The Samaritan Pentateuch and the Dead Sea Scrolls 9042937831, 9789042937833

Citation preview

C O N T R I BU T I O N S TO

BIBLICAL EXEGESIS & THEOLOGY

94

Michael Langlois (ed.)

The Samaritan Pentateuch and the Dead Sea Scrolls

PEETERS

THE SAMARITAN PENTATEUCH AND THE DEAD SEA SCROLLS

CONTRIBUTIONS TO BIBLICAL EXEGESIS AND THEOLOGY

SERIES EDITORS K. De Troyer (Salzburg) G. van Oyen (Louvain-la-Neuve)

ADVISORY BOARD Reimund Bieringer (Leuven) Lutz Doering (Münster) Mark Goodacre (Duke) Bas ter Haar Romeny (Amsterdam) Annette Merz (Groningen) Madhavi Nevader (St Andrews) Thomas Römer (Lausanne) Jack Sasson (Nashville) Tammi Schneider (Claremont)

Michael LANGLOIS (ed.)

THE SAMARITAN PENTATEUCH AND THE DEAD SEA SCROLLS

PEETERS LEUVEN – PARIS – BRISTOL, 2019

CT

A catalogue record for this book is available from the Library of Congress. © 2019 — Peeters, Bondgenotenlaan 153, B-3000 Leuven ISBN 978-90-429-3783-3 eISBN 978-90-429-3784-0 D/2019/0602/17 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher.

In Memoriam Gary N. Knoppers (1956–2018)

FOREWORD Seventy years after their discovery, the Dead Sea Scrolls continue to shed light on the Samaritan Pentateuch. The textual features, orthography, script, variant readings and even theology of the Samaritan Pentateuch have parallels in various manuscripts found in the Judaean desert and copied during the Second Temple period. After a statusquestionis by Magnar Kartveit (p. 1), Emanuel Tov (p. 19) traces the textual history of the Samaritan Pentateuch back to its common ancestry with the Septuagint. The phenomenon of “harmonization,” for which the Samaritan Pentateuch is so famous, is studied in detail by Michaël N. van der Meer (p. 41), who compares it to similar features found in the Dead Sea scrolls. This phenomenon may even account for the mention of Mount Gerizim in the Samaritan version of the Ten Commandments, according to Stefan Schorch (p. 77), who shows that this insertion is in fact not typically Samaritan. The issue of altars and cult centralization is taken up by Gary N. Knoppers † (p. 99), who combines archaeological and textual data to shed light on the political and religious history of Samaria and Judaea. On the basis Deuteronomy’s oldest textual witnesses – found in the Judaean desert – as well as the Samaritan Pentateuch, the Septuagint and the Masoretic text, Benjamin Ziemer (p. 127) offers a new stemma for the Book of Deuteronomy as a whole. Innocent Himbaza (p. 199) focuses on legal material in Leviticus and Deuteronomy and its harmonization both by the Dead Sea scrolls and the Samaritan Pentateuch. Another case of harmonization is studied by Jonathan Ben-Dov (p. 217), who compares Num 20–21 and Deut 2–3. Beyond the Torah, Abraham Tal (p. 243) shows the affinities between the scribal and editorial activities at work in 1QIsaa and those that underlie the Samaritan Pentateuch. Michael Langlois (p. 255) also focuses on scribal features and transitions from text-type to history by means of palaeography, offering a new dating of the Dead Sea scrolls associated to the Samaritan Pentateuch or written in the (palaeo-)Hebrew script. Thanks to scribal practices, and orthography in particular, Christian Stadel (p. 287) shows the linguistic affinities between the Hebrew dialects of the Dead Sea Scrolls and the Samaritan Pentateuch. Finally, beyond Hebrew, Jan Joosten (p. 313) identifies Samaritan interpretation hidden in marginal readings of Greek manuscripts.

VIII

FOREWORD

These twelve contributions have been peer-reviewed and are the outcome of an international conference held on May 26–27, 2016 at the University of Strasbourg with funding by the University Institute of France. The preparation of this volume has also been supported by the Helsinki Collegium for Advanced Studies and a French-Israeli PHC grant. I want to thank all the participants of the conference for their excellent scholarship and the staff of the university, together with my student, Thibault Foulon, for their tremendous help with the logistics. We hope that this conference and its proceedings have pushed the boundaries of our knowledge on the Samaritan Pentateuch and the Dead Sea scrolls. Michael Langlois

Addendum: It is with deep sadness that I learnt of Gary Knoppers’ sudden death on December 22, 2018. I am honored to dedicate this volume to his memory.

TABLE OF CONTENTS Magnar KARTVEIT Scholars’ Assessments of the Relationship between the Pre-Samaritan Texts and the Samaritan Pentateuch. . . . . . . . . . . . . . . . .

1

Emanuel TOV From Popular Jewish LXX-SP Texts to Separate Sectarian Texts: Insights from the Dead Sea Scrolls . . . . . . . . . . . . . . . . .

19

Michaël N. VAN DER MEER Exclusion and Expansion: Harmonisations in the Samaritan Pentateuch, Pre-Samaritan Pentateuchal Manuscripts and Non-Pentateuchal Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41

Stefan SCHORCH The So-Called Gerizim Commandment in the Samaritan Pentateuch

77

Gary N. KNOPPERS † Altared States: The Altar Laws in the Samaritan and Jewish Pentateuchs, and Their Early Interpreters . . . . . . . . . . . . . . . . .

99

Benjamin ZIEMER A Stemma for Deuteronomy . . . . . . . . . . . . . . . . . . . . . 127 Innocent HIMBAZA Looking at the Samaritan Pentateuch from Qumran: Legal Material of Leviticus and Deuteronomy . . . . . . . . . . . . . . . . . . . 199 Jonathan BEN-DOV Text Duplications between Higher and Lower Criticism: Num 20-21 and Deut 2-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Abraham TAL Do the Samaritan Pentateuch and 1QIsaᵃ Follow the Same Model? 243 Michael LANGLOIS Dead Sea Scrolls Palaeography and the Samaritan Pentateuch . . 255

X

TABLE OF CONTENTS

Christian STADEL Variegation in Second Temple Period Hebrew: Passive t-Stems, the ‫ הלז‬Demonstrative Series, and ‫ אפוא‬in Samaritan Hebrew and in the Dead Sea Scrolls . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 Jan JOOSTEN Biblical Interpretation in the Samareitikon as Exemplified in Anonymous Readings in Leviticus Attested in M′ . . . . . . . . . . . . . 313

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP BETWEEN THE PRE-SAMARITAN TEXTS AND THE SAMARITAN PENTATEUCH Magnar KARTVEIT

In which ways do the pre-Samaritan texts and the Samaritan Pentateuch shed light on each other? The research history of this question opens with an article by Patrick W. Skehan from 1955. Here he presented a “Samaritan” scroll from Qumran that in one respect was a “surprise”, but in other ways confirmed Wilhelm Gesenius’ appraisal of the SP from 1815. Much of the investigations into the question undertaken after 1955 also built on the insights of Gesenius. As the DSS were published and the study of them increased in the 20th century, new discoveries and new approaches led to a more nuanced evaluation of the relationship between the DSS and the SP. Standing on the shoulders of Gesenius, Skehan and many others, we are now able to see better what the situation was like when these texts were transmitted and transformed in the centuries around the turn of the eras.

THE OPENING: PATRICK W. SKEHAN The 1955 article by Patrick W. Skehan presented what he called “Exodus in the Samaritan Recension from Qumran”.1 This article announced that among the Qumran texts there was a scroll of Exodus with features that were previously only known from the SP: 4QpaleoExodm. One can perceive the author’s joy of discovery in the presentation of the scroll: The Exodus scroll … has been judged of sufficient interest to make a preliminary notice desirable in advance of the full edition of the biblical texts from Cave 4 at Khirbet Qumran… The recension in question is the “Samaritan” recension, with all the essential characteristics of that fuller text, including its repetitious manner of recounting the plague episodes, its 1

Patrick W. Skehan, “Exodus in the Samaritan Recension from Qumran”, JBL 74 (1955): 182-187.

2

MAGNAR KARTVEIT

borrowings from Deuteronomy and its transpositions; this is true at almost every point where the extant fragments make verification possible.2

The significance of this discovery was enhanced by these observations: “The script cannot by any stretch of the imagination be called Samaritan … Neither is the orthography Samaritan.”3 Skehan’s “surprise” was therefore that “the Samaritan recension … is shown by this scroll to have been preserved with a measure of fidelity … that compares not unfavourably with the fidelity of transmission of MT itself…”4 On the assumption that the text in significant respects was the “Samaritan recension” he concluded that this recension had been preserved well over the ages. But as he surmised that the recension was Samaritan, yet the script and the orthography were not, he cautiously concluded that a suspended judgement was called for on the question of the scroll’s Samaritan character.5 In this article Skehan briefly introduced the scroll and then presented a photograph of one column containing Exod 32:10-30 (later known as col. XXXVIII) with Hebrew transcription and comments. Skehan’s overall evaluation of the scroll is twofold: By way of conclusion, the writer wishes to underline the judgment implicit in what has been said above: that this new evidence for the antiquity and for the constancy of transmission of the Samaritan recension of Exodus does not alter the internal evidence for the secondary character of much that is proper to that recension, and that the general appraisal of the recension itself remains in outline substantially what it has been since the time of Gesenius.6

Skehan referred to Wilhelm Gesenius’ study, but failed to see the significance of the fact that the “essential characteristics” of the “Samaritan recension” had been found at Qumran.

THE BACKGROUND: WILHELM GESENIUS If we take a look at Wilhelm Gesenius’ study of the SP, 4QpaleoExodm is even more significant than Skehan realized.7 Gesenius studied the particular readings of the SP and placed them in eight categories: a) Readings 2 3 4 5 6 7

Skehan, “Exodus”, 182. Ibid., 182f. Ibid., 183. Ibid., 183. Ibid., 187. Guilielmus Gesenius, De Pentateuchi Samaritani origine, indole et auctoritate: commentatiophilologico-critica (Halle: Rengerianae, 1815).

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

3

that have been adjusted according to the grammatical norm; b) Interpretations or glosses received into the text; c) Conjectural emendations; d) Readings corrected or added on the basis of parallel passages; e) Large additions interpolated from parallel passages; f) Emendations of passages that are difficult in subject matter, mostly of a historical type; g) Morphological adjustments to the Samaritan dialect; and h) Passages conformed to Samaritan theology and hermeneutics.8 The names of the categories are telling: they imply that the SP mainly contains secondary readings in relation to the MT. Gesenius’ conclusion is well known: only in four instances is the SP to be considered an older text: Gen 4:8 (Cain said to his brother, “Letus goouttothefield.”[MT: nil]); 14:14 (Abram counted [MT: armed] his followers in order to free Lot); 22:13 (Abraham saw one [MT: behind] ram); 49:14 (Issachar is a bony, i.e. strong, donkey [MT: donkey of bones, i.e. of strength]).9 In all other cases he found the text to be inferior to the MT. Skehan referred to Gesenius’ conclusion and overall estimation of the secondary character of SP, but he did not take into consideration that Gesenius studied manuscripts he estimated to be later than the 13th century CE,10 whereas Skehan stated that texts with the script of 4QpaleoExodm “were being currently produced in the second and es first centuriB. C. [sic, = first centuries BCE].”11 He therefore dealt with a text that was older than those studied by Gesenius by more than a millennium. Skehan focused on phenomena covered by Gesenius’ category e), but if e.g. categories g) and h) describe phenomena of the SP that are found in Qumran texts, they are less Samaritan than Gesenius could imagine and more significant than Skehan noted. A brief look at the scholarly situation at the time of Gesenius is helpful for evaluating his stance. The SP-manuscript that Pietro della Valle brought to Europe in 161612 was made accessible in the Paris Polyglot 8

9 10 11 12

“a) Lectiones, quae ad normam grammaticae exactae sunt; b) Interpretamenta s. glossae in textum receptae; c) Emendationes coniecturales; d) Lectiones e locis parallelis vel correctae vel suppletae; e) Additamenta maiora e locis parallelis interpolata; f) Emendationes locorum difficultate quadam in rebus, et maxime quidem historico in genere, laborantium; g) Formae vocabulorum ad dialectum samaritanam attemperatae; h) Loca ad theologiam et hermeneuticam Samaritanorum domesticam conformata,” Gesenius, PentateuchiSamaritani, v-vi, with some explanations of these categories on pages 26-60. Gesenius, PentateuchiSamaritani, 61-64. Ibid., 17. Skehan, “Exodus,” 183. And another with the Samaritan Targum to the Pentateuch.

4

MAGNAR KARTVEIT

of 163213, which again was used as a base text for the London Polyglot of 1655-5714. In the Paris Polyglot the whole Samaritan text was printed in Samaritan characters, and the Samaritan Targum underneath, with Latin translations. A good century later, the London Polyglot’s presentation of SP in turn became the source for Benjamin Kennicott’s edition which appeared in 177615. Kennicott printed only the variants in a left hand column with the Masoretic text in the right hand column. It was estimated at the time that the SP has 6000 variants in relation to the MT.16 In 1900 of these cases the SP agrees with the LXX against the MT. The arrival of the SP in Europe happened at a time with disputes between Catholic and Protestant scholars about the original text of the Old Testament, and the SP became entrapped in questions of original and authoritative text. This situation became the fate of the SP for a long time. Pope Clement VIII declared in 1592 the canon of the LXX and the text of the Vulgate to be decisive for faith and life (Editio Sixto-Clementina). Forty years later, Jean Morin/Iohannes Morinus not only published the Samaritan text in the Paris Polyglot, but he also considered it to be “of considerable purity,… a witness to the fragility of the Massoretic [sic] text and a proof of the authenticity of the Septuagint”17. This conclusion was criticized by the Protestant scholars Johannes Buxtorf in 1620, Simon de Muis in 1631, and Henricus Hottinger in 1644, all of whom defended the Masoretic tradition. The Protestants even claimed infallibility and inspiration for the Masoretic text, and the attacks on the Catholic position were vehement.18 We note the positive evaluation of SP against MT by the Catholics. 13

14

15

16 17

18

BibliaHebraica,Samaritana,Chaldaica,Graeca,Syriaca,Latina,Arabica,quibustextus originalestotiusScripturaeSacrae…. Paris: Vitré & Gallicanus, 1629-1645; Vol. 6, ed., J. Morinus. B. Walton, “Animadversiones Samariticae, Textum hebraeum et samaritanum,”, in S. S.BibliaPolyglottaCompletentiaTextusOriginales:HebraicoscumPentat.Samarit: etc., vol. 1, London: 1657. Vetus Testamentum Hebraicum, cum variis lectionibus, edidit Benjaminus Kennicott, 1776. Kennicott used Walton’s polyglot, but in all he used 18 Samaritan manuscripts from different libraries, 11 that were “per totum collati” for Gen, 106, Ex, 203, Num, 360, and Deut, 443, and 10 for Lev, 266, and 7 that were “in locis selectis collati” for all the books. Cf. B. Blayney, Pentateuchus Hebraeo-Samaritanus charactere HebraeoChaldaicoeditus, 1790. Cf. Walton, S.S.BibliaPolyglotta, London 1657. G[illis] F[irmin]/P[aul] L[ester] S[tenhouse], “Morin, Jean (1591-1659),” ACompanion toSamaritanStudies, ed. Alan D. Crown, Reinhard Pummer, Abraham Tal (Tübingen: Mohr Siebeck), 1993, 160; Morin’s main works on the SP include the Excercitationes from 1631, his Diatribe from 1639 and his Opuscula, published 1657. R. T. Anderson, “Jean Morin as a Catalyst in Modern Textual Criticism”, ForaTestimony:EssaysinHonorofJohnH.Wilson, ed. M. Haykin (Toronto: Central Baptist Seminary and Bible College), 1989, 62-73.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

5

Gesenius laid the foundation for the negative evaluation of it against MT, and this has prevailed. Gesenius was part of the Protestant-Catholic discussion, and sided with the Protestants in their high esteem of the MT. This way of treating the SP as a part in the discussion on the oldest, “best” and most authoritative text has prevailed and can be seen in Bruce Waltke’s dissertation on the SP from 1965, and even in Judith Sanderson’s dissertation on 4QpaleoExodm from 1986 and her publication of it in DJD, vol. IX in 1992.19 After Skehan’s brief introduction to 4QpaleoExodm one might expect that the interest in the possible “Samaritan” features of this text and other texts from Qumran would increase, e.g. those covered by Gesenius’ categories g) and h). The linguistic phenomena were, in fact, increasingly taken into account in the three grammars of Samaritan Hebrew that appeared after 1955.

THE LINGUISTIC EVIDENCE: GRAMMARS

AND MORE

The first grammar after 1955 was written by Aimo E. Murtonen and published in 1964. He studied the text of the SP and used recordings of contemporary Samaritan reading of the Torah, but did not take an interest in the findings from Qumran.20 If he had followed up on Gesenius’ category g) and studied the Qumran material available at his time, he might have reached new conclusions, especially as his goal was to contribute to historical linguistics. The situation is different in Rudolf Macuch’s grammar from 1969.21 A number of times he refers to Qumran material, often to “die JesajaRolle (the Isaiah scroll)” (= 1QIsaa). From the latter he mentions the forms attimma, imma, -imma, and -kimma (personal pronouns and pronominal suffixes for 2. and 3. p. pl.), as already noticed by Paul Kahle and Ze᾿ev BenḤayyim.22 These, and some other phenomena, are considered “altertümliche Elemente (ancient elements)” in Samaritan Hebrew, but this observation has 19

20

21

22

B. Waltke, Prolegomena to the Samaritan Pentateuch, PhD diss. Harvard, 1965; J. Sanderson, AnExodusScrollfromQumran:4QpaleoExodmandtheSamaritanTradition, HSS 30, 1986; eadem, “4QpaleoExodusm,” 1992. Aimo E. Murtonen, MaterialsforaNon-MasoreticHebrewGrammarIII:AGrammar oftheSamaritanDialectofHebrew, StOr 29 (Helsinki: Societas Orientalis Fennica), 1964. Rudolf Macuch, GrammatikdessamaritanischenHebräisch, StSam 1 (Berlin: de Gruyter), 1969. Ibid., 91.

6

MAGNAR KARTVEIT

not triggered an investigation into the Pentateuchal texts from Qumran. That the Samaritans have preserved forms known from “the Isaiah scroll” and other Qumran texts is not subject to further reflections in this grammar – apart from the suggestion that they are ancient. Ze᾿ev Ben-Ḥayyim’s grammar from 2000 is much more aware of the Qumran material.23 The linguistic results from his life-long occupation with matters Samaritan are worth quoting. …many of the features of S[amaritan]H[ebrew] are the same as those evident in non-Biblical Hebrew literature among the Jews, such as rabbinic literature. Furthermore, now that we have access to a number of biblical works, whole or fragmentary, preserved among the Dead Sea Scrolls, and find there linguistic features similar to SH, it is entirely certain that we cannot ascribe the differences between the Biblical Hebrew of the Jewish Pentateuch and those of SP to differences of place alone, but rather to differences of time. SP in its written form displays some features of the language as we know it from Second Temple times – more specifically, from the end of that period. As a result of these discoveries, things that were considered generations ago to be peculiar to the Samaritan Pentateuchal tradition turn out to be not so. In summary, then, SP, the sole literary source extant among the Samaritans that dates from the First Temple period, is presented to us in a linguistic redaction that reveals, to the extent possible, features particular to Hebrew of the Second Temple period…, even though its external appearance, the formation of the letters and the division of words by means of a dot, antedates that of the Jewish Pentateuch.24

These insights represent a fundamental shift from Gesenius and show how much linguistic study has contributed to the question of the relation between the SP on the one hand and the pre-Samaritan manuscripts and other DSS on the other. As more and more material from Qumran and other places was published, a closer study of this relation became possible. Especially interesting in this respect is 4QNumb, published by Nathan Jastram.25 He describes this manuscript as having a “close relationship” with SP and “the most significant are the major interpolations shared with” SP.26 “In table 4 of this edition are presented morpho-phonological features in this manuscript 23

24 25

26

Ze᾿ev Ben-Ḥayyim, with assistance from Abraham Tal, A Grammar of Samaritan Hebrew:BasedontheRecitationoftheLawinComparisonwiththeTiberianandOther Jewish Traditions, Revised Edition in English (Jerusalem: Magnes; Winona Lake, In: Eisenbrauns), 2000. Ibid., 3-4. Nathan Jastram, “4QNumb”, QumranCave4VII:GenesistoNumbers, DJD 12 (Oxford: Clarendon), 1994, 205-267. Ibid., 213, 215.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

7

different from MT, but found in the Samaritan tradition.”27 Among those features are the forms in “the Isaiah scroll” noted by Kahle, Ben-Ḥayyim, and Macuch, and mentioned above. Jastram does not discuss further these relations. Judith Sanderson has not presented similar material in the case of 4QpaleoExodm, nor has Frank More Cross for 4QExod-Levf. Judging from the studies published on the linguistic situation, one gets the impression that one of the pre-Samaritan manuscripts and certain other texts from Qumran reveal features of the Hebrew language that are preserved in the Samaritan tradition. This phenomenon has been studied to some extent, and the features that are noted by scholars tie the Samaritans closer to the Jewish development of the centuries around the turn of the eras. Gesenius’ category g) would therefore have to be conceived of differently today. More research would be needed in order to get a more precise impression of this area. FROM “SAMARITAN” QUMRAN TEXT TO THE SP-GROUP OF TEXTS Skehan termed 4QpaleoExodm “Samaritan,” but he changed his opinion when he learned that the scroll did not contain the Samaritan tenth commandment.28 Skehan’s original stance was taken up by Maurice Baillet. After a number of studies on Qumran manuscripts and the SP, he published in 1970 an article where he discussed variants and text divisions in 17 manuscripts from Qumran, containing text from Exodus, and compared them with the SP. To some extent he followed the classification used by Gesenius, and his conclusion was that certain manuscripts from Qumran contain, not only stray Samaritan elements, but the Samaritan text itself.29 Taking into consideration the script, he even raises the question if the Qumran Pentateuchal manuscripts in ancient script cannot simply be considered Samaritan manuscripts.30 In 1988 Baillet presented a study of the SP in the Hebrew, Greek, Aramaic, and Arabic versions.31 For the Hebrew SP he concluded that 27 28

29

30 31

Ibid., 214. Patrick W. Skehan, “Qumran and the Present State of Old Testament Studies: The Masoretic Text,” JBL 78 (1959): 21-25. Maurice Baillet, “Le texte samaritain de l’Exode dans les manuscrits de Qumrân”, Hommages à André Dupont-Sommer, ed. A. Caquot and M. Philonenko (Paris: Libraire d’Amérique et d’Orient Adrien-Maisonneuve, 1971), pp. 363-381; “certains manuscrits de Qumrân contiennent, non seulement des éléments samaritains disséminés, mais le texte samaritain lui-même,” p. 380. Ibid., 381. Maurice Baillet, “Les divers états du Pentateuque samaritain, “ RB 13 (1988): 531-545.

8

MAGNAR KARTVEIT

1. The Samaritan script descends from the pre-exilic Paleo-Hebrew script; 2. Many SP scribal details are attested at Qumran; 3. The Samaritan pronunciation is partly a vestige of the ancient Palestinian pronunciation; 4. The Samaritan recension is the daughter of a pre-masoretic recension; 5. The decisive period for the fixation of the Samaritan text and letters is between the end of the 1st and the 3rd cent. CE. In this period the addition of the tenth commandment also took place.32 The terminology for this text and other texts with the same characteristics has been “proto-Samaritan”33, but now they are termed ”pre-Samaritan” because of the difference to the “proto-Masoretic” texts.34 “Whereas the proto-Masoretic texts derive from the same circles as the medieval [MT] … the pre-Samaritan texts, while agreeing much with [SP], lack the ideological Samaritan features.”35 On this background, Emanuel Tov speaks of the “SPgroup” that is constituted by the pre-Samaritan texts and the SP.36 “At a second stage, after the content of [SP] had been fixed, no major changes were inserted when it was copied but, unlike the [MT]-group, scribes created many orthographical differences between the manuscripts.”37 By the time of Gesenius, a number of Samaritan manuscripts were known in Europe, mainly through Kennicott’s listing of variants from 1776-80.38 Gesenius had already noted that the Samaritan manuscripts varied considerably among themselves and von Gall concluded that it was impossible to reconstruct textual families, and even the idea of an archetype was difficult to retain.39 In this respect, the SP-group is different from the MT-texts.

THE SP: A

SCHOLARLY EDITION, A VULGAR TEXT, A LOCAL TEXT, A RECENSION, OR A LITERARY EDITION?

The variation among Samaritan manuscripts together with the differences to the other textual witnesses laid the foundation for the thesis 32 33

34

35 36 37 38 39

Ibid., 539f. E[manuel] T[ov], “IV. Proto-Samaritan Texts», Companion to Samaritan Studies, 182f. Emanuel Tov, TextualCriticismoftheHebrewBible, third edition revised and expanded, (Minneapolis, MN: Fortress), 2012, 74-92. Ibid., 76. Ibid., 74. Ibid., 79. E. Tov, TextualCriticism, 2001, plate 17, gives an impression of Kennicott’s procedure. A. Fr. von Gall, DerHebräischePentateuchderSamaritaner (Giessen: Töpelmann), 1914-18: “etwa 80”, Preface, dated Christmas 1913, p. VII*.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

9

advocated by Paul Kahle, Gillis Gerleman and Shmaryahu Talmon. They held that the SP constitutes a vulgar version compared to the standard text of the Masoretes. “S[P] is originally a vulgar version of the Torah in which popular trends were systematised and which, at the crucial point of its history, was provided with a ‘typical Samaritan’ superstructure.”40 Sh. Talmon here takes issue with Gerleman and refines the studies by Kahle. “Vulgar” refers to a text which is smooth and reworked to suit popular needs for understanding the text. A shift from Gesenius is here evident: Gesenius considered the Samaritans as critics, scholars, who worked on the text to make it acceptable to a critical mind of its day. This idea was followed up by Waltke, even though both he and Gesenius supposed that the MT was the more original text, and therefore to be preferred. The idea of the SP as a vulgar text is a different approach to it. It relegates the SP to the status of interesting old, but for its quality, long superseded textual witnesses. All previous observations lie behind F. M. Cross’ hypothesis of textual families. SP was seen as a Palestinian text, the LXX as an Egyptian or even Alexandrian text, and the MT as a Babylonian text.41 The geographical distance would explain the textual distance, and the import of texts to Palestine would explain the similarities. Whereas Skehan spoke about the Samaritan recension, Tov now thinks in the direction of an edition: “The changes should be considered editorial rather than harmonizing.”42

POSSIBLE IDEOLOGICAL CONCERNS Was the character of the SP-group determined by ideological concerns? A number of scholars have made different proposals in this direction. Gesenius set the track for later research: the large interpolations are to be “referred to the continuous work of the Samaritans, to render their text fuller and plainer.”43 The parlance was later to be “major harmonistic 40

41

42 43

Sh. Talmon, “The Samaritan Pentateuch,” JJS 2 (1950-51): 150; cf. G. Gerleman, Synoptic studies in the Old Testament, Lund universitets årsskrift, N. F., Avd.1, 44:5 (Lund: Gleerup), 1948; P. Kahle, “Untersuchungen zur Geschichte des Pentateuchtextes,” ThSK 38 (1915): 399-439. Frank M. Cross, “The Evolution of a Theory of Local Texts,” SeptuagintandCognate Studies, ed. R. A. Kraft. 2. (Missoula, Mont.: SBL) 1972, 108-126. Tov, TextualCriticism, 80. Gesenius, PentateuchiSamaritani, 47: “Malumus itaque has etiam interpolationes ad perpetuum Samaritanorum studium, textum suum pleniorem planioremque reddendi, referre...” Waltke’s translation: “We prefer, therefore, to see these interpolations as the

10

MAGNAR KARTVEIT

expansions,” in the words of Eugene Ulrich, “intentionally systematically developed.”44 This is further made precise as follows, the expansions “reflect a tendency not to leave in the Pentateuchal text any internal contradiction or irregularity which could be taken as harmful to the sanctity of the text.”45 Abraham Tal suggested that a large number of the variants are of a literary nature, designed to make the text more comprehensible (cf. Exod 7:18.29; 8:19; 9:5.19; 10:2; 11:3) whenever the redactor considered the narrative too segmental.46 On the other hand, James Purvis characterized them as an “unnecessary redundancy”: “[t]hey are the result of the growth of a textual tradition whose development covered several centuries.”47 The idea of harmonization has to come to terms with some other features of the text, e.g. of Exod 6-11: the two stories of the death of the animals in Egypt were not harmonized; the firstborn of the cattle die twice in the SP also (cf. Exod 9:6 compared with Exod 9:25; 12:29). Moses was to meet Pharaoh a last time in Exod 10:29, but there is another meeting in Exod 11:8. The people believed Moses according to Exod 4:31, and yet in 6:9 he complains that they do not listen to him. A number of inconsistencies are not harmonized. Tov more recently expresses his views on this question carefully: The most characteristic readings of the [SP]-group were created by substantial editorial changes inserted in the earlier text… These changes involve additions (duplications) of other Torah verses (with changes in names and verbal forms) and a few rearrangements but no omissions, following a strong inclination in [SP] not to alter the divine word. The changes should be considered editorial rather than harmonizing … characteristic of the [SP]-group only... Ultimately, the changes reflect theological concerns.48

44 45

46 47

48

endeavour of the Samaritans to render their text more plain and more complete,” “The Samaritan Pentateuch and the Text of the Old Testament,” 1970, 224. E. Ulrich, “The Dead Sea Scrolls and the Biblical Text,” 1998, 87. E. Tov, TextualCriticism, 2001, 85f; id., DerTextderHebräischenBibel, chapter 2, I, B, 4, alfa (3): the expansions “zeigt die Absicht der Revisoren, die exakte Durchführung der Anordnung zu betonen.” Cf. F. Dexinger, ”Das Garizimgebot im Dekalog der Samaritaner,” Studien zum Pentateuch, W. Kornfeld zum 60. Geburtstag, ed. G. Braulik (Vienna: Herder), 1977: ”Man wird nicht fehlgehen, wenn man – ganz allgemein gesprochen – eine Tendenz zur (theologischen) Systematisierung mittels Textwiederholung sieht,” 130. A. Tal, art. “Pentateuch,” A. D. Crown et al., Companion, 1993, 179 J. Purvis, TheSamaritanPentateuch, HSM 2 (Cambridge, Mass.: Harvard University Press) 1968, 72, with reference to F. M. Cross, TheAncientLibraryofQumran, 193. Tov, TextualCriticism, 80.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

11

One notes the development in this assessment: the major expansions reveal an editorial activity rather than being harmonizations. Earlier, Hans Gerhard Kippenberg suggested that one of the three additions to Exod 20 SP was made in order to express the expectation for an eschatological prophet like Moses.49 Deut 5:28b.29 and 5:30-31 were first inserted after Exod 20:21a, and later Deut 18:18-22 was inserted into this insertion. The reason for the first insertion was to emphasize the primary position of the Decalogue in Exodus over that in Deuteronomy, and the second insertion was made because the possibility for false prophets had arisen in the form of the Judean and Israelite prophets, i.e. the prophetical books of the HB. Against the prophetical books, where words by Moses were not found, the Samaritans mobilized the eschatological prophet who would speak God’s commandments, statutes and ordinances in the way that Moses did. Kippenberg was aware of 4QpaleoExodm and 4QTest and accepted that the former text proved the fidelity of the transmission of the Samaritan text form, whereas the latter received the SP’s expectation for an eschatological prophet by quoting the beginning of Deut 5:28f; 18:18f. in the form given by the SP-addition in Exod 20:21b.50 The Samaritan expectation for an eschatological prophet was later abandoned by the Samaritans when the prophetic movement in Samaritanism, The Dositheans, appeared. Dositheus probably lived in the first century CE, and after this period the expectation for an eschatological prophet was partly transferred to Joshua and partly replaced by belief in Moses and Torah.51 Kippenberg’s interest in prophecy in the pre-Samaritan texts and the SP was taken up and developed further by Magnar Kartveit.52 Although we are here concerned with the SP-group, it should not be forgotten that we often have to take into consideration more of the available material. One example is Deut 27:4. Eugene Ulrich’s understanding is illustrative.53 His solution is this: in the first stage there was no name for the place of the altar in Deut 27:4; it was only said that the building should take place immediately after the crossing of the Jordan. In correspondence to this, the narrative with the building of the altar in the Book 49

50 51 52

53

H. G. Kippenberg, Garizim und Synagoge, RGVV 30 (Berlin: de Gruyter), 1971, 312. Ibid, 70f. Ibid, 326. Magnar Kartveit, TheOriginoftheSamaritans, VTS 128 (Leiden: Brill), 2009, 313349. E. Ulrich, “4QJosha,” DJD XIV, 1995, 145f.

12

MAGNAR KARTVEIT

of Joshua constituted the end of chapter 4. This arrangement is witnessed to by 4Q47 (4QJosha) and Josephus’ Antiquities 5:16-20. In the second stage the Samaritans inserted “on Mount Gerizim” into Deut 27:4, as witnessed by Vetus Latina and the SP. This second stage in Ulrich’s theory has, most likely, to be adjusted into a Jewish stage, not a Samaritan, but it was adopted by the Samaritans. The final stage is represented by the MT and LXX, which introduced “on Mount Ebal” into Deut 27:4 and into the narrative with the building of the altar in the Book of Joshua as a counterclaim to the Samaritan version. The pericope with the altar-building was then transposed to its present location in Josh 8 in MT and in Josh 9 in the LXX.54 The altar at Ebal “seems to make sense only as a countermove to the Samaritans’ claim for Mt. Gerizim.”55 A more complex theory is developed by Michaël Nicolaas van der Meer.56 His point of departure is that Deut 27:4-7 and Deut 17:11-13 were unrelated: one command to build an altar on Mount Ebal and another concerning a ceremony reflecting the political situation under Josiah. From this point the different texts developed, with one result attested by 4QJosha, and another attested in the LXX. Josephus represents a third line of development.57 A Hebrew text for Deut 27:4 has appeared with the reading “Mount Gerizim”. The provenance and authenticity of this text has yet to be confirmed. In the meantime, we are left with the SP, Vetus Latina and a Greek manuscript with the reading “Mount Gerizim”, probably meaning that an early text had this reading.58 Judith Sanderson based her explanation of the reason for the major expansions on Exod 7-11, and suggested that they were made because the point of these expansions is not to emphasize Moses’ strict obedience to Yahweh. The point is to emphasize Yahweh’s threats to Pharaoh. This is a narrative of warfare between two deities. Various scribes used various means to heighten the tension and to highlight the victory of Yahweh. Some scribes made minor changes to enhance the subplot of the conflict between Moses and the magicians (see 9:8[gimel] and 9:10). This particular scribe decided to make major expansions to enhance the central plot of the conflict between Yahweh and Pharaoh.59 54 55 56

57 58 59

Ibid. Ibid. Michaël Nicolaas van der Meer, FormationandReformulation:TheRedactionofthe BookofJoshuaintheLightoftheOldestTextualWitnesses, Diss., Leiden 2001. Ibid., 434f. Discussion in Kartveit, Origin, 300-309. J. Sanderson, AnExodusScroll, 1986, 203-4.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

13

She made an attempt at understanding the expansions from a theological viewpoint, but this explanation does not take into account the other expansions, and it is not further substantiated. It is an interesting suggestion, however, that the expansions in Exod 7-11 may be seen in light of a communication between Yahweh and Pharaoh. One wonders if this idea can be underpinned by finding a situation in which the expanded text would communicate into a conflict between Yahweh and a contemporary “Pharaoh”. In the Hellenistic period some foreign ruler would be a possible counterpart for the God of Israel. Sanderson thought that “While one individual may be responsible for the creation of the typological characteristics that establish the existence of this text-type, perhaps a series of scribes are responsible for the introduction of the many minor differences between the two members of this text-type”60, and she goes on to state that Perhaps expansionist texts were especially favored by groups hostile to Jerusalem, because circles in leadership positions in Jerusalem were – again, perhaps – attempting, possibly with a good degree of success, to establish a more conservative text – at least as far as the Torah is concerned – as authoritative. As some refused to worship in Jerusalem, so some may also have preferred to use biblical texts that were frowned on in Jerusalem... The major expansions in Exodus may have been made by one individual at Shechem or at Qumran or at some third place; or several scribes of a similar bent, if not necessarily working in the same place, may have been involved. Presumably texts with these major expansions were, if not created, at least popular, at Shechem for some time before the “sectarian” expansion was made, since that last expansion reveals the same approach to Scripture. For such an expansion to be conceived in the first place and to be accepted as both legitimate and powerful in the second place, the community must already have been accustomed to and must have valued the text as we know it from Qumran.61

If we combine these ideas with that of an enhanced conflict between Yahweh and Pharaoh in the expansions in Exod 7-11, we should look for groups and situations where the expansions would have a contemporary message. To undertake the editing solely because of the original conflict in Egypt would, in the logic of Sanderson, probably not be reason enough for this work.

60 61

Ibid., 314. Ibid., 317.

14

MAGNAR KARTVEIT

Alan D. Crown made the following suggestion: Comparisons result in the redating of some of the “expansionist” Samaritan readings and interpretations. It may well prove to be the case that expansions in the pre-Samaritan and Samaritan texts will be found to have nothing to do with matters of late, textual harmonization but rather reflect and preserve the halakhic traditions of those Jews who did not accept the Oral Law. In other words, they are the textual deposit of shared laws agreed by “One Torah” Jews in their common past.62

These Jews would be the Qumran community and the Samaritans. However, it is difficult to see any halakhic interest in the large expansions. One of the conspicuous traits about them is that they have not affected the legal material in the Pentateuch. On the other hand, there are omissions and additions in the Covenant Code that may reveal an interest in updating of the Torah, but there are no major expansions in the SP-group with reworkings of the legal material. In his study of the question, Magnar Kartveit found 39 larger expansions in the SP-group, and the expansions are unevenly distributed: ten of them are concentrated in Exod 6-11, and 15 are copied from Deut 1-3, one into Exod, the rest into Num. Exod 20 has received two long insertions, together covering 13 verses. Expansions and transpositions numbers 16-19 and 21 are in the large section dealing with the construction of the tabernacle, Exod 25-40. Expansion number 22 also has to do with the Tabernacle, this time supplementing a regulation for its transport. In the main, we are, therefore, dealing with four sets of expansions and transpositions: the plague narrative, the theophany at Sinai, the tabernacle construction, and the narrative counterparts of the historical survey in Deut 1-3.63

He further suggested that [the expansions] form one distinct layer in the pre-Samaritan texts and in the SP. This layer had Moses as its primary figure, and we may term it a ‘Moses-layer.’ This editing of the Pentateuch is in line with a series of other texts in the same corpus, emphasizing the status of Moses as a prophet. They are in presumed historical order, starting with the youngest: Deut 34:10; Num 12:6-8; Deut 18:9-22; Num 11:14-17, 24b-30.64 62

63 64

A. D. Crown, “Samaritans,” EncyclopediaoftheDeadSeaScrolls, (Oxford: University Press), 2000, 818. Kartveit, Origin, 276-7. Ibid., 280-81.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

15

In this way “the prophetic primacy of Moses” was secured, and One may hear behind these texts the discussion between Moses and the prophets, between the traditions behind the Pentateuch and later prophecy. In the pre-Samaritan texts and in the SP the result of the discussion is clear: Moses is the absolute authority from which the other prophets had to derive theirs.65

SUMMARY Skehan’s observation on the fidelity of transmission inside the SP-group has been retained and refined in later scholarship. Some observations have been made on the linguistic characteristics of the SP-group, but more research is needed in order to arrive at a precise understanding of the non-Samaritan, pre-Samaritan and Samaritan phenomena in this area, and how they are to be assessed within the larger picture. The parlance has shifted from Skehan’s “Samaritan recension” to Tov’s “SP-group”, and this reflects the observations on stability and variation inside the transmission. Several interesting suggestions have been made concerning the possible ideological interests behind the SP-group. In this area also, more investigation is needed. To some critics, the manuscript situation has removed the boundaries between literary criticism in the sense of Literarkritik and textual criticism. This creates a new methodological situation. Magnar Kartveit, prof. em. VID Specialized University, Stavanger, Norway

BIBLIOGRAPHY ANDERSON, R. T. “Jean Morin as a Catalyst in Modern Textual Criticism.” Pages 62-73 in ForaTestimony:EssaysinHonorofJohnH.Wilson. Edited by M. Haykin. Toronto: Central Baptist Seminary and Bible College, 1989. BAILLET, Maurice. “Le texts samaritain de l’Exode dans les manuscrits de Qumrân.” Pages 363-381 in HommagesàAndréDupont-Sommer. Edited by A. Caquot and M. Philonenko. Paris: Librairie d’Amérique et d’Orient Adrien-Maisonneuve, 1971. —. “Les divers états du Pentateuque samaritain.” RB 13 (1988): 531-545. 65

Ibid., 299.

16

MAGNAR KARTVEIT

BEN-ḤAYYIM, Ze᾿ev, with assistance from Abraham Tal. AGrammarofSamaritan Hebrew:BasedontheRecitationoftheLawinComparisonwiththeTiberian andOtherJewishTraditions. Revised Edition in English. Jerusalem: Magnes; Winona Lake, In: Eisenbrauns, 2000. BLAYNEY, Benjamin. Pentateuchus Hebraeo-Samaritanus charactere HebraeoChaldaicoeditus. Oxonii: Clarendinano, 1790. CROSS, Jr., FRANK M. “The Evolution of a Theory of Local Texts.” Pages 108126 in SeptuagintandCognateStudies. No. 2, edited by Robert A. Kraft. Missoula, Mont.: SBL, 1972. CROWN, Alan D. “Samaritans.” Pages 817-818 in EncyclopediaoftheDeadSea Scrolls. Edited by Lawrence Schiffman and James C. VanderKam. Oxford: University Press, 2000. DEXINGER, Ferdinand. “Das Garizimgebot im Dekalog der Samaritaner.” Pages 111-133 in StudienzumPentateuch. WalterKornfeldzum60.Geburtstag. Edited by Georg Braulik. Vienna: Herder, 1977. VON GALL, August Freiherr. DerHebräischePentateuchderSamaritaner. Giessen: Töpelmann, 1914-18. GERLEMAN, Gillis. Synoptic studies in the Old Testament. Lund universitets årsskrift, N. F., Avd. 1, 44:5. Lund: Gleerup, 1948. GESENIUS, Guilielmus [Wilhelm]. DePentateuchiSamaritaniorigine,indoleet auctoritate:commentatiophilologico-critica. Halle: Rengerianae, 1815. JASTRAM, Nathan Ray. “4QNumb.” Pages 205-267 in QumranCave4.VII:Genesis toNumbers. Edited by F. M. Cross and E. Ulrich. DJD 12. Oxford: Clarendon, 1994. KAHLE, Paul. “Untersuchungen zur Geschichte des Pentateuchtextes.” ThSK 38 (1915): 399-439. KARTVEIT, Magnar. TheOriginoftheSamaritans. VTS 128. Leiden: Brill, 2009. KENNICOTT,Benjaminus. VetusTestamentumHebraicum cumvariislectionibus. Oxonii: Clarendoniano, 1776. KIPPENBERG, Hans Gerhard. GarizimundSynagoge. RGVV 30. Berlin: de Gruyter, 1971. MORINUS, Ioannis. Biblia Hebraica, Samaritana, Chaldaica, Graeca, Syriaca, Latina,Arabica,quibustextusoriginalestotiusScripturaeSacrae. Vol. 6. Paris: Vitré & Gallicanus, 1629-1645. MACUCH, Rudolf. GrammatikdessamaritanischenHebräisch. Studia Samaritana 1. Berlin: de Gruyter, 1969. VAN DER MEER, Michaël Nicolaas. FormationandReformulation:TheRedaction oftheBookofJoshuaintheLightoftheOldestTextualWitnesses. Diss., Leiden, 2001. MURTONEN, Aimo E. Materials for a Non-Masoretic Hebrew Grammar III: A GrammaroftheSamaritanDialectofHebrew. StOr 29. Helsinki: Societas Orientalis Fennica, 1964. PURVIS, James. TheSamaritanPentateuch. HSM 2. Cambridge, Mass.: Harvard University Press, 1968. SANDERSON, Judith E. AnExodusScrollfromQumran:4QpaleoExodmandthe SamaritanTradition. HSS 30. Atlanta: Georgia: Scholars Press, 1986. —. “4QpaleoExodusm.” Pages 53-103 in QumranCave4.IV:Paleo-Hebrew andGreekBiblicalManuscripts. Edited by P. W. Skehan, J. E. Sanderson, and E. Ulrich. DJD 9. Oxford: Clarendon, 1992.

SCHOLARS’ ASSESSMENTS OF THE RELATIONSHIP

17

SKEHAN, Patrick W. “Exodus in the Samaritan Recension from Qumran.” JBL 74 (1955): 182-187. —. “Qumran and the Present State of Old Testament Studies: The Masoretic Text.” JBL 78 (1959): 21-25. S[TENHOUSE], G[illis] F[irmin]/P[aul] L[ester]. “Morin, Jean (1591-1659).” Pages 160-161 in ACompaniontoSamaritanStudies. Edited by Alan D. Crown, Reinhard Pummer, and Abraham Tal. Tübingen: Mohr Siebeck, 1993. TAL, Abraham. “Pentateuch. I. General Discussion.” Pages 177-179 in ACompaniontoSamaritanStudies. Edited by Alan D. Crown, Reinhard Pummer, and Abraham Tal. Tübingen: Mohr Siebeck, 1993. TALMON, Shemaryahu. “The Samaritan Pentateuch.” JJS 2 (1950-51): 144-150. T[OV], E[manuel]. “Pentateuch. IV. Proto-Samaritan Texts.” Pages 182-183 in A CompaniontoSamaritanStudies. Edited by Alan D. Crown, Reinhard Pummer, and Abraham Tal. Tübingen: Mohr Siebeck, 1993. TOV, Emanuel. DerTextderHebräischenBibel:HandbuchderTextkritik. Stuttgart: Kohlhammer, 1997. —. TextualCriticismoftheHebrewBible. Third edition revised and expanded. Minneapolis, MN: Fortress, 2012. ULRICH, Eugene. “The Dead Sea Scrolls and the Biblical Text.” Pages 79-100 in TheDeadSeaScrollsafterFiftyYears:AComprehensiveAssessment. Vol 1. Edited by P. W. Flint and J. C. VanderKam. Leiden: Brill, 1998. —. “4QJosha.” Pages 143-152 and plates XXXII-XXXIV in QumranCave4.IX: Deuteronomy,Joshua,Judges,Kings. Edited by E. Ulrich et al. DJD 14. Oxford: Clarendon, 1995. WALTKE, Bruce. ProlegomenatotheSamaritanPentateuch. PhD diss. Harvard, 1965. WALTON, Brian. “Animadversiones Samariticae,Textum hebraeum et samaritanum.” In S.S.BibliaPolyglottaComplectentiaTextusOriginales:Hebraicos, cumPentateuchoSamaritano… Vol. 1. London: Thomas Roycroft, 1657.

FROM POPULAR JEWISH LXX-SP TEXTS TO SEPARATE SECTARIAN TEXTS: INSIGHTS FROM THE DEAD SEA SCROLLS Emanuel TOV

1. INTRODUCTION The fate of texts changes over the course of the centuries because of historical developments. This happens all the time, also in our corner. For example, the Greek translation of the Septuagint was very much a Jewish enterprise, but in due course it became the major vehicle for the Christian Old Testament in Greek. Likewise, the Jewish Bible, the so-called Old Testament, was adopted not only by the Jewish people, but also by the early Christians, and thus became part of the Christian Bible as well. Simply put, at that point the Jewish community abandoned its original Greek translation of Hebrew Scripture, and turned to a number of so-called Jewish revisions that had been prepared in the meantime. Furthermore, in the period for which we have textual evidence, from the third to the first pre-Christian centuries, we witness textual plurality with regard to Hebrew Scripture, especially with reference to the Torah, which is the topic of this study. It may well be that such textual plurality characterized the fifth and fourth pre-Christian centuries as well, but these centuries are beyond our horizon. For the Torah alone, I count some ten different textual branches from the third century BCE onwards, while for the other books we witness only one, two, or sometimes three textual branches. Thus, in Judges, Job, Ruth, Qohelet and Lamentations, only a single textual tradition can be identified, since the textual witnesses do not differ significantly from one another. All this is in remarkable contrast to the period following the destruction of the Temple, when the Jewish people were thrown into disarray, the new religion of Christianity had arisen, and as a result the textual situation had changed completely with regard to the Hebrew Bible. At that time, the Jewish people adhered to only a single text, the Masoretic text (MT), and this situation has been explained in different ways by modern scholars. Two other major texts had been adopted by other religions, the Greek Septuagint had become the exclusive

20

EMANUEL TOV

possession of the new Christian religion and the Samaritan Pentateuch (SP) had become the central text for the Samaritans. The latter text continued the tradition of one of several pre-Samaritan texts and differed only very slightly from its Vorlage, and gradually became associated firmly with the new religion of the Samaritans. Beyond these groups, the Qumran community, once the proud owner of a major collection of biblical manuscripts, was now annihilated, and there was no one to continue its cultural heritage. When summarizing the textual situation after the year 70 CE, to the best of my knowledge, there were no organized groups in ancient Israel who could have fostered textual traditions other than MT, the Greek LXX, and the Samaritan SP. These textual conditions had been determined by the historical situation following the destruction of the temple. At the end of this study I will return to these generalizing remarks.

2. TEXTUAL HISTORY OF THE PENTATEUCH The main topic of the present study is not MT, but the non-Masoretic texts, especially the Samaritan Pentateuch, but I cannot develop this theme without discussing the full range of non-Masoretic texts. Among these texts, the SP and the pre-Samaritan texts were of special importance, more so than I had thought ten years ago. They form the so-called SP group, which included several additional Hebrew texts. In my view, at an earlier stage it was closely linked to the Vorlage of the LXX as these two texts derived from the same text block. The textualfeatures of this text block are the topic of this study. In an earlier study,1 I spoke about the division of the textual witnesses of the Torah into two major text blocks, and I now focus on their textual character. Before doing so, I return to some of the basic facts on which my view on the two text blocks is based. 1. LXX-SP. A major argument in my view is the assumption that the SP group and the LXX are closely connected. The assumption of a common ancestor of the LXX and the SP group was first surmised in the 1815 monograph by W. Gesenius, who guided the discussion of the SP and LXX in a sound direction.2 In Gesenius’ view, the two traditions derived from a common source that he named the “Alexandrino-Samaritan edition.”3 1 2 3

Tov, “The Development of the Text of the Torah in Two Major Text Blocks.” Gesenius, DePentateuchiSamaritaniorigineindole. Ibid., p. 14. Gesenius explained the background of the similarity between SP and the LXX by saying that “the Alexandrian translation and the Samaritan text derived from Judean codices which were similar to each other.” This text, adopted by both the Jews

INSIGHTS FROM THE DEAD SEA SCROLLS

21

Central in my analysis are both the large number of agreements between the SP and LXX, and their special nature.4 These two sources agree frequently in secondary readings in all the books of the Torah. For example, in most of the differences between the SP and MT in Jacob’s blessing in Genesis 49, the SP agrees with the LXX.5 This closeness is visible especially in their shared and separate harmonizing pluses, but also in individual readings. In each of the books of the Torah, the LXX contains even more harmonizations than the SP. Until one does a word-for-word analysis of each of the Pentateuchal books one does not realize how often the LXX and SP agree in secondary readings (see below). This agreement is extended to the so-called pre-Samaritan Qumran scrolls. Compared with MT, the two sources also have in common a revision of the genealogical lists in Genesis 5 and 11, in which revisional and hence secondary traits are recognizable.6 These combined data lead to the suggestion that the LXX and SP have a common background in secondary readings, even though they actually disagree as often as they agree.7 Although the books of the Torah differ in content, the LXX and SP must have undergone a similar textual development or they were based on a common base text in all five books, although at a later stage the two texts went in separate directions. CompositionsbasedonthecommontextbaseofLXX-SP. The assumption that the LXX and SP derived from a common text base is supported by the fact that several rewritten Bible compositions are closer to the common text of the LXX and the SP than to MT (11QTa, 4Q252 [4QComm

4

5

6 7

of Alexandria and the Samaritans in Palestine, removed many problems from the original text, and should therefore be characterized as secondary. Gesenius’ approach was followed by Kohn, De Pentateucho Samaritano and Nutt, Fragments of a Samaritan Targum, 98. For a detailed analysis of the close relation between the LXX and the SP group, see my study “The Shared Tradition of the Septuagint and the Samaritan Pentateuch.” This pertains to fourteen of the twenty different content differences between MT and the SP in verses 3, 5, 6, 7, 8, 10, 11, 12, 13, 14, 22, 23, 26. Tov, “The Genealogical Lists in Genesis 5 and 11 in Three Different Versions.” The use of secondary readings as a guiding principle in composing the stemma follows Paul Maas’ principle of Leitfehler (indicative errors): Maas, Textual Criticism, 42–49; trans. of “Textkritik,” in A. Gercke and E. Norden, eds., EinleitungindieAltertumswissenschaft, I, VII. These common secondary readings are so significant that the occurrence of a good number of them suffices to characterize textual witnesses. By the same token, the occurrence of a good number of common harmonizations in SP and the LXX suffices to characterize these two sources as textually close to each other. When this is recognized, the large deviations of the SP can be ascribed easily to a secondary factor (subsequent content editing of SP) even though these editorial manipulations are of a greater magnitude than the harmonizations themselves.

22

EMANUEL TOV

Gen A], Jubilees in its Ethiopic versions, Pseudo-Philo, Genesis Apocryphon, as well as 4QTestimonia). In fact, there are no rewritten Bible compositions that are based clearly on MT instead of the LXX and SP.8 An additional group of texts based on the common LXX-SP base are the liturgical texts, to be discussed below: two different branches of tefillin from Qumran9 and three liturgical Qumran texts that contain the same pericopes as the tefillin (4QDeutj,k1,n).10 Structure of the LXX-SP group. Determining relationships between manuscripts is precarious when so many ancient texts have been lost, but in the case of the Torah we can attempt to do so because the evidence seems to be reliable. It seems to me that the SP group distanced itself further from the common LXX-SPtradition than did the Vorlage of the LXX. This is suggested by the large editorial changes in Exodus 7–11 in the SP group, the addition of parallels to Deuteronomy 1–3 in Exodus and Numbers, and the addition of the tenth Samaritan commandment.11 At the same time, there are no unmistakable instances of major secondary elements in the Vorlage of the LXX,12 with the possible exception of Exodus 35–40.13 I therefore think that the LXX remained in closer proximity to the common LXX-SP base than did the pre-Samaritan texts and the SP. 2. Vorlage oftheLXX. The first textual tradition to branch off from the LXX-SP base was the Vorlage of the LXX, which reflects a greater number of late than early elements in the Torah. 8

9

10

11

12

13

For details, see my study “The Textual Base of the Biblical Quotations in Second Temple Compositions.” See my study “The Tefillin from the Judean Desert and the Textual Criticism of the Hebrew Bible.” Two groups of tefillin are closely related to the LXX-SP text base as opposed to MT, one written in a conservative spelling pattern (4QPhyl C, D-E-F, R, S; 4QPhyl 4) and the other written in the pattern of the Qumran scribal practice (4QPhyl A, B, G-H-I, J-K, L-N, O, P, Q). 4QDeutj contains sections from Deuteronomy 5, 8, 10, 11, 32 and Exodus 12, 13; 4QDeutk1 contains sections from Deuteronomy 5, 11, 32. 4QDeutn contains sections from Deuteronomy 8 and 5. In this list the sections from Deuteronomy 8 are not covered by the tefillin. See the study quoted in n. 4 and see my HebrewBible,GreekBible,andQumran, 57– 70. However, I note that the LXX, which in my view reflects an earlier stage in the development than the SP, contains many small harmonizing pluses in Numbers that adapt the text to Exodus, Leviticus, and Deuteronomy. See 9:14, 15 (cf. Lev 19:33); 14:10 (cf. Exod 16:10); 21:2 (Exod 34:15); 27:12 (Deut 32:49). See also the examples gathered and analyzed in Teeter, ScribalLaws, 118–74. See Aejmelaeus,“Septuagintal Translation Techniques,” 116–30 (118). See further my analysis in “The Source of Source Criticism.”

INSIGHTS FROM THE DEAD SEA SCROLLS

23

The reconstructed Hebrew source of the LXX translation of the Torah reflects a free approach to the text, like that of its base, the common LXXSP text. This freedom is reflected in the large number of small contextual harmonizations, far more so than in the SP group,14 which until recently was considered to be the most harmonizing text among the Scripture texts.15 The identification of these secondary features in the LXX of the Torah, not recognized by most scholars, is quite surprising. Remarkably, the nature of the Vorlage of the LXX in the Torah differs totally from the Vorlage of the LXX in all other Scripture books. However, the LXX also reflects important original readings in the Torah, such as in the song in Deuteronomy 32, where it is joined by 4QDeutq.16 3–4. SPgroup. The SP group branched off from the common LXX-SP source. According to our present knowledge, that SP group consists of three layers, in historical or typological sequence: a single pre-Samaritan text 4QNumb (3) reflecting a transition stage between the LXX and the SP group,17 the other pre-Samaritan texts (4), and the medieval texts (4a) that continued on from the pre-Samaritan texts.18 The pre-Samaritan nature of this group is recognizable in a number of pre-SP texts that are best described as pre-SP twigs (4) of the SP branch: 4QpaleoExodm, 4QExod-Levf, and possibly also 4QLevd.19 These three scrolls never lack an editorial addition of SP, but conversely, in one instance, 4QNumb (the combination of Numbers 27 and 36)20 contains an editorial intervention not found in SP. Group 4 is pre-Samaritan as it foreshadows the medieval SP text (4a). The pre-Samaritan texts contain a greater amount of harmonization than SP.21 The following are textual branches that depended on the SP group: 14

15 16 17

18

19 20

21

A large number of harmonizations of this type are also found in the pre-Samaritan texts 4QExod-Levf, 4QNumb, and 4QRPb, but statistical calculations need to be carried out cautiously as these texts are fragmentary. The common base of the harmonizing pluses in the LXX and SP makes it likely that the LXX pluses were rendered from a Hebrew source and were not created by the translators, as is also suggested by inner-Greek differences. For examples, see Appendix. See n. 32. See Tov, TCHB, 249–50 with literature. Typologically, 4QNumb probably presents the oldest representative of the SP-LXX group, reflecting more significant agreements with the LXX than the other texts. For the general background of the parting of the ways between Jews and Samaritans, see Knoppers, JewsandSamaritans, 178–216. 4QDeutn is not a pre-Samaritan text; see Owen, “4QDeutn: A Pre-Samaritan Text?” Partially reconstructed but supported by the juxtaposition of these chapters in 4QRPb, frg. 36. See my study “The Samaritan Pentateuch and the Dead Sea Scrolls.”

24

EMANUEL TOV

5. 4QReworkedPentateucha,b (4Q158, 4Q364) are very close to SP, but I cannot determine whether they are true pre-Samaritan texts or based on pre-Samaritan texts. At the same time, these texts differ substantially from SP since, unlike group 3–4, they inserted several elements not found in MT. We possess insufficient information regarding the exact relation between the two texts.22 6–7. 4QReworkedPentateuchc,d (4Q365–366). Two exegetical texts that branched off from the LXX-SP group form another cluster, each of them carrying individual features. They contain running biblical texts intertwined with large and small exegetical additions such as an expanded Song of Miriam in 4QRPc 6a ii and 6c, not equaled by any other source. Both texts are clearly closer to LXX-SP than to MT, but they are not members of that group. 8. Liturgical texts and tefillin (based on group 2–4). Several liturgical texts displaying more agreements with LXX-SP than with MT were probably based on the former group. These sources reflect a very free and harmonizing approach to the text: two different clusters of tefillin from Qumran (containing the required passages according to rabbinic sources and a mixture of required and non-required passages)23 and three liturgical Qumran texts that contain the same pericopes as the tefillin (4QDeutj,k1,n).24 In all these texts, harmonization, including the addition of small pericopes, is the main textual-editorial feature.25 These texts carried authority as liturgical texts, but not as Scripture texts. Interimsummary. When trying to understand the textual situation of the Torah in the Second Temple period we are faced with a veritable textual plurality, probably reflected in some ten different textual branches, which is more than in the other Scripture books. Moving beyond earlier textual theories, I suggested in the study quoted in n. 1 that this plurality should be reduced to a pattern of two large tradition blocks, MT and the remaining texts. Among these other texts, the (probably Palestinian) block consisting of the LXX and the SP group, together with their congeners, are the major component. This assumption is supported by the binary division of the 22

23 24

25

For groups 5–7, see the analysis in Tov and White, “364–367. 4QReworkedPentateuchb–e and 4QTemple”; Zahn, RethinkingRewrittenScripture. See n. 9. 4QDeutj contains sections from Deuteronomy 5, 8, 10, 11, 32 and Exodus 12, 13; 4QDeutk1 contains sections from Deuteronomy 5, 11, 32. The liturgical character of 4QDeutj is supported by its small size. See Tov, Hebrew Bible,GreekBible,andQumran, 37. Note further that both 4QDeutj and 4QDeutn start with Deut 5:1 and continue until the beginning of chapter 6. Both texts also contain a fragment that covers Deut 8:5–10. See Eshel, “4QDeutn,” 151.

