The Legacy of Soisalon-Soininen: Towards a Syntax of Septuagint Greek [1 ed.] 9783666564871, 9783525564875

Citation preview

De Septuaginta Investigationes (DSI) Edited by Anneli Aejmelaeus, Kristin De Troyer, Wolfgang Kraus, Emanuel Tov In Co-operation with Kai Brodersen (Erfurt, Germany), Cécile Dogniez (Paris, France), Peter Gentry (Louisville, USA), Anna Kharanauli (Tbilisi, Georgia), Armin Lange (Wien, Austria), Alison Salvesen (Oxford, UK), David Andrew Teeter (Cambridge, USA), Julio Trebolle (Madrid, Spain), Florian Wilk (Göttingen, Germany) Volume 13

Tuukka Kauhanen / Hanna Vanonen (eds.)

The Legacy of Soisalon-Soininen Towards a Syntax of Septuagint Greek

Vandenhoeck & Ruprecht

Bibliographic information published by the Deutsche Nationalbibliothek: The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data available online: https://dnb.de. © 2020, Vandenhoeck & Ruprecht GmbH & Co. KG, Theaterstraße 13, D-37073 Göttingen All rights reserved. No part of this work may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without prior written permission from the publisher. Typesetting: NEUNPLUS1, Berlin Vandenhoeck & Ruprecht Verlage | www.vandenhoeck-ruprecht-verlage.com ISSN 2197-0912 ISBN 978-3-666-56487-1

Table of Contents Introduction�������������������������������������������������������������������������������������������������������������������7 John A.L. Lee Back to the Question of Greek Idiom�����������������������������������������������������������������������13 Seppo Sipilä Soisalon-Soininen Meets Grice: The Cooperative Principle and the Septuagint Syntax�������������������������������������������������������������������������������������������������27 Raija Sollamo The Usage of the Article with Nouns Defined by a Nominal or Pronominal Genitive in LXX Genesis����������������������������������������������������������������������43 Jan Joosten Grammar and Style in the Septuagint: On Some Uses of Preverbs����������������������55 Theo A.W. van der Louw The Dynamics of Segmentation in the Greek Pentateuch�������������������������������������65 Anssi Voitila Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch��������������������������������������������������������������������������������������������������������81 Patrick Pouchelle Did the Translators of the Pentateuch Differentiate the Piel from the Qal?�������������������������������������������������������������������������������������������������97 Philippe Le Moigne Une tournure syntaxique fréquente en Ésaïe-LXX: Substantif abstrait + λήμψεται + complément d’objet direct�������������������������������119 Miika Tucker The Translations of Infinitives in Septuagint Jeremiah���������������������������������������139 Anneli Aejmelaeus Translation Technique and the Recensions: A Late Review of Ilmari Soisalon-Soininen’s Doctoral Thesis on the Text-Forms of Judges��������159

6

Table of Contents

José Manuel Cañas Reíllo Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges������������������������������������������������������������������������������������������������������������175 William A. Ross Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt�����������������������189 Srećko Koralija παράδεισος as/and κῆπος�����������������������������������������������������������������������������������������207 Ilmari Soisalon-Soininen The Use of the Genitive Absolute in the Septuagint���������������������������������������������223 Ilmari Soisalon-Soininen A Palette of Renderings for the Hebrew Construct State Relation in the Greek Pentateuch�������������������������������������������������������������������������������������������229 Ilmari Soisalon-Soininen The Alleged Interchangeability of ἐν and εἰς in the Septuagint��������������������������241 Ilmari Soisalon-Soininen Segmentation in the Greek Pentateuch������������������������������������������������������������������251 Ilmari Soisalon-Soininen The Rendering of the Hebrew Personal Pronoun as Subject in the Greek Pentateuch�������������������������������������������������������������������������������������������263 Ilmari Soisalon-Soininen Back to the Question of Hebraisms������������������������������������������������������������������������283 Bibliography���������������������������������������������������������������������������������������������������������������301 Indexes������������������������������������������������������������������������������������������������������������������������323

Introduction Ilmari Soisalon-Soininen was born shortly before Finland gained its independence in 1917. Already in school, he demonstrated an exceptional talent for languages in teaching himself Arabic and Hungarian from German textbooks. His competence in Hungarian led him to a student exchange in Hungary that resulted in life-long friendships. As a young student of theology, he served on the front in the Winter War (1939–1940) and later as a military chaplain. After the Second World War, he worked as a hospital chaplain and family counsellor as well as a secondary school teacher, at the same time doing research for his doctoral dissertation on the Septuagint of the Book of Judges. Often, he did his research at night, as the days were devoted to pastoral care and teaching. Soisalon-Soininen received his doctorate in 1951, becoming a Senior Research Fellow in 1962 and an Assistant Professor in 1964. He was appointed to full professorship in Old Testament Exegesis in 1967 and held that position until his retirement in 1984. He was a member of the committee for the most recent Finnish Bible translation (published 1992). When Soisalon-Soininen began his doctoral research, Septuagint studies were more or less unknown in Finland. His important innovation was to concentrate on the study of translation technique, particularly on the role of syntax in describing the nature of different translations. In his Die Infinitive in der Septuaginta, often considered his most influential work, he proceeded from the observation that when it comes to morphology, the translators follow the conventions of the target language; but when it comes to syntax, the translation tends to reflect structures of the source language. From this it follows that it is impossible to describe the syntax of the Septuagint without researching the translation technique employed by different translators, and in this endeavour, the characteristics of both the Hebrew and Greek languages need to be taken into consideration. These insights led Soisalon-Soininen to launch a long-term research project to describe the syntax and translation technique of the individual translators of the Septuagint. Soisalon-Soininen and his students, Raija Sollamo and Anneli ­Aejmelaeus, soon became a familiar sight at international conferences. In addition, Soisalon-Soininen became a prominent figure in the International Organization for Septuagint and Cognate Studies (IOSCS). He continued doing research during his retirement until his death in 2002. To honour the centennial of Soisalon-Soininen’s birth, June 2017, a symposium was held at the University of Helsinki. The thirteen articles in the present volume include most of the papers given at that symposium. In addition, the volume contains six previously published articles by Soisalon-Soininen, translated from German into English.

8

Introduction

One of the characteristics of the Septuagint as translation is the appearance of Hebraisms. Soisalon-Soininen defined a Hebraism as “an expression calqued from Hebrew that is out of tune with contemporary Greek usage” (“Back to the Question of Hebraisms” in this volume). The counterpart of a Hebraistic rendering is natural or idiomatic Greek. The typology of literal vs. free translations has to take into account what is or is not natural in the target language. John A. L. Lee’s “Back to the Question of Greek Idiom” aims to clarify the concepts behind Soisalon-Soininen’s treatment of Hebraisms. He proposes an approach to determine Hebrew interference that has caused non-idiomatic renderings. In the Greek Pentateuch there are clear cases of the translators following the Greek idiom, which demonstrates the translators’ competence. The Greek of the Septuagint Pentateuch can be described as “Greek with Hebrew interference” rather than as “Hebraic Greek”. Some linguistic features, such as the iterative use of ἄν, can be considered to reflect contemporary Koine usage even though the earliest evidence can only be found in the Greek Pentateuch itself. In “Soisalon-Soininen Meets Grice: The Cooperational Principle and the Septuagint Syntax”, Seppo Sipilä provides a historical overview of the development of Soisalon-Soininen’s translation-technical method. He then proceeds to adapt Paul Grice’s Cooperative Principle to understand the Septuagint as a translation: “the ancient translators when conducting the translation work made assumptions as to what their audience was expecting when it comes to translating an important text.” The reason for often unidiomatic formal renderings must be looked for in the translators’ aim for an effortless way of translating; not in their putative Hebrew mind-set, as suggested early by Soisalon-Soininen. Conversely, when abandoning a literal alternative for a more idiomatic rendering, the translators did so to facilitate communication in a cooperative fashion. The Hebrew genitive construction does not use the definite article in connection with the thing owned; nevertheless, the thing owned is considered to be defined. In Greek too, the thing owned is usually considered defined and thus marked with the definite article. When translating from Hebrew into Greek, the translators had to choose whether to mechanistically follow the Hebrew structure and leave out the article, or to use the article in accordance with good Greek practice. Raija Sollamo investigates this in “The Usage of the Article with Nouns Defined by a Nominal or Pronominal Genitive in LXX Genesis”. She finds that such nouns have an article in a surprisingly high 78.5% of the cases. In this respect, the Greek Genesis does not display Hebrew interference. Rather, the richness of idiomatic variation in some renderings of the Hebrew genitive structure suggests that the translator must have been a native Greek speaker. In his translation-technical studies, Soisalon-Soininen often set out from Hebrew features that have no natural counterpart in Greek. The studies described how the Septuagint translators transposed these grammatical structures into Greek. It is also possible to adopt a reverse approach: to investigate the use of

Introduction

9

a characteristically Greek feature that has no direct counterpart in Hebrew. Jan Joosten (“Grammar and Style in the Septuagint: On Some Uses of Preverbs”) holds that the use of such characteristically Greek features opens up a window onto the style of the Septuagint. His study demonstrates that the translators were not insensitive to requirements of style in their treatment of preverbs. They used preverbs to achieve a rhetorical flourish or a desired nuance, such as gradation. Although not always successful, the translators were capable of making use of the possibilities of the Greek language to create a pleasing effect. Soisalon-Soininen suggested that the Septuagint translators rendered the Hebrew text segment by segment. Occasionally, this led to losing sight of the context and resulted in awkward renderings. (“Segmentation in the Greek Pentateuch” in this volume.) Theo van der Louw revisits “The Dynamics of Segmentation in the Greek Pentateuch”. He reviews three phenomena that can be used as evidence of segmentation: losing touch with the context, translating freely within segment boundaries, and unsuccessfully anticipating the following segments. After having refuted two alternative explanations of these phenomena, van der Louw concludes that the dynamics of segmentation explain those best “through its appeal to the workings of the human brain”. In addition, he provides a typology of “intrasegmental and intersegmental inconsistency”: the former is caused by short-term memory overload, the latter by other factors such as losing touch with the preceding context. Building upon Ricardo Maldonado’s theory, Anssi Voitila investigates the “Middle Voice as Depiction of the Subject’s Dominion in the Greek Pentateuch” with the verb τιθέναι. The nine occurrences of the middle aorist of τιθέναι indicate relocation, putting someone in prison, giving laws, and making covenants. The use of the middle aorist in these cases does not imply the subject being affected by the action. Rather, it indicates that the subject is highly involved in the action and that the action has a focusing function. While the subject’s affectedness remains a good interpretation of the Greek middle voice in many instances, one must not overlook the possibility of the middle emphasizing the agent’s role as the initiator of the action, superior to the patient or the beneficiary. Patrick Pouchelle asks “Did the Translators of the Pentateuch Differentiate the Piel from the Qal?” He provides a survey of current research on the meaning of the piel and the qal stems, and the Greek translators’ treatment of the Hebrew binyanim in general. He observes that when the qal has a stative and the piel a fientive meaning, the translators discerned the difference using the passive for qal and the active for piel—or two different Greek verbs for the two Hebrew stems. For a fientive qal, the difference is less clear; e.g., both stems of ‫‘ ׁשלח‬to send’ can be rendered with either ἀποστέλλω or ἐξαποστέλλω, but the latter is used more frequently for the piel. However, the use of the piel often indicates “dismissal” rather than “sending with a mission” and the distinction used by the translators may be regulated by the context rather than a perception of the underlying Hebrew stem.

10

Introduction

Philippe Le Moigne’s “Une tournure syntaxique fréquente en Ésaïe-LXX: substantif abstrait + λήμψεται + complément d’objet direct” investigates the usage of λαμβάνω ‘to take, receive’ with an abstract noun subject and a direct object in the Septuagint of Isaiah. The translator consciously favoured the formulation even in some cases where the putative Hebrew Vorlage did not readily suggest it. In some of these cases the semantics remain the same even though the Greek syntax clearly differs from the Hebrew (e.g., “they shall obtain joy and gladness” Isa 35:10 MT / “and joy shall take hold of them” LXX). In others, the translation seems to abandon the Hebrew semantics as well (e.g., “the mountains would quake at your presence” Isa 64:1 MT / “trembling from you would seize the mountains” 63:19 LXX). Le Moigne observes that the rendering is used in particular for climatic phenomena and it produces semantic and stylistic unity for such passages. The study contributes to the understanding of the Greek translator of Isaiah as one having a global vision of the work: the translator knows the book of Isaiah in Hebrew perfectly, and does not need to resort to a word-for-word translation style. The nature of the Septuagint Jeremiah is distinctively different between its earlier (chapters 1–28: Jer a’) and later (29–52: Jer b’) parts. Miika Tucker (“The Infinitive in Septuagint Jeremiah”) investigates how the renderings of the Hebrew infinitive differ between the two parts. A noteworthy feature is the rendering of the Hebrew construction ‫ל‬+inf. with the Greek infinitive either with or without the definite article. Both renderings appear in both of the two sections, Jer a’ and Jer b’, but in the former section the articular infinitive is more frequent than in the latter section. The difference is most notable in infinitives expressing purpose: in Jer a’, these have an article in 84% of the possible cases while in Jer b’ the rate is only 45%. As the renderings of Hebrew infinitives with other prepositions yields similar results, it can be concluded that the renderings of infinitives in the two sections display a marked difference in their character. Scholars of the syntax of the Septuagint cannot avoid the question of which Greek text form it is they are actually studying. Today, there is a full critical edition for about two-thirds of the text in the series Septuaginta: Vetus Testamentum Graecum by the Göttingen Academy of Sciences and Humanities. The books for which an edition has not yet come out comprise Joshua, Judges, 1–4 Kingdoms (Samuel-Kings), 1 Chronicles, Song of Songs, Proverbs, and 4 Maccabees. In addition, a new edition of the Psalms is in planning. For the aforementioned books, the textual traditions occasionally differ considerably. A major reason for this is recensions that aimed at bringing the text into closer conformity with the Hebrew text or enhancing the literary style of the translation. While the recensions were often unsystematic, they nevertheless brought about noteworthy lexical and syntactic changes. Such recensions feature in two articles that combine text-historical and linguistic study of the Septuagint version of the Book of Judges. Anneli ­Aejmelaeus

Introduction

11

revisits Soisalon-Soininen’s dissertation in “Translation Technique and the Recensions: A Late Review of Ilmari Soisalon-Soininen’s Doctoral Thesis on the Text-Forms of Judges”. When Soisalon-Soininen began his research, it was unclear whether the A and B text forms of Judges represented two different translations, or whether one was dependent on the other. Soisalon-Soininen first looked for translation-technical similarities between the two text forms. This resulted in a characterization of the translation in areas such as word order, conjunctions, prepositions and semi-prepositions, verb construction, pronouns, the Hebrew infinitive, and nominal clauses. Features known from ­Aquila, ‫ אנכי‬rendered with ἐγώ εἰμι when followed by a finite verb and translating the object marker ‫את‬ by σύν + accusative, he considered not to be features of the original translation. When the original translational features are duly characterized, the recensional features must be investigated against those. Soisalon-Soininen’s study demonstrated that the B text presents a recension according to the Hebrew text: this explains its greater degree of literalness when compared with the A text. José Manuel Cañas Reíllo, the editor of Judges for the Göttingen Septuagint, provides an overview of “Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges”. His study builds upon Soisalon-Soininen’s dissertation, refining it on various issues. Cañas Reíllo defines four main textual groups: B, Hexaplaric (O), Lucianic (L), and mixed (M), and analyses how their vocabularies differ. The distinct lexical features of B result from a Hebraizing recension that went beyond the kaige recension. O has some, but not many, distinct Hebraizing lexemes. L has the greatest number of lexemes found only in that group. This phenomenon is partly due to its retaining ancient vocabulary via its proto-Lucianic stage; this makes it a particularly important witness for the original translation. M oscillates between B and AOL when those differ in lexemes; in those cases, AOLM often, but not always, attest to the old reading. The translators of the Septuagint did not use some syntactic features for the sake of syntax; they used them in order to convey meaning. Meaning is tied to both syntax and the semantics of the lexemes used. Syntax helps to orientate with regard to the semantic field of a lexeme: for instance, λαμβάνω means quite different things when regulated by contextual subjects and objects. Conversely, sometimes it is only the semantics of the lexemes that allow the reader to correctly recognize the constituents of a sentence. The study of the syntax of the Septuagint should not be entirely separated from the study of semantics. In this volume, this fact is acknowledged by including two articles on lexicography: William A. Ross, “Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt”, and Srecko ­Koralija, “Παράδεισος as/and Κῆπος”. The use of Greek words for young persons was not so much informed by the actual age of the person, but rather by their perceived age and by sociocultural factors. This makes it difficult to differentiate between terms such as παῖς and

12

Introduction

παιδίον or νεός and νεανίας. Ross examines the use of two such words, παιδάριον and νεανίσκος, in Greek literature, documentary papyri, and inscriptions of Ptolemaic Egypt. He concludes that παιδάριον means ‘a semi-skilled labourer’ and νεανίσκος ‘a civic officer’. A similar distinction can be observed in the use of these terms in LXX-Ruth and LXX-Ezekiel. This observation provides further evidence that the translators of those books lived in an Egyptian milieu. The skill of the translators is demonstrated by their ability to use these terms in ways appropriate to their social and linguistic context. Παράδεισος is an Avestan/Old Persian loan word denoting ‘an orchard’ while the native Greek word κῆπος means rather ‘a garden’ with vegetables and plants. Koralija observes that in Classical Greek, παράδεισος seems to imply a degree of grandeur, while in Hellenistic times it can be used to denote various kinds of orchards without an implication of exceptional splendour. The Septuagint reflects the general Hellenistic usage: κῆπος mainly designates a place containing vegetables and seeds and is often associated with personal property, while παράδεισος may carry associations of prestige, as can be seen in its being the preferred term for the garden of Eden in the Genesis creation narrative. Subtle differences in meaning between the two words call for special attention on the part of lexicographers. The volume concludes with six articles by Soisalon-Soininen, originally written in German and translated into English by Theo van der Louw: “The Use of the Genitive Absolute in the Septuagint” (1973), “A Palette of Renderings for the Hebrew Construct State in the Greek Pentateuch” (1977), “The Alleged Interchangeability of ἐν and εἰς in the Septuagint” (1982), “Segmentation in the Greek Pentateuch” (1983), “The Rendering of the Hebrew Personal Pronoun as Subject in the Greek Pentateuch” (1984), and “Back to the Question of Hebraisms” (1990). These studies have lasting value, not only as material for history of research, but as analyses of significant translation-syntactic phenomena. All of them were pioneering studies of their time, and many of them remain the only analysis of the subject. Together with Soisalon-Soininen’s monographs, they crystallize the translationtechnical method. Helpfully, every article is preceded by a short abstract. At the end of this volume, before the list of literature cited in the articles, the complete bibliography of Ilmari Soisalon-Soininen is given. Thanks are due to Mr. Aki Ranta for compiling the bibliography and the indexes. The editors of the present volume hope that this book can serve as inspiration for more research on the syntax of the Septuagint and to uphold the legacy of Ilmari Soisalon-Soininen. Tuukka Kauhanen

John A.L. Lee Back to the Question of Greek Idiom 1. Soisalon-Soininen and Greek Idiom The title of my paper intentionally echoes that of a 1990 paper by Soisalon-­ Soininen, “Back to the Question of Hebraisms”, providing both a link to his work and a starting-point for the present discussion.1 That paper was, naturally, about Hebraisms, which he saw as neglected in recent work on literal versus free rendering in the LXX. He did not only discuss Hebraisms, however, but made significant mention of idiomatic renderings: “… nearly all the translators occasionally produce very skilful, idiomatic renderings”. He also said that “our work is based on (the question) which renderings are Hebraistic and which are idiomatic”. By idiomatic renderings he evidently meant those that use natural, normal Greek; and the work to which he referred was “to study the differences between individual translators”.2 Soisalon-Soininen thus recognised the presence of idiomatic Greek in the LXX, and his paper gave some illustrations, as well as examples of what he termed Hebraisms. But it seems to me he did not really set out a means of identifying the two, though he did give a definition of a Hebraism.3 As to idiomatic renderings, he only said that they are “renderings which involve, to a greater or lesser degree, a break with the customary word for word approach”4—which is rather limited and I think not true: a “word for word” rendering may still be idiomatic Greek. (I admit to some trouble following his thinking here, as at times elsewhere). 1 I. Soisalon-Soininen, “Zurück zur Hebraismenfrage”, in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien zur Septuaginta. Robert Hanhart zu Ehren (Göttingen: Vandenhoeck & Ruprecht, 1990) 35–51; transl. by Theo van der Louw: “Back to the Question of Hebraisms” in this volume. Page refs. are to the original; translations are van der Louw’s with occasional revisions. 2  “Auf der anderen Seite bringen fast alle Übersetzer gelegentlich sehr geschickte idiomatische Wiedergaben” (Soisalon-Soininen, “Zurück”, 43); “[Ob wir es geradeheraus sagen oder nicht,] so gründet sich unsere Arbeit dann darauf, welche Wiedergaben hebraistisch, welche idiomatisch sind” … “[Haben wir denn überhaupt die Möglichkeit,] die Unterschiede zwischen den verschiedenen Übersetzern zu untersuchen?” (“Zurück”, 51). 3  “Ein Hebraismus in der Septuaginta ist ein aufgrund des Hebräischen entstandener Ausdruck, der nicht mit dem Sprachgebrauch des gleichzeitigen Griechisch im Einklang steht oder der vom Hebräischen aus erklärliche Gebrauch möglicher Vokabeln oder Ausdrücke im Griechischen in Zusammenhängen, in denen sie nach dem griechischen Sprachgebrauch nicht passend sind” (Soisalon-Soininen, “Zurück”, 39). 4  “…Wiedergaben, die einen größeren oder kleineren Eingriff in die übliche Wiedergabe der einzelnen Teile des Ausdrucks voraussetzen” (“Zurück”, 43).

14

John A.L. Lee

In other publications also, Soisalon-Soininen made observations on the presence of idiomatic Greek in the LXX. I take two further papers that show this and reveal more about the significance he attached to these features. A 1973 paper dealt with the genitive absolute.5 This, as Soisalon-Soininen of course knew well, is a purely Greek construction without any equivalent construction in Hebrew, and therefore a case of idiomatic Greek. It appears only sporadically in the LXX as a whole, though more frequently in books like Job and Proverbs (as one would expect). When it does appear, it gives us (he says) a glimpse of “the true language competence of the translators”.6 Though he does not quite spell it out, by “competence” Soisalon-Soininen evidently means competence or skill in Greek, not competence in translation. He adds, moreover, in his summing up, that “we should give more weight to these rare cases”,7 a sentiment that is in harmony with the aim of this paper—though I would not agree that they are rare. Clearly, then, Soisalon-Soininen recognised the presence of Greek idiom in the occasional appearances of the genitive absolute. Yet he adds the (to me) curious interpretation that “in those cases, the constraining factors stepped back for some reason”.8 What are those “constraining factors”? What does he mean? Evidently they are the constraints imposed by attempting to produce a close rendering of the original Hebrew. This is deducible from Soisalon-Soininen’s preceding discussion, where he speaks of the syntax of the LXX as “translationese” and asserts that “one could say that it is not Greek but rather Hebrew with Greek words”.9 Free renderings, then, are not expected, they are not supposed to occur; the translators were supposed to render the Hebrew exactly and not depart from it. And why “for some reason”? Could the answer be in doubt? Soisalon-Soininen does not explain further or seem to want to explore why a translator might depart from the strict letter of the original.

5 I. Soisalon-Soininen, “Der Gebrauch des genetivus absolutus in der Septuaginta”, in The Fifth World Congress of Jewish Studies (vol. 4; Jerusalem: World Union of Jewish Studies, 1973), 131–6; repr. in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur ­Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 175–80; transl. by Theo van der Louw: “The Use of the Genitive Absolute in the Septuagint” in this volume. Page refs. are to Studien; translations are van der Louw’s. 6  “[D]ie wirklichen Sprachkenntnisse der Übersetzer” (Soisalon-Soininen, “Gebrauch”, 180). He also speaks of “ihr wirkliches Sprachgefühl and ihr Können” (176) and “die Geschicklichkeit … der Übersetzer” (180). 7  “[M]üssen wir das Gewicht eben auf diese seltenen Fälle legen” (Soisalon-Soininen, “Gebrauch”, 180). 8  “Es sind aus irgendeinem Grunde die hemmenden Faktoren zurückgetreten” (SoisalonSoininen, “Gebrauch”, 176). 9  “Übersetzungssprache”; “Man könnte sagen, es ist kein Griechisch, sondern Hebräisch mit griechischen Worten” (Soisalon-Soininen, “Gebrauch”, 175).

Back to the Question of Greek Idiom

15

My third sample is Soisalon-Soininen’s 1977 paper on the relative pronoun.10 The main interest here is in what Soisalon-Soininen calls “nominal relative clauses”, that is, those in which a Hebrew relative clause introduced by ‫ אׁשר‬is translated as a noun phrase with the article, as for example: Gen 43:16 ‫ = ויאמר לאׁשר על־ביתו‬and he said to the (one) who (was) over his house. καὶ εἶπεν τῷ ἐπὶ τῆς οἰκίας αὐτοῦ = and he said to the (one) over his house. Gen 19:11 ‫ = את־האנׁשים אׁשר־פתח הבית‬the men who (were) at the door of the house. τοὺς ἄνδρας τοὺς ὄντας ἐπὶ τῆς θύρας τοῦ οἴκου = the men the (ones) being at the door of the house.

The renderings are natural Greek idiom; the literal equivalent with a relative pronoun would have been possible but less natural. Soisalon-Soininen describes such forms of expression as “idiomatic, one might even say stylish from the viewpoint of the Greek language” and “good Greek style” (57). His appreciation of these as Greek idiom is thus plainly and well stated. But he nevertheless seems to want to push—he does push—the Greek into a closer match to the Hebrew. He does this by an argument that equates the article with the relative pronoun: “the [Greek] relative pronoun is replaced by the almost identical article” (56), and “in some cases the form is actually identical” (57). One may respond that to equate the Greek article and relative pronoun is to ignore their difference of function; and it is not really true that the forms are “almost identical”.11 But on this basis Soisalon-Soininen goes on to assert that “this type of translation is not to be taken as a significant indication of the liberties taken by the Septuagint translators” (57). We are thus back to the position that the translator is following his Semitic original exactly and not really departing from it, even when he turns a Hebrew relative clause into a natural Greek equivalent with different syntax—and does depart from it.

2. A Different View This brief exploration of three of Soisalon-Soininen’s papers suffices to give an indication of his thinking on the subject of Greek idiom in the LXX. Now my 10  I. Soisalon-Soininen, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch”, in Proceedings of the Sixth World Congress of Jewish Studies Held at the Hebrew University of Jerusalem, 13–19 August, 1973 (vol. 1; Jerusalem: World Union of Jewish Studies, 1977) 401–6; repr. in Studien zur Septuaginta-Syntax, 55–61. Page refs. are to Studien. The original is in English. 11  The only forms that are the same are ὁ, ἡ, οἱ, αἱ, and these are distinguished by the accent (rel. pron. ὅ, ἥ, οἵ, αἵ). Further, ὅ is neuter nom./acc., so cannot function like masc. ὁ.

16

John A.L. Lee

aim here is not to criticise everything he said, or to diminish the value of all the fine work he did. Rather, I wish to bring the focus on to the element that he himself identified, the idiomatic Greek renderings. He and I agree that they are present; how we ought to view them is where we differ. He evidently sees them as somehow unexpected or anomalous, even difficult to explain, in translations presumed to follow the original as exactly as possible. To me this is to approach the matter around the wrong way. The LXX translators, at least in the Pentateuch, were not tied to producing a close, exact translation, which they departed from only when forced to—for some strange reason. Rather, natural Greek, of which they had an intimate knowledge, was their starting-point, and they deviated from it because of their generally literal method. They follow this method with great inconsistency: they are just as likely to produce natural Greek as to produce Hebraic Greek. That is why natural Greek idiom keeps appearing. It may be said, of this or that Hebrew feature, that an idiomatic rendering is statistically infrequent compared with a Hebraistic one, and therefore the former is of little or no significance: the translators’ aim was still to make the translation as literal as possible, and deviations are in the nature of anomalies. Again I take a contrary view. That a certain feature of Greek idiom appears at all is significant, all the more so if it involves an unrequired or unforced change, as in the case of a shift of word-order. One instance of that kind is just as telling as many. Whether one agrees with all this or not, the question of the presence of idiomatic Greek in the LXX is a very important one, and I suggest is worth pursuing for its own sake. The aim of this paper is to demonstrate the presence of incontestable cases of Greek idiom in the LXX-Pentateuch. The enclitic pronoun/adjective τις and the particle ἄν will be our examples for this purpose. But we need to begin with sound criteria for identifying Greek idiom.

3. Identifying Greek Idiom The question of Hebrew interference in the Greek of the LXX has been the subject of enduring discussion, in various guises, since the days of Thackeray. But the sort of methodological framework suited to our purpose has not yet been proposed (as far as I know). I suggest that it is helpful for the present purpose to divide all LXX renderings into three categories, as set out below. This is not, of course, a complete system for analysing all features of a rendering: it is aimed at finding renderings free of interference. I recognise also that the categories are broad and the terminology imprecise. But further refinements can follow, and some will emerge in the discussion below. The illustrations all involve Hebrew ‫ איׁש‬in some combination, and all have been chosen from the Leviticus translator, deliberately so, to show that he has no less control of Greek idiom than his fellowtranslators. These are the three categories of rendering that I propose:

Back to the Question of Greek Idiom

17

1. The rendering matches the Hebrew and is not natural Greek: interference is present. Example: (1) Lev 17:3 ἄνθρωπος ἄνθρωπος τῶν υἱῶν Ἰσραήλ, ὃς ἂν σφάξῃ μόσχον … (4) καὶ λογισθήσεται τῷ ἀνθρώπῳ ἐκείνῳ αἷμα A man a man of the sons of Israel who slaughters a calf … (4) then bloodshed shall be reckoned to that man. ‫( דם יחׁשב לאיׁש ההוא‬4) ... ‫איׁש איׁש מבית יׂשראל אׁשר יׁשחט ׁשור‬

2. The rendering matches the Hebrew and is consistent with natural Greek: interference may or may not be present. Example: (2) Lev 25:17 μὴ θλιβέτω ἄνθρωπος τὸν πλησίον· καὶ φοβηθήσῃ κύριον τὸν θεόν σου. Let a man not oppress his neighbour; and you shall fear the Lord your God.



‫ולא תונו איׁש את־עמיתו ויראת מאלהיך‬

3. The rendering does not match the Hebrew and is natural Greek: interference is not present. Example: (3) Lev 6:40(7:10) καὶ πᾶσα θυσία ἀναπεποιημένη ἐν ἐλαίῳ καὶ μὴ ἀναπεποιημένη πᾶσιν τοῖς υἱοῖς Ἀαρὼν ἔσται, ἑκάστῳ τὸ ἴσον. And every sacrifice made up with oil or not made up shall be for all the sons of Aaron, to each an equal portion.



‫וכל־מנחה בלולה־בׁשמן וחרבה לכל־בני אהרן תהיה איׁש כאחיו׃‬

Brief explanations will be useful. Example (1) shows a rendering that is not acceptable as natural Greek, even though the meaning would not be totally opaque to a Greek speaker. Comprehension would be aided by “distributive doubling”, a known though rare phenomenon in Greek, as well as by context. But a natural idiomatic Greek equivalent would be different, and one is produced by the Leviticus translator himself in Lev 20:2 ἐάν τις ἀπὸ τῶν υἱῶν Ἰσραὴλ … (‫איׁש איׁש‬ ‫)מבני ישׂראל‬. Renderings in this first category can be labelled “Hebraisms”, if one wants to use the term. In renderings that fall into category 2—and they would probably be the great majority of all renderings—the question of interference remains open. Although they match the Hebrew, they are not inconsistent with natural Greek, that is, they are capable of being read as natural Greek, and are therefore not certainly Hebraisms. Example (2) is in fact not an entirely natural Greek form of expression: we

18

John A.L. Lee

would be unlikely to find it in a free Greek text. But one cannot say it would be impossible in terms of Greek grammar, syntax, or word-usage. Renderings of type 3 are by definition one thing and not the other. They are undoubted cases of Greek idiom because they do not match the Hebrew. In example (3) the Hebrew literally means “a man as/like his brother”. The translation departs from a literal rendering and gives an idiomatic Greek equivalent that conveys the sense of the original in different words, literally “to each the equal (portion)”. The insertion of τό is Greek idiom (where English uses “an”). No better proof of the Leviticus translator’s grip on Greek could be found than this rendering. Some further observations arising from the discussion in the Helsinki Symposium and afterwards may be added here.12 As regards the term “Hebraism”, I would reserve it for individual renderings of type 1 alone, where natural Greek is by definition not present; I would not use it for those of type 2, where there is nothing wrong with the rendering as Greek. This means that a great many renderings are left open; but methodologically that is sounder than classing everything that matches the Hebrew as a “Hebraism”. Other considerations may come into play to help decide the question. Many of the renderings of type 2 seem to me to be a matter of style. That is, they reflect the style of the original, which like the content is carried over into the translation. It would be surprising if this were not so—unless we think that a Hebrew book can be easily converted into a Greek one. Example (2), as we have noted, may not be the sort of thing a Greek writer would say, but it is acceptable Greek. The reason for the oddity is that the form of expression, like the content, comes from the original. Style is also the answer, in my opinion, to the perennial issue of frequency. If a Hebrew expression rendered by a natural Greek equivalent appears more frequently than in a normal Greek text, is it a “Hebraism”? My answer is definitely no. Since each individual instance is not a Hebraism, the frequency only reflects that of the original, so it is a feature of style. Soisalon-Soininen expressed it well in his discussion of ἰδού: “So, even ἰδοὺ φωνὴ μεγάλη is not a Hebraism per se. But the frequency of ἰδού gives the text a Hebraizing flavour”.13 The term “Hebraizing flavour” is of course inexact, but I am not sure of a better alternative. “Stylistic interference” from the Hebrew original might work and is worth a trial. It is only renderings of type 3 that provide us with undoubted idiomatic Greek. We turn now to examples. 12  Discussion of LXX syntax may advance even in an ice-cream shop or at a tram-stop (thank you Theo van der Louw and Jim Aitken). 13  “So ist sogar ἰδοὺ φωνὴ μεγάλη an sich kein Hebraismus. Das sich oft wiederholende ἰδού gibt aber dem Text ein hebraisierendes Kolorit” (Soisalon-Soininen, “Zurück”, 41). Similarly: “… eine wiederholte Verwendung … dem Ganzen ein hebräisches Kolorit geben kann” (39). Cf. R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (Suomalainen Tiedeakatemia: Helsinki, 1979), 299: “… exceptionally frequent usage [of the studied expressions] in the LXX is Hebraistic (a stylistic Hebraism)”.

Back to the Question of Greek Idiom

19

4. The Enclitic Pronoun/Adjective τις The enclitic pronoun/adjective τις is found 79 times in the LXX-Pentateuch.14 A full analysis appears in my forthcoming book and the data presented here are derived from it.15 First let us consider the matches between τις and the Hebrew original. Hebrew has no exact equivalent, since it does not employ enclitics like this at all. The Hebrew word for the idea is ‫איׁש‬, which has the primary meaning “man”, with a subsidiary function to mean “someone, a person”. The Pentateuch translators do sometimes use the “literal” equivalent ἀνήρ or ἄνθρωπος to render ‫ איׁש‬meaning “someone, a person”, as in the renderings by ἄνθρωπος in examples (1) and (2) above. Mostly, however, they employ τις or ἕκαστος (“each”), both of which are a move away from a “literal” rendering in favour of more natural Greek.16 Summary data for τις and the corresponding Hebrew in the Pentateuch are as follows: Total of τις in the Pentateuch: 79 In clauses: 54 In noun phrases: 22 In μή τι: 3 With equivalent in Hebrew: 43 ~ ‫איׁש‬: 40 ~ ‫אחד‬, ‫מאומה‬: 3 Without equivalent in Hebrew, i.e., added: 36 Position of τις: In same position as Hebrew equivalent: 19 Repositioned: 24

The tabulation includes data on the position of τις in its clause—and this is where the story becomes really interesting. Like other enclitics, τις conforms to certain position rules. The main one is that it tends to gravitate to the second position in its clause, or the “second slot” (to use a term of my own). Position in the second slot is a manifestation of Wackernagel’s Law, best known for its effects on the

14  Initial search via Accordance©; readings checked in the Göttingen edition. I do not include Exod 22:9(8) ὅ τι οὖν [sic Gött. edn.], which is an instance of the relative pronoun ὅστις + -οὖν. 15  J.A.L. Lee, The Greek of the Pentateuch: Grinfield Lectures on the Septuagint 2011–2012 (Oxford: Oxford University Press, 2018). 16  ἕκαστος is found 80 times in the LXX-Pent. and corresponds to (‫ איׁש )אׁשה‬or some combination thereof in 77 of them. This word, too, could be used to demonstrate the presence of idiomatic Greek.

20

John A.L. Lee

position of the enclitic personal pronoun.17 For τις I have formulated 11 position rules, based on a close analysis of the 79 occurrences in the Pentateuch, supported by parallels in non-LXX Greek. This list of rules is drawn on here and presented in full in my book. The question of interest is how much the position of τις matches that of its equivalent ‫איׁש‬. Some might expect it to match consistently, being changed only where ‫ איׁש‬is first word and τις has to be moved, or perhaps not even then. In fact, while the position of τις often matches that of ‫איׁש‬, it is more often changed (24 times), the new position being of course in comformity with the Greek position rules. Moreover, in a large proportion of the examples (36 out of 79), τις is added without any match to a word in the original. Why would this be? The answer can only be that the addition of τις is natural Greek idiom. Let us now consider τις in relation to the categories of rendering defined above, with selected examples and parallels. There are no instances of τις in category 1, that is, matches to the Hebrew that are also not natural Greek. In all the cases where τις keeps the same position as ‫איׁש‬, none are unnatural Greek positions. Where keeping the Hebrew order would have led to an unnatural position, as in example (7) below, the translators avoided it by moving τις to a new position. There are instances of τις in category 2, that is, those in which τις keeps the same position as ‫( איׁש‬or another Hebrew equivalent), and is in a natural Greek position. According to my analysis there are 19 of these. Two examples of τις in category 2: (4)

Exod 16:20 ἀλλὰ κατέλιπόν τινες ἀπ᾽ αὐτοῦ



‫ויותרו אנׁשים ממנו‬



‫כי יבער־איׁש ׂשדה‬

BGU 6.1280.17 (210 BC) ἐὰ]ν̣ δὲ μὴ βεβαιοῖ ἀ̣λ̣λ̣ʼ ἐγβάλλῃ τις Διοκλῆν Pl. Theaet. 175c.1 ὅταν δέ γέ τινα αὐτός, ὦ φίλε, ἑλκύσῃ ἄνω, καὶ ἐθελήσῃ τις αὐτῷ ἐκβῆναι (5) Exod 22:5(4) ἐὰν δὲ καταβοσκήσῃ τις ἀγρόν BGU 14.2367.17 (III BC) ἐὰν δʼ ἐπ̣ά̣ναγ̣κ̣ε̣ς̣ δ̣α̣νείζῃ τι̣ς με̣τα̣[---] Pl. Symp. 221e.1 εἰ γὰρ ἐθέλει τις τῶν Σωκράτους ἀκούειν λόγων

Theoretically these could be due to interference, but other considerations apply that make this unlikely to be the main reason for the placing of τις. In example (4), the translator has good reason not to reposition τινες before the verb, though 17  See, e.g., G.C. Horrocks, Greek: A History of the Language and its Speakers (2nd ed.; Chichester: Wiley Blackwell, 2014), 108–9; R. Sollamo, “The Place of the Enclitic Personal Pronouns in the Old Greek Psalter”, in M.K.H. Peters (ed.), XII Congress of the International Organization for Septuagint and Cognate Studies, Leiden 2004 (Atlanta: Society of Biblical Literature, 2006) 153–60.

Back to the Question of Greek Idiom

21

this is a standard position in the case of ἐὰν δέ at the head of the clause, as in examples (6) and (7) below: τις cannot be placed immediately after a co-ordinating conjunction such as ἀλλά. In example (5), in which τις is “postponed” till after the verb when it could have gone before, there are two possible factors to consider. a) Exod 22:5(4) occurs in a rather lengthy string of regulations, many of them expressed in the same form as in example (6) below, that is, ἐὰν δέ τις + verb.18 It is possible the translator placed τις in a different position simply for variety. b) The second factor is a possible link to historical developments in the placement of enclitics. Over time Greek shows a tendency towards relocation of enclitics to a position after the verb.19 As can be seen in the parallel above, this was already happening in the time of Plato. It is possible the translator was unconsciously influenced by this undercurrent. Of course it could have been a little of both factors. At any rate, both these examples are definitely natural Greek and this explains them better than a resort to interference from the Hebrew. This is true of τις throughout the Pentateuch. But it is the instances that fall into category 3 that provide incontestable cases of Greek idiom. In these, τις is repositioned or added, contrary to the Hebrew, while being at the same time natural Greek. Examples of τις in category 3: a. τις corresponds to ‫איׁש‬, but is repositioned (6)

Exod 21:7 ἐὰν δέ τις ἀποδῶται τὴν ἑαυτοῦ θυγατέρα



‫וכי־ימכר איׁש את־בתו‬



‫ואיׁש כי־ימכר בית מוׁשב‬

(7)

Lev 25:29 ἐὰν δέ τις ἀποδῶται οἰκίαν οἰκητήν

SB 18.13256.9 (III BC) ἐὰν δέ τις παραθῆται, ἀποτίνειν αὐτόν … Pl. Phaedo 60b.6 ἐὰν δέ τις διώκῃ τὸ ἕτερον …

b. τις is added, without correspondence in the Hebrew (8)



Gen 6:5 καὶ πᾶς τις (“everyone”) διανοεῖται ἐν τῇ καρδίᾳ αὐτοῦ

‫וכל־יצר מחׁשבת לבו‬

UPZ 1.60.17 (179–168 BC) ἀλλὰ πᾶς τις πειρᾶται Pl. Theaet. 178a.5 ἔτι τοίνυν ἐνθένδε ἂν μᾶλλον πᾶς τις ὁμολογήσειεν

In (6) the translator moves τις to the standard position in the second slot, where the Hebrew equivalent ‫ איׁש‬came later, after the verb. There is another word there as well, the particle δέ, and τις is placed after it. This order is not accidental: there 18  Exod 21:7, 14, 20, 26, 33, 22:1(21:37), 22:7(6), 10(9). With the order ἐὰν δέ + verb + τις: Exod 22:5(4), 14(13), 16(15). 19  See Horrocks, Greek, 108–9.

22

John A.L. Lee

are word-order rules within the second slot. (More will be said on these under ἄν.) In example (7) the position of ‫ איׁש‬could not have been kept without breaking one of the basic rules for τις, that it is not placed first in its clause. Even καί τις (~‫ )ואיׁש‬would not have been idiomatic. In (8) τις is just a natural addition to πᾶς in the manner of idiomatic Greek since Classical times. So τις is a good example of Greek idiom in the LXX-Pentateuch, indubitably in the cases in category 3 and potentially in all cases in category 2. We may note again the point made earlier that the mere choice of τις as opposed to a “literal” equivalent of ‫ איׁש‬is a step towards more idiomatic Greek.

5. The Particle ἄν The particle ἄν occurs just over 400 times in the Greek Pentateuch.20 It is an entirely Greek feature without any counterpart in Hebrew, and therefore an instance of pure Greek idiom. It occurs in the LXX as a component of various clause types, as required by Greek syntax. These are familiar to us, even if we do not always notice ἄν or realise that it is not there by accident. Most of these clause types are illustrated in the examples below; parallels in contemporary (III BC) papyri are added. The “indefinite construction” in temporal and relative clauses is common, as in examples (10), (13), (14), (15), (18). In examples (9) and (17) we have the stylish addition of ἄν to ὅπως, a feature often seen in Ptolemaic officialese. In (11) and (16) there is the highly idiomatic Greek use of ἄν with a verb in the indicative to convey an “unfulfilled” condition (“I would have sent you …”). In example (12) we see the rare use of ὡς ἄν + optative in a simile. (9) (10) (11) (12) (13) (14)

Gen 12:13 ὅπως ἂν εὖ μοι γένηται BGU 8.1738.32 (72 BC) ὁπηνίκʼ ἂν εὖ συνθῶμεν. Gen 21:6 ὃς γὰρ ἂν ἀκούσῃ PCairZen 1.33.7 ἐπόρισεγ γὰρ ἂν αὐτοῖς Gen 31:27 ἐξαπέστειλα ἄν σε μετ᾽ εὐφροσύνης PMich 1.29.3 ἀπέστ[ειλα] | ἄν \σοι/ αὐτήν. Gen 33:10 εἶδον τὸ πρόσωπόν σου, ὡς ἄν τις ἴδοι πρόσωπον θεοῦ. PCairZen 1.93.18 κέχρηται ἡμῖν ὡς ἂν εἴ τις ἐχθρῶι χρήσαιτο. Exod 32:34 ᾗ δ᾽ ἂν ἡμέρᾳ ἐπισκέπτωμαι PHib 1.29.10 ἧι δʼ ἂν ἡμ[έρ]α̣ι̣ Lev 13:51 κατὰ πάντα ὅσα ἂν ποιηθῇ PCairZen 5.815.4 καὶ ὅσα ἂν εὑρεθῆι

20  In the current Göttingen text 394 of ἄν (as counted by Accordance©) + 12 of ἐάν = particle ἄν.

Back to the Question of Greek Idiom

(15) (16) (17) (18)

23

Lev 16:17 ἕως ἂν ἐξέλθῃ PPetr 2.3a.1 διὸ ἐπέχω ἕως ἂν ἐξέλθηι. Num 22:29 εἰ εἶχον μάχαιραν … ἤδη ἂν ἐξεκέντησά σε PCairZen 4.599.8 καὶ εἰ μὴ … ἐπεβοήθει … πάλαι ἂν ἠ̣σ̣χημόνουν. Num 16:40(17:5) ὅπως ἂν μὴ προσέλθῃ μηθείς PLond 7.2033.6 ὅπως ἂν μὴ κατέχητα̣ι ὁ Καλλικῶν. Deut 12:26 ἃ ἄν σοι γένηται PCairZen 3.358.3 ὅσα ἄν σοι δόξῃ

These uses of ἄν in the Pentateuch are interesting enough as instances of Greek idiom; but there is more. This particle exhibits word-order behaviour very similar to that of τις, even though it is not an enclitic. Like τις it gravitates to the second slot in its clause, but it has an even stronger tendency to do so than τις; it also precedes some words that τις does not, and τις itself. The rules can be seen played out in our selection: ἄν follows the particles γάρ (10) and δέ (13), but otherwise comes immediately after the headword of its clause, preceding the enclitic pronouns σε (11), σοι (18), the negative μή (17), and τις (12). These word-order rules can be traced back as far as Homer.21 This of course does not mean that the LXX translators learnt them from Homer; it means that the patterns were established in early Greek and continued to operate all the way to the time of the translators.

6. Conclusions The aim of identifying Greek idiom in the LXX-Pentateuch can be readily achieved. As we have seen, there are undoubted instances according to the criteria we have established. In the case of ἄν, every use falls into this category. In the case of τις, we have found clear instances in which Hebrew interference cannot be involved (60 in all); in the rest the possibility of interference remains open according to our criteria, but other factors apply that can tip the balance against interference. The approach through three defined categories of rendering (see section 3) provides a method applicable to all cases of potential interference, in effect to all renderings in the LXX. There are three further conclusions that I think are warranted. First, the evidence of idiomatic Greek woven into the fabric of the Greek Pentateuch demonstrates the translators’ complete competence in Greek. If one also takes into account other evidence beyond τις and ἄν—though these are significant enough—it is not too much to assert that the Pentateuch translators had native-speaker 21  See C.J. Ruijgh, “La place des enclitiques dans l’ordre des mots chez Homère d’après la loi de Wackernagel”, in H. Eichner/H. Rix (eds.), Sprachwissenschaft und Philologie. Jacob Wackernagel und die Indogermanistik heute (Wiesbaden: Reichert, 1990) 213–33.

24

John A.L. Lee

competence.22 It is hard to see what other explanation there is for their intimate knowledge of the usage of τις and ἄν, including the word-order patterns. We can say further that this observation applies to all the translators of the Pentateuch, however many there were. I believe that there were five, but the point is not affected even if this is not true: the spread of τις and ἄν shows that they were all equally competent in Greek. This is not the same as competence as translators. Nor do these two features enable us to characterise the techniques of individual translators: as far as I can see, they all use them in the same way. A second conclusion is that the presence of Greek idiom as in τις and ἄν supports the contention that the Greek translation of the Pentateuch should be viewed as in essence Greek with Hebrew interference, rather than as Hebraic Greek into which idiomatic Greek occasionally intrudes, for no evident reason. This is to some extent just a matter of viewpoint, but there is a genuine issue underlying it. What we expect to see will have an impact on what we find when we try to analyse the Greek of the Pentateuch and the translators’ methods. Soisalon-Soininen’s approach to relative clauses might be seen as an illustration (see section 1). A third conclusion relates to Koine Greek as a whole. Our usual approach to the Greek of the LXX, founded on a long-standing assumption that there is something anomalous about it and every feature needs to be proved to be acceptable Greek, has been to pursue a quest for parallels outside the LXX to provide the proof. What we have found in τις and ἄν in the Pentateuch, that is, normal usage of characteristically Greek idiom, suggests that the Pentateuch itself has value as evidence for Koine Greek (with of course the usual cautions and provisos). In regard to the uses of τις and ἄν that are already well-attested, as in those presented above, the Pentateuch has little new to offer. But things are different in the case of another use of ἄν, the “iterative” use with indicative in subordinate clauses, as seen in this example: (19) Exod 33:8 ἡνίκα δ᾽ ἂν εἰσεπορεύετο Μωυσῆς εἰς τὴν σκηνήν, … And whenever Moyses went into the tent … … ‫והיה כצאת מׁשה אל־האהל‬ (Lit.) And it happened at the going out of Moses to the tent …

Though this new Koine Greek construction is well attested later, there are no examples in III BC documents and there is a general lack of evidence for it in early Koine Greek.23 The Pentateuch, with some 15 instances, actually provides the (so far) missing proof that it had begun in III BC. As can be seen from the example, it 22  Further evidence supporting this conclusion is offered in Lee, Greek of the Pentateuch. 23  See E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit (2 vols. in 6 parts; Berlin/Leipzig: De Gruyter, 1906–38), 2,1:295; 2,3:79–80; 94; T. Muraoka, A Syntax of Septuagint Greek (Leuven: Peeters, 2016), 773.

Back to the Question of Greek Idiom

25

has nothing to do with the Hebrew original, nor can there be any reason to think the translators were using something that was not current Greek. To conclude. The study of Hebraisms has its place, but rather than go back to the question of Hebraisms I would like to see us go forward to the question of Greek idiom in the LXX. Our evidence has shown that Greek idiom is undoubtedly present in the LXX-Pentateuch, but the pursuit of such evidence has been slow up till now. We could now turn our attention to the many ways in which the translators followed Greek idiom, rather than the ways in which they departed from it.

Seppo Sipilä Soisalon-Soininen Meets Grice The Cooperative Principle and the Septuagint Syntax

1. The Legacy of Ilmari Soisalon-Soininen Professor Ilmari Soisalon-Soininen, whose centenary we celebrate with this symposium, is world-famous for his contribution to the study of the Septuagint syntax. When he started his scholarly career, he lamented that we lack a proper syntax for the Greek we meet in the Septuagint, and he wanted to contribute to efforts in correcting this unfortunate state of affairs. In 2016, Takamitsu Muraoka published a Syntax of Septuagint Greek, but without following the methodology created by Soisalon-Soininen. In Muraoka’s own words “[to] investigate every question from the perspective of translation technique would have been sheer madness.”1 I shall discuss in this paper the method that according to Muraoka leads into “a gigantic project”. How did Soisalon-Soininen develop the translation technical method? What kind of assumptions were there originally? And what can we learn from the development in linguistics since he published the magnum opus Die Infinitive in der Septuaginta? I shall start with a historical overview on the development of the method and introduce some less known writings by Soisalon-Soininen that explain the background of it and list some assumptions that he made while developing the method. The second part of my paper will deal with the Cooperative Principle by Paul Grice and finally I shall point out a few points where the method by SoisalonSoininen and the Cooperative Principle will intersect. Let me, however, start with a short personal reflection. I had the privilege of studying under prof. Soisalon-Soininen. He retired while I was still an undergraduate student, but he took interest in my master’s thesis and my post-graduate studies. It became quickly very clear that Soisalon-Soininen, or “Soiski” as we students used to call him, was a man with a keen interest in language and especially the language used in the Septuagint. Concerning my master’s thesis, Soisalon-Soininen gave me some very telling comments. My thesis was about the relationship between the MT and the LXX at the beginning of the book of Joshua. His main comment was that such a study is merely a duty of the commentaries and therefore not particularly interesting, at least not for him. He was more interested on the grammatical issues in the Greek 1  T. Muraoka, A Syntax of Septuagint Greek (Leuven: Peeters, 2016), xli.

28

Seppo Sipilä

and pointed out cases where the Greek translator used the infinitive instead of the participle, such as in Josh 4:11 ‫—ויהי כאשר תם כל העם לעבור‬καὶ ἐγένετο ὡς συνετέλεσεν πᾶς ὁ λαὸς διαβῆναι (pro διαβαίνων cf. e.g. Josh 3:17).2 Such a legacy tells a lot about my teacher. His interest lay in the formal features of the Biblical text, especially on the Old Testament text. In this light then, there is little wonder that Soisalon-Soininen himself concentrated on the linguistic study of the Biblical text.

2. Background of the translation technical methodology Soisalon-Soininen’s doctoral thesis from 1951 is, as we well know, about the textual history of the Greek Judges. He wrote it soon after the war, while working as a chaplain at a military hospital. I shall not discuss the meaning and results of his thesis, but concentrate on the methodology he employed while conducting his study. This is because already in the doctoral thesis the translation technical methodology is present, though in its early form. I do not think it is far-fetched to believe that after the Second World War, when Soisalon-Soininen prepared his doctoral work, the general tendency in Finland concerning Bible translation was a strong commitment to the formal translation technique, where people saw the translation only as a linguistic operation.

2.1 The Early Form of the Translation Technical Method Coming from this background, it seems only natural that Soisalon-Soininen begins his study by concentrating on the linguistic character of the Greek text of Judges. The key to the proper understanding of the character of the Greek text of Judges is to take the text as a translation. The following three factors influence the character of a translation says Soisalon-Soininen: mainly by the language from which the translation is made (source language), but also by the language into which the translation is made (target language). Finally, the third factor is the translator and his Einstellung ‘attitude’ to both languages.3 Such a short explanation was possible, because Soisalon-Soininen could assume that his readers 2  I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965), 39 says that in the LXX the participle is used in cases like this ziemlich allgemein, even though in Attic the infinitive is the norm with the verb συντελέω. He does not present any exact statistics for this particular case, but in the LXX the participle appears as the complement for συντελέω more than twice as often as the infinitive. 3 I. Soisalon-Soininen, Die Textformen der Septuaginta-Übersetzung des Richterbuches (AASF, ser. B, 72,1; Helsinki: Suomalainen Tiedeakatemia, 1951), 23. When he speaks about the translator, he uses the masculine singular, der Übersetzer. By saying this, he actually anticipates the answer to his main question about the Greek text of Judges.

Soisalon-Soininen Meets Grice

29

would not disagree on the basic starting point. It was enough just to mention it. If we think about the history of translation studies, this is just how it was. The first translation theories were linguistic ones and they appeared in 1950’s.4 Typical to these theories is a concentration on similarities and differences between language pairs by comparing how various linguistic units operate in the two languages. This is done by classification and systematization of differences in grammar and vocabulary making the whole operation look like a codeswitching enterprise.5 Even today many of us still do exactly this. This is also how Soisalon-Soininen presents his discussion of the Greek text of Judges. His main question was, do the various manuscript groups represent the same translation or not? To answer his question, a study of the character of the Greek translation becomes necessary.6 If the main sources of the Greek text share the same translation technical peculiarities (übersetzungstechnische Eigentümlichkeiten), the sources are based on the same Greek starting point and there is only one Urseptuaginta for Judges. Soisalon-Soininen starts his discussion on the translation technique in the spirit of Deissmann by pointing out that the Greek we find in the Septuagint is actually not any special form of Greek, but simply translationese (Übersetzungsgriechisch).7 This is due to the differences between the two languages, Hebrew and Greek. Before venturing into the details, Soisalon-Soininen says that even though the Hebrew text in Judges is not particularly difficult, the translator nevertheless had problems with it and must have occasionally recht hilftlos gewesen sein.8 The next step in Soisalon-Soininen’s discussion is the syntax. He discusses additions and omissions, word order, translations of the conjunctions ‫ ו‬and ‫כי‬, the grammatical subject, predicate and object, some prepositions and cases, pronouns, infinitives etc.9 It is to be noted that in this discussion Soisalon-Soininen starts from the grammatical elements in the Hebrew text and then studies, what are the Greek translations for each Hebrew element in focus. Therefore, we can see in Soisalon-Soininen’s thesis the methodology in its early form. The very basis of it is that when studying the translation the starting point must be the parent text and its wording. Nothing else will do.

4  Among the first published theories is Andrey V. Fedorov’s Введение в теорию перевода (An Introduction to Translation Theory), published in Moscow 1953. 5 See e.g. I. Vehmas-Lehto, Kopiointia vai kommunikointia: Johdatus käännösteoriaan (Helsinki: Finn Lectura, 2002), 35–36. 6 Soisalon-Soininen, Die Textformen, 14. 7 Soisalon-Soininen, Die Textformen, 23–24. 8 Soisalon-Soininen, Die Textformen, 30. 9 Soisalon-Soininen, Die Textformen, 31–60.

30

Seppo Sipilä

2.2 The Fully Developed Method Soisalon-Soininen turned to a proper grammatical analysis of the Septuagint as a whole during early 1960’s. In 1962 he received a three year research grant, which made it possible to work full time on the subject. In 1964 he was nominated as the associate professor of the Old Testament studies and in 1967 as the full professor. In 1963 appeared an essay on the methodology Septuagintan lauseopin tutkimuksen perusteista ‘On the basics of the study of the Septuagint syntax’. In this essay Soisalon-Soininen explains how one should conduct a study of the syntax. As the essay is published only in Finnish, I shall briefly explain its main content. Soisalon-Soininen starts by describing several important tools for studying the Septuagint syntax. He laments the fact that the grammars published by Helbing and Thackeray only include the morphology. He then continues in discussing several individual studies concentrating on how these handled the infinitive. Because these studies start from the Greek text, not from the Hebrew one, one cannot get the full picture of the phenomenon if one uses these as the basis, he says. When it comes to the use of the infinitive, it is not enough to describe how the Greek infinitive is used, because then one cannot take into account the differences between the Hebrew texts in various books nor the individual way the translators handled the Hebrew text.10 On the other hand, with some studies where the Hebrew background of the Septuagint is taken as the starting point, several other weaknesses emerge. Soi­ salon-Soininen specifically mentions how the organization of the Hebrew phenomena does not take into account the translation challenge. This leads into a fragmented discussion. The same problem appears in several different instances and the result is that the thread of the discussion is lost.11 As the term ‘­translation challenge’ is one of the key terms Soisalon-Soininen uses, I will come back to this later. After discussing several studies and mainly criticising them, Soisalon-Soininen turns then to the methodology. He thinks that it has not been clear to the authors of these studies, what is the nature of the study on the Septuagint syntax.12 The Septuagint is, says Soisalon-Soininen, almost without exception a formal translation. Considerable freedom in translating is rare. Moreover, the work was done by Greek speaking Jews for other Greek speaking Jews. This makes it understandable, Soisalon-Soininen continues, that the wording of the Greek translation heavily depends on the original language, Hebrew. 10  I. Soisalon-Soininen, “Septuagintan lauseopin tutkimuksen perusteita”, Teologinen Aikakauskirja 68 (1963) 216–31, on pp. 217–19 on Votaw’s The Use of the Infinitive in Biblical Greek (1896). 11 Soisalon-Soininen, “Septuagintan lauseopin”, 222–3. 12 Soisalon-Soininen, “Septuagintan lauseopin”, 225.

Soisalon-Soininen Meets Grice

31

This claim of his calls for an explanation. Ten years earlier, in 1953, SoisalonSoininen had published another short essay in Finnish titled Septuagintan kielenkäyttö ‘The use of language in the Septuagint.’ He discusses in this earlier essay the fact that the Septuagint includes a particular kind of Greek. He says that the nature of the Greek in the Septuagint as a translationese is understandable for several different reasons. It is difficult for any translator to avoid certain influences of the source language to the target text. Moreover, the fact that Hebrew and Greek belong to two entirely different language families, made the task for the Septuagint translators particularly difficult.13 There is yet another fact that according to Soisalon-Soininen had a major influence on the language the translators used. This concerns the religion of the translators. The translators were Jews who still could understand Hebrew and Aramaic. They knew their Old Testament in Hebrew. Soisalon-Soininen then concludes that the religious thinking of the translators was deeply rooted in the Hebrew text and therefore the Hebrew way of addressing the religious ideas and texts must have been natural to them.14 The Hebrew influence, thus, came almost automatically. That this was natural to the translators does not mean that it would have been natural to their readers, too. On the contrary, Soisalon-Soininen claims that for its ancient readers the Greek text must have appeared as quite particular religious language that preserved the ‘Hebrew spirit’ of the Old Testament.15 I am not quite certain that the background of the translators would this easily explain the nature of the translation. The Greek we encounter in the Septuagint, in those parts that are translations, that is, does include a Semitic colouring, but I think that the problem is more complicated than Soisalon-Soininen in the early 1950’s thought. It is not at all certain that the translators always used peculiar constructions in their texts. On the contrary, sometimes they do, sometimes they do not. In any case, in 1963 he repeats his earlier view on the main factors influencing the translation and especially the syntax of a translation: the syntax of the source language, the syntax of the target language, and the translator, who decides how to take the requirements of each language into account. Now he says that the main factor is the translator and this is why the study of the syntax becomes a study of the translation technique.16 Because the translator decided the wording 13  I. Soisalon-Soininen, “Septuagintan kielenkäyttö”, in Talenta quinque: Commentationes in honorem Ilmari Salomies, Eino Sormunen, E. G. Gulin, Rafael Gyllenberg, G. O. Rosenqvist (Helsinki: Otava, 1953), 46–52, on p. 47. Being a child of his own age Soisalon-Soininen also spoke about the different races and their different psychologies. 14 Soisalon-Soininen, “Septuagintan kielenkäyttö”, 47. 15 Soisalon-Soininen, “Septuagintan kielenkäyttö”, 50–51. 16 Soisalon-Soininen, “Septuagintan lauseopin”, 226–7. There are areas of the syntax, where the source language has no effect, he continues. He mentions especially verbs, both moods and tenses of them.

32

Seppo Sipilä

of the Greek text, the way he did this became the most important factor that influenced the translation. As an example on how to study the syntax, Soisalon-Soininen turns to the infinitive that he was interested in while writing this essay. He says that firstly one has to collect all the cases where an infinitive is used. Then one must organize the material according to the Hebrew. How did the writers of the Hebrew text use the infinitive and how are these different usages handled in the translation process? Then, one must base a more detailed organising of the material on the challenge— the feature in question—set to the Greek translator.17 As this crucial point in the analysis, he gives an explanation of what he means. Not every Hebrew infinitive set the same kind of challenge to the translator. In Gen 2:4 the Hebrew infinitive is translated by using a relative clause (‫ביום‬ ‫—עׂשות יהוה אלהים ארץ וׁשמים‬ᾗ ἡμέρᾳ ἐποίησεν ὁ θεὸς τὸν οὐρανὸν καὶ τὴν γῆν). The formal translation by using a Greek infinitive (*ἐν τῆ ἡμέρα ποιεῖν/ποιῆσαι ὁ θεός…) would be out of the question, because the Hebrew infinitive is used as a (genitive) attribute. The same applies to Isa 58:3 where the Hebrew ‫הן ביום‬ ‫ צמכם‬is translated into Greek as ἐν γὰρ ταῖς ἡμέραις τῶν νηστειῶν ὑμῶν.18 However, in Num 14:3 the Greek infinitive in translating ‫ טוב לנו ׁשוב מצרימה‬using the words βέλτιον ἡμῖν ἐστιν ἀποστραφῆναι εἰς Αἴγυπτον is unproblematic, because the Hebrew infinitive is the grammatical subject of the expression. Cases where the Hebrew infinitive acts as the subject are not all equal, however. If there is a modifier attached to the infinitive, the Greek infinitive is not the best rendering for the Hebrew one. This is the case in 2 Sam 15:20 ‫—תמול בואך‬εἰ ἐχθὲς παραγέγονας (pro *ἐχθὲς παραγενέσθαι σέ) where the literal translator avoided using the infinitive in Greek.19 These examples illustrate the thinking behind SoisalonSoininen’s methodology. He starts with the assumption that a formal literal translation is what the ancient translators aimed at, but could not always use, because of the differences between Hebrew and Greek. It is very important that scholars take these differences into account. That is to say, one must first collect all the relevant examples in the Hebrew text with their Greek translations. Then one must organize the material according to the challenge the Hebrew formulation in question sets for the translator. This will reveal the various ways the translators have represented the Hebrew text. These various ways one can then compare to the contemporary Greek texts, the apocrypha and the New Testament. The comparison will show the peculiarities of

17 Soisalon-Soininen, “Septuagintan lauseopin”, 228–9. 18  On the Hebrew infinitive, see e.g. HALOT. One could, in principle understand ‫ צום‬as a noun, too. 19 Soisalon-Soininen, “Septuagintan lauseopin”, 229.

Soisalon-Soininen Meets Grice

33

the translationese used in the Septuagint. Once one has completed this part, she or he can proceed and compare individual translators to one another.20 Soisalon-Soininen’s magnum opus Die Infinitive in der Septuaginta is a report on a translation technical study where Soisalon-Soininen employed his methodology. His starting point is the question how various translators handled the Hebrew infinitive. While speaking about the infinitive (die Infinitive) he excludes the Hebrew infinitive absolute. This is because the Hebrew infinitive absolute deviates strongly from the Greek infinitive.21 I do not think he actually says this, but the crucial issue is that the difference is based on the literal translation. The Greek infinitive can be used as a literal translation for the Hebrew infinitive construct, but not so easily for the Hebrew infinitive absolute.22 The occurrences where the Hebrew infinitive is attached to the preposition ‫ל‬ give us a good illustration about Soisalon-Soininen’s methodology. He says that adverbial constructions form the largest group of infinitives with ‫ ל‬in Hebrew. The Hebrew adverbial expression with the infinitive operates much the same way than the infinitive without any preposition. Translating the Hebrew infinitive with the Greek one is in many cases quite natural.23 However, when the infinitive + ‫ ל‬is used adnominally the difference to the usage of the infinitive without a preposition becomes visible. These different cases then set a different challenge for the translator and must therefore be treated separately. The challenge here means the difficulty in using a literal translation. Thus, Soisalon-Soininen divides the cases into three groups: ‘independent’, adnominal, and adverbial ones. Independent cases include occurrences where the Hebrew infinitive with ‫ ל‬acts as the subject or the predicate. Adverbial cases he divides further into two groups, those where the Hebrew expression acts as a direct object and those where the expression acts as an adverbial proper.24 The most common rendering for the Hebrew infinitive + ‫ ל‬used by the translators is the Greek infinitive with the article τοῦ, the ‘genitive infinitive’, as SoisalonSoininen calls it. There are, however, some differences between the groups and their subgroups when it comes to the use of various possibilities. Final subordinate clauses, for instance, are used only when the Hebrew construction also has a final function.25

20 Soisalon-Soininen, “Septuagintan lauseopin”, 229. 21 Soisalon-Soininen, Die Infinitive, 16 n. 1. 22 On the differences between the two infinitives in Hebrew, see e.g. B.K. Waltke/M. O’Connor, An Introduction to Biblical Hebrew Syntax (Winona Lake: Eisenbrauns, 1990) § 35.2.2. 23 Soisalon-Soininen, Die Infinitive, 20. 24 Soisalon-Soininen, Die Infinitive, 27–28. 25 Soisalon-Soininen, Die Infinitive, 75.

34

Seppo Sipilä

Soisalon-Soininen names the Greek infinitive with τοῦ as a formal rendering (mechanisch wortgetreu) of the Hebrew infinitive with the preposition ‫ל‬,26 and uses it as one of the means to compare different translators to one another. He sets the use of the Greek infinitive with τοῦ against the other possibilities as the rendering of Hebrew constructions with the infinitive.27 The differences between various translations becomes visible especially when the Greek infinitive used in a final function is a rendering for the Hebrew infinitive with ‫ל‬. When the number of cases of the Greek infinitive with τοῦ is compared to the number of renderings either with the plain Greek infinitive or with the infinitive with ὥστε, several groups of translations emerge. The statistical differences are not only numbers, but they tell us how formal a translation is. Hence, when the Greek article τοῦ appears as a rendering of the Hebrew preposition ‫ ל‬the result is a formal rendering. When Greek article τοῦ does not appear the result is a less formal rendering.

2.3 Why This Method? The translation technical method starts from the general well-known observation that the Greek language we encounter in the Septuagint is translationese. The method seeks to identify the different translation possibilities actually used and separate the formal renderings from the non-formal ones.28 At the same time, the method allows us to describe various translators especially by comparing them to one another. Such a comparison is not only interesting and revealing but it is actually the ethical duty of scholars working in this field. Any generalization concerning the Septuagint must be set vis-à-vis the fact that the Septuagint is a collection of different translations that were composed at different times by different people. The study of the syntax is no exception to this. For Soisalon-Soininen the method was intended originally to find answers to three questions: a) How does the Greek we encounter in the Septuagint differ from contemporary Koine? b) How can we compare different translators to one another?, and c) What are the Hebraic expressions, the so-called Hebraisms, that give the Septuagint texts their Semitic colouring?29 26 Soisalon-Soininen, Die Infinitive, 187. 27 Soisalon-Soininen, Die Infinitive, 179–86. 28 Muraoka, A Syntax, xl: “I … entertain no shadow of doubt that such an approach [= the method by Soisalon-Soininen] is capable in revealing many important facets of the syntax of SG [= Septuagint Greek]”. 29 Soisalon-Soininen, Die Infinitive, 12. In his last published work on the Septuagint, titled as Zurück zur Hebraismenfrage, he deals with the third question. His conclusion is that it is not easy to say exactly when we can label a formal rendering met in the LXX as a Hebraism. This is because a formal rendering can also be an idiomatic Greek expression. At the end of the day, the Hebraism looks like a phenomenon that only the context of an expression can reveal (I. Soisa­ lon-Soininen, “Zurück zur Hebraismenfrage” in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien

Soisalon-Soininen Meets Grice

35

2.4 Later Development of the Method Since the publication of Soisalon-Soininen’s study on the infinitive, two important additions have been made to the methodology. Whereas Soisalon-Soininen originally started from a dichotomy between the formal and free translations, Sollamo in her study of the semiprepositions made a remarkable contribution to the analysis by dividing the Greek expressions used into three groups: slavish, literal, and free.30 The helpful category ‘slavish’ is due to the special nature of her material, but it beautifully illustrates the difficulty in approaching the peculiar language of the Septuagint: the closer you look at it, the more complicated the challenge appears. Aejmelaeus in her studies on the translations of the Hebrew particle ‫ כי‬made the other important methodological improvement. As the Hebrew conjunction has a variety of usages or meanings, there can hardly be a literal translation of it in the same way than the Greek infinitive operates as a literal translation of the Hebrew infinitive construct. The Greek rendering for ‫ כי‬is very often the conjunction ὅτι, but we cannot designate ὅτι as a literal translation of ‫כי‬.31 The translation of a multi-purpose conjunction always requires consideration of the context where the conjunction is used.32 The same then holds true in case of the methodology. For the analysis one must first clearly separate different usages of the conjunction before going any further. Despite the methodological development, the basis of the translation-technical method has remained unchanged. Later contributions have been refinements sharpening and defining more clearly how to use the method. Originally, Soisalon-Soininen based the method on certain assumptions on the way the translators worked. He pointed out the religious and linguistic background of the translators as an explanation and why they acted in a way they did. Since 1950’s many things have changed in our understanding on how the human mind works, on how languages operate and on how human beings communicate. I will therefore turn to these issues in the second part of my paper.

zur Septuaginta. Robert Hanhart zu Ehren (Abhandlungen der Akademie der Wissenschaften in Göttingen / MSU 20; Göttingen: Vandenhoeck & Ruprecht, 1990) 35–51, on pp. 38–39). 30  See e.g. R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (AASF, Diss. Hum. Litt. 19; Helsinki: Suomalainen Tiedeakatemia, 1979), 69 on translations of ‫ לפני‬referring to living beings. 31  A. Aejmelaeus, “The Significance of Clause Connectors in the Syntactical and Translation-Technical Study of the Septuagint” in On the Trail of the Septuagint Translators: Collected Essays (Kampen: Kok Pharos, 1993), 43–58, on p. 49. 32 Aejmelaeus “The Significance”, 57. Cf. also S. Sipilä, Between Literalness and Freedom: Translation technique in the Septuagint of Joshua and Judges regarding the clause connections introduced by ‫ ו‬and ‫( כי‬SESJ 75; Helsinki: The Finnish Exegetical Society, 1999), 141.

36

Seppo Sipilä

3. Grice’s Cooperative Principle During 1960’s when Soisalon-Soininen worked on the infinitive in the Septuagint, an important development in linguistics took place. This development had far-reaching consequences not only by causing a paradigm shift in traditional linguistics, but also by leading into the development of new fields of study. One of the new fields that began to emerge was pragmatics. Behind pragmatics was a recognition that the social setting of human beings have an important part in the way the language is used. Thus, the focus of the study shifted out of the formal study and use of syntax as describing the ‘rules of the language’.33 One of the figures that had a part in this development was a British language philosopher Herbert Paul Grice. Grice created the Cooperative Principle of communication in order to explain how humans communicate in a successful way while having an informal conversation. An informal conversation means here an ordinary discussion that does not follow the rules of formal logic. The principle has since gained wide acceptance in social science and in linguistics especially. Grice approaches the conversation as a formal logician. His thinking was based on the obvious observation that people in their ordinary conversation used language in illogical—hence his title ‘Logic and Conversation’—or inappropriate ways. Nevertheless, we may suppose that such behaviour is rational, as he puts it.34 According to Grice this is only possible because of several general conditions. The article ‘Logic and Conversation’ based on his lectures in 1967 makes an inquiry into these general conditions.35 Grice’s starting point is an observation that people during a regular conversation often say more than just the conventional meaning of words they use. Suppose I am in London and someone asks me: “Can you tell me where the bus to Heathrow airport leaves from?” I might well answer: “Sorry, I am a stranger here myself.” Assuming that from this response the enquirer could understand that I do not know the where the bus leaves from.36 People often speak like this and imply things. Grice is especially interested in those implicated things that help the hearer to work out from the way something was said what the speaker meant. He labelled this kind of implicature as a ‘conversional implicature’. Grice proceeds in claiming that parts of the implied parts of the communication are connected with certain general features of any conversation. The most important of all of these is the fact that when people talk to each other they follow 33  J.L. Mey, Pragmatics: An Introduction (Oxford: Blackwell, 2001), 3–4. 34  P. Grice, Studies in the Way of Words (Cambridge: Harvard University Press, 1989), 20. 35  H.P. Grice, “Logic and Conversation”, in P. Cole/J. Morgan (ed.), Speech Acts (New York: Academic Press, 1975), 41–58, on p. 43. 36 Example is based on E.-A. Gutt, Translation and Relevance: Cognition and Context (2nd ed.; Manchester: St. Jerome, 2000), 25.

Soisalon-Soininen Meets Grice

37

‘a mutually accepted direction’. Hence, he suggests that there is a general principle that people follow when engaging in conversation: “Make your conversational contribution such as is required, at the stage at which it occurs, by accepted purpose or direction of the talk exchange in which you are engaged.” This general principle he labels as the ‘Cooperative Principle.’37 The principle will lead into several maxims connected with quantity, quality, relation, and manner. These technical expressions refer to optimum amount of information, truthfulness of it, relevance of it to the talk exchange, and clarity of the utterance. Once we think about the situations where we failed in making our point understood, the principle seems to work well. Failure is often a result of breaking the principle. If we fail to fulfil a maxim, but not deliberately, we will mislead others. Sometimes we might fail to fulfil a maxim because there is a clash between them, because we might have to sacrifice one in order to fulfil another. We may also chose not to follow any of them for a reason.38 This is why, for instance, lying is successful: people suppose that we always speak the truth. Grice himself gave a famous example on how people may violate the principle in order to imply something they might not want to put blatantly. The example is about a teacher writing about a student who is a candidate for a philosophy job. This is what the teacher writes: Dear Sir, Mr. X’s command of English is excellent, and his attendance at tutorials has been regular. Yours etc.

Using the maxims, we may conclude the following. The teacher is cooperative, because he has taken the trouble of writing in the first place. Thus, we may expect him to follow the principle. He knows the student, so what he says is not because of ignorance, but he also knows that more information than this is wanted. Thus, we may conclude that there is a reason to be this brief in his letter. This reason lies in the teacher’s reluctance to say what he actually wants to say (the candidate is no good at philosophy).39 Grice does not claim that we always follow the principle or the maxims purposefully and consciously. Quite the contrary, much of the application is subconscious. The principle describes the cognitive processes, when we participate in a conversation. Grice does not even claim that the maxims he lists in his article are exactly those and only those that are meaningful here. His is also aware that the Cooperative Principle applies only in situations where “a maximally effective exchange of information” is the main goal.40 37 Grice, “Logic and Conversation”, 45. 38 Grice, “Logic and Conversation”, 49. 39 Grice, “Logic and Conversation”, 52. 40 Grice, “Logic and Conversation”, 46–47.

38

Seppo Sipilä

4. The Cooperative Principle and General Communication Theory Grice in his lectures and in his article starts discussing the principles of mutual talking from the point of view of the speaker. The power of Grice’s principle lies in two additional facts. While Grice starts from the speaker and the principle according to which the speaker has to formulate the utterance, one can always apply the principle to the hearer. In that case, the maxims describe the expectations of the hearer, the way the hearer approaches what is being said. A hearer expects that the speaker intends to express her or his ideas using good quality language and suitable amount of information in such a way that the expression becomes relevant and understandable. The principle and the maxims are not limited to spoken communication only. It seems that the same principle governs almost any communication between human beings. Yet another observation is in order. Grice’s Cooperative Principle presupposes strong ties between the individual acts of communication and the situation under which the communication takes place. The situation helps us to apply the principle in a successful way. The particular situation tells us, what the optimal amount of information is and how to express that in order to avoid any ambiguity helping the communication towards the accepted direction.41 Since one can see translation as an act of communication, one can conclude that in the case of translation the very situation, where it takes place, does influence the way the translation is done. The translator follows Grice’s principle and does so at the particular situation he or she is working in. We apply the Cooperative Principle based on assumptions. We assume that the others are following the principle. Based on the assumptions, we can calculate, as Grice puts it, what others are trying to say.42 This calculus we link with the situation and our knowledge about the situation in question. When the principle was later developed further, it became evident that our idea about the situation is a psychological construct, a kind of cognitive model based on our assumptions about the immediate environment but also about the world.43 In other words, the hearer will conclude what the speaker is saying by using various mental tools and resources at her or his disposal.

41 Grice, “Logic and Conversation”, 56. 42 Grice, “Logic and Conversation”, 57–58. 43  D. Sperber/D. Wilson, Relevance: Communication and Cognition (Oxford: Blackwell, 1986), 15–16.

Soisalon-Soininen Meets Grice

39

5. The Cooperative Principle and the Septuagint This approach should not be alien to us, who try to read and understand the Septuagint. We have always struggled, I suppose, with the elementary challenges connected to the Greek texts, the words and structures the translators chose to employ. Since our aim is to understand what has happened, we seldom conclude that the way the translators worked was simply erroneous, illogical and unacceptable.44 Instead, we—at least at some level—try to start from Grice’s observation: the translators used Greek in “illogical” or “inappropriate” but still rational ways. Grice’s principle does not help us solve our challenge with the Greek text, but the principle makes it more understandable: the challenge is connected to our interpretational processes while reading the Greek text. These processes are based on our assumptions about the environment of the translations. Grice’s principle even claims that the translators followed some general principles. Even though our exact knowledge of the translation situation is limited, we may nevertheless safely assume something about the way the translators proceeded in their task. Starting from the assumption that the translators tried to be rational in their work, we may then say that they expressed the contents of the Hebrew (or Aramaic) text in front of them as well as they could. Despite the fact that there are passages that were difficult for them, or as Soisalon-Soininen put it, ‘there were passages in front of which the translators must have felt helplessness,’ one should not overestimate the difficulty. As far as I can see, the Greek translators understood most of the Biblical text without any problem. The Cooperative Principle would then mean this: the ancient translators when conducting the translation work made assumptions as to what their audience was expecting when it comes to translating an important text. I think we would not err much in assuming that the translators tried to be as informative as necessary. Very often this meant following Hebrew constructions as closely as possible, but not always. Consider, e.g., Josh 4:9 ‫—ושתים עשרה אבנים הקים יהושע בתוך הירדן‬ ἔστησεν δὲ Ἰησοῦς καὶ ἄλλους δώδεκα λίθους ἐν αὐτῷ τῷ Ιορδάνῃ where the translator “added” the expression καὶ ἄλλους and the pronoun αὐτῷ obviously in order to clarify the meaning of the passage.45 We may also assume that the translators did not try to lead the reader astray by intentionally distorting the meaning 44 As far as I can see, most Septuagint scholars would agree with M. Harl, “La Bible d’Alexandrie”, in B.A. Taylor (ed.), X Congress of the International Organization for Septuagint and Cognate Studies: Oslo, 1998 (SBLSCS 51; Atlanta, GA: Scholars Press, 2001), 187: “Close observation of the translators’ work convinces us that they were competent and conscientious.” 45  S. Sipilä, Jordanin ylityskertomus. Septuagintan ja Masoreettisen tekstin väliset erot (M. Th. thesis, Helsingin yliopisto, 1988), 91–92, cf. C.G. den Hertog, Studien zur griechischen Übersetzung des Buches Josua (PhD diss., Justus-Liebig Universität, 1996), 158 or A.G. Auld, J­oshua: Jesus Son of Nauē in Codex Vaticanus (Septuagint Commentary Series; Leiden: Brill, 2005), 115–16.

40

Seppo Sipilä

of the text. We may further conclude that any translator who is translating an important text naturally thinks that his or her work is relevant to the audience, simply because the text by definition is relevant. The concept of clarity though is a little different. Brief and orderly the translations may well be, but they do not always avoid ambiguity or obscurity of expression. Soisalon-Soininen in his early essay suggested that the reason lies in the translators’ familiarity on the Hebrew idiom in the Old Testament. The Semitic way of expressing ideas matched well with their way of thinking.46 Today, I believe, we would explain such peculiarities by the translation process rather by the translator’s linguistic background. This is because there seems to be no serious weakness in the translators’ knowledge of Greek. I will now suggest that the reason for using the formal way of translating lies in the Grice’s idea of implicature. Grice claims that all humans base their communication on implication. We express only what we think is needed. Nothing more, nothing less. The rest of the meaning we simply imply counting on the knowledge of the person with whom we are talking. In deciding what is needed, we apply the Cooperative Principle. After Grice created and published the Cooperative Principle, we have learned that much of the decision making in applying the Cooperative Principle is based on trying to achieve successful communication by using only as little effort as necessary.47 The labour in reducing the effort naturally leads into employing standard renderings and an easy technique.48 Let us consider the use of the Greek conjunction καί as an example. καί was used as the standard rendering of the Hebrew conjunction ‫ ו‬throughout the Septuagint just because it was suitable in many occasions. καί has the widest applicability in this case, as Aejmelaeus has put it.49 When using καί to render the Hebrew conjunction ‫ו‬, the translator had an effortless job and he could trust that it would be as easy for his audience to interpret the text correctly.

46 Soisalon-Soininen, “Septuagintan kielenkäyttö”, 47. 47 Cf. Sperber/Wilson, Relevance, 155–61 and D. Wilson/D. Sperber, “Pragmatics and Time” in R. Carston/S. Uchida (ed.), Relevance Theory: Applications and Implications (Pragmatics & Beyond New Series 37; Amsterdam: John Benjamins, 1998), 8–9. 48  The term “an easy technique” was coined by James Barr in his seminal work on the literalism, see especially J. Barr, The Typology of Literalism in Ancient Biblical Translations (MSU XV; Göttingen: Vandenhoeck & Ruprecht, 1979), 26. 49  A. Aejmelaeus, Parataxis in the Septuagint: A Study of the Renderings of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASF, Diss.hum.litt. 31; Helsinki: Suomalainen Tiedeakatemia, 1982), 29–30.

Soisalon-Soininen Meets Grice

41

6. Conclusions Soisalon-Soininen when developing the translation technical method was a child of his own time. However, he also was able to take a fresh view on how to study the language and especially the syntax of the Septuagint. By developing the translation technical method, he broke the barriers of traditional theoretical study of syntax being clearly ahead of his own time. As he could not have the same tools and ideas that we now have, some of his suggestions did not withstand the ravages of time. If we can accept Grice’s way of thinking, the consequences for the translation technical method are clear. The use of formal translations that forms a basis to the translation technical method becomes readily understandable. The translators used formal translations because it required less effort. The use of the literal translations was not only easy for them but also light in terms of their cognitive processing effort. As the translators naturally looked for an effortless way of translating, wherever they found the literal translation as a possible candidate, they needed nothing else. The use of literal translations creates, as a side product, the Semitic colouring and the Hebraisms. I do not think that translators aimed at translating formally, but they chose those particular renderings in the same way that we process communication and that led to those formal results. We can abandon Soisalon-Soininen’s early explanation about the Hebrew mind-set as an explanation for the literalness of the Septuagint. It is outdated and today we have a better explanation for the phenomenon. The method itself is still valid. When the translators did abandon their literal translations, the reason lies in the same cognitive process, too. The translator calculated in those cases that he could not use the literal translation. The literal translation would in that case distort the communication too much. It would make the communication too demanding for the audience to process. What, of course, is surprising and remains as such, are the differences between the translators. Some translators seem to calculate the effort of the process differently than their colleagues. This may be due to the result of segmentation. Narrow segmentation makes it difficult for a translator to estimate the amount of processing effort. If one keeps only a minimal amount of text actively in mind, one cannot notice the consequences to the overall meaning or clarity. However, I do not think that segmentation is the only explanation for the differences between the translators.50 It is clear that a translator, ancient or modern, will handle a common syntactic issue quicker and with less effort than a rare one. Acceptable literal translations will eventually become second nature to the translator when dealing with common issues. It may also be that the translators calculated the effort in different ways, or should I say, some of them were less interested in reducing the effort of their readers than others. This is true for all human beings. 50  Cf. Sipilä, Literalness and Freedom, 207–9.

Raija Sollamo The Usage of the Article with Nouns Defined by a Nominal or Pronominal Genitive in LXX Genesis 1. Introduction The usage of the article is an example of grammatical differences between H ­ ebrew and Greek. In Greek, the noun is usually understood as defined if it is connected with a genitive, be it a noun or pronoun in the genitive, and this understanding is marked by the use of the definite article.1 In Hebrew, the noun is also understood as defined if it is in the status constructus form before another noun in the status absolutus, which is the way the genitive is expressed in Hebrew. But o ­ pposite to the Greek practice, the Hebrew locution leaves the article out with the understanding that the definition is self-evident.2 One does not need to mark it with an article unless in connection with an adjective attribute. The main focus of my paper is not, however, on the contrastive grammar, even though it might be very interesting. Rather, it is the translation technique of the Genesis translator that comes to the fore. The question is whether the translator followed the Greek grammatical rules in the usage of the definite article or whether he worked more mechanically on the basis of the Hebrew and left the article out whenever there was no article in the Hebrew Vorlage. Because the translation technique is my final focus, I also include those very freely rendered cases that display no genitive in the Greek and accordingly, the renderings do not actually belong under the title “The Usage of the Definite Article with Nouns Defined by the Nominal or Pronominal Genitive in Greek”. This concerns my last section “Free Renderings without the Genitive”. Altogether, the number of cases to be examined is huge. Therefore, I had to confine this paper to the Greek Genesis. In order to interpret the translator’s working habits and thinking correctly, Muraoka’s new Syntax of Septuagint Greek is a helpful tool in providing some categories of the usage of the article in the Septuagint on the basis of Greek.3 1  Schwyzer, E./Debrunner A., Griechische Grammatik II: Syntax und syntaktische Stilistik (München: C.H. Beck, 1950), 19–27; J. Blomqvist/P.O. Jastrup, Grekisk/graesk grammatik (Odense: Akademisk Forlag, 1990), § 224–5 (pp. 169–71). 2  E. Kautzsch, ed., Gesenius’ Hebrew Grammar (2nd ed.; Oxford: Clarendon Press, 1990), § 127a (p. 410): “In a construct chain, the construct can never be prefixed with the article. The definiteness of the genitive specifies the definiteness of the phrase;” Waltke B.K./O’Connor M. P., An Introduction to Biblical Hebrew Syntax (Winona Lake, Indiana: Eisenbrauns, 1990), § 9.7 (pp. 156–7). 3  T. Muraoka, A Syntax of Septuagint Greek (Leuven/Paris/Bristol, CT: Peeters, 2016).

44

Raija Sollamo

2. The Definite Article The most common practice the Genesis translator follows is to use the definite article before a noun which is followed by a nominal or pronominal genitive.4 This happens in circa 1940 out of 2470 cases (78.5%) in the Greek Genesis. The examples are of various kinds and there seems to be no difference for the translator whether the Hebrew noun is determined by another noun in the genitive or by a pronominal suffix. Examples: Gen 3:24 ‫ת־ּד ֶרְך ֵעץ ַה ַחּיִ ים‬ ֶ ‫— ִל ְׁשמֹר ֶא‬φυλάσσειν τὴν ὁδὸν τοῦ ξύλου τῆς ζωῆς. Gen 9:12 ‫ר־אנִ י נ ֵֹתן‬ ֲ ‫אֹות־ה ְּב ִרית ֲא ֶׁש‬ ַ ‫—זֹאת‬τοῦτο τὸ σημεῖον τῆς διαθήκης ὃ ἐγὼ ­δίδωμι. The definition of τὸ σημεῖον is reinforced by both the genitive and the relative clause. Gen 9:22 ‫י־א ָחיו‬ ֶ ֵ‫—וַ ּיַ ְרא ָחם ֲא ִבי ְכנַ ַען ֵאת ֶע ְרוַ ת ָא ִביו וַ ּיַ ּגֵ ד ִל ְׁשנ‬καὶ εἶδεν Χαμ ὁ πατὴρ ­Χανααν τὴν γύμνωσιν τοῦ πατρὸς αὐτοῦ καὶ ἐξελθὼν ἀνήγγειλεν τοῖς δυσὶν ἀδελφοῖς αὐτοῦ. Gen 14:13 ‫—וְ הּוא ׁש ֵֹכן ְּב ֵאֹלנֵ י ַמ ְמ ֵרא‬αὐτὸς δὲ κατῴκει πρὸς τῇ δρυὶ τῇ Μαμβρη. The usual Classical Greek practice of repeating the article of the regens before a genitive attribute5 seldom appears in LXX Genesis, mostly in connection with the genitives of Hebrew proper names (see also Gen 14:1). Gen 20:11 ‫ל־ּד ַבר ִא ְׁש ִּתי‬ ְ ‫— ַע‬ἕνεκεν τῆς γυναικός μου. Gen 21:33 ‫— ִּב ְב ֵאר ָׁש ַבע‬ἐπὶ τῷ φρέατι τοῦ ὅρκου Gen 23:16 ‫י־חת‬ ֵ ֵ‫ת־ה ֶּכ ֶסף ֲא ֶׁשר ִּד ֶּבר ְּב ָאזְ נֵ י ְבנ‬ ַ ‫— ֶא‬τὸ ἀργύριον ὃ ἐλάλησεν εἰς τὰ ὦτα τῶν υἱῶν Χετ. Gen 31:3 ‫מֹול ְד ֶּתָך‬ ַ ‫ּול‬ ְ ‫בֹותיָך‬ ֶ ‫ל־א ֶרץ ֲא‬ ֶ ‫—ׁשּוב ֶא‬ἀποστρέφου εἰς τὴν γῆν τοῦ πατρός σου καὶ εἰς τὴν γενεάν σου. Gen 38:19 ‫נּותּה‬ ָ ‫—וַ ִּת ְל ַּבׁש ִּבגְ ֵדי ַא ְל ְמ‬καὶ ἐνεδύσατο τὰ ἱμάτια τῆς χηρεύσεως αὐτῆς. Gen 39:8 ‫ל־א ֶׁשת ֲאד ֹנָ יו‬ ֵ ‫אמר ֶא‬ ֶ ֹ ‫—וּי‬εἶπεν δὲ τῇ γυναικὶ τοῦ κυρίου αὐτοῦ.

The use of the definite article with a noun determined by another noun in the genitive or by a possessive pronoun (or another pronoun) shows the translator’s good competence in Koine Greek. His competence also guides him in instances where the parent text has a shorter or lengthy chain of coordinate expressions with the same suffix. He then understands that he does not have to repeat the

4  This is the normal word order. The usual practice of using the definite article does not change in cases where the genitive is placed before the noun. In these few cases the genitive is situated either before the arthrous noun or between the article and the noun. With the genitive of the relative pronoun the article and the noun follow the relative pronoun. 5  Schwyzer/Debrunner, II, B, 1 f γ (p. 26).

The Usage of the Article with Nouns Defined by a Genitive

45

possessive pronoun with every item, but the definite article suffices [e.g. Gen 32:23(22) cited at the beginning of Special Cases below].6 It has also been suggested that the translator adhered to the length of the Hebrew expression in that he easily used the definite article in instances where the Hebrew has the nota accusativi or a preposition, such as ‫ל‬,ְ before the noun in the status constructus. This practice was characteristic of Aquila.7 To form an opinion I have haphazardly checked 700 cases where the definite article appears. In them the nota accusativi is found 102 times and ‫ ְל‬53 times. Thus, the cases with ‫ ֶאת‬form only 14.6% of all the instances with the definite article in my checked 700 cases. Since the correspondence of the other prepositions with the use of the definite article is minor, we can safely conclude that for the Genesis translator the Hebrew nota accusativi or basic prepositions did not play any significant role for the use of the definite article in his Greek translation. The real reason for the usage of the definite article by the translator seems to be his good command of Greek. He shows no endeavour in the direction of a slavish translation in the style of Aquila.

Special cases Gen 32:23 (22) ‫ת־א ַחד ָע ָׂשר יְ ָל ָדיו‬ ַ ‫ת־ׁש ֵּתי ִׁש ְפח ָֹתיו וְ ֶא‬ ְ ‫ת־ׁש ֵּתי נָ ָׁשיו וְ ֶא‬ ְ ‫—וַ ּיִ ַּקח ֶא‬ἔλαβεν τὰς δύο γυναῖκας καὶ τὰς δύο παιδίσκας καὶ τὰ ἕνδεκα παιδία αὐτοῦ.

The genitive αὐτοῦ determines all the items, even though it is not repeated. Gen 32:10 ‫מֹול ְד ְּתָך‬ ַ ‫ּול‬ ְ ‫—ׁשּוב ְל ַא ְר ְצָך‬ἀπότρεχε εἰς τὴν γῆν τῆς γενέσεώς σου (freely). Gen 22:16 ‫ת־ּבנְ ָך ֶאת־יְ ִח ֶידָך‬ ִ ‫—וְ לֹא ָח ַׂש ְכ ָּת ֶא‬καὶ οὐκ ἐφείσω τοῦ υἱοῦ σου τοῦ ­ἀγαπητοῦ.

Instead of the noun ‫את־יְ ִח ֶידָך‬, ֶ an adjective attribute is used by the translator. The result is excellent Greek and corresponds well to the meaning of the Hebrew clause. Gen 25:24 ‫יה ָל ֶל ֶדת‬ ָ ‫יָמ‬ ֶ ‫—וַ ּיִ ְמ ְלאּו‬καὶ ἐπληρώθησαν αἱ ἡμέραι τοῦ τεκεῖν αὐτήν.

A fine Greek rendering with a genitive of an infinitive. Another example appears in Gen 2:9. 6  R. Sollamo, Repetition of the Possessive Pronouns in the Septuagint (SCS 40; Atlanta, ­Georgia: Scholars Press, 1995), 22–9. 7 I. Soisalon-Soininen, “Einige Merkmale der Ûbersetzungsweise von Aquila”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 19–27. Soisalon-Soininen writes: “Wenn im Hebräischen kein Artikel steht, wird er auch bei Aquila weggelassen, wenn nicht ein anderes hebräisches Wörtchen (nota acc. oder ‫)ל‬ ְ mit ihm wiedergegeben ist” (p. 21).

46

Raija Sollamo

Gen 28:12 ‫יְמה‬ ָ ‫— ִהּנֵ ה ֻס ָּלם ֻמ ָּצב ַא ְר ָצה וְ רֹאׁשֹו ַמּגִ ַיע ַה ָּׁש ָמ‬καὶ ἰδοὺ κλίμαξ ἐστηριγμένη ἐν τῇ γῇ ἧς ἡ κεφαλὴ ἀφικνεῖτο εἰς τὸν οὐρανόν.

This is an idiomatic Greek translation, the relative pronoun in the genitive referring to the ladder. Furthermore, it changes the Hebrew parataxis into a more refined Greek clause structure. These and the other special cases attest the freedom of the Genesis translator as regards his parent text but they also demonstrate his great competence in producing good Koine Greek. If we evaluate the possible Hebraic flavor or Hebrew interference in his translation, we could direct our attention to the abundant use of the possessive pronouns, i.e. the genitives of the personal pronouns and the genitives of the third person pronouns (αὐτοῦ, αὐτῶν) in particular.

3. No Article with the Noun Defined by the Genitive The second largest group consists of instances where the noun determined by a nominal or pronominal genitive remains without an article in the Greek translation. The structure strictly adheres to the Hebrew expression of the genitive where the noun in status constructus (the regens) does not have an article. These cases appear 530 times (21.5%) in the Greek Genesis. As the use of the definite article is overwhelming (78.5%) with the nouns defined by a nominal or pronominal genitive, the omission of the article is a marked alternative according to modern linguistics. Even though the translator did not master modern linguistics, it is reasonable to suppose that the omission of the article was a choice that had good reasons in his translating. Of course, the absence of an article now and then might be a slip, due to the influence of the Hebrew parent text (interference). Nevertheless, in most cases the translator’s thinking and his language ­ability come clearly to the fore. This can be concluded from the groups of instances following the same anarthrous translational patterns. A very marked rule the translator follows is the usual Greek practice of leaving the predicate of a descriptive nominal clause anarthrous if it is contextually indeterminate.8 There appear a number of instances of this phenomenon, such as: Gen 4:2 ‫י־ה ֶבל ר ֵֹעה צֹאן‬ ֶ ‫—וַ יְ ִה‬καὶ ἐγένετο Αβελ ποιμὴν προβάτων. Gen 14:18 ‫—וְ הּוא כ ֵֹהן ְל ֵאל ֶע ְליֹון‬ἦν δὲ ἱερεὺς τοῦ θεοῦ τοῦ ὑψίστου. Gen 36:30 ‫ּלּופי ַהח ִֹרי‬ ֵ ‫— ֵא ֶּלה ַא‬οὗτοι ἡγεμόνες Χορρι. Gen 46:1 ‫—וְ ֵא ֶּלה ְּבנֵ י זִ ְל ָּפה‬οὗτοι υἱοὶ Ζελφας.

8 Muraoka, A Syntax, § 3 c (pp. 13–14).

The Usage of the Article with Nouns Defined by a Genitive

47

In Genesis 36 the idioms (οὗτοι) υἱοί + a proper name in the genitive and οὗτοι ἡγεμόνες + a proper name in the genitive are frequently repeated. The idiom (οὗτοι) υἱοί + a proper name in the genitive recurs a number of times in the long list of Jacob’s sons and grandsons in Genesis 46, too. Some manuscripts add the definite article in these instances. The names of family members are often used without an article without any strict logic as, for instance, in Gen 46:7 ‫ּובנֹות ָּבנָ יו‬ ְ ‫ּובנֵ י ָבנָ יו ִאּתֹו ְּבנ ָֹתיו‬ ְ ‫— ָּבנָ יו‬υἱοὶ καὶ οἱ υἱοὶ τῶν υἱῶν αὐτοῦ μετ᾽ αὐτοῦ, θυγατέρες καὶ θυγατέρες τῶν υἱῶν αὐτοῦ.9 Gen 36:17 ‫ן־ע ָׂשו‬ ֵ ‫עּואל ֶּב‬ ֵ ‫—וְ ֵא ֶּלה ְּבנֵ י ְר‬καὶ οὗτοι υἱοὶ Ραγουηλ υἱοῦ Ησαυ.

If the proper name is followed by a patronymic, matronymic or an apposition of the similar type, the noun denoting the family connection is mostly without the article. This happens in the genealogical lists in particular:10 Gen 36:39 ‫ת־מ ְט ֵרד ַּבת ֵמי זָ ָהב‬ ַ ‫יט ְב ֵאל ַּב‬ ַ ‫—וְ ֵׁשם ִא ְׁשּתֹו ְמ ֵה‬ὄνομα δὲ τῇ γυναικὶ αὐτοῦ Μαιτεβεηλ θυγάτηρ Ματραιθ υἱοῦ Μαιζοοβ. Gen 36:43 ‫—הּוא ֵע ָׂשו ֲא ִבי ֱאדֹום‬οὗτος Ησαυ πατὴρ Εδωμ.

In the long list of the sons and grandsons of Jacob in Genesis 46, the nouns are usually anarthrous: υἱοὶ δὲ Ρουβην Ενωχ καὶ Φαλλους Ασρων καὶ Χαρμι (Gen 46:9), etc. Nevertheless, the introduction to this list includes definite articles in Gen 46:8 ταῦτα δὲ τὰ ὀνόματα τῶν υἱῶν Ισραηλ τῶν εἰσελθόντων εἰς Αἴγυπτον. Ιακωβ καὶ οἱ υἱοὶ αὐτοῦ. One category of anarthrous nouns consists of such titles as βασιλεύς and ἡγεμών. They are used without an article before a personal or geographical proper name in LXX Genesis. The anarthrous use of βασιλεύς as a title is also attested in contemporary Koine sources.11 Examples: Gen 14:8 ‫ּומ ֶלְך ֲעמ ָֹרה‬ ֶ ‫ְך־סד ֹם‬ ְ ‫—וַ ּיֵ ֵצא ֶמ ֶל‬ἐξῆλθεν δὲ βασιλεὺς Σοδομων καὶ βασιλεὺς Γομορρας. Gen 36:40 ‫— ַאּלּוף ִּת ְמנָ ע ַאּלּוף ַע ְלוָ ה ַאּלּוף יְ ֵתת‬ἡγεμὼν Θαμνα, ἡγεμὼν Γωλα, ἡγεμὼν Ιεθερ. 9  E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit: Band II 2 Satzlehre, Analytischer Teil, zweite Hälfte (Berlin und Leipzig: Walter de Gruyter & Co, 1933), § 57, 2 b Anmerk (p. 23). His examples of the use of the article in (ἡ) γυνὴ καὶ (οἱ) υἱοὶ are illustrative. 10  Muraoka (A Syntax, § 3 b [pp. 12–13]) does not consider the patronymics or matronymics. In expressions of origin by birth the nouns for patronymics and matronymics often appear without the article as the examples of Schwyzer/Debrunner (II, e III 2 γ [p. 124]) show. See also Mayser, Grammatik II 2, § 57, 2a (p. 22). 11 Mayser, Grammatik II 2, § 58 2 b (p. 27).

48

Raija Sollamo

One fascinating group of instances consists of geographical proper names in ­Hebrew and their Greek renderings, and in particular those that are formed with the noun γῆ “country” (regens) and an anarthrous proper name (rectum). There appear 60 cases altogether in which the proper name Αἰγυπτος, Χανααν, Εδωμ, etc. is in the genitive after the anarthrous noun γῆ in the genitive, dative or accusative12 and 11 cases where the “country” is in the dative and the proper name Αἰγυπτος is in the dative Αἰγύπτῳ after the anarthrous dative γῇ. Examples: Gen 31:18 ‫— ָלבֹוא ֶאל־יִ ְצ ָחק ָא ִביו ַא ְר ָצה ְּכנָ ַען‬ἀπελθεῖν πρὸς Ισαακ τὸν πατέρα αὐτοῦ εἰς γῆν Χανααν. Gen 32:3(4) ‫— ַא ְר ָצה ֵׂש ִעיר ְׂש ֵדה ֱאדֹום‬εἰς γῆν Σηιρ εἰς χώραν Εδωμ. Gen 47:4 ‫י־כ ֵבד ָה ָר ָעב ְּב ֶא ֶרץ ְּכנָ ַען‬ ָ ‫— ִּכ‬ἐνίσχυσεν γὰρ ὁ λιμὸς ἐν γῇ Χανααν. Gen 21:21 ‫— ֵמ ֶא ֶרץ ִמ ְצ ָריִם‬ἐκ γῆς Αἰγύπτου. Gen 46:20 ‫— ְּב ֶא ֶרץ ִמ ְצ ַריִם‬ἐν γῇ Αἰγύπτῳ. Note the dative Αἰγύπτῳ!

The proper noun Αἴγυπτος is found both in the dative and in the genitive after the anarthrous dative γῇ (dative in 41:19; 41:29; 41:30; 41:36; 41:53; 46:20; 46:27; 47:11; 47:14; 47:27; 47:28; genitive in 41:44; 41:48; 41:54), while after the anarthrous genitive and accusative it appears in the genitive (after a genitive 21:21; 41:33; 41:43; 45:8; 45:9; 45:26; 47:15; after an accusative 41:46). Which is the older or more original reading the dative or the genitive after the dative γῇ, is a challenging text-critical problem. Soisalon-Soininen, when studying the status constructus cases, presented statistics for the relation of the occurrences of the dative and genitive after the dative γῇ and clearly demonstrated that the dative is more common than the genitive in all books of the Pentateuch, the relations are 10:3 in Genesis, 16:10 in Exodus, 1:0 in Leviticus, 3:1 in Numeri and 9:0 in Deuteronomy.13 The witnesses are strongly divided and accordingly, it is impossible to reach one and the same solution in every occurrence, but the variation between the dative and the genitive remains.14 I give here an example of the unanimity of the mss. 12  The translator is not fully consistent in the use of the article. Two examples of γῆ with the article: Gen 36:21 ‫—ּב ֶא ֶרץ ֱאדֹום‬ἐν ְ τῇ γῇ Εδωμ, Gen 23:19 ‫—ּב ֶא ֶרץ ְּכנָ ַען‬ἐν ְ τῇ γῇ Χανααν, see also Gen 10:10, and 36:34. In the nominative the article usually appears (Gen 13:10, 41:55, 47:6, 47:13 bis, e.g. Gen 13:10 ὡς ἡ γῆ Αἰγύπτου), occasionally also in the genitive and accusative if they are not connected with a preposition (Gen 2:11, 2:13, 17:8, 41:34, and 50:11). 13  I. Soisalon-Soininen, “Verschiedene Wiedergaben der hebräischen status constructusVerbindung im griechischen Pentateuch”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari SoisalonSoininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 62–70. For Genesis the figures should be 11:3, as the text above demonstrates. 14  I have followed Wevers’ decisions in his critical text of the Göttingen Septuaginta. See J.W. Wevers (ed.), Septuaginta Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum vol. I Genesis, (Göttingen: Vandenhoeck & Ruprecht, 1974).

The Usage of the Article with Nouns Defined by a Genitive

49

Gen 41:19 ‫ל־א ֶרץ ִמ ְצ ַריִם‬ ֶ ‫— ְּב ָכ‬ἐν ὅλῃ γῇ Αἰγύπτῳ. Αἰγύπτῳ A D 961 72-135 Bs 44-107-370-610*(vid) 75 121-318-619 319 LaI ] 962 αιπτῳ 962; >408*; -πτου LaS rell.

Fixed idioms and grammaticalized expressions consisting of a preposition and a noun usually appear without an article in LXX Genesis and in Koine sources. The renderings of Hebrew semiprepositions that consist of a preposition and a noun belong to this group of instances (19 cases).15 Examples: Gen 17:11 :‫יכם‬ ֶ ֵ‫ּובינ‬ ֵ ‫—וְ ָהיָ ה ְלאֹות ְּב ִרית ֵּבינִ י‬καὶ ἔσται ἐν σημείῳ διαθήκης ἀνὰ μέσον ἐμοῦ καὶ ὑμῶν. Gen 40:20 ‫— ְּבתֹוְך ֲע ָב ָדיו‬ἐν μέσῳ τῶν παίδων αὐτοῦ. Gen 32:12 ‫ילנִ י נָ א ִמּיַ ד ָא ִחי ִמּיַ ד ֵע ָׂשו‬ ֵ ‫— ַה ִּצ‬ἐξελοῦ με ἐκ χειρὸς τοῦ ἀδελφοῦ μου Ησαυ. Gen 16:8 ‫אמר ִמ ְּפנֵ י ָׂש ַרי ּגְ ִב ְר ִּתי ָאנ ִֹכי ּב ַֹר ַחת‬ ֶ ֹ ‫—וַ ּת‬καὶ εἶπεν ἀπὸ προσώπου Σαρας τῆς κυρίας μου ἐγὼ ἀποδιδράσκω. Gen 38:20 ‫הּודה ֶאת־ּגְ ִדי ָה ִעּזִ ים ְּביַ ד ֵר ֵעהו‬ ָ ְ‫—וַ ּיִ ְׁש ַלח י‬ἀπέστειλεν δὲ Ιουδας τὸν ἔριφον ἐξ αἰγῶν ἐν χειρὶ τοῦ ποιμένος αὐτοῦ.

Other examples of fixed expressions used without an article also appear. The absence of the article could be considered due to Hebrew interference or “reinforced by the underlying Semitic syntagm with the necessarily anarthrous NP (Noun Phrase) in the st.c…”16 Nevertheless, this is not very likely, since the principal method of the Genesis translator was to use the definite article in rendering nouns in the status constructus defined by another noun in the status absolutus or defined by a pronominal suffix. Thus, the omission of the article was—at least mostly—a marked alternative based on the translator’s command of Greek.17 Examples: Gen 8:19 ‫ן־ה ֵּת ָבה‬ ַ ‫יהם יָ ְצאּו ִמ‬ ֶ ‫— ְל ִמ ְׁש ְּפח ֵֹת‬κατὰ γένος αὐτῶν ἐξήλθοσαν ἐκ τῆς κιβωτοῦ. The nouns in distributive expressions after the preposition κατά are usually anarthrous in Greek sources.18 Gen 10:3 ‫י־ׁשם ְל ִמ ְׁש ְּפח ָֹתם ִל ְלׁשֹנ ָֹתם ְּב ַא ְרצ ָֹתם‬ ֵ ֵ‫— ֵא ֶּלה ְבנ‬οὗτοι υἱοὶ Σημ ἐν ταῖς φυλαῖς αὐτῶν κατὰ γλώσσας αὐτῶν ἐν ταῖς χώραις αὐτῶν. Gen 17:11 ‫—וְ ָהיָה ְלאֹות ְּב ִרית‬καὶ ἔσται ἐν σημείῳ διαθήκης. Here apparently ἐν pro εἰς. 15  See more in R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (AASF Diss. hum. litt. 19 Helsinki: Suomalainen Tiedeakatemia, 1979). 16  Cf. Muraoka, A Syntax, § 3 a (p. 12). See also, however, note 2 (p. 12). 17 R. Kühner/B. Gerth, Ausführliche Grammatik der griechischen Sprache II:1 Satzlehre (3. Aufl; Hannover: Hansche, 1898), § 462 (pp. 598–610). After a preposition the article is often omitted § 462, f (p. 605). 18  See the examples in Schwyzer/Debrunner, Syntax, 2b α 10 (pp. 473–81). Cf. Mayser, Grammatik II 2, § 57, 3 b (p. 24).

50

Raija Sollamo

Gen 47:24 ‫—יִ ְהיֶ ה ָל ֶכם ְלזֶ ַרע ַה ָּׂש ֶדה ּוֽ ְל ָא ְכ ְל ֶכם‬ἔσται ὑμῖν αὐτοῖς εἰς σπέρμα τῇ γῇ καὶ εἰς βρῶσιν ὑμῖν.

To sum up, the great number of the arthrous examples in relation to the anarthrous examples was a surprise, at least for me. These examples suffice to lead us to the understanding that the instances with no definite article are not at all indifferent or Hebraistic Greek even if they are slavish renderings from the point of view of grammatical structure. The absence of the article is not due to a slavish adherence to the Hebrew syntactic structure, but due to paying attention to the qualities of the target language. Even the anarthrous examples follow good Koine practice and idiom in not using the article. The translator was capable of seeing in which instances an anarthrous noun is idiomatic in Greek. Finally, both the use and non-use of the definite article attest the Genesis translator’s con­ siderable competence in Greek. This situation is rare because usually when one studies translation technique he/she is able to find both genuine Koine Greek and Hebraistic translation Greek side by side by the same translator. If and when a Hebraistic instance or an instance of Hebrew interference occurs in this material, it really comprises an individual case and is an exception or a slip in the midst of otherwise correct Koine Greek.

4. Free Renderings without the Genitive in Greek My title was “The Usage of the Article with Nouns Defined by a Nominal or Pronominal Genitive in LXX Genesis”. The task is now fulfilled, but only if we consider the Greek translation. If we should like to conduct a more thorough investigation of translation technique, we should take into the loop all the 2844 Hebrew instances where there appears a status constructus—status absolutus structure or a pronominal suffix attached to a noun in the parent text and find out what has happened to those cases in the Greek translation even when it displays a different structure and freer rendering without the genitive. In other words, these instances are translated so freely that one cannot apply the definitions of my title to them. This is my problem, but it has wider relevance for the study of Septuagint syntax. In my opinion, we should not study the Septuagint syntax paying attention primarily to the Greek target language as Muraoka does in his new syntax and as my title now does. Rather, we should prefer starting from the Hebrew parent text and its structures and features in order to learn to understand the subtle differences between the two languages and in order to give an adequate description of what the translator has achieved. Therefore, I now refer in brief at the end to those cases where the translator has not used a genitive at all. This kind of free renderings comprises 374 instances (circa 13%).

The Usage of the Article with Nouns Defined by a Genitive

51

The most common model for a free rendering is to leave the genitive out and use only the definite article (93 instances). This practice is in accord with Greek idiom in cases where the referent is clear from the context. In the Septuagint the genitives of the third persons of the possessive pronoun are anyway found with unidiomatic frequency. Leaving a number of such genitives out shows the great competence of the Genesis translator in Greek.19 Examples: Gen 24:45 ‫ל־ׁש ְכ ָמּה‬ ִ ‫—וְ ִהּנֵ ה ִר ְב ָקה י ֵֹצאת וְ ַכ ָּדּה ַע‬εὐθὺς Ρεβεκκα ἐξεπορεύετο ἔχουσα τὴν ὑδρίαν ἐπὶ τῶν ὤμων. Gen 27:31 ‫אמר ְל ָא ִביו‬ ֶ ֹ ‫—וַ ּי‬καὶ εἶπεν τῷ πατρί. Gen 33:10 ‫—וְ ָל ַק ְח ָּת ִמנְ ָח ִתי ִמּיָ ִדי‬δέξαι τὰ δῶρα διὰ τῶν ἐμῶν χειρῶν.

One way of deleting genitives that the translator resorts to is to replace them with datives in Greek. It happens sporadically here and there (as in Gen 47:24 cited above), but most consequently in the cases of the onoma rule (25 instances in LXX Genesis). The onoma rule is valid in Greek idiomatic expressions where the name of a person is introduced: onoma (without the article) + the person whose name is in question in the dative and then the name.20 Examples: Gen 11:29 ‫ת־א ְב ָרם ָׂש ָרי‬ ַ ‫— ֵׁשם ֵא ֶׁש‬ὄνομα τῇ γυναικὶ Αβραμ Σαρα. Gen 36:32 ‫—וְ ֵׁשם ִעירֹו ִּדנְ ָה ָבה‬καὶ ὄνομα τῇ πόλει αὐτοῦ Δενναβα. Gen 26:33 ‫ם־ה ִעיר ְּב ֵאר ֶׁש ַבע‬ ָ ‫ל־ּכן ֵׁש‬ ֵ ‫— ַע‬διὰ τοῦτο ὄνομα τῇ πόλει Φρέαρ ὅρκου.

With the aid of the status constructus Hebrew also expresses material and quality (in at least 43 cases), not only possession and belonging. Often this implies that a genitive of quality is translated with an adjective. Examples: Gen 6:4 ‫— ַאנְ ֵׁשי ַה ֵּׁשם‬οἱ ἄνθρωποι οἱ ὀνομαστοί Gen 9:12 ‫עֹולם‬ ָ ‫— ְלדֹר ֹת‬εἰς γενεὰς αἰωνίους

One way to avoid using a genitive is to form the words in the status constructus and status absolutus into a compound word and use it with or without the article (at least 44 instances). Examples:

19  I. Soisalon-Soininen, “Die Auslassung des Possessivpronomens im griehischen Pentateuch” in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur SeptuagintaSyntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237 Helsinki: Suomalainen Tiedeakatemia, 1987) 86–92. 20  I have given a fuller analysis of these cases in my article under the working title “The Onoma Rule and the Vocative in LXX Genesis” to appear in 2020. See also J.A.L. Lee, “The Onoma Rule” NovT 56 (2014) 411–21.

52

Raija Sollamo

Gen 12:11 ‫ת־מ ְר ֶאה ָא ְּת‬ ַ ‫—יָ ַד ְע ִּתי ִּכי ִא ָּׁשה יְ ַפ‬γινώσκω ἐγὼ ὅτι γυνὴ εὐπρόσωπος εἶ. Gen 19:4 ‫ל־ה ַּביִת‬ ַ ‫—וְ ַאנְ ֵׁשי ָה ִעיר ַאנְ ֵׁשי ְסד ֹם נָ ַסּבּו ַע‬καὶ οἱ ἄνδρες τῆς πόλεως οἱ Σοδομῖται περιεκύκλωσαν τὴν οἰκίαν. Gen 40:4 ‫ת־יֹוסף‬ ֵ ‫—וַ ּיִ ְפקֹד ַׂשר ַה ַּט ָּב ִחים ֶא‬καὶ συνέστησεν ὁ ἀρχιδεσμώτης τῷ Ιωσηφ. Gen 49:13 ‫יַּמים יִ ְׁשּכֹן‬ ִ ‫בּולן ְלחֹוף‬ ֻ ְ‫—ז‬Ζαβουλων παράλιος κατοικήσει.

Furthermore, the pronominal suffix of the first person singular often indicates an address corresponding to the Greek vocative case (36 cases in LXX Genesis).21 Examples: Gen 19:7 ‫אמר ַאל־נָ א ַא ַחי‬ ַ ֹ ‫—וַ ּי‬εἶπεν δὲ πρὸς αὐτούς Μηδαμῶς, ἀδελφοί. Gen 27:20 ‫— ַמה־ּזֶ ה ִמ ַה ְר ָּת ִל ְמצֹא ְּבנִ י‬τί τοῦτο, ὃ ταχὺ εὗρες, ὦ τέκνον. Gen 44:18 ‫— ִּבי ֲאד ֹנִ י‬δέομαι, κύριε.

In addition to these instances there appear 53 miscellaneous free renderings which occur sporadically, such as Gen 7:2 and 39:19. Gen 7:2 ‫ח־לָך ִׁש ְב ָעה ִׁש ְב ָעה ִאיׁש וְ ִא ְׁשּתֹו‬ ְ ‫הֹורה ִּת ַּק‬ ָ ‫— ִמּכֹל׀ ַה ְּב ֵה ָמה ַה ְּט‬ἀπὸ δὲ τῶν κτηνῶν τῶν καθαρῶν εἰσάγαγε πρὸς σὲ ἑπτὰ ἑπτά ἄρσεν καὶ θῆλυ. Gen 39:19 ‫—וַ ּיִ ַחר ַאּפֹו‬καὶ ἐθυμώθη ὀργῇ.

These free renderings demonstrate that the translator felt, at least to a certain extent that his Hebrew parent text had more genitives and possessive suffixes in different uses than what he could afford in his translation in idiomatic Greek. Thus, he changed the structure and rendered freely. The main thing is to be aware of these two different approaches: one keeping within the boundaries of Greek grammar, and another one following the Hebrew grammatical patterns and their Greek renderings even if these are scattered in different grammatical categories. The problem is similar to what Ilmari Soisalon-Soininen met when investigating the Hebrew infinitives and their renderings in the Septuagint.22 He had to leave the infinitivus absolutus cases out of his research because the Hebrew infinitivus absolutus was never translated by an infinitive in Greek. My solution here differs from his in that I refer to the free translations not following the grammatical structure of Hebrew. This is done in order to illuminate the vast selection of choices the translator had and mastered. The free renderings are very typical of his way of translating.

21  Here again I refer to my article under the working title “The Onoma Rule and the Vocative in LXX Genesis” to be published in 2020. 22  I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132:1; Helsinki: Suomalainen Tiedeakatemia, 1965).

The Usage of the Article with Nouns Defined by a Genitive

53

5. Conclusions To sum up in brief, the use of the definite article dominates the field before nouns defined by a nominal or pronominal genitive in Greek appearing in 1940 out of 2470 cases (78.5%). This is a result I did not expect. I anticipated in my mind that the translator followed more strictly the Hebrew usage of the article. This can only be explained by the Genesis translator’s proficiency in Greek. In this respect his translation is free from Hebrew interference. Certainly, there are also found cases without the article before nouns defined by a nominal or pronominal genitive in Greek, 530 out of 2470 cases (21.5%). Evaluated from a purely formal viewpoint these instances strictly follow the Hebrew expression in not using the article. It is, however, too hasty to regard all of these cases as witnesses of Hebrew interference. Some of them might show Hebrew interference, but most of them follow Greek practice and idiom as marked renderings chosen by the translator. At any rate, in this group there also exist some inconsistent uses of the anarthrous and arthrous alternatives which might reflect an uncertainty felt by the translator about whether to use or not to use the article or the possibility he regarded both alternatives as equally suitable because the vacillation between the use and non-use of the article in a number of phrases was typical of the contemporary Koine. In these anarthrous cases the influence of Hebrew interference is difficult either to prove or to totally exclude. The third group of instances consists of the very free renderings which break the Hebrew structure in 374 cases out of 2844 (13.1%). They are absolutely good Greek and show beyond doubt the translator’s superb command of Greek and good command of Hebrew too. He recognizes Hebrew addressing forms and renders them with Greek vocatives, he knows the onoma rule, he diminishes the number of genitives and of genitives of the third person possessive pronouns and is familiar with Greek idioms. He simply writes good idiomatic Greek. Even in his free renderings he adheres to the meaning of the Hebrew clauses and expressions but renders them by utilizing genuine Greek means. His linguistic competence is so extensive and rich in idiomatic variation that he must have been a native Greek speaker.

Jan Joosten Grammar and Style in the Septuagint: On Some Uses of Preverbs 1. Introduction In a broad definition, “Translation Technique” covers everything translators do when they render a text from one language into another. In a more restricted definition, it has to do with the way translators negotiate characteristic differences, particularly in the area of grammar, between the source language and the target language. In Ilmari Soisalon-Soininen’s pathbreaking papers, gathered in his Studien zur Septuaginta-Syntax, he generally sets out from syntactic or morphosyntactic features of the Hebrew source that find no ready equivalent in Koine Greek: the structure of the relative clause, nominal phrases with a construct state, temporal phrases with a construct infinitive, and other phenomena.1 He then describes how the various Septuagint translators transposed these grammatical structures into Greek while attempting to stay as close as possible to both the form and the meaning of the Hebrew. The analysis illuminates various aspects of the translational process. Several of Soisalon-Soininen’s students have further developed this approach, setting out from syntactic features in the Hebrew Bible and tracing their treatment in the Septuagint.2 Studies of translation technique can also adopt a reverse orientation. In some of his papers, Soisalon-Soininen focuses on Greek features that have no precise equivalent in Hebrew, such as the genitive absolute or the verb ἔχειν “to have”. These studies too have found followers in the Finnish School. Notably Anssi Voitila’s thesis on the verbal system in the Pentateuch includes highly interesting research on such typically Greek phenomena as the historical present.3 The results of these studies are less systematic. The translators’ use of characteristically Greek features is not induced by the Hebrew source text, but reflects their creativity, and their competence in the target language. Thus, although SoisalonSoininen does not, unless I err, state this explicitly, attention to typically Greek features opens up a window onto the style of the version. The style of the Septuagint has attracted relatively little attention. Only very recently have scholars 1  A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987). 2  See, e.g., the special issue SJOT 10 (1996) 159–237, and R. Sollamo/S. Sipilä (ed.), Helsinki Perspectives on the Translation Technique of the Septuagint. Proceedings of the IOSCS Congress in Helsinki 1999 (PFES 82; Helsinki: Finnish Exegetical Society/Göttingen: Vandenhoeck & Ruprecht, 2001). 3  Anssi Voitila, Présent et imparfait de l’indicatif dans le Pentateuque grec (PFES 79; Helsinki: Finnish Exegetical Society/Göttingen: Vandenhoeck & Ruprecht, 2001).

56

Jan Joosten

begun to discuss what may be defined as the style of the Septuagint translation, how it should be studied, and why it is important.4 In the footsteps of Soisalon-Soininen, I will in the present paper explore aspects of another feature peculiar to ancient Greek, namely the use of “preverbs” attached to simple verbs and influencing their meaning.

2. General Remarks on Preverbs in Ancient Greek Like many other languages, Greek allows the formation of compound verbs consisting of a “preverb” and a simple verb: the compound εἰσέρχομαι “to enter” is derived from simple ἔρχομαι “to come”.5 Even in the earliest Greek dialects, verbs compounded with preverbs are well represented. In Hellenistic Greek, they become very frequent, and compounds at times tend to compete with simple verbs or even to supersede them completely. Hellenistic Greek has a plethora of compound verbs, hundreds of verbs with two preverbs, and even some with three.6 In diachronic perspective, Greek preverbs may have started out as spatial adverbs that later specialized as prepositions, with nouns or pronouns, or as preverbs with verbs.7 Thus the adverb πρός “besides, over and above”, still somewhat frequent in Homer, is to be regarded as the ancestor of the well-known preposition and preverb of the same form. The phenomenon of tmesis, by which a preverb appears to be separated from the verb with which it interacts, may reflect an intermediate stage in which the two elements have not entirely combined into one.

4  See, e.g., the collection of papers: E. Bons/Th. J. Kraus (ed.), Et sapienter et eloquenter. Studies on Rhetorical and Stylistic Features of the Septuagint (FRLANT 241; Göttingen: Vandenhoeck & Ruprecht, 2011); see also the section on “The Style of the Septuagint” in E. Bons/ J. Joosten (ed.), Die Sprache der Septuaginta/The Language of the Septuagint (LXX.H 3; Gütersloh: Gütersloher, 2016) 357–428. 5 On preverbation in general, see the articles in G. Booij/J. Van Marle (ed.), Yearbook of Morphology 2003 (Dordrecht: Kluwer, 2003), 1–212. On the preverb in early Greek, see C. Viti, “Coding Spatial Relations in Homeric Greek: Preverbs vs. Prepositions”, Historische Sprachforschung / Historical Linguistics 121 (2008) 114–161; C. Zanchi, “New Evidence for the Source–Goal Asymmetry: Ancient Greek Preverbs”, in S. Luraghi/T. Nikitina/C. Zanchi (ed.), Space in Diachrony (Amsterdam: John Benjamins, 2017) 147–78. 6  An annotated list of composites with two preverbs can be found in R. Helbing, Grammatik der Septuaginta. Laut- und Wortlehre (Göttingen: Vandenhoeck & Ruprecht, 1907), 128–136. The list is limited to verbs attested in the Septuagint, but for the verbs included it does provide information on the attestations in Hellenistic literature and documentary papyri. 7  See Zanchi, “Greek Preverbs”.

Grammar and Style in the Septuagint: On Some Uses of Preverbs

57

Il. 5.632 Τὸν καὶ Τληπόλεμος πρότερος πρὸς μῦθον ἔειπεν “Tlepolemos first said a word to him”

Whatever their origin, preverbs and prepositions clearly go hand in hand. Practically all preverbs are also attested as prepositions and vice versa. In Greek grammars, preverbs are generally discussed in sections on word formation. They are regarded as a morphological means of producing a new verb. This view is correct where the semantics of the compound clearly differ from the simple verb: the compound is lexicalized. Compound ἀποστέλλω “to send” and ἀποκρίνομαι “to answer” are etymologically related to στέλλω and κρίνω, but they are semantically distinct. Lexicalization accounts for double or triple preverbation: when a compound verb expresses a global meaning and is no longer felt to be composite, an additional preverb may be added to it: ἐξαποστέλλω is derived from ἀποστέλλω “to send”, not from στέλλω “to array”. This could only happen when the force of the first preverb (i.e. the second) was no longer felt. Similarly, a third preverb could be added to a doubly prefixed verb whose meaning had been lexicalized: προεξαπέστειλεν in 2 Macc 12:21 could only be derived from ἐξαποστέλλω. Not all compound verbs are lexicalized, however. At times the preverb modifies the verb without changing its lexical meaning. Adding the preverb is in this case more or less optional, as is shown by the fact that simple verbs and compound verbs often alternate, in parallel passages or in different manuscripts of the same writing.8 The preverb may qualify the meaning of the verb almost like an adverb: περιβλέπω is “to look around,” συγκαλέω “to call together,” etc. In other cases, the preverb focuses on one phase of the verbal process (Phasenaktionsart): ἀναμιμνῄσκω is “to begin to remember” and ἐκθερίζω is “to harvest to the end.” In many cases the meaning of the preverb is hard to define: περικυκλόω “to surround” is more or less the same as simple κυκλόω, and προσκολλάω “to attach to” is very close to κολλάω. One might say in these cases that the preverb somehow reinforces the meaning of the verb. Others will prefer to see these preverbs as more or less redundant. Lexicalized compounds and “free” uses of the preverb are not wholly distinct categories. There is a sliding scale. Also, one and the same compound can at times function in different ways according to the context.

8  For the Septuagint this has been partially documented by M. L. Margolis, “The Greek Preverb and Its Hebrew-Aramaic Equivalent”, AJSL 26 (1909) 33–61.

58

Jan Joosten

3. Preverbs in the Septuagint As expected in light of the situation in Hellenistic Greek generally, preverbs are prominent in the Septuagint. Several simple verbs of classical Greek are replaced with compounds.9 The well-known Greek verb ἀγγέλλω is never attested in the Septuagint. In contrast, compounds of ἀγγέλλω with the preverbs ανα, απο, δια, εξ, επι, κατα, παρα, and προς are all attested multiple times. Some of the compounds express distinct meanings: thus, middle ἐπαγγέλλομαι specializes in the meaning “to promise”. Other compounds of ἀγγέλλω correspond to well-defined Hebrew equivalents: for instance, διαγγέλλω mostly renders Hebrew ‫ ספר‬piel. But some of the remaining compounds are hard to tell apart. Both ἀναγγέλλω and ἀπαγγέλλω occur very often, in the same contexts, for the same Hebrew equivalent, ‫“ הגיד‬to announce”. In several instances, it seems the change from one to the other simply reflects the need for stylistic variation. While the use of lexicalized compounds in the correct meaning is a mark of basic competence, the ease with which the translators vary “free” preverbs testifies to their intimate familiarity with the kind of Greek they write. The translators display a fine sensitivity to the nuances preverbs are capable of. Beyond the more or less spontaneous use to express distinct meanings, or to vary the style, the Septuagint translators develop the use of the preverb in highly profiled ways. Two special uses will be presented in what follows.

3.1 Simple Verb and Compound Expressing Gradation A beautiful illustration of the competence of the Septuagint translators is the way they use the preverb ἐκ to express gradation. While translating the Septuagint of Hosea for La Bible d’Alexandrie some fifteen years ago, the research group was surprised to find the compound ἐκπορνεύω to be used alongside simple πορνεύω in the meaning “to fornicate, to commit sexual impropriety”. The compound is found throughout the Septuagint and may in some passages express the nuance that the fornication happens outside the expected social group.10 In Hosea, however, the simple and composite forms appear to alternate without clear motivation. Since the preverb ἐκ can express completion as well as removal, we decided the nuance of the compound was one of excess: “to fornicate exceedingly”. This interpretation is particularly fitting in passages where the two verbs follow closely upon one another:

9 Margolis, “Preverb”. 10  See, e.g., Gen 38:24.

Grammar and Style in the Septuagint: On Some Uses of Preverbs

59

Hos 4:18 ᾑρέτισεν Χαναναίους· πορνεύοντες ἐξεπόρνευσαν, ἠγάπησαν ἀτιμίαν ἐκ φρυάγματος αὐτῶν. He chose Chananites; fornicating they committed fornication. They loved dishonor because of its insolence.11

Although, as NETS demonstrates, it is hard to translate into English, the Greek seems to imply that the fornicating went on to a point where it became excessive: “fornicating they have gone over the top in their fornication.” Whatever the precise nuance of the preverb, the translator is clearly trying to rise to a challenge presented to him by the paronomastic expression ‫ַהזְ נֵ ה ִהזְ נוּ‬ in the Hebrew text. Verbal repetition of various kinds is characteristic of good Hebrew style, but it posed a problem to the Greek translators. Often, they render repetitions literally, presumably in the name of faithfulness to the source text, but in a number of cases they find ways to attenuate or do away with it. The technique found in Hos 4:18 is particularly ingenious in that it exploits a feature in the target language that had no equivalent in the source language. Both πορνεύω and ἐκπορνεύω are correct renderings of Hebrew ‫זנה‬, but their combination in the same verse adds a nuance that goes beyond the Hebrew. That the combination of the simple verb with the compound is indeed a mark of gradation is confirmed by a number of parallels. Similar examples are found in several other books of the Septuagint: Cant 2:7 ὥρκισα ὑμᾶς, θυγατέρες Ιερουσαλημ, ἐν ταῖς δυνάμεσιν καὶ ἐν ταῖς ἰσχύσεσιν τοῦ ἀγροῦ, ἐὰν ἐγείρητε καὶ ἐξεγείρητε (‫עֹוררוּ‬ ְ ‫ם־תּ‬ ְ ‫ם־תּ ִעירוּ וְ ִא‬ ָ ‫ ) ִא‬τὴν ἀγάπην, ἕως οὗ θελήσῃ. I have adjured you, O daughters of Ierousalem, by the powers and by the forces of the field that you do not stir up or awaken love until it wish!

In this example, the Hebrew and the Greek arguably do something similar: in Hebrew the use of the polel alongside the hiphil of the same verb creates an expression of intensity in the same way as the use of the simple and the composite in Greek. Ezek 16:54 and 61 ὅπως κομίσῃ τὴν βάσανόν σου καὶ ἀτιμωθήσῃ … καὶ μνησθήσῃ τὴν ὁδόν σου καὶ ἐξατιμωθήσῃ … (‫ וְ נִ ְכ ַל ְמ ְתּ‬in both verses)

11  English translations of the Septuagint follow NETS except where indicated.

60

Jan Joosten

in order that you may bear your trial and be dishonoured … And you will remember your way, and you shall be completely dishonoured …

Although these verses are not adjacent, they clearly link up with one another in the rhetoric of Ezek 16. Identical verbal forms in Hebrew are rendered first with the simple, then with the compound. The implication is one of gradation, as NETS correctly reflects. Note also the following examples in non-translated Greek: Dan 3:29 LXX (Theodotion idem)12 ὅτι ἡμάρτομεν ἐν πᾶσι καὶ ἠνομήσαμεν ἀποστῆναι ἀπὸ σοῦ καὶ ἐξημάρτομεν ἐν πᾶσι For we have sinned in everything and broken your law in turning away from you, and in all matters we have sinned grievously. 3 Macc 6:27 λύσατε ἐκλύσατε ἄδικα δεσμά Loose and untie their unjust bonds.

Although they present only a small sample, these examples nicely illustrate the creativity and competence of the Septuagint translators. The translators are able to draw upon the resources of Greek even where the Hebrew source text does not invite doing so.

3.2 A Rhetorical Flourish While preparing the article on ἀπωθέω “to reject” for the Historical and Theological Lexicon of the Septuagint, we came across the double compound προσαπωθέω in Sirach: Sir 13:21 πλούσιος σαλευόμενος στηρίζεται ὑπὸ φίλων, ταπεινὸς δὲ πεσὼν προσαπωθεῖται ὑπὸ φίλων. When a rich person totters, he is supported by friends, but when a humble person falls, he is pushed away by friends.

Dictionaries and translations of the Septuagint neglect the preverb in this verse, perhaps submitting that it simply reinforces the movement expressed by the verb

12  On the prayer of Azariah as an original Greek composition, see J. Joosten, “The Prayer of Azariah (DanLXX 3): Sources and Origin”, in J. Cook (ed.), Septuagint and Reception (VTS 127; Leiden: Brill, 2009), 5–16.

Grammar and Style in the Septuagint: On Some Uses of Preverbs

61

ἀπωθέω. In that case, however, one would expect the goal of the movement to be stated. A different nuance seems to be meant. According to LSJ, προς in composite verbs may express motion toward, as in προσάγω, or being on, at, by, or beside, as in πρόσειμι. It may also express addition as in προστίθημι “to add” and προσδίδωμι “to give in addition”. Where προς expresses addition, reiteration of the same process is usually implied. Leah says: Gen 29:33 ὅτι ἤκουσεν κύριος ὅτι μισοῦμαι, καὶ προσέδωκέν μοι καὶ τοῦτον Because the Lord has heard that I am hated, he has also in addition given me this one too

After having given her sons, the Lord has given Leah another son. The verb προσδίδωμι “to give in addition” is well established in Hellenistic Greek. But even in cases where the derivation is more spontaneous, reiteration of the same process is usually meant. Note the law on criminal neglect: Exod 21:29 ὁ ταῦρος λιθοβοληθήσεται, καὶ ὁ κύριος αὐτοῦ προσαποθανεῖται the bull shall be stoned, and its owner shall die as well.

The verb προσαποθνῄσκω is very rare in Greek and seems to have been produced freely in view of the context in Exod 21:29. But it still expresses reiteration of the same process: the goring bull will die, and so will its owner.13 There is, however, a more evolved usage where the preverb no longer refers to the process expressed by the verb, but to the entire situation described in the discourse. In his lexical studies of the Septuagint, Caird pointed out a possible instance in passing:14 Sir 3:14 ἐλεημοσύνη γὰρ πατρὸς οὐκ ἐπιλησθήσεται καὶ ἀντὶ ἁμαρτιῶν προσανοικοδομηθήσεταί σοι. For charity for a father will not be forgotten, and it will be credited to you against sins.

13 Other examples of the more spontaneous use of this preverb: Exod 22:5(6): ἐὰν δὲ ἐ­ ξελθὸν πῦρ εὕρῃ ἀκάνθας καὶ προσεμπρήσῃ ἅλωνα ἢ στάχυς ἢ πεδίον “Now if fire breaks out and finds thorns and [in addition, JJ] burns a threshing floor or ears of corn or a plain”; προσαναλεξάμενος “in addition, he recounted” 2 Macc 8:19. 14  G. B. Caird, “Towards a Lexicon of the Septuagint. II”, JThS (1969) 21–40.

62

Jan Joosten

Caird rendered this rather freely: “If you support your father, it will not be forgotten, but will go to build up extra credit for you to offset your sins.”15 In this verse, the compound does not imply reiteration of the same process. Instead, it designates that the building of credit will come on top of what else is implied in the verse. The preverb functions in a similar way in Sir 13:21. The humble man, when he stumbles, will in addition be thrust down by his friends. The second process does not repeat the first, but is added to it as an extra. Further confirmation of this function is forthcoming from the context. The same usage seems to occur again in the next two verses in Sirach: Sir 13:22–23 πλουσίου σφαλέντος πολλοὶ ἀντιλήμπτορες· ἐλάλησεν ἀπόρρητα, καὶ ἐδικαίωσαν αὐτόν. ταπεινὸς ἔσφαλεν, καὶ προσεπετίμησαν αὐτῷ· ἐφθέγξατο σύνεσιν, καὶ οὐκ ἐδόθη αὐτῷ τόπος. πλούσιος ἐλάλησεν, καὶ πάντες ἐσίγησαν καὶ τὸν λόγον αὐτοῦ ἀνύψωσαν ἕως τῶν νεφελῶν. πτωχὸς ἐλάλησεν καὶ εἶπαν Τίς οὗτος; κἂν προσκόψῃ, προσανατρέψουσιν αὐτόν. When a rich person staggers, many are his helpers; he spoke things not to be spoken, and they justified him. A humble person staggered, and in addition they rebuked him; he uttered sense, and no place was given to him. A rich person spoke, and all kept silent, and they exalted his word up to the clouds. A poor person spoke, and they said, “Who is this?” And if he should stumble, they will even overturn him.

The thread of the whole section is that the errors of the rich are glossed over by their retinue, while the mishaps of the poor are an occasion for their friends to distance themselves from them. The preverb corresponds to nothing in the Hebrew, but fits the context nicely and enhances its overall rhetoric. The best translation into English would be something like “for full measure” or “to cap it all”. I have not found instances of this usage in the other books of the Septuagint. The translator of Ben Sira seems to be particularly fond of it. The usage is encountered in two more passages:

15 Caird, “Lexicon”, 30.

Grammar and Style in the Septuagint: On Some Uses of Preverbs

63

Sir 4:3 καρδίαν παρωργισμένην μὴ προσταράξῃς An angry heart do not [in addition, JJ] trouble. Sir 13:3 πλούσιος ἠδίκησεν, καὶ αὐτὸς προσενεβριμήσατο· A rich person did wrong, and he was angry to boot.

Much work remains to be done on the use of preverbs in the Septuagint. The present case studies are tentative and need to be extended and followed up. And many questions linked to the syntax and semantics of preverbs remain unanswered in the present state of our knowledge.

4. Conclusions On first impression the Septuagint does not seem to live up to the classical canons of style. According to Greek rhetoric, style should be correct, clear, and ornamented. On these criteria, the Greek version falls short in many passages. The word-for-word approach to translation has led to some incorrectness, and much lack of clarity. A more patient examination, however, will also discover stylistic flourishes among the weeds of literalism. Our exploration of the use of the preverb in the Septuagint links up with earlier studies on variation, alliteration and metre to show that the Septuagint translators are not oblivious to the demands of style. They were able to make use of the resources of the Greek language to create a pleasing effect. One may judge that the translators were not successful. Some readers may think of the Septuagint as a sort of monstrosity. But one cannot accuse the translators of being indifferent to style. Their intention was to produce a Greek text that would be pleasing to the reader, while at the same time accurately rendering the Hebrew source. In addition, the present study demonstrates the enduring fruitfulness of the programme of Septuagint studies defined by Ilmari Soisalon-Soininen. His focus on language, and grammar in particular, and his plea for conjoint consideration of the source text and the translation, make for a daunting enterprise. Translation technique as practiced in the Finnish school is a tremendous amount of work. But the results are often revealing, as the present volume of essays illustrates.

Theo A.W. van der Louw The Dynamics of Segmentation in the Greek Pentateuch 1. Introduction and State of the Question Dedicating a symposium to the memory of Ilmari Soisalon-Soininen is more than appropriate. Soisalon-Soininen was gifted with an unrivalled insight into what went on in the mind of the Septuagint translators, and pioneered observations that we have not even begun to delve into. A good example is “Segmentation in the Greek Pentateuch” (1983), which has held me enthralled since I first read it.1 But three decades after this classic paper, scholarly opinion on the segmentation hypothesis shows a stalemate. Segmentation is either posited as a given,2 is ignored, or is denied.3 Its opponents have never subjected Soisalon-Soininen’s paper to scrutiny. Schaper, for example, calls the notion of segmentation with respect to the Septuagint “an unwarranted assumption”. According to him, a mere look at LXX-Exod 5:2 should be enough to refute it, since the elegance of that verse “betrays the hand of a highly confident translator.”4 The adherents of segmentation theory, for their part, have never bothered to fill in its evident lacunae and mental leaps. In other words, the phenomenon itself has not become clearer.5 So it happens that scholars who sympathize with the

1  I. Soisalon-Soininen, “Segmentation in the Greek Pentateuch” (in this volume). SoisalonSoininen gave the segmentation hypothesis its classical formulation, building on the work of others, notably A. Aejmelaeus, Parataxis in the Septuagint: A Study of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASF, Diss. Hum. Litt. 31; Helsinki: Suomalainen Tiedeakatemia, 1982), 145–6. 2  E.g., 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… July 2000 (Leiden/Boston: Brill, 2002) 337–64, on pp. 355–6; A. Tjen, On Conditionals in the Greek Pentateuch: A Study of Translation Syntax (LHBOTS 515; London: T&T Clark, 2010), 179, 224. 3  E.g., T. V. Evans, Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference (Oxford: OUP, 2001), 140, n. 39; J. Schaper (see below). 4  J. Schaper, “The Concept of the Translator(s) in the Contemporary Study of the Septuagint”, in K. De Troyer/T. Law/M. Liljeström (ed.), In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014) 31–46, on p. 36. 5  Cf. van Klinken’s judgement: “In my view, this theory and its implications have not yet been evaluated sufficiently.” W. van Klinken, “From Literal to Free? A Study of Development in the Genesis Translation of the Septuagint” (M.A. Thesis; Free University of Amsterdam, 2006), 11–13, online at https://www.academia.edu/8197331/From_Literal_to_Free_A_Study_ of_Development_in_the_Genesis_Translation_of_the_Septuagint.

66

Theo A.W. van der Louw

notion of segmentation do not really know what to do with it. Take the following observation: G, through what may seem a rather loose approach to reflecting Hebrew syntax, in fact prompts his reader to ask each time what exactly is being mapped by the Greek. Here I would suggest, is not variation for its own sake or as a result of uneven, episodic work with smaller chunks of text, although this was indeed how translators went about their business.6 I think it is more appropriate to ascribe to G some intentionality. He draws the reader’s attention to Hebrew syntax by providing a variety of representations for it.7

Büchner thus acknowledges that segmentation defined the translation procedure, yet denies its practical consequences. Let us first define what we are talking about. Within Septuagint research, the term “segmentation” commonly refers to the procedure of translating a source text in units below sentence level, i.e., segments.8 These varied in length, from 2 to 7 words at a time. According to the segmentation hypothesis, each segment was read, translated and written down before the next segment was begun. It also assumes that segment boundaries coincided with clause boundaries, and if these were too long, with phrase boundaries.9 For example, Deut 10:22 could have been segmented as follows:10 ‫ֹכוכ ֵבי ַה ָּׁש ַמיִם ָלר ֹב׃‬ ְ ‫ֹלהיָך | ְּכ‬ ֶ ‫יְמה | וְ ַע ָּתה ָׂש ְמָך יְ הוָ ה ֱא‬ ָ ‫ְּב ִׁש ְב ִעים נֶ ֶפׁש יָ ְרדּו ֲאב ֶֹתיָך ִמ ְצ ָר‬ Seen in this light, translation segments are the molecules of the Greek Pentateuch, and the words contained in them are their atoms. Physicists have long known how enlightening a study of molecules is. It is now common knowledge that what is going on inside and between molecules is intimately related to the secrets of the universe. If the segmentation hypothesis has any credibility, the same might be true for the molecules of the Septuagint. In the following, I want to review phenomena that have been ascribed to three aspects of segmentation before venturing an evaluation of this hypothesis. 6  In a footnote, Büchner refers to Soisalon-Soininen, “Segmentation”, 29. 7  D. Büchner, “Writing a Commentary on the Septuagint”, in M.K.H. Peters (ed.), XIV Congress of the International Organization for Septuagint and Cognate Studies, Helsinki 2010 (SBLSCS 59; Atlanta: SBL, 2013) 525–38, on p. 533 (italics added). 8  Note that J. Barr, The Typology of Literalism in Ancient Biblical Translation (MSU 15; NAWG I, Phil-Hist. Kl.; Göttingen: Vandenhoeck & Ruprecht, 1979), 20–29, defines “segmentation” differently, viz. as the division of words into morphemes. 9  The issues of segment boundaries and segment length must be reserved for future treatment. 10  For proposals of segment boundaries by Soisalon-Soininen, cf. “Segmentation”, 35, 37, 38.

The Dynamics of Segmentation in the Greek Pentateuch

67

1.1 Losing Touch with the Content of Preceding Segments Segmentation has often been invoked to explain incorrect syntactic structures: the translator, it is said, forgot the preceding segments and lost touch with the sentence construction he was rendering, which resulted in all sorts of syntactical oddities. a) In conditional sentences, the indicative sometimes occurs too soon, viz. in the protasis. This happens in, e.g., Lev 6:2–3 [MT 5:21–22],11 Ψυχὴ έὰν ἁμάρτῃ καὶ παριδὼν παρίδῃ τὰς ἐντολὰς κυρίου, καὶ ψεύσηται τὰ πρὸς τὸν πλησίον ἐν παραθήκῃ ἢ περὶ κοινωνίας ἢ περὶ ἁρπαγῆς, | ἢ ἠδίκησέν τι τὸν πλησίον …

The same mechanism is operative in long sentences where the translator forgets he is rendering a grammatical object and lapses into the nominative.12 b) In Lev 17:8 the translator kept the general partitive idea in mind, but switched from a partitive genitive to ἀπό + genitive,13 Ἄνθρωπος ἄνθρωπος τῶν νἱῶν Ἰσραὴλ | καὶ ἀπὸ τῶν υἱῶν τῶν προσηλύτων τῶν προσκειμένων ἐν ὑμῖν …

c) Quite frequent are cases of apodotic καί, e.g., in Lev 5:4–5, 11, 15, 17–18,14 whereby the general pattern is that apodotic καί is more liable to occur the longer the protasis is.15 ψυχὴ ἐὰν λάθῃ αὐτὸν λήθῃ καὶ ἁμάρτῃ ἀκουσίως ἀπὸ τῶν ἁγίων κυρίου, | καὶ οἴσει τῆς πλημμελείας αὐτοῦ τῷ κυρίῳ κριὸν ἄμωμον … (Lev 5:15)

d) Cases of disagreement: in Exod 29:29 the rendering of the Hebrew plural with a singular caused problems further on in the sentence: ‫א־בם ֶאת־יָ ָדם׃‬ ָ ‫ּול ַמ ֵּל‬ ְ ‫ּובגְ ֵדי ַהּק ֶֹדׁש ֲא ֶׁשר ְל ַא ֲהר ֹן יִ ְהיּו ְל ָבנָ יו ַא ֲח ָריו ְל ָמ ְׁש ָחה ָב ֶהם‬ ִ

καὶ ἡ στολὴ τοῦ ἁγίου, ἥ ἐστιν Ἀαρών, ἔσται τοῖς υἱοῖς αὐτοῦ μετʼ αὐτόν, | χρισθῆναι αὐτοὺς ἐν αὐτοῖς καὶ τελειῶσαι τὰς χεῖρας αὐτῶν.

11 Soisalon-Soininen, “Segmentation”, 34; more extensively, Tjen, Conditionals, 152–8. 12  Examples: T. Muraoka, A Syntax of Septuagint Greek (Leuven: Peeters, 2016), §90f. 13 Soisalon-Soininen, “Segmentation”, 31. Interestingly, this example of segmentation occurs in a long sentence that is otherwise correct, i.e., without apodotic καί. 14 Further examples: Soisalon-Soininen, “Segmentation”, 33–35; Aejmelaeus, Parataxis, 126–47; Tjen, Conditionals, 215–218; Muraoka, Syntax, §89d, 90g. 15  Van Klinken, “From Literal to Free”, 23.

68

Theo A.W. van der Louw

e) Variation that is not of a stylistic nature. In Gen 9:2 ἐπί is first construed with the dative, and then shifts to the accusative:16 καὶ ὁ τρόμος ὑμῶν καὶ ὁ φόβος ἔσται ἐπὶ πᾶσιν τοῖς θηρίοις τῆς γῆς καὶ ἐπὶ πάντα τὰ ὄρνεα τοῦ οὐρανοῦ καὶ ἐπὶ πάντα τὰ κινούμενα … (+ 3 acc.)

Although this variation tells something about the style (substandard), it is not likely to have been employed for stylistic reasons. In that case, the writer would have varied the preposition rather than the case.17 f) The pleonastic (or resumptive) pronoun in relative clauses has been explained through segmentation.18 However, since the pleonastic pronoun often occurs in short relative clauses, and thus plausibly within a segment, it cannot be the result of the translator losing track.19

1.2 Free Translation within Segment Boundaries Segmentation has long been considered a constraining force, and rightly so. But there is another side to it, for within segment boundaries there is nothing to constrain the translator. It is there in that limited space that we find participium ­coniunctum, genitive absolute, creative renderings of construct states, explicitations, etc. To the examples found elsewhere,20 let me add a few. ‫ת־הּיֶ ֶלד וְ ִהּנֵ ה־נַ ַער ּב ֶֹכה‬ ַ ‫וַ ִּת ְפ ַּתח וַ ִּת ְר ֵאהּו ֶא‬

ἀνοίξασα δὲ ὁρᾷ παιδίον κλαῖον ἐν τῇ θίβει (Exod 2:6)

‫אתי ָלְך ִא ָּׁשה ֵמינֶ ֶקת ִמן ָה ִע ְב ִרּיֹת‬ ִ ‫ַה ֵא ֵלְך וְ ָק ָר‬

Θέλεις καλέσω σοι γυναῖκα τροφεύουσαν ἐκ τῶν Ἑβραίων (Exod 2:7)

‫ּומ ְל ֶּתם ֵאת ָע ְר ַלת ְל ַב ְב ֶכם‬ ַ

καὶ περιτεμεῖσθε τὴν σκληροκαρδίαν ὑμῶν (Deut 10:16)

‫ִּכי ָענִ י הּוא וְ ֵא ָליו הּוא נ ֵֹׂשא ֶאת־נַ ְפֹׁשו‬

ὅτι πένης ἐστὶν καὶ ἐν αὐτῷ ἔχει τὴν ἐλπίδα (Deut 24:15)

16  Cf. also Gen 49:26 (ἐπί + acc./gen.); Deut 14:26 (ἐπί + gen./dat./gen.). 17  According to J.A.L. Lee (personal communication). 18 Soisalon-Soininen, “Segmentation”, 36–38. 19  That is why R. Sollamo did not invoke segmentation to explain the pleonastic pronoun in “The Pleonastic Use of the Pronoun in Connection with the Relative Pronoun in the Greek Pentateuch”, in C.E. Cox (ed.), VII Congress of the IOSCS, Leuven 1989 (SBLSCS 31; Atlanta: SBL, 1991), 75–85; idem, “The Pleonastic Use of the Pronoun in Connection with the Relative Pronoun in the LXX of Leviticus, Numbers and Deuteronomy,” in L. Greenspoon/O. Munnich (ed.), VIII Congress of the IOSCS, Paris 1992 (SBLSCS 41; Atlanta: SBL, 1995), 43–62. 20 Soisalon-Soininen, “Back to the Question of Hebraisms” (in this volume), 43; “Segmentation”, 32; “A Palette of Renderings for the Hebrew Construct State” (in this volume), 62–70.

The Dynamics of Segmentation in the Greek Pentateuch

69

Quantitative representation, i.e., the adherence to the number of words of the source text, is often considered a general trait of the Septuagint. But it does not appear to have been a systematic concern of the Pentateuch translators on the segment level. Additions and omissions are quite frequent, actually. Additions outnumber omissions, which creates a longer text.21 For example, the total of additions and omissions in Gen 41 is: + 40 − 32. Soisalon-Soininen explains this preponderance of additions as follows: From a psychological viewpoint, this is understandable. If something drops out, the translator instinctively feels that the translation is incomplete. But if he adds something, he is only perfecting his rendering from the perspective of Greek.22

An increasing focus on communication with the target audience results in a greater number of additions as the translator employs more explications and clarifying additions. The difference between Gen 2 (+ 27 − 24)23 and Exod 2 (+ 48 − 20) speaks volumes with respect to the translational outlook of these two chapters.24 The bulk of the figures just mentioned concern function words (conjunctions, particles, pronouns, etc.) rather than content words (verbs, nouns, adjectives), which the translators were reluctant to leave untranslated. See the omissions (Ø) and additions (underlined) in Gen 44:26: ‫נּוכל ָל ֶר ֶדת ִאם־יֵ ׁש‬ ַ ‫לֹא‬ ‫ָא ִחינּו ַה ָקטֹן ִא ָתנּו וְ יָ ַר ְדנּו‬ ‫נּוכל ִל ְראֹות ְפנֵ י ָה ִאיׁש‬ ַ ‫ִכי־לֹא‬ ‫וְ ָא ִחינּו ַה ָקטֹן ֵאינֶ ּנּו ִא ָתנּו‬

Οὐ δυνησόμεθα καταβῆναι· ἀλλʼ εἰ μὲν ὁ ἀδελφὸς ἡμῶν ὁ νεώτερος καταβαίνει μεθʼ ἡμῶν, Ø καταβησόμεθα· οὐ γὰρ δυνησόμεθα ἰδεῖν τὸ ­πρόσωπον τοῦ ἀνθρώπου, Ø τοῦ ἀδελφοῦ τοῦ νεωτέρου μὴ ὄντος μεθʼ ἡμῶν.

21  That translations are generally longer than their source texts has been suggested as a “translation universal”. Cf. 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, Ljubljana 2007 (SBLSCS 55; Atlanta: SBL, 2008), 339–51. 22 Soisalon-Soininen, “Der Gebrauch des Verbs ἔχειν in der Septuaginta”, in Studien zur Septuaginta-Syntax, 181−8, on p. 184 (my translation). 23  The untypical rate + 27 − 31 for Gen 2 becomes + 27 − 24 if we ignore the translator’s handling of the divine name; cf. T.A.W. van der Louw, Transformations in the Septuagint: Towards an Interaction of Septuagint Studies and Translation Studies (CBET 47; Leuven: Peeters, 2007), 146–7, 362. 24  Genesis 2 and Exodus 2 are of comparable length (329 and 341 Hebrew words respectively).

70

Theo A.W. van der Louw

1.3 Unsuccessful Anticipation of Coming Segments The idea that the translator did not always correctly anticipate coming segments was formulated by Soisalon-Soininen as follows: We may also surmise that the influence of the preceding text was stronger than that of the following. What had already been translated remained in the translator’s memory. But he did not see the need of consulting in advance a passage that would follow.25

A good example is Num 22:2–3, where the translator uses a participium coniunctum, unaware that the subject of his main clause will not be identical with the subject of the part. coni. (Moab ≠ Balak).26 ‫ן־צּפֹור ֵאת ָכל־‬ ִ ‫וַ יַ ְרא ָב ָלק ֶב‬ ‫ר־ע ָשה יִ ְש ָר ֵאל ָל ֱאמ ִֹרי׃‬ ָ ‫ֲא ֶש‬ ‫מֹואב ִמ ְפנֵ י ָה ָעם ְמאֹד‬ ָ ‫ | וַ יָ גָ ר‬3

ἰδὼν Βαλὰκ υἱὸς Σεπφὼρ πάντα, ὅσα ἐποίησεν Ἰσραὴλ τῷ Ἀμορραίῳ, 3 | καὶ ἐφοβήθη Μωὰβ τὸν λαὸν σφόδρα 2 Καὶ

Ignorance of coming segments explains why part. coni. (and genitive absolute) occurs mainly towards the end of the sentence. Compare three renderings of ‫ב‬ + inf. cs. in the same book (underlined), followed by Aejmelaeus’ elucidation of their rationale. ἐν τῷ συμπορεύεσθαι πάντα Ἰσραὴλ ὀφθῆναι ἐνώπιον κυρίου τοῦ θεοῦ σου ἐν τῷ τόπῳ, ᾧ ἂν ἐκλέξηται κύριος, ἀναγνώσεσθε τὸν νόμον τοῦτον (…) (Deut 31:11) καὶ ἔσται ἡνίκα ἐὰν καταπαύσῃ σε κύριος ὁ θεός σου ἀπὸ πάντων τῶν ἐχθρῶν σου (…), ἐξαλείψεις τὸ ὄνομα Ἀμαλὴκ ἐκ τῆς ὑπὸ τὸν οὐρανὸν καὶ οὐ μὴ ἐπιλάθῃ. (Deut 25:17) μνήσθητι ὅσα ἐποίησεν κύριος ὁ θεός σου τῇ Μαριὰμ ἐν τῇ ὁδῷ ἐκπορευομένων ὑμῶν ἐξ Αἰγύπτου. (Deut 24:9) Before starting a sentence with a participle—in most cases the participle precedes the main verb—he had to acquaint himself with the following words in advance in order to find out whether he could use the participle or not.27

25 Soisalon-Soininen, “Segmentation”, 30; A. Aejmelaeus, “Participium Coniunctum as a Criterion of Translation Technique”, in On the Trail of the Septuagint Translators: Collected Essays (CBET 50; Leuven: Peeters, 2007) 1–10, on p. 5. 26  The same happens in Exod 12:30. 27 Aejmelaeus, Parataxis, 108.

The Dynamics of Segmentation in the Greek Pentateuch

71

It was easier to fit in the part. coni. as a free rendering when the main clause was already there. The text that has already been translated has more influence upon a given rendering than the text that lies ahead, yet to be translated.28 In other words, the translator was constantly anticipating the sequel of the segment at hand. The more of the sentence he had already processed, the easier it became for him to anticipate correctly how the sentence construction would turn out. Therefore, he used a part. coni. only in the third example above (Deut 24:9), where he already knew the sentence as a whole. Uninformed anticipation often tripped the translator up in conditional sentences. Short ones posed no problem, but if the protasis contained two or more clauses, the indicative often occurred too soon. Two factors affected the translator. Not only did he lose his awareness of the protasis, which would require subjunctives (cf. section 1), but at the same time he could not anticipate when the apodosis would begin. This is nicely illustrated by Num 32:20–23. The short conditional sentence in v. 23 is rendered correctly, whereas in the long protasis of vv. 20–22a the translator used the indicative too soon. He therefore incorrectly turned v. 21a into the apodosis (underlined). But—as the parallel with v. 23 suggests—the apodosis rather begins in v. 22c “you shall be clear before the Lord.”29 Compare the New English Translation of the Septuagint (NETS) with a correct interpretation of the Hebrew syntax (NJPS). Ἐὰν ποιήσητε κατὰ τὸ ῥῆμα τοῦτο, ἐὰν ἐξοπλίσησθε ἔναντι κυρίου εἰς πόλεμον 21καὶ παρελεύσεται ὑμῶν πᾶς ὁπλίτης τὸν Ἰορδάνην ἔναντι κυρίου, ἕως ἂν ἐκτριβῇ ὁ ἐχθρὸς αὐτοῦ ἀπὸ προσώπου αὐτοῦ 22καὶ κατακυριευθῇ ἡ γῆ ἔναντι κυρίου, καὶ μετὰ ταῦτα ἀποστραφήσεσθε, καὶ ἔσεσθε ἀθῷοι ἔναντι κυρίου καὶ ἀπὸ Ισραήλ, καὶ ἔσται ἡ γῆ αὕτη ὑμῖν ἐν κατασχέσει ἔναντι κυρίου. 23ἐὰν δὲ μὴ ποιήσητε οὕτως, ἁμαρτήσεσθε ἔναντι κυρίου καὶ γνώσεσθε τὴν ἁμαρτίαν ὑμῶν, ὅταν ὑμᾶς καταλάβῃ τὰ κακά. LXX/NETS (20 …) If you do according to this word—if you arm yourselves before the Lord for war, (21) then every armed soldier of you shall pass the Jordan before the Lord until his enemy is wiped out from before him (22) and the land is dominated before the Lord. And after this you shall turn back and you shall be innocent before the Lord and from Israel, and this land shall be to you as a possession before the Lord. 23 But if you will not do so, you will sin against the Lord, and you will know your sin when misfortunes overtake you. MT/NJPS (20…) If you do this, if you go to battle as shock-troops, at the instance of the Lord, (21) and every shock-fighter among you crosses the Jordan, at the instance of the Lord, until He has dispossessed His enemies before Him, (22) and the land has been subdued, at the instance of the Lord, and then you return, you shall 28 Aejmelaeus, “Participium Coniunctum”, 5. 29  Some versions (REB, NIV, ESV) start the apodosis one clause before that, viz. in 22b.

72

Theo A.W. van der Louw

be clear before the Lord and before Israel; and this land shall be your holding under the Lord. (23) But if you do not do so, you will have sinned against the Lord; and know that your sin will overtake you.

Authors sometimes speak of losing touch with the preceding segment, where rather incorrect anticipation is the issue. Such is the case with Deut 11:21:30 ‫יכם‬ ֶ ֵ‫ימי ְבנ‬ ֵ ִ‫יכם | ו‬ ֶ ‫יְמ‬ ֵ ‫ְל ַמ ַען יִ ְרּבּו‬

ἵνα πολυημερεύσητε | καὶ αἱ ἡμέραι τῶν υἱῶν ὑμῶν

The translator joined the verb ‫ רבה‬and the noun ‫ ימים‬into πολυημερεύειν—a free rendering within segment boundaries—because he did not expect the following segment. Otherwise he would have chosen a verb that would have left room for adding “the days of your sons” to the subject.31 Such renderings prompt the question, “Why did the translator not go back to correct this rendering?”32 which is tantamount to suggesting that in hindsight the translator would have put something different. The notion of hindsight leads to the diagnosis of “unsuccessful anticipation”.

2. Alternative Theories If the array of phenomena listed above can be explained otherwise, the segmentation hypothesis becomes superfluous. Let us therefore review two alternative explanations. a) During the symposium the objection was raised that word-for-word translation—which presupposes one-word segments, really—as a strategy would yield the same results as segmentation as envisioned in this paper. However, the free renderings described in section 2 (above) and the supple attitude towards quantitative representation make it doubtful that word-for-word translation was a conscious strategy. Besides, the translation of some features varies according to syntactic parameters. For example, the likelihood of apodotic καί increases in proportion to the length of the protasis. In a word-for-word approach, the length of the protasis would obviously play no role. 30  Quoted by Soisalon-Soininen, “Segmentation”, 30. 31  For this reason, I accept Soisalon-Soininen’s assumption of two unusually short segments in this case. It is hard to see how this incorrect syntax could have arisen within a segment. 32 Soisalon-Soininen, “Segmentation”, 29; Aejmelaeus, Parataxis, 105; Emanuel Tov, “The Septuagint Translation of Genesis as the First Scripture Translation”, in K. De Troyer/T. Law/ M. Liljeström (ed.), In the Footsteps of Sherlock Holmes. Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014) 47–64, on p. 50.

The Dynamics of Segmentation in the Greek Pentateuch

73

b) Others point out that many of the phenomena listed above have their parallels in original Greek texts. They isolate a syntactical feature and compare it with (near) parallels from Greek texts. Le Moigne seeks to demonstrate that in conditional relative clauses a subjunctive is sometimes followed by an indicative both in the Septuagint and in classical Greek.33 Apart from the fact that these authors seek their parallels in disparate corners, they only account for subsets of what are in effect larger categories. Instances of discord in lists are part and parcel of the overall phenomenon of grammatical discord in the LXX and thus of a piece with cases of discord that have no parallels in Greek texts, such as Exod 29:29: καὶ ἡ στολὴ τοῦ ἁγίου, ἥ ἐστιν Ἀαρών, ἔσται τοῖς υἱοῖς αὐτοῦ μετʼ αὐτόν, | χρισθῆναι αὐτοὺς ἐν αὐτοῖς καὶ τελειῶσαι τὰς χεῖρας αὐτῶν.

Although instances of a subjunctive followed by an indicative in the same clause may have near parallels in classical Greek and might allow for the sophisticated interpretation that Le Moigne gives them, they cannot be separated from the issue of subjunctives followed by indicatives in the LXX as a whole. And if we see that the translators often “corrected” themselves later in similar clauses of the same context, it appears that these instances of syntactical discord are not so idiomatic as one might make it appear. The example of Num 32:20–23 (section 3) makes that clear. Given the parallel nature of both conditionals it is far-fetched to postulate a “pré-actualisation” in Num 32:21, as Le Moigne would do. Parallel clauses in Lev 21:17–21 push Le Moigne to the limits of his approach. His problem is that subjunctives and indicatives appear alternately in conditional relative clauses that are syntactically on a par. That makes it recherché to explain the alternation with the distinction between “éventuel” and “réel definitionel,” as Le Moigne acknowledges. ‫ל־א ֲהר ֹן‬ ַ ‫ַד ֵבר ֶא‬ ‫ִאיׁש ִמזַ ְר ֲעָך ְלדֹר ָֹתם‬ ‫ֲא ֶשר יִ ְהיֶ ה בֹו מּום‬ ‫לֹא יִ ְק ַרב ְל ַה ְק ִריב‬ ‫ֹלהיו׃‬ ָ ‫ֶל ֶחם ֱא‬ ‫ל־איׁש ֲא ֶשר־ּבֹו מּום‬ ִ ‫ִכי ָכ‬ ‫לֹא יִ ְק ָרב‬ ‫ִאיׁש ִעוֵ ר אֹו ִפ ֵס ַח‬ ‫רּוע׃‬ ַ ‫אֹו ָח ֻרם אֹו ָש‬ ‫אֹו ִאיׁש‬

17 Εἶπον Ἀαρών∙ Ἄνθρωπος ἐκ τοῦ γένους σου εἰς τὰς γενεὰς ὑμῶν, τίνι ἐὰν ᾖ ἐν αὐτῷ μῶμος, οὐ προσελεύσεται προσφέρειν τὰ δῶρα τοῦ θεοῦ αὐτοῦ. 18 πᾶς ἄνθρωπος, ᾧ ἂν ᾖ ἐν αὐτῷ μῶμος, οὐ προσελεύσεται, ἄνθρωπος χωλὸς ἢ τυφλὸς ἢ κολοβόρριν ἢ ὠτότμητος 19 ἢ ἄνθρωπος,

33  Ph. Le Moigne, “Au delà du réel: recherches sur le subjonctif éventuel de l’Octateuque” (habilitation, Université de Strasbourg, 2014), 231.

74

Theo A.W. van der Louw

‫ֲא ֶשר־יִ ְהיֶ ה בֹו‬ ‫ֶש ֶבר ָרגֶ ל אֹו ֶש ֶבר יָ ד׃‬ ‫אֹו־דק אֹו ְת ַב ֻלל ְב ֵעינֹו‬ ַ ‫אֹו־גִ ֵבן‬ ‫רֹוח ָא ֶשְך׃‬ ַ ‫אֹו גָ ָרב אֹו יַ ֶל ֶפת אֹו ְמ‬ ‫ל־איׁש ֲא ֶשר־ּבֹו מּום‬ ִ ‫ָכ‬ ‫ִמזֶ ַרע ַא ֲהר ֹן ַהכ ֵֹהן‬ ‫ת־א ֵשי יהוה‬ ִ ‫לֹא יִ גַ ׁש ְל ַה ְק ִריב ֶא‬ ‫מּום ּבֹו‬ ‫ֹלהיו לֹא יִ גַ ׁש‬ ָ ‫ֵאת ֶל ֶחם ֱא‬ ‫ְל ַה ְק ִריב׃‬

ᾧ ἐστιν ἐν αὐτῷ σύντριμμα χειρὸς ἢ σύντριμμα ποδός, 20 ἢ κυρτὸς ἢ ἔφηλος ἢ πτίλος τοὺς ὀφθαλμοὺς ἢ ἄνθρωπος, ᾧ ἂν ᾖ ἐν αὐτῷ ψώρα ἀγρία ἢ λιχήν, ἢ μόνορχις, 21 πᾶς, ᾧ ἐστιν ἐν αὐτῷ μῶμος, ἐκ τοῦ σπέρματος Ααρων τοῦ ἱερέως, οὐκ ἐγγιεῖ τοῦ προσενεγκεῖν τὰς θυσίας τῷ θεῷ σου, ὅτι μῶμος ἐν αὐτῷ, τὰ δῶρα τοῦ θεοῦ οὐ προσελεύσεται προσενεγκεῖν.

The passage opens with two relative conditionals that correctly have the subjunctive. Verse 17, with its quotation formula, opens a new legislative section, which guided the translator in understanding ‫ ִאיׁש ִמזַ ְר ֲעָך ְלדֹר ָֹתם‬as a casuistic casus pendens that awaits an apodosis. He correctly anticipated the sentence structure, which resulted in a protasis with subjunctive and an apodosis in the indicative. In v. 18, a verse with a similar content and short clauses, he had no trouble repeating this procedure. In v. 19, the indicative ᾧ ἐστιν is irregular, proven by the widely attested correction ᾧ ἂν ᾖ.34 Where Le Moigne wrestles with it, and candidly admits that a “réel definitionel” is a shaky explanation of this indicative, the segmentation hypothesis allows us to go the second mile. It does more than trying to make sense of the product; it also explains the process. Verse 18 closes with a string of words for handicapped people separated by ‫אֹו‬, “a person who is lame or blind, or having a mutilated nose or a cut ear”. This string of adjectives made the translator focus on a very different handicap each time, the last one even (probably) leading him to coin the neologism κολοβόρριν. This made him loose contact with the overall syntax. Receiving his text piecemeal, he then interpreted v. 19a (which starts with ‫)אֹו‬, as a continuation of the list of handicaps (18d). And since he had been “transported” to the world of handicaps, he was describing these attributes as he mentally pictured them. Thus, Le Moigne’s term “réel” for ᾧ ἐστιν in 19a is close to the mark, not as a feature of Greek grammar, but as a description of a mental process. In v. 20 the translator realized his mistake and now put the correct subjunctive ᾧ ἂν ᾖ. Such micro-developments are typical of the Pentateuch.35 In v. 21 the manuscripts accept the indicative ᾧ ἐστιν without

34  Cf. Wevers’ apparatus and Le Moigne, “Au delà du réel”, 231–2. 35  Van der Louw, “The Evolution of the Genesis Translator”, in M. Meiser et al. (ed.), Die Septuaginta. Geschichte—Wirkung—Relevanz; 6. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal, 21–24. Juli 2016 (WUNT 405; Tübingen: Mohr Siebeck, 2018) 146–57.

The Dynamics of Segmentation in the Greek Pentateuch

75

demur. Therefore, I concur with Le Moigne that the indicative can be appropriate here.36 This verse no longer deals with details but is a closing statement that recapitulates the drift of the section.

3. Segmentation and the Workings of the Human Brain Identifying linguistic parallels of syntactical “oddities” in ancient sources helpfully shows that such features did occur in original Greek texts. Nevertheless, it fails to explain the distribution of such features in LXX Greek. Why do they occur in some places and not in others (often in the same context)? That is the ­challenge of the parallel clauses in Lev 21, Num 22, the distribution of apodotic καί, and many other examples listed above. Segmentation is able to account for all those features through its appeal to the workings of the human brain. It gives a unified explanation of a wide variety of phenomena, especially when linked to an actual practice of translation. Source texts must be segmented before they can be translated. Because of memory limitations it is next to impossible to render a text of even two paragraphs without segmenting it, first into paragraphs, then into sentences etc. Process-oriented research reveals that translators often render clause by clause instead of sentence by sentence, and when the clauses are long or complex, they work by units below clause level.37 This can be captured through recordings of translators who have been trained to think aloud (Think Aloud Protocols). But it usually leaves no visible traces in the written target text, as it is revised one or more times to make it sound natural. Segmentation thus being a given, we shall now review the three features traditionally ascribed to segmentation to explore their relation to human brain ­capacity. (1) What does it imply that the translator lost touch with the preceding segment? To resume the example of the conditional clause, we have evidence that in rendering a long protasis, the translator lost his awareness of its syntactical nature which would have required a continuation of the subjunctive. So, he began a translational strategy but lapsed into the unmarked form, the indicative. The translator lost touch with the preceding segments because his brain capacity became overloaded. He had to keep retaining information from preceding

36  Cf. H.W. Smyth, Greek Grammar (Cambridge, MA: Harvard University Press, 1956), §2572, 2573. 37  K. McTait/M. Olohan/A. Trujillo, “A Building Blocks Approach to Translation Memory”, in Translation and the Computer 21: Proceedings of the 21st International Conference on Translating and the Computer (London: Aslib, 1999).

76

Theo A.W. van der Louw

s­egments, absorb and analyse a new segment, and then produce a translation while monitoring his output. Such an overload has been demonstrated to be the classic problem, not of translators but of interpreters.38 Translators do not have to bother about k­ eeping track of what they have written, since the preceding segment is right there under their very eyes, on a fresh sheet. And if they want to review the position of the segment in the sentence as a whole, they can simply take another look at the source text to read the sentence again. This, interpreters cannot do. As I see it, then, there is a direct line from the effects of segmentation caused by brain capacity overload to the hypothesis that the LXX is the product of a process whereby the source text was recited, interpreted orally, and taken down.39 Recent research confirms the correlation between verbalizing and segmentation length: reading out the text leads to shorter segments.40 The type of interpreting that has segmentation as its characteristic feature is known as “short consec(utive interpreting).”41 An example from a sermon: Well, the good news is that because the price was so high Et la bonne nouvelle c’est parce que ce prix était tellement élèvé every sin we’ve ever committed n’importe quel péché nous ayons pû commettre

38  This is especially the case for simultaneous interpreting, where the listening/analysing/ memorizing phase and the production/monitoring phase occur simultaneously; cf. D. Gile, “Conference Interpreting as a Cognitive Management Problem”, in F. Pöchhacker/M. Shlesinger (eds.), The Interpreting Studies Reader (London/New York; Routledge, 2002) 163–76. Although in (short) consecutive interpreting the phases are separated timewise, the pressure on the brain capacity is still significant; cf. ibid. 167–8 and A. Kriston, “The Importance of Memory Training in Interpretation”, Professional Communication and Translation Studies 5,1–2 (2012) 79–86. For the constraining force of the cognitive load on oral Bible translation, see J. Karlik, “Interpreter-mediated Scripture: Expectation and Performance”, Interpreting 12 (2010) 160–85, on pp. 164–5. 39  T.A.W. van der Louw, “The Dictation of the Septuagint Version”, JSJ 39 (2008) 211–29. The idea that the translator heard the text was prefigured by Soisalon-Soininen, “Back to the Question”, 36: “The translator hears or reads the source text in small segments.” 40  A correlation between verbalizing and segmentation length was found by A.L. Jakobsen, “Effects of Thinking Aloud on Translation Speed, Revision and Segmentation”, in F. Alves (ed.), Triangulating Translation (Amsterdam: John Benjamins, 2003) 69–95, whereby thinking aloud during translation led to significantly shorter segments. The ancient practice of reading aloud makes such findings relevant to LXX studies. 41  F. Pöchhacker, “Consecutive Interpreting”, in K. Malmkjaer/K. Windle (eds.), The Oxford Handbook of Translation Studies (Oxford Handbooks in Linguistics; Oxford: Oxford University Press, 2011) 294–306, on pp. 295–6. “Consecutive interpreting” processes larger stretches of discourse, and relies on note-taking to ensure accuracy.

The Dynamics of Segmentation in the Greek Pentateuch

77

has been forgiven il a été pardonné.42

(2) Free renderings within segment boundaries not only occur frequently, they also occur side by side with literal renderings. Let us call this intrasegmental inconsistency. An interesting example comes from our first experiment, where the interpreter went to great lengths to make the cupbearer address Pharaoh in a polite way (Gen 41:9). The result is an elegant sentence with only one flaw, the final position of the time adjunct vandaag “today”: ‫ת־ח ָט ַאי ֲאנִ י ַמזְ ִּכיר ַהֹּיום‬ ֲ ‫ֶא‬

Als ik zo vrij mag zijn mijn fouten bij u in herinnering te roepen vandaag. “If I may make bold as to call my mistakes to memory with you today.”

The interpreter employed so many transformations that his brain capacity probably had no room left for changing the word order.43 Similar examples can be found in the Greek Pentateuch: ‫יה וַ ִּת ְב ָל ֵעם‬ ָ ‫ת־ּפ‬ ִ ‫ֲא ֶׁשר ָּפ ְצ ָתה ָה ָא ֶרץ ֶא‬

οὓς ἀνοίξασα ἡ γῆ τὸ στόμα αὐτῆς κατέπιεν αὐτούς (Deut 11:6)

An idiomatic participium coniunctum is juxtaposed to a pleonastic pronoun (οὕς … αὐτούς for the Hebrew ‫)א ֶׁשר‬. ֲ It stands to reason that in both cases the translator focused on the elegant rendering and had no mental space left to do the rest in equal style. (3) Anticipation is not a challenge for regular translators since they can first read the sentence as a whole and then render it clause by clause, but it is for interpreters, who receive the source text piecemeal.44 A classic example is interpreting relative clauses from German to English. The German verb comes at the end of the clause, but for the sake of grammatical correctness an English interpreter needs to produce it at the beginning. Great dexterity is required to guess the correct verb, and examples of (funny) mistakes abound.

42 J. Downie, “Intervention Length and Interpreter Performance in Short Intervention Consecutive Interpreting”, 2010, online at https://www.academia.edu/4504695/Intervention_ Length_and_Interpreter_Performance_in_Short_Intervention_Consecutive_Interpreting, 5. 43  Van der Louw, “Analysis of the First Interpreting Experiment”, 2007, online at https:// www.academia.edu/3080833/Analysis_of_the_First_Interpreting_Experiment_held_June_ 2007, 16. 44  F. Pöchhacker, Introducing Interpreting Studies (London/New York: Routledge, 2004), 133f.; idem, The Interpreting Studies Reader, 115, 138–40, 261.

78

Theo A.W. van der Louw

Soisalon-Soininen implausibly theorized that “the translator did not see the need of consulting in advance a passage that would follow” (quoted above). ­Rather, the interpreter processed pieces that were orally delivered to him. He had no access to the context apart from his previous familiarity with it. Examples like Deut 11:21 can be clarifying factors in the debate on LXX origins, because the idea that the source text could take the translator by surprise seems to suggest that the translators did not know the source text well, and that they cannot be regarded as exegetes in the standard use of that term.45 To my mind, the controversy on this point is largely terminological. Knowing the source text in advance as an exegete—which the translators probably did—does not mean that the linguistic sequel could always be successfully anticipated in moments of multitasking. For when a translator heard ‫יכם‬ ֶ ‫יְמ‬ ֵ ‫ל ַמ ַען יִ ְרּבּו‬,ְ then his knowledge of this particular linguistic context must have been profound indeed to produce ‫יכם‬ ֶ ֵ‫ימי ְבנ‬ ֵ ִ‫ו‬. In the light of parallel and similar formulas he might as well have anticipated ‫יכם נ ֵֹתן ָל ֶכם‬ ֶ ‫ֹלה‬ ֵ ‫ּב ָא ֶרץ ֲא ֶׁשר יְ הוָ ה ֱא‬.ָ 46

4. Excursus: Intrasegmental and Intersegmental Inconsistency The occurrence of literal and free renderings side by side—so characteristic of the Septuagint—is usually ascribed to inconsistency in the sense of thoughtlessness. But from the point of view of segmentation there must be two types of “inconsistency”. First there is intrasegmental inconsistency: literal and free renderings occurring side by side within segment boundaries (cf. section 2). A second type is intersegmental inconsistency which entails free renderings in some segments and literal ones in others. An idiomatic feature is followed by a Hebraism, which makes for two subsequent segments exhibiting different translation styles, e.g., a genitive absolute followed by apodotic καί (cf. section 1). Other examples are free renderings that occur in real life situations such as greetings (starring Πῶς ἔχετε in Gen 43:27), in passages that were familiar to the translator (Exod 2), and in vivid scenes. Such renderings stand in contrast to literal renderings in the surrounding context. A third category of intersegmental “inconsistency” occurs in those cases where a translator abandons a rendering and starts to use another, often a more natural rendering to represent the same Hebrew word. In the following example the translator first uses the literal ἰδού for ‫וְ ִהּנֵ ה‬, and a bit further on, in a similar context, the idiomatic εἰ.47

45  Cf. Schaper, “Concept”, 35–37 contra Pietersma, “New Paradigm”, 355–6. 46  Parallel: Deut 11:21, 30:16; similar: Deut 4:40; 5:16, 33; 11:9; 25:15 etc. 47  Cf. Tjen, Conditionals, 92–93.

The Dynamics of Segmentation in the Greek Pentateuch

79

‫וְ ָר ָאה וְ ִהּנֵ ה ָּפ ָׂשה ַהּנֶ גַ ע ְּב ִקיר ֹת ַה ָּביִת׃ וְ ִצּוָ ה ַהּכ ֵֹהן‬

καὶ ὄψεται τὴν οἰκίαν, καὶ ἰδοὺ διεχύθη ἡ ἁφὴ ἐν τοῖς τοίχοις τῆς οἰκίας· καὶ προστάξει ὁ ἱερεύς (Lev 14:39–40)

‫וְ ָר ָאה וְ ִהּנֵ ה ָּפ ָׂשה ַהּנֶ גַ ע ַּב ָּביִת ָצ ַר ַעת ַמ ְמ ֶא ֶרת ִהוא ַּב ַּביִת ָט ֵמא הּוא׃‬

καὶ ὄψεται· εἰ διακέχυται ἡ ἁφὴ ἐν τῇ οἰκίᾳ, λέπρα ἔμμονός ἐστιν ἐν τῇ οἰκίᾳ, ἀκάθαρτός ἐστιν (Lev 14:44)

Another series occurs in Gen 27, where Isaac and Rebecca address their sons in an increasingly natural way, υἱέ μου (1), υἱέ (2), τέκνον (8, 13, 18), ὦ τέκνον (20), all for ‫בני‬, but the translator finally settles on τέκνον in 21, 25.48 Such cases testify to translational development rather than to mere variation.49

5. Conclusions The segmentation theory is capable of explaining 1. how the translator lost touch with preceding segments through brain capacity overload in the effort involved in intake, analysis and output; 2. how intrasegmental inconsistency came about through brain capacity overload; 3. the translator’s unsuccessful anticipation of coming segments. These issues are typical of (consecutive) interpreting, which suggests that the Greek Pentateuch was translated orally in a setup whereby the source text was recited in segments, translated orally and taken down in writing. Combining the twin concepts of intrasegmental and intersegmental inconsistency helps reduce what we usually call “inconsistency” to manageable proportions. For it appears that intrasegmental inconsistency is related to limitations of the short-term memory while intersegmental inconsistency is related to, inter alia, losing touch with the preceding segments (section 1), or a translational development.50 Thus, our study of translational molecules is able to shed light on 48  More examples in T.A.W. van der Louw, “Did the Septuagint Translators Really Intend the Greek Text as it is?”, in S. Kreuzer/M. Meiser/M. Sigismund (eds.), Die Septuaginta. Orte— Intentionen; 5. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal, 24–27. Juli 2014 (WUNT 361; Tübingen: Mohr Siebeck, 2016) 449–66, on p. 458. 49  For more on this issue, cf. van der Louw, “Evolution”. 50  For the sake of completeness, I must add a factor, indeed an avenue of research, viz. that the Old Greek Pentateuch probably underwent a literalizing revision; thus (building on the work of Ulrich, Tov and myself) I. Himbaza, “What are the Consequences if 4QLXXLeva Contains the Earliest Formulation of the Septuagint?” in Die Septuaginta—Orte und Intentionen, 294–308. Such a revision may have changed numerous idiomatic renderings for Hebraizing renderings that were germane to its purpose, but not in a rigid way, first because “trivial”

80

Theo A.W. van der Louw

one of the most elusive features of the Greek Pentateuch, viz. its “inconsistency”. I do not doubt that there a remains a category where thoughtlessness and lack of concentration did play a role, but its proportion is much smaller than is usually thought.

i­diomatic renderings were maintained and second because such an endeavour could never be consistent. In this way, another factor of inconsistency comes into view.

Anssi Voitila Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch 1. Introduction In this article, I will discuss the usage of the middle aorist form of the verb τιθέναι (‘to set, put, place s.th./s.o. somewhere’), comparing it with the related active and middle usages. The intention is to establish more firmly the meaning of this middle form and how it is used in the Septuagint Pentateuch. The middle aorist is chosen because in the Greek language there is a distinguishable passive form in the aorist (τεθῆναι). The verb τιθέναι is selected because its indirect reflexive middle (‘to put for oneself ’) meaning does not differ from the active as clearly as the meanings of verbs in the so-called “direct reflexive” (e.g. λούομαι ‘I wash [myself]’) or “reciprocal middle” (e.g. μάχομεθα ‘we fight [with each other]’) would. Therefore, the middle aorist of the verb τιθέναι lends itself to a more specific analysis of the other relevant factors that may separate it from its active and passive aorist counterparts. Likewise, it offers valuable information concerning the translators’ methods. The middle future (which also provides a distinguishable passive counterpart) would also be useful in this analysis, but as time and space are limited here, only some excursions into the middle future will be made. In recent linguistic research, the middle voice has been seen as indicating subject affectedness. Klaiman states that in classical Indo-European, voice indicates whether the Subject is to be construed as the participant most affected in consequence of the sententially denoted action, irrespective of its thematic relation to the verb.1

With “thematic relation,” she is referring to whether the subject of the verb functions as the controller/volitional animate agent of the action,2 as in I broke a window, where the subject is the agent and non-affected, or as the patient/victim of the action, as in The window broke, in which the grammatical subject is the patient and is affected. This means that voice is “unrelated to thematic role structure,” as it “signals not the subject’s thematic relation to the action, but its

1  M.H. Klaiman, “Affectedness and Control: A Typology of Voice Systems”, in M. Shibatani (ed.), Passive and Voice (Amsterdam/Philadelphia: Benjamins, 1988) 25–83, on p. 37. 2 Klaiman, “Affectedness and Control”, 27 and 29.

82

Anssi Voitila

c­ onceptual status as affected or non-affected”.3 In defining the different interpretations of middle meaning, Kemmer, Bakker and Allan have further developed this perspective, highlighting the importance of the verb’s lexical meaning, the event type of the verb (e.g., state, process, motion, emotion) and the type of sentence (e.g., transitive/intransitive, one- or two-participant) in which the middle form appears.4 Bakker sees the affectedness of the subject as the basic value of the middle, but he also argues that “the middle voice in ancient Greek is constrained by Aktionsart, both as to its meaning, and as to its relation with aspect.”5 He further explains that Aktionsart refers to the event-type, “differentiated on the basis of the transitivityrelated features of volition, agency and causation” (emphasis in original).6 Similarly, for the middle voice in Greek, Allan7 considers the affectedness of the subject to be the basic meaning, or “abstract” or “schematic” meaning, in Langacker’s terms.8 However, because the affectedness is often a rather abstract generalization, he points out the essentially polysemous nature of the middle voice category in Greek, referring to it as a complex network category.9 Barber offers the diagrams in Figure 1, in which the full middle corresponds to the more classical definition of indirect reflexive middle (SS = Sentence Subject):10

Figure 1: The active and middle situation types represented graphically. 3 Klaiman, “Affectedness and Control”, 37. 4  S. Kemmer, The Middle Voice (Amsterdam/Philadelphia: Benjamins, 1993); E.J. Bakker, “Voice, Aspect, and Aktionsart: Middle and Passive in Ancient Greek”, in B. Fox/P.J. Hopper (ed.), Voice: Form and Function (Amsterdam/Philadelphia: Benjamins, 1994) 23–47; R.J. Allan, The Middle Voice in Ancient Greek (Amsterdam: Gieben, 2003). 5 Bakker, “Voice, Aspect, and Aktionsart”, 44. 6 Bakker, “Voice, Aspect, and Aktionsart”, 44. 7 Allan, The Middle Voice in Ancient Greek, 39. 8  R.W. Langacker, Foundations of Cognitive Grammar, Vol. 1 (Stanford, CA: Stanford University Press, 1987), cited in Allan, The Middle Voice, 32–4. 9 Allan, The Middle Voice in Ancient Greek, 30–42, 118–24. 10  E.J.W. Barber, “Voice—Beyond the Passive” in Annual Meeting of the Berkeley Linguistics Society 1 (2014) 16–24, on p. 21.

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

83

Before this, Smyth had already stated that the middle voice indicated that the object entered into the “sphere” of the subject.11 Similarly, Benveniste noted that the focus was not so much on the affectedness of the subject, but on indicating that the subject and the action belonged together because the subject was interior to the process.12 This interiority of the subject is graphically represented in A and B in Figure 2, while C illustrates the active voice without subject interiority. In A and B in Figure 2, the square represents the subject S’s dominion as an area (domain) in which the event is situated, and S is shown as interior to the event. In A, B and C, then, the subject “volitionally initiate[s] physical activity resulting, through physical contact, in the transfer of energy to an external object”.13 The object is internal when the verb is in middle form, that is, it also is inside the square in figures A and B (Figure 2). The incoming arrow in square A represents the affectedness of the subject by the action; in B, the subject is unaffected. In contrast, in the active, represented in C, the connexion between the subject and the action are not in the focus.

Figure 2: Middle and active voice schemas for the verb τιθέναι.

Maldonado has gone one step further, claiming that the middle voice “focuses on the subject’s dominion” and “corresponds to situation types implying only the subject”,14 while the use of reflexive pronouns indicates a split representation of the subject’s/agent’s self,15 corresponding to situation types implying both the 11  H.W. Smyth, Greek Grammar (Cambridge, MA: Harvard University Press, 1956), 390– 92, cited in Klaiman, “Affectedness and Control”, 32. 12  “[L]e verbe indique un procès dont le sujet est le siège; le sujet est intérieur au procès”; É. Benveniste, “Actif et Moyen dans le Verbe”, in É. Benveniste, Problèmes de Linguistique Générale I (Paris: Gallimard, 1966) 168–75, on p. 172. 13 R.W. Langacker, Foundations of Cognitive Grammar, Vol. 2: Descriptive Application (Stanford, CA: Stanford University Press, 1991), 285. 14 R. Maldonado, “Middle as a Basic Voice System”, in E. Guerrero/L. Sergio Ibáñez/ V. Belloro (ed.), Studies in Role and Reference Grammar (México: Instituto de Investigaciones Filológicas, UNAM, 2009) 69–112: 70. 15 Kemmer, Middle Voice, 95–150.

84

Anssi Voitila

subject and the object. (Maldonado is referring to Spanish, in which the reflexive pronoun is used with middle voice.) In this view, the A and B in Figure 2 may also represent the middle aorist of the verb τιθέναι. In this article, I will demonstrate how this conception of the meaning of the middle voice explains more convincingly, in my mind, the use of the middle aorist forms of the verb τιθέναι in the Greek Pentateuch. Though I do not deny that subject affectedness is the most salient element most of the time, I will show that there are instances in which the interpretation of this form as indicating the subject’s dominion better explains the context as a whole. In my analysis, we will see that the translator has used the middle construction to highlight the subject’s position as an agent (that is, an actor with a relatively high level of power) and his/her close relation to the action and to the beneficiary of the action.

2. Statistics There are nine occurrences of the middle aorist of the verb τιθέναι (indic. ἔθετο) in Genesis, and eleven of active aorist forms (indic. ἔθηκεν, -αν, impv. θές). In Exodus, two middle aorist forms (indic. ἔθετο, impv. θέσθε) are found, along with seven active aorist forms (indic. ἔθηκεν, -αν, subjunctive θῇς). Finally, in Numbers, there are only three active forms, with no middle aorist forms. There are no occurrences of the passive aorist form of this verb in the Pentateuch. All the occurrences are given in the following table. Table 1. The occurrences of the aorist of the verb τιθέναι in the Pentateuch

Genesis Exodus Leviticus Numbers Deuteronomy

Middle occurrences

Active occurrences

1:17; 2:8,15; 4:15; 40:3; 41:10; 42:17, 30; 47:26 15:25; 32:27

15:10; 24:2, 9; 28:11; 30:40, 41; 31:37; 41:48; 48:20; 50:26 2:3; 34:12, 15; 40:20, 22, 24, 26 21:8; 24:21, 23

In these contexts, the corresponding Hebrew forms are the ground stem qatal (perfect), wayyiqtol (consecutive/inverted imperfect), weqatal (consecutive/inverted perfect), yiqtol (imperfect) or imperative of the verbs nātan (in the sense of ‘set, place s.th./s.o. somewhere’) and śīm (‘to set, put, place s.th./s.o. somewhere’). On one occasion, the causative stem wayyiqtol of the verb nūah (hiphil impf. cons.) and ground stem wayyiqtol of the verb ´āsaf (Gen 42:17, ‘take away to

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

85

prison’) appear as the Hebrew equivalents. None of these showed any particular correlation, however, with the voice chosen for the Greek translation. Since all of these Hebrew verb forms’ meanings may be considered active and transitive, and as the Hebrew language also has a reflexive stem, niphal, it may be concluded that the Hebrew verbs in the Vorlage were not what led the translator to render them as he did. Consequently, it seems obvious that the translator, as a native Greek speaker, followed his own linguistic knowledge here. It is curious that he resorted to the middle form, and in this article I hope to explain why he made this choice. Moreover, I shall examine other instances in which the translators of the Greek Pentateuch employed the middle voice instead of the active. The conceptualization of the event “Agent puts/places something or someone somewhere” differs from the prototypical transitive clause in which, according to Langacker’s definition, the subject (agent) deliberately (intention, volition, control) puts energy into an activity that, through physical contact, causes an object (patient) to undergo an internal change of state.16 Here, the patient does not undergo any internal physical change, but rather is relocated from one place to another.

3. τιθέναι Indicating Relocation We shall start with the instances in which τιθέναι signals, in its concrete meaning, the agent’s act of putting or placing something or someone somewhere, without the middle verb being part of an idiom or having any metaphorical meaning. At the beginning of Genesis, we encounter four instances in which God puts, sets or places something somewhere in the primeval narratives and in which the verbs indicating relocation are translated into Greek as middle aorist forms, although the Hebrew counterparts are actives both formally and semantically. God, as the agent or controller of the action, has full power over the realization of the action. (1) Gen 1:17 καὶ ἔθετο αὐτοὺς ὁ θεὸς ἐν τῷ στερεώματι τοῦ οὐρανοῦ ὥστε φαίνειν ἐπὶ τῆς γῆς. (wayyitten) And God set them in the firmament of the sky so as to give light upon the Earth. (2) Gen 2:8 Καὶ ἐφύτευσεν κύριος ὁ θεὸς παράδεισον ἐν Ἔδεμ κατὰ ἀνατολάς, καὶ ἔθετο ἐκεῖ τὸν ἄνθρωπον, ὃν ἔπλασεν. (wayyaśæm) And the Lord God planted a garden in Edem toward the east; and there he put the man whom he had formed.

16 Langacker, Foundations, Vol. 2, 285.

86

Anssi Voitila

(3) Gen 2:15 Καὶ ἔλαβεν κύριος ὁ θεὸς τὸν ἄνθρωπον, ὃν ἔπλασεν, καὶ ἔθετο αὐτὸν ἐν τῷ παραδείσῳ ἐργάζεσθαι αὐτὸν καὶ φυλάσσειν. (wayyanniḥ e-hū) And the Lord God took the man, whom he had formed, and put him into the garden to till and to keep it. (4) Gen 4:15 καὶ ἔθετο κύριος ὁ θεὸς σημεῖον τῷ Κάιν τοῦ μὴ ἀνελεῖν αὐτὸν πάντα τὸν εὑρίσκοντα αὐτόν. (wayyaśæm) And the Lord God set a mark for Cain, lest any finding him should kill him. (The formulation in (4) may suggest that the mark was not “upon Cain”, as the preposition le is generally interpreted in this verse in Hebrew, but “for him” as a separate object, for example, a written document.)

The patients of the action—the sun, the moon and the stars in (1), the man in (2) and (3), and the mark in (4)—do not undergo any change in their physical or mental form. Therefore, they are not prototypical patients, but themes in Saeed’s terms;17 instead, they are simply relocated to another place. It is also significant that the patients are all relocated to perform a function that defines their existence, which in (1)–(3) is to fulfil their purpose in God’s created world and in (4) is to prevent anybody from killing Cain.18 The traditional interpretations of the verb as ‘put someone/something for oneself ’19 or of the middle voice as indicating “the affectedness of the subject of the verb in, or by, the event denoted by the verb” seem untenable.20 The intention cannot be to indicate that God did these things for His own benefit, or that He underwent some sort of change by doing them. To establish the meaning of the middle form in (1)–(4), we shall first compare them to some active cases. Interestingly, there are no active aorist forms of ­τιθέναι with God as the subject in the Greek Pentateuch. We have to turn, first, to two instances, in (4a) and (4b), in which a human being functions as the subject.

17  “PATIENT is reserved for entities acted upon and changed by the verb’s action while THEME describes an entity moved in literal or figurative space by the action of the verb, but constitutionally unchanged” (emphasis in original); J.I. Saeed, Semantics (4th ed.; Chichester: Wiley Blackwell, 2015), 152. 18  The meaning ‘set, put, place’ is known elsewhere in Greek literature for the middle as well; see, for instance, in Homer, Odyssey, 20.387: ‘set a stool or chair for oneself ’, ἡ δὲ κατ᾽ ­ἄντηστιν ­θεμένη περικαλλέα δίφρον; Euripides, Hecuba 711: ἐμὸς ἐμὸς ξένος, Θρῄκιος ἱππότας, ἵν᾽ ὁ γέρων πατὴρ ἔθετό νιν κρύψας ‘My, my friend, Thracian horseman with whom his old father placed him (Hecuba’s son) in hiding’. Aesop. Fable 62 A farmer and a Snake, version 1.2: γ­ εωργὸς … ὄφιν … καὶ λαβὼν ὑπὸ κόλπον ἔθετο ‘a farmer … taking the snake up, placed it under his coat.’ 19  H.G. Liddell/R. Scott/H.S. Jones, A Greek-English Lexicon (Oxford: Oxford University Press, 1996), cited in Allan, The Middle Voice, 170. 20 Bakker, “Voice, Aspect, and Aktionsart”, 24.

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

87

(4a)  Gen 15:10 ἔλαβεν δὲ αὐτῷ πάντα ταῦτα καὶ διεῖλεν αὐτὰ μέσα, καὶ ἔθηκεν αὐτὰ ἀντιπρόσωπα ἀλλήλοις, τὰ δὲ ὄρνεα οὐ διεῖλεν. And he took for him all these and divided them in the middle and placed (wayyitten) them facing one another, but he did not divide the birds. (4b)  Gen 28:11 καὶ ἔλαβεν ἀπὸ τῶν λίθων τοῦ τόπου καὶ ἔθηκεν πρὸς κεφαλῆς αὐτοῦ, καὶ ἐκοιμήθη ἐν τῷ τόπῳ ἐκείνῳ. (wayyaśæm) And he took one of the stones of the place and put it at his head and lay down in that place.

In (4a), the narrator relates how Abraham carries out the orders of God. Abraham does not necessarily know the full purpose of his actions. The second instance, in (4b), enumerates subsequent events, in which the agent has a neutral role, having control over the event but without this control being the focus. To compare the aforementioned aorist middle instances in (1)–(4) with the active form of τιθέναι in contexts in which God acts as the agent, we have to turn, first, to two indicative future cases: in (5), God addresses Abraham, and in (6), Jacob. In (6), Jacob is repeating God’s previous words. (5) Gen 17:6 καὶ αὐξανῶ σε σφόδρα σφόδρα, καὶ θήσω σε εἰς ἔθνη, καὶ βασιλεῖς ἐκ σοῦ ἐξελεύσονται. (ūnetattī-kā) And I will make you exceedingly fruitful, and I will make nations of you, and kings will come out of you. (6) Gen 32:12 (13) Καλῶς εὖ σε ποιήσω καὶ θήσω τὸ σπέρμα σου ὡς τὴν ἄμμον τῆς θαλάσσης. (weśamtī) I will surely do you good, and make your seed as the sand of the sea.

In both instances, God is making a promise to a named patriarch, stating what He will do in the future. The meaning of the verb τιθέναι is still ‘to put something somewhere’, but now the relocation of the patient is more abstract. In fact, the patient is not relocated, but reformed (‘you are put into this form of appearance’), thus semantically functioning as a real patient undergoing a tangible change in its nature. The ‘you’ (σε) in (5) becomes more than one nation (‘you will be made nations’), and ‘the seed’ (τὸ σπέρμα) in (6) is multiplied (‘your seed will be as the sand of the sea’). The sentence focus is found there, in the content of the promise, not in the agent and His dominion. Next, we turn to (7). (7) Gen 9:13 τὸ τόξον μου τίθημι ἐν τῇ νεφέλῃ, καὶ ἔσται εἰς σημεῖον διαθήκης ἀνὰ μέσον ἐμοῦ καὶ τῆς γῆς. (nātattī) I do set my bow in the cloud, and it shall be for a token of a covenant between me and the Earth.

88

Anssi Voitila

(ἔσται middle future is an activity: “it will be after coming into existence, it will happen.”)

In (7), God is speaking using the first-person singular. The verb form is present indicative. At first glance, the phrase’s meaning seems identical to the ‘putting something somewhere’ meaning, at least to the one found in (1), ‘to put something in the sky’. With a second look, however, we note that although it is in the present tense, the verb form is punctual and instantaneous, and thus is a so-called “performative” (that is, an action that is completed by the very act of saying it). At the very same instant that the verb and its arguments are spoken, the act is performed; the subject/agent is not affected by the action during the performance of the words spoken, because the affectedness lies not inside but outside the placing event, at the moment the performative act is completed; that is why there is no room to express the subject’s dominion.21 If it were in the middle voice, this sentence would not be performative. Consequently, the usage of the middle form in (1)–(4) corresponds, in my view, to Maldonado’s interpretation of the middle construction as one that “focuses on the subject’s dominion”. The focus is not so much on affecting the subject, but on indicating that the subject and the action belong together. The middle, focusing on the subject/agent participant, indicates the subject’s high involvement in the action and condenses the event via its focusing function. In other words, the subject/agent participant of the middle sentence, albeit not entirely the same as the object/patient participant of the active sentence, is more connected to the event described by the middle verb, and as such, the middle verb gives more prominence to what is happening in the narrative. Maldonado terms such uses of the middle form “energetic middles”, which represent a construal “where some type of energy is profiled”.22 In these constructions, “the event is compressed to the crucial moment of change”, and the subject is acting upon the object to induce a change of state.23 In the cases examined here, this is a change in the location of the object. In (1)–(3), this means that God is personally committed to, 21  “High-transitivity causative events like ‘kill,’ on the other hand, lack inherent affectedness of the subject/agent: the event consists in an effect produced on an object/patient, the subject/agent standing, as an external cause, outside the event.” (Bakker, “Voice, Aspect, and Aktionsart”, 26) For similar reasoning regarding the nonexistence of aorist middle forms of the verb ἀποθνῄσκω and πίπτω, he states: “this affectedness is not the sort of affectedness that is coded by middle morphology: it lies not inside but outside the dying- or falling-event … ‘die’ is no more than the transitional point between the state of being alive and the state of being dead, not an event in/by which someone is affected,” see Bakker, “Voice, Aspect, and Aktionsart”, 29. 22 Maldonado, “Middle”, 96. Energetic constructions are defined as those that describe events in which some type of energy is supplied to bring about the event; Langacker, Foundations, Vol. 2, 389–93; see also “force dynamic” in L. Talmy, “Force Dynamics in Language and Thought”, Cognitive Science 12 (1988) 49–100. 23 Maldonado, “Middle”, 97.

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

89

and involved in, the acts of creation. In (4), the activity of “marking with a sign” equally involves God’s personal commitment; he enters into a sort of covenant with Cain. This conclusion is further confirmed by the fact that the middle form of τιθέναι often refers to the making of agreements, contracts, testaments, etc., in the documentary papyri of the period, as they appear in the Duke Database of Documentary Papyri.24

4. θέσθαι τὸν δεῖνα εἰς φυλακήν One statistically common instance of the middle aorist of τιθέναι in the Greek Pentateuch turns out to be the expression ‘to put someone in prison/a prison cell/custody’. It appears in Gen 40:3; 41:10; 42:1725, 30. It may be that the middle voice instantiates the personal nature of the event: the agent who puts the theme/ patient in prison is personally committed to the event, or the process happens in his/her interest, so the agent also takes on the semantic role of the beneficiary. The agent has all the control in the situation. In Gen 40:3 and 41:10, Pharaoh has become angry with his servants, and in Gen 42:17 and 30, Joseph, as ruler of Egypt, has put his brothers in prison to test them. The agent’s personal sphere is particularly salient in the middle usage when the patient is speaking about what has happened to him (Gen 41:10 and 42:30). Thus, the middle form depicts the agent’s domain, in which there is a close connection between the agent and the event. In contemporary Greek, the idea of ‘putting someone in prison’ is more fre­ quently expressed with an active aorist clause using the verb ἀν-, ἀπαγαγεῖν or βαλεῖν εἰς φυλακήν/ἐν φυλακῇ. It may be that in these cases the perspective is not that of the agent, as in the examples from Genesis, but that of the speaker, who is often complaining, among other things, about having been put in prison. Therefore, there is no need to highlight the agent’s perspective in the complaint, but the use of the active aorist clause appears more appropriate. Further, the active voice, as the unmarked member of the voice opposition, is used “even though, from a semantic point of view, the occurrence of the middle voice would be considered possible”.26 This may be the reason why the active form replaced the middle form in Gen 42:17 and Exod 32:27 in some manuscripts. Nevertheless, I have found two examples of middle voice forms of τιθέναι in the Greek Papyri that may have the agent’s personal sphere meaning; moreover, they are not aorist, but present, P.Petr. 2.5[a] (εἰς φυλακὴν τίθεσθαι), and perfect infinitives (8). 24  See http://papyri.info. 25  ἔθετο] ἔθηκεν 19´. 26 Allan, The Middle Voice, 24.

90

Anssi Voitila

(8) PSI. 4.367 ἀναγγέλλουσι γὰρ ἡμῖν τεθεῖσθαί σε αὐτὸν εἰς φυλακήν ‘they inform us that you have taken him into custody’

This expression appears more frequently with the preverb ἀπο (i.e., ἀποθέσθαι), as in (9), both in contemporary Greek literature (e.g., “ἀ. τινὰ εἰς φυλακήν” Plb.23.10.8; “ἐν φυλακῇ” Matt 14:3) and the papyri. (9) P.Eleph. 12 καθάπερ ὤιου δεῖν, γεγράφαμεν Πλειστάρχωι τῶι φυλακίτηι περὶ τῶν παροινησάντων σε ἀποθέσθαι αὐτοὺς εἰς τὴν φυλακήν ‘Just as you thought it necessary, we have written to Pleistarchus, the police agent, about those that have mistreated you to put them away in prison.’ (Other instances are P.Gur. 10.8; P.Tebt. 3.1.769.)

Furthermore, this formulation occurs in Lev 24:12 and Num 15:34, as seen in (10). (10) Lev 24:12 καὶ ἀπέθεντο αὐτὸν εἰς φυλακὴν διακρῖναι αὐτὸν διὰ προστάγματος κυρίου. (wayyanniḥ e-hu) ‘And they put him away in prison, to decide about him by the ordinance of the Lord.’

As the agent of the ‘putting someone in prison’ in these texts is the Israelite community in the wilderness, its acts are done by the authority of God. In the narrator’s perspective, it is in the interest of the Israelites to perform the juridical procedures just as God ordered. Here, we may conclude, the author uses the middle voice to indicate the agent’s sphere of action, but also to emulate a common phrase of the period. The agent’s personal commitment to the event is indicated in (11), in which the sons of Levi are taking an oath to act as God tells them to do. (11) Exod 32:27 Τάδε λέγει κύριος ὁ θεὸς Ἰσραὴλ Θέσθε27 (mid.impv.aor.pl.2.) ἕκαστος τὴν ἑαυτοῦ ῥομφαίαν ἐπὶ τὸν μηρόν. (śīmū) ‘This is what the Lord, the God of Israel, says, “Each one put his own sword on his thigh.’”

The subject’s dominion in this situation may be conceptualized such that the subjects (pl.2.) are in full control of the action; they are free men who are also capable of refusing to act as the speaker wishes. It is interesting that Maldonado’s “energetic” middles, which occur in Spanish, indicate, among other meanings,

27  Θέσθε] θέτε 707 n 71´ 426.

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

91

the subject’s high involvement in the action.28 This reading aligns very well with the oath-taking performance profiled in (11). Interestingly, the genitive of the reflexive pronoun is used as a determiner for the sword, making it ‘his own’. Similar oath-taking expressions are found in (11a) and (11b), in which an order and its execution are instantiated. Abraham orders his servant to show his obedience by not letting his son marry a Canaanite woman. (11a) Gen 24:2 Θὲς (act.impv.aor.sg.2.) τὴν χεῖρά σου ὑπὸ τὸν μηρόν μου …. (śīm) ‘Put your hand under my thigh.’ (11b) Gen 24:9 καὶ ἔθηκεν ὁ παῖς τὴν χεῖρα αὐτοῦ ὑπὸ τὸν μηρὸν Ἀβραὰμ τοῦ κυρίου αὐτοῦ καὶ ὤμοσεν αὐτῷ περὶ τοῦ ῥήματος τούτου. (wayyaśæm) ‘And the servant put his hand under the thigh of Abraham his lord and swore to him concerning this matter.’

The active voice is used because Abraham is the initiator and controller of the event, the agent is not highly involved and the feeling is neutral. Based on these instances, we may further conclude that the middle voice of τιθέναι indicates the agent’s dominion: he has full control over the execution of the action referred to and is often acting within his domain of power.

5. Giving Laws and Making Covenants The middle voice of τιθέναι often occurs when referring to the giving of laws, statutes, judgements and similar situations, already in Herodotus (12) and Isocrates (13), as well as in papyri (13a). (12) Herodotus 2.117.11 Σόλων δὲ ὁ Ἀθηναῖος λαβὼν ἐξ Αἰγύπτου τοῦτον τὸν νόμον Ἀθηναίοισι ἔθετο. ‘Solon, the Athenian, took this law from Egypt and gave to the Athenians’ (13) Isocrates, Panegyricus 40.1 πρώτη γὰρ καὶ νόμους ἔθετο καὶ πολιτείαν κατεστήσατο. ‘for she (The city of Athens) was the first to lay down laws and establish a polity.’ (13a) P.Lond. 7.2193 νόμος ὃν ἔθεντο [κα]τὰ κοινὸν οἱ ἐκ τῆς τοῦ Διὸς Ὑψίστου συνόδου τοῦτον εἶναι κύριον ‘The law which those of the association of Zeus the highest made in common, that it should be authoritative.’29 28 Maldonado, “Middle”, 105. 29  R.S. Bagnall/P. Derow, Greek Historical Documents: The Hellenistic Period (Chico: Scholars Press, 1981), § 162.

92

Anssi Voitila

In all three instances, the laws, statutes and judgements are made for the beneficiary, but they also benefit the agent, the lawgiver. Not only is he committed to the action as an authority who makes the law, but also, as a community member, the individual lawgiver profits when law and order prevail in the community. Further, the agent is in control of the law-giving; hence, the middle voice highlights the fact that the action belongs within the subject’s sphere. In this light, it is only natural that we also encounter such a use of the middle in the Septuagint Pentateuch, as in (14) and (15). (14)  Gen 47:26 καὶ ἔθετο αὐτοῖς Ἰωσὴφ εἰς πρόσταγμα ἕως τῆς ἡμέρας ταύτης ἐπὶ γῆν Αἰγύπτου ἀποπεμπτοῦν τῷ Φαραώ (wayyaśæm) ‘And Joseph imposed it for them as an ordinance upon the land of Egypt, to this day, to give one-fifth to Pharaoh’ (15)  Exod 15:25 ἐκεῖ ἔθετο αὐτῷ δικαιώματα καὶ κρίσεις, καὶ ἐκεῖ ἐπείρασεν αὐτόν· (śām) ‘there He [God] made for him [Moses] statutes and judgements, and there he tested him’

In (14), Joseph is presented as an attentive ruler of his people, a benefactor whose impact on Egypt continues up to the author’s present. In (15), God makes laws for Moses, but His role as a beneficiary is not so obvious. However, His commitment to Moses and to the people, demonstrated by His giving of laws, is encoded with the middle voice. The translator, as a scribe, would have been familiar with the administrative language of the time and thus was naturally acquainted with the common expressions νόμον θέσθαι ‘to enact a law’ or τὸν γραπτὸν λόγον θέσθαι ‘present a written document (in court),’ as in (15a). (15a) P.Gur. 2 [καὶ Δωσιθέου] … οὔτε τὸ[ν] γραπτὸν λόγον θ[εμένου … Ἡράκλειας … καὶ [ἅμα τε] [γραπ]τὸν λόγον θ[ε]μένης καὶ τὰ δικ[α]ιώματα ‘[and Dositheos] … nor put in a written statement … Herakleia … and put in both a written statement and justificatory documents’30

Studied more closely, the instances taken from the contemporary documentary papyri bring us to yet another aspect of the middle voice. The middle aorist indicative of the verb τιθέναι appears frequently in various agreements, contracts, deeds, testaments and so on, signalling that someone has drawn up such and such a contract with someone else. Let us consider (15b) and (15c).

30  R.S. Kraemer, Women’s Religions in the Greco-Roman World (Oxford, Oxford University Press, 2004), Nr. 49.

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

93

(15b) P.Eleph. 2 (284 b.c.e.) τὴν δὲ συγγραφὴν ἑκόντες ἔθεντο παρὰ συγγραφύλα Ἡράκλειτον. ‘They have of their own free will placed the contract in the keeping of Herakleitos.’31 (15c)  UPZ 1.123 (157–156 b.c.e.) τῆς μητρός μου Ἀσκληπιάδος συνούσης (5) ­Ἰσιδώρωι τινὶ τῶν ἐκ Πίτου, καθʼ ἣν ἔθετο αὐ[τ]ῆι συγγραφὴν ὁμολογίας, διʼ ἧς διομολογεῖται ἄλλα τε καὶ ἔχειν πα[ρʼ] αὐτῆς ἣν προσενήνεκτο φερνὴν χαλκοῦ (ταλάντων) β ‘My mother Asklepias lived with a certain Isidoros, a resident of [the village of] Pitos pursuant to a memorandum of understanding which he drew up for her benefit. By its terms he acknowledged inter alia that he had received from her the dowry she brought, worth two copper talents.’32

In (15b), the subjects of the middle verb ἔθεντο have placed the document with their partner(s) for safekeeping for their mutual benefit. The memorandum in (15c) was made by the husband (ἔθετο) for his wife. This middle usage corresponds to what traditional grammars call the indirect reflexive middle use. In Bakker’s words, “the subject is not only the agent, but also the beneficiary of the action”, and “s/he is the one who gives orders for an event to happen in which s/he has an interest” (emphasis in original).33 Thus, the subject is in control of this placing, although rather often the subject is the “party” who pays the buyer or debtor; for example, in (15c), the husband should return the dowry in case of divorce. Thus, the middle of τιθέναι may denote the making of reciprocally beneficiary agreements, but it also allows the subject’s dominion interpretation, in that the subject is in control of the process (θέσθαι) of the document. In the next two instances, in (15d) and (16), we may infer a similar reading. (15d) P.Enteux. 22 (218 b.c.e.) [μὴ ἐχούσης] μου κύριον μεθʼ οὗ τὰς περὶ τούτων οἰκονομίας θήσομαι ‘[since I have] no guardian with whom I can make arrangements about these things (i.e., her legacy left by her late husband by a will)’34 (16) Gen 17:2 καὶ θήσομαι τὴν διαθήκην μου ἀνὰ μέσον ἐμοῦ καὶ ἀνὰ μέσον σοῦ καὶ πληθυνῶ σε σφόδρα. (ūnetattī-kā) ‘And I will make my covenant between me and you, and will multiply you exceedingly’—This future middle is what Bakker calls volitional.35 Thus, we

31 Bagnall/Derow, Greek Historical Documents, § 124: Will/testament. 32  N. Lewis, Greeks in Ptolemaic Egypt (Oakville, CT: American Society of Papyrologists, 1986), 28. 33 Bakker, “Voice, Aspect, and Aktionsart”, 36–7. 34 Bagnall/Derow, Greek Historical Documents, § 123. 35 Bakker, “Voice, Aspect, and Aktionsart”, 29.

94

Anssi Voitila

may translate: ‘I want to/will realize this covenant with you so as to multiply you’ (Cf. Exod. 34:12, 15 tikrot—act.subj.aor.)

The use of the middle verb in (15d) implies that the act belongs to the subject/ agent’s personal sphere, it is done in her interest and/or that she is committed to the contract. The middle in (16) denotes the reciprocity of the act: God is committed to the covenant with Abraham. This act is then followed by indicative futures, as in (5) (and also in (6)), that highlight the acts that God is willing to do as tokens of the covenant. As active verb forms (θήσω), the future indicatives signal mere action in which the agent will perform the act upon the patient, with agent and patient being two distinguishable participants; no confusion of agent and patient is to be detected, as is the case with middle forms. Furthermore, although (5) and (6) are part of a covenant text, the verbs constitute a mere promise to a patriarch; they are not statements of a covenant made between two parties. These promises are made solely by God and only to the patriarchs. In (7), though the putting of the bow in the clouds functions as a sign for the first covenant between God and the people, this idea is not being expressed in this sentence. Finally, the middle form often appears to express the making of a testament (διαθήκη). The middle form of τιθέναι is occasionally encountered in these contexts, but the most frequent verb is διατίθεναι. The middle form may be conceptualized as signalling the personal connection between the agent of the will and the beneficiary, at least during the lifetime of the agent. This is substantiated by the frequently expressed wish that the agent not die for a long time, as in (17). (17) P.Dryton 4 εἴ̣ηι μέμ μοι ὑγιαίνοντι τῶν ἐμαυτοῦ κύριον εἶναι· ἐὰν δέ τι ­ἀνθρώπινον πάθω ‘I wish it were possible for me to stay healthy and to remain the master of my own goods. But if I suffer the human destiny …’

This middle use is common in the Septuagint Pentateuch, as it is employed in reference to the making of the covenant between God and His people/patriarch(s) (19), beginning with Noah in Gen 9:17 (18). (18) Gen 9:17 καὶ εἶπεν ὁ θεὸς τῷ Νῶε Τοῦτο τὸ σημεῖον τῆς διαθήκης, ἧς ­διεθέμην ἀνὰ μέσον ἐμοῦ καὶ ἀνὰ μέσον πάσης σαρκός, ἥ ἐστιν ἐπὶ τῆς γῆς. (haqimōtī) ‘And God said unto Noah: “This is the token of the covenant, which I have established between me and all flesh that is upon the Earth.’” (Other covenants with the same “formulation”: Gen 21:27 (Abraham + Abimelech); 21:32; 31:44.)

Middle Voice as Depiction of Subject’s Dominion in the Greek Pentateuch

95

(19) Deut 5:2 κύριος ὁ θεὸς ὑμῶν διέθετο πρὸς ὑμᾶς διαθήκην ἐν Χωρήβ· (kārat) ‘The Lord our God made a covenant with us in Horeb.’

At this point, it is somewhat self-evident that this covenant-making happens between the agent and the addressees. The middle form enacts God’s particular way of existing with His people, His personal commitment to it. This language chosen by the author reminds the readers of the particularity of testament making, the power of the agent to do what He wishes with His possessions, to give them to whomever He wants. The making of covenants remains in His domain and belongs to His dominion.

6. Preliminary Conclusions The use of the middle aorist of τιθέναι in the Septuagint Pentateuch is not directly motivated by any grammatical element in the verb form it translates in the Hebrew Vorlage. Nonetheless, the above analysis of the middle instances demonstrates that the subject matter in this Vorlage inspired the translator to render the verb form as he did.36 There was some slight variation between translators: the translator of Genesis, in particular, used middle forms. However, this phenomenon embodies, at least in part, differences in the text type (no putting in prison or making covenants), as well as individuals’ predilections for different turns of phrase. Furthermore, the translators were familiar with certain turns of phrase that influenced their use of the middle aorist, such as (ἀπο)θέσθαι τινα εἰς φυλακήν οr διαθέσθαι τὴν διαθήκην (the last one occurred seven times in Deuteronomy). In this article, I have studied how the conception of the meaning of the middle voice presented by Maldonado and others aligns with various middle aorist uses of τιθέναι in the Greek Pentateuch. This short study has shown that the subject’s dominion is a more suitable explanation than the subject’s affectedness for the aorist middle voice usage of τιθέναι in the Septuagint and the other literary evidence treated here. The subject’s affectedness applies in many instances, but not everywhere. Likewise, this interpretation prevents us from realizing the role of the agent as the controller of the entire process that falls under the scope of the middle verb and its arguments. On the contrary, it makes the interpreter regard— too easily—the agent as a mere patient, and not as the initiator of the a­ ction. The 36  My article, A. Voitila, “Middle Voice in the Speech of and about God’s Power and Dominion over the Creation in the Greek Pentateuch”, in Evangelia G. Dafni (ed.), Divine Kingdom and Kingdoms of Men / Gottesreich und Reiche der Menschen (WUNT 432; Tübingen: Mohr Siebeck, 2019) 35–47, further supports the conclusions concerning the middle meaning presented in the present paper.

96

Anssi Voitila

agent in these instances appeared often, if not always, as superior to the patient or the beneficiary, as in control and as highly involved in his own actions. It has been demonstrated that when God puts/places something/someone or makes something in creation, the translator often interpreted these acts through the use of the middle aorist of τίθεναι. God is not primarily presented as affected by His own actions. Instead, the middle form is used to depict God’s dominion, to show His close relation with creation and human beings, and perhaps even His power. I am convinced that further study of middle usage, including that of other middle verbs, in the Septuagint will confirm this interpretation of the middle meaning. In addition, it will show that there were subtler but still relevant factors that influenced the use of the middle voice. A deeper understanding of these factors will lead to less confusion between the active and the middle meanings in instances in which a subject affectedness interpretation is not readily available.

Patrick Pouchelle Did the Translators of the Pentateuch Differentiate the Piel from the Qal? 1. Introduction The way Greek translators rendered tenses and moods was the subject of several monographs or articles in recent years.1 The analysis of their appreciation of the Hebrew stems or binyanim, on the contrary, has been overlooked. The main reason for this avoidance is the fact that the stems are mainly differentiated by vocalization only.2 Evans stated that “a technical sensitivity to the Hebrew verbal stems among the translators of the LXX […] is at best doubtful […] for which we have no positive evidence.”3 He recalls the fact that the “traditional” distinction into 7 stems is attributed to David Qimḥ i (1160–1235), whereas older grammarians ignore many of them.4 Yet, he made a plea for the study of how the stems were rendered in Greek, assuming that these renderings will only be semantic.5 This contribution to the Soisalon-Soininen Symposium on the Septuagint will aim at assessing this view, particularly for the rendering of the piel. The first part will be 1  E.g., I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965); J. Barr, “Translators’ handling of verb tense in semantically ambiguous contexts”, in C.E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (SBLSCS 23; Atlanta, GA: Scholars Press, 1987) 381–403; T.P. Schehr, “Syntax of the Moods and Tenses of the Greek Verb in Septuagint Genesis 1–15” (PhD diss., Hebrew Union College, OH, 1990); J.H. Sailhamer, The Translational Technique of the Greek Septuagint for the Hebrew Verbs and Participles in Psalms 3–41 (Studies in Biblical Greek 2; New York: Peter Lang, 1991); P.J. Gentry, The Asterisked Materials in the Greek Job (Atlanta, GA: Scholar’s Press, 1995), esp. 176–293; A. Voitila, “What the Translation of Tenses Tells about the Septuagint Translators”, SJOT 10 (1996) 183–96; J.A. Beck, Translators as Storytellers: A Study in Septuagint Translation Technique (StBibLit 25; New York: Peter Lang, 2000); T.V. Evans, Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference (Oxford: University Press, 2001); 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; Göttingen: Vandenhoeck & Ruprecht, 2001). S.-H. Woo, Etudes sur le système verbal dans la Septante de Job (Strasbourg: Éditions européennes universitaires, 2011). 2  See, e.g., Sailhamer, The Translational Technique, 18–19; Evans, Verbal Syntax, 63. 3 Evans, Verbal Syntax, 157. 4  Ibid., 158, after L. McFall, The Enigma of the Hebrew Verbal System: Solutions from Ewald to the Present Day (Historic Texts and Interpreters in Biblical Scholarship 2; Sheffield: Almond Press, 1982), 2–9 and B.K. Waltke/M.P. O’Connor, An Introduction to Biblical Hebrew Syntax (Winona Lake, IN: Eisenbrauns, 1990), 32–36. 5 Evans, Verbal Syntax, 83, 156.

98

Patrick Pouchelle

dedicated to the status quaestionis and how it could help to define a method to assess the sensitiveness of the Greek translator(s) to the stems. The second part will present some case studies, classified according to different characteristics of the qal stem.

2. Status Quaestionis Some scholars attempted to study the rendering of Hebrew binyanim into Greek, mainly the rendering of the causative effect of the hiphil.6 In his seminal work on the infinitive in the LXX, Soisalon-Soininen devoted one paragraph to studying the expression ποιέω+infinitive as a rendering of the piel or hiphil.7 In a more recent paper, Tov has studied the way Greek translators render the causative nuance of the hiphil.8 He excluded Greek verbs genuinely expressing causality, focusing notably on verbs specifically suffixed with endings like -όω, -ίζω or -άζω, and the use of auxiliary verbs like ποιέω.9 Exploring a new field, Thompson, in a monograph dedicated to the language of the Apocalypse, has tried to show that derived stems are indicated by the translator by stretching the classical Greek tense and mood. Specifically, he thought that the presence of a perfect indicative when its meaning does not fit the context according to the classical usage was a marker of a “non-qal” stem.10 Two studies on Aquila and Symmachus deal with the rendering of Hebrew stems. Among other observations, Reider noted that Aquila has a preference for rendering the piel and the hiphil with verbs ending with -όω, -ίζω or -άζω.11 He also asserted that when the qal is intransitive, it is rendered by the passive voice, whereas the piel or hiphil is rendered by the active. Additionally, the hiphil is sometimes rendered by the auxiliary ποιέω or a compound based on ποιέω.12

6  It seems that Katz was the one of the first to deal with this subject in an article where he refuted Smend, who saw in the Greek Sirach some conscious elimination of the causative effect of the hiphil; P. Katz, “Zur Übersetzungstechnik der Septuaginta”, Die Welt des Orients 2 (1956) 267–73, on pp. 269–72. He did not suggest any theory of the rendering of the causative effect. 7 Soisalon-Soininen, Die Infinitive, 134–5. 8  E. Tov, “The Representation of the Causative Aspects of the Hiph‘il in the LXX: A Study in Translation Technique”, in The Greek and Hebrew Bible: Collected Essays on the Septuagint (SVTSupp 72; Leiden: Brill, 1999) 195–202. 9 Tov, “The Representation of the Causative Aspects”, 197–202. 10  S. Thompson, The Apocalypse and Semitic Syntax (SNTSMS 52; Cambridge: University Press, 1985), 42–43. Later, he analyses the correspondence of the qal, niphal, pual and hophal participle in Greek (ibid. 71–73). 11  Along with others, such as -εύω, -έω, -ύω, -άνω, -αίνω, -ύνω, -άω. 12  J. Reider, Prolegomena to a Greek-Hebrew & Hebrew-Greek Index to Aquila (Oxford: University Press, 1916), 40–42.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

99

In his monograph devoted to Symmachus, Busto-Saiz, dealt with the translation technique of Symmachus regarding the Hebrew stems. He concluded that the construction ποιέω+adjective or participle or infinitive occurs with greater frequency in correspondence to a piel rather than to a hiphil. The qal may be rendered with an adjective or a participle introduced by the use of εἰμί or γίγνομαι. He also noticed a greater use of the suffixes -ίζω or -άζω for a concrete specialization of specific forms, as well as -ύνω for the intensive forms. As for Hebrew roots that are used either in qal and piel or hiphil, the middle voice rather corresponds to the qal, whereas the active corresponds to the piel and hiphil. He concluded with an observation of the great plurality of techniques to render the different Hebrew stems.13 After this survey of the status quaestionis, one could note that scholarship faced different issues when dealing with the rendering of the binyanim. The question of the translators should be dealt with first. Indeed, the Septuagint could not be studied as a whole as if it was translated by one translator.14 Reider and BustoSaiz assume that their translators were unique. This might be debated at least as regards Aquila who may have based his work on the so-called kaige revision.15 Moreover, both authors acknowledge that it is frequently difficult to attribute a fragment to Aquila or Symmachus owing to the issue of the transmission of the revisors.16 Thompson does not deal with this issue at all. Similarly, Tov did not raise this issue, as he only presents some translation techniques for the hiphil, regardless of the book in which they occur. This paper suggests limiting the enquiry to the Pentateuch only. However, such a limitation may raise some questions. First, this will limit the number of studied verbs. Indeed, a correlation between Hebrew stems and Greek syntagms would be detected for a given Hebrew verb if a given Hebrew verb occurs sufficiently often in the Pentateuch to be able to apply a statistical tool.17 Second, it could not be taken for granted that each book of the Pentateuch was translated by the same team of translators. However, if some similar features could be detected 13  J.R. Busto-Saiz, La traducción de Símaco en el Libro de los Salmos (Textos y estudios “Cardenal Cisneros”; Madrid: Instituto Arias Montano, Consejo Superior de Investigaciones Científicas 1985), 108–15. 14  See, for instance, I. Soisalon-Soininen, “Methodologische Fragen der Erforschung der Septuaginta-Syntax”, in C.E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (SBLSCS 23; Atlanta, GA: Scholars Press, 1987) 425–44. 15  D. Barthélemy, Les devanciers d’Aquila: première publication intégrale du texte des fragments du Dodécaprophéton trouvés dans le désert de Juda, précédée d’une étude sur les traductions et recensions grecques de la Bible réalisées au premier siècle de notre ère sous l’influence du rabbinat palestinien (Leiden: Brill, 1963). 16  For an introduction to these issues, see, for instance, K.H. Jobes/M. Silva, Invitation to the Septuagint (Grand Rapids, MI: Baker Academic, 2015), 24–33. 17  Sometimes even as low as 10 for applying the Fisher’s Exact test; see note 83.

100

Patrick Pouchelle

in all of these translations, whoever their translators might be, then it could be a translation technique shared by every translator. This could apply to the whole Septuagint as well, but it is safer to begin the investigation with only one coherent corpus so as to be able to sketch some conclusions, even if limited. This is an obvious methodological choice for all LXX studies.18 More crucial is the question of the Vorlage and the way it was read. When a specific Greek word corresponds to a specific stem in the MT in one specific verse, how could we be certain that the Hebrew Vorlage of the Greek text equates to the MT? The issue is even more difficult for stems which differ by vocalization only (like qal and piel). Reider and Busto-Saiz rely on historical considerations only. As Aquila and Symmachus are reputedly late translators, their Vorlage should be close to the MT. This assumption could not be taken for granted regarding the Pentateuch as they both acknowledge that these Vorlagen were actually not identical to the MT.19 On a different level, Thompson seems to assume that if a perfect indicative is used outside the “natural” usage of the koinê and corresponds to a derived stem in the MT then that is sufficient to prove that the Vorlage of the Greek was identical to it for this specific verb. This is highly problematic. For instance, he assumes that the use of λελάληκα in Ezek 3:10 is due to ‫א ַד ֵבר‬, ֲ a piel form of ‫דבר‬. Why not consider the possibility that the translator read ‫ד ַב ְר ִתי‬,ִ still a piel form but qatal and corresponding to the same Greek form in Ezek 17:24, which better corresponds to the MT, here? The “improper usage” here would be owing to the Vorlage and not to the stem. Moreover, his notion of the “proper sense” of the perfect indicative is also highly debatable. His three examples say nothing about a real translation technique. Evans elaborated a quick protocol to test Thompson’s theory by analysing the use of the perfect indicative in Genesis. Seventy-five occurrences correspond to a qal, 22 to a non-qal stem with no noticeable improper usage.20 For his part, Tov does not really deal with this issue. He seems to consider that a Greek causative construction which corresponds to a hiphil stem is enough to demonstrate the presence of a hiphil in a Vorlage here identical to the MT. However, by starting with how the hiphil may have been translated, and noting that the corresponding Greek expression conveys a strong causative nuance, he failed to check whether this kind of expression is also used for other stems. For example, φωτίζω “to shine” is taken by Tov as corresponding to ‫ה ִאיר‬,ֵ hiphil of ‫אור‬. He gives no data.21 According to the concordance of Hatch and Redpath, in

18 Soisalon-Soininen, “Methodologische Fragen”, 426. He criticizes the statistics collected by J.M. Rife, “The Mechanics of Translation Greek”, JBL 52 (1933) 244–52. 19 Reider, Prolegomena, 83–95 and Busto-Saiz, La traducción de Símaco, 287–92. 20 Evans, Verbal Syntax, 156–8. 21  As he warned his readers in Tov, “The Representation of the Causative Aspects”, 198.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

101

addition to ‫אּורים‬ ִ (the famous Urim)22, ‫“( ָמֹאור‬luminary” in Num 4:9), and once to ‫“( ֹאור‬light” in Mic 7:8), φωτίζω corresponds 16 times to the root ‫אור‬, 11 times to the hiphil of ‫אור‬, but also 3 times to qal, and twice to niphal. It raises the following question: could the non hiphil correspondence to φωτίζω be neglected or could it be concluded that the translators read the hiphil, instead of the qal or the niphal? The verb ‫ אור‬occurs 50 times in the Bible, 9 times in qal,23 twice in niphal, and 34 times in hiphil. The ratio between the hiphil and the qal is 3.77, and the ratio between φωτίζω corresponding to hiphil and φωτίζω corresponding to qal is 3.66, which is almost the same. It seems that there is not a specific correlation between the use of φωτίζω and the stem of ‫אור‬. Otherwise said, assuming that the Vorlage was identical to the MT, and that the translators had knowledge of the stem, it looks like they did not really take that into account. Hence, the proof of a correlation between the LXX and the MT could need some statistical tools.24 This field of mathematics aims at comparing two sets of data and checking their relationship. A “null hypothesis,” which is “there is no correlation between two sets of data,” is assessed according to the risk taken by rejecting this “null-hypothesis” when it is actually true. Here, the general null hypothesis is: For one Hebrew root, when, in the LXX Pentateuch, two Greek verbs or expressions correspond to the qal and piel of the Hebrew root in the MT, there is no correlation between them.

Some formula could be used to compute the probability of rejecting the “null hypothesis” when it should not be.25 When this probability is less than a specific threshold, typically 5%, the “null hypothesis” could safely be rejected. The main thing to figure out is, first, that the null hypothesis (i.e. the absence of correlation) could never be proven by this method, and, two, that this says nothing about the reason why the alternative hypothesis (“there is a correlation between the two sets of data”) may be true. Indeed, the translators might have chosen a given Greek word or expression according to the general context which implies the 22  In 2 Esd 2:63 and in 2 Esd 17:65. 23  Leaving apart the possible confusion between the noun and the infinitive construct qal: both ‫אֹור‬. 24  It should be acknowledged how dangerous statistics could be in biblical studies; see for instance A. Aejmelaeus, “What We Talk About When We talk About Translation Technique”, in B.A. Taylor (ed.), X congress of the International Organization for Septuagint and Cognate Studies Oslo, 1998 (SBLSCS 51; Atlanta: SBL, 2001) 531–52 (esp. 546–7 and 552), repr. in A. Aejmelaeus, On the Trail of the Septuagint Translators: Collected Essays (CBET 50; Leuven: Peeters, 2007), 205–22; see also T. McLay, The Use of the Septuagint in New Testament Research (Grand Rapids, MI: Eerdmans, 2003), 14. 25  E.g., Pearson’s Chi square test or Fisher’s Exact test, among others.

102

Patrick Pouchelle

use of a given stem in the MT. Therefore, once a correlation is proved, a rigorous analysis should be conducted to determine what the reason for a given choice of the translator could be. Such an analysis should be based on a good knowledge of the stems and their use, and especially of the relationship between the stems. This paper suggests dealing with the qal and piel stems only. As these stems are differentiated by the vocalization only, except for the participle, it is not necessary to discuss whether or not the consonants in the Vorlage and in the MT were the same: in most case they were, but nothing could be said about the vocalization. The set of studied Hebrew verbs will obviously decrease, as a correlation between Hebrew qal and piel and Greek syntagms could be detected if and only if the verb occurs both in qal and piel in the given corpus.26 Such a correlation will ultimately be explained by the relationship between qal and piel. This is still an open question as recent syntheses produced in the field of the Hebrew verbal system have been more dedicated to tenses and moods.27 Although the qal stem is qualified as “simple,” mainly because it does not have any consonantal affixes,28 its morphological, syntactic and semantic uses are not so simple.29 It is surprising that no recent monographs have been devoted to that.30 Joüon divides the qal into two categories: the action verbs and the stative verbs and presents the adjectival nature of the latter category.31 Waltke and O’Connor 26  For example, ‫ ׁשמע‬is used widely in the Pentateuch, but never in the piel (1 Sam 15:4, 23:8). 27  E.g., L. McFall, The Enigma of the Hebrew Verbal System: Solutions from Ewald to Present Day (Sheffield: Almond Press, 1982); Y. Endo, The Verbal System of Classical Hebrew in the Joseph Story: An Approach from Discourse Analysis (SSN 32; Assen: Van Gorcum, 1996); W.Th. van Peursen, The Verbal System in the Hebrew Text of Ben Sira (Studies in Semitic Languages and Linguistics 41; Leiden: Brill, 2004); J. Joosten, The Verbal System of Biblical Hebrew: A New Synthesis Elaborated on the Basis of Classical Prose (Jerusalem: Simor, 2012). K.M. Penner, The Verbal System of the Dead Sea Scrolls: Tense, Aspect, and Modality in Qumran Hebrew Texts (SSN 64; Leiden: Brill, 2015). 28 Waltke/O’Connor, §22.1, 363. 29  Waltke and O’Connor suggest up to 6 patterns for the qal morphological formation (§22.3, 367–71). 30  In contrast to the piel, obviously; see below. But also to the niphal: see P.A. Siebesma, The Function of the Niph’al in Biblical Hebrew: In Relationship to Other Passive-Reflexive Verbal Stems and to the Pu’al and Hoph’al in Particular (Assen: Van Gorcum, 1991); G.L. Klein, “The Meaning of the Niphal in Biblical Hebrew” (PhD diss., Annenberg Research Institute; 1992); R. Benton, “Aspect and the Biblical Hebrew Niphal and Hitpael” (PhD diss., University of Wis­ consin—Madison, 2009); J.T. Baden, “Hithpael and Niphal in Biblical Hebrew: Semantic and Morphological Overlap”, VT 60 (2010) 33–44; Y. Takahashi, “The Historical Development of Niphal Verbal Forms in Biblical Hebrew”, BMECCJ 56 (2013) 83–95; and hiphil: W.Th. Claassen, “The Hiph’il verbal theme in biblical Hebrew” (PhD diss., University of Stellenbosch, 1971); J.M. Léonard, “Enquête sur l’émergence du Hifil en hébreu biblique” (PhD diss., Montpellier, 1990); or a recent thesis about hithpael: M. Arnold, “Categorization of the hitpa’ēl of classical Hebrew” (PhD diss., Harvard University, 2005). 31  Joüon, §41a–b.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

103

also offer to group the qal stem into two main categories: the stative (“one that describes a circumstance or state”) and the fientive (“one that designates a dynamic situation”). Some of the roots belong to both groups; see for instance ‫מלא‬ or ‫אהב‬.32 Joosten acknowledges this diathesis, and speaks of active and middle voices.33 Apparently, the plurality of meaning of the qal stem is a clue to understanding the plurality of the piel. Traditionally, the piel has been viewed as intensifying the qal.34 This view was challenged by Goetze in 1942.35 In his analysis of the ­Akkadian language he reached the conclusion that the D-Stem (equivalent to the piel in Semitic studies) conveys the denominative or factitive nuance. This view has been followed with some nuances by Jenni,36 Ryder,37 Waltke and O’Connor,38 and Creason.39 All of them categorize the function of the piel according to that of the qal. Goetze analysed the D-Stem in Arabic, regarding the transitivity of the GStem (or B-Stem, equivalent to qal in Semitic studies) and he observed that in Assyrian, the D-Stem always corresponds to stative verbs: durative,40 perfect41 and passive42 to which correspond different meanings in the D-Stem.43 Jenni categorizes the piel into two main groups:44 (1) the factitive where the qal denotes the state of a subject, whereas the piel describes the action that makes an object be in a specific state (I am clean vs I clean him = I make him clean) and (2) the resultative where the qal denotes an action performed by a subject over an object and the piel expresses the process coming to an end (I break the jar vs I break up the jar = the jar is actually broken). Ryder45 also suggested two groups (1) the transformative (that transforms the subject of qal into something else) and 32 Waltke/O’Connor, §22.2.3, 365–7. 33  J. Joosten, “The Functions of the Semitic D Stem: Biblical Hebrew Materials for a Comparative-Historical Approach”, Orientalia 67 (1998) 202–30, esp. 206–16. 34  This view relies mainly on ancient Arab grammarians, see S.A. Ryder, The D-Stem in Western Semitic (Janua Linguarum 131; The Hague: Mouton, 1974), 12. 35  A. Goetze, “The So-Called Intensive of the Semitic Languages”, JOAS 62 (1942) 1–8. 36  E. Jenni, Das hebräische Piʿel. Syntaktische Untersuchung einer Verbalform im Alten Testament (Zürich: EVZ, 1968). 37 Ryder, The D-Stem. 38 Waltke/O’Connor, §24.1.b. 39  S.A. Creason, “Semantic Classes of Hebrew Verbs: A Study of Aktionsart in the Hebrew Verbal System” (PhD diss., University of Chicago, 1995). 40  A state inherent to the subject “to be good.” 41  A state resulting from an action of the subject “to learn” = “to be accustomed to.” 42  A state resulting from an action of a third party “to be born.” 43 Goetze, “The So-Called Intensive”, 5. 44  For verbs having no qal; for those whose qal probably never existed, he suggests a denominative function. 45 Ryder, The D-Stem, 94.

104

Patrick Pouchelle

(2) the non-transformative (where he groups the roots whose piel is denominative or is more or less synonymous with qal). Ryder remains aware that for some roots, the relationship between qal and piel is difficult to categorize.46 Waltke and O’Connor side with Jenni while being cautious about some of his conclusions.47 Creason also groups the piel function according to the function of the qal: intransitive stative, transitive stative and fientive.48 Kouwenberg tried to revitalize the ancient notion of intensive piel. This is the degree of transitiveness of the G-stem which leads to the plurality of nuances conveyed by the D-stem. For transitive verbs in qal, piel conveys the nuance of intensiveness or of all kinds of plurality. For intransitive or lower transitive verbs, Kouwenberg suggests either that the intensiveness implies a higher level of transitiveness and that the G-Stem of these roots may have had two nuances, one intransitive and one transitive. The latter stopped being used under the competition with the piel.49 Joosten develops a similar line of argumentation by noting that originally the qal stem developed a middle and an active voice for which the piel stem conveyed the intensive nuance. For many verbs, the fientive qal stopped being used while for others it was the middle voice which was infrequently if not ever used. Finally, for some verbs, the fientive qal never existed and the piel became the intensifier for the middle voice of qal.50 Recently, Beckman51 has refuted the thesis of Waltke and O’Connor, mainly owing to the presence of intransitive verbs, like ‫הלך‬, in piel and supporting the Kouwenberg view, while acknowledging it as difficult to falsify. It is beyond the scope of this article to arbitrate between these irreconcilable views. Nevertheless, it could be interesting to note that all these authors start from the use of qal, and especially whether it is stative or fientive. Unfortunately, the number of Hebrew verbs occurring in qal and piel in the Pentateuch is still important. The examples given in the following sections have been chosen because they offer clear case studies: They occur in the Pentateuch both in qal and piel and they have few correspondent Greek syntagms.52 They were collected in a research survey, as this study does not aim to be exhaustive. 46 Ryder, The D-Stem, 97. For him, the piel and qal emerge independently and were associated at a later stage. A view not accepted by Waltke/O’Connor, 24.1g but not so far away from that of Joosten; see below. 47 Waltke/O’Connor, §24.1h, 399–400. 48 Creason, “Semantic Classes of Hebrew Verbs”, 170–88. 49  J.C. Kouwenberg, Gemination in the Akkadian Verb (SSN 32; Assen: Van Gorcum, 1997), 92–99, 144–236, 435–43. 50 Joosten, “The Functions of the Semitic D Stem”, 225–7. 51  J.C. Beckman, “Toward the Meaning of the Biblical Hebrew Piel Stem” (PhD diss., Harvard University, 2015), 244–54. 52  For example, ‫ אחר‬occurs in the Pentateuch both in qal (once in Gen 32:5 correspondent: χρονίζω) and in piel (Gen 24:56 corr. κατέχω; Gen 34:19, corr. χρονίζω; Exod 22:28, corr. καθυ-

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

105

3. Qal stative When the qal has a stative meaning, whether intransitive, like ‫טהר‬, or transitive, like ‫למד‬, the piel conveys a fientive meaning which could be easily discerned by the translators, even if they did not know anything about the stems. It is therefore not surprising that a clear correlation exists between the stem in the LXX and in the MT. Two categories of correspondence have been found: the use of the passive voice corresponding to qal and the active voice to piel (typical example: ‫)טהר‬, and two different Greek verbs for both stems (typical example: ‫)למד‬. The first example is ‫טהר‬. The following table offers the occurrences of the qal “to be pure” and the piel “to purify” or “to declare pure” of the root ‫טהר‬. Its qal is stative, and its piel is declarative according to Jenni.53 Table 1: Occurrences of the corresponding expression to ‫טהר‬

καθαρίζω active voice καθαρίζω passive voice καθαρός εἶναι Sum

Qal

Piel

sum

154 9 10 20

11 0 155 12

12 9 11 32

The qal is distributed evenly between the verb καθαρίζω and the syntagm καθαρὸς εἶναι. A closer look shows that the passive voice of καθαρίζω corresponds to the qal only, whereas the active voice corresponds to the piel, with one exception. Similarly, καθαρός εἶναι does not correspond to a piel except for one occurrence. The translators seem to have chosen either καθαρὸς εἶναι or the passive voice of καθαρίζω for the qal. Even in the same verse, the two expressions occur, as in Lev 15:13: ἐὰν δὲ καθαρισθῇ (‫ )יִ ְט ַהר‬ὁ γονορρυὴς ἐκ τῆς ῥύσεως αὐτοῦ, καὶ ἐξαριθμήσεται αὐτῷ ἑπτὰ ἡμέρας εἰς τὸν καθαρισμὸν καὶ πλυνεῖ τὰ ἱμάτια αὐτοῦ καὶ λούσεται τὸ σῶμα ὕδατι καὶ καθαρὸς ἔσται (‫)וְ ָט ֵהר‬.

στερέω; Deut 7:10, corr. βραδύνω; Deut 23:22, corr. χρονίζω). However, the number of Greek correspondents precludes any possibility of studying the correlation. 53  Declarative, according to Jenni “für rein erklären” (Jenni, Das hebräische Piʿel, 21, 41, 83). 54  Lev 12:7. 55  Lev 14:7.

106

Patrick Pouchelle

The two occurrences of ‫ טהר‬are in the qal stem in the MT. In a similar context, Lev 15:28, offers two occurrences in the passive voice: ἐὰν δὲ καθαρισθῇ (‫ )יִ ְט ַהר‬ἀπὸ τῆς ῥύσεως, καὶ ἐξαριθμήσεται αὐτῇ ἑπτὰ ἡμέρας καὶ μετὰ ταῦτα καθαρισθήσεται (‫) ִת ְט ָהר‬.

The main difference is that the syntagm καθαρός εἶναι corresponds to the weqatal of qal ‫ טהר‬in the third pers. masc. sing. Indeed, this is true for almost all the occurrences of this syntagm when corresponding to ‫טהר‬.56 This could explain the correspondence of this syntagm with a piel in the MT in Lev 14:7, where the final waw of the piel form (‫ )וְ ִט ֲהרֹו‬may have dropped owing to the fact that the following form begins with a waw. Hence, it seems that the translators interpreted the form ‫ וטהר‬not as a verb but as an adjective.57 Indeed in the MT, the vocalization ‫ וְ ָט ֵהר‬of the qal is the same as that of the presumed adjective, not attested in our modern dictionaries. However, the fact that there is no clear attestation of the adjective ‫ ָט ֵהר‬in the current state of the MT does not preclude that the translator recognized the form ‫ וטהר‬as mainly adjectival. Two exceptions occur in Lev 14:20 and Num 31:23 where the passive voice corresponds to ‫וטהר‬. Here, the translators either had a different Vorlage, or they stylistically changed their translation, but the elucidation of this would lead this essay too far from its original aim. When the translators interpreted the form as a conjugated verb, they chose the passive voice58 for the intransitive and stative qal and the active voice for the transitive and factitive piel. There is only one exception, in Lev 12:7. ‫יה‬ ָ ‫יה וְ ָט ֲה ָרה ִמ ְמקֹר ָד ֶמ‬ ָ ‫וְ ִכ ֶפר ָע ֶל‬

καὶ ἐξιλάσεται περὶ αὐτῆς ὁ ἱερεὺς καὶ καθαριεῖ αὐτὴν ἀπὸ τῆς πηγῆς τοῦ αἵματος αὐτῆς

The LXX here has ὁ ἱερεὺς as a subject, in addition to the MT. This could date from the Vorlage or may, less likely, have been added by the translators.59 The form ‫ וטהרה‬is also present in Lev 12:8 in a gloss added to verse 7, with the mention of the priest: 56  Lev 11:32; 13:6, 34, 58; 14:8, 9; 15:13; 17:15; 22:7; exception in Lev 14:7 (where the syntagm corresponds to the piel ‫וְ ִט ֲהֹרו‬, see below) and in Num 19:12 twice corresponding to a yiqtol (see BHS app. cr.). In Lev 14:20 and Num 31:23, the form ‫ וְ ָט ֵהר‬corresponds to καθαρισθήσεται; see also Num 19:19 where a weqatal of ‫ טהר‬has no Greek correspondence. 57  This is true for the Pentateuch, not for other books a priori. Hence, in Prov 20:9, the qal ‫ ָט ַה ְר ִתי‬corresponds to the syntagm καθαρὸς εἶναι. 58  The only unambiguous middle voice is in Gen 35:2, corresponding to the hithpael of ‫טהר‬. The passive voice corresponding to the qal is always unambiguous. 59  If the translators dared add ‫הכהן‬, for harmonization with v. 8, it could be expected that they also harmonized the form ‫וטהרה‬.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

107

‫יה ַהכ ֵֹהן וְ ָט ֵה ָרה‬ ָ ‫וְ ִכ ֶפר ָע ֶל‬

καὶ ἐξιλάσεται περὶ αὐτῆς ὁ ἱερεύς, καὶ καθαρισθήσεται

Of course, the form ‫ וטהרה‬could be interpreted either as a qal 3rd pers. fem. sing., or a piel 3rd pers. masc. sing. followed by the suffix for the 3rd pers. fem. sing. (with a mappiq). The MT interprets it as a qal in both verses, whereas the LXX in the state extant to us has an active voice in verse 7 and a passive voice in verse 8, as if the translators had read what we could call now a piel in verse 7 (Cf. BHS app. cr.) and a qal in verse 8.60 Another possibility is that the translators recognized an actual consonant in the final ‫ה‬. Therefore, as the difference between the two verses is somewhat unexpected, the translators either modified their translation for a presumed stylistic interest, or they read both verses with a specific tradition of vocalization. I believe the second possibility to be simpler and more reasonable, although this tradition could be based on the stem or on the consonantal interpretation of the letter ‫ה‬. Therefore, the translations more likely have to do with transitiveness rather than with the stem.61 This is also true with ‫ טמא‬where we observe the same phenomenon with the verb (ἐκ)μιαίνω. The active voice corresponds to the piel of ‫( טמא‬18 times) and the passive to the qal (7) with only three exceptions.62 The suffix ἐκ is used only four times, always corresponding to the qal (and so in the passive voice). Therefore, the following hypothesis on the translation technique could be put forward. For a given stem where the qal is stative and intransitive, the translators used the passive voice of a Greek verb when the Hebrew verb is used intransitively (qal) and the active voice when used transitively (piel).

60  As a matter of fact, few witnesses of the LXX offers καθαρισθήσεται, which could be either a revision toward a pre-Masoretic text or more probably a harmonization with verse 8. It should also be noted that the translators did not interpret ‫ וטהרה‬as an adjective. In fact, the syntagm καθαρὸς εἶναι is mainly attested in the masc. sing. Exceptions: in Lev 14:53, probably because the house is feminine in Greek, contrary to Hebrew; in Lev 15:12, in the neuter, with no correspondence in Hebrew; and in Num 8:7 in the masculine plural, corresponding to a hiphil weqatal of ‫טהר‬. This preference could be owing to the fact that, for the translators, as in the MT, only the vocalization ‫ וְ ָט ֵהר‬could be ambiguous whereas there is a different vocalization between the qal conjugated form (‫ )וְ ָט ֵה ָרה‬and the derived adjective (‫ּוט ֵה ָרה‬, ְ unattested). 61  As with English “to close”: “close the door” is transitive and fientive but “the shop closes at 5.30 pm” means that the shop is closed at this time. 62  Lev 22:5 (bis) where the active voice corresponds to the qal and Num 6:9 where the passive voice corresponds to the piel.

108

Patrick Pouchelle

This shift in voice is nevertheless not systematic, as proven by the verb ‫כבד‬. It occurs 5 times in piel63 in the Pentateuch, always corresponding to the active voice of τιμάω. It occurs 10 times in qal (including the adjectival form). It corresponds once to the adjective μέγας, 4 times to βαρύς, 3 times to the passive of βαρύνω, once to the passive of βαρέω and once to the active of the intransitive βαρυωπέω.64 It is rather clear that the piel of ‫ כבד‬was understood by the translators as lexicalized and meaning “to honour.” The grammatical difference between the qal and the piel is also obvious, the qal is intransitive whereas the piel is transitive, linking two animate persons. Another example of this lexicalized translation technique is ‫למד‬. In qal, a stative nuance means “to become accustomed to,” hence “to learn.”65 In piel, a possible factitive nuance could be detected “to make someone become accustomed to,” hence “to teach.” In the Pentateuch, the qal corresponds to μανθάνω 7 times and the piel corresponds to διδάσκω 9 times and once to δεικνύω/­δείκνυμι (Deut 4:5). This last correspondence may be due to the fact that the Vorlage had ‫ראה‬,66 the root to which δεικνύω/δείκνυμι corresponds the most, and accordingly will be disregarded in this study. Table 2: Occurrences of μανθάνω / διδάσκω corresponding to ‫למד‬

μανθάνω διδάσκω Sum

Qal

Piel

Sum

7 0 7

0 9 9

7 9 16

The correlation is obvious. Both qal and piel are transitive but although the qal links an animate subject to something learnt, the piel links an animate subject with an animate object.67 A clear case could be found in Deut 4:10: 63  Declarative according to Jenni, Das hebräische Piʿel, 41, 105. 64  See Hesychius, Lexicon, s.v. 65  Causative, see Jenni, Das hebräische Piʿel, 22. 66  See M. Karrer/W. Kraus (ed.), Septuaginta Deutsch. Erläuterungen und Kommentare; Band 1, Genesis bis Maccabäer (Stuttgart: Deutsche Bibelgesellschaft, 2011), 543. 67  Sometimes with the thing taught either as a second accusative or introduced by a preposition. There is a unique exception with a piel linking an animate subject and a matter taught in Job 21:22: ‫ד־ד ַעת‬ ָ ‫ה ְל ֵאל יְ ַל ֶמ‬.ַ Here the LXX alters the meaning of the sentence: πότερον οὐχὶ ὁ κύριός ἐστιν ὁ διδάσκων σύνεσιν καὶ ἐπιστήμην; This alteration is usually understood as denying the idea that God could be taught, but this alteration could come from the grammatical difficulty the translator faced. Accordingly, this is the only occurrence in the LXX of διδάσκω in the active voice with no animate object.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

109

καὶ ἀκουσάτωσαν τὰ ῥήματά μου, ὅπως μάθωσιν (‫ )יִ ְל ְמדּון‬φοβεῖσθαί με πάσας τὰς ἡμέρας, ἃς αὐτοὶ ζῶσιν ἐπὶ τῆς γῆς, καὶ τοὺς υἱοὺς αὐτῶν διδάξωσιν (‫)יְ ַל ֵמדּון‬.

If the forms ‫ יִ ְל ְמדּון‬and ‫ יְ ַל ֵמדּון‬are differentiated by vocalization only, they also convey a different transitivity. The first verbal form has the matter taught as object (an infinitive), the second has the people who have been taught. Greek translators may well have had the same issue with ‫ למד‬as English translators may have with the French verb “apprendre” which follows a line similar to ‫למד‬, meaning “to learn” when the object is the matter taught and “to teach” when an animate indirect object is introduced by the preposition “à.”68 Interestingly enough, the translators could have applied this same technique (a passive voice for the qal and an active one for the piel). Indeed, the passive of διδάσκω is attested in classical Greek as meaning “to learn,” even with the accusative of what is taught.69 It is used eight times in this voice within the LXX, and corresponds to ‫ למד‬in Jer 38[31]:18, Hos 10:11, Cant 3:8, and 1 Chr 25:7.70 Here, the MT always uses the pual stem.71 Apart from Jer 38[31]:18, the MT has pual participle with a prefixed ‫ מ‬which differs from the piel participle in the vocalization but also in transitiveness.72 The occurrence of ‫ למד‬in Jeremiah is intransitive and may explain by itself the use of the passive of διδάσκω according to the technique described above in this section. This could be compared to Deut 31:12, where ‫ למד‬is used intransitively in the MT: ‫“ יִ ְל ְמדּו וְ יָ ְראּו‬they learn and they fear.” In the Septuagint, the following expression is used: μάθωσιν φοβεῖσθαι “they learn to fear.”73 The Vorlage may have had ‫ ליראה‬instead of ‫( ויראו‬as in Deut 4:10), or the translators were sensitive to the stem, translated ‫ למד‬qal with μανθάνω which is transitive and were accordingly compelled to follow it by the infinitive of φοβέω. The first possibility appears simpler to me. At least its plausibility avoids seeing in this specific correspondence the proof of a sensitivity of the translators to the stems.

68  See the TOB translation of Deut 4:10: “Je leur ferai entendre mes paroles pour qu’ils apprennent à me craindre tous les jours qu’ils vivront sur la terre, et pour qu’ils l’apprennent à leurs fils” to be compared to “I will let them hear my words, so that they may learn to fear me as long as they live on the earth, and may teach their children so” (NRSV). 69  E.g., Herodotus, Hist. 3.81; Euripides, Andr. 739. 70  It corresponds once to ‫ ידע‬hiphil (1 Esd 9:55//Neh 8:12). It has no clear correspondence in Isa 55:12 (the factual correspondence is ‫)יבל‬. Wisdom uses it twice in Wis 6:10 and 9:18. 71  In Isa 29:13 the pual participle corresponds to the substantive διδασκαλία. 72 Cf. Deut 4:1; 2 Sam 22:35 (//Ps 17[18]:35); Isa 48:17; Ps 93[94]:10, 119[118]:99, 143[144]:1; Prov 5:13. 73  Cf. Jenni, Das hebräische Piʿel, 22, n. 29.

110

Patrick Pouchelle

Hence, the Greek translators may have used the passive of διδάσκω when ‫למד‬ is intransitive and the active voice when the object is the person taught, but they preferred μανθάνω when the object is the matter to be taught. The basic reason for this choice is currently not known. However, the mutual relationship between qal (stative and transitive), piel (fientive and transitive), and pual (passive and intransitive) would be worth studying further.74 In conclusion, the difference between the qal and the piel for ‫ למד‬could be identified by the translators by the transitiveness and the context. Their strategy for this root was to use two different verbs rather than trying to keep to the same family of the Greek verb at all costs.75

4. Qal fientive The presumed intensive or resultative meaning of a piel for the verb of these categories is sometimes subtle or artificial.76 If the Greek Pentateuch offers a correlation with the stems of a given root, that would make it a good candidate to demonstrate that the translators knew the stem. Nevertheless, many cases offer no proof of correlation.

74  For example, ‫ טהר‬is hardly attested in the pual (two conjectural readings in 1 Sam 20:26, where the LXX rather reads an adjective, and in Ezek 22:24, where the LXX seems to have read a form of ‫)מטר‬. ‫ טמא‬is attested once in the pual in Ezek 4:14 where the Septuagint offers a doublet οὐ μεμίανται ἐν ἀκαθαρσίᾳ corresponding to ‫לֹא ְמ ֻט ָמ ָאה‬. ‫ כבד‬is attested three times in the pual (Isa 58:13, without Greek correspondence, Prov 13:18 with the passive voice of δοξάζω, Prov 27:18 with the passive voice of τιμάω); these occurrences are intransitive as the qal stem, but the context is clear that honour is at stake rather than physical weight. 75  Another clear example is ‫ילד‬. In the qal, it means “to give birth, to beget.” In piel, it means “to help someone to give birth.” (Causative, cf., Jenni, Das hebräische Piʿel, 23, n. 32, 210–11) The substantive μαῖα corresponds to the piel participle 9 times (Gen 35:17; 38:28; Exod 1:15–21), whereas the verb μαιόομαι correspond to the only form of the verb in the piel which is not a participle (Exod 1:16). The qal corresponds to τίκτω 91 times. The correlation is obvious. Again, it is not difficult for the translator to detect that when the subject is a parent and the object a child τίκτω should be used, whereas when a woman is the subject and a pregnant woman the object, μαιόομαι should be used. 76 See Goetze, “The So-Called Intensive”, 6; Ryder, The D-Stem 97; Waltke/O’Connor, §23.3, p. 404–10; or Beckman, “Toward the Meaning”, 187–90.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

111

The key example is the verb ‫ׁשבר‬. It presumably means “to break” in qal and “to break completely,” or “to make broken” in piel.77 The Greek translators of the Pentateuch invariably make these forms correspond to the same verb συντρίβω, regardless of the stem. There are other cases of such obvious non-correlation between the choice of the translators and the stems, such as ‫גרׁש‬.78 More often, a Hebrew verb corresponds to several Greek verbs or expressions without a clear correlation with the stems, as witnessed in the MT. An example is the verb ‫ נׁשק‬denoting “to kiss.” It occurs 13 times in the Pentateuch, 9 times in qal, and 4 in piel. In the piel, it may denote the intensity of the kiss “to kiss long and intensely,”79 “to kiss in turn,”80 as three occurrences of the piel denote the kissing of several people. One of the qal occurrences corresponds to ὑπακούω.81 All the other occurrences correspond to φιλέω (7 times in qal, once in piel—Gen 29:13) or to καταφιλέω (once in qal—Exod 4:27, three time in piel).82 Table 3: Occurrences of φιλέω/καταφιλέω corresponding to ‫נׁשק‬

φιλέω καταφιλέω Sum

Qal

Piel

Sum

7 1 8

1 3 4

8 4 12

There could therefore be a correlation. However, when observing the whole MT, ‫ נׁשק‬is used in piel only once in the difficult Ps 2:12 (LXX: δράσσομαι). Otherwise, in qal, it corresponds 7 times to φιλέω and 10 times to καταφιλέω. Hence, considering the whole LXX, it looks like there is no correlation

77 Jenni, Das hebräische Piʿel, 141–42. 78 Indeed, ‫( גרׁש‬5 qal and 15 piel), “to expel,” always corresponds to ἐκβάλλω (and once to ἐκβολή for the infinitive absolute). 79  TDOT s.v. 80  HAL; see also Jenni, Das hebräische Piʿel, §148. 81  Gen 41:40 where the Egyptian people kiss Joseph, denoting their veneration of him. In the LXX, the people will obey him. The Septuagint Deutsch characterizes this choice as a free translation according to the context. See Karrer/ Kraus, Septuaginta Deutsch, 236. This occurrence will be disregarded in this study. 82  For the difference between φιλέω and καταφιλέω, see Philo, Her. 41.

112

Patrick Pouchelle

between the Greek text and the MT. A statistical tool would be of some help for assessing if there is a correlation specifically in the Pentateuch. Given the small size of the sample, Fisher’s exact test is particularly well suited.83 Unfortunately, it offers a probability of 6.4% for falsely rejecting the following null hypothesis: When φιλέω and καταφιλέω corresponds to ‫ נׁשק‬there is no correlation of their distribution with the stem of the root in the MT.

This is too high according to the usual criterion of 5%. Several other Hebrew verbs with two usual Greek correspondents follow this example.84 The verb ‫“ ָׁש ַלח‬to send” will be a good example to finish with. It occurs 874 times in the MT (564 times in qal, 1 in nifal, 267 in piel, 10 in pual, and 5 in hiphil).85 The relationship between ‫ ׁשלח‬and the verb ἀποστέλλω and its com-

83  This test was invented by Ronald Fisher (1890–1962) to test the claim of Muriel Bristol of being able to detect which of the tea or the milk was put first in a given cup of tea. Contrary to the chi-squared test, it offers an exact value of the deviation from the null-hypothesis and not an approximation which tends to be closer to the exact value according to the size of the set. Unfortunately, the computation of this test becomes very complicated as the set of data become bigger. Hence, Fisher’s Exact test is ideal for testing a correlation between a small set of data (typically from 10 upward to around 1000) and two categories, whereas the chi-squared test could be applied for larger sets. For a good reference to the usage of this test, please refer to P. Armitage/ P.J. Berry/J.N.S. Matthews, Statistical Methods in Medical Research (4th ed.; Oxford: Blackwell Publishing, 2002). I would like to thank Mikhail Seleznev who suggested this technique to me. 84  Similarly, two Greek verbs correspond to ‫קבר‬, “to bury,” θάπτω and κατορύσσω. In Hebrew the piel of ‫קבר‬, occurring once in the Pentateuch (Num 33:4), presumably means to bury several corpses (Jenni, Das hebräische Piʿel, 145–46). All these forms but one correspond to θάπτω (25 to the qal, once to the piel). The only exception is in Gen 48:7 where the qal corresponds to κατορύσσω. With a p close to 100%, it is obviously not safe to reject the following null-hypothesis: “the occurrence of θάπτω and κατορύσσω is not correlated to the stem of ‫קבר‬.” For ‫ׂשנא‬, “to hate:” The piel occurs three times in the Pentateuch (Jenni, Das hebräische Piʿel, 224), always in participle and always corresponds to μισέω (Num 10:35; Deut 32:41, 33:11). This Greek verb also correspond 26 times to ‫ ׂשנא‬in qal (attested 29 times in MT). The Greek passive voice of μισέω corresponds to the passive participle only (4 times in Gen 29:31, 33; Deut 21:15 [bis], 16, 17.). The active participle qal corresponds once to ἐχθρός (Exod 23:5) and twice to ὑπεναντίος (Gen 24:60; Exod 1:10). 85 Jenni, Das hebräische Piʿel, 193–199. Jenni noticed the difference between “schicken” and “fortschicken” and attributed it to a resultative nuance.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

113

pound, and notably ἐξαποστέλλω, is so strong that only 28 occurrences of ­ἀποστέλλω out of nearly 500 do not correspond to ‫ ׁשלח‬in the MT;86 conversely, about one eighth of the occurrences of ‫ ׁשלח‬correspond to a verb other than ἀποστέλλω and its compounds.87 In the Pentateuch, the distribution of ἀποστέλλω and ἐξαποστέλλω seems to depend on the stems.88 Table 4: Occurrences of ἀποστέλλω and ἐξαποστέλλω corresponding to ‫ׁשלח‬, according to its stem

ἀποστέλλω ἐξαποστέλλω sum

Qal

Piel

Sum

71 6 77

24 74 98

95 80 175

We could apply the statistical test to the following “null hypothesis”: When ἀποστέλλω and ἐξαποστέλλω in the LXX Pentateuch, corresponds to ‫ ׁשלח‬in the MT, there is no correlation between the stem of ‫ ׁשלח‬and the repartition of the Greek words.

The probability is here less than 0.1%.89 This leads to the conclusion that the “null hypothesis” could be rejected with a 99.9% reliability. The alternative hypothesis is therefore true, but are we allowed to build the following model? The translators have knowledge of the stems of ‫ׁשלח‬, they preferably translate it by ἀποστέλλω in a qal stem, by ἐξαποστέλλω in a piel.

86  In the Pentateuch: Gen 32:13, 45:1; Exod 3:12, 15:10, 21:2; Lev 25:21; Deut 28:8, 29:22. 87  Notably when ‫ ׁשלח‬conveys the meaning of “to extend.” It usually corresponds to ἐκτείνω, e.g. Gen 3:22. 88  Recently, J.A.L. Lee, “Ἐξαποστέλλω”, in J. Joosten/P.J. Thompson (ed.), Voces Biblicae: Septuagint Greek and its Significance for the New Testament (CBET 49; Leuven: Peeters, 2007) 99–113, has suggested that the translators used ἀποστέλλω or ἐξαποστέλλω according to the dating of a given book. He argued on the basis of Hellenistic writers, like Polybius or Diodorus of Sicily, as well as papyrological and epigraphical data, that ἐξαποστέλλω competed with ἀποστέλλω during the 3rd to the 2nd century b.c.e. but finally faded around the eve of the Roman era. Assuming that the Pentateuch as a whole was translated within less than a generation, this effect could be neglected in this study. 89  With a set of 175 occurrences, Fisher’s exact test could still be applied.

114

Patrick Pouchelle

That a prefixed verb with ἐκ may convey a specific nuance conveyed by a specific stem had been suggested as early as Thackeray.90 However, it seems that this is not a systematic translation technique.91 Some other explanation could also be suggested: ἀποστέλλω was used by the translator(s) to denote the action of sending with a mission or sending an object. By contrast, ἐξαποστέλλω rather expressed the dismissal of an animate object from somewhere. In this case, the use of ἐξαποστέλλω along with ἐκ also corresponds to a specific usage of the piel stem of ‫ׁשלח‬. The qal of ‫ ׁשלח‬is almost always constructed in a “trivalent” way:92 a subject sends an object for a specific purpose,93 even if the second or the third, and rarely both, are sometimes omitted.94 In the piel, a specific meaning exists alongside the trivalent use: a bivalent construction, sometimes collocating with the preposition ‫מין‬, ִ often translated by ἐκ, conveys the specific meaning of “to dismiss,”95 or “to release.”96 Some cases could be ambiguous. Indeed, in Gen 8:7, ‫ת־הע ֵֹרב‬ ָ ‫“וַ יְ ַש ַלח ֶא‬and he sent out the raven,” the piel may well have been used to convey this nuance. However, in Gen 8:8, ‫ת־הּיֹונָ ה ֵמ ִאּתֹו ִל ְראֹות‬ ַ ‫“ וַ יְ ַש ַלח ֶא‬and he sent out the dove from him, to see …,” the bivalent use of the piel with the preposition ‫ ִמן‬could or should be understood as “to send forth” rather than “to release,” owing to the fact that a mission is specified. In Greek, the verb ἀποστέλλω is used in both cases. Moreover, in v. 7, the syntagm τοῦ ἰδεῖν97 is a plus regarding the MT, as if the translators had ‫ לראות‬in their Vorlage as in verse 8. Moreover, the preposition ‫ ִמן‬has an unexpected Greek correspondent (ὀπίσω αὐτοῦ). In contrast, in Gen 8:10, ‫ן־ה ֵת ָבה‬ ַ ‫ת־הּיֹונָ ה ִמ‬ ַ ‫“ וַ י ֶֹסף ַש ַלח ֶא‬and again he sent out the dove from the ark” corresponds to ἐξαπέστειλεν (piel) τὴν περιστερὰν ἐκ τῆς κιβωτοῦ (Gen 8:10). Regarding this nuance of ‫ׁשלח‬, the distribution of the occurrences of ἀποστέλλω and ἐξαποστέλλω in the Pentateuch is as follows:

90  H.St.J. Thackeray, A Grammar of the Old Testament in Greek According to the Septuagint (Cambridge: University Press, 1909), 289. 91  For example, regarding ‫טמא‬, ἐκμιαίνω corresponds to qal only, μιαίνω to qal and piel (see also n. 56). 92  For example, see the specific bivalent use: to send a gift, also present in piel (cf. Gen 38:20 in qal and Gen 38:17 in piel concerning the same sending of a kid). 93  This purpose could be expressed by an infinitive (e.g. Gen 8:7), or a final clause (e.g. Gen 24:40). 94  TDOT 15:58–67. 95  TDOT 15:67–68. 96  See also TLOT, 1332. 97  Followed by εἰ κεκόπακεν τὸ ὕδωρ which is also a plus corresponding to verse 8.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

115

Table 5: Occurrences of ἀποστέλλω and ἐξαποστέλλω corresponding to ‫ ָׁש ַלח‬, according to its meaning

ἀποστέλλω ἐξαποστέλλω sum columns

General meaning of ‫ׁשלח‬ “to send”

Specific meaning of ‫ׁשלח‬ “to dismiss”

Sum lines

8698 7100 93

1099 72101 82

96 79 175

A translation technique without the knowledge of the stem could then be elaborated. When ‫ ׁשלח‬denotes a dismissal, especially with the mention of where the dismissal is done from (mainly piel in the MT with the preposition ‫)מין‬, ִ then 98  Seventy-one times with qal: Gen 8:7, 8; 19:13; 20:2; 24:7, 40 (here an error in H&R); 27:45; 28:5; 31:4; 32:4, 6, 19; 37:13, 14; 38:17, 20, 23, 25; 41:8, 14; 42:4, 16; 43:8; 45:5, 7, 8, 23, 27; 46:5 (here, the subject is Joseph in Greek, Pharaoh in MT), 28; Exod 2:5; 3:10, 13, 14, 15; 4:13, 28; 5:22; 7:16; 8:24 (Pharaoh as subject, but in Alexandrinus ἐξαποστέλλω); 9:15, 27; 23:20; 23:28 (sending of plagues by God); Num 13:2(bis), 16, 17, 27; 14:36; 16:12, 28, 29; 20:14, 16; 21:6 (sending of plagues), 20, 32; 22:5, 10, 15, 37, 40; 24:12; 31:4, 6; 32:8; Deut 1:22; 2:26; 19:12; 32:24 (but in Alexandrinus ἐξαποστέλλω, sending of plagues); 34:11 and 14 times with piel: Gen 19:13; 37:32; 38:17; 43:4–5, 14; Exod 15:7; 23:27 (sending of plagues by God); Lev 26:22; Num 21:6; 22:40; Deut 7:20; 28:20 (but in Alexandrinus ἐξαποστέλλω); 32:24 (but in Alexandrinus ἐξαποστέλλω). 99 No qal, in piel, only twice with the preposition ‫( ִמן‬Gen 8:8 ἀπό corresponds here to ‫מן‬, ִ see above, Gen 26:27, ἀπό also corresponds here to ‫מן‬, ִ but see the Hexaplaric edition with ἐξαποστέλλω; cf. Gen 25:6); otherwise in Gen 8:7, see above. For Gen 21:14, the meaning of the piel is clearly “to dismiss” (Hagar). In Gen 30:25, the context is also “to release,” but the Greek translators clearly understand a nuance of finality by using ἵνα which may have been interpreted as a kind of mission. In Gen 32:26 (twice), the context is clearly “to release” (Jacob). Exodus 10:10 in which Pharaoh does not want to release the people (see Exod 8:28 with ἐξαποστέλλω— in Alexandrinus or Exod 11:1). Leviticus 16:10, the ritual of the day of atonement with the sending away of the goat (the other occurrences of this passage are with ἐξαποστέλλω; see Lev 16:21.22.26). Deuteronomy 22:7 in which one is required to release the mother bird in a nest. 100  Six times in qal: Exod 3:12 (but ἀποστέλλω in Alexandrinus; see also ἐξ Αἰγύπτου some words later); 9:14 (God sending plagues); 24:5; Num 13:3 (but ἀποστέλλω in Alexandrinus; see ἐκ τῆς ἐρήμου Φαραν some words later), Deut 9:23 (but ἀποστέλλω in Alexandrinus with ἐκ Καδης Βαρνη some words later), and only twice in piel: Lev 26:25; Deut 28:20 (in both cases, God sends plagues). 101 No qal, in piel, 18 times with the preposition ‫מן‬: ִ Gen 3:23; 8:10; 19:29; 25:6 (but ἀπό corresponds here to ‫;)מן‬ ִ Exod 6:11, 13; 7:2; 11:10; Lev 18:24 (prep. πρό); 20:23 (prep. ἀπό); Num 5:2, 3, 4; Deut 15:12 (prep. ἀπό), 13(prep. ἀπό), 18(prep. ἀπό); 24:1.3. Otherwise in Gen 8:12; 26:29; 26:31; 31:27; 31:42; 45:24; Exod 3:12 (see note above), 3:20; 4:21, 23(bis); 5:1, 2(bis); 6:1 (‫ ִמן‬is present in the parallel hemistich); 7:14, 16, 26, 27; 8:4, 16, 17.24 (but ἀποστέλλω in Vaticanus); 8:25, 28; 9:1, 2, 7, 13, 17, 28, 35; 10:3, 4, 7, 20, 27; 11:1(bis); 13:15, 17; 14:5; 18:27; 21:26.27; Lev 14:7.53; 16:21, 22, 26; Deut 21:14; 22:19, 29; 24:4.

116

Patrick Pouchelle

ἐξαποστέλλω is preferred,102 possibly as the use of a verb with a prefix followed by the related preposition.103 This could, by itself, explain the over-representation of the correspondence with the piel in the MT.

5. Conclusion The question this essay attempted to address is “did the translators really care about the difference between qal and piel?” The status quaestionis suggested some solutions such as the use of the passive voice for the qal and the active voice for the piel while staying cautious about the problem of knowing what the Vorlage of the translators was, how they vocalized it, and in what sense they analysed given forms in a similar way to our modern notions of qal and piel. The three main limitations of this investigation, i.e., the study of the Pentateuch only, the study of the rendering of qal and piel only, and a non-exhaustive survey of the Hebrew verbs, are an impetus for further research. However, some provisional conclusions could be given. When a given verb is stative in qal, it is usually fientive in the piel. In this case, it is not very difficult to observe a correlation between the Septuagint and the stems in the MT, as the translators probably constructed their translations according to their semantic analysis of the context and as the Masoretes vocalized104 the verb according to a similar context. For ‫טהר‬, the translators chose the same Greek verb καθαρίζω: in the passive voice, corresponding to what we interpret as the stative status of ‫( טהר‬i.e. qal in the MT), and in the active voice corresponding to the fientive status of ‫( טהר‬i.e. piel in the MT). When they interpret a given form of the stative verb as an adjective, they use the syntagm καθαρὸς εἶναι.

102  See notably the dismissal of the people by Pharaoh in Exodus (e.g., Exod 3:20, 4:21, …). Exodus 7:16 is interesting: Κύριος ὁ θεὸς τῶν Εβραίων ἀπέσταλκέν (‫ ְש ָל ַחנִ י‬qal) με πρὸς σὲ λ­ έγων Ἐξαπόστειλον (‫ ַש ַלח‬piel) τὸν λαόν μου, ἵνα μοι λατρεύσῃ ἐν τῇ ἐρήμῳ. The first ἀποστέλλω denotes the mission given by God to Moses, ἐξαποστέλλω denotes the release of the people, even if we were to interpret ‫ וְ ַי ַֽע ְב ֻדנִ י ַב ִמ ְד ָבר‬as a final clause and accordingly give ‫ ׁשלח‬a nuance of mission. However, this is difficult to accept, Pharaoh is not the sender of the people. This is probably why the MT uses the piel here and why the LXX uses ἐξαποστέλλω. Hence these cases will be classified under the category of bivalent piel meaning to “dismiss” (e.g. Exod 8:4). 103  See notably the contribution of Joosten in this volume. This is not systematic, as ‫ ִמן‬could also be rendered by ἀπό either with ἀποστέλλω (see n. 88) or ἐξαποστέλλω (see n. 90). Another possible case of this technique is to be found in Exod 8:17: ἰδοὺ ἐγὼ ἐπαποστέλλω ἐπὶ σὲ […] κυνόμυιαν, corresponding to ‫ ׁשלח‬governing the preposition ‫ב‬. 104  Or put into writing an oral tradition of reading that vocalized this way.

Did the Translators of the Pentateuch Differentiate the Piel from the Qal?

117

Regarding ‫למד‬, the translators also constructed their translation around the semantic context. They offered one Greek verb, μανθάνω, for what we interpret as the stative status and διδάσκω for the fientive nature of the Hebrew verb. There is no reason to postulate a specific knowledge of the vocalization or the way of interpreting it as a specific conjugation. When both qal and piel are fientive, a correlation between the LXX and the MT would have been proof that the translators could analyse a given vocalization as a specific stem with a specific meaning, such as intensification, for instance. The non-exhaustive analysis of some verbs of this kind shows that there either is no correlation between the LXX and the MT, as for ‫ׁשבר‬, or that the correlation could be explained otherwise, as for ‫ׁשלח‬. If we accept that the translators knew the vocalization of the text and were aware of a difference between what we now call qal and piel, then the absence of correlation found for the fientive verbs in both stems could be an indication that there were no more differences between these stems in those times and those areas. Indeed, these verbs are sometimes used synonymously and the choice between the two stems might be stylistic.105 Historically, it explains why the qal of these stems fell into disuse. In late biblical Hebrew, from Sira onward to Mishnaic Hebrew, as well as in the Dead Sea scrolls and the Samaritan Pentateuch, one may have observed a tendency to replace the qal stem by the piel.106 It is difficult to date the occurrence of this phenomenon, but the closeness in classical Hebrew of the two stems for some roots may indicate that this evolution dates back a long time and is possibly compatible with the time of the translation of the Pentateuch. Hence, for ‫ׁשלח‬, we cannot exclude that the translators had a tradition of reading close to the Samaritan one where this verb is always in the piel;107 this will not alter their translation technique. This remains to be studied more thoroughly, as the translators may well have also acknowledged the different vocalizations for these forms but did not give them any 105  See Ryder, The D-Stem, 121–122. 106  See Z. Ben-Hayyim, “La tradition samaritaine et sa parenté avec les autres traditions de la langue hébraïque”, Mélanges de philosophie et de littérature juives 3–5 (1958–1962) 89–128 (112–20, where he suggests this phenomenon in 1QIsaa [p. 113]). See also the examples given by E.Y. Kutscher, The Language and Linguistic Background of the Isaiah Scroll (1 QIsaa) (STDJ 6; Leiden: Brill, 1974), 358, based on the polel of ‫ רנן‬in Isa 65:14, with ‫ ירננו‬instead of ‫ירנו‬, but the qal is exceptional in the MT. See also ‫ תעופפנה‬instead of ‫ תעופינה‬in Isa 50:8, but E. Qimron, The Hebrew of the Dead Sea Scrolls (HSS 29; Atlanta: Scholars Press, 1985), §311.2, does not observe any specific treatment in the piel stem between Biblical Hebrew and Qumran Hebrew. See also Z. Ben-Hayyim/A. Tal, A Grammar of Samaritan Hebrew: Based on the Recitation of the Law in Comparison with the Tiberian and other Jewish traditions (Winona Lake, IN: Eisenbrauns, 2000), §2.15.5. See also S.E. Fassberg, “The Movement from Qal to Pi“el in Hebrew and the Disappearance of the Qal Internal Passive”, HS 42 (2001) 243–5 and T. Zurawel, “The Qal Conjugation in Samaritan Hebrew”, in M. Bar-Asher (ed.), Massorot: Studies in Language Traditions (Jerusalem: Magnes Press, 1984) 135–51. 107  See Ben-Hayyim, “La tradition samaritaine”, 114.

118

Patrick Pouchelle

specific semantic nuance, similarly to this assertion of Kaufman: “[I]t is pointless to try to find a single … explanation to account for all of the transformative power of the D Stem. It is simply a form.”108

108  S.A. Kaufman “Semitics: Directions and Re-directions”, in J.S. Cooper/G.M. Schwartz (ed.), The Study of the Ancient Near East in the Twenty-First Century (Winona Lake, IN: Eisenbrauns, 1996) 273–82, on pp. 282.

Philippe Le Moigne Une tournure syntaxique fréquente en Ésaïe-LXX Substantif abstrait + λήμψεται + complément d’objet direct

1. Introduction Le verbe λαμβάνω (essentiellement ‹ prendre › ou ‹ recevoir ›)1, des plus fréquents en grec classique, l’est aussi dans la version de la LXX, puisque, selon le dictionnaire LEH2, on le lit 1335 fois dans l’ensemble du corpus. En ce qui concerne le livre d’Ésaïe, il ne se rencontre pas moins de 48 fois, où il correspond, d’après la concordance de Hatch/Redapth, à 12 termes hébreux différents;3 si l’on ajoute six † de la concordance, signes qui indiquent que la correspondance mot à mot est impossible selon ses auteurs,4 ainsi que deux—, qui signifient que le verbe grec ne correspond purement et simplement à rien dans le TM,5 on se rend facilement compte que ce verbe fait partie des mots favoris de l’auteur de l’Ésaïe grec, auquel il recourt dans des contextes différents, face à des Vorlagen de différents visages. Parmi toutes ces occurrences de ce verbe, nous en sélectionnerons quelquesunes, onze plus précisément, qui présentent une similitude syntaxique frappante; en effet, dans chacun de ces passages, le verbe λαμβάνω se présente à la troisième personne du singulier du futur de l’indicatif, sous la forme λήμψεται donc, avec un complément d’objet direct qui, à une seule exception près,6 désigne un animé (des personnes vivantes), et un sujet qui est un nom abstrait ou, plus rarement, un nom de chose. Si l’on ajoute les deux préverbés ἀναλαμβάνω (une occurrence dans la tournure étudiée) et καταλαμβάνω (trois occurrences, dont une sous la 1  Voici ce qu’en dit P. Chantraine dans son dictionnaire étymologique: « Sens: ‹ prendre › (mais n’est pas exactement synonyme de αἱρέω ‹ prendre, enlever, supprimer ›), avec les emplois particuliers ‹ recevoir ›, ‹ posséder ›, ‹ gagner › ‹ comprendre ›, etc. ». Aucune mention n’est faite par l’auteur d’une quelconque valeur spécifique du futur: Dictionnaire étymologique de la langue grecque: Histoire des mots (Librairie Klincksieck—Série linguistique; Paris: Klincksieck, 2009), 616a. 2  Voir la bibliographie à la fin du présent article. 3  Compte non tenu des différences de binyanim des verbes hébreux que la concordance note par des lettres minuscules; ainsi pour ‫ לקח‬trouve-t-on en Ésaïe, pour correspondre à λαμβάνω, le qal, le pual, le hophal, et la forme ‫ ;מלקוח‬pour ‫נׂשא‬, il s’agit soit du qal soit du niphal. 4  Isa 10.9 bis; 10.10; 11.5; 28.4; 30.28. 5  Isa 26.11, qui fait partie de notre corpus et que nous discuterons au fil de notre étude, et 31.4: ἐπὶ τῇ θήρᾳ, ᾗ ἔλαβεν « sur la proie qu’il a saisie » en face de ‫« על־טרפו‬ sur sa proie ». 6  Deux en fait, mais les deux occurrences sont strictement identiques, à deux versets d’écart (64.1,3); voir infra, exemples 14 et 15.

120

Philippe Le Moigne

forme de la négation emphatique οὐ μὴ καταλάβῃ, en 59.9), nous avons là un corpus de quinze phrases qui, en grec, présentent une unité indéniable7, là où le texte massorétique, comme cela se devine à partir de ce que nous avons dit des correspondances observées dans la concordance Hatch/Redpath, ne permet pas nécessairement un tel rapprochement. Il s’agit donc indéniablement d’une formulation aimée du traducteur, et nous allons tâcher tout d’abord de formuler des hypothèses sur la genèse de ce tour en Ésaïe-LXX, puis d’en observer la généralisation dans des cas où le TM n’offrait pas de circonstances manifestement favorables à l’emploi d’un tel tour, auquel a pourtant recouru l’auteur de la LXX d’Ésaïe—même si, sémantiquement, on peut trouver des analogies—, pour finir par des passages où le tour s’observe alors que le TM est en bien éloigné.

2. Les matrices du tour 1 41.168 15. ἰδοὺ ἐποίησά σε ὡς τροχοὺς ἁμάξης ἀλοῶντας καινοὺς πριστηροειδεῖς, καὶ ἀλοήσεις ὄρη καὶ λεπτυνεῖς βουνοὺς καὶ ὡς χνοῦν θήσεις·

16. καὶ λικμήσεις, καὶ ἄνεμος λήμψεται αὐτούς, καὶ καταιγὶς διασπερεῖ αὐτούς, σὺ δὲ εὐφρανθήσῃ ἐν τοῖς ἁγίοις Ισραηλ. καὶ ἀγαλλιάσονται

15 Voici,

je t’ai fait comme les roues d’un chariot qui battent l’aire, nouvelles, pareilles à des scies, et tu battras les montagnes, tu broieras les collines et feras d’elles comme un duvet; 16 tu vanneras et le vent les saisira, le tourbillon les dispersera; mais toi, tu te réjouiras dans les lieux saints d’Israël.

15. Voici que je fais de toi une herse, un traîneau neuf à deux tranchants; tu fouleras les montagnes, tu les broieras, et tu rendras les collines pareilles à la bale. 16. Tu les vanneras, et le vent les emportera ‫ורוח תשׂאם‬, la tempête les dispersera; et toi, tu jubileras en Yahvé dans le Saint d’Israël tu te loueras.

7  Ce que Bybee appelle « prefabricated expression »: voir J. Bybee, Language, Usage and Cognition (Cambridge: Cambridge Université Press, 2010). 8  Dans les synopses de cet essai, la première colonne représente le texte d’Ésaïe édité par J. Ziegler—Duodecim prophetae. Septuaginta. Vetus Testamentum Graecum auctoritate Academiae Scientiarum Gottingensis editum. Vol. XIX, Isaias, (Göttingen: Vandenhoeck & Ruprecht, 1939, 21967— la deuxième, sa traduction parue dans Vision que vit Isaïe, par A. Le Boulluec et P. Le Moigne—Vision que vit Isaïe. Traduction du texte du prophète Isaïe selon la Septante. Index littéraire des noms propres et glossaire (Paris: Cerf, 2014) — la troisième, la traduction du TM que l’on lit dans La Bible Osty—Osty, E./Trinquet, J., La Bible. Traduction française sur les textes

Une tournure syntaxique fréquente en Ésaïe-LXX

17. οἱ πτωχοὶ καὶ οἱ ἐνδεεῖς· ζητήσουσιν γὰρ ὕδωρ, καὶ οὐκ ἔσται, ἡ γλῶσσα αὐτῶν ἀπὸ τῆς δίψης ἐξηράνθη· ἐγὼ κύριος ὁ θεός, ἐγὼ ἐπακούσομαι, ὁ θεὸς Ισραηλ, καὶ οὐκ ἐγκαταλείψω αὐτούς, 18. ἀλλὰ ἀνοίξω ἐπὶ τῶν ὀρέων ποταμοὺς καὶ ἐν μέσῳ πεδίων πηγάς, ποιήσω τὴν ἔρημον εἰς ἕλη καὶ τὴν διψῶσαν γῆν ἐν ὑδραγωγοῖς,

17 Et

les pauvres et les indigents exulteront, car ils auront cherché de l’eau; il n’y en aura pas eu, leur langue aura été desséchée par la soif: moi, le Seigneur Dieu, moi j’écouterai, le Dieu d’Israël, et je ne les abandonnerai pas, 18 mais j’ouvrirai des fleuves sur les montagnes, et des sources au milieu des plaines; je ferai du désert un marais, et de la terre assoiffée, des conduites d’eau.

121

17. Les malheureux et les indigents cherchent de l’eau; et… rien! Leur langue est desséchée par la soif. Moi, Yahvé, je les exaucerai, [Moi], le Dieu d’Israël, je ne les abandonnerai pas. 18. Sur les monts dénudés je ferai couler des fleuves, et des sources au milieu des vallées; je changerai le désert en étangs, la terre aride en sources d’eau.

Si l’on observe le TM, on constate deux moments différents dans les quatre versets cités. 15–16 opposent le sort de l’allocutaire, le peuple de Dieu, qui vanne les montagnes / collines, et qui finit par se réjouir, à ces dernières; 17–18 passent à un tout autre sujet, le sort des humiliés, qui cherchent de l’eau sans qu’il n’y en ait—en dehors, bien sûr, de l’intervention du Seigneur qui « ne les abandonner[a] pas ». On notera l’isotopie de l’eau dans ces deux derniers versets, eau d’abord manquante, puis donnée à profusion par le bienfaiteur divin. On retrouve cette isotopie de l’eau dans la LXX, mais le passage ne y nous paraît exactement superposable à son correspondant massorétique. En effet, tout se joue autour du αὐτούς (v. 16) qui touche la tournure qui intéresse la présente étude. On peut certes, d’après un emploi très courant dès le grec classique, considérer ici l’anaphorique comme renvoyant aux « habitants » des territoires déjà nommés, à savoir les montagnes et les collines, αὐτός renvoyant ad sensum aux lieux précédemment évoqués.9 Cependant, le découpage syntaxique différent, qui s’observe au tournant des versets 16–17, avec notamment l’expression καὶ ἀγαλλιάσονται / οἱ πτωχοὶ καὶ οἱ ἐνδεεῖς, originaux par Émile Osty avec la collaboration de Joseph Trinquet. Introductions et notes d’Émile Osty et de Joseph Trinquet (sl: Seuil, 1973). 9  Par exemple Iliade 24.499 εἴρυτο δὲ ἄστυ καὶ αὐτούς, « et il protégeait la ville et ses habitants »; Hérodote 3.72 ἐξηγέο αὐτὸς ὅτεῳ τρόπῳ πάριμεν ἐς τὰ βασιλήια καὶ ἐπιχειρήσομεν αὐτοῖσι « explique, toi en personne, comment nous parviendrons au palais et comment nous attaquerons ses occupants » (traductions PLM).

122

Philippe Le Moigne

conduit à une interprétation différente en ce qui concerne la LXX. On notera notamment l’absence de complément d’objet direct pour le verbe λικμήσεις du v. 16; dès lors, les compléments d’objets directs qui suivent, les deux αὐτούς, peuvent être considérés comme cataphoriques, annonçant le sort des « pauvres » et des « indigents ». En effet, il s’observe dans les versets 15–16 le même double mouvement que dans les versets 17–18: détresse puis salut. On aurait donc, dans la LXX de ces quatre versets, une unité sémantique que l’on ne retrouve pas dans le TM. 2 57.13 ὅταν ἀναβοήσῃς, ἐξελέσθωσάν σε ἐν τῇ θλίψει σου· τούτους γὰρ πάντας ἄνεμος λήμψεται καὶ ἀποίσει καταιγίς. οἱ δὲ ἀντεχόμενοί μου κτήσονται γῆν καὶ κληρονομήσουσιν τὸ ὄρος τὸ ἅγιόν μου.

Lorsque tu pousseras un cri, qu’ils te délivrent dans ton oppression; car ceux-là, un vent les saisira tous, et un tourbillon les emportera; mais ceux qui s’attachent à moi posséderont la terre et hériteront de ma montagne sainte.

Quand tu crieras, qu’elles te délivrent, tes Ordures ! Le vent les enlèvera toutes,

‫ואת כלם יׂשא־רוח‬

un souffle les emportera. Mais qui s’abrite en moi héritera du pays et il possédera ma montagne sainte.

La LXX est globalement semblable au TM; on note une différence dans l’expression ἐν τῇ θλίψει σου, à quoi correspond dans le TM ‫« קבוציך‬ ceux qui t’ont rassemblé ». On remarquera cependant que le sujet de l’impératif ἐξελέσθωσαν n’est pas exprimé dans la LXX, ce qui confirme le néant des faux sauveurs, qui ne méritent même pas d’être nommés; rappelons que le démonstratif οὗτος, à quoi rien ne correspond dans le TM, peut avoir une valeur péjorative. De même le pluriel à la fin du verset, en grec, fait référence à l’ensemble du peuple, alors que le TM n’envisage qu’un cas particulier. La traduction de ‫ רוח‬par ἄνεμος, qui est la règle dans la LXX,10 plutôt que par πνεῦμα, préfigure le superlatif du vent destructeur, qui est la καταιγίς; remarquons que nous avons la même succession dans le passage précédemment étudié. On notera en outre que la LXX présente un chiasme, là où le TM préfère un parallélisme synonymique (second terme: ‫ ;)יקח־הבל‬or le chiasme est une figure de la clôture, employé ici à parfait escient par l’auteur d’Ésaïe-LXX. En effet, cette figure parfait le sort réservé aux prétendus sauveurs, après quoi on passe à un autre sujet, avec le ‫ ו‬rendu par le δέ oppositif.

10  Il n’y a pas d’autre mot hébreu qui corresponde au mot grec d’après la concordance.

Une tournure syntaxique fréquente en Ésaïe-LXX

3 40.24 οὐ γὰρ μὴ σπείρωσιν οὐδὲ μὴ φυτεύσωσιν, οὐδὲ μὴ ῥιζωθῇ εἰς τὴν γῆν ἡ ῥίζα αὐτῶν· ἔπνευσεν ἐπ’ αὐτοὺς καὶ ἐξηράνθησαν, καὶ καταιγὶς ὡς φρύγανα ἀναλήμψεται αὐτούς.

Car ils ne sèmeront ni ne planteront, et leur racine ne s’enracinera pas en terre; il a soufflé sur eux et ils se sont desséchés, et un tourbillon les emportera comme des brindilles.

123

À peine sont-ils plantés, à peine sont-ils semés, à peine leur racine a-telle pris racine en terre qu’il souffle sur eux: ils sèchent et la tempête les emporte comme de la paille ‫וסערה כקׁש כׂשאם‬.

Nous sommes dans un passage éminemment métaphorique; il est question des souverains,11 et l’on note la différence de voix entre les deux textes, actifs dans la LXX (σπείρωσιν / φυτεύσωσιν), et passifs dans le TM (‫נטעו‬, ‫)זרעו‬. Une nouvelle fois la καταιγίς est le superlatif du πνεῦμα, et, conformément à l’usage général de la tournure que nous étudions, nous avons le futur ἀναλήμψεται, qui opère un retour vers un procès situé dans l’avenir, après un passage où se lisent deux verbes à l’aoriste. Ce détour par l’aoriste, avec l’asyndète avant ἔπνευσεν (vs TM ‫וגמ־נׁשף‬ ‫)בהם‬, peut avoir une valeur causale: s’ils ne doivent être ni plantés ni semés, etc., c’est parce que le souverain suprême, Dieu, a soufflé sur eux. La fin du verset, par la particule ὡς, marque une démétaphorisation du propos (nous sommes désormais dans la comparaison, et non plus dans l’implicite de la métaphore), et confère à nouveau un sens propre au complément du verbe de la tournure étudiée.

11  40.23 LXX « lui qui donne des souverains pour qu’ils soient souverains du néant, et il a fait de la terre comme un néant », ὁ διδοὺς ἄρχοντας εἰς οὐδὲν ἄρχειν, τὴν δὲ γῆν ὡς οὐδὲν ἐποίησεν.

124 4 59.9 8. καὶ ὁδὸν εἰρήνης οὐκ οἴδασιν, καὶ οὐκ ἔστιν κρίσις ἐν ταῖς ὁδοῖς αὐτῶν· αἱ γὰρ τρίβοι αὐτῶν διεστραμμέναι, ἃς διοδεύουσιν, καὶ οὐκ οἴδασιν εἰρήνην. 9. διὰ τοῦτο ἀπέστη ἡ κρίσις ἀπ’ αὐτῶν, καὶ οὐ μὴ καταλάβῃ αὐτοὺς δικαιοσύνη· ὑπομεινάντων αὐτῶν φῶς ἐγένετο αὐτοῖς σκότος, μείναντες αὐγὴν ἐν ἀωρίᾳ περιεπάτησαν.

Philippe Le Moigne 8 Et

ils ne connaissent pas la voie de la paix, et dans leurs voies il n’est pas de droit, car les chemins qu’ils frayent sont tortueux et ils ne connaissent pas la paix. 9 C’est pourquoi le droit s’est éloigné d’eux, et la justice ne les trouvera pas. Ils s’attendaient à la lumière, et l’obscurité est venue sur eux; ils attendaient la clarté et ils ont piétiné dans la nuit noire.

8. Ils ne connaissent pas le chemin de la paix, le droit n’est pas dans leurs sentiers ; ils se font des voies tortueuses, quiconque y marche ne connaît point la paix. 9. Voilà pourquoi le droit reste éloigné de nous et la justice ne nous atteint pas  ‫ ;ולא תׂשיגנו צדקה‬nous attendions la lumière, et voici les ténèbres, la clarté, et nous marchons dans l’obscurité !

Ce qui attire l’attention lorsque l’on lit les deux textes, c’est bien évidemment la différence de personne qui s’observe dans le second verset. Le TM passe de la troisième personne du pluriel à la première personne du pluriel, tandis que la LXX conserve la troisième personne du pluriel du premier verset.12 Cela permet la présence de la tournure étudiée, toujours au futur, ici sous la forme du futur emphatique par οὐ μή suivi du subjonctif aoriste, et un complément à la troisième personne. Ce futur est d’autant plus remarquable que l’ensemble des autres verbes du verset est au passé; cela dit, le traducteur a utilisé l’aoriste ἀπέστη qui vaut, sémantiquement, quasiment un parfait: si « le droit s’est éloigné d’eux », c’est qu’à présent il est, précisément, loin d’eux; après le tour étudié, le retour à l’aoriste se fait par le « tampon » du génitif absolu, que l’on peut d’ailleurs lire de deux manières différentes: celle de la traduction proposée, avec φῶς comme COD du participe ὑπομεινάντων, et une autre, plus originale, où ce dernier participe serait construit absolument (« alors qu’ils étaient dans l’attente »), et φῶς comme sujet de ἐγένετο (« la lumière est devenue obscurité »), sur le modèle, par exemple de 5.20 οἱ τιθέντες τὸ σκότος φῶς καὶ τὸ φῶς σκότος ou de 42.16 ποιήσω αὐτοῖς τὸ σκότος εἰς φῶς.

12  Il est frappant de constater que cette différence de personne se poursuit, de manière extrêmement cohérente, jusqu’au verset 11a inclus.

Une tournure syntaxique fréquente en Ésaïe-LXX

5 28.19 17. καὶ θήσω κρίσιν εἰς ἐλπίδα, ἡ δὲ ἐλεημοσύνη μου εἰς σταθμούς, καὶ οἱ πεποιθότες μάτην ψεύδει· ὅτι οὐ μὴ παρέλθῃ ὑμᾶς καταιγίς, 18. μὴ καὶ ἀφέλῃ ὑμῶν τὴν διαθήκην τοῦ θανάτου, καὶ ἡ ἐλπὶς ὑμῶν ἡ πρὸς τὸν ᾅδην οὐ μὴ ἐμμείνῃ· καταιγὶς φερομένη ἐὰν ἐπέλθῃ, ἔσεσθε αὐτῇ εἰς καταπάτημα. 19. ὅταν παρέλθῃ, λήμψεται ὑμᾶς· πρωὶ πρωὶ παρελεύσεται ἡμέρας, καὶ ἐν νυκτὶ ἔσται ἐλπὶς πονηρά· μάθετε ἀκούειν.

17 Et

je ferai de la justice un espoir et de ma miséricorde, des piliers. Et vous qui mettez vainement votre confiance dans le mensonge (« le tourbillon ne passera pas sur vous, 18 il ne supprimera pas non plus votre alliance avec la Mort »), eh bien votre espoir dans l’Hadès ne demeurera pas; le tourbillon déchaîné, s’il passe, vous serez pour lui bons à piétiner. 19 Quand il passera, il vous saisira; tôt, très tôt il passera de jour, et dans la nuit l’espoir sera cruel. Apprenez à écouter !

125

17. Je ferai du droit un cordeau, et de la justice un niveau. La grêle balaiera l’abri du mensonge et les eaux entraîneront votre asile. 18. Votre alliance avec la Mort sera rompue, et votre pacte avec le Chéol ne tiendra pas. Quand passera le fléau torrentiel, vous en serez écrasés; 19. chaque fois qu’il passera, il vous saisira; car il passera chaque matin, le jour et la nuit, et il n’y aura qu’épouvante à comprendre le message.

L’ensemble du passage se comprend à la lumière du v. 15 du même chapitre, qui exprime les paroles des insensés qui ne mettent pas leur espoir dans le Seigneur; la LXX, que nous citons ici, est proche du TM: « 15. Puisque vous avez dit: ‹ Nous avons fait alliance avec l’Hadès, et avec la Mort nous avons conclu des pactes; le tourbillon déchaîné, s’il passe, ne viendra pas sur nous. Nous avons fait de notre espoir un mensonge et le mensonge sera notre abri ›, 16. c’est pourquoi ainsi parle le Seigneur: ‹ Voici, moi je poserai sur les assises de Sion une pierre de grand prix, de choix, une pierre d’angle, précieuse, sur ses assises; et qui croit en elle ne sera pas couvert de confusion ›». Il y a une différence frappante entre les deux textes, qui consiste dans la présence d’une négation au début du v. 18 dans la LXX, là où le TM a un parallélisme synonymique comprenant deux propositions allant dans le même sens (« Votre alliance avec la Mort sera rompue, et votre pacte avec le Chéol ne tiendra pas »). Alain Le Boulluec et nous-même avons compris, pour sauvegarder la cohérence du propos, que la proposition par ὅτι qui fait la transition entre les versets 17 et 18 était une citation, bien entendue ironique, des propos antérieurs de ces insensés, propos que nous avons rapportés en mentionnant le

126

Philippe Le Moigne

contenu du verset 1513. Selon cette analyse, le καὶ de καὶ ἡ ἐλπίς en 18 est une conjonction apodotique, et, plus subtilement encore, le οἱ πεποιθότες μάτην ψεύδει du verset 17 est un nominatif absolu (voire un vocatif), repris par le génitif ὑμῶν au verset suivant. Une différence lexicale qui s’observait déjà aux versets 15–16 se trouve amplifiée dans la synopse dont nous proposons la lecture. En effet, là où la LXX disait, au v. 15, ἐθήκαμεν ψεῦδος τὴν ἐλπίδα ἡμῶν « nous avons fait de notre espoir un mensonge », le TM lit ceci: « car nous avons fait du mensonge notre abri », ‫כי ׂשנמו כזב מחסנו‬. Or ce mot ‹ espoir › du grec, ἐλπίς, qui ne se lisait déjà pas dans ce verset 15 du TM,14 ne se trouve pas moins que trois fois dans le passage cité en synopse, toujours sans correspondant massorétique immédiat. Il est évident que ce terme, ou plutôt la notion qu’il dénote, est ici central dans le projet littéraire du traducteur. Cet ‹ espoir › est bien évidemment le pacte passé avec le shéol; ainsi, face au parallélisme synonymique du TM déjà mentionné, au v. 18, où l’on a deux termes équivalents (‫ בריתכם‬puis ‫)חזותכם‬, le grec, reprenant le propos des insensés, mentionne bien en premier la διαθήκη, l’‹ alliance ›, mais parle ensuite d’ἐλπίς, tout cela en lien avec la rupture discursive que nous proposons de lire dans la version grecque, avec la succession du discours indirect libre et de la reprise de la narration prophétique. Dans ce contexte, comment interpréter le tour avec λαμβάνω au futur plus COD? Il est à noter que, exceptionnellement dans notre corpus, le sujet du verbe n’est pas exprimé; ou plutôt il n’est pas réexprimé; en effet, si nous relisons le grec, indépendamment de la division en versets qui dépend de la structure du texte hébreu, nous trouvons ceci: καταιγὶς φερομένη ἐὰν ἐπέλθῃ, ἔσεσθε αὐτῇ εἰς καταπάτημα. ὅταν παρέλθῃ, λήμψεται ὑμᾶς· Il y donc la succession d’une hypothétique, qui laisse ouvert le champ du possible—la chose arrivera ou n’arrivera pas—, et d’une temporelle, où la chose est présentée comme certaine, sa venue n’étant qu’une question de temps. Le fait que le sujet, la καταιγίς, n’est pas explicitement répété confère à l’ensemble du passage une note de dérision: ce qui ‘vous emportera’, λήμψεται ὑμᾶς, est un néant syntaxique, ce qui confirme poétiquement l’inanité de l’espoir des personnes en question, thématique très présente dans la LXX, alors qu’elle est absente, nous l’avons vu, dans le TM correspondant. Nous lisons bien l’expression d’ἐλπὶς πονηρά à la fin de l’extrait. Et même la finale 13  « In his reaction to the bold statements of the people, God makes use of some of their words in order to contradict them. By the reformulation of verse 17b this repetition has been increased by the translator »: M. van der Vorm-Croughs, The Old Greek of Isaiah: An Analysis of its Pluses and Minuses (Atlanta: SBL Press, 2014), 324–5. 14  28.15 TM: « Vous dites: ‹ Nous avons conclu une alliance avec la Mort, et avec le Chéol nous avons fait un pacte; quand passera le fléau torrentiel, il ne nous atteindra pas, car nous avons fait du mensonge notre abri, et dans la tromperie nous nous sommes réfugiés. ›» Voir R.L. Troxel, LXX-Isaiah as Translation and Interpretation: The Strategies of the Translator of the Septuagnt of Isaiah (Leiden – Boston: Brill, 2008), 259; 263–5.

Une tournure syntaxique fréquente en Ésaïe-LXX

127

absolue de notre passage, l’impératif μάθετε ἀκούειν, à quoi rien ne correspond vraiment de manière claire dans le TM, peut être considérée comme une critique de cette espérance infondée. Elle rappelle 1.17 μάθετε καλὸν ποιεῖν ou encore 26.9 δικαιοσύνην μάθετε, et oppose le bon enseignement venu de Dieu aux croyances néfastes des humains égarés. 6 33.14 ἀπέστησαν οἱ ἐν Σιων ἄνομοι, λήμψεται τρόμος τοὺς ἀσεβεῖς· τίς ἀναγγελεῖ ὑμῖν ὅτι πῦρ καίεται; τίς ἀναγγελεῖ ὑμῖν τὸν τόπον τὸν αἰώνιον;

Ils sont partis, les transgresseurs de Sion; le tremblement saisira les impies. Qui vous annoncera que le feu s’embrase? Qui vous annoncera le lieu éternel?

À Sion les pécheurs sont effrayés ‫פחדו‬, un tremblement a saisi les impies ‫אחזה רעדה חנפים‬: « Qui de nous séjournera dans un feu dévorant? Qui de nous séjournera dans un brasier éternel? »

Ici, à la différence de la plupart des passages étudiés précédemment, les différences entre les deux textes sont nombreuses15. D’un point de vue strictement grammatical, la LXX passe au futur avant le TM (un verbe à l’aoriste puis trois verbes au futur dans la LXX, contre 2 verbes au qatal dans le TM, suivis de deux verbes au yiqtol). Cette anticipation du futur permet de retrouver la structure en question dans la présente étude, le verbe λήμψεται avec un sujet abstrait et un complément d’objet direct. Il est probable que le choix du futur de la part de l’auteur d’Ésaïe, en face du qatal du TM, provient de la prégnance de la formule étudiée. Cela dit il n’est pas à exclure que, dans la Vorlage, il y ait pu se trouver un yiqtol, induit par les mêmes raisons qui ont poussé l’auteur d’Ésaïe-LXX à proposer un futur. Si l’on se penche maintenant plus précisément sur le visage propre qu’offre chacune des deux versions, on note dans l’une et l’autre un double parallélisme synonymique. La présence, dans la LXX, d’un verbe qui indique l’éloignement (ἀπέστησαν), jointe à la différence de personne observée dans la seconde partie du verset (« vous » en grec, « nous » en hébreu) souligne l’idée d’absence dans la LXX; nul n’est présent pour avertir le peuple du danger immédiat; la première partie du verset répond ainsi au verset précédent, 33.13: ἀκούσονται οἱ πόρρωθεν ἃ ἐποίησα, γνώσονται οἱ ἐγγίζοντες τὴν ἰσχύν μου « ceux qui sont loin entendront ce que j’ai fait, ceux qui sont proches connaîtront ma force ». Tandis que le TM, avec sa première personne du pluriel (‫ לנו‬X 2), exprime des interrogations 15  « The ad hoc character of this translation is evident in several ways »: Troxel, LXX-Isaiah, 114.

128

Philippe Le Moigne

angoissées, que l’on peut comprendre comme étant le discours des « pécheurs » et des « impies », en raison du lien que l’on y observe avec la frayeur doublement présente dans le premier stique, la LXX, de son côté, prépare la place aux versets qui suivent, 33.15–16: πορευόμενος ἐν δικαιοσύνῃ, λαλῶν εὐθεῖαν ὁδόν, μισῶν ἀνομίαν καὶ ἀδικίαν καὶ τὰς χεῖρας ἀποσειόμενος ἀπὸ δώρων, βαρύνων τὰ ὦτα ἵνα μὴ ἀκούσῃ κρίσιν αἵματος, καμμύων τοὺς ὀφθαλμοὺς ἵνα μὴ ἴδῃ ἀδικίαν, / οὗτος οἰκήσει ἐν ὑψηλῷ σπηλαίῳ πέτρας ἰσχυρᾶς· ἄρτος αὐτῷ δοθήσεται, καὶ τὸ ὕδωρ αὐτοῦ πιστόν « marchant dans la justice, disant une voie droite, haïssant la transgression et l’injustice, chassant des mains les cadeaux, se bouchant les oreilles pour ne pas entendre le droit du sang, fermant les yeux pour ne pas voir l’injustice, / c’est lui qui habitera dans une grotte élevée, de roche dure; on lui donnera du pain, et son eau sera fidèle. » Ainsi, le λήμψεται du v. 14, objet de la figure étudiée, introduit toute la suite des futurs et confère à donner au texte de la LXX un visage particulièrement cohérent, à condition de dépasser la limite d’un seul verset. Dans cette première étape de notre parcours, nous avons vu six passages où une correspondance terme à terme est possible en ce qui concerne la tournure qui nous intéresse. Désormais nous allons nous pencher sur d’autres extraits, où la synopse n’est pas aussi immédiatement perceptible; en d’autres termes, il s’agira d’équivalences sémantiques, mais non syntaxiques entre les deux textes.

3. Équivalences sémantiques mais non syntaxiques Dans un premier temps, nous allons nous pencher sur deux passages analogues, où la présence de la tournure étudiée se fait à la faveur d’un renversement des relations entre sujet et COD. 7 35.10 καὶ συνηγμένοι διὰ κύριον ἀποστραφήσονται καὶ ἥξουσιν εἰς Σιων μετ’ εὐφροσύνης, καὶ εὐφροσύνη αἰώνιος ὑπὲρ κεφαλῆς αὐτῶν· ἐπὶ γὰρ κεφαλῆς αὐτῶν αἴνεσις καὶ ἀγαλλίαμα, καὶ εὐφροσύνη καταλήμψεται αὐτούς, ἀπέδρα ὀδύνη καὶ λύπη καὶ στεναγμός.

Et, rassemblés à cause du Seigneur, ils reviendront et arriveront à Sion dans la joie. Joie éternelle sur leur tête! Oui, sur leur tête louange et exultation, et la joie s’emparera d’eux, et souffrance, chagrin et gémissement ont fui.

Les libérés de Yahvé reviendront, ils entreront à Sion avec des cris de joie, une joie éternelle sera sur leur tête; l’allégresse et la joie les accompagneront ‫ׂשׂשון‬ ‫וׂשמחה יׂשיגו‬, l’affliction et les gémissements s’enfuiront.

Une tournure syntaxique fréquente en Ésaïe-LXX

129

Comme souvent dans ce que nous avons déjà vu, les deux textes sont globalement semblables, et les différences se font dans le détail. Ce qui nous intéressera particulièrement sera l’emploi du terme εὐφροσύνη, « joie », que l’on trouve à trois reprises dans le verset, contre deux occurrences seulement de ‫ׂשמחה‬. Il est dit tout d’abord qu’ils entreront dans Sion « dans la joie », littéralement « avec de la joie »; puis on lit « joie éternelle sur leur tête », avec une phrase nominale, sans verbe qui exprimerait un temps: l’absence de verbe corrobore l’emploi de l’adjectif αἰώνιος. Cela dit on a une synecdoque; la partie (la tête) vaut bien évidemment pour le tout (la personne), mais le texte, à ce moment-là, reste partiel, même si la synecdoque est répétée dans le « plus » de la LXX, ἐπὶ γὰρ κεφαλῆς αὐτῶν αἴνεσις. Il y a ensuite une répartition différente de la matière textuelle dans les deux versions. Dans le TM, l’« allégresse et la joie » peuvent être COD du verbe « atteindre »16, tandis que le premier substantif, dans la LXX, est rattaché à son « plus », alors que l’εὐφροσύνη devient sujet du verbe καταλήμψεται au futur, conformément au modèle de la formulation présentement étudiée. Le traducteur a donc changé la syntaxe de son modèle au profit d’une nouvelle occurrence de ce tour. On notera la valeur intensive du préverbe κατα-: à la différence de la synecdoque précédente, cette fois-ci la « joie » les prendra dans leur intégralité. Cette tonalité euphorique est corroborée par l’asyndète devant ἀπέδρα17, quand le TM présente bel et bien une conjonction de coordination (‫)ונסו‬. Les mêmes arguments peuvent être développés dans un passage visiblement très inspiré de celui-ci: 8 51.11 10. ἡ ἐρημοῦσα θάλασσαν, ὕδωρ ἀβύσσου πλῆθος; ἡ θεῖσα τὰ βάθη τῆς θαλάσσης ὁδὸν διαβάσεως ῥυομένοις

N’est-ce pas toi 10 qui dévastes la mer, la masse d’eau de l’abîme, qui des profondeurs de la mer fis une voie de passage pour ceux qui sont délivrés

10. N’est-ce pas toi qui desséchas la mer, les eaux du grand Abîme? qui fis des profondeurs de la mer un chemin pour le passage des rachetés?

16  Toutefois cela n’est pas nécessairement le cas; on peut en effet traduire le TM ainsi: « l’allégresse et la joie s’approcheront », avec les substantifs comme sujets, et le verbe au pluriel. 17  Non rendue dans la traduction proposée.

130

9 23.5

Philippe Le Moigne

11. καὶ λελυτρωμένοις; ὑπὸ γὰρ κυρίου ἀποστραφήσονται καὶ ἥξουσιν εἰς Σιων μετ’ εὐφροσύνης καὶ ἀγαλλιάματος αἰωνίου· ἐπὶ γὰρ τῆς κεφαλῆς αὐτῶν ἀγαλλίασις καὶ αἴνεσις, καὶ εὐφροσύνη καταλήμψεται αὐτούς, ἀπέδρα ὀδύνη καὶ λύπη καὶ στεναγμός.

11 et

rachetés? Car le Seigneur les fera revenir et ils arriveront à Sion dans une joie et une exultation éternelles: sur leur tête liesse et louange, et la joie se saisira d’eux. Au loin souffrance, chagrin et gémissement!

11. Les libérés de Yahvé reviendront ils entreront à Sion avec des cris de joie, une joie perpétuelle sera sur leur tête, l’allégresse et la joie les accompagneront, l’affliction et les gémissements s’enfuiront !

4. αἰσχύνθητι, Σιδών, εἶπεν ἡ θάλασσα· ἡ δὲ ἰσχὺς τῆς θαλάσσης εἶπεν Οὐκ ὤδινον οὐδὲ ἔτεκον οὐδὲ ἐξέθρεψα νεανίσκους οὐδὲ ὕψωσα παρθένους.

4 Sois

4. Sois honteuse, Sidon, car elle a dit, la forteresse de la mer: « Je n’ai pas été en travail, je n’ai pas enfanté, je n’ai pas fait grandir de garçons, je n’ai pas élevé de jeunes filles. » 5. Quand on l’apprendra en Égypte

dans la confusion, Sidon, a dit la mer. La force de la mer a dit: « Je n’ai pas été dans les douleurs, je n’ai pas enfanté, je n’ai pas élevé de jeunes gens, je n’ai pas fait grandir de jeunes filles. » 5 Lorsqu’on l’entendra 5. ὅταν δὲ ἀκουστὸν γένηται Αἰγύπτῳ, en Égypte, la douleur λήμψεται αὐτοὺς ὀδύνη les saisira à propos de περὶ Τύρου. Tyr.

‫כאׁשר ֵׁש ַמע למצרים‬ on frémira ‫ יחילו‬des nouvelles de Tyr.

Le yiqtol ‫ יחילו‬est traduit par Darby par un passé18, en dépit de l’absence de verbe conjugué au qatal dans la subordonnée; d’autres versions françaises de la Bible hébraïque traduisent l’ensemble du verset 5 au futur, comme c’est le cas en grec19. Le grec a choisi de rendre le hiphil du verbe ‫ חול‬en deux temps, avec le verbe λαμβάνω au futur et le substantif ὀδύνη, ce qui reproduit l’écho sonore présent dans le TM; en effet au verset 4 ce dernier présente la forme ‫לא־חלתי‬, avec le même

18  « Ils ont été dans l’angoisse à l’ouïe des nouvelles de Tyr ». 19  Bible de Jérusalem: « Quand la nouvelle parviendra en Égypte, on tremblera en apprenant le sort de Tyr »; Segond: « Quand les Égyptiens sauront la nouvelle, Ils trembleront en apprenant la chute de Tyr »; Bible en français courant: « À cette nouvelle, l’Égypte sera prise de douleurs, comme lorsqu’elle apprendra ce qui est arrivé à Tyr », etc.

Une tournure syntaxique fréquente en Ésaïe-LXX

131

verbe qu’en 23.5, tandis que le ὀδύνη du grec rappelle de près le Οὐκ ὤδινον du discours de la mer. Concernant l’aspect sémantique, le traducteur a pu comprendre ‫ חול‬à partir d’un des sens du qal: « souffrir, avoir des douleurs », si bien qu’il aurait une valeur factitive au hiphil; interprétation encouragée par le fait qu’au verset précédent, il est question d’enfants (non) mis au monde. 10 19.9 8. καὶ στενάξουσιν οἱ ἁλεεῖς, καὶ στενάξουσιν πάντες οἱ βάλλοντες ἄγκιστρον εἰς τὸν ποταμόν, καὶ οἱ βάλλοντες σαγήνας καὶ οἱ ἀμφιβολεῖς πενθήσουσιν. 9. καὶ αἰσχύνη λήμψεται τοὺς ἐργαζομένους τὸ λίνον τὸ σχιστὸν καὶ τοὺς ἐργαζομένους τὴν βύσσον, 10. καὶ ἔσονται οἱ διαζόμενοι αὐτὰ ἐν ὀδύνῃ, καὶ πάντες οἱ τὸν ζῦθον ποιοῦντες λυπηθήσονται καὶ τὰς ψυχὰς πονέσουσιν.

8 Les

marins pêcheurs gémiront, et tous ceux qui jettent un hameçon dans le fleuve gémiront, et ceux qui jettent des seines et les jeteurs de filets seront endeuillés. 9 Et la confusion saisira ceux qui travaillent le lin cardé, et ceux qui travaillent le lin fin, 10 et ceux qui les montent en chaîne seront dans la douleur, et tous ceux qui font la bière seront affligés et souffriront dans leur âme.

8. Les pêcheurs gémiront, tous ceux qui jettent l’hameçon dans le Nil seront endeuillés, et ceux qui tendent le filet à la surface des eaux dépériront. 9. Ceux qui travaillent le lin seront déçus ‫ובׁשו עבדי פׁשתים‬, cardeuses et tisseurs blêmiront, 10. ses tisserands20 seront prostrés, tous les artisans à gages auront l’âme affligée.

Ici encore, face au verbe unique, au weqatal, ‫ובׁשו‬, correspond une tournure avec un substantif abstrait, αἰσχύνη, proche sémantiquement du verbe hébreu, mais sujet de λήμψεται21. L’avantage, si l’on peut dire, de la tournure du grec, est d’éviter l’hyperbate, car le reste du verset 9, dans la LXX, représente le COD de la tournure étudiée. En outre, si l’on dépasse les frontières du seul verset 9, la configuration du grec crée un chiasme nominal avec καὶ ἔσονται οἱ διαζόμενοι αὐτὰ ἐν ὀδύνῃ au début du verset 10, avec l’écho en -ύνη des deux substantifs, alors que le TM dit: « Et les colonnes du pays sont brisées », avec le verbe « être » et

20  Note Osty: « texte légèrement corrigé et traduction hypothétique »: TM ‫ׁשתתיה‬. 21  Voir van der Vorm-Croughs, The Old Greek of Isaiah, 105n18.

132

Philippe Le Moigne

un participe passif: ‫והיו ׁשתתיה מדכאים‬.22 On remarque du reste que le verset 10 continue le futur dans la LXX, alors que le TM n’a que des tournures nominales, la fin du verset présentant la tournure, avec hapax, ‫כל־עׂשי ׂשכר אגמי־נפׁש‬. Ce futur est, selon nous, induit par la prégnance du tour étudié, c’est-à-dire qu’il dépend de l’emploi de λήμψεται. En outre le verset 10, dans la LXX, est relié au précédent plus étroitement que ce que l’on observe dans le TM, en raison de l’anaphorique αὐτά, concurremment au choix du verbe concret διαζόμενοι en face de la métaphore ‫‘ ׁשתות‬fondement, colonne’, si l’on accepte la correction de la BHS, reprise par Osty. 11 10.29 καὶ παρελεύσεται εἰς Μαγεδω καὶ ἐν Μαχμας θήσει τὰ σκεύη αὐτοῦ· καὶ παρελεύσεται φάραγγα καὶ ἥξει εἰς Αγγαι, φόβος λήμψεται Ραμα πόλιν Σαουλ.

Il passera vers Magedo, et à Machmas il posera ses affaires. Il passera le ravin et viendra vers Aggai, la crainte saisira Rama la ville de Saül.

Ils passent la passe, à Guéba ils bivouaquent, Rama tremble ‫חרדה‬, Guibéa de Saül prend la fuite.

Par la suite de l’omission de « Guibéa » dans la LXX, Rama devient, pour cette dernière version, la « ville de Saül », c’est-à-dire une localité reliée de près à l’histoire du peuple juif, ce qui la distingue nettement des autres toponymes mentionnés dans le passage. Peut-être est-ce la raison du choix, de la part du traducteur, du terme φόβος, qui est à double entente: la ‘crainte’ au sens de la ‘peur’, 23 mais aussi la ‘crainte’ révérencielle, au sens du sentiment religieux de respect envers la divinité,24 ce qui est approprié à propos de la ville du premier roi d’Israël. Conformément au modèle micro-syntaxique qui constitue notre sujet d’étude, le verbe ‘prendre’ est ici au futur, alors que l’on lit des qatal dans le TM. Notons cependant qu’au verset précédent, 10.28, le TM porte un yiqtol, ‫ ;יפקיד‬il n’est pas à exclure qu’au v. 29 la Vorlage du traducteur portait, pour les deux παρελεύσεται, des weqatal (cf. la double présence du καί, sans correspondant massorétique). Après cette deuxième étape de notre parcours, où nous avons constaté que le sens global était comparable dans les deux textes, hébreu et grec, nous allons 22  Alors qu’ailleurs dans la LXX d’Ésaïe le verbe hébreu au pual est rendu par un verbe de la même nature, par exemple en 53.5, καὶ μεμαλάκισται διὰ τὰς ἁμαρτίας ἡμῶν. 23  Entre de nombreux exemples, on peut citer 8.12: τὸν δὲ φόβον αὐτοῦ οὐ μὴ φοβηθῆτε οὐδὲ μὴ ταραχθῆτε. 24  Il suffira de citer le verset qui suit immédiatement celui que nous venons de mentionner dans la note qui précède: 8.13 κύριον αὐτὸν ἁγιάσατε, καὶ αὐτὸς ἔσται σου φόβος.

Une tournure syntaxique fréquente en Ésaïe-LXX

133

nous pencher sur des extraits plus difficiles, où la concordance de Hatch et Redpath n’a trouvé d’autre ressource que d’indiquer par † la difficulté d’une comparaison synoptique.

4. Tours originaux 12 30.28 28. καὶ τὸ πνεῦμα αὐτοῦ ὡς ὕδωρ ἐν φάραγγι σῦρον ἥξει ἕως τοῦ τραχήλου καὶ διαιρεθήσεται τοῦ ἔθνη ταράξαι ἐπὶ πλανήσει ματαίᾳ, καὶ διώξεται αὐτοὺς πλάνησις καὶ λήμψεται αὐτοὺς κατὰ πρόσωπον αὐτῶν. 29. μὴ διὰ παντὸς δεῖ ὑμᾶς εὐφραίνεσθαι καὶ εἰσπορεύεσθαι εἰς τὰ ἅγιά μου διὰ παντὸς ὡσεὶ ἑορτάζοντας καὶ ὡσεὶ εὐφραινομένους εἰσελθεῖν μετὰ αὐλοῦ εἰς τὸ ὄρος τοῦ κυρίου πρὸς τὸν θεὸν τοῦ Ισραηλ;

28 Et

son souffle, comme un torrent violent dans un ravin, arrivera jusqu’à la nuque, et il se divisera afin de bouleverser les nations pour leur égarement vain, et l’égarement les poursuivra et les saisira à la face. 29 Faut-il que tout le temps vous vous réjouissiez et que vous pénétriez dans mon sanctuaire, tout le temps, comme des gens qui célèbrent une fête, comme des gens qui se réjouissent d’aller au son de la flûte sur la montagne du Seigneur, auprès du Dieu d’Israël?

28. Son souffle est comme un torrent débordé qui atteint jusqu’au cou, pour cribler les nations au crible destructeur et mettre un frein d’égarement à la mâchoire des peuples

‫ורסן מתעה על לחיי‬ ‫עמים‬

29. Vous chanterez alors comme dans la nuit où se célèbre la Fête, avec la joie au cœur, comme lorsqu’on marche au son de la flûte pour aller à la montagne de Yahvé, vers le Rocher d’Israël.

Les deux états du verset 28 présentent des divergences remarquables. Si son début peut à la rigueur se prêter à une synopse, sa fin, dans la LXX, est très originale par rapport au substrat proto-massorétique. Alors que le TM parle successivement des ‹ nations ›, ‫גוים‬, et des ‹ peuples ›, ‫ עמים‬la LXX, en recourant à une syllepse, anaphorise ἔθνη par αὐτούς à trois reprises. Et tandis que le TM parle, dans un polyptote, du ‹ crible › (‫ להנפה‬/ ‫)בנפת ׁשוא‬, la LXX redouble le terme d’‹ égarement ›, πλάνησις, alors que ‫ מתעה‬ne figure qu’une fois en hébreu, et en finissant par en faire le sujet de λαμβάνω au futur. Il y a donc une progression assez claire dans le texte grec; d’abord il est dit que les nations seront bouleversées en raison

134

Philippe Le Moigne

de l’égarement (ἐπὶ πλανήσει), puis l’égarement les poursuivra et les prendra tout entiers, un peu à la manière du souffle divin, au début du verset. L’égarement n’est donc pas un simple état de fait, c’est une puissance agissante; la mention de la ‹ face ›, comme c’est le cas dans le TM avec la ‹ mâchoire ›, poursuit l’isotopie du visage introduite par ἕως τοῦ τραχήλου. Si l’on prête désormais attention au verset suivant, on remarque une différence de taille. En effet, le TM est positif, parlant de ‹ chant › et de ‹ fête ›, et s’achevant sur la mention d’une rencontre avec le Seigneur; la rupture de ton avec le verset 28 est complète. Par sa syntaxe, la LXX se distingue radicalement de ce brusque revirement euphorique. En effet, μή plus indicatif ne peut se comprendre que comme introduisant une proposition interrogative; un doute est donc émis, d’autant plus que, en principe, une question introduite par μή laisse attendre une réponse négative. Nous restons donc, en ce qui concerne la lecture de la LXX, dans une tonalité négative, qui se poursuit d’ailleurs au verset suivant, où, cette fois-ci, les deux versions sont proches; ne citons donc que la LXX: 30.30 καὶ ἀκουστὴν ποιήσει ὁ θεὸς τὴν δόξαν τῆς φωνῆς αὐτοῦ καὶ τὸν θυμὸν τοῦ βραχίονος αὐτοῦ δείξει μετὰ θυμοῦ καὶ ὀργῆς καὶ φλογὸς κατεσθιούσης· κεραυνώσει βιαίως καὶ ὡς ὕδωρ καὶ χάλαζα συγκαταφερομένη βίᾳ « et Dieu fera entendre la gloire de sa voix; il montrera l’emportement de son bras, avec emportement, colère et flamme dévorante, il sera foudre violente et comme l’eau et la grêle qui s’abat avec violence ». La démétaphorisation que l’on avait observée à la fin du verset 29 (τὸν θεὸν τοῦ Ισραηλ versus ‫)צור יׂשראל‬ permet une anadiplose au verset 30, qui contribue à renforcer les liens, dans la LXX, entre ces trois versets; ce lien est confirmée par la répétition, en 30, du ὡς ὕδωρ du verset 28, alors que le TM présente deux mots différents. 13 26.11 κύριε, ὑψηλός σου ὁ βραχίων, καὶ οὐκ ᾔδεισαν, γνόντες δὲ αἰσχυνθήσονται· ζῆλος λήμψεται λαὸν ἀπαίδευτον, καὶ νῦν πῦρ τοὺς ὑπεναντίους ἔδεται.

Seigneur, ton bras est élevé, et ils ne le savaient pas; et quand ils le connaîtront, ils seront dans la confusion. L’ardeur saisira un peuple malappris et le feu va dévorer les adversaires.

Yahvé, ta main levée, ils ne la voient pas; qu’ils voient, honteux, ton amour jaloux pour ton peuple ‫יחזו ויבׁשו‬ ‫ ;קנאת־עם‬que les dévore le feu qui attend tes adversaires!

Hors de tout contexte, ce verset peut paraître obscur. Si l’on se réfère à ce qui précède, nous sommes dans un passage qui oppose la ‹ voie des gens pieux › (ἡ ὁδὸς τῶν εὐσεβῶν, 26.7 bis) à la conduite de l’‹ impie › (ὁ ἀσεβής, 26.10 bis). Il semble bien que l’ensemble du verset 11 se poursuive dans la tonalité du verset précédent, c’est-à-dire qu’il s’agisse ici des ‹ impies › qui ne connaissent pas le chemin

Une tournure syntaxique fréquente en Ésaïe-LXX

135

du Seigneur. Dans ce contexte, la différence de construction qui s’observe entre les deux textes est instructive. La ‫ קנאת־עם‬la ‹ jalousie pour le peuple ›, qui est objet du verbe ‫ חזה‬dans le TM, est dissociée dans la LXX, au profit de la tournure qui forme le sujet de la présente étude, le verbe λαμβάνω au futur avec un sujet abstrait et un COD animé. De manière classique, la ‫קנאה‬, dans le TM, concerne l’ardeur exclusive qui scelle les relations entre Dieu et son peuple; cela est corroboré par son complément du nom, ‫ עם‬et non ‫גוי‬. La LXX réinterprète complètement le matériau textuel à sa disposition en introduisant, dans un premier temps, l’adjectif ἀπαίδευτον, qui n’a aucun correspondant formel dans le TM. D’après ses autres emplois septantiques25, il renvoie toujours à celui qui ne marche pas dans la voie de l’enseignement divin26; ne citons que Proverbes 14.15: καρδία ὀρθὴ ζητεῖ αἴσθησιν, στόμα δὲ ἀπαιδεύτων γνώσεται κακά27. Il devient évident que le ζῆλος de Dieu va ici toucher les non-juifs ou, à la rigueur, les mauvais juifs. Nous aboutissons ici à un aspect extrêmement intéressant du point de vue théologique, puisque le ζῆλος, qui est en principe réservé à l’action de l’Éternel envers son peuple28, va désormais se tourner vers les non-juifs. On peut s’interroger sur l’interférence entre les sentiments ordinaires de Dieu envers les juifs et cette action dirigée contre les nations. Ce qui est certain, c’est qu’elle est entraînée, du strict point de vue de la syntaxe, par l’occurrence du tour aimé du traducteur qui forme l’objet de ce parcours.

