Reading the Sealed Book: Old Greek Isaiah and the Problem of Septuagint Hermeneutics 9781602589803, 9783161525575, 9783161578151, 1602589801

Citation preview

Forschungen zum Alten Testament Herausgegeben von Konrad Schmid (Zürich) • Mark S. Smith (New York) Hermann Spieckermann (Göttingen)

88

J. Ross Wagner

Reading the Sealed Book Old Greek Isaiah and the Problem of Septuagint Hermeneutics

Mohr Siebeck Baylor University Press

J. Ross Wagner, born 1966; 1999 PhD from Duke University; currently Associate Professor of New Testament, Duke Divinity School.

Distributors For the United States and Canada Baylor University Press One Bear Place #97363 Waco, Texas 76798 USA

For Europe and the UK Mohr Siebeck Wilhelmstraße 18, Postfach 20 40 D-72010 Tübingen Germany

All other countries are served by both publishers. Library of Congress Cataloging-in-Publication Data is on file at the Library of Congress, Washington, D.C. ISBN 978-1-60258-980-3 The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. ISBN 978-3-16-152557-5

978-3-16-157815-1 Unveränderte 2019 ISSN 0940-4155 (Forschungen zumeBook-Ausgabe Alten Testament) ©2013 by Mohr Siebeck, Tübingen. This book may not be reproduced, in whole or in part, in any form (beyond that permitted by copyright law) without the publisher's written permission. This applies particularly to reproductions, translations, microfilms and storage and processing in electronic systems. The fonts used to print this work (SymbolGreekTU, NewJerusalemCU, and TranslitLSU) are available from Linguist's Software, Inc., PO Box 580, Edmonds, WA 98020-0580 USA. Tel (425) 775-1130; www.linguistsoftware.com. The Unicode Syriac font used in this book (Estrangelo Edessa) is freely available from http://www.bethmardutho.org. The book was printed on non-aging paper by Guide-Druck in Tübingen and bound by Großbuchbinderei Spinner in Ottersweier. Printed in Germany.

Meiner Frau So ist's ja besser zu zweien als allein; denn sie haben guten Lohn für ihre Mühe. Wenn sie fallen, so hilft der eine dem andern auf.

Foreword This book began to take shape during a sabbatical year spent as a Humboldt Research Fellow at the Georg-August-Universität Göttingen in 2006-2007. My sincere thanks go to the Alexander von Humboldt-Stiftung, the University of Göttingen and Princeton Theological Seminary for the generous support that funded that year and a subsequent research stay in the summer of 2010.1 offer my deepest gratitude to my hosts, Prof. Dr. Hermann Spieckermann and Prof. Dr. Florian Wilk, for their warm welcome, generous friendship and continuing collaboration over these past seven years. My thanks go also to Dr. Bernhard Neuschäfer for graciously providing access to the resources of the Septuaginta-Unternehmen. Our family's time in Germany was greatly enriched by the hospitality of kind friends in Göttingen and Bremen, including the families of Karsten and Juliane Ahlers, Gisela Apel, Branislav and Jasna Beocanin, Eberhard and Beate Busch, Markus and Doro Frenz, Rainer and Barbara Hirsch-Luipold, and Florian and Eimelt Wilk. In Göttingen, Michael Grimmsmann and Judith Krawelitzki helped trouble-shoot everything from residence permits and die Schulpflicht to transportation, housing and Handys, while my colleague and friend Shane Berg provided invaluable support back in Princeton. I owe a particular debt of gratitude to Darrell Guder, my former dean at Princeton Theological Seminary. Without his vision, encouragement and assistance, my residence in Germany would not have been possible. Parts of the present work were written during my tenure as a member of the Center of Theological Inquiry in Princeton from 2009 to 2010. I would like to thank the Director, Dr. William Storrar, Dr. Thomas Hastings, former Director of Research, and my fellow members-in-residence at the Center for fostering a collégial and stimulating environment in which to think and work. A week of intense writing in the guest cottage of the Franciscan Friars, T.O.R., in Loretto, PA, proved crucial to the completion of this volume. I am grateful to my hosts, Fr. Daniel Mulkern, Fr. James Morman and Fr. Malachi Van Tassel for their generous hospitality. I also wish to express here my gratitude to the editors of Forschungen zum Alten Testament for accepting this volume into the series and to Dr. Henning Ziebritzki and his staff at Mohr Siebeck for their courteous and professional assistance during the process of readying the manuscript for the press. Thanks also to Dr. Carey Newman of Baylor University Press, whose encouragement over the many years we have known one another has been a great gift.

VIII

Foreword

A wonderful group of students at Princeton Theological Seminary provided research assistance at various stages of my work: Colin Yuckman, Toby Long, Amy Peeler, Christopher Holmes, and Thomas Dixon. A special word of thanks goes to Joel Estes, now a doctoral student at PTS, for his insightful criticism of the penultimate draft of this book and for his sharp editorial eye. Many conversation partners have influenced my thinking about Old Greek Isaiah during this project's long period of gestation. I would like to thank them for their willingness to discuss and debate "Septuagint hermeneutics" with me and, at the same time, absolve them of any guilt by association with what I have written: Anneli Aejmelaeus, Gary Anderson, Frank Austermann, Friedrich Avemariet, Shane Berg, Markus Bockmuehl, Chip Dobbs-Allsopp, Peter Gentry, Richard Hays, Arie van der Kooij, Reinhard Kratz, Nathan MacDonald, Mel Peters, Tessa Rajak, Leong Seow, Hermann Spieckermann, Loren Stuckenbruck, Ronald Troxel, and Florian Wilk. Finally, I would like to say a public word of thanks to my family: to my parents, Jackson and Janett Wagner, and my in-laws, Adolf and Linda Webel, for their wise counsel and unflagging support; to my dear children, Nathaniel, Caleb, Naomi and Claire, for the joy and chaos and laughter they bring to our home; and above all to Ronda, for her steadfast faith and hope and love as we journey through life together.

Table of Contents Foreword

1. A Book with Seven Seals: The Problem of Septuagint Hermeneutics The Shape of the Controversy A Theory of Translation Describing Translation: Function, Process, Product The 'Acceptability ' of a Translation in the Target Culture Classifying Translations Overview In Search of the 'Typical' Septuagint Translation: Aquila 'Atypical' Translations in the Septuagint Corpus OG Reigns OGJob Categorizing Old Greek Isaiah Characteristics of the Translation Conflicting Models of Translation Breaking the Seals: The Plan of This Study

2. Opening the Sealed Book: Interpreting a Translated Text A Framework for Interpretation The Cultural Encyclopedia The Model Reader and the Intention of the Text Model Author and Model Translator Interpreting a Translated Text Investigating the Process of Translation Analyzing the Product of Translation Constructing the Cultural Encyclopedia Résumé: Reading the Sealed Book

VII

1 2 6 6 8 11 11 12 17 17 22 29 29 31 34

37 37 37 39 43 45 46 52 56 62

X

Table of Contents

3. "Give Heed to God's Law": Translation and Interpretation in OG Isaiah 1:1-20 Isaiah's Opening Vision Overview (1:1-31) Superscription (1:1) Israel Accused (1:2-20) First Address: "How Long Will You Practice Lawlessness? " (1:2-7) Summons to Heaven and Earth (1:2a) Complaint: Israel Does Not Know Me (l:2b-3) Woe to the Disobedient Children (1:4) Appeal to Israel: Debilitated Body, Devastated Land (1:5-7) Israel's Response: Lament and Hope (1:8-9) Second Address: "Give Heed to God's Law " (1:10-20) Summons to Leaders and People (1:10) Complaint: Barren Worship, Bloodstained Hands (1:11-15) Call for Repentance and Social Justice (1:16-17) Appeal to Israel: Restoration or Ruination (1:18-20)

4. The Purification of Zion: Translation and Interpretation in OG Isaiah 1:21-31

64 64 64 67 70 70 71 72 80 84 93 98 98 103 138 142

148

Zion's Doom Pronounced: Trial by Fire (1:21-31) 148 The City Cleansed (1:21-27) 150 Lament for Zion (1:21-23) 150 Rebellious Rulers Ruined, Righteous Rule Restored (1:24-27).... 165 The Lawless Consumed (1. 28-31) 196 The End of the Wicked (1:28) 199 Withered Tree, Waterless Garden (1:29-30) 202 Excursus: Translation of Verbs in OG Isaiah 1-5 205 An Unquenchable Conflagration (1:31) 223

5. Characterizing Old Greek Isaiah Constitutive Character Linguistic Acceptability Textual Acceptability Literary Acceptability Prospective Function Isaiah with a Greek Accent: Interpretation in OG Isaiah 1

227 227 227 229 232 234 235

Table of Contents

XI

Epilogue: Open Book, Overflowing Fountain

239

Bibliography

241

Index of Ancient Sources

269

Index of Modern Authors

291

Chapter 1

A Book with Seven Seals: The Problem of Septuagint Hermeneutics Kai eaovxai ij|j.iv ra'tvxa xa p r p a x a x a i x a ox; oi Xoyoi xoi pxpAiou xo-0 ec^payianevou xotixou, o e a v S O K T I V a u t o avOpamcp ¿ T T I A X A N E V C P ypaji(iaxa keyovxeq Avayvw8i xaixa- Kai e p e i OIJ 8ijva|iai ravrayvoivrai, eo^paytoxax yap. OG Isaiah 29:11

The past two decades have witnessed something of a renaissance in the study of the Septuagint. The "first major translation in western culture,"1 the Greek Pentateuch quickly became the center of a growing body of Jewish-Greek literature comprising original works alongside translations of additional 'scriptural' texts.2 These texts offer eloquent testimony to the cultural creativity and religious vitality of hellenistic Judaism. Moreover, because of their Nachleben as scripture both for diaspora Jews and for emerging Christian communities, the writings collected under the rubric of 'the Septuagint' are of signal importance for investigating early Jewish and Christian practice and belief. The contemporary confluence of large-scale translation projects in French (La Bible d'Alexandrie),3 German (Septuaginta Deutsch),4 Spanish (La Biblia Griega Septuaginta)5 and English (A New English Translation of the Septuagint)6 has sparked a vigorous debate over how to understand and interpret the 1

Rajak 2009: 1. Sebastian Brock observes that "the Greek translation of the Pentateuch was an undertaking totally without precedent in the Hellenistic world" (Brock 1972: 12; see also Brock 1974; Brock 1984; Brock 1992). Dell'Acqua 2010 surveys the evidence for translation in Ptolemaic Egypt. 2 Accounts of Septuagint origins can be found in Swete 1914; Jellicoe 1968; Dorival et al. 1988; Fernández Marcos 1998 (ET, Fernández Marcos 2000); Jobes and Silva 2000; Siegert 2001; Dines 2004; Tilly 2005; Law 2013. 3 Numerous volumes of this translation and commentary have already appeared (e.g., Dogniez et al. 2001), with more to follow. See also Harl 1992. 4 Translation: Kraus and Karrer 2009; Notes and Commentary (two vols): Karrer and Kraus 2011. See also Kraus 2006; Kraus 2010. 5 Fernández Marcos and Spottorno Díaz-Caro 2008. See also Fernández Marcos 2008. 6 Pietersma and Wright 2007. The introduction to NETS offers a clear statement of the principles behind the translation (Pietersma and Wright 2007a). There are also two Septuagint commentary series in the works in English: the Society of Biblical Literature Commentary on

2

Chapter 1: The Problem of Septuagint

Hermeneutics

translational literature within this corpus.7 The complexity of the issues has, in turn, given rise to sharply divergent approaches to what might be termed 'Septuagint hermeneutics.'8 Despite considerable advances elsewhere in the field, there is still no broad consensus concerning either the character of the translated texts within the Septuagint corpus or the proper methods for their interpretation. In this respect, at least, Hans Hiibner's wry observation from twenty years ago would still seem to hold: the translation of the Seventy remains "a book with seven seals."9

The Shape of the Controversy The debate over Septuagint hermeneutics circles around a number of interrelated questions. While we find a spectrum of positions on each issue, opinions have tended to cluster at one end of the continuum or the other. Thus, for the purposes of this brief overview, I will risk oversimplification by speaking in each case of two primary alternatives.10

the Septuagint, connected with NETS and sponsored by the International Organization for Septuagint and Cognate Studies (http://ccat.sas.upenn.edu/ioscs/commentary/prospectus.html; see also Pietersma 2004); and the Septuagint Commentary Series published by Brill (http:// www.brill.nl/publications/septuagint-commentary-series). A brief statement of its principles may be found in Porter and Pearson 1997: 546. 7 See, for example, the different perspectives represented by the essayists in B. Taylor 2001: 181-240, Ausloos 2008, and Hiebert 2010. In the present work, I use 'Septuagint corpus' to refer to the set of Greek texts (both translations and new compositions) collected in early Christian codexes of the 'Old Testament' and conveniently collected in the Handausgabe of the Septuaginta Unternehmen in Gottingen (Rahlfs and Hanhart 2006). The limits of the collection can remain somewhat loosely defined, for there is no reason to believe that there ever existed a distinct 'Alexandrian canon' of scripture (Sundberg 1964). I reserve the designation ' L X X ' for the original translations of the Books of Moses (cf. Peters 1992: 109394), using ' O G ' to refer to the earliest translations of other 'scriptural' texts (cf. Wright 2008:104; Greenspoon 1987). Unless otherwise noted, I take the critical editions in the Gottingen Septuagint (1931) to represent the closest approximation of the LXX/OG texts readily available to scholars. 8 I take 'Septuagint hermeneutics' to be concerned with two interrelated questions: how a modern reader is to interpret the translated text, and how to characterize the translator's own interpretation of his source. 9 Hiibner 1990, 1:62. Albert Pietersma, the doyen of North American Septuagint Studies and co-editor of NETS, avers: "At the risk of being proven a 'pseudoprophetes,' I predict that the hermeneutics of the Septuagint will become one of the central issues (if not the central issue) in the discipline for some time to come" (Pietersma 2005: 2; emphasis original). 10 As will become clear, the two alternatives are best represented by the 'interlinear paradigm' underlying NETS, on the one hand, and by the approach behind La Bible d'Alexandrie, on the other. Despite the intention of the editors of the Septuaginta Deutsch to find a 'middle