INSIGHTS FROM THE DEAD SEA SCROLLS

25

textual character of the tefillin, which belong to either the MT or the LXXSP groups. It is further supported by the fact that the MT tradition is quoted only in the rabbinic literature, while the LXX-SP block served as the basis for the rewritten Bible compositions. Unlike block II, the MT block remained relatively clean of secondary readings. It is remarkable that among the biblical and non-biblical texts at Qumran, there is more evidence for block II than for block I.

3. THE CHARACTER OF THE

TWO TEXT BLOCKS

I now come to the main focus of this study. The two tradition blocks differ not only in content but also in character. The texts of block II are closely connected by links in common secondary features as opposed to mainly primary features in block I, MT. However, I stress that MT also contains some secondary features. The novel idea of subdividing the textual witnesses of the Torah into two text blocks is closely connected to the perception of two different scribal approaches, conservative and popularizing. The approaches themselves have been mentioned in the scholarly literature, but without connection to specific texts or text groups beyond the occasional mentioning of SP, 1QIsaa, and the Qumran scribal practice as examples of the popularizing texts, and MT as an example of a conservative text. The description of these approaches was not yet developed when Kahle described SP as a popularizing vulgar recension26 and MT as an official and reliable recension.27 It was more developed in the subsequent research, such as in my own writing in which I contrasted the conservative and popularizing approaches of copying and translating.28 A number of other scholars wrote about different scribal approaches, and a good summarizing analysis 26

27 28

Kahle, “Untersuchungen,” especially 5–12. Various scholars accepted from Kahle’s writings the concept of “vulgar” texts, albeit with certain changes. Nyberg, “Das textkritische Problem des Alten Testaments am Hoseabuche demonstriert”; Gerleman, SynopticStudies intheOldTestament; Greenberg, “Stabilization of the Text of the Hebrew Bible”; Lieberman, HellenisminJewishPalestine; and Kutscher, TheLanguageandLinguistic BackgroundoftheIsaiahScroll(1QIsa) (“vernacular and model texts”) posited in their descriptions the “careful” tradition of MT alongside “vulgar” texts. The scribes of these “vulgar” texts (e.g., 1QIsaa and SP) approached the biblical text in a free manner and inserted changes of various kinds, including orthography. Kahle, “Untersuchungen,” 35. Tov, TCHB, 184–85 and idem, “Approaches towards Scripture Embraced by the Ancient Greek Translators.”

26

EMANUEL TOV

was presented by D. A. Teeter.29 Teeter distinguished between two main approaches to the biblical witnesses that were given various names (conservative/official/standard, etc., as opposed to vulgar/popular/harmonizing/ interpretive, etc.). However, all these analyses were theoretical. I now connect the conservative approach with block I and the popularizing approach with block II in the Torah, and will attempt to assign most known texts to one of these two text blocks. In this binary division, the primary nature of the texts cannot be proven. The discussion thus moves to the presence of secondary readings, among which harmonizations take a central position. The texts of block I are characterized by the absence of secondary features, and those of block II are characterized by their presence. When stressing the secondary features of block II, I not only focus on elements that enable the characterization of these texts, but I also try to grasp their central features. It so happens that harmonizing additions represent the most characteristic textual feature of the LXX in the Torah. In a similar fashion, E. Eshel has argued that the pre-Samaritan scrolls should be named “harmonistic” and not “pre-Samaritan,” and she expanded that group to include texts such as 4Q158, tefillin, and mezuzot.30 I expand that group even further. What is harmonization? Harmonization is recognized when a detail in source A is changed to source B because they differ. Scribes adapted many elements in the text to other details in the same verse, the immediate or a similar context, the same book, or in parallel sections elsewhere in Scripture. Some of the examples pertain to strikingly different texts, but most of the examples of harmonization in the Torah are of a different type. They pertain to very similar texts in which harmonization seems to be rather inappropriate. Found frequently in direct speech, they often pertain to small literary differences in formulation that were harmonized because of a formal approach to Scripture, according to which a Godinspired text should only include formulations that are perfectly in harmony with one another. Some such changes must have been inserted unconsciously, but most were inserted because of a theological concern for perfection, especially in harmonizing pluses. Harmonization is by definition a secondary feature, and in SP and LXX it is coupled with similar secondary features such as various forms of adaptation of details to the context.

29 30

Teeter, ScribalLaws, 240 and passim. Eshel, “4QDeutn.”

INSIGHTS FROM THE DEAD SEA SCROLLS

27

My working hypothesis is that the texts that I have assigned to text block II are characterized by secondary textual features and the one text that is assigned to block I, MT, carries far fewer such features. When reviewing the various witnesses in light of this working hypothesis, we wonder how we can prove that a text carries secondary characteristics. Kutscher considered the large Isaiah scroll (1QIsaa) secondary as compared to MT.31 He made a good case, since he subjected most readings of that scroll to a detailed linguistic analysis and expressed a judgment on the basis of that analysis. However, that analysis was only linguistic, and exegetical changes were not included. His coverage was incomplete, and therefore his judgment was also incomplete. Nevertheless, I agree with Kutscher’s judgment, since the linguistic analysis covered most of the characteristic readings. In a similar way I will remark on what I consider to be the most characteristic groups of readings in the main texts of the second block. 3.1. LXX and SP: Harmonizations In my research of the last few years, it came as a great surprise to find so many secondary features in the LXX of the Torah. The LXX reflects many primary readings, but the textual feature that characterizes the Vorlage of that translation most is its many harmonizing pluses. These harmonizations (mainly: harmonizing pluses) appear more frequently in the LXX than any other textual phenomenon such as interchanges of letters, scribal mistakes, literary variants, or anything else. Further, these harmonizing pluses appear more frequently in the LXX than in any other textual witness in the Torah or any other Scripture book with the possible exception of the pre-Samaritan scrolls. Several of the harmonizations are clearly secondary, and many of them are shared with the SP. The common base of the harmonizations of the SP and LXX makes it likely that the Greek pluses were rendered from Hebrew, and were not created by the Greek translator. In addition, internal differences between these Greek renderings also indicate that the harmonizations were not carried out at the Greek level (for all these, see Appendix). Most other groups of secondary readings that are found in SP cannot be identified in the LXX because the LXX is a translation in which subtle differences between Hebrew readings, such as grammatical adaptations, are not recognizable. 31

Kutscher, TheLanguageandLinguisticBackgroundoftheIsaiahScroll(1QIsa).

28

EMANUEL TOV

The working hypothesis can be tested via statistics provided in my studies on Genesis and Deuteronomy,32 and those in studies on Exodus Leviticus, and Numbers.33 For example, the total numbers of the harmonizing readings are quoted as follows for Genesis: 1. 2. 3. 4.

LXX ≠ MT SP (196) SP LXX ≠ MT (63) SP ≠ MT LXX (40) MT SP ≠ LXX (47)

These figures show the distribution of the harmonizations in Genesis. The LXX contains by far the greatest number of unique harmonizations (196) in Genesis, with SP (40) trailing far behind, and with no instances for MT. Since all three sources also share harmonizations with the other two, these instances have to be added to each source’s unique features, in which case the numbers for the three sources are:34 LXX SP MT

259 150 47

A similar pattern is found in the other Pentateuchal books:35 Table 1: Combined Figures of Harmonizations in the Pentateuch

Genesis 1–11 Genesis 12–50 Genesis (total) Exodus 1–2436 Leviticus Numbers Deuteronomy 32

33

34

35 36

LXX

SP

MT

61 196 259 169 281 224 134

31 120 150 66 108 105 93

11 36 47 18 25 44 54

“Textual Harmonizations in the Ancient Texts of Deuteronomy”; “Textual Harmonization in the Stories of the Patriarchs”; “The Harmonizing Character of the Septuagint of Genesis 1–11.” For earlier studies, see especially Kim, “Studies in the Relationship between the Samaritan Pentateuch and the Septuagint” and further Hendel, TheTextof Genesis1–11. “Textual Harmonizations in Leviticus,” forthcoming; “The Septuagint of Numbers as a Harmonizing Text”; “Textual Harmonization in Exodus 1–24.” Summarizing article: “Textual Harmonization in the Five Books of the Torah: A Summary,” TheBible,Qumran,andtheSamaritans, ed. Magnar Kartveit and Gary N. Knoppers, SJ 104, Studia Samaritana 10 (Berlin: De Gruyter, 2018), 31–56. These numbers do not constitute the combined number of harmonizations in these chapters, since many instances are shared by two sources. Sources: the studies quoted in nn. 32–33. Based on my study “Textual Harmonizations in Exodus 1–24.”

29

INSIGHTS FROM THE DEAD SEA SCROLLS

These figures provide the total numbers of harmonizations in the books of the Torah, among which it is important to recognize the unique occurrences: Table 2: Unique Harmonizations

Genesis 1–11 Genesis 12–50 Genesis Exodus Leviticus Numbers Deuteronomy

LXX

SP

MT

51 145 196 139 201 179 99

9 31 40 18 8 16 22

0 0 0 2 5 1 2

The LXX stands out not only regarding the number of its harmonizations, but also with relation to their nature. The harmonizations in that source are more extensive than those in the SP and the MT. For example, most of the common harmonizations in the SP and MT consist of single words supplying a subject or object, while the LXX also contains longer harmonizations, sometimes shared with the SP.37 It should be noted that MT contains very few cases of unique harmonization together with several significant instances of harmonization common with other sources, such as the plus of Deut 1:39 ‫אשר אמרתם לבז יהיה‬ in MT SP, which was added to the text of LXX and 4QDeuth (apparently) based on the parallel verse Num 14:31.38 3.2. Pre-Samaritan Texts: Harmonizations Harmonizations can be examined in the pre-Samaritan texts, albeit not in Genesis because no such texts have been preserved for that book.39 The main stumbling block in this analysis is the fragmentary nature of the Qumran scrolls, which complicates statistical analysis. I found more harmonizations in the pre-Samaritan scrolls than in the other sources, as indicated in Table 3:

37

38 39

See Tov, “Textual Harmonization in the Stories of the Patriarchs,” § 4c and the studies on Exodus, Leviticus, and Numbers quoted in n. 33. See the analysis of Rofé, “Historico-Literary Aspects of the Qumran Biblical Scrolls,” 33. See my study “The Samaritan Pentateuch and the Dead Sea Scrolls.”

30

EMANUEL TOV

Table 3: Harmonizations in the Pre-Samaritan Scrolls Compared with Other Texts Scroll m

4QpaleoExod

4QExod-Levf 4QNumb 4QRPa (4Q158) 4QRPb (4Q364)

SP

LXX

MT

8

12

9

5

8 15 0 10

6 1 0 5

3 6 0 7

4 1 0 5

Some of these harmonizations are common to two or more texts, while others are unique. The phenomenon is well rooted in these texts, and it was already found in the pre-Samaritan texts, on one of which SP was based. These figures show that the largest amount of harmonization is found in 4QExod-Levf, 4QNumb, and 4QRPb, while for the data covered by 4QpaleoExodm, SP contains the largest amount.40 3.3. Samaritan Pentateuch: Secondary Readings Since the epochmaking study of Gesenius in 1815,41 SP has been described as reflecting many secondary features. Gesenius himself found only four original readings in that version, and what is more significant is that SP is replete with several groups of secondary readings that have been defined in different ways by scholars. In my analysis of the typical secondary readings of SP, I include many instances of small harmonizing changes (mainly pluses) that are discussed above (see further the Appendix), and elsewhere.42 In this analysis, I do not include the large editorial changes in SP for which it is famous, viz., the narrative duplications in Exodus and Numbers based on Deuteronomy 1–3 and the execution of commands in the story 40

41 42

Coincidence plays a part in the analysis of the fragmentary scrolls since harmonizations do not appear with the same frequency in all chapters. For example, no cases of harmonization are spotted in the fragments of 4QRPa. Gesenius, DePentateuchiSamaritaniorigineindole,61–64. See the replacement of unusual forms with regular ones in my TCHB (Table 12, chapter 2), grammatical adaptations (Table 13), and facilitating readings (some examples in Table 14). The secondary nature of other groups of readings is less clear: correction of unusual spellings (Table 11), correction of morphological forms (Table 15), and synonymous words (Table 16). In these instances, it cannot be proven that SP indeed replaced an earlier reading like that in MT as SP could also reflect an alternative reading.

INSIGHTS FROM THE DEAD SEA SCROLLS

31

of the plagues in Exodus 7–10.43 Traditionally, these editorial changes have been named “harmonizing,” and some form of harmonizing is indeed involved, but it would be more correct to regard them as part of the overall editing of SP, probably executed by a single person who followed his literary intuition. Obviously, all these duplications are secondary when compared with the original text, but they are exegetical and not typical secondary features. They may be considered sophisticated literary exercises in rewriting, not considered as being “vulgar.” It was suggested by Ben-Dov in a stimulating study that this rewriting should be thought of as academic and was probably influenced by similar expansions of Homeric songs in contemporary Alexandrian scholarship.44 Indeed, even though the nature of the expansions in the Ptolemaic Iliad papyri is very different from those in the SP group,45 many of the Homeric expansions do repeat verses occurring elsewhere46 and the two literatures have in common a literary and almost academic type of expansion. In the case of the SP group, its base text, together with the Vorlage of the LXX, contains a popular text, but these expansions were not popular. In any event, these major editorial changes are not included in my characterization of SP as a secondary text, and neither are the orthographical, phonological, and morphological features of SP, although many of them are also secondary.

43

44

45

46

For a detailed analysis, see my study “Rewritten Bible Compositions and Biblical Manuscripts, with Special Attention to the Samaritan Pentateuch.” A similar view on the nature of these editorial interventions was adopted by Segal, “The Text of the Hebrew Bible in Light of the Dead Sea Scrolls,” 13–16; Ben-Dov, “Early Texts,” 221; Zahn, RethinkingRewrittenScripture, 147. Ben-Dov, “Early Texts.” A similar suggestion has been made independently by Michael van der Meer elsewhere in this volume (pp. 47ss), based on recent studies on expansions in papyri of the Iliad. On the other hand, Zahn, “The Samaritan Pentateuch” is not certain that the Alexandrian milieu played a decisive role in the creation of these editorial changes. Zahn wonders against which background the pre-Samaritan texts developed their harmonizing tendencies. For a recent analysis, see Bird, Multitextuality. I am grateful to Michael van der Meer for bringing this book to my attention during the Strasbourg meeting. For an earlier analysis and for the text of the papyri themselves, see West, ThePtolemaicPapyriof Homer. For details, see the listings and analysis of Bird, Multitextuality, 75–100. Therefore, for the Homeric expansions, the discussion centers round the question of whether or not they are authentic and how such authenticity needs to be defined in a literature that grew through oral expansion, while the SP expansions are clearly not authentic. The Homeric expansions are spotted in the early Ptolemaic papyri starting with the third century BCE and their presence decreases in the following two centuries. See Bird, Multitextuality, 61–100.

32

EMANUEL TOV

3.4. Liturgical Torah Scrolls and Tefillin: Harmonizations Among the Qumran texts there are three liturgical texts that contain the same pericopes as the tefillin (4QDeutj,k1,n); a large number of such tefillin were found in the Judean Desert, especially at Qumran. These tefillin are often treated as idiosyncratic biblical texts, but actually they are ordinary biblical texts, since the textual data in the tefillin provide evidence of the Bible text that is as good as any other fragment from the Judean Desert. No liturgical texts based on MT are in evidence, but a few MT-like and proto-MT tefillin reflect MT, and contain only the pericopes required by the rabbis (MurPhyl, 34SePhyl, 8QPhyl I). These tefillin presumably derived from proto-rabbinic circles, and they do not reflect any harmonizations. A larger group of tefillin, based on the LXX-SP text block, contained two different branches of tefillin incorporating a mixture of required and non-required passages.47 The distinction between the two groups is not watertight since a few tefillin that contain only the required passages occupy an intermediate position.48 The LXX-SP tefillin reflect many harmonizations and, as in the case of the LXX, the occurrence of harmonizations is their main textual feature. For example, most of the differences between 4QPhyl A and the other texts may be ascribed to various types of harmonization.49 In the case of 4QDeutn, this was shown in detail by Esther Eshel, who analyzed that scroll’s many harmonizing changes (17), especially in the Decalogue in Deuteronomy 5.50 A striking example is the long addition of the argument for the Sabbath in Exod 20:8–11 (the creation of the world in six days) in the Deuteronomy scroll, added after the argument of Deut 5:15 (remembrance of the slavery in Egypt). This harmonizing addition in 4QDeutn links us directly to the tefillin, since two tefillin and the Nash Papyrus51 replace the argument for the Shabbat in the text of Deuteronomy with that in Exodus, viz., 4QPhyl G and 8QPhyl III.52 47 48

49

50 51 52

See n. 9. 4QPhyl C and 4QPhyl D-E-F (SP-LXX). In the last centuries BCE and the first two centuries CE the rabbinic circles were apparently less strict than in later times. I classify twelve of the variants (frequently shared with the LXX and/or SP) as harmonistic: Deut 10:13, 21; 11:2, 4, 6, 7, 8, 10, 11, 13; Exod 12:50; 13:3, 5 and nine as representing various textual categories: Deut 5:3, 32; 10:13, 18, 18, 22; 11:10, 12, 16. Eshel, “4QDeutn,” 142–47. See Albright, “Biblical Fragment”; Cook, “Pre-Massoretic Biblical Papyrus.” For the texts from cave 4, see Milik, DJD VI, 33–79; for 8QPhyl, see idem, DJD III, 154.

INSIGHTS FROM THE DEAD SEA SCROLLS

33

The group of tefillin thus supports the basic distinction between the two text blocks in ancient Israel, which may be linked to different religious environments. 3.5. Exegetical Texts In the previous four paragraphs, harmonization and other secondary features have been described as being characteristic of texts in the second text block. In other cases, the secondary features are expressed by the insertion of exegetical elements. 4QReworkedPentateucha,b (4Q158, 4Q364) have been described as very close to the SP group and possibly as part of it, but these texts further developed that tradition with exegetical additions. The same pertains to 4QRPc,d (4Q365–366), which move further away from SP with their exegetical changes. 3.6. Developed Orthographical and Morphological Features I left orthographical and morphological features to the end because it is difficult to evaluate this category. However, I do realize that the orthography of the Hebrew Bible developed in a particular way that left a more original form of orthography in the first text block and different and later forms of orthography in the second text block. My point of departure is that the Hebrew orthography developed from defective to full. For Hebrew Scripture, the most defective form is evidenced outside the Torah in 4QSamb. The stages of development have been indicated elsewhere,53 and there is a special interest in practices that differ greatly from MT and SP. Indeed, several Torah scrolls reflect the same orthographical and morphological features as the large Isaiah scroll (the “Qumran scribal practice”), namely 1QDeuta, 4QExodb, 4QNumb, 4QDeutj,k1,k2,m, and 4QRPa,b,c (4Q158, 364, 365).54 In these scrolls, orthography, morphology and secondary content features go together. As an example, I take 4QExodb: most of the deviations from MT in this scroll, which is written in the Qumran scribal practice, are secondary.55 This scroll, as well as the other mentioned Torah scrolls, 53

54

55

Tov, TCHB, 211–12 with bibliography; “Orthographic Practices of the Biblical Texts,” forthcoming. Some of these scrolls have been analyzed above in other categories among the preSamaritan scrolls (4QNumb, 4QRPa,b,c), while the others have not been mentioned above. In the analysis of Cross, “13. 4QExodb,” 83, this scroll contains ten secondary readings as opposed to eight primary readings, while eight cases are undecided. In my own

34

EMANUEL TOV

represents a free approach in which the scribes renewed the orthography and morphology and incorporated exegetical and harmonizing elements. When summarizing the features of the second text block, I note that harmonization is the central textual phenomenon in several texts, often joined by exegetical changes. The insertion of harmonizations presented holy Scripture more consistently. The addition of exegetical elements, the removal of difficult forms, and the use of spelling systems that were meant to facilitate the reading made the text more suitable for general use. All these tendencies existed also in MT, but in such a minimal way that enables us to distinguish between the two text blocks. The main means of differentiation between the two text blocks in the Torah remains the fact that MT was transmitted with a great deal of conservatism, preserving several archaic features that are not known from the other Scripture books,56 while the second text block is one of innovation and change. 4. IMPLICATIONS The beginning and end of the story are clear, but the middle part is speculative. The SP group started off as a non-sectarian and, if I am right, a popular text. Together with the source of the LXX, the SP group may have been the majority text, but it is not easy to prove this aspect. However, it cannot be coincidental that all the rewritten Bible texts are based on texts that reflect SP, LXX, or a combination of the two. These two texts represent also the majority of the texts of the Torah found at Qumran, outnumbering MT.57 The fact that these scrolls have been preserved in two different scripts may also be an indication of their distribution, although I have no explanation for this situation.58

56

57

58

analysis, this scroll contains twenty secondary readings and a few primary readings. Most of the secondary readings are pluses, and most of them are shared with the LXX. Of these twenty cases, eight are harmonizing pluses (1:18; 2:6, 11; 3:16, 16, 18; 4:8, 8), ten are explicative pluses or changes (1:1; 2:3, 3, 13, 14 [4x], 16, 16), and two are simplified linguistic forms (2:6; 3:19). See my studies “Orthographic Practices of the Biblical Texts,” and “The Enigma of the Masoretic Text,” Theologie und Textgeschichte, Septuaginta und Masoretischer Text alsÄußerungentheologischerReflexion,ed. Frank Ueberschaer, Thomas Wagner, and Jonathan Miles Robker,WUNT 407 (Tübingen: Mohr Siebeck, 2018), 45–70. There are more texts from block II than from block I at Qumran: there are more straightforward texts of LXX-SP (6) than of MT (4), and a large number of texts that reflect indirectly the tradition of block II: three liturgical texts (4QDeutj,k1,n); seven tefillin written in a conservative spelling pattern (4QPhyl C, D, E, F, R, S; 4QPhyl 4); and several rewritten Bible compositions (11QTa, 4Q252, GenApocr, 4QTest). In addition, there are twenty fragmentary texts that could reflect either MT or SP; in the past I included them with MT, but I now realize that they cannot be included in any statistics. All the scrolls are written in the square Aramaic script with the exception of 4QpaleoExodm, which is written in paleo-Hebrew.

INSIGHTS FROM THE DEAD SEA SCROLLS

35

At the outset, the SP group belonged to the same milieu as the source of the LXX. At a later stage, possibly in the fourth or third century BCE, their paths diverged. The earliest pre-Samaritan text is 4QExod-Levf, which is dated to the middle of the third century BCE. At some point, probably in the second century, one of the pre-Samaritan texts was adopted by the Samaritan community as authoritative Scripture. The acceptance of such a text was natural as the Samaritans took the popular text that was in the hands of the people. No choice of a text was involved, except for the fact that MT was not chosen; that was a deliberate act since the Samaritans separated themselves from the Judaism that cherished MT. In the words of M. Zahn, until the time of the schism, the Samaritan scribes owned the same Torah as the Jewish scribes.59 Returning to the title of this study, the SP started off in a completely different fashion from the way it developed during the last centuries BCE and became known in the Middle Ages. This pertains also to the source of the LXX, which in my view had a common text base with the SP before they parted ways. Subsequently, that text was translated into Greek and in due course it became the major vehicle for the Christian Old Testament in Greek. That is why I named my study “From Popular Jewish LXX-SP Texts to Separate Sectarian Texts: Insights from the Dead Sea Scrolls.” The Dead Sea Scrolls have aided us in realizing that the two texts that are at the center of two religious movements, the SP and LXX, once had a central position in Judaism as popular texts. 5. APPENDIX This appendix exemplifies harmonizing pluses in the LXX and SP group. This phenomenon is illustrated by these examples, but not proven by them. It would be hard to provide absolute proof of the existence of any textual feature, because that can be proven best by a conglomeration of examples that are not contradicted by a large number of counterexamples.

EXAMPLES OF HARMONIZING PLUSES

IN

LXX AND THE SP GROUP

a. SP LXX Gen 20:14 MT SP LXX (‫;ויקח אבימלך )צאן ובקר ועבדים ושׁפחת‬ SP LXX + ‫ אלף כסף‬+ (χίλια δίδραχμα). Based on v. 16 MT SP LXX.

59

Zahn, “The Samaritan Pentateuch,” 306–07.

36

EMANUEL TOV

The harmonization in this verse reveals its secondary nature. According to v. 14 MT, Abimelech gave Abraham “sheep and oxen, and male and female slaves,” but according to v. 16 MT SP LXX he told Sarah that he had given him “a thousand pieces of silver.” That monetary unit probably represented the monetary value of the items he had given Abraham according to v. 14. However, the SP LXX version of v. 14 added this detail from v. 16, and thus according to that version Abraham received twice as much in reparation. Lev 5:6 MT SP LXX ‫ ;וכפר עליו הכהן מחטאתו‬SP LXX + ‫אשר חטא‬ ‫ ונסלח לו‬+ (ἧς ἥμαρτεν, καὶ ἀφεθήσεται αὐτῷ). Based on v. 10. Num 21:21 MT SP LXX ‫וישלח ישראל מלאכים אל סיח)ו(ן מלך האמרי‬ (LXX: Μωυσῆς) SP LXX + ‫( דברי שלום‬λόγοις εἰρηνικοῖς); cf. Deut 2:26 MT SP LXX ‫ואשלח)ה( מלאכים ממדבר קדמות אל סיחון מלך חשבון דברי שלום‬

b. LXX Gen 24:14 MT SP LXX ‫ ;אשׁקה‬LXX + ‫ עד אם כלו לשׁתת‬+ (ἕως ἂν παύσωνται πίνουσαι). Based on v. 19 MT SP LXX ‫)אשׁאב( עד אם כלו‬ ‫לשׁתת‬.

The influence did not take place at the Greek level: the LXX in v. 19 probably reflects etymological exegesis of (‫( כל)ם‬ἕως ἂν πᾶσαι πίωσιν) as opposed to v. 14 ‫כלו‬. Gen 24:44 MT SP LXX ‫ ;אשׁר הכיח יהוה לבן אדני‬LXX + ‫לעבדו ליצחק‬ ‫ ובה אדע כי עשׂית חסד עם אדני אברהם‬+ (τῷ ἑαυτοῦ θεράποντι Ισαάκ καὶ ἐν τούτῳ γνώσομαι ὅτι πεποίηκας ἔλεος τῷ κυρίῳ μου Αβράαμ). Based on v. 14 MT SP LXX ‫הכחת לעבדך ליצחק ובה אדע כי עשׂית חסד‬ ‫עם אדני אברהם‬.

The influence took place at the Hebrew level since the Greek equivalents differ: In v. 14, ‫ עבד‬is rendered with τῷ παιδί σου, and for πεποίηκας v. 14 has ἐποίησας. Gen 32:20 (19) MT SP LXX (‫ ;ויצו )גם את השׁני גם את השׁלישׁי‬LXX + ‫ את הראשׁון‬+ (τῷ πρώτῳ). Based on v. 18(17) MT SP LXX ‫ויצו את‬ ‫הראשׁון‬.

This verse quotes Jacob’s words to the second and third servants, but the Hebrew parent text of the LXX found it necessary to complete the picture by adding the “first.” However, the “first one” was already mentioned in v. 18 (17), making this addition superfluous.

INSIGHTS FROM THE DEAD SEA SCROLLS

37

Num 15:36 MT SP LXX ‫ויציאו אתו כל העדה אל מחוץ למחנה וירגמו אתו‬ ‫ ;באבנים‬LXX + ‫ מחוץ למחנה‬+ (ἔξω τῆς παρεμβολῆς). Based on the context. Same addition in MT SP in the previous verse, 15:35. Good example of a superfluous addition.

c. SP Gen 10:19 MT LXX ‫מצידן באכה גררה עד עזה באכה סדמה ועמרה‬ ‫ ;ואדמה וצבים עד לשע‬SP ‫מנהר מצרים עד הנהר הגדול נהר פרת ועד הים‬ ‫האחרון‬. For the borders of the Canaanites, SP combines Gen 15:18 and Deut 11:24. Gen 11:31 MT LXX ‫ ;ואת שרי כלתו אשת אברם בנו‬SP ‫ ואת‬+ ‫ואת שרי‬ ‫ בניו‬+ ‫ ונחור‬+ ‫ כלותו אשת אברם‬+ ‫ = מלכה‬v. 29.

The plus of SP is based on v. 29, where the two daughters-in-law of Terah are mentioned together. SP could not imagine that Abraham would have left Ur Kasdim with his wife while Nahor left without his wife Milkah, and therefore Milkah was added in the context. However, the position of the plus in SP betrays its secondary nature: MT LXX “Sarai, his daughter-in-law, the wife of his son Abraham” has been changed in SP to “Sarai and Milkah, his daughters-in-law [note the hybrid form ‫]כלותו‬, the wife of Abram and Nahor his sons.” The addition of Nahor in SP is equally as secondary as the change of “his son” to “his sons.” Deut 4:49 MT SP ‫ ;ועד ים הערבה‬SP + ‫ ים המלח‬+ based on 3:17 ‫ועד ים‬ ‫הערבה ים המלח‬ Deut 11:6 MT SP LXX ‫ואשר עשה לדתן ולאבירם בני אליאב בן ראובן‬ ‫ ;אשר פצתה הארץ את פיה ותבלעם‬SP + ‫ ואת כל האדם אשר לקרח‬+ = Num 16:32

d. Pre-Samaritan Scrolls Exod 39:21a10 ‫ האפד‬MT 4Q SP LXX ] 4QpaleoExodm + ‫כאשר צוה‬ ‫ ;יהוה את משה‬cf. v. 21b. Beyond v. 21b, where this phrase is found in all sources, it is found eight times elsewhere in chapter 39, more than in any other chapter in Scripture. Exodus 40 (7×) and Leviticus 8 (altogether 6×) come close. Exod 40:17 ‫ בשנה השנית‬MT 4QExod-Levf SP LXX ] 4QExod-Levf SP LXX + ‫ ;לצאתם ממצרים‬cf. 16:1 ‫לחדש השני לצאתם מארץ מצרים‬. Num 20:20 ‫ לא תעבר‬MT SP LXX ] 4QNumb + ‫;פן בחרב אצא לקראתכה‬ cf. v. 18 ‫ לא תעבר בי פן בחרב אצא לקראתך‬MT. The words ‫ לא תעבר‬in v. 20 triggered the addition from v. 18 in 4QNumb. Num 22:11 ‫ ויכס את עין הארץ‬MT SP (LXX) ] 4QNumb + ‫והואה יושב‬ [‫ = ]ממולי‬LXX; cf. v. 5 ‫הנה כסה את עין הארץ והוא ישב ממלי‬.

38

EMANUEL TOV

Num 22:16 ‫ בלק בן צפור‬MT SP LXX ] 4QNumb + ‫ ;מלך מואב‬cf. v. 10 ‫בלק בן צפר מלך מואב‬ Num 22:19 ‫ מה יסף יהוה דבר עמי‬MT SP LXX ] 4QNumb + [‫וישבו ]שרי‬ ‫ ;מואב עם בלעם‬cf. v. 8 ‫כאשר ידבר יהוה אלי וישבו שרי מואב עם בלעם‬. Num 35:21 MT SP LXX ‫ ] וימת מות יומת המכה רצח הוא‬LXX 4QNumb +‫ מות יומת הרצח‬+ (θανάτῳ θανατούσθω ὁ φονευτής). Based on v. 18 ‫( מות יומת הרצח‬θανάτῳ θανατούσθω ὁ φονευτής). Even though the verdict has already been pronounced in this verse, the plus, which repeats the formulation of v. 18, is tautological.

BIBLIOGRAPHY AEJMELAEUS, Anneli. “Septuagintal Translation Techniques: A Solution to the Problem of the Tabernacle Account.” Pages 116–30 in eadem, OntheTrail oftheSeptuagintTranslators:CollectedEssays.Kampen: Kok Pharos, 1993. ALBRIGHT, William Foxwell. “A Biblical Fragment from the Maccabaean Age: The Nash Papyrus.” JBL 56 (1937): 145–76. BEN-DOV, Jonathan. “Early Texts of the Torah: Revisiting the Greek Scholarly Context.” JAJ 4 (2013): 210–34. BIRD, Graeme D. MultitextualityintheHomericIliad:TheWitnessofthePtolemaicPapyri. Hellenic Studies 43. Center for Hellenic Studies. Cambridge: Harvard University Press, 2010. COOK, Stanley A. “A Pre-Massoretic Biblical Papyrus.” Proceedings of the Society of Biblical Archaeology 25 (1903): 34–56. CROSS, Frank Moore, Jr. “13. 4QExodb.” Pages 79–95 in QumranCave4.VII: GenesistoNumbers. Edited by Eugene Ulrich and Frank M. Cross. DJD XII. Oxford: Clarendon, 1994. ESHEL, Esther. “4QDeutn: A Text That Has Undergone Harmonistic Editing.” HUCA62 (1991): 117–54. GERLEMAN, Gillis. SynopticStudies intheOldTestament. Lund: Gleerup, 1948. GESENIUS, Wilhelm. DePentateuchiSamaritaniorigineindoleetauctoritatecommentatiophilologico-critica.Halle: Bibliotheca Rengeriana, 1815. GREENBERG, Moshe. “The Stabilization of the Text of the Hebrew Bible Reviewed in the Light of the Biblical Materials from the Judean Desert.” JAOS 76 (1956): 157–67. HENDEL, Ronald S. TheTextofGenesis1–11:TextualStudiesandCriticalEdition. New York/Oxford: Oxford University Press, 1998. KAHLE, Paul. “Untersuchungen zur Geschichte des Pentateuchtextes.” Pages 3–37 in OperaMinora. Leiden: Brill, 1956. KIM, Kyung-Rae. “Studies in the Relationship between the Samaritan Pentateuch and the Septuagint.” PhD diss., Hebrew University of Jerusalem, 1994. KNOPPERS, Gary N. JewsandSamaritans:TheOriginsandHistoryofTheirEarly Relations. Oxford: Oxford University Press, 2013. KOHN, Samuel. De Pentateucho Samaritano ejusque cum versionibus antiquis nexu.Leipzig: Kreysing, 1865. KUTSCHER, Edward Y. The Language and Linguistic Background of the Isaiah Scroll(1QIsa). STDJ 6. Leiden: Brill, 1974.

INSIGHTS FROM THE DEAD SEA SCROLLS

39

LIEBERMAN, Saul. Hellenism in Jewish Palestine. 2nd ed. New York: Jewish Theological Seminary, 1962. MAAS, Paul. Textual Criticism. Translated by Barbara Flower. Oxford: Clarendon, 1958. Translation of “Textkritik.” in Einleitung in die Altertumswissenschaft, I, VII. Edited by Alfred Gercke & Eduard Norden. 3rd ed. Leipzig: Teubner, 1957. MILIK, Józef T. in Maurice Baillet et al. Les‘petitesgrottes’deQumrân.DJD III. Oxford: Clarendon, 1962. —. in Roland de Vaux and Józef T. Milik. Qumrângrotte4.II:I.Archéologie, II.Tefillin,MezuzotetTargums(4Q128–4Q157).DJD VI. Oxford: Clarendon, 1977, 33–79. NUTT, John W. FragmentsofaSamaritanTargum:EditedfromaBodleianManuscript, with an Introduction, Containing a Sketch of Samaritan History, DogmaandLiterature.London: Trübner, 1874. NYBERG, Henrik S. “Das textkritische Problem des Alten Testaments am Hoseabuche demonstriert.” ZAW 52 (1934): 241–54. OWEN, Elizabeth. “4QDeutn: A Pre-Samaritan Text?” DSD 4 (1997): 162–78. ROFÉ, Alexander. “Historico-Literary Aspects of the Qumran Biblical Scrolls.” Pages 30–39 in The Dead Sea Scrolls: Fifty Years after Their Discovery: ProceedingsoftheJerusalemCongress,July20–25,1997. Edited by Lawrence H. Schiffman, Emanuel Tov, and James C. VanderKam. Jerusalem: Israel Exploration Society and the Shrine of the Book, Israel Museum, 2000. SEGAL, Michael. “The Text of the Hebrew Bible in Light of the Dead Sea Scrolls.” MateriagiudaicaXII (2007): 5–20. TEETER, David A. ScribalLaws:ExegeticalVariationintheTextualTransmissionofBiblicalLawintheLateSecondTemplePeriod.FAT 92. Tübingen: Mohr Siebeck, 2014. TOV, Emanuel. “Approaches towards Scripture Embraced by the Ancient Greek Translators.” Pages 213–28 in Der Mensch vor Gott: Forschungen zum MenschenbildinBibel,antikemJudentumundKoran.FestschriftfürHerrmannLichtenbergerzum60.Geburtstag.Edited by Ulrike Mittmann-Richert et al. Neukirchen-Vluyn: Neukirchener, 2003. —. HebrewBible,GreekBible,andQumran:CollectedEssays.TSAJ 121. Tübingen: Mohr Siebeck, 2008. —. “Rewritten Bible Compositions and Biblical Manuscripts, with Special Attention to the Samaritan Pentateuch.” DSD5 (1998): 334–54. Revised version: Pages 57–70 in HebrewBible,GreekBible,andQumran. —. “Textual Harmonizations in the Ancient Texts of Deuteronomy.” Pages 271– 82 in HebrewBible,GreekBible,andQumran. —. TextualCriticismoftheHebrewBible(TCHB). Minneapolis: Augsburg Fortress, 2012. —. “The Samaritan Pentateuch and the Dead Sea Scrolls: The Proximity of the Pre-Samaritan Qumran Scrolls to the SP.” Pages 59–88 in KeterShemTov: EssaysontheDeadSeaScrollsinMemoryofAlanCrown. Edited by Shani Tzoref and Ian Young. PHSC 20. Piscataway, NJ: Gorgias Press, 2013. Revised version: Pages 387–410 in Emanuel Tov, Textual Criticism of the Hebrew Bible,Qumran,Septuagint:CollectedWritings,Volume3.VTSup 167. Leiden: Brill, 2015. —. “The Source of Source Criticism: The Relevance of Non-Masoretic Textual Witnesses.” Pages 283–301 in Text–Textgeschichte–Textwirkung,

40

EMANUEL TOV

Festschriftzum65.GeburtstagvonSiegfriedKreuzer. Edited by Thomas Wagner et al. AOAT 419. Münster: Ugarit-Verlag, 2015. —. “Textual Harmonization in the Stories of the Patriarchs.” Pages 166–88 in CollectedWritings,Volume3. —. “The Harmonizing Character of the Septuagint of Genesis 1–11.” Pages 470– 89 in CollectedWritings,Volume3. —. “The Genealogical Lists in Genesis 5 and 11 in Three Different Versions.” Pages 37–52 in FromAuthortoCopyist:EssaysontheComposition,Redaction,andTransmissionoftheHebrewBibleinHonorofZipiTalshir.Edited by Cana Werman. Winona Lake, IN: Eisenbrauns, 2015. Revised version: Pages 221–38 in CollectedWritings,Volume3. —. “The Shared Tradition of the Septuagint and the Samaritan Pentateuch.” Pages 277–93 in DieSeptuaginta:OrteundIntentionen. Edited by S. Kreuzer et al. WUNT 361. Tübingen: Mohr Siebeck, 2016. —. “The Development of the Text of the Torah in Two Major Text Blocks.” Text 26 (2016): 1–27. —. “Orthographic Practices of the Biblical Texts.” Forthcoming. —. “The Enigma of the Masoretic Text.” Pages 45–70 in TheologieundTextgeschichte,SeptuagintaundMasoretischerTextalsÄußerungentheologischer Reflexion. Edited by Frank Ueberschaer, Thomas Wagner, and Jonathan Miles Robker. WUNT 407. Tübingen: Mohr Siebeck, 2018. —. “The Textual Base of the Biblical Quotations in Second Temple Compositions.” Pages 280–302 in Hā-’îsh Mōshe: Studies in Scriptural Interpretation in theDeadSeaScrollsandRelatedLiteratureinHonorofMosheJ.Bernstein. Edited by Binyamin Y. Goldstein, Michael Segal, and George J. Brooke. STDJ 122. Leiden: Brill, 2017. —. “The Tefillin from the Judean Desert and the Textual Criticism of the Hebrew Bible.” Pages 277–92 in IsThereaTextinThisCave?StudiesintheTextualityoftheDeadSeaScrollsinHonourofGeorgeJ.Brooke. Edited by Ariel Feldman, Maria Cioată, and Charlotte Hempel. STDJ 119. Leiden: Brill, 2017. —. “Textual Harmonizations in Leviticus.” Forthcoming. —. “The Septuagint of Numbers as a Harmonizing Text.” Pages 181–201 in Die Septuaginta–Geschichte,Wirkung,Relevanz,6.InternationaleFachtagung veranstaltetvonSeptuagintaDeutsch(LXX.D),Wuppertal21.–24.Juli2016. Edited by Martin Meiser et al. WUNT 405. Tübingen: Mohr-Siebeck, 2018. —. “Textual Harmonization in Exodus 1–24.” TC:AJournalofBiblicalTextual Criticism 22 (2017). http://rosetta.reltech.org/TC/v22/TC-2017-Tov.pdf. TOV, Emanuel, and Sidney WHITE. “364–367. 4QReworked Pentateuchb–e and 4QTemple.” Pages 192–96 in QumranCave 4.VIII, ParabiblicalTexts,Part 1. Harold Attridge et al. in consultation with James VanderKam. DJD XIII. Oxford: Clarendon, 1994. WEST, Stephanie. ThePtolemaicPapyriofHomer. Papyrologica Coloniensia III. Cologne: Westdeutscher Verlag, 1967. ZAHN, Molly M. RethinkingRewrittenScripture:CompositionandExegesisin the4QReworkedPentateuchManuscripts. STDJ 95. Leiden: Brill, 2011. —. “The Samaritan Pentateuch and the Scribal Culture of Second Temple Judaism.” JSJ 46 (2015): 285–313.

EXCLUSION AND EXPANSION: HARMONISATIONS IN THE SAMARITAN PENTATEUCH, PRE-SAMARITAN PENTATEUCHAL MANUSCRIPTS AND NON-PENTATEUCHAL MANUSCRIPTS Michaël N. VAN DER MEER*

1. INTRODUCTION Since the publication of 4QpaleoExodm some thirty years ago it has become customary to distinguish between pre-Samaritan Pentateuchal manuscripts and a later supplementary Samaritan layer as found in the Samaritan Pentateuch.1 It has become clear that many of the large expansions in the Samaritan Pentateuch vis-à-vis the Masoretic Text of the Pentateuch are not the result of specific Samaritan scribal interventions, but part of a wider phenomenon of supplementing the authoritative text with parallel passages drawn from other parts of the same corpus. Thus, 4QpaleoExodm and the Samaritan Pentateuch share the major expansions in the plague narratives (Exod 7–11) and the second part of the book (Exod 18–39), but 4QpaleoExodm lacks the specific Samaritan version of the Ten Commandments in which a combination of Deut 11:29-30 and 27:2b-7 supplements Exod 20:17 and its parallel in Deut 5:21 in order to present the commandment to sacrifice at mount Gerizim (not Ebal as in MT) as the tenth commandment.2 The Exodus scroll from Qumran also lacks * It is my pleasant duty to thank Michael Langlois for his kind invitation to this conference

1

2

and to the participants of the conference for their stimulating reactions to an earlier version of this paper. As always I owe a special debt to my former Doktervater, Arie van der Kooij, for his useful and critical comments. Judith E. Sanderson, AnExodusScrollfromQumran:4QpaleoExodmandtheSamaritan Tradition, HSM 30 (Atlanta: Scholars, 1986); idem, “4QpaleoExodm,” in Patrick W. Skehan, Judith E. Sanderson, and Eugene Ulrich, eds., QumranCave4.IV:Palaeo-Hebrewand GreekBiblicalManuscripts, DJD IX (Oxford: Clarendon, 1992), 53-103. Ferdinand Dexinger, “Das Garizimgebot im Dekalog der Samaritaner“ in Georg Braulik (ed.), StudienzumPentateuch:WalterKornfeldzum60.Geburtstage (Wien: Herder, 1977), 111-33; see also Magnar Kartveit, TheOriginoftheSamaritans, VTSup 128 (Leiden: Brill, 2009), 290-95.

42

MICHAËL N. VAN DER MEER

the small but highly significant variant of the so-called missing yod in the statement regarding Yhwh’s election of the place for His worship. According to the Masoretic Text that decision still lays in the future of the narrative world of the Pentateuch aiming at the Temple in Jerusalem (‫)יבחר‬, whereas according to the Samaritan Pentateuch Yhwh had already chosen (‫ )בחר‬this site when Abraham built an altar at Shechem (Gen 12:6-7; 33:18-20). Hence, we can distinguish between three stages in the textual development of the Hebrew book of Exodus, one attested by the Masoretic text, a second with major expansions attested by 4QpaleoExodm and thirdly an even later, less extensive specifically Samaritan adaptation responsible for the expansion of the Ten Commandments (Exod 20:17) and the elision of the yod in statements regarding Yhwh’s election of the place for his sanctuary (22 times in Deut 12–31).3 Further study and publication of the biblical Dead Sea scrolls made clear that other Pentateuchal manuscripts from Qumran reflect the same layer of pre-Samaritan major expansions. Manuscript 4QNumb reflects the same interweaving of fragments from Moses’ retrospective speech found in Deut 1–3 as we find in the Samaritan Pentateuch. This scroll also combines elements from the parallel accounts of the narratives about Zelophehad’s daughters (Num 27:1-11; 36:1-13). Unlike the Samaritan Pentateuch and 4QpaleoExodm this Numbers manuscript is not written in the paleo-Hebrew script. This also applies to 4QExod-Levf which contains an expansion of Exod 39:20-23 based on the parallel passage in Exod 28:30, dealing with the production of the Urim and Thummim. It may well be that the manuscripts that are labelled ReworkedPentateuch (4Q158, 4Q364-367) in fact reflect the same kind of expansions found in the Samaritan Pentateuch and its pre-Samaritan precursors found among the Dead Sea Scrolls. Finally, manuscript 4QLevd seems to contain the same expansion after Lev 17:4 that is found in the Samaritan Pentateuch, but not in other witnesses like the paleo-Hebrew Leviticus scroll from Qumran cave 11.

3

Emanuel Tov, Textual Criticism of the Hebrew Bible, Third Edition, Revised and Expanded (Minneapolis: Fortress, 2012), 74-93, idem, “Foreword,” in Benyamin Tsedaka, Sharon Sullivan, eds., TheIsraeliteSamaritanVersionoftheTorah.FirstEnglishTranslationComparedwiththeMasoreticVersion(Grand Rapids: Eerdmans, 2013), vii-xii. Eugene Ulrich, “Rising Recognition of the Samaritan Pentateuch,” in The Dead Sea ScrollsandtheDevelopmentalCompositionoftheBible, ed. Eugene Ulrich (VTSup 169; Leiden: Brill, 2015), 215-27.

43

EXCLUSION AND EXPANSION

Table 1: List of duplications in pre-Samaritan Qumran Pentateuchal scrolls4 MS (date) 4QpaleoExodm (100-25 BCE)

duplication

source of addition in MT

alignment

Exod 7:18+

Exod 7:16-18

SamP cf. SyhL (ex Sam)] 4QGen-Exoda

Exod 7:29+

Exod 7:7:26-29

SamP cf. SyhL (ex Sam)] 4QExodc

(Exod 8:1+)

Exod 8:1

SamP cf. SyhL (ex Sam)

Exod 8:19+

Exod 8:16b-19

SamP cf. SyhL (ex Sam)

Exod 9:5+

Exod 9:1-5

SamP cf. SyhL (ex Sam)

Exod 9:19+

Exod 9:13-19

SamP cf. SyhL (ex Sam)

Exod 10:2+

Exod 10:3-6

SamP cf. SyhL (ex Sam)

(Exod 11:3+)

Exod 4:22f

SamP

Exod 18:25SamP

Deut 1:9b-18

SamP cf. SyhL (ex Sam)

Exod 20:19+

Deut 5:24-27

SamP cf. SyhL (ex Sam)

Exod 20:21+

Deut 5:28f.; SamP cf. SyhL (ex Sam) 18:18-22; 5:30-31

Exod 27:19+

Exod 39:1

SamP LXX53.58.56mg.129. 246.527.664.707

4QExod-Levf (c. 250 BCE)

Exod 32:10+

Deut 9:20

Exod 39:21+

Exod 28:20

SamP LXX58.767.318

4QLevd Lev 17:4+ ? (30 BCE – 20 CE)

Lev 17:3

SamP] 11QpaleoLeva

4QNumb (c. 30-1 BCE)

(Num 12:16+)

Deut 1:20-22

SamP Samareitikon

Num 20:13+

Deut 3:24-25, 26b-28; 2:2-6

SamP

Num 21:11+ (12a) Deut 2:9 4

SamP

Based on data provided by the editors in the DJD series and further Emanuel Tov, “Rewritten Bible Compositions and Biblical Manuscripts, with Special Attention Paid to the Samaritan Pentateuch”, DSD 5 (1998): 334-54 = HebrewBible,GreekBible,and Qumran.CollectedEssays, TSAJ 121 (Tübingen: Mohr-Siebeck, 2008), 57-70; Kartveit, TheOriginoftheSamaritans, 310-312. The major expansion is indicated by the + sign. Reconstructed variants are placed between brackets. In the column the alignment with the Samaritan Pentateuch and overlapping biblical scrolls from Qumran are indicated.

44

MICHAËL N. VAN DER MEER

MS (date)

4QRPa (75-25 BCE)

4QRPb (75-25 BCE)

duplication

source of addition in MT

alignment

Num 21:12+ (13a) Deut 2:17-19

SamP

Num 21:20+

Deut 2:24-25

SamP

(Num 21:22+)

Deut 2:28-29a

SamP

(Num 21:23+)

Deut 2:31

SamP

Num 27:23+

Deut 3:21b-22

SamP

(Num 31:20+)

Num 31:21-24

SamP

Num 36:5+

Num 27:2-11

–

?Gen 32:32+

Exod 4:27-28

–

?Exod 3:12+

Exod 24:4-6

–

?Exod 20:22+

Deut 5:29; 18:18-20, 22

cf. SamP

?Exod 20:12, 16, 17+

Deut 5:30-31

+Gen 28:6 Gen 30:36+

cf. Jub 27:14, 17 Gen 31:11-13

SamP

?Exod 19:17?

Exod 24:12-14

Exod 24:18+ ?

Exod 25:1-2

cf. Jub 1:4

Deut 2:8-9

SamP LXX767

Num 14:20+ Num 20:17-18+? 4QRPc (75-25 BCE)

Exod 15:21+ Exod 15:1-20 (song of Miriam?) (song of Moses)

–

Lev 24:2+

–

Num 27:11+

Num 36:1-2

cf. 4QNumb

4QRPd (75-25 BCE)

Num 30:1+ ?

Deut 16:13-14

cf. 11QTa col. xxix

4RPe (75-25 BCE)

+Num 20:13; 27:30-34

–

What combines these manuscripts is the phenomenon of deliberate duplication of relatively large sections from elsewhere within the same corpus of authoritative texts, viz. the Torah. This phenomenon presupposes not only the establishment of a corpus of such authoritative writings

EXCLUSION AND EXPANSION

45

but also a number of scholarly scribes so thoroughly familiar with its contents and discrepancies that they were able and willing to counterbalance perceived omissions and inconsistencies by duplicating material from rather remote parts of the same corpus. It is also noteworthy that this phenomenon occurs already relatively early within the textual transmission of the Pentateuch. In fact, it can be found already in the oldest extant manuscript of the Pentateuch, 4QExod-Levf, dating from the third century BCE, although the date of many other witnesses is younger, viz. the first century BCE (4QpaleoExodm, 4QNumb, 4QRPa-e). Manuscript 4QpaleoExodm shares with the Samaritan Pentateuch the feature of paleoHebrew script, but other paleo-Hebrew manuscripts, 11QpaleoLeva, 4QpaleoGen-Exodl, 4QpaleoGenm, 4QpaleoDeutr, 4QpaleoDeuts and 4QpaleoJobc, do not contain the duplications as found in the former category. This phenomenon of expansion by means of duplication can be distinguished from other scribal phenomena in the so-called biblical and parabiblical manuscripts, even though the distinction may be somewhat artificial. These phenomena include the rearrangement or transposition of existing material without duplication or the addition of completely new material, as found for instance in the Great Psalms scroll from cave 11, 11QPsalmsa, the rewriting of extant material as found for instance in the Temple Scroll, the Genesis Apocryphon or the book of Jubilees, substantial curtailment of existing material as found for instance in 4QCanta and 4QCantb, or related phenomena of excerpted and abbreviated texts such as 4QTest, 4QExodd, 4QDeutj, 4QDeutk1, 4QDeutn, 4QDeutq.5 For the sake of the argument I also make a distinction between this type of deliberate extensive duplication and unintentional forms of harmonization and textual alterations, even though this distinction (and the other mentioned above) may be somewhat artificial.6 Since we lack substantial parts of the Qumran manuscripts it is often difficult to tell whether a major expansion within a scroll actually reflects duplication, rearrangement, considerable abbreviation or accidental omission of biblical texts. Nevertheless, we do have enough material to prove that what we find in the Samaritan Pentateuch was part of a wider scribal 5

6

Emanuel Tov, “Excerpted and Abbreviated Biblical Texts from Qumran,” RevQ 16 (1995): 581-600, repr. in HebrewBible,GreekBible,andQumran, ed. Emanuel Tov, TSAJ 121 (Tübingen: Mohr-Siebeck, 2008), 27-41. See also Emanuel Tov, “The Nature and Background of Harmonizations in Biblical Manuscripts,” JSOT 31 (1985): 3-29.

46

MICHAËL N. VAN DER MEER

phenomenon and not restricted to the scribal activity of Samaritan scholars after the parting of the ways between Judaism and Samaritanism following the conquest of the Temple on Mount Gerizim by John Hyrcanus in 128 BCE7 and its final destruction in 110 BCE.8 But how widespread was this scribal phenomenon? Was it restricted in place, for instance to Palestine, as Cross’ theory of local texts would have it?9 Was it restricted to a particular scribal circle, for instance that of the Sadduccees, as surmised by Tov for the paleo-Hebrew manuscripts?10 Was it restricted to a specific period in time, for instance the Second Temple Period which ended with the destruction of the Temple in Jerusalem in 70 CE, as advocated by Eugene Ulrich?11 and finally: Was it restricted to a specific part of the Bible, e.g. the Torah, which was the sole canon for both the Samaritans and the Sadducees, or can we detect examples of the same phenomenon beyond the Pentateuchal scrolls from Qumran and Samaria dating to the Hellenistic and early Roman periods? In short: how exclusive were these expansionistic biblical texts?

7

8

9

10

11

See James D. Purvis, TheSamaritanPentateuchandtheOriginoftheSamaritanSect, HSM 2 (Cambridge, MA: Harvard University Press, 1968); Robert T. Anderson, Terry Giles, TheSamaritanPentateuch.AnIntroductiontoItsOrigin,History,andSignificanceforBiblicalStudies, SBLSBS 71 (Atlanta: Scholars, 2012). The excavations at mount Gerizim have revolutionized Samaritan studies. It is now clear that the sacred precinct at that mountain were built much earlier than previously thought (i.e. already in the Persian period, mid-fifth century BCE rather than the times of Alexander the Great as claimed by Flavius Josephus) and was destroyed only around 111-110 BCE (instead of 128 BCE), see Yitzhak Magen, Haggai Misgav, Levana Tsfania, eds., MountGerizimExcavations. 1: TheAramaic,HebrewandSamaritanInscriptions. Judea and Samaria Publications 2 (Jerusalem: Israel Antiquities Authority, 2004) and 2. ATempleCity, Judea and Samaria Publications 4 (Jerusalem: Israel Antiquitates Authority, 2008). See also Esther Eshel and Hanan Eshel, “Dating the Samaritan Pentateuch’s Compilation in Light of the Biblical Scrolls,” in Emanuel. Studies in HebrewBible,Septuagint,andDeadSeaScrollsinHonorofEmanuelTov, eds. Shalom M. Paul, Robert A. Kraft, Lawrence H. Schiffmanm and Weston W. Fields, VTSup 94 (Leiden: Brill, 2003), 215-40. The relationship between Jews and Samaritans even after the schism in the Hasmonean period is described in detail by Gary N. Knoppers, Jews andSamaritans:TheOriginsandHistoryofTheirEarlyRelationsHardcover(Oxford: Oxford University Press, 2013). See e.g. Frank Moore Cross, Shemaryahu Talmon, eds., QumranandtheHistoryofthe BiblicalText (Cambridge, MA: Harvard University Press, 1975). Emanuel Tov, “The Socio-Religious Background of the Paleo-Hebrew Biblical Texts Found at Qumran,” in Geschichte-Tradition-Reflexion. Festschrift für Martin Hengel zum70.Geburtstag.BandI:Judentum, eds. Hubert Cancik, Hermann Lichtenberger, Peter Schäfer (Tübingen: Mohr-Siebeck, 1996), 353-374. Eugene Ulrich, The Dead Sea Scrolls and the Origins of the Bible (Grand Rapids: Eerdmans; Leiden: Brill, 1999); idem, “The Developmental Composition of the Biblical Text,” in Ulrich, DevelopmentalComposition, 1-14.

EXCLUSION AND EXPANSION

2. DUPLICATION

47

IN NON-PENTATEUCHAL SOURCES

It is my contention that we can trace this phenomenon of duplication of biblical material in biblical scrolls beyond the boundaries of the Pentateuch. In this section I would like to discuss some examples from the Former Prophets both in Hebrew and Greek. 2.1 4QJoshuaa Manuscript 4QJoshuaa, written during the Hasmonean period (c. 15050 BCE), contains a significant variant at the beginning of the extant text. Here, parts of the passage now found in the Masoretic Text in Josh 8:3035, but in the Greek text after Josh 9:2, appear before Josh 5:2. Since the passage about Joshua’s altar at mount Ebal, MT-Josh 8:30-35, clearly interrupts the sequence of the stories of the Fall of Ai (Josh 8:1-29) and the ruse of the Gibeonites (Josh 9:1-27), scholars had long concluded that this passage reflects a later Deuteronomistic addition to the book. Since the passage is found at two different places within the Masoretic and Septuagint traditions (either before or after Josh 9:1-2) and yet at an even different position after the conquest narratives in the rewritten versions of Flavius Josephus (Ant. 5.68-70) and Pseudo-Philo (LAB 21:7-10), it has long been assumed that 4QJosha also reflects a transposition of this passage, dubbed by Graeme Auld as “a latecomer looking for a suitable home”.12 As a result, the scholarly discussion concerning this remarkable major variant has mainly revolved around the question whether the position of the passage in MT-Josh 8:30-35 is secondary, reflecting antiSamaritan polemics, as Eugene Ulrich argues,13 or for nomistic reasons, 12

13

A. Graeme Auld, “Reading Joshua after Kings,” in WordsRemembered,TextsRenewed. Essays in Honour of John F. Sawyer, eds. J. Davies, G. Harvey, W.G.E. Watson, JSOTSup 195 (Sheffield: Sheffield Academic Press, 1995), 167-181, repr. in Joshua Retold.SynopticPerspectives, ed. A. Graeme Auld, OTS (Edinburgh: T & T Clark, 1998), 102-12; see already Carl Steuernagel, ÜbersetzungundErklärungderBücherDeuteronomiumundJosuaundallgemeineEinleitungindenHexateuch, HKAT 1.3 (Göttingen: Vandenhoeck & Ruprecht, 1900), 186. For a comparable view see Emanuel Tov, “Some Sequence Differences between the MT and LXX and their Ramification for the Literary Criticism of the Bible,” JNSL 13 (1987): 151-60, repr. in TheGreekandHebrewBible. CollectedEssaysontheSeptuagint, ed. Emanuel Tov, VTSup 72 (Leiden: Brill, 1999), 411-18. For a history of research concerning this passage, see my Formulation and Reformulation.TheRedactionoftheBookofJoshuaintheLightoftheOldestTextual Witnesses, VTSup 102 (Leiden: Brill, 2004), 479-96. Eugene Ulrich, “4QJoshuaa and Joshua’s First Altar in the Promised Land,” in New QumranTextsandStudies.ProceedingsoftheFirstMeetingoftheInternationalOrganizationforQumranStudies,Paris1992, eds. George J. Brooke and Florentino García

48

MICHAËL N. VAN DER MEER

as argued differently by either Alexander Rofé, Heinz-Josef Fabry, Kristin De Troyer and Ed Noort.14 Yet, the extant text of 4QJoshuaa does not contain the whole text of MT-Josh 8:30-35, but only a few words of verse 34, the whole of verse 35 and an additional verse (5:x ‫אחר אשר נתקו̇ ] [ל] [את פפר התורה אחר‬ [ ]‫הארון‬ ̇ ‫)כן ] [ל נושאי‬ ̇ that is not found in any other textual tradition, but which seems to embed the preceding lines within the narrative context of Joshua 4 and 5, especially Josh 4:18a (‫ויהי כבעלות )קכעלות( הכהנים‬ ‫)נשאי ארון ברית יהוה מתוך הירדן נתקו כפות רגלי הכהנים אל החרבה‬. Moreover, we cannot be sure whether in 4QJoshuaa this passage was missing between Josh 8:29 and 9:3. As a matter of fact, we do have fragments of Josh 8, 9 and 10 from this 4QJosha scroll including a top margin (fragment 9 ii) and bottom margin (fragment 21) which allows for a tentative reconstruction of the whole of Joshua 8:3-10:11. As I attempted to demonstrate in my 2001 Leiden dissertation such a reconstruction is much more plausible with than without the passage found in MT-Joshua 8:30-35.15 What we then find in 4QJosha is to my mind not evidence of transposition of this passage either to its current position in MT or to its allegedly secondary position in 4QJosha. Rather only those verses of Josh 8:30-35 that deal with the writing and reading of the torah, i.e. verses 32, 34 and 35, have been duplicated by the scribe of 4QJosha or his predecessor before Josh 5:2. The reason for this duplication should be sought in the related passage in Deut 27:2-8. Here we find instructions to build an altar on mount Ebal and write and read the torah from the plastered stones erected there, corresponding to MT-Joshua 8:30-35. This parallel passage, however, is no less problematic than its counterpart in Joshua, since it seems to combine two conflicting instructions, one dealing with the

14

15

Martínez, STDJ 15 (Leiden: Brill, 1994), 89-104, repr. in Ulrich, DevelopmentalComposition, 47-65 (revised and expanded edition). Alexander Rofé, “The Editing of the Book of Joshua in the Light of 4QJosha,” in New QumranTextsandStudies, 73-80; Heinz-Josef Fabry, “Der Altarbau der Samaritaner-ein Produkt der Text- und Literaturgeschichte,” in DieTextfundevomTotenMeerundder TextderHebräischenBibel, ed. Ulrich Dahmen, Armin Lange, Hermann Lichtenberger (Neukirchen: Neukirchener, 2000), 35-52; Ed Noort, Een plek om te zijn. Over de theologievanhetlandvanJozua8:30-35, inaugural lecture Rijksuniversiteit Groningen (Kampen: Kok, 1993); idem, “The Traditions of Ebal and Gerizim. Theological Positions in the Book of Joshua,” in DeuteronomyandDeuteronomisticLiterature.Festschrift C.H.W.Brekelmans, BETL 133 (Leuven: Peeters, 1997), 161-80; idem, “4QJoshuaa and the History of Tradition in the Book of Joshua,” in JNSL 24 (1998), 127-44; Kristin De Troyer, “Building the Altar and Reading the Law. The Journeys of Joshua 8:30-35,” in ReadingthePresentintheQumranLibrary:ThePerceptionoftheContemporaryby Means of Scriptural Interpretations, ed. Kristin De Troyer, Armin Lange, SBLSS 30 (Atlanta: Scholars, 2005), 141-62. FormationandReformulation, 511-19.

