50. 51; 2:3. 17 (a plus in L). 19. 21. 22. 23. 24. 25 (> MT). 41. 42. 43. 46h. 49. 50. 51. 53 (plus in L).

72

Pablo A. Torijano/Julio Trebolle

kaige B Χερεθι that follows MT ‫ה ְּכ ֵר ִתי‬.ַ Ιωαδ (L) is the old form of the name, reflected in the Georgian, against Ιωιαδαε (MT ‫)יְ הֹויָ ָדע‬.64 According to Rahlfs the Lucianic recensor replaced the transcriptions present mainly in the kaige sections in the B text with the corresponding translations.65 As stated by E. Tov the transliteration of Hebrew Words in the Greek is a further characteristic of the kaige-Th Revision,66 but it has to be added that the pre/Lucianic Antiochene text has preserved the OG translation, reflected sometimes by OL, Aeth, Arm, Georg or Josephus. Rahlfs attributes the change of μαναα (‫ )מנחה‬in δῶρον/δῶρα to the desire of stylistic variation on the part of the Lucianic recensor,67 but the Antiochene δῶρον/δῶρα is the OG translation attested by Josephus in 4 Kgdms 8:8, reflected by the OL munera in 17:3–4 and by the Armenian version in 20:12. The original translation was replaced in the kaige recension with the transliteration μαναα.68 Other cases of kaige transcriptions that replaced the OG translation are: 4 Kgdms 11:12 νεζερ (‫)נזר‬ – ἀγίασμα L (σ’ ἅγιον); 12:10 ιαμιβιν – ἐν δεξιᾷ L; 14:25 τῆς Αραβα (‫)הערבה‬ – τῆς πρὸς ἑσπέραν L; 23:4 ἐν σαδημωθ – ἐν τῷ ἐμπυρισμῷ L; 23:5 χωμαριμ (‫)כמרים‬ – ἱερεῖς L (OL); 23:7 χεττιιν – στολάς L (OL Arm Georg); 25:4 τὴν Αραβα (‫)ערבה‬ – τὴν ἐπὶ δυσμάς L; 25:5 ἐν Αραβωθ (‫)בערבות‬ – κατὰ δυσμάς L; 25:12 γαβιν (‫)גבים‬ – γεωργούς L (Aeth Arm Georg); 25:13.16 μεχωνωθ (‫)מכנות‬ – βάσεις L; 25:14 ιαμιν (‫)היעים‬ – κρεάγρας L; 25:17 χωθαρ (‫)כתרת‬ – ἐπίθεμα L (3 Kgdms 7:16–18); 25:17 σαβαα (‫)שבכה‬ – δίκτιον L (3 Kgdms 7:17.41–42).

2.4  Doublets and Alternative Readings of the Antiochene text

The typical Lucianic doublets appear mostly in the καίγε sections.69 According to A. Rahlfs they derive from the corresponding readings in the B majority text, but in fact they are made of the OG reading preserved by the pre-Lucianic text plus the kaige reading juxtaposed to the old one. The critical edition incorporates in the lemma the OG and marks in bold type the kaige reading in the apparatus. In the list that follows the pre/Lucianic/OG precedes the juxtaposed kaige reading, although L may contain these readings in reverse order:

64  Ιωαδ is found in kaige section βγ in 2 Kgdms 23:20.22; 3 Kgdms 1:8.26.32.36.38.44 and in non-kaige sections in 2 Kgdms 8:18 (ββ); 3 Kgdms 2:25.29.34.35.46; 4:4 (γγ). 65 Rahlfs, Lucians Rezension, 182, 208 and 248–250. 66  E. Tov, “Transliterations of Hebrew Words in the Greek Versions of the Old Testament: A Further Characteristic of the kaige-Th Revision?” Textus 8 (1973): 78–92. 67 Rahlfs, Lucians Rezension, 140. 68  P. Torijano, “Textual Criticism and the Text-Critical Edition of IV Regnorum: The Case of 17:2–6,” in After Qumran: Old and Modern Editions of the Biblical Texts – The Historical Books, ed. H. Ausloos et alii (Leuven: Peeters, 2012), 195–212, esp. 204–205. 69  Only 9 doublets belong to the non-kaige section (γγ): 3 Kingdoms 2:26; 7:14; 8:33; 8:66; 10:15; 11:17; 15:15; 16:11; 18:21.

The Edition of III–IV Kingdoms

73

2:23 καὶ ἐλίθαζον αὐτόν (‫ )ויסקלו‬+ καὶ κατέπαιζον αὐτοῦ (‫ | )ויתקלסו‬3:20 ἐξ ὁδοῦ τῆς ἐρήμου Σοὺρ + ἐξ ὁδοῦ Εδωμ (‫)מדרך אדום‬, de via eremi Sur (‫| )מדרך מדבר צור‬ 4:4 καὶ τὸ πληρωθὲν ἀρεῖς (‫ )והמלא תסיעי‬+ καὶ αὐτὸ οὐκ ἀποστήσεται (‫)ולא יעמד‬ | 4:31 καὶ ἀπήγγειλεν αὐτῷ (‫ )ויגד לו‬+ καὶ εἶπεν αὐτῷ (‫ | )ויאמר לו‬4:37 ἐπὶ τὰ γόνατα αὐτῆς (‫ )ברכיה‬+ πρὸς τοὺς πόδας αὐτοῦ (‫ | )רגליו‬6:30 ἐπὶ τῆς ὀσφύος αὐτοῦ (‫)מתניו‬ + ἐπὶ τῆς σαρκὸς αὐτοῦ (‫ | )בשרו‬7:5 ἤδη διαυγάζοντος + ἐν τῷ σκότει | 7:10 καὶ ἐκαλεσαν τοὺς στρατηγούς (‫ )שרי‬+ καὶ ἐβόησαν πρὸς τὴν πύλην (‫ | )שער‬8:1 καὶ παρέσται ἐπὶ τὴν γῆν ἑπτὰ ἔτη + καί γε ἦλθεν ἐπὶ τὴν γῆν ἑπτὰ ἔτη | 9:37 καὶ οὐκ ἔσται ὁ λέγων Οἴμμοι L, et non est qui dicat Vae mihi (‫ )ולא יהיה האמר אהה‬+ ὥστε μὴ εἰπεῖν Αὕτη Ιεζαβελ (‫ | )אשר לא יאמרו זאת איזבל‬11:4 καὶ τὸν Χορρεὶ καὶ τὸν ‘Ρασείμ + τοὺς ἑκατοντάρχους τῶν παρατρεχόντων | 11:14 καὶ οἱ στρατηγοί + καὶ οἱ ᾠδοὶ καὶ αἱ σάλπιγγες | 18:17 ἐν τῇ ἀναβάσει + ἐν τῷ ὑδραγωγῷ | 21:23 καὶ ἐπεβούλευσαν αὐτῷ + καὶ συνεστράφησαν … ἐπ᾽ αὐτόν | 23:16 καὶ ἀπέστρεψεν (‫ )וישב‬+ καὶ ἐξένευσεν (‫ | )וַ ּיִ ֶפן‬24:11 καὶ αὐτὸς ἐπεκάθητο ἐπὶ τὴν πόλιν + καὶ οἱ παῖδες αὐτοῦ ἐπολιόρκουν (ἐπ᾽) αὐτήν | 25:19 Σαφαν τὸν ἀρχιστράτηγον (‫ )ספן‬+ τὸν γραμματέα (‫ )הספר‬τοῦ ἄρχοντος τῆς δυνάμεως.

The larger and more significant duplicate of LXX is found in 3 Kgdms 16:28a–h (OG) // 22:41–51 (kaige).70 According to Tov, “the rewritten text of 3 Kingdoms repeated 1 Kgs 22:41–51 (description of Jehoshaphat’s activities) in 3 Kingdoms 16:28a–h.”71 However the Hebrew Vorlage of LXX knew only the text translated in 16:28a–h, just as MT only knows the text located in 22:41–51, causing in this way a duplicate between two sections of the Greek text (γγ and γδ) that correspond to two different scrolls of the Hebrew text. Rahlfs kept in his edition both passages, but the original Greek translation could not contain such a duplicate nor can a critical edition. The following list picks up significant cases where our edition follows the pre-Lucianic/OG reading and reproduces in the critical apparatus the kaige reading: 3 Kgdms 22:10 ἐν ὁδῷ πύλης (ad viam portae OL 91–95) / ἐν ταῖς πύλαις (‫ֶּפ ַתח‬ ‫ ַׁש ַער‬MT) | 4 Kgdms 3:23 αἷμα τοῦτο ἐρίσαντες γὰρ ἤρισαν οἱ τρεῖς βασιλεῖς (καὶ ἐπάταξαν ἕκαστος) / αἷμα τοῦτο ῥομφαίας καὶ ἐμαχέσαντο οἱ βασιλεῖς (καὶ ἐπάταξαν ἀνήρ) (‫ | ) ָּדם זֶ ה ָה ֳח ֵרב נֶ ֶח ְרבּו ַה ְּמ ָל ִכים‬4:20 καὶ ἐκάθισεν αὐτόν (deposuit eum Aeth) ἐπὶ τὰ γόνατα αὐτῆς / καὶ ἐκοιμήθη ἐπὶ τῶν γονάτων αὐτῆς (‫וַ ּיֵ ֶׁשב ַעל־‬ ‫יה‬ ָ ‫ | ) ִּב ְר ֶּכ‬6:10 οὐχ ἅπαξ οὐδὲ δίς / οὐ μίαν οὐδὲ δύο (‫ ;) ַא ַחת וְ לֹא ְׁש ָּתיִ ם‬6:30 εἱσθήκει 70  J.D. Shenkel, Chronology and Recensional Development in the Greek Texts of Kings (Cambridge MA: Harvard University Press, 1968), 42–60. 71  E. Tov, “The Many Forms of Hebrew Scripture: Reflections in Light of the LXX and 4Q Reworked Pentateuch,” in From Qumran to Aleppo: a discussion with Emmanuel Tov about textual history of Jewish scriptures in honor of his 65th birthday, ed. A. Lange (Göttingen: Vandenhoeck & Ruprecht, 2009) 11–28, esp. 16.

74

Pablo A. Torijano/Julio Trebolle

(‫)עמד‬ / διεπορεύετο (‫ | )ע ֵֹבר‬8:12 ὅσα ποιήσεις τοῖς υἱοῖς Ισραηλ σὺ τὰς πόλεις αὐτῶν τάς ἐστερεωμένας ἐμπρήσεις ἐν πυρί / ὅσα ποιήσεις τοῖς υἱοῖς Ισραηλ κακά τὰ ὀχυρώματα αὐτῶν ἐξαποστελεῖς ἐν πυρί (‫ר־ּת ֲע ֶׂשה ִל ְבנֵ י יִ ְׂש ָר ֵאל‬ ַ ‫ֵאת ֲא ֶׁש‬ ‫יהם ְּת ַׁש ַּלח ָּב ֵאׁש‬ ֶ ‫ | ) ָר ָעה ִמ ְב ְצ ֵר‬8:22 τοῦ μὴ δοῦναι χειρά (‫ )מתת יד‬/ ὑποκάτωθεν χειρός (‫) ִמ ַּת ַחת יַ ד‬ | 9:32 τοὺς ὀφθαλμοὺς αὐτοῦ (‫ )עניו‬/ τὸ πρόσωπον αὐτοῦ (‫ | ) ָפנָ יו‬10:11 καὶ πάντας τοὺς ἀγχιστεύοντες αὐτοῦ (‫ )גאליו‬/ καὶ τοὺς ἁδροὺς αὐτοῦ (‫ | )ּגְ ד ָֹליו‬10:19 μὴ ἀπολειφθῆτο (‫ )שאר‬/ μὴ ἐπισκεπήτω (‫יִּפ ֵקד‬ ָ ) | 17:2 παρὰ πάντας τοὺς γενομένους (OL prae omnes qui fuerunt) (‫ )מכל אשר היו‬/ πλὴν οὐχ ὡς βασιλεῖς Ισραηλ οἳ ἦσαν (‫ | ) ַרק לֹא ְּכ ַמ ְל ֵכי יִ ְׂש ָר ֵאל ֲא ֶׁשר ָהיּו‬23:4 τοῖς ἱερεῦσι τοῖς δευτερεύουσι / τοῖς ἱερεῦσιν τῆς δευτερώσεως (‫ | )ּכ ֲֹהנֵ י ַה ִּמ ְׁשנֶ ה‬25:18 τὸν ἱερέα τὸν δεύτερον / υἱὸν τῆς δευτερώσεως (‫)ּכ ֵֹהן ִמ ְׁשנֶ ה‬.

2.5  The different arrangement of the text in the OG and in the kaige and Hexaplaric recensions LXX and MT present a different order of literary units in 3 Kgdms 2–14 in the non-kaige section. Some authors consider that LXX represents an edition anterior rather than subsequent to MT;72 others attribute the change in order to the Greek translator,73 or to the editor of the Hebrew original who reorganized the previous proto-MT text.74 In the kaige section of 3 Kgdms 22–4 Kingdoms the OG represented by L presents also considerable transpositions versus MT. The section on Jehoshaphat’s reign was originally “in the right place” attested by LXXBL in 3 Kgdms 16:28a–h. LXXB presents a double reading: the OG of 3 Kgdms 16:28a–h and the parallel kaige text of 22:41–46.51. The section of 22:47–50 is an Hexaplaric addition. Rahlfs’ LXX edition contains the OG text of 3 Kgdms 16:28a–h as well as the parallel kaige text of 3 Kgdms 22:41–51, omitted by the Antiochene text which represents the OG here. A critical edition of 3 Kingdoms should assign the B kaige text to the critical apparatus. However, given the 72  J. Trebolle, Salomón y Jeroboán. Historia de la recensión y redacción de 1 Rey 2–12; 14 (Salamanca: Universidad Pontificia, 1980), esp. 278 and 321; J. Trebolle, “Redaction, Recension, and Midrash in the Books of Kings,” BIOSCS 15 (1982): 12–35; A. Schenker, Septante et texte Masssorétique dans l’histoire la plus ancienne du texte de 1 Rois 2–14 (Paris: Gabalda, 2000); Ph. Hugo, Les deux visages d’Élie: Texte massorétique et Septante dans l’histoire la plus ancienne du texte de 1 Rois 17–18 (Fribourg: Academic Press – Göttingen: Vandenhoeck & Ruprecht, 2006). 73  D.W. Gooding, Relics of Ancient Exegesis: A Study of the Miscellanies in 3 Reigns 2 (Cambridge: Cambridge University Press, 1976); P.S. Van Keulen, Two Versions of the Solomon Narrative: An Inquiry into the Relationship between MT 1 Kgs 2–11 and LXX 3 Reg 2:11 (Leiden: Brill, 2005). 74 Talshir, The Alternative Story. Polak attributes the differences to a revision in G as well as (in parts) to its differing Hebrew Vorlage, E.H. Polak, “The Septuagint Account of Solomon’s Reign: Revision and Ancient Recension,” in X Congress of the International Organization for Septuagint and Cognate Studies, Oslo 1998, ed. B.A. Taylor (Atlanta: Scholars Press, 2001), 139–164.

The Edition of III–IV Kingdoms

75

t­ ransmission conditions of the LXX text – OG in some sections or passages and kaige in others in which it may be the oldest recoverable text – it is surely advisable to include also the text of 3 Kgdms 22:41–51 in the lemma, but indicating with some graphic sign that it is in fact a kaige text. The edition of 3–4 Kingdoms will reflect the history of the text by reproducing the OG of 3 Kgdms 16:28a–h, the text of 22:41–46.51 with some graphic indication that identifies it as kaige and the Hexaplaric addition of 22:47–50 in the apparatus as attested by A Arm SyrH (sub * α′ vid). Textual variants usually accumulate at the points where a shift or insert has occurred. The Greek textual variants reflect changes produced in the editing process of the Hebrew text. Thus, additions to the Hebrew text operated via resumptive repetition bring about satellite variants at the insertion points. For example, the hexaplaric addition in 3 Kgdms 16:29 ἐν ἔτει τριακοστῷ καὶ ὀγδόῳ τοῦ Aσα βασιλέως Iουδα βασιλεύσας δὲ Aχααβ υἱὸς Zαμβρι (A Arm SyrH*) does not include the opening sentence of this verse MT ‫ן־ע ְמ ִרי ָמ ַלְך ַעל־יִ ְׂש ָר ֵאל‬ ָ ‫וְ ַא ְח ָאב ֶּב‬, although it is absent in the OG. But the whole addition is transmitted by other Hexaplaric testimonies after 16:28: ὁ δὲ Aχααβ υἱὸς Zαμβρι ἐβασίλευσεν ἐπὶ Ισραηλ ἐν ἔτει τριακοστῷ καὶ ὀγδόῳ τοῦ Aσα βασιλέως Iουδα βασιλεύει δε Aχααβ υἱὸς Zαμβρι ἐπὶ Iσραηλ (A V 247 121 488 z 71).75 The readings of A Arm SyrH and 247 121 488 z 71 frame the unit of 16:28a–h as a type of Wiederaufnahme. In the same way at 3 Kgdms 16:8 the Hexaplaric reading ἐν ἔτει εἰκοστῷ καὶ ἐκτῷ ἐπὶ τοῦ Aσα βασιλέως Iουδα (A 71 Arm SyrH sub * γ′) supplies the supposed omission of OG. However, the corresponding OG text is found earlier in 16:6: ἐν τῷ εἰκοστῷ ἔτει / ἔτει εἰκοστῷ / ἐν ἔτει εἰκοστῷ καὶ ὀγδόῳ βασιλέως Ασα / τοῦ Aσα βασιλέως Iουδα. The two readings frame like a resumptive repetition the verse 16:7 which is considered an addition in the Hebrew.76 Also, the Hexaplaric addition at the end of 3 Kgdms 7:38 καὶ συνετέλεσεν Σαλωμων ὅλον τὸν οἶκον αὐτοῦ (A Arm SyrH sub *) is no other than the OG of 7:50 καὶ συνετέλεσεν Σαλωμων ὅλον τὸν οἶκον αὐτοῦ. The two readings frame the literary unit on the construction of the royal palace found in different locations in MT and LXX. The “supplements” of LXX (3 Kgdms 2:35a–o; 36–46; 46a–l) appear framed by the sentence καὶ Σαλωμων υἱὸς Δαυιδ ἐβασίλευσεν ἐπὶ Ισραηλ καὶ Ιουδα ἐν Ιερουσαλημ attested by the manuscripts: M V 121 o x z 58 342 in 2: 35a and by the whole manuscript tradition but 245 in 2:46l.

75  The historical present βασιλεύει (247 121) is an OG residue against the Hexaplaric βασιλεύσας (A V). 76  E. Würthwein, Die Bücher der Könige 1. Könige 1–16 (ATD 11,1; Göttingen: Vandenhoeck & Ruprecht, 1977), 192.

76

Pablo A. Torijano/Julio Trebolle

The Hexaplaric reading καὶ ἰάσατο τὸ θυσιαστήριον (κυρίου = MT) τὸ κατεσκαμμένον that appears before 3 Kgdms 18:30–32 is taken from the OG that is found after this clause considered a gloss of the Hebrew. In 4 Kgdms 25:23 the B kaige reading καὶ ἦλθον πρὸς Γοδολιαν εἰς Μασσηφαθ precedes the list of those who came with their men to Gedaliah at Mizpah. The OG transmitted by L has the same reading after the list. This repetition functions as a sort of Wiederaufnahme which may betray the previous insertion of this list into the Hebrew text. The above cases show how the Hexaplaric text reproduces the OG albeit at the MT site. Therefore, it becomes one further testimony of OG so that the critical apparatus should record it. On the other hand, it is necessary to clarify in each case which is the original location of the reading, before or after the interposed text. The OG represented by L differs considerably from MT in 4 Kgdms 13. The OG places 13:23 after 13:7 as part of the Dtr material of 13:2–7.23. The OG presents the conclusive formula of Joash’ reign (2 Kgs 13:12–13) after the notice of 13:22.24–25, “in the right place” as recognized by Montgomery.77 In MT the initial formula of the reign of Joash (13:10–11) and the conclusive one (13:12–13) follow each other, leaving the narrative about Elisha’s death and the notice on the Aramean war outside the frame of the reign of Joash. In this way the textual layout in MT breaks a composition rule of Kings according to which every literary unit must be framed by the initial and final formulas of the corresponding reign.78 Furthermore, MT repeats the conclusive formula of 13:12–13 later on in 14:15–16 (LXX ignores v. 15). L (and Josephus) attests the OG which presents the conclusive formulae of Joash after 13:25 according to the aforementioned composition rule of the books of Kings.79 Besides, the narratives about Elisha’s death and burial, which in MT/LXXB appear in 13:14–21, are found in OL (Vindobonensis) in chapter 10, between v. 30 “And the Lord said unto Jehu …” and v. 31 “But Jehu took no heed to walk in the law of the Lord …”. The frequent transpositions between the OG and kaige variants or between the OG and Hexaplaric additions, as well as the incorporation of OL in the edition of 3–4 Kingdoms, pose difficulties not foreseen in the edition of other LXX books, in particular as regards the construction of the critical apparatus. The edition will reproduce the OG according to the text and the text arrangement of L and the OL. On the other hand, it incorporates the kaige text with some graphic sign that identifies it as a kaige text that reflects the arrangement of the MT. In 77 Montgomery, Kings, 434. 78  S.R. Driver, An Introduction to the Literature of the Old Testament (Edinburgh: T&T Clark, 19139), 179. 79  M. Richelle, Le testament d’Elisée: Texte massorétique et Septante en 2 R 13,10–14,16 (Paris: Gabalda, 2010).

The Edition of III–IV Kingdoms

77

this way the reader can recognize kaige readings, which would otherwise be lost among other variants in the critical apparatus. The inclusion of OL in the edited text implies giving up an established principle of the Göttingen edition, as Hanhart states in relation to Paralipomena.80 The ultimate aim of the edition is not the reconstruction of the “original text,” unrecoverable in all its smallest details, but of the “oldest recoverable text.”81 On some occasions, this text may be not a Greek text but a fossilized translation of an old Greek such as the OL. For this reason, we propose to include in the lemma of the edition the OL of 10:30a–i in its own location, enclosed in square brackets to indicate that it is the translation of a Greek text older than the oldest recoverable Greek text. On the other hand, the OG text of 13:14–21 will appear in its own location, giving preference to pre-Lucianic readings over those of the kaige B text highlighted in bold. The layout of the edition would be as follows: 4 Kgdms OL 10:30 …υἱοὶ τέταρτοι καθήσονταί σοι ἐπὶ θρόνου Ισραηλ. [OL Et mortus est helisseus et sepelierunt eum et piratae moab uenerunt in terram illam. Et factum est cum sepeliirent piratae hominem unum accesserunt ad monumentum et proiecerunt hominem in monumentum helissei et fugerunt et adplicitus est homo ossibus helissei et uixit homo et surrexit super pedes suos.] 10:31 καὶ Ιου οὐκ ἐφύλαξεν πορεύεσθαι ἐν νόμῳ κυρίου θεοῦ Ισραηλ ἐν ὅλῃ καρδίᾳ αὐτοῦ …

4 Kgdms 13:20–21 20καὶ

ἀπέθανεν Ελισαιε καὶ θάπτουσιν αὐτόν καὶ πειράται Μωαβ ἦλθον ἐν τῇ γῇ ἐλθόντος τοῦ ἐνιαυτοῦ. 21καὶ ἐγένετο αὐτῶν θαπτόντων ἄνθρωπον ἕνα καὶ ἤγγισε τὸ πειρατήριον αὐτοῖς καὶ ἔρριψαν τὸν ἄνδρα ἐν τῷ τάφῳ Ελισαιε και ἔφυγον καὶ ἥψατο τῶν ὀστέων Ελισαιε καὶ ἔζησε καὶ ἀνέστη ἐπὶ τοὺς πόδας αὐτοῦ. B M V O L CI CII b d f o s t x z al [>] Aeth Arm Sa Lat115 SyrH 20 ἀπέθανεν] post Ελισαιε tr. L 460 700 | Ελισαιε Bc M 127 64 130 55 158 372 Ra] ελεισαιε B*; ελισσεαι 93; ελισεαι 108; ελισσαιος 700 342 460; ελισσαιε rel; Helisseus 80  “… für die Relativierung eines Gesetzes, das in der Göttinger Edition grundsätzlich eingehalten werden muss: für den Verzicht auf die Aufnahme in den Sekundärübersetzungen überlieferter Textformen, die griechisch nicht mitbezeugt sind,” Hanhart, Paralipomenon liber II, 4. 81  “Statt dem ursprünglichen Text definierte Rahlfs als das Editionsziel seiner Septuaginta-Arbeit somit die (methodisch begründbare) Rekonstruktion des ursprünglichsten Texts, d.h. des ältesten erreichbaren Textes resp. Archetypus der Septuaginta, i.e. des dem ‘Urtext’ “möglichst nahekommenden [ihm aber nicht entsprechenden; CS] Texte;” cf. Ch. Schäfer, Alfred Rahlfs (1865–1935) und die kritische Edition der Septuaginta. Eine biographisch-wissenschaftsgeschichtliche Studie (Berlin: De Gruyter, 2016), 332.

78

Pablo A. Torijano/Julio Trebolle

Lat115 (post 4 Reg 10:30) | θάπτουσιν L 460 700] εθαψαν rel | πειραταί scripsi] μονοζωνοι rel; cf. 4 Reg 10: 30 piratae La115 (post 4 Reg 10:309 | Μωαβ] μοαβ 313* | ἐλθόντος] -τες 247 488* | om. τοῦ 52* f 21 om ἐγένετο αὐτῶν 44 | αὐτῶν θαπτόντων] Cf cum sepelierunt piratae Lat115 (post 4 Reg 10: 30) | ἄνθρωπον ἕνα L 460 700 ] > o; τον ανδρα rel; cf. hominem unum Lat115 (post 4 Reg 10:30) | καὶ 2º – ἄνδρα 2] post Ελισαιε tr. 246; > 74 | ἤγγισε τὸ πειρατήριον αὐτοῖς L 460 700] ιδου ειδον τον μονοζωνον rel; cf. acceserunt ad monumentum La115 (post 4 Reg 10:30) | ἔρριψαν] εκρυψαν 799 | τὸν ἄνδρα 2] αυτου 71 | om τῷ 19 | Ελισαιε Bc V 127 64 55 158 372 Ra] ελισσεαι 93 | ελισεαι 108; ελεισαιε B*; ελισσαιου 342 700; ελισσεαι 93; > 460; ελισσαιε rel | om καί 4º – Ελισαιε 2 52* | ἔφυγον eds.] εφυγων 158; εφυγον και ηλθε L 460 700 (cf. fugerunt Lat115 post 4 Reg 10:30); επορευθησαν 71 342 Cop SyrH; επορευθη (+ ο ανος o 372) rel | ἥψατο] + ο ανηρ ο θαπτομενος L 460 700; + ο ανος 55 245 342 554 Cop; + is vir SyrH ⸕; cf. + homo Lat115 (post 4 Reg 10:30)‘ | Ελισαιε Bc V 127 64 55 158 372 Ra] ελισσεαι 93 | ελισεαι 108; ελεισαιε B*; ελισσαιου 342 460 700; ελισσαιε rel | ἔζησεν καί ἔστη] εστη και ανεζησε 125 | ἔζησε] + homo Aeth; + homo Lat115 (post 4 Reg 10:30) | ἀνέστη] εστη L 460 700; + et stetit Aeth | ἐπί ] υπο 82 d–106

The difficulties that the edition of 3–4 Kingdoms presents underline the complexity of the textual history of the LXX, of its recensions and secondary versions, which is but a reflection of the no less complex history of the Hebrew text of 1–2 Kings. M. West links the problems raised by the Homer and LXX editions pointing out “the importance of always keeping the whole history of the transmission in mind, in so far as it is known or can be plausibly hypothesized: a problem that manifests itself in a late phase of the tradition may sometimes find its solution in an earlier phase.”82 The readings attested by late witnesses such as the secondary versions or the duplicates and transpositions attested differently in some witnesses or in others may sometimes find their origin and their solution in earlier phases of the textual history known or plausibly hypothesized of the LXX and of the Hebrew text.

3. Bibliography Barthélemy, Dominique, Les devanciers d’Aquila. VTSup 10 (Leiden: Brill, 1963). Hanhart, Robert, ed. Septuaginta: Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum. Vol. 7.2, Paralipomenon liber II (Göttingen: Vandenhoeck & Ruprecht, 2014). Haupert, Raymond Samuel, The Relation of Codex Vaticanus and the Lucianic Text in the Books of the Kings from the Viewpoint of the Old Latin and the Ethiopic Versions (Leipzig: Drugulin, 1930). 82 West, “Critical Editing,” 26.

The Edition of III–IV Kingdoms

79

Kauhanen, Tuukka, Lucifer of Cagliari and the Text of 1–2 Kings (Atlanta: SBL Press, 2018). Moreno Hernández, A., Las glosas marginales de Vetus Latina en las biblias vulgatas españolas: 1 y 2 Reyes, TECC 49 (Madrid: CSIC, 1992). Piquer, Andrés, Pablo Torijano, and Julio Trebolle, “Versions of the Septuagint, Greek Recensions and Hebrew Editions: Text-Critical Evaluation of the Old Latin, Armenian and Georgian Versions in III–IV Regnorum” in Translating a Translation: The LXX and Its Modern Translations in the Context of Early Judaism, ed. H. Ausloos (Leuven: Peeters, 2008), 251–282. Rahlfs, Alfred, Lucians Rezension der Königsbücher. Septuaginta Studien III (Göttingen: Vandenhoeck & Ruprecht, 1911). Talshir, Zipora, The Alternative Story of the Division of the Kingdom 3 Kingdoms 12:24a–z (Jerusalem: Simor, 1993). Tov, Emanuel, Textual Criticism of the Hebrew Bible, Third Edition, Revised and Expanded (Minneapolis: Fortress Press, 2012). Wevers, John William, “The Gottingen Pentateuch: Some Post-Partem Reflections,” in VII Congress of the International Organization for Septuagint and Cognate Studies, Leuven 1989, ed. C. E. Cox (Atlanta: Scholars Press, 1991), 51–60.

Robert J. V. Hiebert A Synopsis of the Textual History of 4 Maccabees1 1. Introduction: Titles, Authorship, Purpose, Genre, Provenance The manuscript tradition associated with the book of 4 Maccabees attests to a variety of titles. Codices Sinaiticus (fourth century CE), Alexandrinus (fifth century CE), and Venetus (eighth century CE), along with a few other witnesses (542 46 771) contain what is undoubtedly the earliest one – Μακκαβαίων δ′ “4 Maccabees.” Eusebius, who entitled this treatise Περὶ αὐτοκράτορος λογισμοῦ “On the Supremacy of Reason,” attributed it to Josephus2 as did Jerome, who was dependent upon Eusebius.3 In fact, many Greek and Syriac manuscripts of 4 Maccabees associate it with Josephus, though the wording in these titles is diverse, as the following examples demonstrate:4 Ἰωσήπου εἰς τοὺς Μακκαβαίους (q′–370c 3002 325) “Of Josephus regarding the Maccabees”; διήγησις Ἰωσίππου εἰς τὸ μαρτύριον τῶν ἁγίων Μακκαβαίων (m–325 473 607 686 714*(vid)–455c 577) “Narration of Josephus regarding the Testimony of the Holy Maccabees”; Ἰωσήπου περὶ σώφρονος λογισμοῦ (534–728txt 62 340) “Of Josephus on Temperate Reason”; ̈ ‫ܘܒܢܝܗ‬ ‫( ܡܐܡܪܐ ܕܝܘܣܝܦܘܣ ܥܠ ܐܠܝܥܙܪ ܘܫܡܘܢܝ‬SyB*a–f = 12a1* 17a1–a4 17/15a1 f Sy ) “A Discourse of Josephus concerning Eleazar and Shamoni and Her Sons”; ̈ ‫ܘܒܢܝܗ‬ ‫ܕܡܟܬܒ ܠܗ ܥܠ ܐܠܝܥܙܪ ܘܫܡܘܢܝ‬ ݂ ‫( ܡܐܡܪܐ ܕܝܘܣܝܦܘܣ ܚܟܝܡܐ‬SyC = 12h1) “A Discourse of the Learned Josephus That Was Written by Him concerning Eleazar and Shamoni and Her Sons.” The claim that Josephus was the author of 4 Maccabees is, however, implausible. For one thing, whereas the 4 Maccabees author reports that Antiochus IV Epiphanes is the son of Seleucus IV Philopator (4:15), Josephus knows that they are brothers.5 Furthermore, while the author of 4 Maccabees eulogizes those who steadfastly refuse to yield to their foreign overlord, Antiochus – ἀνὴρ ὑπερήφανος καὶ δεινός “an arrogant and terrible man” (4:15) – due to their unshakeable commitment to God and his laws (chapters 5–18), Josephus characterizes the Jewish zealots who went to war with Rome in 66–73 CE as seditious rebels rather than as stalwart defenders of their faith, and the Romans at times as compassionate opponents.6 Thus, although 4 Maccabees may well have 1  For further information on this topic, see Hiebert, “4 Maccabees,” 306–319; Hiebert, “Makkabaion IV / Das vierte Buch der Makkabäer,” 322–329; Hiebert, “Textual History of 4 Maccabees,” “4 Maccabees: Greek,” “4 Maccabees: Latin,” “4 Maccabees: Syriac.” 2 Eusebius, Historia ecclesiastica 3.10.6. 3 Jerome, De viris illustribus, s. v. “XIII. Josephus Matthiae filius.” 4  Variant readings are not noted here. 5  Antiquitates judaicae 12.234. 6  Bellum judaicum 1.10, 27–28; 2.258–265, 648–651; 4.121–161.

82

Robert J. V. Hiebert

been written during the period of turmoil that culminated in the First Jewish-Roman War chronicled by Josephus in Bellum judaicum, the identity of the author of the treatise that celebrates the Maccabean martyrs has been lost to history.7 The purpose of that author, whatever the title of this discourse might be, is to make the case for mastering human emotions/passions (τὰ πάθη) through the exercise of pious reason (ὁ εὐσεβὴς λογισμός). Grounding his argument on both Jewish νόμος “law” and Greek philosophical ideas, the author recounts stories about biblical characters who met various kinds of challenges with fortitude and resolve, but then for most of the book focuses on a Jewish ἱερεύς “priest” and νομικός “lawyer” named Eleazar and seven brothers and their mother who, in the course of the attempts by ὁ τύραννος “the tyrant” Antiochus to get them to abandon their faith and ancestral traditions, unflinchingly resisted the inducements offered them and endured the threats and torture meted out against them. The literary genre of 4 Maccabees continues to be a matter of debate because elements of literary forms such as sermon, diatribe, encomium, funeral oration, and protreptic address have variously been detected by scholars. That the book was composed in Greek seems clear from its elegant syntax. Where it was written remains, like the identity of its author, unknown. Suggestions have included Alexandria, Jerusalem, and Antioch of Syria, where a Maccabean martyr cult was established. Whatever the exact location, most scholars have concluded that Syria or Asia Minor would most likely have been the general locale.8 The fact that 4 Maccabees was included in two of the great early Bible codices – Sinaiticus and Alexandrinus – is an indication of the high regard in which it was held in the communities that produced them. Found in more than seventy extant Greek manuscripts penned by Christian scribes from the fourth to sixteenth centuries, this book became surprisingly popular in Christian circles, as is also attested by the influence it exerted on the orations and writings of Origen, Gregory of Nazianzus, John Chrysostom, and Ambrose of Milan.9

7 Hiebert, “4 Maccabees,” 307–309. 8 Hadas, The Third and Fourth Books of Maccabees, 95–115; Anderson, “4 Maccabees,” 533–536; Klauck, 4. Makkabäerbuch, 659–669; deSilva, 4 Maccabees: Introduction and Commentary on the Greek Text, xi-xxv; Rajak, “The Fourth Book of Maccabees in a Multi-Cultural City,” 134–150. 9 Rahlfs, Verzeichnis der griechischen Handschriften des Alten Testaments, 387–390; Origen, Exhortatio ad martyrium (TLG 22–27); Gregory of Nazianzus, In Machabaeorum laudem, oration 15 (TLG 35.912–933); John Chrysostom, De Maccabeis, homilies 1–3 (TLG 50.617– 628) and De Eleazaro et septem pueris (TLG 63.523–530); Ambrose, De officiis minoribus (PL 16.83–84) and De Jacob et vita beata (PL 14.597–638). See also Freudenthal, Die Flavius Josephus beigelegte Schrift Ueber die Herrschaft der Vernunft, 29–36, 157–159.

A Synopsis of the Textual History of 4 Maccabees

83

2. Extant Versions 2.1 Greek 2.1.1  Early Editions The earliest printed text of Greek 4 Maccabees, which was based on a single manuscript, appeared in 1526, and similar versions, based on only one or two witnesses, were published over the next three and a half centuries.10 In 1871, Fritzsche produced an edition of the book that is primarily dependent on the texts of Alexandrinus and Sinaiticus, though in his introduction he referred to a dozen more manuscripts from which he drew variant readings recorded in the apparatus.11 Some twenty-three years later, Swete published 4 Maccabees as part of The Old Testament in Greek according to the Septuagint. This was a diplomatic edition based largely on Alexandrinus,12 but that included variant readings from Sinaiticus and Venetus. Rahlfs’ Septuaginta, first published in 1935, contains an eclectic text based on both Sinaiticus, which he tended to favour, and Alexandrinus, although variants gleaned from those two uncials as well as from Venetus, are recorded in the apparatus.13 2.1.2  Greek Witnesses Uncials: A S V A 542 (11:5 – fin libri) L = 236 491 534 728 q = 71 74 120 370 380 452 731 3002 q1 = 44 107 610 q2 = 55 747 m = 316 317 322 325 391 397 446 457 467 472 473 586 591 592 594 595 596 597 607 617 639 640 641 656 677 682 683 686 695 699 713 714 774 778 782 789 m1 = 455 585 m2 = 587 738 m3 (init. libri – 11:4) = 62 542 747c/mg 11:5 – fin libri: m′’ 62 747c/mg Codices mixti: 46 58 340 577 668 690 741 771 773 930 Most of the Greek manuscripts listed above can be grouped according to characteristic textual affiliations. Codices Alexandrinus, Sinaiticus, and Venetus, which 10 Hadas, The Third and Fourth Books of Maccabees, 136–137; Townshend, “The Fourth Book of Maccabees,” 664; Anderson, “4 Maccabees,” 532. 11 Fritzsche, “ΜΑΚΚΑΒΑΙΩΝ ΤΕΤΑΡΤΟΣ,” xx-xxii. 12 Swete, Hosea – 4 Maccabees, Psalms of Solomon, Enoch, The Odes, vol. 3, vi and n. 2, acknowledged that, because of the many errors in A, quite a number of corrections from other manuscripts were admitted. 13 Rahlfs, Septuaginta: Id est Vetus Testamentum graece iuxta LXX interpretes.

84

Robert J. V. Hiebert

are normally among the best witnesses to the Old Greek, tend not to exhibit significant agreement with one another when they diverge from the original text. From 11:5 onward, minuscule manuscript 542 is a textual congener of Alexandrinus. Another uncial, the very fragmentary manuscript 930, should be classified with the codices mixti (46 58 340 577 668 690 741 771 773 930), which are normally not aligned with other manuscripts or groups when they deviate from the Old Greek. Manuscripts 236 491 534 728 attest a text type that the editors of the Göttingen editions of 1–3 Maccabees have assigned the siglum L and have associated with the recension of Lucian of Antioch.14 Some of the characteristics that typify Lucian’s work (e.g., addition of particles, conjunctions, and pronouns; Atticisms) do occur in this group in 4 Maccabees, but it is debatable whether this recensional activity in these four manuscripts is attributable to Lucian since there is no literary substantiation of his involvement with, let alone his interest in, any of the books of Maccabees.15 The textual histories of the Göttingen editions of 1–3 Maccabees all include a manuscript group that their editors have designated as q.16 The situation in 4 Maccabees is more complex, however, given the fact that, in addition to a main q group (71 74 120 370 380 452 731 3002), there are two subgroups, q1 (44 107 610) and q2 (55 747). The q and q1 groups exhibit certain Atticistic features, such as the elision of particles, conjunctions, and prepositions that end in short vowels before words beginning with vowels, and the replacement of οὕτως with οὕτω before words that begin with consonants.17 Who might be responsible for this recensional activity remains unknown. The greatest number of Greek manuscripts of 4 Maccabees are either menologia – i.e., manuscripts associated with the commemoration of Christian saints in conjunction with their feast days – or ones that exhibit the same text type as the menologia. These manuscripts, which are divided into a cluster of related groups that are not, however, part of the textual histories of the other books of Maccabees, have been assigned the siglum m. As in the case of the q cluster, there is a main m group (316 317 322 325 391 397 446 457 467 472 473 586 591 592 594 595 596 597 607 617 639 640 641 656 677 682 683 686 695 699 713 714 774 778 782 789) and related subgroups: m1 (455 585), m2 (587 738), and m3 (62 542 747c/mg). 14 Kappler, Maccabaeorum liber I, 26–27; Kappler and Hanhart, Maccabaeorum liber II, 18–23; Hanhart, Maccabaeorum liber III, 17–24. 15  H. Dörrie, Passio SS. Machabaeorum, 8 n. 2. 16 Kappler, Maccabaeorum liber I, 24–26; Kappler and Hanhart, Maccabaeorum liber II, 24–26; Hanhart, Maccabaeorum liber III, 28–32. 17 Smyth, Greek Grammar, §§46, 70–75, 136. In Hellenistic/Koine Greek, however, οὕτως can appear before words beginning with either vowels or consonants, and elision occurs much less frequently than in Attic Greek, cf. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, §§3 (p. 22), 9, 9–10 (pp. 136–137).

A Synopsis of the Textual History of 4 Maccabees

85

The m3 group, however, is extant only as far as 11:4, after which ­manuscript 542 comes to be aligned with Codex Alexandrinus, while manuscripts 62 and 747c/mg continue to be affiliated with the remaining m textual tradition. The m groups display some agreement with the L group in regard to certain textual characteristics (such as the addition of pronouns, conjunctions, and the vocative particle ὦ) and to other divergences from the Old Greek, but all in all they evince considerably more comprehensive recensional activity than appears in the other textual traditions.18 In a number of Greek manuscripts, certain parts of the text appear to have been omitted intentionally. The kind of material that has been excised is often that which is expressed in a similar fashion elsewhere in the book and could therefore have been regarded as redundant. The following omissions in Greek witnesses are noteworthy: 71: 16:25–17:4; 17:13–18:1 Ἰσραηλῖται q1 747c2: 5:36; 6:21–24 ἀνάγκας; 7:17–23; 14:15–17; 15:22–24; 17:20 οὐ – (22) ἁμαρτίας 317: 6:31–9:9; 12:13–16:21 ἐβλήθη 58: 7:15, 17; 13:6–10, 15, 18–22, 26–14:1; 14:3; 15:1–4, 6 ἥτις – 15, 19 οὐκ – 20, 24–32; 16:2–13, 18–19 ὑπομένειν; 17:2–22; 18:3–4 690: 15:15 καί 3° – 20 τέκνων 1°, 21–28; 16:4 ἡ – 23; 17:3; 17:20 οὐ – (22) ἁμαρτίας

2.1.3  Literary Character As indicated above, the book of 4 Maccabees was written in Greek, not translated from a Semitic original. The author wrote with literary sophistication and rhetorical flourish. The vocabulary that he employed was extensive, with more than one-quarter of the words in the book – apart from names – not occurring elsewhere in the Septuagint. Included in this lexicon are a good number of neologisms, many of which are compound words: ἀδελφοπρεπῶς “worthy of brothers” (10:12), ἀλλοφυλέω “become an allophyle” (18:5), ἀνατίκτω “bring forth again, give a new birth” (16:13), ἀντιρρητορεύω “dispute with” (6:1), ἀποξαίνω “tear” (6:6), ἀρθρέμβολον “torture instrument, joint dislocator” (8:13; 10:5), ἀσθενόψυχος “of tender spirit” (15:5), αὐτοδέσποτος “absolute master” (1:1, 30; 13:1), δειλόψυχος “fainthearted” (8:16; 16:5), ἐθνοπάτωρ “father of a nation” (16:20), ἐθνόπληθος “throng of people” (7:11), ἐκμελίζω “dismember” (10:5, 8; 11:10), ἐκπολιτεύω “alter the form of government” (4:19), ἐκριζωτής “uprooter” (3:5), ἐναγκάλισμα “embrace, that which embraces” (13:21), ἐννοσσοποιέομαι “build a nest” (14:16), ἐπασθμαίνω “breathe hard, gasp for breath” (6:11), ἐπικαρπολογέομαι “glean” (2:9), ἐπιρρωγολογέομαι “glean grapes” (2:9), ἑπταμήτωρ “mother of seven” (16:24), 18 Hiebert, “Makkabaion IV / Das vierte Buch der Makkabäer,” 324–325; Hiebert, “In Search of the Old Greek Text of 4 Maccabees,” 127–129.

86

Robert J. V. Hiebert

ἱ­ ερόψυχος “holy-minded” (17:4), ἰσάστερος “star-like” (17:5), καρτεροψυχία “strength of spirit” (9:26), κηρογονία “making of honeycombs” (14:19), κοσμοπληθής “filling the world” (15:31), κοσμοφορέω “sustain the world” (15:31), μαλακοψυχέω “play the coward” (6:17), μεταδιαιτάω “change one’s way of life” (8:8), μεταπαιδεύω “train differently” (2:7), μιαροφαγία “eating of defiling foods” (5:27; 6:19; 7:6; 11:25), μονοφαγία “private gormandizing” (1:27), οἰστρηλασία “frenzied urge” (2:4), ὀλεθροφόρος “destruction-bringing, calamitous” (8:19), ὁμοζηλία “common zeal” (13:25), παγγέωργος “master cultivator” (1:29), παθοκράτεια “control of passions” (13:5, 16), παντοφαγία “indiscriminate eating” (1:27), περιλακίζομαι “be torn all around” (10:8), περιχαλάω “relax” (7:13), προκακόομαι “be previously afflicted” (17:22), προσεπικατατείνω “tighten further” (9:19), προσημειόομαι “prognosticate, forebode” (15:19), ὑπερασπίστρια “one who holds a shield over, champion” (15:29). In the light of the declining occurrence of the optative in post-classical literature and common usage, the author’s rather frequent use of it is noteworthy. He frequently resorted to melodramatic excess in his employment of figures of speech such as apostrophe (7:10) and prosopopoeia (16:5–6) and in his proliferation of vivid metaphors and similes involving horticulture (1:28–29), athletics (6:10–11; 15:29; 17:11–16; 18:23), sailing (7:1–3), siege warfare (7:4), structures that withstand the ravages of storm and earthquake (7:5; 13:6–7, 13; 17:3), performing arts (8:4; 14:3–8; 15:21; 18:23), animal attack (9:28), animal nurture (14:13–20), and politics (15:25–28).19

2.2 Syriac 2.2.1  Syriac Witnesses The only early translation of 4 Maccabees that is fully extant is the Syriac version, attested in twelve Peshitta manuscripts ranging in date from the seventh to the nineteenth centuries:20 7a1 = A; 12a1 = B; 12h1 = C; 17a2 = a; 17a4 = b; 17a3 = c; 17/15a1 = d; 17a1 = e; an undated, possibly Florentine, manuscript = f; 15a3; 17a10; 19g3. 2.2.2 Edition An edition published by Bensly and Barnes21 contains an eclectic text for which the first nine witnesses listed above were collated but which is reliant to a significant extent on the first one, 7a1 = A. 19 Breitenstein, Beobachtungen zu Sprache, Stil und Gedankengut, 15–17, 28–32, 177–178; Hiebert, “4 Maccabees,” 310–311. 20  The numerical designations (7a1 etc.) are those assigned by the Peshitta Institute. The alphabetical ones (A B C a b c d e f) are the ones used in the edition described below. 21  Bensly and Barnes, The Fourth Book of Maccabees and Kindred Documents.

A Synopsis of the Textual History of 4 Maccabees

87

2.2.3  Literary-Critical Considerations In the same volume that contains their edition of Syriac 4 Maccabees, Bensly and Barnes include an anonymous memra or poem, structured in twelve-syllable verse, which is described as a paraphrase of much of 4 Maccabees.22 Peterson, who designates the memra 6 Maccabees, argues that it embodies a significant component of what is a complex literary history of 4 Maccabees. Her contention is that, like chapters 6–7 of 2 Maccabees, 6 Maccabees – which she maintains is based on Greek, Aramaic, and Syriac source materials from between the second century BCE and the sixth or seventh century CE – served, presumably in its constitutive earliest stages, as a source for 4 Maccabees.23 Freudenthal, however, suggests that 4 Maccabees is likely dependent, not specifically on 2 Maccabees, but on the long-lost, five-volume work of Jason of Cyrene, whose account of the Maccabean period the author of 2 Maccabees undertook to condense into a single volume (2 Macc. 2:19–23).24 These different hypotheses regarding the literary history of 4 Maccabees point to the need for further investigation of the relationships between these documents. 2.2.4  Translation Technique While it is fair to say that the Syriac translation of 4 Maccabees “may be generally described as faithful,”25 it is not a stilted rendering of its Greek source text in the manner of Paul of Tella’s translation of Origen’s Hexapla. The technique of the Syriac translator of 4 Maccabees ranges “from the reasonably literal to the freely interpretative.”26 Indeed, Hanhart’s comment that the text of Syriac 3 Maccabees is “oft frei ausgestaltet, so daß die Textform der Vorlage nur noch schwer erkennbar ist”27 is also applicable to Syriac 4 Maccabees. That being said, the Greek Vorlage of Syriac 4 Maccabees seems more often than not to have been the Old Greek text. Furthermore, in its divergences from the Old Greek, the Syriac version does not appear to be closely aligned with any of the Greek manuscripts or groups.28 The translation technique of Syriac 4 Maccabees can be illustrated by surveying equivalences for commonly-occurring Greek words. The term λογισμός “reason,” which is at the heart of the theme of the book, is rendered by ‫ܪܥܝܢܐ‬ “mind, intellect, way of thinking” (53x), ‫“ ܚܘܫܒܐ‬thinking, thought” (7x), ‫ܗܘܢܐ‬ 22  Bensly and Barnes, The Fourth Book of Maccabees and Kindred Documents, xxv, xlviii-lxxii, ‫ܩܢܕ‬-‫ܩܟܘ‬. 23 Peterson, “Martha Shamoni: A Jewish Syriac Rhymed Liturgical Poem about the Maccabean Martyrdoms (Sixth Maccabees),” vi, 2–3, 7, 25–27, 128–135, 214–215. 24 Freudenthal, Die Flavius Josephus beigelegte Schrift Ueber die Herrschaft der Vernunft, 72–90. 25  Bensly and Barnes, The Fourth Book of Maccabees and Kindred Documents, xiv. 26 Hiebert, “Preparing a Critical Edition of IV Maccabees: The Syriac Translation,” 202. 27 Hanhart, Maccabaeorum liber III, 9. 28 Hiebert, “In Search of the Old Greek Text of 4 Maccabees,” 130.

88

Robert J. V. Hiebert

“mind, reason” (4x), and ‫“ ܡܚܫܒܬܐ‬thought, reasoning” (1x). The equivalents for εὐσεβεία “piety,” a quality frequently paired by the Greek author with λογισμός, are ‫“ ܫܪܪܐ‬truth” (22x), ‫“ ܪܚܡܬ ܫܪܪܐ‬love of truth” (3x), ‫“ ܕܚܠܬ ܐܠܗܐ‬fear of God” ̈ (17x), ‫“ ܗܝܡܢܘܬܐ‬faithfulness” (1x), and ‫ܢܛܘܖܬܐ ܕܢܡܘܣܐ‬ “observances of the law” (1x). The Syriac translator evidently considers ἀρετή “virtue” to belong to the same semantic domain as εὐσεβεία “piety” because he makes use of some of the same terms – ‫“ ܫܪܪܐ‬truth” (3x) and ‫“ ܕܚܠܬ ܐܠܗܐ‬fear of God” (1x) – to render ̈ ̈ it, though other counterparts are ‫“ ܡܝܬܪܘܬܐ‬virtue” (7x), ‫ܡܝܬܖܐ‬ ‫ܕܘܒܖܐ‬ “excellent habits” (1x), and ‫“ ܢܨܝܚܘܬܐ‬splendor, excellence” (2x). The book’s oft-repeated thesis is that λογισμός “reason” masters τὰ πάθη “the passions/emotions,”29 and in most cases the Syriac equivalent for πάθος “passion, emotion” is ‫“ ܚܫܐ‬passion, emotion” (55x); other counterparts, however, are ‫“ ܟܐܒܐ‬pain” (3x), ‫“ ܫܢܕܐ‬torment” (1x), ‫“ ܣܘܥܪܢܐ‬deed” (1x), and ‫“ ܪܚܡܐ‬womb,” plural: “tenderness, compassion” (1x). One of the instruments that is said to have been used to inflict pain and torment is the καταπέλτης “catapult,” a term that the Syriac translator renders quite differently in its various contexts: ‫“ ܩܝܣܐ‬the stake” (11:9); ‫ܦܪܙܐܠ‬ “iron [tool]” (11:26); ‫“ ܫܦܘܕܐ‬skewer” (18:20). In another case (9:26), the dissimilarity between the Greek and Syriac texts is such that no specific counterpart to καταπέλτης is discernible: ὀργάνῳ καὶ καταπέλτῃ προσέδησαν αὐτόν “they bound him to the torture machine and catapult” (NETS); ‫ܟܕ ܬܐܠ ܓܙܡܝܢ ܗܘܘ‬ ‫“ ܕܢܣܪܩܘܢܝܗܝ‬while he was suspended they were threatening to lacerate him with combs.”