The Shape of the Controversy

3

The first question concerns the degree to which the textual-linguistic character of the LXX/OG translations conforms to target-language models. That in these Greek texts we encounter passages of literary beauty, along with not a few that border on incoherence, is beyond dispute.11 But one viewpoint holds that, due to pervasive linguistic interference from the source texts, "unintelligibility of the Greek text qua Greek text is one of [the] inherent characteristics" of the Septuagint."12 At the other end of the spectrum lies the view that the L X X / O G translators "produce a text, if not easy to read, in any case almost always of good 'greekness,' comprehensible and coherent - at least just as much so as the MT, and sometimes more so."13 A second area of disagreement centers on the nature of the relationship between the LXX/OG translations and their parent texts. One perspective locates the typical translated text in a position of subservience to its source. By consistently subjecting the target text to the form of the parent, such a translation attempts to bring its target audience to the source text, rather than the source to the audience.14 As a result, it is argued, the reader may be required in some cases to turn to the parent text in order to puzzle out the meaning intended by the Greek translators.15 In contrast, others regard the typical L X X / OG version as "distinct and independent from its parent text," a translation

way,' Benjamin Wright alleges that they do not always succeed in doing so in actual practice (Wright2008: 111). 11 As Albert Pietersma explains, "It is the aspect of unintelligibility as well as that of intelligibility (even literary beauty) that an explanatory model has to be able to accommodate" (Pietersma 2002: 351). 12 Pietersma 2001: 220; for support he appeals to Conybeare and Stock 1988: 21, who claim that the language (particularly the syntax) of the Septuagint "is so deeply affected by Semitic influence as often to be hardly Greek at all, but rather Hebrew in disguise" (Pietersma 2001: 219). Ziegler similarly observes, "Als Übersetzung wies die LXX von vorneherein verschiedene grammatikalische und lexikalische Eigentümlichkeiten auf, die zwar dem alexandrinischen Juden nicht unverständlich waren, aber doch erst im Verleich mit der hebr. Vorlage richtig erkannt werden konnten" (Ziegler 1934: 175-176). 13 Harl 2001: 187. "The translators use in an intelligent and creative way the syntactic flexibility and lexical richness of the everyday language of their time" (ibid., 188). 14 Pietersma 2001: 219; for this distinction, see Brock 1972: 28; Brock 1984: 73. It is important to recognize that the linguistic competence of the translators need not be at issue here. One can agree with Marguerite Harl that the translators' knowledge of Greek is "sound" (Harl 2001: 187) yet argue that their method of translation led them consistently to adhere to the form of the source text, even at the cost of "good greekness" (so Pietersma 2001: 223). 15 Pietersma claims, "For some essential linguistic information, the parent text needs to be consulted, since the text as we have it cannot stand on its own feet" (Pietersma 2002: 350); similarly, he asserts, "What the Septuagint says, and how it says it, can only be understood in its entirety with the help of the Hebrew" (Pietersma 2001: 220). Harl cautions, however: "The meaning of the Hebrew text of the MT is often obscure. ... For that reason the meaning of [the] Hebrew does not impose itself on us as something evident. It cannot always serve as a criterion to evaluate semantic and lexical divergences of the Greek text" (Harl 2001: 191).

4

Chapter 1: The Problem of Septuagint

Hermeneutics

that aims to bring an interpretation of its source to the target audience.16 A more extreme version of this view argues that in certain instances the interpretive reworking of the source text by the translator is so extensive as to justify speaking of the Greek version as, in some sense, "a new composition."17 A third controversy arises over the proper focus of the modern interpreter's attention. Drawing a sharp distinction between 'text production' (the translator's understanding of the source at the time of translation) and 'text reception' (any later interpretation of the text the translator produced), some take the principal object of study to be the Greek text in its relation to its source.18 Others place the Greek text itself at the center of the investigation. Whether examined from the point of view of the translator, from the perspective of later interpreters, or 'on its own terms,' the translation is interpreted solely within the target language and culture, "sans contamination avec l'hebreu."19 In this approach, the entire Septuagint corpus becomes a primary context for interpretation: "Septuagint Greek (syntax and vocabulary) is clarified essentially by itself, by referring one text to another one. ... The recurrence of some un-greek usages ("hebraisms") makes them more comprehensible due to accumulation of like contexts."20

16

Harl 2001: 185. Speaking on behalf of the translators of La Bible d'Alexandrie, Harl explains, "[Our] approach to the Septuagint is based upon our conception of what it is as a text. We do not ignore its nature as a translation, nevertheless we take it primarily for what it says in Greek. We are convinced that every act of translating results in a text which receives a new life within the domain of the translation language. We acknowledge the fundamental axiom of linguistics: a text written in any language should be read and analysed only in the context of this language" (Harl 2001:184). 17 Van der Kooij 1997: 529. Van der Kooij argues that OG Isa 8:11-16 represents "a free rendering of a whole passage (and not only of single words, or a single clause) which presents itself as a new text with a coherence of its own" (519). Similarly, at Isa 25:1-5, "the passage produced by the translator turns out to be, to some extent, a new text or composition" (van der Kooij 1998: 13; cf. Coste 1954; van der Kooij 2010a). 18 See Pietersma 2008b. 19 Dogniez et al. 2001: 19, cited in Wright 2008: 100. Walser 2008 discusses a number of passages in OG Jeremiah where "the subsequent reader is likely to have understood the text in a different way than the translator of the text" (356). 20 Harl 2001: 186. Harl spells out the consequences for Septuagint lexicography as follows: "The criterion for determining the meaning of words in the LXX is not the meaning of their counterparts in Hebrew. It is their meaning in the Koine, or more precisely, the sense they acquire in the context of the LXX sentences, according to the use the translators make of them, following their choices and habits. The meanings of words are specified by the study of their recurrence in the LXX, within similar contexts. ... The Greek of one passage is explained by the Greek of another. Translation of one book presupposes reference to the entire LXX" (ibid.). The problem with this approach to Septuagint words, according to Pietersma, is that it ignores the role interference from the source plays in the production of the Greek translation. In some cases it is possible to show "that word X was used not because the context of the Greek target text demanded it but because a lexeme of the Hebrew source text suggested

The Shape of the

Controversy

5

Implicit in all of these disputes is the problem of how to identify and evaluate 'interpretation' of the source text in the translation. In attempting to answer this most basic question of Septuagint hermeneutics, one must offer a reasoned account of her position on each of the preceding issues. The identification of 'interpretation' in a translated text simply cannot be disentangled from judgments about the 'fit' of that text within the target culture, on the one hand, and about its relationship as a translation to the source text, on the other.21 Nor can the problem be addressed apart from a clear sense of where the primary interest of one's investigation lies - with the translator's understanding of the source text, as this is manifested in the translation he has produced, or with the rich and varied Nachleben of the translated text among its many readers past and present.22 No one believes that broad generalizations about the translated texts in the Septuagint corpus will resolve these important questions. All sides recognize that there exist 'family resemblances' as well as notable differences among the LXX/OG translations. What Septuagint hermeneutics needs is a theoretical framework and a corresponding methodology for interpreting translated texts that will enable meaningful analysis of both the deep affinities and the significant dissimilarites among the LXX/OG translations. To advance the discussion beyond its current impasse will require methodical study of the individual members of the corpus and careful comparison between them. The present monograph seeks to contribute to this project through a close investigation of the opening vision of the Book of Isaiah (Isaiah 1). My purpose is twofold: (1) to characterize Old Greek (OG) Isaiah as a translation, that is to say, as a re-presentation of its source text in the language of the target culture, and (2) to model an approach to its interpretation appropriate to its character as a translated text.23

it" (Pietersma 2012: 1). Thus, he argues, "It is a basic principle of LXX lexicography that, in order to establish the existence of a new sense of a given [Greekl word, incontrovertible examples of that sense must be found, and one must be able to exclude the source text from being the de facto context" (ibid., 9). On Septuagint lexicography, see further Lee 1969; Kraft 1972; Tov 1976; Lee 1983; Muraoka 1987; Muraoka 1990; Lefebvre 1995; Lee 2003; Muraoka 2008; Joosten and Bons 2011. 21 For a broad array of viewpoints on 'exegesis' in the Septuagint, see the essays in Kraus and Wooden 2006. 22 These stark alternative require further clarification and nuancing. See chapter 2 below on the interaction of translator, text and reader in the production of meaning. 23 This is not to deny that there are other legitimate and fruitful ways to approach a translated text. For example, exploration of the Wirkungsgeschichte of a translation, quite apart from consideration of the original-language text, can offer important insights into the impact of the translated text on the receiving culture. One need only think of the numerous conference papers, essays and books published in 2011 tracing the immense influence the King James Version of the Bible has exercised for four centuries on the shape of the English lan-

6

Chapter 1: The Problem of Septuagint

Hermeneutics

A Theory of Translation It is necessary to begin by spelling out in some detail just what interpretation of OG Isaiah "as a translated text" will entail. This task is made much lighter by the recent publication of Cameron Boyd-Taylor's weighty book, Reading between the Lines.2* Drawing heavily on insights from Descriptive Translation Studies, Boyd-Taylor further refines this descriptive-explanatory framework in a way that holds great promise for the study of the translational literature within the Septuagint corpus. In the following pages, I sketch the outlines of this theoretical approach, demonstrating its capacity not only to describe a wide range of translations but also to characterize the differences among them. This then leads to a discussion of the long-standing debate over the particular character of OG Isaiah as a translated text and a proposal for moving the conversation forward. Describing Translation: Function, Process,

Product

As developed by Gideon Toury, Descriptive Translation Studies (DTS) conceives of translation as an event25 within the literary system of the target culture: Toury identifies three interdependent dimensions of translation, 1) the position or function of the text within the target culture (function); 2) the process through which it is derived from the parent (process); and 3) the textual linguistic make-up of the product (product). Any fully adequate descriptive-explanatory study of a translation will attempt to account for the observed interrelationship of these variables. 26

guage and its literature. In a number of previous publications, I have investigated the significance of the Greek version of Isaiah - as a scriptural text in its own right - for Paul and other early Christian writers (Wagner 2002; Wagner 2005; Wagner 2006; Wagner 2008). On the importance of reception-history for Septuagint studies, see further Meiser 2012. 24 Boyd-Taylor 2011, incorporating insights from a number of earlier studies, including Boyd-Taylor 1998, Boyd-Taylor 2001, Boyd-Taylor 2004, Boyd-Taylor 2006a, Boyd-Taylor 2006b. Although the book is subtitled, "The Interlinear Paradigm for Septuagint Studies," it is possible to adopt Boyd-Taylor's theoretical framework without accepting his contention that the interlinear paradigm most adequately accounts for the character of the 'typical' translation in the Septuagint corpus. His careful and learned study does go a long way toward making the case for the explanatory power and empirical adequacy of the paradigm, however. Many of Boyd-Taylor's ideas have been worked out in conversation with Albert Pietersma, whose own publications have contributed significantly to the emergence and development of the interlinear paradigm. See especially Pietersma 2001; Pietersma 2002; Pietersma 2004; Pietersma and Wright 2007a; Pietersma 2010. 25 Toury 1999; 18. 26 Boyd-Taylor 2011: 39, drawing on Toury 1995: 11-14. See also Toury 1978; Toury 1999. A brief critical introduction to DTS is provided by Munday 2008: 107-123.

A Theory of Translation: Function, Process,

Product

1

With its focus on the prospective function of a translated text within the target culture, Toury's model recognizes that translation takes place in a social context. Shared cultural expectations regarding aims and methods both guide and constrain the translator's work.27 To become a translator is to assume a social role, that is, "to fulfill a function allotted by a community - to the activity, its practitioners and/or their products - in a way that is deemed appropriate in its own terms of reference."28 Thus, the role or function envisioned for the translation in the target culture will prove a strong governing factor in its surface realization or textual-linguistic make-up [product]. The translator will aim at producing a text with the make-up requisite to its intended location, and will thus be working from a sort of paradigm. This in turn will govern the relationship between the target text and its source. For it is with reference to such a paradigm that the translator will select the linguistic strategies by which the translation is produced [process]. In this way, the process of translation is itself conditioned by the prospective function of the product. 29

The 'paradigm' or 'model of translation' from which a translator takes his or her bearings can be conceptualized as a set of socially constructed 'norms.'30 "Insofar as translators make decisions that are open to assessment within the target culture, and hence are liable to sanctions of one sort or another (e.g. the praise or blame implicit in how the translation is received), their work is governed in some manner by norms."31 These "prescriptions, proscriptions, preferences and permissions" are neither static nor absolute;, sitting somewhere between 'rules' and the 'idiosyncracies' of the translator, norms reflect social conventions that are themselves fluid and changeable, and they constitute a spectrum of negotiable behaviors.32 As Theo Hermans explains, "A norms-

27 "Given the translators' participation in the literate culture of a certain time and place, we would expect them to have internalized assumptions both as to what was expected of them and how best to proceed" (Boyd-Taylor 2011: 34; see further Hermans 1997). 28 Toury 1995: 53, cited in Boyd-Taylor 2011: 57. 29 Boyd-Taylor 2011:56. 30 For a range of perspectives on the meaning and utility of 'norm' as a concept, see Schaffner 1999. Theo Hermans (Hermans 1999: 50) notes that norms are "legitimated] in terms of institutional values" and handed on "through precept and example," i.e., through processes of socialization into a particular group and activity (cf. Toury 1999: 16). 31 Boyd-Taylor 2011: 66. Toury speaks of three types of norms: a 'basic' or 'primary norm' governs behavior that is "more or less mandatory for all instances of a certain phenomenon"; a 'secondary norm' or 'tendency' represents "common, but not mandatory" behavior; and what we might call a 'tertiary norm,' which Toury describes as "other tolerated (permitted) behaviour." According to this schema, "the more frequent a phenomenon ... the more it is likely to represent (in this order) a more permitted (tolerated) activity, a more decisive tendency, a more basic norm" (Toury 1978: 95). 32 Hermans 1999: 50. According to Hermans, "a culture's value system together with the norm complexes which serve to hold it in its place, see to it that translation is governed by at least three normative levels: general cultural and ideological norms which may be held to ap-

8

Chapter 1: The Problem of Septuagint

Hermeneutics

based approach to translation starts from the assumption that the translation process involves decision-making on the part of the translator,"33 a process of selecting one option from among "a limited range of realistically available alternatives."34 Thus, the investigator asks questions "not only about what is there on the page but also about what might have been there but, for one reason or another, is not."35 The concept of translational norms allows one to describe in a systematic way the principles guiding the translator as he navigates the challenges of re-presenting the source text in a form that will be 'acceptable' to the target culture (or a particular sub-culture thereof) as a translation.36 The 'Acceptability' of a Translation in the Target Culture The qualification in the previous sentence, "acceptable as a translation," must not be overlooked. 37 Within a given culture or community, the standards of

ply throughout the larger part of a community; translational norms arising from general concepts of translatability and cross-linguistic representation alive in that community; and the textual and other appropriateness norms which prevail in the particular client system for which individual translations cater" (ibid., 59). 33 Hermans 1999: 52. 34 Hermans 1999: 57. "[A]t every point in the life of a societal group, especially a complex and/or heterogeneous one, there tends to be more than one norm with respect to any behavioural dimension. Consequently, the need to choose between alternative modes of behaviour tends to be built into the very system, so that socialisation as concerns translating often includes acquisition of the ability to manoeuvre efficiently between the alternatives" (Toury 1999: 27). 35 Hermans 1999: 57. Toury acknowledges that norms must be extraploated from regularities of behavior that we assume to be governed by "recurrent underlying motives" (Toury 1999: 16). "The texts present us with the results of actual norm-regulated behaviour, that is, with a primary product of their activity, out of which the norms themselves are to be (and can be) reconstructed (Toury 1978: 91; emphasis original). Thus, "for the researcher, norms ... emerge as explanatory hypotheses" (Toury 1999: 16; emphasis original). 36 The notion of 'acceptability' plays a key role both in DTS and in Boyd-Taylor's adaptation of the approach. Toury defines acceptability as conformity "to the norms active in the target culture, or in that section of it which would host the end product" of translation (Toury 1995: 56; cf. Boyd-Taylor 2011: 58-59, 68-71). With respect to the translation process, it is useful to distinguish between two types of norms, 'regulative' and 'constitutive.' As BoydTaylor (2011: 71) explains, the former "guide the translator's selection from his repertoire of strategies so that he may solve the problems he encounters in [a culturally] acceptable way"; the latter "reflect what a particular culture (or community) accepts as a translation (as opposed to e.g. an adaptation)." In distinction from 'norm,' then, the term 'strategy' describes the means by which a translator 'solves' a perceived 'problem' in replacing a particular feature of the source text with a target-language equivalent (cf. Toury 1999: 28). 37 Andrew Chesterman develops the notion of 'product (or expectancy) norms,' which "are established by the expectations of readers of a translation (of a given type) concerning what a translation (of this type) should be like" (Chesterman 1997; 64).

A Theory of Translation:

'Acceptability'

9

'acceptability' for a translation may not be the same as those for a literary work composed in the target language. Toury observes, "There are often good reasons to regard translations as constituting a special system (Dressier 1972), or 'genre' (James 1989:35-36) of their own within a culture."38 This is because translations tend to deviate from [the] sanctioned patterns [of the target culture], on one level or another, not least because of the postulate of retaining invariant at least some features of the source text - which seems to be part of any culture-internal notion of translation .... This tendency often renders translations quite different from non-translational texts, and not necessarily as a mere production mishap either; it is not unusual for a certain amount of deviance to be regarded not only as justifiable, or even acceptable, but as actually preferable to complete normality, on all levels at once. Moreover, even if they are not culturally favoured, deviations even when they manifest themselves in the very make-up of the texts - do not necessarily disturb the 'persons-in-the-culture.'39

Such is likely to be the case especially where the community places a high value on the perceived fidelity of the translation to its source. As Boyd-Taylor notes, "To the extent that its location [in the literary system of the target culture] is central, a translation will tend to break with target models and adhere instead to the textual relations of its source."40 We will return to this point in the following chapter as we consider the question of the 'target culture' for OG Isaiah and its prospective location within that system.41 With this important caveat, target models do serve as key points of reference for determining the 'fit' of a translated text within the overall literary system of the target culture.42 In Toury's schema, such an assessment takes place at three distinct levels of the discourse hierarchy: the linguistic, the textual and the literary.43 At the linguistic level, one considers the extent to which the equivalents chosen by the translator adhere to the grammatical and syntactical conventions of the target language. Interference from the source language at this level may take the form of 'negative transfer,' in which the translator chooses to represent a feature of the source text in her translation in a way that contravenes the norms of the target language.44 Conversely, interference may be felt

38

Toury 1995: 28. Ibid. 40 Boyd-Taylor 2011: 64, summarizing the findings of Even-Zohar 1978. 41 See pp. 56-62 below. 42 Boyd-Taylor 2011: 59. 43 Toury 1995: 170-171. 44 For the terminology see Boyd-Taylor 2011: 58-59, following Toury 1995: 275. James Barr (1979: 293) draws attention to an example of negative transfer in the rendering of the Hebrew particle 'a ('please') as ev e^oi in Judges 6:15; 13:8; 1 Reigns 1:26; 3 Reigns 3:17, 26. Contrast the idiomatic translation of 'a with 5eo|xai/8eo|xe9a in the Pentateuch (e.g., Gen 43:20; 44:18; Exod4:10, 13). 39

10

Chapter 1: The Problem of Septuagint

Hermeneutics

through 'positive transfer.' In this case, the translator represents a particular feature of the source text with a grammatical or syntactical construction native to the target language; due to the influence of the source text, however, the translation ends up showing a markedly higher concentration of this construction than one finds in texts composed in the target language.45 At the textual level, the goal is to assess the extent to which the translation conforms to the target culture's expectations of a well-formed text. Here we ask whether "the translation is now not only grammatical but recognisable as a particular type of discourse in the target language."46 Attention focuses not only on the degree of thematic coherence evident in the translation, but also on the cohesiveness of its discourse.47 Evaluating a translation at the literary level requires us to appraise its adherence to "the rhetorical and stylistic conventions" of the target language as well as to "the norms governing such phenomena as intertextuality and cultural referencing."48 A given translation may compare favorably to target models in one respect - for example, linguistic well-formedness - and yet fall short in another, such as textual cohesion. 49 Conversely, a translation that shows a high degree of interference from the source at the linguistic level may yet find ways to engage the cultural and literary traditions of the target system with a fair degree of sophistication.50

45 Georg Walser has observed, for example, that in the Pentateuch and in other narrative texts within the Septuagint corpus, predicative aorist participles uniformly "precede their main verb, and all words between the participle and the main verb belong to the participle" (Walser 2001a: 509). As Walser notes, "This word order is of course not alien to other varieties of Greek; what is strange is that [in the Pentateuch] we find only this word order and no other" (ibid., 503). See further Walser 2001; Walser 2008a. Similarly, Trevor Evans attributes the high ratio of volitive to potential optatives in the Greek Pentateuch to interference from the source language: The volitive optative "is always employed idiomatically and can be regarded as exhibiting bilingual interference only in terms of its frequency comparative to the optative's other functions" (T. Evans 2001: 175-197, here 197). 46

Boyd-Taylor 2011: 59. Textual 'cohesion' refers to "the way the text hangs together lexically, including the use of pronouns, ellipsis, collocation, repetition, etc." (Munday 2008: 91). "Languages differ considerably with respect to the degree to which they permit or oblige their users to connect text-units in sequence by means of explicit indications of cohesion" (Lyons 1995: 264). 48 Boyd-Taylor 2011: 59. 49 Summarizing the results of his study of OG Gen 11:1-9, Boyd-Taylor comments, "The conditions for acceptability under which the translator produced his text were not those underlying literary composition within the target culture; rather, insofar as his text was acceptable as a cultural product it met a very different sort of criterion. While adhering to the requirements of grammaticality, the translation appears decidedly isomorphic to the source text in its textual linguistic make-up" (Boyd-Taylor 2011: 308). 50 So, for example, Aquila's sophisticated Greek vocabulary suggests that he was learned and well read, even though the syntax of his translation shows strong interference from the Hebrew source. See pp. 12-16 below. 47

Classifying Translations:

Overview

11

Classifying Translations Overview To delineate the interrelatedness of function, process and product with respect to a particular translation is to specify what Boyd-Taylor and Pietersma have come to call the 'constitutive character' of the translated text.51 In the case of ancient translations, such as those in the Septuagint corpus, however, we have very little external evidence to help us reconstruct the aims and methods of the translators, including the 'slot' they intended the translation to occupy in the target culture.52 Study of other translations - particularly other ancient versions of biblical texts, whether in Greek (Aquila, Symmachus, Theodotion) or in other languages (Targum, Peshitta, Vulgate) - provides an important reference point both for mapping the strategies and norms [process] of the LXX/ OG translators and for assessing the results of their efforts [product].53 Nevertheless, it is the textual-linguistic character of the translated text itself that constitutes the primary evidence for the function, process and product of the translation. In this sense, the constitutive character of a translation is, as Pietersma has put it, "embedded in the text."54

51

Boyd-Taylor 2011: 39-40; Pietersma 2005a: 446. 'Constitutive character' is probably best understood as "a descriptive stance" (Boyd-Taylor 2011: 53). "It is a perspective one adopts in order to conceptualize the [translated] text as a specific object of study. ... To approach a text in terms of its constitutive character is to view it as the product of a socially intelligible undertaking, and thus against the backround of 'human customs and institutions' (Wittgenstein). In this respect, the notion is inherently teleological; it will give rise to explanations which locate the translation within a larger context of goal-oriented behaviour" (ibid.). 52 On the debate over the value of the Letter of Aristeas as evidence for Septuagint origins, see the balanced treatment of Rajak 2009: 24-91. DTS treats extra-textual statments about normative principles - including the translator's own statements regarding purpose or method - as secondary evidence for reconstructing translational norms. "[Normative pronouncements] are partial and biased, and should therefore be treated with every possible circumspection; all the more so since - emanating from interested parties - they are likely to lean toward propaganda and persuasion" (Toury 1995: 65). 53 "Toury suggests that the most illuminating approach [to the task of reconstructing the norms underlying a translation] is the comparative one. This would involve, preeminently, the comparison of translations with their sources, but also the comparison of different translations of the same source, and different recensions of the same translation (Boyd-Taylor 2011: 61). 54 Pietersma 2005a: 446. Cf. Boyd-Taylor 2011: 40: "From the verbal make-up of a translation one can draw inferences regarding the specific relationship between function, process and product inherent to it."

12

Chapter 1: The Problem of Septuagint

Hermeneutics

In Search of the 'Typical' Septuagint Translation: Aquila Reading between the Lines develops an extended argument for the 'interlinear paradigm,' 55 theorizing that "the (prospective) systemic position and function of the typical Septuagintal translation was conceived of as the Greek half of a Greek-Hebrew diglot."56 The prototypical example of such an 'interlinear' translation is Aquila's version, which represents "a systematic attempt to put the Greek language entirely in the service of the source text." In this "programmatically isomorphic" translation, "the formal features of [Hebrew] words and their [syntactical] relations are consistently mirrored in the Greek."57 Aquila regularly attempts to represent in translation each word and not infrequently each morpheme - of the parent text (quantitative fidelity) as well as to follow the word order of the source (serial fidelity), even when doing so requires the suspension of target-language norms. In addition, Aquila tends to maintain morphosyntactical correspondence between the translated text and its source; that is, the translator typically represents a participle by a participle, a prepositional phrase by a prepositional phrase, and so on.58 At the word level, Aquila's translation displays a strong preference both for lexical stock-pairing, the rendering of Hebrew words with their conventional Greek glosses,59 and for lexical correspondence, where a particular Greek term is consistently (and often exclusively) used to translate a single Hebrew lexeme.60 At the same time, Aquila displays an impressive range in his vocabulary, drawing freely on classical sources such as Homer and Herodotus.61 All 55

See n. 24 above. Boyd-Taylor 2011: 101. Proponents of the theory insist that interlinearity is merely a metaphor; there need never have been actual Greek-Hebrew texts laid out in the form of an interlinear or of a diglot (ibid., 100). Nevertheless, suggestive ancient analogies can be found; see Brock 1972: 29-31. 57 Boyd-Taylor 2011: 171-172. 58 None of these translational norms is inviolable, of course. So, for example, Aquila may, in deference to Greek usage, choose a singular Greek noun to represent a plural noun in the source or vice versa (Reider 1916: 36). 59 "The essence of stock pairing is its conventional nature, i.e. well established glosses are regularly selected as matches" (Boyd-Taylor 2011: 123 n. 24). Such glosses trade on the unmarked or 'stock' meanings of their respective Hebrew terms. 60 This feature of Aquila's version can easily be seen by perusing the index prepared by Joseph Reider and Nigel Turner (Reider and Turner 1966). Nevertheless, as James Barr notes, "If the mass of our evidence is reliable, Aquila did not press the stereotyping tendency very far and on very many words did not achieve a one-to-one relationship between Hebrew and Greek, in either direction" (Barr 1979: 312). Instead, as Reider claims, Aquila sought to "minimize the number of synonymous equivalencies" (Reider 1916: 26). I owe these references to Boyd-Taylor 2011: 126. 61 Reider 1916: 18. Although Aquila is often suspected of creating neologisms, one cannot be sure that he did not obtain his hapax legomena "from some nook or corner which our sources fail to lay bare" (Reider 1916: 33). 56

Classifying Translations:

13

Aquila

this suggests that Aquila's normal unit of replacement is the word. 62 This atomistic approach to the process of translation frequently leads to semantic oddities (such as collocations of words that are atypical for the target language) as well as to ambiguities and, occasionally, to apparent nonsense. 63 A brief example from 3 Reigns 21:7 and 11-12 illustrates many of these features of the translation (see Fig. I). 64 Figure 1:3 Reigns 21:7b Aquila (!1T, 1 Kings b

B p a o NT r u n "O MIL fcU I N 'Bab ^ S n*7!D

[Kaieijiev] YVCOTE 8fi Kod i S e x e o i l K a K i a v o w o i ; t/nxei,

c

OTI djcEOTEiXev itpoi; | i e ei6po :rnirr 'D^a

in OG Isa

Paoiteia

O^iou K a i Iaxi0a(i K a i A^a^ K a i E ^ e k i o v o'v epaai/xuaav xf|q louSaiai;

Whereas the Hebrew text sets Isaiah's vision in the "days" ("iTD) of Uzziah, Jotham, Ahaz and Hezekiah, the Greek version speaks more precisely of their "reign" (ev PacnAeia).13 This move finds its complement in G's rendering of "O^O, pointed by HI as the noun "Obi?, 'kings.' 14 G translates with the relative clause, "who reigned" (o'i epacriAeuaav), thus naming the activity associated with the role.15 The significance of these subtle transformations comes fully into view only retrospectively. Considered in light of the distinctive themes that emerge from the translator's handling of the chapter as a whole, they

12 In this and similar figures I give fU (according to BHS) and OG (according to Ziegler 1939) in parallel columns. Layout and punctuation of the Greek text is based on my own analysis of the syntax; the Hebrew is lined up to match the Greek, even where the Masoretic accents indicate a different segmentation of the text. Because the shape of the translator's Vorlage is to be determined through investigation of his Ubersetzungsweise (see pp. 46-52 above), I have not attempted to represent it through retroversion. Instead, I have included the vowels and accents of IK along with the consonantal text so that it will be possible to compare one ancient interpretation of the Hebrew text with that offered by the Old Greek. 13 Goshen-Gottstein 1995: 1 notes a parallel equivalency in Gen 14:1, where ev xrj pacnXeia tfi Ap.apa.A represents Contrast OG Isa 7:1, where G offers ev xaiq rpEpaii; A^ai^ for Trtx Predictably, a ' a ' 0' read [ev xaTi;] i p e p a i ^ at Isa 1:1. Like other books in the Septuagint corpus, G transliterates Ioxx9a|i and A ^ a ^ but employs the Graecized forms Ofyaq and E£ekux|j(XTfi>v, 'practices.' Compare Isa 3:10, where G renders DrrbbBQ '"id straightforwardly as xà yevrmaxa xwv ëpyœv aùxwv. In 3:8, (iExà àvo)itaç appears to represent DirbbBD, understood as nnb *7BD (Seeligmann 1948: 54; cf. Wilk 2003: 24 n. 31c). It is possible that G read DD-biJQ in Isa 1:16 (cf. Koenig 1982: 110; Goshen-Gottstein 1995: 3 notes that this variant appears in one manuscript collated by Kennicott); he would then have rendered the pronomial suffix expansively as yuxoi Ù|ig>v (cf. Ezek 18:31, where DD'^BQ = ana èauxiûv). Ottley's hypothesis (1906: 107; so also Seeligmann 1948: 54; Wilk 2003: 24 n. 31c) that G somewhat creatively analyzed n a - ^ r n as a ^ b bun deserves consideration as well (cf. Prov 6:21; 2 Chron 7:11). Compare Isa 59:13, where G renders as mtô KapSiaç fuimv, supplying the pronoun from the context (for yux 1 ! = 3*7,

Appeal to Israel: Debilitated

Body, Devastated

Land

(1:5-7)

89

from one's deeds to one's soul in OG Isa 1:16c corresponds to the interiorization of Israel's plight effected by G's bold transmutation of the figure in 1:6b.90 Israel's ailment lies beyond the reach of any human leechcraft; only the Kyrios can heal their wounds.91 The Greek translator's rendering of v. 6c vividly illustrates the complex set of negotiations required - at every level of the discourse hierarchy - to reconcile the claims of the Vorlage with the potentialities of the target language in order to produce a translation that faithfully conveys the meaning of its source and yet communicates lucidly and powerfully to its Greek-speaking audience. On the one hand, G replicates the overarching form of v. 6c, producing a sentence comprising three coordinate members, each introduced with a negative particle.92 Like his source, the translator describes a medical protocol for treating wounds. On the other hand, he reconfigures the individual elements of v. 6c in accordance with the larger conception he has formed of the image as a whole.93 Elsewhere in Isaiah, G interprets the root tZQn (v. 6c|3) as a general reference to healing, and he may so have understood the verb here.94 This verb, together with the references to olive oil (}DB) and softening (V"p~i) in v. 6cy, would have led G to understand the prophet to be speaking in v. 6c

see Isa 7:2 [bis], 4; 10:7; 13:7; 24:7; 33:18; 42:25. In 44:19 we find either a double translation of by G [first as KapSia, then as y v / r \ I or a doublet that arose during the transmission of the Greek version). 90 The Vulgate stands alone among ancient versions in offering a translation of 1:16c that approaches the sense of the Old Greek: auferte malum cogitationum vestrarum ab oculis meis. In Isa 1:16, however, Jerome's choice of cogitationes for c ^ b r c was likely influenced by OG Isaiah. His commentary on the verse recalls the diction of Isa 1:16 in one strand of the Old Latin: "... ut qui Christus baptisma consecutus est. auferat malum de corde suo" (In Isaiam 1.22; Gryson 1993: 169). 91 Similarly, Ps-Basil: o Xoyoq xi T|ai Ttepi xdiv oXatv 8i' xfi ctaQeveia Kaxei^rinL^evwv; ... ETteiSfi \mep xpaiinnaxa. OTJ Xpeia, (|>r|ai, xfi0T|aea9c(i ¡Uryovxai (Eusebius, In Isaiam 13; Ziegler 1975: 7). This intratextual link appears also in D, which translates both 2VJ and T P as derelinquo. In contrast, C and 5 employ different verbs in v. 8 f r o m those selected for v. 4. G ' s choice of eyKaxoiAEUtetv further strengthens intratextual echoes between 1:8 and later prophecies of the reversal of Z i o n ' s abandonment in 60:15 and 62:12 (in both instances, eyKaxaXei7teiv translates V^TSI, as it does in 1:4). 112 The notion that Israel's punishment is commensurate with their sin is not foreign to Hebrew Isaiah, of course (see below on vv. 2 9 - 3 0 , for example), but at a number of points (as here) the Greek translator appears to emphasize the fittingness and proportionality of the L o r d ' s judgment on his disobedient people. A much-studied instance of this is G ' s interpretation of Isa 6:9-10, which frames the prophet's call to 'hear and see without perceiving' as a response to the people's prior indifference to and, indeed, rejection of (xoup(xi|i xoù Mavaocrri, òxi &(ia TtoXiopKriaouaiv xòv IouSav. There is no reason to believe that G ' s Vorlage carried a text different from that uniformly attested by m, lQIsa a and other witnesses in Isa 9:20: "both of them together against Judah" (nnn , " l 7JJ n a n HIT). Rather, the translator fleshes out the hostile sense of the preposition bo in 9:20 with the verb icoXiopKelv. As he does so, he draws not on the immediate context, which speaks of eating and devouring, but on the wider setting of the Syro-Ephraimitic crisis that has been in view since 7:1. The Isaiah Targum similarly looks to the narrative of Isa 7 to interpret 9:20, expanding the latter with the verbal phrase, "they made an alliance to come against [Judah]" (WD1? ]l"nnrr), inspired by the T a r g u m ' s own rendering of 7:2 (KD^n " n n n x binàri iobn D"1X"T). A common tradition may lie behind O G Isa and C here, but it is just as likely that the correspondence reflects their similar exegetical approaches, in which one part of the book may be employed to illuminate another. 118 By representing the Syro-Ephraimitic crisis as involving an attempted siege of Jerusalem, G heightens the threat to the Davidic line in Isa 7 - 9 . He further tightens the parallelism between Ahaz and his descendant Hezekiah, during whose reign Jerusalem will again face the prospect of a siege, this time at the hands of the Assryrians. On the importance given to the contrast between Ahaz and Hezekiah in the final form of Hebrew Isaiah, see Ackroyd 1982; Sweeney 1988: 12-17; Seitz 1991; Seitz 1993: 6 0 - 7 1 .

On oapottoG, see p. 166 below.

Israel 's Response: Lament and Hope

(1:8-9)

97

sets the themes of judgment and restoration in counterpoint.12" The translator tightens the parallelism of his source a touch in v. 9b by constructing both comparisons with (bq, rather than representing the b in his Vorlage (v. 9bP) with a dative. This adjustment has the further effect of bringing the syntax of v. 9b into alignment with that of v. 8ba-p. 12 ' In the remainder of the sentence, G capitalizes on the link he has already created between v. 4 and v. 8 via eYKatoc^eiTieiv. His rendering of "P~1D by GTtEpiia in v. 9122 recalls the accusation in v. 4 that God's children are ojiepfioc 7iovT|p6v who have abandoned (eyKaxa^eiTteiv) their Lord and father.123 At the same time, however, 'offspring' evokes the covenant promises made to Abraham, Isaac and Jacob, and to their cmep|!a after them - commitments to which the Kyrios has freely pledged himself. 124 Although the Kyrios has quite fittingly responded to betrayal by forsaking daughter Zion, he has nevertheless graciously preserved a future for his people, in faithfulness to his own character.125 And so, despite the somber undertones sounded by the synkrisis

120 The switch to an aorist verb in v. 9a directly after the future in v. 8a is not as jarring as Wilk maintains (2003: 33-34). Just as the future fits the imagery of imminent destruction developed in v. 7 (see n. 110 above), so the perfective aspect of the aorist suits the perspective of the speakers, who have thus far survived the judgment whose unfolding is described in v. 7. Their statement expresses hope in the Lord's fixed purpose to preserve his people. As the form of address in v. 10 indicates, the prophet does not offer unqualified support to this point of view. 121 The selection of o*; Tonoppa instead of xo> Fo|ioppa. represents a genuine choice on the part of the translator, as his employment of ojioiotiv + dative elsewhere attests (40:18, 25; 46:5). While C and D, like G, choose the same construction for the comparison in both lines (with 3 and with quasi, respectively), 5 adheres to the syntactical pattern of the Hebrew exactly: r^icrauAo p o i m vypC. The k«i in v. 9bp likely represents a Vorlage reading (or taken to read) ... r n a c ^ l T T ... . Though neither !R nor lQIsa a has the conjunction, it is found in many Masoretic MSS as well as in E, 5 and D. 122 The Hebrew term occurs only here in Isaiah. Cf. Deut 3:3, the only other occurrence of this equivalency in the Greek Bible, where eax; Toft jjti KaTaA.i7te.tv «ijxoij artep^a translates TIE iV^xrn (Ziegler 1934: 106 also notes this possible intertextual link). In Isa 1:9 a' offers M'.Ifipti, a word belonging to the semantic field from which the L X X / O G translators normally select equivalents for T I E (see Muraoka 2010: 356-57). 123 So also Eusebius, in Isaiam 14 (Ziegler 1975: 7). While the oldest form of S reads rix-ii» ,A iiwK' (Brock 1987), the variant in MS 916 rd^ii ,A xl=h. replicates both the noun and the verb n-rr, from v. 4 (re.i=*A v o i u m i . ... re^-u), suggesting that later tradents perceived the intratextual link betwen v. 10 and v. 4 in the Greek version and sought to reproduce it in Syriac. Note also the rendering of D, semen, reflecting the influence of the Greek via the Old Latin translation. 124 See, e.g., Gen 12:7; 13:15-16; 15:5, 18; 17:7-8, 19; 22:17-18; 24:7; 26:3^1, 24; 28:13-14; 32:12; 35:12; 48:4; Exod 32:13; 33:1; Deut 1:8; 4:37; 10:15; 11:9; 34:4. 125 Clements comments, "The purpose of [1:91 is clearly to offer some element of alleviation of the preceding threat (vv. 5 - 8 ) and to suggest the idea of a remnant through whom the future would be secured" (Clements 1980: 425). For the preservation of "seed" as a sign of

98

Chapter 3: Translation

and Interpretation

in OG Isa 1 :l-20

with Sodom and Gomorrah, Israel's lament modulates for a brief moment into a more hopeful key. For the first time in Isaiah's dark vision, there is the faintest flicker of a promise. Jerusalem has not yet shared the fate of the Cities of the Plain.126 They have, in truth, arrived at the very brink of destruction. But because of the Lord's mercy, barren Zion may yet send out shoots that will bear fruit once more.127 Second Address: Give Heed to God's Law

(1:10-20)

Summons to Leaders and People (1:10) This brief flash of hope quickly fades, however, as the prophet deftly parries Israel's lament, turning the pathos of their reference to Sodom and Gomorrah against them with a stinging riposte: "Hear the word of the Kyrios, you rulers of Sodom! Give heed to the law of God, you people of Gomorrah!" (v. 10; see Fig. 10).128 Isaiah daringly applies the names 'Sodom' and 'Gomorrah' to his audience in an attempt to jolt them into awareness of the true magnitude of their rebellion against the Kyrios.129 Just beneath the surface of his bold

God's unswerving commitment to Israel, even in the midst of exile, see Isa 41:8-10; 43:5; 44:2-3; 45:25; 65:9, 22. On the development of this theme in Second Temple literature, see Wagner 2002: 112-115. 126 With the particle av, the translator adeptly represents the sense of the Hebrew construction with (cf. Gen 31:42; 43:10; Judg 14:18; OG Ps 118:92; 123:1-3; contrast 1 Reigns 25:34; 2 Reigns 2:27; OG Ps 93:17). 127 See, e.g., Isa 49:18-21; 54:1-3; 62:1-5. The absence of a lexical equivalent in OG Isaiah for E0i)D3 (attested by ITT and lQIsa 3 ) could be a consequence of haplography in the Vorlage or of parablepsis on the part of the translator due to homoiarchton (D7CO CDJJOD). The fact that ç in Isa 1:10 as a reference to the Pentateuch fits exceedingly well with the following speech (Isa 1:11-20), whose central concerns are for right worship and social justice (just these elements are emphasized in the roughly contemporaneous account of the vo^oç in Aristeas\ see Feldmeier 1994; Zuntz 1959). In any case, we must not make the mistake of assuming that G would have attributed to m m in Isa 1:10 the meaning that modern critics identify as its 'original' sense in this context (for the latter, see Williamson 2006: 85-87; Jensen 1973; Fischer 1995; Sweeney 1996b).

Complaint: Barren Worship, Bloodstained Hands

(1:11-15)

103

Complaint: Barren Worship, Bloodstained Hands (1:11-15) What, then, does it mean to "attend to God's Law"? This is the burden of the Lord's second address. The Kyrios first takes the people to task for flagrantly disregarding what the divine Law demands (vv. 11-15). He follows this with a clarion call to national repentance, manifested by the renewed practice of social justice (vv. 16-17). The speech concludes with an offer of amnesty to those who will hearken - and a grim word of warning to those who refuse (vv. 18-20). Figure 11: Isaiah

1:11-15

nirv "lOS"1 bDTQnri '^"nab

lla

zi not jtXfjQcx; xwv Ououov tj^wv Xeyei ropioc;

n,1rx rrfti; n,8,HD a'pni fflWBl D'teDl D,-1S DT) ^RSBn iib rrisn1? lion "3 D3"ra riitT Di?3",,a

c

12a b

Kai cxeap apvwv Kai al|ia xaiiprav Kai xpaycov o-u poi)Xo(jai ov>5' av epXTicsGe 60f|vai noi ziq yap e!;e£f]xr|csev xoruxa ek xwv xeipcov ij(i(Bv

«•'Din Hb 13a ^"ixn Obi

c

~nran ¡rnrt iOtS

13b

¡Til rQBin rnipp

0

Ouniajia pSeXvyna noi eoxiv

Bin haEI IfipD bpwsb

d

xac; vonnriviai; \)(i(ov Kai xa oappaxa Kai ftnepav neyodr|v o\>k avexo|iai

e

vpaxeiav Kai apyiav . ~ Kai. xaq vov|xr]via5 v|icov Kai xa"ns'p

D , 1 RK NI1?!) TUJ3ÌD

•'Jena n'pm D-na nil

xi )j.oi 7iXfj0O(; tfflv OuaiMv

IJ^MV

Xéyei Kiipioq

TtXfipriq cip! 0^OKaiix0)|iaT0)v KpiCDv Kav axeap apvcov Kai aijia taiipcov Kaì xpàycov

nniFioi :"Tissn vfr 'is nim1? Miin -o

oi!) ßoV)A.opai oi!)8' av epxnaöe ò6fivai |ioi

D 3 T D FLXT BJ?3 _ , Q

xinó>v

G selects equivalents for m (v. 1 la) and Tlinto (v. 1 lb) - 7i^fj6oq and Trvaprn;, respectively - whose semantic ranges overlap with those of the corresponding Hebrew lexemes. In one sense, then, we find little 'difference' between the translation and the source text at this point. But it would be premature to conclude that there is little of significance here for understanding the Ubersetzungsweise of the Greek translator. As we follow the contours of the text that he has produced, it becomes increasingly apparent that G's choice of these particular terms (7cX.fi0oq /TrA,r\pr|q) sets up a play on words highlighting the stark contrast in vv. 11-15 between the people's attitude toward sacrifice and the Lord's. Paronomasia with the Greek root *TIA.T|- again features prominently at the close of this section of the speech (vv. 14b-15c). It thus appears that G has all of vv. 11-15 in view when he translates v. l i b , for in choosing 162 In the Greek text, 'you' last appeared as subject in 13b. Nevertheless, the intervening three clauses (13c, 13d, 13e-14a) have expressed 'your' agency indirectly, while still maintaining the contrast between 'your' [the people's] actions and 'my' [the Lord's] response.

108

Chapter 3: Translation and Interpretation in OG lsa 1:1-20

7iA,ripriq he adeptly fashions an inclusio that captures the central message of the passage it encloses: "I am full of your sacrifices and supplications (jtX,f|pr||J.8' âv Kân\|/T|ç cbç KpÎKov xôv ipâxri^ôv aou Kai GOIKKOV Kai OTTOÔÔV imoaxpoxrri, o\)S' oikcoç KaXéaexe vr|axeiav 8EKTT|V. Compared to simple coordination, the concessive construction lends greater intensity and movement to the oracle (cf. Smyth §§ 2372, 2381). 172 See, e.g., OG Ps 41:3; 62:3; 1 Reigns 1:22; cf. Deut 31:11. 173 Exodus 23:15, 17; 34:23 (cf. v. 20); Deut 16:16; cf. Sir 35:6. So Kimhi: "This appearance in the Temple [occurs] during the three pilgrimmage festivals" (¡"Pinn If ClL>;~ tD*7B3). 174 The sole Masoretic manuscript that vocalizes the form as a qal belongs to the fourteenth century; another ancient scribe later repointed the form as a niphal (De Rossi 1969, 3:1; cf. Barthélémy 1986: 4). 1 interprets the verb as passive in voice (nxmnK1?; cf. D, cum ueneritis ante conspectum meum). S takes it as active (rc*u»=A), as it does at Ps 42:3 (rc'vurc'a), even though, as Weitzman notes (1999: 29), the expression 'see God's face' is "more often replaced (as in m) by 'appearing before God' (e.g. Exod 34:23, Deut 16:16)." 175 In Exodus 33:20, the Kyrios tells Moses, oij 8WT|CTT] iSeîv U M I xö rcpoaomov- ov yàp Ht) ï5r| avÔpGmoç xô jcpôccojcov nou Kai Çr|aexat. Cf. Isa 6:5.

110

Chapter 3: Translation and Interpretation in OG Isa 1:1-20

this fundamental tenet has left its mark elsewhere in the Greek translations of Israel's scriptures.'76 We should not read too much into G's decision to render the figurative "OS with the more pedestrian (though idiomatic) |ioi, however. 177 Given his retention of o^Gafyioi in vv. 15a, 16c, oxo|j.a in v. 20b and %eip in v. 25a, G's choice of (xoi in v. 12a can hardly be explained by a distaste for depicting the deity in anthropomorphic terms.178 Finally, by supplying yap in v. 12b, the translator guides his listeners to hear the rhetorical question in this line as a justification for the Lord's sweeping repudiation of their sacrifices (xaita): "Who has sought (xi.r|v |j.ou 13a on> 7tpoa0riOEa6e

eav (|>epT|Te ae|iiSa>av (laxaiov G-u^iana pSeXv/mx |ioi ecmv

G might have taken the infinitive Iran in v. 13b as the subject of the next clause, but instead he renders it as a finite verb, giving it a second-person subject in continuity with the previous verb, 7ipoo0f|oeo0e. This names the people explicitly as agents, an interpretive move that coheres with a whole series of subtle alterations and additions to the source text in Isa 1:10-20. By supplying first- and second-person pronouns181 and by shifting verbs into firstand second-person forms, G subtly highlights the interpersonal dimensions of the discourse.182 This sustained strategy of translation appears to reflect his sense that the contrast between 'you' and 'I' (established in the Vorlage in vv. 11-12) plays a key role throughout the Lord's second address. Taken together, these transformations deepen the pathos of the speech. It is important to recognize, however, that G is amplifying a feature already present in the source text rather than creating an entirely new effect.

language' of the Septuagint corpus (see pp. 5 8 - 6 2 above). G also employs this construction in Isa 7:10; 11:11; 23:12; 29:14; 47:1; 51:22; 52:1. 180 Sweeney 1988: 120 recognizes this catchword connection in the Hebrew text. 181 G supplies 7tpoi; |j.e in v. 15aa as a counterpart to a' 141 |id)v, 14a), 14c, 15b and 16d (in parallel with unoiv [16cfi] and in counterpoint to |iou [16ey]). 182 The Greek translator shifts the verbs in vv. 13b and 14b into the second person and transforms those in vv. 18b and 18c from third to first person.

112

Chapter 3: Translation and Interpretation

in OG lsa I :I-20

Contextual considerations have likely played a signficant role in G's choice of oen.i8aA,iav (cf. Frankel 1851: 97-98). For the Greek translator, it is not the particular sacrifices - let alone sacrifice itself - that is the crucial point at issue here, but the disposition of the worshipper, as the translator's provision of an explicit subject for these participles, o avo|a,on in Ezra 9:5 (cf. p. 121 n. 224). Fernández Marcos et al. 2005 rightly doubt Hatch-Redpath's equivalency vrioxeijeiv = N~,p at 3 Reigns 20:9, where vriaxeúactxe VT|axeiav most likely represents the translator's contextual rendering of a Vorlage reading DlS'linp with tit. 236 So a ' (ávaxfetéq), a ' 9' (áSiKÍov), it (3in), S D (iniquus). 237 On the necessity of such informed "guessing" by readers of ancient manuscripts, see Barr 1990; Tov 1984 and pp. 49-50 above. Fischer (1930: 18) suggests that N might have been read as ü, and 31 as 0, producing (so also van der Louw 2007: 197-198; cf. Weiss 1963). While one cannot rule out the possibility that G's Vorlage already read D1S (Gray 1912: 20; Ziegler 1934: 106; Koenig 1982: 414—424) several considerations speak against this hypothesis. First, no other witness attests a Hebrew text reading mx. Second, the argument that an original m s was changed to "IN to avoid offense cannot explain why the similarly problematic v. 14a was left untouched. Third, although the two terms occur in close proximity in Joel 1:14 and 2:15, mxj?l CIS does not itself appear to be a fixed phrase; thus, the variant •IX is no more likely to have arisen in the course of Hebrew transmission than during the process of translation. Finally, since some level of interpretive activity clearly lies behind the reading Clü/vr|axEÍa, there is no compelling reason to attribute the variant to an unknown Hebrew tradent rather than to the Greek translator, whose willingness to make interpretive adjustments to his source text, however modest, is displayed in v. 13d (f)|iépa neyóAri) and, indeed, throughout the chapter. So similarly Williamson 2006: 78-79; cf. Wildberger 1991: 35. 238 Compare Jer 14:12: cm éáv VTIOTFA'JOOKTIV, OIJK eíaaKoúao|ica Tfjc; Seriaeax; awtbv [cf. Isa 1:15], Kai éáv TtpoaevéyKoxnv ÓAOKA.ijxcóiiaxa Kai Guaía.^ [cf. Isa 1:11-13], OÚK ei> 8OKTIOCO év aijxoi>. Wilk (2003: 22 n. 28h) suggests that G derived VT|axeía from read as (BSí) a creative construal influenced by Isa 58:3 and 5, where ES3 tTUi) (xa.Tteivoijv tr]v x|/t>XTiv) stands parallel to mx (vr)axeia). Although in both Isa 58:3 and 5 the Greek translator

124

Chapter 3: Translation and Interpretation

in OG Isa I :l-20

practice of fasting in response to personal distress or national crises finds wide attestation both in Israel's scriptures and in the broader literature of the Second Temple period. Naturally, the temple courts served as a focal point for public fasts, though such observances were not restricted to the holy city or its temple.239 Although G's selection of apyia for m^B is without parallel in other Old Greek translations,240 there is no reason to suppose that he read a noun other than msi? here.241 Rather, as in the cases of nmo (v. 13b) and Xipo m p (v. 13d) examined above, G's rendering suggests both that he was familiar with the use of the Hebrew term in cultic contexts and that he chose an equivalent in accordance with his sense of the wider Isaian passage. In the Pentateuch, mXB appears as the term for the holy convocations (cf. Lev 23:36) that bring the festivals of Sukkot (Lev 23:36; Num 29:35) and Unleavened Bread (Deut 16:8) to a close. 242 The equivalent chosen by the translators of the Greek Pentateuch, e£,o8iov, apparently became a technical designation for the final day of Sukkot at some point, for it appears with precisely this sense in 2 Chr 7:9 and Neh 8:18. Philo attests the currency of this usage in

renders pu by xajieivoüv, it is not impossible that G hit upon the translation vricrteia by extension; compare Barn. 7:3/Lev 23:29 (discussed in n. 222 above) and note the extended sense of r n c n as 'fast' in Mishnaic Hebrew (n. 224 above). Such a complex chain of interpretive moves cannot be ruled out, but more immediate contextual factors strike me as providing an adequate (and simpler) explanation for G ' s procedure here. 239 Zechariah 7 - 8 mentions annual fasts held during the post-exilic period in the fourth, fifth, seventh and tenth months. See also 2 Chron 20:3-19; 1 Esdr 8:49; Neh 9:1; 1 Macc 3:47; 2 Macc 13:12; Joel 1:14; 2:12-17; Jer 43(36):9; Bar 1:5; Dan 2:18; Jdt 4:8-15; Josephus A.J. 5.166; 11.134; 12.290; 20.86-89; C. Ap. 2.282; Vita 290-303; Philo Contempl. 3 4 35; m. la an. Famously, one fast day was later associated with the translation of the Torah into Greek (Megillat Ta'anit Batra 21; see Veltri 1991). However, according to Hacham, "with the exception of Yom Kippur, the communal fast did not exist in Qumran" (2001: 145). 240 Elsewhere txpyia translates rotB (Exod 21:19) and m b a o (Eccl 10:18). The related verb, apyeiv, renders (Eccl 12:3; Ezra 4:24 [/>/.*]; cf. 1 Esdr 2:26). 241 The noun m s i J (til) is attested by a (e7iiaxeai, 13:13). And Philo (Spec. 2.60), sensitive perhaps to criticism from outsiders on this score, stridently insists that in requiring a weekly Sabbath the Law offers no inducement to laziness (paGupia).249 On the contrary, the Law directs people to labor strenuously (evspyevv) for six days every week. Moreover, it censures

243

"To the seven days he adds an eighth as a seal, calling it 'finale' (e^oSiov) - not only of this feast, it would seem, but of all the yearly feasts that we have enumerated and explained - for it is the final [feast] and the conclusion of the year" (Spec. 2.211). 244 Others who followed the pioneering translators of the Pentateuch similarly avoided the term eqoSiov. Some opted for Greek equivalents with specifically cultic connotations: 4 Reigns 10:20 (iepeia); Joel 1:14; 2:15 (Sepaiteia); Amos 5:21 (navriyupiq). The translator of Jeremiah selects a more neutral term for an assembly: e.g., Jer 9:1, oiivodoc,. So also in Isa 1:13, c ' renders niSJJ by auaipor|, 0' by ¿Tucruaxeaiq. See further Dogniez 2006. 245 Apyia occurs with this sense in Exod 21:19 (time lost from work due to injury), and it is often so used in Ptolemaic papyri. In Ezra 4:24/1 Esdr 2:26 apyetv refers to a stoppage of work during the rebuilding of the Temple; in Eccl 12:3 it denotes the cessation of bodily functions due to old age. Cf. Aquila's choice of eiticxecriq, 'stoppage, cessation,' in Isa 1:13. 246 Lev 23:36, n&v epyov Xaxpeuxov ov jiovriaexe; Num 29:35 rcav epyov Xaxpeuxov oi'i jcoifioexe EV aij-rfi; Deut 16:8, OTJ Jtoiriaeii; ev C A M I icav epyov itXfiv o a a 7tovr|0T|aexai *)A)%TI247 With reference to the Sabbath, see 2 Macc 5:25; Jos. A J. 12.4, 274; 14.63; 18.319, 323, 354; BJ. 2.392; 7.52; C. Ap. 1.209, 282; with regard to other festal days, see Jos. A J. 3.254; C. Ap. 2.232; in connection to the sabbatical year, see Jos. BJ. 1.60; A J. 13.234. Similarly, in pagan authors the noun can bear the positive sense of 'leisure' (PI. Leg. 61a) or 'holidays' (Arrian Epict. diss. 4.8.33; App. B. Civ. 1.56; P. Oxy. 31.2586 [264 BCE]). 248 Sir 33:28: TCOXXTIV Kaidav ¿SiSa^ev f] apyia. 249 "Seneca ... declares that their practice [of the Sabbath] is inexpedient, because by introducing one day of rest in every seven [the Jews] lose in idleness almost a seventh of their life" (Aug. Civ. 6.11, citing Seneca, De superstitione\ translation in Stern 1974, 1: 431). Similar criticisms are voiced by Tacitus (Hist. 5.4.3) and Juvenal (Sat. 14.105-106).

126

Chapter 3: Translation and Interpretation in OG Isa 1:1-20

"those who desire to remain idle (apyeiv) and just sit around (o%o^.d^eiv)." Far from inculcating slothful habits, Philo continues, the weekly Sabbath actually increases the worker's productivity. For by prescribing regular, carefully apportioned periods of rest (|ie|o.exprmevai dveoeiq), the Law provides a weekly respite from unceasing toil and allows the body to reinvigorate itself for the strenuous labors (rcpoi; xaq avxaq evepyeiai;) of the subsequent six days. Similar warnings about the dangers of apyia are commonplace among moral philosophers, both Jewish and Greek. 250 Once again the notion of translation as a necessary 'negotiation' of losses and gains proves illuminating. 251 With apyia, G lights on an equivalent that maintains a discernable connection with the source text through its association with m s a in cultic contexts. At the same time, the potential for apyia to carry pejorative connotations - especially, as here, within the framework of a reproof - allows the translator to recover something of the negative valence of the source text that he lost by selecting vriaxeia for ]1X. The rendering vr|oxeia Kai apyia results in a striking intratextual connection with the prophet's further fulminations against fasting in Isaiah 58. There are good reasons to believe this link is not coincidental. In all of Isaiah (apart from 1:13), it is only in Isaiah 58 that we again encounter the vocabulary of fasting (Vdiu, translated vrioxeia, 58:3, 5, 6; VTICTXETJEW, 58:3, 4).252 Moreover, the argumentative logic of Isaiah 58 closely parallels that of Isaiah 1:11-15. In both passages, Isaiah declares that it is futile for Israel to seek to draw near to God by observing its customary religious practices. 253 Despite their public displays of devotion, the Kyrios refuses to take notice of their pretensions to piety or pay heed to their fervent supplications. 254 The root of the problem in each case turns out to be not the Lord's indifference, but the "sins" (ajiap-

250 See Philo, Flacc. 33, 41; Legat. 128; Her. 77; Prob. 69; Jos. A J. 19.180, 248; BJ. 2.182; C. Ap. 2.228, 291. Among Greek moralists, apyia frequently appears in the company of other vices, such as naXaida ('effeminacy'), |ie6r| ('drunkenness'), pa.6u(j.ia ('laziness'), xpxxtrii ('self-indulgence'). See, e.g., PI. Leg. 873c, 901e; Resp. 398e, 405d, 422a; Arist. Eth. nic. 9.4.8; Plut. Adol. poet. aud. 24A; Adul. amic. 69B; An seni 791D. 251 See above, p. 46 with n. 46 and p. 89. 252 Note also larceivouv \|/uxriv in Isa 58:3, 5, recalling the diction of Lev. 23:27, 32. 253 Isaiah 58:2, eyyi^eiv 9eq> ¿TUOU^OOCTIV; cf. 1:12, oi>8' eav EPXT|a9e 606f|vai (xoi. For eyyi^eiv [9eq>] in a cultic setting, see Exod 19:21; Lev 10:3; 21:21; Isa 29:13; Ezek 40:46; 42:13; 43:19; 44:13; 45:4. 254 Compare, ti o n evricrteijaajiev Kai OVK elSeq (58:3) with cixav xixq X e ^P a i ¿Kxeivr|xe npoq ne, anooxpeyo) xoiiq o9aA.|ioij' 141 OJV (1:15a); and iva xi (ioi vriaxeijexe. ox; OT|(iepov aKO-ua&rjvai ev Kpavyfj xt^v (jicovfiv i)|xcov (58:4) with Kai eav TtA.ri9iJVT|xe xf)v 8er|aiv, 01I1K eiaaKoi)ao|Kxi {j(id)v (1:15b). If the people repent, however, the Kyrios promises to respond with a graciousness far exceeding their deserts: xoxe Pof|n%ai). Thus, the cleansing and purification that the Kyrios demands is the radical internal reorientation of repentance. 323 Although the Kyrios has resolved to "turn [his] eyes away" from them when they pray (15a), in reality there is nothing that does not appear "before [God's] eyes" (16c) - including the evil deeds that stain their souls.324 anevavxi

The Kyrios offers a twofold prescription for their purification, corresponding to the twofold nature of the voiioq as prohibition and command: "desist from your evil deeds" (16d) and "learn to do good" (17a).325 G elsewhere develops the theme of learning, already important in the Hebrew original, beyond the bounds of his source text. Notably, in 8:16 he castigates those who

321 With the sequence and ... a n e v a v x i . . . ano, G manages to vary his prepositions while maintaining the sound play. 1 322 Here G preserves a verbal link he finds in his source (in, 16c; inn, 16d; • inn, 4a). 323 Compare the command that frames the discussion of Yom Kippur in Lev 23:27-32: •STUDS] nK Drrai/icai Tajieivcbaexe xaq nivya^ u|id>v (Lev 23:27, 32). Note also Philo's claim that the scapegoat carries away the curses resulting from sins, but only for those who cleanse themselves through a repentance that leads to a renewed devotion to the practice of the Law: oi irerapoXaii; xalq npoq xo pe)aiov eKa9ap6r|aav, eilvo|iia Kaivfi itataiuiv avoniav EKvtyanevoi (Spec. 1.188). The Letter of Aristeas (234) similarly makes the point that purity of soul honors God more than sacrifice and offering: To xijiav xov 8eov ... eoxiv OIJ 8(ipoi8pà | i è v o i j v Xpóa r c a p e i K a ^ o v T c a a i a d a p t i o n f i n w v , 8 i à to ipoviKai elvai x o ) v \fuxwv (PC 30:204A). 350 So Duhm 1892; Scott 1956; H0genhaven 1988: 207. Others read these lines as questions, though without the sarcasm (Kutsch 1982; Kaiser 1983; Wildberger 1991). See further Willis 1983; Williamson 2006: 115-116). 351 For this sense of k o i ì èàv, see Smyth §2372; cf. our comments above on vv. 12a and 15b. In contrast, the initial m i of Kod è à v in v. 19 serves as a simple conjunction joining the second pair of conditional clauses, vv. 19-20a, to the one that preceeds (vv. 18b-c; on this construal of rat èàv, see Smyth §2373). Instead of ( J i o i v i k o ù v , a ' and 0' trace D'JB to V n x s ('change,' 'be different'), translating 8uxopa and àM-oiou^svov, respectively. 352 Contrast the passive verb forms found in D (quasi nix dealbabuntur) and S (^cncvuiu).

Appeal to Israel: Restoration

or Ruination

(1:18-20)

145

v. 18c|J.353 The Lord's adamant refusal to forgive his people's sins while they persist in practicing lawlessness (v. 14c) is more than matched by his readiness to cleanse them completely, if only they will turn from oppression and pursue justice. 354 Two paths now lie open before the people (vv. 19-20a). The Greek translator makes the contrast between them more straightforward, but also more prosaic. Rather than reproduce the lexical diversity of his source text, where v. 20aoc carries different verbs (jXO and m a ) from those found in the parallel stich v. 19a (H3S and BQO), G simply repeats the two verbs from v. 19a in v. 20aa. He then adds negative particles in order to construct a precise opposition between two possible responses to the Lord's call for repentance: Oeaeiv kou eioaKoi)eiv or |ifi 9eA£iv |xr|8e sioaKOTjeiv. 355 Ziegler judges this to be particularly effective from a rhetorical point of view: "In diesem Falle wirkt die Wiederholung sicherlich recht ansprechend und eindrucksvoll." 356 As in the original, the reference to hearing recalls the summons that prefaced this speech: "Hear (axoijeiv) the word of the Lord ... pay attention to the Law of God" (1:10). The addition of |iOi> to both v. 19 and v. 20a once again emphasizes the interpersonal nature of the conflict between the Kyrios and his addressees. 357 Should the people hearken to God's appeal and obey his commandments, they will "eat the good things of the land" (xa a y a 6 a xfjq yf\q

353

G further tightens the parallelism through the repetition of gktiv (v. 18ba) in v. 18ca; contrast a ' 0' TtuppoOcOai (cf. jiuppoijaOai for DIN in OG Lam 4:7 and Exod 25:5 a ' 0 ' 0 ). Alternatively, cociv (ox;) at v. 18ca may reflect G ' s interpretation of the verb in his Vorlage as a form of V n m , as perhaps in lQIsa", which reads 1DVT (so Ulrich and Flint 2010, 2:120). In this case also, however, the choice of axriv will represent a decision on the part of the translator, who elsewhere renders i1Q7 by 6 | i o i o w (1:9; 40:18, 25; 46:5) or 6(j.oioc elvat (14:14). In any event, it is reasonable to infer that concern for parallelism with v. 18ba motivated G ' s selection of the equivalent qxjvv (ox;) in v. 18ca (cf. a ' rnjppai raaiv). While the translator's singular ({joiviko'uv (v. 18bct) corresponds to the singular ,3tB in lQIsa", it is also possible that G read the plural DTD (as in !Tt) but chose to translate with the singular in keeping with normal Greek usage. Either way, the resulting pair, ooivikoOv/kokkivov, further enhances the parallelism between vv. 18ba and 18ca. 354 Only an atomistic reading of the Greek text leads to the conclusion that the translator finds in 1:18 an unconditional promise of forgiveness (pace Williamson 2006: 113). Read in the context of vv. 10-17 and vv. 19-20, it is clear that in OG Isaiah 1 reconciliation requires a return to the just practices of the vo|io. 17 The strict quantitative fidelity of Aquila's translation offers an illuminating contrast here: predictably, a translates b ¡Til in v. 22a with yeyovev eiq (as does a'). 18 I say "greater weight" because trouble with the Vorlage cannot be excluded as a factor in G's rendering; indeed, difficulty in reading the manuscript may have prompted the translator to consider the context more carefully in order to intuit the "correct" reading.

Lament for Zion (1:2l-23)

153

tax; it has the further effect of enhancing the epigrammatic quality of this short clause: TO apyupiov \)|i.cov a5oia|j.ov. 19 How, then, might we explain the plural ij|icov? Van der Louw suggests that the Greek translator simply gets caught napping here. Lulled by the steady succession of second-person plural forms in the preceding verses, he renders "J3DD unthinkingly as TO apyupiov ii|j.cov. Only when he proceeds to the parallel stich in v. 22b does he suddenly catch on to the fact that his Vorlage has switched to the second-person singular. 20 He does not think to look back to the line he has just translated (v. 22a), however, and so he misses his opportunity to resolve the discrepancy between the pronouns. A momentary lapse of this sort is not unthinkable, of course, especially if the Greek translator worked in a strictly linear fashion, never glancing back or looking too far ahead, as van der Louw's reconstruction seems to presume. Yet to this point in Isaiah 1, G has given every impression of attending closely to the continuity of the discourse he is producing, altering verbs from third- to second-person in two cases, supplying pronouns in several others - all for the purpose of bringing the relationship between speaker and addressees into sharper relief. Very shortly we will once again witness the translator intervening to heighten the cohesion of the discourse, this time by transforming second-person verbs to third person. 21 Moreover, we have had frequent occasion to observe that G has an ear for parallelism, a stylistic concern that demands the translator keep more than one colon at a time within his purview. 22 It seems quite unlikely, then, that G would have remained oblivious to the shift from i)|j.6)v to 00u in continguous lines or have failed to rectify the variance if t>|ia>v in v. 22a had been merely a momentary lapse. Rather than regard \j|ia>v as an inadvertent error, then, we should consider how the Greek translator might have understood the transition from plural to singular in v. 22, whether iJ(atov simply reproduces the reading of his Vorlage23 or whether it reflects his own transformation of the source text. A quick glance through OG Isaiah suggests that when it comes to second-person pronouns and verbs, at least, G is rather tolerant of sudden shifts of number. Sometimes the alternation between singular and plural can be traced to the Vorlage (e.g., 30:19-20; 48:5-6); in other instances, the fluctuation of number is probably attributable to the translator (e.g., 3:24-26; 10:27; 16:4). In each instance, the significance of the variation can only be appraised in the context

19 Note further the sonic resemblance between apyupiov and äööicinov. Contrast the rendering of a', which, though it preserves a close formal relationship to the source text, sounds rather clunky by comparison: TO apyupiov aou yeyovev eiiAra. 20 Van der Louw 2007: 211-212. 21 See below on vv. 29aß-30, pp. 197-199. 22 See above, e.g., pp. 72, 73, 78, 86, 91, 97, 106, 133, 146. 23 I.e., a text read as (n)QDSOD. See p. 152 and n. 16.

154

Chapter 4: Translation

and Interpretation

in OG Isa

1:21-31

of the larger discourse in which it occurs. Within a speech addressed to a colletive entity, such as a city or nation, the change of number may function rhetorically to engage the attention of the individual members of the group more directly. In the Greek version of Isa 3:16-26, for example, a tirade against the "ruling daughters of Zion" (3:17) that begins in the third person moves abruptly into the second-person singular for the dramatic portrayal of the violence and destitution that awaits each of them in captivity (3:24). The singular forms stress personal, not simply corporate, culpability,24 and they bring into view individual as well as collective suffering.25 Conversely, the address to Zion as a barren woman in Isa 54 employs second-singular verbs and pronouns throughout vv. 1—17b, while its concluding promise switches to the plural (54:17c: i)|j.ei appears in v v . 22b, 23a, 25a, 25b (an O G plus), 25d,

26a; a second-singular verb, k/VT|Gi]0T1, appears in v. 26b. In contrast, a variety o f epithets in the plural designate Z i o n ' s ruling class: o i Kci7rnX,oi aou, 22b; o i apxovxei; aou, 23a-f; o i iaxix>vT£ yap itaiSiá xó npáy^a; aXka KOIKÓV trji; iieyiaxr^ Komyyopiai; fi^iü)(iévov (PG 30:209A). Israel's vónopio TT|V yexpa. (xov erci oe. The Greek translator's interest in the theme of the Lord's anger extends beyond Isa 1 12 as well. As Troxel has shown (2008: 262), G's interpretive translation of Isa 28:2 makes the Lord's Owfioi; the topic of the entire chapter. G supplies 811(105 a s a n interpretive rendering also in 27:8 (Ttveiiiicm Ouiioii for m p , 'east wind,' parallel to 7cveunan aKXripco); in 30:27 (f) opyfi TOIJ 6-141011 for UIK1?, picking up two previous references to anger in the Hebrew text of v. 27: and nut, both translated opyr|); and in 31:4 (an interpretive addition as part of the translator's free recasting of the entire simile). Cf. his addition of opyfi in 26:21 for }0)1I>, attested by m, lQIsa 1 , 4QIsab (4Q56) [vid].

Rebellious

Rulers Ruined, Righteous

Rule Restored

(1.24-27)

173

redeemed will sing a song celebrating the cessation of the Lord's anger: "I will bless you, Lord, because you were wrathful against me, but you turned your anger aside (á7téaxpe\|/aq xóv 0D|ióv GOD) and showed me mercy" (12:1).105 Thanks to the translator's introduction of into the opening vision, Isa 1-12 is now bounded by references to the main "plot-line" of this first section of the book: the imposition of the Lord's wrath in judgment (1:24) followed by its removal in mercy (12:1). The trail followed by the translator in rendering v. 24cp as Kod Kpioiv EK xrav éxOpójv |iou 7ioitictco proves much simpler to trace. G clearly understands the noun Dp] to denote just recompense for wrongdoing. He translates Dp] by Kpioiq in 34:8, where the term stands in parallel to ávianóSoaiq (C'm^S). In 35:4, KIT Dp] becomes Kpioiv avxouioSiScooiv Kai ávxowioSíóoEi. Terms for vengeance appear as equivalents for Dp] in its remaining occurrences as well: É K 8 Í K T I < T I I ; (59:17), ávxanóSoaiq (61:2; 63:4) and xó S Í K O U O V (47:3, referring to retribution against Babylon). G handles the single appearance of the verb (1:24) in an comparable manner, rendering it with the idiom Kpioiv EK xivoq Ttoiefv, meaning 'to exact justice from someone.' 106 The reappearance here of the key term Kpioiq enhances the bitter irony that flavors this warning. Zion's rulers have refused to "learn to do good" or to "search out a just verdict" (náGexe Kodóv rcoieiv, ekí^IXTIOOCXE Kpioiv, v. 17); with their judgment perverted by bribes, they pay no regard to the just suits of orphans and widows (óp (v.

28), both rendered

with

eyKaxa^iTteiv. T o these he adds a third, translating •TTTTOQ (v. 4) and (v. 28) as avo)ioq. With the triad, oi &vo|ioi ... oi apaptwAoi ... oi ¿YKaxaX^iTtovxei; xov Kijpiov, G strengthens the connections between the end of this vision

and

its

beginning

(otiai

eOvog

ap.apxwA.ov

...

uioi

avo|ioi-

¿YKaxEAircaxe xov Kt>piov, v. 4).216 Thus, in his translation of 1:28, G again demonstrates profound sensitivity to the shape of the larger discourse, once more displaying his considerable skill in producing a Greek text that communicates the message of his source with clarity and eloquence.

212

This represents a stylistic enhancement of the chiasm in his source text at v. 28a-b.

213

See, for example, 10:12 (translating

10:22

28:22 ( r f o ) , and 44:24, 46:10,

55:11 (all rendering forms of ¡TOB). 214

For translational equivalents for VutBS in O G Isaiah, see p. 77 n. 42 above.

215

That the terms avonoi and aOexotivxE«; can function as near synonyms for the Greek

translator is suggested by 21:2, where G, rendering his Vorlage rather freely, places the two terms in parallel: o aOexcov aBexet, o avojicbv avojiei (THE n i E m "1:13 u i a n , !TC and lQIsa a ). See further 24:16: Oiiod xol^ aOe-uoiciv, oi dSexoiv-cEi; xov vopov (fil and lQIsa", O'lin ^ 'IK 1132

H33). W e suggested earlier (p. 77 above) that G ' s singular translation of

VuDS in 1:2 by aOexevv was influenced by the intertext in Isa 63:8. 216

Van der Louw (2007: 226) thus rightly identifies the articles in v. 28a (an O G plus) as

anaphoric. I cannot follow him in considering these additions "obligatory," however. The syntax of the Greek clause does not strictly require them; rather, they represent G's deliberate choice to observe a translational norm of textual acceptability,

which entails fostering coher-

ence at the discourse level. As already noted (p. 54 n. 91), van der Louw's overly broad conception of "obligatory" linguistic features skews his account of the Isaiah translator's work (for similar criticisms, see above, p. 70 n. 19; p. 132 n. 285; p. 134 n. 295; p. 145 n. 209).

202

Chapter 4: Translation and Interpretation

in OG Isa

1:21-31

Withered Tree, Waterless Garden (1:29-30) In the Hebrew text, vv. 29 and 30 each begin with the particle "O (see Fig. 15). The Greek translator interprets both of these as explicative, 217 but he orders them logically, assigning greater weight to the first with the more emphatic particle 8 u m and then using yap to mark v. 30 as a further explanation and expansion of v. 29. 218 So construed, these verses elucidate the preceding pronouncement of doom (v. 28) by linking the disgrace and destruction of the lawless directly to their ardent devotion to idolatry. Figure 15: Isaiah

1:29-30

r T 8 ? itia; -s •rnnn i p ? niaann r i s m

8 IOTI KATAIOXWQRIAOVTAI é r c ì x o ì t ; e ì S o ' j À o i q a / ù x o j v

a carrai riPoijXovTO rai

n b s n'pni ròs?3 "3 :rò mp—iox najDi

ÉITRIORXUV0TIOAV é r c ì XOÌQ KTITCOK; o r à x è v

a è7te0ij|jriaav

:AN-M IDS 30a

b

eoovxca yàp àx; xepépiv0ot)>Aa Kaì àx; itapàSeiaoc ijScop pi] éxcov

G largely preserves the parallelism of the two complex sentences that he finds in Isa 1:29. He selects forms of aiox'uveiv for both main clauses while varying the prefix (Kaxatox-uvew/eTiaioxweiv), 219 and he chooses synonymous verbs for the subordinate clauses (Pot>A£o0ou/e7u0u|j.£!v). In this way, he mimics the sense of dynamic parallelism produced by the lexical variation in

217 So, e.g., NRSV, Elberfelder. In contrast, JPS understands the first as asseverative, "Truly, you shall be shamed"; so also Buber-Rosenzweig: "Ja, zuschanden wird man an den Gotteichen." 218 D similarly translates the first as enim and the second as cum, thereby subordinating the second sentence to the first. Contrast 5 (a j and C ('"IS ... 'IK), which employ the same equivalent for both instances of 'D. Just within the confines of Isaiah 1 we see that G finds a range of nuances in the particle In addition to Sioxi (v. 29) and yap (vv. 20, 30), he also renders by oxi (v. 12) and e a v (vv. 12, 15). 219 The form KAXAIAX^veiv, attested by MSS A Q 710, is probably the original reading, with the simple form cdaxiweiv, found in MS S and in the majority of other witnesses, including hexaplaric and Lucianic manuscripts, representing a later correction, perhaps in light of a ' and a ' (so also Ziegler 1934: 44). We may note a similar tendency in the transmission-history of v. 29b to simplify the form ETtaiaxuveiv (a hapax in OG Isaiah) to aiaxvveiv (B, Lucianic manuscripts and many other witnesses). Observing that Kaxoaaxiiveiv (A Q) is the reading of 6' (apud Q), Rahlfs judged it to be a secondary intrusion into the Old Greek. His reconstructed text, a i o x u v e i v / e i t a i a x w e i v , agrees essentially with MS S. If Rahlfs's reconstruction were preferred, OG Isa 1:29 would provide another example of step-parallelism, a rhetorical figure we observed above in vv. 19-20, where in adjacent lines G offers first the simplex and then a prefixed form of the same verb.

Withered Tree, Waterless Garden

(1:29-30)

203

the parent text ("lsn/013; ~ira/mn). 2 2 0 For the complements of both main verbs G opts to employ the more idiomatic Greek construction with èjiì rather than try to represent the underlying |Q with the formally equivalent, but unidiomatic, preposition arcò.221 In v. 29b, this results in the striking sound play, énfi cry •uvGnqav èrcì xolq icf)7ioiq aùxràv a grreQugriaav. G further maintains the correspondence between the two sentences by adding the pronoun ocùxcòv both to v. 29aa and to v. 29ba. 222 His use of the neuter pronoun a in v. 29bp should likewise be seen as a deliberate strategy intended to mirror the syntax of the corresponding clause in v. 29a(3, where a has the neuter eiScoÀa for its antecedent. 223 Given G's obvious attention to parallelism in his rendering of v. 29, it is noteworthy that his translation of the verbs in the main clauses (vv. 29aa and 29ba) appears to weaken the correspondence between the two halves of the verse. Maintaining continuity with the verbs in v. 28 (cruvxpipfiaovxai,

220 Wilk (2003: 28 nn. 44b, e) believes the translator has reversed the order of the verbs (similarly van der Louw 2007: 229), noting that "ion translates E7ti0i>nevv in Deut 7:25 (see also Exod 20:17 [bis]\ Deut 5:21[18]; Song 2:3; Micah 2:2). Compare G's rendering of i n n by erti8t)(iux in Isa 27:2 (with Kakoq, a double translation of IQn [cf. 53:2, KaXtax;]) and in Isa 32:12. However, the equivalency i m = (5oijX£a9ai never appears elsewhere in OG Isaiah, or in the rest of the Septuagint corpus for that matter. Since in OG Isaiah both EJtiOuneiv (Isa 52:2 [bis]) and Poi3Xecs9ai (Isa 1:11; 53:10; 65:12; 66:4) render f s n , there is no reason why G could not have used either term for i o n in 1:29. Whatever the case, it is clear that in this verse the Greek translator has selected equivalents connoting 'desire' more than 'choice' (cf. Troxel 2010: 154, who argues that in OG Isaiah, "(ioij/xaGai and OeXeiv [regularly] express predilection"; contrast G's use of 8KA£yeiv for every other instance of "irn). Moreover, as a careful stylist, he has chosen to place the stronger term (e7ci0D(j.eiv) last. 221 For the construction with eni, see Xen. Mem. 2.2.8 (aioxuveiv ejii); Jos. A.J. 2.163; 1 Macc 4:31; Jer 10:14; Zech 9:5; OG Ps 68:7. In contrast, a ' a ' 9' translate with oato; cf. Jer

2:36, duo Aiywtxou (ruino) Kx6T|. G appears to have read a wyqtl (nnntzn VP) in 2:8 where m and lQIsa" have nnr.S" i t . 270 Van der Louw sets Isa 2:3 aside on the grounds that here "w + imperfect follows an imperative, so that the cohortative sense is suggested very strongly" (2007: 230 n. 249). 271 See pp. 211-212 above. 212 That is to say, the "cohortative sense" communicated by (van der Louw 2007: 230 n. 249) does not, in fact, determine the tense G selects for 13TV in Isa 2:3c, just as it does not govern his choice of the future for rD^T) in 2:3d. Contrast Isa 2:5, where after G translates the identical form rD*?31 with a hortatory subjunctive. 213 One or more of these verbs in G's Vorlage was in all likelihood a non-lengthened wyqtl form, although our extant witnesses do not entirely agree. In l:24c-26a, !T! offers the seF o u r of the five ries nrtpsi ... m/ONl - ni?P?$l prefix forms with waw are lengthened with -n. In place of ncpiX" in v. 24c(3, however, lQIsa" has cp:N1. In addition, the scroll reads the wyqtl form n'iom in v. 25aa where in offers nD'tDNt Due to a flaw in the manuscript, it is unclear how the form in v. 25ap (a wyqtl in Hi, r p¥K!) ended in lQIsa', which now reads In v. 25b(3, lQIsa" (TON!) and 4QIsa f (I'Ofsi]) each attest a wyqtl where ITT has the lengthened form nTONl.

Withered Tree, Waterless Garden

(1:29-30)

215

lator is not initially wed to the fixed equivalencies wqtl = future and wyqtl aorist.274 As early as Isa 1:2, he translates a wqtl form with an aorist verb, and in 1:24-26 and 2:3 he chooses future verbs to render wyqtl forms. In every instance we have examined, the context in which a given verb form occurs plays a large part in the translator's determination of its Greek equivalent. This observation only confirms the difficulty of the aorist G selects for the wyqtl n a n m in Isa l:29ba, however (see Fig. 17). Time and again in our study of Isaiah's opening vision we have seen that G is carefully attuned to the texture of the Greek discourse he is crafting as he translates his source. It is difficult to imagine, then, that he could have remained oblivious to the fact that the aorist in 1:29ba disrupts the otherwise unbroken sequence of future verbs in the main clauses of 1:28-31 and damages the parallelism between the two halves of v. 29.275 Figure 17: Isaiah

•'^•'Nl? "B31 •rnan "IDS N I A N O L'ÌSNRI'!

:Dnin3 "IBS

1.29 29a

B

Sióxi KATAI.AX,ov0f|AOVTOU èrti TOÌI; eiSwXoiq a/ùxràv a cràtoi f|p0i>>.0VT0 KOÙ ¿TRPCXIJV&RIAAV È N Ì TOÌpiov; cf. 1:3), and they have become (eyevri&ri) a heap of corpses because of famine and thirst for water." Together with v. 13, 5:25a envisions a past event. Yet the very next statement indicates that this earlier disaster is but the harbinger of a far greater calamity to come: "And in all this [his] anger has not turned away (OTJK ot7t£cn;pd(|>R|), but [his] hand [is] still upraised" (v. 25b). Supplying the connective particle xoiyapow to v. 26, the translator abruptly reverts to the future tense (last used in v. 24a) as he paints a vivid picture of the impending invasion of Judah by an implacable enemy from far away (5:26-30). Stepping back to survey the overall shape of the discourse in 5:24-30 allows us to appreciate the effect created by sandwiching a past-tense account of divine vengeance (vv. 24b-25a) between two future prophecies (v. 24a and vv. 26-30). A recollection of divine retribution executed in the past lends vividness and urgency to the prophet's warnings about further judgment that is soon to fall. The Lord's demonstrated willingness to mete out punishment to his people in the past makes the prediction of his future action all the more credible. The Greek translator, I suggest, produces a similar effect in 1:28-31 with his switch from future to aorist and back in 1:29-30. The transition is a rapid one, giving it a rather clunky feel, but the impact is much the same. While the future verbs in l:28-29a look forward to the decisive intervention in Zion's affairs by the Kyrios that has been in view since 1:24, G suddenly shifts perspective in 1:29b, recalling the past in order to evoke in his hearers a deep confidence that the final downfall of the idolators is certain. With the resumption of the future in v. 30 (eoovxou) the translator returns to his depiction of the retribution that yet lies ahead for the "lawless and sinfful] ones" who continue to plague the Lord's people. If Isa 1:29b looks backward, to what does it refer? A range of possible answers lie open to hearers of the Greek version. Within the context of Isaiah 1, "they have been put to shame" (¿7rr|GXijv8r|CTav) might hark back to the great distress Zion has already experienced as foreigners laid waste to the towns and countryside of Judah (v. 7). If the entire book of Isaiah constitutes the horizon for interpretation, then the change of perspective between past and future in v. 29 might signal an awareness that in the final form of the book ear-

281

I have translated these aorists, which reflect a Hebrew qtl - wyqtl sequence, as preterites, but they could also be construed as constative ("has become ... has set," etc.).

218

Chapter 4: Translation and Interpretation in OG Isa 1:21-31

lier crises presage calamities to come.282 So, for example, within the book of Isaiah the Syro-Ephraimitic war in the time of Ahaz (Isa 7-8) foreshadows the Assyrian invasion during Hezekiah's reign (Isa 36-39), and both together betoken the Babylonian conquest a century later. When Isaiah's opening vision is read from the standpoint of exile and return adopted in the latter half of the book (cf. already G's rendering of 1:27), Isa 1:29b calls to mind the fact that the idolatrous rulers of pre-exilic Jerusalem were indeed put to shame by the utter inability of their gods to avert Assyrians aggression or to halt the Babylonian juggernaut. Correspondingly, the promise of 1:29a, "they will be put to shame," anticipates the fierce anti-idol polemic that permeates Isa 40-55, even as it looks ahead to the final recompense of the wicked so chillingly portrayed in the book's final lines.283 Because the ultimate horizon of the prophet's vision - the restoration of Zion and the creation of a new heavens and a new earth - lies beyond the confines of the scroll, those who received and transmitted the book regarded many of its oracles as yet open to fulfillment. Isaiah's prophecies offered a trustworthy - if often enigmatic - vision of the future into which these later tradents and their communities believed themselves to be living under the watchful care of their gracious and sovereign God. Interpeters like those responsible for the book of Daniel, the sectarian writings of the Yahad and the New Testament sought through various techniques to "open the sealed book," that through it they might understand their own present and future.284 The Greek translator of Isaiah no doubt believed that much of the book of Isaiah spoke of a time yet to come. He may have believed that certain events in his own day represented, in some sense, fulfillments of particular Isaianic oracles; he may even have had a quite specific contemporary referent in mind when translating 1:29b, "They have been put to shame." However, if G encoded his translation of Isaiah with allusions to events in his own day, he has not left clear enough signs in the text to enable us to identify them with any degree of confidence. Thus, while I agree with Wilk, against van der Louw, that the Greek translator's selection of the aorist in l:29ba represents a considered choice rather than simply a default rendering of the wyqtl form in his source, I believe the explanations I have offered here and at 1:8 better account for the translator's Ubersetzungsweise in these passages than does the hypothesis of 'actualizing exegesis.'

282 By 'final form' I mean the form of Isaiah known to the Greek translator and to his second-century BCE contemporaries, which surviving evidence indicates was virtually identical to the shape of the book of Isaiah as we know it. 283 As noted above (p. 78 n. 46), the closing verses of Isaiah 1 contain multiple intratextual links to the final chapters of the book, suggesting that the final compiler(s) intended the opening and closing sections of the book to be read in light of one another. 284 See Blenkinsopp 2006.

Withered Tree, Waterless Garden (1:29-30)

219

Any explanation of the alternation of verb tense from future to aorist in Isa 1:29 remains hypothetical, of course. We have no unmediated access to G's intentions or to his strategies of translation. Both must be recovered, to the degree possible, from the text that he has produced - which is itself an artifact requiring interpretation. Indeed, the history of transmission of OG Isaiah puts a question mark on my own reconstruction of G's Ubersetzungsweise, for some scribes clearly found the syntax in v. 29 too choppy to tolerate. They sought to smooth things out, either by transforming the second verb into a future285 or by changing the first into an aorist.286 Still, the explanation I have proposed has the decided advantage of cohering with the picture of the Isaiah translator that has emerged throughout this investigation: G shows himself to be both conversant with the full sweep of the book of Isaiah and sensitive to the rhetorical effects of his translational decisions. Although implied from the beginning in the charge of "forsaking the Kyrios" (v. 4; cf. v. 28), the sin of idolatry is not named openly until Isa 1:29. G's Vorlage almost certainly read c'TXO (so ill) or D^ND (so lQIsa" and many Masoretic manuscripts) in v. 29a. The targumist, who understands C'TX to mean "trees," 287 clarifies the contextual meaning for his hearers with the translation xniUCD ' j ' t n , "the trees of the idols." It is possible that e'iScotax represents a similar metonymic translation of a word G understood to mean "tree." 288 But given his rendering of other passages in which similar forms appear, it is perhaps more likely that he took in Isa 1:29a to be the plural of bx, "god," as did, apparently, the Vulgate (ab idolis) and the Peshitta (ri'H^a). 289

285 This solution is found in a wide range of manuscripts: Alexandrian (A-106 26), Lucianic (nearly all such witnesses, including Sca, whose corrections reflect the influence of the Lucianic recension) and members of the catena group (377 564 565). As far as the original text of OG Isaiah goes, however, the seemingly problematic aorist form of the verb in v. 29b has weightier external evidence, being attested by MSS Q S* Sc and B. 286 The sole witness for this variant is the tenth-century Alexandrian manuscript 710. One thirteenth-century manuscript, 109 (categorized as hexaplaric by Ziegler 1939), omits the second verb altogether. 287 See HALOT, s.v. n * ^ ; cf. Williamson 2006: 149. 288 Symmachus translates aito xtov 5p\)|xwv, "from the oaks," while Aquila has duo ic^'updiv (on the latter, see Liitkemann and Rahlfs 1915: 258). 289 G renders n ^ f ) « in Isa 57:5 as eiti xa eiSwtax (so I , Kmiffl1?; 5, K - H ! ^ ; and D in diis), and he translates N^XBL in 41:28 (perhaps read as ri^XQI) as Kod ano TOW eiSw^wv aurtov (contrast 1TC, n'psm; £6|J.evoi; OIJK e a x a i 300

avxqj.

See pp. 174-175 above on G ' s employment of the " K a i style" in v v . 25-26.

301

See above, pp. 196-199.

302

In contrast, in a'

(Svo

avzoi),

a' (dfiijioxepa) and 8' (a(«|>6xepoi), as in the Hebrew

text, the nearer referents |0n and ibuE appear to be the more likely antecedents of QrntB. 303

G interprets ]Onn as a noun (cf. the Aramaic cognates |OQ and |0(i)n 'strength': Dan

2:37 [ O G loxii?]; 4:27 [ 6 ' io%\x;; O G Kpaxoq). 3 into the phrase KAAD|ir| GTUTTTUOU, an expression designating 'tow,' the short, coarse fibers of the flax plant that are stripped away when flax is prepared for spinning.304 The image recalls the earlier metaphor of the refiner's fire (v. 25). Just as one dresses and combs flax to separate the straw and tow from the long inner fibers used to make fine linen, so the Lord's judgment of Zion will remove the lawless and sinners from among God's faithful people. A highly combustible substance that burns rapidly, tow comes to epitomize that which is insubstantial and evanescent. So, for example, as soon as Samson flexes his muscles, his bonds dissolve "like tow when it smells fire" (rooei 0X171711)0v, rivim av 6o