EXCLUSION AND EXPANSION

49

erection of an altar on mount Ebal (Deut 27:5-7), the other with erecting large stones and writing the torah on it as soon as Israel has crossed the river Jordan (Deut 27:2-4, 8): Table 2: Redaction-critical comparison between Deut 27:2-4, 8 and 27:5-7 crossing of the Jordan

erection of stones

plastering of stones purpose of stones

Deut 27:2-4, 8 ‫והיה ביום אשׁר תעברו‬ ‫את הירדן אל הארץ אשׁר‬ ‫יהוה אלהיך נתן לך‬ ‫והקמת לך אבנים גדולות‬ ‫ושׂדת אתם בשׂיד‬ ‫וכתבת עליהן את כל‬ ‫דברי התורה הזאת‬ ‫בעברך למען אשׁר תבא‬ ‫אל הארץ אשׁר יהוה‬ ‫אלהיך נתן לך ארץ זבת‬ ‫חלב ודבשׁ כאשׁר שדבר‬ ‫יהוה אלהי אבתיך לך‬ ‫וכתבת על האבנים אי כל‬ ‫דברי התורה הזאת באר‬ ‫היטב‬

Deut 27:5-7 ‫והיה בעברכם את הירדן‬ ‫תקימו את האבנים האלה‬ ‫אשׁר אנכי מצוה אתכם‬  ‫היום בהר עיבל‬ ‫ושׂדת אתם בשׂיד‬ ‫ובנית שׁם מזבח ליהוה‬ ‫אלהיך מזבח אבנים לא‬ ‫תניף עליהם ברזל אבנים‬ ‫שׁלמות תבנה את מזבח‬ ‫יהוה אלהיך והעלית עליו‬ ‫עולת ליהוה אלהיך וזבחת‬ ‫שׁלמים ואכלת שׁם‬ ‫ושׂמחת לפני יהוה אלהיך‬

As I see it, the scribe responsible for the second edition of Joshua 4-5 as reflected by 4QJoshuaa read the commandments of Deut 27:2-8 as two different instructions: one applying to the day of the crossing of the Jordan (Deut 27:2) and the instruction to write the torah on the stones, apparently taken from the riverbed of the Jordan (Josh 4); the other dealing with the erection of an altar on mount Ebal (Deut 27:5-7) on any given moment after the crossing of the river Jordan (Deut 27:4). Since Josh 8:30-35 reports the fulfillment of this second set of instructions, the scribe copied Josh 8:32, 34-35 in order to have an execution of the other commandment (Deut 27:2-4, 8) as well. Balancing divine command and its execution is exactly what we find in the Samaritan Pentateuch and its pre-Samaritan precursors. For that reason I contend that the phenomenon which we find in these Pentateuchal witnesses is attested beyond the Pentateuch as well, in this case in 4QJoshuaa. The problematic position of Josh 8:30-35 should not be dismissed as a rather unfortunate secondary transposition, nor can it plausibly be explained as an anti-Samaritan textual move. Rather it should be seen within the

50

MICHAËL N. VAN DER MEER

wider nomistic redaction of the book, encompassing Josh 1:7-8, 22:5 and 23, which transforms the older Deuteronomistic themes of loyalty to Yhwh under a sole leader into a polemic against affiliation with foreign people. Seen in this way, Josh 8:30-35 marks the end of the period of absolute compliance to God’s torah (hence the somewhat enigmatic phrase in Josh 8:33 ‫ )לברך את העם ישראל בראשנה‬and the beginning of Israel’s Lapse (inaugurated by the treaty with the Gibeonites).16 The fact that the passage is found a few verses further in the narrative in the Old Greek tradition then requires an explanation of its own. To my mind, it makes good sense to see this transposition as a deliberate attempt of the Greek translator to enhance the probability of the narrative. Josh 9:1-2 narrates the gathering of the hostile forces. Within the Deuteronomistic stratum of the book this short episode serves to underline the successfulness of Joshua’s campaign over so many enemies. Within the Greek text it serves as an interlude, a time of truce and reinforcement of the troops, allowing Joshua to perform his religious duties.17 The development of Deut 27:5-7, 2-4, 8, Josh 8:30-35 in MT, 4QJosha and LXX can therefore be outlined as follows: Table 3: The Formation Deut 27:5-7, 2-4, 8, Josh 8:30-35 in MT, 4QJoshuaa and LXX Dtr layer Deut?

DtrN layer Deut?

DtrN layer Josh (MT)

4QJosha

Josh 1:1-5:1

LXX-Josh

Josh 1:1-5:1

Josh 5:1+ (duplication of Josh 8:32, 34-35 + additional verse) Josh 5:2-8:29 Deut 27:2-4 Deut 27:5-7

17

Josh 5:2-8:29

Josh 8:30-35 Josh 9:1-2

Deut 27:8

16

Josh 8:30-35

Josh 9:1-2 Josh 9:2+

FormationandReformulation, 498-511. Formation and Reformulation, 519; Michaël N. van der Meer, “Provenance, Profile and Purpose of the Greek Joshua”, in XIICongressoftheInternationalOrganization forSeptuagintandCognateStudies,Leiden2004, ed. Melvin K.H. Peters, SBLSCS 54 (Atlanta: Scholars, 2006), 55-80.

EXCLUSION AND EXPANSION

51

2.2 The Septuagint of Joshua Although the Old Greek or Septuagint version of Joshua does not provide direct corroboration for the literary and textual history of Josh 8:30-35, it does have a number of remarkable duplications of authoritative material: – In LXX-Josh 6:26 we find a note appended to Joshua’s curse over the rebuilder of Jericho describing the actual fulfillment of that curse in the time of king Ahab. The source of this addition is found in MT-1 Kgs (LXX-3 RegB) 16:34 (but absent from the so-called “Lucianic” or “Antiochene” text of 3 Regum). – Likewise, the note about the Judahites’ failure to conquer Gezer in MTJosh 16:10 is complemented with a short note found in MT-1 Kgs 9:16 (but in LXX-3 Reg at 5:14b). – Material from Judg 1:34-35 dealing with the settlement of the Danite tribe is also found in LXX-Josh 19:47+. In the Masoretic version of the book of Joshua the section about the division and distribution of the land of Israel west of the Jordan ends at Josh 19:49-51. – The following two chapters (Josh 20-21) contain priestly provisions about the cities of refuge and the cities for the Levites. In the Septuagint these chapters are embedded within the previous part by means of a duplication of Josh 19:49-51 after Josh 21:42. – This major expansion concluding the whole of Josh 13-21 also contains a short note not found elsewhere. It deals with the flint knives Joshua made when he was ordered to reinstall the rite of circumcision as narrated in Josh 5:2-9. – A similar additional note is found after LXX-Josh 24:31 (MT-Josh 24:30). – The final chapter of the book also contains duplications of material found also in Deut 26:5-6 (LXX-Josh 24:4b-5a) and Judges 2:6, 11-14; 3:12, 14 (LXX-Josh 24:33a-b). With respect to this last major expansion in the Septuagint of Joshua it is also possible to argue for the originality of the Old Greek version vis-à-vis the much longer version now found in Judg 1-3, which parallels parts of the material found at the end of Joshua 24. As is well-known, Alexander Rofé has argued that the Septuagint of Joshua has retained here elements of an Old Ephraimite historical work, that was later superseded by the Deuteronomistic history.18 18

Alexander Rofé, “‫סיומו שׁל יהושׁע לפי תרגום השׁבעים‬,” Shnaton 2 (1977): 217-27, transl. as “The End of the Book of Joshua According to the Septuagint,” Hen 4 (1982): 17-35; see also Birgit Lucassen, “Josua, Richter und CD,” RevQ 18 (1998): 373-96; Tov, TextualCriticism, 297-98;Van der Meer, FormationandReformulation, 60-62.

52

MICHAËL N. VAN DER MEER

Attractive as this attempt to delve into the prehistory of the ancient Israelite historical works may be,19 it is much more likely that the Septuagint contains a later duplication of parallel material, as argued already by Hartmut Rösel and Erhrard Blum.20 Table 4: Major Expansions in 4QJoshuaa and LXX-Joshua witness 4QJosha (150-50 LXX-Josh

(220-200

duplication BCE)

source of addition in MT

alignment

Josh 5:1+

Josh 8:32, 34-35 cf. 4:19

Josh 6:26+

1 Kgs (3 RegB) 16:34

cf. 4QapocrJoshc, 4QTest

Josh 16:10+

MT-1

cf. LXX-3 Reg 5:14

Josh 19:47+

Judg 1:34-35

Josh 21:42+

Josh 19:49-50

Josh 24:4+

Deut 26:5-6

BCE?)

Kgs 9:16

Josh 24:30+ Josh 24:33+

cf. Josh 5:2-9 cf. Josh 5:2-9; 21:42+

Judg 2:6, 11-14; 3:12, 14

These major expansions in the book of Joshua resemble the major expansions in the Pentateuch as found in the Samaritan Pentateuch and the pre-Samaritan biblical scrolls from Qumran in many ways: they too reflect the tendency to coordinate command and execution. The curse over anyone who rebuilds Jericho, which originally anticipated the ominous reign of king Ahab, finds its fulfillment almost immediately in the expanded version of Josh 6:26+. The other expansions are also complementary and concluding in character: the flint knives made in Josh 5:3 are taken and 19

20

See also C.F. Burney, The Book of Judges (London: Rivingtons, 1918), xli-l; Eduard Nielsen, Shechem:ATraditio-HistoricalInvestigation (Copenhagen: Gad, 1959), 134-37; Alexander Rofé, “Ephraimite versus Deuteronomistic History,” in StoriaeTradizionedi Israele.ScrittiinOnorediJ.AlbertoSoggin, ed. D. Garrone, F. Israel (Brescia: Paideia, 1991), 221-35. Hartmut N. Rösel, “Die Überlieferungen vom Josua- ins Richterbuch,” VT 30 (1980): 342-50; Erhard Blum, “Der kompositionelle Knoten am Übergang von Josua zu Richter: Ein Entflechtungsvorschlag,” in DeuteronomyandDeuteronomicLiterature.Festschrift ChrisH.W.Brekelmans, ed. Marc Vervenne and Johan Lust, BETL 133 (Leuven: Peeters, 1997), 181-212.

EXCLUSION AND EXPANSION

53

buried along with Joshua (21:42+; 24:30+), Gezer is captured after all (16:10+), the tribe of Dan eventually gains some foothold in Israel (Josh 19:47+), the sections about division of land and cities find their conclusion after the priestly appendices in Josh 20-21 (21:42+) and the period of Israel’s loyalty to Yhwh, which for the nomistic editor of Josh 8:30-35 may have ended already with the treaty with the Gibeonites, finally finds its conclusion with the death of Joshua and Eleazar (Josh 24:33+). To my mind, therefore, there is every reason to believe that the scribal activity responsible for the harmonizing duplications found in the Pentateuch the Samaritans chose and found in 4QpaleoExodm, 4QExod-Levf, 4QLevd, 4QNumb and 4QReworkedPentateuch was not restricted to the Pentateuch, but also included the Former Prophets, at least the book of Joshua. The oldest extant textual witnesses, 4QJosha, LXX-Joshua, each in their own way reflect examples of this type of duplication. The differences also show that there was not a single unified expansion or re-edition of the book, but rather several attempts to expand the book by means of duplication of material found elsewhere in the torah and Former Prophets. The Hebraistic character of the major expansions found in the Septuagint suggests that the expansions reflected by the Old Greek version were made by a Hebrew rather than a Greek scribe, but the possibility of a Greek creative translator-scribe can not be dismissed easily, I would argue. Anyhow, if the Greek translation was made already during the last decades of the third century BCE, as I have argued earlier,21 the duplications found in the Septuagint must date from before the second century BCE (cf. the early date of 4QExod-Levf). 2.4 4QSamuela My proposal to see the major expansions in the oldest textual witnesses of Joshua in the light of the duplications found in the Samaritan Pentateuch and affiliated manuscripts from Qumran has not found much response yet. Emanuel Tov does interact with my views on 4QJoshuaa, but according to him the link between the plus in that manuscript before Josh 5:2 and 4:18 indicates that the whole of Josh 4:19-5:1 constitutes a later textual addition.22 Other scholars have pointed to the hypothetical 21 22

Van der Meer, “Provenance, Profile and Purpose”. Emanuel Tov, “Literary Development of the Book of Joshua as Reflected in the MT, the LXX, and 4QJosha,” in TheBookofJoshua, ed. Ed Noort, BETL 250 (Leuven: Peeters, 2012), 65-85, repr. in Textual Criticism of the Hebrew Bible, Qumran,

54

MICHAËL N. VAN DER MEER

character of my reconstruction of 4QJosha, something I readily admit.23 Nevertheless, the organizers of this congress kindly invited me to elaborate on my views on the Dead Sea Scrolls and the Samaritan Pentateuch. So, I am very happy to pursue my line of thinking and adduce more examples that reflect precisely the same kind of harmonizing duplication as found in the manuscripts and witnesses discussed thus far. Manuscript 4QSamuela, written around 50-25 BCE, also contains major expansions vis-à-vis the other extant witnesses of the book of Samuel. In quite a number of cases the pluses in that manuscript are probably more original than the shorter Hebrew text preserved in MT, which probably suffered from haplography in many instances. This probably applies to the pluses found in 1 Samuel 1 and at the end of 1 Samuel 10.24 The Song of Hannah, however, presents a different case. The end of the song as reflected in 4QSama and the Old Greek version contains a major expansion of a passage also found in Jer 9:22-23. This passage contains a warning of Yhwh to the wise not to boast in his wisdom, to the mighty not to boast in his might, to the wealthy not to boast in his wealth, but to boast in the understanding of Yhwh. This admonition ties in nicely with the warning in 1 Sam 2:9 that man does not prevail by strength (‫)כי לא בכח יגבר איש‬: Table 5: Synopsis of 1 Sam 2:9-10 in Vs

LXX-1

Regum 2

9 διδοὺς εὐχὴν τῷ εὐχομένῳ καὶ εὐλόγησεν ἔτη δικαίου ὅτι οὐκ ἐν ἰσχύι δυνατὸς ἀνήρ

23

24

LXX,

4QSamuela col. ii

4QSamuela and MT-1

MT

Samuel 2 Vs

‫מר ודרך ח]סידיו ישמור‬ ֹ ֔ ‫]ח ִס ָיד ֙יו[ יִ ְשׁ‬ ֲ (‫)ח ִסידוֹ‬ ֲ ‫ַרגְ ֵל֤י‬ [‫ורשעים בחשך ידמו‬ ‫יִדּמּוּ‬ ֑ ָ ‫ְוּר ָשׁ ִ ֖עים ַבּ ֣חֹ ֶשְׁך‬ ‫]נוד[ר‬ ̊ ‫נד]ר ̊[ל‬ ̇ ‫נתן‬

9

‫ויברך ש]נות צדיק‬ [‫כי לוא בכח יגבר איש‬

‫ר־אישׁ׃‬ ֽ ִ ‫י־ל ֹא ְב ֖כֹ ַח יִגְ ַבּ‬ ֥ ‫ִ ֽכּ‬

Septuagint.CollectedEssays,Volume3, ed. Emanuel Tov, VTSup 167 (Leiden: Brill, 2015), 132-53. See the reviews of my Formation and Reformulation in ZAW 117 (2005): 324-25 (Martin Rösel), JSOT 29 (2005): 58-59 (George J. Brooke), and RBL (2005) (Steven L. McKenzie). See Frank Moore Cross, Donald W. Parry, Richard J. Saley, Eugene Ulrich, eds., Qumran Cave4.XII:1-2Samuel, DJD XVII (Oxford: Clarendon, 2005) and the numerous publications mentioned in that volume.

55

EXCLUSION AND EXPANSION

Vs

LXX-1

Regum 2

10 κύριος ἀσθενῆ ποιήσει ἀντίδικον αὐτοῦ κύριος ἅγιος

μὴ καυχάσθω ὁ φρόνιμος ἐν τῇ φρονήσει αὐτοῦ καὶ μὴ καυχάσθω ὁ δυνατὸς ἐν τῇ δυνάμει αὐτοῦ καὶ μὴ καυχάσθω ὁ πλούσιος ἐν τῷ πλούτῳ αὐτοῦ ἀλλ᾽ ἢ ἐν τούτῳ καυχάσθω ὁ καυχώμενος συνίειν καὶ γινώσκειν τὸν κύριον καὶ ποιεῖν κρίμα καὶ δικαιοσύνην ἐν μέσῳ τῆς γῆς κύριος ἀνέβη εἰς οὐρανοὺς καὶ ἐβρόντησεν αὐτὸς κρινεῖ ἄκρα γῆς καὶ δίδωσιν ἰσχὺν τοῖς βασιλεῦσιν ἡμῶν καὶ ὑψώσει κέρας χριστοῦ αὐτοῦ

4QSamuela col. ii

MT-1

Samuel 2 Vs

‫]מ ִר ָ֗יביו[ יהוה י̊ ̇חת מר]י[בו‬ ְ (‫)מ ִריבוֹ‬ ְ ‫יֵחתּוּ‬ ֣ ַ ‫יְהוה‬ ָ֞ 10 [‫מי ̊ק]דוש כיהוה‬ [ ] ]◦‫◦תם בשלמ‬ ̊ [ ] ‫אל יתהלל חכם‬ [ ‫בחכמתו‬ ‫ו̇ ̇אל ̊י̊ת]ה[ל]ל הגבור‬ ‫בגבורתו‬ ‫ואל יתהלל עשיר‬ ‫בעשרו‬ ‫כי בזאת יתהלל‬ ‫המתהלל‬ [‫השכל וידע את יהוה‬ ‫מש]פט‬ ̊ ‫]ולעשו[ת‬ ̊ ‫וצדקה בתוך הארץ‬ [‫יהוה עלה בשמים‬ ‫וירעם‬

‫]ע ָל ֙יו[ ַבּ ָשּׁ ַ ֣מיִם ְיַר ֔ ֵעם‬ ָ (‫)עלוֹ‬ ָ

‫]יהוה ידין אספי ארץ‬ ‫ויתן עז למלכנו‬

‫י־א ֶרץ‬ ֑ ָ ‫יָדין ַא ְפ ֵס‬ ֣ ִ ‫יְהו֖ה‬ ָ ‫וְ ֶיִתּן־ ֣עֹז ְל ַמ ְל ֔כּוֹ‬

]‫וירם קרן[ משיחו‬

‫יָרם ֶ ֥ק ֶרן ְמ ִשׁ ֽיחוֹ׃‬ ֖ ֵ ְ‫ו‬

Although there is a thematic link between the verses found in 4QSamuela and LXX-1 Regum 2:9-10, it is also evident that these verses form a later addition to this early Hebrew poem.25 The plus seems to import a theme 25

Thus according to the editors of DJD XVII. See also Emanuel Tov, “Different Editions of the Song of Hannah and its Narrative Framework,” in Tehillah le-Mosheh. Biblical andJudaicStudiesinHonorofMosheGreenberg, eds. Michael Cogan, Barry L. Eichler, Jeffry H. Tigay (Winona Lake: Eisenbrauns, 1997), 149-70 repr. in Tov, TheGreekand

56

MICHAËL N. VAN DER MEER

from (post-exilic?) wisdom traditions into a song that exalts the strength of seemingly insignificant Israelite kings.26 Like the examples from the Pentateuch we find here a duplication of material from a different composition, in this case the prophetic writings attributed to Jeremiah, attached to thematically related material in a completely different literary setting, in this case the book of Samuel. Piet de Boer suggested that the two poetical compositions, 1 Sam 2:1-10 and Jer 9:22-23 formed part of the liturgy of the synagogue and hence became mixed up.27 Yet we do not know anything of synagogue liturgy in the pre-Christian period and from what we do know from the later period is that these two texts did not constitute a fixed liturgical unity. If the two texts had been conflated because of liturgical use, we might have expected more creative adaptations. Instead the one passage is carefully quoted and copied into the other literary composition. Ernst Kutsch proposed to reverse the order of literary dependence regard Jer 9:22-23 as a reworking of the material found in the plus in 1 Sam 2:10. He argued that the passage in 1 Samuel stands closer to the sapiential tradition from which this wisdom text is obviously derived.28 Although it is obvious that the inclusion of this sapiential admonition in the corpus of Jeremiah traditions transforms this wisdom text is now transformed into divine prophetic speech, it is not so obvious that the plus in 1 Sam 2:10

26

27

28

HebrewBible, 433-55. See further the commentaries to 1 Samuel, e.g. Samuel R. Driver, NotesontheHebrewTextoftheBooksofSamuel (Oxford: Clarendon, 1890), 20; Henry R. Smith, ACriticalandExegeticalCommentaryontheBooksofSamuel, ICC (Edinburgh: T & T Clark, 1899), 17; Hans Joachim Stoebe, DasersteBuchSamuelis, KAT VIII/1 (Gütersloh: Gütersloher Verlagshaus, 1983), 106-7;Ralph W. Klein, 1 Samuel, WBC 10 (Waco: Word, 1983), 13; Walter Dietrich, 1Samuel1-12, BKAT VIII/1 (Neukirchen: Neukirchener Verlag, 2011), 69, 96-97 with references to relevant other secondary literature. For a relatively late date of Jer 9:22-23, see e.g. Bernhard Duhm, DasBuchJeremia, KHCAT 11 (Tübingen: Mohr-Siebeck, 1901), 97; Paul Volz, Der Prophet Jeremia übersetztunderklärt, KAT 10 (Leipzig: Scholl, 1928),119-20, and William McKane, ACriticalandExegeticalCommentaryonJeremiah 1, ICC (Edinburgh: T & T Clark, 1986), 212-13. More recent scholars ascribe these verses to the prophet and date them to the early sixth century, but nevertheless recognize the sapiential character, see e.g. Wilhelm Rudolph, Jeremia, HAT 12 (Tübingen: Mohr-Siebeck, 1947), 59; Peter C. Craigie, Page H. Kelley, Joel F. Drinkard, Jeremiah1-25, WBC 26 (Waco: Word, 1991), 152-54; William L. Holladay, Jeremiah1.ACommentaryontheBookofthe ProphetJeremiahChapters1-25, Hermeneia (Minneapolis: Fortress, 1986), 316-18; Jack R. Lundbom, Jeremiah1-20 AncB 21A (New York; Doubleday, 1999),568-575. Pieter A.H. de Boer, ResearchintotheTextof1SamuelI-XVI.AContributiontothe StudyoftheBooksofSamuel(Amsterdam: Paris, 1938), 58. Ernst Kutsch, “Weisheitsspruch und Prophetenwort. Zur Traditionsgeschichte des Spruches Jer 9,22-23,” BZ 25 (1981): 161-79.

57

EXCLUSION AND EXPANSION

formed the indispensable interim stage. If Jer 9:22-23 was derived from 1 Sam 2:10+ one wonders why the Samuel plus was dropped already so early in the transmission history. Early interpretations of the Samuel text as found in Targum Jonathan and Liber Antiquitatum Biblicarum already attest to the shorter Hebrew version without the plus.29 Hence, I find it much more likely that we are dealing in 1 Samuel 2:10+ with the same type of scribal activity that underlies the duplications found in the Pentateuch and the book of Joshua. To my mind, we also find an example of scribal duplication at the end of the scroll, viz. in 2 Sam 24:16-20, where the story of David and the angel at the threshing floor of Araunah contains elements found in the rival version of 1 Chronicles 21. Particularly the statements about David seeing an angel gave rise to coordination by means of duplication. In the Samuel version the angel seems to be a human figure, whereas in Chronicles he is elevated to a heavenly figure, suspended between heaven and earth. In 4QSamuela both versions seem to be combined: Table 6: Synopsis of 2 Sam 24:17-20 in 4QSamuela and with 1 Chr 21:16 Vs 16

MT-1

4QSamuela (DJD)

Chronicles 21

‫ת־ע ָ֗יניו‬ ֵ ‫וַ יִּ ָ ֙שּׂא ָד ִ ֜ויד ֶא‬

‫וי̊ ̇ש ̊א ]דו[י̇ ̊ד] את[ ]עיניו‬

‫וַ ַ֞יּ ְרא‬

‫וירא‬

‫ת־מ ְל ַ ֤אְך יְהוָ ֙ה‬ ַ ‫ֶא‬

MT-2

MT

Compared

Samuel 24

Vs

‫את מלאך יהוה‬

‫ע ֵֹ֗מד ֵבּ֤ין ָה ָ֙א ֶר ֙ץ ֵוּב֣ין‬ ‫ַה ָשּׁ ַ֔מיִם‬

̇‫עומד בין[ ̊הארץ ובין‬ ‫]ש[מ]י[ם‬ ̊ ‫̊ה‬

‫לוּפ ֙ה ְבּ ֔יָדוֹ‬ ָ ‫וְ ַח ְר ֤בּוֹ ְשׁ‬

‫וחר]ב[ו̊ ̇שלופה בידו‬

ִ‫ל־יְרוּשׁ ָל֑ם‬ ָ ‫טוּי֖ה ַע‬ ָ ְ‫נ‬

‫ונטוא]ה[ ]על ירושלים‬ ̇

‫וְ ַהזְּ ֵק ִנ֛ים ְמ ֻכ ִ ֥סּים ַבּ ַשּׂ ִ ֖קּים‬ ‫נֵיהם׃‬ ֽ ֶ ‫ל־פּ‬ ְ ‫ַע‬

‫ויפלו הזקנים ע[ל‬ ‫]פנ[י̇ הם‬ ̊ [‫מתכ]סים‬ ̊ ‫ב[ש ̊קים‬

17

29

‫ֹלהים‬ ִ֡ ‫ל־ה ֱא‬ ָ ‫אמר ָ ִדּו֣יד ֶ ֽא‬ ֶ ֹ ‫וַ ֣יּ‬

‫ו]י[אמר דויד אל יהוה‬17

‫ל־יְהוה‬ ָ֜ ‫אמר֩ ָ ֙דּ ִוד ֶא‬ ֶ ֹ ‫ וַ יּ‬17

[‫בראת]ו‬ ̊

‫ִבּ ְרא ֹ֣תוֹ׀‬

‫]את המלאך המכה בעם‬

‫ת־ה ַמּ ְל ָ ֣אְך׀ ַה ַמּ ֶכּ֣ה ָב ֗ ָעם‬ ַ ‫ֶ ֽא‬

‫ויאמר‬

‫אמ ֙ר‬ ֶ ֹ ‫֙ ַויּ‬

Dietrich, 1Samuel1-12, 100-1.

‫‪58‬‬

‫‪MICHAËL N. VAN DER MEER‬‬

‫‪Samuel 24‬‬

‫‪Vs‬‬

‫)‪4QSamuela (DJD‬‬

‫‪MT-2‬‬

‫‪Chronicles 21‬‬

‫ִהנֵּ֙ ה‬

‫הנה‬

‫אתי‬ ‫ָאנ ִ ֹ֤כי ָח ָ֙ט ִ ֙‬

‫אנכי חטאתי‬

‫אתי‬ ‫ר־ח ָ֙ט ִ ֙‬ ‫ֲא ֶשׁ ָ‬

‫וְ ָאנ ִ ֹ֣כי ֶה ֱע ֵ ֔ו ִיתי‬

‫הרעתי‬ ‫הרעה ̇‬ ‫וא[נ̊ כי ̇‬

‫וֹתי‬ ‫וְ ָה ֵ ֣ר ַע ֲה ֵר ֔ע ִ‬

‫וְ ֵ ֥א ֶלּה ַה ֖צּ ֹאן‬

‫]צ[אן̇‬ ‫ואלה ̊ה ̇‬

‫וְ ֵ ֥א ֶלּה ַה ֖צּ ֹאן‬

‫ֶ ֣מה ָע ֑שׂוּ‬

‫מה ]עשו‬ ‫̇‬

‫ֶ ֣מה ָע ֑שׂוּ‬

‫‪MT-1‬‬

‫‪Vs‬‬

‫ֲהל ֹ ֩א ֲאנִ֙ י ָא ַ֜מ ְר ִתּי‬ ‫ִל ְמנ֣ וֹת ָבּ ֗ ָעם‬ ‫נִי־הוּא‬ ‫וַ ֲא ֤‬

‫ֹלהי‬ ‫יְהו֣ה ֱא ַ֗‬ ‫ָ‬ ‫ְתּ ִ֙הי ָנ֥א ְיָדָך֛ ִ ֖בּי ְוּב ֵ ֥בית ָא ִ ֽבי׃ תהי ידך בי ובבית אבי‬

‫ְתּ ִ֙הי ָנ֤א ָי ְֽ ד ָ֙ך ִ ֚בּי ְוּב ֵב֣ית ָא ִ֔בי‬ ‫וּֽ ְב ַע ְמָּך֖ ֥ל ֹא ְל ַמגֵּ ָ ֽפה׃‬

‫פ‬ ‫ל־דּו֖ד ַבּיּ֣ וֹם‬ ‫א־ג֥ד ֶא ָ ִ‬ ‫‪ 18‬וַ יָּב ֹ ָ‬ ‫ַה ֑הוּא‬

‫ביו[ם‬ ‫̊‬ ‫‪18‬ויבוא גד אל דויד‬ ‫ההוא‬

‫ל־גּ֖ד‬ ‫יְהו֛ה ָא ַ ֥מר ֶא ָ‬ ‫וּמ ְל ַ ֧אְך ָ‬ ‫ַ‬

‫אמר ֗לוֹ‬ ‫וַ ֣יּ ֹ ֶ‬

‫ויאמר‬

‫מר ְל ָ ִדו֑יד‬ ‫ֵלא ֣ ֹ‬

‫ֲע ֵל ֙ה‬

‫עלה‬

‫ִ ֣כּי׀ ֲיַע ֶל֣ה ָד ִ ֗ויד‬

‫ָה ֵ ֤ קם ַ ֽליהוָ ֙ה ִמזְ ֵ֔בּ ַח ְבּ ֹ֖ג ֶרן‬ ‫)א ַרנְיָה( ֲ‬ ‫ֲ‬ ‫]א ַ ֥רוְ נָ ה[ ַה ֻיְב ִ ֽסי׃‬

‫ו̊ ]הקם ליהוה[ ]מזבח בגרן ְל ָה ִ ֤ קים ִמזְ ֵ֙בּ ַ֙ח ַל ָ ֔‬ ‫יהוה ְבּ ֹ֖ג ֶרן‬ ‫ָא ְר ָנ֥ ן ַה ֻיְב ִ ֽסי׃‬ ‫ארנא היבסי‬

‫ר־גּד‬ ‫‪ 19‬וַ יַּ ַ֤על ָדּוִ ֙ד ִכּ ְד ַב ֔ ָ‬ ‫יְהוֽה׃‬ ‫ַכּ ֲא ֶ ֖שׁר ִ ָצוּ֥ה ָ‬ ‫‪ 20‬וַ יַּ ְשׁ ֵ ֣ קף ֲא ַ ֗רוְ נָ ה‬ ‫ת־ה ֶ֙מּ ֶל ְ֙ך‬ ‫וַ יַּ ֤ ְ רא ֶא ַ‬ ‫ת־ע ָב ָ ֔דיו‬ ‫וְ ֶא ֲ‬

‫‪19‬ויעל דויד כדבר גד‬

‫ר־גּד‬ ‫וַ יַּ ַ֤על ָדּוִ ֙יד ִבּ ְד ַב ֔ ָ‬

‫כא[שר צוה יהוה‬

‫יְהוֽה׃‬ ‫ֲא ֶ ֥שׁר ִדּ ֶ ֖בּר ְבּ ֵ ֥שׁם ָ‬

‫‪20‬‬

‫̊‬ ‫וישק ̊ף] ארנא‬

‫וַ יָּ ָ֣שׁב ָא ְר ָ֗נן‬

‫וירא[ ]את המלך דויד‬

‫ת־ה ַמּ ְל ָ֔אְך‬ ‫וַ ְיַּר ֙א ֶא ַ‬

‫ואת ארבעת בניו עמו‬ ‫מתחבאים[ בשקים‬

‫וְ ַא ְר ַ ֧בּ ַעת ָבּ ָנ֛יו ִע ֖מּוֹ‬ ‫ִ ֽמ ְת ַח ְבּ ִ ֑אים‬

‫וארנא דש ̇ח ̇ט ̊ים‬

‫וְ ָא ְר ָנ֖ ן ָ ֥דּשׁ ִח ִ ֽטּים׃‬

‫] ויבא דויד עד[ ]ארנא[‬

‫ד־א ְר ָנ֑ ן‬ ‫וַ ֥יָּב ֹא ָ ִדו֖יד ַע ָ‬

‫ו̊ ירא ̊א]ת המלך‬ ‫ע ְֹב ִ ֖רים ָע ָל֑יו‬

‫ועבדיו ערבים אליו‬ ‫בשקי̇ם‬ ‫̇‬ ‫מתכ[סים‬ ‫בא]ים אליו‬ ‫ת־דּ ִ ֔ויד‬ ‫וַ יַּ ֣ ְ רא ֶא ָ‬

‫וַ ֵיֵּצ֣א ֲא ַ ֔רוְ נָ ה‬

‫ויצא ארנא[‬

‫וַ יִּ ְשׁ ַ ֧תּחוּ ַל ֶ ֛מּ ֶלְך ַא ָ ֖פּיו ָ ֽא ְר ָצה׃ ]וישתחו[ לדו̊ י]ד על אפיו‬ ‫ארצא‬

‫ן־ה ֔גּ ֹ ֶרן‬ ‫וַ ֵיֵּצ ֙א ִמ ַ‬ ‫וַ יִּ ְשׁ ַ ֧תּחוּ ְל ָ ִדו֛יד ַא ַ ֖פּיִם‬ ‫ָ ֽא ְר ָצה׃‬

‫‪18‬‬

‫‪19‬‬ ‫‪20‬‬

‫‪21‬‬

EXCLUSION AND EXPANSION

59

Frank Moore Cross and Eugene Ulrich have argued that both 4QSamuela and 1 Chronicles derive from an older version of 1 Samuel 24 and that this version was lost in MT due to haplography.30 Many other scholars, however, hold that 4QSamuela reflects a harmonization of the parallel accounts in 2 Samuel 24 and 1 Chronicles 21 by means of duplication.31 I concur with this opinion. Although 4QSamuela does not systematically duplicate material from Chronicles into the Samuel narratives, this passage about the divine appointment for the place of the Temple apparently did attract the attention of scholarly scribes. 2.5 The Old Greek version of 1 Kings (3 Reigns) Within the Septuagint tradition the books of Samuel and Kings constituted a unity called 1-4 Regum or Kingdoms.32 The character of the Greek translation, however, is not homogenous, since major parts of these books show the same characteristics as the Greek Minor Prophets Scroll from Naḥal Ḥever (8ḤevXIIgr),33 which have become known as the kaigerecension since the groundbreaking work of Dominique Barthélemy.34 The original Greek translation of 1-4 Regum has probably been best preserved by Codex Vaticanus in the sections 1 Regum 1-2 Regum 9 and 3 Regum 30

31

32

33

34

Cross e.a., QumranCave4.XII.1-2Samuel, 193-94; see already Frank Moore Cross, The Ancient library of Qumran and Modern Biblical Studies. The Haskell Lectures 1956-1957 (New York: Doubleday, 1958 [1961]), 189; Eugene C. Ulrich, TheQumran TextandJosephus, HSM 19 (Georgia: Scholars, 1978), 156-57; Eugene Ulrich, “David, the Plague, and the Angel: 2 Samuel 24 Revisited,” in AfterQumran.OldandModern Editions of the Biblical Texts-The Historical Books, eds. Hans Ausloos, Bénédicte Lemmelijn, Julio Trebolle Barrera, BETL 246 (Leuven: Peeters, 2012), 63-79, idem, “The Samuel Scrolls,” in idem, DevelopmentalComposition, 73-108; P. Kyle McCarter, IISamuel, AncB 9 (New York: Doubleday, 1984), 507. See e.g. Stephen J. Pisano, Additions and Omissions in the Books of Samuel, OBO 57 (Freiburg: Universitätsverlag; Göttingen: Vandenhoeck & Ruprecht, 1984), 112-14; Alexander Rofé, “4QSama in the Light of Historico-literary Criticism: The Case of 2 Sam 24 and 1 Chr 21,” in BiblischeundJudaistischeStudien.FestschriftfürPaolo Sacchi, ed. Angelo Vivian, Judentum und Umwelt 29 (Frankfurt am Main: Peter Lang, 1990), 109-19; Anneli Aejmelaeus, “Lost in Reconstruction? On Hebrew and Greek Reconstructions in 2 Sam 24,” BIOSCS 40 (2007): 89-106; Zipora Talshir, “The Relationship between Sam-MT, 4QSama and Chr and the Case of 2 Sam 24,” in IntheFootsteps ofSherlockHolmes.StudiesintheBiblicalTextinHonourofAnneliAejmelaeus, eds. Kristin De Troyer, T. Michael Law, and Marketta Liljeström, CBET 72 (Leuven: Peeters, 2014), 273-98. Philippe Hugo, “1-2 Kingdoms (1-2 Samuel),” in TheT&TClarkCompaniontothe Septuagint, ed. James K. Aitken (London: Bloomsbury, 2015), 127-46. Emanuel Tov, TheGreekMinorProphetsScrollfromNaḥalḤever(8ḤevXIIgr), DJD VIII (Oxford: Clarendon, 1990). Dominique Barthélemy, Lesdevanciersd’Aquila, VTSup 10 (Leiden: Brill, 1963).

60

MICHAËL N. VAN DER MEER

(1 Kgs) 2:12-21:43 and possibly the Lucianic or Antiochene recension for the other parts. Particularly this latter section is full of differences and duplications vis-à-vis the corresponding material in MT-1 Kings. Over the past decades much scholarly attention has been dedicated to the so-called Miscellanies after verse 35 and 46 of 1 Kings 2, where material found elsewhere in the Solomon narratives has been collected in a relatively loose manner,35 as well as the alternative story of Jeroboam’s rise to power found after 3 Reg 12:24.36 Other scholars have taken into account the whole array of additions, omissions, and transpositions found in the Old Greek of 1 Kings.37 Yet, even when we leave out of consideration the textual problems related to the miscellanies and the alternative story, which may reflect transposition rather than literal duplication and reflect a rival version of Jeroboam’s rise to power, we still find several other undisputed examples of duplications. For instance in 3 Reg 16:28a-h we find the material concerning king Jehoshaphat (MT-1 Kgs 22:41-51 = LXX-3 Reg 22:41-51) duplicated. In this way the narratives about king Ahab of northern Israel (16:29-22:41) are now embedded within the repeated reference to the contemporary reign of king Jehoshaphat of Judah. Likewise the information about the reign of king Joram, son of Ahab (MT-2 Kgs 3:1-3 = LXX4 Reg 3:1-3), is found twice in the Septuagint (here the kaige-section), 35

36

37

See e.g. J.A. Montgomery, “The Septuagint at End of 3 Kingdoms 2 (I Reg. 2),” ZAW 50 (1932): 124-29 David W. Gooding, RelicsofAncientExegesis.AStudyofthe Miscellaniesin3Reigns2, SOTMS 4 (Cambridge: Cambridge University Press, 1976); Gerhard Krautwurst, “Studien zu den Septuagintazusätzen in 1. (3.) Könige 2 und ihren Paralleltexten,” (PhD diss., Johannes Gutenburg-Universität Mainz, 1977); Emanuel Tov, “The Septuagint Additions (‘Miscellanies’) in 1 Kings 2,” Textus 11 (1984): 89112; repr. in Tov, TheGreekandHebrewBible, 549-70. See e.g. Jörg Debus, DieSündeJerobeams.StudienzurDarstellingJerobeamsundder GeschichtedesNordreichsinderdeuteronomistischenGeschichtsschreibung, FRLANT 93 (Göttingen: Vandenhoeck & Ruprecht, 1967); David W. Gooding, “The Septuagint’s Rival Versions of Jeroboam’s Rise to Power,” VT 17 (1967): 173-89; Zipora Talshir, TheAlternativeStory.3Kingdoms12:24a-z, JBS 6 (Jerusalem: Simor, 1993). See e.g. Julio Trebolle Barrera, SolomónyJeroboán.Historiadelarecensiónyredacción de1Reyes2-12,14, Institución San Jerónimo 10 (Valencia, 1980); idem, Centenain LibrosSamuelisetRegum.Variantestextualesycomposiciónliterariaenloslibros deSamuelyReyes, TECC 47 (Madrid: CSIC, 1989) ; Adrian Schenker, Septanteet textemassorétiquedansl’histoirelaplusanciennedutextede1Rois2-14, CahRB 48 (Paris: Gabalda, 2000); idem, DieältesteTextgeschichtederKönigsbücher:Diehebräische Vorlage der ursprüngliche Septuaginta als älteste Textform der Königsbücher, OBO 199 (Freiburg: Universitätsverlag; Göttingen: Vandenhock & Ruprecht, 2004); Percy S.F. van Keulen, TwoVersionsoftheSolomonNarrative:AnInquiryintotheRelationshipbetweenMT1Kgs.2-11andLXX3Reg.2-11, VTSup 104 (Leiden: Brill, 2005); A. Turkanik, OfKingsandReigns:AStudyofTranslationTechniqueintheGamma/ GammaSectionof3Reigns(1Kings), FAT 2/30 (Tübingen: Mohr-Siebeck, 2008).

61

EXCLUSION AND EXPANSION

i.e. before (LXX-4 Reg 1:18a-d) and after the story of Eliah’s ascension and succession by Elisha (2 Kings [4 Regum] 2). Just as in the case of the duplications found in LXX-Joshua the style of the Greek is very Hebraistic. Hence it is well possible that the duplications were already made by Hebrew scribes: Table 7: Major Expansions in Samuel-Kings: 4QSamuela and LXX-Regum witness

duplication

source of addition in MT

alignment

4QSamuela (50-20 BCE)

1 Sam 2:9-10

Jer 9:22-23

LXX-1

Reg 2:9-10

2 Sam 24:16-20 1 Chron 21:17-21 LXX-3-4 Reg 3 Reg 2:35a-b (150 BCE ?)38

3 Reg 2:35c

MT-1

Kgs 5:9-10

cf. LXX-3 Reg 5:9-10

MT-1

Kgs 3:1;

cf. LXX-3 Reg 5:14a cf. LXX-3 Reg 5:29

6:38 3 Reg 2:35d-e

MT-1

Kgs 5:29

3 Reg 2:35f-g

MT-1

Kgs 9:24-25 cf. LXX-3 Reg 11:27, 2 Chron 8:12

3 Reg 2:35h

MT-1

Kgs 9:23

3 Reg 2:35i-k

MT-1

Kgs 9:15-18 cf. LXX-3 Reg 10:22; 2 Chron 8:5a

3 Reg 2:35l-o

MT-1

Kgs 2:8-9

(3 Reg 2:46a)

(MT-1 Kgs 4:20)

cf. LXX-3 Reg 2:8-9 (kaige!) transposition ? cf. LXX-3 Reg 2:46k; 2 Chron 9:26; 8:4a

3 Reg 2:46b-d

38

cf. LXX-3 Reg 5:30; 2 Chron 8:10

3 Reg 2:46e-g

MT-1

Kgs 5:2-5

cf. LXX-3 Reg 5:2-5

3 Reg 2:46h

MT-1

Kgs 4:2-6

cf. LXX-3 Reg 4:2-6

(3 Reg 2:46i)

(MT-1 Kgs 5:6)

transposition? cf. 2 Chron 9:25

See for the tentative dating of the non-kaige sections of Kingdoms to the middle of the second century BCE: Marguerite Harl, Gilles Dorival, Olivier Munnich, LaBiblegrecque des Septante du judaïsme hellénistique au christianisme ancien (Paris: Cerf, 1994), 96-97; Timothy Michael Law, “3-4 Kingdoms (1-2 Kings),” in TheT&TClarkCompaniontotheSeptuagint, ed. James K. Aitken (London: Bloomsbury, 2015), 149.

62 witness

MICHAËL N. VAN DER MEER

duplication

source of addition in MT

alignment

3 Reg 12:24a

cf. MT-1 Kgs 11:43; 14:21b-22a

3 Reg 12:24b-c

cf. MT-1 Kgs 11:26, 28, 27

3 Reg 12:24c

cf. MT-1 Kgs 11:40

3 Reg 12:24d

cf. MT-1 Kgs 11:21-22a

3 Reg 12:24e

cf. MT-1 Kgs 11:19b-20

3 Reg 12:24f

cf. MT-1 Kgs 11:22b

3 Reg 12:24g-n

cf. MT-1 Kgs 14:1-7, 12, 10-11, 13, 17

3 Reg 12:24n-o

cf. MT-1 Kgs 12:1; 11:29-31

3 Reg 12:24p-u

cf. MT-1 Kgs 12:3-16, 18b, 20b

3 Reg 12:24x-z

cf. MT-1 Kgs 12:21-24

3 Reg 16:28a-h

MT-1 Kgs 22:41-51

cf. LXX-3 Reg 22:41-51

4 Reg 1:18a-d

MT-2

cf. LXX-4 Reg 3:1-3

Kgs 3:1-3

3. INTERIM

BALANCE

Thus far I have discussed a number of examples of the same type of duplication of authoritative material within the Qumran and Septuagint versions of Joshua and 1-4 Kingdoms that we also find in the Samaritan Pentateuch and pre-Samaritan Qumran manuscripts. I believe these parallels indicate that the scribal activity underlying the harmonistic expansions in the 4QpaleoExodm, 4QNumb and related manuscripts was not restricted exclusively to the Pentateuch, an impression that is still only enforced upon us because of the adoption of one such Pentateuchal manuscript by the Samaritans at the end of the second century BCE. Rather what we see is a more widespread phenomenon attested in Hebrew biblical scrolls from the Hellenistic, Hasmonean and Herodian periods for Joshua (4QJoshuaa) and Samuel (4QSamuela) as well. The same phenomena can be traced back to the Greek versions of Joshua and Kingdoms as well. This should not surprise us since duplication of authoritative material is not uncommon in the later textual transmission of the Bible. In my study of the textual variations in Num 14:23, 31 and related

EXCLUSION AND EXPANSION

63

passages in Num 32:11, and Deut 1:35, 39 I found an example of such a duplication in such a late textual witness as the Peshitta to Num 14:23.39 It is tempting to continue this quest for expanded biblical texts beyond the corpus of Genesis-2 Kings. The great Isaiah scroll (1QIsaa) seems to be a self-evident candidate for such an enterprise40, although the expansions are not as sizable as we find in the Samaritan Pentateuch and the longer Qumran scrolls 4QpaleoExodm and 4QNumb.41 For now I will leave that investigation to others. Instead I would like to conclude this all too sketchy survey of expansionistic texts by looking at contemporary literary papyri outside the realm of Jewish and Samaritan religious traditions. I have in mind the Ptolemaic papyri of Homer that to a considerable extent reflect similar phenomena.

4. THE PTOLEMAIC PAPYRI OF HOMER Since the nineties of the nineteenth century an enormous amount of Greek papyri of the Iliad and Odyssey have come to light that are considerably older than the extant Medieval manuscripts. These papyri were uncovered during archaeological campaigns in Egypt. At present some 1600 fragments of the Homeric writings on papyrus, parchment, ostraca, wax boards and wooden tablets have been found and published.42 As is well-known, the Iliad formed part of the standard curriculum of any Greek education 39

40 41

42

Michaël N. van der Meer, “The Next Generation. Textual Moves in Numbers 14,23 and Related Passages,” in TheBooksofLeviticusandNumbers, ed. Thomas Römer, BETL 215 (Leuven: Peeters, 2008), 399-416. See Avraham Tal’s contribution to this volume, pp. 263ff. See e.g. Eduard Y. Kutscher, The Language and Linguistic Background of the Isaiah Scroll(1QIsaa), STDJ 6 (Leiden: Brill, 1974); Jean Koenig, L’herméneutiqueanalogique du Judaïsme antique d’après les témoignes textuels d’Isaïe, VTSup 33 (Leiden: Brill, 1982); Arie van der Kooij, Die alten Textzeugen des Jesajabuches. Ein Beitrag zur TextgeschichtedesAltenTestaments, OBO 35 (Freiburg: Universitätsverlag; Göttingen: Vandenhoeck & Ruprecht, 1981), 74-119; Paulson Pulikottil, TransmissionofBiblical TextsinQumran.TheCaseoftheLargeIsaiahScroll1QIsaa, JSPSup 34 (Sheffield: Sheffield Academic, 2001); Eugene Ulrich, Peter W. Flint, and Martin Abegg jr., eds., QumranCave1.IITheIsaiahScrolls.Part2:Introductions,Commentary,andTextual Variants, DJD XXXII.2 (Oxford: Clarendon, 2010),88-92. Graeme D. Bird, Mulitextuality in the Homeric Iliad. The Witness of the Ptolemaic Papyri, Hellenic Studies 43(Cambridge, MA: Harvard University Press, 2010), 64. See further Michael Haslam, “Homeric Papyri and Transmission of the Text,” in A NewCompaniontoHomer, eds. Ian Morris and Barry Powell, Mnemosyne 163 (Leiden: Brill, 1997), 55-100; Antonios Rengakos, “Die Überlieferungsgeschichte der homerischen Epen,” in Homer-Handbuch.Leben-Wirken-Wirkung, eds. Antonios Rengakos and Bernhard Zimmermann (Stuttgart: Metlzer, 2011), 167b-175a.

64

MICHAËL N. VAN DER MEER

in Antiquity, hence the multitude of writing exercises and private and public, official and semi-official forms of the Iliad.43 Table 8: Distribution of textual witnesses of the Iliad (after G.D. Bird)44

The earliest of these fragments date from the third, perhaps even the fourth century BCE, which makes them contemporaries of, or even older than, the Dead Sea Scrolls. As is the case with the textual witnesses of the writings of the Hebrew Bible, we also see often that the older the Greek papyrus of the Iliad is, the more deviations it displays vis-à-vis the later, established text of the medieval manuscripts. In modern scholarship devoted to the Homeric writings these papyri are often designated as “eccentric manuscripts” as opposed to what is called the “vulgate version of the Iliad”.45 Many scholars consider them to be ancient testimonies to 43

44 45

See e.g. Henri-Irénée Marrou, Histoire de l’éducation dans l’Antiquité (Paris: Seuil, 1948); Martin P. Nilsson, DiehellenistischeSchule (München: Beck, 1955); Raffaella Cribiore, Gymnastics of the Mind. Greek Education in Hellenistic and Roman Egypt (Princeton: Princeton University Press, 2001). Bird, Multitextuality, 64; online: http://chs.harvard.edu/CHS/article/display/4853. See Stephanie West, The Ptolemaic Papyri of Homer, Papyrologica Coloniensia III (Köln: Westdeutscher Verlag, 1967); Rudolf Pfeiffer, HistoryofClassicalScholarship 1. FromtheBeginningstotheendoftheHellenisticAge (Oxford: Clarendon, 1968), 105-22; L.D. Reynolds, N.G. Wilson, ScribesandScholars.AGuidetotheTransmissionofGreekandLatinLiterature (Oxford: Clarendon, 1968); Egert Pöhlmann, EinführungindieÜberlieferungsgeschichteundindieTextkritikderantikenLiteratur 1. Altertum, Die Altertumswissenschaft (Darmstadt: Wissenschaftliche Buchgesellschaft, 1994).

EXCLUSION AND EXPANSION

65

a period of textual fluidity and multitextuality, a time when every oral performance by a rhapsode was a single event and the time of absolute textual uniformity.46 Since manuscripts written later than the middle of the second century BCE hardly show the same type of variation, it is assumed that a process of stabilization and fixation took place around 150 BCE. In schematic form these models can be displayed as follows: Table 9: Models of Homeric Transmission (after G.D. Bird)47 Conventional view of Homeric transmission

Multitextuality model of Homeric transmission

These discussions in the field of Homer studies remind the biblical scholar of discussions about the canonization of the writings of the Hebrew Bible, but also of the debate between the position of Paul de Lagarde and Paul Kahle about the original state of the biblical text: from single archetype to plurality of textual forms (de Lagarde) or from initial multiplicity towards standardization (Kahle). This discussion seemed to be settled in favor of the former position, but nevertheless lingers on underneath many recent publications. I bring up these Ptolemaic papyri of Homer in a discussion about the Samaritan Pentateuch and the Dead Sea Scrolls because the special characteristic of these papyri is the high number of what is called “plus verses”. 46

47

See e.g. the well-known works by Albert B. Lord, TheSingerofTales (Cambridge, MA: Harvard Univerity Press, 1960) and Milman Parry, TheMakingofHomericVerse.The CollectedPapersofMilmanParry, ed. Adam Perry (Oxford: Oxford University Press, 1987); Gregory Nagy, Homer’sTextandLanguage (Urbana: University of Illinois Press, 2004), Bird, Multitextuality. Bird, Multitextuality, 41, 43, online: http://chs.harvard.edu/CHS/article/display/4853

66

MICHAËL N. VAN DER MEER

Although these “eccentric” Ptolemaic papyri of Homer occasionally also display elements of “minus verses” and transposition of verses, they overwhelmingly contain extra verses when compared to the later textual tradition. In some cases these plus verses may contain more original lines that were deliberately deleted by the Alexandrian Homeric scholars such as Zenodotus of Ephesus (fl. c. 280 BCE), Aristophanes of Byzance (c. 257 BCE–c. 185 BCE), and Aristarchus of Samothrace (ca. 220 BCE– ca. 143 BCE). In some cases the plus verses are actually transpositions of verses from a context that is now lost due to the fragmentary state of these early papyri. However, in the majority of cases these plus verses look very much like the phenomenon we have discussed so far, i.e. deliberate duplications of material found elsewhere in the Homeric corpus, including not only the Iliad and Odyssey, but also the Homeric hymns and even the epic literature of Hesiod. In order to illustrate this phenomenon I will briefly discuss some examples found in one of the oldest extant manuscript of the Iliad, papyrus 7 (c. 285-260 BCE), divided over the London (P.Tebt. I 21) and Heidelberg (P.Heid. 1261) collections.48 Book 8 of the Iliad opens with a council of the Olympian deities (8.1-41) who receive a warning by Zeus not to interfere with the Trojan war and Zeus’ departure to mount Ida in order to overlook the course of the war personally (8.42-53). After that follows a description of the usual course during the Trojan War. After lines 50, 52, 54 and 55, 65 we find interpolations from Iliad 5.777 (8.50a), 21.449 (8.51a), 11.83 (8.52a), 8.1 (8.52b), 2.477-479 (54a-d), 11.57-60 (55a-d) and 18.535-537, 540 (8.65a-d) 4.439-441, 444-445 (8.65e-i): line Papyrus 7 Iliad 8.49-67 (after Bolling and West)

Translation (after Wyatt)49

49

[ἔνθ’ ἵππους ἔστησε Κρόνου πάϊς There did thesonofcrookedἀγκυλομ]ήτεω counselingCronos (ǁ 2.205, 309; 4.75; 9.37; 12.450; 16.431; 18.293; Od. 21.415) stay his horses,

50

[λύσας ἐξ ὀχέων, κατὰ δ’ ἠέρα πουλὺν ἔχευ]ὲν.

48

49

and loose them from his chariot, and shed thick mist over them (ǁ 5.776)

West, PtolemaicPapyri, 74-90; Bernard P. Grenfell and Arthur S. Hunt, eds., TheHibeh Papyri.Part1 (London: Egypt Exploration Society, 1906), 88-96; Gustav A. Gerhard, PtolemäischeHomerfragmente, P.Heid. IV.1 (Heidelberg: Winter, 1911), 8-19. See further George M. Bolling, TheExternalEvidenceforInterpolationinHomer (Oxford: Clarendon, 1925), 100-5. Willliam F. Wyatt, Homer.Iliad.Books1-12;Books13-14, LCL 170-71 (Cambridge: Massachusetts, 1999).

EXCLUSION AND EXPANSION

67

line Papyrus 7 Iliad 8.49-67 (after Bolling and West)

Translation (after Wyatt)49

50a [τοῖσιν δ’ ἀμβροσίην Σιμόεις ἀνέτειλε νέμεσθαι·]

andSimoïsmadeambrosiatospring upforthemtograzeon (=5.777)

51

[αὐτὸς δ’ ἐν κορυφῆισι καθέζετο κύδεϊ γαί]ὼν,

51a [Ἴδης ἐν κορυφῆισι πολυπτύχου, ὑληέσσης,] 52

[εἰσορόων Τρώων τε πόλιν καὶ νῆας Ἀχαιῶν]

And he himself sat exulting in his glory among the mountain peaks amongthepeaksofwoodedIda, themany-ridged (=21.449) looking upon the city of the Trojans and the ships of the Achaeans (ǁ 11.82),

52a [χαλκοῦ τε στεροπῆν ὀλλύντας τ’ oftheflashingofbronze,andon ὀλλυμένους τε.] theslayersandtheslain (=11.83) 52b [Ἠὼς μὲν κροκόπεπλος ἐκίδνατο πᾶσαν ἐπ’ αἶαν,]

NowDawnthesaffron-robedwas spreadingoverthefaceoftheearth (=8.1)

53

[οἱ δ’ ἄρα δεῖπνον ἕλοντο κάρη κομόωντες Ἀχαιο]ὶ

54

[ῥίμφα κατὰ κλισίας, ἀπὸ δ’ αὐτοῦ hastily throughout the huts, and after θωρήσσον]το. eating arrayed themselves in armor,

but the long-haired Achaeans took their meals,

54a [τοὺς διεκόσμεον πάντες, ὅσοι πάρος ἦσαν ἄριστο]ι,

theywouldmarshalallthose (=2.476a) whowerebestmen before(=11.825b),

54b [πάντη ἐποιχόμενοι· μετὰ δε κρείων] Ἀγα[μ]έμνων,

goingeverywhere(=10.167a); amongthemlordAgamemnon (=2.477b)

54c [ὄμματα καὶ κεφαλὴν ἴκελος Διὶ τερ]πικερ[αύ]νωι,

hiseyesandheadlikeZeuswho hurlsthethunderbolt (=2.478)

54d [Ἄ]ρεϊ δὲ ζ[ώνην, στέρνον δὲ Ποσειδάω]νι.

hiswaistlikeAres,andhisbreast likePoseidon (=2.479)

Τρῶες δ’ [αὖθ’ ἑτέρωθεν ἀνὰ πτόλι]ν ὡπλί[ζο]ντο

In like manner the Trojans on their side armed themselves throughout the city (ǁ 11.56);

55

55a Ἕκτορά τ’ [ἀμφὶ μέγαν καὶ ἀμύμον]α Που[λυ]δάμαντα

andwerearrayedaboutgreat Hectorandincomparable Polydamas (=11.57)

55b Αἰνεία[ν θ’, ὃς Τρωσὶ θεὸς ὣς τίετο δή]μω[ι,]

andAeneas,whowashonoredby thepeopleoftheTrojanslikea god (=11.58),

55c τρεῖς τ’ Ἀ[ντηνορίδας, Πόλυβον καὶ Ἀγήνορα δῖον]

andthethreesonsofAtenor, PolybusandnobleAgenor (=11.59)

68

MICHAËL N. VAN DER MEER

line Papyrus 7 Iliad 8.49-67 (after Bolling and West)

Translation (after Wyatt)49

55d ἠίθεόν τε Ἀκά[μαντ’, ἐπιείκελον ἀθανάτοισιν,] 56 παυρότεροι· μέμ[ασαν δὲ καὶ ὣς ὑσμῖνι μάχεσθαι] 57 χρηῆι ἀναγκα[ίηι, πρό τε παίδων καὶ πρὸ γυναικῶν.] 58 πᾶσαι δὲ ὠΐγον[το πύλαι, ἐκ δ’ ἔσσυτο λαός,] 59 πεζο[ί θ’ ἱ]ππῆ[ές τε· πολὺς δ’ ὀρυμαγδὸς ὀρώρει.] 60 οἳ δ’ ὅτε [δή] ῥ’ ἐ[ς] χ[ῶρον ἕνα ξυνιόντες ἵκοντο,] 61 σύρ ῥ’ ἔβ[α]λον ῥ[ινούς, σὺν δ’ ἔγχεα καὶ μένε ἀνδρῶν]

andyoungAcamas,peerofthe immortals (=11.60) fewer they were, but even so they were eager to contend in battle through utter necessity, to protect their children and their wives. And all the gates were opened, and the army rushed out, both foot soldiers and charioteers; and a great din arose. Now when they had met together and come into one place, then they dashed together their shields and spears and the fury of bronze-mailed warriors; and the bossed shields pressed

χαλκ[εοθ]ωρήκ[ων· ἀτὰρ ἀσπίδες ὀμφαλόεσσαι] 63 ἔπλη[ντ’ ἀλλ]ή[ληισι, πολὺς δ’ ὀρυμαγδὸς ὀρώσει.] 64 ἔνθα δ’ [ἅμ’] οἰμω[γή τε καὶ εὐχωλὴ πέλεν ἀνδρῶν] 65 ὀλλύντων τ[ε καὶ ὀλλυμένον, ῥέε δ’ αἴματι γαῖα.] 65a ἐν δ’ Ἔρις, [ἐ]ν δὲ Κ[υδοιμὸς ὁμίλεον, ἐν δ’ ὀλοὴ Κήρ,] 62

65b ἄλλον ζ[ω]{ι}ὸν ἔ[χουσα νεούτατον, ἄλλον ἄουτον,] 65c ἄλλον τε[θ]ν[ηῶτα κατὰ μόθον ἕλκε ποδοῖιν·] 65d ν[εκρούς] τ’ [ἀνθρώπων ἔρυον κατατεθνηώτων.] 65e [ὦρσε δὲ τοὺς μὲν Ἄρης, τοὺς δὲ γλαυκῶπις Ἀθήνη] 65f [Δεῖμός τ’ ἠδὲ Φόβος καὶ Ἔρις, ἄμοτον μεμαυῖα,]

one on another, and a great din arose. Then were heard alike the sound of groaning and the cry of triumph of the slayers and the slain, and the earth flowed with blood. AmongthemStrifeandTumult joined,anddestructiveFate (=18.535; Hesiod, Shield 156), graspingonemanalive,freshwounded,anotherwithoutawound (=18.536; Hesiod, Shield 157), andanothershedraggeddead throughthemeleebythefeet (=18.537; Hesiod, Shield 158) andtheyeachweredraggingaway thebodiesoftheother’sslain (=18.540). AndtheTrojanswereurgedby Ares,andtheAchaeansbyflashingeyedAthene (=4.439), andTerror,andRout,andStrife whoragesincessantly (=4.440)

69

EXCLUSION AND EXPANSION

line Papyrus 7 Iliad 8.49-67 (after Bolling and West)

Translation (after Wyatt)49

65g [Ἄρεος ἀνδροφόνοιο κασιγνήτη ἑτάρη τε,]

sisterandcomradeofman-slaying Ares (=4.441);

65h [ἥ σφιν καὶ τότε μέσσωι ὁμοίιον ἧκ]εν ὄλεθρο[ν]

Sheitwaswhonowcastevilstrife intotheirmidst (=4.444)

65i [ἐρχομένη καθ’ ὅμιλον, ὀφέλλουσ’] εὖ στόν[ον ἀνδρῶν.]

asshewentthroughthethrong, makingthegroaningsofmento increase (=4.445)

66

ὄφρ[α] μὲ[ν ἠὼ]ς ἦν [καὶ ἀέξετο ἱ]ε[ρὸν ἦμαρ,

Now as long as it was morning and the holy day was waxing,

67

τόφρα μά[λ’ ἀ]μφοτ[έρων βέλε’ ἤπτετο, πῖπτε δὲ λαός.

so long the missiles of each side reached their mark, and the men kept falling.