2.3  Coptic Sahidic 2.3.1  Coptic Sahidic Witnesses Another early translation of 4 Maccabees is the Coptic Sahidic version, which is unfortunately now only extant in six leaves of two codices: Sa 2115 = Paris, National Library, BnF Copte 1313 f. 28, two leaves containing 1:1–16; University of Michigan, Mich. Ms. 158 (34) a–f, three leaves containing 15:16–21, 16:14–23, 17:7–12, and 18:6–15; Sa 2116 = a leaf containing 5:2–6 from a miniature codex discovered during excavations in Antinoë in 2005. 2.3.2 Publication The leaves of Sa 2115 were published by Ivan Miroshnikov in an article that included a transcription of their text, an English translation, and photographs of the fragments in the University of Michigan Library.30 A seventh or eighth century ostracon that mentions Coptic 4 Maccabees in a list of non-canonical books 29  In all but two (1:14[1°], 24) of the sixty-two occurrences of πάθος in 4 Maccabees, it is in the plural number. In just one case (1:1) is there no Syriac counterpart. 30 Miroshnikov, “The Sahidic Coptic Version of 4 Maccabees,” 69–92.

A Synopsis of the Textual History of 4 Maccabees

89

furnishes proof that this translation occurred some centuries before the tenth century, the date assigned by Miroshnikov to the codex from which these leaves come.31 The text of Sa 2116 was published by Alain Delattre, who dated this manuscript fragment to the fifth century CE.32

2.3.3  Noteworthy Features The surviving remnants of the Coptic version exhibit evidence of intentional omissions in comparison to the Old Greek text. There is partial overlapping with the curtailments in some of the Greek manuscripts, though it would appear that this sort of activity occurred independently in each version: Coptic: 15:22–16:13 Greek: 15:22–24 (q1 747c2); 12:13–16:21 ἐβλήθη (317); 15:24–32 and 16:2–13 (58); 15:21–28 and 16:4 ἡ – 23 (690) Coptic: 16:24–17:6 Greek: 16:25–17:4 (71); 17:2–22 (58); 17:3 (690) Coptic: 17:13–18:5 Greek: 17:13–18:1 Ἰσραηλῖται (71); 17:2–22 and 18:3–4 (58); 17:20 οὐ – (22) ἁμαρτίας (q1 747c2 690)

A distinguishing mark of the Coptic version pertains to how the martyrs are said to be related to one another. In the Coptic text, Eleazar, the aged priest, is called the father of the seven young men and the husband of their mother (16:15, 17:9, 18:9), while in the Greek version nothing is said about them being from the same family. Coptic 4 Maccabees features a significant number of unique variant readings. Because it is a quite free translation of the Old Greek, it is of limited value for reconstructing the original text.33

2.4 Latin 2.4.1 Edition The Latin work entitled Passio Sanctorum Machabaeorum is, strictly speaking, not a translation of 4 Maccabees but rather a free, Christianized adaptation of it. Dörrie, who dated Passio to the fourth century CE, published a critical edition 31 Miroshnikov, “4 Maccabees: Coptic;” Ostracon P 1069, transcribed in Crum, “The Literary Material,” 197. 32 Delattre, “Textes coptes et grecs d’Antinoé.” 33  Feder, in fact, argues that the books of Maccabees were not part of the standard Coptic Sahidic translation and, therefore, were not considered to be canonical in an official sense. The few early translated portions that are extant are free renderings and were probably used in liturgy in conjunction with the feasts of martyrs who were held in high esteem by the Coptic monks (Feder, “The Coptic Canon”).

90

Robert J. V. Hiebert

for which he collated thirty-nine manuscripts whose dates range from the ninth to the fifteenth centuries.34

2.4.2  Latin Witnesses The unabridged text of Passio is found in only ten manuscripts, a group that Dörrie designated Archetypus 𝔄. It consists of five “direkten Zeugen” (manuscripts i q t y z, tenth to thirteenth centuries) and five members of Subarchetypus 𝔞, which stands for textus 𝔞ustriacus (manuscripts w n h l m, twelfth to fifteenth centuries).35 The remaining twenty-nine manuscripts contain abridged texts, a group that he called Archetypus 𝔙 to denote “Hss. des 𝔳erkürzten Textes” or the “𝔙ulgatÜberlieferung.” It is comprised of manuscript e (ninth century) and the Subarchetypi 𝔥 (= textus 𝔥elveticus, the “süddeutsch-schweizerische Familie,” manuscripts G K O, ninth to eleventh centuries), 𝔠 (= manuscripts b r c g p p' p'', “vorwiegend aus ℭistercienser-Klöstern,” eleventh to thirteenth centuries), and 𝔣 (= manuscripts L N S V W D F B C E Q P d o s X H a, the “𝔣ranzösische Familie,” eleventh to fifteenth centuries).36 What was omitted in Archetypus 𝔙 amounts to almost half the text of Archetypus 𝔄 and, according to Dörrie, involves “alles Philosophische” and “ein großer Teil der rhetorischen Lobpreisungen.”37 The truncating activity that occurred during the course of the textual history of Passio was clearly an inner-Latin phenomenon, unrelated textually to what took place in certain Greek manuscripts and in the Coptic Sahidic version. Moreover, the loose literary relationship between Passio and 4 Maccabees makes it difficult to reconstruct the Greek source text of Passio, and consequently, this Latin composition is of little use in establishing the Old Greek text of 4 Maccabees.

3. Bibliography Anderson, Hugh, “4 Maccabees,” in The Old Testament Pseudepigrapha, vol. 2: Expansions of the “Old Testament” and Legends, Wisdom and Philosophical Literature, Prayers, Psalms, and Odes, Fragments of Lost Judeo-Hellenistic Works, ed. by J. H. Charlesworth (Garden City: Doubleday, 1985), 531–564. 34 Dörrie, Passio. His conclusions regarding the date of its composition are stated on pp. 36 and 43. 35  Upper and lower case Gothic characters are used to designate Archetypi and Subarchetypi, respectively. Upper case Latin letters are used for Bible manuscripts, and lower case Latin letters for manuscripts that contain “Passionalia und Legendarien.” The order in which manuscripts are listed is in accordance with their text-critical importance (Dörrie, Passio, 2, 63–64). 36 Dörrie, Passio, 2–3. 37 Dörrie, Passio, 10.

A Synopsis of the Textual History of 4 Maccabees

91

Bensly, Robert L. and William E. Barnes, eds., The Fourth Book of Maccabees and Kindred Documents in Syriac (Cambridge: University Press, 1895). Breitenstein, Urs, Beobachtungen zu Sprache, Stil und Gedankengut der Vierten Makkabäerbuchs (Basel/Stuttgart: Schwabe, 1976). Crum, Walter Ewing, “The Literary Material,” in The Monastery of Epiphanius at Thebes, Part I (New York: The Metropolitan Museum of Art, 1926). Deissmann, Adolf, “Das vierte Makkabäerbuch,” in Die Apokryphen und Pseudepigraphen des Alten Testaments, 2 vols., vol. 2, ed. E. Kautzsch (Hildesheim: Georg Olms, 1900), 149–176. Delattre, Alain, “Textes coptes et grecs d’Antinoé,” in Antinoupolis, vol. 1, ed. R. Pintaudi, Scavi e Materiali 1 (Florence: Istituto Papirologico “G. Vitelli,” 2008), 131–162. DeSilva, David Arthur, 4 Maccabees: Introduction and Commentary on the Greek Text in Codex Sinaiticus. Septuagint Commentary Series (Leiden/Boston: Brill, 2006). Dörrie, Heinrich, Passio SS. Machabaeorum. Abhandlungen der Gesellschaft der Wissenschaften zu Göttingen. Philologisch-Historische Klasse, 3.22 (Göttingen: Vandenhoeck & Ruprecht, 1938). Dupont-Sommer, André, Le quatrième livre des Machabées (Paris: Honoré Champion, 1939). Erasmus, Desiderius, Flavii Iosepi viri iudaei περὶ αὐτοκράτορος λογισμοῦ. Hoc est de imperatrice ratione, deque inclyto septem fratrum Macabaeorum, ac fortissimae eorum matris diuae Solomonae martyrio liber (Cologne: E. Cervicornus, 1517). Feder, Frank, “1.1.6 The Coptic Canon,” in Textual History of the Bible, vol. 2A: The Deuterocanonical Scriptures, ed. by F. Feder and M. Henze (Leiden: Brill, 2020), 213–239. Freudenthal, Jacob, Die Flavius Josephus beigelegte Schrift Ueber die Herrschaft der Vernunft (IV Makkabäerbuch) (Breslau: Schletter’sche Buchhandlung, 1869). Fritzsche, Otto Fridolin, “ΜΑΚΚΑΒΑΙΩΝ ΤΕΤΑΡΤΟΣ,” in Libri apocryphi Veteris Testamenti graece (Leipzig: F. A. Brockhaus, 1871), 351–386. Grimm, Carl Ludwig Willibald, “Viertes Buch der Maccabäer,” in Kurzgefasstes exegetisches Handbuch zu den Apokryphen des Alten Testaments, 6 vols., vol. 4, ed. O. F. Fritzsche and C. L. W. Grimm (Leipzig: S. Hirzel, 1851–1860), 283–370. Hadas, Moses, ed. and trans., The Third and Fourth Books of Maccabees. Jewish Apocryphal Literature (New York: Ktav, 1951). Hanhart, Robert, ed., Maccabaeorum liber III. Septuaginta: Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum, vol. 9.3 (2nd ed., Göttingen: Vandenhoeck & Ruprecht, 1980). Hiebert, Robert J. V. “4 Maccabees,” in The T&T Clark Companion to the Septuagint, ed. J. K. Aitken (London: Bloomsbury T&T Clark, 2015), 306–319. —, “10.4.1 Textual History of 4 Maccabees,” “10.4.2 4 Maccabees: Greek,” “10.4.3 4 Maccabees: Syriac,” “10.4.4 4 Maccabees: Latin” (with David J. Sigrist), in Textual History of the Bible, vol. 2C: The Deuterocanonical Scriptures, ed. F. Feder and M. Henze (Leiden: Brill, 2019), 187–213. —, “4 Maccabees 18,6–19 – Original Text or Secondary Interpolation?” in Die Septuaginta – Texte, Kontexte, Lebenswelten: Internationale Fachtagung veranstaltet von Septuaginta

92

Robert J. V. Hiebert

Deutsch (LXX.D), Wuppertal 20.–23. Juli 2006, ed. M. Karrer and W. Kraus. WUNT 219 (Tübingen: Mohr Siebeck, 2008), 439–449. —, “A New Critical Edition of Greek IV Maccabees,” in Die Septuaginta – Orte und Intentionen: 5. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 24.–27. Juli 2014, ed. S. Kreuzer, M. Meiser, and M. Sigismund. WUNT 361 (Tübingen: Mohr Siebeck, 2016), 389–403. —, “In Search of the Old Greek Text of 4 Maccabees,” in Text-Critical and Hermeneutical Studies in the Septuagint, ed. J. Cook and H.-J. Stipp. VTSup 157 (Leiden: Brill, 2012), 127–143. —, “Makkabaion IV / Das vierte Buch der Makkabäer,” in Handbuch zur Septuaginta / Handbook of the Septuagint, LXX.H 1: Einleitung in die Septuaginta, ed. S. Kreuzer (Gütersloh: Gütersloher Verlagshaus, 2016), 322–329. —, “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, ed. F. García Martinez and M. Vervenne. BETL 192 (Leuven: Peters, 2005), 193–216. Kappler, Werner, ed., Maccabaeorum liber I. Septuaginta: Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum, vol. 9.1 (3rd ed., Göttingen: Vandenhoeck & Ruprecht, 1990). Kappler, Werner and Robert Hanhart, eds., Maccabaeorum liber II. Septuaginta: Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum, vol. 9.2 (2nd ed., Göttingen: Vandenhoeck & Ruprecht, 1976). Klauck, Hans-Josef, 4. Makkabäerbuch. JSHRZ 3.6 (Gütersloh: Gerd Mohn, 1989). —, “Makkabaion IV: Das vierte Buch der Makkabäer.” in Septuaginta Deutsch: Das griechische Alte Testament in deutscher Übersetzung, ed. W. Kraus and M. Karrer (Stuttgart: Deutsche Bibelgesellschaft, 2009), 730–746. —, “Makkabaion IV: Das vierte Buch der Makkabäer,” in Septuaginta Deutsch: Erläuterungen und Kommentare zum griechischen Alten Testament, vol. I: Genesis bis Makkabäer, ed. W. Kraus and M. Karrer (Stuttgart: Deutsche Bibelgesellschaft, 2011), 1445–1475. Lucchesi, Enzo, “Découverte d’une traduction copte du quatrième livre des Maccabées (BHG 1006),” AnBoll 99 (1981): 302. —, “Encore trois feuillets coptes du Quatrième Livre des Maccabées,” in Écritures et traditions dans la littérature copte: Journée d’études coptes, Strasbourg, 28 mai 1982. Cahiers de la Bibliothèque Copte 1 (Louvain: Peeters, 1983), 21–22. Miroshnikov, Ivan, “10.4.5 4 Maccabees: Coptic,” Textual History of the Bible, vol. 2C: The Deuterocanonical Scriptures, ed. F. Feder and M. Henze (Leiden: Brill, 2019), 213–218. —, “The Sahidic Coptic Version of 4 Maccabees,” VT 64 (2014): 69–92. Peterson, Sigrid, “Martha Shamoni: A Jewish Syriac Rhymed Liturgical Poem about the Maccabean Martyrdoms (Sixth Maccabees)” (PhD diss., University of Pennsylvania, 2006). Rahlfs, Alfred, Verzeichnis der griechischen Handschriften des Alten Testaments. MSU 2. (Berlin: Weidmannsche Buchhandlung, 1914). —, Septuaginta: Id est Vetus Testamentum graece iuxta LXX interpretes (Stuttgart: Deutsche Bibelgesellschaft, 1935. Rev. ed. R. Hanhart, 2006).

A Synopsis of the Textual History of 4 Maccabees

93

Rajak, Tessa, “The Fourth Book of Maccabees in a Multi-Cultural City,” in Jewish and Christian Communal Identities in the Roman World. Ancient Judaism and Early Christianity 94, ed. Y. Furstenberg (Leiden/Boston: Brill, 2016), 134–150. Renehan, Robert, “The Greek Philosophic Background of Fourth Maccabees,” Rheinisches Museum für Philologie 115 (1972): 223–238. Scarpat, Giuseppe, Quarto libro dei Maccabei. Biblica testi e studi 9 (Brescia: Paideia Editrice, 2006). Smyth, Herbert Weir, Greek Grammar (Cambridge, Mass.: Harvard University Press, 1956). Swete, Henry Barclay, ed., The Old Testament in Greek according to the Septuagint, vol. 3: Hosea – 4 Maccabees, Psalms of Solomon, Enoch, The Odes (Cambridge: University Press, 1894; 3rd ed., 1905). Thackeray, Henry St. John, A Grammar of the Old Testament in Greek according to the Septuagint (Cambridge: University Press, 1909). Townshend, Richard B., “The Fourth Book of Maccabees,” in The Apocrypha and Pseudepigrapha of the Old Testament in English, vol 2: Pseudepigrapha, ed. R. H. Charles (Oxford: Clarendon, 1913), 653–685. Westerholm, Stephen, “4 Makkabees,” in A New English Translation of the Septuagint, ed. A. Pietersma and B. G. Wright (New York / Oxford: Oxford University Press, 2007), 530– 541. Wright, William, A Catalogue of the Syriac Manuscripts Preserved in the Library of the University of Cambridge, 2 vols. (Cambridge: University Press, 1901).

Peter J. Gentry The Göttingen Edition of Ecclesiastes The difference between creating critical editions of classical texts and of the text of the Septuagint is primarily twofold: (1) the text of the Septuagint is a translation, and (2) the text of the Septuagint is a canonical and sacred text for both Jews and Christians. The fact that the text is a translation establishes immediately two different approaches to the study of the text: (1) one approach aims to determine the meaning intended by the translator, and (2) another aims to determine the meaning as understood by the communities reading and using this translation throughout the entire history of textual transmission. The canonical character of the text affects both approaches to the study of the text. For both translators and tradents, the text was the Word of God. The apparatus, therefore, of a critical and scientific reconstruction of the earliest form of the text does not simply serve scholars and students by providing information crucial to the constitution of the critical text. A complete and full history of the transmission of the text is necessary to serve the aims and approaches just stated. Each of my presentations in the last decade entailed discussion of the advantages and distinctive aims of the Göttingen Septuagint.1 The focus in this article is on the importance of the Göttingen Septuagint for critical editions of the Hebrew Bible such as the Biblia Hebraica Quinta. Nonetheless, in order to provide context, it is useful to begin with a brief survey of the major points.2

1. Inner-Greek Corruptions The person who produced the translation in Greek we now know as Ecclesiastes followed an approach of extreme formal and quantitative correspondence between Greek and Hebrew so that for almost one hundred years his work was mistaken for that of Aquila himself. As a result, there are parts of this translation that are difficult even for a native speaker of Hellenistic Greek to comprehend unless they also knew Hebrew and could consult the Hebrew source text. Since 1  See Gentry, “The Distinctive Aims of the Göttingen Apparatus: Examples from Ecclesiastes — An Edition in Preparation,” for a presentation given in 2008 at the 100 Year Jubilee of the Septuaginta-Unternehmen. At IOSCS and SBL in San Antonio, TX, 2016, I described the process of preparing the edition of Ecclesiastes in an unpublished paper entitled “Post-Partum Reflections on Preparing a Critical Edition of LXX Ecclesiastes.” See Gentry, “Ecclesiastes and Jerome’s Trifaria Varietas,” for a presentation at Wuppertal, Germany, 2018. 2  For detailed analyses, see the publications Gentry, “Special Problems in the Septuagint Text History of Ecclesiastes,” and Gentry, “Ecclesiastes and Jerome’s Trifaria Varietas.”

96

Peter J. Gentry

most scribes who copied the text did not know Hebrew and had no recourse to the parent text, sometimes they corrected the text on the basis of the context in order to improve the sense.

4:9–10a

MT 4:9 4:10a

‫ן־ה ֶא ָחד‬ ָ ‫טֹובים ַה ְּׁשנַ יִם ִמ‬ ִ ‫ׁש־ל ֶהם ָׂש ָכר טֹוב ַּב ֲע ָמ ָלם‬ ָ ֶ‫ֲא ֶׁשר י‬ ‫ת־ח ֵברֹו‬ ֲ ‫ִּכי ִאם־יִ ּפֹלּו ָה ֶא ָחד יָ ִקים ֶא‬

Göttingen Ecclesiastes 4:9 4:10a

ἀγαθοὶ οἱ δύο ὑπὲρ τὸν ἕνα, οἷς ἔστιν αὐτοῖς μισθὸς ἀγαθὸς ἐν μόχθῳ αὐτῶν· ὅτι ἐὰν πέσωσιν, ὁ εἷς ἐγερεῖ τὸν μέτοχον αὐτοῦ

————— A B C S (870) 998 O L C′’ d k min verss ————— 10 πέσωσιν] absc 870 Didlem et com 126,9 126,10; πεση 411 L(–125) (-σει 261) C′–299 (‑σει 609) 443 La160 Hilem et com 286,116 287,140 Aeth Arm Didlem 124,12 Ps.CatA 1103 Amb Ep 81,3.6 ter Inst virg 11,74 Chrom Matth 22,3,5.6 ter PetrChr 170,5 = Ald Vulg | ὁ εἷς] inc C; > 534; + ο ετερος 411 L(–125) C′–299 Amb Inst virg 11,74 Chrom Matth 22,3,5 = Ald ————— 10 πέσωσιν] α′ σ′ θ′ ὁμοίως τοῖς οʹ πέσωσιν Syh

The Hebrew Text can be rendered as follows: 9 10a

Two are better than one when they have a good reward for their labour; for if they fall, the one can lift up his partner.

In verse 10a, the end of the protasis and the beginning of the apodosis is clearly marked in MT by the fact that the verb ‫ יִ ּפֹלּו‬is in pause. According to the critical

The Göttingen Edition of Ecclesiastes

97

reconstruction of the Greek text, the translation is literal and the translator has fastidiously followed the order of the words in his source text: 9 10

Better are two than one, because they have a good reward in their toil. For if they fall, the one will raise his partner up3

Nonetheless, for scribes copying the Greek text without any knowledge of the Hebrew, delimiting the protasis and apodosis was not so transparent. Instead of both people falling down and each having a partner to help him get up, the position of ὁ εἷς naturally suggests a singular verb so that one of them falls and the other person who remains standing is the one who raises his fallen comrade. In this case ὁ εἷς belongs to the protasis and is construed as the subject of the verb πέσωσιν and the form is corrected from third person plural to third person singular πέσῃ. Since ὁ εἷς is now no longer understood as the subject of ἐγερεῖ, the same scribe(s) supplied ὁ ἕτερος as an explicit subject. It is now abundantly clear, contrary to the source text, that the benefit of two is that when only one falls, the other person who is still standing can then raise his fallen comrade.

2. Variants Preserved Almost Entirely in Early Daughter Versions Many translations of the Old Testament were made early and were not made from the original Hebrew, but rather from the Greek. Each daughter version represents at least one early Greek manuscript as its source. Of 113 Greek manuscripts analysed for the critical edition, only four are as old as, for example, the Armenian translation made in the fifth century CE. This does not include three fragmentary papyri. The source of the Armenian, then, is an important witness. Claude Cox, in his collation of the Armenian daughter version of Ecclesiastes was able to show that πονηρούς in place of πολλούς at 7:29, known otherwise only from patristic quotations apart from its preservation in Arm indicates that it was once to be found in a Greek MS, namely the parent text of Arm 1.

3 Gentry, “Ecclesiast.”

98

Peter J. Gentry

3. Variants Preserved Almost Entirely in Patristic Sources 7:20 MT

‫ִּכי ָא ָדם ֵאין ַצ ִּדיק ָּב ָא ֶרץ‬ ‫ֲא ֶׁשר יַ ֲע ֶׂשה־טֹוב וְ לֹא יֶ ֱח ָטא‬

Göttingen Ecclesiastes 7:20

ὅτι ἄνθρωπος οὐκ ἔστιν δίκαιος ἐν τῇ γῇ, ὃς ποιήσει ἀγαθὸν καὶ οὐχ ἁμαρτήσεται.

————— A B C S 998 O L C′’ d k min verss ————— 20 ἄνθρωπος] pr ὁ 371; post ἔστιν tr 336′ 443 Didcom 221,8 Ps.CatA 539 (sed hab Dam Metlem et com VII.1,3,81 PsChr Hi Gal 384 Hilem 309,307); post δίκαιος tr V; > Ath IV 92 Or IV 377,9: cf 3 Reg 846 | οὐκ ἔστιν] post δίκαιος tr L Ol | om δίκαιος Agnellus 45 Coll Avell 97,9 ConcilCarth Reg 115 ConcilMilev 7 Lucul 810 826 PsAug Ful 197,21 202,12 PsSalo Ecl 1005 Ruf Lev 459 SedScot Eph 1005 Armte

The quantitative approach of the Greek translator provides as a literal rendering of the Hebrew text: “for as to humanity, there is not a just person in the earth who will do good and not sin.” In the Hebrew, the sentence structure entails left dislocation or y movement – an element is removed from the clause and given first position.4 Such a literal rendering produces an awkward construction in Greek. Some scribes resolved this by moving ἄνθρωπος after the copula. Some scribes omitted ἄνθρωπος while others omitted δίκαιος. This last variation is attested only by 11 citations in 9 Latin fathers and by the Armenian. One of the main principles in the Göttingen editions is that citations from patristic sources are only given when they support a Greek manuscript. Here the citations from the Latin fathers do not support any Greek witness. Nonetheless, in conjunction with the Armenian, there is the distinct possibility that a Greek witness did attest the omission just as Greek witnesses also attested other inner

4  If an element is removed and not replaced by a pronoun, the fronting is called y movement. See Givón, Syntax.

The Göttingen Edition of Ecclesiastes

99

Greek “improvements” to an awkward sentence structure. The omission of the word “just” may also represent the Old Latin at this point.

4. Interdependence of LXX Text-History and Text-History of the Jewish Revisers Due to the influence of Origen’s Hexapla, the textual history of the Old Greek and the textual history of the text of the Jewish revisers is interdependent. Examples in 6:5 and 7:12 were discussed.5 In one case an entire line from Symmachus corrupted the textual transmission of the LXX. In the other instance, manuscripts of the LXX actually preserved a reading of the Three otherwise unknown. Thus, one cannot untangle the problem in the manuscripts of the Old Greek without recourse to the text of the Jewish revisers, and at other times one cannot discover the text of the Three without analysis of the textual transmission of the Old Greek.

5. Relation to Ancient Early Versions (e.g. Peshitta) The main purpose of the Göttingen edition is to reconstruct the earliest form of the Old Greek prior to the recensions. Nonetheless, in the process of pursuing this task one may clarify the relation of the Old Greek to other versions such as the Peshitta. In the past, some scholars have seen the Peshitta as a daughter version of the LXX. Factors such as polygenesis in the process of translation, common access to interpretive traditions of Second Temple Judaism, and common Vorlage, can account for agreements besides direct dependence. Only if factors such as these are eliminated as possible reasons for agreement between Peshitta and LXX can dependence on LXX be clearly demonstrated. In a few places in Ecclesiastes, the Peshitta shows sporadic and non-systematic dependence upon the Septuagint.

6. Diachronic Development of Greek (Hellenistic / Byzantine Periods) Lastly, the Göttingen editions may be a valuable source of information for grammatical developments in Hellenistic and Byzantine Greek. For example, the

5  See Gentry, “The Distinctive Aims of the Göttingen Apparatus: Examples from Ecclesiastes — An Edition in Preparation,” 82–86.

100

Peter J. Gentry

manuscripts in the textual transmission of the Septuagint show the development from Classical ἄν to ἐάν in relative clauses.

7. The Early Egyptian Text Group One of the main points of the publication “Ecclesiastes and Jerome’s Trifaria Varietas,” was to characterise, in describing the textual transmission of the Greek Ecclesiastes, what I have identified as the early Egyptian text group.6 The Egyptian text-type is characterised by careful editing to improve the text. I already demonstrated this in 2007.7 Since the Greek translator was so completely devoted to formal equivalence, frequently the resultant translation contravenes good grammar and style in Greek. Deliverance from the disease of “Vaticanitis” – always bowing down before the combination of B–S – was finally granted to me. Moreover, there is good and sufficient reason to consider the early Egyptian group as coming from Hesychius of Alexandria. Felix Albrecht has come to an identical conclusion concerning his research on the same group in the Twelve Prophets.8 The advantage of the critical editions of the Göttingen Septuagint can especially be seen in relation to editions of the Hebrew Bible such as the new Biblia Hebraica Quinta. The first volume of the BHQ was Volume 18: Megilloth. The editor of Ecclesiastes was Goldman. He asked me for help on the text of the Septuagint, but unfortunately, I was only beginning my research and unable to grant him any wisdom. He did visit the Septuaginta-Unternehmen in Göttingen before I was assigned Ecclesiastes and was allowed to look at the collation books for Ecclesiastes. He also had a preliminary copy of what is now the publication of volume 5 of Critique textuelle de l’Ancien Testament by Dominique Barthélemy, completed by Clemens Locher, Stephen D. Ryan, and Adrian Schenker in 2015.9 Goldman cites the Septuagint as a source in three ways. He employs the siglum G when the textual transmission of the Septuagint is fairly unified. When the textual tradition is not unified, he used the siglum G* for what he decided was the original text and GMss for those witnesses in the textual tradition of G that supported what he decided was not the original text.

6  See Gentry, “Ecclesiastes and Jerome’s Trifaria Varietas,” 601–610. 7 Gentry, “Special Problems in the Septuagint Text History of Ecclesiastes.” 8  See Felix Albrecht, “Hesychius of Alexandria (the Exegete),” in Encyclopedia of the Bible and Its Reception 11, (Berlin: De Gruyter, 2015), 994–95; Felix Albrecht, “Die alexandrinische Überlieferung und die Rezension des Hesych von Alexandrien in den Prophetenbüchern der Septuaginta,” in Die Septuaginta. Orte und Intentionen, ed. S. Kreuzer, et al. WUNT 361 (Tübingen: Mohr Siebeck, 2016), 337–62. 9 Barthélemy, Critique textuelle de l’ancien Testament.

The Göttingen Edition of Ecclesiastes

101

Goldman employs the siglum G* in 47 instances in his apparatus. Twenty-two of these correspond to the lemma of the critical edition and 25 do not. These cases are briefly listed as follows: Chapter/Verse Critical Edition G* in BHQ Apparatus 2ː8a καὶ χρυσίον] καί γε χρυσίον G* 2ː12c  τίς ὁ ἄνθρωπος] τίς ἄνθρωπος G* [good one to discuss – stylistic improvement in B] 2ː22a ὅτι τί γίνεται] ὅτι γίνεται G* 2ː24 ἐν ἀνθρώπῳ] τῷ ἀνθρώπῳ G* 2ː23 ὃ φάγεται] εἰ μὴ ὃ φάγεται G* 3ː11a σὺν τὰ πάντα] σύμπαντα G* 3ː11c ὃ ἐποίησεν] ἃ ἐποίησεν G* 3ː16c εὐσεβής] ἀσεβής G 4ː8b καὶ ἀδελφός] καί γε ἀδελφός G 5ː6b ὅτι σὺν] ὅτι σύ G* 5ː18b ὁ ἄνθρωπος] ἄνθρωπος G* 6ː5b ἀνάπαυσις] ἀναπαύσεις G* 7ː1b γεννήσεως αὐτοῦ] om αὐτοῦ G* 7ː2 τοῦ ἀνθρώπου] om τοῦ G* 7ː6 ἀπολλύει] ἀπόλλυσι G* 7ː14a καί γε V Sc L 338 CPA] pr ιδε rel G* 7ː14b καί γε σὺν τοῦτο σύμφωνον τούτῳ] καί γε τοῦτο σύμφωνον τούτῳ G* 8ː4b λαλεῖ] om G* 8ː10b καὶ ἐκ τόπου ἁγίου] καὶ ἐκ ἁγίου G* [8ː12b αὐτῶν] ἀυτοῦ 543 ἀυτῷ Hi (ei) = Gra. Ra. 𝔐 G* Now agree with BHQ] 9ː9b πάσας τὰς ἡμέρας ματαιότητός σου] > G* [Counts as 3 Cases] 10ː5b ὃ ἐξῆλθεν] om ὅ G* 10ː11 ὁ ὄφις] om ὁ G* 11ː9b  ἐν ὁδοῖς καρδίας σου καὶ ἐν ὁράσει ὀφθαλμῶν σου] ἐν ὁδοῖς καρδίας σου ἄμωμος καὶ μἐν μὴ ὁράσει ὀφθαλμῶν σου = G* 12ː13b ἀκούεται] ἄκου 637; ἀκούετε 475 Syh; ἄκουε = G*

We have opportunity to consider six of these cases here. Three of them entail the presence or absence of the article.

2ː12c BHQ Gött

• ‫ ֶ ֣מה ָה ָא ָ ֗דם‬GMss σ′ V Hielem S | τίς ἄνθρωπος G* (hapl) | ‫ מה הנאה אית לגבר‬T ὁ O C′–298 522 798 = 𝔐] om rel (Fa1 GregNy 355,6 Ol Met II.3,3 Syn)

102

Peter J. Gentry

Goldman states: Rahlfs includes the article with the hexaplaric and minority text of Codex Venetus, bringing G nearer to M, but Β 998 S A Complutensis and the majority of Greek MSS. have no article, so G probably did not read an article in its source text, the translator being very careful to render the presence or absence of determiners. V quid est inquam homo translates σ′ τί δὲ ὁ ἄνθρωπος (where δέ is for the initial ‫) ִּכי‬. G*’s Vorlage probably resulted from a haplography of ‫ה‬.10

Jerome may have been influenced by Symmachus, but one cannot assert that Latin witnesses such as the lemma of Jerome’s Commentary or the Vulgate are evidence for articulation in Hebrew. Furthermore, if the absence of the article is original, then M represents a dittography; one cannot say that the Vorlage of G had a haplography. In essence, Goldman is impressed by 998 B-S and A. He does not know how these witnesses fit into the text history nor is he aware that the early Egyptian text is highly recensional and contains many stylistic corrections, a large number of which concern articulation. It is also possible that the early Egyptian text was corrected to a Vorlage that had haplography.

5ː18b BHQ Gött

• ‫ ָה ָא ָ ֡דם‬GMss | > art G* | V S T (indet) ὁ 1° O–637 336′ 698(α) = Ra. 𝔐] > rel (539)

Goldman’s reasoning is as follows: G ὁ ἄνθρωπος, as edited by Rahlfs, is taken from two hexaplaric manuscripts, V and 253. G* read a text without the article. The situation is similar to 3:13 where Old Greek is not Rahlfs’ text but the one without the article. At first sight, the article of Μ is lectio difficilior, particularly in view of the following relative clause in 3:13 and 5:18. However, the problem is probably a literary one and this double variation of 3:13; 5:18 should be approached with a look at 7:2, where G no longer reflects the article before ‫אדם‬.11

10  Y. A. P. Goldman, “Qohelet,” in Biblia Hebraica: quinta editione cum apparatu critico novis curis elaborato 18: General Introduction and Megilloth ed. Robert Althann and Adrian Schenker (Stuttgart: Deutsche Bibelgesellschaft, 2004), 72*. 11  Goldman, “Qohelet,” 87*.

The Göttingen Edition of Ecclesiastes

103

10ː11b BHQ Gött

• ‫ ַהּנָ ָ ֖חׁש‬GMss (spont) | > art G* | pl T | V S (indet) || pref ‫ נָ ָחׁש‬see G* ὁ ὄφις] om ὁ A B-S-68 L–130 C′’–260 d k 338 547 645 766 Fa1 (absc Fa2; sed hab SaI Did Met IX.11,2)

Goldman gives the following explanation: ‫ ַהּנָ ָ ֖חׁש‬G is difficilior, an article being spontaneously added in M. Moreover, G reflects better assonance. The text of Rahlfs betrays once again the tendency to choose the text nearest to M, with a minority of Greek witnesses.12

Once again, in 5:18b and 10:11b, Goldman is influenced by the combination B-S and A and the majority of witnesses. There are examples in Ecclesiastes where corrections of a recensional and stylistic nature in the early Egyptian text have influenced a large part of the textual transmission. The O text is closer to the earliest form of the text in these instances.

11ː09b BHQ Gött

• ‫ּוב ַמ ְר ֵ ֖אי‬ ְ ‫ ְּב ַד ְר ֵכ֣י ִל ְּב ָ֔ך‬V Hielem S | ἐν ὁδοῖς καρδίας σου ἄμωμος καὶ μὴ ἐν ὁράσει ὀφθαλμῶν σου G* (gloss) | ‫ בענותנותא עם ארחי לבך ותהי זהיר בחיזו‬T καρδίας σου 357 542 Hilem et com 348,156 351,68 Damap = Ra Vulg] καρδία σου 788; ἀμώμων ἡ καρδία σου 539; (+ σου SaI II AGeo) ἄμωμος B-68′-998 336′ Fa1 2 SaI II AGeo Did 333,4s 335,26s PsChr Antioch 1485 (ἀμώμως); ἄμωμοις 766; > σου 338*; + ἄμωμα 613; + ἀμώμως 254′; + (∽ Syh) ἄμωμος rel (788 Syh CPA Aeth Arm Arab Ol Met 334,2 Anton 960 1057 1208 Damte Max II 968 Amb Exh virg 10,69 Spec 473,15 Hi Pach 144,13 Ald Sixt) | καί 3°] > 125-261*; + μή B-S*-68′-998 C′–298 571* d 336′ CPA Fa1 2 SaI (absc SaII) Aethte (sed hab Aethap) Arab  ‫ܘܗܠܟ ܒܐܘܪܚܬܐ ܕܠܒܐ ܕܝܠܟ ∽ܕܐܠ ܡܘܡܐ‬ ܿ ‫ ܂ܐ܂ܣ܂ܬ܂‬Index super ‫ܕܝܠܟ‬ Syhmg: ‫ܒܗ ܒܕܡܘܬܐ܀‬

12  Goldman, “Qohelet,” 107*.

104

Peter J. Gentry

The explanation given by Goldman is as follows: ‫ּוב ַמ ְר ֵ ֖אי‬ ְ ‫ ְּב ַד ְר ֵכ֣י ִל ְּב ָ֔ך‬Most of the Greek witnesses reflect an “orthodox gloss”(McNeile,

An Introduction to Ecclesiastes, 167): ἐν ὁδοῖς καρδίας σου ἄμωμος καὶ μὴ ἐν ὁράσει, adding “blameless” after “the ways of your heart“ and a negation before “the sight,” creating the following text: “walk in the ways of your heart blameless and not in the sight of your eyes.” In any case, this must be regarded as Old Greek, probably reflecting a Hebrew Vorlage in view of the very literalistic approach of G Qoh. Rahlfs presents as G the proto-M text via the lemma of Jerome, characterized as “Old Latin.” The word ἄμωμος is under obelos in Syh. M is supported by V S Τ and the three revisions of G according to Syh, where α′-σ′-θ′ are characterized as being “like G.” It is not surprising to find a warning in T’s translation: “walk in humility in the ways of your heart and be careful in the sight of your eyes.”13

The case of 11:9 is complicated by comparison. The words in Syriac for which ἄμωμος is the parent text are preceded by a sign like an S rotated ninety degrees to the right and followed by a metobelus. Called a lemnisk (λημνίσκος or λιμνίσκος, Epiphanius, de Mensuris et Ponderibus, § 8) by Epiphanius,14 it is better known as an antisigma (∽), a sigma rotated counter-clockwise, i.e. minus ninety degrees.15 Descriptions and discussions of the use of this sign in the Syro-Hexapla by Gardthausen, Field, Rahlfs, Ziegler, Munnich and Wevers demonstrate that scholars are entirely unclear and uncertain about the identity and purpose of this critical sign.16 The first accurate explanation of this sign among Septuagint scholars was given in my contribution to the FS for Anneli Aejmelaeus and presented 13 Goldman, “Qohelet,” 109*. 14  Μουτσούλας, Τὸ “Περὶ μέτρων καὶ σταθμῶν” ἔργον Ἐπιφανίου τοῦ Σαλαμῖνος. 15  Wevers, not recognising the sign as antisigma, says it is ninety degrees to the right (see John W. Wevers and Detlef Fraenkel, Studies in the Text Histories of Deuteronomy and Ezekiel [MSU 26; Göttingen: Vandenhoeck & Ruprecht, 2003], 100). The earliest form of the sign is a backwards lunate sigma. Since the form in Syh is based on the later minuscule sigma, whether the sigma is rotated clockwise or counterclockwise, the result would be the same. Neither Ziegler nor Wevers recognised the form as an antisigma. Epiphanius appears to confuse the sign with what we know as an obelus. 16  Most scholars devote only two or three sentences to this topic. See Victor Gardthausen, Griechische Palaeographie (2nd ed.; Leipzig: Veit, 1913), 2:414; Frederick Field, Origenis Hexaplorum quae supersunt, 2 vols. (Oxford: Clarendon Press, 1875), lxiv-lxvii; Alfred Rahlfs, Studie über den griechischen Text des Buches Ruth (MSU 3; Berlin: Weidmannsche Buchhandlung, 1922); Joseph Ziegler ed., Ezechiel, Septuaginta Vetus Testamentum Graecum 16.1 (2nd ed.; Göttingen: Vandenhoeck & Ruprecht, 1952, 1977), 41–44; Joseph Ziegler ed., Susanna, Daniel, Bel et Draco, Septuaginta Vetus Testamentum Graecum 16.2 (2nd ed.; Versionis iuxta LXX interpretes textum plane novum constituit O. Munnich, Versionis iuxta “Theodotionem” fragmenta adiecit D. Fraenkel (Göttingen: Vandenhoeck & Ruprecht, 1999), 41–42; see Wevers and Fraenkel, Studies in the Text Histories, 100.

The Göttingen Edition of Ecclesiastes

105

at SBL in Chicago in 2012.17 It derives from the signs developed by the Alexandrian grammarians and librarians as they constructed a scholarly text of Homer. This antisigma normally marks a misplaced text and I demonstrated that all the cases in the Syro-Hexapla of Ecclesiastes conform to this function. Thus, the Aristarchian sign in the Syh indicates ἄμωμος is a word found in several places in various witnesses: a sure sign of something secondary. The Index in Syh is placed over the word ‫ ܕܝܠܟ‬to show that the hexaplaric text to which the texts of all of the Three are identical does not have ἄμωμος. The author of the marginal note is aware that there is a variation in the text-tradition which omits ἄμωμος. A few early witnesses have ἄμωμος instead of καρδίας σου (B′’ 998 Did alii). There is a difference, however, in that almost no witnesses have preserved precisely this variant text. Yet in the new critical edition three minuscules do not have ἄμωμος. Of these 357 and 788 are witnesses extremely close to the root of the genealogical tree and therefore are of far greater weight than B-S and A. It is a perilous task to construct an edition of the Hebrew Bible without a Göttingen Edition. MT 9:9

9:9  Error in Hebrew or LXX Textual Transmission

‫ל־יְמי ַחּיֵ י ֶה ְב ֶלך‬ ֵ ‫ר־א ַה ְב ָּת ָּכ‬ ָ ‫ם־א ָּׁשה ֲא ֶׁש‬ ִ ‫ְר ֵאה ַחּיִ ים ִע‬ ‫ל־יְמי ֶה ְב ֶלָך‬ ֵ ‫ן־לָך ַּת ַחת ַה ֶּׁש ֶמׁש ָּכ‬ ְ ‫ֲא ֶׁשר נָ ַת‬ ‫ִּכי הּוא ֶח ְל ְקָך ַּב ַחּיִ ים‬ ‫ר־א ָּתה ָע ֵמל ַּת ַחת ַה ֶּׁש ֶמׁש‬ ַ ‫ּוב ֲע ָמ ְלָך ֲא ֶׁש‬ ַ

Göttingen Ecclesiastes 9:9

ἰδὲ ζωὴν μετὰ γυναικός, ἧς ἠγάπησας, πάσας τὰς ἡμέρας ζωῆς ματαιότητός σου τὰς δοθείσας σοι ὑπὸ τὸν ἥλιον, πάσας τὰς ἡμέρας ματαιότητός σου, ὅτι αὐτὸ μερίς σου ἐν τῇ ζωῇ σου καὶ ἐν τῷ μόχθῳ σου, ᾧ σὺ μοχθεῖς ὑπὸ τὸν ἥλιον.

17  Peter J. Gentry, “The Aristarchian Signs in the Textual Tradition of LXX Ecclesiastes.” In In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus ed. by Kristin De Troyer, T. Michael Law, and Marketta Liljeström (Contributions to Biblical Exegesis and Theology; Leuven: Peeters, 2014), 463–478.

106

Peter J. Gentry

————— A B C S 998 O C′’–299 b d–342 k min verss ————— 9 init Fa Hi = Ra] pr και rel (Hi Mal 2) = Pesch | ἰδέ] ειδε 637 706 Did 274,5; vive Hi te ap Mal 2 (sed hab vide Hilem 327,209); perfruere Hilem 327,209 = Vulg ↓ | ζωήν] ζων 357; + tuam Fa | μετά] επι 125′ | γυναικός] + σου Arm | ἧς ἠγάπησας] > 147-159-560; + και 296′ | ἧς] ην OlΜ; > 295-752 | πάσας 1°] ∩ 2° k OlΒΗ; > 425 | τὰς 1° A S 637 C′’–390 425 130 254 155 161 248c 260′ 296′ 311 339 359 547 645 698 706 795 Did 277,1 GregAg 1073 OlΑΓΔΕΖΙΚ = Ald Sixt] absc 998; > B C O–637-411 390-425 L–130 357-754 68 125II 248* 252 336′ 338 443 534-602-613 542 543 549 766 | ζωῆς] pr της GregAg 1073; > 336′ LatHi Mal 2 Hite; + tuae Fa | ματαιότητός 1°] νεοτητος C′: ex Prov 518b; νηπιοτητος 698 | txt σου 1° ∩ 2° 253 797txt L 252txt 359 543 OlΖ = Pesch | τὰς δοθείσας] της δοθεισης 336′; αι εδοθησαν 637-411 338 542 GregAg 1073 | ἥλιον 1° ∩ 2° 766I | πάσας τὰς ἡμέρας ματαιότητός σου C S V-411 C′(797mg) 125II 155 161 248 252mg 260′ 296′ 311 336′ 339 443 542 549 645 698 706 766II 795 Did 277,1–2 OlΕΙΜ Arm Hi SaII Syh = Ald] πασας τας ημερας ματαιοτητας σου k; πασαι ημεραι ημεραι (> 534 Fa1.2) ατμου σου B* 998 534 Fa1.2 ↓ ; > A Bc 637 797txt cII d 68 338 547 602-613 OlΑΓΔΖmgΚ = Sixt | τὰς 2° V C′(797mg) 161 248 260′ 296′ 311 336′ 339 645 698 706 795] > C S 411 125II 155 252mg 443 542 549 766II OlΕΙΜ | ματαιότητός 2°] pr της 411 147; νεοτητος 155: ex Prov 518b | αὐτό] αυτω 637 390 OlΑΖ; αυτη 125 766II; αυτος S; + est Arm Hi | σου 3° – σου 4°] ζωης σου 797 | om σου 3° Didlem 278,4 | om τῇ C 637 645 | ζωῇ] ζωης OlΑ: dittog | om σου 4° Hi = Vulg | om καί Fa | om ἐν 2° S 645 | om τῷ k | om σου 5° 998 411 k 336′ Didcomm 278,9 Hi = 𝔐 | ᾧ σὺ μοχθεῖς] ο εμοχθης 637 | ᾧ] pr εν Sc; ο 637 147; ος 698 | σύ] σοι 563-571* 548 Didlem et comm 278,6 278,9; > A 797 443 728 OlΑ Arm Fa = Vulg | μοχθεῖς] -θης 296′ 336′ 359; -θις 998 V; προεμοχθησας GregAg 1073 | ἥλιον 2° ∩ ἥλιον 911a 357 ————— 9 ἰδὲ ζωήν] σʹ ἀπόλαυσον ζωῆς 161 248 | (9b) πάσας τὰς ἡμέρας ζωῆς ματαιότητός σου] α′ ἀτμοῦ σου (πᾶσαι αἱ ἡμέραι) 161 248 (ind ad ματαιότητός 2°) | (9d) πάσας τὰς ἡμέρας ματαιότητός σου] αʹ πᾶσαι αἱ ἡμέραι (+ ἡμέραι 248) ἀτμοῦ σου 161 248 (ind ad ματαιότητός 2°) | ματαιότητός 2°] οἱ λʹ ἀτμοῦ 252

Before considering 9:9, a brief discussion of the actual readings in MSS 161, 248 and 252 in the Second Apparatus is necessary. The marginal note for each manuscript is as follows: 161 248

α′ ἀτμοῦ σου πᾶσαι αἱ ἡμέραι ἀτμοῦ σου (Ind ad ματαιοτητος 2°) α′ ἀτμοῦ σου πᾶσαι αἱ ἡμέραι ἡμέραι ἀτμοῦ σου (Ind ad ματαιοτητος 2°)

252txt is lacking 9:9cd. This is supplied in the margin opposite 9:9b. In the margin opposite 9:9e and below the previous marginal note are the words οἱ λο ἀτμοῦ.

The Göttingen Edition of Ecclesiastes

107

Marshall, in his provisional critical edition of hexaplaric remains for Ecclesiastes, has observed rightly that the marginal notes of 161 and 248 constitute in fact two distinct readings, one corresponding to 9:9b and the other to 9:9d, even though the index is tied only to ματαιότητός in 9:9d.18 His suggestion, however, that the repetition of ἡμέραι in 248mg is due to dittography is probably not right.19 In a discussion of the problem with Detlef Fraenkel he proposed more plausibly that the doubled ἡμέραι in 248mg, also in B* 998, and secondarily removed in 161mg and 534 Fa1.2 points to full Aquila readings for both 9:9b and d. This is how the information has been represented above in the Second Apparatus. The main issue in 9:9 may now be considered. This last problem is difficult and perplexing. There certainly is corruption in the textual transmission, but it is not certain whether the corruption occurred already in the Hebrew Vorlage of the Greek Translator or in the course of textual transmission after the Greek Translation was made. Y. A. P. Goldman, editor of Qoheleth for Biblia Hebraica Quinta argues for the former position as follows: ‫ ֲא ֶׁשר‬2  to ‫ ֶה ְב ֶלך‬2  S [Syriac] attests the shortest form of the verse, without v. 9aβ. That part is attested in G, but in a dubious way: (1) τὰς δοθείσας for ‫אשר נתן‬, as indicat-

ed by McNeile (An Introduction, 150), is “foreign to the style of the translation”; (2) the translation of the last three words (‫ )כל ימי הבלך‬πάσας (τὰς) ἡμέρας ματαιότητός σου is omitted in Codex Alexandrinus and part of the medieval manuscripts; and they are imported (in the nominative) from the Aquila version in the text of B and 998. The omission in S and in G or their Vorlage might have resulted from an accidental parablepsis due to homoioteleuton, but on the other hand that part of the verse is rather redundant. ‫ל־יְמי ֶה ְב ֶלך‬ ֵ ‫ ָּכ‬Codex Alexandrinus and a majority of witnesses of the Greek tradition do not have these words, which in Codex Vaticanus (first hand) and 998 seem to have been imported from the Aquila version.20

In order to deal fairly with the analysis proposed by Goldman, the arguments proposed by McNeile need to be cited in full: v. 9. καὶ ἴδε MSS. S.H. Pesh | M. ‫ ראה‬Σ Hier. τὰς δοθείσας … ἀτμοῦ σου. It is probable that the two clauses (a) "‫השמש ת" ל" נ‬ ‫אשר‬, and (b) ‫כל ימי הבלך‬, were absent from the pre-Aḳ iban text used by the translator. In the case of (b) this amounts to a certainty.

18 Marshall, “A Critical Edition of the Hexaplaric Fragments of Ecclesiastes,” 265. 19  Although made in consultation with myself (as his supervisor) at the time. 20 Goldman, “Qohelet,” 104*.

108

Peter J. Gentry

For: (i) B (alone) has πᾶσαι ἡμέραι ἡμέραι (sic) ἀτμοῦ σου, which is clearly a corruption of Aq. πᾶσαι αἱ ἡμέραι ἀ. σ. (ii) In CSV 147-157-159. 161-248. 296 S.H. a literal rendering has been supplied from the earlier similar clause in the verse – πάσας [τὰς] ἡμέρας [τῆς] ματαιότητός σου. (iii) The clause is omitted in A curss. rel. Pesh. Hier. Tg. and in some Heb. MSS. K. and de R. Clause (a) is found in all Greek MSS. (exc. 106-261), and in S.H, but Pesh. om. The rendering τὰς δοθείσας is foreign to the style of the translation, and may have been supplied from Σ or Θ, as clause (b) was from Aq. It is omitted, together with the first clause, in ten MSS. K. de R. ἐν τῇ ζώῃ σου MSS. S.H. | om. pron. M. Pesh. Hier. (om. pron. with μόχθῳ also). This may, however, be a mistake of a Greek scribe, owing to the occurrence of σου with nine other words in vv. 7–9.21

Parts of this complicated problem can be resolved through exhaustive study of the translator’s technique. First, at the beginning of v. 9, for the Göttingen Edition, I am proposing to follow Rahlfs in reconstructing an original text without the conjunction καί. This is based only on the Fayyumic Version and the lemma of Jerome’s Commentary. The Fayyumic is inconsistent in rendering καί. Possibly the lemma of the Commentary by Jerome is not a correction to the Hebrew here. This evidence is weak. From the perspective of translation technique, it is possible that the translator rendered asyndeton or zero in his parent text by καί (e.g. 1:8c, 3:21a, 8:9a). Normally, however, waw is rendered by καί and zero by zero in Greek. The correspondence is 99%. Also, from the point of view of stylistic structure, we have a series of proverbs each marked at the onset by asyndeton (v. 7a, 8a, 9a, 10a) and by a copula conjunction in between, e.g. καὶ πίε v. 7b. The combination of stylistic structure and translation technique form a strong argument, while the addition of a καί is the lectio facilior – an inner Greek corruption that doubtless occurred very early. Second, McNeile followed by Goldman claim that rendering ‫ן־לָך‬ ְ ‫ ֲא ֶׁשר נָ ַת‬by τὰς δοθείσας σοι is “foreign to the style of the translation.” Yet study of translation technique shows that the Greek translator employed a participle for qatal

21 McNeile, An Introduction, 150.

The Göttingen Edition of Ecclesiastes

109

in 18 instances22 and for yiqtol in eight,23 all in relative clauses.24 The majority of these cases involve the verb ‫היה‬, but the construction itself is well within norms for the style of the translator. This shows the importance of exhaustive analysis of translation technique and not relying on Probeschriften for results. Goldman and McNeile, then, have no reason on the basis of translation technique to suggest that 9c is secondary and not part of the original translation. What about the arguments of Goldman and McNeile based on external evidence that 9:9c and d were not in the original Greek translation? Most of the evidence recorded for the Göttingen Edition was available to both of them. This case is a good opportunity to show how to read the Apparatus of the Göttingen Edition. One must eliminate from consideration minor mistakes and the issue of articulation. This leaves the following apparatus: 9 πάσας 1°] ∩ 2° k OlΒΗ; > 425 | σου 1° ∩ 2° 253 797txt L 252txt 359 543 OlΖtxt = Pesch | ἥλιον 1° ∩ 2° 766I | πάσας τὰς ἡμέρας ματαιότητός σου C S V-411 C′(797mg) 125II 155 161 248 252mg 260′ 296′ 311 336′ 339 443 542 549 645 698 706 766II 795 Did 277,1–2 OlΕΙΜ Arm Hi SaII Syh = Ald] πασας τας ημερας ματαιοτητας σου k; πασαι ημεραι ημεραι (> 534 Fa1.2) ατμου σου B* 998 534 Fa1.2 ↓ ; > A Bc 637 cII d 68 338 547 602-613 OlΑΓΔΖmgΚ = Sixt | σου 3° – σου 4°] ζωης σου 797 | ἥλιον 2° ∩ ἥλιον 911a 357

Omissions and parablepses are noted at five different locations in the apparatus: Omitting 9b and 9c: Omitting 9c and 9d: Omitting 9d: Omitting 9def: Omitting 9f – 11a:

k OlΒΗ 253 797txt L 252txt 359 543 OlΖtxt = Pesch mg A Bc 637 cII d 68 338 547 602-613 OlΑΓΔΖ Κ = Sixt 766I 357

Since the reading of B* 998 534 Fa1.2 is clearly a correction added from Aquila, the archetype of these witnesses also omitted 9d. Note carefully the witnesses that attest to 9:9c and d: C S V-411 C′(797mg) 125II 155 161 248 252mg 260′ 296′ 311 336′ 339 443 542 549 645 698 706 766II 795 Did 277,1–2 OlΕΙΜ Arm Hi SaII Syh = Ald

22  1:9 (bis), 1:10, 1:13, 1:14, 2:7, 2:9, 2:17, 3:15, 4:2, 4:3, 5:8, 7:19, 8:16, 8:17, 9:6, 9:9, 11:8. 23  1:9 (bis), 1:11 (bis), 2:18, 8:7, 10:14 (bis). 24  For complete discussion, see Yun Yeong Yi, “Translation Technique of the Greek Ecclesiastes” (PhD diss., Southern Baptist Theological Seminary, 2005), 151–154, 167–171.

110

Peter J. Gentry

McNeile’s supposition that 9c is secondary is groundless and his argument that 9d is supplied from 9b is not sound. The O Group is V-253-637 and sometimes 411. MS 542, closely related to Syh also attests the fifth column of the Hexapla. Since V-411-542 and Syh have 9c and d, this is surely the text received by Origen in his fifth column before filling in the minuses vis à vis the Hebrew from the Three. This is as early a witness as 998 which attests the parablepsis. The argument of Goldman and McNeile that the original text of LXX or even the Hebrew Vorlage omitted 9c and 9d is simplistic and not supported by the complex of parablepses and omissions in the textual tradition in Greek. The best explanation accounting for the data must assume a Hebrew parent text for LXX identical to MT and parablepses occurring in the textual transmission of the LXX. First, a parablepsis resulted in the omission of 9c and d. Second, a partial correction added only 9c. The precursor of 998 and B* added 9d from Aquila. mg Only one manuscript has 9c as a correction, i.e. OlΖ and this is certainly secondary. Therefore, since no serious witness has the text of 9c as a correction, one cannot explain the omission of only 9d. This is the crux criticorum of the thesis of Goldman and McNeile (i.e. that which completely crucifies it). From examples such as these, then, sound methodology in Septuagint text criticism must involve more than counting manuscripts or even weighing important witnesses. Knowledge of translation technique and issues in the text history play a crucial role in the decisions made. Hopefully, these examples also show how the economic format recording the information in the Göttingen apparatus can be read and interpreted.

12ː13b 12:13

‫סוף דבר הכל נׁשמע‬

τέλος λόγου τὸ πᾶν ἀκούεται ἀκούεται (ακου 637) 998 O 359] ἄκουε B′’ S A 411 C C′’ 3 L d k 125II 149 161 248 252 260 296 311 336 338 339 357s 443 542 543 547 548 549 602 613 645 698 706 728 766I 766II 795 Sa = Sixt ‫ܫܘܠܡܐ ܟܠܗ ܡܠܬܐ ܫܡܥܘ܂‬ Syhmg: ‫ ܂ܐ܂ܬ܂ ܒܕܡܘܬ ܫܒܥܝܢ܀‬Index super ‫ܫܡܥܘ‬

The final instance is like that of the previous case in that according to the marginal note in Syh, only two of the Three have texts identical to that of ο′, but this time it is α′ and θ′ instead of α′ and σ′. Presumably the text of σ′ is different, as one might well expect, but it is not supplied. What is interesting here is the reading the text of Syh – a m p Imperative of ‫ܫܡܥ‬. This is clearly based on a parent text reading ἀκούετε which in turn is an

The Göttingen Edition of Ecclesiastes

111

obvious etacistic corruption of ἀκούεται. The Hexaplaric group uniformly attests to ἀκούεται and this is also likely original, as it correctly renders the 3rd person masc. singular Niphal perfect in the Hebrew. The 2nd person plural Imperative found in the majority of witnesses is secondary. Once again, the marginal note makes sense if it originated in a non-Hexaplaric text. These examples are not intended to fault the work of Goldman, but only to show that preparing an edition of Biblia Hebraica Quinta for a book that has no critical edition of the Septuagint is a perilous task. Finally, a list of differences between the Handausgabe of Rahlfs and the Göttingen edition show about 72 instances improving the text.25 Some of these are not significant, but in a good number of places, a better knowledge of the text history and how to assess our known witnesses has resulted in a better text.

8. Bibliography Althann, Robert and Adrian Schenker, Biblia Hebraica: quinta editione cum apparatu critico novis curis elaborato 18: General Introduction and Megilloth (Stuttgart: Deutsche Bibelgesellschaft, 2004). Barthélemy, Dominique, Critique textuelle de l’Ancien Testament, eds. C. Locher, Stephan D. Ryan, and Adrian Schenker. OBO 50.5 (Göttingen: Vandenhoeck & Ruprecht, 2015). Gentry, Peter J. “Ecclesiast,” in A New English Translation of the Septuagint, eds. Albert Pietersma and Benjamin G. Wright (Oxford: Oxford University Press, 2007). —, “Special Problems in the Septuagint Text History of Ecclesiastes,” in XIII Congress of the International Organization for Septuagint and Cognate Studies: Ljubljana, 2007, ed. Melvin K. H. Peters. Society of Biblical Literature Septuagint and Cognate Studies 55 (Atlanta, Georgia: Scholars Press, 2008), 133–153. —, “The Distinctive Aims of the Göttingen Apparatus: Examples from Ecclesiastes – An Edition in Preparation,” in Die Göttinger Septuaginta: Ein editorisches Jahrhundertprojekt, eds. Reinhard G. Kratz and Bernhard Neuschäfer. Abhandlungen der Akademie der Wissenschaften zu Göttingen, NF 22, Mitteilungen des Septuaginta-Unternehmens 30 (Berlin: De Gruyter, 2013), 73–105. —, “Post-Partum Reflections on Preparing a Critical Edition of LXX Ecclesiastes.” Paper Presented at the Annual Meeting of the Society of Biblical Literature, San Antonio, Texas, November 20, 2016. —, Ecclesiastes. Septuaginta. Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum 11.2 (Göttingen: Vandenhoeck & Ruprecht, 2019). —, “Ecclesiastes and Jerome’s Trifaria Varietas,” in Die Septuaginta. Themen, Manuskripte, Wirkungen, ed. Eberhart Bons, et al. WUNT 444 (Tübingen: Mohr Siebeck, 2020), 601–610.

25  This list will be given in text history of the Greek Ecclesiastes.

112

Peter J. Gentry

—, Text History of the Greek Ecclesiastes. DSI (Göttingen, Vandenhoeck & Ruprecht, forthcoming). Givón, Talmy, Syntax: An Introduction, 2 vols. (Philadelphia: John Benjamins, 1984, 1990). McNeile, Alan Hugh, An Introduction to Ecclesiastes (Cambridge: University Press, 1904). Marshall, Phillip S., A Critical Edition of the Hexaplaric Fragments of Ecclesiastes (PhD diss., Southern Baptist Theological Seminary, 2007). Μουτσούλας, Ἠλίας Δ. Τὸ “Περὶ μέτρων καὶ σταθμῶν” ἔργον Ἐπιφανίου τοῦ Σαλαμῖνος. Εἰσαγωγή, Κριτικὴ ἔκδοσις, Σχόλια (Athens, 1971), 149–151.

Felix Albrecht The Göttingen Edition of the Psalms of Solomon The Psalms of Solomon (PsSol) is the common name for a collection of 18 Psalms, largely from the first century BCE, that have been transmitted in the context of the Septuagint under the name of King Solomon.1

1. Editorial History2 The Psalms of Solomon have a long editorial history. The Greek version was first published in 1626 by the Spanish Jesuit Juan Luis de la Cerda, who had made a name for himself as a commentator of Virgil. In 1713, the edition of de la Cerda was reprinted by Fabricius. This was followed by eleven other editions, attesting to exceptional interest in the text: In 1869, an edition was published by Adolf Hilgenfeld, which took up the conjectures of Paul Anton de Lagarde, in 1871 the two editions of Otto Fridolin Fritzsche and Eduard Ephraem Geiger were published, in 1883 that of Bernhard Pick, in 1891 that of Herbert Edward Ryle and Montague Rhodes James, in 1894 that of Henry Barclay Swete in his concise edition of the Septuagint and in 1895 that of Oscar von Gebhardt. In 1911, Joseph Viteau published a revision, and in 1935 Alfred Rahlfs published his Handausgabe, which is – in the case of the Psalms of Solomon – essentially based on the critical edition of von Gebhardt, as was the 1978 edition by Sabbas Agourides. In 2007, Robert Bradley Wright published a deficient edition of the text.3 Finally, in 2018, the Göttingen edition of the Psalms of Solomon was published. In 2020, a separate volume with German translation, short notes and indices appeared. The Göttingen edition opens with an introduction that is based on the well-proven model of the Göttingen Editio critica maior. It has, however, two peculiarities, due to the peculiar nature of the Psalms of Solomon: its focus is, firstly, on the linguistic features of the Greek tradition in light of the dating of the textual witnesses and, secondly, on the textual history of the collection.

1  For a brief overview of the Psalms of Solomon, see e.g. Albrecht, Psalmen Salomos, 3–17. 2  For an extensive editorial history, cf. Albrecht, Psalmi Salomonis, 29–33 (‘Die Druckausgaben’); and Albrecht, “Notwendigkeit,” 110–112 (‘Zur Editionsgeschichte’). 3 Wright, Psalms of Solomon. For a critical evaluation of Wright’s edition see Albrecht, “Notwendigkeit.”

114

Felix Albrecht

2. Editorial Challenges4 It is assumed that the Psalms of Solomon were composed in Hebrew, but only the Greek tradition has been preserved. Later, a Syriac translation branched off from the Greek tradition. The Greek might have been based on a Hebrew Vorlage in its oldest nucleus, which, however, can hardly be determined with certainty. Unlike most books in the Septuagint, no comparison is possible with a Hebrew parent text. This unique situation makes the critical constitution of the text a special challenge. For those books of the Septuagint, which are not written genuinely in Greek but whose Vorlage is known or can be reconstructed, the critical constitution of the oldest attainable text, due to the textual history of the Septuagint and its transmission, is based on three interrelated steps: 1. Establishing the Greek text on the basis of the Greek tradition, which includes the primary tradition (Greek manuscripts), the secondary tradition (daughter versions) and the tertiary tradition (patristic citations in Greek and Latin). 2. Comparing the Greek text with the Hebrew tradition in the form of the primary tradition (Hebrew manuscripts) and secondary tradition (daughter versions). The relationship between Greek and Hebrew must be determined through an analysis of the translation technique, the manner of which vacillates between formal and functional correspondence to the Vorlage. 3. Identifying readings stemming from Greek recensions (Jewish and Christian) and, subsequently, eliminating those elements linked to recension. In the case of the Psalms of Solomon the quantitative range and the qualitative nature of the tradition are limited: 1. In quantitative terms, the edition must focus, according to the historical circumstances of transmission, on the Greek tradition, which consists of the Greek manuscripts and the Greek secondary tradition – namely, the Syriac daughter version. The Hebrew tradition and any further reception of the Greek are missing. There are also no detectable characteristics of recension present in the Greek text. 2. In qualitative terms, the tradition is poor. The text of the Greek Bible has been revised several times. For most canonical books, traces of such a revision or recension are detectable and associated with Origen, Hesychius and Lucian. These revisions or recensions were based on different motivations: while Origen sought to develop a philologically secured text for biblical exegesis, Hesychius and especially Lucian seem to have been focused on 4  Section 2 (‘Editorial Challenges’) is essentially based on the edition’s prolegomena, for which cf. Albrecht, Psalmi Salomonis, 3–9 (‘Prolegomena’).

The Göttingen Edition of the Psalms of Solomon

115

the edition of the biblical text for ecclesiastical or liturgical purposes. As for the deuterocanonical books, a revision can be detected only with reservations. The parabiblical scriptures are nearly free of standardization: the lower the canonical reputation, the lower is the subsequent probability that an authoritative edition of that book influenced the tradition, as, for instance, the Antiochene edition did in the canonical books of the Old and New Testament.

3. The Psalms of Solomon: Textual History and Linguistic Features The influence of Origen’s work on the text of the Greek Bible can hardly be overestimated: Lucian, for example, seems to depend grosso modo on Origen. The books of the Septuagint written and composed in Greek, such as the Wisdom of Solomon, had a special position in their treatment by Origen. The influence of recension can almost be ruled out, as Origen likely left these books unedited, probably because of the absence of a Hebrew Vorlage. The same seems true with a high degree of probability for the Psalms of Solomon. In its Greek tradition, an influence of recension is not detectable; moreover, the tradition even appears quite ‘feral’, so that we can safely exclude the possibility of an early diorthosis of the Greek text – for example, at the hands of Origen or Lucian. This lack of diorthosis is very fortunate, since it allows the study and reconstruction of the original linguistic shape of the Psalms of Solomon. The text has been composed and transmitted in Koine Greek, a language variety of such simplicity that many scholars have simply dismissed the work altogether. Felix Perles, for example, disparaged the work as literarily ‘worthless’. Nevertheless, it is precisely the linguistic shape of the Psalms of Solomon that is peculiar and remarkable: even though almost one thousand years separate the archetype and the oldest Greek textual witnesses, not only can the light of inquiry illuminate the darkness of the – heretofore largely unexplored – textual history, but it is also possible to reconstruct cum grano salis the oldest attainable linguistic shape of the text of the Psalms of Solomon, which can be dated – in my estimation – to around 50 CE. This is the starting point for the subsequent tradition, for which the principal points are the following: on the one hand, Hyparchetype β, which may have arisen at the earliest in the 2nd half of the 3rd century CE, as it testifies to the composition of Psalms of Solomon and Odes of Solomon, and, on the other hand, Hyparchetype γ.5 It is of further interest that Hyparchetype ε, depending on Hyparchetype γ, must be an ancestor of Codex Alexandrinus.

5  For the stemma codicum, cf. Albrecht, Psalmi Salomonis, 180, and Albrecht, Psalmen Salomos, 142 (slightly revised).

116

Felix Albrecht

3.1 Hyparchetype ζ and Codex Alexandrinus

It is important to bear in mind that Codex Alexandrinus was part of the former transmission of the Psalms of Solomon. The resemblance of the nomina sacra (ms. 769, and also ms. 336, compared to Codex Alexandrinus) and the special form of the paragraphos a coda ondulata (ms. 336, compared to Codex Alexandrinus) are striking.6 The Psalms of Solomon were originally placed at the end of the codex. Unfortunately, the final part of the manuscript disappeared before the 14th century CE. However, how does Codex Alexandrinus fit into the transmission history of the Psalms of Solomon? The origin of Codex Alexandrinus is usually located in Alexandria, around 450 CE. The further history of the manuscript is largely unknown, but Skeat assumes that the codex was in Constantinople in Byzantine times. In any case, he supposed that the Alexandrian Patriarch Athanasius II (sic!)7, who resided in Constantinople, acquired the manuscript sometime between 1278 and 1308 in Constantinople and brought it to Alexandria, from where it came, again via Constantinople, to London in 1627/28.8 Of course, the assumption that Codex Alexandrinus was in Constantinople before the 14th century suits my hypothesis that Codex Alexandrinus (which, before the 14th century, comprised two volumes) coincides with or should, for stemmatic reasons, be set immediately before Hyparchetype ζ (450 CE). Hyparchetype ζ is the ancestor of two manuscripts that fit this picture well: ms. 336 (copied in Cyprus in the 14th century – on the way, so to speak, from Constantinople to Alexandria) and ms. 769 (which belonged to the environment of the Byzantine State Chancellery). Is it too conjectural to assume that the final part of Codex Alexandrinus, lost today, circulated separately at least since the 11th century, if not earlier? This would explain well why ms. 769, copied in Constantinople in the year 1060, 6  On the nomina sacra in the Psalms of Solomon, cf. Albrecht, Psalmi Salomonis, 272–275. For the paragraphos a coda ondulata in Codex Alexandrinus (and ms. 336), cf. Albrecht and Matera, “Testimonianze,” 11–23. 7  “Usually called Athanasius III […]” (Skeat, “Provenance,” 233 n. 3; cf. A. Failler, “Le séjour d’Athanase II d’Alexandrie à Constantinople,” REByz 35 [1977]: 43–71, 43 n. 1). On this patriarch, see A. von Gutschmid, “Verzeichniss der Patriarchen von Alexandrien,” in Kleine Schriften vol. 2, ed. F. Rühl, Leipzig: Teubner, 1890, 395–525: 490–491 (Nr. 72, Athanasios III.); furthermore G. Fedalto, Hierarchia ecclesiastica orientalis. Series episcoporum ecclesiarum christianarum orientalium vol. 2, Patriarchatus Alexandrinus, Antiochenus, Hierosolymitanus, Padua: Edizioni messaggero, 1988, 584 (Athanasius, 1276–1316). – An Arabic curse at the beginning of Codex Alexandrinus, which is written by the hand of “Athanasius the humble”, who is identified with Patriarch Athanasius II by Skeat, “Provenance,” 234–235, or with Patriarch Athanasius III by Spinka, “Acquisition,” 26, is the only historical evidence. 8  Cf. Skeat, “Provenance,” 234–235. The ms. was donated to the King of England in 1627, and came to England in 1628. For the history of Codex Alexandrinus, cf. Albrecht, “Überlie­ ferung,” 343–344; Albrecht and Matera, “Testimonianze,” 11–15. Following Skeat, I assume that Codex Alexandrinus originated in Egypt, but soon after came to Constantinople.

The Göttingen Edition of the Psalms of Solomon

117

transmits the Psalms of Solomon in a strange codicological context – namely, after a psalter catena. This would also explain why ms. 336, copied in Cyprus in the 14th century, likewise attests a special codicological position for the Psalms of Solomon – namely, not following the Wisdom of Solomon, which was its original place in the canon of the Septuagint, but following Sirach.9

3.2 Hyparchetype η and the Bible of Niketas

Another cornerstone is Hyparchetype η, which derives from Hyparchetype ε. Hyparchetype η can be dated for good reasons to around 535 CE, during the reign of Emperor Justinian the Great. The Bible codex, as one can imagine, must have been ornately set with illustrations, the Vorlage of the so-called ‘Bible of Niketas’, a 10th-century copy of the Justinian Bible. One part of Niketa’s Bible is ms. 260, better known in Septuagint scholarship as ‘Catena Hauniensis’, a manuscript to which I will return later. For Hyparchetype ε, a tendency of atticising correction can be observed. In my edition, I have evaluated and described this atticising correction in full detail. In many places, it covers over the ancient language of the Psalms of Solomon. In the critical reconstruction of the oldest attainable text, such correction must be removed, and for this purpose, I have identified two categories:10 1. In its patterns of declension, a koine-typical metaplasm appears.11 For instance, in the case of πᾶς, the koine-typical acc. sg. masc. πᾶν is found: πᾶν ἄνδρα (PsSol 3:8), πᾶν σοφόν (PsSol 8:20).12 Von Gebhardt, however, rejects the koine-form πᾶν, as it occurs only in ms. 253; Rahlfs, on the other hand, rightly picks it up in his Handausgabe.13

9  For the original place of the Psalms of Solomon in the canon of the Septuagint, cf. Albrecht, Psalmi Salomonis, 236–241, esp. 236–237. 10  Cf. Albrecht, Psalmen Salomos, 5. Also, in the area of syntax, one can observe the phenomenon that a koine-typical construction is replaced by a common construction, cf. Albrecht, Psalmi Salomonis, 123–132 (“Syntax”). For example, PsSol 5:10: It is typical of the Psalms of Solomon, as of the Septuagint, that possessive suffixes in the plural are constructed with nouns in the singular; here, the original πρόσωπον αὐτῶν has been corrected in ms. 260 and its descendants to πρόσωπα αὐτῶν (cf. Albrecht, Psalmi Salomonis, 124). Another example is PsSol 8:30: A linguistic peculiarity of the Psalms of Solomon is the construction of ἔθνος in the neuter plural with the plural of the predicate; here, the original καταπίωσιν […] ἔθνη is preserved by ms. 253, whereas the other witnesses have corrected it into καταπίῃ […] ἔθνη (Albrecht, Psalmi Salomonis, 124). 11  Cf. Albrecht, Psalmi Salomonis, 88–89. 12  Another example (cf. Albrecht, Psalmi Salomonis, 88–89): the nouns ὁ ζῆλος and ὁ ἔλεος appear as τὸ ζῆλος and τὸ ἔλεος (PsSol 2:24; 14:9). 13  von Gebhardt, Ψαλμοὶ Σολομῶντος, 99. 112; Rahlfs and Hanhart, Septuaginta II:474. 479.

118

Felix Albrecht

2. In its patterns of conjugation, koine-typical forms appear.14 One of Koine’s peculiarities is to transform the endings of the 3rd pers. pl. ind. of the sigmatic aorist to the strong aorist and the imperfect. For instance, for τίλλω (‘to pluck’): the imperfect ἐτίλλοσαν is used instead of ἔτιλλον (PsSol 13:3); for μιαίνω (‘to pollute’), the imperfect ἐμιαίνοσαν is used instead of ἐμίαινον (PsSol 2:13); for ὁράω (‘to see’), the aorist εἴδοσαν is used instead of εἶδον (PsSol 8:25); for φεύγω (‘to flee’), the aorist ἐφύγοσαν is used instead of ἔφυγον (PsSol 11:4; 17:16).

3.3  Textual Variation and Corruption

In addition, the text is simply corrupt in many places, a problem that derives from three causes: majuscule errors, contextual variants and erroneous clause division:15 1. Majuscule errors:16 in my edition, I have tried to tally the number of variants based solely on majuscule errors. This yields an increasing frequency of textual corruption from the oldest level of transmission onwards.17 Particularly significant is the perspective that emerges from ms. 260, a magnificent manuscript that formed a part of the tripartite ‘Bible of Niketas’. Ms. 260 derives from the early Hyparchetype ε, mediated by Hyparchetype η. The latter can be identified as the Vorlage of Niketas’ Bible. Hyparchetype ε is the ascendant of two branches – namely, Hyparchetype η and ζ – and must have been of high quality. Hyparchetype ζ, which derives from it quite early on, only once misreads its Vorlage. However, the picture changes decisively with ms. 260, which is characterized by numerous misreadings. The best explanation for this is that Hyparchetype η was in a rather poor condition at the time when a copy of it was made for ms. 260. This impression is consistent with Cavallo’s suggestion that the ‘Bible of Niketas’, written in the 10th century, is a direct copy of a Vorlage that can be dated to the year 535 CE.18 I have identified this Justinian Vorlage with Hyparchetype η. 14  Cf. Albrecht, Psalmi Salomonis, 90–107. 15  Cf. Albrecht, Psalmen Salomos, 5–7. 16  One illustrative example of a misreading based on majuscule error (ΝΟ —> Ω) can be found in PsSol 2:13, for which cf. Albrecht, Psalmi Salomonis, 77. 94–95. I emend here ἐμιαίνοσαν, whereas von Gebhardt and Rahlfs read, with ms. 253, ἐμιαίωσαν. T. Muraoka, A Greek-English Lexicon of the Septuagint (Louvain: Peeters, 2009), 461 explains this form as going back to the base form μιαιόω, a hapax legomenon in the whole of Greek, though synonymous with μιαίνω. I, however, assume that the form ἐμιαίωσαν is based on majuscule error and stands for the koine-typical imperfect form ἐμιαίνοσαν. 17  Cf. esp. Albrecht, Psalmi Salomonis, 45–46. 18  Ms. 260 forms part of the famous ‘Bible of Niketas’ (ms. 90; ms. 719 + ms. 260). This connection has been until now totally neglected in former studies on the Psalms of Solomon. It is, however, of great importance, as the direct Vorlage of Niketas’s Bible can be dated exactly to the year 535 CE, according to Guglielmo Cavallo (Belting and Cavallo, Bibel).

The Göttingen Edition of the Psalms of Solomon

119

2. Contextual variants: PsSol 9:6b provides a good example of this sort of variation. All manuscripts testify to the sigmatic future καθαρίσει, with sigmatic forms like χρηστεύσει (v. 6a) and ἀφήσει (v. 7a) occurring in the immediate context. Here, I suppose that the asigmatic, contracted future (the so-called futurum atticum) is original but has been secondarily aligned with its context in the course of the transmission history. Therefore, I emend the text to καθαριεῖ. This asigmatic, contracted koine-typical future (futurum atticum) is typical of the Septuagint. This is also shown in PsSol 17:30, where a well-documented καθαριεῖ is attested and is considered original.19 3. Incorrect sentence division:20 the stichometric structure of the Psalms of Solomon has caused unintended mistakes by the copyists, as well as intentional and unintentional alterations by previous editors.21

4. The Psalm Collection and Its Formation22 In the introduction to my edition, I have also tried to shed light on the formation of the Psalms of Solomon as a collection. In my estimation, the text can be divided into two sub-collections: I. PsSol 2–8, and II. PsSol 9–16. All in all, three stages of development can be observed: 1. The core of the collection (sub-collections I–II) comes from Hasmonean times (165–63 BCE) and is related to the circle of the ‘pious’ (Chasidim); cf. the Ἁσιδαῖοι in 1Macc 2:42; 7:13; 2Macc 14:6. 2. A continuation of the first sub-collection (Ps. 2–8) was provoked by the Roman conquest of Judea by Pompey (63 BCE), as is reflected in PsSol 2 and 8. 19  Cf. Albrecht, Psalmi Salomonis, 46–47; 103–05. 20  Cf. Albrecht, Psalmi Salomonis, 47–52. 21  One illustrative example of a misreading based on the incorrect sentence division can be found in PsSol 2:4b–6b, for which cf. Albrecht, Psalmi Salomonis, 48–49. Ms. 336, written in stichoi, attests to three parallel formed stichoi, introduced by the sentence: “He (sc., God) did not grant them (sc., the sons of Jerusalem) the beauty of her (sc., Jerusalem’s) glory.” Here, the verb εὐοδοῦν (‘to grant’, or ‘to prosper’, lit. ‘to lead a good way’) is constructed with the dative of the person (αὐτοῖς, ‘them’) and the accusative of the object (τὸ κάλλος, ‘the beauty’). Hilgenfeld—followed by von Gebhardt, Ψαλμοὶ Σολομῶντος, and Rahlfs and Hanhart, Septuaginta—rejects, however, the attested εὐόδωκεν conjecturing instead εὐδοκῶ ἐν. The phrase is thus understood as the direct speech of God: ‘Throw them away from me! I do not like them.’ This conjecture would mean that the original structure—namely, the following three parallel stichoi—would no longer form a parallel structure, and the meaning of the text becomes obscure. If one now considers the attested stichometry, any conjecture of this sort becomes unnecessary, as the meaning of the verse becomes clear. 22  Cf. the short overview in Albrecht, Psalmen Salomos, 7–8. For an extensive overview of this topic, cf. Albrecht, Psalmi Salomonis, 191–192 (‘Die Psalmensammlung’); and 197–235 (‘Zur Entstehungsgeschichte’).

120

Felix Albrecht

At this stage, the Psalms of Solomon might already have been transmitted in Pharisaic circles. 3. The compilation of both sub-collections and the consolidation of the Psalms of Solomon into a single collection took place finally in Herodian times. In my estimation, this can be demonstrated by the final psalms, which, in my view, were added successively in Herodian times and were completed under Agrippa I (41–44 CE). These final psalms evidence a clear anti-Herodian sentiment, directed against Herod the Great. This may have been stimulated by Herod the Great’s self-declaration as a Solomonic ruler. It is therefore likely that criticism would have emerged against Herod the Great in Pharisaic circles. The picture of a Davidic Messiah, drawn in the final psalms (17–18), is, in my opinion, the last layer of revision of the whole collection. This includes the second part of PsSol 17 (vv. 30–46) and the first part of PsSol 18 (vv. 1–9).23 The final psalms are certainly the highlight of the whole collection. Indeed, as Wilhelm Bousset wrote, “nowhere is painted such a brilliant and elaborate picture of the Davidic Messiah as in the Solomonic Psalms (17–18).”24 The Psalms of Solomon bear witness to a Davidic messianism that is theologically significant, especially with regard to the development of the Davidic vision of the Messiah unfolded in the New Testament and linked to Jesus.25 For this reason, as well as for their place in the historical development of messianism, the Psalms of Solomon are an important theological document of the Palestinian Judaism of that time. The new Göttingen edition of the Psalms of Solomon provides a reliable textual basis that accounts for all relevant textual witnesses and that is based on an extensive consideration of the Greek and Syriac traditions. The introduction to this edition discusses both the linguistic characteristics of the text and the textual history of the work.

23  In my chapter on the textual history, I argue that the final messianic revision of the concluding Psalms (PsSol 17–18) likely fell in the reign of Agrippa I (41–44 CE), for which cf. Albrecht, Psalmi Salomonis, 212–234 (‘Die Endredaktion in Herodianischer Zeit’). This agrees with Bousset’s observations, especially on the reign of Caligula (37–41 CE); cf. Bousset, Religion4, 204–206. 24  “Nirgends ist ein so glänzendes und ausführliches Gemälde vom davidischen Messias entworfen wie in den salomonischen Psalmen (17–18).” Bousset, Religion4, 223, and 228: “Will man sich ein Bild davon machen, wie die nationale volkstümliche Hoffnung sich etwa die Gestalt des Messias dachte, so liegt ein Zeugnis ersten Ranges im siebzehnten salomonischen Psalm vor.” On the importance of PsSol 17, cf. also Atkinson, Lord, 129–179, esp. 129. 25 Willitts, “Matthew,” for instance, has shown that the concept of the Messiah in the Gospel of Matthew is not unlike that in the Psalms of Salomon.

The Göttingen Edition of the Psalms of Solomon

121

5. Bibliography Albrecht, Felix, “Die alexandrinische Überlieferung und die Rezension des Hesych von Alexandrien in den Prophetenbüchern der Septuaginta,” in Die Septuaginta, Orte und Intentionen, eds. S. Kreuzer, M. Meiser, and M. Sigismund. WUNT 361 (Tübingen: Mohr Siebeck, 2016), 337–62. —, Die Psalmen Salomos. Griechischer Text nebst deutscher Übersetzung und Gesamtregister (Göttingen: Vandenhoeck & Ruprecht, 2020). —, Psalmi Salomonis. Septuaginta. Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum 12.3 (Göttingen: Vandenhoeck & Ruprecht, 2018). —, “Zur Notwendigkeit einer Neuedition der Psalmen Salomos,” in Die Septuaginta: Text, Wirkung, Rezeption, ed. W. Kraus et al. WUNT 325 (Tübingen: Mohr Siebeck, 2014), 110–123. Albrecht, Felix, and Margherita Matera, “Testimonianze di παράγραφοι ‘a coda ondulata’ in alcuni manoscritti Greci e Copti,” Νέα Ῥώμη. Rivista di ricerche bizantinistiche 14 (2017): 5–35. Atkinson, Kenneth, I cried to the Lord. A Study of the Psalms of Solomon’s Historical Background and Social Setting. JSJ.S 84 (Leiden: Brill, 2004). Belting, Hans, and Guglielmo Cavallo, Die Bibel des Niketas. Ein Werk der höfischen Buchkunst in Byzanz und sein antikes Vorbild. Veröffentlichungen der Heidelberger Akademie der Wissenschaften, Philosophisch-Historische Klasse (Wiesbaden: Reichert, 1979). Bousset, Wilhelm, Die Religion des Judentums im späthellenistischen Zeitalter. In dritter verbesserter Aufl. hg. v. H. Gressmann. 4., photomech. gedruckte Auflage mit einem Vorwort von E. Lohse. HNT 21 (Tübingen: Mohr, 41966). Rahlfs, Alfred, and Robert Hanhart, Septuaginta. Id est Vetus Testamentum Graece iuxta LXX interpretes. Duo volumina in uno. Editio altera (Stuttgart: Deutsche Bibelgesellschaft, 2006). Skeat, Theodore Cressy, “The Provenance of the Codex Alexandrinus,” JThS 6 (1955): 233–235. Spinka, Matthew, “Acquisition of the Codex Alexandrinus by England,” JR 16 (1936): 10–29. von Gebhardt, Oscar, Ψαλμοὶ Σολομῶντος, Die Psalmen Salomo’s. Zum ersten Male mit Benutzung der Athoshandschriften und des Codex Casanatensis herausgegeben. TU 13.2 (Leipzig: J.C. Hinrichs’sche Buchhandlung, 1895). Willitts, Joel, “Matthew and ‘Psalms of Solomon’s’ Messianism. A Comparative Study in First-century Messianology,” BBR 22 (2012): 27–50. Wright, Robert Bradley, The Psalms of Solomon: A Critical Edition of the Greek Text. Jewish and Christian Texts in Contexts and Related Studies 1 (New York: T&T Clark, 2007).

Frank Feder A New Textual Witness of the Sahidic Version of Jeremiah and Its Text Historical Assessment 1. Introduction In a recent Festschrift1, I had the opportunity to publish a quite old parchment leaf from the British Museum, London, containing a hitherto non extant text portion from the beginning of chapter 22 of the Sahidic version of Jeremiah. While the Coptic manuscripts in London, counted among the ‘Western Manuscripts’, were transferred in the 1970s from the British Museum to the British Library the single leaf in question was acquired in 1996 by the British Museum at an auction and constitutes, together with a number of other Coptic manuscripts, a “refreshed” Coptic manuscript pool for the British Museum.2 Meanwhile, the BM leaf is also available in the Virtual Manuscript Room (VMR), the manuscript repository and publication platform of the Göttingen Academy project Complete Digital Edition and Translation of the Coptic-Sahidic Old Testament that represents the most actual text edition (manuscript: sa 2100).3 This is also a very good example for how quickly print publications become outdated, and for the advantages of a digital edition which can be updated at any time. The parchment leaf is of unknown provenance and can be dated to the fifthsixth century C.E.4 It is of a twofold importance for the Sahidic version of the Septuagint since it preserves a text section that was not extant so far, and it covers a time period that has been only very fragmentarily recorded by manuscript witnesses.5 In the critical edition of the Sahidic Corpus Ieremiae,6 the most substantial witnesses, which constitute the text, are the (early) fifth century P.Bodmer XXII7 and two fragmentary medieval manuscripts (ninth-tenth century)8 from the so

1 Feder, “Ein Pergamentblatt aus dem British Museum zu London.“ 2  E. O‘Connell, “Greek and Coptic Manuscripts from the First Millennium CE Egypt (still) in the British Museum,” in Proceedings of the 28th Congress of Papyrology, Barcelona 1–6 August 2016 (eds. A. Nodar and S. Torallas Tovar; Scripta Orientalia 3; Barcelona: Publicacions de l’Abadia de Monserrat, 2019), 68–80, see esp. 72–73. For the manuscript, see: https://www.britishmuseum.org/collection/object/Y_EA75338; accessed 06/14/2020. 3  http://coptot.manuscriptroom.com/manuscript-workspace/?docID=622100; accessed 06/14/2020. 4 Feder, “Ein Pergamentblatt aus dem British Museum zu London,“ 77–78. 5  Cf. Feder, Biblica Sahidica, 27–39. 6 Feder, Biblica Sahidica. 7 Feder, Biblica Sahidica, 28–30. 8 Feder, Biblica Sahidica, 30–34.

124

Frank Feder

called White Monastery near the old Upper Egyptian town Panopolis.9 As the best and most comprehensive manuscript witness, P.Bodmer XXII, sets in only in Jer 40:3, the first part of the book is quite badly preserved. All the more the new witness is very welcome. The conclusion of the text critical evaluation of all manuscript witnesses used for the critical edition revealed for the book of Jeremiah and Lamentations that all extant Sahidic witnesses, from the fourth-fifth century to the tenth century, show an outstanding textual stability, and that the Sahidic version was based on a Greek Vorlage which was characterized by a text form sharing many alterations (mostly additions) in its text with the Hexaplaric (and Lucianic) recension without being identical with it since these alterations appear more rarely and with a much lesser consequence in the Sahidic version. So, the Greek Vorlage of the Sahidic version could not be assigned to any of the known LXX text groups that transmit the book of Jeremiah.10 The extraordinary textual stability of the Sahidic transmission, which must go back to an initial translation event that I described as ‘Sahidic Standard Translation’,11 unfortunately, cannot be verified for the new witness from London since it has no Sahidic parallel text. Besides our new witness, there are only the beginning of verse Jer 22:4 (in a codex), and Jer 22:13, Jer 22:18*–19 (in quotations from Patristic literature) extant.12 Therefore, we will focus here on its text critical evaluation in comparison with the LXX text of the Göttingen edition by Joseph Ziegler.13 Our interest will be if the text of the London parchment leaf shows the same recensional traces as the other Sahidic witnesses as observed in my critical edition of the Sahidic book of Jeremiah?14

2. Variant Readings of sa 2100 I quote the variant readings from the Apparatus of Ziegler’s Göttingen Edition15 adding Sa (bold face) to the witnesses which show the same variant. 22:6 πόλεις] pr. sicut LaW Sa Aeth

9 Cf.: https://atlas.paths-erc.eu/places/112; accessed 06/14/2020. 10 Feder, Biblica Sahidica, 69–78; for the text groups, cf. Ziegler, Jeremias, 41–98. 11  F. Feder, “1.1.6. The Coptic Canon,” in Textual History of the Bible 2A: The Deuterocanonical Scriptures, eds. F. Feder and M. Henze; (Leiden: Brill, 2020), 213–239. 12 Feder, Biblica Sahidica, 143. 13 Ziegler, Jeremias. 14 Feder, Biblica Sahidica, 69–78. 15 Ziegler, Jeremias, 256–259.

A New Textual Witness of the Sahidic Version of Jeremiah

125

Though the passage is only partly readable in sa 2100, there can be no doubt that Sa read “ὡς” before πόλεις with the Old Latin manuscript Codex Wirceburgensis and the Ethiopic version. 22:7 ἄνδρα ὀλεθρεύντα B–S–130 A–106‘ Q–613 LaW Sa Bo Aeth Arab τὰς ἐκλεκτὰς κέδρους σου] ⲛⲟⲩⲕⲉⲛⲇⲣⲟⲥ ⲉⲧⲥⲟⲧⲡ Sa (ἐκλεκτὴν κέδρον) In this case there is no other witness to confirm this reading, so it might be caused by the translator. 22:8 ἔθνη] + πολλά rel. incl. Sa (ϩⲉⲛϩⲉⲑⲛⲟⲥ ⲉⲛⲁϣⲱⲟⲩ) = MT 22:11 Ιωσία] + τοῦ βασιλέως ιουδα L‘ 26 Sa Chr. Tht. = MT 22:12 μετῴκισα] -σαν Q–V–130–239–534–544–710 O L‘–538 Sa Bo Aeth Arm Tht. = MT 22:13 παρὰ τῷ πλησίον αὐτοῦ] παρ᾽ ἀυτῷ ὀ πλησίον O 51c-l Sa Tht. Hi

However, this variation can be explained as caused by translation technique as well. 22:14 ἐν κέδροις 534 Sa 22:15 βασιλεὺς εἶ L-311 Sa (ⲉⲕⲛⲁϭⲱ ⲉⲕⲟ ⲛⲣⲣⲟ) Tht.p | βέλτιον ἦν σε] om. ἦν B Sa (cf. Bo)

It is not certain that Sa really did not have ἦν in its Vorlage, however, the more literal Bohairic version suggests this. 22:16 om. κρίσιν ταπεινῷ οὐδέ Sa: homoioteleuton

This might well be an omission that occurred in the inner Sahidic transmission process; if we had other Sahidic witnesses this passage would probably reappear. 22:17  ἰδοὺ οὔκ εἰσιν οἱ ὀφθαλμοί σου] ἰδοὺ οὔκ εἰσιν οἱ ὀφθαλμοί σου εἰς ὄρθον Sa (cf. ὄρθοι) οὐδέ] καί Sa Bo Cyr. X 292 = MT Sa + Bo render this phrase correctly with a negation so that οὐδέ very likely was in their Vorlage. Therefore, Sa and Bo have to be deleted from the apparatus for this lemma. (ἡ καρδία σου) καλὴ ἀλλ(ά) rel. incl. Sa εἰς 1° 2°] επί Sa L‘ = MT; Sa reads ἐπὶ τὴν πλεονεξίαν (omitting σου) om. τό 1° Sa 26 = MT; om. τό 2° Sa 26 538 om. ἀυτό] Sa Hi. = MT καὶ εἰς ἀδίκημα] om. εἰς Sa LaW Sa read: καὶ αἶμα ἀθῶον τοῦ ἐκχέειν καὶ ἀδίκημα καὶ εἰς φόνον] om. εἰς Sa LaW

126

Frank Feder

22:18 επὶ Ιωακιμ] επὶ τὸν ἄνδρα τοῦτον Ιωακιμ Qmg Sa Bo om. οὐαὶ ἐπὶ τὸν ἄνδρα τοῦτον O L‘–538 Sa Bo Chr. Tht. = MT οὐ μὴ κόψωνταί σε (αυτον Sa-QAth) λέγοντες Sa ἄδελφε] + μου Syh Sa Bo Arab: cf. MT κλαύσονται ἀυτόν] et (+ non Aeth) dicent Sa-QAth Aeth Arm οἴμμοι κύριε μου, οἴμμοι ἄδελφέ μου Sa  κύριε] + μου Syh Sa Bo (om. Sa-QAth) Aeth; + οἴμμοι ἄδελφε (+ μου Sa Aeth Arab) L‘–613 Sa Aeth Arab Tht.; + καὶ οἴμμοι ἄδελφε (+ μου Syh) Q–V–46–130–534–710 O–233 c = MT

3. Conclusion From the relatively few verses of chapter 22 of Jeremiah it becomes obvious that also the London manuscript sa 2100 shares the same text character as the other Sahidic witnesses transmitting the Jeremianic corpus. We encounter again the same kind of changes towards the LXX text, which I described and listed in my critical edition.16 These are mostly additions, and they regularly have the tendency to adapt the text closer to MT. My then conclusion was that the Sahidic version of Jeremiah was based on a Greek Vorlage which was characterized by a Pre-Hexaplaric recension sharing many alterations in its text with the Hexaplaric (and Lucianic) recension without being identical with it since these alterations appear more rarely and with a much lesser consequence in the Sahidic version. So, the Greek Vorlage of the Sahidic version could not be assigned to any of the known LXX text groups that transmit the book of Jeremiah. Before I will suggest a different explanation some details from the text critical notes above may further be highlighted. It is very significant, again, that the Greek minuscule manuscripts 26, 534, and 538, also in these few passages of chapter 22, share regularly variant readings with Sa.17 Further investigation should take into account that they probably preserve traces of a Greek text form that could have been the Vorlage for Sa. The same might be true for the Old Latin version which is represented here by LaW, i.e. Codex Wirceburgensis. Some variant readings are more likely to be based on translation technique, cf. Jer 22:7, 22:13. A typical addition based on translation technique is the Coptic possessive article, which is often preferred to translate a Greek vocative since such a form does not exist in Coptic (cf. 22:18 the addition of μου).18 One may wonder if Aeth and Arab follow here Sa (or Bo)? However, there is also a case of 16 Feder, Biblica Sahidica, 69–78. 17  Cf. Feder, Biblica Sahidica, 75–78. 18 Feder, Biblica Sahidica, 98–99.

A New Textual Witness of the Sahidic Version of Jeremiah

127

a wrongly noted variation in the apparatus by Ziegler (cf. Jer 22:17). Sa and Bo, though both Coptic versions, at least for the book of Jeremiah are text historically independent versions. When they share a variant reading, very often, this is caused by translation technique, or is a pseudo variation. At the occasion of reviewing the study on translation technique and its implication on the evaluation of the text critical value of Sahidic 1 Samuel by Elina Perttilä,19 I underlined the importance of the results of her study both for the Sahidic version of Jeremiah and very likely for the Sahidic version in general. The Sahidic version of 1 Sam shows some influence from Hexaplaric-type readings, mostly these are also encountered in the witnesses of the Lucianic recension which is dependent on the Hexapla. However, only some of these Hexaplaric-type corrections enter the Sahidic text. In some cases, the Sahidic text is closer to MT than any preserved Greek witness. Thus, a now-lost Greek source lies behind some Sahidic variants. These observations exactly parallel my own observations in the Jeremianic Corpus.20 My original assumption of a Greek Pre-Hexaplaric recension being behind this phenomenon is certainly incorrect. Consequently, Perttilä formulates more cautiously that “one cannot know whether scribes working on the Sahidic tradition acquired these readings from one source, from several margins or from several manuscripts over the years. The Sahidic tradition contains some early corrections according to the MT, but it does not consistently reflect any particular Greek text type. This textual diversity is typical of early Greek manuscripts as well.”21 Recent research especially on the Coptic versions of the Minor Prophets, which are preserved in various dialects, seems to confirm that the Sahidic translators could have taken those Greek readings from manuscripts with marginal glosses stemming from the fifth column of the Hexapla.22 As for the inner Sahidic and inner Coptic textual history, not much can be said here about the Sahidic transmission as sa 2100 is practically the only witness. However, as I mentioned at the beginning, the few Sahidic parallel verses (Jer 22:13; 22:18:19) from patristic authors, Athanasius of Alexandria and Shenoute,23 show 19  F. Feder, Review of Perttilä, Sahidic 1 Samuel – A Daughter Version of the Septuagint (Göttingen: Vandenhoeck & Ruprecht, 2017), JSCS 53 (2020): 147–57. 20 Feder, Biblica Sahidica, 74–78. 21 Perttilä, Sahidic 1 Samuel, 225. 22  This will be outlined in the forthcoming re-edition of the Minor Prophets in the Göttingen Septuagint series: J. Ziegler and F. Albrecht, Duodecim Prophetae (Septuaginta: Vetus Testamentum Graecum XIII; 4th completely revised edition; Göttingen: Vandenhoeck & Ruprecht), forthcoming; cf. also N. Bosson, “Jonas: La version sahidique du codex Crosby-Schøyen ms 193 et ses liens avec la version Paléo-Bohairique du Papyrus Vatican Copte 9 des Petits Prophètes in Coptic Society,” in Literature and Religion from Late Antiquity to Modern Times: Proceedings of the Tenth International Congress of Coptic Studies, Rome, September 17th–22nd, 2012 (eds. P. Buzi, A. Camplani, and F. Contardi; OLA 247; Leuven: Peeters, 2016), 821–37. 23  Cf. Feder, Biblica Sahidica, 143 and 46–50.

128

Frank Feder

almost no variation. They seem, nevertheless, to confirm the outstanding textual stability of the Sahidic version. Since Athanasius wrote in Greek we still need to investigate if the Coptic translator of his festal letters24 used for the biblical quotations the Sahidic Bible, or if he translated them directly from Greek. However, the variation towards sa 2100 is marginal. The only completely preserved Coptic version of the book of Jeremiah is the classical Bohairic version (Bo). The extant manuscripts of the Major Prophets date not earlier than the 14th century.25 Since there is no modern edition of any of the manuscripts, we still have to rely on the edition by Henry Tattam from 1852.26 It remains, however, unclear on which manuscript(s) his edition was based.27 The text character of Bo, apparently independent from Sa,28 and its striving for a more literal translation29 make it very worthwhile for a more thorough investigation.

4. Bibliography Feder, Frank, “Ein Pergamentblatt aus dem British Museum zu London (EA 75338) mit Jer 22,4–22,20,” in Ägypten und der Christliche Orient: Peter Nagel zum 80. Geburtstag, eds. H. Behlmer, U. Pietruschka, and F. Feder. Texte und Studien zur Koptischen Bibel 1 (Wiesbaden: Harrassowitz, 2018), 77–84. —, Biblia Sahidica: Ieremias, Lamentationes (Threni), Epistula Ieremiae et Baruch. TUGAL 147 (Berlin: De Gruyter, 2002). Perttilä, Elina, Sahidic 1 Samuel – A Daughter Version of the Septuagint 1 Reigns. De Septuaginta Investigationes 8 (Göttingen: Vandenhoeck & Ruprecht, 2017). Tattam, Henry, Prophetae Majores: in Dialecto Linguae Aegyptiacae Memphitica seu Coptica, vol. 1 (Oxford: E Typographeo Academico, 1852). Ziegler, Joseph, Ieremias, Baruch, Threni, Epistula Ieremiae. Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum 15 (4th edition; Göttingen: Vandenhoeck & Ruprecht, 2013).

24  L.-Th. Lefort, Lettres festales et pastorals en Copte (CSCO 150–151; Louvain: Durbercq, 1955). 25  Cf. H. Takla, An Introduction to the Coptic Old Testament, Coptica 6 (2007): 87–88. 26 Tattam, Prophetae Majores. 27 Tattam, Prophetae Majores. Judging from his Preface (p. v) in Latin his transcription “e Codice Copto-Arabico, …, eodem pulchro sed non antiquo, in Bibliotheca Episcopi Ecclesiae Latinae apud Cairenses adservato” could have been based on the MS. now London British Library Or. 1319 (dated 1806), and the manuscript he used for comparison “Manuscripto vetustissimo in Bibliotheca Patriarchae Alexandrini” might have been MS. Cairo, Patriarchal Library Bible 11 (14th century?). 28  Cf. Feder, Biblica Sahidica, 53–78. 29 Feder, Biblica Sahidica, 79–103.

II.  Hexapla and Recensions

John D. Meade The Dream for a ‘New Field’ Comes True A Description and Defense of the New Critical Edition of Job 22–42 1. Introduction Work on the “New Field for the 21st Century” has been ongoing since the Rich Seminar in 1994, and now finally its first volume, A Critical Edition of the Hexaplaric Fragments of Job 22–42, edited by John D. Meade was published with Peeters in 2020.1 Although this is a momentous occasion for me, the Hexapla Project’s editorial committee of Peter Gentry, Alison Salvesen, and Bas ter Haar Romeny and other advisers along the way deserve much praise for casting the vision for this project and now for finally producing volumes of this ambitious critical edition. In this article, first, I want to review the objectives of the Hexapla Project and compare and contrast them with the stated aims of the magisterial Göttingen Septuaginta. Second, in order to describe the Hexapla Project’s New Field, I will look at a few representative examples where the “New Field” advances beyond not only the ‘Old Field’ but also the second apparatus of the Iob edition by Joseph Ziegler.2 Third, I want to address Olivier Munnich’s criticisms of the Hexapla Project’s New Field, which he made in 2014 to show that, at least for Job, they are off the mark.

2. Aims of the Hexapla Project From the beginning, the Hexapla Project has set out to produce “A New Field for the 21st century.”3 Due to the newer manuscript collations of the Göttingen Septuaginta-Unternehmen, which uncovered new hexaplaric materials for its second apparatus, and new critical editions of patristic works and catenae, the founders of the Hexapla Project realized that Field’s work from 1875 had become outdated. What was needed was not a new work based exactly on Field’s previous edition but a new work that used his method and combined it with the newer collations included in the Göttingen Septuaginta.4 1 Meade, A Critical Edition. 2 Ziegler, Iob,; Field, Origenis Hexaplorum. 3  According to the website: http://hexapla.org/. Accessed on 10/14/2020. 4  Cf. Romeny and Gentry, “Towards a New Collection of Hexaplaric Material for the Book of Genesis;” Norton, “Collecting Data for a New Edition of the Fragments of the Hexapla.”

132

John D. Meade

Primarily, Field used and expanded the work of Montfaucon by supplying retroversions from the Syro-hexapla and occasional readings from the Holmes-Parsons Edition and passim, but he still attempted to reconstruct the original readings of the Three in his edition noting variants and providing commentary on these readings in Latin in the footnotes.5 The readings of the Three were preceded by the relevant wording of the Hebrew and Septuagint texts so the user of the edition could better appreciate the revision of the Septuagint vis-à-vis the Hebrew Text. The Göttingen Septuaginta chiefly aims to reconstruct the Old Greek translation.6 The second apparatus in all these editions, except the first and provisional volume of Psalms which has no second apparatus, contains hexaplaric readings, but as I will show below at least in the case of the Job volume, it did not attempt a critical reconstruction of many of these readings. In many cases, readings are simply left side by side with the editor making no decision between the more original text and the variant. This practice is consistent with the aims of the Septuaginta-Unternehmen (1908–2015) to reconstruct the Old Greek, not its revisions. The Hexapla Project’s “New Field”, therefore, employs Field’s method and aim for establishing the wording of the fragments and makes great use of the collation of the hexaplaric fragments in Göttingen’s second apparatus as well as other editions and manuscripts that have appeared since the publication of these editions. The description of the new series, Origen’s Hexapla: A Critical Edition of the Extant Fragments, is as follows: “Frederick Field’s marvellous late Victorian edition (1875) of the remains of Origen’s Hexapla is now outdated. Field rearranged earlier collections, and added new material, notably retroversions into Greek from Syriac sources. In the course of work on critical editions of the Septuagint, new manuscripts and patristic sources have become available, as well as new editions of Church Fathers and catenae. Some of these contain better readings and even previously unknown material from Origen’s Hexapla. This new critical reconstruction of all known hexaplaric materials is being prepared by the Hexapla Project, a project spawned by the Hexapla Institute under

5 Montfaucon, Hexaplorum Origenis quae supersunt; Holmes/Parsons, Vetus Testamentum Graecum cum variis lectionibus,. For a history of past editions of Origen’s Hexapla, see Law, “A History of Research on Origen’s Hexapla: From Masius to the Hexapla Project.” 6  Cf. the website of the Septuaginta-Unternehmen (1908–2015): https://adw-goe.de/en/ research/completed-research-projects/akademienprogramm/septuaginta-unternehmen/. Accessed 10/14/2020: “The Göttingen Septuaginta-Unternehmen had set itself the task of reconstructing—in the form of a critical edition—the earliest form of the text that can be reasonably reconstructed, that is: a text that precedes all recensions. In doing so, it also aimed to document all alterations in an apparatus that accompanied the critical text.”

The Dream for a ‘New Field’ Comes True

133

the aegis of “The International Organization for Septuagint and Cognate Studies” (IOSCS).”7

What does the “New Field” look like and what does it do? Since the appearance of Joseph Ziegler’s Iob in 1982, Ursula and Dieter Hagedorn published their Nachlese (1991) and a critical edition of the oldest catena of Job (1994–2004).8 Since the appearance of their edition, Reinhart Ceulemans informed us of an important catena manuscript, Tyrnavos 25 (Rahlfs 788). Subsequent work on this manuscript showed that it is the most faithful witness to the text of the oldest catena of Job, and therefore, it often transmitted the copious hexaplaric remains in that tradition more faithfully than other known witnesses.9 Before the addition of 788, the Hagedorns had shown that Rahlfs 250 (= G in their edition) was considered the purest witness of the C group (= α-Vorläuferkatene in their edition). Since 788 is the ancestor of 250, the new edition cites 788-250 as 788′ and lists variants contained in 250 in its apparatus. The introduction to the new critical edition provides a thorough description and update to this scholarship. In what follows, I describe the features of the new edition and compare it with the previous ones.10

3. Description of the ‘New Field’ In comparison with Ziegler’s Iob, the “New Field” for Job 22–42 contains twenty-two previously unknown fragments, some thirty revised attributions and lemmas, some twenty-five previously known fragments received revised attributions, and the wording of about sixty-five fragments was revised. A few fragments that Ziegler included in his second apparatus were shown not to be related to the Hexapla and were therefore removed. My focus in this article, however, is not merely to give an update. I want to compare Field, Ziegler’s Iob edition, and the New Field in order to show how the new critical edition advances our under7 Meade, A Critical Edition, ii. 8  Hagedorn and Hagedorn, Nachlese; Hagedorn and Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band I; Hagedorn and Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band II; Hagedorn and Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band III; Hagedorn and Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band IV. 9 Meade, “The Significance of Ra 788 for a Critical Edition of the Hexaplaric Fragments of Job.” 10  The Appendix below provides a summary of the advances of Meade’s edition over prior editions. In the Notes sections below, “Kommentar” refers to Hagedorn and Hagedorn, Olympiodor Diakon von Alexandria: Kommentar zu Hiob; “Nic” refers to the Nicetas catena in the edition by Young, Catena Gaecorum Patrum in beatum Iob; „AGK“ refers to Hagedorn and Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band I–IV.

134

John D. Meade

standing of the textual history of the hexaplaric remains and their reception into Christian sources. Thus, the New Field puts both the hexaplaric readings and their provenance on a surer foundation than previous collections.

3.1  Job 39:24b

Field Σ. καὶ οὐ δηλωθήσεται ὑπὸ ἤχους σάλπιγγος11 Iob, ed. Ziegler 24b] σ′ καὶ οὐ δηλωθήσεται ὑπὸ ἤχους (ηχουp; ηλιου 256) σάλπιγγος C New Field (Job 22–42, ed. Meade) HT LXX σ′ Wit1: Attr: Var:

‫ׁשֹופר׃‬ ָ ‫וְ לֹא־יַ ֲא ִמין ִּכי־קֹול‬ καὶ οὐ μὴ πιστεύσῃ, ἕως ἂν σημάνῃ σάλπιγξ· καὶ οὐ δειλωθήσεται ὑπὸ ἤχους σάλπιγγος ↓C12 (= 788′ 3005) ↓cI–[680] 732 σ′] > 250 139 643 δειλωθήσεται 788′ 260] δηλωθήσεται 3005 cI–260 [680] 732; δηλωνθήσεται 395 | ὑπό] ἀπό 788′ | ἤχους] ἤχου 788′ 3005 137 139 260 395 643 3006

Notes: Ziegler does not list the evidence of the variant spelling of manuscript Rahlfs 250 δείλω- (now also attested in manuscript Rahlfs 788), and thus he presents all the manuscripts as having δήλω-. The Nachlese reports that manuscript Rahlfs 250 has δείλω-, which would be from δειλόομαι “to be afraid,” “to fear.”13 This reading fits the context more suitably, the interchange between ει and η is due to itacism.14 This confusion accounts for the variant, which entered the textual tradition quite early. The conjecture of Kreysigius, reported by Schleusner, has now found significant support in 788′, despite Schleusner’s own rejection of it (Schleusner I, 540, s.v. δηλόω).15 The Syriac Peshitta has .‫“( ܘܐܠ ݁ܕܚܠ ܡܢ ܩܐܠ ܕܩܪܢܐ‬and he was not afraid of the sound of the trumpet”). Symmachus is not dependent on Pesh, and Pesh is not dependent on Symmachus, but since there is a reference to the horse’s scorn of 11  Field added, “Kreyss. tentat δειλωθήσεται” (p. 73), referring to the conjectural emendation by Kreissig, Observationes Philologico-Criticae in Jobi Cap. XXXIX. vers. 19–25, 41. 12  For the new edition, the C group contains the following manuscripts: 249-257-395-406788′-3005-Ω, while the cI group contains: 110-137-138-139-147-476-251-255-256-258-260395-474-559(3004)-612-643-680-705-732-740-754-765-3006. 13  Hagedorn and Hagedorn, Nachlese, 408. 14  For this well-documented interchange, see Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, 240–2. 15 Schleusner, Novus Thesaurus Philologico-Criticus.

The Dream for a ‘New Field’ Comes True

135

fear in 39:22, each translator/reviser probably came to the same conclusion in verse 24 independently.

Explanation and Summary First, each entry in the New Field begins with the Hebrew Text, which is the vocalized Masoretic Text without cantillation marks or accents. Origen did not have a text with the vocalization graphemically represented but the Hexapla Project includes it because the revisers frequently reflect the consonantal text of proto-MT and their translations often, though not always, anticipate the vocalization of the MT. Second, the Old Greek from the Göttingen Septuaginta is provided. Origen may or may not have had this text in front of him but once again it is provided to facilitate modern research of the revisions. Third, the reconstructed attribution and hexaplaric fragment are given. Fourth, the Wit1 line contains all the manuscripts that contain evidence for this fragment. In this case, the C group and most of the cI group contain the reading. The down arrow next to a manuscript or manuscript group indicates that these witnesses appear below in another apparatus. Many entries include a Wit2 apparatus which presents relevant information from Ziegler’s first apparatus, especially those places where the readings of the Three have entered MSS of the transmission of the Old Greek text. Fifth, the Attr apparatus provides variants to the attribution. In this case, 250 139 643 lack the attribution to Symmachus which is found in all other witnesses. Sixth, the Var apparatus supplies all variants to the established text. Seventh, if there had been evidence in Latin, Syriac, or Armenian for this fragment, the relevant text would have been given in the original script in the NonGr line. Eighth, most entries have a Notes section that provides commentary on the textual decisions and other matters of interest. The New Field presents a clear, reconstructed text with variants in the apparatus. Ziegler used parentheses to show variants but could not show economically the divided evidence between ἤχους/ἤχου as the Var apparatus in the New Field does. Therefore, Ziegler lists ηχουp for part of the C group but the researcher does not know what part or what manuscripts exactly have this reading. The New Field shows that members of the newly established C group and the lower members of the cI group that have ἤχου, a variant caused by haplography of sigma. Furthermore, the New Field provides a fresher collation compared to Ziegler here and elsewhere. Ziegler had access to manuscript Rahlfs 250 but he did not note its text for this fragment. Now, manuscript Rahlfs 788 confirms that δειλωθήσεται, not δηλωθήσεται, is the reading of the oldest Greek catena and probably the original wording of Symmachus.

136

John D. Meade

As explained above, Symmachus attempts a contextual rendering of the Hebrew Text by recourse to v. 22, where he rendered the noun ‫ימה‬ ָ ‫ ֵא‬with φόβος. In the Hebrew Bible, only the noun ‫ימה‬ ָ ‫ ֵא‬appears with no verbal forms from ‫אים‬ appearing until later in the history of Hebrew, when a Piel denominative verb surfaces. Symmachus would have to be rendering the Pual stem of this late verb for one to suggest that he rendered a different Hebrew Text than proto-MT in this instance.

3.2  Job 40:8ab

Field Σ. μήτι καὶ παρακυρώσεις κρίσιν μου; ἄδικον ποιήσεις με ἵνα δικαιωθῇς;16 Iob, ed. Ziegler 8a] σ′ (leg. α′) μήτι καίπερ ἀκυρώσεις (και παρακυρωσης 248; σ′ ἀκυρώσεις C) κρίσίν μου 248 252 8b] σ′ ἄδικον ποιήσεις με, ἵνα δικαιωθῇς 248 252 (sub α′); anon μὴ ἄδικον ποιήσῃς με (sic 250; > rel), ἵνα δικαιωθῇς C New Field (Job 22–42, ed. Meade) Job 40:8a HT LXX ⟨α′⟩ Wit1: Attr: Var:

‫ַה ַאף ָּת ֵפר ִמ ְׁש ָּפ ִטי‬

μὴ ἀποποιοῦ μου τὸ κρίμα. μήτι καίπερ ἀκυρώσεις κρίσίν μου lemma] ↓248 ↓252 | ἀκυρώσεις] ↓C (= 788′ 3005) ↓cI–138 260 [680] 732 3006 ⟨α′⟩] α′ σ′ 788 3005; σ′ 137 139 248 252 255 395 559 612vid 643 740; > 250 καίπερ ἀκυρώσεις] καὶ παρακυρώσης 248

Notes: 3005 preserves a double attribution to Aquila and Symmachus, while the rest of the witnesses preserve an attribution to Symmachus.17 788 also attests the double attribution, indicating that the attribution to Aquila is part of the history of the text. Already, Ziegler noted that the lexemes in this fragment show the attribution to Aquila as correct and he favored him as the author in his edition.18 Based upon the vocabulary and translation technique (e.g. μήτι καίπερ), Aquila is probably the author of this fragment, and already in the Vorlagen of 3005 and 788, the attribution to Symmachus was added to the attribution to Aquila.19 After 16  Field notes, “From Nobil. and Drusio, Montef says, but he wrongly omitted καὶ after μήτι. Parsons, from Cod. 252 reports: Σ. μήτι καίπερ ἀκυρώσεις κρίσιν μου. Ἀ. ἄδικον κ. τ. ἑ. We rewrote as παρακυρώσεις instead of -σῃς [the reading of Ra 248]” (p. 75n1). 17  Hagedorn and Hagedorn, Nachlese, 409. 18 Ziegler, Beiträge, 47; Ziegler, Iob, 397. 19  Symmachus uses καίπερ for ‫ גַ ם‬in Eccl 4:14. By contrast, Aquila uses καίπερ for ‫ וְ ַאף‬in Lev 26:39 and Ps 67(68):19 and ‫ ַאף‬in 4 Kgdms 2:14 and Prov 21:27 (Field’s “Nobil.” here = man-

The Dream for a ‘New Field’ Comes True

137

this intermediate stage, the rest of the catena lost the attribution to Aquila and only the one to Symmachus remained. The reading in 248 arose from an error of word division and subsequent corruption.

Job 40:8b HT LXX σ′ Wit1: Attr: Var:

‫יענִ י ְל ַמ ַען ִּת ְצ ָּדק׃‬ ֵ ‫ַּת ְר ִׁש‬ οἴει δέ με ἄλλως σοι κεχρηματικέναι ἢ ἵνα ἀναφανῇς δίκαιος; μὴ ἄδικον ποιήσεις με, ἵνα δικαιωθῇς ↓C (= 788′ 3005) ↓cI–[680]↓248 ↓252 σ′ 788 3005 248] α′ 252; > 250 cI–[680] μή] > 248 252 | ποιήσεις 248 252] ‑σης 788′ 3005 cI–3006; ποίησις 3006 | με 788′ 3005 248 252] > cI–[680]

Notes: 3005 confirms that 248 preserved the correct attribution to Symmachus against 252, which has Aquila.20 788 also supports this attribution. Ziegler already favored Symmachus as the author based on 248, but considered this lemma different than the one preserved in the catena (Iob, ed. Ziegler, 397). Since the fragment is now identified in the catena, it is clear that these fragments derive from the same source.

Explanation Field’s edition attributed the fragments of chapter 40, verses 8a and b to Symmachus on the basis of Nobilius and Drusius.21 In this case, Nobilius’ source is manuscript Rahlfs 248, which already contained a variant reading to what the catenae and manuscript Rahlfs 252 preserved for v. 8a. The analysis of translation equivalents, however, caused Ziegler already to conclude that the fragment in 40:8a should be attributed to Aquila and to treat it as a separate fragment from the one in v. 8b. The New Field presents the manuscript evidence of the double attribution to Aquila and Symmachus in Rahlfs 788 and 3005 for 40:8a showing that an attribution to Aquila was part of the textual history of that fragment. uscript Rahlfs 248 fol. 40v). Symmachus employs μήτι for ‫ ֲה‬once in Job 6:6 (outside of Job, cf. Exod 2:14b σ′, Amos 9:7 σ′, and Isa 66:9 σ′). However, Aquila uses μήτι for ‫ ֲה‬many places in Job (4:2, 10:3 11:7, 14:14, 15:2, 15:8, 23:6, 26:5, 38:28, and 40:31). See Reider / Turner, An Index to Aquila, 157–8. Thus, the attribution to Aquila is most probable, and it is one of the attributions preserved in 788 3005. 20  Hagedorn and Hagedorn, Nachlese, 409. 21 Field, Origenis Hexaplorum, 75. For the readings in Nobillius, see Nobilius, Vetus Testamentum secundum LXX Latine redditum, 740; for the readings in Drusius, see Drusius, Veterum Interpretum Graecorum in totum vetus Testamentum Fragmenta, 855.

138

John D. Meade

Thus, the New Field updates our knowledge of the textual history of this fragment significantly. Ziegler’s treatment of the two fragments in 40:8b shows the necessity of the New Field. There is one fragment attributed to Symmachus for which variants exist. That conclusion seemed clear even before the new evidence of Rahlfs 788 and 3005, which preserve the original attribution to Symmachus, and therefore, show that the fragment in the catena is the same as the one in Rahlfs 248 and 252 albeit with variants. But Ziegler’s second apparatus does not decide the original text of the fragment here or in many other places like this one. Rather, Ziegler often leaves textual decisions of this nature undecided and presents them in parallel fashion. To be clear, I do not criticize the Göttingen Septuaginta here, since its project’s aim is to establish the Old Greek translation – not the text of the revisers. But the New Field’s aim does attempt to reconstruct the text of the revisers from all the available evidence and puts variants to this reading in its apparatus.

3.3  Job 30:7a

Field ’Α …φυτῶν…22 Σ. μεταξὺ φυτῶν ἀγρίων ἠχοῦντες λιμῶδες (s. λιμώδεις).23 Iob, ed. Ziegler 7a] σ′ μεταξὺ φυτῶν (των 138–680; σ′ φυτῶν 612; α′ φυτῶν 138-255-680-740) ἀγρίων ἠχοῦντες λιμῶδεις (-δες c) C′ (Olymp): cf 6:5a New Field (Job 22–42, ed. Meade) HT E σ′ Wit1: Attr: Var:

‫יחים יִ נְ ָהקּו‬ ִ ‫ין־ׂש‬ ִ ‫ֵּב‬

ἀνὰ μέσον εὐήχων βοήσονται· μεταξὺ φυτῶν ἀγρίων ἠχοῦντες λιμῶδες C (= 788′ 3005) ↓cI–137 [559] 732 cII 161 σ′] θ′ 139; > 788′ 643 740 φυτῶν] τῶν 138 | λιμῶδες 788′ 3005 cII 161] λιμώδεις 138 139 255 260 395 612 643 680 740; λιμώδης 3006

22  Field cites Cod. 138, 255. 23  Field says, “Nicet. and “omnes” for the reading. According to 138, 139, 255; from which Cod. 138 has τῶν instead of φυτῶν, and Cod 139, 255 Θ. instead of Σ. Furthermore, λιμῶδες is in Nicet. and Montefaucon except for Reg. unum [i.e. Ra 560]; λιμῶδεις in Reg. and Parson’s three [i.e. 138, 139, 255].”

The Dream for a ‘New Field’ Comes True

139

Notes: The Nachlese comments that Ziegler was not quite clear that the oldest Greek catena contains two separate fragments, one attributed to Aquila (see infra) and the other attributed to Symmachus.24 Therefore, there is no variant attribution to Aquila. The reading of 138 is dependent on the Hagedorns, who have read φ. των. Furthermore, 680 has φυτῶν and not τῶν as in Ziegler’s edition. The neuter λιμῶδες is considered the right reading in the Nachlese pace Ziegler’s edition, since (1) this is the reading in the oldest Greek catena (manuscript Rahlfs 788 confirms this reading) and (2) Job 6:5a shows that Symmachus uses the adverbial function of the neuter singular, and wider Greek literature also confirms this usage (see esp. 389–90, where the Hagedorns cite Plutarch, Amatorius 751A).25 Ziegler’s attribution to Olympiodorus is not supported by Kommentar or AGK but only by Nic (Kommentar, 252; AGK 3:103; Nic, 454–5).26 HT E α′ Wit1: Attr:

(‫יחים )יִ נְ ָהקּו‬ ִ ‫ין־(ׂש‬ ִ ‫)ּב‬ ֵ

(ἀνὰ μέσον) εὐήχων (βοήσονται·) φυτῶν ↓C (= 788′ 3005) ↓cI–137 139 260 [559] 643 732 3006 α′] σ′ 612; > 250

Notes: The catena had two separate readings for Aquila and Symmachus.27

Explanation The New Field’s text of the hexaplaric reading sometimes confirms the text of Field against the Göttingen Septuaginta. Field already noted the two fragments of Aquila and Symmachus for this verse, even though he only had a few manuscripts upon which to base these readings. His edition still expressed doubt over the decision between the neut. singular λιμῶδες or the masc./fem. plural λιμώδεις. The New Field clarifies the problems here and decides on λιμῶδες because the best manuscripts have this reading and the usage of the neuter singular as an adverb is attested in Symmachus at Job 6:5 and Plutarch. Ziegler attempted to convey the same material as Field, but his method of placing variants and shorter lemmas, which contain a word or two of longer lemmas in parentheses, has obscured the Aquila reading for this verse and now it

24  Hagedorn and Hagedorn, Nachlese, 403. 25 Ziegler, Iob, 340; Hagedorn and Hagedorn, Nachlese, 403. 26  For abbreviations and bibliographic information see n. 288 above. 27  Hagedorn and Hagedorn, Nachlese, 403.

140

John D. Meade

could be mistaken for a variant to the Symmachus fragment. The New Field not only restores the two readings for this verse but also updates the evidence register showing that both the readings and the attributions are traced to the C group. Finally, Ziegler noted that the Symmachus reading was found in Olympiodorus. The New Field verifies that this reading is not found in the Hagedorns’ Olympiodorus Commentary on Job nor is it found in their edition of the oldest Greek Catenae of Job. The attribution to Olympiodorus enters the textual transmission of the catenae at the stage of the Nicetas catena or cII and is thus suspect. Therefore, the New Field attempts to verify the provenance of the hexaplaric fragments as carefully as possible, in order to avoid the levelling of the varied sources that preserve them.

3.4  Job 41:4b

Field Verbum autem etiam valde potens non opponet se mihi.28 Iob, ed. Ziegler σ′ λόγος δὲ ἱκανοῦ ἀδιάπτωτος 252; σ′ λόγος δὲ καὶ ὁ ὑπερίσχυρος οὐκ ἀντικείσεταί μοι Syh New Field (Job 22–42, ed. Meade) HT E σ′ Wit1: Wit2: Attr: NonGr:

‫ְּוד ַבר־ּגְ בּורֹות וְ ִחין ֶע ְרּכֹו׃‬ καὶ λόγον δυνάμεως ἐλεήσει τὸν ἴσον αὐτοῦ. λόγος δὲ κἂν ὁ δυνατώτατος οὐκ ἀντιτεθήσεταί μοι lemma] ↓C (= 788′ 3005) Syh | λόγος δὲ ἱκανοῦ ἀδιάπτωτος] 252 λόγον] -ος O A-V-637-336′-754c Aeth Arm σ′] > 788′ Syh

‫ ܡܠܬܐ ܕܝܢ ܐܦ‬.‫ܣ‬. ̇ ‫ܐܝܕܐ ̇ܕܣܓܝ ܚܝܠܬܢܝܐ‬ ‫ܐܠ ܬܬܣܝܡ‬ ‫ܠܘܩܒܠܝ܀‬

Notes: The Greek wording for this fragment was cited for the first time in Ziegler’s Beiträge, and therefore Field’s retroversion of Syh was shown to be invalid.29 Sig28  After citing the Syriac text, Field says, “Middeld. retroverted: Σ. λόγον δὲ καὶ αὐτὸν πολλῆς ἰσχύος οὐκ ἀντιστήσεις μοι. I attempted: Σ. λόγος δὲ καὶ ὁ ὑπερίσχυρος οὐκ ἀντικείσεταί μοι. But Parsons. from Cod. 252 conveys: Σ. λόγος δὲ ἱκανοῦ ἀδιάπτωτος. Moreover, neither οὐκ ἀντικείσεταί μοι nor ἀδιάπτωτος pertains to the Hebrew, but rather I suppose it to be from a scholiast” (Field, 78, n8). 29  Hagedorn and Hagedorn, Nachlese, 410; Ziegler, Beiträge, 48; Field, Origenis Hexaplorum, 78.

The Dream for a ‘New Field’ Comes True

141

nificantly, manuscript Rahlfs 788 confirms the Greek wording. The reading of manuscript Rahlfs 252 does not align with the Hebrew, and it is doubtful whether it belongs to Symmachus, given the evidence of C and Syh.

Explanation Field retroverted Syh into Latin and not his usual Greek in smaller font. In his footnote, he expressed some doubt about the retroversion and whether οὐκ ἀντικείσεταί μοι pertained not to the Hebrew but to “a scholiast.” He was correct to wonder whether ἀδιάπτωτος in manuscript Rahlfs 252 related to the Hebrew. Ziegler recorded side by side the readings of manuscript Rahlfs 252 and Field’s retroversion from Syh in accordance with Göttingen’s aim for the second apparatus only to document the evidence and not to provide critical reconstruction. Ziegler’s Beiträge was the first to record the Greek reading and attribution of 3005 250 and thus show Field’s retroversion of Syh to be invalid.30 The New Field bases the reading of Symmachus on the C group, now including 788, and Syh. Manuscript Rahlfs 252 preserved the attribution to Symmachus, but its reading is not related to Symmachus but is probably an exegetical fragment. I included the reading of manuscript Rahlfs 252 in the Wit1 line because of the attribution to Symmachus and because it is traditionally associated with the textual history of this fragment but can now be shown not to be from Symmachus. It is not a variant of the Symmachus fragment of C and Syh. Finally, this Symmachus fragment illustrates why a critical edition of the hexaplaric fragments remains crucial for ongoing projects of Hebrew Bible textual criticism. In 2015, the fifth volume of Critique textuelle de l’Ancien Testament (CTAT) was published, and it did not include this fragment in its discussion of the textual problem for 41:4b. CTAT conveyed the discussion in Field where he suspected that neither manuscript Rahlfs 252 nor the Syrohexapla pertained to the Hebrew Text but to a scholiast.31 The problem with CTAT’s commentary is that it did not cite Ziegler’s Beiträge nor the Hagedorns’ Nachlese, which already presented the Symmachus fragment from manuscript Rahlfs 3005. The New Field publishes this fragment as a text that reflects Hebrew ‫וְ ֵאין ערכו‬, instead of MT’s ‫וְ ִחין‬, an observation already made by Ziegler.32 A new critical edition of the hexaplaric fragments should keep future researchers from omitting these important readings in the course of their research or at least one can hope. 3.5 Summary These few examples show the advance of the New Field upon earlier collections. First, the New Field is based upon a superior collation of the catenae manuscripts 30 Ziegler, Beiträge, 48. 31 Barthélemy, Critique textuelle de l’Ancien Testament: Tome 5, 437–8. 32 Ziegler, Beiträge, 48.

142

John D. Meade

of Job. Not only did this new collation lead to the discovery and inclusion of new hexaplaric readings, but also the rechecking of previously known readings in the manuscripts revealed that the lemmas in the second apparatus of Ziegler’s Iob were sometimes incorrect or not recorded in the first place. Second, Iob’s second apparatus, according to its aim, leaves too many hexaplaric readings in parallel format to be considered a critical edition of the hexaplaric fragments. That is, in too many cases Ziegler did not decide between readings but merely let them remain side by side. This method is suitable for a critical edition of the Old Greek but it is not suitable for a critical edition of the hexaplaric fragments. Third, despite some praise for the parenthetical approach to variants to be discussed below, the use of parentheses for variants is often too abbreviated to give the researcher a clear picture of the evidence. The New Field supplies a clean, critically reconstructed text of the hexaplaric fragment with variants and each manuscript supporting the variants in its apparatus. Given the descriptions and aims of Field, Ziegler, and Meade, i.e. the New Field for Job 22–42, and showing how they compare with each other, I turn now to Olivier Munnich’s interaction with the Hexapla Project’s New Field.

4. Criticism of the Hexapla Project’s “New Field” In an article from 2014, Olivier Munnich argued that the second apparatus of the Göttingen Septuaginta is a New Field, while also making several criticisms of the Hexapla Project’s New Field.33 First, after quoting Gerard Norton’s desire for an expanded Field, as Field was an expanded Montfaucon, Munnich emphasized that in the Göttingen second apparatus “un tel outil de travail existe déjà.”34 He gives several examples from some of the Göttingen Pentateuch and Prophets volumes where he believes the second apparatus handles these matters sufficiently. In an earlier response to Munnich, Alison Salvesen mentioned that the readings of the Three in Exodus 1:19 and 1:21 are not easily read and understood by researchers “due to the visual complexity of those entries.”35 Above, I showed a few examples where, at least for Job 22–42, the second apparatus handled the readings of the Three insufficiently, that is, if establishing the text of the Three was Ziegler’s aim, which it was not. Munnich believes that the aim of the Göttingen editions was to edit all forms of the Greek Old Testament including the revisions.36 Though I cannot see whether or not he indicates so in his article, Munnich’s criticism appears to be dependent on Detlef Fraenkel’s idea that the 33 Munnich, “Les révisions juives,” 141–90, esp. 175–90. 34 Munnich, “Les révisions juives,” 178. 35 Salvesen, “A ‘New Field’ for the Twenty-First Century?,” 300. 36 Munnich, “Les révisions juives,” 187.

The Dream for a ‘New Field’ Comes True

143

second apparatus should not only document the hexaplaric materials but also evaluate them critically.37 But Fraenkel’s view is not the principle for the second apparatus of the Göttingen editions, which, as stated above, is simply to document the hexaplaric readings and not evaluate them critically.38 Indeed, Ziegler’s Iob edition from 1982 does not attempt to establish the text of many fragments in the second apparatus, since it was not part of the aim of his project. The Hexapla Project’s New Field meets this need in a critical edition that expands upon Field. After attempting to defend the Göttingen second apparatus as a New Field, Munnich offers three specific criticisms of the Hexapla Project: (1) a choice between an edition of the hexaplaric fragments and a study of their manuscript tradition, (2) the place of the Hebrew Text, and (3) the relationship between the first and second apparatus, data from manuscripts and patristic sources. First, given Göttingen’s second apparatus, Munnich questions whether an exhaustive critical edition of the hexaplaric remains is needed or only an edition containing the longest, more original version of the fragment, not placing all the data in the apparatus. Furthermore, Munnich believes the New Field to be too analytic to the point where the whole fragment of a reviser cannot be appreciated because its apparatus also includes those sources where only a word is represented (see example Job 40:8ab above).39 I do not know why the New Field must choose between a critical edition and a study of the manuscript tradition. The recently published edition of the remains of Job 22–42 established the wording of the revisers and placed all variants in the apparatus. It has also relied on the Hagedorns’ new collation and manuscript stemma of the oldest Greek catenae. The study of the manuscripts and their stemma aided the editing of the materials. Munnich expressed concern over a too analytical edition that would lose sight of the sources preserving a longer fragment for the sources that preserve only a word or two of that same fragment. The examples given above show that he is concerned over previous iterations of the layout. The critically reconstructed lemma of the revisers can be clearly seen. The Wit1 line makes clear from what sources the longer lemma was taken and those sources which preserve shorter versions or a single word of the longer fragment. There is no need to choose between a critical edition and a study of the manuscripts. The revised layout of the New Field allows the lemma to be quite readable and unencumbered and gives the researcher access to the variants to the text.

37 Munnich, “Les révisions juives,” 187; Fraenkel, „Hexapla-Probleme im Psalter,“ 309–22, esp. 311. 38 Neuschäfer, “Die Göttinger Septuaginta-Ausgabe;” Neuschäfer, “Die kritische Edition des griechischen Alten Testaments.” 39 Munnich, “Les révisions juives,” 185–6.

144

John D. Meade

Second, although the original plan of the Hexapla Project was to include only the Hebrew consonantal text, as Munnich indicated,40 the recent editions in the form of dissertations except Phillip Marshall’s on Ecclesiastes, which was the first one (see bibliography below), have included the vocalized Masoretic Hebrew Text. This move facilitates research of the recentiores who mainly revised their Greek text according to the proto-Masoretic text and whose renderings mainly anticipate the reading tradition represented in the MT, even though the Three still preserve some variants to this tradition. Although I would agree with Munnich’s criticism of the original plan for the presentation of the unvocalized Hebrew text, the Hexapla Project had already revised its approach to the Hebrew some time before 2014. Munnich had criticized earlier samples of the New Field. The critical edition of Job 22–42 presents the Hebrew Text with traditional vocalization, which was the reading the revisers and Origen anticipated as shown from the remains of their versions. Third, Munnich questions whether the New Field obscures the relationship of the transmission of the Septuagint’s textual history and the hexaplaric remains in the margins of those manuscripts. Furthermore, and more acutely, he is concerned that the New Field “levels” these sources by disassociating the data in the second apparatus of Göttingen from the data in the first apparatus.41 Scholars continue to research from where the hexaplaric, marginal readings came. Someone excerpted them from Origen’s Hexapla, or more accurately, from his own revision of the Seventy with readings from the Three in the margins. But who excerpted Origen’s work, and when and where is not clear. And were there secondary collections of the hexaplaric marginalia?42 Thus, Munnich may place too much significance on the relationship between text and margin, since we do not yet know the nature of the relationship. What we do know is that the marginal readings ultimately came from the Hexapla, but how they migrated from Origen’s tome to the margins of Byzantine catenae manuscripts is still an open question. Be that as it may, the New Field does include the Hebrew Text and the Göttingen editions’ text for the purpose of situating the hexaplaric fragments. Furthermore, the New Field includes a Wit2 line for the purpose of relating all relevant data from the first apparatus to the hexaplaric fragment. That is, when a hexaplaric reading has interfered with or corrupted the transmission of the Old Greek text and the Göttingen editor has included that information in the first apparatus, the Hexapla Project now includes that information in its Wit2 line. This line was 40 Munnich, “Les révisions juives,” 186; cf. Romeny and Gentry, “Towards a New Collection,” 292. 41 Munnich, “Les révisions juives,” 187–8. 42 Meade, “An Analysis of the Syro-Hexapla of Job and Its Relationship to Other Ancient Sources.” For example, Syh Job and the Job catenae probably accessed different collections of hexaplaric fragments, since their respective sets of marginalia differ in no minor way.

The Dream for a ‘New Field’ Comes True

145

developed after the original descriptions of the Project, and unless Munnich has looked at the recent dissertations that have been completed, he probably would not be aware of this welcomed development. Lastly, the edition for Job 22–42 could not be criticized for “flattening” or “putting all the elements on the same plane.” The edition is dependent on the Hagedorns’ manuscript stemma of the oldest Greek catenae of Job and as a result its readings can be evaluated according to whether they appear in manuscripts higher or lower on the stemma. Furthermore, hexaplaric readings and patristic attributions appearing in the cII group (the Nicetas catena) were evaluated according to the newer critical editions of the commentaries so that we might be more certain of the provenance of the hexaplaric readings.43 In conclusion, it appears that Munnich has critiqued older samples and descriptions of the Hexapla Project and was not aware of much of the internal developments within the project itself. Göttingen’s second apparatus is very useful and serves the overall aim of establishing the text of the Old Greek by showing the remaining evidence of the revisions and especially those places where the revisions have entered the textual transmission of the Old Greek. But the second apparatus is not a sufficient critical edition of these fragments according to its own aim. For Job 22–42, the New Field’s fresh collation of the MSS and use of newer critical editions of the catenae and patristic commentaries combined with the method of Field for establishing the text of the revisers constitutes a real advance on the Göttingen Edition of Iob.

5. Conclusions Frederick Field’s magisterial work has become outdated. But the New Field is a real tribute to him and his work, for it has adopted his overall method for establishing the text of the hexaplaric remains even as the New Field adopts the fresher collation of the Göttingen editions and revisions. The New Field aims for a critical edition of the hexaplaric remains. It is not a mere documenting of the various readings of the Three in the manuscripts as the Göttingen editions often do in their second apparatus. The Göttingen editions include readings of the Three since that material serves to establish the text of the Old Greek and shows the later revision of that text. These aims are quite different, and therefore, the methods and approaches to these fragments are different. This article has tried to illustrate the differences between Field’s collection, the Göttingen editions, and 43 Henrichs, Hagedorn, Hagedorn, and Koenen, Kommentar zu Hiob: (Tura-Papyrus); Hagedorn, Der Hiob Kommentar des Arianers Julian; Hagedorn and Hagedorn, Olympiodor Diakon von Alexandria Kommentar zu Hiob; Hagedorn and Hagedorn, Johannes Chrysostomos Kommentar zu Hiob.

146

John D. Meade

the New Field and has attempted to show how the New Field, at least for Job 22–42, constitutes an advance in our knowledge of the original wording of the hexaplaric fragments and also their reception into Christian sources.

6. Appendix: Catalogue of Revised Readings In cursory fashion below, I catalogue the revisions and changes between the new edition of the Hexaplaric fragments of Job 22–42 and Ziegler’s second apparatus. In most cases, I supply the references so the reader can consult the new edition for further context. At times, I supply the attribution and lemma of the revised reading. In all cases, the reader should consult the edition for the evidence and commentary.

6.1  Newly Included Hexaplaric Readings and Attributions

The new edition includes twenty-two new hexaplaric fragments. The following fragments are not listed in Ziegler’s second apparatus and all of them except one have not been included in an edition until now. Field listed Job 39:18a σ′ πετομένη, but in a footnote he considered it to be a scholion, and Ziegler did not list it in his second apparatus. Initially, only the cI group preserved this fragment and attribution to Symmachus, and therefore it could have been interpreted as a scholion. The C (= 788 3005) group now confirms that this fragment and attribution belong to the oldest Greek catena, and therefore it should be listed among the hexaplaric materials of Job. Of these fragments, fourteen have an attribution preserved in the witnesses, while eight were assigned an attribution through considerations of translation technique. The newly included fragments are as follows: 23:3b 23:14a 24:1b–2a 24:10a 24:24ab

〈θ′〉 α′ θ′ 〈σ′〉 γ′ γ′

27:12b 27:14b 30:21a 31:9a 31:11ab 31:12a 34:18a 34:18a

σ′ 〈σ′〉 α′ θ′ σ′ σ′ σ′ ⟨α′⟩ ⟨σ′⟩

ἕως ἑτοιμασίας αὐτοῦ ὅτι ἀπαρτίσει ἀκριβασμόν καὶ γινώσκοντες αὐτὸν οὐκ οἴδασι τὰς ἡμέρας αὐτοῦ ἄνευ ἐνδύματος ὀλίγον καὶ οὐχ ὑπάρχει, καὶ ταπεινώσει αὐτὸν καθὼς πάντα τὰ συγκλασθησόμενα ματαιολογεῖτε οἱ ἀπόγονοι ἐστράφης εἰς ἀνελεήμονά μοι ὥρμησεν ὅπερ ἐστὶ μύσος καὶ τοῦτο ὑπερβάλλουσα ἀδικία πῦρ δὲ τὸ ἕως συντελείας ἀναλίσκον ἀποστάτα ἀπαίδευτε

The Dream for a ‘New Field’ Comes True

34:18a 36:2b 36:28a 38:12a 39:18a 40:17a 40:22b 42:17aα 42:17eδ

⟨θ′⟩ ⟨σ′⟩ ⟨α′⟩ γ′ σ′ σ′ λ′ θ′ θ′

147

ἄδικε [ἔτ]ι γὰρ περὶ θεοῦ [εἰ]σὶ λόγοι ⟨ῥοπαί⟩ ἦ ἀφ’ ἡμερῶν σου πετομένη τονώσει ἰτέαι μεθ’ ἡμῶν ὅτι θεὸς ἀναστήσει ὁ Ἰεμιναίων

6.2  Revision of Attribution and Lemma This category covers two types of revision: (1) five actual fragments which required revision of attribution and lemma of previously identified hexaplaric fragments, and (2) twenty-six places in Ziegler’s edition, where multiple attributions and lemmata are listed, which have been reassessed in order to ascertain the original attribution and lemma.44 There were five instances of the first type, and two of these instances require further explanation. The attribution to Aquila in Job 37:1a was a typographical error in Ziegler’s edition, which has been corrected here. Job 30:7a α′ appears as a variant in Ziegler’s second apparatus, but my edition has clarified that there are two fragments for that verse; thus, the Aquila fragment appears as a new fragment here. The five instances of this type are listed as follows with the attribution of Ziegler’s edition in parentheses: 23:8

α′ (⟨α′⟩)

28:16a 30:7a 31:13a 37:1a

⟨σ′⟩ (α′) α′ (σ′) σ′ (⟨σ′⟩) σ′ (α′)

ἰδοὺ ἀρχῆθεν πορεύσομαι καὶ οὐχ ὑπάρχει, καὶ ὀπίσω καὶ οὐ συνήσω αὐτόν οὐκ ἀναβληθήσεται φυτῶν εἰ ὑπερεφρόνησα καὶ περὶ τούτου

Two examples of the second type should be sufficient for this summary. (1) At 31:31a, Ziegler lists two fragments as follows: α′ εἰ μὴ εἶπον and σ′ θ′ οἱ ἄνδρες τῆς σκηνῆς μου, but there is no manuscript which attests these readings and at44  They are listed as follows: 22:18a α′ θ′, 24:2a α′, 27:20a α′ σ′, 28:1b λ′, 28:15a σ′, 29:25b σ′, 30:12c α′, 31:21a σ′, 31:31a α′ σ′ θ′, 31:40a θ′, 32:11c α′, 34:5b, 37:3a α′, 37:9a α′, 37:9b α′ θ′, 38:25a ⟨θ′⟩, 38:25a α′, 38:35b α′, 39:13a σ′, 39:19b σ′, 40:8b σ′, 41:1a θ′, 41:10a α′, 42:11g σ′, 42:14a α′, 42:14c α′ σ′.

148

John D. Meade

tributions. The C (= 788 3005) group attests one fragment with an original triple attribution. From this starting point, it is easier to see how the longer fragment was divided and the attributions were distributed the way they were in the cI group. (2) At 41:1a, Ziegler lists the following two fragments: σ′ ἰδοὺ ἡ ἀρχὴ αὐτοῦ ⟨διαψεύσεται⟩ 252; θ′ ἰδοὺ ἡ ἀρχὴ αὐτοῦ διεψεύσατο C′. The fragments are the same (manuscript Rahlfs 252 omits the verb), but the attributions are different. The new evidence of manuscript Rahlfs 788 shows that the attribution to Theodotion is part of the oldest Greek catena, and therefore, the attribution to Symmachus for the same fragment in manuscript Rahlfs 252 is incorrect. Manuscript Rahlfs 252 has also misplaced the fragment beside 41:3a, further suggesting its text and attribution are secondary.

6.3  Revision of Attribution

The new edition of the hexaplaric fragments of Job 22–42 revised the attributions of twenty-five previously established hexaplaric fragments. Of these fragments, twenty-three have attributions which are attested in manuscript evidence, while two are conjectures based on other considerations. In fourteen cases, there is a revision of an existing attribution (e.g. 29:2b). In six cases, old conjectures are confirmed by new manuscript evidence (e.g. 22:16b). In three cases, a new attribution was assigned based on manuscript evidence (e.g. 38:40a). At Job 42:11f, a new attribution was chosen by conjecture through new manuscript evidence and translation technique. Previously, Ziegler listed two fragments attributed to Sym, one based on manuscript Rahlfs 248 and the other based on the catena. The new evidence of manuscripts Rahlfs 788 and 3005 brings this conclusion into question. Rahlfs 3005 has an attribution to Aquila and Rahlfs 788 preserves a double attribution to Symmachus and Theodotion. Rahlfs 788 preserves the original double attribution, but it had already been corrupted by the insertion of Symmachus and the omission of Aquila, which Rahlfs 3005 confirms as genuine. These two manuscripts preserve the attributions to Aquila and Theodotion, and translation equivalents confirmed this textual reconstruction, since Symmachus usually translates ‫ רעה‬with κακός as the fragment in manuscript Rahlfs 248 attests, and Aquila and Theodotion (to a lesser degree) render ‫ רעה‬with κακία, as the oldest Greek catena attests. The revised attributions in the new edition are as follows: 22:16b 22:22a 23:9a 29:2a 29:2b

σ′ (⟨σ′⟩) α′ θ′ (α′) σ′ () σ′ (α′ σ′) θ′ (α′ σ′)

ποταμὸς παρασύρει τὸν θεμέλιον αὐτῶν νόμον καὶ οὐκ ἐπιλήψομαι τίς δῴη μοι κατὰ τοὺς μῆνας τοὺς πρώτους κατὰ τὰς ἡμέρας ἃς ὁ θεὸς ἐφύλασσέν με

The Dream for a ‘New Field’ Comes True

29:11b 30:1d 31:9b 33:7a 33:26b 37:20b 38:7b 38:12b 38:16b 38:24a 38:28a 38:39a 38:40a 39:1a 39:21a 40:8a 40:10a 42:2b 42:10c 42:10c 42:11f

α′ σ′ ([α′] σ′) α′ θ′ σ′ (λ′) λ′ (α′ σ′) σ′ (α′) α′ θ′ (α′) σ′ (⟨σ′⟩) α′ σ′ θ′ (α′ θ′) α′ (α′ θ′) σ′ (α′) α′ σ′ θ′ (α′ θ′) σ′ (⟨σ′⟩) α′ σ′ (σ′ θ′) σ′ () σ′ (θ′ (leg σ′)) α′ σ′ θ′ (α′ θ′) ⟨α′⟩ (σ′ (leg α′) α′ (⟨α′⟩) θ′ (οἱ ἄλλοι) σ′ θ′ () α′ (⟨α′⟩) ⟨α′ θ′⟩ (σ′)

149

ἐμαρτύρησεν περὶ ἐμοῦ μετὰ τῶν κυνῶν τῶν προβάτων μου ἐνήδρευσα ἐκθαμβήσει ἐν ἀλαλαγμῷ ἐὰν εἴπῃ ἀνὴρ καταποθήσεται υἱοὶ θεοῦ ἐγνώρισας τῷ ὄρθρῷ τόπον αὐτοῦ ἐν ἐξιχνιασμῷ φῶς ἆρά ἐστιν ὑετῷ πατήρ θήραν ὅτι φωλεύσουσιν νεβρῶν ἐν πέτρᾳ ἐν κοιλάδι μήτι καίπερ ἀκυρώσεις κρίσιν μου κόσμησαι δὴ ὑπερφερείᾳ καὶ μετεωρότητι καὶ οὐκ ἀφαιρεθήσεται ἀπὸ σοῦ ἐνθύμημα ※ εἰς διπλασιασμόν εἰς δευτέρωσιν περὶ πάσης τῆς κακίας

6.4  Revision of Lemma

There were sixty-five places where the lemma of Ziegler’s second apparatus was revised but the attribution was confirmed.45 Not all these revisions are equally significant: some consist of minor changes, while others consist of more substantive ones. Job 23:6a σ′ furnishes an example of minor change. Ziegler’s second apparatus had: διαδικάζεταί μοι, while this edition has δικάζεταί μοι based on manuscript Rahlfs 3005 and consideration of translation equivalents. Job 27:9a λ′ was not revised with regard to the wording of the lemma but the placement of the fragment itself. The Edition listed it at 9b, but the evidence of manuscript

45  They are listed as follows: 22:3b σ′, 22:17b θ′, 23:6a σ′, 23:8 σ′, 24:5b σ′, 24:12c σ′, 24:15c σ′, 24:22b σ′, 24:23 σ′, 24:25b σ′, 25:2a α′ θ′, 25:3b σ′, 25:5a α′, 26:5 α′, 26:5 σ′, 26:14bc σ′, 27:1 α′, 27:9a λ′, 27:23a α′, 28:18a σ′, 28:19b σ′, 28:25–26a α′, 28:26b σ′, 29:4a σ′, 29:4b σ′, 29:25c σ′, 30:2b–3a α′, 30:2b–3a σ′, 30:4a α′, 30:7a σ′, 30:17a λ′, 30:27a σ′, 30:30a λ′, 31:6a α′, 31:29b σ′, 32:1b σ′, 33:13 σ′, 34:11b σ′, 34:17a σ′, 34:33b σ′, 35:6a σ′, 35:16b σ′, 36:5b–6a α′, 36:23a σ′, 36:33a σ′, 37:6b σ′, 37:10a σ′, 37:13a σ′, 37:14b θ′, 37:15a σ′, 37:18a α′, 38:9a σ′, 38:25b σ′, 38:32a σ′, 38:38 α′, 39:4b σ′, 39:24b σ′, 39:30b σ′, 40:16a σ′, 40:30a σ′, 41:2a α′ θ′, 41:4b σ′, 42:5a λ′, 42:7c θ′, 42:8c α′ θ′.

150

John D. Meade

Rahlfs 3005 and considerations of translation technique show that it should be listed at 9a. Some revisions were more substantive. For example, at 28:25–26a α′, the Edition has the following: καὶ ὕδωρ ἐσταθμίσατο ἐν καταμετρήσει, while the new Edition based on the new evidence of 257 has: καὶ ὕδωρ ἐσταθμίσατο ἐν καταμετρήσει ἐν τῷ ποιῆσαι αὐτὸν ὑετῷ ἀκριβασμόν. The expanded lemma follows the Hebrew text and is found in a reliable witness. At 30:17a λ′, the Edition has νυκτός, while this new edition, based on the new evidence of 257, has: συγκέκοπται νυκτός. The expansion follows the Hebrew and is preserved in a reliable witness.

6.5  Unattributed Scholia

In Ziegler’s edition, there were eleven unattributed scholia.46 Of these instances 23:9a and 38:40a were found to be hexaplaric (see Revision of Attribution and Revision of Attribution and Lemma above). In manuscript Rahlfs 157, Job 23:16b (bis) contains two fragments, which are probably related to the Nicetas catena (see discussion in chapter three of my edition). When the Hagedorns re-collated manuscript Rahlfs 523, they found that this manuscript supported the lemma in 23:15c.47 Job 26:13b probably records a hexaplaric reading, but there is no way to determine the author. The rest of the instances are listed in chapter three.

6.6  Removed Readings

There were also four readings in Ziegler’s second apparatus that needed to be removed from the corpus for various reasons. Job 24:13b σ′ is not hexaplaric. The fragment of Job 31:2a α′ was caused by misreading the attribution to Aquila (see chapter two). Job 38:27b was not found in any manuscript which the Hagedorns collated.48 Job 41:21a χόρτος α′ θ′ arose from a misreading of manuscript Rahlfs 255.

7. Bibliography Barthélemy, Dominique, Critique textuelle de l’Ancien Testament: Tome 5: Job, Proverbes, Qohélet et Cantique des Cantiques (Fribourg: Academic Press; Göttingen: Vandenhoeck & Ruprecht, 2015).

46  The references are as follows: 22:12a, 22:29b, 23:9a, 23:15c, 23:16b (2x), 26:13b, 28:16b, 30:3a, 30:16a, 38:40a. 47  Hagedorn and Hagedorn, Nachlese, 400. 48  The Hagedorns shared their unpublished Kollationen der hexaplarischen Fragmente des Buchs Hiob with me, which contains no entry for ἐκβλαστήσει (cf. Ziegler’s, Iob, 388) from any of their manuscripts. I have also rechecked Ra 788 p. 170, and the reading is not there.

The Dream for a ‘New Field’ Comes True

151

Burris, Kevin, A Critical Edition of the Hexaplaric Fragments of Numbers 1–18 (Ph.D. diss., The Southern Baptist Theological Seminary, 2009). Ceulemans, Reinhart, A Critical Edition of the Hexaplaric Fragments of the Book of Canticles, with Emphasis on their Reception in Greek Christian Exegesis (Ph.D. diss., Katholieke Universiteit Leuven, 2009). Drusius, Johannes, Veterum interpretum Graecorum in totum Vetus Testamentum fragmenta (Arnhemiae: apud Iohannem Ianssonium Bibliopolam, 1622). Field, Frederick, Origenis Hexaplorum quae supersunt sive veterum interpretum Graecorum in totum Vetus Testamentum fragmenta (Oxford: Clarendon, 1875). Fraenkel, Detlef, “Hexapla-Probleme im Psalter,” in Der Septuaginta-Psalter und seine Tochterübersetzungen, eds. A. Aejmelaeus and U. Quast. MSU 24 (Göttingen: Vandenhoeck & Ruprecht, 2000), 309–322. Gignac, Francis T, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, Vol. 1. Phonology. Testi e documenti per lo studio dell‘antichità 55 (Milano: Istituto Editoriale Cisalpino, 1976). Hagedorn, Dieter, Der Hiob Kommentar des Arianers Julian. PTS 14 (Berlin: De Gruyter, 1973). Hagedorn, Ursula and Dieter Hagedorn, Die älteren griechischen Katenen zum Buch Hiob: Band I Einleitung, Prologe und Epiloge, Fragmente zu Hiob 1,1–8,22. PTS 40 (Berlin: De Gruyter, 1994). —, Die älteren griechischen Katenen zum Buch Hiob: Band II Fragmente zu Hiob 9,1–22,30. PTS 48 (Berlin: De Gruyter, 1997). —, Die älteren griechischen Katenen zum Buch Hiob: Band III Fragmente zu Hiob 23,1–42,17. PTS 53 (Berlin: De Gruyter, 2000). —, Die älteren griechischen Katenen zum Buch Hiob: Band IV Register Nachträge und Anhänge. PTS 59 (Berlin: De Gruyter, 2004). —, Johannes Chrysostomos Kommentar zu Hiob. PTS 35 (Berlin: De Gruyter, 1990). —, Nachlese zu den Fragmenten der jüngeren griechischen Übersetzer des Buches Hiob. NAWG 10 (Göttingen: Vandenhoeck & Ruprecht, 1991). —, Olympiodor Diakon von Alexandria: Kommentar zu Hiob. PTS 24 (Berlin: De Gruyter, 1984). Henrichs, Albert, Ursula Hagedorn, Dieter Hagedorn, and Ludwig Koenen, Kommentar zu Hiob: (Tura-Papyrus). Papyrologische Texte und Abhandlungen, 1–3, 33,1 (Bonn: Rudolf Habelt, 1968). Holmes, Robert / James Parsons, Vetus Testamentum Graecum cum variis lectionibus, 5 vols. (Oxford: Clarendon, 1798–1827). Kreissig, Johann Gottlieb, Observationes Philologico-Criticae in Jobi Cap. XXXIX. vers. 19–25 (Leipzig: Goethio Commissum, 1802). Law, T. M, “A History of Research on Origen’s Hexapla: From Masius to the Hexapla Project,” BIOSCS 40 (2007): 30–48. Marshall, Phillip S., A Critical Edition of the Hexaplaric Fragments of Ecclesiastes (Ph.D. diss., The Southern Baptist Theological Seminary, 2007).

152

John D. Meade

McClurg, Andrew H, A Critical Edition of the Hexaplaric Fragments of Numbers 19–36 (Ph.D. diss., The Southern Baptist Theological Seminary, 2011). Meade, John D, “An Analysis of the Syro-Hexapla of Job and Its Relationship to Other Ancient Sources,” Aramaic Studies 14.2 (2016): 212–41. —, “The Significance of Ra 788 for a Critical Edition of the Hexaplaric Fragments of Job,” in XV Congress of the International Organization for Septuagint and Cognate Studies: Munich, 2013, ed. W. Kraus et al. SBLSCS 64 (Atlanta: Scholars Press, 2016), 109–131. —, A Critical Edition of the Hexaplaric Fragments of Job 22–42. Origen’s Hexapla: A Critical Edition of the Extant Fragments 1 (Peeters: Leuven, 2020). Montfaucon, D. Bernard de, Hexaplorum Origenis quae supersunt, multis partibus auctiora, quam a Flaminio Nobilio et Joanne Drusio edita fuerint. Ex manuscriptis & ex libris editis eruit & notis illustravit, 2 vols (Paris: Ludovicus Guerin, 1713). Munnich, Olivier, “Les révisions juives de la Septante. Modalités et fonctions de leur transmission. Enjeux éditoriaux contemporains,” in La Bible juive dans l’Antiquité, eds. Rémi Gounelle and Jan Joosten. HTB 9 (Éditions du Zébre; Prahins, 2014), 141–190. Neuschäfer, Bernhard, “Die Göttinger Septuaginta-Ausgabe. Standortbestimmung eines editorischen Jahrhundertprojekts. Internationale Fachtagung, Göttingen, 28.–30. April 2008.” Editio 22 (2008): 241–245. —, “Die kritische Edition des griechischen Alten Testaments. Anspruch und Aufgabe des Göttinger Septuaginta-Unternehmens,” in Jahrbuch der Akademie der Wissenschaften zu Göttingen (Vandenhoek & Ruprecht: Göttingen, 2004), 129–139. Nobilius, Flaminius, Vetus Testamentum secundum LXX Latine redditum (Rome: Georgius Ferrarius, 1588). Norton O.P., Gerard J., “Collecting Data for a New Edition of the Fragments of the Hexapla,” in IX Congress of the International Organization for Septuagint and Cognate Studies, Cambridge 1995, ed. Bernard A. Taylor. SBLSCS 45 (Atlanta: Scholars Press, 1997), 251–262. Reider, Joseph / Nigel Turner, An Index to Aquila: Greek-Hebrew, Hebrew-Greek, Latin-Hebrew. With the Syriac and Armenian Evidence. VTSup 12 (Leiden: Brill, 1966). Romeny, Bas ter Haar and Peter J. Gentry, “Towards a New Collection of Hexaplaric Material for the Book of Genesis,” in X Congress of the International Organization for Septuagint and Cognate Studies, Oslo 1998, ed. Bernard A. Taylor. SBLSCS 51 (Atlanta: Society of Biblical Literature, 2001), 285–299. Salvesen, Alison, “A ‘New Field’ for the Twenty-First Century? Rationale for the Hexapla Project, and a Report on Its Progress,” in The Text of the Hebrew Bible and Its Editions: Studies in Celebration of the Fifth Centennial of the Complutensian Polyglot, eds. Andres Piquer Otero and Pablo Torijano Morales (Leiden: Brill, 2017), 286–309. Schleusner, Johann Friedrich, Novus Thesaurus Philologico-Criticus. Lexicon in LXX et Reliquos Interpretes Graecos, ac Scriptores Apocryphos Veteris Testamenti. 3 vols. (London: Priestly, 1822). Woods, Nancy, A Critical Edition of the Hexaplaric Fragments of Job: Chapters 1–21 (Ph.D. diss., The Southern Baptist Theological Seminary, 2009).

The Dream for a ‘New Field’ Comes True

153

Young, Patrick, Catena Graecorum Patrum in beatum Iob: collectore Niceta Heracleae metropolita ex duobus mss. Bibliothecae Bodleianae codicibus, Graecè nunc primùm in lucem edita, & Latinè versa. (London: Ex typographio Regio, 1637). Ziegler, Joseph, Beiträge zum griechischen Iob. MSU 18 (Göttingen: Vandenhoeck & Ruprecht, 1985). —, Iob. Septuaginta Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum 11.4 (Göttingen: Vandenhoeck & Ruprecht, 1982).

Benjamin Kantor Discovering the Secunda Insights from Preparing a New Critical Edition of the Second Column of Origen’s Hexapla The second column (Secunda) of Origen’s (185–254 CE) Hexapla, which contains Greek transcriptions of a Biblical Hebrew tradition from the Roman period, constitutes the earliest and most extensive evidence for how Biblical Hebrew was pronounced in antiquity. Indirectly, it also constitutes a witness to a type of the Hebrew consonantal text circulating in Roman-period Palestine. It is a marvel, then, that a critical edition has never been published of such an important text. The need for a new critical edition of the second column of the Hexapla – and of a “judiciously retroverted” first column of the Hebrew consonantal text alongside it – was already touched on by Flint in the Rich Seminar on the Hexapla held at Oxford in 1994.1 This gaping desideratum, relevant for both Hexapla scholars and historical Hebrew linguists, is finally being rectified in the form of the first comprehensive critical edition of the Secunda, set to appear in the not-so-distant future as part of the Hexapla Project series published with Peeters. The rest of this paper will provide the reader with some background on previous text-critical work on the Secunda, outline some of the most important methodological considerations in producing a new critical edition and share a selection of new insights coming out of the work on the new critical edition. The paper will conclude by outlining a number of ways in which this new edition may impact a number of disparate fields, ranging from textual criticism, to historical Hebrew linguistics, to patrology. Because this paper is designed to serve as a sort of introduction to the text critical issues surrounding the Secunda and the new critical edition thereof, the various issues taken up in each section (1. background, 2. methodological considerations, 3. new insights, 4. conclusions) are presented in question and answer form.

1. Background The present discussion on the Secunda begins by surveying a number of important issues about its ancient background, its treatment by scholars in the modern era, and why the present state of scholarship necessitates the publication of a new critical edition: 1 Flint, “Columns I and II,” 131–132.

156

Benjamin Kantor

1.1  What is the Secunda?

As noted earlier, the Secunda constitutes a Greek transcriptional text of the Hebrew Bible, in which, e.g., the word ‫ ָׁשלֹום‬is written as σαλωμ and the word ‫ַּביִת‬ is written as βαϊθ. Its very designation as “the Secunda” (i.e., ‘second’) conveys something about its original literary and codicological context. The Secunda was the second of six columns in what may be considered the world’s first parallel Bible, known as the Hexapla (i.e., ‘six-fold’). The first column of the Hexapla (I) contained Hebrew in Hebrew letters, the second column (II) contained Hebrew in Greek letters, the third column (III) contained Aquila’s Greek translation of the Bible, the fourth column (IV) contained Symmachus’ Greek translation of the Bible, the fifth column (V) contained a recension of the Septuagint, and the sixth column (VI) contained Theodotion’s Greek translation of the Bible.2 A visual representation of the Hexapla, based on the oldest manuscript witness to its columnar and codex form, known as the Cairo Genizah palimpsest (Cambridge T–S 12.182), is depicted below (Figure 1): Figure 1

At this point, it is necessary to make an important terminological distinction, since many scholars fail to do so regularly. In many instances, scholars refer to the “transcriptions of the Hexapla” or the “Hexapla transcriptions” as if these terms are synonymous with the “transcriptions of the second column”. This, of course, is not at all the case. Aquila (III), Symmachus (IV), the Septuagint (V), and Theodotion (VI) all transcribe Hebrew words on occasion instead of translating them. Where they do opt for transcription over translation, they often 2  Though the Hexapla is generally thought to have had six columns, this is a bit of an over-simplification. Some portions of the Hexapla actually contained an additional one or two columns, which have been called the “Quinta” (i.e., ‘fifth’) and “Sexta” (i.e., ‘sixth’), so named because of their order among the translations (and not among the Hexapla as a whole.

Discovering the Secunda

157

e­ xhibit a transcription tradition that is distinct from that of the Secunda. Therefore, the phrase “transcriptions of the Hexapla” or “Hexapla transcriptions” is vague at best and misleading at worst. A proper distinction between “the Secunda transcriptions,” on one hand, and “the Transcriptions of the Three,” “the transcriptions of the LXX,” etc., on the other hand, should be maintained. Although the Hexapla was composed by Origen in Caesarea Maritima during the third century CE, none of the texts supplying the material for its six columns were composed by Origen himself. Despite the fact that column II does not have a separate (i.e., non-Hexaplaric) attestation in antiquity like columns I and III–VI do, Origen lacked the requisite skills to produce such a text. Rather, he probably encountered the practice of transcribing Hebrew into Greek among the Jewish community of Caesarea and appropriated it for his biblical synopsis known as the Hexapla. Whether he found a fully transcribed text of the Bible among the Caesarean Jews or only partial extracts, it is likely that the transcription texts underlying the Secunda ultimately have a Jewish origin. If these texts did not cover the entirety of the Old Testament, it is likely that Origen enlisted Jewish aids in applying the same convention to the entire Bible for the purposes of the Hexapla, forming what would come to be known as the Secunda.3

1.2  Where can the Secunda be found in print today? Many students and scholars alike attempting to find a simple and straightforward presentation of the text of the Secunda only end up frustrated and dismayed. Such interested persons are typically first referred to Field’s 1875 tomes on the Hexapla, through which one must comb meticulously only to find a Secunda citation about once every twenty or thirty pages.4 Moreover, when the reader does finally find such a citation, most of which are taken from the church fathers or LXX scholia, they can have little confidence that it represents a reliable textual witness. Finally, because most of the extant Secunda material is contained in the Ambrosiana palimpsest (Ambrosiana O. 39 sup.), which was not discovered until two decades after Field’s edition was published (see below), there are actually relatively few transcriptions in Field. All in all, for those interested in appreciating the Secunda as a text in its own right, Field’s volumes are impractical, unreliable, and incomplete. Although the more recent Göttingen editions of the Septuagint have included more up-to-date and reliable text critical references to the Secunda in the second apparatus, they remain scattered across many volumes and sometimes incomprehensive with respect to the material derived from the church fathers. Further, 3  The original Sitz im Leben for such transcription texts among the Jewish community is not entirely clear, though they were probably used as a sort of vocalization aid, whether for didactic, liturgical, or exegetical purposes. For more, see Kantor, The Second Column, 10–47. 4  Field, Origenis Hexaplorum quae supersunt.

158

Benjamin Kantor

the data from the Ambrosiana palimpsest is not included in the critical edition of the Psalms.5 Though Field’s volumes did not facilitate easy access to the text of the Secunda, all of this would begin to change with the discovery of the Ambrosiana palimpsest in Milan by Giovanni Mercati at the end of the nineteenth century. This discovery immediately led to a number of publications that would make the text of the Secunda more accessible. In the coming decades, several scholars would publish grammatical treatments of the Secunda material from a historical-Hebrew perspective, based both on the citations in Field and on the text of the Ambrosiana palimpsest.6 Because these works focus specifically on the grammar of the second column, unlike Field, the Secunda material is presented in a concentrated and accessible fashion. However, none of these early scholars include a comprehensive presentation of the text of the second column itself, but rather organize the Secunda material around phonological and morphological categories. Finally, it should also be noted that the Secunda material was included in Hatch and Redpath’s index to the Septuagint, though it is mixed with other sources and sometimes unreliably transcribed.7 It was not actually until 1958/1965 that Giovanni Mercati finally published the editio princeps of the Ambrosiana palimpsest and furnished the discipline with its first print edition of a continuous text of the Secunda.8 This edition would continue to serve as the main textual resource for the Secunda for decades to come. It has only been in the last decade that Yuditsky has re-examined the Ambrosiana palimpsest and suggested a number of corrections.9 His transcription of the second column portion of the Ambrosiana palimpsest is published as an appendix in his grammatical treatment of the Secunda.10 While both Mercati’s text and Yuditsky’s update remain the most helpful (and only) presentations of the Secunda material as a continuous text, unfortunately, there are not a few transcription 5  There is actually a monumental long-term project currently underway in Göttingen, which began in early 2020, to produce a new critical edition of the Psalms and Odes known as the Psalter Project (“Editio critica maior des griechischen Psalters”). This will undoubtedly resolve such issues with the current Göttingen edition. 6 Margolis, “The pronunciation of the shewa according to new Hexaplaric material”; Pretzl, “Hexapla des Origenes”; Speiser, “Hexapla”; Sperber, “Hebrew based upon Greek and Latin transliterations”; Brønno, Studien über hebräische Morphologie und Vokalismus. For more recent treatments, see Janssens, Origen’s Secunda; Yuditsky, Diqduq; Kantor, “The Second Column.” 7  Hatch and Redpath, A Concordance to the Septuagint. There are a number of readings incorrectly transcribed in the appendix of vol. 2 of Hatch and Redpath’s concordance: e.g., ισαββουνι (wrong) vs. ϊσ̓ο′ββουνι (correct) (Ps. 48:6 [fol. 29v–28r]); οὐβανγαϊβ (wrong) vs. ου᾽βανγαϊμ (correct) (Ps. 89:33 [fol. 17v–24r]). 8 Mercati, Psalterii Hexapli Reliquiae; Mercati, Psalterii Hexapli Reliquiae, osservazioni. 9 Yuditsky, “ʕIyun ḥ adash bexitve yad O 39 misifriyat ambrosiana shebemilano.” 10 Yuditsky, Diqduq, 235–263.

Discovering the Secunda

159

errors in both.11 Moreover, it must also be reiterated here that the Ambrosiana palimpsest only contains roughly two-thirds of the total Secunda material.

1.3  Why is there a need for a (new) critical edition of the Secunda? As may be clear from the preceding survey, all previous textual work that has been done on the Secunda falls short in one or more of the following three areas: (i) comprehensiveness, (ii) reliability, and (iii) accessibility. The new critical edition will rectify each of these issues. In terms of comprehensiveness, the new critical edition will include not only the material from the Ambrosiana palimpsest, but will also gather all other columnar witnesses, quotations from the church fathers, scholia in LXX manuscripts, and citations in the Syro-Hexapla. (For more on these four main categories of witnesses, see 2.1 below.) In terms of reliability, the latest and most up-to-date manuscript readings – scores of these are based on personal examination – will serve as a basis for the text. In terms of access and ease of use, the continuous text of the Secunda will finally be presented as a standalone text with a conventional critical apparatus (on the format, see 2.7 below).

2. Methodological Considerations for a New Critical Edition Because much of the Secunda material is attested only secondarily as brief quotations in various sources such as the church fathers or scholia in LXX manuscripts, it is not always clear if a Greek transcription of Hebrew contained in such quotations should be regarded as deriving from the second column. Surely not all Greek transcriptions of Hebrew quoted in early Christian sources were derived from the Secunda. One of the biggest challenges, then, in gathering the material for the Secunda edition is determining which transcriptions to include in the text. The most pertinent issues tied to this problem are outlined below (see 2.5), through which it is possible to better understand both how to gather Greek transcriptions of Hebrew among early sources and what sort of criteria are necessary for identifying genuine Secunda material among such transcriptions.

2.1  What types of witnesses are extant for the Secunda?

Although portions of the Hexapla were frequently used, copied, or cited for several centuries after its original composition during Origen’s lifetime, it was eventually lost or destroyed, probably between the fourth and seventh centuries CE. Only a handful of “direct” or “columnar” witnesses remain – the terms “direct” or “columnar” in this context refer merely to those witnesses of the Hexapla that 11  e.g., Mer. οὐβαμοτ (wrong) and Yud. ουβαμωτ (wrong) vs. οὐβομωτ (correct) (Ps. 46:3 [fol. 74r–69v]) and Mer./Yud. ιοβαδου (wrong) vs. ιεβαδου (correct) (Ps. 49:11 [fol. 27r–30v]).

160

Benjamin Kantor

preserve the original columnar format, even if not all of these witnesses are direct descendants (in the purest sense of the word) of the Hexapla.12 As such, the term “columnar” should be preferred. There are five such “columnar” witnesses, all of which contain at least part of the second column:13 1. Cambridge, Cambridge University Library T–S 12.182 (Rahlfs 2005) – VI/VII 2. Milan, Ambrosiana O. 39 sup. (Rahlfs 1098) – IX 3. Rome, Codex Vaticanus Barberinus Graecus 549 (Rahlfs 86) – X 4. Milan, Ambrosiana B. 106 sup. (Rahlfs 113) – 966/967 CE 5. Rome, Vaticanus Graecus 1747 (Rahlfs 271) – X/XI14 In addition to these five columnar witnesses, quotations of the Secunda are also found in the writings of the following church fathers: Origen (3rd c. CE), Eusebius of Caesarea (3rd–4th c. CE), Epiphanius of Salamis (4th c. CE), John Chrysostom (4th–5th c. CE), Jerome (4th–5th c. CE), Cyrill of Alexandria (4th–5th c. CE), Theodoret of Cyrus (5th c. CE), and Procopius of Gaza (5th–6th c. CE). In addition to the writings of these fathers, there may also be Secunda quotations – or at least valuable transcriptions for comparison – in the writings of Basil of Caesarea (4th c. CE) and Evagrius (4th c. CE). The rest of the Secunda material is found in marginal notes in manuscripts of the Septuagint and the 12  In reality, it is only Rahlfs 2005 that can securely be regarded (codicologically) as a direct descendant of the Hexapla. In every other case, the columnar text is presented in the context of a commentary or catena. As such, in some cases, such as Rahlfs 86, one might even argue that the columnar form is reconstructed and not a copy of a text with a columnar form. On the other hand, with the exception of Rahlfs 271, all the other columnar witnesses exhibit characteristics that would imply an authentic copy from a columnar form (which the later copies preserved) of the Hexapla at some point in transmission history. This is probably most apparent from the tendency to write multiple transcribed words on the same line (see 3.2 below). Also—again with the possible exception of Rahlfs 271—the so-called “direct witnesses” appear to exhibit the most reliable (even if difficult) readings in comparison with other sources. Nevertheless, for the sake of clarity and objectivity, these witnesses will henceforth be referred to as “columnar” witnesses. 13  On Rahlfs 113 and Rahlfs 271, cf. Albrecht, “Göttingen Septuagint.” 14  Alongside the other columnar witnesses, it might be appropriate to look a bit askance at Vaticanus Graecus 1747 (Rahlfs 271). Although it presents Ps 1:1–2a in two adjacent columns, the left column containing Hebrew in Greek letters and the right column containing a Greek translation of the Hebrew, the Greek translation does not appear to match any of the other Hexaplaric columns, exhibiting features that suggest it was modified by the scribe to better match the transcription column. In fact, this witness has the appearance of some kind of scribal exercise rather than a formal citation of the Hexapla. This is supported by the fact that a Greek transcription of the Hebrew alphabet is also written elsewhere on the page. This is in addition to the very brief quotation of Ps 112:1 that is also presented in columnar form. Nevertheless, notwithstanding the “scribal-exercise” features of the witness, the form and transcription conventions suggest that at least the left columns of Ps 1:1–2a and Ps 112:1 are indeed taken from the Hexapla.

Discovering the Secunda

161

Syro-Hexapla. All in all, this means that there are four main categories of witnesses for the Secunda material: (i) columnar witnesses, (ii) quotations in the church fathers, (iii) marginal notes in Septuagint manuscripts, and (iv) marginal notes in the Syro-Hexapla.

2.2  How is a Secunda quotation introduced/cited among the church fathers?

This question is especially important considering the fact that around one-third or one-fourth of all Secunda material is found in the church fathers, even though it is strewn across hundreds of passages as only brief quotations. The patristic authors tend to introduce Secunda transcriptions by some variation on one of the phrases listed below: 1.

2.

3. 4.

5.

6. 7.

τὸ ἑβραϊκόν ‘the Hebrew’ a. τὸ ἑβραϊκόν (ἔχει/περιέχει/εἶπεν) ‘the Hebrew (has/contains/says)’ b. ἐν τῷ ἑβραϊκῷ ‘in the Hebrew’ ἡ ἑβραϊκὴ λέξις ‘the Hebrew word/text’ a. ἡ ἑβραϊκὴ λέξις (περιέχει) ‘the Hebrew word/text (contains)’ b. ἐν τῇ ἑβραϊκῇ ‘in the Hebrew’ ὁ ἑβραῖος ‘the Hebrew’ a. ὁ ἑβραῖος (ἔχει/φησιν) ‘the Hebrew (has/says)’ οἱ ἑβραῖοι ‘the Hebrews’ a. παρὰ (τοῖς) ἑβραίοις ‘among (the) Hebrews’ b. κατὰ (τοὺς) ἑβραίους ‘according to (the) Hebrews’ ἡ ἑβραϊκὴ φωνή/γλῶττα ‘the Hebrew sound/language’ a. ἡ ἑβραϊκὴ φωνή ‘the Hebrew sound/language’ b. κατὰ τὴν ἑβραϊκὴν φωνήν ‘according to the Hebrews sound/language’ c. κατὰ τὴν ἑβραίων γλῶτταν ‘according to the sound/language of the Hebrews’ d. κατὰ τὴν ἑβραίων φωνήν ‘according to the sound/language of the Hebrews’ e. τῇ ἑβραΐδι διαλέκτῳ ‘in the Hebrew dialect/language’ ἑβραϊστί ‘in Hebrew’ Ø – i.e., no introductory phrase

Sometimes variation in the usage of these terms may indicate a substantial difference in meaning. For example, when phrases 1–2 introduce a Greek transcription of Hebrew, they almost always refer to the Secunda. Phrases 3–6 are not as specific, often being used for general Hebrew linguistic traditions apart from the text. On the other hand, there also appear to be many instances in which the choice of one term over another is merely stylistic. For example, the

162

Benjamin Kantor

use of ὁ ἑβραῖος ‘the Hebrew’ (masc. sg.) rather than τὸ ἑβραϊκόν ‘the Hebrew’ (neutr. sg.) to introduce a Secunda quotation can be a stylistic choice, seeing as it occurs almost exclusively and overwhelmingly in Chrysostom’s Expositiones in Psalmos. Moreover, references to ἡ ἑβραϊκὴ λέξις ‘the Hebrew word/text’ seem to be a favorite of Eusebius. The absence of any reference to ‘the Hebrew’ (phrase 7) before a transcription is not common, but it does occur. In such cases, the context (e.g., a list of translations) may elucidate the source of the transcription. While this list is not comprehensive, it accounts for the overwhelming majority of Secunda quotations in the writings of the church fathers. Several illustrative passages of such citations of the Secunda among the church fathers are cited below (Text 1–4):15 Text 1: Epiphanius, Ancoratus, 97.1–2 “Ῥουβεὶμ πρωτότοκός μου καὶ ἀρχὴ τέκνων μου, ἐξύβρισας ὡς ὕδωρ μὴ ἐκζέσῃς. ἀνέβης γὰρ εἰς τὴν κοίτην τοῦ πατρός σου καὶ ἀπεκάλυψας φησίν, οὗ ἀνέβης”. ἐν δὲ τῷ Ἑβραϊκῷ ἐλθωθάρ, ὅπερ ἐστὶν ἑρμηνευόμενον μὴ ἀνακάμψῃς ἢ μὴ προστεθείης ἢ πάλιν μὴ περισσεύσῃς, ψιλούμενον δὲ τῇ λέξει μὴ ἐκζέσῃς οἱ δὲ ἑρμηνευταὶ ἐξέδωκαν οὕτως.

“Reuben, my firstborn and beginning of my children, you have broken out insolently as water, do not break out! For you went up on the bed of your father and you revealed, he said, where you went up”. And in the Hebrew ἐλθωθάρ, which is interpreted as ‘do not return’ or ‘do not add’ or even ‘do not excel’, but the simple explanation for the word is ‘do not break out’. And the interpreters translated thus.

15  For the text of Epiphanius, see Holl, Epiphanius I: Ancoratus und Panarion haer. For the text of Eusebius’ Commentarius in Isaiam, see Ziegler, Eusebius Werke 9: Der Jesajakommentar. For the texts of Eusebius’ Commentaria in Psalmos and Chrysostom’s Expositiones in Psalmos, see Migne’s editions (vol. 23 and vol. 55).

Discovering the Secunda

163

Text 2: Eusebius, Commentaria in Psalmos, 23.1149 τηρητέον πρῶτον μέν, ὡς ἡ πέμπτη ἔκδοσις συνᾴδει τοῖς Ἑβδομήκοντα, φήσασα καὶ αὐτή· Ὅτι σὺ, Κύριε, ἐλπίς μου, τὸν Ὕψιστον ἔθου καταφυγήν σου·δεύτερον, ὡς ἡ Ἑβραϊκὴ φωνή, Ἐλίων περιέχουσα τὸν Ὕψιστον δηλοῖ. Εὕρομεν γοῦν ἐν νϛʹ ψαλμῷ κείμενον τό, Κεκράξομαι πρὸς τὸν Θεὸν τὸν ὕψιστον· ἔνθα ἀντὶ τοῦ, τὸν ὕψιστον, τὸ Ἑβραϊκὸν Ἐλίων περιέχει, ὡς καὶ ἐπὶ τοῦ παρόντος.

It should be observed first, that the fifth edition agrees with the Seventy, itself also saying, That you, Lord, are my hope, you have set the Most High as your refuge. Second, as the Hebrew phrase, containing Ἐλίων, signifies the Most High. We also found in the 56th Psalm the following text: I will call out to God, the Most High. There, instead of, the Most High, the Hebrew contains Ἐλίων, just as in the present passage.

Text 3: Eusebius, Commentarius in Isaiam, 1.64 “ἰδοὺ ὁ θεὸς ὁ σωτήρ μου, πεποιθὼς ἔσομαι ἐπ’ αὐτῷ” ἀντὶ γὰρ ‘τοῦ σωτῆρός μου’ ἡ Ἑβραϊκὴ λέξις ‘ἠσουαθὶ’ περιέχει, καὶ πάλιν “ἐκ τῶν πηγῶν τοῦ σωτηρίου” ἀντὶ “τοῦ σωτηρίου”, “ἠσουὰ” λέλεκται.

“Behold, God my Savior, I will be trusting in Him”. For instead of ‘my savior’, the Hebrew contains the word ‘ἠσουαθί’, and then again, “from the springs of salvation”, instead of “of salvation”, “ἠσουά” is said.

Text 4: Chrysostom, Expositiones in Psalmos, 55.90 Ὑψώθητι ἐν τοῖς πέρασι τῶν ἐχθρῶν σου. Ἄλλος φησὶν, Ἐν θυμῷ ἐπὶ τοὺς ἐχθρούς σου. Ἕτερος, Ἐν χόλῳ τῶν θλιβόντων με. Ἕτερος, Ἐν ἀνυπερθεσίαις ἐνδεσμούντων. Ὁ δὲ Ἑβραῖος τὸ, Ἐν τοῖς πέρασι, Βεβαρώθ φησιν.

Be exalted in the boundaries of your enemies. Another says, In anger against your enemies. One, In wrath on those who afflict me. Another, In haste on those who bind. But the Hebrew renders in the boundaries as Βεβαρώθ.

2.3  How is the Secunda cited in scholia?

There are scores of instances in which a potential quotation of the Secunda is found in a marginal note of a Septuagint manuscript. Typically, a unique symbol will be written above a word or phrase in the main text of the Septuagint elsewhere on the page. In the margin, this symbol will be repeated near the note

164

Benjamin Kantor

containing the Greek transcription of Hebrew that elucidates the word or phrase marked with the symbol. When other Hexaplaric translations are quoted to elucidate this same key word or phrase, the entire list of Hexaplaric quotations may be subsumed under the single symbol, each column preceded by a particular identifier (e.g., α′ for Aquila, θ′ for Theodotion). The Secunda quotation itself is often introduced by an abbreviated form of ὁ ἑβραῖος or τὸ ἑβραϊκόν (see above, § 2.2), such as ο εβραι, ο εβρ′, το εβρ′, or εβρ′. In other cases, non-abbreviated identifiers similar to those used in the church fathers, such as τὸ ἑβραϊκόν, precede a transcription. Sometimes the identifier is entirely absent, with only an isolated transcription written in the margin. In such instances, it cannot be clear from the note alone if the transcription should be attributed to the Secunda or to one of the other Hexaplaric translations such as Aquila, Symmachus, or Theodotion. In either case, such a note may have been taken directly from the Hexapla or it may have been taken from a note in the Tetrapla (see § 2.4 below). Finally, a marginal note containing a Secunda transcription does not have to come directly from the Hexapla/Tetrapla; sometimes a brief extract from a patristic author – it is not always possible to identify the author – is written in the margin (in non-catena manuscripts). In such cases, the transcription tends to be introduced with one of the phrases discussed earlier (§ 2.2). Several examples of Secunda quotations in marginal notes in LXX manuscripts are cited below (Text 5–Text 7): Text 5: Rome, Vaticanus Barberinus Graecus 549 (Rahlfs 86), fol.6v Main Text

Margin

(Hebrew Reference)

⩫ γόμορ κριθῶν· καὶ νέβελ οἴνου·

ὁ ἐβραῖ ουομρσε ωριμ· οὐλεχθσε ωριμ:· ⩫

(Hos 3:2: ‫וחמר שערים ולתך‬ ‫)שערים‬

165

Discovering the Secunda

Text 6: Rome, Vaticanus Graecus 747 (Rahlfs 57), fol. 50r Main Text

Margin

⩫ καὶ οὐαλὰμ ἦν ὄνομα τῆι ⩫ πόλει … ἐβραῖ οὐλαλοὺζ σεμαεἴρp.c. (σεμαηρa.c.)

(Hebrew Reference) (Gen 28:19: ‫ואולם לוז שם‬ ‫)העיר‬

Text 7: Paris, BnF, Coisl. 8 (Rahlfs 243), fol. 122r Main Text

Margin

(Hebrew Reference)

καὶ ⩫συνεκάμψεν

⩫ … εβρ̾ ουϊεγαρ

(2Kgs 4:35: ‫)ויגהר‬

2.4  How is the Secunda cited in the Syro-Hexapla?

As Gentry and others have argued in recent years, the Syro-Hexapla may be regarded essentially as a Syriac translation of the Tetrapla.16 The Tetrapla itself, a Greek composition of Origen and the Library at Caesarea (i.e., Pamphilius and Eusebius), was a recension of the Septuagint (column V) with extensive annotations. The annotations served to compare the recensed Septuagint text with the original Hebrew and other Hexaplaric sources such as Aquila, Symmachus, and Theodotion. While the most famous of these annotations entail the use of the Aristarchian signs to mark omissions and additions, brief extracts from the other Hexaplaric columns were also written in marginal notes.17 Presumably, then, the annotations and marginal notes in the Syro-Hexapla should roughly approximate those present in the original form of the Tetrapla. Transcriptions from the second column also seem to have been included among these notes and annotations. Much like the citations of the Secunda in manuscripts of the Septuagint, these Secunda quotations in the Syro-Hexapla 16 See Gentry, “Aristarchian Signs.“ See also, regarding how the Syro-Hexapla/Tetrapla question relates to Job, John Meade, “An Analysis of the Syro-Hexapla of Job and Its Relationship to Other Ancient Sources.” 17 Gentry, “Aristarchian Signs.”

166

Benjamin Kantor

are often keyed to a particular word or phrase of the main text and cited in the margin, with or without a symbol. Though such transcriptions are not always introduced by an identifier, the following abbreviations are often found: (i) ‫ ܥܒ‬ʕb; (ii) ‫ ܥܒܪ‬ʕbr; (iii) ‫ ܥܒܪܐ‬ʕbrʔ; (iv) ‫ ܥܒܪܝܐ‬ʕbryʔ. Because of the use of Syriac ‫ ܥ‬ʕ rather than ‫ ܐ‬ʔ for the first letter of these monikers, it may be concluded that such abbreviations are natively Syriac and not transliterations of Greek εβρ. The definite article in identifier (iii), however, may indicate a precise rendering of ὁ ἑβραῖος or τὸ ἑβραϊκόν. It may also be the case, however, that Syriac ‫ ܥܒܪ‬ʕbr reflects a sort of compromise between an original Greek εβρ′ and the Syriac cognate root ‫ ܥܒܪ‬ʕbr, with the scribe being guided by the original Greek identifying abbreviation but conforming it to the root of the Syriac cognate. In light of the compositional history of the Syro-Hexapla, the transcription of the Hebrew itself should be regarded as a Syriac rendering of an original Greek transcription. In some manuscripts, there are even notes that contain both a Syriac transliteration of the Greek transcription and the original Greek transcription (see below the example of ϊωβηλ, Text 8). Finally, it should also be noted that just because a transcription is marked with with the moniker ‫ ܥܒܪ‬ʕbr does not necessarily mean that it is derived from the Secunda. At least in some instances, it seems that such a moniker may also precede transcriptions taken from the other Hexaplaric translations such as Aquila, Symmachus, or Theodotion. Several examples of potential quotations of the Secunda in marginal notes of the Syro-Hexapla are cited below (Text 8–Text 10): Text 8: Princeton Scheide MS 150 (MS Midyat), fol. 97v Main Text

Margin

(Hebrew Reference)

+ ‫ܕܫܘܒܡܢܐ‬

+ ‫ܗܬ ܥܒ ܝܘܒܝܠ‬ ϊωβηλ

(Lev 25:11: ‫יֹוב֣ל‬ ֵ )

167

Discovering the Secunda

Text 9: Ambrosiana Mediolanensis, fol. 120v18 Main Text

Margin

(Hebrew Reference)

⤄ ‫ܘܕܒܚܐ‬ ܼ

⤄ .‫ ܡܢܐܐ‬.‫ܥܒ‬

(Jer 14:12: ‫וּמנְ ָ ֖חה‬ ִ )

Text 10: Ambrosiana Mediolanensis, fol. 138r Main Text

Margin

(Hebrew Reference)

⤄ ܵ ‫ܘܠܒܣܝܣ‬

⤄ ܵ ‫ܡܐܟܢܘܬ‬ .‫ܥܒ‬

(Jer 52:17: ‫ת־ה ְמּכֹנ֞ ֹות‬ ַ ‫)ְ ֽו ֶא‬

2.5  How can it be determined if a Greek transcription of Hebrew from one of these sources (church fathers, LXX manuscripts, Syro-Hexapla) is from the Secunda? This question is especially important considering the fact that potentially almost one-third of the Secunda material is attested in quotations in the church fathers, Septuagint manuscripts, and the Syro-Hexapla. Because the Secunda was not the only source for Greek transcriptions of Hebrew in antiquity, or even in the Hexapla itself, it is necessary to form a number of criteria by which one may best determine – though often not decisively – whether a particular transcription is derived from the Secunda. Although there are more nuanced considerations to be taken into account, the likelihood that a particular transcription comes from the Secunda is generally directly related to how many of the following three conditions it meets and how well it meets each of them: 1. An identifier with some form of ‘Hebrew’ – e.g., τὸ ἑβραϊκόν, ὁ ἑβραῖος, ἡ ἑβραϊκὴ λέξις, εβρ′, ‫ܥܒܪ‬, etc. – precedes or introduces the transcription in the passage/note. 2. The transcription is found alongside quotations/citations of other Hexaplaric columns, such as Aquila, Symmachus, and Theodotion. 3. The phonological and orthographic conventions exhibited in the transcription are consistent with those used in the columnar witnesses of the Secunda. 18  Foliation here and in the following example according to the facsimile: Ceriani, Monumenta sacra et profana: Tom. VII.

168

Benjamin Kantor

The first criterion indicates that the author (or the author’s source) regarded the transcription as reflecting (the) Hebrew. Some of these identifiers, however, are less likely than others to point specifically to the text of the Secunda. It should be noted that the moniker ‫ ܥܒܪ‬ʕbr (and its various iterations) in the Syro-Hexapla may be a less reliable indicator of a second column source, since it might also refer to transcriptions found in the Hexaplaric translations. The second criterion is used to determine the likelihood that the transcription was derived as the result of a more general consultation of the Hexapla. Presumably, a Greek transcription of Hebrew found nestled in a list alongside the translations of Aquila, Symmachus, and Theodotion was also likely derived from the Hexapla. The third and final criterion assumes that, generally speaking, the phonology and spelling of the transcription should be consistent with parallels found elsewhere in the Secunda. The columnar witnesses, of course, constitute the most reliable evidence for the phonological and orthographic conventions found elsewhere in the Secunda to which uncertain forms may be compared. The application of these criteria to any particular transcription may result in four degrees of strength/likelihood that the transcription may be attributed to the Secunda. If it meets all three conditions, it may be regarded as very likely (Text 11). If it meets two of the three conditions, it may be regarded as likely (Text 12). If it meets one of the three conditions, it may be regarded as plausible (Text 13). If it meets none of the conditions, it may be regarded as uncertain-tounlikely. Illustrative examples are cited below (Text 11–13): Text 11: Theodoretus, Interpretatio in Psalmos, PG 80, 920 “Εἰς τὸ τέλος ὑπὲρ τῶν κρυφίων τοῦ υἱοῦ, ψαλμὸς τῷ Δαβίδ.” Ὁ μὲν Σύμμαχος, ἐπινίκιον περὶ τοῦ θανάτου τοῦ υἱοῦ· ὁ δὲ Ἀκύλας, τῷ νικοποιῷ νεανεότητος τοῦ υἱοῦ, καὶ ὁ Θεοδοτίων, ὑπὲρ ἀκμῆς τοῦ υἱοῦ. τὸ ἑβραϊκὸν αλμωθ βεν.

“For the end, for the hidden things of the son, a psalm of David.” Symmachus, of victory regarding the death of the son. Aquila, to the one who causes victory, of the youth of the son, and Theodotion, for the zenith of the son. The Hebrew, αλμωθ βεν.

Text 12: Origen, Selecta in Genesim, PG 12, 133 Καὶ ἐκήρυσσεν ἔμπροσθεν αὐτοῦ κήρυξ. And a herald was announcing before him. The Τὸ Ἑβραϊκὸν ἔχει Ἀβρὴχ, ὃ κυρίως Hebrew has Ἀβρὴχ, which properly signifies σημαίνει, πατὴρ ἁπαλός. ‘tender father’.

Discovering the Secunda

169

Text 13: Paris, BnF, Coisl. 1 (Rahlfs M), fol. 78r Main Text

Margin

(Hebrew Reference)

καὶ ϊδοὺ ο̾υλὴ λευ |

̾σηθ

(Lev 13:10: ‫) ְשׂ ֵאת־‬

2.6  Should transcriptions of the uncertain-to-unlikely category be included?

In light of the discussion in the preceding section, it is necessary to confront the obvious question regarding whether or not to include transcriptions that are not necessarily – but possibly could be – derived from the Secunda. There are two possible ways of dealing with this issue. On one hand, it might be reasonable to exclude any transcription that is not at least plausibly attributed to the Secunda. While this would make the text more consistently reliable, it would also eliminate helpful data that may turn out to be relevant when more evidence comes to light. On the other hand, it might be beneficial to include any transcription that is at least possibly attributed to the Secunda, even if its status is only uncertain. While this approach has the benefit of casting a wide net to include all possible quotations, it runs the risk of presenting too much material and diluting the authentic text. The new critical edition thus opts for a bit of a compromise between these two approaches. In wanting to avoid leaving out any possible references to the Secunda and possibly sacrificing comprehensiveness, it was decided to include all possible quotations of the Secunda. However, in order to mitigate against the dilution of the text, transcriptions that are unlikely (or uncertain) to belong to the Secunda are so indicated in the edition, whether explicitly in the introduction (Einleitung) or in the textual notes on the page of the critical text itself. In this way, all potential transcriptions will be included, but it will still be made known to the reader when the authenticity of a particular transcription is in question.

2.7  How should the text of the Secunda be presented in the new critical edition? It seems obvious to the present author that, in the name of authenticity, the new critical edition should replicate the original format of the second column as closely as possible. This means that the Greek transcriptions of the Secunda should be presented in columnar form with a Hebrew column (I) to the left. Because column I of the Hexapla (Hebrew in Hebrew letters) is unattested, it was necessary to make some important decisions about how the Hebrew text thereof should be represented, especially when the vocalization of the Secunda is contrary to the vocalization of the Masoretic Text. Already at the 1994 Rich Seminar on the Hexapla at Oxford, Flint pointed out how merely transcribing the vocaliza-

170

Benjamin Kantor

tion of the Masoretic Text is not always the best recourse, while making several recommendations for new critical editions of the Hexapla: “It is clear that the Hebrew column of the Hexapla cannot automatically be equated with MT, even in its consonantal form. To reproduce the Masoretic Text automatically, as Field and others have done, is not only inaccurate but even misleading … Although not always ideal, MT … constitutes our practical standard of reference … the form of this cited text should sometimes be adjusted with recourse to the second column where the latter is extant … judicious retroversions from the [second column] may occasionally be undertaken to determine the form of the Hebrew text …”19 This problem may be further illustrated by depicting several different ways that the Hebrew column could be presented when corresponding to a variant vocalization in the Secunda. The following chart presents four possible ways of dealing with the Secunda’s variant vocalization of Masoretic ‫‘ ָבּ ֵ֨תּימֹו‬their houses’ (Ps 49:12) (Table 1): Table 1 Re-Consonantal

MT-Consonantal

Re-Vocalized

‫קרבם‬ καρβαμ ‫קרבם‬ καρβαμ ‫ִק ְר ָּבם‬ ‫ ביתמו‬βηθαμου ‫ בתימו‬βηθαμου ‫יתמֹו‬ ָ ‫ֵּב‬ ‫לעולם‬ λωλαμ ‫לעולם‬ λωλαμ ‫עֹולם‬ ָ ‫ְל‬ ‫ משכנתם‬μισχνωθαμ ‫ משכנתם‬μισχνωθαμ ‫ִמ ְׁש ְּכנ ָֹתם‬ ‫לדר‬ αδωρ ‫לדר‬ αδωρ ‫ְלד ֹר‬ ‫ודר‬ ουαδωρ ‫ודר‬ ουαδωρ ‫וָ ד ֹר‬

MT-Vocalized

καρβαμ ‫ִק ְר ָ ֤בּם‬ καρβαμ βηθαμου ‫ ָבּ ֵ֨תּימֹו‬βηθαμου λωλαμ ‫עֹולם‬ ָ ֗ ‫ְ ֽל‬ λωλαμ μισχνωθαμ ‫ ִ ֭מ ְשׁ ְכּנ ָֹתם‬μισχνωθαμ αδωρ ‫ְל ֣ד ֹר‬ αδωρ ουαδωρ ‫וָ ֑ד ֹר‬ ουαδωρ

There are positives and negatives to each of these approaches. The further to the right one goes in this chart (i.e., “Re-Vocalized” and “MT-Vocalized”), the more familiar and accessible the forms will be to most students and scholars. The further one goes to the left in this chart (i.e., “Re-Consonantal” and “MT-Consonantal”), the closer one is, presumably, to the original format of the Hexapla. The unvocalized Hebrew layouts actually force the modern scholar to interact with the Secunda just as the ancients likely would have, drawing on the Greek transcriptions as an aid for vocalizing the Hebrew consonantal text. It should be noted, however, that opting for “Re-Consonantal” over “MT Consonantal” actually runs the risk of elevating the vocalization/reading tradition above the consonantal text, which might render such “judicious retroversions” of the consonantal text superfluous. In other words, it is possible that the Hebrew column (I) might indeed have read as the “MT-Consonantal” column does, with the vocalization (i.e., oral) tradition 19 Flint, “Columns I and II,” 131–132.

171

Discovering the Secunda

preserving a variant that was not found in the consonantal text – the hypothesized Hebrew column (I) of “Re-Consonantal” as ‫ ביתמו‬may never have existed for this word in the Secunda. The transcription βηθαμου may only reflect an oral tradition (if it does not reflect a scribal error: βαθημου → βηθαμου). All in all, each of these formats leaves something to be desired and does not quite provide all the information, flexibility, and/or authenticity that would be desired in such an edition. To solve this problem, it was determined to present two of these formats in a facing-page layout. For the sake of authenticity and for the benefit of highlighting possible textual variants in the Hebrew consonantal text, the “Re-Consonantal” format will be used for the main text of the edition on the left-facing page. However, for the sake of ease of use and comparison, the “MT-Vocalized” format (with verse numbers indicated instead of repeating the Greek) will be presented in a facing-page layout on the right-facing page. A simplified version of this layout (obviously not proportional) is depicted below (Figure 2): Figure 2 Psalms 49

‫קרבם‬ ‫ביתמו‬ ‫לעולם‬ ‫משכנתם‬ ‫לדר‬ ‫ודר‬

καρβαμ βηθαμου λωλαμ μισχνωθαμ αδωρ ουαδωρ

– Source list (Kopfleiste) – Critical Apparatus – Quotations

Psalms 49

‫ִק ְר ָ ֤בּם‬ ‫ָבּ ֵ֨תּימֹו‬ ‫עֹולם‬ ָ ֗ ‫ְ ֽל‬ ‫ִ ֭מ ְשׁ ְכּנ ָֹתם‬ ‫ְל ֣ד ֹר‬ ‫וָ ֑ד ֹר‬

49:12

– Translations of quotations

This layout allows the reader to appreciate the text in a form that is as close as possible to its original layout in the Hexapla by reading the left-facing page, but to compare it with the more standard and familiar form of Hebrew represented by the Masoretic text by reading the right-facing page. It should also be noted that at the bottom of the left-facing page are apparatuses and quotations of the passages in which the Secunda transcriptions are found (where applicable). In the corresponding space at the bottom of the right-facing page are translations

172

Benjamin Kantor

of the same passages. In short, all of the ancient data falls on the left-facing page and all of the scholarly aids fall on the right-facing page.

3. A Selection of New Insights from Working on the Critical Edition A comprehensive critical edition of the second column of the Hexapla has never before been published nor has all the Secunda material ever been gathered into one place. As a result, simply gathering and compiling all of the extant evidence for the first time has led to a deeper understanding of and new insights into the original nature of the Secunda, its place in the compositional history of the Hexapla, and its early reception. A selection of such insights is presented below:

3.1  At what point did Origen come to have access to the Greek transcriptions that would form the basis of the second column of the Hexapla? The current scholarly consensus holds that the Hexapla was produced by Origen in Caesarea at some point between 234 CE and 250 CE.20 This by no means implies, however, that the text underlying the second column did not have earlier roots. It is, of course, entirely possible that Origen first encountered the Secunda text long before he produced the Hexapla. Some have even suggested that Origen might have gotten the Secunda text from Egypt, which would likely imply that he acquired it relatively early on in his career while he was still living in Alexandria. What the work on the new critical edition has called attention to, however, is the fact that Origen does not seem to have even been familiar with using the Secunda text until towards the end of his life. While isolated Hebrew words in Greek transcription are peppered throughout his writings, only three of Origen’s works contain unambiguous quotations of the Secunda: Selecta in Psalmos (PG 12.1084.7–39; 12.1057.42–12.1060.35), the Commentary on Matthew, and the Letter to Africanus. All three of these works are dated to around 246–250 CE, just several years before Origen’s death.21 In my view, this confluence points to the following conclusion: It seems that Origen was not able to use the Greek transcriptions of the Secunda in the Hexa20 Gentry, “Origen’s Hexapla.” 21 Nautin, Origène: Sa vie et son oeuvre, 183, 261–292, 409–412; Heine, Origen: Scholarship in the Service of the Church, 115–122, 189. It should be noted, however, that Origen’s commentary on Psalms was composed over several decades, with the first section (Ps 1–25) being composed in Alexandria between 222–229 CE and the larger commentary in Caesarea around 246–247 CE. After Origen completed his large commentary on the Psalms, he also composed an abridged ἐγχειρίδιον ‘handbook; manual’ of the work for more manageable consumption. All three separate works—or stages of the work—have been collected and subsumed under the title Selecta in Psalmos in modern editions. The earlier portions are not relevant here, because “unambiguous/certain quotations of the Secunda” occur only in the ἐγχειρίδιον.

173

Discovering the Secunda

pla until the final years of his life. While he had always been interested in Jewish interpretations of scripture and the meanings of Hebrew words – as evidenced by his earlier writings – the philological use of the Secunda against the other columns of the Hexapla is only first evidenced in the late 240s CE. This finding seems to be highly relevant for understanding the compositional history of both the Secunda and the Hexapla as a whole.

3.2  How many words were written per line in the Secunda originally?

Although it is common for scholars to describe the Hexapla as containing one Hebrew word with its Greek equivalents per line,22 this is actually not entirely accurate. Even though one Hebrew word is written on each line in most cases, it is by no means uncommon for two or three Hebrew words to be written on the same line, as can be seen in the brief extract from fol. 73v–70r + 30r–27v of the Ambrosiana palimpsest, MS. Rahlfs 1098 (Table 2): Table 2 Col. I

Col. II

(Tiberian)

(Verse)

‫למנצח‬ ‫לבני קרח‬ ‫מזמור‬ ‫שמעו זאת‬ ‫כל העמים‬ ‫האזינו כל‬ ‫ישבי‬ ‫חלד‬ ‫גם בני‬ ‫אדם‬ ‫גם בני איש‬ ‫יחד עשיר‬ ‫ואביון‬

λαμανασση αβνη·κὸρ μαζμωρ σιμου·ζωθ χολ·ααμιν σεζινου·χολ εωσηβ ολδ γαμ·βνη ἀδαμ γαμ·βνη·εἴς ιααδσιρ ουεβιῶν

‫ִשׁ ְמעוּ־ז ֹ֭את‬ ‫ִק ְר ָ ֤בּם‬ ‫עֹולם‬ ָ ֗ ‫ְ ֽל‬ ‫ִשׁ ְמעוּ־ז ֹ֭את‬ ‫ל־ה ַע ִ ֑מּים‬ ֽ ָ ‫ָכּ‬ ‫ַ֝ה ֲא ִ֗זינוּ ָכּל־‬ ‫ֹ֥י ְשׁ ֵבי‬ ‫ָ ֽח ֶלד׃‬ ‫ם־בּ ֵנ֣י‬ ְ ַ‫גּ‬ ‫אָ ָ֭דם‬ ‫י־אישׁ‬ ֑ ִ ֵ‫ם־בּנ‬ ְ ַ‫גּ‬ ‫ַ֗ ֝י ַחד ָע ִ ֥שׁיר‬ ‫וְ ֶא ְביֽ ֹון׃‬

49:1

49:2

49:3

Our only sources for analyzing the distribution of such a phenomenon are the five columnar witnesses mentioned earlier. A statistical analysis of all roughly 1090 words attested in the columnar witnesses shows that about one out of every five words was written in a multi-word line, whether part of a pair (i.e., two words per line) or a triad (i.e., three words per line). These statistics are depicted in the following chart (Table 3): 22  Grafton and Williams, Christianity and the Transformation of the Book, 101.

174

Benjamin Kantor

Table 3 words part of… Ambrosiana Palimpsest (Rahlfs 1098) Genizah Palimpsest (Rahlfs 2005) Var. Barb. gr 549 (Rahlfs 86) Ambros. B. 106. sup. (Rahlfs 113) Vat. gr. 1747 (Rahlfs 271) Total

One-word line Pair Triad ~850 16 4 2 17 82%

172 15 0 0 0 3 0 0 8 0 18%

Total ~1040 16 7 2 25 ~1090

The fact that this phenomenon is attested in three out of the five columnar witnesses, and abundantly in the Ambrosiana palimpsest, suggests that it was almost certainly original to the form of the second column as it was written in the Hexapla. It is also worth noting that most cases of pairs and triads in the extract above correspond with conjunctive accents in the Tiberian tradition. Indeed, the distribution of pair- and triad- lines in the columnar witnesses is by no means arbitrary. In fact, pairs and triads generally seem to correspond with sense units and/or prosodic units. Roughly 85% of all instances of pair- and triad- lines correspond to conjunctive accents in the Tiberian tradition of Biblical Hebrew. Of these, roughly half are connected with the strongest sort of binding accent, namely, the maqqef. Only 15% of the examples correspond with a disjunctive accent, and some of these may be the result of a different accentuation pattern. The sum total of the evidence, which is far too extensive to be enumerated here, points to the pair- and triad- lines reflecting a Roman-period Jewish tradition of dividing the biblical text into sense and/or prosodic units. Whether this reflects a conscious endeavor or merely the faithful transmission of an oral tradition of pauses is unclear. Nevertheless, this general practice may be regarded as a precursor to the accent system which would come to be represented graphically several centuries later. On this point, it is worth noting that Revell has found hints of an early Jewish “accent” system in early Septuagint manuscripts as well.23 Such a phenomenon also finds parallels in bilingual Graeco-Latin literary texts used as language-learning material from the ancient world. Works of Latin literature, such as the Aeneid, were sometimes presented in columnar form in order to help Greek-speakers learn Latin. The Latin would be written in one column and the corresponding Greek translation or gloss would be written in the adjacent column. This allowed the Greek-speaking language learner to match the Latin with his more familiar Greek when needed. Although it was common for only one word 23 Revell, “The Oldest Evidence for the Hebrew Accent System”; Revell, “Biblical Punctuation and Chant in the Second Temple Period.”

Discovering the Secunda

175

to be written per line, there were also plenty of instances in which two or more words were written on the same line. According to Dickey, the line breaks “are positioned to divide up meaningful units; the translator can use them both to show the reader how the original text is to be construed and to organize groupings that can be successfully translated as a unit.”24 This may be illustrated by an extract from a Graeco-Latin papyrus of the Aeneid from the 4th/5th c. CE. Lines containing more than one word (excluding the enclitic -que) are bolded (Table 4):25 Table 4 [illa t]ib[i] Italiae εκ[ει]νη σοι της Ιταλιας [po]pulos τους δημους [ve]ntura vella και τους ελευσομενους πολεμους [et quo q]uemque και οποιουσδηποτε [modo] τροπω 65 [fugias]que και φευξιας [fera]sque και οισιας l[ab]orem τον καματον expediet ευλυτωση cursusque secundos και δρομους εσι[ους] 70 da[b]it δωσει venerata προσκυνηθεισα haec sunt ταυτα εστιν cuae liceat te α εξες[τ]ι σε nostra voce τη ημετερα φωνη moneri υπομνησθηνα[ι]

60

She, to you, Italy’s peoples, (and) the coming wars, and in what (each ↓) way [you might] both avoid and bear (each↑) trouble she will recount and a favorable journey will give having been reverenced these are [the things] about which you may by my voice be warned

We should first note that the traditional word order can be rearranged in such texts to make the syntax more tractable for a language learner. Although more than half the lines contain just one word (excluding the enclitic -que), there are seven instances of pairs and/or triads. Typically, a clear syntactic or prosodic relationship can be discerned between the two or three words written on the same line, such as a noun with its attributive adjective (line 69: cursusque secundos ‘and a favorable journey’; line 74: nostra voce ‘by our/my voice’), a noun with an attributive participle (line 63: ventura vella ‘the coming wars’), a noun with its verb (line 72: haec sunt ‘these are … ’), or the principle part of a relative clause (line 74: cuae liceat te ‘about which you may … ’). Although not all instances of multiple 24 Dickey, “Columnar Translation: An Ancient Interpretive Tool that the Romans Gave the Greeks.” 25 Cavenaile, Corpus Papyrorum Latinarum, 20–22.

176

Benjamin Kantor

words being written on the same line can be explained this way, the essential principle is that each line contains the most concise meaningful unit that can be isolated in a helpful way for the learner.26 It would seem, then, that there are very good ancient grounds for assuming that the pair and triad lines attested in the columnar witnesses go back to the original form of the first and second columns in the Hexapla. The fact that the rest of the columns of the Hexapla are keyed to the Hebrew further suggests that these line divisions reflect a Jewish tradition of dividing the Hebrew, rather than a Greek tradition of dividing the translations or some arbitrary divisions created purely for the Hexapla.

3.3  Were diacritics originally part of the Secunda? Unfortunately, the relevance of diacritics for understanding the Secunda transcriptions has been minimized following Janssens’ erroneous comment in his work on the Secunda that the diaeresis (¨) on iota (i.e., ϊ), the spiritus lenis (᾽), and the spiritus asper (῾) were “only added in the 8th or 9th century.”27 Although such signs were not used very frequently in the ancient period, they were certainly available for the knowledgeable scribe to be used on occasion already in the Hellenistic period.28 The question of the originality of the diacritic marks in the second column should thus be approached with an open mind. Although a full treatment of the topic would require another article in itself, it may be advantageous to make mention here of a number of relevant points regarding the nature of diacritics in the Secunda. There are two primary questions that should be asked to determine if a particular diacritic was originally part of the Secunda: First, does its use reflect Hebrew knowledge rather than typical Greek scribal practice? Second, is it attested across the spectrum of the columnar witnesses? With these two questions kept in mind, each of the following topics may be treated in turn: (i) diaeresis; (ii) breathing marks; (iii) accents. In Greek scribal practice, diaeresis (¨) on iota (i.e., ϊ) is often used at the beginning of a word, to indicate that ι = [i] should be read distinctly from the preceding vocalic grapheme, or to distinguish the letter graphically so that iota is not lost/omitted in copying.29 Though its use is not always consistent, diaeresis (¨) on iota (i.e., ϊ) is commonly used in the Ambrosiana palimpsest to mark consonantal yod /j/: e.g., ωϊηβ ‫אֹוי֑ב‬ ֵ (Ps. 31.9), ϊαδω ‫( ֑יָדֹו‬Ps. 89.26), μαϊμ ‫( ַמ֣יִם‬Ps. 32.6), and αλαϊ ‫( ָע ַל֣י‬Ps. 35.16).30 26  I would like to thank Eleanor Dickey for her personal correspondence with me on this topic. 27 Janssens, Origen’s Secunda, 39–40. 28 Thompson, Handbook, 71–72. 29 Thompson, Handbook, 73. 30  This statement requires further qualification: Although both ι and ϊ can indicate [i], [iː], [j], [(i)j(i)], there is much more of a tendency for ϊ to indicate consonantal [j] and ι to indicate a vocalic [i(ː)]. Moreover, geminated consonantal [jj] is almost exclusively indicated by ϊ. Finally,

177

Discovering the Secunda

Of the other columnar witnesses, consonantal yod /j/ is attested in just two: Cod. Vaticanus Barberinus Graecus 549 (Rahlfs 86) and Cod. Vaticanus Graecus 1747 (Rahlfs 271). In the former, diaeresis is absent: ισ·ραηλ· ‫יִשׂ ָר ֵ ֖אל‬ ְ (Hos 11:1); οὐμεμμισ̔ραιμ· ‫וּמ ִמּ ְצ ַ ֖ריִם‬ ִ (Hos 11:1). In the latter, diaeresis is present: ϊασὰβ ‫יָשׁב׃‬ ֽ ָ (Ps 1:1). The evidence thus remains ambiguous as to whether or not diaeresis was originally present in the Secunda. However, the fact that its usage in both the Ambrosiana palimpsest (Rahlfs 1098) and Cod. Vaticanus Barberinus Graecus 549 (Rahlfs 86) reflects Hebrew knowledge as well as the fact that it is attested in two out of three columnar witnesses suggests that it may be a feature original to the second column of the Hexapla. In more modern Greek orthography, each word beginning with a vowel is marked with either a spiritus lenis (᾽), which indicates no consonantal value before the following vowel, or a spiritus asper (῾), which indicates the consonantal value [h] before the following vowel. There are nearly 500 words beginning with a Greek vocalic grapheme in the Secunda of the Ambrosiana palimpsest. A spiritus marks roughly 100 of these. In light of all the evidence – for which the reader must be referred to the forthcoming edition – it is unlikely that such word-initial breathing marks were original to the second column. They probably reflect a later scribal addition in conformity with typical Greek practice. On the other hand, instances of spiritus asper on a word-medial grapheme, which occur only a handful of times, are far more likely to be original (Table 5): Table 5 Verse

Tiberian

Hebrew Letter

Reading

Ps 18:35 Ps 46:4 Ps 89:28 Ps 35:27 Ps 49:15 Ps 32:7 Ps. 46.5

‫חוּשׁה‬ ָ֗ ‫ְ֝נ‬ ‫יֶ ְח ְמ ֣רוּ‬ ‫ֶא ְתּ ֵנ֑הוּ‬ ‫חוּ‬ ֮ ‫וְ יִ ְשׂ ְמ‬ ‫וַ יִּ ְר ֘דּוּ‬ ‫ֹוב ֵב֣נִ י‬ ְ ‫ְתּ ֖ס‬ ‫ִמ ְשׁ ְכּ ֵנ֥י‬

 Heth  Heth Heh Yod  Yod   Samekh  Shin

νεόυσα31 ιε῾μρου ἐθνη῾ου32 ουειˈεσομου33 ουι῾αρδου θσὡβαβηνι μσ῾χνη

There may be a Hebrew phonological distribution attested in these examples. A word-medial spiritus mark occurs on h-like gutturals, on yod, and on sibilants. word-initial consonantal [j] can be represented with ϊ Brønno, Hexapla des Origenes, 376–383; Yuditsky, Diqduq, 32–33; Kantor, “The Second Column,” 232–234, 305–310. 31  The acute accent in νεόυσα is interpreted as a modification from an earlier spiritus asper. 32  Note that this breathing mark may also be interpreted as an accute accent (see below). 33  The mark transcribed as ˈ is best interpreted as a spiritus asper.

178

Benjamin Kantor

In the case of the gutturals, it may be that spiritus asper was used to indicate the pronunciation [h] or [ħ], which could not be conveyed by the Greek letters themselves. In the case of yod, the spiritus may indicate the consonantal [j], rather than the vocalic [i], pronunciation. In the case of sibilants, the spiritus may be intended to somehow differentiate between the various sibilants that might be reflected by σ. Because Greek σ is used to represent four separate Hebrew letters (‫ ׁש‬,‫ ׂש‬,‫ צ‬,‫)ס‬, a scribe with Hebrew knowledge may have wanted to help distinguish between these sounds.34 Interestingly, the use of spiritus asper both to indicate Hebrew /h/ in word-medial position and to differentiate σ = ‫ צ‬from σ = ‫ ׂש‬,‫ ס‬seems to be attested in one of the other columnar witnesses as well, namely, Cod. Vaticanus Barberinus Graecus 549 (Rahlfs 86). The only word containing h-like gutturals in this witness, ‫וָ א ֲֹה ֵ ֑בהוּ‬, uses spiritus asper to mark each of them: ουεἁβηὁυ ‫( וָ א ֲֹה ֵ ֑בהוּ‬Hos 11). Only two sibilants are attested in this witness as well, one pronounced as /s/ (‫ )ׂש‬and the other as /ṣ/ (‫)צ‬. Interestingly, while the former is unmarked, the latter is marked with a spiritus asper: ισ·ραηλ· ‫( יִ ְשׂ ָר ֵ ֖אל‬Hos 11:1); οὐμεμμισ̔ραιμ· ‫וּמ ִמּ ְצ ַ ֖ריִם‬ ִ (Hos 11:1). It may be that spiritus asper served to mark the emphatic sibilant in this case. While it is possible that word-medial spiritus asper to indicate h-like gutturals was a later addition by a scribe with knowledge of Hebrew, the fact that the same sort of convention is attested in two distinct witnesses suggests that it may be a feature original to the second column of the Hexapla. Unfortunately, the evidence is insufficient to make a determination between these two possibilities. In Greek scribal practice, the primary stress/accent of a word is indicated by an acute accent (´), a grave accent (`), or a circumflex accent (῀). These three signs are also used in the Ambrosiana palimpsest, though they only correspond to the Hebrew stress a little more than half the time. An acute accent is generally used in word-initial or word-medial position and a grave accent is generally used in word-final position, though this distribution is not without exceptions. A circumflex tends to be used over long /ō/ or long /ū/, but there are also exceptions to this rule. Long /ō/ or /ū/ can also be marked with an acute or grave accent. The statistical correspondence between the various Greek accent signs and the Tiberian primary stress is depicted in the following chart (Table 6):

34  The possibility of a spiritus distinguishing the sibilants is discussed in Pretzl “Hexapla des Origenes,” 18 and Brønno, Hexapla des Origenes, 414.

179

Discovering the Secunda

Table 6 before Tib. Stress

= Tib. Stress

after Tib. Stress

Total

36 0 0 36 (31%)

33 35 3 71 (61%)

2 4 3 9 (8%)

71 39 6 116

Acute Grave Circumflex Total

What is most apparent from these statistics is that only the grave accent replicates the Tiberian primary stress with any degree of consistency. This may be due to the fact that Hebrew is normally stressed word-finally and the Greek grave accent is meant to appear only in word-final position. In fact, it is possible to highlight certain patterns more clearly by manipulating the presentation of the data in a different way. If the data are tabulated according to whether or not Tiberian ultimate or pen-ultimate stress corresponds to the placement of the Greek accent, they yield the following statistics (Table 7):35 Table 7

Acute = Acute before Acute after Acute Total Grave = Grave after Grave Total Circumflex = Circumflex after Circumflex Total Total

Tib. Penultimate Stress

Tib. Ultimate Stress

Total

17 7 2 26 1 4 5 0 0 0 31 (27%)

16 29 0 45 34 0 34 3 3 6 85 (73%)

33 36 2 71 35 4 39 3 3 6 116

What this second set of statistics demonstrates is that the accents only seem to correspond with the Tiberian primary stress when they are both word-final. When the primary stress is penultimate, there is no statistical correlation at all. 35  It should be noted that several transcriptions have been eliminated from these statistics due to the possible incongruence of traditions.

180

Benjamin Kantor

This may indicate that there was a general scribal awareness that Biblical Hebrew had a preference for word-final stress. Such a claim would be supported by transcriptions of Hebrew segholate nouns with a Greek accent on the ultima, even though they should normally be stressed on the penultima: e.g., μεββεσε` ‫ה־בּ ַצע‬ ֥ ֶ ‫ַמ‬ (Ps. 30.10) and ασσααθ̀ ‫( ַה ָ ֽשּׁ ַחת‬Ps. 49.10). In sum, the Greek accents in the Ambrosiana palimpsest are not a reliable indicator of primary stress in the Hebrew. Moreover, of the five columnar witnesses, accent marks are only present in two: the Ambrosiana palimpsest (Rahlfs 1098) and Vaticanus Graecus 1747 (Rahlfs 271). There is thus no reason to suppose that they were original to the second column of the Hexapla.36

3.4  Where is the Secunda material attested today? In an earlier selection, it was noted that there are four main categories in which Secunda material has been preserved: (i) columnar witnesses; (ii) quotations in the church fathers; (iii) marginal notes in Septuagint manuscripts; (iv) marginal notes in the Syro-Hexapla. What gathering the material for a critical edition has done, however, is provide an idea of the relative distribution of the material across these four categories. Although an exact number is not yet possible, there are probably over 1500 transcribed words in ancient sources that may ultimately be derived from the Secunda. The overwhelming majority of these, roughly 1090 words, are attested among the columnar witnesses (i). The Ambrosiana palimpsest (Rahlfs 1098) itself, however, accounts for around 1040 words. The combined total of the other four columnar witnesses is around 50 words. If the Ambrosiana palimpsest is not factored in to the statistics, the overwhelming majority of Secunda material comes from the church fathers (ii), who account for around 400 words, though an exact number is difficult to determine because many of the church fathers seem to quote the same words and passages (a very important observation!).37 If recycled quotations and transcriptions of dubious attribution are excluded, this number may decrease significantly. Secunda transcriptions in marginal notes in 36  It is possible that they were added later on the basis of a limited knowledge of Hebrew, one facet of which being that Hebrew tended to stress words on the ultima. Still, there may be a more sophisticated explanation for the distribution of the Greek accent signs or one that does not necessarily relate to Tiberian primary stress. It may be that they were never intended to reflect primary stress. It is possible that some of the accent signs may have been intended to indicate some other phonological or prosodic phenomenon. This remains a topic for future research. 37  This tends to occur in passages with Messianic prophecies and in texts with important and/or perplexing names. For example, the name ‫ גיבור אל‬in Isaiah 9 receives comment (with transcription) by multiple authors. Similarly, the terms ‫ ציים‬and ‫( איים‬Isa 13:21–13:22) tend to receive comment (with transcriptions) by multiple patristic authors. It is possible, then, that some of the later authors were not quoting directly from the Hexapla but only from an earlier patristic author.

181

Discovering the Secunda

Septuagint manuscripts (iii) probably account for more than 100 words in total, though this count also requires further attention with difficulties of its own; e.g., there are some marginal notes containing transcriptions that are probably from patristic commentaries, even though they lack explicit attribution. The Syro-Hexapla (iv) accounts for the smallest portion of Secunda material, totaling around or less than 50 transcribed words. Finally, it should be noted that these statistics are only preliminary. More stable numbers will be available only toward the end of the project. Nevertheless, preliminary numbers regarding the distribution of the Secunda material may be depicted in chart form, the left chart indicating the total distribution and the right chart indicating the distribution if the Ambrosiana palimpsest (Rahlfs 1098) is excepted from the statistics (Figure 3): Figure 3 Distribution

Distribution (-Amb.)

Direct Witnesses

Church Fathers

Direct Witnesses

Church Fathers

LXX Manuscripts

Syro-Hexapla

LXX Manuscripts

Syro-Hexapla

It becomes quite clear from these charts that, aside from the Ambrosiana palimpsest, it is the church fathers who have preserved most of the Secunda material for future generations. Further, the overwhelming majority of Secunda material in the patristic writings is concentrated among just five authors (in descending order): Chrysostom, Origen, Epiphanius, Eusebius, and Theodoret. Jerome might also be added to this list, but it has not yet been fully ascertained what percentage of his transcriptions are derived from the Secunda and what percentage are original to him. Finally, due to its relevance for understanding the early reception history of the Hexapla, it is also worthwhile to make brief mention of the relative distribution of transcriptions among different biblical books. Psalms is by far the most commonly attested book, with roughly 1420 attested words. Isaiah is the next most commonly attested book, with roughly 90 attested words. Genesis and Kings follow, with around 35 words each. The fifth most commonly attested book is Jeremiah, with roughly 25 attested words. A chart of the distribution of transcriptions among the top five attested books in the Secunda, one with the

182

Benjamin Kantor

data from the Ambrosiana palimpsest (Rahlfs 1098) and one without, is depicted below (Figure 3): Figure 4 Word Count

Psalms

Isaiah

Genesis

Kings

Word Count (-Amb.)

Jeremiah

Psalms

Isaiah

Genesis

Kings

Jeremiah

Even if the data from the Ambrosiana palimpsest is eliminated, Psalms still accounts for the overwhelming majority of attested Secunda transcriptions. It is also worth noting that four out of the five columnar witnesses contain portions of Psalms. All of this evidence leads me to speculate that there might have been a “Hexaplaric Psalter” – i.e., a copy of the Psalms portion of the Hexapla – circulating in Syro-Palestine in the early first millennium CE, but this must remain merely speculation. At the same time, however, one might reasonably argue that the Cairo Genizah palimpsest currently housed in Cambridge (T–S 12.182 [Rahlfs 2005]), which contains a portion of Psalm 22, was just such a copy. On the other hand, one has to be careful about reading too much into the distribution of transcriptions among different biblical books, which may be a mere historical accident. On this point, it is worth noting that Psalms and Isaiah are also the most commonly quoted Old Testament books in the New Testament and in other Second Temple period Jewish documents like the Qumran scrolls. The high attestation of transcriptions from these books among the patristic writings may thus simply reflect the portions of scripture which contained the most messianic prophecies and to which they devoted most of their attention.

3.5  How reliable are the transcriptions found in each type of source (columnar witnesses, church fathers, LXX manuscripts, Syro-Hexapla)? It should go without saying that the attestations of the Secunda in the columnar witnesses, with the possible exception of Cod. Vaticanus Graecus 1747 (Rahlfs 271), are by far the most reliable sources available. In short, anything found in the other four columnar witnesses (Rahlfs 1098, Rahlfs 2005, Rahlfs 86, Rahlfs 113) should almost always take precedence over readings found in other sources.

Discovering the Secunda

183

Even though the church fathers account for the majority of transcriptional material outside the Ambrosiana palimpsest (Rahlfs 1098), they can actually be some of the least reliable sources for Secunda transcriptions, since the manuscript witnesses for each work of the church fathers have a much more complex and extensive transmission history than the columnar witnesses, Septuagint manuscripts, or the Syro-Hexapla. Moreover, it is not uncommon for later scribes to conform the Greek transcription of the Secunda to the phonology of a more “standard” Hebrew tradition (see more on this in § 3.6 below). Print editions of the late nineteenth and early twentieth centuries are notorious for preferring the “standard” – and thus incorrect – reading in such instances. Moreover, several of the church fathers, such as Origen and Chrysostom, have various works with no critical editions at all. Migne’s volumes are sometimes all the modern scholar has to adjudicate between various readings. As such, each patristic work must be treated with as thorough a text-critical analysis as possible in order to arrive at the best possible reading for every transcription. In the case of Chrysostom, it was necessary to carry out primary text-critical work almost completely from scratch, examining over 20 manuscripts personally and consulting with various libraries throughout Europe. Secunda quotations in marginal notes in Septuagint manuscripts are often cited with no clear attribution. As a result, is not always apparent what sort of source the scribe used to derive a particular transcription. Further, each transcription has to be treated in isolation, since there is no guarantee that a scribe used the same source for all the transcriptions in their marginal notes. In a couple of cases, it seems that individual scribes used a mix of reliable and corrupt sources to make notes on the same manuscript. Some of the transcriptions in Septuagint scholia are likely the result of a far more direct line of transmission from the Tetrapla/Hexapla than most of the quotations in the church fathers. Of particular note here are Cod. Coislinianus 1 (Rahlfs M), Cod. Vaticanus Barberinus Graecus 549 (Rahlfs 86),38 and Cod. Coislinianus 8 (Rahlfs 243), all of which seem to be fairly reliable transmitters of Secunda transcriptions in most such scholia. Even though the Syro-Hexapla constitutes some of the earliest extant evidence, it is, by nature, perhaps the least helpful of all the sources. This is because the Syriac script masks a significant portion of what must have been present in the original Greek transcription. Although a judicious use of matres lectionis (e.g., ‫ܐ‬ for α/ε, ‫ ܝ‬for η/ι, ‫ ܘ‬for ω) on the part of the scribe may successfully transmit part of the original transcription, the reader is often left guessing as to the vowels of the remaining part. Nevertheless, the evidence from the Syro-Hexapla is not to be neglected, since it frequently provides helpful information. 38  In addition to fol. 17v of Vaticanus Barberinus Graecus 549 (Rahlfs 86), which contains the columnar presentation of Hos. 11.1 from the Hexapla (cited as a columnar witness above), there are also numerous marginal notes in Rahlfs 86 containing quotations of the Secunda.

184

Benjamin Kantor

3.6  Do scribes with knowledge of “standard” Hebrew ever emend citations of the Secunda quoted in the writings of the church fathers? The relationship of Greek scribes to Greek transcriptions of Hebrew quoted in the patristic sources is a tenuous one. It seems that they did not always know how they should handle them. In some instances, they mistake the Greek transcription of Hebrew for a misspelled Greek word and thus correct it to a similarly-spelled Greek word. In other instances, they seem to have become flustered by their infamiliarity with the transcription so as to omit it entirely! In still other instances, however – and these are the most difficult to detect – a scribe with Hebrew knowledge seems to have altered a reading ever so slightly to conform it to “standard” Hebrew phonology. While the former two scribal alterations are relatively straightforward, the last of these can be highly problematic, since it is not always possible to determine if such a change has taken place. Two of the many instances in which such a change has taken place are described below: σεφρ → σεφερ || ‫( ֵס ֶפר‬Psalms Title): In the transcribed list of biblical book names found in Origen’s Selecta in Psalmos, the title of the book of Psalms, ‫ֵס ֶפר‬ ‫)ּת ִה ִּלים‬ ְ ‫(ה‬, ַ is presented in Migne’s edition as σεφερ θιλλιμ (PG 12, 1084). Hebrew nouns of the pattern *qVtl (i.e., “segholate” nouns), however, are not normally vocalized with an epenthetic vowel in the Secunda.39 Moreover, ḥ ireq in a closed unstressed syllable normally corresponds with Greek ε, rather than ι, in the Secunda. Therefore, the reading of σεφερ θιλλιμ is highly unlikely to be original. Indeed, the corresponding title of the book of Psalms in the columnar witnesses – in this case Cod. Ambrosianus B. 106 sup. (Rahlfs 113) – reads σεφρ αθελλιμ, providing the original Secunda reading on each of these points. It may thus be concluded that a later scribe (or editor) who knew a more “standard” form of Hebrew corrected the original to match the form of Hebrew more familiar to him.40 αων ακοββαϊ ϊσοββουνι → αων ακουββαει ισουββουνει || ‫סוּבּנִ י׃‬ ֽ ֵ ְ‫ֲעֹו֖ ן ֲע ֵק ַב֣י י‬ (Ps 49:6): In Chrysostom’s Expositiones in Psalmos, the phrase ‫סוּבּנִ י׃‬ ֽ ֵ ְ‫ֲעֹו֖ ן ֲע ֵק ַב֣י י‬ (Ps 49:6) is presented in Migne’s edition as Ἀὼν ἀκουββαεὶ ἰσουββουνεί (PG 55, 226). There is just one recurring problem in this transcription as presented in Migne’s edition. In the Secunda, the parallel to šureq/qibbuṣ in a closed unstressed syllable normally corresponds with Greek ο and not ου. Thus, the presence of ου before geminated ββ in both ἀκουββαεὶ and ἰσουββουνεί is unlikely to be original. While this is confirmed by the attestation of this phrase among the columnar 39 Kantor, “The Second Column,” 354–357. 40  It should be noted that this passage is also quoted in Eusebius’ Historia Ecclesiastica. Moreover, the list of transcribed biblical book names has multiple independent manuscript attestations that are not directly connected to Origen or Eusebius. For a fuller treatment of these issues, see the forthcoming critical edition and Albrecht, “Göttingen Septuagint.”

185

Discovering the Secunda

witnesses – the Ambrosiana palimpsest (Rahlfs 1098) has αὼν ακοββαϊ ϊσ̓όββουνι – this can be demonstrated even from an overview of just the Chrysostom manuscripts attesting to this passage (Table 8):41 Table 8 Group A Cod. Messanensis S. Salvatoris 33 Cod. Caesenas Malatestianus plut. 28.2 Cod. Parisinus gr. 658

Group B

αων ακοββαει ισοββουνει

Cod. Ambrosianus A. 176 sup.

ἀών˙ ἀκουββαεί˙ ἰσουββουνεί˙

ἀὼν ἀκὼβ βαϊσὸβ βουνεὶ

Cod. Parisinus gr. 145

ἀών ἀκουββάει ἰσουββουνεί

ἀωνακὼβ βαεὶ ἰσοββουνεὶ

Cod. Bodleianus Auct. E.3.2

ἀων (→ Ἀων) ἀκοὺθβαεί ἰσουββουνεί

While it lies beyond the scope of this article to outline all of the considerations that went into classifying the manuscripts of Chrysostom’s Expositiones in Psalmos into these two groups (A and B) – this is explained in more detail in the coming edition – it will suffice to say that group A is generally the older group and group B is generally the younger group. Although group A usually contains the original reading, there are some corruptions that affect all of group A that do not affect group B. In such cases, group B may preserve the original reading. In this particular case, then, because there is no apparent corruption in the readings of group A, it is likely that the readings with Greek ο, rather than ου, are original. This would suggest that at some point early on in the transmission of group B, a scribe that knew Hebrew intentionally changed the vowel from ο → ου to better correspond to the phonology of “standard” Hebrew. In each of these two cases, the questionable reading in the patristic source was able to be compared to a parallel reading in one of the columnar witnesses. This made it easy to ultimately determine which reading is original and which reading reflects a scribal emendation in conformity with more “standard” Hebrew. Making such a determination, however, becomes much more difficult when there is no attested parallel reading in one of the columnar witnesses with which 41  I would like to especially thank Stefano Serventi of the Biblioteca Ambrosiana in Milan for checking the readings of Chrysostom in Cod. Ambrosianus A 176. sup. for me.

186

Benjamin Kantor

the transcription in the patristic sources may be compared. Nevertheless, even in such instances, conducting careful analysis of all the attested readings in the patristic’s witnesses and paying close attention to the morphophonological patterns of the Secunda can reveal the original reading. The final example of this section may illustrate just such a case: ουαλλα → ουαλεα || ‫יה‬ ָ ‫( וְ ָ֝ע ֗ ֶל‬Ps. 7.8): In Chrysostom’s Expositiones in Psalmos, the phrase ‫יה‬ ָ ‫( וְ ָ֝ע ֗ ֶל‬Ps 49:6) is cited in Migne’s edition as Οὐαλεά (PG 55, 90). The transcription is attested as follows among the Chrysostom manuscripts (Table 8): Table 9 Group A

Group B

Ἰωάννου τοῦ Θεολόγου 171

οὔ∙ἀλλὰ

Cod. Ambrosianus A. 176 sup.

οὐαλεά

Cod. Parisinus gr. 6584

οὖ∙ ἀλλὰ

Cod. Bodleianus gr. th. C.2

οὐαλεά

Cod. Messanensis S. Salvatoris 33

οὐάλλα

Cod. Parisinus gr. 145

οὐαλεά

Cod. Parisinus gr. 658

οὐαλλὰ

Cod. Bodleianus Auct. E.3.2

οὐαλέα

Rahlfs 1717

οὐα̅λ̅λο̅ς̅

On the surface, it may appear that the readings in group A are corrupt while group B has preserved the original Hebrew shape of the word. Indeed, the form ουαλλα in group A looks as though it has taken on the shape of two Greek words, οὗ ‘who/which’ and ἀλλά ‘but/rather’ (or even ἄλλος ‘other’ in Rahlfs 1717). The form ουαλεα, on the other hand, closely reflects the form in the Masoretic Text ‫יה‬ ָ ‫וְ ָ֝ע ֗ ֶל‬. There are two reasons, however, one based on historical Hebrew linguistics and the other on parallels found elsewhere in the Secunda, that ουαλεα is unlikely to be the original reading. First, the seghol [ε] vowel in ‫יה‬ ָ ‫ וְ ָ֝ע ֗ ֶל‬is typically explained as the result of assimilation in vowel height to the following qameṣ.42 However, it is by no means clear that etymological long */ā/ had shifted in quality to */ɔ̄/ by the time of the Secunda. Moreover, there is evidence that the sort of assimilation attested here (e > ε / _ɔ; a > ε / _ɔ), resulting in seghol in the Tiberian 42 Khan, “Pronominal Suffixes,” 3:268.

Discovering the Secunda

187

t­ radition, had not yet occurred in the Secunda: e.g., ιεσαχα ‫‘ ִ֫י ְש ֶ ֥עָך‬your (ms) salvation’ (Ps 18:36); αα ‫‘ ֶהאָ֣ח‬aha!’ (Ps 35:25); ααφής* (Rahlfs 1098: αλφη′ς) ‫‘ ֶ֝ה ָח ֗ ֵפץ‬who is pleased’ (Ps 35:27). The expected parallel form at the time of the Secunda, then, would have been *ʕālēhā, which would have been rendered as **αληα according to typical Secunda conventions. Second, the 3rd feminine singular suffix on a plural base is normally rendered in the Secunda as –α, similar to Aramaic (e.g., αμουδα ‫יה‬ ָ ‫מּוּד‬ ֣ ֶ ‫[ ַע‬Ps 75:4]).43 Evidence that this suffix was in use in contemporary Palestinian Aramaic is found in one of the Judaean Desert legal texts: e.g., ‫וקים‬ ‫‘ עלה‬and it is established for her’ (XHev/Se13, f1R:9).44 Therefore, in light of both historical considerations and parallel forms attested elsewhere in the Secunda, the expected form of the transcription should be **ουαλα. Indeed, one can make a strong case that ουαλλα is actually just a minor corruption of **ουαλα, which would parallel the Aramaic form ‫ ֲע ָלּה‬and cohere with the 3fs suffix in the form αμουδα ‫יה‬ ָ ‫מּוּד‬ ֣ ֶ ‫( ַע‬Ps 75:4). It is, of course, also possible that the unexpected gemination in ουαλλα might be original. The fact that the form with gemination was still recognized as Hebrew (i.e., not Greek οὗ + ἀλλά) is confirmed by the fact that it is accented as one word in Cod. Messanensis S. Salvatoris 33 and Cod. Parisinus gr. 658 and that an explanatory Hebrew gloss (‫יה‬ ָ ‫ )וְ ָע ְל‬is written in the margin of Cod. Parisinus gr. 658. Therefore, it may be concluded that ουαλλα is the original reading, being a possible corruption only of **ουαλα (or perhaps **ουαλαα). The form ουαλεα is a later emendation in conformity with Tiberian Hebrew. This conclusion actually has quite an important implication for understanding scribal knowledge of Hebrew in the early Middle Ages. The date of the earliest witness from group B with the reading ουαλεα, Cod. Ambrosianus A. 176 sup., is the twelfth century. Because the reading is shared across all four witnesses of group B that attest to this word, it is possible that the change occurred at least a century or two earlier. This suggests that, already in the first millennium, medieval scribes may have already been emending Greek transcriptions of Hebrew exhibiting non-standard phonology/morphology in conformity with a more “standard” (perhaps the Tiberian?) tradition of Hebrew. Such a scenario has important ramifications for understanding the historical development of the medieval Hebrew reading traditions during this same period.

43  In Origen’s Selecta in Psalmos, he specifically states that αμουδα is the Hebrew for στύλους αὐτῆς ‘its pillars’ in the plural (Selecta in Psalmos, PG 12, 1060, l.11). 44 Kantor, “The Second Column,” 275–276, 286. XHev/Se13 is published in Cotton and Yardeni, Aramaic, Hebrew, and Greek Documentary Texts from Naḥ al Ḥ ever and other Sites, 65–70.

188

Benjamin Kantor

4. Conclusions A comprehensive critical edition of the second column of the Hexapla has never before been published. Despite this, the material of the second column has still been used in a wide variety of disciplines. Therefore, the potential of a new comprehensive critical edition to make this impact even more deep and profound is great. This article will be concluded by outlining a number of ways in which the new critical edition may impact several related fields of scholarship, namely (1) Textual Criticism of the Bible, (2) Historical Hebrew linguistics, and (3) Patristic Scholarship.

4.1  Textual Criticism of the Bible After the publication of the new critical edition, for the first time, any student or scholar desiring to look up the rendering of a particular verse in the Secunda will have a single resource to which they can turn for a reliable and comprehensive presentation of the text of the second column. There are two layers to this: First, the “judicious retroversion” of the Hebrew consonantal text of column I, based on the attested text of the second column, can be compared with other forms of the consonantal text discussed in textual criticism of the Hebrew Bible (e.g., the Dead Sea Scrolls, the Masoretic Text, Hebrew Vorlage of the Septuagint). Second, the vocalization of the Hebrew text of column II, reliably presented in light of the best available manuscript evidence, can be compared with other vocalization traditions of Hebrew, whether directly attested in the medieval traditions (e.g., Tiberian, Babylonian, Palestinian, Samaritan) or implied by various translations of the Hebrew Bible (e.g., Aquila, Symmachus, Theodotion, the Vulgate, the Peshitta). All in all, though not much of it remains, the extant Secunda material may begin to serve a more prominent role in the discussion of text types and variants in the landscape of the Hebrew Bible’s text and vocalization in late antiquity. 4.2 Historical Hebrew Linguistics

The phonology and morphology of the Secunda has been well-covered by a number of Hebrew linguists over the past century.45 However, the corpus of analysis for these scholars has been made up of only the Ambrosiana palimpsest and/ or unreliable readings of Secunda quotations in other sources (e.g., the church fathers). The material for a phonological and morphological description of Secunda Hebrew, then, has really only been limited to about 70% of what is actu45 Speiser, “Hexapla”; Sperber, “Hebrew based upon Greek and Latin transliterations”; Brønno, Hexapla des Origenes; Janssens, “Origen’s Secunda”; Yuditsky, Diqduq; Kantor, “The Second Column”.

Discovering the Secunda

189

ally attested. As a result, there remain some gaps in the field’s understanding of both the phonology and morphology of Secunda Hebrew. The critical edition will thus provide historical Hebrew linguists with several hundred more words to factor in to phonological and morphological description. More importantly, however, it will make clear which among these words should be regarded as more secure readings, upon which it is possible to build phonological and morphological arguments, and which among these words are more tenuous readings, which should not be considered sufficient for informing on phonological or morphological matters.

4.3 Patristic Scholarship

There are two primary contexts in which the patristic authors quote the Secunda (or other Greek transcriptions of Hebrew) in their works: (i) in Jewish-Christian debates about the correct interpretation and/or translation of messianic prophecies and (ii) when interpreting exegetically difficult passages (or individual words that are difficult to translate). These two contexts often overlap. The new critical edition of the Secunda – the introduction of which includes a quotation of each patristic passage with accompanying translation and analysis – can thus serve as a sort of handbook for understanding the relationship of Christian scholars to Hebrew from the third to sixth centuries CE. Treating each church father separately, as the edition does, helps facilitate ascertaining the role that Hebrew played in each author’s scholarship, ideology, and cultural context. These details, of course, are important for anyone studying Jewish-Christian relations in the early church. Finally, it should also be noted that, because many of the same Greek transcriptions of Hebrew appear among multiple patristic authors, the presentation and analysis of the material in one resource allows threads of interdependence to become more conspicuous. The legacy of the Rich seminar on the Hexapla held at Oxford in 1994 continues to live on. After all, it is perhaps the second column (more than any of the other columns) that has been most in need of a critical edition. For a text in which every single letter – not just every single word – has paramount importance for the significance of each transcription, a critical edition that establishes the text has long been overdue. Fortunately for Hexapla scholars, the coming publication of the new critical edition of the second column is not an isolated occurrence. It comes on the back of roughly a decade now of new critical editions of the Hexaplaric fragments (for various books) coming to light. It is thus a pleasure and an honor to be able to contribute to this worldwide scholarly endeavor with the publication of a new critical edition of the Secunda in the near future. This paper has shared only a handful of insights gained through the preparation of such an edition.

190

Benjamin Kantor

5. Bibliography Albrecht, Felix, “Report on the Göttingen Septuagint,” Textus 29 (2020): 201–220. Brønno, Einar, Studien über hebräische Morphologie und Vokalismus auf Grundlage der Mercatischen Fragmente der zweiten Kolumne der Hexapla des Origenes (Leipzig: D MG, 1943). Cavenaile, Robert, Corpus Papyrorum Latinarum (Wiesbaden: Otto Harrassowitz, 1958). Ceriani, A. M, Monumenta sacra et profana: Tom. VII: Codex Syro-Hexaplaris Ambrosianus: Photolithographice editus (Milan: Bibliotheca Ambrosiana, 1874). Cotton, Hannah, and Ada Yardeni, eds., Aramaic, Hebrew, and Greek Documentary Texts from Naḥ al Ḥ ever and other Sites, with an Appendix Containing Alleged Qumran Texts. Part 2: Cave 4.XXIV. DJD XXVII (Oxford: Clarendon Press, 1999). Dickey, Eleanor, “Columnar Translation: An Ancient Interpretive Tool that the Romans Gave the Greeks,” The Classical Quarterly 65/2 (2015): 807–821. Field, Frederick, Origenis Hexaplorum quae supersunt (Oxford: Clarendon, 1875). Flint, Peter W., “Columns I and II of the Hexapla: The Evidence of the Milan Palimpsest (Rahlfs 1098).” Pages 125–132 in Origen’s Hexapla and Fragments: Papers presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25th–3rd August 1994, ed. Alison Salvesen (Tübingen: Mohr Siebeck, 1998). Gentry, Peter, “Did Origen Use the Aristarchian Signs in the Hexapla?” in XV Congress of the International Organization for Septuagint and Cognate Studies: Munich, 2013, ed. Wolfgang Kraus, Michael N. van der Meer, and Martin Meiser (Atlanta: SBL Press, 2016), 133–147. –, “Origen’s Hexapla,” in The Oxford Handbook of the Septuagint, eds. Alison Salvesen and Michael Law (Oxford: Oxford University Press, 2021). Grafton, Anthony, and Megan Williams, Christianity and the Transformation of the Book: Origen, Eusebius, and the Library of Caesarea (Cambridge, MA.: The Belknap Press of Harvard University Press, 2006). Hatch, Edwin, and Henry A. Redpath, A Concordance to the Septuagint and the Other Greek Versions of the Old Testament. Including the Apocryphal Books (Oxford: Clarendon, 1897). Heine, Ronald E., Origen: Scholarship in the Service of the Church (Oxford: Oxford University Press, 2010). Holl, Karl, Epiphanius I: Ancoratus und Panarion haer. 1–33. GCS 10.1 (Berlin: De Gruyter, 2013). Janssens, Gerard, Studies in Hebrew Historical Linguistics based on Origen’s Secunda. (Leuven: Peeters, 1982). Kantor, Benjamin, The Second Column (Secunda) of Origen’s Hexapla in Light of Greek Pronunciation (PhD diss., The University of Texas at Austin, 2017). Khan, Geoffrey, “Pronominal Suffixes,” in EHLL 3 (Leiden: Brill, 2013), 267–272. Margolis, Max L., “The pronunciation of the shewa according to new Hexaplaric material,” American Journal of Semitic Languages 26 (1909): 62–70. Meade, John, “An Analysis of the Syro-Hexapla of Job and Its Relationship to Other Ancient Sources,” ARST 14 (2016): 212–241. Mercati, Giovanni, Psalterii Hexapli Reliquiae (Vatican: Bibliotheca Vaticana, 1958).

Discovering the Secunda

191

—, Psalterii Hexapli Reliquiae, osservazioni: Commento critico al testo dei frammenti esaplari (Vatican: Bibliotheca Vaticana, 1965). Migne, J. P., Origenis: opera omnia: tomus secundus. Vol. 12 of Patrologiae cursus completus: series graeca (Paris: Migne 1862). —, Eusebii Pamphili: opera omnia quae exstant: tomus quintus. Vol. 23 of Patrologiae cursus completus: series graeca (Paris: Migne 1857). —, S. P. N. Joannis Chrysostomi: opera omnia quae exstant: tomus quintus. Vol. 55 of Patrologiae cursus completus: series graeca (Paris: Migne 1862). —, Theodoreti: opera omnia: tomus primus. Vol. 80 of Patrologiae cursus completus: series graeca (Paris: Migne 1860). Nautin, Pierre, Origène: Sa vie et son oeuvre (Paris: Beauchesne, 1977). Pretzl, Otto, “Die Aussprache des Hebräischen nach der zweiten Kolumne der Hexapla des Origenes,” Biblische Zeitschrift 20 (1932): 4–22. Revell, E. J., “The Oldest Evidence for the Hebrew Accent System.” Bulletin of the John Rylands Library 54/1 (1971): 214–222. —, “Biblical Punctuation and Chant in the Second Temple Period,” Journal for the Study of Judaism 7/2 (1976): 181–198. Speiser, Ephraim A., “The pronunciation of Hebrew according to the transliterations in the Hexapla.” Jewish Quarterly Review 16/4 (1926): 343–382; 23/3 (1933): 233–265; 24/1 (1933): 9–46. Sperber, Alexander, “Hebrew based upon Greek and Latin transliterations,” Hebrew Union College Annual 12–13 (1937): 103–274. Thompson, Edward Maunde. A Handbook of Greek and Latin Paleography (New York: D. Appleton and Company, 1893). Yuditsky, Alexey, “ʕIyun ḥ adash bexitve yad O 39 misifriyat ambrosiana shebemilano,” Leshonenu 68 (2006): 63–71. —, Diqduq haʕivrit shel taʕatiqe ʔorigenes (Jerusalem: The Academy of the Hebrew Language, 2017). Ziegler, Joseph, Eusebius Werke 9: Der Jesajakommentar (Berlin: De Gruyter, 2013).

Daniel Olariu Recensional Additions Insights from Theodotion Daniel* 1. Preliminary Remarks The Septuagint studies in the book of Daniel afford the rare opportunity to investigate two parallel versions for the entire book, namely, the Old Greek (OG/ OG-Dan) and Theodotion (Th-Dan).1 This study explores recensional exegesis, seeking to analyze the contextual additions extant in the two versions, within the framework of a translation-revision relationship. Previous recensional studies take little account of the exegetical elements in the recensional units which have been investigated and make no attempt at systematization. Consequently, the need for such an inquiry, together with the fortunate situation to benefit form a complete Theodotionic text which reflects a recension, provide the ideal context for this study. The nature of the topic in question necessitates a priori decisions related to at least two areas of research. The first relates to the unity of OG-Dan. Starting with Alberts,2 this unity was challenged, since he advanced the theory that Daniel 4–6 reflects the work of a distinct translator, his style differing from the reminder of the book (i.e. chaps. 1–3, 7–12). McLay has reached a similar decision, maintaining that “The analysis of OG supported the thesis of Albertz that chaps. 4–6 originate from a translator different from the person(s) who translated 1–3; 7–12.”3 However, our view regarding the unity of the OG’s translation is informed by the latest research on this issue which was conducted by Amara. After systematical*  This study is partly based on segments of the author’s PhD diss., supervised by Prof. Emanuel Tov and Prof. Michael Segal: An Analysis of the Revisional Process in Theodotion’s Greek Text of Daniel (Hebrew University of Jerusalem, 2021). 1  The analysis makes use of the latest critical editions of the Göttingen series: Ziegler, Munnich, and Fraenkel, Susanna, Daniel, Bel et Draco; 2nd rev. ed. of Joseph Ziegler (ed.), Susanna, Daniel, Bel et Draco, Septuaginta: Vetus Testamentum Graecum XVI/2 (Göttingen: Vandenhoeck and Ruprecht, 1954). The Old Greek text was updated by Munnich in light of the newly published Papyrus 967, which was not available for the first edition prepared by Ziegler in 1954. Similarly, Fraenkel has updated the Theodotion text with readings from fragments not included in the previous edition. Occasionally, Rahlfs’ edition was also consulted where necessary. Cf. Rahlfs and Hanhart, Septuaginta. 2 Albertz, Der Gott des Daniel. 3 McLay, The OG and Th Versions of Daniel, 242; see also McLay, “The Old Greek Translation of Daniel IV–VI and the Formation of the Book of Daniel;” McLay “The Greek Translations of Daniel 4–6.”

194

Daniel Olariu

ly comparing the translation technique of Daniel 4–6 with the reminder of the book, she unequivocally concluded that OG-Dan is the product of a single hand.4 Another area of research that has even more bearings on our endeavor relates to the proper characterization of the relationship between OG-Dan and Th-Dan. The history of research indicates the following intriguing situation: scholars for years have taken as granted that Th-Dan is in the nature of a revision of the OG.5 However, the first systematic studies on the topic concluded the opposite, that Th-Dan represents a de novo translation.6 This issue is being investigated anew by Olariu, his results indicating that Th-Dan reflects the specimen of a literal revision.7 Consequently, the assumption that Th-Dan constitutes a revision permeates through this analysis and, hence, the goal of discussing exegesis in Th-Dan is tantamount to investigating recensional exegesis. To capture the full range of aspects related to the topic of recensional additions, Part 2 will review relevant literature which discusses exegetical aspects in the Septuagint of Daniel. The review will set the stage for the next part: recensional contextual additions. Part 3 consists of three sections, each of them providing important information for the overall assessment of Th-Dan’s exegesis. We will first exemplify the influence of the OG on Th-Dan in keeping some of its additions, arguing about the interference of the former on the latter’s lexical choices. We then evidence Th-Dan’s tendency to reject various contextual ad4  In Amara’s words, “our findings show that there are no distinctive disagreements between Chapters 4–6 and the other chapters of OG-Daniel. Consequently, we can attribute the translation of these chapters to the translator of the rest of the book.” Amara, The Old Greek Version of Daniel, iii (abstract). See also Amara, “Bel and the Dragon,” 126, n. 5. 5  Among those that have directly or indirectly expressed views that Th-Dan reflects a revision, we mention: Barthélemy, Les devanciers d’Aquila; Delcor, Le livre de Daniel, 21–22; Munnich, “Le text de la Septante,” 150–57; Schmitt, Stammt der sogenannte “θ”–Text bei Daniel wirklich von Theodotion?, 112; Schmitt, “Die griechischen Danieltexte (“θ” und o′) und das Theodotionproblem;” Jeansonne, The Old Greek Translation of Daniel 7–12, 32–57. 6  Credit is due to McLay’s challenge to the common opinion and his case for Th-Dan as an independent translation. In addition to his literature already mentioned above (cf. n. 3), see further McLay’s following studies: McLay, “Double Translations in the Greek Versions of Daniel;” McLay, “Daniel (Old Greek and Theodotion);” McLay, “It’s a Question of Influence: The Theodotion and the Old Greek Texts of Daniel;” McLay, “Syntactic Profiles and the Characteristics of Revision: A Response to Karen Jobes;” McLay, “The Relationship Between Greek Translations of Daniel 1–3.” Demonstrably, McLay’s characterization of Th-Dan as an independent translation is to a lesser or grater extend shown by the studies of Obiajunwa, Semitic Interference in Theodotion-Daniel; Amara, The Old Greek Version of Daniel; and Braasch, Die LXX-Übersetzung des Danielbuches. 7  In addition to his doctoral dissertation mentioned above, see further Olariu, The Quest for the Common Basis; Olariu, “18.1 Textual History of Daniel;” Olariu, “Criteria for Determining the Common Basis of the Greek Versions of Daniel;” Olariu, “The Mechanics of Recensional Process;” Olariu, “How Does a Reviser Work? Insights from Theodotion’s Recension of Daniel.”

Recensional Additions

195

ditions present in OG-Dan. This fact points to the character of Th-Dan as a literal revision. The unique contextual additions in Th-Dan are lastly considered. These elements, which are independently introduced by Th-Dan, are interpreted as genuine Theodotionic exegetical touches.

2. History of Research The current state of literature8 reveals that the OG has captured the scholarly attention for years, whereas Th-Dan “has often been neglected in the research like a younger sibling following in the footsteps of the successful older brother.”9 Two incentives have presumably determined this status quaestionis. The first relates to the intriguing fact that, during its transmission history, OG-Dan has been almost obliterated. Indeed, because of the decision of the church fathers to replace the OG with Th-Dan in the use of the church,10 the best codices contain Th-Dan, the OG being preserved through the ages only in Codex Chisianus (MS. 88)11 and in the Syriac translation of the Syro-Hexapla (Syh).12 Both attest Origen’s Hexaplaric recension. However, the discovery of Papyrus 967 in 1931 boosted interest

8  For surveys on the state of literature from a textual standpoint, see Di Lella, “The Textual History of Septuagint-Daniel and Theodotion-Daniel;” Olariu, “18.1 Textual History of Daniel;” Bledsoe, “The Relationship of the Different Editions of Daniel: A History of Scholarship.” 9 McLay, Translation Technique and Textual Studies in the Old Greek and Theodotion Versions of Daniel, 332. 10  Regarding the social and historical incentives which have led to the OG’s replacement, Jerome comments in the preface to his translation of Daniel: “The Septuagint version of Daniel the prophet is not read by the Churches of our Lord and Saviour. They use Theodotion’s version, but how this came to pass I cannot tell. Whether it be that the Language is Chaldee, which differs in certain peculiarities from our speech, and the Seventy were unwilling to follow those deviations in a translation; or that the book was published in the name of the Seventy, by some one or other not familiar with Chaldee, or if there be some other reason, I know not; this one thing I can affirm – that it differs widely from the original, and is rightly rejected.” Weber and Gryson, Biblia sacra iuxta vulgatam versionem, 1341 (lines 1–7). Translation quoted from The Principal Works of St. Jerome, trans. W. H. Fremantle, NPNF 2/6 (Grand Rapids: Eerdmans, 1979), 492. 11  According to Swete, The Old Testament in Greek, 3.xii–xiii, MS 88 belonged to Pope Alexander VII who was part of the Chigi family. Alexander entrusted the manuscript to the Vatican librarian Leo Allatius for publication. However, the publication was delayed approximately 100 years until after the death of Bianchini and Regibus who continued the work of Allatius. It was finally published by Magistris, Daniel secundum Septuaginta ex tetraplis Origenis. 12  Syh is a translation into Syriac of the fifth column included in Origen’s Hexapla. The translation was carried out from 615 to 617 C.E. and is attributed to Paul of Tella. Syh of Daniel has been preserved in Codex Ambrosianus (ca. ninth century), which was published by Ceriani, Codex syro-hexaplaris Ambrosianus photolithographice editus, folios 143a–151b.

196

Daniel Olariu

in OG-Dan in twentieth century,13 the investigation getting advantage for the first time of readings which antedated Origen’s Hexapla.14 In addition to its intriguing transmission history, the second OG feature that attracted the scholarly attention was its many translational disparities with the Masoretic Text (MT).15 The analysis of these disparities announced promising results, especially considering the fortunate situation that the OG translational style could be checked against the literal style of Th-Dan. Among the most exciting findings was the presence of numerous exegetical elements in the OG. However, a side-effect of the attention to the conspicuous differences between OG-Dan and MT is the complete absence of any study to single out and interpret the exegetical elements in Th-Dan. At the beginning of the twentieth century, scholarly reflections on the exegetical elements in the Greek versions of Daniel entirely concerned the OG. Beside the extensive analyses of Bludau, Rießler, and Jahn,16 the other comments were dispersed and mainly confined to commentaries such as those produced by Driver,17 Montgomery,18 and Charles.19 For many of these differences, the scholarly opinions at the turn of twentieth century have pointed to the “free,” “paraphrastic,” and “midrashic” style of the OG’s translator. Notwithstanding the fact that in many instances he also holds the OG translator responsible for a midrashic and theological Tendez, Bludau’s analysis her13 The editiones principes of Papyrus 967 for the Book of Daniel have been prepared by Hamm, Daniel 1–2; Hamm, Daniel 3–4; Kenyon, Daniel 3,72–6,18; Geißen, Daniel 5–12; Roca-­ Puig, Daniel: dos semifolios del còdex 967; Roca-Puig, “Daniel: Due Semifogli del Codex 967.” 14  In addition to the studies produced by Amara, Obiajunwa, and Jeansonne (cf. nn. 5–6), other important analyses that took advantage of the discovery of Papyrus 967 are Ashley, The Book of Daniel Chapters I–VI; McCrystall, Studies in the Old Greek Translation of Daniel; Wenthe, The Old Greek Translation of Daniel 1–6. 15  The OG differs from MT displaying smaller deviations such as pluses (e.g., Dan 4:9[12]; 6:4[3]) and minuses (e.g., Dan 4:17[20]; 5:2; 6:11[10]), which occur repeatedly over against MT. Furthermore, the OG also displays long minuses (e.g., Dan 4:3–6[6–9]; 4:32[35]; 5:13–16, 18–22; 6:16[15]17[16]), long pluses (e.g., Dan 4:37[34]; 5:0 [the brief abstract of Daniel 5]), differences in layout (the epistolary introduction that precedes Daniel 4 of MT [Dan 3:31–33] is reflected in LXX at the end of the story, cf. Dan 4:34[37b,c]), doublets (e.g., Dan 6:13[12a]; 4:14[17a]), and overall disparity in length (i.e., Daniel 4 is estimated as one-quarter longer than mt whereas Daniel 5 reflects a shorter version of MT). Note also the intriguing absence in OG of Daniel’s qualification according to which he purportedly owns “the Spirit of the Holy God” (Dan 4:5–6, 15; 5:11, 14). 16 Bludau, Die alexandrinische Übersetzung des Buches Daniel; Paul Rießler, Das Buch Daniel, Textkritische Untersuchung; Paul Rießler, Das Buch Daniel, Kurzgefasster wissenschaftlicher Kommentar; Gustav Jahn, Das Buch Daniel. 17 Driver, The Book of Daniel. 18 Montgomery, Daniel. 19 Charles, A Critical and Exegetical Commentary. See also his earlier commentary, Charles, Daniel.

Recensional Additions

197

alded a positive approach to the OG, making an important distinction between Daniel 4–6 and the reminder of the chapters. He basically argued that, in these three chapters, the OG reflects a Vorlage much different from MT. As a result, the nature of the Vorlage of these chapters proved to be of interest in the twentieth century with scholars debating whether the OG’s Vorlage reflects a superior text to MT or the opposite. To a high degree, the debate over the nature of the OG’s Vorlage has diverted attention from the translator as a producer of exegetical readings. However, the studies of Bruce20 and McCrystall21 have revived Bludau’s claim that the translator rendered his Semitic text from a deliberate theological viewpoint. They emphasized the role of the translator as an interpreter who created readings that are valuable for the history of exegesis and reception of the text. The renewal of the claim that the OG manifests tendentious exegesis has attracted a response from Jeansonne who has disputed the examples proposed by Bludau, Bruce and McCrystall.22 Instead, Jeansonne underscored the accurate character of the OG and suggested alternative explanations for the alleged tendentious rendering, e.g., mechanical variants and errors,23 post-translational changes,24 variants illuminated through contextual comparison.25 However, in spite of Jeansonne’s positive affirmation, Amara’s recent study alleges that the OG translation imbeds interpretative as well as theological elements, e.g., monotheism, uniqueness of God, God’s angels, etc.26 Since these deviations from the MT ultimately reflect on its free translational style, she writes, “The extremely non-literal nature of this translation [OG-Dan] turns it almost worthless for text-critical purposes, since one can never be certain whether a variant text actually goes back to a different Vorlage or originated from the hand of the translator.”27 The literature review indicates that OG-Dan has been preferred over Th-Dan for investigating exegesis. This preference towards the OG is somehow expected, considering both its intriguing transmission history and prominent disparities over against MT. As a result, there is no systematic study that tackles the issue of exegesis in Th-Dan. The time is ripe for such an investigation, especially into a text which demonstrably displays the idiosyncrasies of a revision.

20 Bruce, “The Earliest Old Testament Interpretation;” Bruce, “The Oldest Greek Version of Daniel;” Bruce, “Prophetic Interpretation in the Septuagint.” 21 McCrystall, “Studies in the Old Greek Translation of Daniel.” 22 Jeansonne, The Old Greek Translation of Daniel 7–12, 24–31. 23 Jeansonne, The Old Greek Translation of Daniel 7–12, 70–82. 24 Jeansonne, The Old Greek Translation of Daniel 7–12, 83–102. 25 Jeansonne, The Old Greek Translation of Daniel 7–12, 103–130. 26 Amara, The Old Greek Version of Daniel, 154–190. 27 Amara, The Old Greek Version of Daniel, iv (English abstract).

198

Daniel Olariu

3. Recensional Contextual Additions In its broader definition, exegesis refers to the practice of translators to deviate from a literal representation of the underlying text, by adding, omitting, or substituting its elements. Furthermore, by means of these techniques, the translators embed their linguistic, exegetical, and theological ideas. It is important to note that these departures do not reflect variant readings, “though from a formal point of view, exegetical elements stand in the same relationship to MT as variant readings.”28 That is to say, the presence of these elements reflect rather the style of the translator. Consequently, the more acute the translator’s propensity towards literalism is, the fewer instances of exegetical elements transpire throughout his work. Recensions are not free of such elements. However, their nature and extent significantly differ from a translation. For instance, a comparative analysis of the Greek versions of Daniel convincingly argues that the OG is a free translation while Th-Dan reflects the idiosyncrasies of a literal revision. This section begins to document the interference of the OG’s style on Th-Dan’s agenda. The fact that OG contextual additions are singled out in Th-Dan’s revision argues for the OG’s influence on the reviser’s lexical choices. In the second part of this section, the recensional corrections are exemplified with a select number of cases, highlighting the reviser’s tendency to eliminate OG-Dan additions. As part of the recensional process, Th-Dan endeavored to eliminate exegetical elements from its base text – the OG – in order to bring it to conform quantitatively and qualitatively with its MT-like Vorlage. Lastly, and in agreement with the study’s goal to document recensional exegesis, several relevant cases are discussed which typify Th-Dan’s unique contextual additions. The last type of added elements affords rare glimpses into the reviser’s linguistic and exegetical concerns while undertaking his work. It is important to be noted at this point that all added elements discussed in this study share the common denominator that they are the result of contextual exegesis. In this respect, Tov explains:29 “In a way, all forms of exegesis might be called ‘contextual exegesis’, because the translators’ concept of ‘context’ was wider than ours. Translators created relationships between words not only when they occurred in the immediate context, but also when they occurred in remote contexts. Furthermore, the translator might insert into the translation any idea the source text called to mind.” However, in presenting the data, both form and content criteria are taken into consideration. By form, the contextual elements are described first in relationship to its source text. Accordingly, the elements can be described as additions, 28 Tov, The Text-Critical Use of the Septuagint in Biblical Research, 50. 29 Tov, The Text-Critical Use of the Septuagint in Biblical Research, 50.

Recensional Additions

199

omissions, and substitutions. By content, we mainly refer to the probable factors that lead to their formation. As such, it can be distinguished between linguistic, exegetical, and theological incentives. This study confines to investigate contextual additions in Th-Dan within the framework of a translation-revision theory. The organizing principle of presenting the data in this section is by charts, grouping the readings from MT-Dan, OG-Dan, and Th-Dan in parallel columns. The examples adduced to support each tendency do not follow a descending pattern. Since the same phenomenon shines through each of them, the items below are listed in the sequence of how they occur in Daniel’s book. The string of words under discussion appears as underlined text in the columns of each chart. The contextual elements which appear as plusses in OG-Dan over against MT-Dan and Th-Dan are marked as underlined text in the middle column. When Th-Dan adopts a contextual plus from the OG, it is equally marked as underlined text in both the second and the third column. Finally, Th-Dan’s unique contextual additions over against MT and the OG are marked as underlined text in the third column.30

3.1  Preservation in Th-Dan of Contextual Additions from OG-Dan

The reviser has sometimes adopted contextual additions from the OG such as particles, prepositions, pronouns, and helping verbs. However, many such deviations share the common denominator that they, to a certain degree, are expected to show up in conformity with the rules of the grammar of the Greek language.31 They represent linguistic additions pertaining to style. However, most of the examples below are of another kind: they represent distinctive agreements. These are cases in which the OG’s influence on Th-Dan for selecting its equivalents can be demonstrated.

3.1.1  Linguistic Elements The reviser sometimes has retained linguistic additions which demonstrably pertains to the OG’s style. 30  For the excerpts quoted in the subsequent charts we have used the following editions: for the English translation—JPS (JPS Hebrew–English Tanach: The Traditional Hebrew Text and the New JPS Translation, 2nd ed.); for both the OG’s and Th-Dan’s translation—NETS (Albert Pietersma and Benjamin G. Wright, eds., A New English Translation of the Septuagint). 31  E.g., OG-Dan and Th-Dan agree with each other thrice adding the preposition εἰς (+acc) “into,” “to,” “for” in Dan 1:1, 2; 11:8. However, it can be shown that they compensate the lack of a preposition attached to the verb ‫“ בוא‬to come.” The verb occurs over forty times in MTDan, being followed most of the time by a preposition. However, since the verb is devoid of a preposition in the cases in question, Th-Dan has expectedly accepted the OG’s addition of the preposition εἰς. Th-Dan’s decision is in accordance to the requirements of the verb that he employs, namely, ἔρχομαι “to come,” especially since it was used with a noun of place, indicating the destination.

200

Daniel Olariu

No. 1. Dan 2:24 MT

LXX

Th

‫ ]…[ […] לחכימי בבל אל תהובד‬Τοὺς μὲν σοφιστὰς τῆς […] Τοὺς σοφοὺς ‫ העלני קדם מלכא […]׃‬Βαβυλωνίας μὴ ἀπολέσῃς, Βαβυλῶνος μὴ ἀπολέσῃς,

εἰσάγαγε δέ με πρὸς τὸν βασιλέα, […]. […] “Do not do away with […] “Do not destroy the the wise men of Babylon; savants of Babylon, but bring me to the king […]. bring me in to the king […]. 

εἰσάγαγε δέ με ἐνώπιον τοῦ βασιλέως, […]. […] “Do not destroy the wise men of Babylon, but bring me in before the king […]. 

The marker of contrast δέ “but” can be used independently or as part of the idiomatic construction μέν…δέ, denoting contrast in a set of items, i.e. “on the other hand…but on the other hand.”32 A cursory check suggests that δέ belongs to the style of OG-Dan occurring about forty-seven times, independently or idiomatically.33 It expectedly renders eleven times the Semitic particle ְ‫“ ו‬and” (1:7, 15, 18; 2:6, 13, 16, 41, 42Ra; 3:15; 7:72, 16) but also once each ‫“ כל קבל דנה‬thereupon” (2:241); 0‫“ די־‬who,” “which” (2:262); 0‫“ אדין־‬then” (4:16[19]); and 0‫“ אף־‬also” (6:23[22]1). However, the fact that δέ most of the time was stylistically added in the OG (2:5, 7, 242, 261, 27, 30, 332X, 36, 43, 44; 3:12, 16, 28[95]; 6:11[10], 13[12]2X, 17[16]; 7:71; 8:4; 12:22X) argues that the term belongs to its translational style.34 Th-Dan contrasts sharply with the OG: δέ occurs only ten times. However, out of ten, Th-Dan agrees with the OG in rendering five times both ְ‫( ו‬2:6, 41–42; 3:15) and 0‫“ אף־‬also” (6:23[22]). Furthermore, in three instances wherein the OG=0, Th-Dan renders 0‫“ אדין־‬then” (2:15); ְ‫( ו‬4:15[18]); and 0‫“ ברם־‬but.” Consequently, Th-Dan’s style was to minimize as much as possible the marker of contrast, repeatedly replacing δέ with καί “and,” “also,” “even,” “and yet,” “but” from the base text when δέ rendered various Semitic words. In addition, in the three

32  Louw and Nida, Greek-English Lexicon of the New Testament Based on Semantic Domains, 794–795. 33  Excepting the additions, δέ appears three times in chap. 1 (vv. 7, 15, 18); eighteen times in chap. 2 (vv. 5–7, 13, 16, 242X, 262X–27, 30, 332X, 36, 41, 42Ra–44); five times in chap. 3 (vv. 12, 15–16, 23, 28[95); six times in chap. 4 (vv. 16[19], 19[22], 28[31], 30[33], 34[37b,c]2X); once in chap. 5 (v. 0[abs.]); eight times in chap. 6 (vv. 5[4]–6[5], 11[10], 13[12]2X, 17[16], 23[22]2X); trice in chap. 7 (vv. 72X, 16; 8:4); and twice in chap. 12 (v. 22X). 34  In the remaining instances—ten times, δέ occurs in OG-Dan as part of long plusses: MT=0: Dan 3:23; 4:19[22], 28[31], 30[33], 34[37b,c]2X; 5:0; 6:5[4]–6[5], 23[22]2).

201

Recensional Additions

cases when OG=0, Th-Dan always uses δέ for a Semitic word. The only exceptions are Dan 2:24, 30 where influenced by the OG, Th-Dan adds δέ stylistically.35 No. 2. Dan 5:29 MT

LXX

Th

‫ ]…[ […] והלבישו לדניאל‬ἐνέδυσε τὸν Δανιηλ […] καὶ ἐνέδυσαν τὸν ‫ ארגונא והמונכא די דהבא‬πορφύραν καὶ μανιάκην Δανιηλ πορφύραν καὶ τὸν ‫ על צוארה […]׃‬χρυσοῦν περιέθηκεν αὐτῷ μανιάκην τὸν χρυσοῦν […].

[…] they clothed Daniel in purple, placed a golden chain on his neck, […].

[…] clothed Daniel in purple, and he put a gold torque on him, […]. 

περιέθηκαν περὶ τὸν τράχηλον αὐτοῦ […]. […] and they clothed Daniel in purple, and the gold torque they put around his neck […]. 

The language of v. 29 reflects the literal fulfilment of the king’s promise repeated twice in vv. 7, 16, for the one who would be able to read and interpret the writing on the wall. In v. 7, Th-Dan eliminates the OG’s 3rd sing. fut. act. ind. verb περιθήσει “he will put on,” which was added to compensate the elliptical clause in the source language.36 In v. 16, not benefiting from a parallel translation in the OG, the reviser adds 3rd. sing. fut. mid. ind. ἔσται “[the torque] will be,” interpreting the wording ‫ והמונכא די דהבא על צוארך‬as a nominal clause.37 Consequently, ἔσται is expected. However, in v. 29, the presence of περιέθηκαν in Th-Dan indicates dependence on the OG. Not only the equivalent does not reflect Th-Dan’s style 35  Further evidence for the dependence of Th-Dan on the OG emerges by analyzing the idiomatic ‫“ מנהון )…( ומנהון‬they were partly this (…) partly that” (HALOT 5:1919), which was rendered in OG-Dan 2:33, 41 with the unique LXX construction μέρος μέν τι (…) μέρος δέ τι. In 2:33, Th has μέρος τι but lacks μέν…δέ of the idiomatic construction. However, Th-Dan retains the full idiomatic expression in v. 41 and uses it further in v. 42. Additional evidence of Th-Dan’s dependence on the OG derives from its use of μέν (…) δέ. The collocation is used eight times in OG-Dan both before and after 2:41, and consequently it can be assigned to the OG style. Its presence in Th-Dan at v. 41 reveals the OG’s influence on Th-Dan’s lexical choices. See Olariu, The Quest for the Common Basis, 77–78. 36  The Semitic clause ‫ די דהבא על צוארה והמונכא‬was rendered in the versions as καὶ μανιάκην χρυσοῦν περιθήσει αὐτῷ “and the gold torque he will put on him” (OG-Dan) and καὶ ὁ μανιάκης ὁ χρυσοῦς ἐπὶ τὸν τράχηλον αὐτοῦ “and [have] the gold torque around his neck” (Th-Dan). 37  Th-Dan renders the phrase literally: καὶ ὁ μανιάκης ὁ χρυσοῦς ἔσται ἐπὶ τὸν τράχηλόν σου “and the gold torque will be around your neck.”

202

Daniel Olariu

displayed in vv. 7, 16, but also περιέθηκαν demonstrably stems from the OG (cf. v. 7).

3.1.2  Exegetical Elements In many instances, the reviser follows the OG’s lead in adding exegetical elements. They were generally introduced to clarify the source text and to make explicit what is implied in its immediate or larger context. No. 3. Dan 2:32 MT

LXX

Th

‫ חזה הוית עד די התגזרת‬ἑώρακας ἕως ὅτου ἐτμήθη ἐθεώρεις, ἕως οὗ ἐτμήθη ‫ אבן די לא בידין […]׃‬λίθος ἐξ ὄρους ἄνευ χειρῶν λίθος ἐξ ὄρους ἄνευ χειρῶν As you looked on, a stone was hewn out, not by hands, […].

[…]. And you saw until when a stone was cut from a mountain, without hands, […].

[…]. you kept looking until a stone was severed from a mountain, not by hands, […]. 

Notwithstanding that the agreement between the OG and Th-Dan in adding ἐξ ὄρους would suggest at the first glance a different Vorlage, the versions provide a valuable piece of information which convincingly indicates that the reading reflects contextual harmonization in OG-Dan. The clue is provided in v. 45 wherein the Semitic phrase ‫ מטורא אתגזרת אבן‬was translated with ἐξ ὄρους τμηθῆναι λίθον in the OG while Th-Dan has ἀπὸ ὄρους ἐτμήθη λίθος. As Montgomery has aptly remarked, ἐξ ὄρους reflects “an intrusion from LXX” in v. 32, since Th has ἀπὸ ὄρους in v. 45 while the OG displays ἐξ ὄρους as in v. 32.38

38  Cf. Montgomery, Daniel, 169. Collins, Daniel, 151, follows Barthélemy in interpreting the addition as an “explanatory gloss;” Barthélemy, Critique textuelle de l’Ancien Testament, 438. However, not only that their view is more speculative as compared to the wide attested practice of harmonization, but even though someone admit that ἐξ ὄρους entered the OG as an interpolation, its presence in Th-Dan still argues for its dependence on the OG.

203

Recensional Additions

No. 4. Dan 7:24 MT

LXX

‫ ]…[ […] ואחרן יקום אחריהון‬καὶ ὁ ἄλλος βασιλεὺς ‫ והוא ישנא מן קדמיא […]׃‬μετὰ τούτους στήσεται, καὶ αὐτὸς διοίσει κακοῖς ὑπὲρ τοὺς πρώτους […]. […] and after them […] and another king shall another will arise. He rise after these. And he will be different from the shall excel more than the former ones, […]. former ones in evil, […]. 

Th […] καὶ ὀπίσω αὐτῶν ἀναστήσεται ἕτερος, ὃς ὑπεροίσει κακοῖς πάντας τοὺς ἔμπροσθεν, […]. […] and another shall arise behind them who shall surpass in evil all the previous ones, […]. 

V. 24 is part of the lengthier interpretation section dealing with the little horn and the fourth beast. Th-Dan incorporates in its text the OG exegetical addition of κακοῖς, which was likely meant to underscore that the little horn’s surpassing power over the other horns regards its malevolent actions. No. 5. Dan 8:13 MT

LXX

‫ ]…[ […] עד מתי החזון התמיד‬Ἕως τίνος τὸ ὅραμα ‫ והפשע שמם תת וקדש‬στήσεται καὶ ἡ θυσία ἡ ‫ וצבא מרמס׃‬ἀρθεῖσα καὶ ἡ ἁμαρτία

Th

[…] Ἕως πότε ἡ ὅρασις στήσεται, ἡ θυσία ἡ ἀρθεῖσα καὶ ἡ ἁμαρτία ἐρημώσεως ἡ ἐρημώσεως ἡ δοθεῖσα, καὶ δοθεῖσα, καὶ τὸ ἅγιον καὶ ἡ τὰ ἅγια ἐρημωθήσεται εἰς δύναμις συμπατηθήσεται; καταπάτημα; […] “How long will [what […] “How long will this […] “For how long will was seen in] the vision vision continue: even the this vision continue: the last – the regular offering sacrifice, which has been sacrifice, which has been be forsaken because taken away, and the sin of taken away, and the sin of of transgression; the desolation that has been desolation that has been sanctuary be surrendered given and the sanctuaries given and the sanctuary and and the [heavenly] host be will be desolated unto the host will be trampled trampled?” trampling?” […].  under foot?”

204

Daniel Olariu

There is no doubt that the concise Hebrew question in v. 13b has posed problems in translation. Th-Dan follows the OG’s lead to exegetically add terms that enhance readability. In this case, the reviser borrows from OG-Dan the e­ xegetical additions στήσεται (which govern ἡ ὅρασις) and ἡ ἀρθεῖσα (which govern ἡ θυσία). No. 6. Dan 8:14 MT

LXX

Th

‫ ויאמר אלי עד ערב בקר‬καὶ εἶπεν αὐτῷ Ἕως ‫ אלפים ושלש מאות ונצדק‬ἑσπέρας καὶ πρωὶ ἡμέραι ‫ קדש׃‬δισχίλιαι τριακόσιαι, καὶ

καὶ εἶπεν αὐτῷ Ἕως ἑσπέρας καὶ πρωὶ ἡμέραι δισχίλιαι καὶ τριακόσιαι, καὶ καθαρισθήσεται τὸ ἅγιον. καθαρισθήσεται τὸ ἅγιον. He answered me, “For And he said to him, And he said to him, twenty-three hundred “Two thousand three “Two thousand and three evenings and mornings; hundred days, evenings hundred days, evenings and then the sanctuary shall be and mornings, and the mornings, and the sanctuary cleansed.” sanctuary will be purified.” will be purified.”

Most of the modern commentators interpret the 2300 “evenings and mornings” as referring to 1150 days.39 However, Th-Dan seemingly adopts from the OG the exegetical addition ἡμέραι “days,”40 suggesting that, similar to the OG and Vulgate, it has understood the time period as tantamount to 2300 days.41 No. 7. Dan 8:17 MT

LXX

Th

‫ ויבא אצל עמדי ]…[ כי לעת‬καὶ ἦλθε καὶ ἔστη ἐχόμενός καὶ ἦλθεν καὶ ἔστη ἐχόμενος ‫ קץ החזון׃‬μου τῆς στάσεως, […] ἔτι τῆς στάσεώς μου, […] ἔτι γὰρ εἰς ὥραν καιροῦ τοῦτο γὰρ εἰς καιροῦ πέρας ἡ τὸ ὅραμα. ὅρασις.

39  E.g., Segal, “Calculating the End: Inner-Danielic Chronological Developments;” Collins, Daniel, 336; Hartman and Di Lella, Daniel, 227; Montgomery, Daniel, 242–243; Lacocque, The Book of Daniel, 164; Newsom and Breed, Daniel, A Commentary, 267. 40  A further indication of Th-Dan’s dependence on the OG in this verse is the shared significant equivalent καθαρισθήσεται for ‫ונצדק‬. Cf. Olariu, The Quest for the Common Basis, 70. 41  For modern scholars that have understood the time period of 2300 as days in light of Genesis 1, see Goldingay, Daniel, 213; and Seow, Daniel, 125.

205

Recensional Additions

MT He came near to where I was standing, […] that the vision refers to the time of the end.”

LXX And he came and stood near where I stood, […] for this vision is yet for an appropriate time.”

Th And he came and stood near where I stood, […] for the vision is yet for the end of time.”

The incentive for the addition of καὶ ἔστη reflects most likely contextual exegesis, the OG’s translator being probably inspired by the Hebrew collocation ‫ ועמד‬+ ‫בוא‬ “to come and to stand” which frequently appear in MT.42 It can be further argued that the translator may have had been directly influenced by Ezek 9:2; 10:6.43 Indeed, both passages have a phraseology similar to the longer OG text, i.e. ‫‏ויבא‬ ‫ ויעמד אצל‬and ‫‏ויבאו ויעמדו אצל‬, respectively, and the LXX-Ezek’s rendering of each is very similar to OG-Dan 8:17, namely, καὶ εἰσήλθοσαν καὶ ἔστησαν ἐχόμενοι and καὶ εἰσῆλθεν καὶ ἔστη ἐχόμενος. The presence of ‫ עמדי‬in the proximity may have been a trigger for the OG translator to recall the Semitic collocation.44 Th-Dan has taken in the OG’s exegetical addition.

42  The habitual juxtaposition of these words mirrors the logical sequence of action, namely, that a person first comes and then stands in front of another entity. The collocation occurs in 1 Kgs 3:15; 2 Kgs 5:15, 25; 8:9; 18:17; Jer 7:10; Ezek 9:2; 10:6; Ruth 2:7; Dan 2:2; 8:17. Since there is no clear contextual indication to suspect that a scribal error has occurred (e.g., the omission because of parablepsis of the purported ‫)ויעמד‬, it is methodologically likely to assume that the addition of καὶ ἔστη reflects contextual harmonization and adaptation in view of the frequent Semitic collocation. 43  Another possibility is to postulate contextual harmonization with Dan 2:2. However, notwithstanding that the collocation ‫ ויבאו ויעמדו‏‬is attested in this verse, the OG renders differently (καὶ παραγενόμενοι ἔστησαν “and when they arrived, they stood”) than 8:17. In addition, the collocation in 2:2 lacks ‫אצל‬, making Ezek 9:2; 10:6 more likely source on whom the OG has depended. 44  In addition, the collocation ‫ ע ֶֹמד‬+ ‫ עמד‬appears twice in the immediate context (i.e., Dan 8:18; 10:11) and several times in the remote MT context (i.e., Neh 13:11; 2 Chr 30:16; 34:31; 35:10). Noting its presence, Charles uses this piece of information to suggest that both the OG and Th-Dan reflect the original Vorlage, the reading presupposed by καὶ ἔστη being “lost in MT.” Cf. Charles, Daniel, 196 (see also 214–215). However, he fails to provide a reason for its disappearance as well as to note that the OG’s addition which entered Th-Dan (and further in Vulgate) reflects translational harmonization.

206

Daniel Olariu

No. 8. Dan 8:22 MT

LXX

Th

‫ והנשברת ותעמדנה ארבע‬καὶ τὰ συντριβέντα καὶ ‫ תחתיה […]׃‬ἀναβάντα ὀπίσω αὐτοῦ One was broken and four came in its stead – that [means]: […].

καὶ τοῦ συντριβέντος, οὗ ἔστησαν τέσσαρα ὑποκάτω τέσσαρα κέρατα, […]. κέρατα, […]. And as for the four horns And as for the crushed one, that were crushed and came where four horns arose up after it: […].  underneath: […]. 

In the interpretation section of Daniel 8, v. 22 parallels v. 8 of the vision. The def. fem. sing. part. ‫ הנשברת‬along with the clause ‫ ותעמדנה ארבע תחתיה‬function as headings, aiming to resume features from the vision that are to be explained.45 OG-Dan has mistakenly taken both phrases as forming a single clause, wherein the fem. sg. part. was interpreted attributively as determining ‫ ארבע‬as subject. Consequently, the translator made all four horns as being crushed and not only the conspicuous horn of the he-goat. The imprecision was seemingly triggered by the different vocalization of the part. by the translator., i.e. ‫( ונשבר ֹת‬fem. pl.) vs. ‫ונשברת‬ ֶ (fem. sg.). The reviser corrected the imprecisions toward a literal representation of the MT but still showing dependence, copying the explicating addition κέρατα from OG.

45  As such, ‫ הנשברת‬resumes ‫ וכעצמו נשברה הקרן הגדולה‬while ‫ ותעמדנה ארבע תחתיה‬parallels ‫ ותעלנה חזות ארבע תחתיה לארבע רוחות השמים‬of v. 8. The use of headings is amply attested in the pesharim literature found at Qumran. See Collins, Daniel, 339. It is less likely that v. 22b reflects a casus pendence clause as suggested by Montgomery (Daniel, 353), and who was rightly criticized by Charles (Daniel, 1929, 216), or that that the Hebrew “would be unintelligible without recourse to the underlying Aramaic” as Hartman and Di Lella (Daniel, 228) emphatically claim.

207

Recensional Additions

No. 9. Dan 12:7 MT

LXX

Th

‫ ]…[ […] וירם ימינו ושמאלו אל‬καὶ ὕψωσε τὴν δεξιὰν […] καὶ ὕψωσεν τὴν δεξιὰν ‫ השמים וישבע בחי העולם כי‬καὶ τὴν ἀριστερὰν εἰς τὸν αὐτοῦ καὶ τὴν ἀριστερὰν ‫ למועד מועדים וחצי […]׃‬οὐρανὸν καὶ ὤμοσε τὸν αὐτοῦ εἰς τὸν οὐρανὸν καὶ

ζῶντα εἰς τὸν αἰῶνα θεὸν ὅτι εἰς καιρὸν καὶ καιροὺς καὶ ἥμισυ καιροῦ […]. […] as he lifted his right […] And he raised hand and his left hand to the right hand and left heaven: “For a time, times, hand toward heaven, and half a time […]. and he swore by God, who lives forever, [“The consummation] … will be at a time and times and half a time […]. 

ὤμοσεν ἐν τῷ ζῶντι τὸν αἰῶνα ὅτι Εἰς καιρὸν καιρῶν καὶ ἥμισυ καιροῦ […]. […] and he raised his right hand and his left hand toward heaven. And he swore by means of the one who lives forever: [“Pertaining] to a time of times and half a time; […]. 

Th-Dan follows the OG in adding καιροῦ which is implied by the Hebrew idiomatic phrase ‫למועד מועדים וחצי‬.

3.2  Revision of Contextual Additions in OG-Dan

OG-Dan features numerous exegetical additions, omissions, and substitutions, which were meant to exegetically clarify tensions in the Semitic Vorlage or overtly utter the religious beliefs of the translator. Contrastingly, Th-Dan features the systematic tendency to eliminate such elements in conformity to MT.

3.2.1  Linguistic Elements The OG translator has added elements reflecting a different linguistic interpretation of the MT. That is to say, such deviations suggest his problematic decoding of the grammar of the Semitic language, either at the level of syntax, morphology or semantics.

208

Daniel Olariu

No. 10. Dan 7:23 MT

LXX

Th

‫ כן אמר חיותא רביעיתא‬καὶ ἐρρέθη μοι περὶ τοῦ καὶ εἶπεν Τὸ θηρίον τὸ ]‫ מלכו רביעיא [רביע][אה‬τετάρτου θηρίου, ὅτι τέταρτον, βασιλεία τετάρτη ‫ תהוא בארעא […]׃‬βασιλεία τετάρτη ἔσται ἐπὶ ἔσται ἐν τῇ γῇ […]. This is what he said: ‘The fourth beast [means] – there will be a fourth kingdom upon the earth […].

τῆς γῆς […] […] And it was said to me concerning the fourth beast: “There shall be a fourth kingdom upon the earth.

And he said: “As for the fourth beast, there shall be a fourth kingdom on the earth, […].

The OG translator interpreted ‫ כן אמר חיותא רביעיתא‬as forming a verbal clause. Consequently, the formula ‫כּן ֲא ַמר‬,ֵ which technically introduces the quote of the angelus interpretum, was transformed as introducing indirect speech. Furthermore, the translator has changed the verbal aspect (act. → pass.), then added the explicating μοι περὶ, and finally the conjunction ὅτι, which connects the first clause to the reminder.46 Th-Dan eliminates the additions and make the necessary changes so that the OG conforms to MT. No. 11. Dan 11:8 MT

LXX

Th

‫ וגם ֱאלהיהם עם נסכיהם עם‬καὶ τοὺς θεοὺς αὐτῶν ‫ כלי חמדתם כסף וזהב בשבי‬καταστρέψει μετὰ τῶν ‫ יבא מצרים […]׃‬χωνευτῶν αὐτῶν καὶ τοὺς

καί γε τοὺς θεοὺς αὐτῶν μετὰ τῶν χωνευτῶν αὐτῶν, πᾶν σκεῦος ἐπιθυμητὸν ὄχλους αὐτῶν μετὰ τῶν αὐτῶν ἀργυρίου καὶ σκευῶν τῶν ἐπιθυμημάτων χρυσίου, μετὰ αἰχμαλωσίας αὐτῶν, τὸ ἀργύριον καὶ τὸ οἴσει εἰς Αἴγυπτον· […]. χρυσίον, ἐν αἰχμαλωσίᾳ ἀποίσουσιν εἰς Αἴγυπτον· […].

46  For the idiosyncrasy of OG-Dan to transform active verbal forms into passive, see further Amara, “Old Greek Version of Daniel,” 57–70.

209

Recensional Additions

MT He will also take their gods with their molten images and their precious vessels of silver and gold back to Egypt.

LXX And he will overthrow their gods with their cast images and their crowds with their precious vessels. They will carry off the silver and gold to Egypt […].

Th And indeed their gods, with their cast images, all of their precious vessels of silver and gold, he will carry to Egypt […].

OG-Dan has divided the long Hebrew clause in v. 8 into two, adding καταστρέφω “to overthrow” as an explicating verb in the first clause. However, καὶ τοὺς ὄχλους αὐτῶν was probably inserted to fill in the list which completely overlooked “people” as affected by the devastating attack of the king of the south (cf. vv. 7–8).

3.2.2  Exegetical Elements The OG translator has added elements which were influenced from his interpretation of the immediate context. The additions were meant primarily to clarify the meaning of the source text or solve its presumed exegetical tensions. The reviser eliminates most of these. No. 12. Dan 1:6 MT

LXX

‫ ויהי בהם מבני יהודה דניאל‬καὶ ἦσαν ἐκ τοῦ γένους ‫ חנניה מישאל ועזריה׃‬τῶν υἱῶν Ισραηλ τῶν

Among them were the Judahites Daniel, Hananiah, Mishael and Azariah.

ἀπὸ τῆς Ιουδαίας Δανιηλ, Ανανιας, Μισαηλ, Αζαριας. And there were of the race of the sons of Israel who were from Judea: Daniel, Hananias, Misael, Azarias.

Th καὶ ἐγένετο ἐν αὐτοῖς ἐκ τῶν υἱῶν Ιουδα Δανιηλ καὶ Ανανιας καὶ Μισαηλ καὶ Αζαριας. And among them from the sons of Iouda were Daniel and Hananias and Misael and Azarias.

The OG adds words presumably to clarify the precise ethnicity of the main characters of the story. The translator probably believed that a literal translation of the phrase ‫ מבני יהודה‬would confuse the readers, creating an exegetical tension with

210

Daniel Olariu

v. 3 which informs that the youth are ‫“ מבני ישראל ומזרע המלוכה‬Israelites of royal descent.” Consequently, the translator elaborates that they are the descendants of their forefather Israel, from the “province of Judah” (cf. Hag 1:1, 14; 2:2, 21; Neh 5:14), instead of the “tribe of Judah.”47 No. 13. Dan 9:23 MT

LXX

‫ בתחלת תחנוניך יצא דבר‬ἐν ἀρχῇ τῆς δεήσεώς σου

‫ […]׃‬ἐξῆλθε πρόσταγμα παρὰ κυρίου, […]. A word went forth as you At the beginning of your began your plea, […]. supplication an ordinance went out from the Lord, […].

Th ἐν ἀρχῇ τῆς δεήσεώς σου ἐξῆλθεν λόγος, […]. At the beginning of your supplication a word went out, […].

The addition παρὰ κυρίου seemingly has the exegetical function to distinguish between two types of ordinances within the chapter: the “ordinance” which went out straight “from the Lord” and mediated by the angel Gabriel (v. 23), and “the ordinance of the Lord [that] came to the prophet Ieremias” (v. 2).

3.2.3  Theological Elements The OG translator introduces elements that express his theological worldview. The ideas primarily regard the exclusiveness of God of Israel, monotheism, angels, etc. The reviser eliminates all of these elements.

47  The phrase ‫ בני יהודה‬occurs fifty-four times and it mainly refers to the descendants from the tribe of Judah. This may have probably made Collins (Daniel, 140) to note that ‫מבני יהודה‬ signify “from the tribe of Judah, because they have already been characterized as Israelites.” However, it appears that none of the modern interpreters have noted the exegetical tension that the OG translator attempted to solve in v. 6.

211

Recensional Additions

No. 14. Dan 2:47 MT

LXX

Th

‫ ]…[ ]…[וגלה רזין די יכלת‬ὁ ἐκφαίνων μυστήρια […] καὶ ἀποκαλύπτων ‫ למגלא רזה דנה׃‬κρυπτὰ μόνος, ὅτι μυστήρια, ὅτι ἠδυνήθης […] and the revealer of mysteries to have enabled you to reveal this mystery.

ἐδυνάσθης δηλῶσαι τὸ μυστήριον τοῦτο. […] who alone brings to light hidden mysteries, because you have been able to disclose this mystery!

ἀποκαλύψαι τὸ μυστήριον τοῦτο. […] and a revealer of mysteries, because you have been able to reveal this mystery!

As tempting as may be, the reviser eliminates the OG inserted words which theologically emphasizes the exclusiveness of God of Israel in untangling impenetrable secrets.48 The words belong to Nebuchadnezzar’s confession which acknowledge the God of Daniel who can reveal secrets. No. 15. Dan 3:17 MT

LXX

Th

‫ הן איתי אלהנא די אנחנא‬ἔστι γὰρ ὁ θεὸς ὁ ἐν ἔστιν γὰρ θεός, ᾧ ἡμεῖς ‫ פלחין […]׃‬οὐρανοῖς εἷς κύριος ἡμῶν, λατρεύομεν […]. for if so it must be, our God whom we serve […].

ὃν φοβούμεθα […]. for there is God who is in heaven, our one Lord, whom we fear […]

for there is a god whom we serve […]

The OG translator reworks the response of the three youth into a monotheistic confession. The phrase ἐν οὐρανοῖς appears to be influenced by Daniel’s confession in the book in 2:28, whereas εἷς κύριος ἡμῶν recalls Deut 6:4. The reviser has rejected the exegetical plus.

48  See Amara, Old Greek Version of Daniel, 178.

212

Daniel Olariu

3.3  Unique Contextual Additions in Th-Dan

The examples below share the common vantage point of being unique contextual additions in Th-Dan. These additions reflect reviser’s touches which are different from the OG. However, as it shall be seen below, the OG’s influence on Th-Dan’s additions is still detected in some instances. Th-Dan’s plusses are marked as single underlined text in the right column. In the event that a certain addition in the OG has inspired the reviser to use a similar addition, the OG’s plus would be also marked as a single, underlined text. In such instances, wherein the OG’s pluses may have influenced Th-Dan, they differ in the words added. Consequently, the OG’s influence is likely at the level of translation technique and not at the level of lexical choices.

3.3.1 Linguistic Elements Th-Dan introduces sometimes unique linguistic additions over against MT generated either by the nature of the source text or by the requirements of the target language. In the former case, the problematic nature of the Semitic text has caused the reviser to add certain words contextually. In the latter case, the requirements of the Greek language have determined the addition of certain elements, which do not alter the source-text’s meaning, e.g., stylistic addition of helping verbs, personal pronouns, prepositions, connective particle καί, etc. Since such additions are expected and should be properly described as variants/non-variants,49 we would limit to mention them later in passing. The majority of the cases below exemplify how the reviser has added elements in order to circumvent linguistic issues raised by the nature of the source text. No. 16. Dan 2:15 MT

LXX

Th

‫ ]…[ […] על מה דתא מהחצפה‬Περὶ τίνος […] περὶ τίνος ἐξῆλθεν ‫ מן קדם מלכא […]׃‬δογματίζεται πικρῶς παρὰ ἡ γνώμη ἡ ἀναιδὴς ἐκ τοῦ βασιλέως; […]

[…] Why is the decree of the king so urgent?” […]

[…] “Why is it being decreed so harshly by the king?”

προσώπου τοῦ βασιλέως; […] […] why has the shameless opinion gone out from before the king?”

49  See Tov, The Text-Critical Use of the Septuagint in Biblical Research, 169–178.

213

Recensional Additions

In the interrogative question above, the OG has freely changed the grammatical category from noun to verb, i.e. 0‫“ דת־‬law” – δογματίζω “to ordain,” and the participle derived from 0‫“ חצף־‬to be harsh,” “to be urgent” was taken adverbially. Th-Dan has corrected the grammatical category while, instead of the expected εἰμί “to be” in the nominal clause, it has employed ἐξέρχομαι “to come out,” “to go out.” His decision was demonstrably influenced by the idiomatic phrase ‫דתא‬ ‫ נפקת‬in the immediate context (v. 13). The OG renders the phrase in v. 13 with ἐδογματίσθη “it was decreed,” whereas Th-Dan has τὸ δόγμα ἐξῆλθεν “the decree went out.” Consequently, each version maintains the same equivalents in vv. 13, 15, a translation technique known as textual levelling. No. 17. Dan 2:38 MT

LXX

Th

‫ ]…[ […]‏יהב בידך והשלטך‬παρέδωκεν εἰς τὰς ‫ בכלהון […]׃‬χεῖράς σου κυριεύειν

[…] ἔδωκεν ἐν τῇ χειρί σου καὶ κατέστησέν σε κύριον πάντων, […] πάντων, […]. […] into whose hands He […] he has delivered them […] he has given into your has given … and to whom into your hands to rule hand, and he has established He has given dominion over all […]. you as lord of all […]. over them all […]

The OG departs from a literal representation of the root ‫ ׁשלט‬in causative haphel stem, i.e. ‫“ השלטך‬he made you ruler.”50 The reviser was quick to notice the deviation and aimed at a literal translation. However, given the limitation of the Greek language in finding a word-for-word equivalent, he employed two-for-one-word technique. Th-Dan’s dependence on the OG emerges from the use of the noun κύριον that was likely inspired by the OG’s equivalent κυριεύω “to rule over.”51

50  HALOT 5:1995. 51  There are other hints that indicate Th-Dan consultation of the OG: (1) Th-Dan has followed the OG in not representing both the preposition and the pronominal suffix of the immediate phrase ‫“ בכלהון‬over them all; (2) Th-Dan’s use of the equivalent καθίστημι “to put,” “to put in charge,” “to make” in v. 38 was seemingly inspired from v. 48 were the OG uses it for ‫ׁשלט‬ haphel. Th-Dan keeps the same equivalent in v. 48 as well. However, like the OG, it does not add κύριος “lord” as in v. 38, since the word was reserved by the reviser only for God and in reference to the king.

214

Daniel Olariu

No. 18. Dan 4:5[8] MT

LXX

‫ […] ודי רוח אלהין קדישין‬LXX=0 ‫בה […]׃‬

Th […] ὃς πνεῦμα θεοῦ ἅγιον ἐν ἑαυτῷ ἔχει […]. […] who has a holy, divine spirit in himself […].

[…] in whom the spirit of the holy gods was […].

For rendering the nominal, subordinate clause above, Th-Dan adds the helping verb ἔχω “to have” and not εἰμί “to be,” likely because it renders in the target language more adequately the meaning of the source text. In the other two instances where the clause appears in the same chapter, Th-Dan renders literally, making no use of helping verbs: ὅτι πνεῦμα θεοῦ ἅγιον ἐν σοὶ (vv. 6[9], 15[18]. In both instances the OG reflects a different Vorlage. No. 19. Dan 4:24[27] MT

LXX

Th

‫ ]…[ […] הן תהוא ארכה‬ἵνα ἐπιείκεια δοθῇ σοι […] ἴσως ἔσται μακρόθυμος ‫]…[ לשלותך׃‬. τοῖς παραπτώμασίν σου ὁ […] then your serenity may be extended.”

[…] so that equity might be given to you […].

θεός. […] Perhaps God will show forbearance for your transgressions.”

Th-Dan adds θεός due to his problematic understanding of the Aramaic clause. Demonstrably, the reviser struggled to decode the meaning of 0‫“ שלוה־‬prosperity.” As a result, he has contextually interpreted the clause as referring to “God” who might show restraint for the king’s transgressions. To do so, he has inter-

215

Recensional Additions

preted 0‫“ ארכה־‬extension” in light of ‫“ ארך אפים‬indulgent”52 while 0‫ שלוה־‬was contextually guessed.53

3.3.2  Exegetical Elements No. 20. Dan 1:3 MT

‫ויאמר המלך לאשפנז‬ ‫רב סריסיו להביא מבני‬ ‫ישראל ומזרע המלוכה ומן‬ ‫הפרתמים׃‬

LXX

αὶ εἶπεν ὁ βασιλεὺς Αβιεσδρι τῷ ἑαυτοῦ ἀρχιευνούχῳ ἀγαγεῖν αὐτῷ ἐκ τῶν υἱῶν τῶν μεγιστάνων τοῦ Ισραηλ καὶ ἐκ τοῦ βασιλικοῦ γένους καὶ ἐκ τῶν ἐπιλέκτων Then the king ordered And the king told Abiesdri, Ashpenaz, his chief officer, his own chief eunuch, to to bring some Israelites of bring to him some of the royal descent and of the sons of the nobles of Israel nobility. and of royal descent and of the aristocracy:

Th καὶ εἶπεν ὁ βασιλεὺς τῷ Ασφανεζ τῷ ἀρχιευνούχῳ αὐτοῦ εἰσαγαγεῖν ἀπὸ τῶν υἱῶν τῆς αἰχμαλωσίας Ισραηλ καὶ ἀπὸ τοῦ σπέρματος τῆς βασιλείας καὶ ἀπὸ τῶν φορθομμιν And the king told Asphanez, his chief eunuch, to bring in some of the sons of the captivity of Israel, both of the seed of the kingdom and of the Phorthommin:

In contrast to the MT reading ‫מבני ישראל‬, both Greek versions reflect the addition of an extra element between the nouns of the construct chain, namely, τῶν μεγιστάνων in the OG and τῆς αἰχμαλωσίας in Th-Dan. The plus can be interpreted as an exegetical explanatory addition. Given the fact that v. 2 refers to the deportation of only the king, the translators needed to add precision and clarify the target group intended by the king’s command. While OG perhaps freely added a word based on the context of the verse, Th-Dan harmonized the 52  Μακρόθυμος “patient,” “long-suffering,” “enduring” appears about nineteen times in LXX literature. It renders ‫ ארך אפים‬thirteen times (Exod 34:6; Num 14:18; Neh 9:17; Pss 86[85]:15; 103[102]:8; 145[144]:8; Prov 14:29; 15:18; 16:32; Sir 5:4; Joel 2:13; Jonah 4:2; Nah 1:3); once ‫( יקר רוח‬Prov 17:27); and once ‫( ארך רוח‬Eccl 7:8). MT=0: Ps 7:12; Odes 12:7; Sir 1:23; Wis 15:1. 53  The Hebrew cognate ‫ ַשׁ ְלוָ ה‬occurs five times and its meaning has been deduced erroneously thrice (Jer 22:21; Prov 1:32; 17:1). In the other two instances LXX renders it with εὐθηνία (Ezek 16:49; Ps 122[121]:7). In Daniel, ‫ ַשׁ ְלוָ ה‬occurs thrice: in 8:25, Th-Dan adopts the erroneous equivalent δόλος “deceit” from the OG; in 11:21, 24, because he singled out an accurate equivalent, Th-Dan corrects the contextual OG guess ἐξάπινα “suddenly” with the precise εὐθηνία “prosperity.”

216

Daniel Olariu

expression with 2:25,54 identifying the target group as the “sons of the captivity of Israel.” No. 21. Dan 6:21[20] MT

LXX

‫…היכל‬ ‫ ]…[ […] חיא אלהך‬καὶ ὁ θεός σου … ‫ לשיזבותך מן אריותא׃‬σέσωκέ σε ἀπὸ τῶν

λεόντων, καὶ οὐκ ἠχρείωκάν σε; […] was God … able to […] and has your God … deliver you from the lions? saved you from the lions, and have they not injured you?”

Th […] ὁ θεός σου … εἰ ἠδυνήθη ἐξελέσθαι σε ἐκ στόματος τῶν λεόντων; […] […] has your God … been able to deliver you from the mouth of the lions?” […]

The addition of ἐκ στόματος constitutes most likely a contextual harmonization with v. 23[22]. However, it is perhaps impossible to determine whether the harmonization occurred in Th-Dan’s Vorlage or during the revising process. No. 22. Dan 7:17 MT

LXX

Th

‫ אלין חיותא רברבתא די אנין‬Ταῦτα τὰ θηρία τὰ μεγάλα Ταῦτα τὰ θηρία [τὰ μεγάλα]Ra ‫ ארבע ארבעה מלכין יקומון‬εἰσὶ τέσσαρες βασιλεῖαι, αἳ τὰ τέσσαρα, τέσσαρες ‫ מן ארעא׃‬ἀπολοῦνται ἀπὸ τῆς γῆς· βασιλεῖαι ἀναστήσονται ἐπὶ ‘These great beasts, four in number [mean] four kingdoms will arise out of the earth;

τῆς γῆς, αἳ ἀρθήσονται· “These great beasts are “These four beasts: four four kingdoms, which shall kingdoms will arise upon perish from the earth. the earth, which will be destroyed

54  The Semitic phrase ‫ מן בני גלותא די יהוד‬was rendered similarly in both versions: ἐκ τῶν υἱῶν τῆς αἰχμαλωσίας τῆς Ιουδαίας (Th-Dan); ἐκ τῆς αἰχμαλωσίας τῶν υἱῶν τῆς Ιουδαίας (OG-Dan).

217

Recensional Additions

The addition of αἳ ἀρθήσονται in Th-Dan is contextually added, likely inspired from the OG’s exegesis. Vv. 17–18 contain the first brief explanation of the vision recorded in vv. 2–14. Consequently, the condensed interpretation did not cover all elements of the vision. The first to take issue with the sequence of events in vv. 17–18 was the OG’s translator. In his view, the “holy ones of the Most High” receive the kingdom only after the four kingdoms are destroyed. This is most likely the reason why he renders ‫ יקומון‬with αἳ ἀπολοῦνται “which shall perish.” The reviser corrects αἳ ἀπολοῦνται with the literal ἀναστήσονται, whereas under the influence of the OG, he has demonstrably added αἳ ἀρθήσονται55 with the same effect: to exegetically clarify that the event of receiving the kingdom is preceded by the destruction of the earthly kingdoms.56 No. 23. Dan 12:1 MT

LXX

Th

‫ ]…[ […] והיתה עת צרה אשר‬ἐκείνη ἡ ἡμέρα […] καὶ ἔσται καιρὸς ‫ לא נהיתה מהיות גוי עד‬θλίψεως, οἵα οὐκ ἐγενήθη θλίψεως, οἵα οὐ γέγονεν ἀφ᾿ ‫ העת ההיא […]׃‬ἀφ᾿ οὗ ἐγενήθησαν ἕως τῆς οὗ γεγένηται ἔθνος ἐπὶ τῆς ἡμέρας ἐκείνης· […]

[…] It will be a time of trouble, the like of which has never been since the nation came into being. […]

[…] That is a day of affliction, which will be such as has not occurred since they were born until that day […].

γῆς ἕως τοῦ καιροῦ ἐκείνου· […] […] And there will be a time of affliction such as had not occurred since a nation first came into existence until that time. […]

The addition of ἐπὶ τῆς γῆς in Th-Dan 12:1 was probably introduced to spell out a subtle exegetical nuance lost in the OG’s translation but implied by the source text: while the OG suggests that the intensity of the anticipated time of affliction would not find a correspondent in the time of one single generation witnessing 55  The equivalent αἴρω “to take up” occurs five times in Th-Dan. Twice it renders ‫“ נשא‬to lift,” “to carry,” “to take” in the idiomatic phrase ‫“ ואשא (את) עיני‬and I raised my eyes.” The reviser rejects ἀπόλλυμι “to destroy,” “to ruin” for αἴρω in Dan 7:17, presumably because he has restricted the former only for 0‫אבד־‬. It is significant to note that in the other two occurrences αἴρω was exegetically employed by OG-Dan and further taken in by the reviser in Dan 8:13 and 9:27. 56  Th-Dan has also followed OG-Dan in rendering ‫“ מלכין‬kings” with βασιλεῖαι “kingdoms,” exegetically influenced by the context of vv. 23, 27. The section of vv. 23–27 represents the second, more elaborate interpretation of the vision.

218

Daniel Olariu

the events, Th-Dan explains that the afflictions cannot be compared with any past events in the history of nations.

4.  Interpretation of the Data The aim of this study was to investigate recensional exegesis, specifically focusing on contextual additions.57 As it was intimated above, not only that a study of this kind is lacking within the framework of recensional studies, but also the two complete, parallel versions of the book of Daniel provide the ideal context to pursue such a quest. However, the task is complex since we have to assess the reviser’s attitudes toward both the source text (MT) and the base text (OG-Dan). It is the thesis of this study that only by considering each of these aspects we can assess recensional exegesis in Th-Dan. The analysis of contextual additions in the Greek versions of Daniel supports the theory of a translation-revision relationship between the two texts. The standard methodology for testing the character of a given text as a revision posits two criteria: (1) the confirmation of a common basis between the two texts by means of shared significant lexical choices; (2) the confirmation of recensional tendencies in the alleged revised text.58 The close investigation of contextual additions affirms both criteria. The common basis between OG-Dan and Th-Dan is demonstrated by the fact that the latter has adopted a high number of OG lexical choices which were added in order to compensate the MT’s concise, elliptical, and sometimes problematic language (Nos. 1–9). It was contended that such borrowings are tantamount to significant contextual additions shared between OG-Dan and Th-Dan. Several significant additions pile up in Daniel 8 (Nos. 5–8), evidencing the common basis in this chapter. These results are particularly relevant considering that previous studies have disputed the existence of a shared significant equivalents in this chapter.59 It was further contended that the influence of the OG on Th-Dan did not show up only when the problematic nature of the source text gave rise to contextual additions (No. 17). Th-Dan has been sometimes influenced by the OG’s translation style. The shared particle δέ as a plus typifies this tendency (No. 1). Not only that the particle is cherished by the OG and, thus, typifies its translational style, but also its use in Th-Dan convincingly points to the same assessment. In addition, there are certain cases in which the reviser adds unique elements following the 57  This study will be followed by two other investigations that will explore the reviser’s attitude toward exegetical minuses and exegetical substitutions. 58 Tov, Textual Criticism of the Hebrew Bible, 141; Tov, The Septuagint Translation of Jeremiah and Baruch, 43. 59  Cf. McLay, The OG and Th Versions of Daniel, 153–174.

Recensional Additions

219

OG’s pattern (Nos. 20, 22).60 That is to say the lexical choices he uses differs from the OG’s equivalents but the latter’s influence on the former can still be discerned. The study further confirms the second criterion of the methodology that tests the quality of a text as a revision. Section 3.2 documents with a select number of cases the reviser’s tendency to eliminate contextual additions from his base text (Nos. 10–15). He aimed literalism, reworking the OG to correspond word-forword to his Semitic Vorlage. The number of such instances is far bigger than the constrains of this study would allow to reproduce. The reviser had in many cases a better linguistic knowledge, successfully correcting the OG’s linguistic deviations toward Proto-MT (Nos. 10–11). He has also rejected the OG’s exegetical additions which were introduced to remove interpretative tensions or to spell out much more clearly the alleged intent of the source text (Nos. 12–13). The OG’s theological additions have received a similar treatment (Nos. 14–15)). The reviser was not persuaded to integrate OG additions related to monotheism, angelology, God’s temple, or God’s people.61 For him, these would simply not justify his programmatic agenda to produce a literal revision. The rejection of such elements corroborated with the absence in Th-Dan of any theological additions62 become even more revealing once there is a high degree of probability that both the OG translator and Th-Dan’s reviser shared the same religious, Jewish background.63 The unique additions embedded by the reviser in his text are minimal in quantity and insignificant in content as compared to the OG’s. Foremost, they reflect short plusses,64 which can be divided into two groups: (1) unique additions which are best described as variants/non-variants, e.g., article, pronouns, preposition, etc.; (2) unique additions which reflect on the style of the reviser. Of the two groups, the analysis has focused on the latter, the elements of the first group being expected to appear in both literal and free translational units. The examples discussed typifying group two indicate that the reviser had his own linguistic limitations as well. Sometimes he introduced elements in order to produce an acceptable rendering when he came across with difficult language in the source 60  A similar case can be made for the additions recorded in Dan 3:28[95] (εἰς ἐμπυρισμόν OG/εἰς πῦρ Th-Dan); 10:13 (τοῦ στρατηγοῦ OG/τοῦ ἄρχοντος Th-Dan); and possibly in 7:7 ([διαφόρως] χρώμενον OG/[διάφορον] περισσῶς Th-Dan). 61  Many of the Amara’s cases that she brings to substantiate the OG’s interest in theological exegesis have been reworked by the reviser to correspond word-for-word to MT. Cf. Amara, The Old Greek Version of Daniel, 154–190. 62  Cf. sections 3.1.3; 3.3.3. 63  For relevant summaries regarding the issues surrounding the background of Th-Dan text, see Marcos, The Septuagint in Context: Introduction to the Greek Versions of the Bible, 142–154; Dines, The Septuagint, 84–87; Jellicoe, The Septuagint and Modern Study, 83–93. 64  It should be also stated that the reviser’s tendency toward literalism is reflected in the absence of long additions in Th-Dan. We have identified three such additions in 3:24[91], 30[97] and 12:13. While the latter addition is equally shared by both the OG and Th-Dan, the former ones are unique and reflect variant readings. They will be discussed in a future study.

220

Daniel Olariu

text (No. 19). Other unique linguistic additions pertain to the requirements of the Greek language. In such cases, the reviser made use of helping verbs and words, making explicit what was implicitly embedded in the syntax or the idiomatic expressions of the source text (Nos. 16–18). These types of linguistic additions are expected even in a revision. On the one side, some examples remind us that (literal) revisers are humans too and can show linguistic limitation. On the other hand, the other linguistic additions inform us that the reviser’s literalness is circumscribed to the grammar and rules of the target language. Significantly, he did both with minimal additions. Th-Dan’s unique exegetical additions are insignificant in content, likely reflecting harmonizing elements (Nos. 20–21). In addition, this study warrants the following intriguing conclusion: we can learn more about the reviser’s exegesis from the added elements he accepts from the OG than of the elements he imbeds. Consequently, the adoption of κακοῖς from the OG in 7:24 may spell his desire to underscore the wickedness of the little horn (No. 4). Further support for that may be drawn from the unique addition of πάντας in Th-Dan, with the effect of creating a stronger contrast in order to vilify even more the horn. Likewise, through the addition of αἳ ἀρθήσονται in 7:17 which was likely influenced by the OG transpires the exegetical desire to clarify that the event of receiving the kingdom by the saints is preceded by the destruction of the earthly kingdoms (No. 22). Furthermore, the acceptance of ἡμέραι in 8:14 seemingly signals the reviser’s support for the OG’s interpretation of the “evenings and mornings” as signifying 2300 full days (No. 6). Lastly, Th-Dan adds τῆς αἰχμαλωσίας in 1:3 while the OG displays τῶν μεγιστάνων. Whereas the latter addition restricts the identification of ‫ מבני ישראל‬to “the sons of the nobles of Israel,” the reviser finds exegetically appropriate to explicate to the readers that the recruitment was done among the “sons of the captivity of Israel” (No. 20).

5. Conclusions This study has investigated contextual additions in Theodotion’s text of Daniel within the framework of a translation-revision theory. The data indicate that ThDan’s reviser had limited interest in exegesis as compared to the OG’s translator. Theological additions are completely absent from his text. Most of his additions are linguistic in nature that are either required by the Greek grammar or seldom motivated by the reviser’s lack of understanding of the source text. As far as the exegetical additions are concerned, more can be learned from the OG’s elements that the reviser accepts than from the unique elements he embeds. These observations corroborated with the high number of significant additions shared between the OG and Th-Dan further affirm the character of Th-Da as a literal

Recensional Additions

221

revision. This, in turn, presumably provides the best explanation for the minimal number of unique exegetical additions in Th-Dan.

6. Bibliography Albertz, Rainer, Der Gott des Daniel: Untersuchungen zu Daniel 4–6 in der Septuagintafassung sowie zu Komposition und Theologie des aramäischen Danielbuches. SBS 131 (Stuttgart: Katholisches Bibelwerk, 1988). Amara, Dalia, The Old Greek Version of Daniel: The Translation, the Vorlage and the Redaction/ ‫ המצע והעריכה‬,‫ התרגום‬:‫( תרגום השבעים לספר דניאל‬PhD diss., Ben-Gurion University of the Negev, 2006). —, “Bel and the Dragon: The Relationship Between Theodotion and the Old Greek,” in From Author to Copyist: Essays on the Composition, Redaction, and Transmission of the Hebrew Bible in Honor of Zipi Talshir, ed. Cana Werman (Winona Lake, IN: Eisenbrauns, 2015), 125–147. Ashley, Timothy R., The Book of Daniel Chapters I–VI Text, Versions and Problems of Exegesis (PhD diss., University of St. Andrews, 1975). Barthélemy, Dominique, Les devanciers d’Aquila: Premiére publication intégrale du texte des fragments du Dodécaprophéton. VTSup 10 (Leiden: Brill, 1963). —, Critique textuelle de l’Ancien Testament. Tome 3: Ézéchiel, Daniel et les 12 Prophètes. OBO 50/3 (Göttingen: Vandenhoeck & Ruprecht, 1992). Bledsoe, Amanda M. Davis, “The Relationship of the Different Editions of Daniel: A History of Scholarship.” CBR 13.2 (2015): 175–190. Bludau, August, Die Alexandrinische Übersetzung des Buches Daniel und ihr Verhältnis zum massorethischen Text. Biblische Studien II 2/3 (Freiburg: Herder, 1897). Braasch, Birte, Die LXX-Übersetzung des Danielbuches – eine Orientierungshilfe für das religiöse und politisch-gesellschaftliche Leben in der ptolemäischen Diaspora: Eine rezeptionsgeschichtliche Untersuchung von Dan 1–7 (PhD diss., Hamburg University, 2003). Bruce, F. F., “The Earliest Old Testament Interpretation.” OtSt 17 (1972): 37–52. —, “Prophetic Interpretation in the Septuagint,” BIOSCS 12 (1979): 17–26. —, “The Oldest Greek Version of Daniel,” OtSt 20 (1977): 22–40. Ceriani, Antonio Maria, Codex syro-hexaplaris Ambrosianus photolithographice editus. Monumenta sacra et profana 7 (Milan: Bibliothecae Ambrosianae, 1874). Charles, Robert Henry, The Book of Daniel. The Century Bible (Edinburgh: T.C. and E.C. Jack, 1921). —, A Critical and Exegetical Commentary on the Book of Daniel (Oxford: Clarendon, 1929). Collins, John J., Daniel. Hermeneia. A Critical and Historical Commentary on the Bible (Minneapolis, MN: Fortress Press, 1993). Delcor, Mathias, Le Livre De Daniel. SB (Paris: Gabalda, 1971).

222

Daniel Olariu

Di Lella, Alexander A., “The Textual History of Septuagint-Daniel and Theodotion-Daniel,” in The Book of Daniel: Composition and Reception, eds. John J. Collins and Peter W. Flint. VTSup 83.2 (Leiden: Brill, 2001), 586–607. Dines, Jennifer M., The Septuagint (London: T&T Clark, 2004). Driver, Samuel Rolles, The Book of Daniel: With Introduction and Notes. The Cambridge Bible for Schools and Colleges (Cambridge: University Press, 1900). Geißen, A., Daniel 5–12; Susanna, Bel et Draco; Esther: Der Septuaginta-Text des Buches Daniel, Kap. 5–12, zusammen mit Susanna, Bel et Draco, sowie Esther Kap. 1,1a–2,15 nach dem Kölner Teil des Papyrus 967. Papyrologische Texte und Abhandlungen 5 (Bonn: Rudolf Habelt, 1968). Goldingay, John E., Daniel. WBC 30 (Dallas, Texas: Word Books, 1989). Hamm, W., Daniel 1–2: Der Septuaginta-Text des Buches Daniel, Kap. 1–2, nach dem Kölner Teil des Papyrus 967. Papyrologische Texte und Abhandlungen 10 (Bonn: Rudolf Habelt, 1969). —, Daniel 3–4: Der Septuaginta-Text des Buches Daniel, Kap. 3–4, nach dem Kölner Teil des Papyrus 967. Papyrologische Texte und Abhandlungen 21 (Bonn: Rudolf Habelt, 1977). Hartman, Louis F., and Alexander A. Di Lella, The Book of Daniel. AB 23 (Garden City, NY: Doubleday, 1978). Jahn, Gustav, Das Buch Daniel nach der Septuaginta hergestellt, übersetzt und kritisch erklärt (Leipzig: Pfeiffer, 1904). Jeansonne, Pace Sharon, The Old Greek Translation of Daniel 7–12. CBQMS 19 (Washington, DC: The Catholic Biblical Association of America, 1988). Jellicoe, Sidney T., Septuagint and Modern Study (Winona Lake, IN: Eisenbrauns, 1989). JPS Hebrew–English Tanach: The Traditional Hebrew Text and the New JPS Translation, 2nd ed. (Philadelphia: The Jewish Publication Society, 1999). Kenyon, Frederic G, Daniel 3,72–6,18: The Chester Beatty Biblical Papyri, Description and Texts of Twelve Manuscripts on Papyrus of the Greek Bible, Fasc. VII Ezekiel, Daniel, Esther (Plates). 2 vols. (London: E. Walker, 1938). Lacocque, André, ed. The Book of Daniel. Translated by David Pellauer (Atlanta, GA: John Knox, 1979). Louw, Johannes P., and Eugene A. Nida, eds. Greek-English Lexicon of the New Testament Based on Semantic Domains, 2nd ed. (New York, NY: United Bible Societies, 1989). Magistris, Simon de, Daniel secundum Septuaginta ex tetraplis Origenis nunc primum editus a singulari Chrisiano codice (Rome: Typis Propagandae Fidei, 1772). Marcos, Natalio Fernández, The Septuagint in Context. Introduction to the Greek Versions of the Bible. Translated by Wilfred G. E. Watson (Leiden: Brill, 2000). McCrystall, Andrew, Studies in the Old Greek Translation of Daniel (PhD diss., Oxford University, 1980). McLay, Timothy R., Translation Technique and Textual Studies in the Old Greek and Theodotion Versions of Daniel (PhD diss., University of Durham, 1994). —, The OG and Th Versions of Daniel. SCS 43 (Atlanta, GE: Scholars Press, 1996). —, “Syntactic Profiles and the Characteristics of Revision: A Response to Karen Jobes,” BIOSCS 29 (1996): 15–21.

Recensional Additions

223

—, “It’s a Question of Influence: The Theodotion and the Old Greek Texts of Daniel,” in Origen’s Hexapla and Fragments: Papers presented at the Rich Seminar on the Hexapla, ed. Alison Salvesen. TSAJ 58 (Tübingen: Mohr Siebeck, 1998), 231–254. —, “The Relationship Between Greek Translations of Daniel 1–3,” BIOSCS 37 (2004): 29–53. —, “Double Translations in the Greek Versions of Daniel,” in Interpreting Translation: Studies on the LXX and Ezekiel in Honour of Johan Lust, eds. F. García Martinez and M. Vervenne (Leuven: University Press, 2005), 255–267. —, “The Old Greek Translation of Daniel IV–VI and the Formation of the Book of Daniel,” VT 55/3 (2005): 304–323. —, “The Greek Translations of Daniel 4–6,” in The Temple in Text and Tradition: A Festschrift in Honour of Robert Hayward, ed. Timothy R. McLay (London: Bloomsbury, 2015), 187–214. —, “Daniel (Old Greek and Theodotion),” in T&T Clark Companion to the Septuagint, ed. James K. Aitken. Bloomsbury Companions (London: Bloomsbury T&T Clark, 2015), 544–554. Montgomery, James A., A Critical and Exegetical Commentary on the Book of Daniel. ICC (Edinburgh: T&T Clark, 1964). Munnich, Olivier, “Le text de la Septante,” in La Bible grecque des Septante du judaïsme hellénistique à la patristique grecque, eds. Marguerite Harl, Gilles Dorival, and Olivier Munnich (Paris: Cerf, 1994), 129–200. Newsom, Carol A., and Brennan W. Breed, Daniel, A Commentary. OTL (Louisville, KY: Westminster John Knox, 2014). Obiajunwa, Chukwudi J., Semitic Interference in Theodotion-Daniel (PhD diss., The Catholic University of America, 1999). Olariu, Daniel, The Quest for the Common Basis in the Greek Versions of the Book of Daniel. M.A. thesis (The Hebrew University of Jerusalem, 2015). —, “18.1 Textual History of Daniel.” in Textual History of the Bible: The Hebrew Bible, eds. Armin Lange and Emanuel Tov. Vol. 1C Writings (Leiden: Brill, 2017), 517–527. —, “Criteria for Determining the Common Basis of the Greek Versions of Daniel,” Textus 28 (2019): 105–124. —, “How Does a Reviser Work? Insights from Theodotion’s Recension of Daniel,” Paper presented at the International Symposium “Explorări în Tradiția Biblică Românească și Europeană,” 9th ed. Universitatea “Alexandru Ioan Cuza” din Iaşi, 9–11 May 2019, 1–15. —, “The Mechanics of the Recensional Process: Theodotion’s Treatment of First-Found Equivalents in Old Greek Daniel,” JSCS 52 (2019): 173–191. —, An Analysis of the Revisional Process in Theodotion’s Greek Text of Daniel (PhD diss., Hebrew University of Jerusalem, 2021). Pietersma, Albert, and Benjamin G. Wright, eds. A New English Translation of the Septuagint. (Oxford: Oxford University Press, 2007). Rahlfs, Alfred, and Robert Hanhart, eds. Septuaginta, id est Vetus Testamentum graece iuxta LXX interpretes. 2nd ed. (Editio Altera; Stuttgart: Deutsche Bibelgesellschaft, 2006). Rießler, Paul, Das Buch Daniel, Textkritische Untersuchung (Stuttgart/Wien: Roth, 1899). —, Das Buch Daniel, Kurzgefasster wissenschaftlicher Kommentar zu den Heiligen Schriften des Alten Testamentes III/3/2 (Wien: Mayer, 1902).

224

Daniel Olariu

Roca-Puig, R., Daniel: dos semifolios del còdex 967: papir de Barcelona, Inv. nº. 42 i 43. (Barcelona: Grafos, 1974). Roca-Puig, R., “Daniel: Due Semifogli del Codex 967,” Aeg 56 (1976): 3–18. Schmitt, Armin, Stammt der sogenannte “θ”-Text bei Daniel wirklich von Theodotion? MSU 9 (Göttingen: Vandenhoeck & Ruprecht, 1966). –, “Die griechischen Danieltexte (“θ’” und ο′) und das Theodotionproblem,” BZ 36 (1992): 1–29. Segal, Michael, “Calculating the End: Inner-Danielic Chronological Developments,” VT 68 (2018): 272–296. Seow, C. Leong, Daniel. Westminster Bible Companion (Louisville, KY: Westminster John Knox, 2003). Swete, Henry Barclay, The Old Testament in Greek. 3 vols. (Cambridge: The University Press, 1894–1912). The Nicene and Post-Nicene Fathers, ed. Philip Schaff and Henry Wace. 1886–1889. 14 vols. (Repr., Grand Rapids, MI: Eerdmans, 1979). Tov, Emanuel, The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of Jeremiah 29–52 and Baruch 1:1–3:8. HSM 8 (Missoula, MT: Scholars Press, 1976). —, Textual Criticism of the Hebrew Bible. 3rd rev. and enl. ed. (Minneapolis, MN: Fortress, 2012). —, The Text-Critical Use of the Septuagint in Biblical Research. 3rd. rev. and enl. ed. (Winona Lake, IN: Eisenbrauns, 2015). Weber, Robert, and Roger Gryson, eds., Biblia sacra iuxta vulgatam versionem (5th ed.; Stuttgart: Deutsche Bibelgesellschaft, 2007). Wenthe, Dean Orrin, The Old Greek Translation of Daniel 1–6 (PhD diss., University of Notre Dame, 1991). Ziegler, Joseph, Olivier Munnich and Detlef Fraenkel, eds., Susanna, Daniel, Bel et Draco. Septuaginta. Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum 16.2 (Göttingen: Vandenhoeck & Ruprecht, 1999).

Indices Ancient Sources Hebrew Bible/Old Testament Gen 30:22

20; 21

Exod 1:19 142 1:21 142 34:6 215 Lev 13:10 169 25:11 166 Num 14:18 215 23:19 21 Deut 6:4 211 33:9 42 Ruth 2:7 205 I Regnorum 1:13 20 2:14 13; 18 2:18, 28 13 4:14–16 16 6:20 65 8:2 66 12:9 55 13:13 18 14:3 13 14:10 55 14:18 13

14:32 13 14:37 55 15:3, 8 13 15:11 16; 20 15:28 17 15:29 21 15:30–32 28 16:22 65 17:13 66 17:47 55 18:23 67 21:3 55 22:41–51 73 23:4 55 23:12 55 23:14 55 23:17 66 24:5 55 24:11 55 28:17 18 30:8, 15, 23 13 30:10 13; 23 30:21 13; 22; 23 II Regnorum 3:28 44 5:11 36, 44 5:19 55 8:17 71 10:6 28 10:8 59 10:15 58 11:2 27 11:4 44 11:23 59 12:15 44 12:16 45 12:17 45 13:3 45

226 II Regnorum (cont.) 13:24 45 13:32 45 14:7 28; 49 14:11 49 14:20 49 14:21 49 14:27 49 14:32 38; 49 15:1–12 40 15:2 39; 46; 49 15:15 28 16:1 58 16:8 55 18:9 46 19:16 58 19:17 58 20:10 46 20:13 46 20:23 71 21:9 55 21:16 47 22:41 47 22:43 47 22:46 41; 42; 43; 47 22:49 41; 42; 43; 47 22:51 48 23:1 48 23:3 48 23:20.22 72 24:14 55 24:16 16 24:18 48 III Regnorum 1:1–2:11 53 1:8.26.32.36.38.44 72 1:2 65 1:17 64 1:28 65 1:34 68f. 1:35 64

Indices 1:38 71 1:44 71 1:46 64 1:47 64 2:5 65 2:8 58 2:11 27 2:19 58; 64 2:23 73 2:25.29.34.35.46 72 2:34 65 2:26 72 2:35 71 2:35a–o 75 2:35c 64 2:35n 58 2:36–46 75 2:46a–l 75 2:42 65 2:46 64; 65 2:46e 64 2:46h 71 3:2 64 3:15 65; 205 3:16 65 3:20 73 3:24 64 4:2 71 4:4 64; 71; 72; 73 4:31 73 4:37 73 6:10 65 6:23 65 6:30 73 7:5 73 7:10 73 7:14 72 7:16–18 72 7:17,41–42 72 7:38 75 7:50 75 8:1 73

227

Indices 8:22 65 8:33 72 8:66 72 9:37 73 10:8 65 10:15 72 11:4 73 11:14 73 11:17 72 11:26 59 12:6 65 12:12a.b 59 12:24a–z 56; 58; 61; 70 12:24a.b 59 12:24k.l 60 12:24r 58 14:21 59 14:31 59 15:2 59 15:10 59 15:15 72 15:18 55 16:8 75 16:11 72 16:15 65 16:28 75 16:28a 59 16:28a–h 73; 74; 75 16:29 75 16:31 67 17:1 65 18:2 59 18:9 55 18:15 65 18:16 58 18:17 73 18:21 72 18:29 65 18:30–32 76 18:46 63 19:26 58 20:1 63

20:10 63; 67 20:18 58 21:1 59 21:13 55 21:14–17 62 21:14 62 21:15 62: 63 21:15 63 21:19 59 21:23 73 21:28 55 22 27 22:1 53; 59 22:6.12 55 22:10 68; 73 22:15 56 22:41–51 73; 74; 75 22:42 59 23:16 73 23:31 59 23:36 59 24:8 59 24:11 73 24:18 59 25:17 72 25:19 73 IV Regnorum 1:2 69 1:3.6.7 58 1:11 64 1:13 64 1:14 63; 64: 65 1:16 64 2:3 66 2:10 64 2:14 64 2:15 58 2:18 63; 65 3:4 70 3:9 69 3:10.18 55

228 IV Regnorum (cont.) 3:12 64 3:13 55; 63; 66 3:14 65 3:18 66; 67 3:23 73 3:25 64; 71 3:27 70 4 69 4:2 64; 69 4:3 69 4:4 71 4:5 69; 71 4:7 69 4:8 69 4:12 65 4:19 69 4:20 73 4:23 69 4:24 69 4:26 58; 69 4:27 69 4:28 69 4:29 69 4.31 58 4:36 69 4:39 69 4:41 69 4:42 69 5:15 65; 205 5:16 65 5:21 58 5:25 205 5:26 58 6:8 70 6:10 73 6:30 73 7:2 70 7:6 70; 71 7:8 54 7:10 60 8:8 72

Indices 8:9 58; 65; 205 8:10 69 8:12 74 8:22 74 9:18 58 9:32 74 10:4 65 10:8 59 10:11 74 10:15 58 10:19 74 10:30–31 74; 77 10:30 77; 78 11:4 71 11:9 68 11:12 72 11:19 71 13:2–7.23 76 13:3 55 13:7 76 13:10–11 76 13:12–13 76 13:14–21 76; 77 13:20–21 77 13:22 76 13:23 76 13:24–25 76 14:6 71 14:13 70 14:15–16 76 17:2 68; 74 17:3–4 72 17:20 55 17:31 71 18:17 205 18:30 55 19:10 55 20:12 72 21:6 71 21:14 55 21:27 58 23:4 66; 74

229

Indices 23:5 72 23:6 70 23:7 72 23:8 59; 60 23:29 58 23:33 69 25:4 72 25:8 65 25:12 72 25:13.16 72 25:14 72 25:17 72 25:18 66; 74 25:23 76 25:30 53 II Par 23:8 68 Neh 5:14 210 9:17 215 I Mac 2:42 119 7:13 119 II Mac 2:19–23 87 14:6 119 IV Mac 1:1–16 88 1:1.30 85 1:27 86 1:28–29 86 1:29 86 2:4 86 2:7 86 2:9 85 3:5 85 4:19 85

5:2–6 88 5:27 86 5:36 85 6:1 85 6:6 85 6:10–11 86 6:11 85 6:17 86 6:19 86 6:21–24 85 6:31–9:9 85 7:1–3 86 7:4 86 7:5 86 7:5.17 85 7:6 86 7:10 86 7:11 85 7:13 86 7:17–23 85 8:4 86 8:8 86 8:13 85 8:16 85 8:19 86 9:19 86 9:26 86 9:28 86 10:5 85 10:8 85; 86 10:12 85 11:10 85 11:4 85 11:5 84 11:25 86 12:13–16:21 85 13:1 85 13:5.16 86 13:6–7.13 86 13:6–10 85 13:15 85 13:18–22 85

230 IV Mac (cont.) 13:22–14:1 85 13:21 85 13:25 86 14:3 85 14:3–8 86 14:13–20 86 14:19 86 14:15–17 85 14:16 85 15:1–4 85 15:5 85 15:15.20.24–32 85 15:16–21 88 15:19 85; 86 15:21 86 15:22–24 85 15:25–28 86 15:29 86 15:31 86 16:5 85 16:5–6 86 16:13 85 16:14–23 88 16:20 85 16:24 85 16:25–17:4 85 17:13–18:1 85 17:2–22 85 17:3 85; 86 17:4 86 17:5 86 17:7–12 88 17:11–16 86 17:20 85 17:22 86 18:3–4 85 18:5 85 18:6–15 88 18:23 86

Indices Ps 1:1 177 1:1–2a 160 7:8 186 12:8 42 18:35 177 18:36 187 22 182 30:10 180 31:9 176 32:6 176 32:7 177 35:16 176 35:25 187 35:27 177; 187 46:4 177 46:5 177 49:1 173 49:2 173 49:3 159; 173 49:6 184; 186 49:10 180 49:11 159 49:12 170; 171 49:15 177 75:4 187 86(85):15 215 89:26 176 89:28 177 103(102):8 215 112:1 160 122(121):7 215 145(144):8 215 Prov 1:32 215 14:29 215 15:18 215 16:32 215 17:1 215 17:27 215 22:12 42

Indices Eccl 2:8a 101 2:12c 101 2:22a 101 2:23 101 3:11a 101 3:11c 101 3:13 102 3:16c 101 4:8b 101 4:9–10a 96; 97 5:6b 101 5:18 102 5:18b 101 6:5 99 6:5b 101 7:1b 101 7:2 101; 102 7:6 101 7:12 99 7:14a 101 7:14b 101 7:20 98 7:29 97 8:4b 101 8:10b 101 8:12b 101 9:9 105; 106 9:9b 101; 107 9:9c 109 9:9d 107 10:5b 101 10:11 101 10:11b 103 11:9 104 11:9b 101; 103 12:13 110 12:13b 101; 110 Iob 4:2 137 6:5a 139

231 10: 3 137 11:7 137 14:14 137 15:2 137 15:8 137 22:3b 149 22:12a 149 22:17b 149 22:18a 147 22:29b 149 22:42 131; 133; 142; 143; 144; 145; 146; 148 23:3b 146 23:6 137 23:6a 149 23:8 147; 149 23:9a 148; 149; 150 23:14a 146 23:15c 149; 150 23:16b 149; 150 24:1b–2a 146; 147 24:5b 149 24:10a 146 24:12c 149 24:13b 150 24:15c 149 24:22b 149 24:23 149 24:24ab 146 24:25b 149 25:2a 149 25:3b 149 25:5a 149 26:5 137; 149 26:13b 149; 150 26:14bc 149 27:1 149 27:9a 149 27:12b 146 27:14b 146 27:20a 147 27:23a 149

232 Iob (cont.) 28:1b 147 28:15a 147 28:16a 147 28:16b 149 28:18b 149 28:25–26a 149; 150 28:26b 149 29:2a 148 29:2b 148 29:4a 149 29:4b 149 29:11b 149 29:25b 147 29:25c 149 30:1d 149 30:2b–3a 149 30:3a 149 30:4a 149 30:7a 139; 147; 149 30:12c 147 30:16a 149 30:17a 149; 150 30:21a 146 30:27a 149 30:30 149 31:2a 150 31:6a 149 31:9a 146 31:9b 149 31:11ab 146 31:13a 147 31:12a 146 31:21a 147 31:29b 149 31:31a 147 31:40a 147 32:1b 149 32:11c 147 33:13 149 33:26b 149 34:5b 147

Indices 34:11b 149 34:17a 149 34:18a 146; 147 34:33b 149 35:6a 149 35:16b 149 36:2b 147 36:5b–6a 149 36:23a 149 36:28a 147 36:33a 149 37:1a 147 37:3a 147 37:6b 149 37: 9a 147 37:10a 149 37:13a 149 37:14b 149 37:15a 149 37:18a 149 37:20b 149 38:7b 149 38:9a 149 38:12a 147 38:12b 149 38:16b 149 38:24a 149 38:25a 147 38:25b 149 38:27b 150 38:28 137 38:28a 149 38:32a 149 38:35b 147 38:38a 149 38:39a 149 38:40a 148; 149; 150 39:1a 149 39:4b 149 39:13a 147 39:18a 145; 147 39:21a 149

233

Indices 39:22 135 39:24 135 39:24b 134; 149 39:30b 149 40:8ab 136; 137 40:8a 149 40:8b 138; 143; 147 40:10a 149 40:16a 149 40:17a 147 40:22b 147 40:30a 149 40:31 137 41:1a 148 41:2a 149 41:3a 148 41:4a 148 41:4b 140; 141 41:10a 147 41:21a 150 42:2b 149 42:5a 149 42:7c 149 42:8c 149 42:10c 149 42:11f 148; 149 42:11g 147 42:14a 147 42:14c 147 42:17a 147 Sir

2:24 117 3:8 117 5:10 117 8 119 8:20 117 8:25 118 8:30 117 9–16 119 9:6a 119 9:6b 119 9:7a 119 11:4 118 13:3 118 14:9 117 17:16 118 17:30 119 17:30–46 119 18:1–9 119 Os 11 178 11:1 177; 178 Mich 7:17 42 Ioel 2:13 215 Ion 3:9, 10 16 4:2 215

5:4 215 11:21 67 42:1 66

Nah 1:3 215

PsSal 2 119 2–8 119 2:4b–6b 119 2:13 118

Agg 1:1 210 1:14 210 2:2 210 2:21 210

234 Is 13:21–22 180 49:6 67 Ier 7:10 205 22:4 124 22:6 124 22:7 125; 126 22:8 125 22:11 125 22:12 125 22:13 124; 125; 126; 127 22:14 125 22:15 125 22:16 125 22:16b 148 22:17 125; 127 22:18–19 124; 127 22:22a 148 40:3 124 Ez 9:2 205 10:6 205 16:49 215 Dan 1–3 193 1:1 199 1:2 199 1:3 215; 220 1:6 209 1:7 200 1:15 200 1:18 200 2:2 205 2:5 200 2:6 200 2:7 200 2:13 200; 213 2:15 212; 213

Indices 2:16 200 2:24 200; 201 2:25 216 2:26 200 2:27 200 2:28 211 2:30 200; 201 2:32 202 2:33 200; 201 2:36 200 2:38 213 2:41 200 2:42 200 2:43 200 2:44 200 2:45 202 2:47 211 2:48 213 3:12 200 3:15 200 3:16 200 3:17 211 3:23 200 3:28 200; 219 3:31–33 196 4–6 193; 194 4:3–6(6–9) 196 4:5–6 196 4:5(8) 214 4:6(9) 214 4:9(12) 196 4:15(18) 196; 214 4:16(19) 200 4:17(20) 196 4:19(22) 200 4:24(27) 214 4:28(31) 200 4:30(33) 200 4:32(35) 196 4:34(37b,c) 200 4:37(34) 196 5:0 196; 200

235

Indices 5:2 196 5:7 201; 202 5:11 196 5:13–16 196 5:14 196 5:16 201; 202 5: 18–22 196 5:29 201 6:4(3) 196 6:5(4)–6(5) 200 6:11(10) 196; 200 6:13(12) 196; 200 6:16(15) 196 6:17(16) 196 6:21(20) 216 6:23(22) 200 7–12 193 7:2–14 217 7:7 200 7:16 200 7:17 216; 217; 220 7:18 217 7:23 208 7:24 203 8 218 8:4 200 8:8 206 8:13 203 8:14 204; 220 8:17 204; 205 8:22 206 9:2 210 9:23 210 10:13 219 11:7 209 11:8 199; 208; 209 12:1 217 12:2 200 12:7 207 12:13 219

Dead Sea Scrolls 4QJosa 54 4QJudga 54 4QSama,c 27; 28; 30; 34; 43 4QSama 28; 30; 31; 32; 33; 34; 35; 36; 38; 40; 41; 42; 43; 44; 46; 47; 48; 53; 54 4QSamc 28; 31; 33; 34; 35; 38; 39; 40; 41; 43; 49 4QSam 2:7 33 4QSam 6:9 33 4QSam 11:6 33 4QSam 11:8 33 4QSam 12:1 33 4QSam 12:16 32 4QSam 13:16 32 4QSam 13:37 33 4QSam 13:39 32 4QSam 14:9 33 4QSam 15:12 33 4QSam 17:25 32 4QSam 18:9 33 4QSam 18:11 32 4QSam 21:4 32 4QSam 23:4 33 1 4QSam 24:17 32 4QSam 24:172 32

Patristic Sources Chrysostomus, Expositiones in Psalmos (CPG 4413) 55.90 Epiphanius, Ancoratus (CPG 3744) 97.1–2 Eusebius, Commentarii in Psalmos (CPG 3467) 23.1149 Eusebius, Commentarii in Isaiam (CPG 3468) 1.64

163 162 163 163

236

Indices

Origenes, Commentarii in Matthaeum (CPG 1450) 172 Origenes, Epistula ad Iulium Africanum (CPG 1494) 172 Origenes, Selecta in Genesim (CPG 1410) PG 12, 133 168 Origenes, Selecta in Psalmos (CPG 1425) PG 12, 1084; 12, 1057–1060 172 Theodoretus, Interpretatio in Psalmos (CPG 6202) PG 80,920 168

Manuscripts Septuagint Manuscripts Sigla according to Rahlfs, cf. https://septua�ginta.uni-goettingen.de/catalogue A

B

C M S V

11; 16; 17; 18; 20; 21; 22; 23; 30; 32; 33; 36; 39; 48; 58; 63; 64; 65; 66; 83; 85; 96; 98; 103; 106; 110; 125; 140 11; 16; 17; 18; 20; 21; 22; 23; 29; 32; 33; 39; 45; 46; 48; 49; 55; 56; 57; 58; 59; 60; 61; 62; 63; 64; 65; 68; 69; 70; 71; 76; 77; 85; 96; 98; 103; 106; 107; 125 96; 98; 106; 110 11; 29; 31; 32; 45; 46; 47; 49; 64; 75; 77; 169; 183 70; 83; 96; 98; 103; 106; 110; 125 11; 17; 21; 29; 32; 44; 45; 47; 48; 49; 58; 60; 61; 64; 75; 77; 83; 140

L

16; 17; 18; 21; 32; 44; 45; 47; 48; 56; 58; 59; 61; 64; 65; 68; 69; 70; 71; 72; 76; 77; 101; 103; 110

19 26

11; 17; 29; 53; 56; 58; 60; 62 125; 126

29 44 46 52 55

11; 17; 31; 32; 45; 46; 48; 49; 69 11; 21; 22; 63; 64; 78; 83; 84 11; 61; 63; 81; 83; 84; 126 11; 58; 65; 78 11; 17; 20; 29; 32; 33; 45; 48; 49; 57; 58; 59; 60; 61; 63; 65; 71; 77; 78; 83; 84 56 11; 48; 55; 58; 64 57 165 58 75; 83; 84; 85 62 81; 83; 84; 85 64 11; 17; 29; 60; 77; 78 68 11; 29; 63; 103; 106; 109 71 11; 17; 31; 32; 45; 46; 48; 57; 58; 59; 60; 61; 62; 63; 65; 75; 78; 83; 84; 85 74 11; 17; 31: 58; 63; 65; 78; 83; 84 82 11; 17; 29; 53; 56; 58; 59; 60; 64; 65; 78 86 160; 164; 174; 177; 183 88 195 90 118 92 11; 31; 58; 63; 65 93 11; 29; 47; 53; 56; 58; 60; 62; 65; 69; 77; 78 98 11; 31; 36 106 11; 17; 31; 58; 60; 61; 64; 65; 125 107 11; 17; 56; 61; 63; 65; 83; 84 108 11; 29; 53; 58; 60; 62; 77; 78 110 134 113 160; 174 119 11; 58 120 11; 31; 65; 83; 84 121 11; 15; 16; 18; 19; 21; 29; 46; 48; 53; 58; 59; 62; 63; 65; 69; 75 122 11; 29 125 11; 17; 46; 49; 63; 65; 78; 103; 106; 109; 110 127 11; 29; 47; 53; 56; 58; 60; 64; 65; 69; 77; 78; 125; 126 130 11; 31; 47; 58; 77; 103; 106 134 11; 31; 58; 65 137 134; 136

Indices 138 134; 138; 139 139 134; 135; 136; 138 147 134 149 110 155 106; 109 157 150 158 11; 17; 18; 29; 32; 44; 45; 46; 47; 48; 49; 58; 60; 61; 63; 64; 65; 77; 78 161 106; 107; 109; 110; 138 214 63 236 11; 62; 78; 84 239 125 242 11; 46; 58; 63; 64 243 11; 16; 31; 44; 62; 63; 65; 183 244 11; 16; 17; 29; 32; 39; 45; 48; 57; 58; 59; 60; 61; 62; 63; 64; 165 245 11; 17; 20; 21; 29; 32; 39; 44; 45; 47; 48; 57; 58; 59; 60; 61; 63; 64; 65; 75; 78 246 11; 46; 57; 59; 61; 63; 78 247 11; 16; 19; 20; 30; 32; 36; 44; 47; 48; 58; 59; 62; 63; 65; 66; 69; 71: 75; 78 248 106; 109; 110; 136; 137 249 134 250 133; 134; 135; 139 251 134 252 106; 109; 110; 136; 137; 140; 141; 148 253 106; 109; 110; 117 254 106 255 134; 136; 138; 150 256 134 257 134; 150 258 134 248 106; 107; 148 260 110; 117; 118; 134; 138 261 103 271 160; 174; 177; 180 296 110 311 106; 109; 110 313 11; 39; 78 314 11; 31; 65

316 317 318

237

83; 84 83; 84; 85 11; 17; 29; 32; 39; 45; 46; 47; 48; 49; 57; 58; 59; 60; 61; 62; 63; 64; 65; 69 322 83; 84 325 81; 83; 84 328 11; 44; 47; 48; 59; 60; 61; 63 336 110; 116; 117; 119 338 101; 103; 106; 109; 110 339 106; 109; 110 340 81; 83; 84 342 11; 22; 32; 39; 45; 47; 48; 57; 58; 59; 60; 61; 65; 69; 71; 75; 77; 78 357 105; 106; 109; 110 359 106; 109 370 11; 58; 83; 84 372 39; 45; 46; 49; 57; 58; 59; 60; 61; 63; 64; 69; 77; 78 371 98 376 11; 20; 30; 32; 33; 36; 44 379 11; 19; 31; 36; 45; 58 380 83; 84 381 11; 17; 29; 60 390 106 391 83; 84 395 134; 136; 138 397 83; 84 406 134 411 96; 106; 110 418 58 425 106; 109 443 98; 106; 109; 110 446 83; 84 452 83; 84 455 81; 83; 84 457 83; 84 460 11; 16; 17; 22; 29; 32; 33; 39; 46; 47; 48; 56; 58; 59; 60; 61; 63; 64; 65; 69; 71; 77; 78 467 83; 84 472 83; 84

238 473 83; 84 474 134 475 101 476 134 488 11; 31; 32; 39; 47; 58; 49; 63; 65; 75; 78 491 83; 84 489 11; 31; 32; 65 509 11; 15; 16; 17; 18; 19; 21; 32; 39; 47; 49; 53; 55; 56; 58; 59; 60; 61; 64; 65 523 150 527 11; 32; 45; 46; 49 530 11; 46; 47; 49; 58; 63; 65 534 81; 83; 84; 96; 106; 107; 109; 125; 126 538 125; 126 539 102; 103 542 81; 84; 85; 106; 109; 110 543 101; 106; 109; 110 544 125 547 103; 106; 109; 110 548 106; 110 549 106; 110 554 11; 17; 29; 39; 45; 46; 55; 58; 59; 61; 63; 65; 69; 78 559 134; 136 563 106 571 106 577 81; 83; 84 585 83; 84 586 83; 84 587 83; 84 591 83; 84 592 83; 84 594 83; 84 595 83; 84 596 83; 84 597 83; 84 598 58 602 106; 109; 110

Indices 607 83; 84 609 96 610 11; 17; 61; 63; 83; 84 612 134; 136; 138; 139 613 106; 109; 110; 125; 126 617 83; 84 627 57; 60; 61 637 101; 106; 109; 110; 140 639 83; 84 640 83; 84 641 83; 84 643 134; 135; 136; 138 645 103; 106; 109; 110 656 83; 84 668 83; 84 677 83 680 134; 138; 139 682 83; 84 683 83; 84 686 83; 84 690 83; 84; 85 695 83; 84 698 102; 106; 109; 110 699 83; 84 700 32; 58; 60; 69; 70; 77; 78 705 134 706 106; 109; 110 707 11; 17; 20; 21; 29; 32; 39; 47; 48; 49; 55; 57; 58; 60; 61; 63; 65 710 125; 126 713 83; 84 714 83; 84 719 118 728 81; 83; 84; 110 731 11; 16; 17; 31; 36; 63; 83; 84 732 134 738 83; 84 740 136; 138 741 83; 84 747 83; 84; 85 754 106

239

Indices 762 11; 31 765 134 766 103; 106; 109; 110 769 116 771 81; 83; 84 773 83; 84 774 83; 84 778 83; 84 782 83; 84 788 105; 133; 134; 135; 136; 137; 138; 141; 145; 148 789 83; 84 795 106; 109; 110 797 106; 109 799 11; 17; 78 842 11 845 11 846 11 867 11 870 93 930 83; 84 967 195 998 96; 98; 103; 105; 106; 107; 109; 110 1098 173; 174; 177; 180; 181; 182; 183; 185 1717 186 2005 174; 182 3002 83; 84 3005 134; 136; 137; 138; 139; 140; 141; 145, 148; 149; 150 3006 134; 138

Other Manuscripts Cairo, Patriarchal Library, Bible Ms. 11 Cologny, Bodmer Library, Papyrus Bodmer XXII London, British Library, Ms. Or. 1319 Michigan, University of Michigan, Ms. 158 (34) a–f

128 123–124 128 88

Milan, Bibliotheca Ambrosiana, Syrohexapla (Ambrosiana Mediolanensis) 167 Paris, BnF, Ms. Copte 1313 88 Princeton, Scheide Ms. 150 (Ms. Midyat) 166 Würzburg, Univ.-Bibl., Cod. Wirceburgensis M.p.th.f.64a 125–126

Words Greek ἀδιάπτωτος 141 ἀλλοφυλέω 85 ἄμωμος 105 ἀνατίκτω 85 ἀντιρρητορεύω 85 ἀποξαίνω 85 ἀρθρέμβολον 85 ἀσθενόψυχος 85 αὐτοδέσποτος 85 ἀφίημι 119 βαϊθ 156 βασιλεύω 20 γεδδούρ 13 δειλόψυχος 85 δεύτερος 66 ἑβραῖος, ἑβραϊκόν 164; 167 ἐγχειρίδιον 172 ἐθνοπάτωρ 85 ἐθνόπληθος 85 ἔθνος 117 ἐκμελίζω 85 ἐκπολιτεύω 85 ἐκριζωτής 85 ἔλεος 117 ἐναγκάλισμα 85 ἐννοσσοποιέομαι 85 ἐνώπιον 56

240 ἐπασθμαίνω 85 ἐπικαρπολογέομαι 85 ἐπιρρωγολογέομαι 85 ἑπταμήτωρ 85 εὐδοκέω 119 εὐοδόω 119 ἐφούδ 13 ζῆλος 117 θάπτεται 56 ἱερόψυχος 86 ἰσάστερος 86 καθαρίζω 119 καθίημι 18 κάλλος 119 κηρογονία 86 κοιμάω 56 κοσμοπληθής 86 κοσμοφορέω 86 λίθος 36 μαλακοψυχέω 86 μεταδιαιτάω 86 μεταμέλομαι 16 μεταπαιδεύω 86 μιαίνω 118 μιαροφαγία 86 μονοφαγία 86 οἰστρηλασία 86 ὀλεθροφόρος 86 ὁμοζηλία 86 ὁράω 118 παγγέωργος 86 παθοκράτεια 86 παντοφαγία 86 παρακαλοῦμαι 16 παρακέκλημαι 20 πᾶς 117 πατάσσω 18 περιτίθημι 201 περιλακίζομαι 86 περιχαλάω 86 προκακόομαι 86 πορεύεται 56

Indices προσεπικατατείνω 86 προσημειόομαι 86 σαλωμ 156 τίλλω 118 τοῖχος 36 ὑπερασπίστρια 86 φεύγω 118 φόβος 136 χρηστεύομαι 119

Hebrew ‫ אבן‬36–37 ‫ אימה‬136 ‫ אמר‬41 ‫ אפוד‬13 ‫ אראה‬38–39 ‫ בית‬156 ‫ גדוד‬13 ‫ חרם‬13 ‫ מׁשח‬16 ‫ מׁשנה‬66 ‫ נא‬38 ‫ נכה‬18 ‫ נחם‬16 ‫ נטה‬13 ‫ עיט‬13 ‫ ענה‬41 ‫ פגר‬23 ‫ קיר‬36–37 ‫ ראה‬39 ‫ ׁשלום‬156

Persons Historical Persons Agrippa I Alexander VII Ambrose of Milan

120 195 82

Indices Antiochus IV Epiphanes 81 Aquila 53; 63; 65–66; 95; 107; 109–110; 136–137; 139; 147–148; 150; 156; 164–168; 188 Athanasius II 116 Caligula 120 Eusebius 81 Gregory of Nazianzus 82 Herod the Great 120 Hesychius 114 Jason of Cyrene 87 Jesus 120 John Chrysostom 82 Josephus 81; 82 Justinian the Great 117 Lucian 114–115 Origen 19; 65; 82; 114–115; 131 Paul of Tella 87; 195 Pompey 119 Seleucus IV Philopator 81 Solomon 113 Symmachus 30; 53; 63; 65; 99; 102; 134; 135; 137–140; 146; 148; 156; 165–168; 188 Theodotion 29; 53; 63; 66; 148; 156; 164–168; 188; 193; 195; 220 Virgil 113

Modern Persons Aejmelaeus, Anneli 104 Albrecht, Felix 100 Allatius, Leo 195 Barthelemy, Dominique 14; 27; 54; 57; 65; 100

241

Bludau, August 196; 197 Bruce, Frederick Fyvie 197 Cavallo, Guglielmo 118 Cerda, Juan Luis de la 113 Ceulemans, Reinhart 133 Charles, Robert Henry 196 Cox, Claude 97 Delattre, Alain 89 Dieu, Léon 68 Driver, Samuel Rolles 196 Drusius, Johannes 137 Field, Frederick 104; 132; 133; 134; 137; 138; 142; 155; 156; 158 Flint, Peter W. 155; 169 Fraenkel, Detlef 142; 143 Fritzsche, Otto Fridolin 83; 113 Gardthausen, Victor 104 Gebhardt, Oscar von 113; 117 Geiger, Eduard Ephraem 113 Gentry, Peter 131 Goldman, Yohanan A.P. 100; 101; 107; 108; 109; 110 Hagedorn, Ursula and Dieter 133; 138; 140; 145 Hilgenfeld, Adolf 113 Jahn, Gustav 196 James, Montague Rhodes 113 Janssens, Gerard 176 Jeansonne, Pace Sharon 196 Kauhanen, Tuukka 34; 59 Lagarde, Paul de 11; 14; 113 Locher, Clemens 100 Marshall, Phillip 144 McCrystall, Andrew 197 McNeile, Alan Hugh 107; 108; 109; 110 Meade, John D. 131; 142 Mercati, Giovanni 158 Montfaucon, Bernard de 142 Montgomery, James A. 196 Munnich, Olivier 104; 142; 143; 144 Nobilius, Flaminius 137

242

Indices

Norton, Gerard 142 Perles, Felix 115 Perttilä, Elina 127 Pick, Bernhard 113 Rahlfs, Alfred 12; 67; 68; 72; 73; 104 Rießler, Paul 196 Ryan, Stephan D. 100 Ryle, Herbert Edward 113 Swete, Henry Barclay 113 Salvesen, Alison 131; 142 Schenker, Adrian 100

Skeat, Theodore 116 Tattam, Henry 128 ter Haar Romeny, Bas 131 Tov, Emanuel 73; 198 West, Martin L. 61 Wevers, John William 67; 104 Wright, Robert Bradley 113 Ziegler, Joseph 104; 124; 127; 133; 134; 135; 137; 138; 139; 140; 141; 143; 149