25 Dans la LXX on le lit—mis à part la péricope des Odes qui reprend notre passage d’Ésaïe—encore une fois en Sophonie (2.1, avec précisément l’emploi du mot ‹ nation ›, τὸ ἔθνος τὸ ἀπαίδευτον) mais il est surtout fréquent dans les écrits sapientiaux: 7 fois en Proverbes, une fois en Sagesse (17.1) et 7 fois dans le Siracide. 26  Cela est très clair quelques lignes plus loin dans notre livre: 26.16 κύριε, ἐν θλίψει ἐμνήσθην σου, ἐν θλίψει μικρᾷ ἡ παιδεία σου ἡμῖν « Seigneur, dans l’oppression je me suis souvenu de toi, dans un moment d’oppression ton instruction fut sur nous ». 27  Même si parfois une rémission semble envisageable: Siracide 51.23 ἐγγίσατε πρός με, ἀπαίδευτοι, καὶ αὐλίσθητε ἐν οἴκῳ παιδείας, avec la figura etymologica créée par l’occurrence de παιδεία. 28  Dans notre livre, cela est particulièrement clair en 63.15 b ποῦ ἐστιν ὁ ζῆλός σου καὶ ἡ ἰσχύς σου; ποῦ ἐστιν τὸ πλῆθος τοῦ ἐλέους σου καὶ τῶν οἰκτιρμῶν σου, ὅτι ἀνέσχου ἡμῶν; « Où est ton ardeur et où ta force? Où est l’abondance de ta tendresse et de tes compassions, que tu nous aies soutenus? »

136 14 et 15

Philippe Le Moigne

64.1,3 1. ἐὰν ἀνοίξῃς τὸν [63.19, οὐρανόν, τρόμος 64.2] λήμψεται ἀπὸ σοῦ ὄρη, καὶ τακήσονται,

2. ὡς κηρὸς ἀπὸ πυρὸς τήκεται. καὶ κατακαύσει πῦρ τοὺς ὑπεναντίους, καὶ φανερὸν ἔσται τὸ ὄνομα κυρίου ἐν τοῖς ὑπεναντίοις· ἀπὸ προσώπου σου ἔθνη ταραχθήσονται. 3. ὅταν ποιῇς τὰ ἔνδοξα, τρόμος λήμψεται ἀπὸ σοῦ ὄρη.

1. Si tu ouvres le ciel, un tremblement venu de toi saisira les montagnes, et elles fondront,

1. Ah ! Si tu déchirais les cieux, si tu descendais, devant toi les montagnes ruisselleraient

2. comme cire fond au feu. Et le feu embrasera les adversaires, et le nom du Seigneur se manifestera parmi les adversaires: devant ta face les nations seront bouleversées. 3. Lorsque tu accompliras tes actes de gloire, un tremblement venu de toi saisira les montagnes.

2. comme le feu allume du bois sec, comme le feu fait bouillonner l’eau ! Tu ferais connaître ton Nom à tes adversaires, devant toi les nations frémiraient, 3. te voyant accomplir de terribles actions que nous n’attendions pas; devant toi les montagnes ruisselleraient29. ‫ירדת מפניך הרים נזלו‬.

‫לוא־קרעת ׁשמים ירדת‬ ‫מפניך הרים נזלו‬,

Du strict point de vue la synopse à établir entre les deux textes30, et si l’on se concentre sur le verset 1, qui n’est que (partiellement) répété au verset 3, on constate l’absence de correspondance, dans la LXX, pour le verbe ‫‹ ירד‬ descendre › du TM, à quoi correspond matériellement le substantif τρόμος; pour le reste on peut supposer une équivalence entre ἀπὸ σοῦ et ‫מפניך‬. Il n’y a par ailleurs rien qui motive, en apparence, l’apparition de λήμψεται, si ce n’est, précisément, la prégnance, dans l’esprit du traducteur, du tour recourant à ce verbe avec un substantif abstrait. Le futur induit par l’occurrence de cette tournure se poursuit dans tout le passage, alors que nous avons des qatal dans le TM au verset 3; au début de celui-ci, la tournure ‫ביׂשותך‬, qui est neutre du point de vue du temps, a été rendue par l’éventuel dans la LXX, avec une valeur future manifeste. Au verset 2, la comparaison prend de moindres dimensions dans la LXX, au profit, précisé29  Ces cinq derniers mots de la traduction, qu’il juge être une inutile répétition, sont rejetés en note par Osty. 30  Pour une brève étude philologique de ce passage, voir van der Vorm-Croughs, The Old Greek of Isaiah, 405, qui le rapproche de 33.14, notre exemple 6.

Une tournure syntaxique fréquente en Ésaïe-LXX

137

ment, de l’action future du Seigneur, avec la double mention des « adversaires », là où le TM parle, la première fois, de « l’eau » dans la comparative. Ainsi, notre exemple final, où rien ne correspond dans le TM au verbe « prendre » de la LXX, indique bien les conséquences syntaxiques assez larges qui sont consécutives à l’emploi de la formulation étudiée.

5. Conclusion La présente étude a porté, de manière exhaustive pour l’Ésaïe grec, sur le tour sujet abstrait + λαμβάνω au futur + COD. Il existe d’après notre relevé quinze occurrences de cette formulation, si l’on y inclut les préverbés de λαμβάνω. On aura remarqué que le traducteur y recourt volontiers quand il est question de vent ou d’autres phénomènes climatiques, plus ou moins violents, et de leurs extensions métaphoriques; cela confère à la locution une incontestable unité sémantique et stylistique, qui s’ajoute au « refrain » syntaxique. Après avoir parcouru des passages où une synopse entre TM et LXX était relativement facile à établir, mais où nous avons constaté le recours systématique au verbe λαμβάνω face à des substrats hébreux fort divergents, ce que nous avons appelé les « matrices » du tour, nous nous sommes penché sur des péricopes où la LXX divergeait plus nettement de son original. Ce qu’il y a de remarquable, c’est que dans ce deuxième type d’exemple, par exemple 23.5 ou 19.9, il peut s’agir de versets qui se situent plus tôt dans le cadre d’une lecture linéaire de ce livre; autrement dit, les sept passages étudiés dans le premier moment de ce parcours, qui sont traduits de manière à pouvoir établir une synopse, et notamment une correspondance entre sujet, verbe et COD, sont situés après des extraits de nos deuxième et troisième parties. Le fait que l’on trouve cette phraséologie dénuée de substrat proto-massorétique direct avant les matrices prouve—s’il était encore besoin de le démontrer—que notre traducteur agit de manière concertée, c’est-à-dire qu’il a une vision globale de son travail, qu’il connaît parfaitement le livre d’Ésaïe en hébreu et qu’il ne traduit pas de manière myope, mot pour mot. Il ne s’agit pas là d’une tournure qu’il connaît, qu’il a découverte dans un premier temps dans le cadre d’une traduction calque puis étendue ensuite de manière plus libre. Les matrices sont postérieures ou bien, autre manière de le dire, la mère est plus jeune que la fille.

Miika Tucker The Translations of Infinitives in Septuagint Jeremiah 1. Introduction A notable aspect of research to which Ilmari Soisalon-Soininen has greatly contributed is the translation of infinitives in the Septuagint. His momentous study, Die Infinitive in der Septuaginta,1 though published over 50 years ago, continues to be the definitive work on the subject. In his study, Soisalon-Soininen classifies the different ways in which Hebrew infinitives are translated into Greek in the Septuagint. He discusses each type of translation with respect to its implications for the character of the translation, and provides an ample amount of examples. This article draws from Soisalon-Soininen’s research for the purpose of evaluating translations of infinitives in Septuagint Jeremiah in light of theories of bisectioning the book. Most notably, Henry St. John Thackeray and Emanuel Tov have posited that the translation exhibits the work of more than one person,2 whether that be multiple translators or the presence of a revision. Both authors make reference to the discrepancy between the two halves of the book with regard to their use of the infinitive. Thackeray observes that in the first half, chapters 1–28 (Jer a’), the anarthrous infinitive is rare and the infinitive with the genitive article τοῦ is usual. On the other hand, in the second half, chapters 29–52 (Jer b’), the anarthrous infinitive is common, but the genitive articular infinitive is also used.3 Tov restricts his comments to the renderings of the Hebrew construction ‫ל‬+inf. ְ construct. According to his calculations, these are rendered 56 times by an articular infinitive and 16 times by an anarthrous infinitive in Jer a’. In Jer b’, the corresponding figures are 35 times an articular infinitive and 59 times an anarthrous infinitive. He concludes that Jer a’ prefers the article as a formal equivalent for the preposition ‫ל‬,ְ while Jer b’ has often omitted the article because of its “un-Greek employment.”4 Soisalon-Soininen, writing after Thackeray and before Tov, notes the difference between the two halves of Septuagint Jeremiah with regard to their use of the genitive articular infinitives, and mentions that a break point does occur between chapters 28 and 29. For the purposes of his own study, however, he p ­ roceeds to 1  I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965). 2  H.St.J. Thackeray, “The Greek Translators of Jeremiah”, JTS 4 (1903) 245–66; E. Tov, The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of the LXX of Jeremiah 29–52 and Baruch 1:1–3:8 (HSM 8; Missoula, MT: Scholars Press, 1976). 3 Thackeray, “The Greek Translators of Jeremiah”, 251. 4 Tov, The Septuagint Translation of Jeremiah and Baruch, 97.

140

Miika Tucker

address Septuagint Jeremiah as a single unit in comparison to other LXX books. The material, he observes, does point toward the work of two translators, but as some of Thackeray’s other multiple translator theories had been disproven at the time, Soisalon-Soininen suggests a cautious approach to Septuagint Jeremiah as well.5 Other explanations of the evidence have also been proposed, most notably by Hermann-Josef Stipp and by Albert Pietersma. They both argue that the differences are explainable within the framework of one translator, though their proposals slightly differ from each other. Stipp’s6 main design is to argue against the notion of a revision in Jer b’ rather than to advocate for a single translator, which he only briefly proposes in his article. He presents four criticisms of Tov’s proposal. First, he draws attention to the mixed nature of Septuagint Jeremiah 29, pointing out that equivalents that are characteristic to both Jer a’ and Jer b’ are located in this chapter. Second, Stipp sees the lack of a well-defined motive for revision as a challenge, since a goal or motives are essential to distinguishing a revised text. Third, the distribution of the revised equivalents proposed by Tov do not always match each other. One revised equivalent, ῥομφαία, occurs in a block of text (Septuagint Jeremiah 44–51) which simultaneously employs the equivalent ἐντέλλομαι, which elsewhere in Jer b’ is revised to συντάσσω. Fourth, Stipp claims that some of Tov’s distinct equivalents to which he appeals for identifying a common base text between Jer a’ and Jer b’ are not suitable to the task, thus diminishing the list of elements that connect the two halves to each other. Pietersma’s most recent articles on Septuagint Jeremiah articulate his position in detail. He proposes that the translator recognized a certain duology in his Hebrew Vorlage. This duology divides the LXX Vorlage into chapters 1–32 and chapters 33–52. The first part is a conceptual overview of history, beginning in Israel’s exodus from Egypt and ending in a cosmic judgement of the nations, while the second part portrays a shorter period of history, from Jehoiakim’s reign to the flight of the Judean refugees to Egypt.7 This division into two types of material, according to Pietersma, is the motive behind the different equivalents used in the two halves. The translator has practiced a degree of “textual accommodation and exegesis,” which is reflected in the different equivalents that he chooses.8 5 Soisalon-Soininen, Die Infinitive, 174. 6  H.-J. Stipp, “Offene Fragen zur Übersetzungskritik des antiken griechischen Jeremiabuches”, JNSL 17 (1991) 117–28. 7  A. Pietersma, “An Excursus on Bisectioning Ieremias”, in A. Pietersma/B.G. Wright (ed.), Electronic Edition of The New English Translation of the Septuagint (Oxford: Oxford University Press, 2009) 7–9. 8  A. Pietersma, “Of Translation and Revision: From Greek Isaiah to Greek Jeremiah”, in M. van der Meer et al. (ed.), Isaiah in Context: Studies in Honour of Arie van der Kooij on the Occasion of his Sixty-Fifth Birthday (VTSup 138; Leiden: Brill 2010) 359–88, on p. 386.

The Translations of Infinitives in Septuagint Jeremiah

141

In the following, I present my own research on the infinitives in Septuagint Jeremiah and conclude with a discussion on how this material reflects on theories of bisectioning the book. I limit my focus to specific forms of the infinitive construct, particularly to the prepositional expressions ‫ל‬+inf., ְ ‫ב‬+inf., ְ and ‫כ‬+inf., ְ as the renderings of these forms display a degree of difference between Jer a’ and Jer b’. The renderings of infinitives absolute are not examined in this article because their character and meaning are so far removed from Greek infinitives, as Soisalon-Soininen has pointed out in his monograph.9 I will also not include translations of the expression ‫לאמֹר‬,ֵ which has assumed a fairly standard translation equivalent throughout the Septuagint, including Septuagint Jeremiah, in the form of the participle λέγων. Instances in which the LXX lacks a quantitative equivalent of a Hebrew infinitive are also not considered, since most of the MT plusses are understood to be later additions to the text.10 This leaves 239 occurrences11 to consider. Due to a lack of space, only a general overview is presented here.

2. Analysis By far the most common manifestation of the infinitive in Jeremiah is that with the preposition ‫ל‬,ְ which accounts for over 70% (357 out of 501) of all infinitives construct in the book. The majority of these are rendered by the genitive articular infinitive (τοῦ+inf.), but a significant number are also rendered by an infinitive without an article. Other renderings, such as finite verbs, prepositional phrases and participles, occur in small numbers. In the following tables, the data is presented in two rows in order to distinguish between the renderings in Jer a’ and in Jer b’. Table 1 displays the renderings of ‫ל‬+inf. ְ according to the limits defined in the introduction:

9 Soisalon-Soininen, Die Infinitive, 16. 10  For discussions on the nature of the textual plusses and minuses in Jer MT and Jer LXX, cf. J.G. Janzen, Studies in the Text of Jeremiah (Cambridge, MA: Harvard University Press 1973); Y.-J. Min, “The Minuses and Pluses of the LXX Translation of Jeremiah as Compared with the Massoretic Text: Their Classification and Possible Origins” (PhD diss., Hebrew University of Jerusalem, 1977); E. Tov, “Some Aspects of the Textual and Literary History of the Book of Jeremiah”, in P.-M. Bogaert (ed.), Le Livre de Jérémie. Le prophète et son milieu, les oracles et leur transmission (BETL 54; Leuven: Leuven University Press, 1981) 145–67; and E. Tov, “The Literary History of the Book of Jeremiah in the Light of Its Textual History”, in J.H. Tigay (ed.), Empirical Models for Biblical Criticism (Philadelphia: University of Pennsylvania Press, 1985) 211–37. 11 According to my calculations, there are 215 cases of ‫ל‬+inf., ְ 14 cases of ‫כ‬+inf., ְ and 10 cases of ‫ב‬+inf. ְ

142

Miika Tucker

Table 1: Translations of ‫ ְל‬+inf. in Septuagint Jeremiah12

Jer a’ Jer b’

0+inf.

τοῦ+inf.

finite verb

prep.+noun

other

total

18 52

63 39

3 5

11 7

6 11

101 114

This data brings Thackeray, Soisalon-Soininen and Tov’s observations to the fore. When the two halves of the book are compared, one can clearly see that the anarthrous infinitive, represented by 0+inf. in the table, is infrequent in Jer a’ relative to Jer b’. The frequency of τοῦ+inf., the genitive articular infinitive, also fluctuates between the two halves, though to a lesser extent. In the following, I shall examine each type of rendering in turn with a view towards this difference.

2.1 The Anarthrous Infinitive An anarthrous infinitive is the equivalent of ‫ל‬+inf. ְ 70 times in Septuagint Jeremiah. Twenty of these function as the direct object of finite verbs, usually translating the Hebrew expressions ‫יכל‬+inf. and ‫מאן‬+inf. According to Soisalon-Soininen, the anarthrous infinitive is the usual equivalent in such cases in the Septuagint.13 Examples include the following: 6:10 ‫יוּכלוּ ְל ַה ְק ִשׁיב‬ ְ ‫— ִהנֵּ ה ֲע ֵר ָלה אָזְ נָ ם וְ לֹא‬ἰδοὺ ἀπερίτμητα τὰ ὦτα αὐτῶν, καὶ οὐ δύνανται ἀκούειν 43(36):26 ‫ת־שׁ ֶל ְמיָ הוּ ֶבּן־‬ ֶ ‫יאל וְ ֶא‬ ֵ ‫ן־עזְ ִר‬ ַ ‫ת־שׂ ָריָ הוּ ֶב‬ ְ ‫ן־ה ֶמּ ֶלְך וְ ֶא‬ ַ ‫וַ יְ ַצוֶּ ה ַה ֶמּ ֶלְך ֶאת־יְ ַר ְח ְמ ֵאל ֶבּ‬ ‫ת־בּרוְּך ַהסּ ֵֹפר וְ ֵאת יִ ְר ְמיָ הוּ ַהנָּ ִביא‬ ָ ‫— ַע ְב ְדּ ֵאל ָל ַק ַחת ֶא‬καὶ ἐνετείλατο ὁ βασιλεὺς τῷ Ιερεμεηλ υἱῷ τοῦ βασιλέως καὶ τῷ Σαραια υἱῷ Εσριηλ συλλαβεῖν τὸν Βαρουχ καὶ τὸν Ιερεμιαν

Such equivalents are dispersed quite evenly between the two halves of the book. Nine occur in Jer a’ and eleven in Jer b’.14 Most of these anarthrous infinitives, that is 40 cases, serve to express the purpose of verbs of action that mainly denote movement, sending, giving or doing something. Generally such infinitives occur in the Septuagint as genitive articular infinitives, but Soisalon-Soininen points out that the anarthrous infinitive is a 12  These totals do not include the 114 cases of ‫ לאמר‬and the 28 cases that appear in MT plusses. 13 Soisalon-Soininen, Die Infinitive, 38–39. 14  Jer a’ 5:3; 6:10; 11:10, 11; 12:16; 13:10; 14:9, 10 and 19:11; Jer b’ 32:14, 15; 33:2, 8; 38:15; 43:3, 5, 26; 45:21; 49:22 and 51:22.

The Translations of Infinitives in Septuagint Jeremiah

143

verbatim equivalent of ‫ל‬+inf. ְ in such cases as well.15 The distribution of these equivalents is very uneven between the two halves of the book. Only eight occur in Jer a’ and 32 in Jer b’.16 1:10 ‫וּל ַה ֲא ִביד וְ ַל ֲהרֹוס‬ ְ ‫ל־ה ַמּ ְמ ָלכֹות ִלנְ תֹושׁ וְ ִלנְ תֹוץ‬ ַ ‫ל־הגֹּויִם וְ ַע‬ ַ ‫ְר ֵאה ִה ְפ ַק ְד ִתּיָך ַהיֹּום ַהזֶּ ה ַע‬ ‫טֹוע‬ ַ ְ‫— ִל ְבנֹות וְ ִלנ‬ἰδοὺ κατέστακά σε σήμερον ἐπὶ ἔθνη καὶ ἐπὶ βασιλείας ἐκριζοῦν καὶ κατασκάπτειν καὶ ἀπολλύειν καὶ ἀνοικοδομεῖν καὶ καταφυτεύειν 33(26):2 ‫הוּדה ַה ָבּ ִאים ְל ִה ְשׁ ַתּ ֲחֹות ֵבּית־יְ הוָ ה‬ ָ ְ‫ל־ע ֵרי י‬ ָ ‫ל־כּ‬ ָ ‫—וְ ִד ַבּ ְר ָתּ ַע‬καὶ χρηματιεῖς ἅπασι τοῖς Ἰουδαίοις τοῖς ἐρχομένοις προσκυνεῖν ἐν οἴκῳ κυρίου 43(36):21 ‫ת־ה ְמּגִ ָלּה‬ ַ ‫הוּדי ָל ַק ַחת ֶא‬ ִ ְ‫—וַ יִּ ְשׁ ַלח ַה ֶמּ ֶלְך ֶאת־י‬καὶ ἀπέστειλεν ὁ βασιλεὺς τὸν Ιουδιν λαβεῖν τὸ χαρτίον

There are three contexts in which most of the infinitives in this group occur. The first is as renderings of ‫ל‬+inf. that complement the verb ‫בוא‬. These account for 10 cases (e.g. 33[26]:21). The verb ‫ לגור‬complements ‫ בוא‬in the context of taking refuge in Egypt from the Babylonian army a total of 8 times, all in chapters 49–51(42–45), and in each case it is rendered without an article, save for two occurrences (51[44]:12 and 14). The second context is as a complement to the verb ‫ ׁשלח‬in 33(26):12, 15, 43(36):21, and 47(40):14, and the third as part of the expression ‫“ אׁשר עשו להכעסני‬that which they did to provoke me to anger” in 39(32):32, 51(44):3 and 8. The remaining cases of the anarthrous infinitive as an equivalent of ‫ל‬+inf. ְ are mostly epexegetical in nature, that is, infinitives used to specify the meaning of exhortations of obedience or grounds of accusations. Only one of these occurs in Jer a’, and nine occur in Jer b’.17 17:24 ‫ם־שׁמ ַֹע ִתּ ְשׁ ְמעוּן ֵא ַלי נְ ֻאם־יְ הוָ ה ְל ִב ְל ִתּי ָה ִביא ַמ ָשּׂא ְבּ ַשׁ ֲע ֵרי ָה ִעיר ַהזֹּאת ְבּיֹום‬ ָ ‫וְ ָהיָ ה ִא‬ ‫וּל ַק ֵדּשׁ ֶאת־יֹום ַה ַשּׁ ָבּת‬ ְ ‫— ַה ַשּׁ ָבּת‬καὶ ἔσται ἐὰν ἀκοῇ ἀκούσητέ μου, λέγει κύριος, τοῦ μὴ εἰσφέρειν βαστάγματα διὰ τῶν πυλῶν τῆς πόλεως ταύτης ἐν τῇ ἡμέρᾳ τῶν σαββάτων καὶ ἁγιάζειν τὴν ἡμέραν τῶν σαββάτων 51(44):8 ‫יכם‬ ֶ ‫) ְל ַה ְכ ִע ֵסנִ י ְבּ ַמ ֲע ֵשׂי יְ ֵד‬8( … ‫דֹולה ֶאל־נַ ְפשׁ ֵֹת ֶכם‬ ָ ְ‫אַתּם ע ִֹשׂים ָר ָעה ג‬ ֶ ‫) ָל ָמה‬7( (7) ‫אֹלהים ֲא ֵח ִרים ְבּ ֶא ֶרץ ִמ ְצ ַריִם‬ ִ ‫ — ְל ַק ֵטּר ֵל‬Ἵνα τί ὑμεῖς ποιεῖτε κακὰ μεγάλα ἐπὶ ψυχαῖς ὑμῶν … (8)παραπικρᾶναί με ἐν τοῖς ἔργοις τῶν χειρῶν ὑμῶν θυμιᾶν θεοῖς ἑτέροις ἐν γῇ Αἰγύπτῳ

15 Soisalon-Soininen, Die Infinitive, 49. 16  Jer a’ 1:10 (5x); 11:13; 15:20 and 16:8; Jer b’ 33:2, 12, 15; 34:5; 35:14; 37:21; 38:28 (2x), 32; 39:19, 24, 31, 32, 33, 39; 43:21; 45:26; 47:10, 14; 48:12; 49:15, 17, 22; 50:2, 5; 51:3 (2x), 5, 7, 8 (2x) and 28. 17  Jer a’ 17:24; Jer b’ 29:13; 31:16; 33:5; 47:4; 50:4; 51:8, 17 and 25(2x).

144

Miika Tucker

The infinitives in both examples serve to further explain what has been said before. Among the anarthrous infinitives that form equivalents of the Hebrew ‫ל‬+inf., ְ the disparity between Jer a’ and Jer b’ is mainly evident with regard to purpose expressing infinitives and epexegetical and other infinitives (cf. Table 2 below). Purpose expressing ‫ל‬+inf. ְ are rendered only eight times by an anarthrous infinitive in Jer a’ in contrast to 32 times in Jer b’. Similarly, epexegetical and other infinitives are rendered by an anarthrous infinitive once in Jer a’ and nine times in Jer b’. The more common rendering of similar expressions in Jer a’ in both of these categories is the genitive articular infinitive, which is used 41 times to render purpose expressing ‫ל‬+inf. ְ and 13 times to render epexegetical infinitives. The distribution pattern of these two equivalents is also uneven with regard to the object infinitives. Though the anarthrous infinitive is used to render a fairly equal amount of object infinitives in Jer a’ and Jer b’, the genitive articular infinitive is employed more often in Jer a’ than in b, that is, nine times in Jer a’ and only twice in Jer b’. These figures are depicted in Table 2 below.

2.2 The Genitive Articular Infinitive The genitive articular infinitive, τοῦ+inf, is the equivalent of the Hebrew ‫ל‬+inf. ְ 102 times in Jeremiah. Of these, twelve are direct objects of other verbs. Generally these consist of singular cases within Septuagint Jeremiah: οὐκ θέλω+inf. translating ‫מאן‬+inf., δύναμαι+inf. translating ‫יכל‬+inf., and the bitransitive ἐντέλλομαι+inf. translating ‫צוה‬+inf., equivalences which all occur, in more numbers, as anarthrous infinitives. 8:5 ‫רוּשׁ ַלםִ ְמ ֻשׁ ָבה נִ ַצּ ַחת ֶה ֱחזִ יקוּ ַבּ ַתּ ְר ִמית ֵמ ֲאנוּ ָלשׁוּב‬ ָ ְ‫שֹׁוב ָבה ָה ָעם ַהזֶּ ה י‬ ְ ‫דּוּע‬ ַ ‫— ַמ‬διὰ τί ἀπέστρεψεν ὁ λαός μου ἀποστροφὴν ἀναιδῆ καὶ κατεκρατήθησαν ἐν τῇ προαιρέσει αὐτῶν καὶ οὐκ ἠθέλησαν τοῦ ἐπιστρέψαι 18:6 ‫א־אוּכל ַל ֲעשֹׂות ָל ֶכם ֵבּית יִ ְשׂ ָר ֵאל נְ ֻאם־יְ הוָ ה‬ ַ ֹ ‫יֹּוצר ַהזֶּ ה ל‬ ֵ ‫— ֲה ַכ‬Εἰ καθὼς ὁ κεραμεὺς οὗτος οὐ δυνήσομαι τοῦ ποιῆσαι ὑμᾶς, οἶκος Ισραηλ

As mentioned above, there is a slight imbalance in the distribution of these renderings between Jer a’ and Jer b’. In Jer a’ there are ten cases, and in Jer b’ only two.18 Most of the genitive articular infinitives express the purpose of verbs of action, just as their counterparts among the anarthrous infinitives. Again, the actions denoted are generally movement, sending, giving or doing something. There are 69 of these in Jer.

18  Jer a’ 8:5; 12:16; 13:6, 18:4, 6, 8, 10; 23:27; 28:30 and 63. Jer b’ 33:3 and 51:14.

The Translations of Infinitives in Septuagint Jeremiah

145

25:6 ‫וּל ִה ְשׁ ַתּ ֲחֹות ָל ֶהם‬ ְ ‫ֹלהים ֲא ֵח ִרים ְל ָע ְב ָדם‬ ִ ‫אַח ֵרי ֱא‬ ֲ ‫אַל־תּ ְלכוּ‬ ֵ ְ‫—ו‬μὴ πορεύεσθε ὀπίσω θεῶν ἀλλοτρίων τοῦ δουλεύειν αὐτοῖς καὶ τοῦ προσκυνεῖν αὐτοῖς 19:14 ‫—וַ יָּבֹא יִ ְר ְמיָ הוּ ֵמ ַהתּ ֶֹפת ֲא ֶשׁר ְשׁ ָלחֹו יְ הוָ ה ָשׁם ְל ִהנָּ ֵבא‬καὶ ἦλθεν Ιερεμίας ἀπὸ τῆς Διαπτώσεως, οὗ ἀπέστειλεν αὐτὸν κύριος ἐκεῖ τοῦ προφητεῦσαι 38(31):28 ‫וּל ָה ֵר ַע ֵכּן ֶא ְשׁקֹד‬ ְ ‫וּל ַה ֲא ִביד‬ ְ ‫יהם ִלנְ תֹושׁ וְ ִלנְ תֹוץ וְ ַל ֲהר ֹס‬ ֶ ‫וְ ָהיָ ה ַכּ ֲא ֶשׁר ָשׁ ַק ְד ִתּי ֲע ֵל‬ ‫טֹוע נְ ֻאם־יְ הוָ ה‬ ַ ְ‫יהם ִל ְבנֹות וְ ִלנ‬ ֶ ‫— ֲע ֵל‬καὶ ἔσται ὥσπερ ἐγρηγόρουν ἐπ᾿ αὐτοὺς καθαιρεῖν καὶ κακοῦν, οὕτως γρηγορήσω ἐπ᾿ αὐτοὺς τοῦ οἰκοδομεῖν καὶ καταφυτεύειν, φησὶ κύριος19

Forty-three such renderings are found in Jer a’ and 26 in Jer b’.20 This distribution is also uneven, as was the case for the anarthrous infinitives, but with more cases in Jer a’ than in Jer b’. The remaining 21 cases are not uniform. They include epexegetical infinitives, but also infinitives that qualify other words besides verbs, such as nouns, adjectives and pronouns, and infinitives that function as object clauses in indirect speech. 41(34):15 ‫ת־היָּ ָשׁר ְבּ ֵעינַ י ִל ְקרֹא ְדרֹור ִאישׁ ְל ֵר ֵעהוּ‬ ַ ‫אַתּם ַהיֹּום וַ ַתּ ֲעשׂוּ ֶא‬ ֶ ‫—וַ ָתּ ֻשׁבוּ‬καὶ ἐπέστρεψαν σήμερον ποιῆσαι τὸ εὐθὲς πρὸ ὀφθαλμῶν μου τοῦ καλέσαι ἄφεσιν ἕκαστον τοῦ πλησίον αὐτοῦ 43(36):8 ‫ר־צוָּ הוּ יִ ְר ְמיָ הוּ ַהנָּ ִביא ִל ְקרֹא ַב ֵסּ ֶפר ִדּ ְב ֵרי יְ הוָ ה ֵבּית‬ ִ ‫וַ יַּ ַעשׂ ָבּרוְּך ֶבּן־נֵ ִריָּ ה ְכּכֹל ֲא ֶשׁ‬ ‫—יְ הֹוָ ה‬καὶ ἐποίησε Βαρουχ κατὰ πάντα, ἃ ἐνετείλατο αὐτῷ Ιερεμίας τοῦ ἀναγνῶναι ἐν τῷ βιβλίῳ λόγους κυρίου ἐν οἴκῳ κυρίου

These are quite evenly distributed between the two halves of Septuagint ­Jeremiah. Ten of them are in Jer a’21 and eleven in Jer b’.22 Like the anarthrous infinitive, the genitive articular infinitive is distributed between the two halves of Septuagint Jeremiah in an uneven manner. When rendering both object infinitives and infinitives that express purpose, the articular infinitive predominantly occurs in Jer a’. In the case of purpose expressing infinitives, the opposite occurs with the anarthrous infinitives, which are predominantly used in Jer b’. Another difference in relation to the anarthrous infinitives regards the object infinitives. The anarthrous infinitive is used evenly in both Jer a’ and Jer b’ to render ‫ל‬+inf. ְ as a direct object, but τοῦ+inf. is used more often in Jer a’ than in b’. 19  This verse also contains two cases of the anarthrous infinitive as renderings of ‫ל‬+inf. ְ 20  Jer a’ 1:8, 12, 19; 2:7, 18(2x), 33, 36; 4:7; 7:18, 30, 31; 9:21; 10:22; 11:10, 17, 19; 13:10 (2x), 11; 15:20; 16:5, 8 (2x); 17:10; 18:15, 16, 20 (2x); 19:5, 14; 20:18; 22:17 (2x); 24:7; 25:6 (2x); 26:10, 13; 27:28; 28:11, 29 and 31. Jer b’ 29:4; 32:4; 33:24; 36:11, 26; 38:28 (2x); 39:35, 41; 40:2, 5 (2x); 41:8; 42:9, 15; 44:7, 12 (2x); 48:5, 17; 49:11 (2x); 50:3 (2x); 51:11 and 14. 21  Jer LXX 4:22 1°; 11:5; 17:27; 18:7(2x), 9(2x); 19:12; 26:13 and 28:62. 22  Jer LXX 33:4; 35:6; 36:10; 39:35; 41:9–10, 15, 17; 42:13; 43:8 and 47:5.

146

Miika Tucker

2.3 Other Renderings of ‫ ְל‬+ Infinitive Septuagint Jeremiah attests a finite verb as the translation of ‫ל‬+inf. in eight cases.23 Six of these renderings function as the main verb of a sentence, e.g.: 2:13 ‫—א ִֹתי ָעזְ בוּ ְמקֹור ַמיִם ַחיִּ ים ַל ְחצ ֹב ָל ֶהם בֹּארֹות‬ἐμὲ ἐγκατέλιπον, πηγὴν ὕδατος ζωῆς, καὶ ὤρυξαν ἑαυτοῖς λάκκους 29(47):4 ‫וּל ִצידֹון כֹּל ָשׂ ִריד עֹזֵ ר‬ ְ ‫— ְל ַה ְכ ִרית ְלצֹר‬καὶ ἀφανιῶ τὴν Τύρον καὶ τὴν Σιδῶνα καὶ πάντας τοὺς καταλοίπους τῆς βοηθείας αὐτῶν 51(44):19 ‫וּל ַה ֵסְּך ָלהּ נְ ָס ִכים‬ ְ ‫י־אנַ ְחנוּ ְמ ַק ְטּ ִרים ִל ְמ ֶל ֶכת ַה ָשּׁ ַמיִם‬ ֲ ‫—וְ ִכ‬καὶ ὅτι ἡμεῖς ἐθυμιῶμεν τῇ βασιλίσσῃ τοῦ οὐρανοῦ καὶ ἐσπείσαμεν αὐτῇ σπονδάς

In the dialogue in 1:12, the verb in the phrase ‫יט ְב ָתּ ִל ְראֹות‬ ַ ‫ ֵה‬is rendered as an adverb, while the inf. is transformed into the main verb: καλῶς ἑόρακας. A similar Hebrew construction is found in 16:12 ‫יכם‬ ֶ ‫בֹות‬ ֵ ‫אַתּם ֲה ֵרע ֶֹתם ַל ֲעשֹׂות ֵמ ֲא‬ ֶ ְ‫ו‬, but here Septuagint Jeremiah represents the two verbs with only one equivalent καὶ ὑμεῖς ἐπονηρεύσασθε ὑπὲρ τοὺς πατέρας ὑμῶν. It is uncertain whether the inf. was represented in the translator’s Vorlage, or whether this may be attributed to the character of his translation.24 The Hebrew preposition ‫ ְל‬is generally rendered in the Septuagint by the Greek preposition εἰς. A number of the ‫ל‬+inf. ְ forms in Jer are also translated by a prepositional phrase, especially with the preposition εἰς. Within this group, the preposition is usually followed by a noun, but in two cases it is followed by an adjective (4:11[2x]), once by an adverb (48[41]:10), and once by a dative articular infinitive (11:7). Most of these cases translate ‫ל‬+inf. when it functions as an adverbal expressing purpose. The final aspect of the Hebrew is retained in six cases, four of which occur in the same verse: Jer 15:3 ‫ת־ה ְכּ ָל ִבים ִל ְסחֹב‬ ַ ‫ת־ה ֶח ֶרב ַל ֲהר ֹג וְ ֶא‬ ַ ‫אַר ַבּע ִמ ְשׁ ָפּחֹות נְ ֻאם־יְ הוָ ה ֶא‬ ְ ‫יהם‬ ֶ ‫וּפ ַק ְד ִתּי ֲע ֵל‬ ָ ‫וּל ַה ְשׁ ִחית‬ ְ ‫אָרץ ֶל ֱאכֹל‬ ֶ ‫ת־בּ ֱה ַמת ָה‬ ֶ ‫—וְ ֶאת־עֹוף ַה ָשּׁ ַמיִם וְ ֶא‬καὶ ἐκδικήσω ἐπ᾿ αὐτοὺς τέσσαρα εἴδη, λέγει κύριος, τὴν μάχαιραν εἰς σφαγὴν καὶ τοὺς κύνας εἰς διασπασμὸν καὶ τὰ θηρία τῆς γῆς καὶ τὰ πετεινὰ τοῦ οὐρανοῦ εἰς βρῶσιν καὶ εἰς διαφθοράν.

In each of these instances the preposition εἰς indicates the purpose of the action: “for slaughter,” “for tearing apart,” “for eating” and “for decay.”

23  Jer a’ 1:12; 2:13 and 7:22; Jer b’ 29:4; 38:2, 12; 46:14 and 51:19. 24  Jer 38(31):12 also renders two verbs with one verb: ‫א־יֹוסיפוּ ְל ַד ֲא ָבה עֹוד‬ ִ ֹ ‫—וְ ל‬καὶ οὐ πεινάσουσιν ἔτι.

The Translations of Infinitives in Septuagint Jeremiah

147

In a number of cases the rendering does not translate the final aspect of the Hebrew prepositional phrase. Seven times the preposition εἰς assumes a directional function, in each case in connection to a verb of motion: 14:8 ‫וּכא ֵֹר ַח נָ ָטה ָללוּן‬ ְ ‫אָרץ‬ ֶ ‫— ָל ָמּה ִת ְהיֶ ה ְכּגֵ ר ָבּ‬ἵνα τί ἐγενήθης ὡσεὶ πάροικος ἐπὶ τῆς γῆς καὶ ὡς αὐτόχθων ἐκκλίνων εἰς κατάλυμα. 28(51):40 ‫בֹוח‬ ַ ‫אֹור ֵידם ְכּ ָכ ִרים ִל ְט‬ ִ —καταβιβάσω αὐτοὺς ὡς ἄρνας εἰς σφαγὴν 48(41):6 ‫אתם‬ ָ ‫—וַ יֵּ ֵצא יִ ְשׁ ָמ ֵעאל ֶבּן־נְ ַתנְ יָ ה ִל ְק ָר‬καὶ ἐξῆλθεν εἰς ἀπάντησιν αὐτοῖς ­Ισμαηλ

The rendering εἰς+noun is used twice in Septuagint Jeremiah to render an adverbal ‫ל‬+inf. that functions as a direct object. In both instances, Jer 29(49):9 and 30:8(49:30), the phrase and its translation are identical: ‫—ה ְע ִמיקוּ ָל ֶשׁ ֶבת‬ἀθύνατε ֶ εἰς κάθισιν. Είς+noun is the rendering of a final adnominal ‫ל‬+inf. twice ‫רוּח ַצח‬ ַ ‫ת־ע ִמּי לֹוא ִלזְ רֹות וְ לֹוא ְל ָה ַבר‬ ַ ‫—שׁ ָפיִ ים ַבּ ִמּ ְד ָבּר ֶדּ ֶרְך ַבּ‬Πνεῦμα ְ πλανήσεως ἐν τῇ ἐρήμῳ, ὁδὸς τῆς θυγατρὸς τοῦ λαοῦ μου οὐκ εἰς καθαρὸν οὐδ᾿ εἰς ἅγιον (4:11), and it translates ‫ל‬+inf. functioning as a genitive attribute once ‫בֹוח‬ ַ ‫יכם ִל ְט‬ ֶ ‫יְמ‬ ֵ ‫י־מ ְלאוּ‬ ָ ‫—כּ‬ὅτι ִ ἐπληρώθησαν αἱ ἡμέραι ὑμῶν εἰς σφαγήν (32:20[25:34]). There are a few other means by which the Hebrew ‫ל‬+inf. is translated in Septuagint Jeremiah. Participles are used on several occasions. These include a predicative participle (51[44]:18, below), modal participles (45[38]:4 and 51[44]:3), a genitive attribute (51[44]:14), and three other instances in which a participle is used (33[26]:8, 50[43]:1 and 48[41]:4) 51(44):18 ‫וּמן־אָז ָח ַד ְלנוּ ְל ַק ֵטּר ִל ְמ ֶל ֶכת ַה ָשּׁ ַמיִם‬ ִ —καὶ ὡς διελίπομεν θυμιῶντες τῇ βασιλίσσῃ τοῦ οὐρανοῦ 45(38):4 ‫יהם ַכּ ְדּ ָב ִרים ָה ֵא ֶלּה‬ ֶ ‫… ְל ַד ֵבּר ֲא ֵל‬ ‫הוּא־מ ַר ֵפּא ֶאת־יְ ֵדי אַנְ ֵשׁי ַה ִמּ ְל ָח ָמה‬ ְ —αὐτὸς ἐκλύει τὰς χεῖρας τῶν ἀνθρώπων τῶν πολεμούντων … λαλῶν πρὸς αὐτοὺς κατὰ τοὺς λόγους τούτους 51(44):3 ‫אֹלהים ֲא ֵח ִרים‬ ִ ‫— ִמ ְפּנֵ י ָר ָע ָתם ֲא ֶשׁר ָעשׂוּ ְל ַה ְכ ִע ֵסנִ י ָל ֶל ֶכת ְל ַק ֵטּר ַל ֲעבֹד ֵל‬ἀπὸ προσώπου πονηρίας αὐτῶν, ἧς ἐποίησαν παραπικρᾶναί με πορευθέντες θυμιᾶν θεοῖς ἑτέροις 51(44):14 ‫גוּר־שׁם ְבּ ֶא ֶרץ ִמ ְצ ָריִם‬ ָ ‫הוּדה ַה ָבּ ִאים ָלגוּר‬ ָ ְ‫—וְ לֹא יִ ְהיֶ ה ָפּ ִליט וְ ָשׂ ִריד ִל ְשׁ ֵא ִרית י‬καὶ οὐκ ἔσται σεσῳσμένος οὐθεὶς τῶν ἐπιλοίπων Ιουδα τῶν παροικούντων ἐν γῇ Αἰγύπτῳ 33(26):8 ‫ר־צוָּ ה יְ הוָ ה‬ ִ ‫ל־א ֶשׁ‬ ֲ ‫—וַ יְ ִהי ְכּ ַכלֹּות יִ ְר ְמיָ הוּ ְל ַד ֵבּר ֵאת ָכּ‬καὶ ἐγένετο Ιερεμίου παυσαμένου λαλοῦντος πάντα, ἃ συνέταξεν αὐτῷ κύριος 50(43):1 ‫ל־ה ָעם‬ ָ ‫ל־כּ‬ ָ ‫—וַ יְ ִהי ְכּ ַכלֹּות יִ ְר ְמיָ הוּ ְל ַד ֵבּר ֶא‬Καὶ ἐγενήθη ὡς ἐπαύσατο Ιερεμίας λέγων πρὸς τὸν λαὸν 48(41):4 ‫—וַ יְ ִהי ַבּיֹּום ַה ֵשּׁנִ י ְל ָה ִמית ֶאת־גְּ ַד ְליָ הוּ‬καὶ ἐγένετο τῇ ἡμέρᾳ τῇ δευτέρᾳ πατάξαντος αὐτοῦ τὸν Γοδολίαν

The participles in 33(26):8 and 50(43):1 are onset by the use of the verb παύω, which is often aligned with a participle, and sometimes even a genitive. Jer

148

Miika Tucker

48(41):4 displays the translator’s use of the genitive absolute, which is quite infrequent in Septuagint Jeremiah. The causative aspect of the hiphil is reflected in the explicit subject αὐτοῦ, which otherwise does not have an equivalent in the MT. In addition, there are a few exceptional renderings of ‫ל‬+inf. The construct ‫ל ָמ ְלכֹו‬,ְ when functioning as a genitive attribute, is twice translated by a noun in the genitive: ‫—בּ ְשׁנַ ת ָה ְר ִב ִעית ְל ָמ ְלכֹו‬ἐν ִ τῷ ἔτει τῷ τετάρτῳ τῆς βασιλείας αὐτοῦ (28[51]:59) and ‫—וַ יְ ִהי ַב ָשּׁנָ ה ַה ְתּ ִשׁ ִעית ְל ָמ ְלכֹו‬καὶ ἐγένετο ἐν τῷ ἔτει τῷ ἐνάτῳ τῆς βασιλείας αὐτοῦ (52:4). The conjunction ὥστε+inf., which is considered to be a more literary rendition, is found once as a translation of ‫ל‬+inf.: ְ ‫יְמ ֲאנוּ‬ ָ ‫וְ ָהיָ ה ִכּי‬ ‫ת־הכֹּוס ִמיָּ ְדָך ִל ְשׁתֹּות‬ ַ ‫—ל ַק ַח‬καὶ ָ ἔσται ὅταν μὴ βούλωνται δέξασθαι τὸ ποτήριον ἐκ τῆς χειρός σου ὥστε πιεῖν (32:14[25:28]). In sum, the renderings of ‫ל‬+inf. ְ in Septuagint Jeremiah can be divided into three categories: object infinitives, purpose expressing infinitives, and other infinitives. Table 2 depicts the manner in which the anarthrous infinitive and the genitive articular infinitive are used to render ‫ל‬+inf. ְ in these categories in these functions in Jer a’ and Jer b’. Table 2: The distribution of the anarthrous infinitive (0+inf.) and the genitive articular infinitive (τοῦ+inf.), according to their function, as renderings of ‫ ְל‬+inf. in Septuagint Jeremiah Obj. inf.

Jer a’ Jer b’

Purpose expressing inf.

Other inf.

0+inf.

τοῦ+inf.

0+inf.

τοῦ+inf.

0+inf.

τοῦ+inf.

Total

9 11

10 2

8 32

43 26

1 9

10 11

81 91

As noted earlier, the most significant differences between Jer a’ and Jer b’ are evident in the use of the articular infinitive as a direct object, and the renderings of purpose expressing infinitives on the whole. Especially regarding the purpose expressing infinitives, there is a major drop in the use of the articular infinitive, from 43 out of 51 (84%) possible cases to 26 out of 58 (45%) possible cases. At the same time, the use of the anarthrous infinitive increases from 8 cases (16%) to 32 (55%). There are two main differences in the category of other infinitives. Most cases in which a prepositional phrase is used as the rendering of ‫ל‬+inf. ְ are located in Jer a’, and conversely all the cases in which a participle is used are located in Jer b’. These differences are discussed further below.

149

The Translations of Infinitives in Septuagint Jeremiah

2.4 ‫ ְב‬+ Infinitive and ‫ ְכ‬+ Infinitive There are 14 translated cases of ‫ב‬+inf. ְ in Jeremiah, and all of them are temporal in nature. Similarly, there are only 13 cases of ‫כ‬+inf., ְ eleven of which function as temporal and two of which are comparative. The prepositional phrase ‫ב‬+inf. ְ is translated in several different ways in the Septuagint. All of the common renditions are found also in Septuagint Jeremiah. Table 3: Translations of ‫ ְב‬+inf. in Septuagint Jeremiah

Jer a’ Jer b’

ἐν τῷ+inf.

ἐν+noun

ὅτε -clause

ὅταν -clause

gen. abs.

other

Total

1 2

1 –

2 4

– 1

– 2

– 1

4 10

Table 3 shows that there are more instances of the prepositional phrase in Jer b’. Both halves include cases where the more literal rendering ἐν τῷ+inf. has been used and cases where ὅτε -clauses have been used. The major difference is the employment of the genitivus absolutus -construction on two occasions in Jer b’. Here are the two cases: 35(28):9 ‫ר־שׁ ָלחֹו יְ הוָ ה‬ ְ ‫ַהנָּ ִביא ֲא ֶשׁר יִ נָּ ֵבא ְל ָשׁלֹום ְבּבֹא ְדּ ַבר ַהנָּ ִביא יִ וָּ ַדע ַהנָּ ִביא ֲא ֶשׁ‬ ‫— ֶבּ ֱא ֶמת‬ὁ προφήτης ὁ προφητεύσας εἰς εἰρήνην, ἐλθόντος τοῦ λόγου γνώσονται τὸν προφήτην, ὃν ἀπέστειλεν αὐτὸν κύριος ἐν πίστει 49(42):18 ‫ֹלהי יִ ְׂש ָר ֵאל ַּכ ֲא ֶׁשר נִ ַּתְך ַא ִּפי וַ ֲח ָמ ִתי ַעל־י ְֹׁש ֵבי יְ רּו־‬ ֵ ‫ּכי כֹה ָא ַמר יְ הוָ ה ְצ ָבאֹות ֱא‬ ‫יכם ְּבב ֲֹא ֶכם ִמ ְצ ָריִם‬ ֶ ‫— ָׁש ַלםִ ֵּכן ִּת ַּתְך ֲח ָמ ִתי ֲע ֵל‬ὅτι οὕτως εἶπε κύριος Καθὼς ἔσταξεν ὁ θυμός μου ἐπὶ τοὺς κατοικοῦντας Ιερουσαλημ, οὕτως στάξει ὁ θυμός μου ἐφ᾽ ὑμᾶς ­εἰσελθόντων ὑμῶν εἰς Αἴγυπτον

The use of the genitivus absolutus is generally seen as exceptional in the context of the Septuagint,25 especially within a translation that is considered to be fairly literal. Similar Hebrew expressions are rendered by ὅτε -clauses elsewhere in Septuagint Jeremiah, and the renderings with the preposition ἐν are generally used in cases where the translator has likely interpreted the text in a different manner from the later Masoretic text or where he has difficulty understanding the text. The following two verses are examples of this phenomenon.

25 Soisalon-Soininen, Die Infinitive, 89.

150

Miika Tucker

28(51):39 ‫יהם‬ ֶ ‫ת־מ ְשׁ ֵתּ‬ ִ ‫אָשׁית ֶא‬ ִ ‫— ְבּ ֻח ָמּם‬ἐν τῇ θερμασίᾳ αὐτῶν δώσω πότημα αὐτοῖς 22:23 ‫א־לְך ֲח ָב ִלים ִחיל ַכּיּ ֵֹל ָדה‬ ָ ‫—י ַֹשׁ ְב ְתּי ַבּ ְלּ ָבנֹון ְמ ֻקנַּ נְ ְתּי ָבּ ֲא ָרזִ ים ַמה־נֵּ ַחנְ ְתּ ְבּבֹא‬κατοικοῦσα ἐν τῷ Λιβάνῳ ἐννοσσεύουσα ἐν ταῖς κέδροις καταστενάξεις ἐν τῷ ἐλθεῖν σοι ὀδύνας, ὠδῖνας ὡς τικτούσης

Jer 28(51):39 displays a rare case of ἐν+noun instead of ἐν τῷ+inf.26 The translator understood ‫ חמם‬as the noun ‫ חם‬with a plural suffix instead of an infinitive. The meaning of the Hebrew text in 22:23 is difficult even for modern readers. It transmits two Ketiv/Qere readings in ‫ י ַֹשׁ ְב ְתּי‬and ‫מ ֻקנַּ נְ ְתּי‬. ְ The verb ‫ נֵּ ַחנְ ְתּ‬is rendered by καταστενάξεις, which reflects an understanding of the verb as ‫ אנח‬instead of the niphal form of ‫חנן‬.27 The prepositional phrase ‫כ‬+inf. ְ is also translated by several different means in Septuagint Jeremiah. Table 4: Translations of ‫ ְכ‬+inf. in Septuagint Jeremiah28

Jer a’ Jer b’

ἐν τῷ+inf.

ὡς -clause

ὅτε -clause

ὅταν -clause

gen. abs.

Total

1 –

1 2

– 1

2 –

– 3

4 6

Throughout the Septuagint, the most common equivalent of ‫כ‬+inf. ְ is a ὡς -clause.29 There are three such renderings in Septuagint Jeremiah. The genetivus absolutus appears again as a special rendering on three occasions, and again all within Jer b’. Soisalon-Soininen notes that Jeremiah is exceptional among the books of the Septuagint by the fact that it alone attests so many cases of this equivalence.30 It is important to note that ‫כ‬+inf. ְ shares a certain number of equivalents with ‫ב‬+inf. ְ in the Septuagint. Since the two Hebrew prepositions are orthographically similar, one cannot always be so sure which of the two the translator thought he was translating. When the equivalent is the prepositional phrase ἐν τῷ+inf., as in Jer 25:12, it seems even more likely that he understood the text as ‫ב‬+inf.: ְ 26 Soisalon-Soininen, Die Infinitive, 83. 27  This verb elicits confusion among commentators as well. The Dictionary of Classical Hebrew (ed. by D.J.A. Clines; Sheffield: Sheffield Phoenix Press 1996) considers the verb to be the niphal form of ‫חנן‬, but adds that it might be a corrupted form of ‫( אנח‬see volume 3, p. 273). The latter solution may be traced to H. Bauer and P. Leander, Historische Grammatik der Hebräischen Sprache des Alten Testaments (2 vols.; Halle: M. Niemeyer, 1918), 315, who suggest that ‫ נֵּ ַחנְ ְתּ‬is the result of a metathesis of ‫ נ‬and ‫ח‬. 28  Three occurrences of ‫כ‬+inf. ְ (Jer MT 5:26; 17:2 and 41:6) do not have Greek equivalents. 29 Soisalon-Soininen, Die Infinitive, 95. 30 Soisalon-Soininen, Die Infinitive, 97.

The Translations of Infinitives in Septuagint Jeremiah

151

25:12 ‫—וְ ָהיָ ה ִכ ְמלֹאות ִשׁ ְב ִעים ָשׁנָ ה‬καὶ ἐν τῷ πληρωθῆναι ἑβδομήκοντα ἔτη

The only comparative ‫כ‬+inf. ְ that has an equivalent in Septuagint Jeremiah is in 6:7. It is translated by a ὡς -clause, which is very appropriate in the context. 6:7 ‫יה ֵכּן ֵה ֵק ָרה ָר ָע ָתהּ‬ ָ ‫ימ‬ ֶ ‫— ְכּ ָה ִקיר ַבּוִ ר ֵמ‬ὡς ψύχει λάκκος ὕδωρ, οὕτως ψύχει κακίαν αὐτῆς

The three genitivus absolutus -constructions all appear in Jer b’. Each one occurs as part of a temporal expression, preceded by the formula ‫וַ יְ ִהי‬. Here are two examples:31 33(26):8 ‫ר־צוָּ ה יְ הוָ ה‬ ִ ‫ל־א ֶשׁ‬ ֲ ‫—וַ יְ ִהי ְכּ ַכלֹּות יִ ְר ְמיָ הוּ ְל ַד ֵבּר ֵאת ָכּ‬καὶ ἐγένετο Ιερεμίου παυσαμένου λαλοῦντος πάντα, ἃ συνέταξεν αὐτῷ κύριος 43(36):23 ‫אַר ָבּ ָעה יִ ְק ָר ֶע ָה ְבּ ַת ַער ַהסּ ֵֹפר‬ ְ ְ‫הוּדי ָשֹׁלשׁ ְדּ ָלתֹות ו‬ ִ ְ‫—וַ יְ ִהי ִכּ ְקרֹוא י‬καὶ ἐγενήθη ἀναγινώσκοντος Ιουδιν τρεῖς σελίδας καὶ τέσσαρας, ἀπέτεμνεν αὐτὰς τῷ ξυρῷ τοῦ γραμματέως

The other cases of ‫כ‬+inf., ְ which are rendered by ὥς, ὅτε and ὅταν -clauses, also occur after the temporal formula ‫וַ יְ ִהי‬. Especially noteworthy is 50(43):1, which translates the exact same Hebrew expression as 33(26):8, but renders it with a ὥς -clause instead of the genetivus absolutus: 50(43):1 ‫ל־ה ָעם‬ ָ ‫ל־כּ‬ ָ ‫—וַ יְ ִהי ְכּ ַכלֹּות יִ ְר ְמיָ הוּ ְל ַד ֵבּר ֶא‬Καὶ ἐγενήθη ὡς ἐπαύσατο Ιερεμίας λέγων πρὸς τὸν λαὸν

Though the data attained from the translations of the prepositional phrases ‫ב‬+inf. ְ and ‫כ‬+inf. ְ is not very voluminous, the occurrence of five cases of the genetivus absolutus in Jer b’ forms a striking contrast with the absence of the equivalent in Jer a’. Especially significant is the fact that the temporal expressions which are rendered by the genetivus absolutus in Jer b’ have parallel expressions in Jer a’ that are rendered by other means. This difference effects the impression of a different translation character.

3. Discussion A glimpse at the distribution of the two main manifestations of infinitives in Septuagint Jeremiah, the genitive articular infinitive (τοῦ+inf) and the anarthrous

31  The third case is in Jer 48(41):7.

152

Miika Tucker

infinitive, might help to visualize the distinction between the two halves of the book. The distribution is presented in the following two charts. 14 12 10 8 6 4 2 0

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Chart 1: Τοῦ+inf. in each chapter of Septuagint Jeremiah32

Chart 1 shows the cases of τοῦ+inf. in each chapter of Septuagint Jeremiah. There is a spike in occurrences in chapter 18, but otherwise τοῦ+inf. appears consistently throughout the translation. The total number of occurrences in each half is 73 cases in Jer a’ and 42 cases in Jer b’. 16 14 12 10 8 6 4 2 0

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Chart 2: The anarthrous infinitive in each chapter of Septuagint Jeremiah33

32  These figures include the equivalent τοῦ+inf. both as a rendering of ‫ל‬+inf. ְ as well as the inf. without a preposition, but exclude renderings of ‫לבלתי‬+inf., which often are rendered with a genitive article (τοῦ μή). 33  These figures include all instances in which the anarthrous infinitive appears in Jer LXX. In 70 of these cases the Hebrew equivalent is ‫ל‬+inf. ְ

The Translations of Infinitives in Septuagint Jeremiah

153

Chart 2 shows the cases of the infinitive without an article in Septuagint Jeremiah. There is a clear increase in the usage of the anarthrous infinitive in the latter half of the book. The number of cases more than doubles from 31 in Jer a’ to 70 in Jer b’. Why does the use of τοῦ+inf. slightly decrease as we move toward the end of the book, and why does the use of the anarthrous infinitive increase at the same time? The previous analysis has identified certain categories of the Hebrew infinitive to which this shift can be attributed. The use of ‫ל‬+inf. ְ to express the purpose of the main verb is the most obvious. Such cases are mainly rendered by the articular infinitive in Jer a’ (43 out of 51 cases [84%]) and by the anarthrous infinitive in Jer b’ (32 out of 58 cases [55%]). On a smaller scale, this shift is evident from the renderings of ‫ל‬+inf. ְ as a direct object and of ‫ל‬+inf. ְ in other functions. The equivalents of the direct objects in Jer a’ are quite evenly divided between the two translation options, but in Jer b’ the anarthrous infinitive is used almost exclusively (11 out of 13 times [85%]). When rendering other uses of ‫ל‬+inf., ְ Jer a’ mainly employs the articular infinitive (10 out of 11 cases), while Jer b’ displays a more equal balance of the two equivalents in using the anarthrous infinitive nine times and the articular infinitive eleven times.34 The remainder of this paper consists of a closer look at these infinitives with the possibility of bisectioning in mind. The evidence will be applied to both the hypothesis of a revised text in Jer b’,35 and to the hypothesis that the whole book is the work of a single translator. This procedure is adopted in order to ascertain how well these possibilities explain the evidence. The renderings found in Jer a’ can act as a starting point, as they figure to represent the OG in each hypothesis. From there, the change that takes place when one moves on to the renderings in Jer b’ can be considered.

3.1 Purpose Expressing ‫ ְל‬+ Infinitives In Jer a’, 8 cases of ‫ל‬+inf. ְ that express purpose are rendered by the anarthrous infinitive, and 43 cases are rendered by τοῦ+inf. Furthermore, 5 of the anarthrous infinitives occur in a single sentence (1:10), so that Jer a’ attests this equivalence in only 4 verses (1:10, 11:13, 15:21[20]36 and 16:8). In these four verses, the infinitives express the purpose of the verbs καθίστημι, τάσσω and εἰσέρχομαι, verbs of action and movement, and of the verb ‘to be’ (15:21[20]). The articular ­infinitives 34  Cf. table 2 above. 35  The notion of two translators will not be examined here, as it is felt that Tov has sufficiently demonstrated distinctive agreements between the two halves of the translation. For Tov’s evidence of distinctive agreements between Jer a’ and Jer b’, cf. Tov, The Septuagint Translation, 19–40. 36 The Vorlage of the Greek καὶ ἐξαιρεῖσθαί σε at the beginning of Jer LXX 15:21 is located at the end of Jer MT 15:20 ‫וּל ַה ִצּ ֶילָך‬. ְ

154

Miika Tucker

are used in connection with verbs of action and movement as well, but they are also used with verbs of giving and sending. The Hebrew equivalents of these verbs are common: ‫בא‬, ‫בנה‬, ‫הלך‬, ‫יצא‬, ‫נתן‬, ‫ עׂשה‬and ‫ׁשלח‬. The translator of Jer a’ clearly preferred to use the articular infinitive when rendering purpose expressing infinitives with the preposition ‫ל‬,ְ but he used the anarthrous infinitive on rare occasions to express the purpose of verbs of action and movement. As the translator seems to aim towards isomorphism,37 it may be assumed that the article is used as a formal equivalent for the Hebrew preposition. In Jer b’, the preferred rendering shifts to the anarthrous infinitive, but not exclusively. The anarthrous infinitive renders purpose expressing ‫ל‬+inf. ְ in 32 cases, and the articular infinitive renders it in 26 cases. The articular infinitives in Jer b’ are used to render the same expressions as in Jer a’, that is verbs of movement, verbs of action, and verbs of giving and sending. The Hebrew counterparts are the exact same verbs: ‫נתן‬, ‫בא עׂשה‬, ‫בנה‬, ‫הלך‬, ‫ יצא‬and ‫ׁשלח‬. This gives a slight impression of continuity between the two halves. However, the anarthrous infinitive is used in similar cases, rendering infinitives that express the purpose of the same verbs of movement, action, giving and sending. Some differences can be observed, but such cases are only few in number. The evidence regarding purpose expressing infinitives can be described as follows. The articular infinitive is predominant in Jer a’, but in Jer b’ both the articular infinitive and the anarthrous infinitive are used, with a slight preference for the anarthrous infinitive. The most dramatic difference is the rise in the use of the anarthrous infinitive, from only 16% of cases in Jer a’ to 55% of cases in Jer b’. If this change is assumed to be the work of a reviser, he would seem to prefer to use the anarthrous infinitive, which would be in line with better Greek style. However, since early revisions generally strived to be more consequent and rigorous,38 and since a tendency toward pedantic literalism can already be observed in some books of the LXX,39 we are left with the puzzling fact that similar Hebrew expressions are rendered by both equivalents in Jer b’. Both the anarthrous infinitive and the articular infinitive express the purpose of movement, action, giving and sending. It is difficult to find any criteria that a reviser would have based his changes on, as Stipp has suggested. Also, with this frame of early revision in mind, one would expect the articular infinitive to be more condu37  For discussion on isomorphism as a characteristic trait of the translation of Jer LXX, cf. A. Pietersma/M. Saunders, “Ieremias: To the Reader”, in A. Pietersma/B. G. Wright (ed.), Electronic Edition of The New English Translation of the Septuagint (Oxford: Oxford University Press, 2009); and A. Shead, “Jeremiah”, in J.K. Aitken (ed.), T&T Clark Companion to the Septuagint (London: Bloomsbury 2015) 469–86, on p. 472. 38  J. Barr, The Typology of Literalism in Ancient Biblical Translations (MSU 15; Göttingen: Vandenhoeck & Ruprecht 1979), 281. 39  H.St.J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint (Cambridge: Cambridge University Press, 1909), 1:9.

The Translations of Infinitives in Septuagint Jeremiah

155

cive to a revisional tendency, as it represents a formal equivalent of the Hebrew preposition ‫ל‬.ְ This inconsistency, however, does allow the conclusion that the continued use of the articular infinitive into Jer b’ may be understood as part of the common textual basis between the OG text and the revised text, which is a necessary step in identifying a revised text.40 Assuming that the change in equivalents between Jer a’ and Jer b’ is the work of a single translator would lead to the conclusion that the translator changed his disposition with regard to the process of translation after he completed the translation of Septuagint Jeremiah 28. The possibility of an exegetical reason for the differences must be disregarded since the articular infinitives in Jer a’ and both the articular and anarthrous infinitives in Jer b’ render the same types of Hebrew expressions, i.e. ‫ל‬+inf. ְ constructions that express the purpose of verbs of action, including movement, giving and sending. Stipp’s proposal, that the same translator made a break in translation after chapter 28 and varied his equivalents to a greater degree once he assumed the translation process, can explain the differences between Jer a’ and Jer b’.41 However, this solution is not so different from that of a revision, since both identify similarities and differences in the translation character of the two halves.

3.2 ‫ ְל‬+inf. as Direct Object There are not as many occurrences of ‫ל‬+inf. ְ that function as direct objects as there are occurrences that express purpose. In Jer a’, nine such object infinitives are rendered by the anarthrous infinitive and ten by the genitive articular infinitive. The Hebrew expressions in which these translations are used are mainly ‫מאן‬+inf. and ‫יכל‬+inf., and both equivalents are used to render these expressions. Infinitival objects of other verbs are generally rendered by τοῦ+inf, with only the occasional anarthrous infinitive. In Jer b’, the evidence is slightly different. The articular infinitive is used to render only two cases of ‫ל‬+inf. ְ as a direct object of a verb, these verbs being ‫חׁשב‬ and ‫ נׂשא‬piel.42 The anarthrous infinitive, on the other hand, is used to render both the infinitive objects of the frequently occurring verbs ‫ מאן‬and ‫ יכל‬and the infinitive objects of other verbs as well, some of which are cases that are rendered by the articular infinitive in Jer a’.

40  The identification of a shared common textual basis between two texts is a prerequisite for identifying a relation of dependence between the two texts. For a discussion on the criteria for identifying a revision, cf. E. Tov, Textual Criticism of the Hebrew Bible (Minneapolis MN: Fortress Press 2012), 141. 41 Stipp, Offene Frage, 153–4. 42  ‫ חׁשב‬in 33(26):3 and ‫ נׂשא‬in 51(44):14.

156

Miika Tucker

The translations of the infinitives functioning as direct objects give the following picture. In Jer a’, the renderings are split nearly half and half between the anarthrous infinitive and the genitive articular infinitive, with nine cases of the former and ten cases of the latter. In Jer b’, on the other hand, the anarthrous infinitive is predominantly used, with a ratio of 11 cases to 2 in comparison to the articular infinitive. The situation is the opposite to that of the purpose expressing infinitives, for which Jer a’ displayed a strong preference for one of the equivalents, and Jer b’ showed more balance between the two equivalents. In this case, it would seem reasonable to argue that the equivalents in Jer b’ might be the result of revision. The equivalents are more consistently of one kind and the second equivalent is almost completely extinguished. The only possible objection would be that the more consistent use of the anarthrous infinitive does not suit the tendency of revisions as aiming toward a more literal representation of the Vorlage. An anarthrous infinitive does not contain a formal representation of the Hebrew preposition ‫ל‬.ְ On the other hand, the use of the anarthrous infinitive to express a direct object is more natural Greek than the use of the articular infinitive, which might reflect a more obvious way of producing a formal equivalent of the Hebrew expression. To suggest that a single translator made a break in translation only to assume the task with a different disposition, i.e. in a more natural Greek style with regard to object infinitives, seems like a ready explanation of the differences. Yet here again, the base conclusion of a different translation character in Jer b’ is the same as in the argument for a revision.

3.3 ‫ ְב‬+ Infinitive and ‫ ְכ‬+ Infinitive The occurrences of the prepositional phrases ‫ב‬+inf. ְ and ‫כ‬+inf. ְ are not very numerous in Jer, yet the small differences of a few cases of the genitivus absolutus in Jer b’ in comparison to the absence of such in Jer a’ can be used in arguing for a different character between the two sections. The difference comes to the fore with regard to the temporal use of the infinitival expressions. Such cases are rendered by the genitivus absolutus five times in Jer b’ and not a single time in Jer a’. To argue that these equivalents in Jer b’ have been revised does not appropriate well with models of early revision, as the genitivus absolutus is thought to be a trait of a more free and natural disposition toward translation. If revision proves to be the state of the text in Jer b’, then these equivalents must either be considered untouched OG renderings, or the revision must be characterized as very rigorous in part, yet tending toward very natural Greek expressions on occasion. The possibility of a single translator varying his renderings after a break is here also an applicable solution. One would be hard pressed, however, to explain the different character of the renderings in the two sections by appealing to an

The Translations of Infinitives in Septuagint Jeremiah

157

exegetical reason, as the underlying Hebrew consists of similar temporal expressions in both cases.

4. Conclusions This article has surveyed the renderings of the Hebrew expressions ‫ל‬+inf., ְ ‫ב‬+inf., ְ and ‫כ‬+inf. ְ in Septuagint Jeremiah. These expressions were chosen because their renderings demonstrate marked differences between the two halves of the translation, Jer a’ and Jer b’. The analysis has identified the following categories to which the differences pertain: The use of the genitive articular infinitive as a representation of ‫ל‬+inf. ְ when the infinitive functions as a direct object, the use of both the genitive articular infinitive and the anarthrous infinitive as renderings of ‫ל‬+inf. ְ when the infinitive expresses the purpose of verbs of action and movement, and finally the use of the genitivus absolutus as a rendering of the temporal aspects of both ‫ב‬+inf. ְ and ‫כ‬+inf. ְ The renderings of these infinitival expressions are very beneficial for the characterization of the translation. First, the renderings show that the use of the genitive articular infinitive significantly decreases after Septuagint Jeremiah 28, while at the same time the use of the anarthrous infinitive increases. This change would seem to correspond to more natural Greek expression, particularly with regard to infinitives that serve as direct objects. The use of the genitivus absolutus as a rendering of temporal ‫ב‬+inf. ְ and ‫כ‬+inf. ְ also seems to suggest this. These phenomena, however, go against the grain of the assumed tendencies of early revisions of the Septuagint, which are rather understood to strive toward more consistent and consequent equivalents of the Hebrew text. Regardless of what one thinks to be the cause of the differences between Jer a’ and Jer b’, whether that be revision or a change in the disposition of a single translator, this analysis has shown that the renderings of infinitives in Jer a’ and Jer b’ do indeed display a marked difference in their character. This difference is difficult to explain by appealing to textual accommodation or to an exegetical approach by a single translator, as similar Hebrew expressions are rendered by different means in the two halves of the book.

Anneli Aejmelaeus Translation Technique and the Recensions A Late Review of Ilmari Soisalon-Soininen’s Doctoral Thesis on the Text-Forms of Judges

1. Introduction In his dissertation, Ilmari Soisalon-Soininen was knocking at a door that would not even begin to open before one more decade had elapsed.1 The key to the lock was not found until the discovery of the Naḥ al Ḥ ever Minor Prophets scroll and its interpretation by Dominique Barthélemy as a “missing link” that explained textual phenomena that had puzzled scholars in other books of the Septuagint.2 It is, however, interesting to see what Soisalon-Soininen was able to achieve working on the text-forms of Judges alone without knowledge of the “missing link.” His goal was not to prepare a critical edition of the book, but as a matter of fact, he worked under similar conditions, as editors of critical editions do, trying to find out what happened in the transmission process—trying to define the phenomena that were at work in the textual history of the text at hand—without however having previous knowledge of all the various possibilities.

2. Defining Phenomena of Textual History With the phenomena that need to be defined I mean different tendencies, exemplified by numerous individual cases, to change or correct the text in a certain direction, different revisions or recensions according to certain principles— 1  Ilmari Soisalon-Soininen, Die Textformen der Septuaginta-Übersetzung des Richterbuches (AASF, ser. B, 72,1; Helsinki: Suomalainen Tiedeakatemia, 1951). References to this work are given as page numbers within parentheses. 2  The greater part of the Naḥ al Ḥ ever scroll came into the hands of Dominique Barthélemy 1952, a year later than Soisalon-Soininen’s dissertation was published. Barthélemy made the first announcement of the discovery in “Redécouverte d’un chainon manquant de l’histoire de la Septante”, RB 60 (1953) 18–29, and published the fragments in Les devanciers d’Aquila: Première publication intégrale du texte des fragments du Dodécaprophéton (VTSup 10; Leiden: Brill, 1963). The final publication of the fragments: E. Tov/R.A. Kraft/P.J. Parsons, The Greek Minor Prophets Scroll from Naḥ al Ḥ ever (8Ḥ evXIIgr) (DJD 8; Oxford: Clarendon Press, 1990). See also A. Aejmelaeus/T. Kauhanen (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (De Septuaginta Investigationes 9; Göttingen: Vandenhoeck & Ruprecht, 2017).

160

Anneli Aejmelaeus

whether translation-technical, linguistic, or stylistic—found in manuscript families. When dealing with a text-critical problem case—for instance, the text-forms of Judges—there are often different theoretical possibilities to explain what happened in the textual history, so that the existing alternative forms of the text could emerge: whether the change had happened in this or that direction, and what motivations were the driving forces in the minds of the revisers.3 Not all explanations are equally valid. The highest probability for an explanation can be reached when numerous analogous cases can be found in the same text or another related text. But what is one to do if there are cases that do not fit the patterns and phenomena that are known previously? Theoretically, there are often cases in which you have an either-or decision to make. Just to take a simple example, in many books of the Hebrew Bible, you are often confronted with the problem whether the Septuagint had a Hebrew Vorlage frequently differing from the MT or alternatively the translator was being excessively free. Deciding how to solve these problem cases is like launching a domino effect: your previous decisions function as evidence for the following ones, and you find more and more cases of the same kind. Even if this is just an example, we know very well that certain famous scholars have been axiomatically of the opinion that the solution is practically always found in the translator’s way of translating, whereas today we are much more open to the other alternative, the Vorlage having been different. Having made this decision, we need to deal with the alternatives whether the change had happened in the Vorlage of the Septuagint or in the textual line of the MT. Again, many scholars have been reluctant to see the changes in the MT, that is, to consider the MT to represent a later stage of the textual history. Also, in this respect, the times have changed—and the dominoes are falling in the other direction. The question is how to know which way the dominoes should go, if you are dealing with a new or yet unsolved problem, as Soisalon-Soininen was dealing with in his dissertation project. In scholarly literature, there are always certain models of how to solve the problem cases. There are certain phenomena of textual history that have been established beyond doubt. For Soisalon-Soininen, Origen’s endeavours in comparing the Greek and Hebrew texts of his time were the most well-known phenomenon and influential factor in the textual history of the Septuagint, although there are still today plenty of questions about the Hexapla to be discussed. However, if you are confronted with cases that do not conform with any of the phenomena known so far, you need to collect your evidence very carefully. One or two examples do not make a text-historical phenomenon. You need 3  See my “Corruption or Correction? Textual Development in the MT of 1 Samuel 1”, in P.A. Torijano Morales/A. Piquer Otero (ed.), Textual Criticism and Dead Sea Scrolls Studies in Honour of Julio Trebolle Barrera—Florilegium Complutense (Supplements to the Journal for the Study of Judaism 157; Leiden: Brill, 2012) 1–17.

Translation Technique and the Recensions

161

several examples of which at least some are indisputable. If it is a recensional phenomenon that you are working on, you also need a plausible explanation to it as well as a motivation behind it. An example of this, if I may take one from my own work, would be the discovery of kaige-type corrections according to the Hebrew text in the B text of 1 Samuel (= 1 Kingdoms), which I have discussed time and again during the recent years.4 Even this discovery would not have been possible without the “missing link.” From the theoretical point-of-view, the method to be used in a certain inquiry always needs to correspond to the reality to be studied, and the phenomena I am talking about are part of that reality. It may be a hypothetical reality at first if you do not know too much about it, but as soon as the material begins to reveal its secrets and a more correct picture of the reality emerges, the method needs to be adjusted to it—and not the other way around. In my example, this would mean that the significance of the B text, when establishing the critical text, is no longer the same as before.

3. Discovering the Translation-Technical Method When Ilmari Soisalon-Soininen embarked upon a study of the different textforms of Judges, he knew very little about the reality behind the text-forms. The only major text-historical phenomenon known at that time was the work of Origen, that is, the Hexapla and Hexaplaric influence on the Greek manuscripts. Of the two other Christian text-forms mentioned by Jerome, the Lucianic and the Hesychian, the former was only vaguely known, the latter was a mystery. More than that was not known about the recensions. Instead, the interest of scholars seemed to go in a different direction: several scholarly works had been written concerning the division of certain books between different translators.5 This seems to have been a fashion in the field before the time of Soisalon-Soininen. The state of scholarship did not give a very good starting-point to his inquiry. Innovation was called for.

4 See my recent article “Kaige Readings in a Non-Kaige Section in 1 Samuel”, in A. Aejmelaeus/T. Kauhanen (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (De Septuaginta Investigationes 9; Göttingen: Vandenhoeck & Ruprecht, 2017) 169–84. 5  For instance, H.St.J. Thackeray, “The Greek Translators of Jeremiah”, JTS 4 (1903) 245– 66; and “The Greek Translators of the four Books of Kings”, JTS 8 (1907) 262–78; F. Baumgärtel, “Die Septuaginta zu Jesaja das Werk zweier Übersetzer”, in J. Herrmann/F. Baumgärtel (ed.), Beiträge zur Entstehungsgeschichte der Septuaginta (Beiträge zur Wissenschaft vom Alten Testament, Neue Folge Heft 5; Leipzig: Kohlhammer, 1923) 20–31; J. Herrmann/F. Baumgärtel, “Die Septuaginta zum Zwölfprophetenbuch das Werk zweier Übersetzer”, ibid., 32–38.

162

Anneli Aejmelaeus

The main point of orientation for Soisalon-Soininen was the fact that the Septuagint is a translation, produced by translators who showed varying degrees of literalism. This was the reality behind his object of study that mattered for him. Consequently, he was led by his observations on the text, both Hebrew and Greek, to formulate his methodology, which proved to be successful. Both at the beginning and at the end of his thesis he emphasizes the utmost importance of approaching the Septuagint from a translation-technical point of view, researching it in relation to its Hebrew Vorlage. This is the point-of-view from which he suggests that the syntax of the Greek language used in the Septuagint should be studied. Syntax is the area of language that—along with idioms—is most of all affected by the process of translation, the area in which there is the most interference from the source text. Soisalon-Soininen deplores that so little had been done in this area so far. He also refers to the studies concerning different translators in different parts of books, seeing in them the weakness that they only deal with the vocabulary usage. He wisely remarks that the lexical items could just as well have been changed by revisers and need not indicate different translators. As we know, most—if not all—of those theories about different translators within one book have been replaced by theories about original translators and subsequent revisers. How to distinguish between translation and revision was the first problem Soisalon-Soininen tackled in the case of Judges.

4. One or Two Translations in Judges? There had been a great deal of discussion concerning the text-forms of Judges, whether there had been different translations of the same text or just one translation and its revisions. Rahlfs could not solve the problem and consequently printed in his edition both main text-types, the A text on the upper half of the page and the B text below it. In the edition of Brooke-McLean the main text is taken as usual from the manuscript B, and the divergences of the A text have been printed in the apparatus in bold face. There was no consensus on the matter, and the opinions were divided. Instead of trying to describe the various text-forms through long lists of readings, as it had been done before, Soisalon-Soininen approached the problem through the translation-technical method, comparing the Greek text-forms not only with each other but with the Hebrew text.6 First of all, he was looking for translation-technical similarities between the text-forms. At the same time, he gave some general characterizations of the translator. No doubt, the translator’s knowledge of Hebrew was deficient, but precisely his false translations and trans6  The text to be compared was naturally the MT, but in a number of cases the Old Greek seems to have had a different Vorlage or at least to have read the consonants differently.

Translation Technique and the Recensions

163

literations that were transmitted in the various manuscripts revealed the common origin behind the text-forms (p. 29). For instance: Judg 3:15 ‫ד־יְמינ֑ ֹו‬ ִ ַ‫יְמ ִ֔יני ִ ֥איׁש ִא ֵ ּ֖טר י‬ ִ ‫ן־ה‬ ַ ‫ ֶּב‬υἱὸν τοῦ Ιεμενι ἄνδρα ἀμφοτεροδέξιον (an error and a unique formulation) Judg 4:11 ‫ וְ ֶ ֤ח ֶבר ַה ֵּקינִ י‬καὶ οἱ πλησίον τοῦ Κιναίου (A) – καὶ Χαβερ ὁ Κιναῖος (B) (an error and its correction)

No one would have changed the text in the direction of the false translations, but the divergences from such renderings could be seen as corrections. This very literal translator had employed some Hebraistic renderings that are otherwise rare in the Septuagint, and when such formulations occur in all manuscript groups, they also confirm the existence of a common background (p. 28). For instance: Judg 11:40 and 21:19 ‫ימה‬ ָ ‫יָמ‬ ִ֗ ‫ ִמּיָ ִ ֣מים‬ἀφ᾿ (ἐξ) ἡμερῶν εἰς ἡμέρας (a fairly rare literal rendering)7

On the other hand, the small-scale freedom shown by the translator in translation-technical additions and omissions,8 which could not have come about independently in different translations, were mostly found in all manuscript groups (p. 32). For instance: Judg 21:21 ‫נֹות־ׁשילֹו‬ ִ ‫ ְב‬αἱ θυγατέρες τῶν (κατ)οἰκούντων Σηλων (a free addition)

These kinds of observations made it obvious that there had been one translation from which all the existing text-forms diverge, whereas corrections in order to bring the text closer to the MT were later improvements and proof of a revision or revisions. In the following chapter of his dissertation, Soisalon-Soininen makes a thorough survey of various syntactical features of the translation in Judges, mainly features in which the difference in the structures of the two languages concerned, Hebrew and Greek, could be expected to have caused different renderings, either Hebraistic formulations, produced by the word-for-word method, or more natural Greek expressions which however demanded a somewhat freer approach.9 A 7 Cf. ἐξ ἡμερῶν εἰς ἡμέρας in 1 Sam 1:3, 25; 2:19. 8 Translation-technical additions and omissions are cases in which the translator has added a minor element of the text (e.g. an article or an explanatory word, or made the subject explicit) or omitted something unnecessary in Greek (e.g. repetition of prepositions) without changing the meaning of the text. 9  Hebraisms or Hebraistic expressions in Septuagint Greek were in the focus of SoisalonSoininen’s interest from the very beginning. He returned to this question in one of his latest articles, “Zurück zur Hebraismenfrage,” in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien zur

164

Anneli Aejmelaeus

major problem in this study was that the translation of Judges had been literal to such an extent that there were very few cases revealing the latter, somewhat freer attitude. Soisalon-Soininen characterizes the translation of Judges “with respect to the language, the weakest in the whole Septuagint” (p. 48).

5. Translation-Technical Survey of the Greek Text-Forms of Judges The translation-technical survey of syntactical features in Judges was conducted in the following areas, to mention the most important ones: (1) Word-order, which is free in Greek, but follows certain patterns in Hebrew (for instance, predicate-object-subject and attribute following the main word). The original translator was found to follow meticulously the word-order of the Hebrew text, with just a few exceptions in which the manuscript groups agree with each other, showing that they derive from one translation. However, the Lucianic text (A II) was occasionally found to change to a more natural word-order in Greek (pp. 33–37). (2) Conjunctions, which are numerous in Greek but scanty in Hebrew. The translation follows the sentence structure of the source text even in cases where the Greek language would have demanded different constructions. The conjunctions ‫ ו‬and ‫ כי‬are mainly rendered by καί and ὅτι respectively, even more consistently than elsewhere in the Septuagint. Rare exceptions mainly occur in single manuscript groups (pp. 37–40). (3) Prepositions, which are employed in all languages in their own idiosyncratic ways. Divergences from the word-for-word translation do occasionally occur in this very large and varied material, which in a few cases allows for conclusions concerning original translation or revision (pp. 43–45). (4) Semi-prepositions that include a noun (for instance, “face” or “eye”) tend to be translated in a Hebraistic manner, and there is large agreement among the manuscript groups, whereas the freer alternatives show differences between them (pp. 45–46). (5) Construction of verbs, which reveals many Hebraistic constructions that are rare elsewhere in the Septuagint (e.g. πολεμεῖν ἐν in A; παρατάσσεσθαι ἐν in B). A comparison of the Hebraistic usage in Judges with other books of the Septuagint shows similarities especially with 1–4 Kingdoms (pp. 46–50). (6) Pronomina, especially in cases in which Septuaginta. Robert Hanhart zu Ehren (MSU 20; Göttingen: Vandenhoeck & Ruprecht, 1990) 35–51, and gave the following definition: “Ein Hebraismus in der Septuaginta ist ein aufgrund des Hebräischen entstandener Ausdruck, der nicht mit dem Sprachgebrauch des gleichzeitigen Griechisch im Einklang steht, oder der vom Hebräischen aus erklärliche Gebrauch möglicher Vokabeln oder Ausdrücke im Griechischen in Zusammenhängen, in denen sie nach dem griechischen Sprachgebrauch nicht passend sind.” He emphasizes that an expression that is new in Greek need not be regarded as Hebraistic, if it accords with Greek idiom to such a degree that it could be accepted as genuine Greek (p. 39).

Translation Technique and the Recensions

165

Hebrew employs nouns instead of indefinite pronouns (e.g. ‫ איׁש‬translated by ἀνήρ instead of ἕκαστος or τις—in B ἀνήρ consistently, A with some variation) (pp. 50–52). (7) Translation of Hebrew infinitive constructions that results in various Hebraistic expressions. In the case of the infinitive absolute, translated either by a participle or a noun in the dative, differences can be observed between the manuscript groups (e.g. 11:30 παραδώσει παραδῷς A, διδοὺς δῷς B). The infinitive construct is mostly rendered by an infinitive in Greek, although the construction is not always good Greek. An arthrous infinitive with a preposition (e.g. 13:20 ἐν τῷ ἀναβῆναι τὴν φλόγα) does occur in genuine Greek, but its frequent occurrence as rendering of Hebrew temporal expressions must be regarded as Hebraistic (pp. 52–56). (8) Concerning Hebrew nominal clauses, the copula is fairly common, whether or not corresponding to the Hebrew pronoun, but surprising is the translation of ‫ אנכי‬by ἐγώ εἰμι in cases with a following finite verb (in B, occasionally also in other groups) (p. 41). (10) Translation of the Hebrew object marker ‫ את‬by σύν + accusative (p. 42). The two last-mentioned phenomena, known from Aquila, were not regarded as features of the original translation of Judges. Most of these topics are familiar from Soisalon-Soininen’s later studies as well as from those of his students. In his dissertation, he already had a program for a major research project. When these topics have been studied on a broader basis, either in the whole Septuagint or in the Greek Pentateuch, significant differences between the various translators have been discovered. Soisalon-Soininen already makes reference to such differences in this first study of his, but within Judges, very little difference could be observed between the text-forms. Nevertheless, Soisalon-Soininen was able to demonstrate that the uniformity of the text-forms in many uniquely literal or defective translations, as well as their divergence from strict literalism in the same rare cases, clearly speaks for the common origin in one translation and that the divergence of the B text from the A text or its subgroups (or the other way around) could be best explained as the result of a revision or revisions. In all the different categories of the survey, it was found that the B text represents the most intensive revision according to the Hebrew, practically the MT, and this revision had produced such Hebraistic expressions that are otherwise rare in the Septuagint. Soisalon-Soininen also recognized as revisional features those Hebraistic formulations that Barthélemy listed as characteristics of the kaige group (such as the translation of ‫ אנכי‬by ἐγώ εἰμι, of the object marker ‫את‬ by σύν + accusative, and the translation of ‫ איׁש‬by ἀνήρ instead of ἕκαστος), but he connected them with Aquila, which meant a later dating for the revision. Compared with the survey of the syntax, Soisalon-Soininen deals with the vocabulary very briefly. He counts nearly 600 cases in which there is a clear-cut difference in choice of words between the A text and the B text and in addition hundreds of cases where either the A manuscripts or the B manuscripts are d ­ ivided

166

Anneli Aejmelaeus

(pp. 62–70). The following examples reveal differences between the text-forms (disregarding, however, diverging manuscripts): ‫—נַ ַער‬παιδάριον A; νεανίας B ‫— ִמ ְל ָח ָמה‬πόλεμος A; παράταξις B ‫—הלך‬ἀπέρχεσθαι A; πορεύεσθαι B ‫—שׁוב‬ἀποστρέφειν A; ἐπιστρέφειν B The equivalents used by the B text suggest that it is mainly in the B text that the changes have happened. The special vocabulary found in the B text is in many cases rare or totally absent in other books of the Septuagint,10 so that it is more probable that the equivalents also found in other books represent the original translation.

6. Approaching the Solution However, revision was not only found in the B text but also in the various branches of the A text, namely, influence from the Hexaplaric recension, which was proven by the Aristarchian signs, asterisks and obeli, preserved in some manuscripts as well as in the Syrohexapla. One of the subgroups of A (A II) had been recognized as Lucianic and had recensional features of its own. Quantitative differences, plusses and minuses according to the Hebrew text, is where the Hexaplaric recension becomes most clearly visible. However, doublets that occur in all witnesses made Soisalon-Soininen hesitant: he thought that the doublets could be to a great extent pre-Hexaplaric, having their origin in marginal readings (pp. 77, 84).11 In fact, here we have a similarity between the textual histories of the Books of Samuel and Judges: doublets that witness the earliest phase of the textual history represented in the manuscripts.12 However, Soisalon‑Soininen did not regard this as conclusive evidence, and was thus inclined to think that the doublets, too, were Hexaplaric. A very puzzling fact was that the readings of A and B were in close agreement in cases where minuses of the Septuagint had been filled in, with asterisks being transmitted in witnesses of the A text (pp. 97–101). For instance:

10  Soisalon-Soininen (pp. 65–69) thus refutes the close relationship between A III and B presupposed by O. Pretzl, “Septuagintaprobleme im Buch der Richter”, Biblica 1 (1926) 233–69. It must be A III (partly also A II) that has taken over material from B. 11  In this context, the word doublet refers to double translations that consist of the Old Greek translation of a phrase or a short passage and its correction. 12  See below.

Translation Technique and the Recensions

167

Judg 4:1 ‫ וְ ֵא ֖הּוד ֵ ֽמת‬καὶ Αωδ ἀπέθανεν AI–a B Arm Aeth Syh(sub ) Judg 10:6 ‫ֹלהי ֲא ָרם‬ ֣ ֵ ‫ת־א‬ ֱ ‫ וְ ֶא‬καὶ τοῖς θεοῖς Συριας (Αραδ B) AI–a ptv AIII B Arm Sa Aeth La Syh(sub ) Judg 5:3 ‫ ָאנ ִ ֹ֣כי ָא ִׁ֔ש ָירה‬ἐγώ εἰμι ᾄσομαι cx B Sa Syh(sub ) Judg 12:11–12 ‫ּפט ֶאת־יִ ְׂש ָר ֵ ֖אל ֶ ֥ע ֶׂשר ָׁש ִנֽים׃‬ ֹ ֥ ‫ּוֹלנ֑י וַ ּיִ ְׁש‬ ִ ‫וַ ּיִ ְׁש ֤ ֹּפט ַ ֽא ֲח ָר ֙יו ֶאת־יִ ְׂש ָר ֵ֔אל ֵא ֖ילֹון ַהּזְ ֽב‬ ‫בּולן׃‬ ֽ ֻ ְ‫ּוֹלנ֑י וַ ּיִ ָּק ֵ ֥בר ְּב ַאּיָ ֖לֹון ְּב ֶ ֥א ֶרץ ז‬ ִ ‫ וַ ָּי ָ֖מת ֵא ֣לֹון ַהּזְ ֽב‬καὶ ἔκρινεν μετ᾽ αὐτὸν τὸν Ισραηλ Αιλων ὁ Ζαβουλωνίτης καὶ ἔκρινεν τὸν Ισραηλ δέκα ἔτη· καὶ ἀπέθανεν Αιλων ὁ Ζαβουλωνίτης ἐν Αιλιμ, καὶ ἔθαψαν αὐτὸν ἐν γῇ Ζαβουλων A B Arm Sa Aeth La Syh (sub )

This, if anything, seemed to connect the B text with the Hexaplaric recension. But how was this connection to be explained? The alternative models for solution, available in scholarly literature, were not many. Pretzl was one of the main interlocutors for Soisalon-Soininen, but his own survey of the plusses and minuses did not allow him to go with Pretzl, who saw the dependencies between the text-forms in a very different way (partly even through hypothetical differences between Hexapla and Tetrapla) and maintained that the B text had nothing to do with Origen.13 Even if a pre-Hexaplaric or a preChristian revision was occasionally mentioned in literature as one possibility, such intensive, Jewish revisional activity as we know from Naḥ al Ḥ ever was not yet in vision. In a footnote, Soisalon-Soininen presents as a theoretical possibility that both Origen and the B reviser could have complemented their text independently from the same source, primarily Theodotion, but he thought that it would have been too much of a coincidence, and those features that connected the B text with Aquila seemed to point to another direction (p. 102). The remaining alternatives for the B text were either Hexaplaric or later than that, and because he did not consider later Christian scribes to have been capable of such thorough comparison with Hebrew texts, he decided for Hexaplaric origin—a decision made in the first place by elimination of improbable or seemingly improbable alternatives. Pre-Hexaplaric Jewish revision was not yet available as a solution. Considering the result of Barthélemy’s study that there was a continuum from the kaige revision to Theodotion and Aquila, one cannot help admiring how close Soisalon-Soininen was to the correct solution already before the discovery of the Naḥ al Ḥ ever scroll. When I read Soisalon-Soininen’s study from the point-of-view of 1 Samuel (= 1 Kingdoms), I can see some parallel phenomena. The situation is much simpler in 1 Samuel, and this precisely makes it easier to see the parallels. Interestingly, the original translation of Judges is found to share several features with 1 Samuel (e.g. ‫ נלחם ב‬πολεμεῖν ἐν ‘to fight against’; ‫ נַ ַער‬παιδάριον ‘boy/ 13  O. Pretzl, “Septuagintaprobleme”, 233–8; “Der hexaplarische und tetraplarische Septuagintatext des Origenes in den Büchern Josue und Richter”, ByzZ 30 (1929) 262–68.

168

Anneli Aejmelaeus

youth’), including a number of erroneous translations and transliterations (e.g. ‫ ֵאפֹוד‬ἐφούδ). There are also in 1 Samuel numerous doublets that occur in practically all manuscripts; they must have originated with early marginal readings which were moved into the text—just as Soisalon-Soininen suggested at least for a greater part of the doublets in Judges.14 In 1 Samuel, as I see it, these early doublets represent the earliest stage of the manuscript tradition that is within our reach.15 The second earliest stage are the kaige-type corrections that I have found to occur sporadically in the B text of 1 Samuel. These kaige-type readings are often transmitted also by the Hexaplaric manuscripts. This is not indicative of the B text being Hexaplaric, but shows that Origen knew and used as one of his basic texts a manuscript of the B text, which must have existed some hundred and fifty years before Codex Vaticanus.16 There are also other scholars—for instance, Rahlfs—who have observed similarities between Origen’s basic text and the B text in Samuel and Kings.17 Something similar may have happened in Judges, too. If in Judges the B text was more thoroughly influenced by the kaige revision, a manuscript of this kind could very well have been among those used by Origen. His main interest was in the plusses and minuses, but if he had a manuscript that showed closer adherence to the Hebrew text in its word variants, he could choose such readings into his fifth column, considering the reading closest to the Hebrew to be original Septuagint.18 On the other hand, the B text that shows such intensive correction according to the MT must have been very close to Theodotion, which Origen had in one of his columns. In the end, Soisalon-Soininen gathered that all the various text-forms in Judges were dependent on the Hexapla, whereas a decade later Barthélemy connected the B text with the revision discovered at Naḥ al Ḥ ever, the so-called kaige group or revision. Nevertheless, Soisalon-Soininen made a pioneering study on recensions of the Septuagint and developed a methodology by which to study the Septuagint in all those areas of inquiry in which the translation character of the Septuagint plays a role. This is the case in the study of the syntax of Septua14  That the doublets originated with marginal readings is revealed by the location of the secondary part of the doublet, the correction: sometimes before the original counterpart, sometimes after it, and sometimes at a different location. 15  See my article “Textual History of the Septuagint and the Principles of Critical Editing”, in A. Piquer Otero/P. Torijano Morales (ed.), The Text of the Hebrew Bible and Its Editions: Studies in Celebration of the Fifth Centennial of the Complutensian Polyglot (Supplements to the Textual History of the Bible 1; Leiden/Boston: Brill, 2016) 160–79, esp. on p. 177. 16  See my “Kaige Readings in a Non-Kaige Section in 1 Samuel”, 172. 17  See A. Rahlfs, Septuaginta-Studien I. Studien zu den Königsbüchern (2nd ed.; Göttingen: Vandenhoeck & Ruprecht, 1965) 85; H.B. Swete, An Introduction to the Old Testament in Greek (2nd ed.; Cambridge: University Press, 1914), 487; B. Johnson, Die Hexaplarische Rezension des 1. Samuelbuches der Septuaginta (Studia Theologica Lundensia 22; Lund: Gleerup, 1963). 18  See my “Hexaplaric Recension and Hexaplaric Readings in 1 Samuel”, forthcoming.

Translation Technique and the Recensions

169

gint Greek, as pointed out by Soisalon-Soininen already in his dissertation. The same is true for textual criticism of the Septuagint. Working, for instance, on the critical edition of the Septuagint of 1 Samuel, one constantly needs to take into account the Hebrew Vorlage, its possible divergences from the MT, and the mode of translation characteristic of this particular translator. And as demonstrated by Soisalon-Soininen, the translation-technical method is also needed in the study of the recensions of the Septuagint. Distinguishing revisions from original translations, and characterizing the principles of the revisions, the most conclusive evidence is to be found—not in the vocabulary usage—but in the way the structure of the text and the syntactical features of the language were dealt with.

7. Methodology for Researching Recensions Soisalon-Soininen pointed out that the theories of different translators within one book could not be argued for in a conclusive manner on the basis of the evidence of vocabulary usage only. In the present research situation, the corresponding theories concern the differentiation between original translation and subsequent revision, but the concentration on vocabulary usage is still strong. However, it is worth noting that those characteristics of the kaige recension that Barthélemy lists in his epoch-making study include some that concern the syntax (in a broad sense of the word):19 (1) the translation of ‫ אנכי‬by ἐγώ εἰμι regardless of a following finite verb, (2) the elimination of the historical present, (3) the elimination of ἕκαστος as a rendering of ‫ איׁש‬and preference for ἀνήρ, (4) translation of ‫ אין‬by οὐκ ἔστιν without regard to the tense, and (5) the employment of καίγε for ‫וגם‬ (or ‫)גם‬, which often concerns the connection of sentences. The other listed items are, however, lexical: (6) translation of ‫ מעל‬by ἐπάνωθεν, (7) etymological translation of the roots ‫יצב‬/‫ נצב‬by στηλόω (from στήλη), (8) translation of ‫ לקראת‬by εἰς συνάντησιν or εἰς ἀπαντήν (instead of εἰς ἀπάντησιν), (9) the differentiation between ‫ ׁשופר‬κερατίνη and ‫ חצצרה‬σάλπιγξ. The list is not and was not meant to be exhaustive. By the listed features, Barthélemy demonstrated the connection between the different texts in which he could recognize the same phenomenon, the so-called kaige group or revision. After Barthélemy, all of the studies that have followed in his footsteps, looking for more characteristic features of kaige/Theodotion, have collected more lexical items. Tim McLay published 1998 an article with a list of all suggested translation equivalents, which add up to 96 in number, and more could be added to it from more recent publications.20 Not all of them are equally widely distributed, or even 19 Barthelemy, Les devanciers d’Aquila, 31–80. 20  T. McLay, “Kaige and Septuagint Research”, Textus 19 (1998) 127–39. The features suggested after Barthélemy are all lexical with the exception of one or two cases.

170

Anneli Aejmelaeus

indisputable, but this long list shows the trend in this kind of study. A further example can be found in Emanuel Tov’s study on the differences between the two halves of the Book of Jeremiah, which also deals with a recensional phenomenon, although of a different type, and is based on differences in vocabulary usage.21 One of the reasons for this state of affairs might be that it is less complicated to collect the material for a study on lexical features. Before the time of computers and search programs, you could collect this kind of material on the basis of a concordance. Syntactical phenomena could be collected only by reading through the whole text, as Soisalon-Soininen did for his study on Judges as well as later for his study on the infinitives.22 With modern search programs, it is possible to collect all kinds of materials, words not listed in concordances and grammatical forms as well as different combinations. You only need to have an idea what to search for. I would like to mention a study on the translation of the Books of Samuel by Raimund Wirth as an example.23 His dissertation is a description of the mode of translation in the Books of Samuel, based mainly on a translation-technical survey of syntactical features, among them the use of tenses in certain special cases. The translator of the Books of Samuel is a fairly literal one, but he proves to have a special talent for the use of verbal forms, especially the historical present in translation of normal Hebrew narrative forms and the imperfect in translation of the Hebrew past iterative as well as a few other forms. The translator also had a very creative way of alternating both the present and the imperfect with the aorist. Already Thackeray had paid attention to the historical present in Samuel– Kings and noticed that the sections where he believed to see a different translator (2 Sam 11:2–1 Kings 2:11; 1 Kings 22–2 Kings 25) avoided the use of the historical present, using instead the aorist. The elimination of the historical present is also listed by Barthélemy as one of the characteristics of kaige. By contrast, the use of the imperfect has drawn little attention.24

21  E. Tov, The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of the LXX of Jeremiah 29–52 and Baruch 1:1–3:8 (HSM 8; Missoula, MT: Scholars Press, 1976). 22  I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965). 23  R. Wirth, Die Septuaginta der Samuelbücher. Untersucht unter Einbeziehung ihrer Rezensionen (De Septuaginta Investigationes 7; Göttingen: Vandenhoeck & Ruprecht, 2016). 24  See however my “The Septuagint of 1 Samuel”, in L. Greenspoon/O. Munnich (ed.), VIII Congress of the International Organization for Septuagint and Cognate Studies, Paris 1992, (SBLSCS 41; Atlanta, GA: SBL, 1995) 109–29; repr. in A. Aejmelaeus, On the Trail of the Septuagint Translator: Collected Essays (Revised and Expanded Edition; CBET 50; Leuven: Peeters, 2007) 123–41. See also A. Voitila, “The Use of Tenses in the L- and B-Texts in the Kaige-Section of 2 Reigns”, in S. Kreuzer/M. Meiser/ M. Sigismund (ed.), Die Septuaginta—­Entstehung,

Translation Technique and the Recensions

171

More details on both phenomena can be found in Wirth’s study. In the kaige section of 2 Samuel, the historical present was changed to the aorist, which for these revisers was the correct rendering for the Hebrew narrative forms, the imperfect consecutive and the perfect. 2 Sam 10:17 καὶ παρετάξατο (παρατάσσεται L) Συρία ἀπέναντι Δαυὶδ καὶ ἐπολέμησαν μετ᾽ αὐτοῦ. Cf. 1 Chron 19:17 καὶ παρατάσσεται Σύρος ἐξ ἐναντίας Δαυὶδ καὶ ἐπολέμησαν αὐτόν.

Also in the case of the imperfect, change into the aorist could be seen, but Wirth was able to observe that the change did not happen in all cases. If the Greek imperfect was translating a Hebrew participle or an imperfect, it was tolerated, but in all other cases it was changed to the aorist. 2 Sam 17:20 ‫יְב ְקׁשּו וְ לֹא ָמ ָצאּו‬ ַ ַ‫ ו‬καὶ ἐζήτησαν (ἐζήτουν L) καὶ οὐχ εὗραν (εὕρισκον L) 2 Sam 15:37 ִ‫רּוׁש ָלם‬ ָ ְ‫חּוׁשי ֵר ֶעה ָדוִ ד ָה ִעיר וְ ַא ְב ָׁשֹלם יָבֹא י‬ ַ ‫ וַ ּיָבֹא‬καὶ εἰσῆλθεν (εἰσπορεύεται L) Χουσὶ ὁ ἑταῖρος Δαυὶδ εἰς τὴν πόλιν, καὶ Ἀβεσσαλὼμ εἰσεπορεύετο εἰς Ἰερουσαλήμ.

Thus, the translator had his patterns for the use of the tenses, and the kaige reviser had his, but different ones.25 Concerning the kaige sections in Samuel–Kings, there has been some controversy in scholarship about the role of the Lucianic text and to what extent it preserves the Old Greek readings. Wirth extended his survey of the use of tenses to the patterns of the Lucianic reviser—a methodological necessity in this kind of study. As for the historical present, the Lucianic reviser was found to be more reserved than the translator, changing it to the aorist in a number of cases. 1 Sam 6:14 ‫ת־ה ָּפרֹות ֶה ֱעלּו ע ָֹלה‬ ַ ‫ת־ע ֵצי ָה ֲעגָ ָלה וְ ֶא‬ ֲ ‫דֹולה וַ ַיְב ְּקעּו ֶא‬ ָ ְ‫וַ ַּת ֲעמֹד ָׁשם וְ ָׁשם ֶא ֶבן ּג‬ ‫ ַליהוָ ה‬καὶ ἔστησαν ἐκεῖ παρ᾽ αὐτῇ λίθον μέγαν∙ καὶ σχίζουσιν (ἔσχισαν L) τὰ ξύλα τῆς ἁμάξης, καὶ τὰς βόας ἀνήνεγκαν εἰς ὁλοκαύτωσιν τῷ Κυρίῳ.

Nevertheless, the Lucianic reviser is occasionally also found to be responsible for a change in the opposite direction for the sake of consistency: Sprache, Geschichte. 3. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 22.–25. Juli 2010 (WUNT 286; Tübingen: Mohr Siebeck, 2012) 213–37. 25 R. Wirth, “Das Praesens Historicum in den griechischen Samuelbüchern”, in K. De Troyer/T.M. Law/M. Liljeström (ed.), In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014); “Dealing with Tenses in the Kaige Section of Samuel”, in A. Aejmelaeus/T. Kauhanen (ed.), The Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (De Septuaginta Investigationes 9; Göttingen: Vandenhoeck & Ruprecht, 2017).

172

Anneli Aejmelaeus

1 Sam 7:6 ‫בּו־מיִם וַ ּיִ ְׁש ְּפכּו ִל ְפנֵ י יְ הוָ ה‬ ַ ‫ ּיִ ָּק ְבצּו ַה ִּמ ְצ ָּפ ָתה וַ ּיִ ְׁש ֲא‬καὶ συνήχθησαν εἰς Μασσηφάθ, καὶ ὑδρεύονται (= L) ὕδωρ καὶ ἐξέχεαν (ἐκχέουσιν L) ἐνώπιον Κυρίου ἐπὶ τὴν γῆν.

As for the kaige section of 2 Samuel, the Lucianic text at times—but not always— preserves the historical present of the original translator. In the case of the imperfect, the situation was the opposite: the Lucianic reviser increased their number (cf. 2 Sam 17:20 above). In both kinds of cases the change was motivated by a need to write good Greek narrative style. All in all, in the kaige section of 2 Samuel, part of the original historical presents have been preserved in the Lucianic text, and part of the imperfects of the Lucianic text represent the original translation. On the basis of his survey, Wirth came to the conclusion that the beginning of the kaige section was not at 2 Sam 11:2—as suggested by Thackeray and accepted by Barthélemy without question—but rather at 2 Sam 10:6.26 This is the point where the elimination of the historical presents begins, and this view is supported by other typical features of kaige after the turn and of the Old Greek before it.27 Wirth’s study serves as demonstration of what can be achieved by a translationtechnical study of syntactical features in the case of recensions.

8. Conclusions Soisalon-Soininen created his translation-technical methodology as a side-product of his dissertation while coping with the text-historical problems of Judges. The advantage of the study of syntactical features over lexical ones is that they are frequent and occur in all kinds of texts, whereas vocabulary to a certain extent is dependent on the subject matter of the text. The frequency of the cases to be studied helps to see the differences between translators as well as changes within texts much more sharply. Indeed, Soisalon-Soininen recognized that syntactical phenomena are good indicators of the quality of translation, and they also make it possible to compare different text-units with each other. As for the original translators of the Septuagint, most of them were not conscious of their own patterns of dealing with certain Hebrew structures and could also spontaneously vary their way of translation, if a better equivalent occurred to them. The later

26  See especially Wirth, “Dealing with Tenses in the Kaige Section of Samuel”. 27  There are three changes of the historical present to the aorist (1 Sam 10:6, 14, 17), three lexical changes (vv. 8, 16, 18), and two changes from perfect to the aorist (vv. 14, 19). The preceding verses (vv. 3, 5) contain typical Old Greek alternatives to expressions that were regularly changed in the kaige section.

Translation Technique and the Recensions

173

revisers, however, followed certain patterns of correcting the translation more consistently, thus revealing their overall attitude to translating scripture.28 What can be achieved by a study of the recensions, of course, depends on several factors. The first one is the original translator and the kind of translation that he left behind. If there were numerous false translations and/or extremely free renderings that departed from the Hebrew text, there was, of course, from the view-point of the revisers, more need for revision. Another factor that made revision necessary, at least in some books, is the Hebrew text, which was obviously still edited until quite late, so that comparison of the Greek text with the proto-MT revealed many differences. A further problem is how and why the early Jewish revisional readings were distributed into Christian manuscripts: in some cases, there is a small manuscript group, or perhaps a bigger one, showing fairly thorough correction according to the principles of the revision, in some other cases, there are just sporadic changes that can, however, be connected with the principles of the revision. Recognizing the principles, interests, and motives of the revisers is of crucial importance, but in the end, all we can do is interpret the traces left by the translators and their revisers and preserved by the manuscript tradition—and try to understand what happened.

28 The kaige revision was not only concerned with extreme literalism, but also with exegesis and theology. For the theological aspect, examples can certainly be found more easily in changes of vocabulary. See my “Does God Regret? A Theological Problem that Concerned the Kaige Revisers”, in A. Aejmelaeus/T. Kauhanen (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (De Septuaginta Investigationes 9; Göttingen: Vandenhoeck & Ruprecht, 2017) 41–53.

José Manuel Cañas Reíllo Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges 1. Introduction An important part of the legacy of Prof. Ilmari Soisalon-Soininen is his textual study on the Greek book of Judges. In it, questions involving vocabulary, recensions, and text types are fundamental. Since Soisalon-Soininen,1 the question of the double text of Judges, A and B, has been solved: the arguments of SoisalonSoininen made clear that the Greek texts of Judges preserved by the manuscript evidence go back to a single, not two translations. The greater part of A and B have the same text; the differences, though relevant, are not as important as previously thought. If we take into account other books, such as Samuel, Kings and Chronicles, we find a differentiation as striking as in Judges, or perhaps more. In these latter there is a kaige text very well delimited vis-à-vis an Antiochian text that could open the access to textual stages very close to the original, as reconstructed by Fernández Marcos and Busto Sáiz.2 Thus, Judges, Samuel, Kings, or Chronicles do not present the same problem as Esther or Daniel, where research does accept (at least for now) the duality of texts.

* Author’s note: This paper was written within the framework of the research project Edición y estudio de textos bíblicos y parabíblicos (FFI2017-86726-P) funded by the Ministerio de Economía y Competitividad (Government of Spain). I would like to thank the former Septuaginta-Unternehmen and the present Kommission zur Edition und Erforschung der Septuaginta of the Göttingen Academy of Sciences and Humanities for permission to use the collations of Greek manuscripts and bibliographical materials. I would also like to thank Kevin J. Zilverberg (University of Saint Thomas, Saint Paul, MN) for reviewing the English language of this work and for having improved it with his linguistic observations. 1  I. Soisalon-Soininen, Die Textformen der Septuaginta-Übersetzungen des Richterbuches (AASF, ser. B, 72.1; Helsinki: Druckerei-A.G. der Finnischen Literaturgesellschaft, 1951). 2  N. Fernández Marcos/J.R. Busto Sáiz, El texto antioqueno de la Biblia Griega. I. 1–2 Samuel (Textos y Estudios “Cardenal Cisneros” 50; Madrid: CSIC, 1989); N. Fernández Marcos/J.R. Busto Sáiz, El texto antioqueno de la Biblia Griega. II. 1–2 Reyes, (Textos y Estudios “Cardenal Cisneros” 53; Madrid: CSIC, 1992); N. Fernández Marcos/J.R. Busto Sáiz, El texto antioqueno de la Biblia Griega. III. 1–2 Crónicas (Textos y Estudios “Cardenal Cisneros” 60; Madrid: CSIC, 1996).

176

José Manuel Cañas Reíllo

Some years after the work of Soisalon-Soininen was published, the appearance of the work of Barthélemy on the kaige recension,3 although not directly related to Judges, brought light to the interpretation of the status and character of the B text in this book. In this matter, the vocabulary is the most important criterion. For J.A.L. Lee,4 the vocabulary of the Pentateuch shows that the text of MS A is older than the text of MS B, and such an observation is also valid for Joshua and Judges. In the framework of this article, vocabulary is the best way to detect revisions and deliberate variations in the text. Yet, vocabulary is not the only important field of study; other fields, such as translation technique, syntax, stylistics, and history of the language must be taken into account for the reconstruction of the text, but they are more unstable than vocabulary since they are very exposed to free changes throughout the history of transmission. In the most recent research,5 the view that dominates the study on the Greek Judges is that the A and B texts derive from a single original. As a result of this history of research since Soisalon-Soininen, there is a consensus in the classification of text types in Judges: 1. There is a B group witnessed by a large number of manuscripts headed by codex Vaticanus (B). 2. Besides B, there are other text types (printed under “A” in Rahlfs’ edition6): codex Alexandrinus (A), the Lucianic (L) and the Hexaplaric (O) groups, and a group of mixed texts headed by the uncials M and V.

3  D. Barthélemy, Les devanciers d’Aquila. Première publication intégrale du texte des fragments du Dodécaprophéton (VTSup 10; Leiden: Brill, 1963). 4  J.A.L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SCS 14; Chico, CA: Scholars Press, 1983). 5  For the history of research, see W.R. Bodine, The Greek Text of Judges: Recensional Developments (HSM 23; Ann Arbor, MI: Edwards Brothers, 1980), 1–4; J. Targarona Borrás, “Historia del texto griego del libro de los Jueces” (PhD diss., Universidad Complutense Madrid, 1983), 6–68. For the status quaestionis, see N. Fernández Marcos, Judges (BHQ 7; Stuttgart: Deutsche Bibelgesellschaft, 2011), 6*–9*; N. Fernández Marcos, “The B-Text of Judges: Kaige Revision and Beyond”, in H. Ausloss/B. Lemmelijn/J. Trebolle Barrera (ed.), After Qumran. Old and Modern Editions of the Biblical Texts. The Historical Books, (BETL 246; Leuven/Paris/ Walpole, MA: Peeters, 2012) 161–70, on pp. 161–9; N. Fernández Marcos, “Kritai/Iudices/Das Buch der Richter”, in S. Kreuzer (ed.), Einleitung in die Septuaginta, Handbuch zur Septuaginta/ Handbook of the Septuagint. LXX.H 1 (Gütersloh: Gütersloher Verlagshaus, 2016) 188–98; J.M. Cañas Reíllo, “LXX-Judges: The Value of Secondary Translations for Its Textual History”, in M. Meiser et al. (ed.), Die Septuaginta: Geschichte—Wirkung—Relevanz. 6. Internationale Fachtagung Wuppertal, 21.–24. Juli 2016 (WUNT 405; Tübingen: Mohr Siebeck, 2018) 230–43, on pp. 230–2. 6  A. Rahlfs, Septuaginta. Id est Vetus Testamentum graece iuxta LXX interpretes (Stuttgart: Deutsche Bibelgesellschaft, 2006).

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

177

2. Vocabulary and the Textual Groups: The B Group In 2013 I began the work of editing the Greek text of this book for the Septuaginta-Kommission of the Göttingen Academy of Sciences and Humanities. I had at my disposal a much wider and richer collection of material than previous researchers had.7 MSS and editions collated for Greek Judges are the following (* = fragmentary; C = manuscripts that contain catenae): Uncials: A B G* K* M S* V Z* Papyri: 876* 968* Minuscules: 15 16 18 19 29 30 44 46 52 53 54 55 56 57C 58 59 64 68 71 72 73C 74 75 76 77C 82 84 85 106 107 108 118* 120 121 122 125 126 127 128 129 130 131 134 236 246 313C 314 318 319 320C 321 328C 343 344 346 370 376 381 407 413C 414C 417C 422C 426 458* 488 489 500C 509* 527 528C 529C 530C 537 550C 551C 552C 610 616C 618 628C 630 646C 669 707 730C 739C 761C 799 932* 937* Editions: Ald (= Aldina) Comp (= Polyglotta Complutensis) Sixt (= Sixtina)

Having reviewed all the material available on the basis of the Göttingen collations and the status quaestionis, I am sure that it is possible to edit a single text of Judges, even for the most difficult chapters in this book, such as The Song of Deborah (Judg 5). For example, there is a critical edition of this text by N. LaMontage in his dissertation, defended at the Catholic University of America in Washington, D.C., in 2013.8 He uses the Lucianic text as his basis and the result is very interesting. This recension is the way to the original Greek text of Judges, but external criteria, such as the support of the Vetus Latina, are fundamental. Indeed, the most recent research holds that the Lucianic group, especially with the support of the Vetus Latina, should be taken as the key to the original Greek text of Judges, but the Lucianic text itself is not the original. For most of the text 7  For the numbering of Greek manuscripts, see A. Rahlfs, Verzeichnis der griechischen Handschriften des Alten Testaments. Für das Septuaginta-Unternehmen aufgestellt von Alfred Rahlfs (MSU 2; Berlin: Weidmannsche Buchhandlung, 1914), A. Rahlfs, Verzeichnis der griechischen Handschriften des Alten Testaments. Band I, Die Überlieferung bis zum VIII. Jahrhundert (bearbeitet von D. Fraenkel; Septuaginta. Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum, Supplementum I,1 (Göttingen: Vandenhoeck & Ruprecht, 2004), and the updated list of Greek manuscripts: Offizielles Verzeichnis der RahlfsSiegeln. Herausgegeben vom Septuaginta-Unternehmen der Akademie der Wissenschaften zu Göttingen (Göttingen 2012), online at http://hdl.handle.net/11858/00-001S-0000-0022-A30C-8. 8  N. LaMontagne, “The Song of Debora (Judges 5): Meaning and Poetry in the Septuagint” (PhD diss., The Catholic University of America, 2013).

178

José Manuel Cañas Reíllo

of Judges the Lucianic group is part of a large group whose text is found in Alexandrinus, the Hexaplaric group, and mixed texts like M and V. They constitute a group vis-à-vis the B revision, as the differentiation of vocabulary shows in the following selection of examples: MT

A O L mixed texts

B group

‫ פגע‬qal ‫לקראת‬ ‫ הלך‬qal

ἀπαντάω ἀπάντησις ἀπέρχομαι

συναντάω συνάντησις πορεύομαι

‫ שוב‬qal / hiphil ‫אלה‬ ‫ סור‬qal ‫לפנים‬ ‫בעיני‬ ‫ שלח‬qal / piel ‫ שום‬qal ‫ ראה‬qal ‫רעה‬ ‫ נתץ‬qal

ἀποστρέφω δρῦς ἐκνεύω ἔμπροσθεν ἔναντι / ἐναντίον ἐξαποστέλλω ἐπιτίθημι θεωρέω κακός / κακία κατασκάπτω

‫ עבד‬qal ‫חרב‬

λατρεύω μάχαιρα

ἐπιστρέφω τερέμινθος ἐκκλίνω τὸ πρότερον ἐνώπιον ἀποστέλλω τίθημι βλέπω πονηρία καθαιρέω / καταστρέφω δουλεύω ῥομφαία

‫ ראה‬qal ‫נער‬

ὁράω παιδάριον

‫ נתן‬qal

παραδίδωμι

βλέπω νεανίας / νεανίσκος δίδωμι

‫ בוא‬qal

παραγίνομαι

ἔρχομαι

‫ נכה‬hiphil ‫חתן‬ ‫ לחם‬niphal

πατάσσω πενθερός πολεμέω / ἐκπολεμέω

κόπτω γαμβρός παρατάσσω

‫שכר‬

σίκερα

μέθυσμα

8:21; 18:25 4:18; 6:35; 14:5; 15:14 1:26; 2:6; 6:21; 9:55; 19:7, 9, 10, 14, 27, 28; 21:21 7:3 (2 times); 9:56 6:11, 19 4:18 (3 times) 1:10, 11 3:12; 4:1; 6:1; 10:6; 13:1 6:35; 11:17; 19:29; 20:12 6:19; 9:24, 48; 16:3 13:19, 20 15:3; 20:3, 12, 13 2:2; 6:30, 31; 8:17 10:6, 10, 13 1:8, 25; 4:5, 16; 7:22; 8:20; 9:54 9:36; 19:30 17:7, 11: 18:3; 19:9, 13, 19 6:1; 7:7, 9; 12:3; 15:12; 16:23; 18:10 6:5; 9:31; 11:18; 13:9; 18:2, 7, 8 1:4, 5, 17 1:16; 19:9 1:3, 5; 8:1; 9:17, 38, 39, 45, 52; 10:9; 11:4, 6, 8, 9, 12, 25, 27, 32; 12:1, 3, 4 13:4, 7, 14

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

179

This picture of different vocabulary is much too simple, however. It is a communis opinio to consider the recension in B as very close to the kaige recension,9 but the research has shown that the revision of the vocabulary goes beyond the kaige recension. Fernández Marcos explains many of the phenomena involved in lexical changes in B Judges.10 The approach to the Masoretic Hebrew is the most important feature, though there are some instances where the revision is difficult to explain, as Fernández Marcos says, and some solutions could be found in the history of the Greek language and its evolution. Furthermore, not all the manuscripts of this group keep the peculiarities of the B revision. There are indications that part of the B text was successively retouched in syntax, and in some instances it is certain that B can preserve old, unrevised readings, especially when it has the support of the Vetus Latina and its vocabulary moves away from the lexical guidelines observed for this type of text. Since Soisalon-Soininen, it is accepted that the B revision is found in the following manuscripts: B 52 53 56 57 (72) 85 120 129 130 407 509

The vocabulary of the B revision and its evolution could be best known and best explained with the incorporation of the new manuscripts collated at the Septuaginta-Unternehmen, especially the fragments of Sinaiticus (S). A new image of the B text arose: B is not a revision but a set of revisional processes that sometimes are contradictory and do not affect the whole book. And in many cases the syntactical differentiation must be taken into account, especially for the use of prepositions. So, the new grouping of the B text is the following: B1 B2

B S 120 318 707 Sixt 53 85 130 321 346 56 129 246 319 407 417 489 509 57 343 413 551 761 73 320 16 44 46 131 236 313 328 530 52 77 74 422 500 646 76 370 414 528

9  As, for example, in M. Karrer, “The New Leaves of Sinaiticus Judges”, in S. Kreuzer/ M. Meiser/M. Sigismund (ed.), Die Septuaginta—Entstehung, Sprache, Geschichte. 3. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 22.–25. Juli 2010 (Tübingen: Mohr Siebeck, 2012) 600–17, on pp. 608–12, for the first hand of the Sinaiticus. 10  Fernández Marcos, “The B-Text of Judges”, 164–6.

180

José Manuel Cañas Reíllo

529 550 552 616 739 Mixed text (B/O): 72

The most important element is that B is divided into two main groups: B1 and B2. B1 is more consistent in vocabulary; it is a very compact group. The second one (B2) shows interference from other textual groups, such as the Hexaplaric and Lucianic, and perhaps has been affected by minor reworking. Thus, B is not as compact as it might seem looking at Rahlfs’ text: it is a confusing web of revisional tendencies that sometimes are contradictory. A type of B text was used in Egypt. It is found in papyrus 968 from Oxyrhynchus, and the Coptic translation had a Greek Vorlage of the B text type, as I concluded in a study on the matter.11 Considering the non-B texts, it is difficult to isolate the vocabulary of each recension, especially of the Hexaplaric and the Lucianic ones. For only a few words is it possible to isolate what is Hexaplaric and what is Lucianic and what could be the typical lexical variants for some revisional processes detected in the mixed texts.

3. The Hexaplaric Group For the Hexaplaric text (O), there is no well-defined vocabulary in Judges. Only a few Greek-Hebrew renderings are characteristic of this group:

7:11 9:26 15:14 17:1 17:3 17:8 18:17 19:12 19:13 20:28

MT

O

Reliqui

‫ חזק‬qal ‫ בוא‬qal ‫ מסס‬niphal ‫ לקח‬pual ‫ שוב‬hiphil ‫ הלך‬qal ‫ הלך‬qal ‫ עבר‬qal ‫ קרב‬qal ‫ חדל‬qal

ἐνισχύω ἔρχομαι διαλύω (with A) συλλαμβάνω (mss.) ἐπιστρέφω ἔρχομαι παραπορεύομαι πορεύομαι εἰσέρχομαι κοπάζω

ἰσχύω παρέρχομαι τήκω λαμβάνω ἀποδίδωμι παραγίνομαι πορεύομαι παρέρχομαι ἐγγίζω ἐπίσχω

11  J.M. Cañas Reíllo, “El papiro de Florencia, Bibl. Laur. PSI II 127 (Rahlfs 968): Su lugar en la historia textual del libro griego de Jueces y su relación con las versiones coptas”, in Τί ἡμῖν καὶ σοί; Lo que hay entre tú y nosotros. Estudios en honor de María Victoria Spottorno (Córdoba: UCOPress, 2016) 43–57, on pp. 56–7.

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

20:30 20:32 20:43 21:20

MT

O

Reliqui

‫ עלה‬qal ‫ נגף‬niphal ‫ רדף‬hiphil ‫ הלך‬qal

τάσσω προσκόπτω καταδιώκω ἀπέρχομαι

ἀναβαίνω πίπτω, τροπόω διώκω πορεύομαι

181

Most of the differences are related to motion verbs (πορεύομαι, ἔρχομαι, παραγίνομαι) and to the use of preverbs. A few examples are related to word variations, like εἰσέρχομαι / ἐγγίζω and ἐπιστρέφω / ἀποδίδωμι. The incorporation of new manuscripts brings a new approach to the study of this group, and the use of unique lexical variants is a relevant criterion to isolate it in the frame of the whole textual history of Judges. The new classification is the following: O1 (A) G 15 19 84 108 376 426 Compl O2 18 126 488 669 799 O3 128 630 O4 58 72 Mixed text: 876 and 937 (AO/L)

The largest Hexaplaric group, keeping the Origenian signs, is O1, while O4 is the least Hexaplaric. It shows interference from the Lucianic and the B groups.

4. The Lucianic Group Here we arrive at the key to the question: that is, the Lucianic text. There are many words that are found in Judges only in the manuscripts of the group considered Lucianic. I place them in the following list, in front of the vocabulary of the rest of the Greek evidence. The support of the Vetus Latina (VL)12 is added, 12 With “VL” I refer to the MS of Lyon, Bibliothèque de la Ville 403 [329] + 1964 [1840], number 100 at R. Gryson, Altlateinische Handschriften / Manuscripts Latins, VL 1/2A (Freiburg 1999) 159–60. Other evidence, such as the fragment of New Haven, Yale University, Beinecke Rare Book and Manuscript Library, Ms. 482/1A (nº 107 at R. Gryson, Altlateinische, 168) and patristic quotations are not taken into account for this study. For the text of the MS of Lyon, I use the edition by U. Robert, Heptateuchi partis posterioris versio latina antiquissima e codice Lugdunensi. Version latine du Deutéronome, de Josué et de Juges antérieure à Saint Jérôme (Lyon: Librairie de A. Rey et Cie, 1900). For the value of the Vetus Latina evidence for the reconstruction of the Old Greek of Judges, see Fernández Marcos, Judges, 9*–10*, and J. Trebolle, “The Textual History and the Text Critical Value of the Old Latin Version in the Book of Judges”, in

182

José Manuel Cañas Reíllo

and with an asteriscus I point out the lexical selections attested in the Antiochian (Ant) text of Samuel-Kings and Chronicles. For this comparison, the Índice Antioqueno published by the CSIC in 2005,13 in which I participated, serves as the basis. In the following examples, the Hebrew word that is the basis for the Greek readings, the reading of the Lucianic text (L) and those of the rest of the evidence (reliqui = rel.) are given. Because of the importance of the VL, sometimes it is given as support for Greek readings. Other abbreviations: Thdt = Theodoretus of Cyr, Quaestiones in Octateuchum;14 LXX* = the Old Greek. *2:1 ‫ עלה‬hiphil: ἀνάγω L (= VL), Thdt / ἀναβιβάζω rel. 2:3  ‫ אמר‬qal: φημί L / εἶπον rel. Ant uses φημί for ‫ נעם‬qal. Not with Thdt. 3:10 ‫ עזז‬qal: βαρύνω L (= VL) / κραταιόω rel. 4:11 ‫ פרד‬niphal: ἀφίστημι L, perhaps LXX*15 (= VL) / χωρίζω rel. 4:21 ‫ שים‬hiphil: λαμβάνω L (= VL) / τίθημι rel. *6:16 ‫ נכה‬hiphil: ἀποκτείνω L (= 1 Sam 17:46), with Thdt / πατάσσω rel. (= VL). 6:31 ‫ עמד‬qal: συνίστημι L / ἐπανίστημι rel. 6:31  ‫ ריב‬qal: ἀδικέω L (= VL) / δικάζω, ἀντιδικάζω rel. Ant uses ἀδικέω for ‫ מעל‬qal and ‫ עוה‬hiphil; ‫ ריב‬qal is the basis for κρίνω. The translator of Greek Judges could have read ‫ ריב‬hiphil, ἀντίδικος in Ant. 6:34  ‫ זעק‬qal: καλέω L (= VL) / βοάω rel. In Ant καλέω is translation of ‫ קרא‬qal. *7:5  ‫ ירד‬hiphil: κατάγω L / καταφέρω, καταβιβάζω rel. *7:8, 16, 18, 19, 20, 22 ‫שופר‬: σάλπιγξ VL (= tuba VL 7:16, 18–20, 22) with ThdtCommentary / κερατίνη rel. (= cornea VL 7:8). 7:20 ‫ שבר‬qal: ἀπορρίπτω L (= VL) / συντρίβω rel. 7:21  ‫ רוץ‬qal: συντρέχω L / τρέχω rel. Ant translates the Hebrew word with ἐξέρχομαι, παρατρέχω, τρέχω and φεύγω. *7:22 ‫ שים‬qal: ἐμβάλλω L / τίθημι rel. (= VL). W. Kraus et al. (ed.), Die Septuaginta—Text, Wirkung, Rezeption. 4. Internationale Fachtagung veranstaltet von Septuaginta Deutsch /LXX.D), Wuppertal 19.–22. Juli 2012 (Tübingen: Mohr Siebeck, 2014) 53–72. 13  N. Fernández Marcos/M.V. Spottorno Díaz-Caro/J.M. Cañas Reíllo, Índice griego-hebreo del texto antioqueno en los libros históricos. (2 vols.; Textos y Estudios “Cardenal Cisneros” 75; Madrid: CSIC, 2005). 14  N. Fernández Marcos/A. Sáenz-Badillos, Theodoreti Cyrensis Quaestiones in Octateuchum. Editio critica (Textos y Estudios “Cardenal Cisneros” 17; Madrid: CSIC, 1979). References to Thdt correspond to the lemma; when corresponding to commentary “Thdt-Commentary” is used. For this commentary and its importance for the textual history of the Greek Bible, see N. Fernández Marcos, “Theodoret’s Biblical Text in the Octateuch”, BIOSCS 11 (1978) 27–43. 15  Fernández Marcos, Judges, 12.

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

183

8:9  ‫ אמר‬qal: ὀμνύω L (= VL) / εἶπον L. In Ant ὀμνύω is translation of ‫ שבע‬niphal and hiphil. 8:15 ‫כף‬: κεφαλή L (= VL) / χείρ rel. 9:4  ‫פחז‬: ἐπανιστάμενος L (= VL?) / δειλός, θαμβούμενος rel. *9:19 ‫תמים‬: ὁσιότης L (= VL) / τελειότης rel. *9:26 ‫בעל‬: ἄρχων L (= VL), and in Ant (2 Sam 6:2) / ἀνήρ rel. 9:28 ‫פקיד‬: τετάγμενος L / ἐπίσκοπος rel. (= VL). 9:33 ‫ שכם‬hiphil: ἀνίστημι L / ὀρθρίζω rel. (= VL). 9:33 ‫ מצא‬qal: δύναμαι L (perhaps for ‫ )?עצר‬/ εὑρίσκω rel. (= VL). *9:45 ‫ נתץ‬qal: κατασκάπτω L (= VL) / καθαιρέω rel. *11:11 ‫ראש‬: ἄρχων L (= VL) / κεφαλή, ἡγούμενος rel. But in 11:8, 9: κεφαλή (AOL mixti) / ἄρχων (B). *11:26 ‫ ישב‬qal: κατοικέω L / οἰκέω rel. 11:26  ‫בת‬: περίοικος L / ὅριον rel. (= VL). In 1 Kings 7:33 Ant uses περίοικος for ‫ככר‬. *11:33 ‫ כנע‬niphal: ταπεινόω L (= VL) / συστέλλω, ἐντρέπω rel. 12:1 ‫ צעק‬niphal: συναθροίζω L / συνάγω rel. 13:16  ‫ עצר‬qal: παραβιάζομαι L, with Thdt / κατέχω, βιάζομαι rel. Παραβιάζομαι is translation for ‫ פצר‬qal in 2 Kings 2:17; 5:16, and for ‫פרץ‬ qal in 1 Sam 28:23. 14:10 ‫משתה‬: δοχή L / ποτόν or πότος rel. 14:14  ‫עז‬: σκληρός L (but πικρός in ms. 75) / ἰσχυρός rel. In Ant σκληρός is translation of ‫ הרה‬and ‫קשה‬. 15:4  ‫ מסה‬hiphil: διαρρήγνυμι L / τήκω, διαλύω rel. Ant has διαρρήγνυμι for ‫ בקע‬and ‫קרע‬. In 16:9 διαρρήγνυμι is the word of the group AOL and some mixed texts for the Hebrew verb ‫ נתק‬piel; the B group has διασπάω. A few Greek mss. (55 82 121 125 314) have διαρρίπτω. 15:8  ‫ ישב‬qal: παροικέω L / κατοικέω rel. In Ant παροικέω translates the Hebrew ‫ גור‬qal. Πάροικος in Ant is translation of ‫ גור‬qal, ‫ גר‬and of ‫( תושב‬from ‫ )ישב‬in 1 Chron 29:15(2º). 16:3 ‫ אחז‬qal: λαμβάνω L / ἐπιλαμβάνω rel. 16:9  ‫פתיל‬: σπαρτίον L (= VL) / κλῶσμα rel. The Ηebrew word is not attested in Samuel-Kings. Σπαρτίον in Ant translates ‫( חוט‬1 Kings 7:3) and ‫( קו‬2 Chron 4:2). 16:13  ‫ התל‬piel: καταμωκάομαι L / πλανάω, παραλογίζομαι rel. (= VL). Ant uses καταμωκάομαι for ‫ לעג‬hiphil in 2 Chron 30:10, doublet with μυκτηρίζω. 16:13  ‫מחלפות‬: βόστρυχος L / σειρά rel. In 16:14 βόστρυχος is found in the AOL group. These Greek words are not attested Samuel and Kings. 16:22  ‫ צמח‬piel (“to grow”) or qal (“to germinate”): ἀναβαίνω L (only in the Greek mss. 54 and 75) / βλαστάνω, ἀνατέλλω rel. (= VL). The

184

José Manuel Cañas Reíllo

Greek word ἀναβαίνω is not attested in Ant; but ἀνατέλλω in 2 Sam 10:5 and 1 Chron 19:5. 17:4  ‫ שוב‬qal: δίδωμι L / ἀποδίδωμι rel. (= VL). Δίδωμι as translation of ‫ שוב‬qal is not attested in Ant, but it can be found as translation of ‫ שים‬qal in 1 Sam 22:15 and 2 Chron 23:15. *17:7, 10 ‫היה‬: εἰμί L (= VL) / γίνομαι rel. The equivalence ‫היה‬/εἰμί is very common in Ant and is attested in Thdt (17:7). 17:8  ‫ בוא‬qal: γίνομαι L / ἔρχομαι, παραγίνομαι rel. (= VL). Παραγίνομαι is common in Ant for ‫ בוא‬qal. 17:11  ‫ מלא‬piel: εὐδοκέω L / πορεύομαι rel. (= VL). In Ant εὐδοκέω is translation of ‫חפץ‬, ‫ חצל‬and ‫רצה‬. *18:2  ‫ לין‬qal: καταλύω L, perhaps LXX*?16 (= VL, or from καταπαύω?) / αὐλίζομαι, καταπαύω rel. The equivalence ‫ לין‬qal / καταλύω is found in Ant 1 Kings 19:9. *18:21 ‫ שים‬qal: τάσσω L (= VL) / τίθημι, πατάσσω rel. The equivalence ‫ שים‬qal / τάσσω is found in Ant 1 Sam 22:7 and 2 Kings 10:24.

On this list the following observations can be made: 1. There is an important agreement between the Lucianic text and the VL. This could help us detect the proto-Lucianic stratum in this text and, with it, to approach a stage of the Greek text very close to the original, free from the Hexaplaric recension and the B revision. This support of the VL is similar for words attested in the Antiochian text of Samuel, Kings, and Chronicles, and for those not attested, but can in any case belong to an early stage of the text. In the same way, there are Lucianic words without support of the VL in both groups. Perhaps these words belong to a revised Lucianic stage. But it must be taken into account that the VL does not always preserve a good text, or it sometimes supports a B text. 2. In these examples, only 14 (verse numbers marked with *) contain lexical selections common to the Antiochian text in Samuel, Kings and Chronicles. Another question arises: In Samuel, Kings, and Chronicles the name “Antiochian” is applied to the Lucianic group. In recent times, it is usual to find the use of “Lucianic” and “Antiochian” as synonyms for the group in question.17 For the moment I prefer to be cautious and keep the name “Lucianic” for this group in Judges until we have more criteria that allow us to apply the name “Antiochian” 16  Fernández Marcos, Judges, 101*. 17  For example, in S. Kreuzer, “Der Antiochenische Text der Septuaginta in seiner Bezeugung und Bedeutung”, in S. Kreuzer/M. Sigismund (ed.), Der Antiochenische Text der Septuaginta in seiner Bezeugung und seiner Bedeutung (De Septuaginta Investigationes 4; Göttingen: Vandenhoeck & Ruprecht, 2013) 23–56.

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

185

to this textual group with the same value that is found, for example, in Samuel, Kings, and Chronicles: that is to say, considering this Lucianic text in Judges as an Antiochian text born of editorial activity that could have been begun in the first century A.D. in the area of Antioch. This text would have very quickly arrived in the Roman milieu, where it served as the Vorlage for the translation of the VL. We do not have data that allow us to make such an identification. The few readings of Theodoret’s Quaestiones in Octateuchum that exist for Judges do not always support this group, so their testimony is not decisive when establishing a relationship.18 In addition, for Judges we do not have the testimony of Antiochian Fathers, like Asterius, Diodorus of Tarsus, Eustathius, and Severus of Antioch. Regarding the VL, its testimony is not unitary: The VL text of the manuscript of Lyon has agreements with the Lucianic text in Judges, but often goes with other texts, such as B or some mixed manuscripts. The text-critical use of the VL in Judges therefore requires a certain caution. I believe that it is premature to speak of an “Antiochian recension”, since Lucian and Antioch are not the same realities, and in Judges there are not enough criteria to accept an Antiochian revision similar to that found in Samuel, Kings, and Chronicles. I therefore prefer to remain cautious and to continue speaking of Lucianic, or proto-Lucianic, whose best evidence the traditional research recognized in MSS K Z 54 59 75 and 314 in the first Lucianic group, and in 106 134 and 344 in the second group, with sporadic support of 44. The unique readings help us know the manuscript tradition of this group better because of the incorporation of new manuscript evidence into the collation. At the same time it is possible to do a new classification of the Lucianic manuscripts based on the vocabulary and other criteria, which for the moment is provisional: L1 K Z 54 59 75 (82) 314 (932?) L2 118 127 458 537 L3 (44) 106 134 344 L4 30 730 Mixed texts: 876 and 937 (L/AO)

The L1 group very faithfully preserves the Lucianic text, although the manuscripts which could be their best witnesses, K and Z, are very fragmentary and have no text for the whole book of Judges. Even so, the remaining four manuscripts are a sufficient basis. The four newly collated manuscripts that we grouped in L2 allow us to find a text very close to L1, as indicated by the close relation of 458 to 54 and to 75, and of 537 to 314. In many cases in which Lucianic readings are not preserved in the L1 group for various reasons, the L2 group allows one to restore it, although 18  See Fernández Marcos, “Theodoret’s Biblical”, 39.

186

José Manuel Cañas Reíllo

it is necessary to take into account the limitations presented by the fragmentary nature of two of its manuscripts: 118 and 458. Among the newly incorporated manuscripts, 127 (= c2 in Brooke McLean: Moscow, Syn. Library, Gr. 31) was used by Fernández Marcos and Busto Sáiz in their edition of the Antiochian text of the historical books as containing good evidence for this text. These authors tell us that it is a manuscript highly esteemed for its numerous marginal notes of Hexaplaric content, and that it is also one of the most faithful mss. in providing evidence of the diacritical signs of the Hexapla; it was used by Field for this purpose.19 Regarding groups L1 and L3, the status of 82 and 44 is very unstable. We have been able to detect that both generally transmit a mixed Lucianic/B text. The fourth group frequently supports L1 or L2 when it does not follow the B text, although this is very common.

5. The Mixed Group We also have to deal briefly with the mixed texts, an undefined group headed by the uncials M and V that is mostly aligned with the group AOL, although it sometimes supports the B text. Since Soisalon-Soininen, the manuscripts M V 29 55 and 121 are considered mixed texts (M). They have an undefined place between the B text and the AOL, and share coincidences of both groups. The collation of new manuscripts enriches this picture: M1 M2 M3

M V 29 55 64 381 618 628 Ald 68 121 122 71 107 125 527 610 968

Its close relationship to the uncial A, the Hexaplaric, and the Lucianic texts supposes that M can preserve old stages of the text. It is possible that, in the case of corrupted texts, whether Hexaplaric or Lucianic, the mixed manuscripts could be the only key to the old text, but external support, such as the Vetus Latina, is necessary. But there is an additional problem in this group: in contrast to the Hexaplaric and Lucianic groups, it has great intrinsic diversity. Thus, it is possible to isolate a subgroup in mss. 107 125 610, sometimes with ms. 44 (which in many instances preserves Lucianic readings). This group shows many omissions that can be explained by homoioteleuton; its text is the worst in Judges. The same trait is found in Samuel and Kings.20 19  Fernández Marcos/Busto-Sáiz, El texto antioqueno de la Biblia Griega. I. 1–2 Samuel, XIX. 20  For example, in Samuel, these manuscripts (with the exception of 527), are part of the so-called d group (classification of manuscripts by Anneli Aejmelaeus). See R. Wirth, Die Sep-

Recensions, Textual Groups, and Vocabulary Differentiation in LXX-Judges

187

Finally, the last question concerns the catena manuscripts. In most editions of the Septuaginta-Unternehmen the editors have established a catena group. In the case of Judges, I have not been able to isolate such a group; accordingly, the unique word variants of the catena manuscripts are included mostly in the B group: B1 417 73 320 57 413 551 761 B2 313 328 530 77 422 500 646 414 528 529 550 552 616 739

The catena-manuscript 730 has a mixture of Lucianic and B texts, and 628 has an M text.

6. Conclusions On the basis of the above considerations, the following conclusions can be made: 1. With the available materials, I judge that it is possible to reconstruct the textual history of Judges going back to a single text, which, if not the original, is very close to it. 2. Unique readings are the fundamental criterion. They are more reliable than syntactic changes, which are more exposed to stylistic touch-ups. 3. The closest vocabulary to the original should be sought in the AOL and mixed groups. Still, B could sometimes preserve unrevised sections older than AOL. 4. Within AOL, the Lucianic group is the safest way to the old stages, but not always, for it is necessary to distinguish the proto-Lucianic and Lucianic stages. 5. Currently there is no certainty that the Lucianic text in Judges can be placed at the same level as the Antiochian text in Samuel, Kings, and Chronicles. 6. It is necessary to give importance to the mixed group, since it could sometimes preserve old readings along with the uncial A, the Hexaplaric group and the Lucianic group. tuaginta der Samuelbücher. Untersucht unter Einbeziehung ihrer Rezensionen (De Septuaginta Investigationes 7; Göttingen: Vandenhoeck & Ruprecht, 2016), 28.

188

José Manuel Cañas Reíllo

7. I cannot isolate a catena group in Judges, as in other biblical books published in the Septuaginta-Unternehmen. Most of the catena manuscripts transmit the B text type. It can be concluded that the path to the Old Greek is clear, yet difficult!

William A. Ross Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt 1. Introduction Like most languages, post-classical Greek had no shortage of words that were used to refer to males and females in terms of their stage of life.1 Despite the attempts of some ancient authors at describing the numerical ranges and sequence of such stages—known as the ἡλικίαι—there was apparently no universal system in either the classical or post-classical periods.2 Modern lexicographers thus face the unenviable task of attempting to draw distinctions between, for example, παῖς and παιδίον, or νεός and νεανίας, and often do so with less than satisfying results. To describe the meaning of these words is to face the challenges of overlapping semantic ranges and scarce attestation in surviving sources. Add to this situation the fact that the use of these words was informed, not so much by numerical values associated with a date of birth, but rather perceived age and sociocultural factors no longer fully understood.3 Such pragmatic factors were manifested by an individual’s position in society, for instance, “in the order of the family, in the resulting distribution of economic resources, and in the distribution of power within the political system.”4 As with all analysis of ancient Greek vocabulary, understanding the social context of the evidence for a lexical item is paramount to understanding its use and meaning. For example, ancient Greek artwork actually illustrates how slavery was socially analogous to “a state of permanent childhood” in that “when a slave iconography develops it involves small size and youth as marks of lower status.5 1  It is a privilege to offer this essay in honour of the scholarly legacy of Prof. I. SoisalonSoininen, whose work in Greek Judges has informed the preliminary research for this lexical study. 2  C.A. Forbes, Neoi: A Contribution to the Study of Greek Associations (Middletown: The American Philological Association, 1933), 2; M. Golden, Childhood in Classical Athens (Baltimore: Johns Hopkins University Press, 1990), 12–16. For ancient authors, see Xenophon, Cyr. 1.2.4, 1.5.4, 8.7.6; Aristotle, Ath. pol. 42.1–2; Plutarch, Lyc. 21; Mor. 238a–b, 544e; Philo, Opif. 105; Aristophanes of Byzantium, Περὶ ὀνομασίας ἡλικιῶν 37–66, etc. 3  It is unlikely that many in the ancient world knew their precise birthdate or numerical age, and regardless there are complexities involved with determining the accuracy of age when it is mentioned in ancient sources. See T.G. Parkin, Old Age in the Roman World: A Cultural and Social History (Baltimore: Johns Hopkins University Press, 2003), 26–35. 4  J. Timmer, “Age”, in S. Bagnall et al. (ed.), The Encyclopedia of Ancient History (Oxford: Blackwell, 2013) 173–8, on p. 174. 5  S. Lewis, The Athenian Woman: An Iconographic Handbook (London: Routledge, 2002), 83. She continues, “This could be a reason for the painters’ reluctance to portray children

190

William A. Ross

From this vantagepoint, the polysemy of παῖς—meaning both ‘child’ and ‘slave’— is more easily comprehensible to the modern reader. The objective of this essay is to gain a better understanding of the social context of the evidence for the lexical items παιδάριον and νεανίσκος, in order that their meaning and use within the Septuagint corpus will likewise become more fully understood. Given that the translations of many books of the Septuagint derive from a Ptolemaic Egyptian context, the corresponding documentary evidence in papyri and inscriptions will receive the bulk of attention.6 The analysis undertaken here of παιδάριον and νεανίσκος in these sources helps explain their use in the Septuagint, as illustrated by case studies in Ruth and Ezekiel for each word. Moreover, this lexical study not only provides further evidence in support of the idea that these books were translated within an Egyptian milieu, but also demonstrates the linguistic subtlety and skill of the translators.

2. The Παιδάριον as a Semiskilled Labourer in Ptolemaic Egypt To begin with distributional observations on παιδάριον, its attestation in sources up to the end of the 2nd century c.e. is heavily skewed towards documentary evidence (Table 1).7

­ orking alongside [slave] adults, since it would render their status ambiguous, and markers w to distinguish between free and slave children are virtually impossible to find.” Similarly, the fact that “boys, slaves and pathics” were all addressed as παῖς indicates “how closely their social identities were conflated”; R. Garland, The Greek Way of Life: From Conception to Old Age (London: Duckworth, 1990), 106. 6  The analysis presented here is based upon a more in-depth study of all extant Greek sources, focusing on those dated between the 3rd century b.c.e. and the 2nd century c.e. That investigation included evidence normally classified as literature, along with the nonliterary sources in all their linguistic variety found in papyri and inscriptions. See William A. Ross, “Septuagint Lexicography and Language Change in Greek Judges” (PhD diss, University of Cambridge, 2018) 111–57. 7  All statistics for literature derive from manual counts of searches performed with the TLG database and include only sources that can be confidently dated or attributed. Figures exclude citations from spurious or fragmentary works, along with testimonia, scholia, the Septuagint (παιδάριον 234×; νεανίσκος 110×), and the New Testament (παιδάριον 1×; νεανίσκος 11×). Papyri were examined by cross-checking http://papyri.info/ and https://www.trismegistos.org/ words/, while inscriptions were examined using https://epigraphy.packhum.org/ and the online database for Supplementum Epigraphicum Graecum.

191

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

Table 1: Distribution of παιδάριον

Literature Papyri Inscriptions8 Total

5th

4th

3rd

2nd

1st

1st c.e.

2nd c.e.

14 – – 14

31 – – 31

7 100 1 108

7 39 73 119

1 14 16 31

62 15 7 84

81 44 9 134

From the beginning of the 3rd through the end of the 1st centuries b.c.e., παιδάριον appears just fifteen times in literary sources. In contrast, in papyri and inscriptions dating to the same time frame, παιδάριον is attested well over two hundred times. While allowing for possible misrepresentation owing to chance survival of material evidence, this trend strongly suggests that παιδάριον was more commonly used within non-literary varieties of post-classical Greek. Examining the documentary evidence, we find that παιδάριον is most frequently used to refer to paid members of staff who undertook semi-skilled labour on an estate, and who were in a position of economic dependence and lower social status. Some examples can illustrate the typical social context of a παιδάριον. In one 3rd century b.c.e. papyrus, a certain Jew named Toubias writes to inform Apollonius that he is sending several males to him:  ουβίας Ἀπ̣[ολλωνίωι χαίρειν]. εἰ σύ τε ἔρρωσαι καὶ τὰ σὰ πάντα | καὶ τὰ λοιΤ πά σο[ι κατὰ νοῦν ἐστὶν, πο]λλὴ χάρις τοῖς θεοῖς· καὶ αὐτὸς δὲ | ὑγίαινον, σοῦ διὰ π[αντὸς μνείαν ποι]ούμενος, ὥσπερ δίκαιον ἦν. ἀπέσταλ|κά σοι ἄγοντα Αἰνέ̣[αν εὐνοῦχον ἕ]να καὶ παιδά[ρια οἰκε]τικά τε | καὶ τῶν εὐγενῶν τέσσαρα, ὦν [ἐστὶν] ἀ̣περίτμητα δύο. ὑπογεγράφαμεν | δέ σοι καὶ τὰς εἰκόνας ⟦αὐ⟧τῶν π̣[αιδαρ]ί̣ω̣ν ἵνα εἰδῆις.| ἔρρωσο. (ἔτους) κθ, Ξανδικοῦ ι. Toubias to Apollonius, greetings. If both you are well and all your dealings and the rest of your affairs are in order, many thanks to the gods. I too am in health, always keeping you in mind, as is right. I have sent you Aineias bringing one eunuch and four παιδάρια, who are both locals and well built, of whom two are uncircumcised. And we have attached for you also the descriptions of the παιδάρια so you know. Be well. (Year) 29, Xandikos 10 (P. Cair. Zen. 1.59076 [TM 731]).9 (1)

8  In addition to those inscriptions included in the tabulation, there are a further eighteen where παιδάριον appears in an undated source. 9 Translation adapted from J.L. White, Light from Ancient Letters (Philadelphia: Fortress, 1986), 39–40. Other translations by the author unless otherwise mentioned. Note that “TM [#]” refers to the stable identifier of a source in the Trismegistos online database:

192

William A. Ross

The individuals in (1) are in some capacity at the disposal of Toubias, who transfers their circumstances to the control of Apollonius. The four παιδάρια are subsequently described in some detail—so that Apollonius will know he has received the correct individuals—with age estimates between seven and ten years old. Caution is needed here against making this numerical age range prescriptive for the meaning of the lexical item παιδάριον. As noted in the introduction, it is both the (perception of) age as well as an individual’s social status that are the most significant aspects of how a language user might have categorized someone in such terms. The seven- to ten-year-olds described in example (1) may have been considered and called παιδάρια well into their teenage years, perhaps beyond, if their social status remained unchanged.10 Other sources provide further clarity regarding the ways in which a typical παιδάριον was dependent upon and at the disposal of an agent or estate for their welfare and assignments. The examples below are drawn from two other 3rd century Zenon papyri from Egypt: (2)  ὑπόμνημα Ζήνωνι πα|ρὰ Ἀλκίμου. καλῶς ποή|σεις τὰ παιδάριά μοι ἀπο|δούς, καθότι διωμολό|γητο ἡμεῖν· τὰ γὰρ ἔργα | μοι ἐνέστηκε. περὶ δὲ | τοῦ ἱματισμοῦ ἐπίκρινον | αὐτὸς ὅσον δεῖ δοθῆναι, | θεωρήσας καὶ εἰς τὸ ἐμό\ν/. Memorandum to Zenon from Alcimus. Would you please send me the παιδάρια, as we agreed; for my labours have come to a halt. And about the clothing, decide yourself however much is necessary to provide, seeing also to mine. (P. Cair Zen. 3.59378 [TM 1021], lines 1–9) (3)  Ζήνωνι χαίρειν Πετοβάσ|τις περιστεροτρόφος. καλῶς | ἂν ποιήσαις, εἰ καί σοι δοκεῖ, | συν̣ τάξας δοθῆναί μοι τὸ | ὀψώνιον· στενῶς γὰρ διά|κειμαι· ὀφείλεται δέ μοι τὸ | ὀψώνιον τετραμήνου. | καὶ εἰς τὴν σιτομετρίαν ἔχω | κριθόπυρα ἀχρεῖα· οὐ δυνά|μεθα καταχρήσασθαι. | χρείας παρέχομαι καὶ τὰ παιδά|ριά http://www.trismegistos.org/. See M. Depauw and T. Gheldof, “Trismegistos: An Interdisciplinary Platform for Ancient World Texts and Related Information,” in Ł. Bolikowski et al. (ed.), Theory and Practice of Digital Libraries—TPDL 2013 Selected Workshops (CCIS 416; Cham: Springer, 2014) 40–52. My translation ‘locals’ rather than ‘home-bred’ or an equivalent is based upon the frequent use of οἰκογενῆ in the Delphic manumission inscriptions. In these sources, seventeen percent of the individuals manumitted are called a παιδάριον. See K. Hopkins, Sociological Studies in Roman History (2 vols.; Cambridge: Cambridge University Press, 1978), 1:139 and Table III.1 on p. 40. Many of these παιδάρια are described as either “home born” (οἰκογενῆ) or “foreign born” (ἐνδογενῆ), so that the οἰκο- affix does not constrain personal origins to the household or estate, but rather the region/country, however construed. 10  G.S. McGregor notes that, while in classical sources age may have been a more prominent aspect of the meaning of παιδάριον, in the post-classical period the word was often used to describe an individual’s “status as a slave or servant, irrespective of his age”; “A Semantic Study of Words for Young Person, Servant and Child in the Septuagint and Other Early Koine Greek” (M.A. Thesis; University of Sydney, 1976) 102–3, quote on p. 103.

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

193

μου. δέομαι οὖν σου, | εἰ καί σοι δοκεῖ, \συντάξας/ δοθῆναί μοι τὴν | σιτομετρίαν καὶ τὸ ὀψώνιον | εὐκαίρως, ἵνα σοι τὰς χρείας παρέσ|χωμαι. | εὐτύχει. Petobastis the pigeon-keeper to Zenon, greetings. Would you please, if you are so minded, arrange for wages to be given to me, as I am in a tight fix. Four months of wages are owed to me. And I have lousy barley-wheat for rations; we cannot be depleted. I have needs and so do my παιδάρια. So I request of you, if you are so minded, arrange rations and wages to be given to me promptly, so I can provide for your needs. Be well. (P. Cair. Zen. 3.59498 [TM 1136])

In (2) and (3), παιδάρια are again under the authority of their superior, who puts them to work within a labour force. However, the παιδάρια in view are not chattel, but rather receive remuneration and support in the form of clothing, foodstuffs, and possibly monetary payment.11 Some three centuries later, the lexical item παιδάριον is still attested in a similar way in documents from Egypt, though with a spelling variant. In example (4), an Alexandrian merchant named Sarapion writes to Herakleides, one of his staff members who evidently made a bad business deal with a certain Ptollarion. Sarapion, who is elsewhere described as an ἔμπορος or ‘retail merchant’ (BGU 4.1078 [TM 9455]), instructs the following course of action: (4)  ἀκολούθει δὲ Πτολλ|αρίωνι πᾶσαν ὥραν· τά|χα δύναταί σε εὔλυτ|ον ποισαι· λέγε | αὐτῷ· ἄ|λλο ἐγώ, ἄλλο πάντες, | ἐγὼ παιδάριν εἰμί. παρὰ | τάλαντόν σοι πέπρακα | τὰ φο[ρτ]ία μου· οὐκ οἶδα | τί μ[ε ὁ] πάτρων ποισ̣ε̣ι,̣ | πολλοὺς δανιστὰς ἔχο|μεν. μὴ ἵνα ἀναστατώ|σῃς ἡμᾶς, | ἐρώτα αὐτὸ\ν/ | καθʼ ἡμέραν· Now, follow Ptollarion all the time; maybe he is able to resolve your problem. Tell him, “It is one thing for me, another for everyone else, I am a παιδάρι[ο]ν. I have sold you my merchandise for a talent too little. I do not know what my patron will do, as we have many creditors. Do not drive us out of business!” Ask him every day. (BGU 4.1079 [TM 9456], lines 10–22; 1st c. c.e.)12

Here we find Sarapion speaking to Ptollarion through the mouth of Herakleides, who Sarapion considers a παιδάριον. As such, Herakleides is apparently exposed 11  See also P. Cair. Zen. 3.59509 (TM 1147), lines 12–14. Many labourers in the Hellenistic period received payment of some kind, making a binary opposition between free and slave impossible. It is more accurate to speak of “varying degrees of unfreedom”; D.J. Thompson, “Slavery in the Hellenistic World”, in Keith Bradley and Paul Cartledge (ed.), The Cambridge World History of Slavery (Cambridge: Cambridge University Press) 194–213, on p. 195. Likewise, E.N. Gardiner states that “[f]orced labor was part of Egyptian tradition… but forced labor is not slavery”; “A School in Ptolemaic Egypt”, TCR 44 (1930) 211–13, on p. 212. 12  Translation adapted from White, Ancient Letters, 129–30.

194

William A. Ross

to commercial risk that others are not, indicating his lower social status. Herakleides has also been dispatched by his superior like the other παιδάρια in examples (1) and (2). Yet example (4) more clearly indicates the shared interests of this παιδάριον with his patron (πάτρων, line 18), on whose behalf Herakleides acts as a business agent. Indeed, Sarapion himself presumes their mutual concern for the fortunes of their trade (N.B. ἡμᾶς, line 21). Space prevents further exploration of relevant sources in equal detail. However, a summary of a considerable amount of remaining documentary evidence is possible. In general, the lexical item παιδάριον appears most frequently in administrative documents—many of which fragmentary—dealing with various forms of estate labour, payment of wages, and supplies acquisition.13 Although in some sources it is not made explicit, the evidence strongly suggests that παιδάρια were ordinarily remunerated for their work, typically in the form of food, the quantity of which was determined by the gender, age, and position of the labourer.14 In terms of the actual labour conducted, in papyri we find παιδάρια engaged in animal husbandry.15 They also participated in seasonal tasks such as weeding wheat and flax, tending castor trees, and in cultivating various crops like hemp and olives.16 The semiskilled labour common to παιδάρια was not limited to agriculture, however, as they are also found at work on Apollonius’s ships. There παιδάρια received the same payment in both kind and quantity as the ναυτικοί who operated the vessel.17 In several papyri preserving long records of daily expenses, παιδάρια are listed alongside ἐργάται, who are known to have been temporary day-labourers.18 Yet other sources demonstrate that a promising 13  E.g., BGU 6.1290 (TM 4555), 3rd/2nd c.; BGU 7.1512 (TM 4762), 3rd/2nd c.; BGU 7.1518 (TM 4768), 2nd c.; BGU 7.5121 (TM 4771), 3rd/2nd c.; BGU 7.1526 (TM 4776), 3rd/2nd c; BGU 7.1527 (TM 4777), 3rd/2nd c; BGU 7.1531 (TM 4781), 3rd/2nd c; O.Strasb. 1.603 (TM 76187), 3rd c.; P. Cair. Zen. 2.59292 (TM 936), 3rd c.; P. Mich. 1.49 (TM 1949), 3rd c.; P. Cair. Zen. 3.59435 (TM 1075), 3rd c.; P. Cair. Zen. 4.59698 (TM 1325), 3rd c.; P. Cair. Zen. 2.59292 (ΤΜ 936), 3rd c.; P.Hamb. 1.114 (TM 2366), 3rd c., etc. 14  See the study by T. Reekmans, La Sitométrie Dans Les Archives De Zénon (Pap.Brux. 3; Brussels: Fondation épygtologique Reine Elisabeth, 1966). 15  P. Cair. Zen. 3.59406 (TM 1048), 3rd c.; P. Cair. Zen. 2.59195 (TM 841), 3rd c. In the latter papyrus, Apollonius dispatched παιδάρια from the estate specifically to learn that trade for themselves (ὅπως μανθάνωσ[ιν], line 8). See I. Bieżuńska-Małowist, L’esclavage Dans L’egypte Gréco-Romain (2 vols.; Archiwum filologiczne 30; Wrocław: Polska Akademia Nauk, Komitet Nauk o Kulturze Antycznej, 1974) 1:63. 16 See P. Lond. 7.2164 (TM 1724), 3rd c.; Bieżuńska-Małowist, L’esclavage, 62–3. 17  P. Cair. Zen. 4.59677 (TM 1305), 3rd c.; P. Cair. Zen 3.59406 (TM 1048), 3rd c.; P. Cair. Zen. 4.59698 (TM 1325), 3rd c. See R. Scholl, Sklaverei in den Zenonpapyri: Eine Untersuchung zu den Sklaventermini, zum Sklavenerwerb und zur Sklavenflucht (Trierer Historische Forschungen 4; Trier: Verlag Trierer Historische Forschungen, 1983), 10. 18  P. Cair. Zen. 2.59176 (TM 822), 3rd c.; P. Cair. Zen. 4.59729 (TM 1356), 3rd c.; P. Heid. Gr. 3.227 (TM 78313), 3rd c. In the latter, for παιδάρια see lines 84, 89, 90, 119, 149, 154, 163, and ἐργάται lines 16, 23, 41, 56, 80, 105, 147, 152, 157, 160, 178, etc. According to

195

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

παιδάριον could be sponsored for athletic or even military training, suggesting long-term and mutually beneficial relationships between παιδάρια and the giftestate were also common, if not the norm.19

3. The Νεανίσκος as a Civic Officer in Ptolemaic Egypt Moving on to analysis of the lexical item νεανίσκος, again we begin with distributional observations (Table 2). Table 2: Distribution of νεανίσκος

Literature Papyri Inscriptions20 Total

5th

4th

3rd

2nd

1st

1st c.e.

2nd c.e.

45 – 5 50

89 – 3 92

2 17 8 27

67 3 16 86

83 1 1 85

277 2 5 284

473 7 5 485

In general, νεανίσκος is more commonly attested within literature of both the classical and post-classical periods, with its frequency markedly increasing after the turn of the era.21 Compared to its use in literature, this lexical item is less common among the papyri and inscriptions, which is the reverse of trends seen above for παιδάριον. Focus falls here upon the evidence germane to Ptolemaic Egypt, since in that context νεανίσκος was associated with a specific concept. This association is made clear in the documentary evidence from that region, but it is also supported by literature from other regions that specifically discusses νεανίσκοι in Egypt. It Bieżuńska-Małowist, ἐργάται were “paysans s’engageant de temps à autre comme journaliers” (L’esclavage, 63). 19  For athletic training, see P. Iand. VI 92 (= P. Lugd. Bat. 20.51); P. Lond. 7.1941 (TM 2384), cf. 3rd c.; cf. P. Pt. VI. 17146 and P. Cair. Zen. 1.59098 (TM 750), 3rd c. For military training, see P. Cair. Zen. 3.59298 (TM 942), 3rd c.; cf. P. Cair. Zen. 3.59488 (TM 1126), 3rd c. See also Scholl, Sklaverei, 11; Gardiner, “School”, 212; C.C. Edgar (ed.), Catalogue général des antiquités égyptiennes du Musée du Caire, nos. 59298–59531. Zenon Papyri III (Cairo: L’Institut français d’archéologie orientale, 1928), 1. According to Thompson, the poorer class of Greeks would have had a “variety of statuses” and are known to have looked to local representatives like Zenon for provision and protection (“Slavery”, 199). 20  A further twenty-one inscriptions are undated, and there are seven in which νεανίσκος may appear but which are too fragmentary to be reliable. 21  Note that almost all of the literary attestations of νεανίσκος in the 3rd through 1st centuries b.c.e. occur in Polybius (67×), Diodorus Siculus (49×), and Dionysius of Halicarnassus (31×).

196

William A. Ross

was there that the lexical item came to be associated with a Greek civic institution that had developed in 4th century Athens to provide highly structured civic training for young men, known as ephebeia (ἐφηβεία). This system was designed to “integrate young men into the community of adults” and to provide military training and experience in service to the state.22 The ephebic system was closely tied to the gymnasium and transformed over time, becoming one of the most important elements in Hellenizing the ancient Mediterranean region as it spread to “every corner of the Greek world” by the 2nd century b.c.e.23 Evidence for ephebic training in Ptolemaic Alexandria, as well as in the rural chora, indicates that it was very similar to the Athenian model and present there as early as the first half of the 3rd century b.c.e.24 Within the social context of Ptolemaic Egypt we find that the lexical item νεανίσκος is most frequently used to refer to those who completed ephebic military training—or were close to completion—to serve as a kind of civic officer.25 They appear to have formed “a separate class of young soldiers who were charged with police duties and participated in municipal life.”26 These males were granted full citizenship and typically 20–30 years of age (Polybius, Hist. 4.20.7). As is apparent in the sources below, the Egyptian νεανίσκοι 22  A.S. Chankowski, “Age-Class”, in R.S. Bagnall et al. (ed.), The Encyclopedia of Ancient History (Oxford: Blackwell, 2013) 178–80, on p. 179. Starting in Athens, this military training physically organized males into distinct spaces in the gymnasium according to ἡλικία (Aristotle, Pol. 7.1331a37-38). The typical age range was from 18–20 years old, and their training typically included use of a bow, javelin, and other elements of hoplite warfare, as well as border patrol; E. Casey, “Educating the Youth: The Athenian Ephebeia in the Early Hellenistic Era”, in J.E. Grubbs, T. Parkin, and R. Bell (ed.), The Oxford Handbook of Childhood and Education in the Classical World (Oxford: Oxford University Press) 418–43, on p. 421; Garland, Greek Way, 183–4. The ephebic system has been especially well documented by epigraphers such as N.M. Kennell, who notes that in Athens it was part of “wide-ranging military reforms that radically altered the method of conscription for hoplites;” Ephebeia: A Register of Greek Cities with Citizen Training Systems in the Hellenistic and Roman Periods (Nikephoros Beihefte 12; Hildesheim: Weidmann, 2006), ix. See also A.S. Chankowski, L’éphébie Hellénistique. Étude D’une Institution Civique Dans Les Cités Grecques Des Îles De La Mer Égée Et De L’asie Mineure (Culture et cité 4; Paris: De Boccard, 2011). 23  See Kennell, Ephebeia, vii–viii, on p. xi; See also Chankowski, L’éphébie, 114–234; N.M. Kennell, “The Ephebeia in the Hellenistic Period,” in W. Martin Bloomer (ed.), A Companion to Ancient Education (Oxford: Wiley Blackwell, 2015), 172–5; idem., “New Light on 2 Macabees 4:7–15,” JJS 56.1 (2005) 10–24; Garland, Greek Way, 185. 24  See, e.g., IG XII.9 Supp. 646; CPJ 41; IG Fayum I.8; SEG VIII.694; P.Oxy. 2465, fr. 2, col. 1, ll. 10–11; IG VII 2715–21; Chankowski, L’éphébie, 173, 229; Kennell, “Ephebeia in the Hellenistic Period”, 173, 181–2; M. Launey, Recherches Sur Les Armées Hellénistiques (2 vols.; La Bibliothèque des Écoles françaises d’Athènes et de Rome 169; Paris: E. de Boccard, 1950) 2:859 n. 1. 25 Forbes, Neoi, 61–4; E. Cantarella, “Neaniskoi. Classi di età e passaggi di « Status » del diritto ateniese”, MEFRA 102 (1990) 37–51, on pp. 43–5. 26  N.M. Kennell, “Who Were the Neoi?”, in P. Martzavou and N. Papazarkadas (ed.), Epigraphical Approaches to the Post-Classical Polis: Fourth Century BC to Second Century AD (Oxford Studies in Ancient Documents; Oxford: Oxford University Press, 2013) 217–32, on p. 218.

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

197

were qualified for various duties and often associated with a military unit called a σημείον, as shown in (9) below.27 Egyptian νεανίσκοι were trained by a drill sergeant (ὁπλομάχος), led by a military commander (ἡγεμών), were occasionally categorized as soldiers (στρατιῶται), and were dispatched on peacekeeping assignments.28 Several examples from papyri can demonstrate the kinds of benefits and duties that were entailed by the ephebic training of an Egyptian νεανίσκος. Firstly, we find that in the 3rd century νεανίσκοι took oaths of loyalty to the state and were granted cleruchic lands: (5)  Φανίας Ζήνωνι χαίρειν. τοὺς κεκληρουχημένους ἐν τῶι Ἀρ[σινοίτηι] | νομῶι νεανίσκους \διέγνωκα/ πάντας ἀρ̣ [ι]θμήσω καὶ ὁρκιῶ ἐν Φιλαδελφ̣[είαι]. | καλῶς οὖν ποιήσεις καταλυμάτιόν μοι ἑτοιμάσας· | τῶι γὰρ σωματίωι ἐτύγχανον ἀσθενῶς διακείμενος, | ἅμα δὲ καί σε ἰδεῖν βούλομαι ὅσον ἐπιδέχεται ⟦πλεῖ̣σ̣τ[̣ ον] ⟧ | χρόνον. ἔρρωσο. Phanias to Zenon, greetings. The νεανίσκους who were made cleruchs in the Arsinoite nome I have determined I will count them all and administer oaths in Philadelphia. So would you please prepare a little room for me, as I happen to be poorly disposed in body, and so I would also like to see you for as long a time as possible. Farewell. (P.Cair.Zen 2.59254 [TM 899])

In the same century, νεανίσκοι were also entitled to land allotments (κλήροι) in which they held some authority, and were also bound by oath to the king:29 (6) [Φα]νίας Ἀντιπάτρωι χαίρειν. πρότερον μέν σοι ὑποθεὶς τῆς παρὰ Διοτίμου | ἐπιστολῆς τὸ ἀντίγραφον ἔγραψα ἐπιμεληθῆναι ἱππέ[ων] | ὅσοις καταμεμέτρηται γῆ δυναμένη σπείρεσθαι εἰς τ[ὸ] ε καὶ λ (ἔτος) [ὡς] | πᾶσα σπαρῆι κ[α]ὶ δυνηθῶσιν οἱ ἐν τῆι ἐπιστατείαι ν[εανίσκοι] | ἀπὸ τῶν γενομένων καρπῶν χορηγηθέντες καταβαίνειν πρὸς τὸν βασιλέα | ἔφιπποι καὶ τοῖς ἄλλοις \ἀναγκαίοις/ κατεσκευασμένοι. Phanias to Antipatros, greetings. First, enclosed for you is a copy I wrote of a letter from Diotimos to oversee the cavalry, to whom land is allotted that is able to be sown from year 35, so all is sown and the νεανίσκοι in charge are

27  See L. D’Amore, “Ginnasio e difesa civica nelle poleis d’Asia minore (IV–I sec. A.C.)”, Revue de Études Anciennes 109 (2007) 147–73, esp. pp. 169–71; Forbes, Neoi, 20, 60. On the σημείον, see J. Lesquier, Les Institutions Militaires De L’égypte Sous Les Lagides (Paris, E. Leroux, 1911), 103–4; Launey, Armées Hellénistiques, 861. 28  See various sources cited in Forbes, Neoi, 65–6, who notes that some scholars believe Egyptian νεανίσκοι would have belonged to the upper level of the population. 29 On κλήροι, see lines 8, 9, and 12, not cited here and Launey, Armées Hellénistiques, 860.

198

William A. Ross

able, having been supplied from the crops produced, to go down to the king by horse, and are fully equipped with whatever else they need. (P. Freib 1.7 [TM 5644], lines 5–7)

Another text preserves an account of the arrival of Ptolemy III at Antioch in 246 b.c.e., where he was jubilantly received by various civic officials, including: (7) [οἵ τε] σατράπαι καὶ οἱ ἄλλοι ἡγε- | μόν[ες καὶ οἱ στρατιῶ]ται καὶ οἱ ἱερεῖς καὶ αἱ | συναρχίαι | καὶ [πάντες οἱ ἀπ]ὸ τοῦ γυμνασίου νεανίσκοι καὶ ἄλλος | ὄχ[λος | ἐστεφ]ανωμένος … the satraps and other leaders, and the soldiers and the priests and the magistrates, and all the νεανίσκοι from the gymnasium, and a surrounding crowd besides. (P.Petr. 2.45, col. 3, lines 19–23)30

The association of νεανίσκοι with the gymnasium and their civic responsibilities appears also in a late 2nd century papyrus:  ξιῶ μὴ ὑπερ|ιδεῖν με ἀγνωμονούμενον | ἀλλὰ ἐπανενέγκαι ἐπί τε τὸν | γυμναἀ σίαρχον καὶ [ἐ]πὶ τοὺς | ἐκ τοῦ ἐν τῆι Φιλαδελφείαι | γυμνασίου νεανίσκους I request that you not to allow me to be unfairly treated but to refer my case to the gymnasiarch and to the νεανίσκους of the gymnasium (BGU 6.1256 [TM 4543], lines 24–29)31 (8)

A similar instance of the regulatory and policing activity of νεανίσκοι from 3rd century Egypt appears in P.Cair.Zen. 1.59018 (TM 678), where we read of their duty in collecting a debt as far away as Palestine (lines 4–9).32 Finally, as shown below, at least some νεανίσκοι were grouped among soldiers to form part of a military unit called a σημείον, a “kind of military equivalent of οἱ ἐκ τοῦ ­γυμνασίου.”33 30  Discussed in Chankowski, L’éphébie, 422–3. 31  Translation adapted from A.S. Hunt, Select Papyri: Public Documents (3 vols.; LCL 282; Cambridge, Harvard University Press, 1934), 2:253. 32  See J.S. Kloppenborg, The Tenants in the Vineyard: Ideology, Economics, and Agrarian Conflict in Jewish Palestine (WUNT 195; Tübingen: Mohr Siebeck, 2006), 365. 33  R.W. Daniel, “The Military [Cημεϊον] in P.Amh. Ii 39,” ZPE 52 (1983) 269–71, on p. 269. He notes that it is possible the lacunae [ἐκ] τοῦ in line 1 could be restored as [ἑαυ]τοῦ, which would render the phrase “captain of … the νεανίσκοι of his own standard.” This particular rank of νεανίσκοι is reflected in J.H. Moulton and G. Milligan, who in reference to this papyrus state that the νεανίσκοι seem to have been “members of a chosen band of youths;” The Vocabulary of the Greek Testament Illustrated from the Papyri and Other Non-Literary Sources (London: Hodder and Stoughton, 1929), 423.

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

199

(9)  Πόρτεις ἡ̣γε̣ μὼν τῶν ἐν προχειρισ|μῶι̣ καὶ οἱ̣ [ἐκ] τοῦ σημείου νεανίσκοι | Πατῆτι [καὶ] Παχράτηι καὶ τοῖς ἄλλοις | [στ]ρ̣ [α]τιώται[ς] πᾶσι χαίρειν Porteis, captain of those in training and the νεανίσκοι of the standard, to Patet and Pachrate and the other soldiers, greetings to all… (P. Amh. 2.39 [TM 164], lines 1–4)

Chankowski concludes that “[l]a documentation papyrologique montre à l’évidence que le groupe des neaniskoi en Égypte ptolémaïque était une composante de l’armée royale et participait à la vie du gymnase. Il s’agit donc de gens certes adultes, mais pas encore très avancés en âge, disons entre 20 et 30 ans.”34 Space constraints limit final observations to the use of the lexical item in Polybius. Frequently in his work, νεανίσκοι are mentioned in contexts that make clear their military training (e.g., Hist. 4.16.6; 4.35.1–3; 4.76.8–9; 6.20.1–3; 21.3b) and status as civic officers (5.30.1). In one relevant text, Polybius recounts the fall of the Aetolian general Scopas to Aristomenes (Hist 18.53.1–11). Upon learning of a plot for usurping power of the Ptolemaic administration, Aristomenes surrounded a house where Scopas was located with soldiers (στρατιῶται) and sends from among them one Ptolemy to extract Scopas with the help of νεανίσκοι (7–11; cf. 5.96.6–8).35

4. Case Studies The analyses above for the lexical items παιδάριον and νεανίσκος provide clarity regarding their usage in post-classical Greek. Specifically, these words are associated in post-classical sources with particular categories of male individuals within the Ptolemaic Egyptian social context. On the one hand, those described as a παιδάριον were often of lower social status and economically dependent upon a patron or estate in long-term positions of semi-skilled labour. On the other hand, individuals described as a νεανίσκος were products of the Greek ephebic system

34  L’éphébie, 357. 35  For other relevant texts see νεανίκος in A. Mauersberger and H. Helms (ed)., PolybiosLexikon (4 vols.; Berlin: Akademie Verlag, 22006) 1.4:1671. Although 43 citations are given for “non-military” uses of νεανίσκος, 24 are listed as specifically military related, for which the glosses provided include “als Waffenträger”, “Rekruten”, and “Soldaten”. Launey concludes that the νεανίσκοι “forment donc une classe de jeunes soldats … chargés de fonctions de police, et participant aux fêtes, voire, à l’óccasion, à la vie municipale, un peu à la manière des éphèbes dans les cités grecques. Mais forment-ils un corps de futures officiers, de cadets … ou seulement des soldats des jeunes classes, des recrues … . La place éminente qu’ils tiennent au gymnase suggère un recrutement d’élite qui serait bien celui de futurs officiers;” Armées Hellénistiques, 861–2. Cf. McGregor, “Semantic Study”, 75–6.

200

William A. Ross

as it was manifested in Egypt, receiving military training through the gymnasium and assigned to various duties as a civic officer. Of course, the semantics of the lexical items παιδάριον and νεανίσκος allowed for their use in less specific ways than those described above. When facing such cases, it is appropriate to translate using the typical glosses for these words in modern lexicons: sometimes a παιδάριον is simply a ‘child’ and a νεανίσκος a ‘young man’. Often we lack sufficient knowledge of the context of use (or lack an equivalent term in English) to do anything else. The same is true of many instances where these lexical items are used in the Septuagint. But in other cases, παιδάριον or νεανίσκος appear in the translation of texts that present situations very similar to those where semi-skilled labourers and civic officers were found in the Ptolemaic social context. The following case studies present two such texts and point to the likelihood that the translation of these books was carried out by Jews who were apparently at home within that context.

4.1 Παιδάρια in LXX-Ruth The translator of the book of Ruth had few occasions to use vocabulary within the semantic range of παιδάριον. Of course, this book is relatively short compared with others in the Septuagint. But there are no occurrences of νεός or νεανίσκος, for example. Otherwise, there is one instance apiece of παιδίον (4:16), and νεανίας (3:10), only the latter two of which refer to males, one being an infant.36 There is, however, no lack of males in the book of Ruth, most notably the farm hands under Boaz’s charge who are mentioned on six occasions.37 The Greek translator selected παιδάριον in all six instances where farm hands are discussed, and in every case it is used to render ‫נער‬. (10)  καὶ εἶπεν Βοός τῷ παιδαρίῳ αὐτοῦ τῷ ἐφεστῶτι ἐπὶ τοὺς θερίζοντας Τίνος ἡ νεᾶνις αὕτη;38 καὶ ἀπεκρίθη τὸ παιδάριον τὸ ἐφεστὸς ἐπὶ τοὺς θερίζοντας καὶ εἶπεν Ἡ παῖς ἡ Μωαβῖτίς ἐστιν ἡ ἀποστραφεῖσα μετὰ Νωεμὶν ἐξ ἀγροῦ Μωάβ And Boaz said to his παιδάριον who was supervisor over the reapers, “Whose is this young woman?” And the παιδάριον who was supervisor over the reapers answered and said “She is the Moabite girl who came back with Naomi from the field of Moab.” (Ruth 2:5–6)

36  The lexical items παῖς (2:6) and νεᾶνις (2:5) also appear, but both are used in reference to Ruth, otherwise called a γυνή (3:8, 11, 14; 4:11) and παιδίσκη (4:12). 37  See E. Bons, “Sur la traduction du vocabulaire de la servitude dans la Septante du livre de Ruth”, JSJ 33 (2002) 153–63, esp. p. 156. 38  LXX-Ruth text from U. Quast (ed.), Ruth (Septuaginta: Vetus Testamentum Graecum auctoritae Academiae Scientiarum Gottingensis editium 4.3; Göttingen: Vandenhoeck & Ruprecht, 2006).

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

201

(11)  οἱ ὀφθαλμοί σου εἰς τὸν ἀγρόν, οὗ ἐὰν θερίζωσιν, καὶ πορεύσῃ κατόπισθεν αὐτῶν· ἰδοὺ ἐνετειλάμην τοῖς παιδαρίοις τοῦ μὴ ἅψασθαί σου· καὶ ὅ τι διψήσεις, καὶ πορευθήσῃ εἰς τὰ σκεύη καὶ πίεσαι ὅθεν ἂν ὑδρεύωνται τὰ παιδάρια. Keep eyes towards the field, wherever they are reaping, and proceed after them. Behold I commanded the παιδάρια not to touch you. And if you should want something to drink, then you will go to the containers and drink whatever the παιδάρια draw. (2:9)39 (12)  καὶ ἀνέστη τοῦ συλλέγειν, καὶ ἐνετείλατο Βοός τοῖς παιδαρίοις αὐτοῦ λέγων Καί γε ἀνὰ μέσον τῶν δραγμάτων συλλεγέτω, καὶ μὴ καταισχύνητε αὐτήν· And she rose to glean, and Boaz commanded his παιδάρια saying, “Let her gather even among the sheaves, and do not violate her.” (2:15) (13)  καὶ εἶπεν ‘Ροὺθ πρὸς τὴν πενθερὰν αὐτῆς Καί γε ὅτι εἶπεν πρός με Μετὰ τῶν παιδαρίων μου προσκολλήθητι, ἕως ἂν τελέσωσιν ὅλον τὸν ἀμητόν, ὃς ὑπάρχει μοι. And Ruth said to her mother-in-law “He even said to me ‘Stay close with my παιδάρια, until they finish the entire harvest, which belongs to me.” (2:21)

What is striking about these males in the contexts they appear is their position over other agricultural labourers, namely the reapers, in example (10). Perhaps because that status entails some level of power over others or privileges, Boaz feels he must circumscribe their contact with Ruth while she is in their company during the workday (11–13). Their status does not exempt them from hard work, however, as the παιδάρια apparently bear the responsibility for drawing water (11). The examples in (10) through (13) describe an agricultural setting in the narrative of Ruth that is markedly similar to that of Ptolemaic Egypt. There, Greekspeaking settlers known as cleruchs were granted land allotments on which tax was due, and which were often leased back to Egyptian natives for management and labour.40 It is within this social context that we find παιδάρια, as described above, working as semi-skilled paid labourers and occasionally entrusted with responsibilities of financial consequence to the estate. That situation is closely paralleled by the one in the literary context in Ruth 2 and the males described there.41 Given the characteristics of the individuals designated by the term παιδάριον in Ptolemaic Egypt, the translator’s selection of this lexical item in translating ‫ נער‬appears to be a natural consequence of his familiarity with that context. 39  On the use of κοράσιον in 2:8, see Bons, “Servitude”, 158–9. 40  G. Shipley, The Greek World After Alexander: 323–30 b.c. (London: Routledge, 2000), 216–19. 41  Bons states that “à la lumière des papyrus, rien n’empêche de supposer qu’un παιδάριον puisse se voir confier des responsabilités dans les domaines du commerce et de l’agriculture” (“Servitude”, 159).

202

William A. Ross

Analysis of the lexical item παιδάριον in Greek sources, then, provides supporting evidence for the book of Ruth having been translated in Egypt, as some scholars have suggested elsewhere.42 At the very least, the translator demonstrates his linguistic subtlety and skill insofar as he conveys the meaning of his source text in Greek by consistently using vocabulary that is semantically appropriate within his social context.

4.2 Νεανίσκοι in LXX-Ezekiel There is little age- or status-related Hebrew vocabulary in the book of Ezekiel. Thus, within the Greek version of Ezekiel, there is just one instance each of νήπιος (9:6) and παῖς (46:17), as well as five attestations of νεανίσκος.43 In one of the latter texts, νεανίσκος is used in a general way in a phrase commanding the indiscriminate slaughter of all idolaters, including the πρεσβύτερος, νεανίσκος, παρθένος, νήπιος, and γυνή (9:6). In all four other instances of νεανίσκος in LXX-Ezekiel, however, it appears in contexts that very closely resemble that in which νεανίσκοι appear in Ptolemaic Egypt. Three of these occur in close proximity to one another, in a chapter describing Oholah and Oholibah, who metaphorically represent Samaria and Jerusalem, respectively (23:4). Notably, these characters are portrayed as having “played the whore in Egypt” (v. 3; ‫)ותזנינה במצרים‬44. Oholah is said to have “lusted after her lovers the Assyrians” there, who were “warriors” (‫)קרובים‬. These males are then described as: (14) … ἐνδεδυκότας ὑακίνθινα, ἡγουμένους καὶ στρατηγούς· νεανίσκοι ἐπίλεκτοι πάντες, ἱππεῖς ἱππαζόμενοι ἐφ᾽ ἵππων.45 … clothed in blue, commanders and captains. All were elite νεανίσκοι, cavalrymen riding on horses. (23:6)

Oholibah too went after these same Assyrians in Egypt (cf. v. 19).

42  Ibid., 163; A. Rahlfs, Studie über den griechischen Text des Buches Ruth (Berlin: Buchhandlung, 1922); idem, Das Buch Ruth griechisch, als Probe einer kritischen Handausgabe der LXX (Stuttgart: Privileg. Württ Bibselanstalt, 1922). The question is a complex one owing to the textual situation, on which see E. Bons, “Ruth”, in S. Kreuzer (ed.), Einleitung in die Septuaginta (3 vols.; LXX.H; Gütersloh: Gütersloher Verlagshaus, 2016) 1:199–204, on p. 202. 43  Additionally, there are well over one hundred instances of both ὑιός and ἄνθρωπος, in large part owing to the repeated ‫ בן־אדם‬formula in the book (93×). 44  All translations of the Hebrew Bible drawn from the NRSV. 45  LXX-Ezekiel text from J. Ziegler (ed.), Ezechiel (SVTG 16,1; Vandenhoeck & Ruprecht, 32006).

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

203

(15)  ἐπὶ τοὺς υἱοὺς τῶν Ἀσσυρίων ἐπέθετο, ἡγουμένους καὶ στρατηγοὺς τοὺς ἐγγὺς αὐτῆς ἐνδεδυκότας εὐπάρυφα, ἱππεῖς ἱππαζομένους ἐφ᾽ ἵππων· νεανίσκοι ἐπίλεκτοι πάντες. She plied herself upon the sons of the Assyrians, commanders and captains who were near her clothed in fine garments, cavalrymen riding on horses. All were elite νεανίσκοι. (23:12)

Her sexual activity continues and intensifies, until God rouses all of Oholibah’s lovers against her. (16) … υἱοὺς Βαβυλῶνος καὶ πάντας τοὺς Χαλδαίους, Φακουδ καὶ Σουε καὶ Κουε καὶ πάντας υἱοὺς Ἀσσυρίων μετ᾽ αὐτῶν, νεανίσκους ἐπιλέκτους, ἡγεμόνας καὶ στρατηγοὺς πάντας, τρισσοὺς καὶ ὀνομαστοὺς ἱππεύοντας ἐφ᾽ ἵππων· … sons of Babylon and all the Chaldeans, Pekod and Shoa and Koa and all the sons of Assyrians with them, elite νεανίσκοι, all commanders and captains, triple ranked (?) and renowned men riding on horses. (23:23)46

Moreover, this host of military figures will come against Oholibah in chariots armed with oblong shields (θυρεός) and light shields (πέλτη; v. 24). In this way, God turns back her whoring “from the land of Egypt” (ἐκ γῆς Αἰγύπτου) so Oholibah no longer remembers that land (v. 27). The last instance of νεανίσκος in LXX-Ezekiel appears in a verse in which the translator has specified geographical locations to those known in the Ptolemaic Egyptian context. Within a lament over the coming judgment against Egypt in general, we read: (17)  νεανίσκοι Ἡλιουπόλεως καὶ Βουβάστου ἐν μαχαίρᾳ πεσοῦνται, καὶ αἱ γυναῖκες ἐν αἰχμαλωσίᾳ πορεύσονται. The νεανίσκοι of Heliopolis and of Bubastis will fall by the sword, and the women will depart among captives (30:17)

The two cities mentioned in the Greek version were located in the Nile Delta of Lower Egypt. Here the translator has rendered the Hebrew toponyms ‫ און‬and ‫פי־בסת‬, which appear to be transcriptions of the Egyptian names Onu and Tell Bastah (Per-Bast), respectively, using their appropriate Greek names.47 In examples (14) through (17), the literary context that was presented to the translator is clearly an Egyptian one. Egypt is the explicit location of both the 46  Note that Ziegler lists τριστάτης as an alternative reading for τρισσοὺς among various witnesses, the former of which Muraoka defines as “military officer of high rank”; GELS, ad loc. 47  On the Egyptian names, see “Heliopolis” and “Tell Bastah,” in J. Everett-Heath (ed.), The Concise Dictionary of World Place-Names (Oxford: Oxford University Press).

204

William A. Ross

metaphorical sexual activity and the object of lament in the chapters discussed. More specifically, the males described in each of these four texts are involved explicitly in military activity. Given the typical characteristics of those described as νεανίσκοι in Ptolemaic Egypt, the translator’s selection of this lexical item for such contexts is well-suited, and demonstrates remarkable linguistic sensitivity. In light of the analysis of documentary sources presented above, the use of νεανίσκος in these texts provides evidence that the translator worked and wrote for a readership situated within the Egyptian social context. The transposition of toponyms from Egyptian to Greek in (17) lends further support to this idea, which indeed has been posited in preliminary fashion by others.48

5. Conclusion Semantic analysis of Greek lexical items used to refer to individuals owing to their perceived age and social status is a complicated task. On top of the challenges presented by the state of the evidence, both in literary and documentary sources, there is the added difficulty of sufficiently understanding the social context in which such words were used to categorize people according to criteria not always well understood. Something of an interpretive circle exists in analysis of ancient documents, using the words to understand the social context and the social context to understand the words. However, as analysis proceeds a picture does emerge. This essay has presented an examination of the lexical items παιδάριον and νεανίσκος in order to better understand their usage and meaning within Greek sources. Both are frequently found within documentary evidence related to Ptolemaic Egypt, used there in reference to different categories of males. These males may be characterized as a semi-skilled labourer and a civic officer, respectively. Certain Septuagint translators dealt with texts in which the narrative context presented a situation very similar to those described in Greek sources where a παιδάριον or νεανίσκος often appear. In light of the analysis above, use of these lexical items in such contexts in LXX-Ruth and LXX-Ezekiel provides 48  See H.St.J. Thackeray, The Septuagint and Jewish Worship: A Study in Origins (London: Oxford University Press, 1921), 9–39, esp. pp. 37–39; G. Dorival, M. Harl, and O. Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien (Initiations au christianisme ancen 3; Paris: Éditions du Cerf, 1988), 105; K. Hauspie, “Ezekiel”, in J.K. Aitken (ed.), T&T Clark Companion to the Septuagint (London: Bloomsbury T&T Clark, 2015) 528– 43, on p. 532. Also note that the oldest witness to LXX-Ezekiel is papyrus 967, found near Aphroditopolis in Egypt and dated to the 2nd century c.e.; S. Kreuzer, “Papyrus 967: Its Significance for Codex Formation, Textual History, and Canon History”, in The Bible in Greek: Translation, Transmission, and Theology of the Septuagint (SBLSCS 63; Atlanta: SBL Press, 2015) 255–71, on p. 255.

Some Aspects of Παιδάριον and Νεανίσκος in Ptolemaic Egypt

205

further evidence that their translators lived and worked in an Egyptian milieu. The translators of both books, moreover, chose to use these lexical items in ways appropriate to their social and linguistic context, demonstrating their subtlety and skill as translators, and thereby providing the lexicographer with further evidence.

Srećko Koralija παράδεισος as/and κῆπος 1. Introduction οἱ δὲ παράδεισοι, βαρβαρικὸν εἶναι δοκοῦν τοὔνομα ἥκει κατὰ συνήθειαν εἰς χρῆσιν Ἑλληνικήν, ὡς καὶ ἄλλα πολλὰ τῶν Περσικῶν paradeisoi: thought to be barbarian, the word has been adapted by custom to Greek usage, like many other (words) of the Persians1

This passage from Pollux’s (second century c.e.) Onomasticon reflects the sophisticated biography of the Greek word παράδεισος. From the Classical period and Xenophon’s comparison of the meaning of παράδεισος with the Greek ­κῆπος,2 up to the Hellenistic period, the Greek use of παράδεισος went through a gradual increase, including in the Septuagint.3 Over the last decades scholars like Deissmann,4 Lee,5 Montevecchi,6 Joosten,7 Evans,8 and Aitken9 have examined the place of the Septuagint grammar and vocabulary within the Koine, and its distinctiveness in contrast to Classical Greek. In this respect, some attention has already been paid to the LXX understanding of παράδεισος.10 However, the 1  P. Grammaticus/E. Bethe (ed.), Onomasticon (Leipzig: Teubner, 1931), 150. 2 Xenophon, Oec. 4, 13.14.20. 3  In the Classical period, παράδεισος is found exclusively in Xenophon’s corpus, used to explain its Persian setting (that is, in Xenophon’s Classical Greek παράδεισος is a loan word). As we will demonstrate with examples, in both the Hellenistic period and in the Septuagint, the word became Greek and, consequently, is used much more frequently. The fact that it is not used in only one corpus (as it was in the case of Xenophon) but in different corpora from different periods, supports the argument about the increase in its usage. 4  A. Deissmann/A.J. Grieve, 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 (Edinburgh: T&T Clark, 1901). 5  J.A.L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (Chico, CA: Scholars Press, 1983). 6  O. Montevecchi, Bibbia e papiri. Luce dai papiri sulla bibbia greca (Barcelona: Institut de Teologia Fondamental, Seminari de Papirologia, 1999). 7  J. Joosten, “The Vocabulary of the Septuagint and Its Historical Context”, in J. Joosten/ E. Bons (ed.), Septuagint Vocabulary: Pre-History, Usage, Reception (Atlanta: Society of Biblical Literature, 2011) 1–11. 8  T.V. Evans, Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference (Oxford: Oxford University Press, 2001). 9  J.K. Aitken, No Stone Unturned: Greek Inscriptions and Septuagint Vocabulary (Winona Lake, IN: Eisenbrauns, 2014). 10  The principal study is: Lee, A Lexical Study of the Septuagint, 53–56.

208

Srećko Koralija

Hellenistic and LXX use of the word κῆπος in comparison to παράδεισος has not yet been investigated. Since a full examination of all the occurrences would form a major study in itself, the present proposal will bring out what lies behind their biographies in comparison with the LXX and current lexicographical ­explanations. It is demonstrated that in Classical Greek, in contrast to κῆπος, the word παράδεισος is not in general use, whilst in the Hellenistic period and the Septuagint both παράδεισος and κῆπος are used in different contexts. Namely, from the evidence we have at our disposal, in the Hellenistic period a παράδεισος was not an enclosed place for hunting animals, but rather an orchard,11 while κῆπος was more like a garden with vegetables and young plants. The paper is divided in two sections: The first examines the Old Persian variants of παράδεισος along with its reception in Classical Greek, and contexts in which παράδεισος and κῆπος are used in the papyri and the Hellenistic period. The second part investigates their features in the Septuagint, later periods, and lexicons.

2. Etymology Lexicons of the Greek language and some commentaries on the Greek literature of the Classical period quite often inform the user that παράδεισος is etymologically connected with variants of Old Persian. They analyse it as either Avestan,12 Old Iranian,13 Persian14 or oriental15 pairidaēza and claim it as the source of both the Hebrew ‫ פרדס‬and Greek παράδεισος. However, the differences between these descriptions are small but significant. Avestan dictionaries explain that the first part of the word, i.e., pairi, signifies ‘with, around, over; under; all round, completely, thoroughly’ while the word pairidaēza implies a ‘walled enclosure’.16 These conclusions are supported by the Avestan texts themselves.17 The transition of pairidaēza18 from Avestan to the Greek παράδεισος probably happened during the Achaemenid period when it came to the Greek audience via Xenophon’s works, if we presume they were transmitted to the Greek audience during 11  For their English equivalents, I mainly follow Lee, A Lexical Study of the Septuagint, 55. 12  H. Liddell/R. Scott/H. Jones, Greek-English Lexicon (Oxford: Clarendon Press, 1996). 13  R.S.P. Beekes, Etymological Dictionary of Greek (Leiden: Brill, 2010). 14 E. Eynikel/K. Hauspie /J. Lust, Greek-English Lexicon of the Septuagint (Stuttgart: Deutsche Bibelgesellschaft, 2003). 15  E.g., A.H.N. Sewell’s commentary The Oeconomicus (In Greek with Introduction and Notes in English) (Cambridge: University Press, 1925). 16  J.H. Peterson, Dictionary of Most Common Avesta Words (online edition, 1995), available at: http://www.avesta.org/avdict/avdict.htm#dctp letter P, line 5. 17  M. Fakour, “Garden i. Achaemenid Period”, in E. Yarshater (ed.), Encyclopaedia Iranica (New York: Bibliotheca Persica Press 2001) 10:297–8. 18  Old Persian paridaida and Median paridaiza (walled-around, a walled garden).

παράδεισος as/and κῆπος

209

Xenophon’s life or shortly after his death (354 b.c.e.). Specifically, Xenophon is the only writer from the Classical period who uses that word. Concerning the Hebrew word ‫פרדס‬, lexicons connect it either with παράδεισος (Muraoka19) or to the Avestan (Köhler-Baumgartner). The latter explains that the Hebrew ‫ פרדס‬comes from the Avestan pairidaēza (rampart) which designates a domain of the king in the Achaemenid period.20 This seems to be inappropriate for two reasons: first, the Avestan pairidaēza denotes a barren plot of land surrounded by a wall; second, the meaning Köhler-Baumgartner gives to the Avestan word is actually Old Persian in which the form would be the Late Achaemenid paradayadama for paridaidam,21 rather than pairidaēza. That is, the Iranian loan word in the Greek and Semitic sources has the shape of the Avestan word but the meaning of the Old Persian. In this respect, the Avestan word pairidaēza does not overlap either in meaning or in context with the Greek παράδεισος and its counterpart κῆπος. Consequently, putting the Avestan, or any other Old Persian variant into a Greek dictionary for the word παράδεισος can mislead the readers since the Old Persian usage does not allude to any kind of garden or orchard in the Koine sense of the word παράδεισος. Nonetheless, the etymology of the word is correct, but the meaning and the context are different. We will now have a look at some of the data from the Classical and Post-Classical period.

2.1 Classical Greek While κῆπος is the usual Greek word for ‘garden’22 already used by Homer in the sense of the ‘prepared ground’,23 its counterpart παράδεισος entered the scene when it was used for the first time by Xenophon, who was also the only writer of the Classical period to use it. Accordingly, the data about the use of παράδεισος in the Classical period are confined to the relatively small corpus of Xenophon. It is also probable that Xenophon was not an eyewitness of the Persian environment he described.24 19  T. Muraoka, A Greek-English Lexicon of the Septuagint (Louvain: Peeters, 2009). 20  L. Köhler/W. Baumgartner, The Hebrew and Aramaic Lexicon of the Old Testament (Leiden: E.J. Brill, 1994–2000), 963. 21 A. Panaino, “No Room for the ‘Paradise’? About Old Persian paradayadama”, in G.P. Basello/A.V. Rossi (ed.), Persepolis and Its Settlements: Territorial System and Ideology in the Achaemenid State (Napoli: Universita degli studi di Napoli “L’Orientale”, 2012) 139–52. I express my gratitude to Almut Hintze who brought my attention to some references to the scholarship of Old Avestan and Persian. 22 Beekes, Etymological Dictionary of Greek, 688. 23 Iliad Θ 306; R.J. Cunliffe/N. Okla, A Lexicon of the Homeric Dialect (Norman: University of Oklahoma Press, 1963), 226. 24  It is worth saying here that in the Anabasis Xenophon does not draw attention to his status as an eyewitness. According to Grethlein, the Anabasis is told by a hetero-diegetic narrator: J. Grethlein, “Xenophon’s Anabasis from character to narrator”, JHS 132 (2012) 23–40, on p. 23.

210

Srećko Koralija

In Xenophon’s Economics, during one of the dialogues, Critoboulus asks Socrates about the Persian king’s everyday customs. In his answer, comparing the Persian παράδεισος to the Greek κῆπος, Socrates answers that the Persian king25 takes care that wherever he goes there be a παράδεισος full of the good and beautiful things that the soil produces:26 ἐν ὁπόσαις τε χώραις ἐνοικεῖ καὶ εἰς ὁπόσας ἐπιστρέφεται, ἐπιμελεῖται τούτων ὅπως κῆποί τε ἔσονται, οἱ παράδεισοι καλούμενοι, πάντων καλῶν τε κἀγαθῶν μεστοὶ ὅσα ἡ γῆ φύειν θέλει27 among the various places he inhabits and visits, he takes care there be gardens, called παράδεισοι, full of the good and beautiful things that the soil produces.

From that context, one can conclude that a παράδεισος was a place arranged for Persian kings. In his Hellenika, a παράδεισος is described as an ‘enclosed’ place with animals: καὶ θῆραι αἱ μὲν καὶ ἐν περιειργμένοις28 παραδείσοις29 and wild animals both in the enclosed παραδείσοις and in open spaces.30

This is also the case in Anabasis, which is one of the main references in lexicons, where a παράδεισος is described as a place with wild animals: ἐνταῦθα Κύρῳ βασίλεια ἦν καὶ παράδεισος μέγας ἀγρίων θηρίων πλήρης, ἃ ἐκεῖνος ἐθήρευεν ἀπὸ ἵππου, ὁπότε γυμνάσαι βούλοιτο ἑαυτόν τε καὶ τοὺς ἵππους31 Here Cyrus owned a palace and a large park full of wild beasts, which he used to hunt on horseback whenever he wished to give himself or his horses exercise.32

In the same text, παράδεισος is a place full of trees:

25  In this particular context: Pharnabazos 26 Xenophon, Oec. 4, 13. 27 Xenophon, Oec. 4, 13.14.20. In his commentary, Juan Gil asserts that in the late Greek period παράδεισος signified ‘jardin a secas’ (a dry garden); cf. J. Gil (ed.), Jenofonte, Económico (Madrid: Sociedad de Estudios y Publicaciones, 1967), 300. 28  ‘To attach around’ in: F. Montanari, The Brill Dictionary of Ancient Greek (Leiden/Boston: Brill, 2015), 1624. 29 Xenophon, Hell. 4, 1, 15, 8. 30 Xenophon, Hell. II.3.11–IV.2.8. Cf. P. Krentz (ed.), Xenophon, Hellenika (Warminster: Aris & Phillips, 1995), 111; Generally useful, but not the best translation. Some words are missing in the edition. 31 Xenophon, Anab. I.2.7. 32 English translation available at: https://www.gutenberg.org/files/1170/1170-h/1170-h. htm#link2H_4_0014 book II, para. 2.

παράδεισος as/and κῆπος

211

οἱ μὲν οὖν Ἕλληνες παρ’ αὐτὴν ἐσκήνησαν ἐγγὺς παραδείσου μεγάλου καὶ καλοῦ καὶ δασέος παντοίων δένδρων33 Greeks camped by the side of a big and nice παράδεισος which was full with all sorts of trees.

Furthermore, in Cyropaedia, a παράδεισος is a place where the young Cyrus learnt hunting: τά τε νῦν ἐν τῷ παραδείσῳ θηρία δίδωμί σοι34 I give you the wild animals that are in the παράδεισος.35

Accordingly, παράδεισος was contextually understood as a ‘hunting park’, ‘hunting lodge’ or ‘timber-reserve’.36 From Xenophon’s writings, we see that παράδεισος was a Greek rendering for a Persian enclosed space containing things that the soil produces (fruits, vegetables and trees) and that it was used as a place for kings and a park for hunting. In this respect, it has nothing to do with the notion of ‘orchard’ in the later (Hellenistic and LXX) sense of the word because it is used to explain the Persian context of its usage and is not attested in other corpora of Classical Greek. Although Xenophon says that a παράδεισος contained many beautiful things that the soil produces,37 there is a lack of additional information and a difference between his and the Hellenistic usage of the word in later periods. Apparently, Xenophon’s ultimate intention was to describe to the Greek audience, in the form of a Socratic dialogue, what was going on in Persia.

2.2 Papyri38 and Hellenistic Usage During the Hellenistic period, there is an increase in the use of παράδεισος, which is also reflected in the post-Septuagintal and Late Antiquity periods. Namely, the evidence from the papyri reveals that παράδεισος was used in the language of agriculture, which appears in administrative contexts. Beyond the evidence from the papyri and the Septuagint we know little about the context in which this increase happened. Although the papyri are much shorter and very often do not 33 Xenophon, Anab. II.4.14. 34 Xenophon, Cyr. 1.3.14; also 1.4.5; 8.1.38; 8.6.12. 35  H.G. Dakyns (ed.), Xenophon, The Education of Cyrus (London: J.M. Dent & Sons Ltd, 1992), 15. 36  J.P. Brown (ed.), Israel and Hellas: Vol. 3, The Legacy of Iranian Imperialism and the Invividual? (BZAW 299; Berlin/New York: De Gruyter, 2001), 122. 37 Xenophon, Oec. 4, 13.14.20. 38  All the references to the Greek texts (and their English renderings) of the papyri are available at: http://papyri.info/browse/ddbdp/.

212

Srećko Koralija

provide enough contextual information, it can be said with certainty that the papyri evidence escaped the problems of scribal transmission present in works of Classical Greek literature [oral traditions, manuscripts, and the like]. From papyrus Oxyrhynchus 22739 we know that Xenophon’s work Economics was circulating in Egypt, but it is difficult to connect that information with the increase of use of παράδεισος in the Hellenistic period. One of the papyri (a memorandum) speaks of arrangements for work to be done in the παράδ(ε)ισος. The context of the papyrus reveals that a παράδεισος contains trees, olive-groves and water-channels: πιδὴ ὑμᾶς δῖ τὰ ἔργα συντελοῦντας τὰ ἐν τῶι παραδίσωι ἀνηλίσκιν, παρὰ τούτων τὰ βέβαια λαβεῖν. ὡς ἂν κατεργασθῇ ὁ παράδισος, καὶ καθαρῶ αὐτόν. […]περὶ Μενάνδρου καὶ Ἀγάθωνος, ἐργᾶται Στοτοῆτις, Νεχθοσῖρις, Διόδωρος πρὸς τοῖς δένδρεσιν καὶ ὑδραγωγοῖς […] τὸν ἐλαῶνα Ὀννῶφρις τὸν ἐν τῶι παραδίσωι. Since it is necessary for you to spend money to finish the works in the orchard, take assurance from these. Once the orchard has been finished, I will cleanse it. […] Concerning Menandros and Agathon, the following will work at the trees and water-channels, Stotoetis, Nechthosiris, and Diodoros. […]. Onnophris will do the olive-grove in the orchard.40

From this evidence, one can say that a παράδεισος was cultivated primarily for its produce rather than for decoration, as was already observed by Lee in his analysis of the Zenon papyri. He suggests that the best English equivalent of παράδεισος is an orchard, since ‘garden’ implies an area planted mainly with vegetables and flowers.41 Some other papyri (including damaged examples) do not offer more contextual evidence on παράδεισος, although they confirm the widespread use of the noun during the Hellenistic and Roman periods. I propose here to group the attestation of the words under discussion according to the context of usage, covering the period from the third century b.c.e. to the third century c.e. I also decided to give as many references as possible, to allow the reader to get a fuller sense of the usage of the word. Although for many of these papyri we do not have enough information on their provenance except their title and the dating, they shed some light on the contexts and meanings. In many cases the word παράδεισος is attested in legal decrees, land surveys, lists of personal properties, and official receipts for payment of money, mainly in the context of a plot of land to be bought or sold. In some of these, παράδεισος is described as ἔρημος (παράδεισος ἔρημος): 39 B.P. Grenfell/A.S. Hunt (ed.), The Oxyrhynchus Papyri (London: Egypt Exploration Fund, 1898–1924), 120–123. 40  Memorandum to Zenon, 252–251 b.c.e.; P. Mich. 1 45. 41 Lee, A Lexical Study of the Septuagint, 54.

παράδεισος as/and κῆπος

213

a) τοῦ Ἀπολλωνίου παράδεισος ἔρημ[ος42 uncultivated43 orchard of Apollonius b) Σωστρά(του) Μαστικῶντός ἐστιν ⟦ε̣ σ̣κυ⟧ παρά(δεισος) ἐν μισ(θώσει) ἔρη(μος) … Ἀ̣ μ̣ βο τοῦ Ἰάσονος παρά(δεισος) ἔρη(μος)44 uncultivated orchard in lease of Sostrates Mastikon; uncultivated orchard of Ambos Iason.

Some papyri also mention the owner’s name whereby it becomes evident that a παράδεισος could be a personal property: c) π̣ [α]ρ̣ ά̣δ̣εισος Ἐμβείους45 Embeious’ orchard d) Ἀναξαγόρας ὁ γαμβρός σου πωλεῖ ἀρούρας τριάκοντα ἐν Σιναπὴ ἐν οἷς ἐστὶν ὁ παράδεισ[ο]\ς ἐξ ἀπηλιώτου τῶν σῶν γ ἀρουρῶν46 Anaxagoras, your brother in law, sells thirty fields in Sinape where there is an orchard on the east of your fields which you intend to turn into a vineyard

Later papyri inform us about what a παράδεισος looked like: e) παράδεισος περικεκρηπειδωμένος47 orchard surrounded by many springs f) περὶ κώ(μην) Ἀλαβανθ(ίδα) παράδ(εισος) ἐν ἱερ(εῖ) τάξι [ -ca.?- ]48 Around the village of Alabanthida, an orchard in a suitable order49

42  Decree of Euergetes II concerning Associations, and Purchase of Property, Tebtynis, 124/123 b.c.e.; P. Tebt . 3.1 700. 43  Because of the lack of more contextual information, it is not easy to render ἔρημ[ος into English. If a παράδεισος was cultivated for its produce, it is difficult to render ἔρημ[ος by ‘uncultivated’, though it is possible. We do not have enough contextual information. The reason I chose to render ἔρημος by ‘uncultivated’ is the distinction between the meaning and the context. It is possible that once cultivated, a παράδεισος became uncultivated. More examples: Testament eines Kleruchen, Herakleopolis, 110 b.c.e.; BGU 6 1285, Official Accounts, Thmuis (Mendesios), 101–200 c.e., P. Ryl . 2 215. 44  Account of Land and Survey of Crops, Arsinoites, 178–138 b.c.e.; P. Tebt. 3.2 1001 R. 45  List of House Property, Tebtynis, 225/176 b.c.e.; P. Tebt. 3.2 834. Similar examples: Bodenpachtvertrag, Arsinoites, 158/157 c.e.; SB 16 12373, Land Survey at Arsinoe, Krokodilopolis, 125–101 b.c.e.; P. Tebt . 1 86 R, Document on the Episkepsis at Karanis, Karanis (Arsinoites), 80/81 c.e.; P XV. Congr.15. 46  Brief des Epaphrodeitos an Apollonios, Hermopolites, 116–120 c.e.; P. Giss. 1 13. 47  Concerning House Property, Oxyrhynchos, 101/200 c.e., P. Osl o 3 108. 48  Κτηματικὸς κατάλογος, Arsinoites, 131 c.e.; P. At hen. 43 R. 49  More examples: Fragment einer Liste von Steuereingängen, Karanis, 183/184 c.e.; SB 12 11115; Eine Steuerliste aus Pheretnuis, Pheretnuis, 195/196 c.e.; P. Pher.; Untitled, Hermopolis, after 249 c.e.; P. Lond. 3 1157 V(a) (S. 109); Abstracts of Contracts, Oxyrhynchos, 176–200 c.e.;

214

Srećko Koralija

In the later period (second century c.e.), the word is also used as a personal name: g)

Ὡρίων ὁ καὶ Παράδει[σο]ς̣ καὶ κοιν̣(ωνοὶ) διὰ50 Orion and Paradeisos51 and companions

It is clear from consultation of the papyri that a παράδεισος could contain trees, olive groves, springs and water-channels and that it could be uncultivated, watered, and furnished. While many of them do not contain much evidence about the meaning, they certainly reveal the context of its usage: i.e., that it was a private, cultivated property. The strongest reason why it should be preferred to translate it as ‘orchard’ is the above-mentioned distinction from the English notion of ‘garden’ and, consequently, the Greek κῆπος. The noun κῆπος is also attested in the papyri as a private property: ἔκ τε τοῦ ἡμετέρου κήπου [καὶ ἐ]κ τῆς ἄκρας52 from our garden and from the hill.

A letter from Apollonios to Zenon reveals that a κῆπος could contain pear trees and young plants: ἀπίων καὶ φυτὰ … ἔκ τε τοῦ ἡμετέρου κήπου pear(s) and plants… from our garden.53

Some other papyri also mention κῆπος but it is impossible to see the context of usage: τῆι γυναικὶ εἰς κῆπον54 (with) the woman to the garden.55

That κῆπος can contain fruits is also supported by other evidence from two ­papyri: P Oxy. 14 1648; Bulletin de recensement, Oxyrhynchos, 159–160 c.e.; P. Rein. 2 93; Calculation of the Land Tax in Kind, Sko (Oxyrhynchites), 101–300 c.e.; P. Oxy. Hel s. 22. 50  Grundsteuerausfalle, Hermopolis, 1–200 c.e.; P. St rasb. 1 23. 51  It is probably not surprising that ‘Paradeisos’ was a proper name. For example, in Strabo and Pliny paradise is used as a proper noun (cf. Brown, The Legacy of Iranian Imperialism and the Individual, 126). 52  Letter from Apollonios to Zenon, Kharabet el Gerza, 256–255/ b.c.e.; P. Mich. 5 282. 53  Letter from Apollonios to Zenon, 256 b.c.e., P. Cair.Zen. 2 59156. 54  Untitled, III–II b.c.e., O. St ras. 1 585. 55  Similar cases are: Privatbrief, Alexandrien, 14/13 b.c.e.; BGU 4 1141 and Pachtzinsquittung aus Theadelphia, Theadelphia, 169–70.

παράδεισος as/and κῆπος

215

κ̣ήπου φοι̣ν̣ι̣κῶνο[ς]56 a garden with date-palms κήπου λαχανευομένου57 a garden with vegetables.

The context is similar to a passage in the Septuagint; see the next section.58 Furthermore, compound words may suggest that the Hellenistic context distinguished between κῆπος and παράδεισος. Although we do not have enough contextual evidence for a firm claim, the very fact that compound words exist strongly suggests the possibility of such a distinction: The word κηποπαράδεισος refers to a garden-orchard as a private property with palms and olives: a) ὁ πρ̣ [ογε]γ̣ρ̣[αμμένος κηποπαράδεισος59 prescribed garden-orchard b) ἐκ τοῦ πρὸς νότον μέρους τοῦ ὑπάρχοντός συ τῇ Εἱστίᾳ ἐν κόμῃ Κερκεσούχον Ὄρους κηποπαραδίσω60 on the south side of the garden-orchard that belongs to you, Hestia, in the village of Kerkesoucha Orous c) ἀπη(λιώτου) Ἡρωίδου τοῦ Ἀπολλωνίου ἐλαιονοπαράδει61 on the east an olive-grove orchard of Herodes (son) of Apollonius d) Φοινικωνοπαραδίσου … καὶ διο͂ρυξ δ̣ι̣ʼ οὗ ποτίζετα(ι ὁ) παρά(δεισος)62 palm orchard […] and the canal through which the orchard is watered.

Apart from the papyri, παράδεισος and κῆπος are used in literary compositions which also point to a distinction between the two words. For example, Diodorus Siculus (90–30 b.c.e.) mentions both κῆπος and παράδεισος in the same context: ἔφθειρε τούς τε κήπους καὶ τὸν παράδεισον τὸν Τισσαφέρνους63 he destroyed gardens and Tissaphernes’ orchard.

56  Summons, P. Yadin 123, 130 c.e. 57  Affitto di terreno, P. Flor. 1 16, 239 c.e. 58  Deut 11:10. 59  Vendita di aree, Tebtynis, 1–100 c.e. [PSI 8 917]; a similar example is: Sale of Part of a Vacant Lot, Arsinoites, 1–100 c.e. 60  Sale of Part of a Vacant Lot, Tebtynis, Polemonos meris, 1–99 c.e.; P Mich.:5:282. 61  Tax Receipt on the Sale of a Vineyard, Ptolemais Euergetis (Arsinoites), 78 b.c.e.; P. Ludg. Bat. 25 21. 62  Sale of Part of a Palm Grove, Arsinoites, 45/46 c.e.; P. Mich. 5 272. 63  Library, book 14, ch 80, sec 2; Diodorus Siculus, The Library of History (London: Heinemman; Cambridge, MA: Harvard University Press, 1933–67).

216

Srećko Koralija

Here the contextual distinction between the two words is apparent;64 otherwise he would not have used both words in the same sentence. In comparison to the word’s appearance in Classical Greek, it is evident that the word was used more frequently and that it became closer, though distinct in context from the word κῆπος.

3. Josephus With the Septuagint (third/second century b.c.e.) and its reception, παράδεισος also started to be used in theological contexts with biblical allusions. For example, Flavius Josephus (37–100 c.e.), who was acquainted with the Septuagint, refers to the Genesis narrative: Φησὶ δὲ τὸν θεὸν καὶ παράδεισον πρὸς τὴν ἀνατολὴν καταφυτεῦσαι παντοίῳ τεθηλότα φυτῷ65 He says that God planted a paradise in the east, flourishing with all sorts of trees. Παράδεισος is also a place for kings: Ἀδωνία παρασκευασαμένου δεῖπνον ἔξω τῆς πόλεως παρὰ τὴν πηγὴν τὴν ἐν τῷ βασιλικῷ παραδείσῳ.66 Adonijah prepared a meal outside the city near the source in the royal orchard.

One can say that the later period continues the Hellenistic usage of the word, which was now used in biblical translations and theological discussions where appropriate (e.g. Flavius Josephus), without losing its usual meaning.

3.1 Septuagint67 As we have seen in the papyri, παράδεισος was mainly used for ‘orchard’ in the Hellenistic period. In this respect, the LXX usage of παράδεισος is in close agreement with the Koine words. This view has been given its most detailed presen-

64  Library, book 16, ch 41, sec 5. 65 Josephus, A.J. 1.37. 66 Josephus, A.J. 7.347. See also 9.225 (τοὺς παραδείσους τοὺς βασιλικούς) and 10.226 (καὶ κατεσκεύασε τὸν καλούμενον κρεμαστὸν παράδεισον “and erected a pleasant suspended orchard”). 67  Unless otherwise indicated, the version used for the Greek Biblical text is A. Rahlfs/ R. Hanhart, Septuaginta Editio Altera (Stuttgart: Deutsche Bibelgeselschaft, 2006), and English renderings are taken from the online edition of NETS available at: http://ccat.sas.upenn.edu/ nets/edition/.

παράδεισος as/and κῆπος

217

tation in Lee’s work.68 In the LXX, Hebrew69 ‫ גן‬and ‫ גנה‬are rendered by both παράδεισος and κῆπος while ‫ פרדס‬is rendered by παράδεισος70 only. In the book of Genesis, παράδεισος renders the Hebrew ‫ גן‬in meanings where ‫ גן‬is either connected with the Hebrew ‫( עדן‬e.g., Gen 2:8, 15; 3:23–24 ‫ע ֶדן‬-‫ן‬ ֵ ַ‫ ּג‬/ ‫ּב ֵע ֶדן‬-‫ן‬ ְ ַ‫ )ּג‬or alludes to it (e.g., Gen 2:9, 16; 3:1, 2, 3, 8, 10). It is used to render the Hebrew notion of the ‘garden of Eden’ by ‘orchard of delight’71 (e.g. Gen 3:23). The context of the Gen 2:8–9 narrative requires the sense of God’s perfect orchard, given to human beings who did not have to toil to arrange it, but are asked to preserve it: Gen 2:15: καὶ ἔθετο αὐτὸν ἐν τῷ παραδείσῳ ἐργάζεσθαι αὐτὸν καὶ φυλάσσειν and put him in the orchard to till and keep it.

We have seen in the papyri that a παράδεισος was cultivated. The context of the book of Genesis does not allow us to say that παράδεισος was cultivated by human beings, but it does suggest that God planted it and that, inter alia, it contained fruits. Therefore, one can translate παράδεισος as ‘orchard’ in the Septuagint as well. In the same context (Gen 2:8), παράδεισος is also said to be planted by God: Καὶ ἐφύτευσεν κύριος ὁ θεὸς παράδεισον ἐν Εδεμ κατὰ ἀνατολὰς καὶ ἔθετο ἐκεῖ τὸν ἄνθρωπον, ὃν ἔπλασεν παράδεισον ἐν Εδεμ and the Lord God planted an orchard in Edem toward the east, and there he put the man whom he had formed.

In the book of Jeremiah, God is the subject of the sentence ordering the planting of orchards and gardens: Οἰκοδομήσατε οἴκους καὶ κατοικήσατε καὶ φυτεύσατε παραδείσους καὶ φάγετε τοὺς καρποὺς αὐτῶν (Jer 29:5)72

Here it is clear, as in the case of some of the papyri we discussed, that the Septuagint distinguishes between the two words. Furthermore, the word παράδεισος is used in the Septuagint in the context of ethical parables:

68 Lee, A Lexical Study of the Septuagint, 53. 69 For the Hebrew text I use A. Tal (ed.), Biblia Hebraica Quinta, Genesis (Stuttgart: Deutsche Bibelgeselschaft, 2016). 70  Only in Song 4:16. 71 NETS. 72  Jer 36:5 MT.

218

Srećko Koralija

χάρις ὡς παράδεισος ἐν εὐλογίαις, καὶ ἐλεημοσύνη εἰς τὸν αἰῶνα διαμενεῖ. (Sir 40:17) kindness is like an orchard with blessings, but an act of charity will endure forever.73

The Septuagint also uses κῆπος to render ‫ גן‬and ‫ גנה‬in LXX passages, but not in the book of Genesis. In Deut 11:10 ‫ גן‬is rendered by κῆπος to designate a place for vegetables: κῆπον λαχανείας garden of herbs;74 garden of vegetables. In some other places, κῆπος is used in the context of personal property: ἐτάφη ἐν τῷ κήπῳ τοῦ οἴκου αὐτοῦ (2 Kgs 21:18) was buried in the garden of his house;75 a place of sacrifices (in the negative sense): αὐτοὶ θυσιάζουσιν ἐν τοῖς κήποις (Isa 65:3) they sacrifice in the gardens.

We have seen that in the Septuagint κῆπος is mainly used to designate a place containing vegetables and seeds and is found in the context of personal property. There is also sufficient attestation that the creation narrative of the book of Genesis prefers παράδεισος to κῆπος in the creation narrative and in contexts where God is the subject.

3.2 Lexicon More recent lexicons76 of the Greek language (e.g., LSJ, Montanari, Baily, Beekes, Chantraine, Muraoka) recognize levels of παράδεισος and κῆπος usage, but are not always accurate about the period they are covering. LSJ defines παράδεισος as 73  Additional examples are Sir 40:27, Isa 1:30, Est 7:7–8. 74 Muraoka, A Greek-English Lexicon of the Septuagint. 75  Also in Song 4:16, 5:1. 76  For the sake of space and clarity, I do not discuss ancient lexicons. However, I think it is appropriate to mention one example here in the note: the word παράδεισος is attested in the Hesychius lexicon (fifth/sixth century c.e.) where it is defined in t he fol l owing way: παράδεισος: τίθεται ἐπὶ τῶν ἀναισθήτων, οἷς ἐστιν ἐμπεριπατεῖν (com.ad.fr. 523 K.—A). ἢ τόπος [ἒνυνδρος ἢτοι] εὒυδρος, ἐν ᾧ περίπατοι. καὶ ἡ βασιλέως κατάλυσις paradeisos: it means one of the imperceptible ones; to them to whom it is to walk about; or a [well-watered] place abounding in waters, in which there are walks; and the resting place of the kings. In the same lexicon, παράδεισος is the translation of κῆπος that is defined as παράδεισος παρά Πέρσοις for more information, see P.A. Hansen/I.C. Cunningham (ed.), Hesychii Alexandrini Lexicon: editionem post Kurt Latte continuans recensuit et emendavit (Berlin: De Gruyter, 2005).

παράδεισος as/and κῆπος

219

a) ‘enclosed park’ or ‘pleasure-ground’ where it is used to describe a) the parks of the Persian kings and nobles; b) ‘garden’, ‘orchard’, c) the garden of Eden, and d) ‘a stupid fellow’.77 In his lexicon of Classical Greek, Montanari,78 actually following LSJ, explains παράδεισος as: ‘park’ (esp. of the Persian king or nobility), ‘place planted with trees and stocked with game’ (where he gives Xenophon as the main reference) Furthermore he defines it as: a ‘garden’ or ‘orchard’ (giving some biblical references), and ‘terrestrial paradise’ (referring mainly to Philo of Alexandria and to the NT). He also gives a meaning already attested in Hesychius: ‘stupid’, ‘obtuse’.79 Muraoka’s definition is that παράδεισος is an ‘enclosed spacious plantation for cultivating fruits’, ‘orchard’, ‘luxuriant garden’ and the like.80 According to Baily, παράδεισος is ‘parc, lieu planté d’arbres où l’on entretient des animaux’.81 Furthermore, Beekes defines παράδεισος as an ‘enclosed park with animals’, ‘garden’ in general, and garden of Eden (LXX).82 It is curious how data from the literature and the papyri can easily be overlooked or discounted in lexicons. In contrast to Muraoka, Beekes would thus miss the discussion about παράδεισος meaning orchard in the LXX and give it only the meaning of the garden of Eden. Concerning κῆπος Montanari defines it as: garden, orchard, plantation, enclosure (for the Olympic games), cepus (type of hairstyle), and female genitals.83 Chantraine explains κῆπος as ‘jardin, verger’.84 Muraoka defines κῆπος as ‘orchard’85 but supports it with ‘garden of herbs’ from Deut 11:10. Furthermore, Beekes defines κῆπος: ‘garden, orchard, plantation’ or ‘uncultivated/barren piece of land’.86 It seems to me that lexicographers understood παράδεισος and κῆπος as synonyms. One can say that all lexicons, no matter whether of Classical, Hellenistic or Septuagint Greek, actually offer similar solutions to the reader of the entry, and easily overlook subtleties of their meaning and context. Many recent lexicons also follow LSJ, which does not always do justice to the LXX, although it uses it to support the meanings. During the Koine period, κῆπος and παράδεισος were common words used in many contexts, and it is hard to delineate their meanings if we do not take into consideration as much context of their use as possible.

77 Liddle/Scott/Jones, Greek-English Lexicon. 78  F. Montanari, The Brill Dictionary of Ancient Greek (Leiden: Brill, 2015). 79 Montanari, The Brill Dictionary of Ancient Greek, 1546. 80 Muraoka, A Greek-English Lexicon of the Septuagint, 525. 81  A. Baily, Dictionnaire Grec-Français (Paris: Hachette, 1950), 1461. 82 Beekes, Etymological Dictionary of Greek, 1151. 83 Montanari, The Brill Dictionary of Ancient Greek, 1123–4. 84  P. Chantraine, Dictionnaire étymologique de la langue grecque. Histoire des mots (Paris: Librairie Klincksieck, 2009), 505. 85 Muraoka, A Greek-English Lexicon of the Septuagint, 396–7. 86 Beekes, Etymological Dictionary of Greek, 688.

220

Srećko Koralija

4. Conclusion In this brief survey we have had the opportunity to consider some contexts of the usage of παράδεισος and κῆπος from Xenophon’s period and their usage in the Septuagint up to their usage in Late Antiquity. We have seen that by the 3rd century b.c.e. παράδεισος had become an ordinary term as has already been recognized by Lee87 and have examined the usage of κῆπος in the Septuagint. We can summarize our conclusions as follows: 1. Putting Old Persian variants as etymological explanations into a Greek dictionary for the word παράδεισος can mislead the readers since the Old Persian usage does not allude to any kind of garden or orchard in the Koine sense of the word. The Iranian loan word in the Greek and Semitic sources has the shape of the Avestan word but the meaning of the Old Persian paridaidam (a pleasant retreat88). the Avestan pairidaēza denotes a barren plot of land surrounded by a wall. In this respect, the Avestan word pairidaēza, strictly speaking, does not overlap with the Greek παράδεισος. Depending on the context, etymological explanations for παράδεισος probably make sense in dictionaries of Classical Greek where the word is still not in general use, but should not appear in dictionaries of Hellenistic/Koine and Septuagint Greek where the word is actually Greek. 2. We have also seen that in Xenophon’s writings παράδεισος was an enclosed space containing things that the soil produces (fruits, vegetables and trees)89 and that it was used as a place for kings and a park for hunting.90 In this respect, it has nothing to do with the notion of ‘orchard’ in the later (Hellenistic and LXX) sense of the word because it is used to explain the Persian context of its usage and is not attested in other corpora of Classical Greek. During the same period κῆπος was a widely used word for ‘garden’, as already attested in the Homeric corpus.91 3. It is clear from consultation of the papyri we covered in this study92 that a παράδεισος could contain trees, olive-groves, springs and water-channels and that it could be uncultivated, watered, and furnished. On the other hand, a κῆπος was understood in a similar way as παράδεισος. However, compound words may suggest that later Koine distinguished between κῆπος and παράδεισος. In the Hellenistic period we do not find clear evidence that a παράδεισος was an enclosed place for hunting animals, as was the case 87 Lee, A Lexical Study of the Septuagint, 54. 88 Panaino, “No Room for the ‘Paradise’? About Old Persian paradayadama”, 151. 89 Xenophon, Oec. 4, 13.14.20. 90 Xenophon, Cyr. 1.3.14; also 1.4.5; 8.1.38; 8.6.12. 91 Iliad Θ 306; Cunliffe/ Okla, A Lexicon of the Homeric Dialect, 226. 92  See the section “Classical Greek” above.

παράδεισος as/and κῆπος

221

in Xenophon’s writings. It denoted an orchard, while κῆπος was more like a garden with vegetables and young plants. This means that the two words are not synonyms and that they should be contextually distinguished. 4. We have seen that in the Septuagint there are a good number of examples where παράδεισος is used in the sense of ‘orchard of delight’,93 planted by God (reflecting the Hebrew ‘garden of Eden’), in contexts where God is the subject.94 In some Biblical books it is also used in the context of ethical parables.95 The word κῆπος is mainly used to designate a place containing vegetables and seeds and is found in the context of personal property. There is also sufficient attestation that the book of Genesis prefers παράδεισος to κῆπος in the creation narrative and in contexts where God is the subject. Regarding lexicons, one can say that the majority of them, no matter whether of Classical, Hellenistic or Septuagint Greek, actually offer similar solutions to the reader of the entry, and easily overlook subtleties of their meaning and context. Given the evidence from the papyri and contextual analysis of biblical books, one should distinguish between κῆπος and παράδεισος, depending on the context of usage.

93  E.g. Gen 3:23. 94  E.g. Gen 2:8. 95  E.g. Sir 40:27, Isa 1:30.

Ilmari Soisalon-Soininen The Use of the Genitive Absolute in the Septuagint1 1973 Abstract The occasional instances of the genitive absolute in the Septuagint allow us glimpses of the translators’ competence, which otherwise was constrained by factors that encouraged a literal output. This holds true for even the most literal translators. The genitive absolute, which has no formal counterpart in Hebrew, is used mainly for the rendering of Hebrew time designations of various kinds, expressed through paratactic clauses, ‫ ב‬+ inf. cs., ‫ כ‬+ inf. cs., and other constructions. It is also used to bring concessive or conditional nuances across. Finally, prepositional phrases (often time designations) have been rendered as a genitive absolute. A study of idiomatic Greek features that lack a Hebrew counterpart will give us more insight into the translators’ competence.

1. Introduction [175] The syntax of the Septuagint has been studied rather sparsely. Although several topics have already been treated in separate publications, it will take much time and effort before we can even begin to draw a comprehensive picture. But every individual piece of research, as great or small as it may be, brings us one step closer to the grand objective we will hopefully reach one day, viz. the comprehensive presentation of Septuagint syntax. Some basic features have been clear since Thackeray’s Grammar.2 Septuagint Greek is translation Greek. With respect to morphology and generally also to vocabulary, the Septuagint fits into contemporary Koine. The syntax, however, 1  This is a translation of “Der Gebrauch des genitivus absolutus in der Septuaginta,” in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 175–80, by Theo van der Louw. 2  H.St.J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint I (Cambridge: University Press, 1909).

224

Ilmari Soisalon-Soininen

clearly shows that we are dealing with translationese. Hebrew interference is so strong that in many passages every sentence testifies to it. Unidiomatic expressions often neighbour on each other. One could say that it is not Greek but rather Hebrew with Greek words. Some expressions must definitely have been unintelligible to a Greek. And yet, from the language of the Septuagint it is a long way to that of Aquila, who wanted to reflect the character of the Hebrew parent text. [176] Thackeray also studied the character of the individual LXX books and categorized them according to their linguistic features. Some parts have been translated literally, others more freely, but everywhere the Hebrew background is palpable. It would be a mistake to measure the Septuagint translation against our present-day standards. After all, it is the first translation of a major literary corpus from one language to another. The translators did groundbreaking work. They did not have philological training, and we should assume that at first they operated purely intuitively. Besides, it was Holy Writ that was to be translated, which required the greatest fidelity possible. To the intended audience this posed no problem, by the way: the Septuagint is not a translation by a Greek for Greeks, but by a Greek-speaking Jew for a Jewish audience. Many expressions incomprehensible to Greeks were probably transparent to Greek-speaking Jews. These factors constrained the translators in the execution of their task so that their true feel for the language and their competence are not that visible. If one carefully surveys the renderings of any Hebrew idiom, occasional free renderings that testify to a subtle feel for the language show up among the literal renderings. In those cases, the constraining factors stepped back for some reason. These cases are particularly interesting because they allow us a glimpse of the translator’s personal traits, which otherwise would have remained hidden. Conversely, if we choose a typically Greek expression that has no counterpart in Hebrew, the resulting small range of instances will provide examples of free renderings. Such a linguistic phenomenon is the Greek genitive absolute, which is the topic of our paper. It would be proper to look at the participium conjunctum in connection with it, but that would take up too much space. For the same reason we will limit ourselves to the canonical books, in other words, those that have been translated from Hebrew and whose parent text is known. In the books of the Septuagint included in the Hebrew canon, the genitive absolute occurs a little more than 200 times. Thus, it occurs, but rarely. There is a noticeable difference between the various books. For example, in the rather [177] freely translated books of Job and Proverbs it occurs 21 and 22 times, respectively. This means one genitive absolute per 43 and 41 verses, respectively. In contrast, the book of Psalms features only 5 cases, in other words, one genitive absolute per 505 verses. The frequencies are so low that these figures alone do not warrant any conclusions. On the other hand, even the books that have the reputation of being exceptionally literal contain some cases. Judges and Ezekiel feature

The Use of the Genitive Absolute in the Septuagint

225

6 cases each. In sum, the genitive absolute occurs only sporadically, but it seems that none of the translators avoided it on principle.

2. Temporal Use of the Genitive Absolute Almost all cases of the genitive absolute in the Septuagint have a temporal meaning (for other senses, see below). The rule for classical Greek, that the subject of a genitive absolute does not occur in the main clause, is not observed, but this is very common in Koine Greek in any case. It is exceedingly rare for the main clause and the genitive absolute to have the same subject. A phenomenon that is peculiar from the perspective of Greek is that, in the books from Judges to 2 Kings, a main clause, when following after the genitive absolute, often begins with καί. So, e.g., 1 Sam 9:27 αὐτῶν καταβαινόντων εἰς μέρος τῆς πόλεως καὶ Σαμουὴλ εἶπεν τῷ Σαούλ. It then always reflects Hebrew ‫ו‬. In this way, a characteristic feature of literal translation, apodotic καί, crept into an otherwise free rendering. Almost all instances of the genitive absolute in the canonical books of the Septuagint are good renderings of the corresponding Hebrew text. Some are minor additions to the text, and sometimes a genitive absolute occurs in a way that does not reflect the source text. But these are exceptions. As the mainly temporal sense of the genitive absolute in the Septuagint suggests, the majority of the renderings reflect Hebrew designations of time. The clearest category concerns renderings of ‫ ב‬+ infinitive construct, approximately 60 cases. Some examples: ‫ר־ע ָשה‬ ָ ‫זָ כֹור ֵאת ֲא ֶש‬ ‫ֹלהיָך ְל ִמ ְריָם ַב ֶד ֶרְך‬ ֶ ‫יְ הוָ ה ֱא‬ ‫את ֶכם ִמ ִמ ְצ ָריִם‬ ְ ‫ְב ֵצ‬ ‫יקים‬ ִ ‫ִב ְרבֹות ַצ ִד‬ ‫יִ ְש ַמח ָה ָעם‬ ‫ּוב ְמש ֹל ָר ָשע‬ ִ ‫יֵ ָאנַ ח ָעם‬

μνήσθητι ὅσα ἐποίησεν κύριος ὁ θεὸς τῇ Μαριὰμ ἐν τῇ ὁδῷ ἐκπορευομένων ὑμῶν ἐξ Αἰγύπτου. ἐγκωμιαζομένων δικαίων εὐφρανθήσονται λαοί, ἀρχόντων δὲ ἀσεβῶν στένουσιν ἄνδρες.

Deut 24:9

Prov 29:2

A genitive absolute or in some cases a participium conjunctum would almost always be an appropriate rendering of ‫ ב‬+ inf. cs. But none of the LXX translators made it a rule [178] to render it so. The literalistic rendering ἐν τῷ + inf. is used almost everywhere in great quantities. In addition, temporal clauses occur very frequently. These renderings are especially numerous in the free translations, such as Job and Proverbs, which both avoid ἐν τῷ + inf. Although the temporal clause is a correct rendering, the genitive absolute is more elegant and evinces a good feel for the language.

226

Ilmari Soisalon-Soininen

Table 1: Renderings of ‫ ב‬+ inf. cs. in literally and freely translated books

Prov Job Judg Ezk

ἐν τῷ + inf.

subord. clause

gen. abs.

10 88

16 10 2 14

5 2 -

part. coni. total ‫ ב‬+ inf. cs. 1 1 1

26 16 17 127

The data provided in this table suggest that the relative rarity of the genitive absolute is an indicator of the translator’s feel for the language. Very close to ‫ ב‬+ inf. cs. is the temporal expression ‫ כ‬+ inf. cs. These two may often have been confused in the manuscripts. That explains why ‫ כ‬+ inf. cs. is sometimes rendered as ἐν τῷ + inf. However, the most common rendering is a temporal clause, mostly with ὡς, a conjunction that rarely renders ‫ ב‬+ inf. cs. Sometimes the genitive absolute is employed, but these cases are so rare that one cannot determine if the translator perhaps read ‫ ב‬+ inf. cs. there. In any case, the genitive absolute is very rare as a rendering of ‫ כ‬+ inf. cs. In Hebrew, a coordinated clause with ‫ ו‬and sometimes also an asyndetic clause may have a function that in our languages is best expressed with a temporal subordinate clause. Most instances of the genitive absolute in the canonical books of the Septuagint are renderings of such clauses. We cannot possibly present the total figure of how often these Hebrew clauses occur or the various ways they have been rendered in the Septuagint; we can only take note that, sporadically, such clauses have been rendered very elegantly with a genitive absolute in the most diverse books. The usual rendering is the literal one, namely a coordinated clause, which does not express the original idea very adequately. Since the function of these Hebrew clauses is much wider than that of strictly temporal clauses, it was the translators’ understanding of the function of the clauses in question that prompted them to come up with a genitive absolute. Some examples: ‫ֵה ָמה ָבאּו ְב ֶא ֶרץ צּוף וְ ָשאּול‬ ‫ר־עּמֹו‬ ִ ‫ָא ַמר ְלנַ ֲערֹו ֲא ֶש‬ ָ‫ַּת ְע ִּתיר ֵא ָליו וְ יִ ְׁש ָמ ֶעּך‬

αὐτῶν ἐλθόντων εἰς τὴν Σὶφ καὶ Σαοὺλ εἶπεν τῷ παιδαρίῳ αὐτοῦ τῷ μετʼ αὐτοῦ εὐξαμένου δέ σου πρὸς αὐτὸν εἰσακούσεταί σου

1 Sam 9:5 Job 22:27

[179] In sentences with ‫ ו‬+ subject + participle (= predicate) the temporal function stands out particularly. Renderings of such clauses with a genitive absolute occur in various books, as the following examples show:

The Use of the Genitive Absolute in the Septuagint

‫וַ יֵ ָרא ֵא ָליו יהוה ְב ֵאֹלנֵ י‬ ‫ַמ ְמ ֵרא וְ הּוא י ֵֹשב ֶפ ַתח־‬ ‫ָהא ֶֹהל ְכחֹם ַהּיֹום׃‬ ‫וַ יְ ִהי ֵהם ק ְֹב ִרים ִאיׁש וְ ִהנֵ ה‬ ‫ת־הגְ דּוד‬ ַ ‫ָראּו ֶא‬

Ὤφθη δὲ αὐτῷ ὁ θεὸς πρὸς τῇ δρυὶ τῇ Μαμβρή, καθημένου αὐτοῦ ἐπὶ τῆς θύρας τῆς σκηνῆς αὐτοῦ μεσημβρίας. καὶ ἐγένετο αὐτῶν θαπτόντων τὸν ἄνδρα καὶ ἰδοὺ εἶδον τὸν μονόζωνον

227 Gen 18:1

2 Kgs 13:21

The temporal meaning stands out even clearer if the clause also has ‫עוד‬. In fact, a clause with ‫ בעוד‬already corresponds to ‫ ב‬+ inf. cs. Around 25 cases have been rendered with a genitive absolute in the whole OT. Compare the following examples, the first two with ‫עוד‬, the second pair with ‫בעוד‬: ‫יתה‬ ָ ‫הּודה וְ ֶא ָחיו ֵב‬ ָ ְ‫וַ יָבֹא י‬ ‫עֹודּנּו ָשם‬ ֶ ‫יֹוסף וְ הּוא‬ ֵ ‫זֶ ה ְמ ַד ֵּבר וְ זֶ ה ָּבא‬ ‫עֹוד‬ ‫וַ יְ ַׁש ְּל ֵחם ֵמ ַעל יִ ְצ ָחק ְּבנֹו‬ ‫עֹודּנּו ַחי‬ ֶ ‫ְּב‬ ‫עֹודנִ י ַחי ִע ָמ ֶכם ַהּיֹום‬ ֶ ‫ְב‬ ‫יִתם ִעם־יהוה‬ ֶ ‫ַמ ְמ ִרים ֱה‬

Εἰσῆλθεν δὲ Ἰούδας καὶ οἱ ἀδελφοὶ αὐτοῦ πρὸς Ἰωσὴφ ἔτι αὐτοῦ ὄντος ἐκεῖ Ἔτι τούτου λαλοῦντος ἦλθεν ἕτερος ἄγγελος καὶ ἐξαπέστειλεν αὐτοὺς ἀπὸ Ἰσαὰκ τοῦ υἱοῦ αὐτοῦ ἔτι ζῶντος αὐτοῦ ἔτι γὰρ ἐμοῦ ζῶντος μεθʼ ὑμῶν σήμερον παραπικραίνοντες ἦτε τὰ πρὸς τὸν θεόν

Gen 44:14 Job 1:16, 17, 18 Gen 25:6 Deut 31:27

3. Other Uses of the Genitive Absolute There are only a few cases of a conditional,3 concessive4 or causal5 sense. Some sentences with ‫ אין‬have been rendered with a genitive absolute. These cases sometimes have a conditional or concessive meaning. Compare these two very adequate renderings of the conditional and concessive sense, respectively: ‫נּוכל ִל ְראֹות ְפנֵ י‬ ַ ‫ִכי־לֹא‬ ‫ָה ִאיׁש וְ ָא ִחינּו ַה ָקטֹן ֵאינֶ ּנּו‬ ‫ִא ָתנּו‬ ‫ת־ח ֶרב וְ נָ ְפלּו‬ ֶ ‫וְ נָ סּו ְמנֻ ַס‬ ‫וְ ֵאין ר ֵֹדף׃‬ ‫יׁש־ב ָא ִחיו‬ ְ ‫וְ ָכ ְשלּו ִא‬ ‫י־ח ֶרב וְ ר ֵֹדף ָאיִ ן‬ ֶ ֵ‫ְכ ִמ ְפנ‬

οὐ γὰρ δυνησόμεθα ἰδεῖν τὸ πρόσωπον τοῦ ἀνθρώπου, τοῦ ἀδελφοῦ τοῦ νεωτέρου μὴ ὄντος μεθʼ ἡμῶν καὶ φεύξονται ὡς φεύγοντες ἀπὸ πολέμου, καὶ πεσοῦνται οὐθενὸς διώκοντος· καὶ ὑπερόψεται ὁ ἀδελφὸς τὸν ἀδελφὸν ὡσεὶ ἐν πολέμῳ οὐθενὸς κατατρέχοντος

3  Gen 44:26, 34; Prov 29:12, 14. 4  Lev 26:17, 36, 37. 5  Prov 3:28.

Gen 44:26

Lev 26:36–37

228

Ilmari Soisalon-Soininen

A great number of other expressions have been rendered with a genitive absolute. In most of these cases, Hebrew has a noun with a preposition. Typical examples are time designations: ‫ַּב ֲח ִצי ַה ַּליְ ָלה‬ ‫ׁשּובת ַה ָּׁשנָ ה‬ ַ ‫ִל ְת‬ ‫יקים ְּב ַמ ַּפ ְל ָּתם יִ ְראּו‬ ִ ‫וְ ַצ ִּד‬ ‫ְּב ֻחּמֹו‬

μεσούσης τῆς νυκτὸς ἐπιστρέψαντος τοῦ ἐνιαυτοῦ οἱ δὲ δίκαιοι ἐκείνων πιπτόντων κατάφοβοι γίνονται. θέρμης γενομένης

Exod 12:29 2 Sam 11:16 Prov 29:16 Job 6:17

Occasionally ‫ ְּב ָאזְ נֵ י‬has been rendered as ἀκουόντων / ἀκούοντος,7 e.g., ‫וַ ּיַ ַען ֶע ְפרֹון ַה ִח ִּתי ֶאת־‬ ‫י־חת ְלכֹל‬ ֵ ֵ‫ַא ְב ָר ָהם ְּב ָאזְ נֵ י ְבנ‬ ‫ר־עירֹו ֵלאמֹר‬ ִ ‫ָּב ֵאי ַׁש ַע‬

ἀποκριθεὶς δὲ Εφρων ὁ Χετταῖος πρὸς Ἀβραὰμ εἶπεν ἀκουόντων τῶν υἱῶν Χὲτ καὶ πάντων τῶν εἰσπορευομένων εἰς τὴν πόλιν λέγων

Gen 23:10

A special case is our last example, where a paraphrased construct state (BDB 513b) has been turned into a time designation [180]: ‫ישית‬ ִ ‫וַ יְ ִהי ַב ָשנָ ה ַה ֲח ִמ‬ ‫ַל ֶמ ֶלְך ְר ַח ְב ָעם‬

καὶ ἐγένετο ἐν τῷ ἐνιαυτῷ τῷ πέμπτῳ βασιλεύοντος Ῥοβοάμ

1 Kgs 14:25

4. Concluding Remarks When surveying the use of the genitive absolute in the Septuagint, one cannot but admire the competence of the translators and their subtle feeling for the language. In the study of the translation technique of the Septuagint, the focus has been on the literal renderings and the question of whether the resulting expressions were possible in Greek or not. And it is certainly true that, generally speaking, the language of the Septuagint is thoroughly stamped by Hebrew. The sporadic elegant renderings, however, make it clear that that there is another side to it. We should not forget that the true language competence of the translators surfaces so rarely because it was often restricted by the factors we started this paper with. If we want to learn more about the translators, we should give more weight to these rare cases. That will yield valuable insights.

6  Cf. also 1 Kgs 20 (21):22, 26; 2 Chr 36:10. 7  Cf. also Job 13:17; Ezk 9:5; 10:13.

Ilmari Soisalon-Soininen A Palette of Renderings for the Hebrew Construct State Relation in the Greek Pentateuch1 1977

Abstract The renderings of the Hebrew construct relation in the Greek Pentateuch reveal a kaleidoscopic variety of choices. For a large part, these are bound up with semantic categories, viz. when the nomen rectum denotes the material from which something is made, expresses an attribute or when the construct relation denotes a geographical entity or consists of an active participle and its object. Yet other renderings are found in scattered cases. What we find are noun + genitive, noun + dative, noun + adjective, noun + apposition, noun + prepositional phrase, participle + accusative, participle + genitive, one noun instead of two nouns (through omission or the use of a compound word), verb or relative clause. The translators regarded the construct relation as a translation unit that was rendered as a whole, and made no attempts at a uniform rendering for this construction.

1. Introduction [62] If we wish to study how the Septuagint translators proceeded, then, instead of merely analyzing the syntax of the final product, we must take Hebrew categories as our point of departure. Through this approach, important cases that are usually overlooked will surface. It is significant that Frankel already followed this principle, e.g., in his listing of the variegated renderings of genitive constructions.2 Yet it has often been neglected since then.3 It is also important to select grammatical categories that we identify nowadays but that the Septuagint 1  This is a translation of “Verschiedene Wiedergaben des hebräischen Status-ConstructusVerbindung im griechischen Pentateuch,” in A. Aejmelaeus/R. Sollamo (ed.), Ilmari SoisalonSoininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 62–70, by Theo van der Louw. 2  Z. Frankel, Vorstudien zu der Septuaginta (Leipzig: F.C. Vogel, 1841), 134–6. 3 Frankel, Vorstudien, 132–203.

230

Ilmari Soisalon-Soininen

translators did not consciously delineate. But it stands to reason that they sensed intuitively which expressions belonged to the same category. In such cases, they often opted for a single rendering even if it was unidiomatic Greek,4 or, [63] exceptionally, they could employ a free rendering. If that is so, these categories yield valuable insights for the study of the translation process. In other cases, our categories do not run parallel to those of the translators. We will keep this constantly in mind when we turn to the ways in which the Hebrew construct relation has been rendered. It is quite natural that the majority of the construct relations have been rendered with genitives. Not having detailed statistics at my disposal, I would estimate these cases in the Pentateuch at 90–95%. An array of different renderings covers the remaining instances. The following sections, which follow semantic functions that the construct relation may have, describe such renderings.

2. Nomen Rectum Denoting a Material The largest and clearest group that seems to have called for free renderings is those construct relations that denote the material from which something is manufactured or constructed (GKC §128o). Almost without exception, a Greek adjective is employed in these cases. ‫ּוכ ֵלי זָ ָהב‬ ְ ‫י־כ ֶסף‬ ֶ ‫ְּכ ֵל‬ ‫ת־בגְ ֵדי ַה ָבד‬ ִ ‫וְ ָל ַבׁש ֶא‬ ‫ל־מ ֲע ֵשה‬ ַ ‫ל־כ ִלי־עֹור וְ ָכ‬ ְ ‫וְ ָכ‬ ‫י־עץ‬ ֵ ‫ל־כ ִל‬ ְ ‫ִעזִ ים וְ ָכ‬ ‫עֹל ַב ְרזֶ ל‬

σκεύη ἀργυρᾶ καὶ χρυσᾶ καὶ ἐνδύσεται τὴν στολὴν τὴν λινῆν καὶ πᾶν σκεῦος δερμάτινον (…) καὶ πᾶν σκεῦος ξύλινον (ἀφαγνιεῖτε) κλοιὸν σιδηροῦν

Gen 24:53 Lev 16:32 Num 31:20 Deut 28:48

Such instances number around 140 in the Greek Pentateuch. Some Hebrew phrases are repeated often and are mostly rendered with the same phrase in Greek. The use of the preposition ἐκ is the second most common translation strategy:5 θυσιαστήριον ἐκ γῆς θυσιαστήριον ἐκ λίθων οἱ κόσυμβοι τῶν χιτώνων ἐκ βύσσου

Exod 20:24 Exod 20:25; Deut 27:5 Exod 28:39

4  For example, the tendency to use an infinitive instead of a participle after verbs meaning “to stop”; cf. I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965), 38–39. 5  When unnecessary, the Hebrew text will not be printed in the tables.

Renderings for the Hebrew Construct State Relation

σεμίδαλιν ἐκ πυρῶν ποιήσεις αὐτά ἐγκρὶς ἐξ ἐλαίου

231

Exod 29:2 Num 11:86

On rare occasions a genitive is employed to denote the material:7 τὰ πέταλα τοῦ χρυσίου ἀφαιρέματα χρυσίου ἐν σκεύει σιδήρου

Exod 36:10 (39:3) Exod 35:22 Num 35:16

It appears that the translators struggled most in cases where the adjectival attribute does not specify the construction as a whole but merely the nomen rectum. Such instances are infrequent (approximately 7), but they exhibit three different renderings: 1. Adjective: δύο δακτυλίους χρυσοῦς καθαρούς (Exod 30:4). Note that the adjective specifies the wrong word: the gold is pure, not the rings. 2. Genitive with attribute: καὶ ποιήσεις ἱλαστήριον ἐπίθεμα χρυσίου καθαροῦ (Exod 25:17) 3. Prepositional phrase with ἐκ, as in καὶ ποιήσεις ἀσπιδίσκας ἐκ χρυσίου καθαροῦ (Exod 28:13). A phrase with ἐκ is also used for ‫( ֲארֹון ֲע ֵצי ִש ִטים‬Exod 25:10; Deut 10:3): ποιήσεις κιβωτὸν μαρτυρίου ἐκ ξύλων ἀσήπτων.8 Under the category of “material” I subsume phrases with ‫ ָשנִ י‬and ‫ת ֵכ ֶלת‬, ְ which express colours although they originally denoted the material from which the colour was made. ‫ד־כ ִליל‬ ְ ֶ‫ְת ֵכ ֶלת ֶבג‬ ‫ְת ֵכ ֶלת ֻל ְלאֹת‬ ‫ּתֹול ַעת ֶבגֶ ד‬ ַ ‫ָשנִ י‬ ‫ַא ְרגָ ָמן ֶבגֶ ד‬

ἱμάτιον ὅλον ὑακίνθινον ἀγκύλας ὑακινθίνας ἱμάτιον κόκκινον ἱμάτιον ὁλοπόρφυρον

Num 4:6, 99 Exod 26:4 Num 4:8 Num 4:13

The frequently occurring ‫ ַת ַחׁש‬has erroneously been rendered as ὑακίνθινος, e.g., in Num 4:8, ‫ > ְב ִמ ְכ ֵסה עֹור ָת ַחׁש‬καλύμματι δερματίνῳ ὑακινθίνῳ.

6  Cf. also Exod 29:33; Lev 7:12, 8:26. 7  In the Ptolemaic papyri the genitive of the material is often replaced by ἐκ so that the language of the Greek Pentateuch reflects contemporary usage in this respect. Cf. E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften: 2,2, Satzlehre, analytische Teil (Berlin: De Gruyter, 1934), 345. 8  ‫“ שטים‬acacia tree” is rendered erroneously as ἄσηπτος “incorruptible.” 9  In Num 4:7 ἱμάτιον ὁλοπόρφυρον.

232

Ilmari Soisalon-Soininen

3. Nomen Rectum Expressing an Attribute The Hebrew construct relation is very frequently used in the attributive sense (GKC §128p), and rendered accordingly in Greek. The nomen rectum then expresses belonging or a characteristic. The great variety of the cases makes it virtually impossible to classify them. We can only give some examples. Frequently occurring nouns that are rendered as adjectives are ‫עֹולם‬ ָ and ‫ק ֶֹדׁש‬. The former is rendered αἰώνιος on many occasions, sometimes ἀέναος,10 and in Deut 32:7 we find a genitive, μνήσθητε ἡμέρας αἰώνιος, which is a good rendering of ‫זְ כֹר‬ ‫עֹולם‬ ָ ‫יְמֹות‬. The adjective αἰώνιος is collocated with various nouns, νόμος αἰώνιος, κατάσχεσις αἰωνία, θεὸς αἰώνιος, but adjectives occur too, e.g., ἐπ’ εὐλογίαις θινῶν ἀενάων (Gen 49:26), ἀπὸ κορυφῆς βουνῶν ἀενάων (Deut 33:15), ὑπὸ ἰσχὺν βραχιόνων ἀενάων (Deut 33:27). In Num 18:19 the adjective is connected to the wrong noun, διαθήκη ἁλὸς αἰωνίου (instead of αἰώνιος) ἐστίν. Depending on the context, the noun ‫ ק ֶֹדׁש‬is rendered quite inconsistently either with an adjective or a genitive. Many instances can be interpreted both ways. As a rule, ἅγιος is employed, as in στολὴ ἁγία and ἱμάτια ἅγια,11 but we also find genitives like αἱ φυλακαὶ τῶν ἁγίων and τὰ λειτουργήματα τοῦ ἁγίου.12 Note the following contrasting pairs: ‫ִבּגְ ֵדי־ק ֶֹדׁש ( ֲא ֶשר) ְל ַא ֲהר ֹן‬

‫רּומת ַה ֳק ָד ִשים לֹא‬ ַ ‫ִהוא ִב ְת‬ ‫אכל‬ ֵ ֹ‫ת‬ ‫רּומת‬ ַ ‫אכה ִל ְת‬ ָ ‫ַאל־יַ ֲעׂשּו־עֹוד ְמ ָל‬ ‫ַהק ֶֹדׁש‬ ‫ֶש ֶקל ַהק ֶֹדׁש‬

ποιήσουσιν τὴν στολὴν τὴν ἁγίαν Ἀαρών καὶ ἡ στολὴ τοῦ ἁγίου, ἥ ἐστιν Ἀαρών

Exod 28:4

αὐτὴ τῶν ἀπαρχῶν τῶν ἁγίων οὐ φάγεται μηκέτι ἐργαζέσθωσαν εἰς τὰς ἀπαρχὰς τοῦ ἁγίου

Lev 22:1214

τὸ δίδραχμον τὸ ἅγιον τῷ σίκλῳ τῶν ἁγίων

Exod 30:13 Lev 5:15

Exod 29:2913

Exod 36:6

10  Gen 49:26; Deut 33:15, 27. 11  E.g., Num 4:14; 18:3; 31:6 τὰ σκεύη τὰ ἅγια; Lev 4:6 τὸ καταπέτασμα τὸ ἅγιον (where τοῦ ἁγίου would actually be better; cf. Arm Eth sanctitatis). 12  In Num 3:28, 32; 18:5 and 7:9, respectively. 13  The same rendering occurs in Exod 39:18 (MT 41). 14  In Num 18:19 ‫ ְּתרּומֹת ַה ֳק ָד ִשים‬is also rendered as a genitive: πᾶν ἀφαίρεμα τῶν ἁγίων.

233

Renderings for the Hebrew Construct State Relation

‫(עב ַֹדת) ַהק ֶֹדׁש‬ ֲ ‫אכת‬ ֶ ‫ִל ְמ ֶל‬ [64–65]

‫ֵשם ָק ְד ִשי‬

εἰς πάντα τὰ ἔργα τοῦ ἁγίου κατὰ πᾶσαν τὴν ἐργασίαν τῶν ἁγίων

Exod 36:6 Exod 39:1 (MT 38:24)

τὸ ὄνομα τὸ ἅγιόν μου τὸ ὄνομα τοῦ ἁγίου15

Lev 22:2 Lev 22:32

If ‫ ק ֶֹדׁש‬cannot denote the sanctuary, the adjective is almost always used: τὸ ἔλαιον τὸ ἅγιον ὁ οἶκος ὁ ἅγιος

Num 35:25 etc. Deut 26:15

ἐπίκλητος ἁγία κλητὴ ἁγία

Num 28:18 etc. Lev 23:3 etc.

The intensive phrase ‫ ק ֶֹדׁש ַה ֳק ָד ִשים‬is rendered hebraistically as ἅγια (ἅγιον) τῶν ἁγίων / ἅγιον τοῦ ἁγίου. Thus, the same Hebraizing construction was used all the time, but it did not become so stereotyped as to preclude variations between singular and plural. Several nouns such as ‫צ ֶדק‬,ֶ ‫א ֶמת‬, ֱ ‫ ֶש ֶקר‬etc., occurring as a nomen rectum, typically express an attribute, and are rendered accordingly as adjectives. Some examples: ‫יפת‬ ַ ‫י־צ ֶדק ֵא‬ ֶ ֵ‫מֹאזְ נֵ י ֶצ ֶדק ַא ְבנ‬ ‫ֶצ ֶדק וְ ִהין ֶצ ֶדק‬ ‫ד־ש ֶקר‬ ֶ ‫ ֵע‬/ ‫ֵעד ָח ָמס‬

in contrast to

ζυγὰ δίκαια καὶ στάθμια δίκαια καὶ χοῦς δίκαιος μάρτυς ἄδικος

‫ֵעד ָשוְ א‬ ‫ַאנְ ֵשי ֱא ֶמת‬ ‫י־חיִ ל‬ ַ ‫ַאנְ ֵש‬

μάρτυς ψευδής ἄνδρες δίκαιοι ἄνδρες δυνατοί

‫י־צ ֶדק‬ ֶ ‫זִ ְב ֵח‬ ‫ְב ֶד ֶרְך ֱא ֶמת‬

θυσία δικαιοσύνης ἐν ὁδῷ ἀληθείας

Lev 19:36 Exod 23:1; Deut 19:18 Deut 5:20 Exod 18:21 Gen 47:5 (MT 6); Exod 18:21, 25 Deut 33:19 Gen 24:48

While there are numerous other phrases that are occasionally rendered as adjectives, we must limit ourselves to certain examples. Although in construct relations names of nations or tribes are rendered as nouns most of the time (υἱοὶ Ἰσραήλ), we do find adjectives as well: εἰσῆλθεν ὀπίσω τοῦ ἀνθρώπου τοῦ Ἰσραηλίτου ἐπὶ παρθένον Ἰσραηλῖτιν ἐν ἕλκει Αἰγυπτίῳ

Num 25:8, 14 Deut 22:19 Deut 28:27

15  Possibly the translator read it without the suffix. It was subsequently corrected to τὸ ὄνομα τὸ ἅγιόν μου in mss.

234

Ilmari Soisalon-Soininen

Sometimes cardinal points or time designations are presented as adjectives: ‫רּוח ָק ִדים‬ ַ ‫ְב ַא ְשמ ֶֹרת ַהב ֶֹקר‬ ‫ְכ ִמנְ ַחת ַהב ֶֹקר‬

ὁ ἄνεμος ὁ νότος16 ἐν τῇ φυλακῇ τῇ ἑωθινῇ κατὰ τὴν θυσίαν τὴν πρωινήν

Exod 10:13 (bis); 14:21 Exod 14:12 Exod 29:41

Other examples include: [66] ‫ְל ַמ ַען יִ ְירׁשּו (…) נַ ֲח ַלת‬ ‫ֲאב ָֹתיו‬ ‫ַאנְ ֵשי ַה ִמ ְל ָח ָמה‬ ‫ַאנְ ֵשי ִמ ְקנֶ ה‬ ‫ׁשֹוק ַהיָ ִמין‬ ‫ֹלהי ַהנֵ ָכר‬ ֵ ‫ֱא‬ ‫ֶא ֶרץ ַה ִמיש ֹר‬ ‫אכת ֲעב ָֹדה‬ ֶ ‫ְמ ֶל‬ ‫ִש ְמ ַלת ִא ָשה‬ ‫ַתּּנּור ָע ָשן‬ ‫ַמ ֲע ֵשה ח ֵֹשב‬

ἵνα ἀγχιστεύσωσιν (…) τὴν κληρονομίαν τὴν πατρικὴν αὐτοῦ οἱ ἄνθρωποι (ἄνδρες) οἱ πολεμισταί ἄνδρες κτηνοτρόφοι ὁ βραχίων ὁ δεξιός οἱ θεοὶ οἱ ἀλλότριοι ἡ γῆ ἡ πεδινή ἔργον λατρευτόν στολὴ γυναικεία κλίβανος καπνιζόμενος ἔργον ὑφαντόν

Num 36:8; cf. Gen 50:8; Lev 25:41 Num 31:28–53; Deut 2:14, 16 Gen 46:32, 34 Exod 29:22; Num 18:18 Gen 35:2, 417 Deut 4:43; cf. 32:10 Lev 23:718 Deut 22:5 Gen 15:17 Exod 28:6; 36:10 (39:9)19

Especially noteworthy are those instances where one would have expected a genitive: ‫ַחטֹאת ָה ָא ָדם‬ ‫ֶע ֶצם ָא ָדם‬ ‫ֶד ֶרְך ַה ֶמ ֶלְך‬ ‫ם־הנָ ִקי‬ ַ ‫ַד‬

αἱ ἁμαρτίαι αἱ ἀνθρώπιναι ὀστέον ἀνθρώπινον ὁδὸς βασιλική τὸ αἷμα τὸ ἀναίτιον

Num 5:6 Num 19:16 Num 20:17; 21:22 Deut 19:13

The last example shows a slight shift in meaning: “innocent blood” instead of “blood of the innocent” (if the translators did not read ‫הדם‬, i.e., with article). Adjectives are also employed in the renderings of some geographical designations:

16  The use of the article makes it clear that νότος is an adjective. 17  Cf. also Deut 32:12. 18  Also Lev 23:8, 21, 25, 35, 36; Num 28:18, 25, 26; 29:12, 35. See, however, Lev 25:39 ‫ֲעב ַֹדת‬ ‫ < ָע ֶבד‬δουλεία οἰκέτου. 19  But Exod 26:1 ἐργασία ὑφάντου (with a noun).

235

Renderings for the Hebrew Construct State Relation

‫ֵע ֶמק ַה ִש ִדים‬ ‫יַ ם־סּוף‬ ‫ם־ה ֶמ ַלח‬ ַ ָ‫י‬

ἡ κοιλὰς ἡ ἁλυκή ἡ ἐρυθρὰ θάλασσα ἡ θάλασσα ἡ ἁλυκή

Gen 14:8, 10 passim Num 34:1220

In light of the foregoing examples, it is surprising that ‫יחֹוח‬ ַ ִ‫( ֵר ַיח־נ‬which occurs approximately 40 times) is regularly rendered as ὀσμὴ εὐωδίας, i.e., with a noun.21 A different translation strategy is represented by those instances where the nomen regens instead of the nomen rectum has been rendered as an adjective. Renderings of this kind are excellent for the most part, although they reflect an even greater departure from the otherwise literal strategy than do the categories mentioned above. Sometimes even the word order was changed, but the translators did not hesitate to use such renderings and often they effectively achieved good results. ‫ֲע ִליֹלת ְד ָב ִרים‬ ‫ַח ְל ִמיׁש צּור‬ ‫ֶע ְרוַ ת ָד ָבר‬ ‫א־ה ַמ ְח ָתה‬ ַ ֹ ‫ְמל‬ ‫ְבחֹזֶ ק יָ ד‬ ‫ַכ ֲחצֹת ַה ַליְ ָלה‬ ‫ֶפ ֶרא ָא ָדם‬ ‫ַמ ְמ ֶל ֶכת כ ֲֹהנִ ים‬ ‫ם־ל ָב ִבי‬ ְ ‫ְב ָת‬ ‫יטב ָה ָא ֶרץ‬ ַ ‫ְב ֵמ‬ ‫ְכג ֶֹדל ַח ְס ֶדָך‬

προφασιστικοὶ λόγοι στερεὰ πέτρα ἄσχημον πρᾶγμα22 τὸ πυρεῖον πλῆρες (ἀνθράκων πυρὸς) ἐν χειρὶ κραταιᾷ23 περὶ μέσας νύκτας ἄγροικος ἄνθρωπος βασίλειον ἱεράτευμα ἐν καθαρᾷ καρδίᾳ ἐν τῇ βελτίστῃ γῇ [67] κατὰ τὸ μέγα ἔλεός σου

Deut 22:14, 17 Deut 32:13 Deut 24:1 Lev 16:12 Exod 13:3, 14, 15 Exod 11:4 Gen 16:12 Exod 19:6 Gen 20:5 Gen 47:6, 11 Num 14:19

If the nomen regens is an adjective in Hebrew, it is rendered as a Greek adjective quite naturally. These cases will be included in our discussion of free renderings of the nomen rectum.

4. Construct Relations Denoting Geographical Entities Geographical entities are frequently designated by construct relations that consist of words like ‫ארץ‬, ‫הר‬, ‫ עמק‬+ proper noun. Such cases are rendered by genitives or, quite often, by appositions. E.g., ‫ ֶא ֶרץ ִמ ְצ ַריִם‬is rendered as ἐν τῇ γῇ Αἰγύπτου or ἐν τῇ γῇ Αἰγύπτῳ. If the name of the country is indeclinable, we cannot determine 20  Contrast Gen 14:3 ἡ θάλασσα τῶν ἁλῶν. 21  Contrast Exod 30:23 κιννάμωμον εὐῶδες (‫ן־ב ֶשם‬ ֶ ‫)קנְ ָמ‬ ִ and κάλαμος εὐῶδες (‫)קנֵ ה־ב ֶֹשם‬. ְ 22  Contrast Deut 23:15 ἀσχημοσύνη πράγματος. 23  Contrast Exod 15:16 μεγέθει βραχίονός σου (‫רֹועָך‬ ֲ ְ‫)בגְ ד ֹל ז‬. ִ

236

Ilmari Soisalon-Soininen

which option was chosen since these names lack the article. If γῆ appears in the nominative or accusative, it is combined with the genitive Αἰγύπτου. However, the dative γῇ can be combined with either dative Αἰγύπτῳ or genitive Αἰγύπτου (the dative making up ¾ of the instances). The ratio dative : genitive is as follows, Genesis 10 : 3, Exodus 16 : 10,24 Leviticus 1 : 0, Numbers 3 : 1, Deuteronomy 9 : 0.25 The Hebrew proper nouns following ἔρημος are all indeclinable and are rarely preceded by an article. Some instances allow the conclusion that the dative ἐρήμῳ is combined both with genitive and dative. ‫ארן‬ ָ ‫ְב ִמ ְד ַבר ָפ‬ ‫ְב ִמ ְד ַבר ִסינָ י‬

ἐν τῇ ἐρήμῳ τῇ Φαράν ἐν τῇ ἐρήμῳ τοῦ Φαράν ἐν τῇ ἐρήμῳ τῇ Σινά26 ἐν τῇ ἐρήμῳ τοῦ Σινά (Rahlfs)27

Gen 21:21 Num 12:16 Num 1:19 Num 9:5

When names of mountains occur (with ‫)הר‬, their renderings show a more frequent use of the Greek article, combined with an apposition in every grammatical case: ἐν τῷ ὄρει τῷ Γαλαάδ τὸ ὄρος τὸ Σινά

Gen 31:23 Exod 19:18

τὸ ὄρος τὸ Σηίρ

Deut 2:5

Of the remaining geographical designations, the only other instance that allows us to deduce the grammatical case from the article is Gen 14:13 τῇ δρυὶ τῇ Μαμβρή.28 Thus, we may safely conclude that the translators did not feel obliged to use genitives. The use of appositions instead of genitives is quite common in other cases too: ‫ת־מ ְכ ֵסה עֹור ֹת ָה ֵא ִילם‬ ִ ‫וְ ֶא‬ ‫ִא ֵשה ֵר ַיח נִ יח ַֹח ַליהוה‬ ‫ָק ְר ַבן ִמנְ ָחה ַליהוה‬

καὶ τὰς διφθέρας δέρματα κριῶν θυσία (κάρπωμα) ὀσμὴ εὐωδίας τῷ κυρίῳ δῶρον θυσίαν τῷ κυρίῳ

Exod 39:20 (34) Lev 1:13, 17; 2:2, 9; 3:5 Lev 2:1

24  Beginning with Exod 11, all instances are with dative. 25  In one case γῇ has been added; in 15:15 I read Αἰγύπτῳ with LXXGö contra Rahlfs. 26  A significant number of mss. read τοῦ. 27 LXXGö without τοῦ. 28  Cf. Deut 11:24 ἀπὸ τοῦ ποταμοῦ τοῦ μεγάλου, ποταμοῦ Εὐφράτου (genitive).

237

Renderings for the Hebrew Construct State Relation

5. Construct Relations with Participle and Object [68] In Hebrew, a participle of a transitive verb may be followed by an object. The participle may be in the absolute state, governing an “accusative” as its object. But the participle and its object may also occur as a construct relation.29 Such construct relations are generally rendered two ways. Greek either has a participle with an accusative, or the participle is rendered as a noun with a genitive. The former predominates in the Pentateuch (viz. #1–4) but the latter also occurs (viz. #5–630). (1) (2) (3) (4) (5) (6)

‫י ְֹצ ֵאי ַש ַער‬ ‫י ְֹש ֵבי ָה ָא ֶרץ‬ ‫י־מ ֲע ֵלה גֵ ָרה הּוא‬ ַ ‫ִכ‬ ‫נ ְֹש ֵאי ֲארֹון ְב ִרית־‬ ‫יהוה‬ ‫יֹושב ָה ָא ֶרץ ַה ְכנַ ֲענִ י‬ ֵ ‫ג ֵֹאל ַה ָדם‬

οἱ ἐκπορευόμενοι τὴν πύλην τοῦ κατοικοῦντος τὴν γῆν ὅτι ἀνάγει μηρυκισμὸν τοῦτο τοῖς αἴρουσιν τὴν κιβωτὸν τῆς διαθήκης κυρίου οἱ κάτοικοι τῆς γῆς Χανάαν ὁ ἀγχιστεύων τοῦ αἵματος

Gen 34:24 Gen 36:20 Lev 11:5 Deut 31:25 Gen 50:11 Deut 19:6ff.

6. A Palette of Renderings for the Construct Relation The variety of renderings of the Hebrew construct relation is not yet exhausted. There are more options than those we have seen up to now. Some of them are reserved to specific semantic collocations. In many cases, a dative instead of a genitive is used in the rendering of the construct relation. Thus, ὄνομα is often combined with the dative, as is ὅμοιος (< ‫ למין‬+ suffix); cf. examples #1–3. The dative is also used in many other expressions (#4–6). Sometimes the free rendering of the whole phrase requires a dative (#7–8). Conversely, the use of a dative often leads to a modified clause construction (#9–10). (1) ‫וְ ֵשם ִא ְשּתֹו‬ (2) ‫ֵשם ָה ֶא ָחד… וְ ֵשם ָה ֶא ָחד‬ (3) ‫ְל ִמינָ ּה‬ (4) ‫רּומת יהוה‬ ַ ‫יא ָה ֵאת ְת‬ ֶ ‫ִיְב‬ (5) ‫מֹוע ֵדי יהוה‬ ֲ ‫ֵא ֶלה‬

ὄνομα δὲ τῇ γυναικὶ αὐτοῦ ὄνομα τῷ ἑνὶ … καὶ τὸ ὄνομα τοῦ δευτέρου … καὶ τὰ ὅμοια αὐτῷ οἴσουσιν τὰς ἀπαρχὰς κυρίῳ αὗται αἱ ἑορταὶ τῷ κυρίῳ

Gen 36:39 Exod 18:3–4 Lev 11:14 Exod 35:5 Lev 23:4, 37

29  Suffixes can be interpreted both ways. 30  Note of the translator: Examples #4–6 I added for the sake of transparency. The noun + genitive rendering is particularly frequent with participles that denote professional occupations and function practically as nouns, such as ‫“ שופט‬judge” and ‫“ רועה‬shepherd.”

238

Ilmari Soisalon-Soininen

(6) (7) (8) (9) (10)

‫ֹלהיָך‬ ֶ ‫תֹוע ַבת יהוה ֱא‬ ֲ ‫ִכי‬ ‫הּוא‬ ‫ְב ֻע ַלת ָב ַעל‬ ‫ִל ְרצֹנְ ֶכם ִתזְ ָבחּו‬ ‫ת־ה ִאיׁש‬ ָ ‫ָה ֵשב ֵא ֶש‬ ‫ית ֵאת ֲח ַצר ַה ִמ ְש ָכן‬ ָ ‫וְ ָע ִש‬

ὅτι βδέλυγμα κυρίῳ τῷ θεῷ σού ἐστιν συνῳκηκυῖα ἀνδρί εἰς δεκτὸν ὑμῖν θύσετε αὐτό ἀπόδος τὴν γυναῖκα τῷ ἀνθρώπῳ καὶ ποιήσεις αὐλὴν τῇ σκηνῇ

Deut 17:131 Gen 20:3 Lev 22:29 Gen 20:7 Exod 27:9

If the regens is an adjective, the rectum is often rendered as a dative, as in πᾶς σοφὸς τῇ καρδίᾳ (Exod 35:10 [9]). Prepositions are very frequently added in the rendering of the construct relation. Above, I noted instances where the material from which something is made is referred to by the preposition ἐκ. A prepositional phrase may be added to the noun, as in λυτρωταὶ διὰ παντὸς (< ‫ ;עולם‬Lev 25:32) or it can be connected to the noun with an article, thus forming an adjectival attribute, as in οἱ ἄρτοι οἱ διὰ παντός (< ‫ ;תמיד‬Num 4:7). Note the following selection: [69] ἀπό διά εἰς

θυσία ἀπὸ ἐσχάρας ὅρια ἀπὸ βορρᾶ cf. Lev 25:32; Num 4:732 τὸ ἐκπορευόμενον διὰ στομάτος θεοῦ νόμιμον εἰς τὸν αἰῶνα ξύλα εἰς ὁλοκαρπώσιν οἱ εἰσπορευόμενοι εἰς τῆν πόλιν ὁδὸν εἰς τὸν ἔρημον

ἐκ

λίθοι εἰς τὴν γλυφήν ὁ κριὸς εἰς ὁλοκαύτωμα τὸ ἱκανὸν εἰς τὸ πρόβατον καρπώματα εἰς ὀσμὴν εὐωδίας μόσχον ἐκ βοῶν καὶ ἀμνὸν ἐκ προβάτων καὶ χίμαρον ἐξ αἰγῶν34 ἀδελφή μού ἐστιν ἐκ (‫ )בת‬πατρός ὁ ἐκ (‫ )בן‬τῆς Ἰσραηλίτιδος ῥήματα ἐκ στόματός μου

Lev 2:7 Num 34:9 Deut 8:3, cf. 34:24 Lev 3:17; cf. Ex 40:15; Deut 15:17 Gen 22:333 Gen 23:10, 18 Exod 13:18; cf. Num 21:33; Deut 3:1 Exod 25:7; 35:9 (8) Lev 8:18 Lev 5:7; cf. 12:9 Num 29:13 Deut 14:4 Gen 20:1235 Lev 24:10 Deut 32:1 (cf. Num 30:13)

31  Cf. also Deut 18:12; 22:5; 23:19; 25:16. 32  Almost all cases in Leviticus and Numbers concern διὰ παντός. 33 Contrast τῆς ὁλοκαρπώσεως in 22:6 (!). 34  This is fairly standard even where Hebrew ‫ בן‬is omitted. Genesis is somewhat inconsistent, as witness ἔριφος αἰγῶν (37:31, 38) as opposed to ἔριφος ἐξ αἰγῶν (38:20). Νεοσσοὶ περιστερῶν (< ‫ )בני יונה‬is always without ἐκ. 35  Likewise Lev 18:9; 20:17; Deut 13:7; 27:22.

Renderings for the Hebrew Construct State Relation

ἐν

ἐπί

ἕως

τὰ ἀνατέλλοντα (‫ )צמח‬ἐκ τῆς γῆς ὕδωρ ἐκ φρέατός σου οἱ ἄνθρωποι οἱ ἐκ τοῦ τόπου τὸ ξύλον τὸ ἐν τῷ παραδείσῳ οἱ κατοικοῦντες ἐν ταῖς πόλεσιν τὸ ὄρος τὸ ἐν τῷ πέραν ἡ γυνῆ ἐν κόλπῳ σου οἱ ἐγκαθημένοι ἐπὶ τῆς γῆς τὸ στέαρ τὸ ἐπ’ αὐτῶν ὁδὸς ἐπ’ ὄρος Σηίρ παιδία ἕως τρίτης γενεᾶς

239

Gen 19:25 Num 21:22 (cf. 20:17) Gen 38:22 Gen 2:16 Gen 19:25 Num 27:12 (& often) Deut 13:7; 28:54, 56 Exod 34:12 Lev 8:25 Deut 1:2 Gen 50:23

The prepositions κατά, παρά, περί and ὑπό are used in a similar fashion. Ὑπό is used 4 times to designate the agent, e.g., εὐλογητὸς ὑπὸ κυρίου (< ‫ברוך יהוה‬, Gen 26:29). In many cases it is quite natural to render the two words of a construct relation as one word. Such renderings are not infrequent. Sometimes one word is omitted as superfluous (#1–7), sometimes a compound word is employed (#8–12). In some cases, ‫ בן‬has simply been rendered with an article, as in Ἐφρὼν τοῦ Σάαρ (< ‫ע ְפר ֹן ֶבן־צ ַֹחר‬, ֶ Gen 25:9). (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

‫ף־רגְ ָלּה‬ ַ ‫ְל ַכ‬ ‫ית־א ִביְך‬ ָ ‫ֵב‬ ‫ֲא ֶשר ְב ֵבית ַהּבֹור‬ ‫ִב ְק ֵצה ַה ִמ ְד ָבר‬ ‫וְ ֵאת גַ בֹת ֵעינָ יו‬ ‫ָע ֵרי ִמ ְק ָלט‬ ‫ד־לב ַה ָש ַמיִם‬ ֵ ‫ַע‬ [70]‫ַשר ַה ַט ָב ִחים‬ ‫ת־כ ֶסף‬ ֶ ַ‫ּומ ְקנ‬ ִ ‫ח ֵֹטב ֵעץ‬ ‫וַ ֲאבֹות ֲאב ֶֹתיָך‬ ‫מֹול ֶדת ָא ִביָך‬ ֶ

τοῖς ποσὶν αὐτῆς παρὰ τῷ πατρί σου ἐν τῷ λάκκῳ παρὰ τὴν ἔρημον καὶ τὰς ὀφρύας φυγαδευτήρια ἕως τοῦ οὐρανοῦ ἀρχιμαγείρος36 ἀργυρώνητος ξυλοκόπος οἱ πρόπαπποι αὐτῶν ὁμοπατρία

Gen 8:9 Gen 24:23 Exod 12:29 Exod 13:20 Lev 14:9 Num 35:11 Deut 4:11 Gen 37:36 Gen 17:12 Deut 29:10 (11) Exod 10:6 Lev 18:11

Compounds are also used if the nomen regens is an adjective.

36  The Hebrew ‫ ַשר‬is often rendered with compounds containing ἀρχι- as well as in cases like ‫ > ָש ֵרי ֲח ִמ ִשים‬πεντηκονταρχαί.

240

Ilmari Soisalon-Soininen

‫ּוכ ַבד ָלׁשֹון ָאנ ִֹכי‬ ְ ‫ד־פה‬ ֶ ‫ְכ ַב‬ ‫ֹלהים‬ ִ ‫יִ ְר ֵאי ֱא‬ ‫ם־ק ֵשה־ע ֶֹרף‬ ְ ‫ַע‬

ἰσχνόφωνος καὶ βραδύγλωσσος ἐγώ εἰμι θεοσεβεῖς σκληροτράχηλος

Exod 4:10 Exod 18:21 Exod 33:3, 537

There is a remaining category of cases, not at all rare, where one part of the construct relation is rendered as a verb. The Greek of such renderings is often very skilful. ‫מֹול ְדּתֹו‬ ַ ‫ְב ֶא ֶרץ‬ ‫יהם‬ ֶ ‫ֵאת ֶא ֶרץ ְמגֻ ֵר‬ ‫ּובכֹל ִמ ְש ַלח יָ ֶדָך‬ ְ

ἐν τῇ γῇ, ᾗ ἐγενήθη τὴν γῆν, ἣν παρῳκήκασιν ἐν πᾶσιν, οὗ ἂν ἐπιβάλῃς τὴν χεῖρά σου

Gen 11:28 Exod 6:4 Deut 15:10

7. Concluding Remarks Hebrew construct relations have been rendered in diverse ways. Considering the fact that the genitive is usually the most obvious rendering, the proportion of other renderings is relatively high. One could say that the translators did not feel tied to the obvious standard rendering of the construct relation, the genitive. Rather, they understood the construct relation as a unit which they then rendered as a whole. There are numerous indications that the Septuagint translators had small segments of text before them. Many Hebraisms or otherwise awkward renderings, which abound in the Pentateuch,38 came into being, in all likelihood, because the words that belong together were far removed from one another.39 This is a question that merits further research. But the ways the translators rendered the closely linked words of the construct relation show that they regarded it as a single unit of translation. This might explain the great variety of their renderings. It is, however, important to note that certain specific, frequently occurring construct relation phrases were rendered consistently. For the translators, the construct relation was not a category of the kind that would have called for a uniform treatment.

37  Also Exod 34:9; Deut 9:6, 13. 38  E.g., the pleonastic pronoun in relative clauses. 39 Note of the translator: Soisalon-Soininen refers here to the effects of segmentation, through which the translators lost touch with the segments they had already translated. This resulted in later segments referring back incorrectly or not at all to elements from preceding segments, an issue that Soisalon-Soininen treated six years later in “Segmentation in the Greek Pentateuch” (in this volume).

Ilmari Soisalon-Soininen The Alleged Interchangeability of ἐν and εἰς in the Septuagint1 1982 Abstract The so-called interchangeability of ἐν and εἰς is often not properly distinguished from the issue of whether the collocations with these prepositions express movement or rest. For example, an unexpected collocation with ἐν may not be due to its interchangeability with εἰς but rather to the author’s focus on the state of rest which is the end result of the action. A survey of the Pentateuch material shows that this, indeed, is the rationale of the use of ἐν, and that the semantic distinction between ἐν and εἰς was still recognized. This is also the background of the papyri. The boundaries between ἐν and εἰς begin to blur in the NT. We should therefore use Johannessohn’s work with caution. The acquired insight can serve as a useful guideline in textual criticism.

1. Confusing Treatments in Greek Grammars [131] The Greek preposition εἰς is merely a variant of ἐν, formed with the ending -ς (ενς > εις). Originally, ἐν expressed both rest and movement. In the former meaning it was expressed with the dative, in the latter with the accusative. Ἐν + acc. occurs in many dialects. After the first centuries CE, the use of εἰς + acc. expanded until ἐν completely disappeared in modern Greek, and εἰς + acc. had absorbed the meaning of rest. Since in Koine Greek ἐν + dat. is frequently used where one would expect εἰς + acc., many grammars claim that ἐν and εἰς are interchangeable. Nigel Turner, for example, puts it like this:

1  This is a translation of “Ἐν für εἰς in der Septuaginta”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987, AASF, ser. B, 237 (Helsinki: Suomalainen Tiedeakatemia, 1987) 131–40, by Theo van der Louw.

242

Ilmari Soisalon-Soininen

In the Koine ἐν and εἰς are freely interchanged, until in MGr εἰς has absorbed ἐν completely, consistently with the disappearance of the dat.2

But matters are not that simple. If ἐν and εἰς were freely interchangeable, both should answer the question of “where” as well as the question of “where to.” In other words, both should then express rest as well as movement. But if in expressions with certain verbs the focus is sometimes on movement and sometimes on rest, that is a different thing. [132] The former suggests that the prepositions have merged semantically (hence the ubiquitous use of εἰς + acc. in modern Greek), the latter points to a different psycholinguistic nuance. For example, in English we say either “to arrive at a place” or “to arrive somewhere,” but in Finnish the second possibility is not possible. Or we say “to hide somewhere,” although hiding normally entails movement (it is obviously also possible to hide something by covering it, i.e. without movement). In this case Finnish would also say “to hide to some place.” If such use is limited to expressions of a certain kind, we must account for it by different psycholinguistic reasoning: the end result (“where”) is in focus instead of the movement (“where to”) or vice versa. Some expressions allow both possibilities for certain verbs. The confusion of interchangebility and mental focus is visible in Moulton’s Grammar of New Testament Greek, where it says (255): In the LXX (e.g. Ge 31:33 Nu 35:33), Diodorus (e.g. 3, 44) and in Hermas (e.g. V, I 2, 2; II 4, 3; S I 2), εἰς = ἐν is common. Note especially κρύπτω εἰς Jb 40:8 (13), Pr 1:11 Isa 2:10 Je 4:29 Ps 88:40.

But as already pointed out above, κρύπτω normally entails movement. In Greek, this verb can answer both of the questions “where” and “where to.” The other examples from the LXX can be explained in a similar fashion. And since Diodorus and Hermas represent later texts, we must conclude that Moulton’s examples are quite wide of the mark. There are, however, grammars that observe that the focus sometimes shifts between rest and movement in Greek. Schwyzer writes: Already in older Greek the focus varies between rest (ἐν + dat.) and movement (εἰς + acc.), e.g, with τίθημι etc.3 Kühner and Gerth express it very eloquently:

2  J.H. Moulton/N. Turner, A Grammar of New Testament Greek: Vol. III, Syntax (Edinburgh: T&T Clark, 1963), 254. 3  E. Schwyzer/A. Debrunner, Griechische Grammatik auf der Grundlage von Karl Brugmanns griechischer Grammatik: Teil 2, Syntax (München: C.H. Beck, 1950), 461.

The Alleged Interchangeability of ἐν and εἰς in the Septuagint

243

It is a particularity of the Greek, especially Homeric, language that it often joins verbs that express the direction where to to prepositions governing the dative (rarely the genitive, see note 3) and, vice versa, [133] collocates verbs that do not contain the idea of movement with prepositions governing the accusative. These constructions arose from summarizing two moments of an action or from the merging of two ideas, whereby next to the moment of movement also the moment of rest that logically follows it or, conversely, next to the moment of rest simultaneously the moment of preceding or resulting movement was conceived and expressed. We call this type of construction pregnant. The semantic force and picturesque shortness of this construction go without saying. It serves to present our minds with two images when we view either the movement and the subsequent state of rest together or the state of rest (πίπτειν ἐν κοινίησιν) and the preceding or following movement (λὶς ἐφάνη εἰς ὁδόν) together.4

That is a bit overstated. It is clear that the viewpoint of rest or movement could shift, but that it served a stylistic purpose is less probable. Mayser remarks with respect to earlier Koine that the different viewpoint of rest or movement led to a change between ἐν and εἰς in a single sentence.5 His examples are few and limited to certain expressions, especially γράφειν, ἀνα-, ἐγγράφειν ἐν or εἰς. Sometimes ἐν is used with a perfect tense, e.g. προσεπέπτωκεν ἡμῖν ἀπῆχθαι αὐτὸν ἐν τῇ ἐν Κροκοδίλων πόλει φυλακῇ. This example is very understandable because of the meaning of the Greek perfect. Likewise, the examples of εἰς for ἐν can be understood readily with the help of the notion of movement, e.g. προσορμῆσαι εἰς Μέμφιν, παρεῖναι εἰς (373). As is obvious, ἐν and εἰς are not freely interchanged, rather the question of either rest or movement comes in, as Mayser acknowledges. But in a footnote he points to the merging of εἰς and ἐν in general and refers to Hatzidakis’ Einleitung in die neugriechische Grammatik (1892) and similar works that treat this question. The majority of their material concerns late texts [134] where εἰς is already clearly used for ἐν. In this context, Mayser mentions Johannessohn’s Gebrauch der Präpositionen in der Septuaginta, which shows that in reality he did not clearly distinguish the two issues of ἐν or εἰς on the one hand and the focus on rest or movement on the other hand. The phenomenon he is discussing lies outside the scope of the literature he quotes. With respect to the New Testament, Blass—Debrunner—Rehkopf treat this question very clearly.6 The merging of ἐν and εἰς begins to make itself felt where 4  R. Kühner/B. Gerth, Ausführliche Grammatik der griechischen Sprache: Teil 2, Satzlehre (3rd ed.; Hannover und Leipzig: Hahn, 1898), §447. 5  E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften: 2,2, Satzlehre, analytische Teil (Berlin: de Gruyter, 1934), 371. 6  F. Blass/A. Debrunner/F. Rehkopf, Grammatik des neutestamentlichen Griechisch (Göttingen: Vandenhoeck & Ruprecht, 1976), 167, 177–8.

244

Ilmari Soisalon-Soininen

εἰς stands for ἐν in a local sense, albeit only in Mark, Luke-Acts, rarely in John. The Letters and even Revelation show the correct distinction of ἐν and εἰς in a local sense. Most of the examples are convincing, e.g., John 1:18 ὁ ὢν εἰς τὸν κόλπον τοῦ πατρός. The interpretation of ἐν for εἰς as a hypercorrect ἐν is probably unsubstantiated. Most of the examples of ἐν instead of εἰς can be explained as a different focus on either rest or movement. It is mentioned that classical writers could also use ἐν with τιθέναι and ἱστάναι, and διδόναι ἐν τῇ χειρί τινος is compared with it. Some examples could be explained otherwise, e.g., Luke 4:1 ἤγετο … ἐν τῇ ἐρήμῳ “he was led around in the desert.” But it is clear that in the NT, partly, εἰς seems to be used for ἐν, even in those cases where rest is in focus. It is not clear whether ἐν is used in the sense of εἰς or if the focus is on rest instead of movement in these cases. This question will be discussed below. That Blass— Debrunner—Rehkopf refer to Johannessohn and Mayser indiscriminately suggests that in the NT on the one hand, and in the LXX and the Ptolemaic papyri on the other we are dealing with basically the same phenomenon. As said above, Mayser expresses himself clearly on the subject of movement and rest, although in a footnote he refers to later usage without mentioning that crucial distinction. Johannessohn expresses himself very vaguely, “ἐν is often used for εἰς with verbs of movement.”7 But about εἰς for ἐν he says, “Conversely, εἰς sometimes answers the question ‘where?’” The word “conversely” suggests that he explains the cases of ἐν with verbs of movement in the same vein. I will return to Johannessohn’s examples later. [135] All in all, it looks as if none of the authors explicitly distinguished between the questions of whether εἰς is simply used for ἐν or whether the focus is on movement or rest.

2. Ἐν for εἰς in the Septuagint? In Hebrew, the preposition ‫ ב‬in a local sense can answer the questions “where” and “where to.” Therefore, it can be rendered as ἐν or εἰς according to the context. But the Hebrew prepositions ‫ אל‬and ‫ ל‬are also used very frequently in the sense of “into,” and are then rendered as εἰς, naturally enough. Generally speaking, the meaning “where” is the most common meaning of ‫ב‬. Thus, one could expect mechanical renderings ‫ → ב‬ἐν, especially when the verb is farther removed. It shall be our task to investigate how far the use of ἐν goes in those cases where we would expect εἰς. We shall pay special attention to the question of whether or not this usage is limited to expressions of a certain kind, and what the boundaries of this usage are. We shall also investigate if εἰς can have the meaning “where.”

7  M. Johannessohn, Der Gebrauch der Präpositionen in der Septuaginta (MSU 3,3; Berlin: Weidmann, 1926), 330–32.

The Alleged Interchangeability of ἐν and εἰς in the Septuagint

245

For the LXX, I surveyed all the data for the Pentateuch, the oldest part of the LXX and representative of the earliest phase. For the rest, I had to limit myself to the examples of Johannessohn and other authors. In the Pentateuch, ἐν is quite often used in cases where we would expect εἰς. Not infrequently we find exactly the same expression with εἰς in some cases, with ἐν in others. Most cases belong to the category where classical usage could already vary between a focus on rest or movement. This is predominant for τιθέναι: ‫ֹלהים ִּב ְר ִק ַיע‬ ִ ‫וַ ּיִ ֵּתן א ָֹתם ֱא‬ ‫ַה ָּׁש ָמיִם‬

καὶ ἔθετο αὐτοὺς ὁ θεὸς ἐν τῷ στερεώματι τοῦ οὐρανοῦ καὶ ἔθετο αὐτὸν ἐν τῷ παραδείσῳ καὶ ἔθετο αὐτοὺς ἐν φυλακῇ καὶ ἔθηκεν τὰ βρώματα ἐν ταῖς πόλεσιν οὐδὲ λίθον σκοπὸν θήσετε ἐν τῇ γῇ ὑμῶν

Gen 1:17 Gen 2:158 Gen 40:3 Gen 41:48 Lev 26:19

That these cases cannot have been prompted by Hebrew [136] is confirmed by the fact that in Gen 42:17 ἐν also appears as a rendering of ‫( אל‬ἔθετο αὐτοὺς ἐν φυλακῇ). Ἐν is also used with verbs other than τιθέναι that have a similar meaning: ἀποτιθέναι διδόναι

λαμβάνειν ἱστάναι

καὶ ἀπέθηκεν Μωυσῆς τὰς ράβδους ἔναντι κυρίου ἐν τῇ σκηνῇ τοῦ μαρτυρίου ἔδωκεν ὑμῖν θησαυροὺς ἐν τοῖς μαρσίπποις ὑμῶν ᾧ ἔδωκεν ὁ θεὸς ἐπιστήμην ἐν τῇ καρδίᾳ ἔδωκέν τις τὴν κοίτην αὐτοῦ ἐν σοί λάβετε ἀπὸ τῶν καρπῶν τῆς γῆς ἐν τοῖς ἀγγείοις ὑμῶν στήσετε τοὺς λίθους τούτους, οὓς ἐγὼ ἐντέλλομαί σοι σήμερον, ἐν ὄρει Γαιβάλ Other passages that belong here: Διὰ τὸ ἀργύριον τὸ ἀποστραφὲν ἐν τοῖς μαρσίπποις ἡμῶν κατεπόντισεν ἐν ἐρυθρᾷ θαλάσσῃ ὧν ἐὰν σπείρῃς ἐν τῷ ἀγρῷ σου ὅτι ἐν σκηναῖς κατῴκισα τοὺς υἱοὺς Ισραηλ βάψει ἐν ἐλαίῳ τὸν πόδα αὐτοῦ

Num 17:22 Gen 43:23 Exod 36:2 Num 5:2010 Gen 43:11 Deut 27:4

Gen 43:18 Exod 15:4 Exod 23:16 Lev 23:43 Deut 33:24

8  If possible, I omit the Hebrew text. If a preposition other than ‫ ב‬is being rendered, this is duly indicated. 9  Similar cases are Gen 30:38, 41; 41:10; Exod 30:36; 40:26 (24); Deut 14:28. 10  Other cases with διδόναι are Lev 14:34; Deut 18:18.

246

Ilmari Soisalon-Soininen

ἔθαψεν Ἀβρὰμ Σάρραν τὴν γυναῖκα αὐτοῦ ἐν (‫ )אל‬τῷ σπηλαίῳ θάψατέ με μετὰ τῶν πατέρων μου ἐν (‫ )אל‬τῷ σπηλαίῳ

Gen 23:19 Gen 49:29

Γράφειν ἐν / εἰς belongs to the examples adduced by Kühner-Gerth and Mayser. The examples below are somewhat similar to that usage. Κατάγραψον τοῦτο … ἐν βιβλίῳ καὶ ἐντομίδας ἐπὶ ψυχῇ οὐ ποιήσετε ἐν τῷ σώματι ὑμῶν ὅσα ἐγὼ λαλῶ ἐν τοῖς ὠσὶν ὑμῶν

Exod 17:14 Lev 19:28 Deut 5:1

A special group is the numerous renderings with ‫ → ביד‬ἐν χειρί (ἐν ταῖς χερσίν) or εἰς τὰς χεῖρας respectively. Ἐν is also used with verbs of movement: πάντα τὰ τέρατα, ἃ ἔδωκα ἐν ταῖς χερσίν σου καὶ τὴν ῥάβδον … λαβὲ ἐν τῇ χειρί σου καὶ ἐλάβον ἐν ταῖς χερσὶν αὐτῶν ἀπὸ τοῦ καρποῦ τῆς γῆς

Exod 4:21 Exod 17:5 Deut 1:25

The last two examples could be explained as ἐν instrumenti, although this was not intended. As expected, εἰς also occurs, e.g. Exod 5:21, δοῦναι ῥομφαίαν εἰς τὰς χεῖρας αὐτοῦ.11 When χείρ is not used in its concrete sense but in the sense of “in someone’s power,” the LXX Pentateuch uses εἰς all the time: καὶ παραδώσω εἰς τὰς χεῖρας ὑμῶν τοὺς ἐγκαθημένους ἐν τῇ γῇ καὶ παραδώσει τοὺς βασιλεῖς αὐτῶν εἰς τὰς χεῖρας ὑμῶν

Exod 23:31 Deut 7:2412

In the other books of the LXX, especially Judges, [137] ἐν τῇ χειρί also occurs with παραδιδόναι.13 In Judges, ἐν (τῇ) χειρί occurs 22 times with (παρα-/ἀπο-) διδόναι and another 5 times with τιθέναι and λαμβάνειν. But εἰς τὰς χεῖρας occurs only 3 times in both the A and B texts. Hebrew ‫ ב‬clearly led to this outcome. Ἐν μέσῳ is used both for “where” and “where to” in the NT, whereas εἰς μέσον + genitive does not occur (perhaps Mat 10:16 B).14 In the LXX, εἰς μέσον + gen. is relatively common. As could be expected, τιθέναι also occurs with ἐν μέσῳ, e.g. 2 Chr 6:13; Ezek 5:5; 37:1, 26.15 It remains uncertain whether we 11  Cf. Gen 27:27; 40:13; Deut 24:1, 3. 12  Further cases occur in Lev 26:25, Num 21:34, Deut 1:27; 2:24, 30; 3:2, 3; 7:24; 19:12; 20:13; 21:10. Note, incidentally, the use of παραδιδόναι for ‫נתן‬. 13  R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (AASF 19; Helsinki: Suomalainen Tiedeakatemia, 1979), 161. 14 Blass/Debrunner/Rehkopf, Grammatik des ntl. Griechisch, 161. 15 Sollamo, Renderings of Hebrew Semiprepositions, 254.

The Alleged Interchangeability of ἐν and εἰς in the Septuagint

247

should understand Zech 5:8, καὶ ἔρριψεν αὐτὴν ἐν μέσῳ τοῦ μέτρου, in the same vein or whether the translator imagined that the woman, personifying evil, was thrown somewhere within the ephah. Second Esdras 6:10 συναχθῶμεν εἰς οἶκον τοῦ θεοῦ ἐν μέσῳ αὐτοῦ would have to be explained thus: first the focus was on entering the temple, and then on gathering in the sanctuary. A clearer case of ἐν μέσῳ instead of εἰς μέσον can hardly be found, with the exception of Judg 7:17 (see below). That Blass—Debrunner—Rehkopf refer to p. 325f. of Johannessohn is probably because he only treats the renderings of ‫ בתוך‬and ‫ בקרב‬as ἐν (μέσῳ), but does not mention cases with εἰς μέσον. In his index, εἰς μέσον, frequent enough according to Hatch & Redpath and Sollamo, is absent. It is logical that εἰς μέσον is much rarer than ἐν μέσῳ in the LXX because contexts with ‫ בתוך‬and ‫ בקרב‬focus more on rest than on movement. With ‫ ביד‬the opposite is the case. Significantly, almost all cases of ἐν where we would expect εἰς correspond closely to τιθέναι ἐν or γράφειν ἐν. It is no easy matter to draw a boundary between verbs of movement that use ἐν and those that do not. Negatively speaking, at least in the Pentateuch ἐν is not used with intransitive verbs of movement in the sense of ‘into.’ But of course ἐν is used with intransitive verbs of movement in the sense of ‘moving within certain boundaries.’ That is how Lev 19:33, [138] Ἐὰν δέ τις προσέλθῃ προσήλυτος ὑμῖν ἐν τῇ γῇ ὑμῶν, should be understood. It is equally natural to say, as in Exod 15:10, ἔδυσαν ὡσεὶ μόλιβος ἐν ὕδατι σφοδρῷ. It is also natural that with intransitive περιβάλλειν both ἐν and εἰς are used. In fact, they are used promiscuously in the list of campsites in Num 33. In Exod 7:28 there is a case where the translator apparently forgot how he began his sentence. He started with εἰς but continued with ἐν: καὶ ἐξερεύξεται ὁ ποταμὸς βατράχους, καὶ ἀναβάντες εἰσελεύσονται εἰς τοὺς οἴκους σου καὶ εἰς τὰ ταμιεῖα τῶν κοιτώνων σου καὶ ἐπὶ τῶν κλινῶν σου καὶ εἰς τοὺς οἴκους τῶν θεραπόντων σου καὶ τοῦ λαοῦ σου καὶ ἐν τοῖς φυράμασίν σου καὶ ἐν τοῖς κλιβάνοις σου.

The rest of the LXX cannot be too different from the Pentateuch. I must limit myself to a survey of the cases with ἐν χειρί—εἰς χεῖρας and ἐν μέσῳ—εἰς μέσον. An intransitive verb is very rare with ‫ביד‬, and logically so. An expression like ‫נפל‬ ‫( → ביד‬ἐμ)πίπτειν εἰς χεῖρας occurs only rarely, and I found no instances of ἐν in these cases. Ἐν μέσῳ does occur with intransitive verbs of movement, but not in the sense of ‘into the midst of.’ In Judg 18:20, ‫וַ ּיָבֹא ְּב ֶק ֶרב ָה ָעם‬

καὶ εἰσῆλθεν ἐν μέσῳ τοῦ λαοῦ

the Hebrew text presupposes the sense of “he went into the midst of the people,” but the LXX rendering should be understood as “he went in the midst of, i.e.

248

Ilmari Soisalon-Soininen

together with, the people.” This meaning is also present, and clearly suggested in Hebrew, in the two following examples. ‫ת־אֹרון‬ ֲ ‫נִ ְק ָחה ֵא ֵלינּו ִמ ִּׁשֹלה ֶא‬ ‫ְּב ִרית יְ הוָ ה וְ יָבֹא ְב ִק ְר ֵּבנּו‬ ‫ן־א ָּׁשה יִ ְׂש ְר ֵא ִלית וְ הּוא‬ ִ ‫וַ ּיֵ ֵצא ֶּב‬ ‫ן־איׁש ִמ ְצ ִרי ְּבֹתוְך ְּבנֵ י יִ ְׂש ָר ֵאל‬ ִ ‫ֶּב‬

λάβωμεν τὴν κιβωτὸν τοῦ θεοῦ ἡμῶν ἐκ Σηλώμ, καὶ ἐξελθέτω ἐν μέσῳ ἡμῶν Καὶ ἐξῆλθεν υἱὸς γυναικὸς Ισραηλίτιδος καὶ οὗτος ἦν υἱὸς Αἰγυπτίου ἐν τοῖς υἱοῖς Ἰσραήλ

1 Kgdms 4:3

Lev 24:10

I did an additional survey of Judges, because there one could expect mechanical renderings ‫ → ב‬ἐν. Judges is an extremely literal translation and deviates from the rest of the LXX in this respect. Effectively, ‫ ב‬is rendered as ἐν nearly always, and the most awkward Hebraisms are related to this phenomenon (e.g. πολεμεῖν, παρατάσσεσθαι ἐν). But with intransitive verbs of movement εἰς is mostly used. Otherwise, εἰς is sometimes used for ‫ ב‬with a transitive verb of movement, or for ‫ אל‬or ‫ ל‬or ‫ ה‬locale. However, in Judg 7:11, 17, 19, all instances [139] close together, we find ἐν (μέσῳ) with καταβαίνειν and εἰσπορεύεσθαι. In 7:13 κυλιομένη ἐν τῇ παρεμβολῇ can be understood as ‘rolling in the camp.’ Similarly natural expressions are also ἐπορεύθη Ἰσραήλ ἐν τῇ ἐρήμῳ (11:16) and διῆλθεν ἐν τῇ ἐρήμῳ (11:18). In 7:11, 17, 19 ἐν is used mechanically, and the same holds true for 1 Kgdms 9:25 καὶ κατέβη ἐκ τῆς Βαμὰ ἐν τῇ πόλει, which is Johannessohn’s only example of ἐν with an intransitive verb of movement. So, it is very probable that what I said about the Pentateuch also holds true for the rest of the LXX. In principle, the use of ἐν with intranstive verbs of movement could have been possible in the same way as with transitive verbs. But this seems to have been a boundary for the LXX translators and probably also for earlier Koine in general. That such a boundary exists is proven by the fact that the clear semantic distinction between ἐν and εἰς was still preserved, and that we are dealing with a different focus on either rest or movement.

3. Εἰς for ἐν in the Septuagint? With respect to the use of εἰς for ἐν, Blass—Debrunner—Rehkopf give examples from the NT, but only from Mark, Luke-Acts and John.16 They refer to examples from the LXX, Diodorus etc. in the works by Johannessohn, Mayser and others.17 16 Blass/Debrunner/Rehkopf, Grammatik des ntl. Griechisch, 205. 17  Namely, A.N. Jannaris, An Historical Greek Grammar, Chiefly of the Attic Dialect (London: Macmillan, 1897); J. Humbert, Syntaxe grecque (3rd ed.; Paris: Klincksieck, 1960); A. Oepke, articles “εἰς” and “ἐν” in G. Kittel (ed.), Theological Dictionary to the New Testament:

The Alleged Interchangeability of ἐν and εἰς in the Septuagint

249

The examples Mayser gives from the older texts are not relevant; they merely illustrate a shift in focus from rest to movement: for example, UPZ 81 II, 6 προσορμῆσαι εἰς Μέμφιν; P. Zen 59, 226, 3 ἡμᾶς ἴσθι παρεσομένους εἰς Πτολεμαίδα. One easily gets a false impression from Johannessohn. He claims that εἰς sometimes answers “where” in the LXX, i.e. in the sense of ἐν. The examples he gives are the following: καὶ εὗρεν αὐτοὺς εἰς Δωθάειμ ὁ δὲ μικρότερος μετὰ τοῦ πατρὸς ἡμῶν σήμερον εἰς γῆν Χανάαν

Gen 37:17 Gen 42:32

He mentions only later on that the reading is uncertain. Indeed, both are now rightly recognized as later variants for an original ἐν.18 The same holds true for 1 Macc 3:46. The other examples from Maccabees are not more convincing, καὶ οἱ εἰς τὸ ἀριστερὸν κέρας εἶδον εἰς τὸν τόπον, οὗ τὸ πῦρ ἔκρυψαν …, τὸ ὕδωρ ἐφάνη παρῆσαν οἱ ἄρχοντες Δημητρίου εἰς Κηδες

1 Macc 9:16 2 Macc 1:33 1 Macc 11:63

With respect to the third example, Johannessohn acknowledges that παρεῖναι is often used with εἰς by Polybius.19 A more convincing example is Gen 31:33: εἰσελθὼν δὲ Λαβὰν ἠρεύνησεν εἰς τὸν οἶκον Λείας, καὶ οὐχ εὗρεν· καὶ [140] ἐξῆλθεν ἐκ τοῦ οἴκου Λείας καὶ ἠρεύνησεν εἰς τὸν οἶκον Ἰακὼβ καὶ ἐν τῷ οἴκῳ τῶν δύο παιδισκῶν.

The verb forms εἰσελθών and ἐξῆλθεν influenced the rendering here. We should further note that ἠρεύνησεν has no counterpart in the MT. This means that ‫ ב‬appears here in the sense of ‘into.’ One cannot and should not refer to Johannessohn here. Although I do not have the entire LXX material at my fingertips, I have the impression that examples of εἰς for ἐν are very hard to find, or at best in the latest parts of the Septuagint. In both the LXX and the Ptolemaic papyri, the prepositions in question are used according to the focus on “where” or “where to.” There is no evidence that εἰς could answer “where” or ἐν could answer “where to.”

Volume 2, δ–η, trans. W. Bromley (Grand Rapids: Eerdmans, 1964) 420–33, 537–42; J. Vogeser, Zur Sprache der griechischen Heiligenlegenden (München: Seitz, 1907). 18  The Göttingen edition of Genesis (ed. Wevers) has ἐν in both cases: J.W. Wevers (ed.), Genesis (SVTG 1; Göttingen: Vandenhoeck & Ruprecht, 1974). 19  See also Mayser’s example mentioned above on p. [134].

250

Ilmari Soisalon-Soininen

In the first centuries CE, εἰς started to be used for ἐν, until it had absorbed its function completely in Modern Greek. Starting with some of the NT books, perhaps a bit earlier, there occur examples of ἐν with intransitive verbs of movement. Preisigke mentions several such instances,20 but all from the 3rd to 7th centuries CE. Although it is not in itself impossible that with intransitive verbs of movement the situation of rest is in focus as a result of the movement, it stands more to reason that the expansion of εἰς caused uncertainty in the use of ἐν as well. To the best of our knowledge, several NT books are the oldest testimonies to this process of merging. That its origin could already be found in the LXX and the Ptolemaic papyri cannot be documented from the examples given by Johannessohn and Mayser.

4. Conclusions My aim was to emphasize that from a psycholinguistic point of view there is a crucial difference between synonymy or interchangeability of εἰς and ἐν on the one hand and a varying focus on rest or movement on the other. For the LXX our discussion has the following implications. 1) Great caution is needed in the case of ἐν with intransitive verbs of movement if the sense of “moving within certain boundaries” is impossible. 2) A similar reserve is in order with respect to readings of εἰς for ἐν if they cannot be explained in the sense of “where to.” 3) Johannessohn’s Untersuchungen should be used with due caution.

20  F. Preisigke, Wörterbuch der griechischen Papyrusurkunden mit Einschluss der griechischen Inschriften, Aufschriften, Ostraka, Mumienschildern usw. aus Ägypten: Band 1, A–K (Berlin: Selbstverlag der Erben, 1925).

Ilmari Soisalon-Soininen Segmentation in the Greek Pentateuch1 1983

Abstract The Septuagint translators differ from present-day translators in that they considered their first draft the definitive version and in that their work was stamped by segmentation. The translators rendered their text segment by segment, which made them lose contact with the preceding sentence construction, with awkward renderings as a result. Within segment boundaries, they translated quite freely. The impact of segmentation is illustrated in four syntactic phenomena, (1) repetition, (2) apodotic καί, (3) ‫ ב‬+ inf. constr. and (4) the pleonastic pronoun in relative clauses. Segmentation is an essential factor, albeit not the only one, that helps explain how the translation Greek of the Septuagint came about.

1. Introduction In the Septuagint we encounter very different ways of translating. There is a wide discrepancy between the most literally and the most freely translated books. But within literal translations one occasionally also finds elegant free renderings of individual Hebrew expressions, and, conversely, Hebraistic renderings in free translations. It is especially noteworthy how certain phrases have been rendered quite frequently in a Hebraistic fashion, but sometimes have also been translated freely. We cannot explain this phenomenon by assuming that the translators were intentionally striving after a faithful word-for-word rendering of the text. Rather, we should rather imagine that they found certain expressions

1  This is a translation of “Beobachtungen zur Arbeitsweise der Septuaginta-Übersetzer,” in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 28–39, by Theo van der Louw. The original page numbers are indicated between square brackets.

252

Ilmari Soisalon-Soininen

­ ifficult, and sometimes arrived at free renderings and at other times at literal d renderings. But there is more to it. In the following, I will attempt to outline various characteristics of the approach of the Septuagint translators in which they differ from present-day translators. Additionally, I will show which phenomena can be explained, at least partially, to lie in the background of this approach. [29] A modern translator does not normally regard the first draft as the definitive version. Even if he has a firm grasp of the larger textual units, he knows that the context will occasion changes later in the process. If he then reads a further portion, he often comes upon expressions that will have to be changed in the light of the wider context. He corrects, even repeatedly, and does it without restrictions, since he knows that a clean copy of the text has to be made in the end. He may put alternatives side by side and choose the best option for the definitive text later on. For ancient translators, this flexible approach was not possible. They could not use their writing materials too lavishly. But, more importantly, it had never occurred to them that such an approach would be necessary. If we worked like them, awkward phrases and outright errors would arise to the same extent. In cases where a linguistic phenomenon is caused by the specific approach, it is pointless to ask how the translator intended the rendering. He did not create it on purpose; it just came about. A decisive factor in the approach of the LXX translators is that they did not work with long translation units at a time. They read a longer or shorter section of a sentence, translated it and wrote it down, and then went on to the next section. The length of these segments may have been different for the various translators, but may also have varied considerably for the same translator. Small segments were a factor leading to a more literal translation, albeit not the only factor. For in parts of the Septuagint we already find literal renderings as a result of a conscious attempt. The smaller the segments, the greater the danger that certain expressions would be rendered without regard for the context. And when the translator did not feel the need to correct the outcome, the rendering remained unchanged. One has to keep in mind that the segmentation [30] did not lead to a total loss of contact with the preceding or the following, but the contact was considerably weakened. We may also surmise that the influence of the preceding text was stronger than that of the following. What had already been translated remained in the translator’s memory. But he did not see the need of consulting in advance a passage that would follow. In the following section I will take some examples from the Pentateuch, mainly from Genesis, that will show what the limits of the translator’s attention to the context were. The Pentateuch is quite a good translation, in good Koine Greek, but there are occasional renderings that call for explanation. Certain expressions have attracted renderings that seem to be contrary to the linguistic character of

Segmentation in the Greek Pentateuch

253

these books as a whole.2 I will first give some general examples, and then discuss frequent expressions that exhibit special features.

2. Segmentation: General Examples Sometimes the translator begins with a free rendering but does not continue in the same vein. A good example can be found in Deut 11:21, ‫יכם‬ ֶ ֵ‫ימי ְבנ‬ ֵ ִ‫יכם ו‬ ֶ ‫יְמ‬ ֵ ‫ְל ַמ ַען יִ ְרּבּו‬

ἵνα πολυημερεύσητε καὶ αἱ ἡμέραι τῶν υἱῶν ὑμῶν

The translator has used the verb πολυημερεύειν, incorporating ‫יכם‬ ֶ ‫יְמ‬, ֵ so that the nd 2 person plur. became the subject. But, in the following, αἱ ἡμέραι is employed as subject. Such anacolutha are typical of careless speech, but in this case it shows how even a short sentence has been translated in two segments, and how easily the translator lost touch with the preceding, already translated text. It often happens in a series of paratactic clauses depending on the same verb that the construction is changed suddenly. See, e.g., Exod 7:28, where it says of the frogs, ‫ּוב ֲח ַדר‬ ַ ‫יתָך‬ ֶ ‫ּובאּו ְּב ֵב‬ ָ ‫ל־מ ָּט ֶתָך‬ ִ ‫ִמ ְׁש ָּכ ְבָך וְ ַע‬ ‫ּוב ַע ֶּמָך‬ ְ ‫ּוב ֵבית ֲע ָב ֶדיָך‬ ְ ‫ֹרותיָך‬ ֶ ‫ּוב ִמ ְׁש ֲא‬ ְ ‫ּנּוריָך‬ ֶ ‫ּוב ַת‬ ְ

καὶ ἀναβάντες εἰσελεύσονται εἰς τοὺς οἴκους σου καὶ εἰς τὰ ταμιεῖα τῶν κοιτώνων σου καὶ ἐπὶ τῶν κλινῶν σου καὶ εἰς τοὺς οἴκους τῶν θεραπόντων σου καὶ τοῦ λαοῦ σου καὶ ἐν τοῖς φυράμασίν σου καὶ ἐν τοῖς κλιβάνοις σου·

Obviously, we should not ask the translator what he wanted to express with the difference between ἐν and εἰς. He [31] simply lost touch with the preceding and kept following the Hebrew. A similar case occurs in Gen 14:8–9, ‫וַ ּיַ ַע ְרכּו ִא ָּתם ִמ ְל ָח ָמה ְּב ֵע ֶמק‬ ‫ַה ִּׂש ִּדים׃ ֵאת ְּכ ָד ְר ָלע ֶֹמר ֶ ֣מ ֶלְך‬ ‫ֵע ָילם וְ ִת ְד ָעל ֶ ֣מ ֶלְך ֹּגויִם… ַא ְר ָּב ָעה‬ ‫ת־ה ֲח ִמ ָּׁשה‬ ַ ‫ְמ ָל ִכים ֶא‬

καὶ παρετάξαντο αὐτοῖς εἰς πόλεμον ἐν τῇ κοιλάδι τῇ ἁλυκῇ, πρὸς Χοδολλογόμορ βασιλέα Αἰλὰμ καὶ Θαργὰλ βασιλέα ἐθνῶν …, οἱ τέσσαρες βασιλεῖς πρὸς τοὺς πέντε.

In both cases, ‫ את‬has been rendered freely (παρατάσσεσθαι μετά also occurs in the Septuagint), and both expressions, παρατάσσεσθαι + dative and παρατάσσεσθαι πρός are current in the Septuagint and other Greek writings. The relationship with the verb was thus preserved, but not with its preceding construction. 2  See, e.g., H.J. Gehman, “Hebraisms of the Old Greek Version of Genesis”, VT 3 (1953) 141–8.

254

Ilmari Soisalon-Soininen

Another such case occurs in Lev 17:8, ‫ּומן־‬ ִ ‫ִאיׁש ִאיׁש ִמ ֵּבית יִ ְׂש ָר ֵאל‬ ‫ֹתוכם‬ ָ ‫ַהּגֵ ר ֲא ֶׁשר־יָ גּור ְּב‬

Ἄνθρωπος ἄνθρωπος τῶν νἱῶν Ἰσραὴλ καὶ ἀπὸ τῶν υἱῶν τῶν προσηλύτων τῶν προσκειμένων ἐν ὑμῖν

Both renderings, with partitive genitive and with ἀπό, are common in the Pentateuch. One could imagine that the translator lost touch with the preceding and merely proceeded mechanically. That is not too probable, however, in view of the translation of the book as a whole and especially of the preceding example. The contact with the preceding context merely became so weak that the translator was no longer aware of the preceding construction. There are other cases, too, where the rendering of prepositions has been influenced by the distance from the reference word. Prepositions have been rendered in a contextually adequate way in the Pentateuch, generally speaking. Several exceptional cases have posed difficulties for the translators. The following rendering of Lev 20:4 was probably caused by the distance from the reference word. The verb form ὑπερίδωσιν is not followed by the expected accusative, but by ἀπό, which shows that the Greek translator had lost touch with the reference word. ‫וְ ִאם ַה ְע ֵלם יַ ְע ִלימּו ַעם ָה ָא ֶרץ‬ ‫ן־ה ִאיׁש ַההּוא‬ ָ ‫יהם ִמ‬ ֶ ֵ‫ת־עינ‬ ֵ ‫ֶא‬

ἐὰν δὲ ὑπερόψει ὑπερίδωσιν οἱ αὐτόχθονες τῆς γῆς τοῖς ὀφθαλμοῖς αὐτῶν ἀπὸ τοῦ ἀνθρώπου ἐκείνου

By the way, note the free rendering ὑπερίδωσιν … τοῖς ὀφθαλμοῖς αὐτῶν. In Exod 29:29 the rendering of the Hebrew plural with a singular caused problems further on in the sentence: ‫ּובגְ ֵדי ַהּק ֶֹדׁש ֲא ֶׁשר ְל ַא ֲהר ֹן יִ ְהיּו‬ ִ ‫ְל ָבנָ יו ַא ֲח ָריו ְל ָמ ְׁש ָחה ָב ֶהם‬ ‫א־בם ֶאת־יָ ָדם׃‬ ָ ‫ּול ַמ ֵּל‬ ְ

καὶ ἡ στολὴ τοῦ ἁγίου, ἥ ἐστιν Ἀαρών, ἔσται τοῖς υἱοῖς αὐτοῦ μετ᾽ αὐτόν, χρισθῆναι αὐτοὺς ἐν αὐτοῖς καὶ τελειῶσαι τὰς χεῖρας αὐτῶν.

It is possible that the second “in them” (‫ )בם‬was not rendered as ἐν αὐτοῖς again after τελειῶσαι because its position in the sentence would then have remained undefined. The present ἐν αὐτοῖς can only be interpreted as ‘in their midst’ with difficulty. We should rather [32] assume that the translator was simply more preoccupied with the Hebrew plural than with the Greek text. But the attentive reader was almost bound to read the meaning ‘in their midst’ into the text. We have just seen that in long sentences the translator lost contact with the preceding context. However, within word groups that belong closely together he often translated freely. We may think of construct relationships in those cases where a mere genitive is not the most natural rendering. When the nomen rectum indicates the material of which something is made, an adjective is normally used,

Segmentation in the Greek Pentateuch

255

not infrequently, however, also the preposition ἐκ etc. The use of adjectives or addition of prepositions also occurs in other cases. Sometimes a dative is used instead of a genitive. Often, a unit of two words is rendered as one word. I will give just a few examples that represent a great number of instances.3 ‫ית ִמ ְׁש ְּבצֹת זָ ָהב‬ ָ ‫וְ ָע ִׂש‬ ‫ַאנְ ֵׁשי ֱא ֶמת‬ ‫ַאנְ ֵׁשי ִמ ְקנֶ ה‬ ‫ֶע ְרוַ ת ָּד ָבר‬ ‫י ְֹׁש ֵבי ָה ָא ֶרץ‬ ‫ֵא ֶׁשת ֵח ֶיקָך‬ ‫ּוכ ַבד ָלֹׁשון‬ ְ ‫ד־ּפה‬ ֶ ‫ְכ ַב‬ ‫ת־א ִּמי‬ ִ ‫ת־א ִבי ִהוא ַאְך לֹא ַב‬ ָ ‫ַב‬

καὶ ποιήσεις ἀσπιδίσκας ἐκ χρυσίου καθαροῦ ἄνδρες δίκαιοι ἄνδρες κτηνοτρόφοι ἄσχημον πρᾶγμα ὁ κατοικῶν τὴν γῆν ἡ γυνὴ ἡ ἐν κόλπῳ σου ἰσχνόφωνος καὶ βραδύγλωσσος ἀδελφή μού ἐστιν ἐκ πατρός, ἀλλ᾽ οὐκ ἐκ μητρός

Exod 28:13 Exod 18:21 Gen 46:32 Deut 24:1 Gen 36:20 Deut 13:7 Exod 4:10 Gen 20:12

These cases demonstrate that, under the right circumstances, the translators chose various free renderings without hesitation. This happened very often with small units that were understood as a whole.

3. Segmentation: Samples from Hebrew Syntax (1) Repetition. An interesting phenomenon comes to light in the rendering of the typically Hebrew repetitions. Let us begin with some general observations. Very often the repetitions are translated literally, but not always. For example, we find ‫ מאד מאד‬rendered both as σφόδρα σφόδρα and as simple σφόδρα. In such cases the segmentation obviously did not affect the rendering. The repetition of the preposition before coordinated nouns is normally replicated in the translation, but sometimes it is not. This type of repetition is quite common in Greek too. The matter is different for [33] suffixes that are repeated after coordinated nouns. In translation, the Greek possessive pronoun is often repeated, but very often it is rendered only once. But cases like Gen 28:7 (τοῦ πατρὸς καὶ τῆς μητρὸς αὐτοῦ) and 47:1 (τὰ κτήνη καὶ οἱ βόες αὐτῶν) are rare—I counted eight cases in Genesis. Much more often an order is used that is rarer in Greek, but idiomatic and elegant, e.g., ὁ τρόμος ὑμῶν καὶ ὁ φόβος (Gen 9:2); πάντες οἱ υἱοὶ αὐτοῦ καὶ αἱ θυγατέρες (Gen 37:35), which amounts to 19 cases in Genesis. The same approach is used where an object suffix is repeated after two coordinated verbs.

3  A more extensive discussion of these cases is found in Soisalon-Soininen, “A Palette of Renderings for the Hebrew Construct State Relation in the Greek Pentateuch” (in this volume).

256

Ilmari Soisalon-Soininen

‫ּול ָׁש ְמ ָרּה‬ ְ ‫ְל ָע ְב ָדּה‬ ‫וַ יְ ַח ְּב ֵקהּו וַ ּיִ ָּׁש ֵקהּו‬ ‫ּוב ֻק ָמּה‬ ְ ‫וְ לֹא־יָ ַדע ְּב ִׁש ְכ ָבּה‬

ἐργάζεσθαι αὐτὸν καὶ φυλάσσειν καὶ περιλαβὼν αὐτὸν ἐφίλησεν καὶ οὐκ ᾔδει ἐν τῷ κοιμηθῆναι αὐτὴν καὶ ἀναστῆναι

Gen 2:15 Gen 33:4 Gen 19:35

It would be rash to conclude that the translators preferred this order for stylistic reasons. It is more probable that in such enumerations every part served as a unit to the translator. The most natural practice was then to repeat the pronoun. If the translator had already rendered the first part and had not yet lost touch with the preceding phrase, omission of the repeated pronoun would have suggested itself. However, he did not return to the preceding phrase to erase the pronoun, but simply omitted the pronoun of the second and third members. That implies that the rendering of a unit was more readily influenced by what preceded than by what followed. More rarely, the translator treated such a series as a unit and omitted the first pronoun. (2) Apodotic καί. One of the most frequently occurring features—one that contrasts with normal Pentateuch usage—is the Hebraistic use of καί in the apodosis.4 Some cases are attested in classical usage, but these should rather be explained primarily by the [34] meaning ‘also.’5 In the Pentateuch, apodotic ְ‫ ו‬was not replicated in half of the cases, whereby I exclude apodoses of participium coniunctum because of their special position.6 In the other half it was rendered with καί (rarely δέ). Let me give two pairs of contrasting examples. ‫אבה ָה ִא ָּׁשה ָל ֶל ֶכת‬ ֶ ֹ ‫וְ ִאם־לֹא ת‬ ‫ית ִמ ְּׁש ֻב ָע ִתי זֹאת‬ ָ ‫ַא ֲח ֶריָך וְ נִ ִּק‬ ‫אתי‬ ִ ‫אּולי יִ ָּק ֵרה יְ הוָ ה ִל ְק ָר‬ ַ ‫ְּוד ַבר ַמה־ּיַ ְר ֵאנִ י וְ ִהּגַ ְד ִּתי ָלְך‬

ἐὰν δὲ μὴ θέλῃ ἡ γυνὴ πορευθῆναι μετὰ σοῦ εἰς τὴν γῆν ταύτην, Ø καθαρὸς ἔσῃ ἀπὸ τοῦ ὅρκου τούτου· εἴ μοι φανεῖται ὁ θεὸς ἐν συναντήσει, καὶ ρῆμα, ὃ ἐάν μοι δείξῃ, ἀναγγελῶ σοι

Gen 24:8

Num 23:3

4  I am using here the material that my pupil Anneli Aejmelaeus has collected and presented in an MA workshop, Apodoottinen καί Pentateukin Septuagintakäännöksessä (“Apodotic καί in the Septuagint translation of the Pentateuch”). See now her dissertation Parataxis in the Septuagint. A Study of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASF, Diss. Hum. Litt. 31; Helsinki: Suomalainen Tiedeakatemia, 1982), 126–47. The manuscript of this article was submitted in 1977. 5  R. Kühner/B. Gerth, Ausführliche Grammatik der griechischen Sprache: Teil 2, Satzlehre (4th ed.; Leverkusen: Gottschalksche Verlagsbuchhandlung, 1955), 225; Aejmelaeus, Parataxis, 1f. (127). 6 Aejmelaeus, Parataxis, 3 (128), the table.

Segmentation in the Greek Pentateuch

‫ת־ע ֵרי‬ ָ ‫ֹלהים ֶא‬ ִ ‫וַ יְ ִהי ְּב ַׁש ֵחת ֱא‬ ‫ֹלהים ֶאת־‬ ִ ‫ַה ִּכ ָּכר וַ ּיִ זְ ּכֹר ֱא‬ ‫ַא ְב ָר ָהם‬ ‫ת־אב ֶֹתיָך‬ ֲ ‫וְ ַת ַחת ִּכי ָא ַהב ֶא‬ ‫ֹּיוצ ֲאָך‬ ִ ַ‫וַ ְּיִב ַחר ְּבזַ ְרֹעו ַא ֲח ָריו ו‬ ‫ְּב ָפנָ יו ְּבכֹֹחו ַהּגָ ד ֹל ִמ ִּמ ְצ ָריִם׃‬

καὶ ἐγένετο ἐν τῷ ἐκτρῖψαι κύριον πάσας τὰς πόλεις τῆς περιοίκου ἐμνήσθη ὁ θεὸς τοῦ Ἀβραάμ διὰ τὸ ἀγαπῆσαι αὐτὸν τοὺς πατέρας σου καὶ ἐξελέξατο τὸ σπέρμα αὐτῶν μετ᾽ αὐτοὺς ὑμᾶς καὶ ἐξήγαγέν σε

257 Gen 19:29

Deut 4:37

Long sentences made it especially difficult for the translator to get a grasp of the whole and to find the right rendering.7 This is proven by constructions that are unnatural. But apodotic καί also occurs in short sentences. We should assume that subordinate and main clauses, to mention just one example, belonged to different translation units, and, at the same time, that the translator took pains not to lose contact with what preceded. Often he succeeded in rendering the apodosis without replicating Hebrew ְ‫ו‬, but sometimes, lacking a good grasp of the whole, he rendered it with καί. Aejmelaeus has pointed out that such καί-sentences were not considered coordinated clauses, but apodoses, because the translators shifted from the subjunctive in the coordinate clause to the indicative in the main clause.8 In long sentences, the indicative sometimes occurs too soon (e.g., Lev 19:23; 14:34–35; Num 32:20–21), further proof that the translator did not always have a grasp of the whole. When the first of two coordinated clauses is rendered as a participium coniunctum, Hebrew normally has ְ‫ ו‬in both clauses. The striking thing is that after a part. coni. apodotic καί is quite rare, viz. 20 out of 416 instances in the Pentateuch, [35] thus in roughly 5% of the cases.9 In most of these cases the apodosis is far removed from the part. coni. Possibly, apodotic καί was even less tolerable for the translators in these cases than it was after a subordinate clause. But perhaps another factor played a greater role. When the first of two coordinated ְ‫ו‬-clauses was rendered with a part. coni., the translator had to process it together with the second clause or a part of it (the predicate). Otherwise he obviously could not employ the part. coni. In the following example, two instances of part. coni. occur in one verse. The segmentation can be pictured as follows. ‫וַ ּיִ ָּׂשא ֵעינָ יו וַ ּיַ ְרא ׀ וְ ִהּנֵ ה‬ ‫ֹלׁשה ֲאנָ ִׁשים נִ ָּצ ִבים ָע ָליו‬ ָ ‫ְׁש‬ ‫אתם‬ ָ ‫׀ וַ ּיַ ְרא וַ ּיָ ָרץ ִל ְק ָר‬ ‫ִמ ֶּפ ַתח ָהא ֶֹהל ׀ וַ ּיִ ְׁש ַּתחּו‬ ‫ָא ְר ָצה׃‬

ἀναβλέψας δὲ τοῖς ὀφθαλμοῖς αὐτοῦ εἶδεν, | καὶ ἰδοὺ Ø τρεῖς ἄνδρες εἱστήκεισαν ἐπάνω αὐτοῦ | καὶ ἰδὼν Ø προσέδραμεν εἰς συνάντησιν αὐτοῖς | ἀπὸ τῆς θύρας τῆς σκηνῆς αὐτοῦ | καὶ προσεκύνησεν ἐπὶ τὴν γῆν

7 Aejmelaeus, Parataxis, 12. 8 Aejmelaeus, Parataxis, 13 (136f.). 9 Aejmelaeus, Parataxis, 17 (104).

Gen 18:2

258

Ilmari Soisalon-Soininen

It would be natural that within these units ְ‫ ו‬would not be rendered with apodotic καί. A different case is the following example. ‫יהם וְ ֶאת־‬ ֶ ֵ‫ת־מ ְקנ‬ ִ ‫וַ ּיִ ְקחּו ֶא‬ ‫כּוׁשם ֲא ֶׁשר ָר ְכׁשּו ְּב ֶא ֶרץ‬ ָ ‫ְר‬ ‫יְמה יַ ֲעקֹב‬ ָ ‫ְּכנַ ַען וַ ּיָ בֹאּו ִמ ְצ ָר‬ ‫וְ ָכל־זַ ְרֹעו ִאֹּתו׃‬

καὶ ἀναλαβόντες τὰ ὑπάρχοντα αὐτῶν καὶ πᾶσαν τὴν κτῆσιν, ἣν ἐκτήσαντο ἐν γῇ Χανάαν, καὶ εἰσῆλθον10 εἰς Αἴγυπτον, Ἰακὼβ καὶ πᾶν τὸ σπέρμα αὐτοῦ μετ᾽ αὐτοῦ

Gen 46:6

Here the translator lost contact with what preceded and used καί. But why did he employ a part. coni. at all? What I consider most likely is the following. The long protasis of v. 5 speaks of the preparations for the departure. Then the translator sees that the following part continues the description of the preparations and feels the need to speed up the narration. That is how he employs a part. coni., being under the influence of what precedes but lacking a grasp of the whole. If, with Wevers, we should read the singular εἰσῆλθεν, the contrast is even greater. In sum, it is very probable that apodotic καί in apodoses after a part. coni. is so rare just because in those cases the translator had a more natural grasp of the context. (3) ‫ ב‬+ inf. constr. In all LXX books, the Hebrew ‫ ב‬+ inf. constr. [36] is rendered in various ways. Everywhere we find ἐν τῷ + inf. but also coordinate clauses with ὅτε, ὁπότε, ἡνίκα, ὡς, and, last but not least, genitive absolute and participium coniunctum.11 It is quite possible that the rendering with ἐν τῷ + inf. suggested itself especially when the infinitive construction together with its adjuncts constituted a translation unit and the main clause followed the infinitive. Of course, one could consider the translator capable of rendering such a unit with a subordinate clause. But in Genesis, for example, all cases where a subordinate clause or a genitive absolute is used, are sentences where the main clause precedes the infinitive. Those cases with ‫ ויהי )והיה( ב‬+ inf. constr. that occur at the beginning of the sentence have been translated as ἐγένετο ἐν τῷ + inf., with only one exception. (‫ ויהי‬before the infinitive increases the possibility that the infinitive will not form a unity with the following main clause.) An example of both cases: ‫ן־ׁשמֹנִ ים ָׁשנָ ה‬ ְ ‫וְ ַא ְב ָרם ֶּב‬ ‫ת־הגָ ר‬ ָ ‫וְ ֵׁשׁש ָׁשנִ ים ְּב ֶל ֶד‬ ‫ֶאת־יִ ְׁש ָמ ֵעאל ְל ַא ְב ָרם‬

Ἀβρὰμ δὲ ἦν ὀγδοήκοντα ἓξ ἐτῶν, ἡνίκα ἔτεκεν Ἁγὰρ τὸν Ἰσμαὴλ τῷ Ἀβράμ.

Gen 16:16

10  Pace Wevers, I read the plural with Rahlfs. The sing. occurs in the old papyrus 962, as well as A, but I deem it more probable that the singular is a secondary correction towards the following subject Ἰακώβ than that the translator would have employed a singular in deviation from the Hebrew plural and the preceding plural part. coni. 11  I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965), 80ff, 188ff.

Segmentation in the Greek Pentateuch

‫ֹלהים ֶאת־‬ ִ ‫וַ יְ ִהי ְּב ַׁש ֵחת ֱא‬ ‫ֹלהים‬ ִ ‫ָע ֵרי ַה ִּכ ָּכר וַ ּיִ זְ ּכֹר ֱא‬ ‫ת־א ְב ָר ָהם‬ ַ ‫ֶא‬

καὶ ἐγένετο ἐν τῷ ἐκτρῖψαι κύριον πάσας τὰς πόλεις τῆς περιοίκου ἐμνήσθη ὁ θεὸς τοῦ Ἀβραάμ

259 Gen 19:29

In Exodus, the categories cannot be distinguished with the same clarity, but it is only natural that the distinction between the two categories is not so clear cut. (4) Pleonastic pronoun. A special case with respect to our subject is the use of the pleonastic pronoun in relative clauses. Although it occurs occasionally in normal Greek, at least in careless speech, the cases in the Septuagint are often expressions that could not have occurred outside translation Greek. In fact, they stand in stark contrast to the Greek of the Pentateuch in general. When I treated this subject before,12 I could not provide an answer to the question of how to explain that the translator deviated so noticeably from their use of Greek in general. Here I want to make up for that. I limit myself to Genesis, from which the other Pentateuch books only differ with respect to the proportion in which the various cases occur. First, some remarks about relative clauses in general. It [37] is to be assumed that relative clauses were not rendered as independent units without antecedent. On the other hand, the antecedent is such an integral part of the main clause, that it could not be separated from it. But the two together constituted units that were too long. One could imagine that the main clause was translated first, then the relative clause, but in such a way that the already translated antecedent was still connected to this unit. Some examples will illustrate this. ‫וַ ּיִ ְׁש ְּכנּו ֵמ ֲחוִ ָילה ַעד־ׁשּור ׀‬ 13‫ל־ּפנֵ י ִמ ְצ ַריִם‬ ְ ‫ֲא ֶׁשר ַע‬

κατῴκησεν δὲ ἀπὸ Ἑυιλὰτ ἕως Σούρ, ἥ ἐστιν κατὰ πρόσωπον Αἰγύπτου

Gen 25:18

Sometimes the connection with the preceding may have been quite loose. In the following example the relative clause has been transformed into an attribute. ‫י־ב ֵרְך ֲא ָב ֶר ְכָך ׀ וְ ַה ְר ָּבה‬ ָ ‫ִּכ‬ ‫ֹכוכ ֵבי‬ ְ ‫ַא ְר ֶּבה ֶאת־זַ ְר ֲעָך ׀ ְּכ‬ ‫ַה ָּׁש ַמיִם ׀ וְ ַכֹחול ֲא ֶׁשר ַעל־‬ ‫ְׂש ַפת ַהּיָ ם‬

ἦ μὴν εὐλογῶν εὐλογήσω σε καὶ πληθύνων πληθυνῶ τὸ σπέρμα σου ὡς τοὺς ἀστέρας τοῦ οὐρανοῦ καὶ ὡς τὴν ἄμμον τὴν παρὰ τὸ χεῖλος τῆς θαλάσσης

Gen 22:17

The translator often renders the finer nuances of ‫ אשר‬in relative sentences quite well, especially in short ones. Sometimes he gives a free rendering of the whole. 12  I. Soisalon-Soininen, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch,” in Studien zur Septuaginta-Syntax, 55–61. 13  The segmentation could also have been: ‫ל־ּפנֵ י ִמ ְצ ַריִם‬ ְ ‫וַ ּיִ ְׁש ְּכנּו ֵמ ֲחוִ ָילה ׀ ַעד־ׁשּור ֲא ֶׁשר ַע‬

260

Ilmari Soisalon-Soininen

‫י־עץ‬ ֵ ‫ל־ה ֵעץ ֲא ֶׁשר־ֹּבו ְפ ִר‬ ָ ‫ת־ּכ‬ ָ ‫וְ ֶא‬ ‫ז ֵֹר ַע זָ ַרע‬ ‫ל־ה ָא ֶרץ ֲא ֶׁשר־‬ ָ ‫ֹרומׂש ַע‬ ֵ ‫ּולכֹל‬ ְ ‫ֹּבו נֶ ֶפׁש ַחּיָ ה‬

καὶ πᾶν ξύλον, ὃ ἔχει ἐν ἑαυτῷ καρπὸν σπέρματος σπορίμου

Gen 1:29

καὶ παντὶ ἑρπετῷ τῷ ἕρποντι ἐπὶ τῆς γῆς, ὃ ἔχει ἐν ἑαυτῷ ψυχὴν ζωῆς

Gen 1:30

The same construction with ἔχειν we also find in Gen 7:22; 34:14; 41:38. The pronoun within the relative clause is not replicated in the following example. ‫ר־א ֶּלה ֹּלו ָאנ ִֹכי ָה ָרה‬ ֵ ‫ְל ִאיׁש ֲא ֶׁש‬

Ἐκ τοῦ ἀνθρώπου, τίνος ταῦτά ἐστιν, ἐγὼ ἐν γαστρὶ ἔχω

Gen 38:25

Very often the pronoun is processed together with ‫ אשר‬and not repeated again. This happens especially often in short relative sentences. ‫ּוק ָר ָאהּו ָאֹסון ַּב ֶּד ֶרְך ֲא ֶׁשר‬ ְ ‫כּו־בּה‬ ָ ‫ֵּת ְל‬

καὶ συμβήσεται αὐτὸν μαλακισθῆναι ἐν τῇ ὁδῷ, ᾗ ἂν πορεύησθε

Gen 42:38

‫ם־אנַ ְחנּו‬ ֲ ַ‫ִהּנֶ ּנּו ֲע ָב ִדים ַלאד ֹנִ י ּג‬ ‫ּגַ ם ֲא ֶׁשר־נִ ְמ ָצא ַהּגָ ִב ַיע ְּביָ ֹדו׃‬

ἰδού ἐσμεν οἰκέται τῷ κυρίῳ ἡμῶν, καὶ ἡμεῖς καὶ παρ᾽ ᾧ εὑρέθη τὸ κόνδυ

Gen 44:16

In the second example, ‫ ְּביָ ֹדו‬was not replicated but taken into account in παρ᾽ ᾧ. However, in many cases the pleonastic pronoun (or οὗ … ἐκεῖ, ὅθεν … ἐκεῖθεν) is used, even in short sentences. But in general these sentences are longer than those that are rendered freely, at least in Genesis. Note, in the first of the following examples, the construction ἡ γῆ … σοὶ δώσω αὐτὴν. ‫יה‬ ָ ‫ָה ָא ֶרץ ֲא ֶׁשר ַא ָּתה ׁש ֵֹכב ָע ֶל‬ ‫ּולזַ ְר ֶעָך׃‬ ְ ‫ְלָך ֶא ְּתנֶ ּנָ ה‬ ‫ֹיוסף ָאסּור ָׁשם׃‬ ֵ ‫ְמֹקום ֲא ֶׁשר‬

[38] ἡ γῆ, ἐφ᾽ ἧς σὺ καθεύδεις ἐπ᾽ αὐτῆς, σοὶ δώσω αὐτὴν καὶ τῷ σπέρματί σου

Gen 28:13

εἰς τὸν τόπον, οὗ Ιωσηφ ἀπῆκτο ἐκεῖ

Gen 40:3

The combination ‫ אשר…שם‬was rendered pleonastically in numerous cases. I would explain the translator’s approach as follows. First, he considers the whole, but while rendering the antecedent and especially undeclined ‫ אשר‬he partly loses sight of the end of the sentence and renders it again. He feels there is something he hasn’t translated yet and redoes the end of the relative clause in an isolated fashion. This picture is confirmed by those cases where the pleonastic pronoun does not fit well with the relative pronoun.

Segmentation in the Greek Pentateuch

‫יתי ִע ְּמָך‬ ִ ‫ר־ע ִׂש‬ ָ ‫ַּכ ֶח ֶסד ֲא ֶׁש‬ ‫ם־ה ָא ֶרץ‬ ָ ‫ַּת ֲע ֶׂשה ִע ָּמ ִדי וְ ִע‬ ‫ֲא ֶׁשר־ּגַ ְר ָּתה ָּבּה׃‬

ἀλλὰ κατὰ τὴν δικαιοσύνην, ἣν ἐποίησα μετὰ σοῦ, ποιήσεις μετ᾽ ἐμοῦ καὶ τῇ γῇ, ᾗ σὺ παρῴκησας ἐν αὐτῇ

261 Gen 21:23

Παροικεῖν can be construed with the dative, but in the meaning ‘to live close to something.’ In the above rendering the translator did not take ‫ בה‬into account. ‫ר־ה ְת ַה ַּל ְכ ִּתי ְל ָפנָ יו‬ ִ ‫יְ הוָ ה ֲא ֶׁש‬ ‫ת־ׁשם ַה ָּמֹקום‬ ֵ ‫וַ ּיִ ְק ָרא יַ ֲעקֹב ֶא‬ ‫ֹלהים‬ ִ ‫ֲא ֶׁשר ִּד ֶּבר ִאֹּתו ָׁשם ֱא‬ ‫ית־אל׃‬ ֵ ‫ֵּב‬

Κύριος, ᾧ εὐηρέστησα ἐναντίον αὐτοῦ

Gen 24:40

καὶ ἐκάλεσεν Ιακωβ τὸ ὄνομα τοῦ τόπου, ἐν ᾧ ἐλάλησεν μετ᾽ αὐτοῦ ἐκεῖ ὁ θεός, Βαιθήλ.

Gen 35:15

Sometimes this procedure could result in a new meaning that was hardly intended by the translator: ‫א־ת ַּקח ִא ָּׁשה ִל ְבנִ י ִמ ְּבֹנות‬ ִ ֹ‫ל‬ ‫ַה ְּכנַ ֲענִ י ֲא ֶׁשר ָאנ ִֹכי י ֵֹׁשב ְּב ַא ְרֹצו‬

Οὐ λήμψῃ γυναῖκα τῷ υἱῷ μου ἀπὸ τῶν θυγατέρων τῶν Χαναναίων, ἐν οἷς ἐγὼ παροικῶ ἐν τῇ γῇ αὐτῶν

Gen 24:37

The unintended meaning is ‘in whose midst I live in their land.’ In this case, too, the end of the sentence was not taken into account in the rendering of ‫אשר‬. The relative pronoun was given a fitting meaning too rashly. In my last example, the translator rendered ‫ אשר‬simply with a nominative, although it did not fit at all into the context. ‫ּכֹל ֲא ֶׁשר־ֹלו ְסנַ ִּפיר וְ ַק ְׂש ֶק ֶׂשת‬

πάντα, ὅσα ἐστὶν ἐν αὐτοῖς πτερύγια καὶ λεπίδες

Deut 14:9

The peculiarities of the relative sentences are, to my mind, best explained in the light of segmentation, although we should beware of simplification. It is psychologically understandable that in this way unidiomatic expressions arose that were not representative of the translator’s general usage.

4. Concluding Remarks Let me conclude with the caveat that the division into [39] segments that were rendered without adequate contact with the context should in no way be regarded as an all-sufficient explanation for the characteristics of Septuagint Greek. It is only one of the factors to take into account in the most divergent cases. We

262

Ilmari Soisalon-Soininen

should also bear in mind that in similar cases the segmentation could have been different. The degree of contact with the context certainly varied. A couple of words that were difficult to translate may have broken the contact more easily than an easy text. Be that as it may, it is certain that backward contact was stronger than forward contact. It is also possible that the segments were divided into even smaller fragments, and were even rendered word for word. The fact that most LXX translators very rarely deviated from the word order of the source text points to that. In spite of all these qualifications, I believe that consideration of the factors just discussed can serve linguistic research of the Septuagint.

Ilmari Soisalon-Soininen The Rendering of the Hebrew Personal Pronoun as Subject in the Greek Pentateuch1 1984 Abstract In clauses with finite verbs, Hebrew personal pronouns that occur as a subject are commonly rendered with a Greek pronoun (and αὐτός). Pronouns are rarely omitted but frequently added for stylistic reasons if Hebrew only has a finite verb without a pronominal subject. Verbless clauses are interesting since the pronominal subject cannot be omitted. The most common rendering of a pronominal subject is the copula εἶναι. This is not always possible or desirable, however. Often the pronoun is rendered as (any type of) pronoun, equally often with or without an added copula. In exceptional cases, the sentence structure is changed. If the predicate of the verbless clause is a participle, it is almost always rendered as a finite verb. More often than not, the subject is rendered explicitly. This raises the question of whether the subject was intended to be emphasized, the question of interference, and the question of whether the result does justice to the sentence perspective of the original. For the LXX translators, literal rendering of the pronominal subject seems to have been preferred over omission, in spite of the often-resulting Hebraisms. The statistics show that Genesis and Exodus represent a freer approach than the rest of the Pentateuch, especially Numbers.

1  This is a translation of “Die Wiedergabe des hebräischen Personalpronomens als Subjekt im griechischen Pentateuch,” in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 71–85, by Theo van der Louw.

264

Ilmari Soisalon-Soininen

0. Introduction [71] In Hebrew, a personal pronoun occurs as a subject with a finite verb only if the subject is emphasized.2 Because personal pronouns as subjects are likewise mainly used to mark contrast or emphasis in the Greek language,3 it is natural that the Septuagint translators commonly followed the Hebrew parent text in this respect. Exceptions however, can be found. This paper deals with the rendering of the pronominal subject of the Hebrew verbless clause. Its distinguishing feature is that the subject must necessarily be present, emphasized or not. [72] Obviously, the same holds true for the Greek verbless clause, but the tendency to render verbless clauses as verbal clauses makes the rendering of the subject problematic. The present paper only covers Pentateuch material.

1. The Rendering of the Pronoun as the Subject of a Finite Verb 1.1 Pronoun Rendered as Pronoun As would be expected, a personal pronoun collocated with a perfect or imperfect form is commonly rendered literally, e.g., ‫ׁשּופָך רֹאׁש וְ ַא ָּתה‬ ְ ְ‫הּוא י‬ ‫ׁשּופּנּו ָע ֵקב‬ ֶ ‫ְּת‬

αὐτός σου τηρήσει κεφαλήν, καὶ σὺ τηρήσεις αὐτοῦ πτέρναν

Gen 3:15

In the 3rd person, αὐτός or ἐκεῖνος is sometimes used: ‫ֹר־ציִ ד‬ ַ ‫הּוא־היָ ה גִ ּב‬ ָ ‫ֵה ָּמה יַ ֲעלּו ָב ָהר‬

οὗτος ἦν γίγας κυνηγός

Gen 10:9

ἐκεῖνοι ἀναβήσονται ἐπὶ τὸ ὄρος

Exod 19:13

Often the pronoun is used with an imperative. In some of these cases, the pronoun is rendered as αὐτός (Gen 42:19).4 In Exod 5:11, ὑμεῖς is added to αὐτοί.

2  W. Gesenius, Gesenius’ Hebrew Grammar (ed. E. Kautzsch; transl. A.E. Cowley; Oxford: Clarendon, 1910), §135. 3  E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften: 2,1, Satzlehre, Analytische Teil, erste Hälfte (Leipzig: De Gruyter, 1926), 62; cf. F. Blass/A. Debrunner/F. Rehkopf, Grammatik des neutestamentlichen Griechisch (Göttingen: Vandenhoeck & Ruprecht, 1976), 228. 4  This sometimes also happens when the pronoun is used with other verb forms, e.g., Exod 19:4 ‫יתי ְל ִמ ְצ ָריִם‬ ִ ‫יתם ֲא ֶׁשר ָע ִׂש‬ ֶ ‫ ַא ֶּתם ְר ִא‬Αὐτοὶ ἑωράκατε ὅσα πεποίηκα τοῖς Αἰγυπτίοις.

The Rendering of the Personal Pronoun as Subject

‫וְ ַא ֶּתם ְּפרּו ְּורבּו‬ ‫וְ ַא ֶּתם ְלכּו‬ ‫ַא ֶתם ְלכּו ְקחּו ָל ֶכם ֶת ֶבן‬

265

ὑμεῖς δὲ αὐξάνεσθε καὶ πληθύνεσθε

Gen 9:7

αὐτοὶ δὲ βαδίσατε

Gen 42:19

αὐτοὶ ὑμεῖς πορευόμενοι συλλέγετε ἑαυτοῖς ἄχυρα

Exod 5:11

In Hebrew, the emphasis is sometimes strengthened by ‫גם‬, in which case the pronoun often follows the verb. Normally ‫ גם‬is rendered as καί. But at the start of a sentence, καί can easily be taken for a conjunction (Gen 48:19). ‫ֶפן־יָמּות גַ ם־הּוא‬ ‫ה־ל ָעם‬ ְ ֶ‫גַ ם־הּוא יִ ְהי‬

Μήποτε ἀποθάνῃ καὶ οὗτος

Gen 38:11

καὶ οὗτος ἔσται εἰς λαόν

Gen 48:19

In one case the copula was left untranslated, with a verbless clause as a result: ‫ל־חי‬ ָ ‫יְתה ֵאם ָּכ‬ ָ ‫ִּכי ִהוא ָה‬

ὅτι αὕτη μήτηρ πάντων τῶν ζώντων

Gen 3:20.

Sometimes the pronoun has to be employed because another subject occurs alongside it (so Lev 10:14 below). In such cases, the verb is conjugated just as in Hebrew (Gen 31:44). If the verb precedes, it can appear in the singular, because it is quite natural in Greek as well that a preceding verb agrees only with the first member of a compound subject (Lev 25:54).5 If this agreement leads to a 1st or 2nd person verb form, the verb does not agree with the compound subject as a whole (Gen 20:7; Num 20:8 below). ‫נּופה‬ ָ ‫וְ ֵאת ֲחזֵ ה ַה ְּת‬ ‫אכלּו‬ ְ ֹ ‫רּומה ּת‬ ָ ‫ֹׁשוק ַה ְּת‬ ‫וְ ֵאת‬ ‫ּובנֶ יָך‬ ָ ‫ְּב ָמֹקום ָטֹהור ַא ָּתה‬ ‫ּובנ ֶֹתיָך ִא ָּתְך‬ ְ

καὶ τὸ στηθύνιον τοῦ ἀφορίσματος καὶ τὸν βραχίονα τοῦ ἀφαιρέματος φάγεσθε ἐν τόπῳ ἁγίῳ, σὺ καὶ οἱ υἱοί σου καὶ ὁ οἶκός σου μετὰ σοῦ

Lev 10:14

‫ְל ָכה נִ ְכ ְר ָתה ְב ִרית ֲאנִ י‬ ‫וָ ָא ָּתה‬

διαθώμεθα διαθήκην ἐγὼ καὶ σύ

Gen 31:44

‫וְ יָ ָצא ִּב ְׁשנַ ת ַהּי ֵֹבל הּוא‬ ‫ּובנָ יו ִעֹּמו‬ ָ

ἐξελεύσεται ἐν τῷ ἔτει τῆς ἀφέσεως αὐτὸς καὶ τὰ παιδία αὐτοῦ μετ᾽ αὐτοῦ

Lev 25:54

5  E. Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit. Mit Einschluss der gleichzeitigen Ostraka und der in Ägypten verfassten Inschriften. 2,3, Satzlehre: Syntetischer Teil (Leipzig: De Gruyter, 1934), 30.

266

Ilmari Soisalon-Soininen

‫ַּדע ִּכי־ֹמות ָּתמּות ַא ָּתה‬ ‫ר־לְך׃‬ ָ ‫ל־א ֶׁש‬ ֲ ‫וְ ָכ‬

[73] γνῶθι ὅτι ἀποθανῇ σὺ καὶ πάντα τὰ σά

Gen 20:7

‫ת־ה ֵע ָדה ַא ָּתה‬ ָ ‫וְ ַה ְק ֵהל ֶא‬ ‫וְ ַא ֲהר ֹן‬

καὶ ἐκκλησίασον τὴν συναγωγὴν σὺ καὶ Ἀαρών

Num 20:8

The last two examples have to be understood so that γνῶθι ὅτι ἀποθανῇ and καὶ ἐκκλησίασον τὴν συναγωγήν are complete sentences that were subsequently supplied with a compound subject—an interpretation that could be enforced by placing a comma between both parts. Sometimes such cases of disagreement are avoided by using a plural verb form: ‫ל־ּת ְׁש ְּת ַא ָּתה‬ ֵ ‫יַ יִ ן וְ ֵׁש ָכר ַא‬ ‫ּובנֶ יָך ִא ָּתְך‬ ָ

Οἶνον καὶ σίκερα οὐ πίεσθε, σὺ καὶ οἱ υἱοί σου μετὰ σοῦ

Lev 10:9

‫ְל ַמ ַען ִּת ָירא ֶאת־יְ הוָ ה‬ ֣‫ּובנְ ך‬ ִ ‫… ַא ָּתה‬ ‫ֹלהיָך‬ ֶ ‫ֱא‬ ‫ן־ּבנְ ָך‬ ִ ‫ּוב‬ ֶ

ἵνα φοβῆσθε κύριον τὸν θεὸν ὑμῶν …, σὺ καὶ οἱ υἱοί σου καὶ οἱ υἱοὶ τῶν υἱῶν σου

Deut 6:2

If the verb follows the compound subject, it is plural in both Hebrew and Greek. Sometimes the first member of the compound subject is mentioned in full and is then repeated in the form of a pronoun and followed by the other subjects. Next to Num 31:19, I would like to quote Deut 3:1, and point to the peculiar rendering in Num 23:6, where ὅδε renders ‫הּנֵ ה‬,ִ and ‫ הּוא‬was left untranslated. ‫ְך־ה ָּב ָׁשן‬ ַ ‫וַ ּיֵ ֵצא ֹעוג ֶמ ֶל‬ ‫ל־עֹּמו‬ ַ ‫אתנּו הּוא וְ ָכ‬ ֵ ‫ִל ְק ָר‬ ‫וְ ִהּנֵ ה נִ ָּצב ַעל־ע ָֹלֹתו הּוא‬ ‫ֹמואב׃‬ ָ ‫ל־ׂש ֵרי‬ ָ ‫וְ ָכ‬

καὶ ἐξῆλθεν Ὢγ βασιλεὺς τῆς Βασὰν εἰς συνάντησιν ἡμῖν, αὐτὸς καὶ πᾶς ὁ λαὸς αὐτοῦ

Deut 3:16

καὶ ὅδε ἐφειστήκει ἐπὶ τῶν ὁλοκαυτωμάτων αὐτοῦ, καὶ πάντες οἱ ἄρχοντες Μωὰβ μετ᾽ αὐτοῦ

Num 23:6

As in the case of the additional subjects, the single subject is sometimes emphasized with the help of a pronoun: ‫וְ ֶה ֶבל ֵה ִביא גַ ם־הּוא‬ ‫ִמ ְּבכֹֹרות צֹאֹנו‬

καὶ Ἅβελ ἤνεγκεν καὶ αὐτὸς ἀπὸ τῶν πρωτοτόκων τῶν προβάτων αὐτοῦ

Gen 4:4

6  In a parallel passage, Num 21:33, the resumptive pronoun is absent, καὶ ἐξῆλθεν Ὢγ … καὶ πᾶς ὁ λαὸς αὐτοῦ, but αὐτός is supplied in Ncdgnptx Syr.

267

The Rendering of the Personal Pronoun as Subject

‫ם־הוא יָ ְל ָדה‬ ִ ַ‫וְ ִצ ָּלה ג‬

Σελλὰ δὲ ἔτεκεν καὶ αὐτή

Gen 4:22

If both the subject and its pronominal resumption that gives it emphasis occur together either before or after the verb, the pronoun is not rendered: ‫ם־הוא יָ ְל ָדה ֵּבן‬ ִ ַ‫וְ ַה ְּצ ִע ָירה ג‬ ‫ם־הם ַח ְר ֻט ֵּמי‬ ֵ ַ‫וַ ּיַ ֲעׂשּו ג‬ ‫ִמ ְצ ַריִם‬

ἔτεκεν δὲ καὶ ἡ νεωτέρα υἱόν

Gen 19:38

καὶ ἐποίησαν καὶ οἱ ἐπαοιδοὶ τῶν Αἰγυπτίων

Exod 7:11

More than once, a subject is resumed pronominally after a relative clause:7 ‫ָה ִא ָּׁשה ֲא ֶׁשר נָ ַת ָּתה ִע ָּמ ִדי‬ ‫ן־ה ֵעץ‬ ָ ‫ה־ּלי ִמ‬ ִ ָ‫ִהוא נָ ְתנ‬ ‫וָ א ֵֹכל׃‬

Ἡ γυνή, ἣν ἔδωκας μετ᾽ ἐμοῦ, αὕτη μοι ἔδωκεν ἀπὸ τοῦ ξύλου, καὶ ἔφαγον.

Gen 3:12

In Gen 44:4, the pronominal subject becomes a genitive as a corollary of the genitive absolute. Something similar happens in Deut 4:34, where the first two words of v. 35 are now joined to it in the form of a prepositional phrase.

1.2 Pronoun Omitted Omission of the pronoun is very rare. We encountered some scattered cases above.8 Some of these cases are due to special circumstances. In Gen 24:45 ‫ויהי‬ was read as ‫ ;אני‬in 47:19 the clause construction was changed. [74] In Exod 18:19 the repeated ‫ ַא ָּתה‬is reduced to one σύ, and something similar happens to repeated ‫ הּוא‬in Deut 32:6. When rendering laws, the translator is more focused on clearly delimiting the stages of the case than on representing each pronoun, especially when the participant remains the same: … ‫ֹאו ִכי יִ ּגַ ע ְּב ֻט ְמ ַאת ָא ָדם‬ ‫וְ נֶ ְע ַלם ִמ ֶּמּנּו וְ הּוא יָ ַדע וְ ָא ֵׁשם‬

‫ימי‬ ֵ ‫ל־יְמי ֹזוב ֻט ְמ ָא ָתּה ִּכ‬ ֵ ‫ָּכ‬ ‫נִ ָּד ָתּה ִּת ְהיֶ ה ְט ֵמ ָאה ִהוא׃‬

ἢ ἅψηται ἀπὸ ἀκαθαρσίας ἀνθρώπου (…) καὶ ἔλαθεν αὐτόν, μετὰ τοῦτο δὲ γνῷ καὶ πλημμελήσῃ,

Lev 5:3

πᾶσαι αἱ ἡμέραι ῥύσεως ἀκαθαρσίας αὐτῆς καθάπερ αἱ ἡμέραι τῆς ἀφέδρου, ἀκάθαρτος ἔσται

Lev 15:25

7  In Gen 24:7 this happens after a very long relative clause. 8  Gen 19:38; Exod 7:11; Num 23:6 (and 21:33).

268

Ilmari Soisalon-Soininen

The remaining cases are the following: ‫וַ ּיָ ָקם ַּפ ְרעֹה ַליְ ָלה הּוא‬ ‫ל־מ ְצ ַריִם‬ ִ ‫ל־ע ָב ָדיו וְ ָכ‬ ֲ ‫וְ ָכ‬

καὶ ἀναστὰς Φαραὼ νυκτὸς καὶ πάντες οἱ θεράποντες αὐτοῦ καὶ πάντες οἱ Αἰγύπτιοι

Exod 12:30

‫ּובנָ יו‬ ָ ‫וְ יָ ָצא ֵמ ִע ָּמְך הּוא‬ ‫ִעֹּמו‬

καὶ ἐξελεύσεται τῇ ἀφέσει καὶ τὰ τέκνα αὐτοῦ μετ᾽ αὐτοῦ (+ αὐτός bcw Arm Boh Eth)

Lev 25:41

‫ּוׁש ַמ ְר ֶּתם ַא ֶּתם ֶאת־‬ ְ ‫ת־מ ְׁש ָּפ ַטי‬ ִ ‫ֻחּק ַֹתי וְ ֶא‬

καὶ φυλάξεσθε πάντα τὰ νόμιμά μου καὶ πάντα τὰ προστάγματά μου

Lev 18:26

‫ת־ּפנַ י‬ ָ ‫וְ ַׂש ְמ ִּתי ֲאנִ י ֶא‬ ‫ָּב ִאיׁש ַההּוא‬

καὶ ἐπιστήσω τὸ πρόσωπόν μου ἐπὶ τὸν ἄνθρωπον ἐκεῖνον

Lev 20:5

Ἵνα τί ὑμεῖς παραβαίνετε τὸ ρῆμα κυρίου; οὐκ εὔοδα ἔσται ὑμῖν.

Num 14:41

τὸ γὰρ αἷμα αὐτοῦ ἀντὶ τῆς ψυχῆς ἐξιλάσεται

Lev 17:11

‫ָל ָּמה ּזֶ ה ַא ֶּתם ע ְֹב ִרים‬ ‫ת־ּפי יְ הוָ ה וְ ִהוא לֹא‬ ִ ‫ֶא‬ ‫ִת ְצ ָלח׃‬ ‫י־ה ָּדם הּוא ַּבּנֶ ֶפׁש יְ ַכ ֵּפר‬ ַ ‫ִּכ‬

1.3 Pronoun Added Addition of a pronoun occurs quite often. Sometimes this is due to a different Vorlage, but most instances should be explained otherwise. The pronoun is certainly an addition in cases where ὁ δὲ εἶπεν, ἡ δὲ εἶπεν, οἱ δὲ εἶπαν occur as renderings of ‫אמר‬ ֶ ֹ ‫וַ ּי‬, ‫אמר‬ ֶ ֹ ‫וַ תּ‬, ‫אמרּו‬ ְ ֹ ‫וַ ּי‬. The translator presumably wanted to express the gender implied in the Hebrew verb form. But this does not explain cases like Gen 31:11 ‫ → וָ א ַֹמר‬ἐγὼ δὲ εἶπα and 44:26 ‫אמר‬ ֶ ֹ ‫ → וַ ּנ‬ἡμεῖς δὲ εἴπαμεν. These additions should be understood as follows: the translator wanted to employ δέ, which is very appropriate here. But since εἶπεν δέ does not have the right stress, he chose to add ὁ, ἡ etc. These renderings show the translator’s fine feeling for the language. For a statistical presentation, see Table 1 below. In Gen 38:23 the rendering of ‫ ִהּנֵ ה‬as μέν caused the addition of ἐγώ: ‫ִהּנֵ ה ָׁש ַל ְח ִּתי ַהּגְ ִדי ַהּזֶ ה‬ ‫אתּה‬ ָ ‫וְ ַא ָּתה לֹא ְמ ָצ‬ ‫וַ ּיִ ְל ְקטּו ַה ַּמ ְר ֶּבה וְ ַה ַּמ ְמ ִעיט׃‬

ἐγὼ μὲν ἀπέσταλκα τὸν ἔριφον τοῦτον, σὺ δὲ οὐχ εὕρηκας.

Gen 38:23

καὶ συνέλεξαν, ὁ τὸ πολὺ καὶ ὁ τὸ ἔλαττον

Exod 16:17

269

The Rendering of the Personal Pronoun as Subject

Special instances are Gen 44:19, where the polite address was rendered as a vocative, and 41:40, where a Hebrew text with ‫ אני‬would be inconceivable: ‫ת־ע ָב ָדיו‬ ֲ ‫ֲאד ֹנִ י ָׁש ַאל ֶא‬

κύριε, σὺ ἠρώτησας τοὺς παῖδάς σου

Gen 44:19

ָ‫ַרק ַה ִּכ ֵּסא ֶאגְ ַּדל ִמ ֶּמּך‬

πλὴν τὸν θρόνον ὑπερέξω σου ἐγώ

Gen 41:40

In the other passages too, the absence of a pronoun is generally quite natural, but its presence simply better Greek: ‫ם־ה ָא ֶרץ‬ ָ ‫ַּת ֲע ֶׂשה ִע ָּמ ִדי וְ ִע‬ ‫ֲא ֶׁשר־ּגַ ְר ָּתה ָּבּה׃‬ ‫ת־ה ַּל ַחץ‬ ַ ‫יתי ֶא‬ ִ ‫ם־ר ִא‬ ָ ַ‫וְ ג‬ ‫יתָך יַ ֲעׂשּו‬ ִ ִ‫ר־צּו‬ ִ ‫ְּככֹל ֲא ֶׁש‬

ποιήσεις μετ᾽ ἐμοῦ καὶ τῇ γῇ, ᾗ σὺ παρῴκησας ἐν αὐτῇ

Gen 21:23

κἀγὼ ἑώρακα τὸν θλιμμόν

Exod 3:9

κατὰ πάντα, ὅσα ἐγὼ ἐνετειλάμην σοι, ποιήσουσιν

Exod 31:11

Sometimes a pronoun is lacking in Hebrew before an additional subject if the first subject is expressed with a noun. In those cases [75] a pronoun has been added in Greek: ‫ל־א ֶׁשר־ֹלו‬ ֲ ‫וַ ּיִ ַּסע יִ ְׂש ָר ֵאל וְ ָכ‬

Ἀπάρας δὲ Ισραηλ, αὐτὸς καὶ πάντα τὰ αὐτοῦ

Gen 46:1

Table 1: Addition of pronominal subjects Gen

Exod

Lev

Num

Deut

total

41 14

7 7

0 7

2 3

0 3

50 34

of ὁ δέ, ἡ δέ etc. other additions

From “other additions” I excluded the instances where the Greek pronoun reflects a Hebrew dativus ethicus. This rendering, which nicely expresses the emphasis, occurs once in Lev and 7 times in Deut. The dativus ethicus is also rendered as a dative in Greek, which gives the wrong sense, but sometimes, quite appropriately, as οὖν: ‫ְּפנּו ָל ֶכם ָצפֹנָ ה‬

ἐπιστράφητε οὖν ἐπὶ βορρᾶν

Deut 2:3.

270

Ilmari Soisalon-Soininen

2. Verbless Clauses 2.1 The Predicate is a Noun, an Adjective or a Localization We now turn to the rendering of the pronoun in verbless clauses. These can be divided into two categories. To the first category belong those instances where the predicate is a noun, an adjective or a localization, to the second the cases where it is a participle. The boundary between these two groups is not always clear cut since participles of stative verbs can often be considered adjectives. In general, participles are often used as nouns or adjectives. Sentences whose subject is not expressed by an independent pronoun but by a suffix need to be treated separately. Greek also has its verbless clauses, but their use is more restricted. Only the forms ἐστίν, εἰσίν, ἔστω, εἶναι can be omitted. In main clauses, omission can take place in the following cases: 1) in generally valid facts derived from experience; 2) in brief parenthetical remarks; 3) in statements of necessity and duty and with verbal adjectives ending in -τέος; 4) in the concise style characteristic of charts, in titles, and in official forms.9 A tendency to avoid verbless clauses in Greek is thus to be expected.

2.1.1 Rendering of the Pronoun as a Mere Copula (εἶναι) The most common rendering of a pronoun that functions as a subject in a verbless clause is εἶναι. In this way, a verbless clause becomes a verbal clause, the subject becoming implicit. In other words, it yields a truly Greek result. As the Hebrew personal pronoun has the same function here as the copula in Greek, it is no wonder the translators intuitively seized on this option. Aquila also employed this type of rendering and considered it, apparently, a literal [76] translation.10 Some examples from the LXX Pentateuch: ‫י־לי ָא ִחי הּוא‬ ִ ‫ִא ְמ ִר‬ ‫ת־מי ַא ְּת‬ ִ ‫ַּב‬ ‫ֵאל ַקּנָ א הּוא‬ ‫וְ לֹא יָ ְכלּו ִל ְׁשּתֹת ַמיִם‬ ‫ִמ ָּמ ָרה ִּכי ָמ ִרים ֵהם‬ ‫לֹא ַב ָּׁש ַמיִם ִהוא‬

εἰπὸν ἐμὲ ὅτι Ἀδελφός μού ἐστιν

Gen 20:13

Τίνος εἶ θυγάτηρ;

Gen 24:47

θεὸς ζηλωτής ἐστιν

Exod 34:14

καὶ οὐκ ἠδύναντο πιεῖν ἐκ Μέρρας, πικρὸν γὰρ ἦν

Exod 15:23

οὐκ ἐν τῷ οὐρανῷ ἄνω ἐστίν

Deut 30:12

9 Mayser, Grammatik 2,3, 16–19. 10  I. Soisalon-Soininen, “Einige Merkmale der Übersetzungsweise von Aquila,” in Studien zur Septuaginta-Syntax, 26.

The Rendering of the Personal Pronoun as Subject

271

Especially in a sentence with a casus pendens, which is repeated together with the pronoun, the pronoun has the function of a copula: | ‫ָה ֲאנָ ִׁשים ָה ֵא ֶּלה ְׁש ֵל ִמים‬

‫ֵהם ִא ָּתנּו‬

‫ֹלהים‬ ִ ‫וְ ַה ֻּלחֹת ַמ ֲע ֵׂשה ֱא‬ ‫ֵה ָּמה וְ ַה ִּמ ְכ ָּתב ִמ ְכ ַּתב‬ ‫ֹלהים הּוא‬ ִ ‫ֱא‬

Οἱ ἄνθρωποι οὗτοι εἰρηνικοί εἰσιν μεθ᾽ ἡμῶν

Gen 34:21

καὶ αἱ πλάκες ἔργον θεοῦ ἦσαν, καὶ ἡ γραφὴ γραφὴ θεοῦ ἐστιν

Exod 32:16

The same holds true for relative clauses, where ‫ ֲא ֶׁשר‬is the subject and is followed by a resumptive pronoun: ‫הּוא־חי‬ ַ ‫ל־ר ֶמׂש ֲא ֶׁשר‬ ֶ ‫ָּכ‬ ‫א־א ִחיָך‬ ָ ֹ ‫ִאיׁש נָ ְכ ִרי ֲא ֶׁשר ל‬ ‫הּוא‬

καὶ πᾶν ἑρπετόν, ὅ ἐστιν ζῶν

Gen 9:3

ἄνθρωπον ἀλλότριον, ὅτι οὐκ ἀδελφός σού ἐστιν

Deut 17:15

Two adjacent clauses may share the copula: ‫ם־ּבן הּוא וַ ֲה ִמ ֶּתן אֹֹתו‬ ֵ ‫ִא‬ ‫ם־ּבת ִהיא וָ ָחיָ ה‬ ַ ‫וְ ִא‬

ἐὰν μὲν ἄρσεν ᾖ, ἀποκτείνατε αὐτό, ἐὰν δὲ θῆλυ, περιποιεῖσθε αὐτό

Exod 1:16

As the examples show, various moods and tenses are used according to the content of the sentence. The nature of the Hebrew verbless clause often demands a present tense in translation, but imperfect and future occur too. In Genesis, for example, the imperfect occurs 8 times and the future 5 times: ‫וַ ּיֵ ְדעּו ִּכי ֵע ֻיר ִּמם ֵהם‬ ‫ל־א ֶרץ ִמ ְצ ַריִם‬ ֶ ‫ִּכי־טּוב ָּכ‬ ‫ָל ֶכם הּוא‬

καὶ ἔγνωσαν ὅτι γυμνοὶ ἦσαν

Gen 3:7

τὰ γὰρ πάντα ἀγαθὰ Αἰγύπτου ὑμῖν ἔσται

Gen 45:20

In the legal parts of the Pentateuch, verbless conditional clauses occur, which the translator renders with a subjunctive: ‫וְ ִאם ֹׁשור נַ ּגָ ח הּוא‬

ἐὰν δὲ ὁ ταῦρος κερατιστὴς ᾖ

Exod 21:29

A few times, verbless clauses are turned into participial constructions so that the participle comes to represent the Hebrew pronoun (see examples below). In Gen 38:21; 25:29 an attributive participle is used, and in Lev 15:25 a genitive absolute:

272

Ilmari Soisalon-Soininen

‫וְ ָׂש ָרה ׁש ַֹמ ַעת ֶּפ ַתח‬ ‫ָהא ֶֹהל וְ הּוא ַא ֲח ָריו‬ ‫וְ הּוא נַ ַער‬ ‫ַאּיֵ ה ַה ְּק ֵד ָׁשה ִהוא ָב ֵעינַ יִם‬ ‫ן־ה ָּׂש ֶדה‬ ַ ‫וַ ּיָבֹא ֵע ָׂשו ִמ‬ ‫וְ הּוא ָעיֵ ף‬ ‫ל־ה ִּמ ְׁש ָּכב‬ ַ ‫וְ ִאם ַע‬

Σὰρρα δὲ ἤκουσεν πρὸς τῇ θύρᾳ τῆς σκηνῆς, οὖσα ὄπισθεν αὐτοῦ.

Gen 18:10

ὢν νέος

Gen 37:2

Ποῦ ἐστιν ἡ πόρνη ἡ γενομένη ἐν Αἰνάν;

Gen 38:21

ἦλθεν δὲ Ἠσαὺ ἐκ τοῦ πεδίου ἐκλείπων

Gen 25:29

ἐὰν δὲ ἐν τῇ κοίτῃ αὐτῆς οὔσης

Lev 15:23

The subject expressed by way of a suffix is sometimes expressed with a [77] form of εἶναι, too: ‫ַאּיֶ ָּכה‬ ‫ִהּנֶ ּנּו ֲע ָב ִדים ַלאד ֹנִ י‬ ‫וְ ִהּנָ ם ְּב ֶא ֶרץ ּג ֶֹׁשן‬

Ἀδάμ, ποῦ εἶ;

Gen 3:9

ἰδού ἐσμεν οἰκέται τῷ κυρίῳ ἡμῶν

Gen 44:16

καὶ ἰδού εἰσιν ἐν γῇ Γέσεμ

Gen 47:111

With respect to ‫ ַאיִ ן‬+ suffix we should note that ‫ ַאיִ ן‬itself is already rendered as οὐκ ἔστιν. The suffix determines the person, and so the finite form of εἶναι, but it is not expressed separately apart from the already explicit subject, if present. In the last two examples below, verbs other than εἶναι are employed: ‫ֹיוסף ֵאינֶ ּנּו וְ ִׁש ְמֹעון ֵאינֶ ּנּו‬ ֵ ‫ִּכי ֵאינֶ ּנִ י ְּב ִק ְר ְּב ֶכם‬ ‫ָה ֶא ָחד ֵאינֶ ּנּו‬ ‫וַ ּיִ ְת ַה ֵּלְך ֲחֹנוְך ֶאת־‬ ‫ֹלהים וְ ֵאינֶ ּנּו‬ ִ ‫ָה ֱא‬

Ἰωσὴφ οὐκ ἔστιν, Συμεὼν οὐκ ἔστιν

Gen 42:36

οὐ γάρ εἰμι μεθ᾽ ὑμῶν·

Deut 1:42

ὁ εἷς οὐχ ὑπάρχει

Gen 42:32 cf. 42:13 Gen 5:24

καὶ εὐηρέστησεν Ἑνὼχ τῷ θεῷ καὶ οὐχ ηὑρίσκετο

2.1.2 Rendering of the Pronoun by a Pronoun Translating the personal pronoun as a copula is not always possible. It is limited to clauses that say what, who, or where the subject is. The predicate is an indefinite noun, an adjective or a localization. If the predicate is a definite noun, the subject must be present. As the translators follow this rule, it may be considered a 11  Cf. also Exod 2:20; Deut 1:10.

The Rendering of the Personal Pronoun as Subject

273

valid rule for Greek, at least in its Koine form. It is quite obvious that one can say, e.g., βασιλεὺς εἶ but not ὁ βασιλεὺς εἶ. ‘You are king’ should be σῦ εἶ ὁ βασιλεύς. This is because of the emphasis on the subject in these cases (“you, no one else”). In other cases where the subject is emphasized, it must also be expressed. The if and how of this emphasis are determined by the content of the sentence, but the translator often has the option of interpreting the sentence differently. It is notable that a pronoun occurring at the beginning of a clause is rarely rendered as a mere copula. In contrast, a pronoun occurring further on in a clause can be seen as either with or without emphasis. If rendering a pronoun as a mere copula is ruled out, it is often translated as a mere pronoun, with a verbless clause as a result,12 or else the copula is added. Verbless clauses occur far beyond the boundaries described by Mayser. Firstly, verbless clauses are also used for 1st and 2nd persons. Secondly, 3rd person verbless clauses do not fit into the categories defined by Mayser. Adding a word or rendering one word by two words [78] was so alien to the translators that they often opted for less natural renderings. It is very probable that the translators did not sense a big difference between the verbless clauses they were familiar with in Greek and the ones they themselves employed. But the frequent addition of the copula shows that their use of verbless clauses was not uninhibited. It seems impossible to find any difference between verbless and verbal clauses: similar clauses in similar contexts are often rendered variously, i.e., with or without copula. The balance between an orientation towards the source or the target language must have been such that the translators sometimes had recourse to a more idiomatic, sometimes to a Hebraistic solution. Let me add that I do not think of conscious deliberations on the part of the translators, I just want to describe the factors that led to their intuitive decisions. A typical example of the instances where a definite predicate appears is Gen 2:10–14: ‫אׁשים‬ ִ ‫ּומ ָּׁשם יִ ָּפ ֵרד וְ ָהיָ ה ְל ַא ְר ָּב ָעה ָר‬ ִ 10 ‫ ֵׁשם ָה ֶא ָחד ִּפיֹׁשון הּוא ַהּס ֵֹבב ֵאת‬11 … ‫ילה‬ ָ ִ‫ל־א ֶרץ ַה ֲחו‬ ֶ ‫ָּכ‬ ‫ֹּסובב ֵאת‬ ֵ ‫ם־הּנָ ָהר ַה ֵּׁשנִ י ּגִ יֹחון הּוא ַה‬ ַ ‫ וְ ֵׁש‬13 ‫ל־א ֶרץ ּכּוׁש‬ ֶ ‫ָּכ‬ ‫יׁשי ִח ֶּד ֶקל הּוא ַהה ֵֹלְך‬ ִ ‫ וְ ֵׁשם ַהּנָ ָהר ַה ְּׁש ִל‬14 ‫ִק ְד ַמת ַאּׁשּור‬ ‫יעי הּוא ְפ ָרת‬ ִ ‫וְ ַהּנָ ָהר ָה ְר ִב‬

ἐκεῖθεν ἀφορίζεται εἰς τέσσαρας ἀρχάς. 11 ὄνομα τῷ ἑνὶ Φισών· οὗτος ὁ κυκλῶν πᾶσαν τὴν γῆν Εὑιλάτ … 13 καὶ ὄνομα τῷ ποταμῷ τῷ δευτέρῳ Γηών· οὗτος ὁ κυκλῶν πᾶσαν τὴν γῆν Αἰθιοπίας. 14 καὶ ὁ ποταμὸς ὁ τρίτος Τίγρις· οὗτος ὁ πορευόμενος κατέναντι Ἀσσυρίων. ὁ δὲ ποταμὸς ὁ τέταρτος, οὗτος Εὐφράτης.

12  In Deut 32:4 (‫ → ַצ ִדיק וְ יָ ָשר הּוא‬δίκαιος καὶ ὅσιος κύριος), where the Tetragram appears for a pronoun, I assume a deviating Vorlage, viz. ‫ יהוה‬for ‫הּוא‬.

274

Ilmari Soisalon-Soininen

As happens very frequently, ‫ הּוא‬is rendered as the demonstrative pronoun οὗτος, so that the personal pronoun is not, strictly speaking, the subject in Greek. In the next set of examples of a definite predicate one can see how cases (the first two) with and (the last two) without the added copula alternate: ‫ָאנ ִֹכי ֵע ָׂשו ְּבכ ֶֹרָך‬ ‫ֲהֹלוא ָאנ ִֹכי ֲאתֹנְ ָך‬ ‫ן־מ ְל ָּכה‬ ִ ‫ָאנ ִֹכי ֶּב‬ ‫אמר ֲאנִ י ִּבנְ ָך ְבכ ְֹרָך ֵע ָׂשו‬ ֶ ֹ ‫וַ ּי‬

Ἐγὼ Ἠσαὺ ὁ πρωτότοκός σου

Gen 27:19

Οὐκ ἐγὼ ἡ ὄνος σου

Num 22:30

Θυγάτηρ Βαθουὴλ εἰμὶ ἐγώ

Gen 24:24

Ἐγώ εἰμι ὁ υἱός σου ὁ πρωτότοκος Ἠσαύ

Gen 27:32

The verbless clause is especially appropriate in parenthetical remarks, and here too we find the copula added most of the time (last two examples): ‫ת־ר ֶסן‬ ֶ ‫ת־נ֣ינְ וֵ ה… וְ ֶא‬ ִ ‫וַ ֶּיִבן ֶא‬ ‫ּובין ָּכ ַלח ִהוא ָה ִעיר‬ ֵ ‫ֵּבין נִ ינְ וֵ ה‬ ‫ַהּגְ ד ָֹלה‬

καὶ ᾠκοδόμησεν τὴν Νινευή … καὶ τὴν Δάσεμ ἀνὰ μέσον Νινευὴ καὶ ἀνὰ μέσον Χαλάχ· αὕτη ἡ πόλις ἡ μεγάλη.

Gen 10:12

…‫ר־רבּו‬ ָ ‫ֵה ָּמה ֵמי ְמ ִר ָיבה ֲא ֶׁש‬

τοῦτο ὕδωρ ἀντιλογίας, ὅτι ἐλοιδορήθησαν…

Num 20:13

καὶ βασιλέως Βάλα (αὕτη ἐστὶν Σήγωρ).

Gen 14:2

καὶ ἐτάφη ἐν τῇ ὁδῷ Ἐφράθα (αὕτη ἐστὶν Βηθλέεμ)

Gen 35:19

[79] ‫ּומ ֶלְך ֶּב ַלע ִהיא־צ ַֹער‬ ֶ

‫וַ ִּת ָּק ֵבר ְּב ֶד ֶרְך ֶא ְפ ָר ָתה ִהוא‬ ‫ ָל ֶחם‬ ‫ֵּבית‬

An often-repeated phrase is the formula ‫אנִ י יהוה‬, ֲ often with additional attributes. In Deuteronomy the formula occurs in the 3rd person, e.g., 10:21 ‫הּוא ְת ִה ָל ְתָך וְ הּוא‬ ‫ֹלהיָך‬ ֶ ‫א‬. ֱ In these cases the predicate is definite, so it cannot be rendered with a mere copula. Renderings are equally divided between a verbless clause and one with a copula added, ‫ֹלהי ַא ְב ָר ָהם‬ ֵ ‫ֲאנִ י יהוה ֱא‬ ‫ָא ִביָך‬

Ἐγὼ κύριος ὁ θεὸς Ἀβραὰμ τοῦ πατρός σου

Gen 28:13

275

The Rendering of the Personal Pronoun as Subject

‫ֹלהי ַא ְב ָר ָהם ָא ִביָך‬ ֵ ‫ָאנ ִֹכי ֱא‬

Ἐγώ εἰμι ὁ θεὸς Ἀβραὰμ τοῦ πατρός σου

Gen 26:2413

ὅτι ἐγὼ κύριος ὁ ἁγιάζων ὑμᾶς

Exod 31:13

‫ִּכי ֲאנִ י יְ הוָ ה ר ְֹפ ֶאָך‬

ἐγὼ γάρ εἰμι κύριος ὁ ἰώμενός σε

Exod 15:26

‫יכם‬ ֶ ‫ֹלה‬ ֵ ‫ֲאנִ י יְ הוָ ה ֱא‬

ἐγὼ κύριος ὁ θεὸς ὑμῶν

Lev 20:24

‫ִּכי ֲאנִ י יְ הוָ ה ְמ ַק ִּד ְׁש ֶכם‬

Table 2: Renderings of ‫ ֲאנִ י יהוה‬with a pronoun

without copula with copula

Gen

Exod

Lev

Num

Deut

total

4 5

11 9

25 26

5 1

6 1

51 42

Verbless clauses occur, with adjectives as a predicate, in blessings and curses. ‫ית ּזֹאת ָארּור ַא ָּתה‬ ָ ‫ִּכי ָע ִׂש‬ ‫וְ ַע ָּתה ָארּור ָא ָּתה‬ ‫ּוברּוְך‬ ָ ‫ָּברּוְך ַא ָּתה ָּב ִעיר‬ ‫ַא ָּתה ַּב ָּׂש ֶדה‬

Ὅτι ἐποίησας τοῦτο, ἐπικατάρατος σύ

Gen 3:14

καὶ νῦν ἐπικατάρατος σύ

Gen 4:11

εὐλογημένος σὺ ἐν πόλει, καὶ εὐλογημένος σὺ ἐν ἀγρῷ

Deut 28:314

In principle, εἶ could be used here, but the pronoun is emphasized and the verbless clause works very well here. Some other examples with an adjective as predicate: ‫טמא(ה) הּוא ָל ֶכם‬

ἀκάθαρτον τοῦτο ὑμῖν

Lev 11:4, 5, 6, 7, 815

‫ְט ֵמ ִאים ֵה ָּמה ָל ֶכם‬

ἀκάθαρτα ταῦτα ὑμῖν ἐστιν

Lev 11:2816

An indefinite noun as a predicate occurs in the first two examples, among others, and a localization in the second two:

13  One could imagine that ‫ ֲאנִ י‬was rendered as ἐγώ, ‫ ָאנ ִֹכי‬as ἐγώ εἰμι (as would become the rule for Aquila), but this appears not to have been the case. 14  Cf. also vv. 6, 16, 19. 15  Lev 11:8 in plur. 16  Similarly Lev 11:35; 11:12 (βδέλυγμα τοῦτό ἐστιν ὑμῖν).

276

Ilmari Soisalon-Soininen

‫וְ ָאנ ִֹכי ָע ָפר וָ ֵא ֶפר‬

ἐγὼ δέ εἰμι γῆ καὶ σποδός

Gen 18:27

Ὑμεῖς λαὸς σκληροτράχηλος

Exod 33:5

‫וְ ִהּנֵ ה ָאנ ִֹכי ִע ָּמְך‬

καὶ ἰδοὺ ἐγὼ μετὰ σοῦ

Gen 28:15

‫ַא ַחי ֵמ ַאיִ ן ַא ֶּתם‬

Ἀδελφοί, πόθεν ἐστὲ ὑμεῖς;

Gen 29:4

‫ם־ק ֵׁשה־ע ֶֹרף‬ ְ ‫ַא ֶּתם ַע‬

[80] The addition of a copula is necessary if the syntax requires an imperfect or future tense, or a subjunctive. These cases are rare, however: ‫ית־א ְב ָרם‬ ַ ‫וְ ֵהם ַב ֲע ֵלי ְב ִר‬ ‫יכם‬ ֶ ‫וְ ַא ֶתם ְב ֶא ֶרץ א ֵֹיְב‬ ‫ה־הוא‬ ִ ‫ּובינְ ָך ַמ‬ ֵ ‫ֵבינִ י‬

οἳ ἦσαν συνωμόται τοῦ Ἀβράμ

Gen 14:13

καὶ ὑμεῖς ἔσεσθε ἐν τῇ γῇ τῶν ἐχθρῶν ὑμῶν

Lev 26:34

ἀνὰ μέσον ἐμοῦ καὶ σοῦ τί ἂν εἴη τοῦτο;

Gen 23:15

The proportion between the two modes of translation in the Pentateuch, either without or with a copula in the present tense, is as follows: Table 3: Renderings of the pronominal subject with a pronoun

without copula with copula

Gen

Exod

Lev

Num

Deut

total

33 33

18 25

35 37

16 9

29 9

131 113

There is no indication, by the way, that the translators of these books avoided verbless clauses in the 1st or 2nd person. In Genesis and Exodus, verbal clauses are preferred, and verbless clauses in Numbers and Deuteronomy, while Leviticus holds a middle position. The former books display a freer, the latter books a more literal translation style. The study of other phenomena has led to a similar picture.

2.1.3 The Sentence Character is Changed There is a category of exceptional cases where the character of the sentence is changed. If the personal pronoun repeats the nominal or pronominal (‫זה‬, ‫מי‬ ‫אשר‬,) subject of a clause, it is sometimes omitted altogether instead of being rendered as a copula. This results in a verbless clause whose subject is a noun or a non-personal pronoun:

The Rendering of the Personal Pronoun as Subject

‫ד־ציִ ד‬ ַ ‫י־אֹפוא הּוא ַה ָּצ‬ ֵ ‫ִמ‬ ‫ִּכי ָב ֵּתי ָע ֵרי ַה ְלוִ ּיִ ם ִהוא‬ ‫ֲא ֻחּזָ ָתם‬ ‫ם־ה ֵא ֶּלה‬ ָ ִ‫א־מ ָע ֵרי ַהֹּגוי‬ ֵ ֹ ‫ֲא ֶׁשר ל‬ ‫ֵהּנָ ה‬ ‫ִּכי ַה ָּדם הּוא ַהּנָ ֶפׁש‬

277

Τίς οὖν ὁ θηρεύσας μοι θήραν…;

Gen 27:33

ὅτι οἰκίαι τῶν πόλεων τῶν Λευιτῶν κατάσχεσις αὐτῶν

Lev 25:33

αἳ οὐχὶ ἐκ τῶν πόλεων τῶν ἐθνῶν τούτων

Deut 20:15

ὅτι τὸ αἷμα αὐτοῦ ψυχή

Deut 12:23

In the last example, ‫ הּוא‬has been rendered as αὐτοῦ, with a slightly different meaning as a result. Sometimes a verbless subordinate clause is rendered in combination with the main clause. The instances below are excellent examples of adequate and elegant renderings: ‫וְ ִאם־נְ ֵק ָבה ִהוא וְ ָהיָ ה ֶע ְר ְּכָך‬ ‫ֹלׁשים ָׁש ֶקל‬ ִ ‫ְׁש‬

τῆς δὲ θηλείας ἔσται ἡ συντίμησις τριάκοντα δίδραχμα

Lev 27:4

ἀπὸ δὲ τῶν κτηνῶν τῶν μὴ καθαρῶν δύο δύο, ἄρσεν καὶ θῆλυ

Gen 7:2

‫וַ ֵּת ֶרא אֹֹתו ִּכי־ֹטוב הּוא‬

ἰδόντες δὲ αὐτὸ ἀστεῖον

Exod 2:2

‫ם־ק ֵׁשה־ע ֶֹרף ַא ָּתה‬ ְ ‫ִּכי ַע‬

διὰ τὸ λαὸν σκληροτράχηλόν σε εἶναι

Exod 33:3

‫ן־ה ְּב ֵה ָמה ֲא ֶׁשר לֹא ְטה ָֹרה‬ ַ ‫ּומ‬ ִ ‫ִהוא ְׁשנַ יִם ִאיׁש וְ ִא ְׁשֹּתו‬

[81] Sometimes the adjectival predicate is rendered as a verb so that a verbless clause turns into a verbal one (occasionally with omission of the subject):17 ‫ִכי ָעיֵ ף ָאנ ִֹכי‬ ‫י־חיֹות ֵהנָ ה ְב ֶט ֶרם ָתבֹוא‬ ָ ‫ִכ‬ ‫ֲא ֵל ֶהן ַה ְמיַ ֶל ֶדת‬ ‫ם־דל הּוא‬ ַ ‫וְ ִא‬ ‫י־עצּום הּוא ִמ ֶמנִ י‬ ָ ‫ִכ‬ ‫וְ ַא ָתה ָעיֵ ף וְ יָ גֵ ַע‬

ὅτι ἐκλείπω

Gen 25:30

τίκτουσιν γὰρ πρὶν ἢ εἰσελθεῖν πρὸς αὐτὰς τὰς μαίας

Exod 1:19

Ἐὰν δὲ πένηται

Lev 14:21

ὅτι ἰσχύει οὗτος ἢ ἡμεῖς

Num 22:6

σὺ δὲ ἐπείνας καὶ ἐκοπίας

Deut 25:18

17  Occasionally a nominalized participle (with article) is also rendered as a finite verb, e.g., Deut 8:18 ‫ → ִכי הּוא ַהנ ֵֹתן ְלָך כ ַֹח‬ὅτι αὐτός σοι δίδωσιν ἰσχύν. Cf. also Deut 3:22; 9:3.

278

Ilmari Soisalon-Soininen

‫ם־איׁש ָענִ י הּוא‬ ִ ‫וְ ִא‬ ‫ְפ ֶח ֶתת ִהוא ְב ָק ַר ְחּתֹו אֹו‬ ‫ְבגַ ַב ְחּתֹו‬ ‫ר־א ֶלה ּלֹו ָאנ ִֹכי‬ ֵ ‫ְל ִאיׁש ֲא ֶש‬ ‫ָה ָרה‬ ‫ִמ ֶמּנּו הּוא‬

ἐὰν δὲ ὁ ἄνθρωπος πένηται

Deut 24:12

ἐστήρικται ἐν τῷ ἱματίῳ, ἢ ἐν τῷ στήμονι ἢ ἐν τῇ κρόκῃ

Lev 13:55

Ἐκ τοῦ ἀνθρώπου, τίνος ταῦτά ἐστιν, ἐγὼ ἐν γαστρὶ ἔχω

Gen 38:25

ἐξ αὐτοῦ ἐποίησαν

Exod 39:5 (LXX 36:12)

The last example is a free but very apt rendering. A subject expressed as a suffix can be omitted more easily than an independent pronoun. Nevertheless, it is often preserved, ‫ִהנָ ְך ָה ָרה וְ י ַֹל ְד ְת ֵבן‬ ‫עֹודָך ָחי‬ ְ ‫ִכי‬

Ἰδοὺ σὺ ἐν γαστρὶ ἔχεις, καὶ τέξῃ υἱόν

Gen 16:11

ἔτι γὰρ σὺ ζῇς

46:3018

If ‫ ִהנֵ נִ י‬occurs as the reply to an address, it is frequently rendered unnaturally as ἰδοὺ ἐγώ,19 but sometimes we find the appropriate τί ἐστιν;20

2.2 The Predicate is a Participle A verbless clause with a participle as a predicate—under the condition that the participle is not adjectival or nominalized—has a character different from the verbless clauses just mentioned. Although it formally is a verbless clause, the participle is used verbally, and functionally we may consider it a verbal clause. The translators understood the character of these clauses, and almost always rendered the participle as a finite form. This also holds true of Aquila, in spite of his extreme literalism.21 Because a finite form is used, the subject receives a different position than it has in Hebrew. In a [82] verbless clause, the pronominal subject that receives no emphasis must also be expressed, in a verbal clause only the emphasized subject. We already saw that the unemphasized subject is mostly rendered as a copula. But in the cases we now review, that option does not exist. The translator can only choose between using an unemphasized subject or omitting 18  In Gen 43:27, 28; Exod 4:18, where the same phrase occurs, the pronominal subject is not rendered. 19  Gen 22:1, 11; 27:1, 18; 37:13. 20  Gen 22:7; 31:11; 46:2; Exod 3:4. 21 Soisalon-Soininen, “Einige Merkmale der Übersetzungsweise von Aquila”, 19–27.

279

The Rendering of the Personal Pronoun as Subject

it—if we do not consider the rare deviating renderings. The question arises to what extent omissions of the subjects, i.e., free renderings, occur. In some of the cases, obviously, a contrast or other emphasis of the subject may come into play, which makes omission of the pronoun impossible: ‫ן־ה ָב ָקר‬ ַ ‫ּוב‬ ֶ ‫וַ יִ ַקח ֶח ְמ ָאה וְ ָח ָלב‬ ‫יהם‬ ֶ ֵ‫ֲא ֶשר ָע ָשה וַ יִ ֵתן ִל ְפנ‬ ‫יהם ַת ַחת ָה ֵעץ‬ ֶ ‫וְ הּוא־ע ֵֹמד ֲע ֵל‬ ‫אכלּו‬ ֵ ֹ ‫וַ י‬ ‫ל־פ ְרעֹה ֶמ ֶלְך ִמ ְצ ַריִם‬ ַ ‫ַד ֵבר ֶא‬

‫ל־א ֶשר ֲאנִ י ד ֵֹבר ֵא ֶליָך׃‬ ֲ ‫ֵאת ָכ‬

‫ץ־כנַ ַען ֲא ֶשר ֲאנִ י ֵמ ִביא‬ ְ ‫ֶא ֶר‬ ‫ֶא ְת ֶכם ָש ָמה‬

ἔλαβεν δὲ βούτυρον καὶ γάλα καὶ τὸ μοσχάριον, ὃ ἐποίησεν, καὶ παρέθηκεν αὐτοῖς, καὶ ἐφάγοσαν· αὐτὸς δὲ παρειστήκει αὐτοῖς ὑπὸ τὸ δένδρον.

Gen 18:8

λάλησον πρὸς Φαραὼ βασιλέα Αἰγύπτου ὅσα ἐγὼ λέγω πρὸς σέ

Exod 6:29

… γῆς Χανάαν, εἰς ἣν ἐγὼ εἰσάγω ὑμᾶς ἐκεῖ

Lev 18:3

In other instances, it is open to debate whether the subject is emphasized or not (the first two examples below), and many times clearly does not receive emphasis at all. ‫ל־ה ָא ֶרץ ֲא ֶשר ַא ֶתם י ְֹש ִבים‬ ָ ‫ַע‬ ‫ָבּה‬

ἐπὶ τῆς γῆς, ἐφʼ ἣν ὑμεῖς κατοικήσετε

Num 33:55

‫ֹלהיָך ֶאת־‬ ֶ ‫ִכי־יַ ְכ ִרית יהוה ֱא‬ ‫א־ש ָמה‬ ָ ‫ַהּגֹויִם ֲא ֶשר ַא ָתה ָב‬ … ‫אֹותם‬ ָ ‫ָל ֶר ֶשת‬

Ἐὰν δὲ ἐξολεθρεύσῃ κύριος ὁ θεός σου τὰ ἔθνη, εἰς οὓς σὺ εἰσπορεύῃ ἐκεῖ κληρονομῆσαι τὴν γῆν αὐτῶν, …

Deut 12:2922

‫ַה ְמ ַכ ֶסה ֲאנִ י ֵמ ַא ְב ָר ָהם ֲא ֶשר‬ ‫ֲאנִ י ע ֶֹשה‬

Μὴ κρύψω ἐγὼ ἀπὸ Ἀβραὰμ τοῦ παιδός μου ἃ ἐγὼ ποιῶ;

Gen 18:17

μὴ ἀνελεῖν με σὺ θέλεις, ὃν τρόπον ἀνεῖλες ἐχθὲς τὸν Αἰγύπτιον;

Exod 2:14

οὐ φάγεσθε… τὸν κάμηλον, ὅτι ἀνάγει μηρυκισμὸν τοῦτο

Lev 11:4

διὰ τοῦτο αὐτὸς ἔδωκεν ὑμῖν τῇ ἡμέρᾳ τῇ ἕκτῃ ἄρτους δύο ἡμερῶν

Exod 16:29

‫ַה ְל ָה ְרגֵ נִ י ַא ָתה א ֵֹמר ַכ ֲא ֶשר‬ ‫ת־ה ִמ ְצ ִרי‬ ַ ‫ָה ַרגְ ָת ֶא‬ ‫ת־הגָ ָמל ִכי־‬ ַ ‫… ֶא‬ ‫אכלּו‬ ְ ֹ ‫לֹא ת‬ ‫ַמ ֲע ֵלה גֵ ָרה הּוא‬ ‫ל־כן הּוא נ ֵֹתן ָל ֶכם ַבּיֹום‬ ֵ ‫ַע‬ ‫יֹומיִם‬ ָ ‫ַה ִש ִשי ֶל ֶחם‬

22  In B oI f129 392+ (cprm) 18′–630′ Aeth Bo σύ is lacking.

280

Ilmari Soisalon-Soininen

‫ּובי ֶֹשר ְל ָב ְבָך‬ ְ ‫לֹא ְב ִצ ְד ָק ְתָך‬ ‫ת־א ְר ָצם‬ ַ ‫ַא ָתה ָבא ָל ֶר ֶשת ֶא‬

οὐχὶ διὰ τὴν δικαιοσύνην σου οὐδὲ διὰ τὴν ὁσιότητα τῆς καρδίας σου σὺ εἰσπορεύῃ κληρονομῆσαι τὴν γῆν αὐτῶν

Deut 9:5

In the last two examples above, the literal rendering of the pronoun creates an emphasis that distorts the sentence perspective. The instances of participles that were rendered as finite verbs [83] display notable Hebrew interference with respect to the pronominal subject. We still must investigate how often the subject is omitted in these cases. Ideally, every case would have to be reviewed. But this would unavoidably lead to subjective judgements. In any case, the cases where the use of the subject is obligatory are so rare that they would not materially alter our results. Since it is difficult to draw exact boundaries, we have to content ourselves with results that are somewhat imprecise. Table 4: Renderings of participle predicates as finite verbs

without subject with subject

Gen

Exod

Lev

Num

Deut

total

8 41

9 35

2 19

2 30

14 85

35 210

Again, Genesis and Exodus have the greatest and Leviticus and Numbers the smallest number of free renderings. The figures for Deuteronomy are not as significant as they seem since most of the cases without pronoun concern (almost) the same expression23 for which renderings with pronoun also occur. In contrast, the phrase )‫(מ ַצוֶ ה ֶא ְת ֶכם‬ ְ ‫ ֲא ֶשר ָאנ ִֹכי ְמ ַצוְ ָך‬is rendered with a pronoun 37 times. The subject that is expressed as a suffix is often rendered as a pronoun (p. [81]), but also quite frequently omitted. The relationship between the two options is as follows: Table 5: Renderings of subjects expressed as a suffix

with pronoun without pronoun

Gen

Exod

Lev

Num

Deut

total

5 3

8 3

-

1 1

1 6

15 13

23  I mean the expressions ‫ ֲא ֶשר ַא ֶתם ע ְֹב ִרים ָש ָמה‬and ‫א־ש ָמה‬ ָ ‫ר־א ָתה ָב‬ ַ ‫א ֶש‬, ֲ rendered as εἰς ἣν εἰσπορεύῃ (εἰσέλθῃς, δια-, ἀναβαίνεις) ἐκεῖ.

281

The Rendering of the Personal Pronoun as Subject

Occasionally, verbal clauses with a pronoun as subject and a participle as predicate are rendered in alternative ways. In some instances, a participle + ἦν is used.24 In Gen 42:35, ἐν τῷ + infinitive is used. Another interesting case is Gen 37:7. ‫יהם‬ ֶ ‫יקים ַׂש ֵּק‬ ִ ‫וַ יְ ִהי ֵהם ְמ ִר‬ ‫וְ ִהּנֵ ה ֲאנַ ְחנּו ְמ ַא ְּל ִמים‬ ‫ֲא ֻל ִּמים‬

ἐγένετο δὲ ἐν τῷ κατακενοῦν αὐτοὺς τοὺς σάκκους αὐτῶν

Gen 42:35

ᾤμην ἡμᾶς δεσμεύειν δράγματα

Gen 37:725

In exceptional cases the subject can be added in the rendering of a verbless clause. The following example is an adequate rendering, whereby the notion of “who follow you” has been rendered as “whom you lead,” ‫ר־ב ַרגְ ֶליָך‬ ְ ‫ל־ה ָעם ֲא ֶש‬ ָ ‫וְ ָכ‬

καὶ πᾶς ὁ λαός σου, οὗ σὺ ἀφηγῇ

Exod 11:8

2.3 Excursus: The Subject Expressed as Suffix of an Infinitive We will briefly review the rendering of the subject if expressed as the suffix [84] of an infinitive. Such subjects occur frequently with ‫ב‬, ‫כ‬, ‫ עד‬+ infinitive.26 These infinitive constructions are occasionally rendered as a subordinate clause. As a result, the subject is only visible by the ending of the conjugated verb and not rendered separately. ‫וְ ָהיָ ה ִב ְכזִ יב ְב ִל ְד ָתּה אֹתֹו‬ ‫ַעד־בֹאֹו ְל ַד ֵבר ִאּתֹו‬

αὐτὴ δὲ ἦν ἐν Χασβί, ἡνίκα ἔτεκεν αὐτούς

Gen 38:5

ἕως ἂν εἰσέλθῃ συλλαλεῖν αὐτῷ.

Exod 34:35

Similar cases occur, although rarely, in the rendering of an infinitive with suffix but without a preposition, ‫ל־ה ֲא ָד ָמה‬ ָ ‫יֹותם ַע‬ ָ ‫ִמּיֹום ֱה‬ ‫ַעד ַהּיֹום ַהזֶ ה‬

ἀφʼ ἧς ἡμέρας γεγόνασιν ἐπὶ τῆς γῆς ἕως τῆς ἡμέρας ταύτης

Exod 10:6

The standard rendering is obviously an infinitive for an infinitive. The subject is then often rendered with a pronoun in the accusative, but also frequently omitted:

24  Gen 14:12; Num 5:13 (twice); Deut 19:6, and in Deut 4:42; 19:2 without copula. 25  Cf. 41:7 (without subject). 26  Rarely with ‫ ל‬+ infinitive.

282

Ilmari Soisalon-Soininen

‫ְל ַל ֵמד ֶא ְת ֶכם ֻח ִקים‬ ‫ּומ ְש ָפ ִטים ַל ֲעש ְֹת ֶכם א ָֹתם‬ ִ … ‫וְ ַא ָתה ַה ְק ֵרב ֵא ֶליָך‬

‫נֹו־לי‬ ִ ‫ְל ַכ ֲה‬

διδάξαι ὑμᾶς δικαιώματα καὶ κρίσεις, ποιεῖν ὑμᾶς αὐτά

Deut 4:14

Καὶ σὺ προσαγάγου πρὸς σεαυτὸν τόν τε Ἀαρὼν … ἱερατεύειν μοι

Exod 28:1

These omissions occur in the most diverse situations, and are equally distributed among the books of the Pentateuch, so that they cannot serve to distinguish different ways of translating. The frequency of omission shows that it must have been a quite natural thing to do for the translators. Obviously, the subject is also lacking when a participium coniunctum is used for the infinitive construction, but not when a genitive absolute is employed.

3. Conclusion Our material shows the extent of literal and free renderings in the translation of the personal pronoun that is used as a subject. If the pronoun had been rendered literally in all cases, the result would have been even more un-Greek than it is now. An unnatural use of the pronoun has often been avoided, sometimes by simply omitting it, sometimes by rendering it as a copula. For the translators, omission seems to have been harder to swallow than employing a word with a different meaning. In contrast, omission of a subject expressed by a suffix appears to have been a trifle, probably because a suffix is not a [85] separate word.27 In some cases it was possible to make a comparison between books of the Pentateuch. We saw that Genesis and Exodus represent a freer approach than the other books. Numbers and Deuteronomy have been translated more literally, whereas Leviticus holds a middle position. With respect to the omission of the subject in verbless clauses with a participle as a predicate, Genesis and Exodus are also much freer than Leviticus and Numbers. Deuteronomy holds a middle position but it is not easy to pin down due to certain factors.

27  Aquila and Theodotion strove to render the suffixes explicitly at all times, except when finite forms correspond to an infinitive construction.

Ilmari Soisalon-Soininen Back to the Question of Hebraisms1 1990 Abstract In analyzing the linguistic character of the LXX and the differences between the individual translators, it is not enough to simply measure literalness or freedom of translation. More germane is the question if and how the translator follows the source text if the result is not idiomatic. Here the question of Hebraisms comes in. Hebraisms are thrown into relief by normal Greek usage in a particular type of context. Many Hebraisms have parallels in nontranslated Greek, but in different contexts or frequencies. It is difficult to draw sharp boundaries between Hebraisms and non-Hebraisms. Recently, there have been attempts to analyze literalness and freedom without reference to Hebraisms. The parameters suggested in Barr’s Typology of Literalism are of limited value for the comparison of translated books. It makes no sense to study consistency (stereotyping), for instance, if this is done without reference to idiomatic or Hebraistic renderings. This mistake underlies articles by Tov/ Wright and Marquis. Since their data are lumped together without discernment, the resulting statistics give a misleading impression of reliability. Rather, we should study linguistic phenomena that occur through all sorts of texts, and that lend themselves to either idiomatic or Hebraistic renderings. Only by taking idiomatic or Hebraistic usage as our point of departure, will we be able to make reliable comparisons between the individual translators.

1. Name and Nature of Literal and Free Translation [35] In research on the linguistic character of the Septuagint translation and the differences between the individual translators, there frequently occur concepts 1 This is a translation by Theo van der Louw of “Zurück zur Hebraismenfrage,” in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien zur Septuaginta. Robert Hanhart zu Ehren (Abhandlungen der Akademie der Wissenschaften in Göttingen / MSU 20; Göttingen: Vandenhoeck & Ruprecht, 1990), 35–51.

284

Ilmari Soisalon-Soininen

that are defined insufficiently or not at all. One could imagine that terms like “translation technique,” “Hebraism,” “free” and “literal” translations are so evident that they are not in need of definition. There are indications, however, that a precise delimitation of these concepts is desirable. It is regrettable that, within our own discipline, research is already diverging to such an extent that some researchers barely use the results of a different line of research. I originally took up the plan to write a syntax of the Septuagint to complete the work of Thackeray2 and Helbing,3 whose grammars only treated phonology and morphology, and that of Helbing4 and Johannessohn,5 who conducted specialist studies along these lines. But it became clear that it is much more effective to study the influence of the source language on the language of the translators than to draw up a general syntax of the translation. It allows us to concentrate on the details of the concrete usage in the translation and to compare the approaches of the different translators. The term translation technique refers to the approach of the different translators, i.e., the way in which they rendered Hebrew expressions into Greek, especially those that could not be rendered mechanically. This use of the term translation technique has a disadvantage. In principle, it should not express whether the renderings were chosen consciously or originated spontaneously. But the term translation technique easily suggests that we are dealing with a conscious choice between different possibilities. With respect to the Septuagint one should beware of this misleading way of thinking. It is not necessary, either, to hold the opposite viewpoint; it is enough to treat those questions [36] neutrally. For my part I am convinced that the oldest translators proceeded instinctively. Otherwise it would be difficult to explain that the same expression was rendered in a variety of ways and that occasionally elegant, truly Greek renderings occur. Possibly, the later Septuagint translators already had a conscious preference for a certain approach. A conscious tendency toward slavish literalism is evident in Theodotion’s translation and even more radically so in Aquila’s work.

2  H. St. J. Thackeray, A Grammar of the Old Testament in Greek according to the Septuagint (Cambridge: University Press, 1909). 3  R. Helbing, Grammatik der Septuaginta. Laut- und Wortlehre (Göttingen: Vandenhoeck & Ruprecht, 1907). 4  R. Helbing, Die Kasussyntax der Verba bei den Septuaginta. Ein Beitrag zur Hebraismenfrage und zur Syntax der Koine (Göttingen: Vandenhoeck & Ruprecht, 1927). 5  M. Johannessohn, Der Gebrauch der Präpositionen in der Septuaginta (MSU 3,3; Berlin: Weidmann, 1926).

Back to the Question of Hebraisms

285

James Barr describes the translation process aptly: “There are two sides to the process of translation, which we may call the input side and the expression side.”6 How the content in the source language (SL) is expressed in the target language (TL), what it then looks like, in other words, what happens in the translator’s mind between input and expression, that defines the translation process. The translator hears or reads the source text (ST) in small segments. He understands the content and expresses the same thought in the TL. Someone who has learned two languages in a natural environment lacks some skills that are acquired through theoretical language acquisition. If you ask him what a word means in the other language, he normally answers with one word. Only if you lay further examples of this word before him, will he adduce other counterparts. For the bilingual person, then, a word seems to have a core meaning, which may vary according to the person and, in the case of the same person, may vary according to the situation. If he starts to translate and encounters the same word or expression over and over again, it easily happens that he feels a growing urge to render it in the same fashion all the time. Or he limits himself to a few obvious possibilities and uses them repeatedly without regard to the context. But the context may also influence him to such an extent that he uses a new, contextually adequate rendering. It is very natural that translators gradually started to systematize the translation of frequent words or expressions. Because of their impression that a translation that formally replicated the source text was more precise than a freer rendering, they started, both unconsciously and consciously, to aim at the ideal of literal translation. Theodotion and more radically Aquila put this into practice. However, there is another line which is represented by the translations of Job and Proverbs, and later by Symmachus (though differently). These translations are much freer than those of the [37] Pentateuch, which was almost certainly translated first. The various translators all proceeded very differently. They noticed the tension between literal and free translation. We should keep in mind, however, that within both free and literal translation there is a scale of possibilities. For example, ‫ ל‬+ inf. cstr. was usually rendered as an infinitive without the article, but the desire for literalness led translators to render ‫ ל‬separately. That is why literal translators often use τοῦ before the infinitive, which is otherwise rare in Greek. Aquila made it a rule. However, ‫ לאמר‬was rendered as λέγων, although otherwise participles rarely occur as renderings of ‫ ל‬+ inf. cstr. But this was seen as a literal rendering, too. Only Aquila used τῷ λέγειν for ‫לאמר‬, which is different from τοῦ + infinitive and even more literal, if that is at all possible.

6  J. Barr, The Typology of Literalism in Ancient Biblical Translations (MSU 15; NAWG I, Phil-Hist. Kl.; Göttingen: Vandenhoeck & Ruprecht, 1979), 296.

286

Ilmari Soisalon-Soininen

The question is, what really is a free or literal translation? Marquis gives the following definition: Literalness in a translation is the degree of adherence to the source language reflected in the language of the translation, measured relative to a perfectly literal translation.7

Marquis thus wants to define the literal translation with the help of the perfectly literal translation. This is not only a circulus vitiosus, but it also contains the imaginary concept of a perfectly literal translation. Such a thing does not exist and cannot exist. I prefer to take free translation as my point of departure and define it as follows: A free translation expresses the content of the SL in such a way that the resulting rendering does not in any way distinguish itself from normal TL usage. For a literal translation it is not enough to express the content, it also seeks to follow the word order, the word classes and the grammatical constructions of the SL as slavishly as possible.

This definition leads us from the notion of “free translation” to the notion of “idiomatic translation” and from “literal” to “slavish.” If one investigates the character of a translation or analyzes the differences between individual translators, it is very much to the purpose to focus on the extent to which the translation follows idiomatic TL usage or deviates from it. Every translator follows SL conventions to a certain extent. But if that leads to an expression [38] that goes too much against his feel for the language, he has to find an alternative. No Septuagint translator says καὶ ἐλάλησεν τοῦ λέγειν for ‫ וידבר לאמר‬but, if he does not leave ‫ לאמר‬out altogether, καὶ ἐλάλησεν λέγων.8 Some shifts did not feel like changes, e.g., ‫וַ ּיִ ֹּלנּו ָה ָעם‬ ‫ת־הּיֶ ֶלד‬ ַ ‫וַ ִּת ַּקח נָ ֳע ִמי ֶא‬ ‫וַ ְּת ִׁש ֵתהּו ְב ֵח ָיקּה‬

καὶ διεγόγγυζεν ὁ λαός

Exod 15:24

καὶ ἔλαβεν Νωεμιν τὸ παιδίον καὶ ἔθηκεν εἰς τὸν κόλπον αὐτῆς, Origenes + ⁜ αὐτό ∕

Ruth 4:16

7  G. Marquis, “Consistency of Lexical Equivalents as a Criterion for the Evaluation of Translation Technique as Exemplified in the LXX of Ezekiel”, in C.E. Cox (ed.), VI Congress of the International Organization of Septuagint and Cognate Studies, Jerusalem 1986 (SBL SCS 23; Atlanta: SBL, 1987), 405. 8  With very few exceptions. Aquila (but not yet Theodotion) uses καὶ ἐλάλησεν τῷ λέγειν.

Back to the Question of Hebraisms

287

The first example features a shift from plural to singular, and in the second the redundant pronominal object was omitted. Sometimes a translator cannot find a good rendering and sticks to a literalistic rendering. That is how I would explain Gen 12:3: ‫וְ נִ ְב ְרכּו ְבָך ּכֹל ִמ ְׁש ְּפחֹת ָה ֲא ָד ָמה‬

καὶ ἐνευλογηθήσονται ἐν σοὶ πᾶσαι αἱ φυλαὶ τῆς γῆς

In Genesis, ‫ ב‬instrumenti is often rendered with a dative. But in Greek an instrumental dative cannot be used with a person, even less so if the person is referred to with a pronoun. Consequently, ἐν σοί cannot be understood as instrumental. Rather, the translator merely decided on a literal rendering, with a somewhat obscure result in this case. The character of the ST partly determines the approach of the translator. An extreme example would be a list of goods with exact quantities mentioned alongside. If the translator manages to translate the mentioned goods correctly, we are dealing with a perfectly literal translation in the sense of Marquis. Idiomatic and slavish translation, as I define them, then coincide. Therefore, such material makes it impossible to determine the translational approach. But even in less extreme examples the character of the ST has a pervasive influence on the extent to which the approach of the translator comes to light. The psalms, with their simple, parallel clauses, provide the translator with material very different from prophetic indictments or Deuteronomy with its complex constructions. Yet it is our task to analyze the different approaches of the individual translators on the basis of heterogeneous material. As the example of a list shows, the question is not how faithfully the translator follows his ST, but in what cases and to what extent he does so when it renders the translation unidiomatic. This is because a literal or a slavish translation is recognized by its deviation from TL usage. The tension arises between the requirements of the TL and the adherence to SL expressions. There is no point in talking about a literal or a slavish translation if a word-for-word rendering is perfectly idiomatic.

2. Definition and Illustrations of Hebraisms If the problem is seen in this light, we must address the question of [39] Hebraisms. Let me say straightaway that we touch upon a problem that is not easily solved. Some decades ago, Robert Hanhart told me that he disliked the term “Hebraisms” and preferred to speak of “Hebraizing expressions” instead. For it is difficult to draw a line between what is a Hebraism and what is not. Around the turn of the century, Adolf Deissmann sought to demonstrate that practically all biblical Greek expressions were attested in Koine. It appeared that the similarities

288

Ilmari Soisalon-Soininen

that could be traced were quite unexpected. Thackeray,9 Psichari10 and Helbing11 studied Hebraisms in more depth. In the domain of New Testament Greek, the question of Hebraisms was treated by Moulton and Blass(–Debrunner–­Rehkopf). Sollamo discussed the issue in great detail, having worked through an extensive corpus of Koine material in order to make her comparisons as reliable as possible.12 The question of Hebraisms is also explicitly dealt with in my book Die Infinitive in der Septuaginta13 and in Aejmelaeus’ dissertation.14 Helbing distinguishes between lexical, phraseological, stylistic and syntactic Hebraisms. This division, to which I shall return below, is basically followed by Sollamo. But let us first define what a Hebraism is: A Hebraism in the Septuagint is an expression calqued from Hebrew that is out of tune with contemporary Greek usage or it concerns the use of words and phrases that are prompted by Hebrew and occur in contexts where they do not fit according to normal Greek usage.

An expression that has not been used in Greek before is not necessarily a Hebraism. It may accord so well with Greek usage that it may be taken for genuine Greek. Also, a Hebraism can be accepted gradually into Greek usage so that it is no longer recognized as a Hebraism. Let us further note that if expressions that are not Hebraisms but not very common in Greek are repeated often enough, they can give a Hebraistic colour to the text as a whole. [40] This definition does not help us much in distinguishing Hebraisms from non-Hebraisms in Septuagint usage. We can hardly draw such boundaries in our own mother tongue, let alone in a dead language. Nevertheless, in several cases we can say with certainty that an expression or its use is Hebraistic. In certain areas we can draw reliable boundaries. A good example, in my opinion, is Sollamo’s study of the renderings of semiprepositions in the Septuagint. But there are other clear examples of Hebraisms as well, although the comparative material from the papyri has made us very cautious. Helbing, who gives a very wide interpretation of Hebrew influence and Hebrew colouring, nonetheless defines Hebraisms very narrowly: “On the basis of my results I maintain that we can only

9 Thackeray, Grammar, Introduction. 10  J. Psichari, “Essai sur le grec de la Septante”, REJ 55 (1908) 161–208. 11 Helbing, Kasussyntax. 12  R. Sollamo, Renderings of Hebrew Semiprepositions in the Septuagint (AASF 19; Helsinki: Suomalainen Tiedeakatemia, 1979), 161. 13 I. Soisalon-Soininen, Die Infinitive in der Septuaginta (AASF, ser. B, 132; Helsinki: Suomalainen Tiedeakatemia, 1965). 14  A. Aejmelaeus, Parataxis in the Septuagint: A Study of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASF, Diss. Hum. Litt. 31; Helsinki: Suomalainen Tiedeakatemia, 1982).

289

Back to the Question of Hebraisms

speak of a Semitism if the phrase cannot be elucidated with help of Greek usage.”15 It remains unclear what “elucidate” means here. At the end of his book he prints a list of Hebraisms involving grammatical cases that go along with certain verbs. It gives evidence for his narrow interpretation of “Hebraisms.” I would like to mention some scattered examples which show that many expressions are doubtlessly Hebraisms. ‫ֹּיולד ֶאת־‬ ֶ ַ‫ן־מ ַאת ָׁשנָ ה ו‬ ְ ‫ֵׁשם ֶּב‬ ‫ַא ְר ַּפ ְכ ָׁשד‬ …‫ִאיׁש ִאיׁש ִּכי‬ ‫וְ ָע ָׂשה ֶפ ַסח ַליהוָ ה׃‬

‫נּו־לי ִאיׁש נֶ זֶ ם ְׁש ָלֹלו‬ ִ ‫ּות‬ ְ ‫… ֵלאמֹר‬ ‫ָהיָ ה ְד ַבר־יְ הוָ ה‬

Σὴμ υἱὸς ἑκατὸν ἐτῶν, ὅτε ἐγέννησεν τὸν Ἀρφαξάδ

Gen 11:10

Ἄνθρωπος ἄνθρωπος, ὃς ἐὰν … καὶ ποιήσει τὸ πάσχα κυρίῳ

Num 9:10

καὶ δότε μοι ἀνὴρ ἐνώτιον τῶν σκύλων αὐτοῦ

Jdg 8:24

ἐγενήθη ρῆμα κυρίου … λέγων

Gen 15:1

In many cases the expression is possible but takes on a different meaning in Greek: ‫ימ ֶלְך‬ ֶ ‫ּכֹל ָּב ִעיר נִ ְל ָחם וַ ֲא ִב‬ ‫ַההּוא ַהֹּיום‬

καὶ Ἀβιμέλεχ ἐπολέμει ἐν τῇ πόλει ὅλην τὴν ἡμέραν ἐκείνην

Jdg 9:45

The Hebrew text means “against the city,” and this may well have been what the translator wanted to say. But in Greek it can only be understood locally, “in the city,” which obviously runs against the context. Such a case should be considered a Hebraism, too. Let me mention Exod 14:31: ‫ת־הּיָ ד ַהּגְ ד ָֹלה‬ ַ ‫וַ ּיַ ְרא יִ ְׂש ָר ֵאל ֶא‬ ‫ֲא ֶׁשר ָע ָׂשה יְ הוָ ה ְּב ִמ ְצ ַריִם‬

εἶδεν δὲ Ἰσραὴλ τὴν χεῖρα τὴν μεγάλην, ἃ ἐποίησεν κύριος τοῖς Αἰγυπτίοις

[41] The relative pronoun ἅ is neuter plural, whereas τὴν χεῖρα is feminine singular. The phrase ἣν ἐποίησεν* would distort the sense. In Hebrew, ‫ ַהּיָ ד ַהּגְ ד ָֹלה‬refers to God’s acts, and therefore ἅ can be used. The translator renders the sentence correspondingly because in Greek it is possible to express the idea as follows: “And Israel saw the great hand, (namely) that which the Lord did to Egypt.” Most Hebraisms in the Septuagint concern expressions that are possible as such, yet not in the particular context. I will give some examples. 15 Helbing, Kasussyntax, VII.

290

Ilmari Soisalon-Soininen

(1) Ἰδού is already used adverbially by Aristophanes. Having lost the meaning ‘behold,’ it could also be used with verbs of hearing. So, even ἰδοὺ φώνη μεγάλη is not a Hebraism per se. But the frequency of ἰδού gives the text a Hebraizing flavour. This word even occurs in complicated constructions, fully in accordance with Greek usage, e.g. 1 Kgdms (Sam) 3:11:16 ‫מּואל ִהּנֵ ה‬ ֵ ‫ל־ׁש‬ ְ ‫אמר יְ הוָ ה ֶא‬ ֶ ֹ ‫וַ ּי‬ ‫ָאנ ִֹכי ע ֶֹׂשה ָד ָבר ְּביִ ְׂש ָר ֵאל‬

καὶ εἶπεν κύριος πρὸς Σαμουήλ Ἰδοὺ ἐγὼ ποιῶ τὰ ῥήματά μου ἐν Ισραήλ …

The rendering ἰδοὺ ἐγώ becomes very unnatural when it renders ‫ ִהּנֵ נִ י‬as a response to an address. Compare these two examples: ‫אמר‬ ֶ ֹ ‫אמר ֵא ָליו ַא ְב ָר ָהם וַ ּי‬ ֶ ֹ ‫וַ ּי‬ ‫ִהּנֵ נִ י׃‬

καὶ εἶπεν πρὸς αὐτόν Ἀβραάμ, Ἀβραάμ· ὁ δὲ εἶπεν Ἰδοὺ ἐγώ

Gen 22:1

‫ל־א ְב ָר ָהם‬ ַ ‫אמר יִ ְצ ָחק ֶא‬ ֶ ֹ ‫וַ ּי‬ ‫אמר ִהּנֶ ּנִ י‬ ֶ ֹ ‫אמר ָא ִבי וַ ּי‬ ֶ ֹ ‫ָא ִביו וַ ּי‬ ‫ְבנִ י‬

εἶπεν δὲ Ισαὰκ πρὸς Ἀβραὰμ τὸν πατέρα αὐτοῦ εἴπας Πάτερ. ὁ δὲ εἶπεν Τί ἐστιν, τέκνον;

Gen 22:7

The free, idiomatic rendering in v. 7 proves that the LXX translators themselves sensed that ἰδοὺ ἐγώ was unnatural. In other cases too, they wondered about ἰδού and occasionaly omitted it. (2) The temporal construction ‫ ויהי… ו‬is generally rendered as και ἐγένετο (ἐγενήθη)… καί. Among other things, γίνεσθαι means ‘to happen’ (collocated with a noun but also with ὥστε + infinitive). In the Septuagint, and partly in the NT, it is used in a new construction, calqued from Hebrew. In this way it is a Hebraism, though well comprehensible in Greek, e.g., Gen 19:29: ‫ת־ע ֵרי‬ ָ ‫ֹלהים ֶא‬ ִ ‫וַ יְ ִהי ְּב ַׁש ֵחת ֱא‬ ‫ת־א ְב ָר ָהם‬ ַ ‫ֹלהים ֶא‬ ִ ‫ַה ִּכ ָּכר וַ ּיִ זְ ּכֹר ֱא‬

καὶ ἐγένετο ἐν τῷ ἐκτρῖψαι κύριον πάσας τὰς πόλεις τῆς περιοίκου ἐμνήσθη ὁ θεὸς τοῦ Ἀβραάμ

(3) Regarding προστιθέναι + inf. as a rendering of Hebrew ‫ויוסף‬, Helbing writes: I would like to state quite clearly that one of [42] the last remaining examples that Thumb still considers Hebraisms, viz. προστιθέναι + inf. for Hebrew ‫ ויוסף ל‬+ inf. (‘continually, again’), can now be termed a good Greek construction in view of Pseudo-­Callisthenes II, 41 (final part) οὐκέτι οὖν προσεθέμην ἀδύνατα ἐπιχειρεῖν.

16  Note in this example the reference to a future occurrence.

Back to the Question of Hebraisms

291

But that is an exaggeration. Good Greek though the construction may be, its use in the Septuagint is not. The first example below cannot be good Greek by any means. ‫ת־ה ֶבל‬ ָ ‫ת־א ִחיו ֶא‬ ָ ‫וַ ּת ֶֹסף ָל ֶל ֶדת ֶא‬ ‫אמר‬ ַ ֹ ‫וַ ּי ֶֹסף ֹעוד ְל ַד ֵּבר ֵא ָליו וַ ּי‬

καὶ προσέθηκεν τεκεῖν τὸν ἀδελφὸν αὐτοῦ τὸν Ἀβέλ

Gen 4:2

καὶ προσέθηκεν ἔτι λαλῆσαι πρὸς αὐτὸν καὶ εἶπεν

Gen 18:29

The expression is Hebraistic both in 4:2 and in 18:29 because, to my mind, it cannot be used in the sense of ‘further’ if the action is in fact unique. In 18:29 the Hebrew sentence means ‘he further said to him,’ but the Greek sentence could mean at most ‘he continued speaking to him and said to him, among other things.’ (4) It is disputed whether φοβεῖσθαι ἀπό should be considered a Hebraism. To my mind, BDF exaggerate when they write: In addition to the accusative, the NT also employs ἀπό with the genitive with verbs of ‘fearing, fleeing, avoiding’ etc. which was in part possible already in classical, but was encouraged by Semitic influence (Johannessohn II 245 n. 7, 276f.; Helb., Kas. 24–36, 71f.).17

Helbing, to whom they refer, writes: φοβεῖσθαι likewise contains originally the idea of separation. As Homer shows, it really means “to be scared away from, to flee,” and could be construed with ἀπό = “away from,” cf. Ilias 16, 303 φοβέοντο ἀπὸ νηῶν. The original sense was certainly never extinguished—which is why ἀπό did not strike anyone as odd.18

Helbing oversimplifies the matter. For he admits: Apart from the LXX and the NT I could not find this usage attested, but ἀπό turns up in Modern Greek; cf. Psichari 186.

He probably refers to the expression ἀπό τινος ‘whoever’ here. Mayser mentions no cases from his papyrical material. To my mind, LSJ’s estimate of φοβεῖσθαι ἀπό τινος as “prob. a Hebraism” is nearer the mark.

17  F. Blass/A. Debrunner/R. Funk, A Greek Grammar of the New Testament and Other Early Christian Writings (Cambridge: Cambridge University Press, 1961), §149. 18 Helbing, Kasussyntax, 294.

292

Ilmari Soisalon-Soininen

(5) As I have demonstrated elsewhere,19 ἐν instrumenti does occur in Koine, but in a limited fashion. In the Septuagint, however, it is [43] widespread. There are cases where it is improperly used where there are many alternative options, e.g., the instrumental dative. Although one could use ἐν ὅπλοις, ‘armed’ on the analogy of ἐν στολῇ, that does not authorize us to consider the following use of ἐν good Greek: ‫ת־הּיַ ְר ֵּדן‬ ַ ‫ִּכי ְב ַמ ְק ִלי ָע ַב ְר ִּתי ֶא‬ ‫יתי ִל ְׁשנֵ י ַמ ֲחֹנות׃‬ ִ ִ‫ַהּזֶ ה וְ ַע ָּתה ָהי‬

ἐν γὰρ τῇ ῥάβδῳ μου διέβην τὸν Ἰορδάνην τοῦτον, νῦν δὲ γέγονα εἰς δύο παρεμβολάς

Gen 32:11 (10)

This sentence is incomprehensible to a Greek if its continuation does not guide him to get the right sense. So far, I have given some scattered examples, but any page of the Septuagint will yield more. The examples show how difficult it is to draw a clear line between Hebraisms and non-Hebraisms. If we consider that Hebraisms can be idiomatic to a greater or lesser degree, we must conclude that, as a category, they form no basis for the comparison of the various translators. Our focus on Hebraisms should not cause us to lose sight of the other side of the coin, viz. the (occasional) presence of very apt, idiomatic renderings. The latter involve, to a greater or lesser degree, a break with the customary word-forword approach. I already mentioned τί ἐστιν for ‫ ִהּנֵ נִ י‬as a response to an address. Let me give some more examples. ‫וַ ּיִ ְׁש ַאל ָל ֶהם ְל ָׁשֹלום‬

ἠρώτησεν δὲ αὐτούς Πῶς ἔχετε;

Gen 43:27

Οἶδά σε παρὰ πάντας

Exod 33:12, 17

‫לב‬

στῆθος (not καρδία)

‫וְ ָׂש ַמח ְּב ִלֹּבו‬

διάνοια χαρήσεται ἐν ἑαυτῷ

Exod 28:23 (29), 28:26 (30) Exod 9:21 Exod 4:14

‫יְ ַד ְע ִּתיָך ְב ֵׁשם‬

‫ן־ׁשנָ ֹתו ( ַכ ְב ָׂשה‬ ְ ‫ֶּכ ֶבׂש ֶּב‬ )‫ת־ׁשנָ ָתּה‬ ְ ‫ַּב‬

ἄμνος (πρόβατος) ἐνιαύτος

Lev 12:6; 14:10; 23:12

19  I. Soisalon-Soininen, “Die Wiedergabe des ‫ ב‬instrumenti im griechischen Pentateuch”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987), 28–39.

293

Back to the Question of Hebraisms

‫ל־א ִחיו‬ ָ ‫אמרּו ִאיׁש ֶא‬ ְ ֹ ‫וַ ּי‬

καὶ εἶπαν ἕτερος τῷ ἑτέρῳ

Num 14:4

‫ ֶאת־‬ ‫וַ ּיָ ָׁשב יִ ְצ ָחק וַ ּיַ ְחּפֹר‬ ‫ְּב ֵאר ֹת ַה ַּמיִם‬

καὶ πάλιν Ισαακ ὤρυξεν τὰ φρέατα τοῦ ὕδατος

Gen 26:18

‫ׁשּובה ֶא ְר ֶעה צֹאנְ ָך‬ ָ ‫ָא‬ ‫ֶא ְׁשמֹר‬

πάλιν ποιμανῶ τὰ πρόβατά σου καὶ φυλάξω

Gen 30:31

3. Hebraisms Ignored in Recent Research Interestingly, recent years have seen attempts to treat literal and free (idiomatic) translation independently from the question of Hebraisms. James Barr reviewed the characteristics of literal translation from various perspectives.20 His observations outline literal translation in a well-considered and multi-faceted way. But for the study of individual translations and their mutual comparison, his characterization barely offers new perspectives—that was not his focus. Barr observes the following tendencies in the translations: [44] 1) The divisions 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.21

With these observations, Barr gives a very apt and multi-faceted sketch of literal and free translation. With respect to segmentation (1), I have pointed out peculiarities in the Septuagint text which illustrate that the translator usually rendered small segments at a time.22 This phenomenon is connected to the interesting question of word order (see below). 20 Barr, Typology of Literalism. 21 Barr, Typology of Literalism, 294. 22  I. Soisalon-Soininen, “Segmentation in the Greek Pentateuch” (in this volume). The manuscript was handed in before the publication of Barr’s article.

294

Ilmari Soisalon-Soininen

Omissions and additions (2) are treated very briefly by Barr. Apparently, he merely considered those additions and omissions that are semantically significant. However, additions and omissions are highly important features for translationtechnical reasons. The ancient translators themselves understood that too. It is such minuses that a considerable part of Origen’s asterized additions sought to redress.23 Let me give some examples: • redundant repetition of pronominal subject or object, of the genitive, of the possessive pronoun;24 • verbs that are normally construed with an accusative or a dative are now purposefully provided with a preposition, with a Hebraistic result. It was the ideal of Theodotion, Aquila and also Origen that even the smallest SL detail should be reflected in the target text. This quantitative precision is, naturally, a characteristic of slavish translation. It can be measured, but only if all categories25 are studied apart, and the results are considered together afterwards. In this case, the various categories are so similar that, taken together, they can be used to compare different [45] translators. Of course, the result should be checked through a detailed analysis. This type of research is already underway for the Pentateuch. Consistency (3) in the rendering of the same word or expression is a logical feature of slavish translation, but it is not without its complexities. It is immediately clear that the translators who consciously aimed at morphemic translations also strove for stereotyped renderings. Aquila was the champion of this tendency, although even he was not always consistent.26 The issue of consistency becomes more complex in the case of translators whose slavish or idiomatic approach was rather unconscious. A translator may be consistent in the use of a free rendering. Such is the case in LXX-Job, which is very consistent in the rendering of ‫ ב‬+ inf. cstr., but in such a manner that it uses a clause with ὅτε ten times and the literal ἐν τῷ + inf. only once. And conversely, a slavish translator may freely alternate synonyms.27 Simple and compound verbs may alternate, for example. The slavish translation reveals itself when the use of stock equivalents leads to a Hebraism, e.g., διδόναι εἰς βασιλέα. Such cases abound in the Septuagint. It is equally slavish when ‫ נתן‬in the sense of ‘allow’ is not rendered with the normal Greek words for this meaning, but always with the 23  I. Soisalon-Soininen, Der Charakter der asterisierten Zusätze in der Septuaginta (AASF, ser. B, 114; Helsinki: Suomalainen Tiedeakatemia, 1959), 46–106. 24  After certain words, such as body parts or family members, the personal pronoun is wholly redundant. 25  Note of the translator: according to Raija Sollamo (personal communication), SoisalonSoininen refers here to categories of phenomena that were being studied under his direction, such as repetition of possessive pronouns, partitive ‫מן‬, ‫ ב‬instrumenti, etc. 26  Cf. the example on p. [37]. 27  In Judges, e.g., παρατάσσεσθαι πολεμεῖν, μάχαιρα ῥομφαία etc.

Back to the Question of Hebraisms

295

also-possible διδόναι. Research in this area requires very detailed work in order to lay a solid foundation as a starting point. Barr offers several examples that lead the way in this direction. He also shows awareness of how important the conscious attempt at consistency is. This has the following implications. When studying the rendering of certain expressions and comparing translators, one should study not only consistency but also the use of both truly Greek and Hebraistic expressions. As I mentioned above,28 these perspectives do not always run parallel. A focus on mere consistency easily distorts the picture. The dimension idiomatic—slavish should be decisive. Barr’s three other features do not offer new avenues for a meticulous study of the material that could be useful for comparing translators. In sum, although Barr’s study starts from a different viewpoint, it helps establish some principles for detailed study of the character of translations. [46] The features of literal and free translations, as presented by Tov29 and Tov/Wright30 are basically those of Barr and need no separate treatment. But I do want to mention that, here too, a treatment of Hebraizing renderings is totally lacking. In the article by Tov/Wright, this omission leads to such a one-dimensional treatment of the material that the statistics become meaningless.31 With respect to ‫ → ב‬ἐν Tov/Wright observe: Probably one of the best criteria for investigating the literalness of translation units is the rendering of ‫ ב‬by ἐν creating numerous Hebraisms.32

This sentence ought to have led to a discussion of the extent to which different translators employ Hebraizing renderings. This question, however, is completely neglected, and all the different cases of ‫ ב‬are lumped together. That all translators employ renderings different from ἐν on numerous occasions should have led to the question in which cases ‫ → ב‬ἐν is avoided. If all cases, viz. local, temporal, instrumental, ‫ ב‬+ inf. cs. etc. are lumped together, the material becomes heterogeneous, and, hence, incommensurable. Therefore, a more appropriate question would be: How often do the translators use ἐν or other renderings in those cases where ἐν is not the most natural rendering? Such cases would amount to 40% in some books, and a mere 10% in others. If one overlooks the basic rules of

28  In the case of of ‫ ב‬+ inf. cstr. in LXX-Job. 29  E. Tov, The Text-Criticial Use of the Septuagint (Jerusalem: Simor, 1981), 54–60. 30  E. Tov/B.G. Wright, “Computer-Assisted Study of the Criteria for Assessing the Literalness of Translation Units in the LXX”, Textus 12 (1985) 152 n. 3. 31  Cf. the caveats expressed in I. Soisalon-Soininen, “Methodologische Fragen zur Erforschung der Septuaginta-Syntax,” in Studien zur Septuaginta-Syntax, 40–52. 32 Tov/Wright, “Computer-Assisted Study”, 159.

296

Ilmari Soisalon-Soininen

s­ tatistics and makes examples of the heterogeneous material invisible to the public, one cannot expect the intelligent reader to use the statistics profitably. Equally unreliable are the statistics regarding the renderings of ‫ ו‬and ‫כי‬. The caveats for the treatment of this material become clear upon consulting ­Aejmelaeus’ work.33 The other statistics adduced by Tov/Wright also need a close scrutiny. For example, those suffixes [47] which an idiomatic translation would leave untranslated, such as repetitions, should be treated separately. If it had been taken into account that slavish translation is related to the use of Hebraistic expressions, a more nuanced treatment of the material would have followed. With respect to the significance of word order for the study of translation technique Marquis writes:34 In the case of word order, the task is relatively simple: one compares the two texts, the presumed Hebrew source and the Greek translation, lists differences in word order between the two and expresses the resultant number relative to the total number of verses as a percentage of subservience or non-subservience to the word order of the source.35

To me, this position comes as a surprise, since I have always held word order in the Septuagint for the one of the most complicated and difficult questions. In some cases, e.g., postpositional conjunctions, the word order is changed without exception. One could object that slavish translations use fewer postpositional conjunctions than idiomatic ones. This means that more is involved than just word order. I would argue that the cases that do not permit a choice in word order should be omitted from the statistics. For the remaining phenomena, the possibilities of changing the word order varied from case to case, or rather, the drive to change word order varied. In the entire Septuagint, translators only spar33 A. Aejmelaeus, “OTI causale in Septuagintal Greek,” in N. Fernández Marcos (ed.), La Septuaginta en la investigación contemporánea. V congreso de la IOSCS (Textos y estudios “Cardenal Cisneros” 34; Madrid: Instituto Arias Montano, 1985) 115–32; idem, Parataxis in the Septuagint. With respect to the use of ὅτι, one should take special care—which Tov and Wright did not do—that it could already be used in a causal sense in classical Greek and especially in Koine, but that its frequent use in the LXX and NT exceeds the limits of its use in contemporary Greek. 34 Marquis, “Consistency of Lexical Equivalents,” 405. 35  It is strange that the number of word order deviations with respect to the number of verses is expressed in percentages. These items are incommensurable, but commensurability is a requirement of the calculation of percentages. In reality, Marquis wants to compare the number of verses without deviations to the total number of verses. In one verse there can be several deviations. Marquis speaks of “97.3% subservience,” which cannot be correct. For Ezekiel he indicates a percentage of 90.1, for Isaiah 66.4, a very significant difference. The percentages do not really work. It is much clearer to say that Ezek has 13.3 deviations per 10 pages BHS, and Isa 34.0, or to say that they have 9.9 and 33.6 deviations per 100 verses, respectively.

Back to the Question of Hebraisms

297

ingly resorted to this option. In nearly every sentence they could have introduced several word order changes, had they realized they could have produced better Greek that way. I investigated the word order changes in Judges and also those connected with the rendering of the Hebrew infinitive. Almost all cases I found are changes that could come about unintentionally. In his paper on word order Marquis classified the changes.36 Of the cases mentioned by him, [48] many words are so closely linked to other words37 that a change in word order—if natural in Greek—can take place unintentionally. In a number of cases Marquis does not place the instances in the right category.38 Further, subjects and objects that are expressed by pronouns or suffixes should be treated separately. But, most importantly, we would need to know to what extent the various changes occur in the individual books or translation units. Only equivalent material can be compared. Otherwise one cannot compare the different translators in a reliable way. Marquis would have been perfectly capable of finding a methodological example. In the area of word order, there is Wifstrand’s excellent study.39 He chose a phenomenon limited in scope, but one in which natural Greek word order commonly deviates from Hebrew. Wifstrand’s material is, in its entirety, homogeneous and commensurable. In some books, the material is somewhat scanty, but otherwise his method led to very reliable results. The nature of the examples which Marquis adduced shows that it would have been possible to categorize the material more adequately. It is inadmissible to investigate a language, with all its intricacies and peculiarities, as mechanically as Marquis did. Even more hazardous is his article on consistency.40 As said above, I doubt whether consistency is indicative of a slavish, and non-consistency of an idiomatic, translation. Both phenomena often go hand in hand though. When the ancient translators, notably Aquila, wanted to produce a translation as precise as possible, they aimed for systematic stereotyping. But when a translator did not, he could not always remember how he had rendered a certain word before. In spite of stereotyping tendencies, the renderings could vary considerably, especially with synonyms. That does not render the translation as a whole idiomatic though. What really makes a difference is if a translator selects [49] the counterparts in order to meet Greek stylistic requirements. Then he translates freely or 36 Marquis, “Word Order.” 37  E.g., possessive suffixes, numbers, and words like ‫עוד‬, ‫שם‬, ‫לא‬, ‫אין‬, ‫היה‬, ‫כל‬. 38  E.g., in Marquis, “Word Order,” 69 he claims that in Ezek 1:4 and 36:7 the position of ‫ סביב‬was changed, whereas in reality it was ‫ לו‬and ‫ לכם‬that changed positions: ‫→ וְ נֹגַ ּה ֹלו ָס ִביב‬ καὶ φέγγος κύκλῳ αὐτοῦ; ‫ → ֲא ֶׁשר ָל ֶכם ִמ ָּס ִביב‬τὰ περικύκλῳ ὑμῶν, where the rendering of ‫ ל‬with a genitive also played a role. 39 A. Wifstrand, Die Stellung der enklitischen Personalpronomina bei den Septuaginta (K. Humanistiska Vetenskapssamfundets i Lund Årsberättelse 1949–1950 II; Lund: Gleerup, 1950), 44–70. 40  See note on p. [37].

298

Ilmari Soisalon-Soininen

idiomatically. But there is something that often happens in the Septuagint, viz. that the use of stock equivalents gradually extends itself to cases where their use is no longer idiomatic. That is a true sign of slavish or literal translation. Here we touch again upon the problem of Hebraizing expressions. We need to know both languages very well to recognize unidiomatic use of a word. By way of example I refer to the rendering of ‫( נתן‬see p. [45]). Because διδόναι could be used in the sense of ‘allow,’ it became a temptation to render ‫‘ נתן‬allow’ as διδόναι at all times, and that is what some LXX translators did. Others, more idiomatically oriented, employ ἐφιέναι or ἐᾶν as well. Although according to our modern understanding almost all books of the Septuagint have been translated slavishly, the variety in renderings is considerable.41 These comprise synonyms and contextual renderings. But merely varying synonyms is not nearly as idiomatic as heeding TL requirements. Marquis correctly saw that different contexts gave rise to different renderings and that Hebraisms occurred if the context was neglected. But he believes the translators had a list with various counterparts listed for each word, if not written, then at least in their mind. He further postulates that when they stuck to the counterparts from their “list,” they translated consistently, and non-consistently when they deviated from it. Since we do not have these “lists,” we have to theorize about them. Marquis thinks that all renderings of a word that occur at least twice, must have been in the list. These are “consistent.” Of the unique renderings, a part is “consistent” too, while the other part is formed by either the free renderings or renderings of a deviating Vorlage. Marquis adduces an example, the rendering of ‫( עבר‬lumping together qal and hiphil), which actually shows more random variation than his theory allows.42 The number of unique renderings is determined by various factors, inter alia, the [50] size of the text, the frequency of the corresponding Hebrew word, the lexical density of the ST, etc. But even here Marquis wants to proceed purely mathematically. He starts with the rates of consistent and unique renderings, which he believes produce the degree of consistency. This he uses as a basis to calculate the degree of literalness, since consistency and literalness are not identical, in his view. A complicated mathematical formula is developed, where a = the number of renderings that occur twice or more; b = the number of unique renderings; a + b = the total number of renderings = 100. So, the calculation runs as follows: a ba + a + b (a + b)2 41  H.B. Swete, An Introduction to the Old Testament in Greek (Cambridge: Cambridge University Press, 1914), 328f., counts 32 different renderings for ‫נתן‬. 42 Marquis, “Consistency”, 414. Examples are διέρχεσθαι, διοδεύειν, διάγειν, διαπορεύεσθαι, which occur (for qal and hiphil of ‫ עבר‬together) 17 times.

Back to the Question of Hebraisms

299

which is the degree of literalness expressed in a percentage (the multiplier covers both parts of the binomial). My first objection is that it could be done much more easily, viz. by using simple percentages (a and b are percentages here): a+

ab 100

My second objection is more fundamental: what does all this really mean? The figures for Ezekiel β43 turn out as follows: 17.8% are unique renderings, 82.2% are “consistent.” Now Marquis takes 17.8% of 17.8%, resulting in 3.2%. And then he seriously claims that the “literalness” of Ezek β is 100% − 3.2% = 96.8%. This result does not modify the proportions that arise from a simple percentage of the unique renderings. It only gives a misleading impression of reliability. Marquis tries to substantiate his method, but we need not repeat and refute it again. It is immediately obvious that this oversimplified method is not applicable. It is true that the number of unique renderings may reflect the freedom of the translation, but it is highly dependent upon the overall translation technique. Marquis’ method would only make sense if applied to the comparison of different translators who had all translated the same text. What role word studies should play in the analysis of translation technique is not easy to say because of the complexity of the material. What is certain, though, is that such research cannot be conducted without painstaking labour, even if computers are used. It is essential to ask how (un)idiomatic the use of words and expressions is in various contexts.

4. A Way Ahead [51] Is it at all possible for us to study the differences between individual translators, given the fact that they translated very different texts and in view of the pivotal role of Hebraisms? It can only be done if we look for linguistic phenomena that are of a very similar type, i.e., that allow for idiomatic or slavish translation in different contexts to the same degree. A very good example is the position of enclitic personal pronouns. Another example of uniform material is temporal ‫ ב‬+ inf. cstr., a bit less so is ‫ ב‬instrumenti. These phenomena have as an advantage that the complete material can be surveyed to find distinguishing traits. These can then be divided into smaller categories. In the case of ‫ ל‬+ inf. cstr. as an 43  It is hard to understand why Marquis refers to the division by Thackeray, Grammar, 11 of Ezekiel in α = 1:1–27:36 and 40:1–48:35; β = 28:1–36:23 and 37:1–39:29; ββ = 36:24–28, without even mentioning Ziegler’s defense of the unity of LXX-Ezek (with the exception of the secondary passage 36:24–28).

300

Ilmari Soisalon-Soininen

object, two categories would be (a) the collocation with ‫כלה‬, e.g. ‫‘ ויכל לעשות‬he stopped doing,’ for which the natural rendering would be παύειν + participle; and (b) the collocation with ‫שוב‬, e.g. ‫‘ וישב לעשות‬he did again.’ The material cannot be completely uniform, but it can be so to a certain extent, enough to allow a r­ eliable investigation. Thus, comparing the renderings of different translators, one often finds significant differences. Some books clearly avoid slavish renderings of the phenomena under review, and why? Because a literal rendering ran against the translators’ feel for the language. We may then conclude that the translators considered those renderings Hebraistic. Whether we admit it or not, this makes the question which renderings are Hebraistic or idiomatic a key factor in our research. Not the researcher’s judgement, but the subject matter itself reveals the Hebraistic expressions, and in a way that is objective and independent of mere assumptions. It is true that even in this way we still cannot draw exact boundaries between Hebraisms and non-Hebraisms. But we can to a degree that enables us to take Hebraizing and idiomatic usage as a point of departure for comparing the work of different translators.

Bibliography Abbreviations AASF Annales Academiae Scientiarum Fennicae AJSL American Journal of Semitic Languages and Literatures BMECCJ Bulletin of the Middle Eastern Culture Center in Japan BHQ Biblia Hebraica Quinta BETL Bibliotheca Ephemeridum Theologicarum Lovaniensium BIOSCS Bulletin of the International Organization for Septuagint and Cognate Studies (Journal of Septuagint and Cognate Studies) ByzZ Byzantinische Zeitschrift BZAW Beihefte zur Zeitschrift für die alttestamentliche Wissenschaft‎ CBET Contributions to Biblical Exegesis & Theology DJD Discoveries in the Judaean Desert DSI De Septuaginta Investigationes FRLANT Forschungen zur Religion und Literatur des Alten und Neuen Testaments HSM Harvard Semitic Monographs HSS Harvard Semitic Studies JBL Journal of Biblical Literature JHS Journal of Hellenic Studies JNSL Journal of Northwest Semitic Languages JOAS Journal of the American Oriental Society JSJ Journal for the Study of Judaism JThS Journal of Theological Studies LCL Loeb Classical Library LHBOTS The Library of Hebrew Bible/Old Testament Studies LXX.H Handbuch zur Septuaginta MSU Mitteilungen des Septuaginta-Unternehmens NovT Novum Testamentum PFES Publications of the Finnish Exegetical Society RB Revue Biblique REA Revue des Etudes Anciennes REJ Revue des études juives SBLSCS Society of Biblical Literature Septuagint and Cognate Studies SJOT Scandinavian Journal of the Old Testament SNTSMS Society for New Testament Studies Monograph Series SSN Studia Semitica Neerlandica StBibLit Studies in Biblical Literature STDJ Studies on the Texts of the Desert of Judah

302 SVTG TCR TECC VT VTSup WO WUNT ZPE

Bibliography Septuaginta: Vetus Testamentum Graecum The Classical Review Textos y Estudios «Cardenal Cisneros» Vetus Testamentum Supplements to Vetus Testamentum Die Welt des Orients Wissenschaftliche Untersuchungen zum Neuen Testament Zeitschrift für Papyrologie und Epigraphik

Bibliography of Ilmari Soisalon-Soininen Soisalon-Soininen, I., “Heprean kielen ääntäminen”, Teologinen Aikakauskirja 52 (1947) 71–76, 115–127, 161–172. —, “Septuagintatutkimuksen suhde Vanhan testamentin tutkimukseen. Lectio praecursoria 22.5.1951”, Teologinen Aikakauskirja 56 (1951) 208–214. —, Die Textformen der Septuaginta-Übersetzung des Richterbuches (AASF, ser. B, 72,1; Helsinki: Suomalainen Tiedeakatemia, 1951). —, “Septuagintan synnyn ongelmia”, Teologinen Aikakauskirja 57 (1952) 151–160. —, “Septuagintan kielenkäyttö”, in Talenta quinque: Commentationes in hororem Ilmari Salomies, Eino Sormunen, E. G. Gulin, Rafael Gyllenberg, G. O. Rosenqvist (Helsinki: Otava, 1953) 46–52. —, Vanhan testamentin alkuteksti, (Suomen eksegeettisen seuran julkaisuja. Helsinki: Otava, 1953). —, “Wutzin teoria Septuagintan synnystä”, Teologinen Aikakauskirja 59 (1954) 180–195. —, “Uusi Septuaginta-löytö”, Teologinen Aikakauskirja 60 (1955) 83–94. —, “Profeetta Naahum runoilijana”, in P. Aalto (ed.) Kuunsirppi (Suomen itämaisen seuran kansantajuisia julkaisuja 12; Porvoo: WSOY, 1956). —, Der Charakter der asterisierten Zusätze in der Septuaginta (AASF, ser. B, 114; Helsinki: Suomalainen Tiedeakatemia, 1959) 46–106. —, “Kärsimyksen tiellä. Laskiaissunnuntai. Joh. 12: 25–33”, in M. Simojoki (ed.) Suomalainen saarnakirja II (Porvoo: WSOY, 1960) 78–82. —, “Pala vanhinta Raamattua. Löydöt ja Raamatun tekstin historia”, in E. Haapa (ed.) Qumran. Kuolleen meren löydöt 1950-luvun tutkimuksessa (Porvoo: WSOY, 1960) 22–54. —, “Palestiina pyhiinvaelluksen ja opintomatkan kohteena”, Teologinen Aikakauskirja 67 (1962) 293–301. —, “Heprean ekspegeettisen infinitiivin kääntämisestä suomalaisessa Raamatussa”, Teologinen aikakauskirja 68 (1963) 107–16. —, “Septuagintan lauseopin tutkimuksen perusteita”, Teologinen Aikakauskirja 68 (1963) 216–31. —, “Kolmanteen ja neljänteen polveen”, Teologinen Aikakauskirja 69 (vol. 3, 1964) 213–226.

Bibliography

303

—, “Palestiina Israelin heimojen maahantulon aikaan”, Teologinen Aikakauskirja 69 (vol. 3, 1964) 158–171. —, Aabrahamista Joosefiin: patriarkkatraditioiden historiaa (Helsinki: Kirjayhtymä, 1965). —, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965) —, “Kristillisen kirkon Vanha testamentti”, Teologinen Aikakauskirja 72 (1967) 183–189. —, “A teremtés gondolata Deuteroezsaiás igehirdetésében”, Lelkipásztor 42 (Budapest, 1967) 721–730. —, “Begreppet Funktion I Gammaltestamentlig Traditionsforskning”, Svensk Exegetisk Årsbok 33 (1968) 55–67. —, “Eettinen momentti Israelin varhaisajan uskonnossa”, Teologinen Aikakauskirja 73 (1968) 349–358. —, “Historia ja ilmoitus Vanhassa testamentissa”, Pöytäkirja Tampereen hiippakunnan lakimääräisessä pappeinkokouksessa Tampereella lokakuun 17–19 päivinä 1967 sekä liitteet (Tampere, 1968) 172–182. —, “Der Charakter Der Ältesten Alttestamentlichen Erzähltraditionen”, Temenos—Nordic Journal of Comparative Religion 4 (1969). —, “Das ethische Moment in der frühisraelitischen Religion”, Zeitschrift für evangelische Ethik 13 (1969) 146–153. —, Israelin kansan historia (Helsinki: Kirjayhtymä, 1969). —, “Exegetikens roll i det teologiska studiet och dess relation till övriga ämnen. Inledning vid den nordiska fakultetskonferensen i Uppsala 29.8.1969”, Teologinen Aikakauskirja 75 (1970) 232–238. —, “Die Urgeschichte Im Geschichtswerk Des Jahwisten”, Temenos—Nordic Journal of Comparative Religion 6 (1970). —, “Einige Merkmale der Übersetzungsweise von Aquila”, in J. Schreiner (ed.) Wort, Lied und Gottesspruch I. Beiträge zur Septuaginta. Festschrift für Joseph Ziegler (Forschung zur Bibel 1, Würzburg, 1972) 177–184. —, “Historia ja ilmoitus Vanhassa testamentissa”, in E. Koskenvesa (ed.) Johdatus teologisiin oppiaineisiin: Studia Theologica (Helsinki: Ylioppilastuki, 1972) 17–25. —, “Der Infinitivus constructus mit l im Hebräischen”, Vetus Testamentum 22 (1972) 82–90. —, “Septuagintan julkaisutyön periaatteita”, in K. Silvola/E. Wilhelms Sana, kirkko ja yhteys. Professori Aimo T. Nikolaisen juhlakirja (Helsinki, 1972) 37–47. —, “Syntax of translation-technique”, Bulletin of the International organization for Septuagint and cognate studies 5 (1972) 9–11. —, “Der Gebrauch des genetivus absolutus in der Septuaginta”, in The Fifth World Congress of Jewish Studies (vol. 4; Jerusalem: World Union of Jewish Studies, 1973) 131–6. —, “Septuaginta käännöksenä”, in Suomalainen tiedeakatemia. Esitelmät ja pöytäkirjat 1973 (Helsinki, 1975) 125–135. —, “Der Gebrauch des Verbs ἔχειν in der Septuaginta”, Vetus Testamentum 28 (1978) 92–99.

304

Bibliography

—, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch”, in Proceedings of the Sixth World Congress of Jewish Studies Held at the Hebrew University of Jerusalem, 13–19 August, 1973 (vol. 1; Jerusalem: World Union of Jewish Studies, 1977) 401–6. —, “Verschiedene Wiedergaben der hebräischen status constructus. Verbindung im griechischen Pentateuch”, Svensk Exegetisk Årsbok 41–42 (1976–1977) 214–223. —, “Die Wiedergabe einiger hebräischer Zeitangaben mit der Präposition ‫ ב‬in der Septuaginta”, in Annual of the Swedish Theological Institute XI (1978) 138–146. —, “Die Konstruktion des Verbs bei einem Neutrum Plural im griechischen Pentateuch”, Vetus Testamentum 29 (1979) 189–199. —, “Renderings of Hebrew Comparative Expressions with ‫ מן‬in the Greek Pentateuch”, Bulletin of the International Organization for Septuagint and Cognate Studies 12 (1979) 27–42. —, “ἐν für εἰς in der Septuaginta”, Vetus Testamentum, 32(2), (1982) 190–200. doi:10.2307/ 1518444 —, “Vanhan testamentin perikooppitekstin ongelmia”, Teologinen aikakauskirja 87 (vol. 4, 1982) 268–82. —, “Die Wiedergabe des ‫ ב‬instrumenti im griechischen Pentateuch”, in Glaube un Gerechtigkeit, In memoriam Rafael Gyllenberg (Schriften der Finnischen Exegetischen Gesellschaft 38, Helsinki, 1983) 31–46. —, “Eksiilin ajan merkitys Israelin teologiselle ajattelulle”, Teologinen aikakauskirja 89 (vol. 4, 1984) 260–64. —, “Iob”, Theologische Literaturzeitung 109 (vol. 2, 1984) 260–111. —, “Die Wiedergabe des hebräischen, als Subjekt stehenden Personalpronomens im griechi­ schen Pentateuch”, in C. Cox/A. Pietersma (ed.) De Septuaginta: Studies in honour of John ­William Wevers on his sixty-fifth birthday (Mississauga: Benben Publications, 1984) 115–128. —, “Die Auslassung des Possessivpronomens im griehischen Pentateuch”, Studia Orientalia 55 (1984) 279–294. —, “Jussi Aro: muistopuhe“, Vuosikirja 1983–84 (Helsinki: Suomalainen Tiedeakatemia, 1985) 95–98. —, “Die Wiedergabe des partitiven ‫ מן‬im griechischen Pentateuch”, in La Septuaginta en la investigación contemporánea (V Congreso de la IOSCS) (Textos y estudios “Cardenal Cisneros” 34; Madrid: Instituto “Arias Montano”, 1985) 83–100. —, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch”, in Proceedings of the Sixth World Congress of Jewish Studies I (Jerusalem: World Union of Jewish Studies, 1977) 401–406. —, “Methodologische Fragen der Erforschung der Septuaginta-Syntax”, in C.E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (SBLSCS 23; Atlanta, GA: Scholars Press, 1987) 425–44. —, “Zurück zur Hebraismenfrage”, in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien zur Septuaginta. Robert Hanhart zu Ehren (Abhandlungen der Akademie der Wissenschaften in Göttingen / MSU 20; Göttingen: Vandenhoeck & Ruprecht, 1990) 35–51. —, “Übersetzen—der Sprache Gewalt antun”, in L. Greenspoon/O. Munnich (ed.), VIII Congress of the IOSCS, Paris 1992 (SBLSCS 41; Atlanta, GA: SBL, 1995) 63–72.

Bibliography

305

In A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987): Soisalon-Soininen, I., “Einleitung”, 11–18 (originally pages 7–15 in Die Infinitive in der Septuaginta). —, “Einige Merkmale der Übersetzungsweise von Aquila”, 19–27. —, “Beobachtungen zur Arbeitsweise der Septuaginta-Übersetzer”, 28–39. —, “Methodologische Fragen der Erforschung der Septuaginta-Syntax”, 40–52. —, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch”, 55–61. —, “Verschiedene Wiedergaben der hebräischen status constructus: Verbindung im griechischen Pentateuch”, 62–70. —, “Die Wiedergabe des hebräischen, als Subjekt stehenden Personalpronomens im griechischen Pentateuch”, 71–85. —, “Die Auslassung des Possessivpronomens im griechischen Pentateuch”, 86–103. —, “Die Wiedergabe einiger hebräischer Zeitangaben mit der Präposition ‫ ב‬in der Septuaginta”, 107–115. —, “Die Wiedergabe des ‫ ב‬instrumenti im griechischen Pentateuch”, 116–130. —, “ἐν für εἰς in der Septuaginta”, 131–140. —, “Renderings of Hebrew Comparative Expressions with ‫ מן‬in the Greek Pentateuch”, 141– 153. —, “Die Wiedergabe des partitiven ‫ מן‬im griechischen Pentateuch”, 154–171. —, “Der Gebrauch des genetivus absolutus in der Septuaginta”, 175–180. —, “Der Gebrauch des Verbs ἔχειν in der Septuaginta”, 181−188. —, “Die Konstruktion des Verbs bei einem Neutrum Plural im griechischen Pentateuch”, 189–199. —, “Der infinitivus constructus mit ‫ ל‬im Hebräischen”, 203–211. In T. Kauhanen/H. Vanonen (ed.), The Legacy of Soisalon-Soininen: Towards a Syntax of Septuagint Greek (De Septuaginta Investigationes 13; Göttingen: Vandenhoeck & Ruprecht, 2020); transl. Theo van der Louw: Soisalon-Soininen, I., “The Use of the Genitive Absolute in the Septuagint” —, “A Palette of Renderings for the Hebrew Construct State in the Greek Pentateuch” —, “The Alleged Interchangeability of ἐν and εἰς in the Septuagint” —, “Segmentation in the Greek Pentateuch” —, “The Rendering of the Hebrew Personal Pronoun as Subject in the Greek Pentateuch” —, “Back to the Question of Hebraisms”

Literature Cited Aejmelaeus, A., Parataxis in the Septuagint: A Study of the Hebrew Coordinate Clauses in the Greek Pentateuch (AASF, Diss. Hum. Litt. 31; Helsinki: Suomalainen Tiedeakatemia, 1982).

306

Bibliography

—, “OTI causale in Septuagintal Greek,” in N. Fernández Marcos (ed.), La Septuaginta en la investigación contemporánea. V congreso de la IOSCS (TECC 34; Madrid: Instituto Arias Montano, 1985) 115–32. —, “The Significance of Clause Connectors in the Syntactical and Translation-Technical Study of the Septuagint,” in On the Trail of the Septuagint Translators: Collected Essays (Kampen: Kok Pharos, 1993) 43–58. —, “The Septuagint of 1 Samuel”, in L. Greenspoon/O. Munnich (ed.), VIII Congress of the International Organization for Septuagint and Cognate Studies, Paris 1992, (SBLSCS 41; Atlanta, GA: SBL, 1995) 109–29; repr. in A. Aejmelaeus, On the Trail of the Septuagint Translator: Collected Essays (Revised and Expanded Edition; CBET 50; Leuven: Peeters, 2007) 123–41. —, “What We Talk About When We talk About Translation Technique”, in B.A. Taylor (ed.), X Congress of the International Organization for Septuagint and Cognate Studies Oslo, 1998 (SBLSCS 51; Atlanta, GA: SBL, 2001) 531–52; repr. in A. Aejmelaeus, On the Trail of the Septuagint Translators: Collected Essays (CBET 50; Leuven: Peeters, 2007) 205–22. —, “Participium Coniunctum as a Criterion of Translation Technique”, in On the Trail of the Septuagint Translators: Collected Essays (CBET 50; Leuven: Peeters, 2007). —, “Corruption or Correction? Textual Development in the MT of 1 Samuel 1”, in P.A. Torijano Morales/A. Piquer Otero (ed.), Textual Criticism and Dead Sea Scrolls Studies in Honour of Julio Trebolle Barrera—Florilegium Complutense (Supplements to the Journal for the Study of Judaism 157; Leiden: Brill, 2012) 1–17. —, “Textual History of the Septuagint and the Principles of Critical Editing”, in A. Piquer Otero/P. Torijano Morales (ed.), The Text of the Hebrew Bible and Its Editions: Studies in Celebration of the Fifth Centennial of the Complutensian Polyglot (Supplements to the Textual History of the Bible 1; Leiden/Boston: Brill, 2016) 160–79. —, “Does God Regret? A Theological Problem that Concerned the Kaige Revisers”, in A. Aejmelaeus/T. Kauhanen (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (DSI 9; Göttingen: Vandenhoeck & Ruprecht, 2017) 41–53. —, “Kaige Readings in a Non-Kaige Section in 1 Samuel”, in A. Aejmelaeus/T. Kauhanen (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (DSI 9; Göttingen: Vandenhoeck & Ruprecht, 2017) 169–84. —, “Hexaplaric Recension and Hexaplaric Readings in 1 Samuel”, forthcoming. Aejmelaeus A./Kauhanen, T. (ed.), Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (DSI 9; Göttingen: Vandenhoeck & Ruprecht, 2017). Aejmelaeus A./Sollamo R. (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987). Aitken, J.K., No Stone Unturned: Greek Inscriptions and Septuagint Vocabulary (Winona Lake, IN: Eisenbrauns, 2014). Allan, R.J., The Middle Voice in Ancient Greek: A Study in Polysemy (Amsterdam: Gieben, 2003). D’Amore, L., “Ginnasio e difesa civica nelle poleis d’Asia minore (IV–I sec. A.C.)”, REA 109 (2007) 147–73. Armitage, P./Berry, P.J./Matthews, J.N.S., Statistical Methods in Medical Research (4th edition; Oxford: Blackwell Publishing, 2002).

Bibliography

307

Arnold, M., “Categorization of the Hitpa’ēl of Classical Hebrew” (PhD diss., Harvard University, 2005). Auld, A. G., Joshua: Jesus Son of Nauē in Codex Vaticanus (Septuagint Commentary Series 1; Leiden: Brill, 2005). Ausloos H./Lemmelijn, B., “Characterizing the LXX Translation of Judges on the Basis of Content-Related Criteria”, in H. Ausloos/B. Lemmelijn/J. Trebolle Barrera (ed.), After Qumran. Old and Modern Editions of the Biblical Texts—The Historical Books (BETL 246; Leuven/ Paris/Walpole, MA: Peeters, 2012) 171–92. Baden, J.T., “Hithpael and Niphal in Biblical Hebrew: Semantic and Morphological Overlap”, VT 60 (2010) 33–44. Bagnall, R.S./Derow, P., Greek Historical Documents: The Hellenistic Period (Chico: Scholars Press, 1981). Bakker, E.J., “Voice, Aspect, and Aktionsart: Middle and Passive in Ancient Greek”, in B. Fox/ P.J. Hopper (ed.), Voice: Form and Function (Amsterdam/Philadelphia: Benjamins, 1994) 23–47. Barber, E.J.W., “Voice—Beyond the Passive”, in Annual Meeting of the Berkeley Linguistics Society 1 (2014) 16–24. Barr, J., The Typology of Literalism in Ancient Biblical Translation (MSU 15; NAWG I, Phil-Hist. Kl.; Göttingen: Vandenhoeck & Ruprecht, 1979). —, “Translators’ Handling of Verb Tense in Semantically Ambiguous Contexts”, in C.E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (SBLSCS 23; Atlanta, GA: Scholars Press, 1987) 381–403. Barthélemy, D., “Redécouverte d’un chainon manquant de l’histoire de la Septante”, RB 60 (1953) 18–29. —, Les devanciers d’Aquila. Première publication intégrale du texte des fragments du Dodécaprophéton trouvés dans le désert de Juda, précédée d’une étude sur les traductions et recensions grecques de la Bible réalisées au premier siècle de notre ère sous l’influence du rabbinat palestinien (VTSup 10; Leiden: Brill, 1963). Bauer, H./Leander, P., Historische Grammatik der Hebräischen Sprache des Alten Testaments (2 vols.; Halle: M. Niemeyer, 1918). Baumgärtel, F., “Die Septuaginta zu Jesaja das Werk zweier Übersetzer”, in J. Herrmann/ F. Baumgärtel (ed.), Beiträge zur Entstehungsgeschichte der Septuaginta (Beiträge zur Wissenschaft vom Alten Testament, Neue Folge Heft 5; Leipzig: Kohlhammer, 1923) 20–31. Beck, J.A., Translators as Storytellers: A Study in Septuagint Translation Technique (StBibLit 25; New York: Peter Lang, 2000). Beckman, J.C., “Toward the Meaning of the Biblical Hebrew Piel Stem” (PhD diss., Harvard University, 2015). Beekes, R.S.P., Etymological Dictionary of Greek (Leiden: Brill, 2010). Ben-Hayyim, Z., “La tradition samaritaine et sa parenté avec les autres traditions de la langue hébraïque”, Mélanges de philosophie et de littérature juives 3–5 (1958–1962) 89–128.

308

Bibliography

Ben-Hayyim, Z./Tal, A., A Grammar of Samaritan Hebrew: Based on the Recitation of the Law in Comparison with the Tiberian and other Jewish Traditions (Winona Lake, IN: Eisenbrauns, 2000). Benton, R., “Aspect and the Biblical Hebrew Niphal and Hitpael” (PhD diss., University of Wisconsin—Madison, 2009). Benveniste, É., “Actif et Moyen dans le Verbe”, in Problèmes de Linguistique Générale I (Paris: Gallimard, 1966) 168–75. Bieżuńska-Małowist, I., L’esclavage dans l’egypte gréco-romaine. Première partie: Période ptolémaïque (Archiwum Filologiczne 30; Wrocław: Polska Akademia Nauk, Komitet Nauk o Kulturze Antycznej, 1974). Blass F./Debrunner, A./Funk R., A Greek Grammar of the New Testament and Other Early Christian Writings (Cambridge: Cambridge University Press, 1961). Blass, F./Debrunner, A./Rehkopf, F., Grammatik des neutestamentlichen Griechisch (Göttingen: Vandenhoeck & Ruprecht, 1976). Blomqvist, J./Jastrup, P.O., Grekisk/græsk grammatik (Odense: Akademisk Forlag, 1990). Bodine, W.R., The Greek Text of Judges: Recensional Developments (HSM 23; Ann Arbor MI: Edwards Brothers, 1980). Bons, E., “Ruth”, in S. Kreuzer (ed.), Einleitung in die Septuaginta (LXX.H 1; Gütersloh: Gütersloher Verlagshaus, 2016) 199–204. —, “Sur la traduction du vocabulaire de la servitude dans la Septante du livre de Ruth”, JSJ 33 (2002) 153–63. Bons E./Joosten J. (ed.), Die Sprache der Septuaginta/The Language of the Septuagint (LXX.H 3; Gütersloh: Gütersloher, 2016). Bons E./Kraus Th.J. (ed.), Et sapienter et eloquenter: Studies on Rhetorical and Stylistic Features of the Septuagint (FRLANT 241; Göttingen: Vandenhoeck & Ruprecht, 2011). Booij G./Marle J. van (ed.), Yearbook of Morphology 2003 (Dordrecht: Kluwer, 2003). Brown, J.P. (ed.), Israel and Hellas: Vol. 3, The Legacy of Iranian Imperialism and the Invividual? (BZAW 299; Berlin/New York: De Gruyter, 2001). Büchner, D., “Writing a Commentary on the Septuagint”, in M.K.H. Peters (ed.), XIV Congress of the International Organization for Septuagint and Cognate Studies, Helsinki 2010 (SBLSCS 59; Atlanta, GA: SBL, 2013) 525–38. Busto-Sáiz, J.R., La traducción de Símaco en el Libro de los Salmos (TECC; Madrid: Instituto Arias Montano, Consejo Superior de Investigaciones Científicas, 1985). Bybee, J., Language, Usage and Cognition (Cambridge: Cambridge Université Press, 2010). Caird, G.B., “Towards a Lexicon of the Septuagint. II”, JThS (1969) 21–40. Cañas Reíllo, J.M., “El papiro de Florencia, Bibl. Laur. PSI II 127 (Rahlfs 968): Su lugar en la historia textual del libro griego de Jueces y su relación con las versiones coptas”, in Τί ἡμῖν καὶ σοί; Lo que hay entre tú y nosotros. Estudios en honor de María Victoria Spottorno (Córdoba: UCOPress, 2016) 43–57. —, “LXX-Judges: The Value of Secondary Translations for Its Textual History”, in M. Meiser et al. (ed.), Die Septuaginta: Geschichte—Wirkung—Relevanz. 6. Internationale Fachtagung Wuppertal, 21.–24. Juli 2016 (WUNT 405; Tübingen : Mohr Siebeck, 2018) 230–43.

Bibliography

309

Cantarella, E., “Neaniskoi. Classi Di Età E Passaggi Di «Status» Nel Diritto Ateniese”, MERF— Antiquité 102 (1990) 37–51. Casey, E., “Educating the Youth: The Athenian Ephebeia in the Early Hellenistic Era”, in J.E. Grubbs/T. Parkin/R. Bell (ed.), The Oxford Handbook of Childhood and Education in the Classical World (Oxford: Oxford University Press) 418–43. Chankowski, A.S., L’éphébie hellénistique. Étude d’une institution civique dans les cités grecques des îles de la Mer Égée et de l’Asie Mineure (Culture & Cité 4; Paris: De Boccard, 2011). —, “Age-Class”, in R.S. Bagnall et al. (ed.), The Encyclopedia of Ancient History (Oxford: Blackwell, 2013) 178–80. Chantraine, P., Dictionnaire étymologique de la langue grecque. Histoire des mots (Librairie Klincksieck—Série linguistique; Paris: Klincksieck, 2009). Claassen, W.Th., “The Hiph‘il Verbal Theme in Biblical Hebrew” (PhD diss., University of Stellenbosch, 1971). Clines, D.J.A., (ed.), The Dictionary of Classical Hebrew (Sheffield: Sheffield Phoenix Press 1996). Creason, S.A., “Semantic Classes of Hebrew Verbs: A Study of Aktionsart in the Hebrew Verbal System” (PhD diss., University of Chicago, 1995). Cunliffe R.J./Okla N., A Lexicon of the Homeric Dialect (Norman: University of Oklahoma Press, 1963). Dakyns, H.G. (ed.), Xenophon, The Education of Cyrus (London: J.M. Dent & Sons Ltd, 1992). Daniel, R.W., “The Military [Cημεϊον] in P.Amh. Ii 39”, ZPE 52 (1983) 269–71. Deissmann A./Grieve A.J., 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 (Edinburgh: T&T Clark, 1901). Depauw, M./Gheldof, T., “Trismegistos: An Interdisciplinary Platform for Ancient World Texts and Related Information”, in Ł. Bolikowski/V. Casarosa/P. Goodale/N. Houssos/P. Manghi/ J. Schirrwagen (ed.), Theory and Practice of Digital Libraries—Tpdl 2013 Selected Workshops (Communications in Computer and Information Science 416; Charm: Springer, 2014) 40–52. Diodorus Siculus, The Library of History (London: Heinemann/Cambridge, MA: Harvard University Press, 1933–67). Dorival, G./Harl, M./Munnich O., La Bible greque des Septante. Du judaïsme hellénistique au christianisme ancien (Initiations au christianisme ancien 3; Paris: Éditions du Cerf, 1988). Downie, J., “Intervention Length and Interpreter Performance in Short Intervention Consecutive Interpreting”, 2010, online at https://www.academia.edu/4504695/Intervention_ Length_and_Interpreter_Performance_in_Short_Intervention_Consecutive_Interpreting. Edgar, C.C. (ed.), Catalogue général des antiquités égyptiennes du Musée du Caire, nos. 59298– 59531. Zenon Papyri III (Cairo: L’Institut français d’archéologie orientale, 1928). Endo, Y., The Verbal System of Classical Hebrew in the Joseph Story: An Approach from Discourse Analysis (SSN 32; Assen: Van Gorcum, 1996). Evans, T.V., Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference (Oxford: Oxford University Press, 2001).

310

Bibliography

Everett-Heath, J. (ed.), The Concise Dictionary of World Place-Names (Oxford: Oxford University Press). Eynikel E./Hauspie K./Lust J., Greek-English Lexicon of the Septuagint (Stuttgart: Deutsche Bibelgesellschaft, 2003). Fakour, M., “Garden i. Achaemenid Period”, in E. Yarshater (ed.), Encyclopaedia Iranica (New York: Bibliotheca Persica Press, 2001) 10:297–8. Fassberg, S.E., “The Movement from Qal to Pi“el in Hebrew and the Disappearance of the Qal Internal Passive”, HS 42 (2001) 243–5. Fedorov, A.V., Введение в теорию перевода (An Introduction to Translation Theory) (Moscow: Изд-во лит-ры на иностранных языках, 1953). Fernández Marcos, N., Judges (BHQ 7; Stuttgart: Deutsche Bibelgesellschaft, 2011). —, “Kritai/Iudices/Das Buch der Richter”, in S. Kreuzer (ed.), Einleitung in die Septuaginta (LXX.H 1; Gütersloh: Gütersloher Verlagshaus, 2016) 188–98. —, “The Antiochene Edition in the Text History of the Greek Bible”, in S. Kreuzer/M. Sigismund (ed.), Der Antiochenische Text der Septuaginta in seiner Bezeugung und seiner Bedeutung (DSI 4; Göttingen: Vandenhoeck & Ruprecht, 2013) 57–73. —, “The B-Text of Judges: Kaige Revision and Beyond”, in H. Ausloos/B. Lemmelijn/J. Trebolle Barrera (ed.), After Qumran. Old and Modern Editions of the Biblical Texts—The Historical Books (BETL 246; Leuven/Paris/Walpole, MA: Peeters, 2012) 161–9. —, “Theodoret’s Biblical Text in the Octateuch”, BIOSCS 11 (1978) 27–43. Fernández Marcos, N./Busto Sáiz, J.R., El texto antioqueno de la Biblia Griega. I. 1–2 Samuel (TECC 50; Madrid: CSIC, 1989). —, El texto antioqueno de la Biblia Griega. II. 1–2 Reyes (TECC 53; Madrid: CSIC, 1992). —, El texto antioqueno de la Biblia Griega. III. 1–2 Crónicas (TECC 60; Madrid: CSIC, 1996). Fernández Marcos, N./Sáenz-Badillos, A., Theodoreti Cyrensis Quaestiones in Octateuchum. Editio critica (TECC 17; Madrid: CSIC, 1979). Fernández Marcos, N./Spottorno Díaz-Caro, M.V./Cañas Reíllo, J.M., Índice griego-hebreo del texto antioqueno en los libros históricos. (2 vols; TECC 75; Madrid: CSIC, 2005). Forbes, C.A., Neoi: A Contribution to the Study of Greek Associations (Middletown: The American Philological Association, 1933). Frankel, Z., Vorstudien zu der Septuaginta (Leipzig: F.C. Vogel, 1841). Gardiner, E.N., “A School in Ptolemaic Egypt”, TCR 44 (1930) 211–13. Garland, R., The Greek Way of Life: From Conception to Old Age (London: Duckworth, 1990). Gehman, H.J., “Hebraisms of the Old Greek Version of Genesis”, VT 3 (1953) 141–8. Gentry, P.J., The Asterisked Materials in the Greek Job (SBLSCS 38; Atlanta, GA: Scholar’s Press, 1995). Gesenius W., Gesenius’ Hebrew Grammar (ed. E. Kautzsch; transl. A.E. Cowley; 2nd edition; Oxford: Clarendon Press, 1982; repr. 1990). Gil, J. (ed.), Jenofonte, Económico (Madrid: Sociedad de Estudios y Publicaciones, 1967). Gile, D., “Conference Interpreting as a Cognitive Management Problem”, in F. Pöchhacker/ M. Shlesinger (ed.), The Interpreting Studies Reader (London/New York; Routledge, 2002) 163–76.

Bibliography

311

Goetze, A., “The So-Called Intensive of the Semitic Languages”, JOAS 62 (1942) 1–8. Golden, M., Childhood in Classical Athens (Baltimore: Johns Hopkins University Press, 1990). Grammaticus P./Bethe E. (ed.), Onomasticon (Leipzig: Teubner, 1931). Grenfell, B.P./Hunt A.S. (ed.), The Oxyrhynchus Papyri (London: Egypt Exploration Fund, 1898–1924). Grethlein, J., “Xenophon’s Anabasis from Character to Narrator”, JHS 132 (2012) 23–40. Grice, H. P., “Logic and Conversation”, in P. Cole/J. Morgan (ed.), Speech acts (New York: Academic Press, 1975) 41–58. Grice, P., Studies in the Way of Words (Cambridge: Harvard University Press, 1989). Gryson, R., Altlateinische Handschriften / Manuscripts Latins (Vetus Latina: Die Reste der altlateinischen Bibel 1/2A, Freiburg: Herder, 1999). Gutt, E.-A., Translation and Relevance: Cognition and Context (2nd edition; Manchester: St. Jerome, 2000). Hatch, E./Redpath, H.A., A Concordance to the Septuagint and the Other Versions of the Old Testament (Including the Apocryphal Books) (2nd augmented edition; Grands Rapids: Baker Books, 1998). Hansen P.A./Cunningham, I.C. (ed.), Hesychii Alexandrini Lexicon: Editionem post Kurt Latte continuans recensuit et emendavit (Berlin: De Gruyter, 2005). Harl, M., “La Bible d’Alexandrie”, in B.A. Taylor (ed.), X Congress of the International Organization for Septuagint and Cognate Studies: Oslo, 1998 (SBLSCS 51; Atlanta, GA: Scholars Press, 2001) 181–97. Hauspie, K., “Ezekiel”, in J.K. Aitken (ed.), T&T Clark Companion to the Septuagint (London: Bloomsbury T&T Clark, 2015) 528–43. Helbing, R., Grammatik der Septuaginta. Laut- und Wortlehre (Göttingen: Vandenhoeck & Ruprecht, 1907). —, Die Kasussyntax der Verba bei den Septuaginta. Ein Beitrag zur Hebraismenfrage und zur Syntax der Koine (Göttingen: Vandenhoeck & Ruprecht, 1927). Herrmann, J./Baumgärtel, F., (ed.), Beiträge zur Entstehungsgeschichte der Septuaginta (Beiträge zur Wissenschaft vom Alten Testament, Neue Folge Heft 5; Leipzig: Kohlhammer, 1923). —, “Die Septuaginta zum Zwölfprophetenbuch das Werk zweier Übersetzer”, in J. Herrmann/F. Baumgärtel (ed.), Beiträge zur Entstehungsgeschichte der Septuaginta (Beiträge zur Wissenschaft vom Alten Testament, Neue Folge Heft 5; Leipzig: Kohlhammer, 1923) 32–38. Hertog, C.G. den, “Studien zur griechischen Übersetzung des Buches Josua” (PhD diss., JustusLiebig Universität, 1996). Himbaza, I., “What are the Consequences if 4QLXXLeva Contains the Earliest Formulation of the Septuagint?”, in S. Kreuzer/M. Meiser/M. Sigismund (ed.), Die Septuaginta—Orte und Intentionen (WUNT 361; Tübingen: Mohr Siebeck, 2016) 294–308. Hopkins, K., Conquerors and Slaves (Sociological Studies in Roman History 1; Cambridge: Cambridge University Press, 1978). Horrocks, G.C., Greek: A History of the Language and its Speakers (2nd edition; Chichester: Wiley Blackwell, 2014). Humbert, J., Syntaxe grecque (3rd edition; Paris: Klincksieck, 1960).

312

Bibliography

Hunt, A.S./Edgard, C.C (transl.), Select Papyri, Volume II: Public Documents (LCL 282; Cambridge: Harvard University Press, 1934). Jakobsen, A.L., “Effects of Thinking Aloud on Translation Speed, Revision and Segmentation”, in F. Alves (ed.), Triangulating Translation (Amsterdam: John Benjamins, 2003) 69–95. Jannaris, A.N., An Historical Greek Grammar, Chiefly of the Attic Dialect (London: Macmillan, 1897). Janzen, J.G., Studies in the Text of Jeremiah (Cambridge, MA: Harvard University Press 1973). Jenni, E., Das hebräische Piʿel. Syntaktische Untersuchung einer Verbalform im Alten Testament (Zürich: EVZ, 1968). Jobes K.H./Silva M., Invitation to the Septuagint (Grand Rapids, MI: Baker Academic, 2015). Johannessohn, M., Der Gebrauch der Präpositionen in der Septuaginta (MSU 3,3; Berlin: Weidmann, 1926). Johnson, B., Die Hexaplarische Rezension des 1. Samuelbuches der Septuaginta (Studia Theologica Lundensia 22; Lund: Gleerup, 1963). Joosten, J., “The Functions of the Semitic D Stem: Biblical Hebrew Materials for a ComparativeHistorical Approach”, Orientalia 67 (1998) 202–30. —, “The Prayer of Azariah (DanLXX 3): Sources and Origin”, in J. Cook (ed.), Septuagint and Reception (VTSup 127; Leiden: Brill, 2009) 5–16. —, “The Vocabulary of the Septuagint and Its Historical Context”, in J. Joosten/E. Bons (ed.), Septuagint Vocabulary: Pre-History, Usage, Reception (Atlanta, GA: SBL, 2011) 1–11. —, The Verbal System of Biblical Hebrew: A New Synthesis Elaborated on the Basis of Classical Prose (Jerusalem: Simor, 2012). Karlik, J., “Interpreter-Mediated Scripture: Expectation and Performance”, Interpreting 12 (2010) 160–85. Karrer, M., “The New Leaves of Sinaiticus Judges”, in S. Kreuzer/M. Meiser/ M. Sigismund (ed.), Die Septuaginta—Entstehung, Sprache, Geschichte. 3. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 22.–25. Juli 2010 (WUNT 286; Tübingen: Mohr Siebeck, 2012) 600–17. Karrer, M./Kraus W. (ed.), Septuaginta Deutsch. Erläuterungen und Kommentare; Band 1, Genesis bis Maccabäer (Stuttgart: Deutsche Bibelgesellschaft, 2011). Katz, P., “Zur Übersetzungstechnik der Septuaginta”, WO 2 (1956) 267–73. Kaufman, S.A., “Semitics: Directions and Re-directions”, in J.S. Cooper/G.M. Schwartz (ed.), The Study of the Ancient Near East in the Twenty-First Century (Winona Lake, IN: Eisenbrauns, 1996) 273–82. Kemmer, S., The Middle Voice (Amsterdam/Philadelphia: Benjamins, 1993). —, “Middle Voice, Transitivity, and the Elaboration of Events”, in B. Fox/P.J. Hopper (ed.), Voice: Form and Function (Amsterdam/Philadelphia: Benjamins, 1994) 179–229. Kennell, N.M., “The Ephebeia in the Hellenistic Period”, in W.M. Bloomer (ed.), A Companion to Ancient Education (Oxford: Wiley Blackwell, 2015) 172–83. —, “Who Were the Neoi?”, in P. Martzavou and N. Papazarkadas (ed.), Epigraphical Approaches to the Post-Classical Polis: Fourth Century BC to Second Century AD (Oxford Studies in Ancient Documents; Oxford: Oxford University Press) 217–32.

Bibliography

313

—, Ephebeia: A Register of Greek Cities with Citizen Training Systems in the Hellenistic and Roman Periods (Nikephoros-Beihefte 12; Hildesheim: Weidmann, 2006). —, “New Light on 2 Maccabees 4:7–15”, JSJ 56 (2005) 10–24. Klaiman, M.H., “Affectedness and Control: A Typology of Voice Systems”, in M. Shibatani (ed.), Passive and Voice (Amsterdam/Philadelphia: Benjamins, 1988) 25–83. Klein, G.L., “The meaning of the Niphal in Biblical Hebrew” (PhD diss., Annenberg Research Institute, 1992). Klinken, W. van, “From Literal to Free? A Study of Development in the Genesis Translation of the Septuagint” (M.A. thesis, Free University of Amsterdam, 2006). Kloppenborg, J.S., The Tenants in the Vineyard: Ideology, Economics, and Agrarian Conflict in Jewish Palestine (WUNT 195; Tübingen: Mohr Siebeck, 2006). Köhler L./Baumgartner W., The Hebrew and Aramaic Lexicon of the Old Testament (Leiden: Brill, 1994–2000). Kouwenberg, J.C., Gemination in the Akkadian Verb (SSN 32; Assen: Van Gorcum, 1997). Kraemer, R.S., Women’s Religions in the Greco-Roman World (Oxford, Oxford University Press, 2004). Krentz, P. (ed.), Xenophon, Hellenika (Warminster: Aris & Phillips, 1995). Kreuzer, S., “Der Antiochenische Text der Septuaginta. Forschungsgeschichte und eine neue Perspektive”, in S. Kreuzer/M. Sigismund (ed.), Der Antiochenische Text der Septuaginta in seiner Bezeugung und seiner Bedeutung (DSI 4; Göttingen: Vandenhoeck & Ruprecht, 2013) 23–56. —, “Papyrus 967: Its Significance for Codex Formation, Textual History, and Canon History,” in The Bible in Greek: Translation, Transmission, and Theology of the Septuagint (SBLSCS 63; Atlanta, GA: SBL Press, 2015) 255–71. Kriston, A., “The Importance of Memory Training in Interpretation”, Professional Communication and Translation Studies 5,1–2 (2012) 79–86. Kühner R./Gerth B., Ausführliche Grammatik der griechischen Sprache: Teil 2, Satzlehre (3rd edition; Hannover und Leipzig: Hahn, 1898; 4th edition; Leverkusen: Gottschalksche Verlagsbuchhandlung, 1955). Kutscher, E.Y., The Language and Linguistic Background of the Isaiah Scroll (1 QIsaa) (STDJ 6; Leiden: Brill, 1974). LaMontagne, N., “The Song of Deborah (Judges 5): Meaning and Poetry in the Septuagint” (PhD diss., The Catholic University of America, 2013). Langacker, R.W., Foundations of Cognitive Grammar, Vol. 1 (Stanford, CA: Stanford University Press, 1987). —, Foundations of Cognitive Grammar, Vol. 2: Descriptive Application (Stanford, CA: Stanford University Press, 1991). Launey, M., Recherches sur les armées hellénistiques. Volume 2 (Bibliothèque des Écoles Françaises d’Athènes et de Rome 169; Paris: E. de Boccard, 1950). Le Boulluec, A./Le Moigne, P., Vision que vit Isaïe. Traduction du texte du prophète Isaïe selon la Septante. Index littéraire des noms propres et glossaire (Paris: Cerf, 2014).

314

Bibliography

Lee, J.A.L., A Lexical Study of the Septuagint Version of the Pentateuch (SBLSCS 14; Chico, CA: Scholars Press, 1983). —, “Ἐξαποστέλλω”, in J. Joosten/P.J. Tomson (ed.), Voces Biblicae: Septuagint Greek and its Significance for the New Testament (CBET 49; Leuven: Peeters, 2007) 99–113. —, “The Onoma Rule” NovT 56 (2014) 411–21. —, The Greek of the Pentateuch: Grinfield Lectures on the Septuagint, 2011–2012 (Oxford: Oxford University Press, 2018). Le Moigne, P., “Au delà du reel. Recherches sur le subjonctif éventuel de l’Octateuque” (habilitation, Université de Strasbourg, 2014). Léonard, J.M., “Enquête sur l’émergence du Hifil en hébreu biblique” (PhD diss., Montpellier, 1990). Lesquier, J., Les institutions militaires de l’Egypte sous les Lagides (Paris: E. Leroux, 1911). Lewis, N., Greeks in Ptolemaic Egypt (Oakville, CT: American Society of Papyrologists, 1986). Lewis, S., The Athenian Woman: An Iconographic Handbook (London: Routledge, 2002). Liddell, H.G./Scott, R./Jones, H.S., A Greek-English Lexicon (Oxford: Oxford University Press, 1996). Louw, T.A.W. van der, “Analysis of the First Interpreting Experiment”, 2007, online at https:// www.academia.edu/3080833/Analysis_of_the_First_Interpreting_Experiment_held_ June_2007. —, Transformations in the Septuagint: Towards an Interaction of Septuagint Studies and Translation Studies (CBET 47; Leuven: Peeters, 2007). —, “The Dictation of the Septuagint Version”, JSJ 39 (2008) 211–29. —, “Did the Septuagint Translators Really Intend the Greek Text as It Is?”, in S. Kreuzer/ M. Meiser/M. Sigismund (ed.), Die Septuaginta. Orte—Intentionen; 5. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal, 24–27. Juli 2014 (WUNT 361; Tübingen: Mohr Siebeck, 2016) 449–66. —, “The Evolution of the Genesis Translator”, in M. Meiser et al. (ed.), Die Septuaginta. Geschichte—Wirkung—Relevanz; 6. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal, 21–24. Juli 2016 (WUNT 405; Tübingen: Mohr Siebeck, 2018) 146–57. Lust, J./Eynikel, E./Hauspie, K., A Greek-English Lexicon of the Septuagint (Stuttgart: Deutsche Bibelgesellschaft, 1992 [part 1], 1996 [part 2]). Maldonado, R., “Middle as a Basic Voice System”, in E. Guerrero/L. Sergio Ibáñez/V. Belloro (ed.), Studies in Role and Reference Grammar (México: Instituto de Investigaciones Filológicas, UNAM, 2009) 69–112. Margolis, M.L., “The Greek Preverb and Its Hebrew-Aramaic Equivalent”, AJSL 26 (1909) 33–61. Marquis, G., “Word Order as a Criterion for the Evaluation of Translation Technique in the LXX and the Evaluation of Word-Order Variants as Exemplified in LXX-Ezekiel”, Textus 13 (1986) 59–84. —, “Consistency of Lexical Equivalents as a Criterion for the Evaluation of Translation Technique as Exemplified in the LXX of Ezekiel”, in C.E. Cox (ed.), VI Congress of the Interna-

Bibliography

315

tional Organization of Septuagint and Cognate Studies, Jerusalem 1986 (SBL SCS 23; Atlanta, GA: SBL, 1987). Mauersberger, A./Helms, H. (ed.), Polybios-Lexikon: Volume 1.4 [λ-ο] (2nd edition; Berlin: Akademie Verlag, 2006). Mayser, E., Grammatik der griechischen Papyri aus der Ptolemäerzeit (2 vols. in 6 parts; Berlin/ Leipzig: De Gruyter, 1906–38). McFall, L., The Enigma of the Hebrew Verbal System: Solutions from Ewald to the Present Day (Historic Texts and Interpreters in Biblical Scholarship 2; Sheffield: Almond Press, 1982). McGregor, G.S., “A Semantic Study of Words for Young Person, Servant and Child in the Septuagint and Other Early Koine Greek” (M.A. thesis, University of Sydney, 1976). McLay, T., “Kaige and Septuagint Research”, Textus 19 (1998) 127–39. —, The Use of the Septuagint in New Testament Research (Grand Rapids, MI: Eerdmans, 2003). McTait K./Olohan M./Trujillo A., “A Building Blocks Approach to Translation Memory”, in Translation and the Computer 21: Proceedings of the 21st International Conference on Translating and the Computer (London: Aslib, 1999). Mey, J.L., Pragmatics: An Introduction (Oxford: Blackwell, 2001). Min, Y.-J., “The Minuses and Pluses of the LXX Translation of Jeremiah as Compared with the Massoretic Text: Their Classification and Possible Origins” (PhD diss., Hebrew University of Jerusalem, 1977). Montanari, F., The Brill Dictionary of Ancient Greek (Leiden/Boston: Brill, 2015). Montevecchi, O., Bibbia e papiri. Luce dai papiri sulla bibbia greca (Barcelona: Institut de Teologia Fondamental, Seminari de Papirologia, 1999). Moulton, J.H./Milligan, G., The Vocabulary of the Greek Testament Illustrated from the Papyri and Other Non-Literary Sources (London: Hodder and Stoughton, 1929). Moulton J.H./Turner, N., A Grammar of New Testament Greek: Vol. III, Syntax (Edinburgh: T&T Clark, 1963). Muraoka, T., A Greek-English Lexicon of the Septuagint (Louvain: Peeters, 2009). —, A Syntax of Septuagint Greek (Leuven: Peeters, 2016). Oepke, A., “εἰς”, in G. Kittel (ed.), Theological Dictionary to the New Testament: Volume 2, δ–η (trans. W. Bromley; Grand Rapids: Eerdmans, 1964) 420–33. —, “ἐν”, in G. Kittel (ed.), Theological Dictionary to the New Testament: Volume 2, δ–η (trans. W. Bromley; Grand Rapids: Eerdmans, 1964) 537–42. Offizielles Verzeichnis der Rahlfs-Siegeln. Herausgegeben vom Septuaginta-Unternehmen der Akademie der Wissenschaften zu Göttingen (Göttingen 2012), online at http://hdl.handle. net/11858/00-001S-0000-0022-A30C-8. Osty, E./Trinquet, J., La Bible. Traduction française sur les textes originaux par Émile Osty avec la collaboration de Joseph Trinquet. Introductions et notes d’Émile Osty et de Joseph Trinquet (S. l.:Seuil, 1973). Panaino, A., “No Room for the ‘Paradise’? About Old Persian paradayadama”, in G.P. Basello/ A.V. Rossi (ed.), Persepolis and Its Settlements: Territorial System and Ideology in the Achaemenid State (Napoli: Universita degli studi di Napoli “L’Orientale”, 2012) 139–52.

316

Bibliography

Parkin, T.G., Old Age in the Roman World: A Cultural and Social History (Baltimore: Johns Hopkins University Press, 2003). Penner, K.M., The Verbal System of the Dead Sea Scrolls: Tense, Aspect, and Modality in Qumran Hebrew Texts (SSN 64; Leiden: Brill, 2015). Pestman, P.W et al., A Guide to the Zenon Archive (P. L. Bat. 21): Lists and Surveys (Papyrologica Lugduno-Batava 21; Leiden: Brill, 1981). Peterson, J.H., Dictionary of Most Common Avesta Words (1995), online at http://www.avesta. org/avdict/avdict.htm#dctp. Peursen, W.Th. van, The Verbal System in the Hebrew Text of Ben Sira (Studies in Semitic Languages and Linguistics 41; Leiden: Brill, 2004). Pietersma, A., “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… July 2000 (Leiden/Boston: Brill, 2002) 337–64. —, “An Excursus on Bisectioning Ieremias”, in A. Pietersma/B.G. Wright (ed.), Electronic Edition of The New English Translation of the Septuagint (Oxford: Oxford University Press, 2009) 7–9. —, “Of Translation and Revision: From Greek Isaiah to Greek Jeremiah”, in M. van der Meer et al. (ed.), Isaiah in Context: Studies in Honour of Arie van der Kooij on the Occasion of his Sixty-Fifth Birthday (VTSup 138; Leiden: Brill 2010) 359–88. Pietersma, A./Saunders, M., “Ieremias: To the Reader”, in A. Pietersma/B. G. Wright (ed.), Electronic Edition of The New English Translation of the Septuagint (Oxford: Oxford University Press, 2009). Pöchhacker, F., Introducing Interpreting Studies (London/New York: Routledge, 2004). —, “Consecutive Interpreting”, in K. Malmkjaer/K. Windle (ed.), The Oxford Handbook of Translation Studies (Oxford Handbooks in Linguistics; Oxford: Oxford University Press, 2011) 295–6. Preisigke, F., Wörterbuch der griechischen Papyrusurkunden mit Einschluss der griechischen Inschriften, Aufschriften, Ostraka, Mumienschildern usw. aus Ägypten. Band 1, A–K (Berlin: Selbstverlag der Erben, 1925). Pretzl, O., “Septuagintaprobleme im Buch der Richter”, Biblica 1 (1926) 233–69. —, “Der hexaplarische und tetraplarische Septuagintatext des Origenes in den Büchern Josue und Richter”, ByzZ 30 (1929) 262–68. Psichari, J., “Essai sur le grec de la Septante”, REJ 55 (1908) 161–208. Qimron, E., The Hebrew of the Dead Sea Scrolls (HSS 29; Atlanta, GA: Scholars Press, 1985). Quast, U. (ed.), Ruth (SVTG 4.3; Vandenhoeck & Ruprecht, 2006). Rahlfs, A., Verzeichnis der griechischen Handschriften des Alten Testaments. Für das Septuaginta-Unternehmen aufgestellt von Alfred Rahlfs (MSU 2; Berlin: Weidmannsche Buchhandlung 1914). —, Das Buch Ruth griechisch, als Probe einer kritischen Handausgabe der LXX (Stuttgart: Privileg. Württ Bibselanstalt, 1922). —, Studie über den griechischen Text des Buches Ruth (Berlin: Buchhandlung, 1922).

Bibliography

317

—, Septuaginta-Studien I. Studien zu den Königsbüchern (2nd ed.; Göttingen: Vandenhoeck & Ruprecht, 1965). —, Verzeichnis der griechischen Handschriften des Alten Testaments. Band I, Die Überlieferung bis zum VIII. Jahrhundert (bearbeitet von D. Fraenkel; Göttingen: Vandenhoeck & Ruprecht 2004). —, Septuaginta. Id est Vetus Testamentum graece iuxta LXX interpretes. Editio altera quam recognovit et emendavit Robert Hanhart. Duo volumina in uno (Stuttgart: Deutsche Bibelgesellschaft, 2006). Reekmans, T., La sitométrie dans les archives de Zénon (Pap. Brux. 3; Brussels: Fondation épygtologique Reine Elisabeth, 1966). Reider, J., Prolegomena to a Greek-Hebrew & Hebrew-Greek Index to Aquila (Oxford: University Press, 1916). Rife, J.M., “The Mechanics of Translation Greek”, JBL 52 (1933) 244–52. Robert, U., Heptateuchi partis posterioris versio latina antiquissima e codice Lugdunensi. Version latine du Deutérome, de Josué et de Juges antérieure à Saint Jérôme (Lyon: Librairie de A. Rey et Cie, 1900). Ross, William A., “Septuagint Lexicography and Language Change in Greek Judges” (PhD diss, University of Cambridge, 2018). Ruijgh, C.J., “La place des enclitiques dans l’ordre des mots chez Homère d’après la loi de Wackernagel”, in H. Eichner/H. Rix (ed.), Sprachwissenschaft und Philologie. Jacob Wackernagel und die Indogermanistik heute (Wiesbaden: Reichert, 1990). Ryder, S.A., The D-Stem in Western Semitic (Janua Linguarum 131; The Hague: Mouton, 1974). Saeed, J.I., Semantics (4th edition; Chichester: Wiley Blackwell, 2015). Sailhamer, J.H., The Translational Technique of the Greek Septuagint for the Hebrew Verbs and Participles in Psalms 3–41 (Studies in Biblical Greek 2; New York: Peter Lang, 1991). Schaper, J., “The Concept of the Translator(s) in the Contemporary Study of the Septuagint”, in K. De Troyer/T. Law/M. Liljeström (ed.), In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014). Schehr, T.P., “Syntax of the Moods and Tenses of the Greek Verb in Septuagint Genesis 1–15” (PhD diss., Hebrew Union College, OH, 1990). Scholl, R., Sklaverei in den Zenonpapyri. Eine Untersuchung zu den Sklaventermini, zum Sklavenerwerb und zur Sklavenflucht (Trierer Historische Forschungen 4; Trier: Verlag Trierer Historische Forschungen, 1983). Schwyzer E./Debrunner A., Griechische Grammatik II. Syntax und syntaktische Stilistik (München: C.H. Beck, 1950). Sewell, A.H.N., (ed.), Xenophon, Oeconomicus (In Greek with Introduction and Notes in English) (Cambridge: University Press, 1925). Shead, A., “Jeremiah”, in J.K. Aitken (ed.), T&T Clark Companion to the Septuagint (London: Bloomsbury 2015) 469–86. Shipley, G., The Greek World After Alexander: 323–30 b.c. (London: Routledge, 2000). Siebesma, P.A., The Function of the Niph’al in Biblical Hebrew: In Relationship to Other PassiveReflexive Verbal Stems and to the Pu’al and Hoph’al in Particular (Assen: Van Gorcum, 1991).

318

Bibliography

Sipilä, S., Jordanin ylityskertomus: Septuagintan ja Masoreettisen tekstin väliset erot. (M. Th. thesis, Helsingin yliopisto, 1988). —, Between Literalness and Freedom: Translation technique in the Septuagint of Joshua and Judges regarding the clause connections introduced by ‫ ו‬and ‫( כי‬SESJ 75; Helsinki: The Finnish Exegetical Society, 1999). Smyth, H.W., Greek Grammar (Cambridge, MA: Harvard University Press, 1956). Soisalon-Soininen, I., Die Textformen der Septuaginta-Übersetzung des Richterbuches (AASF, ser. B, 72,1; Helsinki: Suomalainen Tiedeakatemia, 1951). —, “Septuagintan kielenkäyttö”, in Talenta quinque: Commentationes in hororem Ilmari Salomies, Eino Sormunen, E. G. Gulin, Rafael Gyllenberg, G. O. Rosenqvist (Helsinki: Otava, 1953) 46–52. —, Der Charakter der asterisierten Zusätze in der Septuaginta (AASF, ser. B, 114; Helsinki: Suomalainen Tiedeakatemia, 1959) 46–106. —, “Septuagintan lauseopin tutkimuksen perusteita”, Teologinen Aikakauskirja 68 (1963) 216– 31. —, “Der Gebrauch des genetivus absolutus in der Septuaginta”, in The Fifth World Congress of Jewish Studies (vol. 4; Jerusalem: World Union of Jewish Studies, 1973) 131–6. —, “The Rendering of the Hebrew Relative Clause in the Greek Pentateuch”, in Proceedings of the Sixth World Congress of Jewish Studies Held at the Hebrew University of Jerusalem, 13–19 August, 1973 (vol. 1; Jerusalem: World Union of Jewish Studies, 1977) 401–6. —, Die Infinitive in der Septuaginta (AASF, ser. B, 132,1; Helsinki: Suomalainen Tiedeakatemia, 1965). —, “Die Auslassung des Possessivpronomens im griehischen Pentateuch”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 86–92 —, “Einige Merkmale der Ûbersetzungsweise von Aquila”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 19–27. —, “Der Gebrauch des Verbs ἔχειν in der Septuaginta”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 181−8. —, “Methodologische Fragen der Erforschung der Septuaginta-Syntax”, in C.E. Cox (ed.), VI Congress of the International Organization for Septuagint and Cognate Studies, Jerusalem, 1986 (SBLSCS 23; Atlanta, GA: Scholars Press, 1987) 425–44. —, “Verschiedene Wiedergaben der hebräischen status constructus. Verbindung im griechischen Pentateuch”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 62–70. —, “Die Wiedergabe des ‫ ב‬instrumenti im griechischen Pentateuch”, in A. Aejmelaeus/R. Sollamo (ed.), Ilmari Soisalon-Soininen, Studien zur Septuaginta-Syntax. Zu seinem 70. Geburtstag am 4. Juni 1987 (AASF, ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987) 28–39.

Bibliography

319

—, “Zurück zur Hebraismenfrage”, in D. Fraenkel/U. Quast/J.W. Wevers (ed.), Studien zur Septuaginta. Robert Hanhart zu Ehren (Abhandlungen der Akademie der Wissenschaften in Göttingen / MSU 20; Göttingen: Vandenhoeck & Ruprecht, 1990) 35–51. —, “Back to the Question of Hebraisms” (in this volume). —, “A Palette of Renderings for the Hebrew Construct State Relation in the Greek Pentateuch” (in this volume). —, “Segmentation in the Greek Pentateuch” (in this volume). Sollamo, R., Renderings of Hebrew Semiprepositions in the Septuagint (AASF, Diss. Hum. Litt. 19; Helsinki: Suomalainen Tiedeakatemia, 1979). —, “The Pleonastic Use of the Pronoun in Connection with the Relative Pronoun in the Greek Pentateuch”, in C.E. Cox (ed.), VII Congress of the IOSCS, Leuven 1989 (SBLSCS 31; Atlanta, GA: SBL, 1991) 75–85. —, “The Pleonastic Use of the Pronoun in Connection with the Relative Pronoun in the LXX of Leviticus, Numbers and Deuteronomy”, in L. Greenspoon/O. Munnich (ed.), VIII Congress of the IOSCS, Paris 1992 (SBLSCS 41; Atlanta, GA: SBL, 1995) 43–62. —, Repetition of the Possessive Pronouns in the Septuagint (SBLSCS 40; Atlanta, GA: Scholars Press, 1995). —, “The Place of the Enclitic Personal Pronouns in the Old Greek Psalter”, in M.K.H. Peters (ed.), XII Congress of the International Organization for Septuagint and Cognate Studies, Leiden 2004 (SBLSCS 54; Atlanta, GA: SBL, 2006) 153–60. —, “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, Ljubljana 2007 (SBLSCS 55; Atlanta, GA: SBL, 2008) 339–51. —, “The Onoma Rule and the Vocative in LXX Genesis”, forthcoming 2020. Sollamo R./Sipilä S. (ed.), Helsinki Perspectives on the Translation Technique of the Septuagint: Proceedings of the IOSCS Congress in Helsinki 1999 (PFES 82; Helsinki: Finnish Exegetical Society/Göttingen: Vandenhoeck & Ruprecht, 2001). Spottorno, V., “The Status of the Antiochene Text in the First Century A.D.”, in S. Kreuzer/ M. Sigismund (ed.), Der Antiochenische Text der Septuaginta in seiner Bezeugung und seiner Bedeutung (DSI 4; Göttingen: Vandenhoeck & Ruprecht, 2013) 74–83. Sperber, D./Wilson, D., Relevance: Communication and Cognition (Oxford: Blackwell, 1986). Stipp, H.-J., “Offene Fragen zur Übersetzungskritik des antiken griechischen Jeremiabuches”, JNSL 17 (1991) 117–28. Swete, H.B., An Introduction to the Old Testament in Greek (2nd ed.; Cambridge: Cambridge University Press, 1914). Takahashi, Y., “The Historical Development of Niphal Verbal Forms in Biblical Hebrew”, BMECCJ 56 (2013) 83–95. Tal, A. (ed.), Biblia Hebraica Quinta: Genesis (Stuttgart: Deutsche Bibelgesellschaft, 2016). Talmy, L., “Force Dynamics in Language and Thought”, Cognitive Science 12 (1988) 49–100. Targarona Borrás, J., “Historia del texto griego del libro de los Jueces” (PhD diss., Universidad Complutense, 1983). Thackeray, H.St.J., “The Greek Translators of Jeremiah”, JThS 4 (1903) 245–66.

320

Bibliography

—, “The Greek Translators of the four Books of Kings”, JThS 8 (1907) 262–78. —, A Grammar of the Old Testament in Greek according to the Septuagint (Cambridge: University Press, 1909). —, The Septuagint and Jewish Worship: A Study in Origins (London: Oxford University Press, 1921). Thompson, D.J., “Slavery in the Hellenistic World”, in K. Bradley/P. Cartledge (ed.), The Cambridge World History of Slavery (Cambridge: Cambridge University Press, 2011) 194–213. Thompson, S., The Apocalypse and Semitic Syntax (SNTSMS 52; Cambridge: University Press, 1985). Timmer, J., “Age”, in R.S. Bagnall et al. (ed.), The Encyclopedia of Ancient History (Oxford: Blackwell, 2013) 173–8. Tjen, A., On Conditionals in the Greek Pentateuch: A Study of Translation Syntax (LHBOTS 515; London: T&T Clark, 2010). Tov, E., The Septuagint Translation of Jeremiah and Baruch: A Discussion of an Early Revision of the LXX of Jeremiah 29–52 and Baruch 1:1–3:8 (HSM 8; Missoula, MT: Scholars Press, 1976). —, “Some Aspects of the Textual and Literary History of the Book of Jeremiah”, in P.-M. Bogaert (ed.), Le Livre de Jérémie. Le prophète et son milieu, les oracles et leur transmission (BETL 54; Leuven: Leuven University Press, 1981) 145–67. —, The Text-Criticial Use of the Septuagint (Jerusalem: Simor, 1981). —, “The Literary History of the Book of Jeremiah in the Light of Its Textual History”, in J.H. Tigay (ed.), Empirical Models for Biblical Criticism (Philadelphia: University of Pennsylvania Press, 1985) 211–37. —, “The Representation of the Causative Aspects of the Hiph‘il in the LXX: A Study in Translation Technique”, in The Greek and Hebrew Bible: Collected Essays on the Septuagint (VTSup 72; Leiden: Brill, 1999) 195–202. —, Textual Criticism of the Hebrew Bible (Minneapolis MN: Fortress Press 2012). —, “The Septuagint Translation of Genesis as the First Scripture Translation”, in K. De Troyer/ T. Law/M. Liljeström (ed.), In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014). Tov, E./Kraft, R.A./Parsons P.J., The Greek Minor Prophets Scroll from Naḥ al Ḥ ever (8Ḥ evXIIgr) (DJD 8; Oxford: Clarendon Press, 1990). Tov, E./Wright B.G., “Computer-Assisted Study of the Criteria for Assessing the Literalness of Translation Units in the LXX”, Textus 12 (1985) 149–87. Trebolle, J., “The Textual History and the Text Critical Value of the Old Latin Version in the Book of Judges”, in W. Kraus et al. (ed.), Die Septuaginta—Text, Wirkung, Rezeption. 4. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 19.–22. Juli 2012 (WUNT 325; Tübingen: Mohr Siebeck, 2014) 53–72. Troxel, R.L., LXX-Isaiah as Translation and Interpretation: The Strategies of the Translator of the Septuagnt of Isaiah (Leiden/Boston: Brill, 2008). Vehmas-Lehto, I., Kopiointia vai kommunikointia: Johdatus käännösteoriaan (Helsinki: Finn Lectura, 2002).

Bibliography

321

Viti, C., “Coding Spatial Relations in Homeric Greek: Preverbs vs. Prepositions”, Historische Sprachforschung / Historical Linguistics 121 (2008) 114–61. Vogeser, J., Zur Sprache der griechischen Heiligenlegenden (München: Seitz, 1907). Voitila, A., “What the Translation of Tenses Tells about the Septuagint Translators”, SJOT 10 (1996) 183–96. —, Présent et imparfait de l’indicatif dans le Pentateuque grec: Une étude sur la syntaxe de traduction (PFES 79; Helsinki: Finnish Exegetical Society/Göttingen: Vandenhoeck & Ruprecht, 2001). —, “The Use of Tenses in the L- and B-Texts in the Kaige-Section of 2 Reigns”, in S. Kreuzer/M. Meiser/ M. Sigismund (ed.), Die Septuaginta—Entstehung, Sprache, Geschichte. 3. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 22.–25. Juli 2010 (WUNT 286; Tübingen: Mohr Siebeck, 2012) 213–37. —, “Middle Voice in the Speech of and about God’s Power and Dominion over the Creation in the Greek Pentateuch”, in Evangelia G. Dafni (ed.), Divine Kingdom and Kingdoms of Men / Gottesreich und Reiche der Menschen (WUNT 432; Tübingen: Mohr Siebeck, 2019) 35–47. Vorm-Croughs, M. van der, The Old Greek of Isaiah: An Analysis of its Pluses and Minuses (SBLSCS 61; Atlanta, GA: SBL Press, 2014). Waltke B.K./O’Connor M.P., An Introduction to Biblical Hebrew Syntax (Winona Lake, IN: Eisenbrauns, 1990). Wevers, J.W. (ed.), Genesis (SVTG 1; Göttingen: Vandenhoeck & Ruprecht, 1974). White, J.L., Light from Ancient Letters (Philadelphia: Fortress, 1986). Wifstrand, A., “Die Stellung der enklitischen Personalpronomina bei den Septuaginta”, in K. Humanistiska Vetenskapssamfundets i Lund Årsberättelse 1949–1950 II (Lund: Gleerup, 1950) 44–70. Wilson D./Sperber D., “Pragmatics and Time”, in R. Carston/S. Uchida (ed.), Relevance Theory: Applications and Implications (Pragmatics & Beyond New Series 37; Amsterdam: John Benjamins, 1998) 1–22. Wirth, R., “Das Praesens Historicum in den griechischen Samuelbüchern”, in K. De Troyer/T.M. Law/M. Liljeström (ed.), In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus (CBET 72; Leuven: Peeters, 2014) —, Die Septuaginta der Samuelbücher. Untersucht unter Einbeziehung ihrer Rezensionen (DSI 7; Göttingen: Vandenhoeck & Ruprecht, 2016). —, “Dealing with Tenses in the Kaige Section of Samuel”, in A. Aejmelaeus/T. Kauhanen (ed.), The Legacy of Barthélemy: 50 Years after Les Devanciers d’Aquila (DSI 9; Göttingen: Vandenhoeck & Ruprecht, 2017). Woo, S.-H., Etudes sur le système verbal dans la Septante de Job (Strasbourg: Éditions européennes universitaires, 2011). Zanchi, C., “New Evidence for the Source–Goal Asymmetry: Ancient Greek Preverbs”, in S. Luraghi/T. Nikitina/C. Zanchi (ed.), Space in Diachrony (Amsterdam: John Benjamins, 2017) 147–78. Ziegler, J. (ed.), Duodecim prophetae (SVTG 13; 2nd ed.; Göttingen: Vandenhoeck und Ruprecht, 1967).

322

Bibliography

—, Ezechiel (SVTG 16.1; 3rd ed.; Göttingen: Vandenhoeck und Ruprecht, 2006). —, Isaias (SVTG 14; 3rd ed.; Göttingen: Vandenhoeck und Ruprecht, 1983). Zurawel, T., “The Qal conjugation in Samaritan Hebrew”, in M. Bar-Asher (ed.), Massorot: Studies in Language Traditions (Jerusalem: Magnes Press, 1984) 135–51.

Indexes Ancient Sources The biblical index does not contain references to verses mentioned only in lists (e.g., instances of translation-technical phenomena; long reading lists pp. 178–184). The index of modern authors does not contain references to Ilmari Soisalon-Soininen.

Hebrew Bible/Old Testament Pentateuch 6, 19, 21, 22–25, 48, 65–66, 79–80, 81, 84, 86, 89, 95, 99–102, 104, 108, 110–117, 165, 230, 237, 241, 245–248, 252, 256, 258–259, 263–264, 270–271, 276, 282, 285, 294 Genesis 43–53, 85, 95, 236, 252, 259–260, 263, 269, 271, 275–276, 280, 282, 287 1:17 85, 245 1:29, 30 260 2 69 2:4 32 2:8 85, 217, 221 2:9 45, 217 2:10–14 273 2:11, 13 48 2:15 86, 217, 245, 256 2:16 217, 239 3:1–3 217 3:7 271 3:8 217

Genesis (cont.) 3:9 272 3:10 217 3:12 267 3:14 275 3:15 264 3:20 265 3:22 113 3:23 217, 221 3:24 44, 217 4:2 46, 291 4:4 266 4:11 275 4:15 86 4:22 267 5:24 272 6:4 51 6:5 21 7:2 52, 277 8:7, 8 114–115 8:9 239 8:10 114 8:19 49 9:2 68, 255 9:3 271 9:7 265 9:12 44, 51 9:13 87 9:17 94 9:22 44 10:3 49 10:9 264 10:10 48 10:12 274 11:10 289 11:28 240 11:29 51 12:3 287 12:11 52

324 Genesis (cont.) 12:13 22 13:10 48 14:2 274 14:3 235 14:8 47, 235, 253 14:9 253 14:10 235 14:12 281 14:13 44, 236, 276 14:18 46 15:1 289 15:10 87 15:17 234 16:8 49 16:11 278 16:12 235 16:16 258 17:2 93 17:6 87 17:8 48 17:11 49 17:12 239 18:1 227 18:2 257 18:8 279 18:10 272 18:17 279 18:27 276 18:29 291 19:4, 7 52 19:11 15 19:25 239 19:29 257, 259, 290 19:35 256 19:38 267 20:3 238 20:5 235 20:7 238, 265–266 20:11 44 20:12 238, 255 20:13 270

Indexes Genesis (cont.) 21:6 22 21:14 115 21:21 48, 236 21:23 261, 269 21:33 44 22:1 290 22:3, 6 238 22:7 290 22:16 45 22:17 259 23:10 228, 238 23:15 276 23:16 44 23:18 238 23:19 48, 246 24:2 91 24:7 267 24:8 256 24:9 91 24:23 239 24:24 274 24:37 261 24:40 114, 261 24:45 51, 267 24:47 270 24:48 233 24:53 230 24:56 104 24:60 112 25:6 115, 227 25:9 239 25:18 259 25:24 45 25:29 271–272 25:30 277 26:18 293 26:24 275 26:27 115 26:29 239 26:33 51 27 79

Indexes Genesis (cont.) 27:19 274 27:20 52 27:31 51 27:32 274 27:33 277 28:7 255 28:11 87 28:12 46 28:13 260, 274 28:15 276 29:4 276 29:31 112 29:33 61, 112 30:25 115 30:31 293 30:38, 41 245 31:3 44 31:11 268 31:18 48 31:23 236 31:27 22 31:33 249 31:44 265 32:3 (4) 48 32:5 104 32:10 45 32:11 (10) 292 32:12 49 32:12 (13) 87 32:13 113 32:23 (22) 45 32:24 90 32:26 115 33:4 256 33:10 22, 51 34:19 104 34:21 271 34:24 237 35:2, 4 234 35:15 261 35:17 110

Genesis (cont.) 35:19 274 36 47 36:17 47 36:20 237, 255 36:21 48 36:30 46 36:32 51 36:34 48 36:39 47, 237 36:40 47 36:43 47 37:2 272 37:7 281 37:17 249 37:31 238 37:35 255 37:36 239 37:38 238 38:5 281 38:11 265 38:17 114 38:19 44 38:20 49, 114, 238 38:21 271, 272 38:22 239 38:23 268 38:24 58 38:25 260, 278 38:28 110 39:8 44 39:19 43, 52 40:3 89, 245, 260 40:4 52 40:20 49 41 69 41:9 77 41:10 89, 245 41:19 49 41:33–34 48 41:40 111, 269 41:46 48

325

326 Genesis (cont.) 41:48 245 41:55 48 42:13 272 42:17 84, 89, 245 42:19 264, 265 42:30 89 42:32 249, 272 42:35 281 42:36 272 42:38 260 43:11 245 43:16 15 43:18, 23 245 43:27 78, 292 44:4 267 44:14 227 44:16 260, 272 44:18 52 44:19 269 44:26 69, 227, 268 44:34 227 45:1 113 45:20 271 46 47 46:1 46, 269 46:5–6 258 46:5 115 46:7 47 46:20 48 46:30 278 46:32 234, 255 46:34 234 47:1 255, 272 47:4 48 47:5 (6) 233 47:6 48, 235 47:11 235 47:13 48 47:19 267 47:26 92 48:7 112

Indexes Genesis (cont.) 48:19 265 49:13 52 49:26 232 49:29 246 50:11 48, 237 50:23 239 Exodus 48, 236, 259, 263, 269, 275–276, 280, 282 1:10 112 1:15–21 110 1:16 110, 271 1:19 277 2 69, 78 2:2 277 2:6, 7 68 2:14 279 3:9 269 3:12 113 3:20 116 4:10 240, 255 4:14 292 4:21 116, 246 4:27 111 5:2 65 5:11 264, 265 5:21 246 6:4 240 6:29 279 7:11 267 7:16 116 7:28 247, 253 8:4, 17 116 8:24 115 9:21 292 10 (9) 21 10:6 239, 281 10:13 234 11:4 235 11:8 281

Indexes Exodus (cont.) 12:29 228, 239 12:30 70, 268 13:3, 14, 15 235 13:18 238 13:20 239 14 (13) 21 14:12, 21 234 14:31 289 15:4 245 15:10 113, 247 15:16 235 15:23 270 15:24 286 15:25 92 15:26 275 16 (15) 21 16:17 268 16:20 20 16:29 279 17:5 246 17:14 246 18:3–4 237 18:11 239 18:19 267 18:21 232–233, 240, 255 18:25 233 19:4 264 19:6 235 19:13 264 19:18 236 20 21 20:24, 25 230 21:2 113 21:7 21 21:29 61, 271 22:1 (21:37) 21 22:5 (4) 20, 21 22:5 (6) 61 22:7(6) 21 22:9 (8) 19 22:28 104

Exodus (cont.) 23:1 233 23:5 112 23:16 245 23:31 246 25:7 238 25:10, 17 231 26 21 26:1 234 26:4 231 27:9 238 28:1 282 28:4 232 28:6 234 28:13 231, 255 28:23 (29), 26 (30) 292 28:39 230 29:2 231 29:22 234 29:29 67, 73, 232, 254 29:41 234 30:4 231 30:13 232 30:23 235 30:36 245 31:11 269 31:13 275 32:16 271 32:27 90 32:34 22 33 21 33:3 240, 277 33:5 240, 276 33:8 24 33:12, 17 292 34:9 240 34:12 239 34:14 270 34:35 281 35:5 237 35:9 (8), 10 (9) 238 35:22 231

327

328

Indexes

Exodus (cont.) 36:2 245 36:6 232–233 36:10 (39:3) 231, 234 36:12 (39:5) 278 39:1 (38:24) 233 39:18 (41) 232 39:20 (34) 236 40:26 (24) 245 Leviticus 16, 18, 48, 236, 269, 275–276, 280, 282 1:13 236 1:17 236 2:1, 2 236 2:7 238 2:9 236 3:5 236 3:17 238 4:6 232 5:3 267 5:7 238 5:15 67, 232 6:2–3 (5:21–22) 67 6:40 (7:10) 17 8:18 238 8:25 239 10:9 266 10:14 265 11:4–8 275 11:4 279 11:5 237, 275 11:14 237 11:28 275 12:6 292 12:7 105–106 12:8 106–107 13:51 22 13:55 278 14:7 105–106 14:9 239

Leviticus (cont.) 14:10 292 14:20 106 14:21 277 14:34 245 14:39–40 79 14:44 79 14:53 107 15:12 107 15:13 105 15:23 272 15:25 267, 271 15:28 106 16:10 115 16:12 235 16:17 23 16:32 230 17:3–4 17 17:8 67, 254 17:11 268 18:3 279 18:26 268 19:28 246 19:33 247 19:36 233 20:2 17 20:4 254 20:5 268 20:24 275 21 75 21:17–21 73–74 22:2 233 22:5 107 22:12 232 22:29 238 22:32 233 23:3 233 23:4 237 23:7 234 23:12 292 23:37 237

Indexes Leviticus (cont.) 23:43 245 24:10 238, 248 24:12 90 25:17 17 25:21 113 25:29 21 25:32 238 25:33 277 25:39 234 25:41 234, 268 25:54 265 26:1 245 26:17 227 26:25 246 26:34 276 26:36–37 227 27:4 277 Numbers 48, 236, 263, 269, 275–276, 280, 282 1:19 236 3:28, 32 232 4:6 231 4:7 231, 238 4:8 231 4:9 101, 231 4:13 231 4:14 232 5:6 234 5:20 245 6:9 107 7:9 232 8:7 107 9:5 236 9:10 289 10:35 112 11:8 231 12:16 236 14:3 32 14:4 293 14:19 235

Numbers (cont.) 14:41 268 17:22 245 16:40 (17:5) 23 18:3, 5 232 18:18 234 18:19 232 19:16 234 20:8 265, 266 20:13 274 20:17 234 21:22 234, 239 21:33 266 21:34 246 22 75 22:2–3 70 22:6 277 22:29 23 22:30 274 23:3 256 23:6 266 25:8, 14 233 27:12 239 28:18 233 29:13 238 31:6 232 31:19 266 31:20 230 31:23 106 31:28–53 234 32:20–23 71 32:21 73 32:24 115 33 247 33:4 112 33:55 279 34:9 238 34:12 235 35:11 239 35:16 231 35:25 233 36:8 234

329

330 Deuteronomy 48, 236, 269, 275–276, 280, 282, 287 1:2 239 1:25 246 1:27 246 1:42 272 2:3 269 2:5 236 2:14, 16 234 2:24, 30 246 3:1 266 3:2, 3 246 4:10 108–109 4:11 239 4:14 282 4:34 267 4:37 257 4:43 234 5:1 246 5:2 95 5:20 233 6:2 266 7:10 105 7:24 246 8:3 238 8:18 277 9:5 280 9:6, 13 240 10:3 231 10:16 68 10:21 274 10:22 66 11:6 77 11:10 215, 218–219 11:21 72, 78, 253 12:23 277 12:26 23 12:29 279 13:7 239, 255 14:4 238 14:9 261

Indexes Deuteronomy (cont.) 14:28 245 15:10 240 17:1 238 17:15 271 18:18 245 19:12 246 19:13 234 19:6ff 237 19:18 233 20:15 277 21:15 112 22:5 234 22:7 115 22:14 235 22:16 112 22:17 112, 235 22:19 233 23:15 235 23:22 105 24:1 235, 255 24:9 70, 225 24:12 278 24:15 68 25:17 70 25:18 277 26:15 233 27:4 245 27:5 230 28:3 275 28:8 113 28:27 233 28:48 230 28:54, 56 239 29:10(11) 239 29:22 113 30:12 270 30:16 78 31:11 70 31:12 109 31:25 237 31:27 227

331

Indexes Deuteronomy (cont.) 32:1 238 32:4 273 32:6 267 32:7 232 32:13 235 32:41 112 33:11 112 33:15 232 33:19 233 33:24 245 33:27 232 Joshua 27, 176 3:17 28 4:9 39 4:11 28 Judges–2 Kings (4 Kingdoms)

Judges (cont.) 12:11–12 167 13:20 165 14:20 165 18:20 247 21:21 163 Ruth 190, 200–202 2 201 2:5–6 200 2:9, 15, 21 201 3:8, 11, 14 200 4:11, 12 200 4:16 286 Samuel-Kings 164, 170–171, 175, (Books of 182, 184–187 Kingdoms)

225

Judges 29, 159, 161–162, 164–168, 170, 172, 175–181, 184–189, 224–226, 246, 248, 294, 297; reading lists pp. 178–184 are not indexed 3:15 163 4:1 167 4:11 163 5 177 5:3 167 7:11, 13 248 7:17 247–248 7:19 248 8:24 289 9:45 289 10:6 167 11:16, 18 248 11:30 165 11:40 163

1–2 Samuel (1–2 Kingdoms)

166, 170, 186

1 Samuel 161, 167–169 (1 Kingdoms) 3:11 290 4:3 248 6:14 171 7:6 172 9:5 226 9:25 248 9:27 225 10:3, 5, 6, 8, 14 172 10:16–19 172 20:26 110 2 Samuel 11:2–1 Kings 2:11

170–172

2 Samuel 171–172 10:6 172 10:17 171 11:1 228

332

Indexes

2 Samuel (cont.) 11:2 172 15:20 32 15:37 171 17:20 171–172 1 Kings (3 Kingdoms) 14:25 228 1 Kings 22–2 Kings 25 170 2 Kings (4 Kingdoms) 21:18 218 13:21 227 Nehemiah 8:12 109 Esther 175 7:7–8 218 Job 14, 224–226, 285, 294–295 1:16–18 227 6:17 228 21:22 108 22:27 226 Psalms 224, 287 2:12 111 Proverbs 14, 135, 224–226, 285 3:24 64 3:28 227 4:16 218 5:1 218 13:18 110 14:15 135 27:18 110 29:2 225 29:12, 14 227 29:16 228

Canticum (Song of Songs) 2:7 59 3:8 109 4:16 217 Isaiah 119–120, 122, 127, 132, 137, 296 1:17 127 1:30 218, 221 5:20 124 8:12–13 132 10:28–29 132 19:8 131 19:9 131, 135 19:10 131–132 23:4 130 23:5 130–131, 137 26:7 134 26:9 127 26:10 134 26:11 119, 134 26:16 135 28:15, 17–19 125–126 28:16 125 29:13 109 30:28–29 133–134 30:30 134 33:13 127 33:14 127–128, 136 33:15–16 128 35:10 128 40:23–24 123 41:15–18 120–122 42:16 124 50:8 117 51:10–11 129–130 53:5 132 55:12 109 57:13 122 58:3 32 58:13 110 59:8–9 124

Indexes Isaiah (cont.) 59:9 120 63:1–3 (63:19, 64:2) 136 63:15 135 65:3 218 65:14 117 Jeremiah 139–142, 144–157, 170 1–32 140 1–28 139–146, 148–157 1:10 143, 153 1:12 146 2:13 146 4:11 146 5:26 (MT) 150 6:7 151 6:10 142 8:5 144 11:7 146 11:13 153 14:8 147 15:3 146 15:21[20] 153 16:8 153 16:12 146 17:2 (MT) 150 17:24 143 18:6 144 19:14 145 22:23 150 25:6 145 25:12 150–151 28 155, 157 28(51):39 150 28(51):40 147 28[51]:59 148 29–52 139–146, 148–157 29 140 29(47):4 146 29:5 (36:5) 217

Jeremiah (cont.) 29(49):9 147 30:8 (49:30) 147 32:14 (25:28) 148 32:20 (25:34) 147 33–52 140 33(26):2 143 33(26):3 155 33(26):8 147 33(26):8 147, 151 33(26):12, 15, 21 143 35(28):9 149 38(31):12 146 38(31):18 109 38(31):28 145 39(32):32 143 41(34):6 150 41(34):15 145 43(36):8 145 43(36):21 143 43(36):23 151 43(36):26 142 44–51 140 45(38):4 147 47(40):14 143 48(41):4 147 48(41):4 147–148 48(41):6 147 48(41):7 151 48(41):10 146 49–51 (42–45) 143 49(42):18 149 50(43):1 147, 151 51(44):3 143, 147 51(44):8 143 51(44):12 143 51(44):14 143, 147, 155 51(44):18 147 51(44):19 146 52:4 148

333

334

Indexes

Ezekiel 190, 202–204, 224, 226, 296, 299 1:1–27:36, 299 40:1–48:35 (Ezek α) 1:4 297 3:10 100 4:14 110 5:5 246 9:6 202 16 60 16:54, 61 59 17:24 100 22:24 110 23:4, 6 202 23:12 203 23:19 202 23:23, 24, 27 203 28:1–36:23, 299 37:1–39:29 (Ezek β) 30:17 203 36:7 297 36:24–28 (Ezek ββ) 299 37:1 246 37:26 246 46:17 202

Zechariah 5:8 247

Daniel 175 3:29 60

2 Maccabees 1:33 249 8:19 61 12:21 57

Hosea 58 4:18 59 10:11 109 Micah 7:8 101 Zephaniah 2:1 135

1–2 Chronicles 175, 182, 184–185, 187 1 Chronicles 25:7 109 2 Chronicles 6:13 246 Apocryphal/Deuterocanonical Books 1 Esdras 9:55 109 2 Esdras 2:63 101 6:10 247 9:55 109 17:65 101 1 Maccabees 3:46 249 9:16 249 11:63 249

3 Maccabees 6:27 60 Wisdom 6:10 109 9:18 109 17:1 135

335

Indexes Sirach 135 3:14 61 4:3 63 13:3 63 13:21–23 62 40:17 218 40:27 218, 221 51:23 135

Diodorus of Sicily, The Library of History 14.80.2 215 16.41.5 216

New Testament

Euripides, Hecuba 711 86

Dionysius of Halicarnassus

195

Euripides, Andromache 739 109

Matthew 10:16 B

246

Flavius Josephus

Mark

244, 248

Flavius Josephus, Jewish Antiquities 1.37 216 7.347 216 9.225 216 10.226 216

Luke–Acts 244, 248 4:1 244 John 244, 248 1:18 244 Revelation 244

216

Herodotus, Histories 2.117.11 91 3.81 109 Hesychius, Lexicon

108, 218, 219

Homer

23, 209, 220

Greco-Roman Literature Aesop, Fable 62 86 Aristophanes of Byzantium, Περὶ ὀνομασίας ἡλικιῶν 37–66 189

Homer, Odyssey 20.387 86

Aristotle, The Athenian Constitution 42.1–2 189

Homer, Iliad 5.632 57 16.303 291 24.499 121 Θ 306 209, 220

Aristotle, Politics 7.1331a37–38 196

Isocrates, Panegyricus 40.1 91

Diodorus of Sicily

Philo 219

113, 195, 215

336

Indexes

Philo, Quis Rerum Divinarum Heres Sit 41 111 Philo, De Opificio Mundi 105 189 Plato 21 Plato, Phaedo 60b.6 21 Plato, Symposium 221e.1 20 Plato, Theaetetus 175c.1 20 178a.5 21 Plutarch, Lycurgus 21 189 Plutarch, Moralia 238a–b 189 544e 189

Xenophon

207–211, 219–221

Xenophon, Anabasis 209 I.2.7 210 II.4.14 211 Xenophon, Cyropaedia 211 1.2.4 189 1.3.14 211, 220 1.4.5 211, 220 1.5.4 189 8.1.38 211, 220 8.6.12 211, 220 8.7.6. 189 Xenophon, Hellenika 4, 1, 15, 8 210 II.3.11–IV.2.8. 210 Xenophon, Oeconomicus 4, 13 4, 13.14.20

208, 210 210 207, 210–211, 220

Pollux, Onomasticon 207 Papyri and Ostraca Polybius

113, 195, 249

Polybius, The Histories 4.16.6 199 4.20.7 196 4.35.1–3 199 4.76.8–9 199 5.30.1 199 6.20.1–3 199 18.53.1–11 199 21.3b 199 Theodoret of Cyr, Quaestiones in Octateuchum

182

BGU (Berliner Griechische Urkunden) 4.1078 (TM 9455) 193 4.1079 (TM 9456) 193 4.1141 214 6.1256 (TM 4543) 198 6.1280.17 20 6.1285 213 6.1290 (TM 4555) 194 7.1512 (TM 4762) 194 7.1518 (TM 4768) 194 7.1527 (TM 4777) 194 7.1526 (TM 4776) 194 7.1531 (TM 4781) 194 7.5121 (TM 4771) 194

Indexes BGU (cont.) 8.1738.32 22 14.2367.17 21 Inscriptiones Graecae VII 2715–21 196 Inscriptiones Graecae XII.9 Supp. 646 196 SB (Sammelbuch griechischer Urkunden aus Aegypten) 18.13256.9 21 O.Stras. (Griechische und griechischdemotische Ostraka der Universitäts- und Landesbibliothek zu Strassburg im Elsass) 1.585 214 1.603 (TM 76187) 194 P.XV.Congr.15. 213 P.Amh. (The Amherst Papyri) 2.39 (TM 164) 199 P.Athen. (Papyri Societatis Archaeologicae Atheniensis) 43 R 213 P.Cair.Zen. (Zenon Papyri, Catalogue général des antiquités égyptiennes du Musée du Caire) 212 1.33.7 22 1.59076 (TM 731) 191 1.93.18 22 2 59156 214 2.59176 (TM 822) 194 2.59195 (TM 841) 194 2.59226 249 2.59254 (TM 899) 197 2.59292 (TM 936) 194 3.59298 (TM 942) 195

337

P.Cair.Zen. (cont.) 3.358.3 23 3.59378 (TM 1021) 192 3.59406 (TM 1048) 194 3.59435 (TM 1075) 194 3.59498 (TM 1136) 192–193 3.59509 (TM 1147) 193 4.59677 (TM 1305) 194 4.59698 (TM 1325) 194 4.59729 (TM 1356) 194 4.599.8 23 5.815.4 22 P.Dryton (Family Archive of Dryton) 4 94 P.Eleph. (Aegyptische Urkunden aus den Königlichen Museen in Berlin: Griechische Urkunden) 2 93 12 90 P.Enteux. (ΕΝΤΕΥΞΕΙΣ: Requêtes et plaintes adressées au Roi d’Égypte au IIIe siècle avant J.-C., Cairo) 22 93 P.Flor. (Papiri greco-egizii, Papiri Fiorentini) 1.16 215 P.Freib. (Mitteilungen aus der Freiburger Papyrussammlung) 1.7 (TM 5644) 197–198 P.Giss. (Griechische Papyri im Museum des oberhessischen Geschichtsvereins zu Giessen) 1.13 213 P.Gur. (Greek Papyri from Gurob) 2 92 10.8 90

338

Indexes

P.Hamb. (Griechische Papyrusurkunden der Hamburger Staats- und Universitätsbibliothek) 1.114 (TM 2366)

P.Oxy. (The Oxyrhynchus Papyri) 14.1648 213–214 227 212–213 2465 196

P.Heid.Gr. (Veröffentlichungen aus der Heidelberger Papyrussammlung) 3.227 (TM 78313) 194–195

P.Oxy.Hels. (Fifty Oxyrhynchus Papyri) 22 214

P.Pher. Untitled, Hermopolis, after 249 c.e. 213 P.Hib (The Hibeh Papyri) 1.29.10 22 P.Iand. (Papyri Iandanae) VI 92 195 P.Leid.Inst. (Papyri, Ostraca, Parchments and Waxed Tablets in the Leiden Papyrological Institute) (= P.Ludg.Bat. 25) 21 215 P.Lond. (Greek Papyri in the British Museum) 3 1157 V(a) (S. 109) 213 7.1941 (TM 2384) 195 7.2033.6 23 7.2164 (TM 1724) 194 7.2193 91 25 21 P.Mich. (Michigan Papyri) 1.29.3 22 1.45 212 1.49 (TM 1949) 194 5.272 215 5.282 214, 215 P.Oslo (Papyri Osloenses) 3.108 213

P.Petr. (The Flinders Petrie Papyri) 2.3a.1 23 2.45 198 2.5[a] 89 P.Rein. (Papyrus grecs et démotiques recueillis en Égypte) 2.93 214 P.Ryl. (Catalogue of the Greek and Latin Papyri in the John Rylands Library, Manchester) 2.215 213 P.Stras. (Griechische Papyrus der Kaiserlichen Universitäts- und Landes-bibliothek zu Strassburg) 1.23 214 P.Tebt. (The Tebtunis Papyri) 1.86 R 213 3.1.700 213 3.1.769 90 3.2.834 213 3.2.1001 R 213 P.Yadin (The Documents from the Bar Kochba Period in the Cave of Letters) 123 215 PSI. (Papiri greci e latini. Florence.) 4.367 90 8.917 215

Indexes Sammelbuch griechischer Urkunden aus Ägypten 12.11115 213 16.12373 213

Aejmelaeus, A.  14, 35, 40, 45, 48, 51, 55, 65, 67, 70–71, 101, 159–161, 168, 170–171, 173, 186, 223, 229, 241, 251, 256–257, 263, 288, 292, 296 Aitken, J.K.  18, 154, 204, 207 Allan, R.J.  82, 86, 89 Alves, F.  76 Armitage, P.  112 Arnold, M.  102 Auld, A.G.  39 Ausloos, H.  176

Belloro, V.  83 Ben-Hayyim, Z.  117 Benton, R.  102 Benveniste, É.  83 Berry, P.J.  112 Bethe, E.  207 Bieżuńska-Małowist, I.  194–195 Blass, F.  243–244, 246–248, 264, 288, 291 Blomqvist, J.  43 Bloomer, W.M.  196 Bodine, W.R.  176 Bogaert, P.-M.  141 Bolikowski, Ł.  192 Bons, E.  56, 200–202, 207 Booij, G.  56 Boulluec, A. Le  120 Bradley, K.  193 Bromley, W.  249 Brooke, A.E.  162, 186 Brown, J.P.  211, 214 Busto-Saiz, J.R.  99–100, 175, 186 Büchner, D.  66 Bybee, J.  120

Baden, J.T.  102 Bagnall, R.S.  91, 93, 196 Baily, A.  218–219 Bakker, E.J.  82, 86, 88, 93 Bar-Asher, M.  117 Barber, E.J.W.  82 Barr, J.  40, 66, 97, 154, 283, 285, 293–295 Barthélemy, D.  99, 159, 165, 167, 169, 172, 176 Basello, G.P.  209 Bauer, H.  150 Baumgartner, W.  209 Baumgärtel, F.  161 Beck, J.A.  97 Beckman, J.C.  104, 110 Beekes, R.S.P.  208–209, 218–219 Bell, R.  196

Caird, G.B.  61–62 Cañas Reíllo, J.M.  176, 180, 182 Cantarella, E.  196 Gardiner, E.N.  193, 195 Cartledge, P.  193 Carston, R.  40 Casey, E.  196 Chankowski, A.S.  196, 198–199 Chantraine, P.  119, 218–219 Claassen, W.Th.  102 Clines, D.J.A.  150 Cole, P.  36 Cook, J.  60, 65 Cooper, J.S.  118 Cowley, A.E.  264 Cox, C.E.  286 Creason, S.A.  103–104

UPZ (Urkunden der Ptolemäerzeit) 1.123 93 1.60.17 21 81.2.6 249

Modern Authors

339

340 Cunliffe, R.J.  209, 220 Cunningham, I.C.  218 D’Amore, L.  197 Dafni, E.G.  95 Dakyns, H.G.  211 Daniel, R.W.  198 Debrunner, A.  43–44, 49, 242–244, 246–248, 264, 288, 291 Deissmann, A.  29, 207, 287 Depauw, M.  192 Derow, P.  91, 93 Dorival, G.  204 Downie, J.  77 Edgar, C.C.  195 Eichner H.  23 Endo, Y.  102 Evans, T.V.  65, 97, 100, 207 Everett-Heath, J.  203 Eynikel, E.  208 Fakour, M.  208 Fassberg, S.E.  117 Fedorov, A. V.  29 Fernández Marcos, N.  175–176, 179, 181–182, 184–186, 296 Field, F.  186 Fisher, R.  99, 112, 113 Forbes, C.A.  189, 196–197 Fraenkel, D.  13, 34, 163, 177, 283 Frankel, Z.  229 Funk, R.  291 Gardiner, E.N.  195 Garland, R.  190, 196 Gehman, H.J.  253 Gentry, P.J.  97 Gerth, B.  49, 242–243, 246, 256 Gesenius, W.  230, 232, 264 Gheldof, T.  192

Indexes Gil, J.  210 Gile, D.  76 Goetze, A.  103, 110 Golden, M.  189 Grammaticus, P.  207 Greenspoon, L.  68, 170 Grenfell, B.P.  212 Grethlein, J.  209 Grice, P.  27, 36–41 Grieve, A.J.  207 Gryson, R.  181 Grubbs, J.E.  196 Guerrero, E.  83 Gutt, E.-A.  36 Hanhart, R.  216, 287 Hansen, P.A.  218 Harl, M.  39, 204 Hatch, E.  100, 119–120, 133, 247 Hatzidakis, G.N.  243 Hauspie, K.  204, 208 Helbing, R.  30, 56, 284, 288–289, 291 Helms, H.  199 Herrmann, J.  161 Hertog, C.G. den  39 Himbaza, I.  79 Hintze, A.  209 Hopkins, K.  192 Horrocks, G. C.  20, 21 Humbert, J.  248 Hunt, A.S.  198, 212 Jakobsen, A.L.  76 Jannaris, A.N.  248 Janzen, J.G.  141 Jastrup, P.O.  43 Jenni, E.  103–105, 108–112 Jobes, K.H.  99 Jones, H.S.  86, 208, 218–219, 291 Johannessohn, M.  241, 243–245, 247–250, 284, 291

Indexes Johnson, B.  168 Joosten, J.  56, 60, 102–104, 113, 116, 207 Joüon, P.  102 Karlik, J.  76 Karrer, M.  108, 111, 179 Katz, P.  98 Kaufman, S.A.  118 Kauhanen, T.  159, 161, 171, 173 Kautzsch, E.  43, 264 Kemmer, S.  82–83 Kennell, N.M.  196 Kittel, G.  248 Klaiman, M.H.  81–82 Klein, G.L.  102 Klinken, W. van  65, 67 Kloppenborg, J.S.  198 Kouwenberg, J.C.  104 Kraemer, S.  92 Kraft, R.A.  159 Kraus, T.J.  56 Kraus, W.  108, 111, 181–182 Krentz, P.  210 Kreuzer, S.  79, 170, 176, 179, 184, 202, 204 Kriston, A.  76 Kutscher, E.Y.  117 Kühner, R.  49, 242–243, 246, 256 Köhler, L.  209 LaMontage, N.  177 Langacker, R.W.  82–83, 85 Launey, M.  196, 197 Law, T.M.  65, 72, 171 Lee, J.A.L.  19, 24, 51, 68, 113, 176, 207–208, 212, 217, 220 Leander, P.  150 Lemmelijn, B.  176 Léonard, J.M.  102 Lesquier, J.  197 Lewis, N.  93 Lewis, S.  189–190

341

Liddell, H.G.  86, 208, 218–219, 291 Liljeström, M.  65, 72, 171 Louw, T.A.W. van der  13–14, 18, 69, 74, 76–77, 79, 221, 223, 229, 241, 251, 263, 283 Luraghi, S.  56 Lust, J.  208 Maldonado, R.  83–84, 88, 90, 91, 95 Malmkjaer, K.  76 Margolis, M.L.  57–58 Marle, J. van  56 Marquis, G.  283, 286–287, 296–299 Martzavou, P.  196 Matthews, J.N.S.  112 Mauersberger, A.  199 Mayser, E.  24, 47, 231, 243–244, 246, 248–250, 264–265, 270, 273 McFall, L.  97, 102 McGregor, G.S.  192, 199 McLay, T.  101, 169 McLean, N.  162, 186 McTait, K.  75 Meiser, M.  74, 79, 170, 176, 179 Mey, J.L.  36 Milligan, G.  198 Min, Y.-J.  141 Moigne, P. Le  73–75, 120 Montanari, F.  210, 218–219 Montevecchi, O.  207 Morgan, J.  36 Moulton, J.H.  198, 242, 288 Munnich, O.  68, 170, 204 Muraoka, T.  24, 27, 34, 43, 46–47, 49–50, 67, 209, 218–219 Nikitina, T.  56 O’Connor, M.P.  33, 43, 97, 102–104, 110 Oepke, A.  248 Okla, N.  209, 220

342

Indexes

Olohan, M.  75 Osty, E.  120, 131, 136 Panaino, A.  209, 220 Papazarkadas, N.  196 Parkin, T.G.  189, 196 Parsons, P.J.  159 Pearson, K.  101 Penner, K.M.  102 Peters, M.K.H.  20, 66, 69 Peursen, W.Th. van  102 Pietersma, A.  65, 140, 154 Peterson, J.H.  208 Piquer Otero, A.  160, 168 Preisigke, F.  250 Pretzl, O.  166–167 Psichari, J.  288, 291 Pöchhacker, F.  76–77 Qimron, E.  117 Quast, U.  13, 34, 163, 200, 283 Rahlfs, A.  162, 168, 176–177, 180, 202, 216, 236, 258 Redpath, H.A.  100, 119–120, 133, 247 Rehkopf, F.  243–244, 246–248, 264, 288 Reider, J.  98–100 Reekmans, T.  194 Rife, J.M.  100 Rix, H.  23 Robert, U.  181 Ross, W.A.  190 Rossi, A.V.  209 Ruijgh, C.J.  23 Ryder, S.A.  103–104, 110, 117 Saeed, J.I.  86 Sáenz-Badillos, A.  182 Sailhamer, J.H.  97 Saunders, M.  154 Schaper, J.  65, 78 Schehr, T.P.  97

Scholl, R.  194 Schwartz, G.M.  118 Schwyzer, E.  43–44, 49, 242 Scott, R.  86, 208, 218–219, 291 Seleznev, M.  112 Sergio Ibáñez, L.  83 Sewell, A.H.N.  208 Shead, A.  154 Shibatani, M.  81 Shipley, G.  201 Shlesinger, M.  76 Sigismund, M.  79, 170, 179, 184 Siebesma, P.A.  102 Silva, M.  99 Sipilä, S.  35, 39, 41, 55 Smend, R.  98 Smyth, H.W.  75, 83 Sollamo, R.  14, 18, 20, 35, 45, 48–49, 51–52, 55, 68–69, 223, 229, 241, 246, 247, 251, 263, 288, 292, 294 Sperber, D.  38, 40 Spottorno Díaz-Caro, M.V.  182 Stipp, H.-J.  140, 155 Swete, H.B.  168, 298 Takahashi, Y.  102 Tal, A.  117, 217 Talmy, L.  88 Targarona Borrás, J.  176 Taylor, B.A.  39, 101 Thackeray, H. St.J.  16, 30, 114, 139–140, 142, 154, 161, 172, 204, 223–224, 284, 288, 299 Thompson, D.J.  193, 195 Thompson, S.  98–100, 113 Tigay, J.H.  141 Timmer, J.  189 Torijano Morales, P.A.  160, 168 Tov, E.  72, 79, 98–100, 139–142, 153, 155, 159, 170, 283, 295–296 Trebolle Barrera, J.  176, 181–182 Trinquet, J.  120, 139

Indexes Troyer, K. De  65, 72, 171 Troxel, R.L.  126–127 Trujillo, A.  75 Turner, N.  241–242 Uchida, S.  40 Ulrich, E.  79 Vehmas-Lehto, I.  29 Viti, C.  56 Vogeser, J.  249 Voitila, A.  55, 95, 97, 170 Vorm-Croughs, M. van der  126, 131, 136 Waltke, B.K.  33, 43, 97, 102–104, 110 Wevers, J.W.  13, 34, 48, 74, 163, 249, 258, 283

White, J.L.  191, 193 Wifstrand, A.  297 Wilson, D.  38, 40 Windle, K.  76 Wirth, R.  170–172, 186–187 Woo, S.-H.  97 Wright, B.G.  154, 283, 295–296 Yarshater, E.  208 Zanchi, C.  56 Ziegler, J.  120, 202–203, 299 Zilverberg, K.J.  175 Zurawel, T.  117

343