Whereas the first part of this section contains expansions of only a single line (8.50a, 51a, 52a-b), the latter part includes whole sections of passages found elsewhere in the Ilias. The description of the course of battle is enriched with descriptions of the leaders of both camps: Agamemnon on the Achaean side (8.54b-d = 2.477-479) and the Trojan heroes, Hector, Polydamas, Aeneas, Atenor, Polybus, Agenor and Acamas (8.55a-d = 11.57-60) on the other. The expansion attested in this early Ptolemaic papyrus is facilitated by the fact that the same line occurs here and in expanded form elsewhere in the Ilias (8.50 || 5.776+777 > 8.50a; 8.52 || 11.82+83 > 8.52a; 8.55 || 11.56+57-60 > 8.55a-d). The inclusion of 11.83 as 8.52a frames the description of the battle as arena for slayers and slain (ὀλλύντας τ’ ὀλλυμένους τε cf. 8.65) as does the reference to the increasing groan of falling warriors (4.445 as 8.65i harking back to 8.64). The table below provides a provisional attempt to list the plus verses in the Ptolemaic papyri of both the Iliad and Odyssey with the “source” of the alleged duplications: Table 11: Expansions in the Ptolemaic Papyri of Iliad (after S. West) witness

date

duplication

P.53 (P.Vitelli)

second half 1.484(a-)z second century BCE

P.269 (P.Tebt. III/2.898)

late third century BCE

1.543a

source of addition ≈ Homeric hymn to Apollo (503-07) cf. P.121 12.250a; P.12 22.126a

70

MICHAËL N. VAN DER MEER

witness

date

P.40 (P.Hibeh I 19)

285-250 BCE

duplication 2.794a 2.855a-b 3.283a 3.302a-d

3.304a 3.334-337 3.339a-c

3.366a

source of addition cf. 2.465; 3.252; 21.300; 24.332; Odyssey 3.485 cf. Strabo (542); Sch. T ad 20.329; Eust. 1210,46 = 3.75, 258 = 2.39-40; 14.280; Odyssey 2.378; 10.346; h.Ap. 89; cf. Odyssey 12.304; 15.438; 18.59 cf. 7.68; 7.349; Odyssey 7.187; 17.469; 18.352 transposition? cf. Odyssey 1.256; Iliad 6.322; 13.527; 18.459; 3.331; 2.45; 3.334; 16.135; 19.372 = 16.766; cf. 3.329; 7.355; 8.82

P.41 (P.Hibeh I 20)

280-240 BCE

4.69a

cf. 4.515; Odyssey 3.378

P.Mich. 5 (anthology)

early third century BCE

3.425a 3.429a

cf. 19.39 cf. 3.53

P.7 (P.Grenf. 2,2; P.Hibeh I 21; P.Heid. 1261

285-260 BCE

8.38a 8.52a-d

= 5.426; 15.47 = 7.421-423 & Odyssey 3.2-3 = 2.477-479 = 11.57-60 = 4.439-445 cf. 15.224 cf. 12.162; 15.113f; 15.397f, 16.124f.; Odyssey 15.198f.; h.Cer. 245f = 8.354-355 cf. 7.21; 13.347; 16.121; 23.682 ? (too fragmentary for reconstruction) = 8.130; 11.310 cf. 8.397, 438; 11.544 ? (too fragmentary for reconstruction)

8.54a-d 8.55a-d 8.65a-i 8.197a 8.199a

8.202a-b 8.204a 8.206a 8.216a 8.252a-b 8.255a

71

EXCLUSION AND EXPANSION

witness P.432 (P.Hamb. I 153)

date latter half of third cent. BCE

duplication 11.266a-z 11.272a 12.130a

source of addition cf. 11.531-533; 17.456-458; h.Cer. 88f; Hes. Sc. 341f 10.15 2.746

12.189a-b

cf. 4.504; 5.540; 13.187; 17.50, 311; Odyssey 24.525

12.190a

= 2.746 cf. 12.130a

12.193a

= 12.140; cf. 12.139

P.8 (P.Lit.Lond. 21) third century BCE

11.504a 11.509a 11.513a 11.514a

– – cf. 14.429 –

P.5 (P.Gen. I 90)

11.795a-b 11.804a

= 1.538; 1.556 cf. 16.52; 23.223; h.Cer. 56 – cf. 18.3, 19.344 = 8.235; cf. 15.507; 22.374; 11.664-665; 11.622

latter half of third cent. BCE

11.805a 11.807a 11.827a-c

P.121+342 (P.Lit Lond. 251+ P.Harris 36)

second century BCE 12.183a

cf. 16.346; replaces 12.184-186

12.189a-b

cf. 4.528 cf. P. 432

12.190a

cf. P.432 (2.746 cf. 12.130a)

12.250a 12.360a

19.139 ? (too fragmentary for reconstruction) ? (too fragmentary for reconstruction)

12.363a P.51 (P.Berlin 9774) first century BCE

18.606a 18.608a-d

cf. 18.494f = Hes. Sc. 207-215

P.12 (P.Grenf. 2,4l P.Hibeh I 22; P.Heid. 1262-66

21.382a 22.99a

cf. 12.33 – cf. Hes. Sc. 366

22.126a

? (too fragmentary for reconstruction)

22.259a-b 22.316a-c

= 22.342-343 = 7.79-80 = 22.133-135 (transposition)

280-240 BCE

72

MICHAËL N. VAN DER MEER

witness

date

duplication

source of addition

22.392a

– cf. 24.20; 10.52; 23.183a

23.93a

cf. Odyssey 4.809

23.130a

? (too fragmentary for reconstruction)

23.136a

cf. 18.23, 27

23.155a

? (too fragmentary for reconstruction)

23.160a

?

23.162a

= 2.399

23.165a

cf. 24.165; Odyssey 9.247

23.183a

–

23.195a

4.102, 120; 23.864

23.209a 23.223a-b

cf. 17.36 cf. Plutarch, Consol.ad Apoll. 117D

23.278a-b

cf. Odyssey 5.212f.

Table 12: Expansions in the Ptolemaic Papyri of Odyssey (after S. West) witness

date

duplication

source of addition

P.22 (P.Tebt. III/1 696)

1.92a early second century BCE

cf. 1.368, 4.321; 17.370, 468; 18.351; 21.275

P.30 (P.Tebt. III/1 697)

second century BCE 5.27a

? (too fragmentary for reconstruction) ? (too fragmentary for reconstruction) cf. 5.113 cf. Iliad 11.201; Odyssey 4.829 – 4.825; cf. h.Ven. 193-194; Iliad 12.246; Odyssey 5.347 Cf. Iliad 14.184-185 –

5.32a-b 5.40a 5.104a 5.111a 5.183a-b 5.232a-b 5.259a

73

EXCLUSION AND EXPANSION

witness

date

duplication

source of addition

P.110 (P.Merton 1)

second century BCE 6.206a

P.31 (P.Sorbonne inv. 2245A)

250-200 BCE

P.126 (P.Hamburg 154)

second century BCE 16.176a 16.177a

P.19 (P.Hibeh I 23)

285-250 BCE

20.51a 20.55a 20.58a

cf. Iliad 14.47 cf. Iliad 1.222 cf. Callimachus frg. 238 v.9

P.146 (P.Lefort 1)

second half of third century BCE

21.19a

? (too fragmentary for reconstruction)

9.412a 9.519a 9.522a 9.537a

? (too fragmentary for reconstruction) = 9.519 cf. 9.529 = 9.511 = 9.501 cf. Iliad 11.669; Odyssey 11.394; 21.283; h.Ven. 157, 238 = 6.231 ?

The analogy with the expanded biblical text is striking, especially in those cases where larger sections from elsewhere in the Homeric corpus have parallels in the extended version attested by these “eccentric” Ptolemaic papyri. In many cases modern scholars hold the expansions to be later additions to the more original text preserved in the later so-called “vulgate text”. These expansions are taken not only from the stock epic phraseology or poetic bridges between the various episodes, but also include complete descriptions taken from the Homeric hymns (to Apollo, Ceres or Venus) as well as the work of Hesiod (see for example above Ilias 8.65a-c || Hesiod, Shield, 156-158). The fact that two different ancient manuscripts (P. Hamb. I 153 and P.Lit.Lond.251+P.Harris 36) reflect the same expansions (12.189a-b, 190a) makes it clear that these additions are not random interpolations of individual scribes, but point to a common literary ancestor from which individual manuscripts were copied. I doubt therefore whether all the variants in these early Ptolemaic Homer manuscripts should be ascribed to different recordings of ancient oral performances, as the Nagy school seems to believe. I find it more likely to ascribe duplications of Homeric material within the same corpus to the literary activity of the Homeric scholars at Alexandria. I think it is legitimate to go a step further: Since the Iliad formed the core of the curriculum of any Greek education in Antiquity and since Jewish elites in the Greco-Roman period could not afford not to provide

74

MICHAËL N. VAN DER MEER

at least some Greek education for their children as the many Jewish Greek writings from the Hellenistic and Roman period make clear (e.g. undoubtedly bilingual high officials in the Hasmonean Empire like Eupolemos and the Greek translators of the biblical books), it is probable that the phenomenon of expansion of authoritative writings by means of duplication of material from that same corpus as found in the ancient biblical manuscripts has a Hellenistic-Alexandrian background. After all, both the Iliad and the Hebrew Bible had gained what we would now call a canonical status during the Hellenistic period.50 The second century BCE composition Exagoge written by Ezekiel Tragicus provides a fascinating blending of biblical narrative (as told in the book of Exodus) and Homeric language.51 Such an influence of Homeric scholarship upon the transmission and interpretation of the Hebrew Bible has been postulated by several scholars over the last years. Sylvie Honigman has assumed a direct influence of Homeric scholarship not only on the creation of the so-called Letter of Aristeas during the middle of the second century BCE, but also on the production of the Septuagint already at the beginning of the third century BCE.52 Maren Niehoff sees Homeric scholarship rather influencing Jewish exegesis of the Pentateuch just as the Homeric scholars in Alexandria were producing their scholia on the Homeric writings.53 In both cases it remains difficult to pinpoint a direct relationship between the work of the Alexandrian scholars and their Jewish contemporaries. It has even been suggested that the number of the 24 books of the Hebrew Bible as reflected in early Jewish sources as 4 Ezra 12:45 reflects the paradigm set by the Homeric epics which also consist of 24 books.54 In a recent article Jonathan Ben-Dov argues for a correlation between Alexandrian scholarly editions of Homer and the academic edition of the Pentateuch.55 Ben-Dov does not pay attention to the non-Pentateuchal 50

51

52

53

54

55

Margalit Finkelstein and Guy G. Stroumsa, eds., Homer,theBible,andBeyond.Literary andReligiousCanonsintheAncientWorld, Jerusalem Studies in Religion and Culture 2 (Leiden: Brill, 2003). See e.g. Howard Jacobson, TheExagogeofEzekiel (Cambridge: Cambridge University Press, 1983). See now P.Oxy. LXXXIII 5348. Sylvie Honigman, TheSeptuagintandHomericScholarshipinAlexandria:aStudyin theNarrativeoftheLetterofAristeas (London: Routledge, 2003). Maren Niehoff, JewishExegesisandHomericScholarshipinAlexandria (Cambridge: Cambridge University Press, 2011). Guy Darshan, “The Twenty-Four Books of the Hebrew Bible and Alexandrian Scribal Methods,” in Homer and the Bible in the Eyes of Ancient Interpreters, ed. Maren R. Niehoff, Jerusalem Studies in Religion and Culture 16 (Leiden: Brill, 2012), 221-44. Jonathan Ben-Dov, “Early Texts of the Torah. Revisting the Greek Scholarly Context,” JournalofAncientJudaism 4 (2013): 210-34. I thank the author for bringing this very important article to my attention during the symposium.

EXCLUSION AND EXPANSION

75

biblical manuscripts or the early Ptolemaic Homeric papyri, but does point to the Greek academic background against which many of the duplications in pre-Samaritan Pentateuchal manuscripts can be understood. He points to the philological tradition of explaining Homer within his own context (Ὅμηρον ἐξ Ὁμήρου σαφηνίζειν) that started with Aristotle’s AporemataHomerica and continued in the Alexandrian commentaries on Homer. Ben-Dov does not postulate a direct literary dependence of the academic biblical text upon Aristotelian and Alexandrian Greek scholarship, but argues for a common intellectual background. I find myself in agreement with almost everything written in his highly important scholarly contribution with the exception of his statement that the “solution adopted” (viz. by Hebrew scholarly scribes) “i.e., the supplementation of large sections from other parts of the Pentateuch is not attested in Greek Homeric scholarship.”56 The evidence adduced above makes clear that there is reason to see this very same phenomenon of duplication of parallel material attested already in the earliest Homeric papyri. Needless to say, establishing a relationship for the expanded Homeric manuscripts and the expanded biblical manuscripts has to remain hypothetical until further evidence comes to light. Nevertheless we do have early epigraphical material (Homeric papyri) to work from instead of later testimonies (Letter of Aristeas, Philo). Hence, I believe there is sufficient reason to investigate this parallel between the Greek and Hebrew manuscripts from the third and second century BCE further.

5. CONCLUSION: EXCLUSION AND EXPANSION The Bible used by the Samaritans can be characterized with the terms “exclusion” and “expansion.” On the one hand, the Samaritan canon excludes all ancient Hebrew religious literature apart from the Pentateuch. On the other hand, the text of the Samaritan Pentateuch contains a large number of harmonistic expansions. We now know that the Samaritans are not solely responsible for most of these expansions, but rather adopted one of several expanded versions of the Pentateuch current and cultivated among priestly circles during the Ptolemaic and Hasmonean periods. Just as there is no compelling reason anymore to ascribe the expansions in the plague narratives (Exod 7-11) and the itinerary narratives (passages in Exodus-Numbers corresponding to Deut 1-3) only to the Samaritans, I also think there is no compelling reason to see the scribal activity reflected 56

Ben-Dov, “Early Texts of the Torah,” 224.

76

MICHAËL N. VAN DER MEER

in these writings restricted exclusively to the Pentateuch. Rather, I believe there are good reasons to explain a number of unusual textual phenomena in 4QJoshuaa and 4QSamuela as well as the Old Greek versions of Samuel and Kingdoms as examples of the same expansionistic scribal concerns. It may well be that this phenomenon is not exclusively either Jewish, Sadducean or even Palestinian after all, but rather the result of the influence of the cultural decidedly non-Jewish elite at the Alexandrian courts and cultural institutions.57

57

See now: Pieter B. Hartog, PesherandHypomnema:AComparisonofTwoCommentaryTraditionsfromtheHellenistic-RomanPeriod, STDJ 121 (Leiden: Brill, 2017).

THE SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH Stefan SCHORCH

1. INTRODUCTION: PRE-SAMARITAN

VERSUS

SAMARITAN

The question whether the Samaritan Pentateuch contains a textual layer of ideological features that are distinctively “Samaritan” has been an important topic of research throughout the last decades. According to a suggestion of Emanuel Tov and others, such a presumed layer would be the most important distinction between the Samaritan Pentateuch in the stricter sense, on the one hand, and the texts found in the pre-Samaritan manuscripts, on the other. In the second edition of his “Textual criticism of the Hebrew Bible” from 2001, Tov writes as follows:1 One can now distinguish between the pre-Samaritan substratum and a second, Samaritan, layer added in ⅏. This second layer is relatively thin and if we could remove it, we would clearly see the pre-Samaritan text, upon which the Samaritan text was based. The criteria for this separation between the two layers are on the one hand, the features of the preSamaritan texts, with whose help we can isolate the substratum of ⅏, and on the other hand, the characteristics of the Samaritan religion, literature, and language, by means of which we can isolate the second layer of their Torah.

According to the same book, the characteristics of the pre-Samaritan text consist of “Harmonizing Alterations”, “Linguistic corrections,” “Content differences,” and “Linguistic Differences,”2 among which the first seems by far the most important, while the linguistic peculiarities of both layers, in relation to other textual traditions as well as in comparison with each other, seem much less coherent and distinctive. Consequently, the criteria identified by Tov as decisive for each of the two subsequent layers, in terms of literary history, focus on quite different dimensions of the textual corpus: while the pre-Samaritan substratum is above all characterized by literary features, the Samaritan adstratum is determined predominantly by references to Samaritan beliefs. 1 2

Tov, TextualCriticismoftheHebrewBible2, 84–85. Ibidem, 85–94.

78

STEFAN SCHORCH

To some extent, this use of different reference frames is due to the diversity of the two main corpora of sources itself: while the corpus of pre-Samaritan texts comprises exclusively manuscripts from Qumran, of which scarcely any historical context beyond the other manuscripts and textual traditions from that period is known, acquaintance with the Samaritan Pentateuch stems mainly from medieval manuscripts copied and transmitted by Samaritans. Nevertheless, from a methodological point of view, the application of two different reference frames for the separation of two presumed textual layers found within one textual body is problematic, since the procedure requires in fact two separate analytical operations that lack a common denominator, and it is thus rather susceptible for misinterpretation of data. A few glimpses at the history of research in the Samaritan Pentateuch seem only to reinforce these doubts about the reliability of the distinction between pre-Samaritan and Samaritan on the basis of the above-mentioned criteria: the first scholar who systematically used the concept of ideological changes as a means to discern the Samaritan Pentateuch from other textual traditions of the Hebrew Bible was Wilhelm Gesenius in his study “De Pentateuchi Samaritani origine, indole et auctoritate commentatio philologico-critica” from 1815. The book contains a whole paragraph dealing with “Loca ad theologiam et hermeneuticam Samaritanorum domesticam conformata” (“Readings adapted to Samaritan exegesis”), subdivided into five different categories, in accordance with respective tenets of the Samaritan beliefs. Due to a much better awareness of the ideological and textual variety of Judaism and its canonical texts in the Late Second Temple period, the list of Samaritan ideologisms in the Samaritan Pentateuch has shrunk to three cases in the second edition of Tov’s handbook, published in 2001,3 namely the supposed Samaritan changes of the text from Ebal to Gerizim in Deut 27:4, from ‫ יבחר‬to ‫ בחר‬in the Deuteronomic centralization formula (Deut 12:5 et al.), and the insertion of a passage devoted to Mount Gerizim after the Ten commandments, the so-called Gerizim-commandment. In the third edition of 2012, the first case of this list is revoked, as a consequence of new textual evidence.4 In the meantime, many scholars of the textual history of the Hebrew Bible are inclined to discard also the reading ‫בחר‬.5 Today, the so-called Tenth commandment 3 4 5

Ibidem, 94–95. Tov, TextualCriticismoftheHebrewBible3, 88 note 140. See e.g. Schenker, “Le Seigneur choisira-t-il le lieu de son nom”; Schorch, “The Samaritan Version of Deuteronomy and the Origin of Deuteronomy,” 31–32; Gallagher, “Is the Samaritan Pentateuch a Sectarian Text,” 97.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

79

of the Samaritan Torah appears to be the last instance most students of the Samaritan Pentateuch are prepared to accept as a true Samaritan ideologism, although this also has already been questioned.6 Most obviously, therefore, the assumption of Samaritan ideology as the motive for textual changes has been turned out erroneous on several occasions in light of new or previously overlooked textual evidence. This demonstrates that a reading which supposedly fits into the Samaritan mindset was not necessarily born out of a Samaritan mind. In light of this caveat, supposed cases of secondary textual “Samaritanisms” should be subjected to the same methods of inquiry as the reminder of the respective texts. Above all, however, the attribution of a certain reading or a whole passage to an alleged Samaritan adstratum inevitably requires proof that this respective case does not belong to the pre-Samaritan substratum, according to the criteria determining the latter. Following this insight, the present article approaches the so-called Tenth (Gerizim-)commandment from the Samaritan Pentateuch from a literary point of view, and applies a comparison with literary techniques found in the pre-Samaritan texts. 2. THE TEXT OF THE GERIZIM-COMMANDMENT The passage in question consists of a considerable plus of 92 or 93 words7 in the Samaritan Pentateuch as opposed to the other known textual witnesses of the Hebrew Bible, following the traditional Tenth commandment ‫“( לא תחמד בית רעך ולא תחמד אשת רעך וג׳‬You shall not covet your neighbor’s house and you shall not covet your neighbor’s wife” etc.), in both the SP version of Exod 20 and Deut 5, and focusing on the publication of “this Torah” and the erection of an altar on Mount Gerizim immediately after entering the Holy Land. In the following presentation of the text, the verses of the additional passage are numbered with letters (a)–(h), for convenience and further reference, in accordance with the verse division of the Masoretic text in their respective original co-texts. The source for each verse is indicated 6

7

See Gallagher, “Is the Samaritan Pentateuch a Sectarian Text,” 101–104; Zahn, “The Samaritan Pentateuch and the scribal culture of Second Temple Judaism,” 301–307. Depending on the question whether ‫הרגריזים‬, which is generally written as one word in the Samaritan tradition, is counted as one or two words. It should be noted, however, that the spelling ‫ הר גריזים‬is not entirely absent from the Samaritan Hebrew tradition either; e.g., it is attested in MS Nablus 6 in Exod 20:17d and Deut 11:29 (but not in the other occurrences of Mount Gerizim in this manuscript).

‫‪80‬‬

‫‪STEFAN SCHORCH‬‬

‫‪on the right margin in square brackets, and in one case, the source for a‬‬ ‫‪contaminated quote is additionally indicated in curly brackets. Passages‬‬ ‫‪of the original that were omitted in the Gerizim composition are printed‬‬ ‫‪in a smaller type on grey background, in order to provide information‬‬ ‫‪about the original context, and words that were added to the original‬‬ ‫‪appear with a dotted underline:‬‬ ‫]‪[Exod 20:17 ∥ Deut 5:21‬‬ ‫לא תחמד בית רעך ולא תחמד אשת רעך שדהו עבדו ואמתו שורו וחמורו וכל אשר לרעך‬ ‫)‪[Deut 11:29] (a‬‬

‫והיה כי יביאך יהוה אלהיך אל ]הארץ[ ארץ הכנעני אשר אתה בא‬ ‫שמה לרשתה‬ ‫]ונתתה את הברכה על הר גריזים ואת הקללה על הר עיבל[‬

‫)‪[Deut 27:2] (b‬‬

‫]והיה ביום אשר תעברו את הירדן אל הארץ אשר יהיה אלהיך נתן לך[‬ ‫והקמת לך אבנים גדלות ושדת אתם בשיד‬

‫)‪[Deut 27:3] (c‬‬ ‫}‪{cf. Deut 27:8‬‬

‫וכתבת ]עליהן[ על האבנים את כל דברי התורה הזאת‬ ‫]בעברך למען אשר תבוא אל הארץ אשר יהוה אלהיך נתן לך‬ ‫ארץ זבת חלב ודבש כאשר דבר יהוה אלהי אבותיך לך[‬

‫)‪[Deut 27:4] (d‬‬

‫והיה בעברכם את הירדן תקימו את האבנים האלה אשר אנכי מצוה‬ ‫אתכם היום בהרגריזים‬ ‫]ושדת אתם בשיד[‬

‫)‪[Deut 27:5] (e‬‬

‫ובנית שם מזבח ליהוה אלהיך מזבח אבנים לא תניף עליהם ברזל‬

‫)‪[Deut 27:6] (f‬‬

‫אבנים שלמות תבנה את מזבח יהוה אלהיך והעלית עליו עלות ליהוה‬ ‫אלהיך‬

‫)‪[Deut 27:7] (g‬‬

‫וזבחת שלמים ואכלת שם ושמחת לפני יהוה אלהיך‬

‫)‪[Deut 11:30] (h‬‬

‫]הלא הם[ ההר ההוא בעבר הירדן אחרי דרך מבוא השמש בארץ‬ ‫הכנעני הישב בערבה‬ ‫מול הגלגל אצל אלון מורא מול שכם ‪:‬‬

‫]‪[Exod 20:18‬‬

‫]‪[Deut 5:22‬‬

‫וכל העם שמע את הקולות ואת קול‬ ‫השופר‬

‫את הדברים האלה דבר יהוה אל כל קהלכם‬ ‫בהר‬

‫וראים את הלפידים ואת ההר עשן‬

‫מתוך האש חשך ענן וערפל קול גדול‬ ‫ולא יסף‬

‫ויראו כל העם וינעו ויעמדו מרחק‬

‫ויכתבם על שני לוחת אבנים ויתנם אלי‬

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

81

As can be well observed from this presentation of the text, the socalled Gerizim-commandment of the Samaritan Pentateuch is inserted into the original text, immediately following the end of the traditional Decalogue in both of its occurrences in Exodus and Deuteronomy. None of the textual blocks used for the composition of this passage is new or original, on the contrary: Except for the omission of some verse parts and some minor modifications in wording, the text is generally composed of verses taken from other parts of the Book of Deuteronomy, forming in fact a florilegium of verses dealing with Mount Gerizim. Most noteworthy, however, the sequence of these verses in the new composition does not follow their original order in Deuteronomy, but the verses are rearranged (see below, 4). The focal point of the text is without any doubt Mount Gerizim and its central importance for Israel, and it seems from the outset that the insertion that followed aimed at connecting this passage with the Decalogue, although this first impression will have to be analyzed more in depth in the following.

3. TEXTUAL HISTORY OF THE GERIZIM COMPOSITION? The Gerizim composition is known, above all, from the manuscripts of the Samaritan Torah, the oldest of which date to the 12th century. Since none of the manuscripts from the Judean Desert preserves any part of this passage, we have very few sources that could provide insights into its internal textual history before the 12th century, a perspective which should be carefully kept apart from the question of the text-historical origin of the Gerizim composition.8 One relevant corpus in this regard comprises Samaritan inscriptions on stone, some of which have been dated as early as the early Byzantine era and contain abridged versions of the Decalogue.9 In an article published jointly in 1951, John Bowman and Shemaryahu Talmon compared a number of these stone inscriptions. Based on their interpretation, they concluded that the inscriptions attest a textual stage in which the text of the so-called Gerizim commandment (and the Samaritan Decalogue as a whole) was still fluid:10 “These facts seem to suggest that the stones 8 9

10

See below, 6. For an overview, see Dexinger, Das Garizimgebot im Dekalog der Samaritaner, 111– 124, and Kartveit, The origin of the Samaritans, 226–228. Bowman / Talmon, Samaritan Decalogue Inscriptions, 224.

82

STEFAN SCHORCH

date from early days and record a version of the Samaritan Decalogue which had not yet been definitely fixed.” Of central importance for this claim is an inscription reportedly found in Nablus in 1856 and brought to the Leeds City Museum in 1862. The inscription reads as follows:11 [‫]השבת לק[דשה ]ו כבד את‬ [‫]אביך[ ואת אמ]ך לא תרצח‬ [‫]לא ת[נאף לא ]תגנב לא‬ [‫]תע[נה ברעך ]עד שקר לא‬ [‫]ת[חמד בית ]רעך ולא‬ [‫]ת[חמד אשת ר]עך ובנית‬ [‫]שם[ מזבח לי]הוה אלהיך‬ [‫וכתבת על האבנ]ים את כל‬ [‫התורה ה]ז[את בא]ר היטב‬ [‫]תורה צו[ה לנו משה ]מורשה קהילת יעקב‬

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

The sequence (from line 6 onwards) ‫( לא תחמד אשת רעך‬i.e. the traditional 10th commandment) ‒ ‫( ובנית שם מזבח ליהוה אלהיך‬Deut 27:5) ‒ ‫( וכתבת על האבנים האלה באר היטב‬Deut 27:8) ‒ ‫תורה צוה לנו משה מורשה‬ ‫( קהילת יעקב‬Deut 33:4) is understood by Bowman and Talmon as an older textual version of the end of the Samaritan Decalogue. This interpretation has been contested by Ferdinand Dexinger, who pointed out, in accordance with the genre characteristics of these inscriptions as mezuzot, that lines 8–10 of the Leeds inscription were not conceived as part of the Decalogue, but as a complementary closing of the inscription as a whole.12 Due to Dexinger’s refutation of the Samaritan mezuza-inscriptions as a reliable source for the reconstruction of the textual history of the Tenth commandment, only one textual witness seems relevant: the Syro-Hexapla, which is the oldest attestation of its text and context. According to a recent detailed examination by Bradley Marsh, Jr., “the Samaritan passages, i.e., the ‘expansions’ such as those recorded by Masius, […] part of the old Eusebian layer – or what might more neutrally be called the ‘basic Caesarean layer’ – of the Greek, hexaplaric εκδοσις Paul [of Tella, first half of the 7th century] translated.”13 In other words, the Hexaplaric tradition of the Gerizim composition leads back to the 11

12 13

Leeds City Museum (LEEDM.D.2008.0005). An excellent image of this inscription can be found on the website of the museum: http://www.leeds.gov.uk/museumsandgalleries/ pages/object.aspx?oid=177020 (accessed 20 October 2017). Dexinger, Das Garizimgebot im Dekalog der Samaritaner, 114–115. Marsh, Early Christian Scripture and the Samaritan Pentateuch I, 10.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

83

3rd century CE. Marsh provides an excellent edition of the Syriac text together with a detailed commentary.14 It shows no deviation from the Samaritan manuscript evidence.

4. LITERARY ANALYSIS

OF THE

TEXT OF THE GERIZIM COMPOSITION

As the above presentation of the text shows, the Gerizim composition combines all in all eight verses (according to the counting of the Masoretic text), taken from Deut 11 and 27. As regards the arrangement of these verses, the passage is built in a ring-like or frame-structure: The verses Deut 11:29–30 frame the sequence Deut 27:2–7, creating a narrative continuum between these two texts, which in the book of Deuteronomy are separated by 16 chapters, although they are clearly paralleled to each other, in terms of the literary structure of the book.15 As regards the content of the new composition, the scribes entered into Deut 11:29–30 an information that is lacking in this passage as compared with Deut 27, but at the same time provided a another detail to the account in Deut 27, which is found only in Deut 11, as follows: In Deut 11:29–30, the narrative proceeds straight from the (future) crossing of the Jordan in verse 29a to the blessing and cursing of the covenant ceremony in verse 29b, followed by a short geographical localization of the two mountains, on which this ceremony takes place (verse 30). According to Deut 27, after crossing the Jordan and before the covenant ceremony, a cultic site is founded and an altar is erected on Mount Gerizim. In opposition to Deut 11:29–30, however, Deut 27 itself lacks any explicit indication of the mountain’s location. Thus, the conflation16 of Deut 11:29– 30 and Deut 27:2–7 produces a new text that contains information not present in either of the two original texts alone. As to the literary technique of this conflation, the latter was carried out through inserting the later text – in terms of the original sequence of the Book of Deuteronomy – into the former, i.e. Deut 27:2–7 was entered into Deut 11:29–30, and not the opposite, as can be seen from the fact that Deut 11:29 and Deut 11:30 form a frame around Deut 27:2–7. Thus, the 14

15 16

Edition: Marsh, Early Christian Scripture and the Samaritan Pentateuch II, 6–7; commentary: Marsh, Early Christian Scripture and the Samaritan Pentateuch I, 97–107. See Schorch, The Samaritan Version of Deuteronomy, 26–28. A definition of this term with regard to Biblical texts is provided by Salzer, Die Magie der Anspielung, 35, and her book provides a large number of examples taken from medieval Hebrew texts.

84

STEFAN SCHORCH

Gerizim composition was created through a retroverted insertion of one textual passage into another, resulting in a literary structure in which the former passage of the original text becomes the frame for the later. As already mentioned before, the single verses of the Gerizim composition appear generally in accordance with the wording in their original contexts, but some minor changes have been carried out in order to accommodate the function of the new composition as well as its textual coherence, and, most importantly, some parts of the quoted material have been removed entirely: – Of the first verse (Deut 11:29), the Gerizim composition uses only the first half (= a), while the second half – about the ceremony of blessing and cursing on Mounts Gerizim and Ebal – has been removed. As a result, the new text contains no reference to the covenant ceremony. This seems to demonstrate that, from the outset, the focus of this composition was the cultic site and altar on Mount Gerizim rather than the ritual landscape of the covenant ceremony comprising Mount Gerizim and Mount Ebal. – From the second conflated verse (Deut 27:2), the Gerizim composition removes the first half, which would have repeated information already found in the preceding verse Deut 11:29a (= a); the same is true for everything, except for the first sentence of Deut 27:3 (= c), from which even a phrase of such prominent status as ‫ ארץ זבת חלב ודבש‬was left out. The last sentence of Deut 27:4, ‫ושדת אתם בשיד‬, repeats what has already been said in Deut 27:2b (= b) and is likewise removed. All these changes seem to reveal further characteristics of the scribal work leading to the Gerizim composition: As regards the narrative surrounding the foundation of the cultic site and the altar, the aim was brevity and focus on the basic structure of the narrative, rather than repetition or insistence on ornamental details. – In light of this observation, the passages that remained intact in the course of their re-application in the Gerizim composition deserve special interest. This is the case with the first part of Deut 27:4a (= d), which indeed repeats details already found in a–b, namely the future crossing of the Jordan and erecting of the stones. However, Deut 27:4a is the only verse providing the exact location of the cultic site and altar, which is the point of the present composition, and it therefore appears here as one of its core elements. Moreover, the three subsequent verses, Deut 27:5–7, which are devoted to the erecting of the cultic site and the altar itself, entered the new composition untouched (= e–g) and thus provide further information from the perspective of literary history.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

85

– The removal of the first two words ‫ הלא הם‬from the verse that closes the frame, Deut 11:30 (= h), obviously follows the requirements of the new context. Its replacement with the phrase ‫ ההר ההוא‬creates a syntactically and semantically smooth transition from Deut 27:4–7 (= d–g) to Deut 11:30 (= h), but it also achieves a change in the meaning of the latter: While the verse in its original context, i.e. ch. 11, refers to the cultic landscape of the covenant ceremony, it provides a description of the position of Mount Gerizim in the context of the Gerizim composition. – This leads to a further important literary characteristic of the Gerizim composition: Changes in wording as in Deut 11:30 produce in fact two different versions of the same textual passage. E.g., verse h of the Gerizim composition is different from Deut 11:30, although the two are not only immediately related in terms of literary history, but also become literary parallels due to the insertion of the Gerizim composition into the Book of Deuteronomy. The same phenomenon can be observed in the other changes carried out in the text of the Gerizim composition, but the change from ‫ הארץ‬to ‫ ארץ הכנעני‬in Deut 11:29 (= a) seems especially interesting with regard to the literary motives that may have caused these changes: While the above-mentioned change produces a disharmonic version of the verse itself, it also creates a link within the frame of the new composition, because as a result of that change, verses a and h conform in the designation of the land. Most obviously, therefore, the composition of the new text was more led by the aim to create coherence within this text itself, and much less by the constraints of literary harmonization within the horizons of the Book of Deuteronomy. In other words, the Gerizim composition is conceived above all as a separate textual unit, and not primarily as an expansion of the receiving textual continuum.

5. THE CONCEPTUAL CONTEXT OF THE GERIZIM COMPOSITION Since the Gerizim composition is attested twice in the Samaritan Pentateuch, in Exodus 20 and in Deuteronomy 5, the question arises of how this double tradition emerged, in terms of literary history, as well as in terms of the conceptualization of the Torah as whole: Was the Gerizim composition first inserted into the text of Exodus and then via harmonization into the text of Deuteronomy, or vice versa, did the Gerizim composition originate in Deuteronomy, or was the composition rather inserted into the text of both books at the same time?

86

STEFAN SCHORCH

Although a conclusive solution to this problem can hardly be provided on the basis of the evidence presently at hand, it seems plausible to assume that the creation of the Gerizim composition and its insertion into the text of the Pentateuch responded primarily to a textual problem emerging from the Book of Exodus, and not from Deuteronomy: The election of the chosen place, Mount Gerizim, according to SP, is found only in Deuteronomy, while it is absent from the revelation on Mount Sinai as told in Exodus through Numbers. In this case as in numerous others, it appears to be just in line with the textual characteristics exhibited in the pre-Samaritan texts as well as in the Samaritan Pentateuch to supplement the text of Exodus with passages taken from Deuteronomy. As is well known, the larger editorial changes of the pre-Samaritan/Samaritan text of the Pentateuch generally have been carried out with the help of passages taken from Deuteronomy (mainly from the first three chapters of Deuteronomy) inserted into Exodus or Numbers, with only one exception (Deut 2:7, expanded with material from Num 20:14.17f). Thus, at least conceptually, the creation and insertion of the Gerizim composition seems to have been primarily aiming at the text of the Book of Exodus, in order to align it with Deuteronomy as regards the main narrative content. Moreover, the Book of Exodus posed a further textual challenge to the ways in which the two accounts of the Sinai revelation in Exodus– Leviticus–Numbers vis-à-vis Deuteronomy were apparently read in parallel within the scribal circles that produced the (pre-)Samaritan version of the Pentateuch, namely the altar law of Exod 20:24–26, which does not provide any specific location for the altar and apparently originally referred to a potential multitude of altars.17 The insertion of the Gerizim composition before this altar law made sure that the latter could be read in light of the former. The insertion of the Gerizim composition into the text of Deuteronomy, on the other hand, seems to have followed the logic of textual harmonization: The Gerizim composition was inserted at the end of the Ten commandments and obviously became part of that text, and though the two texts of the Ten commandments in the Samaritan Pentateuch contain some textual differences, especially in the Sabbath commandment,18 a major difference like the absence versus presence of the Gerizim composition would 17

18

The textual difference between ‫( בכל המקום אשר אזכיר את שמי‬MT) and ‫במקום אשר‬ ‫( אזכרתי את שמי‬SP) may reflect the scribal endeavor to change the text in accordance with the notion that only one legitimate altar exists, see Knoppers, Jews and Samaritans, 209. See Zahn, The Samaritan Pentateuch, 300–301.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

87

have been singular in that context. Thus, with the Gerizim composition becoming part of the Ten commandments in one of the two versions, the other one had to be harmonized.

6. THE GERIZIM COMPOSITION AND OTHER LARGE EDITORIAL CHANGES IN SP In order to understand the literary techniques by means of which the Gerizim composition became part of the Book of Exodus, a literary analysis of the broader context of this passage in the pre-Samaritan/Samaritan version of Exod 20 seems required, especially in light of the observation that this chapter as a whole was reworked within the pre-Samaritan/Samaritan tradition. As is well known, the Gerizim composition, inserted in SP between Exod 20:17 and 18 (= insertion I), is only the first of three of large-scale insertions in the SP version of Exod 20. In fact, only one verse, Exod 20:18, although in a slightly changed wording, separates the Gerizim composition from the following insertion, a four-verse-long passage adopted from Deut 5:24–27 (= insertion II):19 The LORD our God has just shown us His majestic Presence, etc. […] You go closer and hear all that the LORD our God says, and then you tell us everything that the LORD our God tells you, and we will willingly do it.

In this passage, the Israelites express the huge and terrifying impression that the first revelation on Mount Sinai made on them and ask Moses to act as their intermediary with God. Insertion II follows the sequence of Deut 5, where this passage (i.e. Deut 5:24–27) also comes after the Ten commandments (Deut 5:6–21). However, though the scribes did rework Exod 20 here in accordance with Deut 5 through the insertion of these verses, they do not seem to have aimed at harmonizing the two accounts. This can be seen from the fact (among others) that the account provided in Deut 5:22 – according to which God wrote his words on two tablets of stone – is lacking in both SP and the known manuscripts affiliated with the pre-Samaritan type. Thus, Moses’ words ‫ויכתבם על שני לחת אבנים ויתנם אלי‬ have no counterpart in either version of Ex 20. In addition, both concerns of Deut 5:24–27, namely the people’s fear of a direct encounter with God as well as their request to Moses that he should act as an intermediary on 19

English translations of the Biblical text generally follow JPS Tanakh (Philadelphia 1985).

88

STEFAN SCHORCH

their behalf, seem to have been already present in the unrevised version of Exod 20:19; therefore, the insertion of this passage is rather to be described as a replacement for the short version of this speech as found in the unrevised text of Exod 20:19, with the longer version contained in Deut 5:24–27. The purpose of this textual change was most probably to provide a more detailed and explicit narrative in preparation for the next larger insertion, which is 9 verses long and focuses on the assignment of Moses by God as his prophetic interlocutor. The opening section of this composition is as follows: I have heard the plea that this people made to you; they did well to speak thus. (Deut 5:25)

In the short version of the people’s speech, i.e. in the unrevised version of Exod 20:19 as preserved in MT, the plea referred to here would simply have said ‫“ – דבר אתה עמנו ונשמעה ועל ידבר עמנו אלהים פן נמות‬You speak to us and we will obey; but let not God speak to us, lest we die,” and the longer version, therefore, might seem to clarify the intermediary role expected of Moses. Nevertheless, there is hardly anything contained in the longer version of the speech that could not have already been understood in the shorter original as regards the narrative preparation of continuation, including the following insertion of Deut 5:28f - 18:18–20 5:30f (= insertion III). This observation seems relevant insofar as it shows that the insertion of Deut 5:24–27 is not immediately nor necessarily linked with the insertion of Deut 5:28f - 18:18–20 - 5:30f. Most importantly, the former is neither a requirement for nor a consequence of the latter. Therefore, insertions II and III should be regarded as independent from the other in terms of literary history and function. As for insertion III, this passage is composed in a ring-like structure, with closely resembles the structure of the Gerizim composition: ‫]ויאמר[ וידבר יהיה]אלי[אל משה לאמר‬ ‫שמעתי את קול דברי העם הזה אשר דברו אליך היטיבו כל אשר דברו‬

[Deut 5:28] (a)

‫מי יתן והיה ]לבבכם[ לבבם זה להם ליראה אתי ולשמר את מצותי‬ ‫כל הימים‬ ‫למען ייטב להם ולבניהם לעולם‬

[Deut 5:29] (b)

‫[ נביא אקים להם מקרב אחיהם כמוך‬Deut 18:18] (c) ‫ונתתי דברי בפיו ודבר אליהם את כל אשר אצונו‬ ‫[ והיה האיש אשר לא ישמע אל דבריו אשר ידבר בשמי‬Deut 18:19] (d) ‫אנכי אדרש מעמו‬ ‫[ אך הנביא אשר יזיד לדבר דבר בשמי את אשר לא צויתיו לדבר‬Deut 18:20] (e) ‫ואשר ידבר בשם אלהים אחרים‬ ‫ומת הנביא ההוא‬

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

89

‫[ וכי תאמר בלבבך‬Deut 18:21] (f) ‫איך נודע את הדבר אשר לא דברו יהוה‬ ‫[ אשר ידבר הנביא בשם יהוה לא יהיה הדבר ולא יבוא הוא הדבר‬Deut 18:22] (g) ‫אשר לא דברו יהוה‬ ‫בזידון דברו הנביא לא תגור ממנו‬ ‫לך אמר להם שובו לכם לאהליכם‬

[Deut 5:30] (h)

‫ואתה פה עמד עמדי ואדברה אליך את כל המצוה החקים‬ ‫והמשפטים אשר תלמדם‬ ‫ועשו בארץ אשר אנכי נתן להם לרשתה‬

[Deut 5:31] (i)

As can be seen in this chart, Deut 5:28–29 and 30–31 form a frame around Deut 18:18–22 in order to achieve – as in the Gerizim composition – the conflation of the two texts in order to overcome the absence of certain details in either of the two earlier texts. The connection of the two texts proceeds from the observation that Deut 18:16–17 explicitly contextualizes the following verses 18–22 in the same event as the one already referred to in Deut 5:24–31: This is just what you asked of the LORD your God at Horeb, on the day of the Assembly, saying, Let me not hear the voice of the LORD my God any longer or see this wondrous fire any more, lest I die. Whereupon the LORD said to me, They have done well in speaking thus. (Deut 18:16–17)

A parallel reading of the two accounts makes it obvious, however, that Deut 5 contains no equivalent to 18:18–22, i.e. the announcement of a prophet like Moses. On the other hand, the original version of Deut 18 seems to speak about the future prophet only, not about Moses himself and his task, as Deut 5 does. The insertion of the verses from Deut 18 into Deut 5 follows therefore not only the literary purpose of creating a new harmonistic version of the two accounts, but it is a hermeneutical means comparable to the Gerizim composition: It creates an explicit intertext of the two accounts, illuminating each of the two earlier with the help of the other. Like the Gerizim composition, the present composition was created through a retroverted insertion of the later text into the former, according to the order of the Book of Deuteronomy in its present form. And as in the case of the Gerizim composition, the new text did not replace one or both of its predecessors, but was conceived from the outset as an independent textual composition, as can be seen from the fact that the explicit version of the intertext was not inserted into the Book of Deuteronomy, but only into the Book of Exodus, at least in known textual witnesses. The insertion

90

STEFAN SCHORCH

of the Moses composition in Exodus, on the other hand, solves again a problem: Without this insertion, the text of Exodus does not contain any designation or appointment of Moses as prophet. Thus, most obviously, the composition of this explicit intertext must have been motivated by the Book of Exodus and, more specifically, by the beginning of the passage on the revelation at Sinai. In light of the numerous parallels between the Gerizim and Moses compositions in terms of literary technique, literary history, and hermeneutics, it would seem rather difficult to suggest that they emerged independently from the other. Instead, it is much more plausible to assume that both compositions originate in the same scribal circles. The fact that the Moses composition is found not only in the Samaritan Pentateuch but also in the pre-Samaritan scrolls 4QRPa (4Q158) and 4QTest (4Q175),20 and apparently also in the pre-Samaritan “Biblical” scroll 4QpaleoExodm (4Q22),21 shows that it did not originate in the Samaritan community, but belongs to the pre-Samaritan substrate of the preSamaritan/Samaritan type. Therefore, the above conclusion that the Gerizim and Moses compositions originated in the same context implies that the former belongs much more plausibly to a pre-Samaritan than Samaritan layer of the Samaritan Pentateuch, in terms of literary history.

7. THE INSERTION OF THE GERIZIM COMPOSITION INTO THE TEN COMMANDMENTS The attribution of the Gerizim composition to the pre-Samaritan substratum of the Samaritan Pentateuch leads to two further problems: The first is of text-critical nature and relates to the attestation or absence of the Gerizim composition in pre-Samaritan witnesses. The second problem is literary and relates to the fact that the verses conflated in the Gerizim composition contain words of Moses, while the composition in its new context may seem to have been inserted into a passage presenting words of God.

20 21

See Tov, The Samaritan Pentateuch and the Dead Sea Scrolls, 67. See the conclusion reached by Ulrich and Sanderson in their edition of the text, DJD IX, 102, although in this case, the Moses composition (= insertion III of SP) itself is not preserved, but only the insertion from Deut 5:24–27 (= insertion II), followed by a lacuna which would match the extent of insertion III.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

91

7.1. The text-critical question While none of the known fragments found in the Judean Desert preserves the Gerizim composition itself or parts of it, at least two preSamaritan manuscripts – 4QpaleoExodm (4Q22) and 4QRPa (4Q158) – seem to provide some evidence that they originally contained the Gerizim composition.22 The relevant passage in 4QpaleoExodm is found in columns XX–XXII of Ulrich’s and Sanderson’s edition in DJD IX and the text, as far as it is preserved in the fragments, comprises Exod 20:1 [lacuna A] 20:19 [lacuna B] 21:5–6, in three subsequent columns. Proceeding from the Samaritan Pentateuch, we would expect to localize the Gerizim composition in lacuna A, which leads to the question of the exact length of this column. According to Ulrich’s and Sanderson’s estimate, the size of this lacuna provides not enough room for the Gerizim composition,23 but this conclusion does not seem entirely convincing: Lacuna A, in their own reconstruction of the scroll, comprises 22 lines, which could have accommodated c. 1080 letters, if we take the average of letters per line as found in the scroll as point of departure, which is 49.24 The space required for Exod 20:2–19a25 is 1078 letters, if we count according to the Samaritan Pentateuch.26 Moreover, the calculation of the length of lacuna A depends on the place of the small fragments preserving remnants of column XXI within the column as a whole. As pointed out by Ulrich and Sanderson, this place is to be determined with the help of the last 3 lines of column XXII, which are preserved and contain Exod 21:5–6. Therefore, lacuna B should be expected to have originally contained Exod 20:19b–21:4. Within the Samaritan Pentateuch, this passage comprises 1398 letters, equaling c. 28 lines, which would mean that this whole lacking passage was originally contained in column XXII, while Exod 20:19a ended at the bottom of column XXI, implying that lacuna A seems to have been 5 lines 22

23 24 25

26

For a recent overview over the pre-Samaritan scrolls and their research from a textcritical point of view see Tov, Textual criticism, 90–93, Tov, The Samaritan Pentateuch and the Dead Sea Scrolls, and Segal, The Text of the Hebrew Bible in light of the Dead Sea Scrolls, 10–17. Skehan / Ulrich / Sanderson, DJD IX, 102. Ibidem, 57. The first word in col. XXI that seems to provide a clear indication of its source, is ‫הראנו‬ (Exod 20:19a/Deut 5:24). The text follows Tal / Florentin, The Pentateuch, which contains a transcription of MS Nablus 6 (copied 1204).

92

STEFAN SCHORCH

longer than suggested by Ulrich and Anderson. The result of this reestimate is therefore opposite to Ulrich’s and Anderson’s conclusion: There seems to be no reason to doubt that the lacuna in 4QExodm between Exod 20:1 (column XX) and Exod 20:19a (column XXI) is big enough for the Ten commandments and the Gerizim composition, as contained in the Samaritan Pentateuch. This observation is of course no proof that the scroll did indeed contain the Gerizim composition. On the other hand, however, the Gerizim composition does seem by far the most likely candidate with regard to textual passages that might have been originally present in this destroyed part of the scroll. A further text-critical problem is posed by 4QRPa: As can be seen from the chart provided by Michael Segal27 and the improved re-edition of the text provided by Molly Zahn,28 this scroll contains a thoroughly reworked version of Exod 20, apparently proceeding from a pre-Samaritan version of this passage, as suggested by the fact that the text contains many of the additions from Deuteronomy that are known to have been part of the pre-Samaritan/Samaritan texts: (fragment 6:) Deut 5:24–27 Exod 20:19b– 21 Deut 5:28–29 18:18–22 (fragment 7:) Decalogue Deut 5:30–31 Exod 20:22–26.29 In this light, the absence of the Gerizim composition after the Decalogue seems notable and has been cautiously interpreted as support for the hypothesis that the pre-Samaritan stage did not contain the Gerizim composition.30 However, the evidence found in 4Q158 appears rather difficult in this regard and complicates any clear conclusion, since most obviously the whole passage has undergone a comprehensive process of reworking, including above all an intervention within the Moses composition, namely the separation of Deut 18:22 and 5:30 through the insertion of the Decalogue, the relocation of the Decalogue, in comparison to the original narrative continuum, and possibly even the partitioning of the Decalogue into different sections, as suggested by Michael Segal himself.31 The usability of 4Q158 as a witness for the original pre-Samaritan arrangement of the different parts of the text is therefore rather questionable, and it seems indeed possible that the scribe of the scroll inserted the Gerizim composition at a different place as part of his rewriting of the revelation at Mount Sinai. Furthermore, the literary links between the traditional Tenth commandment (‫ )לא תחמד בית רעך ולא תחמד אשת רעך‬and the Gerizim 27 28 29 30 31

Segal, Biblical Exegesis in 4Q158, 56. Zahn, Rethinking rewritten scripture, 245–258. Segal, Biblical Exegesis in 4Q158, 56. Segal, Biblical Exegesis in 4Q158, 56 n. 31. Segal, Biblical Exegesis in 4Q158, 57–58.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

93

composition seem rather weak, and the early textual text-historical evidence attests no sign that the Ten commandments and the Gerizim composition formed one unit (see above, 3). This literary as well as text-historical observation is even supported by the Samaritan manuscripts of the Torah, which place, following an entirely coherent and uniform scribal tradition, a qiṣṣa-sign between these two passages, marking them as separate units of the text. The analysis of the text-critical evidence demonstrates that the presence of the Gerizim composition in the pre-Samaritan does not seem to be contradicted by any of the scrolls from the Judean Desert; on the contrary, it is supported by the evidence emerging from 4QpaleoExodm. 7.2. Inserting the Gerizim composition – the literary aspects In spite of the obvious literary parallels between the Gerizim and Moses compositions, as shown above (4.), at least one clear difference exists between them: While the verses forming the Moses composition were spoken by God, according to both their original contexts in Deut 5 and 11 and their text itself, the Gerizim composition quite obviously contains Moses’ words. The problem in this regard is not that words of Moses were copied to another place, since that happened in others cases as well,32 but that Moses’ words were inserted in a context that may imply God as the speaker.33 The latter reading of the text, however, is based on two presuppositions that are rather controversial, namely that the entire Decalogue was announced by God himself, and that the Gerizim composition was conceived from the outset as a part of the Decalogue itself. As to the first point, only the First of the Ten commandments (according to the traditional Samaritan counting), uses indeed the divine first person: You shall have no other gods besides Me. You shall not make for yourself a sculptured image, or any likeness of what is in the heavens above, or on the earth below, or in the waters under the earth. You shall not bow down to them or serve them. For I the LORD your God am an impassioned God, visiting the guilt of the parents upon the children, upon the third and upon the fourth generations of those who reject Me, but showing kindness to the thousandth generation of those who love Me and keep My commandments. (Exod. 20:3-6) 32 33

See the chart in Kartveit, The origin of the Samaritans, 311–312. See Zahn, The Samaritan Pentateuch and the scribal culture of Second Temple Judaism, 304.

94

STEFAN SCHORCH

Beginning with the Second commandment (Samaritan counting), God is referred to in the third person, where applicable: You shall not swear falsely by the name of the LORD your God; for the LORD will not clear one who swears falsely by His name. etc. (Exod. 20:7)

The literary difference between these two parts is obvious and was already observed by early Jewish Midrashim preserved in Pesikta de-Rab Kahane and the Babylonian Talmud, which solved the problem by suggesting that only the beginning of the Decalogue reflects God’s own speech, while the reminder is attributed to Moses and refers to God in the third person.34 From that perspective, the Gerizim composition, inevitably continuing the mode of speaking introduced already with the Second commandment (Exod 20:7), is more plausibly understood as being inserted in the context of Moses’ words rather than God’s. Another problem with a strong impact on the perception of the Gerizim composition in its literary context is the question whether the Gerizim composition in the pre-Samaritan/Samaritan tradition was indeed conceived as part of the Ten commandments from the outset. It is of course true that that this was indeed the case in the medieval Samaritan tradition, as suggested by marginal peritexts found in some Samaritan manuscripts of the Torah as early as the 13th century.35 However, the Samaritan scribal transmission of the Ten commandments provides very few indications in the text itself that could possibly be used in support of this assumed literary conjunction.36 Moreover, and most notably, from a text-historical point of view, the Syro-Hexapla as the oldest textual witness of the text of the Gerizim composition and insertion after Exod 20:17 contains no hint that this addition could have been understood as part of the Ten commandments, as pointed out by Marsh:37 [N]o evidence from S[yro-]H[exapla] Exod (or J[acob of Edessa’s revision of]Exod) suggests that Exod 20.17(13) ‫ד‬-‫ ו‬was understood to be a part of the Samaritan Decalogue. One would certainly expect the special Samaritan reckoning to have been explained in the attribution annotation, or some other scholion attached to the readings, if this was understood. 34 35

36

37

See Segal, Biblical Exegesis in 4Q158, 57. The earliest manuscript known to me that contains a Yud for “10” in the margins of the passage in question is MS Jerusalem, National Library of Israel Sam. 2° 6, copied in the year 1215. In the medieval manuscripts of the Samaritan Torah, the whole passage is generally subdivided into four paragraphs: Exod 20:1–7; 8–11; 12–17; 17a–17f. Marsh, Early Christian Scripture and the Samaritan Pentateuch I, 99.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

95

The fact that the Syro-Hexapla does otherwise provide numbers for the single commandments makes this absence conspicuous:38 Insofar as SH Exod is concerned, each command is assiduously numbered (1-4 in-text; 1-9 marginally) save the tenth; there the scribe of SHL opted instead to use the margin to record the readings of the Three at v 17.

Thus, the available text-historical evidence at least calls into question the perception of the Gerizim composition as part of the Ten commandments for the 3rd century CE. From a literary point of view, it should be noted that the Gerizim composition is clearly set apart from the preceding commandments, each of which is introduced by an imperative, either in the positive form of a command, or in the negative form of a prohibition. The Gerizim composition, on the other hand, commences with a temporal clause: When the LORD your God brings you into the land that you are about to enter and possess […] (Deut 11:29)

Only one part of the whole passage deviates from the imperative sequence dominating the Decalogue in a comparable way, and this is the introductory sentence: I the LORD am your God who brought you out of the land of Egypt, the house of bondage. (Exod. 20:2)

Whether this sentence is part of the First commandment or simply an introduction does not seem to be entirely obvious. The medieval Samaritan tradition insists that the first sentence is not part of the commandments. Following this perspective, a possible correspondence emerges between the first sentence and the Gerizim composition, which might have been an attempt to create (and understood in antiquity as) a narrative frame around the Ten commandments, relating them to the foundational narrative of the people of Israel, from the Exodus to the possession of the promised land.

CONCLUSION The Gerizim composition found in the Samaritan text of the Ten commandments in Exod 20 and Deut 5, according to medieval manuscripts, presents a deliberate scribal attempt to stress some main points 38

Ibidem, 98–99.

96

STEFAN SCHORCH

of Pentateuchal hermeneutics, completely in line with the hermeneutics and the literary techniques attested in the pre-Samaritan textual layer. Nothing in the Gerizim composition itself, or in its insertion at these points of the Pentateuch narrative, is specifically “Samaritan.” Rather, the authors of this insertion seem to have been part of the scribal culture shared by followers of both the Jerusalem and Gerizim sanctuary.

BIBLIOGRAPHY BOWMAN, John / Shemaryahu TALMON. Samaritan Decalogue Inscriptions. Bulletin of the John Rylands Library 33 (1951), 211–236. DEXINGER, Ferdinand. Das Garizimgebot im Dekalog der Samaritaner. In: Georg Braulik (ed.): Studien zum Pentateuch: Walter Kornfeld zum 60. Geburtstag. Wien: Herder, 1977. 111–134. SKEHAN, Patrick W. / Eugen ULRICH / Judith E. SANDERSON. Qumran Cave 4. IV: Paleo-Hebrew and Greek Biblical Manuscripts (Discoveries in the Judean Desert; IX). Oxford: Clarendon Press, 1992. GALLAGHER, Edmon L. Is the Samaritan Pentateuch a Sectarian Text? ZAW 127 (2015), 96–107. KARTVEIT, Magnar. The origin of the Samaritans. Leiden / Boston: Brill, 2009 (VT. Suppl.; 128). KNOPPERS, Gary N. Jews and Samaritans: The origins and history of their early relations. New York, NY: Oxford University Press, 2013. MARSH, Bradley John, Jr. Early Christian Scripture and the Samaritan Pentateuch: A study in Hexaplaric Manuscript Activity. A thesis submitted to the Faculty of Theology and Religion in fulfilment towards the degree of Doctor of Philosophy. Oxford: University of Oxford, 2016 (2 volumes). SALZER, Dorothea M. Die Magie der Anspielung: Form und Funktion der biblischen Anspielungen in den magischen Texten der Kairoer Geniza. Tübingen: Mohr Siebeck, 2012 (Texts and studies in ancient Judaism; 134). SCHENKER, Adrian. Le Seigneur choisira-t-il le lieu de son nom ou l’a-t-il choisi?: l’apport de la Bible grecque ancienne à l’histoire du texte samaritain et massorétique. In: Anssi Voitila (ed.): Scripture in transition: Essays on Septuagint, Hebrew Bible, and Dead Sea Scrolls in Honour of Raija Sollamo. Leiden / Boston: Brill, 2008 (JSJ. Suppl.; 126), 339–351. SCHORCH, Stefan. The Samaritan Version of Deuteronomy and the Origin of Deuteronomy. In: József Zsengellér (ed.): Samaria, Samarians, Samaritans. Studies on Bible, history and linguistics; [papers … presented at the sixth international conference of the Société d’Études Samaritaines held at … Pápa, Hungary in July 17 - 25, 2008]. Berlin: de Gruyter, 2011 (Studia Samaritana; 6), 23–37. SEGAL, Michael. Biblical Exegesis in 4Q158: Techniques and Genre. Textus 19 (1998), 45–62. —. The Text of the Hebrew Bible in light of the Dead Sea Scrolls. Materia giudaica 12 (2007), 5–20.

SO-CALLED GERIZIM COMMANDMENT IN THE SAMARITAN PENTATEUCH

97

TAL, Abraham / Moshe FLORENTIN (ed.). The Pentateuch: The Samaritan version and the Masoretic version. Tel Aviv: The Haim Rubin Tel Aviv University Press, 2010. TOV, Emanuel. Textual Criticism of the Hebrew Bible. Minneapolis: Fortress Press / Maastricht: Royal Van Gorcum, 22001. —. Textual Criticism of the Hebrew Bible. Minneapolis: Fortress Press. Third, fully revised edition, 2012. —. The Samaritan Pentateuch and the Dead Sea Scrolls: The Proximity of the Pre-Samaritan Qumran Scrolls to the SP. In: Shani Tzoref / Ian Young (eds.): Keter Shem Tov: Essays on the Dead Sea Scrolls in Memory of Alan Crown. Piscataway, NJ: Gorgias Press, 2013. 59–88. ZAHN, Molly Marie. Rethinking rewritten scripture. Composition and exegesis in the 4QReworked Pentateuch manuscripts. Leiden / Boston: Brill, 2011 (Studies on the texts of the desert of Judah; 95). —. “The Samaritan Pentateuch and the scribal culture of Second Temple Judaism.” JSJ 46 (2015), 285–313.

ALTARED STATES: THE ALTAR LAWS IN THE SAMARITAN AND JEWISH PENTATEUCHS, AND THEIR EARLY INTERPRETERS Gary N. KNOPPERS

This essay has four parts. The first part reviews the multiple and discrepant major altar laws in the Jewish and Samaritan Pentateuchs, particularly those in the Covenant Code and Deuteronomy, to demonstrate intertextual relationships among three of them. The second and third parts explore how the commonalities and differences among these three sets of statutes were appropriated in particular ways by the Judean and Samarian communities in favor of their own sanctuaries.1 In so doing, the study concentrates attention on the texts of Exodus and Deuteronomy, as those texts were likely shared by Samarians and Judeans in the late Persian and Hellenistic eras prior to the introduction of sectarian additions and small changes in the late second and first centuries BCE.2 1

2

Terminologically, I am referring to the residents of Yehud (Judah) and Samerina (Samaria) during the Neo-Babylonian, Persian, and Hellenistic periods as Judeans and Samarians to distinguish them from the later Jews and Samaritans of the Roman period. In both cases, one can discern lines of continuity from one period to the next. See further G. N. Knoppers, JewsandSamaritans:TheOriginsandHistoryofTheirEarlyRelations (New York: Oxford University Press, 2013). This procedure requires some speculation, but fortunately the textual variants to the major witnesses provide important clues to the growth and expansion of older texts. See F. M. Cross, “The History of the Biblical Text in the Light of the Discoveries in the Judaean Desert,” HTR 57 (1964) 281–99; idem, “Aspects of Samaritan and Jewish History in Late Persian and Hellenistic Times,” HTR 59 (1966) 201–11; J. D. Purvis, TheSamaritanPentateuchandtheOriginsoftheSamaritanSect (HSM 2; Cambridge: Harvard University Press, 1968); J. E. Sanderson, AnExodusScrollfromQumran (HSS 30; Atlanta: Scholars Press, 1986); M. Baillet, “Les divers états du Pentateuque Samaritain,” RevQ 13 (1988) 531–45; E. Tov, “The Proto-Samaritan Texts and the Samaritan Pentateuch,” in TheSamaritans (ed. A. D. Crown; Tübingen: Mohr Siebeck, 1989) 397–407; idem, TextualCriticismoftheHebrewBible (3rd rev. ed.; Minneapolis: Fortress, 2012) 74–93; S. Schorch, “Die Bedeutung der samaritanischen mündlichen Tradition für die Textgeschichte des Pentateuch (II),” MitteilungenundBeiträgederForschungsstelle Judentum,TheologischeFakultätLeipzig 12/13 (1997) 53–64; idem, DieVokaledes Gesetzes:DiesamaritanischeLesetraditionalsTextzeuginderTora,1 (BZAW 339; Berlin: de Gruyter, 2004); idem, “La formation de la communauté samaritaine au 2e siècle avant J.-Chr. et la culture de lecture du Judaïsme,” in Uncarrefourdans

100

GARY N. KNOPPERS

The fourth part of this study assesses one important contemporary theory about the relationship between the altar laws of Exodus and Deuteronomy and the historical development of cultic practices in ancient Israel and Judah. Some scholars have perceived a progression from altars of unfinished stones, as presented both in the Covenant Code and in the Mt. Gerizim/Mt. Ebal legislation to altars constructed with fine ashlar masonry, as purportedly found in Deuteronomy’s centralization legislation (11:31–12:31).3 In this theory, the style of altar construction in Deuteronomy’s mandate for cultic unity and cultic purity (11:31–12:31) typifies a later or urban setting in contrast with the purportedly early or rural historical setting imagined for the Covenant Code and Deuteronomy’s Mt. Gerizim/Mt. Ebal statutes (Deut 27:2–8).4 The alterations in legislative mandates for different types of stone altar construction are thought to provide, therefore, more than incidental or arcane details about the sacrificial altars they (re)configure. Rather, the newer laws are drafted with knowledge of various types of sacrificial installations and purportedly

3

4

l’histoiredelaBible(ed. I. Himbaza and A. Schenker; OBO 233; Göttingen: Vandenhoeck & Ruprecht, 2007), 5–20; R. Pummer, “The Samaritans and their Pentateuch,” in ThePentateuchasTorah (ed. G. N. Knoppers and B. M. Levinson; Winona Lake, IN: Eisenbrauns, 2009) 237–69; idem, TheSamaritans:AProfile (Grand Rapids: Eerdmans, 2015) 195–207; J. Zsengellér, “Origin or Originality of the Torah? The Historical and Textcritical Value of the Samaritan Pentateuch,” in FromQumrantoAleppo:ADiscussion with Emanuel Tov about the Textual History of Jewish Scriptures in Honor of his 65th Birthday (ed. A. Lange, M. Weigold, and J. Zsengellér; FRLANT 230; Göttingen: Vandenhoeck & Ruprecht, 2009), 189–202; G. N. Knoppers, “Parallel Torahs and Inner-Scriptural Interpretation: The Jewish and Samaritan Pentateuchs in Historical Perspective,” in ThePentateuch:InternationalPerspectivesonCurrentResearch (ed. T. B. Dozeman, K. Schmid, and B. J. Schwartz; FAT 78; Tübingen: Mohr Siebeck, 2011) 507–31. In the SP, Deut 11:31–32 introduce the material mandating centralization that follows (the qiṣṣa appears at the end v. 30). The SP reading tradition thus interprets the initial conjunction kî in v. 31 as temporal: “When you cross the Jordan to enter to possess the land . . . be careful to practice all of the statutes and judgments, which I, I am delivering before you this day.” Deut 11:31–32 fulfill a janus role, concluding the material that precedes (the prologue of Deuteronomy 1–11) and introducing the material that follows (the laws of Deuteronomy 12–26). See also M. Weinfeld, Deuteronomy 1–11 (AB 5; New York: Doubleday, 1991) 453. On the SP text, see A. Tal and M. Florentin, The Pentateuch:TheSamaritanVersionandtheMasoreticVersion (Tel Aviv: Haim Rubin Tel Aviv University Press, 2010). All of my quotations of the SP will be taken from this edition. Z. Zevit’s definition of an altar as a “manufactured, solid-based installation or artifact on which something is burnt for, or displayed before, or presented to a deity” covers all of these different stone altar types, TheReligionsofAncientIsrael:ASynthesisof ParallacticApproaches (London: Routledge, 2001), 276.

ALTARED STATES

101

include critical hermeneutical clues about how the new laws relate to older laws, governing stone installations. Yet, is such a theory borne out by the evidence?5 To be clear, my concern in this study is not with small and portable altars of various sorts, employed for libations, grain offerings, and incense, all of which are amply attested in the literary and material remains, but rather with large stone altars employed for animal sacrifice.6

I. DISTINCTIVE YET RELATED: THE STONE ALTAR LAWS IN THE SAMARITAN AND JEWISH PENTATEUCHS Of the four pieces of major altar legislation appearing in the Samaritan and Jewish Pentateuchs, the most famous is the centralization legislation of Deuteronomy (11:31–12:31), mandating the establishment of one sacrificial installation at the unnamed place of God’s own choosing and calling for the elimination of all other cultic precincts – Yahwistic or otherwise. Additional texts mandating the construction of major altars in the Samaritan and Jewish Pentateuchs appear in the Covenant Code (Exod 20:24–26) and the Priestly writing (Exod 27:1–7), respectively. Paradoxically, yet another piece of major altar legislation appears within Deuteronomy itself (27:1–8).7 Given that the main Tent of Meeting altar is constructed from acacia wood and plated with bronze (Exod 27:1–7), it will not comprise a main part of this investigation.8 5

6

7

8

The focus of this study is on the literary remains. The material remains will be dealt with in a future essay. S. Gitin, “The Four-Horned Altar and Sacred Place: An Archaeological Perspective,” in SacredTime,SacredPlace:ArchaeologyandtheReligionofIsrael (ed. B. M. Gittlen; Winona Lake, IN: Eisenbrauns, 2002) 95–123; I. Ziffer, “Altars and Art Objects,” in Beer-ShebaIII:TheEarlyIronIIAEnclosedSettlementandtheLateIronIIA-IronIIB Cities (3 vols.; ed. Z. Herzog and L. Singer-Avitz; Winona Lake, IN: Eisenbrauns, 2016), 3.1229–57. “The little section [Deut 27:1–8] bristles with difficulties, textual, critical, historical,” A. C. Welch, TheCodeofDeuteronomy:ANewTheoryofItsOrigin (London: J. Clarke & Co., 1924), 179. For Welch, the inclusion of Deut 27:1–8 indicated that the author did not believe in cultic centralization, Code of Deuteronomy, 181. Welch recognizes the profound problems with the content and sequence of the material in Deuteronomy 27, but for most modern commentators, there is no Deuteronomy without centralization. See E. Otto, Deuteronomium12–34 (2 vols.; HThKAT; Freiburg: Herder, 2016–2017), 2: 1910–58. P. Heger, TheThreeBiblicalAltarLaws–DevelopmentsintheSacrificialCultinPractice andTheology:PoliticalandEconomicBackground (BZAW 279; Berlin: de Gruyter,

102

GARY N. KNOPPERS

The three remaining pieces of Pentateuchal altar legislation are all related in some fashion to each other. Thus, matters of interpretation, application, and revision are not simply something characterizing the early Jewish and Samaritan reception of the stone altar laws, but are also embedded within the Pentateuch itself. As is well-known, Judeans in antiquity applied Deuteronomy’s vaguely formulated centralization legislation (Deut 11:31–12:31) to the benefit of the Jerusalem temple, while Samarians applied it to the benefit of the Mt. Gerizim temple.9 This fact is all the more significant, because the statutes calling for cultic unity and cultic purity are virtually identical in the Jewish and Samaritan Torahs. Aside from the variation between the consistent use of ‫בחר‬, “he chose,” in the Samaritan Pentateuch and the consistent use of ‫יבחר‬, “he will choose,” in the Masoretic Text, the two textual witnesses are virtually identical to one another.10 This means, among other things, that both communities could lay claim to the same prestigious texts, but understand them differently. In making the case for cultic unity and cultic purity, the centralization laws do not stand without precedent: they repeatedly allude to and revise older Covenant Code altar statutes (Exod 20:24–26).11 Deuteronomy’s mandate for one centralized altar implicitly cites an older standard, which it studiously rewrites.

1999) 171–206; C. Houtman, Exodus, 3: Chapters 20–40 (HCOT; Leuven: Peeters, 2000) 441–51; Zevit, Religions, 288–94; W. H. C. Propp, Exodus19–40 (AB 2A; New York: Doubleday, 2006) 334–37, 420–24. 9 Knoppers, JewsandSamaritans, 194–216. 10 Although the variation is sometimes taken as determinative to pointing to Mt. Gerizim, on the one hand, and to Jerusalem, on the other hand, the literary evidence is more complex. In neither version is a toponym or sanctuary named. In reading ‫( יבחר‬MT) or ‫( בחר‬SP), each major witness is internally self-consistent (Deut 12:5, 11, 14, 18, 21, 26; 14:23, 24, 25; 15:20; 16:2, 6, 7, 11, 15, 16; 17:8, 10; 18:6; 26:2; 31:11). See further, my JewsandSamaritans, 178–88. For a somewhat different perspective, see M. Kartveit, “The Place That the Lord Your God Will Choose,” HeBAI 4 (2015) 205–18. 11 J. Van Seters contends that the Covenant Code is a later (exilic) response to Deuteronomy, ALawBookfortheDiaspora:RevisionintheStudyoftheCovenantCode (New York: Oxford University Press, 2003). But see B. M. Levinson, “Is the Covenant Code an Exilic Composition? A Response to John Van Seters,” in In Search of Pre-exilic Israel:ProceedingsoftheOxfordOldTestamentSeminar (ed. J. Day; JSOTSup 406; London: T. & T. Clark, 2004) 297–315 (repr. in his “TheRightChorale”:Studiesin BiblicalLawandInterpretation [FAT 54; Tübingen: Mohr Siebeck, 2009] 276–330). J. Van Seters replies, “Revision in the Study of the Covenant Code and a Response to My Critics,” SJOT 21 (2007) 5–28.

ALTARED STATES

103

The Citation and Revision of the Covenant Code in Deuteronomy’s Centralization Legislation Deut 12:13–14 ‫ השמר לך פן תעלה עלתיך‬Deut 12:13 ‫בכל מקום אשר תראה‬ ‫ כי אם במקום אשר‬Deut 12:14 ‫)י(בחר יהוה באחד שבטיך שם תעלה עלתיך‬ ‫ושם תעשה כל אשר אנכי מצוך‬ Deut 12:13–14 Be careful lest you presentyourburntofferingsatany place that you happen to see, but ratherattheplaceYhwhwill choose/haschosen in one of your tribes. There you will present your burnt offerings and there you will practice all that I am commanding you.

Exod 20:24 ‫ מזבח אדמה תעשה לי‬Exod 20:24 ‫וזבחת עליו את עלתיך ואת שלמיך‬ ‫ בכל‬13‫את צאנך ואת בבקרך‬12 16 ‫ את שמי שמה‬15‫ אשר אזכיר‬14‫המקום‬ ‫אבוא אליך וברכתיך‬ Exod 20:24 An altar of earth you will make for me and youwillsacrifice uponityourburntofferings and your well-being offerings, your flock and your livestock; inevery place at which I shall cause my name to be remembered there, I shall come to you and bless you.

Whereas the Covenant Code allows for a multiplicity of earthen and stone altars, wherever Yhwh causes his name to be invoked, Deuteronomy’s rewriting of this legislation allows for only one major altar.17 Older statutes are carefully employed, in this case, to authorize a major transformation in ritual practice.18 But there is a complication. Even though the 12

13

14

15

16

17

18

So MT and SP (lectio brevior). Many medieval Heb. MSS, LXXB, and Syr. preface conjunctive wāw. So the MT (maximum differentiation). SP, Sam.Tg., Tg.Onq., and Tg.Ps.–J. Exod 20:24 read the partitive min ‫מצאנך ומבקרך‬, in conformity with the lemma of Deut 12:21 ‫ומצאנך‬ ‫מבקרך‬. So the MT (lectiodifficilior). Some medieval Heb. MSS, the SP, Sam.Tg., and a marginal gloss to Tg.Neof. read ‫במקום‬, “at the place.” The SP lemma, although the shorter reading, likely corrects toward the standard reading of Deuteronomy (‫ ;במקום‬Deut 12:5, 11, 14, 18, 21, 26). So the MT and LXX (ἐπονομάσω). The SP lemma, “(the place at which) I have caused (‫( )אזכרתי‬my name) to be remembered,” alludes to Gen 12:6–7. Syr. and some witnesses toTg.Neof. reflect ‫תזכיר‬. So the LXX and the SP. The MT (‫ )את שמי‬lacks ‫ שמה‬because of haplography (homoioarkton). The open-ended nature of the formula, ‫בכל המקום אשר אזכיר את שמי אבוא אליך וברכתיך‬, in Exod 20:24, which does not mention priests, human kings, or temples, is stressed by S. Chavel’s recent study, “A Kingdom of Priests and Its Earthen Altars,” VT 65 (2016) 169–222. Chavel argues that the narratives in Exodus 19–24 programmatically react against hierarchical royal polities in northern and southern Israel, legitimating potentially free access of all Israelite subjects directly to the divine king. B. M. Levinson explores in great detail the innovative reuse of the Covenant Code altar laws in Deuteronomy’s centralization legislation, DeuteronomyandtheHermeneutics ofLegalInnovation (New York: Oxford University Press, 1997) 23–52.

104

GARY N. KNOPPERS

centralization laws stress that only one particular site is to be used for all Israelite sacred donations, holy vows, and animal sacrifices, Deuteronomy contains a second piece of altar legislation, remanding Israelites, upon entering the land, to travel to Mt. Gerizim and Mt. Ebal and instructing them, among other things, to construct an altar and to erect large stones and inscribe the Torah upon them (Deut 27:1–26). At this sacrificial installation, to be constructed at Mt. Gerizim (SP; MT Mt. Ebal) of whole stones untouched by iron (tools), the Israelites are to present their burnt offerings and offerings of well-being.19 There, the people are told to feast and rejoice before Yhwh their God (Deut 27:7). As with the centralization legislation, the Mt. Gerizim/Mt. Ebal legislation recalls earlier textual precedent, namely the Covenant Code regulations governing the creation of earthen and stone altars (Exod 20:24–26).20 19

20

In Deut 27:4, the lemma ‫ הרגריזים‬of the SP, Papyrus Giessen 19 (argar[i]zim), the Samareitikon (argarizim), and the OL (inmonteGarzin) is likely to be earlier than that of the MT, “Mt. Ebal” (‫ )הר עיבל‬and most witnesses to the LXX (orei Gaibal). A. Schenker demonstrates the antiquity of the lemma Mt. Gerizim in LXX witnesses and the other Versions (e.g., the Old Boharic), “Le Seigneur choisira–t-il le lieu de son nom ou l’a-t-il choisi ?,” in ScriptureinTransition (ed. A. Voitila and J. Jokiranta; JSJSup 126; Leiden: Brill, 2008) 339–51; idem, “Textgeschichtliches zum Samaritischen Pentateuch und Samareitikon,” in Samaritans:PastandPresent (ed. M. Mor and F. V. Reiterer; Studia Samaritana 5; Berlin: de Gruyter, 2010) 105–20. The same reading as that of the SP, written in scriptiocontinua, ‫בהרגרזים‬, appears on a late-2nd or 1st century BCE DSS fragment of Deut 27:4b–6, J. H. Charlesworth, “What Is a Variant? Announcing a Dead Sea Scroll Fragment of Deuteronomy,” Maarav16 (2009) 201–12, 273–74 (Pls. ix–x). On the two textual variants (Mt. Gerizim vs. Mt. Ebal), C. Nihan comments, “it is entirely unlikely that Mount Ebal, which stands as the place for curses in v. 13, should be chosen for the erection of stones engraved with the Torah and the building of an altar as commanded in vv. 4–8,” “The Torah between Samaria and Judah: Shechem and Gerizim in Deuteronomy and Joshua,” in ThePentateuchasTorah:NewModelsforUnderstanding Its Promulgation and Acceptance (ed. G. N. Knoppers and B. M. Levinson; Winona Lake, IN: Eisenbrauns, 2007) 187–223 (213; italics his). For a review of longstanding theories suggesting a northern origin of this material, see C. Edenburg and R. Müller, “A Northern Provenance for Deuteronomy?” HeBAI 4 (2015) 148–61, and the further references cited there. For a recent exposition of this perspective, see S. Schorch, “The Samaritan Version of Deuteronomy and the Origin of Deuteronomy,” in Samaria,Samarians,Samaritans (ed. J. Zsengellér; Studia Samaritana 6; Berlin: de Gruyter, 2011) 23–37. In my view, the relevant material in Deuteronomy 11 and 27, whatever its complicated origins, does not belong to the oldest layers of Deuteronomy, G. N. Knoppers, “The Northern Context of the Law-Code in Deuteronomy,” HeBAI 4 (2015) 162–83. Similarly, J. L’Hour, “L’alliance de Sichem,” RB 69, 1962, 5–36, 161–184, 349–368 (168–169); E. Nielsen, Deuteronomium (HAT 6; Tübingen: Mohr-Siebeck, 1995), 128–130; C. Nihan, “The Torah between Samaria and Judah,” in ThePentateuchasTorah (ed. G. N. Knoppers and B. M. Levinson; Winona Lake, IN: Eisenbrauns, 2007), 187–223 (209); R. Müller, “TheAltaronMountGerizim (Deut27:1–8):CenterorPeriphery?” in CentresandPeripheriesintheEarlySecond TemplePeriod (ed. E. Ben Zvi and C. Levin; FAT 108; Tübingen: Mohr Siebeck, 2016) 197–214; Otto, Deuteronomium23,16–34,12, 1925–30.

ALTARED STATES

105

The Citation of the Covenant Code Altar Law in Deuteronomy 27 Deut 27:5–7

Exod 20:24–25

‫ והעלית עליו עולת ליהוה אלהיך‬Deut 27:6

‫ מזבח אדמה תעשה לי‬Exod 20:24

‫ וזבחת שלמים ואכלת שם‬Deut 27:7 ‫ושמחת לפני יהוה אלהיך‬

‫וזבחת עליו את עלתיך ואת שלמיך‬ ‫את צאנך ואת בקרך בכל המקום‬ ‫אשר אזכיר את שמי שמה אבוא אליך‬ ‫וברכתיך‬

‫ ובנית שם מזבח ליהוה אלהיך‬Deut 27:5 ‫מזבח אבנים לא תניף עליהם ברזל‬ ‫ אבנים שלמות תבנה את‬Deut 27:6a ‫מזבח יהוה אלהיך‬

‫ ואם מזבח אבנים תעשה‬Exod 20:25 ‫לי לא תבנה אתהן גזית כי חרבך הנפת‬ ‫ ותחללה‬21‫עליה‬

Deut 27:6b–7 Youwillofferuponit burntofferings to Yhwh your God andyouwillsacrificewell-being offerings and you will eat there and rejoice before Yhwh your God

Exod 20:24 An altar of earth you will make for me andyouwillsacrifice uponityourburntofferingsand yourwell-beingofferings, your flock and your livestock; in every place at which I shall cause my name to be remembered, I shall come to you and bless you.

Deut 27:5–6a And you will build there an altar to Yhwh your God, analtarofstones.Donotwield(an) iron(tool)uponthem. (Of) whole stonesyouwillbuildthealtar of Yhwh your God.

Exod 20:25 And if analtarofstones you make for me, donotbuilditof hewnstones. Ifyouwieldyourchisel uponit,youwillprofaneit.

As the studies by Nihan demonstrate, the inverted quotation of select Covenant Code statutes in Deuteronomy 27 calls attention to the literary dependence of the Mt. Gerizim altar legislation upon the Covenant Code altar legislation.22 Although the language used for the mode of altar construction is not exactly identical, both legal drafts authorize the introduction and deployment of altars consisting of unfinished stones. In this case, the 21

22

The masculine singular suffixes in SP (‫ )עליו ותחללהו‬refer to the (masculine) ‫מזבח‬. Cf. LXXAFMh, the Sahidic, Cyr., and Thdt. ἐπ’ αὐτό καὶ μεμίανται, “upon it and they have been defiled.” The feminine singular suffixes in MT Exod 20:25 (‫ )עליה ותחללה‬lack a clear referent, unless it is ‫גזית‬. In the LXXB, the “stones” (λίθων) are the referent, ἐπ’ αὐτούς καὶ μεμίανται, “upon them and they have been defiled.” Some witnesses to Tg.Neof. also have the stones as the referent. Nihan, “Samaria and Judah,” 216; idem, “L’autel sur le mont Garizim: Deutéronome 27 et la redaction de la torah entre Samaritains et Judéens à l’époque achéménide,” Transeu 36 (2008) 97–124; idem, “Garizim et Ébal dans le Pentateuque: Quelques remarques en marge de la publication d’un fragment du Deutéronome,” Sem 54 (2012) 185–210.

106

GARY N. KNOPPERS

text legislates the erection of a stone altar that basically conforms to Covenant Code precedent at a specific site. In the Covenant Code the proscription of well-cut stones is evidently linked to a more general concern to separate the deity and his sacred space from defilement through physical modification.23 Hewn (‫)גזית‬, dressed stones would ipsofacto be defiled stones, whereas whole (‫)שלמות‬, undressed stones would be undefiled stones.24 In Deuteronomy the Covenant Code prohibition of using steps to ascend the altar, “because you must not reveal your nakedness upon it” (Exod 20:26), does not reappear, but otherwise the altar decrees in Deuteronomy 27 are profoundly indebted to those in the Covenant Code.25 In my recent monograph, I explore some of the complexities of early Judean and Samarian interpretations of the altar laws in the Covenant Code and Deuteronomy.26 Early Judean and Samarian scribes could and did construe the relations among these different altar laws in different ways. Judean and Samarian interpreters engaged the multiplicity of Covenant Code statutes (Exod 20:24–26), centralization regulations (Deut 12:13–14, 27), and Mt. Gerizim/Mt. Ebal commandments by picking up on different cues within the texts of Exodus and Deuteronomy to defend the sanctity of their own sanctuaries. More than that, I argue that the edition of Deuteronomy shared by the Judean and Samarian communities accommodated both traditions. That discussion may be developed and extended here.

II. SAMARIAN READING STRATEGIES

IN

APPROACHING THE ALTAR LAWS

Proponents of the Samarian position could cite the dependence of the Mt. Gerizim altar instructions in Deut 27:4–8 upon the Covenant Code stone altar instructions to validate Mt. Gerizim’s status. The Mt. Gerizim 23

24

25

26

S. Olyan, “Why an Altar of Unfinished Stones? Some Thoughts on Ex 20,25 and Dtn 27,5–6,” ZAW 108 (1996) 161–71 (163). Similar concerns about alterations causing defilement, whether initiated by human agents or not, are evident in statutes governing sacrificial animals and priests, Olyan, “Unfinished Stones,” 167–70. Animals presented for sacrifice must lack any blemish (‫)מום‬, such as blindness, lameness, physical injury, or disease (Deut 15:21; 17:1; Lev 1:3, 10; 3:1; 4:32; 22:17–25). Priests, according to the so-called Holiness Code (Lev 21:16–24), must also be without bodily imperfections, such as blindness, lameness, injury, and disease. See S. M. Olyan, Disability in the Hebrew Bible: Interpreting Mental and Physical Differences (Cambridge: Cambridge University Press, 2008) 27–36. Reading with the MT and the LXX, ‫( עליו‬lectiodifficilior). SP “to him” (‫)אליו‬, i.e., to the deity. Knoppers, JewsandSamaritans, 194–212.

ALTARED STATES

107

altar could be considered as continuous with the older pattern. Moreover, early Samarian interpreters could cite the literary contextualization of the central law code within Deuteronomy to point to the privileged status given to the Shechem area. Instructions about all-Israelite ceremonies to be conducted at Mt. Gerizim and Mt. Ebal frame the introduction and the conclusion to the central law code. Ceremonies at Mt. Gerizim and Mt. Ebal Framing the Central Law Code A Blessings and Curses at Gerizim and Ebal (11:29–30) B Introduction to the Law Collection (11:31–12:1) C Laws of the Covenant (12:2–26:15) B1 Conclusion to the Law Collection (26:16–19) A1 Ceremonies at Gerizim and Ebal (27:1–26)

Since instructions directing all Israelites to Mt. Gerizim and Mt. Ebal both preface and conclude Deuteronomy’s central law code, these instructions could be employed as a hermeneutical key to construing the force of the centralization instructions themselves, which ambiguously do not name a specific site. Samarian interpreters could also cite literary parallels between the Mt. Gerizim/Mt. Ebal statutes and the centralization legislation. Studies by Rofé and Schorch have called attention to many of these.27 These are, for example, the only two cases in biblical literature in which a sacred place is dedicated by a direct order, rather than by an epiphany or an angelophany.28 27

28

A. Rofé, Deuteronomy (London: T. & T. Clark, 2002) 98–99; Schorch, “Samaritan Version,” 23–37. Yet, I would argue that there are also some important differences. See Section III below. Rofé, Deuteronomy, 100; Schorch, “Samaritan Version,” 28. In the case of Deuteronomy 27, the command is, however, delivered by Moses (who is accompanied by the elders; Deut 27:1), rather than by Yhwh. That this was likely a point of contention already within antiquity may be seen in the formulation of the Samaritan tenth commandment (largely a pastiche of quotes from other Pentateuchal texts), which presents the commands drawn from Deuteronomy 27 as stemming from a divine voice. Forms of divine-human communication are debated within other Pentateuchal texts. Responding, e.g., to Aaron and Miriam’s questioning whether Yhwh speaks (dibbēr) through them or only through Moses (Num 12:2), Yhwh pointedly declares that he speaks with his servant Moses “mouth to mouth” (peh᾿elpeh᾿ădabbēr-bô) and not through prophetic visions or dreams (Num 12:6–7). In great detail, see R. Achenbach, DieVollendungder Tora (BZAR 3; Wiesbaden: Harrossowitz, 2003) 267–301. The insistence on Moses’ unrivalled prophetic status reflects debates among the authors responsible for Numbers, Deuteronomy, and the prophetic books about issues of divine authority, mediation, and communication. In this respect, one should also acknowledge the incorporation of Deut 5:28b–29 and Deut 18:18–22 into some pre-Samaritan DSS witnesses of Exodus 20 and into the SP, which underscores, among other things, Moses’ intermediary role, M. Kartveit, TheOriginoftheSamaritans (VTSup 128; Leiden: Brill, 2009) 259–312.

108

GARY N. KNOPPERS

In both cases, an altar is specifically mentioned. The timing of the implementation is also important, “when you cross the Jordan” (Deut 11:31; 27:4).29 Both passages stress the importance of feasting and rejoicing in the presence of Yhwh. To this point, it might be objected that the phraseology is stereotypically Deuteronomic and granted the literary setting of the events depicted on the Steppes of Moab, such verbiage is to be expected. On this basis, the parallels between the temporal subordinate clauses found in the two passages would hold no exceptional value. But while the prepositional phrase “across the Jordan” (e.g., Deut 1:1, 5; 2:29; 3:8, 25, 27; 4:41, 46) and the participial phrase, “crossing the Jordan” (e.g., Deut 4:22, 26; 9:1; 30:18; 31:13; 32:47), are quite common, the deployment of the temporal subordinate clause (“when you cross the Jordan” or “upon your crossing the Jordan”) as an introduction to an injunction (or set of injunctions) is fairly rare in Deuteronomy.30 In fact, to the best of my knowledge, it only occurs in these two sets of passages (Deut 11:31; 27:4, 12). 29

30

The yet more specific temporal delimitation, appearing in Deut 27:2, “on the day, you cross the Jordan,” pertains to the construction and inscription of a public stele (Section III below). The introduction of Deut 12:10, ‫ועברתם את הירדן וישבתם בארץ‬, “And (when) you cross over the Jordan and reside in the land,” is also relevant, but differs from these other instances in that it delays the fulfillment of the implementation of the centralization decree until that time, “when he (Yhwh) grants you rest from all of your surrounding enemies and you dwell securely” (‫)והניח לכם מכל איביכם מסביב וישבתם בטח‬. Precisely when this occurred becomes a point of contention within the history of the Deuteronomistic work itself. Did Israel achieve this peace in the time of Joshua (Josh 11:23; 21:43–45; Josephus, Ant. 7.342) or in the time of Solomon (1 Kgs 5:18; 6:1)? The latter interpretation dominates the Deuteronomistic depiction of Solomon’s reign, while the former interpretation becomes a major emphasis in the Samaritan chronicles, G. N. Knoppers, “Samaritan Conceptions of Jewish Origins and Jewish Conceptions of Samaritan Origins: Any Common Ground?” in DieSamaritanerunddieBibel:Die SamaritanerinderbiblischenTradition–diejüdischeundfrühchristlicheGeschichte insamaritanischenQuellen (ed. J. Frey, U. Schattner-Rieser, and K. Schmid; Studia Samaritana 7; Berlin: de Gruyter, 2012) 81–118. It is interesting that there is a minor tradition within Joshua itself (e.g., Josh 13:1b–6; 23:7–10), insisting that there were sites and nations yet to be conquered. In the analysis of R. Smend, these texts belong to the late (postexilic) layer(s) of DtrN, “Das Gesetz und die Völker: Ein Beitrag zur deuteronomistischen Redaktionsgeschichte,” in ProblemebiblischerTheologie:Festschrift GerhardvonRad (ed. H. W. Wolff; München: Kaiser, 1971) 494–509 [transl. “The Law and the Nations: A Contribution to Deuteronomistic Tradition History,” in ReconsideringIsraelandJudah:TheDeuteronomisticHistoryinRecentThought, (SBTS 8; ed. G. N. Knoppers and J. G. McConville; Winona Lake, IN: Eisenbrauns, 2000) 95–110]. If the medieval Samaritan traditions about a completely successful conquest in the time of Joshua have ancient antecedents, it would be tempting to tie the origin of these particular texts in Joshua to Persian/Hellenistic Judean and Samarian debates about the extent of the imagined Israelite conquest.

ALTARED STATES

109

The hortatory employment of the verbal root ‫שמח‬, along with the cultic formula ‫לפני יהוה‬, is also noteworthy.31 Within Deuteronomy such usage occurs in literary settings (re)structuring common meals within a larger liturgical celebration (Deut 12:7, 12, 18; 14:26; 16:11; 27:7).32 Leaving aside, for the moment, the case of rejoicing before Yhwh at Mt. Gerizim (Deut 27:7), each of these other commandments (or sets of commandments) is tied to the central sanctuary: sacrificial offerings (12:7, 12), tithes, firstlings, votive offerings, and freewill offerings (12:16–18), the centralized tithe (14:22–29), and the Festival of Weeks (16:9–12).33 One can understand, therefore, why the selective deployment of such verbiage in the Mt. Gerizim/Mt. Ebal legislation should be viewed as deliberate. Parallels between the Centralization Laws and the Gerizim/Ebal Laws Deut 27:2–7

Deut 11:31–12:7, 27

‫ והיה בעברכם את הירדן תקימו‬Deut. 27:4 ‫את האבנים האלה אשר אנכי מצוה אתכם‬ ‫היום בהרגריזים ושדת אותם בשיד‬

‫ כי אתם עברים את־הירדן‬Deut. 11:31 ‫לבא לרשת את־הארץ אשר יהוה אלהיכם‬  ‫נתן לכם‬

‫ ובנית שם מזבח ליהוה אלהיך‬Deut. 27:5 ‫מזבח אבנים לא תניף עליהם‬ ‫ברזל‬

‫ כי אם אל המקום אשר‬Deut. 12:5 ‫)י(בחר יהוה אלהיכם מכל שבטיכם‬ ‫לשים את שמו שם לשכנו תדרשו‬  ‫ובאתם שמה‬

‫ אבנים שלמות תבנה את מזבח‬Deut. 27:6  ‫יהוה אלהיך והעלית עליו עולת ליהוה אלהיך‬

‫ והבאתם שמה עלתיכם‬Deut. 12:6 ‫וזבחיכם ואת מעשרתיכם ואת תרומתיכם‬ ‫ונדריכם ונדבתיכם ובכרת בקרכם וצאנכם‬

‫ וזבחת שלמים ואכלת שם‬Deut. 27:7  ‫ושמחת לפני יהוה אלהיך‬

 ‫ ואכלתם שם לפני יהוה‬Deut. 12:7 ‫אלהיכם ושמחתם בכל משלח ידיכם‬ ‫אתם ובתיכם אשר ברכך יהוה אלהיך‬ ‫ ועשית עלתיך הבשר והדם‬Deut. 12:27 ‫על־מזבח יהוה אלהיך ודם־זבחיך ישפך‬ ‫על־מזבח יהוה אלהיך והבשר תאכל‬

31

32

33

On the theme of rejoicing as part of a larger festal theology in Deuteronomy, see G. Braulik, “Die Freude des Festes: Das Kultverständnis des Deuteronomium – die älteste biblische Festtheorie,” in his StudienzurTheologiedesDeuteronomiums (SBAB 2; Stuttgart: Verlag Katholisches Bibelwerk, 1988), 161–218; P. Altmann, Festive Meals in AncientIsrael:Deuteronomy’sIdentityPoliticsintheirAncientNearEasternContext (BZAW 424; Berlin: de Gruyter, 2011). G. Vanoni, Literarkritik und Grammatik: Untersuchung der Wiederholungen und Spannungen in 1 Kön 11-12 (Arbeiten zu Text und Sprache im Alten Testament 21; St. Ottilien: EOS, 1984) 176–80; Braulik, “Freude des Festes,” 212–14. A similar combination may be implied in the legislation restructuring the Festival of Booths (Deut 16:13–15). The feast is to be dedicated to Yhwh and the people are told: “in all of the works of your hands, you will only be joyful” (‫)שמח והיית אך‬.

110

GARY N. KNOPPERS

If the Mt. Gerizim directives are indeed written to comport with the Deuteronomic centralization commands, would this not effectively indicate that the Mt. Gerizim altar was the one and only legitimate locus of sacrifice? On the basis of these considerations and others, the place for the name of Yhwh, which consistently goes unnamed in Deuteronomy, could be considered as none other than Mt. Gerizim itself. Such an interpretation of the Mt. Gerizim altar law is codified within the (later) Samaritan tenth commandment, which consists almost entirely of selections selectively drawn from other passages found in the Pentateuch (Exod 13:11a, Deut 11:29b, 27:2b–3a, 4a, 5–7, and 11:30, in this particular order).34 The Altar in the Samaritan Tenth Commandment Translation

SP Exod 20:13‫∥ ו–ג‬ Deut 5:13‫ז–ד‬

And when you cross the Jordan, you are to erect these stones about which I am commanding you this day upon Mount Gerizim (Deut 27:4a).

‫והיה בעברכם את הירדן תקימו‬ ‫את האבנים האלה אשר אנכי‬ ‫מצוה אתכם היום בהרגריזים‬

And you will build there an altar to Yhwh your God, an altar of stones. Do not wield (an) iron (tool) upon them (Deut 27:5).

‫ובנית שם מזבח ליהוה אלהיך‬ ‫מזבח אבנים לא תניף עליהם‬ ‫ברזל‬

Of whole stones you will build the altar of Yhwh your God. You will lift up upon it burnt offerings to Yhwh your God (Deut 27:6).

‫אבנים שלמות תבנה את מזבח‬ ‫יהוה אלהיך‬ ‫והעלית עליו עלות ליהוה‬ ‫אלהיך‬

And you will sacrifice well-being offerings and you will feast there and rejoice before Yhwh your God (Deut 27:7). 34

‫וזבחת שלמים ואכלת שם‬ ‫ושמחת לפני יהוה אלהיך‬

F. Dexinger, “Das Garizimgebot im Dekalog der Samaritaner, in Studien zum Pentateuch (ed. G. Braulik; Vienna: Herder, 1977) 111–33; J. E. Sanderson, AnExodusScrollfrom Qumran:4QpaleoExodmandtheSamaritanTradition (HSS 30; Atlanta, GA: Scholars Press, 1986) 235–37, 317–20; Z. Ben-Ḥayyim, “The Tenth Commandment in Samaritan Research,” in EssaysinHonourofG.D.Sixdenier:NewSamaritanStudiesoftheSociété d’étudessamaritaines (ed. A. D. Crown and L. Davey; Studies in Judaica 5; Sydney: Mandelbaum, 1995), 487–92; J. Zsengellér, GerizimasIsrael (Utrechtse Theologische Reeks 38; Utrecht: University of Utrecht, 1998) 37–42; I. Himbaza, Le Décalogue et l’histoiredutexte:ÉtudesdesformestextuellesduDécalogueetleursimplicationsdans l’histoiredutextedel’AncienTestament (OBO 207; Göttingen: Vandenhoeck & Ruprecht, 2004) 63–66, 183–219; Kartveit, Origin, 290–96; Pummer, “Samaritans and their Pentateuch,” 244–47. More generally, see E. C. Ulrich, “The Developmental Growth of the Pentateuch in the Second Temple Period,” in TheDeadSeaScrollsandtheDevelopmentalCompositionoftheBible (VTSup 169; Leiden: Brill, 2015) 29–45.

ALTARED STATES

111

By including the Mt. Gerizim diktats among the ten words spoken by Yhwh to his people on Mt. Sinai, Samaritan scribes granted these decrees a privileged place within the Torah. Whereas in Deut 27:1, the directives are given by Moses (accompanied by the elders) to the people, the literary recontextualization of the material drawn from Deut 27:4a, 5– 7 within the Ten Commandments elevates the Mt. Gerizim altar commands into orders given directly by Yhwh to his people.35 In so doing, Samaritan scribes accord the (new) tenth commandment an extraordinary status. Yet, the literary contextualization of this pastiche of instructions is important in another way. By placing these decrees in a literary context (Exodus 20 ∥ Deuteronomy 5) prior to the Covenant Code altar laws (Exod 20:24–26), the centralization decrees (Deut 11:31–12:31), and the Mt. Gerizim/Mt. Ebal instructions (Deut 27:1–26), the writers effectively suggest a continuity of identity among the Covenant Code altar instructions, the Mt. Gerizim altar, and the central altar. Elements within the Covenant Code decrees, the centralization instructions, and the Mt. Gerizim/Mt. Ebal commands echo what was previously revealed in the Ten Commandments.36 In the literary flow of the Samaritan Pentateuch, one set of decrees anticipates the others.37 In so doing, the Samaritan tenth commandment effectively sanctions the mode of natural fieldstone construction called for in the composition of the Mt. Gerizim sacrificial altar as the compositional design called for in building the central altar. 35

36

37

Although Moses is grouped with the Israelite elders, the directives to the Israelites stem from him: “Be careful to observe the entire commandment, which I am commanding you this day” (‫ ;שמר את כל המצוה אשר אנכי מצוה אתכם היום‬Deut 27:1). The context, composition, and content of the Samaritan tenth commandment will be dealt with more fully in a later study. It is no wonder, then, that later (medieval) Samaritan chronicles speak of the fulfillment of these instructions by Joshua and the people at Mt. Gerizim, after they have successfully conquered the land. One of the Samaritan Chronicles, the Chronicon Samaritanum (21), depicts a pan-Israelite feast occurring at “the blessed mountain” On the so-called Samaritan Joshua, see T. W. J. Juynboll,ChroniconSamaritanum: Arabiceconscriptum,cuititulusestLiberJosuae (Leiden: Luchtmans, 1848) [transl. O.T. Crane, The Samaritan Chronicle or the Book of Joshua the Son of Nun (New York: Alden, 1890)]. There, Joshua offers sacrifices for the people before the land apportionment begins (ChroniconSamaritanum 22). In another medieval chronicle (Abū ᾿lFatḥ 14,16), Joshua builds an altar of stones at Mt. Gerizim, sacrifices there, and reads the Torah in its entirety with half the people facing Mt. Gerizim and the other half facing Mt. Ebal. On the Kitābal–TarīkhofAbū᾿lFatḥ, see E. Vilmar, AbulfathiAnnales Samaritani:quosArabiceediditcumprolegomenis (Gotha: Perthes, 1865) [transl. P. L. Stenhouse, TheKitābal-tarīkhofAbū᾿lFatḥ (Studies in Judaica 1; Sydney: Mandelbaum, 1985)].

112

GARY N. KNOPPERS

III. JUDEAN READING STRATEGIES IN APPROACHING THE ALTAR LAWS We have been discussing various ways by which early Samarian and Samaritan interpreters could construe the force of various altar regulations in favor of the Mt. Gerizim sanctuary. But early Judean readers of Deuteronomy could and did read the same texts differently. Judeans could view the dependence of the Mt. Gerizim altar instructions upon the Covenant Code instructions as a tangible hint that the Mt. Gerizim altar was, in fact, not the permanent altar called for in the centralization legislation. Because the Covenant Code legislation allowed for multiple altars, the stone altar to be built on Mt. Gerizim represented one instantiation of the Covenant Code altar decrees and was, therefore, representative of the old order, not a harbinger of the new order.38 From this Judean perspective, the Mt. Gerizim altar was a temporary phenomenon, a commemorative altar to celebrate the successful entrance into the promised land, rather than the mandated central altar called for in the Deuteronomic law-code. Is the instruction, for instance, to assemble large stones, cover them with plaster, and “write distinctly all other words of this Torah” upon them (27:2–3, 8) a one-time command or a repetitive duty? If the former holds, would not the altar command also refer to a one-time event?39 To take a second example, are the Israelites expected to assemble continually on Mt. Gerizim to pronounce the blessing and upon Mt. Ebal for the pronouncement of the curse (Deut 11:29–30) or are they supposed to do so, when they first enter the land? If the latter holds, the people’s participation in national ceremonies in the Shechem area consists of a one-time commemorative public liturgy. Judean literati could also observe that the critical Deuteronomic language of divine election (‫יבחר‬/‫ )בחר‬does not appear in either set of instructions directing Israelites to the Shechem area. Nor does either passage speak of Mt. Gerizim as “the place” (‫ )המקום‬at which Yhwh causes his name to reside (Deut 12:11; 14:23; 16:2, 6, 11; 26:2).40 From a Judean perspective, the absence of such critical features would speak volumes. 38

39 40

The later reworking of Exod 20:24 as ‫ במקום‬in the SP reacts, in part, to this ambiguity. The change from ‫ בכל המקום‬to ‫ במקום‬is not unique to the SP, but appears in some nonSamaritan medieval textual witnesses, as well (see n. 14 above). So also Chavel, “Kingdom of Priests,” 204–5. M. Weinfeld, DeuteronomyandtheDeuteronomicSchool (Oxford: Clarendon, 1972) 325 (nos. 3, 4); A. Schenker suggests another possibility for construing the force of the name formula in Deuteronomy, based on the translation of that phrase in LXX 1 Kings 18 as reflecting a generic meaning, rather than a numerical meaning, “Der Ort, an dem Jhwhs wohnt: Eine oder mehre Stätten?” HeBAI 4 (2015) 219–29. On the use of the formula within a broader comparative context, see S. L. Richter, TheDeuteronomisticHistory and the Name Theology: lešakkēn šemô šām in the Bible and the Ancient Near East (BZAW 318; Berlin: de Gruyter, 2002).

ALTARED STATES

113

To be sure, the Mt. Gerizim/Mt. Ebal texts play a very positive role. The pan-Israelite ceremonies dedicate the land to the deity, who instructed the Israelites on the Steppes of Moab and granted them safe passage. The erection of an official stele in the central hill country proclaims Yhwh’s ownership of the land, while the inscription of the Torah upon that stele marks the Torah’s authority over the people living within that land. The proceedings memorialize, therefore, a signal moment in the people’s past, a formative experience in shaping the people’s identity before God.41 Nevertheless, if the public liturgies the Israelites were to implement are inherently occasional in nature, the Mt. Gerizim altar would not be the single altar called for in Deuteronomy’s centralization mandate.42 The Shechem-area ceremonies represent one stage in a longer journey, not the culmination of that journey. Such a southern interpretation is embedded in the biblical book of Joshua and in a somewhat different way in the Dead Sea Scrolls text of 4QJoshuaᵃ.43 The altar Joshua dedicates at Mt. Ebal is made from whole 41

42

43

From a Judean literary perspective, such a long-term view extends beyond Joshua and is enshrined within the rest of the Former Prophets (or the Enneateuch), C. Nihan, “L’autel sur le mont Garizim,” 97–124. That Deuteronomy contains instructions directing Israelites to specific northern Israelite sites is noteworthy, because the work does not mention many Israelite toponyms, Knoppers, “Northern Context, 167–69. Paradoxically, one function of the instructions favorable to the Shechem area could be to indicate that none of the named northern sites was the designated place of God’s own choosing. The LXX is an important textual witness, but also a critical tool in gaining a better sense of literary growth in the text of Joshua during the last centuries BCE. See further A. Rofé, “The Editing of the Book of Joshua in the Light of 4QJoshᵃ,” in NewQumran TextsandStudies:ProceedingsoftheFirstMeetingoftheInternationalOrganization forQumranStudies,Paris1992 (ed. G. J. Brooke; STDJ 15; Leiden: Brill, 1994) 73–80; R. D. Nelson, Joshua:ACommentary (OTL; Louisville: John Knox, 1997) 22–24, 115– 20; E. Noort, “The Traditions of Ebal and Gerizim: Theological Positions in the Book of Joshua,” in DeuteronomyandDeuteronomicLiterature:FestschriftC.H.W.Brekelmans (ed. M. Vervenne and J. Lust; BETL 133; Leuven: Peeters, 1997) 161–80; idem, “4QJoshua and the History of Tradition in the Book of Joshua,” JNSL 24/2 (1998) 127–44; idem, DasBuchJosua:ForschungsgeschichteundProblemfelder (Erträge der Forschung 292; Darmstadt: Wissenschaftliche Buchgesellschaft, 1998) 46–98; A. G. Auld, “Joshua: The Hebrew and Greek Texts,” in JoshuaRetold:SynopticPerspectives (Edinburgh: T. & T. Clark, 1998) 7–18; E. Tov, “The Growth of the Book of Joshua in Light of the Evidence of the Septuagint,” in TheGreekandHebrewBible:CollectedEssays ontheSeptuagint (VTSup 72; Leiden: Brill, 1999) 385–96; idem, TextualCriticism, 294–99; idem, “The Literary Development of the Book of Joshua as Reflected in the Masoretic Text, the LXX, and 4QJoshᵃ,” in TextualCriticismoftheHebrewBible,Qumran, Septuagint: Collected Essays (VTSup 167; Leiden: Brill, 2015) 132–53; K. de Troyer, “Building the Altar and Reading the Law: The Journeys of Joshua 8:30–35,” in ReadingthePresentintheQumranLibrary:ThePerceptionoftheContemporarybyMeans ofScripturalInterpretation (ed. K. de Troyer and A. Lange; SBLSymS 30; Atlanta: Society of Biblical Literature, 2005) 141–62.

114

GARY N. KNOPPERS

stones (‫)אבנים שלמות‬, untouched by an iron tool (‫;לא־הניף עליהן ברזל‬ MT Josh 8:30–31), in explicit conformity with the directives of (MT) Deut 27:4–8. Given that the passage also occurs in a different literary context in the MT (8:30–35) from that of the LXX (9:2a–f), the pericope seems to have been interpolated into Joshua at a late stage in the work’s compositional history. In any event, the Joshua text is carefully written in such a way so as to underscore the fulfillment of the Mt. Gerizim/Mt. Ebal instructions in a text closely resembling MT Deuteronomy.44 That the nature, sequence, and import of the instructions governing Israel’s ritual conduct, upon entering the land, were debated in Judean circles within antiquity is evident in the DSS witness of 4QJoshuaᵃ.45 In this fragmentary text, dating to the 2nd or 1st century BCE, Joshua and the Israelite leaders conduct a commemorative rite earlier than that depicted either in the MT (8:30–35) or in the LXX (9:2a–f).46 The Setting and Sequence of 4QJoshuaᵃ 4QJoshuaᵃ

44

45

46

MT Joshua

Versification

‫ויכתב שם על האבנים את‬ ‫משנה תורת משה אשר כתב‬ ‫לפני בני ישראל‬

Josh 8:32

‫וכל ישראל וזקניו ושטרים‬ ‫ושפטיו עמדים מזה ומזה לארון‬ ‫נגד הכהנים הלוים נשאי ארון‬ ‫ברית יהוה כגר כאזרח חציו אל‬ ‫מול הר גרזים והחציו אל מול‬ ‫הר עיבל כאשר צוה משה עבד‬ ‫יהוה לברך את העם ישראל‬ ‫בראשנה‬

Josh 8:33

The double citation formula: ‫כאשר צוה משה עבד־יהוה‬, “as Moses the servant of Yhwh commanded,” and: ‫ככתוב בספר תורת משה‬, “as is written in the instruction of Moses,” underlines the point (Josh 8:31, 33). The text resonates with certain features found among the instructions for covenant renewal delivered on the Steppes of Moab (Deut 31:9–13), M. N. van der Meer, Formation and Reformulation: The Redaction of the Book of JoshuaintheLightoftheOldestTextualWitnesses (VTSup 102; Leiden: Brill, 2004), 503–4; H. N. Rösel, Joshua (HCOT; Leuven: Peeters, 2011), 131–39. E. C. Ulrich, “4QJoshuaᵃ and Joshua’s First Altar in the Promised Land,” in NewQumranTextsandStudies (ed. G. J. Brooke; STDJ 15; Leiden: Brill, 1994) 89–104; idem, “4QJoshᵃ (Pls. XXXII–XXXIV),” in QumranCave4:Vol.9,Deuteronomy,Joshua, Judges,Kings (ed. E. C. Ulrich, F. M. Cross, et al.; DJD 14; Oxford: Clarendon, 1995), 143–52; idem, “Joshua’s First Altar in the Promised Land,” in TheDeadSeaScrollsand theDevelopmentalCompositionoftheBible (VTSup 169; Leiden: Brill, 2015) 47–65. The sequence in 4QJoshᵃ may find a reflex in the testimony of Josephus (Ant. 5.16–20), although Josephus’s work also includes a later episode, situated at Shechem (Ant. 5.68–70), that self-reflectively refers back to the instructions of Deuteronomy 27 (Ant. 4.305–308).

115

ALTARED STATES

4QJoshuaᵃ

MT Joshua

Josh 8:34

‫]בספר [ התורה‬

‫ואחרי כן קרא את כל דברי‬ ‫התורה הברכה והקללה ככל‬ ‫הכתוב בספר התורה‬

‫לא היה דבר מכל צוה משה‬ ‫] את יה[ושוע אשר לא קרא‬ ‫יהשע נגד כל ]ישראל בעברו‬ ‫[את הירד]ן [והנשים והטף‬ ‫והג]ר[ ההולך בקרבם‬

‫לא היה דבר מכל אשר צוה‬ ‫משה‬ ‫אשר לא קרא יהושע נגד כל‬ ‫קהל ישראל והנשים והטף‬ ‫והגר ההלך בקרבם‬

Josh 8:35

[ -- ] ‫ [ל‬-- ] ‫אחר אשר נתקו‬ ‫את ספר התורה אחר כן] [ל‬ [ --] [--] ‫◦ נושאי הארון‬ ‫]בעת [ ההיא אמר יהוה‬ ‫אליהש]ע ע[ש]ה לך חרבות‬ ‫צרים[ ]ושוב מל את בני‬ [. ‫ישראל‬

Versification

x:x47

‫בעת ההיא אמר יהוה אל‬ ‫יהושע עשה לך חרבות‬ ‫צרים ושוב מל את בני‬ ‫ישראל שנית‬

Josh 5:2

As the table indicates, the Israelites celebrate a series of rites, when they enter the promised land, recalling one of the commands given by Moses (‫ ;והיה ביום אשר תעברו את הירדן אל הארץ‬Deut 27:2). That much seems clear, but many other issues about 4QJoshuaᵃ are contested. In Ulrich’s view, the DSS fragments indicate that the Israelites constructed an altar and celebrated the rites called for in Deuteronomy’s Mt. Ebal/Mt. Gerizim legislation as soon as the Israelites entered the land. The evidence of 4QJoshuaᵃ would present, therefore, an alternate scenario to those presented in MT and LXX Joshua. Indeed, in Ulrich’s view, 4QJoshuaᵃ represents a third Joshua recension, historically antecedent to those of MT and LXX Joshua. But, to complicate matters, the fragmentary evidence supplied by 4QJoshuaᵃ does not necessarily include the construction of an altar.48 Van der Meer argues, in fact, that the reuse of Josh 8:30–35 in this part of 4QJoshuaᵃ was limited to vv. 32, 34–35. In this theory, 4QJoshuaᵃ never reported any altar construction following the Israelite crossing of 47

48

Ulrich interprets these portions of fragment 1 (2:2–4) of 4Q47 as “Josh 5:0,” “Joshua’s First Altar in the Promised Land,” 50–52. E. Tov views the same textual lemma as a modified version of Josh 4:18, “The Literary Development of the Book of Joshua,” in Textual Criticism of the Hebrew Bible, Qumran, Septuagint: Collected Essays (VTSup 167; Leiden: Brill, 2015) 132–53. The reconstruction in my table cannot do full justice to the linear arrangements, uncertain letter forms, and spacing found within the Qumran fragments. Readers are urged to consult the editioprinceps, Ulrich, “4QJoshᵃ,” 147 (Pl. XXXII). Van der Meer, Formation, 511–14.

116

GARY N. KNOPPERS

the Jordan River. Quite the contrary, fragments 1–2 of 4QJoshuaᵃ (4Q47) only mentioned Torah inscription upon the stones (Josh 8:32) and Torah recitation (Josh 8:34–35) after the priests had emerged from the midst of the Jordan River (Josh 4:18).49 As for the directives of Deut 27:4–8, mandating the construction of an altar and the offering of public sacrifices, their fulfillment would ostensibly have been recounted in (the now lost) 4QJoshuaᵃ 8:30–35, in accordance with the sequence found in the MT.50 Tov agrees with some of van der Meer’s analysis, but limits the scope of the relevant 4QJoshᵃ fragments even further to vv. 34–35.51 He thinks that these lines in 4QJoshuaᵃ were followed by a rewritten version of Josh 4:18, relating to the crossing of the Jordan (see x:x above), and the enigmatic tale about Joshua’s circumcision of the Israelites (Josh 5:2–7). Inserted into and adapted to Israel’s crossing of the Jordan River (Josh 4:18), the public recitation of the Torah (Josh 8:34–35) underscores the foundational importance of the Torah and implies a likeness between Moses and Joshua.52 Yet, a question may be raised whether Tov’s reconstruction requires some further scrutiny. In so far as the text of Josh 8:34–35 has parallels in Deuteronomy, those parallels most strongly align with Deut 31:9–13, which mentions “women, the child, and your alien” (‫ )והנשים ו הטף וגרך‬as part of the gathered national assembly commanded to listen to the recitation of “this Torah.” If so, the authors of 4QJoshuaᵃ would be sequencing the timing of Deut 27:2, when the Israelites enter the land, the Torah’s public recitation by Joshua (Josh 8:34–35), in conformity with the terms used in the demand for a national assembly (Deut 31:9–13), and the mass circumcision of Israelite males by Joshua (5:2ff.).53 The Qumran text would thus link Joshua’s circumcision of the Israelites to the commands of the Torah recited within the corporate assembly, but in so doing presents a sequence at variance with the demands of Deut 27:1–26, 31:9–13, and the pattern of Josh 8:30–35.54 In so far as 4QJoshuaᵃ would have any parallels at all with the demands of Deut 11:29–30 and Deut 27:1–26, the 49 50

51 52

53 54

Van der Meer, Formation, 515–19. Because there would be no Josh 8:33 in this text, mentioning Mt. Gerizim and Mt. Ebal, van der Meer’s reconstruction separates any geographical link between Gilgal and the area of Mt. Gerizim and Mt. Ebal in 4QJoshuaᵃ at this point in the narrative. Tov, “Literary Development,” 147–53. In this reconstruction, the recitation of the Torah occurs, while the Israelites are crossing the Jordan, Tov, “Literary Development,” 148. But the timing in Deut 27:2 is explicitly linked to constructing a public stele in the land. The national assembly demanded every seventh year in the land is to occur at “the place” of God’s own choosing (Deut 31:9–13).

ALTARED STATES

117

parallels would be limited to timing (Deut 27:2) and the verbal pronouncement of the blessing and the curse (Josh 8:34).55 My point in mentioning the different theories about the original nature and extent of 4QJoshuaᵃ is not to decide among them. Rather, it is to suggest that even if one accepts van der Meer’s reconstruction, the text of 4QJoshuaᵃ represents a two stage fulfillment of the directives of Deut 27:2–8. If one accepts Tov’s reconstruction, the text of 4QJoshuaᵃ borrows slightly from Deut 27:2 and incorporates Josh 8:34–35 to preface Josh 4:18 and Josh 5:2ff. Both scenarios are unparalleled in MT Joshua, LXX Joshua, and Josephus (Ant. 5.16–20). That is, the Israelites would be conducting one set of rites, upon entering the land and a second set of rites upon their arrival in the vicinity of Shechem.56 Briefly surveying the evidence presented in MT, LXX, and the DSS fragments of 4QJoshuaᵃ indicates, then, the presence of ongoing discussions among Judean literati during the last centuries BCE about the time and place of Israel’s fulfillment of Deuteronomy’s Mt. Gerizim/ Mt. Ebal legislation.

IV. PROGRESSIVE ALTAR TECHNOLOGIES: A WAY OUT OF THE IMPASSE? We have seen that a number of arguments and counter-arguments could be advanced both in favor of and in refutation of the proposition that the similarity between the Covenant Code altar stipulations and the Mt. Gerizim altar stipulations points toward the Mt. Gerizim altar being the central altar at the place of God’s own choosing repeatedly alluded to in Deuteronomy. Judean and Samarian interpreters could thus appropriate the centralization instructions in very different ways. Is there a way out of this seeming impasse? Some scholars maintain that the type of unhewn stone construction depicted in Exod 20:24–25 and reworked in Deut 27:5–7 is to be contrasted with the cut-stone construction of the Jerusalem Temple.57 In this reading of the Pentateuch and the Former Prophets, the call for an exclusive centralized altar in Deuteronomy is simultaneously code for a metamorphosis in the nature of the altar itself. The centralized altar 55

56

57

The timing in Deut 11:29, “when you enter the land” (‫)והיה כי יביאך יהוה אלהיך אל הארץ‬ is more vague than that of Deut 27:2. G. N. Knoppers, “The Sacred Sites of Gilgal, Mt. Gerizim, and Mt. Ebal and their Sacred Rites: Evidence from the Masoretic Text, the Samaritan Pentateuch, and the Dead Sea Scrolls,” in TheSamaritansandTheirPentateuch(ed. C. Nihan; Winona Lake, IN: Eisenbrauns, forthcoming). E.g., G. Fohrer, HistoryofIsraeliteReligion (Nashville: Abingdon, 1972) 62–63.

118

GARY N. KNOPPERS

legislation of Deuteronomy 12 is implicitly aligned with the arrangements made at the Jerusalem Temple, while the Covenant Code and Mt. Gerizim/ Mt. Ebal legislation are aligned with the stone altar constructed by Joshua. Whereas the Mt. Gerizim/Mt. Ebal altar is considered to be an example of an early, basic, temporary, or rural sacrificial installation, the sanctuary altar is considered to be a later, finely-crafted, permanent, or urban installation. Indeed, one recent scholar has argued that the writer of Exod 20:24–26 has the legislation of Deuteronomy 12 already in view, casting “Israel’s worship in the desert period in the tradition of the patriarchs to underscore its temporary and transient nature. After use, the nature altar reverts to nature and can never be a competitor of YHWH’s definite abode.”58 By contrast, the existence of well-dressed ashlar masonry is attested at a variety of ancient Israelite Iron II urban sites.59 The claim about a progression in altar design technologies deserves, however, a thorough review and assessment. In what follows, I shall begin with the literary basis for the theory, before questioning crucial aspects of the theory. In my conclusions, I shall return to the commonalities and contrasts between the Samarian and Judean positions. The debates about altar types illumine, in my view, the particular characteristics of the centralization legislation shared by Samarians and Judeans. The literary basis for speaking of a major transformation in the composition of major sacrificial installations resides in the claim that King Solomon “ordered that large, costly stones (‫ )אבנים גדלות אבנים יקרות‬be quarried (‫)ויסעו‬ to establish the foundation of the temple with hewn stones” (‫;גזית אבני‬ 1 Kgs 5:31).60 The assertion about the deployment of expensive, well-cut stones stresses the monumental and sophisticated nature of the royal undertaking.61 58 59

60

61

Houtman, Exodus, 103. A. Mazar, ArchaeologyoftheLandoftheBible,10,000–586BCE (ABRL; New York: Doubleday, 1990) 379–87, 406–15, 471–75; R. Reich, “Building Materials and Architectural Elements in Ancient Israel,” in TheArchitectureofAncientIsrael:Fromthe PrehistorictothePersianPeriods (ed. A. Kempinski and R. Reich; Jerusalem: Israel Exploration Society, 1992) 1–16 (3–5); idem, “Palaces and Residencies in the Iron Age,” in TheArchitectureofAncientIsrael, 202–30 (210–12). See also 1 Kgs 7:9–11, which relate the use of valuable, cut stones in the construction of Solomon’s palace and related public buildings. In the singular (‫ ;אבן יקרה‬cf. Akk. abnu aqartu), see 2 Sam 12:30; 1 Chr 20:2. Yet, the evidence is hardly straightforward and univocal. Another (probably later) text in Kings, unparalleled in Chronicles, portrays a more nuanced picture, namely that when the Jerusalem Temple was constructed, it was built with “quarried whole (undressed) stone (‫ ;)אבן־שלמה מסע‬neither hammer nor ax nor any iron tool (‫ )כלי ברזל‬was heard, while the temple was being built” (1 Kgs 6:7). This passage is evidently written with Deut 27:5–7 in mind, because the text in Deuteronomy decrees that ‫לא תניף עליהם ברזל‬, “Do not wield

ALTARED STATES

119

Similarly, in the later work of Chronicles, when David makes initial preparations for the temple to be built by his divinely-elected son, he “gave orders to gather the resident aliens in the land of Israel and appointed masons to cut hewn stones (‫ )גזית לחצוב אבני‬to build the House of God (1 Chr 22:2).62 In a private speech, he subsequently informs his son Solomon that he has not only provisioned tremendous amounts of gold, silver, bronze, iron, wood, and stones for the House of God, but also innumerable artisans, including masons (‫)חצבים‬, stone cutters (‫)חרשי אבן‬, and carpenters (1 Chr 22:14–15).63 The Hewn Stone Construction of the Davidic-Solomonic Temple in Chronicles Translation And David gave orders to gather the resident aliens in the land of Israel and appointed masons to cut hewn stones to build the House of God. In my humility, I have provided for the House of God – 100,000 talents of gold and 1,000,000 talents of silver, as well as bronze and iron beyond weighing, because there was so much of it. I have also provided wood and stones and you will add to them. And with you in abundance are artisans – masons, stone cutters, and carpenters – with every kind of skill in every kind of work.

1 Chronicles 22:2, 14–15 ‫ויאמר דויד לכנוס את‬ ‫הגרים אשר בארץ ישראל‬ ‫ויעמד חצבים לחצוב אבני‬ ‫גזית לבנות בית האלהים‬ ‫והנה בעניי הכינותי לבית‬ ‫יהוה זהב ככרים מאה אלף‬ ‫וכסף אלף אלפים ככרים‬ ‫ולנחשת ולברזל אין משקל‬ ‫כי לרב היה ועצים ואבנים‬ ‫הכינותי ועליהם תוסיף‬ ‫ועמך לרב עשי מלאכה‬ ‫חצבים וחרשי אבן ועץ וכל‬ ‫חכם בכל מלאכה‬

This text, like the earlier text in Kings, quite readily admits the employment of well-dressed ashlar masonry in the construction of the Jerusalem temple. Indeed, given the considerable expense involved in undertaking such monumental building construction, they promote David and Solomon’s munificence.64 Seen from this perspective, the arrival of the Jerusalem Temple signals a new and authoritative state of affairs. The Covenant Code provisions purportedly represent a transient and provisional period, whereas

62 63

64

(an) iron (tool) upon them” and ‫אבנים שלמות תבנה את מזבח יהוה‬, “(Of) whole stones you will build the altar of Yhwh.” On the grammatical construction, see further M. J. Mulder, 1Kings,1:1Kings1–11 (HCOT; Leuven: Peeters, 1998) 243–44. The material is unique to Chronicles, Knoppers, IChronicles10–29, 765–88. Cf. 2 Chr 2:6; Josephus, Ant. 7.336; G. N. Knoppers, IChronicles10–29 (AB 12A; New York: Doubleday, 2004), 777–79. V. Fritz, 1and2Kings:AContinentalCommentary (Minneapolis: Fortress, 2003) 64.

120

GARY N. KNOPPERS

the centralization provisions in Deuteronomy represent a permanent and definitive development.65 If so, one could contend that the unfinished stone altars mandated by the Covenant Code and the Mt. Gerizim legislation are both relics of an older era, whereas the (presumably) finished stone altar tradition of the Jerusalem temple influenced the formulation of the Deuteronomic centralization laws.66 Yet, there are three debilitating flaws in the aforementioned interpretation of the literary remains that need to be discussed. First, Deuteronomy’s centralization prescriptions mandate the establishment of an altar at the place of God’s own choosing, but do not mandate a specific altar type. In this respect, the Deuteronomic call for the monopolization of animal sacrifice at one unnamed site differs both from the Covenant Code legislation mandating an unfinished stone altar and from the Priestly legislation mandating a bronze altar (Exod 27:1–7). The two altar references in Deut 12:27 are the only explicit references to the main Yahwistic altar in the entire centralization legislation (Deut 11:31–13:1). The legal draft leaves the manner and mode of altar construction open. The only other mention of altars within the centralization directives pertains to those the Israelites are to eliminate from the land: “And you will surely eradicate all of the places (‫)אבד תאבדון את כל המקמות‬, at which the nations serve their gods … and you will tear down their altars” (‫;ונתצתם את מזבחתם‬ Deut 12:2–3). In calling for the elimination of rival altars in the land and in specifying that the one altar at the divinely elect site is the “altar of Yhwh your God” (Deut 12:27), the writers do not discuss altar types. 65 66

See Houtman, Exodus,98–103 (with further references). Admittedly, the relationship among the Covenant Code legislation, Deuteronomy, and the history of the Jerusalem temple altar can be conceptualized in yet other ways. In N. Lohfink’s view, allowing a Yahwistic altar to be made of unhewn stones may be understood as a reaction to the hegemony of the hewn altar tradition of the Jerusalem temple, “Zur deuteronomistischen Zentralisationsformel,” Bib 65 (1984) 297–328 (repr. in his StudienzumDeuteronomiumundzurdeuteronomistischenLiteratur [SBAB 121; Stuttgart: Katholisches Bibelwerk, 1991] 125–42). Hence, Lohfink views the Covenant Code provisions as a response to a preexisting historical precedent. The theories of Van Seters (n. 11 above) and C. Levin offer yet another interpretive possibility. They regard the Covenant Code regulations as appropriate for conditions in the early Judean Diaspora, DieVerheissungdesneuenBundes:inihremtheologiegeschichtlichenZusammenhang ausgelegt (FRLANT 137; Göttingen: Vandenhoeck & Ruprecht, 1985) 96. Van Seters and Levin thus view the Covenant Code stipulations as an adaptation to the international realities of Judean life in the Neo-Babylonian and early Persian eras. Both views assume much the same point, namely that the Covenant Code provisions contrast with the Temple altar’s mode of construction.

121

ALTARED STATES

Second, the supposition that the altar built in the courtyard of the First Temple conforms to the fine-ashlar type requires some criticism. While the authors of Kings speak of Solomon’s employment of well-dressed stones in the construction of the sanctuary and palace, they do not attend to the mode of construction employed in the establishment of the sacrificial altar. Indeed, they do not mention Solomon’s construction of a main altar, unless such a mention was inadvertently lost by haplography.67 There is, however, one mention of the temple altar in the lightly Priestly-edited account of Solomon’s temple dedication, when the text speaks of Solomon’s offering sacrifices upon “the bronze altar” (1 Kgs 8:64).68 Yet, even so, the reference is indirect; it does not go on to discuss the precise makeup of the bronze altar. The Central Altar in Use (1 Kgs 8:64 ∥ 2 Chr 7:7) Translation On that day the king consecrated the centrecourt, which was before the House of Yhwh, for he offered there the burnt offering, the cereal offering, and the fatty pieces of the well-being offerings, because the bronze altar, which was before Yhwh, was too small (to receive) the burnt offering, the cereal offering, and the fatty pieces of the well-being offerings.

67

68

1 Kgs 8:64 ‫ביום ההוא קדש המלך את‬ ‫תוך החצר אשר לפני בית‬ ‫יהוה כי עשה שם את העלה‬ ‫ואת המנחה ואת חלבי‬ ‫השלמים כי מזבח הנחשת‬ ‫אשר לפני יהוה קטן מהכיל‬ ‫את העלה ואת המנחה ואת‬ ‫חלבי השלמים‬

The relevant texts in Kings (1 Kgs 8:22, 31, 54, 64; 9:25; 2 Kgs 12:10; 16:10–16; 23:9), relating to the main altar do not specify its precise mode of construction. On the lack of such description, see V. A. Hurowitz, “Yhwh’s Exalted House – Aspects of the Design and Symbolism of Solomon’s Temple,” in Temple and Worship in Biblical Israel (ed. J. Day; LHBOTS 422; London: T. & T. Clark, 2005) 63–110 (67, 77–78), who notes that in 2 Chr 4:1 the bronze altar is described at some length and may preserve a detail missing from 1 Kgs 7:23, lost through haplography (homoioarkton from ‫ ויעש‬to ‫)ויעש‬. So also K. Hognesius, TheTextof2Chronicles1-16 (ConBOT 51; Stockholm: Almqvist & Wiksell, 2003) 88; R. W. Klein, 2 Chronicles (Hermeneia; Minneapolis: Augsburg Fortress, 2012) 60. The argument has some merit, because later references within Kings (e.g., 2 Kgs 16:10–16) assume that the altar is there. Yet, it is doubtful that the Chronicler’s Vorlage contained the stupendous dimensions of the main altar found in 2 Chr 4:1 (see below). These would seem to be a typical Chronistic flourish. On the Priestly-style additions attested in 1 Kings 8, such as the incorporation of the “Tent of Meeting” into the temple’s “Holy of Holies” (vv. 4, 6), see M. Noth, Könige1:I.Könige1–16 (BK IX/1; Neukirchen-Vluyn: Neukirchener Verlag, 1968) 168–93.

122

GARY N. KNOPPERS

Translation

2 Chr 7:7

And Solomon consecrated the centrecourt, which was before the House of Yhwh, for he offered there the burnt offerings and the fatty pieces of the well-being offerings, because the bronze altar that Solomon had made was not able to hold the burnt offering, the cereal offering, and the fatty pieces.

‫ויקדש שלמה את תוך‬ ‫החצר אשר לפני בית יהוה‬ ‫כי עשה שם העלות ואת‬ ‫חלבי השלמים כי מזבח‬ ‫הנחשת אשר עשה שלמה‬ ‫לא יכול להכיל את העלה‬ ‫ואת המנחה ואת החלבים‬

Third, although Chronicles depicts David’s investment in fine ashlar masonry for the planned temple at some length, the work also speaks of the construction of a gigantic altar, 20 × 20 cubits in length and width and 10 cubits high (2 Chr 4:1). Yet, the work distinguishes between the construction of this sacrificial installation and that of the temple itself. Chronicles overtly categorizes the new structure as a bronze altar (‫)מזבח נחשחת‬, rather than as a stone altar.69 The Central Altar Construction in Chronicles Translation He made a bronze altar; 20 cubits was its length, 20 cubits was its width, and 10 cubits was its height.

2 Chronicles 4:1 ‫ויעש מזבח נחשת עשרים אמה‬ ‫ארכו ועשרים אמה רחבו ועשר‬ ‫אמות קומתו‬

I would argue that the Chronistic narrative attempts to adjust a particular feature of his Vorlage, namely a bronze altar (cf. Exod 27:1–7), to the writer’s grandiose portrait of the Davidic-Solomonic temple.70 Certainly, the deliberate archaizing of temple provisions, features, and furnishings in accordance with select features of the Priestly tabernacle is well-attested elsewhere within the temple construction story told in Chronicles.71 The archaizing is only partial, however, because the Chronistic altar is some 53 times larger in volume than the tabernacle altar.72 Whatever one might 69 70

71

72

The relevant texts are 2 Chr 4:1; 5:12; 6:12, 22; 8:12; 15:9; 29:18–19, 21; 33:16; 35:16. To be sure, the depiction of the Priestly bronze altar (Exod 27:1–7) may have been based to some measure, as some have argued, on the Jerusalem temple altar. In any case, the Chronicler will not have failed to notice that the basic altar type found in his Vorlage was the type attested in one of his beloved Pentateuchal sources. J. Van Seters, “The Chronicler’s Account of Solomon’s Temple-Building: A Continuity Theme,” in The Chronicler as Historian (ed. M. P. Graham, K. G. Hoglund, and S. L. McKenzie; JSOTSup 238; Sheffield: JSOT Press, 1997) 283–300. Given that the Chronicler wrote at a time in which the temple at Mt. Gerizim had been extant for some time, his archaizing of critical dimensions of the First Temple may have

ALTARED STATES

123

all say about this gigantic sacrificial installation, it seems safe to conclude that it was not designed, unlike the Priestly altar, to be portable.73 It also seems apt to recognize the distinction that the work itself makes between the composition of the main altar and the composition of the temple building.

CONCLUSIONS We have seen that the available literary evidence does not fall into any single clear and discernable pattern. Analysis of the textual evidence found within the Jewish and Samaritan Pentateuchs, Joshua, 4QJoshuaᵃ, Kings, and Chronicles undermines the supposition that the writers envisioned some sort of progression from primitive (or rural) altars made from natural fieldstones to advanced (or urban) altars made from well-dressed ashlar masonry.74 Yet, analysis of the literary evidence also results in a greater appreciation of the subtlety and complexity of the centralization legislation in

73

74

been deliberately undertaken with the northern sanctuary in view. In this scenario, it may not have been enough to posit continuity between the First and Second Temples, although this argument is implicitly present throughout the Chronistic narratives. Rather, an additional, even more far-reaching literary appeal needed to be made to the arrangements introduced at Mt. Sinai and on the Steppes of Moab. If Samarian altar traditions appealed to the Covenant Code (Exod 20:24–26) and the Mt. Gerizim/Mt. Ebal legislation (Deut 11:29–30; Deut 27:1–26), the Chronistic Judean altar traditions appealed to the Priestly-style legislation. In this environment in which both temples claimed critical connections with the (imagined) Israel of old, each sought to buttress its position by recourse to prestigious antecedent. On the Chronistic propensity to deploy large and stylized numbers to magnify what the work fashions as Israel’s classical past, especially the united monarchy, see my I Chronicles10–29, 569–71. The monumental altar of Ezekiel’s visionary program, consisting of a succession of square, tiered levels of progressively smaller size, resting upon a platform base with its steps (‫ )מעלתהו‬leading down toward the east (Ezek 43:13–17), is an interesting exception to earlier patterns in Exodus, Deuteronomy (27), Joshua, and Kings, and may be a reaction to them. The burnt offerings and offerings of well-being (Ezek 43:27) to be presented upon the altar hearth (‫ ;האראיל‬Ezek 43:16) correspond, however, to earlier precedent. See further J. D. Levenson, TheologyoftheProgramofRestoration ofEzekiel40–48 (HSM 10; Missoula, MT: Scholars Press, 1976) 38–43; D. I. Block, TheBookofEzekiel,Chapters25–48 (NICOT; Grand Rapids: Eerdmans, 1998) 531; S. S. Tuell, TheLawoftheTempleinEzekiel40–48 (HSM 49; Atlanta: Scholars Press, 1992) 46–51; J. Milgrom and D. I. Block, Ezekiel’sHope (Eugene, OR: Cascade, 2012, 122, 214–18. In Zevit’s theory (Religions, 291–92), the new altar ordered by King Ahaz, following a design seen in Damascus, served as a historical precedent for the design of Ezekiel’s altar. Nevertheless, the composition and dimensions of Ahaz’s new altar are not addressed in 2 Kgs 16:10–16.

124

GARY N. KNOPPERS

Deuteronomy, whose writers leave the particular type of altar construction unspecified. Of the four major altar laws found within the Pentateuch, there is only one – the centralization legislation – that does not prescribe a particular mode of altar construction. This is true, as we have seen, for both the Jewish and the Samaritan Pentateuchs. Such an anomaly hardly seems accidental, considering how much attention the writers devote to revising the feasts, festivals, juridical institutions, and the like in light of centralization.75 Of great concern in the centralization statutes is the establishment of cultic purity and cultic unity, but not a particular style of altar manufacture.76 Presumably, the writers and redactors of Deuteronomy could have dictated a particular type of construction to clarify matters, if they had wished to do so, given the pattern of other altar drafts.77 They certainly do provide directions about a number of other matters in the centralization legislation itself, including the provision of secular slaughter (Deut 12:14–16, 20–25). Yet, in the matter of altar design, they chose to leave the matter open. Such a vague formulation permits multiple interpretations, including an altar of unhewn stones, an altar of finely-finished stones, and even a quasi-Priestly bronze altar, as one finds in Chronicles. 75

76

77

N. Lohfink, “Die Sicherung der Wirksamkeit des Gotteswortes durch das Prinzip der Schriftlichkeit der Tora und durch das Prinzip der Gewaltenteilung nach den Ämtergesetzen des Buches Deuteronomium (Dt 16,18–18,22),” in StudienzumDeuteronomium und zur deuteronomistischen Literatur I (SBAB 8; Stuttgart: Katholisches Bibelwerk, 1990) 305–23; E. Otto, “Von der Programmschrift einer Rechtsreform zum Verfassungsentwurf des Neuen Israel: Die Stellung des Deuteronomiums in der Rechtsgeschichte Israels,” in BundesdokumenutundGesetz:StudienzumDeuteronomium (ed. G. Braulik; HBS 4; Freiburg: Herder, 1995), 92–104; idem, Deuteronomium12–34, 1: 1417–1503; G. N. Knoppers, “The Deuteronomist and the Deuteronomic Law of the King: A Reexamination of a Relationship,” ZAW 108 (1996) 329–46; idem, “Rethinking the Relationship between Deuteronomy and the Deuteronomistic History: The Case of Kings,” CBQ 63 (2001) 393–415; B. M. Levinson, “The Reconceptualization of Kingship in Deuteronomy and the Deuteronomistic History’s Transformation of Torah,” VT 51 (2001) 511–34. This is also true of the reference to an altar in Deut 16:21. The legislation prohibits planting “an asherah of any kind of tree next to the altar of Yhwh your God, which you will make for yourself” (‫)אצל מזבח יהוה אלהיך אשר תעשה לך‬. Yet, the text does not address any particular style or mode of construction for the altar. On the particular kind of asherah in question, see S. M. Olyan, Asherah and the Cult of Yahweh in Israel (Atlanta: Scholars Press, 1988) 9. If, as seems likely, the centralization legislation has a long history of literary development, extending back into the preexilic period and perhaps continuing in the Persian period (see, e.g., T. C. Römer, TheSo-calledDeuteronomisticHistory:ASociological, Historical,andLiteraryIntroduction [London: T. & T. Clark, 2005] 56–65), the lack of such specificity is all the more remarkable. The editors of Deuteronomy and the Pentateuch had numerous opportunities to elaborate upon the altar instructions in Deut 12:27, but chose not to do so.

ALTARED STATES

125

Comparison of the Pentateuch’s altar laws in the Samaritan and Jewish Pentateuchs provides, therefore, further evidence that Deuteronomy, in particular, and the Pentateuch, in general, functioned in the Persian and Hellenistic period as a compromise document. The Pentateuch was not simply a Judean project nor simply a Samaritan project, but rather a literary work that bound both groups together. In such a literary corpus, the altar legislation within the centralization decrees could be read by each community with its own sacred traditions and institutions in view. The fundamental concern was, therefore, not Samaria perse or Judah perse, but the larger Israel of which they both were part.

A STEMMA FOR DEUTERONOMY Benjamin ZIEMER

1. THE USE OF THE

STEMMATIC METHOD

1.1. There has to be a stemma for a book like Deuteronomy All manuscripts of the Book of Deuteronomy we know are genetically related: They agree in the general order of chapters, sections and verses. They share most words, their order and their form. They have in common many idiosyncrasies which disagree to other works: In contrast to the book of Leviticus or the Temple Scroll, the law is not spoken by God but by Moses. In contrast to the book of Genesis or the book of Jubilees, the patriarchs are not told to build altars or to anoint stelae. In contrast to the book of Kings or the Halakhic Letter 4QMMT, the chosen place or chosen city is not explicitly called by name. All these manuscripts have so much in common that we have no choice but to speak of one big family of manuscripts, with the family name “Deuteronomy.” Therefore there has to be a family tree, a stemmafor Deuteronomy, notwithstanding the doubts about the possibility to reconstruct it. The family of Deuteronomy manuscripts has got other relatives: e.g., the book of Leviticus contains a list of clean and unclean animals very close to that in Deut; the books of Exodus and Numbers bear close parallels too.1 So called 4QDeutq, containing an archaic version of the Song of Moses, can even be understood as a late witness to the age-long separate transmission of one of the sources of Deuteronomy.2 The tefillinand 1

2

Gen, Ex, Lev and Num manuscripts are excluded from my analysis, despite the fact that both in the Samaritan and in the Masoretic tradition the Deuteronomy manuscripts are actually Pentateuchal manuscripts. For understanding the genetical relationship between Deut and Lev or Ex it is important to compare the original text of Deuteronomy since this relationship is of a literary nature, cp. Ziemer, Prophetenrede. However, in the parallel passages, like the Decalogue (Deut 5) or the list of clean and unclean animals (Deut 14), there is danger of contamination in the textual history. Therefore my argument will not be based on these chapters. 4QDeutq is a case of special interest because its theologically archaic version of Deut 32:43 is very close to a tradition intervowen by 𝔊 with the less archaic version transmitted by 𝔐 and ⅏. I will come back to this issue later.

128

BENJAMIN ZIEMER

mezuzotfrom Qumran as well as the Papyrus Nash and excerpted texts3 like 4QDeutʲ, 4QDeutk1, 4QDeutn, 5QDeut and 4QTestimonia, are other relatives. These texts seem to be clearly dependent on the book of Deuteronomy and could be placed theoretically in the “family tree,” the stemmacodicum. But they are strongly influenced by liturgical tradition, they are eclectic in character, sometimes they seem to combine several Vorlagen, and they show a lot of memorial variants. So they cannot serve to establish the stemma. Since Deuteronomy has been one of the most appreciated books in Qumran and elsewhere in the societies of Achaemenid and Hellenistic Judah and Samariah, there are many other works dependent on Deuteronomy, like the Temple Scroll, CD, or 4QMMT. Since they all combine several sources they too have to be treated with caution. Be it as it may, enough actual manuscripts of Deuteronomy remain to draw a stemmacodicum. I will focus especially on 4QDeutᵇ, 4QDeutᶜ and 4QpaleoDeutʳ. 4QDeutᵇ and 4QDeutᶜ stem from the 2nd c. BCE and are therefore two of the oldest manuscripts of Deuteronomy. Among the oldest Deuteronomy manuscripts they are the best preserved ones. 4QpaleoDeutʳ is the largest preserved paleo-Hebrew Deuteronomy manuscript, it is therefore also of interest for the early textual history of Deuteronomy. If these manuscripts were the only Deuteronomy manuscripts we possess one could say nothing about their relationship. They overlap only in some single words where they agree one with the other.4 These three manuscripts together cover only 1000 words of a book which contains more than 14000 words in the Masoretic or the Samaritan text. Therefore, if we are interested in the earliest textual history we will have no choice but to start with the only three extant independent witnesses for the wholebook of Deuteronomy: 𝔐, ⅏ and 𝔊. Although there is a plurality of medieval manuscripts for each of these three groups, they are but threewitnesses since for the stemmaof Deuteronomy only their earliest predecessors are of interest. First, the Masoretic text is transmitted, as generally known, very homogenously. It is now clear that we can be sure its letters go back at least to the first c. BCE, shown by manuscripts like MasDeut and 4QDeutᵍ. Hence we can take the letters of the medieval Masoretic text (𝔐)5 as if it were a complete Deuteronomy manuscript from the first c. BCE. 3 4

5

Cp. Duncan, Excerpted Texts. The first ‫ ביום ההוא‬in Dtn 31:17 is present both in 4QDeutᵇ and 4QDeutᶜ; three parts of words in Dtn 28:20 are present both in 4QDeutᶜ and 4QpaleoDeutʳ. “𝔐” means the agreement of the main Tiberian manuscripts, their Masorot and the t.r. See Mordechai Breuer, Biblical Text, for the differences between the textus receptus,

A STEMMA FOR DEUTERONOMY

129

The situation is somewhat similar in the case of the Samaritan Pentateuch. I will take its consonantal framework (⅏), somewhat optimistically, as a witness from the second6 or first c. BCE. ⅏ is not as clearly witnessed in BCE times as the letters of 𝔐. But its tradition is also very homogenous. ⅏ is of utmost value for the stemmabecause it has been transmitted completely independent of 𝔐. Many of its minor and major differences to 𝔐 are attested now in various Qumran scrolls. The third independent witness is the Septuagint, that means the unrevised Old Greek (𝔊). Actually its Hebrew Vorlagehas to be compared with the other witnesses – and this bears some problems. Sometimes the translator took one word for two, sometimes two words for one, sometimes he changed words. So he regularly translated ‫“( בשעריך‬within your gates”), from Deut 12 on, as ἐν ταῖς πόλεσίν σου / ἐν ταῖς πόλεσιν ὑμῶν (“within your cities”),7 and he regularly renders ‫כל המצוה הזאת‬ (“this whole commandment”) as plural πάσας τὰς ἐντολὰς ταύτας (“all these commandments”).8 Therefore we cannot know with certainty if at some place in his Hebrew Vorlage he found ‫ בעריך‬instead of ‫בשעריך‬, or ‫ כל המצות האלה‬instead of ‫כל המצוה הזאת‬. Beyond that, in a lot of minor cases it is impossible to conclude which Hebrew text is translated – for instance if diverse Hebrew witnesses differ in the presence or absence of the accusative marker ‫את‬. Finally it is of course difficult to decide which Hebrew Vorlage lay in front of the translator in the many cases where the wording of the Old Greek itself is doubtful. Nevertheless, in general the Old Greek has been a faithful word-for-word representation of its Vorlage, which differed obviously from 𝔐 and ⅏, hence it can be used as the third main witness. So we have got three completewitnesses for the book of Deuteronomy, contemporary with the Qumran scrolls: The proto-masoretic text, determined by 𝔐, at least of the first c. BCE; the prototype of the Samaritan version of the Pentateuch, determined by ⅏, from approximately the same time, and the 𝔊 Vorlagefrom the third c. BCE.

6 7

8

the Codex L and other ancient manuscripts and masorot. Since these differences are, at least in Deut, solely orthographical, the Masoretic text (𝔐) can be counted as a single witness for all textual variants discussed in this paper. The deviations collected by Kennicott and others from late medieval Hebrew manuscripts include textual variants, but they can be explained generally as typical copyists’ errors. The second century BCE dating has been proposed by Eshel and Eshel, Dating. Deut 12:17-18, 21; 14:21, 27-29; 15:22; 16:11, 14; 17:8; 24:14; 26:12; 28:57; 31:12. Deut 6:25; 11:22; 15:5; 19:9.

130

BENJAMIN ZIEMER

This would constitute a provisional family tree, a stemmacodicum, with the archetype α and three branches:

Į

╱_╲

  Every line in a stemma means dependence in such a way that the manuscript below any line is dependent on the manuscript above the line in direct lineage. Ideally, every reading of a given manuscript included in the stemmais either inherited by its parent text or an innovation of the copyist. In the first case, the reading helps to know the parent text and, finally, to determine the archetype itself. In the second case, the reading reveals the character of the specific manuscript or its exclusive forerunners and helps to understand the textual history. 1.2. The applicability of the stemmatic method for the textual criticism of the Hebrew Bible According to the stemmatic method as described by Paul Maas9 and others for editions of Greek and Latin classical texts, any such line is determined by errors made by the copyists and reproduced by all subsequent scribes. This principle led to two fundamental problems causing this method not to play a big role for studies in the textual history of the Hebrew Bible to date: First, in most cases it is difficult to decide on internal grounds which reading is erroneous and which is correct. And, second, the scribes of Biblical texts are thought to have been able to reverse the deviations found in their Vorlagen, therefore blurring the true relationship between the manuscripts. But both problems can be solved. 1.2.1. Weneednotjudgewhichreadingiserroneousbutlookonagreements The designation of a reading as correct or erroneous, as superior or inferior, is highly subjective and in many cases even impossible to answer. In fact, the three main witnesses of Deuteronomy as well as the Qumran fragments – as far as they are preserved – as a rule present a readable and understandable text. Hence it is more appropriate to speak of “deviations” or “divergences” than of “errors.” By far most deviations may have been 9

Maas, Textkritik.

A STEMMA FOR DEUTERONOMY

131

arisen in both directions. But it is a misunderstanding of the stemmatic method that errors are its main category. In fact, only agreements are essential for defining textual families, i.e., readings sharedby some witnesses against others. This is true whether these shared readings are “original” or “secondary.” Shared readings are, as a rule, inherited, so determining the text of a common ancestor, either the text of the archetype or the innovations of a later parent manuscript of the witnesses with the shared readings. There are two types of exceptions: Firstly, agreements may arise by accident – typical minor deviations like addition or omission of a ‫ו‬-copulativum or a notaaccusativihappen to occur several times independently, or they are reversed by chance. The greater the number of deviations, the greater the certainty to differentiate between the rule (inherited readings) and the exceptions.10 Secondly, there is the possibility of contamination as mentioned already for the category of the “excerpted texts” in Qumran: If a scribe uses more than one written Vorlage or is influenced by another tradition beyond its immediate Vorlage, one has to treat the readings of this manuscript for the stemmawith caution. This leads us to the second objection against the stemmatic method, that contamination would be omnipresent in the Textual history of the Hebrew Bible. But this doesn’t stand up to scrutiny. 1.2.2. Howwidespreadwascontamination? One has to differentiate between the main branches of the textual history of the Hebrew Bible to answer this question crucial for the applicability of the stemmatic method. First, concerning 𝔐, the Jews invented a very effective system to detect and to correct deviations: the Masora.11 Since a scribe was able to reverse the deviations introduced unconsciously by his forerunner, we cannot hope to define family relations between manuscripts by clusters of shared errors. For the most part, the massive amount of deviations collected by Kennicott and de Rossi shows individual errors of the manuscripts later corrected either by the scribes themselves, by subsequent readers, or by the next copyist. Hence, these deviations rather reveal the inclination of a given manuscript to err than any family relationship between manuscripts. Shared errors between certain 10

11

Polak, Statistics, 224: “What counts, is number. Agreements (common variants) are significant, if their number is ‘significantly’ larger than expected.” Cp. Ofer and Lubotzky, Masora.

132

BENJAMIN ZIEMER

manuscripts most often occur by chance, between manuscripts with a high inclination to deviations. That means, on the one hand, the stemmatic method is not applicable at all for the internal relations of the Masoretic manuscripts. But, on the other hand, this has no relevance because the stemmatic method is not necessaryfor the Masoretic text! The few differences between the best Masoretic manuscripts are limited to matters of orthography,12 if not of vocalisation or accentuation. Therefore it is possible to treat 𝔐 as one single witness for the stemma. The situation is similar for ⅏: here, in a relatively narrow community, the oral tradition enables every scribe to detect and correct deviations from the traditional text. In contrast, the limitations of the stemmatic method lie somewhere else for the Septuagint, the third branch of the textual tradition. Especially through the famous work of the Hexapla, but in some cases even before Origen, the original Greek translation has been contaminated either directly or indirectly by a Hebrew version nearly identical with 𝔐. Here, the main task is to exclude the contamination and to reconstruct the Old Greek (𝔊), which is the aim of the Göttingen Septuagint. For this, the stemmatic method is helpful: e.g., if a manuscript of the Old Latin agrees with some Greek manuscripts in a variant which is not imaginable to have been arisen twice independently, such agreement proves the prehexaplaric origin of a certain Greek reading. Hence the reconstructed Old Greek (𝔊), albeit not as certain as 𝔐 and ⅏, can serve as the third main witness for the stemma. Anyway, compared with the Qumran manuscripts, the transmission of the Septuagint is more prone to contamination, because the scribes, at least in the earliest times, knew they had before themselves a translation, and sometimes they corrected towards what they considered to be the correct Hebrew text. So if you have two Greek manuscripts, one agreeing in a certain detail with 𝔐, and one disagreeing with 𝔐, there are always two possibilities: either the agreement is inherited through the original Old Greek, and the disagreement is due to later deterioration, like parablepsis, doublet, assimilation to parallels or any other reason; or the disagreement is inherited through the original Old Greek, and the agreement with 𝔐 came 12

Cp. the lists by Breuer, Biblical Text. Beyond this, there are indeed a lot of minor textual variants in less exact late medieval Hebrew manuscripts, as shown by the collections of Kennicott, de Rossi and Ginsburg. The BHS mentions these variants in cases they agree with some of the ancient versions, and therefore they are mentioned in some of the DJD volumes too. Generally speaking, such agreements indicate that the variant in question may well have been arisen several times independently, by accident. A ‫ו‬-copulativum within a three- or four-membered itemisation or at the beginning of a verse can be added and omitted several times, covering any tracks of dependence.

A STEMMA FOR DEUTERONOMY

133

into existence only later by contamination, e.g., through the Hexaplaric revision. However, such revisions, combining two (or more) separate traditions in order to produce one revised text, were a relatively rare phenomenon in textual history and as a rule restricted to translations.13 In contrast, it can fairly be assumed that the scribes who copied a Hebrewtext of Deuteronomy in BCE times, whether Samaritan or Jewish, were confident to copy the originalversionof this book even though there may have been some minor copying errors. There was no reason for a scribe to conform his Vorlageto another written version as it was the case for Origen who attempted to bring together the traditional Greek translation with the Hebrew text generally accepted in his time. In case a Hebrew scribe had doubts on the authenticity of his Vorlage, he need not mix readings from several manuscripts but would copy the preferred version instead. Emanuel Tov has shown convincingly that there is no empirical evidence for correcting a manuscript of one textual tradition towards another textual tradition in the scribal practice of Qumran.14 Beyond the combination of two or more written manuscripts there is a second possibility of contamination, the influence of a tradition. The textual variants within the Decalogue give a good example for the so-called contamination by memory.15 This fits as an explanation in cases a passage is well known and used in liturgy within a given community. If in Alexandria the children learned the short commandments of the Decalogue in the order “You shall not commit adultery” before “You shall not murder,” it is easy to explain that an Alexandrian scribe who copied the book of Deuteronomy may have changed the order of the commandments in this way. Likewise, in the Samaritan community it is easy to explain that the altar commandment has been introduced in the Decalogue both in Ex 20 and in Deut 5, and the name ‫ הרגרזים‬in Deut 27:4. It is in this way not difficult to understand why the variation between the textual witnesses of the Deuteronomy Decalogue is much larger than anywhere else in Deuteronomy. This is to be explained by contamination! Any scribe of Deuteronomy knew the Decalogue by heart in at least one certain wording. And if it comes to copy this text any scribe would be tempted to conform it to the wording he knows by heart. 13

14 15

Such revisions by comparison of texts must not be confused with the somewhat more usual rewriting or editing, as visible in the 𝔐 versions of Samuel, Jeremiah or Ezekiel, in the 𝔊 versions of Daniel, Esther, or Kings, or in Qumran manuscripts like 4QSamᵃ, 11QPsa or so-called 4QRP. Tov, Scribal Practices, 224. Cp. Brooke, Phylacteries, 59; Duncan, Excerpted Texts, 60.

134

BENJAMIN ZIEMER

The versions of the Decalogue are legion. In the traditional Masoretic text alone no less than six slightly different versions of this text are transmitted. The manuscripts and the editions show two different reading traditions (qere) for each of the two versions, in Exodus and Deuteronomy: the upper and the lower accentuation (‫ טעם עליון‬and ‫)טעם תחתון‬. The written tradition (ketiv) differs from both, concerning the arrangement of the setumotand petuhot, and some cases where the pronounced words do not agree with the written consonants (‫ ֲאד ֹנָ י‬read permanently for the ketiv‫יהוה‬, and ‫וֹתי‬ ָ ‫ ִמ ְצ‬read in Deut 5:10 for the ketiv‫)מצותו‬. It is not possible to draw a useful stemmaof the different versions of the Decalogue to be found in 𝔐, ⅏, the 𝔊 manuscripts, the tefillinand mezuzotfrom Qumran, the Papyrus Nash and the excerpted Deut manuscripts. In contrast, contamination is a very good explanation for the confusion to be found there. Contamination as explanation is almost irrefutable if two versions were conflated. The Sabbath commandment in the Papyrus Nash, in the excerpted manuscript 4QDeutn, and in the first hand reading of 𝔊B respectively conflates the two motivations – God’s creation and Israel’s slavery in Egypt – otherwise known separately from the Exodus and the Deuteronomy version of the Decalogue. This is obviously the result of contamination which may happen more than once in a similar manner. The Decalogue is not the only part of Deuteronomy known and transmitted on its own in antiquity and therefore prone to contamination. The whole sections of Deut 5:1–6:9 and Deut 10:12–11:21 can be found on tefillin and mezuzot, and parts of Deut 8; 30 and 32 are included in excerpted texts.16 The latter, the Song of Moses, is a unique case in transmission history. It is included in the excerpt collections of 4QDeutʲ and 4QDeutk1 and is the only text present in 4QDeutq and in 4QPhylN. In the Greek manuscript tradition it is transmitted at two distinct places, as part of Deuteronomy and as the second of the Odes. So in the Hebrew as well as in the Greek tradition this text had a twofold SitzimLeben: as part of the Book of Deuteronomy, and as part of the liturgical tradition.17 That’s why the wording of this text is especially prone to contamination. I will come back later to this issue, in the section dealing with the 𝔊 version of the Song of Moses (4.12). As a result, there is no reason to assume from the outset that contamination will play a big role if we limit the stemmato the three complete main witnesses, 𝔐, ⅏, and 𝔊, and the actual (i.e., not excerpted) Deuteronomy manuscripts from the Judean desert. 16 17

See DJD XIV, 79. Cp. Thiessen, Form.

A STEMMA FOR DEUTERONOMY

135

It will be shown that virtually all readings shared between these witnesses can be explained without the assumption of contamination, in accordance with a concrete stemmacodicum. Of the agreements, only a small number demands an explanation as accidental, whereas the vast majority can be explained as inherited agreements. 1.3. Methodological and terminological issues Frank Polak names the “stemmatic method” as the only approach which “can provide the foundations on which to base textual criticism.”18 Regrettably it is common practice in textual criticism of the Hebrew Bible to look at every single textual variant independently, as if every scribe at any time had a free choice to choose among all variants extant in any manuscript. But here the perspective of the copyist and the perspective of the textual critic must not be confused. The textualcritichas all readings of all preserved manuscripts at his disposal, but the copyisthad not. So, if three manuscripts agree in a certain reading against a fourth manuscript, in general their three scribes did not choose this reading deliberately but simply copied it from their parent text. The reading of the fourth manuscript can be the reading of the archetype only in specific cases: if there are reasons for the assumption that the first three manuscripts stem from a common ancestor which is secondary in relation to the ancestor of the fourth manuscript, or if the majority reading can be explained by parallel revision (most plausible if the majority readings are only similar, and not identical), by coincidence (plausible only for very small differences or for clear homoioteleuta), or by contamination (plausible only within a community). If none of these explanations fits, the shared reading of the three manuscripts is the original reading – and it is irrelevant whether it is shorter or longer, and whether it seems to be a more or less difficult reading. Finally, it is important to remember that the pure number of deviations cannot be crucial for the localisation of a manuscript in the stemmaor for the value of a given manuscript to reconstruct the archetype. In contrast, only agreements are decisive. 1.3.1. “Independent”readingsin“dependent”manuscripts–somenecessary clarifications Its clear terminology is one of the most serious advantages of the stemmatic method. For a book like Deuteronomy, there hastobea real stemma. 18

Polak, Statistics, 224.

136

BENJAMIN ZIEMER

That means, any Deuteronomy manuscript is connected to any other Deuteronomy manuscript either by direct dependence or by common parent texts. This remains true, regardless of the fact that it is not always possible to determine the exact relationship of all known manuscripts. A fundamental category for drawing a stemmais the question of dependence. But what means “dependent,” and what means “independent”? In Qumran studies, several completely different phenomena are commonly described by these antonyms, easily leading to confusion: the “independence” of manuscripts, the “independence” of scribes, and the “independence” of readings.19 If we speak about the manuscripts within the family of Deuteronomy manuscripts, there has been, as a matter of definition, only onemanuscript absolutelyindependentof all other Deut manuscripts: the archetype itself. It was dependent on other literary works and made use of some immediate Vorlagen. Older, lost versions of the entire book might have been among them. Furthermore, certainly some texts have been transmitted separately for a long time, whether oral or written, like the song of Moses (Deut 32). But these supposed Vorlagenare not Deuteronomy manuscripts in the strict sense of the word. Allother (later) Deuteronomy manuscripts are dependenton their parent texts, ultimately on the archetype, clearly shown by the abundant similarities in structure and wording. Hence, “independence” of manuscripts is always relative and in need of specification. ⅏, 𝔐 and the Old Greek all are independent oneoftheother. The Qumran manuscripts too are independent ofthesethreewitnesses – they can neither be dependent on medieval manuscript traditions (⅏, 𝔐) nor on a Greek translation (𝔊). One may speak of the relative “independence” of a manuscript, if it does not share errors of related witnesses and therefore sparks the interest of a textual critic who aims to reconstruct certain hyparchetypes or the archetype itself. In contrast, if a scribe“maintained a certain level of independence,”20 this is another expression for the inclination of the scribe to deviate from his Vorlage. The higher his inclination to deviations against his Vorlage, the lower the value of his manuscript for the reconstruction of the archetype. To avoid confusion, it would be better to say that a scribe took a certain level of liberty. 19

20

E.g., Frank Polak (Statistics, 241) states in his discussion of the textual relations between 4QSamᵃ, 𝔐 and 𝔊 that “the text of the scroll is slightly more independent” in one certain category of deviations than in others. Tov, Textual Criticism, 3rd ed., 109, FN 190, similarly speaks about “independent features” reflected in 4QSamᵃ. Tov, Textual Criticism, 3rd ed., 103f., cp. 109.

A STEMMA FOR DEUTERONOMY

137

If it comes to mention “independent” readingsin a certain manuscript, the confusion is inevitable. The relative independence of a manuscript denotes a high value, but the relative liberty taken by its scribe denotes a low value of the readings of a given manuscript for the reconstruction of the archetype. We have to bear in mind that every manuscript reproduces, as a rule, the text of his Vorlage, except for its innovations. These innovations may be, one the one hand, deliberate corrections, whether conjectures in an attempt to correct a Vorlageregarded as erroneous, or exegetical or linguistic actualisation or rewriting. These innovations may be, on the other hand, scribal errors, made unconsciously by any scribe more or less often. At the surface, both kinds of deviations include the same types: addition, omission, substitution or reordering concerning letters, words, groups of words, sentences or entire paragraphs. All these deviations have one thing in common which distinguishes them from other elements of the text of a certain manuscript: the innovations are made independentof the immediate Vorlage.21 As a consequence, the “independence” of a reading is nearly the same as the “error” in the classical stemmatic method – a reading which distinguishes a manuscript from its forerunners and, thereby, from all those relatives which are not themselves descendants of the given manuscript. However, in the discussion of the Qumran scrolls, it has become usual to call those readings “independent” which are not attested in ⅏, 𝔐 or 𝔊. But this is misleading, such readings should better be called isolated readings or singularity readings. In case a manuscript contains at least one such singularity reading, one has to reject the hypothesis that ⅏, 𝔐 or 𝔊 are dependent on this specific manuscript. E.g., 4QDeutᶜ has ‫ ה[י̇ רדן‬before [‫ שמה ]לרשתה‬at the end of Deut 4:14, a word not attested for this verse in ⅏, 𝔐 or 𝔊. It can be excluded almost certainly that ⅏, 𝔐 and 𝔊 all omitted the word ‫ הירדן‬from this verse, since they agree in reading ‫ הירדן‬resp. τὸν Ἰορδάνην in the parallel formulations in Deut 4:26 and 31:13.22 It is far more plausible that ⅏, 𝔐 and 𝔊 21

22

This includes changes by contamination that are independent of the immediate parent text but dependent on others. As Polak rightly emphasises, the probability of contamination is relatively high in matters of exegetical revision, and can be neglected in matters of small, synonymic changes which constitute the bulk of variants discussed in this paper. The phenomenon of double renderings, abundant in the textual history of the book of Samuel, does not play a significant role in the textual history of Deuteronomy, besides the Decalogue and the Song of Moses which are influenced by their continued separate transmission, outside of Deuteronomy. 𝔊 even attests, against the shorter reading of ⅏ and 𝔐, an additional record of τὸν Ἰορδάνην in a further parallel formulation in Deut 11:8, there supported by 4QPhylA.K

138

BENJAMIN ZIEMER

inherited their shorter reading from the archetype of Deuteronomy whereas 4QDeutᶜ (or its lost Vorlage) added the word in assimilation to the parallel verses.23 I.e., the so-called “independent” reading of 4QDeutᶜ reveals the independence of ⅏, 𝔐 and 𝔊 of this manuscript, and brings out (together with similar readings) a specific tendency of this manuscript towards harmonisation. That 4QDeutᶜ itself is not dependent on ⅏, 𝔐 and 𝔊 is defined on grounds of chronology and language. But the question if 4QDeutᶜ is dependent on a 4th, 3rd or 2nd century BCE manuscript containing exactly the letters of ⅏, 𝔐 or the Hebrew Vorlageof 𝔊 remains open. If we look at the isolated readings of 4QDeutᶜ, we learn nothing to answer this question! The scribe of 4QDeutᶜ was able to add ‫ הירדן‬in Deut 4:14 against a parent text like 𝔐 or ⅏ as well as against the Hebrew Vorlageof 𝔊. As in this case, he was able to add, to omit or to change words in other cases against the text presupposed by these three main witnesses regardless which Vorlagehe had in front of him.24 To answer the question if 4QDeutᶜ is independent of the assumed forerunners of ⅏, 𝔐 or 𝔊, the specific readings of thelatterare decisive. E.g., in Deut 13:7, 4QDeutᶜ reads ‫ בן [אבי֯ ֯ך] או‬in agreement with ⅏ and 𝔊, while these words are not present in 𝔐. This proves that 4QDeutᶜ is independent of the 𝔐 tradition and isolates the short reading of 𝔐.25 In Deut 10:6–7, 4QDeutᶜ attests the order of 𝔐 and 𝔊 (first the burial of

23 24

25

(4Q128; 4Q138) and by the excerpted manuscript 4QDeutᵏ¹ (4Q 38). So 𝔊 is not suspected to shorten this formula. Cp. a similar addition attested in 𝔊 in Deut 27:3. In Deut 7:4, 4QDeutᶜ attests the longer reading ‫ יהוה [אלהי]כם‬against the shorter reading ‫ יהוה‬in the other three witnesses; in Deut 16:8, 4QDeutᶜ reads ‫עת ימים מצות תאכלו‬ ̇ ‫ש[ב‬ ̇ against ‫ ששת ימים תאכל מצות‬witnessed by all other witnesses; in Deut 13:7, 4QDeutᶜ reads [‫ ֯או֯ בנך או אש]ת חיקך‬against the longer reading ‫ או בנך או בתך או אשת חיקך‬presupposed by all other witnesses. These isolated variants of 4QDeutᶜ can easily be explained as secondary readings – the first by a harmonising addition (the Divine name is followed by a suffixed form of ‫ אלהים‬in a majority of the cases in all textual witnesses of Deuteronomy, so 𝔐 ⅏ 𝔊 share the lectiobreviorand difficilior), the second by a harmonising assimilation (towards Ex 12:5 and Lev 23:6), the third by omission through parablepsis (the eye jumped from ‫ בנך או‬to the similar ‫)בתך או‬. The same is true for most, if not all other singularity readings of this manuscript. An isolated reading may be original, if this is in accordance with the stemma. That means, the common reading of ⅏, 𝔊 and 4QDeutᶜ may be secondary against the isolated reading of 𝔐 if there are reasons to suppose a common ancestor of these three witnesses. Since there are no longer readings of 𝔐 against the combined testimony of ⅏ and 𝔊 in the text covered by 4QDeutᶜ, and no other special readings of 𝔐 which are clearly secondary, it cannot be excluded that this (and other) isolated reading(s) of 𝔐 is (are) inherited from the archetype. (On the contrary, in many cases this is the most suitable explanation for the variants.)

A STEMMA FOR DEUTERONOMY

139

Aaron, then the encamping in Gudgod/Gadgad) against ⅏, proving its own independence of the ⅏ tradition and isolating the ⅏ reading. This is supplemented by other cases where 4QDeutᶜ shares the common reading of 𝔐 and 𝔊 against a specific reading of ⅏.26 𝔊 shows far more idiosyncratic readings than 𝔐 or ⅏, and again 4QDeutᶜ helps to isolate some of these readings: in Deut 27:26, 4QDeutᶜ does not support the 𝔊 reading πᾶς ἄνθρωπος = ‫ כל איש‬before ‫ארור‬, proving the manuscript’s independence of the Hebrew Vorlageof 𝔊 and isolating the latter. This is supplemented by other cases where 4QDeutᶜ shares the shorter reading of 𝔐 and ⅏ against a longer reading translated by 𝔊 and viceversa.27 As a result of this short survey on isolated readings we can conclude that we have got four witnesses of Deuteronomy independent one of the other: 𝔐, ⅏, 𝔊 and 4QDeutᶜ. The isolated readings of a witness are crucial in order to exclude the possibility that the other witnesses are directly depending on it. 1.3.2. Classificationofvariantsaccordingtotheagreement/disagreement pattern In the present paper, I call such isolating readings “singularity variants” (S), and use the following abbreviations, using “1,” “2” and “3” as indications for the three main witnesses 𝔐, ⅏ and 𝔊: S1 = Singularity variant isolating a 𝔐 reading by the agreement of ⅏, 𝔊 and a Qumran manuscript against 𝔐. S2 = Singularity variant isolating a ⅏ reading by the agreement of 𝔐, 𝔊 and a Qumran manuscript against ⅏. S3 = Singularity variant isolating a 𝔊 reading by the agreement of 𝔐, ⅏ and a Qumran manuscript against 𝔊. S4 = Singularity variant isolating the reading of a specific Qumran manuscript by the agreement of 𝔐, ⅏ and 𝔊 against this Qumran manuscript. 26

27

In Deut 10:7, where ⅏ names, in parallel to the account in Num 33, several stations of the wilderness itinerary not present in the Deuteronomy text of 𝔐 or 𝔊, 4QDeutᶜ supports the shorter reading of the latter. In contrast, in Deut 11:3, where ‫ מלך מצרים‬is present in all witnesses except ⅏, 4QDeutᶜ shares the longer reading of the majority text. So, in both cases, 4QDeutᶜ isolates ⅏. Two more examples where 4QDeutᶜ supports the shorter reading of 𝔐 and ⅏, isolating the longer reading of 𝔊: In Deut 9:18 𝔊 alone has δεύτερον = ‫ שנית‬before ‫כראשנה‬, and in Deut 28:7 𝔊 alone has ὁ θεός σου = ‫ אלהיך‬after ‫יהוה‬. Two examples where 4QDeutᶜ supports the longer reading of 𝔐 and ⅏ so isolating the short reading of 𝔊: in Deut 4:15, 4QDeutᶜ attests the word ‫ כל‬before ‫תמונה‬, in agreement with 𝔐 and ⅏ against the shorter reading of 𝔊; in Deut 8:2, 4QDeutᶜ attests the words ‫זה ארבעים שנה‬, in agreement with 𝔐 and ⅏ against the shorter reading of 𝔊.

140

BENJAMIN ZIEMER

All these readings tell nothing about the positive family relationships between the witnesses. Supposedly by reason of the high number of such variants,28 Emanuel Tov uses 4QDeutᶜ as an example for a “statistically independent” manuscript, whose “few preserved readings are insignificant agreements and disagreements with the other sources.”29 But this is a deficient approach. In order to establish the stemma, the readings shared by at least two witnesses against at least two other witnesses are most important. I call them “family readings” since they alone may define textual families. I use the following abbreviations: F1 = Family variant indicating that a given Qumran manuscript is related to 𝔐 through agreement of the Qumran manuscript with 𝔐 against a reading shared by ⅏ and 𝔊. F2 = Family variant indicating that a given Qumran manuscript is related to ⅏ through agreement of the Qumran manuscript with ⅏ against a reading shared by 𝔐 and 𝔊. F3 = Family variant indicating that a given Qumran manuscript is related to 𝔊 through agreement of the Qumran manuscript with 𝔊 against a reading shared by 𝔐 and ⅏.

Sidnie Ann White, the DJD editor of this manuscript, made use of such family variants30 for the conclusion that 4QDeutᶜ “appears to be a manuscript located in the textual tradition of 𝔊.”31 E.g., 4QDeutᶜ and 𝔊 agree in reading the word ‫ דברי‬before ‫ השירה הזאת‬in Deut 31:28 against the shorter reading of 𝔐 and ⅏ (F3). Another F3 variant is the absence of ‫ ואת רעהו‬in Deut 15:2 shared by 4QDeutᶜ and 𝔊 against the longer reading of 𝔐 and ⅏. Both variants indicate the same stemmatic relationship: there were two different parent texts, one with the readings shared by 𝔐 and ⅏ and the other with the readings shared by 4QDeutᶜ and 𝔊. Nevertheless, it remains possible that one of these two parent texts depends on the other.32 The readings included in such family variants are, as a rule, inherited readings, telling about the idiosyncrasies of the common parent text. 28

29 30

31 32

Since 4QDeutᶜ is the Judean desert manuscript of Deuteronomy with most preserved words, it comes at no surprise that the number of variant readings covered by this manuscript is the greatest too. Among the 86 variants listed in DJD for 4QDeutᶜ there are more than 50 singularity variants, about a third of them S4 variants. Tov, Textual Criticism, 3rd ed., 109. In White’s terminology this means in most cases “shared error,” the “Leitfehler” of the classical stemmatic method. White, Special Features, 162. E.g., in our stemma(see below) the “𝔐+⅏ parent text” will be the archetype itself, α, whereas the “4QDeutᶜ+𝔊 parent text” will be δ, a manuscript finally dependent, by way of β and γ, on α.

A STEMMA FOR DEUTERONOMY

2. A FIRST

γ, THE 𝔊 AND ⅏

STEP TOWARDS THE STEMMA: SHARED BY

141

COMMON ANCESTOR

Before including the Qumran manuscripts in the stemma, it is helpful to refine the stemmafor the three main witnesses. With our three branch stemma drawn above, any agreement between two witnesses would define, as a rule, the text of the archetype. But there is a number of cases where this seems unlikely. A good example is the election of the tribe of Levi as described in Deut 18:5. Here 𝔊 and ⅏ (supported by the Temple Scroll) show some agreement against 𝔐: 𝔐

⅏ 𝔊GÖ retroverted 𝔊GÖ to Hebrew

‫ִ ֣כּי‬

‫כי‬

‫ כי‬ὅτι

‫֗בוֹ‬

‫בו‬

‫ בו‬αὐτὸν

‫ָבּ ַ ֛חר‬

‫בחר‬

‫ בחר‬ἐξελέξατο

‫יְהו֥ה‬ ָ

‫יהוה‬

‫ יהוה‬κύριος

‫ֹלהיָך‬ ֖ ֶ ‫ֱא‬

‫אלהיך‬

‫ִמ ָכּל־‬

‫מכל‬

‫שבטיך ְשׁ ָב ֶ ֑טיָך‬

‫כי‬ ‫במה‬ ‫בחרתי‬

– – ‫ מכל‬ἐκ πασῶν ‫ שבטיך‬τῶν φυλῶν σου

‫מד‬ ֹ ֨ ‫ַל ֲע‬

‫לעמד‬

–

‫לפני‬

‫ לפני‬ἔναντι

–

‫יהוה‬

‫ יהוה‬κυρίου

–

‫אלהיך‬

‫ולשרתו ְל ָשׁ ֵ ֧רת‬

Temple Scroll (11Q19, 60:10–11)

‫ לעמד‬παραστῆναι

‫מן כול‬ ‫שבטיכה‬ ‫לעמוד‬ ‫לפני‬

‫ אלהיך‬τοῦ θεοῦ σου ‫ לשרת‬λειτουργεῖν

‫ולשרת‬

–

‫ולברך‬

‫ ולברך‬καὶ εὐλογεῖν

‫ולברך‬

‫ם־יְהו֛ה‬ ָ ‫ְבּ ֵשׁ‬

‫בשמו‬

‫ בשמו‬ἐπὶ τῷ ὀνόματι αὐτοῦ

‫בשמי‬

‫֥הוּא‬

‫הוא‬

‫ הוא‬αὐτὸς

‫וּב ָנ֖יו‬ ָ

‫ובניו‬

‫ ובניו‬καὶ οἱ υἱοὶ αὐτοῦ

‫ָכּל־‬

‫כל‬

‫יָּמים׃‬ ֽ ִ ‫ַה‬

‫הימים‬

‫ כל‬πάσας ‫ הימים‬τὰς ἡμέρας

‫הוא‬ ‫וכול בניו‬ ‫כול‬ ‫הימים‬

Both ⅏ and 𝔊 have two longer readings (‫ לפני יהוה אלהיך‬and ‫ )ולברך‬and one shorter reading (‫ )בשמו‬in common, against 𝔐. In case these shared readings are secondary they can easily be explained as assimilating the verse to a close parallel in Deut 10:8 where the Hebrew text attested by 𝔐

142

BENJAMIN ZIEMER

and ⅏ reads 33 ‫לעמד לפני יהוה לשרתו ולברך בשמו‬. Against it, in case the isolated readings of 𝔐 in Deut 18:5 are secondary, it would be very difficult to explain their emergence. There is no visible occasion for parablepsisin this verse. Moreover, two unconscious omissions are unlikely since the correlation of the two shorter readings of 𝔐 with the longer reading ‫ בשם יהוה‬in place of ‫ בשמו‬resulting in exactly one record of the Divine name in the description of the function of the tribe of Levi in 𝔐 as well as in ⅏ and 𝔊 looks more like a conscious modification of the text, in whichever direction. But why should 𝔐 have consciously omitted the function of the tribe of Levi to bless in the name of the Lord? No theological reasons are to be found for this, since in the parallel, in Deut 10:8, 𝔐 agrees with ⅏ in the words ‫ולברך בשמו‬. So it is highly probable that the shorter reading of 𝔐 ‫ לעמד לשרת בשם יהוה‬is original for Deut 18:5, and that both ⅏ and 𝔊 replaced it by the longer, more fluent reading, which was not new in content but assimilating with Deut 10:8. Now we have to decide whether the agreement of ⅏ and 𝔊, even supported by 11QT, is inherited, or has arisen by accident. If ⅏ and 𝔊 made this harmonisation independently it is astonishing that both agree in Deut 10:8 in the short reading ‫ לעמד לפני יהוה‬but in Deut 18:5 in the longer reading ‫לעמד לפני יהוה אלהיך‬, against Deut 10:8. It is not very probable that this happens twice in the same manner. Moreover, since the addition of ‫ ולברך‬is completely understandable, it would not have been necessary to change ‫ לשם יהוה‬to ‫לשמו‬, but again both witnesses agree in this point. To summarise, it is by far the most easy explanation to suppose an inherited agreement in a secondary variant. For this we have to refine the stemma, using γ as designation for the common ancestor of ⅏ and 𝔊:

Į

╱╲

Ȗ

╱æ╲

11QT  If this stemmais correct, the text of γ is determined by the agreement of ⅏ and 𝔊, or by 𝔐 with either ⅏ or 𝔊. So, if we like to exclude accidental agreement in this verse, the text of γ, resulting from the 33

In Deut 10:8, Tal’s edition of ⅏, according to Ms. Shekhem/Nablus B, lacks the word ‫ לשרתו‬present in all Samaritan manuscripts collated by Kennicott and von Gall, so it seems to be a secondary omission by this manuscript.

A STEMMA FOR DEUTERONOMY

143

described assimilation to Deut 10:8 (addition of ‫לפני יהוה אלהיך‬, and replacing of ‫ בשם יהוה‬by ‫)ולברך בשמו‬, can be reconstructed for Deut 18:5 as follows: ‫(לשרת ולברך‬34?‫כי בו בחר יהוה אלהיך מכל שבטיך לעמד לפני יהוה אלהיך )ו‬ .‫בשמו הוא ובניו כל הימים‬

The three witnesses of γ made further changes. 𝔊 (or one of its Hebrew parent texts after γ, i.e. δ, ε or ζ, defined below) omitted the first ‫אלהיך‬ (and, perhaps, the ‫ ו‬before ‫)לשרת‬. ⅏ (or a further parent text between γ and ⅏, like η, defined below) replaced ‫( לשרת‬or ‫ )ולשרת‬by ‫ולשרתו‬. Both changes may have arisen consciously or unconsciously. The author of the Temple Scroll has changed consciously the perspective into a first person speech of God, so we cannot be sure what exactly was his Vorlagefor ‫בחרתי‬, ‫ לפני‬and ‫ בשמי‬since he would render ‫לפניו‬, ‫ לפני יהוה‬or ‫לפני יהוה‬ ‫ אלהיך‬in any case by ‫)ל ָפנַי =( לפני‬. ְ Moreover, there are several minor changes made by 11QT or its Vorlage: Replacement of ‫ בו‬by ‫במה‬, addition of ‫כול‬ before ‫בניו‬, and (presumably) addition of ‫ ו‬before ‫לשרת‬. So we can learn from this case that in the course of textual history there are additions, replacements and omissions. It is impossible to explain all quantitative divergences by addition. If the assimilation of Deut 18:5 towards Deut 10:8 would be the only reading shared by ⅏ and 𝔊, we had to choose an alternative explanation. But in fact there is a pattern of similar cases which consolidates this stemma, with a presumed hyparchetype γ as parent text for ⅏ and 𝔊. I can show this by some statistics assembled by Emanuel Tov. He collected quantitative differences between the three main witnesses which could be explained according to Tov as “textual harmonizations.”35 Due to the fact that it is not clear by itself that all these differences are harmonisations, and that the direction of the changes can be explained in many cases in both directions as addition oras omission, I will describe the evidence in a more objective manner. First, the differences concern the 34

35

It is not clearly determined whether γ read ‫ לשרת‬or ‫ולשרת‬. In any case, the presence of ‫ ו‬fits better with the longer text shared by 𝔊, ⅏ and 11QT. Most likely, ⅏ and 11QT added the ‫ ו‬independently, by accident. Their agreement is only partial, typical for accidental agreement: 11QT has ‫ולשרת‬, ⅏ ‫ולשרתו‬. Possibly, but less likely, γ might have added ‫ו‬, and 𝔊 omitted it, leading to agreement with 𝔐 by accident, while ⅏ later added the suffix. Least likely is the third possibility: α and γ had ‫ולשרת‬, 𝔐 and 𝔊 both omitted the ‫ו‬, ⅏ later added the suffix, and just the least reliable witness 11QT retains the original reading. Tov, Textual Harmonizations.

144

BENJAMIN ZIEMER

presence or absence of at least one content word. This is stated by Tov since he explicitly excludes some “additions of small details,” like suffixed ‫ אלהים‬as apposition to ‫יהוה‬, the words ‫כל‬, ‫נא‬, ‫כן‬, ‫גם‬, ‫כה‬, pronominal prepositions such as ‫בך‬, ‫ לך‬and similar.36 Therefore the remaining variants are large enough to make accidentalagreementunlikely. Second, all these parts of text, present at a certain place in one or two of the main witnesses but absent in the remaining witness(es), have a counterpart elsewhere in the text common to all three witnesses. This fact puts Emanuel Tov in the position to describe the variants as harmonisations, but it bears another implication important for our task: even though these changes concern content words they do not change the content of the book. For that reason, contamination – the conscious aligning of a manuscript to another tradition – is unlikelyfor these cases. That is why the evidence collected by Emanuel Tov is typical for variants with inherited agreements, explainable solely by stemmatic relationship. Tov found 202 such cases, covering at least one content word, in the prose sections of Deuteronomy. He arranged the evidence as follows, supposing that in each and every case the shorter reading is original: Types of variants, according to Tov

𝔐

⅏

𝔊

exclusive harmonising additions

2 (type A)

22 (type B)

99 (type C)

shared harmonising additions

27 with 𝔊 (type D) 27 with ⅏ (type D) 8 with 𝔊 (type E) 8 with 𝔐 (type E) 44 with ⅏ (type F) 44 with 𝔐 (type F)

Sum of harmonising “54” additions37

“93”

“134”

Tov’s intention was to show that all textual traditions of Deuteronomy show harmonising additions. This explanation fits well for the stemmatic explanation of our example in Deut 18:5: the presence of ‫לפני יהוה אלהיך‬ in ⅏ and 𝔊 is just one of the 27 cases of type D in Tov’s statistics. But the variant types E and F, taken as additions, would be incongruent with 36 37

Tov, Textual Harmonizations, 273f. These figures are given explicitely by Tov, Textual Harmonizations, 281.

A STEMMA FOR DEUTERONOMY

145

the stemma, leading to serious contradictions. However, they can easily be explained if we concede, against Tov, that omissions and substitutions occurred as well as additions. The same verse, Deut 18:5, gives welcome evidence for this. Tov, based on the text of 𝔊RA, states the “addition” of ‫ כל הימים‬at the end of Deut 18:5 as a “harmonizing plus of MT SP” “to the short text of the LXX.”38 Indeed, 𝔊RA (based on 𝔊B) has ἐν τοῖς υἱοῖς Ισραηλ at the end of this verse, which seems to be a translation of ‫ בבני ישראל‬instead of ‫כל‬ ‫הימים‬. But this is, as now shown by one of the oldest 𝔊 witnesses, Pap. 848 from the first c. BCE, a singularity reading of 𝔊B. Most Greek witnesses, including Pap. 848, read πάσας τὰς ἡμέρας, consequently taken as text of the Old Greek by 𝔊GÖ. A large number of Greek witnesses added ἐν τοῖς υἱοῖς Ισραηλ before this, assimilating the verse to the wording of Deut 17:20 (καὶ οἱ υἱοὶ αὐτοῦ ἐν τοῖς υἱοῖς Ἰσραήλ). 𝔊B apparently inherited this longer reading but omitted πάσας τὰς ἡμέρας for no discernible reason, so leading in two stages to the substitution of πάσας τὰς ἡμέρας (the original text, not an addition!) by ἐν τοῖς υἱοῖς Ἰσραήλ in Deut 18:5.39 So, within the Greek manuscript tradition we see a harmonising addition followed by an omission of a string of the original text. In contrast, the omission of the first ‫ אלהיך‬in this verse by the Old Greek (𝔊GÖ shows this shorter reading, based on Pap. 848, 𝔊B and the Old Latin) has been followed by the addition of ὁ θεός σου in the majority of Greek manuscripts (= 𝔊RA), leading to agreement with ⅏ and 𝔐 by accident or via Hexaplaric influence. If we reckon with the possibility of additions andomissions we may group the evidence collected by Emanuel Tov in another way. For genetic explanation, it is of utmost importance to know which reading is shared by two witnesses against the third, so as to determine the special, singularity readings of each of the three witnesses. 38 39

Tov, Textual Harmonizations, 274f. The witnesses collated for 𝔊GÖ show the following readings in Deut 18:5, with additions and omissions: 848 A F 82 oʹ’ Cʹ’ 56txt πάσας τὰς ἡμέρας 127 syzAeth Arm La = 𝔊GÖ πάσας τὰς ἡμέρας κυρίῳ 376ʹ bArab πάσας τὰς ἡμέρας ÷κυρίῳp Syh ἐν υἱοῖς Ἰσραήλ πάσας τὰς ἡμέρας M n-127 85mg-321ʹ mg ἐν Ἰσραήλ πάσας τὰς ἡμέρας V ἐν τοῖς υἱοῖς Ἰσραήλ πάσας τὰς ἡμέρας df-56txt tCo ἐν τοῖς υἱοῖς Ἰσραήλ B = 𝔊ᴿᴬ

146

BENJAMIN ZIEMER

Types of variants exclusive pluses

exclusive minuses exclusive readings

𝔐 〛 𝔊=⅏

2 (type A)

27 (type D)41

29 (A+D)

⅏ 〛 𝔐=𝔊

22 (type B)

8 (type E)

30 (B+E)

𝔊 〛 𝔐=⅏

99 (type C)

44 (type F)

143 (C+F)

40

Looking at this chart, two facts are striking. First, 𝔊 has got by far most exclusive readings: 143,42 against 30 exclusive readings of ⅏, and 29 exclusive readings of 𝔐. Second, 𝔐 has virtually no exclusive pluses whereas the other two witnesses both have more pluses than minuses. If we were assuming a stemmawith three branches, the reading of the majority would define, as a rule, the original reading. In this case the exclusive readings show the tendency of each witness: 𝔊 should have added harmonisations in 99 cases and omitted words in 44 cases irrespective of whether they are understandable as harmonisation. Which means that the textual tradition behind 𝔊 was not as faithful to their Vorlagenas the traditions behind ⅏ and 𝔐. ⅏ shows a similar tendency towards expansion, but on a lower level, with a ratio of 22 additions vs. 8 omissions. Against this, 𝔐 would show a clear tendency towards shortening, with 27 cases of omission against 2 cases of addition. This sounds implausible. But if we assume a stemmawith γ as the common hyparchetype for ⅏ and 𝔊, the complete picture will be coherent (and we can retain our explanation of Deut 18:5 as a paradigmatic type D variant): Types of changed readings

additions

omissions

ratio additions: omissions

𝔐〛α

0? (type A)

0? (type D)

not defined

γ〛α

27? (type D)

2? (type A)

≈ 13.5 : 1

⅏〛γ

22 (type B)

8 (type E)

2.75 : 1

𝔊〛γ

99 (type C)

44 (type F)

2.25 : 1

sum of changes

148–150

52–54

ca. 2.8 : 1

40

41

42

That means: In a type A variant the string of text in question is present in 𝔐, but absent from ⅏ and 𝔊. That means: In a type D variant the string of text in question is absent in 𝔐, but present in ⅏ and 𝔊. The picture would differ in some detail if he had taken 𝔊GÖ as textual base for 𝔊 instead of 𝔊ᴿᴬ, but this would not affect the general conclusions. As shown for Deut 18:5, additions, substitutions and omissions occurred both in the Hebrew pre-history of 𝔊 and in the textual history of 𝔊 itself.

A STEMMA FOR DEUTERONOMY

147

Since Biblical scholarship is accustomed to acknowledge textual and literary additions anytime and everywhere but to doubt the evidence of omissions,43 I would like to show that omissions may be a good explanation for textual variants. Let us take the eight cases of type E where 𝔐 and 𝔊 agree in a longer reading against ⅏. Tov considered these cases as harmonising additions by a shared base text of 𝔐 and 𝔊. But this is a contradiction in itself, if we explain the 27 cases of type D with a shared base text of 𝔊 and ⅏. If our stemmais correct, the eight type E differences musthave been come into existence as omissions of ⅏ against γ. Such omission may or may not by itself be a “harmonistic” one. It could well be an accidental omission by parablepsis, for example through homoioteleuton. Indeed, the largest of the eight differences Tov counted as common harmonisation by 𝔐 and 𝔊 against ⅏ (type E) is suitable as paradigm for homoioteleuton, Deut 2:10f.: ‫אַף־הם ָכּ ֲענָ ִ ֑ קים‬ ֵ֖ ‫𝔐 ָה ֵא ִ ֥מים ְל ָפ ִנ֖ים ָי ְ֣שׁבוּ ָ ֑בהּ ַע֣ם גָּ ֥דוֹל וְ ַ ֛רב וָ ָ ֖רם ָכּ ֲענָ ִ ֽ קים׃ ְר ָפ ִ ֛אים יֵ ָח ְשׁ ֥בוּ‬ ‫אָבים יִ ְק ְר ֥אוּ ָל ֶ ֖הם ֵא ִ ֽמים׃‬ ִ֔ ‫מּ‬ ֹ ֣ ‫וְ ַה‬ ‫⅏ האימים לפנים ישבו בה עם גדול ורב ורם כענקים׃‬ ‫והמואבים יקראו להם אימים׃‬

𝔐, supported by 𝔊, reads at the end of verse 10 ‫ עם גדול ורב ורם כענקים‬and in the middle of the following verse ‫אף הם כענקים‬. The eye of a scribe in the textual tradition inherited by the Samaritan Pentateuch just jumped from the first ‫ כענקים‬to the second. So he omitted by accident the whole sentence ‫רפאים יחשבו אף הם כענקים‬. This sentence is a remark of archeological interest not needed for the syntax nor for the understanding. So its omission left no traces, and remained uncorrected in the Samaritan tradition.44 Of the eight type E differences, three can easily be explained by parablepsis (Deut 2:11; 9:11; 30:5), four others as negative assimilations (i.e., assimilation by omission: Deut 2:9; 3:21; 24:22; 28:15), and just one case remains unclear, with both possibilities of explanation (Deut 11:3). Moreover, in the three cases of Qumran evidence, only the longer text of 𝔐 and 𝔊 is attested in Qumran manuscripts respectively (Deut 3:21; 11:3; 28:15), isolating in any case the Samaritan reading. I list the eight type E differences counted by Emanuel Tov as common harmonising additions by 𝔐 and 𝔊 against ⅏, with suggestions on how or why the differences can easily be explained as omissions by ⅏. The cited text strings are present 43 44

But cp. Pakkala, God’s Word. Note this same omission happened in Kennicott’s Ms. 184, proving the occasion for homoioteleuton.

148

BENJAMIN ZIEMER

in 𝔐 and translated by 𝔊 but absent from ⅏ (the words in brackets show the context present in all witnesses): Deut 2:9

Deut 2:11

‫ מלחמה‬negative assimilation to Deut 2:5 ∥ Num 20:13⅏ and to Deut 2:19 ∥ Num 21;13⅏, in accordance with the parallel added in Num 21:12⅏

45

‫ רפאים יחשבו אף הם כענקים‬homoioteleuton‫( –ם כענקים‬see above)

Deut 3:21

‫ אלהיכם‬and ‫ כל‬these are minor omissions (both words must not be counted according to Tov’s own criteria) but they are also negative assimilations to Deut 4:3 ⅏+𝔐46, in accordance with the parallel added in Num 27:23 ⅏+4QNumb; the presence of these words in Deut 3:21 is confirmed by 4QDeutᵈ, a Qumran Deut manuscript relatively close to ⅏

Deut 9:11

(‫ לחת האבנים )לוחת הברית‬homoioarcton(–‫ ל]ו[ח]ו[ת ה‬/ –‫)ל]ו[ח]ו[ת ה‬

Deut 11:3

Deut 24:22 Deut 28:15

Deut 30:5

‫ )לפרעה( מלך מצרים‬occasion of the omission remains unclear;47 anyway, ‫ מלך מצרים‬is present in 4QDeutᶜ and several phylacteries, so this accidental omission by ⅏ is obviously an isolated one ‫ לעשות‬negative assimilation to Deut 15:15 ‫ לעשות את כל‬negative assimilation to Deut 28:15 and 39:10; these words were apparently present in 4QpalDeutr, a Qumran Deut manuscript relatively close to ⅏, so the omission by ⅏ is an isolated one (‫ והיטבך )והרבך‬homoioarctonor homoioteleuton (‫ וה–בך‬/ ‫)וה–בך‬

These omissions, presumably made unconsciously, are few. In general, ⅏ tended to expansion, but (beside the Garizim commandment in the Decalogue) only very slightly. So all isolated readings of ⅏ listed by 45

46

47

Tov counts ‫ בם מלחמה‬as the “harmonizing addition” but ⅏ has got ‫ב‬+suff., hence the quantitative difference affects only one word. 𝔊 assimilates both verses in opposite direction, by adding πάντα and ὁ θεὸς ἡμῶν in Deut 4:3, reading Οἱ ὀφθαλμοὶ ὑμῶν ἑωράκασιν πάντα, ὅσα ἐποίησεν κύριος ὁ θεὸς ὑμῶν/ἡμῶν in both verses. It is either a partial negative assimilation to the close parallels in Deut 7:18; 29:1 and 34:11, or a conjecture, since nowhere else the pleonasm ‫ פרעה מלך מצרים‬follows immediately the word ‫מצרים‬.

A STEMMA FOR DEUTERONOMY

149

Tov (types B and E), whether concerning the presence or the absence of words, can easily be explained as secondary readings, coherent with our stemma. The same is true, as a rule, for the isolated omissions of 𝔊 (type F). I just take the first of the 44 type F variants listed by Emanuel Tov as “harmonizing additions in the combined text of MT SP to the short text of the LXX”48 as an example, ‫ ראשי שבטיכם‬in Deut 1:15. In this verse, 𝔐 and ⅏ agree in reading ‫ואקח את ראשי שבטיכם אנשים חכמים וידעים‬, whereas 𝔊 (καὶ ἔλαβον ἐξ ὑμῶν ἄνδρας σοφοὺς καὶ ἐπιστήμονας καὶ συνετοὺς) apparently translates a different Hebrew text: ‫ואקח מכם אנשים‬ ‫חכמים ונבנים וידעים‬.49 Tov does not mention the addition of ‫ מכם‬by 𝔊50 and he does not explain why ⅏ and 𝔐 should have introduced an expression from a completely different context (Deut 5:23).51 His motivation lies in a preference for the rule of the lectiobrevior. But it is easier and more consistent to explain this difference with the stemmadrawn above. Hence, ⅏ and 𝔐 copied faithfully the wording of their parent texts. This has to be the text of the archetype α, since ⅏ inherited this reading through γ which is determined in this case by the agreement of 𝔐 and ⅏. In contrast, the tradition behind 𝔊 changed the text, adding one word (‫ )ונבנים‬and substituting three words (‫ )את ראשי שבטיכם‬by another word (‫)מכם‬, a conclusion supported by Qumran evidence. Both changes are plausible, whether consciously or not, as assimilations to similar verses in the 𝔊 branch of the stemma. The first change, the plus of καὶ ἐπιστήμονας (‫ ונבנים‬in the assumed Hebrew Vorlage),can easily be explained as a harmonising addition, since the phrase ἄνδρας σοφοὺς καὶ ἐπιστήμονας καὶ συνετοὺς is an exact reduplication of the wording in Deut 1:13 where καὶ ἐπιστήμονας renders the Hebrew word ‫וּנְ בֹנִ ים‬.52 Apparently, this addition was shared by 4QDeutʰ, according to calculations of space. The second change, the replacement of ‫את ראשי שבטיכם‬, had not yet occurred in the common parent text of 4QDeutʰ and 𝔊.53 Nevertheless it solves an inherent problem 48 49

50

51 52

53

Tov, Textual Harmonizations, 274. The Qumran evidence isolates the shorter 𝔊 reading at the beginning as a singularity reading: 4QDeutʰ reads ‫ואקח את ראשי שבטיכם אנשים חכמ]ים‬, like 𝔐 and ⅏. If one takes this into account, even a harmonising addition of ‫ מכם‬and a following homoioteleuton is possible – whether in the Hebrew (‫ )מ]כם את ראשי שבטי[כם‬or in the Greek (ἐξ [ὑμῶν τοὺς ἡγουμένους τῶν φυλῶν] ὑμῶν). Tov, harmonization, 274. Tov does not mention this addition in his article. But according to his principles it could be added to the 99 cases of exclusive harmonistic additions of 𝔊 (type C). With the testimony of 4QDeutʰ, placed in the stemmabelow (with some hesitation) as a descendant of ε, we are able to locate the changes. The addition of ‫ ונב)ו(נים‬took place

150

BENJAMIN ZIEMER

of the verse in the text shared by ⅏ and 𝔐. If the men Moses “took” were already “the heads of the tribes,” why Moses had to make them to “heads over you,” and how many heads of the tribes were there if Moses could make of them heads over thousands, hundreds, fifties and tens, as is said in the second half of this verse (‫ואתן אתם ראשים עליכם שרי אלפים‬ …‫ ?)ושרי מאות‬The formulation 𝔊 used instead (καὶ ἔλαβον ἐξ ὑμῶν ἄνδρας) may well be taken from Deut 1:23 (καὶ ἔλαβον ἐξ ὑμῶν δώδεκα ἄνδρας).54 In that verse ‫ ואקח‬is followed by ‫מכם‬, and ‫שנים עשר אנשים‬ functions as the direct object instead of ‫ראשי שבטיכם‬. The solution of 𝔊 or its Vorlageto shorten the text by assimilation (καὶ ἔλαβον ἐξ ὑμῶν ἄνδρας…) may be seen as a real enhancement of the text – but not as original. I hope this example suffices to show that isolated minuses of 𝔊 or its Vorlage, as a rule, may be explained as omissions. However, nearly three quarters of the two hundred variants collected by Tov will be explained as additions by our stemma. In the chart above, question marks remain for the type A and type D variants. This is due to the fact that these variants show the only case the text of the archetype is not determined automatically. But this uncertainty does not affect the advantages of the stemmatic method. This is important to remember if we now take a step forward and include the Qumran manuscripts in the discussion: we need not know whether a reading is superior or inferior for placing our manuscript in the stemma. We have to look solely at the agreements or disagreements. Say, we have a series of superior readings in 𝔊 and ⅏ against a series of inferior readings in 𝔐. If we find some of these inferior readings in a manuscript we should place this manuscript at the 𝔐 side of the stemma. But the same is true if we have got a series of inferiorreadings in 𝔊 and ⅏ against superior readings in 𝔐. In case we will find some of these superior readings we have to place the manuscript at the 𝔐 side of the stemma again, since otherwise we cannot explain from where our manuscript got the superior reading. In both cases the place in the stemmais determined solely by the agreement, not by a supposed superiority.

54

in δ or ε, whereas the replacement of ‫ ראשי שבטיכם‬by ‫ מכם‬was an innovation not introduced until ζ. So the decision of McCarthy (BHQ, adloco).

A STEMMA FOR DEUTERONOMY

151

3. THE IMPORTANCE OF WEIGHING THE AGREEMENTS: THE EVIDENCE OF THE VULGATE If we are looking at and counting the agreements and disagreements, it is of utmost importance to weigh them according to the discernible tendency of a certain witness. Let us take an example: the Divine name ‫ יהוה‬is mentioned in 𝔐 of Deut 550 times, and in ⅏ 555 times. There are 545 records common to 𝔐 and ⅏, and thirteen cases of deviation; in four cases ‫ יהוה‬is present in 𝔐 but absent from ⅏,55 and in nine cases ‫יהוה‬ is present in ⅏ but absent from 𝔐, among them the five records in the Garizim commandment of Deut 5.56 So it is obvious that both witnesses are very close one to the other in textual details. If we take a third witness into account, we are able to get a picture of the exclusivereadingsof each of the witnesses. The Greek Deuteronomy (according to the Göttingen edition) uses the different inflected forms of κύριος 553 times. Among them are 18 records of κύριος where 𝔐 and ⅏ lack ‫( יהוה‬among them 6 cases where κύριος renders forms of ‫)אדון‬,57 two records where 𝔊 and 𝔐 share the presence of ‫ יהוה‬/ κύριος against an isolated reading of ⅏,58 and two records where 𝔊 and ⅏ share the presence of ‫ יהוה‬/ κύριος against an isolated reading of 𝔐.59 In contrast, there are 15 cases where 𝔊 alone lacks κύριος while ‫ יהוה‬is present in 𝔐 and ⅏,60 seven cases where 𝔊 and 𝔐 share the absence of ‫ יהוה‬/ κύριος against an isolated reading of ⅏,61 and two cases where 𝔊 and ⅏ share the absence of ‫ יהוה‬/ κύριος against an isolated reading of 𝔐.62 55

56

57

58

59

60

61 62

Deut 1:8; 9:5 [2]; 18:5 [2]; 28:53. In two of these four cases 𝔊 agrees with ⅏ against 𝔐 (Deut 1:8; 18:5). Deut 2:12; 5:18B (= five records within the Garizim commandment not extant in 𝔐); 6:23; 14:23; 18:5 [1]. In two of these nine cases 𝔊 agrees with ⅏ against 𝔐 (Deut 14:23; 18:5). In Deut 3:24; 9:26; 10:17 (bis); 23:16 (bis) the “additional” κύριος is the equivalent for various forms of ‫אדון‬, so in these cases there was presumably no difference between 𝔐 and the Hebrew Vorlageof 𝔊. The remaining twelve cases are in Deut 6:4; 7:13; 11:25; 13:18; 16:16; 30:16; 31:11, 17, 23; 32:4, 37, 43. Deut 9:5 [2]; 28:53. In Deut 9:5 [2] 𝔊GÖ has κύριος, whereas 𝔊ᴿᴬ has the shorter reading in accidental agreement with ⅏. Deut 14:23 (⅏ and 𝔊 have ‫ יהוה אלהיך‬as subject in the election formula against 𝔐); 18:5 (⅏ and 𝔊: ‫𝔐 ;לעמד לפני יהוה אלהיך‬: ‫)לעמד‬. So these readings turn out to be exclusive short readings of 𝔐. Deut 2:15; 4:20; 6:19; 8:3; 9:26; 10:9; 12:11, 21; 16:1; 19:1; 21:5; 23:3; 26:17; 29:19; 31:27. Among these cases ‫ יהוה‬is rendered by θεος nine times, and omitted six times. These 15 cases show exclusive readings of 𝔊. The five records in the Garizim commandment, and Deut 2:12; 6:23. In Deut 1:8, within a cited Divine speech, 𝔐 reads ‫נשבע יהוה‬, a lectiodifficilioragainst ‫ נשבעתי‬attested by ⅏ and 𝔊. In Deut 18:5, 𝔐 reads ‫ בשם יהוה‬against ‫ בשמו‬attested by

152

BENJAMIN ZIEMER

Altogether, these three witnesses agree in the presence of the Divine name in 531 cases; 𝔊 shows 33 exclusive readings (presence of κύριος: 18, among them 6 times as translation for ‫אדניו‬/‫אדונים‬/‫אדון‬/‫ ;אדני‬absence: 15); ⅏ shows 9 exclusive readings (presence: 7, among them 5 within the Garizim commandment; absence: 2); and 𝔐 shows 4 exclusive readings (presence: 2; absence: 2). It seems obvious that these three witnesses of Deuteronomy are relatively close one to the other, agreeing in more than 95 % of the cases. However, 𝔐 and ⅏ (at least beside the Garizim commandment)63 prove to be more faithful to the supposed archetype than the textual tradition behind 𝔊. It is also obvious that the differences must not be explained solely by expansions. All this fits in the picture drawn above concerning the evidence assembled by Tov under the misleading title “harmonizing additions.” If we now take the Vulgate into consideration, the result may come as a surprise. The Vulgate of Deuteronomy shows only 477 records of Dominus even though this word is the unrivaled equivalent for ‫ יהוה‬and ‫אדני‬. The Vulgate shows, in matters of the Divine name alone, more than 70 shorter readings against the common text of 𝔐, ⅏ and 𝔊. In the following chart I summarize the data collected above concerning the exclusive readings of 𝔐, ⅏ and 𝔊, and add the records in the Vulgate.

63

⅏ and 𝔊: Obviously, this variant is connected with the exclusive absence of ‫לפני יהוה‬ in the same verse in 𝔐. See above for the discussion of this verse. In fact, the Garizim commandment is a composition of Biblical verses and follows the common archetype in most textual details exactly. All five occurrences of the divine name are in complete correspondence to the five references of ‫ יהוה אלהיך‬in Deut 11:29 and 27:5–7 which are there supported by the common testimony of 𝔐=⅏=𝔊. Within the Garizim commandments, only the two geographical designations ‫( בהרגריזים‬against ‫בהר עיבל‬/ἐν ὄρει Γαιβάλ) as localisation for the stones and the altar (the central issue of the pericope), and ‫( מול שכם‬exclusive plus, in allusion to Gen 12:6) correspond to differences between ⅏ on the one hand and the combined testimony of 𝔐 and 𝔊 on the other in Deut 11:30 (𝔐 𝔊 without ‫ )מול שכם‬and in Deut 27:4 (𝔐 𝔊 ‫בהר עיבל‬/ἐν ὄρει Γαιβάλ). Two further geographical specifications are not even attested in the respective verses of ⅏: First, ‫( אל ארץ הכנעני‬perhaps influenced by Ex 13:11, or an allusion to Gen 12:6) in the first sentence of the commandment faces ‫ אל הארץ‬attested by all witnesses of Deut 11:29. Second, ‫ ההר ההוא‬as introduction of the last sentence, referring to the Garizim alone, faces the plural ‫ הלוא הם‬/ ‫ הלא המה‬/ οὐκ ἰδοὺ ταῦτα which is attested in Deut 11:30 by all witnesses, including ⅏, referring to the two mountains, Garizim and Ebal. So the text of the Garizim commandment shows the same characteristics as ⅏ as a whole: Isolated readings against the combined testimony of 𝔐 and 𝔊 in matters of the geographical designation of the promised land (cp. Gen 10:19; Deut 34:1–3) and the chosen place (cp. Deut 12:5, 11 etc.), but conservatism in textual details.

153

A STEMMA FOR DEUTERONOMY

Exclusive readings of 𝔐, 𝔊 or ⅏ regarding ‫ יהוה‬resp. κύριος in Deuteronomy

Exclusive presence of ‫( יהוה‬κύριος, Dominus)

Exclusive absence of ‫( יהוה‬κύριος, Dominus)

Sum of exclusive readings, and classification

Cases Cases shared Cases Cases shared Cases (total) by Vulgate (total) by Vulgate (total) Yes

No

2

0

⅏ altogether 7 (within Garizim comm.) (5)

0 (0)

𝔊 altogether 18 (– translations of ‫)אדני‬64 (6) (= remaining cases) (12)

𝔐 (Ketiv)

2

69

Vulgate against all three 1 Sum of cases Dominuspresent or… … not present in Vulgate 64

65

66

67

68

69

Cases shared by Vulgate

Yes

No

2

2

0

4

4

7 (5)

2

0

2

9 (5)

0 F2 9 S2 (0) (5)

5 (4)65 (1)66

13 (2)67 (11)

15

768

8

(15)

(7)

(8)

33 (6) (27)

21 12 (2) (4) (8) F3 (19) S3

1

–

72

72

–

73

73 S4 –

28 (22)

91

No F1 0

119 (113)

8 (4)

10 20 (18)

Yes

81

18

(14)

101

(99)

The translation of ‫ יהוה‬and ‫ אדני‬with the same word for “the Lord” reveals no difference in the Hebrew Vorlage, since it is in exact agreement with the traditional Qere of 𝔐. These cases have only been included in the statistics in order to match all cases of ‫יהוה‬, κύριος and Dominus in Deuteronomy. In two verses (Deut 3:24; 9:26) the Vulgate reads Dominefor ‫ אדני‬and Deusfor ‫ ;יהוה‬these are counted as two cases of presence and two cases of absence of Dominusagainst the Ketivof 𝔐, though the reading of the Vulgate is in exact agreement with the Qereof the Masoretic text. Deut 3:24; 9:26; 10:17 [1]; 23:16 [1]. Three cases show a rendering of the elements of ‫ אדני יהוה‬resp. ‫ אדני האדנים‬with partial agreement since 𝔊 reads Κύριε κύριε resp. κύριος τῶν κυρίων but the Vulgate reads Domine Deus (3:24; 9:26) resp. Dominus dominantium (10:17). In the fourth case, Deut 23:16 [1], 𝔊 and Vulgate both use κύριος resp. Dominusfor ‫אדניו‬. Deut 31:11. The partial agreement of Vulgate and 𝔊 most likely arose independently (see below). For ‫ האדנים‬in Deut 10:17 𝔊 has a form of κύριος but the Vulg. has a form of the participle dominans.The phrase ‫ מעם אדניו‬at the end of Deut 23:16, with the second record of ‫ אדניו‬in this verse, is rendered by 𝔊 through παρὰ τοῦ κυρίου αὐτοῦ but is omitted by the Vulgate. Deut 2:15; 6:19; 9:26; 12:21; 19:1; 21:5; 23:3. Among these agreements in the absence of the Divine name there are only two real agreements (Deut 6:19; 21:5). In the five remaining cases the Vulgate and 𝔊 disagree one with the other. In Deut 2:15; 9:26; 12:21 𝔊 reads θεός while the Vulgate has a shorter reading without Dominusor Deus; in Deut 19:1 [2] the Vulgate again has a different and shorter reading than 𝔊 (𝔐=⅏ ‫הגוים אשר יהוה אלהיך נתן לך‬, 𝔊 ὰ ἔθνη, ἃ ὁ θεός σου δίδωσίν σοι; Vulg. gentes quarumtibitraditurusest); in Deut 23:3 𝔊 omits a whole sentence whereas the Vulg. omits only the second ‫בקהל יהוה‬. Deut 12:26 [1].

S1

154

BENJAMIN ZIEMER

With the Vulgate as fourth witness, there remain only 459 agreements between all witnesses, besides 119 variants. If we would count the differences as Armin Lange does in his Handbuchfor the Judean desert manuscripts we would get the following statistics (in brackets without the six cases of ‫אדני‬/‫)אדון‬: there are 119 (113) variants concerning the Divine name. The Vulgate reads with 𝔐 in 34 (32) cases, against 𝔐 in 85 (81) cases, with ⅏ in 21 (19) cases, against ⅏ in 98 (94) cases, with 𝔊 in 21 (17) cases, against 𝔊 in 98 (96) cases, and reads independently (“eigenständig”) in 73 cases. So the Vulgate would turn out to be an “independent” witness for Lange.70 But, obviously, this would be misleading! If we eliminate the cases where forms of ‫ אדון‬are rendered by κύριος or Dominus, we get 113 cases where there has obviously been once an option to use or not to use the Divine name – insofar as it is useful to count the absolute number of variants as Lange does. Among those 113 cases, the Divine name is attested by 𝔐 91 times, by ⅏ 96 times, by 𝔊 88 times and by the Vulgate 14 times. Which means that the Vulgate shows a strong preference for shorter readings against all other witnesses. As shown in the chart, on the one hand, 72 of the 73 exclusive readings of the Vulgate are shorter readings, i.e., implying the absence of Dominus. On the other hand, the only exclusive presence of Dominusis caused obviously by confusion of ‫ יהוה‬with ‫יהיו‬.71 With this preference in mind, one has to admit the chance of random agreement with exclusive readings of the other witnesses. As shown in the chart, there are four cases in which 𝔐 shows exclusive readings against the combined witness of ⅏ and 𝔊. It is noticeable that 70

71

Lange counts the agreements for 40 Qumran Pentateuchal manuscripts in such a way and gives the following classifications (including the named Deut manuscripts): 2 “protomasoretisch” (4QDeutᵉ·ᵍ; this is untenable for 4QDeutᵉ which never reads with 𝔐 against ⅏ 𝔊 [F1], but once with 𝔊 against 𝔐 ⅏ [F3]), 2 “präsamaritanisch” (no Deut manuscript), 2 “semimasoretisch” (no Deut manuscript in this dubious category), 2 “Handschriften, die der hebräischen Vorlage von 𝔊 nahestehen” (among them 4QDeutq – strictly speaking, no Deut manuscript), 11 “stehen 𝔐 und ⅏ gleich nahe” (among them 4QDeutᵈ·ᶠ·i·ᵒ; 5QDeut; this is untenable for 4QDeuti which follows 𝔐 in allcases where the texts of 𝔐 and ⅏ differ [3×F1, 2×S2]); and 21 (“52.5 %”) are classified as “eigenständig” (“independent”) because of the high number of “independent” readings (among them 1QDeutᵇ; 4QDeutᵇ·ᶜ·ʰ·ʲ·ᵏ¹·ᵏ²·n, 4QpaleoDeutʳ) which is a misleading terminology (Lange, Handbuch, 169). Obviously, the designation as a so-called “independent” manuscript means nothing for its stemmatic relationship, apart from the fact that dependence of 𝔐 or ⅏ or 𝔊 on one of these manuscripts is excluded because of their singularity variants (S4). I will show that all these manuscripts insofar they show enough variants (this is not the case for 4QDeutᵏ²·ᵒ) and are not excerpted ones (like 4QDeutʲ·ᵏ¹·q; 5QDeut) can be related with the concrete textual prehistory of 𝔐, ⅏ and 𝔊. Deut 12:26 [1].

A STEMMA FOR DEUTERONOMY

155

the Vulgate agrees with 𝔐 in all four cases. The agreement with the two longer readings of 𝔐 is especially striking. In contrast, the Vulgate supports none of the nine exclusive readings of ⅏. Finally, the Vulgate seems to agree with 𝔊 at first glance in 12 cases (5 presence, 7 absence), but four of the five agreements in presence of Dominus / κύριος are due to the translation of ‫אדני‬, so only 8 agreements (1 presence, 7 absence) remain. Five of the seven agreements in absence of the Divine name show in fact different changes made in 𝔊 and Vulgate independently (in three cases 𝔊 uses θεός against an omission in Vulgate, in two cases 𝔊 and Vulgate omit different strings of text).72 So real agreement can be stated in but 1 of the 12 cases where the Divine name is presupposed exclusively in 𝔊,73 and in 2 of the 15 cases where the Divine name is exclusively absent from 𝔊.74 72 73

74

See above the notes to the chart for details. Deut 31:11 [2]. Since this is the only variant inconsistent with the supposed stemmatic position of the Vulgate, discussion is necessary. Deut 31:11 is the last record of the formula for the chosen place in Deut, a formula which is followed by the Divine name as the subject of the election in most cases. 𝔐 lacks the Divine name in three of the 21 records of this formula (14:23; 16:16; 31:11), whereas 𝔊 has the Divine name in all 21 cases. ⅏ and the Vulgate both agree with 𝔐 in the 18 cases where the Divine name is present, and agree also in the absence of the Divine name in two of the three remaining cases, while in one case each of these show the lectiofaciliorwith the Divine name present in the formula (⅏: 14:23; Vulg.: 31:11). This distribution is most easily explained with the assumption that 𝔐 shows the original readings, and that the other witnesses added the Divine name, whether systematically (𝔊) or not (⅏, Vulgate). Otherwise it would be very difficult to explain the agreement between ⅏ and 𝔐 in the absence of the Divine name in 31:11. This means that the agreement of Vulgate with 𝔊 in 31:11 most likely came into being by accident. Another aspect in favour of this is the exclusive reading of the Vulgate at the end of the verse: for ‫התורה‬, which is attested by all other witnesses, the Vulgate has not the usual rendering legem, but shows a longer reading, verbalegis, which usually renders ‫דברי התורה‬. Seemingly, since the addition of Dominusand verbain the context “inloco quemelegeritDominuslegesverbalegishuius” does not affect the content of the verse, it simply served as clarification to avoid the possible misreading of a sequence of three very similar Latin words “… elegeritlegeslegem…” which would render exactly the words ‫יבחר תקרא את התורה‬. The named two “real” agreements (Deut 6:19; 21:5) likely arose by accident, since there is different motivation for the omission in 𝔊 and Vulgate: in the first case, Deut 6:19, the Vulgate renders ‫ כאשר דבר יהוה‬by sicutlocutusestin Deut 6:19 as well as in Deut 9:3 according to its shortening tendency, so assimilating to shorter formulations like Deut 1:11 or 12:20. On the contrary, the Hebrew Vorlageof 𝔊 seems to have shortened by parablepsis, since the hyparchetype γ read ‫ יהוה‬at the end of Deut 6:19 immediately followed by ‫ והיה‬at the beginning of Deut 6:20 (‫ והיה‬being an addition of γ witnessed by ⅏ and 𝔊 [Καὶ ἔσται] against α, witnessed by 𝔐 and Vulgate). In the second case, Deut 21:5, 𝔊 (καὶ εὐλογεῖν ἐπὶ τῷ ὀνόματι αὐτοῦ) or its Vorlageassimilates towards the reading of γ in Deut 18:5 which is a longer reading against 𝔐 and Vulgate. Vulgate again shortens in accordance to its general tendency. Deut 21:5 is just one of many cases where the Vulgate omits the second appearance of the Divine name in a verse (cp. Deut 1:45; 3:20.21.26; 4:3.21; 5:5.11.27.28 etc.), a tendency not paralleled in 𝔊.

156

BENJAMIN ZIEMER

This is approximately what can fairly be expected by chance, bearing in mind the tendency of the Vulgate to shorten the text by omitting the Divine name. It can therefore be concluded with certainty that the Vulgate is a witness to the 𝔐 branch of the stemma. The abundant number of isolated readings of the Vulgate does not affect this conclusion – but it excludes the theoretical possibility that 𝔐 depends on the Vulgate. I treated the Vulgate as an example since it is generally accepted that Jerome used a Hebrew Vorlagevery close to, if not identical with, 𝔐 for his translation. Therefore, it is not necessary to include further types of variants into the discussion. They will show the same picture: the Vulgate shows a lot of isolated, especially shortened, readings. But if there is a deviation between 𝔐, on the one hand, and the combined testimony of ⅏ and 𝔊, on the other hand, the Vulgate supports these isolated readings of 𝔐 in most cases.75 So, the pure number of exclusive readings alone says nothing about the age or the value of a textual witness, nor about its position in the stemma codicum. But if the number of such exclusive readings is relatively high, we can learn something about the tendency of the witness – e.g., shortening, expanding, assimilating – and therefore draw conclusions as to how far certain agreements with other witnesses may have arisen by accident or show a family relationship. This is very important for the evaluation of the testimony of 𝔊. We know that it is one of the earliest witnesses, and we know that it shows more exclusive variants than ⅏ or 𝔐. The evidence discussed so far shows that these variants include additions, substitutions and omissions, with a slight preference for longer readings. The inclusion of the Qumran 75

One example is the formula ‫ מצוה אתכם‬/ ‫אשר אנכי מצוך‬, attested 32 times by the three main witnesses of Deuteronomy. In most cases, it is followed by ‫היום‬/σήμερον/hodie; for every attestation of the formula there is at least one witness for this longer reading. It appears in 𝔐 24 times, in ⅏ 30 times (including an additional formula in the Garizim commandment in Deut 5:18B, taken from Deut 27:4), in 𝔊 30 times, but in the Vulgate only 18 times. If we compare the attestations of the former three witnesses, ‫ היום‬or its equivalent is present in all three witnesses in 24 cases; it is present both in 𝔊 and ⅏ but absent from 𝔐 in three cases (Deut 6:2; 11:22; 13:1); it is present exclusively in ⅏ in three cases (Deut 4:2 [1]; 12:28; and the additional formula in 5:18B); and exclusively in 𝔊 in three cases (Deut 4:2 [2]; 12:11, 14). So there are nine differences between 𝔐 on the one hand and 𝔊 and/or ⅏ on the other. The Vulgate agrees in all these nine differences with the shorter reading of 𝔐, which confirms the family relationship between these two witnesses. Nevertheless, beyond 𝔐, the Vulgate omits hodie three times (Deut 4:40; 11:27; 27:10), omits the complete formula two times (28:14; 30:16), and in other cases the formula is translated so freely that it is no longer recognisable (4:2 [1]; 11:28; 12:11).

157

A STEMMA FOR DEUTERONOMY

evidence will show that these variants did not arise at once but came into being at several stages. Against it, the common ancestor of ⅏ and 𝔊, γ, seems to show a slight tendency towards harmonisation by addition. ⅏, in turn, shows no specific tendency beside the themes of the chosen place and the promised land. The example of the Vulgate shows that it is not possible nor necessary to explain every deviation of a certain witness, whether additions or omissions. In contrast, it is crucial to explain the agreements between certain witnesses. Ideally, the vast majority of variants can be explained in accordance with the stemmatic relationship, i.e. as inherited readings or changes made by the witness itself, whereas only rare exceptions demand other explanations.

4. PLACING

THE

QUMRAN MANUSCRIPTS IN THE

STEMMA CODICUM

If we turn to the Qumran manuscripts, far less variants will be available by reason of their fragmentary state. So we will have to include variants below word level. Anyway, I will focus on the “family variants” (F1, F2, F3) as the most important and only decisive for stemmatic relationship: agreements of a specific Qumran manuscript with one of the major witnesses against the two other major witnesses (F1= exclusive agreement with 𝔐, F2 = exclusive agreement with ⅏, F3 = exclusive agreement with 𝔊). In theory, a manuscript which is not contaminated should have such family readings either with one or with none of the major witnesses. In practice we speak about very small variants, mostly not more than one word, so that we have to concede that agreement by accident is possible – as an exception. We have to reckon with this last possibility especially for two types of minor variants: both the presence vs. absence of ‫ו‬-copulativum and the use of second person in the plural vs. singular are types of variants where any scribe had exactly two options and no direction of change is more or less plausible; therefore such changes can easily be made twice independently or reversed by accident. In contrast, an addition, omission or exchange of a content word is, as a rule, significant. If the fragment is very small and contains only two or three minor variants, it is impossible to place it in the stemmawith certainty. So, if we have got only one minor (e.g., presence vs. absence of ‫ו‬-copulativum) family reading within the variants covered by a fragment, we cannot know whether it confirms the rule or happens to be an exception.

158

BENJAMIN ZIEMER

4.1. 4QpaleoDeuts (4Q46), showing 1 variant: S3 E.g., the oldest Deut manuscript, 4QpaleoDeutˢ (4Q46)76 is at the same time one of the smallest fragments. ‫ל[מ ֯ת ֯שמעתי בקול י‬ ̇ line 1 (Deut 26:14) ‫ ממע[ו̇ ן קדשך מן השמי‬line 2 (Deut 26:15) ‫ נתתה [לנו כאשר נשבע‬line 3 (Deut 26:15) ⟦ ⟧ […

line 4 (break)

“Both the text and the orthography agree with those of 𝔐 and ⅏ for the six complete and five partial words that are extant.”77 Not noted in the DJD edition, there is one variant against 𝔊 in Deut 26:15: ‫ לנו‬agrees with 𝔐 and ⅏ against ‫ להם‬or ‫ למו‬required by 𝔊 (ἣν ἔδωκας αὐτοῖς instead of ‫א ֶ ֥שׁר נָ ַ ֖ת ָתּה ָל֑נוּ‬, ֲ S3). (1) S3 (singularity reading of 𝔊): minor (suffix) Deut 26:15 ‫ לנו‬4QpaleoDeutˢ 𝔐 ⅏ 〛 ‫𝔊 להם‬

Unfortunately, we cannot be sure about where to place this manuscript in the stemma. It is dated by the DJD editors to the third c. BCE. With the new palaeographic classification by Michael Langlois, who hypothesises for this manuscript a date about 400 BCE,78 it could theoretically be identical with the archetype itself or one of its earliest copies – α, β, or γ in our stemma. Anyway, the one variant is quite suggestive: the earliest extant Deuteronomy manuscript underlines the isolation of a 𝔊 reading which is already determined as secondary through the agreement of 𝔐 and ⅏ against it. If there is no textual variance, the agreement in orthography is also of interest. It shows that the orthography of 𝔐 and ⅏ with its mixture of full and defective readings is not a result of later systematisation but a true reflection of orthographical practice in the fourth or third c. BCE. There were no less than six possibilities to opt for another orthography in the eleven words and parts of words preserved in the manuscript, and in every case the orthography is in accordance with 𝔐 and ⅏: first, ‫ָשׁ ַמ ְע ִתּי‬ could have been written defective as ‫ – שמעת‬such archaic orthography is attested in inscriptions (Kuntillet Ajrud 18:1 ‫ברכת‬, Murabbaat 1:1 ‫)שלחת‬ 76

77 78

Publication: Patrick W. Skehan, Eugene Ulrich and Judith E. Sanderson, DJD IX, pp. 153–154 and pl. XXXVII. DJD IX, 153. See his article in the present volume, pp. 275ff.

A STEMMA FOR DEUTERONOMY

159

as well as in 𝔐ᴷᵉᵗ (1 Kings 8:48 ‫ ;בנית‬Ezek 16:59 ‫)ועשית‬, and more frequently in suffixed forms in 𝔐 (Deut 9:12 ‫יתם‬ ִ ִ‫)צוּ‬. ִ Second, a fuller orthography was a option at least for other Qumran manuscripts in the following cases: ‫ ֵמת‬could have been written ‫( מית‬so regularly in 1QIsaa, e.g. Isa 38:1). ‫ ָק ְד ְשָׁך‬is consistantly written by 1QIsaa with ‫ו‬, so ‫קודשך‬ (Isa 63:18), compare the irregular plenewriting of qamaṣqatanin some cases in 𝔐ᴷᵉᵗ (e.g., Deut 32:13 ‫)במותי ארץ‬, and the suffix could have been written plene like ‫ קודשכה‬again in 1QIsaa (Isa 63:15), compare some cases in 𝔐 (Deut 28:22ff. ‫יַ ְכּ ָכה‬, Ex 13:16 ‫)יָ ְד ָכה‬. Thirdly, there are two cases of plene writing not required necessarily: ‫ ְבּקוֹל‬and ‫ ִמ ְמּעוֹן‬are written plene in Deut 26:14f. by 𝔐 ⅏ 4QpaleoDeutˢ, but 𝔐 shows at other places the defective writing for ‫ ַהקֹּל( קול‬Gen 27:22; ‫ ְלקֹל‬Ex 4:8; ‫ וְ קֹל‬Ex 19:16) and for words similar to ‫ ְב ָמקֹם( ָמעוֹן‬Ex 29:31; ‫ ַל ָמּאֹר‬Ex 26:6; ‫ ִמ ְמּקֹר‬Lev 12:7). The latter cases are especially instructive, since other Qumran manuscripts, like 4QDeutᵈ or 4QSamb, use the materlectionis ‫ ו‬for the word internal vowel ofar less frequently. The testimony of 4QpaleoDeutʳ shows that in cases where 𝔐 and ⅏ agree in matters of orthography and morphology, this agreement may well be inherited from the first Pentateuchal manuscripts, or from the archetype itself. Moreover, the age of this fragment found in a Qumran cave together with much younger manuscripts remembers us that such biblical manuscripts written on leather may have been in use for several centuries. Therefore we have to reckon with the possibility that manuscripts written in the first c. CE copied manuscripts even from the Persian era. 4.2. 4QDeutᵃ (4Q28), showing five variants, the largest among them showing exclusive agreement with 𝔊 (F3) For the second oldest extant Deuteronomy manuscript, 4QDeutᵃ (4Q28)79 from the first half of the 2nd c. BCE, the DJD edition states only five variants. Among them are two variants at the word level – one shorter reading with ⅏ and 𝔐 against an addition by 𝔊 (S3), thus strengthening the definition of the original text, and one shorter reading shared with 𝔊 against ⅏ and 𝔐 (F3) at a place with clear occasion of homoioteleuton. So it seems plausible that this manuscript is related to 𝔊 but the number of variants is too small to bear an unequivocal decision. 79

Publication: Sidnie White Crawford, DJD XIV, pp. 7–8 and pl. I. Cp. White, Special Features, 158–60; id., Three Deuteronomy Manuscripts, 23–28.

160

BENJAMIN ZIEMER

(1) F3 (exclusive agreement with 𝔊): major (absence of two words) Deut 24:1–2 ‫ והלכה‬2 ‫ ו[שלחה מביתו‬4QDeutᵃ 𝔊 〛 ֽ ‫⅏ 𝔐 וְ ִשׁ ְלּ ָ ֖חהּ ִמ ֵבּ‬ ‫יתוֹ וְ ָה ְל ָ ֖כה‬ ֑ ‫ וְ יָ ְצאָ֖ה ִמ ֵבּ‬2 ‫יתוֹ׃‬

This Minus of two words, ‫ויצאה מביתו‬, is clearly the most important variant within the few lines of this manuscript. The content is not affected, because it is clear without this short sentence that the divorced woman leaves the house of her former husband. So it cannot be a clarification, but merely seems to be an unintentional variant. The unintentional adding of a sentence is not very likely, and it is next to impossible that two scribes – the parent texts of ⅏ and 𝔐 in our stemma – independently added this sentence in the same manner. The unintentional omitting of two words is by far the easiest explanation. There is a clear cause for parablepsis: ‫שלחה מביתו ויצאה מביתו והלכה‬. The word ‫ מביתו‬is twice preceded by a ‫ ה‬and followed by a ‫ ו‬so the eye could easily jump from the first ‫ מביתו‬to the second. Since the fragment is very small (about 100 words) we cannot say if its scribe was especially prone to parablepsis. If this was the case, perhaps the agreement with 𝔊 came into being accidentally. But it is more plausible that both witnesses are genetically related. So this reading helps us to place 4QDeutᵃ in the stemma, tentatively: it had a common parent text with 𝔊, from which ⅏ and 𝔐 are independent. (2) S3 (singularity reading of 𝔊): major (absence of one word) Deut 24:4 ‫ יהוה‬4QDeutᵃ 𝔐 ⅏ 〛 ‫𝔊 יהוה אלהיך‬

This variant shows that our manuscript did not share all variants with 𝔊. The addition of ‫ אלהיך‬is a very frequent one in the Greek translation of Deuteronomy. There are a number of further minor agreements between 4QDeutᵃ, 𝔐 and ⅏ against 𝔊 which are not mentioned in the DJD series. They do not affect the relatedness of 4QDeutᵃ with 𝔊 but show that the parent text of 4QDeutᵃ has to be placed in the stemmabetween the Vorlageof ⅏ and the immediate Vorlageof 𝔊. (3) F1 (exclusive agreement with 𝔐): minor (change in number), ⅏ 𝔊 at variance Deut 24:4 ‫ תחטא‬4QDeutᵃ ‫יא‬ ֙ ‫ 〛 𝔐 ַת ֲח ִט‬80 ‫ ;⅏ תחט]י[או‬μιανεῖτε 𝔊 (= ‫)תטמאו‬ 80

The Samaritan reading cited in DJD is ‫ תחטיאו‬according to v. Gall, against ‫ תחטאו‬in Tal’s edition. The variance is easy to be explained since according to the Samaritan pronunciation tradition one can understand the form tāṭṭiyyu either as Piel (regularly written defectively) or as Hifil(regularly written plene), cp. Ben-Hayyim, Tradition.

A STEMMA FOR DEUTERONOMY

161

All three main witnesses are at variance, so the text of γ is unknown. However, 4QDeutᵃ agrees in number with 𝔐 against ⅏ and 𝔊. This is a very frequent type of variant in Deut, which can easily arise independently by chance: at the end of the verse the sec. m. singular (‫אלהיך‬, ‫)לך‬ prevails, hence 4QDeutᵃ possibly changed a plural form found in its parent text towards the singular. (4) F3 (exclusive agreement with 𝔊): minor (absence of a proclitic preposition) Deut 24:5 ‫ כל‬4QDeutᵃ 𝔊 〛 ‫⅏ 𝔐 לכל‬

Again there is agreement with 𝔊 in a shorter reading congruent with the stemma. (5) S3 (singularity reading of 𝔊): minor (presence of syndesis) Deut 24:8 ‫ ול]עשות‬4QDeutᵃ 𝔐 ⅏Tal 〛 ‫⅏ לעשות‬v.Gall 𝔊

This variant appears in the DJD edition as exclusive agreement with 𝔐 since von Gall’s edition is used as citation standard. But Tal’s edition of ⅏ shows the reading ‫ולעשות‬, like 𝔐 and 4QDeutᵃ. In fact the whole phrase is different in 𝔊: φυλάξῃ σφόδρα ποιεῖν seems to render ‫ְשׁמֹר מאד‬ ‫ לעשות‬or ‫ ִתּ ְשׁמֹר מאד לעשות‬rather than ‫ לשמור מאד )ו(לעשות‬as present in 4QDeutᵃ, 𝔐 and ⅏. Anyway, the presence of the ‫ ו‬seems to be a lectio difficilior, suggesting by itself to omit the ‫ ו‬as is supposed by 𝔊 as well as in many Samaritan and some Masoretic manuscripts. Since omissions, especially by homoioteleuton,are not typical for δ as determined below by other Deuteronomy manuscripts, I place this manuscript tentatively as an offspring of ε, halfway between γ and 𝔊, but retaining a question mark. 4.3. 4QDeutᵇ (4Q29), showing 17–18 variants, among them 6 exclusive agreements with 𝔊 (F3) The next manuscript, 4QDeutᵇ (4Q29),81 also from the 2nd c. BCE, shows more variants and can therefore be placed in the stemmawith more certainty. For this manuscript the DJD edition lists 18 variants (I give them numbers according to their order in DJD), four of them are unanimously to be classified as “family variants”: 81

Publication: Julie Ann Duncan, DJD XIV, pp. 9–14 and pl. II.

162

BENJAMIN ZIEMER

(8) F3 (exclusive agreement with 𝔊): major (presence of one word) Deut 30:11 ‫[רחוקה היא ממך‬ ֯ ‫ ולא‬4QDeutᵇ 𝔊〛 ‫𝔐 וְ ֥ל ֹא ְרח ָ ֹ֖קה ִ ֽהוא‬, ‫ולא‬ ‫⅏ רחקה היא‬

A word absent from ⅏ or 𝔐 but present in 4QDeutᵇ and translated by 𝔊 is a first argument for placing this manuscript on the 𝔊 side of the stemma, tentatively. (10) F3 (exclusive agreement with 𝔊): major (presence of one word) Deut 30:14 ‫ ובידך‬4QDeutᵇ 𝔊 (καὶ ἐν ταῖς χερσίν σου ‫> 〛 )?וביד]י[ך‬ ⅏𝔐

The context is, in 𝔐 and ⅏, “The word is very close to you, in your mouth and in your heart (‫ )בפיך ובלבבך‬to do it.” 4QDeutᵇ 𝔊 add to this list that the word is “in/by your hand(s),” a clear bridge to doing the word, whereas the physical presence of the written form of the word in the tefillinis usually expressed by ‫( על ידך‬Ex 13:9, 16; Dtn 6:8). (12) F3 (exclusive agreement with 𝔊): minor (number) Deut 31:11 ‫ תקראו‬4QDeutᵇ 𝔊 (ἀναγνώσεσθε) 〛 ‫𝔐 ִתּ ְק ָ ֞רא‬, ‫⅏ יקרא‬

This reading shows agreement with 𝔊 and 𝔐 against ⅏ in the second person against the third person, and exclusive agreement with 𝔊 in number. Since change in number is very frequent in Deuteronomy, this agreement is not decisive for family relationship, but it fits in the pattern shown by the two former variants. (17) F3 (exclusive agreement with 𝔊): major (presence of [at least] one word) Deut 31:28 ‫ זקניכם[ ושפטיכם‬4QDeutᵇ 𝔊 〛 > 𝔐 ⅏

Again there is one word clearly present (another reconstructed) in the manuscript which is presupposed by 𝔊 but absent from ⅏ and 𝔐, so this variant confirms the exclusive relationship between 4QDeutᵇ and 𝔊. In the DJD edition, another variant is designated as exclusive agreement of 4QDeutᵇ with 𝔐 (in our terminology: F1) against the combined witness of ⅏ and 𝔊. A closer examination shows that this is not tenable: (14) S2 (singularity reading of ⅏): minor (grammar) Deut 31:26 ‫ לקח‬4QDeutᵇ, ‫𝔐 ָלק ַֹח‬, cp. Λαβόντες 𝔊 (?) 〛 ‫⅏ לקחו‬

This is a very instructive variant. 𝔐, supported by 4QDeutᵇ, uses an infinitivusabsolutusin the function of an imperative, a feature restricted to classical Hebrew. From the context it is clear that a plural is meant

A STEMMA FOR DEUTERONOMY

163

because of the following ‫𝔊 ;⅏ = 𝔐( וְ ַשׂ ְמ ֶ ֣תּם‬: θήσετε). Neither the Samaritan nor the Greek grammars know about an infinitivusabsolutus. 𝔊 frequently renders the infinitivus absolutus as a participle, including here, apparently (understanding ‫ לקח‬as a modal verb). ⅏ reads an imperative albeit a very rare form of it: lēqā᾿u (‫ )לקחו‬with the lamed even though this letter is regularly disappearing in the qal imperative of ‫לקח‬. The regular form in Samaritan Hebrew is qā᾿u, written ‫( קחו‬equivalent to ‫ְקחוּ‬ in 𝔐).82 So ‫ לקחו‬here is clearly a singularity reading of ⅏ and not at all likely as Vorlageof 𝔊.83 Among the remaining variants stated in the DJD edition there are two further (minor) cases with an exclusive agreement between 4QDeutᵇ and 𝔊 not classified in the DJD edition as such: (6) F3 (exclusive agreement with 𝔊): minor (number) Deut 30:10 ‫ הכתובים‬4QDeutᵇ 𝔊 〛 ‫⅏ 𝔐 הכתובה‬

𝔊 has τὰς γεγραμμένας, agreeing in number with 4QDeutᵇ. However, this agreement may have arisen by chance, since the preceding noun in 𝔊, καὶ τὰς κρίσεις αὐτοῦ (‫)ומשפטיו‬, governing sex and number of the participle, has no equivalent in 4QDeutᵇ. Therefore it is justifiable to classify this as an inner-Hebrew singularity reading, as the DJD edition does. (15) F3 (exclusive agreement with 𝔊): minor (sex of demonstrative pronoun) Deut 31:26 ‫ הזאת‬4QDeutᵇ 𝔊 〛 ‫⅏ 𝔐 הזה‬

After ‫ספר התורה‬, the feminine form of the demonstrative pronoun corresponds to ‫“( התורה‬the book of this Torah”), the masculine form to ‫ספר‬ (“this book of Torah”). 𝔊 has τὸ βιβλίον τοῦ νόμου τούτου, “the book of this law.” τούτου agrees definitely, in sex and case, with νόμου, and not with βιβλίον. So we have got six exclusive agreements of 4QDeutᵇ with 𝔊, among them three major variants (8, 10 and 17). These three pluses of at least one word shared by 4QDeutᵇ and 𝔊 within the 175 words covered by this manuscript lead to a clear conclusion: since it is most plausible that 𝔐 and ⅏ inherited the agreement in their shorter readings by their parent texts, the additional words were not yet present in γ (the common ancestor of ⅏ and 𝔊) but only later, in δ, a common ancestor of 4QDeutᵇ and 𝔊 on which ⅏ is not depending. 82 83

I am indebted to Stefan Schorch for introducing me to Samaritan grammar. By the way, most antique and modern translations from the Hebrew translate an imperative plural here, so one cannot see if they had an absolute infinitive before them or an imperative.

164

BENJAMIN ZIEMER

The remaining twelve variants are congruent with this stemma. They are distributed as follows: 6 majority readings of 4QDeutᵇ isolate either 𝔐 (S1, once)84, ⅏ (S2, twice)85 or 𝔊 (S3, three times),86 thereby determining the text of γ and δ. Two exclusive readings of 4QDeutᵇ not shared by any other witness (S4)87 exclude the dependence of 𝔊 on the text of 4QDeutᵇ. 2 inner-Hebrew variants shared by ⅏ (H2)88 determine the orthography of γ and δ, whereas the only exclusive inner-Hebrew variant (H4)89 does not give any additional information for the stemma, as does the one inner-Greek variant mentioned in DJD.90 That alltextual variants witnessed by 4QDeutᵇ are congruent with its place in the stemmais even more than can fairly be expected. In addition, the stemmatic method uncovers the individual characteristics of a manuscript: it seems clear that 4QDeutᵇ by itself was not prone to additions, since both singularity readings of this manuscript (S4) are minuses against all other witnesses. All pluses of 4QDeutᵇ against 𝔐 or ⅏ are in fact inherited additions of δ against γ. We have thus made a serious step forward to a stemma codicum of Deuteronomy. 4QDeutᵇ turns out to be a close relative of 𝔊, but nonetheless there must have been a manuscript δ between γ and the Vorlageof 𝔊. The text of δ is determined by any agreement of 4QDeutᵇ with any main witness, but in cases of isolated readings of 4QDeutᵇ it is determined by the agreement of 𝔊 with ⅏ and/or 𝔐. δ shows an inclination to additions against its parent text γ. 84

85

86

87

88

89 90

(3) Deut 30:9, changed arrangement of a list. The order ‫בפרי אדמתך ובפרי[ בהמתך‬, shared by 4QDeutᵇ, ⅏ and 𝔊, determines the text of γ and δ against 𝔐 (‫)ובפרי בהמתך ובפרי אדמתך‬. (11) Deut 31:11, ‫ יבחר‬against ‫בחר‬. Thereby 4QDeutᵇ confirms the common reading ‫יבחר‬ as that of α, β, γ and δ, isolating the variant reading of ⅏ which recurs on the Garizim commandment in the Samaritan version of the Decalogue. For variant (14), ‫ לקח‬vs. ‫לקחו‬ in Deut 31:26 (grammar), see above. (4) Deut 30:9, isolated minus of ‫ לטובה‬in 𝔊. (5) Deut 30:9, isolated plus of ‫ ומשפטיו‬presupposed by 𝔊. (16) Deut 31:27, changed designation of God, ‫ האלהים‬presupposed by 𝔊 instead of ‫יהוה‬. These variants show that 4QDeutᵇ is not dependent directly on the Vorlageof 𝔊 but on an earlier parent text, called δ in our stemma. (1) Deut 29:24, absence of ‫ אשר‬before ‫ כרת‬in 4QDeutᵇ. (13) Deut 31:15, absence of ‫ על‬before ‫ פתח‬in 4QDeutᵇ. (7) Deut 30:11, and (9) Deut 30:13, 4QDeutᵇ has twice ‫ היא‬agreeing with 𝔐Qere, against ‫𝔐 הוא‬Ketiv. Actually, these are orthographic variations that should not be counted as textual variants. (18) Deut 32:3 ‫ גדול]ה‬4QDeutᶜ 〛 ‫ ;⅏ 𝔐 גדל‬this may be an orthographic or a textual variant. (2) Deut 30:4. A number of old and weighty Greek witnesses read an additional κύριος ὁ θεός σου (= ‫ )יהוה אלהיך‬at the end of the verse which is therefore stated in DJD as variant. But in this case the Göttingen edition differs from Rahlfs’, so the short reading of 4QDeutᵇ is shared not only by 𝔐 ⅏ but also by 𝔊GÖ against 𝔊ᴿᴬ.

A STEMMA FOR DEUTERONOMY

165

4.4. 4QDeutᶜ (4Q30), showing more than 80 variants, among them 11 exclusive agreements with 𝔊 (F3) 4QDeutᶜ (4Q30)91 is the largest extant manuscript of Deuteronomy from Qumran, and also one of the oldest, from the second half of the second c. BCE.92 Therefore this is a very good test case for our stemma. Sidnie Ann White (Crawford), the editor of this manuscript in DJD, states explicitly: “4QDtᶜappears to be a manuscript located in the textual tradition of 𝔊. In cases of shared error,93 it agrees with 𝔊 most often against either 𝔐 and/or ⅏, although there is some overlap among the witnesses. In most of these cases, it might be argued that the errors could have arisen independently and do not point to recensional differences. However, we have five cases of explicating plus which are unlikely to arise independently and to be identical at the same time.”94

Since there are more than 80 variants listed in the DJD edition I will not name them all but concentrate on the family readings and the quantitative variants. I do not agree with White when she judges on superiority, inferiority or errors. Nevertheless, she based her argument on some exclusive agreements. Let us take an example: F3 (𝔊-family): major (presence of one word) Deut 31:28 ‫דב[רי השירה ]הזאת‬ ֯ ]‫את‬ ̇ 4QDeutᶜ = 𝔊 〛 ‫ירה ַה ֔זּ ֹאת‬ ֣ ָ ‫ת־ה ִשּׁ‬ ַ ‫ֶא‬ 𝔐=⅏

The last two letters of the word ‫ דברי‬are clearly visible. An equivalent is present in 𝔊 (τὰ ῥήματα τῆς ᾠδῆς ταύτης) but absent in ⅏ and 𝔐, so this variant is likely to be explained with an exclusive relationship between 4QDeutᶜ and 𝔊. Since this is a small scale addition, the common ancestor of 4QDeutᶜ and 𝔊 may well be the hyparchetype δ postulated as the common ancestor of 4QDeutᵇ and 𝔊. There are, in my counting, 11 F3 cases in which 4QDeutᶜ agrees exclusively with 𝔊 against the common text of ⅏ and 𝔐 (8 times completely, 3 times partially).95 On the word level, there are 4 cases of 91

92 93

94 95

Publication: Sidnie White Crawford, DJD XIV, pp. 15–34 and pls. III–IX. Cp. White, Special Features, 160–62. DJD XIV: “c.150–100 BCE”, White, Special Features, 161: “ca. 125 BCE”. We cannot be sure what is an erroneous or a correct reading. In our terminology these are family readings, exclusively shared by 4QDeutᶜ and 𝔊. White, Special Features, 162. The eleven exclusive agreements of 4QDeutᶜ and 𝔊 (F3) against the readings of γ defined by the agreement of ⅏ and 𝔐 are:

166

BENJAMIN ZIEMER

agreement in pluses of one word,96 and 1 case of agreement in a minus of two words (‫ ואת רעהו‬in Deut 15:2).97 White concludes: “… there is both positive and negative evidence to suggest that 4Q belongs in the textual tradition of 𝔊. […] Therefore we believe that 4QDtcis related to the tradition of the Old Greek of Deuteronomy.”98

In our stemmathese agreements do not establish the original text, but the text of the Vorlageof the Old Greek (ζ) and that of the common parent text of 4QDeutᶜ and 𝔊 (δ). If the placing of 4QDeutᶜ within the “family” of the Old Greek is correct, we would expect that in cases where 𝔊 and ⅏ agree against the reading of 𝔐, our manuscript will share these readings too, as a rule. The reason is to be seen in the stemma: if ⅏ and 𝔊 inherited a shared reading, they got this reading from their parent text. So the manuscripts γ and δ are defined by such agreement, as a rule – and these are ancestors of 4QDeutᶜ. Now, there are 10 such agreements of ⅏ and 𝔊 against 𝔐 in the text covered by 4QDeutᶜ. Among them we find 4 quantitative differences concerning at least one word – in all these cases 𝔊 and ⅏ witness the longer text – and 6 minor differences. In 3 minor cases the agreement between

96 97 98

(1) absence of the preposition ‫ ל‬before ‫ נסותך‬in Deut 8:2 (but partially disagreement: 𝔊 adds καὶ); (2) presence of ‫ ו‬before ‫ לדעת‬in Deut 8:2; (3) absence of the suffix of ‫ אדמתך‬in Deut 12:19; (4) absence of ‫ ואת רעהו‬and agreement in word order in Deut 15:2; (5) presence of ‫ בו‬after ‫ לא תעשה‬in Deut 16:8; (6) agreement in word order ‫ו[ל ֯תהלה ֯ולתפ]ארת‬ ֯ in Deut 26:19 (𝔊: καὶ καύχημα καὶ δοξαστόν) against ‫ ⅏( לתהלה לשם ולתפארת‬and 𝔐); (7) presence of the demonstrative pronoun (‫ )הזאת‬after ‫ המצוה‬in Deut 27:1 (𝔊: πάσας τὰς ἐντολὰς ταύτας; 𝔊 always translates ‫ כל המצוה‬as plural in Deut [Deut 5:31; 6:25; 11:8, 22; 15:5; 19:9; 27:1; 31:5], so the difference in number may be traced back to the translator); (8) replacement of √‫ רבץ‬by √‫𝔊( דבק‬: κολλάω, cf. Deut 10:20; 28:60) in Deut 29:19 (but disagreement in number); (9) presence of ‫ יהוה‬before ‫ אלהי‬in Deut 31:17; (10) presence of ‫ דברי‬before ‫ השירה‬in Deut 31:28 (see above); (11) agreement in number: ‫[ =( שימוה‬καὶ] ἐμβαλεῖτε αὐτὴν; the words cited by DJD, καὶ διδάξετε αὐτὴν, are in fact the rendering of ‫ )וְ ַל ְמּ ָ ֥דהּ‬against ‫ ⅏( שימה‬and 𝔐). Partial disagreement: the copula is present exclusively in 𝔊. Cases 5, 7, 9, 10 in the preceding note. Case 4. White, Special Features, 162. Tov, Contribution, 33, does not share this belief, but the distribution of the variants analysed in the stemmatic method supports White’s argument.

A STEMMA FOR DEUTERONOMY

167

the Vorlageof 𝔊 and 4QDeutᶜ remains unclear, therefore 7 cases remain with a clear situation. In 6 of these 7 cases 4QDeutᶜ agrees with 𝔊 and ⅏ as expected, among them in the largest such difference, the presence of three additional words in Deut 13:7. Only in one case 4QDeutᶜ does not support the shared reading of 𝔊 and ⅏ against 𝔐, since the additional ‫עוד‬ after ‫ לא יוסיפו‬in Deut 13:12 is not present in 4QDeutᶜ.99 Sidnie White wrote: “it might be argued that the errors could have arisen independently.” This is indeed correct. Better, we only need this last resort for 1 of 7 cases, where the “error” just seems to have arisen independently in 𝔊 and ⅏, and not in 4QDeutᶜ. Only one further proof is necessary for the accuracy of our stemmawith respect to 4QDeutᶜ. If 𝔐 and 𝔊 agree against ⅏, their common readings determine the text of the manuscripts α, γ, and δ. It is therefore to be expected that our manuscript, 4QDeutᶜ, agrees in those shared readings against ⅏, as a rule. There is one difference between ⅏ and the two other main witnesses in a passage covered by 4QDeutᶜ which goes beyond the variants discussed so far: the complete different order and wording of Deut 10:6–8 99

The agreements between ⅏ and 𝔊 against 𝔐 in the text covered by 4QDeutᶜ are the following (in brackets the classification and a remark on the congruence with the stemma): (1) plural ‫ ברגליך‬against singular ֖‫ ְב ַרגְ ְלָך‬in Deut 11:10; 4QDeutᶜ corrected towards ‫ברגלך‬, but it remains possible that 𝔊 translated such a defectively written form as plural (DJD: “orth. or var.”?, unclear situation); (2) presence of ‫ בן אביך או‬before ‫ בן אמך‬in Deut 13:7, supported by 4QDeutᶜ (= S1, congruent); (3) presence of ‫ עוד‬after ‫ יוסיפו‬in Deut 13:12, not supported by 4QDeutᶜ (= F1, incongruent; see below note 102); (4) presence of ‫ כל‬before ‫ מלאכה‬in Deut 16:8, supported by 4QDeutᶜ (= S1, congruent); (5) plural ‫ ואמרו‬against ‫ ואמר‬in Deut 27:26 (and in the eleven preceding verses), supported by 4QDeutᶜ which is extant only for this last case (= S1, congruent); (6) replacement of ‫𝔐 ישגלנה‬Ketib by an euphemism (⅏: ‫𝔊 ;ישכב עמה‬: ἕξει) in Deut 28:30, not supported by 4QDeutᶜ (]‫;)ישג[לנ֯ ה‬ ֯ since 𝔐Qere reads an euphemism too (‫ )יִ ְשׁ ָכּ ֶ֔בנָּ ה‬and there is no way to define exactly the written Vorlageof the Old Greek (‫ ישגלנה‬or perhaps something like ‫)?יש לה‬, so one can say nothing about the agreement or disagreement of 4QDeutᶜ with the Vorlageof 𝔊 (unclear situation); (7) plural of the verb ‫ )⅏( ורבצו‬or ‫𝔊( ודבקו‬, apparently) against the singular ‫ורבצה‬ (𝔐) in Deut 29:19, not supported by 4QDeutᶜ which read yet another form, ‫ודבקה‬ agreeing partially with 𝔐 and 𝔊 (unclear situation); (8) plural ‫ ועזבוני‬against ‫ ועזבני‬in Deut 31:16, supported by 4QDeutᶜ (= S1, congruent); (9) plural ‫ והפרו‬against ‫ והפר‬in Deut 31:16, supported by 4QDeutᶜ (= S1, congruent); (10) presence of ‫ מהם‬after ‫ פני‬in Deut 31:18, supported by 4QDeutᶜ while changing the numerus, ‫ =( ממנ֯ ]ו‬S1, congruent).

168

BENJAMIN ZIEMER

in ⅏ against 𝔐 and 𝔊. In this case, 4QDeutᶜ agrees with the common text of 𝔐 and 𝔊, counted as two variants by Sydnie White. There are five other such variants; only in one of them does 4QDeutᶜ agree with ⅏ against 𝔐 and 𝔊 in a change between waw and yod, which has to be explained by accident. In all other cases, among them the presence of the two words ‫מלך‬ ‫ מצרים‬in the common text of Deut 11:3, 4QDeutᶜ reads with the common text of 𝔐 and 𝔊 against ⅏ and confirms thereby the definition of γ in the stemma.100 So, the variants of the largest Qumran Deut manuscript can be explained with the stemmain a satisfying way. It is not possible to say how closely related 4QDeutᵇ and 4QDeutᶜ are. Both must have had a parent text anywhere between γ and the Hebrew Vorlage of 𝔊. Had they been preserved completely, we would likely have seen that one of these two manuscripts was closer to or more distant from γ. Since this is not the case we can but say there must have been a manuscript δ which underwent the changes shared by 4QDeutᵇ and 𝔊 and those shared by 4QDeutᶜ and 𝔊, and from which, apparently, 4QDeutᵇ, 4QDeutᶜ and 𝔊 are dependent.101 I hereby give an overview of the types of the major quantitative deviations for any manuscript in the stemmaagainst its parent text so far as the text is covered by 4QDeutᶜ:102

100

101

102

The agreements between 𝔊 and 𝔐 against ⅏ in the text covered by 4QDeutᶜ are the following: (1) major difference in Deut 10:6, supported by 4QDeutᶜ (= S2, congruent); (2) major difference in Deut 10:7, supported by 4QDeutᶜ (= S2, congruent); (3) the presence of ‫ מלך מצרים‬in Deut 11:3, supported by 4QDeutᶜ (= S2, congruent); (4) ‫ תאכל‬against ‫ תאכלו‬in Deut 16:3, supported by 4QDeutᶜ (= S2, congruent); (5) ‫ ידך‬against ‫ ידיך‬in Deut 16:10, supported by 4QDeutᶜ (= S2, congruent); (6) ‫ ורדפוך‬against ‫ ירדפוך‬in Deut 28:22, not supported by 4QDeutᶜ (= F2, incongruent); (7) ‫ הבו‬against ‫ והבו‬in Deut 32:3; supported by 4QDeutᶜ (= S2, congruent). Since 4QDeutᵇ·ᶜ are in a fragmentary state, we cannot be sure whether there were perhaps two succeeding manuscripts δ1 and δ2, one determined by the agreements of 4QDeutᵇ with 𝔊 against ⅏ and 𝔐, and the other determined by the agreements of 4QDeutᶜ with 𝔊 against ⅏ and 𝔐. Anyway, if 4QDeutᵇ and 4QDeutᶜ “branched off” at different times, we have no possibility to determine which of them branched off earlier. Therefore I prefer to speak of the onemanuscript δ whose existence is sure since there had to be at least one common ancestor of 𝔊, 4QDeutᵇ and 4QDeutᶜ beyond γ. The presence of ‫ עוד‬in Deut 13:12 shared by ⅏ and 𝔊 against 𝔐 and 4QDeutᶜ (apparent F1) is taken as accidental agreement between ⅏ and 𝔊. That is why this variant appears twice in the chart for the supposed textual development (as S2 and S3, asterisked in the chart).

169

A STEMMA FOR DEUTERONOMY

additions Major quantitative deviations specific for any manuscript, in the text covered by 4QDeutᶜ

omissions

additions vs. omissions

γ 〛 α (S1) – γ is 13:7 ‫בן אביך או‬ determined in these cases 16:8 ‫כל‬ by agreement of ⅏ with 31:18 ‫מהם‬ 4QDeutᶜ or 𝔊

–

3 : 0 cases (5 : 0 words)

𝔐 〛 α (S1)

–

–

⅏ 〛 γ (S2) – γ is a) 33 additional determined in these cases words in 10:7 by agreement of 𝔐 with 4QDeutᶜ or 𝔊 a) recensional variant, verses partially covered by 4QDeutᶜ b) other variants at word level b) 13:12 ‫*עוד‬

a) eleven words lacking in 10:6, among them three present in 4QDeutᶜ (‫ ;)ויקבר שם ויכהן‬one word lacking in 10:7, present in 4QDeutᶜ (‫)הגדגדה‬ b) 11:3 ‫מלך מצרים‬

a) major var. in Deut 10:6–7: 33 : 12 words;

δ 〛 γ (F3) – in these cases δ is determined by agreement of 4QDeutᶜ with 𝔊; γ is determined by agreement of ⅏ with 𝔐

16:8 ‫בו‬ 27:1 ‫הזאת‬ 31:17 ‫יהוה‬ 31:19 ‫דברי‬

15:2 ‫את רעהו‬

4 : 1 cases (4 : 2 words)

4QDeutᶜ 〛 δ (S4) – δ is determined in these cases by agreement of 𝔊 with ⅏ or 𝔐

4:14 ‫הירדן‬ 7:4 ‫אלהי]כם‬ 10:2 ‫יהוה‬ 26:19 ‫לך‬

13:7 ‫או בתך‬ 16:11 ‫בשערך‬ 28:1 ‫לשמר‬

4 : 3 cases (4 : 4 words)

𝔊 〛 δ (S3) – δ is determined in these cases by agreement of 4QDeutᶜ with ⅏ or 𝔐; some of these 𝔊 readings are inherited from ε, e.g. the omission of 13 words in Deut 17:5

8:3 ρηματι (‫)דבר‬ 9:18 δευτερον (‫)שנית‬ 13:12 ετι* (‫)עוד‬ 15:3 παρ αυτω (‫)מאתו‬ 16:9 ολοκληρους (‫)שלמים‬ 27:26 πας ανθρωπος (‫)כל איש‬ 28:7 ο θεος σου (‫)אלהיך‬

7 : 8 cases 4:15 ‫כל‬ (8 : 25 words) 8:2 ‫זה ארבעים שנה‬ 8:3 ‫ידעת ולא‬ 13:16 ‫ואת בהמתה‬ 15:2 ‫בעל‬ 17:5103 ‫אש]ר עשו‬ ‫את הדבר‬ ‫הר[ע הזה ]אל שעריך‬ [‫את האיש או את האשה‬ 28:4 ‫ופרי בהמתך‬ 31:19 ‫לכם‬

103

–

b) other var.: 1 : 1 cases (1 : 2 words)

This shorter reading of 𝔊 is supported by the Temple Scroll, so it presumably goes back to a further hyparchetype, ε.

170

BENJAMIN ZIEMER

Since we included the indirect testimony of the Temple Scroll in our discussion of Deut 18:5, it should be mentioned that, at least in three cases included in the chart, the Temple scroll gives additional evidence for the stemmatic relations. First, the longer reading including ‫ בן אביכה או‬is present in the citing/rewriting of Deut 13:7 by 11QT 44:19[26], thus strengthening the definition of γ against 𝔐. Second, the presence of ‫ או בתכה‬in the same verse/line in 11QT strengthens the definition of δ against 4QDeutᶜ. Third, in the citing/rewriting of Deut 17:5 by 11QT 55:20f.[27f.] the shorter reading of 𝔊 is shared by 11QT. Since the longer text of γ and δ is determined for this case by the agreement of 𝔐, ⅏ and 4QDeutᶜ, this shared abbreviation of 𝔊 and 11QT (caused obviously by homoioteleuton) has to be inherited from a further common parent text of those witnesses.105 Since 4QDeutᵃ too agrees with 𝔊 in the omission of a clause by parablepsiswhereas omissions of whole clauses are not attested for γ or δ, we place 4QDeutᵃ in our stemma, tentatively and with a question mark, as an offspring of ε. So we get the following refined stemma:

Į4SDOHR'HXWV"  Ȗ4SDOHR'HXWV"   į _

4'HXWE4'HXWF İ æ 4'HXWD" 47 ȗ

To show the implications of such a stemmafor the reconstruction of texts, the beginning of Deut 13:7 may serve as an example. The text of α, γ, δ and ε is clearly defined. In contrast, indeed, it is inevitable to postulate the existence of a manuscript ζ as actual Hebrew Vorlageof 𝔊, 104

105

The parablepsis is caused by the repetition of ‫ את האיש או את האשה‬in the text of the archetype witnessed by 𝔐, ⅏ and 4QDeutᶜ. The question of how this “verbose Hebrew text” (so McCarthy, Deuteronomy [BHQ], 102*) came into being does not concern us here since we do not deal with the literary history of Deuteronomy. In any case, the omission shared by 𝔊 and 11QT (apparently inherited from ε) affects the words ‫אל‬ ‫שעריך‬, leading to a deficient text since these words are required as a destination for ‫( והוצאת‬cp. Deut 22:21, 24 for similar cases). So in this case we have no choice but to refer to the agreement of 𝔊 and 11QT as a shared error. For the relation between 𝔊 and 11QT cp. Schiffmann, Septuagint, and Tov, ScriptureLike Texts, 212–216.

A STEMMA FOR DEUTERONOMY

171

but its text is not clearly determined by the stemmasince we know only one descendant of it, namely 𝔊. Therefore the text given for ζ is only a raw guess, positioning ζ anywhere between ε and 𝔊. Finally, it is impossible to reconstruct the readings of manuscripts fragmentarily preserved with certainty, since they may have omitted, added or substituted portions of the text. The text of all witnesses named in the stemmaabove is determined as follows for the beginning of Deut 13:7: α, the reconstructed archetype:

‫כי יסיתך אחיך בן אמך או בנך או בתך‬ ‫או אשת חיקך או רעך אשר כנפשך‬

4QpaleoDeutˢ, not preserved:

[…………………………………… ……………]

𝔐, adding vowel signs and accents:

‫וֹ־ב ְתּ ָ֜ך‬ ִ ‫וֹ־בנְ ָ֨ך ֽא‬ ִ ‫ן־א ֶמָּך ֽא‬ ִ ֠ ‫אָחיָך ֶב‬ ֣ ִ ‫ית ָ֡ך‬ ְ ‫ִ ֣כּי יְ ִ ֽס‬ ֗ ֶ ‫֣אוֹ ׀ ֵ ֣א ֶשׁת ֵח‬ ֖‫יקָך ֧אוֹ ֵ ֽר ֲעָך֛ ֲא ֶ ֥שׁר ְכּנַ ְפ ְשָׁך‬

γ, with clarifying addition:

‫כי יסיתך אחיך בן אביך או בן אמך או‬ ‫בנך או בתך או אשת חיקך או רעך אשר‬ ‫כנפשך‬

⅏, no change against γ:

‫כי יסיתך אחיך בן אביך או בן אמך או‬ ‫בנך או בתך או אשת חיקך או רעך אשר‬ ‫כנפשך‬

δ, no change against γ:

‫כי יסיתך אחיך בן אביך או בן אמך או‬ ‫בנך או בתך או אשת חיקך או רעך אשר‬ ‫כנפשך‬

4QDeutᵇ not preserved:

[…………………………………… ………………………]

4QDeutᶜ, with omission:

‫ בן[ אבי֯ ֯ך ]…………[ ֯או֯ בנך או‬...] [‫כנפ]שך‬ ̇ ‫אש]ת……………[אשר‬

ε, with substitution (?):

‫ יסיתך אחיך בן אביך או בן אמך‬106‫ואם‬ ‫או בנך או בתך או אשת חיקך או רעך‬ ‫אשר כנפשך‬

4QDeutᵃ not preserved:

[…………………………………… ……………………]

11QT, rewriting the passage from ε, with many minor changes, mainly in orthography: 106

‫ואם ישיתכ   ה אחיכ   ה בן אביכ   ה או בן‬ ‫אמכ   ה או בנכ   ה או בתכ   ה או אשת‬ ‫חיקכ   ה או ר י ע י כ   ה אשר כנפשכ   ה‬

Since replacing ‫ כי‬by ‫ אם‬in 11QT is an inner-Hebrew variant, it is not clearly determined whether ε and ζ read ‫ ואם‬or ‫וכי‬. Further, the rendering by ἐὰν δὲ does not necessarily imply that the conjunction ‫ ו‬was present in ζ. 𝔊 has ἐὰν δὲ 65 times in Deuteronomy. γ, determined by agreement of 𝔐 and ⅏, has ‫ כי‬in 47 of these cases, ‫ וכי‬in 5 cases, ‫ ואם‬in 7 cases, and ‫ אם‬and ‫ פן‬in one case each. In one case the string of text translated by 𝔊 including ἐὰν δὲ is not present in 𝔐 or ⅏ (Deut 30:16), in three cases 𝔐 and ⅏ are at variance (19:11, 16; 21:22). So ζ may have read ‫כי‬, ‫ וכי‬or ‫ואם‬. The reading ‫ ואם‬would fit in the stemmamost easily.

172

BENJAMIN ZIEMER

[[ζ, the Hebrew Vorlageof 𝔊, was either identical with ε or slightly modifying, perhaps as follows:]]

‫]]ואם יסיתך אחיך בן אביך או בן אמך‬ ‫או בנך או בתך או   האשה אשר בחיקך‬ [[‫או הרע אשר כנפשך‬

𝔊, translating ζ, at least with a free rendering for ‫ בן‬before ‫ אביך‬and ‫אמך‬:

Ἐὰν δὲ παρακαλέσῃ σε ὁ ἀδελφός σου ἐκ πατρός σου ἢ ἐκ μητρός σου ἢ ὁ υἱός σου ἢ ἡ θυγάτηρ σου ἢ ἡ γυνὴ ἡ ἐν κόλπῳ σου ἢ ὁ φίλος ὁ ἴσος τῆς ψυχῆς σου

If we don’t look only at one verse, but at the whole book, we see additions and omissions at almost every stage of the transmission history. This is typical for a stemmatic relationship. However, an inclination towards the addition of single words seems to be characteristic for γ and δ. This fits with the observations made earlier: the data collected by Emanuel Tov showed 27 additions vs. 2 omissions for γ against α, which corresponds well with the 3 additions vs. 0 omissions covered by 4QDeutᶜ. Summarising the evidence gathered by Tov for the three main witnesses and the observations made on the basis of 4QDeutᵇ and 4QDeutᶜ, some specifics of the manuscripts in our stemmabecome visible: Types of major quantitative revealed by deviations for any manuscript in the stemma against its parent text γ〛α 4QDeutᵇ 4QDeutᶜ Tov statistics 𝔐〛α 4QDeutᵇ 4QDeutᶜ Tov statistics ⅏〛γ 4QDeutᵇ 4QDeutᶜ – major – minor Tov statistics δ〛γ 4QDeutᵇ 4QDeutᶜ 4QDeutᵇ 𝔊 〛 δ (sum of the changes made by ε, ζ and 𝔊 against δ) 4QDeutᶜ 𝔊 〛 γ (sum of the changes 4QDeutᵇ and made by δ, ε, ζ and 𝔊 4QDeutᶜ against γ) Tov statistics 4QDeutᵇ 〛 δ 4QDeutᵇ 4QDeutᶜ 〛 δ 4QDeutᶜ

additions

omissions ratio

– 3 27 – – – – (33 words) 1 22 3 4 1 7 15

– – 2 – – – – (12 words) 1 8 0 1 1 8 10

– 3:0 13.5 : 1 – – – – (2.75 : 1) 1:1 2.75 : 1 3:0 4:1 1:1 0.87 : 1 1.5 : 1

99 0 4

44 2 3

2.25 : 1 0:2 1.33 : 1

A STEMMA FOR DEUTERONOMY

173

One surprising fact catches the eye: 𝔊 is not by itself prone to additions! If the portions of text randomly covered by 4QDeutᵇ·ᶜ are taken as representative, γ and δ both had an inclination towards small scale additions, 4QDeutᵇ was prone to omissions, whereas 4QDeutᶜ and 𝔊 (resp. ε, ζ and 𝔊) show a relatively high level of deviation against δ, but no visible preference for longer or shorter readings. Let us now have a look at more Qumran Deuteronomy manuscripts. 4.5. 4QDeutᵈ (4Q31), showing about 25 variants, among them no major family variant In the text covered by 4QDeutᵈ (4Q31),107 there is only one clear divergence of 𝔐 against the combined testimony of ⅏ and 𝔊. In this case 4QDeutᵈ supports γ by isolating 𝔐 (S1).108 At the same time, 4QDeutᵈ shares none of the “major” special readings of either 𝔊 or ⅏ concerning the presence or absence of words. In all these cases 4QDeutᵈ shares the majority reading, either that of 𝔐 𝔊 against ⅏ (S2)109 or that of 𝔐 ⅏ against 𝔊 (S3).110 So far, with respect to the “major” variants, 4QDeutᵈ turns out to be a third witness to γ, independent of ⅏ and δ, helping to determine the text of γ. But is this positioning coherent with the minor variants too? There are, next to the usually large number of isolated 𝔊 readings (twelve are listed in DJD), four variants which appear to be exclusive agreements with either ⅏ or 𝔊 (F2 or F3). But none of them is strong enough to determine exclusive relationship with one of these witnesses. The two agreements with ⅏ against 𝔐 (F2?) are designed as “orth. or var.”: we do not know from the translation how the Vorlage of 𝔊 has been written.111 If we see them as inner-Hebrew variants, they help to determine the Hebrew text of γ. Of the 107

108

109

110

111

Publication: Sidnie White Crawford, DJD XIV, pp. 35–38 and pl. X. Cp. White, Special Features, 163–164; id., Three Deuteronomy Manuscripts, 28–34. 4QDeutᵈ reads in Deut 3:27 ‫על ראש‬, which agrees with 𝔊 (ἐπὶ κορυφὴν) in the presence of the preposition ‫על‬. Such longer reading is presupposed also by ⅏ (‫)אל ראש‬, admittedly with the very frequent change between ‫ אל‬and ‫על‬. Thus, the shorter reading of 𝔐 (‫ )ראש‬is isolated (S1). 4QDeutᵈ isolates, by agreement with 𝔐 𝔊, the following ⅏ readings (S2): ⅏ shows change in word order in 3:19 (‫ טפכם‬before ‫ )נשיכם‬and omission of one word (‫)אלהיכם‬ in 3:21 (S2). 4QDeutᵈ isolates, by agreement with 𝔐 ⅏, the following 𝔊 readings (S3): 𝔊 shows omission of one word in 2:27 (‫בדרך‬, haplography), in 2:31 (‫ )רש‬and in 3:26 (‫)אלי‬, addition of one word in 2:36 (‫)הר‬, in 3:20 (‫ )אלהיכם‬and in 3:25 (‫)הזאת‬. These shared readings are ‫ תיראם‬in 3:22 against ‫𝔐 תיראום‬, and ‫ ינחל אתם‬in 3:29 against ‫𝔐 ינחיל אותם‬.

174

BENJAMIN ZIEMER

two exclusive agreements with 𝔊, one was already shown as determining the reading of γ, isolating both 𝔐 and ⅏.112 The other, the only remaining “family variant,” is a very good candidate for accidental agreement in a minor detail: Whereas in Deut 3:16 ⅏ and 𝔐 read ‫ועד נחל‬, 4QDeutᵈ reads ‫נח]ל‬ ̇ ‫עד‬, like 𝔊 (ἕως χειμάρρου). In this case, the agreement of ⅏ and 𝔐 determines the reading of γ whereas 4QDeutᵈ and 𝔊 apparently made the same change twice independently, for the following reasons: first, this shorter reading is clearly a lectiofaciliorwhich arose several times independently – several ⅏ manuscripts as well as Masoretic manuscripts omitted this ‫ו‬, and the Vulgate, a witness of the 𝔐 family, omitted the syndesis as well as most modern translations. Second, 4QDeutᵈ shows a slight tendency to prefer asyndetic connections. Out of the three singularity readings of this manuscript (S4) there are two omissions of ‫ו‬-copulativum.113The third and last S4 variant reading has been a ‫ד‬-‫ ר‬change: ‫ ויתעבד‬in Deut 3:26 against ‫ ויתעבר‬testified by all other witnesses. After all, 4QDeutᵈ turns to be a faithful witness to γ.114 This is underlined by the fact that in the seven textual variants designated as innerHebrew variants 4QDeutᵈ never goes its own ways but supports either 𝔐115 or ⅏.116 So it determines in all these cases the Hebrew text of γ and can be placed with certainty in our stemma. 4.6. 4QDeutᵉ (4Q32), showing only one family variant: F3 Among the few variants of 4QDeutᵉ (4Q32) stated in DJD,117 there is only one family reading: ‫ בידך‬in Deut 7:23, an exchange of one word shared by the Hebrew text presupposed by 𝔊, against ‫ לפניך‬determined as the text of γ by the agreement of 𝔐, ⅏ and 4QpaleoDeutʳ. On the 112

113

114

115

116 117

‫ על ראש‬in Deut 3:27 is the reading of γ, witnessed by 4QDeutᵈ and 𝔊, and later changed to ‫ אל ראש‬by ⅏, whereas ‫ראש‬, the reading of 𝔐, was likely the reading of α. The ‫ ו‬is omitted against the combined testimony of 𝔐 ⅏ 𝔊 in Deut 3:19 before ‫טפכם‬ (the second member of a three-membered itemisation), and in 3:27 before ‫( תימנה‬the third member of a four-membered itemisation), again facilitating readings, leading to a more fluent text and paralleled by medieval Hebrew manuscripts as well as by modern translations. A further case of an omitted ‫ ו‬was the object of a supralinear correction, presumably by the scribe himself, since it concerned the second member of a twomembered itemisation (‫]ר[ץ‬ ֯ ‫ ו ֯ב ֯א‬in 3:24). To summarise, the textual variants are distributed as follows (in brackets the apparent opinion in the DJD edition): 1(0)×S1, 2×S2, 12×S3, 4(3)×S4; 0(2)×F2, 0(2)×F3, 9(7)×inner-Hebrew. ‫ישמעו֯ ן‬ ̇ in 2:25 against ‫⅏ ישמעו‬, ‫ ואצו‬in 3:18 against ‫⅏ ואצוה‬, ‫ ומקנכם‬in 3:19 against ‫⅏ ומקניכם‬, and ‫ וירשו‬in 3:20 against ‫⅏ ויירשו‬. ‫ ̇א]ת שמעך‬in 2:25 against ‫𝔐 ִשׁ ְמ ֲעָך‬, and ‫ ההיא‬in 2:34 and 3:23 against ‫𝔐 ההוא‬ᴷᵉᵗ. Publication: Julie Ann Duncan, DJD XIV, pp. 39–44 and pl. XI.

A STEMMA FOR DEUTERONOMY

175

other hand, none of the two major singularity readings of 𝔊 is shared by 4QDeutᵉ.118 Since there is some overlap with other Qumran manuscripts, 4QDeutᵉ helps determine the archetype by agreeing with the majority text (⅏ 𝔐 𝔊) against certain singularity readings of other manuscripts.119 One case might be an orthographic variant: at Deut 7:23, 4QDeutᵉ agrees with 𝔐 and 4QDeutᶠ in the defective spelling ‫ השמדם‬which is understood as Nifal by 𝔐, against the Hifil ‫ השמידם‬in 4QpaleoDeutʳ and ⅏ and apparently translated by 𝔊. We cannot know if the scribe of 4QDeutᵉ read Hifil or Nifal. However, if we have to decide if he or the scribes responsible for ⅏ and 4QpaleoDeutʳ made a change in orthography, it is worth to be noted that 4QDeutᵉ itself shows a slight tendency to retain or to create unusual defective readings against 𝔐 (‫ החלת‬Deut 4:23)120 or against ⅏ (‫ כלתם‬Deut 7:22;121 ‫ מצותו‬Deut 8:2122) whereas ⅏ shows a tendency towards adapting the spelling to the reading tradition, and 4QpaleoDeutʳ shows a tendency towards a fuller spelling. To summarise, we can see 4QDeutᵉ as an additional witness for δ, so explaining all non-orthographic variants in a coherent manner. But this assignment should be accompanied by a question mark, since there are only seven textual variants covered by this manuscript. 4.7. 4QDeutᶠ (4Q33), showing more than 40 variants, among them 5 exclusive agreements with 𝔐 (F1) Within the text covered by 4QDeutᶠ (4Q33),123 there are enough idiosyncratic readings of 𝔐, ⅏ and 𝔊 to determine the place of this manuscript in the stemma. 118

119

120

121

122

123

In Deut 7:22 𝔊 alone (γένηται ἡ γῆ ἔρημος) presupposes the addition of ‫תהיה הארץ‬ ‫ שממה‬in assimilation to Ex 23:29, a clear singularity reading since 𝔐, ⅏ and 4QpaleoDeutʳ agree on the shorter reading, so determining γ, with 4QDeutᵉ. In contrast, in Deut 8:2, 𝔊 alone presupposes the omission of ‫זה ארבעים שנה‬, again a clear singularity reading since 𝔐, ⅏, 4QDeutᶜ and 4QDeutᶠ (thus determining β, γ and δ) agree in the longer reading with 4QDeutᵉ. First, ‫כל]ת[ם‬ ֯ 4QDeutᵉ in 7:22 against the singularity reading of 4QDeutᵐ, ‫;לכלותמה‬ second, ‫ תחמד‬4QDeutᵉ in 7:25 against the singularity reading of 4QpaleoDeutʳ, ̇‫;תחמדו‬ third, ‫ שמלתך‬4QDeutᵉ in 8:4 against the singularity reading of 4QDeutᶜ, ‫( שלמתך‬innerHebrew). Here, the defective spelling is determined as the spelling of γ, since 4QDeutᶠ agrees with ⅏ and 4QDeutᵉ. Here again, the defective spelling is determined as the spelling of γ, since 4QDeutᶠ agrees with 𝔐. Here, the spelling of γ is not determined, since 4QDeutᶠ agrees with 𝔐Qere and ⅏ against 𝔐ᴷᵉᵗ and 4QDeutᵉ. Publication: Sidnie White Crawford, DJD XIV, pp. 45–54 and pls. XII–XV.

176

BENJAMIN ZIEMER

Out of 25 cases where 𝔐 and ⅏ agree against 𝔊, 4QDeutᶠ agrees 24 times against 𝔊. Among these agreements are all major deviations of 𝔊. Against 𝔊, 4QDeutᶠ shows, on the one hand, the presence of ‫ זה ארבעים שנה‬in Deut 8:2, and the presence of ‫ ידעת ולא‬in Deut 8:3.124 The longer text of 4QDeutᶠ is in both cases already determined as the text of α, β, γ and δ by the agreement of 𝔐, ⅏ and 4QDeutᶜ. On the other hand, 4QDeutᶠ shares with 𝔐 and ⅏ a shorter text concerning the absence of ‫ ושבעת‬before ‫ ושמחת‬in Deut 27:7 against a harmonising addition in 𝔊 (καὶ ἐμπλησθήσῃ). Among the many minor deviations of 𝔊, 4QDeutᶠ agrees only once with 𝔊, in the addition of a ‫ו‬-copulativum (‫ ולא תחסר‬in Deut 8:9), presumably by accident.125 There are 11 cases where 𝔐 stands alone against ⅏ and 𝔊. In contrast to 4QDeutᵃ·ᵇ·ᶜ·ᵈ, 4QDeutᶠ supports a significant number of these 𝔐 readings (five cases), agreeing with ⅏ and 𝔊 in the remaining cases. We can explain this pattern by assuming that 4QDeutᶠ is dependent on a manuscript β, halfway between α and γ. The five exclusive agreements with 𝔐 (F1) show that between β and γ some minor changes occurred, typical for textual history and explainable in both directions: change in number, change in word order, presence vs. absence of ‫ו‬-copulativum. However, they determine the place of β, the parent text of 4QDeutᶠ, between α and γ. The six variants where 4QDeutᶠ agrees with γ against 𝔐 are far more symptomatic. They isolate, as a rule, difficult readings of the 𝔐 tradition against all other branches of the stemma, revealing the character of 𝔐 most clearly. These six isolated readings (S1) are the following: in Deut 7:24, the unusual ‫ )𝔐( בפניך‬stands isolated against the more usual ‫לפניך‬, the latter shared by 4QDeutᶠ, 4QpaleoDeutʳ, ⅏ and 𝔊.126 In Deut 8:7, ‫ ורחבה‬is 124

125

126

The shorter reading is likely caused by homoioarkton in 𝔊 since the common text reads ‫לא ידעת ולא ידעו‬. The same addition occurred in one medieval 𝔐 manuscript, and is made by the Vulgate. Such distribution is typical for readings arising more than once independently. Sidnie White, in the DJD edition, states that 𝔊 (κατὰ πρόσωπόν σου) supports ‫בפניך‬, the reading of 𝔐, against ‫ לפנ֯ ̇יך‬in 4QpaleoDeutʳ and ‫ לפניך‬in ⅏ and 4QDeutᶠ. But κατὰ πρόσωπόν is an equivalent frequently used for ‫ לפני‬in 𝔊, more than 50 times in the Old Testament, among them Deut 9:2 and 31:21 where 𝔐 ⅏ both have ‫לפני‬. In contrast, ‫בפני‬, far less frequent in the Hebrew text, is rendered most often (four times) by εἰς (τὸ) πρόσωπον, with one case in Deut where 𝔐 ⅏ agree in the reading (Deut 25:9). In one further case ‫ בפני‬testified by both 𝔐 ⅏ (Deut 4:37) is not translated by 𝔊. In contrast, there are two cases in Deut where 𝔐 and ⅏ disagree: ‫ בפני‬in 𝔐 stands against ‫לפני‬ in ⅏ (Deut 7:24; 11:25). 𝔊 has κατὰ πρόσωπόν in both cases, which can easily be explained with the assumption that its Vorlagewas ‫ לפני‬in these two cases, as in ⅏, thus determining γ.

A STEMMA FOR DEUTERONOMY

177

present in 4QDeutᶠ, ⅏ and 𝔊 (as well as in the excerpted texts 4QDeutʲ·n), in assimilation to Ex 3:8, against the shorter reading of 𝔐, supported by the Vulgate, the Peshitta and Targum Onkelos. In two cases of syndesis differences, 4QDeutᶠ, ⅏ and 𝔊 show the lectiofaciliorby omitting the ‫ ו‬at the beginning of Deut 19:21, and by adding a ‫ ו‬before ‫ עם רב ממך‬in Deut 20:1. The last two S1 variants are even more significant: here, the medieval 𝔐 manuscripts testify a tradition split between an erratic consonantal text (Ketiv) and a facilitating reading tradition (Qere). In Deut 21:8, 𝔐Ket shows clear disgruence between the subject (‫ = ידינו‬plural) and the predicate (‫ שפכה‬apparently fem. singular). Virtually all branches of the tradition read the verb in the plural, so eliminating the tension. While 𝔐 retains the erratic form in the written text (𝔐ᴷᵉᵗ) limiting the correction to the oral tradition, marked by the punctuation and, usually, by an annotation in the masoraparva(𝔐Qere), the tension is not any longer visible in the Vulgate or the Targums, translating the qere. The agreement of 4QDeutᶠ with ⅏ and 𝔊, however, shows that β, according to our stemmaone of the earliest copies of Deuteronomy, parent text for 4QDeutᶠ and γ, corrected the erroneous spelling. The same is true for the 𝔐ᴷᵉᵗ spelling ‫ הנער‬designating a young woman, which is read ‫ ַהנֲּ ָע ָרה‬by 𝔐Qere as a qere perpetuum. Again, this reading tradition agrees with all translations which use female equivalents, and with the written and oral tradition of the Samaritans (‫)⅏ הנערה‬. The only case of ‫נערה‬/‫ נער‬in Deut covered by Qumran evidence is Deut 22:15, partially extant in 4QDeutᶠ. This manuscript spells ‫ הנערה‬with a supralinear ‫ ה‬supplemented by a later hand or at least with a different shape.127 Regrettably, we cannot know how the other cases of ‫נערה‬/‫ נער‬were written in 4QDeutᶠ. However, we see in this case how written ‫ הנער‬was used to be read and/or corrected as ‫הנערה‬. There are similar cases, noted in DJD as inner-Hebrew variants or as orthographic differences: First, the qereperpetuum‫ ִהיא‬where 𝔐ᴷᵉᵗ ‫הוא‬ denotes a femininum, is another case where the unusual spelling of 𝔐 is isolated as singularity variant by 4QDeutᶠ. In Deut 21:4, 4QDeutᶠ has ‫ההיא‬, like ⅏, against 𝔐ᴷᵉᵗ ‫ההוא‬. Second, in Deut 8:2, 𝔐ᴷᵉᵗ ‫ מצותו‬looks like a singularic form ‫ ִמצוָ תוֹ‬although it is to be read according to the Masora like ‫ווֹתיו‬ ָ ‫( ִמ ְצ‬plur). This 𝔐Qᵉʳᵉ tradition again is in accordance with the spelling in ⅏ and 4QDeutᶠ, and presupposed by 𝔊. Third, finally, instead of the unusal form ‫ ָה ֵאל‬for the pluralic demonstrative pronoun, passed down by 127

Cp. the supralinear ‫ ה‬with the ‫ ה‬in the following word ‫ ואמה‬in the photograph (DJD XIV, plate XIV, fragment 20, last line, online: http://www.deadseascrolls.org.il/explore-thearchive/image/B-473839).

178

BENJAMIN ZIEMER

𝔐 as one of the sebirinreadings, both ⅏ and 4QDeutᶠ use in Deut 7:22 the more common ‫( האלה‬cp. ‫ ָה ֵא ֶלּה‬elsewhere in 𝔐). None of these readings has any theological significance. In contrast, the singularity readings of 𝔐ᴷᵉᵗ, isolated with the aid of 4QDeutᶠ, show the specifics of the writing tradition of 𝔐: it is a form of the text conserving erratic peculiarities not understandable for later generations. The spelling of hi᾿as ‫הוא‬, of na῾arāas ‫נער‬, of šapeḥūas ‫ שפחה‬contradicts all rules of grammar but is conserved in the writing tradition. The same is true for such unusual spellings as ‫ מצותו‬for the plural, or ‫ האל‬for the demonstrative pronoun. So the characteristics of 𝔐 show not, as often asserted, any standardisation of content, style, morphology or orthography but the opposite: the singularity readings of 𝔐 typically maintain unusual forms, and show reservation against harmonising, assimilating and facilitating changes. It is comprehensible that these extravagant spellings were not copied unless a special interest arose in single words and letters of the original text among the Pharisees or the proto-Rabbinic movement, assured by the belief that the letters of the original text were counted from the beginning. If we see 4QDeutᶠ as an offspring of β, this manuscript β turns out to be one of the earliest copies which, in contrast to 𝔐, did not hesitate to improve the text in minor details but avoided to change the content. Together with manuscripts γ and δ which are characterised by small assimilating additions, β appears to be a member of a sequence of thoroughly copied manuscripts which avoided accidental omissions but facilitated the reading of the text in some minor details, and from which almost all manuscripts until the early Heriodianic period were copied. 4.8. 4QDeutᵍ (4Q34), showing about 15 variants, agreeing in all details with 𝔐 4QDeutᵍ (4Q34)128 is a deluxemanuscript129 which shows in its corrected text not even one single orthographic deviation in 133 words or parts of words against 𝔐.130 In case of the two textual differences between ⅏ and 𝔐 in the text covered by 4QDeutᵍ, 4QDeutᵍ agrees with 𝔐 in both cases (S2); the same is true for the two orthographic differences between 128

129 130

Publication: Sidnie White Crawford, DJD XIV, pp. 55–59 and pl. XVI. Cp. White, Special Features, 165–66; id., Three Deuteronomy Manuscripts, 35–42. Tov, Scribal Practices, 126. The Masoretic text for these 133 words shows no difference between Codex L and the textusreceptus.

A STEMMA FOR DEUTERONOMY

179

⅏ and 𝔐. Concerning the twelve variant readings of 𝔊 against ⅏ and 𝔐, 4QDeutᵍ agrees with 𝔐 and ⅏ in all cases (S3). In the case of one reading shared by ⅏ and 𝔊 against 𝔐, 4QDeutᵍ supports 𝔐 (F1). It is especially noteworthy that the manuscript agrees with 𝔐 in all questions of defective and plene writing.131 Such agreement is not explainable by chance or by accident. On the contrary, this manuscript, as well as MasDeut and MurDeut, gives reason to the assumption that the form of its Vorlagewas so highly esteemed that even orthographical details were copied consciously. This makes sense only in case the scribes believed every single letter has been counted. Such Vorlagen, identical with the letters of the later 𝔐, were made accessible for the public at the latest in the first c. BCE (cp. MasDeut) but may have been exact copies of a much earlier archetype. 4.9. 4QDeuth (4Q35), showing more than 40 variants, among them 8 exclusive agreements with 𝔊 (F3) 4QDeutʰ (4Q35)132 is famous for its closeness to the wording of Deut 33 testified by 4QTestimonia.133 Next to this, it is noticeable that all major deviations 4QDeutʰ shows in comparison to ⅏ or 𝔐 (i.e., presence or absence of more than one word) are shared with 𝔊 (F3).134 This, together with minor exclusive agreements with 𝔊, is a strong argument for placing the manuscript as an offspring of δ and/or ε in the stemma. However, in two cases where ⅏ and 𝔊 agree against 𝔐, 4QDeutʰ has the reading of 𝔐. We have to explain these apparent F1 variants as accidental agreement 131

132 133

134

This includes minority readings recorded by the Masora, like ‫ גדולה‬written plene in Deut 25:14, which is unusual in the Pentateuch. In 𝔐, 18 of the 22 records of gedōlâ in the Pentateuch are written defective, ‫גדלה‬, and only four plene, ‫גדולה‬, among them Deut 25:14 (cp. the Mm in 𝔐L). ⅏ writes consistently defective ‫ גדלה‬according to the Samaritan pronunciation gādalla. Manuscripts written with less accuracy tend to introduce the majority (in this case, the defective) spelling. It is thus no wonder that Kennicott lists 12 Masoretic manuscripts with the defective writing ‫ גדלה‬in Deut 25:14. Publication: Julie Ann Duncan, DJD XIV, pp. 61–70 and pls. XVII–XVIII. Cp. Duncan, Readings. Within Deut 33:8–11 (the text string cited by 4QTest), 4QTest agrees completely with the text witnessed by 4QDeutʰ in any case where the main witnesses 𝔐, ⅏ or 𝔊 are at variance. Anyway, both manuscripts show further idiosyncrasies: 4QTest has, obviously erroneously, ‫ דעתיכהי‬against ‫ ראיתך‬4QDeutʰ in v. 9, besides some errors made and subsequently corrected by the scribe himself, whereas 4QDeutʰ alone omits the ‫ ו‬of ‫ ובריתך‬in v. 9. Two pluses of two words against 𝔐 and ⅏ are supported by 𝔊: ‫ ̇ע ֯ל] ספר‬in Deut 31:9 and ‫ללו֯ ]י‬ ̇ ‫ הבו‬in 33:8. In Deut 1:39 a minor agreement with 𝔊 against 𝔐 and ⅏ is connected with a major quantitative variant: according to calculations of space, 4QDeutʰ seems to have shared the omission of the five words ‫ אמרתם לבז יהיה ובניכם אשר‬by homoioteleutonpresupposed in 𝔊 (see DJD).

180

BENJAMIN ZIEMER

or as a result of contamination by memory, not surprising in the case of the very beginning of the book and in the case of the blessing of Moses, presumably well known to the scribes by heart.135 There is one minor exclusive agreement with ⅏ explainable as accident.136 Therefore I retain a question mark while placing this manuscript in the stemma. The specific agreements of 4QDeutʰ and 4QTest against the three main witnesses define a manuscript θ as parent text of both manuscripts. θ itself is determined as an offspring of ε by reason of the proximity of 4QDeutʰ and 𝔊, including shared additions and omissions against γ. 4.10. 4QDeuti (4Q36), showing about 10 variants, among them 2 exclusive agreements with 𝔐 A different case is 4QDeuti (4Q36).137 This manuscript agrees with 𝔐 in all textual variants where ⅏ or 𝔊 disagree with 𝔐, among them two family variants (F1), and is thereby a clear member of the 𝔐 family. But there is one inner-Hebrew textual variant extant – ‫( שלמת‬4QDeuti) instead of ‫ )⅏ 𝔐( שמלת‬in Dtn 22:5138 – besides four orthographic deviations 135

136

137 138

In Deut 1:8, within a cited Divine speech, 4QDeutʰ agrees with 𝔐 in the lectiodifficilior‫נשבע יהוה‬. This may have been the reading of γ and δ, whereas ⅏ and 𝔊 could have changed this by the more convenient ‫ נשבעתי‬twice independently. In Deut 33:17, 4QDeutʰ agrees with 𝔐 on reading ‫ שורו‬against the shorter reading ‫שור‬ (determined as the reading of γ, δ and ε by the agreement of ⅏ and 𝔊). For Deut 33:19 – a further variant which, at first sight, could be classified as F1 – there are in fact four different readings: ‫𝔐 יינקו‬, ‫⅏ יינק‬, ‫ ינקך‬presupposed by 𝔊, and ‫ ינקו‬in 4QDeutʰ. ‫ יינקו‬may have been the text of γ, explaining the agreement in orthography between ⅏ and 𝔐. ‫ינקו‬, the spelling of 4QDeutʰ, may well have been the text of δ and ε, misread by 𝔊 or its Vorlageas ‫ינקך‬, thereby determining ‫ יינק‬as singularity reading of ⅏ and the partial agreement of ⅏ and 𝔊 in number as random. In Deut 1:7, 4QDeutʰ agrees with ⅏ in the asyndetic reading ‫בנגב‬. This is a facilitating reading, since it concerns the fourth member of a five-membered itemisation. 𝔐 and 𝔊 agree on the very unusual pattern a, b + c + d + e, a clear lectiodifficilior(retained, e.g., by Targum Onkelos and KJV), thus defining the text of α–ζ. The Peshitta replaces this unusual pattern by a + b + c + d + e, clearly facilitating. In contrast, ⅏ (and, e.g., one Greek manuscript, the Armenian translation and the German Luther translation) has a, b + c, d + e; the Vulgate has a + b + c, d + e; and some modern translations (German: Zürcher, Einheitsübersetzung) have a, b, c, d + e. All these versions and translations ignore the ‫ ו‬before ‫ בנגב‬in order to produce a more fluent text, thus causing agreement by accident. 4QDeutʰ may have done so, but is preserved only for the last part of the itemisation: [a ? b ? c], d + e. Publication: Sidnie White Crawford, DJD XIV, pp. 71–74 and pl. XIX. This metathesis suggests itself since the two words ‫ ִשׂ ְמ ָלה‬and ‫ ַשׂ ְל ָמה‬are used synonymously. It is a good example for a deviation which may appear several times independently. So 4QDeutᶜ has ‫ שלמתך‬against ‫ שמלתך‬in 𝔐 and ⅏ in Deut 8:4, whereas 1QIsaᵃ uses ‫ שמלה‬two times where 𝔐 has ‫( שלמה‬Isa 3:7; 4:1). Against it, ⅏ uses ‫ שמלה‬consistently in all cases where 𝔐 uses ‫( שלמה‬e.g., Deut 24:13; 29:4).

A STEMMA FOR DEUTERONOMY

181

in 102 words or parts of words. So this manuscript itself cannot be the Vorlageof later 𝔐 manuscripts. 4.11. Manuscripts not included in the stemma I do not include the manuscripts 1QDeutᵃ; 4QDeutᵏ²·ᵏ³·ˡ·ᵐ·ᵒ·ᵖ and the Deuteronomy manuscripts from Caves 2, 6 and 11 in the stemmasince they do not cover enough text and/or enough significant agreements for an educated guess about their place in the stemma. All these manuscripts may have been descendants of α, β, γ or δ since they either show a random distribution of variants (4QDeutˡ),139 no family variant at all (4QDeutᵏ², 4QDeutᵒ) or only one family variant (1QDeutᵃ, 4QDeutᵐ). In the latter cases the exclusive agreement with 𝔊 shown by the one family variant respectively is not strong enough to lead to a decision: In 4QDeutᵐ the agreement concerns the supralinear addition of suffixed ‫אלהים‬, leaving the status of the original text unclear,140 and in 1QDeutᵃ it concerns a change in number, a minor variant which often occurs twice independently in the course of textual history.141 However, the about ten variants covered by each of the manuscripts 1QDeutