Die Sprache der Septuaginta: The Language of the Septuagint 9783641310950

Citation preview

HANDBUCH ZUR SEPTUAGINTA HANDBOOK OF THE SEPTUAGINT

LXX.H Herausgegeben von Martin Karrer, Wolfgang Kraus und Siegfried Kreuzer

HANDBUCH ZUR SEPTUAGINTA HANDBOOK OF THE SEPTUAGINT

LXX.H

BAND VOLUME

3

Eberhard Bons / Jan Joosten (Hg./eds.)

Die Sprache der Septuaginta The Language of the Septuagint

Gütersloher Verlagshaus

Sollte diese Publikation Links auf Webseiten Dritter enthalten, so übernehmen wir für deren Inhalte keine Haftung, da wir uns diese nicht zu eigen machen, sondern lediglich auf deren Stand zum Zeitpunkt der Erstveröffentlichung verweisen. Der Inhalt dieses E-Books ist urheberrechtlich geschützt und enthält technische Sicherungsmaßnahmen gegen unbefugte Nutzung. Die Entfernung dieser Sicherung sowie die Nutzung durch unbefugte Verarbeitung, Vervielfältigung, Verbreitung oder öffentliche Zugänglichmachung, insbesondere in elektronischer Form, ist untersagt und kann straf- und zivilrechtliche Sanktionen nach sich ziehen. Wir haben uns bemüht, alle Rechteinhaber an den aufgeführten Zitaten ausfindig zu machen, verlagsüblich zu nennen und zu honorieren. Sollte uns dies im Einzelfall nicht gelungen sein, bitten wir um Nachricht durch den Rechteinhaber.

1. Auflage Copyright © 2023 Gütersloher Verlagshaus, Gütersloh, in der Penguin Random House Verlagsgruppe GmbH, Neumarkter Str. 28, 81673 München ISBN 978-3-641-31095-0 www.gtvh.de

Inhaltsverzeichnis Vorwort zum Handbuch zur Septuaginta/Handbook of the Septuagint . . . . . . . .

9

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

Preface

I

Prolegomena 1.

History of Scholarship on the Language of the Septuagint . . . . . . . . . . Stanley E. Porter

15

2.

Die Diversität des Griechischen in der Septuaginta . . . . . . . . . . . . . . Peter Prestel

39

II

The Greek of the Septuagint as Hellenistic Greek

3.

The Phases of the Greek Language . . . . . . . . . . . . . . . . . . . . . . Geoffrey Horrocks

71

4.

Phonology and Morphology of Septuagint Greek Gerard Mussies

. . . . . . . . . . . . . .

89

5.

The Vocabulary of the Septuagint and Documentary Evidence . . . . . . . . John A. L. Lee

98

6.

Septuagint Syntax and Hellenistic Greek . . . . . . . . . . . . . . . . . . . Anssi Voitila

109

7.

Literarische Anspielungen in der LXX Knut Usener

. . . . . . . . . . . . . . . . . . . .

119

III

The Greek of the Septuagint as translation Greek

8.

Limitations of Greek in Representing Hebrew . . . . . . . . . . . . . . . . Takamitsu Muraoka

129

9.

The Classification of Literalism in Ancient Hebrew-Greek Translation . . . . Cameron Boyd-Taylor

139

10.

The Study of Translation Technique Raija Sollamo

. . . . . . . . . . . . . . . . . . . . .

161

11.

Transcriptions of Hebrew Words . . . . . . . . . . . . . . . . . . . . . . . Katrin Hauspie

172

12.

Hebraisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrés Piquer Otero

182

5

Inhaltsverzeichnis

13.

Etymological Translations in the Septuagint Hans Ausloos / Bénédicte Lemmelijn

. . . . . . . . . . . . . . . . .

193

14.

New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila . . Peter J. Gentry

202

15.

Statistical Differences between Translation Greek and Non-Translation Greek Texts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Georg A. Walser

221

IV

Local influences

16.

Ägyptische Elemente im Griechischen der LXX Stefan Pfeiffer

. . . . . . . . . . . . . . .

231

17.

Septuagint Greek and the Jewish Sociolect in Egypt . . . . . . . . . . . . . Jan Joosten

246

18.

Do Some Books of the Septuagint Contain Elements of Palestinian Greek? Carlo Consani

.

257

V

The vocabulary of the Septuagint

19.

La lexicographie de la Septante . . . . . . . . . . . . . . . . . . . . . . . . Gilles Dorival

271

20.

Innovative Elements in the Vocabulary of the Septuagint Pentateuch Robert J. V. Hiebert

. . . .

306

21.

The Septuagint Translation of the Torah as a Source and Resource for the Post-Pentateuchal Translators . . . . . . . . . . . . . . . . . . . . . . . . . Emanuel Tov

316

22.1 Der Kult . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martin Vahrenhorst

329

22.2 Sin and Forgiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anna Passoni dell’Acqua

335

. . . . . . . . . . . . . . .

340

22.4 Le vocabulaire de la loi dans la Septante . . . . . . . . . . . . . . . . . . . Cécile Dogniez

350

22.3 Le lexique de l’administration et de la politique Gilles Dorival

VI

The style of the Septuagint

23.

The language of the Septuagint between Greek and Hebrew . . . . . . . . . Alexis Léonas

6

357

Inhaltsverzeichnis

Stylistic Features of the Septuagint . . . . . . . . . . . . . . . . . . . . . . Jennifer Dines

375

. . . . . . . . . . . . . . . . . . . . . .

386

. . . . . . . . . . . . . . . . . . . . .

393

25.3 The Language of Second Maccabees . . . . . . . . . . . . . . . . . . . . . Frank Shaw

407

25.4 Die Sprache des 3. Makkabäerbuches . . . . . . . . . . . . . . . . . . . . . Wolfgang Orth

416

25.5 Sprache und Stil des 4. Makkabäerbuches . . . . . . . . . . . . . . . . . . Christoph Kugelmeier

421

24.

25.1 The Style of the Book of Wisdom Luca Mazzinghi 25.2 The Language of the Book of Judith Eberhard Bons

VII The language of the Septuagint and the New Testament 26.

The Language of the Septuagint and the New Testament: Prolegomena . . . Moisés Silva

431

27.

Le vocabulaire de la Septante dans le Nouveau Testament . . . . . . . . . . Madeleine Wieger

440

28.

Die Sprache der Septuaginta und das Neue Testament: Grammatik . . . . . Thomas J. Kraus

451

29.

Die Sprache der Septuaginta und das Neue Testament: Stil Ralph Brucker

. . . . . . . . .

460

Indexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

473

1.

Scripture references

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

473

2.

Greek words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

489

3.

Modern authors

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

491

4.

General index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

497

Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

501

7

Vorwort zum Handbuch zur Septuaginta / Handbook of the Septuagint Das Handbuch zur Septuaginta will eine umfassende Darstellung der derzeitigen Forschungen um die Septuaginta geben. Es ist damit Hinführung zu den vielfältigen Fragen und Ergebnissen der Septuagintaforschung, Bilanz des aktuellen Standes und Grundlage für die weitere Forschung. Folgende Bände sind vorgesehen: Einleitung in die Septuaginta, Textgeschichte der Septuaginta, Sprache der Septuaginta, der historische Kontext der Septuaginta, Theologie der Septuaginta, Wirkungsgeschichte. Die Planungen für das Handbuch entstanden auf dem Hintergrund von »Septuaginta-Deutsch«. Schon die Übersetzung Septuaginta Deutsch. Das griechische Alte Testament in deutscher Übersetzung (hg. von Wolfgang Kraus und Martin Karrer, Stuttgart: Deutsche Bibelgesellschaft 2009; 22010) und die damit verbundenen Bände Septuaginta Deutsch. Erläuterungen und Kommentare (2 Bde. hg. von Martin Karrer und Wolfgang Kraus, 2011) waren international orientiert. In den Bänden des Handbuches spiegelt sich dieses Anliegen in der internationalen und interdisziplinären Zusammensetzung des Herausgeberkreises und der Autorenschaft. Die Septuagintaforschung erlebt in jüngster Zeit eine eindrucksvolle Blüte. Ein Ausdruck dafür sind die zahlreichen Übersetzungsprojekte. Während zuvor nur zwei schon ältere englische Übersetzungen existierten, gibt es nun bzw. sind in Bearbeitung eine neue Übersetzung ins Englische, eine französische Übersetzung, die deutsche Übersetzung, aber auch eine Übersetzung ins Rumänische, ins Spanische, ins Italienische, ins Neuhebräische und Neugriechische sowie Übersetzungen in das Japanische und Koreanische. Die Übersetzungen erleichtern den Zugang zur Septuaginta und fördern ihre Wahrnehmung nicht nur im Bereich der Theologie, sondern auch in anderen Fachgebieten wie etwa der Geschichte, der Judaistik, der Sprachwissenschaft oder der Übersetzungs- und der Editionswissenschaft. Zugleich ergeben sich immer wieder neue Fragestellungen. Die verschiedenen Teilbände des Handbuchs zur Septuaginta wollen hier die bisherigen Forschungen bündeln, neue Fragestellungen aufnehmen und sowohl Basis als auch Impuls für die weitere Forschung geben. Nachdem zu Beginn des Jahres 2016 mit LXX.H 1, »Einleitung in die Septuaginta«, der erste Band erschienen ist, wird hiermit LXX.H 3, »Sprache der Septuaginta / Language of the Septuagint«, vorgelegt. Die Hauptherausgeber danken den Herausgebern der Bände, in diesem Fall Eberhard Bons, Straßburg, und Jan Joosten, Oxford, und den zahlreichen Autorinnen und Autoren für ihre engagierte Arbeit und dem Gütersloher Verlagshaus für den Mut, dieses große Projekt auf den Weg zu bringen und zu realisieren. Martin Karrer, Wolfgang Kraus und Siegfried Kreuzer

9

Preface The Greek language used in the Septuagint elicited comments and debate already in Antiquity. Some ancient authors, such as Philo, praise it to high heaven, while others voice criticism, discretely (e. g. the translator of Ben Sira) or openly (e. g. Celsus). In the modern period it was commonly held that the language of the Septuagint represented some form of dialect, a Judaized Greek or, in a more theological mode: a type of Greek especially prepared for the expression of the Gospel. More recently, characterizations of this type have given way to the observation that Septuagint Greek is in actual fact extraordinarily diverse. The corpus contains books translated literally, books translated more freely, and books written from the start in Greek. The stylistic register of the different books varies between a rather good literary Koine and an almost vulgar level of language. Texts from the third century BCE rub shoulders with texts from later periods, some of them as late perhaps as the second century CE. In spite of this diversity, however, a family relationship can be recognized among the different Septuagint books in regard to their language. A measure of distinctness of this language must in any case be recognized because of the linguistic influence of the Septuagint on later writings: the notion of “Septuagintisms” in New Testament Greek or in other ancient writings indicates that the Septuagint origin of certain expressions leaps to the eye. The language of the Septuagint is a complex field of research, a field in motion with a constant supply of new data (e. g. papyri, inscriptions) and the development of new approaches. For all these reasons, it may appear daunting. It is nevertheless of interest to all Septuagint scholars, for no textual interpretation is possible without correct linguistic analysis. The present volume is designed as an introduction for nonspecialists to the state of the art in linguistic research on the Septuagint. The first section seeks to define the notion of “Septuagint Greek” through a consideration of the history of research and an exploration of different dimensions of diversity that need to be taken into account in any approach to the subject. The second section situates Septuagint Greek in its language-historical context, which is that of the Koine: Hellenistic Greek as a world language. In comparison to classical Greek, a number of developments, in both grammar and vocabulary, have to be taken into account, although the literary language also shows a measure of continuity with classical texts, particularly Homer. Features that characterize Koine Greek naturally turn up in the Septuagint, and should not be considered distinctive of “biblical Greek”. Most of the books of the Septuagint are translations from Hebrew or Aramaic, consequently the translation process and the way it affected the language of the Septuagint is investigated in the third section. The fourth section inquires to what extent it is possible to recognize local and cultural influences in Septuagint Greek. Most of the corpus must have been created in Egypt, as is demonstrated by linguistic features, but some books may have a Palestinian background. The question of the Jewish sociolect is also explored in this section. The fifth and sixth sections consider areas of language where the Septuagint is generally agreed to show distinctive features: vocabulary, and style. The section on 11

Preface

style presents a number of general characteristics of the Septuagint, but also discusses a number of specific writings whose style sets them apart from the rest of the corpus. Finally, section seven explores how the Greek of the Septuagint influenced the Greek of the New Testament. No effort has been made to harmonize different approaches elaborated in the individual chapters, or to present a unified picture. The dynamics of the debate are more instructive than a mere enumeration of assured results, which in any case are few in number. Some overlap between the various chapters has been tolerated on the view that complex issues (e. g. the question of “Hebraisms”) deserve to be presented in different perspectives. The plurality of scientific languages, German, English and French, was maintained on the same principle. We are grateful first and foremost to the authors of the individual chapters. It has been a privilege for us to work with such distinguished scholars. Preparing the manuscript has been a long and exacting process, but we think it was worth the wait. Thanks are due also to our research groups (EA 4377 and EA 4378) at the University of Strasbourg, to the University of Oxford, and to the “Septuaginta Deutsch” project, all of which supported the editorial process at various points. We also thank the publishing house “Gütersloher Verlagshaus”, particularly Mr Diedrich Steen and Ms Tanja Scheifele. Jason Dean, Séverin Schneider and Matthew Albanese took on specific editorial tasks, such as copy-editing and producing the indexes. Prof. Hans Schmoll went through the Greek quotations and checked the accentuation. To them too go our thanks. Jan Joosten Eberhard Bons

12

I Prolegomena

1. History of Scholarship on the Language of the Septuagint Stanley E. Porter

1. Introduction The history of scholarship on the nature of the Greek of the Septuagint is very similar to the history of scholarship on the Greek of the New Testament, for a number of reasons. These include the use of Greek in the Septuagint and in the New Testament as a result of the Hellenistic conquest of the Mediterranean world by Alexander the Great, the use of a form of Attic Greek (sometimes called Great Attic) by Alexander and his successors as an administrative language, the unified Mediterranean region (and beyond) created by the Hellenistic empires and later Rome that enabled widespread communication, and the need for a lingua franca to linguistically unify this expanding and diverse multi-cultural and multi-lingual region, as well as others. There are also a variety of factors that differentiate the Greek of the Septuagint from the Greek of the New Testament. These factors include the facts that the Septuagint is mostly, though not entirely, a translated document, the Septuagint has a (debatable) number of acknowledged Semitisms because of its Hebrew substrate, the divergent translation types found in the varying books (and even within books) create linguistic diversity, knowledge of this translational and creative process is obscure so that those involved and their levels of linguistic competence are unknown, the status of the Septuagint as a sacred text was different from that of the New Testament at its inception, the Septuagint was produced over a much lengthier period of time than the writing of the New Testament (even for skeptics regarding its extent of composition), the Septuagint was the product of a diverse translational and writing process that occurred in a different and developing set of linguistic contexts than the New Testament, the relationship of prestige to non-prestige languages within the diverse cultures differed (e. g. Egypt vs. Palestine or the northern Mediterranean), the multi-lingual context of the Septuagint, in terms of both its translators/writers and its readers, varied from that of the New Testament (even if they perhaps overlapped), and the uses of the Septuagint were different from those of the New Testament, among others. 1 These factors enter into discussion of the question concerning the nature of the Greek of both the New Testament and Septuagint, but are especially important for discussing the Septuagint as a translated religious text. As a result, it is rightly acknowledged that, like the New Testament, the Septuagint emerged out of a complex multi-lingual milieu over a period of time to be used by 1.

Many of these factors are mentioned—though rarely in systematic fashion—in introductions to the Septuagint. For some of the linguistic issues, see S. E. Porter, Verbal Aspect in the Greek of the New Testament, with Reference to Tense and Mood (Studies in Biblical Greek 1), New York, NY 1989, 111-156. 1. Introduction

15

1. History of Scholarship on the Language of the Septuagint

those for whom Greek was sometimes, if not often, not their primary language. Unlike the New Testament, however, it contained both translated documents and those originally written in Greek, to be used in a wide variety of especially religious contexts as sacred texts. Therefore, it is important to trace the history of the discussion of views of the Greek of the Septuagint in its own right, even if the topics and some of the approaches and people involved are similar to those of the New Testament debate.

2. Periods in the Study of the Greek of the Septuagint The history of debate over the Greek of the New Testament goes back to the sixteenth century. Two early debates were those between the Greek purists and Hebraists and between the sacred and common Greek advocates. 2 The Hebraists won out in their short-lived battle in the late sixteenth and early seventeenth centuries because of the noteworthy differences of New Testament from Classical Greek, and then the advocates of a form of sacred or Holy Ghost Greek came strongly into the fore, as these features were more widely recognized and discussed. These debates then became more focused in the late nineteenth and early twentieth centuries with the discovery of the documentary Greek papyri in Egypt, especially with the work of Adolf Deissmann and James Hope Moulton. 3 The history of debate over the nature of the Greek of the Septuagint is much more recent. An early work was Friedrich W. Sturz’s on the Alexandrian dialect, 4 but fuller accounts did not begin to be written until roughly the mid nineteenth century. This history of discussion can be divided into the following rough major periods: Semitic Greek, Hellenistic/Koine Greek, Hebraic/Jewish Greek, and revival of the Koine Greek hypothesis, with the last two being coterminous, even if the Koine Greek hypothesis is still prevalent.

2.1 Semitic Greek Hypothesis At the time that the discussions of the Greek of the New Testament as a sacred language (Holy Ghost Greek) were on-going, discussion of what was sometimes called Biblical Greek (including the New Testament and the Septuagint) and then especially 2.

3.

4.

These debates are recounted in G. B. Winer, A Treatise on the Grammar of New Testament Greek (translated by W. F. Moulton, third edition), Edinburgh 1882, 12-41; G. B. Winer, Grammatik des neutestamentlichen Sprachidioms (edited by P. W. Schmiedel), part 1, Göttingen 1894, 4-30; J. W. Voelz, “The Language of the New Testament” in: W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, II.25.2, Berlin 1984, 893-977, especially 894-906; and S. Wong, “The Nature of the Greek of the New Testament—Its Past and Present” Scriptura 32 (1900), 1-27. Cf. also G. Friedrich, “Pre-History of the Theological Dictionary of the New Testament” in: G. Kittel / G. Friedrich (eds.), Theological Dictionary of the New Testament (translated by Geoffrey W. Bromiley), vol. 10, Grand Rapids, MI 1976, 613-661. This debate is traced in some detail in Porter, Verbal Aspect, 112-117 and S. E. Porter, “Introduction” in: S. E. Porter (ed.), The Language of the New Testament: Classic Essays (JSNTSup 60), Sheffield 1991, 11-38, along with supporting essays. F. W. Sturz, De Dialecto Alexandrina, Leipzig, 1784 (expanded to De dialecto Macedonica et Alexandrina, Leipzig 1808).

16

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

the Greek of the Septuagint emerged in its own right. Because of numerous recognizable differences between the Greek of the classical authors of several centuries earlier than the Greek of the New Testament (and even the Septuagint), as well as broad similarities between them, there came to be a characterization of Biblical Greek as its own distinct type of Greek. This theory developed in part as a reaction against those who, noting the differences between Classical and Septuagint Greek, denigrated the latter as in some way sub-standard. One of the first such studies of a distinct Biblical Greek, though limited to the Pentateuch, was by the German scholar Heinrich (H. G. J.) Thiersch. Although he makes some limited use of early papyri, 5 Thiersch systematically examines the Greek of the Pentateuch and compares it to Attic Greek. He identifies numerous Hebraisms, which he attributes to the fact that they reflect the Hebraic genius and were appropriate language for divine revelation. His volume is for the most part a chronicling of the perceived non-Attic uses in the Pentateuch. 6 In many ways similar, Zechariah Frankel offers a short grammar of Hebrew in the Septuagint. He classifies and gives representative examples of how various Hebrew constructions are manifested in the Septuagint, including nouns, pronouns, verbs, and particles. 7 Eduard Reuss calls this distinct Greek the “Hellenistic Idiom,” by which he means “the Greek language which the Jews settled in foreign lands learned to understand and speak, or which met them in Palestine itself.” The Hellenistic idiom was used outside of day to day conversation: “Whatever lay without this sphere [of daily conversation], especially religious ideas, and in general whatever was directly connected with the spirit of the Orient, was rather translated literally, with reference to thought, without regard to Greek usage, and the construction was sometimes modeled very strikingly after Shemitic syntax.” 8 Although Reuss offers a brief and accurate summary of the development of the Greek language, and appreciates that conversation 5.

6. 7.

8.

And is commended for this by Adolf Deissmann. See A. Deissmann, Light from the Ancient East: The New Testament Illustrated by Recently Discovered Texts of the Graeco-Roman World (translated by L. R. M. Strachan, fourth edition), London 1927 (1910), 48 (translation of Licht vom Osten: Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt (fourth edition), Tübingen 1923 [1908]). H. G. J. Thiersch, De Pentateuchi versione Alexandrina libri tres, Erlangen 1841, 52, cf. 65-188, for detailed comparisons. Z. Frankel, Vorstudien zu der Septuaginta (Historisch-kritische Studien zu der Septuaginta), Leipzig 1841, 132-163, followed by a section on hermeneutics and exegesis of the Septuagint (163-203). Emanuel Tov calls this year, 1841, and the work of Thiersch and Frankel, the beginning of the modern “study of translation technique.” See E. Tov, “The Nature and Study of the Translation Technique of the LXX in the Past and Present” in: C. E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (Septuagint and Cognate Studies 23), Atlanta, GA 1987, 342. E. Reuss, History of the Sacred Scriptures of the New Testament (translated by E. L. Houghton), Edinburgh 1884, 34, cf. 30-40, (translation of the fourth German edition of 1842). This is similar to what is found in E. Reuss, “Hellenistisches Idiom” in: J. J. Herzog / D. G. T. Plitt (eds.), Realencyklopädie für protestantische Theologie und Kirche (second edition), vol. 5, Leipzig 1879, 741-749. Adolf Deissmann claims that such a distinction regarding Greek was not found in linguistic circles, but only in theological ones, apart from G. Meyer, Griechische Grammatik, (Indogermanische Grammatiken 3, third edition), Leipzig 1896, 26. See A. Deissmann, “Hellenistic Greek with Special Consideration of the Greek Bible” in: Porter, Language of the New 2. Periods in the Study of the Greek of the Septuagint

17

1. History of Scholarship on the Language of the Septuagint

might be different from writing, he clearly has no category for the Greek of the Septuagint (or New Testament) within this development, except to give it its own label. The Semitic Greek hypothesis was more widely promoted in one of the first monographs on the topic of Biblical Greek by Edwin Hatch. In his book, the first essay of which is devoted to the value and use of the Septuagint, Hatch recognizes that there are many similarities between Attic Greek and Biblical Greek, but also that there are many differences. These are caused by a variety of factors, such as the lapse of time between the two language types, the Septuagint being a collection of books, the different social locations, and the Septuagint’s translational characteristics. As a result, Hatch concludes that the individual books “afford clear internal evidence that their writers, in most cases, were men whose thoughts were cast in a Semitic and not in a Hellenic mould.” 9 He thus continues, “Biblical Greek is thus a language which stands by itself,” reflective of a “Semitic mind.” 10 There are therefore numerous instances in which knowledge of specific examples of language use in the Septuagint enable interpreters to understand the Greek New Testament (which is Greek of the same type), including both new words and especially words already known but used in different ways. In a work focused upon the New Testament, but with an extensive introduction to the broader topic, Joseph Viteau provides one of the fullest expositions of the hypothesis that Biblical Greek is a unique form of Semitic Greek. 11 Viteau begins by surveying the history of Greek, dependent upon the work of the lexicographer Evangelinus Apostolides Sophocles. 12 He especially notes that Greek was not indigenous to Egypt (or Asia Minor), but that Jews of the Diaspora generally adopted the language of their environment as their spoken or colloquial language, as opposed to learning a literary language. This language is called “langue judéo-grecque, langue grecque hébraïsante, grec hébraïsant,” 13 as represented in the Septuagint. As Viteau states, “La langue judéogrecque ou grec hébraïsant est le grec post-classique, modifié dans sa couleur générale par l’hébreu et l’araméen, et mélangé d’hébraïsmes et d’aramaïsmes.” 14 This position came to represent the general tenor of belief regarding the Greek of the Septuagint for the nineteenth and even into the twentieth centuries, and was still maintained in a number of works, even after the discovery and appreciation of the Greek documentary papyri. Two of these worth mentioning are the grammar of the Septuagint written by F. C. Conybeare and St. George Stock and the handbook to the Septuagint by Richard Rusden Ottley. In an introduction to the original publication of the grammar by Conybeare and Stock, deleted from some subsequent reprints, they want to accept that the Greek of the Septuagint is Alexandrian in nature, but also that it is Biblical Greek because of the

9. 10. 11. 12. 13. 14.

Testament, 39 (translation of “Hellenistisches Griechisch” in: A. Hauck [ed.], Realencyklopädie für protestantische Theologie und Kirche [third edition], vol. 7, Leipzig 1899, 627-639). E. Hatch, Essays in Biblical Greek, Oxford 1889, 10, cf. 1-35. Hatch, Biblical Greek, 11. J. Viteau, Étude sur le grec du Nouveau Testament, Paris 1893. E. A. Sophocles, Greek Lexicon of the Roman and Byzantine Periods from B.C. 146 to A.D. 1100, Boston, MA 1870. Viteau, Étude, vi (in italics in the original). Viteau, Étude, xxii (emphasis in the original).

18

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

significant number of non-Hellenic Semitisms (many more than in Josephus). Conybeare and Stock go so far as to refer to the “uncompromising Hebraism of the Septuagint,” and attribute this feature “to the reverence felt by the translators for the Sacred Text.” 15 However, they also note that the Greek seems to be puerile or perhaps even, on numerous occasions, not understood by those translating it. The short and often reprinted grammar reflects this orientation at numerous places where Hebraisms are noted. 16 Ottley positions his work in a similar way—recognizing the convincing work of Deissmann and Moulton (see below) regarding the Greek of the New Testament and non-translation Greek (especially not wishing to denigrate Koine Greek) but also wishing to emphasize the Hebraic elements in the translation Greek of the Septuagint. In fact, Ottley spends much of his chapter on the language and style of the Septuagint defining his position in opposition to that of Deissmann and Moulton. As a result, he wishes to re-establish that those features explained by Moulton are still to be regarded as Semitisms: e. g. the relative pronoun with following redundant pronoun, προστιθέναι as an auxiliary indicating repeated action, the syntax of a verb with cognate object, a participle used for a finite verb, uses of ἐν, πᾶς + negative as “none,” and other phrases. 17 The Semitic Greek position holds that, even though written in Greek form, there is a distinctive Jewish Greek, possibly created under divine guidance for revelatory purposes, that is distinguished by particular linguistic constructions. These constructions are foreign to Greek but directly dependent upon Hebrew. This position has had a continued residual influence upon study of the Greek Bible, especially the Septuagint, to the point that even works that accept the force of later research (see the next section) continue to promote it in varying ways. 18

2.2 Hellenistic/Koine Greek Hypothesis Around the year 1895, scholarship on the language of the Septuagint took a significant change in direction. In light of the discoveries of numerous documentary papyri in Egypt, a number of scholars began to argue that the form of Greek used in the Septuagint (along with that of the New Testament, though not as translation Greek) was remarkably syntactically similar to the Greek of the papyri, and that what had been 15. F. C. Conybeare / St. G. Stock, Selections from the Septuagint According to the Text of Swete, Boston, MA 1905, 22-23, especially 23, for quotations (the entire introduction is on pages 1-24). 16. Conybeare / Stock, Selections, 25-100. 17. R. R. Ottley, A Handbook to the Septuagint, London 1920, 163 (see the entire chapter, pages 159-178). For a summary of research on these constructions, see Porter, Verbal Aspect, 119-141. 18. See, for example, B. F. Atkinson, The Greek Language, London 1931, 273-275, where he says many of the phrases in the Septuagint “were totally foreign to Greek minds” (274). This position is also maintained because of its presence in works on the New Testament. A distinct form of Jewish Greek was proposed by Friedrich Blass in the first edition of his grammar (Grammatik des Neutestamentlichen Griechisch, Göttingen 1896, 4-5) and this position is also maintained in the well-known English translation (F. Blass / A. Debrunner, A Greek Grammar of the New Testament and Other Early Christian Literature [translated by R. W. Funk], Chicago, IL 1963, 34) and is still found in the latest edition (F. Blass / A. Debrunner, Grammatik des neutestamentlichen Griechisch [seventeenth edition edited by F. Rehkopf], Göttingen 1990, 4-6). 2. Periods in the Study of the Greek of the Septuagint

19

1. History of Scholarship on the Language of the Septuagint

considered anomalous uses of Greek vocabulary could be explained by their similar use in these same papyri. Prior to the discovery of the papyri and the tangible evidence that they provided, there had been occasional arguments made by a number of scholars that Koine Greek was actually Attic Greek as it had been modified after the time of Alexander by its use throughout the Greco-Roman world. In his introduction to his translation of Georg Winer’s grammar, Edward Masson accounts for the Hebraisms within the Septuagint within the context of its being Koine Greek used in a way to preserve the “force and beauty” of the “inspired” original text. 19 Similarly, J. B. Lightfoot delivered lectures in 1863, in which he (rightly) conjectured that Greek words thought to have fallen from use were still part of the “common speech,” as would have been discovered if “letters that ordinary people wrote to each other” had been known. 20 The discovery of these very letters prompted a thorough re-examination of the koine language, including that of the Septuagint. This reformulation argued that Koine Greek was a developed form of the Attic-Greek based language spread abroad by Alexander in his conquests, 21 that it constituted a single dialect even if it had regional variations, and that, though there may have been some Semitic influence, it was recognizably Greek, as the evidence of the extra-biblical Greek, especially the documentary papyri, abundantly demonstrated. There are two major figures who argued most strongly for this position and influenced a large number of other scholars. The first major proponent was Adolf Deissmann. First in his Bible Studies (published in German in 1895 and 1897), and then later in his Light from the Ancient East (1908), as well as in a number of other works, Deissmann was the first to make widely known the importance of the then recently-discovered Egyptian documentary papyri, as well as other sources such as inscriptions. Deissmann argued against especially the Biblical Greek hypothesis noted above, contending that the Greek of the Septuagint (along with the New Testament) was part of the body of Egyptian and popular Greek of the Greco-Roman period. Deissmann preferred the term Hellenistic Greek to describe this Greek that was used from 300 BCE to 500/600 CE. Deissmann recognized the place of the Septuagint as a translated document, but he nevertheless discounted any form of Semitic Greek as being a spoken or written language. He admitted that there were Semitic or Hebraic elements in the Septuagint, and that the morphology and phonology had been affected. Nevertheless, he contended that the syntax was Hellenistic and not Semitic. Deissmann’s work was for the most part devoted to investigation of lexicographical items, where he maintained that 19. E. Masson, “Translator’s Prolegomena” in: G. B. Winer, A Grammar of the New Testament Diction, (sixth edition), Edinburgh 1866, vii. See J. R. Harris, “The So-Called Biblical Greek” The Expository Times 25 (1913), 54-55. 20. Cited in J. H. Moulton, Prolegomena, vol. 1 of his A Grammar of New Testament Greek (third edition), Edinburgh 1908 (1906), 242 (a note added to the second edition). See also J. Wellhausen, Einleitung in die drei ersten Evangelien, Berlin 1905, 9, who sees an idiomatic Greek used by the Gospel writers. 21. See J. Blomqvist, “The Nature of Post-Classical (Hellenistic) Greek” in: C. C. Caragounis (ed.), Greek: A Language in Evolution. Essays in Honour of Antonios N. Jannaris, Hildesheim 2010, 139-152, especially 139, where he selects the word “expansion” to describe what happened to Greek during the Hellenistic period.

20

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

given lexical items could generally be paralleled in abundance in the Hellenistic Greek of the papyri and inscriptions. 22 James Hope Moulton has the distinction of having done the most work to apply Deissmann’s theories regarding Koine (or Hellenistic) Greek to syntax, as well as being concerned with lexis. Moulton wrote a number of articles in the Classical Review and The Expositor, and his work culminated in his Prolegomena to his grammar of New Testament Greek. 23 The development of Moulton’s ideas under the influence of the papyrological evidence is reflected in the fact that, in the first edition of 1895 of his elementary New Testament Greek grammar, he refers to Hellenistic Greek as “Hebraic Greek, colloquial Greek, and late Greek.” 24 By the second edition of 1903, this Greek is “common Greek, colloquial Greek, and late Greek.” 25 As he explains in his Prolegomena, the use of Greek by Jews, especially in a bilingual environment, made more prominent constructions that otherwise would have remained comparatively infrequent. Examples include paratactic use of καί, various uses of prepositions such as ἐν, types of periphrasis, and use of various pronouns, among others. What we find in the Septuagint, Moulton thinks, is instances of “abnormal Greek produced by the effort of Greekspeaking men to translate the already obsolete and imperfectly understood Hebrew,” 22. A. Deissmann, Bibelstudien: Beiträge, zumeist aus den Papyri und Inschriften, zur Geschichte der Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums, Marburg 1895; Neue Bibelstudien: Sprachgeschichtliche Beiträge, zumeist aus den Papyri und Inschriften, zur Erklärung des Neuen Testaments, Marburg 1897, translated together as Bible Studies: Contributions Chiefly from Papyri and Inscriptions to the History of the Language, the Literature, and the Religion of Hellenistic Judaism and Primitive Christianity (translated by A. Grieve, second edition), Edinburgh 1909 (1901); Licht vom Osten (1908)/Light from the Ancient East (1910), which went through several editions, the fourth being published in 1923 (German)/1927 (English). Cf. also “Hellenistic Greek” (above note 8); The Philology of the Greek Bible: Its Present and Future (translated by L. R. M. Strachan), London 1908; and Die Urgeschichte des Christentums im Lichte der Sprachforschung, Tübingen 1910; among others. 23. J. H. Moulton, “New Lights on Biblical Greek” The Biblical World 19 (1902), 190-196; “Grammatical Notes from the Papyri” Classical Review 15 (1901), 31-38, 434-442; 18 (1904), 106-112, 151155; “Notes from the Papyri” The Expositor Sixth Series 3 (1901), 271-282; 7 (1903), 104-121; 8 (1903), 423-439; “Characteristics of New Testament Greek” The Expositor Sixth Series 9 (1904), 67-75, 215-225, 310-320, 359-368, 461-472; 10 (1904), 124-134, 168-174, 276-283, 353-364, 440-450; with many examples finding their way into his (and George Milligan’s) The Vocabulary of the Greek Testament: Illustrated from the Papyri and Other Non-Literary Sources, London 1914-1929; Prolegomena, especially 1-41 (re-issued in a German edition, Einleitung in die Sprache des Neuen Testaments, Heidelberg 1911); cf. The Science of Language and the Study of the New Testament (Inaugural Lecture), Manchester 1906; “Language of the New Testament” in: J. Hastings (ed.), Dictionary of the Bible, Edinburgh 1909, 528-530; “New Testament Greek in the Light of Modern Discovery” in: H. B. Swete (ed.), Essays on Some Biblical Questions of the Day: By Members of the University of Cambridge, London 1909, 461-505 (reprint in: Porter, Language of the New Testament, 60-97); From Egyptian Rubbish Heaps, London 1916; and “The Language of the New Testament” in: A. S. Peake (ed.), A Commentary on the Bible (with Supplement by A. J. Grieve), London 1919, 591-593. 24. J. H. Moulton, An Introduction to the Study of New Testament Greek, London 1895, 2 (emphasis in the original). 25. J. H. Moulton, An Introduction to the Study of New Testament Greek (second edition), London 1903, 2 (emphasis in the original). 2. Periods in the Study of the Greek of the Septuagint

21

1. History of Scholarship on the Language of the Septuagint

but such instances did not influence or have a lasting effect on the Greek language as a whole. 26 Whereas there probably were some differences in the Greek of the time, insofar as differences in pronunciation are concerned, there was not significant difference in the written varieties, as evidenced by letters being sent throughout the Roman Empire, so as to justify calling them dialects. 27 Deissmann and Moulton were followed in their findings by numerous other scholars. 28 The works of some of these scholars can be conveniently divided into three major groups of writings: general works on Koine Greek, those specifically on the Septuagint, and Septuagint grammars. The first category of general works on Koine Greek includes works by such scholars as H. A. A. Kennedy and Albert Thumb. In 1895, Kennedy published a book on the sources of the New Testament, in particular the Septuagint as a source for the New Testament vocabulary. 29 This book, which began from the standpoint of Hatch, ended up at very different conclusions. Kennedy traces the history and development of the Greek language, from the Attic of Xenophon through the influence of Alexander upon the use and spread of Greek to its use in the Septuagint. He examines the Greek of Egypt in relation to the common or colloquial Greek of the time, and especially examines the vocabulary of the Septuagint and that of the New Testament. He concludes that the Septuagint “is the first entire group of writings composed in the colloquial 26. Moulton, Prolegomena, 13. 27. This does not mean that there are not differences between written and spoken (H. Paul, Prinzipien der Sprachgeschichte [fourth edition], Halle 1909, 33-89) or literary and non-literary language. See Deissmann, “Hellenistic Greek,” 43. Both Moulton and Deissmann reject the notion of dialects within Hellenistic or Koine Greek (that is, broadly phonologically, morphologically and even syntactically differing varieties of the language). Cf. K. Dietrich, Untersuchungen zur Geschichte der griechischen Sprache von der hellenistischen Zeit bis zum 10. Jahrhundert nach Christus (Byzantinisches Archiv 1), Leipzig 1898, 306 ff., who differentiated dialects in Egypt, Asia Minor, and Greece. 28. One of the responses to the Deissmann and Moulton hypothesis was to note that the Greek of Alexandria had already come under some kind of Semitic or other influence. Those who argued for Semitic, primarily Hebrew, influence because of the large Jewish population in Egypt include: Ottley, Handbook, 165; G. Dalman, The Words of Jesus: Considered in the Light of Post-Biblical Jewish Writings and the Aramaic Language (translated by D. M. Kay), Edinburgh 1902, 17; J. C. James, The Language of Palestine and Adjacent Regions, Edinburgh 1920, 57-75; and C. F. Burney, The Aramaic Origin of the Fourth Gospel, Oxford 1922, 5-6; cf. also those below arguing for the Jewish Greek hypothesis. Others argued for the influence of Egyptian Coptic. These include: L.-T. Lefort, “Pour une grammaire des LXX” Le Muséon 41 (1928), 152160; J. Vergote, “Grec Biblique” in: L. Pirot (ed.), Dictionnaire de la Bible, Supplément, vol. 3, Paris 1938, columns 1320-1369, especially columns 1353-1360; F. T. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, vol. 1, Milan 1976, 46-48; F. T. Gignac, “The Language of the Non-Literary Greek Papyri” in: D. H. Samuel (ed.), Proceedings of the Twelfth International Congress of Papyrology (American Studies in Papyrology 7), Toronto 1970, 139152; and F. T. Gignac, “The Papyri and the Greek Language” Yale Classical Studies 28 (1985), 155-165. These hypotheses had already been addressed by Thumb and Moulton. See also S.-T. Teodorsson, The Phonology of Ptolemaic Koine (Studia Graeca et Latina Gothoburgensia 36), Gothenburg 1977, 25-35. 29. H. A. A. Kennedy, Sources of New Testament Greek, or The Influence of the Septuagint on the Vocabulary of the New Testament, Edinburgh 1895.

22

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

language of everyday life.” He admits that there are “Semitic characteristics,” but these are directly attributable to its being a literal translation from Hebrew. In this sense, it occupies a mediating position of reflecting Koine Greek and the original, while also creating a technical vocabulary necessary for the subject matter. This vocabulary had an influence upon the Greek of the New Testament, but “must not be exaggerated,” as the New Testament reached a “higher plane.” 30 T. K. Abbott, writing in 1891, had already arrived at somewhat similar conclusions regarding the relationship of the vocabulary of the New Testament to that of the Septuagint. 31 In what remains as one of the most important books written on the Greek of the Hellenistic period, Albert Thumb offered a thorough examination that has in many ways never been equaled. 32 Many, if not most, of Thumb’s conclusions match those of Deissmann and Moulton, in that he sees Koine Greek as the “natural development” of Attic Greek into an essentially dialectless language, or at least one that did not evidence the continued existence of the earlier classical dialects. Thumb also, however, emphasizes several other characteristics of the language. He believes that the languages with which Koine Greek came into contact, apart possibly from instances of some vocabulary items, did not have an influence upon the language. He further believes that there were local but not systematic variations in Koine Greek. In fact, he posits that, if one knew sufficiently well the local characteristics, one could identify the place of origin of a given Greek text. He further believes that knowledge of modern Greek, in that sense forming a trajectory of Greek development, could aid in understanding linguistic phenomena of an earlier period. 33 The Greek of the Septuagint, Thumb argues, was a local variety of koine, although it also reflects the fact that it was a translated document. In any case, it was not a special or particular kind of Jewish Greek. In fact, only in the Septuagint (not the New Testament) are features of Hebrew syntax identifiable, but even here numerous supposed instances can be accounted for as instances of Koine Greek. The second category of monographs dedicated to the Septuagint includes the introduction by Henry Barclay Swete. At the time that he wrote, there was no grammar dedicated to the Greek of the Septuagint (see the category below). 34 As a result, Swete 30. See Kennedy, Sources, 164, for a useful summary of his findings (and the source of the quotations). 31. T. K. Abbott, Essays, Chiefly on the Original Texts of the Old and New Testaments, London 1891, 65-109, on New Testament lexicography, especially 67. 32. A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus: Beiträge zur Geschichte und Beurteilung der Κοινή, Straßburg 1901; cf. also A. Thumb, “Hellenistic and Biblical Greek” in: J. Hastings (ed.), Dictionary of the Apostolic Church, vol. 1, Edinburgh 1915, 551-560. See also F. Büchsel, “Die griechische Sprache der Juden in der Zeit der Septuaginta und des Neuen Testaments” Zeitschrift für die alttestamentliche Wissenschaft 60 (1944), 132-149. 33. Thumb’s interest in modern Greek is reflected in the fact that he wrote a grammar of it. See A. Thumb, Handbook of the Modern Greek Vernacular: Grammar, Texts, Glossary (translated by S. Angus), Edinburgh 1912. See also A. Thumb, “On the Value of the Modern Greek for the Study of Ancient Greek” Classical Quarterly 8 (1914), 181-205. 34. H. B. Swete, An Introduction to the Old Testament in Greek, Cambridge 1902, 19-20. The revised edition of 1914 by R. R. Ottley is essentially the same, with the addition of a few footnotes with reference to later secondary literature, especially Thackeray’s grammar (see below). 2. Periods in the Study of the Greek of the Septuagint

23

1. History of Scholarship on the Language of the Septuagint

to some extent is feeling his way through the subject. Consequently, he ends up being somewhat caught between two different views of the Greek of the Septuagint (and may well not belong fully in this category). On the one hand, he recognizes the history of discussion of the Greek of the Septuagint. He summarizes the views expressed in the standard grammars of New Testament Greek, and recognizes the (then) recent work of Deissmann in regard to the documentary papyri. On the other hand, Swete also wishes to recognize some of what he perceives to be the unique characteristics of the Septuagint. 35 As a result, he sees the common Greek of Alexandria as a “mongrel patois,” the Alexandrian dialect, made up of elements of many of the earlier Greek dialects. However, he also recognizes that “[n]o monument of the Alexandrian ‘dialect’ remains, unless we may seek it in the earlier books of the Alexandrian Greek Bible.” 36 He sees this dialect as best represented in both the Septuagint and the New Testament, as forms of “colloquial Greek” used in Alexandria and Palestine. 37 This language was what was used by the translators of the Septuagint, as a type of local dialect. He especially sees evidence of this in the new meanings that are given to various lexical items under Hebraic influence. The third category includes the Septuagint grammars of Robert Helbing and Henry St. John Thackeray (though neither of them is a complete grammar of the Septuagint), and the grammar of Félix-Marie Abel. 38 The first of these volumes to appear, at least in part, was the first part of the grammar by Robert Helbing, on phonology and morphology, in 1907. The second part, on the syntax of cases and the verb, appeared in 1928. Neither together constitutes a complete grammar. Despite this, Helbing describes the Septuagint as a special source for Koine Greek and places it in the midst of the original works of what he characterizes as vulgar Greek. After surveying previous research, he acknowledges the great work of Deissmann in relating the language of the New Testament and the Septuagint to the papyri and inscriptions, as well as to Hellenistic literature. Because of Deissmann’s work, the language of the Greek Bible was removed from the isolation in which it had been located for decades, and shown to be written by common people in the language of their time, Koine Greek. Helbing recognizes that the Septuagint is, first, a translation from Hebrew, and then, second, a Koine Greek document. As for the first, following Thumb, he sees the supposed Hebraisms as understandable within the Hellenistic Greek of the time. As for the second, he places it within the developments that 35. This position comes out more strongly in his commentary on Revelation. See H. B. Swete, The Apocalypse of St John, London 1906, cxx, note 1. 36. Swete, Introduction, 292. 37. Swete, Introduction, 293. 38. R. Helbing, Grammatik der Septuaginta: Laut- und Wortlehre, Göttingen 1907; R. Helbing, Die Kasussyntax der Verba bei den Septuaginta: Ein Beitrag zur Hebraismenfrage und zur Syntax der Κοινή, Göttingen 1928; H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, vol. 1: Introduction, Orthography and Accidence, Cambridge 1909; F.-M. Abel, Grammaire du Grec Biblique suivie d’un choix de papyrus (Étude Biblique), Paris 1927. For a short review of these works, in the context of other syntactical works on the language of the Septuagint, see R. Sollamo, “Prolegomena to the Syntax of the Septuagint” in: R. Sollamo / S. Sipilä (eds.), Helsinki Perspectives on the Translation of the Septuagint, Helsinki/Göttingen 2001, 23-41, especially 27.

24

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

occurred from the time of Alexander the Great. 39 His second volume continues with a similar perspective, while entering into discussion of Hebraic features. Soon after the appearance of Helbing’s grammar, in 1909, Henry St. John Thackeray published the first volume of his grammar of the Septuagint on matters of introduction, orthography, and accidence (in other words, with much the same scope as Helbing’s first volume). Despite not writing a complete syntax of Septuagint Greek— note, however, that he does include a brief syntax within his discussion of Koine Greek —Thackeray’s volume is more valuable than Helbing’s because of the introduction, which includes a discussion of Koine Greek as the “basis” of Septuagint Greek and an extensive treatment of Semitisms (the introduction also contains Thackeray’s valuable assessment of the types of translational styles found in the Septuagint). Thackeray argues that the “Septuagint, considered as a whole, is the most extensive work which we possess written in the vernacular of the κοινή or Hellenistic language, and is therefore of primary importance for a study of later Greek, and the main function of a grammar of LXX Greek is to serve as a contribution to the larger subject, the grammar of the κοινή.” 40 He attributes this view to the knowledge gained through study of the Egyptian papyri. While recognizing that the Septuagint, as a translated document, “has naturally a Semitic coloring,” he downplays its significance because of the appearance in the papyri of similar features. 41 Instead, he paints a picture of the Greek of the time that relies heavily upon the views of Thumb regarding the general development and description of Koine Greek, and Moulton regarding the question of Semitisms. Thackeray’s discussion of Semitisms includes extended treatment of individual elements where he attempts to show that they are paralleled by extra-biblical examples. A similar position was argued for by Jean Psichari, who wrote a lengthy article from the Koine Greek perspective, describing the Greek of the Septuagint. 42 Abel’s grammar is not specifically of the Septuagint, but is of Biblical Greek as illustrated by examples from the papyri, and liberally illustrated with references from the Septuagint. The formulation of the work itself evidences its approach. From the outset, Abel characterizes Koine Greek as the development of Greek following on from Alexander, based upon Attic, and replacing the ancient dialects (he disputes an Alexandrian dialect). He minimizes the influence of local dialects, and follows Thumb, Thackeray, and Moulton regarding Semitisms in the Septuagint. 43 From the earliest twentieth century to the present, there have been a number of histories of the Greek language written that tend to follow the koine hypothesis regarding the Greek of the Septuagint. 44 Procope Costas says that both “the Septuagint

Helbing, Grammatik, i-v. Thackeray, Grammar, 16. Thackeray, Grammar, 16. J. Psichari, “Sur le Grec de la Septante” Revue des études juives 5 (1908), 161-208. See also H. Pernot, “Observations sur la langue de la Septante” Revue des études grecques 42 (1929), 411-425. 43. Abel, Grammaire, xvii-xl. 44. Among those writing in dictionaries, see E. C. Colwell, “Greek Language” in: G. A. Buttrick (ed.), The Interpreter’s Dictionary of the Bible, vol. 2, Nashville, TN 1962, 479-487, especially 484-485. 39. 40. 41. 42.

2. Periods in the Study of the Greek of the Septuagint

25

1. History of Scholarship on the Language of the Septuagint

and the New Testament are monuments of the vernacular Koine constituting, with Epictetus, the most important literary sources of our knowledge of the popular language of the period,” although he nuances this position in regard to linguistic registers. 45 Robert Browning lists the Septuagint as one of the five major sources for our knowledge of Koine Greek. 46 More cautious in recognizing Semitisms, though accounting for them in light of the sacred nature of the Old Testament, is L. R. Palmer. 47 Geoffrey Horrocks states that the “Greek translation of the Old Testament […] constitutes one of our most important examples of surviving ‘vernacular’ literature of the period,” following the work of Thackeray. 48 F. R. Adrados seems to hold to a similar position, while recognizing debate over Hebraisms and Semitisms. 49 Finally, Nicholas de Lange rejects any notion of a “distinct Jewish dialect of Greek,” and places Septuagint Greek (especially of the Pentateuch) within the ambit of authors “with a perfectly sound command of Greek.” 50 The Hellenistic/Koine Greek position holds that Hellenistic Greek was the natural result of linguistic development and was essentially a single variety used throughout the Mediterranean world of the time. The Greek of the Septuagint, not to be seen as a unitary entity, was a form of translational Greek but remained recognizably a type of Koine Greek, which can be demonstrated through comparison of significant examples, whereby supposed instances of Semitic influence can for the most part be seen as enhancements of constructions already found in Koine Greek. 51

2.3 Hebraic/Jewish Greek Hypothesis Despite the strength of opinion that swept through Hellenistic Greek studies in the first part of the twentieth century and that came to dominate the first half of the century, there was for some the residual issue that the Greek of the Septuagint was in particular and definable ways different from the Greek of other koine authors, even when the issue of translation was taken into account. In some circles, this resulted in a

45. P. S. Costas, An Outline of the History of the Greek Language with Particular Emphasis on the Koine and the Subsequent Periods, Chicago, IL 1937 (reprint Chicago, IL 1997), 41-71, here 55. 46. R. Browning, Medieval and Modern Greek, London 1969, 30. Cf. his “Von der Koine bis zu den Anfängen des modernen Griechisch” in: H.-G. Nesselrath (ed.), Einleitung in die griechische Philologie, Stuttgart 1997, 156-168. 47. L. R. Palmer, The Greek Language, London 1980 (reprint London 1995), 196. 48. G. Horrocks, Greek: A History of the Language and its Speakers, London 1999, 56-59, here 56. 49. F. R. Adrados, A History of the Greek Language: From Its Origins to the Present (translation Francisca Rojas del Canto), Leiden 2005 (1999), 186-187. 50. N. de Lange, “Jewish Greek” in: A.-F. Christidis (ed.), A History of Ancient Greek: From the Beginnings to Late Antiquity, Cambridge 2007 (2001), 638-645, here 640. Cf. C. H. George, “Jewish and Christian Greek” in: E. J. Bakker (ed.), A Companion to the Ancient Greek Language, Oxford 2010, 167-180, who emphasizes Semitic interference. 51. See Porter, Verbal Aspect, 118, where he distinguishes between direct translation, intervention, and enhancement, arguing that only intervention can be considered a Semitism. This distinction is similar to the one found in J. H. Moulton / W. F. Howard, Accidence and Word-Formation, vol. 2 of J. H. Moulton, A Grammar of New Testament Greek, Edinburgh 1929, 16.

26

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

re-evaluation of the Koine Greek hypothesis and the re-assertion of the Septuagint as a form of Hebraic or Jewish Greek. Advocates of this position argued that this Jewish Greek was found not only in the Septuagint, but possibly in other ancient documents written in Greek by Jewish authors (including the New Testament)—to the point of positing, at least by some, a particular type of Jewish Greek dialect. The two major early proponents of this view were Henry Gehman and Nigel Turner, followed also by Matthew Black. More recently, they have been succeeded for the most part, though with their own emphases and variations, by the Finnish school of Septuagint studies and Georg Walser. In 1951, Gehman published a relatively short article on what he called the “Hebraic Character of Septuagint Greek.” 52 Gehman begins from the premise that the “object of a translator obviously is to render a document clearly into the vernacular.” 53 While acknowledging that Septuagint Greek is Koine Greek, he also notes that there are difficulties in reading the Greek without using the Hebrew and that it differs from the koine at various points. Although not wanting to re-introduce the idea of a “JewishGreek jargon,” Gehman does wish to dispute Thackeray and speak of a “Jewish-Greek, which was in use in the synagogues and in religious circles.” To illustrate his point, he marshals examples from Septuagint Greek: paratactic καί, other types of clauses, uses of the article (e. g. indicating the direct object), use of the preposition ἐν and other prepositions, the participle for the infinitive absolute, use of προστίθημι and other similar expressions, various explicatives, and some other phrases. However, he then goes on to cite fifteen Greek words that are “adapted to OT usage.” He concludes that “it is clear that LXX Greek has numerous cases of grammar and vocabulary which are Hebraic,” and that this language “would have caused trouble to a Greek who was not acquainted with the psychology of the Hebrew language, its idioms, and its construction.” This is because “the translators had no intention of making a book to be used for textual studies.” However, “we may suppose that its language made sense to Greekspeaking Jews.” Gehman repeats that he does not want to claim that there was “a Jewish-Greek jargon, but there was a Greek with a pronounced Semitic cast that was used and understood in religious circles. If the LXX made sense to Hellenistic Jews, we may infer that there was a Jewish Greek which was understood apart from the Hebrew language.” 54 Gehman was soon followed by Nigel Turner. Whereas Gehman focused almost exclusively upon the Septuagint, Turner dealt with both the Septuagint and, even more so, the New Testament. Turner wrote a short article on what he called the “unique character of Biblical Greek” in response to the first two of Gehman’s Vetus Testamentum articles. In this article, Turner cites as examples placement of ἕνεκα and use of πᾶς and a substantive, purporting to show that Septuagint usage deviates statistically 52. H. S. Gehman, “The Hebraic Character of Septuagint Greek” Vetus Testamentum 1 (1951), 8190 (reprint in: Porter, Language of the New Testament, 163-173—from which I quote). A similar opinion is expressed in H. S. Gehman, “Hebraisms of the Old Greek Version of Genesis,” Vetus Testamentum 3 (1953), 141-148; and in H. S. Gehman, “Ἅγιος in the Septuagint, and its Relation to the Hebrew Original” Vetus Testamentum 4 (1954), 337-348. 53. Gehman, “Hebraic Character,” 163. 54. Gehman, “Hebraic Character,” 172-173. 2. Periods in the Study of the Greek of the Septuagint

27

1. History of Scholarship on the Language of the Septuagint

from extra-biblical Greek usage. 55 He cites occasional instances of Septuagint usage in others of his works, though the bulk of them is devoted to the Greek New Testament. 56 There is very little new in the arguments of Gehman or Turner. 57 The evidence that they cite as evidence of the distinctive character of Biblical Greek was already acknowledged, examined, and explained by Deissmann, Moulton, Thumb, and others. This is especially true with regard to the Septuagint. Whatever one may think of the discussion regarding the Greek of the New Testament, these scholars had already admitted the fact that the Septuagint was a translation and that, as a translation of Hebrew, it contained constructions that in some way reflected the substrate Hebrew language. The major difference seems to be the overall linguistic framework in which the evidence is placed. Deissmann, Moulton, Thumb, and others took their entry point as the Greek language and traced a trajectory in the development of Greek from classical to post-classical times. They then located the Greek of the Septuagint (as well as that of the New Testament) in relation to this trajectory. Post-classical Greek in its widely used form, the koine, became the type of Greek of which the Greek of the Septuagint was a translational variety. For Gehman and Turner, however, their point of departure appears to have been the Hebrew Bible, or perhaps the Jewish synagogue. Rather than beginning with continuity and inclusion, they stressed discontinuity and separation. They desired to treat Biblical Greek, in particular Septuagint Greek, as a linguistic attempt in Greek to capture the Hebrew Bible. The result was a non-idiomatic Greek created for religious purposes. 58 55. N. Turner, “The Unique Character of Biblical Greek” Vetus Testamentum 5 (1955), 208-213. 56. Those with some significant reference to the Septuagint include: “The ‘Testament of Abraham’ : Problems in Biblical Greek” New Testament Studies 1 (1954-1955), 219-223; “The Relation of Luke I and II to Hebraic Sources and to the Rest of Luke-Acts” New Testament Studies 2 (1955-1956), 100-109; “The Language of the New Testament” in: M. Black / H. H. Rowley (eds.), Peake’s Commentary on the Bible, London 1963, 576-579; “Jewish and Christian Influence in the New Testament Vocabulary” Novum Testamentum 16 (1974), 149-160; Syntax, vol. 3 of J. H. Moulton, A Grammar of New Testament Greek, Edinburgh 1963, especially 2-9; “Second Thoughts. VII. Papyrus Finds” The Expository Times 76 (1964-1965), 44-48; “The Language of Jesus and His Disciples” in his Grammatical Insights into the New Testament, Edinburgh 1965, 174-188 (reprint in: Porter, Language of the New Testament, 174-190); and Christian Words, Edinburgh 1981. In this regard, Turner and Gehman are followed by D. Hill, Greek Words with Hebrew Meanings (SNTS Monograph Series 5), Cambridge 1967, especially 16; cf. also E. Tov, “Greek Words and Hebrew Meanings” in: T. Muraoka (ed.), Melbourne Symposium on Septuagint Lexicography (Septuagint and Cognate Studies 28), Atlanta, GA 1990, 83126. 57. In much discussion, the issues raised here are automatically transferred over to issues regarding the translation of the Septuagint itself. This does not necessarily follow, as was ably indicated by Chaim Rabin in “The Translation Process and the Character of the Septuagint” Textus 6 (1968), 1-26, especially 25. For an examination of the study of translation technique, see Tov, “Nature and Study,” 337-359. 58. I cannot help but notice that there are similarities between the perspective of Gehman and others and the Biblical Theology movement, especially the distinctive Hebraic mentality. See B. S. Childs, Biblical Theology in Crisis, Philadelphia, PA 1970, 13-50. To a great extent, this was what James Barr objected to in his The Semantics of Biblical Language, Oxford 1961. Despite Barr’s criticism, there are those who continue to defend the Semitic mind and the work found in Kittel and Friedrich’s Theological Dictionary, even for Septuagint studies.

28

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

This viewpoint, nevertheless, has gained momentum during the second half of the last century and into this century. Matthew Black, who devoted most of his efforts to discussing Semitisms in the Greek New Testament, 59 clearly inclined toward this position. When examining the biblical languages, he notes instances where the papyri cannot account for biblical usage that is to be reckoned in light of “the fundamentally Semitic ways of thought, impressed on language and idiom.” 60 He cites Turner, Gehman, and Klaus Beyer 61 as advocates for such a position. He concludes that there was a “Jews’ Greek” found in the Septuagint and spoken in the synagogue as a type of Koine Greek: “this language, like the Hebrew of the Old Testament which molded it, was a language apart from the beginning; biblical Greek is a peculiar language, the language of a peculiar people.” 62 From the 1960s to the present, there has been significant work on the Septuagint done in Finland by a number of scholars, to the point of there being an identifiable Finnish school of Septuagint studies. Their work often focuses upon particular syntactical phenomena (e. g. a particular conjunction or tense-form). In fact, to a large extent their work is on translation technique, rather than the question of the Greek language itself, although they seem to draw such implications from their work. 63 Whereas those within this school do not go so far as to identify a particular Jewish Greek, they are convinced that the Greek of the Septuagint cannot be explained simply by comparison with Koine Greek, in particular regarding its Hebraisms. These Hebraisms loom large in the discussion by the Finnish scholars, as if their major agenda is to attempt to refute the Koine Greek hypothesis by means of accumulating individual examples. Ilmari Soisalon-Soininen cautions about the difficulties of the “hebräische Hintergrund,” which makes it extremely difficult to compare books of the Septuagint to those written in non-biblical Greek. 64 In fact, he summarizes the situation in this way: “Die 59. See M. Black, An Aramaic Approach to the Gospels and Acts (third edition), Oxford 1967 (1946). 60. M. Black, “The Biblical Languages” in: P. R. Ackroyd / C. F. Evans (eds.), The Cambridge History of the Bible, vol. 1, From the Beginnings to Jerome, Cambridge 1970, 1-11, here 8. See also his “Second Thoughts. IX. The Semitic Element in the New Testament” The Expository Times 77 (1965-1966), 20-23, with similar language. 61. See K. Beyer, Semitische Syntax im Neuen Testament (second edition), vol. 1, Göttingen 1968. 62. Black, “Biblical Languages,” 11 (similarly his “Second Thoughts,” 23), citing Turner, Syntax, 9: “[…] the strongly Semitic character of Biblical Greek, and therefore its remarkable unity within itself, do seem to me to have contemporary significance at a time when many are finding their way back to the Bible as a living book and perhaps are pondering afresh the old question of a ‘Holy Ghost Language.’” These comments, besides reflecting the Biblical Theology movement, are also suspect in terms of linguistic determinism. 63. See R. Sollamo, “Translation Technique as a Method” in: H. Ausloos et al. (eds.), Translating a Translation: The LXX and Its Modern Translations in the Context of Early Judaism (BETL 213), Leuven 2008, 35-41, especially 35. I question whether Sollamo is correct in stating that “in order to be able to write a syntax of the Septuagint, it is necessary to study first the translation technique followed by the translators in rendering different syntactic items and phenomena of the source language” (p. 35). This contradicts the notion of language as system. 64. I. Soisalon-Soininen, Die Infinitive in der Septuaginta (Annales Academiae Scientiarum Fennicae B 132,1), Helsinki 1965, 7-15 (reprint “Einleitung,” in his Studien zur Septuaginta-Syntax: Zu seinem 70. Geburtstag am 4. Juni 1987 [edited by A. Aejmelaeus and R. Sollamo, Suomalai2. Periods in the Study of the Greek of the Septuagint

29

1. History of Scholarship on the Language of the Septuagint

Sprache der Septuaginta ist in ziemlich großem Maße Hebräisch mit griechischen Wörtern.” 65 Raija Sollamo states that, “[a]t times, however, evidence from the Koine is inadequate in deciding whether an expression or a construction is Hebraistic or not.” As a result, “where no Greek parallel is available I regard the LXX expression as Hebraistic,” although she recognizes that new discoveries in the papyri might change this situation. 66 Recently, Anssi Voitila has made a number of studies of the Greek tense-forms, in an attempt to establish the relationship between Hebrew and Greek usage. He contends that there are sufficient examples—although he rejects the simple use of statistical analysis—to argue that there is both an unpredictable relationship between Hebrew and Greek usage and that the resulting Greek usage, contrary to scholars such as James Barr, 67 is not idiomatic Greek. He characterizes the results as reflecting a “stereotyping tendency.” He has done such studies on the present, the imperfect, and the perfect tense-forms. His analysis seems to be based primarily upon a time-based conception of both the Hebrew and the Greek verbal systems. 68 This posi-

65.

66.

67. 68.

sen Tiedeakatemian Toimituksia Annales Academiae Scientiarum Fennicae B 237], Helsinki 1987, 11-18, here 12). I. Soisalon-Soininen, “Methodologische Fragen der Erforschung der Septuaginta-Syntax” in: Cox, VI Congress, 425-444 (reprint in: Studien zur Septuaginta-Syntax, 40-52, here 42). See also his “Zurück zur Hebraismenfrage” in: D. Fraenkel / U. Quast / J. W. Wevers (eds.), Studien zur Septuaginta—Robert Hanhart zu Ehren: Aus Anlass seines 65. Geburtstages (Abhandlungen der Akademie der Wissenschaften in Göttingen Philologisch-historische Klasse dritte Folge 190, Mitteilungen des Septuagint-Unternehmens [MSU] 20), Göttingen 1990, 35-51. This view is similar to that of D. Tabachovitz, Die Septuaginta und das Neue Testament: Stilstudien, Lund 1956. R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (Annales Academiae Scientiarum Fennicae Dissertationes Humanarum Litterarum 19), Helsinki 1979, 4 and 5. See also A. Aejmelaeus, Parataxis in the Septuagint: A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch (Annales Academiae Scientiarum Fennicae Dissertationes Humanarum Litterarum 31), Helsinki 1982, 1. Cf. A. Aejmelaeus, On the Trail of the Septuagint Translators: Collected Essays (Contributions to Biblical Exegesis & Theology 50, revised edition), Leuven 2007 and R. Sollamo, “Why Translation Technique and Literalness Again? The Renderings of the Infinitive Absolute in the Septuagint of Jeremiah” in: M. Nissinen (ed.), Congress Volume Helsinki 2010 (VTSup 148), Leiden 2012, 1-20; among other articles. However, Sollamo seems to take a step back from this position and acknowledge the koine hypothesis in R. Sollamo, Repetition of the Possessive Pronouns in the Septuagint (Septuagint and Cognate Studies 40), Atlanta, GA 1995, 2-3; and “The Koine Background for the Repetition and Non-Repetition of the Possessive Pronoun in Co-Ordinate Items” in: Fraenkel / Quast / Wevers, Studien zur Septuaginta, 52-63. J. Barr, “Translators’ Handling of Verb Tense in Semantically Ambiguous Contexts” in: Cox, VI Congress, 381-403. See A. Voitila, “La Technique de traduction du Yiqtol (l’imparfait hébreu) dans l’Histoire du Joseph grecque (Gen 37,39-50)” in: C. E. Cox (ed.), VII Congress of the International Organization for Septuagint and Cognate Studies, Leuven, 1989 (Septuagint and Cognate Studies 31), Atlanta, GA 1991, 223-237; “Some Remarks on the Perfect Indicative in the LXX” Bulletin of the International Organization for Septuagint and Cognate Studies 26 (1993), 11-16; “What the Translation of Tenses Tells About the Septuagint Translators” Scandinavian Journal of the Old Testament 10.2 (1996), 183-196; “The Translator of the Greek Numbers,” in: B. A. Taylor (ed.), IX Congress of the International Organization for Septuagint and Cognate Studies, Cambridge,

30

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

tion has been maintained by a number of other Septuagint scholars as well, including others influenced by the Finnish perspective. 69 A very detailed study using a variety of statistical methods has been performed by Georg Walser. 70 Walser is not necessarily attempting to revitalize the Jewish-Greek hypothesis as it was promoted by Gehman and Turner. Nevertheless, he does argue that there was a variety of Greek that was heavily influenced by the Septuagint translation of the Pentateuch and that was used, as one of several varieties of Greek, within the synagogue. 71 Walser’s study is distinguished from most others, especially those that preceded him and attempted to defend various types of Semitic Greek, by the rigor of his methodology. He first defines two categories of documents for study: those 1995 (Septuagint and Cognate Studies 45), Atlanta, GA 1997, 109-121; Présent et imparfait de l’indicatif dans le pentateuque grec: Une étude sur la syntaxe de traduction (Publications de la société d’exégèse de Finlande 79), Helsinki/Göttingen 2001; “The Perfect Indicative in the Greek Pentateuch and the Hebrew Qatal” Studia Orientalia 99 (2004), 415-423; “La Septante: un document linguistique de la koiné grecque antique?” in: P. Le Moigne (ed.), L’apport de la septante aux études sur l’antiquité: Actes du colloque de Strasbourg 8-9 novembre 2002, Paris 2005, 17-35. Besides the problem with the time-based view of the tense-forms (which Voitila recognizes not all would agree with for Hebrew), there are other major problems in his work (e. g. characterizing systemic linguistics as a type of transformational generative grammar in Voitila, Présent, vi and note 7). 69. See, for example, R. A. Martin, “Some Syntactical Criteria of Translation Greek” Vetus Testamentum 10 (1960), 295-310; J. Lust, “Introduction” in LEH, vol. 1, i-xv, especially viii-ix; J. Lust, “Syntax and Translation Greek” Ephemerides Theologicae Lovanienses 77 (2001), 395-401; and, apparently, J. Joosten, “The Vocabulary of the Septuagint and its Historical Context” and “Le Vocabulaire de la Septante et la Question du Sociolecte des Juifs Alexandrins: Le Cas du Verbe εὐλογέω, ‘Bénir’” in: J. Joosten / E. Bons (eds.), Septuagint Vocabulary: Pre-History, Usage, Reception (Septuagint and Cognate Studies, 58), Atlanta, GA 2011, 1-11 and 13-23. 70. G. Walser, The Greek of the Ancient Synagogue: An Investigation on the Greek of the Septuagint, Pseudepigrapha and the New Testament (Studia Graeca et Latina Lundensia 8), Stockholm 2001. This work is summarized in G. Walser, “The Greek of the Ancient Synagogue” in: B. Olsson / M. Zetterholm (eds.), The Ancient Synagogue from Its Origins until 200 C.E.: Papers Presented at an International Conference at Lund University, October 14-17, 2001 (Coniectanea Biblica New Testament 39), Stockholm 2003, 260-276. See also G. Walser, “A Peculiar Word-Order Rule for the Septuagint and for Cognate Texts” in: B. A. Taylor (ed.), X Congress of the International Organization for Septuagint and Cognate Studies, Oslo, 1998 (Septuagint and Cognate Studies 51), Atlanta, GA 2001, 499-511; “Die Wortfolde der Septuaginta” in: M. Karrer / W. Kraus (with M. Meiser) (eds.), Die Septuaginta—Texte, Kontexte, Lebenswelten: Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 20.–23. Juli 2006 (WUNT 219), Tübingen 2008, 258-266. 71. Walser shows the influence of a Scandinavian interest in the synagogue as the linguistic center of Jewish usage of Greek (as well as the synagogue being important for this view of Septuagint Greek). See A. Wifstrand, “Stylistic Problems in the Epistles of James and Peter” Studia Theologica 1 (1947), 170-182 (reprint in: A. Wifstrand, Epochs and Styles: Selected Writings on the New Testament, Greek Language and Greek Culture in the Post-Classical Era [edited by L. Rydbeck and S. E. Porter, translated by Denis Searby, WUNT 179], Tübingen 2005, 46-58); cf. also A. Wifstrand, “Lukas och Septuaginta” Svensk Teologisk Kvartalskrift 16 (1940), 243262 (translated as “Luke and the Septuagint” in: Wifstrand, Epochs and Styles, 28-45); and Olsson / Zetterholm, Ancient Synagogue, which contains five essays on language (including the one by Walser noted above). 2. Periods in the Study of the Greek of the Septuagint

31

1. History of Scholarship on the Language of the Septuagint

connected with the synagogue and those not. He then performs a number of syntactical studies of this corpus of material. As he points out, rather than starting with later Greek and attempting to account for it in relation to the Pentateuch, he begins with the Pentateuch as his norm, and then compares all of the other Greek in his corpus with it. As a result, he concludes that “the language of the Pentateuch served as some kind of a model for subsequent texts written in the same genre as the Pentateuch, intended for a similar audience and dealing with the same subject-matter. This peculiar ‘Pentateuchal’ variety of Greek was used within the context of the synagogue side by side with other varieties in use in the Greek-speaking world at the time of the ancient synagogue.” 72 He accounts for the differences in use of these varieties on the basis of diglossia, in which, in a polyglossic environment, there are varieties of high and low dialects. His use of modern linguistic categories is a step forward in the discussion, although there are possible other explanations of his evidence, which, despite its corpus construction, has limitations. 73 The Hebraic hypothesis holds that a blending of Greek and Hebrew is a linguistically recognizable phenomenon that occurred in Egypt beginning in the third century BCE, and that clear instances of such a combination can be detected in the documents available and can only be adequately explained by such a theory.

2.4 Revival of the Koine Greek Hypothesis In 1980, there was a revival of the Koine Greek hypothesis. In one sense, this was not a revival, in that many had continued to hold to the Koine Greek hypothesis, as is evidenced in the histories of Greek written throughout last century. Nevertheless, there was an increase in scholarly justification of the traditional Deissmann and Moulton perspective, but along new lines of defense. This revival is characterized by recognition of features of Semitic syntax and vocabulary in the Greek of the Septuagint, but within the framework of it being a translated document within the wider world of Koine or Hellenistic Greek. This period is also marked by an increased attention to framing the discussion in modern linguistic terms. In 1980, in Biblica, Moisés Silva published an article on bilingualism in which he directly addressed the issues raised by Jean Vergote in opposition to the Deissmann hypothesis. 74 Silva characterizes the strong difference of opinion between Vergote and Thumb over the influence of native Semitic languages on users as a failure to distinguish between two fundamental linguistic concepts, langue (or language) and parole (or speech). 75 He believes that the issues involved are many and complex (he lists 72. Walser, Greek, 173. 73. S. Wahlgren, “The Languages of the Synagogue: An Evaluation” in: Olsson / Zetterholm, Ancient Synagogue, 298-302. 74. M. Silva, “Bilingualism and the Character of Palestinian Greek” Biblica 61 (1980), 198-219 (reprint in: Porter, Language of the New Testament, 205-226). See also M. Silva, “New Lexical Semitisms?” Zeitschrift für die neutestamentliche Wissenschaft 69 (1978), 253-257; M. Silva, Biblical Words and their Meaning: An Introduction to Lexical Semantics, Grand Rapids, MI 1983, especially 53-73; M. Silva / K. H. Jobes, Invitation to the Septuagint, Grand Rapids, MI 2000, 105-118. 75. These are fundamental concepts in modern linguistics, usually attributed to Ferdinand de

32

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

twelve of them), and he determines to set some of them straight. Silva also believes that Deissmann and his followers had not been fairly represented, and that many of the responses to him—such as Vergote’s appeal to the influence of Egyptian languages on the papyri (more currently made by Gignac), 76 the Scandinavian synagogue argument, and the Jewish Greek hypothesis—had already been adequately addressed or can be addressed through modern linguistics. This includes especially a refined notion of dialect and realization of issues surrounding bilingualism. In addressing the situation of Alexandrian bilingualism, Silva directly responds to Gehman, disputing his approach and his findings. For example, he rejects Gehman’s characterization of “Septuagint Greek” as a unified entity and the idea that only reference to Hebrew can resolve particular linguistic issues. Most of all, he rejects the idea of any kind of a transitional language between Hebrew and Greek on the basis of Alexandrian bilingualism. For Silva, this all finds resolution in the concepts of langue and parole. He believes that much of the discussion, especially by those who argue for a Semitic Greek hypothesis, is of style, or parole, whereas Deissmann and Moulton were concerned with langue, that is the grammar of the language. 77 A number of works have continued to develop this perspective. I will discuss them under the categories of monographs, general studies, and grammatical works. In a thorough assessment of the Jewish Greek hypothesis, originally finished in 1970 but only published in 1983, John A. L. Lee finds the Deissmann and Thackeray view sound. 78 He believes that there has been no adequate response to the appearance of Semitic Greek only in translations (it would be interesting to hear his response to Walser’s work) and confusion over Hebraisms and Aramaisms (they tend to be lumped together). In 1989, I surveyed the various positions, including an analysis of the major proposals for Semitisms in the Greek Bible (applying my view of verbal aspect in a number of instances). 79 I respond to Gehman’s hypothesis by questioning both whether the object of a translator is to render the source text into the vernacular and whether the Septuagint translators always even understood their original text. Rather than positing a Jewish Greek, as do Gehman and others, I believe that such is a non sequitur and that

76. 77.

78.

79.

Saussure. See F. de Saussure, Course in General Linguistics (edited by C. Bally and A. Sechehaye, with A. Reidlinger, translated by W. Baskin), London 1959, 7-17. Gignac, Grammar, vol. 1, 46-48. He is followed, to a large extent (although at places he appears to be confused over issues, such as appealing to Wifstrand), by S. Oloffson, The LXX Version: A Guide to the Translation Technique of the Septuagint (Coniectanea Biblica Old Testament 30), Stockholm 1990, 33-40. J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (Septuagint and Cognate Studies 14), Chico, CA 1983, 11-19. Cf., however, J. Joosten / P. J. Tomson (eds.), Voces Biblicae: Septuagint Greek and its Significance for the New Testament (Contributions to Biblical Exegesis & Theology 49), Leuven 2007, especially vii-viii. Lee has a study within this volume (“Ἐξαποστέλλω,” 99-113), but not all of the essays are written from this perspective. Porter, Verbal Aspect, 113-156, especially 145-147, 154-155. See also S. E. Porter, “The Functional Distribution of Koine Greek in First-Century Palestine” in: S. E. Porter (ed.), Diglossia and Other Topics in New Testament Linguistics (JSNTSup 193), Sheffield 2000, 53-78 (and other essays in this volume). Cf. also J. Frösén, Prolegomena to a Study of the Greek Language in the First Centuries A.D.: The Problem of Koiné and Atticism, Helsinki 1974. 2. Periods in the Study of the Greek of the Septuagint

33

1. History of Scholarship on the Language of the Septuagint

we must recognize, as did Deissmann, Thackeray, and others, that there are common linguistic elements of even translation Greek. 80 Drawing upon sociolinguistics, I contend that there has been a failure to define “dialect” or “language,” to the extent that significant irregularities lead some to identify a unique dialect or language. I instead differentiate between code and text (somewhat similar to Silva’s use of langue and parole). Deissmann identifies the Septuagint, as well as the Greek of the New Testament, as sharing the code of Koine Greek, even though there may be various manifestations of this code in a variety of texts, such as the Septuagint, on the basis of such factors as register variance. I further note that within the environment of Alexandrian multilingualism, the prestige language was Koine Greek, as the lingua franca of the dominant political and economic powers. In such a linguistic environment, even though lexical transfer might occur from the secondary (Hebrew) to the primary (Greek) language to fill particular religious-terminological needs, syntactical linguistic interference would move from the primary to the secondary language, as is evidenced in development of the Hebrew language. 81 At the same time, in 1989, Greg Horsley published a direct rebuttal of the Jewish Greek hypothesis, entitled “The Fiction of ‘Jewish Greek.’” 82 Directly focused upon the work of Turner and Steven Thompson, 83 Horsley takes a linguistic approach by discussing the nature of bilingualism, including such issues as interference, diglossia, code-switching, and dialect. He then applies these to the languages of the first century, especially in Palestine, where he distinguishes between preferred and first languages, primary and secondary bilingualism, and receptive and productive bilingualism. On the basis of these categories, he examines the notion of Semitisms, and concludes that “the edifice of Jewish Greek lacks foundation in reality, neither does it have any cogent linguistic framework.” 84 Most current introductions to the Septuagint that discuss the matter of language generally conclude with some form of the Deissmann and Moulton koine hypothesis —often summarized in terms of the original formulation rather than appropriating more recent linguistic analysis of the issue. In a brief treatment more concerned with bibliography than detailed analysis, Sidney Jellicoe overall endorses the koine hypothesis, but makes clear that he wishes to avoid what he believes are its extremes. 85 As a 80. Porter follows J. W. Olley, “Righteousness” in the Septuagint of Isaiah: A Contextual Study (Septuagint and Cognate Studies 8), Missoula, MT 1979, 11. 81. See C. Rabin, “Hebrew and Aramaic in the First Century” in: S. Safrai / M. Stern (eds.), Compendia Rerum Judaicarum ad Novum Testamentum, Section I, The Jewish People in the First Century, vol. 2, Assen 1976, 1007-1039, here 1024. 82. G. H. R. Horsley, New Documents Illustrating Early Christianity, vol. 5, Linguistic Essays, Sydney 1989, 5-40, with extensive bibliographical (though often incomplete) references. See also G. H. R. Horsley, “Divergent Views on the Nature of the Greek of the Bible” Biblica 65 (1984), 393-403. 83. S. Thompson, The Apocalypse and Semitic Syntax (SNTS Monograph Series 52), Cambridge 1985. For a direct response to Thompson, see S. E. Porter, “The Language of the Apocalypse in Recent Discussion” New Testament Studies 35.4 (1989), 582-603, who also surveys proposals regarding the Greek of Revelation. 84. Horsley, New Documents, 40. 85. S. Jellicoe, The Septuagint and Modern Study, Oxford 1968, 329-337; cf. C. F. D. Moule, An

34

2. Periods in the Study of the Greek of the Septuagint

1. History of Scholarship on the Language of the Septuagint

result, he includes a qualified endorsement of the contribution of James Barr in limiting some of the extremes found in the Jewish-Greek hypothesis. Much better is the chapter on the language of the Septuagint in La Bible grecque des Septante written by Marguerite Harl, Gilles Dorival, and Olivier Munnich. This full discussion has two concise pages (plus bibliography) on the nature of the Greek of the Septuagint. The authors say that Deissmann caused a “véritable revolution,” and essentially support his position—recognizing that in the course of the century scholars have identified lexical and syntactical Semitisms but that there is no Jewish Greek. 86 Julio Trebolle Barrera takes a very similar position, stating the findings of others. 87 Natalio Fernández Marcos offers one of the best summaries of the debate regarding the Greek of the Septuagint, starting from the early church and moving to the relatively recent past—although he includes the discussion of the Greek of the New Testament within his survey and thus somewhat obscures some of the distinct issues for the Septuagint. Nevertheless, he concludes that Deissmann’s proposal moved discussion of the language of the Septuagint beyond the earlier Purist/Hebraist days, even if more work remains to be done. He identifies this work as recognizing that we know relatively little about the koine, appreciating the role of bilingualism within the koine environment, and noting the need to move beyond discussion of Jewish Greek. 88 Finally, Jennifer Dines provides a useful summary of the issues. She acknowledges the importance of the Koine Greek hypothesis, but also recognizes that “there is still considerable disagreement as to how far Hebrew idiom has affected LXX Greek,” with some arguing for its influence being “all-pervasive” and others that it is relatively infrequent—and that this is an issue that will continue to be discussed. 89 It has often been noted that Septuagint studies suffers from the lack of a dedicated grammar of the Greek of the Septuagint. Even though several have been started (noted above), no major one has been completed. The situation persists. However, in recent years, there have been a number of grammatical studies dedicated to the Greek of the Septuagint, especially the question of its relation to non-biblical Greek and possible Semitic influence. In his introduction to the Septuagint that includes a section on grammar, Mario Cimosa, even though he wishes to see God speak through the Septuagint, also sees him using Koine Greek. His short conspectus of Septuagint grammar is based on Conybeare and Stock and Thackeray. 90 Lino Cignelli and Rosario Pierri simply accept the koine hypothesis in their brief concordance of the syntax of the Septua-

86. 87.

88. 89. 90.

Idiom Book of New Testament Greek (second edition), Cambridge 1959, 3-4, a statement made on the basis of no evidence. M. Harl / G. Dorival / O. Munnich, La Bible grecque des Septante: Du judaïsme hellénistique au christianisme ancien, Paris 1988, 223-266, especially 233-235. J. Trebolle Barrera, The Jewish Bible and the Christian Bible: An Introduction to the History of the Bible (translated by W. G. E. Watson), Leiden/Grand Rapids, MI 1998 (1993), 71-72 (with Deissmann’s name misspelled here and on page 302). N. Fernández Marcos, The Septuagint in Context: Introduction to the Greek Versions of the Bible (translated by W. G. E. Watson), Leiden 2000, 3-17. J. M. Dines, The Septuagint (edited by M. A. Knibb), London 2004, 110-115, citing 114. M. Cimosa, Guida allo Studio della Bibbia Greca (LXX): Storia – Lingua – Testi, Rome 1995, 7274. 2. Periods in the Study of the Greek of the Septuagint

35

1. History of Scholarship on the Language of the Septuagint

gint. 91 Whereas these two works attempt to discuss the breadth of the Septuagint language, Trevor Evans is more narrowly concerned with the verbal syntax of the Septuagint Pentateuch. 92 In his summary of the character of Septuagint Greek, Evans states: “Over the last century it has become possible, in the light of the newly discovered evidence of Greek papyri coupled with methodological advances, to demonstrate more and more clearly in various respects the affinities of LXX Greek with the contemporary Koine vernacular of Egypt. This has led in large measure to a resolution of the old dispute.” 93 Whether he is entirely accurate in his final statement is beside the point and to some extent belied by this study, but his position is clear. As a result of his analysis of the verbal systems of both Hebrew and Greek in the Pentateuch, he concludes: “The method of translation adequately explains the Hebraistic case of the LXX. It is unnecessary to propose the existence of a special Jewish Greek dialect to explain the abnormalities. Nevertheless, the notion of Jewish Greek continues to find its advocates, especially in the broader sphere of biblical Greek studies. The present work will supply further clear evidence of ordinary Koine characteristics in the translation Greek of the LXX.” 94 After examining the Greek and Hebrew verbal systems (he takes a binary aspectual view of each, but without obvious iconicity), Evans examines how the two systems match up regarding the Greek perfect, the optative mood, the imperfect and aorist indicatives, and periphrastic tense-forms. He concludes that his findings contradict the “LXX syntax equals Hebrew syntax” perspective for the Pentateuch. 95

3. Conclusion There will no doubt be continuing discussion of the Greek of the Septuagint. At this point, there are two major orientations to the nature of its Greek: the revived form of the Koine Greek hypothesis and the modified form of the Jewish Greek hypothesis. The first begins with Koine Greek as its basis but admits that there are Semitisms due to the translational nature of the Septuagint. The other, even if recognizing that the Greek of the Hellenistic world, and even possibly of the Septuagint, was the koine, emphasizes the Hebraic elements of syntax and lexis. However, in their latest manifestations, in many (though not all) instances, these views are more a difference in perspective and approach than one of major substance. One approaches the issues from the standpoint of a Greek orientation and the other from a Semitic one. In many specific instances, they tend to converge around recognizable instances of Hebraic influence upon phenomena of the translation. This is not always the case, however, as some 91. L. Cignelli / R. Pierri, Sintassi di Greco Biblico (LXX e NT) (Studium Biblicum Franciscanum Analecta 61), Jerusalem 2003. 92. T. V. Evans, Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference, Oxford 2001. See also “The Comparative Optative: A Homeric Reminiscence in the Greek Pentateuch?” Vetus Testamentum 49.4 (1999), 487-504 and “Approaches to the Language of the Septuagint” Journal of Jewish Studies 56.1 (2005), 25-33. 93. Evans, Verbal Syntax, 4. 94. Evans, Verbal Syntax, 4. 95. Evans, Verbal Syntax, 259. I note that Evans’s own characterization of Greek is not without its problems, especially its bi-aspectuality.

36

3. Conclusion

1. History of Scholarship on the Language of the Septuagint

of those who argue for the koine hypothesis maintain the independence of Septuagint Greek, at least in its Pentateuchal form, from Hebraic interference. The difference in perspective remains, and tends to determine how the evidence is interpreted. As we have also seen, a remaining desideratum is a full and complete grammar of the Septuagint, although this is admittedly not without problems, as the Septuagint is a collection of translated, for the most part, and some original Greek works. The collection was written and compiled over a period of time and reflects different translational techniques. The above survey indicates the course of discussion of the issue, including whether the emphasis of such a grammar will be upon that of the substrate Hebrew or the Greek text itself. There has, nevertheless, been a general trend towards a more linguistic approach to the issue of the Greek of the Septuagint. The earliest discussions were limited in their comparative Greek material, and tended to use theological categories, such as a particular kind of Jewish Greek or Holy Ghost Greek, to describe the Greek that they saw. They did this as a way of explaining its distinct differences from Classical Greek. Because it made available a set of comparable Greek texts, the discovery of the documentary papyri opened up new categories of thought, forcing scholars to re-conceptualize the development of the Greek language and the relationship of the Greek of the Septuagint, even if as a translated work, to it. In this sense, there is a retrogressive thrust to the rejuvenation of the Jewish Greek hypothesis. This movement tends to focus upon the differences between the Greek Septuagint and other varieties of Greek, often concentrating upon particular grammatical phenomena. The revival of the Koine Greek hypothesis has shifted the ground again, in that it has taken a much more linguistically oriented approach. Utilizing work done in sociolinguistics and other areas, there is now a consideration of such constraints as bilingualism, dialect and idiolect, langue and parole, code and text, diglossia and code-switching, register, and prestige languages, as well as recognizing the most recent categories for discussion of the language, such as verbal aspect. Future studies that attempt to move the discussion forward will need to frame their analyses in these terms. Issues still to consider in this discussion include the following. There must be an increased awareness of the nature, breadth, and depth of the questions in the debate. Those who have led the revival of the Koine Greek hypothesis appear to have grasped this more fully than many of those promoting the Jewish Greek hypothesis. The questions are complex and involve more than simply the opposition between Greek and Hebrew. Proponents of the Jewish Greek hypothesis must also shed some of the residual theological baggage that associates language with a particular mindset. The second issue is the need for a more nuanced and sophisticated view of language. As a result, it is not enough any more simply to cite a particular linguistic phenomenon, without taking into account the language system and linguistic milieu in which it is used. This means that there must be a greater recognition of how the question of the nature of Septuagint Greek in relation to Hebrew must be seen as part of a larger question of what constitutes a language, a variety, a dialect, and a register. There is a wealth of discussion of these issues outside of biblical studies that could offer insights into this fundamental debate. The third and final issue is that of determining what counts for evidence and what constitutes the basic questions that are being answered. At various times in the on-going debate, the proponents seem to be at cross-purposes to each other. Sometimes they seem to be waging a proxy battle over the relationship 3. Conclusion

37

1. History of Scholarship on the Language of the Septuagint

between Judaism and Christianity. At other times, they seem to be engaged in a subtle dispute over cultural superiority. In only a relatively few instances does it appear that the proponents are actually discussing a linguistic issue on the basis of explicit linguistic criteria. The nature of the Greek of the Septuagint is first and foremost a linguistic question—whatever other issues may be involved or surround the debate. Future study needs to engage the question at that level, even if some of the conclusions are not those that would be welcome because of anticipated adverse results.

38

3. Conclusion

2. Die Diversität des Griechischen in der Septuaginta Peter Prestel

Das Textkorpus, das wir heute unter LXX verstehen, hat eine etwa 300jährige Entstehungszeit, es vereinigt in sich verschiedene Textgattungen und –sorten, es erfährt schon in der Entstehungszeit verschiedene Revisionen, die zu divergierenden Parallel-Versionen geführt haben, es enthält aus dem Hebräischen (oder Aramäischen) übersetzte Texte, die auf vom Masoretischen Text abweichenden Vorlagen beruhen können, und es enthält original auf Griechisch verfasste Texte; auch einige der übersetzten Texte enthalten original griechische Zusätze. Diese komplizierte Entstehungsund Überlieferungsgeschichte muss zu Diversität in der Sprachform innerhalb des Korpus führen. In diesem Beitrag sollen solche Diversitäten benannt werden; da diese im Einzelnen aber fast zahllos sind und sich doch typisch wiederholen, sollen vor allem auch der Typ und die Faktoren aufgezeigt werden, die zu solchen Diversitäten führen. Und da sich Diversität nur vor Konstanz abhebt, sollen auch die Konstanten sichtbar werden, die es durchaus durch das Textkorpus hindurch gibt. Abgesehen wird hier von textkritischen Fragen, die die ausgangssprachliche Vorlage betreffen. Der fundamentalste Faktor ist die Grundunterscheidung in Texte, die original griechisch verfasst wurden (insbesondere 2-4Makk und Weish), und solche, die aus dem Hebräischen übersetzt wurden – das ist die überwiegende Mehrzahl. Da Übersetzung nun sehr unterschiedlich vorgenommen werden kann, das Verhältnis zwischen Ausgangs- und Zielsprache sehr unterschiedlich tariert werden kann, wird es in diesem Beitrag auch – allgemein und konkret – um Fragen der Übersetzungstechnik und um verschiedene Übersetzungstypen gehen, und um die Frage von Übersetzung in der hellenistisch-römischen Antike überhaupt: In der Übersetzungstechnik ist einer der Hauptfaktoren für Diversität innerhalb der übersetzten Texte zu finden, durch den Vergleich mit anderer antiker Übersetzungspraxis zeigt sich nach außen die Besonderheit des Übersetzungstyps, der für die LXX insgesamt gilt. Eine anschauliche Hinführung in das Thema kann der Sirach-Prolog bieten: Er stellt eine »beispiellose« Besonderheit in der LXX dar 1, die einzige Stelle, wo ein LXX-Übersetzer sich explizit über die Übersetzung und deren Problematik äußert: Er ist – griechisch verfasst – der Weisheitsschrift vorangestellt, die zwischen 190 und 175 vom jüdischen Weisheitslehrer Jesus Sirach, dem »Großvater«, in hebräischer Sprache in Jerusalem verfasst wurde und zwischen 132 und 117 vom »Enkel« in Alexandria ins Griechische übersetzt wird. Wir haben also hier die Grundopposition: griechisch verfasst vs. übersetzt in einer Schrift, und eine Reflexion auf diesen Übersetzungsprozess.

1.

C. Wagner, Die Septuaginta-Hapaxlegomena im Buch Jesus Sirach. Untersuchungen zu Wortwahl und Wortbildung unter besonderer Berücksichtigung des textkritischen und übersetzungstechnischen Aspekts, Berlin 1999, 22. 3. Conclusion

39

2. Die Diversität des Griechischen in der Septuaginta

1. Sirachprolog I: Griechische und hebräische Kultur: Originales Griechisch vs. übersetztes Griechisch Interessant ist für unseren Zusammenhang zweierlei: Was der Übersetzer im Prolog sagt, vor allem aber, was er sprachlich tut, also wie er das sagt, was er sagt. In dieser Kombination liegt für unser Thema eine der wesentlichen Botschaften des Prologs. Zunächst zu dem, was er sagt: 1. Die vorliegenden Schriften, der Pentateuch, die Propheten und die übrigen Schriften machen die jüdische παιδεία und σοφία aus. Durch ihr Studium entsteht Verständnis und Gelehrsamkeit; da diese Gelehrsamkeit zur Weitergabe in Wort und Schrift auffordert, hat der Großvater, ein jüdischer Weisheitslehrer, aus genauer Kenntnis dieses Schrifttums selbst eine eigene Schrift in hebräischer Sprache verfasst und den bestehenden Schriften hinzugefügt; er steht also in einer Tradition und verfolgt dabei einen bestimmten erzieherischen Zweck: Ein gesetzestreues, an der Tora orientiertes 2 Leben soll durch Belehrung befördert werden (V. 1-14). 2. Die Bitte des Enkels an die Adressaten, evtl. (ziel-)sprachlichen Mängeln (λέξεις 3) seiner Übersetzung dieser Schrift mit Nachsicht zu begegnen. Es ist viel Mühe und Übersetzungskunst aufgewendet worden. Aber: Die Übersetzung stößt auf ein zentrales Problem: Ausgangs- und Zielsprache verfügen nicht über denselben Code – so könnte man ἰσοδυναμεῖ (V. 21) im Bezug auf Syntax wie Semantik auffassen. Es scheint also für griechisch Sprechende Befremdlichkeiten zu geben, man könnte sagen: Diversitäten im griechischen zielsprachlichen Duktus, der durch die Übersetzung bedingt ist. Und darüber hinaus: Schon im hebräischen Original weisen die kanonischen Schriften selbst erhebliche Unterschiede auf. Es gibt also schon im Original ausgangssprachlich Diversität, ohne dass diese hier näher bestimmt würde; sie scheint aber in einen – wenn auch etwas unklaren, eher apologetischen – Bezug zu den zielsprachlichen Divergenzen gesetzt zu sein 4 (V. 15-26). 3. Der Enkel gibt eine genauere Zeit-, Orts- und Adressatenangabe seiner Übersetzungsarbeit: In Alexandria zwischen 132-117; Adressat sind die Juden in der παροικία. Er betont erneut die viele Mühe und Kenntnis, die er seinerseits auf die Übersetzung verwandt hat, andererseits auch die hohe Bildungsaffinität und das diesbezügliche Interesse in seiner Zielgruppe in Alexandria. Er verspricht sich, in Aufnahme der Intention des Großvaters, einen großen erzieherischen Nutzen für 2. 3.

4.

Wagner, Hapaxlegomena, 123. Wagner, Hapaxlegomena, 118: λέξις meint im engeren Sinn einzelne Ausdrücke oder Redewendungen, im weiteren Sinn die stilistisch-literarische Gestaltung überhaupt. Hier wird der Plural auf das erstere verweisen. Schwierig die Deutung von ἐν ἑαυτοῖς λεγόμενα (V. 22 und 26). Die Übersetzung in W. Kraus / M. Karrer, Septuaginta Deutsch. Das griechische Alte Testament in deutscher Übersetzung, Stuttgart 2010, 1091 deutet beide Vorkommen unterschiedlich, den zweiten kaum richtig. Der Bezug in V. 26 von λεγόμενα als Acc.Graec. (»in Bezug auf das in ihnen Gesagte«) mit der Einengung auf die Inhaltsseite scheint mir nicht korrekt. Man müsste es beides Mal auf die hebräische Fassung beziehen: »wenn es bei sich gesagt ist«, d. h. nicht in einer Übersetzung (ἑτέραν γλῶσσαν). Man würde also auch in V. 26 verstehen wie in V. 22: »auf Hebräisch gesagt, verfasst«. Richtig F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta (Münsteraner Judaistische Studien 9), Münster 2001, 34.

40

1. Sirachprolog I: Griechische und hebräische Kultur

2. Die Diversität des Griechischen in der Septuaginta

diese Zielgruppe: Ein gesetzes- d. h. Tora-treuer Lebenswandel wird durch die jetzt möglich gewordene Lektüre der Schrift des »Großvaters« befördert (V. 27-36). Die wesentlichen Punkte der Aussage sind für uns: Inhalt und Bezug der Schrift, die Aussagen zur Übersetzungsproblematik und zur Zielgruppe. Hier sind wichtige Konstituenten der Pragmatik einer Übersetzung genannt, also der Textverwendung. Zur Sprachhandlung, der Form, in der diese Aussagen gebracht sind: Der Übersetzer zieht hier alle Register der griechischen literarischen Sprache und Kunstprosa seiner Zeit. Die drei Aussagen sind auf drei weitausgreifende Blöcke verteilt, deren erster und letzter nur aus einem einzigen Satz bestehen, also jeweils eine komplexe Periode bilden. Viele Partizipialkonstruktionen, variiert mit Hypotaxe, viele Parallelismen und ein Chiasmus (V. 7-11), Homoioteleuton, Isokolie, Prosarhythmus 5, eine sorgfältige Wort- und Satzverknüpfung durch Adjunktionen, geradezu virtuose syntaktische Konstruktionen mit der Attractio Relativi in V. 18-21 und der sehr komplexen Infinitivkonstruktion in V. 33-36, dazu im semantischen Bereich eine hohe Verdichtung der Felder durch sorgfältige Rekurrenz (Bildung, Weisheit, Anstrengung, gesetzestreues Leben). Eine Passage wie für das Lehrbuch der literarischen (asianischen 6) Kunstprosa wie auch der Rhetorik: Der Autor bewegt sich hier souverän innerhalb der gattungsspezifischen Proömiums-Topik – Bescheidenheits- und gleichzeitig Schwierigkeitstopos, captatio benevolentiae, Bezug auf die eigene Person wie den Text, dem der Prolog vorangestellt ist 7 – wie der dafür vorgesehen Stilebene, dem für ein Proömium empfohlenen genus medium. Wenn wir beide Informationsebenen, die sachliche wie die stilistisch-formale, zusammennehmen, können wir einige Grundzüge feststellen, die uns an das Thema der Diversität innerhalb der LXX insgesamt vorbereitend heranführen und sich dort als Faktoren von Diversität ausziehen lassen: Der Übersetzer gibt eine Information über seine griechische παιδεία; dies tut er durch die gewählte elaborierte Form; er zeigt seine Fertigkeit in der griechischen Sprache, der Zielsprache seiner Übersetzung und in der griechischen formalen παιδεία. Inhaltlich gibt er in dieser Form einen Ausblick auf jüdische παιδεία; diese wird er in einer Übersetzung darbieten, die offensichtlich trotz der erwiesenen Kompetenz und trotz aller Mühe zielsprachliche Probleme aufweist, also stark von dem elaborierten Griechisch differiert, das er im Prolog schreibt und dessen er fähig ist. Ein modernes Urteil über den sprachlichen Charakter des übersetzten Textes lautet denn auch: »The 5.

6.

7.

Das von Rahlfs in seiner LXX-Ausgabe Bd. II, 377 im Apparat gekennzeichnete poetische Metron des Hexameters wird von Siegert, Hebräische Bibel, 181 zurückgewiesen. Siegerts dortige Ausführungen sind leicht modifiziert wiederabgedruckt in M. Karrer / W. Kraus (Hg.), Septuaginta Deutsch. Erläuterungen und Kommentare, Stuttgart 2011, Bd. I, 53-64. Zum sog. Asianismus als dem auf Einfachheit und Klarheit bedachten Attizismus entgegengesetzte Stilrichtung E. Norden, Antike Kunstprosa, Leipzig/Berlin 1909 (Nachdruck Darmstadt 1974), Bd. I, 126-147. Vergleichbar mit diesem Prolog, wenn auch stilistisch nicht so raffiniert, ist der Prolog 2Makk 19-32, wo V. 19-28 aus drei langen Perioden bestehen; dort überwiegen Parallelismen. Asianisch ist dann vor allem 4Makk. Wagner, Hapaxlegomena, 25; S. Kreuzer, »Der Prolog des Buches Ben Sira (Weisheit des Jesus Sirach) im Horizont seiner Gattung – Ein Vergleich mit dem Euagoras des Isokrates« in: J.-F. Eckholdt / M. Sigismund / S. Sigismund (Hg.), Geschehen und Gedächtnis. Die hellenistische Welt und ihre Wirkung (Festschrift für W. Orth), Berlin 2009, 135-160. 1. Sirachprolog I: Griechische und hebräische Kultur

41

2. Die Diversität des Griechischen in der Septuaginta

translation, by contrast, is executed in a more or less stilted translationese that is often at pains to repeat certain formal aspects of the Hebrew very closely – matters such as word order …« 8. Der Übersetzer zeigt also im Prolog, was er eigentlich zielsprachlich kann und er zeigt, auf was er verzichtet: Er begibt sich seiner elaborierten Sprache, seiner literarischen Virtuosität, seiner griechischen formalen παιδεία, weil er übersetzt 9. Griechische formale Kultur und Literatur tritt hinter eine Übersetzung zurück, in der die jüdische παιδεία dargeboten wird und offensichtlich dargeboten werden muss. Der Autor bewegt sich in zwei Welten, der griechischen literarisch-formalen Kultur und der hebräischen inhaltlichen Kultur und trennt sie durch unterschiedliches sprachliches Handeln. Was er verbindet, ist jüdische Lehre mit jüdischer Adressatengruppe; dies tut er mit einer Übersetzung, die offensichtlich die Ausgangssprache nicht vergessen lässt, sondern durch eine gewisse Sperrigkeit auf sie verweist und sie so auch in der Zielsprache bewahrt. Hier lässt sich also eine enge Bindung der jüdischen παιδεία an die Sprache, in der sie ursprünglich verfasst ist, konstatieren; sie wird in einer ausgangssprachlichen Übersetzung wiedergegeben und damit in einer sprachlichen Bindung; die Form ist »unfrei« gegenüber einer freien griechischen Komposition. Die griechische Sprache schlägt die Brücke zwischen jüdischer Lehre und jüdischem Adressatenkreis, hält aber die griechische literarische Kultur offensichtlich fern. Diese Paradoxie, die der Kontrast zwischen dem Sirachprolog und der Sirach-Übersetzung vermittelt, gilt im Prinzip für die gesamte übersetzte LXX.

2. Übersetzungspraxis in der griechisch-römischen Antike: Eine andere Pragmatik: Varianz- vs. Invarianzforderung Ein kleiner Exkurs über die Übersetzungspraxis in der Antike wird die spezifische Pragmatik der LXX-Übersetzung – gebildet durch die Faktoren, die im Sirachprolog aufgeführt sind: Ausgangstext, Zielsprache, Adressatenkreis, Übersetzung als Mittlung und mittelnder Übersetzer – und ihre singuläre Besonderheit in ihrem historischen Kontext verdeutlichen. Wir werden dabei kontrastiv ein anderes Verhältnis zwischen Ausgangssprache und Ausgangskultur auf der einen und Zielsprache und Zielkultur auf der anderen Seite finden, als sie stellvertretend für die LXX im Sirachprolog erscheint. Die Besonderheit der LXX-Übersetzung besteht in zwei Hinsichten, nämlich dass übersetzt wurde, und wie übersetzt wurde. Die Bedeutung des für uns nicht ungewöhnlichen Sachverhaltes einer Übersetzung in eine andere Sprache liegt vor allem in der Zielsprache begründet: Dass Texte 8.

9.

B. G. Wright III, »Why a prologue? Ben Sira’s Grandson and his Greek Translation« in: S. M. Paul / R. A. Kraft / E. Ben-David / L. H. Schiffman / W. W. Fields, Emanuel. Studies in the Hebrew Bible, Septuagint and Dead Sea Scrolls. In Honor of Emanuel Tov (VTS 94), Leiden 2003, 637-638. Zu Besonderheiten der Sirach-Übersetzung, vor allem unter dem Aspekt des Wortschatzes vgl. Wagner, Hapaxlegomena. In dem nicht übersetzten, griechisch verfassten Buch Weish stellt sich die Frage paradoxerweise umgekehrt: Dort scheint ein gebildeter Autor in einem natürlich-griechischen Text absichtlich auf die gleichzeitig demonstrierte sprachlich-kulturelle Bildung zu verzichten und teilweise zu schreiben, als ob er übersetzte.

42

2. Übersetzungspraxis in der griechisch-römischen Antike

2. Die Diversität des Griechischen in der Septuaginta

in der griechisch-römischen Antike ins Griechische übersetzt wurden, ist außerordentlich ungewöhnlich. Man hat in der Antike auch übersetzt, aber normalerweise nicht ins Griechische; die griechische Sprache und Kultur ist nicht die nehmende, sondern die gebende 10. Wenn man das griechische Publikum erreichen wollte, schrieb man original griechisch in einem der literarisch-rhetorisch zur Verfügung stehenden Register – Beispiele sind Berossos mit seiner babylonischen, Manetho mit seiner ägyptischen und schließlich Flavius Josephus mit seiner jüdischen Geschichte, auch wenn man dazu, wie Josephus in seiner Anfangszeit, vielleicht eine Art Redakteur brauchte. Allein die Tatsache also, dass ein Text ins Griechische übersetzt wurde, sagt etwas über die Pragmatik der LXX-Übersetzung aus. Die Römer haben viel und früh aus dem Griechischen übersetzt. Allerdings würde man diese Übersetzungen nach heutigen Maßstäben nicht als Übersetzung bezeichnen, sondern als freie Bearbeitung. Übersetzung in Rom speist sich aus zwei Interessen und bewegt sich zwischen zwei Polen 11: Die griechische Ausgangssprache und -kultur ist objektiv höher entwickelt, es wird durch Übersetzung versucht, die eigene Sprache zu befördern und Anschluss an die höhere Kultur und Sprache zu finden. Es gibt keine Gleichzeitigkeit oder Ebenbürtigkeit der beiden Sprachen und Kulturen. Es liegt eine vertikale »Übersetzung« vor – primär als descensus von oben nach unten, von einer höher eingeschätzten und von den objektiven Gegebenheiten wie Vokabular, Syntax und Stilistik her differenzierter ausgebildeten Sprache nach unten in eine Volkssprache, die dadurch entwickelt, erweitert, differenziert, kultiviert wurde, oder umgekehrt als ascensus von einer Volkssprache nach oben in die Kultursprache, dies vor allem später aus den Volkssprachen ins Lateinische, aber eben in der Antike nicht ins Griechische. Die vertikale Übersetzung ist so konstitutiv für die Ausbildung der römischen Literatursprache und Literatur wie auch später für die der modernen Nationalsprachen und -literaturen. Man hat diese vertikale Übersetzung lange Zeit für die einzig kunstvolle und lohnende Form des Übersetzens gehalten. Der kunstvolle Übersetzer soll, wie Cervantes sagt, nicht wie einer sein, der nur ein Papier vom anderen abschreibt 12. Der andere Pol ist die so genannte aemulatio, die konkurrierende und auf Überbietung angelegte Nachbildung des Ausgangstextes in der »Übersetzung«. Das führt zu einer Amplifikation oder Reduktion des Ausgangstextes in vielerlei Hinsicht. Als Invariante des Ausgangstextes wird oft nur das nackte Handlungs- oder Faktengerüst respektiert. Diese Art der übersetzenden Bearbeitung gilt heute als Intertextualität. In Rom sind es Äußerungen eines Kulturpatriotismus, der sich aus einem gewissen Nachholbedürfnis und der vertikalen Situation gegenüber der griechischen Kultur speist. Die beiden Pole der römischen Übersetzung – Kultivierung der eigenen Sprache und schöpferische Überbietung des Originals – liegen also auf der Achse der Kon10. H. St. J. Thackeray, A Grammar of the Old Testament in Greek, Cambridge 1909 (Nachdruck Hildesheim 1987), 21. 11. Die folgenden Ausführungen stützen sich auf A. Seele, Römische Übersetzer. Nöte, Freiheiten, Absichten. Verfahren des literarischen Übersetzens in der griechisch-römischen Antike, Darmstadt 1995, und J. Albrecht, Literarische Übersetzung. Geschichte – Theorie – Kulturelle Wirkung, Darmstadt 1998. 12. Albrecht, Literarische Übersetzung, 143. 2. Übersetzungspraxis in der griechisch-römischen Antike

43

2. Die Diversität des Griechischen in der Septuaginta

kurrenz einer weniger entwickelten Zielsprache und Zielkultur mit einer höher entwickelten Ausgangssprache und Ausgangskultur. Charakteristisch für den in Rom ausgebildeten Übersetzungstyp ist somit: Kollektiv stehen Bedürfnisse der Zielsprache und -kultur eindeutig im Vordergrund, individuell die Bedürfnisse des Übersetzers selbst. Es geht um sein Künstlertum, seine Kreativität, seinen Rang als Literat, nicht als Dolmetscher. Das am wenigsten Interessante und geradezu Hemmende ist aus diesem Blickwinkel eine zu enge Bindung an den Ausgangstext und seinen genauen Wortlaut. Es tritt etwas hinzu, das diesen Typ der einbürgernden oder, wie Nietzsche sagt 13, erobernden Übersetzung in Rom begleitet und vielleicht auch befördert: Es ist der in gebildeten Kreisen übliche Bilingualismus. Das Zielpublikum konnte griechische Texte ohne weiteres im Original lesen; Kenntnis des Originals konnte vorausgesetzt werden, war zumindest möglich. Das entlastet den Übersetzer von einer Informationspflicht und schafft Raum für seine künstlerische Freiheit der Bearbeitung. Vor diesem Hintergrund werden signifikante Unterschiede in der Pragmatik der LXX-Übersetzung deutlich, die das Entstehen eines ganz anderen Übersetzungstyps bedingen oder zumindest befördern: Denn hier liegt zum ersten Mal, soweit wir wissen, ein Übersetzungstyp vor, der dem modernen Übersetzungs-Ideal, seit dem 18. Jh. und von Schleiermacher formuliert 14, entspricht: Invariante ist nicht wie in Rom, oder später bei Josephus, der das Alte Testament frei nacherzählt, das nackte Handlungs- oder Faktengerüst, sondern der Wortlaut, sogar die Wortstellung, manchmal sogar die Wortzahl. Damit ist eine Bedingung in hohem Maße erfüllt, die ein moderner Übersetzungstheoretiker als konstitutiv für »Übersetzung« postuliert zur Unterscheidung von »Bearbeitung«: Eine Übersetzung beruht auf aufgestellten Invarianzforderungen, d. h. sie orientiert sich textintern an Merkmalen der Ausgangssprache, die Bearbeitung beruht hingegen auf Varianzforderungen gegenüber der Ausgangssprache mit einer stärkeren Berücksichtigung textexterner Aspekte 15. Diese »moderne« Auffassung von Übersetzung, die die LXX-Übersetzer entwickeln, lässt sich auch an der anderen Ausrichtung der gerade beschriebenen Achse der Konkurrenz von Kulturen und ihrer Sprachen beschreiben: Es gibt kein objektives Bedürfnis der griechischen Zielsprache nach Differenzierung und Kultivierung durch eine Übersetzung aus dem Hebräischen. Die griechische Sprache ist morphologisch, syntaktisch, lexikalisch reich differenziert; ebenso ist eine griechische Literatursprache in allen Varianten und Gattungen voll entwickelt. Insofern ist eher verwunderlich, dass in der LXX-Übersetzung die römische Übersetzungsrichtung nicht umgekehrt wird: nämlich als ascensus von unten nach oben in die existente Literatursprache. Aber das ist bezeichnenderweise verweigert, wie der Sirachprolog zeigt: Die Übersetzung erfolgt von unten nach unten, unter der Literatursprache hindurch 16. Man benutzt 13. 14. 15. 16.

Zitiert bei Albrecht, Literarische Übersetzung, 154. Albrecht, Literarische Übersetzung, bes. 74. Albrecht, Literarische Übersetzung, 267. J. Joosten, »Rhetorical Ornamentation in the Septuagint: The case of grammatical variation« in: E. Bons / Th. J. Kraus (Hg.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint, Göttingen 2011, 11-22, bes. 13-14 unterschätzt m. E. die Übersetzer, wenn er ihnen die Fähigkeit zu einem besseren Griechisch abspricht und daran zweifelt, dass sie ausgerechnet den hochprestigehaltigen Text unter dem literarischen Standard übersetzen

44

2. Übersetzungspraxis in der griechisch-römischen Antike

2. Die Diversität des Griechischen in der Septuaginta

die Koine und damit eine unprätentiöse, syntaktisch und stilistisch eher reduzierte nicht-literarische Umgangs- und z. T. Fachsprache, die selbst nicht viel Körper entfaltet 17. Die Übersetzungsrichtung ist nicht vertikal, sondern horizontal. Diese Verweigerung lässt weitere Rückschlüsse auf die Pragmatik zu: Es ist nicht die eigene literarische Ambition, die diese Übersetzung antreibt. Und: Der Adressat kann nicht der griechische Literatur- und Bildungsbetrieb sein, dann hätte man virtuos geschrieben und frei bearbeitet wie Josephus in seinen Antiquitates. Es kann nur eine jüdische Adressatengruppe sein, die jüdische Gemeinde in Alexandria, die der Sirachprolog ja benennt. Damit ist die römische Situation verändert: Die Übersetzung der LXX verlässt zwar die Ausgangssprache, das Hebräische, das in Alexandria nicht verstanden wird – es gibt vermutlich kaum Bilingualismus – und das offensichtlich nicht als Identifikationsfaktor für die jüdische Ethnie dient, aber sie verlässt nicht die hebräische Ausgangskultur. Zwar ist jede Übersetzung in gewisser Weise interkulturell, aber man kann vielleicht doch etwas zugespitzt sagen: Hier findet vergleichsweise eher eine interlinguale denn eine interkulturelle Operation statt. Man kann vermutlich einen Schritt weitergehen: Absicht der Übersetzung ist wahrscheinlich gerade die Bewahrung und vielleicht auch Erneuerung der hebräischen Kultur, um in der dominanten Umgebung einen identitätsstiftenden Stabilisator zu besitzen. Anders als in Rom existiert der Wettbewerbsdruck nicht gegenüber der Kultur des Ausgangstextes, wo ja die eigene jüdische Identität herkommt, wohl aber existiert vermutlich ein Abgrenzungsdruck gegenüber der Kultur, deren Sprache man spricht. Wo die Römer aus Kulturpatriotismus die Texte verändern und damit die Varianzforderung konstituieren, bedingt der jüdische Religionspatriotismus eine Bewahrung des Textes in der Übersetzung: Hier wird die Invarianzforderung konstituiert. Die möglichst genaue Abbildung des hebräischen Textes in die griechische Sprache bildet die identitätsstiftende Grenze zur griechischen Kultur, der Bewahrung der jüdischen παιδεία. Wo die Römer versuchen, sich durch freie Bearbeitung und Überbietung von der Ausgangskultur abzugrenzen, versuchen die Juden durch Verweigerung der literarischen Standards der Zielsprache, durch Unterbietung der griechischen Literatursprache und durch enge Bindung an den Ausgangstext die Ausgangskultur zu konservieren. Es entsteht so soziolinguistisch betrachtet eine Art In-Group-Effekt 18; man hat auch von Soziolekt gesprochen 19. hätten wollen, wenn sie es denn besser gekonnt hätten. Es gibt in der aufgezeigten Pragmatik gute Gründe dafür, und dass zumindest manche es auch besser gekonnt hätten, zeigt z. B. der Sirach-Prolog. 17. Thackeray, Grammar, 16 nennt als Charakteristika eine erhebliche Erweiterung des Vokabulars und Tendenz zu simplification, uniformity, lucidity (21), unter Vernachlässigung des literarischen Stils, wobei es auch in der Koine eine literarische Sprache gibt, z. B. Polybius und Josephus, die allerdings in der übersetzten LXX nicht verwendet ist. Siegert, Hebräische Bibel, 143 bezeichnet die Koine als »modern in dem Sinne, dass ein großes Vokabular kombiniert wird mit einer umso einfacheren Syntax«, »das ›Erfolgsrezept‹ aller heutigen Weltsprachen, am meisten des Englischen«. Zur Koine vgl. auch K. Usener, »Zur Sprache der Septuaginta« in: Septuaginta Deutsch. Erläuterungen, Bd. I, 40-51. 18. Siegert, Hebräische Bibel, 186 und Usener, Septuaginta Deutsch. Erläuterungen, Bd. I, 63. 19. Usener, Septuaginta Deutsch. Erläuterungen, Bd. I, 51 verwendet diesen Begriff. Allerdings ist die Sprache der LXX-Übersetzung kaum die gesprochene Sprache der Juden; man sollte eher 2. Übersetzungspraxis in der griechisch-römischen Antike

45

2. Die Diversität des Griechischen in der Septuaginta

Dazu kommt der alles entscheidende Unterschied: Der Ausgangstext ist kein Theaterstück wie bei den Römern, sondern ein heiliger Text. Damit liegt eine Voraussetzung vor, die in der modernen Theorie als weitere konstitutive Bedingung für »Übersetzung« angeführt wird zusammen mit der schon genannten Invarianzforderung: Funktionskonstanz zwischen Ausgangs- und Zieltext 20. Die Textfunktion eines heiligen Textes bleibt in der Ausgangs- wie Zielkultur am ehesten konstant, im Fall der LXX wird diese Funktionskonstanz durch die Diasporasituation (παροικία) eher noch gesteigert. Diese in Abhebung von der antiken »Übersetzungs« praxis beschriebene Pragmatik der LXX-Übersetzung führt zu einer horizontalen Übersetzung, es ist wie das Abschreiben eines Blattes Papier, wie Cervantes diesen Übersetzungstyp abwertend beschreibt – Hieronymus trägt dem Unterschied in der Übersetzung von heiligen und nicht-heiligen Texten Rechnung, indem er sagt: Man braucht Texte nicht wörtlich zu übersetzen, »mit Ausnahme der hl. Schriften, wo selbst die Wortfolge ein Mysterium ist« 21. Der Übersetzer soll kein eigener Autor sein, sondern er soll den Autor des Ausgangstextes zum Sprechen bringen, nicht zu viel eigene Körperlichkeit entwickeln. Es wird kein eigener künstlerisch-literarischer Anspruch entwickelt, sondern man stellt sich handwerklich-technisch in den Dienst des heiligen Ausgangstextes. Kein Wunder, dass man in der Neuzeit die Bibelübersetzung als Modell für Übersetzung überhaupt nahm: Jeder Ausgangstext soll heilig sein. Das führt uns zurück zum Sirachprolog. Wir betrachten den zweiten Ausgangspunkt nach der Tatsache der Übersetzung ins Griechische: Es ist die Problematik, die sich durch ein so verstandenes Übersetzen, das auf der Basis der Funktionskonstanz der Invarianzforderung unterliegt, ergibt.

3. Sirachprolog II: Übersetzungsprobleme aufgrund der Invarianzforderung: Produzenten- und Rezipientenperspektive Hier gibt es zwei Aussagen: Die Übersetzungsproblematik liegt nicht in der mangelnden Fertigkeit oder Nachlässigkeit des Übersetzers; sein ἀδυναμεῖν (V. 20) ist objektiv begründet in der unterschiedlichen Codierung zweier Sprachen (οὐ ἰσοδυναμεῖ); man kann hinzusetzen: Zumal wenn sie so unterschiedlich codiert sind wie das semitische und primär parataktisch ausgerichtete Hebräisch und das indoeuropäische und syntaktisch so differenzierte Griechisch. von Sprachregistern oder traditionell von Sprachstil sprechen, der für bestimmte Sprachverwendungssituationen typisch, aber auch daran gebunden ist und dort identitätsstiftend sein kann. Vgl. Albrecht, Literarische Übersetzung, 97. 20. Vgl. Albrecht, Literarische Übersetzung, 259: »dass ein Zieltext nur dann als Übersetzung sensu stricto eines Ausgangstextes verstanden werden kann, wenn Funktionskonstanz wenigstens in einem sehr allgemeinen Sinn vorliegt: Es muss erkennbar sein, dass die Funktion des übersetzten Textes nicht ausschließlich im Hinblick auf textexterne Faktoren gewählt, sondern wenigstens teilweise aus Charakteristika des Originals abgeleitet wurde«. 21. Hieronymi ep. 57, 5 »[…] absque scripturis sanctis, ubi et verborum ordo mysterium est, […]«: Text nach G. J. M. Bartelink, Hieronymus, Liber de optimo genere interpretandi (Epistula 57). Ein Kommentar, Leiden 1980.

46

3. Sirachprolog II: Übersetzungsprobleme aufgrund der Invarianzforderung

2. Die Diversität des Griechischen in der Septuaginta

Die Probleme werden genauer verortet in λέξεις τινές (V. 20); der Plural scheint hier darauf hinzuweisen, dass Redewendungen gemeint sind, die zielsprachlich, rezipientenbezogen offensichtlich auffallen müssen 22; an diesen bleibt der Adressat hängen. Das zeigt Abweichungen von einem zielsprachlichen Standard an. Was bedeutet die zweite Stelle, wo ausdrücklich Diversität genannt ist (διαφορά V. 26)? Diese Aussage wird oft auf die übersetzten Versionen von Nomos, Propheten und den »übrigen Büchern« bezogen und als Hinweis genommen, dass deren Übersetzung dann zu dieser Zeit schon vorgelegen habe. Es würde also Diversität in den bereits vorliegenden griechischen Versionen konstatiert, und der Sirach-Übersetzer würde sich gleichsam in eine Tradition der zielsprachlichen Unebenheit stellen, die bereits übersetzten Texte müssten als »Entschuldigung« für zielsprachliche Ungleichheit der eigenen Übersetzung dienen 23. Also läge hier eine Wiederholung und Verstärkung der rezipientenorientierten Argumentation vor. Aber von übersetzten Texten ist nicht explizit die Rede 24. Wozu diente dieses Argument, wenn es auf ausgangssprachliche Diversität bezogen ist? Was hat es mit den Schwierigkeiten der Übersetzung zu tun? Gemeint kann eigentlich nur sein: Es gibt zielsprachlich Diversität, die auf ausgangssprachliche Diversität zurückzuführen ist. Hier würde also dem rezipienten- der produzentenbezogene Aspekt der Übersetzung hinzugefügt werden 25. Zielsprachliche Diversität ist somit in eine Beziehung gesetzt zu ausgangssprachlicher Diversität, die sich in die Zielsprache fortsetzen würde. Es wird also ein enger Konnex weniger für den Inhalt als vielmehr für die sprachliche Entscheidung des Übersetzers hergestellt, was z. B. Stilistik oder andere sprachliche Merkmale des Prätextes betreffen könnte 26. Wieder betont das den engen Anschluss an den Ausgangstext, also den Übersetzungscharakter, und impliziert damit übersetzerische Entscheidungen und Invarianzhierarchien der Übertragung. Der jüdische Leser, der des Hebräischen unkundig ist, muss dies ungeprüft hinnehmen. Der moderne Betrachter schließt daraus: Diversität kann nicht nur durch zielsprachlichen innergriechischen Vergleich – rezipientenbezogen – beurteilt werden, sondern muss auch von der Ausgangssprache her betrachtet werden, von einem genauen Vergleich beider sprachlicher Realisierungen und der daraus zu 22. Albrecht, Literarische Übersetzung, 93: Rezipientenorientierte Beschreibung geht von einer »Erwartungsnorm« des Rezipienten aus, von der es aus Rezipientensicht quantitativ – z. B. durch ungewöhnliche Frequenz – oder qualitativ Abweichungen gibt. 23. Wagner, Hapaxlegomena, 29 liest dies auch als Hinweis auf stärkere inhaltliche Veränderungen, Akzentverschiebungen zum Original, die der Autor wohl vorgenommen hat, Veränderungen, die allerdings das griechisch-sprechende jüdische Zielpublikum nicht beurteilen kann: Inhaltliche Abweichungen oder etwa Sachadäquatheit von Äquivalenzen entziehen sich ohne Kenntnis des hebräischen Ausgangstextes. 24. Wagner, Hapaxlegomena, 81 leitet den Bezug auf vorliegende Übersetzungen unnötig aus dem Unterschied zwischen προφῆται (V. 1) für das Original und προφητεῖαι (V. 24) für die angeblichen Übersetzungen ab. 25. Albrecht, Literarische Übersetzung, 93: Produzentenbezogen werden die Phänomene beschrieben »unter dem Gesichtspunkt der Wahl, die ein Autor aus einem zur Verfügung stehenden Vorrat von Ausdrucksmitteln« in der Zielsprache getroffen hat. 26. Zu denken wäre etwa an parataktische vs. hypotaktische Struktur, parataktisch syndetische vs. asyndetische Strukturen, ganz zu schweigen von unterschiedlichen Textsorten: prosaisch vs. poetisch, narrativ vs. argumentativ, direkte Rede, und natürlich entsprechende Inhalte. 3. Sirachprolog II: Übersetzungsprobleme aufgrund der Invarianzforderung

47

2. Die Diversität des Griechischen in der Septuaginta

entnehmenden Entscheidungen der Übersetzer. Dies hat in der LXX-Forschung den Blick auf die Übersetzungstechnik geöffnet 27. Der Weg geht von einem zielsprachlichen rezipientenorientierten Vergleich zur interlingualen Untersuchung. Hat Thackeray in seiner berühmten und ja auch grundlegenden Klassifizierung der übersetzten Bücher sich noch mehr von zielsprachlich-stilistischen Kriterien leiten lassen (gutes Koine-Griechisch etc.) 28, so wird heute die Frage nach der Diversität vor allem fokussiert auf die Abhängigkeit von der Ausgangssprache. Die Diskussion wird entlang der Termini »frei« vs. »wörtlich« geführt. Da diese Begriffe aber oft sehr mehrdeutig verwendet werden, etwa auch exegetische und sprachliche Kriterien nicht getrennt werden, ist eine einengende Kriterienbildung entscheidend; die Kriterien für »frei« vs. »wörtlich« sollen in sprachlicher, systemlinguistischer Hinsicht bestimmt werden, Fragen der exegetischen Freiheit werden hier nicht verfolgt. Als Rahmen sei hier bewusst ein moderner übersetzungstheoretischer Ansatz für die Kriterienbildung für »freie« und »wörtliche« Übersetzung vorangestellt 29; in ihm werden zudem die äußeren wie inneren Grenzen von »Übersetzung« bestimmt, und für eine Einordnung der Ergebnisse der Erforschung der Übersetzungstechnik kann ein nicht immanenter Standpunkt durchaus fruchtbar sein.

4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht: Kriterienbildung und Anwendung auf die LXX Mit »wörtlich vs. frei« kann man auf sprachlicher Ebene zwischen Ausgangs- und Zieltext eine Skala der Annäherung an die Zielsprache bzw. der Nähe zur Ausgangssprache aufstellen; technisch sind hier die Begriffe syntagmatisch (die horizontale 27. Bahnbrechend sind die Arbeiten von J. Barr, The Typology of Literalism in ancient biblical translations, (Nachrichten der Akademie der Wissenschaften, Phil-Hist. Klasse), Göttingen 1979 und E. Tov, The Text-Critical Use of the Septuagint in Biblical Research, Jerusalem 1981, und andererseits die Publikationen der »finnischen Schule«: I. Soisalon-Soininen, Die Infinitive in der Septuaginta, Helsinki 1965; R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint, Helsinki 1979; A. Aejmelaeus, Parataxis in the Septuagint. A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch, Helsinki 1982; einen kritischen Vergleich der beiden Ansätze nimmt vor B. Lemmelijn, »Two Methodical Trails in Recent Studies on the Translation Technique of the Septuagint« in: R. Sollamo / S. Sippilä (Hg.)., Helsinki Perspectives on the Translation Technique of the Septuagint, Göttingen 2001, 43-63. 28. Thackeray, Grammar, 13 gruppiert die übersetzten Bücher in vier Klassen primär nach ihrer zielsprachlichen Qualität: 1) Good Koine Greek: Pentateuch, Jos, Jes, 1Makk; 2) Indifferent Greek: Teile von Jer, Ez, 1–2Chr und Kgt, sowie Ps, Sir, Jdt; 3) Literal or unintelligent versions: z. B. Teile von Jer, Bar, Kgt sowie Hld und Klgl, Dan (Θ), 2Esdr, Koh; 4) Literary: 1Esdr mit Dan O, Est, Hiob, Spr. Kritik an dieser Klassifizierung und ihren Kriterien bei Barr, Typology, 283, allerdings findet sie modifiziert auch aus dem Fokus der Übersetzungstechnik viel Bestätigung, z. B. bei Soisalon-Soininen, Infinitive, 208. 29. Die folgenden Ausführungen orientieren sich an dem eher pragmatisch-unideologischen, nicht zu komplexen und die Übersetzungstradition bis zur antiken Rhetorik aufnehmenden und weiterführenden Modell von Albrecht, Literarische Übersetzung, bes. 141-142 und 262269.

48

4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

2. Die Diversität des Griechischen in der Septuaginta

»und«-Beziehung, z. B. die Abfolge der Elemente, die Wortstellung) und paradigmatisch (die vertikale »oder«-Beziehung, die Austauschmöglichkeiten nicht nur im semantischen, sondern auch im syntaktischen Bereich: z. B. Kommutation des Genitivattributs mit adjektivischem Attribut oder Relativsatz) einzusetzen. Die größte Annäherung geschieht sicher in einer durchgängig idiomatischen Übersetzung; hier wäre man bei einer »freien« Übersetzung deswegen, weil sie sich in den zwei Beziehungen – paradigmatisch und syntagmatisch – frei macht von grammatischen und lexikalischen Bindungen, die primär der Ausgangssprache gehören und sie durch idiomatische der Zielsprache ersetzt. Diese Übersetzung führt das Original zum Leser. Am anderen Ende der Skala steht eine Übersetzung, die sich auf Kosten der idiomatischen, aber eventuell auch grammatisch-lexikalischen Korrektheit in der Zielsprache mit bloßer Verständlichkeit begnügt; hier sind wir bei der am ehesten »unfreien« Übersetzung, weil sie sich von den Eigenheiten der Ausgangssprache nicht frei machen will, sie vielmehr dokumentieren will; eine solche Übersetzung ist die Interlinearversion, die ja auch die Wortabfolge wiedergibt und dokumentiert. Sie bringt eher den Leser zum Original 30. Dazwischen wäre anzusiedeln eine Strategie, die eine nicht unbedingt idiomatische, aber doch grammatisch und lexikalisch korrekte zielsprachlichen Version erreicht. Hier werden drei Relationen zwischen Ausgangs- und Zieltext beschrieben, die ein Übersetzer wählen kann; für Albrecht sind zwei Bedingungen für die Gattung »Übersetzung« konstitutiv; zum einen die schon genannte Funktionskonstanz des Textes, die für die Pragmatik der LXX-Übersetzer konstitutiv ist, und zum anderen der Bezug aller Wiedergabeentscheidungen in der Zielsprache auf den Ausgangstext, also die Realisierung der Invarianzforderung auf systemlinguistischer Ebene. Wenn man dies zu konkretistisch versteht, scheint nur eine sehr wörtliche Wiedergabe diese Bedingung für Übersetzung zu erfüllen. Es treten aber allgemeinere Relationsbegriffe wie Adäquatheit, Äquivalenz oder die traditionelle »Treue« hinzu. Ein Übersetzer wird – explizit oder implizit – ausgehend von der Funktionskonstanz eine Invarianzhierarchie aufstellen, die darüber entscheidet, was er vom Ausgangstext bewahren will und was nicht. Sodann wird er nach adäquaten und äquivalenten Mitteln der Zielsprache suchen, die diese invariant gesetzten Elemente des Ausgangstextes und der Ausgangssprache in der Zielsprache realisieren können. Das sind nicht unbedingt die gleichen Elemente, die die Ausgangssprache aufweist; z. B. kann die Anfangsstellung des Verbs oder parataktischer Satzbau in einer Sprache normal, also unauffällig, in der anderen aber sehr auffällig sein. Das vermittelnde Kriterium im Sinne der Äquivalenz wäre dann »normal«: Also z. B. die Umsetzung des normalen parataktischen Satzbaus in einen für die Zielsprache typischen und normalen, aber eben nicht unbedingt parataktischen Satzbau. Hier entscheidet der Übersetzer also über den Grad an ausgangs- bzw. zielsprachlicher Orientierung – und über die zielsprachliche Qualität seiner Übersetzung, allerdings nicht über deren »Treue«. Auch eine im obigen Sinn hergestellte freie Übersetzung ist treu, und eine ganz wörtliche kann teilweise untreu sein 31, wenn Mit30. Nach dem bekannten Diktum von F. Schleiermacher, »Über die verschiedenen Methoden des Übersetzens« in: Friedrich Schleiermacher’s sämmtliche Werke. Dritte Abteilung. Zur Philosophie, Zweiter Band, Berlin 1838, 218, zitiert bei Albrecht, Literarische Übersetzung, 74. 31. Vgl. Lemmelijn, »Two Methodical Trails«, 51 mit Anm. 31 und Siegert, Hebräische Bibel, 134. 4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

49

2. Die Diversität des Griechischen in der Septuaginta

tel der Ausgangssprache in der Zielsprache zu wenig äquivalent sind, d. h. eine völlig andere Bedeutung erlangen. Wenn man dieses Schema auf die Schriften der LXX anwendet, die sich auf einen Ausgangstext beziehen, so lassen sich zunächst vier Grundlinien und Faktoren der Diversität aufzeigen: 1. Zum einen gibt es im zu beobachtenden Textkorpus sowohl nach oben wie auch nach unten Überschreitungen der hier bestimmten Grenzen von Übersetzung: nach oben durch quantitative Zusätze oder Subtraktion von Elementen 32; hier sollte man besser nicht mehr von »freier Übersetzung« sprechen; der Begriff »frei« macht ja nur Sinn in der Bindung an einen Ausgangstext. Am anderen Ende ist gelegentlich ein Unterschreiten dieser Skala zu konstatieren. Dies geschieht etwa dort, wo die Bibelübersetzer die eigentlich zur Übersetzungstechnik gehörende Arbeit der Entscheidung für ein Sinnverständnis verweigern, um Interpretation zu vermeiden 33. Es wird sozusagen semantisch Rohmaterial geboten, z. B. Gen 6,14 νοσσιὰς ποιήσεις τὴν κιβωτὸν – »du wirst den Kasten (d. h. die Arche) zu Nestern machen«. Was soll das bedeuten? Der Sinnvollzug soll dem Leser überlassen werden, die Übersetzung ist nicht zum Ende gekommen. Dies trifft in verstärkter Weise auf Aquila zu, dessen Wörtlichkeit eigentlich nicht mehr treu ist. 2. Es ist schwierig, ganze Bücher oder Büchergruppen, wie z. B. den Pentateuch durchgängig einem der drei Übersetzungstypen zuzuordnen. Innerhalb der Bücher, selbst der früh übersetzten, kommt es zu erheblicher Diversität; teilweise von einem Kapitel zum anderen, teilweise von einem Satz zum anderen enthalten sie Elemente aller drei Ebenen. Das trifft selbst für den von Thackeray und gemeinhin eher als zielsprachlich orientiert angesehenen Pentateuch zu. Dazu später. 3. Innerhalb des Schemas gibt es jedoch eine deutliche Tendenz zur zweiten und dritten Übersetzungsweise. Die oberste Ebene, die zielsprachlich-idiomatische, die syntagmatisch wie paradigmatisch vom Ausgangstext frei ist, ist deutlich unterrepräsentiert; sie findet sich teilweise in Ex, Hiob, Spr, Dan, Est, die letzteren alles Bücher, die dem dritten Teil des hebräischen Kanons angehören, auf die vielleicht die Invarianzforderung von Dtn 4,2 nicht so strikt angewendet wurde 34. Der Grund für diese Beschneidung um die obere Ebene liegt primär in der überwiegend vorliegenden syntagmatischen Bindung an die Ausgangssprache: Die Wortabfolge orientiert sich meistens an der hebräischen, die zudem stark parataktisch ausgerichtet ist, und schöpft die Variationsmöglichkeiten des Griechischen, wie sie etwa im Sirach-Prolog zu sehen waren, bei weitem nicht aus. Aufs Ganze gesehen ist die für das Hebräische eher typische Parataxe beibehalten. Dies mag auf eine easy technique 35 zurückzuführen sein, die, ohne heute unverzichtbar erscheinende Hilfsmittel wie Wörterbuch, Lexikon oder Konkordanz, eher in kleinen Einheiten – Wort für Wort – übersetzt als in größeren Blöcken. Weitere Faktoren, die die oberste Ebene ab32. Barr, Typology, 304-305 nennt Targumismen als Zusätze wie z. B. Hiob 2,9a-e und etliche Passagen in Spr, z. B. 1,7bc oder 6,8abc, vgl. H-W. Jüngling / H. von Lips / R. Scoralick, »Paroimiai« in: Septuaginta Deutsch. Erläuterungen, Bd. II, 1950 und bes. 1955. 33. Ein anderer Grund kann sein, dass ein hebräisches Wort nicht verstanden wurde. 34. Thackeray, Grammar, 15. 35. Barr, Typology, bes. 300.

50

4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

2. Die Diversität des Griechischen in der Septuaginta

schneiden, betreffen die Einzelwortebene; die Übersetzung wird »wörtlicher«, wenn und weil sie am Einzelwort hängt. Dies äußert sich 36: a) In Wortkonkordanz, d. h. der Tendenz, Wörter der Ausgangssprache mit immer dem gleichen Wort der Zielsprache zu übersetzen, es entstehen also feste Vokabelgleichungen, die auf Kosten der flexiblen kontextuellen Übersetzung gehen können. Einige dieser festen Wortpaarungen werden aus der Gen, die eher kontextuell übersetzt, in die ganze LXX übernommen wie διαθήκη – »Verfügung« in 99 % der Fälle für ‫ברית‬, oder κύριος für ‫יהוה‬, νόμος für ‫תורה‬, ὁδός »Weg« für ‫דרך‬, ἀρχή für hebräisch ‫»( ראש‬Haupt«). Es können auch mehrere Stereotypen gebildet werden, z. B. wird das hebräisch polyseme ‫»( דבר‬Wort/Sache«) mit λόγος, ῥῆμα, πρᾶγμα wiedergegeben 37; umgekehrt werden bei anderen polysemen Wörtern wie z. B. dem hebräischen ‫»( נפשׁ‬Person/ Leben«, aber auch hebräisch als Personalpronomen verwendet) die verschiedenen Bedeutungen in ein einziges Stereotyp gegossen, wodurch in einer starr-konkordanten Übersetzung eine ausgangssprachliche Polysemie nicht monosemiert ist und zielsprachlich auffällig wird 38 oder Verständnisschwierigkeiten bereiten kann; b) in Segmentierung hebräischer Wörter in bedeutungstragende Elemente und Wiedergabe durch entsprechende individuelle – teilweise wieder stereotypisierte – griechische Elemente, z. B. 2Kgt 5,24 wird das ‫ בשמעך‬zerlegt in ἐν τῷ ἀκοῦσαί σε: alle Konstituenten des hebräischen Wortes sind durch ihre Standardentsprechungen im Griechischen wiedergegeben 39; c) durch quantitative Entsprechung: Wörtliche Übersetzer versuchen jedes Element der Ausgangssprache mit einem Element der Zielsprache wiederzugeben, freie expandieren ein hebräisches Element oder dessen Aussage in mehrere griechische Wörter 40 oder verdichten mehrere hebräische Wörter zu einem griechischen 41, oder lassen kontextuell überflüssig Erscheinendes aus 42. Die wörtlichen Übersetzer zielen eine 1:1 Relation an. Wenn man die Freiheit auf diesen beiden syntagmatisch unfreien und auf die Übersetzung von Einzelelementen reduzierten Ebenen sucht, wird man sie deshalb eher kleinteilig im paradigmatischen Bereich finden, d. h. produzentenbezogen in der Auswahl zwischen verschiedenen für ein Ausgangselement kommutierbaren semantischen und syntaktischen Elementen der Zielsprache, um ein bestimmtes ausgangssprachliches 36. Tov, Text-Critical Use, 54. 37. Zu Vokabelgleichungen mit weiteren Beispielen vgl. Tov, Text-Critical Use, 54-57; M. Harl, »La langue de la Septante«, in: M. Harl / G. Dorival / O. Munnich (Hg.), La Bible grecque des Septante. Du judaisme hellénistique au christianisme ancien, Paris 1988, 246-251 und Siegert, Hebräische Bibel, 220-250. 38. Z. B. Hos 9,4, wo das hebräische Personalpronomen ‫ לנפשם‬stereotyp mit ταῖς ψυχαῖς wiedergegeben wird. 39. Tov, Text-Critical Use, 58; für die standardisierte Wiedergabe von ‫ ב‬durch ἐν, das als Hebraismus gilt, vgl. Soisalon-Soininen, Infinitive, 80. 40. Z. B. Gen 9,22 durch Zusatz des Partizips ἐξελθών oder Ex 32,26 durch Zusatz des Imperativs ἴτω. 41. Z. B. Jos 5,8, wo die hebräische Übergangsformel ‫ ויהי כאשר־תמו‬in das Partizip περιτμηθέντες verdichtet wird. 42. Z. B. Jos 4,14, wo ein hebräischer Vergleichssatz ‫ כאשר‬zusammengezogen wird zu ὥσπερ ohne Wiederholung des Verbs: statt Vorlage »sie fürchteten ihn wie sie Mose gefürchtet hatten«: καὶ ἐφοβοῦντο αὐτὸν ὥσπερ Μωυσῆν. 4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

51

2. Die Diversität des Griechischen in der Septuaginta

Merkmal variiert und doch adäquat, also treu wiederzugeben 43. Wir haben also eine deutliche Einengung der möglichen »Freiheit« durch die syntagmatische Bindung und Konzentration auf kleinteilige Übersetzungseinheiten mit der Tendenz zu festen Wortpaarungen. Diese Übersetzungstechnik ist letztlich durch die beschriebene interlinguale Übersetzungspragmatik – spezifischer Adressatenkreis und eher noch pointierter Funktionskonstanz – bedingt und bringt diese sprachlich zum stärksten Ausdruck. 4. Und doch ist bei aller Diversität diachron eine auffällige konstante Tendenz festzustellen: nämlich im Laufe der LXX-Übersetzungsgeschichte die Entwicklung von der mittleren Ebene zur unteren, teilweise sogar darunter. Später übersetzte Bücher sind unfreier 44, teilweise in einem Ausmaß, dass sie Fehler in der Zielsprache akzeptieren, die teilweise die Verständlichkeit, das Kriterium der untersten Ebene, tangieren 45. Die Unfreiheit und »Wörtlichkeit« in der – konkordant-starren – Übersetzung von Einzelwörtern oder sogar Morphemen eines hebräischen Wortes, die sich auch in den frühen Übersetzungen in den Büchern ja durchaus punktuell findet, scheint hier zum konsistenten Programm zu werden 46: So wird etwa die hebräische Partikel ‫» – בי‬bitte« im Pentateuch und Jos richtig mit δέομαι wiedergegeben, spätere wie Ri und Kgt übersetzen sie dagegen mit ἐν ἐμοί 47. Der Grund für diese auffällige Entwicklung liegt in der allmählichen Herausbildung eines protomasoretischen Textes aus der Perspektive frühjüdisch-rabbinischer Hermeneutik und deren Interpretationstendenzen als hebräischer Standardtext 48. Damit entsteht für die LXX ein neuer

43. Beispiele sind etwa die Variation des in einer Passage wiederholt vorkommenden hebräischen Wortes ‫ בן‬durch griechische Synonyme, wie z. B. in Dtn 22,6-7 τέκνον vs. παιδία, oder in Ex 21,15-17, wo der hebräische Inf. abs. »des Todes sterben« variiert wird durch θανατούσθω vs. τελευτήσει. Weitere Beispiele bei Η. Β. Swete, The Old Testament in Greek, Cambridge 1914, 328, etwa Ps 36/37 wo das hebräische ‫» חטא‬Sünder« mit ἁμαρτωλός bzw. ἀσεβής wiedergegeben wird. Im syntaktischen Bereich: In Ps 3,7 wird der hebräische Relativsatz ‫ אשר סביב‬mit dem griechischen substantivierten Partizip τῶν συνεπιτιθημένων, einem mit dem Relativsatz kommutierbaren Element, wiedergegeben und dadurch auch der in den Relativsätzen häufig vorkommende Hebraismus, der in der redundanten Setzung eines Personalpronomens oder Adverbs besteht, vermieden. 44. Etwa die Chronik und 2Esdras. 45. Z. B. Koh 8,17: καὶ εἶδον σὺν πάντα τὰ ποιήματα, 1Makk 3,15: καὶ προσέθετο καὶ ἀνέβη, 4Kgt 6,10: καὶ ἐφυλάξατο ἐκεῖθεν οὐ μίαν οὐδὲ δύο, oder Valenzverstöße wie ποεῖν μετά (Ri 9,19, Ruth, 2Chr, Sir, 1Makk), ἀσεβεῖν ἀπό (2Kgt 22,22, Ps 17/18,22), πειράζω ἐν (Ri 6,39) oder gar ὠφελεῖν ἐν (Ps 88/89,23, wo noch eine Verwechslung des hebräischen Wortes zugrunde liegt), dürften zielsprachlich teilweise die Verständlichkeitsschwelle unterschritten haben. 46. Barr, Typology, 281. 47. Z. B. Gen 43,20 gegenüber 3Kgt 3,17. Ein weiteres Beispiel stellt die stereotypisierte, kontextfreie Übersetzung des hebräisch polysemen ‫»( שלום‬Friede, Heil, Wohlergehen«) dar; im Pentateuch wird kontextuell mit εἰρήνη, σωτηρία, σωτήριον übersetzt oder die idiomatische Verwendung als Grußformel mit ὑγιαίνειν wiedergegeben (Gen 26,9), in den Ps wird dagegen stereotyp mit εἰρήνη übersetzt. 48. Zusammenfassend bei S. Kreuzer, »Entstehung und Entwicklung der Septuaginta im Kontext alexandrinischer und frühjüdischer Kultur und Bildung« in: Karrer / Kraus, Septuaginta Deutsch. Erläuterungen, Bd. I, 3-36, bes. 26-28.

52

4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

2. Die Diversität des Griechischen in der Septuaginta

hebräischer Bezugstext und das Bedürfnis, die vorhandenen Übersetzungen (Old Greek) dem neuen Text in neuen Bearbeitungen möglichst eng anzugleichen; es werden daraufhin Überarbeitungen der älteren Übersetzungen vorgenommen, insbesondere die kaige-Rezension 49, und z. T. entstehen verschiedene Versionen eines Textes, z. B. in Ri, weite Passagen in 2–4Kgt, wobei die revidierten Übersetzungen gegenüber den nicht revidierten Fassungen stärker die oben genannten Merkmale der Wörtlichkeit zeigen, z. B. Ri 5,3 A: ἐγὼ τῷ κυρίῳ ᾄσομαι – »ich werde dem Herrn singen« gegenüber B, das der kaige-Revision unterliegt: ἐγώ εἰμι τῷ κυρίῳ, ἐγώ εἰμι ᾄσομαι – »ich bin dem Herrn, ich bin ich werde singen«; diese kontextund sinnfreie Übersetzung entsteht durch das Bedürfnis, ein Morphemmerkmal des hebräischen Wortes zu transportieren, nämlich die Langform des Personalpronomens 50; oder z. B. die stereotype Verwendung des hebräischen ‫ שׁלום‬in Ri 18,15 B gegenüber der kontextuellen in A. Und neben der Revision bestehender Texte werden auch neue Übersetzungen in diesem stark ausgangssprachlichen interlinearhebraisierenden Stil hergestellt, z. B. die Psalmen und das Hohelied. Dieser Weg führt dann schließlich zu Aquila, der eine fast absolute Stabilität von Wortgleichungen zwischen dem Hebräischen und Griechischen befolgt 51, wo das Verständnis des Zieltextes eigentlich die Kenntnis der Ausgangssprache verlangt, im Einzelnen die Kenntnis der hebräischen Wortbildung, die der griechische Text abbildet auf Kosten des Sinnes. Es werden Morpheme, die in der hebräischen Wortbildung gebunden sind, in den Rang von Wörtern gesetzt, die ihre griechische selbständige Semantik mit sich führen, z. B. entsteht durch die Übersetzung der hebräischen Akkusativpartikel –et durch das selbständige Wort σύν in Gen 1,1 dieser im Griechischen ungrammatische Satz: ἐν κεφαλαίῳ ἔκτισεν ὁ θεὸς σὺν τὸν οὐρανὸν καὶ σὺν τὴν γῆν – »in der Hauptsache 52 schuf Gott mit den Himmel und mit die Erde«. Aquila transportiert weniger den hebräischen Sinn ins Griechische als vielmehr zeigt er den Weg in die hebräische Formenbildung. Er stellt in seiner Fixierung nur auf das Einzelwort und nicht auf die Verknüpfung, auf die Form und nicht auf den Sinn 53, den Extrem-

49. Nach der charakterisierenden Wiedergabe von hebräisch ‫ גם‬mit καί γε statt einfachem καί. 50. Thackeray, Grammar, 55. Zielsprachlich fast noch krasser 2Kgt 11,5, das auch der kaige-Revision unterliegt: ἐγώ εἰμι ἐν γαστρὶ ἔχω – »ich bin ich habe (ein Kind) im Bauch«. In 2Kgt 12,18.22 lässt sich beobachten, dass die kaige-Rezension gegenüber dem antiochenischen Text den Gen.abs. ἔτι τοῦ παιδίου ζῶντος durch einen subst. AcI ersetzt: ἐν τῷ ἔτι τὸ παιδάριον ζῆν; beide Versionen haben den gleichen semantischen und syntaktischen Wert – »als das Kind noch lebte«, der Gen.abs. ist aber mehr zielsprachlich, der subst. AcI näher der ausgangssprachlichen Formulierung; in 2Kgt 15,32 tilgt kaige den Gen.abs. ersatzlos. An einer längeren Textpassage kann man die Auswirkung von kaige in einem Vergleich von 4Kgt 19–32 (kaige) mit Jes 37,31-33 verfolgen, die einen fast identischen masoretischen Text übersetzen, vgl. Tov, Text-Critical Use, 64. 51. Barr, Typology, 312 und O. Munnich, »Le texte de la Septante et ses problèmes«, in: La Bible grecque des Septante, 146 nennen aber auch für Aquila Ausnahmen von den starren Wortkonkordanzen, z. B. für κεφάλαιον in Gen 1,1, vgl. nächste Anmerkung. 52. Aquila »verbessert« das bisherige ἐν ἀρχῇ, um die hebräische Wurzel ‫» ראש‬Haupt« (griechisch: κεφαλή) wiederzugeben. 53. Barr, Typology, 312. 4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

53

2. Die Diversität des Griechischen in der Septuaginta

fall der Wörtlichkeit dar, die die Gattung der Übersetzung eigentlich vielfach nach unten untersteigt. Das Messen der LXX-Übersetzungen an dieser Skala hat grundsätzlich eine starke Kondizionierung der zielsprachlichen Version durch die Ausgangssprache ergeben. Es sind so negativ die Beschränkungen benannt, die durch die beschriebene Übersetzungspragmatik bedingt sind: Sie zeigt sich im Sprachlichen absteigend von der obersten Ebene in unterschiedlichen Abstufungen von »Wörtlichkeit«, d. h. nach unten die zunehmende Übersetzung auf Wortniveau und teilweise auf Wortkonstituentenniveau, und sie wird greifbar in der Stereotypisierung von Übersetzungsvarianten auf Kosten der Kontextualität. Dadurch wird die Skala um die oberste Ebene, für die syntagmatische wie paradigmatische Freiheit charakteristisch ist, stark beschnitten; wir sehen, was fehlt, auf was die Übersetzer verzichten, was sie sich von ihrer Pragmatik her verbieten. Es lassen sich aber auch gegenläufig positiv von unten nach oben die Bedingungen aufzeigen, die erfüllt sind und von den Übersetzern erfüllt sein müssen, um die mittlere und punktuell auch oberste Ebene zu erreichen. Wie die negative Beschreibung der Beschränkungen die Bindung der Übersetzer an die Ausgangssprache und ihre Pragmatik zeigt, so wird in dieser positiven Beschreibung die zielsprachliche Leistung der Übersetzer stärker gewürdigt 54. Diese ergänzende positive Beschreibung ist auch deswegen sinnvoll und nötig, weil in den Arbeiten zur Übersetzungstechnik mehr oder weniger latent oft noch der alte Hebraismus-Streit zu spüren ist: Hebräischer Stil in griechischen Buchstaben – oder eben ein arriviertes Koine-Griechisch 55. Das setzt sich in Werturteile über »gutes« oder »schlechtes« Griechisch um. Hier liegen meistens Vorentscheidungen vor, die durch den Blickpunkt des Betrachters bedingt sind: Liegt er mehr produzentenbezogen auf der Ausgangssprache, besteht in der Beobachtung der Übersetzungsrelation die Tendenz, im Griechischen überall das Hebräische zu sehen, selbst da, wo der nur zielsprachliche Beobachter, der des Hebräischen nicht kundig ist, oft gar nicht an eine Bedingung durch eine Ausgangssprache denken wird oder muss. Ein Beispiel liegt in Gen 1,4 vor: καὶ εἶδεν ὁ θεὸς τὸ φῶς ὅτι καλόν – »und Gott sah das Licht, dass es gut war«. Hier wird der mehr ausgangssprachlich orientierte Betrachter die hebräische Wortstellung betonen, dem griechischen Rezipienten und dem zielsprachlichen Betrachter fällt hier aber nur auf, dass der Übersetzer eine im griechischen gut beheimatete und durchaus elegante Version erzielt hat, die sog. Prolepsis, das Vorziehen des Nebensatzsubjektes durch syntaktische Anbindung an das nebensatzauslösende Prädikat (εἶδεν). Die adäquat-äquivalente Wiedergabe der ausgangssprachlichen Prolepsis ins Deutsche ist ihre Rücknahme: Und Gott sah, dass das Licht gut 56 war. Hier liegt also für sich genommen in der Zielsprache kein störendes Phänomen vor, eher im Gegenteil. In solchen nicht seltenen Fällen wird eine hohe Geschicklich54. Vgl. Lemmelijn, »Two Methodical Trails«, 50, bes. Anm. 25 mit dem schönen Zitat von Soisalon-Soininen, nach dem die wirklichen (zielsprachlichen) Kompetenzen der Übersetzer nicht unterschätzt werden sollten und besonders an den seltenen freieren Wiedergaben abzulesen sind. 55. J. Joosten, »Rhetorical Ornamentation«, 11 führt Vertreter der beiden Positionen auf. 56. Dass καλός die Bedeutung »gut« annimmt, ist allgemeine Koine-Entwicklung.

54

4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

2. Die Diversität des Griechischen in der Septuaginta

keit der Übersetzer deutlich, die in der Sicht des einsprachigen Rezipienten die Frage nach »frei« oder »wörtlich« überhaupt nicht stellt. Die Übersetzer analysieren offensichtlich die Mittel der Ausgangssprache, und versuchen, die adäquaten Mittel der Zielsprache dafür zu finden. Dass sie dabei einfache Lösungen der ja sehr flexiblen Zielsprache suchen, d. h. Lösungen, die strukturell Ähnlichkeiten mit ausgangssprachlichen Phänomenen aufweisen, z. B. den Infinitivgebrauch 57 – und sie z. T. geringfügig modifizieren, hält die Übersetzung im interlingualen, mittleren Bereich. Wo die Übersetzung aber diese mittlere Ebene erreicht, ist dies auf die Geschicklichkeit und zielsprachliche Kompetenz der Übersetzer zurückzuführen; es wird nach unten deutlich, was sie können, und nach oben, was sie nicht wollen. Sie erfüllen das Kriterium der Adäquatheit im Ganzen erstaunlich gut. Zudem bleiben linguistische Bereiche, die nicht durch die Ausgangssprache kondizioniert sind, in der Würdigung der zielsprachlichen Leistung ausgespart, z. B. weitgehend das griechische Verbalsystem mit seinen Differenzierungen in Modus, Tempus, Aspekt, Diathese 58 und im syntaktischen Bereich die Valenzverknüpfungen und Kasusrektionen, die ja größtenteils respektiert sind 59; Syntax ist mehr als Wortabfolge. Durch die zielsprachliche Realisierung dieser Mittel wird weitgehend die zielsprachliche Korrektheit der mittleren Ebene erreicht. Durch die Betrachtung genau solcher Bereiche, wo keine ausgangssprachliche Bindung vorliegen kann, ergibt sich nun aber ein zielsprachliches binnengriechisches Relief. Wenn der Fokus ergänzend zum übersetzungstechnischen Aspekt auch rezipientenorientiert auf der Zielsprache liegt, gibt es zwei Referenzpunkte: Das Griechische kommt zum einen in seine natürliche Umgebung, das der nicht-literarischen Koine, das die jüdischen Rezipienten ja sprachen 60, zum anderen kommt es in einen Vergleich zwischen den Übersetzern innerhalb der LXX. In der ersten Beziehung wird man vergleichbare Verwendungen außerhalb der LXX, vor allem in den Papyri, suchen – und ja auch in erstaunlichem Umfang finden; die Zahl der eindeutigen und stereotypisierten Hebraismen auf Wort- oder Wortgruppenniveau sind dadurch eher begrenzt worden: Lexikalische calques wie προστίθημι – »fortfahren, wiederholt etwas tun« oder sperrige syntaktische Wendungen wie die typische narrative Übergangsformel καὶ ἐγένετο/ἔσται καί – »und es geschah/wird sein und« fallen besonders auf; aber auch dort gibt es Versuche, sie der Zielsprache anzupassen 61. Als natürlich griechisch erweist sich dagegen der Gebrauch des Verbalsystems, d. h. vor allem Modi und Aspekte; sie können im Pentateuch typisch für die frühe Koine angesehen werden; die attischen Strukturen sind intakt, die beginnenden 57. Soisalon-Soininen, Infinitive, zusammenfassend 207. 58. T. V. Evans, Verbal Syntax in the Greek Pentateuch. Natural Greek Usage and Hebrew Interference, Oxford 2001, 67-68 »[…] in verbal usage there was major scope for independence in translation from Hebrew to Greek: The LXX translators regularly faced options pertaining to Greek syntactical usage which were not specifically motivated by their Hebrew text« (68). Vgl. seine Schlussfolgerungen 259-264. 59. Eine Übersicht über die syntaktischen Hebraismen gibt R. Helbing, Die Kasussyntax der Verba bei den Septuaginta. Ein Beitrag zur Hebraismusfrage und zur Syntax der κοινή, Göttingen 1928, 324. 60. Die Vorstellung, die Juden in Ägypten hätten eine abweichende Sondersprache gesprochen, wird schon von Thackeray, Grammar, 27-28 als unwahrscheinlich zurückgewiesen. 61. Dazu weiter unten im Text. 4. »Wörtlich« vs. »frei« aus systemlinguistischer Sicht

55

2. Die Diversität des Griechischen in der Septuaginta

Veränderungen im Pentateuch setzen nur Entwicklungen fort, die schon in der späten klassischen Periode begonnen haben 62. Der zweite zielsprachliche Referenzpunkt – die LXX-Binnenperspektive – eröffnet den Blick auf Diversitäten und Niveauunterschiede innerhalb des übersetzten Textkorpus 63. Ihnen soll an einigen Beispielen nachgegangen werden.

5. Zielsprachlich realisierte Diversität: Paradigmatische Freiheit in der syntagmatischen Bindung In den folgenden Stellen Hiob 6,8; 14,13; 19,23; 29,2; 31,31.35 und Ps 14/13,7 = 53/52,7 liegt jeweils die gleiche hebräische Wunschformel vor: Auf die Einleitungsformel (wörtlich: »Wer gibt … ?« > idiomatische Bedeutung »Oh dass doch«; wahrscheinlich steht das »Wer gibt … ?« als Ersatz für »Gott möge geben, dass …«) folgen im Hebräischen unterschiedliche Konstruktionen, so dass sich folgende Kombination ergibt: (1) »Oh dass doch (sei):« → (2) (ein beliebiger Satz, der durch den Zusammenhang mit (1) ko-textuell als Wunsch determiniert wird). Also: (1) »Oh dass doch (sei):« → (2) »Meine Bitte wird/soll kommen.« (Hiob 6,8) → (2) »Du wirst/sollst mich im Totenreich verbergen.« (Hiob 14,13) → (2) »Dass meine Worte aufgeschrieben werden/würden.« (Hiob 19,23) → (2) »Ich [im Hebräischen im Akk., also eine Art AcI-Konstruktion, wenngleich ohne Verb, da ein Nominalsatz ohne Kopula vorliegt] (bin) wie in früheren Zeiten.« (Hiob 29,2) → (2) »Von seinem Fleisch wird er nicht satt [Partizip].« (Hiob 31,31) → (2) »Ich habe einen, der mir zuhört [< einen mir Zuhörenden].« (Hiob 31,35) → (2) »Aus Zion (ist) das Heil Israels.« (Ps 14,7) Hintergrund ist letztlich, dass es im Hebräischen keine eigentlichen Nebensatzkonstruktionen gibt, der semantische Gehalt der hebräischen Aussagesätze lässt sich nicht definitiv bestimmen 64. Hier ist in der vom hebräischen Text unabhängigen Verwendung des Optativs Variation zu erkennen: Im Griechischen wird differenziert zwischen dem WunschOptativ (Hiob 6,8 ἔλθοι), wobei auch die Einleitungsformel zielsprachlich etwas auffällig wiedergegeben ist (εἰ γὰρ δῴη), dem Irrealis (14,13, ähnlich Ex 16,3), markiert durch εἰ γὰρ ὄφελον unter zielsprachlich guter Auslassung der Einleitungsformel, dem Potentialis (19,23, ähnlich Dtn 28,67) ausgangssprachlich mit Einleitungsformel (τίς γὰρ ἂν δῴη) und zielsprachlich mit folgendem AcI, in 29,2 erscheint wieder der Potentialis als Frage (τίς ἄν με θείη) ohne Einleitung mit syntaktisch richtiger Anbindung des με in einer ansonsten ausgangssprachlichen Formulierung, ähnlich in 31,31 und 31,35, wo zusätzlich die Modalpartikel ἄν fehlt, wie in Num 11,29; in Ps 14,7 steht in der Frage ein Futur (τίς δώσει, ebenso Dtn 5,29). Wie man diese Variationen 62. Evans, Verbal Syntax, 262. 63. Dies ist der Ansatz der in Bons / Kraus, Et sapienter et eloquenter versammelten Beiträge. 64. Ich verdanke Stefan Schorch diese Erklärungen zum Hebräischen.

56

5. Zielsprachlich realisierte Diversität

2. Die Diversität des Griechischen in der Septuaginta

auch immer erklären will, die unterschiedliche modale Realisierung scheint auf jeden Fall nicht von der Vorlage her bedingt zu sein 65, sondern zielsprachliche Gestaltungsabsicht zu verraten, teilweise ist in Hiob syntagmatische Freiheit, also die höchste Form der Freiheit vom Ausgangstext erreicht 66. Dass die zielsprachlich gelungenen Wendung in unmittelbarer Verbindung zu zielsprachlich eher auffälligen Formulierungen stehen, zeigt das enge Ineinander der mittleren, auf zielsprachliche Korrektheit und Angemessenheit und unteren Übersetzungsweise, die sich mit zielsprachlicher Verständlichkeit begnügt. Ähnliche Variationen finden sich in Formulierungen von Geboten 67, wo zweimaliges hebräisches yiqtol in manchen Texten (Gen 3,3; Ex 20,5; 22,30; 23,13; Dtn 22,5) im zweiten Vorkommen variiert wird durch οὐ μή + Konj. Aor., im Unterschied zu Stellen wie 2Kgt 17,35, dem bereits der späteren neuen Hermeneutik gehorchenden Übersetzungstyp, und Ez 24,16, wo jeweils das Futur variationslos wiederholt wird. Da diese – paradigmatische – Variation allerdings nicht außerhalb der LXX vorkommt, hat man hier mit einer Binnenunterscheidung zu rechnen, die von den frühen Übersetzern gefunden und später wohl wieder aufgegeben wurde. Allerdings könnte zielsprachlich die Variation durch Aspektüberlegungen mit bestimmt sein, da einer temporal bestimmten Form (Futur) 68 mit dem Aorist ein Aspekt hinzugefügt wird und mit der affirmativen Negationskombination οὐ μή eine zusätzliche Nuance, etwa im Sinne einer Verstärkung, entsteht. Eine weitere Auffälligkeit, die nicht ausgangssprachlich motiviert ist, sind einige wenige Homerismen im griechischen Text, d. h. sprachliche oder inhaltliche Bezüge auf die homerischen Epen, vielleicht eine Frucht der in Alexandria blühenden Homerphilologie. Ein formelhaftes Relikt könnte die Wendung ὡς ἂν εἰ oder ὡσεί mit Optativ in Vergleichssätzen sein 69, z. B. Dtn 1,44 im Vergleich mit Homer Ilias 2,780. Diese Wendung ist nach Homer im klassischen Griechisch sehr selten. In der übersetzten LXX gibt es dagegen 25 Vorkommen, gehäuft im Pentateuch mit neun Vorkommen; sieben davon geben hebräische Imperfekte, zwei den Infinitivus constructus wieder. Der Vergleich an sich wie auch seine Distribution – fünf der neun Pentateuchstellen – ist natürlich durch die Vorlage bestimmt, jedoch nicht die Verwendung des Optativs, die von anderen Vergleichssätzen in der Septuaginta abweicht 70. Ein Beispiel für 65. So urteilt für Num 11,29 auch Evans, Verbal Syntax, 188. 66. Die Hiob-Übersetzung hebt sich auch durch ein ungewöhnliches hellenistisches und nicht im Sinne fester Vokabeläquivalenzen konkordant verwendetes Vokabular ab, es gibt zahlreiche Hapaxlegomena und Neologismen; der vergleichsweise reiche Partikelgebrauch ist eher an klassischem Griechisch orientiert. In den aus Theodotion ergänzten Teilen weist die Übersetzung dagegen wieder die Merkmale der Interlinearversion auf, z. B. 22,24. Vgl. M. Kepper / M. Witte, »Job« in: Septuaginta Deutsch. Erläuterungen, Bd. II, 2041-2126, bes. 2054-2056. 67. Joosten, »Rhetorical Ornamentation«, 17-20. 68. Evans, Verbal Syntax, 39 und 262 verteidigt die traditionelle Auffassung des Indikativ Futurs, die etwas Zukünftiges als Faktum vorwegnimmt. Der zielsprachliche Leser wird den Unterschied Faktum vs. Aspekt mit der affirmativen Negation sicher wahrgenommen haben. 69. Evans, Verbal Syntax, 190-197, Usener, Septuaginta Deutsch. Erläuterungen, Bd. I, 49-50 und »Griechisches im Griechisch der Septuaginta« in: Bons / Kraus, Et sapienter et eloquenter, 8198 mit Hinweisen auf weiteres griechisch-poetisches Vokabular, besonders in Ez. 70. J. Joosten, »Elaborate Similes – Hebrew and Greek. A Study in Septuagint Translation Tech5. Zielsprachlich realisierte Diversität

57

2. Die Diversität des Griechischen in der Septuaginta

eine inhaltliche homerische Reminiszenz findet sich in Ez 27,5, wo der kontextuell geradezu kontradiktorische Zusatz ἐλατίνους (»aus Tannenholz«; Zypressenholz wird aus dem Libanon geholt, um Masten aus Tannenholz herzustellen!) an geprägte Odysseestellen (2,424 und 15,289) erinnert. Einen interessanten Umsetzungsprozess vom Ausgangs- in den Zieltext zeigt die Übersetzung poetischer Teile der hebräischen Bibel (bes. Ps, Hiob, Spr). Die Übersetzer haben nicht versucht, die Poesie in Poesie zu übertragen, das hätte neben der ohnedies gegebenen Schwierigkeit einer poetischen Übertragung vielleicht auch zu sehr in die Nähe griechischer Literaturgattungen geführt, also die beschriebene Übersetzungspragmatik verlassen. Sie haben im Sinne der oben beschriebenen Adäquatheit ein Mittel der Zielsprache gesucht, das das ausgangssprachliche ersetzt: Sie haben den Text in Prosa umgesetzt, also ein typisches Merkmal der poetischen Gattung, das poetische Versmaß (Hexameter, Pentameter) außer Acht gelassen, und dafür das ästhetische Mittel der Kunstprosa verwendet, den Prosarhythmus. Gehobene Prosa ist in der griechisch-römischen Literatur rhythmisiert 71, indem bei der Wortreihung auf eine bestimmte Abfolge von Längen und Kürzen der Silben geachtet wird; in der LXX wird so insbesondere der sog. Creticus (– ˙ –) verwendet. Diese Rhythmisierung erfolgt an den Betonungsstellen eines Satzes, also am Satzanfang oder Satzende, oder auch an wichtigen Stellen dazwischen, die so geschmückt und betont werden. Erste Ansätze gibt es im Pentateuch (Gen 49,1; Ex 15,1-21; Dtn 32,1-43), eine starke Ausdehnung erfolgt dann in den Psalmen, wo man in bis zu drei Vierteln der Verse Rhythmisierungen erkennen kann 72. Diese Praxis verweist auf das laute Vorlesen, was zu einem »feierlichen 73 Sprechvortrag vor (oder in) größeren Auditorien« 74 für die Psalmen gut passt. Insbesondere die Rhetorik bedient sich dieser Mittel 75, so dass man mit Siegert sagen kann: »Gehobene Prosa ersetzt einstige Poesie« 76, oder noch zugespitzter: Aus hebräischer Poesie wird rhetorische Prosa 77. Es ist klar, dass dabei Verluste entstehen 78.

71. 72. 73. 74. 75. 76. 77. 78.

nique« Biblica 77 (1996), 227-236 gibt eine Übersicht über vier Arten, in denen hebräische »elaborierte« Vergleichssätze in der übersetzten LXX wiedergegeben werden: Durch Optativ mit der auffälligen Häufung im Pentateuch, durch Indikativ, Partizip und Konjunktiv. Joosten wertet, anders als Evans und Usener, die Verwendung des Optativs nicht als Homerismus, sondern im Anschluss an allgemeine Bemerkungen bei Thackeray, Grammar, VIII, 24 und 193 als ein Merkmal der frühen Koine, das sich im Pentateuch noch erhalten habe und später aussterbe. Außerhalb der LXX findet sich allerdings bislang nur ein einziger Beleg in einem in der Tat mit dem Pentateuch gleichzeitigen ptolemäischen Papyrus, vgl. Evans, Verbal Syntax, 194. Vgl. Norden, Kunstprosa, Bd. I, 50. Die Einzelheiten bei Siegert, Hebräische Bibel, 181-185 und Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 53-64. Der Creticus wird gern im sog. genus grande, dem hohen, feierlichen, »pathetischen« Stil verwendet. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 62-63. Vgl. die kurze Übersicht bei G. Ueding, Einführung in die Rhetorik. Geschichte – Technik – Methode, Stuttgart 1976, bes. 280. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 53. Siegert, Hebräische Bibel, 181. Siegert, Hebräische Bibel, 179 verweist auf Eusebius, der schon im 3. Jh. n. Chr. in seiner Chronik darüber klagt, dass die Schönheit des Alten Testaments in der Septuaginta nicht mehr wahrnehmbar sei. Siegert verweist insbesondere auf den Verlust von Wortspielen in den Pro-

58

5. Zielsprachlich realisierte Diversität

2. Die Diversität des Griechischen in der Septuaginta

Rhythmisierung findet sich – zusammen mit anderen rhetorischen Schmuckmitteln 79 – auch verstärkt in den Weisheitsschriften (Spr, Hld, Jer, Est), darüber hinaus in Oden (Dan, Jes, Oden 12) und in original griechisch verfassten LXX-Schriften, so im eingangs behandelten Sirach-Prolog, EpJer, Weish 1,1-14a, 2Makk 15,39, 3Makk an jedem Abschnittsende, besonders extrem in 3Makk 2,20 und in 4Makk häufig. 80 Das zweite Mittel, das die Übersetzer für die Umsetzung hebräischer Poesie in Prosa adäquat fanden und das nicht ausgangssprachlich bedingt ist, ist die electio verborum, die Verwendung eines (aus-)gesuchten Vokabulars 81. Dies trifft insbesondere für Hiob und Spr zu: Nach Siegert handelt es sich anders als in den Psalmen, die Rhythmisierung mit weitgehend unauffälligem Vokabular verbinden 82, in diesen Schriften um »Prosa unter Verwendung poetischer Wörter« 83. Spr wiesen ein »geradezu pompöses Vokabular« 84 auf, darunter 74 Wörter, die sonst nirgends in der LXX vorkommen 85, in Sir gibt es 273 Hapaxlegomena 86. Merkwürdigerweise weist auch die nicht übersetzte, sondern griechisch verfasste Weish in Rhythmisierung, bestimmten Schmuckmitteln wie der Vorliebe für Neologismen und Hapaxlegomena Ähnlichkeit mit der hier beobachteten Umsetzung von hebräischer Poesie in griechische Prosa auf, obwohl für eine griechisch verfasste Schrift ja die Möglichkeit bestanden hätte, vorhandene griechische Literaturregister zu verwenden, wie das andere jüdische Autoren getan haben. Darauf ist im Zusammenhang der original griechisch verfassten Schriften zurückzukommen, jetzt zunächst zurück zu den übersetzten Schriften. Wir haben nun mit der grundsätzlichen Ansiedlung der LXX-Übersetzungen auf der dritten und zweiten Ebene unserer Skala der Übersetzungsmöglichkeiten zwischen »frei« und »wörtlich« die sprachlich-systemlinguistische Realisierung der spezifischen

79.

80. 81. 82. 83. 84. 85. 86.

pheten und Psalmen; und 171: »Es ist eine Schwäche der jüdischen Tradition, seit sie griechisch (sic!) wurde, das Lachen zu verdächtigen, wo nicht zu verbannen«. In Ps 126/125,2 werde das Lachen in χαρά, in Spr 14,13 in εὐφροσύνη »umgemünzt«, in »Freude« und »Fröhlichkeit«. Z. B. Alliterationen (Ps 49/48,11a; 51/50,4 und 51/50,12), Homoioteleuton (Endreim, Ps 51/50,9), Anapher (Spr 13,9; Hiob 12,14 f.; 28,23 f.), Chiasmus (Hiob 30,15; Spr 10,17), Hyperbaton, (Ps 34/ 33,13b), Paronomasie (Ps 18/17,27). Vgl. im Einzelnen zu den genannten Beispielen: J. A. L. Lee, »Translation of the Old Testament« in: St. E. Porter (Hg.), Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C. – A.D. 400), Leiden 1997, 775-783; Th. J. Kraus, »Translating the Septuagint Psalms – some ›Lesefrüchte‹ and their value for an analysis of the rhetoric (and style) of the Septuagint (Psalms)« und E. Bons, »Rhetorical Devices in the Septuagint Psalter« in: Bons / Kraus, Et sapienter et eloquenter, 49-68 bzw. 69-79; ebendort 23-48 untersucht für das Dodekapropheton zielsprachliche Schmuckmittel J. M. Dines, »Stylistic Invention and Rhetorical Purpose in the Book of the Twelve«. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 60. Norden, Kunstprosa, Bd. I, 43 »[…] dass die rhythmische Rede auch ausgezeichnet war durch hochpoetische Worte und Wortverbindungen«. Zur Wortwahl der Ps im Einzelnen Siegert, Hebräische Bibel, 75. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 54. Seine Urteile allerdings über die »niedere« (Hiob) bzw. »roheste Prosa« (Weish) sind zu apodiktisch. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 54. Siegert, Hebräische Bibel, 75-76. Wagner, Hapaxlegomena, 5 gibt eine Übersicht über sämtliche Hapaxlegomena der LXX in buchspezifischer Verteilung und thematisiert 65-86 auch die Unschärfe des Begriffs, die u. a. zu unterschiedlichen Zählungen führt. 5. Zielsprachlich realisierte Diversität

59

2. Die Diversität des Griechischen in der Septuaginta

Übersetzungspragmatik konstatiert und damit die Mittel der grundsätzlich ausgangssprachlichen Verpflichtung der Übersetzer gegenüber dem heiligen Prätext benannt; dann wurden die Mittel betrachtet, mit denen sich die Übersetzer zielsprachlich orientieren, also eine gewisse Freiheit in der Übertragung des Ausgangstextes zeigen; diese sind überwiegend im paradigmatischen Bereich anzusiedeln, also in der gezielten Kommutation von Mitteln der Zielsprache; in Einzelfällen wie der Wunschformel, chiastischen Strukturen, wo es im Prätext keine gibt, oder einer Ersetzung eines hebräischen Parallelismus durch eine griechische Antithese (Spr 6,1) erreichen sie auch die syntagmatische Ausdehnung. Es sind dies alles zahlreiche, aber doch punktuell verwendete Mittel; ihre Zahl lässt sich sicher vermehren, wie dies in der Literatur durch sorgfältige Beobachtung der verschiedenen Bücher ja auch geschieht. Der tragende Grundton der Übersetzung jedoch, die bei aller Diversität für eine gewisse Einheitlichkeit der übersetzten Texte verantwortlich ist und das Griechische als »Übersetzungsgriechisch« erkennen lässt, ist die Ansiedlung der Übersetzung zwischen der oben genannten dritten und zweiten Ebene, d. h. in der syntagmatischen Gebundenheit, und das Ineinander der Ebenen. Und da die drei Ebenen nicht in Reinform auf Einzelbücher, manchmal nur auf Kapitel oder Sätze oder Wortgruppen verteilt sind, kann eine syntagmatisch durchaus gebundene Übersetzung in paradigmatischer Hinsicht freier sein, oder umgekehrt eine syntagmatisch freie Übersetzung im paradigmatischen Bereich unfrei. Insbesondere liegen oft untere und mittlere Ebene eng beieinander. Es wurde schon gesagt, dass die eigentliche Leistung der Übersetzer in der Erreichung der mittleren Ebene der grammatischen Korrektheit liegt unter z. T. kreativer Ausnutzung von oder Anknüpfung an die in der so flexiblen griechischen Sprache gegebenen Möglichkeiten. Dies kann insbesondere an einem Phänomen betrachtet werden, das in beiden Sprachen nicht denselben Wert hat; ein solches Phänomen ist der parataktische Satzbau, der charakteristisch ist für die hebräische Literatur, besonders die erzählende, und für die griechische eher nicht. Allerdings ist einzuschränken: Auch viele Koine-Texte bis hin in die literarische Koine weisen diese Bauweise auf, nur nicht so konsequent wie die LXX in ihrer syntagmatischen Abhängigkeit vom Prätext. Man spricht hier deswegen gern von »Hebraismus«, man könnte sagen, hier liege ein syntagmatischer Hebraismus vor. Aber auch da ist zu bedenken, dass die Flexibilität des griechischen Satzbaus, in dem Wortstellung ja – in bestimmten Grenzen – kein Kriterium der Richtigkeit, sondern eher der Betonung und der Ästhetik ist, diese Eigenheit aufnehmen kann, ohne dass es grammatisch unkorrekt wird. Das Auffällige aus einsprachiger Rezipientensicht ist deshalb oft, wie auch bei anderen als Hebraismus bezeichneten Eigentümlichkeiten weniger das Einzelphänomen, wie vielmehr seine ungewöhnlich hohe Frequenz; daran wird der Rezipient möglicherweise erkennen, dass es ein übersetzter Text ist, wie uns das heute durchaus bei übersetzten Texten auch geschehen kann, wenn bestimmte zielsprachlich mögliche Wendungen eine auffällig höhere Rekurrenz aufweisen als wir sie in muttersprachlich verfassten Texten gewohnt sind. Die im Folgenden referierten Befunde sind am Pentateuch erhoben; dieses Textkorpus ist einerseits für die Beobachtung von Diversitäten insofern besonders interessant, als der Pentateuch früh übersetzt ist, also noch nicht den Revisionen zur immer mehr wörtlichen, »sklavischen« Wiedergabe unterliegt, also Diversität in einem eher als geschlossen und zielsprachlich auch eher hoch bewerteten Korpus zu beobachten 60

5. Zielsprachlich realisierte Diversität

2. Die Diversität des Griechischen in der Septuaginta

ist; und andererseits bietet er eben für die Beschreibung des Ineinander von Gebundenheit und Freiheit eine ausgedehntere Textbasis als Einzelstellen. Die Einzelphänomene selbst lassen sich auch in anderen Schriften beobachten, sind also gewisse Universalien der LXX-Übersetzung geworden. Zudem liegen für den Pentateuch Untersuchungen der Übersetzungstechnik vor, die beispielhaft für eine Erforschung der Übersetzungstechnik sind, die sowohl die Gebundenheit als auch die Freiheit reflektiert 87; es wird dabei auch deutlich, in welch kleiner Bandbreite sich die Freiheit bewegt.

6. Diversität im Pentateuch I: Bindung der Freiheit durch Parataxe Im hebräischen Pentateuch liegen über 10.000 mit waw parataktisch verbundene Sätze vor, über die Hälfte aller Sätze (56 %) beginnen mit waw. Zusammen mit der stilistischen Monotonie ergibt sich semantisch eine gewisse Unschärfe: Parataxe macht das logische Verhältnis zwischen den beiden Teilen nicht so deutlich wie Hypotaxe mit einer i. A. semantisch stärker markierten Subjunktion. Im Pentateuch wird in 76 % der Fälle die naheliegende Wiedergabe mit dem semantisch offenen καί vorgenommen, insgesamt beginnen 44 % aller Sätze mit καί. Hier äußert sich der grundsätzlich interlinguale Übersetzungscharakter: Diese Wiedergabe ermöglicht auch am ehesten die Beibehaltung der hebräischen Wortabfolge und ist überhaupt das Mittel der einfachen Technik, die auf Wort- oder Wortgruppenniveau übersetzt; Subjunktionen verlangen den Überblick über größere syntaktische Einheiten und semantische Verhältnisse. Es gibt dabei hohe Diversität zwischen den Büchern: Gen hat mit 63 % die wenigsten, Lev mit 90 % die meisten καί-Verbindungen zwischen Sätzen. Zielsprachlich besteht die Auffälligkeit in der Frequenz 88. Es wird allerdings an den Prozentzahlen der waw-Koordination im Hebräischen (56 %) gegenüber den 44 % von καίSatzverknüpfungen auch deutlich, dass es in der Übersetzung des waw auch Abweichungen gibt, es liegt keine 1:1 Entsprechung vor; darüber hinaus gibt es sogar Fälle, in denen καί kein hebräisches Äquivalent hat; es ist wohl gesetzt, um zielsprachlich ein ungriechisches Asyndeton zu vermeiden 89. Unter den verschiedenen Mitteln, die für die abweichende Wiedergabe von waw verwendet sind, stechen quantitativ die Adjunktion δέ und das Participium coniunctum deutlich heraus 90, das erste in original griechischen Texten jeder Stilebene viel häufiger als καί, das zweite eines der typischsten Mittel der griechischen Zielsprache; es liegen hier also eindeutig zielsprachliche Orientierungen vor, das Bemühen, Standards der Zielsprache zu treffen. Gerade in der Part.con.-Verwendung ist einerseits die Wendung zu griechischen Idiomatik und damit zur größten Freiheit der obersten Ebe87. Aejmelaeus, Parataxis, und Evans, Verbal Syntax. 88. Aejmelaeus, Parataxis, 32 nennt Vergleichszahlen zu einem griechischen Koine-Korpus: In 2Makk sind 17 % der Sätze mit καί verbunden, im Vergleich zu anderen originalen KoineTexten die höchste Zahl: Aristeas und Epikur haben 6 %, Polybius 8 %, auch Papyri nur zwischen 6-10 %. 89. Wohl aber findet es sich in etlichen Fällen im samaritanischen Pentateuch, was die Frage nach der übersetzten Vorlage aufwirft (Aejmelaeus, Parataxis, 157). 90. Tabelle bei Aejmelaeus, Parataxis, 123. 6. Diversität im Pentateuch I

61

2. Die Diversität des Griechischen in der Septuaginta

ne zu sehen; es dokumentiert auch die Fähigkeit, größere Einheiten als ein Wortpaar zu überblicken und zu meistern. Wenn es z. B. größere Abstände zwischen Partizip und finitem Verb gibt, muss die Syntax eines ganzen Satzes berücksichtigt werden. Und gleichzeitig sieht man doch auch, dass die Freiheit vom Ausgangstext begrenzt ist: Das Participium coniunctum ist – noch mehr die zweite griechische adverbiale Partizipialkonstruktion Genitivus absolutus – zahlenmäßig limitiert 91, Tempus und Wortstellung zeigen wenig Variation, letztere ist bedingt vom Hebräischen, und es ist eine semantisch-lexikalische Beschränkung zu verzeichnen: Die Verben im Partizip sind begrenzt auf Verben der Bewegung des Typs »sich erheben, aufbrechen« etc., der Wahrnehmung, der Redeeinleitung für hebräisch ‫לאמר‬, und Ausdrücke, die ein Nehmen bezeichnen: λαβών 92, dadurch wird der Gebrauch eher monoton und die zielsprachliche Idiomatik begrenzt. Man sieht also nun an diesen Beispielen der abweichenden Wiedergabe von καί zum einen das Bemühen um eine zielsprachlich adäquate Wiedergabe, zum anderen doch auch die recht engen Grenzen, in denen dies geschieht, die Grenzen der Ausdehnung wie der Mittel. Die Leistung der Übersetzer liegt über weite Strecken in der Realisierung der grammatisch richtigen Übersetzung, die auf unserer Skala die mittlere Ebene bildet. Die Leistung wird bestätigt durch das gelegentliche Unterschreiten der grammatischen zielsprachlichen Richtigkeit in den Fällen, wo die typisch hebräische Eigenart, eine Apodosis nach vorhergehender Protasis mit waw einzuleiten, in einem Verstoß gegen das Griechische nachgeahmt wird. Die Protasis kann dabei unterschiedlich sein: konditionale, temporale, relative Nebensätze, aber auch Infinitivkonstruktionen, nominale Temporalangaben oder Casus pendens (Dtn 21,3). Ebenso kann eine der beiden Übergangsformeln καὶ ἐγένετο/ἔσται die Protasis bilden (Gen 22,20). Der Gebrauch des apodotischen waw beläuft sich im hebräischen Pentateuch auf 62 %. Im griechischen Pentateuch finden sich drei Wiedergabeformen: Die Wiedergabe durch das sehr ungriechische καί 93, das gut griechische δέ, und das griechisch normale Auslassen einer Adjunktion. Der apodotische Gebrauch von καί erreicht mit 211 in 415 Fällen eine Proportion, die in genuin griechischen Texten unbekannt ist; hier liegt in der Zielsprache eindeutig ein Hebraismus vor, verstärkt noch durch die Kombination mit den beiden Übergangsformeln ἔσται bzw. ἐγένετο + finite Verbform, die asyndetisch oder eben häufiger mit καί angeschlossen sind; korrektes Griechisch wäre hier der Anschluss eines untergeordneten Satzes oder sein Äquivalent AcI 94, was in der LXX nicht vorkommt. In ähnlicher Weise verstoßen die Fälle, in denen ein Participium coniunctum mit dem finiten Verb durch καί verbunden ist (z. B. Gen 21,9-10; Ex 12,37-38; Lev 4,5 95), gegen die griechische Syntax; hierher gehört die schon genannte 91. Im Pentateuch ein Partizip pro 15-20 Sätze, am dichtesten in der Genesis, gegenüber einem Partizip auf 2-10 Sätzen im Vergleichskorpus von Koine-Texten. 2Makk und Polybius haben ein Vorkommen eines Partizips auf zwei Sätze, was im Pentateuch selten ist. 92. Aejmelaeus, Parataxis, 89 und 105. 93. In 211 von 415 Vorkommen des apodotischen waw; Gen hat 42/101; Ex 16/73; Lev 66/111; Num 41/64, Dtn 46/66. 94. Thackeray, Grammar, 50-52. 95. In der Gen zählt Ajemelaeus sieben solcher Fälle bei insgesamt 208 Part.con., in Ex zwei auf 89, in Lev drei auf 21, Num vier auf 58, und Dtn vier auf 75. In Gen 29,9 wird sogar ein Gen.abs. mit einem καί verbunden.

62

6. Diversität im Pentateuch I

2. Die Diversität des Griechischen in der Septuaginta

hebraistische Wendung προστίθημι für ‫ ;יסף‬auch hier gibt es kleine zielsprachliche Variationen: προστίθημι mit folgendem Infinitiv (109 Vorkommen, z. B. Gen 4,2), die wörtlichste Abbildung des Hebr., mit Anschluss eines finiten Verbs durch καί (9 Vorkommen), und schließlich die Umwandlung von προστίθημι in ein Partizip, gefolgt von einer finiten Verbform; das letzte entspricht am ehesten griechischem Gebrauch und führt somit bereits in die mittlere Ebene, kommt aber nur dreimal vor 96: Gen 25,1 προσθέμενος ἔλαβεν γυναῖκα – »indem er dazusetzte, nahm er eine Frau«, d. h. »er heiratete erneut«; dazu Hiob 27,1 und Est 8,3. Warum ein Übersetzer, wie an diesem Fall des apodotischen καί zu beobachten, einmal mehr zielsprachlich, das andere mal ausgangssprachlich übersetzt 97, ist schwer zu beurteilen und liegt vielleicht an der jeweiligen Einzelstelle und deren unterschiedlicher Komplexität. Eine kleinteilige Wort- oder Wortgruppenübersetzung kann größere und komplexe Zusammenhänge, wie sie längere Protasis-Apodosis-Kombinationen darstellen, leichter aus dem Blick verlieren; dann wird καί gesetzt 98. Es mag dazu treten, dass die nichtliterarische mündliche Koine dafür auch Vorbilder bot, die den Verstoß gegen den griechischen Satzbau als nicht so gravierend erscheinen ließen 99. Das Kriterium der Verständlichkeit ist in all diesen Fällen nicht wirklich verletzt. Ähnliches gilt für die weiteren bekannten zielsprachlichen Auffälligkeiten wie die »physiognomischen« Wendungen 100, die vielfache Tendenz zu analytischer statt synthetischer Verbvalenz mit teilweise nicht ganz passenden Präpositionen 101, die Übersetzung des Infinitivus absolutus, für den entweder die im Griechischen ja auch bekannte figura etymologica – allerdings mit dem Nomen im Dativ statt des mehr literarischen Akkusativs (Gen 2,17 θανάτῳ ἀποθανεῖσθε) – verwendet wird 102 oder die Doppelung der 96. Thackeray, Grammar, 52. 97. Im Pentateuch betrifft das etwa die Hälfte der Fälle, deutlich am wenigsten in Ex mit 22 %. 98. So zeigt der eigentlich nicht freie Lev im letzten Kapitel bei wenig ausgedehntem Umfang der Protasis einen ungewöhnlich hohen Anteil von nicht übersetztem waw. 99. Aejmelaeus, Parataxis, 127 weist auf vergleichbare Erscheinungen in der Koine hin, wo das apodotische καί als Vulgarismus gilt. In der Literatur verwenden Autoren gerade in komplexem Satzbau mitunter das Stilmittel des Anakoluth, um gesprochene Sprache abzubilden, etwa Platon Apol. 19e4. 100. Im Hebräischen werden insbesondere Präpositionen mit Körperteilen wie »Auge, Mund, Ohr, Hand, Gesicht« verbunden, z. B. »sprechen zum Ohr von jdm.«; sofern dies wörtlich abgebildet wird – mit steigender Frequenz wieder in den späteren, wörtlicheren Übersetzungen – entstehen zielsprachlich eigentlich überflüssige Kombinationen von Präposition + ὀφθαλμός, στόμα, οὖς, χείρ, πρόσωπον + Attribut wie z. B. εἶπεν εἰς τὰ ὦτα τοῦ λάου (Gen 23,13) statt der direkten Wendung εἶπεν πρὸς τὸν λάον, oder ἐρωτήσωμεν τὸ στόμα αὐτῆς (Gen 24,57) statt ἐρωτήσωμεν αὐτήν, oder, übersetzungstechnisch besonders interessant Gen 19,21 ἐθαύμασά σου τὸ πρόσωπον »ich bewunderte dein Gesicht« statt ἐθαύμασά σε. Letztere Wendung versucht, den idiomatischen Ausdruck ‫» נשא פנים‬ich hebe das Gesicht« im Sinne von »ich nehme Rücksicht, zeige Respekt« gleichzeitig ausgangssprachlich abzubilden (πρόσωπον) und zielsprachlich doch verständlich zu machen (θαυμάζω); zur Stelle J. W. Wevers, Notes on the Greek Text of Genesis (Septuagint and Cognate Studies 35), Atalanta 1993, 278; generell Thackeray, Grammar, 42-45, und durchgehend die Monographie von Sollamo, Renderings of Hebrew Semiprepositions. 101. Helbing, Kasussyntax, 324; Thackeray, Grammar, 46-47. 102. Thackeray, Grammar, 47-48. 6. Diversität im Pentateuch I

63

2. Die Diversität des Griechischen in der Septuaginta

finiten Verbform durch das ebenfalls griechische paronomastische Partizip (Gen 3,16: πληθύνων πληθύνω), oder in Relativsätzen die pleonastische Wiederholung des Relativpronomens oder –adverbs durch ein Personalpronomen (Gen 19,29) oder eines Adverbs (Ex 24,10) 103. Für manches finden sich im Griechischen Ansätze oder Analogien.

7. Diversität im Pentateuch II: Förderung der Freiheit durch Texttyp und -inhalt Eine weitere rezipientenbezogene Beobachtung eröffnet einen möglichen zusätzlichen Faktor für Diversität der LXX: Der zielsprachlich-griechische Leser, der des Hebräischen nicht mächtig ist, kann speziell bei der ja überhellenden Tätigkeit der Übersetzung des griechischen Textes der Gen in eine moderne Sprache den Eindruck gewinnen, als gehe das Übersetzen in bestimmten Passagen plötzlich flüssiger, mit weniger Anstößen und erklärungsbedürftigen Hindernissen im griechischen Text. Das ist in der Genesis im zweiten Teil der Fall, insbesondere bei den Josephserzählungen. Nun könnte das auf Gewöhnung und Erfahrung zurückzuführen sein, die man sich im Laufe der Übersetzung erwirbt. Ähnlich könnte es sich produzentenbezogen verhalten: Auch der Übersetzer aus dem Hebräischen könnte von der wachsenden Erfahrung und Kompetenz profitiert haben; er könnte aber auch schon in seiner Vorlage Merkmale gefunden haben, die sein Übersetzen flüssiger, freier, zielsprachlich unauffälliger werden lassen. Dies könnte an bestimmten Texttypen und Inhalten liegen. Das legt die Verteilung freierer Wiedergaben im Pentateuch nahe 104: Sie finden sich besonders in narrativen Texten: An der Spitze der freien Wiedergabe steht Ex Kap. 2, in der Gen, die hauptsächlich aus narrativen Passagen besteht, ist auffällig die Josephsgeschichte (nicht der poetische Text in Kap. 49), und einige andere Kapitel (Kap. 25 und 28), wobei auffälligerweise die Frühgeschichte ausgangssprachlicher ist (Ausnahmen Kap. 14 und 18), also die wenigsten freien Elemente enthält. Im zweiten Teil bilden neben Kapitel 49 auch noch die Kapitel 39 und 40 klare Ausnahmen von der freieren Übersetzungstechnik. Eine gewisse Kontinuität zeigt sich buchübergreifend zwischen dem zweiten Teil der Gen und der ersten Hälfte von Ex, wo die zweite Hälfte mit Gesetzen kaum freie Wiedergaben enthält, mit Ausnahme der eingebetteten Narrationen in Kapitel 32 (goldenes Kalb) und 24 (Bund). Lev hat hauptsächlich Gesetze; freie Wiedergabe ist selten, auch nicht in den narrativen Kapiteln 8-10. Num enthält zur Hälfte Erzählung (Auszug in die Wüste), die andere Hälfte sind Zählungen und Gesetze. Die narrativen Kapitel unterscheiden sich leicht von den Kapiteln ohne freie Wiedergabe. Kapitel 21-24 (Schlange und Balaam), narrativ, enthalten die meisten freien Wiedergaben in Num. Dtn schließlich hat einen narrativen Rahmen in den Kapiteln 1-3 und 34. Die Gesetzespassagen unterscheiden sich von den anderen im Penta-

103. Gen 19,29: τὰς πόλεις, ἐν αἷς κατῴκει ἐν αὐταῖς Λώτ – »die Städte, in denen in ihnen Lot gewohnt hatte«; Ex 24,10: τὸν τόπον, οὗ εἱστήκει ἐκεῖ ὁ θεός – »den Ort, wo der Gott (Israels) dort stand«. 104. Aejmelaeus, Parataxis, 159-184, bes. 164-168.

64

7. Diversität im Pentateuch II

2. Die Diversität des Griechischen in der Septuaginta

teuch durch lange Reden, Aufforderungen und Ermahnungen. Die dort seltenen freien Wiedergaben sind einigermaßen gleichmäßig auf die Inhalte verteilt. Eine Häufung der untersuchten freien Wiedergaben findet sich also in narrativen Passagen; Genealogien, Kultvorschriften und Gesetze bleiben deutlich dahinter oder weisen keine der untersuchten freien Wiedergaben auf. Allerdings gibt es auch in den narrativen Passagen erhebliche Unterschiede zwischen benachbarten Kapiteln, freie Wiedergaben können von einem Kapitel zum anderen zu- und abnehmen. Es lässt sich somit vom Texttyp her allein keine klare Kontinuität über Kapitel hinweg feststellen. Auch die statistisch festgestellte Divergenz in freieren Wiedergaben zwischen den beiden Teilen jeweils von Gen und Ex 105 scheint nicht nur vom narrativen Texttyp und auch nicht von einer konsistent durchgeführten Übersetzungstechnik abzuhängen; eine entscheidende Rolle scheinen bestimmte Inhalte zu spielen, die für die Übersetzer nicht nur interessanter, sondern vielleicht auch bekannter waren; sie betreffen ja auch stärker die in Ägypten lebenden Juden als etwa Tempelvorschriften; möglicherweise sind die Übersetzer mit diesen Erzählungen auch durch bereits vorhandene mündlich kursierende Übersetzungen und Wiedererzählungen vertraut. Abschnitte mit direkter Rede könnten zudem das Einfließen von Duktus und Vokabular der selbst gesprochenen griechischen natürlichen Alltagssprache begünstigt haben 106. Wenn man nun also die unterschiedlichen Kriterien und produzenten- wie rezipientenorientierte Blickweisen zusammennimmt, ergeben sich im Pentateuch, der als eher einheitlich übersetzt angenommen wurde, doch erhebliche Diversitäten 107: Ex kann unter dem Aspekt der syntagmatischen Freiheit von der hebräischen Wortabfolge und der Berücksichtigung zielsprachlicher Standards als der kompetenteste Übersetzer angesehen werden, gefolgt vom Übersetzer der Gen, der ebenfalls zielsprachlich orientiert ist, wenn auch grammatisch nicht so korrekt wie der Ex-Übersetzer; Lev, Num und Deut sind in syntagmatischer Hinsicht als deutlich ausgangssprachlicher anzusehen 108; es scheint also möglicherweise bereits etwa ab der Mitte des doch frühen Pentateuchs die vorher schon genannte diachrone Tendenz zu einer wörtlicheren, d. h. mehr am Einzelwort und der hebräischen Wortabfolge orientierten Übersetzungstechnik aufzutreten. Gleichzeitig lässt sich in paradigmatischer Hinsicht eine größere Beweglichkeit in der individuellen Wiedergabe einzelner Wörter in dem syntagmatisch unfreien Lev feststellen 109.

105. Gen ist im zweiten Teil, Ex im ersten Teil freier übersetzt. 106. Aejmelaeus, Parataxis, 173. Evans, Verbal Syntax, 216 trägt zur Unterstützung dieser Hypothese die Beobachtung einer auffälligen Häufung des Imperfekts in solchen Passagen gegenüber dem sonst im Sinne der easy technique überwiegend verwendeten »unmarkierten« Aorist bei. 107. Aejmelaeus, Parataxis, 180. 108. Bis hin zu Verstößen gegen die griechische Sprache, vgl. Dtn 29,17, wo das hebräische ‫אשר‬ syntaktisch völlig falsch und sinnlos mit dem Interrogativpronomen τίνος abgebildet wird. 109. Aejmelaeus, Parataxis, 180. 7. Diversität im Pentateuch II

65

2. Die Diversität des Griechischen in der Septuaginta

8. Original griechisch verfasste Schriften Wir kehren zum Schluss zurück zu den nicht übersetzten, also original griechisch verfassten Schriften, hier fokussiert auf 2–4Makk und Weisheit des Salomo. In der Sprachgestaltung gibt es einen erstaunlichen Unterschied: Während die Makkabäerbücher – in unterschiedlicher Ausprägung – die Merkmale eines (Koine-)literarisch verfassten Originaltextes zeigen, weist Weish neben rhetorischen Elementen viele Besonderheiten auf, die für die übersetzten Texte der LXX typisch sind, d. h. der Verfasser scheint dieses für die übersetzten Texte charakteristische Ineinander von ausgangssprachlicher Bindung und zielsprachlichen Freiheiten nachzuahmen. Eine Gemeinsamkeit zwischen diesen nicht übersetzten Büchern gegenüber der übrigen LXX ist das Vorkommen von Vokabeln, die sonst eher gemieden oder kaum verwendet erscheinen, vermutlich wegen ihrer paganen Konnotation, z. B. ἡδονή (Weish, 4Makk), κόσμος (Weish), ἀθάνατος (Weish, 4Makk), φιλόσοφος (4Makk), ἀρετή (bes. 4Makk). Zunächst zu den Makkabäerbüchern: Sie weisen alle den für griechische Literatur üblichen hypotaktischen Satzbau auf, mit vielen Partizipialkonstruktionen und komplexem Periodenbau, die Verwendung der Adjunktionen ist variantenreich, ebenso das Vokabular. So zeigen in 2Makk z. B. die Wortfelder »töten« und »sterben« eine breite Auffächerung 110; es gibt zwischen den drei Büchern Vokabelgemeinschaften, die außerhalb der Bücher selten oder gar nicht vorkommen, verteilt nach inhaltlichem Schwerpunkt und Gattungsspezifik: 2Makk ist die Zusammenfassung eines hellenistischen, aus jüdisch-religiöser Sicht verfassten Geschichtswerks mit vielen Fachbegriffen der seleukidischen Verwaltung, 3Makk ist eher eine fromme Legende; es wird ihm ein anspruchsvoller, bemühter, aber wenig schöner Stil attestiert 111: »Schwerlich hat einem anderen biblischen Schriftsteller ein so reicher Sprachschatz zu Gebote gestanden, […]; aber derselbe hat einen üblen Gebrauch davon gemacht« 112. 4Makk ist eine philosophische Abhandlung über den stoischen Satz, dass in der erreichten ἀρετή die Vernunft die Herrscherin über Triebe und Affekte ist, auch im Zustand der höchsten Bedrängnis; dies wird am Martyrium des jüdischen Priesters Eleazar exemplifiziert. Das Buch ist in Form einer Rede strukturiert 113: exordium – narratio mit Definitionen – argumentatio mit ausgedehnten exempla und Amplifikation in der Form von Rededuellen – peroratio; es gehört dem epideiktischen Genus an. Die rhetorischen Stilregeln sind befolgt: Während im ersten Teil (narratio) eher eine einfache und sachliche Sprache, also das sogenannte genus tenue verwendet ist 114, werden im zweiten mehr epideiktischen Teil mit den Reden alle Register des hohen Stils gezogen zur Erregung von Pathos 115. Es ist der ausladende schmuckreiche asianische Stil, den wir schon im Sirach-Prolog kennengelernt haben. Nach Nordens Urteil lässt sich kaum 110. Im Einzelnen T. Nicklas, »Makkabaion II« in: Septuaginta Deutsch. Erläuterungen, Bd. I, 1376-1377. 111. A. Dihle, Griechische Literaturgeschichte, Stuttgart 1967, 413. 112. So ein Urteil aus dem 19. Jh., zitiert bei T. Knöppler, »Makkabaion III« in: Septuaginta Deutsch. Erläuterungen, Bd. I, 1418. 113. Im Einzelnen H. J. Klauck, »Makkabaion IV« in: Septuaginta Deutsch. Erläuterungen, Bd. I, 1445-1456. 114. Etwa zu Beginn der narratio 1,13-30. 115. Z. B. 7,1-3 oder in der peroratio 18,20-21.

66

8. Original griechisch verfasste Schriften

2. Die Diversität des Griechischen in der Septuaginta

eine Schrift nachweisen, »die auf einem so verhältnismäßig kleinen Raum eine so ungeheure Zahl von hochpoetischen und meist mit souveräner Willkür neugebildeten Worten enthält« 116. Der Autor beherrscht sowohl sachlich sein Thema und die Fachterminologie wie die rhetorische Technik souverän, er kann anders als der Verfasser des Sirach-Prologs seine formale wie inhaltliche griechische παιδεία in seiner Schrift vereinen, ausstellen und der jüdischen theologischen Aussage dienstbar machen. Der angezielte Adressatenkreis kann hier sicher auf das gebildete Griechentum ausgedehnt sein, »als Mahnung an griechisch gebildete Juden, als Werbung an gebildete Griechen« 117. Während die Makkabäerbücher also griechische Literaturgattungen und Standards verwenden, wie es andere jüdische Autoren auch für poetische Texte getan haben 118, scheint der Autor von Weish dies zu verweigern. Er verwendet statt eines poetischen Versmaßes den schon beschriebenen Prosarhythmus, der für die Umsetzung hebräischer Poesie in den Psalmen und Hiob typisch ist; und er mischt seine griechische παιδεία, d. h. einen gesuchten hochpoetischen Wortschatz, reichen rhetorischen Schmuck (Paronomasien, sehr viele Hyperbata, Chiasmus, Anaphern, Homoioteleuton, Alliterationen und Assonanzen) mit Elementen, die für die übersetzte LXX typisch sind, insbesondere dem hebräischen parallelismus membrorum, der komplexen griechischen Periodenbau unmöglich macht, sodann mit typischem LXX-Vokabular wie mit Kompositionselementen, z. B. der figura etymologica (9,3 κρίσιν κρίνῃ nach Gen 19,9 und öfter), oder der »physiognomischen« Verwendung von χείρ in Kombination mit einer Präposition 119. Der Autor scheint den LXX-Stil also zu imitieren; bei näherem Hinsehen zeigt es sich aber, dass er diese Elemente so modifiziert, dass sie – anders als in der übersetzen LXX – der griechischen Korrektheit nicht zuwider laufen, und dass er dies sozusagen ornamental, nicht substantiell tut. Léonas spricht von einem pastiche der LXX, einem remake von Sprache und Stil, zur Erzielung eines eigenen ästhetischen Effekts 120; man fühlt sich ein bisschen an postmoderne Verwendung historischer Elemente erinnert. Dies spricht aber auch für den Reiz des Vorbildes, also dem Übersetzungsstil der LXX, das in seiner Künstlichkeit – Siegert spricht von »Geschraubtheit« 121, das Modell für feierlich-religiöse Sprache bis weit in die Neuzeit hinein geblieben ist, wie nicht zuletzt die Nachahmung dieses Übersetzungsstils in modernen Bibelübersetzungen zeigt. Der Grund für diesen eigenartigen Sprachcharakter ist die zweifache Diversität, einmal nach innen in der Mischung ausgangssprachlicher fremder Elemente mit unauffällig zielsprachlichen, nach außen in der Abständigkeit von einem natürlichen Sprachgebrauch. Voraussetzung für diesen Effekt ist die spezifische Übersetzungspragmatik, die die textinternen systemlinguisti116. Norden, Kunstprosa, Bd. I, 420. 117. A. Dihle, Die griechische und lateinische Literatur der Kaiserzeit. Von Augustus bis Justinian, München 1989, 175. 118. Z. B. Ezechiel der Tragiker, der den Auszug aus Ägypten in iambischen Trimetern beschrieb, Philo poeta, der die Geschichte Jerusalems in Hexametern verfasste oder die ebenfalls hexametrischen Sibyllinischen Bücher. 119. Im Einzelnen A. Leónas, »The Poetics of Wisdom: Language and Style in the Wisdom of Solomon« in: Bons / Kraus, Et sapienter et eloquenter, 99-126. 120. Leónas, »Poetics«, 124. 121. Siegert, Septuaginta Deutsch. Erläuterungen, Bd. I, 63. 8. Original griechisch verfasste Schriften

67

2. Die Diversität des Griechischen in der Septuaginta

schen Entscheidungen präjudiziert und die Adäquatheit und Äquivalenz der Mittel leitet. Sie ist dafür verantwortlich, dass die Übersetzer »Treue« zum Ausgangstext enger verstehen als das nach unseren Begriffen technisch nötig wäre, es ist – nicht verwunderlich bei den zu übersetzenden Texten – eine religiös bestimmte Treue. So wird die griechische Sprache eingesetzt, um jüdische Lehre zur jüdischen Gemeinde zu bringen, und sie bildet gleichzeitig eine Abgrenzung gegen die griechisch-pagane literarische Kultur.

68

8. Original griechisch verfasste Schriften

II The Greek of the Septuagint as Hellenistic Greek

3. The Phases of the Greek Language Geoffrey Horrocks

1. Introduction 1 Greek constitutes a branch of the Indo-European language family in its own right, though prehistorically it seems to have formed a particularly close relationship with Armenian. It also has significant points of correspondence with the Indo-Iranian and Balto-Slavic languages. Linguists remain unsure whether the isoglosses involved reflect early genetic subgrouping(s) or result from subsequent areal convergence. 2 Having developed its own distinctive identity in the Balkan peninsula in the period after circa 2000 BCE, Greek was carried both east and west across the Mediterranean by successive waves of colonists in the half millennium after circa 1000 BCE, and the language eventually came to be spoken in a ring of coastal cities from southern Spain, France, Italy, and Sicily, via the Balkans, to Asia Minor, Egypt, and the lands around the Black Sea. Thanks to the conquests of Alexander the Great in the fourth century BCE and the broadly philhellenic policies of the Romans who later took control of many of these territories, Greek quickly became the principal lingua franca of the eastern Mediterranean and was eventually established, alongside Latin, as one of the two official and cultural languages of the Roman Empire. Widely written and spoken as both a first and a second language throughout the eastern provinces, Greek retained its international pre-eminence in the multi-ethnic Byzantine state that continued the eastern Roman Empire into the middle ages (its capital, Constantinople was founded in 324 CE). But the Arab conquest of the middle east and north Africa in the seventh century CE deprived Byzantium of its richest provinces, and despite eventual recovery, Byzantine power again declined sharply from the end of the twelfth century CE under pressure from both the Islamic east and the Catholic west, with further extensive territorial losses. Without the structural support of Byzantine institutions, Greek in former Byzantine lands retreated quickly before resurgent local languages (Coptic, Syriac, and Armenian) and the languages of the conquerors (Arabic, Turkish, and Romance), and its speakers came to be heavily concentrated in the Greek-speaking heartlands of the Balkan Peninsula and Asia Minor. The Byzantine capital, Constantinople, fell to the Ottoman Turks in 1453, and though Greek retained something of its status as a language of diplomacy with western powers, its overall importance declined steadily in the following centuries. When 1. 2.

For a fuller treatment of all the topics covered here see G. C. Horrocks, Greek: A History of the Language and its Speakers (second edition), Chichester/Malden, MA 2010. See J. P. T. Clackson, The Linguistic Relationship between Greek and Armenian, Oxford/ Cambridge, MA 1994, and J. P. T. Clackson, Indo-European Linguistics, Cambridge/New York, NY 2007, 1-27. 1. Introduction

71

3. The Phases of the Greek Language

Greece obtained its independence in 1832, Greek became the official language of the new state, which at this time comprised the lands south of a line from Arta to Volos. Through warfare and negotiation much additional territory was acquired over the next century and many Greek speakers who had previously lived outside Greece eventually came to live within its extended borders (along with speakers of other languages, including Albanian, Bulgarian, and Turkish). But conflict with Turkey after the First World War led to the deportation of the Christian population of Asia Minor, and Greek today is spoken principally in Greece and Cyprus. There are, however, large expatriate communities in Germany, the United Kingdom, the United States, and Australia, chiefly as a result of earlier economic migrations.

2. Ancient Greek 2.1 Origins and Earliest Attestation The language we call Greek is believed to be the product of contact between the indigenous populations of the Balkan Peninsula and Indo-European immigrants beginning around 2000 BCE. Many familiar elements of Greek vocabulary are non-IndoEuropean loan words from this time, though the lack of independent external evidence means that we are now unable to identify the donor languages involved. 3 Greek has the longest continuous recorded history of any modern European language, beginning with the dialect used for administrative purposes by the Mycenaean civilization of the late Bronze Age (so named after Mycenae in the Peloponnese). Mycenaean Greek is attested in several corpora of clay tablets dating from the second half of the second millennium BCE. These documents are written in the Linear B syllabary first deciphered by the British architect Michael Ventris in the nineteen-fifties, and record details of the economic activity of certain regions of southern Greece (Pylos and Mycenae in the Peloponnese, Thebes in Boeotia) and Crete (Knossos), each of which was controlled by an administrative center or “palace.” In each case the preservation of the surviving tablets is due to accidental firing when the palace in question was burned, whether as a result of natural disaster, hostile human intervention, or both. 4 The final collapse of the Mycenaean civilization took place around 1200 BCE and is associated with the general decline of the Bronze Age civilizations of the eastern Mediterranean at that time, including that of the Hittites. In Greece writing disappeared for several centuries and we enter a “dark age” without written documents. During the late ninth and early eighth centuries BCE, however, writing was reintro3.

4.

See R. Drews, The Coming of the Greeks: Indo-European Conquests in the Aegean and the Near East, Princeton 1988; G. Klingenschmitt, “Die Verwandschaftsverhältnisse der indogermanischen Sprachen” in: J. E. Rasmussen (ed.), In honorem Holger Pedersen, Wiesbaden 1994, 235-251; and A. Garrett, “A new model of Indo-European sub-grouping and dispersal” in: S. Chang / L. Liaw / J. Ruppenhofer (eds.), Proceedings of the 25th Annual Meeting of the Berkeley Linguistics Society, Berkeley, CA 1999, 146-156 for a range of views. See M. Ventris / J. Chadwick, Documents in Mycenaean Greek (second edition), Cambridge 1973 and A. Morpurgo Davies / Y. Duhoux (eds.), Linear B: A 1984 Survey, Louvain-laNeuve 1985.

72

2. Ancient Greek

3. The Phases of the Greek Language

duced in the form of an adaptation of the Phoenician writing system in which redundant consonant signs were redeployed to represent vowel sounds, thus creating the first true alphabet. Subsequently, different regions developed different versions of the new writing system, including the various native peoples of Italy, who adopted and then adapted their own alphabets from those brought by Greek colonists. The Ionic variant (traditionally used along the coasts of Asia Minor and in many northern and central Aegean islands) was eventually accepted as the norm for writing Greek from the beginning of the fourth century BCE. The earliest surviving alphabetic inscriptions in Greek, dating from the early eighth century BCE, come from Greek colonies in Italy and are scratched on pottery, but thereafter the volume of epigraphic material increases steadily, and sizeable collections of official and personal inscriptions on both stone and bronze survive from most parts of the Greek-speaking world from the fifth century BCE onwards.

2.2 Dialects in the Classical Period The destruction of the Mycenaean civilization and the advent of a harsher economic climate led to extensive migration during the Greek dark age. Over time, as noted, colonies were established throughout the Aegean, along the coasts of Asia Minor and the Black Sea, in north Africa, in Sicily and southern Italy (Magna Graecia), and as far west as Marseilles and the Mediterranean coast of Spain. These colonies at first spoke the dialects of their mother cities, but changes soon began to develop and spread locally, leading to new varieties, a process reinforced by growing political autonomy and regional population movements. Nonetheless, an acute sense of shared cultural identity, supported by widespread trade and travel, ensured that the many different dialects remained mutually comprehensible. Despite the formation of alliances between different cities at different times, there was no unified Greek state prior to the imposition of Macedonian rule in the fourth century BCE. The earlier political fragmentation of the Greek world, with its competitive city states, meant that there was neither the will nor the means to promote the development of a national standard at that time. Each city simply used its own dialect for official purposes, and we therefore have at our disposal, despite some geographical and chronological unevenness, a rich store of dialect data that is unparalleled in the written record of any other ancient language. The dialects of the classical period, i. e. the fifth and fourth centuries BCE, are standardly divided, according to their linguistic attributes, into “East-Greek” and “West-Greek” subgroups (the names reflect the overall geographical distribution of the majority of their members at this time). The East-Greek group comprises, on the one hand, Attic (the dialect of Athens and Attica) together with the closely related Ionic (the dialects of the central and northern Aegean and much of the Asia Minor coast), and, on the other hand, Arcadian (spoken in the mountainous central Peloponnese) together with the geographically remote Cypriot. The West-Greek group comprises Peloponnesian Doric (spoken in the rest of the Peloponnese apart from Elis, many southern Aegean islands, and much of Magna Graecia) and North West Greek (spoken in Elis and in areas the north of the Corinthian gulf). The Aeolic dialects (spoken in Thessaly, Boeotia, the island of Lesbos, and adjacent territory in Asia Min2. Ancient Greek

73

3. The Phases of the Greek Language

or) seem to have been originally of mainly West-Greek type but with an early admixture of East-Greek elements. They probably once formed a geographical “bridge” between the two major subgroups, though they subsequently developed their own distinctive identity in the post-Mycenaean period:

The Bronze-Age Mycenaean dialect, however, already has a clearly “East-Greek” character. It is also apparent from the find-spots of Linear B tablets that Mycenaean was widespread in the south, both in Crete and on the mainland as far north as Boeotia. This suggests that in the late Bronze Age West-Greek speech was confined to areas north-west of the Mycenaean heartlands and that much of the dialect diversity of the classical period, along with the distribution of the dialects at that time, had resulted from migration and population mixture during the dark age. Thus scholars are largely agreed that what is most distinctive about the Attic-Ionic and Aeolic dialect groups is mainly of post-Mycenaean origin, the former evolving from an East-Greek (i. e. Mycenaean-like) base, perhaps in Attica and adjacent islands, the latter from a West-Greek one, perhaps in Thessaly, with both groups then undergoing further development in the wake of subsequent migrations across the Aegean to Asia Minor. The distinction between northern (North West Greek) and southern (Peloponnesian Doric) branches of West Greek is similarly seen as post-Mycenaean, with the principal differences emerging as migrants moved south after the Mycenaean collapse. On this view, Arcadian is taken to be a close descendant of Mycenaean, evolving among speakers whose forebears had taken local refuge from the influx of West-Greek speakers, while the closely related Cypriot is assumed to have developed among descendants of those compelled by the newcomers to leave their homeland. 5 5.

See S. Colvin, A Historical Greek Reader, Oxford 2007 for an up-to-date survey of the dialects and their characteristics.

74

2. Ancient Greek

3. The Phases of the Greek Language

2.3 The Earliest Literature In the absence of a standard, the earliest Greek literature is composed in a range of “literary” dialects that loosely reflect the speech of the regions where its different genres first developed, or where they received their definitive form. For example, Homer’s epic poems, the Iliad and the Odyssey, represent the culmination of an oral tradition of heroic poetry that climaxed in Ionia towards the end of the dark age (early-middle eighth century BCE). The dialect of the texts transmitted to us by the scholars of Alexandria is therefore fundamentally an early variety of Ionic. The tradition, however, almost certainly originated in the Mycenaean age or before, and there are some features of the language that may reasonably be attributed to the Greek of that time. There is also evidence for a parallel Aeolic tradition from which material was periodically borrowed. This was Ionicized as far as possible, i. e. if the results were compatible with the epic meter, but otherwise words and phrases were left in their Aeolic form. The language of the Homeric poems is therefore an artificial amalgam of Ionic with a range of archaisms and Aeolicisms that unambiguously distanced it from the vernacular of any region. 6 Though the stylized quality of the earliest literary dialect was to have profound and lasting consequences, this Kunstsprache had evolved more for practical than aesthetic reasons. The composition, memorization, and performance of large-scale “literary” works in pre-literate societies are greatly facilitated if the linguistic material is partly prefabricated and has a regular rhythmical structure. Oral traditions tend therefore to be verse traditions, with formulaic phraseology organized into a series of subsystems, each dealing with a recurrent theme and comprising a set of near-synonyms of differing metrical value within the structure of a “line” of verse. We may be quite sure that the epic archaisms and Aeolicisms were tenaciously preserved by the Ionian bards because they were metrically different from Ionic forms and provided invaluable flexibility within the precious formula systems that constituted the foundation of their craft. The prestige of the Homeric poems grew exponentially in the age of alphabetic literacy, and versions of the “text” were soon written down everywhere. As Panhellenic treasures, they came to be widely recited at public festivals throughout the Greekspeaking world and eventually to serve as one of the central components of traditional Greek education. Unsurprisingly, therefore, all other genres of early Greek literature show the influence of the epic in a corresponding distancing of their language from everyday usage through the incorporation of archaism and unusual vocabulary, the elaboration of a characteristic phraseology, and the avoidance of dialectal parochialism: each genre may have had its characteristic dialect, but this was never simply a local vernacular. The developed language of choral lyric, for example, deploys a set of epicisms and Aeolicisms in combination with a non-localizable Doric base that reflects the Dorian role in the early development of the genre, most notably in archaic Sparta. The Attic dialect of fifth-century Athenian tragedy, familiar from the dramas of Aeschylus, Sophocles, and Euripides, similarly employs a characteristic set of archaiz6.

See G. C. Horrocks, “Homer’s dialect” in: I. Morris / B. Powell (eds.), A New Companion to Homer, Leiden 1997, 193-217. 2. Ancient Greek

75

3. The Phases of the Greek Language

ing and foreign vocabulary items and grammatical features, while the choral lyrics continue to display a conventionalized Doric gloss. Even prose writing was affected, once prose began to be used for literary purposes in sixth-century BCE Ionia as a vehicle for the groundbreaking work in science, philosophy, and historiography that ushered in the classical era. The literary Ionic of the historian Herodotus (circa 484 – circa 425 BCE) is characterised inter alia by many apparently Homeric forms and phraseological reminiscences. Though some are due to later editors who “restored” what they took to be correct Ionic on a priori grounds, the impulse to develop a contemporary Ionic for belletristic purposes must have been strong in a culture where poetic Kunstsprachen were already established, and in Ionia the Homeric tradition would surely have seemed a particularly attractive source of inspiration. Accordingly, just as in poetry, so in prose, the prestige of early work, with its associated repertoire of stylistic and phraseological resources, led later practitioners, regardless of their native speech, to follow their predecessors and use the established literary dialect according to genre. A stylized and rather “poetic” Ionic thus became the standard language for prose-writing, just as a stylized Doric was the language of choral lyric, and a stylized Attic the dialect of tragic discourse. The genre-conditioning of language may seem strange in the modern context, but for Greek writers of the period it was a straightforward and logical response to the absence of any other forms of standardization. 7

2.4 The Emergence of a Standard During the fifth century BCE Athens, through its efforts to repulse Persian aggression, had become a major naval power in the eastern Mediterranean, becoming for a time the leader of a league of allied states—in reality an empire—obliged to pay tribute to its “protector.” The acquisition of wealth and power is often associated with a desire for cultural prestige, and Athens was no exception. Through the recycling of its vast imperial income the city soon emerged as the most important cultural center of the Greek world, and began to promote the development of a specifically Athenian literature, comprising tragic and comic drama, historiography, oratory, and philosophy. Initially, much of this development took place under the tutelage of Ionian masters, who were only too glad to sell their expertise to the Athenian elite, and the Attic of the earliest Athenian compositions shows significant Ionic influence in vocabulary, grammar, and style, thereby continuing the tradition of literary stylization. The impact of Ionic is already apparent in early work such as the history of the war between Athens and Sparta written by Thucydides (circa 460 – circa 395 BCE), and its pervasiveness amply confirmed by the fact that elements of Ionicized Attic eventually start to appear even in the city’s official inscriptions. By the fourth century BCE the absolute cultural dominance of Athens in fields such as rhetoric (e. g. Demosthenes and Isocrates) and philosophy (e. g. Plato and Aristotle) was such that literary Attic had become the norm for serious prose writing throughout the Greek world. The progress of Attic towards standard status at the level of belles-lettres was sup7.

For a recent survey of the literary dialects and up-to-date bibliography see A. C. Cassio (ed.), Storia delle Lingue Letterarie Greche, Florence 2008.

76

2. Ancient Greek

3. The Phases of the Greek Language

ported by more practical developments. The Athenian empire consisted largely of Ionic-speaking states, and routine contact between Attic and Ionic led quite quickly to the disappearance of the latter as a truly distinct variety. But that same contact also promoted an internationalized administrative variety of Attic, known as “Great Attic,” to which Ionic had again contributed elements of morphology, syntax, and vocabulary. As the business language of the middle and upper classes, spoken and written, Great Attic, closely related to the literary Attic of high culture, soon expanded beyond the confines of the Athenian empire. Even after the defeat of Athens by Sparta at the end of the fifth century BCE its prestige was preserved, and though Greeks continued for a time to use their own dialects for internal affairs, Attic became the natural language of Greek diplomacy with external powers. But the crucial factor in the standardization of Attic was its adoption as both an official and cultural language by the rapidly expanding Macedonian kingdom during the fifth century BCE. Greeks to the south may have disputed the Greekness of the Macedonians, but the establishment of Attic at the Macedonian court was clearly intended as an unambiguous symbol of the Greek identity that the Macedonians themselves were determined to project. When Macedonia later came to control first the Greek cities and then, following the campaigns of Alexander the Great against the Persian empire, vast territories ranging from Egypt to the borders of India, the masterpieces of Athenian literature were established as the centerpiece of a common higher education system and Attic automatically became the language of administration in the newly acquired empire. After Alexander’s death the empire was divided into a number of hereditary monarchies, in Greece and Macedonia, Asia Minor, Syria, and Egypt. The Koine, or “common dialect” of these “Hellenistic” kingdoms, was simply the developed form of Attic used by the Greco-Macedonian elite for the administration of its territories. For foreigners this became the only form of Greek worth learning, and even in the “old” Greek world an ability to use it alongside a regional dialect became essential for anyone with ambition. Though grammatically and lexically standardized in its higher registers, widespread acquisition of the Koine as a second language or dialect led to interference phenomena and regional differentiation of its spoken varieties, as often reflected in sub-elite writing. We should therefore think of the Koine not simply as the written and spoken language of the Greco-Macedonian elite, but as a superordinate form of Greek standing at the pinnacle of a dialectal pyramid of lower-register varieties that now evolved under its influence and ultimately owed their identity to it. In the newly founded cities of the Hellenistic world the different dialects of colonists from old Greece were rapidly assimilated to locally evolving varieties of the Koine, to which indigenous bilingual populations also contributed significantly. In old Greece the local dialects might still be used for day-to-day communication and local affairs, but the Koine similarly gained ground at their expense as more and more Greeks became diglossic. Over time, the ancient dialects, subject to ever greater Koine penetration, lost many of their distinctive characteristics and themselves became versions of the Koine, albeit with residues of the old speech. The traditional spoken dialects of modern Greek, now much eroded by universal education and the mass media, descend ultimately from such local forms of the Koine, though the actual lines of decent are complex and often obscure because of extensive population movements in 2. Ancient Greek

77

3. The Phases of the Greek Language

Byzantine and Ottoman times. Tsakonian, a residual variety of the south-eastern Peloponnese, is a partial exception to this picture in that it appears to have undergone less thorough Koineization, perhaps because of extreme geographical isolation, and to exhibit features that may be direct developments from an ancient Doric dialect of the region. Evidence for the many different registers and varieties of the Koine in the Hellenistic period comes from imperial inscriptions and correspondence, the unpretentious translation-prose of the Septuagint, official and private inscriptions from different regions, and Egyptian papyri with content ranging from official documentation to private letters and fragments of popular literature. 8 The Hellenistic period was a time of innovation in science and philosophy, and higher forms of the Koine, adapted as necessary, were felt to provide an appropriate medium for such writing. But those fortunate enough to have the opportunity to read classical texts became increasingly conscious of the differences between the classical Attic of the literary canon and the contemporary Koine, which even in its highest registers had begun to reflect changes in the spoken language. In response, the prose of belletristic genres such as history began to detach itself stylistically and grammatically from that of technical writing and to adopt a more conservative form. Towards the end of the first century BCE this trend culminated in the emergence of the Atticist movement, which sought to restore classical “purity” through the careful emulation of the best ancient authors. Though initially not much more than a reaction to a florid rhetorical style known as Asianism (after the area where it first flourished), Atticism eventually evolved into an overarching linguistic ideology that became entrenched in the higher education system and had profound long-term consequences for the future of the Greek language.

2.5 Greek in the Roman Empire The inexorable rise of Atticism is best understood within the context of the encroachment of Rome into the Hellenistic world from the middle of the second century BCE onwards. In some quarters the Romans might present themselves as the bringers of civilization to barbarians but the Hellenistic world was home to the high culture they were currently appropriating for themselves. The Romans therefore constructed a more subtle justification of empire in the east, distinguishing an admirable Greek past, in which the customs and values they prized were first developed, from a decadent Greek present that needed the firm hand of Rome if those customs and values were to be successfully revived. At the same time, alienation from a political system over which they no longer had control led the Greeks to indulge in their own nostalgia for past cultural glories. But it eventually became clear that Rome’s commitment to the Greek past as the fountainhead of its own civilization could be successfully exploited by contemporary Greeks as a source of wealth and influence in the modern world, and

8.

For surveys of the key issues relating to the origins and development of the Koine see C. Brixhe (ed.), La koiné grecque antique, vol. I, Une langue introuvable?, Nancy 1993, and C. Brixhe (ed.), La koiné grecque antique, vol. II, La concurrence, Nancy 1996.

78

2. Ancient Greek

3. The Phases of the Greek Language

the second century CE ushered in a flamboyant reassertion of Hellenism (the Second Sophistic) that soon began to shape the Roman experience in decisive ways. The objective of the resurgent Greek intelligentsia was to restore its cultural leadership through reconnection with the ancient classics and the re-establishment of their “pure” Attic dialect. The idea of language change as a form of decay took firm root, and the Greek aristocracy began to pride itself on the acquisition of a linguistic hallmark that distinguished it decisively from the common herd. Literary Attic, already six hundred years old, quickly came to be seen as the “perfect” and “ideal” form of Greek that had been allowed to decay through ignorance and neglect but which could now be restored to its proper place as the ultimate symbol of the Greek revival. Thereafter an Attic or Atticizing style, often imperfectly controlled, became the aim of all writers of artistic prose, and even those who rejected extreme versions of the doctrine as pedantry felt they must still adopt a style that reflected aspects of classical practice. One immediate consequence was that the higher registers of the written Koine also became more classical in style, and more remote from everyday speech, as any pressure for change from below (the spoken vernacular) came to be inhibited by equal or greater resistance from above (classical Attic). Henceforth evidence for the historical development of Greek has to be sought primarily in lower-level documents where Atticizing influence, whether through ignorance or by design, was minimized. Of particular interest in this connection is the Greek of the New Testament, which reflects a range of low-to-middle Koine styles employed by men who in the main lacked a literary education. After the death of Constantine the Great in 337 CE and the adoption of Christianity as the official religion of the Empire, we might have expected a shift in favor of this ordinary Koine as the language of holy scripture, but instead the progressive conversion of the aristocracy led, during the fourth and fifth centuries CE, to doctrinal debates that were instinctively conducted within the overall intellectual framework of traditional Greek culture, including its commitment to Atticized diction. Though a tradition of lower-level composition persisted for the edification of the masses, the literary Greek so firmly rejected by the early Church Fathers as a pagan irrelevance thus extended its domain into the heart of the Christian establishment. Thus even though the Hellenistic world had been fully incorporated into the Roman political sphere by the end of the first century BCE, the educational and cultural significance of Greek actually expanded under Rome in the context of the Second Sophistic and the Christianization of the empire. Eventually, in line with the steady integration of the Greek-speaking aristocracy with its Roman counterpart, Greek came to be regarded as the “other” Roman language, a perception reinforced at a practical level by the fact that the Romans were for the most part content to employ the established language of administration in their eastern provinces. Latin was therefore never a threat to the linguistic status quo in traditionally Greek-speaking areas, despite its having a formal, and in part symbolic, role in the key Roman institutions of the legal profession and the army. Extensive bilingualism and the realities of Roman rule did, however, mean that everyday Greek absorbed many Latin loanwords and that the two languages began to show signs of grammatical convergence in their sub-literary registers. The dominant position of Greek in the east was naturally strengthened after the division of the empire into western and eastern halves at the end of the fourth century 2. Ancient Greek

79

3. The Phases of the Greek Language

CE, and with the final collapse of the Roman state in the west in the fifth century CE, Greek in the east quickly superseded Latin in even its residual functions. 9

3. Medieval Greek After the short-lived recovery of North Africa and Italy in the sixth century CE, the history of the East-Roman or Byzantine state is largely one of geographical contraction. The crushing of Persia in the seventh century CE was almost immediately followed by the Arab invasions that deprived Constantinople of the Middle East and Egypt, while a contemporary Slavic invasion led to the temporary loss of most of the Balkan Peninsula. Between 850 and 1050, however, the Byzantines staged a remarkable recovery during which Orthodox Christianity and other facets of Byzantine culture were carried to Moravia, Bulgaria and Russia. As part of this process the Greek alphabet was adapted by Byzantine missionaries to form the Cyrillic script used specifically for the writing of Slavic languages. But matters again changed for the worse towards the end of the eleventh century when the Normans invaded southern Italy and Sicily and the Seljuk Turks occupied eastern and central Asia Minor after their victory at Manzikert, near Lake Van in eastern Anatolia (1071). Still worse was to follow. The revival of the west and the growing power of the papacy led to the arrival of crusaders on their way to the Holy Land, closely followed by Italian traders greedy for profit. Illfeeling came to a head in 1204 when Constantinople was sacked by the fourth crusade and the bulk of its empire divided among the victors, most notably the Venetians. Although the capital and part of the Peloponnese were eventually restored to Byzantine rule, the empire never recovered militarily or economically. By the fifteenth century Constantinople was virtually all that remained, and the Ottoman Turks, having already taken control of much of eastern Europe as far as Vienna, stormed the city in May 1453. The early Byzantine period saw inter alia a vast output of Christian literature, in particular church histories, saints’ lives and collections of edifying tales exemplifying the spiritual rewards of a pious life. The historical works were aimed at the elite and employed the Atticizing style of traditional historiography (e. g. the great church history of Eusebius of Caesarea, 263-339 CE), but the second type of writing provided reading material for a wider audience and employed a middle-to-low variety of the contemporary Koine, analogous to that of the gospels (a nice example is provided by the sixth-century Leimon of Ioannes Moschus). After the Arab conquests literary output declined until the recovery of the ninth century, but an important survivor from the earlier period was the universal chronicle, a genre that unified the different components of Byzantine culture through the assimilation of biblical and Graeco-Roman “history” into a single chronological framework from the creation to the present day. The most famous example is the early ninth-century chronicle of Theophanes (which

9.

On Atticism and the Second Sophistic see T. Whitmarsh, The Second Sophistic, Oxford 2005 and L. Kim, “The Literary Heritage as Language: Atticism and the Second Sophistic” in: E. J. Bakker (ed.), A Companion to the Ancient Greek Language, Oxford 2010, 468-482.

80

3. Medieval Greek

3. The Phases of the Greek Language

continued an earlier work by George Syncellus). Compiled as reference works for a general audience from a wide range of non-literary sources, chronicles traditionally employed a contemporary middle-register Koine (as in the well-known sixth-century chronicle of Ioannes Malalas). But with the return of stability and confidence we also witness a series of classical revivals in which an interest in ancient literary forms and a fresh commitment to Atticism reasserted themselves. Significant figures here include Photius, the ninth-century polymath and patriarch of Constantinople, and Michael Psellus, the eleventh-century philosopher, historian, and statesman. Originally a multi-lingual conglomerate in which Greek was not even the sole language of culture, let alone the only spoken language, the residual empire of the period after the Seljuk invasions consisted largely of territories where the majority population were in fact Greek native speakers. This newly diminished status seems to have prompted a desire to rethink the place of Byzantium in the world, with a particular emphasis now on the Greek elements of its inheritance. One predictable manifestation of this reappraisal in the twelth-century was the emergence of a fresh generation of Atticists (e. g. Anna Comnene, daughter of the emperor Alexius I Comnenus and author of a history of her father’s times), who, rather in the manner of the writers of the Second Sophistic, sought a revalidation of the Empire through reconnection with the ancient Greek past. More interestingly, however, the same period also saw, for the first time since late antiquity, a willingness to experiment with forms of contemporary vernacular Greek in literature aimed at an educated audience. One version of the “epic” poem of Digenes Acrites, for example, represents a compilation of popular, in part probably orally transmitted, tales of derring-do from the old eastern frontier, while the anonymous comic poems of “Poor Prodromus” (twelfth-century comic begging poetry addressed to the emperor, and perhaps written in part by the poet and courtier Theodorus Prodromus) combine urban vernacular with a range of higher registers in an amusing parody of the language problems then facing Greek speakers in the capital. Such experimentation came to an abrupt end with the political disintegration at the end of the twelfth century, and a literary language with a strong vernacular base reappears only with the verse romances of the fourteenth century, when, under the impact of western models in a Greek-speaking world now largely controlled by western powers, this form of medieval Kunstsprache was adopted as the standard medium for verse fiction. Outside the field of literature both the imperial and church bureaucracies continued to employ a version of the Koine that was slowly modernized and adapted with the passage of time, mostly by combining a more contemporary syntax with the traditional (ancient-looking) surface morphology and by allowing the use of innovative vocabulary as required. The intelligentsia employed a similar form of language for dealing with academic concerns of a more technical kind, and this updated written Koine continued and evolved through the Ottoman period to emerge as the standard written style of the Greek-speaking elite in the age of Enlightenment, and so ultimately as the basis for a written standard when an independent Greece emerged from the declining Ottoman Empire in the early nineteenth century. At the same time assorted forms of low-level documentation, everyday reference works and examples of church outreach (e. g. wills, dowries, land transactions, monastery archives, letters, home cures, horoscopes, travel guides, homilies, sermons, etc.) 3. Medieval Greek

81

3. The Phases of the Greek Language

start to appear, in ever greater quantity from around the fourteenth century onwards. These texts are normally written in forms of Greek that are comparatively free from the learned conventions of the literary and administrative/academic traditions and so partly reflect the evolving vernacular, sometimes even showing evidence of regional differentiation, especially in areas where centralized Byzantine rule had been replaced and the local form of the language had achieved some measure of autonomy and prestige. But we should not forget that the ability to write necessarily entailed exposure to the learned tradition, however small, and that no one at this time was deliberately trying to base their written practice directly on contemporary speech. 10

4. Modern Greek 4.1 Greek in the Ottoman Period Within the territories of the Ottoman Empire creative writing in Greek soon came to an end. The only written forms of the language that survived and developed were of a more practical kind, most importantly the administrative Koine as employed by the sole remaining Greek-speaking institution, the Orthodox Church, and the closely related academic Koine of the intellectuals who continued to write and to communicate with one another both inside and outside the empire. During the seventeenth and eighteenth centuries, however, the ideas of the European Enlightenment began to infiltrate Ottoman lands via the Italian universities, where many Greeks secured a higher education, through the Greek commercial “colonies” in Egypt, Russia and many western countries, and from the Greek-ruled principalities of Wallachia and Moldavia, which lay open to Vienna and the west via the river Danube. At this time a more “demoticizing,” i. e. syntactically simpler and less archaizing, version of the traditional academic style came to be used specifically for religious and technical works aimed at meeting the needs of a fast-growing readership. Typical text types include not only more traditional bible anthologies and scriptural commentaries but also, for the first time, reference works on modern subjects such as geography, history, politics, and science. This simplified written style seems to have reflected the contemporary formal speech of the educated elite quite closely,

10. For general surveys of work in the period see A. P. Kazhdan, A History of Byzantine Literature (650-850), Athens 1999; M. Alexiou, After Antiquity: Greek Language, Myth, and Metaphor, Ithaca, NY 2002; and M. D. Lauxtermann, Byzantine poetry from Psides to Geometres, Vienna 2003. On a range of more specific topics, the following works are helpful: R. Beaton, The Medieval Greek Romance (second edition), Cambridge 1996; R. Beaton / D. Ricks (eds.), Digenes Akrites: New Approaches to Byzantine Heroic Poetry, Aldershot 1993; H. Eideneier, Ptochoprodromos: Einführung, kritische Ausgabe, deutsche Übersetzung, Glossar, Köln 1991; E. Jeffreys, Digenis Akritis: The Escorial and Grottaferrata Versions, Cambridge 1998; E. Jeffreys / B. Croke / R. Scott (eds.), Studies in John Malalas, Sydney 1990; and C. Mango / R. Scott (translators), The Chronicle of Theophanes Confessor: Byzantine and Near Eastern History AD 284-813, Oxford 1997.

82

4. Modern Greek

3. The Phases of the Greek Language

while still retaining something of the traditional archaizing gloss, especially in its morphology. But it was still the case that any text revealing the spoken norms of the uneducated did so only through ignorance of accepted written styles rather than by design. In the creative sphere only oral folk poetry employed a language close to popular vernaculars, but versions of these songs were not recorded in writing until long after Greek independence. Unsurprisingly, the spoken language at all levels began to absorb Turkish loans, and these quickly start to appear also in texts alluding to Turkish institutions or religious and administrative practices. The popular spoken varieties of geographically isolated districts such as Cappadocia in central Anatolia even began to adopt phonological and grammatical phenomena from Turkish, such as vowel harmony and agglutination of suffixes. In Greek-speaking areas outside the Ottoman Empire the demanding linguistic conventions of the centralized Byzantine state had already been weakened with the advent of “Latin” rule after the Fourth Crusade. In these the local forms of Greek not only underwent Romance (particularly Italian and French) influence, but often acquired significant prestige of their own, evolving over time as official and sometimes literary media alongside the languages of the conquerors. In Venetian Crete in particular, a tradition of dialect-based Greek literature of exceptional quality began to develop, with the remarkable sixteenth- and seventeenth-century poetry of the Cretan renaissance occupying pride of place. This too came to an abrupt end when the island succumbed to the Turks, in 1669. The existence and success of such highly developed local varieties, not only on Crete but also on Cyprus and to a lesser extent in the Ionian islands, provided some support for the idea of a future national language based not on the archaizing written tradition but on some form of contemporary speech. But the Greek-speaking elite in Constantinople and elsewhere, many of whom would soon join the new ruling class of independent Greece, instinctively regarded such varieties as too marginal and too parochial to constitute a credible basis for a written standard. 11

4.2 The Quest for a Modern Standard By the beginning of the nineteenth century all educated Greeks were agreed that a standard written language was essential for the development of mass education and as a vehicle of culture and administration in any future Greek state. There was, however, no agreement as to the form such a language should take. Those within the Ottoman Empire often took a more conservative view, those in the Greek diaspora a more radical one, but there were no hard and fast lines of demarcation. Interestingly, the dispute itself was instinctively conducted by all parties in the common archaizing style

11. Useful surveys of literary genres and language issues include R. Beaton, Introduction to Modern Greek Literature, Oxford 1994; R. Beaton, From Byzantium to Modern Greece: Medieval Texts and their Modern Reception, Aldershot 2008; R. Beaton, Folk Poetry of Modern Greece, Cambridge 1980; and D. W. Holton (ed.), Literature and Society in Renaissance Crete, Cambridge 1991. 4. Modern Greek

83

3. The Phases of the Greek Language

that had evolved amongst the upper classes from the traditional written languages of administration and intellectual discourse. Some advocated this “contemporary educated” variety as the basis for the future standard language, other groups wished to build on the more consistently archaizing diction of Orthodox officialdom. A minority of traditionalists, in thrall to the kind of thinking that had originally taken root in the period of the Second Sophistic, promoted the idea of the “purification” of the learned written language in the direction of ancient Attic. A few extremists, encouraged by the current wave of western enthusiasm for ancient Greece, even started planning for the eventual restoration of classical Attic itself. Finally, and in contrast to all of those who advocated some form of traditional written language, a more radical group, influenced by the situation in western Europe where national languages based on prestigious varieties of contemporary speech had long been the norm, put forward the possibility of a written standard built on the living speech of the Greek people, i. e. on the varieties known collectively as demotic. The Greek war of independence broke out in the Peloponnese in 1821, and in the absence of any realistic opportunity to resolve the issue of a national language before its conclusion, the establishment of the Greek state, in 1832, saw the de facto institutionalization of the contemporary written style of the educated minority, later known as katharevousa (“purifying” language). Almost immediately this already archaizing style of writing was subjected to a process of progressive “restoration” (i. e. re-antiquing on the model of ancient Attic) that gathered pace throughout the remainder of the nineteenth century. The motivation behind this officially sanctioned drive to revive the glories of the linguistic past is best understood in the context not only of the traditions and values that had persisted in many learned circles up to the time of independence but in the light of subsequent developments. The first century of Greek independence was dominated by economic weakness, rivalry between royalists and republicans (groups eventually correlating quite closely with the political right and the political left), and by a deep-seated irredentism driven by the fact that most of the great centers of Greek population (e. g. Thessaloniki, Constantinople, and Smyrna) lay beyond the original borders of the kingdom. In the period before the First World War Greece expanded step by step to something close to its modern frontiers, but disaster struck in 1922 when, in an attempt to exploit the collapse of the Ottoman Empire by launching a campaign deep into Asia Minor, the Greek army was decisively defeated. The result was an enforced exchange of populations, the elimination of the historic Greek presence in Asia Minor, and an influx into Greece of one and a half million refugees. Continued political instability led to military dictatorship, and, after the horrors of German occupation in the Second World War, to civil war between the communists, who had taken much of the initiative in coordinating the resistance, and government forces backed first by Britain and then by the United States. Political troubles continued after the communist defeat and culminated in the ultra-right-wing military dictatorship of 1967-1974. The legacy of anger and bitterness only began to fade with the abolition of the monarchy and Greek membership of the European Economic Community (now the European Union), in 1981. For a quarter of a century this new settlement contributed to the growth of prosperity and to the institutionalization of democratic government, but the world-wide financial crisis beginning in 2007-2008 brought 84

4. Modern Greek

3. The Phases of the Greek Language

the country to the brink of bankruptcy and reopened political divisions as government cuts began to bite. Though mutual suspicion between Greece and Turkey remain, fuelled by on-going territorial disputes in the Aegean and the still festering problem of Cyprus, the early years of the twenty-first century have brought some measure of rapprochement. Against this background, the unresolved language question at first lacked any obvious political dimension. The largely middle-class advocates of demotic objected to katharevousa not because it entrenched class divisions (many new city dwellers in fact regarded it as a hard-won badge of social progress, however badly they actually deployed it) but because of its demonstrable artificiality and its supposed pretentiousness and vulgarity. Many were equally dismayed by its persistently negative impact on the prospects for the eventual adoption of a more “natural” form of language that might express a modern identity for Greece as a progressive, liberal, European nation. Conservative critics in turn pointed to the absence of recognized literary classics in the vernacular and to the considerable regional variability of demotic in its totality, both of which they saw as insuperable obstacles to the credibility of the demoticist cause. But this criticism ignored, in part willfully, the literary precedents from the Middle Ages and early-modern Crete and Cyprus, as well as the thriving tradition of folk song and the body of educational texts in “simple” style from the immediately preceding period. Furthermore, the influx of people from all levels of society and many different parts of the Greek-speaking world into the towns and cities of the new Greece was already leading not only to the crystallization of a spoken norm for the upper and middle classes but to dialect leveling and the emergence of novel urban vernaculars with a more homogeneous character than the opponents of demotic were willing to acknowledge. The fact that the Greek state lay south of the Arta-Volos line meant that it was the regional speech of the southern-mainland that provided the dominant input to this process. This lacked both the radical sound changes characteristic of the north and the archaisms of Cretan or Cypriot, and so was not very different, mutatis mutandis, from the traditional language of vernacular literature or the prestigious speech of Constantinople/Istanbul. The spontaneous reshaping and refocusing of spoken Greek within newly independent Greece, supported by an education system in which the teaching of katharevousa was actively contributing to the leveling process, made the case for demotic far less problematical than its opponents allowed. By the eighteen-eighties, most creative writers had rejected the perceived sterility of katharevousa and were using semi-standardized forms of demotic that drew heavily on the literary forerunners mentioned above. This development encouraged the leaders of the demoticist movement to work for the abolition of katharevousa altogether, and demotic was eventually introduced into primary schools in 1913. But it made little headway thereafter because of the intense politicization of the language question that took place in the period after the Russian revolution. With the growth of the communist party demotic was increasingly associated by the establishment, quite unfairly, with the causes of the revolutionary left, and then ruthlessly stigmatized and caricatured as the language of traitors and weirdoes who would strip the country of its linguistic and cultural heritage. In fact the Greek communist party, like every other political party, had used katharevousa in its early years as the only form of written Greek available for serious debate, but the right-wing propaganda took root and the pro4. Modern Greek

85

3. The Phases of the Greek Language

blems intensified during the nineteen-forties, when the key role played by the communists in the resistance to the German occupation led not to a post-war communist state, as in the rest of eastern Europe, but to exile, internal repression, and demonization in the wake of a bloody civil war. Though demotic was eventually put on an equal footing with katharevousa as a language of education in 1964, the military dictatorship of 1967-1974 immediately reversed even this limited reform, and its eventual adoption as the official language of government and administration, in 1976, had to wait for the demise of this odious regime. Thereafter katharevousa was largely confined to ceremonial status in the law, the army, and the church. Today it has all but disappeared, thus bringing two millennia of diglossia to an end and releasing Greeks from the burden of their linguistic past—albeit at the cost of rendering much earlier literature inaccessible to younger generations. Nonetheless, the relationship between modern Greek and the ancient language remains a hot educational and political potato, and many creative writers feel free to experiment with registers that reflect the complex history of their language. Modern literary figures of international repute include the short story writer Alexandros Papadiamandis (1851-1911, who combined katharevousa narrative with demotic dialogue), the novelist Nikos Kazandzakis (1883-1957), and the poets Konstandinos Kavafis (1863-1933), Yorgos Seferis (1900-1971) and Odiseas Elitis (1911-1996), the last two being Nobel laureates. The term “demotic” is rarely used today, and “common” or “standard” modern Greek has all but replaced it to describe the contemporary form of the spoken and written language used by the majority of educated speakers in the major urban centers. This terminological shift acknowledges that the politically driven polarization of the language debate had for a long time deflected attention away from the fact that the institutionalization of katharevousa had had a marked impact on the development of traditional forms of spoken demotic. If this had not been the case, the eventual shift from official katharevousa to official demotic would have been a great deal more problematical than it proved to be. Even those who promoted demotic at the end of the nineteenth century appreciated that it lacked the uniformity and range required of a true standard, and at first efforts were made to systematize its grammar and enrich its vocabulary with coinages based on ancient roots recast in demotic form. But urban Greeks, brought up with katharevousa, instinctively rejected what seemed to them to be yet another form of “artificiality,” and instead preferred to enrich their speech with direct borrowings from ancient Greek and the learned written tradition. The spoken language of the cities thus developed organically as a mixture of demotic and katharevousa, and a related written variety eventually came to be widely used in all but the most right wing newspapers. Mid-twentieth-century grammars of demotic based on the usage of folk songs and early vernacular literature already represented an anachronistic and romanticized ideal that had been long superseded by developments on the ground. As the spoken standard developed its own internal coherence and came to be used, in written form, in fields once dominated by katharevousa, it naturally attracted further characteristics of the old written language, which then passed in turn into educated spoken discourse. The official “transition” from katharevousa to demotic in 1976 was therefore just the final stage, albeit a critical one, in a process that had long been under way. Though conscientious demoticist writers continue to conform to the old rules, their language is in reality an artificial literary dialect, and both this and the old 86

4. Modern Greek

3. The Phases of the Greek Language

rural dialects are best seen today as minority varieties subsumed under the umbrella of standard modern Greek. 12

4.3 Greek in the Twenty-first Century Greek is the sole official language of the Republic of Greece, with nearly 11 million speakers out of a total population of approximately 11.3 million, and one of the two official languages, alongside Turkish, of the Republic of Cyprus, where nearly 80 % of just over 1 million inhabitants are Greek speakers. There are also large Greek-speaking communities in North America, Australia, Germany, and the United Kingdom as a result of relatively recent emigration, chiefly for economic reasons (perhaps some 3 million speakers in all). At various periods in the past, however, Greek was spoken quite widely elsewhere in the Balkans, in southern Italy and Sicily, in Turkey, in the Middle East and North Africa, and in Russia, Ukraine, and the Caucasus. Relatively small numbers of speakers still survive in some of these regions, most notably in southern Albania, the heel of Italy (where the language is known as Grico), and around the coasts of the Black Sea. But the last ninety years have seen a steady geographical concentration of Greek speakers as a result of population exchanges (between Greece, Bulgaria, and Serbia in 1913-1919; between Greece and Turkey in 1923) and through voluntary or involuntary repatriation (from the Soviet Union in 1919 and subsequently; from Egypt under Nasser in 1952; after anti-Greek riots in 1955, from Istanbul, where the Greek minority had been allowed to stay under the terms of the 1923 exchange). Within Greece this, along with the enforced deportation and murder of the large Ladino-speaking Jewish population of Thessaloniki during the Second World War, has produced something close to linguistic homogeneity, though there are still speakers of south Slavic Macedonian (in Macedonia), Vlach (in Macedonia and Epirus), and Albanian (in Epirus, Attica, Boeotia, and the Peloponnese), as well as large communities of Turkish speakers (approximately 100,000) in Thrace, who were again permitted to stay under the terms of the 1923 treaty. The last twenty years or so have also seen a large influx of economic migrants and refugees into Greece, both from neighboring countries and from further afield, some of these arriving legally, others not. Currently, standard modern Greek is used throughout the education system, and in government, journalism, and broadcasting, though some traces of katharevousa persist, as noted, in the legal profession, the army, and the Orthodox Church. The Byzantine liturgy also remains in use alongside the original text of the New Testament. The surviving regional dialects within Greece can be divided into Peloponnesian-Heptanesian (the old base for the modern standard), Northern, Old Athenian (which survives residually in Megara and central Euboea), Cretan-Cycladic, and South-eastern subgroups (the last of which includes Cypriot as a rather aberrant member); the special case of Tsakonian in the Peloponnese has already been discussed. But universal education and daily exposure to the mass media mean that all these varieties are being stea12. See P. A. Mackridge, “Katharevousa (c.1800-1974): an Obituary for an Official Language” in: M. Sarafis / M. Eve (eds.), Background to Contemporary Greece, London 1990 and P. A. Mackridge, Language and National Identity in Greece, 1766-1976, Oxford 2009. 4. Modern Greek

87

3. The Phases of the Greek Language

dily assimilated to the standard, which is also used in Cyprus alongside the local dialect. Beyond Greece, the dialects of Cappadocia in central Anatolia all but disappeared with the exchange of populations, but Pontic dialects from the southern coast of the Black Sea are still spoken in private contexts by the descendants of refugees (though these are once again subject to steady assimilation), while small groups of Muslim speakers of Pontic (known locally as Romeyka, “Roman”) remain in villages around the Turkish city of Trabzon/Trebizond on the southern coast of the Black Sea. The peripheral south Italian and Asia Minor dialects, until quite recently long isolated from the Greek mainstream, are quite different from the more central dialects, and these, along with Cypriot, might reasonably be classified as separate languages, despite the protests of those who insist on the essential unity and continuity of Greek language and culture from antiquity to the present day. Despite the widespread use of Greek as a lingua franca in Hellenistic, Roman, Byzantine, and even Ottoman times (as a diplomatic language and later as the language of administration in the Danubian principalities, now Romania and Moldova), the concentration of its speakers in the twentieth century has produced a situation in which it is no longer routinely learned as a second language in any non-Greek community. It is, however, one of the official languages of the European Union. 13

13. See P. A. Mackridge, The Modern Greek Language, Oxford 1985 and D. W. Holton / P. A. Mackridge / I. Philippaki-Warburton, Greek: A Comprehensive Grammar of the Modern Language, London 1997.

88

4. Modern Greek

4. Phonology and Morphology of Septuagint Greek Gerard Mussies

1. Introduction Any description of the Greek language of the Septuagint is necessarily of a diachronical character due to the different dates of the Septuagint books, the chronological distance between the autographs and the oldest extant manuscripts, and that between the various manuscripts themselves. If the “Septuagint” is taken to be the equivalent of the so-called “Alexandrian Canon,” it comprises some 50 different writings, translations of Hebrew and Aramaic works and original Greek compositions. These were not all translated or composed at the same time, but rather over a period of about four centuries, from the third century BCE to the second century CE. Moreover, the oldest more or less complete manuscripts of the Septuagint are not autographs, but rather copies dating from the fourth century CE, which, depending on the date of translation or composition of the different books, are at least six to two centuries younger than the originals. Here we assume, by way of a working hypothesis, that the oldest part of the Septuagint, the translation of the Pentateuch, was accomplished in Alexandria between 285/4 and 283/2 BCE, that is during the co-regency of Ptolemy I and Ptolemy II. This time range is wide enough to accommodate sources that date the translation to the reign of the former (so Irenaeus, Adversus Haereses 3, 21, 2), others that date it to the tenure of the latter (Aristeas, To Philocrates, passim; Philo, Life of Moses 2, 31), and Clement of Alexandria who knew of both datings (Stromata 1, 22, 148). 1 On the other hand, 4Maccabees, which was probably the last addition to the Septuagint canon, might date according to some scholars from as late as the reign of the emperor Trajan. 2 Consequently, fourth-century CE manuscripts of Genesis and 4Maccabees reflect, respectively, the third-century BCE and the second-century CE Greek of the originals, but may also contain some elements of the language of the copyists who worked some two to six centuries later, an interval during which the language did not remain wholly 1.

2.

It would also lend some historical credibility to Aristeas’ portrayal of Demetrius of Phaleron as fully present and active at the Alexandrian court, which was only possible under his protector Ptolemy I, as Demetrius was banished to the countryside when the old king died and his son became sole ruler (Diogenes Laertius, Vitae philosoph. 5, 78-79). See A. Dupont-Sommer, Le Quatrième livre des Machabées, Introduction, traduction et notes (Bibliothèque de l’École des Hautes Études 274), Paris: 1939, 81, note 45 (117/118 CE) and U. Breitenstein, Beobachtungen zu Sprache, Stil und Gedankengut des Vierten Makkabäerbuchs, Basel 1976, 179. Other authors, however, prefer a date comprised between 19/20 and 54 CE. See E. J. Bickermann, “The Date of IV Maccabees” in: A. Marx (ed.), Louis Ginzberg Jubilee Volume, New York, NY 1945, 105-112; M. Hadas, The Third and Fourth Books of Maccabees, New York, NY 1953, 95-96; H. Anderson, “Fourth Maccabees” in: D. N. Freedman et al. (eds.), The Anchor Bible Dictionary, vol. IV, New York, NY 1992, 453a. 1. Introduction

89

4. Phonology and Morphology of Septuagint Greek

unchanged. What these two manuscripts may not have in common are linguistic features attested in the third century BCE but no longer in use during the second century CE (possibly the concomitant use of οὐθείς and οὐδείς in Genesis and the absence of οὐθείς in 4Maccabees). Of course, many other language elements are common to all the Septuagint books. Robert Helbing, Henry St. John Thackeray and Félix-Marie Abel were well aware of these various chronological differences when they wrote their “grammars” of Septuagint Greek in the early twentieth century. 3 These works were inventories of the “sounds” and “forms” found in the oldest group of sources then known, the fourthand fifth-century CE uncial parchment manuscripts (the Alexandrinus, the Sinaiticus, and the Vaticanus). Regarding sounds, these authors established which letters alternated with one another in identical words or parts of words as compared with the standard Classical Attic orthography and which letters did not (yet) do so. As to forms, they described those that differed from Classical Attic, e. g. the third person plural imperative endings (-έτωσαν, etc.). These “grammars” also established comparisons with the language of the Egyptian Greek papyri and the inscriptions from Attica and Pergamum—to the extent that these had been published and described by Edwin Mayser, Konrad Meisterhans and Eduard Schwyzer, Wilhelm Lademann and Eduard Schweizer 4—in order to identify those forms that could not yet have figured in the language of the autographs, and thereby produce a picture of what this language might have looked like. What probably came closest in Antiquity to a grammar of Biblical Greek is Hadrian of Antioch’s fifth-century CE Introduction to Holy Scripture, which describes idiom, style, and imagery, using examples from the Septuagint and the New Testament. This work, however, does not discuss orthography or morphology. 5 The question now has to be asked whether the source material upon which the Septuagint “grammars” were based has changed, and if so how this affects the descriptions that were given. As to the differences of date between the supposedly earliest and latest Septuagint books, it appears that practically speaking the cleft has neither narrowed nor widened dramatically. Then as now, the Pentateuch is generally assumed to have been the first part of the Hebrew Bible to have been translated, in the early third century BCE, while the Septuagint as a whole was taken to have been completed before the close of the first century CE or during the early part of the second century CE 6. No new data comparable to the postscript of Esther have since come to light that would 3.

4.

5. 6.

R. Helbing, Grammatik der Septuaginta: Laut- und Wortlehre, Göttingen 1907 (reprint Göttingen 1979); H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, vol. I, Introduction, Orthography and Accidence, Cambridge 1909 (reprint Hildesheim/New York, NY 1978); F.-M. Abel, Grammaire du grec biblique suivie d’un choix de papyrus, Paris 1927. E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften, vol. I, Laut- und Wortlehre, Leipzig 1906–1938; K. Meisterhans / E. Schwyzer, Grammatik der attischen Inschriften (third edition), Berlin 1900; W. Lademann, De titulis Atticis quaestiones orthographicae et grammaticae, Kirchhain 1915; E. Schweizer, Grammatik der pergamenischen Inschriften. Beiträge zur Lautund Flexionslehre der gemeingriechischen Sprache, Berlin 1898. J.-P. Migne (ed.), Patrologiae cursus completus (Series Graeca 98), Paris 1860, 1271-1312. See note 2 and cf. Thackeray, Grammar, vol. I, 6, 61; R. Helbing, Die Kasussyntax der Verba

90

1. Introduction

4. Phonology and Morphology of Septuagint Greek

allow us to date any of the translations or original compositions more precisely than had been possible in the early twentieth century. As to the chronological differences between the oldest extant manuscripts and the autographs, the situation has changed somewhat, since many Septuagint papyri have been found and published after the beginning of the last century. Especially interesting are those papyri that predate the fourth- and fifth-century uncials upon which the above-mentioned grammars were based. None of the manuscripts inventoried by Alfred Rahlfs in the original 1914 edition of his catalog were older than the fourth century CE, 7 whereas the 2004 edition lists some eighty-five pre-fourth century CE manuscripts and papyri. 8 However, with a few exceptions, such as “W” and the Bodmer and Chester Beatty Papyri, the overwhelming majority of these are very short and lacunary fragments dating mainly from the third century CE. 9 In order to take into account the newly published material, the grammars of the Greek papyri and the Attic inscriptions have been re-edited. 10 All this implies that the additional pre-fourth-century CE Septuagint papyri and the updated grammars furnish somewhat earlier or more precisely dated examples of well-known facts.

2. Phonology The interchanges referred to above can be subdivided into variations that affected the phonemic system and those that did not. Variations of hαιi 11 for hεi or of hεi for hαιi —both occurring in αναιβεννεν (ἀνέβαινεν) “went up” (Gen 2:6 A—V CE)—indicate, as their Modern Greek pronunciation and its history show, that the diphthong [aı] was progressively abandoned and replaced by [ε], which ultimately resulted in a change of the phonemic inventory and system. Variations, however, of hεi for hαi before liquid consonants, as in σιελος for σίαλος “grease, spittle” (Isa 40:15 B—IV CE) and μιερος for μιαρός “dirty” (2 and 4Mac A—V CE), did not result in the loss of either /ε/ or /a/ and hence caused no change in the phonemic system, but merely in the distribution of the two vowels. The same holds good of the variation of hρi for hλi or of hλi for hρi, as in φαραγγας/-ος for φάλαγγας/-ος “battle-array” (1Mac 6:35,45 A

7. 8. 9. 10.

11.

bei den Septuaginta: ein Beitrag zur Hebraismenfrage und zur Syntax der Κοινή (second edition), Göttingen 1979, xiii. A. Rahlfs, Verzeichnis der griechischen Handschriften des Alten Testaments (Mitteilungen des Septuaginta-Unternehmens der KGW zu Göttingen 2), Berlin 1914. A. Rahlfs / D. Fraenkel (eds.), Verzeichnis der griechischen Handschriften des Alten Testaments, vol. I,1, Die Überlieferung bis zum VIII. Jahrhundert, Göttingen 2004. A list of the most important earlier ones is given by M. K. H. Peters, “Septuagint” in: Freedman, Anchor Bible Dictionary, vol. V, 1094. E. Mayser / H. Schmoll, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften, vol. I, Einleitung und Lautlehre (second edition), Berlin 1970; F. T. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, Milan, vol. I, Phonology, 1976, vol. II, Morphology, 1981; L. Threatte, The grammar of Attic inscriptions, vol. I, Phonology, Berlin 1980. Graphemes will be placed in angle brackets, e. g. hφi; phonemes in slashes: /ph/; phonetic notations (approximate) in square brackets: [ph]. 2. Phonology

91

4. Phonology and Morphology of Septuagint Greek

—V CE) and φλουραν for φρουράν “watch, guard” (1Mac 11:66 A—V CE). As neither /l/ nor /r/ disappeared, and indeed are still present in Modern Greek, the result was only a slight change in their distribution.

2.1 Vowels In the Hellenistic period the vowel system of common Greek was subject to a twofold gradual development: A. The diphthongs in h-ιi were simplified into monophthongs. B. The difference in length between long and short vowels was leveled. A.1. The short diphthongs [aı, εı, ᴐı] developed into [ε, i, y]. hαι~εi: εστανε (= -ναι) (Gen 41:17), υπολειπεσθε (= -σθαι) (Gen 45:7 962—III CE), μεμιανσε (= -σαι) with supralinear correction (Num 5:20 963—II CE), πεδες (= παῖδες) (Gen 14:15 814—II CE); hει~ιi εισχυρα, επει, εστειν, τεις, οτει, ηρειθμησε (Gen 14:5, 7, 7, 13, 14, 14 814—II CE); hοι~υi: οιπομενοντες (Ps 24:3 2110—II-IV CE), κοινων (Ps 67:24 ibid.), υκιαν (Ps 83:4 2149—IV CE), [ε]πυεις (Sus-Th 52 861—III CE). 12 A.2. The long diphthongs, written hαι, ηι, ωιi, lost their second element and became long vowels, written hα, η, ωi. This simplification has been attested in the Attic inscriptions since 204/3 BCE, 13 in the Pergamene inscriptions since the second century BCE, 14 and in the Ptolemaic papyri since the third century BCE. 15 Prior to the fourth century CE it can be observed passim in e. g. P. Bodmer XXIV (Ps 17-118; = 2110—II-IV CE). The conservative orthography, characterized by so-called iota adscriptum, can however also be found there occasionally, e. g. εν τω ουρανωι (Ps 35:6), τω δε αμαρτωλωι (Ps 49:16), εν ανομιαι συν- (Ps 50:7, BS plural, Masoretic Text singular), εν τοπωι (Ps 67:6), εν νεφελαι (Ps 77:14), θ(ε)ω αλλοτριωι (Ps 80:10), 16 εν αληθειαι (Ps 110:8), [εν τ]ω ουρανωι (Ps 112:6). Still earlier examples of iota adscriptum are provided by the fragments of Deuteronomy in 848 (50 BCE), and of Leviticus and Numbers from Qumran Cave 4 (= 801—II-I BCE, 802—I BCE, 803—I BCE-I CE), e. g. [τ]ηι γηι and καιρωι (Lev 26:4), αυτηι (Lev 4:28), [ο]μοσηι (Lev 5:22), αποδωι (Lev 5:23), επ αυτηι (Num 4:7). Hypercorrect instances such as εξετεθηι and αντεστηι (Esth 8:17; 9:2 996—I-II CE) show, however, that the iota adscript was purely graphical. If the original Pentateuch translation was destined for the royal library, it may well have retained the conservative spelling with the iota adscript. B. The difference in length between short and long vowels was leveled, resulting in a phonetic state sometimes termed “vocalic isochrony”; nevertheless the accented vowels may still have been realized in a way somewhat longer than the unaccented, as in Modern Greek, e. g. δέκα [’ðε:ka]. These developments were gradual and, as far as can be judged from the deviations from the traditional Attic orthography, they varied in first occurrence according to time, region, and social level. This reduction of quantities could not be seen in the letters hα, ι, and υi, which each represented both a short and a 12. 13. 14. 15. 16.

Lacunae restored by the editor(s) are in square brackets. Threatte, Grammar, 358-360. Schweizer, Grammatik, 65, 86-88. Mayser / Schmoll, Grammatik, vol. I, 111-113. Original abbreviations supplied by the editor(s) are in round brackets.

92

2. Phonology

4. Phonology and Morphology of Septuagint Greek

long vowel, but was especially visible in the case of the confusion of hοi and hωi, e. g. in 2110 (II-IV CE): τω (= τὸ) (Ps 25:6), μωνω (Ps 50:6, corrected to ο), στωματος (Ps 53:4), το (= τῷ) (Ps 48:7), εδοκε (Ps 66:7), αιον(ι)α (Ps 76:6), βροματα (Ps 77:18), προ[τοτο]κον (Ps 77:51). The analogous interchange of hεi and hηi occurred also in the Septuagint manuscripts: [εκζ]ετουντες, βασιληων, γενεθητω, ηλπι[δα] (Ps 21:27; 44:10; 68:23; 72:28 2110—II-IV CE); εψημα (Gen 25:29,34 961—IV CE, but 25:30 εψεματος). Along with this expected interchange there was, however, perhaps depending on local diversity or circumstances beyond our knowledge, confusion of hηi and hι/ειi (e. g. [πασ]ει and πασι, both = πάσῃ in Dan 1:17 861—III CE), which presupposes the more closed pronunciation of hηi as [i] that eventually prevailed, resulting in the Modern Greek pronunciation of êta as ita. This latter development can hardly be demonstrated from Septuagint manuscripts before the fourth century CE. Let us now consider the relative chronology of the phenomena described in A.2. and B. Since the short and the long diphthongs were subject to separate evolutions, the long diphthongs only progressively developing via long vowels into short vowels, the loss of [–ı] in the long diphthongs must have preceded the leveling of the vowel quantities. Nevertheless, the processes of short diphthong monophthongization and vowel isochrony seem to have taken place concomitantly. Here we should keep in mind that correct conservative spellings may well have represented the actual pronunciation of some scribes. Indeed, school tablets dating from the third century CE attest to the fact that the traditional, old fashioned phonetic values of the letters were still taught in schools, where hεi and hοi were described as “short”, hηi and hωi as “long”, and hα, ι, υi as “ambivalent.” 17

2.2 Consonants The important phonetic changes that here took place were: C. The gradual loss of /h-/, which is visible in such a frequent combination as ἐφ’ ἔτη. 18 This has often been explained on the analogy of ἡμέρα, but ἐφ’ ἡμέραν/ς was not frequent. Other instances of the incorrect application of ἐπ’/ἐφ’, οὐκ/οὐχ, etc., show that /h-/ had disappeared from pronunciation: [ε]φ εθνος (Mic 4:3 943—I BCE-I CE), καθ ει[κονα] (Gen 5:1 911—III CE), κατιστων (Dan 4:37 967—II-III CE), ουχ [ιδια] (Gen 15:13), ουκ ω[ς] (Gen 31:15), ουκ υπηκουεν (Gen 39:10 961—IV CE), ουχ ιδου (Ps 53:2 2110—II-IV CE), ουχ Ιουδα (Sus-Th 56 861—III CE). D. The spirantization of hφi/[ph] > [f], etc., of hβi/[b] > [v], etc., and of the diphthongs hαυi and hευi/ [aw, εw] > [av, εv], pronounced [af, εf] before [-t-] in e. g. αὐτός, might be visible in such orthographical interchanges as hαβ~αυi, hεβ~ευi, which, however, are hardly attested in Septuagint manuscripts, although elsewhere “Livia” is often spelled Λ(ε)ιβια. The only frequent proof of this phonetic change is the interchange of γιγν- [gign-], later [jijn-], and γιν- or γειν- [gin-] or [jin-], as in γίγνομαι (latest Attic example circa 250 BCE 19) > γ(ε)ίνομαι and ὀλίγος > ὀλίος. In 17. F. G. Kenyon, “Two Greek School-Tablets” JHS 29 (1909), 36. 18. Mayser / Schmoll, Grammatik, vol. I, 174 (earliest instance 261 BCE). 19. Threatte, Grammar, 562. 2. Phonology

93

4. Phonology and Morphology of Septuagint Greek

both cases h-γ-i/ [j] was dropped after the preceding i-vowel, where [j] would have been hardly audible as a separate sound. It must again be remarked that conservative school teaching may have somewhat impeded spontaneous phonetic development, since in the third century CE β, γ, and δ were still described as μέσα “middle”; π, κ, and τ as ψιλά “bald”; φ, χ, and θ as δασέα “rough, hoarse,” just as Dionysius of Halicarnassus (De comp. verb. 14) had done in the first century CE. The description of consonants as “rough” or “hoarse” is not very helpful as it does not permit to distinguish between [kh] and [x], [th] and [θ], and [ph] and [f], although Dionysius does say that the “rough” had “also the addition of breath.” It is telling, however, that θάλασσα, a loanword from Greek in Coptic, was spelled ᴛƨᴀᴧᴀᴄᴄᴀ and analyzed into the feminine article ᴛ- “the” followed by the indeclinable ᴤᴀᴧᴀᴄᴄᴀ, hence with the plural article ⲛ- also: ⲛᴤᴀᴧᴀᴄᴄᴀ “the seas”. As to the pronunciation of hφi, it is likewise revealing that when Egyptian, now called “Coptic”, came to be written in Greek letters, it was necessary to add a special Demotic letter hϥi to note [f], because hφi still represented [ph], sometimes also written hᴨᴤi in Coptic. E. On the other hand the frequent confusion in papyri from Egypt of hγ~κ~χi, of hδ~τ~θi, and less often of hβ~π~φi, forms no part of the general phonetic evolution from Koine Greek to Byzantine Greek, but rather is due to the change in the Egyptian dialects of the opposition voiced/unvoiced into the opposition non-aspirated/aspirated voiceless in the dental and velar plosives (less so in the labials). 20 Pre-fourth century CE examples include: γ~κ: Αικυπτου (Ps 79:9; 80:11), κλωσσαν (Ps 33:14), [η]γου[σα] (Ps 30:14), γαρδια (Ps 39:13; 40:7 2110—II-IV CE); κ~χ: δρακμην (Gen 24:22 961—IV CE), [βρα]κιονα (Gen 24:18 962—III CE); δ~τ: καθετρα (Ps 106:32), ονειτος (Ps 30:12; 78:4), αμαρδια(ς) (Ps 24:11,18); cf., however, [α]μαρτιας (Ps 24:18), αυδη (Ps 23:1, = αὐτῇ; 48:14, = αὕτη); θ~δ: Αιδιοψιν (Ps 73:14), αληδινος (Ps 85:15, corrected); θ~τ: παραθεινον (Ps 35:11), εμφανητι (Ps 79:2 2110—II-IV CE); β~π: βροβα[τον] (Isa 53:7 958—IV CE). F. Finally, the substitution of Attic h-ττ-i by the h-σσ-i of the other dialects in e. g. τέτταρες > τέσσαρες was not due to phonetic evolution but to the weight of the majority of the dialects. Nevertheless, -ττ- is still found in ἐλάττων, ἥττων, κρείττων, and derivatives, in other words also in 2, 3, and 4Maccabees.

3. Morphology During the Hellenistic and Roman periods, Common Greek, as compared with Classical Attic, underwent morphological changes that affected declension and conjugation in different ways, many of which are reflected in the Septuagint.

20. J. A. L. M. Vergote, Phonétique historique de l’égyptien: les consonnes, Louvain 1945, 24-28, 118-119; J. Osing, “Lautsystem” in: W. Helck / E. Otto / W. Westendorf (eds.), Lexikon der Ägyptologie, vol. 3, Wiesbaden 1980, 947.

94

3. Morphology

4. Phonology and Morphology of Septuagint Greek

3.1 Declension G. The dual category completely disappeared, both in declension and conjugation; separate dual pronouns such as νῶι, σφῶι, and σφέ, as well as other words that implied a duality such as οὐδέτερος and πότερος, fell into disuse. Δύω became phonetically ambiguous and was generally spelled δύο; the case forms δυοῖν or δυεῖν are attested only three times. Ἄμφω was replaced by ἀμφότεροι; ἀλλήλω by ἀλλήλους/-ας/-α. A related phenomenon concerns the use of the comparative, which was no longer restricted to the comparison of two items, as is most clearly reflected in instances where Benjamin is referred to as the νεώτερος of his twelve or eleven brothers (Gen 42:13; 43:29 962—III CE; Gen 44:23 A—V CE; cf. Ps 67:28 2110—II-IV CE). H. The gradual disappearance of the reflexive pronouns αὑτόν and σαυτόν, which give way to ἑαυτόν and σεαυτόν. No doubt αὑτόν fell into disuse due to the disappearance of its initial /h-/, which made it indistinguishable from αὐτόν; the trisyllabic σεαυτόν was then preferred, on the analogy of the trisyllabic ἐμαυτόν and ἑαυτόν. Hence σεαυτη (Ez 16:17 967—II-III CE), σεαυτω (Deut 4:9 963—II CE; Esth 4:17g 968 —II-III CE), cf. however σαυτην (Ez 16:52) and σαυτω (Ez 2:3 967—II-III CE). Ἑαυτόν also tended to be used for all persons: second person singular εαυτω (Gen 6:14, but after lacuna; 911—III CE), first person plural εαυ[τοις] (Deut 3:7 963—II CE), second person plural [υ]μειν εαυτοις (Deut 4:23), cf. however [υμ]ειν αυτοις (Deut 4:16 963 —II CE). I. The extension of the accusative singular ending -ν to accusatives singular ending in a vowel, such as ἄνδρα > ἄνδραν, ἀληθῆ > ἀληθῆν, βασιλέα > βασιλέαν: [χει] ραν (Gen 8:9), [σα]ρκαν (Gen 9:15 911—III CE), χειραν (Ps 80:15; 103:28, both corrected), μονογενην (Ps 21:21 2110—II-IV CE), ιερεαν (Hag 2:2, corrected W—III CE); the opposite happened when πᾶν was seen as an accusative singular masculine: παν βουνον (Ez 34:6 967—II-III CE; παντα A—V CE). The use of the accusative plural -ας in υ-stems ἰχθύας instead of the ambiguous ἰχθῦς, which could be a nominative singular, a nominative plural, or an accusative plural, hence ιχθυας (Gen 9:2 961— III CE; Nu 11:5 B—IV CE; Ps 104:29 2110—II-IV CE). J. The incidental replacement of the accusative plural -ας by -ες, which may have started with the accusative τέσσαρες, on the analogy of the nominative-accusative τρεῖς; the incidental replacement of βασιλέας with βασιλεῖς, as in the nominativeaccusative plural πόλεις: [κ]ρειους αναβαινοντες (Gen 31:12), (τας) θυγατερες (Gen 5:4; 6:2), τους βοε[ς] (Gen 34:28 911—III CE), τας χειρες (Isa 56:2 965—III CE). K. The disappearance of some seldom-used declension (sub)types such as λεώς, νοῦς, κρέας, ἠχώ. These words did not disappear, but moved to other, more frequent, types of declension. Λεώς > λαός passim, νεώς > ναός passim (there are only six occurrences of νεώς in 2Mac; on the other hand, ἵλεως was still used); ὀστοῦν > ὀστέoν: oστεοι[ς] (Ps 108:18 2110/II-IV CE); λαγώς > λαγωός: τοις λαγωοις (Ps 103:18 2110— II-IV CE); γῆρας > τὸ γῆρος: γηρους (Gen 44:20 962—III CE; Sir 46:9 964—IV CE), cf. however γηρως (Ps 70:9,18 2110—II-IV CE); γηρ(ε)ι (Gen 21:7 961—IV CE; Dan 6:1 968—II-III CE; Ps 91:15 2110—II-IV CE); κέρας, πέρας, τέρας > κερατ-, περατ-, τερατ- (2110—II-IV CE): κερατα (Ps 68:32), περατα (Ps 47:11; 64:9; 66:8; 94:4; 97:3), περατων (Ps 45:10; 58:14; 60:3; 64:6), τερατα (Ps 45:9), τερατων (Ps 104:27), cf. how-

3. Morphology

95

4. Phonology and Morphology of Septuagint Greek

ever κρέας accusative plural κραια (Ps 49:13); ἡ ἠχώ > ὁ ἦχος: dative singular ηχω μεγαλω (Sir 46:17 964—IV CE).

3.2 Conjugation L. The spread of the third person plural ending -σαν (as in ἦσαν) to the imperfect of contract verbs, so as to remove the homophony of the first person singular and the third person plural –ν: εδολιουσ[αν] (Ps 13:3 2019—III CE), επεινοσαν (Dan 5:3 967— II-III CE), ευλογουσαν (Dan 5:4 967—II-III CE), παρετηρουσαν (Sus-Th 12 861—III CE); the second aorists: ειδοσ[αν] (Hab 3:10 943—I BCE-I CE), εφ[αγ]οσαν (Gen 3:7 911—III CE), ημαρτοσα[ν] (Deut 32:5 963—II CE), ηλθοσαν (Ez 14:1 967—II-III CE); the optatives -οισαν, -σαισαν, -(θ)ειησαν- for -οιεν, -σειαν, and -(θ)ειεν-: εκλιποισαν (Ps 103:35 2110—II-IV CE), ποιησαισαν (Deut 1:44 963—II CE), αποστραφειησαν, καταισχυνθειησαν (Ps 69:3 2110—II-IV CE), with the concomitant introduction of -σαις, -σαι, -σαισαν for -σειας, -σειε(ν), -σειαν (which occur only in Job A—V CE—and 4Mac): οικτειρησαι and ευλογησαι (Ps 66:2 2110—II-IV CE); the new imperatives -έτωσαν/-έσθωσαν; -σάτωσαν/-σάσθωσαν, which generally replaced the older -όντων, -έσθων, -σάτων, –σάσθων, such as in the Ptolemaic papyri and, probably, the Pentateuch autograph. M. The reintroduction of the second person singular middle ending h-ῃi for laterAttic h-ειi: σωθηση (Esth 4:13), βιαζη (Esth 7:8 968—II-III CE and B—IV CE). Still, some verbs remained impervious to this change, such as βουλει (Esth 3:11 B—IV CE). N. The replacement of the thematic endings of the second aorist by those of the sigmatic aorist -ον > -α, -ες > -ας, etc., always without h-σ->, the most likely initial instance being the ambiguous ἔπεσον > ἔπεσα. Such is the case of εἰπεῖν. Indicative: ειπα (Gen 20:11; 24:39,42 905—II-III CE), (Ps 29:7; 30:15; 31:5; 39:8; but 39:11: ειπον), (Ps 40:5; 74:5; 76:11; 81:6; 94:10 2110—II-IV CE), (Jon 2:5 943—I BCE-I CE); ειπας (Ps 88:3; 89:3 2110—II-IV CE); (εἶπε(ν) needed no change); ειπαμεν (Gen 26:28 A—V CE), (Gen 42:31; 44:22 962—III CE); ειπατε (Gen 44:28 962—III CE); ειπαν (Ps 63:6; 70:10; 77:19; 82:5,13; 93:7 2110—II-IV CE), (Ps 82:5 2077—II CE), (Sus-Th 13,20,27,28,50 861—III CE). Imperative: ειπον (= εἰπόν, modeled on λοῦσον, but with the accent of εἰπέ) (Gen 20:13 961—IV CE), (Gen 45:17 A—V CE), (Esth 4:10 869—III-IV CE; B—IV CE), ειπα[τω] (Ps 117:2), ειπατε (Ps 65:3; 95:10), ειπ[ατ]ωσαν (Ps 117:4 2110—II-IV CE). Optative: ειπαι[σ]αν (Ps 34:25 2110—II-IV CE). Participle: ειπας (Gen 46:2 A— V CE; B—IV CE; S—IV-V CE), ειπαντες (Isa 3:10 A—V CE). Infinitive: ειπαι (Ez 33:8,13,14 B—IV CE). Other examples include: υπηνεγκα (Ps 68:8), εξειλατο (Ps 114:8), επεσαν (Ps 26:2 2110—II-IV CE), ηλθαν (Sus-Th 28), ιδαν (Sus-Th 18 861—III CE), imperative ενεγκατε (Ps 95:7bis,8 2110—II-IV CE); participle πεταμενου (Ps 90:5 2110—II-IV CE), infinitive ενενκαι (Sus-Th 18 861—III CE). O. The incidental replacement of the third person plural perfect ending -(κ)ασιν with -(κ)αν and, sometimes, the concomitant fusion of the aorist indicative and perfect indicative: πεϕυτευκαν (Ez 19:13 967—II-III CE); incidental shift of ἥκω to ἧκα: ηκασιν (Gen 46:32 962—III CE; Isa 60:4 965—III CE; Sus-Th 52 861—III CE). P. The gradual replacement of the present and imperfect of the athematic “-μιverbs” with thematic types and endings: ἀνοίγνυμι > ἀνοίγω, συνίημι > συνίω, ἵστημι > ἱστάνω: ανοιγ[ει] (Isa 53:7 958—IV CE), ανοιγων (Ps 37:14), απετιννυον 96

3. Morphology

4. Phonology and Morphology of Septuagint Greek

(Ps 68:5), επαν[ιστανομενων] (Ps 17:49, apparently required by space, 2110—II-IV CE), καθιει (Deut 25:2 957—II BCE), συν(ε)ιων (Ps 13:2 2019—III CE), (Ps 52:3), ωμνυον (Ps 101:9), ομνυων (Ps 62:12 2110—II-IV CE). The verb εἰμί also had some new forms: third person imperative ητω = ἔστω (Ps 103:31 2110—II-IV CE); first person singular imperfect ημην (Dan 4:1 968—II-III CE; Tob 12:18 990—III CE); second person singular ἦσθα, sometimes ης (e. g. Judg 11:35 B —IV CE); first person plural ἦμεν, sometimes ημεθα (e. g. 1Reigns 25:16 B—IV CE; A —V CE). Q. The differentiation of presents/imperfects and aorists that had become homophonous by sound change: εκλιπανον (Zech 11:16 W—III CE), καταλιμπανι (Gen 39:16 962—III CE), cf. however ἔλειψα in απολειψασα (Judg 9:9; 11:13 B—IV CE). The use of ἔσθω for ἐσθίω: εσθων (Ps 40:10), εσθοντος (Ps 105:20), εσθοντες (Ps 52:5 2110 —II-IV CE). R. The gradual replacement of the Attic future in -ῶ, -ᾶς/-εῖς, etc. with the sigmatic future in -σω, -σεις, etc.: ανδρισεσθε (Ps 30:25 2110—II-IV CE); α[πολε]σω (Gen 18:28,30), but [α]πολω and [απολ]εις (Gen 18:28,29,31 961—IV CE); διαμερισω (Ps 59:8 2110—II-IV CE); ε[ξορ]κισω (Gen 24:3 961—IV CE); κα[θ]αρισει (Num 30:13), but also [καθαρ]ιει (Num 30:6 963—II CE); σκεπασεις (Ps 30:21); συνβιβασομαι (Ps 31:8), but also [εγγ]ιουσιν (Ps 31:6) and επιστεριω (Ps 31:8 2110—II-IV CE). S. The usage of -κα in the plurals of such aorists as ἔδομεν: ἔδωκαν (Ps 76:18), συνηκαν (Ps 27:5; 63:10; 81:5; 105:7), and προσεθηκαν (Ps 68:27 2110—II-IV CE). T. The temporal augment is almost generally absent in verbs with initial εὐ- and in compounds with αὐτ(ο)-.

4. Conclusion The above survey is only a partial updating and cannot claim completeness. The older grammars that we have already mentioned should always be consulted for further details.

4. Conclusion

97

5. The Vocabulary of the Septuagint and Documentary Evidence John A. L. Lee

1. Introduction The Septuagint translators translated their Hebrew Bible into Greek. That Greek was the Greek of their time, the third to first centuries BCE. Our knowledge of it depends on evidence in surviving Greek texts. Much of that evidence comes from sources that have long been known and have formed the basis of the material in our lexicons, namely, literary texts of both the translators’ time and the preceding period of ancient Greek (Classical Greek). These texts comprise a range of genres, styles, dialects, and levels of Greek, from the Epic of Homer to the historical prose of Polybius. Everyday language is not well represented, though much can be gleaned from writers like Plato and Aristophanes. The literary sources may provide a fairly accurate picture of the usage of a word in the Greek known to the Septuagint translators, and then again may not. The discovery in modern times of documentary texts brought forth a different and illuminating source of evidence. Documentary texts are those inscribed or written on stone or papyrus, surviving from the day they were set down. The majority are postClassical, though there is a large body of inscriptions that are earlier. They are mainly not literary but official or everyday, using the kind of Greek appropriate to their purpose. They thus give access to another dimension of the language, and fill out much of the incomplete picture that we worked with before. The absorption of this new evidence into our reference books is still in progress today and will be ongoing as more documentary texts are found and published. They have made, and will continue to make, a major contribution to our knowledge of Septuagint Greek.

2. History of Study Investigation of the documentary texts in relation to the Septuagint began in earnest with Adolf Deissmann, shortly after the first major finds of papyri. But there were forerunners. A full list of those who earlier made the connection would be tedious, but the names of Friedrich Sturz, 1 Heinrich Thiersch, 2 and Heinrich Anz 3 deserve mention. All had little documentary material to work with, but realized its importance

1. 2. 3.

F. G. Sturz, De dialecto macedonica et alexandrina, Lipsiae 1808. H. G. J. Thiersch, De pentateuchi versione alexandrina libri tres, Erlangae 1841, 87-90. H. Anz, Subsidia ad cognoscendum graecorum sermonem vulgarem e pentateuchi versione alexandrina repetita, Halle 1894.

98

1. Introduction

5. The Vocabulary of the Septuagint and Documentary Evidence

for the Septuagint. John Mahaffy, the first editor of the Petrie papyri, contemporary with the Septuagint Pentateuch, 4 was struck by the similarities. 5 Henry Swete also, following up Mahaffy’s insight, pointed out a number of Septuagint words found in PPetrie and the just-published BGU, and quoted in full a PPetrie text. 6 These were early straws in the wind. Deissmann’s researches were not systematic but a collection of the most telling data he could find to demolish the notion of a “Biblical Greek,” which had long held sway. 7 He pointed out documentary parallels to the language of both the New Testament and the Septuagint, but the New Testament was the ultimate target and grows more prominent in the later studies. 8 Deissmann drew on what was currently available, both inscriptions and papyri, and though they were few in comparison with subsequent discoveries, they were usually enough to establish his point. His work was a breakthrough at the time: it showed that the documentary texts were the evidence needed to bring the language of the New Testament and Septuagint out of the isolation that had been imposed on them. But there was a long way to go to exploit these sources fully, and some of Deissmann’s conclusions have not worn well, especially his view that the Greek found in the New Testament was the language of “simple, unlearned folk.” 9 Even his statement that the language of the Seventy was “the Egyptian Greek of the period of the Ptolemies” needs considerable nuancing. 10 James Moulton took up the challenge of gathering documentary parallels to the vocabulary of the New Testament, and with his (eventual) collaborator George Milligan produced the now-famous “MM.” 11 As Frederick Danker puts it, “Their work helped open up even more the curtains that Deissmann had drawn aside to expose an exciting new world for New Testament explorers.” 12 But MM had limited application to the Septuagint: its focus was the New Testament, even if Septuagint vocabulary was often mentioned along the way. And there were other limitations: it is not a lexicon, and the analyses of meaning are not well done; it deals with a selection of New Testament words, not the whole vocabulary; it draws mostly on papyri, not inscrip-

4. J. P. Mahaffy, The Flinders Petrie Papyri, Dublin 1891, 1893, 1905. 5. See H. B. Swete, An Introduction to the Old Testament in Greek, Cambridge 1900, 21 for references. 6. Swete, Introduction, 292-293. 7. A. Deissmann, Bibelstudien, Marburg 1895; A. Deissmann, Neue Bibelstudien, Marburg 1897 (English translation by A. Grieve, Bible Studies, second edition, Edinburgh 1909); A. Deissmann, Licht vom Osten (fourth edition), Tübingen 1923 (English translation by L. R. M. Strachan, Light from the Ancient East, New York, NY 1927). 8. On the Septuagint (with New Testament) see, in particular, Bible Studies, 86-169 (about 70 lexemes). With more focus on the New Testament, see Bible Studies, 194-267 (= Neue Bibelstudien) and, generally, Light (Septuagint 81-116 passim). 9. Light, 62; cf. 143-145 (“Book of the People,” etc.). 10. Bible Studies, 70. On the one hand, there is variety within the Greek of Ptolemaic Egypt, and it is not a distinct entity from Greek elsewhere; on the other, the Septuagint translators made adjustments to the language to suit their purpose, and chose differing styles and literary levels. 11. J. H. Moulton / G. Milligan, The Vocabulary of the Greek Testament: Illustrated from the Papyri and Other Non-Literary Sources, 5 vols., London 1914-1929 (one volume edition 1930). 12. F. W. Danker, Multipurpose Tools for Bible Study, Minneapolis, MN 1993, 117. 2. History of Study

99

5. The Vocabulary of the Septuagint and Documentary Evidence

tions; and its material is inevitably out of date and incomplete now. 13 The “new world” has expanded to undreamt-of horizons. Those who use MM today need to be aware that it must be supplemented by further investigation, and that its conclusions are not necessarily reliable. An “MM” for the Septuagint was on no one’s agenda when MM reached completion in 1929. Observations on selected parallels may be found in Joseph Ziegler’s 1934 study of the book of Isaiah, 14 and in a survey article by Orsolina Montevecchi in 1964. 15 That was the scene when the present author’s 1970 dissertation on the Septuagint Pentateuch was undertaken. 16 Its aim was to combat once more the idea of a special “Jewish Greek,” which had shown a resurgence, by gathering evidence for the links between the vocabulary of the Pentateuch and the contemporary language. Use was made especially of third-century BCE papyri from Egypt, which had grown in quantity but been barely exploited. Only a selection of words and uses was studied, namely, new developments, as being especially good indicators of the translators’ familiarity with the Greek of their time. Widening the field to other vocabulary and other books of the Septuagint remained to be done. The years following saw minor studies on selected words or parts of the Septuagint vocabulary by Raija Sollamo, 17 Hélène Cadell, 18 and the Italian papyrologists Orsolina Montevecchi, Francesco Vattioni, and Anna Passoni dell’Acqua. 19 More recently there has been renewed activity in short studies by this author 20 and Trevor Evans, 21 and 13. See C. J. Hemer, “Towards a New Moulton and Milligan” NovT 24 (1982), 97-123; G. H. R. Horsley, New Documents Illustrating Early Christianity, vol. 5, Marrickville 1989, 83-92; J. A. L. Lee, A History of New Testament Lexicography, New York, NY 2003, 123 with note 9. 14. J. Ziegler, Untersuchungen zur Septuaginta des Buches Isaias, Münster 1934, 175-212. Cf. assessment in M. N. van der Meer, “Papyrological Perspectives on the Septuagint of Isaiah” in: A. van der Kooij / M. N. van der Meer (eds.), The Old Greek of Isaiah: Issues and Perspectives, Leuven 2010, 107-133, in particular 109-110. 15. O. Montevecchi, “Continuità ed evoluzione della lingua greca nella Settanta e nei papiri” in: J. Wolski (ed.), Actes du Xe Congrès International de Papyrologues, Warsaw 1964, 39-49. 16. J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch, Chico, CA 1983. 17. R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint, Helsinki 1979, 308-351 (parallels in documents and literary texts). 18. H. Cadell, “Vocabulaire de l’irrigation: la Septante et les papyrus” in: B. Menu (ed.), Les problèmes institutionnels de l’eau en Égypte ancienne et dans l’antiquité méditerranéenne, Cairo 1994, 103-117; H. Cadell, “Vocabulaire de la législation ptolemaïque: problème du sens de dikaiôma dans le Pentateuque” in: G. Dorival / O. Munnich (eds.), Κατὰ τοὺς ο´: Selon les Septante. Trente études sur la Bible grecque des Septante en hommage à Marguerite Harl, Paris 1995, 207-221. 19. Detail in van der Meer, “Papyrological Perspectives,” 110; see also G. Dorival / M. Harl / O. Munnich, La Bible grecque des Septante: Du judaïsme hellénistique au christianisme ancien, Paris 1988, 243. 20. J. A. L. Lee, “A Lexical Study Thirty Years on, with Observations on ‘Order’ Words in the LXX Pentateuch” in: S. M. Paul / R. A. Kraft / L. H. Schiffman / W. W. Fields (eds.), Emanuel: Studies in Hebrew Bible Septuagint and Dead Sea Scrolls in Honor of Emanuel Tov, Leiden 2003, 513-524; J. A. L. Lee, “Ἐξαποστέλλω” in: J. Joosten / P. J. Tomson (eds.), Voces Biblicae: Septuagint Greek and its Significance for the New Testament, Leuven 2007, 99-113; J. A. L. Lee, “A Lexicographical Database for Greek: Can it be Far Off? The Case of amphodon” in: M. Karrer / W. Kraus / M. Meiser (eds.), Die Septuaginta: Texte, Kontexte, Lebenswelten, Tübingen

100

2. History of Study

5. The Vocabulary of the Septuagint and Documentary Evidence

significant results are being obtained in Michael van der Meer’s work on Isaiah, notable for its close engagement with the papyrus documents (and even pictorial remains). 22 All this has continued to show the value of the documentary evidence for elucidating the vocabulary of the Septuagint, but there has been no systematic exploration, and a means of gathering the results of these piecemeal studies into one place has not yet emerged. The two recent lexicons of the Septuagint, LEH and GELS, have to some extent performed this role, noting or using any material up to the date of compilation. But the rate of production has outstripped the ability of any printed work to keep an up-to-date record. These lexicons, moreover, designedly made no attempt at new researches into the documentary evidence, but took in what was on hand, rightly considering further work on that front a task for the future.

3. Nature and Value of the Evidence Documentary texts were engraved on stone, or written on papyrus, ostraka, wooden tablets, and other material. They survive from the time when they were written or inscribed, and are in varying states of preservation, from complete to very fragmentary, from perfectly legible to illegible. Decipherment, i. e., simply reading them, is a specialist task, because of differences in the writing of Greek as compared with today, as well as other factors. The range of dates is from late eighth century BCE for inscriptions and from late fourth century BCE for papyri, down to the end of the Koine period (600 CE) and beyond. The find locations also range greatly and the extremes are remarkable, but it is sufficient to say that Greek inscriptions have been found in quantity all over ancient Greece, Asia Minor, Egypt, and the rest of the eastern Mediterranean as well as Rome, while papyri have survived mainly in upper Egypt, with a small number in Syria, Palestine, and elsewhere; ostraka also come mainly from Egypt, though not all. The number of inscriptions is in the thousands; papyri number more than a million, of which less than a tenth have been published. 23 The nature of the documents is endlessly variable and any generalization is liable to be contradicted by a particular text, but a summary will be useful. The inscriptions tend to be of a formal and public character—not surprisingly, since they were set up for public viewing. They include: decrees of civic bodies, treaties, honorific decrees, 2008, 214-220. Note also J. A. L. Lee, “Aposkeuê in the Septuagint” JTS 23 (1972), 430-437; and M. O’Connor / J. A. L. Lee, “A Problem in Biblical Lexicography: The Case of Hebrew tap and Greek aposkeuê” ZAW 119 (2007), 403-409. 21. T. V. Evans, “The Court Function of the Interpreter in Genesis 42:23 and Early Greek Papyri” in: J. K. Aitken / J. Dines / S. Pearce / T. Rajak (eds.), Jewish Perspectives on Hellenistic Kingship, Berkeley, CA 2007, 238-252. 22. M. N. van der Meer, “Trendy Translations in the Septuagint of Isaiah: A Study of the Vocabulary of the Greek Isaiah 3:18-23 in the Light of Contemporary Sources” in: Karrer / Kraus / Meiser, Die Septuaginta, 581-596; M. N. van der Meer, “Bridge over Troubled Waters? The Γέφυρα in the Old Greek of Isaiah 37:25 and Contemporary Greek Sources” in: M. K. H. Peters (ed.), XIII Congress of the International Organization for Septuagint and Cognate Studies, Ljubljana 2007, Atlanta, GA 2008, 305-324; and van der Meer, “Papyrological Perspectives.” 23. R. S. Bagnall (ed.), The Oxford Handbook of Papyrology, Oxford 2009, 645 (P. van Minnen). 3. Nature and Value of the Evidence

101

5. The Vocabulary of the Septuagint and Documentary Evidence

dedications, epitaphs, manumissions, sacred laws, records of benefactions, accounts, catalogues, and even milestones. 24 Texts on papyrus may be said to vary even more. At the formal end there is, say, a letter of the emperor, at the other extreme a shopping list or a brief receipt on an ostrakon. In between there is an almost inexhaustible range of public and private documents: legal proceedings, contracts, accounts, petitions, tax receipts, reports, lists, and private letters, the last often the most interesting of all for their human appeal and language. 25 The Greek of the documents is almost as varied as their content. 26 We must first separate the texts of the Classical period (to 300 BCE), which are written in varieties of Classical Greek, including of course local dialects. The inscriptions of Athens are in Attic, those of Delphi in Delphian, and so on. Once Koine Greek, which grew primarily out of Attic, emerged, the dialects were gradually eroded, but not entirely: inscriptions in dialect continue to appear as late as the second century CE, usually with an admixture of Koine features. The poetic genres are also an important part of the picture, beginning in the earlier period and continuing right through the later. Epitaphs and the like will often display metrical forms and poetic vocabulary and morphology, though they may also contain Koine features. The researcher needs to be able to discern these differing genres and the features of language that accompany them. The great majority of the texts of the later period, both inscriptions and papyri, are in Koine Greek, but not of one single variety. It is a mistake to regard them all uniformly as witnesses to vernacular or “popular” Greek. While they may and do witness to that, their language offers a range from everyday to literary, and what features appear will depend on the choice of register and the degree of education of the writer. Even an “uneducated” writer may introduce learned features and use stereotyped expressions containing obsolete forms. A sensitivity to the possible variations of “literariness” is a prerequisite in the researcher. Works of literature, whether texts of Classical or of later authors, also survive on papyrus (less often stone) and may be included in modern editions of documents. 27 They are by definition not “documentary,” but in terms of language the line is not so easily drawn. The language of literary works overlaps with that of everyday language. For enquiry into the Septuagint vocabulary this source of evidence is not to be neglected, both for everyday words and for literary vocabulary if “literary” books or sections of the Septuagint are in view. In any case, any occurrence of a word in any source may be useful for filling in the history of that word. The value of the documentary evidence for study of the vocabulary of the Septuagint is essentially that it enables us to understand the Septuagint better. All evidence is valuable, but this body of additional and growing evidence, as was said above, gives access to a dimension of the language that was inadequately known before. The mate24. Full survey in B. H. McLean, An Introduction to Greek Epigraphy of the Hellenistic and Roman Periods from Alexander the Great down to the Reign of Constantine (323 B.C.– A.D. 337), Ann Arbor, MI 2002, 181-210. 25. Cf. survey in Bagnall, Handbook, 358-394 (B. Palme). 26. Besides Greek, there are of course documents in other languages, notably Latin, Egyptian Demotic and Coptic, and also bilingual texts. 27. The categories “subliterary” and “paraliterary” are also used: see Bagnall, Handbook, 282-283 (T. Renner).

102

3. Nature and Value of the Evidence

5. The Vocabulary of the Septuagint and Documentary Evidence

rial that is closest in time to the Septuagint is the most valuable, and there is by good fortune a large body of Ptolemaic papyri. Elucidation involves a quest for parallels, and this has the additional outcome of enabling us to conclude (up to a point) whether a word or use belongs to normal Greek or is peculiar to the Septuagint. A further aspect of such researches is important. The traffic is not all one way: the Septuagint itself, used with due caution, is a witness to Koine Greek. By bringing the Septuagint and documentary evidence together we may elucidate or support either by the other.

4. Practicalities If access to the documentary texts were just a matter of consulting a printed lexicon or other tool, this section would be unnecessary. As it is, the constant growth of the material and rapid developments in technology have meant that printed works could not possibly keep pace. Scholarship itself cannot keep pace. Consulting any of the existing reference works, such as LSJ, the lexicon of Friedrich Preisigke and Emil Kiessling, 28 or MM, gives only a partial and often misleading view. On the other hand, full access to the documentary texts themselves has become easy, transformed in recent years by electronic databases and search tools. Any word can be searched for and found electronically with great speed. But therein lies a difficulty. The results are raw material, a vast body of it, requiring expertise and sheer hard work to make use of. A list of occurrences is just the beginning. The data need thorough analysis if they are to be useful for any purpose. Quick and easy answers will not be found. Until the arrival of digitization, documentary texts were accessible only in the printed volumes in which they were published. Searching could be done only via the indexes supplied (or not) by the editors, or by continuous reading. Whatever the drawbacks of this form, the book had, and still has, the advantage that discussion, notes and translation (if provided) can all be readily consulted in one place on the printed page. 29 With the transfer of the texts to electronic form much of this assistance has been omitted (though no doubt it will in time be added). Print publication of texts is however still the norm and is the source from which electronic texts are derived (though this too could change). The upshot at present is that for an accurate report of the text and aids to comprehension, it is necessary, or at least desirable, to consult the printed volume. 30 Published volumes of documentary texts number many hundreds, and to manage the abundance, lists and abbreviations have been created. 31 The secondary 28. F. Preisigke, Wörterbuch der griechischen Papyrusurkunden, 3 vols., Berlin 1925-1931; E. Kiessling et al., vol. 4, 1944-1993; supplements, 1969-1991. 29. Translations are an important aid. Besides those in the editions, there are selections of translated documents, which can usually be located via one of the websites (see, below, notes 33 and 35). 30. This of course is easier said than done: it is a rare library that possesses even a majority of the volumes, let alone all. Epigraphic volumes are especially intractable. 31. For papyri: J. F. Oates et al., Checklist of Editions of Greek, Latin, Demotic and Coptic Papyri, Ostraca and Tablets (fifth edition), Oakville, CT 2001. Epigraphic abbreviations are in disarray. A start was made on a unified system in G. H. R. Horsley / J. A. L. Lee, “A Preliminary Checklist of Abbreviations of Greek Epigraphic Volumes” Epigraphica 56 (1994), 129-169. This 4. Practicalities

103

5. The Vocabulary of the Septuagint and Documentary Evidence

bibliography, itself enormous, will be found listed to some extent in the printed volumes, but complete only up to the date of publication; the electronic media are in the process of bringing this under better control. Electronic searching, which until recently was possible only via CDs, can now be done online. The current main search locations may be noted here, 32 but further change and development are inevitable. There are also multiple sites gathering different types of information. 33 A significant development has been the creation of the “Trismegistos” database, which aims to gather “metadata” on all published papyrological and epigraphic texts from Egypt, in the period 800 BCE to 800 CE. 34 It also sets out to provide links to all the main projects, and has begun the process of assigning a unique number to each documentary text, an essential device for management of the proliferating information. There are good general introductions to the two fields, by B. Hudson McLean in epigraphy 35 and Roger Bagnall in papyrology. 36 For an introduction to Koine Greek, works by Robert Browning 37 and Geoffrey Horrocks 38 are the best on offer. Important reference grammars are those of Edwin Mayser 39 and Francis Gignac, 40 but vocabulary is treated only incidentally in these works. A lexicon or extended treatment of the Koine Greek vocabulary is non-existent, though MM has a useful collection, provided it is not regarded as the final word. As for the Septuagint, GELS does not offer documentary evidence as such or undertake new research in the field, but simply uses the results of existing studies; but it is the essential guide to the meanings of words in the Septuagint, based on a fresh and complete lexical analysis. LSJ’s material on the Septuagint is patchy and often unreliable; but its coverage of documentary evidence, though limited, is more trustworthy, at least as a starting-point. The need for expertise has been mentioned. Of course this can be said of any subject, but work on documentary texts can be especially challenging for those coming to it from other disciplines. There is first the need for familiarity with a wide range of types of Greek, from Classical to late Koine, from learned to colloquial. This includes

32. 33. 34. 35. 36. 37. 38. 39. 40.

is adopted in McLean, Introduction, 387-472, with additions to 2002. The system in PHI Greek Inscriptions is different; CLAROS (hmoderna1.ih.csic.es/claros/cnc/2cnc.htmi) offers another; LSJ and Revised Supplement, Oxford 1996, inevitably have another. A useful guide to epigraphic works is F. Bérard et al., Guide de l’épigraphiste: Bibliographie choisie des épigraphies antiques et médiévales (third edition), Paris 2000. hpapyri.infoi and hepigraphy.packhum.org/inscriptions/i. See further McLean, Introduction, 383-385; Bagnall, Handbook, xv-xvi. hwww.trismegistos.orgi. McLean, Introduction. Also: A. G. Woodhead, The Study of Greek Inscriptions (second edition), Cambridge 1981 (reprint Norman, OK 1992). Bagnall, Handbook. Also: E. G. Turner, Greek Papyri: An Introduction (second edition), Oxford 1980. R. Browning, Medieval and Modern Greek (second edition), Cambridge 1983, chapter 2. G. C. Horrocks, Greek: A History of the Language and its Speakers (second edition), Oxford 2010. E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit, 6 vols., Berlin/Leipzig 1906-1938 (vol. 1.1: 2, revised edition, H. Schmoll, Berlin 1970). F. T. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, 2 vols., Milan 1976, 1981. Epigraphy lacks any equivalent to Gignac and Mayser.

104

4. Practicalities

5. The Vocabulary of the Septuagint and Documentary Evidence

awareness of the phonetic and other changes in the later period; to approach Koine Greek as a sort of debased Classical Greek is a serious mistake. Even with this knowledge, to understand the technical and other terminology, or even just to grasp the purpose and background of a text, can be demanding. The word being studied will not necessarily yield its meaning quickly, and extensive enquiry may be needed. Some particular difficulties deserve notice. Incomplete texts invite restoration, but not every restoration found in the published editions is to be trusted. The researcher needs to be wary of accepting any word fully or even partially restored, unless it is part of a formula or other predictable expression. 41 Another problem is “correction” of originals by editors, that is to say, when the original presents a non-standard spelling or other feature, the editor offers the “correct” form. Some of the corrections go beyond indicating the standard spelling and alter forms or syntax that actually belong to the Greek of the time. 42 Finally, dating. This is of vital concern to the lexical enquirer, but there are many obstacles put in the way of ascertaining the date of a document quickly. These include total omission of “BCE” and “CE” or equivalents, the use of a variety of dating formulae (some in Latin), 43 and most frustrating of all, failure to offer any date whatever, even within wide parameters.

5. Illustrations 5.1 ποτίζω Let us begin with a straightforward example of a new use in the Septuagint paralleled in contemporary Egyptian papyri. 44 In the Classical period ποτίζω is well attested, but not in the meaning “irrigate” (land); the word for this was ἄρδω. In the Pentateuch and occasionally in later books ποτίζω occurs in this sense, ἄρδω not at all. 45 An example: Gen 2:6 πηγὴ δὲ ἀνέβαινεν ἐκ τῆς γῆς καὶ ἐπότιζεν πᾶν τὸ πρόσωπον τῆς γῆς.

The evidence of the Ptolemaic papyri shows that this use was standard in the time of the Septuagint translators. There are now thirty or so occurrences, of which one sample will suffice: PZenPestm Suppl. B p. 266.2, 3 (256 or 255 BCE) Ζήνων Κλέωνι χαίρειν. τὸ ὕδωρ τὸ ἐ�[ν τῆι διώρυγι οὐκ ἀνα]βέβη[κ]εν πλείω ἢ� [πῆ]χυν | ὥστε μὴ δύνασθαι ἀπ᾽ αὐτῆς ποτίζε[σθαι τὴν γῆν. καλῶς ἂν ο]ὖ�� �[ο]ήσαις ἀνοίξας | τὰς θύρας, ἵνα ποτίζηται ἡ γῆ. Zenon to Kleon greetings. The water in [the canal has not] come up more than a

41. Cf. Turner, Greek Papyri, 67-70; McLean, Introduction, 35-39. Acquaintance with the editorial sigla (square brackets, subscript dots, etc.) is also a requisite: see Bagnall, Handbook, 203 (P. Schubert); McLean, Introduction, 27-35. 42. An example: in PCairZen 2.59155 χερός is corrected to χειρός, but the former is a genuine form, not a misspelling. 43. Some variations border on the ridiculous, e. g.: “sac” = saeculum ante Christum, i. e., BCE, but “ac” = after Christ, i. e., CE; but then “a.C.” = ante Christum, i. e., BCE. 44. Cf. Lee, Lexical Study, 118-122. 45. ποτίζω “irrigate”: Gen 2:6,10; 13:10; Deut 11:10; Sir 24:31; Joel 4:18. 5. Illustrations

105

5. The Vocabulary of the Septuagint and Documentary Evidence

cubit, so it is not possible for [the land] to be irrigated from that source. Please, then, open the sluice-gates, so that the land may be irrigated.

In Deut 11:10 the word appears in a special expression, which is worth looking into briefly: ὅταν σπείρωσιν τὸν σπόρον καὶ ποτίζωσιν τοῖς ποσὶν ὡσεὶ κῆπον λαχανείας (“when they sow the seed and irrigate it with their feet like a vegetable garden”). The documentary evidence does not provide an exact parallel, but it comes tantalizingly close: there are examples of ποτίζω / ποτισμὸς ἀπὸ ποδός (“irrigate/-ation by foot”) in papyri of the first and second centuries CE. 46 This looks like the same practice, but what method of irrigation was it? The debate has been drawn-out, but there is a probable answer. 47 The evidence of the papyri makes an important contribution to tracing the history of ποτίζω. Without it the literary record is not very informative: if we trust LSJ this use appears to occur only in Lucian (II CE). 48 The inscriptions also happen to offer only late occurrences of ποτίζω, none of them in the sense of “irrigate.”

5.2 εὐίλατος This adjective (“kindly disposed,” “merciful”) appears twice in the Septuagint: 1Esdras 8:53 καὶ πάλιν ἐδεήθημεν τοῦ κυρίου ἡμῶν κατὰ ταῦτα καὶ εὐιλάτου ἐτύχομεν And again we entreated our Lord in these matters and found him merciful. Ps 98:8 ὁ θεός, σὺ εὐίλατος ἐγίνου αὐτοῖς καὶ ἐκδικῶν ἐπὶ πάντα τὰ ἐπιτηδεύματα αὐτῶν O God, you were merciful to them, though(?) taking vengeance on all their practices.

What attestation does it have outside the Septuagint? The results of a thorough search are remarkable. Aside from the Septuagint itself, it does not appear in literature until much later, namely, Patristic texts from the second century CE on (mostly quoting Ps 98:8) and the Byzantine lexicographers; 49 its attestation before then is entirely documentary. And this evidence reveals what we would otherwise not have known, that εὐίλατος was in extensive use in the time of the Septuagint, and not just in Egypt. LSJ (1925-40) were able to cite six occurrences in documents; the tally today is almost five times as many. The bulk of the texts fall in the period from the third century BCE 46. The current tally is five: PSoter 1.32 (69 CE), 2.26 (71 CE), PFlor 3.369.7 (139/149 CE), PRyl 2.157.21 (135 CE), PMilVogl 7.308.97 (II CE); similarly Cadell, “Vocabulaire,” 114-115; this has grown from two in Lee, Lexical Study, 121-122. (SIFC 13.366 in LSJ Revised Supplement = PFlor 3.369.) 47. Cf. Lee, Lexical Study, 121-122; C. Dogniez / M. Harl, La Bible d’Alexandrie: Le Deutéronome, Paris 1992, ad loc.; Cadell, “Vocabulaire,” 110-117; J. M. Wevers, Notes on the Greek Text of Deuteronomy, Atlanta, GA 1995, ad loc.; LSJ Revised Supplement, s. v. πούς; GELS, s. v. ποτίζω. Cadell examines all the possibilities and returns to the suggestion (in Lee, from Driver) that it involves use of the feet to control the flow of water on to the garden through channels in the earth, also known in modern parallels. 48. Luc. Abd. 27 “etc.” The truth of course is that there are more, as, e. g., in Strabo the Geographer (I BCE/I CE) at 17.1.3; 17.1.4. These were found by a search in TLG. 49. Results from TLG search. An example attributed to Antimachus (V/IV BCE) in Et. Mag. s. v. Ἀδόροις is to be read as εὐήλατος (Frag. 109, ed. Wyss).

106

5. Illustrations

5. The Vocabulary of the Septuagint and Documentary Evidence

to the first century CE; geographically they range from Egypt, through Asia Minor (notably Knidos in Caria), Attica and Thrace, to Pompeii. 50 A small selection must suffice: IKnidos 1.150.A.9 (Knidos, II-I BCE) μὴ τύχοι Δάματρος καὶ | [Κ]όρας μηδὲ θεῶν τῶν παρὰ Δάματρος εὐιλάτων May he not find Demeter and Kore nor the gods with Demeter well-disposed. 51 IEgVers 175.I.36 (Narmouthis, I BCE?) κλῦθι ἐμῶν εὐχῶν, μεγαλοσθενὲς οὔνομ᾽ ἔχουσα | εὐείλατος ἐμοί τε γείνου, λύπης μ᾽ ἀνάπαυσον ἁπάσης. Hear my prayers, O mighty-named one [= Isis-Thermouthis] | and be merciful to me, give me rest from all pain. IG II2 1365.25 (Attica: Sunium, I CE) = CMRDM 1.12.25 (n. d.) εὐίλατος γένοιτο ὁ θεὸς τοῖς | θεραπεύουσιν ἁπλῇ τῇ ψυχῇ. May the god [Men Tyrannos] be merciful to those who tend him with simplicity of soul. PSI 4.392.6 (242/1 BCE) … τοῦ βασιλέως εὐειλάτου γενομένου καὶ ἐπιγράψαντος τὴν ἄφεσιν. … if the king is merciful and signs the release.

Something further is evident. This is a word with an elevated, religious tone, applied in the documentary texts only to gods or King Ptolemy (also a god). The Psalms and Esdras translators knew what they were doing when they chose it; even the expressions match. 52 The Septuagint also has the verb εὐιλατεύω, derived from εὐίλατος, in Deut 29:19, Jdt 16:15 and Ps 102:3, all with the same religious tone. So far there is no other attestation, except in later Patristic texts and lexicographers. But it would be unwise to conclude that it was a neologism of the Septuagint translators; and if past experience is any guide, there is a good chance that it will turn up.

5.3 ἀγαπῶ A full study of this word would be a large subject. The point to be made here is that documentary evidence supports the view that ἀγαπῶ began to supplant φιλῶ as the word for “love” already in the fourth century BCE and continued on the same path later. This was demonstrated by Robert Joly from the literary evidence. 53 It follows that the Septuagint translators used ἀγαπῶ because it was current usage, not because they intended to reject “pagan” usage and create a new vocabulary for “love” words, passed on in due course to the New Testament. Documentary examples from the time of the 50. There is also a later example in a magical papyrus, PGM 2.166 (IV CE?): δέσποτα Μουσῶ[ν], ἵλαθί μοι, τῷ σῷ ἱκέτῃ, καὶ ἔσο εὐμενὴς καὶ εὐίλατος, φάνηθί μοι καθαρῷ τῷ προσώπῳ. 51. Doric α for η in the goddesses’ names illustrates the continuation of older dialect: Knidos was an old Doric foundation. There are about ten other similar texts from Knidos. 52. Zipora Talshir (1 Esdras: A Text Critical Commentary, Atlanta, GA 2001, on 8:53) missed an opportunity: she quotes just one example, taken from LSJ (without the reference, PPetr. 2 p. 45). 53. R. Joly, Le vocabulaire chrétien de l’amour est-il original? φιλεῖν et ἀγαπᾶν dans le grec antique, Bruxelles 1968. A valuable exposition is found in S. P. Swinn, “Ἀγαπᾶν in the Septuagint” in: T. Muraoka (ed.), Melbourne Symposium on Septuagint Lexicography, Atlanta, GA 1990, 49-81. 5. Illustrations

107

5. The Vocabulary of the Septuagint and Documentary Evidence

Septuagint are very few, but they provide evidence that ἀγαπῶ “love” was used not just in Egypt but in the furthest corners of the Greek-speaking world: OGI 1.90.9 (Rosetta, 196 BCE) ἐπειδὴ βασιλεὺς Πτολεμαῖος αἰωνόβιος, ἠγαπημένος ὑπὸ τοῦ Φθᾶ, θεὸς Ἐπιφανὴς Εὐχάριστος … Whereas King Ptolemy, living eternally, loved by Phtha, god Manifest, Gracious … 54 IGExtrEast 292.9 (Kandahar, mid III BCE) 55 ἔδει … διδάσκαλον καὶ πατέρα καὶ μητέρα | ἐπαισχύνεσθαι καὶ θαυμάζειν, φίλους καὶ ἑταίρους ἀγαπᾶν καὶ μὴ διαψεύδεσθαι, | δούλοις καὶ μισθωτοῖς ὡς κουφότατα χρᾶσθαι … They were to … show modesty and respect to their teacher and father and mother, to love their friends and companions and not deceive them, to treat slaves and hired servants as gently as possible …

6. Future Directions The time of establishing the relevance of the documentary evidence to the language of the Septuagint is past: it has been well and truly proven. Numerous studies over the past century have made forays into this resource and produced good findings. But a thorough, systematic exploration is still needed, one that is ongoing, as the material continues to grow and become even more accessible via electronic tools. For vocabulary we now have a basic tool of investigation, namely, a full lexical analysis of the Septuagint, provided by GELS. The next step is systematic searching of the documents for parallels (or not) to the whole gamut of usage of the Septuagint. We are also in great need of a record of the published discussions on Septuagint words, kept up to date as they appear. At present there is no reliable way of keeping track, and each researcher begins the quest all over again. A way of gathering the data into one accessible resource will be essential. In the past this might have been attempted by a printed lexicon, as it was for example by Bauer/BDAG for the New Testament, but the drawbacks of that type of resource are evident. The obvious tool now is an electronic database, able to gather all the material and keep it up to date. This is the new desideratum of Septuagint studies. In the long run, benefits to all scholarship on the Greek vocabulary will ensue from such a project. Not only will the vocabulary of the Septuagint be seen more fully in its context, but its own witness to the language of its time will be more evident. 56

54. There are several more instances of this formula in the Rosetta stone and copies. In PCairZen 4.580.3 (mid III BCE) ἀγαπῶ occurs without context. 55. The abbreviation problem illustrated. The volume is F. Canali De Rossi (ed.), Iscrizioni dello estremo oriente greco (IK 65), Bonn 2004. Abbreviation in PHI Greek Inscriptions: “IK Estremo oriente”; in CLAROS: “IEOG.” “IGExtrEast” is suggested on the pattern of the Horsley-Lee list. 56. My thanks to Trevor Evans for useful comments on this chapter.

108

6. Future Directions

6. Septuagint Syntax and Hellenistic Greek Anssi Voitila

1. Introduction In this overview, I will discuss Septuagint syntax as a part of the broader Post-Classical, Hellenistic, or Koine Greek syntax of the last three centuries BCE. Written in the Koine vernacular, the Septuagint, which was begun the third century BCE, constitutes one of the most important sources for the study of Post-Classical Greek. The Koine is an Ionicized form of non-parochial Attic, also known as Great Attic. Adopted by the Macedonian royal court, it served as a vehicle for the diffusion of Greek language and civilization throughout the Eastern Mediterranean and Asia during the conquests of Alexander the Great. This “common” dialect was not a fully developed linguistic system at the end of the fourth century BCE. While some of its characteristics were already in place before the Hellenistic period, others emerged only in subsequent centuries. Consequently, the Koine of the Greek Pentateuch or early Ptolemaic Papyri is not the same as the Koine of the New Testament. 1 Early works on Septuagint syntax studied individual books as if they were originally written in Greek. This method still has its followers today, particularly among Greek linguists. According to Geoffrey Horrocks: “The analysis of the ordinary language of contemporary private papyrus documents from Egypt has now demonstrated conclusively that the Septuagint’s general grammatical and lexical make-up is that of the ordinary, everyday written Greek of the times.” 2 This conclusion was preceded by these remarks: “The translation of the Pentateuch … seems to be in a very natural contemporary Koine (Thackeray 1909: 13), though certain other books display a rather mechanical literalness (e. g., Lamentations), while others exemplify a spread of styles ranging from near-vernacular (e. g., Tobit, from the Apocrypha) to consciously ‘literary’ (e. g., Esther, with 4Maccabees being positively Atticizing).” 3 This, however, is only partly correct, since the Septuagint’s “general grammatical and lexical make-up”

1.

2. 3.

See, e. g., C. Brixhe, La Koiné grecque antique, vol. I, Une langue introuvable? (Travaux et mémoires. Études anciennes 10), Nancy 1993; G. Horrocks, Greek: A History of the Language and its Speakers (Longman Linguistics Library), London/New York, NY 2007, 123 and C. Caragounis (ed.), Greek. A Language in Evolution, Hildesheim/Zürich/New York, NY 2010. For further studies on Hellenistic Greek syntax (mainly New Testament syntax), see http://www. greek-language.com/Palmer-bibiography.html. Horrocks, Greek, 107. Horrocks, Greek, 106-107. Horrocks’s views, as expressed in the 1997 edition of Greek, have come under criticism from C. Caragounis. See C. Caragounis, The Development of Greek and the New Testament (Wissenschaftliche Untersuchungen zum Neuen Testament 167), Tübingen 2004; C. Caragounis, “Atticism. Agenda and Achievement” in: Caragounis, Greek. A Language in Evolution, 168. 1. Introduction

109

6. Septuagint Syntax and Hellenistic Greek

does not always sound natural, but often reveals the linguistic characteristics of its parent text. 4 As a result, Septuagintal linguistic material has limitations for the study of Greek historical linguistics. The first of these limitations concerns the use of statistics. Some acceptable Greek linguistic elements may occur more frequently in the Septuagint than in other contemporary Greek corpora, because they were used to render recurrent expressions in the original Hebrew text. This is even more significant when the frequency of certain linguistic elements (e. g., καί-style, verb-subject-object word order, optative of wish) appears to increase over time. Second, literal translation practice does not necessarily produce Hebraistic Greek, or translationese; some literal translations, such as the optative of wish or the rendering of the partitive ‫ מן‬by the prepositions ἀπό and ἐκ, are acceptable Greek. Third, some linguistic elements common in Greek do not have a direct equivalent in Hebrew (e. g., some moods, the verb ἔχειν, and certain participle constructions); their presence in the Septuagint reflects their pervasiveness in the contemporary Greek usage of the translator or his community. For all these reasons, the translation character of the Septuagint’s language must be recognized. One of the most powerful methods for examining this phenomenon is translation technical study. 5 This technique compares all the occurrences of a given linguistic element to its equivalent in the source text. The linguistic material thus obtained is then examined in contrast to contemporary Greek material (papyri, inscriptions, texts) in order to evaluate its idiomaticity. It is therefore possible to determine the source of a given linguistic element, either the original Hebrew text or the natural Greek usage. Elements that cannot be explained as direct equivalents of the Hebrew text can be safely assumed to reflect the natural language competence of the translators, unless they constitute deviations from the original text, which the translation technical method properly applied will also reveal. In the following sections, we focus on syntactic features known to have been evolving during the last three centuries BCE and others that disappeared at the end of Hellenistic-Roman period or that otherwise differed in comparison to Classical or Ancient Greek. We then look at how these features relate to Septuagint syntax. Most examples are taken from the translated books, especially the Pentateuch, for they have received the greatest attention from a syntactic point of view.

2. The Shift toward Verb-Subject-Object Word Order Many scholars have observed an increasing tendency from the fourth century BCE onwards to place clitic pronouns immediately after their governing words, instead of 4.

5.

For a discussion of the methodology of those scholars who treat the Septuagint as a document originally written in Greek, see A. Voitila, “La Septante: un document linguistique de la koiné grecque antique” in: J. Joosten / P. Le Moigne, eds, L’apport de la Septante aux études sur l’antiquité, Paris 2005, 17-35. The methodology of the study of Septuagint syntax has evolved thanks to the work of I. Soisalon-Soininen and his students. See also J. Lust, “Syntax and Translation Greek” EThL 77 (2001), 395-401. For more about this method and the relevant literature, see the article by Raija Sollamo in this volume.

110

2. The Shift toward Verb-Subject-Object Word Order

6. Septuagint Syntax and Hellenistic Greek

immediately following the first accented word of the sentence. This phenomenon, called the “law of Wackernagel,” appears to be related to the displacement of the sentence accent, 6 which causes the clitics to assume the second position in relation to the focus word. 7 For example, σήμερον ἐθεασάμην σε occurs alongside σήμερόν σε ἐθεασάμην (GG II, ii, 183). 8 Similarly, we encounter ἔκτεινον τὴν χεῖρα σου as well as ἔκτεινόν σου τὴν χεῖρα. According to Horrocks, this development places greater focus on the verb, which then moves toward the beginning of the sentence. This tendency culminates in the gradual standardisation of VCliSO (verb-clitic-subject-object) word order, the most common word order in the Septuagint, although not the only one. 9 On the basis of this, Horrocks argues that the word order of the Septuagint reflects a general development in the Greek of the period. This would be a satisfactory explanation if the word order of the Septuagint did not follow that of the Hebrew source text. The only justified conclusion would then be that the tendency to place clitic pronouns after their governing words and the gradual standardisation of the VCliSO word order made it easier for the translators to translate literally by reproducing the Hebrew word order in the Greek text. It should however be noted that verbsubject-object word order is not the only one in Hebrew, a language which has sentence types that afford translators freedom in rendering word order, e. g. verbless clauses. 10

3. Paratactic Style A similar conclusion may be reached concerning the so-called καί-style in the Septuagint. Horrocks argues that “the simple paratactic style […] although this certainly reflects the organization of the original text […] is characteristic in some degree of all mid- to low-level writing in the Koine, and in fact constitutes a feature of unsophisticated non-literary language throughout the history of Greek”; he does however recognize that the redundant use of καί at the beginning of an apodosis constitutes a Hebraism. 11 In the Septuagint, the paratactic style again reproduces Hebrew syntax. 6. See M. Janse, “La phrase segmentée en grec ancien – le témoignage des enclitiques” BSLP 86 (1991), xiv-xvi (citing C. Bally). 7. G. Dunn, “Enclitic Pronoun Movement and the Ancient Greek Sentence Accent” Glotta 67 (1989), 1-19; G. Horrocks, “Clitics in Greek—A Diachronic Review” in: M. Rossou / S. Panteli (ed.), Greek Outside Greece, vol. II, Issues in Language, Literature, and Education, Athens 1990, 35-52; Horrocks, Greek, 59-60; Janse, “La phrase,” xiv-xvi; Horrocks, Greek, 108-109; M. Janse, “La position des pronoms personnels enclitiques en grec néo-testamentaire à la lumière des dialectes néo-helléniques” in: Brixhe, La Koiné, vol. 1, 83-121. 8. These examples are taken from Apollonius Dyscolus (GG II, ii, 183 and 184), quoted in M. Janse, “La position des pronoms,” 88. 9. See R. Sollamo, “Enclitic personal pronouns in the Psalter” in: M. K. H. Peters (ed.), XII Congress of the International Organization for Septuagint and Cognate Studies, Leiden 2004 (SBLSCS 54), Atlanta, GA 2006, 153-160. 10. For statistics and more detailed discussion, see Voitila, “La Septante,” 30-31. 11. Horrocks, Greek, 107; M. Janse, “La koiné au contact des langues sémitiques de la Septante au Nouveau Testament (Questions de méthode)” in: C. Brixhe, La Koiné grecque antique, 3. Paratactic Style

111

6. Septuagint Syntax and Hellenistic Greek

Despite a tendency to limit the number of καί-clauses in the Septuagint, their number remains higher than in even the most colloquial of the contemporary documents, too high indeed to convey a natural impression. 12

4. The Reduction in the Use of Infinitival and Participial Constructions During the last three centuries BCE, the articulate infinitive, particularly in the genitive, gained popularity, although, at the same time, the infinitival and participial constructions—so common in classical Greek literature—tended to be replaced by ὅτιand ἵνα/ὅπως-clauses, 13 partly because the conjunctions ἵνα and ὅπως were deprived of their vitality as means of expressing purpose. In the idiomatic Greek usage of the period, τοῦ + infinitive was used in all types of infinitive phrases, replacing the construction τό + infinitive characteristic of the classical period, e. g., ἠθέλησεν τοῦ γενέσθαι με βασιλέα (1Chron 28:4). 14 As an author writing the literary Koine, Polybius used infinitives more sparingly than Plato or Lysias—but more often than 2Maccabees, Didorus Siculus or the Gospel of Mark—because he favored complex infinitival structures with articulate infinitives. Again, the proportion of subordinate clauses in comparison with articulate, particularly prepositional, infinitival phrases is much lower than in Thucydides or Xenophon. 15 In the Septuagint, τοῦ + infinitive occurs more frequently than the bare infinitive, except in the Pentateuch and in the books that immediately follow in the canonical order. 16 In the Pentateuch, the plain infinitive, τοῦ + infinitive, 17 and ἵνα/ὅπως-clauses 18 are sometimes used to render paratactic purpose clauses. The Hebrew infinitive was also translated by ὅτι- and ἵνα/ὅπωςclauses as well, albeit not often. Moreover, in a few instances, direct speech was replaced by indirect speech. 19 However, the Greek infinitive clearly provides the closest

12.

13.

14.

15. 16. 17. 18. 19.

vol. III, Les Contacts (Études anciennes 17), Nancy/Paris 1998, 99-111, in particular 104. See also S. Trenkner, Le style kai dans le récit attique oral (Bibliotheca Classica Vangorcumiana IX), Assen 1960; J. Blomqvist, “Translation Greek in the Trilingual Inscription of Xanthus” (Opuscula Atheniensia XIV, Skrifter utgivna av svenska institutet i Athen, 4, 29), Stockholm 1982, 1719; K. Dover, Evolution of Greek Prose Style, Oxford 1999, 70-72. A. Aejmelaeus, Parataxis in the Septuagint. A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch (Annales Academiae Scientiarum Fennicae/Dissertationes Humanarum Litterarum 31), Helsinki 1982, 31-33. E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluß der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften (Satzlehre II:1), Berlin/Leipzig 1926, 296-302, 310-314; Horrocks, Greek, 92-96. C. J. Ruijgh, “Sur l’emploi complétif de l’infinitif grec” in: B. Jacquinod, Les complétives en grec ancien (Centre Jean-Palerne CNRS GDR 1038, Memoirs XVIII), Saint Etienne 1999, 215231, in particular 226. J.-A. de Foucault, Recherches sur la langue et le style de Polybe, dissertation, université de Paris, faculté des lettres et sciences humaines, 1972, 157, 161-169, 196, 321. I. Soisalon-Soininen, Die Infinitive in der Septuaginta (Annales Academiae Scientiarum Fennicae Serie B 132.1), Helsinki 1965, 38, 50-54, 180-185. Soisalon-Soininen, Die Infinitive, 122-123; Aejmelaeus, Parataxis, 114. Aejmelaeus, Parataxis, 68-72. Soisalon-Soininen, Die Infinitive, 36, 42, 43, 57-59, 128-129.

112

4. The Reduction in the Use of Infinitival and Participial Constructions

6. Septuagint Syntax and Hellenistic Greek

equivalent to the Hebrew infinitive; this is particularly true in the case of the articulate infinitive, which corresponds to the Hebrew infinitive preceded by the preposition ‫ל‬. Although free renderings of conjunctions and unconstrained transformation of the clause structure are typical of the idiomatic Greek of the time, it is important to note that the predilection for the infinitive corresponds to Polybius’s usage. The use of participles declines throughout the last three centuries BCE in comparison with classical literary Attic. Since participle constructions (accusative and participle, connected participle, genitive absolute) do not have direct counterparts in Hebrew, they are seldom found in the Septuagint. However, their use as equivalents for other Hebrew constructions (parataxis, infinitive preceded by a preposition) shows that participle constructions were still a common feature of the Greek of the period. 20

5. Decline of the Optative Mood The use of the optative declined throughout the Hellenistic period, except for its function of expressing a wish. The optative of subordinate clauses in past contexts was most certainly inexistent in the vernacular; and while the potential optative still appears in the Ptolemaic papyri, it gradually becomes less frequent toward the end of the period. 21 In the Septuagint, due to the Hebrew influence, the most common use of this mood is the optative of wish (82.50 % of all occurrences in the Pentateuch); its most usual equivalent is the jussive. By contrast, there are only six examples of the potential optative in the Greek Pentateuch, a fact which may be explained by the absence of a Hebrew counterpart. 22 Nevertheless, these occurrences demonstrate that the use of the potential optative had not altogether disappeared from the Greek idiom of the translators. At the same time, a new use of the optative emerged from the Greek Pentateuch: the comparative optative, i. e., ὡς εἰ and εἰ in conjunction with the optative in comparative clauses. 23 This usage has no direct equivalent in Hebrew. It is known in the Homeric language but is rare in classical literature, and Edwin Mayser gives only one example in the Ptolemaic papyri. Trevor Evans argues that it represents a Homeric reminiscence of Hellenistic Jewish writers. 24 Indeed, it may be a throwback to the translators’ Greek school education, where Homer was a part of the curriculum. The iterative/indefinite condition (protasis: optative; apodosis: augmented verb form with or without ἄν) in subordinate clauses became increasingly rare throughout

20. See Soisalon-Soininen, Die Infinitive; Aejmelaeus, Parataxis. 21. Mayser, Grammatik, II:1, 288-296; Foucault, Recherches, 145-155; Horrocks, Greek, 82, 97, 102-103, 117. 22. T. Evans, Verbal Syntax in the Greek Pentateuch, Oxford 2001, 174-197. There seems to be some overlap between the potential and the optative of wish in the Greek Job. See S.-H. Woo, Études sur le système verbal dans la Septante de Job, dissertation, université de Strasbourg II, 2006, 113. 23. J. Joosten, “Elaborate Similes – Hebrew and Greek: A Study in Septuagint Translation Technique” Biblica 77 (1996), 227-236; T. Evans, “The Comparative Optative: A Homeric Reminiscence in the Greek Pentateuch?” VT 49 (1999), 487-504. 24. Evans, Verbal Syntax, 192-195. 5. Decline of the Optative Mood

113

6. Septuagint Syntax and Hellenistic Greek

the Hellenistic period, 25 and was no longer used in the Septuagint and the New Testament. 26 The oblique optative was progressively replaced by modes of direct speech. It seldom appears in the Ptolemaic Papyri or in Polybius, and no examples exist in the translated books of the Septuagint, although this may be partly explained by the scarcity of corresponding phrases in Hebrew. On the other hand, a few examples can be found in the non-translated books, e. g. 2Mac 3:37; 4Mac 17:1. 27

6. Verbal Syntax As Evans observed, “verbal syntax in the Greek Pentateuch may be characterized as typical of early Koine vernacular usage. The Attic structures are still largely intact.” 28 For instance, the so-called historic present, a tense that is relatively rare in Hellenistic prose and papyri, is attested, albeit infrequently, in the Septuagint. Furthermore, the Septuagint contains no unambiguous examples of the fusion of the perfect tense and the aorist, although there are signs that such a change might have been impending. 29 The resultant state perfect begins to resemble the extended now of the English perfect, in which “the perfect serves to locate an event within a period of time that began in the past and extends up to the present moment.” 30 Gradually, new uses of verbal syntax were introduced. The indicative present and the imperfect acquire modal value. 31 There is at least one example of contrafactual conditions (irrealis/unreal), i. e., a past tense indicative, without ἄν in the apodosis:

25. There are five occurrences in the first five books of Polybius’ Histories and no examples in Mayser, Grammatik II:1. 26. A. Tjen, On Conditionals in the Greek Pentateuch: A Study of Translation Syntax (Library of Hebrew Bible/Old Testament Studies 515), New York, NY/London 2010, 40. 27. In Job 23:3, the introductory verb is also an optative. See Woo, Etudes, 114. 28. Evans, Verbal Syntax, 262; A. Voitila, “La Technique de traduction du yiqtol (l’imparfait hébreu) dans l’histoire de Joseph grecque (Gen 37, 39-50)” in: C. E. Cox (ed.), VII Congress of the IOSCS, Leuven 1989 (Septuagint and Cognate Studies 31), Atlanta, GA 1991, 223-237; A. Voitila, “Some Remarks on the Perfect Indicative in the Septuagint” Bulletin of IOSCS 23 (1993), 11-16; A. Voitila, Présent et imparfait de l’indicatif dans le Pentateuque grec: une étude sur la syntaxe de traduction (Publications de la Société d’Exégèse de Finlande 79), Helsinki/Göttingen 2001. 29. E.-C. Gerö / A. von Stechow, Tense in Time: The Greek Perfect, Tübingen 2002, 1-40; D. Haug, “Aristotle’s Kinesis/Energeia-Test and the Semantics of the Greek Perfect” Linguistics 42 (2004), 387-418; A. Voitila, “The Use of Tenses in the L- and B-Texts in the KaigeSection of 2 Reigns” in: S. Kreuzer / M. Meiser / M. Sigismund (eds.), Die Septuaginta: Entstehung, Sprache, Geschichte (Wissenschaftliche Untersuchungen zum Neuen Testament 286), Tübingen 2010, 213-237. There are seemingly ambiguous cases in the book of Job, though one must take into account that the instances of “aorist-like” perfect studied by Woo (Etudes, 139143) are mostly first- and second-person forms and that the temporal axis of discourse in Job is the moment of speech. 30. D. R. Dowty, Word Meaning and Montague Grammar, Dordrecht 1979, 341. 31. See Voitila, Présent et imparfait.

114

6. Verbal Syntax

6. Septuagint Syntax and Hellenistic Greek

καὶ εἰ μὴ ἐξέκλινεν, νῦν οὖν σὲ μὲν ἀπέκτεινα, ἐκείνην δὲ περιεποιησάμην (Num 22:33). 32 This phenomenon is also attested in the Ptolemaic papyri. 33 During this period or slightly earlier, the future and subjunctive began to merge, due partly to the loss of vowel length distinctions. 34 There are examples of this phenomenon in the Septuagint Pentateuch (Gen 43:4; 44:16) and in Job (Job 23:15; 37:19). The future indicative has an imperatival function both in the Septuagint and in the Ptolemaic legal documents, although frequency varies within each corpus and from one corpus to the other. 35 The future infinitive and the future participle disappear during this period from the spoken language. 36 The future infinitive occurs eleven times in the Septuagint, three occurrences of which are to be found in the Pentateuch. 37 The future participle occurs three times in the Greek Pentateuch, 38 more frequently in the non-translated books of the Septuagint. An increase in the use of various periphrastic constructions with auxiliary verbs εἶναι, γίνεσθαι, θέλειν and μέλλω + infinitive can be seen in the Koine Greek text material. In comparison, θέλειν and μέλλω + infinitive, used as a future tense periphrasis, are extremely rare in Ptolemaic papyri 39 and in the Greek Pentateuch. 40 The construction μέλλω + infinitive occurs more frequently in the later books of the Septuagint. 41 The form θέλω + infinitive is attested only once as a periphrasis for “mean/ intend” (Exod 2:14). The voice system was progressively simplified during the Hellenistic period. The middle and passive voices merge morphologically and functionally. Although this phenomenon has not yet been properly studied in the Septuagint, it is worth noting that ἐγένετο, ἀπεκρινάμην (versus ἐγενήθη, ἀπεκρίθην) appear in both middle and passive aorist forms with no clear difference in meaning. 42 32. Anwar Tjen (On Conditionals, 141, 203-204, 219) does not mention this example. 33. Mayser, Grammatik, II:1, 227-228. 34. For instance, Horrocks, Greek, 117, 174. See, e. g., Euripides, Ion 758; it is also possible that the future, a relatively latecomer to the Greek verbal system, never fully took its place, particularly, in the modal system. 35. Tjen, On Conditionals, 189-193. 36. V. Magnien, Emplois et origines du futur grec, Paris 1912, 6-7, 59; Foucault, Recherches, 157158, 174-175; Mayser, Grammatik, II:1, 216-17, 384-385; Horrocks, Greek, 97, 117. 37. Magnien, Emplois, 59; Soisalon-Soininen, Die Infinitive, 150-151; Evans, Verbal Syntax, 128. 38. Magnien, Emplois, 6-7; Soisalon-Soininen, Die Infinitive, 59; Evans, Verbal Syntax, 129. 39. Mayser, Grammatik, II:1, 226. 40. Evans, Verbal Syntax, 227, although Theodore Markopoulos (The Future in Greek, Oxford 2009, 48-9) cites statistics that seem to indicate the contrary. 41. I have studied the auxiliary verb construction and its polysemous nature in: A. Voitila, “Deontic meaning of the auxiliary verb construction μέλλω + INF in the Septuagint and Hellenistic Greek” Glotta 89 (2013), 242-252, and A. Voitila, “Mέλλω-Auxiliary Verb Construction in the LXX,” in: T. M. Law / M. Liljeström / K. De Troyer (eds.), On the Trail: Studies in Honour of Anneli Aejmelaeus (CBET 72), Leuven 2014, 193-214. 42. The new understanding of the meaning of the middle voice in Classical Greek that R. J. Allan proposed in his monograph The Middle Voice in Ancient Greek: a Study in Polysemy (Amsterdam Studies in Classical Philology 11), Amsterdam 2003, makes even more urgent the study of the convergence of these two voices in the Septuagint, and in other texts of the Hellenistic period as well. 6. Verbal Syntax

115

6. Septuagint Syntax and Hellenistic Greek

7. Reduction of the Case System and Increase in the Use of Prepositions The case system continued to evolve throughout the Hellenistic period. The value of the dative and the genitive declined and had to be strengthened with prepositions; the dative decayed, disappearing entirely from the spoken language by the end of the Roman period. 43 In addition, there seems to have been a functional overlap between the dative and the genitive, as in P. Flor 127 44 λαμβάνεις μου (possessive: “my letters”) τὰ γράμματα (alternatively dative: “letters to me”). 45 Instead of the genitive we sometimes see prepositional phrases, 46 such as τὴν παρ᾽ ἐμοῦ ἐπιστολήν (IG IX.ii.517) or οἱ παρὰ σὲ θεοὶ “the gods beside you,” i. e., “your gods” (P.Par 47). There are similar instances in the Septuagint: Gen 17:7 displays εἶναί σου θεὸς καὶ τοῦ σπέρματός σου μετὰ σέ where a possessive dative would have been expected and Ps 31:7 (σύ μου εἶ καταφυγή) uses a genitive instead of a dative of interest (commodi). 47 The Hebrew genitive construction is sometimes rendered by prepositional phrases in association with a genitive: 48 τὴν ῥάβδον τὴν παρὰ τοῦ θεοῦ (Exod 4:20) or οἱ παρὰ τοῦ βασιλέως, namely ἄνδρες (1Esdras, 1Mac 2:15,17, etc.). Occasionally, an indirect dative object is replaced by an accusative, e. g. Exod 23:1; 32:14. 49 In Hellenistic Greek, the partitive genitive is frequently replaced by the prepositions ἀπό and ἐκ. 50 Increasingly, the prepositions ἀπό and ἐκ replace the partitive genitive: the latter is more frequent in Septuagint and Hellenistic Greek, whereas the former predominates in the New Testament. 51 As the dative case decayed, the preposition ἐν, used to strengthen the plain dative in temporal (ἐν παντὶ καιρῷ) and adverbial meanings (ἐν τάχει), increased in popularity. Nevertheless, it seems that the instrumental use of ἐν + dative, so common in the Septuagint, was not idiomatic in the earliest stages of the language or in Hellenistic Greek. According to Jean Humbert, even the most vulgar papyri show no signs of it. 43. Mayser, Grammatik, II:2, 336-337; Horrocks, Greek, 92; S. Luraghi, On the Meaning of Prepositions and Cases. The Expression of Semantic Roles in Ancient Greek (Studies in Language Companion Series 67), Philadelphia, PA 2003, 331-333. 44. The texts of the papyri have been taken from http://papyri.info/ (Duke Database of Documentary Papyri) provided by NYU Digital Library Technology Services and Institute for the Study of the Ancient World. 45. Horrocks, Greek, 116. 46. Foucault, Recherches, 109-116; examples taken from Horrocks, Greek, 91-92, 115. 47. Num 27:11. For other examples of God as owner, see: Exod 6:7; 29:45; Lev 11:45, etc. (Private correspondence with Katri Tenhunen). For Psalms see Sollamo, “Enclitic,” 154-155. 48. See also I. Soisalon-Soininen, “Verschiedene Wiedergaben der hebräischen status-constructus-verbindung im griechischen Pentateuch” in: I. Soisalon-Soininen, Studien zur Septuaginta-Syntax (Annales Academiae Scientiarum Fennicae Serie B 237), Helsinki 1987, 62-70, in particular 69. 49. Private correspondence with Katri Tenhunen. 50. Mayser, Grammatik, II:2, 348-354; Foucault, Recherches, 97-99. This development started earlier, see Luraghi, On the Meaning, 122, 124. 51. I. Soisalon-Soininen, “Die Wiedergabe des partitiven ‫ ב‬im griechischen Pentateuch” in: Soisalon-Soininen, Studien, 154-171, 170; C. Seppänen, Min-preposition kääntäminen Septuagintan Pentateukissa [“The rendering of preposition min in the Greek Pentateuch”], master’s thesis, University of Helsinki, (http://www.doria.fi/handle/10024/3009), 2006, 52-84.

116

7. Reduction of the Case System and Increase in the Use of Prepositions

6. Septuagint Syntax and Hellenistic Greek

He explains cases like P.Par 28 = UPZ I.48.13 διαλυόμεναι ἐν τῶι λιμῶι, (cf., P.Par 22 = UPZ I.19.21 τῶι λιμῶι διαλυθῆναι) and PSI IV 333,2 ἑλ�[υσ]θῆναί σε ἐν ἀρρωστίαι 52 as adverbial uses. 53 Humbert seems to imply that its use in the Septuagint is a literal imitation of the Hebrew preposition ‫ב‬, with instrumental meaning. In the Pentateuch, seemingly, the translators tried to reduce the number of occurrences of ἐν + dative by alternating it with the bare dative in order to render the Hebrew prepositional phrase. 54 The instrumental ἐν + dative was not however incomprehensible to Greek-speakers. Rather, it anticipated more concrete uses in the Septuagint. Thus, in the foregoing P.Par examples, the expression “to perish by hunger” combines a tangible meaning (lack of food can kill a person) and a more metaphorical one (hunger is not an instrument of killing in a same way that a sword is) in a way that makes sentences such as ἐν χειρὶ ἐξάγειν δεῖνα “to bring somebody out by/with a hand” (e. g. Exod 13:3,9) perfectively comprehensible for the Greek reader. Linguistic materials dated to the turn of the Common Era give clear evidence that the preposition εἰς (to, into) encroached on the semantic field of the preposition ἐν (in, inside) to such a degree that speakers were no longer able to distinguish between them. Mayser gives a few examples of such cases in the Ptolemaic papyri. 55 According to Ilmari Soisalon-Soininen 56, these cases are limited in number and may be explained by the way the noun governed by the preposition was perceived, either as in motion toward another object (εἰς direction, bewegung) or as occupying a specific area (ἐν location, ruhe), 57 e. g., τὰ ὀνόματα τῶν πραγματευομένων εἰς τοὺς λόγους γραφέτωσ[α]� (P.Rev Laws 7,2); ἐὰν [μ]\ὴ/(*) γράψωσι ἐν τοῖς λό�[οις (P.Pev Laws 5,15; 54,22); cf. καὶ γράψει ἑαυτῷ τὸ δευτερονόμιον τοῦτο εἰς βιβλίον (Deut 17:18) and κατάγραψον τοῦτο […] ἐν βιβλίῳ (Exod 17:14), προσπέπτωκεν οὖν ἡμῖν τὸν ἄ�θρωπ�� [ἀ]�ῆχθαι αὐτὸν ἐν τ\ῆι/ (εις την written above the line) ἐν Κροκοδίλων πόλει [φ]υλακῆι (Sammelb. iii 7202.10-12); ἀπάγεται εἰς φυλακὴν (PSI IV 406.23); καὶ ἀπέθεντο αὐτὸν εἰς φυλακὴν (Lev 24:12); καὶ ἔθετο αὐτοὺς ἐν φυλακῇ (Gen 40:3); πλοῖον παπύρινον προσορμῆσαι εἰς Μέμφιν (UPZ 81 ii 6); προσορμήσαντες τῇ Πεπαρήθῳ (Plb 10.42.1); καὶ εὗρεν αὐτοὺς ἐν Δωθαϊμ (Gen 37:17). 58 In 52. Mayser, Grammatik, II:2, 357-358. 53. J. Humbert, La disparition du datif en grec (du Ier au Xe siècle) (dissertation, université de Paris, Collection linguistique 33), Paris 1930, 103-116. 54. I. Soisalon-Soininen, “Die Wiedergabe des ‫ ב‬instrumenti im griechischen Pentateuch” in: Soisalon-Soininen, Studien, 116-130; K. Hauspie, “Ev with dative indicating instrument in the Septuagint of Ezekiel” in: M. K. H. Peters, XII Congress IOSCS, 201-224. 55. Mayser, Grammatik, II:2, 371-373. 56. I. Soisalon-Soininen, “ἐν für εἰς in der Septuaginta” in: Soisalon-Soininen, Studien, 131140. 57. The difference between direction and location is a conceptual one. Examples of this in other languages include the English expression “to arrive at a place” and its Finnish equivalent “saapua paikkaan” (literally “to arrive to a place”). In English, the event is seen from the perspective of its terminus: the action described by the verb is completed. In Finnish, on the other hand, the case form denotes direction, thus the process of moving toward the destination, not its end, is highlighted. 58. M. Johannessohn, Der Gebrauch der Präpositionen in der Septuaginta (Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen. Philologisch-Historische Klasse. Beiheft), Berlin 1925, 331. 7. Reduction of the Case System and Increase in the Use of Prepositions

117

6. Septuagint Syntax and Hellenistic Greek

these sentences, the verb denotes either direction/motion toward an object or the result of such a movement. The preposition ἐν serves to underscore the end point of the movement. Such is the case in Sammelb. iii 7202.10-12, where ἐν indicates that the person is in prison (the governing verb προσπέπτωκεν being a perfect that denotes a resultant state). The preposition εἰς, on the other hand, denotes a process where “the person is being put to the prison.” 59 It is not always easy to determine whether the writer intended directionality or location. It is possible that the confusion between εἰς and ἐν, which finally resulted in the disappearance of ἐν in the early centuries CE, originated in the semantic “overlap” between motion/direction and location.

8. Conclusions Jerker Blomqvist distinguishes three strata of Hellenistic Greek prose, “one of which was used in literary productions, the other in books and documents written for practical purposes, and the third in the everyday conversations of unlearned people”. 60 The Koine of the Septuagint is closest to the second type. The translators had no great literary ambitions, as they were primarily concerned with intelligibility and faithfulness to the Hebrew original. Nevertheless, they did demonstrate at times that they were familiar with Greek literary style. A method of analysis that documents all relevant cases of an individual linguistic phenomenon in a given corpus and then compares them with their source text equivalents, translation technical study is able to provide a more refined picture of the translators’ linguistic usage than was previously possible by random analysis of isolated instances. This method is therefore to be preferred, as we noted in conjunction with word order and καί-style studies. The general syntactic make-up of the Septuagint appears to be that of the contemporary Koine, with the exception of certain non-Greek features induced by the practice of literal translation. Nevertheless, at least in the earliest translated books, linguistic usages characteristic of the classical period subsist. In this survey, we have reviewed several previously studied features of the Koine of the Septuagint but much research still needs to be done, particularly on the non-Pentateuchal books. Questions of highest priority include the hypothetical merger of the middle and passive voices and the translation of word order. Without such studies, we will not be able to create a muchneeded syntax of Septuagint Greek.

59. See Luraghi, On the Meaning, 83, 87 and her examples. 60. J. Blomqvist, Greek Particles in Hellenistic Prose, dissertation, Lund University, 1969, 20.

118

8. Conclusions

7. Literarische Anspielungen in der LXX Knut Usener

Die Tatsache, dass die Texte der LXX, insofern diese Übersetzungen des hebräischen AT sind, ihrem Prätext eng verpflichtet sind, zeigt sich nicht nur in ihrer thematischsachlichen, sondern auch oft in ihrer sprachlich-stilistischen Gestaltung. Doch die Beobachtung, dass es dennoch inhaltliche Abweichungen, Verkürzungen und Erweiterungen gegenüber dem verbindlichen Prätext gibt, wirft Fragen auf, die nicht nur die Technik des Übersetzens, sondern auch den für die Übersetzung herangezogenen hebräischen Text als schriftliches Dokument betreffen. Es hat sich herausgestellt, dass der unseren modernen Texteditionen des hebräischen AT zugrundeliegende Masoretische Text (MT) bisweilen nicht als Vorlage der LXX zu betrachten ist: Vielmehr hat die LXX den Einblick auch in eine Textausgabe ermöglicht, die dem MT vorausgegangen ist und heute durch die LXX rekonstruiert werden kann. Sodann weist die LXX Texte auf, die von vornherein auf Griechisch verfasst sind: Zu nennen sind hier Esdras I, Judith, 2–4Makkabäer, Weisheit Salomons, Baruch, Brief des Jeremia, Susanna sowie Bel und Drache. Vergleicht man die Übersetzungstexte mit diesen Koine-sprachlichen Originalen, die keine Vorlage hatten, fällt auf, dass Spachduktus und Stil etwas ›griechischer‹ sind als die im Zusammenhang mit der Übersetzung entstandenen Texte. Doch auch bei den ›Übersetzungstexten‹ stößt man auf sprachlich-literarische Phänomene, die eine Differenz zwischen der Vorlage und der griechischen Textur deutlich erkennen lassen. Dieses Phänomen soll hier näher betrachtet werden. Dabei ist zunächst zu definieren, was dieses Phänomen ausmacht und was andererseits nicht in diesen Bereich gehört. Wenn der LXX-Text eine Formulierung aufweist, die offensichtlich vom hebräischen Text abweicht, soweit dieser für uns mit seinen Varianten greifbar ist, und wenn diese Abweichung zugleich auf eine Formulierung in der griechischen Literatur außerhalb der LXX zurückgreift, einen anderen griechischen Text zitiert oder eine motivische bzw. literarische Zusatzinformation aus der griechischen Literatur bietet, dann haben wir es mit diesem Phänomen zu tun, das als Anspielung (allusion) auf oder als Zitat aus der griechischen Literatur zu bezeichnen ist. Von diesen durch die griechische Literatur beeinflussten Phänomenen zu unterscheiden sind Formulierungen, die der griechischen Kultur- und Lebenswelt entstammen, aber kaum durch die Vermittlung der Literatur Eingang in den LXX-Text gefunden haben. Zu nennen ist hier etwa Hiob 42,11, wo der einzige LXX-Beleg für τετράδραχμον (»Tetradrachme«) steht: Das Septuaginta-Hapax steht für hebr. ‫ְקִשׂיָטה‬, das die LXX sonst mit einem Begriff für »Lamm« übersetzt. 1

1.

Gen 33,19; Jos 24,32; Hiob 42,11. Jos 24,32LXX hat ἀμνάδων ἑκατόν »für 100 Lämmer«, ähnlich zuvor Gen 33,19LXX (ἑκατὸν ἀμνῶν). 8. Conclusions

119

7. Literarische Anspielungen in der LXX

Ein inzwischen bekanntes Beispiel soll veranschaulichen, welche Phänomene zu untersuchen sind: Ez 27,5 »… Bretter für Schiffsplanken aus Zypressenholz wurden vom Libanon geholt, um dir Mastbäume aus Tannen zu fertigen.« Die Formulierung »aus Tannen« (ἐλατίνους) stellt gegenüber dem hebräischen Text einen sachlich zunächst irritierenden Zusatz dar, der jedoch vor dem Hintergrund einer literarischen Anregung verständlicher wird – wenn man nämlich hinter der gesamten Wortverbindung »Mastbäume aus Tannen« eine Wortprägung erkennt, die aus der Odyssee stammt und die später als Synonym für »Mastbaum« verstanden werden konnte, so dass die Materialangabe gleichsam ungehört in den Hintergrund trat und nicht mehr als Widerspruch empfunden wurde. 2 Der Grund für die Verwendung dieser nur bei Homer an zwei Stellen (Odyssee 2,424 und 15,289) belegten Wortverbindung in der LXX bleibt funktional und literarisch zunächst unklar: Der Rückgriff auf eine homerische Wendung kann allgemein u. a. dem Zweck der literarischen Erhöhung oder Episierung dienen, an eine im Quelltext vergleichbare, jeweils im Aspekt erst zu bestimmende ähnliche oder konträre Situation anknüpfen oder durch das Zitat einen Motivhintergrund anklingen lassen, um so im LXX-Text eine zweite Verstehens-Ebene zu generieren. Der Rückgriff auf eine auffällige Wortfügung kann dadurch zustande kommen oder begünstigt werden, dass der Quelltext gelesen, gehört oder bereits in der Schule behandelt wurde. In den wenigsten Fällen wird ein Autor »nachgeschlagen« und »nachgelesen« haben – Zitate oder Anspielungen erfolgten meist aus dem Gedächtnis. Im speziellen Entstehungskontext der LXX ist sogar zu fragen, ob nicht in Alexandria insbesondere Kontakte zwischen den Übersetzern der LXX und dem einen oder anderen Philologen, der am Museion antike Texte sammelte, edierte und kommentierte, anzunehmen sind: 3 Markante Formulierungen konnten auch auf dem Weg des mündlichen Diskurses in andere Texte und Kontexte gelangen. Antike Erklärungen zu den beiden hier genannten Versen der Odyssee lagen jedenfalls vor und sind in den heutigen Scholien nachlesbar. 4

2.

3.

4.

Vgl. hierzu K. Usener, »Griechisches im Griechisch der LXX« in: E. Bons / T. J. Kraus (Hg.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint, Göttingen 2011, 87-90; K. Usener, »Zur Sprache der Septuaginta« in: M. Karrer / M. Kraus (Hg.), Septuaginta Deutsch. Erläuterungen und Kommentare zum griechischen Alten Testament, Band I, Stuttgart 2011, 50; A. Hammerstaedt-Löhr / K. Usener, »Jezekiel« in: M. Karrer / M. Kraus (Hg.), Septuaginta Deutsch. Erläuterungen und Kommentare, Band I, 2934. In der Vulgata (cedrum de Libano tulerunt, ut facerent tibi malum) wird keine weitere Holzart genannt – malus, i, m. kann ›Mastbaum‹ bedeuten, als fem. Substaniv (malus, i, f.) aber auch ›Apfel-‹ (oder auch allgemein ›Obstbaum‹). Hier ist jedoch kontextbedingt der Mastbaum gemeint. Damit schließt sich die Vulgata klar dem MT an. Aus anderer Perspektive kommt auch S. Honigman zu einem ähnlichen Ergebnis: S. Honigman, The Septuagint and Homeric Scholarship in Alexandria. A Study in the Narrative of the Letter of Aristeas, London/New York 2003. Die Scholien richten das Augenmerk allerdings nicht auf den in der LXX verwendeten Wortlaut. Zu den Scholien vgl. die Ausgabe von W. Dindorf, Scholia Graeca in Homeri Odysseam. Ex codicibus aucta et emendata, Oxford 1855, Bd. 1, 117 und Bd. 2, 614. Die moderne Ausgabe von N. Ernst, Die D-Scholien zur Odyssee. Kritische Ausgabe, Diss. Köln 2004, 55, im Internet unter http://kups.ub.uni-koeln.de/1831/, verzeichnet die Passage im 15. Buch der Odyssee nicht, da diese nicht unter den D-Scholien überliefert ist.

120

8. Conclusions

7. Literarische Anspielungen in der LXX

Poetisches, insbesondere aber auch lyrisches Vokabular begegnet in der LXX an zahlreichen Stellen. Das bedeutet nicht, dass damit jeder Prosa-Text der LXX, der entsprechendes Vokabular integriert, zugleich und unbedingt ein lyrisches Kolorit erhält. Doch kann dies durchaus eine möglicherweise vom Übersetzer intendierte Begleiterscheinung sein. Interessant sind insbesondere Passagen, in denen auffällig seltenes Vokabular verwendet wird. Dies ist etwa bei ἅγνισμα der Fall: τὸ ἅγνισμα (»Reinigung«, »Reinigungsmittel«) findet seinen ersten Beleg in einem Chorlied bei Aischylos (Eumeniden 326), dann erst wieder in den Scholien zur Stelle und bei Autoren, die deutlich nach der Entstehung der LXX gelebt haben. In der LXX verwendet nur Num 19,9 diesen Begriff als Entsprechung für ‫ַחָטּאת‬. Im Griechischen finden sich ansonsten nur gleichbedeutende Ableitungen wie etwa ὁ ἁγνισμός: Die Aischylos-Passage und die Stelle in der LXX sind somit durch die Gemeinsamkeit des Spezialvokabulars verknüpft. Ferner verbinden Opferungen im Zusammenhang mit dem Totenkult als eine Handlungsund Motiv-Gemeinsamkeit beide Texte: 5 Num 19,2 ff. nennt als Opfer hierbei δάμαλις (»junge Kuh«), ein Wort, das nahezu ausschließlich in lyrischen Texten verwendet wird. 6 Des Weiteren findet sich in Num 22,28 mit der Formulierung πέπαικάς με (»du hast mich geschlagen«, von παίω »schlagen«) ein Verbum, das zwar nicht bei Homer, wohl aber bei den Dramen-Dichtern vorkommt, und das nur sehr selten in Prosa Verwendung findet. Es stellt sich im Zusammenhang mit diesen Beobachtungen die Frage, ob poetisches Vokabular auch in anderen Büchern der LXX zu finden ist. In der Tat gibt es viele Belege in der LXX dafür, dass einem Prosa-Text durch die Verwendung poetischen Vokabulars eine Textoberfläche verliehen wird, die von einem griechisch-sprachigen Publikum zunächst nicht erwartet wird. Kaum wird man versucht sein, mit den jeweiligen Worten zugleich auch einen literarischen Handlungs- oder Motiv-Zusammenhang herstellen zu wollen: Der Text erfährt durch die besondere Wortwahl gleichsam eine Rhetorisierung im Sinne der Erhabenheit und des Feierlichen. Eine darüber hinausweisende inhaltliche Referenz ist im Einzelfall zwar zu prüfen, wird aber in den meisten Fällen die Ausnahme sein. Beispiele für die Verwendung poetischen Vokabulars, das in Prosa auffallend selten belegt ist, sind: 1) Verben: � ἐρυθαίνω »rot färben« (anstelle von ἐρυθραίνω, ἐρεύθω) wird in der LXX nur einmal verwendet (SapSal 13,14); seit Homer, Ilias 10,484 wird das Verbum auch im Hellenismus bei Apollonius Rhodius (1,791 etc.) verwendet; späte Prosa (Arrian, 2. Jh. n. Chr.) greift dann wieder auf den Ausdruck zurück. 5.

6.

Allerdings ist es bei Aischylos der von den Rachegeistern verfolgte Muttermörder Orestes, der als »Opfer« bezeichnet wird (Aisylos, Eumeniden 304 f.), in Num 19,2 ff. hingegen wird eine junge Kuh (δάμαλις) erwähnt, die im Kultzusammenhang für Verstorbene geopfert werden soll. Zur Opferterminologie vgl. auch M. Vahrenhorst, »Exkurs: Hinweise zur Opferterminologie im LXX-Pentateuch« in: M. Karrer / M. Kraus (Hg.), Septuaginta Deutsch. Erläuterungen, Band 1, 335-346. 8. Conclusions

121

7. Literarische Anspielungen in der LXX

� ὠρύομαι ist ein lautmalerisches Verb: Es bezeichnet wie das auch in der LXX häufig belegte, gewöhnliche Wort βοάω das »brüllen«, meist von Löwen, in der griechischen Literatur ab und an auch von Wölfen und Hunden. In der LXX wird das Verb meist auf Löwen angewendet. Ferner kann auch das Schreien von Menschen so bezeichnet werden, so bei Herodot 3,117 und Ps 37,9; EpJer 31 (von den Priestern). Insgesamt wird das Verbum in der LXX an zwölf Stellen verwendet. 7 2) Substantive: � γαῖα ist ein poetisches Wort für γῆ »Erde«: Die LXX verwendet es an sieben Stellen. 8 � προσανάβασις ist der »Aufstieg«: In der LXX kennt nur der Übersetzer von Josua 15,3 dieses Doppelkompositum. Das Wort kommt bei Aischylos und Euripides vor und wird sonst kaum verwendet – das bei Weitem häufigere Wort ist auch in der LXX ἀνάβασις. 3) Adjektive: � Hier ist insbesondere auf δυσκλεής »unrühmlich«, »ruhmlos, ohne Ruhm« hinzuweisen. Das Wort begegnet in der LXX nur in 3Makk 3,23.25. Das Adjektiv wird von Homer (Ilias 9,22 etc.) und späteren Dichtern verwendet und findet kaum Eingang in die Prosa (vgl. Xenophon, Kyrupädie 3,3,53, und spätere Autoren). Häufiger wird etwa ἀκλεής verwendet (auch dies in der LXX nur bei 3Makk 4,12).

Ausgehend von dem ersten Beispiel, das einen Homerismus zum Gegenstand hatte, ist somit zu erkennen, dass zahlreiche weitere Begriffe aus der griechischen Dichtung (Lyrik, Drama) Eingang in verschiedene Texte der LXX gefunden haben. Zweifellos hätten auch die in der Prosa üblichen Worte Verwendung finden können: Wenn die LXX-Übersetzer also poetisches Vokabular einflechten, so ist zu fragen, was sie dazu bewogen haben könnte. Am wahrscheinlichsten ist, dass die Rhetorisierung im Sinne der Erhabenheit und des Feierlichen ihr Hauptanliegen war. Eine derartige bewusste Verwendung ausgesuchter Begriffe allerdings offenbart eine hohe sprachliche Kompetenz. Poetologisch aufgeladene Formulierungen wie etwa Neh 5,14 f. (2Esr 15,14) zeigen, dass mancher Übersetzer durch das Abweichen von der strengen Wortübersetzung eine literarische Auf- oder Umwertung vornimmt: So wird aus dem »prosaischen« Statthalter die βία, die »Kraft« der herrschenden Person – eine Formulierung, die sich in Anlehnung an episch-homerische Formulierung liest: Statt Herakles heißt es bei Homer βίη Ἡρακληείη – »Kraft des Herakles«; 9 auch Frauennamen können – wie in Odyssee 11,296 – mit βία (episch-ionisch dann βίη) umschrieben werden. 10 Hiob 2,13 verwendet das in der LXX 98mal belegte Wort πληγή – allerdings mit einer Besonderheit: Das von Gott geschickte schwere Unglück wird mit einem Rückverweis in Hiob 42,16 erneut durch den gleichen Ausdruck wiedergegeben. Eine solche zugleich rahmende Verweis-Technik hat wohl der Übersetzer des Hiob-Buches eingefügt. 11 Das Wort weist überdies poetische Qualitäten auf – insbesondere Aischylos (Agamemnon 367) und Sophokles (Aias 137) sind hier zu nennen. 7. Ri 14,5; Ps 21,14; 37,9; 103,21; SapSal 17,18; Hos 11,10; Zef 3,3; Sach 11,3; Jer 2,15; EpJer 6,31; Hes 22,25. 8. 4Kgt 18,35; Esr 3,3; 9,1.2.14; Ps 48,12; Hes 36,24. 9. Ilias 2,658; 11,690; Odyssee 11,601. 10. Weitere Beispiele bei Usener, »Griechisches im Griechisch der LXX«, 94 f. 11. So die Erläuterungen z.St., S. 2073 und 2125.

122

8. Conclusions

7. Literarische Anspielungen in der LXX

Überwiegend mit der Bedeutung »traurig« erhält στυγνός eine besondere Bedeutungsrichtung in der LXX: In dieser semantischen Richtung ist das Wort poetisch aufgeladen und wird besonders von den Tragikern verwendet (Aischylos, Prometheus 886 etc.). Die Weiterbildung zu μαιμάω (seit Homer, Ilias 5,670 etc.) durch μαιμάσσω »losstürmen« in Jer 4,19 und Hiob 38,8 ist insofern auffällig, als dass dieses Verbum vor der Septuaginta nur von Homer und (tragischen) Dichtern, nicht aber in Prosa verwendet wird. Im Hellenismus findet es sich im 2. Jh. v. Chr. wieder bei dem Arzt und Dichter Nikandros von Kolophon (in der Form μαιμώσσω, Theriaka 470). Die Spuren der Wortverwendung lassen sich noch weiter verfolgen. So gibt es neben dem Bestreben, seltenes und in literarisch hochstehenden Texten beheimatetes Vokabular zu verwenden, Beispiele mit anderen Funktionen: Der Jesaia-Übersetzer spricht in Jes 3,22 von διαφανῆ Λακωνικά, von »durchscheinenden lakonischen Kleidern«. Die Wiedergabe von ‫ַהַמֲּחָלצֹות ְוַהַמֲּעָטפֹות ְוַהִמְּטָפּחֹות‬ ‫ ְוָהֲח ִריִטים‬durch καὶ τὰ ἐπιβλήματα τὰ κατὰ τὴν οἰκίαν καὶ τὰ διαφανῆ Λακωνικά (»und die leichten Hauskleider und die durchscheinenden lakonischen Kleider«) ist zunächst befremdlich – von »lakonisch« oder »spartanisch« bzw. »aus Sparta importiert« ist im MT keine Rede. Was auch immer der Übersetzer vorgefunden hat – er hat seine eigene Vorstellung in den Text getragen: Durch die Setzung des Zugehörigkeits-Adjektivs »lakonisch« erhält die bereits ausgesprochene Qualitätsbezeichnung des Kleidungsstoffes (διαφανῆ) eine weitere, neue Qualität: Er ist nicht nur nahezu »durchsichtig«, sondern auch noch »spartanisch«: Spartanisch allerdings ist der Stoff nicht in dem Sinne, wie dies etwa die in Sparta verbreiteten Filzschuhe waren (dem widerspricht διαφανής), 12 sondern er ist es gleichsam in moralisch-wertender Hinsicht. Im Hintergrund dieser Formulierung schwingt wahrscheinlich eher die ablehnende Haltung der in Sparta geübten Polyandrie mit. 13 Sions stolze Töchter werden somit ihrer verwerflichen Pracht beraubt. Einem Text können somit auch durch Referenzen auf ethnisch-kulturelle Phänomene wertende Aussagen unterlegt werden, die vom Prätext nicht vorgegeben waren. Kulturexterne Anspielungen oder Referenzen sind nicht auf dieses Beispiel beschränkt: Auch Anspielungen auf Vorstellungen des griechischen Mythos etwa finden sich in der LXX, ohne dass der hebräische Text dies vorgibt. In Hiob 42,14 wird der Name einer Tochter von Hiob mit Ἀμαλθείας κέρας, »Füllhorn der Amaltheia« angegeben. Der MT schreibt ‫» ֶק ֶרן ַהפּוְּך‬Keren Happuch«, und damit entspricht die LXX also nicht der Vorlage. Doch hier kann man sehr klar erkennen, wie der Übersetzer sein Anliegen einer textgetreuen Übertragung zuwege gebracht hat: Die Assonanz von κέρας und ‫ ֶק ֶרן‬ist gut hörbar und somit als »Wegweiser« für die Übersetzung gut vorstellbar. Dabei lässt der Übersetzer dann sein Wissen einfließen: Es gibt für ihn nur eine weibliche Person, die er mit dem vorgegebenen Klang in Verbindung bringen konnte – Amaltheia. Diese junge Frau – oder auch bei manchen Dichtern eine junge Ziege – hatte den griechischen Göttervater Zeus mit ihrer Milch großgezogen und wurde zum Dank von ihm an den Sternenhimmel versetzt, als »Füllhorn«. Dass diese Sage und ihre allegorische Auslegung in Alexandria 12. Aristophanes, Vespae 1157 f. 13. Hierzu vgl. Usener, »Griechisches im Griechisch der LXX«, 94. 8. Conclusions

123

7. Literarische Anspielungen in der LXX

bekannt war, steh außer Frage: Unter dem Namen des im ptolemäischen Alexandria tätigen Wissenschaftlers und Dichters Eratosthenes hat die alte Sage eine neue Popularität erhalten – der Hiob-Übersetzer könnte davon erfahren haben. 14 Eine merkwürdige Gruppe von Fabel- und Mischwesen soll im Zusammenhang der Anspielung auf griechische Mythologie noch erwähnt werden – die Sirenen. Weder die Zahl noch das Aussehen dieser Fabelwesen war bei den Griechen einheitlich – insofern kann es nicht verwundern, wenn Bezugnahmen auf die Sirenen durchaus etwas diffus sind. Bisweilen wurden sie als junge Frauen mit schönen oder auch verderbenbringenden Stimmen gedacht: 15 Hiervon findet sich in der Formulierung »Töchter von Sirenen« ein Reflex bei Jeremia (27,39) und Micha (1,8). In 4Makk 15,21 heißt es: »Nicht melodische Gesänge von Sirenen, und keine Stimmen, die wie Schwanengesang […]« (οὐχ οὕτως σειρήνιοι μελῳδίαι οὐδὲ κύκνειοι […] φωναί […]): Neben einer Platon-Reminiszenz 16 werden hier die durch ihren Gesang berühmten Sirenen genannt. Dass der Übersetzer hier an Homers Odyssee gedacht hat, ist möglich, nicht aber sicher: Die Sage war allgemein bekannt. Die Sirenen wurden aber auch als fabulöse Mischwesen von Mensch und Vogel gedacht – ein Reflex hiervon findet sich in Hiob und Jes: 17 Hiob 30,29 schreibt: ἀδελφὸς γέγονα σειρήνων, ἑταῖρος δὲ στρουθῶν – »ein Bruder der Sirenen bin ich geworden und ein Gefährte der Strauße«. Bei Jes 34,13 heißt es: ἔπαυλις σειρήνων καὶ αὐλὴ στρουθῶν »ein Quartier von Sirenen und ein Hof von Straußen« – jeweils ist von Sirenen und Straußenvögeln die Rede – eine Abwandlung des hebräischen Textes, der wohl von »Straußentöchtern« (‫ )ִלְבנ ֹות ַיֲע ָנה‬spricht, aber nicht von Sirenen, sondern von Schakalen (‫)ַתּן‬. Der Übersetzer hatte also hier weniger den mythologischen Hintergrund im Kopf als vielmehr das Bestreben, eine Entsprechung für »Schakal« zu finden. Warum er statt θώς »Schakal« schließlich die Sirene gewählt hat, ist unklar. Im Zusammenhang mit den Fabelwesen ist auch der ὀνοκένταυρος, der »Eselzentaur« zu nennen, der nur bei Jesaia 13,22 und 34,11.14 vorkommt und der dann erst wieder von Aelian (NA 17.9) erwähnt wird: In der griechischen Mythologie ist der Kentaur bekannt – woher der LXX-Übersetzer seine Kreatur genommen hat, ist unklar: Eine Steigerung des Unwirtlichen möglicherweise – das Fabelwesen wird im unmittelbaren Kontext mit den oben genannten Sirenen angeführt. Das hier vorgelegte Material wäre noch in einigen Aspekten zu erweitern. Wortuntersuchungen sind zunächst die Basis, und es hat sich gezeigt, dass mit Begriffen bisweilen Kontexte verknüpft sind, die dem hebräischen Text eher fern stehen, dem griechischen Assoziationsfeld aber leichter entsprechen. Philosophisches Grundvokabular liegt – neben φιλόσοφος (in Dan 1,20; 4Makk 1,1.5,35.7,7) – mit dem Begriff σοφιστής vor (Ex 7,11; DanLXX 1,20; 2,14.18.24): Allerdings ist der Begriff hier nicht im platonischen Sinne negativ gefärbt, sondern eher im 14. Vgl. zum Material Usener, »Griechisches im Griechisch der LXX«, 96. Auch Hiob 9,9 nennt ein Sternbild: Der Arkturos wird in der LXX nur hier erwähnt. 15. Vgl. hierzu A. Heubeck / A. Hoekstra, A Commentary on Homer’s Odyssey, vol. II, Books IXXVI, Oxford 1989, 118 ff. 16. Zum Schwanengesang vgl. Platon, resp. 10, 620a6 f. 4Makk 15,21 bietet eine Anspielung, nicht aber ein Zitat. 17. Hiob 30,29; Jes 13,21; 34,13; 20.

124

8. Conclusions

7. Literarische Anspielungen in der LXX

vorplatonischen Sinne positiv verstanden als »Weiser«. Dass jedoch nicht σοφός, sondern σοφιστής verwendet wird, ist bemerkenswert: Herodot 2,49 oder auch Pindar (Isthm. 5[4],28) etwa verwenden den Begriff ebenso positiv. 18 In den Bereich der Verwendung philosophisch-wissenschaftlicher Begriffe und Fachtermini gehören Passagen wie 4Makk 14,14.18 und SapSal 11,15: Der Begriff ἄλογα nimmt eine im Hellenismus (Stoa, Aristoteles) geführte Diskussion auf, in der die Frage nach der Vernunftbegabtheit der Tiere kontrovers diskutiert wurde. 19 Dass in der LXX immerhin die oben genannten Passagen sich mit den ἄλογα ζῷα, mit der »unvernünftige Kreatur« oder mit den »nicht vernunftbegabten Tieren« befassen, zeigt, dass der entsprechende wissenschaftliche Diskurs manchem Verfasser von LXX-Texten vertraut war, auch wenn mit ἄλογος in der LXX sonst eher der Mangel an Beredsamkeit fokussiert wird. 20 Etwas spezifischer wird es allerdings bei der Verwendung von ἀόρατος »unsichtbar« in Gen 1,2: ἀόρατος ist dabei durchaus keine äquivoke Übersetzung von ‫ת ֹהוּ ָוב ֹהוּ‬, sondern möglicherweise vom platonischen Schöpfungsmythos (bes. Tim. 36e) beeinflusst, wo die dem Körperlichen zugrundeliegende und in der Schöpfung vorausgehende Seele als »unsichtbar« qualifiziert wird. Ob und wie sich platonischer Einfluss, der hier sehr speziell wäre, einstellen konnte, ist eine interessante Frage, die noch nicht geklärt ist. Ob mit ἀνδρόγυνος ein männlich-weibliches Mischwesen im Sinne des Schöpfungsmythos bei Platon (Symp. 189e) vorliegt, oder eher die herabsetzende Bezeichnung eines verweichlichten Menschen (vgl. Herodot 4,67 etc.), mag unklar bleiben: Prov 18,8 und 19,15 sind jedenfalls die einzigen Passagen der LXX, in denen der Begriff begegnet: Der Zusammenhang mit dem Schöpfungsbericht in Gen 2,21 könnte eine Assoziation herstellen. 21 Überblickt man das hier vorgelegte (und noch erweiterbare) Material, so liegt die Frage nach der Verteilung in der LXX nahe. Dabei fällt insbesondere eine Sonderstellung von Ijob auf: Hier begegnen in hoher Dichte Begriffe, die ein besonderes poetisches Potenzial entfalten. Eine weitere Sonderstellung kommt denjenigen Schriften zu, die von vornherein auf Griechisch entstanden sind: Sie waren nicht an einen Prätext gebunden, tragen daher auch Züge derjenigen Kultur-Umgebung, in der diese Texte entstanden sind. Gleichwohl sind auch hier sprachliche Beobachtungen zu machen, die diese »NichtÜbersetzungen« mit den Übersetzungen als verwandt erscheinen lassen. 18. Weitere Grundbegriffe wie γυμνάσιον (1Makk 1,14; 2Makk 4,9.12; 4Makk 4,20), ἔφηβος (2Makk 4,12) und παλαίστρα (2Makk 4,14) finden sich vereinzelt in der LXX. Dass hierbei keine literarischen Quellen benutzt wurden, sondern die hellenistische Umwelt den Begriff zur Verfügung gestellt hat, ist wohl deutlich. Zu ἁρμονία bei Hes 23,42 und 37,7 vgl. Usener, »Griechisches«, 93. Zu erwähnen ist hierbei noch, dass hier eine Klangnachahmung bei der Übersetzung anzunehmen ist: τὰ ὀστᾶ ἑκάτερον πρὸς τὴν ἁρμονίαν αὐτοῦ – lässt ‫ֲעָצמו ֹת‬ ֹ‫ ֶעֶצם ֶאל־ַעְצמו‬anklingen. 19. Vgl. hierzu u. a. M. Giebel, Tiere der Antike. Von Fabelwesen, Opfertieren und treuen Begleitern, Darmstadt 2003, 199 f. 20. Belege finden sich in Ex 6,12; Num 6,12; 3Makk 5,40; 6,25; 4Makk 14,14.18; SapSal 11,15; PsSal 16,10. 21. Vgl. Erläuterungen 1978 zu Prov 18,1 und S. 1979 f. zu Prov 19,15. 8. Conclusions

125

7. Literarische Anspielungen in der LXX

Das jeweilige Übersetzer-Profil ist in der LXX unterschiedlich: Es gibt keine gleichmäßige Verteilung des Befundes über die gesamte LXX hin. Um eine Aussage über Einzelprobleme wie etwa die Einheitlichkeit und Verfasserfrage bei Hes etc. treffen zu können, müssen weitere Beobachtungen angestellt und ausgewertet werden: Die Fragen nach der Aufnahme griechischer Literatursprache und nach einzelnen motivischen oder handlungsbezogenen Phänomenen kann sicherlich dabei helfen, das Gesamtbild der Entstehungsprozesse der LXX besser zu verstehen.

126

8. Conclusions

III The Greek of the Septuagint as translation Greek

8. Limitations of Greek in Representing Hebrew 1 Takamitsu Muraoka

When considering limitations inherent in Greek with which the Septuagint translators had to cope as they attempted to translate the Hebrew (and Aramaic) 2 texts of the Jewish Bible and some documents from the intertestamental period, one should not assume a priori that one knows everything about the linguistic structure of Hebrew and is in a position to evaluate and assess the performance, failure, or success of the translators. On the contrary, the Septuagint 3 still remains for us a valuable source of information as we strive to better understand Biblical Hebrew. Limitations and incompatibilities across languages are commonplace. This is true even between genetically related languages, Hebrew and Aramaic, for instance, and is all the more the case with genetically unrelated languages such as Hebrew and Greek. The phenomenon can be observed in all parts of language structure: script, phonology, morphology, morphosyntax, syntax, lexicon, style.

1. Script Although no archive has survived in which Septuagint translators, on completion of their work, might have deposited the manuscripts of the Hebrew Bible they used, our knowledge of Hebrew epigraphy for the late Hellenistic and early Roman periods suggests that Septuagint translators must have struggled with pairs of similar-looking letpffi ters, such as ‫ר‬, ‫ד‬, ‫י‬, and ‫ו‬. Thus ‫ אֵַחר‬εἷς (= ‫( )ֶאָחד‬Gen 43:14); ‫ ַח ָוּה‬ζωή (= ‫( )חיי‬Gen 3:20); ‫ ישׁוב‬ἀποστρέψει (Ps 54:7). 4

1.

2.

3.

4.

No attempt will be made here to draw up an exhaustive inventory of all the types of limitations to which Septuagint translators might be said to have been subject, but only some typical categories will be dealt with. Lexicographical issues have been deliberately excluded, as they cover too wide an area to receive an adequate treatment within the scope of this chapter. For the sake of simplicity, we will speak hereafter only of Hebrew. Furthermore, by “Hebrew,” we mean the putative Hebrew language that may be assumed to have prevailed and to have been perceived as such at the time when the Septuagint came into existence. This Hebrew is, therefore, not to be equated with the form of Hebrew that is reflected in the Tiberian tradition. A measure of circularity is unavoidable here, because, as stated above, our knowledge and understanding of Hebrew prior to the emergence of the Tiberian tradition is derived to some extent from what we find in the Septuagint. The “Septuagint” is to some extent a theoretical construct, since the initial translation is not extant in its entirety. Indeed, the initial translation is known to have gone through various stages and forms of revision. Even within the Masoretic tradition there is a ktiv ‫ָיִשׁיב‬ 1. Script

129

8. Limitations of Greek in Representing Hebrew

1.1 Consonants Hebrew had six “guttural” consonant phonemes marked with four graphemes: /ʾ / = h‫א‬i; /h/ = h‫ה‬i; /ʿ / and /ġ/ = h‫ע‬i; /ḥ/ and /ḫ/ = h‫ח‬i. By contrast, Greek has merely one such sound, /h/, which apparently was not considered important enough to merit special graphic representation in the early stages of the development of the Greek alphabet and was only later marked furtively by means of a rough breathing, and that only at the beginning of words. 5 Although its presence in the Septuagint period is confirmed by consistent spellings such as καθ᾽ ἡμέραν “daily,” 6 it is not absolutely certain that a reader of the Septuagint ignorant of the Hebrew name of Adam and Eve’s second child would have pronounced /habel/ (‫)ֶהֶבל‬, as directed by the editor of the critical edition of Septuagint Genesis who transcribed it Ἅβελ. It is fair to say then that in the Septuagint period there was no orthographic means to distinguish between the transcription of this personal name and that of ‫ ֶאֶבל‬or ‫אֶָבל‬, if such a person existed. The fricative velar /ḫ/ is represented by χ, while the fricative pharyngeal /ḥ/ remains graphically unmarked, cf. ‫ ֲאִחיֶמֶלְך‬Αχιμελεχ (1Sam 30:7) versus ‫ ַח ָוּה‬Ευα (Gen 4:1). Whereas the fricative pharyngeal /ʿ / has no corresponding Greek letter, the fricative velar /ġ/ is sometimes transcribed γ, e. g. ‫ ִשְׁמעו ֹן‬Συμεων (Gen 29:33) versus ‫ַע ָזּה‬ Γαζα (Gen 10:19). 7 Hence our search for an etymology of the name of Sodom’s twincity must be guided by this knowledge: ‫ ֲעמ ֹ ָרה‬Γομορρα (Gen 10:19). 8 Greek is incapable of distinguishing the “emphatic” consonant ‫ צ‬from three other sibilants, namely ‫ שׂ‬,‫ס‬, and ‫שׁ‬. All four phonemes are represented by σ: Σεβωιμ ‫ְצבֹ ִים‬ (Gen 10:19); Σαρα ‫( ָשׂ ַרי‬Gen 11:29); Σοδομα ‫( ְסד ֹם‬Gen 13:10); Σημ ‫( ֵשׁם‬Gen 6:1). 9 On the other hand, ‫ ט‬is usually distinguished from ‫ת‬: Αχιτωβ ‫( ֲאִחיטוּב‬1Sam 14:3) versus

5. 6.

7.

8.

9.

On the phonetic value of this “glottal fricative,” see W. S. Allen, Vox Graeca: The Pronunciation of Classical Greek (second edition), London/New York, NY 1974, 50-53. Not infrequent deviations in contemporary papyri such as κατ᾽ ἕκαστον and ἀπέστησα for καθ᾽ ἕκαστον and ἀφέστησα respectively may indicate, however, that Septuagint Greek might not be so much a specimen of the contemporary langue as a testimony to the educational standard of the Septuagint translators and copyists. See E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften (second edition), vol. I, Laut- und Wortlehre, Part 1, Einleitung und Lautlehre, Berlin/Leipzig 1970, 173-176. Whereas the modern Palestinian Arabic pronunciation has preserved /ġ/, Jericho is pronounced /ḥ/ as opposed to Ιεριχω of the Septuagint. Χανααν for ‫ ְכּ ַנַען‬is problematic in its Hebrew form in view of kinaḫḫi in Old Canaanite. We believe that at the time of the translation of the Septuagint these two pairs of guttural fricatives were distinct phonemes. See P. Joüon / T. Muraoka, A Grammar of Biblical Hebrew (second edition, Subsidia biblica 27), Rome 2006, section 5 k-l. Jan Joosten (“The Septuagint as a source of information on Egyptian Aramaic in the Hellenistic period” in: H. Gzella / M. L. Folmer (eds.), Aramaic in its Historical and Linguistic Setting, Wiesbaden 2008, 93-105, in particular 97-99), however, holds that these two modes of transcription in the Septuagint reflect an Aramaic influence on the translators. Τύρος ‫“ צֹר‬Tyre” (2Sam 5:11) as against Σιδων ‫( ִצידוֹן‬Gen 49:13) is remarkable. The former probably represents /ṭ/ as shown by ‫“ טוּר‬mountain” in Aramaic.

130

1. Script

8. Limitations of Greek in Representing Hebrew

Θαιμαν ‫( ֵתּיָמן‬Gen 36:11). Similarly, ‫ ק‬and ‫ כ‬are differentiated: Ακκαρων ‫( ֶעְקרוֹן‬Josh 13:3) versus Χανααν ‫( ְכּ ַנַען‬Gen 9:18). 10 The use of υ in Ιωυαν for ‫( ָי ָון‬Gen 10:2) and in Ησαυ for ‫( ֵעָשׂו‬Gen 25:30) is an approximate representation of the bi-labial /w/. Similarly, ι is used in Ιακωβ for ‫ַיֲעקֹב‬ to represent /y/. However, the convergence of /y/ and its homogeneous vowel /i/ points out a limitation inherent in Greek: Ισμαηλ ‫( ִיְשָׁמֵעאל‬Gen 16:11). The fact that ‫ב‬, ‫ג‬, and ‫ ד‬are consistently transcribed β, γ, and δ respectively, whereas ‫כ‬, ‫פ‬, and ‫ ת‬are almost always noted χ, φ, and θ should not be taken to mean that the spirantisation process had not yet affected the voiced plosives. 11 The Greek alphabet possessed only one set of letters for the voiced plosives, whereas for the unvoiced set κ, π, and τ were dedicated to the emphatic series. If the allophonemic spirantisation had already begun in Hebrew, 12 this would represent another limitation of Greek.

1.2 Vowels That Origen used η and ω in the second column of his Hexapla, where, according to the diachronic phonology of Hebrew, one would have expected a long e and o, suggests that, for the period under consideration, the quantitative opposition of these vowels in Hebrew can be assumed. 13 Since ου, representing a long u, was used to mark a long u in Hebrew, it follows that Greek was incapable of graphically distinguishing the quantitative opposition in the three vowels a, i, and u, although Greek did have two quantitative varieties of these vowels, long versus short. 14 Phonologically speaking, we believe, that the shva, including the so-called shva mobile, marked the absence of a vowel. 15

10. But note Κίτιοι ‫( ִכִּתּים‬Gen 10:4). 11. As argued by H. Torczyner (“Die Aussprache der Begad-kefat in der Geschichte der hebräischen Sprache” MGWJ 81 [1937], 340-351) and accepted by G. Lisowsky (“Die Transkription der hebräischen Eigennamen des Pentateuch in der Septuaginta,” dissertation, University of Basel, 1940, 121). See also G. Janssens, Studies in Hebrew Historical Linguistics Based on Origen’s Secunda (Orientalia Gandensia IX), Leuven 1982, 45-50. Furthermore, according to Allen, Vox Graeca, 16, the aspirated voiceless series was still pronounced with an aspiration, thus φ in the manner of the word-initial /p/ in English, not like /f/. 12. Cf. Joüon / Muraoka, Hebrew Grammar, 28, note 3. 13. See E. Brønno, Studien über hebräische Morphologie und Vokalismus auf Grundlage der Merkatischen Fragmente der zweiten Kolumne der Hexapla des Origenes (Abhandlungen für die Kunde des Morgenlandes 28) Leipzig 1943, 251, 364; Janssens, Studies, 54-57. 14. On the length of ου in our period, see Allen, Vox Graeca, 73-75, 84-89. On the data in the second column of the Hexapla, see Brønno, Studien, 364-366. 15. See Joüon and Muraoka, Hebrew Grammar, section 8. On the graphic representation of the shva in the Septuagint and in the second column of the Hexapla, see Lisowsky, “Die Transkription,” 128; Brønno, Studien, 318-342; Janssens, Studies, 89-110. 1. Script

131

8. Limitations of Greek in Representing Hebrew

2. Morphology 2.1 Personal Pronouns Hebrew personal pronouns distinguish two genders, masculine and feminine, in the second and third persons, both in the singular and the plural, whereas the corresponding Greek pronouns are gender-neutral in the second person. Hence ‫ אְַתּ‬/ ‫ = אַָתּה‬σύ; ‫ אֵַתּ ָנה‬/ ‫ = אֶַתּם‬ὑμεῖς; ‫ אָֹתְך‬/ ‫ = אְֹתָך‬σε; ‫ ִמֶכּן‬/ ‫ = ִמֶכּם‬ἀφ᾽ ὑμῶν. This equally holds for object pronouns directly attached to verbs, e. g. ‫ ְשָׁלחוְּך‬/ ‫ = ְשָׁלחוָּך‬ἀπέστειλάν σε. 16 Hebrew possessive pronouns are directly and synthetically attached to the end of nouns. By contrast, their Greek equivalents, genitive pronouns, are phonetically and morphologically independent. Consequently, they can be positioned either before or after the noun head, from which they can be separated by other phrase constituents. In Classical and Hellenistic Greek, therefore, two sequences are possible: λόγος μου and μου λόγος, or, with the article, ὁ λόγος μου and μου ὁ λόγος. When there is an attributive to go with the noun, the pronoun is placed between the attributive, which precedes, and the noun: ὁ καλός μου λόγος. This pattern is exemplified in τὸ μεγαλεῖον αὐτοῦ κράτος “his magnificent power” (2Mac 7:17), τὸ σχέτλιον αὐτοῦ βούλημα “his abominable design” (2Mac 15:5), and, in the New Testament, ἐν τῷ καινῷ αὐτοῦ μνημείῳ “in his new tomb” (Mt 27:60). The synthetic structure of Hebrew, combined with the stringent syntactic rule that requires modifiers (with the exception of numerals) to follow the noun they qualify, allows no freedom of movement for genitive pronouns and attributive adjectives. Hence, ‫ ִדָּב ִרי ַהטּוֹב‬is the only possible pattern. Most of the time, Septuagint translators operated under this syntactic constraint. For example, τὴν φωνήν σου “your voice” (Gen 3:10) agrees not only with the linear sequence of the two components of the Hebrew text, ‫ קְֹלָך‬or ‫קֵֹלְך‬, but also with one of the two sequences observable in Hellenistic Greek. A departure from Hebrew syntax, as in ὑμῶν οἱ ὀφθαλμοί ‫“ ֵעי ֵניֶכם‬your eyes” (Gen 3:5), shows that the translators did on occasions exercise a measure of freedom. 17 However, when an attribute is added, the Septuagint often meticulously follows the Hebrew syntax: τὸν υἱὸν αὐτοῦ τὸν πρεσβύτερον ‫“ ְבּנוֹ ַה ָגּד ֹל‬his elder son” (Gen 27:1), in contrast to the two examples adduced earlier from 2Maccabees, a text composed directly in Greek. Likewise ἐν τῇ ἰσχύι σου τῇ μεγάλῃ καὶ ἐν τῷ βραχίονί σου τῷ ὑψηλῷ “with your great might and with your uplifted arm” ‫( ְבּכֲֹחָך ַה ָגּד ֹל וִּב ְזר ֲֹעָך ַה ְנּטוּ ָיה‬Deut 9:29). However, in non-translational Greek, a pronoun can begin an entire phrase, as in μου τὴν φιλάνθρωπον παρηγορίαν “my humane exhortation” (4Mac 5:12) or μου τὸ σεμνὸν γήρως στόμα “my honourable mouth of old age” (4Mac 5:36). When a noun is modified by a demonstrative pronoun as well as by an adjective, the Hebrew pattern is reproduced: τὴν ἔρημον τὴν μεγάλην καὶ τὴν φοβερὰν ἐκείνην “that vast and frightful wilderness” 16. The absence of gender distinction in the third person plural of the Hebrew suffix conjugation seems to have misled the Septuagint translator at ‫ֲאֶשׁר ָיבֹאוּ ְבּ ֵני ָהֱאל ִֹהים ֶאל־ְבּנוֹת ָהאָ ָדם ְו ָיְלדוּ ָלֶהם‬ (Gen 6:4), where the subject of ‫ ָיְלדוּ‬is most likely the women, although the translator took it as masculine, hence ὡς ἂν εἰσεπορεύοντο οἱ υἱοὶ τοῦ θεοῦ πρὸς τὰς θυγατέρας τῶν ἀνθρώπων καὶ ἐγεννῶσαν ἑαυτοῖς, unless his Vorlage read ‫הולידו‬, as in the Samaritan Pentateuch. 17. Cf. αὐτῶν δὲ διηνοίχθησαν οἱ ὀφθαλμοὶ καὶ ἐπέγνωσαν αὐτόν (Lk 24:31).

132

2. Morphology

8. Limitations of Greek in Representing Hebrew

(Deut 2:7; MT: ‫)ַהִמּ ְדָבּר ַה ָגּדֹל ַה ֶזּה‬. In ὁ θυμὸς τῆς ὀργῆς ὁ μέγας οὗτος ‫ֳח ִרי ָהאַף ַה ָגּדוֹל‬ ‫( ַה ֶזּה‬Deut 29:23), the syntactic hierarchy in the Hebrew text is slightly ambiguous: are the adjective and the pronoun to be construed with ‫ ָהאַף‬or with ‫ ?ֳח ִרי ָהאַף‬In the case of ἡ φωνὴ τοῦ ἀλαλάγματος ἡ μεγάλη αὕτη “this loud sound of noise” (1Sam 4:6L), the translator must have had a good reason for departing from MT ‫קוֹל ַהְתּרוָּעה ַה ְגּדוָֹלה‬ ‫ַהֹזּאת‬. Likewise ἐν τῷ βιβλίῳ τοῦ νόμου τούτου for ‫( ְבֵּסֶפר ַהתּוֹ ָרה ַה ֶֽזּה‬Deut 29:19.20; 30:10) as against ‫( ְבֵּסֶפר ַהתּוֹ ָרה ַהֹזּאת‬Deut 28:61), which is identically translated. Cf. also Josh 1:8 and 2Kings 23:21. Hebrew also attaches pronominal objects directly and synthetically to verbs, although, in contrast to the case of attached pronouns synthetically attached to nouns and prepositions, the language also has at its disposal an analytic structure using the particle ‫ֵאת‬. Thus ‫ ַוֲאָב ֲרֵכם‬ἵνα εὐλογήσω αὐτούς “so that I may bless them” (Gen 48:9). When non-emphatic dative or accusative singular pronouns are used in Greek, the traditional accentuation indicates that, as in Hebrew, they form an enclitic cohesive complex in conjunction with the preceding verb, e. g. καὶ ηὐλόγηκά σε ‫( וֵּב ַרְכִתּיָך‬Gen 26:24). Such a phonetic mechanism is confined in Greek to the first and second person singular pronouns, whereas in Hebrew any pronoun can be synthetically attached to the verb. This constitutes a limitation in Greek. Just as in Hebrew such a pronominal morpheme is at times detached and positioned before the verb for the sake of contrast, focus, or topicalization, 18 Septuagint translators resort to this same device, even against the Hebrew text, as in οὐ μή σε ἐγκαταλίπω ‘I shall never ever abandon you’ for ‫ ל ֹא ֶאֱע ָזְבָך‬Gen 28:15, where the message comes over as reassuring on account of the fronted pronoun. Note also ἡμεῖς σε οὐκ ἐβδελυξάμεθα “we haven’t acted towards you abominably” ‫( ל ֹא ְנ ַגֲענוָּך‬Gen 26:29), where the added pronoun, ἡμεῖς, adds a little more force to the transposition in the context of two parties pitched against each other. See also αὐτός σου τηρήσει κεφαλήν […] καὶ αὐτός σου κυριεύσει ‫( הוּא ְישׁוְּפָך רֹאשׁ ]…[ ְוהוּא ִיְמָשׁל־ָבְּך‬Gen 3:15-16), where the separation of σου from κεφαλήν and the juxtaposition of αὐτός σου are to be noted. 19

2.2 The relative pronoun The ubiquitous 20 relative particle ‫ ֲאֶשׁר‬has many different uses. It can stand alone or be combined with a variety of prepositions; it can introduce a clause or be connected with what precedes it. In most cases, the Septuagint translators coped well with these diverse situations. However, occasionally they went about their task a shade too mechanically, producing unidiomatic Greek. For instance, in ἔθνος, ὃ οὐκ ἀκούσῃ τῆς φωνῆς αὐτοῦ “a people whose language you will not comprehend” ‫גּוֹי ֲאֶשׁר ל ֹא־ִתְשַׁמע‬ ֹ‫( ְלשֹׁנו‬Deut 28:49), the pronoun αὐτοῦ, a mechanical reflection of the Hebrew retrospective pronoun, sounds pleonastic. Greek would express the idea by using τῆς φωνῆς οὗ or οὗ φωνῆς, without αὐτοῦ. Equally awkward is τῇ γῇ, ᾗ σὺ παρῴκησας ἐν 18. For instance, ‫“ ְוָה ְרגוּ אִֹתי ְואָֹתְך ְיַחיּוּ‬me they will kill, but you will they spare” (Gen 12:12) (ἀποκτενοῦσίν με, σὲ δὲ περιποιήσονται). 19. Cf. also σὲ κακώσομεν “we are going to give you a good hiding” for ‫( ָנ ַרע ְלָך‬Gen 19:9), where the fronted σὲ puts teeth into the threat. 20. It occurs some 5,574 times in the Hebrew Bible. 2. Morphology

133

8. Limitations of Greek in Representing Hebrew

αὐτῇ ‫“ ְוִעם־ָהאָ ֶרץ ֲאֶשׁר־ ַגּ ְרָתּה ָבּהּ‬with the land, in which you resided” (Gen 21:23) for οὗ (or: ἐν ᾗ) σὺ παρῴκησας. The reverse phenomenon, just as awkward, is observed in a case such as ἀπὸ τῆς ἡμέρας ἧς ἀπῆλθεν ὁ βασιλεύς ἕως τῆς ἡμέρας ἧς αὐτὸς παρεγένετο “from the day when the king left until the day he arrived” ‫ְלִמן־ַהיּוֹם ֶלֶכת‬ ‫( ַהֶמֶּלְך ַעד־ַהיּוֹם ֲאֶשׁר־ָבּא‬2Sam 19:25), where ὅτε instead of the genitive relative pronoun as attracted to its antecedent could have been used. In μὴ ἀπὸ τοῦ ξύλου, οὗ ἐνετειλάμην σοι τούτου μόνου μὴ φαγεῖν ἀπ’ αὐτοῦ ἔφαγες ‫ֲהִמן־ָהֵ֗עץ ֲאֶ֧שׁר ִצ ִוּיִ֛תיָך ְלִבְלִ֥תּי‬ ‫( ֲאָכל־ִמֶ֖מּנּוּ אָָֽכְלָתּ‬Gen 3:11) we notice a compromise whereby τούτου μόνου, an authentic ablative, is used as a partitive genitive, while the Hebraistic ἀπ’ αὐτοῦ has been left standing—unless it is to be construed with ἔφαγες. Only rarely has the asyndetic relative clause, alien to Greek, led to some awkward renderings, as in λαός μου, οὗ ὁ νόμος μου ἐν τῇ καρδίᾳ ὑμῶν ‫“ ַעם תּוֹ ָרִתי ְבִלָבּם‬a nation in whose heart is my law” (Isa 51:7). In the case of Isa 51:1, it is likely that the translator’s uncertainty over vocalization accounts for a rendering, ἀκούσατέ μου, οἱ διώκοντες τὸ δίκαιον καὶ ζητοῦντες τὸν κύριον, that is irreconcilable with MT ‫“ ַהִבּיטוּ ֶאל־צוּר ֻחַצְּבֶתּם ְוֶאל־ַמֶקֶּבת בּוֹר ֻנַקּ ְרֶתּם‬Look to the rock from which you were hewn and to the quarry from which you were dug.” Had the Masoretic text used the prepositional phrases ‫ ִמְמּנּוּ‬and ‫ ִמְמּ ָנהּ‬the translator might not have gotten confused. The proleptic or anticipatory use of the personal pronoun is infrequent in Greek, especially in oblique cases. As a result, ‫ְכּתו ֲֹעב ֵֹתיֶהם ַלְכּ ַנֲע ִני ַהִחִתּי ַהְפּ ִר ִזּי ַה ְיבוִּסי ָהַעמֹּ ִני ַהמֹּאִָבי‬ ‫( ַהִמְּצ ִרי ְוָהֱאמ ֹ ִרי‬Ezra 9:1) was mistranslated as ἐν μακρύμμασιν αὐτῶν, τῷ Χανανι, ὁ Εθι, ὁ Φερεζι, ὁ Ιεβουσι, ὁ Αμμωνι, ὁ Μωαβι, ὁ Μοσερι καὶ ὁ Αμορι.

2.3 The demonstrative pronoun Some of the ways in which the Hebrew demonstrative pronouns, ‫ ֶזה‬and ‫ ז ֹאת‬in particular, are used are unfamiliar in Greek. For instance, they can be almost enclitically attached to an interrogative to express the nuance of “(who, what, why, where, how) on earth?” (e. g. ‫ מה־זה ָעִשׂית‬τί τοῦτο ἐποίησας;, which could mean “why have you done this?”). But this interpretation could hardly apply in the case of Judg 2:2 (‫מה־זה‬ ‫ ֲעִשׂיֶתם‬ὅτε (B: ὅτι) ταῦτα ἐποιήσατε) or Gen 27:20 (‫ מה־זה ִמַה ְרָתּ ִלְמצֹא‬Τί τοῦτο, ὃ ταχὺ εὗρες; “what is that which you have found quickly?”). Note ‫ ֶבּן־ִמי־זה ָהָעֶלם‬υἱὸς τίνος ὁ νεανίας οὗτος; “whose son is this lad?” (1Sam 17:56) 21, where the transposition of the pronoun indicates that the translator analyzed it as attributive. 22 For an instance of another syntactic analysis, see ‫ ָלָמּה זה אָנִֹכי‬ἵνα τί μοι τοῦτο; (Gen 25:22). A question such as ἵνα τί τοῦτο ἐρωτᾷς τὸ ὄνομά μου; ‫( ָלָמּה זה ִתְּשׁאַל ִלְשִׁמי‬Gen 32:30) would have sounded odd to the average Greek speaker. Hence the Antiochene version ignores the demonstrative in ‫( ָלָמּה זה ֲא ִני ָצם‬2Sam 12:23) ἵνα τί οὖν νηστεύω ἐγώ;, although the Kaige version duly restores it: ἵνα τί τοῦτο ἐγὼ νηστεύω;. The collocation ‫ ֵאי ֶזה‬was generally assigned by Septuagint translators a value typical of Late and Post-Biblical Hebrew, namely “which?”, e. g. ‫ֵאי ְנָך יוֹ ֵדע ֵאי ֶזה ִיְכָשׁר ֲה ֶזה‬ ‫ אוֹ־ ֶזה‬οὐ γινώσκεις ποῖον στοιχήσει, ἢ τοῦτο ἢ τοῦτο “you do not know which is going to make it, this or that” (Eccl 11:6). This does not appear to be the case, however, 21. In the Antiochene version. 22. Possibly under the influence of Aramaic, cf. ‫“ ְד ָנה ִב ְנ ָי ָנא‬this building” (Ezra 5:4).

134

2. Morphology

8. Limitations of Greek in Representing Hebrew

in ‫ ֵאי־ ֶזה ַב ִית ֲאֶשׁר ִתְּבנוּ־ִלי ְוֵאי־ ֶזה ָמקוֹם ְמנוָּחִתי‬ποῖον οἶκον οἰκοδομήσετέ μοι; ἢ ποῖος τόπος τῆς καταπαύσεώς μου; (Isa 66:1). Another unidiomatic use of the demonstrative pronoun is akin to what has been described above. It can be used as a third constituent in a simple nominal clause, either in the second or third position. E. g. ‫( ַהאַָתּה ֶזה ְבּ ִני ֵעָשׂו‬Gen 27:21), which is virtually equivalent to ‫ַהאַָתּה הוּא ְבּ ִני ֵעָשׂו‬. The Septuagint rendered this as εἰ σὺ εἶ ὁ υἱός μου Ησαυ. A more literal translation, such as Εἰ σὺ εἶ αὐτὸς Ασαηλ; for ‫ַהאַָתּה ֶזה ֲעָשׂהֵאל‬ (2Sam 2:20), might have sounded rather unnatural. 23

2.4 Nouns and adjectives Hebrew has the category of dual only in a small number of substantives and cardinal numerals for “two,” “two hundred,” and “two thousand.” Excepting these numerals, the dual lost its original semantic value and dual substantives were used to designate the plural as well. Hence, to speak of “two eyes,” it became necessary to add a numeral to the dual form: ‫ְשֵׁתּי ֵעי ַנ ִים‬. By contrast, the dual had completely disappeared from Greek before the period under study. As regards case, Greek adequately copes with the directional -ā of Hebrew through the use of a variety of prepositions, e. g. ‫ ַהָבּ ְיָתה‬εἰς τὴν οἰκίαν “into the house” (Gen 43:16); ‫ ַהָשַּׁמ ְיָמה‬εἰς τὸν οὐρανόν “to the sky” (Gen 15:5); ἐπὶ τὴν θάλασσαν ‫“ ָיָמּה‬towards the sea” (Josh 16:3). Note also ‫ ַו ִיּפֹּל אַ ְרָצה‬πεσὼν χαμαί “falling to the ground” (Job 1:20).

2.5 Neuter gender Greek, unlike Hebrew, has three distinct genders. The neuter can be used, not only to designate objects whose substantive is neuter in gender, such as τέκνον, but also to refer to general ideas and notions, e. g. διὰ τοῦτο, “therefore,” (Jon 4:2). When Hebrew uses the feminine forms for this latter purpose, this particularity seems to be a source of difficulties for Septuagint translators. In οὐκ ἔγνωσαν ὅτι παρὰ κυρίου ἐστιν ‫( ל ֹא ָי ְדעוּ ִכּי ֵמ ְיה ָוה ִהיא‬Judg 14:4) the Hebrew pronoun is ambiguous. Although it could refer to Delilah, Jerome understood it as a neuter: nesciebant quod res a Domino fieret. In the case of καὶ ἐλογίσθη αὐτῷ εἰς δικαιοσύνην, the translator resolved the ambiguity of the Hebrew structure by changing the active voice into a passive one: ‫( ַו ַיְּחְשֶׁבָה לּוֹ ְצ ָדָקה‬Gen 15:6). Similar is ‫“ ַוֵתֶּצר ְל ָד ִוד‬it became distressful to David” (1Sam 30:6) καὶ ἐθλίβη Δαυιδ, where Δαυιδ is not in the dative case, as is shown by καὶ ἐθλίβησαν οἱ υἱοὶ Ισραηλ ‫( ַוֵתֶּצר ְל ִיְשׂ ָרֵאל‬Judg 10:9). The translator of Isaiah resorted to a more radical departure: ‫“ ל ֹא ָתקוּם ְול ֹא ִתְה ֶיה‬such shall not stand nor come to pass” becomes οὐ μὴ ἐμμείνῃ ἡ βουλὴ αὕτη οὐδὲ ἔσται “that plan will not endure nor come to pass” (Isa 7:7).

23. On this use of ‫ ֶזה‬, see Joüon / Muraoka, Hebrew Grammar, section 153 ja; T. Muraoka, “The tripartite nominal clause revisited” in: C. L. Miller (ed.), The Verbless Clause in Biblical Hebrew: Linguistic Approaches (Linguistic Studies in Ancient West Semitic 1), Winona Lake, IN 1999, 187-214, in particular 209-210. 2. Morphology

135

8. Limitations of Greek in Representing Hebrew

2.6 Prepositions Although Hebrew compound prepositions can be rendered by composite lexemes in Greek as well, as in ‫ ֵמַעל ִכֵּסּא ִיְשׂ ָרֵאל‬ἐπάνωθεν θρόνου Ισραηλ (1Kings 2:4), there are some limitations, e. g. ֹ ‫ אֲַח ֵרי ִהָפּ ֶרד־לו ֹט ֵמִעמּו‬μετὰ τὸ διαχωρισθῆναι τὸν Λωτ ἀπ᾽ αὐτοῦ (Gen 13:14), where the Hebrew compound preposition expresses a breach of bond and company between the patriarch and his nephew, which is not fully rendered by ἀπό.

2.7 Impersonal The third person masculine plural is commonly used in Hebrew to indicate a vague, non-specific subject. However, the use of the third person masculine singular is also well established. 24 Since this latter usage is unknown in Greek, some translators appear to have found it difficult to handle. For instance, ‫“ ִכּי אַָמר ֵאַלי ָערוֹם ַיְע ִרם הוּא‬for one said to me, ‘He is rather wily’” (1Sam 23:22) is rendered in the Antiochene version as ὅτι εἶπε ὁ Σαουλ Μήποτε πανουργευσάμενος οὗτος πανουργεύσηται, by deleting ‫ֵאַלי‬. The impersonal use of the third person masculine singular passive to focus on an action appears to be foreign to Greek, 25 though it is well attested in Hebrew. 26 This could explain the grammatical transformation observable in ‫ֶע ְגַלת ָבָּקר ֲאֶשׁר ל ֹא־ֻעַבּד‬ ‫“ ָבּהּ‬a heifer with which no agricultural work has been done” (Deut 21:3), where the Septuagint has δάμαλιν ἐκ βοῶν, ἥτις οὐκ εἴργασται, with δάμαλιν as the subject. This compelled the translator to perform another grammatical transformation in the following verse: ֹ‫“ ַנַחל ֵאיָתן ֲאֶשׁר ל ֹא־ ֵיָעֵבד בּו‬a permanently flowing wadi in which no farming shall be done,” which he rendered φάραγγα τραχεῖαν, ἥτις οὐχ εἴργασται “a rugged valley, which has not been worked on,” transforming the future into a preterite (perfect) on the analogy of verse 3. A reader of the Septuagint ignorant of Hebrew will be forgiven for thinking that the subject of ἀφεθήσεται is ἁμαρτία in […] ἀπὸ τῆς ἁμαρτίας αὐτοῦ, καὶ ἀφεθήσεται αὐτῷ “[…] off his sin, and forgiveness will be granted to him” (Lev 4:35), where the Masoretic text has ‫ַעל־ַחָטּאתוֹ ֲאֶשׁר־ָחָטא‬ ֹ‫ ְו ִנְסַלח לו‬. The same reader would supply the same noun as the subject in verse 31: καὶ ἀφεθήσεται αὐτῷ, although it, the noun, is not explicitly mentioned. One can easily imagine the bewilderment of the translator trying to render ‫“ אֲַח ַר ִיְך ל ֹא זוּ ָנּה‬you were not sought after as a whore” (Ez 16:34) with the standard impersonal construction: μετὰ σὲ οὐ πεπορνεύκασιν. 27 See καὶ ἐξαπέστειλε Δαυιδ καὶ ἐξεζήτησε τὴν γυ24. Joüon / Muraoka, Hebrew Grammar, section 155 b. 25. A rare example of this is ἑπτάκις ἐκδεδίκηται ἐκ Καιν “seven times vengeance has been taken against Cain” (Gen 4:24). 26. See Joüon / Muraoka, Hebrew Grammar, section 152 e, f, 155 b-i. 27. The reading μετὰ σὲ οὐ was preferred by Joseph Ziegler on the authority of a single witness, Papyrus 967; all other witnesses have μετὰ σοῦ. Should the former be right, the preposition must have been used pregnantly for ἑπόμενοι μετὰ σὲ. It cannot be rendered: “après toi ils ne se sont pas prostitués” (pace K. Hauspie, La version de la Septante d’Ézéchiel. Traduction annotée d’Ez 1-24 et étude du grec d’Ézéchiel par une sélection de particularités lexicales et grammaticales, dissertation, University of Leuven, 2002, 109) nor can it be paraphrased “nach deiner

136

2. Morphology

8. Limitations of Greek in Representing Hebrew

ναῖκα καὶ εἶπον Οὐχὶ αὕτη ἐστὶ Βηρσαβεε, θυγάτηρ Ηλα, γυνὴ Οὐρίου τοῦ Χετταίου; (2Sam 11:3), where the Kaige recension duly revised the idiomatic impersonal εἶπον, changing it to the singular, εἶπεν, in order to conform to MT ‫ ַויּ ֹאֶמר‬, thereby leaving the average Septuagint reader puzzled. Note also ‫ ֵיאֵָמר‬ἐροῦσιν (Gen 10:9); it is immaterial whether the translator mentally vocalized the Hebrew verb as ‫ יֹאַמר‬or not.

2.8 Root and stem The grammar and lexicon of many languages are structured according to a large number of stems. Each stem can be assigned a general meaning. Stems can consist of a single member, though most have multiple members. One such stem in English is /s-ŋ/, represented by such lexemes and inflected forms as sing, sang, sung, sings, singing, song, singer. In Greek, λαμβάνω, λήμψομαι, ἔλαβον, λήμψις, λήμπτωρ, and tens of other compounds and their literally hundreds of fully inflected forms can be reduced to a single stem. Just as in Semitic “roots,” the majority of these stems consist of a fixed sequence of consonants. An important difference between non-Semitic stems and Semitic roots derives from ways in which the phones that make up the actual words and inflected forms are represented graphically. In most non-Semitic languages there is little one-to-one correspondence between vowel graphemes and phonemes. In Semitic languages, however, vowel marking is extremely limited, regardless of the alphabet used. When Septuagint translators went about the task of translating from an unvocalized Hebrew text that made only infrequent use of matres lectionis, they must have visually perceived the roots more clearly than when they heard the text recited at normal speed. Acrostic poems that treat ‫ שׂ‬and ‫ שׁ‬as a single letter can be properly appreciated against the background of this culture of written texts. An attentive reader may have been struck by the recurrence, at crucial points in the narrative, of the root ‫ שׁכב‬in the story about David and Bathsheba: ‫( משׁכבו‬2Sam 11:2), ‫וישׁכב עמה‬ (11:4), ‫( וישׁכב אוריה‬11:9), ‫( לשׁכב עם־אשׁתי‬11:11), ‫( לשׁכב במשׁכבו עם־עבדי אדניו‬11:13), ‫בחיקו‬ ‫( תשׁכב‬12:3), ‫( ושׁכב עם־נשׁיך‬12:11), ‫( ושׁכב ארצה‬12:16), ‫( וישׁכב עמה‬12:24)—a total of ten occurrences. Even if our translator had noticed this, he would have been unable to reproduce this effect in Greek in a natural fashion. The Greek words he selected were κοίτης, κοιμᾶται, κοιμᾶται, κοιμηθῆναι, ἐκοιμήθη ἐν τῇ κοίτῃ, ἐκάθευδεν, κοιμηθήσεται, ἐκάθευδεν, κοιμᾶται. The choice of καθεύδειν in 2Sam 12:3 was a reasonable one in the context. However, it blurs the contrast between the female lamb Art” (pace Septuaginta Deutsch, Stuttgart 2009, 1377). That our understanding of this phenomenon has text-critical implications is also evident in ‫ַהְלֶּבן ֵמאָה־ָשׁ ָנה ִי ָוֵּלד ְוִאם־ָשׂ ָרה ֲהַבת־ִתְּשִׁעים ָשׁ ָנה‬ ‫( ֵתֵּלד‬Gen 17:17), for which John W. Wevers, in his critical edition (Göttingen, 1974), reads εἰ τῷ ἑκατονταετεῖ γενήσεται υἱός, καὶ εἰ Σαρρα ἐνενήκοντα ἐτῶν οὖσα τέξεται, though Alfred Rahlfs, in his Handausgabe (Stuttgart, 1935), does not print υἱός. If the translator was thinking of the divine promise given one verse before, ‫ ָנַתִתּי ִמֶמּ ָנּה ְלָך ֵבּן‬, where he had used τέκνον, why should he change the noun now? The absence of an object with τέξεται also supports the impersonal analysis: “is a birth of a baby likely to … ?” The reader with no Hebrew may have gone even farther: “Is such a thing likely to happen … ?” Incidentally, the use of γίνομαι in the sense of “to be born,” is firmly attested. See T. Muraoka, A Greek-English Lexicon of the Septuagint, Leuven 2009, s. v. 2 b. 2. Morphology

137

8. Limitations of Greek in Representing Hebrew

—a symbol of Bathsheba—lying in idyllic and innocent security and the adulteress who felt guilt and anxiety. The same holds for the same Greek verb chosen at verse 16, where the Hebrew text impresses upon the reader the extraordinary scene of a king forced to lie on the ground next to a dying infant of his instead of in his king-size bed intent on an illicit relation.

2.9 Binyanim or stems of the verbal conjugation The Hebrew (or Semitic, for that matter) binyanim system overlaps, to some extent, the Greek voice system. Yet while a considerable amount of research has been done on the Greek voice system, we know comparatively little about the morphosyntax of the Hebrew binyanim, namely how the functions of the five principal binyanim differ from one another, both theoretically and with reference to concrete verb lexemes, e. g. precisely how qal ‫ ָר ַרף‬differs from piel ‫ ִר ֵדּף‬. Since the principal purpose of translation is to transmit the message contained in the source text, one ought not to quibble over the use of μανθάνειν to render qal ‫ָלַמד‬ as against the use of διδάσκειν to render piel ‫ִלֵמּד‬. However, there is little doubt that something of the total communicative load of the source language was lost here. In ‫ִאם‬ ‫( ל ֹא ַתֲאִמינוּ ִכּי ל ֹא ֵתאֵָמנוּ‬Isa 7:9), the prophet was undoubtedly playing with words. For him not only the message, the substance, but also the form, how the message was packaged, mattered just as much. Our Septuagint translator found the challenge too daunting, coming up with ἐὰν μὴ πιστεύσητε, οὐδὲ μὴ συνῆτε. By contrast, another translator appears to be facing the challenge head on: ‫ַהֲאִמינוּ‬ ‫ ַבּיה ָוה ֱאל ֵֹהיֶכם ְוֵתאֵָמנוּ ַהֲאִמינוּ ִב ְנִביאָיו ְוַהְצִליחוּ‬ἐμπιστεύσατε ἐν κυρίῳ θεῷ ὑμῶν καὶ ἐμπιστευθήσεσθε, ἐμπιστεύσατε ἐν προφήτῃ αὐτοῦ καὶ εὐοδωθήσεσθε (2Chron 20:20), where, in addition to the use of the same verb in the first half, the selection of the future passive for the second verb in the parallel spot of both halves underlines the benefit and reward to be had by acting on the instruction expressed by their respective imperative, and that against the structure of the source text—imperative/yaqtul and imperative/imperative. 28

28. Cf. E. Tov, “The representation of the causative aspects of the hiphil in the LXX. A study in translation technique,” Bibl. 63 (1982), 417-424.

138

2. Morphology

9. The Classification of Literalism in Ancient Hebrew-Greek Translation Cameron Boyd-Taylor

Introduction How we conceptualize the Septuagint as a translation determines in no small measure what we make of its language. Whether one is describing the relationship between a particular translation and its putative source, or characterizing the translation technique which underlies it, the tendency has been to distinguish the “literal” from the “free,” 1 and much intellectual labour has been invested in classifying the books of the Septuagint under these two headings. 2 In this respect, Septuagint Studies reflects the larger context of Western translation discourse, dominated as it has been by the opposition of “free” and “literal” translation. 3 As Susan Bassnett notes, this dichotomy has arisen time and again in accordance with the prevailing views on language. 4 Traditional theory never quite overcame it, and it continues to play a role in translation studies today. 5 The locus classicus for the modern conception of translation as an “either/or” is Friedrich Schleiermacher’s lecture “Über die verschiedenen Methoden des Übersetzens,” delivered in 1813, in which he asserts: Entweder der Übersetzer läßt den Schriftsteller möglichst in Ruhe, und bewegt den Leser ihm entgegen; oder er läßt den Leser möglichst in Ruhe und bewegt den Schriftsteller ihm entgegen. Beide sind so gänzlich von einander verschieden, daß durchaus einer von beiden so streng als möglich muß verfolgt werden, aus jeder Vermischung aber ein höchst unzuverlässiges Resultat nothwendig hervorgeht, und zu besorgen ist daß Schriftsteller und Leser sich gänzlich verfehlen. 6

Schleiermacher’s dictum has itself been translated and glossed in different ways; however construed, its importance lies in the figure of two opposing movements between the source and target languages, the one foreignizing or literalist (tending towards the word-for-word strategy), the other domesticating or naturalizing (tending towards a

1. 2. 3. 4. 5. 6.

B. Lemmelijn, A Plague of Texts? A Text-Critical Study of the So-called ‘Plagues Narrative’ in Exodus 7:14–11:10, Leiden 2009, 108. J. M. Dines, The Septuagint, London 2004, 120. A. Chesterman, Memes of Translation: The Spread of Ideas in Translation Theory, Amsterdam 1997, 12. S. Bassnett, Translation Studies (third edition), London 2002 [1980, 1991], 48-50. M. Snell-Hornby, Translation Studies: An Integrated Approach (revised edition), Amsterdam 1995 [1988], 9. F. Schleiermacher, “Über die verschiedenen Methoden des Übersetzens” in: H. J. Störig (ed.), Das Problem des Übersetzens, Darmstadt 1963, 38-70. Introduction

139

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

sense-for-sense translation). 7 Translation thus involves a fixed choice, either a movement away from the reader and towards the author or away from the author and towards the reader. While no great claim can be made for Schleiermacher’s originality, his remarks neatly encapsulate the terms of a binary structure that has figured significantly in otherwise quite different theories of translation. 8 Its influence on Septuagint Studies has been pervasive. Recent Septuagint scholarship has been particularly self-conscious about references to the “literal” and “free,” and there have been various attempts to nuance the distinction. Jennifer Dines suggests that we think in terms of a continuum of renderings running from the extremely “literal” to the extremely “free” on which the different translations (or parts thereof) could be located. 9 On a more radical note, Bénédicte Lemmelijn moots the idea that the opposition is unusable in se. 10 Yet judging by its widespread use, one can only conclude that the opposition is ipso facto usable, and will remain a part of how we talk about the Septuagint in the foreseeable future. Nevertheless the current unease is by no means misplaced. Notwithstanding the utility of the “either/or,” one might argue that it seriously risks distorting the object of study. It is moreover ideologically freighted, and so, while use it we will, it is important to do so with some measure of critical reflection. The touchstone for contemporary discussion of “literal” and “free” translation in Septuagint Studies remains James Barr’s 1979 publication, The Typology of Literalism in Ancient Biblical Translations (henceforth, Typology). 11 Barr’s stated aim is to analyse what is meant by the terms “literal” and “free” in the context of the ancient biblical versions to the end of classifying the various modalities of literal translation. The seminal nature of his work cannot be over-emphasized. My intention here is to locate Typology both with respect to the issues it originally addressed, and in relation to more recent discussion, to the end of further clarifying the analysis of literalism. I begin by rehearsing the three-stranded argument of Typology, the burden of which is that the terms “literal” and “free” as commonly used are ill-defined (§ 1). I then delineate Barr’s typological analysis, and its later refinement as a taxonomy of formal equivalency (§ 2). 12 Finally, I discuss the shift towards a more target oriented methodology, concluding with some general remarks on the binary model in light of this shift (§ 3).

7. A. Pym, “Schleiermacher and the Problem of Blendlinge” Translation and Literature 4/1 (1995), 5-30, here 5. 8. Pym, “Schleiermacher,” 6. 9. Dines, Septuagint, 121. 10. Lemmelijn, Plague of Texts, 110. 11. J. Barr, The Typology of Literalism in Ancient Biblical Translations, Göttingen 1979, 280. 12. The terms typology and taxonomy are often used interchangeably in reference to the end result of a process of classification. For the purposes of the present study, a typology is primarily conceptual and qualitative while a taxonomy is empirical and quantitative. See K. D. Bailey, Typologies and Taxonomies: An Introduction to Classification Techniques (Quantitative Applications in the Social Sciences 102), London 1994, 6.

140

Introduction

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

1. The Argument Underlying Barr’s critique of the “either/or” are three distinct lines of argument. The first is historical, and pertains to translation discourse rather than translation as such (§ 1.1). His point is that “literal” and “free” were not “clear and simple terms in the world of ancient biblical translation.” 13 The second line of argument is descriptive and methodological (§ 1.2). Barr maintains that the opposition of “literal” and “free” fails to differentiate translations in an interesting manner. His third point addresses the issue of literalism as it applies to the translation technique of the ancient translators (§ 1.3). 14 According to Barr, the binary model does not fit the evidence. Of the three strands in Barr’s argument, it is the second which bears most directly on the matter of typological analysis. I shall therefore flesh it out somewhat by introducing the concept of preference rules (§ 1.2). While the term is not used by Barr, it helps one conceptualize the various modalities of literalism that he wishes to distinguish in his typology (§ 2.1).

1.1 Cicero and the Septuagint To assume a priori that the binary model represents a universal in translation is questionable, and if one is going to ascribe it to Hellenistic Jewish translators some form of justification is required. Unfortunately modern investigators have little to go on, and so the tendency has been to extrapolate from Roman translation discourse. The most significant effort to date within Septuagint Studies has been that of Sebastian Brock. In an influential paper published in 1972, Brock took up Schleiermacher’s metaphor of opposing movements, distinguishing translations that bring the reader to the author (“literal”), from those that bring the author to the reader (“free”). 15 According to Brock, the ancients tended to adopt the former strategy for non-literary texts, and the latter for literary texts. His touchstone was the classic formulation of Marcus Tullius Cicero (106-43 BCE), non converti ut interpres sed ut orator, construed according to the binary model, that is, as an opposition between literalist and naturalizing translation. 16 While the former characterized the approach of the professional hack, the interpres, the latter befitted the literate man of affairs, the orator. Brock maintained that this distinction informed the practice of the Hebrew-Greek translators of the Hel-

13. Barr, Typology, 323-324. 14. For the term “translation technique,” see A. Aejmelaeus, “What We Talk About When We Talk About Translation Technique” in: A. Aejmelaeus (ed.), On the Trail of the Septuagint Translators: Collected Essays (Contributions to Biblical Exegesis & Theology 50), Leuven 2007, 205-222, here 205-206. Aejmelaeus suggests that “translation technique” may serve as a “neutral term to denote the activity of the translator or the process of translation which led from the Vorlage to the translation.” 15. S. P. Brock, “The Phenomenon of the Septuagint” Old Testament Studies 17 (1972), 11-36, here 17. See also S. P. Brock, “Aspects of Translation Technique in Antiquity” Greek, Roman, and Byzantine Studies 20 (1979), 69-87; S. P. Brock, “The Phenomenon of Biblical Translation in Antiquity” The University of Birmingham Review 2/8 (1969), 96-102. 16. De optimo genere oratorum, 46 BCE. 1. The Argument

141

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

lenistic period. 17 With two basic strategies available to them, the translators chose one or the other according to the genre of the source text. On this hypothesis the earliest translators of the Pentateuch would have been faced with something of a dilemma, as this literature contains both literary and legal texts. According to Brock they struck a compromise, vacillating between the free periphrastic style used for literary texts and the calque translation characteristic of legal documents. 18 Brock’s appeal to Cicero is, on Barr’s view, questionable. Over three centuries separate the translators of the Septuagint from the great orator; moreover, Cicero’s comments were rooted in the practice of contemporary Greek-Latin translation, the social and cultural context of which was quite different from that of third-century BCE Alexandria. One might add that Brock’s extrapolation rests on a misreading of Cicero, for while he is often portrayed as a champion of the sense over the word, what is primarily at issue for him is the concept of imitation (imitatio), and specifically the use of Greek oratory as a model for Roman practice. 19 The opposition on which non converti ut interpres sed ut orator turns is not the “literal” and “free” as such, but rather the rival claims of grammar and rhetoric, each a highly developed academic discipline with distinct methods and aims. As Rita Copeland has argued, Roman translation discourse provided a way of clarifying the differences between these two rival disciplines, and as such reflected the tensions between them. 20 Barr was correct to place a question mark beside the relevance of Cicero’s remarks to the study of the Septuagint. Brock’s picture of ancient Hebrew-Greek translators negotiating Schleiermacher’s “either/or” is anachronistic. There is no reason to believe that they thought about translation in such terms; rather they faced a distinct series of problems. 21

1.2 Methodological Considerations Quite apart from how the ancient translators viewed their task, there is the question of how we ourselves conceptualize it. It has proven useful to classify the translations of the Septuagint according to their characteristic manner of rendering certain Hebrew terms or phrases. Such classification is then used as a basis for arbitrating claims touching a wide range of issues, such as the number of translators involved in a particular book, their methods, their theology, the quality of their Greek, or the form of their Hebrew source. 22 Underlying most classificatory schemes is the opposition of “literal” and “free.” A case in point is the highly influential study by Henry St. John Thackeray, which ostensibly classifies the books of the Septuagint on stylistic

17. Brock, “Phenomenon of the Septuagint,” 20. 18. Brock, “Phenomenon of the Septuagint,” 32. 19. His aim as a translator is aptly described by R. Copeland, Rhetoric, Hermeneutics and Translation in the Middle Ages: Academic Traditions and Vernacular Texts, Cambridge 1991; 31, as one of inventive difference. See also Bassnett, Translation Studies, 50. Cicero’s target audience was comprised of highly sophisticated readers who could compare his translation to the original Greek speeches. 20. Copeland, Rhetoric, Hermeneutics and Translation, 10. 21. Barr, Typology, 282. 22. M. L. Wade, “Evaluating Consistency in the Old Greek Bible” BIOSCS 33 (2000), 53-75, here 53.

142

1. The Argument

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

grounds. 23 While Thackeray’s work remains of value, it is telling that one of his groups is comprised of so-called “literal or unintelligent versions,” while another is referred to as “paraphrases and free renderings.” As Barr observes, one and the same translation can be both “literal” and “free” in different modes or on different levels; due to its lack of flexibility, the binary model fails to account for this complicated set of relations. 24 For the purposes of classification, this is a crucial failure. By way of illustration, let us consider two Greek translations of the same verse of the Hebrew Psalter, both of which have been classified as “literal”: the version that comes down to us in the Septuagint (�Ps), 25 dating to some time in the second century BCE (also known as the Old Greek Psalter), 26 and the translation attributed to Aquila (α´), a product of Tannaitic Judaism. 27 The consonantal form of the Masoretic text (� 30:10) provides an approximation of the Vorlage of � (29:10). 28 � ‫מה בצע בדמי ברדתי אל שחת‬

‫היודך עפר היגיד אמתך‬

�Ps τίς ὠφέλεια ἐν τῷ αἵματί μου ἐν τῷ καταβῆναί με εἰς διαφθοράν μὴ ἐξομολογήσεταί σοι χοῦς ἢ ἀναγγελεῖ τὴν ἀλήθειάν σου α´ τί πλεονέκτημα ἐν αἵματί μου ἐν τῷ καταβῆναί με πρὸς διαφθοράν μὴ ἐξομολογήσεταί σοι χοῦς ἢ ἀναγγελεῖ ἀλήθειάν σου

Both translators adhere closely to the structure of their source, matching each Hebrew element with a corresponding Greek one. Yet both deviate from a point-to-point correspondence; moreover they differ slightly one from the another. What is interesting about these differences is not adequately captured by the opposition of “literal” and “free.” More refined analytic tools are required. One such tool is the notion of preference, which has to do with what an agent does in a situation where he or she registers various courses of action as open: values are assigned to the different options, such that the one with the highest value is the one that is chosen. 29 Such preferences are inferred by the investigator, and take the form of rules. 30 They provide a means of describing 23. H. St. John Thackeray, A Grammar of the Old Testament in Greek According to the Septuagint, vol. 1, Introduction, Orthography and Accidence, Cambridge 1909, 6-16. 24. Barr, Typology, 280. 25. The Greek text (�) is that of A. Rahlfs (ed.), Septuaginta. Id est Vetus Testamentum graece iuxta LXX interpretes, Stuttgart 1935. 26. T. Williams, “Towards a Date for the Old Greek Psalter” in: R. Hiebert / C. Cox / P. Gentry (eds.), Old Greek Psalter: Studies in Honour of Albert Pietersma, Sheffield 2001, 248-276, here 276. The external evidence is consistent with a date sometime in the early second century BCE. 27. The text attributed to Aquila is from Codex rescriptus bybliothecae Ambrosianae O 39 sup. (Ra 1098), a palimpsest manuscript of the Octateuch with a lower text containing about one hundred and fifty verses of the Hexapla Psalter. The edition is G. Mercati, Psalterii Hexapli reliquiae II, Rome 1958. 28. The Hebrew text (�) used throughout the present study is that of K. Elliger / W. Rudolph (eds.), Biblia Hebraica Stuttgartensia (fifth corrected edition), Stuttgart 1997. 29. J. Bennett, Linguistic Behaviour, Cambridge 1976, 66. 30. A preference model provides a way of construing interpretation as a form of competence. See E. Schauber / E. Spolsky, The Bounds of Interpretation: Linguistic Theory and Literary Text, Palo Alto, CA 1986, 7. The approach has been used with great success in lexical analysis by R. Jackendoff, Semantics and Cognition, Boston 1983, 128-160. For an application to the lexical 1. The Argument

143

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

categorization decisions in terms of a range of weighted factors. 31 Within linguistic analysis preference rules complement well-formedness rules. In the case of translation, we may distinguish two sorts of factors bearing on preference: (i) those associated with the representation and transformation of the source text, and (ii) those associated with composition in the target language. Of the former, the most interesting variable is the translator’s preferred equivalencies. In any given instance, the process of translation will involve a trade-off between values, with the result depending upon the relative weighting of preferred equivalencies over-against the preference rules and grammar of the target system. So much is evident in the use of the article by �Ps and α´ at Ps 29:10. Both translations exhibit a preference for formal equivalence or isomorphism. Nevertheless in rendering the Hebrew phrases ‫ ברדתי‬and ‫אמתך‬, �Ps provides arthrous nouns, while α´ does not. The introduction of articles in �Ps, triggered by the attributive genitive pronouns μου and σου respectively (corresponding to pronominal suffixes in the parent), might be said to reflect a preference rule of the target language. The article is not, however, grammatically obligatory in these phrases, and is thus absent in α´, where a higher value is evidently placed on formal equivalence. Yet both versions render ‫ ברדתי‬with an arthrous infinitive. It turns out then that α´ will introduce the article without a formal warrant from the source text when it is obligatory in the target language. We thus infer that α´ places a greater value on grammaticality than on formal equivalence. For the most part both translators employ the same lexical replacements; where they differ is thus telling. 32 While �Ps renders ‫ בצע‬by ὠφέλεια, reflecting the more common use of the Hebrew term in the sense “profit,” i. e. “an obtained benefit or advantage,” α´ supplies πλεονέκτημα, which carries the connotation of “rapacity,” and thus picks up on the etymology of the Hebrew word, “a gain made by violence.” Unlike �Ps, α´ exhibits a strong preference for etymological matches (perhaps reflecting the interests of its Tannaitic milieu). Finally, we note that neither translator renders interrogative ‫ ה‬with a morphosyntactical equivalent (e. g. ἦ). Rather, both replace the first occurrence with the negative particle μή, and the second with ἤ. This suggests a preference for syntactically well-formed clauses. The two questions are rhetorical and anticipate a negative reply, which is aptly signaled in Greek by μή. Since the clauses are coordinate, however, the use of a second μή would be intrusive: the conjunction ἤ better reflects the syntactical relationship, and results in a higher degree of cohesion at the level of discourse. Here then are two highly “literal” translations, yet their syntax is conditioned by factors peculiar to the Greek language.

semantics of Hebrew-Greek translation, see C. Boyd-Taylor, “The Evidentiary Value of the Septuagint for Lexicography—Alice’s Reply to Humpty Dumpty” BIOSCS 34 (2001), 47-80. 31. Schauber / Spolsky, Bounds of Interpretation, 8. 32. For a more detailed descriptive analysis of these texts see C. Boyd-Taylor, Reading between the Lines: The Interlinear Paradigm for Septuagint Studies (Biblical Tools and Studies 8), Leuven 2011, 222-264.

144

1. The Argument

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

1.3 Explanatory Fit Barr’s third argument involves two points. His first is that while the “either/or” implies a fixed choice, many of the ancient translators do not exhibit a definite policy one way or the other, rather they often seem to work in an ad hoc manner, opting for a more or less “literal” rendering according to the character of the original text and its local context. 33 The question of whether Barr is correct on this matter cannot be decided in advance of descriptive analysis, yet, in general terms, the binary model does not seem to fit the evidence very well. To illustrate this, Anneli Aejmelaeus has drawn attention to the inconsistent treatment of Hebrew temporal expressions in which the preposition ‫ ב‬is followed by a suffixed infinitive. 34 She presents Lev 23:43 as an example of one strategy. 35 � ‫כי בסכות הושבתי את בני ישראל בהוציאי אותם מארץ מצרים‬ �Lev ὅτι ἐν σκηναῖς κατῴκισα τοὺς υἰοὺς Ισραηλ ἐν τῷ ἐξαγαγεῖν με αὐτοὺς ἐκ γῆς Αἰγύπτου

The Septuagint version of Leviticus (�Lev), a product of third-century BCE Alexandria, is characterized on the whole by its literalism; it evidently translates a Hebrew text very close to �. 36 Here the bound formation ‫ )�( בהוציאי‬is rendered ἐν τῷ ἐξαγαγεῖν με (�Lev). In this way, the Hebrew expression has been represented as the translator analysed it (namely, as three constituents: preposition + infinitive + pronoun) by providing a Greek match for each element, and retaining the word order. We observe that the morphosyntactical class of each Greek word is the same as that of its Hebrew counterpart. 37 The Greek thus has a fairly recognizable relation to the Hebrew segment it renders. To describe this phenomenon in terms of preferred equivalencies, we might distinguish three preference rules: “match for constituent,” “match for word class,” and “match for sequence.” Here a preferred equivalency trumps a preference rule of the target language—the temporal use of ἐν + the infinitive has a low preference value in Greek 38—though the trade-off can go the other way, as we see in the following example at Lev 16:23. � ‫ופשט את בגדי הבד אשר לבש בבאו אל הקדש‬ �Lev καὶ ἐκδύσεται τὴν στολὴν τὴν λινῆν, ἥν ἐνδεδύκει εἰσπορευομένου αὐτοῦ εἰς τὸ ἅγιον 33. Barr, Typology, 280-281. 34. A. Aejmelaeus, “Translation Technique and the Intention of the Translator” in: Aejmelaeus, On the Trail, 59-70, here 60-61. 35. The Greek text is that of J. W. Wevers (ed.), Leviticus (Setuaginta, Vetus Testamentum Graecum, Auctoritate Academiae Scientiarum Gottingensis editum, vol. II,2), Göttingen 1986. 36. J. W. Wevers, Notes on the Greek Text of Leviticus (Septuagint and Cognate Studies 44), Atlanta, GA 1997, ix. 37. It is not improbable that the translator’s analysis was informed by the concept of word classes (τὰ μέρη τοῦ λόγου) that was being developed at the time. See E. Dickey, Ancient Greek Scholarship, Oxford 2007, 126-127. 38. The temporal function is attested in Ptolemaic papyri and Polybius. See I. Soisalon-Soininen, Die Infinitive in der Septuaginta, Helsinki 1965, 81. It did not, however, become a preferential feature of literary Greek. 1. The Argument

145

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

Again the Hebrew uses an infinitival construction with temporal force (‫)בבאו‬, but here the expression is replaced by a genitive absolute (εἰσπορευομένου αὐτοῦ), a construction peculiar to Greek and carrying a stylistic value. This strategy precludes rendering the enclitic preposition (‫)ב‬. The preference for matching by word class, evident at Lev 23:43, is thus not as heavily weighted here. Note, however, that the translation is still characterized by a high degree of literalism. The use of the less common second attributive construction τὴν στολὴν τὴν λινῆν (article + noun + article + adjective) nicely mimics the structure of ‫( את בגדי הבד‬particle + noun + article + adjective) and avoids changing the word order (as τὴν λινῆν στολὴν would). More significantly, it is evident that the genitive absolute has been introduced by �Lev with little regard for the larger syntactical context; it is, as John William Wevers remarks, infelicitous (though not ungrammatical), since αὐτοῦ refers to the subject of the clause (Aaron), a reflex of the Hebrew. 39 Even though the rendering involves the replacement of one construction by another, certain formal features of the source are more heavily weighted than a preference rule of Greek syntax. This raises an interesting point regarding the translation technique of Greek Leviticus: transformations frequently occur cheek by jowl with isomorphism. 40 Whether such inconsistency is attributable to the “intuition and spontaneity” of the translator (so Aejmelaeus), or is the result of a deliberate strategy of some sort, remains an open question. In this regard, Wevers has suggested that the translator, although a literalist, favored variety as a stylistic feature, a preference consistent with literary composition in the target language. 41 However we account for the phenomenon of varied rendering, Barr’s point stands: the bipolar model is inadequate to the task. A further reason for Barr’s discontent with the “either/or” is his contention that what is today regarded as “free” translation hardly existed in Greco-Roman antiquity. 42 Given the dearth of evidence prior to the Common Era, this claim is difficult to assess. Certainly by the fourth century CE there is evidence for a wide range of Latin-Greek translations, from the utilitarian to the highly literary (including creative adaptations); if we turn to Greek-Latin translation, we also find variety. 43 Thus while it is common to assert that ancient translation was typically literal, this is but an assumption. One might, of course, argue that the discipline of grammar as it was taught in the schools implicitly reinforced the universal tendency of translators to lean on the form of the source text. Hellenistic grammatical analysis (τέχναι γραμματικαί), after all, focused on the word as the central unit of language. 44 As Siobhàn McElduff observes, such a system (even if it did not teach translation) could well have produced translators with a strong inclination towards word-for-word translation; one could thus speak of a “lexicon driven” approach, for which there is some evidence in GrecoWevers, Notes on Leviticus, 253. Aejmelaeus, “Translation Technique,” 60. Wevers, Notes on Leviticus, xi. Barr, Typology, 281. For a comprehensive study of the theory and practice of translation in Rome from 240 BCE to the second century CE, see S. McElduff, Roman Theories of Translation: Surpassing the Source (Routledge Monographs in Classical Studies), New York, NY 2013. 44. C. C. De Jonge, Between Grammar and Rhetoric: Dionysius of Halicarnassus on Language, Leiden 2008, 91.

39. 40. 41. 42. 43.

146

1. The Argument

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

Roman Egypt. 45 Yet whether such habits of mind played a role in Hebrew-Greek translation remains speculative. If the typology of literalism requires justification, it must come from methodological considerations.

2. Towards a Classification of Literalism In Typology Barr explores the logic of literalism, that is, he examines the ways in which “literal” and “free” translation may be opposed. Yet, on his view, the literalism of a translation does not reside simply in its relationship to the source text, but on how the source was handled by the translator. 46 Typological analysis thus requires that the investigator hypothesize the path which led from one text to the other. 47 At each point in the analysis one asks, What has the translator done, and why? 48

2.1 Barr’s Typology Barr isolates “six modes of difference” between translations of the Hebrew text, each of which is a distinct variable within literalism. 49 To illustrate these modalities, I shall look at selected renderings from two versions of 3Reigns 21:12 (1Kings 20:12), both of which could be described as relatively “literal.” 50 �Rgns comes from a non-revised (nonKaige) section of the Septuagint version, likely dating to the second century BCE. 51 Since typological analysis aims at the qualitative differentiation of translations, I compare �Rgns to a version of the same text attributed to Aquila (α´). 52 � serves as a point of reference for the analysis (with the caveat that the Vorlage of � evidently differed from � at various points). I again have recourse to the concept of preference (introduced above in § 2.1) to facilitate the analysis. � ‫ויהי כשמע את הדבר הזה‬ �Rgns καὶ ἐγένετο ὅτε ἀπεκρίθη αὐτῷ τὸν λόγον τοῦτον α´ καὶ ἐγένετο ὡς ἤκουσεν σὺν τὸ ῥῆμα τοῦτο � ‫והוא שתה הוא והמלכים בסכות‬ �Rgns πίνων ἦν αὐτὸς καὶ πάντες οἱ βασιλεῖς μετ’ αὐτοῦ ἐν σκηναῖς α´ καὶ αὐτὸς ἔπινεν αὐτὸς καὶ οἱ βασιλεῖς ἐν συσκιασμοίς 45. S. McElduff, “Living at the level of the word: Cicero’s rejection of the interpreter as translator” Translation Studies 2,2 (2009), 133-146, here 140. 46. Barr, Typology 288. 47. Barr, Typology, 285. 48. See K. J. Vanhoozer, Is There a Meaning in This Text?, Grand Rapids, MI 1998, 234, for whom understanding a text is ultimately a matter of interpreting human action. 49. Barr, Typology, 294. 50. For a descriptive analysis of these texts see Boyd-Taylor, Reading between the Lines, 145-152. The text of � is that of Rahlfs, Septuaginta. 51. The text of � is that of Rahlfs, Septuaginta. 52. The text of α´ is drawn from Cambridge UL T-S 12.184 and UL T-S 20.50, fragments of a palimpsest manuscript. The lower text was written in a Greek uncial hand (fifth-sixth century ce), and contains a version of Reigns identified as Aquila. The edition is that of F. C. Burkitt, Fragments of the Books of Kings according to the Translation of Aquila, Cambridge 1897. 2. Towards a Classification of Literalism

147

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

� ‫ויאמר אל עבדיו שימו וישימו על העיר‬ �Rgns καὶ εἶπεν τοῖς παισὶν αὐτοῦ Οἰκοδομήσατε χάρακα καὶ ἔθεντο χάρακα ἐπὶ τὴν πόλιν α´ καὶ εἶπεν πρὸς δούλους αὐτοῦ Θέτε καὶ ἔθηκαν ἐπὶ τὴν πόλιν

1. The division into elements, and the sequence in which these elements are represented. This modality pertains to segmentation and word order. Both �Rgns and α´ exhibit a preference for analysing the source text word-by-word, matching each constituent with a Greek equivalent, and retaining the order of the parent. They are superficially very similar. The Hebrew verse begins with a ‫ ויהי‬form followed by an adverbial clause comprised of an infinitive (‫ )שמע‬governed by the preposition ‫ כ‬and taking a direct object (‫ )הדבר הזה‬marked by ‫את‬, a construction which establishes a temporal relationship between the action of the infinitival clause, “he heard,” and the matrix clause ‫ויאמר‬, “he said.” Both translations render ‫ ויהי‬by καὶ ἐγένετο, match ‫ כ‬with a temporal particle (ὅτε and ὡς respectively), and replace the infinitive by a finite verb (ἀπεκρίθη and ἤκουσεν). The cost of such literalism is stylistic infelicity (the expression καὶ ἐγένετο is redundant here); nevertheless the two translations successfully represent the temporal relationship marked by the source text. A preference for grammaticality in both versions is evident in their use of a finite verb in the subordinate clause, rather than an infinitive in imitation of the Hebrew. The rendering of ‫ את‬by σύν in α´ is puzzling on this score, as it is apparently ungrammatical. I shall take up this item below (6). 2. The qualitative addition or subtraction of elements. A literal translation will express only those linguistic elements present in the original, and all of them. 53 As a rule α´ prefers neither to add nor subtract elements; at the same time, α´ strongly prefers grammatical renderings. The interplay of these two preferences is evident in the rendering of the second clause in verse 12, where, following the ‫ ויהי‬construction, � continues with a parenthetical comment. Whereas in Hebrew discourse the provision of simultaneous background information is often marked by syndetic nominal clauses, this is not the case in Greek. Therefore the Hebrew participle is rendered in α´ by an imperfect form, reconstruing a nominal clause as a verbal one. �Rgns, on the other hand, retains the participle, but introduces ἦν, likewise recasting the clause. Like α´, the Hellenistic translator shows a preference for rendering the parent quantitatively, but the preference is not as heavily weighted: under certain conditions �Rgns adds or subtracts words and phrases. Thus the conjunction preceding the Hebrew nominal clause, intrusive in Greek, is not rendered; nor is the initial pronoun, which would be redundant; conversely, the ambiguous reference to ‫המלכים‬, “the kings,” is elaborated as πάντες οἱ βασιλεῖς μετ’ αὐτοῦ, “all the kings [who were] with him,” thereby specifying the referent. The result is a text that is at once more cohesive and more coherent than α´. 54

53. Barr, Typology, 304. Barr observes that the measure of equivalence is in part semantic. 54. The distinction between cohesion and coherence relates (loosely speaking) to the difference between form and content. Both are aspects of the connectedness between constituent units which defines a well-formed text. See J. Lyons, Linguistic Semantics, Cambridge 1995, 263-264.

148

2. Towards a Classification of Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

3. Consistency or non-consistency in the rendering. A further modality of literal translation is the degree to which a particular term is used for all (or most) cases of a particular term of the original. 55 The rendering of the Hebrew verb ‫ שים‬in α´ nicely illustrates this phenomenon. In this context ‫ שים‬carries the sense “to take a position (against).” Both occurrences of the verb are rendered by τίθημι in α´, a default for this translator. The match does not do great violence to the semantics of the source text, as τίθημι can carry the sense required by its Hebrew counterpart, yet the Greek is hardly felicitous: one expects the middle form of the verb, whereas α´ has supplied the active; moreover, τίθημι typically collocates with τὰ ὅπλα in such contexts. The Hellenistic translator, while also reliant on default lexical matches, shows a stronger preference for textual coherence. For the Hebrew imperative ‫שימו‬, �Rgns supplies οἰκοδομήσατε and introduces an object, χάρακα. The second instance of the Hebrew verb, ‫וישימו‬, is rendered by the aorist middle form of τίθημι, and the verb is again construed with the object χάρακα. The χάραξ, which denotes a “palisade” here, is first ordered to be “constructed” (Οἰκοδομήσατε χάρακα), and is then “put in place” (καὶ ἔθεντο χάρακα). In this way, the selection of two distinct matches for ‫ שים‬contributes to the overall coherence of the narrative in �Rgns; at the same time, a certain literary impetus is evident: the action has been more vividly described. 4. Accuracy and level of semantic information. This modality of literalism pertains to the degree of overlap between the semantic range of lexical matches (wider or narrower as the case may be), with literality here defined as the correct estimate of the semantic range of the source by the translator. 56 Thus, α´ renders the verb ‫שמע‬, “to hear,” by the active form of a Greek word with a congruent semantic range (ἀκούω). �Rgns, conversely, gives a “non-literal” rendering, replacing the Hebrew word with the aorist passive form of ἀποκρίνομαι, “to answer.” As a result of this rendering the subject of the verb switches. The translator is likely picking up on the use of the same verb in the previous verse (v. 11), giving rise to a more cohesive (if less literal) unit of discourse. 5. Coded indications of formal/semantic relationships obtaining in the vocabulary of the source. Barr speaks of an imitative style of translation, in which the translator aims to represent the form of Hebrew words. 57 Such phenomena include homophonic matches, and those which mark out features of the source language by analogy. More common, however, are renderings which capture the relations between Hebrew forms, especially perceived etymologies. Aquila shows a decided preference for representing such information. This is achieved by translating members of a putative Hebrew word group by members of a corresponding Greek word group. An example in α´ above is the rendering of the Hebrew word ‫ סכות‬by συσκιασμοίς in verse 12. Within the extant Aquilanic version, the word συσκιασμός, “covering,” is matched consistently with ‫סכה‬, “booth.” 58 The Greek noun, a derivative of συσκιάζω, “to throw a shade over,”

55. Barr, Typology, 305-306. Barr notes that consistency in lexical matching is not in itself a sign of literalism, but only when the translator adheres to stereotyping. 56. Barr, Typology, 314. 57. Barr, Typology, 318. 58. Lev 23:42; Deut 31:10; Ps 30/31:21; 60/59:8; Amos 5:26; Isa 1:8. At Ps 26/27:5 it renders ‫סך‬. The 2. Towards a Classification of Literalism

149

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

reflects the etymology of its Hebrew counterpart, which is cognate to ‫סכך‬, “to cover or screen.” 6. Level of text and level of analysis. This refers to the translator’s conception of the source text, including whether his point of departure is its written or vocalized form. For Barr a literal translation adheres to the written form of the source. This category picks out an interesting dimension of Hebrew-Greek translation, but as W. Edward Glenny observes, it does not provide a standard by which one can judge literalism; rather it is a factor which affects it. 59 Under this heading, Barr does, however, make reference to lexically based analysis of the source text, which, as I suggested above (4.), is typologically significant. The rendering of ‫ את‬by σύν in α´ is interesting in this respect. Here ‫ את‬marks the accusative. Since Greek lacks such a marker, α´ normally renders ‫ את‬by the definite article. Where the former occurs together with an article, however, this solution is not possible, and in such cases, α´ provides σύν. As a preposition ‫ את‬typically indicates spatial proximity, and may be glossed “with” or “together with.” The basic semantic relation marked by prepositional ‫ את‬is thus picked up by σύν, but this gives rise to a grammatical anomaly, since σύν is construed exclusively with the dative in Greek. The resulting translation is apparently ill-formed, raising the question of why the preference for grammaticality (otherwise evident in α´) has here been trumped. There are undoubtedly hermeneutic values in play (perhaps related to the Tannaitic context of the translation), 60 but the point is that α´ is representing lexical information present in the source, not only at the expense of the well-formedness of the target text, but also, arguably, in such a way as to transform the semantics of the source (as a modern grammarian would understand it). �Rgns, on the other hand, conveys the semantics of the Hebrew particle precisely in not representing it: its force is captured by the case of τὸν λόγον.

2.2 Tov and Wright’s Taxonomy Barr’s typology has proven a flexible tool for classifying various aspects of translation technique in qualitative terms, greatly enhancing our understanding of the process of translation. Nevertheless the study of literalism in Hebrew-Greek translation typically serves the textual criticism of the Hebrew Bible, and here quantitative methods are wanted—taxonomy rather than typology—as the aim is to make valid inferences regarding the Vorlage. To this end, Emanuel Tov has recast Barr’s modalities as formal criteria for classifying literal renderings, distinguishing five such criteria. 61 word σκηνή, booth, used by �Rgns in the present context to render ‫סכה‬, occurs in α´ as a match for both ‫ אהל‬and ‫משכן‬. 59. W. E. Glenny, Finding Meaning in the Text—Translation Technique and Theology in the Septuagint of Amos, Leiden 2009, 36. 60. See A. Salvesen, “Midrash in Greek? An Exploration of the Versions of Aquila and Symmachus in Exodus” in: J. K. Aitken / K. J. Dell / B. A. Mastin (eds.), On Stone and Scroll: Essays in Honour of Graham Ivor Davies, Göttingen 2011, 523-536, here 535. 61. E. Tov, The Text-Critical Use of the Septuagint in Biblical Research, Jerusalem 1981, 50-66. See also E. Tov / B. G. Wright, “Computer-Assisted Study of the Criteria for Assessing the Literalness of Translation Units in the LXX” Textus 12 (1985), 149-187. B. G. Wright, No Small

150

2. Towards a Classification of Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

1. Consistency. When a translator consistently matches a given Hebrew item with the same Greek equivalent we speak of stereotyping. A thoroughgoing taxonomy will distinguish lexical equivalency from the matching of morphosyntactical and syntactical features. Some ancient Hebrew-Greek translators were more inclined to stereotyping than others, and this difference can be quantified and expressed statistically as a measure of literalism. Tov observes that since a stereotyped rendering may not adequately cover the semantic range of its Hebrew counterpart, a high degree of consistency can result in transfer. 62 As I will argue below (§ 3.2), a translator’s tolerance for such interference is a distinct variable in the analysis of translation technique. 2. Representation of constituents of Hebrew words by individual Greek equivalents. Barr fails to make a principled distinction between his first and second modalities, segmentation and quantitative representation; Tov, on the other hand, differentiates the two variables. 63 The present criterion is defined as the division of the words of the source into their constituent parts in order to represent each in the translation. 64 A striking example is the rendering of ‫ ב‬by ἐν when the Hebrew item is bound to an infinitive and carries a temporal function, an equivalency transferred by analogy from the occurrence of the preposition with nouns. 65 3. Word order. Adherence to the word order of the source is not treated as a separate category by Barr, as he attributes the phenomenon to habit and easy technique, and in this he is followed by Aejmelaeus. 66 Tov is, however, correct to regard it as a distinct criterion. 67 Work by both Dines and Glenny, for example, suggests a conscious effort on the part of the translator of Greek Amos to follow the word order of his Hebrew source. 68 Glenny reports only seven transpositions within the entire translation (six of these are minor stylistic changes; the seventh involves a very difficult text). Given the flexibility of word order in compositional Greek, and the differences between the conventions of Greek and Hebrew word-order, Glenny concludes that one would not expect this degree of agreement if the translator was merely employing an easy technique. 69 4. Quantitative representation. Tov defines this as the representation of each element by one equivalent element, over against the addition or subtraction of elements. 70 Benjamin Wright nuances the variable slightly, defining it as the one-to-one

62. 63. 64. 65. 66. 67. 68. 69. 70.

Difference: Sirach’s Relationship to Its Hebrew Parent Text (Septuagint and Cognate Studies 26), Atlanta, GA 1989, employs these criteria in an important investigation aimed at determining the degree to which one could recover the Vorlage of Sirach by retroversion. Tov, Text-Critical Use, 56-57. Glenny, Finding Meaning, 35. Tov, Text-Critical Use, 57. See also Wright, No Small Difference, 55-56. Soisalon-Soininen, Die Infinitive, 80. Aejmelaeus, “Translation Technique,” 26. Tov, Text-Critical Use, 58. Glenny, Finding Meaning, 45; and J. M. Dines, The Septuagint of Amos: A Study in Interpretation, Ph.D. dissertation, University of London, 1991, 138-139. Glenny, Finding Meaning, 45. Tov, Text-Critical Use, 58. 2. Towards a Classification of Literalism

151

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

representation (or lack thereof) of multi-word phrases, clauses, and sentences. 71 Thus he distinguishes it from the second criterion, which pertains to the division of words. 5. Linguistic adequacy of lexical choices. Tov’s fifth category is problematic—in that “adequacy” is hardly a formal criterion in this context—and rightly drops out of Wright’s analysis. 72 As I shall suggest below (§ 3.2), however, the idea of “adequacy,” understood in terms of the translator’s proclivity towards formal representation of the source, on the one hand, and his tolerance for interference from the source text, on the other, can play a key role in descriptive analysis. The value of taxonomic studies has not gone unchallenged. A fundamental problem with statistical measures is that they fail to address the role of local contextual factors in translation. Furthermore the reliance on purely formal criteria, though methodologically appealing, risks losing sight of the semantics of translation. An important investigation by Martha Wade on the phenomenon of consistency bears this out. As Wade notes, the consistency of lexical matching or concordance is one of the main criteria used in classifying Hebrew-Greek translations. 73 She, however, wants to make a distinction between pseudo-concordance and real concordance. 74 The former refers to the one-to-one correspondence of lexical matches (either from Hebrew to Greek or Greek to Hebrew). According to Wade a statistical investigation of stereotyping, such as that undertaken by Wright on the text of Sirach, measures the degree to which pseudo-correspondence obtains. Real correspondence, on the other hand, occurs when the translator uses the same Greek terms to represent the same meanings derived from the source text. Analysis of this phenomenon obviously involves the judgment of the investigator—one moves away from formal criteria to semantic analysis—yet the results are telling. In a study of Exodus 11–13 that contrasted Wright’s methodology with a contextual analysis, Wade found that what appeared to be a high degree of stereotyping was in large part attributable to real concordance, that is, the translator’s attempt to consistently represent the meaning of Hebrew words in light of other conditioning factors. 75 The question thus arises as to whether a purely quantitative study can provide an accurate picture of translation technique. This is not to say that Tov and Wright have not advanced the discussion. Translation technical studies should undoubtedly include the sort of analysis they outline, one which characterizes the target text in terms of its retention of the formal features of the source, thereby introducing an objective element into the investigation. But there is a sense in which the statistical approach as such has been tried and found wanting. 76 One of its most astute critics has been Aejmelaeus, who argues persuasively that translation technique must be described rather

71. Wright, No Small Difference, 55-56. 72. Tov, Text-Critical Use, 59. Tov notes that the attempt at linguistic precision on the part of the translator often meant that exegetical elements were excluded. 73. Wade, “Evaluating Consistency,” 58. 74. Wade, “Evaluating Consistency,” 58-59. See also S. Olofsson, “Consistency as a Translation Technique” Scandinavian Journal of the Old Testament 6 (1992), 14-30. 75. Wade, “Evaluating Consistency,” 74-75. 76. See R. T. McLay, The Use of the Septuagint in New Testament Research, Grand Rapids, MI 2003, 55-61.

152

2. Towards a Classification of Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

than measured (and described from as many angles as possible). 77 This returns us to Barr’s typology, which stresses the qualitative dimension of the analysis, and which is highly adaptable to changing research priorities. Nevertheless a caveat should be registered here too. The focus on literalism as a technique can lead to a certain myopia, such that the investigator loses sight of the dynamic features of a translation. As Timothy McLay observes, it is the quality and frequency of “non-literal” renderings that are often most interesting. 78 What is wanted is a classificatory scheme that captures the modalities of literalism, but does so in relation to other variables.

3. Beyond Literalism In recent Septuagint scholarship, attention has turned increasingly to the translations as products of Hellenistic Judaism, and therefore as evidence for the social and cultural histories of which they are a part. This shift has affected the way in which the phenomenon of literalism is treated. I shall survey two recent approaches to translation technique, one of which focuses on transformations, the other on the phenomenon of interference. Both are indebted to Barr’s methodological insights.

3.1 Towards the Classification of Transformations Theo A. W. van der Louw has made a persuasive case for a micro-level analysis of ancient Hebrew-Greek translation. 79 His approach, like Barr’s, is fundamentally typological: one isolates the formal and semantic features of the source text and then attempts to categorize their relation to the semantic and formal features of the target text. Yet the focus changes. While Barr classified the distinct ways in which certain features of the source text are preserved under transformation, van der Louw proposes that we classify the transformations. Transformation here refers to the micro-level changes that occur in the transfer from one language to another. Van der Louw argues that without careful study of these changes it is impossible to arrive at reliable conclusions about the translator’s method. 80 By failing to ask why translators employ the transformations they do, one risks mystifying translation technique. Whereas Barr seriously doubted whether typological analysis could be applied to “free” translation, van der Louw demonstrates how it can be done. The key is to treat “literal” translation as a default. On this assumption, every transformation implies a rejected literal rendering. 81 To categorize a transformation, one identifies the literal 77. Aejmelaeus, “What We Talk About,” 217. 78. McLay, Use of the Septuagint, 57-58. 79. See T. van der Louw, Transformations in the Septuagint: Towards an Interaction of Septuagint Studies and Translation Studies (Contributions to Biblical Exegesis and Theology 47), Leuven 2007. See also T. van der Louw, “Linguistic or Ideological Shifts? The Problem-Oriented Study of Transformations as a Methodological Filter” in: A. Voitila / J. Jokiranta (eds.), Scripture in Transition: Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo, Leiden 2008, 107-125. 80. van der Louw, “Linguistic or Ideological Shifts,” 109-110. 81. van der Louw, “Linguistic or Ideological Shifts,” 111. 3. Beyond Literalism

153

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

translation that was not chosen and hypothesizes the rationale underlying the rejection, which may involve the claims of the target language (obligatory transformations), but also such factors as style, logic, communicative purpose, culture, and ideology. Each category of transformation represents the solution to a distinct type of problem arising within literal translation. At the risk of simplification, I shall endeavour to illustrate van der Louw’s classificatory scheme through a tentative analysis the Septuagint version of Esther 2:23 (�Esth). Transformations will be categorized by reference to a hypothetical “literal” translation based upon equivalencies attested in other HebrewGreek translations (�Lit). I shall again have recourse to the notion of preference, although it is not used by van der Louw. � ‫ויבקש הדבר וימצא‬ �Esth ὁ δὲ βασιλεὺς ἤτασεν τοὺς δύο εὐνούχους �Lit καὶ ἐζήτησεν τὸ ῥῆμα καὶ εὗρεν

‫ויתלו שניהם על עץ‬ � �Esth καὶ ἐκρέμασεν αὐτούς �Lit καὶ ἐκρεμάσθησαν ἀμφότεροι ἐπὶ ξύλου ‫ויכתב בספר דברי הימים לפני המלך‬ � �Esth καὶ προσέταξεν ὁ βασιλεὺς καταχωρίσαι εἰς μνημόσυνον ἐν τῇ βασιλικῃ βιβλιοθήκῃ ὑπὲρ τῆς εὐνοίας Μαρδοχαίου ἐν ἐγκωμίῳ �Lit καὶ ἔγραφη ἐν βιβλίῳ λόγων τῶν ἡμερῶν κατὰ πρόσωπον τοῦ βασιλέως 1. Style. A literal translation may be rejected due to a preference for one that is more acceptable (with respect to the conventions of the target language) either from a linguistic or literary viewpoint. Stylistic transformations often involve the avoidance of Hebrew idiom in favor of a more natural expression. Thus (ex hypothesi) in the third clause �Esth rejects the prepositional phrase ἐπὶ ξύλου (�Lit), “on a pole,” as a modifier of κρεμάννυμι, “to hang,” preferring the economical construction καὶ ἐκρέμασεν αὐτούς, “and he hung them” (or “he crucified them”). 2. Coherence. Where there is a preference for thematic coherence, a literal translation that is stylistically acceptable may be rejected in favor of one in which the constituent elements are more closely connected, whether formally or semantically. �Esth thus makes the subject of the first clause (ὁ βασιλεύς, “the king”) explicit, as the king is now the thematic participant of the narrative. The use of the conjunction δέ rather than καί (�Lit) may be intended to mark a thematic development. Note that the agency attributed to the king in the opening clause is carried right through the pericope. Hence the passive forms ἐκρεμάσθησαν (�Lit), “they were hung,” and ἔγραφη, “it was written” (�Lit), are rejected (ex hypothesi) in favor of active verbs construed with the king as subject, ἐκρέμασεν, “he hung,” and προσέταξεν […] καταχωρίσαι, “he ordered […] to make an entry.” The verb καταχωρίζω carries the precise sense required for this context, “set down in a book,” or “place on record.” 3. Communicative purpose. A coherent translation may yet fail to communicate effectively, and where there is a preference for the latter it will undergo a transformation. �Esth thus rejects (ex hypothesi) the oblique καὶ ἐζήτησεν τὸν λόγον καὶ εὗρεν (�Lit), “he investigated the matter and discovered,” in favor of ἤτασεν τοὺς δύο εὐ154

3. Beyond Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

νούχους, “he interrogated the two eunuchs.” The eunuchs are re-introduced into the narrative, neatly anticipating the next clause. 4. Culture and ideology. �Esth rejects (ex hypothesi) the spare phrasing of καὶ ἔγραφη ἐν βιβλίῳ (�Lit), “and it was written in the book,” in favor of a rendering with two elaborations, ὑπὲρ τῆς εὐνοίας Μαρδοχαίου, “concerning Mardochaios’ loyalty,” and ἐν ἐγκωμίῳ, “in an encomium.” As Claudine Cavalier observes, both phrases accord political importance to the episode by echoing vocabulary designating the official status of a benefactor of the king in the Persian court. 82 The translator of �Esth shows a marked preference for such renderings. It is evident that van der Louw’s work on transformations complements Barr’s typology of literalism, filling a gap in the study of translation technique. His proposal is undoubtedly important. In reckoning with literalism in ancient translation, one needs to account for those instances in which the option of a “literal” rendering is not exercised. In so doing, one can begin to characterize the translation with respect to the norms which underlay it. But while van der Louw’s typology is useful, his methodology raises a number of issues. As I have indicated, it rests on the a priori assumption that “literal” translation is a default. This is problematic in a number of respects, especially since van der Louw never adequately defines “literal.” The result is a certain arbitrariness and subjectivity in his categorization. Moreover he seems to take literalism as an unproblematic notion. Yet surely Barr’s analysis has demonstrated that it is not simply a given. Van der Louw acknowledges the role of translational norms in motivating transformations, but does not adequately address the possibility that such norms play a role in literalism as well, i. e. motivating certain types of formal equivalence and not others. Perhaps the underlying problem is that although van der Louw’s ostensible focus is on the target text, his methodology remains somewhat source oriented. These issues are not fatal to his research program; they do, however, underscore the need for further enquiry.

3.2 Literalism and the Phenomenon of Interference Published together with the New English Translation of the Septuagint in 2007 was a document entitled “To the Reader of NETS” in which the editors laid out their rationale, the so-called interlinear paradigm. 83 The paradigm has proven controversial, due in part to its association with a specific theory of Septuagint origins, which locates the translation within a school environment. 84 Here I wish to prescind from the question of origins, and address the implications of the paradigm for the classification of translations with respect to their literalism. To frame the discussion, I shall draw on the work of Gideon Toury, who pioneered a target-oriented research program for the

82. C. Cavalier, Esther (Bible d’Alexandrie 12), Paris 2012, 157. 83. A. Pietersma / B. Wright, “To the Reader of NETS” in: A. Pietersma / B. Wright (eds.), A New English Translation of the Septuagint, New York, NY/Oxford 2007, xiii-xx. 84. See for instance J. Joosten, “Reflections on the Interlinear Paradigm in Septuagint Studies” in: A. Voitila / J. Jokiranta (eds.), Scripture in Transition: Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo, Leiden 2008, 163-178, here 171. 3. Beyond Literalism

155

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

study of translation (Descriptive Translation Studies or DTS). 85 According to Pietersma, DTS provides Septuagint Studies with nothing short of “a new Archimedean point.” 86 Toury’s position is encapsulated in the dictum that a translation is “a fact of the culture that produced it.” 87 The tendency in modern linguistically informed studies of translation, Toury argues, has been to focus on the representation of the source text by the translator. With the shift of perspective to the target culture, one examines the translation in relation to the conventional practices of the literary system within which it was produced. What distinguishes translation as a cultural practice is the phenomenon of interference, that is, its tension with the linguistic, textual-linguistic and literary-cultural norms of the target culture. Depending upon the cultural location of the translator, and the assumptions and practices underlying his work, he will be more or less permissive towards interference, and the nature and degree of his literalism will vary accordingly. On this view, the interlinear paradigm for Septuagint Studies may be seen as a theory of interference. Simply put it holds that a strategy of formal equivalency was normative for many ancient Hebrew-Greek translators, resulting in texts characterized by a high degree of transfer from the source language. The word-picture of interlinearity offers a way of conceptualizing the degree and kind of transfer that obtained. 88 According to the paradigm, one is to imagine a vertical relationship between target and source, such that the formal features of the source text placed constraints on the production of the target text. Where the translator exhibits a strong preference for such constraints over against the claims of the target language, textual well-formedness is sacrificed and some degree of unintelligibility is inevitable. This phenomenon is of theoretical import, as it suggests that certain equivalency preferences had a normative force for the translator, such that all else being equal they trumped the conventions of the target language. As Pietersma stresses, it is crucial not to treat such unintelligibility as an incidental feature of the text; rather, it is an inherent or constitutive feature of any translation in which formal equivalency is normative. 89 Within the framework of DTS, and guided by the metaphor of interlinearity, an integrated approach to the classification of literalism presents itself. Taking up the notion of preference values, one begins by identifying which equivalencies are characteristic of the translator (e. g. “match word-for-word,” “match for word class,” “match for word order,” “match consistently”); at the same time, one considers the weight of these preferences over against the conventions of the target language (to the extent that they can be inferred), i. e. its grammar, preference rules, stylistic norms, and lit85. G. Toury, Descriptive Translation Studies and beyond (revised edition), Amsterdam 2012 [1995]. 86. See A. Pietersma, “LXX and DTS: A New Archimedean Point for Septuagint Studies” in: C. Boyd-Taylor (ed.), A Question of Methodology—Albert Pietersma, Collected Essays on the Septuagint (Biblical Tools and Studies 14), Leuven 2013, 273-282, here 282. 87. Toury, Descriptive Translation Studies, 21. 88. Interlinearity may also be understood as a claim regarding the concept of equivalency underlying a given Hebrew-Greek translation. See Boyd-Taylor, Reading between the Lines, 99. 89. A. Pietersma, “Beyond Literalism: Interlinearity Revisited” in: Boyd-Taylor, Question of Methodology, 359-378, here 362.

156

3. Beyond Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

erary features. One asks two basic questions. First, to what extent and in what manner does a given translator favor formal equivalency over linguistic, textual, and literary well-formedness? Second, under what conditions does he favor such well-formedness over formal equivalency? Having addressed these questions, one then classifies the translation relative to two poles: adequacy, i. e. the accommodation of the target language to formal features of the source text, and acceptability, i. e. the assimilation of features of the source text to target conventions. For each pole one distinguishes between the linguistic, textual-linguistic, and literary-cultural levels of analysis. To sketch out this approach in crude terms, I shall use a handful of examples drawn from the Old Greek Psalter (�Ps). 1. Adequacy. The process under investigation is that of accommodation. As a result of the translator’s equivalency preferences, certain features of the source text will remain invariant under transformation, giving rise to interference. Toury distinguishes two basic types: i) positive transfer, which pertains to anomalies in the distribution of linguistic features in a translation relative to non-translational literature; ii) negative transfer, pertaining to features that do not normally occur in the target language. The burden of proof lies with the investigator, i. e. to demonstrate accomodation, he or she must make a case for interference at one of the three levels of analysis. 1.1. Linguistic adequacy. Linguistic interference pertains to the grammar and lexicon. The translator of �Ps tends to adhere to the formal features of the parent (within the bounds of grammaticality), hence examples of both positive and negative transfer abound. An example of the latter occurs in the rendering of the verb ‫ הגה‬by μελετάω. 90 In five instances the Hebrew verb is construed with a prepositional phrase in ‫ ב‬as complement, which is mirrored in �Ps with a prepositional phrase in ἐν (once in εἰς). 91 In Greek composition, however, the verb μελετάω regularly takes the accusative. Thus, as Pietersma notes, we may conclude that the usage of the Psalter (i. e. μελετάω + ἐν) is due to a tolerance for interference. 92 As it happens, the translator typically matches prepositions with a default lexical equivalent, irrespective of whether it produces idiomatic Greek. 93 1.2. Textual-linguistic adequacy. Here one is primarily looking at disruptions to the coherence and cohesion of the text, but also its pragmatics, and stylistics. Positive transfer is especially evident at this level in the use of particles. In contrast to nontranslational Greek in which particles are used extensively as cohesive links, �Ps makes little use of them. 94 Two preferences underlie this phenomenon: on the one hand, �Ps tends to match ‫ ו‬with καί (“match consistently”); on the other, �Ps is reluctant to add items (“match for quantity”). The want of particles differentiates �Ps from other trans-

90. The example is drawn from Pietersma, “Beyond Literalism,” 377. 91. The Hebrew verb ‫ הגה‬occurs eleven times in the Psalter, six times with a direct object, and five times with a prepositional phrase in ‫ ב‬as complement (1:2; 62/63:7; 76/77:13; 115:7/113:15; 142/ 143:5. Greek renders ‫ ב‬by ἐν in every instance except 63/62:7, where εἰς is supplied. See Pietersma, “Beyond Literalism,” 377. 92. Pietersma, “Beyond Literalism,” 377. 93. A. Pietersma, “To the Reader of Psalms” in: Pietersma and Wright, New English Translation, 542-547, here 547. 94. Pietersma, “LXX and DTS,” 279. 3. Beyond Literalism

157

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

lations. Thus, οὖν occurs but once in the Psalter, while in the Septuagint version of Genesis there are forty instances, one-third of which lack a counterpart in the source text. 95 The pairing of μέν and δέ, used frequently in compositional Greek to mark a contrast, does not occur in �Ps once. 1.3. Literary and cultural adequacy. Here one hypothesizes the transfer of features of the source text in tension with the literary conventions and ideological codes of the target culture. One feature of �Ps that undoubtedly arises from literary interference is the formal structure of the verse. The translator does not, as far as one can tell, employ the value system of Greek poetics, in which the length of vowels determines scansion. On the other hand, as a result of an isomorphic strategy, the Greek translation exhibits a prose rhythm derivative of the Hebrew parent. A product of literary interference, it nevertheless had a profound cultural impact. Early Christian hymnody (upon which the later Byzantine rhythmic hymns developed), was almost certainly based on the prose rhythms of �Ps. 96 2. Acceptability. Here the process of assimilation is under description. One is classifying modalities of the target system which find expression in the translation through the transformation of formal features of the source. The burden of proof shifts accordingly: assimilation to target conventions must be demonstrated against the background of transfer. 2.1. Linguistic acceptability. While the translator of �Ps generally prefers equivalencies that are isomorphic to the source text, he also exhibits a strong preference for grammatical well-formedness. 97 An excellent example of this is found at Ps 29/28:9 where the source reads ‫כלו אמר כבוד‬. 98 The pronominal suffix ‫ ו‬on ‫כל‬, “all,” would have posed a problem for the translator, since his default, an attributive genitive pronoun, would not work in this syntactical context. So instead he added τις, producing the fine Greek idiom πᾶς τις, “every last one.” We glimpse here, writes Pietersma, “the translator’s familiarity with standard Greek usage” 99. His tolerance for interference in other contexts is therefore not owing to linguistic deficiency, but rather to his preference for formal equivalency. 2.2. Textual-linguistic acceptability. Here one identifies features of discourse and style introduced by the translator over against the norm of isomorphism. As I have indicated, although particles play an important role in marking cohesion in Greek composition, �Ps uses them sparingly. Yet the conjunction γάρ represents an exception. It is not used as a stereotype, but translates various different Hebrew items; moreover, it matches causal ‫ כי‬on seven occasions, over against 360 cases of ὅτι, and is thus a marked feature for the Greek text; in eleven instances there is no Hebrew counterpart. 100 It would thus appear that γάρ is a feature of discourse introduced by �Ps in order to enhance the cohesion of the translation. Ps 9:35, though the reading may be secondary. See Pietersma, “LXX and DTS,” 279. C. A. Trypanis, Greek Poetry from Homer to Seferis, London 1981, 414. Boyd-Taylor, Reading between the Lines, 265. The example is drawn from A. Pietersma, “Messianism and the Greek Psalter: In Search of the Messiah” in: Boyd-Taylor, Question of Methodology, 243-270, here 260-261. 99. Pietersma, “Messianism,” 260. 100. Pietersma, “LXX and DTS,” 279. 95. 96. 97. 98.

158

3. Beyond Literalism

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

2.3. Literary and cultural acceptability. This refers to assimilation of features of the source text to forms consistent with literary and thematic values within the target culture. Obviously this level of analysis is highly conjectural, yet in some instances the strategy of the translator is to a degree transparent. Thus whereas the Hebrew Psalter sometimes refers to God as a ‫צור‬, “rock or cliff,” or ‫מעוז‬, “stronghold or fortress,” both figures likening God to a secure place of refuge, �Ps avoids rendering the metaphor literally. 101 For example in such contexts ‫ צור‬is matched with βοηθός, “helper” (17:3/ 18:2; 18/19:15; 77/78:35), and θεός, “God” (17/18:32,47; 27/28:1; 30/31:3); ‫ מעוז‬is matched with ὑπερασπιστής, “protector” (27/28:8; 30/31:3); κραταίωμα, “strength” (42/43:2); and βοηθός, “helper” (51/52:9). Karen Jobes and Moíses Silva suggest that the pattern may reflect a religious taboo of the Hellenistic period. 102

3.3 The Translator’s Dilemma Within DTS Schleiermacher’s “either/or” has become a “both/and,” but the terms of the binary model still dog descriptive analysis of the text in the opposition of acceptability and adequacy. Perhaps this is logically unavoidable. Consider the following French translations of the English sentence: E1 The first word of this sentence has three letters. F1 Le premier mot de cette phrase a trois lettres. F2 Le premier mot de cette phrase a deux lettres.

In one respect F1 is a literal rendering. Unlike F2, it states the same proposition as its source. It also matches each English word with a conventional equivalent. Yet while E1 is true, F1 is false, as it does not take into account the self-referential character of the statement, and thereby violates its truth conditions. Only by replacing “three” by “deux” is this avoided. All translators are faced with such challenges. As Barr realized the dilemma presents itself in distinct modalities: there is not one “either/or” but many, and their resolution will depend upon the translator’s preference values. By classifying these preferences we gain conceptual purchase on how the translator operated. Within the field Septuagint Studies the classification of ancient Hebrew-Greek translations has generally involved recourse to some concept of literalism. As I have suggested, there are persuasive reasons for continuing to do so. No doubt other methodologies will present themselves, but if the primary aim of our analysis is to characterize the process of translation underlying the Greek text, then we will not want to lose sight of how the formal features of its source are manipulated by the translator. This dimension of the translation not only has a decisive role in linguistic analysis,

101. K. H. Jobes / M. Silva, Invitation to the Septuagint, Grand Rapids, MI 2000, 95. See also S. Olofsson, God Is My Rock: A Study of Translation Technique and Theological Exegesis in the Septuagint (CBOTS 31), Stockholm 1990, 55. 102. Jobes / Silva, Invitation, 96. See I. L. Seeligmann, The Septuagint Version of Isaiah: A Discussion of its Problems, Leiden 1948, 100. 3. Beyond Literalism

159

9. The Classification of Literalism in Ancient Hebrew-Greek Translation

which must take the phenomenon of interference into consideration, but also in our interpretation of the text as a fact of the culture that produced it, both as the rendering of a Hebrew source, and as a work in its own right. 103

103. See Boyd-Taylor, Reading between the Lines, 431-438.

160

3. Beyond Literalism

10. The Study of Translation Technique Raija Sollamo

Translation technique study has established itself as a significant approach to the study of the Septuagint. This chapter first defines the term and summarizes its development. Second, it describes translation technique as a method of inquiry that opens up the rich world of the Septuagint by considering it as a translation that must be appreciated in its own right. Third, it demonstrates the essential contribution of translation technique studies to such specialized fields as the textual criticism of the Hebrew Bible, the study of Septuagint syntax, and the theology of the translators. Fourth, it presents a sample of some of the most important results produced by translation technique studies.

1. The Term and Its History The study of translation technique is one of the main fields of modern Septuagint research. The term translation technique was introduced by Ilmari Soisalon-Soininen in his doctoral thesis, Die Textformen des Richterbuches (1951), in which, following the advice of his advisor, Gillis Gerleman of the University of Lund, Sweden, 1 he compared the A- and B-texts of Judges, noting differences in the translations. To describe his approach he not only used such terms as “übersetzungstechnische Eigenheiten,” 2 “die übersetzungstechnische Übereinstimmung der verschiedenen Textformen,” 3 and “übersetzungstechnische Untersuchung,” 4, but also “Übersetzungstechnik.” 5 In his second book, Der Charakter der asterisierten Zusätze in der Septuaginta (1959), SoisalonSoininen examined the additions that Origen made to the Septuagint column of his Hexapla on the basis of the longer Hebrew Vorlage at his disposal. 6 However, it was in his monograph Die Infinitive in der Septuaginta (1965) 7 that Soisalon-Soininen de1. 2. 3. 4. 5.

6. 7.

For more details, see R. Sollamo, “The origins of LXX Studies in Finland” SJOT 10 (1996), 159168. I. Soisalon-Soininen, Die Textformen der Septuaginta-Übersetzung des Richterbuches (AASF Series B 72.1), Helsinki 1951, 14. Soisalon-Soininen, Die Textformen, 23. Soisalon-Soininen, Die Textformen, 25. He calls his doctoral thesis “a translation technical study”. Soisalon-Soininen, Die Textformen, 24. This appears to be the first time he used the term “translation technique” in his scholarly production. At least one reviewer saw the main merit of the book as consisting of his materials and observations on translation technique. See S. Segert, “Semitistische Marginalien I” ArOr 29 (1961), 87-89. I. Soisalon-Soininen, Der Charakter der asterisierten Zusätze in der Septuaginta (AASF Series B 114), Helsinki 1959, 46-160. I. Soisalon-Soininen, Die Infinitive in der Septuaginta. (AASF Series B 132.1), Helsinki 1965. 1. The Term and Its History

161

10. The Study of Translation Technique

fined, for the first time, the term “translation technique” (“Übersetzungstechnik”) as a new methodological approach to the Septuagint. 8 This seminal volume became the standard international reference work for the study of translation technique. Accordingly, Soisalon-Soininen deserves to be known as the founder of the study of translation technique in modern Septuagint research. Of course, Septuagint Greek and translation have been studied by numerous scholars throughout the centuries, but what Soisalon-Soininen accomplished was a methodological breakthrough. Translation technique was, for Soisalon-Soininen, both a research object and a research method. For him, the two could not be separated. Soisalon-Soininen’s research interest was primarily focused on the study of Septuagint syntax. 9 He soon realized however that the syntax of a translation cannot be properly analyzed without considering the syntax of the source language, the particularities of the target language, and the translator’s relation to both. Indeed, how a translator translated seemed to depend on the degree to which he felt obliged to follow the wording of his source text, i. e., the Hebrew Vorlage he was using. This appeared to be the main reason for differences between Septuagint syntax and Greek syntax as manifested in contemporary original Greek literature. To be sure, the Greek competence of the translators varied considerably. Soisalon-Soininen emphasized the importance of comparing different translations and different translators, taking the source text as a point of departure; for only those renderings that correspond to the same Hebrew expressions, grammatical forms, or syntactical constructions can shed light on translation technique. Only in this way was it possible to demonstrate true differences in translation technique. The study of translation technique seeks to describe how translators customarily work when they translate Hebrew into Greek. Soisalon-Soininen was fully aware that translators did not randomly select equivalents, but, on the contrary, instinctively chose identical renderings in similar cases with a high degree of consistency, 10 a phenomenon that had nothing to do with modern computers executing predetermined program codes. On the other hand, Septuagint translators sporadically rendered some Hebrew constructions very freely, using idiomatic Greek expressions in a way that contrasted with their customary literalism. A certain degree of variation was, of course, occasioned by context and particularity. Moreover, the degree of constancy varied to a certain extent from one individual to another. Soisalon-Soininen was convinced that the translator was not aware of using a translation technique. Indeed, the presence of a translation technique can only be shown by examining the final product, the translation. The translation equivalent pairs that the study of translation technique reveals in a given translation reflect the way the translator’s mind worked. Cognitive scientists could perhaps explain these patterns as the result of the architecture of the human 8. The term “Übersetzungstechnik” possibly derived from Peter Katz (Walters), whom SoisalonSoininen had consulted in Britain after the war. Katz’s article “Zur Übersetzungstechnik der Septuaginta” appeared in 1956 (WO II, 267-273). 9. Soisalon-Soininen, Die Infinitive, 12-15. 10. I. Soisalon-Soininen, “Methodologische Fragen der Erforschung der Septuaginta-Syntax” in: I. Soisalon-Soininen, Studien zur Septuaginta-Syntax (edited by A. Aejmelaeus and R. Sollamo, AASF B, 237), Helsinki 1987, 40-52.

162

1. The Term and Its History

10. The Study of Translation Technique

brain. But it is important to note that Soisalon-Soininen only applied his methods to the study of the Septuagint, a corpus translated in Alexandria during the last centuries BCE. Today, translation has been revolutionized by the development of new tools (lexica, concordances, computers, etc.) and contemporary translation studies now shift the focus from the customary study of translation technique to identifying translation universals common to all translations regardless of the source or target languages. 11

2. The Method Since the pioneering studies of Soisalon-Soininen, translation technique study has been associated with the Septuagint school at the University of Helsinki, and rightly so, since his disciples (Raija Sollamo and Anneli Aejmelaeus) and their students (Anssi Voitila and Seppo Sipilä) have continued developing this methodological approach. Increasingly however, scholars around the world have adopted the method, applying its procedures to their own work (Jan Joosten, Arie van der Kooij, Takamitsu Muraoka, Staffan Olofsson, Emanuel Tov, to name but a few). 12 Translation technique study presents three principal advantages: it makes it possible to treat the Septuagint as a translation, to evaluate the quality of Septuagint Greek, and to characterize the translations. I will discuss each of these points in turn.

2.1 The Septuagint as Translation Because translation technique study takes seriously the fact that the Septuagint is a translation, the source text and language form its point of departure. Without this methodological foundation, it would be impossible to correctly understand how the translators worked. Translation technique was originally developed in order to further the study of Septuagint syntax. It is therefore better suited to syntactical analysis than to lexical or phonological studies. It can however be used for lexicographical studies if context and semantics are taken into account in such a way that different fields of meaning are considered. It is important to bear in mind that the purpose of translation 11. R. Sollamo, “Translation Technique and Translation Studies: The Problem of Translation Universals” in: M. K. H. Peters (ed.), XIII Congress of the International Organization for Septuagint and Cognate Studies (Septuagint and Cognate Studies 55), Atlanta, GA 2008, 339-351; A. Chestermann, “Hypotheses about Translation Universals” in: G. Hansen / K. Malmkjaer / D. Gile (eds.), Claims, Changes and Challenges in Translation Studies, Amsterdam 2004, 1-13; and G. Toury, “Probabilistic Explanations in Translation Studies: Welcome as They Are, Would They Qualify as Universals?” in: A. Mauranen / P. Kujanmäki (eds.), Translation Universals: Do They Exist?, Amsterdam 2004, 15-32. 12. For the study of translation technique see E. Tov, “The Nature and Study of the Translation Technique of the LXX in the Past and Present” in: C. E. Cox (ed.), VI Congress of the International Organization for the Septuagint and Cognate Studies, Atlanta, GA 1987, 337-359; and B. Lemmelijn, “Two Methodological Trails in Recent Studies on the Translation Technique of the Septuagint” in: R. Sollamo / S. Sipilä (eds.), Helsinki Perspectives on the Translation Technique of the Septuagint, Helsinki/Göttingen 2001, 43-63. 2. The Method

163

10. The Study of Translation Technique

technique study is not to compile statistics on translation equivalents, but to identify variant renderings and propose explanations for these choices. It is akin to being a detective on the trail of the Septuagint translators, discovering how they chose their translation equivalents. 13 The aim is not to criticize them, but to understand the mental processes that generated the renderings. Knowledge of the target language is indispensable for understanding how the translators interpreted their Vorlage. The most promising Hebrew expressions or constructions for translation technical investigation are those that diverge significantly from Greek language and idiom and, in addition, occur with great frequency. In such cases, the translation technical differences between translators come more clearly to the fore. Randomly selected materials do not constitute a sufficiently solid basis for the analysis of translation technique. If one simply examined a chapter or two here and there, the results could be misleading, as the occurrences of given equivalents are not necessarily distributed evenly throughout an entire book. It is therefore necessary to examine all the occurrences of different equivalents for relevant Hebrew constructions and locutions in a given book before proposing explanations for the choices made by the translator. Translation technical scholars begin by classifying all relevant occurrences into appropriate syntactic or semantic categories. Of course, the translator did not proceed in such a systematic way. Instead, his language competence instinctively instructed him how to translate Hebrew expressions, constructions, or words belonging to these categories. Frequently these categories are not evenly represented. In my study of the renderings of the semiprepositions in the Septuagint, I would not have noticed that, from the point of view of the Greek, the preposition ‫ ִלְפ ֵני‬constituted a distinct category when used to refer to objects or places had I not examined all the books of the Septuagint. 14 Only in this way was I able to accumulate enough material to detect this special group. Since constant equivalents such as ἐνώπιον, ἐναντίον, and ἔναντι were used only to refer to humans and other living beings, in the sense “in the presence of, before,” other equivalents were needed when referring to objects and places in front of which something or someone was placed or situated. The instances of ‫ ִלְפ ֵני‬in a temporal sense also formed a special category in Greek, because most of the locative equivalents could not be used to connate a temporal meaning. For a translation technical study, ‫ ִלְפ ֵני‬and its Greek equivalents provided particularly appropriate materials not only because translators had a number of equivalents from which to choose, ranging from slavish to literal to free (e. g., πρὸ προσώπου, κατὰ πρόσωπον, ἐνώπιον, ἐναντίον, ἔναντι, κατέναντι, ἀπέναντι, ἔμπροσθεν, πρό, πρότερος, πρίν [ἤ], πρός + accusative or dative, παρά + dative, etc.), but also because the preposition ‫ ִלְפ ֵני‬was used in a variety of semantic fields and after verbs in such a way that the verbal rection and preverbs in Greek became interesting (e. g., προσκυνοῦσιν τῷ κυρίῳ 1Sam 1:19, προπορεύεται πρότερος + genitive Deut 1:33). As for etymology and grammar, the preposition alternated between a prepositional 13. A. Aejmelaeus, On the Trail of the Septuagint Translators: Collected Essays (CBET 50), Leuven 2007. 14. R. Sollamo, Renderings of the Hebrew Semiprepositions in the Septuagint (AASF, Diss. Hum. Litt. 19), Helsinki 1979, 13-80.

164

2. The Method

10. The Study of Translation Technique

metaphoric phrase and a new grammaticalized preposition. The most slavish translators preserved the noun πρόσωπον in their equivalents, while the literal translators looked for a more suitable Greek expression, using analogical structures such as ἐνώπιον, which consists of the preposition ἐν and the root -ώπ-, the same root as in the middle of πρόσ-ωπ-ον. The prepositions ἐναντίον, ἔναντι, and ἔμπροσθεν also represent attempts toward an analogical structure. The freest translations, which appeared sporadically, used ordinary Greek prepositions or, more simply, the case taken by the preceding verb. This variation in Greek renderings shows that the translators understood ‫ ִלְפ ֵני‬differently according to the textual circumstances and context. Generally speaking, each translator had a favorite equivalent, but to a certain extent his renderings varied from case to case. Translations can be categorized as slavish or literal or free depending on which types of renderings predominate in a given case. Still, even in books where highly slavish equivalents prevail, very free renderings occasionally appear.

2.2 The Quality of Septuagint Greek Translation technique study makes it possible to evaluate the quality of the Greek language used in the Septuagint. This is particularly important because the translators conceived of their task in different ways. Some strive for as literal a translation as possible; others favor idiomatic language and good Greek style as a means of remaining faithful to the original; still others paraphrase continuously. Often it is not easy to determine if the expressions produced by the translator are good Greek or not. In the case of ‫ִלְפ ֵני‬, the equivalent πρὸ προσώπου would seem to be understandable Greek: the preposition πρό takes a genitive and the expression therefore means “before the face of.” However, a comparison with contemporary Greek documents showed that this locution was never used in non-translational koine literature or in Ptolemaic papyri or inscriptions. I was therefore able to identify it as a Hebraism invented by the translators. All the other equivalents of ‫ ִלְפ ֵני‬can be found in Greek sources, although they are infrequent. For instance, before it was raised to a new literary level by the Septuagint translators, ἐνώπιον was only attested in Egyptian papyri. The preposition ἐναντίον, which belongs to a literary register, was attested not only in Ptolemaic papyri and inscriptions but also in Polybius’ Histories. 15 In general, the most slavish translations are written in Hebraistic or poor Greek, but this is not automatically the case: κατὰ πρόσωπον was good Greek, even though it was mostly used in Greek sources as an independent adverbial, not as a prepositional construction followed by a genitive. The use of prepositions and prepositional phrases taken from Greek literature does not, by itself, constitute a criterion of the quality of the Greek in Septuagint books. Rather, their frequency in the translation must be compared with their frequency in original Greek sources. In the case of semiprepositions, frequencies vary significantly. While such prepositions as ἐνώπιον, ἐναντίον, and ἔναντι occur infrequently in Greek sources, they abound in the translated Greek of the Septuagint. This disproportion is due to the practice of imitating Hebrew constructions and locutions 15. R. Sollamo, Semiprepositions, 328, 311-317. 2. The Method

165

10. The Study of Translation Technique

that gave the Septuagint Greek a Hebraistic flavor. In addition, the overrepresentation of certain expressions restricts the possibility of genuine Greek idioms gaining ground in the translation. Another enlightening example is the infrequent use of the verb ἔχειν in the Septuagint. 16 This anomaly can be explained by the absence of a corresponding verb in Hebrew. Instead, the Septuagint translators used the verb εἶναι with a dative, imitating the Hebrew construction ‫ ְל‬+ ‫ָה ָיה‬. Let us take another example of the difficulty of recognizing what is idiomatic Greek and what is not. It is undeniable that the repetition of pronominal genitives, either before or after the coordinate items, is characteristic of Hebraistic translation Greek. But in order to go beyond this observation it is necessary to know where the pronominal genitive is customarily placed in idiomatic Greek: before the two coordinate items, after the first item, or after the second item. Greek grammars were of no avail for answering this question, for it had received no attention prior to the study of Septuagint translation. My comparative survey of original Greek sources indicated that the three positions occur with almost equal frequency. 17 The survey also confirmed that the repetition of the pronominal genitive in connection with two coordinate items—so frequent in the Septuagint—was unattested in original Greek sources. The syntax of the Septuagint abounds in peculiarities that derive from strict adherence to Hebrew syntax.

2.3 Characterization of Translations The comparison of different translators is an important aspect of translation technique study. In order to be significant, such comparisons must be based upon a large number of similar cases in the Hebrew Vorlage. It is hazardous to compare translation equivalents without classifying them according to uses and meanings. The comparison is meaningful only within similar categories. Temporal cases of ‫ִלְפ ֵני‬, for instance, are comparable, but temporal and locative cases of the same preposition are not. Classification requires common sense. If the categories are too restrictive, there will be too few occurrences in each category for comparison. Statistics are only useful if they are correctly compiled. The consistency of the translators varies considerably. General statistics show that translators have a constant or favorite way of translating a term, which they use automatically in most cases. Here, the preliminary intuitions of Soisalon-Soininen have proved to be true. As almost any instance will show, the most common rendering of a syntactic feature usually accounts for 50 to 90 % of the cases. The coordinator καί is used in 63 % of the instances of ‫ ו‬in Genesis, the ratio is 72 % in Exodus, 90 % in Leviticus, 90 % in Numbers, and 84 % in Deuteronomy. 18 The genitive form of the personal pronouns, used as a possessive pronoun, is repeated with two coordinate nouns in 51 % of the cases in Genesis, the ratio is 40 % in Exodus, 75 % in Leviticus, 86 % in Numbers, 16. I. Soisalon-Soininen, “Der Gebrauch des Verbs ἔχειν in der Septuaginta” in: Soisalon-Soininen, Studien, 181-188. 17. R. Sollamo, Repetition of the Possessive Pronouns in the Septuagint (Septuagint and Cognate Studies 40), Atlanta, GA 1995, 7-18. 18. A. Aejmelaeus, Parataxis in the Septuagint (AASF 31), Helsinki 1982, 13.

166

2. The Method

10. The Study of Translation Technique

and 76 % in Deuteronomy. 19 Semiprepositions constitute an exception in that translators did not restrict themselves to a single equivalent, but used different ones from time to time. Thus, the percentage for the most common equivalent of ‫ִלְפ ֵני‬, in referring to living beings, varies between 30 % in Joshua and 89 % in Judges (B-text). 20 Quantitative comparison must however be complemented by qualitative comparison. Indeed, the latter is more important than the former. Labels such as “slavish,” “literal,” “free,” or “paraphrasing” are typically used to characterize translations. They can also refer to the translator’s “philosophy of translation.” 21 Viewed from the perspective of the translator, the philosophy preceded and determined the general orientation of the work. But when we study translations, we begin by examining the translator’s technique, only then can we hope to grasp his philosophy of translation, i. e., what kind of translation (literal, free, paraphrasing, interpretative, etc.) he aimed to produce and for whom it was intended. In the present state of Septuagint studies a more adequate description of the respective natures of literalness and freedom is a desideratum, much in the same way that an understanding of literalness was for James Barr or freedom for Theo van der Louw. 22 It is of vital importance for translation technical studies to learn to recognize differences between translators far better than in the past. Only in this way, can we hope to deal with the thorny issue that arises when two books or two parts of a single book diverge so drastically from one another that they cannot have been produced by the same translator.

3. The Aims of Translation Technique Study A translation technical approach is essential for the study of Septuagint syntax, the textual criticism of the Hebrew Bible, the comparison of different books and translators, and for a better understanding of the individual translators, their translations, and the special character of Septuagint Greek. It also provides a point of comparison for the analysis of Hellenistic Greek in general and New Testament Greek in particular. Recently, translation technique study has proved useful in identifying the theologies, ideologies, and philosophies that underlie the work of the translators. I will discuss three of these issues in more detail: the syntax of the Septuagint, the textual criticism of the Hebrew Bible, and the theologies of the translators.

3.1 Septuagint Syntax Describing the syntax of the Septuagint can no longer be done in the way that Conybeare and Stock went about it at the beginning of the last century. 23 Their grammar R. Sollamo, Repetition, 81-82. R. Sollamo, Semiprepositions, 70-71. S. Jellicoe, The Septuagint and Modern Study, Oxford 1968, 314-315. J. Barr, The Typology of Literalism (MSU XV), Göttingen 1979. T. van der Louw, Transformations in the Septuagint: Towards an Interaction of Septuagint Studies and Translation Studies (CBET 47), Leuven 2007, 57-92. 23. F. C. Conybeare / St. G. Stock, A Grammar of Septuagint Greek (reprint of Selections from the Septuagint, Boston 1905). Grand Rapids, MI 1980. 19. 20. 21. 22.

3. The Aims of Translation Technique Study

167

10. The Study of Translation Technique

consists of a collection of special features of Septuagint syntax. But aside from these particularities, it fails to give a general understanding of the syntax. Conybeare and Stock’s syntax does not inform the reader how frequently the Septuagint peculiarities appear or how they are distributed in the different books. Any syntax of the Septuagint must take seriously that the Septuagint is not a single entity but a collection of different translations made by different translators over a lengthy period of time (from the middle of the third century to the first century BCE). Hence any attempt to create a syntax of the Septuagint must rely upon detailed translation technical studies. 24 In my opinion, writing a syntax of the Septuagint implies describing the general characteristics and the particularities of the syntax of the different books. Such a syntax should inform the reader of syntactic features in each book or group of books and whether they occur frequently, infrequently, or on a par with normal koine syntax. Until a sufficient number of detailed studies on clause connections, tenses, pronouns, prepositions, word order, etc. are done, a comprehensive syntax of the Septuagint will remain out of reach. Such a project has been delayed from one decade to another. Now is the time to begin.

3.2 Textual Criticism Translation technique is an invaluable tool for the textual criticism of the Hebrew Bible. For although there is always a risk factor involved in retro-translation, particularly between dissimilar languages such as Greek and Hebrew, translation technique study provides a reliable means for identifying the Hebrew Vorlage used by the Septuagint translators. 25 This Vorlage can then be compared with the Masoretic text, the Samaritan Pentateuch, and the Qumran manuscripts. The Septuagint has always played an important role in the textual criticism of the Hebrew Bible as a representative of the pre-Masoretic texts. This role would be greatly increased if the complete Hebrew Vorlage of the Septuagint could be reconstructed. To this purpose, the more literal the translation, the greater the degree of reliability of the Vorlage reconstructed with the aid of translation technique method.

3.3 Theologies of the Translators In the present state of Septuagint studies, the ideologies, or theologies, of the translators are fiercely debated between “minimalists” and “maximalists.” 26 Minimalists are very cautious in detecting possible traces of the translator’s theology in the text; 27 max24. For principles of Septuagint syntax, see R. Sollamo, “Prolegomena to the Syntax of the Septuagint” in R. Sollamo / S. Sipilä (eds.), Helsinki Perspectives on the Translation Technique of the Septuagint, Helsinki/Göttingen 2001, 23-41. 25. A. Aejmelaeus, “What Can We Know about the Hebrew Vorlage of the Septuagint” in: Aejmelaeus, Trail, 71-106. 26. M. Knibb, “Introduction” in: M. Knibb (ed.), The Septuagint and Messianism (BETL 195), Louvain 2006, xiii-xxxi. 27. A. Aejmelaeus, “Von Sprache zur Theologie: Methodologische Überlegungen zur Theologie der Septuaginta” in: Knibb, Septuagint and Messianism, 21-48; A. Pietersma, “Messianism and Greek Psalter: In Search of the Messiah” in: Knibb, Septuagint and Messianism, 49-75.

168

3. The Aims of Translation Technique Study

10. The Study of Translation Technique

imalists on the other hand see telltale signs of the translator’s theology in almost every verse. 28 As a general rule, students of translation technique tend to belong to the minimalist party. They have learned to look at the philological side of the translation first. Only once the translation technique of the translator has been properly analyzed does theology enter into the discussion, if indeed it is still relevant. For often a putative theologically pregnant expression is stripped of its theology by translation technique analysis. One example will illustrate this point. In an early (1953) article, my friend John William Wevers wrote on the exegetical principles underlying the Greek text of 2Sam 11:2–1Kings 2:11. Discussing 2Sam 24:14 ἐμπεσοῦμαι δὴ ἐν χειρὶ κυρίου ὅτι πολλοὶ οἱ οἰκτιρμοὶ αὐτοῦ σφόδρα εἰς δὲ χεῖρας ἀνθρώπου οὐ μὴ ἐμπέσω without knowledge of the translator’s translation technique, he quite naturally supposed that the strange phrase ἐν χειρί was due to the translator’s exegetical or theological principles, while its parallel, εἰς χεῖρας, in the same verse did not require comment. Wevers assumed that the use of ἐν χειρί to refer to God “implies a tendency to remove God as far as possible from mankind.” 29 In fact, in the βγ’ section of the KR recension, ἐν χειρί was consistently (eleven occurrences) used to render the Hebrew ‫ְב ַיד‬, which referred equally to human beings or God. On the other hand, εἰς χεῖρας appears only in our verse. Because εἰς χεῖρας was the usual Old Greek expression, the conclusion can be drawn that the KR recension failed to correct εἰς χεῖρας to ἐν χειρί. There was no theology here at all, only philology, in the form of an attempt at as literal a rendering as possible of the Hebrew ‫ְב ַיד‬. As a methodology, the study of translation technique is a sine qua non for detecting possible traces of the Septuagint translators’ theology. Only those elements that do not derive from translation technique or philology can be said to contain theology, in the sense of a theology of the Septuagint distinct from the theology of the source text. Underlining the importance of translation technique does not imply that the translators had no theological motives. It does however warn against drawing hasty conclusions unsupported by solid argumentation. The theology or ideology must be demonstrated, not presupposed. Much more research still needs to be done into the theology of the Septuagint translators. My conviction is that the ideologies, or theologies, of the translators should be seen against the background of the encounter of a conservative religious tradition with a more modern, Hellenistic, cultural environment. It consisted in the cumulative effect of a number of rather small changes in vocabulary that slightly shifted the meaning of the phrases of the Hebrew Vorlage.

28. M. Rösel, Übersetzung als Vollendung der Auslegung. Studien zur Genesis-Septuaginta, Berlin 1994; J. Schaper, Eschatology in the Greek Psalter, Tübingen 1995. 29. 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. 3. The Aims of Translation Technique Study

169

10. The Study of Translation Technique

4. Results The study of translation technique has produced new knowledge about the Septuagint translators and their characteristic ways of translating. The hypothesis that, as a rule, each book of the Septuagint was the work of a single translator has been strengthened. The book of Twelve Prophets is also now considered to be due to a single translator. Earlier theories that postulated the presence, for certain books, of two translators 30 or one translator and a reviser 31 have not found sufficient evidence in recent research. 32 But there is at least one exception to this: the Books of Kings (1–4Reigns). Henry St. John Thackeray divided the text into five major sections. These were subsequently adopted by Dominique Barthélemy, 33 who was the first to identify the text types as the Old Greek translation and the Kaíge Recension (KR). A subsequent study by James Donald Shenkel then revealed further characteristics of the two translation techniques that he attributed respectively to the OG and KR sections. 34 Old Greek

KR

1Rg 2Rg 1:1–11:1

2Rg 11:2–3Rg 2:11

3Rg 2:12–21:43

3Rg 22:1–4Rg 25:30

One of the most significant results of translation technique study is that the books of the Septuagint can be divided into three categories or types of translations: slavish, literal, and free. On a scale ranging from slavish to literal to free, the most slavish translations are those of Ecclesiastes, 2Esdras, Judges A and B, Jeremiah, Ezekiel, and the KR sections of 1-4Reigns. At the other end of the scale, the freest are Esther o’, Proverbs, Job, Isaiah, and Daniel o’, followed by the books of the Pentateuch, most notably Exodus and Genesis, and by Joshua. The remaining books could be said to have been translated literally, but not slavishly. 35 30. E. g. H. St. J. Thackeray, “The Greek Translators of Ezekiel” JTS 4 (1903), 387-397; H. St. J. Thackeray, “The Greek Translators of Jeremiah” JTS 4 (1903), 245-266; H. St. J. Thackeray, “The Greek Translators of the Prophetical Books” JTS 4 (1903), 578-585. F. Baumgärtel / J. Herrmann, “Die Septuaginta zum Zwölfprophetenbuch das Werk zweier Übersetzer” BWAT NF 5 (1923), 32-38. O. J. Baab, “A Theory of Two Translators for the Greek Genesis” JBL 52 (1933), 239-243. 31. E. g. E. Tov, The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of the LXX of Jeremiah 29-52 and Baruch 1,1–3,8, Missoula, MT 1976. 32. T. Muraoka, “In Defence of the Unity of the Septuagint Minor Prophets” AJBI 15 (1989), 2536; J. Ziegler, “Die Einheit der Septuaginta zum Zwölfprophetenbuch” in: J. Ziegler, Sylloge: Gesammelte Aufsätze zur Septuaginta (MSU 10), Göttingen 1971, 29-42; A. Pietersma, “Of Translation and Revision: From Greek Isaiah to Greek Jeremiah” in: M. N. van der Meer et al. (eds.), Isaiah in Context. Studies in Honour of Arie van der Kooij on the Occasion of his SixtyFifth Birthday, Leiden/Boston 2010, 359-387. 33. H. St. J. Thackeray, “The Greek Translators of the Four Books of Kings” JTS 8 (1907), 262-278; D. Barthélemy, Les devanciers d’Aquila. VT Supplement X, 1963. 34. J. D. Shenkel, Chronology and Recensional Development in the Greek Text of Kings, Cambridge, MA 1968 35. R. Sollamo, Semiprepositions, 280-289; I. Soisalon-Soininen, Die Infinitive, 176-190.

170

4. Results

10. The Study of Translation Technique

Translation technique study has shown that the translators seem to have read and translated their source text in small units of a few words at a time. 36 Free translation deals with larger units, as the book of Proverbs shows. Wide variations in the length of books can generally be attributed to a different Vorlage, the Septuagint version of Jeremiah being a parade example. 37 The existence of Hebrew texts different from the MT has been confirmed by the Qumran manuscripts. Some of these Qumran manuscripts are closely related to the Vorlage of the Septuagint. 38 In general, the translator did not shorten or revise the text, but adhered to the Hebrew Vorlage at his disposition. At times the text is very different from the MT (Jer, Exod 36–40 39); at others very close (the books of the Pentateuch, Isaiah). Thus, the Vorlage of the Septuagint gives us access to pre-Masoretic textual developments of the Hebrew Scriptures.

36. Soisalon-Soininen, “Beobachtungen zur Arbeitsweise der Septuaginta-Übersetzer” in: Soisalon-Soininen, Studien, 28-39. 37. E. Tov, “The Nature of the Large-Scale Differences between the LXX and MT S T V, Compared with Similar Evidence in Other Sources” in: Tov, Hebrew Bible, Greek Bible, and Qumran. Collected Essays, Tübingen 2008, 154-170. 38. E. Tov, “The Biblical Texts from the Judean Desert – An Overview and Analysis” in: Tov, Hebrew Bible, Tübingen 2008, 128-154, esp. 140-154. 39. A. Aejmelaeus, “Septuagintal Translation Techniques – A Solution to the Problem of the Tabernacle Account” in: Aejmelaeus, Trail, 116-130. 4. Results

171

11. Transcriptions of Hebrew Words Katrin Hauspie*

The Septuagint stands at the crossroads of two different languages, and of two different scripts. One of the consequences of this position is the presence of transcriptions and/ or transliterations. “Transcription” designates the mapping of the sounds of one language into the best matching phonetic equivalents of another language; “transliteration” refers to the representation of a word in the closest corresponding letters, or characters, of a different alphabet or language. While there is a fairly large body of research related to transcriptions and/or transliterations in the Septuagint, many studies fail to provide adequate definitions of these terms, “transcription” and “transliteration” often being used indifferently to describe the same phenomenon, the latter being more frequent. The first section of the present study assesses the use of both terms in Septuagint literature, highlighting some curious facts. The second section describes different categories of transcriptions. The third section discusses an issue of theoretical significance: the phonetic basis for the transcription of Hebrew texts.

1. A Question of Terminology Although “transcription” and “transliteration” are distinct terms, Septuagint literature has taken little notice of the distinction. In studies written in English, the term “transliteration” is consistently used to describe Hebrew words that are represented, not translated, in the closest corresponding letters of the Greek alphabet. 1 Henry St. John Thackeray 2 was the first to speak of transliteration, concerning words like γομορ, μαννα, πασχα. Later scholars adopted the term from this classic work and used it to refer to similar cases. Hence it became the standard term for designating Hebrew

*

1.

2.

I would like to express my gratitude to John Lee and Takamitsu Muraoka for making valuable remarks and suggestions on an early draft of this article. Any remaining errors are, of course, solely my responsibility. See, e. g., B. Kedar-Kopfstein, “The Interpretative Element in Transliteration” Textus 8 (1973), 55-77; J. Lust, “A Lexicon of the Three and the Transliterations in Ezekiel” in: A. Salvesen (ed.), Origen’s Hexapla. Papers presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25th July–3rd August 1994 (TSAJ 58), Tübingen 1998, 274301; E. Tov, “Loan-words, Homophony and Transliterations in the Septuagint” in: E. Tov (ed.), The Greek and Hebrew Bible. Collected Essays on the Septuagint (SVT 72), Leiden/Boston, MA 1999, 165-181; E. Tov, “Transliterations of Greek Words in the Greek Versions of the Old Testament. A Further Characteristic of the kaige-Th. Revision” in: Tov, Greek and Hebrew Bible, 501-512. H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, vol. I, Introduction, Orthography and Accidence, Cambridge 1909, 31-38.

172

1. A Question of Terminology

11. Transcriptions of Hebrew Words

words written in Greek characters. The term “transcription” is not however completely absent from the literature. 3 Studies written in French and German speak of “transcription” and “Transkription.” 4 Marguerite Harl draws a clear distinction between “transcription” (“reproduction des caractères hébreux dans le texte grec,” e. g. the tetragram) and “translittération” (“écriture des mots hébreux en caractères grecs”). 5 The difference between “transliteration” and “transcription” appears in studies that use these terms to refer to different phenomena. 6 Jan Joosten explicitly links transcription with the representation of the vocalization—sound representation—of the trigram YHW/YHH. 7 Benjamin Kedar-Kopfstein distinguishes between graphic transliteration and phonetic transcription, but categorizes both as transliteration. 8 On the other hand, Karen Jobes and Moisés Silva define transliteration as “using Greek letters to represent the sound of the Hebrew,” 9 thereby conflating transliteration (which aims at a letter-to-letter correspondence) and transcription (which focusses on sound resemblance). Occasionally, the terms “transcription” and “translation” are used synonymously in the same article, only a few lines apart. Maria V. Spottorno, also, speaks of transliterations in the book of Ezekiel, citing such instances as αιλαμ,

3.

4.

5.

6. 7. 8. 9.

Cf. F. Knobloch, Hebrew Sounds in Greek Script: Transcriptions and Related Phenomena in the Septuagint, with Special Focus on Genesis, dissertation, University of Pennsylvania, 1995; A. F. Rainey, “LXX Toponymy as a Contribution to Linguistic Research” Lingua Aegyptia 9 (2001), 179-192; J. Joosten, “Language as Symptom: Linguistic Clues to the Social Background of the Seventy” Textus 23 (2007), 74-75; T. Harviainen, “The Greek Traditions of Proper Names in the Book of First Esdras and the Problems of Their Transfer into a Modern Translation in the Light of a New Finnish Version” in: A. Voitila / J. Jokiranta (eds.), Scripture in Transition. Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo (SJSJ 126), Leiden/Boston, MA 2008, 485-495. Cf. F. X. Wutz, Die Transkriptionen von der LXX bis zu Hieronymus. Texte und Untersuchungen zur vormasoretischen Grammatik des Hebräischen (BWAT N.F. 34/9), vol. 2, Kohlhammer 1925; É. Masson, Recherches sur les plus anciens emprunts sémitiques en grec, Paris 1967, 115-116; A. Pelletier, “Σαββατα, transcription grecque de l’araméen” VT 22 (1972), 436-447; R. Hanhart, Text und Textgeschichte des 1. Esrabuches, Göttingen 1974, 55; M. Harl, La Genèse (La Bible d’Alexandrie 1), Paris 1986, 51, 103, 323; J.-L. Fournet, “Les emprunts du grec à l’égyptien” Bulletin de la Société de la Linguistique de Paris 84 (1989), 55-80; G. Dorival, Les Nombres (La Bible d’Alexandrie 4), Paris 1994, 271; B. Grillet / M. Lestienne, Premier Livre des Règnes (La Bible d’Alexandrie 9/1), Paris 1997, 108-109; I. Assan-Dhôte / J. Moatti-Fine, Baruch, Lamentations, Lettre de Jérémie (La Bible d’Alexandrie 25/2), Paris 2005, 83, 265; J. Joosten, “Le Dieu IAÔ et les tréfonds araméens des Septante” in: M. Loubet / D. Pralon (eds.), Eukarpa. Études sur la Bible et ses exégètes réunies en hommage à Gilles Dorival, Paris 2011, 118, 119. In Italian and Spanish, see A. Vaccari, “πόλις ασεδεκ Is. 19, 18” Biblica 2 (1921), 353-356; N. Fernández Marcos, “Nombres proprios y etimologias populares en la Septuaginta” Sefarad 37 (1977), 239-259. M. Harl / G. Dorival / O. Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien, Paris 1988, 261. However, page 66, the authors use both terms indistinctly. Joosten, “IAÔ,” 118; Pelletier, “Σαββατα,” 446. Joosten, “IAÔ,” 119. Kedar-Kopfstein, “Interpretative Element,” 55. K. H. Jobes / M. Silva, Invitation to the Septuagint, Grand Rapids, MI 2000, 41. 1. A Question of Terminology

173

11. Transcriptions of Hebrew Words

αιλευ, θαιηλαθα; later on, she refers to these same terms as “transcriptions,” 10 thus conflating transliteration and transcription. The same equation is made explicitly by Emanuel Tov. 11 The interchangeability of our terms is particularly noticeable with respect to the word σαββατα. André Pelletier describes this term as a transcription of the Aramaic ‫ שבתא‬12, but many other authors deem it a transliteration. 13 Even in referring to Pelletier’s article, some scholars use the terms “transliteration” and “transcription” interchangeably, as if they were synonymous. 14 In a text written in English, Joosten calls “Sabbath” a transliteration; elsewhere, in an article published in French, 15 he describes it as a transcription. 16 As this author apparently uses “transcription” or “transliteration” depending on the modern language he is writing in, 17 a possible explanation could be that English favors one term, “transliteration,” while French (and other languages) allows for finer terminological distinctions. 18 Still, a specific use of the term “transcription” in English appears in an article written by Benjamin G. Wright, where it is used in the sense of “copy.” 19 This meaning is close to the use of the derivative “transcriptional,” which designates scribe’s activity. 20 This quick survey seems to indicate that the English language favors the term “transliteration” to designate the process of representing words originating in one language or script with the closest corresponding letters or sounds of another tongue or alphabet and that it reserves the term “transcription” to refer to the work of the copyist. Given the equivocal terminology, it is reasonable to ask whether it is indeed possible to distinguish between transcription and transliteration. Some scholars posit that 10. M. V. Spottorno, “Some Lexical Aspects in the Greek Text of Ezekiel” in: J. Lust (ed.), Ezekiel and his Book (BETL 74), Leuven 1986, 79. 11. Tov, “Loan-words,” 166: “Transliteration, that is, the transcription into Greek characters of Hebrew and Aramaic words”; and 174: “transliteration (transcription).” See also B. Lemmelijn, “Flora in Cantico Canticorum. Towards a More Precise Characterisation of Translation Technique in the LXX of Song of Songs” in: Voitila / Jokiranta, Scripture in Transition, 27-51; see also, above, note 6. 12. Pelletier, “Σαββατα,” 436-447. 13. J. Joosten, “Reflections on the Interlinear Paradigm” in: Voitila / Jokiranta, Scripture in Transition, 173; Tov, “Loan-words,” 165. 14. Harl / Dorival / Munnich, Bible grecque, 262: “A propos des mots paskha et sabbata, A. Pelletier a montré qu’en les translittérant […]”; Tov, “Loan-words,” 177. 15. Joosten, “Reflections,” 173; Joosten, “Language,” 75-76. 16. Joosten, “IAÔ,” 119: “sabbata, paskha – transcrits en grec.” 17. When writing in English, Joosten consistently uses the words “transcription/transcribe” when referring to θιβις, οιφι, αχει: they are “transcriptions” of Egyptian words that entered into the Greek language before prior to the Septuagint; some Hebrew words are not translated but transcribed (e. g. μαν, χερουβιμ, ιν, γομορ). Further on in the same article, he suddenly introduces the term “ad hoc transliterations” to designate the same phenomenon (“Language,” 74-76). 18. Pelletier (“Σαββατα,” 446) calls the transition from the Greek Σαββατα and Πασχα to the Latin Sabbata and Pascha, a clear case of letter by letter representation, “transliteration.” 19. B. G. Wright, “Transcribing, Translating, and Interpreting in the Letter of Aristeas: on the Nature of the Septuagint” in: Voitila / Jokiranta, Scripture in Transition, 147-161, in particular 155 and 157. 20. Jobes and Silva, Invitation, 128-131.

174

1. A Question of Terminology

11. Transcriptions of Hebrew Words

the representation of Hebrew words in Greek translations is situated midway between transliteration and transcription. 21 Indeed, although the terms “transcription” and “transliteration” are theoretically distinct, the line between them is often thin. Keeping this distinction in mind, we will now take a closer look at transcription.

2. Categories of Transcriptions In an article first published in 1979, subsequently revised in 1999, Tov examined a great number of transliterations, dividing them into four categories according to the reasons why the original words were left untranslated. 22 These categories provided the basis for further study of transliterations. Because they are applicable to transcriptions as well, they are worthy of mention here. 23 To these four categories, I add a fifth one not mentioned by Tov. Hebrew or Aramaic words that were not translated into Greek but transcribed are generally classified into four categories: (1) proper names, (2) technical terms, (3) words unknown to the translator, and (4) common nouns erroneously understood as proper names. 24 Occasionally they are accompanied by a translation, which results in a double rendering of the original word. Although some double renderings were introduced subsequently by editors during the process of textual transmission, 25 others can be traced back to the translator who wished, by this method, to indicate a hesitation between two possibilities or efforts for rendering an unfamiliar or difficult term, e. g. 1Esdras 2:19 τῷ γράφοντι τὰ προσπίπτοντα καὶ Βεελτεέμῳ for ‫בעל טעם‬. 26 When this occurs, it constitutes a fifth category of transcription. 27 21. See Z. Frankel, Vorstudien zu der Septuaginta, Leipzig 1841, 107-123; M. Pei, Glossary of Linguistic Terminology, New York, NY 1969, 280; Kedar-Kopfstein, “Interpretative Element,” 55. 22. Tov, “Loan-words.” 23. As will become clear, the examples given by Tov are actually transcriptions, in the sense of the present article. 24. Thackeray, Grammar, 32; Tov, “Loan-words,” 174; Tov, “Transliterations,” 504. For a nonexhaustive list of transliterations in the Septuagint based on HR, see P. N. Simotas, Αἱ ἀμετάφραστοι λέξεις ἐν τῷ κειμένῳ τῶν Ο’), Thessaloniki 1968. Tov (“Loan-words,” 175-181; “Transliterations,” 507) lists some transliterations according to the above mentioned categories. A list of transliterations in Ezekiel can be consulted in Spottorno “Lexical Aspects,” 79 and Lust, “Lexicon,” 281-288. 25. Z. Talshir, “Double Translations in the Septuagint” in: C. E. Cox (ed.), IV Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem 1986 (SCS 23), Atlanta, GA 1987, 34-35. 26. Talshir, “Double Translations,” 43-46. A few verses earlier ‫ בעל טעם‬was translated as ὁ τὰ προσπίπτοντα (1Esdras 2:13) and transcribed as Βεελτέεμος (1Esdras 2:12). Talshir sometimes calls Βεελτέεμος a transliteration, and at other times a transcription. It is in fact neither a pure transliteration nor a transcription, as the ending of the declension is added. 27. Double renderings of Hebrew words, containing a transcription as well as a translation, reveal a distinct use of either an original or a secondary transcription. In the case of original transcription (category 3), the translator deliberately left a Hebrew word untranslated, because he was ignorant of its meaning and wanted to leave room for later generations to replace it with the proper equivalent. Ultimately, the transcription was not replaced, but rather supplemented 2. Categories of Transcriptions

175

11. Transcriptions of Hebrew Words

Up until now, transcriptions of Hebrew words seem to have in common the translator’s inability to find a precise Greek equivalent. Yet this is not the only reason for resorting to transcription: the translator may deliberately opt for a transcription in order to preserve something of the atmosphere of the original. 28 This can be seen in each of Tov’s four categories 29. Proper names (category 1) are likely to be transcribed 30 because they serve to identify individual persons or entities, although there are examples of proper name translation, 31 particularly toponyms with etiological functions. 32 Furthermore, proper names can be alternately translated or transcribed. ‫ ירבעל‬is transcribed in JudgA 6:32 as Ιαρβααλ, but translated in a descriptive way in JudgB 6:32, Δικαστήριον τοῦ Βααλ; 33 the proper name “Philistines,” ‫פלשתים‬, generally transcribed Φυλιστειμ in the Hexateuch, is translated as ἀλλόφυλοι in the later books; 34 the rare word ‫אהלות‬, which is the name of a plant, is rendered as στακτή in Ps 44/45:8 and as αλωθ in Song 4:14; 35 in Ezekiel ‫ צר‬is transcribed as Σορ in chapters 26 to 27 and then translated as

28. 29. 30. 31.

32. 33. 34. 35.

by a translation. Tov (“Loan-words,” 179) gives some examples. In 1Sam 5:4, e. g., he says that τὰ ἐμπρόσθια is an added translation for the original transliteration αμαφεθ. However such a double rendering could be equally satisfactorily explained the other way round. Indeed, Michel Lestienne (Grillet / Lestienne, Premier Livre des Règnes, 173-174) argues that αμαφεθ is probably a gloss entered at a later stage by a scribe who wanted to give some explanatory information on word choices made by the translator by referring to an underlying Hebrew text that was not necessarily the Masoretic one. Defining Amafeq as transcription however is debatable. The consonants reflect the consonants of the Hebrew, but the vowels do not. The label transliteration is questionable as well. For further discussion on this word see below. For another example of double rendering, see also the note on 1Sam 6:8 (Grillet / Lestienne, Premier Livre des Règnes, 180): in 1Sam 6:8,11 and 15 a scribe thrice introduced the underlying Hebrew word ‫—ארגז‬a rare word that occurs only here in the Masoretic text—in two different transcribed forms (βερεχθαν and εργαβ) in addition to the original translation (θέμα). The transcribed forms do not reflect the Hebrew of the Masoretic text, except possibly in the case of ‫ ארגז‬and εργαβ. The scribe wanted to point to the variety of Hebrew forms for θέμα in his Hebrew manuscript. Kedar-Kopfstein, “Interpretative Element,” 56. These categories are thoroughly studied in Tov, “Loan-words,” 174-182. Tov, “Loan-words,” 175-176; Fernández Marcos, “Nombres proprios,” 242; Kedar-Kopfstein, “Interpretative Element,” 57-58. Hebrew proper names bear a lexical value, linked to common nouns, verbal roots, and recognizable morphemes that points to a symbolic interpretation; when transliterated, this semantic information gets lost, concealing a suggestive interpretation, and leaving only a combination of characters. In this procedure the alleged historicity of the narrative seems to gain importance. On the other hand, translating a proper name compromises its exclusive reference to one person or entity. See Kedar-Kopfstein, “Interpretative Element,” 57-58, 61-63. H. Ausloos, “LXX’s Rendering of Hebrew Proper Names and the Characterisation of Translation Technique of the Book of Judges” in: Voitila / Jokiranta, Scripture in Transition, 53-71. Ausloos, “LXX’s Rendering,” 59-61. Tov, “Loan-words,” 176. Lemmelijn, “Flora,” 48. See also Masson, Recherches, 72: “αλωθ is a transcription.” Transcription is a more appropriate term here since hαi renders h‫אה‬i, unlike the proper name ‫אהרן‬, transcribed by Aaron, where hααi = h‫אה‬i. Neither can Aaron be called a transliteration, for hoi represents the unwritten ḥolem.

176

2. Categories of Transcriptions

11. Transcriptions of Hebrew Words

Τύρος from chapter 28 onward. 36 Inconsistency in the rendering of proper names cannot be accounted for simply as hesitation or a desire for variety on the part of the translator. Rather, it is an indication of multiple translators and/or translation units within a single book, as can be seen in the last example. Technical terms (category 2) can be divided into two subgroups: religious terms and weights and measures. 37 Many technical terms of Semitic origin were attested in Greek before the Septuagint was written. 38 Some of these ultimately derived from Egyptian, e. g. ἵν ‫ הין‬hnw. 39 Technical terms such as weights and measures used in daily speech naturally entered the Septuagint; other technical terms, particularly religious ones, had no direct equivalents in Greek. The language of the Jews of Egypt at the time of the Greek translation was not Hebrew—the language of the Vorlage of the Septuagint—but Aramaic. Technical religious terms denoting institutions and realia peculiar to Judaism were in constant use among Jews. Transcriptions of these words are based on their spoken Aramaic form rather than on the written Hebrew one. One example can serve to illustrate this. The Aramaic traditional calendar in use at the time of the translation inspired the translators 40, who, in choosing the names of the feasts as Sabbat and Pesach, retained the Aramaic forms of the traditional calendar, (‫ שבתא‬and ‫)פסחא‬, which they transcribed as σαββατα and πασχα, rather than the Hebrew one, (‫ שבת‬and ‫)פסח‬, they found in their Vorlage. 41 There are of course transcriptions in the area of religion that reflect Hebrew words preserved in the Masoretic text. Words like μαναα, εφουδ, which were no longer in daily use, 42 were transcribed in their Hebrew forms (‫ מנחה‬and ‫)אפוד‬. 43 A large number of Hebrew transcriptions concern architectural terms related to the Temple, e. g. αιλευ, αιλαμ, θαιηλαθα, θεε (see 1Kings, 2 Chron, Ez). 44 Because of their technical 36. H. St. J. Thackeray, The Septuagint and Jewish Worship (The Schweich Lectures 1920), London 1923, 37; L. J. McGregor, The Greek Text of Ezekiel: An Examination of Its Homogeneity (SCS 18), Atlanta, GA 1985, 31-32. 37. Tov, “Loan-words,” 177-178. 38. Thackeray, Grammar, 32, 34-36; Masson, Recherches. 39. For transcriptions from Egyptian, see J.-L. Fournet, “Les emprunts du grec à l’égyptien” Bulletin de la Société de la Linguistique de Paris 84 (1989), 55-80, in particular 78. See also Joosten, “Language,” 74 on θιβις, οιφι and αχει. 40. Kedar-Kopfstein (“Interpretative Element,” 56) has called this the “desire for the preservation of some of the original mood.” 41. Pelletier, “Σαββατα,” 436-447, in particular 444-446. Later σαββατα conformed itself to Greek declension: it was felt as a plural (sabbaths), and σαββατον became the singular form (a sabbath). 42. The fact that ‫ מנחה‬is often translated as well (δῶρον, θυσία, σεμίδαλις, ξενία, προσφορά) supports the assumption that μαναα was not part of everyday usage. See also Spottorno “Lexical Aspects,” 78. 43. Efoud can as well be considered as a transliteration, due to letter-to-letter correspondence (hwi is represented by the digraph houi, which is the closest corresponding “letter”). 44. For a discussion of the terms related to the architecture of Ezekiel’s temple, see Spottorno “Lexical Aspects,” notes 7 and 8; Lust, “Lexicon,” 281-283; and D. M. O’Hare, “Have you seen, Son of Man?” A Study in the Translation and Vorlage of LXX Ezekiel 40-48 (Septuagint and Cognate Studies 57), Atlanta, GA 2010, 61-71. 2. Categories of Transcriptions

177

11. Transcriptions of Hebrew Words

nature, these words were not understood, and the translators therefore left them untranslated. Fulfilling the hope of the translators, subsequent editors such as Aquila, Symmachus, or Theodotion replaced the transcriptions and transliterations with Greek equivalents. 45 Likewise, Church Fathers clarified transcriptions and transliterations or provided them with an adequate translation. 46 Although similar in some respects to the previous group, 47 words unknown to the translator (category 3) do not correspond to a specific class of words. When the translator did not understand a word at all, as was often the case with hapax legomena and rare or difficult words, he decided the best he could do was to leave the word untranslated, often in its exact Hebrew form, including prefixes and suffixes, in the hope that an adequate translation would be found by subsequent translators. 48 Examples that illustrate this phenomenon include Αβαμα for ‫ הבמה‬in Ez 20:29 (inclusion of the Hebrew article), αμασενιθ for ‫ השמינית‬in 1Chron 15:21 (inclusion of the Hebrew article and inversion of h‫מ‬i and h‫ש‬i), and βεζεκ for ‫ בזק‬in EzA 1:14. 49 Common nouns (category 4) unknown to the translator were sometimes rendered as proper names in transcribed form. 50 In the previously given example, that of Αβαμα, the Hebrew ‫במה‬, “high place,” became a toponym. 51 Another example can be found in 3Reigns. The proper name Χεζραθ in the Septuagint reading of 3Reigns 15:20 can hardly be said to reflect the Masoretic text ‫“( כנרות‬Chinneroth,” a city name in the plural form); more likely, it corresponds to the hypothetical ‫ כצרות‬or ‫כזרות‬ (“strongholds,” an unattested common noun), which derives from the Semitic root qṣr that denotes military fortifications (cf., e. g., the Akkadian kṣr, “to fortify, to be strong”). Subsequently, this hypothetical term may have become associated in the Hebrew text with a location in the land of Naphtali, ‫כנרות‬. Thus, it is just possible that the Septuagint evidences the original reading, although the translator did not recognize the word since it had fallen into disuse and was consequently unfamiliar at the time of the Masoretes, a fact which could explain the reading ‫כנרות‬, a better known place. 52

3. Phonetics in Transcription Since the Hebrew writing system at the time of the translation of the Septuagint only marked systematically some of the sounds of the language, it is more appropriate to speak of “transcription” rather than “transliteration” to describe the phenomenon studied here. 45. Lust, “Lexicon,” 289-293. 46. K. Hauspie, “Hebrew Transliterations in the Septuagint version of Ezekiel elucidated: in Search of the Sources of Theodoret of Cyrrhus” in: W. Kraus / M. Karrer (eds.), Die Septuaginta – Texte, Theologien, Einflüsse (WUNT 252), Tübingen 2010, 435-444. 47. For a tentative list see Tov, “Transliterations,” 507-510. 48. Tov, “Loan-words,” 179-180. 49. Lust, “Lexicon,” 288. 50. For a list of examples see Tov, “Loan-words,” 180. 51. For detailed analysis of other examples, see Lust, “Lexicon,” 285-288 with more examples elaborated. 52. A. Finet, “Termes militaires accadiens conservés dans la LXX” Iraq 25 (1963), 191-192.

178

3. Phonetics in Transcription

11. Transcriptions of Hebrew Words

The most complete example of the transcription of a Hebrew text into Greek is undoubtedly the second column of the Hexapla. Franz Xaver Wutz, went so far as to suppose that a transcribed text similar to the Secunda served as the starting point for the Greek translation. 53 Known as the “transcription theory,” this position is no longer maintained today. 54 Nevertheless it is generally agreed that the Secunda is an important source for the study of the pronunciation of Hebrew. 55 This brings us to our last section: the phonetic basis for the transcription of Hebrew texts. The use of the term transcription is not unproblematical. As the examples above have shown, the translator was sometimes confronted with a word completely unclear to him (e. g. αμαφεθ). In such cases he resorted to copying the Hebrew word letter by letter. However, since he had only an unvocalized text at his disposition, he was faced with a second difficulty: Greek requires that vowels be placed in between consonants, and these were not provided by the Hebrew text. Αμαφεθ, which presumably corresponds to ‫המפת‬, differs significantly from the pointed ‫ המפתן‬in the Masoretic text. Therefore the label “transcription” is misleading; but “transliteration” is unsatisfying as well, as the Greek form contains more letters than the original Hebrew. Transcription in the Septuagint implies that the translator was to some degree familiar with the word he was transcribing or that at the very least he knew how to pronounce it, i. e. was aware of the vowels used. Furthermore, Hebrew pronunciation was itself subject to evolution, as is witnessed by the existence of different transcription systems: the Vaticanus and the Alexandrinus reflect different Hebrew pronunciations. 56 As a general rule Greek tends to represent the Hebrew consonants with the closest corresponding Greek ones. 57 Gerhard Lisowsky set up a list of equivalences based on his study of proper names. 58 Until the late Hellenistic period, the Semitic /‫ת‬/ was mainly transcribed as hτi, e. g. σαββατα, and only came to be transcribed as hθi in the late Aramaic dialects, as can be seen in the New Testament, the Masoretic text, and rabbinic literature. 59 But earlier examples of /‫ת‬/ transcribed as hθi do exist, e. g. Σαβαωθ (1Sam 1:3), αλωθ (Song 4:14). Short vowels without stress (shewa) are often transcribed as hαi, e. g. Χαρχαμις for ‫( כרכמש‬Jer 26/46:2) and Μαναημ for ‫( מנחם‬2Kings 15:14). 60 Vowel length was generally respected, thus long medial syllables were usually rendered in the transcription, e. g. Μαναημ. 61 53. Wutz, Die Transkriptionen. 54. In his last work (Das Buch Job, Stuttgart 1939) Wutz began to distance himself from his earlier position. The theory was definitively disproved by Sidney Jellicoe (The Septuagint and Modern Study, Oxford 1968, 70-73). 55. Cf. E. Brønno, Studien über hebräische Morphologie und Vokalismus auf Grundlage der mercatischen Fragmente der zweiten Kolumne der Hexapla des Origenes (Abhandlungen für die Kunde des Morgenlandes 28), Leipzig 1943; G. Janssens, Studies in Hebrew Historical Linguistics Based on Origen’s Secunda (Orientalia Gandensia 9), Leuven 1982. 56. Janssens, Studies, 6-23. 57. Masson, Recherches, 115. 58. G. Lisowsky, Die Transskription der hebraeischen Eigennamen des Pentateuch in der Septuaginta, Basel 1940, 137. 59. Pelletier, “Σαββατα,” 442. Masson, Recherches, 115. 60. Pelletier, “Σαββατα,” 443-444. 61. Pelletier, “Σαββατα,” 440. 3. Phonetics in Transcription

179

11. Transcriptions of Hebrew Words

Transcriptions of Hebrew words in the Septuagint often reflect an older stage of Hebrew pronunciation than the one reproduced in the Masoretic text, which sometimes reflects Aramaic pronunciation and morphology. 62 Αναχερεθ (JoshB 19:19) most probably represents an old Hebrew pronunciation anterior to the one attested in the Masoretic text (‫)אנחרת‬. 63 The difference in the transcription of the first consonant in ‫ עשתרות‬and ‫ עזה‬proves that ayn and ġayn, sounds still preserved in Arabic, were pronounced differently at the time of the Greek translation, as their respective Greek transcriptions, Ασταρωθ (Josh 9:10) and Γάζα (Josh 15:47), show. 64 Μαγδώλου (Exod 14:2) and Μαγδώλῳ (Jer 51:1) reflect the original prefix denoting the location where something takes place, which is preserved in ‫ מקום‬and ‫מעבר‬, as opposed to the Masoretic pointing ‫מגדול‬, which already gives evidence of a shift in the prefix from [ma] to [mi] in closed, unaccented syllables. 65 The toponyms Ακκαρων (Josh 15:11) and Ιερακων (Josh 19:46) harken back to the earlier Hebrew pattern qattalôn/qittalôn, in contrast to the Masoretic text, which has ‫ עקרון‬and ‫( הירקון‬pattern qatlôn/qitlôn). 66 In a like manner, Αβια (2Chron 13:20,21) and Αχια (2Chron 10:15) present the abridged suffix for the Hebrew name for Jahweh (‫יה‬-), which probably figured in the Vorlage, rather than the longer form (‫יהו‬-) found in the Masoretic text. 67 Σαβατια (2Sam 3:4), Εδια (2Kings 22:1), and Ορνια (2Sam 3:4) coincide however with the abridged form. Our last example will demonstrate the usefulness of phonetic rules for reading and interpreting Greek and Hebrew texts. In Isa 19:18, the Septuagint has πόλις ασεδεκ instead of ‫עיר ההרס‬, as in the Masoretic text, and ‫עיר החרס‬, as in some Hebrew manuscripts (Symmachus and Jerome). On the basis of the Greek transcription, some scholars postulated an original reading ‫עיר הצדק‬. A. Vaccari convincingly refutes this proposal by showing that the transcription ασεδεκ confuses h‫ר‬i and h‫ד‬i (a frequent misreading in Isaiah), transcribes h‫ח‬i as hκi (mostly at the end and in the middle of Hebrew words), and inverts characters (as is often the case in Isaiah e. g. ‫ שלח‬for ‫)חולש‬ in such a way that ‫ החרס‬is read as ‫הסדח‬, resulting in the transcription ασεδεκ. 68

Conclusion This brief survey first clarified some of the terminological confusions that arise in discussions about transcriptions. Second, it assessed the contribution of the study of transcriptions to the field of Semitic linguistics. The transcriptions of the Septuagint pro62. See Rainey, “LXX Toponymy,” 179-192, in particular 179-182. See also Harviainen, “Greek Traditions,” 491. 63. ’Anōḫartu (Canaanite, reconstructed form supported by an Egyptian transcription) ’Anāḫeret (Hebrew = LXX) ’anāḥarat (Aramaic = MT) (Rainey, “LXX Toponymy,” 180). 64. Rainey, “LXX Toponymy,” 180. The Egyptian form ’a-s-ta-ar-tu also gives proof of the presence of an ‘ayn h’i. 65. Rainey, “LXX Toponymy,” 181-182. 66. Rainey, “LXX Toponymy,” 182. 67. S. Norin, “Die Wiedergabe JHWH-haltiger Personennamen in der Septuaginta” SJOT 2 (1988), 93-94. 68. Vaccari, “πόλις ασεδεκ,” 353-356.

180

Conclusion

11. Transcriptions of Hebrew Words

vide invaluable information on the pronunciation of words during the period when the Hebrew Bible was translated into Greek, because they often reflect an older stage of Hebrew pronunciation than the one attested in the Masoretic text (e. g. Αναχερεθ). Some sounds, although indiscriminately represented by one and the same Hebrew character, may still be alive in oral tradition and pronounced distinctively, as evidenced in the Septuagint (e. g. Ασταρωθ and Γάζα). Transcriptions of everyday words, such as technical religious terms, reflect their Aramaic form, spoken by the Jewish translators in Egypt of that time, instead of the older Hebrew written form (e. g. σαββατα and πασχα). Third, it demonstrated that the study of transcriptions can greatly increase our knowledge of the Septuagint: chosen because of the translator’s inability to deal appropriately with the Hebrew, transcriptions preserve something of the original setting and mood.

Conclusion

181

12. Hebraisms Andrés Piquer Otero

1. Hebraisms in the Context of Biblical Greek Scholarship The study of Hebraisms in the Septuagint reflects the development of the study of Biblical Greek in general. 1 Although the phenomenon of Hebraisms can be simply defined from a descriptive standpoint as “Greek words, phrases or constructions which transfer characteristic Hebrew elements into Greek in disregard of Greek idiom,” 2 the field has been marked by competing trends in biblical scholarship: the so-called “Hebraist” and “Purist” schools and, in the modern age, the Hebraizers and the Hellenists. 3 Although some of these lines of thought had a markedly theological or literary-theological bias, they nonetheless influenced contemporary scholarship from the nineteenth century onwards. Previously dominated by concerns over the dignity and purity of Scripture or the distinction between Classical (and hence “pagan”) and “Christian Greek” (New Testament and Septuagint Greek, since this latter corpus had been adopted by Christian communities), 4 explanations of the “Semitic elements” 5 in Septuagint Greek became increasingly dependent on linguistic models and methodologies. Three important trends in this evolution can be identified: (1) contextualization of Biblical Greek (Septuagint and New Testament) in order to compare it, not only to 1.

2. 3. 4. 5.

For a summary of the present state of investigation, see the chapter “Biblical Greek and its Position within koiné” by Natalio Fernández Marcos, in his The Septuagint in Context: Introduction to the Greek Version of the Bible, Leiden/Boston, MA/Köln 2000, 3-17. E. Tov, The Text-Critical Use of the Septuagint in Biblical Research (second edition, revised and enlarged, Jerusalem Biblical Studies 8), Jerusalem 1997 [1981], 22. For a detailed account of the history and positions of these schools, see J. Ros, De studie van het bijbelgrieksch van Hugo Grotius tot Adolf Deissmann, Nijmegen 1940. These questions, which go back to the origins of Christian literature and Patristics, lay beyond the scope of this chapter. See Fernández Marcos, Septuagint in Context, 3-5. This expression, used as early as 1909 by Henry St. John Thackeray, is more encompassing than the term “Semitism.” See H. St. J. Thackeray, A Grammar of the Old Testament in Greek According to the Septuagint, vol. I, Introduction, Orthography and Accidence, Cambridge 1909 (reprint Eugene, OR 2008), 25-55. It also connotes a larger array of contacts and exchanges between Greek (Biblical or other) and Semitic languages, historically, geographically, and culturally. See, e. g., P. Walters (formerly Katz), The Text of the Septuagint. Its Corruptions and Their Emendations (edited by D. Gooding), Cambridge 1973, 155-196. Basic works on the Greek-Semitic relationship include M. C. Astour, Hellenosemitica, Leiden 1965; E. Haag, Ugarit und das Alte Testament, Tübingen 1962; E. Mason, Recherches sur les plus anciens emprunts sémitiques en grec, Paris 1967; H. B. Rosen, L’hébreu et ses rapports avec le monde classique. Essai d’une évaluation culturelle, Paris 1979. On a larger cultural level, see the far-reaching study by Walter Burkert, The Orientalizing Revolution: Near Eastern Influence on Greek Culture in Early Archaic Greece (translated by Margaret E. Pinder and Walter Burkert), Cambridge, MA 1992.

182

1. Hebraisms in the Context of Biblical Greek Scholarship

12. Hebraisms

Classical Greek, but also to the language varieties of the Hellenistic period; (2) assessment of Biblical Greek, in particular its Semitism, from the perspective of the languages with which it was in contact (an approach which led to considering Biblical Greek as a manifestation of a form of bilingualism or diglossia, even ultimately as a witness to a particular dialect labeled “Jewish-Greek”); (3) use of translation technique and theory, which resulted in considering Septuagint Greek primarily as a translation language. These developments are not necessarily exclusive or even clearly sequential. Nevertheless, the ever-increasing body of Greek papyri that has been collected since the first discoveries in the eighteenth century provides an extended set of references for scholarly discussion on the nature of Biblical Greek in general and Hebraisms in particular, making it possible to compare the Septuagint, other Jewish texts in Greek, and the New Testament with the language attested in papyri, ostraca, and inscriptions from the Hellenistic period. The re-evaluation of the koine language—hitherto informed only by the traditional division between “literary” (the only one known until the papyri discoveries began) and “popular” koine—made possible by the comparison and contextualization of the so-called Semitic features of Biblical Greek, brought Adolf Deissmann and Albert Thumb to suggest that many of the voces biblicae (Semitic lexicon) 6 and syntactic features 7 (Hebraisms or Semitisms) of the Septuagint and the New Testament languages were not unique to these corpora but could be explained by the natural development of Greek, without postulating the influence of a foreign language. The arguments against the existence of a Jewish-Greek dialect put forth by Deissmann and Thumb found support in subsequent studies on Greek “literary” koine, particularly technical and scientific prose, some of which demonstrated the proximity of that language to Biblical Greek, thereby firmly establishing biblical and Jewish texts in the linguistic world of post-Classical Greek. 8 In reaction to Deissmann’s and Thumb’s proposal, other scholars sought to emphasize the presence of Hebraisms and Aramaisms in the Greek Bible. 9 A key element in this discussion was the evaluation of shared features between Septuagint and New Testament Greek, on the one hand, and the Hellenistic papyri mentioned above, on 6.

7.

8.

9.

A. Deissmann, Bibelstudien, Marburg 1895; A. Deissmann, Neue Bibelstudien, Marburg 1897; A. Deissmann, Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt (fourth edition), Tübingen 1923. A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus. Beiträge zur Geschichte und Beurteilung der Koiné, Strasbourg 1901 (reprint Berlin/New York, NY 1974); A. Thumb, “On the Value of the Modern Greek for the Study of Ancient Greek” CQ 8 (1914), 181-205. See E. K. Simpson’s seminal work Words Worth Weighing in the Greek New Testament (Tyndale Lecture 1944), London 1946. See also, L. Rydback, Fachprosa, vermeintliche Volksprache und Neues Testament. Zur Beurteilung der sprachlichen Niveauunterschiede in nachklassischen Griechisch, Uppsala 1967; and L. Gil, “Sobre el estilo del libro segundo de los Macabeos” Emerita 26 (1958), 11-32. Conducted mainly in the area of New Testament Greek, these studies aimed at proving that a considerable proportion of that text was translated from an Aramaic original. See, e. g., G. Dalman, Die Worte Jesu mit Berücksichtigung des nachkanonischen jüdischen Schrifttums und der aramäischen Sprache erörtert, Leipzig 1930; C. C. Torrey, The Four Gospels: A New Translation, New York, NY 1933. 1. Hebraisms in the Context of Biblical Greek Scholarship

183

12. Hebraisms

the other. Some authors went so far as to propose that the papyri themselves contained Semitisms 10 that could be attributed to the presence of Jewish communities in the area of Egypt where the materials were found. 11 Jozef Vergote refined this hypothesis by suggesting that Egyptian Greek could have been influenced by the Coptic-Egyptian substratum rather than by a Hebrew or an Aramaic one. 12 Vergote emphasized linguistic affinities between Hebrew-Aramaic and Coptic-Egyptian, noting that Semitisms are attested in the Coptic language and arguing that they should not be disregarded simply because they are also present in the papyri. 13

2. Bilingualism and Diglossia, Linguistic Variety, and Translation Language In spite of the emergence of a model that saw Biblical Greek as having developed from within the world of Koine through the diachronic development of the Greek language, many twentieth-century scholars remained committed to the idea of a “Jewish-Greek” dialect characterized by the presence of Semitisms, i. e. Hebraisms and Aramaisms, and continued to produce papers and monographs analyzing the Semitic features of the Septuagint, Jewish Apocrypha, and the New Testament. 14 Nevertheless, approaches that defended the presence of Semitisms and the exis10. Thackeray, following J. H. Moulton, adopts a more nuanced approach: “[…] ‘Hebraism’ of Biblical writings consists in the over-working of and the special prominence given to certain correct, though unidiomatic, modes of speech, because they happen to coincide with Hebrew idioms.” See Thackeray, Grammar, vol. I, 26 (emphasis by the author). 11. See, e. g., H. A. Redpath, “The Present Position of the Study of the Septuagint” AJT 7 (1903), 119; R. R. Ottley, A Handbook to the Septuagint, New York, NY 1920; F.-M. Abel, Grammaire du grec biblique suivie d’un choix de papyrus, Paris 1927; and the review of the latter in: L.-T. Lefort, “Pour une grammaire des LXX” Le Muséon 41 (1928), 152-160. 12. J. Vergote, “Grec Biblique” DBS 3 (1938), 1320-1369. For an application of this proposal to a case in the New Testament, see R. M. Wilson, “Coptisms in the Epistle to the Hebrews” Nov.T 1 (1956), 322-324. 13. Features noted by Vergote include substituting the numeral εἷς for the indefinite article τις (reflecting the Coptic ⲟⲩ); the distributive repetition of the numeral εἷς; the use of the definite article to express the vocative; and the use of the instrumental ἐν, due to the overlap of functions in the Coptic prepositions ⳍⲛ and ⲛ, which roughly echoes the functions of the Hebrew preposition ‫( ְב‬see Vergote, “Grec Biblique,” 1355-1359). This line of research is clearly problematic, as N. Fernández Marcos (Septuagint in Context, 15) has noted: “[…] in each case it needs to be determined which phenomenon is due to the inner development of Hellenistic Greek and which depends on or has traces of the influence of Coptic. And given the difficulty of this distinction in many cases, since Coptic is a language with a very simple construction, it has to be determined in which cases a particular linguistic phenomenon could be the result of both tendencies combined.” This problem has been connected to the presence of Semitisms (see above note 10). 14. See H. S. Gehman, “The Hebraic Character of the Septuagint” VT 1 (1951), 81-90; H. S. Gehman, “Hebraisms of the Old Greek Version of Genesis” VT 3 (1953), 141-148; H. S. Gehman, “Ἅγιος in the Septuagint, and its Relation to the Hebrew Original” VT 4 (1954), 337-348; N. Turner, “The Unique Character of Biblical Greek” VT 5 (1955), 208-213; N. Turner, “The Testament of Abraham: Problems in Biblical Greek” NTS 1 (1955), 219-223; P. Katz, “Zur Übersetzungstechnik der Septuaginta” Die Welt des Orients 2 (1956), 267-273.

184

2. Bilingualism and Diglossia, Linguistic Variety, and Translation Language

12. Hebraisms

tence of Jewish-Greek had to address the question of the function of the proposed dialect within its socio-linguistic context: was it an actual vernacular spoken by Jews in a situation of diglossia—a more accurate appellation than the more common term “bilingualism”—or did it manifest itself as a language variety circumscribed to precise social and cultural contexts, namely a religious usage related to the synagogue service and the reading of Scripture? 15 Ultimately the ever-finer distinctions introduced into the concepts of bilingualism and diglossia transformed the discussion of socio-linguistic usage (“synagogue Greek”) into a new field of inquiry: Biblical Greek as a translation language. Influenced by recent research on translation analysis theories and method, this approach is concerned with translators’ attitudes toward source texts as well as with the relationship between original and target languages; it studies translation choices both in individual units of translated text and in literary works generally. Eugene A. Nida was instrumental, not only in developing this line of research in translation science, 16 but also in defining research orientations in Septuagint translation technique. 17 He elaborated a new theoretical framework for understanding earlier studies on “translation Greek” in the biblical books. 18 Clearly, inquiry into Biblical Greek as a translation language cannot be restricted to the relationship between the individual translator and the text. 15. This second hypothesis was clearly formulated by Thackeray: “[…] a version that would be intelligible to the Greek-speaking Jew when read in the ordinary services of the synagogue.” See Thackeray, Grammar, vol. I, 28-29. Recently, the question was thoroughly analyzed by Georg Walser, who notes: “the overall impression is that the speech community of the ancient synagogue was poliglossic, i. e., there existed several varieties side by side.” See G. Walser, The Greek of the Ancient Synagogue: An Investigation on the Greek of the Septuagint, Pseudoepigrapha, and the New Testament (Studia Graeca et Latina Lundensia 8), Stockholm 2001, 183. His conclusion gives credence to the general idea that Jewish and Christian texts produced between the second century BCE and the second century CE originated within the context of the synagogue. 16. E. A. Nida, Towards a Science of Translating, Leiden 1964. 17. See, e. g., S. P. Brock, “The Phenomenon of the Septuagint” OTS 7 (1972), 11-36; S. P. Brock, “Aspects of Translation Technique in Antiquity” Greek, Roman, and Byzantine Studies 20 (1979), 67-87; H. M. Orlinsky, “The Septuagint as a Holy Writ and the Philosophy of the Translators” HUCA 46 (1975), 89-114; J. Barr, “The Typology of Literalism in Ancient Biblical Translations” Mitteilungen des Septuaginta-Unternehmens 15 (1979), 279-325; E. Tov and B. G. Wright, “Computer-Assisted Study of the Criteria for Assessing the Literalness of Translation Units in the LXX” Textus 12 (1985), 149-187 (reprint in: E. Tov, The Greek & Hebrew Bible. Collected Essays on the Septuagint, Leiden, 1999, 219-237); J. A. L. Lee, “Equivocal and Stereotyped Renderings in the LXX” RB 87 (1980), 104-117; J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SCS 14), Chico, CA 1983; S. Olofsson, “Consistency as a Translation Technique” Scandinavian Journal of the Old Testament 6 (1992), 14-30; S. Olofsson, A Guide to the Translation Technique of the Septuagint (CBOT 30), Stockholm, 1990. A summary of theories and positions may be found in J. K. Aitken, “The Language of the Septuagint: Recent Theories, Future Prospects” BJGS 24 (1999), 24-33. 18. Works that apply these principles to the Septuagint include: D. W. Riddle, “The Logic of the Theory of Translation Greek” JBL 51 (1932), 13-30; J. M. Rife, “The Mechanics of Translation Greek” JBL 52 (1933), 244-252; E. J. Bickerman, “The Septuagint as a Translation” PAAJR 28 (1959), 1-39; R. A. Martin, “Some Syntactical Criteria of Translation Greek” VT 10 (1960), 295310; R. A. Martin, Syntactical Evidence of Semitic Sources in Greek Documents (SCS 3), Mis2. Bilingualism and Diglossia, Linguistic Variety, and Translation Language

185

12. Hebraisms

Linguistic-literary authority, all the more relevant in the case of a sacred book, played an important role in perpetuating features of the Septuagint version of the Pentateuch —the first books to be translated—in later works, both translations of books from the Hebrew Bible and works written directly in Greek, such as the New Testament. 19 The degree to which translators of different Septuagint books imported Hebrew grammatical elements (Hebraisms) into the Greek by adopting stereotyped renderings is crucial to an understanding of the translation technique used in the Greek Bible, as well as for the study of the textual history of the Septuagint and Biblical Greek as a whole. 20 Concerning this last point, it should be noted that variations in the frequency of Hebraisms in the Septuagint argue against the concept of Biblical Greek as a Jewish-Greek vernacular. Indeed, even theories of “synagogue Greek” seem less helpful than an approach which underscores the Semitic features of Septuagint Greek as a translation language. 21 The first two points, Hebraisms in translation technique and in the context of the history of the Septuagint text, will be further analysed in the next two sections.

3. Hebraisms in the Context of Translation Technique Even if one accepts the model of Septuagint Greek as translation language or, to use a more radical expression, a “linguistic scenario” in which the Septuagint is treated as “a corpus, not a language,” 22 the extent and the depth of the influence of the original Hebrew remains problematic. Some scholars consider that its influence was pervasive

19.

20. 21.

22.

soula, MT 1974; C. Rabin, “The Translation Process and Character of the Septuagint” Textus 6 (1968), 1-26. See Fernández Marcos, Septuagint in Context, 15, note 51: “Apparently this translation Greek imposed its own linguistic categories on a series of later religious writers since it was considered to be a sacred language. It is sufficient to see how the translator of the Book of Sira, capable of writing Greek adorned with rhetorical figures as shown by the prologue, turned out to channels of Semitised Greek or translation Greek to begin his version of the Hebrew text. On the other hand, if we compare passages from the Septuagint (Exodus, Kings, Chronicles) with parallel passages from Josephus, there is a clear shift from the semantic calque of Hebrew in translating the Septuagint, to an imitation of Classical Greek which chiefly affected style.” See Tov, Text-Critical Use, 22. See Fernández Marcos, Septuagint in Context, 16, note 52: “[…] there is no evidence for spoken Jewish-Greek; instead there must have been a translation Greek in which some peculiar syntactic features emerged due to the source language, Hebrew-Aramaic.” The most important point to be considered here is that the so-called “translation Greek” is not uniform throughout the Septuagint text and that much of its grammar and lexicon is actually quite similar to Hellenistic Greek. Thus, the expression should not be used too freely to designate a written version of a purported spoken “Jewish-Greek.” This argument is based upon and expands views already expressed by Thackeray: “One of the strongest arguments which may be adduced to disprove the existence of ‘Jewish-Greek’ as a separate dialectical entity is the striking contrast between the unfettered original Greek writings of Jewish authorship and the translations contained in the Greek Bible.” See Thackeray, Grammar, vol. I, 27. See also Aitken, “Language,” 26. Aitken, “Language,” 30.

186

3. Hebraisms in the Context of Translation Technique

12. Hebraisms

and continuous, not only in the field of syntax, 23 but also in vocabulary. 24 Shying away from an all-encompassing Semitized syntax, other authors have adopted a more cautious position that stresses the role of recurrent elements of Semitic construction in creating the impression of a Hebraized style. 25 This question should be considered from several angles, taking into account the tendency of the translators to use increasingly slavish literalisms in order to render Hebrew structures, 26 a propensity in which ideology played a substantial role. 27 Even in the case of a collection of books such as the Pentateuch, the earliest translated books, distinguishing “free translation” from “literalism” remains problematic. This issue is in turn related to the distinction between

23. Johan Lust, Erik Eynikel, and Katrin Hauspie (LEH, xviii) offer the following short assessment of this question: “Although it may be based on it, LXX Greek cannot be simply characterized as Koine Greek. It is first of all translation Greek. This is most obvious at the level of syntax and style. The order of the words in the translation most often closely sticks to that of the Hebrew original. In fact, in many passages, the Hebrew and the Greek can be put in parallel columns, word by word. The result is that the syntax of the LXX is Hebrew rather than Greek. No classical author and hardly any author using Koine Greek would have written sentences the way they are composed in the first Bible translation. The translation obviously paid more attention to the Semitic source language than to the Greek target language. They did not create an artistic Greek literary composition, but chose to stay as closely as possible with the Semitic original. This led to what is usually called ‘Hebraisms’ or ‘Semitisms’, which might be better referred to as ‘translationisms’”. 24. For a short summary, see, again, LEH, xviii-xix. 25. For such a reading of the evidence of the Greek Pentateuch, see T. V. Evans, Verbal Syntax in the Greek Pentateuch. Natural Greek Usage and Hebrew Interference, Oxford 2001. This idea was already expressed by Thackeray: “The Hebraic character of these books consists in the accumulation of a number of just tolerable Greek phrases, which nearly correspond to what is normal and idiomatic in Hebrew.” See Thackeray, Grammar, vol. I, 29. 26. The tendency towards literal renderings should not be generalized, for the situation remains uneven within the books of the Bible. See I. Soisalon-Soininen, “Zurück zur Hebraismenfrage” in: D. Fraenkel / U. Quast / J. W. Wevers (eds.), Studien zur Septuaginta Robert Hanhart zu Ehren aus Anlaß seines 65. Geburtstages (Abhandlunger der Akademie der Wissenschaften in Göttingen 190, Mitteilungen des Septuaginta-Unternehmens 20), Göttingen 1990, 37: “Daneben verläuft aber auch eine andere Linie, vertreten durch die Übersetzungen von Iob und Prov., später in einer anderen Weise durch Symmachos. Diese Übersetzungen sind ja viel freier als die des Pentateuch.” 27. See Fernández Marcos, Septuagint in Context, 24: “The biblical Greek adopted by the translators of the Pentateuch became a sort of sub-language which later the translators or the authors of pseudoepigrapha, if they were bilingual, imitated.” The most conspicuous examples of the tendency toward the development and radicalization of this literal “sub-language” are perhaps the Greek versions of Aquila and Theodotion, as has been noted by Ilmari SoisalonSoininen (“Hebraismenfrage,” 36): “Meinerseits bin ich überzeugt, daß die ältesten Übersetzer instinktmäßig gearbeitet haben. Sonst wäre es schwer, die in ein und demselben Buch vorkommende mannigfaltige Wiedergabe derselben Ausdrücke und die jeweils einzeln auftauchenden, feinsinnigen, echt griechischen Übersetzungen zu erklären. Es ist aber durchaus möglich, daß schon die späteren Septuaginta-Übersetzer eine bewußte Tendenz bezüglich gewisser Übersetzungsweisen hatten. Eine bewußte Tendenz in Richtung einer sklavisch wortgetreuen Wiedergabe ist deutlich in der Übersetzung von Theodotion zu sehen; bei Aquila wird sie dann noch ausgeprägter.” 3. Hebraisms in the Context of Translation Technique

187

12. Hebraisms

“idiomatic” and “unidiomatic”—hence potentially “Hebraizing”—constructions. 28 With these caveats in mind, 29 it is possible to approach the different types of Hebraism from a critical standpoint. Most authors adopt a system of classification similar to the one, already present to some extent in the earliest grammars of Septuagint Greek, 30 devised by Robert Helbing 31 that divides possible Hebraisms into syntactical, lexical, phraseological, and stylistic categories. This classification should be completed by including the related issue of free versus literal renderings of textual units as well as criteria developed to establish literalness, or lack thereof, in a given text. 32 These criteria can be summarized as follows: 33 (1) internal consistency, (2) representation of the constituents of Hebrew words by individual Greek equivalents, (3) word order, (4) quantitative representation, and (5) linguistic adequacy of lexical choices. Some of these criteria are more closely related to the question of Hebraisms, which are more generally associated with the field of translation technique, than others. This is true of items 1, 2, 3, and 5. 34 Caution, however, should be applied regarding several of these points. For instance, it is possible to account for internal consistency in the rendering of a Hebrew structure by a recurrent Greek construction without appealing to Hebraisms. 35 By the same token, books otherwise characterized by slavish literal trans28. For instance, in the translation of Job, the ‫ְב‬+ infinitive construct structure is rendered only once by the typically Hebraizing Greek ἐν τῷ + infinitive, whereas it is translated no less than ten times using an ὅτε clause. This constitutes a good example of how a free translation can become systematic or mechanic, features usually attributed to slavish literal renderings. See Soisalon-Soininen, “Hebraismenfrage,” 45. 29. For an example of this cautious approach in the exposition and critical analysis of categories of Hebraisms see Soisalon-Soininen, “Hebraismenfrage.” 30. R. Helbing, Grammatik der Septuaginta, Göttingen 1907; Thackeray, Grammar. 31. R. Helbing, Die Kasussyntax der Verba bei den LXX. Ein Beitrag zur Hebraismenfrage und zur Syntax der Koine, Göttingen 1928. For an extensive approach to syntactical and lexical Hebraisms, see Walters, Text of the Septuagint, 141-154. 32. The classic work on this question is Barr, Typology. See also Tov / Wright, “ComputerAssisted Study,” and, for a more recent summary, Tov, Text-Critical Use, 20-24. 33. Following Tov, Text-Critical Use, 20-24. James Barr (Typology, 264) gives a similar series of criteria or elements: “1) the division into elements or segments, and the sequence in which these elements are represented; 2) the quantitative addition or subtraction of elements; 3) consistency or non-consistency in the rendering, i. e. the degree to which a particular versional term is used for all (or most) cases of a particular term of the original; 4) accuracy and level of semantic information, especially in cases of metaphor and idiom; 5) coded ‘etymological’ indication of formal/semantic relationships obtaining in the vocabulary of the original language; 6) level of text and level of analysis.” 34. Though Emanuel Tov gives cautionary remarks on the subjective nature of this criterion, lexical Hebraisms have been a fundamental line of research. See Tov, Text-Critical Use, 24. 35. See note 28 above for Soisalon-Soininen’s remarks on this point. Consistency in the rendering of Hebrew words (even when they generate unidiomatic Greek clauses) has also been studied by G. Marquis. See G. Marquis, “Consistency of Lexical Equivalents as a Criterion for the Evaluation of Translation Technique as Exemplified in the LXX of Ezekiel” in: C. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies Jerusalem 1986 (SCSt 23), Atlanta, GA 1987, 405-424. Marquis’ highly quantitative statistical approach to the phenomenon of consistency has been criticized by Soisalon-Soininen, who insists on the need to distinguish factors affecting the number of lexical choices that may alternately appear

188

3. Hebraisms in the Context of Translation Technique

12. Hebraisms

lations may exhibit free alternation of synonyms, and hence inconsistency. 36 In regards to item 2, a point strongly related to the appearance of syntactic Hebraisms, it is necessary to take into account the heterogeneous nature of much of the evidence when analyzing occurrences from a purely statistical point of view. Authors have repeatedly noted the rendering of ‫ ְב‬+ infinitive construct + suffix into Greek (with ἐν in place of the Hebrew preposition, in analogy to its use with nouns) as an example of a persistent motif characteristic of slavish translation producing a syntactic Hebraism. 37 Concerning word order, some authors have presented quite optimistic views on its usefulness as a criterion for identifying slavish translations said to reflect a Hebrew (and hence Hebraizing) word order 38 regarded as unnatural in Greek prose. Lexical adequacy may be visible in extreme renderings, such as the systematic translation of ‫ נתן‬by διδόναι, even in inappropriate contexts. 39 These shifts, together with transcriptions and use of Hebrew loanwords, constitute the main source of lexical Hebraisms in the Septuagint and are usually, though not always, linked to a principle of consistency. These brief remarks point to new lines of research opened up in the area of Hebraisms by the study of translation techniques. They also, mainly through Ilmari SoisalonSoininen’s cautionary observations, give pause for thought concerning the many difficulties inherent in the task of distinguishing between idiomatic and unidiomatic renderings and the risks involved in making far-reaching generalizations on constructions and usages. Finally they point to the need for specialized studies in well-delimited corpora using comparative methods (between different translators and between different texts). 40

36.

37.

38.

39.

40.

“free” or “consistent” and on the need to contextualize and particularize cases and usages. See Soisalon-Soininen, “Hebraismenfrage.” Soisalon-Soininen (“Hebraismenfrage,” 45) comments on lexical alternation within the same book (e. g. the translation of ‫ חרב‬as either μάχαιρα or ῥομφαία in Judges). Although, given its heterogeneous nature, translation changes between books are frequent in the Septuagint, some notable lexical consistencies throughout the entire body of the work can be found. See Tov, Text-Critical Use, 21 and note 33. Tov, Text-Critical Use, 21; Tov and Wright, “Computer-Assisted Study,” 159. The construction is also analyzed by Soisalon-Soininen (“Methodologische Fragen der Erforschung der Septuaginta-Syntax” in: Cox, VI Congress, 1987, 425-444. A critical evaluation of the evidence, taking into account the heterogeneity of the functions underlying the syntactic structure, is presented in Soilason-Soisinen, “Hebraismenfrage,” 46. See LEH (xviii): “[…] in many passages, the Hebrew and the Greek can be put in parallel columns, word by word.” A similar opinion is expressed by G. Marquis. See G. Marquis, “Word Order as a Criterion for the Evaluation of Translation Technique and the Evaluation of Word-Order Variants as Exemplified in LXX-Ezekiel” Textus 13 (1986), 59-90. By offering a differentiated approach to the renderings of ‫נתן‬, Soisalon-Soisinen demonstrates the dangers inherent in drawing quick conclusions and making broad generalizations. See Soisalon-Soisinen, “Hebraismenfrage,” 49. See Soisalon-Soisinen, “Hebraismenfrage,” 51. 3. Hebraisms in the Context of Translation Technique

189

12. Hebraisms

4. Hebraisms: Review and Suggested Research Directions Having presented a theoretical approach to research in Hebraisms from both the historical point of view of Biblical Greek studies and from the current perspective of translation science, I will now, for the sake of illustration, give a short selection of possible Hebraisms that have been analyzed by scholars using different methods, dividing them into two subsections. The examples in the first subsection show that the study of corpus languages, Septuagint Greek and Biblical Hebrew, cannot be separated from the history of the text, either with regard to translation techniques (whether in different books or in successive translations or revisions of the same book) 41 or to the use of possible Hebraisms to detect variant Hebrew texts in the Vorlage of the Septuagint. 42 The examples in the second subsection illustrate the difficulty of identifying Hebraisms solely on the basis of a comparison with the Greek language.

4.1 Possible Hebraisms, Textual Criticism, and Textual History A. 1Sam 2:25 43 ἐὰν ἁμαρτάνων ἁμάρτῃ (LXX) ‫( ִאם־ ֶיֱחָטא‬MT). In this example, the Greek imitates a typical Hebrew construction by means of a paranomastic usage of the infinitive absolute. This typical syntactic Hebraism 44 authorizes the textual critic to propose a variant Vorlage: ‫אם חטוא יחטא‬ B. Exod 20:24 45 ἐν παντὶ τόπῳ οὗ ἐὰν ἐπονομάσω τὸ ὄνομά μου ἐκεῖ (LXX) þ ‫( ְבָּכל־ַהָמּקוֹם ֲאֶשׁר אַ ְזִכּיר ֶאת־ְשִׁמי‬MT). Here, the Greek uses the retrospective adverb ἐκεῖ to render a Hebrew clause, ‫ֲאֶשׁר‬, with a locative meaning; a variant Vorlage regarding the adverb in the Hebrew original, perhaps connected to a parablepsis between ‫ שׁמי‬and ‫שׁם‬, can therefore be proposed: ‫אשׁר אזכיר את שׁמי שׁם‬. C. 1Kings 1:52 εἰ πεσεῖται τῶν τριχῶν αὐτοῦ (LXX) ‫( ל ֹא־ ִיפּ ֹל ִמַשֲּׂע ָרתוֹ אָ ְרָצה‬MT). This instance highlights the importance of Hebraisms as a means for proposing possible variant readings of the Hebrew text. In this case, the use of εἰ in the Septuagint is unidiomatic in Greek, but it would be a literal rendering of the Hebrew “hanging protasis” with ‫ אם‬in oath formulas. Indeed, it is quite hard to explain this translation as a

41. This is especially evident in books, such as Samuel-Kings, that have undergone a καὶ γε revision, a procedure that introduces a higher level of literal and unidiomatic renderings. It is also apparent in fragments of the Hexapla and in Hexaplaric elements interpolated in the Greek main text. Some of these texts give clear evidence of the growing desire of translators and revisers to find a perfect word-by-word agreement between the Greek and the Hebrew. See I. Soisalon-Soininen, “Der Charakter der asterisierten Zusätze in der Septuaginta” AASF Series B 114 (1959), 46-106. 42. As has been shown by Tov (Text-Critical Use, 83-85). 43. Tov, Text-Critical Use, 84. 44. This, in turn, poses a challenge to studies of consistency, as the paranomastic infinitive absolute can be expressed by different Greek constructions, according to the books and the contexts in which it appears. See E. Tov, “Renderings of Combinations of the Infinitive Absolute and Finite Verbs in the LXX – Their Nature and Distribution” in: Fraenkel / Quast / Wevers, Studien zur Septuaginta, 64-73. 45. Tov, Text-Critical Use, 84.

190

4. Hebraisms: Review and Suggested Research Directions

12. Hebraisms

variant introduced by a Greek writer, even though the Masoretic text attests a standard negative, ‫לא‬. 46 D. 2Kings 8:8 λαβὲ ἐν τῇ χειρί σου μαναα (LXX B) λαβὲ εἰς τὴν χεῖρά σου δῶρα (LXXL) ‫( ַקח ְבּ ָי ְדָך ִמ ְנָחה‬MT). This is just one example of the many cases of transcriptions of Hebrew nouns (as opposed to place and personal names) into Greek. In biblical books with several layers of redaction-recension, such as Samuel-Kings, the translation of a word or the introduction of a Hebraizing transcription or neologism into Greek can be used as a criterion for defining textual typology in the history of the Septuagint. In this instance, the reading of the Lucianic text is actually a translation. By comparing this rendering with secondary versions and similar treatments of the same word in Kings, it can be proposed that the Lucianic text is closer to the Old Greek and that Hebraizing transcriptions were a characteristic feature of the καὶ γε revision of Kings. 47

4.2 Possible Hebraisms and the Limits of the Greek Language E. Judg 9:45 48 καὶ Αβιμελεχ ἐπολέμει ἐν τῇ πόλει ὅλην τὴν ἡμέραν ἐκείνην (LXX) ‫( ַוֲאִביֶמֶלְך ִנְלָחם ָבִּעיר כֹּל ַהיּוֹם ַההוּא‬MT). This example illustrates the limits of unidiomatic or incorrect renderings. The Greek expression is perfectly idiomatic, but it would change the intended meaning of the original, “against the city,” to “in the city.” F. Gen 4:2 49 καὶ προσέθηκεν τεκεῖν τὸν ἀδελφὸν αὐτοῦ τὸν Αβελ (LXX) ‫ַותּ ֶֹסף‬ ‫( ָלֶל ֶדת ֶאת־אִָחיו ֶאת־ָהֶבל‬MT). Here, a literal translation of the Hebrew periphrasis produces aberrant Greek due to semantic incompatibility. In other cases, the same rendering of the periphrasis could produce good Greek, although it might induce a semantic change. 50 46. As Tov remarks, some scholars would contest the claim that Hebraisms in the Greek text always reflect a Hebrew element in the Vorlage. In all fairness, cases such as 1Kings 1:52 should be analyzed in their textual and grammatical context, weighing the information provided by the Hebraizing construction with the other textual and linguistic coordinates of the passage, without automatically assuming the presence of a Hebraism. In this particular case, the slavish rendering of the Hebrew ‫ ִמן‬by the Greek partitive genitive, which, unidiomatically, would work as the subject of the sentence, provides additional evidence in support of a Hebraizing clause. See Tov, Text-Critical Use, 84. See also, A. Piquer Otero, “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” Henoch 30 (2008), 80-93. 47. This question was already raised by Thackeray. See Thackeray, Grammar, vol. I, 31-38. More recently, Hellenized Semitic words in the Septuagint, including transcriptions and transliterations, were discussed by Peter Walters (Text of the Septuagint, 155-196). For a more extensive treatment of the question of transcription versus translation with regards to the history of the Septuagint, see A. Piquer Otero / P. Torijano Morales / J. Trebolle Barrera, “Septuagint Versions, Greek Recensions and Hebrew Editions. The Text-Critical Evaluation of the Old Latin, Armenian, and Georgian Versions in III-IV Regnorum” in: H. Ausloos et al. (eds.), Translating a Translation. The LXX and Its Modern Translations in the Context of Early Judaism (BETL 213), Leuven/Paris/Dudley, MA 2008, 251-281. 48. Soisalon-Soisinen, “Hebraismenfrage,” 40. 49. Soisalon-Soisinen, “Hebraismenfrage,” 42. 50. See the discussion of Gen 18:29 in Soisalon-Soisinen, “Hebraismenfrage,” 42. 4. Hebraisms: Review and Suggested Research Directions

191

12. Hebraisms

In conclusion, analysis and classification of Hebraisms, both syntactic and lexical, has not changed substantially since the pioneering research of the early twentieth century. 51 What has become apparent in recent years is the need to understand these Hebraisms in the context of the history the Septuagint, its diversity, and the multifaceted and nuanced practices of the translators who worked in a specific literary, ideological and intertextual milieu.

51. See, e. g., the monographs and articles in notes 14 and 31.

192

4. Hebraisms: Review and Suggested Research Directions

13. Etymological Translations in the Septuagint Hans Ausloos / Bénédicte Lemmelijn

1. Theoretical Prolegomena In his seminal contribution “Etymology and the Old Testament”, James Barr wrote: “Etymology is particularly important for the identification and elucidation of rare words and hapax legomena. The Hebrew Bible has many such rare words, and these can often be elucidated only through comparison with words in Ugaritic, Akkadian, Arabic and other cognate languages.” 1 In practice, one often observes that the Septuagint is used as a source for the etymological reconstruction of problematic Hebrew terms. 2 However, although the Septuagint is undoubtedly a very ancient textual witness, this procedure is not always legitimate. In translating the Hebrew parent text, the Septuagint translators were confronted with problematic Hebrew lexemes. Even for them, ascertaining the precise meanings was a daunting task. In keeping with Barr’s statement that “etymological consciousness was […] very strong in the ancient world, and notably so in the milieu of the Bible, of early Judaism and of early Christianity,” 3 there can be no doubt as to the fact that the Septuagint translators themselves “thought etymologically.” Nevertheless, their method of etymological rendering can hardly be described in terms of modern scholarly etymology in which the historical perspective, the systematic classification of cognate languages and the systematic interest in phonological correspondences between different languages play an important role. 4 Moreover, it is beyond doubt that etymological interpretation was never a goal in its own right for the Septuagint translators. They were not interested in philological research on the Hebrew language system; neither were they concerned with writing historical Hebrew grammars or dictionaries. Their exclusive aim was to translate the Hebrew biblical texts. De facto, translators proceeded etymologically, interpreting each word into the target language. In other words, they looked for the word’s root in order to find its meaning and then searched for a suitable translation equivalent. 5 Nevertheless, we use the term “etymological rendering” in a particu1.

2.

3. 4. 5.

J. Barr, “Etymology and the Old Testament” in: J. Barr et al., Language and Meaning. Studies in Hebrew Language and Biblical Exegesis (Oudtestamentische studiën 19), Leiden 1974, 1-28, in particular 2. Cf. H. Ausloos, “Hapax Legomena, the Septuagint and Hebrew Lexicography” in: M. Peters (ed.), XIV Congress of the International Organization for Septuagint and Cognate Studies – Helsinki 2010 (Society of Biblical Literature. Septuagint and Cognate Studies), Atlanta, GA 2013, 291-300. On this distinction, see Barr, “Etymology,” 2. Barr, “Etymology,” 3. Etymological renderings can therefore be considered a specific part of the so-called linguistic exegesis that characterizes each translation process. Cf. E. Tov, The Text-critical Use of the Septuagint in Biblical Research (second edition revised and enlarged, Jerusalem Biblical Stu1. Theoretical Prolegomena

193

13. Etymological Translations in the Septuagint

lar sense, being related to cases in which the Vorlage was problematic to some extent, such as, for instance, with respect to Hebrew words the translators did not understand, instances in which they were not sure of the precise meaning of a word, and cases in which the common usage of a term seemed to make little sense. However, this does not imply that the Septuagint translators carried out etymological “research” for each problematic term. They sometimes simply chose to leave difficult words untranslated or opted for a transliteration. 6

2. Etymological Renderings Are Root-Linked The tools the Septuagint translators had at their disposal for their etymological “research” were undoubtedly rather scarce. As a consequence, their etymological interpretations may appear “conjectural” to scholars today, although it is possible that they did not seem so to the translators themselves. 7 After all, it is possible that the translators’ knowledge of Hebrew was—at least to some extent—better than ours and that they could have been familiar with lexemes no longer known to us. When confronted with a problematic Hebrew lexeme, the translator’s etymological “procedure”—Emanuel Tov speaks of “etymological exegesis” 8—mainly consisted in analyzing the word’s component morphemes, thus trying to relate it to a (bi- or triconsonantal) root known to the translator. As a result, difficult Hebrew words were often rendered by (similar) Greek words that were used elsewhere in the Septuagint to translate the more familiar Hebrew root. In this respect, one could talk of “root-linked consistency”. 9 In any case, the most fundamental observation one can make at this point is that each etymological rendering is by definition a root-linked interpretation. Barr therefore rightly states: “Etymologizing is the procedure of interpreting a word by reference to the meaning of another (usually a better-known) Hebrew word which had a similarity to it and could […] be taken as its root; or a word may be analysed into separate units from which it is taken to have been made up”. 10 Without a doubt, hapax legomena would be particularly problematic. 11

6.

7. 8. 9. 10. 11.

dies 8), Jerusalem 1997, 45; E. Tov, Textual Criticism of the Hebrew Bible (third edition revised and expanded), Minneapolis, MN 2012, 118-119. In the process of linguistic exegesis, the translator first analyzes the components of the lexeme, in order to identify the root it derives from. Second, he establishes the meaning of the word. Third, he chooses a Greek equivalent. Cf. E. Tov, “Did the Septuagint Translators Always Understand their Hebrew Text?” in: E. Tov, The Greek and Hebrew Bible. Collected Essays on the Septuagint, Leiden/Boston, MA/ Köln 1999, 203-218, in particular 204-205. See, e. g., θαλπιωθ for ‫ תאלפיות‬in Cant 4:4, which, according to A. M. Honeyman (“Two Contributions to Canaanite Toponymy” Journal of Theological Studies 50 (1949), 50-52) could be related to the root ‫לפא‬. Tov (“Septuagint Translators,” 216-218) considers etymological renderings as a type of conjectural rendering. Cf. Tov, Text-critical Use, 172-180. Cf. infra, p. 199. J. Barr, Comparative Philology and the Text of the Old Testament. With Additions and Corrections, Winona Lake, IN, 1987, 253. On the rendering of Hebrew hapax legomena in the Septuagint, see, in particular, H. Aus-

194

2. Etymological Renderings Are Root-Linked

13. Etymological Translations in the Septuagint

Thus, in his etymological rendering, the translator generally tried to link the word with an existing (and more common) root. Sometimes, however, he seems not to have found an existing root that could be related to the term in question. In these cases, one observes that he tried to make conjectural changes to the term. Tov, in this respect, distinguishes between root-linked renderings and etymological guesses. 12 As indicated above, in a root-linked rendering, the translator tries to link a lexeme to an existing root. When “etymologically guessing,” the translator seems to have tried to manipulate the lexeme, for example by conjecturally changing one or more consonants or their sequence in order to relate it to an existing root. In our opinion, Tov’s distinction should therefore be nuanced to some extent. First, “etymological guesses” also seem to be root-linked. Second, in speaking about an “etymological guess,” one risks making the implicit assumption that the problematic Hebrew lexeme represents the “best” textual variant or witness, therefore not taking into account that the Septuagint translator could have been using a “better,” “more original,” and perhaps unproblematic Vorlage, and that the Masoretic text variant could be due to textual corruption. Against the background of these theoretical observations, the following section will focus on concrete examples in the Septuagint text, in order to illustrate and elucidate different aspects and problems in the etymological rendering of difficult Hebrew words.

loos / B. Lemmelijn, “Rendering Love. Hapax Legomena and the Characterisation of the Translation Technique of Song of Songs” in: H. Ausloos / J. Cook / F. García Martínez / B. Lemmelijn / M. Vervenne (eds.), Translating a Translation. The LXX and its Modern Translations in the Context of Early Judaism (Bibliotheca Ephemeridum Theologicarum Lovaniensium 213), Leuven/Paris/Dudley, MA 2008, 43-61; H. Ausloos, “The Septuagint’s Rendering of Hebrew Hapax Legomena and the Characterization of its ‘Translation Technique’. The Case of Exodus” Acta Patristica et Byzantina 20 (2009), 360-376; H. Ausloos / B. Lemmelijn, “Characterizing the LXX Translation of Judges on the Basis of Content-Related Criteria. The Greek Rendering of Hebrew Absolute Hapax Legomena in Judg 3, 12-30” in: H. Ausloos / B. Lemmelijn / J. Trebolle Barrera (eds.), After Qumran: Old and Modern Editions of the Biblical Texts – The Historical Books (Bibliotheca Ephemeridum Theologicarum Lovaniensium 246), Leuven/Paris/Dudley, MA 2012, 171-192. For a theoretical framework, see E. Verbeke, “The Use of Hebrew Hapax Legomena in Septuagint Studies. Preliminary Remarks on Methodology” in: H. Ausloos / B. Lemmelijn / M. Vervenne (eds.), Florilegium Lovaniense. Studies in Septuagint and Textual Criticism in Honour of Florentino García Martínez (Bibliotheca Ephemeridum Theologicarum Lovaniensium 224), Leuven/Paris/Dudley, MA 2008, 507-521. 12. Tov, “Septuagint Translators,” 216-218: “Reliance on etymology is a known procedure for tranlators [sic], and such reliance is called conjectural when the translation is based on a certain manipulation of the consonants, sometimes involving disregard of prefixes or suffixes. In all cases the Hebrew words involved are understandably difficult” (218). See also W. E. Glenny, Finding Meaning in the Text. Translation Technique and Theology in the Septuagint of Amos (Supplements to Vetus Testamentum 126), Leiden/Boston, MA 2009, 100-105. This author follows Tov’s distinction between root-linked renderings and etymological guesses. 2. Etymological Renderings Are Root-Linked

195

13. Etymological Translations in the Septuagint

3. Different Types of Etymological Renderings 3.1 Linking the Term to an Actually Existing Root In Exod 16:31, the manna the Israelites get to eat is described as ‫( כצפיחת בדבש‬NRSV: “like wafers made with honey”). The noun ‫צפיחת‬, which appears only once in the Hebrew Bible, is, in recent commentaries, most often interpreted as “flat cake.” 13 This English translation is in line with the Septuagint translator’s root-linked etymology, indeed it is dependent on it. In fact, the Greek translator seems to have related the “difficult” Hebrew noun to the “more common” root ‫“( צפח‬making flat”), by opting for the equivalent ἐγκρὶς ἐν μέλιτι (“cake [with honey]” 14). The latter term is also used by ancient Greek authors to indicate a sort of cake. 15 In any case, without being able to achieve certainty regarding the precise meaning of the term ‫צפיחת‬, the translation equivalent that the Septuagint suggests does certainly make sense in the context. In Judg 3:23, after having killed the Moabite king Eglon, the Israelite Ehud “went out into the vestibule” (NRSV). The Hebrew formula is difficult to interpret: ‫ויצא אהוד‬ ‫המסדרונה‬. The hapax legomenon ‫ המסדרונה‬is particularly problematic. The B-text of Judg 3:23 reads τοὺς διατεταγμένους. Τοὺς διατεταγμένους can probably be considered as an accusative of direction with ἐξέρχομαι. 16 As such, the Greek equivalent interprets the final consonant (‫ה‬-) as a ‫ה‬-locale. In any case, the first ‫ה‬- is considered as a definite article (τούς). The participle διατεταγμένος is a passive participle perfectum form of διατάσσω, 17 a verb that occurs twenty-five times in the Old Testament 18 and that is used as an equivalent for no fewer than eight Hebrew verbs. 19 Therefore, it is not improbable that the translator related the Hebrew hapax legomenon to the root ‫“( סדר‬to order, to arrange”). 20

13. Cf. C. Houtman, Exodus I: Exodus 1:1–7:13 (Historical Commentary on the Old Testament), Leuven 1993, 352. 14. A. Pietersma / B. G. Wright (eds.), A New English Translation of the Septuagint and the Other Greek Translations Traditionally Included under That Title, New York, NY/Oxford 2007, 63. 15. According to Stesichorus (seventh to sixth century BCE), Pherecrates (fifth century BCE), and Antiphanes (fourth century BCE). See the references in H. G. Liddell / R. Scott, A GreekEnglish Lexicon. With a Revised Supplement, Oxford 1996, 473. Note that J. Lust, E. Eynikel, and K. Hauspie, (LEH, 170) like Liddell and Scott translate the term as “a cake made with oil and honey.” T. Muraoka (GELS, 188) translates it as a “kind of cake.” 16. Cf. W. Kraus / M. Karrer (eds.), Septuaginta Deutsch: Das griechische Alte Testament in deutscher Übersetzung, Stuttgart 2009: “und ging hinaus zu den aufgestellten Posten.” Contra Tov, Text-critical Use, 178. 17. Cf. G. Delling, “διατάσσω” in: G. Kittel / G. Friedrich (eds.) Theologisches Wörterbuch zum Neuen Testament 8 (1969) 34-36. 18. LEH, 147. 19. HR, 313. 20. Cf. Tov, Text-critical Use, 178: “‫ םסדרון‬is rendered as if it were ‫( הםסדרים‬those who arranged)”. Cf. supra with regard to the interpretation of the Hebrew term.

196

3. Different Types of Etymological Renderings

13. Etymological Translations in the Septuagint

In Eccl 10:17, the lexeme ‫ בשתי‬is a hapax legomenon, which is currently often interpreted as a noun (‫)שתי‬, meaning “the act or place of drinking,” 21 preceded by the preposition ‫ב‬. However, the Septuagint translator apparently related it to the root ‫“( בוש‬to be ashamed”), and therefore rendered it by the verb αἰσχύνομαι.

3.2 Existing Roots and the Translator’s Creativity As indicated in the introduction to the present chapter, current etymological research draws heavily on cognate languages. Although the Septuagint translators sometimes made use of their knowledge of cognate languages in order to determine the meaning of problematic terms in their Hebrew parent texts, comparative philology as a discipline did not yet exist at the time of the Greek translators. 22 Nevertheless, Aramaic, in particular, seems sometimes to have been used by translators as a point of reference for their etymological “research.” 23 In Cant 5:3, for example, the root ‫ טנף‬is a hapax legomenon in the Hebrew Bible. The Septuagint translator’s rendering of the verb by μολύνω seems to have been influenced by the Aramaic (and Akkadian) use of the root: “to soil.” Be that as it may, his interpretation fits entirely into the parallel structure of the verse in which the second colon is antithetical to the first: “I have removed my garment, how could I put it on again? / I have washed my feet, how could I soil them again?” 24 Finally, it must be noted that the etymological interest of the translators was not limited exclusively to problematic terms. Sometimes, by relying on an uncommon etymology, the translator seems to have been able to create new meaning in the text. This situation can be illustrated by Gen 5:29. In the Masoretic text, Lamech names his son “Noah,” and offers an etiological interpretation 25 he is called Noah (‫ )נח‬because he will comfort (‫ )ינחמנו‬his people. The reason why Noah is characterised as someone who will comfort is not clear from the immediate context. Only later does it become evident that Noah alone will be able to save humanity. The Septuagint’s rendering however does not play on any sound similarity nor does it render the etiological link between Noah’s name and his comforting role. In the Septuagint, Noah’s proper name is merely transliterated (Νῶε). More important is the fact that the verb ‫ נחם‬was not rendered by the commonly used equivalent παρακαλέω (“to comfort”), but rather by the Greek hapax legomenon διαναπαύω (“to give respite from”). It therefore seems that the Septuagint etymologically related the form ‫ ינחמנו‬to the verb ‫“( נוח‬to rest”), thereby creating new meaning in the text.

21. L. Koehler / W. Baumgartner, Hebräisches und Aramäisches Lexikon zum Alten Testament. (Unveränderter Nachdruck der dritten Auflage 1967–1995), Leiden/Boston, MA 2004, 1539. 22. On this, see Barr, Comparative Philology, 76-94. 23. Cf. Tov, Text-critical Use, 179. 24. Cf. P. W. T. Stoop-van Paridon, The Song of Songs. A Philological Analysis of the Hebrew Book ‫( ִשׁיר ַהִשּׁי ִרים‬Ancient Near Eastern Studies. Supplement 17), Leuven 2005, in particular 264-266. 25. On the difference between etiology end etymology, see infra p. 200-201. 3. Different Types of Etymological Renderings

197

13. Etymological Translations in the Septuagint

3.3 Root-Linked Manipulation? In searching for a possible root, translators seem, sometimes, to have changed the Hebrew text, by playing with the consonants. In some instances, they seem to have been compelled to do so because the Hebrew word was problematic; in others, no such explanation exists for the apparent manipulation. In Cant 7:10, the form ‫ דובב‬is a hapax legomenon. Some scholars interpret it as a causative form of the verb ‫“( דבב‬to speak”). 26 The Septuagint renders the term as ἱκανούμενος, a participle of the verb ἱκανόομαι (“to make sufficient”, “to qualify”). Thus, the translator probably interpreted the Hebrew word as a verbal form of the verbal root ‫“( רבב‬to be plentiful”). And indeed, in Num 16:7; Deut 1:6; 2:3; 3:26; Ezek 44:6; 45:9, ἱκανόομαι is the translation equivalent of this Hebrew term. He could therefore have relied on “root-linked consistency.” 27 One should, however, be careful here. It is also possible that the translator’s Vorlage read a different form, and that the reading of the Masoretic text was caused by the accidental misreading of the verb ‫רבב‬, or even that the reading ‫ דבב‬was due to the interchange between the consonants ‫ד‬- and ‫ר‬-. 28 In Am 2:16, however, the term ‫ אמיץ‬is not at all problematic. It occurs several times in the Hebrew Bible, in the sense of “strong.” This meaning is likewise unproblematic in Am 2:16 (MT): “he who is courageous shall flee away naked.” Nevertheless, the translator renders it as εὑρήσει (third person singular of εὑρίσκω), thus linking it “etymologically” to the root ‫“( מצא‬to find”). For this purpose, it seems that he had to manipulate his Vorlage. 29 Based on these findings, it can be argued that not only the so-called Old Greek, but also the early recensions, seem to have made use of etymologically root-linked renderings and, moreover, seem to have “played” with the lexemes by manipulating their consonants. In Job 19:20, the problematic term ‫ אתמלטה‬is rendered in the Septuagint by the verbal form ἔχεται, which can hardly be considered a root-linked etymological interpretation. It seems here that the translator used a rather vague equivalent because he was not sure about the precise meaning of the Hebrew text. Symmachus, on the contrary, seems to have related the Hebrew form to the root ‫“( מרט‬to pull [out hair]”), thus reflecting two consonants, in translating it as καὶ ἐξέτιλλον. 30

26. 27. 28. 29.

KJV: “causing the lips of those that are asleep to speak.” Cf. infra p. 199. Cf. Tov, Text-critical Use, 245-246. However, it is equally possible that this reading, already on the level of the Septuagint’s Vorlage, was due to an erroneous metathesis. Cf. Glenny, Finding Meaning, 102-103. 30. F. Field, Origenis Hexaplorum quae supersunt, sive veterum interpretum graecorum in totum Vetus Testamentum fragmenta, Oxford 1875, vol. 2, 36. E. P. Dhorme, (A Commentary on the Book of Job, London 1967, 280) mentions the connection between the verbs ‫ מרט‬and ‫מלט‬, which he considers synonymous.

198

3. Different Types of Etymological Renderings

13. Etymological Translations in the Septuagint

3.4 Etymological Renderings and “Root-Linked Consistency” As touched upon above, the so-called “consistency” phenomenon also requires our attention in the context of root-linked etymological interpretations. With respect to the Septuagint’s rendering of hapax legomena in particular, the root-linked type of etymological renderings described above could be seen as part of the broader phenomenon of the so-called stereotyping tendency of the Septuagint translators. 31 It is generally accepted that this tendency was the rule rather than the exception for the Septuagint translators, who strived to “consistently” render a given Hebrew term within a particular translation. Having unsuccessfully searched for a root-linked etymology of a problematic word, the Septuagint translator often opted to render it in line with the usual rendering of the more commonly used root already known to him. In doing so, he appears to have considered the unusual term as a unique derivative of a more commonly used root that he rendered consistently in Greek. 32 Therefore, when the translator was confronted with a difficult Hebrew word, he seems to have looked for a wellknown Hebrew root that he could render with a Greek word belonging to a commonly and consistently used Greek stem. 33 However, there are also cases in which the Septuagint translator renders a Hebrew hapax legomenon with a Greek one. This is, for example, the case in Cant 6:11, where the Hebrew hapax legomenon ‫—אגוז‬a term commonly translated as “nut”—was rendered by καρύας, which is itself a hapax legomenon in the Septuagint.

3.5 “Popular” Root-Linked Etymology In some instances, however, the root-linked etymological interpretation can give the impression of being “unusual” in the light of scholarly lexical research. 34 The Septuagint translator sometimes relates a word to a Hebrew root that current research does not consider to be its etymological background. This phenomenon has been called “popular etymology,” most notably by Tov. 35 Here, the translator does not appear to be looking for a linguistically related etymology—although to a certain extent the consonants of the root will be identical—but rather to play with sound and meaning. 36 By way of illustration, one can refer to the Septuagint’s frequent rendering of ‫ אהל מועד‬as ἡ σκηνή τοῦ μαρτυρίου. Tov considers this an example of popular etymology, based,

31. Cf. E. Tov, “Septuagint Translators,” 216-217. 32. Cf. supra on Cant 7:10. 33. Within root-linked etymological renderings, Tov (“Septuagint Translators,” 216-218) distinguishes between a rendering which “reflects a concern for the consistent representation of Hebrew word groups with equivalent Greek word groups” and a rendering that “disguises the translator’s ignorance of the exact meaning of the word”. 34. Cf. Tov, Text-critical Use, 173. 35. Tov, Text-critical Use, 177. Barr (“Etymology,” 23-24) however uses the terminology in a different way. See infra n. 43. 36. This type of “etymological rendering” seems to be part of the frequently used literary technique of wordplay. In this respect, see V. Kabergs / H. Ausloos, “Paronomasia or Wordplay? A Babylonian Confusion. Towards A Definition of Hebrew Wordplay” Biblica 93 (2012), 1-20. 3. Different Types of Etymological Renderings

199

13. Etymological Translations in the Septuagint

as it is, on the link between the two final consonants ‫עד‬- with the noun ‫“( עד‬witness”), which was often translated as μαρτύριον. 37

3.6 Intentional Etymologizing in Unproblematic Lexical Situations Common Hebrew words sometimes appear to have been used with specific meanings; this can be seen when the usual meanings do not make sense in particular contexts. In such cases, the translator may have perceived the terms as problematic and sought an acceptation that would make better sense in the context. To this end, he may have done some “etymological research” in order to find an appropriate root. A related phenomenon occurred when translators “etymologized” common Hebrew words well-known to them. The term ‫“( מוקש‬trap”) should have been known to the translator of Am 3:5, as it occurs several times in the Pentateuch and the historical books, as well as in the Psalms and in Wisdom literature. Nevertheless, the translator seems to have invented an etymological rendering by connecting the Hebrew word with the verbal root ‫“( יקש‬to catch a bird with a snare”), translating it as ἰξευτής (“fowler”). 38 Most probably he judged this rendering a better fit in the particular context of the verse.

4. Final remarks Two concluding remarks will complete our discussion of etymological renderings in the Septuagint. First, when dealing with etymological renderings in the Septuagint, one should take into account that the translators were working from a non-vocalised text. It is therefore not implausible that the Septuagint’s reading could actually be more original than the medieval vocalisation of the Masoretic text. In Job 39:8, for example, the lexeme ‫יתור‬, as it is vocalized in the Masoretic text (‫) ְיתוּר‬, is a hapax legomenon. As a result of the strange vocalization, scholars do not agree on whether it is a verbal form or a noun. 39 The Septuagint interprets it as a verbal form, reading κατασκέπτομαι (“to look out, to survey”). This verb occurs several times as an equivalent of the Hebrew verb ‫תור‬. Therefore it is clear that the Septuagint translator has interpreted ‫יתור‬ as a verbal form of ‫תור‬. 40 Moreover, if we disregard the Masoretic vocalisation, which, of course, was not available to the Septuagint translator, relating the consonants to the verb ‫ תור‬is not at all problematic. 41 Second, etymological renderings clearly have to be distinguished from etiologies. The term “etiology” usually refers to short passages, generally written in narrative 37. Tov, Text-critical Use, 177. 38. According to Glenny (Finding Meaning, 101), the translator interpreted ‫ מוקש‬as a piel participle of ‫יקש‬. 39. E. g. H. Tur-Sinai, The Book of Job. A New Commentary, Jerusalem 1967, 541 versus M. H. Pope, Job. Introduction, Translation, Notes (Anchor Bible 15), New Haven, CT 2008, 307. 40. The context of the verse (the parallelismus membrorum) also points in this direction. 41. Cf. R. Gordis, The Book of Job. Commentary, New Translation and Special Studies (Moreshet Series. Studies in Jewish History, Literature and Thought 2), New York, NY 1978, 456, who vocalizes the form as ‫ ָיתוּר‬.

200

4. Final remarks

13. Etymological Translations in the Septuagint

fashion, that retrospectively “explain” a toponym or a person’s name. Etiologies are therefore part of Hebrew wordplay. 42 However, although etiologies can give the impression of being etymological renderings within a narrative framework (especially when the wordplay operates on similar-sounding or similar-looking roots), they have in fact nothing to do with the linguistic search for the precise meaning of a term. 43 This chapter will have achieved its purpose if it has become clear that etymological renderings in the Septuagint occur in different lexical situations and literary contexts. The main aim of any translation equivalent is the faithful rendering of the content in the context of the Hebrew Vorlage. However, all translations must contend with the non-identical character of two language systems, and all translators are confronted with their own imperfect knowledge of the source and target languages. Thus, in manifold situations, where words were obscure, awkward, or seemed not to make sense at all, the Septuagint translator found it necessary to search for creative solutions. He therefore explored the possibilities offered by the two languages and their linguistic characteristics. This is the framework within which the origin of etymological renderings must be understood.

42. Cf. H. Ausloos / B. Lemmelijn / V. Kabergs, “The Study of Aetiological Wordplay as a Content-Related Criterion in the Characterisation of LXX Translation Technique” in: S. Kreuzer / M. Meiser / M. Sigismund (eds.), Die Septuaginta: Entstehung, Sprache, Geschichte (Wissenschaftliche Untersuchungen zum Neuen Testament 286), Tübingen 2012, 273-294. 43. Barr, “Etymology,” 23-24 characterizes this etymological explanation of proper nouns as “popular etymology,” which he considers as “a play on word-similarity, rather than a serious analysis of root meanings” (26). In our view, it is better to make a clear distinction between etiology and etymology. See also M. Garsiel, Biblical Names. A Literary Study of Midrashic Derivations and Puns, Ramat Gan 1991, 17: “One should make a clear-cut distinction between linguistic etymology and the name explanations provided by the biblical texts”; E. A. Speiser, Genesis. Introduction, Translation, and Notes (Anchor Bible Second Edition), New York, NY 1964, 232: “The explanations are symbolical; a correct, or even plausible, linguistic derivation would be purely coincidental, since the play on the name was the significant thing – aetiology rather than etymology.” 4. Final remarks

201

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila Peter J. Gentry

Uncertainties about the history of the process of the translation of the Hebrew Scriptures into Greek are responsible for lack of precision in what is meant by the term Septuagint. It is generally agreed that the Pentateuch, or Torah, was translated in Egypt early during the reign of Ptolemy II Philadelphus (285/282-246 BCE), possibly around 280 BCE if reliable patristic testimony is accepted. 1 The books in the Prophets and the Writings were translated later, most of them by 130 BCE, as is clearly indicated by the Prologue to the Greek translation of Ben Sirach (Ecclesiasticus). Questions arise about the date of translation of each of the books in the collection known as Megilloth (Ruth, Song, Ecclesiastes, Lamentations, and Esther). 2 Some of these may have been first translated after 100 BCE. Thus the term Septuagint is applicable in a technical sense only to the Greek Pentateuch, although it is generally employed in a loose manner of speaking for the Greek translation of the Jewish Scriptures as a whole. To further complicate matters, long before all the books had been translated, revisions were being made of already existing translations. The product of revising an already existing translation in a systematic way, based on certain guidelines or principles (e. g. closer quantitative alignment to the current Hebrew parent text), is properly called a recension. 3 Recensions of the original translation(s) were produced from perhaps as early as 200 BCE until 200 CE. The precise line of demarcation between original translations and revisions in this body of texts has, in fact, not yet been clearly established. The issue is particularly acute in the translation of the books of SamuelKings. Scholars are still working to prepare critical editions of the earliest obtainable text of these translations based upon careful study of all available evidence in Greek manuscripts, citations in church fathers, and early daughter translations. In both ancient and modern times different approaches to the task of translation have been adopted. Each language employs a code to “cut up” and represent the “pie” 1.

2. 3.

N. L. Collins, “281 BCE: the Year of the Translation of the Pentateuch in Greek under Ptolemy II” in: G. J. Brooke / B. Lindars (eds.), Septuagint, Scrolls, and Cognate Writings (SCS 33), Atlanta, GA 1992, 403-503. A recent and thorough re-analysis of the Letter of Aristeas and the origins of the Septuagint is in S. Honigman, The Septuagint and Homeric Scholarship in Alexandria: A Study in the Narrative of the Letter of Aristeas, London 2003. Her conclusions do not challenge a date in the early third century BCE as the proposed time of translation. See M. Harl / G. Dorival / O. Munnich (eds.), La Bible grecque des Septante, Paris 1988, 1994. That difficulties exist in analysis of the so-called καίγε group of texts does not mean we cannot accurately define terms such as recensions, as seems to be the case for L. Greenspoon, “The Kaige Recension: The Life, Death, and Postmortem Existence of a Modern—and Ancient— Phenomenon” in: M. K. H. Peters (ed.), XII Congress of the International Organization for Septuagint and Cognate Studies (SCS 54), Atlanta, GA 2006, 5-16.

202

4. Final remarks

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

of reality. The code of one language may overlap with that of another in multiple ways or perhaps not at all in some aspects. Just as light may be refracted as a continuum of colors on a spectrum, so translations may be characterized as a continuum on a spectrum from formal to functional equivalence. Formal equivalence at one end of the spectrum is isomorphic, item for item, word for word. The code of the receptor language is conformed as closely as possible to that of the source language. Functional equivalence at the other end of the spectrum is dynamic, idiomatic, idea for idea, so to speak. The code of the receptor language is followed as closely as possible to clarify the meaning of the source text and maximize effective communication and understanding for the readership. Thus the notion of fidelity in transmitting the word of God motivates both ends of the spectrum. Not just the books in the Prophets and the Writings, but even the books in the Greek Pentateuch, the very first translations, vary within this spectrum. Some are literal and represent formal equivalence in the extreme; others are idiomatic and represent different gradations of functional equivalence. Genesis and Exodus are dynamic translations, while Leviticus, Numbers, and Deuteronomy are quite literal. The translator of Job abbreviated many of the long, windy speeches for his Hellenistic readership so that the book is one-sixth shorter in Greek. Among the revisions of the original translations are “ultra-literal” recensions attributed to Theodotion and Aquila as well as the related texts in the so-called “καίγε group” or “καίγε tradition.” Characterization and identification of these texts is the focus of this chapter. Beginning with the καίγε group, there are five main periods in the history of research on these texts: (1) pre-Barthélemy, (2) Barthélemy, (3) post-Barthélemy studies, (4) Munnich and Pietersma, and (5) more recent studies. A major problem in research on these texts has been a failure to recognize that only exhaustive and systematic studies of translation technique can contribute effectively to advancement of our knowledge. Too much research has relied on assembling a small group of “key” equivalences or patterns in translation and extrapolating results from such inadequate databases. This article focuses on Theodotion Job, Septuagint Lamentations, Septuagint Ecclesiastes, Septuagint 2Reigns 11:1–3Reigns 2:11, Septuagint Canticles, as well as Aquila, because exhaustive studies have recently been made of the translation technique of these texts. Moreover, in corpora containing complete clauses such as these, there is the real possibility of analysis of translation technique, and not merely of assembling a set of lexical equivalences. Finally, neither the problem of correlating the Greek to its Hebrew parent text nor the question of a bilingual parent text plague study of these materials as is the case in what is attributed to Theodotion in the Book of Daniel—another large corpus of continuous text.

1. Pre-Barthélemy Research and the “Kaige Group” More than a hundred years ago, Henry St. John Thackeray distinguished different translators in the Greek version of Samuel-Kings on the basis of translation technique. Two sections he labeled βγ (2Sam 11:2–1Kings 2:11) and γδ (1Kings 22:1–2Kings 25:30)

1. Pre-Barthélemy Research and the Kaige Group

203

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

were characterized by formal equivalence in translation and Thackeray noted that the style had “much in common with that afterwards adopted by Theodotion.” 4

2. Barthélemy and the “Kaige Group” A quantum leap forward was made by the discovery of the Greek Minor Prophets Scroll from Nahal Hever in 1952 and the brilliant interpretation of its significance by Père Dominique Barthélemy in 1963. 5 This text is clearly a recension of the Old Greek of the Twelve Prophets. One of its most striking traits is the consistent rendering of ‫וגם‬/‫ גם‬by καίγε. Yet the book in which Barthélemy presented this to the scholarly world was far more than just a discussion of this manuscript and its text. With prodigious research he showed that the recensional characteristics of the Minor Prophets Scroll are also found elsewhere in the Greek Old Testament, both in recensions of earlier translations and in texts that seem to be the earliest or first Greek translation. The revisions and translations that shared the common set of traits identified by Barthélemy were dubbed the “καίγε group” on the basis of its characteristic treatment of ‫וגם‬/‫גם‬. Some of notable members included in this corpus were the sections from Kings labeled βγ and γδ by Thackeray, 6 the Greek Minor Prophets Scroll, the asterisked lines in the Greek Job, and the books of Song and Lamentations. Significantly, many members of the group were either directly or indirectly connected to Theodotion. Barthélemy not only posited an array of peculiar traits of translation technique which are characteristic of, and therefore the identifying markers of, the members of the καίγε group, he also argued throughout that each of these revealed exegetical methods showing direct influence from Palestinian rabbis of the first century CE. In addition to the basic group of patterns of translation technique employed to identify the members of the καίγε group, Barthélemy isolated a further set of patterns that revealed a kind of approach to revision and translation that culminated eventually in the work of Aquila. It was because of these characteristics that Barthélemy labeled his book Les Devanciers d’Aquila. The basic group of patterns employed by Barthélemy to identify the members of the group I have labeled “core patterns,” 7 while the additional set of traits that adumbrated the approach to revision and translation technique perfected by Aquila I have called “precursor patterns.” 8 These patterns are listed below; the chart of precursor patterns shows not only the equivalent preferred by the καίγε group, but also the equivalent in the more systematized approach of Aquila.

4. 5. 6. 7. 8.

H. St. J. Thackeray, “The Greek Translators of the Four Books of Kings” Journal of Theological Studies 8 (1907), 277. D. Barthélemy, Les Devanciers d’Aquila (Supplements to Vetus Testamentum 10), Leiden 1963. Barthélemy, Devanciers, 63-65. P. J. Gentry, The Asterisked Materials in the Greek Job (SCS 38), Atlanta, GA 1995, 390. Gentry, Asterisked Materials, 390.

204

2. Barthélemy and the Kaige Group

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

Core Patterns 1. ‫וגם‬/‫ = גם‬καίγε 2. ‫ = אישׁ‬ἀνήρ 3. ‫ = מעל‬ἐπάνωθεν/ἀπάνωθεν + genitive 4. ‫ = יצב – נצב‬στηλόω 5. ‫ = חצצרה‬σάλπιγξ / ‫ = שׁופר‬κερατίνη 6. Elimination of historical presents 7. ‫ = אין‬οὐκ ἔστιν 8. ‫ = אנכי‬ἐγώ εἰμι 9. ‫ = לקראת‬εἰς συνάντησιν Precursor Patterns 1. ‫ = גדוד‬μονόζωνος (α´ εὔζωνος) 2. ‫ = יהוה צבאות‬κύριος τῶν δυνάμεων (α´ κύριος τῶν στρατιῶν) 3. ‫ = אל‬ἰσχυρός (α´ ἰσχυρός employed exclusively for ‫ )אל‬9 4. ‫ = נגד‬Various Derivatives of ἔναντι (α´ κατέναντι) 5. ‫זאת‬-‫על‬/‫כן‬-‫על‬/‫ = לכן‬διὰ τοῦτο (α´ separate equivalents) 6. ‫ = לעלם‬εἰς τὸν αἰῶνα (α´ εἰς αἰῶνα) 7. ‫ = הוי‬οὐαί (α´ ὠί) 8. ‫ = אסף‬συνάγω (α´ συλλέγω) 9. ‫ = כמר‬χωμαρείμ (α´ τεμενίτης) 10. ‫ = אפלה‬σκοτία (α´ σκοτασμός) 11. ‫ = חוץ‬ἔξοδος (α´ ἐξώτατον) 12. ‫הדרה‬/‫ = הדר‬εὐπρέπεια (α´ διαπρέπεια) 13. ‫ = מהר‬ταχύνω (α´ ταχύνω 10)

An exhaustive analysis of the character, habits, and patterns of translation exhibited by the materials attributed to Theodotion in Job formed part of my doctoral research. 11 Since the approach of the translator to his task was quantitative, the methodology followed suit. The analysis distinguished between lexical and structural features and examined correspondence between Hebrew and Greek nouns, pronouns, verbs, and particles. The characterization of the materials also considered the level of phrases, clauses, word order, and the problem of transliterations. Once the Joban materials were characterized, they could be correlated and evaluated along with other similar studies. The results are briefly summarized as follows.

2.1 Theodotion Job and the Core Patterns Barthélemy posited nine basic or what we have called core patterns. Three do not occur in the Hebrew for which we have Greek renderings in Theodotion. Of the six patterns which remain, four are found in Theodotion Job while two are not. Before giving a brief evaluation of these results, the details will be provided for one or two patterns. 9. Barthélemy, Devanciers, 83. 10. Barthélemy (Devanciers, 184) notes that while OG translators normally use σπεύδω for ‫מהר‬, the καίγε recension of Kingdoms and the Greek Translator of Psalms use ταχύνω. Aquila further systematizes this using ‫ = מהר‬ταχύνω, ‫ = חושׁ‬σπεύδω and ‫ = בהל‬κατασπουδάζω. 11. Gentry, Asterisked Materials. 2. Barthélemy and the Kaige Group

205

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

2.1.1 ‫וגם‬/‫ = גם‬καίγε Four instances of ‫ גם‬and one instance of ‫ וגם‬occur in the portion rendered by Theodotion Job. He employs καίγε for ‫( גם‬15:10a; 15:10a; 19:18a; 30:2a), but has just καί for ‫וגם‬ (28:27b). 12 Thus Theodotion Job seems to treat ‫ גם‬and ‫ וגם‬differently. Barthélemy argues that earlier Aquila (i. e. Septuagint Ecclesiastes) does not distinguish between ‫גם‬ and ‫ וגם‬while later Aquila does, employing καὶ καίγε for the latter. 13 Perhaps not much should be made of the datum of only one instance of ‫ = וגם‬καί instead of an expected ‫ = וגם‬καίγε, but it is interesting to observe that Theodotion Job also distinguishes ‫ אף‬and ‫ ואף‬by employing καί in four instances (36:29a; 36:33b; 37:1a; 37:11a) and γάρ in one (15:4a) for ‫ אף‬and καὶ προσέτι in one (36:16a) for ‫ואף‬. 14 It should also be noted that some members of the καίγε group offer καὶ γάρ for ‫וגם‬/‫ גם‬rather than καίγε. 15 The pattern in Theodotion Job, then, shows both connection with and independence from the καίγε group. 2.1.2 ‫ = אין‬οὐκ ἔστιν According to Barthélemy, the rabbis believed ‫ אין‬was non-temporal; consequently cases where a future or past of εἰμί is used are corrected to a present by members of the καίγε group. 16 Barthélemy claimed that this typical translation of ‫ אין‬by οὐκ ἔστιν reappears in Job 32:5, thus linking Theodotion Job to the καίγε group. 17 Theodotion Job rendered ‫ אין‬in ten instances, employing οὐκ ἔστιν in five of them (20:21a, 26:6b, 28:14b, 32:5a, 35:15a), οὐκ ἦν once (32:12b), μή once (33:33a), alpha privative once (21:33c), and finally, where ‫ אין‬had pronominal suffixes, οὐκέτι εἰμί (7:8b) and οὐκ ἔστιν (27:19b) once each. The instance mentioned by Barthélemy (32:5a) is the only case of οὐκ ἔστιν where a present tense would seem to be contextually out of place, but the instances rendered by οὐκ ἦν, μή, and alpha privative show contextually sensitive renderings and especially the case of οὐκ ἦν would be a counter-example to that of 32:5a. In point of fact, however, the case of οὐκ ἔστιν in 32:5a does not support

12. Heater (H. Heater, Jr., A Septuagint Translation Technique in the Book of Job, Ph.D. dissertation, The Catholic University of America, 1976, 22-23) noted only three of the five instances (15:10a, 15:10a, 30:2a) as he did not include in his corpus materials relegated in the Göttingen Edition to the Second Apparatus which were available to him only in Field. 13. Barthélemy, Devanciers, 32-33. 14. There are just four other instances of προσέτι in any of the Greek versions according to Hatch-Redpath: καὶ προσέτι for ‫ ואף כי־‬in 2Reigns(Sam) 16:11 (= καίγε group), προσέτι δέ for ‫ גם‬in Ps 48/49:3 (= σ´), and 2Mac 12:14 and 4Mac 14:1 which are not relevant. From the point of view of the Hebrew, elsewhere in the βγ and γδ portions of Samuel-Kings three occurrences of ‫ אף‬and ‫ ואף‬are rendered: ‫ = אף־אחריו‬κατόπισθεν (οπισω b o c2 e2; 2Reigns(Sam) 20:14), ‫ =אף־הוא‬αφφω (4Reigns[2Kings] 2:14), and ‫ = ואף כי־‬καὶ ὅτι (sic Bij; καθοτι b c2 e2; om και A N rell Eth Syh; 4Reign[(2Kings] 5:13). This evidence is scant, but does not negate the conclusion that ‫ אף‬and ‫ ואף‬are handled differently in the βγ and γδ portions respectively, as in Theodotion. 15. Barthélemy, Devanciers, 41-43, 47. 16. Barthélemy, Devanciers, 65-68. See, however, O. Munnich, “Contribution à l’étude de la première révision de la Septante” in: W. Haase (ed.), Aufstieg und Niedergang der Römischen Welt, vol. II, 20:1, New York, NY 1987, 212. 17. Barthélemy, Devanciers, 67.

206

2. Barthélemy and the Kaige Group

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

Barthélemy’s claim. Within a nominal ὅτι-clause—as οὐκ ἔστιν is found in 32:5a— whether in primary or secondary sequence, the present is normal usage. 18 Therefore no example in Theodotion Job conforms to the pattern in the καίγε group. 19

2.1.3 Summary This detailed discussion illustrates the character of the results. Four patterns out of six where analysis is possible are found in Theodotion Job. But evidence is quite scant for two of them and there is a variation on the pattern in one. Abundant and clear evidence exists for only one or two patterns. Certainly Barthélemy did have some reason to connect Theodotion Job with members of his καίγε group, but the nature of this relationship is problematic.

2.2 Theodotion Job and the Precursor Patterns Twelve precursor patterns are presented by Barthélemy. 20 Four do not occur in the Hebrew for which we have Greek renderings in Theod. Of the eight patterns which remain, three are found in Theod, four are not, and the evidence for the eighth is ambiguous. Space permits discussion of two of them.

2.2.1 ‫ = אל‬ἰσχυρός (α’ ἰσχυρός employed exclusively for ‫)אל‬ Theodotion Job employs ὁ θεός for ‫ אלוה‬and ‫ אלהים‬and ὁ ἰσχυρός for ‫אל‬. 21 But Theod also uses ἰσχυρός once in 37:18b as an adjective to render ‫ חזק‬and once in 15:20b as a divine epithet for ‫עריץ‬. Similarly, according to the evidence from RT, ἰσχυρός is employed by α’ mainly for ‫אל‬, but also sporadically for ‫ מבצר‬,‫ חזק‬,‫ גבור‬,‫ איל‬,‫ אדיר‬,‫אביר‬, and ‫עריץ‬, although only instances of ἰσχυρός for ‫איל‬, 22 ‫ חזק‬,‫גבור‬, and ‫ עריץ‬are certainly attested as uniquely attributed to α´. Barthélemy argues that the use of ἰσχυρός for ‫ חזק‬,‫גבור‬, and ‫ עריץ‬by α´ instead of exclusively for ‫ אל‬represents the earlier rather than

18. F. Blass / A. Debrunner, A Greek Grammar of the New Testament and Other Early Christian Literature (ninth and tenth editions, edited and translated by R. W. Funk), Chicago, IL 1961, § 324. A perfect parallel to 32:5a is John 6:24: ὅτε οὖν εἶδεν ὁ ὄχλος ὅτι ᾽Ιησοῦς οὐκ ἔστιν ἐκεῖ οὐδὲ οἱ μαθηταὶ αὐτοῦ, ἐνέβησαν αὐτοὶ εἰς τὰ πλοιάρια καὶ ἦλθον εἰς Καφαρναοὺμ ζητοῦντες τὸν ᾽Ιησοῦν. 19. Heater (Septuagint Translation Technique, 26) does not distinguish the differing contextual and grammatical settings and concludes that “The KR [Kaige Recension] trend is obvious in the translation of this particle, but the fact that old Greek does not hesitate to use a present should warn against too much weight being placed on this characteristic”. The evidence, however, fails to demonstrate any such trend. 20. In fact, Barthélemy has thirteen, since another pattern is noted in a later section of his book. When commenting on the Greek Minor Prophets Scroll, Barthélemy notes that like the καίγε recension of Kingdoms and the Greek Translator of Psalms, the Greek Minor Prophets Scroll prefers ταχύς and ταχέως for ‫ְמֵה ָרה‬. Aquila further systematizes by using ‫ = מהר‬ταχύνειν, ‫ = חושׁ‬σπεύδειν, ‫ = בהל‬κατασπουδάζειν (Barthélemy, Devanciers, 184). ‫ְמֵה ָרה‬, however, does not occur in Job. 21. For complete evidence and analysis, see Gentry, Asterisked Materials, 89, 118-120. 22. According to RT, s. v., only Isa 1:29 and 61:3 (α´ σ´ θ´), and possibly read as ‫אל‬, but note Jb 41:17a. 2. Barthélemy and the Kaige Group

207

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

the later Aquila. 23 In hexaplaric materials from the Second Apparatus of the Göttingen Edition and from the Hagedorns for Job, ἰσχυρός is attributed to α´ only once (41:17a), but it renders ‫אילים‬. 24 At 37:18b, however, where θ´ has ἰσχυραί for ‫חזק;ם‬, α´ has ἐνισχύουσαι. The claim that Aquila, unlike Theodotion Job, employed ἰσχυρός for ‫ אל‬exclusively does not fit all the data and the proposal of an earlier and later Aquila is not entirely satisfactory. What is important to note is that Theodotion Job normally articulates ἰσχυρός while α´ does not. So the characteristic equivalent for the καίγε group is also in Theodotion Job and Aquila refines it by employing ἰσχυρός as a proper noun, i. e. usually unarticulated.

2.2.2 ‫הדרה‬/‫ » הדר‬εὐπρέπεια (α´ διαπρέπεια) According to Barthélemy, the καίγε group frequently employs εὐπρέπεια for /‫הדר‬ ‫הדרה‬, but Aquila used εὐπρέπεια for ‫ נַֹעם‬because he reserved εὐπρεπής for ‫ ָנִעים‬. 25 Thus he created διαπρέπεια as the equivalent for ‫הדרה‬/‫הדר‬. Theodotion Job renders ‫ הדר‬once (40:10b) and uses ὕψος (δόξα C p). It should be noted that in this text Theodotion Job has εὐπρέπειαν καὶ ὕψος for ‫ הוד והדר‬in MT. Possibly ὕψος is employed for stylistic variation since εὐπρέπεια is already used for ‫הוד‬. 26 Elsewhere Theodotion Job uses εὐπρεπής once for ‫( נוהו‬18:15b) and εὐπρέπεια once for ‫( נות‬8:6b)—in both instances as if from ‫—נאוה‬and once for ‫( נעים‬36:11c). Aquila not only uses εὐπρέπεια for ‫נַֹעם‬, but also for ‫נוה‬. 27 Interestingly, however, at 18:15b α´ has ὡραιότης. 28 Thus Theodotion Job does not have the characteristic of the καίγε group for ‫הדרה‬/‫הדר‬, but neither does he show the specialization of Aquila. 2.2.3 Summary Out of eight patterns, then, where analysis is possible only three are clearly found in Theodotion Job and evidence is slim for one of them. In one case, the evidence is insufficient to demonstrate either connection to or independence from the καίγε group. Putting core and precursor Patterns together, where evaluation is possible only 23. Since most occurrences are in Jer and Ez, Barthélemy claims these were early revisions of Aquila (Barthélemy, Devanciers, 83). 24. See J. Ziegler, Iob (Septuaginta Vetus Testamentum Graecum 12.4), Göttingen 1982 and U. Hagedorn / D. Hagedorn, Nachlese zu den Fragmenten der jüngeren griechischen Übersetzer des Buches Hiob (Nachrichten der Akademie der Wissenschaften in Göttingen 1, Philologisch-Historische Klasse, 10), Göttingen 1991, 23. Available to Barthélemy in Field. Also noted in Heater (Septuagint Translation Technique, 41). The plural shows Aquila did not read ‫ איל‬as ‫אל‬. 25. Barthélemy, Devanciers, 87-88. This characteristic is not considered by Heater. 26. Although δόξα is employed for ‫ הוד‬in 37:22b, the equivalence belongs to OG rather than to θ´ (see Gentry, Asterisked Materials, 261-262). 27. RT, 101. 28. Ziegler’s Second Apparatus for 18:15b has θ´ (leg α´) λικμηθήσεται ἐπὶ ὡραιότητι αὐτοῦ (+ το 137) θεῖον C’ (Olymp). As 18:15b is sub ※, Ziegler is surely right in assigning this reading to Aquila. The evidence from Hagedorn now proves it beyond question. 18:15b α´ (θ´ Γ Ν; anon 250 612) λικμηθήσεται (-θήσονται 3006) ἐπὶ (ἐν τῇ 3005) ὡραιότητι αὐτοῦ θεῖον (τὸ θεῖον 137) Γ’ (= 250 3005 Γ–[138] N (Hagedorn / Hagedorn, Nachlese, 23).

208

2. Barthélemy and the Kaige Group

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

half are clearly found in Theodotion Job. Either we need to redefine the relationship of Theodotion to the καίγε group or redefine what is meant by the group itself, or both! If by the καίγε group we refer to a diverse group of recensions or first translations which share a common approach or attitude to translation—an approach that is developing or has developed into a tradition, then Theodotion belongs to it. But the diversity that exists within the καίγε group, as well as the precise line of demarcation between καίγε and the Septuagint, has yet to be taken seriously by modern scholarship.

3. Post-Barthélemy Studies Numerous articles, dissertations, and other studies were spawned by the original and stimulating work of Barthélemy. For a period of some twenty years, from 1963 to 1983, the focus of research was to discover other key equivalences or patterns characteristic of the “καίγε group,” some in texts already identified as belonging to the group and some in texts newly identified as belonging to the group. The culmination of this period was perhaps the dissertation of Leonard Greenspoon, Textual Studies in the Book of Joshua, published in 1983. On the basis of a compilation largely made by Greenspoon, we have a list of some ninety-seven patterns claimed by different scholars during the period 1963-1983 to be characteristics of the “καίγε group.” One may well question the relative value of many of the patterns proposed postBarthélemy as characteristics or markers of the καίγε group. Nonetheless, where comparison is possible with Theodotion Job, analysis revealed that twelve of these patterns are found in Theodotion Job, eight are not, and in three the evidence is inconclusive. Therefore the data are similar to cases already considered. 29 There is obviously some connection between the so-called καίγε group and the asterisked materials, but the correlation between the two is only partial. Certainly Theodotion Job is not part of one monolithic recension as Homer Heater concluded. Consideration of the history of καίγε studies together with the present results suggests several things: (1) post-Barthélemy studies, and even the research of Barthélemy himself, focused so intensively on agreements in an effort to establish existence of the καίγε group that the differences between group members were not considered or examined sufficiently, (2) the καίγε texts were wrongly conceived of as one recension by pushing evidence in a direction Barthélemy himself was careful to avoid, (3) analysis must go beyond consideration of a few certain equivalences selected as key traits of a reviser’s or translator’s work. A broader analysis is necessary which compares Theodotion Job with the Greek Minor Prophets Scroll and Aquila on the basis of an exhaustive investigation and thorough understanding of the translation technique of all three. The analysis must be contrastive as well as comparative. All possible interconnections and relations must be explored. The issues and politics of the diverse groups of Second Temple Judaism must be the larger framework for interpreting the results.

29. For details, see Gentry, Asterisked Materials, 402-417. 3. Post-Barthélemy Studies

209

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

4. The Contributions of Munnich and Pietersma Olivier Munnich’s analysis of the vocabulary of the Greek Psalter has given new directions and orientation to thinking about the καίγε group. 30 First, he has turned attention away from attempting a close link between exegetical and hermeneutical methods of the Palestinian rabbis and characteristic patterns of translation technique exhibited by the καίγε group. Secondly he has turned attention toward considering the key patterns as generalizations and stereotypes developed from earlier sources, instead of focusing solely on these texts as precursors to Aquila, the one who perfected these traits. The first direction is negative. Munnich demonstrated that a number of patterns were not necessarily motivated by peculiarities of rabbinic exegesis as Barthélemy had claimed. In fact, many patterns discovered post-Barthélemy had no obvious link whatever. The second direction is positive. Munnich sought to establish the sources which influenced the καίγε group rather than focus on Aquila as the one influenced by the καίγε group. Munnich began his research by analyzing the vocabulary of the Greek Psalter. His analysis isolated lexemes which he demonstrated were coined and employed for the first time by the translator of the Hebrew Psalms. Next he traced the use of these lexemes by revisers and translators of biblical books that were clearly later than the date of origin of the Greek Psalter. A fair number of these terms had been taken up by members of the καίγε group and turned into stereotypes. Around the same time, Albert Pietersma, working independently of Munnich, made much the same point. 31 Munnich’s contribution suggested that a pilot study ought to be undertaken in which equivalences constituting patterns in Theodotion Job were compared and contrasted not only with the Minor Prophets Scroll as possibly coming from the same or a similar source, or Aquila as the perfecter of these traits, but also with the Greek Psalter as the possible origin or source of them. Unfortunately, such a comparison would have to be limited largely to lexical equivalences since exhaustive studies of patterns of translation technique are unavailable for the Greek Psalter and Aquila and only partially available for the Scroll. Nonetheless, such a study would point the way to developing a typology in which we could place and position Theodotion Job. As a part of my doctoral research, I sought to analyze the possible relationships between Theodotion Job, the Greek Psalter, the Minor Prophets Scroll, and Aquila in a preliminary way by carefully comparing and contrasting all lexical equivalences employed by Theodotion Job for both nouns and verbs occurring three or more times with equivalences employed by the Psalter, the Scroll, and Aquila for the same words. Only a brief review of the results is possible here.

30. O. Munnich, Étude Lexicographique du Psautier des Septante, Ph.D. dissertation, université de Paris-Sorbonne, 1982; O. Munnich, “Indices d’une Septante originelle dans le Psautier Grec” Biblica 63 (1982), 406-416; O. Munnich, “La Septante des Psaumes et le Groupe Kaige” Vetus Testamentum 33 (1983), 75-89; Munnich, “Contribution.” 31. A. Pietersma, “Septuagint Research: A Plea for a Return to Basic Issues” Vetus Testamentum 35 (1985), 296-311.

210

4. The Contributions of Munnich and Pietersma

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

4.1 Theodotion Job and the Minor Prophets Scroll Comparison of Theodotion Job and the Minor Prophets Scroll is not entirely straightforward. The fragmentary state of the Scroll, the relative shortage of common vocabulary, and the problem of comparing a recension with a translation (largely unaffected by a previous translation) preclude bold statements about the connection between Theodotion Job and the Scroll. There is no unique agreement between the two that could point to an incontrovertible inter-relationship, whereas some six or so disagreements (nouns and verbs) point to a considerable degree of independence. If one also looks at equivalences in Theodotion Job where the frequency is less than three, additional disagreements exist between patterns claimed by Barthélemy to be certain characteristics of the καίγε group based upon the Scroll and what we find in Theodotion Job. 32 The most that can be said in view of the nature of the evidence is that both Theodotion Job and the Minor Prophets Scroll had similar attitudes and approaches towards translation and appreciated some of the same type of techniques.

4.2 Theodotion Job and the Psalter Approximately eight out of twenty-nine agreements between Theodotion Job and the Psalter constitute possible evidence for borrowing from the Psalter, and only two or three further instances arise when instances of disagreement were fully investigated. 33 On the other hand, some forty-three disagreements out of fifty-five show Theodotion Job clearly independent from the Psalter. Certainly Theodotion Job was influenced by the Psalter, but the Psalter is hardly the lode from which the Theodotion Job mined his peculiar traits. Thus the proposal of Munnich that the Greek Psalter formed a sort of glossary from which members of the καίγε group galvanized certain equivalences into stereotypes requires further evaluation and revision. Indeed, Pietersma, in his “Plea For A Return To Basic Issues” was more cautious in describing influence from the Greek Psalter upon the καίγε group. 34 Two features of Munnich’s research impeded his arriving at better results. First, he failed to employ contrastive as well as comparative analysis. For the Book of Job, he deals only with the rendering of ‫ צעד‬by διάβημα in 14:16a and 31:4b and claims on the basis of this a dependence of Theodotion Job upon the Psalter. 35 Not only is the basis

32. Gentry, Asterisked Materials, 439-440. 33. A disagreement at the formal level may in fact be evidence for borrowing from the Greek Psalter by Theodotion Job, since, for example, an equivalence for a noun in Theodotion Job may be inspired by an equivalence for a verb in the Psalter. 34. Pietersma, “Septuagint Research,” 309. 35. Munnich (Étude Lexicographique, 476) asserts, “ce constat fournit un premier indice de la dépendence lexicale de Théodotion par rapport au Psautier des Septante”. Free renderings make MT ‫ צעד‬unrecognizable in OG at 18:7a, 31:37a, 34:21b. OG employs τὰ ἐπιτηδεύματα in 14:16a for which the C Group attests τὰ διαβήματά μου attributed to λ´. 31:4b is clearly from Theodotion. Munnich says that in 14:6a, “Théodotion remplace ἐπιτήδευμα par διάβημα” (Ibid.). He does not make his reader aware that, according to the evidence available to him (which is now confirmed by the critical edition), the attribution is somewhat vague. 4. The Contributions of Munnich and Pietersma

211

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

for such an assertion slim, there are, moreover, a number of neologisms in the Greek Psalter not borrowed by Theodotion Job. 36 Secondly, serious methodological problems arise when one attempts to prove influence or relationship on the basis of certain key peculiar traits. Munnich’s analysis of βᾶρις is penetrating and sophisticated, but one cannot hang the identity of the καίγε group on such a slender thread. 37 My pilot study points the way to further research on the relation of the Psalter to the καίγε group by demonstrating the value of assessing a large block of data and basing conclusions on comparative and contrastive analysis of the whole, not just on certain outstanding or peculiar traits. On the positive side, Munnich’s work was helpful in focusing attention on possible sources of the καίγε group. But the Greek Psalter is not necessarily the only or even the main source. It is indeed plausible that the influence of portions of the Septuagint employed liturgically was greater upon later translators than parts not used this way. But it is equally clear that a number of equivalences were drawn by Theodotion Job from the Greek Pentateuch, Old Greek Job, and Septuagint Isaiah. A more comprehensive explanation for the relations of Theod to other recensions and translations is necessary—one much broader than that proposed by Munnich, who in turn had broadened Barthélemy’s thesis. Such an explanation would correlate a history of the development of approaches and attitudes to the translation of the Jewish Scriptures from the third century BCE to the second century CE as seen in the constant interchange between Diaspora and Palestinian Judaism with the evidence from the texts themselves.

4.3 Theodotion Job and Aquila The few agreements and the large number of disagreements as well as the nature of the evidence support the thesis of Barthélemy. The approach of Aquila clearly represents a further refinement of the translation technique employed by Theodotion Job. 38 It is a more rigid and systematic attempt to represent the Hebrew quantitatively. This has involved the development of a comprehensive and more specialized system of equivalents so that each lexeme in the parent text has a different equivalent in Greek. His system is comprehensive in that all parts of speech from a particular Hebrew root are frequently rendered by equivalents all derived from a single Greek stem, even when this runs counter to standard usage in Hellenistic Greek (e. g. ‫ → חזון‬ὁραματισμός and ‫ → חזה‬ὁραματίζομαι). In this way equivalents for nouns and verbs are correlated. His system also involves an etymologizing approach in the search for exact renderings. Aquila is more consistent than Theodotion Job in sticking to his fixed equivalents and thus less sensitive to contextual factors. Such an approach means that there is 36. Some six or so are discussed, see Gentry, Asterisked Materials, 467-469. One such example is the rendering of ‫( גאל‬Qal active participle) by λυτρωτής in the Psalter; ἀγχιστεύς is used in Theodotion. This former lexeme was coined by the Psalter, as Munnich (Étude Lexicographique, 106, 169-170) demonstrates. The latter was well established in the Greek Pentateuch (twelve occurrences). 37. Munnich, “Contribution,” 200-205. Cf. Pietersma, “Septuagint Research,” 309-310. 38. For a serious analysis of Aquila, see K. Hyvärinen, Die Übersetzung von Aquila (Coniectanea Biblica Old Testament Series 10), Uppsala 1977.

212

4. The Contributions of Munnich and Pietersma

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

much less of a tendency to use more than one Greek equivalent for any particular Hebrew lexeme than is the case in Theodotion Job. So there is a development in attitude to translation from the Septuagint to Theodotion Job to Aquila, which is at least typological if not chronological.

5. More Recent Studies Research continues unabated on the καίγε group. Several studies deserve to be mentioned and summarized.

5.1 Theodotion Daniel Tim McLay concluded his doctoral dissertation Translation Technique and Textual Studies in the Old Greek and Theodotion Versions of Daniel in 1994, just after I finished my work. 39 His research, done independently of mine on Job, emphasizes the importance of studying translation technique and offers results remarkably similar. Nonetheless, his study is based only on five segments of ten verses and does not represent an exhaustive description of translation technique.

5.2 Septuagint Psalter In 1995, at the IXTH Congress of the IOSCS in Cambridge, shorter studies by Staffan Olofsson, on “The Καίγε-recension in the Septuagint Book of Psalms,” and Timothy Janz, on “The Second Book of Ezra and the ‘Καίγε Recension’,” were presented. Oloffson seeks a broader framework for examining questions pertinent to the καίγε group by adopting the general picture painted by S. P. Brock of the translator as expositor or interpres operating in the context of the tensions between Diaspora and Palestinian Judaism over the translation of Scripture. He never attempts, however, to clearly describe what he means by the καίγε group and waffles as to whether the present Greek text of Psalms is Old Greek or recension. In interpreting the data, he appears too eager to claim evidence to identification as a καίγε pattern. The differences must be more carefully explored. The καίγε tradition cries out for a clear typology of development in relation both to original texts and recensions of the Septuagint. In a study for the Pietersma Festschrift, I re-evaluated Olofsson’s contribution, reexamining his data as follows. 40 First, Olofsson never clearly defines what he means by the καίγε group. Καίγε is not a monolithic recension, nor even a group of revisers. Rather καίγε is a tradition in which a certain approach to the translation of the Hebrew scriptures into Greek was considered to be the faithful way of performing this task and in which certain equivalents and traits of translation, perhaps original to earlier members of the tradition or

39. Later published as T. McLay, The OG and Th Versions of Daniel (SCS 43), Atlanta, GA 1996. 40. P. J. Gentry, “The Greek Psalter and the καίγε Tradition” in: R. J. V. Hiebert / C. E. Cox / P. J. Gentry (eds.), The Old Greek Psalter, Sheffield 2001, 74-97. 5. More Recent Studies

213

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

translators prior to it, became hallmarks of this tradition. Both original translations and recensions of original translations belong to this tradition. As we consider the diverse group of translations conveniently labeled “the Septuagint,” we should probably envision a continuum, something like the light spectrum, in which some approaches to translation are more dynamic and some more formal. The books of the Greek Pentateuch reveal this continuum as well as renderings of books translated later. Again both original translations and revisions reveal this continuum. Those who appreciated the approach of formal equivalence, whether on opposite sides of the Diaspora-Palestinian debate or on the same side, noticed certain traits which exemplified this approach and began to make more and more systematic use of them. The καίγε tradition, then, is an integral part of this continuum of the whole. If Olofsson had been clearer on just what he meant by the καίγε group and, also, less insistent that strict consistency is unnecessary to identify members as part of this tradition, his diligent efforts would have resulted in relating the Greek Psalter to the larger framework and to the καίγε tradition. Second, raising the possibility, as Olofsson does, that the present text of the Greek Psalter is a revision belonging to the καίγε tradition only confuses the matter. He seems to steer (wisely) away from this at the end of his study. Even though we do not possess a fully critical text of the Greek Psalter, positing this text as a revision and not as OG beclouds the entire question of the relation of the Greek Psalter to καίγε. Certainly the translator of the Greek Psalter belonged to the camp of formal equivalence. Connections, however, with the καίγε tradition are not strong. On the basis of my own preliminary study and the evidence amassed by Olofsson, it seems plausible that the Greek Psalter may be at an early stage of this tradition. However, before we can clarify how the Greek Psalter fits into the continuum and relates to the καίγε tradition much work must be done. We must analyze the translation technique of the Greek Psalter as a whole, comparing this global picture with an evaluation of translations, both original translations and obvious revisions such as the Greek Minor Prophets, done before and after the Psalter. Second, we must look at neologisms created by the Greek Psalms translator and used by later translators in contrastive as well as comparative analysis. 41

5.3 Septuagint Lamentations Septuagint Lamentations was identified by Barthélemy as belonging to the “καίγε group” and Isabelle Assan-Dhote drew the conclusion that the text is, in fact, Theodotion. 42 These conclusions were re-evaluated by Kevin J. Youngblood’s 1994 doctoral dissertation based upon an exhaustive and thorough study of the translation technique. 43 Youngblood then compared and contrasted the materials in Lamentations

41. See Gentry, Asterisked Materials, 475-477. 42. I. Assan-Dhote, La version grecque des Lamentations de Jérémie, Ph.D. dissertation, université de Paris-Sorbonne, 1996. 43. K. J. Youngblood, Translation Technique in the Greek Lamentations, Ph.D. dissertation, The Southern Baptist Theological Seminary, 2004.

214

5. More Recent Studies

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

with the Greek Psalter, Greek Minor Prophets Scroll, Theodotion Job, and Aquila. Youngblood’s conclusion is as follows: A close, thorough study of the relationship of OGL [LXX Lamentations] to the MT reveals that it is the product of a translator with a formal equivalency approach to translation and who, as far as possible, brought his translation into quantitative alignment with the Hebrew parent text. The word order, and surface structure of the source language dominated the translator’s task but not without certain concessions to Greek style and good sense. The translator attempted to render noun for noun, verb for verb, and particle for particle even if the result was less than good Greek. At no point, however, did he produce nonsense. The meaning of his text is almost always clear and those places where it is not are usually due to a difficult Hebrew text rather than to an incompetent translator. Thus, unlike Aquila, the translator never allowed his reverence for he form of the parent text to eclipse the goal of communicating the text’s meaning. 44

Although Lamentations definitely belongs to the καίγε tradition, Youngblood also concludes that the identification of Lamentations as Theodotion is doubtful and emphasizes the diversity and lack of homogeneity within members of the καίγε tradition.

5.4 Septuagint Ecclesiastes The Greek Translation of Ecclesiastes in the Septuagint is characterized by extreme formal equivalence, so much so that scholars have suggested that it is the work of Aquila himself, a reviser of the Septuagint who flourished circa 120 CE. While the identity of the translator is still undetermined according to the most recent research, the character of the translation reveals that in fact some patterns are identical to those considered classical Aquila and others are clearly not Aquila. 45 The approach of the translator to his task is one where faithfulness is defined and measured by the degree of quantitative alignment between Greek translation and Hebrew original. As a result, the text is difficult to read in places and almost incomprehensible at times from the point of view of the native speaker of Hellenistic Greek who has no knowledge of or recourse to the source text in Hebrew. The attempt of the translator to align his translation with the source language and text affects both lexical and syntactic choices in Greek. Choosing the same lexeme in Greek for every occurrence of a particular lexeme in the Hebrew regardless of the contextual meaning results in a high degree of stereotyping. Few calques or isolate renderings, however, are to be found. Like Aquila, then, consistency and uniformity in Hebrew-Greek equivalents is extremely rigid. Unlike Aquila, however, we do not see a set of Hebrew-Greek equivalents that is so highly refined and specialized that distinct equivalents are em44. Youngblood, Translation Technique, 356-357. 45. For an evaluation of the debate as to whether or not Septuagint Ecclesiastes represents the work of Aquila—either as an early edition or an edition in final form—from A. Dillmann in 1992 to the present, see P. J. Gentry, “The Relationship of Aquila and Theodotion to the Old Greek of Ecclesiastes in the Marginal Notes of the Syro-Hexapla” Aramaic Studies 2.1 (2004), 63-84. The most recent research suggests the verdict is non liquet. See J. Jarick, “Aquila’s Koheleth” Textus 15 (1990), 131-139 and bibliography in his footnotes for earlier contributions to the debate. 5. More Recent Studies

215

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

ployed for each Hebrew lexeme, nor does the translator attempt to represent the root system of Hebrew by forming a set of equivalents from a single Greek stem to be used for nouns and verbs derived from a single root. Even more than in lexical equivalents, it is in the morpho-syntactic structures that the attempt to align the Greek and the Hebrew is most apparent and also where the results are so awkward or even absurd. One obvious and outstanding example is the rendering of ‫ את‬by σύν (1:14; 2:12; 2:17; 2:18; 3:10; 3:11bis; 3:17bis; 4:1; 4:2; 4:3; 4:4bis; 4:15; 5:3; 5:6; 7:14; 7:15; 7:18; 7:26; 7:29; 8:8; 8:9; 8:15; 8:17bis; 9:1bis; 9:15; 10:19; 10:20; 11:5; 11:7; 12:9; 12:14). 46 That σύν is found with the dative in 9:11 is normal (σύν πᾶσιν αὐτοῖς); that it is found, however, with the genitive (9:15) or accusative (rel.) to mark the direct object is contrary to Greek grammar. This characteristic of the translator may be further illustrated. In 2:1 ‫ וראה בטוב‬is rendered by καὶ ἰδὲ ἐν ἀγαθῷ. The verb ‫ ראה‬followed by ‫ ב‬here is idiomatic, conveying the sense ‘to enjoy’ (so NRSV). The Greek Translator of Ecclesiastes rigidly renders ‫ ראה‬by εἶδον and ‫ ב‬by ἐν. Contextually, “to see in good” makes no sense in Greek. Again, a construction article + πᾶς + noun normally contrasts the whole with the part. 47 Yet in 3:17 ‫ כי־עת לכל־חפץ‬is rendered ὅτι καιρὸς τῷ παντὶ πράγματι (cf. 3:1; 4:8; 4:16). The article τῷ is employed to represent the ‫ ;ל‬this effect creates difficulties for the Greek reader and skews the meaning of Hebrew while attempting to represent it formally in all respects. Other cases are perhaps readable, but awkwardly so. Note, for example, that ‫ עד אשׁר לא־יבאו ימי הרעה‬in 12:1 is rendered by ἕως ὅτου μὴ ἔθλωσιν ἡμέραι τῆς κακίας. With the negative μὴ, the result is awkward, but not necessarily impossible Greek. Elsewhere the rigid approach of the translator is carried off more cleverly. Hebrew ‫( גם‬1:17; 2:1; 2:7; 2:8; 2:14; 2:15; 2:15; 2:19; 2:21; 2:23bis; 2:24; 2:26; 3:11; 4:4; 4:8ter; 4:11; 4:14; 4:16bis; 5:9; 5:16; 5:18; 6:5; 6:9; 7:14; 7:21; 7:22bis; 8:10; 8:12; 8:14; 8:16; 9:1bis; 9:6ter; 9:12; 9:13; 10:20; 12:5) or ‫( וגם‬1:11; 3:13; 5:15; 6:3; 6:7; 7:6; 7:18; 8:17; 9:3; 9:11ter; 10:3; 11:2) is always rendered by καίγε except in a double translation in 7:22b, where the equivalent is simply καί. Nonetheless, the equivalence ‫וגם‬/‫ = גם‬καίγε, while illustrating the spirit of the Greek translator of Ecclesiastes, is unlikely to have originated with him. Here the translator extends and perpetuates a tradition. Clearly, then, the Greek translator of Ecclesiastes belongs to the so-called καίγε tradition of translators. 48

5.5 Septuagint 2Reigns 11:1–3Reigns 2:11 In 2004 Paul McLean completed a doctoral dissertation at the University of Toronto in which he conducted an exhaustive analysis of translation technique in Thackeray’s

46. 2:12 and 9:11 are included against Alfred Rahlfs, Septuaginta, Id est Vetus Testamentum graece iuxta LXX interpretes, Stuttgart, 1935, ad loc.; see J. Ziegler, “Die Wiedergabe der nota accusativi ’et, ’aet- mit σύν” ZAW 100 Sup (1988), 222-233. 47. Blass / Debrunner, Greek Grammar, § 275.7. 48. Taken from P. J. Gentry, “Ecclesiast” in: A. Pietersma / B. G. Wright (eds.), The New English Translation of the Septuagint, Oxford 2007, 648-649. For an exhaustive treatment of the translation technique in Septuagint Ecclesiastes and its place in the text history see Y. Y. Yi, Translation Technique in the Greek Ecclesiastes, Ph.D. dissertation, The Southern Baptist Theological Seminary, 2005.

216

5. More Recent Studies

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

βγ καίγε section of Samuel-Kings. 49 McLean’s close examination of the linguistic make-up of this καίγε section establishes this translator’s standard approach to the source text: The Hebrew was rendered in a highly word-based isomorphic manner, though in some cases one encounters interpretive moves at the word, phrase and clause levels. 50

McLean also investigates semantic variations between MT and καίγε. After eliminating cases where the parent text is different he analyzes places where variation demonstrates exegetical and interpretive interests of the translator. These illustrate translation techniques such as ancient rabbinic exegesis, semantic leveling, semantic differentiation, exegetical pluses, semantic specification, other kinds of marked exegesis, intertextual exegesis, and theological exegesis. Even the translator’s handling of Hebrew names shows some unexpected variations. McLean’s conclusions corroborate analysis of texts in the καίγε group given here.

5.6 Septuagint Canticles The most recent study is an exhaustive analysis of Septuagint Canticles, a doctoral dissertation by Dries De Crom at Leuven destined to be an important tool for future study. His analysis of Septuagint Canticles is more sophisticated in the method of translation technique, but the conclusions support the thesis of this article. They should be heard in his own words: The textual analysis […] confirms the existing scholarly consensus concerning the translation character of LXX Canticles. The translation is indeed very formal with regard to the quantitative representation of its ST [Source Text]. The translator obviously took care to preserve word order, segmentation and, to a great extent, lexical consistency. And yet the translation displays a considerable degree of freedom and ingenuity on lower-level text segments (words and word groups). […] Thus, genuine semantic transfer in LXX Canticles is rare and far outweighed by context-bound lexical choices. There are of course words which are out of style or out of context, but this is more a matter of (lacking) stylistic homogeneity and textual cohesion than of semantics. On the level of syntax, the translation is considerably more “literal”, sometimes pushing the limits of TL [Target Language] norms to accommodate Hebrew constructions. However, if one distinguishes between positive and negative transfer, it is found that the former is more common by far. Truly ungrammatical syntax is limited to specific cases of stereotyping (e. g. the prepositions ἐν ~ ‫)ב־‬. On the level of text and discourse, LXX Canticles lacks most features of Greek (literary) discourse. Yet there are traces of a textual awareness on the part of the translator, i. e. a preconceived notion of the general content of the book of Canticles which influenced some of the choices he made while translating. 51

49. P. D. McLean, The Greek Kaige Version of 2 Reigns 11:1–3 Reigns 2:11: A Study of Its Constituent Translation Technique and Semantic Variations from its Hebrew Vorlage Using the Interlinear Paradigm for A New English Translation of the Septuagint (NETS), Ph.D. dissertation, University of Toronto, 2004. 50. McLean, Greek Kaige Version, abstract. 51. D. De Crom, The LXX Text of Canticles: A Descriptive Study in Hebrew-Greek Translation, Ph. D. dissertation, Katholieke Universiteit Leuven, 2009, 627. 5. More Recent Studies

217

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

Concerning relationship to the καίγε group: The judgment on LXX Canticles’ relation to Barthélemy’s καίγε group depends on whether one adopts a maximalist or a minimalist view of the καίγε problem, as well as from the status one accords to the Greek Minor Prophets Scroll. For the present research I have adopted a maximalist interpretation, considering alleged καίγε texts within an evolution from LXX Pentateuch towards Aquila. The καίγε phenomenon itself, then, consists not so much in a concerted, historically fixed project as in a non-homogeneous tradition of increasingly formal translation methods. In this sense LXX Canticles does belong to it. Among the congeners examined in chapter 12, the ties with LXX Psalms are strongest; the Greek Minor Prophets Scroll is comparable but certainly not directly related; and LXX Ecclesiastes clearly represents a further stage of the development (but does it also represent a later stage, or was its translator being “avant-garde”?). 52

5.7 Summary One can see from the ultra-literal translation techniques in texts belonging to the καίγε tradition and to Theodotion that the resultant Greek conforms to the patterns of the source language to the extent that the Greek is awkward, stilted, and wooden. Careful analysis of the corruption of the earliest form of these texts might reveal the struggles that the early readers had in understanding them. The transliterations of Hebrew words often found in texts belonging to this tradition may express a kind of faithfulness on the part of the translator in admitting obstacles to translation but they also communicate gibberish to the reader. Nonetheless, at the same time, the language of these texts bears the normal characteristics and features well-documented for the Greek of the Hellenistic Period. For the book of Ecclesiastes, certain features attested in the language according to James Aitken may point to a time for the translation of this book later than the other books of the Septuagint. 53 In 3:13, a subjunctive follows future forms and the subjunctive in 9:15 is followed by a future (cf. 9:14; 12:5,7?). This might indicate that future and subjunctive were deemed equivalent at this time. The decline of the infinitive and replacement by ἵνα plus subjunctive is also attested (e. g., ἵνα πορευθῇ “in order to go”). One may also consider ἐποίησεν ἵνα φοβηθῶσιν in 3:14 (“he made them fear”) and compare with Job 5:18 ἀλγεῖν ποιεῖ. Yet in spite of the literal translation and awkward syntax there is a certain Hellenistic flavor to the work and even attention to metrical considerations in Greek at times. 54 Especially in Aquila, although to a certain extent also in Theodotion, many neologisms exist following the proper principles of nominal and verbal formation for Greek. This was because he wanted renderings of all words from a particular root in Hebrew to be all derived from the same root in Greek. 52. De Crom, LXX Text of Canticles, 629. 53. J. K. Aitken, “Phonological Phenomena in Greek Papyri and Inscriptions and their Significance for the Septuagint” in: J. Corely / V. Skemp (eds.), Studies in the Greek Bible: Essays in Honor of Francis T. Gignac, S.J. (CBQ Monographs 44), Washington, DC 2008, 256-277. 54. G. Bertram, “Hebräischer und griechischer Qohelet. Ein Beitrag zur Theologie der hellenistischen Bibel” ZAW 54 (1952), 26-49.

218

5. More Recent Studies

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

6. Conclusions Recent research and study of texts in the καίγε tradition, Theodotion, and Aquila yields the following conclusions: 1. Analysis must be based on the treatment of the translation technique of complete texts, wherever possible. We must not just look for select lexical equivalences as markers of inter-relationship between texts which are recensional. 2. It is important to do contrastive as well as comparative study to establish the patterns and traits of translation technique. 55 3. In discussing possible inter-relationships, the differences as well as the common ground must be explored. 4. It is necessary to correlate the καίγε tradition with tensions between Diaspora and Palestinian Judaism in the Second Temple Period. 5. To properly establish the early history of the Septuagint and the role of the καίγε tradition, we need to delineate the relation of καίγε tradition to both original texts as well as recensions. From the beginning there were different approaches to the task of translation best described as a continuum ranging from extreme formal to extreme functional equivalence. Among translators committed to formal equivalence, certain equivalences grammatically and lexically were appreciated, such as ‫וגם‬/‫ = גם‬καίγε. This became a tradition. Theodotion was a real person and represents a reviser working within the καίγε tradition. Aquila was also a reviser who later took the approach of formal equivalence to extremes. Symmachus represents a backlash, a reaction to Aquila, where equivalence to the parent text must be tempered by concern for the demands of the target language. This typology, in all probability, represents a chronological sequence and not just a typological development. It is based on analysis of the texts themselves rather than descriptions of the revisers from patristic sources written long after the time when they lived and work. We should place the person and work of Theodotion at the same time as the Greek Minor Prophets Scroll, prior to Aquila, probably sometime in the early first century CE. There is no basis or need to speak of a proto-Theodotion or UrTheodotion. The date traditionally given for Theodotion is 180 CE based on the assertion of Epiphanius that Theodotion published his version in the reign of the Emperor Commodus (180-192 CE). 56 Epiphanius’ statement is also the basis for dating the time of writing of Irenaeus’ Adversus Hæreses, which mentions the versions of Theodotion and Aquila together (III.21.1) in a discussion of Isaiah 7:14. 57 Epiphanius, however, who lived in the fourth century, is not a reliable source for dating Theodotion. Moreover, his writings contain contradictions concerning the chronology of the Three. 58 In 55. D. De Crom / E. Verbeke / R. Ceulemans, “A Hebrew-Greek Index to 8ḤevXIIgr” Revue de Qumran 95 (2010), 331-349. 56. Epiphanius, De Mensuris et Ponderibus, 17. See for example, W. W. Harvey, Sancti Irenaei: Libros quinque adversus Haereses, Cambridge 1857, 110. 57. The text of Irenaeus at this point is preserved in Greek in Eusebius, Historia Ecclesiastica, 5.8 58. Cf. Barthélemy, Devanciers, 144-156. 6. Conclusions

219

14. New Ultra-Literal Translation Techniques in Kaige-Theodotion and Aquila

fact, the truth could be the other way round: perhaps Epiphanius’ date for Theodotion is based on the date of Irenaeus’ Adversus Hæreses—his source for the publication of the version of Theodotion? It is interesting to note that Irenaeus, in complaining about the Jewish revisers’ treatment of Isaiah 7:14, mentions them in the order Theodotion and Aquila, i. e. the same order that best suits both typologically and chronologically. Note also that this is a pre-hexaplaric, pre-Origenic statement. Due to the influence of Origen’s Hexapla, the three revisers are normally given in the order Aquila, Symmachus, Theodotion. But the sequence in the Hexapla is no evidence of chronological order. The only other reference to the date of Theodotion is from Jerome in a letter to Augustine, circa 398 CE: I am amazed how you do not read the books of the Seventy in their pure form, as they were published by the Seventy, but rather as emended by Origen or rather corrupted by the obeli and asterisks, and you do not follow the translation of a Christian man, especially when he (Origen) transferred these, which have been added from an edition of a man, a Jew and a blasphemer, after the Passion of Christ. 59

This dates the recension of Theodotion to after Christ’s crucifixion, but does not specify further. It sounds like he is referring to someone in the first century CE. If the information provided by Jerome was common knowledge, Epiphanius’ creative narrative according to his own agenda probably derived a date deduced from the publication of Irenaeus’ Adversus Hæreses. Thus analysis of patristic sources calls for a reevaluation of the traditional date for Theodotion while analysis based on the texts themselves plus patristic testimony suggests the first century CE.

59. Jerome, Epistula 112.19 (et miror, quomodo septuaginta interpretum libros legas non puros, ut ab eis editi sunt, sed ab Origene emendatos siue corruptos per obelos et asteriscos et Christiani hominis interpretatiunculam non sequaris, praesertim cum ea, quae addita sunt, ex hominis Iudaei atque blasphemi post passionem Christi editione transtulerit).

220

6. Conclusions

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts Georg A. Walser

1. Introduction Statistical differences between translation Greek and non-translation Greek are, of course, nothing else than the statistical evaluation and presentation of the differences that actually occur in the texts under consideration. Since differences do not exist on an abstract statistical level they cannot be detected by statistical methods alone. Instead all differences evaluated and presented by statistical methods have to be detected by other methods. Hence, before evaluating the results statistically, it is important to clearly define what is meant by “translation Greek” and how it differs from non-translation Greek. Further, not all the differences that can be detected between the two corpora are suitable for statistical evaluation; it is therefore necessary to establish criteria for determining what differences are suitable for statistical evaluation. However, before addressing these questions, another basic issue must be touched upon, viz., “Why statistics?” Some scholars are reluctant to use statistics. There are several reasons for this. Although the bad reputation of statistics seems to be the main factor, lack of training in statistical methods also contributes to the problem. Who does not know that “statistics lie”? And who has not heard the phrase: “There are three kinds of lies: lies, damned lies, and statistics”? Not much can be done about the ill repute of statistics, although even someone with a basic knowledge of statistics can benefit from them. 1 But what are the benefits of statistics?

2. The Benefits of Statistics First of all it is important to point out that any two corpora of texts have differences that, at first glance, may seem significant, especially when presented as percentages. However, statistics is not only a matter of figures or percentages; it is above all the use of a statistical model that takes into account the size of the corpora and the number of examples. Only when these factors are included is it possible to ascertain whether a given difference between two corpora is statistically significant or only the result of natural variation. It is clear that if a feature occurs once in the Pentateuch and twice in a corpus of non-translation Greek texts it is not sound to assert that its frequency is twice as high in non-translation Greek texts as in translation Greek texts, or even that the difference is significant. But how many examples of a feature are required in order 1.

There are numerous introductions to statistics, quite a few of which focus on linguistics, e. g., P. Cantos Gómez, Statistical Methods in Language and Linguistic Research, Sheffield 2013. 1. Introduction

221

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

to claim that a difference is significant? And how large must the corpora be in order to compare frequencies in a meaningful way? When, indeed, does a difference become significant? First, it should be underlined that chance can never be totally ruled out; this means that it is always possible that a given variation be due to chance. Second, one must be aware that there are no exact answers to these questions; rather, each scholar must determine these levels for him or herself. As a consequence, it is very important to state clearly what statistical method is being used and what exactly is meant by “significant.” By doing so, scholars make it possible for their colleagues to evaluate their research and to compare results. Nevertheless, it is generally accepted that only very large corpora with high occurrence frequencies are suitable for statistical evaluation and that, for a difference between two corpora of texts to be significant, there should normally be no more than a five per cent probability that the difference is due to chance. Moreover, for the two corpora under discussion here, i. e. translation and nontranslation Greek texts regardless of their length or where they be found, not all differences are immediately apparent to the reader. There are several reasons for this: the differences might be very rare, or, though quite frequent, difficult to see, or, as is most often the case, consist not in what is present in the translation Greek texts but what is absent from them. In many cases statistics can be used to highlight these non-apparent, but nevertheless significant, differences. And since statistics takes into account the size of the corpora and number of occurrences, examples that look like “normal” Greek but for which the distribution differs between translation Greek and non-translation Greek texts can be detected and evaluated to decide whether there is a significant difference between the corpora or not. Statistics therefore is concerned with evaluating the differences identified by other methods and deciding whether these differences are significant or not, i. e. whether they are real differences or only natural variations due to chance. Exactly stating what statistical method was used and what is meant by “significant” also facilitates the use of the material by other scholars. 2

3. The Limits of Statistics Since statistical evaluation of variation between translation and non-translation Greek requires large corpora and high occurrence frequency to be reliable, it follows that statistical evaluation is not suitable for small texts or features of which there are very few examples. Moreover, as has already been pointed, statistics cannot replace or in any way exclude the work of linguists with a good knowledge of both translation Greek texts and original Greek compositions, as well as a command of the source languages. Furthermore, statistics cannot—or at least should not—attempt to show what does not exist.

2.

E. g., a statistical evaluation and presentation of the many detailed works by Raymond Martin would be very helpful.

222

3. The Limits of Statistics

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

4. Translation Greek versus Translated Greek It is important to stress the difference between translation Greek and translated Greek. 3 From one point of view, any Greek text translated from another language can be said to contain translation Greek, since it is translated from another language. And even if it were possible to translate a text into pure idiomatic Greek, the text would, according to this point of view, still contain translation Greek, since the text was translated into Greek. The emphasis of this viewpoint is on the process of translation rather than on the final product. 4 From another point of view, any original Greek composition could include translation Greek. Since features typical of translation Greek are isolatable, they can be used by writers, thereby resulting in original Greek compositions that contain translation Greek. 5 This viewpoint emphasizes the final product over the process of translation. One consequence of this is that the number of translational features in a given text does not reveal very much about its origin. Hence statistical evaluation of these features cannot determine whether a text is translated or not. Statistical evaluation of translational features can determine whether or not one text differs significantly from another, but it cannot tell very much about the reason for the difference. In this respect, the second viewpoint appears more adequate: regardless of how the text was composed or translated, it is the final product that should be evaluated. Translation Greek can therefore be defined as Greek that contains language elements from other tongues. Hence statistical differences between translation Greek and non-translation Greek texts are the differences between translation Greek texts, whether translated or not, and non-translation Greek texts, i. e. original Greek compositions that do not contain foreign language elements. It is important to point out that this does not mean that all aspects of translation Greek contain traces of another language. Ordinarily several aspects are affected at the same time, but this does not have to be the case in order to define it as translation Greek. If only one aspect, such as vocabulary, syntax, or word order has traces of another language, this is sufficient for it to be regarded as translation Greek. Naturally, the opposite is true as well, i. e. if one aspect of the target language remains unaffected by the source language, this does not mean that the translated text is not affected at all. This might seem self-evident, but it is worth pointing out. When, at the beginning of last century, Adolf Deissmann noticed that the vocabulary of the New Testament had much in common with the vocabulary of the vernacular papyri, some scholars drew wide-sweeping conclusions about the language of the New Testament as a whole—although most of the aspects of the language had not been investigated at all —especially such aspects as syntax and word order. 6 A similar reaction occurred when 3.

4. 5. 6.

For a more detailed discussion, see: G. Walser, “The Greek of the Bible, Translated or Translation Greek?” in: A. Voitila / J. Jokiranta (eds.), Scripture in Transition: Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo (SJSJ 126), Leiden/Boston, MA 2008. This is the focus of the works of Raymond Martin. Cf. T. V. Evans, Verbal Syntax in the Greek Pentateuch: Natural Greek usage and Hebrew Interference, Oxford 2001, 261. A. Deissmann, Licht vom Osten. Das neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt (fourth edition), Tubingen 1923. 4. Translation Greek versus Translated Greek

223

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

Robert Helbing wrote his book about the case syntax of the verb in the Septuagint.” 7 Helbing noticed that according to his definition there were very few actual Hebraisms; he therefore concluded that this was the case of the Greek of the Septuagint as a whole. Unfortunately, such misconstructions still happen today. Someone makes observations about one aspect of the language, usually vocabulary, and conclusions are drawn about the whole language. No sooner are observations made about one aspect of the language, usually vocabulary, that conclusions are drawn about the language as a whole.

5. The Nature of Translation Greek in the Septuagint Most books of the Septuagint are translated from a Hebrew source, and since it appears that none of the translators had the ambition to render what they translated into idiomatic Greek, as, e. g., Josephus did, they all contain some amount of translation Greek. 8 But again it has to be underlined that although most aspects of the language are affected in some way, not all aspects are affected in the same way. 9 The main reason for this is most certainly the particular translation technique applied by the translators, which resulted in the text being translated without changing more than was necessary. Instead very small units, such as words or expressions, were rendered by Greek equivalents, without change in word order or syntax. 10 The choice of this translation technique is often attributed to some kind of reverence for the sacred original text. This is probably correct, although the case has been somewhat overstated. The main reason however seems to lie in the nature of the respective languages, Hebrew and Greek. Although Hebrew and Greek are very different in nature, a Hebrew original could be rendered into Greek with a minimum of syntax and word order changes, whereas the reverse is not possible. Hebrew is rather paratactic, presenting a fixed word order; Greek, on the other hand, is syntactic, and the word order is quite free. Therefore, it is most often possible to translate a Hebrew text into Greek without changing much of its structure, while preserving the meaning of the original quite satisfactorily. The result of such a translation technique is a Greek text with Hebrew structure, or, in the words of the great Finnish Septuagint scholar Ilmari Soisalon-Soininen: “Wenn ein Kenner des klassischen Griechisch und auch der hellenistischen Koine die Septuaginta zu lesen beginnt, so erhält er den Eindruck, dass ihm diese Sprache ganz fremd ist, es ist eine fremde Sprache mit griechischen Vokabeln. Besonders die Syntax scheint ihm fremd. Die Sprache der Septuaginta ist in ziemlich großem Maße Hebräisch mit grie-

7. R. Helbing, Die Kasussyntax der Verba bei den Septuaginta, Göttingen 1928. 8. Cf. Walser, “Greek of the Bible.” 9. Cf. G. Walser, The Greek of the Ancient Synagogue. An Investigation on the Greek of the Septuagint, Pseudepigrapha and the New Testament, Stockholm 2001, 143-150. 10. Cf. I. Soisalon-Soininen, “Beobachtungen zur Arbeitsweise der Septuaginta-Übersetzer” in: I. Soisalon-Soininen, Studien zur Septuaginta-Syntax (edited by A. Aejmelaeus and R. Sollamo, AASF B, 237), Helsinki 1987, 29-30; R. A. Martin, Syntactical Evidence of Semitic Sources in Greek Documents, Cambridge, MA 1974, 2, note 1.

224

5. The Nature of Translation Greek in the Septuagint

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

chischen Wörtern.” 11 It follows from this statement that a translation that does not alter the structure of the original does not necessarily affect vocabulary, the use of aspects, moods and tenses, or the case syntax. On the other hand, translating a text word by word, from one language into another, especially if the two languages are not even distantly related, necessarily alters the structure, especially the word order, of the target language. This is precisely what happened when the Hebrew Scriptures were translated into Greek. 12 But just as different aspects of the language are variously affected, not all differences of translation Greek are equally detectable. Some aspects, such as vocabulary, are relatively easy to point out; others, such as word order and features absent from translation Greek but common in non-translation Greek, are harder to grasp. Soisalon-Soininen summarized the problem this way: “Die Sprache der Septuaginta ist in ziemlich großem Maße Hebräisch mit griechischen Wörtern. Wenn man aber diese Sprache näher zu untersuchen beginnt, so ist es nicht leicht, im Einzelnen zu bestimmen, worin die speziellen Kennzeichen dieser ‘Übersetzungssprache’ liegen. Die meisten einzelnen syntaktischen Erscheinungen könnten wenigstens in der Koine vorkommen, ganz deutliche Hebraismen gibt es wenig. Dagegen stellt man schon ziemlich bald fest, dass gewisse, im sonstigen Griechisch sich oft wiederholende Ausdrücke fast gänzlich fehlen, andere dagegen, die im sonstigen Griechisch nur ganz selten vorkommen, derart frequent sind, dass sie dem Text leicht einen besonderen Charakter geben.” 13 This “Charakter” can be observed by anyone familiar with non-translation Greek. Statistics is a highly suitable tool for highlighting and presenting such characteristics in a way that is clear and allows for comparison with other aspects of the language. Another factor that should be pointed out is that later translators copied translation features first introduced by the original translators. This is especially true with respect to the Greek translation of the Pentateuch, which was most certainly one of the first parts of the Hebrew Bible to be rendered into Greek. Several features that can be found in the later books of the Septuagint were apparently created by the first translators and then copied by subsequent translators, since it is hardly likely that features such as, e. g., λέγων for ‫ לאמר‬were invented over and over again. However, to fully understand why translators used translation features found in previously translated texts the whole translational activity has to be put into a wider linguistic context. One very typical feature of the Greek language is the use of specific varieties of Greek for different genres, such as Homeric Greek for epos or early Ionian for epigrams. 14 Hence, when the narrative parts of the Pentateuch were translated into Greek using the translation technique discussed above, a new variety of Greek was created, which was associated with the narrative genre found in the Pentateuch. For the subsequent translators and composers the translation Greek of the Pentateuch was not primarily 11. I. Soisalon-Soininen, “Methodologische Fragen der Erforschung der Septuaginta-Syntax,” in: Soisalon-Soininen, Studien zur Septuaginta-Syntax, 42. 12. Cf. A. Aejmelaeus, Parataxis in the Septuagint: A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch, Helsinki 1982; A. Wifstrand, Die Stellung der enklitischen Personalpronomina bei den Septuaginta, Lund 1949-1950; T. V. Evans, Verbal Syntax, 259. 13. Soisalon-Soininen, “Methodologische Fragen,” 42. 14. Cf. Walser, Greek of the Ancient Synagogue, 175-176. 5. The Nature of Translation Greek in the Septuagint

225

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

translation Greek but the “holy” Greek of the Pentateuch, the most holy of scriptures, just as the Greek of Homer was the “holy” Greek of the epos. This variety of Greek specific to the Pentateuch was later used when texts of the same genre were translated or composed, such as Joshua, 1Maccabees, Tobit, Judith, Testament of Abraham, and even the synoptic Gospels. Another very interesting example that clearly shows the use of different varieties of Greek is the fifth-century “translation” of the Gospel of John into Homeric Greek made by Nonnos of Panopolis. Here Nonnos mixes the Gospel genre with the Homeric variety of Greek. On the other hand, some later translators avoided the variety of Greek found in the Pentateuch, most probably because they felt it was too free or lacked reverence for the holy original text. They notably rejected features that in some way changed the original, e. g., the use of δέ for rendering the connective ‫ו‬, which changes the word order, or the use of Greek participles for Hebrew consecutive forms, a technique that obliged the translator to leave some words untranslated. That the need for a more “accurate” rendering of the original was felt is confirmed by the translation of Aquila.

6. Statistical Differences between Translation Greek and Non-Translation Greek Texts What kind of features then are suitable for statistical evaluation and presentation with a view to demonstrating differences between translation Greek and non-translation Greek texts? 15 Features most likely to show differences are those that can be found in either Hebrew or Greek, but have no equivalents in the other language, such as the infinitive absolute in Hebrew or the genitive absolute in Greek. When the infinitive absolute is translated into Greek it is likely to produce an odd Greek expression, and even if the expression can be found in non-translation Greek texts, its distribution will most certainly differ from the distribution in translated Greek texts. When a genitive absolute is found in a translated text it is likely to be used in a way that is unusual in Greek or, at the very least its distribution will not be the same as in non-translation Greek texts. This is true for both the infinitive absolute and the genitive absolute, but unfortunately this alone is not enough for these features to be of any real value. As already pointed out, high frequencies are required for meaningful statistical evaluation. Therefore, not all features that could be regarded as characteristic of translation Greek are suitable for statistical assessment. There are quite a few words and expressions that can safely be regarded as Hebraisms but for which the frequency is far too low to be reliably evaluated by statistical methods. In addition, many of the books of the Septuagint are far too short for the purpose of statistical comparisons; only the larger books can be used on their own, while shorter ones should be grouped together in collections of books. Of course, corpus size and feature frequency interact so that high frequencies make it possible to use smaller corpora, and vice versa. Consequently, suitable features are hard to find. 15. Cf. G. Walser, “Die Wortfolge der Septuaginta” in: M. Karrer / W. Kraus (eds.), Die Septuaginta. Texte, Kontexte, Lebenswelten: internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 20.-23. Juli 2006 (WUNT 219), Tübingen 2008.

226

6. Statistical Differences between Translation Greek and Non-Translation Greek Texts

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

Features that fit the criteria mentioned above include λέγων introducing direct speech and δή as an emphatic particle used in exhortations, both of which can be found throughout the Septuagint. But there are also features that are typical of narrative passages, particularly those of the Pentateuch. One such feature is the use of predicative aorist participles in the nominative case to render Hebrew consecutive forms. All these features, with the possible exception of λέγων, can be found in original Greek compositions, but their distribution is totally different. 16 The participle λέγων (as well as the plural and feminine forms) is a rendering of the Hebrew infinitive construct of ‫ אמר‬plus the preposition ‫ל‬, which is commonly used as a means of introducing direct speech. However, λέγων introducing direct speech can only very occasionally be found in non-translation Greek, and even more rarely immediately before direct speech, as is habitual for the Hebrew ‫לאמר‬. It is not enough to identify an appropriate translational feature, structure and word order must also be taken into account. When this is done it is not uncommon to discover that frequencies vary greatly from one corpus to another, i. e. between translation and non-translation Greek texts. However, it is not enough to show that there are large differences in relative frequency between the two corpora. Before claiming that a difference is significant, it is essential to evaluate it using a statistical model that integrates feature frequency as well as corpus size. Only then is it safe to say whether the difference is significant or due to chance, although this latter possibility can never be completely eliminated. The emphatic particle δή, which has a relatively wide range of usages in non-translation Greek, is used in the Septuagint almost exclusively to render the Hebrew particle ‫ נא‬in exhortations. 17 And since the position of ‫ נא‬is largely predetermined in Hebrew, the Greek particle δή has a very stereotyped usage in translation Greek relative to nontranslation Greek. Again, when investigating a translation feature, a combination of usage, structure, and word order is likely to be fruitful. It should be noticed also that, since δή is not very frequent, quite large corpora are required if it is to be used as a marker of translation Greek. Predicative aorist participles in the nominative case are used to render Hebrew consecutive forms. 18 These forms adhere to a very fixed word order that the literal Greek translation reproduces with an equally rigid word order. In non-translation Greek texts, on the other hand, the use of aorist participles is not only very frequent but also quite varied. Since the overwhelming majority of aorist participles in the nominative case in the Septuagint are renderings of Hebrew consecutive forms, the use of aorist participles in the nominative case is highly stereotyped. Given the high frequency of these participles in the narrative passages of the Septuagint, this feature is not only typical of translation Greek, but statistically significant as well, even for relatively small corpora. These features illustrate the value of statistics as a tool for highlighting the differences between translation and non-translation Greek. The reason for this is that though isolated examples might look like “normal” Greek, the distribution of these features is highly influenced by the Hebrew original. Because it takes both feature fre16. Walser, Greek of the Ancient Synagogue, 79-105. 17. Walser, Greek of the Ancient Synagogue, 131-134. 18. Walser, Greek of the Ancient Synagogue, 20-79. 6. Statistical Differences between Translation Greek and Non-Translation Greek Texts

227

15. Statistical Differences between Translation Greek and Non-Translation Greek Texts

quency and corpus size into account, statistics provides an effective tool for determining which differences between translation and non-translation Greek texts are significant.

7. Summary and Conclusion In summary, this chapter has emphasized four important points. First, statistical differences between translation and non-translation Greek texts are neither more nor less than the statistical evaluation and presentation of differences that actually occur between these two corpora. Second, only large corpora and high occurrence frequency can provide a reliable basis for statistical evaluation. Third, translation Greek should be defined as Greek that contains traces of another language, regardless of how the text was composed or translated. Fourth, features suitable for statistical evaluation and presentation are those that can be found in either Hebrew or Greek, but that have no equivalents in the other language. When structure, especially word order, is taken into account, these features can often reveal significant differences between translation and non-translation Greek texts.

228

7. Summary and Conclusion

IV Local influences

16. Ägyptische Elemente im Griechischen der LXX* Stefan Pfeiffer

Bereits 1927 vertrat L. Th. Lefort die Ansicht, dass im Griechischen der LXX auch Belege für Einflüsse der ägyptischen Sprache zu finden seien. 1 Die Annahme, dass sich Ägyptizismen finden könnten, setzt voraus, dass die Übersetzer zumindest bestimmter Teile der LXX in Ägypten lebten und dort ihr Griechisch einem äußerst engen Kontakt zur ägyptischen Umwelt ausgesetzt war. 2 Der Einfluss der ägyptischen Sprache auf das Koinegriechisch in Ägypten ist jedoch kaum untersucht und dürfte, betrachtet man die papyrologische Dokumentation, relativ gering sein. Es handelt sich meist um grammatikalische Einflüsse, die sich auf der Ebene der Satzstruktur, aber nur ganz selten auf der Ebene der Lexeme 3 bemerkbar machen. So gibt es auch Forscher, die die Meinung vertreten, dass ein Einfluss Ägyptens auf die Sprache der LXX auszuschließen sei: »Überhaupt stand die LXX im ganzen, … unter dem Einfluß palästinischer Tradition, auch wenn sie überwiegend in Alexandrien entstanden ist.« 4 Es ist also nach einem Mittelweg zu suchen – zwar lassen sich, wie unten zu zeigen sein wird, nur wenige tatsächlich auf das Land am Nil zurückzuführende lexikalische Einwirkungen in der LXX finden, doch eröffnet sich daneben ein breites Feld kulturell-religiös eindeutig auf Ägypten zurückzuführender Einflüsse. Mit einiger Sicherheit lässt sich deshalb festhalten, dass das Griechische der Septuaginta nicht unberührt vom Ort der Übersetzung des hebräischen Textes geblieben ist. Die Bedeutung des zeitgenössischen Sprachgebrauchs der Papyri, vor allem was Verwaltungstermini betrifft, hat zuerst Gustav Adolf Deissmann 5 und in jüngerer Zeit Anna Passoni Dell’Acqua in mehreren Beiträgen nachzuweisen versucht. 6 Es ist diesbezüglich freilich zwischen den verschiedenen Büchern des AT zu unterscheiden – besonders bei * 1. 2. 3.

4. 5.

6.

Sehr für eine kritische Durchsicht habe ich den beiden Ägyptologen Günter Vittmann (Würzburg) und Daniel von Recklinghausen (Tübingen) zu danken. L. Th. Lefort, »Pour une grammaire des LXX« Le Muséon 41 (1928), 152-160. Vgl. J. Vergote, »Grec biblique« in: Dictionnaire de la Bible. Supplément III, Paris 1938, 1353. Vgl. die Zusammenstellung ägyptischer Worte, die auch in griechischen Texten vorkommen, bei S. P. Vleeming (Hg.), Aspects of Demotic Lexicography. Acts of the Second International Conference for Demotic Studies. Leiden, 19-21 September 1984, Löwen 1987. M. Hengel, Judentum und Hellenismus. Studien zu ihrer Begegnung unter besonderer Berücksichtigung Palästinas bis zur Mitte des 2. Jh.s. v. Chr. (dritte Ausgabe), Tübingen 1988, 189. G. A. Deissmann, Bibelstudien. Beiträge, zumeist aus den Papyri und Inschriften, zur Geschichte der Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums, Hildesheim/New York, NY 1977 (Nd von 1895). Vgl. zuletzt A. Passoni Dell’Acqua, »Von der Kanzlei der Lagiden zur Synagoge. Das ptolemäische Vokabular und die Septuaginta« in: M. Karrer / W. Kraus (Hg.), Die Septuaginta – Texte, Kontexte, Lebenswelten. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 20.-23. Juli 2006, Tübingen 2008, 236-247. 7. Summary and Conclusion

231

16. Ägyptische Elemente im Griechischen der LXX

Jesaja fällt im Griechischen das ptolemäisch-ägyptische Milieu des Übersetzers auf (vgl. Jes 19,18-21; s. u.). 7 Beschäftigt man sich hingegen mit den »ägyptischen Elementen« in der LXX jenseits der administrativen Einflüsse, die im Grunde genommen nicht spezifisch ägyptisch sondern vielmehr allgemein in den Nachfolgereichen Alexanders des Großen zu finden sind, so steht man vor dem Problem, dass die Forschung zu diesem Thema in den letzten Jahrzehnten nicht viele neue Ergebnisse gebracht hat. Grundlegend ist immer noch der Beitrag von Siegfried Morenz, den er im Jahr 1964 in der Festschrift für Theodor Klauser unter dem Titel »Ägyptische Spuren in den Septuaginta« veröffentlich hat. 8 Hinzu kommen in neuerer Zeit verschiedene scharfsinnige Beiträge des Ägyptologen und Alttestamentlers Manfred Görg. 9 Da eine monographische Gesamtuntersuchung des Themas immer noch ein Desiderat der Forschung ist, zu dessen Erfüllung bereits Morenz aufgefordert hatte, kann im folgenden Überblicksbeitrag nur anhand einschlägiger Beispiele der Versuch einer Systematisierung vorgenommen werden. Wenn wiederum über die ägyptischen Einflüsse auf das Griechische der Septuaginta gesprochen wird, so ist zunächst festzuhalten, dass sich diese auf zwei Ebenen manifestieren, die gleichsam die Gliederungspunkte bilden werden. Das sind 1. die Ebene der ptolemäisch-ägyptischen (also griechisch-makedonischen) Kultur und Sprache in Ägypten, 2. die Ebene der indigen-ägyptischen Kultur und Sprache. Ein Großteil dieser beiden ägyptischen Einflussebenen ist dabei auf Aktualisierungen des in der Zeit der Übersetzung entweder unverständlich gewordenen Urtextes zurückzuführen oder es sind absichtliche Veränderungen, die auf zeitgenössische religiöse bzw. politische Entwicklungen anspielen.

7. 8.

9.

M. Hengel / A. M. Schwemer, Die Septuaginta zwischen Judentum und Christentum, Tübingen 1994, 245. S. Morenz, »Ägyptische Spuren in den Septuaginta« in: A. Stuiber / A. Hermann (Hg.), Mullus. Festschrift Theodor Klauser (Jahrbuch für Antike und Christentum. Ergänzungsband 1), Münster 1964, 250-258 (nachgedruckt in: E. Blumenthal / S. Herrmann (Hg.), Siegfried Morenz. Religion und Geschichte des alten Ägypten. Gesammelte Aufsätze, Köln/Wien 1975, 417-428); hierzu F. Siegert, Zwischen hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta (MJSt 9), Münster 2001, 186-191. Vgl. M. Görg, »Alttestamentliche Kleidernamen ägyptischen Ursprungs« Göttinger Miszellen 15 (1975), 17-18; ders., »Ptolemäische Theologie in der Septuaginta« Kairos 20 (1978), 208-217; ders., »Der ›Fremde‹ (ger): ein Fremdwort im Alten Testament« BN 25 (1984), 10-13; ders., »Methodological Remarks on Comparative Studies of Egyptian and Biblical Words and Phrases« in: S. I. Groll (Hg.), Pharaonic Egypt. The Bible and Christianity, Jerusalem 1985, 57-64; ders., »Die Amtstitel des Potifar« BN 53 (1990), 14-20; ders., »Potifar und Potifera« BN 85 (1996), 8-10; ders., »Ptah in der Bibel« BN 86 (1997), 24-28; ders., »Faszinierendes Ägypten: Ägyptisches in der Josefsgeschichte« Bibel heute 34, 128-131; ders., »Das Übersetzungsproblem in Gen 2,1« BN 95 (1998), 5-11; ders., »Josef in Untersuchungshaft« BN 104 (2000), 16-19; ders., »Die Septuaginta im Kontext spätägyptischer Kultur. Beispiele lokaler Inspiration bei der Übersetzungsarbeit am Pentateuch« in: H.-J. Fabry / U. Offerhaus (Hg.), Im Brennpunkt: Die Septuaginta. Studien zur Entstehung und Bedeutung der Griechischen Bibel, Beiträge zur Wissenschaft vom Alten und Neuen Testament, (Achte Folge, Heft 13, 153), Stuttgart 2001, 115130.

232

7. Summary and Conclusion

16. Ägyptische Elemente im Griechischen der LXX

1. Ptolemäisch-ägyptische Kultur und Sprache 1.1 Ptolemäische Verwaltungssprache Anna Passoni dell’Acqua hat in zahlreichen Aufsätzen nachzuweisen versucht, dass die LXX ptolemäische und damit eine spezifisch ägyptische Verwaltungssprache übernommen hat. 10 So wird das Handeln des ägyptischen Pharaos genauso wie das Handeln Gottes mit Formen der Verben τάσσω »anordnen/befehlen« und ἐντέλλω »anordnen« wiedergegeben. Der Pharao des Exodus erlässt genauso ein πρόσταγμα wie ein ptolemäischer König. 11 Es sei, wie erwähnt, angemerkt, dass solche verwaltungstechnischen Begriffe nicht zwingend auf ptolemäische Einflüsse zurückgeführt werden können/müssen, denn dem Übersetzer standen keine Alternativübersetzungen zur Verfügung und er verwendete einfach die in der gesamten hellenistischen Welt benutzten termini technici, die nichts Spezifisches mit Ägypten zu tun hatten. Das gleiche gilt für die Wortfamilien ἀδικία, ἀνομία, ἄγνοια, ἁμαρτία, ἀσέβεια, πλημμέλεια oder Verben im Kontext von Schuldenerlassen wie ἀφίημι und ἀπολύω. 12 Diese kommen selbstverständlich nicht nur in ptolemäischen Amnestiedekreten, sondern auch in Bibeltexten vor, doch heißt das keinesfalls, dass der Übersetzer hier bewusst ptolemäische Terminologie übernahm, sondern dass er sich einfach des allgemeinen koine-Sprachgebrauchs bediente. So darf man letztlich von der Verwendung hellenistischer Verwaltungsterminologie nicht zwingend auf einen ägyptischen Einfluss schließen. Papyri, die uns diese am reichhaltigsten bieten, stammen schließlich fast ausschließlich aus Ägypten, doch ist es durchaus möglich, dass Seleukiden, Antigoniden und Attaliden sich in ihren Reichen einer ähnlichen, wahrscheinlich sogar der gleichen Verwaltungssprache bedienten. Viel interessanter ist, dass die Übersetzer mancher Bücher der Bibel sich bewusst von zeitgenössischer Terminologie abwandten. Das fällt vor allem in einer Partie auf, die in Ägypten selbst spielt. Hier ist diese Abkehr wahrscheinlich mit der Intention des Übersetzers zu erklären, die Ereignisse in einer fernen pharaonischen Vergangenheit anzusiedeln. Wenn er in diesem Fall die zeitgenössische Verwaltungssprache verwendet hätte, so wäre die Geschichte nicht glaubwürdig genug gewesen. Aus diesem Grund wird der »Hofbeamte des Pharaos, der Befehlshaber der königlichen Leibwache«, bei dem Joseph als Sklave ist, als εὐνοῦχος bezeichnet (Gen 39,1: ὁ εὐνοῦχος Φαραω), was die genaue Übersetzung für das hebräische ‫ סריס‬ist. Hätte man den Beamten korrekt, was meint, der ptolemäischen Terminologie folgend, bezeichnen wol10. A. Passoni Dell’Acqua, »La terminologia dei reati nei decreti di amnistia tolemaici e nella versione dei LXX« in: F. Festorazzi u. a. (Hg.), Sapienza e Torah, Atti della XXIX Settimana Biblica, Roma 15-20 settembre 1986, Bologna 1987, 237-242; dies., »La terminologia dei reati nei προστάγματα dei Tolemei e nella versione dei LXX« in: B. G. Mandilaras (Hg.), Proceedings of the XVIII International Congress of Papyrology Athens 25-31 May 1986, Athen 1988, 335350. 11. Vgl. z. B. Gen 47,26 (hier erlässt Joseph das πρόσταγμα); siehe hierzu A. Passoni dell’Acqua, »Il Pentateuco dei LXX testimone di istituzioni di età tolemaica« Annali di Scienze Religiose 4 (1999), 185-187. 12. Passoni Dell’Acqua, »Von der Kanzlei der Lagiden zur Synagoge«, 246-247. 1. Ptolemäisch-ägyptische Kultur und Sprache

233

16. Ägyptische Elemente im Griechischen der LXX

len, dann wäre er mit der Kennzeichnung φίλος versehen worden. Das gleiche gilt für die Berufsbezeichnungen »Oberster der Köche/Schlachter« (ἀρχιμάγειρος), »Oberbäcker« (ἀρχισιτοποιός) und »Oberhofschenk« (ἀρχιοινοχόος). Sie sind für den Ptolemäerhof nicht belegt. 13 Das Gefängnis wird mit einem allgemeinen Wort für »Umwallung« oder spezieller auch einen Deich als ὀχύρωμα (Gen 39,20) bezeichnet statt als δεσμωτήριον. 14

1.2 Ptolemäische Kultur und Religion Anders als bei der verwendeten Verwaltungssprache, die, wie gesagt, der gesamten Koine zu eigen gewesen sein dürfte, lassen sich im Bereich Kultur und Religion durchaus spezifisch ägyptische Einflüsse festmachen. In Jes 65,11 werden etwa die alten kanaanäischen Lokalgottheiten Gad und Men im Griechischen mit Daimon und Tyche wiedergegeben. Ganz richtig ist mit Seeligmann davon auszugehen, dass hier eine Anspielung auf und Kritik am alexandrinischen Kult für den Agathos Daimon, den Schutzgott der Stadt, zu sehen ist – es sich also um eine »contemporization« handelt. 15 Als weitere Beispiele für solche »transpositions of biblical idolatry in the contemporaneous religious life of Alexandria« wären nach Seeligmann zu nennen: Dtn 23,18 (τελεσφόρος und τελισκόμενος für ‫ קדשה‬und ‫ ;)קדש‬Jer 16,5 (Μὴ εἰσέλθῃς εἰς θίασον für ‫)אל־תבוא בית מרזח‬. Bei den Begriffen τελεσφόρος, τελισκόμενος und θίασος, die in der ptolemäischen Urkundensprache überhaupt nicht vorkommen, ist es aber, wie bereits bei der Verwendung der Verwaltungsterminologie nicht zwingend nötig, einen ägyptischen Hintergrund zu vermuten – es handelt sich vielmehr um allgemeinhellenistische Kultbegriffe. Auch Seeligmanns Ansicht, dass die griechische Übersetzung von Jes 14,12 »Du Morgenstern, wie konnte es geschehen, dass du vom hohen Himmel niederstürztest?« (‫ )איך נפלת משמים הילל בן־שחר‬mit πῶς ἐξέπεσεν ἐκ τοῦ οὐρανοῦ ὁ ἑωσφόρος ὁ πρωὶ ἀνατέλλων auf das alexandrinische Umfeld zurückzuführen sein könnte, ist nicht nachzuvollziehen. Der »Morgenstern«, wörtlich »Morgenbringer« (ἑωσφόρος), soll, so Seeligmann, auf den ἑωσφόρος in Alexandria anspielen, der nach dem von Athenaios in seinem Gelehrtenmahl wiedergegebenen Bericht (V 197d) die berühmte pompe Ptolemaios’ II. anführte. 16 Da es aber kein anderes griechisches Wort für Morgenstern gibt, stand dem Übersetzer überhaupt kein Alternativbegriff zur Verfügung. Es zeigt sich am Beispiel von Daimon und Tyche also, dass man den Text aktualisierte, um auf gegenwärtige Missstände hinzuweisen. An einer anderen Stelle nahm man möglicherweise aufgrund des der jüdischen Religion gegenüber feindlich eingestellten alexandrinisch-ägyptischen Umfeldes eine ebenfalls nur religiös zu begrün13. Vgl. St. Pfeiffer, »Joseph in Ägypten. Althistorische Beobachtungen zur griechischen Übersetzung und Rezeption von Gen 39–50« in: M. Karrer / W. Kraus (Hg.), Die Septuaginta – Texte, Kontexte, Lebenswelten. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 20.-23. Juli 2006, Tübingen 2008, 314 f. 14. Vgl. St. Pfeiffer, »Joseph in Ägypten«, 316 mit weiteren Beispielen. 15. I. L. Seeligmann, The Septuagint Version of Isaiah. A Discussion of its Problems, Leiden 1948, 99 f. (Nachdruck in: R. Hanhart / H. Spieckermann (Hg.), The Septuagint Version of Isaiah and Cognate Studies, Tübingen 2004, 264). 16. Seeligmann, »The Septuagint Version of Isaiah«, 100.

234

1. Ptolemäisch-ägyptische Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

dende Aktualisierung vor. In Jes 19,25 heißt es »Gesegnet bist du, Ägypten, mein Volk«. Das änderte der Übersetzer in »Gesegnet ist mein Volk, das in Ägypten ist.« Hier ist recht deutlich die Diasporasituation des Übersetzers zu erkennen, war es doch schon lange zu einer Inversion des Exodus gekommen. 17 Parallel dazu gibt es aber auch Veränderungen, die auf eine Rücksichtnahme des Übersetzers gegenüber dem Herrscherhaus hinzuweisen scheinen – die Juden in Ägypten unterschieden also sehr genau zwischen feindlichen Ägyptern und freundlichen Herrschern. Ein schönes Beispiel hierfür bietet das mit der Unreinheit in Zusammenhang gebrachte Substantiv »Hase«, ‫ארנבת‬, in Lev 11,6. Man nutzte hierfür wahrscheinlich deshalb nicht das griechische Wort λαγός/λαγώς, weil es zu sehr an den Namen des Vaters des Dynastiegründers Ptolemaios namens Lagos erinnerte. Die Herrscher als Lagidendynastie in Zusammenhang mit der Unreinheit zu bringen, wäre wenig schmeichelhaft für sie gewesen. Stattdessen wählte man δασύπους, also die »behaarte Pfote«. 18

2. Einflüsse der indigen-ägyptischen Kultur und Sprache 2.1 Onomastik 2.1.1 Personenonomastik Manche Partien der Bibel spielen in Ägypten oder beziehen sich auf Ägypten, und es begegnen uns deshalb auch ägyptische Personennamen. 19 Am bekanntesten dürfte wohl sein, dass der Name des Mose (Μωϋσῆς/‫ )משה‬ägyptischen Ursprungs ist (abgeleitet von msj – »gebären«), die Erklärung in Ex 2,10 also sekundär hebräisch ist. 20 Diese ägyptische Wurzel hat der Übersetzer dieser Stelle in hellenistischer Zeit aber allem Anschein nach nicht erkannt, denn er übersetzt Moses nicht mit Mose/Mosis, wie Pharaonennamen, die aus derselben Wurzel mit msj gebildet wurden (Thut-mose oder Ah-mose). Vielleicht war in dieser Zeit aber auch die griechische Aussprache des 17. Vgl. hierzu St. Pfeiffer / D. von Recklinghausen, »Inversion des Exodus: Aus der Sklaverei in die Freiheit. Juden im frühptolemäischen Ägypten« in: H. Knuf u. a. (Hg.), Honni soit qui mal y pense. Studien zum pharaonischen, griechisch-römischen und spätantiken Ägypten zu Ehren von Heinz-Josef Thissen, Leuven u. a. 2010, 405-418. 18. Vgl. W. Huß, »Die Juden im ptolemaiischen Ägypten. Ein Beitrag zur Geschichte einer multikulturellen Gesellschaft« in: S. Füssel u. a. (Hg.), Artibus. Kulturwissenschaft und deutsche Philologie des Mittelalters und der frühen Neuzeit. Festschrift für D. Wuttke zum 65. Geburtstag, Wiesbaden 1994, 4; Passoni Dell’Acqua, »Von der Kanzlei der Lagiden«, 241. 19. Vgl. hierzu Y. Muchiki, Egyptian Proper Names and Loanwords in North-West Semitic, Atlanta, GA 1999, 207-229 (Personennamen); 229 (Gottesnamen); vgl. die Rezensionen von J. F. Quack, in: Review of Biblical Literature 2000 (www.bookreviews.org) und Th. Schneider, in: JQR 92, 2001, 155-165. 20. M. Görg, »Mose – Name und Namensträger. Versuch einer historischen Annäherung« in: E. Otto (Hg.), Mose. Ägypten und das Alte Testament (SBS 189), Stuttgart 2000, 17-42, mit weiterer Literatur; H.-J. Thissen, »Zum Namen ›Moses‹« Rheinisches Museum für Philologie 147 (2004), 55-62; Th. A. G. Hartmann, »Mose und Maria – »Amuns Kind und Liebling« – Auf den ägyptischen Spuren zweier biblischer Namen« ZAV 116 (2004), 615-618. 2. Einflüsse der indigen-ägyptischen Kultur und Sprache

235

16. Ägyptische Elemente im Griechischen der LXX

Personennamens Mosche zu fest verankert, als dass er nach der ägyptischen Wurzel hätte verändert werden können, oder aber das Hebräische Mosche wurde in dieser Zeit derart ausgesprochen. 21 Die griechische Wiedergabe des Personennamens wiederum könnte, wie manche meinen, ebenfalls eine ägyptische Etymologie aufweisen und sich auf mw-ḥsj zurückführen. 22 Das erste Wort (mw) heißt »Wasser«, das zweite bezeichnet denjenigen, der ein »gepriesener« ist oder der, in dieser Interpretation, aus dem Wasser »gerettet« wurde. Gegen eine solche Ansicht wandte sich jedoch Thissen, der zur gleichen Etymologie unter Hinzunahme der griechischen Wiedergabe eines anderen Verbes kommt. Es ist nicht ḥsj, sondern das Partizip Perfekt Passiv von ṯꜢj – »der Genommene, Getragene«. 23 Moses ist also »der aus dem Wasser genommene«; die Übersetzer haben damit auf die hebräische Vorlage von Ex 2,1-10 angespielt. 24 Anders sieht das bei den drei im Folgenden zu besprechenden Namen aus: So erscheint in der Josephsgeschichte (Gen 39,1) der Sklavenhalter Potiphar (‫– )פוטיפר‬ ein Name der auch ohne masoretische Punktation so gelesen werden muss. Der Übersetzer gab dies wenig ähnlich mit Petephres (Πετεφρης) wieder. Beide Namen sind jedoch identisch: Es handelt sich um lautliche Wiedergaben des ägyptischen Namens PꜢ-dj-pꜢ-Rʿ (»Der, den Re gegeben hat«). Die lautliche Unähnlichkeit erklärt sich allein aus der Tatsache, dass der ägyptische Name in griechischer Zeit so ausgesprochen wurde, wie er in der LXX steht. Der Übersetzer muss erkannt haben, dass sich hinter der hebräischen Namenswiedergabe der aktuelle ägyptische Personnenname Petephres verborgen hat. 25 Ohne Grundkenntnisse der ägyptischen Onomastik wird ihm das wiederum kaum möglich gewesen sein. Auch die beiden anderen in der Josephsgeschichte erwähnten ägyptischen Namen lassen sich als theophore Namen erklären. So hieß der Re-Priester (Gen 41,45), dessen Tochter Joseph später heiraten sollte, Petephre/Πετεφρη (PꜢ-dj-pꜢ-Rʿ ), 26 hebräisch Potiphera (‫)פוטי פרע‬. 27 Die Tochter des Priesters trug den Namen Asenneth/Ασεννεθ, bzw. hebräisch Asenath (‫)אסנת‬, ägyptisch »die zur Göttin Neith zugehörige« (Ns-Nt), 28 21. Für Loprieno, in: Thissen, »Zum Namen«, 58, handelt es sich um die zeitgenössische Aussprache, weil das lange /o/ mit /ōu/ wiedergegeben wurde; Hartmann, »Mose und Maria«, 618, verweist auf die Schreibung des Gottesnamens Θωτ, Θωυτ als Parallele. 22. A. H. Gardiner, »The Egyptian Origin of Some English Personal Names« JAOS 56 (1936), 195, Anm. 28; J. Černý, »Greek Etymology of the Name Moses« ASAE 41 (1942), 349-354; Y. Koenig, »Quelques ›égyptianismes‹ de la Septante« BIFAO 98 (1998), 223 f. 23. Thissen, »Zum Namen«, 55-62. 24. Thissen, »Zum Namen«, 61. 25. Vgl. hierzu Pfeiffer, »Joseph in Ägypten«, 317 f. 26. Görg, »Die Septuaginta im Kontext«, 127, sieht im Namen ebenfalls eine Wiedergabe des ägyptischen Personennamen PꜢ-dj-pꜢ-Rʿ . Das entspricht dem Griechischen Πετεφρη. Weshalb er der Ansicht, ist, dass »die Übersetzer die Kunstbildung des alten Schriftstellers nicht mehr als solche ausmachen konnten«, ist mir unverständlich. 27. Zur der Problematik, weshalb der gleiche Personenname in zwei Aussprachen auftritt, vgl. D. Redford, A Study of the Biblical Story of Joseph (Genesis 37-50), Leiden 1970, 136 f., der hier unterschiedliche Erzähltraditionen wirksam sieht, die später harmonisiert werden mussten. 28. Vgl. A. R. Schulman, »On the Egyptian Name of Joseph: A New Approach« SAK 2 (1975), 238239; Görg, »Die Septuaginta im Kontext«, 127, möchte für das Hebräische nicht den ägyptischen Satznamen »Sie gehört der Neith« annehmen, sondern das ägyptische Wort »Schwester« oder »Gemahlin« (sn.t), als Vorlage sehen. Das scheint mir sehr unwahrscheinlich, denn

236

2. Einflüsse der indigen-ägyptischen Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

wofür es nur einen fraglichen ägyptischen Beleg gibt. 29 Das Griechische könnte auch für den (ansonsten nicht belegten) ägyptischen Namen Jw=s-n-N.t stehen 30 – was das gleiche heißt und bedeuten würde, dass der Übersetzer den Namen in der gleichen Bedeutung anders wiedergegeben hat. Ein eigenes Problem bildet hingegen der ägyptische Name, den Joseph vom Pharao erhielt. Er ist in seiner griechischen Form Psonthomphanech (Ψονθομφανηχ) bzw. hebräisch Zafenat-Paneach (‫ )צפנת פענח‬jeweils eindeutig als ägyptisch zu identifizieren, entzieht sich bisher aber einer überzeugenden Übersetzung. Die hebräische Form zeigt das ägyptische Namensmuster »Gott N.N. hat gesagt: er möge leben« (ḏd-pꜢ-Gottesname-jw=f-ʿ nḫ). 31 Wahrscheinlich ist, dass kein spezifischer Gottesname eingefügt wurde, sondern der Name »Der Gott sagt: er möge leben« (ḏd-pꜢ-nṯr-jw=f-ʿ nḫ) lautet. Damit wäre dieser Name eine Erfindung des hebräischen Verfassers der Geschichte, der seine eigene monotheistische Theologie mit in das Werk eingebracht hat. 32 Diese Etymologie ist aber problematisch und der Name ist ansonsten nicht belegt. Falls es aber stimmen sollte, so bedeutet das zwingend, dass der Verfasser der Geschichte nicht nur ägyptisch konnte, sondern auch kreativ mit dieser Sprache umzugehen wusste. Andererseits sei aber darauf hingewiesen, dass es den ägyptischen Personennamen Ns-pꜢ-nṯr »Er gehört Gott« gibt. 33 Es muss sich also bei der Verwendung des Wortes Gott ohne weitere Kennzeichnung, um welchen Gott es sich handelt, nicht zwingend um eine monotheistische Tendenz beim ägyptischen Namen Josephs handeln. Das Problem des Josephnamens liegt aber darin, dass der Übersetzer, anders als bei den Ägyptern, die in der Geschichte auftauchen, dessen Name nicht einfach in eine moderne Aussprache gebracht hat. Vielmehr hat er ihn durch den anderen, ebenfalls ägyptisch zu erklärenden Namen Psonthomphanech ersetzt. 34 Der Name würde »Der Herzenserfreuer auf ewig« (pꜢ-snḏm-r-nḥḥ) oder »Der Herzenserfreuer mit Leben« (pꜢ-snḏm-jb-ʿ nḫ) lauten. 35

29. 30. 31.

32.

33. 34. 35.

ebenso wie es den Vaternamen der Aseneth Potifera in Ägypten gab, so gab es auch den Namen der Tochter – man ist also nicht dazu gezwungen, auf Umwegen zu einer neuen Deutung zu kommen. E. Lüddeckens u. a., Demotisches Namenbuch, Wiesbaden 1980-2001, 706. Vergote, Joseph en Égypte, 149, Schulman, »On the Egyptian Name of Joseph«, 239, Anm. 19. Zu den möglichen Deutungen des ägyptischen Namens vgl. H. Seebaß, Genesis III. Josephsgeschichte (37,1–50,26), Neukirchen-Vluyn 2000, 71-72; Schulman, »On the Egyptian Name«, 239-243; V. P. Hamilton, The Book of Genesis. Chapters 18–50, Grand Rapids, MI 1995, 507-508; M. Görg, Die Beziehungen zwischen dem alten Israel und Ägypten. Von den Anfängen bis zum Exil, Darmstadt 1997, 121; vgl. auch den alternativen Deutungsversuch von H. L. F. Lutz, »The Egyptian Archetyp of saphenath paeneah and Ψονθομφανηχ (Genesis 41: 45)« University of California Publications in Semitic Philology 10, 15 (1945), 289-294. Görg, »Die Septuaginta im Kontext«, 126; Y. Volokhine, »L’Égypte et la Bible: histoire et mémoire. À propos de la question de l’Exode et de quelques autres thèmes« BSEG 24 (20002001), 86; A. Loprieno, La pensée et l’écriture. Pour une analyse sémiotique de la culture égyptienne, Paris 2001, 127. Lüddeckens u. a., Demotisches Namenbuch, 669. Görg, Beziehungen, 121; Seebaß, Genesis III, 72, deuten die griechische Form als einen, allerdings in der sonstigen Überlieferung unbekannten Ehrentitel. Görg, »Die Septuaginta im Kontext«, 126. 2. Einflüsse der indigen-ägyptischen Kultur und Sprache

237

16. Ägyptische Elemente im Griechischen der LXX

2.1.2 Ortsonomastik Die Ortsonomastik ist für die Frage nach ägyptischen Elementen in der LXX vor allem deshalb von Interesse, weil hier zu sehen ist, inwiefern der Übersetzer intime Kenntnisse über Ägypten besaß. 36 Das zeigt sich recht gut am Beispiel der von den Griechen Heliopolis genannten ägyptischen Stadt Iunu nahe des heutigen Kairo, die im Hebräischen On heißt, ein Name, der direkt aus dem Ägyptischen übernommen wurde. In der LXX können nun beide Versionen des Städtenamens erscheinen. In Gen 41,45 und 51 steht Heliopolis (῾Ηλίου πόλις), in Ex 1,11 hingegen »On, das Heliopolis ist« (Ων, ἥ ἐστιν ῾Ηλίου πόλις). Der Übersetzer musste also wissen, wie die entsprechende ägyptische Stadt von den Griechen benannt wurde. Bemerkenswert ist weiterhin, wie der Übersetzer mit anderen ägyptischen Orts-/ Ländernamen verfährt. So ist an einer Stelle vom »Land Patros« (Jer 44,1: ‫)ארץ פתרוס‬ als Heimat der Ägypter die Rede, was im Griechischen mit ἐν γῇ Παθουρης wiedergegeben wird. Vielleicht leitet sich das hebräische Wort vom ägyptischen pꜢ-tꜢ-rsj – »das Südland« ab. 37 Zahlreiche ägyptische Städtenamen kommen in Hes 30,13-18 vor: Memphis (‫)נף‬, Patros (‫)פתרוס‬, Zoan (‫)צען‬, No (‫)נא‬, Sin (‫)סין‬, On (‫)און‬, Pi-Beset (‫ )פי־בסת‬und Tachpanhes (‫)תחפנחס‬. Nicht mit dem ursprünglichen Namen Jnb.w-ḥḏ transliteriert ist Memphis, das hebräisch als Noph wiedergegeben wird. Dieses Noph (in Hos 9,6 Moph) dürfte aber nicht die griechische Aussprache aufgreifen, sondern eine Wiedergabe des ägyptischen Alternativnamens der Stadt Memphis Mn-nfr aufgreifen. 38 Für vier Städte gibt der Übersetzer die korrekten griechischen Namen, muss sich also entsprechend gut ausgekannt haben, wohingegen im Hebräischen Transliterationen der ägyptischen Ortsnamen vorliegen. Neben dem bereits besprochenen On wären das Tanis für hebräisch Zoan, das vom ägyptischen Namen Ḏʿ n.t abgeleitet ist. 39 Diospolis, als »Stadt des Zeus«, nannten die Griechen das ägyptische Theben. Das hebräische Wort No für diese Stadt leitet sich vom ägyptischen njw.t »die Stadt« ab – denn für die Ägypter war Theben »die Stadt«. 40 Bubastis, die Stadt der Göttin Bastet schließlich war ägyptisch Pr-BꜢst.t. 41 Das auslautende r von Pr fällt im status constructus üblicherweise aus, so dass das hebräische Pi-Beset die halbwegs korrekte lautliche Wiedergabe des ägyptischen Ortsnamens ist (eigentlich müsste es Ubaste lauten). Unbekannt ist hingegen, wie gesagt, Patros (gr. Παθουρης). Für Tachpanhes (Ταφνας) schlägt die Forschung überzeugend eine Identifikation mit Tell Defenneh,

36. Vgl. Muchiki, Egyptian Proper Names, 229-236. 37. Vgl. hierzu G. Vittmann, Der demotische Papyrus Rylands 9, Wiesbaden 1998, 287-290; B. Porten, Archives of Elephantine. The Life of an ancient Jewish Military Colony, Berkeley, CA/Los Angeles, CA 1968, 42 f. 38. G. Fecht, »Wortakzent und Silbenstruktur« Ägyptologische Forschungen 21, 1960, 43 f. § 81, zur hebräischen Wiedergabe 43 Anm. 136; Günter Vittmann wies mich brieflich darauf hin, dass das Hebräische Moph das äg. Menfe o. ä. (also mit auslautendem Kurzvokal) mit Assimilation des n an das f im Hebräischen widerspiegeln dürfte (*monf > moph). »Nof« hingegen ist eindeutig sekundär mit Dissimilation des m > n. 39. Vgl. M. Römer, »Tanis«, in: Lexikon der Ägyptologie VI, 194 40. W. Westendorf, »Theben«, in: Lexikon der Ägyptologie VI, 465. 41. L. Habachi, »Bubastis«, in: Lexikon der Ägyptologie I, 873.

238

2. Einflüsse der indigen-ägyptischen Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

dem von Herodot (II 30, 154) Δάφναι genannten Ort, vor. 42 Problematisch ist zudem die Wiedergabe von Sin – dieses ist in V 15 mit Sais übersetzt und in V 16 mit Syene. Syene wäre hieroglyphisch Swnw, Sais wäre SꜢw – weder der eine noch der andere Ortsname hat Ähnlichkeiten mit dem hebräischen Sin. Der Verfasser des hebräischen Textes hat hiermit aller Wahrscheinlichkeit nach die Grenzfestung und -stadt Pelusion bezeichnen wollen. 43 Diese Festung hieß in pharaonischer Zeit noch Sjn, wurde also im Hebräischen korrekt angesprochen. Wohl erst unter den Ptolemäern hieß die Stadt dann PꜢ-jr-Jmn/Pr-jr-Jmn. Der ursprüngliche Name scheint dem Übersetzer nicht bekannt gewesen zu sein, so dass er zweimal unterschiedliche, ihm gut bekannte Städtenamen verwendete: Die eine befand sich an der Südgrenze, die andere im Norden, im Nildelta – in dem Fluchspruch über Ägypten sollten schließlich möglichst viele bedeutende Städte der Vernichtung anheimfallen. Die Verdeutlichung dieses holistischen Aspektes war dem Übersetzer sogar so wichtig, dass er auch einmal Memphis (V 14) durch Theben/No ersetzt (V 15). Ein Beispiel für die griechische Wiedergabe einer typisch ägyptischen Landesbezeichnung ist Hes 29,10, wo es heißt: »von Migdol, 44 Syene bis an die Grenze von Kusch« (‫)ממגדל סונה ועד־גבול כוש‬, denn Kusch (‫)כוש‬, wie auch die Ägypter es nannten (Kš), wird mit »(Land der) Äthioper« übersetzt (ἀπὸ Μαγδώλου καὶ Συήνης καὶ ἕως ὁρίων Αἰθιόπων). Auch hier hätte sich als griechische Transliteration dieser Zeit Χους angeboten, 45 doch weiß der Übersetzer, dass das von den Ägyptern und Hebräern Kusch ausgesprochene Land von den Griechen Äthiopien genannt wurde. Die Landesbezeichnung Kusch bereitet dem Übersetzer aber auch an anderer Stelle desselben Buches Probleme. In Hes 30,4-5 heißt es: »Und das Schwert soll über Ägypten kommen, und Kusch wird erschrecken, wenn die Erschlagenen in Ägypten fallen und sein Reichtum weggenommen und seine Grundfesten eingerissen werden. Kusch und Put und Lud mit allerlei fremdem Volk und Kub und ihre Verbündeten sollen mit ihnen durchs Schwert fallen.« (‫ובאה חרב במצרים והיתה חלחלה בכוש בנפל חלל‬

‫במצרים ולקחו המונה ונהרסו יסודתיה׃ כוש ופוט ולוד וכל־הערב וכוב ובני ארץ הברית אתם‬ ‫בחרב יפלו׃‬, καὶ ἥξει μάχαιρα ἐπ᾽ Αἰγυπτίους, καὶ ἔσται ταραχὴ ἐν τῇ Αἰθιοπίᾳ,

καὶ πεσοῦνται τετραυματισμένοι ἐν Αἰγύπτῳ, καὶ συμπεσεῖται αὐτῆς τὰ θεμέλια. Πέρσαι καὶ Κρῆτες καὶ Λυδοὶ καὶ Λίβυες καὶ πάντες οἱ ἐπίμικτοι καὶ τῶν υἱῶν τῆς διαθήκης μου μαχαίρᾳ πεσοῦνται ἐν αὐτῇ). Hier sieht man sehr schön, dass der Übersetzer entweder nicht verstanden hat, was er übersetzt hat, oder dass zwei Übersetzer am Werk waren. Der Übersetzer von Hes 29,10 und 30,4 hat das hebräische Wort ‫ כוש‬noch als Umsetzung der ägyptischen geographischen Bezeichnung für Äthiopien erkannt – denn für Ägypter handelte es sich um Kusch. Im vorliegenden Fall hingegen versucht er, die ihm unverständlichen Toponyme oder Ländernamen zu interpretieren. Kusch wird auf diese Weise das Land der Perser, Put das

42. Vgl. F. Contardi, »Tachpanhes«, in: wibilex 2009. 43. Vgl. zu Pelusium: J.-Y. Carrez-Maratray, Péluse et l’angle oriental du Delta égyptien aux époques grecque, romaine et byzantine, Kairo 1999. 44. Vgl. hierzu A. F. Rainey, »LXX Toponymy as a Contribution to Linguistic Research« Lingua Aegyptiaca 9 (2001), 181 f. 45. Vgl. T. Säve-Söderberg, »Kusch«, in: Lexikon der Ägyptologie III, 889. 2. Einflüsse der indigen-ägyptischen Kultur und Sprache

239

16. Ägyptische Elemente im Griechischen der LXX

Land der Kreter, Lud das Land der Lydier und Kub dasjenige der Libyer. In Vers 9 hingegen ist das hebräische ‫ כוש‬wieder korrekt mit Äthiopien übersetzt.

2.2 Ägyptische naturräumliche Begebenheiten Besonders dann, wenn der hebräische Text in Bezug auf naturräumliche Begebenheiten verändert wurde, ist davon auszugehen, dass die Ursache hierfür in der aktuellen Situation im Land der mutmaßlichen Übersetzung zu suchen ist. Manche derartige Aktualisierungen lassen sich in der Tat nur dann verstehen, wenn man den Übersetzer als in Ägypten tätig sieht. In der Liste der für den Verzehr verbotenen Tiere im hebräischen Text von Lev 11,17 ist beispielsweise der Uhu (?) (‫ )ינשוף‬genannt – ein Wort, das man im Judentum hellenistischer Zeit bereits nicht mehr mit einem spezifischen Vogel zu verbinden wusste. Der Übersetzer ersetzte dieses ihm allen Anschein nach ebenfalls unbekannte Tier einfach durch den Ibis (ebenso in Jes 34,11) – den in Ägypten besonders verehrten Vogel, der mit dem Gott der Schrift und Weisheit Thoth verbunden wurde. Zwar ist nicht bekannt, was ein ‫ ינשוף‬ist, doch kann im hebräischen Text kaum der ägyptische Ibis gemeint gewesen sein, weil er in Syrien-Palästina nicht heimisch war. 46 Einen weiteren derartigen Fall bietet das Beispiel des Ostwindes, wörtlich des »von vorne kommenden Windes« (‫)רוח קדים‬, der an zahlreichen Stellen erwähnt wird. Der Ostwind war religiös gesehen das strafende Werkzeug Gottes. 47 Entsprungen ist eine solche Vorstellung der realen meterologischen Erfahrung in Israel-Judäa. 48 In Ägypten hingegen war der Ostwind der Wind, der, wie es in den religiösen Texten heißt, »die Pflanzen erblühen lässt«. 49 Unmöglich hätte man einer in Ägypten lebenden Leserschaft eine vernichtende Wirkung des segen- und fruchtbarkeitbringenden Ostwindes erklären können. Aus diesem Grund führten die Übersetzer andere Wendungen an; entweder ist es der »verbrennende Wind« (Hos 13,15; Jer 18,17; Hes 17,10, 19,12; Jon 4,8: ἄνεμος καύσων, πνεῦμα καύσων, καύσων) oder es ist der Südwind (Ps 77,26; Hiob 38,24, Hes 27,26: ἄνεμος νότος, νότος, πνεῦμα τοῦ νότου). Auch gibt es Wendungen wie »Zorneswind« (Jes 27,8: πνεῦμα θυμοῦ) und »gewaltiger Wind« (Ps 47,8: πνεῦμα βίαιον). Morenz wies darauf hin, dass gerade wenn für den Ostwind das Wort Südwind gesetzt wurde, ganz klar ägyptisches Lokalkolorit mit im Spiel gewesen sein muss, denn in Griechenland brachte der Südwind Regen, und daher verband ein Grieche in seiner Heimat mit dem Südwind Fruchtbarkeit und nicht Trockenheit. 50 In Ägypten wiederum kann der Südwind der »Dörrende« (šhb koptisch schoheb) sein –

46. Die scheinbar »spezifische Polemik gegen ägyptische Kultbestände«, die Morenz, »Ägyptische Spuren«, 423, aufgrund der Bedeutung des heiligen Tiere für die Ägypter geltend macht, ist nicht überzeugend belegt. 47. Vgl. Jes 27,8; Jer 18,17; Ps 48,8; Hiob 27,22. 48. Vgl. Morenz, »Ägyptische Spuren«, 423-425. 49. A. Gutbub, »Die vier Winde im Tempel von Kom Ombo (Oberägypten)«, in: O. Keel (Hg.), Jahwe-Visionen und Siegelkunst. Eine neue Deutung der Majestätsschilderungen in Jes 6, Ez 1 und 10 und Sach 4, Suttgart 1977, 349. 50. Morenz, »Ägyptische Spuren«, 425.

240

2. Einflüsse der indigen-ägyptischen Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

eine Zuschreibung, die dem sonst auch in Ägypten mit positiven Aspekten belegten Wind 51 erst seit der ptolemäischen Zeit zukommt. 52 Eine zu den Windrichtungen konvergierende Beobachtung machte Bogaert in Bezug auf die nach den Himmelsrichtungen angegebene Ausrichtung des Heiligtums von Ex 27,9-13. 53 Im hebräischen Text sind die Längsseiten des Tempels nach Süden (‫ )נגב־תימנה‬und Norden (‫ )צפון‬ausgerichtet. Der Übersetzer gibt für die südliche Längsseite eine Ausrichtung nach λίψ (πρὸς λίβα), das durchaus den Süden bezeichnen kann. Im vorliegenden Fall muss er aber zwingend den Westen meinen, denn die gegenüberliegende Längsseite ist der Osten – ἀπηλιώτης (πρὸς ἀπηλιώτην). Die Verwendung von λίψ ist wiederum insbesondere in den Papyri Ägyptens mit dem (Süd-) Westen verbunden, 54 also nicht dem Süden, weil sich Libyen im Westen Ägyptens befand. Es liegt also nahe, dass der Übersetzer in Ägypten arbeitete. Sieht man ihn in Ägypten ansässig, so erklärt sich auch, weshalb er das hebräische »zum Meer hin« (‫)ים‬ nicht mit dem Osten verbindet, sondern mit dem Norden verbindet, wo aus Sicht Ägyptens das Mittelmeer liegt, wenn er es wörtlich mit κατὰ θάλασσαν übersetzt – hierzu korrespondiert dann in der Tat der Süden (πρὸς νότον). So erklärt sich also die Verwendung des Wortes λίψ aus der Tatsache, dass der Übersetzer das Hebräische »zum Meer hin« nicht als »Westen«, sondern als »Norden« erkannte und daraus folgernd eine Umorientierung des Heiligtums vornahm.

2.3 »Fehlübersetzungen« In Jer 46,14 f. (= LXX 26,14 f.) steht für die hebräische Wendung »Wie geht es zu, dass deine Gewaltigen zu Boden fallen (‫ )נסחף אביריך‬und können nicht standhalten?« die griechische Übersetzung »warum ist der Apis geflohen, dein erwählter Stier ist nicht geblieben« (διὰ τί ἔφυγεν ὁ ῏Απις; ὁ μόσχος ὁ ἐκλεκτός σου οὐκ ἔμεινεν). Der »Gewaltige/Starke« (‫ )אביר‬des Hebräischen ist also im Griechischen mit dem Namen des ägyptischen Gottes Apis übersetzt. Morenz erklärt das damit, dass der Übersetzer statt ‫ נסחף‬ohne masoretische Punktation und in scriptura continua ‫ נס חף‬gelesen habe und daraus sei dann ὁ ῏Απις ὁ μόσχος geworden. So sei der gesamte Vers als parallelismus membrorum aufgefasst worden. 55

2.4 Hebräisch-ägyptische Lehnworte Es gibt auch Fälle, die zeigen dass der Übersetzer die ägyptische Sprache gut beherrschte und sogar ägyptische Lehnworte 56 korrekt erkannte: Das zeigt das aus dem Ägyptischen übernommene Lehnwort für »Riedgras« achu (‫ )אחו‬und dessen Übertra51. Vgl. D. Kurth, »Wind«, in: Lexikon der Ägyptologie VII, 1267. 52. Wörterbuch der ägyptischen Sprache IV, 529,9; vgl. auch die Beobachtungen zur Ausrichtung des Jerusalemer Heiligtums in Ex 27 von P.-M. Bogaert, »L’orientation du parvis du sanctuaire dans la version grecque de l’Exode (Ex., 27, 9-13 LXX)« L’Antiquité classique 50 (1981), 79-85. 53. P. M. Bogaert, »L’orientation du parvis du sanctuaire«, 79-85. 54. F. Preisigke, Wörterbuch der griechischen Papyrusurkunden II, Heidelberg 1927, 25 f. 55. Morenz, »Ägyptische Spuren«, 257; vgl. aber D. Barthélemy, Critique textuelle de l’Ancien Testament. 2. Isaïe, Jérémie, Lamentations, Göttingen 1986, 763 f. 56. Muchiki, Egyptian Proper Names, 236-258. 2. Einflüsse der indigen-ägyptischen Kultur und Sprache

241

16. Ägyptische Elemente im Griechischen der LXX

gung ins Griechische (Gen 41,3, Sir 40,16 und Jes 19,7). 57 Hierfür steht in der LXX undeklinabel ἄχι. Da es sich um einen griechischen Neologismus handelt, geht man davon aus, dass der Übersetzer das im hebräischen Text stehende Lehnwort direkt ins Griechische übernommen hat. Ein Vorgang, den man als Transliteration oder Transkription bezeichnet und der des Öfteren in der LXX zu beobachten ist. 58 Es gibt jedoch auch eine andere Lösung: Das hebräische Wort ‫ אחו‬ist ein Lehnwort, das sich vom ägyptischen Begriff achi (Ꜣḫj) »Pflanzen« oder achet (Ꜣḫ.t) »Fruchtland« ableiten lässt. 59 Im ptolemäerzeitlichen Ägyptisch wurde dieses nun genauso ausgesprochen, wie es der Übersetzer der LXX lautlich wiedergegeben hat, also nicht achu, mit einem hebräischen waw als letztem Konsonaten, sondern achi, mit einem ägyptischen jod als letztem Konsonanten (demotisch: Ꜣḫj oder Ꜣḫr). In der Koine sprach man das auslautende -ei in achei aufgrund des Iotazismus als -i aus. Es liegt uns im griechischen Text folglich ein ägyptisches Wort in der korrekten zeitgenössischen Aussprache vor. Es gibt aber auch noch weitere Lehnworte, die zeigen, wie gut der Übersetzer mancher Partien sich in der ägyptischen Sprache auskannte. In Gen 37,3 heißt es: »Deshalb ließ er (Jakob) ihm (Joseph) ein prächtiges Gewand machen.« Die Septuaginta liest hier ἐποίησεν δὲ αὐτῷ χιτῶνα ποικίλον, also: »er fertigte ein buntes Gewand für ihn an«. Die hebräische Entsprechung für ποικίλον lautet ‫ – פסים‬seine Übersetzung lässt sich in der hebräischen Lexikographie nicht zufriedenstellend erklären. Görg wies nun darauf hin, dass das ägyptische Lexem psj (»kochen, färben«) gut mit der griechischen Wortbedeutung in Übereinstimmung zu bringen ist. Es könnte sich also bei ‫ פסים‬um ein ägyptisches Lehnwort im Hebräischen handeln, dass der Übersetzer als solches erkannt und korrekt ins Griechische übertragen hat. 60 Schließlich sei noch auf das Wort für »Korb« hingewiesen, in dem Moses ausgesetzt wurde und das identisch ist mit dem Wort für die Arche Noah. Das hebräische Wort ‫ תבה‬kann im Griechischen der LXX sowohl mit κιβωτός als auch mit θίβις/ θῖβις wiedergegeben werden (Ex 2,3). 61 Letzterer Begriff scheint eine Transliteration des hebräischen Wortes ‫ תבה‬zu sein. Da dieser Begriff im Griechisch der Papyri als Ägyptizismus durchaus auftaucht, 62 handelt es sich jedoch davon unabhängig ebenfalls um die Transliteration des ägyptischen Wortes für den »Kasten« db.t, ḏbꜢ.t. 63 Das hebräische tebah wiederum ist möglicherweise ein aus dem Ägyptischen übernommenes Lehnwort. 64 Der Übersetzer des Hebräischen hat nun also die gleiche griechische Transliteration für das hebräische Lehnwort verwendet, wie sie davon un57. Vgl. Vergote, Joseph en Égypte, 59-66; Muchiki, Egyptian Proper Names, 238. 58. Vgl. M. Tilly, Einführung die Septuaginta, Darmstadt 2005, 73. 59. J. Černý, Coptic Etymological Dictionary, Cambridge u. a., 1976, 17; J.-L. Fournet, »Les emprunts du grec à l’égyptien« BSL 84,1 (1989), 68 f. 60. Görg, »Septuaginta im Kontext«, 124; vgl. Wörterbuch der ägyptischen Sprache I, 551 f. 61. Vgl. M. Harl, »Le nom grec de l’›arche‹ de Noé«, in: La langue de Japhet. Quinze études sur la Septante et la grecque de chrétiens, Paris 1992, 100; Fournet, »Emprunts du grec«, 73. 62. Etwa P.Oxy. XXIV 2424,3.31 (2./3. Jh. n. Chr.); P.Petr. III 51,4 (3. Jh. v. Chr.); vgl. F. Preisigke, Wörterbuch der griechischen Papyrusurkunden I, Heidelberg 1924, 677, mit weiteren Belegen. 63. Y. Koenig, »Quelques ›égyptianismes‹ de la Septante« BIFAO 98 (1998), 226; Muchiki, Egyptian Proper Names, 258. 64. Problematisch ist allerdings dass die Wiedergabe des anlautenden Dentals irregulär ist. Zu erwarten wäre ein Teth.

242

2. Einflüsse der indigen-ägyptischen Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

abhängig in Ägypten für das ägyptische Urwort genutzt wurde. Es besteht also durchaus die Möglichkeit, dass das griechische Fremdwort thibis den gleichen Weg ins Griechische genommen hat wie das Wort achi. 65 Auch die Wendung οὐ ποιήσετε σισόην ἐκ τῆς κόμης τῆς κεφαλῆς ὑμῶν – »ihr sollte keine sisoe aus eurem Haupthaar machen« in Lev 19,27 steht im Zusammenhang mit einem ägyptischen Einfluss, denn das Wort σισόη ist im Griechischen unbekannt. 66 Es gibt aber einen ägyptischen Personennamen der im Griechisch dieser Zeit Σισόις ausgesprochen wurde. 67 Dieser Name leitet sich ab vom ägyptischen Eigennamen ṯꜢj-ḏj und bedeutet »Träger der Haarlocke (des Horuskindes)«. Das Kennzeichen eines ägyptischen Kindes war nämlich der geschorene Kopf, an dessen Seite nur eine lange Haarlocke herunterhing. Die griechische Aussprache σισόις dürfte natürlich nicht nur für den Personennamen Verwendung gefunden haben, sondern auch für das ägyptische Wort »Haarlocke« selbst, das im Demotischen als ṯꜢj-ḏj 68 und im Koptischen als ϫⲓϫⲱⲓ 69 vorkommt. So ist Lev 19,27 also mit »Ihr sollte keine Haarlocke (= keinen Zopf) aus eurem Haupthaar flechten« zu übersetzen. Ob dies auch für das Wort Pharao gilt, das sowohl hebräisch als auch griechisch eine direkte Übernahme des ägyptischen Herrschertitels Pr-ʿ Ꜣ »großes Haus« ist, muss freilich offen bleiben.

2.5 Inhaltliche Korrekturen Dort, wo der Übersetzer des hebräischen Textes wusste, wie das Wort in der zeitgenössischen Sprache richtig heißt, korrigierte er ihn in seiner Übersetzung. Das macht er etwa im Fall der Einbalsamierung des Jakob: »Dann befahl er (i. e. Joseph) den Ärzten (‫)הרפאים‬, die in seinem Dienst standen, Jakob einzubalsamieren« (Gen 50,2). Natürlich obliegt die Aufgabe der Balsamierung in Ägypten nicht Ärzten sondern dem Berufstand der »Einbalsamierer« (ἐνταφιασταί). 70 Das ist eine Berufsbezeichnung, die seit dem beginnenden 1. Jh. v. Chr. in den Serapeumspapyri belegt ist. 71

65. Ähnliches könnte für die von den Übersetzern genutzte ägyptische Getreidemaßangabe οἰφεί (etwa Lev 5,11; Hes 45,13 u. a. in verschiedenen Schreibungen) gelten, das keine genau Transkription des hebräischen ‫ איפה‬ist, sondern auf das ägyptische Wort jp.t (vgl. Wörterbuch der ägyptischen Sprache I, 67) zurückgeht; hierzu Fournet, »Emprunts du grec«, 71. 66. Vgl. schon Z. Frankel, Über den Einfluss der palästinischen Exegese auf die alexandrinische Hermeutik, Leipzig 1851, S. 151; P. Harlé/D. Pralon, La bible d'Alexandrie. Le Lévitique, Paris 1988, S. 171; zum Hebräischen vgl. Martin Vahrenhorst, Septuaginta Deutsch. Erläuterungen und Kommentare zum griechischen Alten Testament, Band I, Stuttgart 2011, 395. 67. Vgl. Lüddeckens u. a., Demotisches Namenbuch, S. 1354; H. de Meulenaere, »Le nom propre Σισόις et son prototype égyptien« in: CdÉ 66, 1991, S. 129-135. 68. Vgl. W. Erichsen, Demotisches Glossar, Kopenhagen 1954, S. 669. 69. Vgl. W. Westendorf, Koptisches Handwörterbuch, Heidelberg 1965-1977, S. 443. 70. Vgl. Strab. 11,11,3; hierzu Morenz, »Ägyptische Spuren«, 426; Görg, »Septuaginta im Kontext«, 120 f. 71. UPZ I 106,10; 107,12; 108,2 und 22; 109,2; II 190,6. 2. Einflüsse der indigen-ägyptischen Kultur und Sprache

243

16. Ägyptische Elemente im Griechischen der LXX

2.6 Adaptionen ägyptischer Religion? Eine auf den Einfluss der ägyptischen Religion zurückzuführende Aktualisierung führt Morenz an. 72 Es geht um die Übersetzung von Dtn 9,26, das im Griechischen lautet »Herr, Herr, König der Götter« (Κύριε κύριε βασιλεῦ τῶν θεῶν), das für das Hebräische »Herr, Jahwe« (‫ )אדני יהוה‬steht – zumindest die griechische Wendung »König der Götter« ist also eine interpretative Ergänzung. Zweifelsohne muss sich der Übersetzer bei seiner freien Wiedergabe, in der er zweimal den Gottesbegriff Adonai und Jahwe mit κύριος übersetzte 73 und dann noch, um die große Macht Jahwes auszudrücken, die Königsattribuierung hinzufügte, etwas gedacht haben. Trotz der Tatsache, dass Jahwe ähnlich auch an anderer Stelle bezeichnet wird, 74 handelt es sich hier schließlich, anders als beispielsweise in Ps 95, um ein Epitheton, das der Übersetzer nach freier Entscheidung ergänzte. Zwar galt den Griechen Zeus als König der Götter, 75 doch taucht diese Wendung als Epitheton des Zeus in der griechischen Überlieferung nicht auf. Es handelt sich vielmehr um eine genaue Übersetzung des ägyptischen Gottesepithetons nswt-nṯr.w. So geht Morenz davon aus, dass der Übersetzer ägyptische Gottesepitheta kannte und wusste, wie sie ins Griechische zu übersetzen sind. Er wollte seinen Gott Jahwe mit diesem höchsten ägyptischen Gottestitel über die paganen Götter seiner Umwelt stellen. Ob dies freilich gleichzeitig als Einfluss spätägyptischer Theologie zu werten ist, 76 bleibt Spekulation. So muss auch diese Theorie hinterfragt werden, weil die Wendung »König der Götter« häufig in den Qumran-Rollen vorkommt (vgl. 4Q381, 76-77:7; 4Q403 I 30-46:34). Der griechische Text der LXX könnte sich also auch auf eine entsprechende hebräische Vorlage zurückführen lassen. 77 Einen treffenderen Einfluss ägyptischer religiöser Vorstellungen schlug unlängst Jan Joosten vor: Es findet sich an verschiedenen Stellen der Septuaginta die Wendung, dass man Gott sieht, wo im hebräischen Text andere Formulierung stehen (Gen 31,13; 72. Morenz, »Ägyptische Spuren«, 420 f.; vgl. die Übernahme von Görg, »Die Septuaginta im Kontext«, 116 f. 73. M. Rösel, »Die Übersetzung der Gottesnamen in der Genesis-Septuaginta« in: D. R. Daniels u. a. (Hg.), Ernten, was man sät, Festschrift für Klaus Koch zu seinem 65. Geburtstag, Neukirchen-Vluyn 1991, 357-377; weshalb in der doppelten Setzung des κύριος eine »Aufnahme ägyptischer Vorstellungen« zu erkennen ist (ebd., 377), ist nicht nachzuvollziehen. Es sei noch auf Görg, »Septuaginta im Kontext«, 120-122, Spekulation über das »Tohuwabohu«, griechisch ἀόρατος καὶ ἀκατασκεύαστος hingewiesen. Er möchte letzteres mit dem Gottesnamen des Urgottes Amun »der Unsichtbare« in Verbindung bringen, das Tohuwabohu wiederum mit der Kosmologie der hermopolitanischen Urgötter und die beiden ägyptischen Verben »das Ziel verfehlen« (thj) und »kopflos fliehen« (bhA). 74. Vgl. die Wendung in Ps 94,3/95,3: ὅτι θεὸς μέγας κύριος καὶ βασιλεὺς μέγας ἐπὶ πάντας τοὺς θεούς/‫כי אל גדול יהוה ומלך גדול על־כל־אלהים‬. 75. Vgl. Plutarchus Biogr. et Phil. (Stephanus p. 282), Aetia Romana et Graeca: ἦ ὅτι τῶν μὲν ἀοράτων (C.) θεῶν καὶ νοητῶν βασιλεύουσι Ζεὺς καὶ Ἥρα, τῶν δ’ ὁρατῶν ἥλιος καὶ σελήνη. 76. Görg, »Septuaginta im Kontext«, 118. 77. Vgl. M. Philonenko, »La Bible des Septante«, in: J. Leclant / R. Vian des Rives, Colloque Alexandrie: mégapole cosmopolite. Actes. Préambule, Paris 1999 (Cahiers de la Villa »Kérylos« 9), 1999, 150 f.

244

2. Einflüsse der indigen-ägyptischen Kultur und Sprache

16. Ägyptische Elemente im Griechischen der LXX

33,13; Ex 25,8). Joosten ist der Ansicht, dass »it appears there is some external force leading the Greek translators to introduce the notion of ›seeing God‹ into certain contexts.« Diese »external force« wiederum sei auf einen ägyptischen Einfluss zurückzuführen, denn die Wendung »Gott sehen« (mꜢꜢ nṯr) ist fest im ägyptischen Ritualkontext verortet und findet eine Reflektion bei Manetho (Ios. c.Ap. I 232-233). 78 Als geläufige Ritualszene ist dieser Vorgang in den Tempelreliefs bis in römische Zeit gut belegt. 79

3. Ergebnis Die hier vorgestellten Beispiele ägyptischer Elemente im Griechisch der LXX können, allein schon ob des Umfanges des Beitrages, nur einen kleinen Ausschnitt der Möglichkeiten und vor allem der Fähigkeiten der verschiedenen Übersetzer des hebräischen Textes bieten. Es bleibt zunächst festzuhalten, dass sich, abgesehen von der Onomastik, auf sprachlicher Ebene kaum wirklich ägyptische Elemente beobachten lassen, während die kulturell-religiöse Ebene ein durchaus interessantes und breites Feld von Interaktion des jeweiligen Übersetzers mit seiner Umwelt zeigt. Daraus ergibt sich mit einiger Sicherheit, dass zumindest bestimmte Bücher oder Teile von Büchern der LXX nicht unter palästinischem Einfluss übersetzt wurden, sondern dass einige Übersetzer sehr spezifische und genaue Kenntnisse der ägyptischen Kultur und Religion hatten. Sie kannten sich allem Anschein nach auch recht gut mit dem Idiom des Landes selbst aus und verwendeten dieses sogar kreativ zu Aktualisierung von Inhalten.

78. J. Joosten, »To See God. Conflicting Exegetical Tendencies in the Septuagint« in: M. Karrer / W. Kraus (Hg.), Die Septuaginta – Texte, Kontexte, Lebenswelten. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 20.-23. Juli 2006, Tübingen 2008, 295-297, mit Literatur. 79. Vgl. H. Kockelmann, »Götter als Ritualisten. Zu einem speziellen Typ der Opfer- und Verehrungsszenen in ägyptischen Tempeln«, in: J. F. Quack (Hg.), Ägyptische Rituale der griechisch-römischen Zeit, Tübingen 2014, 29-45. 3. Ergebnis

245

17. Septuagint Greek and the Jewish Sociolect in Egypt Jan Joosten

1. Introduction As is confirmed by epigraphic finds, large groups of Jews in Egypt adopted Greek as their principal means of communication from the early Hellenistic period onwards. 1 Although little is known about the linguistic habits of the early Egyptian diaspora, it stands to reason, on the analogy of other religious groups, that Greek-speaking Jews would use forms of the language that were unusual among non-Jews. Jewish customs and institutions, religious ideas, and ritual formulas would be expressed in ways not immediately comprehensible to those who did not belong to the community. 2 It has often been suspected that the Greek of the Septuagint reflects this “Jewish sociolect” in some measure. 3 When the first translators were confronted with Hebrew words designating institutions or ideas of Judaism, they would naturally fall back on the usual terminology current among the group to which the version was addressed. 4 Once a translation of Hebrew Scriptures had come into being, it would in turn influence Jewish Greek phraseology. In prayers and religious exhortation, turns of phrase would be borrowed from the Greek translation. This latter phenomenon can actually be studied in Judaeo-Hellenistic writings from a slightly later period. 5 What is of interest presently, however, is an earlier process, in which the translation was the receptor and Jewish parlance the donor. That such a process took place is not easy to verify empirically. Since no texts going back to Egyptian Jews of the early Hellenistic age—other than the Septuagint itself—have been preserved, the extent to which that 1. 2. 3.

4.

5.

See V. A. Tcherikover, Corpus Papyrorum Judaicarum, vol. 1, Cambridge, MA 1957; W. Horbury / W. Noy / D. Noy, Jewish inscriptions of Graeco-Roman Egypt, Cambridge 1992. On religiously based sociolects, see, e. g., S. Bunin Benor, “Do American Jews Speak a ‘Jewish Language’ ? A Model of Jewish Linguistic Distinctiveness” JQR 99 (2009), 230-269. See the early statement of Wellhausen quoted below in note 6. See also I. L. Seeligmann, The Septuagint Version of Isaiah, Leiden 1948, 44: “Apart from the transcriptions […] the translator uses many word known only, or chiefly, from the idiom of his milieu […]”; M. Harl / G. Dorival / O. Munnich, La Bible grecque des Septante: Du judaïsme hellénistique au christianisme ancien, Paris 1988, 228 (with further literature). Tessa Rajak has argued against the notion that “cultic and spiritual keywords” would have been coined in Greek during the period preceding the translation of the Torah. In her idea, most of the new terminology found in the Septuagint was created by the translators. She seems to imply that before the Septuagint, Jews residing in Egypt would discuss religious matters in Hebrew or Aramaic. See T. Rajak, Translation and Survival. The Greek Bible of the Ancient Jewish Diaspora, Oxford 2009, 163. See G. Walser, The Greek of the Ancient Synagogue, Stockholm 2001. The process takes place within the Septuagint corpus itself (see Chapter 21 in the present work).

246

1. Introduction

17. Septuagint Greek and the Jewish Sociolect in Egypt

group developed distinct religious phraseology is hard to measure. Nevertheless, the postulate of an early Jewish sociolect having influenced the language of the Septuagint is not mere speculation. The Septuagint itself exhibits particularities that suggest that not all the renderings it contains were chosen freely by the translators. 6 Some of the Hebrew-Greek equivalences may, on various grounds, be regarded as having been established before the translation was begun. Thus the question as to the Jewish sociolect and its influence on the language of the Septuagint merits to be addressed. Influence of the Jewish sociolect in Hellenistic Egypt on the language of the Septuagint is easier to imagine if it is supposed that the version was made by Egyptian Jews. 7 This supposition is not a necessary one, however. Even if the translators were Jerusalemites recently arrived in Egypt, they might employ certain terms and expressions in use among the local community. Only if the version had nothing whatsoever to do with Egyptian Judaism would the question of the Jewish sociolect become groundless. Yet few Septuagint scholars today would go so far as to affirm this. Although the identity and provenance of the translators are subject to debate, there is a broad consensus that at least the Pentateuch translation was produced in Egypt.

2. Methodological Remarks Before the end of the nineteenth century, Septuagint Greek, together with New Testament Greek, was often regarded as a distinct variety of language, a type of dialect. 8 This idea was definitively dispelled by the investigations of Adolf Deissmann, who was able to show that many of the peculiarities of “biblical Greek,” unparalleled in literary writings, were in fact attested in documentary sources. The basic cast of Septuagint Greek —its morphology, elementary vocabulary, and syntax—is that of the common, nonliterary language of the Hellenistic period as it was practiced throughout the Greekspeaking world at the time the writings were created. Research done recently by John Lee on the Greek Pentateuch has essentially confirmed this view: “The vocabulary of the Pentateuch has many close links with the vocabulary of contemporary vernacular Greek.” 9 In the nineteen fifties, Henry Gehman attempted to revive the idea that some aspects of Septuagintal language reflect a “Jewish Greek” spoken in the diaspora. 10 His 6. This was already stated by Wellhausen. While explaining how Septuagint Greek might have been easier to understand for Jews of the Hellenistic period than for us, he noted in passing: “[…] zumal da ein Judengriechisch, zunächst für religiöse Dinge in der Synagoge mündlich ausgebildet, schon vorhanden gewesen sein muss, als die erste schriftliche Uebersetzung entstand.” See J. Wellhausen, Der Text der Bücher Samuelis, Göttingen 1871, 10. 7. For a recent review of the debate, see A. van der Kooij, “The Septuagint of the Pentateuch and Ptolemaic Rule” in: G. N. Knoppers / B. M. Levinson (eds.), The Pentateuch as Torah. New Models for Understanding Its Promulgation and Acceptance, Winona Lake, IN 2007, 289-300. Van der Kooij himself sides with those who think the translators came from Jerusalem. 8. See extensive bibliographical references in chapter 1 of the present work. 9. See J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SCS 14), Chico, CA 1983, 145. 10. See, e. g., H. S. Gehman, “The Hebraic Character of the Septuagint” VT 1 (1951), 81-90. 2. Methodological Remarks

247

17. Septuagint Greek and the Jewish Sociolect in Egypt

main argument was that Hebraisms in the Septuagint could only have been understood by a readership that used them in daily speech. Syntactical and lexical peculiarities induced by Hebrew, he reasoned, would be incomprehensible except to those who either knew the Hebrew source text or spoke a language influenced by Hebrew or Aramaic. His approach, however, found little support: Hebraisms in the Septuagint can be adequately explained by the translational process. As to the interpretation of the Septuagint, it appears that Gehman assumed incorrectly that the meaning attributed to the Greek translation by its readership would correspond to the meaning of the Hebrew source text. This is not necessarily the case: at times, literal translation of the Hebrew may lead to the expression of a different meaning; in other instances, it may produce obscure Greek. The case for the existence of a Jewish sociolect influencing the wording of the Septuagint should be based on linguistic usage—of words, forms, phrases—that cannot be satisfactorily explained either from Hellenistic Greek or from literal translation of the Hebrew source text. An expression contained in the Septuagint that has no parallel in Hellenistic Greek and that does not reflect the wording of the Hebrew source may indeed reflect the particular phraseology characteristic of Jewish speech in the diaspora. These methodological guidelines need to be hedged in two directions. First, our knowledge of Hellenistic Greek depends on the available documentation, which is far from exhaustive. Many words found in the Septuagint for the first time, or for the first time with a specific meaning, may nevertheless reflect the common language of its period. Lack of attestation may be due to accident. The mere absence of a usage in Hellenistic Greek cannot be taken as sufficient proof that one is dealing with the Jewish sociolect. 11 Second, the Hebrew source text underlying the Septuagint may have diverged from the Masoretic text and other attested Hebrew texts in many places. Semitisms without counterpart in the received Hebrew text cannot necessarily be attributed to the translators’ linguistic habits, since they may reflect an unattested Hebrew source text. 12 A more substantial criterion is that expressions belonging to the sociolect may a priori be expected to designate realities that would have special relevance in a Jewish milieu. When the linguistic and the substantial criteria go hand in hand, a good case can be made for recognizing items of the Jewish sociolect.

11. An excellent illustration of this principle is afforded by the word προσήλυτος. As long as this word was attested in biblical texts only, scholars could argue that it was a neologism created by the translators or in the Jewish sociolect, see, e. g., E. Tov, “Three Dimensions of LXX Words” Revue Biblique 83 (1976), 529-544, in particular 537-538 (with references to earlier literature). Now, however, the word has turned up in a third century B.C.E. papyrus without religious connotations, see C. J. Butera / D. M. Moffitt, “P.Duk. Inv. 727: A Dispute with ‘Proselytes’ In Egypt” Zeitschrift für Papyrologie und Epigraphik 177 (2011), 201-206. The authors note: “That the word can now be shown to be in use in Egypt within a period roughly contemporary with the initial translation of the Septuagint in Alexandria calls into question the position that προσήλυτος was originally a terminus technicus for a gentile convert to Judaism” (p. 202). 12. For examples, see the chapter on Hebraisms in the present volume.

248

2. Methodological Remarks

17. Septuagint Greek and the Jewish Sociolect in Egypt

3. Inventory Exhaustive reconstitution of the Jewish sociolect on the basis of the Septuagint will remain out of reach. Even the more modest project of tracing all elements of the sociolect that have influenced the wording of the Septuagint will be hard to achieve. What can be done, however, is to identify a number of elements that may, with varying probability, be attributed to the Jewish sociolect.

3.1 Aramaic Loanwords Obvious candidates for inclusion in our list are Aramaic loanwords attested only in the Septuagint or in writings dependent on it. 13 These include the words γ(ε)ιωρας, “proselyte,” (Ex 12:19 and Isa 14:1) and παταχρα, “idol,” (Isa 8:21 and 37:38) and the forms μαννα, “manna,” πασχα, “Pesach,” σαββατα, “Sabbath,” and σικερα, “strong drink,” (all four of them occurring several times in the Pentateuch). 14 These items are not representative of Hellenistic Greek. 15 Nor are they transcriptions of Hebrew words. They can hardly be explained otherwise than by a reference to pre-existing speech habits among Greek-speaking Jews. 16 A consideration of the meaning of these Aramaic words and forms confirms that they reflect the Jewish sociolect. Without exception they are terms whose relevance is determined by religion. Some of them, like manna, Pesach and Sabbath, had no equivalent at all in Greek, while others received a special connotation in the framework of Jewish beliefs: “idol,” “strong drink” (forbidden by Jewish law on some occasions), “convert to the Jewish faith.” As Peter Walters formulated it, they “point to an old and primitive stratum of the LXX, when the translators in an unreflecting and spontaneous way incorporated into their translation what may well have been modes of expression in the vernacular Greek of the Egyptian Jews.” 17

3.2 Aramaisms Alongside Aramaic loanwords, the Septuagint also attests a number of Greek expressions that seem to reflect Aramaic phrases or idioms. A good example is the expression 13. For the distinction between the Aramaic loanwords of Hellenistic Greek generally and Aramaic words attested only in the Septuagint and parabiblical literature, see P. Walters, The Text of the Septuagint. Its Corruptions and their Emendation, Cambridge 1973, 155-175. 14. See Walters, Text of the Septuagint, 169-171, 173-175. 15. The word σαββατα is attested once in an early papyrus (CPJ I, 10,6), clearly reflecting Jewish usage: on the seventh day, there are no bricks because of the Sabbath. The form of the word is the same as in the Septuagint, see A. Pelletier “Σαββατα: Transcription grecque de l’araméen” VT 22 (1972), 436-447, in particular 438. 16. A mere reference to the multilingual background of the translator cannot be used to justify the presence of these words in the Septuagint (pace Rajak, Translation and Survival, 145-152, 163). It is hardly credible that the translators would have used Aramaic words to render Hebrew ones into Greek, unless indeed the Aramaic words in question had already come to be adopted in at least some forms of Greek. 17. Walters, Text of the Septuagint, 170. He adds in parentheses: “here I do not press ‘Egyptian’”. 3. Inventory

249

17. Septuagint Greek and the Jewish Sociolect in Egypt

“to sin before [God].” This turn of phrase—ἁμαρτάνω ἐνώπιον/ἐναντίον/ἔναντι in Greek—is used over thirty times in the Septuagint, but it is unattested in extra-biblical Greek. 18 It corresponds word for word to the Aramaic ‫חטא קדם‬, with the typically Aramaic reverential use of ‫קדם‬, “before,” in statements in which God (or the King) are the object of human actions. Its use in the Septuagint is not induced by the source text: in Biblical Hebrew, this expression is not found. 19

Josh 7:20

‫אְָמ ָנה אָנִֹכי ָחָטאִתי ַליה ָוה ֱאל ֵֹהי ִיְשׂ ָרֵאל‬

ἀληθῶς ἥμαρτον ἐναντίον κυρίου θεοῦ Ισραηλ “To sin before” should be considered an Aramaism in Septuagint Greek. Since it cannot be explained from Hellenistic Greek or from the Hebrew source text, it may reflect a formula used by Jews in religious discourse. The semantics of the expression fit in nicely with this idea. Unfortunately, the identification of Aramaisms in the Greek of the Septuagint is hazardous and can rarely be done in a manner that is entirely satisfactory. The main obstacle is the linguistic proximity of Aramaic and Hebrew: it is very hard to distinguish Aramaic from Hebrew influence in a Greek text. Nevertheless, a few other cases of this phenomenon have been proposed. 20 If some of the most typical elements of the Jewish sociolect are of Aramaic origin, this shows that the group among which the Septuagint came into being was rooted in an Aramaic-speaking milieu. 21 Historically, there are good reasons to see a degree of continuity between the Egyptian diaspora of the early Hellenistic period and earlier groups of Jews settled in Egypt, such as those who left us the Elephantine archives. 22

18. See the extensive discussion in J. Joosten, “L’Agir humain devant Dieu. Remarques sur une tournure remarquable de la Septante” Revue Biblique 113 (2006), 5-17. This article documents use of the expression in writings dependent on the Septuagint. 19. For rare attestations of similar turns of phrase in Hebrew texts, see J. Joosten, “L’araméen de Qumran entre l’araméen d’empire et les targumim: l’emploi de la préposition ‘devant’ pour exprimer le respect dû au roi et à Dieu” in: K. Berthelot / D. Stökl Ben Ezra (eds.), Aramaica Qumranica. Proceedings of the Conference on the Aramaic Texts from Qumran in Aixen-Provence 30 June – 2 July 2008, Leiden 2010, 83-97, in particular 87-89. 20. See T. Muraoka, “Gleanings of a Septuagint Lexicographer” BIOSCS 38 (2005), 101-108, in particular 106; J. Joosten, “‘À Dieu ne plaise’ (Matthieu 16,22). La provenance et l’arrière-plan de l’expression híleôs soi” in: J. Joosten / P. Tomson (eds.), Voces Biblicae. Septuagint Greek and its Significance for the New Testament, Leuven 2007, 155-167. 21. Pelletier, “Σαββατα,” 442: “L’origine araméenne du nom grec du Sabbat s’explique beaucoup mieux à partir des premiers contacts entre Grecs et communautés juives parlant araméen. C’est, en effet, aux vocables spontanément créés à l’occasion de ces contacts que le Judaïsme hellénisé – au premier rang, les traducteurs du Pentateuque – dut, entre autres, la nomenclature grecque de son calendrier.” 22. See J. Joosten, “The Aramaic background of the Seventy. Language, culture, and history” BIOSCS 43 (2010), 53-72.

250

3. Inventory

17. Septuagint Greek and the Jewish Sociolect in Egypt

3.3 Hebrew formulas While the occasional transcription of Hebrew words, like χερουβιμ, “cherubs,” in Gen 3:24, is to be ascribed to influence of the source text, the use of Hebrew liturgical formulas in Greek is perhaps better explained as reflecting existing speech habits among Greek-speaking Jews. 23 Formulaic αμην and αλληλουια are transcriptions of their Hebrew counterparts, but one needs to ask why they were transcribed. The meaning of the Hebrew equivalents of these formulas was well known to the translators: both ‫ אמן‬and the combination of ‫ יה‬+ ‫ הלל‬are translated adequately in several passages. 24 If nevertheless the translators opted to transcribe them in other passages, this tends to indicate that the Hebrew expressions involved had become acceptable in the Greek practiced by their community. Αμην and αλληλουια are attested only in the later books of the Septuagint. The former in Chronicles, 1Esdras and Nehemiah; the latter in Psalms. In the case of αλληλουια this distribution simply reflects the fact that the Hebrew equivalent is not attested elsewhere in the Bible. The Hebrew ‫אמן‬, “so be it,” however, is attested many times from the Pentateuch onward. The fact that it is transcribed only in later books may indicate that αμην was adopted into the Jewish sociolect at a later date.

3.4 Jewish institutions and concepts It stands to reason that salient institutions of the Jewish faith had received Greek names before the Septuagint was created. Thus there are strong reasons to believe that a verb such as περιτέμνω, “to circumcise,” was in use among Greek-speaking Jews before Genesis 17 was translated into Greek. This is not sufficient however to make it a feature of the Jewish sociolect. The verb περιτέμνω was well known as a technical term designating circumcision in non-Jewish Greek. The Jewish usage would have been perfectly understandable to non-Jews. The case is different for Greek terms designating Jewish institutions or ideas whose form or meaning diverges in some way from normal Greek usage. An example of formal divergence is the noun ἀκροβυστία, “foreskin, uncircumcision.” Although it looks like a Greek word, it is unattested outside the biblical corpus, and its second element has no obvious Greek etymology. 25 Unless the lack of extra-biblical attestation should be due to accident, the word fulfils our criteria: it is not explicable from Hellenistic Greek nor is it the result of literal translation of the Hebrew. Although the origin of the word is obscure, its use almost certainly reflects the Jewish sociolect. An interesting divergence in meaning can be observed in the noun εἴδωλον, “idol.” In non-biblical Greek, the word refers to a visible form or likeness. Although 23. See J. Smith, Translated Hallelujahs. A Linguistic and Exegetical Commentary on Select Septuagint Psalms (CBET 56), Leuven 2011, 37-50. 24. For ‫אמן‬, the standard translation is γένοιτο (see, e. g., Num 5:22); for ‫ יה‬+ ‫הלל‬, see Ps 150:6: πᾶσα πνοὴ αἰνεσάτω τὸν κύριον αλληλουια (see also Ps 101/102:19). 25. Some scholars see it as an intentional transformation of ἀκροποσθία, “tip of the foreskin,” (from ποσθία, “foreskin”), see K. L. Schmidt, TWNT, vol. I, 226-227; BDAG. The link to this rare word is not obvious, however. The second element has been connected with Hebrew ‫בשת‬, “shame,” (see, e. g., LSJ), but this connection, too, remains speculative. 3. Inventory

251

17. Septuagint Greek and the Jewish Sociolect in Egypt

it is used a few times in reference to divine statues, it is not really a religious term at all. In the Septuagint, however, the term is charged with theological meaning: εἴδωλα are false gods. The word is never used for other types of images or forms. It corresponds to a great variety of Hebrew equivalents, thus reordering the polemic against divine images around the notion of non-reality. Although this Septuagintal usage might have been understood by non Jews, it was clearly distinctive and would probably have been identified by contemporary readers as typically Jewish. The rendering of many different Hebrew nouns by the single Greek term εἴδωλον cannot be explained from Hellenistic Greek or from the Hebrew source text. It therefore probably reflects pre-existing speech habits among Greek-speaking Jews. Another usage that may belong in this category is that of the noun ἔθνος, “nation,” and its plural, ἔθνη, in reference to non-Jews. Because the noun is also used in its normal sense in the Septuagint, and often enough refers to the Jewish nation, one cannot be absolutely certain that the meaning “heathen, gentiles” is really attested in the earlier parts of the Septuagint. The specialization in meaning might be a later development that originated in the translation itself. Occurrences in the Pentateuch and in Isaiah make it likely, however, that the special meaning was known to the first translators, among them: Ex 33:16 καὶ ἐνδοξασθήσομαι ἐγώ τε καὶ ὁ λαός σου παρὰ πάντα τὰ ἔθνη, ὅσα ἐπὶ τῆς γῆς ἐστιν “And we shall be glorified, both I and your people, above all the nations that are on the earth.” 26

If these attestations are cogent, the conclusion should be that the usage reflects the speech habits of Greek-speaking Jews even before the translation of the Hebrew scriptures was made. 27

3.5 The special vocabulary of the Septuagint A well-known characteristic of the Septuagint is that many Hebrew words are rendered by stereotyped equivalents. The choice of some of these standard equivalents is surprising. Why, for instance, was διαθήκη, meaning “will, testament,” chosen to render the Hebrew ‫ ?ברית‬Similar instances include ἀγαπάω for ‫אהב‬, “to love”; δόξα for ‫כבוד‬, “glory”; κτίζω for ‫ברא‬, “to create”; νόμος for ‫תורה‬. A possible answer is that the association of the Hebrew meaning with the Greek word in question predates the Septuagint. 28 What remains to be explained, in this case, is why the word was selected to express its special meaning by Greek-speaking Jews before they translated the Hebrew scriptures. An interesting example is the verb εὐλογέω and its cognates. In the Septuagint this verb is used in two ways: in many passages it expresses the meaning it has in non-biblical Greek, “to speak well of someone or something, to praise”; in other pas26. In the Hebrew text, the formal equivalent of both λαός and ἔθνη is ‫עם‬. See also Ex 19:5; Isa 10:6. 27. An analogical use exists in non-biblical Greek: the noun ἔθνος is used to refer to foreigners (see LSJ and BDAG, sub verbo). The specific usage in the meaning “non-Jews” is, of course, Jewish. 28. See, e. g., the introduction to NETS, xvii.

252

3. Inventory

17. Septuagint Greek and the Jewish Sociolect in Egypt

sages, however, it appears to take on the signification of its standard Hebrew equivalent, ‫ברך‬, “to bless”; this latter meaning dominates in the Book of Genesis and the rest of the Pentateuch: God blesses the great sea monsters, the human being, the Sabbath, and his elect; the patriarchs bless their offspring; the priests bless the people of Israel. In regard to these usages, the choice of the equivalent εὐλογέω is hard to explain. Could it be that the choice of εὐλογέω was not made by the translators of the Pentateuch, but goes back to an earlier period? This is indeed possible if we assume that the equivalence of εὐλογέω and ‫ ברך‬was established in the language of prayer. In Jewish prayers, the verb ‫ ברך‬is frequent with the divine name as direct object. This usage is aptly rendered with εὐλογέω: “to bless” God is indeed to praise him. Possibly, then, the association of the two verbs occurred before the creation of the Septuagint, when Hebrew and Aramaic prayers began to be translated into Greek. When later the translator of Genesis hit upon the first occurrence of the Hebrew verb in Gen 1:22, he already knew how to translate it. The fact that the Greek meaning did not fit the context in Genesis was not felt to be important. 29 There may be other examples of this kind. Every instance should be treated on its own merits, however. 30 The equivalences between διαθήκη and ‫ ברית‬and between κτίζω and ‫ ברא‬are perhaps more likely to have been established by the translators. 31 * The harvest of possible features belonging to the Jewish sociolect used among Greekspeaking Jews at the dawn of the Hellenistic age is very meagre. Our inventory contains no more than thirteen expressions, some of them uncertain. The actual sociolect of Greek-speaking Jews in Ptolemaic Egypt would have been much richer. Future research may add a few additional examples. Further epigraphic finds, of liturgical or homiletic texts, could change the picture completely. Even this meagre collection suffices, however, to demonstrate the importance of the phenomenon. Although the Greek translation of the Hebrew scriptures was a novum, it was not an absolute beginning. It continued earlier, probably oral, attempts to “say Jewish things in Greek” (Emmanuel Lévinas).

29. For a more detailed analysis see J. Joosten, “Le vocabulaire de la Septante et la question du sociolecte des juifs alexandrins: Le cas du verbe ευλογεω, ‘bénir’” in: E. Bons / J. Joosten (eds.), Septuagint Vocabulary. Pre-History, Usage, Reception (Septuagint and Cognate Studies 58), Atlanta, GA 2011, 13-23. 30. The blanket statement in the introduction to NETS (“Calques may be expected to predate the Septuagint,” xvii) cannot be accepted. 31. For the former, see A. Schenker, “ΔΙΑΘΗΚΗ pour ‫ברית‬. L’option de traduction de la LXX à la double lumière du droit successoral de l’Égypte ptolémaïque et du livre de la Genèse” in: J.-M. Auwers / A. Wénin (eds.), Lectures et relectures de la Bible (Festschrift P.-M. Bogaert), Leuven 1999, 125-131; A. Schenker, “Le contrat successoral en droit gréco-égyptien et la διαθήκη dans la Septante” ZABR 6 (2000), 175-185. For the latter, see E. Bons and A. Passoni dell’Acqua, “A Sample Article: κτίζω – κτίσις – κτίσμα – κτίστης” in: Bons / Joosten, Septuagint Vocabulary, 173-187. 3. Inventory

253

17. Septuagint Greek and the Jewish Sociolect in Egypt

4. Functional Aspects of the Jewish Sociolect Sociolinguistic investigation has led scholars to identify two aspects of sociolects, corresponding to two distinct functions. 32 On the one hand, sociolects manifest similarities to professional languages. They provide a group with terms and distinctions relevant only to them. This is a cognitive function, with the denotational aspect of meaning in the foreground. On the other hand, sociolects have a sectarian facet. They provide members with a strong sense of collective identity, while demarcating the community from other groups. In this respect, connotations are often as important as denotations. This identitary function relies as much on connotations as on denotations. An example may serve to illustrate the twofold function of sociolects. In modern French (and no doubt in other European languages), the Arabic loanword halal is used among Muslims to designate foodstuffs they are permitted to eat. The cognitive aspects of the term are important: it is crucial to know which foods are halal and which are not. At the same time, the use of the word helps to identify members of a community: a butcher who writes halal across the window of his shop will be perceived by Muslims as “one of us” (and by non-Muslims as “one of them”). Taken as a whole, the linguistic elements attributed to the sociolect of Egyptian Jews in the early Hellenistic age clearly assume this double function. They provide essential terminology necessary to communicate efficiently within the parameters of Jewish religion. At the same time, they define a mode of speech proper for distinguishing Jews from non-Jews. Taken individually, some of the expressions identified above align naturally with one or other of these two aspects. – The names of the Sabbath service and the Pesach festival have an obvious cognitive role. Hellenistic Greek simply did not possess words for designating these institutions. The identitary function is not absent in these words, since only Jews would keep the Sabbath and celebrate Pesach, but the main function would be the practical one of designating certain days of the week and the year. – Similarly, the noun γ(ε)ιωρας, “proselyte,” mainly fulfils a denotational function, designating precisely and unambiguously a person converted to Judaism. – The verb εὐλογέω in the sense “to bless” also plays mainly a cognitive role. The Semitic idea of blessing could not be precisely expressed in Hellenistic Greek. By taking a good Greek word and bending it to fit the Hebrew meaning, the Jewish community filled in this blank. – Contrariwise, the function of the noun εἴδωλον may be said to be situated mainly on the identitary level. The reality referred to could easily be designated by other words, such as ἄγαλμα, “statue,” or εἰκών, “image.” The essential quality of the noun εἴδωλον as it is used in the Septuagint is its pejorative connotation. What non-Jews would refer to as divine images or statues, or more simply as gods, Jews would call “idols.” By using the word εἴδωλον in this way, a Greek-speaking Jew would indicate 32. See, e. g., D. Möhn, “Fachsprachen und Gruppensprachen” in: L. Hoffmann et al. (eds.), Fachsprachen. Ein internationales Handbuch zur Fachsprachenforschung und Terminologiewissenschaft (Handbücher zur Sprach- und Kommunikationswissenschaft 14.1), Berlin/New York, NY 1998, 168-181.

254

4. Functional Aspects of the Jewish Sociolect

17. Septuagint Greek and the Jewish Sociolect in Egypt

that he subscribes to the values and the belief system of Judaism. The semantics of the Aramaic loanword παταχρα, “idol,” may have been similar to those of εἴδωλον, but the scarcity of attestations makes this somewhat uncertain. – Similarly, for the noun ἀκροβυστία, “foreskin,” the identitary function appears to predominate. Hellenistic Greek had other terms to refer to the designated reality (notably ποσθία) and the word appears to be derogatory. Not possessing ἀκροβυστία would be an important distinction for Greek-speaking Jews to make. – The parade example of the second aspect of sociolects would be the word ἔθνη, “non-Jews,” if the usage could be identified with certainty in the Greek Pentateuch and Isaiah. – The formulaic expressions “to sin before” and the liturgical phrases αμην and αλληλουια would simply identify their users as belonging to a community where these modes of speech were usual. Other elements mentioned in this article do not strongly participate of either of the two functions of sociolects. The words μαννα, “manna,” and σικερα, “strong drink,” seem to refer to biblical realia without relevance to the time of the translators and their community. Possibly, the word σικερα had religious implications for Hellenistic Jews who had taken a vow, but this is impossible to ascertain.

5. Conclusions The scarcity of the evidence and the lack of external confirmation invite caution in drawing conclusions from the material gathered in the present study. Nevertheless, it would be unwise to disregard the evidence. The Septuagint undeniably contains a number of linguistic elements that cannot be explained either from normal Hellenistic Greek usage or from the influence of the Hebrew source text. Where these elements refer to typical features of Jewish religion it is reasonable to view them as manifestations of existing speech habits among Greek-speaking Jews at the time the Septuagint was produced. The expressions discussed here show—much more clearly than the Hebraisms of the Septuagint—that this translation was made by Jews, for Jews. This conclusion is particularly valid for the Pentateuch. In recent times, several specialists have, in line with the Letter of Aristeas, renewed the argument that the intended readership of the Septuagint Pentateuch may have consisted at least partly of non-Jews. 33 But the use of items reflecting the Jewish sociolect tends to contradict belies this idea. Non-Jews would have experienced serious problems understanding words like πασχα and σικερα; they would have been clueless as to what is meant by the statement that God “spoke well” of the great sea monsters; moreover, they would not have appreciated the fact that the version referred to them as “nations,” to their gods as “idols,” and to their foreskin with a funny and unusual word. The occasional use of sociolectic expressions did not make the Septuagint incomprehensible to pagan readers: it was written in good Hellenistic Greek (although of a non-literary register), not in some kind of eso-

33. See, e. g., van der Kooij, “Septuagint of the Pentateuch,” 289-300. 5. Conclusions

255

17. Septuagint Greek and the Jewish Sociolect in Egypt

teric code. Nevertheless, non-Jewish readers would inevitably note that the text they were reading was not really addressed to them. 34 The Septuagint was born from the encounter between Judaism and Hellenistic culture. The version was not, however, the very first manifestation of this encounter. The Jews who decided to translate their scriptures into Greek had already been engaged in a process of acculturation for roughly a generation. This process implied incorporating Hellenistic ideas and concepts into their own world view. But it also implied, inversely, the importation of Jewish notions and beliefs into the Greek world and language. The Septuagint is the oldest testimony of this process preserved until today. But the Septuagint itself gives witness to the fact that earlier attempts had been made to “translate” Jewish ideas and conceptions into Greek.

34. In contrast, the Antiquities of Flavius Josephus attest, explicitly and implicitly, that their intended readership included non-Jews.

256

5. Conclusions

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek? Carlo Consani

The possibility that Palestinian linguistic elements, Hebrew or Aramaic, are present in the Greek of the Septuagint is a particularly complex issue that raises numerous problems. In order to answer this question, it is necessary to consider external data, historical and cultural, pertaining to the formation and the growth of the Old Testament corpus in Greek and to closely define the object of investigation based on the large number of studies conducted on the language of the Septuagint. Indeed, the expression “the language of the Septuagint” is only a label, since the language of the Septuagint cannot be considered a proper natural language; rather, it is a translation language or a literary sub-language used for a particular corpus, the books of the Old Testament, that have great religious and theological value and that constitute a source of identity for the Jewish community. 1 The main issue here concerns the history of the Greek language in post-classical times. It is, therefore, necessary to ascertain the plausibility of geographic variations in the Hellenistic koine. In order to avoid tautological reasoning, the case for the existence of a koine containing Palestinian elements can only be made by identifying distinctive linguistic traits in sources other than the Greek Septuagint.

1. Formation and Provenance of the Old Testament Corpus in Greek Modern research maintains that the most ancient core of writings, those that comprise the Torah, or Pentateuch, were translated at the end of the third century BCE by the Jewish community in Alexandria. Later, in addition to books of Egyptian origin and others connected with the Jewish diaspora, a limited number of books, including Song of Songs, Lamentations, Ruth, Esther, and Ecclesiastes (Qohelet), were translated into Greek or composed directly in that language by Palestinian Jews. 2 The question of the presence of Palestinian elements, or better still elements reshaped to imitate the original Jewish text, has been dealt with in depth by Dominique Barthélemy, who recognized certain translation practices as typical of the so-called kaige group of translators 1. 2.

See J. M. Dines, The Septuagint (Understanding Bible and its World 7), London/New York, NY 2004, 1-3. For an overview, see, e. g., G. Dorival / M. Harl / O. Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien, Paris 1988, 101-108; N. Fernández Marcos, Septuaginta. La Bibbia di Ebrei e Cristiani (Antico e Nuovo Testamento 6), Brescia 2010, 23-32; Dines, The Septuagint, chapter 3. The proposed Palestinian origin of certain books is based either on external factors or textual elements, such as the Greek translator’s footnote in the final verse of Esther. 1. Formation and Provenance of the Old Testament Corpus in Greek

257

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

active in Jerusalem in the first decades of the Common Era. 3 Today, as a result of the debate sparked by his thesis, there is general consensus that the following features belong to the “Palestinian” linguistic facies: (a) the use of the καίγε connective to translate the particle gam in the Hebrew text, (b) the use of the ἐγώ εἰμι periphrasis followed by a finite verb to translate the Hebrew ’anoki, (c) the generalized use of ἀνήρ to translate all the uses of the Hebrew ’iš, (d) the translation of the Hebrew ’en by οὐκ ἔστι, without observing agreement of tenses. 4 As this summary shows, the identification of possible Palestinian elements requires particular care for two reasons. First, the aforementioned linguistic features of possible Palestinian origin belong to the lexicon and the syntax, linguistic levels where the boundary between the author’s conscious stylistic choices and his native language competence is notoriously hard to draw. Second, these features demonstrate that the translator’s choices were doubly determined by the translation technique he used and by his intent to be faithful to the original Hebrew text. A relevant example can be found in the debate which has arisen over H. J. Venetz’s hypothesis, in line with Barthélemy’s work, that the Psalms were translated in a Palestinian environment. His conclusion is based on lexical and syntactical characteristics. An example of the former is the use of the terms βᾶρις and πυργόβαρις. An example of the latter is the translation of the particle gam by καὶ γάρ, which can be understood as a syntactical anticipatory phenomenon of καίγε. Venetz’s hypothesis has been contested on the grounds that καὶ γάρ is not the only solution provided by the Psalm translator for gam: the particle is translated twelve times using καί and thirteen times using καὶ γάρ. Furthermore, in other texts, such as 2Chronicles and Ezekiel 28–39, whose Palestinian origin is far from certain, the particle gam is translated by the expression καὶ γάρ. In addition, the term βᾶρις does not appear to be characteristic of the Palestine area, either in form or meaning. 5 From a linguistic point of view, the nature and importance of supposed Palestinian linguistic features depends on whether they derive from Ur- and Old Greek texts or were introduced in later versions. In the case of Ur- and Old Greek texts, it is possible that Semitic linguistic elements could have originated in the environment in which the texts were translated; these elements must therefore be analyzed in relation to their cultural context, the translator’s linguistic competence, and his linguistic repertoire. In the second case, the inclusion of Semitic elements does not necessarily reflect the linguistic competence of the translator or the environment in which he worked, but rather the cultural and religious inclinations of a later editor or copyist. In order to verify, independently of textual analysis or tradition, whether certain books of the Septuagint were written either in Palestine or by authors whose mother tongue was Aramaic or Hebrew, a scholar of Ancient Greek must begin by identifying 3. 4. 5.

D. Barthélemy, Les devanciers d’Aquila (SVT 10), Leiden 1963. N. Fernández Marcos, The Septuagint in Context. Introduction to the Greek version of the Bible, Leiden/Boston, MA/Köln 2000, 147-148. H. J. Venetz, Die Quinta des Psalteriums. Ein Beitrag zur Septuaginta- und Hexaplaforschung, Hildesheim 1974; Dorival / Harl / Munnich, Bible grecque, 104; A. Pietersma, “Septuagint Research: A Plea for a Return to the Basic Issues” VT 35 (1985), 307-311. On the meaning of βᾶρις, see also E. Will, “Qu’est-ce qu’une Baris?” Syria 64 (1987), 253-259.

258

1. Formation and Provenance of the Old Testament Corpus in Greek

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

possible features of Palestinian Greek in texts outside of the Greek Old Testament canon.

2. Origin, Nature, and Possible Varieties of Hellenistic Koine The origin and nature of koine have been much debated since the start of the last century. Schematically two types of views can be identified. Some scholars hold that koine resulted from the intermingling of different ancient dialects. Others maintain that it is a development of the Attic dialect. 6 Although the latter view gained dominance, the question remains complex, particularly concerning Jewish Greek. 7 Recent studies tend to show that the Hellenistic koine is the result of the standardization of a variety of Attic, such as the one used in government and by authors such as Thucydides, Xenophon, and Plato. According to this view, this standardized form became the spoken language of literate fifth- and fourth-century BCE Athenians. 8 This language, subsequently chosen by the Macedonian monarchy as the official language, was diffused throughout the empire and had profound repercussions for the populations of the ancient Greek poleis, and beyond. In Greece, for the first time, the emergence of a “common” language caused a decline in the use of local dialects, creating a situation of diglossia, which was to remain a permanent feature of the Greek language till modern times. 9 Throughout the rest of the empire, since the times of Alexander the Great, and increasingly so with his successors, the koine was spoken by a growing number of people who learned it as a second or a foreign language, according to the level of Hellenization of their respective territories. In Greece, as in peripheral areas, local configurations of linguistic repertoire favored the formation of distinct koines (koinai) reflecting the influence of native dialects or languages. It is also possible to identify forms of the Greek koine in territories beyond Greece. Some of these were studied more than a century ago by scholars such as Karl Dieterich and Albert Thumb. To date, three forms, whose existence was confirmed from a theoretical standpoint 10 as well as by documentary evidence, have been identified: the 6. See A. Lopez-Eire, “De l’attique à la koiné” in: C. Brixhe (ed.), La Koiné grecque antique, vol. I, Une langue introuvable?, Nancy/Paris 1993, 41-43. 7. See, e. g., G. H. R. Horsley, New Documents Illustrating Early Christianity, vol. 5, Linguistic Essays, Marrickville 1989, 30-31. 8. For a view on the origins of koine, see, e. g., A. Lopez-Eire, “Fundamentos sociolingüisticos del origen de la koiné” in: A. Lopez-Eire, Estudios de lingüistica, dialectología e historia de la lengua griega, Salamanca 1986, 401-431; Lopez-Eire, “De l’attique à la koiné,” 41-57; C. Consani, “Continuità e discontinuità nel greco post-classico” in: L. Melazzo (ed.), Continuità e discontinuità nella storia del greco. Atti del Convegno della Società Italiana di Glottologia (Palermo 24-26 ottobre 1994), Pisa/Vienna 1998, 95-113. 9. For an overview of this period see V. Bubeník, Hellenistic and Roman Greece as a Sociolinguistic Area, (Current Issues in Linguistic Theory 57), Amsterdam 1989. On the dominance of the koine over the dialects, see C. Brixhe (ed.), La Koiné grecque antique, vol. II, La concurrence, Paris/Nancy 1996. 10. For analogies between external varieties of the koine and regional varieties of standard modern 2. Origin, Nature, and Possible Varieties of Hellenistic Koine

259

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

Egyptian koine, the Anatolic koine, and a Greek variety, often defined as “eastern,” which was used in areas characterized by substratum languages belonging to the Semitic family. A discussion of the Egyptian and Anatolic koines lies beyond the scope of this study, 11 but the third type of koine is relevant because it provides a frame of reference for a hypothetical Palestinian koine.

3. Eastern Koine: Formation and Characteristics Leaving aside the Greek of Egyptian documents, under the label eastern koine usually one includes: (a) Anatolic Greek, and (b) Syro-Palestinian Greek.

3.1 Plurilingualism in the Syro-Palestinian Area In order to verify the presence of an eastern koine in Syria-Palestine and surrounding regions in Hellenistic and Roman times, it is necessary to analyze the linguistic situation in those areas. Fortunately, beginning at the end of the 1970s, a series of studies provided us with an adequate reconstruction of the languages that were spoken there. 12 Since the end of the fourth century BCE, the territory between Syria and Egypt was fought over by the Ptolemy and the Seleucid Kingdoms; three successive wars redefined the borders between the two kingdoms, causing the progressive Hellenization of the surrounding territories. Beginning in the third century BCE, Ptolemaic Egypt became increasingly influenced by the ancient Hellenic culture. It is widely maintained that the substantial Jewish community residing in Alexandria abandoned their original ethnic languages at an early date and adopted a thoroughly Greek linguistic repertoire. 13 languages see C. Consani, “La koiné et les dialectes grecs dans la documentation linguistique et la réflexion métalinguistique des premiers siècles de notre ère” in: Brixhe, Koiné grecque antique, vol. I, 23-37, and Consani, “Continuità e discontinuità,” 99-102. 11. For the first type of koine, see Stefan Pfeiffer in this volume. For the second type, see C. Brixhe, Essai sur le grec anatolien au début de notre ère (second edition), Nancy 1987; Bubeník, Hellenistic and Roman Greece, 237-252. 12. For linguistic varieties present in Palestine in Hellenistic and Roman times see: C. Rabin, “Hebrew and Aramaic in the First Century” in: S. Safrai / M. Stern (eds.), The Jewish People in the First Century, vol. II, Assen 1976, 1007-1039; G. Mussies, “Greek in Palestine and the Diaspora” in: Safrai / Stern, Jewish People, vol. II, 1040-1064; H. B. Rosén, “Die Sprachsituation in römischen Palästina” in: G. Neumann / J. Untermann (eds.), Die Sprachen in römischen Reich des Kaiserzeit, Köln 1980, 215-239; J. Barr, “Aramaic and Greek in Hellenistic Age” in: W. D. Davies / L. Finkelstein, The Cambridge History of Judaism, vol. II, The Hellenistic Age, Cambridge 1989, 79-114. For the Hellenization of Palestine from a broadly cultural point of view, the standard reference is still M. Hengel, Judaism and Hellenism, London 1974; for current debate about the view held by this scholar see J. J. Collins / G. E. Sterling (eds.), Hellenism in the Land of Israel (Christianity and Judaism in Antiquity Series 13), South Bend, IN 2001. 13. See G. J. Sterling, “Judaism between Jerusalem and Alexandria” in: Collins and Sterling, Hellenism, 265-301.

260

3. Eastern Koine: Formation and Characteristics

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

The cultural and linguistic environment from the Phoenician coast to Palestine and the immediate inland areas was more varied, although the use of Greek loanwords in Hebrew, local onomastic, and, above all, Roman and Hellenistic inscriptions indicate that a progressive cultural and linguistic Hellenization took place throughout the area. These elements give us a greater understanding of how the Greek language was used by different strata of populations. 14 As the cited studies show, the linguistic repertoire of the Palestinian area in Roman times was characterized by the simultaneous presence of Hebrew, Aramaic, and Greek. Late Biblical Hebrew continued to be the written language in works such as 2Chronicles, Ecclesiastes, and Esther. Along with this written code, other varieties of Middle Hebrew were used in the Mishnaic Hebrew of the first and second centuries CE. In spite of the fact that little documentation exists from more ancient times, it is probable that various Aramaic dialects of the Syrian type and others characteristic of eastern Palestinian-Samaritanian areas were spoken during the Hellenistic and Roman period. Although documented evidence of Aramaic is somewhat limited in biblical writings, this language would have been spoken by most of the population, particularly in rural areas. 15 In this linguistic environment, Greek would have represented a prestige language for various reasons: it was the primary means of international communication, the language of government, and the idiom of the prestigious Hellenic culture. This explains why Greek learning was sought for, first by the aristocracy and then, increasingly, by other strata of the population. Hebrew was gradually relegated to being the language of religion, losing its status as a spoken language. 16 Of course, the association of Greek with the Palestinian elite and its correlate, the identification of Aramaic and Middle Hebrew dialects with lower classes, greatly oversimplifies a complex situation; a number of sociolinguistic factors, including occupation, place of residence, family ties, gender, and personal history, should be used to refine this theoretical distinction. 17 Greek inscriptions found in Palestine support this conclusion. Although caution is required when inscriptions are used to provide insight into the level of diffusion of a particular language, 18 the conclusion that can be drawn from available inscriptions 14. Hengel, Judaism and Hellenism, 58-106, Mussies, “Greek in Palestine and the Diaspora,” 1040-1064. For a restrictive point of view on the diffusion of Greek language and culture see T. Rajak, Josephus. The Historian and his Society, London 1983, 46-64 and L. H. Feldmann, Judaism and Hellenism Reconsidered (Supplements to the Journal for the Study of Judaism 107), Leiden/Boston, MA 2006, in particular 71-101. In my opinion Hengel’s theoretical framework is correct even if some aspects, such as local onomastic and bilingual legends on Phoenician coins require a more careful evaluation. 15. J. Barr, “Aramaic and Greek,” 91-98. 16. J. N. Sevenster, Do You Know Greek? How much Greek could the early Christians have Known? (NovT Suppl. 19), Leiden 1968, 96-175; J. Barr, “Aramaic and Greek,” 110-114. 17. J. Barr, “Aramaic and Greek,” 112. For a sociolinguistic view, see C. Hezser, Jewish Literacy in Roman Palestine (Studies in Ancient Judaism 81), Tübingen 2001, 227-247. 18. Even after taking into account documents which came to light after the studies done by Gerard Mussies (“Greek in Palestine and the Diaspora”) and H. B. Rosén (“Die Sprachsituation”), P. W. van der Horst concluded that the percentage of Greek inscriptions in the region con3. Eastern Koine: Formation and Characteristics

261

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

confirms Pieter W. van der Horst’s view that “[…] Roman Palestine was a largely bilingual, or even trilingual, society—alongside the vernacular Aramaic (and, to a much lesser extent, Hebrew), Greek was widely used and understood—but we have to add that the degree of use and understanding of the Greek language probably varied strongly according to locality and period, social status, and educational background, occasion and mobility.” 19 In a situation of bi- and trilingualism, with diglossia, where Greek, even though it was used as a second language, was the high language, forms of interference between a first language (Aramaic, Hebrew, or their respective spoken varieties) and the language of culture taught at school were natural. This phenomenon is characteristic of situations of plurilingualism and diglossia and typically results in the production of a second language interference at a phonological and lexical level and, to a lesser extent, at a morphological and syntactical one.

3.2 The Epigraphic Corpus The analysis of inscriptions is an important tool for defining the features of the koine used by speakers of Semitic languages, such as Hebrew and Aramaic. This is particularly true when the inscriptions were made by semi-educated people with little proficiency in Greek, because their writing errors reveal specific linguistic features of the particular variety of koine they spoke. Several factors however impose objective limitations on this area of investigation: the wide geographical dispersion of inscriptions ranging from Syria to the Egyptian border; the short and fragmentary nature of the available documentation, which consists mostly of epitaphs; and the general rarity of inscriptions from antiquity (although these do increase in number from the Roman period to the Byzantine era). When publication of the Corpus Inscriptionum Iudaeae/Palestinae is completed, it will provide the basis for a definitive quantitative analysis. Until then, we must limit ourselves to identifying some linguistic features based on the following edited sources: (a) volume II of Inscriptiones Judaicae Orientis, dedicated to the Syrian area, with additions by SEG and a collection of letters from the third-century CE merchant Aurelius Nebuchel, who was proficient in Aramaic, Greek, and Latin; (b) material from Palestine, including Greek inscriptions from the area surrounding Jerusalem that were documented at the end of the 1980s, the Beth Shearim collection, and successive additions by Leah Di Segni; 20 (c) the corpus edited by Reginetta Canova for the area south of Palestine, which, despite its late (fourth- and fifth-century CE) period of origin, was tinues to range between a minimum of 53 % and a hypothetical maximum of 65 %. See P. W. van der Horst, “Greek in Jewish Palestine in Light of Jewish Epigraphy” in: Collins / Sterling, Hellenism, 2001, 155-159. These figures will be further refined with the complete publication of the plurilingual Palestinian corpus, the first volume of which appeared recently (H. M. Cotton et al. [eds.], Corpus Inscriptionum Iudaeae/Palaestinae, vol. I, Berlin 2010), in order to evaluate the concentration of Greek inscription production compared with that of other languages in the same region during the same period. 19. P. W. van der Horst, “Greek in Jewish Palestine,” 166. 20. Published in G. C. Bottini / L. Di Segni / E. Alliata (eds.), Christian archaeology in Holy Land. New Discoveries, Jerusalem 1990.

262

3. Eastern Koine: Formation and Characteristics

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

still deeply influenced by a Moabite substratum; 21 (d) the Greek corpus from Ghor EsSafi (Byzantine Zoora), which can be dated to the fourth and fifth centuries CE. 22

3.3 Linguistic Features A review of the findings of modern scholarship 23 combined with linguistic analysis of the above mentioned documents enables us to identify some characteristics of the eastern koine belonging to the phonetical and phonological levels, on the one hand, and to the morphosyntaxic and lexical levels, on the other.

3.3.1 Phonetics and Phonology 1. Iotacism. Numerous examples show that iotacism appears late. This not only concerns the conservation of [e:]/__r, as Vít Bubeník has already shown, 24 but also the persistence of semi-closed and semi-open front vowels [e, ε] in all contexts. This phenomenon is not an exclusively Palestinian characteristic, since it also appears in the north and in the south, where it is even more frequent. Moreover it contrasts with the Egyptian koine, where these sounds had undergone iotacism since Hellenistic times. Otherwise, the passage from [ei, oi, y] to [i] is no different from elsewhere. 2. The reduction of the diphthong hαιi to hεi. This phenomenon is largely attested elsewhere. But in Zoora and in the southern area of the Moabite substratum, it results in an unrounded closed front vowel, as is evidenced in the frequent changes between hιi and hαιi. 3. The dephonologization of the diphthongs hαυi and hευi towards the modern pronunciations [af] and [ef] via the production of [av] and [ev]. This trait is attested in the transcription of names and Greek words in the Syro-Palestinian area. 25 4. The writing changes hεi/hιi and hουi/hωi. This phenomenon, frequent in Syria, in Zoora, and in the southern Moabite area, is indicative of closed pronunciation of mid, back, and front vowels. 5. The confusion of [a] and [o]. This is the most widely diffused phenomenon from both a geographical and statistical point of view, a fact which is probably due to the Semitic substratum, where a low rounded back vowel [ɒ] was very difficult to distinguish from a rounded semi-open back vowel [ɔ]. It appears in the Greek translation of 21. R. Canova, Iscrizioni e monumenti protocristiani del paese di Moab, Rome 1954. 22. Y. E. Meimaris and K. I. Kritikakou-Nikolaropoulou, Inscriptions from Palaestina Tertia, Athens 2005. 23. Essentially based on E. Banfi, “Ai margini della koiné greco-romana: tratti sub-standard nei conti di Aurelius Nebouchel, mercante di Dura Europos (235-240 d.C.)” in: P. Cuzzolin (ed.), Studi di linguistica greca, Milan 1993, 13-30; Bubeník, Hellenistic and Roman Greece, 227-237; N. De Lange, “Jewish Greek” in: A.-F. Christidis (ed.), A History of Ancient Greek. From the Beginnings to Late Antiquity, Cambridge 2007, 638-645; M. Janse, “Aspects of Bilingualism in the History of Greek Language” in: J. N. Adams / M. Janse / S. Swain (eds.), Bilingualism in Ancient Society. Language Contact and the Written Word, Oxford 2002, 332-390; M. Janse, “The Greek of the New Testament” in: Christidis, History, 645-653. 24. Bubeník, Hellenistic and Roman Greece, 228. 25. For an analysis of the various ways of writing the second element of the diphthong, hBi, hPi, or hWi, see Bubeník, Hellenistic and Roman Greece, 228-229. 3. Eastern Koine: Formation and Characteristics

263

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

indigenous names 26 and in the spelling of Greek words. Two examples of this are particularly interesting. First, in Zoora, among more than one hundred examples of hαi/ hοi change in the form μνημίαν, more frequent than the form μνημεῖον, iotacism appears in addition to the singular accusative neuter morpheme [a] < [o]. Second, an inscription from the Church of Mary Theotokos on Mount Gerizim, believed to be of Samaritan origin because of the onomastic and the formulas used, provides the form αὔκου for οἴκου, which Di Segni sees as a manifestation of an Aramaic substratum, due to the variation of the initial vowel. 27 6. The phonological distinction between voiced and voiceless stops. The few extant cases of written confusion between voiceless stops and aspirates can be explained by the incipient fricativization of aspirates, 28 similar to what happens elsewhere. 7. Sonorization of stops. Hebrew and Aramaic transcriptions of Greek words show that nasals tend to produce the sonorization of the following stop. 29 On the other hand, the tendency toward nasal preconsonantic weakening, well attested in this region, is a phenomenon that is characteristic not only of this area but also of Hellenistic Egypt and Cyprus. 8. Frequency of vowel prosthesis. The fact that the Aramaic substratum did not allow the s + stop group in the initial position explains the frequent use of vowel prothesis that is attested in local transcriptions of Greek anthroponyms. Moreover, this phenomenon is often found in inscriptions from Palmira and in trilingual secondand third-century CE Dura Europos. 30 A case of prothesis even occurs in Zoora where the form Ἀκλαυδις is used for Κλαυδίου. Forms of vowel anaptyxis, used to eliminate internal consonant groups, are also found in the same area, for example, Καλαπόρνου for Καλπουρνίου, μηνεμῖον for μνημεῖον, and τριακοσσιοσιτοῦ for τριακοσιοστοῦ.

3.3.2 Morphosyntax and Lexicon The epigraphic corpus yields no distinctive features of the eastern koine from a morphological point of view, since all the characteristics at this level are common to contemporary Greek. These characteristics include: extension of plural nominative endings to the accusative in athematic declension, use of the nominative instead of the accusative or the vocative, lack of agreement between apposition and attribute and the noun to which they refer, uncertain use of verbs with different cases (βοηθέω is typical), use of prepositions constructed with cases that differ from the standard. The situation changes entirely when we consider the language of the Septuagint. 26. The most frequent and characteristic example of this phenomenon is the proper noun “Solomon,” which, in the Moabite area, was alternatively transcribed as Σαλαμάνος, Σελαμάνης, Σολομόνη. See Canova, Iscrizioni e monumenti, LXXXII. 27. L. Di Segni, “The Church of Mary Theotokos at Mount Gerizim: the Inscriptions” in: Bottini / Di Segni / Alliata, Christian archaeology, 344-345. 28. Bubeník, Hellenistic and Roman Greece, 232-233. 29. Bubeník, Hellenistic and Roman Greece, 230-231. 30. Interesting cases of prothesis occur in Donaria Templi Deorum Palmyrenorum (SEG VII, 371377, II sec. d.C.): εἰσκότλα (lat. scutella); εἰσπύλλα (lat. spinula); ἰσταρτήγα (SEG XV, 850), which probably corresponds to the Greek στρατηγός. See Banfi, “Ai margini della koiné greco-romana,” 20.

264

3. Eastern Koine: Formation and Characteristics

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

According to a recent study, the most relevant features are the following: (1) preference for parataxis over of hypotaxis, (2) high frequency of nominal sentences, (3) preference for basic verb-subject-object word order, (4) frequent use of καί, (5) incorrect use of prepositions, (6) high frequency of the figura etymologica. 31 Although these phenomena are common in Post-Classical Greek, two important aspects are worth emphasizing. First, none of the above-mentioned elements are found in the epigraphic corpus. Second, the frequency with which they are used in translations from Hebrew or by authors of Septuagint books who wrote directly in Greek seems to indicate a desire to create a particular style, readily recognizable as biblical. 32 Hence, the presence of these morphosyntactical features cannot be attributed solely to first language interference in bilingual speakers whose primary language was Hebrew and who were not fully proficient in Greek. Indeed, these features might have been introduced deliberately, for literary or theological reasons. The will to create a particular style appears, to an even greater degree, at the lexical level, where Greek words acquire new meanings. Examples of this include: ψυχή, παράδεισος, and ἀγάπη. Given this evidence, one can only agree with Nicholas De Lange’s conclusion: “These various peculiarities [i. e. lexical Semitisms] are specific examples of a general approach to translation (often misleadingly called ‘literal’) that is willing to allow the source language to influence the translation to the point of compromising and sometimes risking unintelligibility […].” 33

4. Conclusion Two important conclusions concerning the eastern koine in Roman times and in late Antiquity can be drawn from the analysis above. First, a phonetic/phonological system, characterized by relevant phenomena in vowels (see § 3.3.1 above), appears to be typical of this variety of the language. Second, the absence of distinctive characteristics at the morphosyntactic and lexical levels contrasts with the linguistic peculiarities of some books of the Septuagint. The Greek koine spoken in Palestine in Roman times and in late Antiquity as a second language was not an independent linguistic system (the notorious “Jewish Greek”), but a regional variety easily recognizable by its pronunciation, not from other linguistic characteristics as has been demonstrated above. The witness of ancient literary sources further corroborates this conclusion. Josephus declared that although he had acquired a sufficient level of proficiency in Greek (τὴν γραμματικὴν ἐμπειρίαν ἀναλαβών), he still suffered from problems of pronunciation because of first language interference (Τὴν δὲ περὶ τὴν προφορὰν ἀκρίβειαν ἡ πάτριος ἐκώλυσε συνήθεια —Ant. Iud. XX, 263). 34 This predicament seems to have persisted for some time if we consider that Ermias Sozomenos, a fifth-century CE Gazan-born historian, in describing the spoken language of the fourth-century CE Syrian bishop Severian of Gabala, 31. 32. 33. 34.

De Lange, “Jewish Greek,” 641-642. De Lange, “Jewish Greek,” 640-641. De Lange, “Jewish Greek,” 642-643. For a discussion, see Rajak, Josephus, 46-64. 4. Conclusion

265

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

commented on its “discordance” and defined it as typical of the Greek spoken by a Syrian (Σευριανὸς τὴν Σύρων δασύτητα […] ἐπὶ τῆς γλώσσης ἔφερεν—Historia Ecclesiastica VIII,10). These remarks confirm what has long been thought, namely that both the theory of koine formation and data collected from inscriptions in the Syro-Palestinian area point to the fact that the eastern koine was a regional variety of Greek during the Roman period immediately recognizable by a particular “accent.” 35 It therefore cannot be defined as the independent linguistic system of a particular linguistic community. If this all can be ascertained concerning Palestinian Greek in Roman times and late Antiquity, then the possibility of providing an answer to the question raised at the beginning of this study is quite remote. This is due to the written nature of the extant historical evidence, whereas the Palestinian origin of a well-educated person capable of writing correct Greek was only revealed by his pronunciation. A thorough analysis of unemended papyrus fragments or codices may furnish new insights. The presence in a text of at least one of the phonetic phenomena described above as characteristic of the eastern koine can be taken as evidence in favor of authorship by a native Aramaic- or Hebrew-speaker. The form of the name “Solomon” in Song of Songs illustrates this criterion. Codices A and B consistently use the form Σαλωμων, whereas S, after reproducing this form in the opening verse, employs the form Σολομων in the remaining text. 36 These spellings variations bear witness to the phonetic trait examined above (§ 3.3.1, paragraph 5, note 26). As can be seen, evidence of externally documented Palestinian elements in the Greek of the Septuagint are extremely limited, but only information of this type can be admitted if the hypothesis of a Palestinian origin for these books is to be based on something other than internal stylistic elements. This is all the more important that many of these latter elements may be attributable to a conscious desire on the part of the authors to imitate, independently of their own linguistic repertoire, the style of previous works translated from the Hebrew.

Addendum Between the time of writing (2010) and the time of the final revision (2014) of this chapter the first four volumes of Inscriptionum Iudaeae/Palestinae have been published; this work aims at collecting the whole epigraphic production from the period of Alexander to Muhammad*. The volumes are undoubtedly a valuable work for reconstructing the epigraphic landscape and the multilingualism which characterizes the region from the Hellenistic period until the late Antiquity; however, as far as the aim of my work is concerned, from a first and provisional review of the volumes, it seems that data which modify in a substantial way the basis of the thesis which has been presented here do not appear. Finally, for a complete bibliographic panorama it is useful to refer also to M. Mancini’s work**, where some inscriptions of the first volume of the Corpus were analysed, particularly those on the ossuaries (I BCE–I CE), in which convincing proofs about an

35. Horsley, New Documents, vol. 5, 40 36. See the apparatus of Rahlfs’s edition, vol. II, 260, ad loc.

266

4. Conclusion

18. Do Some Books of the Septuagint Contain Elements of Palestinian Greek?

imperfect acquisition of Greek from speakers of Hebraic and Aramaic as L1 were individuated. *Corpus Inscriptionum Iudaeae/Palestinae, A multi-lingua corpus of the inscriptions from Alexander to Muhammad, H. Cotton, L. Di Segni, W. Eck, B. Issak, A. KushnirStein, H. Misgav, J. Price, I. Roll, A. Yardeni (Eds.). VOL I, Jerusalem Part 1 (1-704), Berlin – New York 2010 VOL I, Jerusalem Part 2 (705-1120) Berlin – New York 2012 VOL II, Caesarea and the Middle Coast (1121-2160) Berlin – New York 2011 VOL III, South Coast (2161-2648) Berlin – New York 2014. **M. Mancini, “L’epigrafia giudaica e la diffusione del greco nella Palestina Romana”, in M. Mancini, L. Lorenzetti (a cura di), Le lingue del Mediterraneo antico. Culture, mutamenti, contatti, Roma 2013, 213-259.

4. Conclusion

267

V The vocabulary of the Septuagint

19. La lexicographie de la Septante Gilles Dorival

1. Les instruments de travail 1.1 Dictionnaires et lexiques de la langue grecque P. Chantraine, Dictionnaire étymologique de la langue grecque. Histoire des mots, Paris 1968-1980, 2009 (nouvelle édition). La Septante est souvent citée et apparaît comme un témoin précieux pour la connaissance de la langue grecque et de son lexique dans leur histoire. G. Liddell / R. Scott / S. Jones, A Greek-English Lexicon, Oxford 1996 (première édition de G. Liddell / R. Scott, 1843; neuvième édition de S. Jones / R. McKenzie, 1940), 2042 pages, suivies d’Addenda et corrigenda, pages 2043-2111; A Supplement d’E. A. Barber, 1968, 153 pages; Revised Supplement de P. G. W. Glare / A. A. Thompson, 1996, 320 pages. Le LSJ est consultable en ligne soit dans le Perseus Project (2007, www.perseus.tufts.edu), soit dans le Thesaurus Linguae Graecae (2011, www.stephanus. tlg.uci.edu). La partie la plus faible de ce dictionnaire de référence concerne précisément la Septante, comme l’a bien montré G. B. Caird 1: les sens donnés comme propres à la Septante résultent de relevés anciens, qui attribuent souvent aux mots grecs le sens des mots hébreux supposés sous-jacents; mais le modèle hébreu de la Septante n’était pas en tous points identique au texte massorétique; de plus, l’étude des papyrus permet de supprimer un grand nombre de ces prétendus sens propres à la Septante. Emanuel Tov cite l’exemple de σύνδεσμος, auquel le LSJ donne en 1S (= 1Règnes) 14,23 le sens de sodomy, parce qu’il traduirait ‫קדש‬, « prostitué mâle »; en fait, les traducteurs lisaient ici ‫קשר‬, qui signifie « conspiration », tout comme le mot grec 2. Robert Renehan apporte de nombreuses améliorations aux notices du LSJ où il est question de la Septante 3. Le Supplement du LSJ de 1968 a suscité des critiques parallèles à celles qui viennent d’être signalées 4. R. Adrados, Diccionario Griego-Espanol, Madrid 1980-2007 (sept volumes parus qui couvrent les lettres α, β, γ, δ et le début d’ε ; une édition révisée et augmentée du volume I α-ἀλλά a été publiée en 2008). 1. 2.

3.

4.

G. B. Caird, « Toward a Lexicon of the Septuagint » JTS 19 (1968), 453-475 et 20 (1969), 21-40. E. Tov, « Some Thoughts on a Lexicon of the Septuagint » BIOSCS 9 (1976), 14-46 (réimpression in E. Tov, The Greek and Hebrew Bible. Collected Essays on the Septuagint, Leiden/Boston, MA/Köln 1999, 95-108). R. Renehan, Greek Lexicographical Notes: A Critical Supplement to the Greek-English Lexicon of Liddell-Scott-Jones, Göttingen 1975, et Greek Lexicographical Notes. Second Series. A Critical Supplement to the Greek-English Lexicon of Liddell-Scott-Jones, Göttingen 1983. J. A. Lee, « A Note on the Septuagint Material in the Supplement of Liddell and Scott » Glotta 47 (1969), 234-242; W. Baars, « Review of The Greek-English Lexicon. A Supplement » VT 20 (1970), 371-379. 1. Les instruments de travail

271

19. La lexicographie de la Septante

On peut consulter avec profit les dictionnaires et lexiques portant sur la langue grecque postérieure à la Septante: — pour le Nouveau Testament: W. Bauer, Griechisch-Deutsches Wörterbuch zu den Schriften des Neuen Testaments und der übrigen urchristlichen Literatur, Berlin 1910, traduction anglaise par W. F. Arndt / F. X. Gingrich, A Greek-English Lexicon of the New Testament, Cambridge 1957: pour tous les mots de la Septante présents dans le Nouveau Testament, ce lexique donne une bibliographie utile. G. Kittel (éd.), Theologisches Wörterbuch zum Neuen Testaments, dix volumes, Stuttgart 1903-1974 (en abrégé TWNT), traduction anglaise G. W. Bromiley, Theological Dictionary of the New Testament, dix volumes, Grand Rapids, MI 1968-1976: les mots de la Septante présents dans le Nouveau Testament sont étudiés tant dans la Septante que dans le Nouveau Testament, avec une riche bibliographie; il faut cependant se méfier de la théologisation excessive de ces notices, comme l’a bien montré James Barr 5. — pour les Pères de l’Église: G. W. H. Lampe, A Patristic Greek Lexicon, Oxford 1961, 1568 pages. Ce dictionnaire grec-anglais n’offre pas tous les mots attestés chez les Pères grecs, mais uniquement ceux qui ont un contenu théologique. Par exemple, l’article ὁ ou le démonstratif οὗτος manquent. Les mots absents du LSJ sont précédés de l’astérisque *. — pour l’époque romaine et byzantine: E. A. Sophocles, Greek Lexicon of the Roman and Byzantine Periods, Cambridge, MA/Leipzig 1914 (réimpression: Hildesheim/Zurich/New York, NY 1983, 1188 pages.

1.2 Concordances La plus ancienne d’entre les concordances, celle de Conrad Kircher, paraît cent cinquante ans avant le premier lexique de la Septante, mais elle peut être utilisée comme lexique, puisqu’elle donne les traductions latines des mots grecs et hébreux. Il en va de même dans la concordance suivante, celle d’Abraham Trommius. À partir de la fin du XIXe siècle, les concordances (Morrish, Hatch-Redpath) cessent d’être utilisables comme lexiques. C. Kircher, Concordantiae Veteris Testamenti Graecae, Ebreis vocibus respondentes, deux volumes, Frankfurt am Main 1607. C’est la première concordance imprimée de la Septante. Les mots hébreux sont donnés dans l’ordre alphabétique. Sous chaque mot hébreu figurent les différentes traductions grecques avec l’indication de leur provenance. À la fin du second volume, les mots grecs sont donnés dans l’ordre alphabétique, avec l’indication des pages où l’on peut les lire dans la concordance. A. Trommius, Concordantiae graecae versionis vulgo dictae LXX interpretum, deux volumes, Amsterdam/Utrecht 1718. À la différence de la concordance de Kircher, qui 5.

J. Barr, Semantics of Biblical Language, Oxford 1961 (traduction française: Sémantique du langage biblique, Paris 1968).

272

1. Les instruments de travail

19. La lexicographie de la Septante

donne d’abord les mots hébreux, Trommius propose les mots grecs dans l’ordre alphabétique, suivis de leur traduction latine. Viennent ensuite les correspondants hébreux, suivis de leur traduction latine quand elle est différente du sens du mot grec; en de tels cas, les versets grecs concernés de la Septante sont donnés. Suivent les leçons des traducteurs; puis, les occurrences du mot grec qui n’ont pas de correspondant en hébreu; enfin les occurrences dans les deutérocanoniques. Cette concordance n’a pas été rendue complètement obsolète par la concordance postérieure d’Edwin Hatch et Henry A. Redpath, parce qu’elle offre des mots absents de cette dernière en provenance de manuscrits ou d’éditions que n’ont pas utilisés Hatch-Redpath. L’index hébreu-grec donne à la fois les mots hébreux et grecs et est donc plus facilement utilisable que le Reverse Index du Supplement d’Hatch-Redpath (voir plus bas). G. Morrish, A Concordance of the Septuagint giving various readings from codices vaticanus, alexandrinus, sinaiticus, and ephraemi; with an appendix of words, from Origen’s hexapla, etc; not found in the above manuscripts, Londres 1887 (réimpression Grand Rapids, MI 1999, 284 pages). Chaque mot grec est suivi de la liste des livres et versets où il apparaît. Mais il ne donne aucune citation de la Septante. Les racines hébraïques correspondant aux mots grecs ne sont pas proposées. Cet instrument de travail est un simple index, qui ne cite même pas les contextes. Son principal atout est son faible prix. Il peut rendre service aux étudiants. Mais ceux-ci devront très vite consulter les concordances dont il va maintenant être question. E. Hatch, H. A. Redpath, A Concordance to the Septuagint and the Other Greek Versions of the Old Testament (Including the Apocryphal Books), deux volumes (α-ι, κω), Oxford 1897, 1504 pages (dont des Addenda et corrigenda en tête du volume 1 et aux pages 1499-1504), Supplement, Oxford 1906, 272 pages; le Supplement est imprimé immédiatement à la suite de la Concordance dans la réimpression faite à Graz 1975 (en abrégé HR). La concordance a été republiée à Grand Rapids, Michigan, en 1999 6. Chaque page de la Concordance est divisée en trois colonnes qui se lisent à la suite. Chaque mot grec est imprimé en gras; il est suivi sur la même ligne et, le cas échéant, sur les lignes suivantes par les mots hébreux auxquelles il correspond; chaque mot en caractères gras est précédé d’un numéro entre parenthèses, (1), (2), etc.; les formes différentes d’un même mot (par exemple, dans le cas d’un verbe, le qal, le piel, le hiphil, etc.) sont distinguées par les lettres a., b., etc. Le mot grec et ses équivalents hébreux sont suivis de toutes les occurrences du mot dans la Septante. On lit, dans l’ordre: l’abréviation du livre biblique; le chapitre et le verset concernés; le mot grec dans son contexte immédiat, de trois mots à une dizaine; enfin le numéro de la racine correspondante. Par exemple, l’adjectif ἄβατος, auquel correspondent onze racines, est suivi de vingt-neuf occurrences, dont la première est « Le 16,22. λήμψεται … τὰς ἀδικίας αὐτῶν εἰς γῆν ἄ. (3) »; « Le » est l’abréviation de Lévitique et l’adjectif est réduit à sa première lettre. Si dans une occurrence, le mot grec n’a pas de correspondant en hébreu, cela est indiqué en fin de ligne par -; si le symbole † figure en fin de ligne, cela signale une difficulté dans l’identification du grec et de l’hébreu et suggère qu’un textualiste devrait regarder les textes de plus près; si un numéro de racine est suivi de?, cela signifie que l’identification de la racine n’est pas sûre. Par exemple, le mot σύν6.

Cette republication contient des contributions de R. A. Kraft et E. Tov (« Introductory Essay ») et de T. Muraoka (« Hébrew/Aramaic Index to the Septuagint »). Voir plus bas. 1. Les instruments de travail

273

19. La lexicographie de la Septante

δεσμος est suivi de six racines et de dix occurrences; la deuxième d’entre elles, celle de 3Règnes 14,24/23, s’achève par †; la concordance signale ainsi que le texte grec et le texte hébreu ne se correspondent pas; c’est en 1976 que Tov montrera, dans un article signalé plus haut, que les traducteurs lisaient ici ‫קשר‬, « conspiration », et non ‫קדש‬, « prostitué mâle ». L’ensemble des occurrences de la Septante est suivi des listes des emplois chez les autres traducteurs, sans mention des mots hébreux correspondants; chaque liste est donnée entre crochets droits, et les abréviations des livres bibliques sont imprimés en petites capitales; par exemple [Aq. Ps 77 (78). 17; Is 41. 18: 52. 2; Je 50 (27). 12.]. Le Supplement contient quatre indices: 1 / Une concordance des noms propres sur trois colonnes par page (pages 1-162); la racine (1) est suivie de la liste des occurrences, sans citation du grec; puis, le cas échéant, la racine (2), etc. 2 / Une concordance des mots de l’Ecclésiastique avec leurs équivalents hébreux sur trois colonnes par page (pages 163-196; la présence de cette concordance tient au fait que le texte hébreu de l’Ecclésiastique avait été trouvé peu de temps auparavant dans la Genizah du Caire). 3 / Une liste supplémentaire de mots trouvés dans les fragments hexaplaires sur trois colonnes par page (pages 197-216). 4 / Un index hébreu de l’ensemble de la concordance disposée sur quatre colonnes par page (pages 217-272): chaque mot hébreu est suivi des références de la page et de la colonne où il figure; les pages de la concordance sont imprimées en caractères droits, celles du supplément en italiques. Par exemple, ‫ סֶֹבא‬983c, 157b, où la première indication renvoie à la page 983 troisième colonne de la Concordance et la seconde, à la page 157 deuxième colonne du Supplement. La concordance de Hatch-Redpath est un remarquable instrument de travail, mais elle a des limites, qui, en s’inspirant des remarques de Tov et de Staffan Olofsson, sont les suivantes 7: 1 / Elle est faite sur le texte du Vaticanus; cependant, elle indique les principales variantes de l’Alexandrinus, du Sinaiticus et de l’édition Sixtine de 1587; mais elle est antérieure à l’editio critica minor publiée par Alfred Rahlfs en 1935 et à l’editio critica major entreprise depuis 1931 par le Septuaginta-Unternehmen de l’Académie des Sciences de Göttingen et aujourd’hui en voie d’achèvement. De la sorte, un certain nombre de mots sont absents de HR: ceux qui figurent dans les manuscrits et éditions que la concordance n’a pas retenus, ainsi que les corrections des éditeurs de Göttingen. 2 / Si tous les mots de la Septante figurent dans HR, certains d’entre eux, comme δέ, καί, οὐκ, les pronoms personnels et démonstratifs, etc., sont suivis simplement de l’indication passim, sans que soient mentionnés ni les correspondants hébreux ni les références scripturaires, tandis que d’autres, comme les adjectifs possessifs, les conjonctions de subordination du type ὅτι, les négations μή et οὔτε, les noms de nombre, les prépositions comme παρά, etc., sont pourvus de la liste des occurrences, mais ni les correspondants hébreux ni les contextes de la Septante ne sont donnés. 3 / Les mots grecs sont donnés isolément, sans indication des mots apparentés, des synonymes et des antonymes, pourtant indispensable pour une réflexion lexicographique. 4 / Les mots hébreux donnés comme correspondant aux mots grecs sont parfois inexacts; lorsque la concordance a été faite, l’idée que le modèle hébreu de la Septante pouvait être légèrement différent, voire assez différent, du texte massorétique 7.

E. Tov, Text-Critical Use of the Septuagint in Biblical Research, Jérusalem 1981, 142-154; S. Olofsson, The LXX Version. A Guide to the Translation Technique of the Septuagint, Stockholm 1990, 75-77.

274

1. Les instruments de travail

19. La lexicographie de la Septante

était acceptée, mais il était affirmé que la concordance n’avait pas pour tâche de reconstituer ce modèle (voir la préface de Redpath, page vi); dans les faits, les mots du texte massorétique sont privilégiés et le modèle hébreu de la Septante, même là où il est aisé à reconstituer, n’est pas indiqué. 5 / L’index hébreu contenu dans le Supplement n’est pas le pendant de la concordance: alors que celle-ci donne les mots hébreux correspondant aux mots grecs, l’index ne donne pas les équivalents grecs correspondant aux mots hébreux, mais seulement les pages où ces équivalents sont cités; pour les connaître, le lecteur doit se reporter à ces pages, ce qui constitue un travail long et malaisé quand une même racine hébraïque a plusieurs correspondants grecs, jusqu’à une cinquantaine dans le cas de ‫אמר‬. 6 / Les citations des autres traducteurs reposent sur l’édition des Hexaples publiée par Frederick Field en 1867-1875; cependant le troisième index du Supplement tient compte des publications postérieures offrant des leçons des autres traducteurs; il a notamment eu accès aux fragments inédits des psaumes trouvés par Giovanni Mercati, qui ne seront publiés qu’en 1958; mais il ignore évidemment les leçons des traducteurs trouvées après 1906 dans les commentaires patristiques et dans les chaînes exégétiques grecques. De plus, l’absence d’indication des mots hébreux correspondant aux mots grecs des traducteurs est une vraie limite dans une concordance. 7 / Du point de vue lexicographique, la limite la plus importante de HR est que, dans les citations de la Septante, seuls quelques mots sont donnés; le contexte est insuffisant pour une analyse en profondeur; la concordance ne dispense pas de se reporter au texte même de la Septante. X. Jacques, Index des mots apparentés dans la Septante. Complément des concordances et dictionnaires, Rome 1972 (traduction anglaise: List of Septuagint Words sharing common elements. Supplement to concordance or dictionary, Rome 1972). Cet ouvrage qui donne la liste de tous les mots appartenant à une même famille (par exemple κλῆρος et ses vingt-trois dérivés) est utile pour quiconque entreprend une analyse lexicographique. E. C. Dos Santos, An Expanded Hebrew Index for the Hatch-Redpath Concordance to the Septuagint, Jerusalem s. d. [1973]. L’index hébreu du Supplement est reproduit photographiquement; chaque mot hébreu est suivi des mots grecs correspondants, écrits à la main; le nombre total des occurrences de chaque mot grec est indiqué entre parenthèses par un chiffre suivi de x, par exemple (16x). Aujourd’hui, le TwoWay Index de T. Muraoka remplace cet instrument de travail, sauf pour la fréquence des mots grecs, qui est absente chez Muraoka lorsque les chiffres sont élevés. T. Muraoka, A Greek-Hebrew/Aramaic Index to I Esdras, Chico, CA 1984. Cet index a été intégré dans les index de 1998 et de 2010 (voir plus bas). T. Muraoka, Hebrew/Aramaic Index of the Septuagint Keyed to the Hatch-Redpath Concordance, Grand Rapids, MI 1998. Ce volume est à présent remplacé par le Two-Way Index du même auteur paru en 2010. T. Muraoka, A Greek Hebrew/Aramaic Two-way Index to the Septuagint, Louvain/Paris/Walpole, MA 2010, 383 pages. Cet index complète le Lexicon de 2009, qui a renoncé à donner les équivalents hébreux des mots grecs. Il se propose aussi de remplacer l’index hébreu du Supplement de HR. Son texte de référence de la Septante est celui qui est établi au Septuaginta-Unternehmen de Göttingen; il intègre aussi le texte antiochien des livres historiques; enfin, il tient compte des textes bibliques et apocryphes/deutérocanoniques trouvés à Qumrân, aussi bien en hébreu qu’en araméen. 1. Les instruments de travail

275

19. La lexicographie de la Septante

Muraoka a procédé à une révision critique des données de HR: des équivalences retenues par ces derniers sont rejetées, de nouvelles correspondances entre les deux lexiques sont proposées. La partie I (pages 1-130) est consacrée au Greek Hebrew/Aramaic Index, imprimé sur deux colonnes. Chaque mot grec est suivi des mots hébreux qui lui correspondent, numérotées 1), 2), etc. Chacun de ces mots est suivi d’un chiffre entre crochets droits, qui indique le nombre des occurrences de l’équivalence. Ce chiffre est parfois suivi de +, qui signifie que le nombre des occurrences est plus élevé, et ce + peut lui-même être complété; par exemple, sous l’entrée διαθήκη, à la suite de la racine 2) ‫ברית‬, on lit ceci: [93+; + 4K 11.42, Sir 45,24], ce qui signifie que le nombre d’attestations est plus élevé que quatre-vingt-treize et que l’équivalence est attestée dans deux citations de 4Règnes et Siracide que ne signalent pas HR; le même signe + suivant une référence scripturaire indique qu’il y a la même équivalence dans d’autres passages de la Septante. Le signe – devant une référence scripturaire indique que la référence en question ne doit pas être retenue pour l’équivalence considérée. En fin de notice figurent les mots hébreux qu’il faut rejeter. Par exemple, chez HR, le mot ἄβατος est suivi de onze mots; ces mots se retrouvent chez Muraoka, sauf 2), qui est rejeté en fin de notice avec la mention del., qui signifie qu’il faut supprimer cette équivalence. L’astérisque * indique qu’on a affaire à une nouveauté par rapport à HR: ainsi un mot absent de la concordance est précédé de * (par exemple * φράξις); ou une équivalence absente de HR est pourvue du même signe (par exemple κοινῶς *1) Ar. ‫[ ַכֲּח ָדא‬1: To. 6.6]; autrement dit, l’équivalence entre le grec et l’araméen attestée une fois en Tobit n’est pas connue de HR). La partie II (pages 131-383) donne le Hebrew/Aramaic Greek Index imprimé sur trois colonnes par page. Chaque mot hébreu ou araméen est suivi de ses équivalents grecs, donnés dans l’ordre alphabétique et pourvus de l’indication de la page de HR où ils figurent. Lorsque l’on a affaire à un mot araméen, l’indication Ar. suit immédiatement ledit mot. L’astérisque * indique les nouveautés par rapport à HR: il est présent devant les mots grecs absents de HR comme devant les nouveaux équivalents. L’indication « Aramaising » à la suite d’une référence scripturaire signifie que le traducteur a probablement lu ici un mot araméen, et non hébreu. Le symbole ^ devant un mot grec indique que ce mot figure dans 1 Esdras. Le symbole § devant un mot grec indique que ce dernier est une translittération. Le signe + devant une citation scripturaire indique que celle-ci est absente de HR. Le signe – devant une citation indique que celle-ci, qui figure dans HR, doit être écartée. Le ? devant un mot grec indique que ce mot, qui figure en HR, est douteux. Les crochets droits [ ] entourant un mot grec indiquent que ce mot est une variante textuelle du codex 248 du Siracide. Les doubles cochets droits entourant un mot grec signalent que le mot en question n’est pas plausible; par exemple, sous l’entrée ‫ַכּר‬, l’indication [[τόπος 1364b (Is 30.23à)]] indique qu’à la page 1364b de HR, il n’est pas plausible que τόπος traduise ‫ כר‬en Es 30,23. Lorsque les doubles crochets droits sont suivis d’une flèche →, ce qui suit la flèche est une suggestion de Muraoka corrigeant HR; cette suggestion peut concerner le mot grec qu’il faut substituer au mot hébreu indiqué par HR, ou bien le mot hébreu ou araméen équivalent du mot grec, ou bien à la fois le mot hébreu et son équivalent grec; si le mot hébreu équivalent à un mot grec n’est pas connu, il est indiqué par XXX. Une différence avec HR doit encore être signalée: HR regroupe dans une seule et même entrée tous les modes d’un verbe; Muraoka donne autant d’entrées qu’il y a de modes, par exemple trois entrées dans le cas de ‫אבד‬, pour le qal, le piel et le 276

1. Les instruments de travail

19. La lexicographie de la Septante

hiphil. Le Two-Way Index est un indispensable instrument de travail, qui a toutefois quelques limites. 1 / Il ne permet plus la comparaison avec Aquila, Symmaque, Théodotion et les autres traductions. 2 / Les noms propres sont absents; toutefois les adjectifs dérivés sur des noms propres sont présents dans les deux parties (ἰδουμαῖος, ἰσραηλῖτις). 3 / Dans l’index grec-hébreu/araméen, le chiffre des occurrences d’une équivalence ne permet pas de mesurer réellement la fréquence de cette équivalence: Muraoka indique 93 + pour l’équivalence διαθήκη-‫ברית‬, alors qu’elle est attestée 273 fois selon Elmar Camilo Dos Santos; 143 + pour l’équivalence βασιλεύς-‫ֶמֶלְך‬, alors qu’elle est attestée 2 551 fois selon Dos Santos. Au total, le Lexicon de 2009 et le Twoway Index de 2010 rendent largement obsolètes les travaux antérieurs. Mais l’index des noms propres de HR reste indispensable. Il faut souhaiter que Muraoka comble cette lacune.

1.3 Lexiques de la Septante J. C. Biel, Novus thesaurus philologicus, sive lexicon in LXX et alios interpretes et scriptores apocryphos Veteris Testamenti, trois volumes, La Haye, 1779-1780. Ce lexique a été édité après la mort de Johann Christian Biel en 1745 par E. H. Mutzenbecher. Il dépend fortement de la concordance de Trommius. Il arrive cependant que Biel fasse des remarques qui lui sont propres et où il discute par exemple d’expressions grecques. Les sens retenus par Biel pour les mots grecs ont fortement influencé les notices du LSJ. J. F. Schleusner, Novus thesaurus philologico-criticus, sive Lexicon in LXX et reliquos interpretes graecos ac scriptores apocryphos Veteris Testamenti, cinq volumes, Leipzig 1820-1821; édition révisée et complétée, trois volumes, Glasgow 1822, London 1829; réimpression, Turnhout 1994 (qui reproduit l’édition de Glasgow). Le Lexicon graecum Veteris Testamenti occupe les deux premiers volumes de l’édition de Glasgow, ainsi que les pages 1-498 du volume 3: chaque mot grec de la Septante et des autres traducteurs est suivi de sa traduction latine, puis du (ou des) mot(s) hébreu(x) qui lui correspond(ent), ensuite des versets où il est cité; le cas échéant, il y a des discussions philologiques avec renvoi aux lexiques et aux auteurs de l’Antiquité, ainsi qu’aux modernes. La fin du volume 3 contient un Index hebraicus: chaque racine hébraïque est suivie de l’indication des pages où elle est citée. Les limites de ce thésaurus tiennent au fait qu’il a été composé avant les grandes découvertes épigraphiques et papyrologiques qui ont renouvelé la connaissance du lexique et de la langue grecs de l’époque hellénistique. De plus il est fait du point de vue des mots hébreux correspondant à la Septante. Johan Lust a raison de souligner qu’il s’agit moins d’un dictionnaire de grec biblique que d’un lexique d’hébreu biblique 8. Néanmoins, le Novus thesaurus contient une masse d’informations utiles. F. Rehkopf, Septuaginta-Vokabular, Göttingen 1989. Il donne pour chaque mot grec une traduction allemande, ainsi que des indications sur le nombre des occurrences dans l’Ancien et le Nouveau Testaments. Ce lexique est susceptible de rendre service aux étudiants. J. Lust / E. Eynikel / K. Hauspie, A Greek-English Lexicon of the Septuagint, deux volumes (α-ι, κ-ω), Stuttgart 1992-1996 (en abrégé LEH). Chaque mot grec donne lieu 8.

J. Lust, « J. F. Schleusner and the Lexicon of the Septuagint » ZAW 102 (1990), 256-262. 1. Les instruments de travail

277

19. La lexicographie de la Septante

à au moins trois lignes. La première ligne offre à gauche le mot grec; s’il est suivi du symbole +, c’est qu’il est présent aussi dans le Nouveau Testament; à droite, figure une lettre majuscule indiquant à quelle catégorie morphologique le mot appartient, V pour verbe, N pour nom, etc.; le cas échéant, cette lettre est suivie par les chiffres 1, 2 ou 3, pour les trois déclinaisons nominales, et par les lettres F, M ou N, pour féminin, masculin ou neutre; viennent alors six chiffres séparés par un trait d’union (-): les cinq premiers donnent le nombre des occurrences du mot dans le Pentateuque, les Prophètes antérieurs (y compris 1–2Paralipomènes, les Prophètes postérieurs, les Écrits et les deutérocanoniques; le sixième chiffre indique le nombre total des occurrences. La deuxième ligne donne les références du mot, limitées aux cinq premières d’entre elles. La troisième ligne offre la traduction anglaise du mot en italiques; s’il y a plusieurs sens, ils sont indiqués avec la mention d’un exemple; si le mot est une translittération, cela est indiqué par = suivi du correspondant hébreu ou araméen. Il y a d’autres signes conventionnels ou abréviations qui correspondent à des indications variées. L’indication « neol. » en fin de notice signifie que l’on a affaire à un néologisme. Sous la notice, l’indication « Cf. » précède les noms des auteurs qui ont traité du mot; on peut facilement les retrouver dans l’imposante bibliographie qui figure en tête des deux volumes. La limite de ce précieux instrument de travail est qu’il envisage le lexique de la Septante avant tout du point de vue de la traduction, sans prendre en compte le fait que la Septante est aussi une œuvre littéraire s’inscrivant dans l’ensemble de la littérature grecque. En conséquence, il a tendance à projeter le sens du mot hébreu dans le mot grec, au lieu de comprendre ce dernier dans un contexte grec. Par exemple, ὁ ἀγχιστεύων τὸ αἷμα de Nb 35,12 est traduit par « avenger of blood » avec la précision qu’il s’agit d’un sémitisme, alors qu’au niveau de la langue grecque, l’expression signifie « le proche parent quant au sang, le proche parent qui réclame le sang » et ne peut être considérée comme un sémitisme. T. Muraoka, A Greek-English Lexicon of the Septuagint (Twelve Prophets), Leuven 1993. T. Muraoka, A Greek-English Lexicon of the Septuagint. Chiefly of the Pentateuch and the Twelve Prophets, Leuven 2002. Ces deux lexiques sont intégrés dans le lexique de 2009. T. Muraoka, A Greek-English Lexicon of the Septuagint, Louvain/Paris/Walpole, MA 2009 (en abrégé GELS). Ce lexique sur deux colonnes par page est établi sur l’editio critica major entreprise depuis 1931 par le Septuaginta-Unternehmen de l’Académie des Sciences de Göttingen et, pour les livres qui ne sont pas encore disponibles, sur l’editio minor de Rahlfs, publiée en 1935. Le texte antiochien des livres historiques (Juges, 1–4Règnes et 1–2Paralipomènes a été incorporé, en particulier grâce à l’index de Natalio Fernández Marcos et al 9. Chaque mot grec donne lieu à trois sections: A. Le mot grec imprimé en gras est suivi d’indications sur sa morphologie: dans le cas des noms, génitif singulier et genre (m., f. ou n.), dans le cas des adjectifs, féminin et neutre; dans le cas des verbes, temps et modes autres que l’imparfait. Ces indications peuvent être suivies du symbole qui signifie que toutes les occurrences du mot sont

9.

N. Fernández Marcos / V. Spottorno / J. M. Cañas Reíllo, Indice griego-hebreo del texto antioqueno en los libros historicos, 2 vol., Madrid 2005.

278

1. Les instruments de travail

19. La lexicographie de la Septante

données dans la notice: 5 548 mots sur 9 548 sont dans ce cas. En fin de ligne, il peut y avoir l’astérisque *, qui indique que la Septante est la plus ancienne attestation du mot, dans l’état actuel de notre information. Selon Muraoka (page xiii), il y aurait environ 1 900 néologismes, soit un cinquième du lexique total. Mais ce chiffre est sûrement erroné. 10 Selon Christian Schröder il n’y aurait que 287 néologismes 11. Mais la comparaison pour la lettre –α entre la liste de Schröder et le Lexicon montre que ce chiffre pèche nettement par défaut: il y a cinquante et un néologismes selon Schröder, 151 selon Muraoka; le premier signale vingt et un néologismes que ne retient pas le second; celui-ci a 121 néologismes absents de la liste du premier; si l’on considère qu’entre un cinquième et un sixième des mots grecs commence par la lettre –α, on peut calculer que le nombre des néologismes doit être d’environ 850 dans le Lexicon. En ce dernier, lorsqu’un mot de la famille du néologisme considéré est attesté avant la Septante, ce mot est indiqué entre parenthèses; par exemple, βδέλυγμα est un néologisme de la Septante mais le verbe βδελύσσω est connu depuis Aristophane et d’autres auteurs: cela est indiqué par (βδελύσσω Aristoph.+). B. Cette section est consacrée à la sémantique du mot: les sens sont classés en sens principaux, numérotés en gras 1., 2., 3., etc., et sous-sens (dans la rubrique 1., b., c., etc., a. étant ellipsé par économie); chaque sens est suivi par diverses indications: par exemple par une citation et sa référence; ou simplement par une ou plusieurs références; ou par des précisions de syntaxe (par exemple + acc., + inf., etc.); etc. Quelques autres symboles sont utilisés devant des précisions utiles. C. Cette section dresse la liste des mots qu’on peut rapprocher sur le plan sémantique (soixante-dix-huit mots dans le cas d’ἀφανίζω, l’ensemble constituant le champ sémantique de la destruction) et donne la littérature secondaire. Au total, ce lexique est d’une précision remarquable. Le sens des mots est établi à l’aide de contextes larges, au minimum la phrase entière où apparaît le mot; le sens est déterminé aussi par les synonymes et les antonymes. Cependant, la plus grande originalité du lexique est l’absence presque totale de l’hébreu, à la différence des lexiques de 1993 et 2002. C’est donc un vrai travail de lexicologue de la langue grecque qu’a accompli Muraoka, comme le montre l’attention portée aux néologismes. On peut se demander toutefois si la correspondance avec les mots hébreux n’aurait pas été utile, dans la mesure où elle aurait évité au lecteur de se reporter à la concordance de Hatch-Redpath et maintenant au Two-way Index de Muraoka (voir plus bas). De plus, les noms propres sont totalement absents. Un lecteur pressé pourrait croire qu’il n’est jamais fait mention d’Israël ou de Juda dans la Septante! De même, certains mots translittérés, comme σαβεκ, ne sont pas donnés.

10. T. Muraoka (« Recent Discussions on the Septuagint Lexicography With Special Reference to the So-called Interlinear Model » in: M. Karrer / W. Kraus (éd.), Die Septuaginta – Texte, Kontexten, Lebenswelten, Tübingen 2008, 221-235) compte, page 230, 1 490 néologismes au total, dont 103 commençant par a-. 11. C. Schröder, « Alphabetische Zusammenstellung auffälliger Neologismen der Septuaginta » in: H.-J. Fabry / U. Offerhaus (éd.), Im Brennpunkt: Die Septuaginta. Studien zur Entstehung und Bedeutung der Griechischen Bibel, Stuttgart/Berlin/Köln 2001, 61-69. 1. Les instruments de travail

279

19. La lexicographie de la Septante

2. Les débats et les notions-clés 2.1 Le débat autour du grec biblique Le lexique de la Septante permet-il d’affirmer qu’il a existé un grec biblique indépendant de la koinè, un judéo-grec constituant un dialecte à part, faisant sens seulement pour des lecteurs juifs d’Alexandrie? La question se pose aussi pour le Nouveau Testament: y a-t-il eu un grec judéo-chrétien, voire chrétien? Le défenseur le plus célèbre du grec biblique est Hatch 12. Pour déterminer le sens des mots de la Septante, il utilise deux critères: 1 / si un mot est utilisé de manière uniforme pour traduire le même mot hébreu, le mot grec a le même sens que le mot hébreu; 2 / si plusieurs mots grecs sont utilisés de manière interchangeable pour traduire un mot hébreu, ces mots ont un sens voisin, voire identique. Hatch privilégie la langue source au point qu’il la voit tout entière dans la langue cible. La position de Hatch a eu des partisans jusqu’à une date récente et elle en a peut-être encore 13. Pourtant, les notions de grec biblique et de judéo-grec ont été depuis longtemps critiquées, notamment par Adolf Deissmann 14. Ce dernier a montré que les inscriptions et les papyrus de l’époque hellénistique fournissent des parallèles syntaxiques et lexicaux avec la Septante: c’est le cas par exemple de διαταγή, « ordonnance », καταπέτασμα, « rideau », πρωτότοκος, « premier-né ». A la même époque, Albert Thumb, en Allemagne, et Henry St John Thackeray, en Angleterre, sont sur les mêmes positions 15. À date plus récente, la notion de grec biblique a fait l’objet de quatre critiques argumentées. Fernández Marcos retrace le rôle central joué par la prise en compte des papyri, inscriptions et ostraca, qui ont fait singulièrement décroître le nombre des voces biblicae, des mots propres à la Septante et au Nouveau Testament 16. Il note que les spécialistes du Nouveau Testament ont longtemps résisté aux conclusions de Deissmann et Thumb et ont développé la théorie de l’origine araméenne de cet ensemble de textes. Il rappelle que, selon d’autres spécialistes, les expressions non-classiques des papyri seraient des sémitismes dus à l’influence des nombreux Juifs habitant 12. E. Hatch, Essays in Biblical Greek, Oxford 1899 (réimpression: Amsterdam 1970). 13. Voir N. Turner, Grammatical Insights into the New Testament, Edinburgh 1965, notamment 163, et J. H. Moulton / N. Turner, A Grammar of New Testament Greek, vol. IV, Style, Edinburgh 1976, 1-2; S. Thompson, The Apocalypse and Semitic Syntax, Cambridge 1985 (L’Apocalypse est un document judéo-grec). 14. A. Deissmann, Bibelstudien, Marbourg 1895; Neue Bibelstudien, Marbourg 1897; Licht vom Osten. Das neuen Testament und die neuentdeckten Texte der hellenistisch-römischen Welt, Tübingen 1908. 15. A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus. Beiträge zur Geschichte und Beurteilung der Koine, Strasbourg 1901; A. Thumb, « Die sprachgeschichtliche Stellung des biblischen Griechisch » Theologische Rundschau 5 (1902), 85-99; H. St John Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, vol. I. Introduction, Orthography and Accidence, Cambridge 1909 (réimpression: Hildesheim/New York, NY 1978), 25-31. 16. N. Fernández Marcos, Introduccion a las versiones griegas de la Biblia, Madrid 1979, 3-35 (traduction anglaise par W. G. E. Watson, The Septuagint in Context. Introduction to the Greek Versions of the Bible, Leiden/Boston, MA/Köln 2000, 3-31.

280

2. Les débats et les notions-clés

19. La lexicographie de la Septante

la vallée du Nil: le bilinguisme serait responsable des particularités linguistiques de la Septante, du Nouveau Testament et des papyri. Mais l’influence du copte est invoquée par d’autres études 17. La thèse du bilinguisme reste à approfondir. Fernández Marcos, qui s’intéresse plus à la syntaxe qu’à la lexicographie, souligne que les études sur la koinè littéraire sont encore en nombre insuffisant, même s’il est acquis qu’il faille distinguer entre la koinè populaire (les papyri) et la prose hellénistique (Polybe, Diodore de Sicile); le Nouveau Testament est plus proche de cette dernière que de la koinè populaire. Pour Greg H. R. Horsley, le judéo-grec est une fiction qui est due à l’acceptation d’une terminologie trop vague (notamment le concept de sémitisme) et à un manque d’information sur les recherches linguistiques, notamment dans le domaine du bilinguisme 18. En fait, la présence de noms propres hébreux en grec ne prouve rien sur le caractère sémitisant de la Septante, même lorsque ces noms sont indéclinables. Il en va de même dans le cas des lexiques techniques reproduisant les réalités hébraïques. De manière générale, la présence de sémitismes ne suffit pas à établir l’existence du judéogrec. Les défenseurs de ce dernier doivent être situés dans le contexte de la querelle qui a opposé, de la Renaissance au XIXe siècle, les puristes, pour qui les particularités du grec du Nouveau Testament relèvent en fait de la langue attique, et les hébraïstes, pour qui elles prouvent l’existence d’un grec biblique, voire d’une langue du Saint Esprit. Il s’agit d’une querelle aujourd’hui dépassée. Selon Moisés Silva, la langue de la Septante doit être fondamentalement décrite comme du grec hellénistique et, seulement dans un deuxième temps, comme du grec hellénistique « juif », mais au sens où l’on parle de grec stoïcien ou d’anglais journalistique 19. Les Juifs d’Alexandrie formaient une communauté partageant des intérêts spécifiques, dont certains étaient reflétés dans le vocabulaire et le style de la Septante; les Juifs d’Alexandrie partageaient la foi dans le Dieu d’Israël et pratiquaient des coutumes juives; il était inévitable que leur vocabulaire reflète ces centres d’intérêt communs. Un traducteur qui écrit dans une langue apprise, même s’il la maîtrise parfaitement, n’évite pas çà et là l’usage de tours qui lui viennent de sa langue maternelle; sa langue de traduction est véritablement du grec, mais dans ce grec se produisent des interférences sémitiques, qui ne créent pas une nouvelle langue, mais définissent simplement un style. Rien d’étonnant donc qu’il y ait une influence sémitique sur le lexique de la Septante, qui se traduit de trois manières. 1 / Certains mots hébreux ou araméens sont purement et simplement translittérés en grec, ainsi le mot πάσχα. 2 / Plus intéressante sur le plan de la signification linguistique est le phénomène que Silva appelle loan translation, c’est-à-dire le phénomène qui consiste à traduire les parties constitutives d’une séquence plutôt que de rendre la signification d’ensemble de la séquence; par exemple, la séquence ‫נשא פנים‬, littéralement « lever les faces de quelqu’un », est une 17. J. Vergote, « Grec biblique » DBSuppl. 3 (1938), 1321-1369. 18. G. H. R. Horsley, « The Fiction of « Jewish Greek » » in: G. H. R. Horsley (éd.), New Documents Illustrating Early Christianity, vol. 5, North Ryde 1989, 5-40. 19. M. Silva, « Bilingualism and the Character of Palestinian Greek » Biblica 61 (1980), 189-219; M. Silva, Biblical Words and Their Meaning. An Introduction to Lexical Semantics, Grand Rapids, MI 1994, chapitres 4-6; K. H. Jobes, M. Silva, Invitation to the Septuagint, Grand Rapids, MI 2000, 105-110, 183-189. 2. Les débats et les notions-clés

281

19. La lexicographie de la Septante

expression métaphorique qui signifie soit « traiter favorablement quelqu’un » (Gn 32,21) soit « manifester du favoritisme ou de la partialité à l’égard de quelqu’un » (Lv 19,15); elle est assez souvent rendue par λαμβάνω πρόσωπόν τινος, littéralement « prendre [en considération] la face de quelqu’un »; selon Silva, cette traduction devait paraître étrange ou peu claire aux lecteurs grecs; cette affirmation est toutefois discutée dans la note sur Lv 19,15 de « La Bible d’Alexandrie » 20. 3 / Du point de vue linguistique, l’influence la plus importante sur le lexique est le semantic borrowing, l’emprunt sémantique; Silva donne l’exemple de ‫ תורה‬traduit par νόμος, alors que le sens du mot hébreu ne signifie pas seulement « loi », mais aussi « instruction »; le lecteur grec de la Septante pouvait-il trouver étranges les passages où νόμος a ce sens? C’est possible, mais, comme l’a montré Laurent Monsengwo Pasinya, le mot grec n’avait pas un sens étroitement juridique, mais débouchait sur la sphère religieuse 21. Silva cite aussi l’exemple de δόξα pour rendre ‫כבוד‬, « poids, importance, gloire »; il signale que le sens habituel en grec classique est soit l’opinion qu’on a sur les autres ou les choses ou l’opinion que les autres ont sur vous, c’est-à-dire la réputation; en fait, dès l’époque classique, le mot a le sens de bonne réputation, gloire, comme le montre la consultation du LSJ. Récemment, la distinction entre langue et style a été réaffirmée avec force par Alexis Léonas 22: l’aspect inhabituel du langage de la Septante ne relève pas d’un système linguistique étranger au grec; il s’agit plutôt d’une stylistique particulière, que l’auteur appelle le registre hiératique.

2.2 Les notions-clés de la lexicographie de la Septante On le constate à la lecture des lignes qui précèdent: le débat autour du grec biblique et du judéo-grec a permis de faire apparaître et de définir plusieurs notions-clés en matière de lexicographie de la Septante, notamment la notion de sémitisme. En fait, l’influence sémitique sur la Septante est un phénomène plus complexe qu’on ne le dit parfois. Certes, les sémitismes jouent un rôle essentiel et il importe de les décrire finement en distinguant: 1 / les translittérations, 2 / les sémitismes au sens strict, 3 / les traductions stéréotypées et 4 / les homophonies. Mais il arrive que la langue source exerce son influence sur la langue cible de manière plus indirecte: pour rendre compte des réalités hébraïques, les traducteurs font jouer diverses possibilités langagières de la langue cible, ce qui les conduit à des innovations lexicales, et cela de trois façons: 5 / ils créent des néologismes, conformément aux possibilités du grec (et des langues en général) de créer des mots nouveaux, 6 / ils emploient des mots grecs d’une manière légèrement décalée, 7 / enfin ils élaborent des traductions multiples d’un même mot hébreu. Cependant, et quelles que soient les contraintes bien réelles exercées par la langue source, les traducteurs proposent en règle générale des correspondants aux mots hébreux qui sont satisfaisants du point de vue grec. Toutefois, certains de ces 20. P. Harlé / D. Pralon, Le Lévitique (La Bible d’Alexandrie 3), Paris 1988, 166-167. 21. L. Monsengwo Pasinya, La Notion de nomos dans le Pentateuque grec, Rome 1973. 22. A. Léonas, Recherches sur le langage de la Septante, Fribourg/Göttingen 2005; voir aussi du même auteur L’aube des traducteurs. De l’hébreu au grec: traducteurs et lecteurs de la Bible des Septante (IIIe s. av. J.-C. – IVe s. ap. J.-C.), Paris 2007.

282

2. Les débats et les notions-clés

19. La lexicographie de la Septante

mots peuvent apparaître comme accordant trop à la langue cible: on peut les appeler 8 / les hellénismes Enfin, en cas de mots difficiles dans la langue source, les traducteurs proposent des équivalents dans la langue cible, qui relèvent de ce que M. Flashar appelait dès 1912 les Verlegenheitsübersetzungen 23, littéralement les « traductions de perplexité », ou mieux 9 / les traductions conjecturales.

2.2.1 Les translittérations Les translittérations: certains mots hébreux sont écrits dans la Septante en lettres grecques; tout se passe donc comme si les traducteurs renonçaient à traduire. Pourquoi? Dans sa présentation des translittérations de la Septante et de Théodotion, Tov établit deux listes: celle des mots qui étaient probablement inconnus des traducteurs et celle des mots inconnus ou difficiles qu’ils ont compris comme des noms propres 24. Si cette deuxième explication est convaincante, parce que l’identification d’un nom propre est à peu près sûre, la première l’est moins, dans la mesure où nous n’avons pas de critère qui permette d’établir l’ignorance des traducteurs en matière de mots hébreux. On pourrait être tenté de penser que les hapax legomena bibliques constituent un de ces critères, puisque les hapax sont les mots inconnus par excellence; mais, si cela était vrai, il faudrait que les traducteurs les translittèrent systématiquement; or c’est loin d’être le cas: la Septante comprend trente-quatre des 140 hapax verbaux de la Bible de la même manière que les Targums, la Peshitta et la Vulgate; elle traduit vingt et un autres hapax verbaux de la même manière que deux de ces trois versions 25. Bref, les versions anciennes attestent une manière commune de comprendre les hapax, qui est à l’opposé de l’ignorance. Cependant, on peut songer à d’autres explications que celles avancées par Tov, de type littéraire. Les traducteurs auraient voulu respecter le caractère technique de certains termes bibliques, qui n’avaient pas d’équivalents grecs, dans le cas des unités de mesure, de poids et de monnaie, pour certaines plantes, dans le cas des objets du culte et des réalités religieuses et institutionnelles juives. Ou bien ils auraient pu vouloir garder des traces d’hébreu dans le grec et orner leur traduction avec quelques mots d’origine, lui donnant en quelque sorte un cachet d’authenticité biblique. Ce que l’on peut dire, c’est que, en règle générale, les noms propres d’hommes et de lieux sont translittérés. Il arrive cependant que les traducteurs considèrent comme un nom propre géographique ce qui est dans le texte massorétique un nom commun, ainsi χαβραθά en Gn 35,16 et 48,7 26, ou βηρώθ en Dt 10,6 (TM ‫בארת‬, « puits » au pluriel). Pour les translittérations de noms communs, plusieurs catégories peuvent être

23. M. Flashar, « Exegetische Studien zum Septuagintapsalter » ZAW 32 (1912), 81-116, 161-189, 241-268, en particulier 94-95. 24. E. Tov, « Transliterations of Hebrew Words in the Greek Versions. A Further Characteristic of the Kaige-Th Revision? » Textus 8 (1973), 78-92 (réimpression in: Tov, Greek and Hebrew Bible, 501-512). 25. Sur la question des hapax legomena dans la Bible et la manière dont elle a été traitée à travers les âges, voir F. E. Greenspahn, Hapax Legomena in Biblical Hebrew. A Study of the Phenomenon and Its Treatment Since Antiquity with Special References to Verba Forms, Chico, CA 1984. Le cas de la Septante est rapidement traité (56-60). 26. Voir M. Harl, La Genèse (La Bible d’Alexandrie 1), Paris 1986, 253. On peut aussi interpréter χαβραθά comme un terme technique de distance. 2. Les débats et les notions-clés

283

19. La lexicographie de la Septante

distinguées: dans le lexique des plantes, il y a σαβέκ (TM ‫)סבך‬, la plante du sacrifice d’Abraham (Gn 22,13); dans le domaine des unités de mesure, il y a, dans le cas des liquides, ἵν (TM ‫)חין‬, qui équivaut à peu près à 6,5 litres, et οἰφί (TM ‫)איפה‬, soit environ quarante-cinq litres, dans le cas des solides, γόμορ (TM ‫)חמר‬, qui est parfois traduit sous la forme κόρος ; dans le domaine de l’alimentation, la nourriture miraculeuse au désert est appelée μάν ou μάννα ; la boisson alcoolisée ‫ שכר‬devient, d’après l’araméen, σίκερα; dans le domaine des réalités religieuses, il y a les théraphins (θεραφίν), les chérubins, χερουβίμ (TM ‫ )כרובים‬et la fête de Pâque, πάσχα, qui provient de l’hébreu ‫ פסח‬via l’araméen ‫ פסחא‬27. Il faut toutefois faire remarquer qu’il y a une relative hellénisation de ces mots translittérés: dans les manuscrits grecs, ils sont normalement accentués; ils sont précédés de l’article neutre (τὸ ἵν, τὸ οἰφί, τὸ γόμορ, τὸ μάν, τὸ μάννα, τὸ σίκερα, τὰ χερουβίμ, τὸ πάσχα); certains d’entre eux entrent dans le système des déclinaisons, comme le nom propre Μωυσῆς (voir plus bas), le quasi-nom propre λευείτης, l’un des mots qui désigne le prosélyte, ὁ γειώρας, ou encore le nom du sabbat, τὸ σάββατον; ils donnent lieu à des dérivations, ainsi, sur σάββατον, le verbe σαββατίζειν. Thackeray donne la liste de ce qu’il appelle les « mots sémitiques hellénisés 28 ». La plupart d’entre eux sont attestés avant la Septante, dès l’époque classique ou dans les papyrus: ἀρραβών, βακχούρια, βάρις au sens de tour, βῖκος, βύσσος (attesté sous la forme adjectivale βύσσινος dès l’époque classique), γαζαρηνός, γειώρας, θῖβις (papyri du IIIe siècle avant notre ère), κάδος, κασία, κιννάμωμον, κινύρα, κόρος, κύμινον, λίβανος, μνᾶ, νάβλα, νίτρον (papyri du IIIe siècle avant notre ère), σάκκος, σαμβύκη, σάπφειρος, σίκλος, συκάμινος. Deux de ces mots seulement apparaissent pour la première fois dans la Septante: σάββατον et χαύων (Jr 7,18). Plusieurs translittérations sont faites, non sur l’hébreu, mais sur l’araméen: γαζαρηνός (Dn 2,27, d’après ‫)גזר‬, γειώρας (Ex 12,19, d’après ‫)גיורא‬, μάννα (mais non μάν, qui est translittéré sur l’hébreu ‫)מן‬, πάσχα (d’après ‫)פסחא‬, πάταχρα / πάταχρον (Es 8,21 et 37,38; d’après ‫)פתכר‬, σάββατα, σίκερα 29. Pour décrire ce type de translittération, on utilise parfois le terme d’aramaïsmes. Deux translittérations sont faites en référence à l’égyptien: Μωυσῆς, qui dérive de deux mots égyptiens, mw, qui désigne l’eau, et hsy, qui désigne l’état du mort glorifié à la suite d’une immersion comme dans le mythe d’Osiris; et θῖβις, qui désigne le coffre de Moïse et qui correspond à deux mots égyptiens de même racine, signifiant l’un « caisse, coffre », l’autre « coffre, sarcophage »; θῖβις est attesté dans trois papyri des IIIe-IIe siècles, au sens de « coffre » plutôt que de « corbeille » 30.

27. Voir les listes données par Thackeray (Grammar, vol. I, 32-34). 28. Thackeray, Grammar, vol. I, 34-36. 29. Sur πάσχα et σάββατα, voir A. Pelletier, « Pour une histoire des noms grecs du Sabbat et de la Pâque » Comptes rendus de l’Académie des Inscriptions et Belles-Lettres, 115 (1971), 71-83; « Sabbata, transcription grecque de l’araméen » VT 22 (1972), 436-447. Sur l’ensemble de ces mots, voir J. Joosten, « The Septuagint as a Source of Information on Egyptian Aramaic in the Hellenistic Period » in: H. Gzella / M. L. Folmer (éd.), Aramaic in its Historical and Linguistic Setting, Wiesbaden 2008, 93-105; il montre aussi que la translittération des mots hébreux a été influencée par l’araméen. 30. Y. Koenig, « Quelques « égyptianismes » de la Septante » BIFAO 98 (1998), 223-233. Les traduc-

284

2. Les débats et les notions-clés

19. La lexicographie de la Septante

Deux phénomènes concernant les translittérations mériteraient d’être mieux expliqués qu’ils ne le sont à l’heure actuelle. Le premier, c’est que certains mots sont tantôt translittérés, tantôt traduits: αἰλάμ/ναός et d’autres mots; βαμά/ἄλσος et d’autres mots; γεδδούρ (TM ‫)גדור‬/λῃστήριον et d’autres mots; γόμορ (TM ‫)חמר‬/κόρος; δαβείρ/ναός; ἐφούδ (TM ‫)אפוד‬/ἐπωμίς; θεραφίν/εἴδωλον et d’autres mots; μάναα (‫)מנחה‬/δῶρον ou θυσία; ναγέβ (‫)נגב‬/ἔρημος et d’autres mots; ναζίρ ou ναζιραῖος/ εὐξάμενος ou un autre mot; νέβελ (‫)נבל‬/ἀγγεῖον ou ἀσκός; σαβαώθ/παντοκράτωρ ou τῶν δυνάμεων; σεφηλά/πεδινή; σίκερα/μέθυσμα ou μέθη. On peut être tenté d’expliquer ce phénomène au fait que les livres bibliques ont été traduits par des traducteurs différents. Cette explication est pertinente dans bon nombre de cas, mais pas toujours. Par exemple, si l’on prend le couple γόμορ/κόρος, l’explication fonctionne pour le Pentateuque: le premier mot est attesté six fois en Exode et le second seulement en Lv 27,16 et Nb 11,32; mais elle ne vaut pas pour Ézéchiel, où les deux mots sont attestés à deux versets de distance (γόμορ en Ez 45,11; κόρος en Ez 45,13); le traducteur aurait-il voulu reproduire les deux traductions du Pentateuque? Le second phénomène qui mériterait une explication, c’est la répartition inégale des translittérations à travers les livres de la Septante: comme l’a remarqué Thackeray, elles sont peu nombreuses dans le Pentateuque, Ésaïe, la première moitié de Jérémie, les douze Prophètes, totalement absentes de la deuxième moitié d’Ézéchiel, de Psaumes, Proverbes, Job, des livres classés par les Sages parmi les Écrits, mais elles sont très fréquentes en Juges, 3 et 4Règnes, 1 et 2Paralipomènes, Esdras. Ce phénomène s’explique-t-il par la chronologie des traductions, par des principes de traduction, par une combinaison de ces deux facteurs, par d’autres raisons? Pour apprécier à leur juste mesure l’importance des translittérations dans la Septante, il faut souligner que le phénomène inverse se produit parfois: la Septante traduit là où les exégètes modernes procèdent à des translittérations. C’est le cas de certains noms propres de personne, comme en Gn 3,20, où Ève est appelée Ζωή, « Vie »; ainsi que de certains noms géographiques, comme en Gn 11,28, où Ur, la cité d’Abraham, devient χώρα, « territoire »; une trentaine de noms géographiques de Genèse font l’objet de traductions 31. C’est le cas aussi de mots difficiles: en Jon 4,6, κολόκυνθα correspond au mystérieux ‫ קיקיון‬du texte massorétique; en Es 28,10, la Septante propose une traduction (« reçois épreuve sur épreuve, espérance sur espérance, encore un peu, encore un peu ») là où la plupart des exégètes modernes préfèrent translittérer le texte massorétique (‫)כי צו לצו צו לצו קו לקו קו לקו זעיר שם זעיר שם‬, dont la traduction littérale est « car ordre en vue d’ordre ordre en vue d’ordre, règle en vue de règle règle en vue de règle, un peu là un peu là ». Dans tous ces cas, les traducteurs considèrent que le texte de la Bible est plus traduisible que nous le considérons aujourd’hui. Ces exemples de traduction contre-balancent-ils les exemples de translittération? Une recherche reste à mener sur ce point.

teurs ont sans doute voulu évoquer le mythe d’Osiris, enfermé dans un coffre par son frère Seth et dérivant au fil de l’eau jusqu’à Byblos. 31. Harl, Genèse, 322-324. 2. Les débats et les notions-clés

285

19. La lexicographie de la Septante

2.2.2 Les sémitismes Les sémitismes, où l’influence de l’hébreu ne concerne pas la forme des mots grecs comme dans les translittérations, mais porte sur leur signification: le mot de la Septante prend peu ou prou le sens du mot biblique correspondant. Les sémitismes se subdivisent en aramaïsmes, lorsque des mots et des tours grecs reflètent une influence de la langue araméenne ou encore lorsque des mots hébreux ont été compris comme des mots araméens, et en hébraïsmes, lorsque l’influence est celle de l’hébreu biblique et postbiblique. Les aramaïsmes restent encore mal connus, mais le Two-Way Index de Muraoka permettra d’avancer dans la connaissance de ce phénomène; des expressions comme « pécher devant Dieu », « regarder devant Dieu », « s’approcher devant Dieu », où « devant » est un ajout qui exprimer le respect pour la personne divine, sont des aramaïsmes; elles sont présentes également dans les targums araméens, qui les généralisent, et elles reflètent l’influence du langage de la cour de l’époque perse, où l’on ne parle pas au roi, mais devant le roi 32. En ce qui concerne les hébraïsmes, voilà plus de cent ans que Henry Swete a dressé une liste d’une vingtaine de mots qui reçoivent dans la Septante de nouvelles significations dues au modèle hébreu, comme ἅγιον, ἁμαρτωλός, ἀρετή, ἀφόρισμα, ἄφρων; en fait, dans tous ces cas, il ne s’agit pas tant de nouvelles significations que de nouveaux contextes d’emploi; et le sens se laisse assez facilement déduire du contexte 33. Le même savant donne une liste de huit expressions faites de deux mots, qui traduisent littéralement les mots hébreux correspondants au lieu de donner un équivalent acceptable en grec, du type λαμβάνειν πρόσωπον 34. On reconnaît ici les loan translations dont parle Silva. Là encore, les exemples ne sont tous convaincants. Thackeray, qui renvoie à Swete, énumère une série d’inflexions de sens de mots ou d’expressions grecs sous l’influence de l’hébreu 35. C’est ainsi que διδόναι, en plus du sens de « donner », prend le sens de « placer » normalement exprimé par τιθέναι, sous l’influence de ‫ ;נתן‬ce phénomène apparaît en Dt 28,1 et se multiplie dans les livres postérieurs. Le mot ἡμέραι, traduction automatique de ‫ימים‬, signifie parfois « années », comme l’hébreu, ainsi en Ex 13,10, où il est question d’un rituel qu’il faut observer de « jours en jours », en fait d’années en années; les traducteurs de BA 2 ont traduit littéralement, à juste titre: dans le contexte, le lecteur comprend qu’il s’agit d’une longue suite de jours. Sauf dans l’Hexateuque, Ésaïe et 1–2Paralipomènes, le mot υἱός traduit automatiquement ‫בן‬, même dans les passages où ce mot dénote l’âge (« fils de cent ans » pour « âgé de cent ans », en Gn 11,10, qui est le seul contre-exemple du Pentateuque) ou des qualités (« fils étranger » pour « étranger »). Il y a aussi les expressions « physiognomiques », qui se réfèrent à une partie du corps humain, l’oreille, l’œil, la face, la main, la bouche. Ces expressions sont rendues littéralement dans certaines parties de la Septante, mais la plupart de ces traductions sont acceptables en 32. J. Joosten, « L’agir humain devant Dieu. Remarques sur une tournure remarquable de la Septante » RB 113 (2006), 5-17. 33. H. B. Swete, An Introduction to the Old Testament in Greek, Cambridge 1902 (réimpression: New York, NY 1968), 307. 34. Voir aussi C. Dogniez, « The Greek Renderings of Hebrew Idiomatic Expressions and their Treatment in the Septuagint Lexica » JNSL 28 (2002), 1-17 (étude de προστίθημι πρὸς τὸν λαόν, λαλέω εἰς / ἐπὶ τὴν καρδίαν τινός, λαμβάνω τὸ πρόσωπον, ἐπιστρέφω / ἀποστρέφω τὴν αἰχμαλωσίαν). 35. Thackeray, Grammar, vol. I, 39-46.

286

2. Les débats et les notions-clés

19. La lexicographie de la Septante

grec, comme « dévoiler l’oreille de quelqu’un » (ἀποκαλύπτειν τὸ οὖς τινος) pour « révéler à quelqu’un » (1Règnes 9,15) ou « interroger la bouche de quelqu’un » (Gn 24,57); d’autres le sont moins, ainsi « admirer la face de quelqu’un » (Gn 19,21) pour « tenir compte de quelqu’un », qui peut toutefois s’appuyer sur le sens d’« honorer » attesté en grec classique pour θαυμάζειν; elles sont toutefois beaucoup plus nombreuses dans la Septante que dans les autres documents: l’hébreu leur a donné une large circulation. Les pages de Thackeray sont précieuses aussi en ce qu’elles montrent que des tours souvent considérés comme des sémitismes ne le sont pas: ainsi, διδόναι + infinitif au sens de « permettre » est classique; ἀριθμῷ, « en petit nombre », a un parallèle en Hérodote 6,58; ῥῆμα traduit ‫דבר‬, qui signifie à la fois « parole » et « chose » (dont on parle), mais il se peut que le mot ait eu le sens de « chose » en grec parlé, tout comme λόγος. À la lecture de ces exemples, on remarque que les sémitismes sont moins nombreux dans le Pentateuque et Josué que dans Juges et 1–4Règnes et que certains types de textes de la Septante contiennent plus de sémitismes que d’autres. La présence de sémitismes en grand nombre constitue un indice d’une traduction littéraliste 36. Un hébraïsme de type particulier est constitué par les exemples où le grec reflète, non l’hébreu biblique, mais l’hébreu postbiblique: c’est le cas d’ἀναβάλλομαι traduisant en Ps 77/78,21 et 88/89,39 ‫ ;התעבר‬en hébreu biblique, ce verbe signifie « devenir furieux », mais, en hébreu postbiblique, il prend le sens de « négliger » ou de « différer »; c’est ce dernier sens qui est retenu dans les Psaumes 37. A ces exemples de sémitismes, on peut en joindre d’autres, qui illustrent le fait que certains sémitismes prennent place, facilement ou assez facilement, dans le lexique grec. Le mot ἄρτος est l’équivalent régulier de ‫לחם‬, qui désigne soit le pain, soit le repas, la nourriture; en grec, ἄρτος a exclusivement le sens de pain; mais les passages de la Septante où ἄρτος renvoie à la nourriture ne constituent pas une vraie difficulté, dans la mesure où il s’agit d’une simple extension de sens. L’emploi en tête des serments de sens négatif de la conjonction de subordination εἰ, qui décalque l’hébreu ‫אם‬, est sans doute plus compliqué à admettre; mais le grec connaît de nombreuses constructions elliptiques; dans cette tournure l’ellipse est la suivante: « [malheur à moi] si je fais telle chose », autrement dit « je jure de ne pas faire telle chose » 38. Un autre exemple délicat est constitué par le mot ἀνάθεμα. Dans la religion grecque, les ἀναθέματα sont les objets commémoratifs déposés dans un temple ou sur un territoire réservés aux dieux et, de ce fait, consacrés à la divinité. Dans la Septante, l’ἀνάθεμα traduit ‫חרם‬, qui n’est pas identique au rite grec: loin de se traduire par la mise en réserve et la conservation de ce qui est consacré, le rite aboutit à la destruction. La Septante donne donc un sens nouveau à un mot de la religion traditionnelle. Mais ce sens se laisse assez facilement déduire du contexte d’emploi du mot. Autre exemple délicat concernant une séquence: l’investiture des prêtres est exprimée en hébreu par ‫מלא‬ ‫ידם‬, « remplir les mains » (Ex 28,41), qui est rendu littéralement par la Septante (ἐμπι36. Thackeray, Grammar, vol. I, 13. 37. J. Joosten, « Source-Language Oriented Remarks on the Lexicography of the Greek Versions of the Bible » ETL 81 (2005), 152-164. La suggestion de Jan Joosten n’est pas retenue par le Lexicon de T. Muraoka. 38. Harl, Genèse, 76. 2. Les débats et les notions-clés

287

19. La lexicographie de la Septante

πλάναι τὰς χεῖρας). Il paraît difficile qu’un lecteur grec puisse donner le sens d’investiture à cette expression, sauf s’il connaît l’hébreu. Cependant, en Ex 32,29, le lecteur comprend que ce remplissage des mains n’est pas un acte ordinaire. Moïse vient d’ordonner aux lévites de tuer les adorateurs du veau d’or. Une fois le châtiment accompli, il leur déclare: « vous avec rempli vos mains pour Seigneur » (πληροῦν τὰς χεῖρας). Le remplissage des mains est donc une métaphore qui désigne l’acceptation d’une responsabilité particulière au nom de Dieu. En Nb 7,88, la Septante affirme que Moïse oint Aaron et lui remplit les mains: le lecteur comprend que le remplissage des mains est synonyme de l’onction sacerdotale. De plus en Ex 29,9, la Septante emploie l’expression τελειοῦν τὰς χεῖρας, « accomplir les mains », c’est-à-dire les consacrer. Il arrive cependant que certains décalques de l’hébreu aboutissent à des expressions à peu près incompréhensibles. En Nb 1,2; 4,22 et 26,2, que signifie l’ordre divin: « Prenez le commencement (ἀρχή/‫ )ראש‬de toute la communauté des fils d’Israël? ». On constate qu’à cette expression obscure, les traducteurs ont substitué en Nb 4,2; 31,26 et 49 la traduction beaucoup plus claire « prendre la somme totale », κεφάλαιον ; cette traduction est même habile, dans la mesure où κεφάλαιον est dérivé de κεφαλή, le nom grec de la tête, qui traduit normalement ‫ראש‬. De la sorte, une variation de traduction a permis de rendre compréhensible ce qui ne l’était pas. Ce procédé n’est pas attesté de la même façon dans le cas du mot εἰρήνη, qui est l’équivalent fixe de ‫שלום‬ en Juges (texte du Vaticanus) et 1–4Règnes, dans des tours qui décalquent des expressions hébraïques de salutation, par lesquelles on demande des nouvelles. En 2Règnes 11,7, David interroge Ouri « pour la paix de Joab et pour la paix du peuple et pour la paix de la guerre »; pour le traducteur bilingue, David demande des nouvelles de Joab, du peuple et de la guerre; « la paix de Joab » et « la paix du peuple » devaient paraître un peu mystérieuses à un lecteur grec, et « la paix de la guerre », franchement énigmatique. On peut se demander d’ailleurs si le mot grec a encore un sens au niveau du grec et s’il ne fonctionne pas plutôt comme le symbole du mot hébreu 39. Ce qu’on note encore, c’est que ce sémitisme n’a pas été admis par les traducteurs du Pentateuque, qui ne rendent pas littéralement ‫שלום‬, mais emploient des verbes comme ἀσπάζεθαι ou ὑγιαίνειν. Au total, on peut rejoindre les conclusions de Chaïm Rabin, qui, à propos des sémitismes, a mis en avant la notion de « tolérance sémantique »: même si les mots sont employés de manière nouvelle, leur sens nouveau est en général compris grâce au contexte. De plus l’acceptabilité de la Septante augmente au fur et à mesure qu’augmente la quantité des textes traduits: il se crée une sorte de sous-langage (il vaudrait mieux dire une stylistique) qui se comprend de mieux en mieux 40. On peut traiter les égyptianismes en même temps que les sémitismes. Ce mot est peu présent dans les études sur la Septante parce que le phénomène est limité. Il a été employé plus haut à propos de deux translittérations qui dénotent une influence du vocabulaire égyptien sur la Septante: Μωυσῆς et θῖβις. Mais l’influence égyptienne explique aussi certains choix de traduction de la Septante, comme « marais » (ἕλος) au lieu de « joncs » du texte massorétique en Ex 2,3; le terme « marais » convient au pay39. Voir E. Tov, « Three dimensions of Words in the Septuagint » RB 83 (1976), 529-544 (réimpression in: Tov, Greek and Hebrew Bible, 85-94, en particulier 88-89). 40. C. Rabin, « The Translation Process and the Character of the Septuagint » VT 11 (1961), 201-221.

288

2. Les débats et les notions-clés

19. La lexicographie de la Septante

sage du delta du Nil; la Septante atteste donc d’un phénomène d’acculturation, qui permet de rendre compte de la plus grande fréquence de certaines expressions dans la Septante en comparaison du texte massorétique, ainsi « par un bras levé » en Ex 6,1 et 6 (Ex 6,6 seul en TM), qui renvoie à un geste de type magique dans la culture égyptienne 41.

2.2.3 Les traductions stéréotypées ou stéréotypes La notion de stéréotype semble être apparue à propos de la Septante dès 1912 chez Flashar 42. Mais elle n’a pris que peu à peu de l’importance, notamment grâce à Tov 43. On désigne par ce terme les deux phénomènes voisins qui relèvent, chez les traducteurs, de la volonté de traduire de manière cohérente. 1 / Un même mot hébreu tend à être rendu par un même mot grec. Par exemple le mot ‫ ברית‬est traduit par διαθήκη plus de 260 fois; il n’y a que trois exceptions (Gn 14,13; 1Règnes 11,11; 2Paralipomènes 16,13) 44. On peut encore citer ἀδελφός/‫אח‬, θεός/‫אלהים‬, κύριος/‫יהוה‬, νόμος/‫תורה‬. /2 Une même famille de mots hébreux tend à être rendue par une même famille de mots grecs. Par exemple, les mots de la famille de ‫ קדש‬sont rendus par ἅγιος et ἁγιάζειν; ceux de la famille de ‫טוב‬, par ἀγαθός et ἀγαθύνειν. Ces traductions stéréotypées aboutissent à des sémitismes sémantiques. C’est le cas d’ἀγαθύνειν, qui veut dire normalement « faire du bien à quelqu’un »; mais, en Rt 3,16, Booz « mangea, but et son cœur ἠγαθύνθη », littéralement « fut rendu bon »; l’hébreu signifie en fait « se réjouit »; le grec n’exclut pas cette signification, mais elle n’a rien d’évident, même dans le contexte. L’exemple de κοίτη est plus difficile encore; en grec, ce mot désigne la couche et, par euphémisme, les relations sexuelles; mais ces sens ne conviennent pas en Nb 5,20, où il est dit, à propos de la femme adultère, qu’un homme autre que son mari lui a « donné sa κοίτη »; le mot grec est la traduction stéréotypée de ‫ שכב‬et de ses dérivés nominaux. Quelle est cette couche que l’homme donne à la femme adultère? En Lv 15,16, il est question de « couche de semence » (κοίτη σπέρματος/‫)שכבת זרע‬, qu’on comprend comme la semence qui est répandue sur la couche, « les pertes séminales », ou la semence émise pendant les relations sexuelles, « épanchement de semence », « sperme »; en Nb 5,20, le mot κοίτη a le sens de sperme à lui tout seul. Certes une telle signification ne saute pas aux yeux, mais le texte massorétique lui-même n’a pas un sens évident, au point qu’on a imaginé l’existence de deux racines ‫שכב‬, l’une signifiant « coucher », l’autre « s’épancher » 45. Un dernier exemple de traduction stéréotypée aboutissant à un sémitisme, pratiquement contraire à la langue grecque: le 41. Koenig, « Quelques ‹ égyptianismes › »; voir aussi M. Görg, « Die Septuaginta im Kontext spätägyptischer Kultur-Beispiele lokaler Inspiration bei der Übersetzungsarbeit am Pentateuch » in: H.-J. Fabry / U. Offerhaus (éd.), Im Brennpunkt: Die Septuaginta. Studien zur Entstehung und Bedeutung der Griechischen Bibel, Stuttgart/Berlin/Köln 2001, 115-130. 42. M. Flashar, « Exegetische Studien zum Septuagintapsalter » ZAW 32 (1912), 81-116, 161-189, 241-268, en particulier 105 (stereotypen Übersetzungen) 43. E. Tov, The Text-Critical Use of the Septuagint in Biblical Research, Jérusalem 1981, 54-57. 44. Sur cette équivalence, voir A. Schenker, « Διαθήκη pour ‫ברית‬. L’option de traduction de la LXX à la double lumière du droit successoral de l’Égypte ptolémaïque et du livre de la Genèse » in: J.-M. Auwers / A. Wénin (éd.), Lectures et relectures de la Bible. Festschrift P.-M. Bogaert, Leuven 1999, 126-131. 45. J. Barr, Comparative Philology and the Text of Old Testament, Oxford 1968, 137. 2. Les débats et les notions-clés

289

19. La lexicographie de la Septante

mot ψυχή traduit presque toujours ‫ ;נפש‬le mot hébreu désigne l’âme, la personne, la vie ou peut être l’équivalent du pronom personnel; ces emplois sont attestés pour ψυχή dans la langue grecque antérieure à la Septante et ne peuvent surprendre un lecteur grec; en revanche, en Nb 5,2, ψυχή désigne la personne morte qui rend impur celui qui est en contact avec elle; le sens de personne morte n’est pas possible en grec et apparaît comme un sémitisme presque incongru; ce qui atténue toutefois cette conclusion, c’est que le lecteur grec connaît Lv 21,11, qui offre l’expression ἐπὶ πάσῃ ψυχῇ τετελευτηκυίᾳ, littéralement « au contact de toute âme morte »: par ellipse, l’âme peut désigner le cadavre.

2.2.4 Les homophonies Dans les pages de sa grammaire où il réfléchit sur l’élément sémitique dans le grec de la Septante, Thackeray a attiré l’attention sur une tendance qu’on peut observer dans les livres qui suivent le Pentateuque et Josué et qui consiste, pour les traducteurs, à préférer parmi les mots grecs ceux qui ont des sonorités semblables à celles des mots hébreux 46. Ici l’influence de l’hébreu sur le grec ne concerne pas la signification, comme dans les sémitismes, mais uniquement les sonorités, autrement dits les mots en tant que phonèmes. Par exemple, les mots ἀλαλάζειν, ἀλαλαγμός, ὀλολύζειν, ὀλυλυγμός, « hurler, hurlement », sont homophones de ‫ילל‬, « crier », au hiphil ‫;יללה‬ or, ces mots se correspondent dans les Prophètes; aussi bien les mots grecs que les mots hébreux sont des onomatopées. Thackeray propose une liste d’une quarantaine de mots homophones, dont il dit qu’elle est sans aucun doute susceptible d’être étendue. De fait, les savants postérieurs à Thackeray ont élargi le nombre des mots homophones 47. La question est de savoir si ces repérages sont pertinents. Tov affirme que la reconnaissance d’une homophonie suppose que le sens des mots hébreu et grec homophones soit différent; car si les mots hébreu et grec ont le même sens, l’homophonie peut être accidentelle, comme par exemple dans le cas de ‫ עולה‬et de ὁλοκαύτωμα, qui désignent une offrande entièrement brûlée: le mot grec est un excellent équivalent du mot hébreu et il a été choisi pour cette raison, et non pour des raisons d’homophonie. Tov ajoute que, dans le cas d’hapax legomena, l’homophonie peut relever non d’un phénomène littéraire, mais d’une technique conjecturale de traduction 48. Il s’interroge par exemple sur l’homophonie ‫תוך‬/τόκος, « oppression/usure » (Ps 54/55,12 et 71/ 72,14), parce que, dans le contexte des deux Psaumes, le mot grec peut être un bon équivalent du mot hébreu. Dans le même esprit, Barr a émis des doutes sur beaucoup d’exemples d’homophonies et ne retient qu’un petit nombre de cas possibles 49. On peut cependant se demander si ces réserves ne sont pas hypercritiques: le traducteur 46. Thackeray, Grammar, vol. I, 36-38. 47. Par exemple C. Fritsch, « Homoeophony in the Septuagint » Proceedings of the Sixth World Congress of Jewish Studies, Jérusalem 1977, 115-120 (liste de soixante-seize mots grecs ayant des consonances voisines de celles des mots hébreux). Voir aussi G. B. Caird, « Homophony in the Septuagint » in: R. Hamerton-Kelly / R. Scroggs (éd.), Jews, Greeks and Christians. Essays in Honor of W. D. Davies, Leyde 1976, 74-88; J. de Waard, « Homoeophony in the Septuagint » Biblica 62 (1981), 551-556. 48. E. Tov, « Loan-Words, Homophony, and Transliterations in the Septuagint » Biblica 60 (1979), 216-238 (réimpression in: Tov, Greek and Hebrew Bible, 165-182). 49. J. Barr, « Doubts about Homophoeophony in the Septuagint » Textus 12 (1985), 1-78.

290

2. Les débats et les notions-clés

19. La lexicographie de la Septante

peut choisir un mot grec parce qu’il est un bon équivalent d’un mot hébreu et aussi parce qu’il lui est homophone. Sans doute la question mériterait-elle d’être de nouveau examinée. Passons aux influences indirectes de la langue source sur la langue cible.

2.2.5 Les néologismes L’attention portée aux néologismes est ancienne, comme le montre l’Introduction de Swete, qui donne une liste de cinq mots « forgés pour exprimer des idées sémitiques »: ἀκροβυστία, ἀναθεματίζειν, ὁλοκαύτωμα, σκανδαλίζειν, σπλαγχνίζειν 50. Aucun signe conventionnel ne les indique dans le HR, mais celle-ci n’est pas un dictionnaire. En revanche, ils sont signalés par l’indication neol. dans le LEH, mais celle-ci figure en fin de notice et n’est pas facile à repérer. Il n’en va pas ainsi dans le GELS, où les néologismes sont indiqués par le signe conventionnel *, qui est particulièrement visible puisqu’il figure en fin de la première ligne. Les cinq mots de Swete sont suivis de l’astérisque (*) dans le GELS, qui précise cependant, à propos d’ὁλοκαύτωμα, que le verbe ὁλοκαυτέω figure chez Xénophon. Muraoka met ainsi en valeur un fait d’importance: les néologismes ne forcent pas la langue grecque, mais prennent appui sur le vocabulaire existant pour forger des mots nouveaux. C’est évidemment le cas du plus célèbre des néologismes de la Septante, θυσιαστήριον: ce mot a été forgé sur θυσία, qui désigne l’offrande apportée par les fidèles au Seigneur; le sens du mot ne fait aucune difficulté par le lecteur grec: c’est le lieu qui recueille les offrandes, l’« autel à offrandes »; en fabriquant ce mot dont la grécité est incontestable, les traducteurs distinguent l’autel du désert, qui est unique et réservé à l’unique Seigneur, du βωμός païen, qui est le mot adopté pour les multiples autels des multiples dieux. Trois des mots de la liste de Swete sont des néologismes conformes aux possibilités du lexique grec: ἀναθεματίζειν est fabriqué sur ἀνάθεμα, qui désigne l’offrande votive; ὁλοκαύτωμα sur ὁλοκαυτέω, qui indique l’action de brûler entièrement une victime; σπλαγχνίζειν sur σπλάγχνον, « entrailles, viscères ». Du point de vue sémantique, un fait mérite d’être souligné: tantôt le sens du néologisme va de soi au niveau de la langue grecque (ὁλοκαύτωμα); tantôt le contexte d’emploi permet de comprendre le sens: en 2M 6,8, σπλαγχνίζειν ne peut que signifier « manger les entrailles d’une bête sacrifiée »; tantôt, enfin, la détermination de la signification demande une vraie réflexion: c’est le cas d’ἀναθεματίζειν, dont le sens « maudire » repose sur une évolution sémantique du substantif ἀνάθεμα propre à la Septante; il désigne les êtres humains, les animaux, les objets qui, dans un contexte de victoire militaire, sont offerts à Dieu au point d’être détruits complètement et qui sont donc maudits. Les deux derniers cas sont apparemment plus difficiles: σκανδαλίζειν est formé sur σκάνδαλον qui est lui-même un néologisme de la Septante, tandis qu’ἀκροβυστία est composé de deux éléments, ἀκρο- qui désigne l’extrémité, et -βυστία, qui n’est pas connu par ailleurs. Mais σκάνδαλον figure dans le P. Cair. Zen. 608, 7 du IIIe siècle avant notre ère au sens de « piège » et l’élément σκανδαλ- est présent dans le composé σκανδάληθρον attesté chez Aristophane, qui désigne la tige qui sert de détente à un piège; il se pourrait bien que σκάνδαλον soit ancien en grec; seuls les hasards de la transmission des textes expliqueraient son absence dans la littérature classique. En ce qui concerne ἀκροβυστία, le contexte d’em50. Swete, Introduction, 307. 2. Les débats et les notions-clés

291

19. La lexicographie de la Septante

ploi permet de comprendre qu’il ne peut s’agir que du prépuce (voir Gn 17,11); de plus, ce mot fait écho à ἀκροποσθία, qui est le nom du prépuce chez Aristote; enfin un Grec ne pouvait pas ne pas rapprocher l’élément -βυστία du verbe βύω, « bourrer, remplir »: de fait le prépuce se situe sur la partie « remplie » du pénis. Il y a environ 850 néologismes dans la Septante sur un total de 9 500 mots, soit un peu plus d’un dixième du vocabulaire total. Le chiffre de 1 900 donné par Muraoka dans le GELS est erroné (voir plus haut). Ces néologismes mériteraient une étude systématique, qui est aujourd’hui possible grâce aux lexiques LEH et GELS. Elle fera sans doute apparaître qu’il y a plusieurs sortes de néologismes: 1 / certains d’entre eux sont propres à la Septante et constituent des hapax legomena, soit au sens strict (une seule occurrence uniquement dans la Septante), soit au sens large (plusieurs occurrences, toutes dans la Septante); 2 / ceux pour lesquels on a des raisons de penser qu’ils existaient en réalité avant la Septante, comme σκάνδαλον; 3 / ceux qui apparaissent après la Septante dans des documents où il ne peut y avoir d’influence directe de la Septante; 4 / ceux dont la formation et le sens ne posent aucun problème au niveau du grec, comme θυσιαστήριον; 5 / ceux qui sont typiques de la Septante, ce qui ne veut pas dire qu’ils ne réapparaissent pas ultérieurement. Le même mot peut apparaître dans diverses catégories. Dans le cas de Nombres, il y treize néologismes 51. On peut les répartir ainsi: 1/ ἐπαξονεῖν (Nb 1,18), καταρρεμβεύειν (Nb 32,13), λαμπηνικός (7,3), συγκατακληρονομεῖσθαι (32,30) sont des hapax absolus; 2 / αἰνιγματιστής, « faiseur d’énigmes », a de bonnes chances d’être antérieur à la Septante, même s’il n’est pas attesté; 3 / sont attestés après la Septante: αἰνιγματιστής; ἀκουσιάζεσθαι, « agir inintentionnellement » (15,28), qui réapparaît chez Eustathe de Thessalonique (Commentaire sur l’Iliade, volume 3, page 252, ligne 15); ὀπτάζεσθαι, « être vu » (14,14), attesté chez Syméon Logothète (Lettres 111, 18); 4 / ἀκουσιάζεσθαι (15,28); λαμπηνικός, « caractéristique de la λαμπήνη », mot classique de sens discuté, qui désigne peut-être un char d’apparat (7,3); ὀπτάζεσθαι (14,14); συγκατακληρονομεῖσθαι, « cohériter avec quelqu’un » (32,20); cinq verbes attestent la tendance de la koinè à préférer le verbe composé ou surcomposé au verbe simple ou composé: ἐκπορνεύειν, « se prostituer » (25,1), κατακληρονομεῖν, « hériter » (13,31); καταρρεμβεύειν, « faire tournoyer » (32,13); καταπρονομεύειν, « faire du butin » (21,1); προσοχθίζειν, « s’irriter » (21,5); 5 / Sont caractéristiques de Nombres: ἐπαξονεῖν, « mettre sur des tablettes » les noms des combattants hébreux pour les enregistrer (1,18) 52; ἀναθεματίζειν, « maudire » (18,34) (voir plus haut); ἐκσπερματίζειν, « faire aboutir la semence » (5,28) dans le cas de la femme soupçonnée à tort d’adultère 53.

2.2.6 Les mots grecs employés de manière légèrement décalée Ce phénomène est rarement décrit par les lexicographes de la Septante, sans doute parce qu’il est d’une grande finesse linguistique et qu’il peut susciter le reproche de 51. G. Dorival, Les Nombres (La Bible d’Alexandrie 4), Paris 1994, 186; le verbe ἐπικαταλαμβάνεσθαι, signalé comme hapax, est en fait attesté à l’époque classique, du moins à l’actif. 52. Dorival, Nombres, 161-162. 53. On trouve parfois l’expression de néologismes sémantiques, pour désigner l’apparition de nouveaux sens au sein des mots grecs. En fait, les néologismes sémantiques sont des sémitismes.

292

2. Les débats et les notions-clés

19. La lexicographie de la Septante

surinterprétation. Il consiste à utiliser des mots grecs, mais en s’écartant légèrement de leur emploi usuel. Par exemple, dans la religion grecque, τὰ σωτήρια désigne le sacrifice du salut, que l’on offre en action de grâces pour une guérison ou un heureux retour. Un sacrifice analogue existe dans la Bible hébraïque, où il porte le nom de ‫שלמים‬. La traduction littérale de l’hébreu serait σωτήρια. Or, la Septante propose toujours le substantif singulier τὸ σωτήριον. De la sorte, par le truchement du radical σωτηρ-, elle évoque le sacrifice du salut, mais, grâce au singulier, elle suggère que le sacrifice biblique du salut n’est pas exactement identique au rite grec. La réalité biblique est hellénisée, mais en même temps elle est différenciée de la réalité grecque. Autre exemple: la phase préparatoire de tout sacrifice consiste à apporter l’offrande. Dans la religion grecque, cet apport est désigné par le verbe ἐπιφέρειν. Ce verbe est employé par la Septante, mais jamais dans un contexte de sacrifice. À sa place, elle utilise les verbes ἀναφέρειν et προσφέρειν, qui font clairement écho au verbe grec tout en s’en distinguant: il est ainsi suggéré que l’apport sacrificiel biblique est différent de l’apport grec, tout en relevant du même grand rituel.

2.2.7 Les traductions multiples Il s’agit d’un phénomène assez répandu, mais qui est rarement pris en compte par les lexicographes de la Septante. Assez souvent, la Septante donne des traductions variables d’un même mot hébreu. Serait-ce que, comme la langue française, le grec n’aime pas la répétition d’un même mot dans le même ensemble textuel et a le souci de varier? Peut-être cette explication vaut-elle dans quelques cas. Mais, dans d’autres cas, elle ne convient pas. En Nombres 35, les fils d’Israël reçoivent l’ordre de réserver aux lévites quarante-huit villes, ainsi que le ‫ מגרש‬qui les entoure; ce substantif est relativement clair en hébreu, où il désigne la zone immédiatement extérieure aux murs des villes, dans laquelle les habitants mènent paître leurs troupeaux (verbe ‫)גרש‬. Il ne semble pas que la langue grecque ait un correspondant pour cette réalité hébraïque. La Septante propose quatre traductions: aux versets 2 et 7, τὰ προάστια, « les banlieues »; au verset 3, τὰ ἀφορίσματα, « les hlieuxi mis à part »; au verset 4, τὰ συγκυροῦντα, « les hlieuxi contigus »; au verset 5, τὰ ὅμορα, « les [lieux] limitrophes ». Les première, troisième et quatrième traductions renvoient à l’organisation de l’espace, la deuxième est de type interprétatif: au lieu d’être décrit en termes d’espace, le ‫ מגרש‬est défini par sa fonction, qui consiste à être mis à part pour les lévites. Le souci de variation pourrait être invoqué pour la première catégorie de traduction; ce n’est pas le cas pour la seconde catégorie, qui correspond à un changement de registre langagier. Un tel procédé doit-il être versé au dossier des limites de l’hellénisation, dans la mesure où, si un même mot hébreu est rendu par plusieurs termes grecs, c’est qu’aux yeux des traducteurs il n’aurait pas de véritable équivalent en grec et qu’il n’est pas traduisible? Mais une telle explication n’est sans doute pas juste, tout d’abord parce qu’il aurait été facile de translittérer ‫מגרש‬, ensuite parce que les quatre traductions sont de vraies tentatives de traduction. Dès lors, on peut suggérer que la multiplicité des traductions s’explique ainsi: face à une réalité biblique considérée comme n’ayant pas de vrai équivalent grec, les traducteurs ont considéré que la moins mauvaise approximation de celle-ci consistait à combiner plusieurs traductions: le ‫ מגרש‬est une sorte de banlieue, un espace qui est contigu et limitrophe aux villes, enfin un lieu qui est mis à part.

2. Les débats et les notions-clés

293

19. La lexicographie de la Septante

2.2.8 Les hellénismes En règle générale, les études lexicales de la Septante soulignent que les traducteurs ont su proposer des correspondants aux termes hébreux qui sont satisfaisants du point de vue lexicographique, mais dont certains peuvent poser problème quand la langue cible est privilégiée. Prenons l’exemple de la correspondance presque systématique entre δῆμος et ‫ משפחה‬dans le livre des Nombres. Elle renvoie à une réalité politique grecque de l’époque classique et hellénistique, où les tribus et les dèmes sont des divisions des cités d’Athènes, de Rhodes et d’Alexandrie 54. De la sorte, la Septante transpose le modèle de l’organisation civique sédentaire en usage dans des grandes cités grecques à la description d’une population nomade errant dans le désert: les fils d’Israël forment ainsi l’équivalent d’une cité en déplacement. On peut trouver que cette transposition est une réussite. On peut aussi estimer qu’elle sollicite trop les réalités bibliques. Il faut noter néanmoins que de tels exemples ne sont pas nombreux. 2.2.9 Les traductions conjecturales Lorsque les traducteurs étaient en présence d’un texte hébreu à leurs yeux difficile ou problématique, ils ont proposé des traductions conjecturales, qui se divisent en plusieurs catégories, comme l’a bien montré Tov 55. On ne peut toutefois retenir la première de ces catégories, les mots non traduits, simplement translittérés, parce qu’une translittération est une non-traduction ou un refus de traduire, non une traduction conjecturale. Les cinq autres catégories sont: 1 / Les traductions contextuelles: par exemple, le mot ‫ ארמון‬est rendu par ἄμφοδον (Jr 17,27; 49,27/30,16), ἄντρον (3Règnes 16,18), βάρις (Ps 47/48,4), βασίλειον (Pr 18,19), γῆ (Jr 9,20/21), ἐναντίον (4Règnes 15,25), θεμέλιον (Es 25,2), ναός (Jr 30/37,18), οἶκος (Es 32,14), πόλις (Es 34,13), πυργόβαρις (Ps 121/122,7), χώρα (Am 3,9.10.11; Mi 5,4/5). Ces traductions conjecturales aboutissent à des traductions multiples, mais elles ne doivent pas être confondues avec les traductions multiples analysées plus haut, qui sont de véritables traductions. 2 / Les manipulations contextuelles, qui consistent à manipuler les consonnes hébraïques en vue d’arriver à des mots mieux adaptés au contexte, comme en Jr 31/38,8. 3 / La prise en considération du parallélisme, notamment en Ésaïe, qui peut être considérée comme une sous-catégorie de la catégorie des traductions contextuelles. 4 / L’emploi d’un terme général qui convient dans le contexte, ainsi παρασκευάζω en Jérémie; de nouveau, il s’agit d’une sous-catégorie de la catégorie des traductions contextuelles. 5 / Les traductions étymologiques; par exemple, en Dt 28,5 et 17, ἐγκατάλειμμα, « reste », a été retenu parce qu’ἐγκαταλείπω correspond ailleurs à ‫ שאר‬56. Tov note avec pertinence que l’établissement d’une traduction conjecturale n’est jamais sûr parce que celle-ci peut en fait dépendre d’un autre modèle hébreu ou encore d’une tradition ancienne d’interprétation. De fait, le nombre de ces traductions conjecturales est ap54. Voir G. Dorival, « Le lexique de l’administration et de la politique » dans le présent volume. 55. E. Tov, « Did the Septuagint Translators Always Understand Their Hebrew Text? » in: A. Pietersma / C. E. Cox (éd.), De Septuaginta. Studies in Honour of John William Wevers on His Sixty-Fifth Birthday, Mississauga 1984, 53-70 (réimpression in: Tov, Greek and Hebrew Bible, 203-218). 56. En fait il y a un seul exemple de cette correspondance, en 2Esdras 9,15, qui a été traduit en grec bien après le Pentateuque. L’explication de Tov n’est guère recevable.

294

2. Les débats et les notions-clés

19. La lexicographie de la Septante

pelé à se réduire au fur et à mesure que la tradition de lecture dont dépend la Septante sera mieux connue.

2.3 Les résurgences récentes du débat sur le grec biblique. Fernández Marcos notait en 1979 que beaucoup de savants n’avaient pas renoncé à l’idée de grec biblique 57. Deux débats récents le prouvent. Le premier porte sur la question de savoir si les mots de la Septante ont plusieurs sens. Ou du moins ont-ils plusieurs dimensions, comme le suggère Tov 58? Il propose de distinguer le sens des mots à l’étape qui précède la Septante, le sens des mots dans la Septante tel qu’il était voulu par les traducteurs et le sens des mots quand ils sont cités d’après la Septante. Il prend trois exemples, dont celui de Gn 1,16, où Dieu crée le grand luminaire εἰς ἀρχάς du jour et le petit luminaire εἰς ἀρχάς de la nuit; εἰς ἀρχάς correspond à ‫ ;לממשלת‬pour les traducteurs, le sens est incontestablement celui de « à titre de gouvernements »; mais la traduction de la Vieille Latine a adopté le mot initium, parce qu’ἀρχάς a le sens de « commencement » dans d’autres passages de Genèse (par exemple en Gn 1,1); en choisissant le mot ἀρχή, qui a le sens de « commencement » et celui de « commandement », les traducteurs ont ouvert la possibilité de comprendre l’expression d’une manière en réalité fautive. Il y a donc deux dimensions des mots bibliques: la signification voulue par le traducteur et la signification appliquée au mot après l’achèvement de la traduction. A ces deux dimensions s’ajoute la troisième dimension du sens que les mots grecs avaient avant leur emploi dans la Septante. Pour Tov, la lexicographie de la Septante doit privilégier le sens voulu par les traducteurs. Le sens dégagé à des époques ultérieures est secondaire. Une attention particulière doit être apportée à ce que Tov appelle les hébraïsmes lexicaux, c’est-à-dire les sémitismes: en ce cas, les mots grecs n’ont pas d’autre sens que le mot hébreu dont ils sont l’équivalent; εἰρήνη n’est pas autre chose que le symbole de ‫ ;שלום‬προσήλυτος n’est pas autre chose que le symbole de ‫גר‬. Avec ces concepts de sens voulus par les traducteurs et d’hébraïsmes lexicaux, on retrouve le jargon judéo-grec cher aux partisans du grec biblique. Cette analyse a été critiquée avec force par Philippe Lefebvre 59. Les sens de « commencement » et de « commandement » ne sont pas étrangers l’un à l’autre; l’ambiguïté de sens de Gn 1,16 est voulue par le traducteur; le premier jour est celui du commencement de la création; le quatrième jour (les luminaires), celui du commencement du temps; le pluriel ἀρχαί de Gn 1,16 renvoie sans doute au monde du multiple. En fait, c’est à l’intérieur des textes et dans les relations tissées par les textes entre eux qu’il convient d’approcher du sens d’un mot. Tout est question de contexte et d’intertextualité. Il n’y a pas de grec biblique. Une autre critique pertinente se lit chez Arie van der Kooij 60. S’appuyant sur les tra57. Fernández Marcos, Introduccion, 7-10. Il signale notamment Nigel Turner, l’auteur du volume III, Syntax, de J. H. Moulton, A Grammar of New Testament Greek, Edinburgh 1963. 58. Tov, « Three dimensions », 529-544. 59. P. Lefebvre, « Les mots de la Septante ont-ils trois dimensions? Φωστῆρεις εἰς ἀρχάς (Gn 1,16) » in: G. Dorival / O. Munnich (éd.), Selon les Septante. Trente études sur la Bible grecque des Septante en hommage à Marguerite Harl, Paris 1995, 299-320. 60. A. van der Kooij, « Schwerpunkte der Septuaginta-Lexikographie » in: S. Kreuzer / J. P. Lesch (éd.), Im Brennpunkt: Die Septuaginta. Studien zur Entstehung und Bedeutung der Griechischen Bibel, Stuttgart/Berlin/Köln 2004, 119-132. 2. Les débats et les notions-clés

295

19. La lexicographie de la Septante

vaux de Flashar (voir note 23), il fait remarquer que les traductions stéréotypées ne doivent pas être confondues avec des traductions mécaniques et il donne l’exemple d’εἰρήνη en 1Règnes 25,5b–6. La seconde réhabilitation du grec biblique tient à la diffusion, surtout dans les pays anglo-saxons, du modèle interlinéaire proposé par Albert Pietersma. Ce modèle a en effet des implications sémantiques et lexicographiques, notamment celle selon laquelle les mots grecs ont le sens des mots hébreux correspondants; cette conception va dans le sens de l’existence d’un jargon judéo-grec. Cameron Boyd-Taylor a développé des idées qui vont dans le même sens. Le modèle interlinéaire de Pietersma prend son point de départ dans le fait que la traduction de la Septante est en relation de dépendance par rapport à l’hébreu 61; cette dépendance se manifeste en ce qu’elle ne peut être lue indépendamment de l’hébreu, parce que les influences qu’elle a subies la rendent souvent incompréhensible pour quiconque ne fait pas référence au texte source. Les deux textes doivent donc être maniés ensemble. La traduction répond en fait à un but pédagogique: celui de servir, aux côtés de l’original hébreu, comme une aide écrite à la fois pour étudier la langue hébraïque et pour comprendre le texte biblique. Le terme d’interlinéarité a renvoyé un temps à l’existence d’un document interlinéaire réel, mais est employé aujourd’hui dans le sens métaphorique d’une relation de dépendance et de soumission. Les traducteurs fournissent des équivalents mot-à-mot de l’hébreu quand cela est possible, paraphrasent les passages difficiles ou obscurs, ajoutent des explications quand l’hébreu risque d’être mal compris. Une formulation voisine du modèle interlinéaire est proposée par Boyd-Taylor, qui caractérise la Septante comme une « métaphrase », c’est-à-dire une traduction mot-à-mot 62. Il la compare à la manière dont le texte d’Homère était manipulé dans les écoles d’Égypte. Les papyrus scolaires proposent des rendus mot-à-mot de vers homériques dans la langue de la koinè et des paraphrases de passages difficiles. Cela le conduit à chercher les origines de la Septante dans les tentatives juives de créer leur tradition littéraire. Il insiste sur le fait que les Juifs hellénophones avaient besoin d’avoir accès à leurs textes fondateurs dans leur langue originale. Dans une autre contribution, le même auteur a proposé d’utiliser le concept d’interlangage dans le cas de la Septante, l’interlangage étant un système de langage utilisé pour décoder un message dans un langage et l’encoder dans un autre langage 63. Le modèle interlinéaire a inspiré l’entreprise anglo-saxonne NETS (New English Translation of the Septuagint): pour respecter la relation de dépendance entre la Septante et l’hébreu, NETS ne propose pas une traduction directe du texte grec, mais part

61. A. Pietersma, « A New Paradigm for Addressing Old Questions: The Relevance of the Interlinear Model for the Study of the Septuagint » in: J. Cook (éd.), Bible and Computer (Actes d’un colloque de juillet 2000), Leiden 2002, 337-364. Voir aussi A. Pietersma / B. G. Wright (éd.), A New English Translation of the Septuagint and other Greek Translations Traditionally Included under that Title, The Psalms, New York, NY/Oxford 2007, xvi-xvii. 62. C. Boyd-Taylor, « A Place in the Sun: The Interpretative Significance of LXX-Psalm 18:5c » BIOSCS 31 (1998), 71-105. 63. C. Boyd-Taylor, « Lexicography and Interlanguage—Gaining our Bearings » BIOSCS 37 (2004), 55-72. Il emprunte le concept d’interlangage à G. Toury, Descriptive Translation Studies and Beyond, Amsterdam 1995.

296

2. Les débats et les notions-clés

19. La lexicographie de la Septante

de la traduction de l’hébreu telle que la propose la New Revised Standard Version et lui apporte les modifications qu’impose la Septante. De nombreuses critiques ont été apportées au modèle interlinéaire 64. La documentation existante montre que les traductions de l’hébreu ont existé indépendamment de ce dernier; en revanche aucun document n’atteste que la Septante ne pouvait fonctionner qu’accompagnée du texte source. En réalité, la Septante n’est que très rarement inintelligible et, quand elle l’est, l’hébreu l’est parfois aussi. On ne peut la caractériser comme traduction interlinéaire, parce qu’elle prend bien des libertés avec l’hébreu, en particulier dans l’ordre des mots, et parce qu’elle offre des traits de recherche littéraire. Le modèle projette sur la Septante ce qui convient à la situation de la version littéraliste d’Aquila, élaborée à une époque où les Sages veulent affirmer la prédominance de l’hébreu. On peut même se demander s’il n’a pas été élaboré pour justifier la démarche de NETS, qui est somme toute étrange, puisqu’elle consiste à traduire la Septante seulement par endroits. La démarche de Boyd-Taylor n’est pas non plus totalement convaincante: les exercices scolaires homériques se situent au sein de la seule la culture grecque. Il n’est pas sûr non plus que la traduction de la Septante ait été réalisée pour donner accès à l’hébreu. En fait, ni Pietersma ni Boyd-Taylor ne signalent le meilleur argument en faveur du modèle interlinéaire: les papyrus bilingues d’Égypte 65. Nous connaissons une quinzaine de glossaires alphabétiques ou thématiques. Ces derniers envisagent les mots relatifs, par exemple, aux poissons, aux légumes, aux vents, aux dieux. Sur la colonne de gauche figurent les termes latins et, sur la colonne de droit, les équivalents grecs. À chaque ligne figurent normalement, à gauche, un mot latin et, à droite, un mot grec; toutefois, il arrive que plusieurs équivalents grecs soient proposés dans les glossaires alphabétiques. Ceux-ci sont très proches des Hermeneumata Pseudodositheana de la tradition médiévale; néanmoins, ceux-ci inversent l’ordre des colonnes: le grec à gauche et le latin à droite. Ces glossaires sont destinés à des Grecs voulant apprendre le latin et, moins souvent, à des Latins voulant apprendre le grec. Leur date d’apparition est discutée, mais pourrait remonter au premier siècle de notre ère. Il existe également des glossaires bilingues d’auteurs latins ou encore des traductions grecques de ces auteurs disposées sur la colonne de droite, l’original latin occupant la colonne de gauche. Sont concernés principalement Cicéron et Virgile, à un degré moindre, Juvénal, Lucain, Salluste, Térence. Les traductions sont littérales et reproduisent l’ordre des mots de l’original. Elles se ressemblent suffisamment pour que l’on puisse affirmer qu’elles dépendent d’un dictionnaire gréco-latin, dont la date de confection semble remonter au premier siècle de notre ère. On note encore que les mots latins sont plus souvent transcrits en lettres grecques qu’ils ne sont donnés en lettres latines: la translittération apparaît comme une étape normale du processus de traduction. Ainsi, les papyrus 64. J. Dines, The Septuagint, Londres/New York, NY 2004, 52-54; R. L. Troxel, LXX-Isaiah as Translation and Interpretation. The Strategy of the Translator of Septuagint of Isaiah, Leiden 2008, 62-72; J. Joosten, « Reflections on the Inerlinear Paradigm in Septuagintal Studies » in: A. Voitila / J. Jokiranta (éd.), Scripture in Transition. Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo, Leiden 2008, 163-178; T. Rajak, Translation and Survival. The Greek Bible of the Ancient Jewish Diaspora, Oxford 2011, 143-144. 65. G. Dorival, « La forme littéraire des Hexaples d’Origène » in: S. Kaczmarek / H. Pietras (éd.), Origeniana decima. Origen as Writer, Louvain/Paris/Walpole, MA 2001, 601-614. 2. Les débats et les notions-clés

297

19. La lexicographie de la Septante

bilingues d’Égypte font connaître un modèle qui est proche du modèle interlinéaire de Pietersma, mais qui s’en distingue par la présence de la translittération. En fait, il correspond mieux à la disposition en colonnes parallèles des Hexaples d’Origène, avec la présence de la deuxième colonne de translittération, qu’à la Septante, pour laquelle la disposition en colonnes du texte hébreu et de la Septante n’est pas attestée avant Origène. Du point de vue lexicographique, le modèle interlinéaire rend problématique l’établissement d’un lexique de la Septante, puisque le grec est sous la dépendance de l’hébreu 66. Il est nécessaire et suffisant de connaître le sens des mots hébreux pour connaître le sens des mots grecs correspondants. En conséquence, un mot grec donné n’a pas forcément la signification qu’il a habituellement en grec, mais celle du mot hébreu qui lui correspond. Ainsi, selon Pietersma, en Psaumes, ἔλεος ne signifie pas « pitié », mais « bienveillance, gentillesse », comme ‫ חסד‬qui lui correspond 67. Mais, comme l’a montré Jan Joosten, en Psaumes, le verbe ἐλεέω et l’adjectif ἐλεήμων évoquent l’idée de pitié, à l’instar des mots hébreux qu’ils traduisent, les verbes ‫ חנן‬et ‫רחם‬, l’adjectif ‫ ;חנון‬il n’y aucune raison de penser qu’ἔλεος ait un autre sens que les mots qui lui sont apparentés 68. Le modèle interlinéaire est incapable de rendre compte de ce type de phénomène, mais aussi d’autres phénomènes, qui ont retenu l’attention de l’éminent lexicographe qu’est Muraoka. Ce dernier critique aussi les vues de BoydTaylor 69. Pour les partisans du modèle interlinéaire, il faut recourir à l’original hébreu en cas de doute; mais l’hébreu peut lui-même être ambigu, comme en Jos 1,8; de plus, les connotations d’un mot hébreu ne sont pas toujours connues avec certitude: ainsi pour ‫ ;ברית‬enfin, notre connaissance elle-même du sens d’un mot hébreu remonte parfois à la manière dont ce mot a été compris par la Septante: en Dt 24,7, καταδυναστεύω correspond à l’hithpaël du verbe ‫עמר‬, dont le sens discuté n’est pas éclairé par les autres langues sémitiques; on comprend en général comme le fait la Septante. En réalité, il y a deux approches possibles de la Septante qu’on peut exprimer sous forme de paires: langue source versus langue cible, amont versus aval, source versus réception, traducteur versus lecteur, mots grecs versus sens hébreux. Le LEH se centre sur l’intention du traducteur, le GELS sur l’interprétation que les lecteurs pouvaient donner de la Septante entre 250 avant notre ère et 100 après notre ère. Il faut lire la Septante comme un texte grec et non comme un texte reflétant l’hébreu sous-jacent. Il faut ajouter que le modèle interlinéaire est incapable de rendre compte des traductions multiples: en Gn 31,19, ‫ תרפים‬est rendu par εἴδωλα; le traducteur de 1Règnes ne se rappelait peut-être plus cette traduction, mais pourquoi a-t-il transcrit par θεραφίν en 15,23 et traduit par κενοτάφια en 19,13 et 16? Dans ce cas comme dans les autres cas similaires (voir plus haut les quatre traductions différentes de ‫ מגרש‬en Nombres 35), le lexicographe doit prendre en considération le mot hébreu correspondant, le mot grec 66. Voir C. Boyd-Taylor, « The Evidentiary Value of Septuagintal Usage for Greek Lexicography » BIOSCS 34 (2001), 47-80, notamment 74. 67. Pietersma, Psalms, xxii. 68. J. Joosten, « ‫ ‹ חסד‬bienveillance › et ἔλεος ‹ pitié ›. Réflexions sur une équivalence lexicale dans la Septante » in: E. Bons, « Car c’est l’amour qui me plaît, non le sacrifice … ». Recherches sur Osée 6,6 et son interprétation juive et chrétienne, Leiden/Boston, MA 2004, 25-42. 69. Muraoka, « Recent Discussions », 221-235.

298

2. Les débats et les notions-clés

19. La lexicographie de la Septante

en dehors de la Septante, le mot grec au sein de la Septante, la méthode de traduction dans le livre considérée, les mots de même champ sémantique, etc. Le modèle interlinéaire ne permet pas non plus de mettre en évidence les innovations lexicographiques: κλητή, en Lv 23,3, correspond à ‫ «( מקרא‬appelé »), mais il ne peut ici avoir ce sens; il est en fait un adjectif substantivé qui désigne un jour où le peuple est appelé à un rassemblement public 70.

2.4 Les mots de la Septante sont-ils caractéristiques d’un milieu de militaires? C’est ce qu’affirme Joosten 71: la traduction n’a pas été faite par une élite proche de la cour et des cercles lettrés d’Alexandrie, par des scribes bilingues rompus aux finesses du texte hébreu tout autant qu’aux nuances de la langue grecque. Des indices linguistiques vont dans le sens d’un milieu plus modeste: la Septante emploie des mots avec le sens qu’ils ont, non dans les textes littéraires, mais seulement dans les papyri (ὑποζύγιον au sens d’âne et non de bête de somme; γόμος au sens de charge et non de charge d’un bateau); elle emploie un grec non littéraire, dénué de recherche de style. La langue de la Septante est celle de la communauté juive d’Alexandrie, où les soldats étaient nombreux. Des mots du vocabulaire militaire sont présents, comme ἀποσκευή désignant la femme et les enfants qui accompagnent le soldat dans ses déplacements, στρατοπεδεύω dans le cas d’Abraham qui n’est pas un militaire (Gn 12,9), παρεμβολή, qui désigne le campement militaire alors que le mot hébreu correspondant désigne un campement provisoire. Ces conclusions peuvent être discutées. La présence du vocabulaire papyrologique est caractéristique d’un document de la koinè. La présence du vocabulaire militaire plus grande que dans l’hébreu ne doit pas surprendre: pour les traducteurs, le peuple hébreu est une cité en marche dans le désert, qui s’organise de manière militaire 72. D’excellents spécialistes de la littérature grecque soulignent la grécité de la Septante 73. La compétence des traducteurs en matière de religion juive ne fait pas de doute. Toutes ces caractéristiques vont dans le sens de la narration d’Aristée, qui parle d’hommes compétents dans les lettres tant grecques que juives (Lettre d’Aristée à Philocrate 121).

3. Les réalisations de la lexicographie de la Septante 3.1 Les principales études lexicales Les études antérieures à 1969 sont répertoriées par S. P. Brock / C. T. Fritsch / S. Jellicoe, A Classified Bibliography of the Septuagint, Leiden 1973, pages 30-34 (soixante70. C. Dogniez, « Les noms de fêtes dans le Pentateuque grec » JSJ 37 (2006), 344-366, en particulier 351. 71. J. Joosten, « Le milieu producteur du Pentateuque grec » REJ 165 (2006), 349-361. 72. Dorival, Nombres, 147, 158-166. 73. M. Casevitz, « D’Homère aux historiens romains: le grec du Pentateuque alexandrin » in: C. Dogniez / M. Harl (éd.), La Bible des Septante. Le Pentateuque d’Alexandrie, Paris 2001, 77-85; D. Pralon, « Le style de la Septante » in: Harlé / Pralon, Lévitique, 47-81. 3. Les réalisations de la lexicographie de la Septante

299

19. La lexicographie de la Septante

dix signalements). Les études parues entre 1970 et 1993 sont énumérées par C. Dogniez, Bibliography of the Septuagint. Bibliographie de la Septante, Leiden/New York, NY/Köln 1995, pages 35-47 (trente-six études lexicales « générales » et 159 études lexicales « particulières »). Paru en 2009, le GELS (voir plus haut) contient une bibliographie de 660 titres environ, qui ne renvoient pas tous à des études lexicales (pages xxv-xl). Dans une production qui s’accroît de manière exponentielle et en se limitant aux cinquante dernières années, il est possible de sélectionner quelques études particulièrement importantes: S. Daniel, Recherches sur le vocabulaire du culte dans la Septante, Paris 1966 (à compléter par G. Dorival, « L’originalité de la Bible grecque des Septante en matière de sacrifice », dans S. Georgoudi / R. Koch Piettre / F. Schmidt (éd.), La cuisine et l’autel. Les sacrifices en questions dans les sociétés de la Méditerranée ancienne, Paris, 2005, pages 309-315: κάρπωμα ne signifie pas le brûlage intégral de l’offrande animale ou végétale par opposition à θυσία, mais désigne l’animal ou le végétal dont la divinité a la jouissance pour elle seule; d’où la traduction par « apanage » dans les volumes de « La Bible d’Alexandrie »; la religion païenne connaît le rite de la κάρπωσις). R. A. Kraft, Septuagintal Lexicography, Missoula, MT 1972, édition révisée 1975. Ce recueil comprend notamment: S. P. Brock / J. A. L. Lee, « A Memorandum on the Proposed LXX Lexicon Project », pages 20-24; H. S. Gehman, « Adventures in Septuagint Lexicology », pages 102-109 (reproduction d’un article paru dans Textus 5 (1966), 125-132); G. B. Caird, « Towards a Lexicon of the Septuagint », pages 110-152 (reproduction des articles signalés plus haut à propos du LSJ); R. A. Kraft, « Towards a Lexicon of Jewish Translation Greek », pages 157-178. L. Monsengwo Pasinya, La Notion de nomos dans le Pentateuque grec, Rome 1973 (le mot νόμος est une bonne traduction de ‫תורה‬, car il n’a pas un sens étroitement juridique, mais débouche sur la sphère religieuse). P. Walters (formerly Katz), The Text of the Septuagint. Its corruptions and their Emendations, Cambridge 1973 (la deuxième partie, pages 141-264, est consacrée aux sémitismes: hébraïsmes lexicaux, translittérations et ce que l’auteur appelle homonymes, et qui sont en fait des mots homophones en grec et en hébreu). J. W. Olley, « Righteousness » in the Septuagint of Isaiah. A Contextual Study, Missoula, MT 1979. J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch, Chico, CA 1983. L’auteur divise le lexique du Pentateuque en deux catégories et sous-catégories: I. Mots attestés en dehors de la littérature biblique et para-biblique, (a) mots attestés en grec archaïque et classique, (b) mots propres à la koinè. II. Mots attestés seulement dans les textes bibliques et para-bibliques, (a) mots qui relèvent probablement du grec « normal », (b) mots qui relèvent probablement du grec biblique. Les conclusions sont que la plus grande partie des mots du Pentateuque appartient au grec de l’époque; les influences de l’hébreu sur le grec viennent de la traduction, et non d’un judéo-grec en usage dans la diaspora. Ces conclusions ont été réaffirmées récemment par le même savant, « A Lexical Study. Thirty Years on, with Observations on « Order » Words in the LXX Pentateuch » in: S. M. Paul / R. A. Kraft / L. H. Schiffman / W. W. Field, Emanuel. Studies in Hebrew Bible, Septuagint, and Dead Sea Scrolls in Honor of Emanuel Tov, Leiden/Boston, MA 2003, pages 513-524: même si le nombre des occurrences des verbes ἐντέλλομαι (156 occurrences), συντάσσω (82), προστάσσω (12) n’est pas 300

3. Les réalisations de la lexicographie de la Septante

19. La lexicographie de la Septante

identique aux emplois des papyri (respectivement 22, 285 et 32), la Septante et les papyri ont en commun d’employer beaucoup συντάσσω et moins souvent προστάσσω; l’emploi fréquent d’ἐντέλλομαι dans la Septante, surtout dans le Deutéronome (quatre-vingt-cinq occurrences) doit être mis en rapport avec l’importance d’ἐντολή. L’absence de κελεύω dans la Septante contraste avec sa présence dans les papyri (vingtcinq occurrences); elle s’explique par le fait que les ordres dans la Septante renvoient à la sphère officielle, et non privée. M. Cimosa, Il vocabolario di Preghiera nel Pentateuco del LXX, Rome 1985. A. Thibaut, L’infidélité du peuple élu. Ἀπειθῶ entre la Bible hébraïque et la Bible latine, Rome/Turnhout 1988 (étude de ‫סרר‬, ‫מרה‬, ‫מרד‬, ἀπειθῶ, παραπικραίνω). T. Muraoka (éd.), Melbourne Symposium on Septuagint Lexicography, Atlanta, GA 1990. Quatre études: J. A. L. Lee, « συνίστημι: A Sample Lexical Entry », pages 116; T. Muraoka, « Septuagintal Lexicography: Some General Issues », pages 17-48; S. P. Swinn, « ἀγαπᾶν in the Septuagint », pages 49-82; E. Tov, « Greek Words and Hebrew Meanings », pages 83-126. S. Olofsson, God is my Rock. A Study of Translation Technique and Theological Exegesis in the Septuagint, Stockholm 1990 (étude des épithètes divines reposant sur des métaphores fondées sur la nature inanimée, du type « roc » dans le texte massorétique et la Septante, principalement les Psaumes). M. Cimosa, La Preghiera nella Biblia graeca. Studi del vocabolario dei LXX, Bologne 1991. C. E. Cox (éd.), VII Congress of the International Organization for Septuagint and Cognate Studies. Leuven 1989, Atlanta, GA 1991. Deux contributions importantes: M. Harl, « Le renouvellement du lexique des Septante d’après le témoignage des recensions, révisions et commentaires grecs anciens », pages 239-259 (réflexion sur les raisons qui font que des mots sont remplacés par d’autres mots, et notamment sur le phénomène d’usure); T. Muraoka, « Hebrew Hapax Legomena and Septuagint Lexicography », pages 205-222. M. Harl, La Langue de Japhet. Quinze études sur la Septante et le grec des chrétiens, Paris 1994 (les études 3 à 8 portent sur le lexique de la Septante, notamment ἐμπαιγμός, κατάνυξις, κιβωτός, συμποδίζειν). C. Dogniez / M. Harl (éd.), La Bible d’Alexandrie. Le Pentateuque d’Alexandrie, Paris 2001 (M. Harl, « Glossaire », pages 861-906: étude de vingt champs lexicaux: abomination, âme, anathème, apaisement, communauté, consacrer, culte, esprit, face, faute, fête, immigrant, investiture, loi, offrandes, rémission, seigneur, service, témoignage, visite). F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta, Münster 2001, pages 196-288. F. Austermann, Von der Tora zum Nomos: Untersuchungen zur Übersetzungsweise und Interpretation im Septuaginta-Psalter, Göttingen 2003. B. A. Taylor / J. A. L. Lee / P. R. Burton / P. E. Whitaker, Biblical Greek Language and Lexicography, Grand Rapids, MI 2004. Les contributions qui concernent le plus la lexicographie de la Septante sont les suivantes: T. Muraoka, « Septuagintal Lexicography », pages 85-90; K. Hauspie, « The LXX Quotations in the LSJ Supplements of 1968 and 1996 », pages 108-125; E. Eynikel / K. Hauspie, « The Use of δράκων in the Septuagint », pages 126-135; B. A. Taylor, « Hebrew to Greek. A Semantic Study 3. Les réalisations de la lexicographie de la Septante

301

19. La lexicographie de la Septante

of σπεύδω in the New English Translation of the Septuagint », pages 136-148; C. BoydTaylor, « Linguistic Register and Septuagintal Lexicography », pages 149-166; B. A. Taylor, « Deponency and Greek Lexicography », pages 167-176. J. Joosten / P. J. Tomson (éd.), Voces Biblicae. Septuagint Greek and its Significance for the New Testament, Leuven 2008. Sept contributions: E. Bons, « Le verbe κτίζω comme terme technique de la création dans la Septante et dans le Nouveau Testament », pages 1-15; J. Dines, « Light from the Septuagint on the New Testament – or Vice Versa? », pages 17-34; P. J. Tomson, « Blessing in Disguise: ΕΥΛΟΓΕΩ and ΕΥΧΑΡΙΣΤΕΩ between « Biblical » and Everyday Greek Usage », pages 35-61; M. Karrer, « ῬΙΖΑ – Wurzel und Geschlecht ein Motiv zwischen griechischer Antike, Septuaginta und Neuem Testament », pages 63-98; J. A. L. Lee, « ΕΞΑΠΟΣΤΕΛΛΩ », pages 99-113; D. Angers, « The Pauline Expressions « Until this Very Day » and « Until Today » (Rom 11:8 and 2 Cor 3:14-15) in the Light of the Septuagint », pages 115154; J. Joosten, « « À Dieu ne plaise » (Matthieu 16,22). La provenance et l’arrière-plan de l’expression ΙΛΕΩΣ ΣΟΙ », pages 155-167. M. Karrer / W. Kraus (éd.), Die Septuaginta – Texte, Kontexte, Lebenswelten, Tübingen 2008. Six articles sont regroupés sous le titre Lexikographie und Grammatik: K. Hauspie, « The Idiolect of the Target Language in the translation Process. A Study of the Calques in the LXX of Ezekiel », pages 205-213; J. A. L. Lee, « A Lexicographical Database for Greek: Can it be Far Off? The Case of amphodon », pages 214-220; T. Muraoka, « Recent Discussion on the Septuagint Lexicography with Special Reference to the So-called Interlinear Model », pages 221-235; A. Passoni dell’Acqua, « Von der Kanzlei der Lagiden zur Synagoge. Das ptolemäische Vokabular und die Septuaginta », pages 236-247; R. Glenn Wooden, « The φορολόγος of 2 Esdras », pages 248-257; G. Walser, « Die Wortfolge der Septuaginta », pages 258-266. M. Karrer / W. Kraus (éd.), Die Septuaginta – Texte, Theologien, Einflüsse, Tübingen 2010. La section Lexikographie und Übersetzungstechnik présente neuf contributions: T. Muraoka, « The Logico-Semantic Analysis of the Genitive Relationship in the LXX Greek (Gen 1-25) », pages 313-321; A. Passoni dell’Acqua, « Translating as a Means of Interpreting: The Septuagint and Translation in Ptolemaic Egypt », pages 322-339; C. Kugelmeier, « Voces biblicae oder voces communes? Zum Sprachgebrauch der Septuaginta im Lichte neuerer Papyrusforschungen », pages 340-356 (étude de καταργέω, ἀντιλήπτωρ, θειότης); H. Ausloos / B. Lemmelijn, « Content-Related Criteria in Characterising the LXX Translation Technique », pages 357-376; J. Joosten, « Al tiqré as hermeneutical device and the Septuagint », pages 377-390; J.-H. Kim, « Die Wiedergabe von ‫ עבר‬mit δοῦλος oder παῖς in der Septuaginta der Samuelund Königbücher », pages 391-403; E. Bons, « Seltene Wörter in der Septuaginta des Amosbuches (Am 3,5.15): ἰξευτής, σχάζομαι, θερινός, περίπτερος », pages 404415; P.-M. Bogaert, « Baal au féminin dans la Septante », pages 416-435; K. Hauspie, « Hebrew Transliterations in the Septuagint version of Ezekiel elucidated in Search of the Sources of Theodoret of Cyrrhus », pages 435-444. Il faut également consulter systématiquement les introductions et les annotations des volumes de « La Bible d’Alexandrie », qui étudient le lexique de chaque livre biblique. Un important ouvrage collectif est en préparation sous la direction d’E. Bons et J. Joosten, le Historical and Theological Lexicon of the Septuagint, chez Mohr et Sie302

3. Les réalisations de la lexicographie de la Septante

19. La lexicographie de la Septante

beck à Tübingen. Il étudiera 500 mots et racines de la Septante, en retraçant l’histoire de ces mots dans la littérature grecque, les papyri et les inscriptions, la Septante, la littérature juive hellénophone, le Nouveau Testament et la littérature chrétienne des premiers siècles.

3.2 L’apport de la papyrologie et de l’épigraphie Les progrès dans la connaissance du lexique de la Septante sont largement dus aux documents papyrologiques et, à un degré moindre, épigraphiques. Des mots considérés comme bibliques ont été retrouvés dans les papyri, les inscriptions et les ostraca et du coup prennent place dans le vocabulaire de la koinè.

3.2.1 Lexiques des papyri S. Daris, Spoglio lessicale papirologico, trois volumes, Milan 1968, 1749 pages (simple liste de références). F. Preisigke, Wörterbuch der griechischen Papyrusurkunden mit Einschluss der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten, volumes I-III, Berlin 1925-1931; pour le volume IV, dû à E. Kiessling, 5 fascicules (se terminant à la lettre ζ) sont parus à Berlin puis Marburg, de 1944 à 1993; E. Kiessling / W. Räbsam, Supplement 1 (1940-1966), Amsterdam 1969-1971; H. A. Rupprecht / A. Jördens, Supplement 2 (1967-1976), Wiesbaden 1991; H. A. Rupprecht / A. Jördens, Supplement 3 (1977-1988), Wiesbaden 2000; les volumes de suppléments dressent simplement la liste des occurrences. Il faut aussi se reporter à la Duke Data Bank of Documentary Papyri (DDBDP, http://idp.atlantides.org). Pour les noms géographiques et topographiques de la Septante, il faut consulter le dictionnaire établi par Aristide Calderini et Sergio Daris à partir des documents littéraires, des papyri, des inscriptions et des ostraca: A. Calderini, Dizionario dei nomi geografici e topografici dell’Egitto Greco-Romano, volume I 1 (A-Halikarnasseus), Le Caire, 1935, volume I 2 (Halikarnassos-Aolph), Madrid 1966; A. Calderini / S. Daris, volume II-V et Supplemento 1 (1935-1986), Milan 1973-1987; Supplemento 2 (1987-1993), Bonn 1996; Supplemento 3 (1994-2001), Pise 2003; S. Daris seul, Supplemento 4 (20022005), Pise 2007; Supplemento 5 (2006-2009), Pise 2010. Voir aussi H. Verreth, A Survey of Toponyms in Egypt in the Graeco-Roman Period, Trismegistos Online Publications II, Köln/Leuven 2008. Sur ἡ Ἀραβία désignant dans la Septante la péninsule arabique et les territoires parcourus par les nomades arabes et Ἀραβία sans article faisant référence au nome « Arabie » (Gn 45,10 et 46,34, sans correspondant dans le texte massorétique; peut-être aussi Es 10,9 et 11,11), voir G. Dorival, « La Septante: une ou deux Arabies? » à paraître dans les Mélanges Christian Robin. Pour les noms de personnes de la Septante, il faut se référer à F. Preisigke, Namenbuch, enthaltend alle griechischen, lateinischen, ägyptischen, hebräischen, arabischen und sonstigen semistischen und nicht semitischen Menschennamen, sowie sie in griechischen Urkunden Ägyptens sich vorfinden, Heidelberg 1922, 526 colonnes (réimpression Amsterdam 1967); ainsi qu’au complément de D. Foraboschi, Onomasticon alterum papyrologicum. Supplemento al Namenbuch di F. Preisigke, Milan 1971, 353 pages. 3. Les réalisations de la lexicographie de la Septante

303

19. La lexicographie de la Septante

Enfin, les théonymes et les noms d’objets sacrés sont répertoriés dans G. Ronchi, Lexicon theonymon rerumque sacrarum et divinarum ad Aegyptum pertinentium quae in papyris ostracis titulis Graecis Latinisque in Aegypto repertis laudantur, cinq volumes, Milan 1974-1977.

3.2.2 Ouvrages de référence G. H. R. Horsley, New Documents Illustrating Early Christianity. A Review of Greek Inscriptions and Papyri Published in 1976, North Ryde (Australie) 1981; le volume 2 (documents publiés en 1977) est paru en 1982; le volume 3 (documents publiés en 1978), en 1983; le volume 4 (documents parus en 1979), en 1987; le volume 5, paru en 1989, est consacré à des Linguistic Essays (par exemple sur la fiction du judéo-grec), il offre des index cumulatifs portant sur les volumes 1-5; les volumes 6-9 sont dus à S. R. Llewelyn; le volume 6 (documents publiés en 1980-1981) est paru en 1992; le volume 7 (documents publiés en 1982-1983) est paru en 1994; le volume 8 (documents parus en 1984-1985) est paru en 1988 à Grand Rapids, MI et Cambridge; le volume 9 (documents publiés en 1986-1987) est paru en 2002 dans les mêmes villes. Un volume 10 est en préparation par les soins de S. R. Llewelyn / J. Harrison / M. Theophilos. Beaucoup des éclairages qu’apportent les inscriptions et les papyri sur le Nouveau Testament concernent aussi la Septante. Les volumes contiennent un index des passages des Écritures qui permet au lecteur de repérer rapidement ce qui est susceptible de concerner la Septante. O. Montevecchi, La papirologia, Turin 1973, Milan 19882. L’édition de 1988 est revue, corrigée et pourvue de copieux addenda (pages 535-613). Le lexique de la Septante doit toujours être confronté à l’index des mots grecs (pages 514-522). On y apprend par exemple que le mot ὑποζύγιον est le nom le plus habituel dans les papyri pour désigner l’âne. Cela jette un éclairage sur un fait de lexique qui intrigue dans la Septante, où ὑποζύγιον et ὄνος alternent pour traduire le mot ‫חמר‬: la présence d’ὑποζύγιον s’explique par l’influence de la langue usuelle. 3.2.3 Quelques articles relatifs au lexique de la Septante P.-M. Bogaert, « L’orientation du parvis du sanctuaire dans la version grecque de l’Exode », L’Antiquité classique 50 (1981), 79-85 (les traducteurs utilisent deux roses des vents dans le Pentateuque; dans la première, la mer, θάλασσα, désigne l’ouest et λίψ le sud; dans la seconde, qui est notamment attestée en Ex 27,9-13, θάλασσα, désigne le nord et λίψ l’ouest, conformément à l’usage des papyri d’Égypte). H. Cadell, « Vocabulaire de l’irrigation: la Septante et les papyrus » in: B. Menu (éd.), Les problèmes institutionnels de l’eau en Égypte ancienne et dans l’Antiquité méditerranéenne, Le Caire 1994, pages 103-117 (l’arrosage avec les pieds au pluriel, et non au singulier comme dans le texte massorétique, renvoie à une technique agricole connue par cinq papyri, et qui consiste à utiliser ses pieds pour régler l’écoulement de l’eau dans les potagers, les vergers et les vignobles); « Vocabulaire de la législation ptolémaïque. Problème du sens de dikaiôma dans le Pentateuque » dans: G. Dorival / O. Munnich (éd.), Κατὰ τοὺς o’. Selon les Septante. Trente études sur la Bible grecque des Septante en hommage à Marguerite Harl, Paris 1995, pages 206-221 (dans les papyri, δικαιώματα désignent les « pièces justificatives jointes au dossier d’un procès »; ces δικαιώματα comprennent les textes ayant force de loi que le défendeur souhaite pro304

3. Les réalisations de la lexicographie de la Septante

19. La lexicographie de la Septante

duire à l’appui de sa cause; dans le Pentateuque, par métonymie, le mot désigne tous les types d’actes normatifs, toutes les formes de disposition législative, toutes les règles de droit). H. Heinen, « Zur Terminologie der Sklaverei in ptolemaischen Ägypten: pais und paidiske in den Papyri und der Septuaginta » in: Atti del XVII Congresso Internazionale di Papirologia, Naples 1984, pages 1287-1299. G. Husson, « Κάμινος dans les papyrus grecs », Archeologia 30 (1981), 17-20 (rapprochement avec Nb 25,8); « Le paradis de délices (Genèse 3,23-24) » REG 101 (1988), 64-73. O. Montevecchi, « Dal paganismo al cristianesimo: aspetti dell’evoluzione della lingua greca nei papyri dell’Egitto » Aegyptus 37 (1957), 41-59; « Quaedam de graecitate Psalmorum cum papyris comparata », Proceedings of the 9th International Congress of Papyrology, Oslo 1961, pages 293-310; « Continuita ed evoluzione della lingua greca nella Settanta e nei papiri », Actes du Xe Congrès international de papyrologie, Varsovie 1964, pages 39-49. A. Passoni dell’Acqua, « Euergetes » Aegyptus 56 (1978), 177-191; « La metafora biblica di Dio come roccia e la sua soppressione nelle antiche versioni » Ephemerides Liturgicae 91 (1977), 417-453; « La versione dei LXX e i papiri: note lessicali », Proceedings of the XVI International Congress of Papyrology, Chico, CA 1981, pages 621-632; « Ricerche sulla versione dei LXX e i papiri, I Pastophorion » Aegyptus 61 (1981), 171-211; « Ricerche sulla versione dei LXX e i papiri, II Nomos, III Andrizesthai » Aegyptus 62 (1982), 173-194; « Precisazione sul valore di δῆμος nella versione dei LXX », Rivista Biblica 30 (1982), 197-214; « La terminologia dei reati nei προστάγματα dei Tolemai e nella versione dei LXX », Proceedings of the XVIII International Congress of Papyrology, t. II, Athènes (1988), 335-350; « Notazioni cromatiche dall’Egitto greco-romano. La versione dei LXX e i papiri » Aegyptus 78 (1998), 77-115; « Il Pentateuco dei LXX testimone di istituzioni di età tolemaica » Annali di Scienze Religiose 3 (1999), 171-200; « Biblion et Biblia: la scrittura nella scrittura. Osservazioni sul lessico del materiale scrittorio nella Biblia » Rivista Biblica 54 (2006), 291-319; « La tradizione della traduzione: riflessioni sul lessico del ‹ tradurre › nella Biblia greca e nel giudaismo-ellenistico » Liber annuus 53 (2008), 195-276; « Colore e traparenze nella haute couture dell’Egitto greco-romano » Semitica et Classica 1 (2008), 113-138; « Von der Kanzlei der Lagiden zur Synagoge. Das ptolemäische Vokabular und die Septuaginta » in: M. Karrer / W. Kraus (éd.), Die Septuaginta – Texte, Kontexte, Lebenswelten, Tübingen 2008, pages 236-247; « Translating as a Means of Interpreting: The Septuagint and Translation in Ptolemaic Egypt » in: M. Karrer / W. Kraus (éd.), Die Septuaginta – Texte, Theologien, Einflüsse, Tübingen 2010, pages 322-339. F. Vattioni, « La lessicografia dei LXX nei papiri » Studia Papyrologica 19 (1980), 39-59.

3. Les réalisations de la lexicographie de la Septante

305

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch Robert J. V. Hiebert

1. Introduction The translation of the Hebrew Bible into Greek, a project that began with the Torah probably sometime in the third century BCE in Alexandria, Egypt, was a significant literary and theological development. This project was initiated perhaps within fifty years after Alexander the Great had marched with his forces through the eastern Mediterranean regions, ushering in an era not only of Greek political domination but also of Hellenistic cultural influence. The Septuagint Pentateuch seems to have been the first major translation of religious literature into Greek ever carried out. 1 This undertaking involved new ways of representing the Jewish scriptures. In the process, the Greek lexicon of that time experienced the kinds of developments that would be expected to occur in a living language. In his important monograph entitled A Lexical Study of the Septuagint Version of the Pentateuch, John A. L. Lee traces the semantic changes with respect to existing words, the introduction of new forms and terms, and the process whereby some words faded into obsolescence. This study, in which he examines primarily “the everyday, non-theological, vocabulary of the Pentateuch” in relation to “the vernacular Koine Greek of its time,” 2 is a model of sound lexical methodology. Lee makes considerable use of papyrological material to support the assertion advanced earlier by the likes of Adolf Deissmann, Albert Thumb, Henry St. John Thackeray, and James H. Moulton that “the language of the LXX translators was essentially the Greek of their time,” though he acknowledges that “the translators, and the Alexandrian Jews generally, introduced some novel features into their vocabulary […] by borrowing words from Hebrew or Aramaic, by forming new words in Greek, and by giving special significations to some current Greek terms.” 3 The present article constitutes a survey of repre1.

2. 3.

J. W. Wevers, “The Interpretative Character and Significance of the Septuagint Version” in: M. Sæbø (ed.), Hebrew Bible/Old Testament: The History of Its Interpretation, vol. 1, From the Beginnings to the Middle Ages (Until 1300), part 1, Antiquity, Göttingen 1996, 91; N. Fernández Marcos, The Septuagint in Context (translated by W. G. E. Watson), Leiden/Atlanta, GA 2000, 18-19. J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch, Chico, CA 1983, 1. Lee, Lexical Study, 1-2; A. Deissmann, Bible Studies (translated by A. Grieve), Edinburgh 1901; A. Deissmann, Light from the Ancient East (translated by L. R. M. Strachan), New York, NY/ London 1910; A. Thumb, Die griechische Sprache im Zeitalter der Hellenismus: Beiträge zur Geschichte und Beurteilung der Koiné, Strasbourg 1901 (reprint Berlin/New York, NY 1974); H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint, Cambridge 1909, 16-31; J. H. Moulton, A Grammar of New Testament Greek, vol. 1, Prolegomena (third edition), Edinburgh 1908.

306

1. Introduction

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

sentative examples of those novel features discussed by Lee and others who have investigated developments in the Koine Greek lexicon. It deals with terminology— whether “everyday” or explicitly theological—that would have been used in conjunction with various aspects of the life and thought of the Jewish community that gave rise to the Septuagint.

2. Semantic Developments in Existing Greek Terms In chapter 4 of A Lexical Study, Lee discusses cases of semantic change in words that had a history of usage prior to the creation of the Old Greek Pentateuch.

2.1 Agricultural Terms Some semantic developments had to do with the use of words in specific agricultural contexts. The term παράδεισος, for example, is first attested in reference to “the parks or pleasure-grounds of the Persian kings and nobles.” 4 These parklands contained all kinds of trees and wild animals that would have been available for hunting. By the third century BCE, however, παράδεισος had come to be used to designate a cultivated area featuring primarily fruit trees, though it could include as well other kinds of trees and plants such as vines. The closest English equivalent to it would seem to be “orchard,” 5 which is how it is rendered in the NETS Pentateuch. 6 In Septuagint Genesis, παράδεισος is the counterpart to ‫ גן‬in fourteen contexts, all of which pertain to the region of the earth in which the Creator planted fruit trees and situated the first human “to till and keep it” (2:8-15). In Septuagint Numbers, it is the rendering on one occasion for ‫ גנה‬in an oracle of Balaam, who extols Israel’s dwelling places with imagery that evokes that original idyllic setting (24:5-6). Another term that came to be used in an agricultural sense is παρίστημι. The earliest and common usages of this word involved variations on the idea of standing by/beside/near, but there is evidence in papyri from the time of the translation of the Pentateuch into Greek of the specialized sense of “be ripe/fully grown.” There is one example of this in the Pentateuch: ἡ γὰρ κριθὴ παρεστηκυῖα “For the barley was ripe” (Exod 9:31). 7 In the earliest attested occurrence of καταφυτεύω, the meaning conveyed is “transplant.” In the Old Greek Pentateuch, as is the case in contemporaneous papyri, however, the sense in which it is employed is as a synonym of the older simplex form φυτεύω “plant”: ἀμπελῶνας καὶ ἐλαιῶνας, οὓς οὐ κατεφύτευσας “vineyards and olive groves that you did not plant” (Deut 6:11); ἀμπελῶνα φυτεύσεις “You shall 4.

5. 6. 7.

Lee, Lexical Study, 53. My constant companions in carrying out the research for this article have been TLG and OLSJ. Virtually all Greek lexemes discussed have been checked in these valuable reference works. I have consulted a variety of other lexica as well, including LEH, GELS, BDAG, MM, BDB, HALOT, and Jastrow. I have accessed LEH, BDAG, BDB, and HALOT via the Accordance 9 Bible Software package. Lee, Lexical Study, 53-56. In this article, English translations of Septuagint texts are those of NETS. Lee, Lexical Study, 56-57. 2. Semantic Developments in Existing Greek Terms

307

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

plant […] a vineyard” (Deut 28:39). In both cases, the Greek is the counterpart to ‫ נטע‬in the Hebrew Bible. 8

2.2 Legal and Other Technical Terms The term ἐκδέχομαι, which generally conveys the idea of “take” or “receive,” came to be used in Koine Greek and in the papyri of that time in the sense of “be surety for.” 9 That is the case in its two occurrences in the Septuagint Pentateuch, where it quite precisely reflects the meaning of ‫ערב‬: ἐγὼ δὲ ἐκδέχομαι αὐτόν “And I am the one who is surety for him” (Gen 43:9); ὁ γὰρ παῖς σου ἐκδέδεκται τὸ παιδίον “For you servant has become surety for the child” (Gen 44:32). As Lee points out, the normal classical Greek term for this concept is ἐγγυάω. 10 A word that, with respect to the Septuagint corpus, occurs only in the Pentateuch is πάροικος. 11 The same is true of its verbal cognate, παροικέω. 12 The earliest attested usage in Greek literature of the παροικ- root in its adjectival, substantival, and verbal forms is associated with the concept of “neighbor.” Hellenistic Greek papyri and the Septuagint, however, provide evidence of a semantic development that occurred, resulting in the use of this root to convey the idea of “resident alien”: καὶ κατέβη Ἀβρὰμ εἰς Αἴγυπτον παροικῆσαι ἐκεῖ “[…] and Abram went down to Egypt to reside there as an alien” (Gen 12:10); οὐ βδελύξῃ Αἰγύπτιον, ὅτι πάροικος ἐγένου ἐν τῇ γῇ αὐτοῦ “You shall not abhor an Egyptian, because you were a resident alien in his land” (Deut 23:7[8]). 13 In more than two-thirds of these Septuagint contexts the corresponding Hebrew root is ‫גר‬, 14 while in the remaining instances it is ‫ישב‬. 15 The comparative form πρεσβύτερος, derived from the noun πρέσβυς “old man,” came to be used in the Hellenistic period not only to designate one who was either rather old or older than someone else (εἶπεν δὲ ἡ πρεσβυτέρα [‫ ]הבכירה‬πρὸς τὴν νεωτέραν Ὁ πατὴρ ἡμῶν πρεσβύτερος [‫“ ]זקן‬Now the elder said to the younger, ‘Our father is elderly’”—Gen 19:31), but also one who functioned in some kind of an official capacity. The Septuagint Pentateuch, in addition to attesting the preceding senses, employs this term for Egyptian officials 16 as well as for Israelite leaders. 17 In 8. Lee, Lexical Study, 57-58. 9. OLSJ, ἐκδέχομαι. 10. Lee, Lexical Study, 59-60. The definition “to take in charge” that is given in GELS for this word in the two contexts in Genesis that are cited above does not appear to reflect the sense that is attested in papyri contemporaneous with Septuagint Genesis and that corresponds as well with the meaning of the Hebrew counterpart (BDB, II. ‫)ערב‬. 11. Gen 15:13; 23:4; Exod 2:22; 12:45; 18:3; Lev 22:10; 25:6,23,35,40,45,47(bis); Num 35:15; Deut 14:20; 23:7. 12. Gen 12:10; 17:8; 19:9; 20:1; 21:23,34; 26:3; 32:4/5; 35:27; 37:1; 47:4,9(bis); Exod 6:4(bis); 20:10; Num 20:15; Deut 18:6; 26:5. 13. Lee, Lexical Study, 60-61. 14. Gen 12:10; 15:13; 17:8; 19:9; 20:1; 21:23,34; 23:4; 26:3; 32:4/5; 35:27; 37:1; 47:4,9(bis); Exod 2:22; 6:4 (bis); 18:3; Deut 18:6; 14:20; 23:7; 26:5. In Exod 20:10, καὶ ὁ προσήλυτος ὁ παροικῶν ἐν σοί is the translator’s rendering of ‫וגרך אשר בשעריך‬. 15. Exod 12:45; Lev 22:10; 25:6,23,35,40,45,47(bis); Num 20:15; 35:15. 16. Gen 50:7. 17. Exod 17:5; 18:12; 19:7; 24:1,14; 34:30; Lev 4:15; Num 11:16,24,25,30; 16:25; Deut 31:9,28. See H. B.

308

2. Semantic Developments in Existing Greek Terms

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

the great majority of cases it is the equivalent for ‫זקן‬, 18 though sometimes it is the counterpart to ‫ בכירה‬19 or ‫גדול‬, 20 and once to ‫בן‬. 21

2.3 Commercial Terms The primary meanings of ὁλκή in Classical Greek texts are “drawing,” “trailing,” “dragging,” “inhalation,” and “attraction.” A subsequent development, reflected in the Septuagint Pentateuch, involves the idea of “drawing down” a scale or “weight.” 22 For example, ἔλαβεν ὁ ἄνθρωπος ἐνώτια χρυσᾶ ἀνὰ δραχμὴν ὁλκῆς καὶ δύο ψέλια ἐπὶ τὰς χεῖρας αὐτῆς, δέκα χρυσῶν ὁλκὴ αὐτῶν “[…] the man took gold earrings a drachma each in weight and two bracelets for her arms, their weight being ten gold pieces” (Gen 24:22). In all fourteen occurrences in the Pentateuch, the corresponding Hebrew term is ‫משקל‬. 23 The term χρυσοῦς is used in the Pentateuch both in its adjectival sense (θεοὺς χρυσοῦς “gold gods”—Exod 20:23) and as a substantive to denote “a measure of value or weight,” which is also how it was employed in third century BCE Ptolemaic Egypt. 24 Examples of the latter denotation are found in Num 7 where each of the twelve rulers (ἄρχοντες) of Israel is depicted offering various gifts in conjunction with the dedication of the altar, including θυίσκην μίαν δέκα χρυσῶν “one censer of ten gold pieces.” 25

2.4 Terms Connoting Illness The Classical senses of ἐνοχλέω such as “trouble” and “annoy” continued to be attested in Koine Greek, but in that period the medio-passive form of the verb took on the additional meaning of “be unwell.” The Pentateuch provides one example of this later semantic development in a report that comes to Joseph regarding his elderly father, Jacob: Ὁ πατήρ σου ἐνοχλεῖται “Your father is ill” (Gen 48:1). 26 The verb μαλακίζομαι likewise experienced semantic development in Koine Greek when the earlier senses of “be softened/made effeminate,” “show weakness/cowardice,” “be softened/appeased” were augmented by “be ill.” 27 The only occurrence of this term in the Pentateuch attests this last meaning: καὶ συμβήσεται αὐτὸν μαλα-

18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Swete, An Introduction to the Old Testament in Greek (revised second edition by R. R. Ottley), New York, NY 1968, 297; Lee, Lexical Study, 61. Gen 18:11, 2; 19:4,31(2o); 24:1,2; 35:29; 43:27; 44:20; 50:7(bis); Exod 10:9; 17:5; 18:12; 19:7; 24:1,14; Lev 4:15; 19:32; Num 11:16(bis),24,25,30; 16:25; Deut 31:9,28; 32:7. Gen 19:31(1o),33,34,37; 29:26. Gen 27:1,15,42; 44:12. Exod 34:30. OLSJ, ὁλκή. Gen 24:22(bis); Num 7:13,19,25,31,37,43,49,55,61,67,73,79; Lee, Lexical Study, 62-63. Lee, Lexical Study, 63-65; OLSJ, χρύσεος. 7:14,20,26,32,38,44,50,56,62,68,74,80. Lee, Lexical Study, 66; OLSJ, ἐνοχλέω. Lee, Lexical Study, 66-67; OLSJ, μαλακίζομαι. 2. Semantic Developments in Existing Greek Terms

309

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

κισθῆναι ἐν τῇ ὁδῷ “And it will turn out that he becomes sick on the journey” (Gen 42:38).

2.5 Terms Associated with Common Activities or Concepts In Classical Greek, the various denotations of κατατείνω included “stretch,” “draw tight,” “rack,” “torture,” or in the intransitive “extend,” “run straight towards.” In the second century BCE, however, it was also employed in papyri in the sense of “overwork,” “strain,” which coincides with how the term is used in Lev 25:46: τὸν ἀδελφὸν αὐτοῦ οὐ κατατενεῖ αὐτὸν ἐν τοῖς μόχθοις “each shall not abuse his brother with toil” (cf. verses 43 and 53). 28 From earliest times, the verb σκεπάζω denoted “cover” or “shelter,” but then it also came to take on the sense of “protect,” as attested in papyri of the Ptolemaic period. Both connotations are represented in the Greek Pentateuch: ὅταν σκεπάσῃ ἡ νεφέλη ἡμέρας ἀριθμῷ ἐπὶ τῆς σκηνῆς “when the cloud casts a shadow over the tent for days in number” (Num 9:20); καὶ ὄψομαι τὸ αἷμα καὶ σκεπάσω ὑμᾶς “and I will see the blood, and I will protect you” (Exod 12:13). 29 The verb συγκρίνω was employed in a number of senses such as “combine,” “compare,” and “decide” in ancient Greek literature, but the earliest clear case discovered to date of another meaning occurs in the Septuagint of Genesis: Ἐνύπνιον εἴδομεν, καὶ ὁ συγκρίνων οὐκ ἔστιν αὐτό “We have seen a dream, and there is no one to interpret it” (Gen 40:8). 30 The core idea of the term σῶμα throughout its history of usage is “body” in its various senses, whether human or animal, living or dead, individual or corporate. A special signification attested commonly in Hellenistic papyri, however, is “slave.” The word is also employed in this sense in the Septuagint of Genesis in connection with the plundering by Jacob’s sons of the inhabitants of the city in which their sister Dinah was violated: καὶ πάντα τὰ σώματα αὐτῶν καὶ πᾶσαν τὴν ἀποσκευὴν αὐτῶν καὶ τὰς γυναῖκας αὐτῶν ᾐχμαλώτευσαν “And all their slaves and all their chattels and their wives they captured” (Gen 34:29; cf. 36:6). 31

3. New Word Formations In chapter 5 of Lee’s A Lexical Study, the focus is on the formation of new words based on existing stems. He groups them on the basis of the various strategies involved in the creation of previously unattested forms.

28. 29. 30. 31.

Lee, Lexical Study, 71-72; OLSJ, κατατείνω. Lee, Lexical Study, 76-77; OLSJ, σκεπάζω. Cf. 40:16,22; 41:12,13,15. Lee, Lexical Study, 78; OLSJ, συγκρίνω. Lee, Lexical Study, 84; OLSJ, σῶμα.

310

3. New Word Formations

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

3.1 Prepositional Compound Words New prepositional compound forms of the verb πορέυομαι that occur in the Septuagint Pentateuch have, in Koine Greek, replaced Classical Greek compounds of ἔρχομαι in the present and imperfect tenses. These include είσπορεύομαι (Exod 28:30 “enter”), ἐκπορέυομαι (Exod 33:7 “go out”), ἐπιπορέυομαι (Lev 26:33 “come by surprise”), παραπορέυομαι (Gen 32:22 “pass by”; Exod 2:5 “walk beside”), and προσπορέυομαι (Exod 28:43 “come near”). 32 Other prepositional compound verbs have been constructed on the basis of simplex forms: εἰσσπάομαι (Gen 19:10 “draw in”), ἐκδανείζω (Exod 22:24 “lend”), ἐκπορνεύω (Gen 38:24 “play the whore”; Lev 20:6 “commit fornication”), ἐκτοκίζω (Deut 23:20 “charge interest”), ἐκτρυγάω (Lev 25:5 “gather in” [grapes]), ἐνευλογέομαι (Gen 12:3 “be blessed” [passive]), ἐξαποστέλλω (Lev 16:21 “send away”), κατακενόω (Gen 42:35 “empty”), κατακυριεύω (Gen 1:28 “subdue”), κατανύσσω (Gen 34:7 “be cut to the quick” [passive]), and προσοίγω/προσοίγνυμι (Gen 19:6 “shut”). 33 Further new prepositional compound words of various types include ἀνθυφαιρέω (Lev 27:18 “make a deduction”), διοδεύω (Gen 12:6 “pass through”), ἐπαύριον (Exod 9:6 “next day”), καταγίγνομαι (Num 5:3 “dwell”), περίζωμα (Gen 3:7 “loincloth”), and συντίμησις (Lev 27:4 “valuation”). 34

3.2 Compound Words of Other Kinds Lee notes that compounds with the prefix ἀρχι- were readily created in Hellenistic Greek, and cites as examples of such terms that occur in the Greek Pentateuch ἀρχιδεσμοφύλαξ (Gen 39:21 “chief jailer”) and ἀρχιοινοχόος (Gen 40:1 “chief cupbearer”). 35 To these two could be added the following ones that first appear in extant Greek literature in the Pentateuch but whose simplex forms occur in earlier Greek: ἀρχιδεσμώτης (Gen 40:4 “chief jailer”), ἀρχιμάγειρος (Gen 37:36 “chief butcher”), ἀρχιοινοχοΐα (Gen 40:13 “chief cupbearership”), and ἀρχισιτοποιός (Gen 40:1 “chief baker”). 36 There are examples of other compounds that involve previously-attested elements 32. Lee, Lexical Study, 85-92. 33. R. J. V. Hiebert, “Textual and Translation Issues in Greek Genesis” in: C. A. Evans / J. N. Lohr / D. L. Petersen (eds.), The Book of Genesis: Composition, Reception, and Interpretation (VT.S 152, Formation and Interpretation of Old Testament Literature), Leiden/Boston 2012, 410-414; Lee, Lexical Study, 92-94. 34. Lee, Lexical Study, 94-96. 35. Lee, Lexical Study, 48-49, 96. 36. Swete, Introduction, 304; Hiebert, “Textual and Translation Issues,” 411-413. The entry for ἀρχιμάγειρος in LEH reads “chief of a royal guard, lit. chief cook (in Egypt) Gn 37:36.” In GELS, after the denotation “chief cook,” Muraoka’s comment is “of a royal court and wielding considerable authority.” There is no evidence in Greek literature from the time when the Septuagint translator of Genesis was at work, or earlier, that either μάγειρος or ἀρχιμάγειρος were used of royal security personnel or high ranking courtiers. See R. J. V. Hiebert, “Lexicography and the Translation of a Translation: The NETS Version and the Septuagint of Genesis” BIOSCS 37 (2004), 75-77. 3. New Word Formations

311

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

besides prepositions, that first occur in Koine Greek, and that are found in the Septuagint Pentateuch: 37 1. Nouns: ἀκροβυστία (Gen 17:14 “foreskin”), γραμματοεισαγωγεύς (Deut 1:15 “recorder”), δευτερονόμιον (Deut 17:18 “second law”), ἐργοδιώκτης (Exod 3:7 “taskmaster”), τοπάρχης (Gen 41:34 “district governor”), φυγαδευτήριον (Num 35:15 “place of refuge”), χοιρογρύλλιον (Lev 11:6 “coney”). 2. Adjectives: ἀνεμόφθορος (Gen 41:6 “wind-blasted”), ἑτερόζυγος (Lev 19:19 “of a different kind”), θηριάλωτος (Gen 31:39 “caught by wild beasts”), μακρόθυμος (Exod 34:6 “patient”; Num 14:18 “longsuffering”). 3. Verbs: βαρυωπέω (Gen 48:10 “be heavy-sighted”), εὐδοκέω (Lev 26:41 “be content”), κλοποφορέω (Gen 31:26 “rob”), λιθοβολέω (Lev 20:2 “stone”), νευροκοπέω (Deut 21:4 “hamstring”), πολυχρονίζω (Deut 4:26 “prolong”), σιτομετρέω (Gen 47:12 “measure out grain”).

3.3 Various Categories of Nouns An assortment of nouns that are based on earlier stems and that first occur in Hellenistic texts, including the Septuagint, may be categorized morphologically according to different terminations. 38 1. Nouns ending in -ή: ἀναζυγή (Exod 40:32[38] “journeying”), ἀποσκευή (Gen 43:8 “dependants”; Exod 12:37 “chattels”; Deut 20:14 “baggage”), ἐπισκοπή (Gen 50:24 “visitation”; Exod 3:16 “concern”; Exod 30:12 “review”; Lev 19:20 “inquiry”; Num 4:16 “oversight”; Num 14:29 “tally”; Num 26:18[22] “enrollment”). 2. Nouns ending in -μα: ἀνταπόδομα (Gen 50:15 “requital”), ἀφαίρεμα (Exod 29:27 “advance deduction”), γένημα (Lev 25:7 “yield”; Num 18:30 “produce”), δόμα (Gen 47:22 “gift”), κατάλυμα (Exod 4:24 “lodging”), λέπισμα (Gen 30:37 “stripe”), πλεόνασμα (Num 31:32 “abundance”), πρωτογένημα (Lev 23:19 “first products”), χόρτασμα (Deut 11:15 “fodder”). 3. Nouns ending in -ριον: θυσιαστήριον (Gen 8:20 “altar”), μοσχάριον (Lev 9:2 “calf”), ποτιστήριον (Gen 24:20 “watering trough”). 4. Nouns ending in -σις: κατάβρωσις (Gen 31:15 “devouring”), ὁλοκάρπωσις (Gen 22:2 “whole burnt offering”), φαῦσις (Gen 1:14 “illumination”). 5. Nouns ending in -σμός: ἀγορασμός (Gen 42:19 “purchase”), ἐμπυρισμός (Lev 10:6 “burning”), ἱματισμός (Exod 3:22 “clothing”). 6. Noun ending in -σύνη: ἐλεημοσύνη (Deut 6:25 “mercy”). 7. Nouns ending in -ών: ἀμπελών (Exod 22:4 “vineyard”), ἐλαιών (Exod 23:11 “olive grove”), πυλών (Gen 43:19 “gateway”), σιτοβολών (Gen 41:56 “granary”). 37. Swete, Introduction, 303-304, 307; Lee, Lexical Study, 96-98; Hiebert, “Textual and Translation Issues,” 410-415; M. K. H. Peters, “Deuteronomion: To the Reader” NETS, 143; E. Tov, “The Impact of the Septuagint Translation of the Torah on the Translation of the Other Books” in: E. Tov, The Greek and Hebrew Bible: Collected Essays on the Septuagint (VTSup 72), Atlanta, GA/Leiden 1999, 188, 190. 38. Swete, Introduction, 303; Lee, Lexical Study, 98-109; Tov, “Impact,” 186; Hiebert, “Textual and Translation Issues,” 410-419.

312

3. New Word Formations

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

Other nouns of different morphological types are derived from existing stems as well. These include θυίσκη (Num 7:14 “censer”), ῥοΐσκος (Exod 36:34[39:26] “little pomegranate”), and θνησιμαῖον (Lev 5:2 “carcass”). 39

3.4 Adjectives Two Hellenistic adjectival formations that appear together in two places in the Pentateuch exemplify the tendency to replace third declension forms with those of the first and second declensions. They are ἀρσενικός “male” and θηλυκός “female” (Num 5:3; Deut 4:16)—synonyms of the Classical Greek forms ἄρσην/ἄρρην and θῆλυς that continued to be used commonly into the late Koine period. Two other adjectives that are derived from earlier stems and that may occur together are δειλινός “evening” and πρωινός “morning” (Exod 29:41). 40 Additional new adjectival formations may be mentioned. Some of them function as attributive modifiers: ἀναφάλαντος (Lev 13:41 “having baldness of the forehead”), ἀπερίτμητος (Lev 26:41 “uncircumcised”), γονορρυής (Lev 15:4 “suffering from spermatorrhoea”; Num 5:2 “having a discharge”), δεκτός (Lev 1:3 “acceptable”), δικτυωτός (Exod 38:24 “latticed”), πολυέλεος (Exod 34:6 “very merciful”), σανιδωτός (Exod 27:8 “planked”), and χωνευτός (Exod 34:17 “molten”). Others are employed as substantives: ἐπαοιδός (Exod 7:11 “enchanter”), μίσθιος (Lev 25:50 “hired laborer”), and παρεπίδημος (Gen 23:4 “sojourner”). 41

3.5 Verbs Lee concludes chapter 5 of A Lexical Study by mentioning new verb formations. The verb ἀροτριάω (Deut 22:10 “plow”) is a Koine counterpart to the Classical form ἀρόω. As for ἐμπυρίζω (Lev 10:6 “be set on fire”—passive; Lev 10:16 “be burned up”—mediopassive), which is attested in papyri that date to the third century BCE, it is a synonym of ἐμπυρεύω, which began to appear in Classical texts. 42 James Moulton and George Milligan comment on the introduction of ἁγιάζω to the Greek lexicon, a verb that first appears in the Septuagint and occurs eighty-six times in the Greek Pentateuch (e. g., Exod 13:2 “consecrate”). The forms ἁγίζω “hallow” and ἁγιστεύω “perform sacred rites,” “be holy,” “purify” were used already in Classical times but never occur in the Septuagint. Moulton and Milligan suggest that ἁγιάζω was created on the basis of the term ἅγιος, which came to be used by Jews to express their particular concept of holiness, and that ἁγιάζω and its cognates 43 provided Judaism with terminology that did not have pagan associations. 44 Tov, “Impact,” 187-88. Lee, Lexical Study, 109-110; Swete, Introduction, 303. Lee, Lexical Study, 111-112; Tov, “Impact,” 186, 188, 190. Lee, Lexical Study, 113. These would include ἁγίασμα “holiness,” ἁγιασμός “consecration, sanctification,” ἁγιαστήριον “sanctuary,” and ἁγιαστία “sacred service,” all of which appear for the first time in the Septuagint. 44. MM, ἁγιάζω; Tov, “Impact,” 185; OLSJ, ἁγιάζω, ἁγίασμα, ἁγιασμός, ἁγιαστήριον, ἁγιαστία, ἁγίζω, ἅγιος, ἁγιστεύω. 39. 40. 41. 42. 43.

3. New Word Formations

313

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

Additional examples of verbs that are based on earlier stems and that first occur in the Septuagint Pentateuch include γαμβρεύω (Gen 38:8 “act the part of a brother-inlaw”; Deut 7:3 “intermarry”), ἑκατοστεύω (Gen 26:12 “bear a hundredfold”), ἐνταφιάζω (Gen 50:2 “prepare [someone] for burial”), κληδονίζομαι (Deut 18:10 “act as diviner”), ὀρθρίζω (Gen 19:2 “rise early”), and παραπικραίνω (Deut 31:27 “be fractious”). 45

4. New Words Some words were added to the Koine lexicon as loan words from other languages or came from Greek dialects other than Attic. 46 Certain words have non-Attic Greek connections, and in some cases they might also go back to terms in other languages, though such associations are tentative. For example, the Doric term βουνός (Gen 31:52 “mound”; Exod 17:9 “hill”), which was in common use in Koine Greek, may or may not have non-Greek origins. The verb γογγύζω (Num 14:29 “complain”) is reputed by the second century CE grammarian Phrynichus to be Ionic, 47 though its etymology too is less than certain. Seemingly not a loan word and first attested in the third century BCE is κάρταλλος (Deut 26:2 “basket”). The term μάρσιππος (Gen 42:27 “bag”) is not Attic in origin and may have come to Koine via Ionic. As for loan words, the term γ(ε)ιώρας (Exod 12:19 “giora”) is derived from the Aramaic ‫“ גיורא‬stranger,” “proselyte.” 48 The word θῖβις (Exod 2:3 “basket”) comes either from the Hebrew term ‫ תבה‬or from an Egyptian original. 49 Like many words that end in –υ, κόνδυ (Gen 44:2 “cup”) is evidently a loan word, though it is of unknown origin. The term κόρος (“kor”) is a loan word, and its Hebrew/Aramaic cognate is ‫ כר‬as is evident when one compares the Septuagint with the Hebrew Bible in relevant contexts outside the Pentateuch. 50 In the case of both occurrences in the Old Greek Pentateuch (Lev 27:16; Num 11:32), however, the Hebrew counterpart is ‫חמר‬, which in postexilic times was equated with the ‫כר‬. 51 Other Semitic loan words include πάσχα and σάββατα, derived from Aramaic/Hebrew ‫פסח‬/‫ פסחא‬and ‫שבת‬/‫שבתא‬, respectively. 52 Curiously, σάββατα is construed both as a singular and a plural form in the Septuagint Pentateuch: ἓξ ἡμέρας ποιήσεις ἔργα, τῇ δὲ ἡμέρᾳ τῇ ἑβδόμῃ κα45. Hiebert, “Textual and Translation Issues,” 411-416; Hiebert, “Lexicography and the Translation of a Translation,” 82-83; Tov, “Impact,” 188, 190. 46. Swete, Introduction, 295-297, 307; Lee, Lexical Study, 114-117. 47. Phrynichus Atticista, Eclogae 335. 48. LEH, γειώρας; OLSJ, γειώρας; Jastrow, ‫ גיורא‬,‫ ;גיור‬Thackeray, Grammar, § 4, p. 34; Lee, Lexical Study, 16, 52. 49. Lee, Lexical Study, 115; OLSJ, θῖβις; BDB, ‫ ;תבה‬HALOT, ‫תבה‬. 50. See 3Reigns(1Kings) 5:2/4:22(bis), 5:25/11; 2Chron 2:9/10(bis); 27:5; 2Esdras(Ezra) 7:22; OLSJ, κόρος (D); BDB, ‫ ;כר‬HALOT, ‫ ;כר‬Lee, Lexical Study, 116-17. The kor is a measure that is equivalent to 220 liters (see NETS, note to Lev 27:16). 51. M. A. Powell, “Weights and Measures” ABD 6, 897-908. The κόρος – ‫ חמר‬equivalence is also found in Ez 45:13. 52. LEH, πάσχα, σάββατον; BDAG, πάσχα, σάββατον; Jastrow, ‫שובתא‬, ‫שבתא‬, ‫שבתא‬, ‫גסחא‬, ‫;פסח‬

314

4. New Words

20. Innovative Elements in the Vocabulary of the Septuagint Pentateuch

τάπαυσις, ἅγιον, σάββατα, ἀνάπαυσις κυρίῳ “Six days you shall do works, but on the seventh day there is a rest, something holy, Sabbata, a repose for the Lord (Exod 35:2); σάββατα σαββάτων ἀνάπαυσις αὕτη ἔσται ὑμῖν “This rest shall be Sabbata of sabbaths to you” (Lev 16:31).

5. Lexical Intrusion and Concomitant Obsolescence Lee’s treatment of the topic of semantic developments in Koine Greek continues in chapter 7 of A Lexical Study with the consideration of examples of a phenomenon involving the intrusion of words into the lexicon and the corresponding discontinuation of the use of other words. One such case pertains to the Classical terms ἄρδω “water,” “give drink to,” “irrigate” and ἀρδεύω “water,” “irrigate,” 53 which by the third century BCE had been replaced in vernacular Greek by ποτίζω. 54 This development is reflected in the Septuagint by virtue of the fact that ἄρδω and ἀρδεύω are never found in that corpus, whereas ποτίζω is attested sixty-five times, twenty-eight occurrences of which are in the Pentateuch (e. g., Gen 2:10 “water” [orchard]; Gen 21:19 “give [someone] a drink”; Gen 29:10 “water” [sheep]; Exod 32:20 “make [someone] drink”).

6. Conclusion This survey of innovative elements in the vocabulary of the Septuagint Pentateuch is, of course, not comprehensive in scope. Instead, it provides representative examples of the kinds of developments in the Greek lexicon to which that translation corpus bears witness. It has been shown that many of the lexical changes that followed the cultural tsunami generated by Alexander the Great were in fact based on stems that existed already in the Classical period. As commonly occurs in living languages, existing words acquired new components of meaning. Koine writers created compounds of various kinds, some of which involved semantic change in comparison to the forms upon which they were based and others of which did not. Completely new terms were brought into the vernacular from other Greek dialects or even different languages. One of the interesting aspects of the study of the Septuagint Pentateuch is the fact that it provides data to the linguist for all these kinds of lexical phenomena.

BDB, ‫ ;פסח‬HALOT, ‫ ;פסח‬Thackeray, Grammar, § 4, p. 35; Lee, Lexical Study, 16, 30, 52; L. Perkins, “Exodus: To the Reader,” NETS, 45; D. Büchner, “Leuitikon: To the Reader,” NETS, 82. 53. OLSJ, ἄρδω, ἀρδεύω. 54. Lee, Lexical Study, 118-122. 5. Lexical Intrusion and Concomitant Obsolescence

315

21. The Septuagint Translation of the Torah as a Source and Resource for the Post-Pentateuchal Translators* Emanuel Tov

According to ancient evidence as well as modern descriptions, the translation of the Torah preceded that of the later books. As might be expected, this translation influenced those that were prepared subsequently, 1 although this assumption cannot be substantiated for all the post-Pentateuchal translations. The Greek Torah was probably used by Jews in Alexandria in their weekly ceremonial reading from the first century BCE onwards. Philo refers to this custom in Alexandria 2 and 4Mac 18:10-18, possibly written in Egypt in the first century CE, alludes to the reading of the Law together with reflections taken from the Prophets, *

1.

2.

This study was originally published as “The Impact of the LXX Translation of the Pentateuch on the Translation of the Other Books” in: P. Casetti / O. Keel / A. Schenker (eds.), Mélanges Dominique Barthélemy (Orbis biblicus et orientalis 38) Fribourg/Göttingen, 1981, 577592. A slightly revised version is found in my The Greek and Hebrew Bible: Collected Essays on the Septuagint (Supplements to Vetus Testamentum 72), Leiden/Boston, MA/Cologne, 1999, 183-194. Thus H. St. J. Thackeray, “The Greek Translators of the Prophetical Books” JTS 4 (1903), 578585, in particular 583; M. Flashar, “Exegetische Studien zum Septuagintapsalter” ZAW 32 (1912), 183-189; J. Ziegler, Untersuchungen zur Septuaginta des Buches Isaias (Alttestamentliche Abhandlungen XII, 3), Münster i. W., 1934, 134-175; G. Gerleman, Studies in the Septuagint, II, Chronicles (Lunds universitets årsskrift I, 43, 3), Lund, 1946, 22-23; I. L. Seeligmann, The Septuagint Version of Isaiah, Leiden, 1948, 45-49; L. C. Allen, The Greek Chronicles, vol. I (Supplements to Vetus Testamentum 25), Leiden, 1974, 23-26, 57-59. The following studies published after the publication of my initial paper (1981) corroborated this assumption with additional examples: G. B. Caird, “Ben Sira and the Dating of the Septuagint” in: E. A. Livingstone (ed.), Studia Evangelica, vol. VII, Berlin 1982, 95-100; C. G. den Hertog, Studien zur griechischen Übersetzung des Buches Josua, Ph.D. dissertation, Universität Giessen, 1996, 111-125 (Den Hertog, pages 124-125, suggests that the translator of Joshua used the translation of Deuteronomy); J. Joosten, “The Impact of the Greek Pentateuch on the Greek Psalms” in: M. K. H. Peters (ed.), XIII Congress of the International Organization for Septuagint and Cognate Studies, Ljubliana, 2007 (Society of Biblical Literature Septuagint and Cognate Studies 55), Atlanta, GA 2008, 197-205 (from Joosten’s examples we quote ἐγκισσάω, ἐνδιαβάλλω, and λοχεύομαι, as well as Ps 116 [114]:9 quoted in his name). Philo, Prob. 81-82: “They use these laws hthose of the Torahi to learn from at all times, but especially each seventh day, since the seventh day is regarded as sacred. On that day they abstain from other work and betake themselves to the sacred places which are called synagogues … Then one of them takes the books and reads.” See further Philo, Hypoth. 7:13; Moses 2:215. The existence of Greek Torah scrolls is also referred to in m. Meg. 1.8; 2.1 and t. Meg. 4.13. See further A. Wasserstein / D. Wasserstein, The Legend of the Septuagint: From Classical Antiquity to Today, Cambridge 2006, 11-12.

316

6. Conclusion

21. The Septuagint Translation of the Torah as a Source and Resource

Psalms, and Proverbs. The Torah also must have been widely known in Greek. In the following discussion, evidence is presented in support of the following four points: 3 (1) the vocabulary of the Greek Torah was maintained in the translation of the later books; (2) the Greek Torah served as a lexicon for the later translators who often turned to that translation when encountering difficult Hebrew words; (3) quotations from and allusions to the Torah in the later books were sometimes phrased in a manner identical to that used in the translation of the Torah; (4) the contents of the Greek Torah often influenced the wording of later translations on an exegetical level.

1. Source for Vocabulary The translators of the Torah created a translation vocabulary of Hebrew-Greek equivalents, the foundations of which were probably laid in the generations that preceded that translation. 4 The nature of this vocabulary has been analyzed in several studies, 5 and it is safe to say that one of its main characteristics is the lack of variation, a lack caused by the translators’ frequent use of fixed equivalents. The five books of the Greek Torah were rendered by different translators who shared a common translation vocabulary; however, they showed their individuality in certain translation options and peculiarities. 6 When investigating agreements between the vocabulary of the Greek Torah and that of the later books, obvious agreements such as ‫ –אשה‬γυνή, ‫ – איש‬ἄνθρωπος, ‫– שמש‬ 3.

4. 5.

6.

Criticisms of my original study (1981) have been offered by J. Lust and J. Barr. J. Lust, “The Vocabulary of LXX Ezekiel and its Dependence upon the Pentateuch” in: M. Vervenne / J. Lust (eds.), Deuteronomy and Deuteronomic Literature, Festschrift C. H. W. Brekelmans (Bibliotheca ephemeridum theologicarum lovaniensium 123), Leuven, 1997, 529-546 focused on the relation between Ezekiel and the Septuagint translation of the Torah (especially Leviticus) and therefore has a narrow perspective. His criticisms are addressed below. J. Barr, “Did the Greek Pentateuch Really Serve as a Dictionary for the Translation of the Later Books?” in: M. F. J. Baasten / W. Th. van Peursen (eds.), Hamlet on a Hill. Semitic and Greek Studies Presented to Professor T. Muraoka on the Occasion of his Sixty-Fifth Birthday (Orientalia Lovaniensia Analecta 118) Leuven, 2003, 523-543 offered important and detailed criticisms that are likewise answered below. See my study “Studies in the Vocabulary of the Septuagint” Tarbiz 47 (1978), 120-138, in particular 137-138 (Hebrew with English summary). See S. Daniel, Recherches sur le vocabulaire du culte dans la Septante, Paris, 1966; J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (Society of Biblical Literature Septuagint and Cognate Studies 14), Chico, CA, 1996. A. Passoni dell’Acqua stressed the Egyptian background of the LXX vocabulary in a long series of studies on individual words appearing in different books of the LXX, e. g. “La versione dei LXX e i papyri: note lessicali” in: R. S. Bagnall et al. (eds.), Proceedings of the Sixteenth International Congress of Papyrology, New York, 24–31 July 1980, Chico, CA, 1981, 621-62; “Notazioni cromatiche dall’Egitto greco-romano. La versione del LXX e i papiri” Aegyptus 78 (1998) 77-115. See further the bibliography given by M. Harl, “La langue de la Septante,” in: G. Dorival / M. Harl / O. Munnich, La Bible grecque des Septante: Du judaïsme hellénistique au christianisme ancien, Paris, 1988, 243. See H. Kim, Multiple Authorship of the Septuagint Pentateuch, Ph.D. dissertation, Hebrew University, 2007. A summary was published in the Bulletin of Judaeo-Greek Studies 40 (2007), 2-3. 1. Source for Vocabulary

317

21. The Septuagint Translation of the Torah as a Source and Resource

ἥλιος and ‫ – צפרדע‬βάτραχος, are disregarded. The analysis concentrates on agreements that are idiosyncratic, and illustrates how the translation vocabulary of the Torah was maintained in the translations of the other units. In a study of the vocabulary of the Septuagint books, the following points are taken into consideration: 1. Although the degree of dependence of the post-Pentateuchal books on the vocabulary of the Torah cannot be expressed in absolute statistical terms, pilot investigations have shown that the vocabulary of certain books is more “Pentateuchal” than others. For example, G. Gerleman described the vocabulary of Chronicles as more “Pentateuchal” than that of the parallel translations of Samuel-Kings, especially as regards its rendering of cultic terms. 7 J. Lust noted that the Septuagint translation of Ezekiel is less “Pentateuchal” than would otherwise have been expected. 8 2. The post-Pentateuchal books were translated by different individuals in Palestine and Egypt, 9 who, despite their differences, 10 all adhered to some extent to the vocabulary of the Greek Torah. 11 One is therefore justified in investigating the influence of its vocabulary on that of the later translations. 3. The dependence of the later translators on the vocabulary of the Septuagint translation of the Torah was inconsistent 12 since it was based on each translator’s memory rather than an organized list of equivalents. 13 In this way post-Pentateuchal translators often developed new translation vocabularies. 14 Several of the equivalents listed below are not the main ones found in the Septuagint translation of the Torah,

7. Gerleman, Chronicles, 22 8. Note especially the concluding remarks of Lust, “Vocabulary,” 545-546. 9. For a detailed analysis, see my study “Reflections on the Septuagint with Special Attention Paid to the Post-Pentateuchal Translations” in: W. Kraus / M. Karrer (eds.), Die Septuaginta – Texte, Theologien, Einflüsse: 2. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 23.–27. 7. 2008 (Wissenschaftliche Untersuchungen zum Neuen Testament 252), Tübingen, 2010, 3-22. 10. These differences justify our neglect of discrepancies between the individual translations. Little can be learned from disagreements in vocabulary between different translators (even in whole verses or sections that are identical in the Hebrew Bible) apart from their evident lack of cooperation with one another and their failure to consult other translation units. See notes 14, 15. 11. See Tov, “Studies in the Vocabulary” and E. Tov, “Three Dimensions of LXX Words” in: Tov, Greek and Hebrew Bible, 85-94. 12. The importance of the lack of consistency in the scribal transmission and translation enterprises was stressed in my study “Some Reflections on Consistency in the Activity of Scribes and Translators” in: U. Dahmen / J. Schnocks (eds.), Juda und Jerusalem in der Seleukidenzeit. Herrschaft – Widerstand – Identität: Festschrift für Heinz-Josef Fabry (Bonner biblische Beiträge 159), Göttingen, 2010, 325-337. 13. This point was probably not clear enough in my original publication, since Barr, “Greek Pentateuch,” 524 described at length his “doubts about the ‘dictionary’ idea.” For this reason, I presume, Barr often refers to my neglect of “negative evidence” (e. g. 536), that is, instances in which the translators did not consult the Septuagint translation of the Torah, or culled from that translation an unusual equivalent. See note 14 below. 14. Pace Lust, “Vocabulary,” 533. For a discussion of the differences between the vocabulary of the Septuagint translation of the Torah and that of the post-Pentateuchal books, see Kim, Multiple Authorship, 54-58.

318

1. Source for Vocabulary

21. The Septuagint Translation of the Torah as a Source and Resource

which sometimes used a variety of equivalents, 15 but for some reason they came to the translator’s mind. The post-Pentateuchal translators did not try to copy systematically all the elements in the Septuagint translation of the Torah, when such borrowing could be accomplished, because they habitually made no attempt to achieve consistency, even within translation units. 16 In the examination of the dependence of the later translations on the vocabulary of the Greek Torah, the examples are subdivided into four categories: religion, legal terminology, central biblical terms, and miscellaneous. The non-exhaustive lists below concentrate on equivalents occurring several times in the Torah. The following conventions are used: (1) when there are two or more occurrences in one book, only one reference, followed by “etc.,” is given; (2) when there are two occurrences in different books, precise references are given for each case; (3) if there are more than two occurrences in different books, p. (= passim) is used. Words denoted with an asterisk (*) were presumably coined by the Septuagint translators (“neologisms”). 17 Several of these neologisms were maintained in the later books, a situation which underlines the dependence of the latter on the former. Finally, it should be noted that the inclusion in the list of certain equivalents does not imply that the Greek word mentioned is the only equivalent used for the Hebrew one.

1.1 Religion 1.1.1

General

ἀγχιστεύω ἁγιαζ-

‫ גאל‬p ‫ נזיר‬,‫ נזר‬Lev 25:11; Num 6:12

p p

ᾅδης

‫ שאול‬p

p

ἀκαθαρσία

‫ טמאה‬p

p

ἀκαθαρτός

‫ טמא‬p

p

15. This argument was used by Lust, “Vocabulary,” 533-537 and Barr, “Greek Pentateuch,” 526530, criticizing my original publication (1981). 16. Pace Lust (“Vocabulary,” 543-546), who looks for such consistency between Leviticus 26 and the Septuagint translation of Ezekiel. At an earlier stage, upon considering the differences between the Septuagint translations of the Pentateuch and Joshua, C. Egli (“Zur Kritik der Septuaginta. Sind die Hermeneuten des Pentateuch und des Buches Josua identisch?” ZWT 5 (1862), 76-96, 287-321) suggested that the two translations were produced by different (groups of) individuals. 17. These words were probably coined by the translators (or a preceding generation) in order to express Hebrew words and concepts that, in their view, could not be expressed adequately by existing Greek words. Our observations are based on the evidence listed by H. G. Liddell / R. Scott / H. S. Jones, A Greek-English Lexicon (ninth edition) Oxford, 1940; E. A. Barber, A Greek-English Lexicon, A Supplement, Oxford, 1968; and P. G. W. Glare, Revised Supplement, Oxford, 1996. The assumption of a “neologism” is subject to limitations and doubts such as those described in E. Tov, “Compound Words in the LXX Representing Two or More Hebrew Words” in: Tov, Greek and Hebrew Bible, 131-152, in particular 139-141. 1. Source for Vocabulary

319

21. The Septuagint Translation of the Torah as a Source and Resource

ἀνάθημα (-εμα) ἀπαρχή ἀφαίρεμα* 18

‫ חרם‬p

p

‫ ראשית‬p

p

‫ תרומה‬p

p

‫ תרומה‬p

Ez 44:30

γλυπτόν

‫ פסל‬p

p

δεκτός*

‫ לרצון‬p

p

διαθήκη

‫ ברית‬p

p

δῶρον

‫ מנחה‬Gen 4:4 etc.

p

‫ קרבן‬p

Neh 13:31

‫ שחד‬p

p

εἴδωλον

‫ גלולים‬Lev 26:30; Deut 29:16

p

ἑκούσιον

‫ נדבה‬p

p

ἐξιλάσκομαι

‫ כפ״ר‬p

p

ἐξόδιον

‫ עצרת‬p

2Chr 7:9; Neh 8:18

ἐπιδέκατον

‫ מעשר‬p

p

ἐπιτίθημι

(‫ סמך )ידים‬p

2Chr 29:23

‫ הניף‬p

2Kings 5:11

θυσιαστήριον*

‫ מזבח‬p

p

κιβωτός

‫ אר ו ן‬p

p

λειτουργέω

‫ שרת‬p

p

νόμος

‫ תורה‬p

p

ὁμολογία

‫ נדבה‬Deut 12:17

Ez 46:12

ὀσμὴ εὐωδίας πλημμέλεια

‫ ריח ניחח‬p

Ez 6:13 etc.

‫ אשם‬p

p

προσκυνέω

‫ השתחוה‬p

p

πρωτογενήματα*

‫ בכורים‬p

2Kings 4:42; Neh 10:36

πρωτότοκος*

‫ בכור‬p

p

στήλη

‫ מצבה‬p

p

συναγωγή

‫ עדה‬p

p

σωτήριον

‫ שלם‬p

p

χειροποίητος

‫ אליל‬Lev 26:1

Isa 2:18, etc.

χωνευτός*

‫ מסכת‬p

p

18. Lust, “Vocabulary,” 534 points out that Pap. 967 of Ezekiel uses ἀφόρισμα instead of ἀφαίρεμα, suggesting that in this case, as well as in others, the equivalent of the main manuscript tradition of the Septuagint translation may have been secondary.

320

1. Source for Vocabulary

21. The Septuagint Translation of the Torah as a Source and Resource

Furthermore, the translators of Joshua and Isaiah adopted the distinction between the Jewish ‫( מזבח‬θυσιαστήριον) and the pagan ‫( מזבח‬βωμός) made by the translators of the Torah. 19

1.1.2 Technical Cultic Terms A natural source of information for the rendering of technical terms would be the Greek Torah. The post-Pentateuchal occurrences of these words are not necessarily found in religious contexts, e. g. κρατήρ in Song 7:3. γωνία

‫ מקצע‬Exod 26:24

2Chr 26:9; Neh 3:19-25

δέρρις

‫ יריעה‬Exod 26:7-13; Num 4:25

p

διχοτόμημα*

‫ נתח‬Exod 29:17; Lev 1:8

Ez 24:4

θυίσκη*

‫ כף‬p

p

θυμίαμα

‫ קטרת‬p

p

καταπέτασμα

‫ פרכת‬p

2Chr 3:14

κίδαρις

‫ מצנפת‬Exod 28:4 etc.; Lev 16:4

Ez 21:31 (26)

κρατήρ κρεάγρα λαβίς λάγανον

‫ אגן‬Exod 24:6 ‫ מזלג‬Exod 27:3 etc.; Num 4:14 ‫ מלקחים‬Exod 37:23 (38:17); Num 4:9 ‫ רקיק‬p

Song 7:3 p Isa 6:6; 2Chr 4:21 1Chr 23:29

λέβης

‫ סיר‬Exod 16:3

p

λουτήρ

‫ כיור‬Exod 30:18 etc.; Lev 8:11

p

λυχνία μέλος

‫ מנורה‬p ‫ נתח‬Exod 29:17; Lev 1:6 etc.

p Judg 19:29B; Ez 24:6

μέτρον

‫ איפה‬Deut 25:14, 15

p

μηρία

‫ כסל‬Lev 3:4 etc.

Job 15:27

μυρεψός

‫ רקח‬Exod 30:25 etc.

πέταλον

‫ ציץ‬Exod 28:36 etc.; Lev 8:9

πλήρωσις

‫ מלאים‬Exod 35:27

1Kings 6:32,35 1Chr 29:2

πυγμή

‫ אגרף‬Exod 21:18

Isa 58:4

πυρεῖον

‫ מחתה‬p

2Kings 25:15; 2Chr 4:21/22

ῥοίσκος*

‫ רמון‬Exod 28:33/29 etc.

2Chr 3:16 etc.

σάρδιον

‫ אדם‬Exod 28:17 etc.

Ez 28:13

σεμίδαλις

‫ סלת‬p

P

σμύρνα

‫ מר‬Exod 30:23

p

19. For an analysis, see Daniel, Recherches, 18-22. 1. Source for Vocabulary

321

21. The Septuagint Translation of the Torah as a Source and Resource

στίχος

‫ טור‬Exod 28:17 etc.

1Kings 6:36 etc.

σύνθεσις

‫ סמים‬p

2Chr 13:11

τήγανον

‫ מחבת‬Lev 2:5 etc.

Ez 4:3; 1Chr 23:29

τορευτός

‫ מקשה‬Exod 25:18 (17) etc.

Jer 10:5 (9)

1.2 Legal Terminology The post-Pentateuchal occurrences of these words are not necessarily found in legal contexts. (τὸ) βιβλίον (τοῦ) ἀποστασίου αὐτόχθων

‫ ספר‬Deut 24:1, 3 ‫כריתות‬

Isa 50:1; Jer 3:8

‫ אזרח‬p

Josh 8:33 (9:2); Ez 47:22

γειώρας

‫ גר‬Exod 12:19

Isa 14:1

γονορρυής*

‫ זוב‬Lev 15:4 etc.; Num 5:2

p

ἐγγαστρίμυθος

‫ אוב‬Lev 19:31 etc.; Deut 18:11

p

ἐνεχυράζω

‫ חבל‬p

p

ἐπαοιδός*

‫ ידעני‬Lev 19:31 etc.

2Chr 33:6

θνησιμαῖον*

‫ נבלה‬Lev 5:2 etc.; Deut 14:8, 21

p

κληδονίζομαι*

‫ עונן‬Deut 18:10

2Kings 21:6; 2Chr 33:6

λέπρα + derivatives

‫ צרעת‬Lev 13–14; Deut 24:8

p

λιθοβολέω*

‫ סק״ל‬p

p

‫ רג״ם‬p

p

μαντεία, μαντεῖον

‫ קסם‬p

p

μῶμος

‫ מום‬p

p

πρᾶσις

‫ ממכר‬Lev 25:14 ff.; Deut 18:8

Neh 13:20

συντίμησις

‫ ערך‬Lev 27:4,18; Num 18:16

φαρμακός*

‫ מכשף‬Exod 7:11 etc.; Deut 18:10

p

φυγαδευτήριον*

‫ מקלט‬Num 35:6 ff.

Josh 20:2 etc.; 1Chr 6:42, 52

2Kings 12:5 etc.

Appendix: Clean and Unclean Animals δορκάς ἔλαφος

‫ צבי‬Deut 12:15 etc. ‫ אילה‬,‫ איל‬Deut 12:15, 22 etc.

p p

ἶβις

‫ ינשוף‬Lev 11:17

Isa 34:11

μῦς

‫ עכבר‬Lev 11:29

1Sam 6:4 ff.; Isa 66:17

322

1. Source for Vocabulary

21. The Septuagint Translation of the Torah as a Source and Resource

νυκτερίς

‫ עטלף‬Lev 11:19; Deut 14:18

Isa 2:20

νυκτικόραξ

‫ כוס‬Lev 11:17

Ps 102 (101):7

χοιρογρύλλιος*

‫ שפן‬Deut 14:7

Ps 104 (103):18; Prov 30:26 (24:61)

1.3 Central Biblical Terms Many of the words in the Torah pertain to central stories or issues to which reference is made in the post-Pentateuchal books (note especially Psalm 78 [77]). The later translators often used the same Greek equivalents as those used in the Torah: διαγογγύζω*

‫ נלון‬,‫ הלין‬p

διέξοδος

‫ תוצאות‬Num 34:4 ff.

Josh 9:18 Josh 15:4 etc.; Ps 68 (67):21

τὰ ἑρπετά

‫ רמש‬p

p

κατακλυσμός

‫ מבול‬Gen 6:17 etc.

Ps 29 (28):10

κυνόμυια

‫ ערב‬Exod 8:17 (21) etc.

Ps 78 (77):45; 105 (104):31

μαννα*

‫ מן‬p

ὀρτυγομήτρα παράδεισος

‫ שלו‬Exod 16:13; Num 11:31, 32 ‫ גן עדן‬Gen 2:8 etc.

πλάξ

‫ לוח‬Exod 31:18 etc.; Deut 4:13 etc.

Ps 78 (77):24 Ps 105 (104):40 p 1Kings 8:9; 2Chr 5:10

στερέωμα

‫ רקיע‬Gen 1:6 ff.

p

φαῦσις*

‫ מאור‬Gen 1:15

Ps 74 (73):16

1.4 Miscellaneous 1.4.1 Technical Terms ἀμνός, ἀμνάς 20

‫ קשיטה‬Gen 33:19

Josh 24:32; Job 42:11

ἀνεμοφθορία*

‫ שדפון‬Deut 28:22

2Chr 6:28

βούτυρον

‫ חמאה‬Gen 18:8; Deut 32:14

p

δίδραχμον

‫ שקל‬p

Josh 7:21; Neh 5:15 etc.

δοκός

‫ קורה‬Gen 19:8

2Kgs 6:2, 5; Song 1:17

ἐγκρυφίας

‫ עוגה‬p

p

ἐδέσματα

‫ מטעמים‬Gen 27:4 etc.

ἐνώτιον

‫ נזם‬Gen 24:22 etc.; Exod 32:2 etc.

Prov 23:3 p

20. The Hebrew denotes an “ancient weight, used as money” (HALOT), while the Septuagint, as well as the other versions (except for Symmachus), translates the word as “lamb.” 1. Source for Vocabulary

323

21. The Septuagint Translation of the Torah as a Source and Resource

ἐπαοιδός

‫ חרטם‬Exod 7:11 etc.

Dan 2:2, 27

ἔπαυλις

‫ טירה‬Gen 25:16; Num 31:10

Ps 69 (68):26

ἐπισιτισμός

‫ צדה‬Gen 42:25 etc.; Exod 12:39

p

λάχανα

‫ ירק‬Gen 9:3

Ps 37 (36):2

μανδραγόρας μόλιβ(δ)ος μονόκερως μώλωψ

‫ דודאים‬Gen 30:14 ff. ‫ עפרת‬Exod 15:10 ‫ ראם‬Num 23:22; Deut 33:17 ‫ חבורה‬Gen 4:23; Exod 21:25

Song 7:14 p Job 39:9; Ps 22 (21):22 etc. Isa 1:6 etc.; Ps 38 (37):6

νυμφαγωγός

‫ מרע‬Gen 26:26

Judg 14:20

ὀβολός

‫ גרה‬p

Ez 45:12

οἰφι*

‫ איפה‬Lev 5:11 etc.; Num 5:15 etc.

p

ὄλυρα

‫ כסמת‬Exod 9:32

Ez 4:9

ὄψιμος

‫ מלקוש‬Deut 11:14

p

παγετός

‫ קרח‬Gen 31:40

Jer 36 (43):30

περιβόλαιον

‫ כסות‬Exod 22:26; Deut 22:12

Isa 50:3; Job 26:6

περίχωρος

‫ ככר‬Gen 13:10 etc.

2Chr 4:17; Neh 12:28

πυρράκης

‫ אדמוני‬Gen 25:25

σπαρτίον

‫ חוט‬Gen 14:23

p

τιθηνός

‫ אמן‬Num 11:12

p

τροφός

‫ מינקת‬Gen 35:8

1Sam 16:12 etc.

p

τρυγών

‫ ת ור‬p

Jer 8:7; Song 1:10 etc.

ὕσσωπος

‫ אזוב‬p

1Kings 5:13; Ps 51 (50):9

φακός

‫ עדשה‬Gen 25:34

2Sam 17:28 etc.; Ez 4:9

χείρ

‫ חפן‬Exod 9:8; Lev 16:12

Ez 10:2, 7

χολή

‫ ראש‬Deut 29:17 etc. ‫רוש‬

p

χόρτασμα

‫ מספוא‬Gen 24:25 etc.

Judg 19:19

χύτρα

‫ פארור‬Num 11:8

p

ψαλτήριον

‫ כנור‬Gen 4:21

Ps 49 (48):5 etc. Ez 26:13

ψέλ(λ)ιον

‫ צמיד‬Gen 24:22 etc.; Num 31:50

Ez 16:11 etc.

ψόγος

‫ דבה‬Gen 37:2

Jer 20:10; Ps 31 (30):14

324

1. Source for Vocabulary

21. The Septuagint Translation of the Torah as a Source and Resource

1.4.2 Other Words ἀντιλίβανον 21 γένεσις (-εις) γραμματεύς δῆμος

‫ לבנון‬Deut 1:7 etc. ‫ תולדות‬p

p

‫ שטר‬p

p

‫ משפחה‬Num 1:20 etc.

διήγημα

‫ שנינה‬Deut 28:37

ἐγκισσάω

‫ יחם‬Gen 31:10

ἐκλεκτός

‫ בריא‬Gen 41:2 ff.

ἐνδιαβάλλω κατάσχεσις λοχεύομαι μακρόθυμος*

Josh 9:1

‫ שטן‬Num 22:22 ‫ אחזה‬p

p 2Chr 7:20 Ps 51 (50):7 (κισσάω) p Ps 38 (37):21 etc. p

‫ עול‬Gen 33:13

Ps 78 (77):1

‫ ארך‬Exod 34:6; Num 14:18

p

‫אפים‬ μόχθος

‫ תלאה‬Exod 18:8; Num 20:14

Neh 9:32

ὁλκή

‫ משקל‬Gen 24:22; Num 7:13 etc.

p

ὀρθρίζω*

‫ השכים‬p

p

ὀσφύς

‫ חלצים‬Gen 35:11

p

‫ בלל בלה‬Deut 29:4

p

‫ המרה‬Deut 31:27

p

παλαιόω παραπικραίνω* πολυέλεος* ποταμός σημασία σκληροτράχηλος*

‫ רב חסד‬Exod 34:6; Num 14:18 ‫ יאר‬Gen 41:1 ff.; Exod 1:22 etc. ‫ תרועה‬Num 10:5 etc. ‫ קשה ערף‬Exod 33:3 etc.; Deut 9:6 ff.

p p 1Chr 15:28 etc.; Ezra 3:12 ff. Prov 29:1

2. Lexical Source for Difficult Words There is no concrete evidence that the translators possessed either dictionaries or word lists. Thus, when attempting to determine the meaning of a word, they resorted to various sources of information. These ranged from exegetical traditions, context, etymology, post-biblical Hebrew, knowledge of Aramaic, to the Septuagint translation of the Torah. The latter was consulted when the translators encountered difficult Hebrew words that also occurred in the Torah. 22 It is unknown how frequently this was done. Obviously, this hypothesis assumes that the learned post-Pentateuchal translator knew that the Hebrew word to translate occurred in a specific context in the Torah. The 21. See den Hertog, Josua, 122-124. 22. Thus already Flashar, Septuagintapsalter; Seeligmann, Isaiah, 48. 2. Lexical Source for Difficult Words

325

21. The Septuagint Translation of the Torah as a Source and Resource

nature of this type of consultation of the Torah necessarily implies that it was executed extremely inconsistently. This procedure is exemplified by the following examples: 1. Deut 32:42 Judg 5:2A Num 5:18 Judg 5:2B

‫ – מראש פרעות אויב‬ἀπὸ κεφαλῆς ἀρχόντων ἐχθρῶν ‫ – בפרע פרעות‬ἐν τῷ ἄρξασθαι ἀρχηγούς ‫ – ופרע‬καὶ ἀποκαλύψει ‫ – בפרע פרעות‬ἀπεκαλύφθη ἀποκάλυμμα

The A and B texts in Judges are based on different interpretations of ‫בפרע פרעות‬, both of which are reflected in the Septuagint translation of the Torah. 2. Lev 5:4 Num 30:7 Ps 106 (105):33

[…] ‫ – לבטא בשפתים‬διαστέλλουσα τοῖς χείλεσιν […] διαστείλῃ ‫יבטא‬

‫ – מבטא שפתיה‬κατὰ τὴν διαστολὴν τῶν χειλέων αὐτῆς

‫ – ויבטא בשפתיו‬καὶ διέστειλεν ἐν τοῖς χείλεσιν αὐτοῦ

‫ בטא‬occurs elsewhere only in Prov 12:18 ‫ – בוטה‬λέγοντες. 3. Deut 4:27

‫ – מתי מספר‬ὀλίγοι ἀριθμῷ

Jer 44 (51):28

‫ – מתי מספר‬ὀλίγοι ἀριθμῷ

Gen 34:30

‫ – מתי מספר‬ὀλιγοστός

1Chr 16:19

‫ – מתי מספר‬ὀλιγοστούς

Ps 105 (104):12

‫ – מתי מספר‬ὀλιγοστούς

The same phrase is rendered differently elsewhere. 23 4. Exod 1:11 2Chr 8:4, 6; 17:12

‫ – ערי מסכנות‬πόλεις ὀχυράς ‫( – ערי)ה(מסכנות‬τὰς) πόλεις (τὰς) ὀχυράς

Elsewhere ‫ ערי מסכנות‬is rendered by περίχωροι (2Chr 16:4) and πόλεις (2Chr 32:28). 5. Gen 49:14 Ps 68 (67):14

‫ – בין המשפתים‬ἀνὰ μέσον τῶν κλήρων ‫ – בין שפתים‬ἀνὰ μέσον τῶν κλήρων

Elsewhere the word occurs only in Judg 5:16 (B: διγομίας, A: μοσφαθαιμ; A in v. 15: χειλέων). 6. Gen 49:6 Josh 11:6, 9

‫ – עקר‬νευροκοπέω ‫ – עקר‬νευροκοπέω

23. Deut 33:6 ‫ –ויהי מתיו מספר‬καὶ ἔστω πολὺς ἐν ἀριθμῷ; Deut 26:5, 28:62 ‫ – במתי מעט‬ἐν ἀριθμῷ βραχεῖ.

326

2. Lexical Source for Difficult Words

21. The Septuagint Translation of the Torah as a Source and Resource

Elsewhere the Hebrew root occurs only in 2Sam 8:4 = 1Chr 18:4.

‫ – באר היטב‬σαφῶς σφόδρα

7. Deut 27:8

‫ – ובאר על הלחות‬καὶ σαφῶς ἐπὶ πυξίον

Hab 2:2

The unusual translation of Habakkuk, which is problematic in the context, is based on Deuteronomy. 8. ‫ מאכלת‬in Judg 19:29 is rendered by μάχαιρα in manuscript A in accordance with Gen 22:6, 10. Elsewhere the word occurs only in Prov 30:14, where it is rendered differently.

3. Quotations and Allusions Quotations from passages from the Torah and allusions to such passages that appear in the later books of the Bible were usually not formulated as in the Septuagint translation of the Torah (see note 15), but sometimes that translation influenced the postPentateuchal books: 1. Num 35:33 Ps 106 (105):38

‫לא תחניפו את הארץ ]…[ כי הדם הוא יחניף את הארץ‬ καὶ οὐ μὴ φονοκτονήσητε τὴν γῆν […] τὸ γὰρ αἷμα τοῦτο φονοκτονεῖ τὴν γῆν ‫ותחנף הארץ בדמים‬

καὶ ἐφονοκτονήθη ἡ γῆ ἐν τοῖς αἵμασιν

φονοκτονέω (LSJ: to pollute with murder or blood) does not occur elsewhere in the Septuagint. As LSJ does not list occurrences other than those in the Septuagint, the agreement between the two texts is remarkable in view of the presumed rarity of this word. 2. Exod 22:1 (2) Jer 2:34

‫אם במחתרת ימצא‬

ἐὰν δὲ ἐν τῷ διορύγματι εὑρεθῇ

‫לא במחתרת מצאתים‬

οὐκ ἐν διορύγμασιν εὗρον αὐτούς

Elsewhere διόρυγμα occurs only in Zeph 2:14 (‫ )חלון‬and διορύσσω reflects ‫( חתר‬four times). 3. Lev 16:13 Ez 8:11

(‫וכסה ענן הקטרת )את הכפרת‬ καὶ καλύψει ἡ ἀτμὶς τοῦ θυμιάματος

‫ועתר ענן הקטרת עלה‬

καὶ ἡ ἀτμὶς τοῦ θυμιάματος ἀνέβαινε

ἀτμίς only very rarely occurs elsewhere in the Septuagint. ‫ ענן‬is rendered mainly by νεφελή. 4. Gen 32:29 (28) Hos 12:4 12:5

‫שרית עם אלהים‬

ἐνίσχυσας μετὰ θεοῦ

‫שרה את אלהים‬

ἐνίσχυσεν πρὸς θεόν

‫וישר אל מלאך‬

καὶ ἐνίσχυσεν μετὰ ἀγγέλου

3. Quotations and Allusions

327

21. The Septuagint Translation of the Torah as a Source and Resource

The Hebrew root does not occur elsewhere in Scripture. 5. Num 25:3 Num 25:5 Ps 106 (105):28

‫ויצמד ישראל לבעל פעור‬

καὶ ἐτελέσθη Ισραηλ τῷ Βεελφεγωρ

‫הנצמדים לבעל פעור‬

τὸν τετελεσμένον τῷ Βεελφεγωρ

‫ויצמדו לבעל פעור‬

καὶ ἐτελέσθησαν τῷ Βεελφεγωρ

Note the unique interpretation of the verb in the Septuagint (‘to be consecrated to’), cf. Hos 4:14 (‫ )קדשות‬and Deut 23:18 (17) ‫ – קדש‬τελεσφόρος.

4. Influence on the Exegetical Level The contents of the Greek Torah often influenced the wording of later translations on an exegetical level. 24 1. In Jer 1:6; 4:10; 14:13 and 32 (39):17 (‫ אהה )עדני יהוה‬has been represented by ὁ ὢν (δέσποτα κύριε). 25 ‫“( אהה‬alas”) in this verse has been derived from ‫ היה‬in Exod 3:14 (a central verse for biblical theology) and rendered in accordance with the Septuagint translation of that verse: ‫ – אהיה אשר אהיה‬ἐγώ εἰμι ὁ ὤν. 2. Prov 24:28 ‫אל תהי עד חנם ברעך‬ μὴ ἴσθι ψευδὴς μάρτυς ἐπὶ σὸν πολίτην The translation of this verse reflects the exegesis of ‫ חנם‬as “false,” mainly on the basis of the ninth commandment in Greek: Exod 20:16 (13) ‫לא תענה ברעך עד שקר‬ οὐ ψευδομαρτυρήσεις κατὰ τοῦ πλησίον σου μαρτυρίαν ψευδῆ Deut 5:20 (18) ‫לא תענה ברעך עד שוא‬ οὐ ψευδομαρτυρήσεις κατὰ τοῦ πλησίον σου μαρτυρίαν ψευδῆ 3. Ps 116 (114):9 ‫אתהלך לפני יהוה‬ εὐαρεστήσω ἐναντίον κυρίου The translation of this verse is based on the same equivalent in Gen 5:22, as suggested by Joosten, “Impact,” 201. This equivalent was also used by the Greek translator of Sir 44:16.

To summarize, this study has presented evidence in support of the following assumptions: (1) the vocabulary of the Greek Torah was maintained in the translation of the later books; (2) the Greek Torah served as a lexicon for the later translators in the translation of difficult Hebrew words; (3) quotations from and allusions to the Torah in the later books were sometimes phrased according to the Greek Torah; (4) the contents of the Greek Torah often influenced the wording of later translations on an exegetical level.

24. See especially Seeligmann, Isaiah, 45-46. 25. In 4:10 only manuscript 26 reads ο ων. For a discussion, see E. Tov, The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of Jeremiah 29–52 and Baruch 1:1–3:8 (Harvard Semitic Monographs 8), Missoula, MT, 1976, 24.

328

4. Influence on the Exegetical Level

22.1 Der Kult Martin Vahrenhorst

Mit der Übersetzung ins Griechische trat die Heilige Schrift Israels in den Raum der griechischsprachigen Kulte des Mittelmeerraums ein. Schon mit der Entscheidung ‫ אלהים‬in Gen 1,1 mit ὁ θεός zu übersetzen, stellten die Übersetzer den Gott Israels den Göttern hellenistischer Kulte sprachlich prinzipiell an die Seite – auch wenn sie inhaltlich an seiner Einzigartigkeit festhielten. Infolge dieser Entscheidung standen die Übersetzer auch vor der Frage, wie sie Begriffe einer wesentlichen Lebensäußerung ihrer Religion, nämlich des Kultes, in einer Sprache ausdrücken sollten, die ein reiches Repertoire an kultischem Vokabular bot, ohne den Kult Israels einfach im weiten Raum hellenistischer Religiosität aufzugehen zu lassen. Neben die Herausforderung, die Besonderheit des Eigenen zu bewahren und sich doch in einer anderen Sprache verständlich zu machen, trat das Problem, dass kultische Vollzüge im alexandrinischen Judentum nicht zur religiösen Praxis gehörten. 1 Die zu übersetzenden biblischen Texte selbst dürften für die Übersetzer wohl die Hauptquelle ihrer Information über kultische Sachverhalte des Pentateuchs gewesen sein. Im Folgenden sollen wesentliche Charakteristika der kultischen Sprache der LXX dargestellt werden. Dabei ist vor allem der Pentateuch im Blick, weil er für die nach ihm übersetzten Bücher sprachbildend gewesen ist. 2 In methodischer Hinsicht stellt sich dabei die Schwierigkeit, dass die Übersetzer selbst über ihre Entscheidungen keine Auskunft gegeben haben. Alles, was über ihre Übersetzungstechnik gesagt werden kann, muss also aus dem Vergleich zwischen hebräischer Textüberlieferung und der griechischen Übersetzung einerseits und dem Vergleich mit dem religionsgeschichtlich verwandten Vokabular der Umwelt andererseits erschlossen werden. Im Blick auf den Umgang mit dem kultischen Vokabular des Hellenismus sind wir in der glücklichen Lage, dass die Übersetzer der LXX nicht die einzigen sind, die in der Antike über nichtgriechische Kulte auf Griechisch berichten. Der ägyptische Priester 1.

2.

Hinsichtlich der Bezeichnung der Feste (z. B. πάσχα für ‫ )פסח‬ist es wahrscheinlich, dass die Übersetzer schon auf in ihrer Gemeinde geprägtes Vokabular zurückgreifen konnten (πάσχα entspricht dem damals gesprochenen Aramäischen ‫ פסחא‬und nicht dem hebräischen Text). Im Blick auf das Vokabular des Kultes gibt es dafür keine Indizien. Das bedeutet nicht, dass die später übersetzten Bücher den Pentateuch gleichsam als Vokabelliste benutzt hätten. Ihre Übersetzer »schöpfen je und je aus seinem Sprachschatz – sei es, um an Stil und Dignität des Pentateuch teilzuhaben, sei es um theologisch gezielte Bezüge zum Pentateuch herzustellen (vgl. dazu B. Meynadier, »Eléments de lexicographie comparée des Règnes et des Paralipomènes« in: W. Kraus / O. Munnich (Hg.), La Septante en Allemagne et en France. Textes de la Septante à traduction double ou à traduction très littérale (OBO 283), Fribourg/Göttingen 2009, 37 ff. und M. Vahrenhorst, »›Mehr als ein Wörterbuch‹. Beobachtungen zum Verhältnis des 2. Chronikbuches zum Pentateuch« in: Kraus / Munnich, La Septante en Allemagne et en France, 52-63. 4. Influence on the Exegetical Level

329

22.1 Der Kult

Manetho, dessen Ausführungen über ägyptische Kulte u. a. Josephus überliefert hat, stand im dritten Jahrhundert vor der gleichen Aufgabe. Wir können seinen Umgang mit dem kultischen Vokabular der Umwelt mit dem der LXX vergleichen, und unsere Schlussfolgerungen so in gewisser Weise absichern. Dennoch bleibt die Möglichkeit bestehen, dass Phänomene, die sich unserem Auge als theologisch bedeutsame Entscheidungen darstellen, sich in Wahrheit reinem Zufall verdanken. Vergleicht man nun Manethos Umgang mit der kultischen Begrifflichkeit, die auch in seiner Umwelt gebräuchlich war, so fällt auf, dass er sie offenbar vorbehaltlos übernehmen kann. 3 Es lassen sich keine signifikanten Differenzen zwischen den von ihm verwendeten Begriffen und denen der hellenistischen Umwelt ausmachen. Das kann man im Blick auf die Übersetzer des Pentateuchs nicht ohne weiteres sagen. Sie haben, wie wir sehen werden, manches übernehmen können, anderes aber vermieden und stattdessen ein eigenes Vokabular entwickelt.

1. Übernahme kultischer Termini aus der hellenistischen Umwelt Zunächst ist festzustellen, dass die Übersetzer des Pentateuchs grundsätzlich keine Scheu vor der Benutzung traditioneller Kultterminologie paganer Kulte haben, sie vollziehen vielmehr sprachliche Entwicklungen in ihrer Umwelt selbstverständlich mit. So wird das in hellenistischer Zeit häufiger zur Bezeichnung des Eigentums der Gottheiten – oder auch ihrer selbst – anzutreffende ἅγιος zum Hauptäquivalent der Wurzel ‫קדש‬, mit der in der hebräischen Bibel von Gottes Heiligkeit und den zu Gott gehörenden Dingen und Menschen gesprochen wird. Zur gleichen Zeit verliert in der paganen Gräzität das eigentlich bedeutungsgleiche ἁγνός an Bedeutung. 4 Die LXX verwendet es ebenfalls seltener – zum Beispiel, um zwischen dauerhaften Weihen und Besitzübertragungen auf Zeit zu unterscheiden. 5 Andere zentrale Begriffe des Kultes, wie ἱερεύς, δῶρον, oder θύειν werden in gleichem Maße in den Wortschatz der Übersetzung übernommen. 6 Kultvollzüge wie Libations- bzw. Besprengungsriten oder Reinigungsrituale tragen zumeist den gleichen Namen, der sich auch in hellenistischen Kulten findet. 7 »Gleiche Gebräuche 3.

4.

5. 6. 7.

Vgl. M. Vahrenhorst, »Greek Religious and Cultic Terminology in the LXX Pentateuch« in: D. Brakke / A. Jacobsen / J. Ulrich (Hg.), Beyond Reception. Mutual Influences between Antique Religion, Judaism, and Early Christianity, Aarhus 2006, 122-123. E. Williger, Hagios. Untersuchungen zur Terminologie des Heiligen in den hellenisch-hellenistischen Religionen, Gießen 1922, 76 ff.; J. Rudhardt, Notions fondamentales de la pensée religieuse et actes constitutifs du culte dans la Grèce classique (zweite Auflage), Paris 1992, 38 ff. Vgl. M. Vahrenhorst, Kultische Sprache in den Paulusbriefen (WUNT 230), Tübingen 2008, 25, 84. Vgl. M. V. Cerutti, »La terminologia religiosa e cultuale nel Penateucho greco« Annali di Scienze Religiose 6 (2001), 194. Das ‫( זרק‬ausschütten, ausgießen, sprengen) wird mit προσχέω übersetzt (ab Ex 24,6). Die Vokabel χέω bezeichnet im Griechischen sehr oft das Ausgießen von Trankspenden (Hom., Od. 10,518; Hdt. 7,43). Das gebräuchliche προσ/περιῥαίνω (Lev 4,6; 14,7; 4,17) gibt das hebr. ‫נזה‬ (hi.) gibt Besprengungsriten wieder, die an Altären vollzogen wurden (Ar., Lys. 1130) oder zur Reinigung dienten (SIG 982,8; Plut., Lykurg 2). Vgl. J. Casabona, Recherches sur le vocabulaire

330

1. Übernahme kultischer Termini aus der hellenistischen Umwelt

22.1 Der Kult

und Einrichtungen wie im Hellenismus heißen auch gleich«, 8 lässt sich im Blick auf die Arbeit der Übersetzer grundsätzlich festhalten. Das, was im Kult Israels geschieht, scheint also von den Übersetzern als mit anderen Kulten durchaus vergleichbar wahrgenommen worden zu sein.

2. Abgrenzungen zu den Kulten der hellenistischen Umwelt Ein entscheidendes Differenzkriterium besteht jedoch wohl hinsichtlich des Altars und – damit untrennbar verbunden – des Adressaten des Kultes. So zumindest lässt sich erklären, warum die Übersetzer den im Griechischen üblichen Begriff βωμός in der Regel nichtjüdischen Kulten vorbehalten (vgl. Ex 34,13; Num 23,1; Dtn 7,5) und für den Altar des Gottes Israels ein neues Wort schaffen: θυσιαστήριον (Gen 8,20). Es gibt aber auch zahlreiche Begriffe, die in paganen Kulten üblich waren, die die Übersetzer des Pentateuchs nicht gewählt oder bewusst vermieden haben. Hier wäre zunächst die klassische Tempelterminologie zu nennen: ναός, ἱερόν, τέμενος oder ἄδυτον finden sich im Pentateuch nicht. Das muss zunächst nicht weiter verwundern, denn das Heiligtum des Pentateuchs ist ja noch kein Tempel, und selbst wenn seine Kultvollzüge transparent auf den Tempelkult in Jerusalem sind, hat die LXX diese schon in der hebr. Bibel bestehende Differenzierung bewahrt. Aber auch in späteren Texten sind die genannten Vokabeln – sieht man einmal von ναός ab – selten oder finden sich in Beschreibungen fremder Kulte. 9 Ebenfalls übergangen werden alle Derivate von ἱερός – außer ἱερεύς. 10 Selten ist ebenfalls das zentrale Verb griechischer Gottesverehrung: σέβομαι. 11 Bei manchen Opfertermini lässt sich beobachten, dass die Übersetzer auf einen Begriff zurückgreifen, der deutliche Ähnlichkeit mit einem analogen Ausdruck der paganen Kultsprache hat, ohne dass sie diesen Ausdruck genau übernähmen. In der hellenistischen Religiosität kennt man zum Beispiel »τὰ σωτηρία«. Das sind Opfer, mit denen man für Rettung aus Gefahren dankt (vgl. Xenophon, Anabasis 3.2.9; 5.1.1; SIG 402,5). Die LXX vermeidet den Plural, verwendet aber »σωτηρίον« im Genitiv als Übersetzung von ‫( שלמים‬vgl. Lev 3). Ähnlich verhält es sich mit der Übersetzung des hebr. »‫«ריח ניחח‬. Dieses Syntagma bezeichnet den »Wohlgeruch«, der von einem Opfer zur Gottheit aufsteigt. Pagane Kulte kennen die Vorstellung des zu den Göttern aufsteigenden Opferrauchs ebenfalls. Einer der dafür gebräuchlichen Ausdrücke ist

8. 9.

10. 11.

des sacrifices en grec, des origines à la fin de l’époque classique, Aix-en-Provence 1966, 279-281, und den Exkurs »Hinweise zur Reinheitsterminologie im LXX Pentateuch« in LXX.D, Bd. I, 366-371. F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta, Münster 2001, 229. Ναός ist später für den Tempel gebräuchlich. Ἱερόν begegnet hingegen nur in 1Esdr, den Makkabäerbüchern und in Dan 9,27. Τέμενος wird immer nur im Blick auf heidnische bzw. illegitime Kultbezirke gebraucht. G. Dorival, »Le sacrifice dans la traduction grecque de la Septante« AnnSE 18 (2001), 77. Der Pentateuch vermeidet es. Es begegnet zuerst in Jos 4,24. Sieht man von Jes 29,13; 66,14; Jona 1,9; Hi 1,9 verteilen sich die 26 Belege auf Texte, von denen sich keine hebr. Vorlage erhalten hat. 2. Abgrenzungen zu den Kulten der hellenistischen Umwelt

331

22.1 Der Kult

ὀσμὴ σεμνή (heiliger Geruch). Die Übersetzung der LXX lautet ähnlich, ohne den vorhandenen griechischen Begriff einfach zu übernehmen (ὀσμὴ εὐωδίας). 12

3. Neologismen und sinngemäße Übertragungen Für Opfer und Rituale, die kein exaktes Äquivalent in paganen Kulten haben, werden – zuweilen in Anlehnung an in anderen Kulten gebräuchliche Verben oder Begriffe – neue Begriffe geschaffen. Dabei übersetzt man die hebr. Begriffe nach Möglichkeit so, dass das Wesentliche des Geschehens, das sie bezeichnen, in der Zielsprache verständlich gemacht wird. Einige Beispiele mögen dies verdeutlichen: Der Altar ist eine Stätte, an der man opfert, darum heißt er θυσιαστήριον (Opferstätte). Die ‫ כפורת‬ist der Ort, an dem Sühne geschaffen wird. Das gibt die Übersetzung ἱλαστήριον (ab Ex 25,17) angemessen wieder. Die ‫( עולה‬erstes Vorkommen Ex 10,25) ist ein Opfer, dessen Teile – abgesehen von seinem Fell, das den Priestern zukommt (Lev 7,8) – menschlicher Nutzung ganz entzogen sind. Es muss ganz verbrannt werden. Dieser Umstand prägt die griech. Übersetzung, in der sich in den meisten Fällen die Vorsilbe ὁλο (»ganz«) findet. Am häufigsten ist ὁλοκαύτωμα. Καύτωμα leitet sich von καίω (»[ver]brennen«) ab. Die Übersetzer deuten dieses Opfer also von dem her, was mit ihm geschieht. Es wird ganz verbrannt, also ist es ein »Ganzbrandopfer«. 13 Die hebr. Etymologie ist hingegen eine andere. Ihr folgend handelt es sich bei der ‫ עולה‬um ein Opfer, das »aufsteigt«. Daran haben sich die Übersetzer jedoch nicht orientiert. Die pagane Religiosität kennt ὁλοκαυτέω und Derivate davon als termini technici der Opfersprache. 14 Darauf konnten die Übersetzer bei ihrer Arbeit zurückgreifen. Die entsprechenden Substantive der LXX sind hingegen allesamt Neologismen. Zu den Besonderheiten der Sprache der hebr. Bibel gehört es, dass für manche Vorgänge, die das Verhältnis von Gott und Mensch stören, und für die Mittel, dieses wieder in Ordnung zu bringen, die gleiche Wurzel benutzt wird. So sagt Lev 4,3, dass jemand etwas darbringen soll ‫»( על חטאתו אשר חטא‬für die Sünde, durch die er sich versündigt hat« [d. h.: die er begangen hat]). Der Anlass für das Opfer ist »seine Sünde« (‫)חטאתו‬. Das, was Gott ihm als Mittel zur Sühne bereitstellt, heißt ‫( חטאת‬Lev 4,8 [meist »Sündopfer«]). Die LXX wählt als Übersetzung der Wurzel ‫ חטא‬und ihrer Derivate, in aller Regel den Stamm ἁμαρτ-. Um zwischen der begangenen Sünde und dem dafür dargebrachten Opfer unterscheiden zu können, prägt die LXX den Ausdruck περὶ τῆς ἁμαρτίας (»Für-die-Sünde«). 15 Im MT steht dafür meist ‫לחטאת‬. 16 Die Übersetzung wahrt so eine große Treue zum hebräischen Text und verdeutlicht zugleich den Sinn des Rituals: es geschieht für die Sünde. Das Bestreben, den Leserinnen und 12. Vgl. M. Rösel, Übersetzung als Vollendung der Auslegung. Studien zu Genesis-Septuaginta (BZAW 223), Berlin/New York, NY 1994, 190; Dorival, »Le sacrifice«, 78 weist darauf hin, dass Nähe und Distanz zur Sprache paganer Kulte sich auch durch Differenzen im Genus verbi ausdrücken können. So verwendet die LXX σφάζειν während in den Kulten der Umwelt σφαγιάζεσθαι gebräuchlich ist. 13. Vgl. S. Daniel, Recherches sur le vocabulaire du culte dans la Septante (EeC 61), Paris 1966, 253. 14. Vgl. Vahrenhorst, Exkurs »Hinweise zur Opferterminologie im LXX Pentateuch«. 15. Vgl. Vahrenhorst, »Hinweise zur Opferterminologie im LXX Pentateuch«. 16. Analog verfahren die Übersetzer beim ‫אשם‬-Opfer (meist »Schuldopfer«).

332

3. Neologismen und sinngemäße Übertragungen

22.1 Der Kult

Lesern das Wesen eines Rituals begreiflich zu machen, scheint die Übersetzer auch da geleitet zu haben, wo die Opfervollzüge ihnen selbst vielleicht nicht ganz deutlich vor Augen standen. Dies könnte bei der ‫תנופה‬, dem sog. Schwingopfer, der Fall gewesen sein. 17 Die Übersetzer wählen verschiedene Äquivalente, die darauf abheben, dass das Opfer als zusätzliche Abgabe verstanden werden konnte oder eine solche Abgabe enthielt. 18

4. Zwischen Konkordanz und Variation Vergleicht man den Text der LXX mit dem MT, so fällt auf, dass die Übersetzter zwischen strenger Konkordanz und spielerisch anmutender Freiheit wechseln können. Was den Bereich der kultischen Sprache anbelangt, gibt es zu den Variationen innerhalb der LXX keine Entsprechungen in der hebräischen Textüberlieferung, so dass man davon ausgehen kann, dass hebr. Begriffe im Laufe der Übersetzung auch in unmittelbarer Nachbarschaft unterschiedlich übersetzt worden sind. So finden sich für ‫ עולה‬im Griechischen ὁλοκαύτωμα, ὁλοκαύτωσις, ὁλοκάρπωμα und ὁλοκάρπωσις. Das Opfer für die Sünde (‫ )לחטאת‬wird mit περὶ τῆς ἁμαρτίας wiedergegeben. Daneben begegnen bei gleicher hebr. Vorlage aber auch περὶ ἁμαρτίας, τῆς ἁμαρτίας oder nur ἁμαρτίας. Die Existenz solcher Variationen lässt sich unterschiedlich deuten. Man kann einerseits vermuten, dass das Ideal einer konkordanten Übersetzung den Übersetzern noch unbekannt war. Dagegen spricht, dass sich die LXX in anderen Fällen durch große Konkordanz auszeichnet. J. W. Wevers vermutet, dass die Variationen auf des Übersetzers »love of variation« zurückzuführen sind. 19 Möglicherweise lässt sich aber auch ein inhaltlicher Grund angeben, der die Übersetzer veranlasst haben könnte, gleichen hebr. Begriffen unterschiedliche griechische Äquivalente zu geben. Dadurch konnten sie signalisieren, dass die Vorgänge des Kultes Israels sich nicht eins zu eins aus dem Hebräischen ins Griechische übertragen lassen. 20

5. Reduzierung von Anthropomorphismen Zu den Übersetzungstendenzen der LXX allgemein gehört die Reduzierung von Anthropomorphismen im Vergleich zum hebräischen Text. 21 Auch im Bereich des Kultes lässt sie sich beobachten. Das schon erwähnte hebr. ‫ ריח ניחח‬ruft die Assoziation wach, der Opfergeruch beruhige die Gottheit. 22 Die LXX wählt für ‫ ניחח‬εὐωδία. Natürlich bleibt der in den 17. Vgl. dazu J. Milgrom, Leviticus 1-16 (The Anchor Bible), New York, NY 1991, 473. 18. Vgl. M. Vahrenhorst, »Hinweise zur Opferterminologie im LXX Pentateuch« in LXX.D, Bd. I, 335-346. 19. Vgl. J. W. Wevers, Notes on the Greek Text of Leviticus, Atlanta, GA 1997, 28. 20. Vgl. G. Dorival, »›Dire en grec les choses juives‹. Quelques choix lexicaux du pentateuque de la septante« REG 109 (1996), 530 f.; Cerutti, »La terminologia religiosa e cultuale«, 196. 21. Vgl. F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta, Münster 2001, 243 ff. 22. Vgl. zur traditionellen Ableitung des Substantivs ‫( ניחח‬behagen, Beruhigung, Beschwichti4. Zwischen Konkordanz und Variation

333

22.1 Der Kult

Erzählkontexten angelegte Gedanke, dass der Geruch des Opfers Gott positiv beeinflusst, erhalten, aber es wird nur noch von einem wohlriechenden nicht mehr von einem beruhigenden Geruch gesprochen. Im hebr. Text findet sich in Opferzusammenhängen gelegentlich ‫( לחם‬z. B. Lev 3,11.16; 22,25). Die LXX lässt dieses Wort meist unübersetzt und vermeidet so den Eindruck, Gott würde sich tatsächlich von Opfern ernähren.

6. Zusammenfassung Mit der Übersetzung kultischer Sachverhalte ins Griechische bereitet die LXX den Boden für die Reflexion kultischer Sachverhalte im hellenistischen Judentum und im frühen Christentum vor. Philo von Alexandriens allegorische Deutungen des Kultes Israels basieren ebenso auf dieser Übersetzung wie die vielfältigen Rezeptionen kultischer Traditionen im Neuen Testament und in der Alten Kirche. In Nähe und Distanz zur kultischen Sprache ihrer Umwelt gelingt es der Übersetzung, das, was in den kultischen Vollzügen der hebräischen Bibel beschrieben wird, »auf Griechisch zu sagen« (G. Dorival nach E. Levinas) – und zugleich darauf zu verweisen, dass Israels Verehrung des einen Gottes mit den kultischen Vollzügen der hellenistischen Welt nicht ohne weiteres zu vergleichen ist.

gung) von ‫( נוח‬ruhen) R. Rendtorff, Leviticus 1,1–10,20 (BK III/1), Neukirchen-Vluyn 2004, 66 ff.

334

6. Zusammenfassung

22.2 Sin and Forgiveness Anna Passoni dell’Acqua 1

The Septuagint translates the vocabulary of fault and remission in a way that is interesting from both a lexical and a conceptual viewpoint. 2 This lexicon was then adopted by the authors of the New Testament to express sin and forgiveness in early Christianity preaching. 3 The terminology of the Septuagint has its roots in law and inherits the classical usage that distinguishes crimes according to their gravity. 4 The Hebrew vocabulary of errors, faults, and violations stems, too, in part from law: 5 the words that compose this semantic field denote lack, deviation, infringement or detriment of rights to both bodily integrity and to property. Such actions break the “horizontal” relationship with other people and the “vertical” one with human or divine authority. In both cases, a distinction is made between intentional and unintentional acts, the latter consisting of errors committed “through ignorance.” 6

1. 2.

3. 4.

5.

6.

Università Cattolica (Milano, Italy). See A. Passoni Dell’Acqua, “La terminologia dei reati nei προστάγματα dei Tolemei e nella versione dei LXX” in: B. Mandilaras (ed.), Proceedings of the XVIIIth International Congress of Papyrology Athens 25-31 May 1986, vol. II, Athens 1988, 335-350. S. Lyonnet, “Péché (III. Dans le Judaïsme)” in: J. Briend / É. Cothenet (eds.), DBS VII, Paris 1966, 480-485. From its origins, Greek attests to the usage of lexical groups that, through the Koine, continue to influence the modern language. Since Homer’s age (Il. XXVI, 86-88), a distinction has been made between “voluntariness” and “involuntariness” regarding murder. Philosophical reflection led to a progressive evolution in both categories, broadening the one of voluntariness, which originally was very limited. Through his investigation of voluntariness and involuntariness and the levels of guilt associated with them, Aristotle succeeded in laying the foundations of a theory of human acts. (Ethica Nicomachea, see below note 5). See Passoni Dell’Acqua, “Terminologia,” 338-339. These include: ‫פשע‬, “to behave as a criminal,” ‫“ רשע‬to be (become) guilty,” ‫ חטא‬and ‫“ עון‬misdeed.” Words that are not specifically legal can be employed in legislative texts, for example ‫“ אשם‬to be guilty.” The modern distinction does not correspond to the ancient one. According to Aristotle an act done “through ignorance” (δι’ἄγνοιαν) is not voluntary, but neither is it an act done “without knowing” (oὐκ εἰδώς) what one was doing; it is an act done “in ignorance,” that is an act done without knowing the particular circumstances of the act itself. Only acts that are regretted and followed by repentance can be qualified as involuntary (Ethica Nicomachea III, I, 13 (1110b, 1520), cf. H. Rackham (ed.), Aristotle. The Nicomachean Ethics, London 1956, 123). In the Septuagint, the hapax legomenon ‫“ בשׁגגה‬inadvertently” (Lev 22:14) is translated as κατ’ἄγνοιαν. The noun ‫שׁגגה‬, “inadvertent sin” is constructed with prepositions other than ‫ב‬. To indicate full conscience and will Hebrew uses the syntagm ‫“ ביד רמה‬with a high hand” (Exod 14:8; Num 15:30; 33:3). 6. Zusammenfassung

335

22.2 Sin and Forgiveness

The Septuagint gives literary evidence of Ptolemaic legal and political institutions and bears witness to the lexicon employed in both aulic and courtly law activities. It provides a clear case of transculturation: the terms of the Ptolemaic criminal law denoting different types of crimes, either committed or suffered, 7 are adapted to express the sins mentioned in the Bible. In this way, Greek legal words took on a theological meaning, as was previously the case for some of their Hebrew counterparts. The exclusive power of remission exercised by the Lagid king was used to convey the infinite mercy of the God of Israel, the only One able to efface the negative consequences of human transgressions. Similarly, the way the peculiar relationship between the God of the Fathers and His people was formulated relies on the categories of Ancient Near Eastern international law, which bound parties of unequal military power and political importance. 8 These linguistic choices emphasize the absolute gratuitousness of God’s action towards mankind. Translators also worked on a conceptual level, analyzing the multiplicity and the variety of the Hebrew lexemes 9 into five Greek lexical groups, ἁμαρτάνω, ἀδικέω, ἀνομέω, ἀσεβέω, and ἀγνοέω (given here in order of frequency), characterized 10 by the privative initial prefix α- expressing the absence of wrongdoing or fault. In the case of the root ‫ חטא‬and the lexical group of ἁμαρτάνω, the translators attempted to preserve the idea of “failure to hit, falling short of the mark.” 11 Moreover, they established no regular correspondences, as they translated different Hebrew roots with a single

7. The petitions handed down by papyri reveal faults and acts of violence; the most recurrent terms are ἀδίκημα “crime” and the cognate verb used in the passive diathesis (ἀδικοῦμαι ὑπό τινος). See e. g. PLille I, 29, 1 and PEleph 27(a), 25 (third century BCE); PAmh II, 33, 13 (circa 157 BCE). 8. Hebrew uses the term ‫“ ברית‬covenant, agreement.” The Greek translation διαθήκη derives from a legal context, but it is limited to the private testamentary law: YHWH, a living testator, grants his people the estate in usufruct. See A. Schenker, “ΔΙΑΘΗΚΗ pour ‫ברית‬. L’option de traduction de la LXX à la double lumière du droit successoral de l’Égypte ptolémaïque et du livre de la Genèse” in: J.-M. Auwers / A. Wénin (eds.), Lectures et relectures de la Bible. Festschrift P.-M. Bogaert, Leuven 1999, 125-131; A. Schenker, “Le contrat successoral en droit gréco-égyptien et la διαθήκη dans la Septante” ZABR 6 (2000), 175-185. 9. Some fifty terms have been identified, but the main roots are: ‫“ חטא‬to mistake,” ‫“ רשע‬to be (become) guilty,” ‫“ עוה‬to do wrong,” ‫“ פשע‬to behave as a criminal,” ‫“ אשם‬to be guilty.” Nouns are used more often than verbs (see, e. g., ‫“ עון‬wrong,” with 331 occurrences). To this list should be added the root ‫ רעע‬I “to be bad,” which occurs with great frequency (781 occurrences) and appears to have been used in a generic way. The Septuagint translates this root with the words πονηρός, κακόν, and their cognates and derivatives). See Š. Porúbčan, Sin in the Old Testament. A Soteriological Study, Rome 1963, 3-107; E. Beaucamp, “Péché (I. Dans l’Ancien Testament: le vocabulaire hébraïque)” DBS VII, Paris 1966, 407-471. 10. Apart from ἁμαρτάνω, the Greeks seemed to consider the initial α as privative, according to the negative meaning of the verb “failure to hit the target.” 11. The Septuagint translation of ‫ חטא‬using the lexical field of ἁμαρτ- and the value of the Arab stem ḫaṭiʾ a led scholars to give the Hebrew root the original meaning “to miss the target.” See Beaucamp, “Péché,” 441-442.

336

6. Zusammenfassung

22.2 Sin and Forgiveness

Greek lexical group. 12 This practice accentuated a process of simplification already initiated by the authors of the Hebrew Bible. 13 Another translation choice in the Septuagint needs to be underlined: the verb ἀφίημι (and cognates) was used to express the remission of debts and the forgiveness of crimes, as can been seen in the contemporary technical lexicon, heir to the classical usage. The Lagids’ amnesty decrees (προστάγματα τῶν φιλανθρώπων), witnessed by second-century BCE Greek papyri found in Egypt, provide a paradigmatic example of this terminology. 14 Although the extant examples are often preserved in a very fragmentary way, the existence of multiple copies of the same text enables us to reconstruct the exact form. These documents show formulaic expressions listing crimes— usually in order of gravity 15—that are echoed in the biblical text, especially in prayers, which are often additions to the Hebrew narratives or sections of books written in Greek, as well as in the Psalms. Combining two or more terms in the same passage was also favored by the structure of the Hebrew poetry, based on parallelismus membrorum. Royal decrees on various matters were well known in Egypt and in the territories under the Ptolemaic rule (the third-century BCE province of “Syria and Phoenicia”), because they were routinely displayed in public places. God, in the Septuagint Pentateuch, is described as a Lagid sovereign, and His actions are written down according to the official bureaucratic terminology. 16 During the Ptolemaic period, lawsuits involving Jews and Ἕλληνες were judged 12. See T. Muraoka, A Greek-Hebrew/Aramaic two-way index to the Septuagint, Leuven 2010, sub vocibus. 13. See Beaucamp, “Péché,” 407-408, 417: “Dans l’élaboration du concept biblique de péché, l’influence de la Septante a dû être déterminante; la terminologie en la matière y est, en effet, moins riche et moins variée […] Mais, on notera qu’une tendance en ce sens se dessinait déjà dans la littérature hebraïque des dernières siècles.” 14. M. T. Lenger, Corpus des Ordonnances des Ptolémées (COrdPtol) (Académie Royale de Belgique, Mém. de la classe des Lettres LXIV, 2), Bruxelles 1980. The violence of the dynastic struggles compelled the winner to issue an amnesty in favor of his enemies. In the decree the announcement of debts (ὀφειλήματα) or other fiscal taxes was emphasized as well. See G. Hölbl, A History of the Ptolemaic Empire, London/New York, NY 2001, 184, 201-202. 15. COrdPtol 34 and 35 (Ptolemy V Epiphanes, 186 BCE); 53. 53bis-ter (121/120-118 BCE) and 54 (122/121-118/117 BCE) issued by Ptolemy VIII Euergetes II, Cleopatra II and Cleopatra III; 71 (Ptolemy XII Neos Dionisos, circa 60 BCE); PTebt III1,739, 43-45 (163/145 BCE). COrdPtol 53bis-ter, 2-5 is the best documented example: […] ἀφιᾶσι τοὺς ὑπὸ τὴν | βασιλήαν πάντας ἀγνοημάτων ἁμαρτημάτων | ἐγκλημάτων καταγνωσμάτων αἰτιῶν πασῶν τῶν ἕως […]. The first two terms indicate two categories of “errors and crimes”; the other three, the principal phases of legal proceedings: charges (suspended), sentences (pending) and lawsuits (ongoing). Φόνος ἐκούσιος “voluntary murder” and ἱεροσυλία “profanation” were excluded from amnesty. 16. For the lexicon related to orders and rules (προστάσσω and cognates, δικαιώματα, etc.), see A. Passoni Dell’Acqua, “Il Pentateuco dei LXX testimone di istituzioni di età tolemaica” in: A. Passoni Dell’Acqua (ed.), Septuaginta, Libri sacri della diaspora giudaica e dei cristiani. Atti della III giornata di studio, Milano Università Cattolica 11. 5. 1999 (ASR 3), Milan 1999, 171200, 185-187. 6. Zusammenfassung

337

22.2 Sin and Forgiveness

in Greek courts, under Greek law; for this reason Jews were compelled to acquire knowledge of the Greek judiciary system. Thus it is not surprising that the Jewish Alexandrian community, well-known for its cultural achievements, included experts on Ptolemaic law.

1. Purification and Expiation The issue of sin and transgression in Jewish thought raises the question of purification and expiation, necessary to restore the relation with God and His ‫ברית‬, which has been interrupted. According to their gravity faults committed required a sacrifice whereby the guilt was transferred to the sacrificial victim. Scholars argue about the differences between the various kinds of sin and the characteristics of the sacrifices required to atone for them. 17 However, the semantic and theological question has no bearing on the lexical choice: in Hebrew and Greek, identical words are used to designate both fault and sacrifice. The Hebrew roots ‫ חטא‬and ‫ אשם‬are respectively translated in Greek by the lexical groups of ἁμαρτάνω and ἀγνοέω. 18 To designate a sacrifice, the Septuagint coined a syntagm, τὸ/ τὰ περὶ …, consisting of the neuter article + noun in genitive proceeded by περί: τὸ περὶ ἁμαρτίας (Lev 14:13), τὸ περὶ [τῆς] ἀγνοίας (Lev 5:18), τὰ περὶ ἁμαρτίας καὶ τὰ περὶ ἀγνοίας (Ez 42:13).

2. Remarks on the Lexicon of the Septuagint The occurrence of ἀγνόημα 19 and ἁμάρτημα 20 in the Greek Pentateuch to indicate, respectively, “errors” and “crimes” testifies to the usage of these terms in third-century BCE Ptolemaic Egypt and guarantees their vitality in the Alexandrian Koine of the time. The terms ἄγνοια and ἁμαρτία, and cognates, occur with greater frequency in the Septuagint than in contemporary papyri documents. Ἄγνοια rarely occurs during the Ptolemaic period; ἁμαρτία has been attested only since the Christian era. It is necessary to emphasize that the lexical group ἀνομέω 21 took on considerable 17. A. Schenker, “Interprétations récentes et dimensions spécifiques du sacrifice ‫ ”ַחָטּאת‬Biblica 75 (1994), 59-70. 18. The root ‫אשם‬, when referring to the sacrifice of the same name, is translated by the lexical group πλημμελέω, meaning “to make a false note in music; metaphorically: to offend, to err and commit a sinful error.” See S. Daniel, Recherches sur le vocabulaire du culte dans la LXX, Paris 1966, 301-328 and 341-361. 19. Gen 43:12, translating ‫משׁגה‬. 20. Translation of ‫“ חטאת‬sin/sin-offering” (Gen 31:36; Lev 4:29; Num 1:53; Deut 9:27; 19:15), of ‫עון‬ “misdeed” (Exod 28:38; Num 18:23; Lev 4:29), and of ‫“ חטא‬guilt/sin” (Deut 22:26). 21. See J. E. Davidson, “ANOMIA and the Question of Antinomian Polemic in Matthew” JBL 104 (1985), 617-635, especially 619-623; S. Olofsson, “Law and Lawbreaking in the LXX Psalms – a Case of Theological Exegesis” in: E. Zenger (ed.), Der Septuaginta – Psalter. Sprachliche und Theologische Aspekte, Freiburg 2001, 291-330.

338

1. Purification and Expiation

22.2 Sin and Forgiveness

significance in the Septuagint, not only because νόμος was used to render ‫“ תורה‬education, law,” but also owing to the fact that the Pentateuch—the first collection of books to be translated, one that includes rules and regulations dictated to Moses by God—was called ὁ νόμος. The Septuagint bears witness to doublets, and larger sets of repeated terms, 22 that occur in the Hebrew text, 23 mainly in legal contexts and in prayers. As seen in the Ptolemaic papyri, the list of errors and/or crimes seems to mark a conceptual crescendo. In Septuagint Greek, and more specifically in the Pentateuch, the distinction could still be felt between the two cognate nouns -ημα “guilty action” and -ία “state of sin.” In Leviticus, ἁμάρτημα is also used to indicate the sacrificial victim, whereas ἁμαρτία was mainly used to designate the expiation sacrifice (approximately a hundred occurrences). The cognate verbs also occur often in the Septuagint. 24 The expressions closest to the Ptolemaic amnesty decrees occur in Septuagint books translated or written from the second century BCE on (cf. Sir 23:2-3; Τob 3:3; Dan 9:5 25.15-16; 1Mac 13:39 26), but also Ps 24 (25 MT):7; 58 (59 MT):4 and 2Par (Chron) 28:23. In Gen 50:17, one of the first texts to be translated, the appeal to God reads as follows: Ἄφες […] τὴν ἀδικίαν καὶ τὴν ἁμαρτίαν αὐτῶν. Exodus (34:7) and Numbers (14:18) celebrate Him as ἀφαιρῶν ἀνομίας καὶ ἀδικίας καὶ ἁμαρτίας. The verb ἀπολύω, which sometimes alternates with ἀφίημι in the amnesty decrees, is rarely used in the Septuagint, mainly occurring in the most recent books (cf. 2Μac 12:45 and 4Mac). Lexical choices in the Septuagint may have been encouraged by both linguistic and theological factors. The semantic groups ἀγνοέω and ἁμαρτάνω, which in legal terminology describe two categories of crimes, were used, on the basis of etymologic analogies between the Hebrew and Greek vocabularies of guilt, to translate the Hebrew roots (‫ חטא‬and ‫)אשם‬, which convey a more generic concept of guilt. In the most recent books of the Septuagint, the appropriation of the lexicon of the amnesty decrees seems to strike a theologically polemical note: the power to remit men’s sins is given to the God of Israel alone, and not to the earthly sovereign. In a similar way, the aulic titles of the Ptolemaic king are used to celebrate God. Ideal virtues of the Hellenistic sovereign, such as Soter and Euergetes, carefully avoided in the Pentateuch, 27 now become God’s attributes.

22. See Passoni Dell’Acqua, “Terminologia,” 345-346, note 75. 23. ‫עון‬, ‫ חטא‬and ‫ פשע‬occur together fourteen times. 24. Ἁμαρτάνω occurs approximately sixty times in the Pentateuch, often built on ἁμαρτία as an internal object. 25. Confession of the sins of Israel: ἡμάρτομεν, ἠδικήσαμεν, ἠσεβήσαμεν. 26. Demetrius II’s letter to Simon, the High Priest, and the Jewish people reads as follows: ἀφίεμεν δὲ ἁγνοήματα καὶ τὰ ἁμαρτήματα ἓως τῆς σήμερον ἡμέρας. 27. A. Passoni Dell’Acqua, “Euergetes” Aeg 56 (1976), 177-191; Passoni Dell’Acqua, “Pentateuco,” 188-190. 2. Remarks on the Lexicon of the Septuagint

339

22.3 Le lexique de l’administration et de la politique Gilles Dorival

Compte tenu de l’abondance de la matière, la présente étude est centrée sur le Pentateuque; mais elle donne parfois quelques aperçus sur les autres livres. Dans la Bible, il n’y a pas de séparation entre ce qui relève du juridico-religieux et ce que nous appelons aujourd’hui l’administration et la politique. De plus, les activités militaires jouent un rôle primordial. Il y a donc des chevauchements entre des champs sémantiques que les modernes distinguent. Par exemple, le « chef », en hébreu et en grec, est une fonction religieuse, militaire, politique et administrative. Il est néanmoins possible de repérer et d’étudier les mots caractéristiques de l’administration et de la politique.

1. Les mots de la communauté Dans la Genèse, quatre mots servent à désigner les hommes vivant en communauté: γένος, ἔθνος, λαός et συναγωγή. Mais, pour certains de ces mots, d’autres significations sont attestées. Ainsi, dans les premiers chapitres, γένος correspond à ‫ מין‬et indique les « espèces » végétales (1,11-12) et animales (1,21-25; 6,20; 7,14; 8,19, où le correspondant hébreu est ‫ )משפחה‬de la création. C’est seulement à partir du chapitre 11 que γένος traduit ‫ עם‬et désigne l’espèce humaine (11,6; 17,14) ou bien une population ou une famille plus restreinte (19,38; 26,10; 34,16; 35,29). Cet emploi s’inscrit dans le droit fil de la racine indo-européenne *ge/on, « naître ». En 40,17, γένος est de nouveau employé au sens d’espèce végétale et animale, mais le correspondant hébreu n’est pas connu. Dans les formules de promesse adressées aux patriarches, le mot grec qui désigne la collectivité humaine est ἔθνος. Dieu annonce à Abraham qu’il deviendra une « grande nation » (12,2; 17,4-6; 18,18); Isaac deviendra des « nations » (17,16); Ismaël engendrera « douze nations » (17,20; 25,16) et deviendra une « grande nation » (21,13.18); Jacob deviendra « des rassemblements de nations » (28,3; 35,11; 48,4) ou une « grande nation » (46,3); Joseph deviendra une « foule de nations » (48,19). Le même mot ἔθνος est employé pour désigner les nations de la terre issues des fils de Noé (10,5.20.31-32), les nations impliquées dans la guerre des quatre rois contre les cinq rois (14,1.5.9), l’Égypte (15,14), les nations étrangères (17,29; 20,4; 22,18; 26,4; 27,29; 36,40; 49,10), la nation issue de Jacob et la nation issue d’Ésaü (25,23). Ces emplois sont conformes à l’usage grec classique, où ἔθνος était rapproché d’ἔθος, « coutume », et désignait tout groupement humain ayant les mêmes coutumes. Ἔθνος correspond très majoritairement à ‫( גוי‬vingt-six exemples), mais il traduit aussi cinq fois ‫( עם‬17,16.27.29; 28,3; 48,4; 49,10) et une fois ‫אמה‬, « peuplade » (25,16). Dans trois cas, la présence d’ἔθνος s’explique par le contexte: en 14,5, l’expression ἔθνη ἰσχυρά, « nations fortes », correspond à ‫זוזים‬, une population vaincue par les quatre rois; en 17,20, le mot rend ‫נשיא‬, « prince »: 340

1. Les mots de la communauté

22.3 Le lexique de l’administration et de la politique

il s’agit des douze nations promises à Ismaël; en 17,27, le mot traduit ‫בן נכר‬, « étranger », littéralement: fils d’étranger). Le mot λαός est un terme homérique et classique qui désigne toutes sortes de groupes d’hommes. C’est le cas aussi dans la Genèse, où λαός désigne les gens de Lot (14,16), les gens de Sodome (19,4), les gens de Khet (23,7.12.13), le peuple d’Abimélekh (26,11), le peuple qui pourrait résulter de l’assimilation entre le peuple de Jacob et la population de Sychem (34,22), le peuple égyptien (41,40.55; 42,6; 47,21). Il désigne aussi les peuples des patriarches: la famille d’Abraham (25,8), le peuple issu de Jacob et le peuple issu d’Esaü (25,23), la famille de Jacob (32,8; 35,6; 49,29.33), les gens d’Ésaü (33,15), le peuple issu de Manassé (48,19), les gens de Dan (49,16), le peuple issu de Joseph (50,20). Λαός traduit très majoritairement ‫( עם‬vingt exemples), mais aussi ‫לאם‬, « peuple » (25,23 ter). Les mots ἔθνος et λαός sont synonymes en 25,23, où ils désignent les deux nations (ἔθνη correspondant à ‫ )גיים‬et les deux peuples (λαός traduisant ‫ )לאם‬présents dans le ventre de Rebecca: Jacob et Ésaü. La même synonymie se retrouve en Ex 33,13, aussi bien pour les mots hébreux ‫ עם‬et ‫ גוי‬que pour les mots grecs correspondants λαός et ἔθνος. Ainsi, ‫עם‬/λαός n’est pas exclusif pour désigner le peuple d’Israël, pas plus que ‫גוי‬/ἔθνος pour désigner les nations. Cela se vérifie dans l’ensemble de la Bible: ‫גוי‬/ ἔθνος renvoie à Israël en Es 1,4 et Ps 32/33,12, tandis que ‫עם‬/λαός désigne des peuples autres qu’Israël en Ps 32/33,10b. Néanmoins, il est vrai que, souvent, ‫ עם‬au singulier désigne Israël et ‫ גוי‬au pluriel, les nations autres qu’Israël; de même, λαός au singulier renvoie souvent au peuple que Dieu s’est choisi, tandis qu’ἔθνος au pluriel désigne le plus souvent les peuples autres qu’Israël. Le dernier mot qui désigne la communauté humaine est συναγωγή 1. En fait, en grec classique et dans la koinè, ce terme s’applique à bien d’autres groupes que les hommes et désigne toute espèce de rassemblement d’hommes ou de choses. Cet emploi se retrouve dans la Genèse, où συναγωγή désigne aussi bien le rassemblement de l’eau sous forme des mers le troisième jour de la création (1,9) que les rassemblements des nations (ἔθνη) que deviendra Jacob selon la promesse de Dieu (28,3; 35,11; 48,4). Le mot correspond avant tout à deux substantifs hébreux: ‫ עדה‬et ‫קהל‬, qui signifient tous deux « assemblée ». Dans le Pentateuque, le premier substantif est toujours rendu par συναγωγή, sauf en Nb 17,5, où la « communauté » de Koré est connotée péjorativement par l’emploi du mot ἐπισύστασις, « coalition », et en 17,11, où la « communauté » devient le « campement » (παρεμβολή), peut-être dans un souci de variation 2. Quant à ‫קהל‬, il est rendu aussi par πλῆθος (Ex 12,6, pour éviter la redondance * συναγωγή συναγωγῆς, le texte massorétique ayant ‫ )עדה קהל‬et par σύστασις (Gn 49,6, où Jacob connote péjorativement l’« association » de Syméon et Lévi), ainsi que par ἐκκλησία dans le Deutéronome. Le mot grec συναγωγή correspondant à ‫קהל‬/‫ עדה‬est attesté en Exode (vingt cas), Lévitique (vingt cas), Nombres (quatre-vingt-neuf cas). En Genèse (trois cas) et Deutéronome (5,22; 33,4), il correspond seulement à ‫קהל‬. On doit noter pour finir que συναγωγή ne désigne jamais la synagogue dans l’ensemble de la Septante. 1. 2.

Voir M. Harl, « Communauté » in: C. Dogniez / M. Harl (éd.), La Bible des Septante. Le Pentateuque d’Alexandrie, Paris 2001, p. 871-872. Voir G. Dorival, Les Nombres (La Bible d’Alexandrie 4), Paris 1994, ad locum. 1. Les mots de la communauté

341

22.3 Le lexique de l’administration et de la politique

Le Deutéronome introduit une innovation dans la désignation de la communauté. À συναγωγή pour rendre ‫קהל‬, qu’il n’ignore pas (5,22; 33,4), il préfère ἐκκλησία (une dizaine d’exemples). Le mot est employé pour la première fois en 4,10, dans l’expression « au jour de l’assemblée », qui est absente du texte massorétique et qui résulte d’une harmonisation avec 18,16. Il n’est pas sorti de rien: le verbe ‫ קהל‬au hiphil est rendu par ἐκκλησιάζειν en Nb 20,8, par ἐξεκκλησιάζειν en Lv 8,4 et Nb 20,10. Cette innovation du Deutéronome est heureuse car, en grec archaïque et classique, ἐκκλησία désigne toute assemblée de guerriers ou de citoyens qui est réunie par une convocation (verbe καλεῖν).

2. Les subdivisions de la communauté La communauté des hommes se divise en φυλαί, « tribus », ‫אמת‬: l’expression « les tribus de la terre » revient deux fois dans la Genèse (12,3 et 28,14). Chacun des descendants de Noé est divisé lui aussi en tribus: les îles des nations, c’est-à-dire les Grecs (10,5), les Chananéens (10,18), les Sémites (Gn 10,31). Il en va de même pour les descendants d’Esaü (36,40). Cependant, le mot φυλή n’est employé qu’une seule fois pour les fils de Jacob, en 49,16, où il traduit ‫שבט‬, « bâton, sceptre, tribu ». L’expression « les douze tribus d’Israël » survient la première fois en Exode, où elle correspond au même mot hébreu. Dans ce livre, φυλή correspond soit à ‫( מטה‬six cas), « branche, tribu », soit à ‫( שבט‬trois cas), soit à ‫בית‬, « maison » (une fois). En Nombres, le mot φυλή correspond massivement à ‫מטה‬. En Deutéronome, qui n’emploie pas ‫מטה‬, il rend exclusivement ‫שבט‬. Ainsi, le Pentateuque considère comme synonymes deux mots hébreux différents. Le mot φυλή est bien attesté en grec classique pour désigner une portion de peuple ou de cité. Dans le cas des fils d’Israël, chaque tribu est à son tour divisée en sections et soussections, appelées en Exode συγγένειαι, « parentés », et πατριαί, « lignages paternels » (ou οἶκοι πατριῶν, « maisons des lignages paternels »). Ces mots relèvent du vocabulaire de la généalogie. L’originalité de la Septante est donc d’utiliser ces mots dans la sphère du politique. Cependant, en Exode, il n’y a pas de correspondance systématique entre les lexiques: συγγένειαι rend ‫משפחה‬, « clan, famille, parenté » et ‫תולדת‬, « descendance »; πατριά, ‫אבות‬, « pères », et ‫משפחה‬. En revanche, Nombres manifeste une cohérence beaucoup plus grande, grâce à une innovation de traduction propre à ce livre et qui consiste à rendre systématiquement ‫ משפחה‬par δῆμος, le mot qui signifie « peuple », mais qui désigne aussi une subdivision de certaines cités grecques. Cependant, il arrive que δῆμος ne corresponde pas à ‫( משפחה‬3,24; 13,3; 18,2; 23,10; 26,41); il arrive aussi que ‫ משפחה‬soit rendu par un autre mot que δῆμος (1,2; 1,18; 27,11; 33,54; 36,1). Au total, cependant, il y a plus de cent cinquante exemples de la correspondance δῆμος/‫משפחה‬. Chaque tribu est ainsi divisée en συγγένειαι, δῆμοι et οἶκοι πατριῶν. Or, la distinction entre tribus et dêmes renvoie à une réalité politique grecque de l’époque classique et hellénistique: l’Athènes de Clisthène était divisée en dix tribus et un nombre de dèmes qui a varié dans le temps (de 139 à plus de 150); c’est au sein des dèmes que les citoyens âgés de plus de dix-huit ans étaient enregistrés. À Rhodes, il y avait trois tribus et trente-six dèmes. Alexandrie elle-même semble avoir été divisée en cinq tribus et soixante dèmes, ainsi qu’en sept cent vingt phratries. 342

2. Les subdivisions de la communauté

22.3 Le lexique de l’administration et de la politique

À Athènes, le dème est une réalité topographique; ce n’est pas forcément le cas à Alexandrie, où il est peut-être une simple réalité numérique. On peut calculer qu’en Nombres, il y a cinquante-quatre dèmes si l’on ne tient pas compte de la tribu de Lévi, mise à part pour le culte. La tribu de Lévi elle-même se compose de trois dèmes, à leur tour subdivisés en huit sections également appelés dèmes. Ces dèmes bibliques sont une réalité numérique, comme le montre le recensement du chapitre 26, qui se fait dème par dème. Mais ils sont peut-être aussi une réalité topographique: à chaque dème correspond une localisation dans le campement et au sein de l’armée en marche. L’originalité de la Septante des Nombres est de transposer le modèle de l’organisation civique sédentaire en usage à Athènes, Rhodes et Alexandrie à la description d’une population nomade errant dans le désert: les fils d’Israël forment ainsi l’équivalent d’une cité en déplacement. Il faut noter que la Septante n’a pas retenu d’autres modèles de l’organisation civique grecque: ni celui qui, à Athènes, divisait les tribus en phratries (φατρίαι) et les phratries en familles (γοναί), ni celui qui, à Alexandrie, subdivisait les dèmes en phratries.

3. Les membres qui participent à la communauté des fils d’Israël Les membres de la communauté sont appelés en règle générale ‫בני ישראל‬, υἱοὶ Ἰσραήλ, « fils d’Israël », dont le premier emploi apparaît en Gn 32,33, à propos de l’interdit de la consommation du nerf sciatique. Ce tour rappelle la formule homérique υἷες Ἀχαιῶν, « fils des Achéens », pour désigner les Grecs. Mais il y a aussi le mot ‫אזרח‬, « natif », qui est traduit par αὐτόχθων, « autochtone » (dix exemples en Exode, Lévitique et Nombres), ou par ἐγχώριος, « qui est du pays, indigène » (quatre exemples en Exode, Lévitique et Nombres). Αὐτόχθων correspond une fois à ‫עם‬, « peuple » (Lv 20,4: « les autochtones du pays » en grec, « le peuple du pays » en hébreu); ἐγχώριος, une fois à ‫ארץ‬, « terre » (Gn 34,1: « les filles des indigènes » en grec, « les filles de la terre » en hébreu). Les deux mots grecs sont d’emploi classique pour désigner les habitants d’un pays. La communauté est organisée de manière hiérarchique. Au sommet, il y a les chefs 3. Le texte massorétique dispose de plusieurs mots pour désigner ce statut: ‫אדון‬, ‫אלוף‬, ‫נשיא‬, ‫שר‬, ‫שליט‬, sans compter ‫ראש‬, « tête », et même ‫מלך‬, « roi ». La Septante offre une seule correspondance systématique: en Gn 36 (quarante-deux exemples) et Ex 15,15, le substantif ‫ אלוף‬désigne les chefs d’Edom; il est traduit par ἡγεμών, « celui qui conduit, guide », qui désigne en grec le chef d’une contrée ou le chef militaire. Il n’y a pas de correspondance automatique pour les autres mots, même s’il est possible de dégager de grandes tendances: la Septante emploie avant tout ἄρχων, « chef » (quatrevingt-dix exemples environ) et ἀρχηγός, « dirigeant » (neuf exemples). Les ἄρχοντες sont parfois des chefs militaires (douze exemples en Nb 2) ou des chefs religieux (Nb 3,32). Mais ils sont souvent des chefs politiques: chefs des nations (Nb 21,18), chefs des tribus (Nb 1,4.16), chefs des lignages paternels (Nb 31,26), chef d’Israël (Nb 16,13: il 3.

Sur le vocabulaire de la chefferie dans les Petits Prophètes, voir C. Dogniez, « « Lost in translation »: la désignation des chefs dans le Dodekapropheton » JSJ 39 (2008), 192-210. 3. Les membres qui participent à la communauté des fils d’Israël

343

22.3 Le lexique de l’administration et de la politique

s’agit de Moïse). Les ἀρχηγοί sont les chefs des maisons des lignages paternels (Ex 6,14); en Nombres, le mot est synonyme d’ἄρχων. Ces emplois sont conformes à l’usage classique et hellénistique. Les autres mots de la chefferie dans la Septante sont moins fréquents: ἀρχίφυλος, « chef de tribu », est un néologisme de la Septante qui ne pose pas de problème de sens pour un lecteur grec; en Dt 29,9, il rend ‫ראש שבט‬, « tête de tribu », c’est-à-dire « chef de tribu »; ἡγούμενος, en Gn 49,10, désigne le « guide » qui sera issu des cuisses de Juda; il correspond au participe poel de ‫הקק‬, ‫מהקק‬, souvent compris au sens de « sceptre », qui est le symbole du chef; en Dt 1,13 et 5,23, le même mot correspond à ‫ ראש‬et désigne les chefs des tribus; enfin, φύλαρχος, en Dt 31,28, traduit ‫זקני שבטיכם‬, « les anciens de vos tribus »; c’est le mot qui, en grec classique, désigne le président d’une des tribus de la cité d’Athènes. En ce qui concerne ‫מלך‬, « roi », presque toujours rendu par βασιλεύς, on note que ce titre est normalement appliqué à des rois étrangers (par exemple en Gn 14, à propos de la guerre des quatre rois contre les cinq rois), ainsi qu’à Pharaon (premier exemple en Gn 39,20); mais il est utilisé aussi pour parler des rois à venir d’Israël (Gn 36,31) qui seront issus d’Abraham (Gn 17,6 et 16) et de Jacob (Gn 35,11). Toutefois, il faut signaler deux correspondances remarquables: en Gn 23,6, βασιλεύς rend ‫נשיא‬, « prince »: c’est le titre par lequel les fils de Khet saluent Abraham. Ἄρχων correspond à ‫ מלך‬en Nb 23,21 (il s’agit de la deuxième prophétie de Balaam: le texte massorétique parle du triomphe à venir d’un roi, YHWH ou un roi davidique, là où la Septante annonce les victoires des chefs hébreux dans le désert et en Chanaan 4) et Dt 17,14-15 (trois exemples); 28,36; 33,5, où il s’agit du roi à venir d’Israël établi sur le modèle des nations. La substitution du chef au roi a suscité diverses hypothèses 5: volonté de réserver le titre de βασιλεύς à YHWH, mais cette explication ne convainc pas, puisque le Pentateuque n’emploie pas βασιλεύς pour Dieu; loyalisme des traducteurs à l’égard des rois Ptolémées, mais il y a bien d’autres rois que les Ptolémées; texte hébreu différent offrant ‫ נשיא‬pour les rois d’Israël, mais c’est une solution de facilité. Moïse est assisté par Josué/Jésus, que le texte massorétique appelle le « serviteur » de Moïse, ‫ משרת‬en Ex 24,13; 33,11; Nb 11,28 (cf. Dt 1,38), la Septante utilise le participe substantivé ὁ παρεστηκώς, celui qui se tient à côté, l’assistant, sauf en Ex 33,11, où elle emploie le substantif θεράπων, « serviteur »; chez Homère, θεράπων désigne un noble qui se met volontairement au service d’un autre noble, comme Patrocle aux côtés d’Achille ou les chefs grecs aux côtés d’Agamemnon. En Nb 11,28, Jésus/Josué est également appelé l’élu, ὁ ἐκλεκτός, comme dans le texte massorétique. Moïse est également assisté par les « anciens d’Israël », ‫( זקני ישראל‬Ex 3,16). La Septante traduit ‫ זקנים‬tantôt par γερουσία, « conseil des anciens » (cinq exemples en Exode, deux en Lévitique, seize en Deutéronome), tantôt par πρεσβύτεροι, « anciens » (Ex 17,5 et sept autres exemples; Lv 4,15; six exemples en Nombres, trois exemples en Deutéronome). Pour un locuteur grec, les deux mots sont synonymes: une γερουσία est composée de γέροντες, « vieillards », un synonyme de πρεσβύτεροι. Parfois, πρεσβύτεροι est pratiquement synonyme de « fils d’Israël », comme le montrent les manuscrits de la Septante qui hésitent entre les « anciens d’Israël » et les « fils d’Israël » en 4. 5.

Voir Dorival, Nombres, 138. Voir C. Dogniez / M. Harl, Le Deutéronome (La Bible d’Alexandrie 5), Paris 1992, note sur 17,14.

344

3. Les membres qui participent à la communauté des fils d’Israël

22.3 Le lexique de l’administration et de la politique

Ex 34,30 et 32; mais ce n’est pas le cas quand les anciens forment un groupe restreint de soixante-dix (Ex 24,1; Nb 11,16): ils reçoivent l’inspiration de l’esprit de Dieu et assistent Moïse dans la conduite de la communauté (Nb 16,25). Le mot γερουσία appartient au vocabulaire traditionnel des institutions politiques: il désigne le conseil des anciens de diverses cités du pourtour méditerranéen d’époque classique et hellénistique. Il existe un autre organe délibératif de la communauté: la βουλή de Nb 16,2; Koré et ses partisans sont qualifiés de σύγκλητοι βουλῆς, « membres du conseil », littéralement « convoqués du conseil » (texte massorétique: « appelés du conseil », ‫ ;)מועד‬ces σύγκλητοι doivent probablement être identifiés avec les ἐπίκλητοι de 1,16 6. La βουλή du chapitre 16 ne doit pas être confondue avec la γερουσία, puisque Koré et les siens, qui sont deux cent cinquante, sont clairement différents des soixante-dix anciens. Il s’agit probablement de la réunion de tous les πρεσβύτεροι faite par une convocation. La Septante utilise le vocabulaire traditionnel des institutions politiques grecques: le σύγκλητος est celui qui est convoqué dans une assemblée délibérante; la βουλή désigne tout conseil délibératif. En Nb 11,16, les soixante-dix anciens sont appelées « scribes » (‫שטרים‬/γραμματεῖς). Faut-il pour autant faire de tous les scribes du Pentateuque des membres de la γερουσία et des πρεσβύτεροι? Il est question de « scribes des fils d’Israël » en Ex 5,14.15.19: ils ont été établis par les chefs de Pharaon pour contrôler les travaux effectués par les Hébreux. Ces γραμματεῖς sont-ils des « employés de rang inférieur, chargés d’inscrire les quantités de briques fabriquées » 7? Ou bien ont-ils été choisis parmi les membres de la γερουσία dont il est question en Ex 3,16? On peut hésiter. En Dt 20,5.8.9, les scribes donnent des instructions orales au peuple: ils sont sans doute identiques aux πρεσβύτεροι composant la γερουσία. Dans le monde grec, les γραμματεῖς étaient des sortes de greffiers de rangs très divers, chargés d’enregistrer les documents, de les conserver et d’en donner lecture dans les assemblées ou dans les tribunaux. A la différence des scribes de la γερουσία, ils n’étaient pas des hommes politiques, même s’ils contribuaient à la bonne marche de la politique et de la justice. La hiérarchie intermédiaire est représentée par les « chefs de mille », χιλίαρχοι, les « chefs de cent », ἑκατόνταρχοι, les « chefs de cinquante », πεντηκόνταρχοι, les « chefs de dix », δεκάδαρχοι. Ces mots font normalement partie du vocabulaire militaire grec; mais ils ne peuvent relever de ce champ sémantique dans le Pentateuque, car leur fonction consiste à assister Moïse et à juger les causes mineures (Ex 18,21-26); en grec, toutefois, la δεκαδαρχία désigne le gouvernement des dix à Athènes, tandis que les chiliarques sont de hauts dignitaires spartiates et perses: le sens non militaire de ces mots était possible, et c’est lui qu’a retenu la Septante. Le premier mot correspond à ‫ראש אלפים‬, « tête de milliers » (Nb 1,16) ou ‫שר אלף‬, « chef de mille » (deux exemples en Exode, quatre en Nombres, un en Deutéronome); ἑκατόνταρχος rend ‫שר מאות‬, « chef de centaines » (cinq exemples en Exode, Nombres et Deutéronome); πεντηκόνταρχος correspond à ‫שר המשם‬, « cent de cinquante » (trois exemples en Exode et Deutéronome); δεκάδαρχος rend ‫שר עשרה‬, « chef de dizaines » (trois exemples en Exode et Deutéronome). Dans le texte massorétique du Deutéronome, il est question de « scribes » pris par6. 7.

Voir Dorival, Nombres, 345-346. Voir A. Le Boulluec, P. Sandevoir, L’Exode (La Bible d’Alexandrie 2), Paris 1989, 109. 3. Les membres qui participent à la communauté des fils d’Israël

345

22.3 Le lexique de l’administration et de la politique

mi les hommes sages (1,15) et établis dans les villes de Judée-Palestine à côté des « juges » (16,18; voir aussi 29,10 et 31,28): manifestement, ils assistent les juges dans leurs fonctions. En ne traduisant pas par γραμματεῖς, mais par γραμματοεισαγωγεῖς, la Septante signale intelligemment qu’il ne faut pas confondre ces scribes subalternes avec les scribes de la γερουσία. Le mot grec n’a pas été retrouvé dans les papyri. Cependant, à Athènes, l’εἰσαγωγεύς est un magistrat qui reçoit une plainte et l’introduit devant un tribunal. Le néologisme de la Septante fait peut-être écho à ce terme tout en s’en distinguant: comme l’εἰσαγωγεύς, le scribe-instructeur contribue à l’instruction d’une affaire, mais il n’est pas lui-même un juge, seulement un collaborateur de justice. Il n’est pas question de « juges » (‫שפטים‬/κριταί) dans le Pentateuque en dehors du Deutéronome. Cette fonction est présentée comme devant exister dans l’avenir (16,18; 17,9.12; 19,17). Ces juges doivent être équitables, ne pas faire acception des personnes et ne pas accepter de cadeaux. Leur fonction paraît purement judiciaire. Mais, dans la période qui va de Josué à Samuel, ces juges seront en fait les chefs militaires et politiques des Hébreux, comme le montre notamment le livre des Juges. Le bas de la hiérarchie se compose de serviteurs à gages, de domestiques et d’esclaves 8. Le serviteur à gages se dit en hébreu ‫שכיר‬, qui est rendu par μισθωτός (huit exemples en Exode, Lévitique et Deutéronome), un terme de la langue classique qui désigne aussi bien le serviteur à gages que le mercenaire. Les lexiques hébreu et grec ne distinguent pas nettement entre la situation du domestique et le sort de l’esclave. Le mot ‫עבד‬, « serviteur, esclave » est rendu avant tout par παῖς, « enfant, serviteur » (quatre-vingt exemples en Genèse, huit en Exode, trois en Lévitique, six en Nombres, sept en Deutéronome). Mais on trouve aussi οἰκέτης, « domestique » (vingt-six exemples répartis dans tous les livres du Pentateuque) et θεράπων, « serviteur » (trente exemples en Genèse, Exode, Nombres et Deutéronome); δοῦλος est plus rare (Ex 21,7 au féminin; Lv 25,44; 26,13; Dt 32,36). On remarque que le mot θεράπων, qui désigne en grec souvent un vassal, est appliqué, d’une part, aux gens de la suite de Pharaon (une vingtaine d’exemples en Exode), d’autre part, aux patriarches: Abraham (Dt 9,27), Isaac (Gn 24,44), Jacob (Dt 9,27), les frères de Joseph (Gn 50,17), Jésus/Josué (Ex 33,11), les fils de Rouben et de Gad (Nb 32,31), et surtout Moïse (Ex 14,31; Nb 11,11; 12,7-8). Le terme habituel de la langue grecque classique pour désigner l’esclave acheté à prix d’argent, ἀργυρώνητος, est attesté en Genèse et Exode (cinq exemples), où il rend ‫מקנת כסף‬, littéralement « acquisition d’argent ». Passons aux mots désignant les esclaves de sexe féminin: ‫ אמה‬est rendu par δούλη en Lv 25,44; par οἰκέτις, « domestique de sexe féminin », en Ex 21,7; par θεράπαινα, « servante » (qui, en grec, ne comporte pas la nuance de vassalité de θεράπων), en Ex 21,26 et 27; et par παιδίσκη, « esclave de sexe féminin » (sept exemples en Genèse, cinq en Exode, deux en Lévitique, six en Deutéronome). Le mot ‫ שפחה‬est rendu par θεράπαινα en Ex 11,5, οἰκέτις en Lv 8.

Voir H. Heinen, « Zur Terminologie der Sklaverei in ptolemaischen Ägypten: pais und paidiske in den Papyri und der Septuaginta » in: Atti del XVII Congresso Internazionale di Papirologia, Naples 1984, 1287-1299; B. G. Wright, « Δοῦλος and Παῖς as translations of ‫עבד‬: Lexical Equivalences and Conceptual Transformations » in: B. A. Taylor (éd.), IX Congress of the International Organization for Septuagint and Cognate Studies. Cambridge, 1995, Atlanta, GA 1997, 263-277; E. Bons, « Le vocabulaire de la servitude dans la Septante du livre de Ruth » JSJ 33 (2002), 133-163.

346

3. Les membres qui participent à la communauté des fils d’Israël

22.3 Le lexique de l’administration et de la politique

19,20 et παιδίσκη (vingt-huit exemples en Genèse et Dt 28,68). Le mot ἀνδράποδον qui désigne l’être humain qui devient esclave du fait de la guerre n’est présent qu’en 3M 7,5.

4. Les actes de la vie de la communauté La communauté se réunit pour écouter Moïse. Les anciens se réunissent autour de Moïse, tout comme la βουλή dont il est question en Nb 16,2. En dehors de ces réunions, qui sont nombreuses, la Septante décrit peu d’actes de la vie politique et administrative des fils d’Israël: elle se centre sur les activités cultuelles et militaires. Il faut néanmoins faire mention du dénombrement de la communauté. Il s’agit d’un acte dont la finalité est guerrière, mais qui a, dans la Septante, une dimension politique. En Nb 1,18, Moïse et Aaron passent en revue les fils d’Israël, accomplissant ainsi un acte de type militaire. Ils se livrent à une opération dont le sens est peu clair dans le texte massorétique (« ils s’enfantèrent eux-mêmes »), mais que les targums et la Septante comprennent comme un enregistrement. La Septante dit qu’ils mettaient sur les tablettes les noms des soldats. Le verbe ἐπαξονεῖν signifie « placer sur des essieux »; le substantif ἄξων, « axe, essieu », prend dans le vocabulaire politique grec une signification particulière: à Athènes les ἄξονες étaient les quatre côtés d’une pièce de bois de section carrée sur lesquels étaient gravées les lois de Solon; ces tablettes tournaient autour d’un pivot de manière à permettre leur consultation. Aristote les appelle κύρβεις (Constitution d’Athènes 7, 1). Le mot κύρβεις est employé pour les lois de Rome. On peut donc retenir qu’à l’instar de ce que certaines cités grecques et latines faisaient pour leurs lois fondamentales, Moïse et Aaron mettent sur des tablettes les noms des soldats hébreux; ils donnent ainsi à un passage en revue militaire une valeur durable: la force d’un texte de loi.

5. Les membres extérieurs à la communauté Le membre d’une autre communauté, l’étranger, est appelé en hébreu ‫ נכר‬,‫ זור‬et ‫נכרי‬. Les mots grecs qui leur correspondent sont ἀλλογενής (quatorze exemples en Genèse, Exode, Lévitique et Nombres) et ἀλλότριος (dix-huit exemples dans les cinq livres du Pentateuque). Le premier mot traduit onze fois ‫זור‬, trois fois ‫ ;נכר‬le second, quatre fois ‫זור‬, quatorze fois ‫נכרי‬/‫נכר‬. Le mot ἀλλότριος peut désigner en grec classique l’étranger, mais il a un sens beaucoup plus large et désigne ce qui concerne autrui. Ἀλλογενής est une création de la Septante, qui signifie littéralement « qui est d’une autre race »: ce mot ne présente aucune difficulté de compréhension pour un locuteur grec. Un troisième terme désigne l’étranger dans le Pentateuque: ἀλλόφυλος, qui apparait en Ex 34,15; ce mot signifie littéralement « qui est d’un autre φῦλον, race » ou peut-être, à cause de l’emploi fréquent de φυλή dans la Septante, « qui est d’une autre tribu »; il n’a pas de correspondant dans le texte massorétique et ne figure que dans une partie de la tradition manuscrite; il s’applique aux populations installées en Chanaan. Dans les livres qui font suite au Pentateuque, ce terme est très fréquent (sauf en Josué). Il rend presque toujours ‫פלשתים‬, « Philistins ». Le mot ἀλλόφυλος est classique; il s’oppose à 4. Les actes de la vie de la communauté

347

22.3 Le lexique de l’administration et de la politique

ὁμόφυλος, « qui est de même race ». Les Grecs entre eux se reconnaissent ὁμόφυλοι par rapport aux autres peuples; les Lacédémoniens considèrent les Romains comme des ἀλλόφυλοι (Polybe, Histoires 9, 37, 7). Le mot le plus habituel chez les Grecs pour désigner l’étranger, ξένος, est absent du Pentateuque, sans doute parce qu’il signifie à la fois l’étranger et l’hôte et que cette ambiguïté n’était pas acceptable en de nombreux passages. Le terme est attesté une quinzaine de fois dans les livres qui font suite au Pentateuque, par exemple en Rt 2,10 (« je suis étrangère »). Un autre terme qui connote l’étranger est presque totalement absent: βάρβαρος, qui n’est attesté que six fois, dont en Ps 113/114,1, où il est un excellent équivalent de ‫לעז‬, « parler une langue barbare ». Pour désigner l’étranger qui habite hors de son pays d’origine, qu’il soit un membre de la famille d’Abraham de passage dans une terre étrangère ou un étranger résident en Israël, le texte massorétique utilise ‫ גר‬et ‫תושב‬. Le premier de ces substantifs est rendu massivement par προσήλυτος, « celui qui s’adjoint, immigré » (neuf exemples en Exode, vingt en Lévitique, onze en Nombres, dix-sept en Deutéronome), mais aussi par πάροικος, « immigrant de passage » (Gn 15,13; 23,4; Ex 2,22; 18,3; Dt 14,21; 23,8) et par γειώρας (Ex 12,19). Le mot ‫ תושב‬est rendu tantôt par παρεπίδημος, « voyageur de passage » (Gn 23,4) tantôt par πάροικος (un exemple en Exode, huit en Nombres). Ainsi, la correspondance entre ‫גר‬/προσήλυτος et entre ‫תושב‬/πάροικος ne souffre d’exception qu’en Lévitique et Nombres. On constate l’absence de μέτοικος qui désigne, en Grèce, l’étranger venu s’établir quelque part et, à Athènes, l’étranger domicilié dans la cité moyennant une redevance: elle s’explique par le fait que le πάροικος de la Bible est un immigrant de passage, non un résident permanent. Le mot πάροικος est attesté dans les inscriptions d’époque hellénistique au sens d’immigrant sans droit politique, mais avec un statut le protégeant. Παρεπίδημος est également attesté à l’époque hellénistique. Quant à προσήλυτος, c’est une création de la Septante forgée sur le modèle d’ἐπήλυτος. Les traducteurs ont voulu faire écho à ce terme qui désigne l’étranger venu dans un pays, tout en s’en distinguant; en effet, à la différence du πάροικος, qui n’a pas part à la Pâque, le προσήλυτος fait la Pâque pour le Seigneur (Ex 12,43-49); en Nb 10,29-32, Hobab, le beau-frère madianite de Moïse, se voit proposer de venir avec les Hébreux: en cas d’acceptation, il deviendrait un προσήλυτος; mais sont également des prosélytes les Juifs venus d’un pays étranger. En tout cas, le mot ne semble pas dans le Pentateuque désigner le converti au Dieu d’Israël. Le mot γειώρας est lui aussi une création de la Septante: c’est la transcription de la forme araméenne correspondant à ‫ ;גר‬grâce à la terminaison -as, le mot entre dans le système des déclinaisons et fait partie du lexique grec.

6. Les communautés étrangères Les nations sont gouvernées par des rois, ‫מלכים‬/βασιλεῖς, qui sont épaulés par des chefs, ἄρχοντες ou ἀρχηγοί ou ἡγεμόνες. Les rois de Madiam et Moab sont assistés par les « anciens », ‫זקנים‬, qui forment dans la Septante une γερουσία (Nb 22,7). Cependant, parmi les pays étrangers, c’est l’Égypte dont le système politique et administratif est le mieux décrit. Le Pharaon est « le roi d’Égypte ». Il est épaulé par des ἄρχοντες (Gn 12,15) et par « les anciens du pays d’Égypte » (Gn 50,8), qui sont 348

6. Les communautés étrangères

22.3 Le lexique de l’administration et de la politique

peut-être les mêmes personnes. Joseph à la tête de l’Égypte est appelé « l’ἄρχων de la terre » (Gn 42,6). Le Pharaon est assisté par les « gens de sa suite », θεράποντες (plus de vingt exemples en Exode). Plusieurs fonctions importantes sont remplies par des « eunuques »; le mot ‫ סריס‬désigne sans doute une fonction à la cour, et non pas un individu réellement castré; il est rendu par σπάδων (Gn 37,36) ou par εὐνοῦχος (39,1; 40,2.7). La fonction mal identifiée de ‫ « שר טבחים‬chef des tueurs », donne droit au titre d’eunuque; le grec a compris qu’il s’agissait d’un ἀρχιμάγειρος, « chef boucher » (Gn 37,36 et trois autres exemples). Eunuque aussi, le ‫ « ;שר המשקים‬chef des échansons », en grec ἀρχιοινοχόος, « chef échanson » (Gn 40,1 et sept autres exemples). Les deux mots grecs figurent chez Plutarque. Eunuque, enfin, le ‫שר המשקים‬, « chef des panetiers », qui est rendu par ἀρχισιτοποιός, « chef panetier » (Gn 40,1 et six autres exemples). En revanche, rien ne prouve que le ‫שר בית הסחר‬, « chef de la maison-tour », en grec ἀρχιδεσμοφύλαξ, « chef geôlier » (Gn 39,21 et quatre autres exemples), ait droit au titre d’eunuque. Il faut sans doute l’identifier à l’ἀρχιδεσμώτης, « chef de la prison » (Gn 40,4), qui remplace en ce verset le « chef des tueurs » du texte massorétique. Les deux mots grecs sont des néologismes de la Septante, parfaitement clair pour un locuteur grec. Toutes les fonctions qui précèdent sont discutées par les historiens: s’agit-il d’une Égypte réelle ou d’une représentation imaginaire de l’Égypte? En tout cas, un détail propre à la Septante prouve que les traducteurs étaient bien informés sur l’Égypte de leur temps. En Gn 45,10, Joseph charge ses frères de proposer à leur père Jacob de venir s’établir « dans la terre de Gésem d’Arabie » (ἐν γῇ Γεσεμ Ἀραβίας); en Gn 46,34, Joseph accueille son père Jacob/Israël et lui annonce qu’il va aller s’installer « dans la terre de Gésem Arabie » (ἐν γῇ Γεσεμ Ἀραβίαι). Dans les deux cas, le texte massorétique, dans lequel Gésem se dit « Goshèn », ne mentionne pas « Arabie ». Ce terme est un plus de la Septante. Or, Ἀραβία s’emploie ici sans article, alors que, normalement, ce mot prend l’article quand il s’agit de la péninsule arabique. En fait, les traducteurs ont en tête une division géographico-administrative de l’Égypte ptolémaïque: le nome Ἀραβία, qui se situe dans le sud est du Delta. Il est encore question de ce nom en Es 10,9 et 11,11 9.

9.

G. Dorival, « La Septante: une ou deux Arabies », à paraître dans les Mélanges Christian Robin. 6. Les communautés étrangères

349

22.4 Le vocabulaire de la loi dans la Septante Cécile Dogniez

1. Le vocable nomos Le choix du vocable νόμος pour traduire le terme hébreu ‫ תורה‬constitue le trait marquant du vocabulaire législatif de la Septante 1 car, outre la prégnance de cette équivalence majoritaire à travers l’ensemble des livres de la Bible grecque (200 fois sur les 220 emplois de ‫)תורה‬, la correspondance plus ou moins adéquate entre ces deux champs sémantiques a depuis longtemps suscité des questions. La ‫ תורה‬n’est pas la loi et notre notion moderne de loi ne recouvre que partiellement le terme νόμος. En effet, on s’accorde généralement pour rattacher le mot hébreu ‫ תורה‬2 à la racine hébraïque ‫ ירה‬qui signifie « montrer », « instruire » et renvoie plutôt dans la Bible au lexique de l’instruction, de l’enseignement, de la doctrine qui sert en réalité toujours à décrire les relations que YHWH entretient avec son peuple, dans le cadre spécifique de l’alliance. Quant au terme νόμος, c’est d’abord en grec classique ce que l’on partage, la coutume, celle de la polis, puis la forme écrite de la coutume, d’où le sens de « loi ». Mais la notion est complexe et possède tout aussi bien un sens politique en rapport avec l’autorité de l’État, un sens religieux exprimant la manifestation du divin, un sens cosmique étendu aux dimensions du monde des hommes et des dieux, sans oublier la dimension royale du νόμος. Ainsi, en grec, νόμος désigne diverses réalités; il signifie certes la loi, sans pour autant toujours inclure l’aspect juridique, légaliste, de celle-ci. Dans la Septante, le terme νόμος au singulier ne fait son apparition qu’en Ex 12,43, non pour traduire l’habituel ‫ תורה‬mais le mot ‫ חקה‬qui nomme plutôt un « rite », une « prescription » divine particulière. Or le traducteur grec de l’Exode a compris le rituel concernant la Pâque et les personnes admises à la consommer comme une exigence de la « loi » dans son ensemble, comme le confirme bien l’emploi de ‫תורה‬, à nouveau traduit par ce même νόμος, en Ex 12,49. Ainsi en Ex 12,43.49, dans la Septante, la circoncision s’inscrit dans la loi générale fondée sur l’alliance divine, de la même façon que cette institution clé de l’alliance qu’est le sabbat est définie en Ex 16,4.28 comme une loi (νόμος) globale (‫ תורה‬dans le texte massorétique). Dans le Deutéronome, l’emploi de νόμος pour ‫ תורה‬a le sens restrictif de la somme des statuts et ordonnances imposés au mont Sinaï. De manière plus précise, ce sont même les « paroles de Moïse » destinées à régler la vie d’Israël après la conquête de Canaan qui sont mises en Dt 1,5; 1.

2.

Sur la terminologie législative dans la Septante, voir en particulier S. H. Blank, « The LXX Renderings of Old Testament Terms for Law » Hebrew Union College Annual 7 (1930), 259283; L. Monsengwo Pasinya, La Notion de nomos dans le Pentateuque grec (Analecta Biblica 52), Rome 1973; A. Passoni Dell’Acqua, « Il Pentateuco dei LXX testimone di istituzioni di età tolemaica » Annali di scienze Religiose 4 (1999), 171-200, spécialement 185-187. Sur ce mot, voir par exemple G. J. Botterweck / H. Ringgren / H. J. Fabry, Theological Dictionary of the Old Testament, vol. 15, Stuttgart 2006, 609, avec l’abondante bibliographie.

350

1. Le vocable nomos

22.4 Le vocabulaire de la loi dans la Septante

4,8.44 sous l’appellation générique de νόμος/‫תורה‬, entendue au sens de l’enseignement divin transmis par Moïse; celui-ci sera codifié dans un livre nommé en Dt 17,18 avec le néologisme grec δευτερονόμιον, là où le texte massorétique parle de « double de cette loi », ‫משנה התורה הזאת‬, au sens de « copie de la loi », comme l’entendait probablement le traducteur du Deutéronome et non au sens de « seconde loi », selon l’interprétation des chrétiens hellénophones 3. Par conséquent le terme νόμος comme le mot hébreu ‫ תורה‬qu’il traduit n’a pas dans le Pentateuque le sens strict et courant de « loi », de norme juridique obligatoire, mais signifie tout à la fois les enseignements et les ordonnances liées à l’alliance divine transmises par Moïse à Israël. Les traducteurs du Pentateuque chargent même délibérément le mot grec νόμος de ce seul sens totalisant, vaste et générique de loi qui inclut tous les commandements particuliers et résume toute la volonté de Dieu pour son peuple, au point qu’ils n’emploient jamais νόμος au pluriel 4, même lorsque le texte massorétique a ‫תורת‬, par exemple en un emploi unique du mot dans la Genèse, en 26,5, pour dire en un sens non technique les « façons de faire », les « justifications » de YHWH. La Septante anticipe en quelque sorte sur le sens propre et spécifique de loi en tant qu’alliance qu’a parfois le pluriel hébreu ‫תורת‬, par exemple en Lv 26,46; Es 24,5 ou même en Ps 104/105,45, rendu par le singulier νόμος. Hors du Pentateuque, les emplois de νόμος sont peut-être une traduction stéréotypée du mot ‫תורה‬, par exemple en Es 1,10 ou 8,16, mais ils ne procèdent pas pour autant d’un légalisme qui serait propre au judaïsme hellénistique 5. Dans le livre des Proverbes, le singulier νόμος renvoie très certainement à la loi de Moïse 6, tandis que le pluriel νόμοι qualifie les enseignements du père (Pr 6,20), ceux de la mère étant désignés par le terme grec classique θεσμός, « règle, ordonnance », d’un emploi très rare dans la Septante 7. La grande souplesse sémantique du mot grec νόμος se perçoit d’ailleurs lorsque le terme prend encore une autre valeur, celle de « révélation divine » qu’a déjà le terme ‫ תורה‬chez les prophètes. En ce sens, en Es 2,3 et 30,9, par exemple, νόμος tire quasiment cette dimension prophétique des termes « parole » (λόγος κυρίου) et « visions » (τά ὁράματα) qui lui sont étroitement associés. Νόμος c’est aussi la prescription cultuelle du prêtre, comme en Ag 2,11 et en Ml 2,6. Mais le mot νόμος, ne l’oublions 3. 4. 5.

6. 7.

Cf. C. Dogniez / M. Harl, Le Deutéronome (La Bible d’Alexandrie 5), Paris 1992, 27. Sur le peu d’occurrences du pluriel νόμοι dans la Septante, voir Blank, « LXX Renderings », 279-280; Monsengwo Pasinya, Notion, 137. On trouve en effet en Esdras 7,12.14.21.25.26, dans les parties qui sont en araméen, l’emploi du mot ‫ דת‬pour signifier clairement la loi de Dieu, νόμος τοῦ θεοῦ. Cf. P. Richardson / S. Westerholm, Law in Religious Communities in the Roman Period. The Debate over Torah and Nomos in Post-Biblical Judaism and Early Christianity (Studies in Christianity and Judaism 4), Waterloo 1991, 48. Sur la confusion entre ‫ תורה‬et loi avec une accentuation progressive de l’aspect légaliste au détriment du caractère doctrinal de la loi dans le judaïsme, voir par exemple Monsengwo Pasinya, Notion, 67. Sur le vocabulaire de la loi en Proverbes LXX, voir J. Cook, « The Law of Moses in Septuagint Proverbs » Vetus Testamentum 49 (1999), 448-461. Θεσμός, employé par exemple pour nommer les « lois » de Dracon, vers 621 avant notre ère, puis peu à peu évincé dans la langue grecque par le terme νόμος, ne se trouve que six fois dans la Septante, en Pr 1,8; 6,20; Sg 14,23; 3M 6,36; 4M 8,7. 1. Le vocable nomos

351

22.4 Le vocabulaire de la loi dans la Septante

pas, servira à désigner le Pentateuque, en tant que « loi civique 8 », pour les Juifs hellénophones d’Égypte.

2. D’autres vocables À côté de cette diversité de sens pour νόμος, qui n’a donc pas dans la Septante, a priori et dans toutes ses occurrences, cette connotation légaliste qui culminera par exemple dans les livres des Maccabées, d’autres termes très variés expriment à travers l’ensemble des livres de la Bible grecque un certain nombre de prescriptions relevant du domaine juridique. L’adjectif substantivé νόμιμον désigne dans la Septante ce qui est conforme à la loi, c’est-à-dire la coutume, la « règle », ou plus précisément l’« article de loi » compris comme un point particulier du νόμος. Ce vocable est abondamment utilisé par les traducteurs grecs pour traduire trois mots hébreux: ‫תורה‬, ‫ חקה‬désignant l’« ordonnance mise par écrit », et le mot apparenté ‫חק‬, qui signifie « le décret, la loi ». Le pluriel νόμιμα nomme à la fois les règles de Dieu (Gn 26,5; Lv 10,11; Za 1,6) et celles des nations (Lv 18,30; 20,23; Jr 10,3), au sens du grec classique de « coutumes, mœurs 9 ». Au singulier, accompagné de l’épithète αἰώνιος, « éternel », en Exode, Lévitique et Nombres, νόμιμον fait plutôt référence à un précepte cultuel qu’il convient de commémorer à jamais. Les traducteurs grecs utilisent un autre mot, δικαίωμα, qui a également ce sens de « règle du droit ». Ce mot grec traduit principalement trois termes hébreux, ‫חק‬, ‫ חקה‬et parfois ‫ משפט‬désignant à l’origine « le jugement », puis « la coutume », d’où le sens de « loi ». D’usage courant dans la législation ptolémaïque, le mot grec δικαίωμα au pluriel renvoie aux « pièces à conviction » utilisées pour instruire un procès. Mais dans la Septante les traducteurs lui attribuent un nouveau sens juridique pour désigner plutôt les règlements, les lois contenues dans ces pièces 10. Que ce soit des règles de justice, concernant par exemple l’héritage (Nb 27,11) ou le meurtre d’un parent (Nb 35,29), ou des règles du droit religieux, concernant les offrandes (Nb 15,16), ou les vœux (Nb 30,17), les δικαιώματα désignent toujours des préceptes provenant de Dieu, considérés comme un droit coutumier pour Israël. L’abondance de ce mot dans la Septante souligne bien l’idée de justice. Les mots grecs κρίμα et κρίσις ont pour équivalent majoritaire le terme hébreu ‫משפט‬, « le jugement ». Alors que κρίμα est un terme assez rare en grec, la Septante l’utilise souvent au sens technique de « décision d’ordre judiciaire » (Dt 21,22), mais aussi, selon le contexte, pour signifier « la cause » (Ex 23,6), « le jugement, le droit » (Os 5,1; Ha 1,4.7.12; Za 7,9) ou « la justice » (So 2,3); au pluriel, il désigne les « sentences » 8. Cf. J. Mélèze Modrzejewski, « La Septante comme nomos. Comment la Torah est devenue une loi civique pour les Juifs d’Égypte » Annali di scienze religiose 2 (1997), 143-158. 9. Voir par exemple les emplois du mot en 1Maccabées. 10. Sur ce mot voir H. Cadell, « Vocabulaire de la législation ptolémaïque. Problème du sens de dikaiôma dans le Pentateuque » in: G. Dorival / O. Munnich (éd.), ΚΑΤΑ ΤΟΥΣ Ο’ Selon les Septante. Trente études sur la Bible grecque des Septante en hommage à Marguerite Harl, Paris 1995, 207-221.

352

2. D’autres vocables

22.4 Le vocabulaire de la loi dans la Septante

lorsqu’il s’agit des lois divines (Lv 18,5; Nb 36,13; Dt 6,20). Κρίσις, plus fréquent que κρίμα dans la Septante, est employé pour signifier tous les aspects de la notion de « jugement », c’est-à-dire « le procès » (Gn 14,7; Lv 19,15.35), « ce qui est juste » (Gn 18,19), « le droit » (Dt 18,19), « le cas à juger » (Nb 27,5; Dt 17,8), mais aussi, au pluriel par exemple, « les procédures judiciaires » définies par Dieu (Ex 15,25; Dt 4,5.14). De sens assez proche, les deux mots κρίμα et κρίσις associent clairement dans la Septante la loi à l’idée de justice. Πρόσταγμα 11 traduit dans la Septante, et même au sein d’un seul livre, toute une série de mots hébreux: ‫דבר‬, « parole »; ‫משמרת‬, « ordonnance »; ‫חק‬, ‫חקה‬, ‫מצוה‬, « ordre »; ‫משפט‬, ‫פה‬, « bouche »; et même ‫( תורה‬à trois reprises seulement en 2Par 19,10; Jr 39/32,23 et 51/44,10). Utilisé pour « l’ordre » ou « la prescription » en grec classique, ce terme abonde dans la documentation papyrologique de l’époque hellénistique où il désigne l’« ordonnance » émanant du roi Lagide, l’ordre administratif d’un fonctionnaire, ou encore la manifestation impérative d’un ordre divin. Dans le Pentateuque, sauf en Gn 47,27 où le mot désigne une ordonnance de Joseph, en tant que fonctionnaire de Pharaon, πρόσταγμα renvoie exclusivement à une injonction divine, par exemple comme unique équivalent 12 de ‫ מצוה‬en Ex 20,6; Lv 4,2; Dt 5,10; Dn 9,4 et 2Par 29,25 et 31,21 ou pour traduire quatre mots hébreux différents en Si 6,37; 39,16.18; 43,13. Malgré la variété des mots hébreux qu’il traduit, πρόσταγμα renvoie dans la Septante au sens constant de loi qui émane de Dieu. Alors que νόμος ne fait son entrée dans la Septante qu’à partir de l’Exode, ἐντολή apparaît dès la Genèse pour signifier « le commandement » (Gn 26,5), au sens juridique de la vraie « loi ». Dérivé du verbe ἐντέλλομαι utilisé par exemple en Gn 2,16 et 6,22 pour dire « l’autorité 13 bienveillante » de Dieu, ἐντολή est courant dans la langue administrative des papyrus de l’époque hellénistique au sens d’« ordre », de « prescription » et conserve dans la Septante cet aspect de norme prescrite et contraignante, en particulier lorsqu’il est employé au pluriel, seul ou associé à d’autres termes de lois, pour signifier les ordres particuliers que Dieu ordonne. Contrairement à πρόσταγμα qui présente une instabilité d’équivalences lexicographiques mais de sens très proche, ἐντολή est choisi dans la Septante comme équivalent constant de l’hébreu ‫מצוה‬. L’usage du pluriel τὰ μαρτύρια, « les témoignages », pour désigner les lois du Décalogue est nouveau dans la langue grecque. Employé plusieurs fois dans le Pentateuque, par exemple en Ex 25,16.21, pour traduire l’hébreu ‫עדות‬, μαρτύρια renvoie à ce qui est écrit sur les tablettes, sert de témoignage et a donc force de loi. Associé à 11. Voir A. Passoni Dell’Acqua, « La terminologia dei reati nei προστάγματα dei Tolemei e nella versione dei LXX » in: B. G. Mandilaras (éd.), Proceedings of the XVIII International Congress Papyrology, Athens 1986, vol. II, Athènes 1988, 335-350; Cadell, « Vocabulaire de la législation ptolémaïque », 208-209. 12. Sur cette équivalence πρόσταγμα/‫מצוה‬, voir par exemple Blank, « LXX Renderings », 262. 13. Cf. A. Pelletier, « L’autorité divine d’après le Pentateuque grec » Vetus Testamentum 32 (1982), 236-242. Sur le lexique du commandement dans le Pentateuque grec, voir aussi A. Pelletier, « Le vocabulaire du commandement dans le Pentateuque des LXX et dans le Nouveau Testament » Recherche de Sciences Religieuses 41 (1953), 519-524. J. A. L. Lee, « A Lexical Study Thirty Years on, with Observations on « Order » Words in the LXX Pentateuch » in: S. M. Paul / R. A. Kraft et al. (éd.), Emanuel. Studies in Hebrew Bible, Septuagint and Dead Sea Scrolls in Honor of Emanuel Tov, Leiden/Boston, MA 2003, 513-524. 2. D’autres vocables

353

22.4 Le vocabulaire de la loi dans la Septante

δικαιώματα en Dt 4,45 et 6,20 ou à ἐντολαί en Dt 6,17, cet autre terme législatif nomme les prescriptions divines énoncées par Moïse ou celles commandées par Dieu. Ce sens spécifique du pluriel μαρτύρια se retrouve ailleurs que dans le Pentateuque, en 4Règnes 17,15; 23,3; 1Par 29,19; 2Par 34,31, et principalement dans le Psautier 14.

3. Conclusion Pour restituer la terminologie variée de ce qui a trait à l’enseignement divin mais aussi aux prescriptions de caractère plus juridique, les traducteurs grecs ont utilisé un vocabulaire issu du grec classique mais aussi fortement influencé par la langue administrative du milieu ptolémaïque. Ils ont en effet su exploiter la souplesse sémantique d’un terme comme νόμος sans le limiter à sa dimension juridique, ou légaliste comme on l’a trop souvent affirmé, et n’ont pas hésité à doter certains termes comme δικαιώματα ou μαρτύρια d’un sens nouveau. Enseignement et préceptes édictés par l’autorité divine à Moïse lors de l’alliance conclue sur le Sinaï, l’esprit des lois bibliques dans la Septante traduit, à travers la complexité du vocabulaire grec, non l’expression d’un droit contraignant et légaliste mais suggère plutôt l’idée de justice, dans toute l’amplitude de son sens, que ce soit les règles du droit – les lois –, ce qui est juste, le jugement, le procès voire les procédures judiciaires.

14. Sur l’emploi du mot μαρτύρια, voir Dogniez / Harl, Deutéronome, 53-54.

354

3. Conclusion

VI The style of the Septuagint

23. The language of the Septuagint between Greek and Hebrew Alexis Léonas

1. Conceptual Framework The study of the linguistic features of the Septuagint has often been reduced to two conflicting positions, which tend to dominate the entire debate. Formulated in a simplistic way, the first view contends that Septuagint Greek is an alien development within the Greek language, an artificial and/or imperfect product, created by the Septuagint translators, while the second view claims that the Septuagint idiom is a part of Greek language and literary history, its occasional divergences from mainstream Greek being due to our imperfect knowledge of the latter. We will begin by taking a closer look at both positions, exploring their origins and methodological implications. Both views originate from Christian and Jewish apologetics in Late Antiquity. The critics of Judaism and Christianity in first centuries CE denounced what they saw as the poor literary quality of the Biblical writings; Jewish and Christian writers consistently denied these accusations, insisting on the literary soundness of the Scriptures (although another line of apologetic argument suggested that the lack of literary sophistication was actually an advantage, a witness to the pristine truth contained in the Writ). 1 Accusations of poor expression and “barbarity” have been preserved in Origen’s Contra Celsum (VII, 9-11; VII, 59), Julian’s Epistle (89,295d), Cyril of Alexandria’s Contra Iulianum Imperatorem (VII, 232), Basil of Cesarea’s Homiliae super Psalmos (44§4), Theodor1.

Cf. C. Andersen, Logos und Nomos. Die Polemik des Kelsos wider das Christentum (AKG 30), Berlin 1955; G. J. M. Bartelink, “Observations de saint Basile sur la langue biblique et théologique” Vigiliae Christianae 17/2 (1963), 85-105; S. Benko, “Pagan Criticism of Christianity During the First Two Centuries AD” in: W. Haase (ed.), Aufstieg und Niedergang der Römischen Welt. Geschichte und Kultur Roms im Spiegel der Neueren Forschung, vol. II, Principat, 23.2, Berlin/New York, NY 1980, 1055-1118; M. Borret, “L’Écriture d’après le païen Celse” in: C. Mondésert (ed.), Le monde grec ancien et la Bible (Bible de tous les temps 1), Paris 1984, 171-193; J.-C. Fredouille, “Rencontre de l’Évangile et de la culture dans l’Antiquité” Les Cahiers protestants 6 (1984), 8-15; J.-N. Guinot, “Théodoret de Cyr. Une lecture critique de la Septante” in: G. Dorival / O. Munnich (eds.),“Selon les Septante”. Hommage à Marguerite Harl, Paris 1995, 393-407; M. Harl, Le Déchiffrement du sens. Études sur l’herméneutique chrétienne d’Origène à Grégoire de Nysse (Collection des Études Augustiniennes, Série Antiquité 135), Paris 1993; A. Meredith, “Porphyrius and Julianus against the Christians” in: Haase, Aufstieg und Niedergang, vol. II.23.2, 1119-1149; B. Neuschäfer, Origenes als Philologe (Schweiz. Beiträge zur Altertumswissenschaft 18), Basel 1987; J. Schwartz, “Celsus redivivus” RHPR 53 (1973), 399-405; F. M. Young, “The Rhetorical Schools and their Influence on Patristic Exegesis” in: R. Williams (ed.), The Making of Orthodoxy, Essays in Honour of Henry Chadwick, Cambridge, 1989, 182-99. 1. Conceptual Framework

357

23. The language of the Septuagint between Greek and Hebrew

etus of Cyrus’ Graecarum Affectionum Curatio (§ 1), and Isidorus of Pelusium’s Epistle (IV, 28; IV, 67). Interesting “impartial” comments on the language of the Apocalypse by Dionysius of Alexandria are quoted in Eusebius’ Historia Ecclesiae. (VII, 24-26). The theme of the appreciation of obscurity can be found in Clement of Alexandria (Stromata VI, 126-127), Origen (De principiis, III, 16; IV, 2,9 and Contra Celsum, II, 6), and John Chrysostom (De prophetiarum obscuritate, I,3, II,5). As early as the fourth century CE, exegetes belonging to the Antiochene school were able, largely because of their familiarity with Syriac, to establish a connection between unusual turns of phrase in the Septuagint and the use of literal translation technique. Cf. Theodoretus of Cyrus Interpretatio in Isaïam, 65, 20 (20, 455) and Interpretatio in Danielem, 11, 13), Theodore of Mopsuestia (Commentarii in Psalmos, 28, 8), and Diodorus of Tarsus (Commentarii in Psalmos, 31,3b; 48,19b).

Beginning in the sixteenth and seventeenth centuries, scientific discourse rearticulated these viewpoints in terms of new knowledge and methodology. The dispute between the Hebraists and the Purists in the seventeenth and eighteenth centuries can be considered the first scholarly attempt to address the nature and origin of Biblical Greek. The polemic began with the publication of Sebastian Pfochen’s Diatribe de Linguae Graecae Novi Testamenti puritate (Amsterdam 1629), in which the author advocated the pure “Greekness” of the biblical idiom, ascribing all its peculiarities to its specific style. Thomas Gataker advanced a counter-argument in his De Novi Instrumenti Stylo Dissertatio … (London 1648), suggesting that the corrupted Greek in the Bible could only be the result of a separate dialect (De Novi, 58, 262-263). This particular dialect would later come to be defined as Judeo-Greek. 2 If the origins of the Hebraism/Hellenism dichotomy can safely be pronounced polemic—that is to say motivated by concerns other than actual study 3—this opposition has proved a useful tool for the description of Septuagint language. In purely linguistic terms, the opposition between Hebraism and Hellenism can be reformulated as an attempt to grasp the linguistic and stylistic peculiarity of the Septuagint. It is also clear that this dichotomy remains useful only if it operates as part of a well-defined, supple language description.

2. Methodology The debate of the Hebraists and the Purists gave rise to a methodology still in use today. One approach consists in extracting Septuagintal words or turns of phrase from 2.

3.

Studies produced in course of this polemic have been collected by T. C. Vander Honert (Syntagma dissertationum de stylo Novi Testamenti graeco, Amsterdam, 1702) and by J. Rhenferd (Dissertationum philologico-theologicarum de stylo Novi T. syntagma, Leuwarden, 1702). For a detailed account of the debate and its later development, see P. Vulliaud, La Clé traditionnelle des Évangiles, Paris, 1936 (reprint 1978); J. Ros, De studie van het Bijbelgrieksch van Hugo Grotius tot Adolf Deissmann, Nijmegen 1940. See also A. Léonas, “The Septuagint in Premodern Study: A Bibliography” BIOSCS 41 (2008), 93-113. As late as in 1638 the University of Wittenberg declared that searching for barbarisms or solecisms in Biblical Greek is a blasphemy against the Holy Ghost. Cf. J. Irmscher, “Der Streit um das Bibelgriechisch” Acta Antiqua 7 (1959), 130.

358

2. Methodology

23. The language of the Septuagint between Greek and Hebrew

their context and pronouncing them “un-Greek,” that is lacking attestation in the extant corpus of original Greek writings and thus incomprehensible to the ancient reader. A contrary approach attempts to find examples of those usages in Greek literature and present these data as indicators that the usages were linguistically safe and sound. The most eminent proponent of the latter approach was Adolf Deissmann, whose contribution 4 consisted in introducing the rich data gleaned from Egyptian papyri and Hellenistic epigraphy into the field of Septuagint and New Testament study. Deissmann was thus able to demonstrate that much of the lexica and phraseology previously considered exclusively biblical did in fact reflect the Koine of its time. Words such as πρωτότοκος, “first-born,” διαταγή, “commandment,” καταπέτασμα, “curtain,” and turns of phrase such as βλέπειν από, “beware of something,” εἶναι εῖς, “become something,” or κρίνειν τὸ δίκαιον, “consider to be right,” have been successfully vindicated as belonging to the common stock of the Greek language of the epoch. The lasting impact of Deissmann’s research on Koine resulted in the universal dismissal of the hypothesis of a Judeo-Greek dialect. 5

Yet, the methodological weakness of both lines of argument is apparent. The absence of a certain word or turn of phrase from the extant corpus of Greek literature is an argument from silence. It cannot be decisive, since a large segment of Greek literature —particularly of the Hellenistic period—is lost to us. On the other hand, the occurrence of a given Septuagint locution somewhere in the vast Greek literary corpus is not

4.

5.

Deissmann’s main books devoted to this subject are Bibelstudien. Beiträge, zumeist aus den Papyri und Inschriften, zur Geschichte der Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums (Marburg 1895), Neue Bibelstudien. Sprachgeschichtliche Beiträge, zumeist aus den Papyri und Inschriften, zur Erklärung des Neuen Testaments (Marburg 1897), and Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistischen Welt (Tübingen 1908, and 1923 for the fourth edition). This approach has been further developed in later study, cf. A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus. Beiträge zur Geschichte und Beurteilung der Koine, Strasbourg 1901; M. J. Rouffiac, Recherches sur les caractères du grec dans le Nouveau Testament d’après les Inscriptions de Pirène, Paris 1911; H. Ljungvik, Beiträge zur Syntax der spätgriechischen Volkssprache, Uppsala-Leipzig 1926; F.-M. Abel, Grammaire du grec biblique suivie d’un choix de papyrus, Paris 1927; L. Rydbeck, Fachprosa, vermeintliche Volkssprache und Neues Testament, Uppsala 1967; G. H. R. Horsley, New Documents Illustrating Early Christianity. A Review of Greek Inscriptions and Papyri published in 1976-1979, 4 volumes, North Ryde/Marrickville, 1981-1987; G. H. R. Horsley, “Divergent views on the Nature of the Greek in the Bible” Biblica 65, (1984), 393-403; O. Montevecchi, “Dal paganismo al christianesimo: aspetti dell’evoluzione della lingua greca nei papyri dell’Egitto” Aegyptus 37 (1957), 41-59; O. Montevecchi, “Quaedam de graecitate Psalmorum cum papyris comparata” in: Proceedings of the 9th International Congress of Papyrology, Oslo 1961, 292-310; O. Montevecchi, “Continuità ed evoluzione della lingua greca nella Settanta e nei papyri” in: Actes de Xe Congrès International de papyrologie, Warsaw 1964, 39-49; A. Passoni dell’Acqua, “Ricerche sulla versione dei LXX e i papyri” Aegyptus 62 (1982), 173-194; F. Vattioni, “La lessicografia dei LXX nei papyri” Studia Papyrologica 19 (1980), 39-59; J. A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SBL Septuagint and Cognate Studies series, 14), Chico, CA 1983. A. Deissmann, Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistischen Welt (fourth edition), Tübingen 1923, 70-71, 80, 93-94, 96-97. 2. Methodology

359

23. The language of the Septuagint between Greek and Hebrew

in itself sufficient proof. Parallels of this type can only be conclusive if one has also examined the context, specifying the content, the genre, and the style. For instance, the figura etymologica βλέποντος ἔβλεπον, which occurs in verse 447 of Aeschylus’ Prometheus Bound (circa 456 BCE), might be used to illustrate similar expressions frequent in the Septuagint (e. g. γίνωσκων γνώσῃ, in Gen 15:13, or ἰδὼν εἶδον, in Ex 3:7). But this example remains a single occurrence in a poetic text already deemed opaque by its contemporaries. Can this instance then be used to clarify similar expressions in entirely different contexts, such as historical narratives or legal texts?

Methodological uncertainty probably accounts for the lack of a definitive map or list of Hebraisms. Identifying Septuagint Hebraisms on the basis of their resemblance to their Hebrew prototypes implies associating expressions and turns of phrase that do not belong together once the analysis is done from the point of view of Greek use. On the other hand, reliance on purely Greek material carries the risk of overlooking the logic behind various textual phenomena in the Septuagint. For example, the Hebrew infinitive absolute in the Septuagint is often translated as a combination of a conjugated verb and a cognate (or semantically close) noun in the dative (e. g. ‫ – אכל תאכל‬βρώσει φάγῃ “you will eat as/by eating,” in Gen 2:16). It is also frequently rendered by the combination of a verb and its participle (‫ – הרבה ארבה‬πληθύνων πληθυνῶ “I will greatly multiply,” in Gen 3:16) 6. From the point of view of the Hebrew, the two Septuagint expressions reflect the same linguistic reality, viz. the infinitive absolute. Can the Greek linguistic description equally rely on this connection? In principle, the projection of elements of one language into another language is not acceptable linguistic methodology. However, the etymological figure as such is not unknown in Greek literature. Thus a comparative analysis of the Greek renderings of the infinitive absolute might not be impossible on a purely Greek basis. Yet if we study similar turns of phrase in the Septuagint we will find analogous formulae with no counterparts in the Hebrew source-text, e. g. διέθεντο διαθήκην for ‫“ ויכרתו ברית‬they made alliance” (Gen 21:32).

At this stage it transpires that Hebraism may sometimes be a misnomer for the specificity of the Septuagint idiom. The main issue of Septuagint language study should be the identification and systematic description of its traits, be they Hebraisms or not. Interestingly, the criteria of such identification have never been seriously discussed in modern research, beyond the broadest assertions of “bad Greek.” 6.

Cf. H. St. J. Thackeray, “Renderings of the infinitive Absolute in the LXX” JTS 9 (1908), 597601; H. Kaupel, “Beobachtungen zur Übersetzung des Infinitivus Absolutus in der Septuaginta (G)” ZAW 20 (1945-1948), 191-201; R. Sollamo, “The LXX Renderings of the Infinitive Absolute used with a paronymous finite verb in the Pentateuch” in: N. Fernández-Marcos (ed.), La Septuaginta en la investigación contemporánea. V Congreso de la IOSCS (Textos y Estudios “Cardenal Cisneros” 34), Madrid 1985, 101-113; E. Tov, “Renderings of Combinations of the Infinitive Absolute and Finite Verbs in the Septuagint – their Nature and Distribution” in: D. Fraenkel et al. (eds.), Studien zur Septuaginta – Robert Hanhart zu Ehren (Mitteilungen der Septuaginta Unternehmung 20), Göttingen, 1990, 64-73; M. S. Krause, “The Finite Verb with Cognate Participle in the New Testament” in: S. E. Porter / D. A. Carson (eds.), Biblical Greek Language and Linguistics: Open Questions in Current Research (JSNT Supplements 80), Sheffield 1993, 187-206; A. Léonas, Recherches sur le langage de la Septante (OBO 211), Fribourg/Göttingen 2005, 194-196.

360

2. Methodology

23. The language of the Septuagint between Greek and Hebrew

3. Toward a Definition of Hebraism There is a degree of terminological uncertainty which needs be clarified here. The secondary literature often hesitates between the terms “Hebraism” and “Semitism.” Other terms, such as “literalism,” “Aramaism,” or even “Septuagintalism,” further diversify the discourse. Without aspiring to provide full-fledged definitions, we will now explore the nuances these terms suggest. Semitism, as opposed to Hebraism, points to the wider linguistic context of a given usage. The Septuagint translators had a wide-ranging knowledge of the Semitic vernacular, which included Aramaic influences (whence Aramaisms) as well as an older Semitic language layer (whose presence becomes apparent in the so-called Arabisms 7). The term “Semitism” highlights the variety of this vernacular knowledge, unrestricted by the boundaries of the Biblical Hebrew. It is particularly appropriate in the study of the New Testament idiom, whose background was most certainly vernacular and cannot be defined as exclusively Hebraic. Aramaism does not in fact refer to the quality of the Biblical Greek: it indicates the influence of Aramaic (and its different forms) on the Septuagint translators’ knowledge of Biblical Hebrew. Recourse to the Aramaic meaning of certain words and expressions 8 and Aramaic vocalization in the transliterations 9 bears witness to the translators’ cultural and linguistic background. Literalism covers practically the same textual reality as “Hebraism.” This term highlights the procedure responsible for the majority of cases of peculiar wording in the Septuagint. However, literal translation is not the unique source of the specificity of the Septuagint Greek. There are cases where literal renderings did not create linguistic tension in the Greek text 10; in other cases, the linguistic peculiarity of Septuagint Greek is only loosely connected to the source text. Thus, “Hebraism” remains the proper term for the general description of the linguistic specificity of the Septuagint, while “literalism” can be retained for the explanation of the genesis of particular cases. Septuagintalism is an awkward term, useful mostly in lexicography, where it denotes usages limited to (or particularly frequent in) the Septuagint, as opposed to the rest of the Greek corpus. Coming back to the Hebraisms, we can see that from a purely linguistic point of view the difficulty of this notion resides in the reference it carries to an extraneous 7. Cf. G. R. Driver, “Supposed Arabisms in the Old Testament” JBL 55 (1936), 50-64. 8. Cf. J. Joosten, “On the LXX Translators’ Knowledge of Hebrew” in: B. A. Taylor (ed.), X Congress of the IOSCS, Oslo 1998 (SBL SCS Series 51), Atlanta, GA 2001, 165-179; J. Joosten, “The Septuagint as a source of information on Egyptian Aramaic in the Hellenistic Period” in: H. Gzella / M. L. Folmer (eds.), Aramaic in its Historical and Linguistic Setting (Akademie der Wissenschaften und der Literatur Mainz, Veröffentlichungen der Orientalistischen Kommission 50), Wiesbaden 2008, 93-105. 9. Cf. Léonas, Recherches, 62-64. 10. For example, even the literary translation of Hebrew idiomatic expressions can sometimes yield perfectly idiomatic Greek: C. Dogniez, “The Greek Renderings of Hebrew Idiomatic Expressions and their Treatment in the Septuagint Lexica” JNSL 28/1 (2002), 6-10. See a remark on the same subject by W. F. Howard in: J. H. Moulton / W. F. Howard / N. Turner, A Grammar of New Testament Greek, vol. 2, Edinburgh 1929, 14. 3. Toward a Definition of Hebraism

361

23. The language of the Septuagint between Greek and Hebrew

linguistic system. The term “Hebraism” implies that certain linguistic peculiarities of the Septuagint are quasi-independent from the rest of the Septuagint linguistic system by virtue of having been motivated by attempts to represent the expressions of the Hebrew tongue. One can see here a conflict between the synchronic and the diachronic approaches. The synchronic approach yields a description of a complete and self-contained linguistic system, where values are established by contrasting different language units, which form an inter-depending network. The diachronic approach would, in this particular case, explore the underlying source-text and elucidate the causes of any anomaly through this connection. However, at this stage, the diachronic approach presents a serious methodological drawback when applied to the Septuagint: because it is a translation, practically all elements of the Septuagint text can be traced back to the Hebrew. This amounts to saying that everything in the Septuagint is a Hebraism. In other words, the diachronic approach does not afford us the possibility of distinguishing between Hebraisms and non-Hebraisms in the Greek text. To be able to pinpoint any such phenomena we must first take recourse to synchronic description. Language description offers two basic concepts relevant at this point to our study: linguistic norm and linguistic system. The two notions are complementary: the expression “linguistic system” describes the basic functioning of a language as a means of communication, whereas the notion “linguistic norm” reflects the speaker’s perception of the language. While the language system aims to optimize the communication process, exploiting to the full the structural potential of a given language, the norm endeavors to integrate language use into a set of socio-cultural values. The system may go unnoticed to the average language user, who acquired the language in the usual way. The norm, on the contrary, is always more or less consciously perceived; it is imposed by society and transmitted through learning. Average language speakers typically perceive breach of norm as a language error, confusing the acquisition of a norm with language learning and generally mistaking the norm for the system. The early stages of linguistic science were dominated by the elaboration of norms; the notion of “Hebraism” was also initially resulted by literary norm-centered research. A clear distinction between system and norm can help us in our attempts to describe the language of the Septuagint. It enables us to provide a twofold definition of “Hebraism” in terms of these concepts. In terms of the linguistic norm, a Hebraism is a word, turn of phrase, or composition feature incompatible with the literary norm of written Greek. Going back to the diachronic description, we can now ask whether such a nonnormative word or turn of phrase reflects an attempt to render a particular Hebrew expression. This may well be so, yet the picture would be incomplete, for the calques of Hebrew that do not contradict the norm would go unnoticed. There could also be turns of phrase incompatible with the norm and yet unrelated to the Hebrew Vorlage. In terms of the language system, Hebraisms can be defined as words, turns of phrase, or composition features that create tension from the point of view of content, i. e. create any kind of semantic spillover or difficulties unsupported by the context in either lexical, grammatical, or compositional features. Systemic description yields a much wider picture of the peculiarities of Septuagint 362

3. Toward a Definition of Hebraism

23. The language of the Septuagint between Greek and Hebrew

language. Although it is out of the scope of the language system to decide whether any particular feature was motivated by an attempt to render a Hebrew expression, the definition it provides of Hebraism opens the way for the diachronic approach. The existence of a textual element unsupported by the context enables us to seek its motivation elsewhere, e. g. in the underlying Hebrew text. An interesting question concerning the relation to the Hebrew arises at this point of our analysis. Are all Hebraisms resulted by literary translation technique? What about cases where the translators were only imitating other calques, producing formulations not vouchsafed by the Vorlage? We have already cited the translation of ‫ ויכרתו ברית‬as διέθεντο διαθήκην in Gen 21:32; other examples of the same tendency can be easily adduced. In LXX Mal 3:9, the Hebrew expression ‫( במארה אתם נארים‬literally, “you will be cursed by a curse”) is translated as ἀποβλέποντες ἀποβλέπετε, literally: “looking you will look.” 11 The translator has probably mistaken a relatively rare root ‫ ארר‬for the well-known ‫ראה‬. Yet his eagerness to introduce the figura etymologica into the Greek text appears, in this particular instance, as a rather voluntary use of a ready-made formula only vaguely supported by the source-text (cf. the equivalence ‫ נארים‬// ἀποβλέπετε). 12 There exists a whole literature of Septuagint related writings originally produced in Greek and yet marked by the presence of the Hebraisms. The most salient case in the Septuagint itself is the Wisdom of Solomon, written originally in Greek in a style close to the Septuagint idiom. No less important is the evidence of many New Testament texts (e. g. Luke or John). 13 Thus, the relation to the Hebrew Vorlage of a given Hebraism does not appear to be rigidly fixed: the translators had a certain freedom to use expressions first coined as literary calques as building blocks to render meanings in new contexts. In this perspective, one should also mention the suggestion of Jozef Vergote, 14 who provided interesting evidence that many “Hebraic” phrases in Septuagint Greek may in fact be “Egyptianisms.” Vergote demonstrated that numerous parallels to Hebraisms in the Greek papyri could actually could actually be due to interference of the Coptic (or the demotic Egyptian) language. For example, the use of εἶς in the sense of τις, the repetition of a word resulting in a distributive sense (e. g. δύο δύο, Gen 6:19), or the instrumental use of ἐν translate expressions that are also typical in Coptic. It is therefore possible that Coptic usages could underlie similar Greek formulations in the papyri and the Septuagint. Indeed, the Septuagint expression mode may actually reflect an Egyptian linguistic background. 15 11. This translation does not address the issue of the meaning of ἀπό, which might need a longer discussion. 12. Cf. L. Vianès, Malachie (La Bible d’Alexandrie 23.12), Paris 2011, 149. 13. Cf. D. Dimant, “The Problem of Non-Translated Biblical Greek” in: C. E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies. Jerusalem 1986 (Septuagint and Cognate Studies 23), Atlanta, GA 1987, 1-19; A. Léonas, “The Poetics of Wisdom: Language and Style in the Wisdom of Solomon” in: E. Bons / T. J. Kraus (eds.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic features of the Septuagint (FRLANT 241), Göttingen 2011, 99-126. 14. J. Vergote, “Grec biblique” in: SDB, vol. 3, 1354-1360. 15. Cf. also Joosten, “Egyptian Aramaic.” 3. Toward a Definition of Hebraism

363

23. The language of the Septuagint between Greek and Hebrew

4. In Search of Hebraisms Trying to assess the criteria that have been traditionally applied in the search for Hebraisms we come to the following list: Closeness to the Hebrew wording. This is arguably the origin of many remarks on the Hebraisms in the Septuagint. 16 However, from a strictly linguistic point of view this resemblance can teach us nothing about the “Greekness” of the expressions in question. The Septuagint translators most probably drew on their active and passive knowledge of Greek to produce those formulations, while the ancient reader would in any case have been reading his Greek version without recourse to the Vorlage. We must also keep in mind that overlaps between the expressive functions of two languages may occur even between widely divergent cultures. This would be especially frequent in languages belonging to the same cultural and historical setting. 17 It has been observed that the rendering of seemingly untranslatable Hebrew idiomatic expressions in the Septuagint often resulted ultimately in intelligible Greek. 18 Departure from standard Greek usage. Use of a word or a turn of phrase absent from standard Greek usage is a valid criterion for identifying Hebraisms, although it must be used with caution. Such departure from mainstream Greek is an indication of the rarity of a given word or expression, but one cannot conclude from this that any given language feature is “un-Greek.” Nevertheless, rarity is in itself an important linguistic indicator. The main problem is how to interpret it, in other words how to fit the evidence into a classification system. We will come back to the classification problems in the following section. External evidence from ancient readers of the oddity of a particular usage. The study of the reception of the Septuagint by early exegetes has brought to light valuable linguistic data. 19 Early commentaries can be used as touchstones of the linguistic quality of the Septuagint. Despite the chronological gap which separates the Septuagint translators from most of the exegetes (Church Fathers generally from the third century CE

16. Cf. H. S. Gehman, “The Hebraic character of LXX Greek” VT 1 (1951), 81-90; H. S. Gehman, “Hebraisms in the Greek version of Genesis” VT 3 (1953), 141-148; U. Rapallo, “Calchi ebraici” Rendi conti dell’Instituto Lombardo 103 (1969), 369-437; U. Rapallo, Calchi ebraici nelle antiche Versioni del ‘Levitico’, Studio sui Settanta, la Vetus Latina e la Vulgata (Studi Semitici 39), Rome 1971; I. Soisalon-Soininen, “Zurück zur Hebraismenfrage” in: D. Fraenkel et al., Studien zur Septuaginta, 35-51. 17. Cf. remarks on similarities between Hebrew and Greek metaphoric systems in J. P. Brown, Israel and Hellas (BZAW 231), Berlin/New York, NY 1995; J. P. Brown, Israel and Hellas, vol. II, Sacred Institutions and Roman Counterparts (BZAW 276), Berlin/New York, NY 2000; J. P. Brown, Israel and Hellas, vol. III, The Legacy of Iranian Imperialism and the Individual (BZAW 299), Berlin/New York, NY 2001. See also H. B. Rosén, L’Hébreu et ses rapports avec le monde classique. Essai d’évaluation culturelle (Comptes rendus du Groupe linguistique d’études chamito-sémitiques, supplément 7), Paris 1979. 18. Dogniez, “Greek Renderings”; J. Joosten, “Translating the Untranslatable: Septuagint Renderings of Hebrew Idioms” in: R. Hiebert (ed.), “Translation Is Required”: The Septuagint in Retrospect and Prospect, Leiden/Atlanta, GA 2010, 59-70. 19. Cf. M. Harl, “L’usage des commentaires patristiques pour l’étude de la Septante” Revue des sciences religieuses 73,2 (1999), 184-201.

364

4. In Search of Hebraisms

23. The language of the Septuagint between Greek and Hebrew

on), one could argue that their understanding, lack of understanding, or misunderstanding of the Septuagint roughly reflects the contemporary appraisal of the translators’ language use. For example, Hadrian (fifth century CE) 20, a representative of the Antiochene exegetical tradition whose Isagogé is by far the most complete assessment of the difficulties of Biblical Greek by a Greek speaker, confirms some of the judgments current in modern scholarship. Hadrian comments on the strangeness of the instrumental use of ἐν, 21 on the energetic repetition of the verb (i. e. the infinitive absolute), 22 and on the unusual metaphoric sense attributed to certain words (γλῶσσα, ὁδός, υἱός). 23 Furthermore, some of his remarks on the peculiarity of certain usages shed new light on the subject matter. Hadrian finds the plural use of the word οὐρανός to be strange, 24 as well as the metaphoric use of the verb κρύπτω, which actually means φυλάσσω 25; both of these usages had gone uncommented by modern scholars. However, even this kind of evidence cannot always be taken at a face value. Hadrian also made commentaries motivated by reasons other than linguistic. For instance, he indicates that the cognate noun and verb ὑπομονή, “perseverance,” and ὑπομένειν, “to endure,” used in relation to God (e. g. “endure the Lord,” Ps 26:14; 36:34; 38:8; 68:21; Jer 14:8; Isa 59:9; Zeph 3:8) should actually denote “expectation” (προσδοκία), not endurance. 26 One could argue that this remark reflects a linguistic reality and that the Septuagint translators, in this particular instance, introduced an alien expression that departed from standard usage. However, Hadrian’s unwillingness to accept ὑπομονή can better be explained by the rationalistic tendency of the Antiochene school, to which he belonged. 27 Other Christian authors of roughly the same epoch, such as the Desert Fathers Barsanuphius and John of Gaza, 28and even Origen 29, had no difficulty understanding the concept of “enduring God,” which fitted their spiritual experience. Therefore, the testimony of the ancient readers cannot be used in an unqualified way. Recourse to this type of evidence also requires a careful study of context, with special attention to literary genre, style, and content.

The above criteria for identifying Hebraisms appear to be of variable soundness. The evidence of ancient exegesis represents an important contribution in that it allows us to discover the turns of phrase that seemed problematical in Antiquity and/or to confirm the validity of our own observations on the language of the Septuagint. However it is greatly limited by the fragmentary nature of the sources, not to mention the biases, theological and other, of the ancient Bible commentators. Resemblance to the Hebrew 20. F. Goessling (ed.), Adrians Εἰσαγωγὴ εἰς τὰς θείας Γραφάς. Aus neu aufgefundenen Handschriften herausgegeben, übersetzt und untersucht, Leipzig 1887; for a linguistic analysis of Hadrian’s remarks, cf. Léonas, Recherches, 145-238. 21. Cf. Goessling, Adrians Εἰσαγωγὴ, § 105. 22. Cf. Goessling, Adrians Εἰσαγωγὴ, § 104. 23. Cf. Goessling, Adrians Εἰσαγωγὴ, § 74, 93. 24. Cf. Goessling, Adrians Εἰσαγωγὴ, § 72. 25. Cf. Goessling, Adrians Εἰσαγωγὴ, § 65. 26. Cf. Goessling, Adrians Εἰσαγωγὴ, § 76. 27. Cf. Léonas, Recherches, 170-174. 28. Cf. F. Neyt / P. De Angelis-Noah / L. Regnault (eds.), Barsanuphe et Jean de Gaza, Correspondance, vol. I.ii, (SC 427), Paris 1998, 354-357. 29. Cf. Excerpta in Psalmos (PG 17), 125C. 4. In Search of Hebraisms

365

23. The language of the Septuagint between Greek and Hebrew

cannot be considered a valid criterion in itself, yet it can contribute to the identification of expressions that appear at variance with the standard Greek usage.

5. Classification Problems Are Hebraisms to be regarded as isolated phenomena, lapses of negligent or overzealous translators, or do they constitute a system? How does this system relate to the general linguistic system of the Septuagint? A traditional approach saw the phenomenon of Hebraisms as something of a sum total of the differences between the Hebrew and the Greek languages. During the process of translation, these differences came to light in a Greek text, the Septuagint, where they can be identified through comparison with Hebrew grammar. The most comprehensive classification using this scheme is the one proposed by Robert Helbing. 30 He recognized four types of Hebraisms within the Septuagint language: 1. syntactic Hebraisms, which could also be called grammatical Hebraisms, e. g. the expressions ἐπικαλεῖν ἐν and ἀκούειν ἐν; 2. lexical Hebraisms, i. e. words used with a meaning unattested elsewhere; 3. phraseological Hebraisms, e. g. formulae such as ἀπὸ προσώπου τινός or εἰς χεῖρά τινος; 4. stylistic Hebraisms, or peculiarities of composition (parataxis, ellipses, etc.). This distribution of the traits of the Septuagint idiom along the main coordinates of language description implies that Septuagint Greek is a language apart. If this is true, we should expect its peculiarities to be distributed on all language levels, from vocabulary to grammar. But is this supposition justified? When Helbing wrote his study, the existence of a Judeo-Greek dialect was no longer admitted. His classification implies nevertheless that the Hebrew interference existed on practically all language planes. This amounts to postulating a specific idiom for the Septuagint; if not a spoken Judeo-Greek, then at least a specific translation language. Such a specialized translation language appears as a sub-system operating within the larger linguistic system of Koine Greek. Such a sub-system could be of two main types: dialect and style. Since there is no historical evidence of the existence of a JudeoGreek dialect, we are left to search for a peculiar writing style. Helbing’s classification system can be revised to accommodate this hypothesis by rearranging the data in the following way: syntactical and phraseological Hebraisms can be merged in a single entry of technical formulae and the unusual lexemes reclassified as neologisms or archaisms. As regards the composition features, Helbing himself recognized them to be traits of style. The style hypothesis is supported by evidence of Septuagint literary devices that cannot be traced back to the Hebrew Vorlage. 31 The existence of literary texts written 30. R. Helbing, Die Kasussyntax der Verba bei den Septuaginta. Ein Beitrag zur Hebraismenfrage und zur Syntax der Κοινή, Göttingen 1928, vi-x. Cf. asimilar division in P. Grelot, “Sémitismes (dans le Nouveau Testament)” in: SDB, vol. 12, fascicule 67, 1992, 352-357, s. v. 31. Cf. T. Muraoka, “Literary Device in the Septuagint” Textus 8 (1973); J. K. Aitken, “Rhetoric and Poetry in Greek Ecclesiastes” BIOSCS 38 (2005), 55-77; J. Joosten, “Rhetorical Ornamen-

366

5. Classification Problems

23. The language of the Septuagint between Greek and Hebrew

in imitation of the Septuagint further corroborates this assumption. Can we then speak of a specifically “Hebraic” or Septuagintal writing style? 32 Before attempting to resolve this question, let us explore its implications. If we admitted the existence of a specific Septuagint style, we could define Hebraisms not as calques of Hebrew but as marks of this particular style (although the two would obviously be closely related). In this way we would be able to account for “Hebraisms” in Septuagint passages that lack a Vorlage. A systematic stylistic description of the Septuagint would then become possible. Such a description would necessarily include Hebraisms, but would not be limited to them. Furthermore, it would allow us to evaluate the role of Hebrew calques in the formation and functioning of this specific style. A study of Septuagint style will necessarily be based on criteria different from those used by Hellenistic literary critics. Although sensitive in many ways to style issues, Late Antiquity limited its notion of style to acknowledged literary works. 33 Texts which were unquestionably marked by a specific style but happened to be out of the literary scope (for example the Elements of Euclid or, on the other end of the spectrum, the Magical Papyri) were not given this type of consideration. And there is little evidence of literary recognition of the Biblical texts in Antiquity. However, the treatise On the Sublime by Pseudo-Longinus (circa first century CE) does discuss the opening lines of the book of Genesis in a purely literary way. 34 According to Pseudo-Longinus, the creation narrative conveys to the reader the literary effect of the “sublime,” much in the same way as the Homeric descriptions of divine manifestations. Nevertheless, despite echoes of a literary reception of the Septuagint in the writings of Galen, Longinus remains a rather isolated case. 35 The evidence of the Church Fathers on this subject is more difficult to assess, because of theological bias.

32.

33.

34.

35.

tation in the Septuagint: The Case of Grammatical Variation” in: Bons / Kraus, Et sapienter et eloquenter, 11-22; J. M. Dines, “Stylistic Invention and Rhetorical Purpose in the Book of the Twelve” in: Bons / Kraus, Et sapienter et eloquenter, 23-48; E. Bons, “Rhetorical Devices in the Septuagint Psalter” in: Bons / Kraus, Et sapienter et eloquenter, 69-79; K. Usener, “Griechisches im Griechisch der LXX” in: Bons / Kraus, Et sapienter et eloquenter, 83-98. M. Harl / G. Dorival / O. Munnich, La Bible grecque des Septante : du judaïsme hellénistique au christianisme ancien, Paris 1988, 259-266; J. A. L. Lee, “Translations of the Old Testament” in: S. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period 330 B.C. – A. D. 400, Leiden 1997, 775-783; J. K. Aitken, “Rhetoric and Poetry”; J. Joosten, “Rhetorical Ornamentation.” R. W. Smith, The Art of Rhetoric in Alexandria, The Hague 1974; M. Landfester, Einführung in die Stilistik der griechischen und lateinischen Literatursprachen, Darmstadt 1977; W. Fuhrmann, Die antike Rhetorik. Eine Einführung (second edition), Düsseldorf 2008. J. Irigoin, “Psalterio nella versione dei Settanta. Alla ricercha di una poetica” in: C. Moreschini / G. Menestrina (eds.), La traduzione dei testi religiosi, Brescia 1994, 23-34; P. Chiron / C. Lévy (eds.), Les noms du style dans l’antiquité gréco-latine (Bibliothèque d’études classiques 57), Louvain 2010. De sublimitate 9.9, cf. D. A. Russel (ed.), Libellus de sublimitate Dionisio Longino fere adscriptus (Scriptorum classicorum bibliotheca oxoniensis), Oxford, 1968; G. Dorival, “La Bible des Septante chez les auteurs païens (jusqu’au Pseudo-Longin)” Cahiers de Biblia Patristica I, Paris/Strasbourg, 1987, 9-26. Cf. R. R. Walzer, Galen on Jews and Christians, London 1949, 11. 5. Classification Problems

367

23. The language of the Septuagint between Greek and Hebrew

Thus, we can speak of a specifically “Hebraic” or Septuagint writing style only in terms of modern linguistic description. 36 Whether this style was applied consciously or not by the Septuagint translators is a question that must be answered specifically for each Septuagint book. Generally speaking, style is always the result of conscious effort, but the degree of awareness involved may vary. It is certainly clear that at least some ancient texts were deliberately composed in a particular Septuagintal style (e. g. the Wisdom of Solomon and the Gospel of Luke). Only a detailed analysis can show whether the translators of a given book were moved to adopt a particular style by their understanding of the original, the authority of earlier translations, or individual taste preference. The advantage of the stylistic approach is that it does not discard the achievements of earlier research in Hebraisms, but rather attempts to integrate this knowledge into a wider framework. At the same time that it maps the general patterns of Septuagint Greek, the stylistic approach provides a firm basis from which to examine the interference of the Vorlage in any particular instance. The term “Hebraism” remains meaningful within stylistic description: it denotes elements of style responsible for the Septuagint’s peculiar coloring, as opposed to those elements that witness to the Septuagint’s assimilation to other writing strategies. A further advantage of this method is the possibility to compare Septuagint style, and its individual elements, with other styles or elements of styles in use during the Greco-Roman period. Traditional Hebraism research made such comparison extremely difficult, since the choice of the elements of Septuagint language to be studied was skewed by the Hebrew perspective. Unsurprisingly, such elements lacked counterparts in Hellenistic literature, which, with only rare exceptions, was not translated from the Hebrew. In this perspective, stylistic analysis constitutes a powerful tool for locating the Septuagint in its literary and cultural environment. Without aspiring to offer a full picture of Septuagint style and its context within Hellenistic writing, we will present below several Hellenistic parallels to illustrate this potential.

6. Outlines for a Stylistic Description We are now able to reconsider the question of classification. Features of style are normally divided into three generic groups: lexeme, lexeme super-structure, and phrase organization. A lexeme can be defined as the minimal segment of text that carries independent meaning. To characterize lexemes, stylistic description uses such binary

36. Cf. B. A. Ouspenskij, “Les problèmes sémiotiques du style à la lumière de la linguistique” in: J. Kristeva (ed.), Essais de sémiotique. Essays in Semiotics (Approaches to Semiotics 4), The Hague/Paris 1971, 447-466; M. Riffaterre, Essais de stylistique structurale, Paris 1971, M. Riffaterre, La production du texte, Paris 1979; J. Anderegg, Literaturwissenschaftliche Stiltheorie (KVR 1429), Göttingen 1977; B. Sowinski, Stilistik, Stiltheorien und Stilanalysen (Sammlung Metzler 261), Stuttgart 1991; G. Molinié, Sémiostylistique. L’effet de l’art (Formes sémiotiques), Paris 1998.

368

6. Outlines for a Stylistic Description

23. The language of the Septuagint between Greek and Hebrew

oppositions as archaism/neologism, commonplace/refined, euphonic/cacophonic, figurative/literal, opaque/clear and, last but not least, sense/nonsense. Recent studies have shown that Septuagint vocabulary contains a significant number of archaisms, which are present not only on the lexical level but also on the level of syntax. 37 The commentaries of the Church Fathers can help us to identify some Septuagint words as neologisms. Origen did not hesitate to declare the term ἐπιούσιος in the Lord’s Prayer as a neologism; he went on to say that this is not an isolated case: “who among the Greeks has ever heard the word ἐνωτίζω (“receive in the ears”) or ἀκουστίθετι (“make heard”)?” 38 Stylistic evaluation is a delicate task. Most neologisms draw on the archaic language stratum and thus bear a resemblance to archaisms. The word πρωτότοκος has been described as a biblical neologism by several scholars 39, although it occurs in Greek epigraphy 40 and in papyri 41. It is true that Hadrian, in his Isagogè, seems puzzled by the metaphoric use of the word, 42 and this may lend support to the idea of a neologism. However Hadrian could have been motivated by theological reasons, given the prominence of the word πρωτότοκος in Christological discourse. Elsewhere, a chance remark in the correspondence of Isidorus of Pelusium 43 qualifies πρωτότοκος as Homeric usage, which may indicate that the word was actually an archaism. The main concern of the lexical description of the Septuagint is the mapping of its vernacular and literary registers. Although it is well known that the Septuagint mixes words from high literature with others from the popular koine (known to us through the papyri), the exact nature of this mix, from book to book, is far from clear. The widespread figurative use of numerous words in the Septuagint was noted already in the ancient commentaries. Hadrian’s Isagogè lists many words that only made sense in an extended meaning. He observes, for example, that ποτήριον, “cup,” can mean τιμωρία, “chastisement”; that ὠτίον, “ear,” means ὑπακοή, “obedience”; and that ὁδός, “way,” signifies πρᾶξις, “behavior.” The presence of transliterations in the Septuagint text indicates that the dimension of nonsense was yet another element of its literary style. The use of neologisms, extension of the metaphoric range of words, and introduction of semantically opaque elements into the text are probably the most conspicuous traits of Septuagint writing. But these devices are not limited to the Septuagint. 44 With37. Cf. M. Casewitz, “D’Homère aux historiens romains: le grec du Pentateuque alexandrin” in: C. Dogniez / M. Harl (eds.), Le Pentateuque d’Alexandrie, Paris 2001, 77-85. 38. Cf. P. Koetschau (ed.), Origenes Werke, vol. II, Die Schrift vom Gebet (GSC), Leipzig 1899, 367.2-7. 39. Cf. C. Spicq, Notes de lexicographie néo-testamentaire, I-II, (OBO 33/1-2), Fribourg/Gottingen 1978, 771-773; W. Michaelis, “Der Beitrag der LXX zur Bedeutungsgeschichte von πρωτότοκος” in: Sprachgeschichte und Wortdeutung. Festschrift A. Debrunner, 1954, 313-320. 40. Cf. Deissmann, Licht vom Osten, 71. 41. Horsley, New Documents, vol. 3, 82; vol. 4, 163. 42. Cf. Goessling, Adrians Εἰσαγωγὴ, § 41. 43. Cf. PG 78, 749c-d. Cf. Léonas, Recherches, 157-158. 44. Cf. U. von Moellendorff, “Asianismus und Atticismus” Hermes 35 (1900), 38 (reprint in: U. von Willamowitz-Moellendorff, Kleine Schriften III: Griechische Prosa, Berlin 1969, 223-273). 6. Outlines for a Stylistic Description

369

23. The language of the Septuagint between Greek and Hebrew

out suggesting a direct dependence, we can easily compare the Septuagint’s writing strategy with other phenomena from the field of Hellenistic literature. Interesting parallels can be found in the Isis aretalogies corpus. Isis texts have been discovered in different areas of the Mediterranean—Egypt, Asia Minor, Thrace, etc. In their effort to convey a religious message originally of oriental inspiration, they resemble the Septuagint, although the end result was often widely divergent. An interesting characteristic of many of these texts is the recourse to rare and compound words, as well as convoluted periphrastic constructions. This is particularly evident in the hexametric inscription from Andros (first century BCE). In this text of 150 lines Werner Peek identified no fewer than forty cases of neologisms, 45 as well as many instances of wide-ranging metonymy 46 and periphrastic constructions bordering on obscurity. 47 Another product of Hellenistic Egypt, the Magical Papyri, freely mixes exotic divine names and abracadabra into a highly original blend, which does not lack expressiveness. 48Although some Magical Papyri betray the influence of the Septuagint, 49 it would seem plausible to suggest that their particular style developed independently and forms an interesting tangent to the Septuagint diction. 50 An inclination for opacity can be also observed at the other end of the literary spectrum. It is well known that Hellenistic poetry was saturated with allusions to Homer and other classics to the point that practically every line carried a quotation. 51 It has now been demonstrated that Hellenistic poets knowingly used variants offering the greatest difficulty to the reader. 52 Furthermore, such difficulties pertained not only to the morphological level, but to the syntactical one as well. 53 Lexeme super-structure, or “presentation,” is a highly complex notion that refers to all aspects of the immediate context of a lexeme which may have a bearing on its functioning. To describe the way lexemes produce meaning by interacting with a specific context it is necessary to take into account such binary oppositions as intensive/ weak, intellectual/emotional, positive/pejorative as well as ungrammatical/normative. 45. W. Peek, Der Isishymnus von Andros und verwandte Texte, Berlin 1930, 89. For a succinct presentation of extant sources, cf. Y. Grandjean, Une nouvelle arétalogie d’Isis à Maronée (EPRO 49), Leiden 1975, 8-11. 46. E. g. βλαστόν, “child”; τρόπις, “ship”; etc. cf. Peek, Isishymnus, 94. 47. E. g. καρποτόκος μάτερ, “the Earth”; ἀτροπιτὸς πλάγκτειρα, “the Zodiac.” Peek, Isishymnus, 95. 48. P. Levi, “The Prose Style of the Magical Papyri” in: R. A. Coles / P. J. Parsons / J. R. Rea / E. G. Turner (eds.), Proceedings of the XIV International Congress of Papyrologists, Oxford, 2431 July 1974, (Graeco-Roman Memoirs 61), London 1975, 211-216; T. Todorov, “Le discours de la magie” in: T. Todorov (ed.), Les genres du discours (Collection poétique), Paris 1978, 246282; F. Graf, “Prayer in Magic and Religious Ritual” in: C. Faraone / D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion, New York, NY/Oxford 1991, 188-213. 49. A. Léonas, “The Septuagint and the Magical Papyri: Some Preliminary Notes” BIOSCS 32 (1999), 51-64. 50. Levi, “Prose Style.” 51. G. Giangrande, “Hellenistic Poetry and Homer” Antiquité classique 39 (1970), 46-77 (reprint G. Giangrande, Scripta Minora Alexandrina (Classical and Byzantine Monographs 7), Amsterdam 1980, 33-64. 52. Giangrande, Scripta, 35. 53. Giangrande, Scripta, 35.

370

6. Outlines for a Stylistic Description

23. The language of the Septuagint between Greek and Hebrew

Lexeme super-structure covers a wide range of textual phenomena, some of which are connected to composition traits. For example, the Septuagint offers a cluster of usages based on repetition. 54 The mechanical repetition of a word is illustrated by σφόδρα σφόδρα, “very much,” (Gen 7:19; 17:6; Ex 1:7; 23:30; et passim) and πρωί πρωί, “at dawn at dawn,” meaning “very early in the morning,” but probably often understood to mean “in haste” 55 (Ex 16:21; 30:7; 36:3; Lev 6:5(B); Ps 45:6; et passim). The figura etymologica is but a case of repetition with variation. Its occurrences were perceived by ancient readers as duplication (διπλασιασμός) or tautology, 56 while its stylistic effect was described, already in Antiquity, as ἐπίτασις, “force” or “intensity.” 57 The expressive potential of repetition was further developed in the parallelistic structures of composition. The peculiarities of lexeme super-structure in the Septuagint have parallels in several domains of Hellenistic writing. For instance, the etymological play on words, although rare in the Classical texts, enjoyed somewhat greater popularity with the Hellenistic poets. Theocritus’ Idyll XVII offers a variety of such turns of phrase: αὐδῶμεν ἀοιδαῖς (XVII, 2), ὑμνήσαιμ’ ὕμνοι (XVII, 8), φιλέων […] φιλεούσης (XVII, 42). 58 Similar expressions can be found in Callimachus: τέκνα τεκέσθαι (Hymn to Delos, 111). 59 The expressive potential of paronomasia has also been explored in texts related to the sphere of the sacred. Plutarch quotes an oracle given at Delphi that instructed Lycurgus to set up sanctuaries, φυλὰς φυλάξαντα καὶ ὠβὰς ὠβάξαντα (“distributing the tribes and parceling the land-parcels”). Paronomastic expressions also occur in orphic texts 60 and in the Isis aretalogies. Phrase organization essentially describes the phenomena of inversion and disjunction. In the case of the Septuagint it is possible to speak of asyndetic composition, of hyperbaton and hyperthesis, and of manifestations of parallelism, rhythm, and even rhyme. It is interesting to note that biblical parallelism can appear as an element of style in its own right, even in Greek dress. Hadrian’s description of the Septuagint idiom did not fail to acknowledge parallelism, although, for Hadrian, its manifestations fell under several entries, such as ταυτολογία, homoiotheleuton, and antistrophe. 61 The influence of Greek composition techniques can be seen in the fact that the Septuagint M. Frédéric, La répétition. Étude linguistique et rhétorique, Tübingen, 1985. This reading is suggested by Hadrian’s Isagogè § 87, cf. Léonas, Recherches, 190-194. Cf. Diodorus of Tarsus, In Ps. 39:2, Hadrian, Isagogè § 104. Cf. Léonas, Recherches, 190-194. M. A. Rossi, Theocritus’ Idyll XVII, A Stylistic Commentary (Classical and Byzantine Monographs 25), Amsterdam 1989, 15, 26, 81. 59. Cf. F. Williams, Callimachus, Hymn to Apollo, A Commentary, Oxford 1978; see the Index s. v. “jeu étymologique.” 60. Cf. ἐκ καθαρῶν καθαρά or παθῶν τὸ πάθημα in: A. Olivieri, Lamellae aureae Orphicae, 1915, 11, 16. 61. Cf. Goessling, Adrians Εἰσαγωγὴ, § 100, 102-103; For an overview of the early treatment of Septuagint poetic element, see A. Hilhorst, “Poésie hébraïque et métrique grecque. Les témoignages des Anciens, de Philon d’Alexandrie à Boniface de Mayence” in: D. Accorinti / P. Chuvin (eds.), Des géants à Dionysos. Mélanges Francis Vian, Alexandria 2003, 305-329. 54. 55. 56. 57. 58.

6. Outlines for a Stylistic Description

371

23. The language of the Septuagint between Greek and Hebrew

translators sometimes introduced antistrophe independently from the Hebrew Vorlage. 62 On the level of composition, a vast field of study awaits researchers wishing to compare Septuagint literary devices with Asianism, a literary current of the Second Sophistic. 63 Asianic prose, marked by such embellishments as chiasmus, opposition, parallelism, and variations of short and long periods, has many features in common with Septuagint style. 64 Its rhythmic pattern was often reinforced by a recurrent metric organization, especially at the end of each period. 65 The Isis ἐγκώμιον from Maronea (circa second and first centuries BCE) offers an example of this type of writing. 66 Its prose is highly rhetorical and strongly marked by Asianic literary techniques. 67 Its rhythmic structure is often accentuated by the introduction of a metric pattern within and at the end of each period not dissimilar to that found in the Septuagint. 68 The parallelism used in the Greek context differs from that of the Septuagint, but the techniques are similar, involving, as they do, semantic apposition as well as metric and phonetic assimilation: 11. εἰ γὰρ ὑπὲρ τῆς ἐμῆς καλουμένη σωτηρίας ἦλθες, πῶς ὑπὲρ τῆς ἰδίας τιμῆς οὐκ ἂν ἔλθοις; If you came when invoked for the sake of my deliverance, how could you not come for the sake of your own praise? 14-15. καὶ πρῶτον ἐπὶ τὸ γένος ἥξω, τῶν ἐγκωμίων ποιησάμενος ἀρχὴν τὴν πρώτην σου τοῦ γένους ἀρχὴν I will first address the matter of your origin, making the beginning of your race the beginning of my praise.

62. Cf. J. A. L. Lee, “Translations,” 779. 63. Cf. U. von Willamowitz-Moellendorff, “Asianismus und Atticismus” Hermes 35 (1900), 152 (reprint in Willamowitz-Moellendorff, Kleine Schriften III, 223-273); E. Norden, Die antike Kunstprosa vom IV. Jahrhundert v. Chr. bis in die Zeit der Renaissance, I-II, Leipzig/ Berlin 1918; A. Boulanger, Aelius Aristide et la sophistique dans la province d’Asie au IIe s. de notre ère (Bibliothèque des Écoles françaises d’Athènes et de Rome 126, second edition), Paris 1968; J. Goeken, Aelius Aristide et la rhétorique de l’hymne en prose (Recherches sur les rhétoriques religieuses, 15), Turnhout, 2012. 64. Cf. Norden, Die antike Kunstprosa; R. W. Smith, The Art of Rhetoric in Alexandria, The Hague 1974; J. W. Welch (ed.), Chiasmus in Antiquity. Structures, Analyses, Exegesis, Hildesheim 1981. 65. Cf. J. V. Bainvel, “La prose métrique et la prose rythmique” Études religieuses 59 (1893), 143155; A. W. De Groot, A Handbook of Antique Prose-Rhythm, vol. I, History of Greek Prose Metre; Demosthenes, Plato, Philo, Plutarch and Others, Groningen/The Hague, 1919; A. W. DE Groot, La prose métrique des Anciens (Collection d’Études Latines, 2a), Paris 1926; J. Irigoin, “Le Psaume 26 dans la Septante. Étude de composition rythmique” in: Dorival and Munnich, Selon les Septante, 288-297; F. Siegert, “Hebräischer Vers und Griechischer Prosarhythmus in der Septuaginta” in: M. Karrer / W. Kraus (eds.), Septuaginta Deutsch. Erläuterungen und Kommentare zum griechischen Alten Testament, vol. 1, Stuttgart 2011, 53-64. 66. Cf. Grandjean, Maronée. 67. Grandjean, Maronée, 108. 68. Grandjean, Maronée, 108, 115-117.

372

6. Outlines for a Stylistic Description

23. The language of the Septuagint between Greek and Hebrew

While the parallelism of the poetic books of the Greek Bible can be likened to the Asianic rhetorical devices, biblical prose, with its particular rhythmic patterns and markers (ἰδού, καί, etc.), can be compared with the καί style of the attic narrative rooted in Classical usage. 69

7. Conclusions and Perspectives A systematic stylistic description of the Septuagint remains to be accomplished. Without a comprehensive framework, the assessment of the role of literal translation technique in the genesis the language of the Septuagint cannot be said to be complete. In other words, we cannot yet measure the impact of the Hebrew Vorlage on the language use of the Greek translators. The fixation on the Hebraisms that characterized much of pre-modern and modern research delayed to a certain extent the task of synthesis, although this research did collate and analyze a considerable amount of data. An important facet of the Hebraism problem is the historical development of Biblical Greek language and style. It has been noted already by Thackeray 70 that the evolution of the Septuagint idiom is rather paradoxical: one would expect that the development of the translation technique would have produced greater clarity, yet this was not the case. Although some later books manifest greater freedom of translation (e. g. Proverbs), the general tendency was towards increased literalism. Linguistic and stylistic description must take this evolution into account. An important milestone in this development was the work of Aquila, which combined an indepth knowledge of the Greek language with painstaking attention to the Hebrew. Owing to the discoveries in Nahal-Hever we can now include his literary activity in the wider context of translation and revision in Palestine and in the diaspora. 71 Present already in several Septuagint Books, 72 elements of this writing style emerge also in the New Testament. The proliferation of this writing technique, particularly its culmination in the work of Aquila, is an important testimony to its linguistic vigor. Other translations of religious texts from the Hellenistic world, such as the Isis aretalogies, fragments of Berossus and of Philo of Byblos, and even later texts such as the Egyptian Mysteries of Iamblichus or the Chaldean Oracles, can offer interesting parallels to Septuagint diction. The Isis aretalogies, due to the variety of the styles they represent, offer interesting examples of the encounter between the parallelistic style inspired by Egyptian models and the typically Greek literary craft. The Isis texts illustrate alternative attempts to bridge the gap between Greek and oriental forms of expression. The solutions adopted by the Septuagint translators when faced with essen69. Cf. S. Trenkner, Le style καί dans le récit attique oral (Cahiers de l’Institut d’études polonaises en Belgique I), Bruxelles 1948; A. Christol, “Édits grecs d’Ashoka et Koinè” in: C. Brixhe (ed.), La koinè grecque antique, vol. III, Les contacts (Collection études anciennes), Nancy 1998. 70. Cf. H. St. J. Thackeray, A Grammar of the Old Testament in Greek, according to the Septuagint, vol. I, Introduction, Orthography and Accidence, Cambridge 1909, 13, 25-55. 71. Cf. D. Barthélemy, Les devanciers d’Aquila (VT Supplements 10), Leiden 1963. 72. Cf. J. Reider, Prolegomena to a Greek-Hebrew and Hebrew-Greek Index to Aquila, Philadelphia, PA 1916, 18-19; H.-J. Stipp, “Bemerkungen zum Griechischen Michabuch” JNSL 29 (2003), 115. 7. Conclusions and Perspectives

373

23. The language of the Septuagint between Greek and Hebrew

tially the same challenges now appear as part of a wider context of literary development. Far from being an isolated phenomenon, translation was widely practiced in Hellenistic Egypt. The Isis aretalogies demonstrate that the idea of transmitting a foreign religious message in Greek was not alien to the Hellenistic mind. In addition to the Isiac corpus, we know of several other translations made from the Egyptian into the Greek (e. g. P. Oxy 1380-1381). These texts and the Septuagint probably represent parts of the same literary current. In the Egyptian world, this current produced the corpus that we now call the Hermeticum and such by-products as the Magical Papyri. From a strictly textual point of view the Septuagint differs from this literary production by its attitude to its source-text. The innovation of the Septuagint translators lies in their rigorous approach to the original. This rigor presents an interesting parallel with the treatment accorded to Homer and other classics in the Greek philology that emerged at the same epoch. 73 While traditional approaches to translation made no distinction between translation and paraphrase (or for that matter between paraphrase and complete rewriting of the parent text), the Septuagint translators and the Greek philologists adopted a common innovative attitude of extreme textual precision. In the course of time, one of the consequences of the Septuagint translators’ rigorous approach was the advent of the practice of translation in its modern form. From the point of view of Greek literary history, one can suggest further analogies that may help us to understand the phenomenon of the Septuagint and to break the spell of its literary isolation. It has often been observed that Hellenistic poetry flees real life in search of the archaic simplicity of a mythic past (Apollonius Rhodius, Callimachus) or of a static idylls (Theocritus). Interestingly, the Septuagint—taken as a literary work—is for the most part also oriented towards a past or sacred reality, describing the world of the Patriarchs, the prophets, and the kings of long ago. Another trend in the Hellenistic poetry experimented with untraditional poetic material presented in the form of didactic poems (Aratus, Nicander). Again, an educational and descriptive dimension is present throughout the Septuagint books, not just in the Wisdom texts. The interest of the Hellenistic literary audience in cryptic writing is well attested (Lycophron); the Septuagint books, in turn, supply degrees of obscurity to suit any taste. In addition to these points of comparisons, the translation itself can be analyzed as a writing technique with a literary context of its own. Hellenistic poetry is known to be an arte allusiva, an art built on a network of echoes of Homer and other classical poets. Its dependence on an earlier corpus suggests a parallel to the procedure of generating a text out of another, earlier one, although in the case of the Septuagint the parent text was written in a different language and this process came to be known as translation.

73. Cf. the interesting comparison along these lines suggested by D. Weissert, “Alexandrian Analogical Word-Analysis and Septuagint Translation Techniques” Textus 8 (1973), 31-44.

374

7. Conclusions and Perspectives

24. Stylistic Features of the Septuagint Jennifer Dines

The majority of translations have quantitative equivalence as their default mode, and follow Hebrew syntax and word order closely, often to the detriment of normal koine Greek. But they also reproduce at least some of the literary effects and “music” of the Hebrew, showing occasional sensitivity to the aesthetics as well as the meaning of the text. Translators even create literary effects independent of the Hebrew, apparently for stylistic reasons. Although these effects are spasmodic, they are frequent enough to merit—indeed to require—attention and explanation. I will return to this at the end. In what follows, I will examine some of the more obvious stylistic “tricks of the trade” employed by various translators that either imitate the source-text or introduce new effects. As it is impossible to cover all aspects of style, or to examine the stylistic profile of any book systematically, 1 I will take representative examples, mainly from the Pentateuch and Prophets, under three headings: Verbal Effects, Variation and Repetition, and Arrangements.

1. Verbal Effects 1.1 Assonance and Alliteration Repetition of vowels (assonance) or consonants (alliteration) is one of the easiest effects to spot and is very widespread in both the Masoretic text and the Septuagint. In the Septuagint, existing effects are sometimes imitated, and new effects are sometimes created.

1.1.1 Imitating Hebrew Occasionally a translator matches sound for sound, as in Lev 21:17 (μῶμος for ‫ ;מום‬cf. Deut 15:21) or Am 1:14 (σεισθήσεται […] συντελείας for ‫)סופה ]…[ סער‬. More often,

1.

See however: D. R. Gera, “Translating Hebrew Poetry into Greek Poetry: The Case of Exodus 15” BIOSCS 40 (2007), 1-14; E. Bons, “Rhetorical Devices in the Septuagint Psalter” in: E. Bons / T. Kraus (eds.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint (FRLANT 241), Göttingen 2011, 69-79; J. K. Aitken, “Rhetoric and Poetry in Greek Ecclesiastes” BIOSCS 38 (2005), 55-78; J. K. Aitken, “The Literary Attainment of the Translator of Greek Sirach” in: J.-S. Rey / J. Joosten (eds.), The Texts and Versions of the Book of Ben Sira: Transmission and Interpretation (JSJSup 150), Leiden 2011, 95-126. 1. Verbal Effects

375

24. Stylistic Features of the Septuagint

different vowels or consonants are used, as in Gen 1:21 (πᾶν πετεινὸν πτερωτὸν for ‫ )כל־עוף כנף‬2 or Hos 12:2 (πονηρὸν πνεῦμα for ‫)רעה רוח‬. 3 Gen 1:2 is particularly striking. 4 Faced with the alliterative rhyming pair ‫תהו ובהו‬, the translator finds two matching adjectives with alpha-privative (the second of which is rare in Greek, and a hapax in the Septuagint): ἀόρατος καὶ ἀκατασκεύαστος, (“unseen and unsorted”), “neither feature inevitable.” 5 Sometimes alliteration is extensive, as in Gen 49:19: Γαδ πειρατήριον πειρατεύσει […] πειρατεύσει αὐτῶν κατὰ πόδας representing ‫גד גדוד יגודנו ]…[ יגד‬: four words with pi instead of four with gimel (though not all matching). In Am 2:13 κυλίω ὑποκάτω […] κυλίεται […] καλάμης represents ‫]…[ עמיר‬ ‫מעיק ]…[ תעיק העגלה‬: four words with kappa instead of four with ayin (again not all matching, and not all initial letters).

1.1.2 New Effects The frequent occurrence in the Septuagint of assonance/alliteration where there is none in Hebrew, and where other choices could have been made, indicates that the translators’ stylistic instincts included the Greek texts they were producing as well as the Hebrew texts they were translating: 1.1.2.1 Gen 2:10 A sustained effect occurs here: ποταμὸς δὲ ἐκπορεύεται […] ποτίζειν τὸν παράδεισον for ‫ונהר יצא ]…[ להשׁקות את־הגן‬. Some of the words with pi are obvious equivalents (πόταμος, ποτίζειν). But this is the first rendering of ‫ יצא‬by ἐκπορεύομαι, not an inevitable choice, although it becomes standard in the Pentateuch. In the Septuagint as a whole the commonest equivalent is ἐξέρχομαι, which could have been used in Gen 2:10, since the translator had as yet no pattern with which to conform. And παράδεισος, although apt, and already used in 2:8 (without alliteration), was not the only possibility for ‫גן‬. Κῆπος could have been selected: later, it renders ‫ גן‬in synonymous parallelism with παράδεισος (Isa 1:29,30; Jer 36/29:5,28; cf. Sir 24:30,31).

2. 3. 4. 5.

Although this alliteration in pi is partly the result of obvious equivalents, a different word could have been chosen instead of πετεινὸν, e. g. ὄρνεον (Gen 6:20). The Septuagint vocalises differently from the Masoretic text. Instead of πονηρόν, the translator (credibly the same throughout the Minor Prophets) could have used κακόν (Am 9:4,10). Cf. J. M. Dines, “Imaging Creation: The Septuagint Translation of Genesis 1:2” HeyJ 36 (1995), 439-450; J. M. Dines, The Septuagint, London/New York, NY 2004, 57. J. K. Aitken, “The Significance of Rhetoric in the Greek Pentateuch” in: J. K. Aitken / K. J. Dell / B. A. Mastin (eds.), On Stone and Scroll: Essays in Honour of Graham Ivor Davies (BZAW 420), Berlin 2011, 507. On the popularity of alpha-privative in Hellenistic writers, see J. R. Reese, Hellenistic Influence on the Book of Wisdom and its Consequences (AnBib 41), Rome 1970, 30. Other striking examples occur in Jer 2:6 (ἀπείρῳ […] ἀβάτῳ […] ἀνύδρῳ […] ἀκάρπῳ); Sir 20:19 (ἄχαρις […] ἄκαιρος […] ἀπαιδεύτων; Aitken, “Literary Attainment,” 110); Ps 48/49:11 (ἄφρων καὶ ἄνους, together with more assonance in alpha; Bons, “Rhetorical Devices,” 72-73, with further examples).

376

1. Verbal Effects

24. Stylistic Features of the Septuagint

1.1.2.2 Lev 11:35 The translator develops Masoretic text’s ‫ תנור וכירים יתץ‬with κλίβανοι καὶ κυθρόποδες καθαιρεθήσονται. Κλίβανος already renders ‫ תנור‬in Gen 15:17, while καθαιρέω renders ‫ נתץ‬in Exod 34:13. But κυθρόπους (“pot”) is a rare, archaic, literary word found, in its Ionic spelling χυτρόπους, in Hesiod. 6 The translator could perhaps have chosen λουτήρ, already used for ‫ כיר‬as “wash-basin” (8:11; cf. Exod 30:18 etc.); this would also have fitted the context of 11:35, since ‫ כיר‬can be used for washing as well as cooking. The learned choice underlines the translator’s ability to enhance an effect already accidentally produced. Several of the previous examples involve pi or kappa. This seems to reflect fashions among Greek authors, which were mimicked by documentary writers as well as Septuagint translators. 7 But there is much variety, for instance: 1.1.2.3 Lev 26:19 τὴν ὕβριν τῆς ὑπερηφανίας ὑμῶν (“the insolence of your arrogance,” NETS) for ‫את־‬ ‫“( גאון עזכם‬the pride of your power” RV 8). ‫ גאון‬is often rendered by both ὕβρις and ὑπερηφανία (although only here in the Pentateuch), but ‫ עז‬is nowhere else rendered by either word. 9 The assonance has apparently been created as a deliberate effect. 10

1.2 End-Rhyme (Homoeoteleuton) Inflected languages often require identical endings for recurring grammatical forms, so care must be exercised before these are counted as stylistically significant. Sometimes, however, a special effect is clearly being obtained.

1.2.1 Gen 4:12,14 The alliteration of ‫“( נע ונד‬wavering and wandering”), is replaced by the assonance of στένων καὶ τρέμων (“wailing and quailing,” literally, “groaning and trembling”), where both first and second syllables “rhyme.” Στένω is a poetic verb, found in Homer and the tragedians, τρέμω largely so. Both Hebrew verbs express movement rather than emotion (although ‫ נוד‬sometimes expresses grief, e. g. Nah 3:7; Isa 51:19; Jer 22:10) and other Greek words could have been chosen; neither verb occurs again in Genesis. 11

6. Works and Days, 748. 7. See Aitken, “Significance,” 511-512, with n. 33 (pi); Gera, “Translating Hebrew Poetry,” 10 (κατα- as a preferred prefix). Other examples (among many) of alliteration in pi include: Deut 1:30, 33 (Aitken, “Significance,” 512); Isa 3:5; 27:3; Jer; 3:3; cf. J. A. L. Lee, “Translations of the Old Testament. I Greek” in: S. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period 330 B.C. – A.D. 400, Leiden 1997, 778. Examples with kappa include: Ps 50/51:12 (Bons, “Rhetorical Devices,” 73-74); Am 3:14; Isa 27:9; Jer 4:30. 8. Note the English alliteration! NETS uses end-rhyme. 9. The most frequent equivalents are δύναμις, ἰσχύς and κράτος. 10. See also Isa 43:4 (epsilon); Sir 14:18 (a striking example involving delta, phi, and double lambda). 11. For possible Stoic associations in the Greek pair, see B d’A 1, 115. 1. Verbal Effects

377

24. Stylistic Features of the Septuagint

1.2.2 Am 6:11 θλάσμασιν […] ῥάγμασιν (“with dents […] with rents”) for ‫“( רסיסים ]…[ בקעים‬with hits … with splits”) achieves a clever matching of vowels and endings for two unusual Hebrew nouns evoking architectural destruction. The translator’s own rhyming choice involves medical terms, perhaps suggesting damage to occupants rather than buildings. The specialized knowledge and ingenuity are impressive. 12 1.2.3 Deut 32:15 End-rhyme proper occurs here: ἐλιπάνθη ἐπαχύνθη ἐπλατύνθη (MT ‫שׁמנת עבית‬ ‫)כשׂית‬, with added assonance and alliteration provided by augment, internal alpha, and repetition of pi; the penultimate syllables of the second and third verbs also rhyme. 13 1.2.4 Ezek 38:23

καὶ μεγαλυνθήσομαι καὶ ἁγιασθήσομαι καὶ ἐνδοξασθήσομαι καὶ γνωσθήσομαι

‫והתגדלתי והתקדשׁתי ונודעתי‬

This run of four “rhyming” verbs constitutes the finale of the first prophecy against Gog and so perhaps seemed to require grandiose language. The Masoretic text has only three verbs, two of them with the same prefix, and all with the same personal ending. With the addition of ἐνδοξασθήσομαι, the Septuagint has created a run of four verbs, an effect found elsewhere. 14

1.2.5 Am 5:2 The synonymous pair ἔπεσεν […] ἔσφαλεν corresponds to ‫ נפלה ]…[ נטשׁה‬in syllable count, assonance or alliteration in the first syllable, and matching endings. Πίπτω is an exact (and standard) equivalent for ‫נפל‬, but σφάλλω (apparently reflexive) for ‫נטשׁ‬ occurs only here. 15 Ἔσφαλεν makes an aurally striking partner to ἔπεσεν but loses the nuance of abandonment implied by ‫נטשׁ‬. The translator could have used another verb, ἐγκατελείφθη, for instance, as in Mal 2:11 (MT ‫)בגדה‬, but sound and rhythm have taken precedence.

12. ‫ רסיסים‬is a hapax, supposedly from II ‫רסס‬, “crush” (coined to reinforce ‫בקיעים‬, “fissures”?). Θλάσμα is also a septuagintal hapax, but occurs in Greek medical texts for a bruise, or a depression in the skin (LSJ): hence “dents.” Ρῆγμα (or ράγμα) is a medical term for a fracture or lesion (“rents”), though it also occurs in other contexts. Another contrived rhyming pair occurs in Am 5:16 (ῥηθήσεται […] κληθήσεται for ‫)יאמרו ]…[ וקראו‬. 13. In fact, the sequence is anticipated with ἐνεπλήσθῃ (‫)ויבעט‬, extending the effect beyond what already exists in Hebrew. For other poetic features in Deut 32, see Aitken, “Significance,” 513514. 14. E. g. Hab 3:2; Job 24:18-20, cf. Lee, “Translations,” 779. 15. Σφάλλω is normally transitive. There is no default rendering; most frequent is ἀπωθέω (eight times).

378

1. Verbal Effects

24. Stylistic Features of the Septuagint

2. Variation and Repetition Variation is a common—almost instinctive—device whereby synonyms are used to avoid monotony. 16 In the Septuagint this means choosing different Greek terms for the same Hebrew word. It shows up when the Septuagint and the Masoretic text are compared, but the effect would not have been lost on readers or hearers of the Septuagint alone, since anyone with some education would be attuned to such ringing of changes for the sake of euphony. Not all instances are for stylistic reasons alone: sometimes the translator has an exegetical point to make. 17 Often, however, the motivation seems to be “for purely stylistic reasons.” 18 The converse effect—repeating a significant word, or grammatical form, for emphasis—also figures in Greek rhetoric and is frequent in the Septuagint.

2.1 Variation Representative examples include Gen 6:17; 7:21 (τελευτήσει and ἀπέθανεν for ‫;)גוע‬ Exod 6:14 (συγγένεια), 6:15,19 (πατρία) and 6:24,25 (γένεσις), for ‫ ;משפחה‬19 Hos 11:10 (ἐρεύξεται and ὠρύσεται for ‫ ;)שאג‬Am 7:8; 8:2 (ὁράω and βλέπω for ‫ ;ראה‬also Hag 2:3; Zech 5:2); 20 Jer 49/42:15,17 (δῶτε and θέντες for forms of ‫)אשים‬. 21

2.2 Repetition 2.2.1 Isa 1:18 The translator remodels the Masoretic text’s variation to create an emphatic final repetition: ὡς χιόνα λευκανῶ (‫ ]…[ )כשׁלג ילבינו‬ὡς ἔριον λευκανῶ (‫)כצמר יהיו‬. Rather than a sign of semantic leveling, use of this classic device (epiphora) suggests rhetorical skill. 2.2.2 Isa 24:7 This time, repetition begins each clause (anaphora): πενθήσει οἶνος (‫ )אבל תירושׁ‬πενθήσει ἄμπελος (‫)אמללה־גפן‬. 22

2.3 Cognate Phrases This form of repetition, also featured in Greek rhetoric, is very frequent in the Septuagint. It mirrors the Hebrew in e. g. Gen 1:11 (σπεῖρον σπέρμα for ‫ ;)מזריע זרע‬Gen 3:20 (ζωή […] ζώντων for ‫ ;)חוה ]…[ חי‬Gen 32:31 (εἶδος […] εἶδον, an original way of 16. The device occurs in Hebrew, e. g. in synonymous parallelism in Psalms and Proverbs. 17. For examples in the Minor Prophets, see J. M. Dines, “Stylistic Invention and Rhetorical Purpose in the Book of the Twelve” in: Bons and Kraus, Et sapienter et eloquenter, 24-25. 18. NETS, 543. 19. Cf. NETS, 43. 20. For these and other examples in the Minor Prophets, see Dines, “Stylistic Invention,” 25-31. 21. Cf. NETS, 878. 22. Another example of anaphora occurs in Isa 24:8,11 (a fourfold πέπαυται). For the important part played by repetition in Exodus 15, see Gera, “Translating Hebrew Poetry,” 10 and 11. 2. Variation and Repetition

379

24. Stylistic Features of the Septuagint

representing the wordplay of ‫)פניאל ]…[ פנים‬. There are no Hebrew precedents in e. g. Gen 2:4 (γενέσεως […] ἐγένετο for ‫ ;)תולדות ]…[ בהבראם‬Exod 15:4-5 (κατεπόντισεν […] πόντῳ for ‫ ;טבעו ]…[ תהמת‬this repetition has been carefully crafted: the noun occurs only here in the Septuagint); Joel 3:2 (δούλους […] δούλας effectively renders ‫ ;העבדים ]…[ השׁפחות‬Ezek 38:15-16 (ἀναβάται […] ἀναβήσῃ for ‫)רכבי ]…[ ועלית‬. 23

2.4 Compound Verbs Prefixes of Greek compound verbs often contribute to both variation and repetition. As has been observed, Hebrew does not have an equivalent form, so that compound verbs, where not essential to the meaning, often betray a translator’s inventiveness. 24 Examples include: Deut 32:21 (four verbs beginning with παρα-); Hos 4:14 (three verbs beginning with συν-); Am 9:11 (four verbs beginning with ἀνα-).

2.5 Variation and Repetition Combined 2.5.1 Num 27:17 ὅστις ἐξελεύσεται (‫ )יצא‬πρὸ προσώπου αὐτῶν καὶ ὅστις εἰσελεύσεται (‫ )יבא‬πρὸ προσώπου αὐτῶν καὶ ὅστις ἐξάξει αὐτοὺς (‫)יוציאם‬ καὶ ὅστις εἰσάξει αὐτούς (‫)יביאם‬

While maintaining exact equivalence for the opening and closing elements each time, the translator has deftly reworked the Hebrew alternation between ‫ יצא‬and ‫ בוא‬by using ἔρχομαι for the first pair and ἄγω for the second pair and by distinguishing between the antithetical meanings (go/bring out; come/bring in) by using the prefixes ἐκ- and εἰσ-.

2.5.2 Am 9:1 οὐ μὴ διαφύγῃ (‫ )לא־ינוס‬ἐξ αὐτῶν φεύγων (‫)נס‬ καὶ οὐ μὴ διασωθῇ (‫ )ולא־ימלט‬ἐξ αὐτῶν ἀνασῳζόμενος (‫)פליט‬

The translator copies the Hebrew cognate expressions (‫פליט‬/‫נס; ימלט‬/‫ )ינוס‬with verb and matching participle. He knits the two clauses even more tightly by prefixing the verbs with δια- (repetition), but introduces variation in the participles: the first has no prefix, the second the prefix ἀνα-, while the final word breaks the monotony of the repeated two-syllable words with one of no less than six syllables. The compounds with δια- are stylistically motivated, since the normal equivalent for ‫ נוס‬is the simple verb φεύγω; διαφεύγω occurs only here. The simple verb σῳζω is more frequent for the niph’al of ‫ מלט‬than διασῴζω (which occurs only here in the Minor Prophets),

23. For striking examples in LXX Psalms, see Bons, “Rhetorical Devices,” 74-79. 24. Aitken, “Literary Attainment,” 119. The correlation of Greek compound verbs with Hebrew derived stems is not constant enough to provide an overarching raison d’être, as has sometimes been proposed; see T. Muraoka in B d’A 23.1, viii.

380

2. Variation and Repetition

24. Stylistic Features of the Septuagint

while the participle ἀνασῳζόμενος occurs ten times in the Septuagint for ‫פליט‬, but only here and Obad 14 in the Minor Prophets. 25

3. Arrangements Sometimes translators extend effects in more developed ways.

3.1 Chiasm Although chiasm, especially in the ABB’A’ form, is a recognized feature of Greek composition, it occurs more widely, and in more complex forms, in Hebrew literature. 26 That simple chiastic arrangements are reproduced in the Septuagint is not surprising, given most translators’ habit of keeping closely to Hebrew word order. An example occurs in Am 5:7: (κύριος) ὁ ποιῶν εἰς ὕψος κρίμα καὶ δικαιοσύνην εἰς γῆν ἔθηκεν

‫ההפכים ללענה משׁפט וצדקה לארץ הניחו‬

The Septuagint construes the consonantal text differently from the Masoretic text, but maintains the word order. The resulting Greek is awkward, but the keywords “judgment” and “justice” remain juxtaposed at the center of the verse. Similar examples occur in Am 6:8; 9:11. In Am 7:12, however, the Masoretic text’s ABCB’A’ arrangement (‫ )ואכל־שׁם לחם ושׁם תנבא‬is not maintained. Instead, the translator prefers anaphora: καὶ ἐκεῖ καταβίου καὶ ἐκεῖ προφητεύσεις. Unusually, he has overridden the Hebrew word order, which could have been maintained without detriment to acceptable Greek, by rendering: καταβίου ἐκεῖ καὶ ἐκεῖ προφητεύσεις. It is easy to see why he did not: it would have sounded horrible. Euphony has taken precedence. New chiasms are sometimes achieved:

3.1.1 Num 21:1,23,26 An ABB’A’ alternation occurs with ἐπολέμησεν (‫ ]…[ )וילחם‬παρατάξασθαι (‫)לקראת‬ […] παρετάξατο (‫ ]…[ )וילחם‬ἐπολέμησεν (‫)נלחם‬. 3.1.2 Jon 2:4 ἀπέρριψάς / με / εἰς βάθη καρδίας θαλάσσης // καὶ ποταμοί / με / ἐκύκλωσαν ‫ יסבבני‬/ ‫ ונהר‬/ ‫ מצולה בלבב ימים‬/ ‫ותשׁליכני‬

The reversed word order that places the second με before, not after, its verb is a stylistic touch as well as providing a needed object-pronoun; it creates a clearer ABCC’B’A’ pattern than the Masoretic text. 27 25. For other examples of variation and repetition combined in the Minor Prophets, see Dines, “Stylistic Invention,” 33; for examples in Sir 4:14; 9:10, Aitken, “Literary Attainment,” 118. 26. For Hebrew examples, see J. R. Lundbom, The Hebrew Prophets: An Introduction, Minneapolis, MN 2010, 175-178. 27. Cf. B d’A 23.4-9, 124. For examples in Nahum, B d’A 23.4-9, 176. 3. Arrangements

381

24. Stylistic Features of the Septuagint

3.1.3 Zeph 2:3 ζητήσατε τὸν κύριον […] καὶ δικαιοσύνην ζητήσατε (chiasm) ‫( בקשׁו את־יהוה ]…[ בקשׁו־צדק‬anaphora)

The Septuagint vocalizes and construes the Hebrew of this complex verse rather differently from the Masoretic text (which may not represent the translator’s exact Vorlage), but the two clauses clearly form an interrupted ABB’A’ chiasm, marked as deliberate by the reversed word order (δικαιοσύνην ζητήσατε / ‫)בקשׁו־צדק‬. 28

3.1.4 Am 1:3–2:6 The most elaborate chiastic arrangements occur here. I have discussed this remarkable passage in detail elsewhere, so will draw attention only to the essentials. 29 Two patterns emerge: an ABCB’A’ pattern with multiple strands in 1:3-15, and a simpler ABA’ pattern in 2:1-6. The motive for such stylistic fine-tuning of a passage already highly organized in Hebrew is probably to render the formulaic repetitions less monotonous to Greek-attuned ears. 30 At the same time, the circular patterns in 1:1-15, revolving round the single feminine suffix and the single aorist in 1:9, perhaps created a kind of Hebraic poetic music congenial to Jewish readers. The patterning seems to serve no exegetical purpose.

3.2 Triads Frequently, a small ABA’ chiasm occurs, a form of repetition with variation. It may come within the same verse, nearby verses, or—more subtly—stretched over several chapters. It seems to be particularly popular in the prophetic books, but other examples probably remain to be added:

3.2.1 Jon 1:9,13; 2:11 In the Masoretic text, ‫“( יבשׁה‬dry land”) occurs each time, but in the Septuagint there is a sequence ξηράν […] γῆν […] ξηράν. There seems no motive other than stylistic for the alternation. 3.2.2 Zech 10:7 With χαρήσεται […] εὐφρανθήσονται […] χαρεῖται 31 for ‫ושׂמח ]…[ ושׂמחו יגל‬, the Septuagint creates a small pattern where the Masoretic text has variation. 3.2.3 Isa 4:2-3 Like Zech 10:7, this triad replaces Hebrew variation with an elegant repetition of one verb with different prefixes: τὸ καταλειφθὲν (‫ ]…[ )לפליטת‬τὸ ὑπολειφθὲν (‫)הנשׁאר‬ […] τὸ καταλειφθὲν (‫)והנותר‬. Similar examples occur in Jer 31/48:26,27,39; Ezek 39:17. 28. For the Hebrew text, see A. Gelston, The Twelve Minor Prophets (BHQ 13), Stuttgart 2010, 106; B d’A 23.4-9, 350. 29. Dines, “Stylistic Invention,“34-37. 30. For a similar suggestion for MT Jer 51:20-23, see Lundbom, Hebrew Prophets, 169. 31. Some witnesses have χαρεῖται.

382

3. Arrangements

24. Stylistic Features of the Septuagint

3.2.4 Am 1:8; 5:15; 9:12 κατάλοιποι […] περιλοίπους […] κατάλοιποι all render ‫ שׁארית‬with no obvious change in meaning. Occurring at beginning, middle, and end, this triad perhaps helps structure the book. 32

3.3 Alternations Another pattern, already discernible in the Pentateuch, consists in alternation between words or phrases, in quatrains (an extension of the variation-repetition effect) or longer sequences spread through, and even (for the Minor Prophets) across, entire books. 33

3.3.1 Gen 36:13-31 There is rudimentary variation between καὶ οὗτοι and οὗτοι δὲ that does not correlate with differences between ‫ אלה‬and ‫ואלה‬. 3.3.2 Num 14:6,30,38 Num 14:6 contains two personal names introduced by ὁ τοῦ, “son of,” while 14:38 has two names introduced by υἱὸς. Between these pairs, 14:30 has two names, the first introduced by υἱὸς, the second by ὁ τοῦ. The Masoretic text has ‫ בן‬throughout. The result is a curious sequence: AA/BA/BB. An earlier, rougher alternation between the two expressions in Num 10:19-27 suggests that the arrangement is deliberate. 34 3.3.3 Num 9:17,18,22; 10:12 Here, ἵστημι and σκιάζω alternate as renderings of ‫ שׁכן‬for the “settling” and “resting” of the Cloud: ἔστη (‫ ]…[ )ישׁכן־שׁם‬σκιάζει (‫ ]…[ )ישׁכן‬σκιαζούσης (‫]…[ )לשׁכן‬ ἔστη (‫)ישׁכן‬. Theological considerations may have played a part, 35 but artistic concerns are also evident in the grammatical repetition-with-variation between infinitive and participle (9:18,22) and in the chiastic shape. 3.3.4 Am 2:14-15; 9:1 A triadic arrangement (2:14-15), οὐ μὴ σώσῃ […] οὐ μὴ διασωθῇ […] οὐ μὴ σώσῃ, rendering ‫לא־ימלט‬, is converted into an ABA’B’ alternation by a further οὐ μὴ διασωθῇ (9:1). This may be accidental but, as other examples have shown, the translator of the Minor Prophets can remember and extend earlier patterns. 3.3.5 Am 9:11 A clearly significant alternation occurs here (involving assonance, variation, repetition, and compound verbs): ἀναστήσω (‫ ]…[ )אקים‬ἀνοικοδομήσω (‫ ]…[ )וגדרתי‬ἀναστήσω (‫ ]…[ )אקים‬ἀνοικοδομήσω (‫)ובניתיה‬. 36

32. 33. 34. 35. 36.

For further examples in the Minor Prophets, see Dines, “Stylistic Invention,” 38-39. Dines, “Stylistic Invention,” 39-41. Cf. B d’A 4, 53-54. B d’A 4, 60. Cf. Dines, Septuagint, 56; for alternation in Sir 4:3, Aitken, “Literary Attainment,” 119-120. 3. Arrangements

383

24. Stylistic Features of the Septuagint

3.3.6 Hag 1:5,7; 2:15,18 Where the Masoretic text consistently has ‫שׂימו לבבכם‬, the Septuagint has a (not quite regular) alternation: τάξατε […] θέσθε […] θέσθε […] ὑποτάξατε […] θέσθε. 37

3.4 Longer Alternations 38 3.4.1 Hos 5:2–13:4 There is an unobtrusive, but regular, alternation between ἐγὼ δὲ and καὶ ἐγὼ that does not match the variation between ‫ ואני‬and ‫ ואנכי‬and does not seem to serve any contextual purpose. 3.4.2 Hab 2:6 – Zeph 3:18 A small, but marked, alternation occurs between the interjections οὐαί and ὦ, linking these books (always adjacent in the Minor Prophets). Both words render ‫ הוי‬and form two distinct patterns, giving ABA’B’A’ for Hab 2:6-19, and ABA’ for Zeph 2:15–3:18; this is reminiscent of the patterning in Am 1:3-15; 2:1-6. 39 3.4.3 Hos 4:12 – Zech 10:1 Renderings of ‫( שׁאל‬only five occurrences in the Minor Prophets) may form an extended ABCA’ B’ alternation: ἐπηρώτων (Hos 4:12) […] αἰτεῖ (Mic 7:3) […] ἀπελέγετο (Jon 4:8) […] ἐπερώτησον (Hag 2:11) […] αἰτεῖσθε (Zech 10:1). Jon 4:8 could be contextually motivated; the rarity of the verb in the Minor Prophets marks the remaining ABA’ B’ pattern as significant. 40

4. Conclusion To some extent, the devices used by translators, such as alliteration, end-rhyme, variation, repetition, chiasm and so on—present from Genesis onwards—can be seen as arising naturally from sensitivity to Hebrew poetics. 41 But the translators had also learned at least the rudiments of Greek literary theory through the schooling they may be presumed to have had: even elementary education involved some exposure to classical literature and current stylistic fashions. 42 Through copying and memorising passages from Homer and other authors, they would have absorbed, to greater or lesser degree, a sense of the kind of language required by different contexts, and the imCf. B d’A 23.1, xix-xx; Dines, “Stylistic Invention,” 42. For fuller discussions, see Dines, “Stylistic Invention,” 40-41. See, 3.1.4 above. Few examples of extended arrangements come from the Pentateuch, perhaps because they remain to be identified, or because these early translators (with the possible exception of the Numbers translator) were not interested in creating this kind of effect; see J. M. Dines, “Was LXX Pentateuch a Style-Setter for LXX Minor Prophets?” in: M. K. H. Peters (ed.), XIV Congress of the International Organization for Septuagint and Cognate Studies. Helsinki 2010 (Septuagint and Cognate Studies 59), Atlanta, GA 2013, 397-411. 41. Usefully surveyed by Lundbom, Hebrew Prophets, 168-181. 42. Aitken, “Significance,” 508-509; Reese, Hellenistic Influence, 25-26. 37. 38. 39. 40.

384

4. Conclusion

24. Stylistic Features of the Septuagint

portance of making texts euphonious as well as accurate. 43 They are thus heirs to a double, often overlapping, tradition, something we should bear in mind when assessing the translators’ stylistic practices. These practices are, however, difficult to quantify, given the current state of Septuagint studies. So far, stylistic issues have mainly surfaced as by-products in modern translations and commentaries and are usually acknowledged only in passing. Although their significance is being recognized, there are as yet little more than collections of striking examples with which to work. Until each book has been systematically studied for its stylistic features, it is impossible to know for sure how widespread or how significant these are. Such study is, however, important for several reasons. First and foremost, it shows that translators have interests beyond producing a straightforward rendering of their source-text, however “literal” the translation may be overall. Secondly, it provides insights into the translators’ educational level and social context. Thirdly, attention to rhetorical effects may shed light on translators’ pedagogical and hermeneutical concerns. 44 Finally, stylistic study of the translations helps to situate them within the varied corpus of Hellenistic Greek writings, and not just as more or less accurate renderings of their parent Hebrew.

43. The translator of Sirach, for instance, shows considerable literary skill, and not only in the prologue; Aitken, “Literary Attainment,” 95-126. 44. For lack of an overarching rhetorical agenda in the Pentateuch, see Aitken, “Significance,” 13, with note 37. For hints of rhetorical intention in the Minor Prophets, see Dines, “Stylistic Invention,” 46-48. 4. Conclusion

385

25.1 The Style of the Book of Wisdom Luca Mazzinghi

The Book of Wisdom originated in Alexandria, Egypt, almost certainly in the last years of the reign of Emperor Augustus (30[27] BCE – 14 CE). 1 Within Jewish literature written in Greek, this book represents perhaps the best example of a well-balanced dialogue between Judaism and Hellenism. After a history of the study of the Book of Wisdom (§ 1), we present the vocabulary (§ 2.1) and literary genre (§ 2.2), followed by a presentation of rhetorical and stylistic usage: word order (§ 3.1), rhetorical figures (§ 3.2), and the problem of prose or poetry (§ 3.3).

1. History of Study Ancient commentators were already aware of the book’s particular style, so different from that of the majority of the other books of the Septuagint and so very Greek. In this connection, it was Jerome who wrote: “ipse stylus [of the Book of Wisdom] graecam eloquentiam redolet.” 2 Jerome’s judgement has been echoed by modern commentators on Wisdom. Brooke Foss Westcott held that the book is typical “of the style of composition which would be produced by the sophistic schools of rhetoric.” 3 Henry Swete stated that “no other book in the Greek Bible is so manifestly Alexandrian in tone and style.” 4 As early as 1860, Carl L. W. Grimm, perhaps the first great modern commentator on Wisdom,

1.

2. 3. 4.

However, some authors favor a date during the reign of Caligula (37-41 CE). Cf. D. Winston, Wisdom of Solomon (Anchor Bible 43), New York, NY 1979, 20-25; G. Scarpat, “Ancora sull’autore del libro della Sapienza” RivBibIt 15 (1967), 171-189, in particular 180-184; G. Scarpat, “Ancora sulla data di composizione della Sapientia Salomonis. Il termine diágnôsis (Sap 3,18; At 25,21)” RivBibIt 36 (1988), 363-375. But see M. Gilbert, “Sagesse de Salomon (ou Livre de la Sagesse)” DBS XI, Paris 1988, 91-93. Recent confirmation of this dating can be found in M. Gilbert, “‘La vostra sovranità viene dal Signore (Sap 6,3)’ : ambivalenza del potere politico nella tradizione sapienziale” RicStoBib 18 (2006), special issue “Il potere politico: bisogno e rifiuto dell’autorità (Atti della XXXVIII Settimana Biblica Nazionale—Roma, 6-10 settembre 2004)” edited by E. Manicardi and L. Mazzinghi, 117-132 (reedited as “Your sovereignity comes from the Lord (Wis 6:4)” in: M. Gilbert, La Sagesse de Salomon. The Wisdom of Solomon. Recueil d’études. Collected Essays (AnBib 190) Rome 2011, 121-140. Praefat. in Sal. libros, PL XXVIII, 1242. B. F. Westcott, “Wisdom of Solomon” in: W. Smith (ed.), Smith’s Dictionary of the Bible, vol. 4, Cambridge 1872, 3547. B. Swete, An Introduction to the Old Testament in Greek (second edition), Cambridge 1914 (reprint New York, NY 1968), 268. Swete adds that “in the style of the originally Greek books [of the Septuagint] there is little to remind us of the Semitic origin of the writers” (313).

386

1. History of Study

25.1 The Style of the Book of Wisdom

offered for the first time a list of typically Greek stylistic usages to be found in the book. 5 In fact, commentators’ attention to the style of Wisdom arose for the most part as a result of the late-eighteenth and early-nineteenth century debate as to the original language of the book: Aramaic/Hebrew or Greek? There had never before been a systematic examination of the style of the book; in any case, it led scholars to conclude that the book could not be considered a translation of a Semitic original, but had been composed directly in Greek. 6 The book is not lacking in Hebraisms, such as the constant use of the parallelismus membrorum, 7 but this is not a sufficient argument to support the hypothesis of a Semitic original. Along these lines, Joseph Reider writes that “[the book of Wisdom] is written in the purest form of Alexandrinian Greek, free from the Hebraisms and anomalies of the Septuagint and full of passages which combine the richest vocabulary with genuine rhetorical eloquence. Compared with the Septuaginta, Wisdom appears to be an original and independent work.” 8 In his work on the influence of Hellenistic culture on the Book of Wisdom, James M. Reese, the author who, until now, has devoted the most attention to style, is unambiguous in his conclusion: “this survey of the vocabulary and style of Wis shows that the sacred writer was trained in Greek rhetoric and was subject to a wide variety of Hellenistic influences.” 9 The works of Crysostome Larcher and Giuseppe Scarpat, unquestionably the two most important academic commentaries on Wisdom, substantially confirm Reese’s conclusions. 10 Nevertheless, the study of the style of the Book of Wisdom remains an open field. Only very recently has the first article entirely devoted to the subject, that of Alexis Léonas, been published. 11 This author sees in the book a 5. C. L. W. Grimm, Das Buch der Weisheit, Leipzig 1860, 6-9, in particular the list on page 7, which can be found simply replicated in successive commentaries. 6. A good summary of this debate and of the history of the study of the style of Wisdom can be found in C. Larcher, Le Livre de la Sagesse ou la Sagesse de Salomon, vol. I, Paris 1983, 91-95, and, more briefly, in Gilbert, “Sagesse de Salomon,” 61-65. The most significant studies include E. Gärtner, Komposition und Wortwahl des Buches der Weisheit, Berlin 1912, a pioneering work on the vocabulary of Wisdom, and, on the grammar and style of the book in relation to its content, F. Focke, Die Entstehung der Weisheit Salomos, Göttingen 1913, who claims, contrary to generally accepted opinion, that chapters 1–5 were originally composed in Hebrew (65-66). 7. Cf. Winston, Wisdom of Solomon, 15. E. g. WisSol 1:1-15, where the text appears as an attempt to imitate biblical parallelismus membrorum; cf. E. D. Reymond, “The Poetry of Wisdom of Solomon Reconsidered” VT 52 (2002), 385-389. The parallelismus membrorum is very rare in the last section of the book: WisSol 12:18-19:22; cf. Focke, Die Entstehung, 53. 8. J. Reider, The Book of Wisdom, New York, NY 1957, 25-26. On the other hand, the same Reider holds that the Book of Wisdom is characterised by “some ignorance and poverty of diction” (27). See also a similar judgement by Larcher, Sagesse, vol. I, 102. But cf. Winston, Wisdom, 17-18, in particular 17: “the hypothesis that Wisdom is a translation of a Hebrew original [is] virtually untenable.” 9. J. M. Reese, Hellenistic Influence on the Book of Wisdom and its Consequences (AnBib 41), Rome 1970, 30. 10. Cf. C. Larcher, Le Livre de la Sagesse ou la Sagesse de Salomon, vols. I–III, Paris 1983-1987; G. Scarpat, Il libro della Sapienza, vols. I–III, Brescia 1989-1999. 11. A. Léonas, “The Poetics of Wisdom. Language and Style in the Book of Wisdom” in: E. Bons / 1. History of Study

387

25.1 The Style of the Book of Wisdom

conscious attempt to imitate Septuagint style and idiom, “as it could have been seen from within the Greek linguistic and literary system,” picking from the Septuagint “only those features of style that matched his aesthetic norm.” 12

2. Vocabulary and Literary Genre 2.1 Vocabulary The vocabulary used in the Book of Wisdom provides a first indication of the exceptional nature of our author’s Greek. In all, there are 1,734 words in the book. Of these, 1,303 appear only once. Moreover, of these, 335 words, approximately 20 % of the book, are hapax legomena in the Septuagint. 13 To this count should be added 126 others that appear only in the later books of the Septuagint such as the Book of Sira and 3 and 4Maccabees. Not only this, but our book contains a good number of hapax totius graecitatis: see, e. g., νηπιοκτόνος (11:7), δυσδιήγητος (17:1), περικομπέω (17:4), ἐφύβριστος (17:7), φυλακίζω (18:4), ἀχανής (19:17). 14 All this bring us to the conclusion that our author is an authentic creator of language. Many of the terms used in Wisdom hail from a vocabulary that is erudite, philosophical, not infrequently poetic, sometimes also of a medical and scientific nature. Moreover, our author is a lover of compound words, which are extremely numerous in the book. 15 Finally, in keeping with Hellenistic taste, we encounter in Wisdom fiftynine occurrences of rare or poetic words with α-privative and compound words with κατα-, περι-, and ευ-. 16 If some stylistic features, such as, for instance, the causal use of ὅτι, especially in chapters 1, 2, and 5, 17 could suggest a Hebrew original, others, such as, for example, the

12. 13. 14.

15. 16. 17.

T. J. Kraus (eds.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint (FRLANT 241), Göttingen 2011, 99-126. Léonas, “Poetics,” 124, 122. Reese, Hellenistic Influence, 3. Larcher (Études sur le Livre de la Sagesse, Paris 1969, 182) identifies 315. Cf. a list of these hapax in G. Scarpat, “Ancora sull’autore” 172-180. See also Gilbert, “Sagesse,” 64 and L. Mazzinghi, Notte di paura e di luce. Esegesi di Sap 17,1-18,4 (AnBib 134), Rome 1995, 274, see also 186 for δυσάλυκτος (listed by Scarpat but present in PGM IV:2858) —a good example of a magic vocabulary used by Wisdom. For her part, Martina Kepper identifies nineteen hapax totius graecitatis. See M. Kepper, Hellenistische Bildung im Buch der Weisheit: Studien zur Sprachgestalt und Theologie der Sapientia Salomonis (BZAW 280) Berlin/New York, NY 1999, 71-72. Cf. lists in Focke, Enstehung, 61; Winston, Wisdom, 14-15, Kepper, Hellenistische Bildung, 5152. Cf. a list in Focke, Entstehung, 60-61. See R. Sollamo, Renderings of Hebrew Semipropositions in the Septuagint (AASF B Diss 19), Helsinki 1979, 296. Cf. L. Ruppert, “Gerechte und Frevler (Gottlose) in Sap 1,1-6,21” in: H. Hübner (ed.), Die Weisheit Salomos im Horizont Biblischer Theologie, Neukirchen-Vluyn 1993, 1-54, who still maintains the existence of a Semitic original for WisSal 2:12-20; 5:1-7.

388

2. Vocabulary and Literary Genre

25.1 The Style of the Book of Wisdom

use of the infinitive 18 or of personal pronouns 19, differ notably from the style of the Septuagint and show themselves to be genuinely Greek.

2.2 Literary Genre The choice of the encomium was dictated by the book’s main aim: to exhort the Jews of Alexandria, particularly the youth, to remain faithful to their tradition, while not shutting themselves off under the cover of their Jewish identity from any contact with the Hellenistic world. 20 Although some authors have suggested that Wisdom represents a logos protreptikos, 21 the literary structure of the book corresponds to the articulation of the encomium, according to the rules of classical rhetoric: exordium, appeal to those addressed, and refutation of opponents (1:1–6:21); announcement of the theme of the elogium (6:22-25); the elogium proper (7–9); synkrisis or comparatio, with exempla drawn from the past (10–19:9); 22 epilogue and recapitulation (19:10-22). The Book of Wisdom also employs other elements typical of the epideictic genre such as the prosôpopoiia. Such is the case of the discourse of the Wicked (WisSal 2:120), which culminates in an accusation (kathegoria—WisSal 2:21-24). Another example can be seen in the speech of the Impious (WisSal 5:3-13). Another case of classical prosopoiia is the personification of Justice and Wisdom (and perhaps Death). 23 Elsewhere (3:1-9), our sage utilizes the genre of the funerary epitaph (epitaphios logos). Another significant Hellenistic feature of the Book of Wisdom’s style is its carefully crafted concentric structure, which is particularly evident in the organization of chapters 1–6, 7–8, 9, 16–19, but can also be seen in shorter pericopes such as 18:20-24. 24 However, the Book of Wisdom differs from the epideictic genre in several ways. Particularly prominent among these is the midrashic style that characterizes the entire work. The author continually refers to the scriptures of Israel, which he aims to reread and actualize in a cultural context that is typically Hellenistic.

18. Cf. I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF B 132,1), Helsinki 1965, 193. 19. Cf. A. Wifstrand, Die Stellung der enklitischen Personalpronomina bei den Septuaginta, Lund 1950, 63. 20. Cf. J. G. Gammie, “Paraenetic Literature: Toward the Morphology of a Secondary Genre” Semeia 50 (1992), special issue “Paraenesis: Act and Form” edited by J. G. Gammie and L. G. Perdue, 52. 21. Cf. Reese, Hellenistic Influence, 90-121; cf. also, D. Winston, “A Century of Research on the Book of Wisdom” in: G. Bellia / A. Passaro (eds.), The Book of Wisdom in Modern Research, Berlin/New York, NY 2005, 2-5. 22. See I. Heinemann, “Synkrisis oder äussere Analogie in der Weisheit Salomos” TZ 4 (1948), 241-251. 23. Cf. J. R. Dodson, The “Powers” of Personification. Rhetorical Purpose in the Book of Wisdom and in the Letter to the Romans (BZNW 161), Berlin/New York, NY 2008. 24. Cf. J. M. Reese, “Plan and Structure in the Book of Wisdom” CBQ 27 (1965), 391-399; A. G. Wright, “The Structure of the Book of Wisdom” Bib 48 (1967), 165-184; Gilbert, “Sagesse de Salomon”, 65-77; and, especially, P. Bizzeti, Il libro della Sapienza. Struttura e genere letterario, Bologna 1984. 2. Vocabulary and Literary Genre

389

25.1 The Style of the Book of Wisdom

3. Rhetorical and Stylistic Usage 3.1 Word Order The style of the book is clearly dependent on the most characteristic features of Greek rhetoric. Undoubtedly the most significant, in relation to word-order—one of the basic features of Greek style—is the use of highly polished periodic sentences. The structure of the phrase is always carefully measured (often two substantives and two verbs; cf. also the frequent use of chiasm: 1:4-8, with different cases; 3:15, etc.). While the simple coordination with καί predominates in the first ten chapters, the use of μέν … δέ becomes more frequent in chapters 11–19, with an accompanying increase in the periodic style (cf. 12:3-7,27; 13:11-15; 15:7; 17:16-19). Generally rather rare in the Septuagint, the repeated use of hyperbaton, a good 240 times in Wisdom, especially in chapters 10–19, should be noted; at times, as happens in WisSal 14:18, it is in the form of a double hyperbaton. 25 The use of hyperbaton is not dictated solely by stylistic factors but also by content, as, for example, when it is used to highlight a specific term, e. g. σοφίαν (9:4a); νύκτα (17:2a); σκότους (17:20b); ἀνάγκη (19:4a).

3.2 Rhetorical Figures The author shows a keen awareness of the more classical figures of Greek rhetoric, tropes and figures, the frequency of which is noticeably higher when compared with the books of the Septuagint that were translated from Hebrew. 26 The following is a list of the most important ones: Metaphor. The metaphor of death is used in chapters 1–6. 27 The metaphor of light and darkness, employed throughout the fifth diptych (17:1–18:4), takes on a value that is simultaneously cosmological, psychological, moral, and eschatological. Litotes is quite frequent, seventeen times according to Reese 28: it occurs at the beginning of the book (τοῖς μή πειράζουσιν—1:2a), at the end (οὐχ ὑπερεῖδες—19:22b), and many times in between (1:11b; 11:7b; 12:9a,10b,13b; 17:4a,12a, etc.). Anaphora. The repetition of αὐτός [αὐτή] in chapter 10 offers a good example. Other instances include 17:13-15 and 18b-19d.

25. For the various kinds of hyperbaton in Wisdom with some examples, cf. Reese, Hellenistic Influence, 26-27; cf. also Mazzinghi, Notte, 278. 26. We are following here the scheme outlined by G. O. Rowe, “Style” in: S. E. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period. 330 B.C. – A.D. 400, Leiden/New York, NY/Köln 1997, 121-157; cf. also J. D. Denniston, Greek Prose Style, Oxford 1952. Further examples of these stylistic usages of the Book of Wisdom are to be found in Winston, Wisdom, 1517; Reese, Hellenistic Influence, 27-28; cf. also Larcher, Sagesse, vol. I, 102-103. 27. Cf. M. Kolarcik, The Ambiguity of Death in the Book of Wisdom 1–6. A study of literary structure and interpretation (AnBib 127), Rome 1991. 28. Reese, Hellenistic Influence, 30.

390

3. Rhetorical and Stylistic Usage

25.1 The Style of the Book of Wisdom

Paranomasia. Examples can be found in 4:2; 5:3b,10c,14,22a; 6:10a,22-23; 8:4; 11:9; 13:19b; 14:5a; 17:12-13; 18:15. In addition, there are a huge number of examples of assonance and alliteration. The book also contains many examples of word play, especially on words of the same root; e. g. 2:23 (ἰδίας ἀϊδιότητος); 5:17 (πανοπλίαν – ὁπλοποιήσει); 11:14-15; 17:19,20,21; 18:4. Accumulatio. A parade example can be found in the twenty-two epithets of wisdom (7:22-23). Double-duty words (apó koinoû). Examples are ἐν ἀγάπῃ (3:9), which can refer either to the faithful or to God, and συμπαροῦσά μοι κοπιάσῃ (9:10), where μοι can refer to both verbs. Asyndeton. Asyndetical constructions are frequently used to indicate a new point of view (4:10,20; 10:12). Isocolon. As in 1:1 and 18:1b. Homoioteleuton. See 1:1,5; 2:3-4; 4:10. A variation of homoioteleuton is used in 3:11-12; 7:17-21; 8:19-20; 10:13-15; 14:11-12. Antithesis: See 3:5a; 4:16; 7:6; 9:16; 18:7b. The Book of Wisdom also employs the classical tropos of sorites in 6:17-20 while the debated text of 7:19-20 may constitute the rhetorical figure of epiphonema. 29

3.3 Prose or Poetry? A final aspect of the particular style of the Book of Wisdom is the use of clauses which resemble classical metre. J. A. F. Gregg discovered examples of iambic or dactylic metre. His work was closely followed by that of Henry St. John Thackeray, who remains the only scholar to have thoroughly studied the poetic aspect of the Book of Wisdom, although his pioneering study requires further research. An exemplary text in this respect is WisSal 17:1–18:4, in which, in addition to the discovery of a typically Asianic style (for example, the prosa fracta), it is possible to recognize to some extent, if not systematically, the use of classical metre, as the hexametric final clause (cf. particularly the passage 17:16-21). 30

4. Conclusion The style of the Book of Wisdom is truly distinctive. It stands out as a highly “Greek” style, especially in comparison with the other books of the Septuagint, even those books that are not translations. We can justly define it as an authentic attempt at in-

29. Cf. J. R. Dodson, The “Powers” of Personification. 30. Cf. J. A. F. Gregg, The Wisdom of Solomon, Cambridge 1909, xv; H. St. J. Thackeray, “Rythm in the Book of Wisdom” JTS 6 (1905), 232-237. For the style of WisSal 17:1–18:4, see Mazzinghi, Notte, 277-280; see also 163-165, for an analysis of the poetic dimension of WisSal 17:16-21. 4. Conclusion

391

25.1 The Style of the Book of Wisdom

culturation. 31 To be sure, our sage remains a Jew who intends to offer in his work a profoundly biblical message about God and man, but he does so by making use of a style that is typical of the Greek rhetoric of his age.

31. Cf. M. Gilbert, “Le livre de la Sagesse et l’inculturation” in: M. Gilbert / P. In Syek Sye / T. Nkéramihigo (eds.), L’inculturation et la sagesse des nations, Rome 1984, 1-11; M. Gilbert, “La Sagesse de Salomon et l’hellénisme” Hieros 4 (1999), 11-22; L. Mazzinghi, “Il libro della Sapienza: elementi culturali” RicStoBib 1-2 (1998), special issue “Il confronto tra le diverse culture nella Bibbia da Esdra a Paolo (XXXIV Settimana Biblica Nazionale)” edited by R. Fabris, 179-197.

392

4. Conclusion

25.2 The Language of the Book of Judith* Eberhard Bons

Introduction One of the most striking features of the Greek of the book of Judith is its affinity to the language of the translated books of the Septuagint. On the level of vocabulary and syntax, this narrative, of which there is no extant ancient Hebrew version, employs an overwhelming number of linguistic features characteristic of the Septuagint and Jewish or Christian literature influenced by it. Not only does the presence of certain elements of translational Greek, e. g. the redundant use of the genitive forms of pronouns (e. g. four occurrences of αὐτοῦ in Jdt 1:13; four instances of αὐτῆς in Jdt 8:5 and four also Jdt 13:3; four occurrences of αὐτῶν in Jdt 5:21) recall Hebrew syntax, but the absence of elements very common in non-biblical Greek texts (e. g. constructions with μέν … δέ, which do not occur even where they might be expected, as in Jdt 15:5-7, particles like οὖν and ἄρα) is conspicuous. This evidence prompted most scholars of the last century to argue that the Greek text of the book of Judith should be considered a translation from a Semitic source, either Hebrew or Aramaic. 1 In recent years however doubts have been raised as to whether this hypothesis is still convincing. 2 The present article will provide an outline of the most salient linguistic features of the book of Judith. Without aiming at completeness, I hope the following observations can lead to a more differentiated assessment of the language of this deutero-canonical narrative. The result can be anticipated as follows: although it cannot be denied that * 1.

2.

I wish to express my thanks to my colleagues Jan Joosten and Romina Vergari with whom I was able to discuss various aspects of this article. See, e. g., R. Hanhart, Text und Textgeschichte des Buches Judith (MSU 14), Göttingen 1979, 9: “Der griechische Text des Buches Iudith ist ein Übersetzungstext. Seine Vorlage war entweder hebräisch oder aramäisch.” See also E. Zenger, Das Buch Judit (JSHRZ I/6), Gütersloh 1981, 430; C. A. Moore, Judith. A New Translation with Introduction and Commentary (AncB 40), Garden City, NY 1985, 66; J. Vilchez Lindez, Tobías y Judit (Nueva Biblia Española), Estella 2000, 235; N. Fernández Marcos, La Biblia griega Septuaginta, vol. II, Libros históricos, Salamanca 2011, 696. H. Engel, “‘Der Herr ist ein Gott, der Kriege zerschlägt’. Zur Frage der griechischen Originalsprache und der Struktur des Buches Judit” in: K. D. Schunck / M. Augustin (eds.), Goldene Äpfel in silbernen Schalen (BEATAJ 20), Frankfurt 1992, 155-168; J. Joosten, “The Language and the Milieu of the Book of Judith” in: Meghillot. Studies in the Dead Sea Scrolls V-VI. A Festschrift for Devorah Dimant (2003), *159-*177; J. Corley, “Septuagintalisms, Semitic interference, and the original language of the Book of Judith” in: J. Corley (ed.), Studies in the Greek Bible, Washington, DC 2008, 65-96; B. Schmitz, “Ιουδιθ und Iudith. Überlegungen zum Verhältnis der Judit-Erzählung in der LXX und der Vulgata” in: J. Cook / H.-J. Stipp (eds.), Text-Critical and Hermeneutical Studies in the Septuagint (VTS 157), Leiden 2012, 359379; B. Schmitz / H. Engel, Judit. Übersetzt und ausgelegt (HThKAT), Freiburg 2014, 40-43; D. L. Gera, Judith (CEJL), Berlin 2014, 79-97. Introduction

393

25.2 The Language of the Book of Judith

the Greek text reflects Hebrew word order and syntax at more than one point and that certain words can be explained as calques from Hebrew, on closer inspection various linguistic features turn out to be uninfluenced by the Hebrew language—indeed there is strong evidence to suggest that they are independent from a putative Hebrew model. The first section of the present article will focus on selected syntactical features, regarding in particular sentence construction, prepositions, and pronouns. The second section will examine a series of examples of specific vocabulary of the book of Judith. The results obtained will shed some further light on the question whether this text was translated from Hebrew as a whole or written in Greek from the outset.

1. Syntax 1.1 Sentence Construction 1.1.1 Parataxis and Syntaxis Parataxis is predominant (e. g. Jdt 2:13-14; 13:9; 14:3); hypotaxis, in particular subordinate clauses formed with the subjunctive, is the exception. Only a few scattered examples can be found, e. g. clauses introduced by conjunctions such as ἐάν (Jdt 7:31; 8:11,15,17,22; 10:16; 11:6,10,16,23; 12:3,12; 14:2), ἵνα (Jdt 7:9,28; 11:11; 12:2; 14:5,13; 15:4), and ὅπως (Jdt 3:8). Subordinate clauses formed with the indicative are not very frequent either, as the following examples show: εἰ (Jdt 5:20,21; 11:2; 12:12), ἐπεί (Jdt 11:12), ἕως (Jdt 15:5), ἡνίκα (Jdt 14:11; 16:18), μέχρις (Jdt 5:10), ὅτε (Jdt 5:18), ὡς (Jdt 13:12; 14:6,7,12,15,19; 15:1,4,9; 16:18). Two cases, however, deserve closer attention: 1. From the point of view of sentence construction, Jdt 11:2 is a salient example: καὶ νῦν ὁ λαός σου ὁ κατοικῶν τὴν ὀρεινὴν εἰ μὴ ἐφαύλισάν με οὐκ ἂν ἦρα τὸ δόρυ μου ἐπ᾽αὐτούς. In this clause, the element that is emphasized (“your people who live in the hill country”) precedes the conjunction εἰ and the following conditional clause. 3 2. From the point of view of Greek syntax, the conjunction καί introducing the apodosis in Judith 5:20 (καὶ ἀναβησόμεθα) is not necessary. Albeit attested in nonbiblical Greek, this construction is a Hebraism. 4 An analogous phenomenon can be observed in constructions formed by καὶ ἐγένετο ὡς. The subordinate clause introduced by ὡς is followed by a main clause introduced by καί: καὶ ἐγόγγυσεν πᾶς ὁ λαός (Jdt 5:22), καὶ ἀνέστη (Jdt 10:2). A similar example can be found in Judges 2:4: καὶ ἐγένετο ὡς ἐλάλησεν ὁ ἄγγελος κυρίου τοὺς λόγους τούτους πρὸς πάντας υἱοὺς Ισραηλ καὶ ἐπῆραν ὁ λαὸς τὴν φωνὴν αὐτῶν. Admittedly, it is remarkable that a narrative text like the book of Judith contains a relatively small number of subordinate clauses. However, the picture does not change radically if another narrative corpus of the Septuagint is taken into consideration: the translation of the books of Samuel and Kings. In fact, in this corpus, which is more

3. 4.

See BDF, § 475. See BDF, § 442, 7; M. S. Enslin, The Book of Judith. Greek text with an English translation, commentary and critical notes (edited with a general introduction and appendices by Solomon Zeitlin), Leiden 1972, 92.

394

1. Syntax

25.2 The Language of the Book of Judith

than eight times the length of the book of Judith, the occurrences of, for example, final ἵνα are no more frequent.

1.1.2 Nominal Clauses This feature, typical of Hebrew syntax, has some equivalents in the book of Judith, in particular κύριος ὄνομά σοι (Jdt 9:8), σὺ μόνος ἀγαθός (Jdt 11:8), μεθ᾽ ἡμῶν ὁ θεός (Jdt 13:11), σὺ ὕψωμα Ιερουσαλημ σὺ γαυρίαμα μέγα τοῦ Ισραηλ σὺ καύχημα μέγα τοῦ γένους ἡμῶν (Jdt 15:9). Nevertheless, clauses with copula are attested as well, e. g. ἀλλὰ ταπεινῶν εἶ θεός ἐλαττόνων εἶ βοηθός (Jdt 9:11) See also Jdt 9:14; 11:17, 23; 14:18; 16:13. 1.1.3 Relative Clauses The book of Judith provides some examples of particular constructions, 5 e. g. the attraction of the relative (ἀφ᾽ ἧς ἡμέρας εἶδεν αὐτήν—Jdt 12:16; see also Jdt 14:8 and 16:22) and the incorporation of the antecedent in the relative clause (καὶ τίνες ἃς κατοικοῦσιν πόλεις [= τίνες πόλεις ἃς κατοικοῦσιν, a varia lectio quoted in the apparatus of Gö]—Jdt 5:3; ἐγύμνωσεν ὃν ἐνεδεδύκει σάκκον [= τὸν σάκκον ὃν ἐνεδεδύκει]—Jdt 9:1; δὸς ἐν χειρί μου […] ὃ διενοήθην κράτος [= τὸ κράτος ὃ διενοήθην]—Jdt 9:9). A more complicated example of the latter category is Judith 8:15: αὐτὸς ἔχει τὴν ἐξουσίαν ἐν αἷς θέλει σκεπάσαι ἡμέραις (= αὐτὸς ἔχει τὴν ἐξουσίαν σκεπάσαι ἐν ταῖς ἡμέραις ἃς θέλει). Of course, relative clauses like ἀφ᾽ ἧς ἡμέρας are frequently attested in the translated books of the Septuagint (e. g. Exod 9:18). The other cases, however, reflect a quite elegant Greek unlikely to have been translated from a corresponding Hebrew source text of departure. On the other hand, several relative clauses have a pleonastic element which appears to betray the influence of Hebrew syntax: 6 ἐκεῖ (Jdt 5:19; 8:22), ἐν αὐτοῖς (Jdt 7:10), ἐν αὐτῷ (Jdt 10:2), αὐτῶν (Jdt 16:3). 1.1.4 Infinitives Instead of subordinate clauses, the book of Judith several times uses a preposition with (articular) infinitives, e. g.: πρὶν ἐλθεῖν τὴν ῥομφαίαν ἐπ᾽ αὐτούς (Jdt 7:14), διὰ τὸ ἐπὶ πλεῖον γεγονέναι τὸν πότον (Jdt 13:1), πρὸ δὲ τοῦ ποιῆσαι ταῦτα (Jdt 14:5), πρὸ τοῦ ἀποθανεῖν αὐτήν (Jdt 16:24), μετὰ τὸ ἀποθανεῖν αὐτήν (Jdt 16:25). In this regard, the Greek style of the book of Judith fits the general tendencies of koine syntax. 7 However, one typical element of Hebraizing style is completely absent: phrases formed according to the model καὶ ἐγένετο ἐν τῷ + infinitive + noun or pronoun in the accusative (e. g. Gen 19:29; Num 10:34; etc.). To be sure, the book of Judith does not refrain from using the typical biblical ἐγένετο (see above), but this verb is always followed by a subordinate clause introduced by ὡς (Jdt 2:4; 5:22, 10:1; etc.; see also Gen 27:30; 39:13; etc.).

5. 6. 7.

See BDF, § 294; for the following examples, see also Joosten, “Language and Milieu,” *161-162. BDF, § 297. A. Debrunner / A. Scherer, Geschichte der griechischen Sprache, vol. II, Grundfragen und Grundzüge des nachklassischen Griechisch, Berlin 1969, § 197. 1. Syntax

395

25.2 The Language of the Book of Judith

1.1.5 The Accusative with the Infinitive and the Simple Infinitive The accusative with the infinitive is an exception in the book of Judith (see Jdt 14:14: ὑπενόει γὰρ καθεύδειν αὐτὸν μετὰ Ιουδιθ). In many cases, the infinitive without the accusative follows a verbum dicendi, especially when the subject is identical with that of the governing verb, 8 e. g. ὡς ὤμοσεν αὐτοῖς παραδώσειν τὴν πόλιν […] τοῖς Ἀσσυρίοις (Jdt 8:9). Interestingly, the future infinitive παραδώσειν, which has no exact counterpart in Hebrew 9, conforms to classical Greek syntax. 1.1.6 Participle Constructions Whereas the genitive absolute is very rare (Jdt 4:7: στενῆς τῆς προσβάσεως οὔσης), the conjunctive participle, albeit far more frequent, is nonetheless an exception among the numerous paratactical clauses, e. g. τὸν Ολοφέρνην […] δεύτερον ὄντα (Jdt 2:4), πᾶς χειμάρρους καὶ ποταμὸς ἐπικλύζων […] πληρωθήσεται (Jdt 2:8), διαβάντες τὸν Ιορδάνην ἐκληρονόμησαν πᾶσαν τὴν ὀρεινήν (Jdt 5:15), ἐπιστρέψαντες […] ἀνέβησαν ἐκ τῆς διασπορᾶς (Jdt 5:19), καταβάντες δὲ οἱ υἱοὶ Ισραηλ […] ἐπέστησαν αὐτῷ καὶ λύσαντες αὐτὸν ἀπήγαγον (Jdt 6:14). 10 It should be noticed that in most of the occurrences the conjunctive participle is in the nominative case 11, as it refers to the subject of the respective clause. Other cases are exceptional, e. g. the dative in Judith 8:26: ὅσα ἐγένετο τῷ Ιακωβ ἐν Μεσοποταμίᾳ τῆς Συρίας ποιμαίνοντι τὰ πρόβατα Λαβαν τοῦ ἀδελφοῦ τῆς μητρὸς αὐτοῦ. Furthermore, some occurrences of the participles deriving from λέγω should be mentioned. Used to introduce direct discourse, they obviously betray a Hebraizing style (see Jdt 3:1 and 6:18: λέγοντες; Jdt 12:6: λέγουσα). 12 However, given the relatively large number of instances of direct discourse in the book of Judith, the percentage of this kind of participle is rather low compared with other narrative texts (e. g. the Book of Exodus, with its approximately fifty instances of λέγων). Finally, three striking phenomena deserve attention: 1. Judith 12:11 reads: πεῖσον δὴ πορευθεὶς τὴν γυναῖκα τὴν Εβραίαν. The word order does not correspond to the logical order of the action, as revealed by several modern translations, e. g. NRSV: “Go and persuade the Hebrew woman.” In this case, the aorist participle πορευθείς expresses the relative past time in relationship to the imperative πεῖσον. 13 Nevertheless, it is clear that it is this imperative that governs the accusative object τὴν γυναῖκα, although πεῖσον is separated from it by πορευθείς. It is hardly conceivable that such a word order represents a word-to-word translation from a Hebrew Vorlage. Further, it should be stressed that occurrences of πείθω (except the forms of πέποιθα) are very rare in the translated books of the Septuagint (1Reigns 24:8; Est 4:4; Prov 26:25). Therefore, the question arises: Which Hebrew verb could have been rendered by πεῖσον? At any rate, a standard equivalent is not available. As a result, a strong case can be made that πεῖσον δὴ πορευθεὶς τὴν γυναῖκα 8. BDF, § 406. 9. For this argument, see Joosten, “Language and Milieu,” *161. 10. Other occurrences of the conjunctive participle include Jdt 7:8,12; 8:17,36; 10:15,23; 11:16,18,22; 12:9,11,12,14,16,19; 13:4,6-7,10,13,15,20; 14:1,3-5; 15:2,7,11. 11. See also Gera, Judith, 80. 12. See BDF, § 420. The singular λέγων is used, in Judith 4:7, to introduce the text of a letter. 13. BDF, § 339. A similar case is Tob 5:3B,A: λαβὲ πορευθεὶς τὸ ἀργύριον.

396

1. Syntax

25.2 The Language of the Book of Judith

τὴν Εβραίαν is not a translation—at least not a literal translation—of an underlying Hebrew text. 2. Occasionally the book of Judith employs future participles, i. e. forms that have no counterpart in the Hebrew language (e. g. τίς εἰμι ἐγὼ ἀντεροῦσα τῷ κυρίῳ μου —Jdt 12:14). In Judith 11:14, the future participle μετακομίσοντας (Rahlfs) is used to express purpose: 14 καὶ ἀπεστάλκασιν εἰς Ιερουσαλημ […] τοὺς μετακομίσοντας αὐτοῖς τὴν ἄφεσιν. The messengers have been sent to obtain an authorization for the inhabitants of Jerusalem to consume the firstfruits (in the Göttingen edition of the Septuagint, Hanhart opts for the variant μετακομίσαντας: the messengers have already brought the authorization). As in the aforementioned case, it is uncertain which Hebrew equivalents could have been rendered by these participles. 3. In some cases however the Greek style of the Book of Judith reflects Hebrew syntactic patterns. In at least one instance, the finite form of a verb is preceded by the participle of the same verb (ἐπιτελῶν ἐπιτελέσεις—Jdt 2:13). A slightly different construction consists not of two verbal forms, but of a verb and a noun of the same root (see Jdt 6:4: ἀπωλείᾳ ἀπολοῦνται, furthermore Jdt 9:4: ἐζήλωσαν τὸν ζῆλόν σου). Be this as it may, this construction recalls the Hebrew infinitive absolute used to emphasize the idea of the corresponding verb. 15

1.2 Prepositions Some peculiarities in the use of prepositions can explained either against a Koine Greek or a specifically Hebrew background. It might suffice to quote some significant examples.

1.2.1 ἀπό The preposition ἀπό can be used in a causal sense, 16 e. g. οὐ γὰρ ἦν ἀριθμὸς ἀπὸ πλήθους αὐτῶν (Jdt 2:20), ἐξέλιπον ἀπὸ τῆς δίψης (Jdt 7:22). Interestingly, the verb φοβέομαι can be either transitive (Jdt 1:11; 8:8), as in non-biblical Greek, or used to govern the preposition ἀπό (Jdt 5:23). The latter use shows the influence of the Hebrew language. 17 1.2.2 εἰς The strict distinction between εἰς and ἐν tended to disappear in the koine. 18 Accordingly, εἰς can appear where ἐν would be expected, e. g. ἀπέθανεν εἰς Βαιτυλουα (Jdt 16:23). Under Hebrew influence, in clauses formed with εἶναι or γίνομαι, the pre-

14. BDF, § 351. 15. See Gera, Judith, 84; furthermore G. Dorival / M. Harl / O. Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien, Paris 1988, 239; G. Mussies, “Greek in Palestine and the Diaspora” in: S. Safrai / M. Stern (eds.), The Jewish People in the First Century (CRI I/2), vol. 2, Assen/Amsterdam 1976, 1040-1064, here 1048. 16. BDF, § 210. 17. See R. Helbing, Die Kasussyntax der Verba bei den Septuaginta: ein Beitrag zur Hebraismenfrage und zur Syntax der Koinē, Göttingen 1928, 29. 18. BDF, § 205. 1. Syntax

397

25.2 The Language of the Book of Judith

position εἰς with accusative appears instead of a predicative nominative: 19 ἐγένοντο ἐκεῖ εἰς πλῆθος πολύ (Jdt 5:10), ὁ ναὸς […] ἐγενήθη εἰς ἔδαφος (Jdt 5:18), ἐσόμεθα εἰς ὀνειδισμόν (Jdt 5:21), ἔσονται εἰς κατάβρωσιν (Jdt 5:24), ἦσαν εἰς πλῆθος πολὺ σφόδρα (Jdt 7:18), κρεῖσσον γὰρ ἡμῖν γενηθῆναι αὐτοῖς εἰς διαρπαγήν ἐσόμεθα γὰρ εἰς δούλους (Jdt 7:27), ἐσόμεθα εἰς πρόσκομμα καὶ εἰς ὄνειδος (Jdt 8:22).

1.2.3 ἐν Instrumental ἐν, which corresponds to the Hebrew preposition -‫ב‬, is attested in Judith 16:4 (τοὺς νεανίσκους μου ἀνελεῖν ἐν ῥομφαίᾳ [v.l.: ῥομφαίᾳ]) and in Judith 16:6 (Ιουδιθ […] ἐν κάλλει προσώπου αὐτῆς παρέλυσεν αὐτόν). Furthermore ἐν is used when speaking of attendant military forces: 20 ἐν ἅρμασι καὶ ἱππεῦσι καὶ πεζοῖς ἐπιλέκτοις αὐτῶν (Jdt 2:19), ἦλθεν ἐν μυριάσι δυνάμεως αὐτου (Jdt 16:3). In Judith 11:18, however, the textual witnesses read either ἐν or σύν: ἐξελεύσῃ σὺν/ἐν πάσῃ τῇ δυνάμει σου. 1.2.4 ἐπί The verb πέποιθα governs ἐπί with dative instead of the mere dative (Jdt 7:10). 21 1.2.5 μετά The proposition μετά is used several times in the sense of “in dealing with,” e. g. ὅσα ἐποίησεν μετὰ Αβρααμ (Jdt 8:26), τελείως πρᾶγμα ποιήσει μετὰ σοῦ ὁ θεός (Jdt 11:6). See also Jdt 11:16; 15:10. 1.2.6 παρά The preposition παρά with accusative serves to express a comparison after a verb or an adjective (Jdt 12:18; 13:18; see Gen 43:34; Eccl 2:9). 22 1.2.7 Improper Prepositions A few unusual verbal phrases with ἐνώπιόν/ἐναντίον τινος should be noted. In two cases, the verb ἀρέσκω is constructed with ἐναντίον instead of governing a dative (see Jdt 7:16 and the quite identical quotation in Jdt 11:20: ἤρεσαν οἱ λόγοι αὐτῆς ἐνώπιον/ἐναντίον Ολοφέρνου καὶ ἐναντίον/ἐνώπιον πάντων τῶν θεραπόντων αὐτοῦ). In the translated books of the Septuagint there is additional evidence for this kind of construction (see, e. g., Gen 34:18; 41:37). Furthermore, the verb ἁμαρτάνω can govern either the preposition εἰς (Jdt 5:20; 11:10) or ἐνώπιον (Jdt 5:17: ἥμαρτον ἐνώπιον τοῦ θεοῦ αὐτῶν). This specific linguistic use of ἐνώπιον/ἐναντίον in Septuagint texts, which is probably foreign to non-Jewish Greek literature, is likely to have been influenced by conventions at use in the Persian court. 23

19. 20. 21. 22. 23.

BDF, § 145. BDF, § 198, 1. See Helbing, Kasussyntax, 197-198. BDF, § 185. See J. Joosten, “L’agir humain devant Dieu : remarques sur une tournure remarquable de la Septante”, RB 113 (2006), 5-17.

398

1. Syntax

25.2 The Language of the Book of Judith

1.2.8 Hebraistic Circumlocutions of Prepositional Concepts 24 Most of the aforementioned syntactical features corroborate the hypothesis of an influence of the Hebrew language on the Greek of the book of Judith. This applies even more to the many prepositional phrases that reflect without any doubt Hebrew compound prepositions, 25 e. g. ἀπὸ προσώπου (Jdt 2:14; 4:2; 5:8,12; 6:20; 10:11; 11:16; 12:13; 14:3; 16:15), ἐκ (τοῦ) προσώπου (e. g. Jdt 1:5; 5:16; 13:1,4), ἐπὶ πρόσωπον (Jdt 11:11), κατὰ πρόσωπον (Jdt 2:23,25; 4:6,11; 7:6; 10:23; 11:5; 12:13; 15:2; 16:20), πρὸ προσώπου (Jdt 8:15; 10:13), ἀνὰ μέσον (Jdt 7:24; 8:11), διὰ μέσου (Jdt 11:19), ἐν μέσῳ (Jdt 6:16,17; 14:8,19), ἐκ μέσου (Jdt 6:11; 7:19), ἐν ὀφθαλμοῖς (Jdt 3:4), ἐκ χειρός (Jdt 16:2), ἐν τῇ ἰσχύι αὐτῆς (Jdt 13:8).

1.3 Pronouns The use of pronouns is various. Although one does encounter some Hebraizing features in the use of pronouns, the Greek text of the book of Judith appears generally to deviate from Hebrew patterns.

1.3.1 Possessive Pronouns The redundant use of the genitive forms αὐτοῦ, αὐτῆς, and αὐτῶν has already been mentioned. Nevertheless, even if these instances are very widespread, at least one instance of a reflexive pronoun in the genitive should be mentioned (Jdt 7:32: ἐσκόρπισεν τὸν λαὸν εἰς τὴν ἑαυτοῦ παρεμβολήν, v.l. εἰς τὴν παρεμβολήν αὐτοῦ). In the majority of the cases, the genitive form of the pronoun is placed after the noun it qualifies. In some scattered cases however it is placed before the noun, a practice that does not reflect standard Hebrew word order: ῥάξον αὐτῶν τὴν ἰσχύν (Jdt 9:8), θραῦσον αὐτῶν τὸ ἀνάστεμα (Jdt 9:10), ὁ θεός σου ἔσται μου θεός (Jdt 11:23), αὐτῶν ἡ ψυχὴ […] καὶ […] αὐτῶν κραυγή (Jdt 14:19). 26 1.3.2 Personal Pronouns In two cases, personal pronouns occur in prepositional attributives placed between the article and the noun: ἐτείχισαν τὰς ἐν αὐτοῖς κώμας (Jdt 4:5), ἀναμένοντες τὴν παρ᾽ αὐτοῦ σωτηρίαν (Jdt 8:17). Such a word order, very common in Greek, is really highly exceptional in the Septuagint and does not fit Hebrew syntax. See, e. g., Isa 59:21: ἡ παρ᾽ ἐμοῦ διαθήκη; Jer 11:20; 20:12: τὴν παρὰ σοῦ ἐκδίκησιν. 1.3.3 Reciprocal Pronouns The reciprocal relationship is expressed by ἕκαστος πρὸς τὸν πλησίον αὐτοῦ (Jdt 7:4; 10:19) instead of πρὸς ἀλλήλους. 27 In Hebrew, ἕκαστος πρὸς τὸν πλησίον αὐτοῦ would correspond to ‫ את רעהו‬/ ‫( אישׁ אל‬e. g. Ps 11:3LXX; Jonah 1:7). 24. BDF, § 217. 25. Dorival, Harl, and Munnich, Bible grecque, 239. 26. In several cases, the pre-position of the adnominal genitive can be found (BDF, § 271): ὃν τρόπον πολέμου πλῆθος συντάσσεται (Jdt 2:16), ταπεινῶν εἶ θεός ἐλαττόνων εἶ βοηθός […] ἀπεγνωσμένων σκεπαστής ἀπηλπισμένων σωτήρ (Jdt 9:11). 27. See, e. g., H. St. J. Thackeray, A Grammar of the Old Testament in Greek, Cambridge 1909 (reprint Hildesheim 2003), 45. 1. Syntax

399

25.2 The Language of the Book of Judith

1.3.4 Indefinite Pronouns Although using the pronoun οὐδείς (e. g. Jdt 12:10: οὐκ ἐκάλεσεν εἰς τὴν κλῆσιν οὐδένα), the book of Judith has a preference for ἀνὴρ εἷς or ἄνθρωπος 28 rather than οὐδείς/μηδείς in negative clauses: τοῦ μὴ ἐξελθεῖν ἐκ τῆς πόλεως ἄνδρα ἕνα (Jdt 7:13), οὐ καλόν ἐστιν ὑπολείπεσθαι ἐξ αὐτῶν ἄνδρα ἕνα (Jdt 10:19), οὐκ ἐκάκωσα ἄνθρωπον (Jdt 11:1), οὐκ ἦν ἄνθρωπος (Jdt 15:2). 29

2. Vocabulary 2.1 Hebraizing Words and Phrases 2.1.1 Demonyms The use of υἱοί followed by the proper name of a country is very frequent: the Midianites are called υἱοὶ Μαδιαμ (Jdt 2:26), the Ammonites υἱοὶ Αμμων (Jdt 5:5), the Moabites υἱοὶ Μωαβ (Jdt 6:1), even the Titans are υἱοὶ τιτάνων (Jdt 16:6). 30 Only once, θυγάτηρ occurs in such formulations (Jdt 10:12). However, an individual, Achior for instance, is referred to using an adjective: Αμμανίτης (Jdt 14:5). As for the Assyrians, the book of Judith fluctuates between υἱοὶ Ασσουρ (Jdt 6:17; 7:17; etc.) and Ἀσσύριοι (e. g. Jdt 1:1,7; etc.). 2.1.2 Stereotyped Use of Certain Verbal Forms Noteworthy is the frequent use of ἰδού in direct discourse, either to introduce it or to mark a turning point (Jdt 2:5; 3:2-4; etc.). Needless to say, this imperative form corresponds to the Hebrew ‫הנה‬. Likewise, the optative γένοιτο, which expresses approval (Jdt 13:20; 15:10), has a Hebrew equivalent in the Septuagint: ‫( אמן‬Num 5:22; etc.). 2.1.3 Anthropological Vocabulary The noun ψυχή is used in the manner of a reflexive pronoun (Jdt 4:9: ἐταπείνωσαν τὰς ψυχὰς αὐτῶν “they humbled their souls” = “they humbled themselves”; see also Lev 16:29,31; Isa 58:3,5) or to designate individual human beings: σέσωκας τὴν ψυχήν σου (Jdt 10:15), εἰς κατόρθωσιν πάσης ψυχῆς (Jdt 11:7), οὐκ ἐφείσω τῆς ψυχῆς σου (Jdt 13:20). Furthermore, a Hebrew meaning underlies the expression, in Jdt 2:3, πᾶσα σάρξ “all flesh,” which is to be understood as “everybody.” Conversely, in Jdt 10:13, οὐ […] σὰρξ μία οὐδὲ πνεῦμα ζωῆς, means “nobody.” A Hebrew influence can also be seen in Judith 13:4: standing before Holophernes’ bed, Judith prays “in her heart” (εἶπεν ἐν τῇ καρδίᾳ αὐτῆς), i. e. silently, without making any noise. Elsewhere in the Septuagint, “to speak in one’s heart” means “to think” (e. g. Deut 8:17), 31 however the closest parallel is Hannah’s silent prayer, in 1Reigns 1:13: καὶ αὐτὴ ἐλάλει ἐν τῇ καρδίᾳ αὐτῆς καὶ τὰ χείλη αὐτῆς ἐκινεῖτο καὶ φωνὴ αὐτῆς οὐκ ἠκούετο. At times 28. Thackeray, Grammar, 45. 29. Another interesting example is Judith 7:21: οὐκ εἶχον πιεῖν […] ὕδωρ ἡμέραν μίαν, which has μίαν where οὐδεμίαν would be expected. 30. For the latter example, see Gera, Judith, 81. 31. T. Muraoka, A Greek-English Lexicon of the Septuagint, Leuven 2009, 363.

400

2. Vocabulary

25.2 The Language of the Book of Judith

however, instead of following the usage of the other Septuagint books, the Book of Judith sometimes goes its own way. To speak of the physical beauty of a person, the Hebrew Bible uses an expression the Septuagint normally renders by καλὸς/καλὴ τῷ εἴδει (e. g. Gen 29:17; 39:6); in Judith 11:23, however, a slightly different phrase appears: ἀστεία εἶ σὺ ἐν τῷ εἴδει σου. Likewise, Susanna is characterized as a woman ἀστεία τῷ εἴδει (Sus 1:7).

2.1.4

Hebrew Expressions in Greek Garments

In order to congratulate Judith people speak “peace” with her (Jdt 15:8: λαλῆσαι μετ᾽ αὐτῆς εἰρήνην). This expression is reminiscent of some scattered examples in the Septuagint, e. g. Ps 27:3: τῶν λαλούντων εἰρήνην μετὰ τῶν πλησίον αὐτῶν, where εἰρήνη is the equivalent of the Hebrew noun ‫שׁלם‬. 32 Nevertheless, in this instance the book of Judith avoids a very common word verb like μακαρίζω (Gen 30:13; etc.). Another example that comes quite close to Judith 15:8 is Judith 7:24, where the adjective εἰρηνικά takes the place of the noun εἰρήνη: οὐ λαλήσαντες εἰρηνικὰ μετὰ υἱῶν Ασσουρ. In this case, λαλέω εἰρηνικά refers to political measures meant to bring about peace for the Israelites. This expression is not a unique case in biblical literature. In Psalms 34:20 and in Jeremiah 9:7 persons are said to speak “peaceful things” (εἰρηνικά), even as they plot against others. Furthermore, two examples of typically biblical Greek expressions should be mentioned. The expression ὑψόω τὴν φωνήν (Jdt 16:11) is to be found quite often in the Septuagint as a stereotyped rendering of a corresponding Hebrew phrase (see, e. g., Gen 39:15,18; Isa 13:2; 37:23; 58:1; Ezek 21:27). Mutatis mutandis, the same applies for another example related to the death of Judith’s husband: ὁ ἀνὴρ αὐτῆς […] προσετέθη πρὸς τὸν λαὸν αὐτοῦ (Jdt 16:22). Quite identical formulations, which are indeed calques from the Hebrew, can be found, in Genesis 25:8 and 49:33, in the context of the death of the patriarchs Abraham and Jacob. The situation is slightly different in the case of the expression ἐν στόματι ῥομφαίας “mouth of the sword” (Jdt 2:27), which designates its edge. Identical or similar examples occur in other biblical texts, e. g. Jos 6:21: ἐν στόματι ῥομφαίας, where the correspondig Hebrew text reads ‫לפי‬, literally “with the mouth,” for ἐν στόματι. Nevertheless, this particular use of στόμα for the edge of a weapon is not limited to biblical texts translated from Hebrew (see, e. g., Luke 21:24). It occurs as well in non-biblical Greek (see, e. g., Homer, Iliad, 15.389).

2.2 Ritual, Cultic and Theological Terminology 2.2.1 Sacrifice The book of Judith employs some words which refer to ritual issues. Two terms of the typical Septuagint sacrificial vocabulary appear in Judith 16:16-18: ὁλοκαύτωμα (Jdt 16:16,18) and τὰ ἑκούσια (Jdt 16:18; see already Jdt 4:14). No doubt the “burnt offerings,” as well as the “freewill offerings,” are to be understood against the backdrop of the Pentateuch sacrificial prescriptions, where the same Greek terminology occurs (e. g. Num 15:3; 29:39). Nevertheless, while these two instances reflect the characteristic 32. See also Corley, “Septuagintalisms,” 79. 2. Vocabulary

401

25.2 The Language of the Book of Judith

biblical use of the sacrifice terminology, the case of ἀνάθημα in Judith 16:19 is quite different. Dedicating Holofernes’ precious canopy to God as ἀνάθημα, Judith does not consider it an object to be destroyed (for this biblical use of ἀνάθημα, see, e. g., Josh 6:17-18), but suited to adorn the temple. In this respect, Judith 16:19 appears to reflect non-biblical rather than biblical connotations of ἀνάθημα. Indeed, in non-biblical texts, this noun refers to different sorts of more or less precious objects laid down by worshippers in a temple. 33

2.2.2 Blessings When speaking of blessing, the Septuagint employs a standard word: εὐλογέω. A closer analysis of the use of εὐλογέω in the book of Judith yields an interesting result. In Judith 15:9,12, εὐλογέω retains its non-biblical meaning “to speak well of somebody,” “to praise somebody,” 34 Judith in this case. The use of the verbal adjectives εὐλογημένος and εὐλογητός however requires greater attention. In the wake of the Septuagint usage of these words, Judith as a human being can be qualified as εὐλογητή (Jdt 13:18, v.l. εὐλογημένη), εὐλογημένη (Jdt 14:7; see also Ruth 2:19; 3:10), or even as εὐλογημένη […] παρὰ τῷ παντοκράτορι κυρίῳ (Jdt 15:10). Furthermore, the use of this terminology is not limited to humans 35; likewise, God is referred to as εὐλογητός (Jdt 13:17) or as εὐλογημένος (Jdt 13:18). From the point of view of non-biblical Greek, this use of εὐλογέω might seem surprising but it is in keeping with many other Septuagint texts (Gen 9:26; 14:20; etc.). 36 Nevertheless, the manuscript evidence is not clear enough to suggest that the one term is used for God and the other for humans. 37 2.2.3 Achior’s Conversion to Judaism In order to join the Jewish community, the former enemy officer Achior, a Ammonite (see above), must undergo circumcision (Jdt 14:10: περιετέμετο τὴν σάρκα τῆς ἀκροβυστίας αὐτοῦ). The Greek terminology is identical with Genesis 17:11 and 14, etc. Conversely, it turns out to be difficult to find more than a handful of Septuagint examples of another term used in Judith 14:10, the verb πιστεύω. In fact, Achior’s relationship to God is described in terms of “believing” (Jdt 14:10: ἐπίστευσεν τῷ θεῷ σφόδρα). To be sure, the Septuagint employs roughly the same terminology in the case of Abraham (Gen 15:6: καὶ ἐπίστευσεν Αβραμ τῷ θεῷ). Nonetheless, the specific use of πιστεύω in the context of conversion is quite exceptional in the Sep33. For further details, see K. Berthelot, “The Notion of Anathema in Ancient Jewish Literature Written in Greek” in: E. Bons / R. Brucker / J. Joosten (eds.), The Reception of Septuagint Words in Jewish-Hellenistic and Christian Literature (WUNT II/367), Tübingen 2014, 35-52, here 47. 34. See also R. J. Ledogar, “Verbs of Praise in the LXX Translation of the Hebrew Canon”, Bib 48 (1967), 29-56, here 52; J. Joosten, “Mixed Blessings. The biblical notion of blessing in the works of Philo and Flavius Josephus” in: Bons / Brucker / Joosten, Reception of Septuagint Words, 105-115, here 109. 35. See also Ledogar, “Verbs of Praise,” 51. 36. J. Joosten, “Le vocabulaire de la Septante et la Question du Sociolecte des Juifs Alexandrins. Le Cas du Verbe εὐλογέω, ‘Bénir’” in: E. Bons / J. Joosten (eds.), Septuagint Vocabulary: PreHistory, Usage, Reception (SCSt 58), Atlanta GA 2011, 13-23, here 18-19. 37. M. Cimosa, Guida allo studio della Bibbia greca (LXX). Storia, lingua, testi, Rome 1995, 129.

402

2. Vocabulary

25.2 The Language of the Book of Judith

tuagint and in Jewish Hellenistic literature as a whole (see also Wis 12:2; Philo, Abr. 69). 38

2.2.4 Speaking about God To begin with, attention should be given to two divine titles: παντοκράτωρ and κτίστης. God is a παντοκράτωρ (Jdt 4:13; 8:13; 15:10; 16:5,17), a divine epitheton characteristic of the Septuagint and later Jewish and Christian literature. 39 Moreover, he is a κτίστης in the sense of “creator.” This noun, which undergoes a change of meaning from non-biblical Greek (“founder”) to biblical Greek (“creator”), occurs in a series of divine titles in Judith’s prayer (Jdt 9:12). As such, God is called the king of the κτίσις (Jdt 9:12), which must serve him (Jdt 16:14). Accordingly, when speaking about God’s function as creator, the verb employed is κτίζω (Jdt 13:18). In this respect, the book of Judith adopts a terminology that appears in the earliest translated books of the Septuagint (e. g. Deut 4:32: κτίζω; 2Reigns 22:32: κτίστης). 40 The same holds true for other divine titles such as βοηθός 41 (Jdt 9:11, see also Exod 15:2; Ps 9:10; 45:2) and ἀντιλήμπτωρ (Jdt 9:11), which occur together in Psalm 17:3 and Psalm 58:18. Finally, the verb ὑπερασπίζω should be mentioned. In Genesis 15:1, God promises Abraham to “hold his shield over him” (ἐγὼ ὑπερασπίζω σου). The same verb, which is attested more than a dozen of times in the translated books of the Septuagint (e. g. Deut 33:29; Ps 19:2; Prov 2:7), appears three times in Judith (5:21; 6:2; 9:14). In each of these instances, God is the subject, and Bethulia or Israel the object of the divine protection. 2.2.5 Speaking of Other Gods The book of Judith mentions other gods only once referring to them as θεοὶ χειροποίητοι (Jdt 8:18), a typical Septuagint word denoting foreign gods (e. g. Lev 26:1; Isa 46:6). 42 To sum up, the abovementioned examples illustrate that the book of Judith is deeply influenced by the specific ritual and theological vocabulary of the Septuagint. However, this result should not hide the fact that this vocabulary is not completely identical with Septuagint vocabulary. A good example is the use of πιστεύω in Judith 14:10 (see above).

38. G. Barth, art. πίστις, πιστεύω in: H. Balz / G. Schneider (eds.), Exegetisches Wörterbuch zum Neuen Testament (second edition), vol. III, Stuttgart 1992, col. 216-231, here 218. 39. M. Bachmann, Göttliche Allmacht und theologische Vorsicht: zu Rezeption, Funktion und Konnotationen des biblisch-frühchristlichen Gottesepithetons pantokrator (SBS 188), Stuttgart 2002, 152. 40. For further details, see E. Bons / A. Passoni Dell’Acqua, “A Sample Article: κτίζω – κτίσις – κτίσμα – κτίστης” in: Bons / Joosten, Septuagint Vocabulary, 173-187. 41. For this divine title, see E. Bons, “The Noun βοηθός as a Divine Title. Prolegomena to a Future HTLS Article” in: E. Bons / R. Brucker / J. Joosten (eds.), The Reception of Septuagint Words in Jewish-Hellenistic and Christian Literature (WUNT II/367), Tübingen 2014, 5366. 42. Engel, “Zur Frage der griechischen Originalsprache,” 158. 2. Vocabulary

403

25.2 The Language of the Book of Judith

2.3 Biblical and Non-Biblical Greek Vocabulary 2.3.1 Characteristic Septuagint Words in the Book of Judith In addition to the examples given above, the book of Judith proves to be familiar with certain terms belonging to the vocabulary of the Septuagint. It will suffice to mention three of them: 1. The only noun or adjective used to denote strangers is ἀλλογενεῖς, a typically Septuagint word (e. g. Exod 12:43; Isa 56:3) absent from contemporary non-biblical texts. In Judith 9:2, ἀλλογενεῖς probably refers to the Sichemites, held collectively responsible for the rape of Dinah (see Gen 34). 2. The inhabitants of a town or a country are called κατοικοῦντες (e. g. Jdt 4:6; 7:13,20; 8:11; see also Gen 14:7; 19:25), whereas the noun πολῖται is missing. 3. Another characteristic Septuagint noun is σκάνδαλον (see, e. g., Lev 19:14; Ps 48:14), which can designate a material obstacle (Jdt 5:1), a sin that could potentially prevent the God of Israel from intervening in favor of his people (Jdt 5:20), or a possible transgression of the purity laws (Jdt 12:2). 2.3.2 Typical Greek Words Used in the Book of Judith It is noteworthy that several very common Greek adjectives and verbs infrequent or unknown in the remaining books of the Septuagint appear in the book of Judith. 1. As for the adjectives, in Judith 8:31 the protagonist is called εὐσεβής: ὅτι γυνὴ εὐσεβὴς εἶ. This very frequent Greek adjective is rare in the translated books of the Septuagint (e. g. Isa 24:16). For a person to be explicitly qualified as “pious” is exceptional in the biblical literature. 43 Somewhat more frequent is the adjective ἀγαθός, used to refer to the virtues of a human being—Judith seeks to gain Holophernes’ favor by calling him ἀγαθός (Jdt 11:8). However, the biblical instances of persons being qualified explicitly as ἀγαθός are exceedingly rare, Saul (1Reigns 9:2), for instance. 2. As for the verbs, the book of Judith sometimes employs very common Greek verbs that are sporadic or completely missing in the translated books of the Septuagint, e. g. ἀφικνέομαι (Jdt 1:14). Elsewhere in the Septuagint, verbs like παραγίνομαι, in the sense of “to arrive” (e. g. Gen 50:10; Ruth 1:19,22), or εἰσπορεύομαι (e. g. 2Reigns 15:37) are much more frequent. Similarly, the verb κελεύω is used in Judith 2:15 and 12:1, whereas the Septuagint translators have a preference for verbs like ἐντέλλομαι (Lev 24:2; etc.). 2.3.3 Technical Terms Various technical terms denoting objects of utility absent from the remaining books of the Septuagint appear in the book of Judith: τὸ κωνώπιον “canopy” (Jdt 13:9,15; 16:19), 44 ἡ πήρα “knapsack, traveler’s bag” (Jdt 10:5; 13:10,15), ἡ χελωνίς “footstool” (Jdt 14:15). Therefore, it would be impossible to argue that their use is at least partly 43. See also M. Wieger, “Eὐσέβεια et ‘crainte de Dieu’ dans la Septante” in: Bons / Joosten, Septuagint Vocabulary, 101-156, here 101. 44. See B. Schmitz, “Holofernes’s Canopy in the Septuagint” in: K. R. Brine / E. Ciletti / H. Lähnemann (eds.), The Sword of Judith. Judith Studies across the Disciplines, Cambridge 2010, 71-80.

404

2. Vocabulary

25.2 The Language of the Book of Judith

determined by an underlying Hebrew text. However, in regard to the possibility of a Hebrew source text for the book of Judith, one noun is of particular interest: ὁ ἀκινάκης “scimitar,” a word of Persian origin used in Judith 13:6 and 16:9. If we assume that this Persian word was first translated into Hebrew and then into Greek, it seems improbable that it would appear in its correct Hellenized form in the extant Greek manuscripts. 45

Concluding Remarks In the last two decades some scholars have argued that the book of Judith reflects an in-depth knowledge of the Septuagint. Not only syntactical features, but vocabulary can be adduced to support this hypothesis. In some cases the book of Judith quotes, or alludes to, Greek biblical texts whose Septuagint version diverges more or less obviously from the extant Hebrew texts. 46 This is the case in Judith 8:16 (ὅτι οὐχ ὡς ἄνθρωπος ὁ θεὸς ἀπειληθῆναι οὐδ᾽ ὡς υἱὸς ἀνθρώπου διαιτηθῆναι, a quotation of Num 23:19), Judith 9:2 (οὐχ οὕτως ἔσται, a quotation of Gen 34:7), Judith 11:19 (οὐ γρύξει κύων τῇ γλώσσῃ αὐτοῦ, a quotation of Exod 11:7), Judith 16:14 (ὅτι εἶπας καὶ ἐγενήθησαν, an allusion to Ps 32:9: ὅτι αὐτὸς εἶπεν καὶ ἐγενήθησαν), and Judith 9:7 and 16:2 (κύριος συντρίβων πολέμους / θεὸς συντρίβων πολέμους, a quotation of Exod 15:3). If the book of Judith was a translation from a lost Hebrew text, we would expect the translator to have rendered these quotations and allusions according to the Hebrew biblical text. Admittedly, in itself this argument is not sufficient to conclude that the original language of the book of Judith was Greek. 47 Nevertheless, many syntactical and terminological features mentioned above prove difficult to explain if we assume that the Greek text is a translation of a lost Hebrew source. Therefore, these syntactical and terminological observations can be taken as converging indications in favor of a hypothesis capable of challenging the prevailing theory of the Hebrew origin of the book of Judith. On the basis of these arguments, it seems possible to contend that the book of Judith was, from the outset, a Greek composition. As for the various elements of Hebraizing language, they can be attributed to “[…] a Greek author, intending to create a ‘biblical’ story, [who] adopted the biblical style he knew from the Septuagint.” 48 In other words, what at first sight appears to be “Hebraizing” in the Greek style of the book of Judith is in reality “Septuagintizing.” However, this hypoth45. For this argument, see H. Y. Priebatsch, “Das Buch Judith und seine hellenistischen Quellen”, ZDPV 90 (1974), 50-60, here 53. 46. See also Engel, “Zur Frage der griechischen Originalsprache,” 157-158; Joosten, “Language and Milieu,” *164-165; Schmitz / Engel, Judit, 42. 47. See also Hanhart, Text und Textgeschichte, 98-99, note 2, who argues that the book of Judith is a translation from a Hebrew or Aramaic text. According to this author, the presence of Septuagint quotations and allusions is to be explained by the translator’s preference for the Septuagint text, which Hanhart supposes to be “canonical” in his milieu. 48. Joosten, “Language and Milieu,” *163. See also Corley, “Septuagintalisms,” 80, who states— with respect to the expression “sons of” denoting the members of a people—that “its usage in Judith may be intended to evoke the LXX and add a biblical sonority to the narrative.” He concludes (87): “In many cases Judith seems to be imitating biblical style, as found in the Concluding Remarks

405

25.2 The Language of the Book of Judith

esis cannot answer another question: Was the Judith narrative available in ancient Aramaic or Hebrew versions that might possibly have diverged from the Septuagint book of Judith? This question still remains open. 49

LXX. Often such constructions also appear in NT writings originally composed in Greek, and the Semitic influence does not necessitate a previous composition in Hebrew or Aramaic.” 49. See also Schmitz, “Ιουδιθ und Iudith,” 363; Gera, Judith, 79.

406

Concluding Remarks

25.3 The Language of Second Maccabees Frank Shaw

1. The Introductory Missives Given the composite character of Second Maccabees, one cannot carelessly generalize about its linguistic nature or literary quality. Being translation Greek of lost Hebrew or Aramaic originals, the two prefatory letters are of a caliber different from the epitome which is an original Greek composition, albeit mainly in a summarized version. The first epistle (1:1-10a), a mere 184 words, looks similar to most of the translated books of the Septuagint: almost complete parataxis (twenty-seven instances of καί), a lone subordinate clause (verse 7), and a sparing use of participles (only four). A few things still stand out, however, including seven optatives of wishing in verses 2-5, an ambiguous use of ἐν in verse 4 (common in Hellenistic Greek: “to” or “by means of”?), and the imperatival ἵνα + subjunctive (verse 9). The second introductory letter (1:10b–2:18) is far more interesting, due to both its size and complexity. Daniel R. Schwartz notes that it is “more convoluted” than the first epistle. 1 His comment focuses on content rather than Greek, but it is just as true on the linguistic level. Three features are prominent: the intensive use of ὡς in a variety of ways, a fondness for correlatives, and especially strained syntax. There are seventeen simple occurrences of ὡς. 2 Eight instances are temporal, “when”; 3 five are synonyms for ὅτι, “that”; 4 two connote “as” (1:23; 2:1); one, “as if/as though” (1:14). The lone example of a causal employment of the conjunction appears in the form of ὡς ἄν + participle in 1:11, a construction rare in the classical period (and not with causal import) but common in literary koine. In addition to plain ὡς two compounds of the word occur: καθώς (quater, 1:29,31; 2:10,18), and ὡσαύτως (bis, 2:12,14). Correlatives line up as follows: τε […] καί (quater, 1:14,18,21,23); καθώς […] οὕτως (2:10); and ὅτε […] τότε (1:19). The use of ὡς καί […] ὡς καί at 2:8 may also be correlative. In contrast to the main portion of 2Maccabees, the second letter’s translator never utilizes μέν […] δέ, probably a sign of its being a translation. The frequent difficult syntax takes diverse forms. Simple examples include 1:31-32 where a preposition needs to be added in translation to λίθους μείζονας καταχεῖν “to pour on very large stones,” and the subject of ἐδαπανήθη in the sentence is unclear. On the last point Jonathan A. Goldstein suspects that words have disappeared in textual transmission, 5 and given the context (v. 36), Robert Doran supplies “naphtha” 1. 2. 3. 4. 5.

D. R. Schwartz, 2 Maccabees (Commentaries on Early Jewish Literature), Berlin/New York, NY 2008, 133. This figure does not include 1:18, which is textually questionable. 1:15,21,22,32,33; 2:7,8bis. 2:2,4bis,9,13. J. A. Goldstein, II Maccabees (AB 41A), New York, NY 1983, 155, 180. 1. The Introductory Missives

407

25.3 The Language of Second Maccabees

as the verb’s subject. 6 The last participial construction at 1:34 appears hastily thrown in, and the next verse seems strangely truncated, perhaps because of textual issues. B. Risberg 7 emends the ἐχαρίζετο to ἐξηρύσαντο, accepted by Schwartz. 8 MS A has no verb here. Christian Habicht declares that the generally printed text “ist in dieser Form jedenfalls korrupt.” 9 The second letter’s end contains a lengthy sentence that has no stated main verb (2:17-19). James Moffat, 10 Félix-Marie Abel, 11 and Schwartz 12 understand anacoluthon here. Goldstein believes that the verb ἀποδόσει has fallen out of the text. 13 The scribe of MS 58 was evidently so disturbed by the lack of a stated main verb that he excised the words ἐλπίζομεν γὰρ ἐπὶ τῷ θεῷ ὅτι and thus provided ἐλεήσει as a verb for the previous ὁ δὲ θεὸς ὁ σώσας. A less surgical solution is to understand the verb ἐστίν with ὁ θεὸς as subject and the ὁ σώσας not as an ordinary attributive participle but as a full predicate. 14 This then means that the following first γάρ clause is a large parenthetical one consisting of twenty words inserted into the main clause, and that the main clause also has a second γάρ clause finishing it. Long parenthetical γάρ clauses like this are found in Aristotle, Polybius, Josephus, and Lucian, and such a parenthetical γάρ clause followed by a simple causal γάρ clause is present in Polybius 4.34.1. According to LEH the second letter contains sixteen Septuagintal hapax legomena and two or three neologisms. Fifty-seven participles are present, and moderate variety in usage and syntax is found in their employment. Variatio is evident in the two ways the translator chooses to express necessity: the accusative absolute construction δέον [ἐστίν] at 1:18 and χρείαν ἔχω at 2:15. To indicate purpose/intent the translator combines the future participle (1:4; 2:15), the infinitive, 15 ἵνα + the subjunctive (1:18; 2:2), and a noun phrase with a preposition (1:20). Periodic structure is found only at 1:14-16. His passion for conjunctive ὡς is occasionally mollified by ὅτε. 16 Three result clauses with ὥστε are present (1:19,22; 2:6). There is one well-known ambiguity, what ὕδωρ παχύ in 1:20 means. Possibilities include “marsh water,” 17 “(a) thick liquid,” 18 “viscous

6. R. Doran, Temple Propaganda: The Purpose and Character of 2 Maccabees (CBQMS 12), Washington, DC 1981, 8. 7. B. Risberg, “Textkritische und exegetische Anmerkungen zu den Makkabäerbüchern” Beiträge zur Religionswissenschaft 2 (1915), 17-18. 8. Schwartz, 2 Maccabees, 158. 9. C. Habicht, 2. Makkabäerbuch (second edition, JSHRZ 1.3), Gütersloh 1979, 205, note 35a. 10. J. Moffat, “The Second Book of Maccabees” APOT 1 (1913), 125-154. 11. F.-M. Abel, Les Livres des Maccabées (EBib), Paris 1949, 309-310. 12. Schwartz, 2 Maccabees, 132. 13. Goldstein, II Maccabees, 187. 14. Articular participial predicates are by no means rare: οἱ ἄνδρες εἰσὶν οἱ ποιοῦντες “the men are the ones performing,” Xen. Anab. 3.2.18; see also Plato, Phd. 96B; 2Mac 15:14; John 6:63. Cf. SHM § 1152. 15. Articular with the “improper” preposition, χάριν τοῦ λαβεῖν, 1:14; μέλλω + an infinitive, 1:18 and 2:16. 16. 1:18,19 (correlative with τότε), 20. 17. Habicht, 2. Makkabäerbuch, 204; LEH, 363. 18. Moffat, “The Second Book,” 133; Abel, Livres des Maccabées, 293; NRSV.

408

1. The Introductory Missives

25.3 The Language of Second Maccabees

water/liquid,” 19 and “oily liquid.” 20 The last three renderings connect this substance with the naphtha discussed in verse 36. It is certain that a single translator is not responsible for the rendition of both epistles, for they are just too different in style. Though quite short, the first missive contains enough data to indicate that it was translated by someone whose linguistic skills were similar to those of the translators of the Septuagint Pentateuch; the second letter was the work of a more highly educated person, though certainly not on a par with the authors of the epitome.

2. The Epitome By far most of the previous linguistic study on Second Maccabees has been on the epitome, but it has still been quite limited. Prior to my recent study, 21 the most detailed survey was that of Doran, and he makes no pretense to being comprehensive. His work is designed to place the epitome within its Hellenistic context, both literarily and historically. Even my own work, certainly the most detailed investigation into the language of the entire apocryphal book to date, regularly highlights various linguistic desiderata, and makes several statements disavowing any claim of being a complete analysis of 2Maccabees’ language. The fairly often cited article of Charles Mugler, 22 a mere four-page study comparing the number of participles in Second Maccabees with a few other ancient works, is fraught with methodological problems and should never be used to say anything about the participial construction in our work. 23 In order to proceed in any meaningful way with a look at the epitome’s Greek, a bit of a fiction needs to be employed. Since, with few exceptions, 24 the words of the epitomizer cannot be disentangled from those of Jason, the only procedure is to view the work as a single unit. Thus the terms “author, writer, composer” below refer to the document as we have it. 25 In several ways the epitome merits the long-held claims that it is among the few Septuagint works composed in Greek that form its highest literary tier (along with Third and Fourth Maccabees and perhaps the Wisdom of Salomon). Most prominent here is the work’s wide-ranging variatio (or μεταβολή) in vocabulary. LEH records 375 Septuagintal hapax legomena in it, fifty-five “confirmed” neologisms, 26 and twenty-one further probable ones. Adverbs with the ending -ηδόν are found six times 19. H. Bévenot, Die beiden Makkabäerbücher (HSAT IV. 4), Bonn 1931, 174; Goldstein, II Maccabees, 155, 177; Schwartz, 2 Maccabees, 130, 153. 20. TEV. 21. F. Shaw, A New Direction in 2 Maccabean Studies: Text, Language, and Style (forthcoming). 22. C. Mugler, “Remarques sur le second livre des Macchabées” RHPR 11 (1931), 419-423. 23. Shaw, New Direction, chapter 1. 24. The introduction (2:19-32), the conclusion (15:38-39), and most likely a few summary statements (4:17; 5:17-20; 6:12-17; 7:42) are the work of the epitomizer. All such further claims are debatable. 25. The five inserted letters (9:19-27; 11:16-21, 22-26, 27-33, and 34-38) are excluded from the analysis given here. 26. On the establishment of neologisms, see LEH, xiv, xxiv. Doubt always exists here; hence “confirmed.” 2. The Epitome

409

25.3 The Language of Second Maccabees

in four forms, more so than in any other Septuagint book. 27 Our author is fond of employing the prefixes δυσ- and εὐ- to indicate negativity and positiveness, again the former being more heavily used than in other Septuagint works. 28 He regularly utilizes synonyms when describing the same phenomena in close proximity: e. g., different verbs for “sell” in the same verse, πωλέω and πιπράσκω (8:14); at 9:5 variations for “incurable,” ἀνιάτος and ἀνήκεστος; and two verses later for “arrogance,” ἀγερωχία and ὑπερηφανία; even “be” is so varied at 14:44-45, γίνομαι, ὐπάρχω, εἰμί in close sequence. Over larger portions of the work we see the same: for “precipitous,” ἀπόκρημνος (13:5) and ἀπορρώξ (14:16,45); for rejoicing/happiness, μακαριστός at 7:24, εὐφροσυνή at 10:6, and χαρά at 15:28. This interest in μεταβολή holds true not only in vocabulary, but more significantly in syntax, where these two categories sometimes merge. The epitome contains the most frequent and ornate usage in the Septuagint of μέν […] δέ to show contrast. 29 It is employed very simply: ὁ μὲν ᾽Ιάσων […] ὁ δὲ Μενέλαος (4:26-27), οἱ μὲν […] ἡμεῖς δέ (8:18), πρῶτον μὲν […] δεύτερον δέ (14:8); and with a bit more sophistication: τῇ μὲν ἐμφάσει […] τῷ πράγματι δέ “in pretext […] but in fact” (3:8), αἰδήμονα μὲν τὴν ἀπάντησιν, πραῢν δὲ τὸν τρόπον “venerable in appearance, but gentle in manner” (15:12). But most outstanding are its various complex uses. Our writer specifies a large area with the μέν and then a component part (or parts) with the δέ: τὸ μὲν γὰρ ἱερόν […] ἔτι δὲ τὰ μὴ καθήκοντα […] τὸ δὲ θυσιαστήριον “the temple […] the unfit [offerings] […] the altar” (6:4-6; cf. the city and the temple, 11:23; the city and its inhabitants, 9:14-15). One μέν […] δέ clause occurs within another (5:22-23) or repeatedly in rapid succession: four times in five verses (4:25-29). At least four times he employs μέν to close out one narrative section and then δέ to introduce another, 30 a practice dating to Herodotus. The work’s most extended utilization of μέν […] δέ […] δέ […] δέ is at 11:16,22,27,34 where our author links the four documents he quotes in this way. Otherwise contrast is shown with ἀλλά, found twelve times, 31 ὅμως thrice, 32 occasionally an adversative καί, 33 and καίπερ once (4:34). Concessive participles occur, both as dependent modifiers of elements within clauses 34 and as genitive absolutes. 35 To show cause our author has chosen γάρ forty-three times, once as τοιγαροῦν (7:23), and once as καὶ […] γάρ (6:13). He is extremely fond of causal διά + an accu27. ἀγεληδόν, 3:18 and 14:14; σπειρηδόν, 5:2 and 12:20; λεοντηδόν, 11:11; κρουνηδόν, 14:45. Elsewhere only in WisSal 18:23 (σωρηδόν) and 4Mac 2:19 (ἐθνηδόν), 15:19 (ταυρηδόν). λεοντηδόν, κρουνηδόν, and ἐθνηδόν are neologisms; σπειρηδόν and σωρηδόν are suspected as such. See the appropriate LEH entries. 28. Of thirty-two such words in the Septuagint, fourteen occur in Second Maccabees. The next nearest this in frequency is Third Maccabees, with ten. 29. There are fifty-one occurrences of it, followed by Fourth Maccabees, twenty-eight, and Wisdom of Salomon, twenty-six. 30. 3:40–4.1; 7:42–8.1; 10:9-10; 15:24-25. 31. 4:14,17; 5:19bis; 6:12,13,26,31; 7:24,29; 8:15; 10:4. 32. 2:27; 14:18; 15:5; elsewhere in the Septuagint only in WisSal 13:6 and 4Mac 13:27 and 15:11. 33. 4:19; 15:5; perhaps also 12:35. 34. E. g., 2:21; 6:16,30; 7:16,20; 14:35. 35. E. g., 9:4; 12:3,35; 14:45bis.

410

2. The Epitome

25.3 The Language of Second Maccabees

sative, so much so that it is a mannerism, probably from the popular speech of his nonliterary koine environment (twenty-three times, not including the formulas discussed below). Sometimes the preposition is combined with αἰτία in the term δι᾽ ἣν αἰτίαν; once more fully, διὰ τήνδε τὴν αἰτίαν. 36 Seven times he employs διό, 37 with another five instances as the strengthened διόπερ. 38 In addition, the author often uses an articular infinitive as the object of this preposition. 39 Other prepositions designating cause include ὑπέρ, ἕνεκα/ἕνεκεν, ὑπό, περί, and χάριν. 40 He conveys cause numerous times with a participle, again both within a grammatically connected clause and as a genitive absolute. 41 The dative of cause is not absent, 42 causal ὡς is used thrice (6:29; 7:12,23), ἐπεί appears twice (13:8; 14:29), but oddly never ἐπειδή 43 or causal ὅτι. When indicating necessity our writer employs ἀναγκάζω + infinitive seven times and a cognate another two, 44 and twice he utilizes δεῖ in the imperfect + infinitive (6:20; 11:18). He never turns to χρείαν ἔχω (as in the second prefatory letter) 45 or the simple χρή, elsewhere merely once in Septuagint Proverbs (25:27). A major surprise is our author’s usage of -τέος verbal adjectives five times to indicate obligation. 46 These are all concentrated in the early part of the epitome: 2:29bis,31; 3:13; 6:17. While four of these are the more common impersonal use, with “subject” datives (technically the dative of agent), dependent infinitives, and the copula ἐστίν understood, the one at 3:13 is the rarer personal construction (and thus passive in meaning), in indirect speech, with the proper form of εἰμί stated, and inflected in the neuter plural to agree with its subject: ὁ δ᾽ ἕτερος […] ἔλεγεν εἰς τὸ βασιλικὸν ἀναλημπτέα ταῦτα εἶναι “the other man […] said that these [deposits] must be confiscated for the royal [treasury].” This is the only such occurrence in the Septuagint. The epitome’s author displays variatio in his expressing finality. Positive purpose clauses with ἵνα + subjunctive occur three times, and their negative counterpart, twice. 47 Other methods include diverse infinitives: of purpose after verbs of motion, 48 the articular infinitive in the simple accusative (12:7) or genitive (5:22) and as the object of the preposition πρός (4:45; 5:27). Present too are a few instances of nouns used to show purpose after εἰς (3:13; 4:19bis; 15:23). περί and its object at 12:43 seem to have a 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

4:28,35,42,49; 8:26; with τήνδε 12:40. 5:4,17; 9:2,4; 10:7,30bis. 5:20; 6:16,27; 7:8; 14:19. 3:18,38; 4:19,30; 6:11,25,29; 8:36; 10:13 (a large example with two infinitives); 15:17. ὑπέρ: 2:21; 4:36; 6:28; 7:9; 8:21; 12:43,44; 14:27,38; 15:30; ἕνεκα/ἕνεκεν: 3:9; 4:20; 8:15; 12:25; ὑπό: 8:25; 9:9; περί: 7:37; 13:25; χάριν: 4:16. E. g., 3:5; 7:9,16,19; 13:3; 14:8bis; 15:21; genitive absolutes: 3:30; 4:39,40; 8:5,27; 11:10; 12:4,7,31,38; 14:28; 15:37bis. At least at 11:4; 14:8; 15:27. According to HR, there are sixteen confirmed instances of ἐπειδή in various other Septuagint books (including four each in Gen, Jer, and 4Mac). 6:1,7,18; 7:1; 8:24; 11:11,14; ἀνάγκη: 6:7; 15:2. And at Dan 3:16 in both versions. Otherwise in the Septuagint only a single time at Prov 26:23 and nine times in the LetJer (twice in verses 39, 44, 56, and 63; once in verse 51). Positive: 6:22,24-25; 7:29; negative: 6:15; 14:22. 6:1; 8:9; 12:39,43. 2. The Epitome

411

25.3 The Language of Second Maccabees

similar function. Lastly, the author of the epitome uses the future participle for this function, spacing out four instances of it: 3:14; 5:9; 10:24; 12:7. At 5:9 and 10:24 the participial clause is introduced by ὡς. These syntactic choices come into play when our composer produces two instances of the same grammatical phenomenon in close proximity. At 12:7 our writer wants to express purpose twice. He first employs the future participle πάλιν ἥξων “intending to come again” and this is immediately followed by an infinitive, καὶ τὸ σύμπαν τῶν ᾽Ιοππιτῶν ἐκριζῶσαι πολίτευμα “in order to root out the entire community of Joppites.” He certainly could have used two parallel subjunctives after a single ἵνα, but the fact is that he did not, and this is not by chance. 4:6 is also a good example. Here he states that Onias saw two future things, that without the king’s attention (1) “it would be impossible (ἀδύνατον εἶναι) for the government to maintain peace” and (2) “Simon would not cease (οὐ λημψόμενον) his folly.” Again, he could have used two purpose infinitives or two participles, as well as two instances of simple οὐ, but he did not. He preferred both semantic (alpha privative, οὐ) and syntactic (infinitive, future participle) variety. Numerous other instances can be found, especially when our author wishes to express cause in close succession. 49 Our composer’s work is uneven when it comes to certain verbal constructions. Although much work remains to be done in this area, some observations can still be made. Unlike Fourth Maccabees the epitome comes up short in optative usage. There are only three (or four) instances of the mood: a wish at 15:24, an indirect question at 3:37, and one (or two) use(s) in simple indirect speech, a single form of εἰμί with two perfect participles—ὁ δὲ […] Σίμων […] ἐκακολόγει τὸν Ὀνίαν ὡς αὐτός τε εἴη τὸν ῾Ηλιόδωρον ἐπισεσεικὼς καὶ τῶν κακῶν δημιουργὸς καθεστηκώς, “but […] Simon […] began to slander Onias saying that he both had terrified Heliodorus and had been the originator of the problems” (4:1). Given that two separate participles are used here, one might say there are two instances of the mood which simply share the same auxiliary verb. When comparing the dearth of the optative in our work to other roughly contemporary authors such as Polybius, Diodorus Siculus, Philo, Josephus, and even Ps.-Aristeas, one sees just how deficient our writer is. Explanations are not easy. Perhaps the lack of direct speech, which, when present, appears in certain parts of Second Maccabees as mainly dialogue, means that few wishes are recorded. Additionally, there are no conditions extant that would require the optative. However, there are certain passages where our author could have utilized the mood in dependent clauses after past tense head verbs/participles, such as purpose or fearing clauses, but failed to do so. 50 The epitome’s composer overuses the imperfect tense, too often mixing it indiscriminately with the aorist. For example, in the account of Heliodorus’ entry into the sanctuary (chapter 3) we see a long string of imperfects (even when Heliodorus finally enters the temple, ἐπετέλει, 3:23, where one might easily expect the aorist) until God’s epiphany in verse 24: ἐπιφάνειαν μαγάλην ἐποίησεν, perhaps for effect. Heliodorus’

49. See 4:20,28; 8:15; 9:2; 12:21bis; 14:28. 50. 3:32; 6:22; 14:22. The five contemporary authors just mentioned did use the optative in such instances, at least at times.

412

2. The Epitome

25.3 The Language of Second Maccabees

companions are then shocked and are turned to fear (aorist participle καταπλαγέντας and aorist infinitive τραπῆναι). When next the horse appears and strikes Heliodorus, we see two aorist indicatives (ὤφθη, ἐνέσεισε), completely suitable. However, the next verb is then imperfect again: “the one sitting on [the horse] with golden panoply ἐφαίνετο” (verse 25). Why the change? Does the punctiliar action suddenly stop? Did he “begin to appear” or are we going back to a process? How could the latter be? Immediately two other young men appear but now back in the aorist (verse 26, προεφάνησαν) for one verb, but then the action continues, and they ἐμαστίγουν Heliodorus. Again, do they “begin flogging” him or what? This inconcinnity occurs elsewhere. At 7:10 we see imperfect-aorist-imperfect in short order. Here the third of the seven martyred sons “was being sported with/tortured” (ἐνεπαίζετο). Suddenly he sticks out his tongue and appropriately an aorist is used for this (τὴν γλῶσσαν αἰτηθεὶς ταχέως προέβαλε), but are his hands then extended as a process (προέτεινε)? Did he “begin to extend his hands” after suddenly sticking out his tongue as a punctiliar action? Thus it is not easy to classify many imperfects as ongoing, conative, inchoative, as denoting customary action, setting up a scene, and so on, mainly because of sloppy overuse. Our work contains a surprisingly high number of pluperfects, a total of nineteen. 51 They are fairly evenly spaced out, though there are no pluperfects in chapters 2, 5, 7, 8, 10, and 11. Of these nineteen almost half are periphrastic. 52 Just what the writer intended with this usage is unclear since often the aorist would have sufficed. Perhaps it is a poor attempt at verbal variatio. Our author is a heavy user of the participle though his employment is, again, sometimes uneven. While the attributive participle is common enough, the supplementary participle both in indirect speech and not in indirect speech is limited. In the case of the latter, usage is especially meager, limited to just two instances and only with verbs indicating continuance. 53 Nowhere is the construction found with the common verbs τυγχάνω, λανθάνω, and φθάνω, or with others meaning “begin,” “stop,” or “endure.” 54 For the former, supplementary usage, while slightly more common, is apparently limited to just two verbs: μεταλαμβάνω with its late meaning “receive information, learn” (11:6; 12:8; 15:1) and ὁράω or a compound of it (4:4,6; 7:20; 8:8; 14:42). Best attested is our author’s use of the circumstantial participle. By far the majority of the non-absolute instances are temporal, but once our writer utilizes a participle as the protasis of a condition embedded in indirect speech (7:24). Most diverse is our composer’s employment of the genitive absolute. A provisional count yields ninetyseven instances of the genitive absolute. Again, given the historic nature of our work, 51. 3:17,29,32,36; 4:1(one εἴη with two participles, so bis), 26; 6:4,5; 9:18; 11:16; 12:39; 13:17; 14:17,24,28 (one ἦν with two participles, so bis); 15:7,9. Technically the instances at 4:1 are perfect, but since Greek lacks a past optative form for εἰμί, and the sense is pluperfect here, they count as pluperfect. 52. 3:36; 4:1bis; 11:16; 14:17,28bis; 15:7,9. 53. 5:27 with διατελέω; 8:26 with μακροτονέω. 54. Contrast 4Mac: καὶ πάσχοντες ἐνεκαρτέρουν “they too endured suffering,” (14:9; also 16:1), and the more literary portions of the New Testament: ἐπαύσατο λαλῶν “he ceased speaking” (Luke 5:4; also Acts 5:42; Heb 10:2). 2. The Epitome

413

25.3 The Language of Second Maccabees

the majority of these are temporal in meaning (“while, when, after”), or simply provide some attendant circumstance. However, at least thirteen are clearly causal 55 with another fourteen possibly being so. 56 A minimum of five are concessive (9:4; 12:3,35; 14:45bis). One instance appears consequential (or epexegetical): ἐκεῖ τὸν μιαιφόνον ἀπεκόσμησε τοῦ κυρίου τὴν ἀξίαν αὐτῷ κόλασιν ἀποδόντος “there he [Epiphanes] killed the polluted murderer [Andronicus]; thus the Lord dispensed a worthy punishment on him” (4:38). Regarding syntax, as expected, the genitive absolute construction most frequently starts the sentence. At 15:20 five genitive absolutes begin the sentence; at 14:20 three do; at 5:5 one genitive absolute precedes a first verb and then two more a second verb; double genitive absolutes begin sentences at 3:1; 4:18,39; 12:22; and 15:37. In a bit of stylishness the writer also likes to bury the main clause within genitive absolutes: τοῦ δὲ βασιλέως ἐπανελθόντος […] ἐνετύγχανον οἱ κατὰ πόλιν ᾽Ιουδαῖοι συμμισοπονηρούντων καὶ ῾Ελλήνων “when the king returned […] the Jews of the city began to make their appeal while even the Greeks felt a common hatred for the wrongdoing” (4:36—also 4:39; 7:4; 14:28). A few times the genitive absolute appears without expressed subjects. While three of these may be formulaic, the instances at the end of 8:27 and at 14:32 are not. 57 The accusative absolute is absent. Second Maccabees is rich in rhetorical devices, and there can be no doubt of Luis Gil’s observation that paronomasia is our author’s “figura favorite.” 58 The present author has done an intensive study of this figure of speech (also called polyptoton) in the epitome, examining forty-one instances, placing the device in its Hellenistic context, and showing where our writer was dependent on earlier, mostly classical, authors, where he was influenced by previous Septuagint usage, and where he was apparently creative in coming up with new instances of the figure. 59 No one has similarly studied other such rhetorical devices in our work, but Doran lists alliteration, homoioteleuton, metaphor, litotes, parallelism, antithesis, chiasm, and repetition. 60 In addition, the tricolon, the simile, and irony all appear; the first of these may well be our composer’s second most common figure and the investigation of this little studied device in Second Maccabees is a desideratum. 61 Likewise, study of the use of the simile in the epitome is needed because of our author’s strange avoidance of the most common introduction of it, ὡς, or more likely in the Hellenistic world, ὥσπερ, and his fixation on the aforementioned -ηδόν adverbs instead, which are literally similes. The presence of intentional asyndeton is doubtful or at least debatable. 62

55. 3:30; 4:39,40; 8:5,27; 11:10; 12:4,7,31,38; 14:28; 15:37bis. 56. 6:16; 7:25; 8:24; 9:2,18; 10:1; 12:22bis,30,36; 14:20ter, 43. 57. Well known in classical authors, SHM § 2072. This is common in Josephus, e. g., AJ 2.147; 8.30; 11.160; Vita 299, and exists in the New Testament: Matt 17:14,26; Luke 12:36. 58. L. Gil, “Sobre el estilo de libro segundo de los Macabeos” Emerita 26 (1958), 24. 59. Shaw, New Direction, chapter 7. 60. Doran, Temple Propaganda, 42-43. 61. I (New Direction, chapter 3) note how this figure is probably fairly steady throughout our work and use a few of its appearances to help offer a likely solution to a long standing textual crux at 12:13. 62. Shaw, New Direction, chapter 9.

414

2. The Epitome

25.3 The Language of Second Maccabees

Finally, Peter Walters’ remark is apropos: “2 Maccabees betrays its vacillation between vernacular and more cultivated speech.” 63 Our author was not an extremely refined individual with extensive training in composition and rhetoric, such as Philo or even Josephus. Too many flaws are present. Our composer’s overuse of δέ alone (whether consecutive or adversative) strains at the reader’s patience, when so many other options—several of which he obviously knew—were available to him. His excessive employment of διά appears a plebian mannerism. Increasing his other modes of demonstrating cause while decreasing his extreme fondness for διά would have gone a long way toward balancing his composition. Likewise his heavy utilization of the imperfect tense is trying, and his mixing it indiscriminately with the aorist cheapens his work. His limited use of the supplementary participle is poor; even certain books of the New Testament show more skill here. One can only wonder at his relatively heavy employment of the pluperfect tense, especially when the simple aorist would often have sufficed. His utilization of the triple compound προεξαποστέλλω at 12:21 calls to mind Ernest C. Colwell’s description of “a constant over-striving for emphasis” that marked non-literary koine. 64 The doubly augmented ἐκατετιθέμην at 9:25 makes one question just how concerned with writing in a high style our author was. The absence of ἐπειδή is striking, given the fact that other, far less literary writers/translators elsewhere in the Septuagint, along with welleducated mainstream authors of the time, use this conjunction with considerable frequency. Lastly, attempts at periodic structure sometimes fail: 3:10-12 has no main clause, and 12:10 is a grammatical jumble, likely from shoddy summarizing. 65 Chapters 12 and 13 are generally the nadir in epitomizing. Thus while our author often produces a high level of language, a serious look at his shortcomings reveals that he was no aficionado of Greek. Whatever his level of education, in no way does it equal that of surviving contemporary historians, whether Jewish or Gentile. It is safe to place him on a level of style above that of other writers/translators of the Septuagint (save perhaps Fourth Maccabees), and possibly Ps.-Aristeas too, but clearly somewhere below Polybius, Diodorus, Philo, and Josephus. The confession in the epitomizer’s epilogue, that his project might have been “poorly done and mediocre” (15:38, NRSV), may well allude to our writer’s awareness of a degree of shifting quality and middling abridgment. The fact that he would include it is certainly not impressive. 66

63. P. Walters (formerly Katz), The Text of the Septuagint: Its Corruptions and their Emendations (edited by D. W. Gooding), Cambridge 1973, 113. 64. E. C. Colwell, “Greek language” IDB 2 (1962), 481. 65. Goldstein, II Maccabees, 438. 66. This article was penned before the appearance of Robert Doran’s 2Maccabees. A Critical Commentary (Hermeneia), Minneapolis, MN 2012. 2. The Epitome

415

25.4 Die Sprache des 3. Makkabäerbuches Wolfgang Orth

Die in irreführender Weise als »Drittes Makkabäerbuch« betitelte Geschichte von Verfolgung und Errettung der ägyptischen Juden in der Regierungszeit Ptolemaios’ IV., verfasst in Alexandreia um 100 v. Chr. 1 und in zwei Unzialhandschriften überliefert, 2 hat in der Forschung seit der Mitte des 19. Jhdts. überwiegend sehr negative Beurteilung erfahren. Die Kritik bezog sich dabei nicht nur auf die Frage der historischen Glaubwürdigkeit, sondern auch auf Darstellungsweise und Stil. Gerügt wurden logische Inkonsistenzen, schwer erträgliche Übersteigerungen, Verstöße gegen angemessene Sprachform und bombastische Ausdrucksweise. 3 Es gibt allerdings eine bemerkenswerte Ausnahme. Der klassische Philologe U. von Wilamo1.

2.

3.

Begründung bei C. W. Emmet, »The Third Book of Maccabees« in: R. H. Charles (Hg.), The Apocrypha and Pseudepigrapha of the Old Testament in English with Introductions and Critical and Explanatory Notes to the Several Books, Vol. I: Apocrypha, Oxford 1913 (Nachdruck 1978), 155-173, speziell 155-158; B. R. Motzo, »Il rifacimento Greco di ›Ester‹ e il ›III Macc.‹« in: Saggi di Storia e Letteratura Giudeo-Ellenistica, Florenz 1924, 272-290, speziell 274; E. Bickermann, »Makkabäerbücher (III)« in: RECA XIV (1928), Sp. 797-800, speziell 798; H. Anderson, »3 Maccabees. A New Translation and Introduction« in: J. H. Charlesworth (Hg.), Old Testament Pseudepigrapha, vol. 2, New York, NY 1985, 509-529, speziell 510-512; A. Passoni dell’Acqua, »Terzo libro dei Maccabei« in: P. Sacchi (Hg.), Apocrifi dell’Antico Testamento, IV, Brescia 2000, 571-664, speziell 613; Ph. Alexander / L. Alexander, »The Image of the Oriental Monarch in the Third Book of Maccabees« in: T. Rajak / S. Pearce / J. Aitken / J. Dines (Hg.), Jewish Perspectives on Hellenistic Rulers, Berkeley, CA 2007, 92-109, speziell 92-94; J. Mélèze Modrzejewski, Troisième livre des Maccabées (BdA), Paris 2008, 119 f. und 123. Codex Alexandrinus (A; 5. Jh.) und Codex Venetus (V; 8.-9. Jh.); dazu kommt Minuskeltradition, in die die Ergebnisse zweier Rezensionen eingegangen sind; vgl. R. Hanhart (Hg.), Maccabaeorum liber III (Vetus Testamentum Graecum, Vol. IX 3,2, zweite Auflage), Göttingen 1980; R. Hanhart, Zum Text des 2. und 3. Makkabäerbuches. Probleme der Überlieferung, der Auslegung und der Ausgabe (Nachr. der Akad. d. Wiss. in Göttingen, I. Philolog.-Hist. Klasse, Jg. 1961, Nr. 13), Göttingen 1961. Groteske Widersprüchlichkeit der Darstellung: Ch. C. Torrey, »Maccabees (Books)« in: Encyclopaedia Biblica III, London 1902, 2857-2886, speziell 2879; Schwierigkeit und Unausgeglichenheit des Sprachcharakters: Hanhart, Text, 3; bombastische Ausdrucksweise/schwülstige Sprache: C. L. W. Grimm, Kurzgefasstes exegetisches Handbuch zu den Apokryphen des Alten Testamentes, Vierte Lieferung. Das zweite, dritte und vierte Buch der Maccabäer, Leipzig 1857, 214 f.; Torrey, »Maccabees (Books)«, 2880; E. Schürer, Geschichte des jüdischen Volkes im Zeitalter Jesu Christi, Bd. III (vierte Auflage) Leipzig 1909 (Nachdruck Hildesheim 1964), 489 f.; E. Schürer, The History of the Jewish People in the Age of Jesus Christ (175 B.C. - A.D. 135), revised and edited by G. Vermes, F. Millar and M. Goodman, vol. III-1, Edinburgh 1986, 538 und 540; B. M. Metzger, »The Apocrypha of the Old Testament« in: H. G. May / B. M. Metzger (Hg.), The Oxford Annotated Bible with the Apocrypha, New York, NY 1977, 294-308, speziell 294; überkünstliche Rhetorik: E. Kautzsch, »Das sogenannte dritte Buch der Mak-

416

2. The Epitome

25.4 Die Sprache des 3. Makkabäerbuches

witz-Moellendorff, führender Vertreter seines Fachs, hat – recht beiläufig freilich und an entlegener Stelle 4 – den außerordentlichen Wert von 3Makk gerade in sprachlicher Hinsicht unterstrichen: »Es sei auf dieses Buch (sc. 3Makk) nachdrücklich hingewiesen, … als einen getrübten Nachklang jener unvergleichlich anschaulichen alexandrinischen Schilderungen dieser Zeit, die bei Polybios entzücken … wie endlich als rare Proben der ›asianisch‹ genannten Schreibart.« Freilich werde stilistische Kritik hier in der Analyse des Texts noch »sehr viel thun müssen«.

1. 3Makk als griechischer Originaltext Das Profil des für uns anonymen Autors kann nur aus seiner Schrift selbst erschlossen werden. Der Text belegt Kenntnis klassischer griechischer Literatur und Vertrautheit mit Methode und Technik der griechischen Rhetorik; von daher lässt sich für den Verf. ein sehr gehobenes Bildungsniveau rekonstruieren, wie es in Alexandreia damals für Angehörige der jüdischen Oberschicht bei günstigen Rahmenbedingungen erreichbar war. 5 Ausgeklügelte sprachliche Ausdrücke und Wendungen weisen die Schrift als einen von Anfang an griechisch konzipierten Text aus; eine hebräische Vorlage ist auszuschließen. 6

2. Die Prägung der Sprache von 3Makk durch Anlehnung an stilistische Vorlagen Nach heutigen Maßstäben kann das Buch der Gattung des historischen Romans zugeordnet werden; 7 ob man damit die Intentionen des Verf. trifft, ist freilich zweifelhaft. Dass dieser den Ehrgeiz hatte, als einer anerkannt zu werden, der Geschichte 8 vermittelte, dafür sprechen schon der Anfangsabschnitt des Werks (1,1-7), ebenso aber auch die Sorgfalt, mit der der Autor in bestimmten von ihm gestalteten Abschnitten fremde Textarten inhaltlich und formal nachahmt: Besonders in Betracht kommen hier zum

4. 5. 6.

7. 8.

kabäer« in: E. Kautzsch (Hg.), Die Apokryphen und Pseudepigraphen des Alten Testaments I, Tübingen 1900 (Nachdruck Hildesheim 2002), 119-135, speziell 121. U. von Wilamowitz-Moellendorff, »Lesefrüchte LI« Hermes 34 (1899), 633-636, hier: 635 Fn. 1. Mélèze Modrzejewski, Livre, 113-115. So Torrey, »Maccabees (Books)«, 2880; M. Hadas, The Third and Fourth Books of Maccabees, New York, NY 1953, 22; Schürer, History, 540; D. J. Harrington, Invitation to the Apocrypha, Grand Rapids, MI/Cambridge 1999, 175; N. C. Croy, 3 Maccabees (Septuagint Commentary Series), Leiden 2006, XIII. 3Makk als »romantic fiction«: dazu Croy, 3 Maccabees, XV (mit der älteren Lit.). Zu den historiographischen Elementen in 3Makk vor allem Croy, Maccabees, XIVf.; Mélèze Modrzejewski, Livre, 48-87. Anders freilich M. Hadas, »Third Maccabees and Greek Romance« RR 13 (1948), 155-162, hier 155: »the author had no thought of presenting his work as sober history«. 1. 3Makk als griechischer Originaltext

417

25.4 Die Sprache des 3. Makkabäerbuches

einen ›amtliche Verlautbarungen‹ der ptolemäischen Monarchie (3,12-29 und 7,1-9) 9 und zum anderen Gebete jüdischer Priester (2,2-20 und 6,1-15). 10 Die Belesenheit des Vf. kommt zum Ausdruck in Übereinstimmungen mit griechischen Texten aus dem jüdischen Umfeld, 11 vor allem aber in Spuren von Vertrautheit mit griechischer Dichtung; für eine ganze Anzahl von Sätzen konnten metrische Struktur oder metrische Vorlage plausibel gemacht werden (z. B. 5,31: Beginn mit einem iambischen Trimeter); 12 die Grundlage bilden hier Verse aus klassischen griechischen Tragödien, vor allem aus Aischylos. 13

9. Bestätigung der formelhaften Elemente durch späthellenistische Papyrusüberlieferung: G. A. Deissmann, Bibelstudien. Beiträge, zumeist aus den Papyri und Inschriften, zur Geschichte der Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums, Marburg 1895 (Nachdruck Hildesheim/New York, NY 1977), 258-261; Emmet, »Book«, 155-158 (mit Wörterliste); A. Passoni dell’Acqua, »Il III libro dei Maccabei e l’amministrazione tolemaica« in: B. Kramer / W. Luppe / H. Maehler et al. (Hg.), Akten des 21. Internationalen Papyrologenkongresses (Berlin 13.-19. 8. 1995), Stuttgart/Leipzig 1997, 786-794 (mit vielen Einzelbeispielen); Passoni dell’Acqua, »Terzo libro«, 602-605; Mélèze Modrzejewski, Livre, 62 und 119. Anachronismen gänzlich auszuschalten, ist dem Vf. freilich nicht gelungen (siehe etwas Croy, 3 Maccabees, 65 f. und 109). 10. Dazu A. Paul, »Le Troisième livre des Macchabées«, in: ANRW II 20,1, Berlin/New York, NY 1987, 298-336, speziell 308-310; A. Passoni dell’Acqua, »Le preghiere del III libro dei Maccabei. Genere letterario e tematica« RivBib 43 (1995), 135-179; C. Boyd-Taylor, »3 Makkabees. To the Reader« in: A. Pietersma / B. G. Wright, A New English Translation of the Septuagint and the Other Greek Translations Traditionally Included under that Title, New York, NY/Oxford 2007, 521 f., speziell 522. Das bedeutet allerdings auch im Fall der Gebetstexte nicht, daß Hebraismen eine nennenswerte Rolle spielten; zu deren weitestgehendem Fehlen in 3Makk siehe schon Grimm, Handbuch, 215. 11. Gedanklich und stilistisch bestehen besonders enge Beziehungen zu 2Makk (dazu Emmet, »Book«, 156 f.[mit Liste der Wörter und Wendungen, die sowohl in 2Makk als auch in 3Makk vorkommen, sonst aber nirgends in LXX]; J. Tromp, »The Formation of the Third Book of Maccabees« Henoch 17 [1995], 311-328, speziell 318-324); bekannt sind dem Autor auch die griechischen Zusätze zu Dan (Dan 3,46-50; vgl. 3Makk 6,6); bei Übereinstimmungen mit Esther (dazu Motzo, »Rifacimento«, 274-282; Passoni dell’Acqua, »Terzo libro«, 594-596 und 598601; N. Hacham, »3 Maccabees and Esther: Parallels, Intertextuality, and Diaspora Identity, JBL 126 (2007), 765-785, speziell 765-780; vgl. dort 773 f. die Liste von Wörtern, die sich sowohl in Esther als auch in 3Makk finden, sonst aber nirgends in LXX; Mélèze Modrzejewski, Livre, 36-37) und dem Aristeas-Brief (Emmet, »Book«, 157; Alexander, »Image«, 92-100 mit Hervorhebung der für beide Schriften wichtigen περὶ βασιλείας-Tradition) muss die Frage der Priorität offenbleiben. 12. Vgl. Grimm, Handbuch, 215; Emmet, »Book«, 170; J. R. Harris, »Metrical Fragments in III Maccabees« BJRL 5 (1919), 195-207: Einzelnachweise poetischer Sprache in den Kapiteln 4 bis 7. 13. Siehe dazu M. Z. Kopidakis, Τὸ Γ’ Μακκαβαίων καὶ ὁ Αἴσχυλος. Αἰσχύλειες μνῆμες στὸ λεκτικὸ καὶ στὴ θεματογραφία τοῦ Γ’ Μακκαβαίων, Thessaloniki 1982, vor allem 35-51 (Aischyl. Ag., Choeph., Pers., Prom., Sept.). Zu Aischyl. Prom. siehe auch schon Harris, »Fragments«, 206 f. Abhängigkeit von Euripides haben nachzuweisen versucht Harris a. a. O. (Eur. Hec.) und J. R. C. Cousland, »Dionysus theomachos? Echoes of the Bacchae in 3 Maccabees« Bib 82 (2001), 539-548. Auch auf hellenistische Lyrik scheint der Autor Bezug genommen zu haben (so S. L. Sørensen, »Eine Anspielung auf Kallimachos im 3 Makkabäerbuch« GRBS 50 [2010], 87-94).

418

2. Die Prägung der Sprache von 3Makk durch Anlehnung an stilistische Vorlagen

25.4 Die Sprache des 3. Makkabäerbuches

3. Sprachliche Besonderheiten 14 3.1 Wortgebrauch 3Makk zeichnet sich im Wortgebrauch durch Variantenreichtum aus; 15 das Buch weist dabei zahlreiche ungewöhnliche Wörter auf. Die Liste der Ausdrücke, die in der LXX nur hier vorkommen, geht über die Zahl 100 weit hinaus. 16 Einige dieser Termini sind bislang auch sonst in der gesamten antiken Überlieferung nicht belegt; 17 hier handelt es sich vor allem um eigenwillige Zusammensetzungen (2,19 ἀμβλακία »Fehler«; 3,21 ἀμνησικακία »Verzeihung«; 6,8 βυθοτρεφής »in der Tiefe lebend«; 6,31δυσαίακτος »höchst bejammernswert«; 3,1 ἐκχολᾶν »wütend sein«; 6,9 μίσυβρις »Überheblichkeit hassend«; 4,6 μυροβρεχής »salbölbefeuchtet« 18; 3,17 νόθως »unaufrichtig«; 6,4 ποντόβροχος »vom Meer überschüttet«; 4,1 προκατασκιρροῦσθαι »zuvor hart werden«; 1,19 προσαρτίως »eben erst«; 2,29 προσυστέλλεσθαι »zuvor festgesetzt werden«; 7,17 ῥοδοφόρος »Rosen tragend«; 6,20 ὑπόφρικος »ein wenig zitternd«; 6,18 φοβεροειδής »schrecklich anzusehen«; 4,20 χαρτηρία »Papyrus«; 5,25 ψυχουλκεῖσθαι »in den letzten Zügen liegen«). Charakteristisch für den Autor ist im Übrigen die Vielfalt der Gottes-Epiklesen. 19

3.2 Grammatik Was die Orientierung an der Formenlehre betrifft, so ergibt sich bei 3Makk ein widersprüchliches Bild. 20 Einerseits werden immer wieder klassische (attische) Muster bevorzugt, gelegentlich durchaus in Abweichung von dem sonst in der LXX üblichen Sprachstil; 21 in diesem Zusammenhang verdient auch etwa die Anwendung der Krasis 14. Dazu der Abschnitt »Lingua e stile« in Passoni dell’Acqua, »Terzo libro«, 578-581. 15. Dazu Mélèze Modrzejewski, Livre, 115 f. Für »Töten« gibt es in 3Makk beispielsweise fünf verschiedene Ausdrücke (siehe Hanhart, Text, 36). 16. Zusammenstellungen solcher Wörter bei Emmet, »Book«, 161; Kopidakis, Γ’ Μακκαβαίων, 11-13. 17. Dazu Mélèze Modrzejewski, Livre, 116 f. Zu streichen sind hier ἀνεπιστρέπτως (vgl. die Belege bei LSJ) und δράκος (vgl. Pollux II 147). Siehe zu den angegebenen Stellen die Fußnoten zur Übersetzung von Passoni dell’Acqua, »Terzo libro«, 634-663. 18. Dass man den Ausdruck auch sonst kannte, scheint freilich aus Suet. Aug. 86,2 hervorzugehen. 19. Dazu Mélèze Modrzejewski, Livre, 88. 20. Vgl. die einschlägigen Belege und ihre Auswertung bei Thackeray (Register 324-325). 21. Z. B. τέλεον (statt τέλειον) in 1,22; ἴστε (statt οἴδατε) in 3,14. Das hat dazu geführt, daß man den Vf. gelegentlich in irreführender Weise als Attizisten bezeichnet hat: vgl. Emmet, »Book«, 161: 3Makk sei ein Beispiel für »pseudo-Classicalism of the Atticists« in seiner extremsten Form. Ähnlich Metzger, »Apocrypha«, 294; Anderson, »3 Maccabees«, 510; Croy, 3 Maccabees, XIII; Boyd-Taylor, »3 Makkabees«, 522. Für bestimmte Besonderheiten lässt auch Passoni dell’Acqua, »Terzo libro«, 579 die Bewertung »pseudo-attizistisch« gelten. Von Bestrebungen eines attizistischen Reformprogramms ist der Autor aber jedenfalls kaum berührt. In diesem Zusammenhang sind folgende Detailbeobachtungen von Interesse: Hiat ist in 3Makk nicht strikt vermieden (vgl. 5,2); οὐθείς (1,13) steht neben οὐδέν (5,42); Einfügung von σ in Passivformen kommt vor (z. B. 5,18: ἰάσθησαν [falls nicht Unzialfehler] für εἰάθησαν); statt attischem τέτυχηκα erscheint in 5,35 das Perfekt τέτευχα. 3. Sprachliche Besonderheiten

419

25.4 Die Sprache des 3. Makkabäerbuches

Erwähnung (3,22; 7,12; 7,19), die in der LXX nicht häufig ist. Andererseits fehlt es nicht an hellenistischen Bildungen, auch da, wo sonst die Koiné noch an älteren Formen des klassischen Griechisch festgehalten hat. 22 In der Syntax stehen Einfachheit und Kompliziertheit einander gegenüber. Da gibt es zum einen eine Vorliebe für die Parataxe: gerade im Anfangsteil wird die Partikel δέ sehr häufig gebraucht (vgl. 1,1-10). Auf der anderen Seite kommt es nicht selten zu Wortstellungen, die dadurch, dass Zusammengehöriges auseinandergerissen wird, das Verständnis erschweren (z. B. 5,31). Eine vollständige Aufzählung weiterer stilistischer Auffälligkeiten ist hier nicht möglich; beispielshalber sei hier nur noch auf so verschiedenartige Phänomene wie die Weglassung von Akkusativobjekten, 23 die Aufblähung von Verbalausdrücken nach dem Muster von πρόσβασιν ποιεῖσθαι (1,26 anstelle von προσβαίνειν, vgl. entsprechend 1,23; 2,9; 4,17; 6,40) 24 und die Verwendung des Aorist- oder Praesensinfinitivs in futurischer Bedeutung 25 hingewiesen.

4. Zur literarischen Qualität von 3Makk 3Makk hebt sich ab von den übersetzten Teilen der LXX durch Übernahme von Sprach- und Stiltendenzen einer hellenistischen Kunstprosa, die möglichste Anschaulichkeit zu erreichen sucht und dabei das Ausmalen des Details bis hin zum Grotesken steigert. 26 Variantenreichtum im Wortgebrauch, manieristisches Streben nach Effekten (vgl. etwa 4,1-21) und eine an rhetorische Tradition und Praxis anknüpfende Erregung von πάθος (vgl. etwa 1,16-29) sind kennzeichnend für diesen Text. Seiner sprachlichen und stilistischen Form begegnet man heute im allgemeinen mit größerem Verständnis als früher; als Werk der Literatur hat das dritte Makkabäerbuch in jüngerer Zeit wiederholt betont positive Würdigung erfahren. 27

Hanhart, Text, 59. Hanhart, Text, 21. Dazu Grimm, Handbuch, 214. Hanhart, Text, 17. So Hanhart, Text, 59, mit Übernahme der von Wilamowitz-Moellendorff (s. oben S. 392) auf 3Makk angewendeten Stil-Klassifizierung ›asianisch‹ ; vgl. auch Passoni dell’Acqua, »Terzo libro«, 580: »tipicamente asiano«. Zum Begriff ›Asianismus‹ im allgemeinen siehe E. Norden, Die antike Kunstprosa (dritte Auflage), Leipzig/Berlin 1915 (Nachdruck Darmstadt 1958), I, 131-149; U. von Wilamowitz-Moellendorff, »Asianismus und Attizismus« Hermes 35 (1900), 1-52; E. R. Curtius, Europäische Literatur und lateinisches Mittelalter (fünfte Auflage), Bern/München 1965, 76. 27. Hadas, Books, 22; Boyd-Taylor, »3 Makkabees«, 522; Mélèze Modrzejewski, Livre, 115: »On doit également apprécier la richesse de son vocabulaire et le haut niveau de sa culture littéraire«. 22. 23. 24. 25. 26.

420

4. Zur literarischen Qualität von 3Makk

25.5 Sprache und Stil des 4. Makkabäerbuches Christoph Kugelmeier

Das Buch gibt nicht wenige Rätsel auf 1. Sein Verfasser ist unbekannt, sein »Sitz im Leben« und der Ort seiner Abfassung sind umstritten 2. Als Entstehungszeit wird heute zumeist das 1. Jh. n. Chr. angenommen 3. Bei allen Versuchen einer Einordnung spielen auch sprachliche und stilistische Indizien eine Rolle. Dass der unbekannte Verfasser gleich mit den ersten Worten seines φιλοσοφώτατος λόγος (1,1) den klassischen Stil der philosophischen Diatribe anklingen lässt, ist lange schon erkannt worden 4. Bereits im Altertum, zu Anfang des 5. Jhs. n. Chr., bezeichnet der Kirchenhistoriker Philostorgios die Schilderung als οὐχ ἱστορίαν μᾶλλον ἢ ἐγκώμιον 5. Nicht geschichtliche Tatsachen stehen also im Vordergrund, sondern ein nach allen Regeln der rhetorischen Kunst aufgebauter Lobpreis auf Märtyrer, die bereit waren, für ihren jüdischen Glauben unter entsetzlichen Foltern ihr Leben hinzugeben, und damit zugleich – darauf legt der Verfasser besonderen Nachdruck – 1.

2. 3. 4. 5.

Allgemein wichtige Literatur (in chronologischer Reihenfolge): J. Freudenthal, Die Flavius Josephus beigelegte Schrift: Ueber die Herrschaft der Vernunft (IV. Makkabäerbuch). Eine Predigt aus dem ersten nachchristlichen Jahrhundert, Breslau 1869; E. Wolscht, De Ps. Iosephi oratione, quae inscribitur περὶ αὐτοκράτορος λογισμοῦ, Diss. Marburg 1881; H. Dörrie, Die Stellung der vier Makkabäerbücher im Kanon der griechischen Bibel (Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen, Philologisch-Historische Klasse, Fachgruppe 5, Religionswissenschaft, N.F. 1,2), Göttingen 1937; A. Dupont-Sommer, Le quatrième livre des Machabées, Paris 1939; M. Hadas, The Third and Fourth Books of Maccabees, New York, NY 1953; J. C. H. Lebram, »Die literarische Form des vierten Makkabäerbuches« VChr 28 (1974), 81-96; U. Breitenstein, Beobachtungen zu Sprache, Stil und Gedankengut des vierten Makkabäerbuchs, Basel 1976, zugl. Diss. Basel 1974; H. Anderson, »4 Maccabees: A New Translation and Introduction« in: J. Charlesworth (Hg.), The Old Testament Pseudepigrapha, London 1985, 531-564; H.-J. Klauck, »Hellenistische Rhetorik im Diasporajudentum. Das Exordium des Vierten Makkabäerbuchs (4 Makk 1.1-12)« NTS 35 (1989), 451-465; R. Weber, »Eusebeia und Logismos. Zum philosophischen Hintergrund von 4. Makkabäer« JSJ 22 (1991), 212234; J. W. van Henten, The Maccabean Martyrs as Saviours of the Jewish People. A Study of 2 and 4 Maccabees, Leiden u. a. 1997, zugl. Diss. Leiden 1986; H. Spieckermann, Martyrium und die Vernunft des Glaubens – Theologie als Philosophie im vierten Makkabäerbuch (Nachrichten der Akademie der Wissenschaften zu Göttingen, Philologisch-Historische Klasse), Göttingen 2004; D. A. deSilva, 4 Maccabees. Introduction and Commentary on the Greek Text in Codex Sinaiticus, Leiden 2006; H. Görgemanns, »Philosophie, Gesetz und Vorbild. Zu einigen Abschnitten des 4. Makkabäerbuches« in: A. Jördens u. a. (Hg.), Quaerite faciem eius semper. Studien zu den geistesgeschichtlichen Beziehungen zwischen Antike und Christentum (FS A. Dihle), Hamburg 2008, 78-90. Einen Überblick gibt deSilva, 4 Maccabees, xiff. Ausführlich dazu jetzt deSilva, 4 Maccabees, xiv-xvii. E. Norden, Die antike Kunstprosa vom VI. Jahrhundert v. Chr. bis in die Zeit der Renaissance, 2 Bde., Leipzig/Berlin 1898, I 129-131 und I 417; so auch Lebram, »Form«, 81 und 83. Bei Photios, PG 65,461; s. Dörrie (wie Anm. 1), 51-52. 4. Zur literarischen Qualität von 3Makk

421

25.5 Sprache und Stil des 4. Makkabäerbuches

exemplarisch den Beweis dafür erbrachten, dass die Vernunft, identifiziert mit dem göttlichen Gebot, alle πάθη zu beherrschen vermag 6. In seiner Wortwahl zeigt schon das Exordium deutliche Einflüsse der hellenistischen Philosophie seiner Zeit 7. Auf den ersten Blick erwecken die einleitenden Sätze mit ihrer Betonung der Herrschaft der Vernunft über die πάθη und der Aufzählung der Kardinaltugenden 8 einen dezidiert stoischen Eindruck. Auch im weiteren Verlauf erinnert manches an die stoische Ethik, sowohl in den Gedankengängen 9 als auch in der Terminologie 10. In Einzelheiten relativiert sich dies allerdings. So findet sich beispielsweise die 1,16 gegebene Definition der σοφία als γνῶσις θείων καὶ ἀνθρωπίνων πραγμάτων καὶ τῶν τούτων αἰτιῶν fast wörtlich in SVF II fr. 35; denselben Satz gebrauchen freilich die Platoniker ebenfalls (vgl. Albinus, Introductio in Platonem, p. 152 Hermann), bisweilen wird er auch Pythagoras zugeschrieben (vgl. Cramer, Anecdota Graeca IV 400,15 f.) 11. Auch erblickt der Verfasser in den πάθη nicht wie die (ältere) Stoa etwas, das eliminiert werden muss; er postuliert vielmehr, mit Begriffen wie παθῶν ἐπικρατεῖν bzw. κυριεύειν (1,3 f.), ihre Zähmung und ihre Unterordnung unter die Vernunft 12. Die vieldiskutierte Problematik wird verständlicher, wenn man sich vor Augen hält, dass sich die Schrift an ihren Rezipienten orientiert, Juden in einem hellenistischen Milieu, das sie in ihrer Anhänglichkeit an die Thora verunsichert. Aus diesem Grunde setzt der Verfasser das griechische Prinzip der Vernunft und das göttliche Gebot des Judentums gleich, wie sich in dem markanten Sprachbild 2,22 zeigt: Die Vernunft (hier νοῦς, anderswo auch λογισμός und φρόνησις – die Terminologie ist keineswegs konsistent), in dieser Weise aufgefasst, »herrscht« wie Gott unumschränkt (αὐτοκράτωρ, αὐτοδέσποτος τῶν παθῶν, vgl. ἐνεθρόνισεν) in der natürlichen Ordnung, und es ist Aufgabe des Menschen, als wahrer φιλόσοφος 6. A.-M. Malingrey, Philosophia: Étude d’un groupe de mots dans la littérature grecque des Présocratiques au IVe siècle après J.-C., Diss. Paris 1961, 93. Vgl. auch 3,3-5. 7. Ähnliche Formulierungen im Aristeas-Brief (221 f. und 256) zeigen überdies die Anbindung an die Tradition der Septuaginta. 8. Vgl. 1,18 f. und 5,23 f. 9. Norden, Kunstprosa, I 417 verweist zur Standhaftigkeit der Märtyrer auf Seneca, epist. 71,5 sowie 419, Anm. 1 auf Seneca Rhetor, contr. II 5 (torta a tyranno pro marito). Zum Bezug von Eleazars Feststellung, noch die Übertretung des kleinsten Gebotes sei ein Bruch des Gesetzes im Ganzen, und dem σμῆνος ἀρετῶν der Stoiker s. Hadas, Maccabees, 118; kritisch dazu R. Renehan, »The Greek Philosophic Background of Fourth Maccabees« RhM 115 (1972), 229-231; T. Rajak, »Dying for the Law: The Martyr’s Portrait in Jewish-Greek Literature« in: M. J. Edwards / S. Swain (Hg.), Portraits: Biographical Representation in the Greek and Latin Literature of the Roman Empire, Oxford 1997, 53. 10. Näheres bei Renehan, »Background«, 224 f. 11. Renehan, »Background«, 228. Den engen Anschluß dieser »hellenistischen Terminologie« an gängige philosophische Definitionen hebt auch Weber, »Eusebeia«, 219 mit Anm. 30 hervor; V. 17 ist dabei (so Hadas, Maccabees, 149) »the keynote of our author’s thought«. 12. D. C. Aune, »Mastery of the Passions: Philo, 4 Maccabees and Earliest Christianity« in: W. E. Helleman (Hg.), Hellenization Revisited: Shaping a Christian Response within the Greco-Roman World, Lanham, MD 1994, 136; Görgemanns, »Philosophie«, 79-81; Hadas, Maccabees, 186; Breitenstein, Beobachtungen, 160; deSilva, 4 Maccabees, xxviii. Renehan, »Background«, 226 f. vermutet einen Anschluss an den gegenüber der altstoischen Rigorosität konzilianteren Poseidonios, vgl. v. a. fr. 34 Edelstein-Kidd (aus dem Traktat Περὶ παθῶν).

422

4. Zur literarischen Qualität von 3Makk

25.5 Sprache und Stil des 4. Makkabäerbuches

diese Weisheit anzuerkennen und handelnd nachzuvollziehen 13. Dies wird gewiss mit Absicht so formuliert, dass sich eine auffällige Parallele mit allgemein bekanntem stoischem Gedankengut ergibt, jedoch mit der signifikanten Abweichung, dass eben auch die πάθη in der vernunftgemäßen gottgeschaffenen Weltordnung ihren Platz haben, wenngleich einen untergeordneten. Dazu passt auch, dass ausgerechnet ein Element der jüdischen Religionsausübung, das in der heidnischen Umwelt den stärksten Anstoß erregte und zur Ausgrenzung der praktizierenden Juden am meisten beitrug, die Speisegesetze 14, in den Mittelpunkt des Konflikts und seiner Erörterung gestellt wird 15. Es ist also sinnvoller, die Begrifflichkeit des Buches nicht ausschließlich auf eine Schule (etwa die Stoa) festzulegen, vielmehr drückt sich der erkennbare – und für die Zeit der Abfassung keineswegs ungewöhnliche 16 – philosophische Eklektizismus sprachlich in einer »philosophic koinē« aus 17. Seine eigentlich paradoxe Aufgabe, den Nachweis zu erbringen, dass in Wirklichkeit »Vernunft« und »Weisheit« eben mit dem von den Hellenen verachteten und verfolgten Gesetz des einen Gottes übereinstimmen (man beachte die eigentümliche und in dieser Wortjunktur singuläre Hervorhebung des εὐσεβὴς λογισμός 18), zwingt den Autor geradezu, eine wahrhaft unerhörte Situation in den kräftigsten Farben zu schildern; diese Spannung erklärt auch die rhetorische Anspannung des Stils, die recht oft als ein massives »Zuviel« anmutet, ebenso wie das ständige Bemühen um eine Umwertung der seinen Zeitgenossen vertrauten hellenistischen Begrifflichkeit, das sich in der oft hervorgehobenen, verblüffenden Vielzahl von gesucht erscheinenden Wortneubildungen niederschlägt 19. Gerade diese starke stilistische und sprachliche Anspannung hat viel Kritik hervorgerufen 20. Der epideiktische Charakter der Schrift verbindet sich untrennbar mit

13. Weber, »Eusebeia«, 226. Die von Breitenstein, Beobachtungen, 29, Anm. 1 im Anschluss an das LSJ vorgenommene Unterscheidung zwischen einem konkreten und einem metaphorischen Gebrauch von ἐνθρονίζω ist übersubtil, weshalb das Wort nicht zu Recht in Breitensteins Liste der nicht vor dem 3. Jh. n. Chr. belegbaren Vokabeln erscheint. Die Parallele Diodorus Siculus 33,15 Πτολεμαίου κατὰ τὴν Μέμφιν ἐνθρονιζομένου τοῖς βασιλείοις κατὰ τοὺς Αἰγυπτίων νόμους zeigt zugleich, dass es sich nicht um einen Neologismus handelt. 14. S. dazu K. Bringmann, Geschichte der Juden im Altertum, Stuttgart 2005, 147-160, bes. 149. 15. Dies hebt vor allem Weber, »Eusebeia«, 221 und 226 f. hervor. 16. Grundriss der Geschichte der Philosophie. Begründet von Friedrich Ueberweg, völlig neu bearb. Ausgabe. Die Philosophie der Antike, Bd. IV: Die hellenistische Philosophie, Basel 1994, 6. 17. Renehan, »Background«, 228; deSilva, 4 Maccabees, xiii spricht von »standard topics of Greco-Roman philosophical ethics«. Weitere Literatur bei Weber, »Eusebeia«, 215, Anm. 8. 18. Weber, »Eusebeia«, 213 und 218. 19. Eine gründliche Zusammenstellung bei Breitenstein, Beobachtungen, 181-188. 20. Norden, Kunstprosa, I 419 kritisiert »bombastischen Schwulst«; Lebram, »Form«, 82: »eine Prunkrede … die sich zu nichts weniger eignet, als zum Beweis für einen philosophischen Lehrsatz«; er spricht von »abstossendem Realismus« in der Darstellung der Martyrien. Freudenthal, Herrschaft, 155 moniert »schlechtes Griechisch«, wogegen deSilva, 4 Maccabees, xii die »Greek compositional skills« rühmt. Zur stilistischen Kritik an der (bisweilen als Interpolation betrachteten) Schlusspartie s. Freudenthal 155 (»Citirwut«), Hadas, Maccabees, 239, Breitenstein, Beobachtungen, 154-156, Görgemanns, »Philosophie«, 87-90. 4. Zur literarischen Qualität von 3Makk

423

25.5 Sprache und Stil des 4. Makkabäerbuches

ihrer von Philostorgios gekennzeichneten Natur als ἔπαινος bzw. ἐγκώμιον 21. Auffallende Parallelen der fiktiven Grabschrift 17,9 f. mit den Grab- und Gedenkreden der griechischen Literaturtradition (man denkt vor allem an die Gefallenenrede des Perikles bei Thukydides II 35-46 und die 2. Rede des Lysias) gaben bisweilen Anlass, das gesamte Buch als einen solchen ἐπιτάφιος für die Märtyrer zu deuten (das Wort wird tatsächlich 17,8 verwendet) 22. Letztlich ist hier kaum eine Entscheidung zu treffen, schon weil die Anbindung an eine topographisch konkretisierbare Märtyrerverehrung Vermutung bleibt. Einig ist man sich mittlerweile darüber, dass (schon angesichts der Überlieferungslage) der Autor das Buch auf Griechisch verfasst hat 23, für einen Rezipientenkreis, der nicht nur Griechisch verstand, sondern auch die hier besonders reichlich verwendeten Gestaltungsmittel der hellenischen Rhetorik zu würdigen wusste. Einige Beispiele mögen hier genügen, um die Schrift in diesen Zusammenhang einzuordnen und die Funktion der entsprechenden rhetorischen Elemente aufzuzeigen: – die Spannung zwischen dem Streben nach Variation des sprachlichen Ausdrucks und dem Bedürfnis, durch bewusste, oft penetrant wirkende Wiederholung 24 besonders wichtige Gedanken herauszustreichen (vor allem in den Reden, vgl. besonders 8,17-26 25) – der Dramatik der Schilderung dienen vor allem die häufigen direkten Reden, die mit allen Kunstmitteln der ἠθοποιία (sermocinatio) ausgeführt sind und auffallend häufig Einwürfe des Verfassers in Frageform darstellen (vgl. 15,4), bisweilen auch als Einwände eines fictus interlocutor gebildet sind, die gleich darauf als »lächerlich«

21. Gleich mit dem ersten Satz äußert der Verfasser seine Absicht φιλοσοφώτατον λόγον ἐπιδείκνυσθαι. Zum γένος ἐπιδεικτικόν vgl. Aristoteles, Rhet. I 3 und Quintilian, inst. III 4,7. Zu den typischen Charakteristika, die die Eingangspartie als exordium ausweisen, s. Klauck, »Rhetorik«, 104 mit Literatur und 107; vgl. (speziell zu den Widerlegungen der gegnerischen Einwände) Rhet.Her. I 8 und Cicero, de inv. I 22 (contemptio), s. H. Lausberg, Handbuch der literarischen Rhetorik (4. Aufl.) Stuttgart 2008, 158. Details auch bei B. Heininger, »Der böse Antiochus. Eine Studie zur Erzähltechnik des 4. Makkabäerbuches« BZ NF 33 (1989), 44 mit Anm. 3, der auf das Defizit der Diatriben-Klassifizierung Nordens hinweist, die den »enkomiastischen Charakter des zweiten Teils« zu wenig berücksichtige; Klauck, »Rhetorik«, 104 mit Hinweis auf die Nennung des Wortes ἔπαινος 1,7; deSilva, 4 Maccabees, xxiff. 22. Dupont-Sommer, Quatrième livre, 24 f., 34 f. und 67; van Henten, Martyrs, 48 f. und 58 f. (er zieht Ehreninschriften zum Vergleich heran); Rajak, »Portrait«, 39-41; Klauck, »Rhetorik«, 105 (mit Literatur in Anm. 23); Lebram, »Form«, 84-86; deSilva, 4 Maccabees, xxi; zum entsprechenden Aufbau im ganzen Lebram 86 ff. und deSilva 79. 23. Hebraismen finden sich so gut wie nicht (deSilva, 4 Maccabees, xii). 24. Zur großen Anzahl mehrgliedriger und parallel gebauter Sätze und zu den überaus häufigen Polysyndeta mit καί s. Breitenstein, Beobachtungen, 89; vgl. besonders 2,10 ff. und 13,20 ff. 25. Breitenstein, Beobachtungen, 97 f., 100 und 121-128; zum Stilprinzip der μεταβολή (variatio) s. Lausberg, Handbuch, 142; zum parallelismus membrorum und seiner unterschiedlichen Realisierung in der hebräischen Stilistik und in der rhetorischen Gestaltung im Griechischen s. E. Norden, Agnostos Theos: Untersuchungen zur Formengeschichte religiöser Rede, Leipzig/ Berlin 1913, 254-263 und 355-364.

424

4. Zur literarischen Qualität von 3Makk

25.5 Sprache und Stil des 4. Makkabäerbuches

widerlegt werden (1,5, 2,24 26) oder die Entschlossenheit der Märtyrer charakterisieren sollen (6,23 οἱ δὲ τοῦ τυράννου δορυφόροι, τί μέλλετε;) 27 – in dieselbe Kategorie fällt auch die von vielen antiken Rednern gern gebrauchte Apostrophe, wie etwa die an den παιδευτὴς νόμος 5,34 f. und an Eleazar 7,6 ff. sowie 15,1: ὦ λογισμὲ τέκνων παθῶν τύραννε; besonders eindrucksvoll wirkt 8,17, wo der Verfasser sich und seine Zuhörer bzw. Leser einschließt und sie dadurch gleichsam in die geschilderte Situation noch einmal besonders hineinnimmt 28 Sprachlich wird die dramatisierende Wirkung unterstützt von Polyptota wie 15,1-3 εὐσέβεια – εὐσεβείας – εὐσέβειαν und 13,20 f. χρόνον – χρόνῳ – χρόνων; nicht von ungefähr finden sich beide Stellen in einer Apostrophe und in einer eingefügten Rede. Dasselbe gilt für eine Paronomasie wie 8,17 βασιλέως ἡμᾶς καλοῦντος καὶ ἐπὶ εὐεργεσίᾳ παρακαλοῦντος, auch hier also in einer Rede und an einer Stelle, an der sich eine Entscheidung anbahnt 29. Die Sprache des Buches lässt die Tendenz erkennen, im bewussten Rückgriff auf derartige Elemente der klassischen Rhetorik möglichst die dazu passende gehobene Bildungssprache seiner Epoche zu verwenden. Im Einklang damit steht die damals verbreitete Neigung zum Attizismus in Syntax und Semantik 30, die 4Makk teilt und mit der sich die Schrift gegen die sonstige sprachliche Gestaltung der Septuaginta deutlich abhebt 31. Allerdings kann der Verfasser gewisse Züge der hellenistischen Koine nicht verhehlen, wie wir sie aus zahlreichen Inschriften und Papyri als Verwaltungssprache, aber auch aus Werken der Literatur kennen, vor allem aus Polybios und eben aus Schriften aus dem Umfeld der Bibel, insbesondere der Septuaginta, dem Neuen Testament, Philon und Josephus 32. Im Vergleich mit diesen literarischen und nichtliterarischen Textdokumenten lassen sich einige Ähnlichkeiten des Sprachgebrauchs von 4Makk feststellen: – der gegenüber dem älteren Griechisch einschließlich des Attischen weit häufigere Gebrauch des substantivierten Infinitivs vom Typ περὶ τοῦ τεθνάναι αὐτόν (4,22), der schon bei Polybios ins Auge springt und eine gewisse Schwerfälligkeit des Duktus zur Folge hat (ein auffälliges Beispiel ist I 4,2 τὸ μηδένα τῶν καθ᾽ ἡμᾶς ἐπιβεβλῆσθαι τῇ τῶν καθόλου πραγμάτων συντάξει, womit sich außergewöhnliche Sperrungen wie 4Makk 17,20 τῷ δι᾽ αὐτοὺς τὸ ἔθνος ἡμῶν τοὺς πολεμίους μὴ 26. Zu dieser »Einwandformel« Breitenstein, Beobachtungen, 54; vgl. Lausberg, Handbuch, 410. 27. Breitenstein, Beobachtungen, 91; Görgemanns, »Philosophie«, 87; Heininger, »Antiochus«, 49 mit Anm. 17 und 55-57; s. Lausberg, Handbuch, 407 ff. und 543 f. 28. Breitenstein, Beobachtungen, 106-109; Lausberg, Handbuch, 377-379. 29. Breitenstein, Beobachtungen, 192; J. D. Denniston, The Greek Particles (2. Aufl.) Oxford 1954, 124-137. 30. S. dazu, neben dem klassischen Werk von W. Schmid, Der Atticismus in seinen Hauptvertretern, 5 Bde., Stuttgart 1887-1897, Ndr. Hildesheim 1964, jetzt A. Dihle, »Der Beginn des Attizismus« A&A 23 (1977), 162-177; F. Rodríguez Adrados, Geschichte der griechischen Sprache, Tübingen/Basel 2002, 192-196. 31. Ein solcher attizistischer Zug ist der Gebrauch der Modi, insbesondere des Optativs in abhängigen Sätzen; s. dazu Breitenstein, Beobachtungen, 51 f. und 177 sowie Schmid, Der Atticismus, I 97 und Dihle, »Attizismus«, 163; ein sicheres Datierungskriterium für das 2. Jh. n. Chr. kann allerdings mit reiner Statistik nicht gewonnen werden. 32. Ein Überblick bei Rodríguez Adrados, Geschichte, 180 f. 4. Zur literarischen Qualität von 3Makk

425

25.5 Sprache und Stil des 4. Makkabäerbuches

ἐπικρατῆσαι vergleichen lassen); es handelt sich um eine »vorwiegend literarische Stileigenheit« 33, die sich wohl dem Kanzleistil verdankt 34 – die Nominalisierung des Infinitivs durch den Artikel entspricht der allgemeinen Tendenz zum Nominalstil, vgl. 4,10 πρὸς τὴν τῶν χρημάτων ἁρπαγήν und 13,13 τῷ θεῷ ἀφιερώσωμεν ἐξ ὅλης τῆς καρδίας τῷ δόντι τὰς ψυχὰς καὶ χρήσωμεν τῇ περὶ τὸν νόμον φυλακῇ τὰ σώματα; an der letzteren Stelle lässt sich als Effekt eine gesteigerte Feierlichkeit beobachten, die zur Situation der pathetischen Aufforderung zum Selbstopfer gut passt. Auch im Wortgebrauch lassen sich Bezüge zur zeitgenössischen Sprachform entdecken, gerade dann, wenn auch die wichtigen außerliterarischen Zeugnisse der Papyri und Inschriften in den Blick genommen werden, ein Gebiet, das zudem, anders als die literarischen Texte, einen stetigen Zuwachs verzeichnet. Interessant ist in diesem Zusammenhang die militärisch-agonistische Metaphorik, mit Begriffen wie ἀγών, ἀγωνίζεσθαι, ἆθλον und dergleichen. Sie erscheint in der Schrift des Öfteren, um das Standhalten der Märtyrer, ganz im Sinne des deuteropaulinischen Briefs 2Tim 4,7 f. (τὸν καλὸν ἀγῶνα ἠγώνισμαι), als »Kampf« zu kennzeichnen 35. Neben diesen bekannten Metaphern sei noch auf das Verb λειποτακτεῖν »desertieren« hingewiesen (9,23 μιμήσασθέ με, ἀδελφοί, λέγων, μή μου τὸν ἀγῶνα λειποτακτήσητε μηδὲ ἐξομόσησθέ μου τὴν τῆς εὐψυχίας ἀδελφότητα). Im »Greek-English Lexicon of the Septuagint« von Lust, Eynikel und Hauspie (2. Aufl. Stuttgart 2003) wird es als Neologismus bezeichnet, und in der Tat fügt sich das Kompositum gut in die zuweilen extravagante Wortwahl des Buches. Man stellt allerdings fest, daß schon Platon, Leg. 943 d von λιποταξίου γραφαί, Anklagen wegen Desertion, spricht 36. Diese Funktion als typisch juristischer Fachterminus lässt sich anhand einer Fülle von Papyruszeugnissen belegen 37, die deutlich machen, daß der Verfasser von 4Makk auch in diesem Falle auf dem Boden der Hochsprache seiner Zeit steht, die auch in den Kanzleien

33. O. Hoffmann, Geschichte der griechischen Sprache (3. Aufl. bearb. von A. Debrunner), Teil II, Berlin 1953, 128. 34. E. Schwyzer, Griechische Grammatik, Zweiter Band: Syntax und syntaktische Stilistik (vervollst. und hrsg. von A. Debrunner), 369 und 383 f.; F. Blass / A. Debrunner, Grammatik des neutestamentlichen Griechisch (bearb. von F. Rehkopf, 18. Aufl.) Göttingen 2001, § 388; Breitenstein, Beobachtungen, 60 f. 35. Breitenstein, Beobachtungen, 188; Klauck, »Rhetorik«, 113, Anm. 48 und 49; vgl. die mehr oder minder zeitgenössischen Parallelen Philon, omn.prob.lib. 26 f., Seneca, dial. II 9,4 und Epiktet I 18,21, I 24,2 und III 22,56 (der Stoiker und der Kyniker als »olympische« Wettkämpfer). 36. Das Durcheinander von λειπ- und λιπ- in den Vordergliedern der Komposita ist sogar eine recht frühe Erscheinung, s. E. Schwyzer, Griechische Grammatik, Erster Band: Allgemeiner Teil, Lautlehre, Wortbildung, Flexion (6. Aufl.) München 1990, 2,442 und zu 4Makk 9,23 P. Walters, The Text of the Septuagint. Its Corruptions and their Emendation, Cambridge 1973, 32 (der für λιπ- eintritt). Auch im Platon-Text bietet der Parisinus graecus 1807 λειπ-, was sich freilich leicht als itazistische Variante erklären lässt. Die entsprechende Überlieferung in 4Makk ist einheitlich. 37. Gerade in der frühesten Bezeugung, POxy 2407,46 (also erst 1957 ediert), in einem Protokoll einer Volksversammlung im 3. Jh. n. Chr., findet sich das Wort in einem ironisch übertragenen Sinne, nämlich bezogen auf Versammlungsmitglieder, die durch Abwesenheit glänzen.

426

4. Zur literarischen Qualität von 3Makk

25.5 Sprache und Stil des 4. Makkabäerbuches

geformt wurde 38. Dasselbe gilt für einen Solitär wie ἀντιρρητορεύω (6,1 τοῦτον τὸν τρόπον ἀντιρρητορεύσαντα ταῖς τοῦ τυράννου παρηγορίαις). Man trifft es auch in der einzigen Literaturparallele, bei Maximos von Tyros, einen Mittelplatoniker des späten 2. Jhs. n. Chr., in einem juristischen Zusammenhang an, nämlich dem des Prozesses gegen Sokrates 39. Möglicherweise wurde der Gebrauch des Wortes in 4Makk durch die in den Papyri öfters anzutreffenden Rechtstermini ἀντίρρησις »Klagebeantwortung« und ἀντιρρητικός »die Klagebeantwortung betreffend« beeinflusst 40, wofür es allerdings die frühesten Belege ebenfalls erst im 2. Jh. n. Chr. gibt. Auch die Wortbildung zeigt große Übereinstimmung mit einer Tendenz zur denominativen Ableitung, die sich bereits im klassischen Attisch beobachten lässt und sich in der Koine fortsetzt 41. Wenn jedoch 15,32 die Mutter der Märtyrer als νομοφύλαξ gepriesen wird, so weicht diese Verwendung als Lob für ihre Treue zum jüdischen Gesetz vom sonstigen Gebrauch dieses verbreiteten Terminus signifikant ab: Er bezeichnete in Athen das siebenköpfige Kontrollgremium für die gesetzgebenden Versammlungen, eine Einrichtung, die es ausweislich vieler literarischer und nichtliterarischer Zeugnisse ähnlich in vielen anderen griechischen Poleis gab und die wir in veränderter Form als Amtsbezeichnung in Platons Gesetzen antreffen 42. Ein gängiges Wort für eine in der gesamten griechischen Welt bekannte Institution wird also umgedeutet, ganz in dem oben erörterten Sinne des Spannungsverhältnisses zwischen jüdischer Zielsetzung und hellenistischer Argumentationsform. Völlig allein steht der Verfasser mit dieser Uminterpretation freilich nicht; daß Philon, de mutatione nominum 43 mit dem abgeleiteten Femininum νομοφυλακίς die Bundeslade meint, deutet darauf hin, dass sich 4Makk damit in der Traditionslinie der jüdischen Exegese befindet 43. Auf die nicht »dogmatische« Verwendung einiger Begriffe aus dem philosophischtheologischen Bereich wurde bereits eingegangen. Auch an anderen Beispielen bestätigt sich der bisher gewonnene Eindruck, dass sich die Wortwahl des Verfassers in der Hauptsache weniger nach einer festen Terminologie als nach den semantischen Konventionen seiner Epoche richtet. Gut lässt sich dies z. B. an der Entwicklung des Wortes αἰσθητήριον vom Konkreten (»Sinnesorgan«, vgl. als frühesten Beleg Aristoteles, de an. 421 b 17, wo damit der kurz zuvor genannte μυκτήρ, »Nase«, paraphrasiert wird) hin zu dem abstrakteren Wortgebrauch »Sinne als Teil der menschlichen Seele« beobachten, wie er 4Makk 2,22 vorliegt (ἡνίκα δὲ ἐπὶ πάντων τὸν ἱερὸν ἡγεμόνα νοῦν διὰ τῶν αἰσθητηρίων ἐνεθρόνισεν) und sowohl bei den Stoikern als auch bei

38. Polybios kennt das Wortfeld nicht; Philon hingegen benutzt es häufig im übertragenen Sinne von 4Makk. 39. III 3, p. 24,63 Trapp. 40. F. Preisigke, Wörterbuch der griechischen Papyrusurkunden, Erster Band, Berlin 1925, s. v., vgl. POxy 68,10 f. (131 n. Chr.) ποιοῦμα[ι τὴν | δαίουσαν ἀντίρησιν, nämlich gegen die Auszahlung aus einer Erbschaft. 41. P. Chantraine, La formation des noms en grec ancien, Paris 1933, 89 f.; Schwyzer (wie Anm. 36), 2,732. 42. Platon, Leg. 755 a u. ö.; vgl. Aristoteles, Pol. 1322 b–1323 a. 43. Nur auf diesem Umweg über Philon ließe sich rechtfertigen, dass Lebram, »Form«, 90 in dem Wort einen »platonisierenden Titel« erblickt; eigentlich greift Platon selbst ja auf eine hergebrachte Amtsbezeichnung zurück. 4. Zur literarischen Qualität von 3Makk

427

25.5 Sprache und Stil des 4. Makkabäerbuches

Epikur häufig anzutreffen ist 44. In ihrer Verbindung mit der Erkenntnistheorie steht das Epikur-Zitat der Makkabäer-Stelle weit näher als der aristotelische und stoische Wortgebrauch, der in beiden Fällen noch sehr von der konkret biologischen Auffassung geprägt ist. Ähnliches gilt für einen im hellenistischen Denken gern gebrauchten Begriff wie εὐλογιστία (5,22) 45, während κενοδοξία (5,10 und 8,24) zu den beliebten Vokabeln auch der hochsprachlichen Koine gehört 46 und so den bereits gewonnenen Eindruck abrundet, dass die Sprache von 4Makk in wichtigen semantischen Elementen von der Koine ihrer Zeit beeinflusst ist. Als Resümee lässt sich festhalten: Die Sprache von 4Makk steht trotz des unverkennbaren Bemühens um attische bzw. attizistische Hochsprachlichkeit und um den entsprechenden rhetorisch geschliffenen Stil keineswegs außerhalb der Koine, der zeitgenössischen moderneren Form des Griechischen. Dies zeigt sich weit weniger in der grammatischen Struktur als vielmehr im Wortschatz. Wenn der Verfasser bestimmte Begriffe verwendet, die einerseits in der bekannten philosophischen Literatur des Hellenismus, andererseits in der zeitgenössischen Verwaltungsterminologie beheimatet waren, so wie sie uns aus den Zeugnissen der Papyri und Inschriften entgegentritt, dann ist von Fall zu Fall zu klären, ob sein Rückgriff auf dieses Vokabular bewusst erfolgt oder ob ihm Wörter dieser Art offensichtlich unbewusst »in die Feder fließen«. Beides läßt sich in 4Makk beobachten: Titel wie νομοφύλαξ oder das militärisch-agonistische Vokabular erscheinen in bewusster Umdeutung zum üblichen Sprachgebrauch an entscheidenden Wendepunkten der Erzählung. Bei den eigentlich philosophischen Termini, bei deren Verwendung in einer theologischen Schrift eher Konsistenz zu erwarten wäre, fällt dagegen geradezu eine Lässigkeit und Austauschbarkeit auf, die es auch unmöglich macht, den Verfasser einer bestimmten Denkschule zuzuordnen. Auch darin ist er freilich ein Kind seiner Zeit.

44. Vgl. SVF II fr. 836 τῶν δὲ ἑπτὰ μερῶν τῆς ψυχῆς πέντε μέν εἰσι τὰ αἰσθητήρια ὅρασις, ὄσφρησις, ἀκοή, γεῦσις καὶ ἁφή Epikur, ep.Her. 50 καὶ ἣν ἂν λάβωμεν φαντασίαν ἐπιβλητικῶς τῇ διανοίᾳ ἢ τοῖς αἰσθητηρίοις. 45. Vgl. SVF III fr. 264, Philodem, Περὶ θεῶν III fr. 81 Diels; Epikur gebraucht ep.Men. 135 das Adverb εὐλογίστως, woran sich im Vergleich mit der engen gedanklichen Parallele Aristoteles, Rhet. 1410 a 6-8 die Austauschbarkeit der Wortfelder εὐλογιστ- und φρονιμ- und Verwandtem außerhalb der Stoa zeigt – ebenso wie in 4Makk (dazu Breitenstein, Beobachtungen, 169, Anm. 4). Interessant ist der Wortgebrauch bei Philon: An den drei Stellen, an denen er das Substantiv verwendet (de conf. 67, de migr. 71 und de mut. 128), hat es ebenfalls die Bedeutung »Klugheit, Vernunft«, wohingegen das Verb εὐλογιστεῖν zumeist »segnen« bedeutet (vgl. leg.all. III 191 f.), das Adjektiv εὐλόγιστος entsprechend »gesegnet« (vgl. leg.all. I 17 f.). 46. Vgl. Epikur, Sent. 30 = fr. 5 XXX Arrighetti sowie Polybios III 81,9 u. ö.; viele Belege bei Philon.

428

4. Zur literarischen Qualität von 3Makk

VII The language of the Septuagint and the New Testament

26. The Language of the Septuagint and the New Testament: Prolegomena Moisés Silva

Even a cursory acquaintance with biblical scholarship would make clear that specialists are heavily dependent on the Septuagint for their understanding of the Greek of the New Testament. Commentaries on New Testament books, for example, rarely fail to cite Septuagint passages as illustrations of specific usages. And there is Walter Bauer’s well-known assessment, “As for the influence of the LXX, every page of this lexicon shows that it outweighs all other influences on our literature.” 1 It may come as a surprise, therefore, to realize that full-length studies of the linguistic connection between the Septuagint and the New Testament are rare. To be sure, some influential works—such as Hatch’s Essays in Biblical Greek near the end of the nineteenth century and Dodd’s The Bible and the Greeks a generation later 2—have provided substantial discussions of the subject, including observations regarding matters of principles and methods. But one can think of only two publications wholly devoted to such questions: a book by H. A. A. Kennedy that was limited to the vocabulary and a relatively brief monograph (approximately 130 pages) by David Tabachovitz dealing with syntactical and stylistic issues, but not in systematic or comprehensive fashion. 3 The numerous works that wholly or in part address the connection between Septuagint and New Testament focus mainly, and understandably, on text and interpretation. 4 It is thus all the more appropriate and necessary that a Handbuch zur Septuaginta devote a section to the linguistic link between the two corpora.

1. Connections between the Septuagint and the New Testament The present chapter is intended to provide an introductory overview, and we should note at the outset that, even when the discussion is narrowed to the linguistic field, various kinds of connections must be kept in mind. In the first place, there is an in1. 2. 3.

4.

BDAG, xviii. E. Hatch, Essays in Biblical Greek, Oxford 1889; C. H. Dodd, The Bible and the Greeks, London 1935. H. A. A. Kennedy, Sources of New Testament Greek: The Influence of the Septuagint on the Vocabulary of the New Testament, Edinburgh 1995; D. Tabachovitz, Die Septuaginta und das Neue Testament: Stilstudien, Lund 1956. For instance, the recent book by R. T. McLay (The Use of the Septuagint in New Testament Research, Grand Rapids, MI 2003) devotes only pages 146-148 to lexical matters and does not deal at all with grammar, syntax, or style. Most general introductions to the Septuagint pay even less attention to this topic, but see H. B. Swete, An Introduction to the Old Testament in Greek, Cambridge 1902, 451-457; K. H. Jobes / M. Silva, Invitation to the Septuagint (second edition), Grand Rapids, MI 2015, 201-206. 1. Connections between the Septuagint and the New Testament

431

26. The Language of the Septuagint and the New Testament: Prolegomena

direct link resulting from the fact that the books of both the Septuagint and the New Testament are significant instantiations of ἡ κοινὴ διάλεκτος. In other words, even if the New Testament writers had never used what we call the Septuagint, their language could still be elucidated with reference to the latter, more or less in the same way that, say, the writings of Plutarch, though produced after Paul’s death, are often exploited to shed linguistic light on the letters of the apostle. Indeed, most of the references to the Septuagint in Walter Bauer’s Greek-English Lexicon should be understood in this way, without assuming that the New Testament usage in view is the result of immediate influence. For that matter, the reverse perspective is valid as well: the language of the New Testament sheds light on that of the Septuagint. In any case, for our understanding of “Biblical Greek” generally within the context of the Koine, we are still heavily dependent on Albert Thumb’s magisterial contribution from more than a century ago. 5 Second, there is the direct connection that would be expected from the fact that the New Testament writers were evidently familiar with the Septuagint. When analyzing linguistic data, however, it is often very difficult to determine whether direct influence has been at work. If a usage is found in both corpora but not in extrabiblical literature, there may indeed be a reasonable presumption that the New Testament authors have been affected by their firsthand acquaintance with the Septuagint. But such evidence is hardly conclusive: in spite of the large number of Greek writings that have survived (including private communications preserved in the papyri), we still have access to only a fraction of the documents that were produced in the Hellenistic and Roman periods—to say nothing of the incalculably vast amount of oral communication that is beyond our reach. On the other hand, the fact that a relatively rare linguistic datum in the Greek Bible may also be attested in extrabiblical writings does not at all rule out the possibility that its presence in the New Testament should be attributed to the author’s knowledge of the Septuagint. A third type of connection derives from the Jewish background common to both corpora (even in the case of the writings attributed to Luke, a Gentile, the cultural context is largely Jewish). One may say that this consideration is a subclass of the previous one, for it too concerns the direct influence of the Septuagint on the New Testament language, focusing specifically on the area of Semitisms. But the matter is more complicated than that, for if a feature of New Testament Greek appears to reflect Semitic influence and also corresponds to Septuagint usage, one must weigh the likelihood that the New Testament writers have been affected by other possible factors: (a) direct and immediate influence (i. e., at the very time of writing) from their own native Semitic tongue (whether Hebrew or Aramaic); (b) influence mediated by a spoken form of Greek that bears the marks of Palestinian bilingualism; (c) influence from the use of a 5.

A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus: Beiträge zur Geschichte und Beurteilung der Κοινή, Strasbourg 1901. Among more recent contributions, note J. Frösén, Prolegomena to a Study of the Greek Language in the First Centuries A.D.: The Problem of Koiné and Atticism, Helsinki 1974, as well as the relevant sections in histories of the language, e. g., F. Rodríguez Adrados, A History of the Greek Language: From Its Origins to the Present, Leiden 2005; A.-F. Christidis (ed.), A History of Ancient Greek: From the Beginnings to Late Antiquity, Cambridge 2007; G. Horrocks, Greek: A History of the Language and Its Speakers (second edition), Chichester 2010.

432

1. Connections between the Septuagint and the New Testament

26. The Language of the Septuagint and the New Testament: Prolegomena

broader form of Greek with features that parallel Semitic usage but that have a different origin (e. g., Coptic); (d) influence from Semitic elements that are cultural or conceptual, rather than linguistic, in character but that are given linguistic expression by the New Testament writers spontaneously.

2. History of Scholarship Keeping in mind this broad and complex range of issues, we may survey the history of scholarship on the linguistic relationship between the Septuagint and the New Testament, 6 beginning with an essay by Edwin Hatch entitled, “On the Value and Use of the Septuagint.” Concerned that students viewed New Testament Greek as identical with the language of the classical period, Hatch wished to point out the differences. In addition to the lapse of time, he argued, there is the fact that biblical Greek was spoken in a different country and, more to the point, by a different “race.” The worldview of the biblical writers was so different that the Greek words they used communicated something quite other than Hellenic ideas. “For every race has its own mass and combinations of ideas; and when one race adopts the language of another, it cannot, from the very nature of the human mind, adopt with it the ideas of which that language is the expression. It takes the words but it cannot take their connotation: and it has ideas of its own for which it only finds in foreign phrases a rough and a partial covering. Biblical Greek is thus a language which stands by itself. What we have to find out in studying it is what meaning certain Greek words conveyed to a Semitic mind.” 7 To achieve this aim, Hatch proposed to study the New Testament language anew, as though it were a recently discovered dialect. And to do so, he argued, one must recognize the immense value of the Septuagint. Since it is the translation of an original that we possess, we can in fact refer to the Hebrew text “and in most cases frame inductions as to their meaning which are as certain as any philological induction can be.” 8 The value of the Septuagint, he claimed, is enhanced by the fact that it is largely targumic in character: it often gives glosses and paraphrases, it does not always adhere to the original metaphors, and it even varies its own rendering of particular words and phrases. Hatch makes much of this last point and calls special attention to those cases where the Septuagint uses a group of synonyms indiscriminately to translate a corresponding group of Hebrew terms. As an example he shows that six different Hebrew words are translated by ἐξαιρεῖν, λυτροῦν, ῥύεσθαι, or σῴζειν without apparent distinction. “It is reasonable to infer that, in their Hellenistic use, the Greek words which one thus used interchangeably for the same Hebrew words did not differ, at least materially, from each other in meaning, and that no substantial argument can be

6.

7. 8.

Some of the material that follows summarizes M. Silva, Biblical Words and Their Meaning: An Introduction to Lexical Semantics (revised edition), Grand Rapids, MI 1994, 56-68, to which the reader is referred for more detail. Hatch, Essays, 11. Hatch, Essays, 14. 2. History of Scholarship

433

26. The Language of the Septuagint and the New Testament: Prolegomena

founded upon the meaning of any one of them unless that meaning be common to it with the other members of the group.” 9 Hatch’s general conclusion is that, although some words do not differ in meaning from their use in non-biblical authors, “the great majority of New Testament words” are those that “express in their Biblical uses the conception of a Semitic race, and which must consequently be examined in the light of the cognate documents which form the LXX.” 10 He then goes so far as to lay down two canons for studying the relation of Septuagint words to the Hebrew terms they translate: (a) if a Greek word renders a Hebrew word uniformly, the two words must have the same meaning; and (b) if several Greek words are used interchangeably as renderings of the same Hebrew word, those Greek words must have “an allied or virtually identical meaning.” 11 This notion of determining the meaning of Septuagint words on the basis of Hebrew-Greek equivalences has been adopted, at least in part, by subsequent scholars, notably C. H. Dodd. 12 The method, however, can easily lead to faulty conclusions unless the following factors are carefully considered: (a) the context of each occurrence, since the choice of one of the synonyms may have been due to the general drift of the passage, to syntactical details, or even to rhythmic considerations; (b) the stylistic preferences of individual translators; (c) polysemy in the Hebrew words themselves, so that the translator may have tried to reproduce the various shades of the original. Hatch’s proposals were strongly criticized by T. K. Abbott in a long review of publications on the subject of “New Testament Lexicography.” 13 From the very start, Abbott minimizes the importance of Hebraic influence on the New Testament: “This is, as far as the language is concerned, less than is sometimes supposed. Expressions characterized as Hebraisms may in not a few instances be paralleled in classical writers, the difference being in their frequency. As these Hebraisms, however, affect the phraseology more than the vocabulary, we shall not dwell on them.” 14 Neither is Abbott impressed with the idea that the Septuagint has greatly influenced New Testament Greek; after all, the Authorized Version, which was much more influential than the Septuagint, did not prevent the loss of old meanings (e. g., quick in the sense of “living”). “For our part, when we read the Septuagint what strikes us is its unlikeness to the language of the New Testament.” 15 For example, Psalm 51, which was surely very familiar to the New Testament writers, contains five words that are not 9. Hatch, Essays, 23. Because the later Greek versions also appear to use the terms interchangeably, he says on page 30 that the evidence “almost amounts to proof, that the words were in common use as synonyms.” 10. Hatch, Essays, 34. 11. Hatch, Essays, 35. 12. Dodd’s approach is best illustrated in part one of The Bible and the Greeks, entitled “The Religious Vocabulary of Hellenistic Judaism,” which discusses the names of God, law, and other theological terms. 13. T. K. Abbott, Essays, Chiefly on the Original Texts of the Old and New Testaments, London 1891, 65-109. 14. Abbott, Essays, 66 (my emphasis). Notice the similarity between this formulation and that of J. H. Moulton two decades—and many discoveries—later, in A Grammar of New Testament Greek (third edition), vol. 1, Edinburgh 1908, 11. 15. Abbott, Essays, 69.

434

2. History of Scholarship

26. The Language of the Septuagint and the New Testament: Prolegomena

found at all in the New Testament and one word with a different meaning. Even when considering the Septuagint and New Testament in general, Abbott finds relatively few “biblical meanings” common to both; further, there are numerous New Testament words and meanings not in the Septuagint, such as ὑπομονή (Septuagint: “expectation”) and ἄφεσις (Septuagint: “release”). Contradicting Hatch’s main principle, Abbott concludes: “Such facts as these show that the influence of the Septuagint version on the vocabulary of the New Testament was not predominant, and that to make the usage of the former determine the interpretation of the latter, except in the case of terms of Hebrew theology, is quite out of the question.” 16 After these general statements, Abbott goes on to criticize specific elements in the work of Hatch. For example, anticipating Barr’s caveats against “illegitimate totality transfer,” 17 he has this to say concerning Hatch’s treatment of ἀρετή: “It is, indeed, a grave fault in a lexicographer or interpreter to assume that because a word has a modified meaning when used in a particular connexion, therefore it may per se bear the same.” 18 He even takes the Septuagint expert to task for a naive understanding of the character of the Septuagint, especially in connection with Hatch’s principle that the interchangeable translations of the same Hebrew word give a reliable clue to meaning: “Apart from difference of judgment as to the rendering of a word, there are very few translators whose work can be safely taken as a standard of the usage of their own language,” and moreover the Septuagint “is really a collection of versions made by a series of independent translators, differing both in their knowledge of Hebrew and in their command of Greek.” 19 In any case, where Hatch “tries to apply to the N.T. a signification peculiar to the LXX, or ascertained according to the maxims he lays down, he is in no one instance successful.” 20 Kennedy’s Sources of New Testament Greek 21 is apparently the only full-scale work devoted to a comprehensive analysis of the relationship between the Septuagint and New Testament vocabularies. After some interesting introductory chapters, he proceeds to analyze those terms common to the Septuagint and the New Testament but not found outside the Bible. Kennedy realizes, however, that judgments concerning the relationship of the New Testament to the Septuagint must be made, not for the vocabulary as a whole, but for different types of words. He therefore classifies the vocabulary into (a) theological and religious terms, (b) actual Hebrew and Aramaic loan words, (c) technical terms for Jewish customs and ideas, (d) everyday words, (e) Alexandrian words, and (f) new compounds. Devoting the whole of chapter 8 to theological and religious terms, the author concludes that (apart from literal imitations of Hebrew expressions) this class of words “is the clearest instance of a direct influence of the LXX on the vocabulary of the New 16. Abbott, Essays, 71. The happy distinction between theological and non-theological terms— ignored by many, but especially by Hatch—is crucial. 17. J. Barr, The Semantics of Biblical Language, Oxford 1961, 218. 18. Abbott, Essays, 78. 19. Abbott, Essays, 86. 20. Abbott, Essays, 98. On page 86 he speaks of Hatch’s method of determining the meaning of New Testament words as unsound, and on page 94, less kindly, as “utterly fallacious.” 21. See above, note 2. 2. History of Scholarship

435

26. The Language of the Septuagint and the New Testament: Prolegomena

Testament,” though he also points out, wisely, that some of these terms may have been common to the region, that the New Testament writers were bound to have a Semiticized vocabulary regardless of the Septuagint, and that “the special theological terms of the New Testament are at the most connected with, not derived from, the usage of the LXX.” 22 Yet in spite of these reservations, Kennedy wants to “admit absolutely” that theological terms (and to a lesser extent those denoting Jewish customs) can be subjected to Hatch’s canons concerning Hebrew-Greek equivalences. 23 The other classes of words do not so clearly reflect Septuagint influence. His final, general conclusions deserve to be quoted in full: The earliest Christian writers, in proclaiming the new faith, had to express in words deep theological ideas, unheard of in the old world. It was natural that, in making this attempt, they should take for their model a vocabulary already formed. These writers, moreover, were Jews. Their whole view of things was penetrated with Hebrew modes of thought. Accordingly, they could not fail to make copious use of a type of language already adapted to their special requirements. But the influence of the LXX. on the vocabulary of the New Testament must not be exaggerated. Caution is necessary in determining that which is to be regarded as usage in Biblical Greek, seeing that the LXX. is a translation done by unskillful hands, and that ignorance of Greek or ignorance of Hebrew is often responsible for phenomena of vocabulary which are peculiar to the Biblical language. When we consider the exceptional importance of the Greek Bible to the New Testament writers, the astonishing fact is that its influence on their vocabulary is not incomparably greater than it is found to be. 24

During the twentieth century, the question of linguistic relationships between the Septuagint and the New Testament received attention from a variety of angles. Further opposition to Hatch was expressed by Adolf Deissmann, who (a) showed persuasively that the Semitisms found in the New Testament were not “sufficient reason for scholars to isolate the language” of the New Testament from the Koine of the time, 25 and (b) argued against attributing to a Septuagint word the meaning of the Hebrew word it translated. He illustrated this latter point with the term ἱλαστήριον, which does not mean “lid” (as Hebrew kapporet does); the evidence from the papyri makes it clear that the Septuagint, instead of translating the Hebrew concept, “has replaced it by another 22. Kennedy, Sources, 108-109. 23. Kennedy, Sources, 136. 24. Kennedy, Sources, 164-165. Some years later Theodor Nägeli, stimulated by the papyrological discoveries and convinced that Paul’s letters were of utmost importance to the development of the Koine, published Der Wortschatz des Apostels Paulus. Beitrag zur sprachgeschichtlichen Erforschung des Neuen Testaments, Göttingen 1905. He included a discussion of the Septuagint, noting Paul’s strong dependence on it as well as some important deviations, and concluding that Paul’s lexical inheritance came from the Hellenistic colloquial language and from the Septuagint and, further, that Paul’s Hebraisms—since almost all of them seem to have been mediated through the Septuagint—should be called “Septuagintalisms” (74). 25. A. Deissmann, The Philology of the Greek Bible: Its Present and Future, London 1908, 65. Other prominent exponents of this viewpoint were A. Thumb and J. H. Moulton (see above, notes 5 and 14). So-called Deissmannism has been criticized by various scholars, but the basic formulation, when properly understood, is unassailable. See M. Silva, “Bilingualism and the Character of New Testament Greek” Bib 69 (1980), 198-219.

436

2. History of Scholarship

26. The Language of the Septuagint and the New Testament: Prolegomena

concept which brings out the sacred purpose of the lid.” 26 Richard Ottley, however, was not convinced by Deissmann’s example and argued that kapporet itself implied “propitiation”; in other words, the translators attempted, through the use of ἱλαστήριον, to capture the nuances of the Hebrew term. 27 Ottley was not reverting to an uncritical acceptance of Hatch’s principles; rather, he wished to point out that there are two valid ways of approaching the Septuagint—by making use of the papyri and by considering the intention of the translators. This issue continues to be a controversial one, as is clear from the differences among recent dictionaries and translations of the Septuagint. For example, Johan Lust, Erik Eynikel, and Katrin Hauspie seek to understand and present the meaning of many or most Septuagint terms by taking into account the Hebrew text being translated. 28 T. Muraoka, on the other hand, states that we should “read the Septuagint as a Greek document and try to find out what sense a reader in a period roughly 250 B.C. – 100 A. D. who was ignorant of Hebrew or Aramaic might have made of the translation”. 29 In addition, Muraoka seems to place Albert Pietersma’s “interlinear” model 30 in the same category as LEH, but in fact Pietersma, in contrast to LEH, believes that we should recognize “new” meanings in Septuagint words only if such meanings can be verified outside non-translation literature. 31 In many instances, as it turns out, Muraoka and LEH agree, against NETS, in recognizing a meaning that appears to be derived from the Hebrew equivalent. 32

3. Conclusions and Perspectives But even if we limit ourselves to those items that are generally believed to reflect Semitic usage, questions remain regarding the influence of the Septuagint on the New Testament. Hardly a book on New Testament studies fails to touch on this issue, even if only indirectly, yet concrete and well-defined formulations are surprisingly difficult to find. A general consensus, however, can easily be detected: while rejecting the more extreme features of Hatch’s position, most biblical scholars would probably agree with Henry Swete that the Septuagint should be the “starting-point in examining the sense of all words and phrases which, though they may have been used in classical Greek or

26. Deissmann, Philology, 92. 27. R. R. Ottley, A Handbook to the Septuagint, London 1920, 167-171. Similarly, “had it been customary to translate the Hebrew shophet ‘leader’ or ‘chief’ instead of ‘judge’, it would have been advisable to render κριτής in the same way when representing shophet: for the Greek translators meant whatever shophet meant” (171). 28. LEH, vi-vii. 29. GELS, viii. 30. NETS, xv-xvi. 31. See A. Pietersma, “A New Paradigm for Addressing Old Questions: The Relevance of the Interlinear Model for the Study of the Septuagint” in: J. Cook (ed.), Bible and Computer: The Stellenbosch AIBI-6 Conference, Leiden 2002, 337-364, in particular 353-355. 32. For further discussion, see my review of Muraoka’s Lexicon in Westminster Theological Journal 72 (2010), 435-438. Note also John A. L. Lee’s valuable analysis in BIOSCS 43 (2010), 115-125. 3. Conclusions and Perspectives

437

26. The Language of the Septuagint and the New Testament: Prolegomena

by the κοινή, passed into Palestinian use through the Greek Old Testament, and in their passage received the impress of Semitic thought and life.” 33 We would do well to follow the lines laid down by Abbott and Kennedy. While the total impact of the Septuagint on the New Testament is beyond controversy, its influence on the New Testament language must be defined within fairly clear bounds. In particular, we must distinguish those uniform elements of a language that form part of its very structure from those variable elements that may be regarded as stylistic. It is important to appreciate the distinction between regularity and variability in language. To a large extent, meaningful language use is a matter of the writer’s or speaker’s choice (variability), but that choice is constrained by linguistic structure (regularity). Now linguistic structure is determined by the dynamic, spoken form of a language, not by its written tradition. Literary monuments like the Septuagint or the English King James Version will often influence stylistic choice, but rarely if ever will the structure of a language in a community change as a result of reading the literary remains of a previous generation. And although stylistic influence is especially obvious if an author or speaker wishes deliberately to imitate a writing that he or she regards highly, 34 mannerisms that are more or less subconscious give evidence of the same phenomenon. These considerations may also help to explain why discussions of Septuagint influence on the New Testament language almost always have to do with the vocabulary, including expressions of various kinds, and only seldom with the “nuts and bolts” of grammar. Consider again the impact of the King James Version on contemporary English speakers. One sometimes hears allusions to this version in the form of phrases such as the quick and the dead (though note that the adjective quick by itself is most unlikely to be used in the sense “alive” as a result of such influence), but even the most avid readers of the King James Version do not make statements like “If thou wilt, thou canst achieve anything” (Mk 1:40) or “When even was come, we returned home” (Mk 6:47). Likewise, we do not normally find in the New Testament grammatical forms or syntactical constructions that are peculiar to the Septuagint (over against the Koine in general), to say nothing of usages like ἄνθρωπος ἄνθρωπος ὃς ἐὰν γένηται ἀκάθαρτος ἐπὶ ψυχῇ ἀνθρώπου (Num 9:10). Indeed, one may say that the few exceptions there are prove the rule. Perhaps the best-known examples are (a) the use of ἐν τῷ with an infinitive to indicate the sense “while” and (b) the introduction of a sentence with (καὶ) ἐγένετο, especially when followed by καί with a finite verb. 35 Without minimizing these and other usages as attesting Septuagint influence, we should note that they lie on the boundary between lexical and syntactical phenomena. To say that they are merely “turns of phrase” (as is certainly true of the King James Version’s “and it came to pass”) may be going too far, 33. Swete, Introduction, 457; note that the statement begs the fundamental question, How do we identify those particular words and phrases? 34. As is probable in the case of Luke. See H. F. D. Sparks, “The Semitisms of St. Luke’s Gospel” JTS 44 (1943), 129-138. Some scholars interpret the evidence differently. 35. Cf. F. Blass / A. Debrunner / F. Rehkopf, Grammatik des neutestamentlichen Griechisch (eighteenth edition), Göttingen 2001, §§ 404.1; 442.5. Both of these constructions are especially common in Luke, who seems to wish to imitate the Septuagint style.

438

3. Conclusions and Perspectives

26. The Language of the Septuagint and the New Testament: Prolegomena

but neither should they be viewed as evidence of significant alteration in the structure of Greek syntax. From a linguistic point of view, these usages reflect changes that are both restricted in scope and relatively superficial. It should be clear, in conclusion, that the history of research can be of significant help as we seek to refine our understanding of the linguistic relationship between the Septuagint and the New Testament. One cannot help but be impressed, for example, by the way that scholars of previous generations were able to anticipate insights that would later be confirmed by specialized and sophisticated methods. But it also gives one pause to realize that some of the most vigorous debates today represent battles that have been fought for a long time. May we learn from those battles and avoid perpetuating the errors of the past.

3. Conclusions and Perspectives

439

27. Le vocabulaire de la Septante dans le Nouveau Testament Madeleine Wieger

Le message du Nouveau Testament est sans doute nouveau – mais les mots employés pour le dire ne constituent pas une grande chaîne de néologismes, voire une langue inouïe. Les livres du Nouveau Testament ne se sont pas élaborés hors sol, mais sur le terreau familier de la langue grecque ancienne, qui prend à cette époque la forme du grec international de la koinê. Aussi le vocabulaire du Nouveau Testament est-il le même que celui du monde méditerranéen hellénophone du Ier siècle après JésusChrist. Dans l’histoire de la lexicologie du Nouveau Testament, ce fait n’a pas été considéré comme acquis d’emblée. Le grec néotestamentaire présente de nombreuses différences par rapport à celui des œuvres classiques. En hellénistes chevronnés, les Pères de l’Église ont tôt fait de les repérer. À la Renaissance, on dresse des listes de voces biblicae attestées uniquement dans la Bible de langue grecque. Les dictionnaires spécialisés de la fin du XIXe siècle recensent dans le Nouveau Testament près de mille mots nouveaux ou sens neufs donnés à des vocables anciens 1: on conclut à l’existence d’un « grec biblique » faisant système, détaché du reste de la langue. Nombre des vocables concernés sont présents déjà dans la Septante, et cela ne fait que renforcer la théorie 2. Elle est pourtant battue en brèche à l’aube du XXe siècle, lorsqu’au lieu de Platon ou d’Isocrate, on rapproche du Nouveau Testament les papyrus d’époque hellénistique et romaine: Adolf Deissmann produit des listes à son tour, mais pour montrer que la plupart des mots ou sens prétendument bibliques sont attestés dans la langue grecque de tous les jours telle qu’en usent les documents non littéraires 3. Les textes des auteurs post-classiques sont examinés à nouveaux frais et fournissent eux aussi des parallèles. L’originalité du Nouveau Testament en matière de vocabulaire n’est plus attachée qu’à quelques mots isolés. Cette opinion radicale a été nuancée depuis lors. Le grec néotestamentaire est sans conteste une variante parmi d’autres de la koinê: il ne constitue pas une langue à part.

1. 2.

3.

Voir J. H. Thayer, A Greek-English Lexicon of the New Testament, being Grimm’s Wilke’s Clavis Novi Testamenti, translated, revised and enlarged, New York 1887, 693-698. Voir la préface à la troisième édition du dictionnaire de H. Cremer: H. Cremer, Biblisch-theologisches Wörterbuch der Neutestamentlichen Gräcität. Dritte, sehr vermehrte und verbeßerte Auflage, Gotha 1883, V-X. A. Deissmann, Bibelstudien. Beiträge, zumeist aus den Papyri und Inschriften, zur Geschichte der Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums, Marbourg 1895; Neue Bibelstudien. Sprachgeschichtliche Beiträge, zumeist aus den Papyri und Inschriften, zur Erklärung des Neuen Testaments, Marbourg 1897; Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt, Tübingen 1908.

440

3. Conclusions and Perspectives

27. Le vocabulaire de la Septante dans le Nouveau Testament

Mais il présente des traits caractéristiques d’un groupe de locuteurs particulier, celui de la chrétienté naissante. La Septante apporte un éclairage précieux à qui veut étudier les spécificités de leur lexique. Avec le Nouveau Testament, elle compte parmi les corpus les plus importants de textes écrits en grec de la koinê: à l’instar de Luc ou de Paul, les traducteurs de la Bible hébraïque se servent eux aussi des mots du grec commun. Mais ces deux corpus sont encore liés plus étroitement: la Septante est la Bible des premiers chrétiens. Ils s’approprient la koinê sous l’influence de la version grecque des Écritures juives, dont ils font leur livre de méditation et de prière. Cela n’est pas sans conséquences sur le plan linguistique. On parle de « septantisme » lorsqu’un trait distinctif du grec néotestamentaire semble repris directement à la Septante. Ce chapitre est consacré aux septantismes que comporte le vocabulaire du Nouveau Testament. Leur repérage n’est pas toujours aisé: lorsqu’un mot de la Septante est employé aussi dans le Nouveau Testament, et ce, dans un sens analogue, il faut envisager d’abord qu’on puise de part et d’autre au même lexique courant du grec de l’époque. La charge de la preuve s’est inversée depuis les travaux de Deissmann: d’autres hypothèses ne seront avancées que si le mot n’est pas attesté ailleurs dans la koinê ou s’il y est rare, tandis que la Septante et le Nouveau Testament en font usage fréquemment. Alors seulement le terme concerné peut être considéré comme un bien propre du judaïsme et du christianisme hellénophones. On ne conclura pas immédiatement à un septantisme: le grec des Juifs et des premiers chrétiens est jalonné de sémitismes qui ne sont pas à proprement parler des septantismes. La distinction entre les uns et les autres ne relève cependant pas de l’évidence. Si le sémitisme constitue un élément de la langue hébraïque ou araméenne entré dans une autre langue, le septantisme est pour sa part un écho du texte grec des Écritures juives. Or ce texte, traduit pour une bonne part d’un original sémitique, est rempli d’hébraïsmes à chaque page. Les sémitismes que le Nouveau Testament recèle à son tour s’analysent donc sur le plan linguistique comme des emprunts à une langue étrangère, mais pourraient aussi constituer, pour certains, une trace du texte de la Septante. Pour ce qui est du vocabulaire, l’une et l’autre explication tendent à se confondre: nombre de mots ou de sens forgés sans doute par les traducteurs sont entrés dans la langue courante du Juif hellénophone et ne sont plus guère perçus comme des septantismes. Lorsque les auteurs du Nouveau Testament emploient aussi bien διάβολος que Σατανᾶς pour désigner l’Adversaire de Dieu, ils n’ont probablement pas conscience que le premier mot vient de la Septante de Job 1–2 et de Zacharie 3, tandis que le second, translittéré de l’araméen ‫ָשָׂט ָנא‬, a dû naître dans le langage oral; que l’usage de διάβολος pour désigner le « satan » soit ou non une création des Septante, il n’en porte plus la marque et fait désormais partie de la langue. Pour distinguer septantismes de vocabulaire et sémitismes, il faudrait faire le départ, parmi les mots que la Septante et le Nouveau Testament ont en commun, entre ceux choisis par les auteurs du Nouveau Testament sous l’influence directe de la version grecque des Écritures et les termes qu’ils héritent de la langue en usage dans le judaïsme hellénophone. Or les spécificités du vocabulaire des Juifs de langue grecque ne s’expliquent pas séparément de la religion qu’ils pratiquent et qui prend appui justement sur la Septante. Quelques critères d’ordre linguistique et littéraire peuvent être

3. Conclusions and Perspectives

441

27. Le vocabulaire de la Septante dans le Nouveau Testament

mis en œuvre néanmoins pour déceler les septantismes lexicaux véritables contenus dans le Nouveau Testament 4.

1. Les sémitismes lexicaux de la Septante et le Nouveau Testament La traduction grecque de la Bible hébraïque est largement sémitisante. Dans son vocabulaire figurent des mots ou des sens d’emprunt entrés dans l’usage, mais aussi des sémitismes qui n’ont pas fait d’émules, parce qu’ils portent trop nettement la marque du processus de traduction et sont impropres en grec. De nombreux sémitismes que la Septante et le Nouveau Testament ont en commun appartiennent à la première catégorie. Il s’agit pour l’essentiel des mots grecs recouvrant le vocabulaire religieux usité en hébreu. On y compte quelques mots translittérés de l’hébreu ou de l’araméen, comme πάσχα, « Pâque », quelquefois adaptés au moule du mot grec et rendus déclinables. D’autres mots existent déjà en grec, mais revêtent une signification nouvelle lorsqu’ils renvoient à une institution, un objet, une notion théologique propres à la religion juive – ainsi νόμος, « loi », mais aussi 4.

Les ouvrages qui traitent du lien entre le vocabulaire de la Septante et celui du Nouveau Testament sont peu nombreux. L’étude ancienne de H. A. A. Kennedy reste valable sur de nombreux points, dans la mesure où elle tient compte du cadre épistémologique qu’on vient d’esquisser, considérant les singularités du grec dit « biblique » comme reflétant un des modes d’appropriation de la koinê par un groupe de locuteurs spécifique: H. A. A. Kennedy, Sources of New Testament Greek or The Influence of the Septuagint on the Vocabulary of the New Testament, Édimbourg 1895. Dans sa grammaire du grec de la Septante, H. B. Swete fournit quelques indications: H. B. Swete, « Part III. Literary Use, Value, and Textual Condition of the Greek Old Testament. Chapter IV. The Greek Versions as aids to Biblical Study. B. » in: idem, An Introduction to the Old Testament in Greek, Cambridge 1902 (deuxième édition), 450-457. Plus récemment, M. Silva s’essaye à la confection d’une liste de mots utilisés d’une manière spécifique dans le Nouveau Testament et précise à chaque entrée si cet usage est repris à la Septante: M. Silva, Semantic Change and Semitic Influence in the Greek Bible. With a Study of the Semantic Field of « Mind » (Ph.D. Thesis), University of Manchester 1972. Voir aussi idem., Biblical Words and their Meaning. An Introduction to Lexical Semantics. Revised and expanded edition, Grand Rapids 1994, 56-68. À l’heure actuelle se multiplient les publications sur un mot ou un champ sémantique du grec néotestamentaire qui poussent l’investigation jusque dans la Septante. Le point de départ est encore constitué par le dictionnaire de G. Kittel et alii (éds.), Theologisches Wörterbuch zum Neuen Testament (ThWNT), Stuttgart 1953-1973 (9 volumes et 2 volumes d’index) – dont on use en tenant compte des réserves émises notamment par J. Barr, The Semantics of Biblical Language, Londres 1961, quant à la tendance des rédacteurs du ThWNT à confondre mot et concept. Le dictionnaire de W. Bauer, dans sa version anglaise surtout, fournit des références bibliques et des indications bibliographiques pour les mots du Nouveau Testament attestés déjà dans la Septante: W. Bauer, Griechisch-deutsches Wörterbuch zu den Schriften des Neuen Testaments und der frühchristlichen Literatur, Berlin/New York 1988 (sixième édition révisée) ou W. Bauer / F. W. Danker / W. F. Arndt / F. W. Gingrich, A Greek-English Lexicon of the New Testament and other Early Christian Literature, Chicago/Londres 2000 (troisième édition révisée). Le Historical and Theological Lexicon of the Septuagint (HTLS), en cours de parution sous la direction d’E. Bons et J. Joosten, rendra sans nul doute de grands services, puisque chaque entrée comportera une section « Nouveau Testament ».

442

1. Les sémitismes lexicaux de la Septante et le Nouveau Testament

27. Le vocabulaire de la Septante dans le Nouveau Testament

« Torah »; ἄγγελος, « messager », désignant précisément l’« ange »; δόξα, « opinion », renvoyant dans la Bible à la « gloire » de Dieu; εὐλογέω, « dire du bien, faire un éloge », devenu « bénir »; διαθήκη, « testament », signifiant désormais « alliance » dans le judaïsme de langue grecque. Le sens de ces mots grecs s’est élargi à celui des termes hébreux auxquels ils correspondent: ‫תּוֹ ָרה‬, ‫ַמְלאְַך‬, ‫ָכּבוֹד‬, ‫ָבּ ַרְך‬, ‫ְבּ ִרית‬. Lorsqu’un auteur du Nouveau Testament use de ces termes, il ne revendique pas l’héritage des Septante. Ce ne sont pas des septantismes, mais des termes d’une langue de spécialité en usage dans le judaïsme de langue grecque, puis parmi les premiers chrétiens. Ils ont un certain degré de technicité, qui se manifeste par une stabilité du signifié, d’un texte et d’un auteur à l’autre, non seulement à travers les livres du Nouveau Testament, mais aussi, plus généralement, dans la littérature juive de langue grecque, notamment dans les écrits intertestamentaires rédigés en grec ou dans les œuvres de Philon d’Alexandrie et de Flavius Josèphe. Le mot καταπέτασμα, utilisé dans la version grecque du Pentateuque pour désigner un voile du sanctuaire, est repris en ce sens dans le Nouveau Testament (Mc 15,38 // Mt 27,51 // Lc 23,45; He 6,19; 9,3), mais aussi dans la Lettre d’Aristée, chez Philon et chez Josèphe 5. Renvoyant immédiatement à une réalité culturelle extralinguistique, ce terme technique du judaïsme hellénophone véhicule son sens indépendamment du contexte où ils s’insère: il n’est pas lié à un livre ou un verset en particulier. Les mots concernés n’ont pas toujours un sens technique dès la Septante. Διάβολος, dans la traduction de Job 1–2 ou de Zacharie 3, ne désignait sans doute pas encore la figure satanique de l’Adversaire: c’est une traduction contextuelle, qui n’évoque le « diable » qu’à une époque plus tardive. Le sens de certains mots s’est encore modifié entre la Septante et le Nouveau Testament, acquerrant un degré de technicité supplémentaire: συναγωγή, l’équivalent de l’hébreu ‫ ָקָהל‬désignant le rassemblement de la communauté, renvoie par la suite au bâtiment qui héberge le culte communautaire (Mc 1,39; Mt 4,23; Lc 4,44, etc.; voir déjà Susanne 28). Ainsi se constitue une tradition terminologique vivante, à partir de mots dont les changements de sens attestent qu’ils sont bien entrés dans l’usage et ne sont pas transposés directement de la Septante dans le Nouveau Testament. Ces mots se distinguent des sémitismes de traduction cantonnés au texte écrit des Septante et qui ne sont jamais entrés dans l’usage. Le Nouveau Testament est composé de livres rédigés librement en grec. Il ne reprend pas, loin s’en faut, tous les sémitismes lexicaux de la Septante, dans la mesure où bon nombre d’entre eux résultent de l’application d’une technique de traduction: tel mot hébreu incompris du traducteur est simplement translittéré; tel autre terme est rendu systématiquement par un unique mot grec, même lorsqu’un des sens de l’hébreu ne coïncide pas avec ceux du vocable grec correspondant; lorsque l’hébreu dispose de plusieurs mots rattachés à une même racine, les traducteurs ont tendance à façonner une famille de mots parallèle en grec, quitte à forger des néologismes par dérivation 6.

5. 6.

Lettre d’Aristée, 86; Philon d’Alexandrie, De vita Mosis, 2, 87; Flavius Josèphe, Antiquités juives, 8, 90. Voir Kennedy, Sources of New Testament Greek, 113. Voir M. Harl / G. Dorival / O. Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien (Initiations au christianisme ancien), Paris 1988, 246-251. 1. Les sémitismes lexicaux de la Septante et le Nouveau Testament

443

27. Le vocabulaire de la Septante dans le Nouveau Testament

Les auteurs du Nouveau Testament ont peu recours aux mots ou aux sens qui sont le produit de ces mécanismes. Le mot εἰρήνη, « paix », adopte régulièrement dans la Septante un des sens de l’hébreu ‫ָשׁלו ֹם‬, « bien-être, prospérité »; ce sens est fréquent aussi dans le Nouveau Testament, mais il ne s’y trouve guère que dans des formules de salutation ou de prise de congé (Mc 5,34; Lc 7,50; 10,5; etc.). Si dans la Bible hébraïque ‫ ָים‬, « mer », désigne régulièrement l’occident, le mot grec θάλασσα n’a pas cette signification. Par équivalence formelle, les Septante utilisent néanmoins θάλασσα dans les passages où ‫ ָים‬renvoie à l’ouest (Gn 12,8; 13,14; etc.). Mais pour désigner le point cardinal, les évangélistes optent pour le mot grec usuel, δυσμαί (Mt 8,11 // Lc 13,29; Mt 24,27; Lc 12,54; voir aussi Apc 21,13), qui signifie « le couchant ». On ne trouvera pas non plus, dans les livres du Nouveau Testament, tous les néologismes liés à la création de nouvelles familles de mots grecs. Le verbe ἁγιάζειν, « sanctifier », dérivé de l’adjectif ἅγιος, est repris aussi bien dans la version grecque du Notre Père (Mt 6,9; Lc 11,2) que dans le corpus paulinien. Mais un autre verbe formé à l’aide du même suffixe, ἀκουτίζειν, est absent du Nouveau Testament. Les traducteurs des Psaumes l’ont tiré du verbe ἀκούω, « entendre », équivalent de ‫שׁמע‬, pour rendre la forme hiphil du verbe hébreu, « faire entendre » (Ps 50(51),10; 65(66),8; 75(76),9). Il eût été en bonne place en 1Jean 1,3, par exemple: ὃ ἑωράκαμεν καὶ ἀκηκόαμεν ἀπαγγέλλομεν καὶ ὑμῖν, « Ce que nous avons vu et entendu, nous vous le faisons entendre à vous aussi ». Mais comme dans ce verset – ou dans la Septante d’Amos et de Nahum (Am 3,9; Na 1,15 (2,1)) –, le Nouveau Testament privilégie les composés bien usités d’ἀγγέλλω pour exprimer l’acte par lequel bonne nouvelle ou jugement sont annoncés aux hommes. Le Nouveau Testament ne reprend pas à la Septante les sémitismes de traduction caractérisés. Ils sont souvent rares, parfois intelligibles seulement dans le contexte où ils s’insèrent et dont ils ne peuvent être dégagés pour une autre utilisation dans un nouveau texte. Les sémitismes de la Septante qu’on retrouvera dans le Nouveau Testament sont les termes réguliers de la langue de spécialité, auxquels s’ajoutent quelques mots non techniques dont le champ sémantique s’est élargi à celui de leur original hébreu, tel ἄρτος, « pain », qui finit par désigner plus largement la nourriture, à l’instar de l’hébreu ‫( ֶ֫לֶחם‬Es 65,25; Mc 3,20). Ces mots ne sont sans doute pas ou plus perçus comme des extraits des Écritures juives. Pour certains, ils ne sont peut-être pas même l’invention des Septante. Les mots grecs du vocabulaire religieux ont pu préexister à la traduction du Pentateuque dans le cadre d’une pratique orale déjà installée, notamment dans le judaïsme alexandrin 7, et lorsque Marc, par exemple, use du mot ἄρτος au sens de « nourriture », il se peut qu’il s’agisse d’un aramaïsme, venu de ‫ַלְחָמא‬, plutôt que d’un hébraïsme hérité des Septante 8. Les auteurs du Nouveau Testament se sont approprié ces termes aussi bien en raison de l’usage oral qui en est fait à la synagogue ou au quotidien qu’à travers la lecture assidue des Écritures dans leur version grecque. C’est sur ce terrain des sémitismes que le vocabulaire du Nouveau Testament se rapproche le plus visiblement de celui de la Septante. C’est aussi sous la rubrique « Sémitismes » des grammaires du Nouveau Testament qu’on classe d’ordinaire les traits 7.

8.

Voir J. Joosten, « Le vocabulaire de la Septante et la question du sociolecte des Juifs alexandrins. Le cas du verbe εὐλογέω, « bénir » » in: E. Bons / J. Joosten, éds, Septuagint Vocabulary. Pre-History, Usage, Reception (Septuagint and Cognate Studies 58), Atlanta 2011, 13-23. Silva, Biblical Words, 77.

444

1. Les sémitismes lexicaux de la Septante et le Nouveau Testament

27. Le vocabulaire de la Septante dans le Nouveau Testament

du grec néotestamentaire évoquant la Septante 9: les septantismes sont souvent considérés comme une catégorie parmi les sémitismes. Tel est bien le premier lien, indirect, qu’on peut tisser entre le vocabulaire de la Septante et le Nouveau Testament. Mais cette façon de les mettre en perspective ne rend pas justice à la manière dont la Septante a exercé son influence sur les auteurs du Nouveau Testament. Elle est un des canaux par lesquels transitent des mots de l’hébreu avant d’entrer dans le Nouveau Testament; mais elle est aussi un texte littéraire à la consistance propre, revêtant pour les premiers chrétiens une autorité particulière en tant qu’Écriture sainte. Étudier le vocabulaire de la Septante dans le Nouveau Testament n’équivaut pas à recenser des hébraïsmes lexicaux: il s’agit aussi de mesurer jusqu’à quel point les auteurs du Nouveau Testament, lorsqu’ils utilisent des termes septantiques, s’inscrivent dans un rapport direct à la Septante en tant que corpus littéraire. C’est alors que le mot de septantisme s’applique à bon droit.

2. Les septantismes lexicaux: l’influence directe de la Septante sur le vocabulaire du Nouveau Testament Le texte de la Septante a pu réguler dans une certaine mesure le choix des mots effectué par les auteurs du Nouveau Testament 10, en raison de leur fréquentation assidue de la version grecque des Écritures. La distinction saussurienne entre « langue » et « parole » est ici opératoire. Pour Ferdinand de Saussure, la « langue », sur le plan du vocabulaire, désigne l’ensemble des mots disponibles; la sélection qu’effectue un locuteur à l’intérieur de cet ensemble caractérise sa « parole » et relève du style 11. En termes saussuriens, les auteurs du Nouveau Testament usent ici ou là d’un vocable en vertu d’un choix stylistique dicté par une intention plus ou moins consciente de faire écho à la Septante. Les éléments de vocabulaire concernés ne s’analysent pas au niveau de la langue, comme des mots et des sens empruntés au sémitique ou des termes techniques, mais sur le plan stylistique, comme des allusions au texte grec de la Septante ou des septantismes. Non que la Septante constituerait le dictionnaire des auteurs du Nouveau Testament: force est de constater qu’un grand nombre de mots de la Septante ne sont pas attestés dans le Nouveau Testament. Outre les sémitismes de traduction non entrés dans l’usage, des mots grecs courants que les traducteurs ont choisis pour exprimer des réalités d’ordre théologique ou spirituel ne sont pourtant pas repris par les auteurs du Nouveau Testament dans des contextes analogues. Ainsi sont absents du Nouveau Testament le verbe δεσπόζω, « gouverner », utilisé dans la Septante pour désigner le gouvernement de Dieu sur la terre, ou des composés comme πολυέλεος « rempli de 9. Voir J. H. Moulton, « Appendix. Semitisms in the New Testament » dans: idem, A Grammar of New Testament Greek. Volume 2. Accidence and Word-Formation with an Appendix on Semitisms in the New Testament, Édimbourg 1919-1929, 411-485; F. Blass / A. Debrunner / F. Rehkopf, « § 4. Semitismen » dans: idem, Grammatik des neutestamentlichen Griechisch, Göttingen 2001 (dix-huitième édition), 4-6. 10. Swete, An Introduction, 453. 11. F. de Saussure, Cours de linguistique générale, Paris 1995 (première édition 1916), chapitres IIIIV, 23-39. 2. Les septantismes lexicaux

445

27. Le vocabulaire de la Septante dans le Nouveau Testament

pitié » 12. Inversement, βαστάζω, « porter », usité depuis Homère, se trouve près de trente fois dans le Nouveau Testament alors qu’on ne le croise qu’à quelques reprises dans la Septante 13. Certes, la Septante constitue un corpus bien plus considérable que le Nouveau Testament, ce qui explique en partie l’écart constaté. Mais s’ajoute que quatre siècles séparent la traduction du Pentateuque et les livres les plus tardifs du Nouveau Testament, un temps durant lequel le stock de mots en usage et leur sémantisme dans la langue courante ne sont pas restés figés. Lorsque les auteurs du Nouveau Testament écrivent, ils ont recours d’habitude aux mots grecs et aux sens usuels à leur époque, sans multiplier les archaïsmes. Il arrive même qu’ils corrigent le vocabulaire de la Septante lorsqu’il leur paraît daté: les remplacements de ἰσχύν, « force », par son synonyme δύναμιν dans la citation d’Exode 9,16 en Romains 9,17, ou de ταχινοί, « rapides », par ὀξεῖς dans celle d’Esaïe 59,7 en Romains 3,15 pourraient être attribués à une intention de ce genre 14. Certains des décalages observés s’expliquent sans doute aussi par des variations dans les niveaux de langue. Les auteurs du Nouveau Testament n’épuisent pas toutes les ressources que la Septante leur offre en matière de vocabulaire. Mais les septantismes lexicaux y sont pourtant nombreux. Les mots concernés, puisqu’ils relèvent du style propre à chaque auteur, sont répartis plus inégalement dans le Nouveau Testament que les termes incontournables de la langue de spécialité. L’enquête lexicologique devrait être menée auteur par auteur: c’est à ce niveau qu’on peut mesurer à quel point et de quelle manière la Septante inspire le choix des mots effectué dans le Nouveau Testament 15 – au même titre que dans d’autres œuvres de langue grecque issues du judaïsme du Second Temple. Le vocabulaire de Paul est différent de celui de Luc, de Philon d’Alexandrie, de Flavius Josèphe ou du Testament de Job, et les uns se distinguent des autres notamment dans leur façon de s’approprier les mots choisis d’abord par les Septante. Cette enquête approfondie ne peut être menée ici. On se contentera de signaler quelques cas représentatifs des façons variées dont la Septante a exercé son influence sur le style des auteurs du Nouveau Testament, tel qu’il se manifeste dans le choix des mots. La Septante a pu faire préférer tel mot à d’autres termes grecs proches pour exprimer une notion théologique; elle laisse aussi son empreinte lorsqu’un auteur use d’un mot charactéristique de la Septante pour rappeler un passage entier des Écritures juives ou pour en imiter le style dans leur version grecque. 12. Kennedy, Sources of New Testament Greek, 141. 13. T. K. Abbott, Essays, Chiefly on the Original Texts of the Old and New Testaments, Londres 1891, 70. 14. D.-A. Koch, Die Schrift als Zeuge des Evangeliums (Beiträge zur historischen Theologie 69), Tübingen 1986, 141 et 144; Ch. D. Stanley, Paul and The Language of Scripture. Citation Technique in the Pauline Epistles and Contemporary Literature (Society for New Testament Studies Monograph Series 69), Cambridge 1992, 96 et 109. 15. Voir par exemple J. H. Moulton, « Part I. General Introduction. Sounds and Writing. Introduction. (2) Aramaisms and Hebraisms » dans: idem, A Grammar of New Testament Greek. Volume 2. Accidence and Word-Formation with an Appendix on Semitisms in the New Testament, Édimbourg 1919-1929, 14-34; Th. Nägeli, Der Wortschatz des Apostels Paulus. Beitrag zur sprachgeschichtlichen Erforschung des Neuen Testaments, Göttingen 1905.

446

2. Les septantismes lexicaux

27. Le vocabulaire de la Septante dans le Nouveau Testament

2.1 Le vocabulaire théologique De manière générale, les auteurs du Nouveau Testament forgent une terminologie qui leur est propre pour proclamer leur message. Mais la parenté notionnelle et conceptuelle entre le judaïsme hellénophone et le christianisme naissant se reflète régulièrement au niveau du lien linguistique qu’entretiennent la Septante et le Nouveau Testament, notamment à travers le vocabulaire théologique. Les traducteurs de la Bible hébraïque privilégient souvent un unique mot parmi plusieurs équivalents grecs possibles pour un terme hébreu, lorsque le terme concerné ne relève pas à l’évidence du vocabulaire religieux spécialisé, mais sert toutefois à décrire les rapports entre Dieu et l’humanité et prend donc, dans le contexte biblique, une connotation théologique. Ainsi des notions abstraites telles que ‫ֱאֶמת‬, ‫ֵאמוּ ָנה‬, ‫ְצ ָדָקה‬ « vérité », « foi », « justice », sont-elles rendues régulièrement, dans le grec des Septante, par ἀλήθεια, πίστις, δικαιοσύνη. Héritiers de la pensée théologique du judaïsme, les auteurs du Nouveau Testament choisissent d’ordinaire, pour évoquer ces notions, le même mot grec que celui privilégié déjà par les Septante, à l’exclusion d’autres termes possibles. La notion de sainteté se traduit par ἅγιος, « saint », et ses composés, plus volontiers que par ἱερός ou ὅσιος, aussi bien dans la Septante que dans le Nouveau Testament. L’équivalent septantique régulier de l’hébreu ‫אהב‬, « aimer », est ἀγαπάω, et c’est sans doute ce choix qui guide Paul ou la communauté johannique lorsqu’ils utilisent le nom ἀγάπη, plutôt que φιλία, pour développer leur propre conception de l’amour divin. Pour parler de la révélation divine faite aux hommes, la Septante utilise ἀποκαλύπτω, « dévoiler » (Dn 2,28; 1S 3,21; Es 56,1), et le Nouveau Testament lui emboîte le pas (Mt 11,25; Ro 1,18; Eph 5,3; etc.), 16 laissant de côté d’autres composés au sens analogue comme ἀνακαλύπτω ou διακαλύπτω. L’usage répété des mêmes termes grecs pour exprimer des notions de théologie juive et chrétienne finit pas leur imprimer une densité particulière, si bien qu’ils sont susceptibles d’acquérir un certain degré de technicité. Il est malaisé de tracer une ligne nette entre ces mots et les termes de la langue de spécialité et de mesurer jusqu’à quel point des notions telles que la foi, l’amour, la vérité acquièrent déjà dans le judaïsme hellénophone un contenu nouveau puis figé, si bien qu’an au point que le signifié nouveau correspondrait, sur le plan du vocabulaire, un signifiant unique qui s’imposerait aux auteurs du Nouveau Testament, à la suite des Septante. Cependant, une comparaison avec d’autres textes issus du judaïsme hellénistique montre qu’on a bien affaire à deux catégories de mots distinctes. Tandis qu’il use des termes techniques δόξα ou νόμος, Philon d’Alexandrie fait d’autres choix que les auteurs du Nouveau Testament pour désigner telle notion théologique, employant abondamment ὅσιος 17, n’hésitant pas à user de φιλία ou ἔρως pour parler de l’amour de Dieu 18, se servant de διακαλύπτω ou ἀνακαλύπτω aussi volontiers que d’ἀποκαλύπτω pour parler du dévoilement des réalités divines 19. Philon s’éloigne ici de l’usage septantique; il fait 16. 17. 18. 19.

Kennedy, Sources of New Testament Greek, 104. Voir par exemple Philon d’Alexandrie, De specialibus legibus, 2, 42. Voir notamment Philon d’Alexandrie, De fuga et inventione, 58-59. Philon d’Alexandrie, De ebrietate, 139; De migratione Abrahami, 126; De somniis, 1, 87; De Josepho, 90 et 106. 2. Les septantismes lexicaux

447

27. Le vocabulaire de la Septante dans le Nouveau Testament

bien voir, a contrario, que le choix fait par les auteurs du Nouveau Testament au moment d’évoquer ces notions est guidé par le texte écrit de la Septante, et non par quelque terminologie standardisée. On peut sans doute parler ici de septantismes.

2.2 Les allusions lexicales Le terme de septantisme est encore mieux approprié lorsque la reprise d’un mot des Septante dans le Nouveau Testament convoque toute une péricope ou un verset précis des Écritures, ou quand un auteur du Nouveau Testament donne une couleur septantique à son vocabulaire en usant régulièrement de mots ou de sens marqués au coin du texte de la Septante, y compris de termes non techniques et non théologiques, par imitation du style des traducteurs. La théorie littéraire de l’intertextualité s’ajoute ici à la distinction du linguiste entre langue et parole, pour rendre compte des faits observés: elle permet d’insister sur les ressources propres du texte écrit, lorsque la Septante ne régule plus seulement le choix des mots, mais devient présente en tant que texte à l’intérieur d’un autre texte. La possibilité qu’un lien d’intertextualité puisse se tisser d’un seul mot est discutée: l’intertexte ne serait décelable qu’au niveau d’éléments déjà structurés, au-delà du vocable isolé 20. Cependant, le repérage d’échos de la Septante à l’échelle du mot est rendu possible dans le Nouveau Testament en raison de la multiplication des renvois, aussi bien à travers d’autres termes que par des citations, une phraséologie ou des expressions idiomatiques reprises à la Septante. Un effet de structuration se produit par le caractère répété du phénomène. Pris dans ce complexe plus vaste, un mot unique suffit à faire affleurer le texte de la Septante dans le Nouveau Testament. Un passage précis de la Septante est évoqué parfois à travers un mot unique dont use un auteur du Nouveau Testament. C’est le cas notamment chez Paul, qui affectionne ce genre d’allusions lorsqu’il entreprend une démonstration par l’Écriture 21. En Romains 9,17 – on l’a vu – il cite Exode 9,16: λέγει γὰρ ἡ γραφὴ τῷ Φαραὼ ὅτι εἰς αὐτὸ τοῦτο ἐξήγειρά σε ὅπως ἐνδείξωμαι ἐν σοὶ τὴν δύναμίν μου καὶ ὅπως διαγγελῇ τὸ ὄνομά μου ἐν πάσῃ τῇ γῇ, « Car l’Écriture dit au pharaon: Je t’ai suscité tout exprès pour montrer en toi ma puissance et pour que mon nom soit annoncé par toute la terre ». Au verset suivant intervient le verbe σκληρύνω: ἄρα οὖν ὃν θέλει ἐλεεῖ, ὃν δὲ θέλει σκληρύνει, « Ainsi donc, il a compassion de qui il veut, et il fait s’obstiner qui il veut ». Or le verbe σκληρύνω est utilisé à quatorze reprises dans la première partie du livre de l’Exode, comme équivalent tantôt de ‫חזק‬, tantôt de ‫קשׁה‬, pour évoquer l’endurcissement de Pharaon face aux entreprises de Moïse et d’Aaron afin de le contraindre à laisser le peuple quitter l’Égypte (Ex 4,21; 7,3.22; 8,15; 9,12.35; 10,1.20.27; 11,10; 13,15; 14,4.8.17). La mention de Pharaon au verset 17 et la citation 20. Comparer G. Genette, Palimpsestes. La littérature au second degré, Paris 1982, 8, qui parle d’intertextualité y compris dans le cas d’une simple « allusion » à un autre texte, et L. Jenny, « La stratégie de la forme », Poétique 27 (1976), 257-281 (262), selon lequel il s’agit de « repérer dans un texte des éléments structurés antérieurement à lui, au-delà du lexème ». 21. H. Vollmer, Die Alttestamentliche Zitate bei Paulus textkritisch und biblisch-theologisch gewürdigt nebst einem Anhang ueber das Verhältnis des Apostels zu Philo, Fribourg-en-Brisgau/ Leipzig 1895, 11-12.

448

2. Les septantismes lexicaux

27. Le vocabulaire de la Septante dans le Nouveau Testament

d’Exode 9,16 appellent ce second lien d’intertextualité qui convoque, d’un seul mot, la geste des dix plaies et de la sortie d’Égypte. Même en l’absence de citation scripturaire proche, un mot repris à la Septante rappelle quelquefois toute une péricope: lorsque Paul affirme en Romains 8,32 que Dieu « n’a pas épargné son propre fils », τοῦ ἰδίου υἱοῦ οὐκ ἐφείσατο, il renvoie peut-être au récit du sacrifice d’Isaac où l’ange dit à Abraham: οὐκ ἐφείσω τοῦ υἱοῦ σου τοῦ ἀγαπητοῦ δι᾽ ἐμέ, « À cause de moi, tu n’as pas épargné ton fils bien-aimé » (Gn 22,12). Ici et là, Paul ou Jean 22 évoquent d’un mot tel passage précis de la Septante à l’appui d’un développement théologique. D’autres auteurs du Nouveau Testament saturent leur propos de mots et de tournures repris à la Septante dans le seul but de recréer à l’aide d’artifices littéraires une langue perçue comme sacrée. Luc est coutumier du fait, notamment dans les premiers chapitres de l’évangile et dans les discours des Actes 23: la reprise d’un vocabulaire perçu comme typique des Septante est un des moyens par lesquels il donne à son texte une couleur biblique. Il fait siens même des termes sans connotation théologique aucune, par exemple le mot ῥῆμα pris au sens de « chose », tel qu’il est usité dans la Septante en tant qu’équivalent de l’hébreu ‫( ָדָּבר‬Lc 1,37.65; 2,19.51; Ac 10,37). L’imitation passe par l’emploi de mots et de sens rares, qui ne sont guère attestés en dehors de la Septante et de l’œuvre lucanienne. Ainsi l’adjectif σκληροτράχηλος (Ac 7,51), que Luc reprend à la version grecque du Pentateuque, où le mot est utilisé métaphoriquement pour désigner le peuple ‫ְקֵשׁה־ֹע ֶרף‬, « à la nuque raide » (Ex 33,3.5; 34,9; Dt 9,6.13; voir aussi Pr 29,1; Si 16,11; Ba 2,30). D’autres mots, au contraire, tels la Septante par leur fréquence inhabituelle chez Luc – rappellent ἐνώπιον, « en face de », ou λαός, « peuple », qui apparaissent aussi à chaque page dans la Septante. Ces éléments de vocabulaire, traces de la Septante à l’intérieur du texte néotestamentaire, n’apparaissent comme des liens d’intertextualité que s’ils sont considérés dans le cadre plus vaste d’un réseau de mots, d’expressions, de locutions, de citations de la Septante. Ils sont plus fréquents chez certains auteurs du Nouveau Testament que chez d’autres 24. La notion de style et la théorie de l’intertextualité, en mettant l’accent sur les choix individuels de chacun, font ressortir les variations qui existent à l’intérieur même du Nouveau Testament quant à la manière dont le vocabulaire de la Septante exerce son influence.

22. Voir Silva, Biblical Words, 93, sur l’usage de ὑψόω dans l’évangile de Jean, en écho à la Septante d’Esaïe 52,13. 23. Sur l’usage que Luc fait de la Septante, on consultera notamment A. Wifstrand, « Luke and the Septuagint » dans: idem, Epochs and Styles. Selected Writings on the New Testament, Greek Language and Greek Culture in the Post-Classical Era (édité par L. Rydbeck et S. E. Porter, traduit par D. Searby) (Wissenschaftliche Untersuchungen zum Neuen Testament 179), Tübingen 2005 (première publication 1940), 28-45; H. F. D. Sparks, « The Semitisms of St. Luke’s Gospel » JTS 44 (1943), 129-138 et « The Semitisms of Acts » JTS ns 1 (1950), 16-28; J. A. Fitzmyer, The Gospel According to Luke (I-IX) (The Anchor Bible), New York 1979, 114-125. 24. À côté du corpus paulinien et de l’œuvre lucanienne, Hébreux et 1 Pierre comportent également de nombreuses allusions de ce type au lexique des Septante (Swete, An Introduction, p. 403-404; Moulton, A Grammar of New Testament Greek. Volume 2, 25-26). 2. Les septantismes lexicaux

449

27. Le vocabulaire de la Septante dans le Nouveau Testament

3. Conclusion La Bible de langue grecque, Septante et Nouveau Testament, ne constitue pas un corpus homogène d’un point de vue lexicologique. Les mots de la Septante sont pour la plupart le fruit d’un travail de traduction effectué sur un original sémitique, achevé pour l’essentiel dès les IIIe et IIe siècles avant J.-C., tandis que le Nouveau Testament rassemble des compositions nouvelles faites directement en grec entre la moitié du Ier siècle et la moitié du IIe siècle après J.-C. La Septante ne tient pas lieu en tant que telle de stock de mots à caractère normatif à l’usage des auteurs du Nouveau Testament. On décèle toutefois des liens entre le lexique de la Septante et celui des auteurs du Nouveau Testament. Les éléments de vocabulaire concernés s’analysent pour les uns au niveau de la langue, comme des mots et sens d’emprunt ou des termes techniques produits par le judaïsme hellénophone: Septante et Nouveau Testament se font le réceptacle d’un usage qu’on devait pouvoir observer aussi à l’oral, notamment dans le cadre du culte synagogal. D’autres mots reflètent une influence directe du texte écrit de la Septante sur les auteurs du Nouveau Testament: les notions de style et d’intertextualité permettent d’en rendre compte en insistant sur la spécificité et la consistance propres de deux monuments littéraires. Ces principes d’analyse se recoupent dans bien des cas. On a fait ressortir ici toutefois ce qui les distingue: l’approche linguistique s’attarde sur les sémitismes communs à la langue des traducteurs et à celle des auteurs du Nouveau Testament; à l’autre bout, la théorie littéraire de l’intertextualité est attentive aux phénomènes de reprise, d’imitation et d’allusion par lesquels le Nouveau Testament rend hommage à la Septante en tant qu’Écriture sainte.

450

3. Conclusion

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik Thomas J. Kraus

1. Vorüberlegungen 1.1 Zielsetzung und Methode Was ist von dem, was der Grammatik der Septuaginta zuzurechnen ist, ebenso im Neuen Testament zu finden? Anders formuliert, was haben beide, LXX und NT, gemein, was als typisch für beide Textkorpora gelten kann? Diese Fragen ziehen bereits eingangs weitere nach sich, die in erster Linie methodischer Art sind und in einem Beitrag wie dem vorliegenden allenfalls angerissen werden können: Was ist grundsätzlich unter Grammatik zu verstehen? Kann es überhaupt die Grammatik der LXX und die Grammatik des NT geben? Können LXX und NT im Hinblick auf Grammatik – und das nicht nur hinsichtlich eines spezifischen Einzelphänomens – generell in Beziehung gesetzt werden? Und wenn ja, auf welche Art und Weise bleibt dies wissenschaftlich vertretbar und geht dabei nicht das Spezifische unter bzw. gewinnt das weniger signifikante Einzelphänomen zu große Bedeutung? 1 In Bezug auf die LXX zeigt ein Überblick über die einschlägigen Titel, dass die wenigen vorhandenen Grammatiken unvollständig und veraltet sind. 2 Darüber hinaus stellt sich als weiteres Problem, welche Textform(en) der Septuaginta denn überhaupt

1.

2.

Zum Folgenden vgl. S. Jellicoe, The Septuagint and Modern Study, Oxford 1968, 332-335; C. T. Fritsch, »The Future of Septuagint Studies: A Brief Survey« BIOSCS 3 (1970), 4-8, bes. 5-6; P. Walters (formerly Katz), The Text of the Septuagint. Its Corruptions and Their Emendation, Cambridge 1973, 1-13; R. Sollamo, »Prolegomena to the Syntax of the Septuagint« in: R. Sollamo / S. Sipilä (Hg.), Helsinki Perspectives on the Translation Technique of the Septuagint. Proceedings of the IOSCS Congress in Helsinki 1999 (Publications of the Finnish Exegetical Society 82), Göttingen 2001, 23-42, bes. 26-28. Ferner vgl. die bibliographischen Angaben in T. J. Kraus, Sprache, Stil und historischer Ort des zweiten Petrusbriefes (WUNT 2,136), Tübingen 2001, 419-421. Vgl. F. C. Conybeare / St. G. Stock, A Grammar of Septuagint Greek (reprint of Selections from the Septuagint, Boston 1905; H. St. J. Thackeray, A Grammar of the Old Testament in Greek According to the Septuagint, vol. I, Introduction, Orthography and Accidence, Cambridge, 1909; R. Helbing, Grammatik der Septuaginta, Göttingen, 1907; R. Helbing, Die Kasussyntax der Verba bei den LXX. Ein Beitrag zur Hebraismenfrage und zur Syntax der Koine, Göttingen 1928; F.-M. Abel, Grammaire du grec biblique suivie d’un choix de papyrus (EtB 7), Paris 1927; Walters, The Text of the Septuagint. S. außerdem den Abschnitt »Kleine Septuaginta-Grammatik. Volksssprache und Hebraismen« in: F. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Eine Einführung in die Septuaginta (Münsteraner Judaistische Studien 9), Münster 2001, 141-164. 1. Vorüberlegungen

451

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

grammatikalisch erfasst werden und wie dies dann zu erfolgen hat. 3 Für das NT ist die Situation besser, liegen doch – neben zahlreichen Spezialuntersuchungen – zumindest Grammatiken vor, die Hilfsmittel und Standardwerke zugleich und die auch für die Erforschung sprachlicher Phänomene der LXX von Relevanz sind. So wurde die Sprache des Neuen Testaments in den Kontext der damaligen sprachlichen Umwelt eingebettet 4, und Grammatiken wie die von Blass, Debrunner und Rehkopf sowie von Moulton sind zu Klassikern geworden. 5 Gerade Arbeiten, die nicht ausschließlich dem NT oder der LXX gewidmet sind, sondern grundsätzlich auf die Sprache der griechischen Bibel abheben, 6 verdienen besondere Beachtung, insbesondere wenn es um Entscheidungen im Einzelfall geht. Ähnliches gilt für die Untersuchung linguistischer Spezifika, wie etwa die griechischen Verba. 7 Da es auch darum geht, unterschiedliche Bereiche von griechischen Textzeugnissen zueinander in Relation zu setzen, sind Grammatiken der Papyri, 8 der Inschriften 9 3. 4.

5.

6.

7.

8.

9.

So T. Muraoka, »A Septuagint Greek Grammar, but of which Text-form or -forms?« EstBib 51 (1993) 433-458. Vgl. L. Radermacher, Neutestamentliche Grammatik. Das Griechisch des Neuen Testaments im Zusammenhang mit der Volkssprache (HNT 1, zweite Auflage), Tübingen 1925; A. T. Robertson, A Grammar of the Greek New Testament in the Light of Historical Research, Nashville, TN 1934. So die Beurteilung durch Sollamo, »Prolegomena«, 27. Vgl. F. Blass / A. Debrunner, Grammatik des neutestamentlichen Griechisch (bearb. v. F. Rehkopf) Göttingen 171990; J. H. Moulton, A Grammar of New Testament Greek, vol. I, Prolegomena, Edinburgh 21906; vol. II, Accidence and Word-Formation. With an Appendix on Semitisms in the New Testament (J. H. Moulton und W. F. Howard), Edinburgh 1920 (Nachdruck 1976); vol. III, Syntax (N. Turner), Edinburgh 1963; vol. III, Style (N. Turner), Edinburgh 1976. Vgl. z. B. M. Zerwick, Biblical Greek Illustrated by Examples, transl. J. Smith (Scripta Pontificii Instituti Biblici 114), Rom 1963; D. Tabachovitz, Die Septuaginta und das Neue Testament. Stilstudien (Skrifter utgivna av svenska institutet i Athen 8o IV), Lund 1956; K. Beyer, Semitische Syntax im Neuen Testament, Bd. I, Satzlehre, Teil 1 (SUNT 1, zweite Auflage), Göttingen 1968. Hierzu vgl. K. L. McKay, A New Syntax of the Verb in New Testament Greek. An Aspectual Approach (Studies in Biblical Greek 5), New York, NY 1994; S. E. Porter, Verbal Aspect in the Greek New Testament with Reference to Tenses and Moods (Studies in Biblical Greek 1), New York, NY 1989; B. M. Fanning, Verbal Aspect in New Testament Greek (OTM), Oxford 1990. Vgl. E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluß der gleichzeitigen Ostraka und der in Ägypten verfaßten Inschriften, I: Einleitung und Lautlehre, Berlin 1906, I,1, Laut- und Wortlehre (H. Schmoll), Berlin 1970; I,2, Laut- und Wortlehre; Flexionslehre (zweite Auflage), Berlin 1938; I,3, Laut und Wortlehre; Stammbildung, Berlin 21936; II,1, Satzlehre. Analytischer Teil. Erste Hälfte, Berlin 1933; II,2, Satzlehre. Analytischer Teil. Zweite Hälfte, Berlin 1934; II,3, Satzlehre. Synthetischer Teil, Berlin 1934. Die Grammatiken von Gignac und Palmer blieben leider unvollständig. Vgl. F. T. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, vol. II, Morphology (Testi e documenti per lo studio dell’antichità 40,2), Mailand 1982; L. R. Palmer, A Grammar of the Post-Ptolemaic Papyri, vol. I, Accidence and Word-Formation, part I, The Suffixes, London 1945. Ferner von Relevanz T. V. Evans / D. D. Obbink (Hg.), The Language of the Papyri, Oxford 2010. Vorliegende Studien sind aber auf eine bestimmte Zeit und/oder eine bestimmte Region beschränkt. Vgl. z. B. K. Hauser, Grammatik der griechischen Inschriften Lykiens, Diss. Zürich/ Basel 1916; G. Laminger-Pascher, Index Grammaticus zu den griechischen Inschriften Kili-

452

1. Vorüberlegungen

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

und der (klassischen) griechischen Sprache 10 zu konsultieren. Dennoch lösen weder diese Studien noch die genannten nützlichen Grammatiken das eigentliche Problem, das sich hinsichtlich einer Grammatik von LXX und NT stellt: Die sprachliche Diskrepanz zwischen den einzelnen Teilen des NT oder zwischen der Übersetzungsliteratur der LXX und den genuin griechisch verfassten Büchern legt nahe, dass eine umfassende Grammatik weder anzustreben ist noch realistisch umgesetzt werden kann. Darüber hinaus gilt es, die LXX und das NT im Kontext ihrer sprachlichen Umwelt und der sprachgeschichtlichen Entwicklung des Griechischen zu betrachten. Deshalb ist es pragmatisch und realistisch, »keine vollständige Grammatik der von den Übersetzern verwendeten Sprache, sondern nur eine Erläuterung der wichtigsten Abweichungen vom damaligen literarischen Griechisch« 11 verfassen zu wollen. Was hier für die LXX formuliert ist, kann analog auch als Zielsetzung für die Behandlung des NT und folglich beider Textkorpora gelten. Anhand von Tendenzen, Häufigkeiten und Auffälligkeiten ergibt sich durch die Auswahl von relevanten grammatikalischen Phänomenen ein Eindruck dessen, was LXX und NT im Bereich der Grammatik gemein haben. Neben dem Verbindenden gilt es aber auch immer das Individuelle der beiden Korpora und einzelner Texte im Blick zu behalten.

1.2 Abgrenzung Was ist Grammatik? Genauer gefragt: Was gehört dazu und was nicht? Ein Blick in die Inhaltsverzeichnisse von Grammatiken zeigt, dass hierüber kein Konsens herrscht. Kern ist zwar stets die Syntax; dazugerechnet werden aber auch Lautlehre, Wort- bzw. Formenlehre und Stil/Stilistik. Die Lautlehre (und damit, zumindest teilweise, auch ihre Auswirkung auf die Orthographie) widmet sich als eigenständige Disziplin der gesprochenen Sprache, dabei der Beschreibung der Laute (Phonologie) bzw. ihrer Äußerung und Wirkungsweise (Phonetik). Ebenfalls sind Wortbildung und Flexionslehre selbständige Teile einer deskriptiven Grammatik. 12 Wortarten, Wortformen und Wortgruppen und ihre jeweilige Bedeutung und Funktion sowie die Satzlehre selbst bilden beispielsweise die Syntax nach Schwyzer und Debrunner. Der letzte Abschnitt ihrer Grammatik (»Syntaktische Stilistik«) deutet aber eine gewisse Nähe von Grammatik und Stilistik an. Allerdings behandeln sie dort auf gerade einmal vierzehn Seiten nur »stilistische Erscheinungen …, die so häufig angewendet werden, daß sie zum Teil geradezu der gewöhnlichen Syntax angehören« 13 und die in anderen Grammatiken

10.

11. 12. 13.

kiens und Isauriens II (Österreichische Akademie der Wissenschaften – Philosophisch-historische Klasse, Sitzungsberichte 298,3), Wien 1974; K. Meisterhans, Grammatik der attischen Inschriften (dritte Auflage), Berlin 1900; E. Nachmanson, »Syntaktische Inschriftenstudien« Er. 9 (1909) 30-81. Standardwerke sind etwa KBG; KGG; E. Schwyzer, Griechische Grammatik, Bd. 1, Allgemeiner Teil. Lautlehre. Wortbildung. Flexion (HAW 2,1,1, dritte Auflage), München 1959 und E. Schwyzer / A. Debrunner, Griechische Grammatik, Bd. 2, Syntax und syntaktische Stilistik (HAW 2,1,2, fünfte Auflage), München 1988. Siegert, Zwischen Hebräischer Bibel und Altem Testament. Entsprechend werden Lautlehre, Wortbildung und Flexion im allgemeinen Teil von Schwyzers Grammatik abgehandelt. Vgl. Schwyzer, Griechische Grammatik 1. Schwyzer / Debrunner, Griechische Grammatik 2, 698. 1. Vorüberlegungen

453

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

generell unter der Überschrift »Syntax« behandelt werden. 14 Ausgeklammert sind damit linguistische Erscheinungen, die als individuell, bewusst oder als rhetorisch bezeichnet werden können. 15 Deshalb ist eine deskriptive Darstellung grammatikalischer Phänomene Voraussetzung für die nähere Betrachtung der Sprachverwendung eines Autors oder eines Textes, also eines spezifischen Stils, der durch Auswahl und Kombination bestimmter sprachlicher Ausdrucksmittel geprägt ist. 16 So verstanden werden Stil und Stilistik als eigenständig aufgefasst und von der Grammatik selbst abgrenzt.

2. Zur Demonstration: Präposition ἐν mit Dativ und ‫ְבּ‬-instrumenti Was ist zu tun, wenn ein konkretes grammatikalisches Phänomen in einer Textpassage des NT auf ein diesbezügliches Spezifikum in der LXX und darüber hinaus auf das Hebräische verweist? Wie ist es zu bewerten, wenn dieses Phänomen ebenso der sprachinternen Entwicklung des Griechischen geschuldet sein könnte bzw. auch in nicht-biblischen Texten zu beobachten ist? 17 An folgendem Beispiel sollen die Möglichkeiten und Grenzen der Identifizierung eines grammatikalischen Phänomens demonstriert werden, das als für die LXX und darüber hinaus als für die hebräische Sprache spezifisch gelten kann. Aus pragmatischer Sicht ist eine Diskussion der wichtigsten Beobachtungen und Tendenzen des Phänomens ‫ְבּ‬-instrumenti 18 sinnvoller als eine umfassende Darbietung und Erörterung aller in Frage kommender Stellen in der LXX und im NT. Die Berücksichtigung des sprachgeschichtlichen Kontexts sowie der Vergleich mit weiteren Sprachmaterial (z. B. klassische Literatur, literarische und semi-literarische Texte mit Alltags- bzw. Volkssprache und dokumentarische Papyri), schließlich die Beachtung relevanter Alternativen zum Ausdruck des Instrumentalis (z. B. bloßer Dativ, Ersatz der klassischen Genitivkonstruktion durch Dativ und, möglicherweise, dativus causae) 14. Vgl. Schwyzer / Debrunner, Griechische Grammatik 2, 698-712. So z. B. Reduplikation, Paronomasie, Asyndeton, Breviloquenz, Aposiopese, Parallelismus, Pleonasmus, Anakoluth und Parenthese. 15. Vgl. Schwyzer / Debrunner, Griechische Grammatik 2, 698. 16. Für einen pragmatischen Stilbegriff vgl. M. Landfester, Einführung in die Stilistik der griechischen und lateinischen Literatursprachen (Die Altertumswissenschaft), Darmstadt 1997, 4-6; Kraus, Sprache, 27-32. 17. Vgl. etwa Meier-Brügger, Griechische Sprachwissenschaft I, 92 (E404): »Es ist zuweilen verführerisch, eine fragliche Form direkt so für ererbt zu halten und ›tel quel‹ in vergangene Zeiten zurück zu transportieren. Man darf aber nie vergessen, sich zu fragen, ob diese eventuell selbst erst Endprodukt einer innergr. Entwicklung ist.« 18. Bezeichnung nach I. Soisalon-Soininen, »Die Wiedergabe des á instrumenti« in: A. Aejmelaeus / R. Sollamo (Hg.), Studien zur Septuaginta-Syntax (AASF.B 237), Helsinki 1987, 116130 (zuvor veröffentlicht in: J. Kijlunen / V. Riekkinen / H. Räisänen [Hg.], Glaube und Gerechtigkeit: In memoriam Rafael Gyllenberg (Schriften der Finnischen Exegetischen Gesellschaft 38), Helsinki 1983, 31-46), und, unter Bezug auf diesen, F. Austermann, Von der Tora zum Nomos. Untersuchungen zur Übersetzungsweise und Interpretation im Septuaginta-Psalter (MSU 27), Göttingen 2003, 60-26; K. Hauspie, »Ἐν with Dative Indicating Instrument in the Septuagint of Ezekiel« in: M. K. H. Peters (Hg.), XII Congress of the International Organization for Septuagint and Cognate Studies, Leiden, 2004 (SCSt 54), Atlanta, GA 2006, 201-224.

454

2 Zur Demonstration: Präposition ἐν mit Dativ und ‫ְבּ־‬-instrumenti

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

sind dabei unerlässlich. Dabei ist unstrittig, dass ein Zusammenhang zwischen hebräischem ‫ְבּ‬-instrumenti und dem griechischen ἐν mit dativus instrumenti in der Septuaginta besteht, der sich auf die geschriebene Sprache des NT und des frühen Christentums (in unterschiedlichem Maße) auswirkte. 19 Der Instrumentalis mit ἐν findet sich etwa in folgenden Ausdrücken im NT: 20 »mit dem Schwert (töten, umkommen usw.«; ἐν μαχαίρῃ, 21 ἐν ῥομφαίᾳ); »würzen mit etwas« (Mt 5,13 ἐν τίνι ἁλισθήσεται), »mit Feuer verbrennen (ἐν πυρί eigentlich ›im Feuer‹)«, »taufen mit« (ἐν ὕδατι, ἐν πνεύματι), »rechtfertigen durch« (δικαιοῦν[-οῦσθαι] mit ἐν), »mischen mit« (μειγνύναι mit ἐν) und »messen mit« (μετρεῖν mit ἐν). Allerdings gibt es auch entsprechende Belege für den Gebrauch des bloßen Dativ als Instrumentalis. 22 Liegt damit deutlicher Einfluss des hebräischen ‫ְבּ‬-instrumenti vor? Das muss nicht zwingend der Fall sein. Im Laufe der Entwicklung der griechischen Sprache übernehmen die Präpositionen 23 immer mehr die Funktionen der selbständigen Kasus oder Adverbien, und dieser Prozess dauert noch in der Koine an. Der Dativ tritt bei Präpositionen immer mehr zurück. Das trifft jedoch gerade auf die beliebte Präposition ἐν nicht bzw. nicht in dem Maße zu. 24 Gleichzeitig ist dieser Wandel durch andere Faktoren bedingt, so durch die Überlappung von ἐν und εἰς an manchen Stellen sowie ihrer wechselseitigen Verwendung, durch die häufige Ersetzung des dativus modi und des lokalen Dativs, mit bestimmten periphrastischen Wendungen und vor allem mit der vermehrten Verwendung des dativus instrumenti (womit? wodurch?), der selbst nicht immer eindeutig vom dativus causae unterschieden werden kann. 25 So wird ἐν zwischenzeitlich zur am meisten verbreiteten Präposition des Griechischen, was an deren auffälliger Häufigkeit in der LXX, im NT und in den Papyri, aber ebenso an ihrer Beliebtheit z. B. bei Epiktet, Dionysius von Halicarnassus, Polybius oder Diodor von Sizilien abzulesen ist. 26 Trotzdem ist die besondere Häufigkeit von ἐν mit Dativ in instrumentaler Funktion in der LXX und im NT auffällig. Denn durch den Einfluss des ‫ְבּ‬-instrumenti 27 wurde eine Bedeutungserweiterung von ἐν mit Dativ in kausalem und/oder instrumentalem Sinn begünstigt. 28 Doch könnte an bestimmten Stellen pleonastisches ἐν als Wiedergabe von hebräischem ‫ ְבּ‬eine reine Verstärkung des Dativs sein, eine Funk19. So Bauer, s. v. ἐν. 20. Vgl. Helbing, Ks, 199-200, 233, 240 u. ö.; BDR § 195,1a-g. 21. Vgl. hierzu die kritischen Anmerkungen hinsichtlich der Verwendung von ἐν μαχαίρῃ in den Papyri bei BDR § 195,1a (4); Kraus, Sprache, 115 Anm. 310. 22. Vgl. BDR § 195,1c (6). 23. Zum Folgenden vgl. Kraus, Sprache, 86-154 und die dort angeführte Literatur. 24. So dezidiert BDR § 203; Kraus, Sprache, 112-114. 25. Vgl. BDR §§ 195-196. 26. Vgl. Turner, Syntax (Moulton, A Grammar III), 254-257, 260-265; Kraus, Sprache, 99-101, 114. 27. Vgl. K. Hauspie, »Periphrastic Tense Forms with εἰμι and γίγνομαι in the Septuagint of Ezekiel« in: E. Bons / T. J. Kraus (gg.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint (FRLANT 241), Göttingen 2011, 127-151, hier 224 (Ezekiel). 28. Allerdings merkt Helbing, Kasussyntax, 199-200, an, dass an einigen Stellen der LXX, an denen ἐν für ‫ ְבּ‬steht, dieses auch bei rein lokaler Bedeutung verständlich ist. 2 Zur Demonstration: Präposition ἐν mit Dativ und ‫ְבּ־‬-instrumenti

455

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

tion, die ἐν schon immer besaß. Dennoch lässt sich festhalten, dass nunmehr in der LXX und im NT »eine marginale Möglichkeit der klassisch-griechischen Syntax zur Regel erhoben« wurde. 29 Auf diesem Hintergrund ist dann der Gebrauch von ἐν etwa in Psalm 88,33 (ἐπισκέψομαι ἐν ῥάβδῳ; der strafende Gott kommt »in einem Stock« daher bzw. wird »mit einem Stab ihre Gesetzlosigkeiten heimsuchen«) und in 1Kor 4,21 (ἐν ῥάβδῳ ἔλθω πρὸς ὑμᾶς) zu verstehen. 30 Allerdings gibt es nach wie vor unterschiedliche Möglichkeiten, den Instrumentalis wiederzugeben. Das lässt sich z. B. anhand der Korrespondenzen zwischen ‫ְבּ‬-instrumenti und ἐν mit Dativ bzw. dem einfachen Dativ im Buch Ezechiel verdeutlichen. 31 Dort steht ἐν mit dativus instrumenti nur an den Stellen, wo der hebräische Text ‫ְבּ‬instrumenti hat. Steht kein ‫ְבּ‬-instrumenti, wird dieser Fall im Septuaginta-Text mit dem bloßen Dativ ausgedrückt. 32 Angesichts der individuellen Verschiedenheit der Texte des NT sowie deren ursprünglicher Abfassung in griechischer Sprache ist ein so klarer Nachweis dort schwieriger. Zudem ist nicht immer eindeutig zu entscheiden, ob wirklich ἐν mit folgendem Dativ instrumental bzw. kausal oder aber räumlich aufzufassen ist. 33 Vor einer automatischen Rückführung von ἐν mit dativus instrumenti (bzw. causae) im NT auf ‫ְבּ‬-instrumenti ist daher zu warnen. In der zeitgenössischen griechischen Literatur nichtjüdischen Ursprungs sowie in den dokumentarischen Papyri finden sich Belege für ἐν mit instrumentalem Dativ, die nicht im Verdacht einer Beeinflussung durch das Hebräische oder Aramäische stehen. 34

3. Weitere bemerkenswerte grammatikalische Phänomene (in Auswahl) Im Folgenden sollen ausgewählte weitere grammatikalische Phänomene vorgestellt werden, die tendenziell und aufgrund ihrer Häufigkeit als für die LXX und das NT spezifisch gelten können: Nicht nur ἐν und sein Gebrauch verblassen in der LXX und im NT, sondern auch die sonstigen Präpositionen. Generell sind die Ersetzung der bloßen Kasus und der

29. Siegert, 1,151. Ferner Conybeare / Stock, Grammar, § 91b; Turner, Syntax (Moulton, Grammar III), 252. 30. Vgl. Siegert, 1,151. Übersetzung nach LXX.D2. 31. Vgl. Hauspie, »Ἐν with Dative«, 205-218. Im Einzelnen benennt sie die Kategorien wie folgt (205) »I. ἐν with the dative construction rendering ‫ ב‬instrumenti«, »II. ἐν with the dative construction versus simple dative rendering ‫ ב‬instrumenti«, »III. the simple dative rendering ‫ב‬ instrumenti«, »IV. the simple dative not rendering ‫ ב‬instrumenti«, »V. ἐν with the dative construction rendering ‫ ב‬instrumenti versus the simple dative not rendering ‫ ב‬instrumenti«. 32. Vgl. Hauspie, »Ἐν with Dative«, 223. Ferner Austermann, Von der Tora zum Nomos, 61-62. 33. Vgl. Kraus, Sprache, 125-127. 34. So Kraus, Sprache, 114-116, mit Belegen aus der Literatur des Hellenismus (Anm. 309) und aus den Papyri (Anm. 311). Vgl. ferner 115 Anm. 312: Gerade die Wendung διαλυθῆναι (ἐν) τῷ λιμῷ u. ä. der Papyri im Verhältnis zu Apk 6,8 mit ἀποκτεῖναι ἐν ῥομφαίᾳ καὶ ἐν λιμῷ καὶ ἐν θανάτῳ ist interessant. Haben häufiges ἐν ῥομφαίᾳ und die Verwendung von ἐν θανάτῳ an dieser Stelle ἐν λιμῷ als Analogie nahe gelegt, steckt irgendwie ‫ ְבּ־‬dahinter oder spiegelt sich hier Alltagsgriechisch wider?

456

3. Weitere bemerkenswerte grammatikalische Phänomene (in Auswahl)

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

Wandel der Präpositionen auch in der griechischen Bibel zu beobachten. 35 Analog stellt die Zunahme von sogenannten unechten oder uneigentlichen Präpositionen eine weitere sprachgeschichtliche Entwicklung dar. 36 Interessant sind noch Umschreibungen bestimmter präpositionaler Begriffe, die auf hebräischen Einfluss zurückgeführt werden können (z. B. ἀπὸ προσώτου und πρὸ προσώπου; ἐκ χειρός, ἐν χειρί und διὰ χειρός [für διά]; διά/ἐκ/ἐπί + στόματος, was aber auch im klassischen Griechisch in diesen und ähnlichen Wendungen vorkommt). 37 Weitere deutliche Hinweise für einen Einfluss der LXX auf das NT gibt es nur vereinzelt in diesem Bereich. 38 Im Hebräischen 39 und danach auch in LXX und NT wird der erste Tag statt mit einer Ordinal- (πρώτη) mit einer Kardinalzahl (μία) benannt, so also etwa εἰς μίαν σαββάτων (Mt 28,1; vgl. aber Mk 16,9 πρώτῃ σαββάτου) und ἐν ἡμέρᾳ μιᾷ τοῦ μηνός (Ex 40,2; weitere Wochentage aber δευτέρα usw.). Auch sind semantische Überlappungen zwischen dem Zahlwort εἷς und dem indefinitem Pronomen τις (z. B. Mt 8,19; 26,69; Mk 11,29; Gen 21,15; 1Esdr 3,5; 2Makk 8,33) offenbar dem hebräischen ‫ אחד‬bzw. dem aramäischen ‫ חד‬geschuldet. Auf aramäischem Einfluss beruhen wohl auch εἷς τὸν ἕνα statt ἀλλήλους (1Thess 5,11) und ἓν τριάκοντα (Mk 4,8.20). 40 Die in der klassischen Literatur so charakteristische adversative Korrelation von μέν und δέ ist in der LXX nur vereinzelt vorhanden. 41 Im NT fehlt sie in 2Thess, 1Tim, Tit, 2Petr, 1–3Joh und Apk, und in in Eph, Kol, 1Thess und Jak kommt sie nur je einmal vor. 42 Der deutlich reduzierte Gebrauch von μέν kann darauf zurückzuführen sein, dass es im Hebräischen ohne Äquivalent ist. 43 Allerdings müssen stets der Einzelfall und der jeweilige Kontext genauer betrachtet werden. Die Setzung von πᾶς mit vorangehendem oder folgendem οὐ bzw. μή anstelle von οὐδείς bzw. μηδείς (»niemand/keiner«) erfolgt in LXX und NT deutlich häufiger als in außerbiblischer Literatur. Sie entspricht der Verwendung von ‫( לא ]…[ כל‬vgl. Jer 32,17; Lk 1,37) bzw. ‫( כל ]…[ לא‬vgl. Ex 12,43; 2Petr 1,20) im Hebräischen und kann durch diese motiviert sein. 44 35. Vgl. etwa die Ersetzung von ὑπό (Urheber) bei Passiv und bei Verben mit passiven Sinn durch ἀπό (mit Genitiv des Agens). So BDR § 210,2; Kraus, Sprache, 90. 36. Hierzu BDR §§ 214-216. 37. Vgl. im Einzelnen BDR § 217. Ferner Helbing, Kasussyntax, 273. 38. So kann etwa μετά mit Genitiv als »in dealing with« (Conybeare / Stock, Grammar, § 93) durchaus ein Hebraismus sein (vgl. Ri 15,3 und Lk 10,37). Neue, d. h. uneigentliche Präpositionen sind ebenso in der LXX und dann im NT zu beobachten (ἔναντι in Ex 28,12.23.24 etc. und Lk 1,8; Apg 8,21; ἀπέναντι Gen 3,24 u. ö. sowie Apg 17,7), zudem die häufige Folge Verb + Präposition, die auf hebräischen Einfluss zurückgehen kann. Vgl. Conybeare / Stock, Grammar, §§ 89-98. 39. Vgl. GesK § 134o und p. 40. Hierzu Conybeare / Stock, Grammar, §§ 48-49; BDR § 247-248; W. F. Howard, »An Appendix on Semitisms in the New Testament« in: Moulton / Howard, A Grammar II, 439. 41. So Conybeare / Stock, Grammar, § 39; Siegert, 1,151. 42. BDR § 447,2 (a); Kraus, Sprache, 167-168; T. J. Kraus, »Grammatisches Problembewusstsein als Regulativ für angemessene Sprachbeurteilung – das Beispiel der griechischen Negation und 2 Petr« FilNeot 14 (2001) 87-99, hier 88. 43. Vgl. Siegert, 1,151. 44. Vgl. Conybeare / Stock, Grammar, § 80; BDR § 302,1; Kraus, Sprache, 202-203. Über das Anakoluth nach πᾶς vgl. BDR § 466,3. 3. Weitere bemerkenswerte grammatikalische Phänomene (in Auswahl)

457

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

Die doppelte Verneinung οὐ μή und andere Verbindungen von Negationen bedürfen jeweils einer Einzelfallanalyse in ihrem jeweiligen Kontext. Nur so wird, falls überhaupt, entscheidbar, ob es sich um einen im Griechischen gängigen Gebrauch handelt oder doch um eine Beeinflussung durch die hebräische Art und Weise, Negationen auszudrücken. 45 Analoges gilt weiterhin für den Gebrauch des Artikels im Griechischen. Sicherlich dienen ὁ, ἡ, τό grundsätzlich dazu, einen Gegenstand zu substantivieren, zu individualisieren und generalisieren und zu determinieren. Doch kann das Fehlen eines zu erwartenden Artikels noch Relikt der artikellosen Zeit des Griechischen sein, und zwar abhängig vom Genre, vomAutor, vom Kontext oder von der jeweiligen sprachgeschichtlichen Phase. 46 Trotz dieser Variablen lassen sich zwei Phänomene beobachten: (a) Unter Einfluss des artikellosen, regierenden Nomens im status constructus gibt die LXX häufig Nomina und die von ihnen regierten Genitive ohne Artikel wieder – oder aber der Artikel steht dann bei beiden. Dies gilt für die LXX zu (z. B. Jes 10,3 ἐν τῇ ἡμέρᾳ τῆς ἐπισκοπῆς für ‫ )ליום פקדה‬wie auch für das NT (z. B. 1Petr 2,12 ἐν ἡμέρᾳ ἐπισκοπῆς für ‫)ליום פקדה‬. 47 (b) Entsprechend dem hebräischen artikellosen Tetragramm ‫יהוה‬steht in der LXX häufig artikelloses κύριος, besonders nach Präpositionen und im Genitiv – dagegen wird das ‫ אדני‬aber mit ὁ κύριος wiedergegeben Die Verwendung von θεός und κύριος im NT spiegelt diese Auffassung grundsätzlich wider, wenn auch mit mit Abweichungen. 48 Allerdings ist bei solchen Fragestellungen zu berücksichtigen, dass die Setzung oder Nicht-Setzung des griechischen Artikels variabel ist und dass Unterschiede in der Funktion des Artikels in den verschiedenen Sprachen bestehen. 49 Die Parataxe bestimmt den Satzbau der LXX. Verbindungen erfolgen insbesondere durch καί, was meist dem hebr. ‫ו‬-consecutivum bzw. ‫ו‬-adversativum geschuldet ist. Auch die Schriftsteller des NT bedienen sich der Konjunktion καί (vgl. bes. Mk; Apk). Häufiges καί wirkt dabei zwar eintönig, ergibt aber mitunter einen volkssprachlichen Charakter. Sein Einsatz und seine Bedeutung sind aber so vielfältig, dass immer der Einzelfall analysiert werden muss. 50 Als Hebraismus (vgl. ‫ )ויהי ]…[ ו‬können auch

45. Hierzu kürzlich J. Joosten, »Rhetorical ornamentation in the Septuagint: The case of grammatical variation« in: Bons / Kraus, Et sapienter et eloquenter, 11-22, bes. 16-20. Ferner BDR § 431; Kraus, »Grammatisches Problembewusstsein«, 94-96. 46. Vgl. A. Svensson, Der Gebrauch des bestimmten Artikels in der nachklassischen griechischen Epik, Lund 1936; Howard, »An Appendix«, 430-431; BDR §§ 249-276; Kraus, Sprache, 55-61; Siegert, 1147. 47. Vgl. BDR §§ 259,1. 48. Vgl. BDR § 254,1; Kraus, Sprache, 70-72; Siegert, 1204-1205. 49. Hierzu besonders T. J. Kraus, »Der Artikel im Griechischen: Nutzen einer syntagmatischen Beschäftigung anhand von ausgewählten Syntagmata (Hab 1,12; Jud 17; Joh 6,32)« RB 107 (2000) 260-272. 50. Vgl. Conybeare / Stock, Grammar, § 40; Abel, Grammaire, § 78; Howard, »An Appendix«, 420-423; BDR § 442; Kraus, Sprache, 159-160; Siegert 1,149-150. Ferner A. Aejmelaeus, Parataxis in the Septuagint. A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASR.DHL 31), Helsinki 1982, 12-33, 126-138, 148-150; K. Titrud, »The Overlooked Kai in the Greek New Testament« Notes on Translation 5 (1991) 1-28.

458

3. Weitere bemerkenswerte grammatikalische Phänomene (in Auswahl)

28. Die Sprache der Septuaginta und das Neue Testament: Grammatik

die satzeinleitende Ausdrücke καὶ ἐγένετο oder ἐγένετο δέ gelten(z. B. Gen 39,19; 41,8; Jos 1,1; Ri 1,1; Ruth 1,1; Mt 9,10; Mk 1,9; Lk 5,1; Apg 5,7). 51 Die pleonastische Setzung eines Personalpronomens (vor allem. αὐτός) zu einem Relativum erfolgt in der LXX häufig (z. B. Num 13,32 τῆς γῆς, ἣν κατεσκέψαντο αὐτήν) und ist auch im NT anzutreffen (z. B. Mt 3,12 = Lk 3,17 οὖ τὸ πτύον ἐν τῇ χειρὶ αὐτοῦ). Analoges gilt für ein zusätzliches Personalpronomen nach einem Partizip, das einem Relativsatz entspricht. 52 Die Häufigkeit des substantivierten Infinitivs, dann vor allem mit pleonastischem τοῦ für die Präposition ‫ְל‬, verweist auf hebräischen Einfluss (vgl. Ez 21,16; Ps 90,11 = Lk 4,10 τοῖς ἀγγέλοις αὐτοῦ ἐντελεῖται περὶ σοῦ τοῦ διαφυλάξαι σε). 53 Im Einzelfall bleibt aber zu prüfen, ob die Substantivierung nicht doch der innersprachlichen Entwicklung der Koine zuzurechnen ist. 54 Die Ausdehnung der Umschreibung mit Formen von εἶναι mit Partizip (periphrastisches Tempus) 55 in LXX und NT ist durch das hebräische Verb ‫ היה‬mit Partizip begünstigt. Die Häufigkeit ist in außerbiblischen Texten bei weitem geringer. 56 Grundsätzlich werden in diesem Kontext die Begriffe Semitismus und Hebraismus in der Literatur oft verwandt. 57 Es ist jedoch davor zu warnen, sprachliche Phänomene leichtfertig mit diesen Kategorien zu bezeichnen, gerade dann, wenn die historische Entwicklung einer Sprache selbst nicht adäquat berücksichtigt wird. Darüber hinaus können die LXX-Zitate im NT, die eine Übernahme syntaktischer Strukturen und anderer grammatikalischer Phänomenen erkennen lassen, weitere Aufschlüsse über den sprachlichen Zusammenhang zwischen der LXX und dem NT ermöglichen. Doch bedarf es in diesem Feld weiterer Bemühungen, vor allem hinsichtlich der Methodik und der Kriterien. 58 Die angeführten Tendenzen und Häufigkeiten ausgewählter sprachlicher Phänomene geben einen Eindruck von der Zusammengehörigkeit von LXX und NT im Bereich der Grammatik im Vergleich zu außerbiblischen Textzeugnissen. Genauere Aussagen sind aber stets nur dann möglich, wenn der Einzelfall gründlich analysiert wird.

51. Vgl. Conybeare / Stock, Grammar, § 40; Howard, »An Appendix«, 420-423; BDR § 442,4(a); Siegert 1,150. Interessant vor allem Mk 1,9 mit καὶ ἐγένετο und in variae lectiones ἐγένετο δέ, ἐγένετο oder nur καί. 52. Vgl. Conybeare / Stock, Grammar, § 69; BDR § 297,1-2. 53. Hierzu Siegert 1,159; BDR §§ 398-404. 54. Vgl. BDR §§ 398,1-2; Howard, »An Appendix«, 448-451, mit Rekurs auf den Einfluss von ‫ְל־‬ und hebräischen Einfluss (so Radermacher, Neutestamentliche Grammatik, 189) auf der einen und die innersprachliche Entwicklung (so Moulton, A Grammar I, 346) auf der anderen Seite. 55. Z. B. stünde statt οὗτοι μαίνονται dann οὗτοι μαινόμενοί εἰσιν. Hauspie, »Periphrastic Tense Forms«, 128. 56. Vgl. vor allem Hauspie, »Periphrastic Tense Forms«, 127-151. Ferner BDR §§ 65,4; 352-356. 57. Vgl. Howard, »An Appendix«, 411-485. 58. Vgl. U. Rüsen-Weinhold, Der Septuagintapsalter im Neuen Testament. Eine textgeschichtliche Untersuchung, Neukirchen-Vluyn 2004. Deutlichere Ergebnisse sind von dem Forschungsprojekt »Septuaginta-Zitate im Neuen Testament« (Leitung: M. Karrer) an der Kirchlichen Hochschule Wuppertal/Bethel zu erwarten. 3. Weitere bemerkenswerte grammatikalische Phänomene (in Auswahl)

459

29. Die Sprache der Septuaginta und das Neue Testament: Stil Ralph Brucker

1. Allgemeine Vorbemerkungen Die verschiedenen neutestamentlichen Autoren sind mehr oder weniger stark von sprachlich-stilistischen Eigenheiten der Septuaginta geprägt. 1 Daher wird im folgenden Überblick nach Autoren differenziert, wobei der Schwerpunkt auf den (synoptischen) Evangelien und Paulus liegt und die übrigen Schriften etwas knapper behandelt werden. Einige sprachliche Phänomene begegnen jedoch generell im Neuen Testament und sollen daher vorab kurz benannt werden. Die gehäufte Verwendung der obliquen Kasus von αὐτός im NT ist auffällig und erinnert an die entsprechende Wiedergabe der hebräischen Personalsuffixe in der Septuaginta; sie könnte sich aber auch der Anlehnung an die Umgangssprache verdanken, die diese Tendenz ebenfalls aufweist. 2 Ähnlich sieht es bei der figura etymologica aus, bei der ein dem Verb verwandtes Substantiv als inneres Objekt fungiert; diese Figur ist auch im Griechischen geläufig, kommt aber besonders regelmäßig in der Septuaginta vor. 3 Ein wohl aus der Septuaginta übernommenes Beispiel ist ἐφοβήθησαν φόβον μέγαν »sie fürchteten sich mit großer Furcht« (Mk 4,41; Lk 2,9; vgl. LXX Jona 1,10.16; 1Makk 10,8); nach diesem Muster gebildet sind auch ἐχάρησαν χαρὰν μεγάλην σφόδρα »sie freuten sich mit großer Freude sehr« (Mt 2,10), ἐξέστησαν ἐκστάσει μεγάλῃ »sie entsetzten sich mit großem Entsetzen« (Mk 5,42) und ἐθαύμασα θαῦμα μέγα »ich wunderte mich mit großer Verwunderung« (Offb 17,6). Auch die Verwendung eines genitivus qualitatis anstelle eines Adjektivs (z. B. ὁ 1.

2. 3.

Wichtige Literatur, die das Neue Testament insgesamt in den Blick nimmt: F. Blass / A. Debrunner, Grammatik des neutestamentlichen Griechisch (bearbeitet von F. Rehkopf), Göttingen 151976, 182001 (zitiert: BDR); J. H. Moulton, A Grammar of New Testament Greek. Vol. 4: Style (by N. Turner), Edinburgh 1976 (zitiert: Turner, Style); D. Tabachovitz, Die Septuaginta und das Neue Testament. Stilstudien (ActaAth-8o 4), Lund 1956; L. Rydbeck, Fachprosa, vermeintliche Volkssprache und Neues Testament. Zur Beurteilung der sprachlichen Niveauunterschiede im nachklassischen Griechisch (SGU 5), Stockholm 1967; M. Reiser, Sprache und literarische Formen des Neuen Testaments, Paderborn 2001. – Grundlegend ist außerdem: W. Bauer, Griechisch-deutsches Wörterbuch zu den Schriften des Neuen Testaments und der frühchristlichen Literatur, 6., völlig neu bearbeitete Auflage im Institut für neutestamentliche Textforschung/Münster unter besonderer Mitwirkung von V. Reichmann hg. v. K. Aland und B. Aland, Berlin/New York 1988 (zitiert: Bauer). Vgl. BDR § 278; Turner, Style, 21.72. Vgl. BDR § 153; Turner, Style, 15; R. Helbing, Die Kasussyntax der Verba bei den Septuaginta. Ein Beitrag zur Hebraismenfrage und zur Syntax der Κοινή, Göttingen 1928, 88-92.

460

1. Allgemeine Vorbemerkungen

29. Die Sprache der Septuaginta und das Neue Testament: Stil

μαμωνᾶς τῆς ἀδικίας Lk 16,9 = ὁ ἄδικος μαμωνᾶς Lk 16,11) kann als Septuagintismus angesehen werden. 4 Die Bezeichnung des ersten Tages der Woche mit μία statt πρώτη (im NT z. B. Mk 16,2 parr.; Apg 20,7; 1Kor 16,2) ist ein Hebraismus und findet sich so immer in der Septuaginta. 5

2. Die Evangelien Die stilistische Analyse der synoptischen Evangelien bestätigt die Zwei-Quellen-Theorie, wonach Matthäus und Lukas das Markusevangelium sowie die hypothetisch erschlossene Logienquelle Q als gemeinsame Quellen benutzt haben sowie darüber hinaus jeweils unabhängiges Sondergut zur Verfügung hatten. Beide haben ihre Quellen stilistisch bearbeitet.

2.1 Markusevangelium Das Markusevangelium 6 ist geprägt durch eine einfache Sprache. Im Satzbau dominiert die Parataxe, d. h. die Aneinanderreihung von Sätzen mit καί (seltener mit δέ), wobei das Verb am Satzanfang steht, gefolgt vom Subjekt. Dies erinnert stark an den Stil der erzählenden Bücher der Septuaginta und dürfte von daher geprägt sein, auch wenn dieses Phänomen sich in volkstümlicher Literatur ebenfalls findet. Sowohl Matthäus als auch Lukas reduzieren die Satzanfänge mit καί zugunsten einer größeren Variation, was sich offenbar dem Wunsch nach einem eleganteren Stil verdankt. Als weitere Charakteristika des markinischen Stils lassen sich folgende Phänomene benennen: Häufig (ca. 150mal) findet sich das praesens historicum als Mittel der Vergegenwärtigung des Erzählten. 7 Dieses Stilmittel ist zwar generell weit verbreitet, aber gerade in der Septuaginta recht häufig (ca. 330mal). 8 Das Evangelium neigt zu Pleonasmen bzw. Redundanz. Dies zeigt sich insbesondere an Stellen, an denen eine Aussage zweimal mit synonymen Wendungen gemacht wird, z. B. Mk 2,25 ὅτε χρείαν ἔσχεν καὶ ἐπείνασεν »als er [David] Mangel hatte und hungerte« (weitere Belege Mk 1,28.32.35; 4,2.39; 5,15.19.39; 6,4.25; 7,21.33; 8,17; 9,2; 12,44; 4. 5. 6.

7. 8.

Vgl. BDR § 4,2 mit Anm. 5; § 165,1 (dort Anm. 2 weitere Beispiele). Vgl. BDR § 247,1; Turner, Style, 22. Literatur: Turner, Style, 11-30; P. Dschulnigg, Sprache, Redaktion und Intention des MarkusEvangeliums. Eigentümlichkeiten der Sprache des Markus-Evangeliums und ihre Bedeutung für die Redaktionskritik (SBB 11), Stuttgart 1986 (hier ausführliche Referate der älteren Literatur); M. Reiser, Syntax und Stil des Markusevangeliums im Licht der hellenistischen Volksliteratur (WUNT II 11), Tübingen 1984; H. Cancik (Hg.), Markus-Philologie. Historische, literargeschichtliche und stilistische Untersuchungen zum zweiten Evangelium (WUNT 33), Tübingen 1984 (darin besonders H.-P. Rüger, »Die lexikalischen Aramaismen im Markusevangelium«, 73-84; M. Reiser, »Der Alexanderroman und das Markusevangelium«, 131-163; G. Lüderitz, »Rhetorik, Poetik, Kompositionstechnik im Markusevangelium«, 165-203). Vgl. BDR § 321,1 (mit Anm. 1-2). Vgl. Turner, Style, 20. 2. Die Evangelien

461

29. Die Sprache der Septuaginta und das Neue Testament: Stil

13,19.20; 14,1.18.30.61; 15,26). Dies könnte ein Reflex des biblisch-hebräischen Parallelismus mit seiner Häufung von Synonymen sein. 9 Matthäus und Lukas reduzieren auch hier deutlich (wenngleich bei Lukas das Phänomen in geringerem Maß ebenfalls anzutreffen ist; vgl. Lk 22,11; Apg 10,15; 18,21). Zur Kategorie der Redundanz gehört auch die Verwendung von ἄρχομαι als Hilfsverb, z. B. Mk 10,28 ἤρξατο λέγειν ὁ Πέτρος »Petrus begann zu sagen« (Mt und Lk: εἶπεν »sagte«). 10 Obwohl Matthäus und Lukas die 26 markinischen Belege auf 6 bzw. 2 reduzieren, kommen bei Lukas wiederum 25 neue hinzu. In der Septuaginta begegnet diese Ausdrucksweise z. B. Dtn 5,1; Ri 1,27.35; 13,25; 20,31; 1Kgt 14,35; 4Kgt 10,32; 15,37. Ein ähnliches Phänomen wie ἤρξα(ν)το + Infinitiv ist die coniugatio periphrastica, besonders mit dem Imperfekt ἦν/ἦσαν + Partizip, die sich gehäuft bei Markus findet (Mk 1,6.13.22 [39 v.l.]; 2,6.18; 3,1; 4,38; 5,5.11; 9,4; 10,22.32; 14,4.40.49.54; 15,40.43; seltener mit Präsens, Perfekt oder Futur) und bei Matthäus und Lukas nur selten übernommen wird (auch wenn Lukas sie sonst ebenfalls oft verwendet). 11 Erwähnenswert ist noch die subjektlose 3. Person Plural im Sinne von »man« (z. B. Mk 1,32; 10,13 »sie brachten« = »man brachte«; vgl. Mk 3,21; 14,1; 15,27) 12. Diese kommt schon klassisch bei Verben wie φασί, λέγουσιν o. ä. vor, erfährt aber im NT eine Ausdehnung. Da sie sowohl in der Septuaginta (als Wiedergabe hebräischer Idiomatik) als auch im zeitgenössischen Aramäisch anzutreffen ist, liegt hier strenggenommen ein Semitismus vor. Einzelne Wendungen, die sich dem biblischen Stil verdanken: Der Beginn des Evangeliums mit den für einen antiken Buchanfang ungewöhnlichen Worten ἀρχὴ τοῦ εὐαγγελίου Ἰησοῦ Χριστοῦ (Mk 1,1) hat eine Analogie in der LXX-Fassung des Buches Hosea, wo es zu Beginn heißt: ἀρχὴ λόγου κυρίου πρὸς Ωσηε »Anfang des Wortes des Herrn an Hosea« (Hos 1,2). In Verbindung mit dem darauf folgenden expliziten Prophetenzitat Mk 1,2-3 wird hier schon deutlich, wie sehr die prophetische Tradition der Septuaginta – zu der auch die Bücher Josua bis Königtümer zu rechnen sind – den Hintergrund für das ganze Buch abgibt. 13 9. Vgl. Turner, Style, 19-20. Ausführlich zu den Doppelungen: F. Neirynck, Duality in Mark. Contributions to the Study of the Markan Redaction (EThL.B 31), Leuven 1972; nach Neirynck liegt keine Redundanz vor, sondern progressive Entwicklung in zwei Schritten. Lüderitz, Rhetorik, 176-180, verwendet für das Phänomen die rhetorischen Begriffe congeries und amplificatio und betont die steigernde sowie die retardierende Funktion dieser Figur (dort auch Hinweise auf den biblischen Parallelismus in den Fußnoten). 10. Vgl. BDR § 392, Anm. 9; Bauer s. v. ἄρχω 2aβ; Turner, Style, 20; J. W. Hunkin, »›Pleonastic‹ ἄρχομαι in the New Testament«, JTS 25 (1924), 390-402. 11. Zur Verwendung der coniugatio periphrastica in der LXX s. F. C. Conybeare / St. G. Stock, Grammar of Septuagint Greek. With Selected Readings, Vocabularies, and Updated Indexes, Peabody, Mass., 1995 [Erstausgabe: Selections from the Septuagint, Boston 1905; durch Register erweiterter Nachdruck], 68-70; zum NT vgl. ebd. 70-71 sowie BDR § 353,1-3 (mit Anm. 4-8); Bauer s. v. εἰμί II.4. 12. Vgl. Turner, Style, 12; BDR § 130,2 (mit Anm. 4-5); ausführlicher Rydbeck, Fachprosa, 27-45. – Turner listet Mk 1,22.30.32.45; 2,3.18; 3,2.21.32; 5,14.35; 6,33.43.54; 7,32 8,22; 10,2.49 13,9.11; 14,12; 15,14 als Stellen auf. 13. Vgl. Lüderitz, Rhetorik, 167-168.

462

2. Die Evangelien

29. Die Sprache der Septuaginta und das Neue Testament: Stil

Nach Mk 5,7 (par. Lk 8,28) wird Jesus von einem Besessenen angeschrien: τί ἐμοὶ καὶ σοί; »Was habe ich mit dir zu schaffen?« 14 Diese Formulierung findet sich in griechischer Literatur sonst nicht 15, wohl aber in der Septuaginta (vgl. Ri 11,12; 3Kgt 17,18; 4Kgt 3,13; 2Chr 35,21; ferner τί ἐμοὶ καὶ ὑμῖν; 2Kgt 16,10; 19,23). Jesus sagt nach Mk 8,12 in pneumatischer Erregung: ἀμὴν λέγω ὑμῖν, εἰ δοθήσεται τῇ γενεᾷ ταύτῃ σημεῖον. Mit der εἰ-Formulierung – wörtlich »Wenn diesem Geschlecht ein Zeichen gegeben wird …« – ohne Nachsatz liegt ein Hebraismus vor: die Wiedergabe einer typisch hebräischen Schwurformel, hier zu verstehen im Sinne von »Niemals soll diesem Geschlecht ein Zeichen gegeben werden!« 16 So findet es sich in der Q-Parallele des Logions, die bei Matthäus und Lukas übernommen wurde: σημεῖον οὐ δοθήσεται αὐτῇ »es wird [diesem Geschlecht] kein Zeichen gegeben werden« (Mt 12,39; 16,4; Lk 11,29).

2.2 Matthäusevangelium Die Sprache des Matthäusevangeliums 17 ist ein gutes, wenn auch nicht literarisch gehobenes Koine-Griechisch. Sein Stil ist im Vergleich zu seinen Quellen Markus und Q geschliffener und differenzierter, was sich in zahlreichen stilistischen Verbesserungen niederschlägt. In den erzählenden Passagen ist er knapper als Markus und strafft seine Vorlage, wobei die Dialoge stärker hervortreten. Andererseits ist sein Stil durch wiederkehrende sprachliche Formeln gekennzeichnet (etwa bei der Einführung der Erfüllungszitate oder am Ende der Reden Jesu 18); diese Formelhaftigkeit erinnert an bestimmte alttestamentliche Texte (v. a. die Priesterschrift und die Chronikbücher). Insgesamt ist sein Stil »stark durch die LXX bestimmt« und »durchweg bibelgriechisch geprägt« 19. Hinzu kommen weitere Semitismen, die auf ein judengriechisches bzw. zweisprachiges Milieu hinweisen: Der Verfasser konnte vielleicht auch Aramäisch. Die bibelgriechische Prägung zeigt sich u. a. in der Vorliebe des Verfassers für Parallelismen; häufig wird in Jesuslogien aus Markus oder Q der Parallelismus verbessert oder sogar erst geschaffen. 20 14. Die Parallelstelle Mt 13,29 hat den Plural τί ἡμῖν καὶ σοί; (weil Matthäus den Besessenen verdoppelt hat); vgl. dazu Mk 1,24 par. Lk 4,34. Zu τί ἐμοὶ καὶ σοί; vgl. auch Joh 2,4. 15. Eine Ausnahme bildet Epiktet, Diss. 2,19,19 (vgl. 1,22,15; 1,27,13; Plural 1,1,16). 16. BDR § 454,5 (mit Anm. 6); Bauer s. v. εἰ IV; Conybeare/Stock, Grammar, 90. Der weggelassene Nachsatz (»dann strafe mich Gott« o. ä.) kann auch vorangestellt gedacht werden, so wie es in der LXX manchmal tatsächlich explizit formuliert wird: τάδε ποιήσαι μοι ὁ θεὸς καὶ τάδε προσθείη, εἰ … »Das tue mir Gott, und das füge er noch hinzu, wenn …« (2Kgt 19,14; 3Kgt 21,10; 4Kgt 6,31; vgl. 1Kgt 25,22). In Ps 88,36 und 94,11 (zitiert in Hebr 3,11; 4,3.5) ist Gott der Schwörende, hier ist der gedachte Nachsatz entsprechend zu modifizieren (»dann will ich nicht Gott sein« o. ä.) 17. Literatur: Turner, Style, 31-44; U. Luz, Das Evangelium nach Matthäus. Teilband 1: Mt 1–7 (EKK I/1), Zürich/Neukirchen-Vluyn 52002, 52-78; W. Schenk, Die Sprache des Matthäus. Die Text-Konstituenten in ihren makro- und mikrostrukturellen Relationen, Göttingen 1987 (hier jeweils weitere Literatur). 18. Erfüllungszitate: 1,22 f.; 2,15.17 f.23; 4,14-16;. 8,17; 12,17-21; 13,14 f.35; 21,4 f.; 27,9 f.; Reden: 5-7; 10; 13; 18; 23-25 (abschließende Formel: 7,28; 11,1; 13,53; 19,1; 26,1). 19. Luz, EKK I/15, 53. 20. Statistische Angaben hierzu bei Luz, EKK I/15, 55 Anm. 144. 2. Die Evangelien

463

29. Die Sprache der Septuaginta und das Neue Testament: Stil

Die unpersönliche 3. Person Plural, die sich häufig bei Markus findet, hat Matthäus unabhängig davon auch (Mt 1,23; 5,15; 7,16; 9,17). Wie oben zum Markusevangelium festgestellt, hat auch das Matthäusevangelium einen für antike Bücher ungewöhnlichen Anfang: βίβλος γενέσεως Ἰησοῦ Χριστοῦ υἱοῦ Δαυὶδ υἱοῦ Ἀβραάμ (Mt 1,1). Das Wort βίβλος scheint zunächst dafür zu sprechen, den Satz als Titel des ganzen Evangelienbuches aufzufassen. Die Kombination βίβλος γενέσεως aber weist deutlich auf Gen 5,1 (vgl. 2,4) als Hintergrund, wo es eine Genealogie bezeichnet; somit scheint sich der Anfangssatz doch nur auf die Genealogie Jesu Mt 1,1-17 zu beziehen. Die Auslegung steht hier vor einem »Dilemma« 21 – zumal das Wort »Genesis« bereits damals als Titel für das erste Buch der griechischen Bibel geläufig war. Wie auch immer man den Satz versteht: Mit Sicherheit wird hier bereits programmatisch der Bezug des Evangeliums auf die Septuaginta zum Ausdruck gebracht. Das typische Septuaginta-Wort ἰδού »siehe!« gehört zu den matthäischen Vorzugswörtern. Es begegnet häufig in der Formel καὶ ἰδού »und siehe!«, die sich etwa gleich häufig bei Lukas findet (28 : 25, dagegen nur einmal bei Mk). Charakteristisch für Matthäus ist aber die Konstruktion Genitivus absolutus + ἰδού + Hauptsatz (Mt 1,20; 2,1.13.19; 9,10.18.32; 12,46; 17,5; 26,47; 28,11; bei Lk nur einmal), die ebenfalls durch die Septuaginta vorgeprägt ist (vgl. 3Kgt 1,42; 4Kgt 6,33; 13,21; Dan 9,21 – allerdings stets mit καί vor ἰδού). Der substantivierte Infinitiv im Genitiv (τοῦ + Infinitiv) im finalen bzw. konsekutiven Sinn gehört der gehobenen Koine an und findet sich häufig in der Septuaginta. 22 Markus hat ihn nicht, Matthäus hat ihn sechsmal (Mt 2,13; 3,13; 11,1; 13,3; 21,32; 24,45), wird aber durch Lukas deutlich übertroffen (s. u.).

2.3 Lukasevangelium und Apostelgeschichte Betrachtet man den Stil des lukanischen Doppelwerkes 23, ergibt sich ein zwiefaches Bild: Auf der einen Seite strebt der Verfasser eine gehobenere Sprache an, die auch den Ansprüchen gebildeterer Leser genügt. Dies ist bereits am Prolog Lk 1,1-4 zu erkennen; 21. Luz, EKK I/15, 118; zur Diskussion und den verschiedenen Positionen ebd. 117-119. 22. Vgl. BDR § 400; Conybeare/Stock, Grammar, 58-59. 23. Literatur: Turner, Style, 45-63; J. Jeremias, Die Sprache des Lukasevangeliums. Redaktion und Tradition im Nicht-Markusstoff des dritten Evangeliums (KEK Sonderband), Göttingen 1980; E. Plümacher, Lukas als hellenistischer Schriftsteller. Studien zur Apostelgeschichte (StUNT 9), Göttingen 1972, 38-72; ders., »Apostelgeschichte«, TRE 3 (1978), 483-528: 489-491; A. Denaux, »Style and Stylist[i]cs, with Special Reference to Luke«, FN 19 (2006), 31-51 (auch zur grundsätzlichen Frage, was »style« eigentlich ist); M. Müller, »Die Lukasschriften und die Septuaginta«, in: S. Kreuzer / M. Meiser / M. Sigismund (Hg.), Die Septuaginta: Entstehung, Sprache, Geschichte (WUNT 286), Tübingen 2012, 465-479 (hier jeweils weitere Literatur). Unter den Kommentaren ist besonders auf die von Joseph Fitzmyer zu verweisen, wo in der Einleitung Sprache und Stil ausführlicher behandelt werden (The Gospel According to Luke I–IX: Introduction, Translation, and Notes [AncB 28], New York 1981, 107-127; The Acts of the Apostles [AncB 31], New York 1998, 114-119). – Als Klassiker ist zu nennen: H. J. Cadbury, The Style and Literary Method of Luke. I: The Diction of Luke and Acts (HThS 6), Cambridge, Mass.,

464

2. Die Evangelien

29. Die Sprache der Septuaginta und das Neue Testament: Stil

der Verfasser stellt sich selbst hier als Historiker hellenistischer Prägung vor. Literarische Reminiszenzen mit Elementen hellenistischer Rhetorik weist auch Apg 17,16-34 auf (Paulus in Athen). 24 Gegenüber seinen Quellen Mk und Q nimmt Lukas durchgängig kleine stilistische Verbesserungen vor. 25 Dabei scheut er sich auch nicht, sogar an Septuaginta-Zitaten noch stilistisch zu feilen. 26 Auf der anderen Seite fällt die Imitation von Sprache und Stil der Septuaginta auf, die bereits unmittelbar nach dem Prolog mit Lk 1,5 einsetzt und in der Apostelgeschichte besonders in der ersten Hälfte dominiert, später allerdings zurücktritt. Anders als bei Mk oder Mt handelt es sich bei diesen »Septuagintismen« also nicht um das eigene Idiom des Verfassers, sondern um ein absichtlich gewähltes Stilmittel, um der Darstellung biblisches Kolorit zu verleihen. Für diese »Septuaginta-Mimesis« lassen sich zahlreiche Beispiele anführen: Die Einleitung der Erzählepisoden mit dem Septuaginta-typischen καὶ ἐγένετο »und es geschah« findet sich zwar gelegentlich auch bei Markus und Matthäus (Mk 1,9; 2,23; Mt 9,10), aber weitaus häufiger bei Lukas. Die Wendung ἐγένετο δέ, die besonders häufig in den Büchern Genesis und Exodus und sonst nur vereinzelt anzutreffen ist, findet sich im NT sogar ausschließlich im lukanischen Doppelwerk (17mal Lk, 20mal Apg). Die Formel καὶ ἰδού »und siehe!« begegnet bei Lukas etwa gleich häufig wie bei Matthäus (25 : 28, s. o.). Der substantivierte Infinitiv im Dativ mit ἐν (in der Septuaginta überaus häufig, fast immer im temporalen Sinne) ist im NT ebenfalls vorwiegend bei Lukas zu finden (2mal Mk, 3mal Mt, 32mal Lk, 7mal Apg; sonst 3mal Paulus, 4mal Hebr; die Wendung ἐγένετο ἐν τῷ mit Infinitiv Mk 4,4 und sonst 23mal Lk). Entsprechend den griechischen Aktionsarten ist der Infinitiv Präsens durativ (»während«), der Infinitiv Aorist punktuell (»als«, »nachdem«) zu verstehen. Das logische Subjekt steht im Akkusa-

1919. – E. Norden, Die antike Kunstprosa vom VI. Jahrhundert v. Chr. bis in die Zeit der Renaissance. Bd. 2, Leipzig/Berlin 21909 (= Darmstadt 91983), 480-492, ist v. a. an einer Bewertung des Stils vom klassizistischen Standpunkt interessiert, wonach Lukas im tabellarischen Vergleich zu Markus und Matthäus sehr häufig »besser« abschneidet. Die LXX wird von Norden nicht erwähnt. 24. Die Analyse der Areopagrede des Paulus Apg 17,22-31 bildet den Aufhänger für E. Norden, Agnostos Theos. Untersuchungen zur Formengeschichte religiöser Rede, Leipzig/Berlin 1913 (= 71996) (1-140). Das berühmt gewordene Buch bietet in seinem zweiten Teil materialreiche »Untersuchungen zur Stilgeschichte der Gebets- und Prädikationsformeln« (143-308) und einige Anhänge (312-400) und hat die neutestamentliche Forschung nachhaltig angeregt. 25. Vgl. hierzu die Übersicht bei Turner, Style, 57-58. 26. Beispiele bei Jeremias, Sprache, 33-35. Die dort zuletzt besprochene Stelle Apg 7,43 mit dem leicht abgewandelten Zitat aus Am 5,27 ist rhetorisch interessant: Durch die Änderung von ἐπέκεινα Δαμασκοῦ zu ἐπέκεινα Βαβυλῶνος am Zitatende ergibt sich statt des in Prosa verpönten Hexameterschlusses eine rhetorisch elegantere Endung aus Kretikus (bzw. Päon) + Trochäus. Ob dies jedoch der Grund für die Änderung durch Lukas war (so die a. a. O., 335 referierte Vermutung von W. L. Knox) oder eher inhaltliche Aspekte im Vordergrund standen (Aktualisierung der Prophetie), kann hier nicht entschieden werden. 2. Die Evangelien

465

29. Die Sprache der Septuaginta und das Neue Testament: Stil

tiv, es liegt also eine Art AcI vor, auch wenn dieser im klassischen Griechisch so nicht gebräuchlich ist. 27 Der in der Septuaginta ebenfalls häufige substantivierte Infinitiv im Genitiv (τοῦ + Infinitiv) im finalen bzw. konsekutiven Sinn, der bei Matthäus 6mal vorkommt (s. o.), ist bei Lukas 24mal im Evangelium und 22mal in der Apostelgeschichte zu verzeichnen. 28 Die oben schon zu Markus erwähnte pleonastische Verwendung von ἄρχομαι als Hilfsverb ist auch bei Lukas bemerkenswert: Zwar werden die 26 markinischen Belege von ihm auf 2 reduziert, dafür kommen aber 25 neue hinzu. 29 Inchoatives ἀναστάς (ἀναστάντες), meist gefolgt von einem Verb der Bewegung oder des Sagens (»NN stand auf und …«), ist gelegentlich bei Markus (6mal) und Matthäus (2mal), aber regelmäßig bei Lukas anzutreffen (17mal Lk, 19mal Apg), sonst im NT überhaupt nicht. 30 Der Septuaginta-Sprache verdanken sich auch Floskeln wie die Zeitangabe ἐν ταῖς ἡμέραις ἐκείναις/ταύταις »in jenen/diesen Tagen« bzw. »in den Tagen des NN« (Lk 1,5; 2,1; 6,12 u. ö.) 31, die Verbindung ἐπαίρω τὴν φωνήν »die Stimme erheben« (Lk 11,27; Apg 2,14; 14,11; 22,22) sowie die Aussage φόβος ἐπέπεσεν ἐπ’ αὐτόν »Furcht fiel auf ihn« (Lk 1,12; vgl. Apg 19,17). 32 Die in der Septuaginta geläufige Präpositionalphrase διὰ χειρός »durch die Hand von« als Pleonasmus für »durch« kommt im NT nur Apg 2,23; 7,25; 11,30; 15,2 vor. Vergleichbar ist hierzu die Wendung, Gott habe »durch den Mund (διὰ στόματος)« eines Propheten gesprochen (Lk 1,70; Apg 1,16; 3,18.21). 33 Neben der Rezeption einzelner Stilelemente der Septuaginta ist gelegentlich auch die Aufnahme eines Gattungsstils zu beobachten: Die Cantica in Lk 1-2 (bes. 1,46-55 und 1,67-79) ahmen den Stil alttestamentlicher Hymnen und Danklieder nach, der sich in ähnlicher Erzählkonstellation auch in 1Kgt [1Sam] 1–2 findet. 34 Die Abschiedsrede des Paulus in Apg 20,17-38 weist deutliche Züge alttestamentlicher Abschiedsreden wie besonders 1Kgt[1Sam] 12 (vgl. auch Jos 23; 1Makk 2,49-70) auf. 35

27. Vgl. BDR § 404. 28. Vgl. BDR § 400 mit Anm. 2. Sonstiges Vorkommen im NT: 17mal bei Paulus, 5mal Hebr, 2mal Jak, 2mal 1Petr (etwas anders Offb 12,7, s. ebd. Anm. 10). 29. Vgl. BDR § 392, Anm. 9; Bauer s. v. ἄρχω 2aβ.; Turner, Style, 20. Nach Jeremias, Sprache, 105, ist das pleonastische ἄρχομαι »kennzeichnend für die vorlukanische Tradition«. 30. Vgl. Jeremias, Sprache, 55. – LXX: Gen 22,3.19; 32,23; Ex 24,13; Num 22,13.21 u. ö. 31. Vgl. Jeremias, Sprache, 15-17, wo die Wendung »in den Tagen des NN« der Tradition, aber die beiden anderen der lukanischen Redaktion zugewiesen werden. A. a. O. 16 Anm. 4-5 findet sich eine Zusammenstellung aller Formeln mit temporalem ἐν. 32. Im NT sonst nur noch Offb 11,11. Diese und ähnliche Wendungen bei Lukas führt Jeremias, Sprache, 32-33, auf. 33. Vgl. BDR § 217 mit Anm. 6-7. 34. Vgl. N. Lohfink, »Psalmen im Neuen Testament. Die Lieder in der Kindheitsgeschichte bei Lukas« in: K. Seybold / E. Zenger (Hg.), Neue Wege der Psalmenforschung (HBS 1), Freiburg 1994, 105-125 (dort weitere Literatur). 35. Hierzu Plümacher, Lukas, 49-50 (dort weitere Literatur).

466

2. Die Evangelien

29. Die Sprache der Septuaginta und das Neue Testament: Stil

2.4 Johannesevangelium Die Sprache des Johannesevangeliums 36 weist zwar zahlreiche Hebraismen und Aramaismen auf, aber inwieweit eine Beeinflussung speziell durch die Septuaginta vorliegt, ist umstritten, zumal die expliziten Schriftzitate öfter von deren Wortlaut abweichen. Immerhin lassen sich ein paar Wendungen benennen, die auch in der Septuaginta bezeugt sind – wenngleich die Bekanntschaft mit diesen Wendungen sich auch einem allgemeinen jüdisch-griechischen Sprachgebrauch verdanken könnte: Joh 1,6; 3,1 ὄνομα αὐτῷ »sein Name (war)« (im NT noch Offb 6,8; 9,11); vgl. LXX Ri 13,2; 17,1; 1Kgt 1,1; 9,1 u. ö. Joh 2,4 die schon oben zu Markus besprochene Wendung τί ἐμοὶ καὶ σοί; »Was habe ich mit dir zu schaffen?« 37 Joh 3,21 ὁ ποιῶν τὴν ἀλήθειαν »wer die Wahrheit tut« (vgl. 1Joh 1,6). In der Septuaginta kommt ποιέω τὴν ἀλήθειαν mehrmals vor (Gen 32,11; 47,29; Jes 26,10; Tob 4,6; 13,6), sonst aber auch Test. Rub. 6,9; Test. Benj. 10,3 sowie auf Hebräisch in Qumran 1QS I 5; V 3; VIII 2. Joh 3,23 ὕδατα πολλά »viele Wasser« (im NT noch Offb 1,15; 14,2; 17,1; 19,6); die Septuaginta hat diese Verbindung (im Plural) mehrmals in den Psalmen (Ps 17,17; 28,3; 31,6; 76,20; 92,4; 106,23; 143,7; sonst Nah 1,12; Jer 28,13.55). Joh 7,18 ἀδικία ἐν αὐτῷ οὐκ ἔστιν »Ungerechtigkeit ist nicht in ihm«; vgl. LXX Ps 91,16. Hier liegt aber wohl eher eine Anspielung auf eine konkrete Schriftstelle als eine stilistische Übernahme vor.

3. Die Paulusbriefe Der Stil des Paulus 38 ist im Ganzen durch die Briefform bestimmt. Dabei sind »echte«, d. h. situationsbezogene Briefe nicht zwangsläufig »unliterarisch«, sondern können durchaus literarische Qualitäten aufweisen. 39 Verschiedene stilistische Analogien und Einflüsse sind für Paulus herausgestellt worden: 36. Literatur: Turner, Style, 64-79; H. Windisch, »Der Johanneische Erzählungsstil«, in: H. Schmidt (Hg.), ΕΥΧΑΡΙΣΤΗΡΙΟΝ (FS H. Gunkel), Bd. II (FRLANT 36/2), Göttingen 1923, 174-213; R. Schnackenburg, Das Johannesevangelium. I. Teil (HThK IV/1), Freiburg 1965, 88-101; E. Ruckstuhl / P. Dschulnigg, Stilkritik und Verfasserfrage im Johannesevangelium. Die johanneischen Sprachmerkmale auf dem Hintergrund des Neuen Testaments und des zeitgenössischen hellenistischen Schrifttums (NTOA 17), Freiburg, Schweiz, 1991. 37. Siehe oben bei Anm. 14; dort auch die Belege. 38. Literatur: Turner, Style, 80-100. Weitere Titel werden in den nächsten Anmerkungen genannt. 39. Die falsche Alternative »unliterarische Briefe oder literarische Episteln« hat A. Deissmann, Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt, Tübingen 41923, 198, unter dem Eindruck der kurz vorher entdeckten Papyrusbriefe aufgestellt (das Buch bleibt dennoch ein grundlegendes Werk). Eine differenzierte Einführung in die Thematik gibt H.-J. Klauck, Die antike Briefliteratur und das Neue Testament, Paderborn 1998 (überarbeitete und erweiterte englische Ausgabe: Ancient Letters and the New Testament. A Guide to Context and Exegesis, Waco 2006). 3. Die Paulusbriefe

467

29. Die Sprache der Septuaginta und das Neue Testament: Stil

Die Lebendigkeit mündlicher Rede mit überwiegend kurzen Sätzen, direkter Anrede der Adressaten, rhetorischen Fragen, ironischen Aufforderungen und gedachten Einwänden (»man/jemand könnte sagen« o. ä.) steht der kynisch-stoischen Diatribe nah, deren Hauptvertreter die Philosophen Teles, Musonius, Epiktet und Seneca sind. 40 Unübersehbar sind die Elemente antiker Rhetorik in den Paulusbriefen; das gilt nicht nur für die Verwendung rhetorischer Stilmittel (»Figuren«) 41, sondern auch schon für den Aufbau der Briefe. 42 Daneben hat vielleicht auch der Stil der jüdisch-hellenistischen Synagogenpredigt seine Wirkung hinterlassen. 43 Insgesamt jedenfalls wirkt der Stil des Paulus, zumal im Vergleich zu den Evangelien, eher hellenistisch. 44 Gleichwohl gibt es doch auch bei Paulus, dem einstigen pharisäischen Juden, der die »Schrift« in Gestalt der Septuaginta für seine Argumentation heranzieht, stilistische Einflüsse der Septuaginta. Beispiele für Septuagintismen sind 45: – die subjektlose 3. Person Plural im Sinne von »man« 46 (1Kor 10,20 ἃ θύουσιν »was sie opfern« = »was man opfert«) 47, 40. Hierzu grundlegend R. Bultmann, Der Stil der paulinischen Predigt und die kynisch-stoische Diatribe (FRLANT 13), Göttingen 1910 (Nachdruck 1984 mit einem Geleitwort von Hans Hübner). Dieser Ansatz ist fortgeführt worden von S. K. Stowers, The Diatribe and Paul’s Letter to the Romans (SBL.DS 57), Chico 1981; T. Schmeller, Paulus und die »Diatribe«. Eine vergleichende Stilinterpretation (NTA 19), Münster 1987. 41. Vgl. BDR § 485-496. 42. Vgl. schon J. Weiss, »Beiträge zur Paulinischen Rhetorik«, in: Theologische Studien (FS Bernhard Weiss), Göttingen 1897, 165-247, der seinerseits bereits auf C. G. Wilke (Die neutestamentliche Rhetorik, ein Seitenstück zur Grammatik des neutestamentlichen Sprachidioms, Dresden 1843) und auf die Kommentare von C. F. G. Heinrici (Das erste Sendschreiben des Apostels Paulus an die Korinthier, Berlin 1880; Das zweite Sendschreiben des Apostels Paulus an die Korinthier, Berlin 1887) verweisen kann. Im 20. Jh. ist die rhetorische Wahrnehmung der Paulusbriefe zeitweilig zurückgetreten, bis sie durch den Kommentar von H. D. Betz, Galatians: A Commentary on Paul’s Letter to the Churches in Galatia (Hermeneia), Philadelphia 1979 (deutsche Ausgabe: Der Galaterbrief. Ein Kommentar zum Brief des Apostels Paulus an die Gemeinden in Galatien, München 1988) neuen Auftrieb mit einer unüberschaubaren Fülle von neuen Studien erhielt. (Vgl. auch den in Anm. 55 genannten Titel.) 43. Vgl. H. Thyen, Der Stil der jüdisch-hellenistischen Homilie (FRLANT 65), Göttingen 1955. 44. Norden, Kunstprosa, 498-510, legt auch für Paulus den klassizistischen Maßstab an (vgl. oben Anm. 23 zu Lukas), und so ist für ihn »sein Stil, als Ganzes betrachtet, unhellenisch« (499). Dennoch findet auch er in den Paulusbriefen Elemente »aus der zünftigen griechischen Kunstprosa« (502) und Mittel »zierlicher griechischer Rhetorik« (506), als deren Quelle er aber die zeitgenössische ›asianische‹ Sophistik (und eben nicht Demosthenes) ausmacht. 45. Vgl. Turner, Style, 89-93, allerdings auch mit unsicheren Fällen und Belegen aus Kol und Eph, die hier übergangen werden. 46. Vgl. oben bei Anm. 12 zu Markus. 47. Zu vergleichen sind LXX-Stellen wie Dtn 32,17; Ps 105,37 u. a. In 1Kor 10,20 haben viele Handschriften (P46, ‫א‬, A, C, P, 33 u. a.) das explizite Subjekt τὰ ἔθνη, was im byzantinischen Text noch die Korrektur der Verbform nach sich gezogen hat (Singular, weil Neutrum Plural: ἃ θύει τὰ ἔθνη »was die Heiden opfern«). Das ist offenbar pedantische Verbesserung; den ursprünglichen Text haben B, D, F, G sowie der Ambrosiaster bewahrt.

468

3. Die Paulusbriefe

29. Die Sprache der Septuaginta und das Neue Testament: Stil

– – –

–

der substantivierte Infinitiv im Genitiv (τοῦ + Infinitiv) im finalen bzw. konsekutiven Sinn (z. B. Röm 1,24; 1Kor 10,13) 48, physiognomische Umschreibungen wie »mit dem Munde bekennen« (Röm 10,9 f.; 15,6) oder »durch die Hand« statt nur »durch« (Gal 3,19 ἐν χειρὶ μεσίτου) 49, der Gebrauch von uneigentlichen Präpositionen oder präpositionalen Wendungen wie κατέναντι (z. B. Röm 4,17), ἐνώπιον (z. B. Röm 3,20), ὀπίσω (Phil 3,13), ἔμπροσθεν (z. B. 1Thess 2,19; 3,13) oder κατὰ πρόσωπον (2Kor 10,1.7; Gal 2,11) 50, der ausgedehnte Gebrauch von eigentlichen Präpositionen wie dem instrumentalen oder kausalen ἐν (z. B. Röm 1,21.24) und dem kausalen ἐκ (z. B. Röm 1,4) 51.

Ein bei Paulus häufig anzutreffendes Phänomen ist der Parallelismus, besonders in Form von Antithesen. So kann z. B. Röm 4,25 als antithetischer Parallelismus (vgl. Röm 2,7-8; 1Kor 1,18; 4,10) und 1Kor 15,53-54 als synonymer Parallelismus (vgl. Röm 9,2; 11,33; 1Kor 15,42-43) beschrieben werden, wie er aus der Septuaginta gut bekannt ist. 52 Hier dürfte aber auch die griechische Rhetorik eine Rolle spielen, die das Phänomen ebenfalls kennt (unter den Bezeichnungen Antitheton, Isokolon und Paromoiosis) und besonders in der Kaiserzeit sehr schätzt. Die vergleichende Gegenüberstellung (Synkrisisreihe) in 2Kor 11,22-23 etwa hat Ähnlichkeit mit einer berühmten Stelle aus Demosthenes (Cor. 265); der Parallelismus in 1Kor 7,27 erinnert an den Kyniker Teles (fr. 2, p. 10, 6-7 Hense). Insbesondere die bei Paulus sehr häufige Form der Antithese mit οὐκ … ἀλλά … »nicht …, sondern … (Correctio) verdankt sich eher der hellenistischen Rhetorik. 53 Passagen in gehobenem Stil, besonders Röm 8,31-39; 11,33-36, ahmen wohl den Stil der Septuaginta-Psalmen nach (kunstvolle, rhythmische Prosa), von dem sie deutlich inspiriert sind. 54 Der Stilwechsel als solcher, also Passagen mit größerer Feierlichkeit in einem sonst stilistisch schlichteren Kontext, ist ein in hellenistischen Texten weit verbreitetes Phänomen und kein Grund, diese Passagen insgesamt als Zitate anzusehen. 55 Die deuteropaulinischen Briefe sind insgesamt noch weniger durch die Septuaginta geprägt als die echten Paulusbriefe. Dies gilt insbesondere für die sachlich-nüchternen Pas-

48. Weitere Stellen nennt Turner, Style, 90; vgl. BDR § 400 und oben bei Anm. 22 und 28 zu Matthäus und Lukas. 49. Vgl. oben bei Anm. 33 zu ähnlichen Wendungen bei Lukas. 50. Weitere Stellen nennt Turner, Style, 92; vgl. BDR § 214-217 und jeweils auch Bauer s. v. 51. Weitere Stellen bei Turner, Style, 93; vgl. BDR § 219 sowie Bauer s. v. ἐν III. und BDR § 212,2 sowie Bauer s. v. ἐκ 3 f. 52. So Turner, Style, 96-97. 53. Vgl. BDR § 489-490; Weiss, Beiträge, 168-195; N. Schneider, Die rhetorische Eigenart der paulinischen Antithese (HUTh 11), Tübingen 1970, bes. 47-52. 54. Röm 8,31-39 enthält z. B. Anspielungen auf Ps 109,1 und vielleicht Ps 117,6, und es wird sogar ein Psalmvers zitiert (Ps 43,23). 55. Und schon gar nicht als »Poesie« im Gegensatz zu »Prosa«. Dies gilt in den Paulusbriefen auch für die eher hellenistisch geprägten Passagen 1Kor 13 und besonders Phil 2,6-11. Vgl. hierzu ausführlich R. Brucker, ›Christushymnen‹ oder ›epideiktische Passagen‹ ? Studien zum Stilwechsel im Neuen Testament und seiner Umwelt (FRLANT 176), Göttingen 1997 (dort 174-210 auch ausführlich zur antiken Stiltheorie). 3. Die Paulusbriefe

469

29. Die Sprache der Septuaginta und das Neue Testament: Stil

toralbriefe 56, aber auch für den Kolosserbrief 57 und den von diesem abhängigen Epheserbrief 58, die mit ihrer Neigung zur Plerophorie und beinah liturgischen Feierlichkeit einerseits Semitismen, andererseits Einflüsse der »asianischen Rhetorik« aufweisen. 59

4. Die übrigen Schriften des Neuen Testaments Aus Raumgründen kann auf die übrigen Schriften des NT nur noch knapp und mit wenigen Literaturhinweisen eingegangen werden; lediglich die Johannesoffenbarung erfordert eine etwas eingehendere Darstellung der kontroversen Forschungspositionen. Nach allgemeiner Einschätzung gilt der anonyme Verfasser des Hebräerbriefs (die Schrift ist nach Hebr 13,22 als Trost- und Mahnrede zu charakterisieren) als der sorgfältigste Stilist des Neuen Testaments neben Lukas. 60 Bereits der Anfang besteht in einer kunstvollen Satzperiode (Hebr 1,1-4), und dieser Stil ist weithin durchgehalten. 61 Dennoch lassen sich in dieser Schrift, die durchgängig die »Schrift« in Gestalt der Septuaginta zitiert, auch etliche Septuagintismen ausmachen. 62 Der Verfasser des Jakobusbriefs schreibt ähnlich wie Lukas ein »gepflegtes Griechisch«, das semitische Einflüsse durch Vermittlung der Septuaginta aufweist. 63 Im 1. Petrusbrief begegnet ein eleganter Stil, der die Sprache der Septuaginta aufgreift und kreativ damit umgeht. Phraseologie und Vokabular basieren auf der Septuaginta, besonders den Makkabäerbüchern. 64 Die drei Johannesbriefe sind in Stil und Vokabular eng mit dem Johannesevangelium verwandt und daher weniger durch Septuagintismen geprägt. 65 56. Vgl. Turner, Style, 101-105. 57. Vgl. grundlegend W. Bujard, Stilanalytische Untersuchungen zum Kolosserbrief als Beitrag zur Methodik von Sprachvergleichen (StUNT 11), Göttingen 1973. 58. Zu Sprache und Stil des Epheserbriefs vgl. A. T. Lincoln, Ephesians (WBC 42), Dallas 1990, xxxv-xlvii; G. Sellin, Der Brief an die Epheser (KEK VIII9), Göttingen 2008, 62-64. 59. Zum Stil des 2. Thessalonicherbriefs, der sowohl sachlich-offiziell als auch feierlich-plerophorisch ist und dabei deutlich von dem der echten Paulusbriefe abweicht, vgl. W. Trilling, Untersuchungen zum zweiten Thessalonicherbrief (EThSt 27), Leipzig 1972, 46-66. 60. Oft wird hierzu auf BDR § 3 verwiesen (vgl. auch § 485 im Zitat von Blass aus der 2. Aufl.: »Kunstprosa«). Eine ausführliche Darstellung von Sprache und Stil des Hebräerbriefs gibt C. Spicq, L’épître aux Hébreux (EtB), Bd. 1, Paris 1952, 152 ff. 351-378. 61. Vgl. BDR 464 mit Anm. 2. 62. Vgl. Turner, Style, 106-113, bes. 109-112. 63. Vgl. Turner, Style, 114-120; M. Dibelius, Der Brief des Jakobus, 6. Aufl. dieser Auslegung, mit Ergänzungen von H. Greeven und einem Nachtrag von F. Hahn (KEK XV12), Göttingen 1984 [zuerst 1921], 53-57; entsprechende Abschnitte finden sich in fast allen neueren Kommentaren. Die ausführlichste Darstellung von Grammatik und Stil des Jakobusbriefs bietet immer noch J. B. Mayor, The Epistle of St. James. The Greek Text, With Introduction Notes and Comments, 31910, ccvi-cclix. 64. Vgl. Turner, Style, 121-131; L. Goppelt, Der erste Petrusbrief (KEK XII/18), Göttingen 1978, 4547. 65. Vgl. Turner, Style, 132-138.

470

4. Die übrigen Schriften des Neuen Testaments

29. Die Sprache der Septuaginta und das Neue Testament: Stil

Der Judasbrief und der 2. Petrusbrief müssen wegen ihrer direkten literarischen Abhängigkeit zusammen betrachtet werden. Trotz Vertrautheit mit der Septuaginta sowie mit weisheitlicher und apokalyptischer Literatur des Judentums ist der Stil beider Briefe insgesamt relativ »unbiblisch«. 66 Die Sprache der Johannesoffenbarung 67 weist zahlreiche schwere Verstöße gegen die griechische Grammatik (Solözismen) 68 auf, vor allem im Bereich der Kongruenz und in der Verwendung der Partizipien. Diese lassen sich zu einem erheblichen Teil als Semitismen nachvollziehen; manche bleiben aber unerklärlich oder schlicht »falsch«. Wie dieser Befund zu deuten ist, ist in der Forschung umstritten: Handelt es sich um einen Autor, der zwar griechisch schreibt, aber eigentlich hebräisch (oder aramäisch) denkt? 69 Reflektiert dies eventuell eine in seinem Umfeld gesprochene jüdische Umgangssprache? 70 Wäre er gar eigentlich durchaus fähig, korrektes Griechisch zu schreiben, bedient sich aber absichtlich einer von den alttestamentlichen Propheten inspirierten literarischen Kunstsprache, um damit einen »heiligen Stil« zu schaffen? 71 Handelt es sich bei der Schrift um ein ursprünglich hebräisch oder aramäisch geschriebenes Dokument, das in einer Weise ins Griechische übersetzt wurde, die stets auf das Original verweisen soll? 72 Oder haben wir es lediglich mit einem literarisch ungebildeten Verfasser zu tun, dessen sprachliche Fehler allesamt Parallelen in den Papyrusbriefen aus Ägypten finden? 73 Die Beurteilung

66. Vgl. Turner, Style, 139-144; R. Bauckham, Jude, 2 Peter (WBC 50), Waco 1983, 6-7 (Jud) und 135-138 (2Petr); zum 2. Petrusbrief auch T. J. Kraus, Sprache, Stil und historischer Ort des zweiten Petrusbriefes (WUNT II 136), Tübingen 2001. 67. Literatur: Turner, Style, 145-158; S. E. Porter, »The Language of the Apocalypse in Recent Discussion«, NTS 35 (1989), 582-603; D. D. Schmidt, »Semitisms and Septuagintalisms in the Book of Revelation«, NTS 37 (1991), 592-603; J. Frey, »Erwägungen zum Verhältnis der Johannesapokalypse zu den übrigen Schriften des Corpus Johanneum, 4. Beobachtungen zu Syntax und Stil«, in: M. Hengel (Hg.), Die johanneische Frage. Ein Lösungsversuch (WUNT 67), Tübingen 1993, 359-382; T. J. Bauer, Das tausendjährige Messiasreich der Johannesoffenbarung. Eine literarkritische Studie zu Offb 19,11–21,8 (BZNW 148), Berlin/New York 2007, 71-102. 68. BDR § 136 ist allein den Solözismen in der Offb gewidmet und führt doch nur ausgewählte Beispiele an. Eine ziemlich vollständige Zusammenstellung der sprachlichen »Eigentümlichkeiten der Apokalypse« gibt bereits W. Bousset, Die Offenbarung Johannis (KEK XVI6), Göttingen 1906, 159-161 (zur Sprache der Apokalypse insgesamt s. ebd. 159-179); von den neueren Kommentaren vgl. bes. D. E. Aune, Revelation 1–5 (WBC 52), Dallas 1997, cxcix-cciii (insgesamt zur Syntax ebd. clx-ccvii; dort weitere Literatur). 69. So zuerst R. H. Charles, A Critical and Exegetical Commentary on the Revelation of St. John. Vol. 1 (ICC), Edinburgh 1920, cxliii (vgl. insgesamt zum »Hebraic Style« ebd. cxlii-clii). Diese Sicht ist von vielen übernommen worden. 70. Vgl. dazu kritisch G. H. R. Horsley, »The Fiction of ›Jewish Greek‹«, NDIEC 5 (1989), 5-40; Porter, Language, 592-603 (dort werden jeweils Vertreter dieser Ansicht genannt). 71. So etwa E. Lohmeyer, Die Offenbarung des Johannes (HNT 16), Tübingen 21953 (zuerst 1926), 197-199; E. Lohse, »Die alttestamentliche Sprache des Sehers Johannes. Textkritische Bemerkungen zur Apokalypse«, ZNW 52 (1961), 122-126. 72. So R. B. Y. Scott, The Original Language of the Apocalypse, Toronto 1928 (hebräisches Original); C. C. Torrey, The Apocalypse of John, New Haven 1958 (aramäisches Original). 73. So dezidiert J. H. Moulton, A Grammar of New Testament Greek. Vol. 1: Prolegomena, Edinburgh 21906, 8-9. 4. Die übrigen Schriften des Neuen Testaments

471

29. Die Sprache der Septuaginta und das Neue Testament: Stil

des Septuaginta-Einflusses auf diese Schrift hängt stark davon ab, wie die generelle Einschätzung ausfällt. Abgesehen davon gibt es, wie oben bereits zu Lukas (Cantica) und Paulus (Röm 8,31-39; 11,33-36) festgestellt, auch in der Johannesoffenbarung einige Passagen, die den Stil der Septuaginta-Psalmen (und zwar der Hymnen) nachahmen (Offb 4,11; 5,9 f.; 11,17 f.; 15,3 f.; 19,1 f.5.6-8); auch bei diesen handelt es sich nicht um liturgische Zitate, sondern um literarische Bildungen des Verfassers. 74

74. Vgl. G. Delling, »Zum gottesdienstlichen Stil der Johannesapokalypse«, NT 3 (1959), 107-137; R. Deichgräber, Gotteshymnus und Christushymnus in der frühen Christenheit. Untersuchungen zur Form, Sprache und Stil der frühchristlichen Hymnen (StUNT 5), Göttingen 1967, 44-59; K.-P. Jörns, Das hymnische Evangelium. Untersuchungen zu Aufbau, Funktion und Herkunft der hymnischen Stücke in der Johannesoffenbarung (StNT 5), Gütersloh 1971. – Die hier erzielten Ergebnisse sind heute Konsens der Forschung.

472

4. Die übrigen Schriften des Neuen Testaments

Indexes 1. Scripture references Genesis 1:1 53, 295, 329 1:2 125 1:4 54 1:6ff. 323 1:9 341 1:11-12 340 1:11 379 1:14 312 1:15 323 1:16 295 1:21-25 340 1:21 376 1:22 253 1:28 311 2:4 380, 464 2:6 91, 105 2:7 63 2:8-15 307 2:8 276, 323 2:10 315, 376 2:16 353, 360 2:17 63 2:21 125 3:3 57 3:5 132 3:6 64 3:7 96, 311 3:10 132 3:11 134 3:15-16 133 3:16 64, 360 3:20 129, 285, 379 3:24 251, 457 4:1 130 4:2 63, 191 4:4 320 4:12 377 4:14 377 4:21 324 4:23 324 4:24 136 5:1 93, 464 5:4 95 5:22 328 1. Scripture references

5:29 197 6:1 130 6:2 95 6:4 132 6:14 50, 95 6:17 323 6:19 363 6:20 340, 376 6:22 353 7:14 340 7:19 371 8:9 95 8:19 340 8:20 312, 331 9:2 95 9:3 324 9:15 95 9:18 131 9:22 51 9:26 402 10:2 131 10:4 131 10:5 340, 342 10:9 137 10:18 342 10:19 130 10:20 340 10:31-32 340 10:31 342 11:6 340 11:10 286 11:28 285 11:29 130 12:2 340 12:3 311, 342 12:6 311 12:8 444 12:9 299 12:10 308 12:12 133 12:15 348 13:10 130, 324 13:14 136, 444 14 64, 344 14:1 340

14:5 340 14:7 353, 404 14:9 340 14:5ff. 92 14:13 289 14:15 92 14:16 341 14:20 402 14:23 324 15:1 403 15:5 135 15:6 135, 402 15:13 93, 308, 348, 360 15:14 340 15:17 377 16:11 131 17:4-6 340 17:6 344, 371 17:7 116 17:8 308 17:11 292, 402 17:14 312, 340, 402 17:16 137, 340, 344 17:17 137 17:20 340 17:27 340 17:29 340 18 64 18:8 323 18:11 309 18:18 340 18:19 353 18:28-31 97 18:29 191 19:2 314 19:4 309, 341 19:6 311 19:8 323 19:9 67, 133, 308 19:10 311 19:21 63, 287 19:25 404 19:29 64, 395 19:31 308, 309 19:33 309

473

Indexes 19:34 309 19:37 309 19:38 340 20:1 308 20:4 340 20:11 96 20:13 96 21:7 95 21:9-10 62 21:13 340 21:15 457 21:18 340 21:19 315 21:23 134, 308 21:32 360, 363 21:34 308 22:2 312 22:3 466 22:6 327 22:10 327 22:12 449 22:13 284 22:18 340 22:19 466 22:20 62 23:4 308, 313, 348 23:6 344 23:7 341 23:12 341 23:13 63, 341 24:1-2 309 24:3 97 24:5-6 307 24:18 94 24:20 312 24:22 94, 309, 323, 324, 325 24:25 324 24:39 96 24:42 96 24:44 346 24:57 63, 287 25 64 25:1 63 25:8 341, 401 25:16 324, 340 25:22 134 25:23 340, 341 25:25 324 25:29 93 25:30 93, 131

474

25:34 93, 324 26:3 308 26:4 340 26:5 351, 352, 353 26:9 52 26:10 340 26:12 314 26:24 133 26:26 324 26:28 96 26:29 133 27:1 132, 309 27:4 323 27:15 309 27:20 134 27:21 135 27:29 340 27:30 395 27:42 309 28 64 28:3 340, 341 28:14 342 28:15 133 29:9 62 29:10 315 29:16 320 29:17 401 29:26 309 29:33 130 30:13 401 30:14ff. 324 30:37 312 31:10 325 31:12 95 31:13 244 31:15 93, 312 31:19 298 31:26 312 31:27 314 31:36 338 31:39 312 31:40 324 31:52 314 32:4-5 308 32:8 341 32:11 467 32:21 282 32:22 311 32:23 466 32:29 327

32:30 134 32:31 379 32:33 311, 343 33:13 245, 325 33:15 341 33:19 119, 323 34 404 34:1 343 34:7 311, 405 34:16 340 34:18 398 34:22 341 34:28 95 34:29 310 34:30 326 35:6 341 35:8 324 35:11 325, 340, 341, 344 35:16 283 35:27 308 35:29 309, 340 36 343 36:6 310 36:11 131 36:13-31 383 36:31 344 36:40 340, 342 37:1 308 37:2 324 37:3 242 37:17 117 37:36 311, 349 37:39-50 114 38:8 314 38:24 311 39 64 39:1 233, 236, 349 39:6 401 39:10 93 39:13 395 39:15 401 39:16 97 39:18 401 39:19 459 39:20 234, 344 39:21 311, 349 40 64 41:1ff. 325 40:1 311, 349 40:2 349 1. Scripture references

Indexes 40:3 117 40:4 311, 349 40:7 349 40:8 310 40:13 311 40:16 310 40:17 340 40:22 310 41:2ff. 325 41:3 242 41:6 312 41:8 459 41:12 310 41:13 310 41:15 310 41:17 92 41:34 312 41:37 398 41:40 341 41:41 238 41:45 236, 238 41:51 238 41:55 341 41:56 312 42:6 341, 349 42:13 95 42:19 312 42:25 324 42:27 314 42:31 96 42:35 311 42:38 310 43:4 115 43:8 312 43:9 308 43:12 338 43:14 129 43:16 135 43:19 312 43:20 52 43:27 309 43:29 95 43:31 96 43:34 398 44:2 314 44:12 309 44:16 115 44:20 95 44:22 96 44:23 95 1. Scripture references

44:28 96 44:32 308 45:7 92 42:10 349 45:17 96 46:2 96 46:3 340 46:32 96 46:34 349 47:4 308 47:9 308 47:12 312 47:21 341 47:22 312 47:26 233 47:27 353 47:29 467 48:1 309 48:4 340, 341 48:7 283 48:9 133 48:10 312 48:19 340, 341 49 64 49:1 58 49:6 326, 341 49:10 340, 344 49:13 130 49:14 326 49:16 341, 342 49:19 376 49:29 341 49:33 341, 401 50:2 243, 314 50:7 308, 309 50:15 312 50:17 339, 346 50:20 341 50:24 312 Exodus 1:4 343 1:7 371 1:11 238, 326 1:16 343 1:22 325 2 64 2:1-10 236 2:3 242, 288, 314 2:5 311

2:10 235 2:14 115 2:22 308, 348 3:7 312, 360 3:14 328 3:16 312, 345 3:22 312 4:20 116 4:21 448 4:24 312 5:14 345 5:15 345 5:19 345 6:1 289 6:4 308 6:6 289 6:7 116 6:12 125 6:14 344, 379 6:15 379 6:19 379 6:24-25 379 7:3 448 7:11 124, 313, 322, 324 7:22 448 8–10 64 8:15 448 8:17 323 8:21 323 9:6 311 9:8 324 9:12 448 9:16 446, 448, 449 9:31 307 9:32 324 9:35 448 10:1 448 10:9 309 10:20 448 10:25 332 10:27 448 11–13 152 11:5 346 11:7 405 11:10 448 12:6 341 12:13 310 12:19 249, 284, 314, 322, 348 12:37-38 62, 312 12:39 324

475

Indexes 12:43-49 348 12:43 350, 404, 457 12:45 308 12:49 350 13:2 313 13:3 117 13:9 117 13:10 286 13:15 448 14:2 180 14:4 448 14:8 335, 448 14:17 448 14:31 346 15 379 15:1-21 58 15:2 403 15:3 405 15:4-5 380 15:10 324 15:15 343 15:25 353 16:3 56, 321 16:4 350 16:13 323 16:21 371 16:28 350 16:31 196 17:5 308, 309, 344 17:9 314 17:14 117 18:3 308, 348 18:8 325 18:12 309 18:21-26 345 19:5 252 19:7 308, 309 19:29 64 20:5 57 20:6 353 20:10 308 20:16 328 20:23 309 20:24 190 21:7 346 21:15-17 52 21:18 321, 343 21:25 324 21:26 346 21:27 346

476

22:1 327 22:4 312 22:24 311 22:26 324 22:30 57 23:1 116 23:6 352 23:11 312 23:13 57 23:30 371 24 64 24:1 308, 309, 345 24:6 321 24:10 64 24:13 344, 466 24:14 308, 309 25:8 245 25:16 353 25:17 332 25:18 322 25:21 353 26:7-13 321 26:24 321 27:3 321 27:8 313 27:9-13 241 28:4 321 28:12 457 28:17 321, 322 28:23-24 457 28:29 321 28:30 311 28:33 321 28:36 321 28:38 338 28:41 287 28:43 311 29:9 288 29:17 321 29:27 312 29:41 313 29:45 116 30:18 321, 377 30:23 321 30:25 321 31:18 323 32 64 32:2 323 32:14 116 32:26 51

32:29 288 33:3 325, 449 33:5 449 33:7 311 33:11 344, 346 33:13 341 33:16 252 34:6 312, 313, 325 34:7 339 34:9 449 34:13 331, 377 34:15 347 34:17 313 34:30 306, 309, 345 34:32 345 35:27 321 36–40 171 37:23 321 38:17 321 38:24 313 40:2 457 40:32(38) 312 Leviticus 1:3 313 1:6 321 1:8 321 2:5 322 3 331 3:4 321 3:11 334 3:16 334 4:2 353 4:3 332 4:5 62 4:6 330 4:8 332 4:15 308, 309, 344 4:17 330 4:28 92 4:29 338 4:31 136 4:35 136 5:2 313, 322 5:4 326 5:11 243, 324 5:18 338 5:22 92 5:23 92 6:5 371 1. Scripture references

Indexes 7:8 332 8:4 342 8:9 321 8:11 321, 377 9:2 312 10:6 312, 313 10:11 352 10:16 313 11:6 235, 312 11:17 240, 322, 323 11:19 323 11:29 322 11:35 377 11:42 116 13–14 322 13:41 313 14:7 330 14:13 338 15:4 313, 322 15:16 289 16:4 321 16:12 324 16:13 327 16:21 311 16:22 273 16:23 145 16:31 315 18:5 353 18:30 352 19:14 404 19:15 282, 353 19:19 312 19:20 312, 347 19:27 243 19:31 322 19:35 353 20:2 312 20:6 311 20:23 352 21:11 290 21:17 375 22:10 308 22:14 335 22:25 334 23:3 299 23:19 312 23:42 149 23:43 145, 146 24:2 404 24:6 330 1. Scripture references

24:12 117 25:5 311 25:6 308 25:7 312 25:11 319 25:14ff. 322 25:23 308 25:35 308 25:40 308 25:43 310 25:44 346 25:45 308 25:46 310 25:47 308 25:50 313 25:53 310 26:1 320, 403 26:4 92 26:13 346 26:19 377 26:30 320 26:33 311 26:41 312, 313 26:46 351 27:4 311, 322 27:16 285, 314 27:18 311, 322 Numbers 1:2 288, 342 1:4 343 1:16 343, 345 1:18 292, 342, 347 1:20 325 1:53 338 2 343 3:16 344 3:24 342 3:32 343 4:2 288 4:7 92 4:9 321 4:14 321 4:16 312 4:22 288 4:25 321 5:2 290, 313, 322 5:3 311, 313 5:15 324 5:18 326

5:20 92, 289 5:22 251, 400 5:28 292 6:12 125, 319 7 309 7:3 292 7:13 325 7:14 313 7:88 288 9:10 438 9:17 383 9:18 383 9:20 310 9:22 383 10:5 325 10:19-27 383 10:22 383 10:29-32 348 10:34 395 11:5 95 11:8 324 11:11 346 11:12 324 11:16 308, 309, 345 11:24 308, 309 11:25 308, 309 11:28 344 11:29 56, 57 11:30 308, 309 11:31 323 11:32 285, 314, 323 12:7-8 346 13:3 342 13:31 292 13:32 459 14:6 383 14:14 292 14:18 312, 325, 339 14:29 312, 314 14:30 383 14:38 383 15:3 401 15:16 352 15:30 335 15:28 292 16:3 344 16:7 198 16:13 343 16:25 308, 309, 345 17:5 341

477

Indexes 17:11 341 18:16 322 18:2 342 18:23 338 18:30 312 18:34 292 19:2ff. 121 19:9 121 20:8 341 20:10 341 20:14 325 20:15 308 21–26 64 21:1 292, 381 21:5 292 21:23 381 21:26 381 22:7 348 22:13 466 22:21 466 22:22 325 22:28 121 22:33 115 23:1 331 23:10 342 23:19 405 23:21 344 23:22 324 25:1 292 25:3 328 25:5 328 26 343 26:2 288 26:18(22) 312 26:41 342 27:5 353 27:11 116, 342, 352 27:17 380 29:39 401 30:6 97 30:7 326 30:13 97 30:17 352 31:10 324 31:21 312 31:26 288, 343 31:49 288 31:50 324 32:13 292 32:30 292

478

32:31 346 33:3 335 33:54 342 34:4ff. 323 35 298 35:2 293 35:3 293 35:4 293 35:5 293 35:6ff. 322 35:7 293 35:15 308, 312 35:29 352 35:33 327 36:13 353 Deuteronomy 1–3 64 1:5 350 1:6 198 1:7 325 1:13 344 1:15 312, 346 1:30 377 1:33 164, 377 1:38 344 1:44 57, 96 2:3 198 2:7 133 2:27 284 3:7 95 3:26 198 4:1 342 4:2 50 4:3 95 4:5 353 4:8 351 4:9 95 4:13 323 4:14 353 4:16 95, 313 4:21 326 4:23 95 4:26 312 4:32 403 4:44 351 4:45 354 5:1 462 5:10 353 5:20 328

5:22 341, 342 5:23 344 5:29 56 6:11 307 6:20 353, 354 6:25 312 7:3 313 7:5 331 8:17 400 9:6ff. 325 9:6 449 9:13 449 9:26 244 9:27 338, 346 9:29 132 10:6 283 11:10 106 11:14 324 11:15 312 12:15 322 12:17 320 12:22 322 14:7 323 14:8 322 14:20 308 14:21 322, 348 16:18 346 17:8 353 17:9 346 17:12 346 17:18 117, 312, 351 18:6 308 18:8 322 18:10 314, 322 18:11 322 18:16 342 18:19 353 19:15 338 19:17 346 20:5 345 20:8 345 20:9 345 20:14 312 21:3 62, 136 21:22 352 22:5 57 22:6-7 52 22:10 313 22:12 324 22:26 338 1. Scripture references

Indexes 23:7(8) 308 23:8 348 23:18 234, 328 23:20 311 24:1 322 24:3 322 24:7 298 24:8 322 25:2 97 25:14 321 25:15 321 26:2 314 26:5 308, 326 27:8 327 28:1 286 28:5 294 28:17 294 28:22 323 28:36 344 28:37 325 28:39 308 28:49 133 28:61 133 28:62 326 28:67 56 28:68 347 29:4 325 29:9 344 29:10 346 29:17 65, 324 29:19 107, 133 29:20 133 29:23 133 30:10 133 31:9 308, 309 31:10 149 31:28 308, 309, 344, 346 31:37 325 32 378 32:1-43 58 32:5 96 32:7 309 32:14 323 32:15 378 32:17 468 32:21 380 32:36 346 32:42 326 33:4 341, 342 33:6 326 1. Scripture references

33:17 324 33:29 403 34 64 35:5 344 Joshua 1:1 459 1:8 133, 298 1:18 133 4:14 51 4:24 331 5:8 51 6:17-18 402 7:20 250 7:21 323 8:33 322 9:1 325 9:2 322 9:10 180 9:18 323 11:6 326 11:9 326 13:3 131 15:3 122 15:4 323 15:11 180 15:47 180 16:3 135 JoshB 19:19 180 19:46 180 20:2 322 23 466 24:32 119, 323 Judges 1:1 459 1:27 462 1:35 462 2:2 134 3:12-30 195 3:23 196 JudgesA 5:2 JudgesB 5:2 5:3 53 JudgesA 6:32 JudgesB 6:32 6:39 52 9:9 97 9:19 52 9:45 191

326 326 176 176

10:9 135 11:12 463 11:13 97 11:35 97 13:2 467 13:25 462 14:4 135 14:5 122 14:20 324 15:3 457 17:1 467 18:15 53 19:19 324 JudgesB 19:29 19:29 327 20:31 462

321

Ruth 1:1 459 1:19 404 1:22 404 2:10 348 3:16 289 1Samuel (1Kingdoms) 1–2 466 1:1 467 1:3 179 1:13 400 1:19 164 2:25 190 3:21 447 4:6 133 5:4 176 6:4ff. 322 6:8 176 6:11 176 6:15 176 9:1 467 9:2 404 11:11 289 12 466 14:3 130 14:23 271 14:35 462 15:23 298 16:12 324 17:56 134 19:13 298 19:16 298

479

Indexes 23:22 136 24:8 396 25:5-6 296 25:21 463 25:16 97 30:6 135 30:7 130 2Samuel (2Kingdoms) 2Sam 2–1Kings 2:11 169 2:20 135 3:4 180 5:11 130 5:24 51 8:4 327 2Sam 11:1–2Kings 2:11 216 11:2 137 16:11 206 11:3 137 11:4 137 11:5 53 11:7 288 11:9 137 11:11 137 11:13 137 12:3 137 12:11 137 12:16 137 12:18 53 12:22 53 12:23 134 12:24 137 15:32 53 15:37 404 16:10 463 16:11 206 17:28 324 17:35 57 19–32 53 19:14 463 19:23 463 19:25 134 20:14 206 22:22 52 22:32 403 24:14 169

480

1Kings (3Kingdoms) 1:42 464 1:52 190, 191 2:4 136 3:17 52 5:13 324 6:32 321 6:35 321 6:36 322 8:9 323 14:24(23) 274 15:20 178 16:18 294 17:18 463 20:12 (*21:12) 147 21:10 463

203, 1Kings 22:1–2Kings 25:30 203 2Kings (4Kingdoms) 2:14 206 2:42 320 3:13 463 5:11 320 5:13 206 6:2 323 6:10 (4Kingdoms) 52 6:31 463 6:33 464 8:8 191 10:32 462 11:4 276 12:5 322 13:21 464 15:14 179 15:25 294 15:37 462 17:15 354 18:35 122 19–34 53 21:6 322 22:1 180 23:3 354 23:21 133 25:15 321 1Chronicles 6:42 322 6:52 322 15:21 178

15:28 16:19 18:4 23:29 28:4 29:2 29:19

325 326 327 321, 322 112 321 354

2Chronicles 3:14 321 3:16 321 4:17 324 4:21 321 4:22 321 5:10 323 6:28 323 7:9 320 7:20 325 8:4 326 8:6 326 10:15 180 13:11 322 13:20 180 13:21 180 16:13 289 17:12 326 19:10 353 20:20 138 26:9 321 28:32 339 29:25 353 29:39 320 31:21 353 33:6 322 34:31 354 35:21 463 1Esdras 1:14 178 2:12 175 2:13 175 2:19 175 3:5 457 8:53 106 2Esdras 9:15 294 15:14 122

1. Scripture references

Indexes Ezra 3:3 122 3:12ff. 325 5:4 134 7:12 351 7:14 351 7:21 351 7:25 351 7:26 351 9:1-2 122 9:1 134 9:14 122 Nehemiah 3:19-25 321 5:14 (2Esd 15:14) 122 5:15 323 8:18 320 9:32 325 10:36 320 12:28 324 13:20 322 13:31 319 Esther 2:23 154 3:11 96 4:4 396 4:10 96 4:13 96 4:17 95 7:8 96 8:3 63 8:17 92 9:2 92 Judith 393-406 16:15 107 Tobit 3:3 339 4:6 467 5:3 396 6:6 276 12:8 97 13:6 467 1Maccabees 1:14 125 2:15 116 1. Scripture references

2:17 116 2:49-70 466 3:15 52 6:35 91 6:45 91 10:8 460 11:66 92 13:39 339 2Maccabees 3:37 114 4:9 125 4:12 125 4:14 125 6:8 291 7:17 132 8:33 457 12:14 206 12:45 339 14:32 14 15:5 132 15:39 59

15:21 124 17:1 114, 388 17:4 388 17:7 388 18:4 388 18:10-18 316 18:20-21 66 19:17 388 407-415

3Maccabees 416-420 2:20 59 3:23 122 3:25 122 4:12 122 5:40 125 6:25 125 6:36 351 7:5 347 4Maccabees 1:1 124 1:13-30 66 4:12 125 4:20 125 5:12 132 5:35 124 5:36 132 7:1-3 66 7:7 124 8:7 351 11:7 388 13:27 410 14:1 206 14:14 125 14:18 125 15:11 410

421-428

Psalms 1:2 157 3:7 52 10(9):10 403 10:14(9:35) 158 12(11):3 399 14(13):2 97 14(13):3 96 14(13):7 56 18(17):2 159 18(17):3 403 18(17):17 467 18(17):22 52 18(17):27 59 18(17):32 159 18(17):47 159 18(17):49 97 19(18):15 159 20(19):2 403 22(21):14 122 22(21):21 95 22(21):22 324 22(21):27 93 24(23):1 94 25(24):3 92 24(24):7 339 25(24):11 94 25(24):18 94 26(25):6 93 27(26):2 96 27(26):5 149 27(26):14 365 28(27):1 159 28(27):3 401 28(27):8 159 28(27):5 97 29(28):3 467 29(28):9 158 29(28):10 323 30(29):10 143, 144 30(29):7 96

481

Indexes 31(30):3 31(30):12 31(30):14 31(30):15 31(30):21 31(30):25 32(31):5 32(31):6 32(31):7 32(31):8 33(32):9 33(32):10 33(32):12 34(33):13 34(33):14 35(34):20 35(34):25 36(35):6 36(35):11 37(36):2 37(36):34 38(37):6 38(37):9 38(37):14 38(37):21 39(38):8 40(39):8 40(39):11 40(39):13 41(40):5 41(40):7 41(40):10 43(42):2 44(43):23 45(44):8 45(44):10 46(45):2 46(45):6 46(45):9 46(45):10 48(47):4 48(47):8 48(47):11 49(48):3 49(48):5 49(48):7 49(48):11 49(48):12 49(48):14 50(49):13

482

159 94 94, 324 96 97, 149 97 96 97, 467 116 97 405 341 341 59 94 401 96 92 94 324 365 324 122 96 325 365 96 96 94 96 94 97 159 468 176 93 403 371 95 95 294 240 95 206 324 93 59, 376 122 94, 404 96

50(49):16 92 51(50):4 59 51(50):6 93 51(50):7 92, 325 51(50):9 59, 324 51(50):10 444 51(50):12 59, 377 52(51):9 159 53(52):3 97 53(52):5 97 53(52):7 56 54(53):2 93 54(53):4 93 54(53):7 129 55(54):12 290 59(58):4 339 59(58):14 95 59(58):18 403 60(59):8 97, 149 61(60):3 95 63(62):7 157 63(62):12 97 64(63):6 96 64(63):10 97 65(64):6 95 65(64):9 95 66(65):3 96 66(65):8 444 67(66):2 96 67(66):7 93 67(66):8 95 68(67):6 92 68(67):14 326 68(67):21 323 68(67):24 92 68(67):28 95 69(68):5 97 69(68):8 96 69(68):21 365 69(68):23 93 69(68):26 324 69(68):27 97 69(68):32 95 70(69):3 96 71(70):9 95 71(70):10 96 71(70):18 95 72(71):14 290 73(72):28 93 74(73):14 94

74(73):16 323 75(74):5 96 76(75):9 444 77(76):6 93 77(76):11 96 77(76):13 157 77(76):18 97 77(76):20 467 78(77):1 325 78(77):14 92 78(77):17 274 78(77):18 93 78(77):19 96 78(77):21 287 78(77)24 323 78(77):26 240 78(77):35 159 78(77):45 323 78(77):51 93 79(78):4 94 80(79):2 94 80(79):9 94 81(80):10 92 81(80):11 94 81(80):15 95 82(81):5 97 82(81):6 96 83(82):5 96 83(82):13 96 84(83):4 92 86(85):15 94 89(88):3 96 89(88):23 52 89(88):33 456 89(88):36 463 89(88):39 287 90(89):3 96 91(90):5 96 91(90):11 459 92(91):15 95 92(91):16 467 93(92):4 467 94(93):7 96 95(94):3 244 95(94):4 95 95(94):10 96 95(94):11 463 96(95):7-8 96 96(95):10 96 98(97):3 95 1. Scripture references

Indexes 99(98):8 106 101(100):8 92 102(101):7 323 102(101):9 97 102(101):19 251 103(102):3 107 104(103):18 95, 323 104(103):21 122 104(103):28 95 104(103):31 97 104(103):35 96 105(104):12 326 105(104):27 95 105(104):29 95 105(104)31 323 105(104):40 323 105(104):45 351 106(105):7 97 106(105):20 97 106(105):28 328 106(105):33 326 106(105):37 468 106(105):38 327 107(106):23 467 107(106):32 94 109(108):18 95 110(109):1 468 111(110):8 92 113(112):6 92 114(113):1 348 115(114):8 96 116(114?):9 328 118(117):2 96 118(117):4 96 118(117):6 468 122(121):7 294 126(125):2 59 143(142):5 157 144(143):7 467 150:6 251 Psalms of Solomon 16:10 125 Proverbs 1:7 50 1:8 351 2:7 403 6:1 60 6:8 50 1. Scripture references

6:20 351 10:17 59 13:9 59 14:13 59 18:1 125 18:8 125 18:19 294 19:15 125 23:3 323 24:28 328 24:61 323 26:23 411 26:25 396 29:1 325, 449 30:26 323 Ecclesiastes 1:11 216 1:14 216 1:17 216 2:1 216 2:7 216 2:8 216 2:9 398 2:12 216 2:14 216 2:15 216 2:17 216 2:18 216 2:19 216 2:21 216 2:23 216 2:24 216 2:26 216 3:1 216 3:10 216 3:11 216 3:13 216, 218 3:14 218 3:17 216 4:1 216 4:2 216 4:3 216 4:4 216 4:8 216 4:11 216 4:14 216 4:15 216 4:16 216 5:3 216

5:6 216 5:9 216 5:15 216 5:16 216 5:18 216 6:3 216 6:5 216 6:7 216 6:9 216 7:6 216 7:14 216 7:15 216 7:18 216 7:21 216 7:22 216 7:26 216 7:29 216 8:8 216 8:9 216 8:10 216 8:12 216 8:14 216 8:15 216 8:16 216 8:17 52, 216 9:1 216 9:3 216 9:6 216 9:11 216 9:12 216 9:13 216 9:14 218 9:15 216, 218 10:3 216 10:17 197 10:19 216 10:20 216 11:2 216 11:5 216 11:6 134 11:7 216 12:1 216 12:5 216 12:7 216 12:9 216 12:14 216 Song of Songs (Canticum) 1:10 324 1:17 323

483

Indexes 4:4 4:14 5:3 6:11 7:3 7:10 7:14

194 176, 179 197 199 321 198 324

Job 1–2 441, 443 1:9 331 1:20 135 2:9 50 3:13 218 5:18 218 6:8 56 7:8 206 8:6 208 9:9 124 9:14 218 12:5 218 12:14-15 59 14:3 56 14:6 211 14:13 56 14:16 211 15:4 206 15:10 206 15:20 207 15:27 321 18:7 211 18:15 208 19:18 206 19:20 198 19:23 56 20:21 206 21:33 206 22:24 57 23:3 114 23:15 115 24:18-20 378 26:6 206, 324 27:1 63 27:19 206 28:14 206 28:23-24 59 28:27 206 29:2 56 30:2 206 30:15 59

484

30:29 31:4 31:31 31:35 31:37 32:5 32:12 33:33 34:21 35:15 36:11 36:16 36:29 36:33 37:1 37:11 37:18 37:19 37:22 38:8 38:24 39:8 39:9 40:10 41:17 42:11 42:14 42:16

Prol. 26 40, 47 Prol. 27-36 41 Prol. 33-36 41 4:14 381 6:37 353 9:10 381 14:18 377 16:11 449 20:19 376 23:2-3 339 24:30-31 376 39:16 353 39:18 353 40:16 242 43:13 353 45:24 276 46:9 95 46:17 96

124 211 56 56 211 206, 207 206 206 211 206 208 206 206 206 206 206 207, 208 115 208 123 240 200 324 208 207, 208 119, 323 123 122

Wisdom of Solomon 392 1:1-14 59 9:3 67 11:15 125 12:2 403 13:6 410 13:14 121 14:23 351 17:18 122 Ben Sira Prol. 1-14 40 Prol. 1 47 Prol. 7-11 41 Prol. 15-26 40 Prol. 18-21 41 Prol. 20 46, 47 Prol. 21 40 Prol. 22 40 Prol. 24 47

386-

Hosea 1:2 462 4:12 384 4:14 328, 380 5:1 352 5:2–13:4 384 9:4 51 9:6 238 11:10 122, 379 12:2 376 12:4 327 12:5 327 13:15 240 Amos 1:3–2:6 382 1:3-15 382 1:9 382 1:8 383 1:14 375 2:1-6 382 2:13 376 2:14-15 383 2:16 198 3:5 200 3:9-11 294 3:9 444 3:14 377 5:2 378 5:7 381 5:15 383 1. Scripture references

Indexes 5:16 5:26 5:27 6:8 6:11 7:8 7:12 8:2 9:1 9:11 9:12

378 149 465 381 378 379 381 379 380, 383 380, 381, 383 383

Micah 1:8 124 4:3 93 5:4-5 399 7:3 384 Joel 3:2 380 Obadiah 14 381 Jonah 1:7 399 1:9 331, 382 1:10 460 1:13 382 1:16 460 2:4 381 2:5 96 2:11 382 4:2 135 4:6 285 4:8 240, 384 Nahum 1:12 467 2:1(1:15) 444 3:7 377 Habakkuk 1:4 352 1:7 352 1:12 352 2:2 327 2:6-19 384 3:2 378 3:10 96 1. Scripture references

Zephaniah 2:3 352, 382 2:14 327 2:15–3:18 384 3:3 122 3:8 365 Haggai 1:5 384 1:7 384 2:2 95 2:3 379 2:8 384 2:11 351, 384 2:15 384 2:18 384 Zechariah 1:6 352 3 441, 443 5:2 379 7:9 352 10:1 384 10:7 382 11:3 122 11:16 97 Malachi 2:6 351 2:11 378 3:9 363 Isaiah 1:4 341 1:6 324 1:8 149 1:10 351 1:18 379 1:29-30 376 1:29 207 2:3 351 2:18 320 2:20 323 3:5 377 3:10 96 3:22 123 4:2-3 382 6:6 321 7:7 135 7:9 138

7:14 219, 220 8:16 351 8:21 249 10:3 458 10:6 252 10:9 349 11:11 349 13:2 401 13:21 124 13:22 124 14:1 249, 322 14:12 234 17:10 240 19:7 242 19:12 240 19:18-21 232 19:18 180 19:25 235 24:5 351 24:7 379 24:8 379 24:11 379 24:16 404 25:2 294 26:10 467 27:3 377 27:8 240 27:9 377 27:26 240 28:10 285 29:10 239 29:13 331 30:4 239 30:9 351 30:13-18 238 30:23 276 32:14 294 34:11 124, 240, 322 34:13 124, 294 34:14 124 37:23 401 37:31-33 53 37:38 249 40:15 91 41:18 274 43:4 377 45:13 243 46:6 403 50:1 322 50:3 324

485

Indexes 51:1 134 51:7 134 51:19 377 52:2 274 53:7 94, 96 56:1 447 56:2 95 56:3 404 58:4 321 59:7 446 59:9 365 59:21 399 60:4 96 61:3 207 65:11 234 65:25 444 66:1 135 66:14 331 66:17 322 Jeremiah 1:6 328 2:6 376 2:15 122 2:34 327 3:3 377 3:8 322 4:10 328 4:19 123 4:30 377 7:18 284 8:7 324 9:7 401 9:20-21 294 10:3 352 10:5 322 10:9 322 11:20 399 14:8 365 14:13 328 17:27 294 18:17 240 20:10 324 20:12 399 22:10 377 27:39 124 28:13 467 28:55 467 32(39):17 328 32:17 457

486

36(29):5 376 36(43):30 324 36(29):28 376 37(30):18 294 38(31):8 294 39(32):23 353 44:1 238 44(51):28 326 46(26):2 179 46(26):14 241 49:27/30:16 294 50(27):12 274 51:1 180 51(44):10 353 51:20-23 382 Baruch 2:30 449 Epistle of Jeremiah 31 122 39 411 44 411 51 411 56 411 63 411 Ezekiel 1:10 351 1:14 178 2:3 95, 351 4:3 322 4:9 324 6:13 320 8:11 327 8:16 351 10:2 324 10:7 324 10:9 349 14:1 96 16:11 324 16:17 95 16:34 136 16:52 95 17:10 240 19:12 240 19:13 96 20:29 178 21:16 459 21:27 401

21:31 321 23:42 125 24:4 321 24:5 351 24:6 321 24:16 57 26–27 176 26:13 324 27:5 58, 120 27:26 240 28–39 258 28 176 28:10 285 28:13 321 29:10 239 30:4-5 239 30:9 351 30:13-18 238 30:14 239 30:15 239 30:16 239 30:23 276 32:42 125 33:8 96 33:13-14 96 34:6 95 36:24 122 37:7 125 37:38 284 38:15-16 380 38:23 378 42:13 338 44:6 198 44:30 320 45:9 198 45:11 285 45:12 324 45:13 243, 285 46:12 320 47:22 322 Daniel 1:20 124 2:2 324 2:14 124 2:18 124 2:24 124 2:27 324 2:28 447 3:16 411 1. Scripture references

Indexes 3:46-50 418 4:1 97 4:37 93 5:3 96 5:4 96 6:1 95 9:4 353 9:5 339 9:15-16 339 9:21 464 9:27 331 Susannah 7 401 12Th 96 13Th 96 18Th 96 20Th 96 27Th 96 28 443 28Th 96 50Th 96 52Th 92, 96 56Th 93 Gospel of Matthew 1:1 464 1:1-17 464 1:20 464 1:22-23 463 1:23 464 2:1 464 2:10 460 2:13 464 2:15 463 2:17-18 463 2:19 464 2:23 463 3:12 459 3:13 464 4:14-16 463 4:23 443 5–7 463 5:13 455 5:15 464 6:9 444 7:16 463 7:28 463 8:11 444 8:17 463 1. Scripture references

8:19 457 9:10 459, 464, 465 9:17 464 9:18 464 9:32 464 10 463 11:1 463, 464 11:25 447 12:17-21 463 12:39 463 12:46 464 13 463 13:3 464 13:14-15 463 13:29 463 13:35 463 13:53 463 16:22 302 17:5 464 17:14 414 17:26 414 18 463 19:1 463 21:4-5 463 21:32 464 23–25 463 24:27 444 24:45 464 26:1 463 26:47 464 26:69 457 27:9-10 463 27:51 443 27:60 132 28:1 457 18:11 464 Gospel of Mark 1:1 462 1:2-3 462 1:6 462 1:9 459, 465 1:13 462 1:22 462 1:24 463 1:28 461 1:30 462 1:32 461, 462 1:35 461 1:39 443

1:40 438 1:45 462 2:3 462 2:6 462 2:18 462 2:23 465 2:25 461 3:1 462 3:2 462 3:20 444 3:21 462 3:32 462 4:2 461 4:4 465 4:8 457 4:20 457 4:23 443 4:38 462 4:39 461 4:41 460 5:5 462 5:7 463 5:11 462 5:14 462 5:15 461 5:19 461 5:34 444 5:35 462 5:39 461 5:42 460 6:4 461 6:25 461 6:33 462 6:43 462 6:47 438 6:54 462 7:21 461 7:32 462 7:33 461 8:12 463 8:17 461 8:22 462 9:2 461 9:4 462 10:2 462 10:13 462 10:22 462 10:28 462 10:32 462 10:49 462

487

Indexes 11:29 12:44 13:9 13:11 13:19 13:20 14:1 14:4 14:12 14:18 14:30 14:40 14:49 14:54 14:61 15:14 15:26 15:27 15:38 15:40 15:43 16:2 16:9

457 461 462 462 462 462 462 462 462 462 462 462 462 462 462 462 462 462 443 462 462 461 457

Gospel of Luke 1–2 466 1:1-4 464 1:5 465, 466 1:8 457 1:12 466 1:37 449, 457 1:46-55 466 1:65 449 1:67-79 466 1:71 466 2:1 466 2:9 460 2:19 449 2:51 449 3:17 459 4:10 459 4:34 463 4:44 443 5:1 459 5:4 413 6:12 466 7:50 444 8:28 463 10:5 444 10:37 457

488

11:2 11:27 11:29 12:36 12:54 13:29 16:9 16:11 21:24 22:11 23:45 24:31

444 466 463 413 444 444 461 461 401 462 443 132

Gospel of John 1:6 467 2:4 463, 467 3:1 467 3:21 467 3:23 467 6:24 207 7:18 467 Acts 1:16 466 2:14 466 2:23 466 3:18 466 3:21 466 5:7 459 5:42 413 6:63 408 7:25 466 7:43 465 7:51 449 8:21 457 10:15 462 10:37 449 11:30 466 14:11 466 15:2 466 17:7 457 17:16-34 465 17:22-31 465 18:21 462 19:17 466 20:7 461 22:22 466 Romans 1:4 469

1:18 447 1:21 469 1:24 469 2:7-8 469 3:15 446 3:20 469 4:17 469 4:25 469 8:31-39 469, 472 8:32 449 9:2 469 9:17 446, 448 10:9-10 468 11:8 302 11:33-36 469 11:33 469, 472 15:6 468 1Corinthians 1:18 469 4:10 469 4:21 456 7:27 469 10:13 469 10:20 468 13 468 15:42-43 469 15:53-54 469 16:2 461 2Corinthians 3:14-15 302 10:1 469 10:7 469 11:22-23 469 Galatians 2:11 469 3:19 469 Ephesians 5,3 447 Philippians 2:6-11 468 3:13 469 1Thessalonians 2:19 469 3:13 469 1. Scripture references

Indexes 5:11 457 2Timothy 4:7-8 426 2Peter 1:20 457 2:12 458 1John 1:3 444 1:6 467 Hebrews 1:1-4 470

3:11 463 4:3 463 4:5 463 6:19 443 9:3 443 10:2 413 13:22 470 Revelation 1:15 467 4:11 472 5,9-10 472 6:8 456, 467 9:11 467

11:17-18 472 12:7 466 14:2 467 15:3-4 472 17:1 467 17:6 460 19:1-2 472 19:5 472 19:6-8 467 19:6 467 21:13 444

2. Greek words ᾅδης 319 ἀγάπη 255, 447 ἀγαπάω 107, 108, 447 ἄγγελος 443 ἅγιος 289, 331, 447 ἅγνισμα 121 ἀγχιστεύω 319 ἀδυναμέω 46 ἀθάνατος 66 ἀκαθαρσία 319 ἀκάθαρτος 319 ἀκροβυστία 251, 255, 292 ἀλήθεια 447 ἁμαρτάνω 250, 336, 338 ἀμνός 323 ἀναβαίνω 91 ἀνάθημα 320 ἄνεμος 240 ἀνεμοφθορία 323 ἄνθρωπος 317 Ἀντιλίβανον 325 ἀόρατος 125 ἀπαρχή 320 ἀπό 387, 397 ἀποκαλύπτω 447 ἀποστρέφω 129 ἄρδω 105 ἀρετή 66, 435 ἀρχή 51, 288, 295 αὐτόχθων 322 2. Greek words

ἄφεσις 435 ἀχανής 388 βασιλεύς 277, 344 βαστάζω 446 βδέλυγμα 279 βδελύσσω 279 βία 122 βοηθός 159 βούτυρον 323 γειώρας 322 γένεσις 325 γλυπτόν 320 γλῶσσα 40 γονορρυής 322 γραμματεύς 325, 345, 346 γυνή 317 γωνία 321 δεκτός 320 δέομαι 52 δέρρις 321 δή 227 δῆμος 325 διάβολος 441, 443 διαγογγύζω 323 διαθήκη 51, 253, 276, 277, 289, 320, 443 διακαλύπτω 447

διαταγή 280, 359 διατάσσω 196 διαφορά 47 δίδραχμον 323 διέξοδος 323 διήγημα 325 δικαιοσύνη 447 διχοτόμημα 321 δοκός 323 δόξα 282, 443, 447 δορκάς 322 δυσδιήγητος 288 δυσκλεής 122 δῶρον 320, 330 ἐγγαστρίμυθος 322 ἐγκισσάω 325 ἐγκρυφίας 323 ἔδεσμα 323 ἔθνος 250, 255, 340 εἴδωλον 251, 252, 254, 298, 320 εἰρήνη 295, 296 εἰς 397 εἷς 129 ἐκλεκτός 325, 344 ἑκούσιος 320 ἐλάτινος 120 ἔλαφος 322 ἐν 19, 21, 27, 398

489

Indexes ἔναντι 164, 250 ἐναντίον 164, 250 ἐνδιαβάλλω 325 ἕνεκα 27 ἐνεχυράζω 322 ἐντέλλω 233 ἐνώπιον 164, 250, 323 ἐξιλάσκομαι 320 ἑξόδιον 320 ἐπαοιδός 322, 324 ἔπαυλις 324 ἐπί 398 ἐπιδέκατος 320 ἐπισιτισμός 324 ἐπιτίθημι 320 ἑρπετόν 323 ἐρυθαίνω 121 εὐιλατεύω 107 εὐίλατος 106, 107 εὐλογέω 31, 253, 254 εὐνοῦχος 233 εὐπρέπεια 208 εὑρίσκω 198 εὐφροσύνη 59 ἐφύβριστος 388 ζωή 129 ἡδονή 66 ἥλιος 318 θέλω 115 θεός 289, 329 θνησιμαῖον 322 θυΐσκη 321 θυμίαμα 321 θυσιαστήριον 320, 331, 332 ἶβις 322 ἱκανόω 198 ἰσχυρός 207, 208 ἰσοδυναμέω 40, 46 καί 21, 27 καί γε 53, 203-219, 258 κατακλυσμός 323 καταπέτασμα 280, 359, 321, 359, 443 κατάσχεσις 325 κέρας 123

490

κιβωτός 320 κίδαρις 321 κληδονίζομαι 322 κοινή 25 κοίτη 289 κόσμος 66 κρατήρ 321 κρεάγρα 321 κρίσις 67 κτίζω 253 κυνόμυια 323 κύριος 51, 289, 458 λαβίς 321 λάγανον 321 λάχανα 324 λέβης 321 λέγω 225, 227 λειτουργέω 320 λέξις 40, 47 λιθοβολέω 322 λόγος 51, 132 λουτήρ 321 λοχεύομαι 325 λυχνία 321 μακρόθυμος 325 μανδραγόρας 324 μαρτύριον 200, 353, 354 μελετάω 157 μέλλω 115 μέλος 321 μετά 398 μέτρον 321 μηρία 321 μιαρός 91 μονόκερως 324 μόχθος 325 μυρεψός 321 μῦς 322 μώλωψ 324 μῶμος 322 νηπιοκτόνος 388 νόμος 51, 282, 320, 350-352, 442, 447 νυκτερίς 323 νυκτικόραξ 323 νυμφαγωγός 324

ὀβολός 324 ὁδός 51 ὁλκή 325 ὄλυρα 324 ὁμολογία 320 ὀρθρίζω 325 ὀρτυγομήτρα 323 ὀσφύς 325 οὐδείς 90 ὄψιμος 324 παγετός 324 παιδεία 40, 41, 42, 67 παίω 121 παλαιόω 325 παρά 398 παράδεισος 265, 307, 323 παρακαλέω 197 παραπικραίνω 325 παροικία 40, 46 πᾶς 19, 27 πάσχα 177, 181, 249, 255, 284, 442 πέταλον 321 περιβόλαιον 324 περικομπέω 388 περιτέμνω 251 περίχωρος 324 πίστις 447 πλάξ 323 πληθύνω 64 πλημμέλεια 320 πλήρωσις 321 πνεῦμα 240 πολυέλεος 325 ποταμός 325 ποτίζω 105, 106 πρᾶγμα 51 πρᾶσις 322 προσανάβασις 122 προσκυνέω 320 πρόσταγμα 233, 353 προστίθημι 19, 27, 55, 63 πρόσωπον 164, 165 προφήτης 47 προφητεία 47 πρωτογένημα 320 πρωτότοκος 280, 320, 359 πυγμή 321 πυρεῖον 321 2. Greek words

Indexes πυρράκης 324 ῥῆμα 51 ῥοΐσκος 321 σάββατα 174, 177, 179, 181, 249, 284 Σατανᾶς 441 σάρδιον 321 σεμίδαλις 321 σημασία 325 σίαλος 91 σίκερα 249, 255, 284 σκληροτράχηλος 325 σμύρνα 321 σοφία 40 σοφός 125 σπαρτίον 324 στερέωμα 323 στήλη 320 στίχος 322 σύ 132 συναγωγή 320, 340-342, 443

σύνδεσμος 271, 274 σύνθεσις 322 συντίμησις 322 σωτήρια 293 σωτήριον 320, 331 τάσσω 233 τήγανον 322 τίθημι 149 τιθηνός 324 τορευτός 322 τροφός 324 τρυγών 324 ὑμεῖς 132 ὑπομονή 365, 435 ὕσσωπος 324 ὕψος 208 φακός 324 φαρμακός 322 φαῦσις 323 φιλία 447

φίλος 234 φιλόσοφος 66, 124 φιλέω 107 φρουρά 92 φυγαδευτήριον 322 φυλακίζω 388 χαρά 59 χείρ 169, 324 χειροποίητος 320 χοιρογρύλλιος 323 χολή 324 χόρτασμα 324 χύτρα 324 χωνευτός 320 ψαλτήριον 324 ψόγος 324 ψυχή 51, 265, 290 ὠρύομαι

122

3. Modern authors Abbott, T. K. 23, 434, 435, 446 Abel, F. M. 24, 25, 90, 184, 359, 408, 451, 458 Accorinti, D. 371 Adams, J. N. 263 Adrados, F. R. 26, 271, 425, 432 Ackroyd, P. R. 29 Aejmelaeus, A. 30, 48, 61-63, 112, 113, 141, 145, 146, 151, 153, 164, 166, 168, 171, 225, 454, 458 Aitken, J. K. 101, 150, 185, 186, 218, 366, 367, 375-378, 380385 Albrecht, J. 43, 44, 48, 49 Alexander, Ph. 416 Alexiou, M. 82 Allen, L. C. 316 Allen, W. S. 130, 131 Anderegg, J. 368 3. Modern authors

Andersen, C. 357 Anderson, H. 89, 416, 419 Angers, D. 302 Anz, H. 98 Assan-Dhôte, I. 173, 214 Astour, M. C. 182 Atkinson, B. F. 19 Aune, D. C. 422, 471 Ausloos, H. 29, 176, 191, 193, 195, 201, 302 Austermann, F. 301, 454, 456 Auwers, J. M. 289, 336 Baab, O. J. 170 Baars, W. 271 Baasten, M. F. J. 317 Bachmann, M. 403 Bagnal, R. S. 101, 102, 104, 105, 317 Bailey, K. D. 140 Bainvel, J. V. 372

Bakker, E. J. 26, 80 Balz, H. 403 Banfi, E. 263, 264 Barber, A. 271 Barr, J. 30, 48, 50, 52, 53, 140143, 145-149, 167, 185, 188, 193, 201, 261, 272, 289, 290, 317319, 435, 442 Barth, G. 403 Barthélemy, D. 170, 204, 205208, 219, 241, 258, 373 Bartelink, G. J. M. 357 Bassnett, S. 139, 142 Bauckham, R. 471 Bauer, W. 272, 442, 455, 460, 462, 463, 466, 469, 471 Baumgärtel, F. 170 Beaton, R. 82, 83 Beaucamp, E. 336, 337 Bellia, G. 389 Benko, S. 357 Benor, S. Bunin 246

491

Indexes Bérard, F. 104 Bévenot, H. 409 Bennett, J. 143 Berthelot, K. 250, 402 Bertram, G. 218 Beyer, K. 29, 452 Bickermann, E. J. 89, 185, 416 Bizzetti, P. 389 Black, M. 28, 29 Blank, S. H. 350, 351, 353 Blass, F. 207, 216, 426, 438, 445, 452, 460, 470 Blomqvist, J. 20, 112, 118 Blumenthal, E. 232 Bogaert, P. M. 241, 304 Bons, E. 44, 56, 57, 59, 67, 120, 298, 302, 363, 367, 375, 376, 379, 380, 388, 402, 403, 404, 418, 442, 444, 455, 458 Borret, M. 357 Botterweck, G. J. 350 Bottini, G. C. 262, 264 Boulanger, A. 372 Le Boulluec, A. 345 Bousset, W. 471 Boyd-Taylor, C. 144, 147, 156, 158, 159, 296, 298, 302, 418, 420 Brakke, D. 330 Breitenstein, U. 89, 421, 423426 Briend, J. 335 Brine, K. R. 404 Bringmann, K. 423 Brixhe, C. 78, 109, 111, 259, 260, 373 Brooke, G. J. 202 Brock, S. P. 141, 142, 185, 300 Brown, J. P. 364 Browning, R. 104 Brønno, E. 131, 179 Browning, R. 26 Brucker, R. 469 Bubeník, V. 259, 263 Büchner, D. 315 Büchsel, F. 23 Bujard, W. 470 Bultmann, R. 468 Burkitt, F. C. 147 Burney, C. F. 22

492

Butera, C. J. 248 Buttrick, G. A. 25

Curtius, E. R. 420 Cuzzolin, P. 263

Cadbury, H. J. 464 Cadell, H. 100, 106, 304 Caird, G. B. 271, 290, 300, 316 Calderini, A. 303 Cancik, H. 461 Canova, R. 263, 264 Caragounis, C. 109 Carrez-Maratray, J. Y. 239 Casabona, J. 330 Casetti, P. 316 Casevitz, M. 299, 369 Cassio, A. C. 76 Cavalier, C. 155 Černý, J. 236, 242 Cerutti, M. V. 330 Chang, S. 72 Chantraine, P. 271, 427 Charles, R. H. 416, 471 Charlesworth, J. H. 416, 421 Chesterman, A. 139, 163 Childs, B. S. 28 Chiron, P. 367 Christidis, A. F. 26, 263, 432 Christoi, A. 373 Cignelli, L. 36 Cimosa, M. 35, 301, 402 Clackson, J. P. T. 71 Coles, R. A. 370 Collins, J. J. 260, 262 Collins, N. L. 202 Colwell, C. E. 25, 415 Colvin, S. 74 Consani, C. 259, 260 Contardi, F. 239 Conybeare, F. C. 19, 167, 451, 456-459, 462-464 Cook, J. 351, 393, 437 Copeland, R. 140 Corley, J. 218, 393, 401, 405 Costas, P. S. 26 Cotton, H. M. 262 Cousland, J. R. C. 418 Cox, C. E. 17, 30, 114, 163, 175, 188, 189, 301, 363 de Crom, D. 217-219 Croy, N. C. 417, 419

Dahmen, U. 318 Dalman, G. 22, 183 Daniel, S. 300, 317, 318, 332, 338 Daniels, D. R. 244 Danker, F. W. 99 Daris, S. 303 Davidson, J. E. 338 Davies, W. D. 260 Debrunner, A. 395 Deichgräber, R. 472 Deissmann, A. 17, 21, 22, 99, 183, 223, 231, 280, 306, 359, 369, 418, 436, 437, 440, 467 Delling, G. 196, 472 Denniston, J. D. 390, 425 DeSilva, D. A. 421, 424 Dibelius, M. 470 Dickey, E. 145 Dietrich, K. 22 Dihle, A. 66, 67, 421, 425 Dimant, D. 363 Dindorf, W. 120 Dines, J. M. 35, 59, 139, 140, 151, 257, 296, 302, 367, 376, 379, 381-385 Dodd, C. H. 431 Dodson, J. R. 389, 391 Dogniez, C. 106, 286, 299, 300, 301, 341, 343, 344, 351, 354, 361, 364, 369 Dorival, G. 35, 51, 100, 173, 202, 246, 257, 258, 292, 293, 295, 297, 299, 300, 303, 304, 317, 331-333, 341, 344, 345, 349, 352, 357, 367, 372, 397, 399, 443 Doran, R. 408, 414 Dörrie, H. 421 Dover, K. 112 Dowty, D. R. 114 Drews, R. 72 Driver, G. R. 361 Dschulnigg, P. 461 Dunn, G. 111 Dupont-Sommer, A. 89, 421, 424 3. Modern authors

Indexes Eckholdt, J. F. 41 Edwards, M. J. 422 Eideneier, H. 82 Elliger, K. 143 Emmet, C. W. 416, 418, 419 Engel, H. 393, 403, 405 Ernst, N. 120 Evans, C. A. 311 Evans, T. V. 36, 55, 56, 57, 58, 65, 101, 113-115, 187, 223, 452 Eve, M. 87 Eynikel, E. 301 Fabry, H. J. 232, 279, 289 Fanning, B. M. 452 Faraone, C. 370 Fecht, G. 238 Feldmann, L. H. 261 Festorazzi, F. 233 Field, F. 198 Finet, A. 178 Fitzmyer, J. A. 449 Flashar, M. 283, 289, 316, 325 Focke, F. 387, 388 Foraboschi, D. 303 de Foucault, J. A. 112, 113, 115, 116 Fournet, J. L. 173, 177, 242 Fraenkel, D. 30, 187, 190, 360, 363 Frankel, Z. 17, 175 Frédéric, M. 371 Fredouille, J. C. 357 Freedman, D. N. 89 Freudenthal, J. 421, 423 Frey, J. 471 Friedrich, G. 16 Fritsch, C. 290, 451 Frösén, J. 33, 432 Fuhrmann, W. 367 Füssel, S. 235 Gammie, J. G. 389 Gardiner, A. H. 236 Garrett, A. 72 Garsiel, M. 201 Gärtner, E. 387 Gelston, A. 382 Genette, G. 448 George, C. H. 26 3. Modern authors

Georgoudi, S. 300 Gera, D. L. 393, 396, 397, 401, 406 Gera, D. R. 375, 379 Gehman, H. S. 27, 184, 247, 300, 364 Gentry, P. J. 204, 205, 207209, 211-216 Gerleman, G. 316, 318 Gerö, E. C. 114 Giangrande, G. 370 Giebel, M. 125 Gignac, F. T. 22, 33, 91, 104, 452 Gil, L. 183, 414 Gilbert, M. 386-389, 392 Glare, P. G. W. 271, 319 Glenny, W. E. 150, 151, 195, 198 Goeken, J. 372 Goessling, F. 365, 369, 371 Goldstein, J. A. 407, 408, 415 Gómez, P. Cantos 221 Gooding, D. W. 415 Goppelt, L. 470 Gordis, R. 200 Görg, M. 232, 235, 236, 237, 242, 244, 289 Görgemanns, H. 421, 422 Graf, F. 370 Greenspahn, F. E. 283 Greenspoon, L. 202 Gregg, A. F. 391 Grelot, P. 366 Grieve, A. 99 Grillet, B. 173, 176 Grimm, C. L. W. 387, 416, 418, 420 Groll, S. I. 232 De Groot, A. W. 372 Guinot, J. N. 357 Gutbub, A. 240 Gzella, H. 130, 284 Haag, E. 182 Haase, W. 16, 206, 357 Habachi, L. 238 Habicht, C. 408 Hacham, N. 418 Hadas, M. 89, 417, 420-423

Hagedorn, U. 208 Hamerton-Kelly, R. 290 Hamilton, V. P. 237 Hammerstaedt-Löhr, A. 120 Hanhart, R. 173, 393, 405, 416, 420 Hanhart, R. 234 Hansen, G. 163 Harl, M. 35, 51, 100, 173, 202, 242, 246, 257, 258, 283, 285, 287, 301, 317, 341, 357, 364, 367, 397, 399, 443 Harlé, P. 282, 299, 363 Hartmann, Th. A. G. 235, 236 Hastings, J. 21 Harrington, D. J. 417 Harris, J. R. 20, 418 Harvey, W. W. 219 Harviainen, T. 173, 180 Hatch, E. 18, 273, 280, 431, 433, 435 Hauck, A. 18 Haug, D. 114 Hauser, B. K. 452 Hauspie, K. 117, 136, 178, 301, 302, 454-456, 459 Heater Jr., H. 206 Heinemann, I. 389 Heinen, H. 305 Heininger, B. 424, 425 Heinrici, C. F. G. 468 Helbing, R. 24, 25, 55, 63, 90, 188, 224, 366, 397, 398, 451, 455, 457, 460 Helck, W. 94 Helleman, W. E. 422 Hemer, C. J. 100 Hengel, M. 231, 232, 261, 471, 471 van Henten, J. W. 421, 424 den Hertog, C. G. 316, 325 Herzog, J. J. 17 Heubeck, A. 124 Hezser, C. 261 Hiebert, R. J. V. 143, 213, 311, 312, 314, 364 Hill, D. 28 Hoffmann, L. 254 Hoffmann, O. 426

493

Indexes Hölbi, G. 337 Holton, D. W. 83 Holton, D. W. 88 Honigman, S. 120, 202 Horbury, W. 246 Horrocks, G. C. 26, 71, 75, 104, 109, 111-113, 115, 116, 432 Horsley, G. H. R. 34, 100, 103, 247, 259, 266, 281, 304, 359, 369, 471 Van der Horst, P. W. 262 Houtman, C. 196 Howard, W. F. 457-459 Hübner, H. 388 Humbert, J. 117 Hunkin, J. W. 462 Huss, W. 235 Husson, G. 305 Hyvärinen, K. 212 Irigoin, J. 372 Irmscher, J. 358 Jackendoff, R. 143 Jacquinod, B. 112 James, J. C. 22 Janse, M. 111, 263 Janssens, G. 131, 179 Jarick, J. 215 Jeffreys, E. 82 Jellicoe, S. 34, 167, 451 Jenny, L. 448 Jeremias, J. 464-466 Jobes, K. H. 32, 159, 173, 174, 281, 431 Johannessohn, M. 117 Joly, R. 107 Jones, S. 196, 271, 319 de Jonge, C. C. 146 Joosten, J. 31, 33, 44, 54, 57, 100, 110, 113, 155, 173, 174, 177, 244, 245, 250, 284, 286, 287, 296, 298, 299, 302, 316, 361, 363, 364, 366, 393, 395, 396, 398, 402, 405, 444, 458 Jördens, A. 421 Joüon, P. 130, 131, 135, 136 Jüngling, H. W. 50

494

Kabergs, V. 199 Kaczmarek, S. 297 Karrer, M. 31, 40, 41, 52, 100, 120, 121, 178, 196, 226, 231, 234, 245, 279, 302, 305, 318, 372 Katz, P. 184, 415 Kaupel, H. 360 Kautzsch, E. 416, 417 Kazhdan, A. P. 82 Kedar-Kopfstein, B. 172, 175, 176 Kennedy, H. A. A. 22, 23, 431, 436, 442, 443, 446, 447 Kenyon, F. G. 93 Kepper, M. 57, 388 Kiessling, E. 303 Kijlunen, J. 454 Kim, H. 317, 318 Kim, L. 80, 302 Kittel, G. 16, 196, 272, 442 Kircher, C. 272 Klauck, H. J. 66, 421, 424, 467 Klingenschmitt, G. 72 Knibb, M. 168 Knobloch, F. 173 Knoppers, G. N. 247 Knöppler, T. 66 Knuf, H. 235 Koch, D. A. 446 Kockelmann, H. 245 Koehler, L. 197 Koenig, Y. 242, 284, 289 Koetschau, P. 369 Kolarcik, M. 390 van der Kooij, A. 247, 255, 295 Kopidakis, M. Z. 418, 419 Kraft, R. A. 300 Kramer, B. 418 Kraus, Th. J. 59, 451, 455-458, 471 Kraus, W. 31, 40, 41, 52, 100, 120, 121, 178, 196, 226, 231, 234, 245, 279, 302, 305, 318, 329, 372 Krause, M. S. 360 Kreuzer, S. 40, 52, 114, 201, 295, 464

Kristeva, J. 368 Kugelmeier, C. 302 Kurth, D. 241 Lademann, W. 90 Laminger-Pascher, G. 452 Lampe, G. W. H. 272 Landfester, M. 367, 454 de Lange, N. 26, 263, 264 Larcher, C. 387, 390 Lausberg, H. 424, 425 Lauxtermann, M. D. 82 Law, T. M. 115 Lebram, J. C. H. 421, 423, 424, 427 Leclant, J. 244 Ledogar, R. J. 402 Lee, J. A. L. 33, 59, 100, 101, 105, 106, 185, 271, 300, 302, 306-315, 317, 353, 359, 367, 372, 377, 378 Lefebvre, P. 295 Lefort, L. Th. 22, 184, 231 Léonas, A. 67, 282, 358, 360, 361, 364, 369-371, 387, 388 Lemmelijn, B. 48, 49, 54, 139, 140, 163, 174, 176 Lenger, M. T. 337 Levi, P. 370 Liddell, H. G. 196, 271, 319 Lincoln, A. T. 470 Lindez, J. Vilchez 393 Lisowsky, G. 131, 179 Livingstone, E. A. 316 Ljungvik, H. 359 Llewelyn, S. R. 304 Lohfink, N. 466 Lohmeyer, E. 471 Lohse, E. 471 Lopez-Eire, A. 259 Loprieno, A. 237 Loubet, M. 173 van der Louw, T. 153, 167 Lüddeckens, E. 237 Lüderitz, G. 461, 462 Lundbom, J. R. 381, 382, 384 Lust, J. 110, 172, 174, 175, 177, 178, 277, 317-320 Lutz, H. L. F. 237 Luraghi, S. 116, 118 3. Modern authors

Indexes Lyonnet, S. 335 Lyons, J. 148 Mackridge, P. A. 87, 88 Magnien, V. 115 Mahaffy, J. P. 99 Meimaris, Y. E. 263 Malingrey, A. M. 422 Mandilaras, B. G. 233, 335, 353 Mango, C. 82 Marcos, N. Fernández 35, 173, 176, 182, 186, 187, 257, 258, 278, 280, 295, 306, 360, 393 Marquis, G. 188, 189 Martin, R. A. 31, 185, 224 Marx, A. 89 May, H. G. 416 Mayor, J. B. 470 Mayser, E. 90-93, 104, 112-117, 130, 452 Masson, É. 20, 173, 176, 179, 182 Mauranen, A. 163 Mazzinghi, L. 388, 390-392 McElduff, S. 146, 147 McKay, K. L. 452 McLay, R. T. 152, 153, 213 McLean, B. H. 102, 104, 105 McLean, P. D. 217 Van der Meer, M. N. 100, 101, 170 Meisterhans, K. 90, 453 Melazzo, L. 259 Mélèze Modrzejewski, J. 352, 416, 418, 419 Menu, B. 100, 304 Merati, G. 143 Meredith, A. 357 Metzger, B. M. 416, 419 Meyer, G. 17 Meynadier, B. 329 Michaelis, W. 369 Migne, J. P. 90 Milgrom, J. 333 Miller, C. L. 135 Moffat, J. 408 Möhn, D. 254 3. Modern authors

le Moigne, P. 31 Molinié, G. 368 Mondésert, C. 357 Monsengwo Pasinya, L. 282, 300, 350, 351 Montevecchi, O. 100, 304, 359 Moore, C. A. 393 Morenz, S. 232, 240, 241, 244 Moreschini, C. 367 Morpurgo, A. Davies 72 Morris, I. 75 Morrish, G. 273 Motzo, B. R. 416 Moule, F. D. 34 Moulton, J. H. 20, 21, 22, 26, 28, 99, 280, 295, 306, 361, 445, 446, 449, 452, 457, 459, 460, 471 Muchiki, Y. 235, 238, 241, 242 Mugler, C. 409 Müller, M. 464 Munnich, O. 35, 51, 53, 100, 173, 202, 206, 210, 212, 246, 257, 258, 317, 367, 397, 399, 443 Muraoka, T. 28, 107, 135, 137, 170, 250, 275, 278, 298, 301, 302, 337, 366, 400, 452 Mussies, G. 397 Nachmanson, E. 453 Nägeli, Th. 446 Neirynck, F. 462 Nesselrath, H. G. 26 Neumann, G. 160 Neuschäfer, B. 357 Neyt, F. 365 Nicklas, T. 66 Nida, E. A. 185 Norden, E. 41, 58, 59, 67, 372, 420-424, 465, 468 Norin, S. 180 Oates, J. F. 103 O’Connor, M. 101 O’Hare, D. M. 177 Olivieri, A. 371

Olley, J. W. 34, 300 Olofsson, S. 33, 152, 159, 185, 274, 301, 338 Olsson, B. 31, 32 Orlinsky, H. M. 185 Osing, J. 94 Ottley, R. R. 19, 22, 184, 437 Otto, E. 235 Ouspenskij, B. B. 368 Palmer, L. R. 26, 452 Passoni Dell’Acqua, A. 231, 233, 235, 302, 305, 335, 337, 339, 350, 353, 359, 416, 418, 419 Paul, A. 418 Paul, H. 22 Paul, S. M. 42, 100, 300, 353 Peake, A. S. 21 Peek, W. 370 Pelletier, A. 173, 174, 177, 179, 249, 250, 284, 353 Perkins, L. 315 Pernot, H. 25 Peters, M. K. H. 91, 101, 111, 117, 163, 193, 202, 312, 316, 384, 454 Pfeiffer, St. 234-236 Philonenko, M. 244 Pierri, R. 36 Pietersma, A. 155-158, 168, 170, 196, 210-212, 216, 258, 294, 296, 298, 418, 437 Piquer Otero, A. 191 Pirot, L. 22 Pope, M. H. 200 Porten, B. 238 Porter, S. E. 15-17, 19, 21, 2628, 32-34, 59, 360, 377, 390, 449, 452, 471 Porúbčan, Š. 336 Powell, M. A. 314 Pralon, D. 299 Preisigke, F. 103, 241, 242, 303, 426 Priebatsch, H. Y. 404 Psichari, J. 25 Pym, A. 140

495

Indexes Quack, J. F. 245 Rabin, C. 34, 186, 260, 288 Rackham, H. 335 Radermacher, L. 452, 459 Rahlfs, A. 41, 91, 143, 147 Rainey, A. F. 173, 180, 239 Rajak, T. 246, 261, 264, 297, 416, 422 Rapallo, U. 364 Rasmussen, J. E. 72 Redpath, H.A. 184 Reese, J. R. 376, 384, 387-390 Rehkopf, F. 277, 426, 452, 460 Reider, J. 373, 387 Reiser, M. 460, 461 Rendtorff, R. 334 Renehan, R. 271, 422, 423 Reuss, E. 17 Rey, J. S. 375 Richardson, P. 351 Riddle, D. W. 185 Rife, J. M. 185 Riffaterre, M. 368 Risberg, B. 408 Robertson, A. T. 452 Ros, J. 182 Römer, M. 238 Ronchi, G. 304 Rösel, M. 169, 244, 332 Rosen, H. B. 182, 364 Rossi, M. A. 371 Rowe, G. O. 390 Rouffiac, M. J. 359 Ruckstuhl, E. 467 Rudhardt, J. 330 Rüger, H. P. 461 Ruijgh, C. J. 112 Ruppert, L. 388 Rupprecht, H. A. 303 Rüsen-Weinhold, U. 459 Russel, D. A. 367 Rydbeck, L. 183, 359, 460, 462 Sacchi, P. 416 Safrai, S. 34, 260, 397 Salvesen, A. 150, 172

496

Samuel, D. H. 22 dos Santos, E. C. 275 Sarafis, M. 87 de Saussure, F. 33, 445 Säve-Söderberg, T. 239 Scarpat, G. 386-388 Schauber, E. 143, 144 Schenker, A. 287, 336, 338, 463 Schleiermacher, F. 49, 139 Schleusner, J. F. 277 Schmeller, T. 468 Schmid, W. 425 Schmidt, D. D. 471 Schmidt, K. L. 251 Schmitz, B. 393, 404, 405 Schneider, N. 469 Schröder, C. 279 Schulmann, A. R. 236, 237 Schürrer, E. 416, 417 Schwartz, D. R. 407, 408 Schwartz, J. 357 Schwyzer, E. 90, 92, 426, 427, 453, 454 Scott, R. 196, 271, 319 Scott, R. B. Y. 471 Seebass, H. 237 Seele, A. 43 Seeligmann, I. L. 159, 234, 246, 316, 325, 328 Segert, S. 161 di Segni, L. 264 Sellin, G. 470 Seppänen, C. 116 Sevenster, J. N. 261 Seybold, K. 466 Shaw, F. 409, 414 Shenkel, J. D. 170 Siegert, F. 40, 45, 49, 50, 58, 59, 67, 232, 301, 331, 333, 372, 451, 456-459 Silva, M. 32, 281, 433, 436, 442, 444, 449 Simotas, P. N. 175 Smith, J. 251 Smith, R. W. 367, 372 Snell-Hornby, M. 139 Soisalon-Soininen, I. 29, 30, 48, 51, 55, 112, 113, 115-117, 145,

151, 161, 166, 171, 188-191, 224, 225, 364, 389, 454 Sollamo, R. 24, 29, 30, 48, 100, 111, 116, 161, 163-168, 170, 360, 388, 451 Sophocles, E. A. 18, 272 Sørensen, S. L. 418 Sowinski, B. 368 Sparks, H. F. D. 438, 449 Speiser, E. A. 201 Spicq, C. 369, 470 Spieckermann, H. 421 Spottorno, M. V. 174, 175, 177 Stanley, Ch. D. 446 Sterling, G. J. 260 Stipp, H. J. 373 Stoop-van Paridon, P. W. T. 197 Störig, H. J. 139 Stowers, S. K. 468 Strachan, L. R. M. 99 Stuiber, A. 232 Sturz, F. W. 16, 98 Svensson, A. 458 Swete, H. B. 21, 23, 24, 99, 286, 291, 309, 311-314, 386, 431, 438, 442, 445, 449 Swinn, S. P. 301 Tabachowits, D. 30, 431, 452, 460 Talshir, Z. 175 Taylor, B. A. 30, 31, 301, 361 Tcherikover, V. A. 246 Teodorsson, S. T. 22 Thackeray, H. St. J. 24, 25, 42, 45, 48, 50, 53, 55, 58, 62, 63, 90, 143, 170, 172, 175, 177, 182, 184-188, 191, 204, 280, 284, 286, 287, 290, 306, 314-316, 360, 373, 391, 399, 400, 419, 451 Thayer, J. H. 440 Thibaut, A. 301 Thiersch, H. G. J. 17, 98 Thissen, H. J. 235, 236 Thompson, S. 34, 280 Threatte, L. 91-93 Thumb, A. 23, 183, 280, 306, 359, 432 3. Modern authors

Indexes Thyen, H. 468 Tilly, M. 242 Titrud, K. 458 Tjen, A. 114, 115 Todorov, T. 370 Tomson, P. J. 302 Torrey, C. C. 183, 416, 417, 471 Toury, G. 156, 163 Tov, E. 17, 28, 48, 51, 53, 138, 150-152, 163, 170, 171, 172, 174178, 182, 185, 186, 188-190, 193-200, 248, 271, 274, 283, 288-290, 293, 312-314, 318, 319, 328, 360 Trebolle Barrera, J. 35 Trenkner, S. 112, 373 Trilling, W. 470 Trommius, A. 272 Tromp, J. 418 Troxel, R. L. 297 Tur-Sinai, H. 200 Turner, N. 28, 29, 104, 105, 184, 280, 455, 456, 460, 462464, 466-469 Ueding, G. 58 Usener, K. 45, 120, 122-125 Vaccari, A. 173, 180 Vahrenhorst, M. 121, 329, 330, 332, 333 Vanhoozer, K. J. 147 Vattioni, F. 305, 359 Ventris, M. 72

Verbeke, E. 185 Vergote, J. A. L. M. 22, 94, 184, 231, 237, 242, 281, 363 Vermes, G. 416 Verreth, H. 303 Vervenne, M. 317 Vianès, L. 363 Viteau, J. 18 Vittmann, G. 238 Vleeming, S. P. 231 Voelz, J. W. 16 Voitila, A. 30, 110, 111, 114, 115, 153, 173, 174, 223, 297 Vollmer, H. 448 Volokhine, Y. 237 Vulliaud, P. 358 Wade, M. L. 142, 152 Wagner, C. 39-42, 59 Wahlgren, S. 32 Walser, G. 31, 32, 185, 223-227, 246, 302 Walters, P. 182, 188, 249, 300, 415, 426, 451 Walzer, R. R. 367 de Waard, J. 290 Wasserstein, A. 316 Weber, R. 421-423 Weiss, J. 468 Weissert, D. 375 Welch, J. W. 372 Wellhausen, J. 20, 247 Westcott, B. F. 386 Westendorf, W. 238

Wevers, J. 106, 145, 146, 169, 306, 333 Whitmarsh, T. 80 Wieger, M. 404 Wifstrand, A. 31, 225, 389, 449 Wilke, C. G. 468 Will, E. 258 Von Wilamowitz-Moellendorff, U. 369, 372, 417, 420 Williams, F. 371 Williams, R. 357 Williams, T. 143 Williger, E. 330 Wilson, R. M. 184 Windisch, H. 467 Winer, G. B. 16, 20 Winston, D. 386, 389, 390 Wolscht, E. 421 Wolski, J. 100 Wong, S. 16 Woo, S. H. 113, 114 Wooden, R. G. 302 Woodhead, A. G. 104 Wright, A. G. 389 Wright, B. G. 42, 150-152, 174 Wutz, F. X. 173, 179 Young, F. M. 357 Youngblood, K. J. 214, 215 Zenger, E. 338, 393 Zerwick, B. M. 452 Ziegler, J. 100, 170, 208, 216, 316

4. General Index Acceptability 157 Adequacy 157 Aemulatio 41 Aeschylus 121, 122, 418 Agricultural terms 307 Alexander 77, 259, 306 Alexandria 16, 18, 40, 45, 89, 120, 234, 257, 281, 299, 306, 342, 386, 417, 435 Alien 347 4. General Index

Andros 370 Anthropomorphisms 333 Apis 241 Aquila 50, 53, 143, 147, 202, 226, 373 Aramaic 177, 284 Aramaisms 183, 249, 361 Aristophanes 98 Asianism 372, 417

Asterisked material 209 Asyndetic relative clause 134 Athens 342 Attic 22, 75, 76 Atticism 78, 79, 425 Bad Greek 360 Barbarity 357 Berossus 42, 373 Biblical Greek 280, 440

497

Indexes Bilingual papyri 297 Bilingualism 32, 34, 44, 79, 183, 184, 261, 432 Celsus 357 Chaldean Oracles 373 Church Fathers 272 Cicero 141 Coherence 154 Colloquial 22 Consistency 149, 151, 199 Contemporizing renderings 232, 234 Coptic 94, 184, 363 Creativity 197 Cultic terms 321, 329, 336, 401 Descriptive Translation Studies 156 Dialect 72, 73, 94, 247 Diaspora 246, 247, 250, 257 Diatribe 421 Difficult words 325 Diglossia 34, 183, 184, 261 Domesticating translation 139 Doric 314 Easy technique 50 Egypt 15, 231, 284, 288, 314, 363 Elephantine 250 Elevated prose 41, 58, 67 Epideictic 389, 423 Epiphanius 219 Esther, postscript 90 Euripides 122 Exegesis 328 Finnish School 29, 163 Foreignizing translation 139 Functional equivalence 203 Genitive absolute 62 Genre 64, 76 Glossaries 297 Grammar 24, 142 Grammaticality 150

498

Guessing 195 Hadrian of Antioch 90, 365 Hapax legomena 178, 193, 194, 197, 198, 388, 408 Hebraists 16, 182, 358 Hebraisms 24, 29, 55, 182, 224, 248, 360, 361, 387, 394, 400, 458 Hellenism 329 Hellenistic poetry 374 Hexapla 179, 220 Holy Ghost Greek 16 Homer 57, 75, 98, 113, 120, 122, 123, 296, 341, 343 Homophony 149, 282, 290, 375 Iamblichus 373 Idiomatic language 165, 188 Imitation 142 Impersonal expressions 136 Indo-European 71 Infinitive absolute 190, 360 Inscriptions 99, 101, 165, 261, 303, 452 Intercultural relations 45 Interference 156 Interlinear paradigm 156, 296, 437 Inter-Septuagintal borrowing 327 Intertextuality 448 Ionic 314 Irenaeus 219 Isis aretalogies 370, 372 Isomorphism 144, 203 Jerome 46, 220, 386 Jewish Greek 16, 26, 34, 100, 183, 247, 280, 281, 295 Josephus Flavius 42, 330 Kaige 170, 202, 257 Koine 16, 19, 23, 79, 94, 102, 165, 183, 259, 280, 432, 440 Kunstsprache 75 Lagos 235 Legal language

308, 322

Letter of Aristeas 255, 299 Literal and free translation 48, 54, 65, 139, 167 Loan translation 281, 363 Loanwords 72, 79, 94, 189, 241, 249, 314, 435 Local colour 240 Magical papyri 370 Manetho 42, 245, 330 Metre 391, 418 Midrashic 389 Military language 299 Minor Prophets Scroll 204, 211 Modern Greek 23, 91 Moses 235, 288 Multilingualism 15, 260 Mythology 124 Neologism 107, 218, 279, 282, 291, 319, 331, 332, 369, 370, 408, 409, 426 New Testament 15, 272 Onomatopoeia 122 Optative 56, 57, 113, 407 Παιδεία 40, 41, 45 Palestine 15, 204, 219, 257 Papyri 16, 17, 91, 96, 99, 105, 165, 183, 231, 303, 432, 452 Parallelismus membrorum 387, 463, 469 Parataxis 61 Paronomasia 63, 360, 371, 391, 414, 460 Participium conjunctum 62, 396 Periphrastic constructions 115 Philo 334 Philo of Byblos 373 Philosophical vocabulary 124, 422 Place names 237, 303 Plato 98, 125 Poetic vocabulary 121, 122 Polybius 98, 165, 408 Post-biblical Hebrew 287 4. General Index

Indexes Prepositions 116, 397-398 Prestige language 261 Proper names 235 Ptolemaic administration 233, 336 Purists 16, 182, 358 Rabbinic hermeneutics 52, 143 Recension 202 Religion 177, 234, 243, 319, 435 Revision 202 Rhetoric 142, 386, 390, 414, 417, 468 Rhodes 342 Roman Empire 78 Samaritan Pentateuch 132 Second column 179 Segmentation 51, 148, 151

4. General Index

Semantic borrowing 282 Semantic change 306 Semitic Greek 18 Semitisms 19, 281, 282, 286, 361, 432, 442 Septuagintism 361, 441 Sirach, Prologue 39 Sociolect 45, 246 Statistics 152, 164, 221 Stereotyped renderings 51, 186, 282, 289, 351 Style 154, 367 Symmachus 198 Synagogue Greek 186 Syntax 109, 162 Textual criticism 167, 168, 190 Theodotion 202

Theology 168, 339, 402, 435, 447 Transcriptions 172, 189 Translation 42 Translation Greek 223, 407 Translation technique 39, 110, 139, 146, 161, 183, 219 Transliteration 172, 282, 283 Verlegenheitsübersetzung 283 Vernacular 26, 102, 114, 415 Vocabulary 210, 223-225, 247, 252, 306, 317, 369, 387-388, 400-401, 405, 435-436 Voces biblicae 440 Voice 115 Word order

148, 151

499

Authors Hans Ausloos is a Professor at the Université catholique de Louvain and Chercheur qualifié F.R.S.-FNRS (Belgium). He is Director of the Research group “Septuagint Studies and Textual Criticism”, and Research Associate at the Department of Old Testament at the University of the Free State (Bloemfontein, South Africa). His research focusses on the so-called Deuteronom(ist)ic redaction in Genesis-Numbers and the translation technique of the Septuagint. Eberhard Bons studied theology, philosophy and Romance languages at the universities of Mainz, Tübingen, Rome (Gregorian University) and Frankfurt (Faculty of Sankt Georgen). In 1988 he obtained a PhD degree from the University of Mainz. In 1993, he also obtained a doctoral degree in theology from the Philosophisch-Theologische Hochschule St. Georgen, Frankfurt. He received his habilitation in 2000 from the University of Strasbourg (France) where he teaches as a Professor of Old Testament Exegesis. Major publications: Psalm 31 – Rettung als Paradigma. Eine leserorientierte Analyse, Frankfurt, 1994; Das Buch Hosea, Stuttgart 1996; Les Douze Prophètes, vol. 1: Osée, Paris 2002 (in collaboration with J. Joosten, St. Kessler, and Ph. Le Moigne); Textkritik und Textgeschichte. Studien zur Septuaginta und zum hebräischen Alten Testament, Tübingen 2014. Cameron Boyd-Taylor is a Research Associate in the John William Wevers Institute for Septuagint Studies, Trinity Western University. His publications include Reading Between the Lines: The Interlinear Paradigm for Septuagint Studies (Leuven 2011). He is Joint Editor-in-Chief of the Society of Biblical Literature Commentary on the Septuagint, and is currently writing a commentary on the book of Esther. Ralph Brucker, Studium der Evangelischen Theologie in Hamburg; 1996 Promotion; 1998–2000 Stipendiat der Deutschen Forschungsgemeinschaft; 2002–2009 Wissenschaftlicher Mitarbeiter bzw. Assistent für Neues Testament am Fachbereich Evangelische Theologie der Universität Hamburg; 2009–2010 Mitarbeiter im Redaktionsteam von „Septuaginta Deutsch: Erläuterungen und Kommentare“ am Theologischen Zentrum Wuppertal; 2010–2014 Lehrkraft für besondere Aufgaben (Altes und Neues Testament) am Institut für Evangelische Theologie der Justus-Liebig-Universität Gießen; seit 2014 Wissenschaftlicher Mitarbeiter (Editor) an der Schleiermacher-Forschungsstelle Kiel; daneben seit 2011 Mitarbeiter im Redaktionsteam des “Historical and Theological Lexicon of the Septuagint” (HTLS), Université de Strasbourg. Carlo Consani is professor of Linguistics and Sociolinguistics at the “G. D’Annunzio” University of Chieti and Pescara (Italy). His research interests are: the writing systems of the Aegean area in the second and first millennium BC; the geographical, historical and sociolinguistic variation of ancient Greek, particularly in the Hellenistic and Roman periods; the problems of language contact with reference to corpus-Sprachen. He has published several studies in various scientific journals and collective works; he is author of 501

Authors

Dialektos. Contributo alla storia del concetto di dialetto (1991) and of Sillabe e sillabari fra competenza fonologica e pratica scrittoria (2003). He has just edited the volume Contatto interlinguistico fra presente e passato (2015). Jennifer Dines has degrees in Classics from the University of Cambridge and in Theology from the University of London. She did a doctoral dissertation on the Septuagint of Amos (PhD London, 1992) and taught biblical studies at Heythrop College, University of London until retirement in 2001. She published The Septuagint (2004) and is currently working with Professor Eberhard Bons on the Amos volume of the Bible d’Alexandrie series. Cécile Dogniez, chargée de recherches au CNRS (UMR 8167 Orient V. Hiebert is the Professor of Old Testament in the Graduate School of Theological Studies at Trinity Western University (Langley, British Columbia, Canada) and the Director of the John William Wevers Institute for Septuagint Studies. He also serves as the Vice President of the International Organization for Septuagint and Cognate Studies and as Co-editor of the Society of Biblical Literature Commentary on the Septuagint series. Geoffrey Horrocks trained both as a classicist and a theoretical linguist, but his primary research focus has always been the history and structure of Greek, ancient, medieval and modern. He is the author of many articles in this field, and of a now standard history of Greek (Greek: a History of the Language and its Speakers, 2nd edition, WileyBlackwell, 2010). After periods as a Research Fellow of Downing College, Cambridge, and University Lecturer at the School of Oriental and African Studies of the University of London, he has, since 1983, worked in the Classics Faculty of the University of Cambridge, first as a University Lecturer and, from 1997, as Professor of Comparative Philology. He is a Fellow of St John's College, Cambridge and holds an honorary doctorate from the National and Capodistrian University of Athens. Jan Joosten, born 1959 in Ekeren, Belgium, studied theology in Brussels and Princeton, and Semitic languages in Jerusalem. He earned a PhD in Semitic languages at the Hebrew University in 1989, a ThD at the Protestant Faculty in Brussels in 1994, and a HDR (Habilitation à diriger des recherches) in Strasbourg in 1994. For twenty years he taught at the Faculty of Protestant Theology of the University of Strasbourg. In 2014 he was appointed Regius Professor of Hebrew in the University of Oxford, and Student of Christ Church. He is Editor-in-chief of Vetus Testamentum since 2010, president of the International Organization for Septuagint and Cognate Studies since 2012, and editor, together with Eberhard Bons, of the forthcoming Historical and Theological Lexicon of the Septuagint. He is married with four children. Thomas J. Kraus, geboren 1965, ist Lehrbeauftragter an der Universität Zürich. Seine Dissertation wurde 2001 von Mohr Siebeck, Tübingen, unter dem Titel Sprache, Stil und historischer Ort des Zweiten Petrusbriefes veröffentlicht. Er ist Research associate/ fellow am Department of New Testament Studies der Universität Prätoria und am Department of Greek, Latin and Classical Studies der University of the Free State, Bloemfontein, Südafrika. Seine hauptsächlichen Forschungsinteressen, wie zahlreiche Publika502

Authors

tionen, selbst verantwortete und herausgegebene Bücher zeigen, liegen im Bereich der (frühchristlichen) Papyrologie, der christlichen Apokryphen, der Sprache und der Textgeschichte der Septuaginta und des Neuen Testaments, der Magie sowie Buchproduktion und des (An-)Alphabetismus in der (Spät-)Antike. Christoph Kugelmeier (geb. 1965), 1987–1992 Studium der Griechischen und Lateinischen Philologie an der Universität Köln, 1995 Promotion über das Thema Reflexe früher und zeitgenössischer Lyrik in der Alten attischen Komödie, 1995–2002 wissenschaftlicher Mitarbeiter und Assistent in Dresden, Potsdam und Saarbrücken, 2002 Habilitation an der Universität des Saarlandes mit einer Untersuchung über Die innere Vergegenwärtigung des Bühnenspiels in Senecas Tragödien, danach Hochschuldozent und seit 2010 außerplanmäßiger Professor an der Universität des Saarlandes. Seit 2008 Beteiligung am von Eberhard Bons und Jan Joosten herausgegebenen Historical and Theological Lexicon of the Septuagint (HTLS). John A. L. Lee is an Honorary Fellow of Macquarie University, Sydney, having retired from the University of Sydney in 2001, where he lectured in Ancient Greek for 28 years. His research interest is in Greek language of all periods, especially Koine Greek, the Septuagint, and Greek lexicography. His doctoral dissertation (Cambridge, 1970) was on the language of the LXX-Pentateuch, and his Grinfield Lectures (Oxford, 2011–2012) carried the same subject further. His major works are A Lexical Study of the Septuagint Version of the Pentateuch (1983) and A History of New Testament Lexicography (2003). Bénédicte Lemmelijn is Professor at the Katholieke universiteit Leuven (Belgium), where she is Director of the Centre for Septuagint Studies and Textual Criticism. Moreover, she is Research Associate at the Department of Old Testament at the University of the Free State (Bloemfontein, South Africa). Her research focusses on the Plague narrative in the Book of Exodus and on the translation technique of the Septuagint. Alexis Léonas, born 1968 in Moscow, and now based in Budapest, Hungary, did his doctoral work on the language of the Septuagint at the Sorbonne (1996–2002). He is a historian and writer who has long been associated with the Bible d'Alexandrie project. He has published two books and a range of articles focusing on the Septuagint. Luca Mazzinghi, parish priest in the Catholic church (diocesis of Florence, Italy), teaches Old Testament exegesis as Ordinary professor in the Theological University of Central Italy and, as invited professor, at the Pontifical Biblical Institute in Rome. He is president of the Associazione Biblica Italiana. He has published many books and studies especially on the wisdom books, including: Storia di Israele. Dalle origini al periodo romano, EDB, Bologna 2007; Ho cercato e ho esplorato: studi sul Qohelet, EDB, Bologna 2009 (2d ed.); Il Pentateuco sapienziale: Proverbi, Giobbe, Qohelet, Siracide, Sapienza, EDB, Bologna 2012; with P. Sacchi, La Bibbia dei Settanta. 1. Il Pentateuco, Morcelliana, Brescia 2013. Takamitsu Muraoka, born in Hiroshima in 1938. Educated at Tokyo Kyoiku University for BA in English (1960), MA in linguistics (1962), The Hebrew University of Jerusalem 503

Authors

for Ph. D. with a dissertation entitled Emphasis in Biblical Hebrew, supervised by Prof. Ch. Rabin (1970). Employed: lecturer in Semitic languages at Manchester University, England (1970-80), professor of Middle Eastern Studies at Melbourne University, Australia (1980-91), Professor of Hebrew, Israelite antiquities and Ugaritic at Leiden University, The Netherlands (1991-2003). Research interests: biblical languages and ancient versions, especially the Septuagint. Gerard Mussies was born in The Hague (Holland) in 1934. He studied Classical and Semitic languages, Ancient-Egyptian and Spanish. He was lecturer in the University of Utrecht in Hellenistic Greek, especially that of the NT. His publications include: The Morphology of Koine Greek as Used in the Apocalypse of St. John (SNT 27), and “Greek in Palestine and the Diaspora” (in CRINT 2). He did research in the project of the Corpus Hellenisticum Novi Testamenti, notably publishing Dio Chrysostom and the New Testament (SCHNT 2). Wolfgang Orth (geb. 1944), ist Professor (i. R.) für Alte Geschichte an der Bergischen Universität Wuppertal. Studium der Klassischen Philologie und der Geschichte in München und Tübingen. Promotion in München 1970; Habilitation in Alter Geschichte in München 1976. 1977–1988 Professor für Alte Geschichte in Münster; seit 1988 in Wuppertal. Mitglied des DFG-Sonderforschungsbereichs »Tübinger Atlas des Vorderen Orients« 1975–1990. Mitherausgeber »Septuaginta Deutsch« 2003–2011. Forschungsschwerpunkte: Hellenismus und römische Kaiserzeit; historische Geographie; Nachleben der Antike. Anna Passoni dell’Acqua teaches Biblical Philology at the Catholic University of Milan. She works on the philological and literary study of the Biblical Text and the apocryphal literature of Old Testament. The main field of her researches is the Septuagint version. She specializes in lexical surveys on Biblical Greek in comparison with Hellenistic Greek, and in particular with the language of Egyptian Greek papyri. She is a member of the Editorial Board of the Historical and Theological Lexicon of the Septuagint, edited by E. Bons / J. Joosten and printed by Mohr Siebeck Verlag, Tübingen. Stefan Pfeiffer (Jahrgang 1974) studierte an der Universität Trier von 1995 bis 2002 Geschichte, Klassische Archäologie, Ägyptologie, Papyrologie und Katholische Theologie (Magister Artium/1. Staatsexamen). 2004 wurde er mit einer Arbeit über das Dekret von Kanopos in Trier promoviert. Mit einer Habilitation über Kaiserkult und Kaiserverehrung in Ägypten erhielt er 2007 die venia legendi für das Fach Alte Geschichte. Von 2008 bis 2009 war er Wissenschaftlicher Mitarbeiter an der Universität Mannheim, von 2009 bis 2010 Akademischer Rat an der Universität Münster, von 2010 bis 2013 Professor für „Antike und Europa“ an der Technischen Universität Chemnitz. Seit 2013 ist er Professor für Alte Geschichte an der Martin-Luther-Universität Halle-Wittenberg. Seine Forschungsschwerpunkte liegen in der Religionsgeschichte des griechisch-römischen Ägypten, der Geschichte des antiken Judentums und der frühen Kaiserzeit. Gerade erschienen ist sein Buch Griechische und lateinische Inschriften zum Ptolemäerreich und zur römischen Provinz Aegyptus.

504

Authors

Andrés Piquer Otero is Professor of Hebrew and Aramaic Studies in Universidad Complutense de Madrid, Spain. His research and publications focus on a historical approach to the historical books of the Bible (Samuel-Kings), with special interest in textual criticism of the Septuagint and its secondary versions, historical linguistics and a comparative study of Biblical Hebrew in its Northwest Semitic context. Stanley E. Porter is President and Dean, and Professor of New Testament, at McMaster Divinity College, Hamilton, Ontario, Canada. He is the author of over twenty volumes and the editor of over ninety others. He has written widely on various topics in biblical studies, especially in the area of Greek language and linguistics. Recent publications include Constantine Tischendorf: The Life and Work of a 19th Century Bible Hunter (2015) and Linguistic Analysis of the Greek New Testament (2015). Peter Prestel, geb. 1952 im Südschwarzwald, Abitur in Lörrach, Studium der Klassischen Philologie in Basel, Heidelberg und Rom. Staatsexamina und Referendariat in Heidelberg, ebendort Promotion mit der Arbeit Die Rezeption der ciceronischen Rhetorik durch Augustinus in ‚De doctrina Christiana‘. Ab 1985 Dozent für Klassische Philologie an der Kirchlichen Hochschule Bethel in Bielefeld, seit 2009 als Koordinator für Griechisch und Latein am Fachsprachenzentrum der Universität Bielefeld. Lehrbuchautor, Publikationen u.a. zu Flavius Josephus und zur Septuaginta, dort Mitarbeit an den Projekten LXX.Deutsch (Genesis) und Lexham Greek-English Interlinear Septuagint (Hohelied). Frank Shaw is an independent scholar in the USA. He has taught classics, history, and religious studies at the University of North Dakota, Xavier University, the University of Dayton, Earlham College, Wright State University, and Ashland University. His monograph The Earliest Non-mystical Jewish Use of Ιαω was published in 2014 (Peeters). Moisés Silva (PhD, University of Manchester) has taught Biblical Studies at Westmont College, Westminster Theological Seminary, and Gordon-Conwell Theological Seminary. He is the author or coauthor of ten books. In recent years he has worked as an independent scholar on various projects, including the revision of the Zondervan Encyclopedia of the Bible and of the New International Dictionary of New Testament Theology and Exegesis. Raija Sollamo, Doctor of Theology 1980, University of Helsinki. Professor of Biblical Languages 1998-2007 University of Helsinki. Vice-Rector of the University of Helsinki 1998-2003. Dean of the Faculty of Theology at the University of Helsinki 1992-1998. President of the International Organization for the Study of the Old Testament (IOSOT) 2007-2010. Emanuel Tov is J.L. Magnes Professor emeritus of Biblical Studies at the Hebrew University, Jerusalem. His research focuses on the textual criticism of the Hebrew and Greek Bible, and he has written handbooks in both areas. He is the past editor-in-chief of the International team publishing the Dead Sea Scrolls.

505

Authors

Knut Usener (Jahrgang 1959) ist Altphilologe und lehrt als außerplanmäßiger Professor an der Kirchlichen Hochschule Wuppertal-Bethel. Sein spezielles Interesse gilt der antiken Literatur- und Kulturwissenschaft. Insbesondere befasst er sich mit der frühgriechischen Dichtung, dem Hellenismus und der Septuaginta. Hierbei sind die besonderen Formen der Rezeption, der Aneignung und der Transformation literarischer Motive und Formen im Medium der Übersetzungsliteratur zu einem Schwerpunkt seiner Forschung geworden. Martin Vahrenhorst, geboren 1967, Studium der evangelischen Theologie in Wuppertal, Göttingen, Jerusalem und Bochum. Promotion 2000 (Neues Testament), Habilitation 2007 (Neues Testament).2007-2015 Studienleiter von „Studium in Israel“ in Jerusalem. Anssi Voitila, Doctor of Theology 2001 (University of Helsinki). Adjunct professor at the University of Eastern Finland. A Project participant: Changes in Sacred Texts and Traditions (CSTT) (Center of Excellence in Research, Academy of Finland, 2014-2019). Georg Walser, born 1964; 1994 BA in Classical Greek at University of Gothenburg; 2001 PhD in Classical Greek at Lund University; 2008 ThB at University of Gothenburg; 2008 ThM at University of Gothenburg; 2013 PhD in Old Testament Exegesis at University of Gothenburg; 2013 PhD in New Testament Exegesis at University of Leicester. Madeleine Wieger, née en 1977 à Saverne, France. Etudes de lettres classiques et de théologie protestante à Paris, Strasbourg et Tübingen. Depuis 2007 Maître de conférences en Philologie biblique à la Faculté de théologie protestante de Strasbourg.

506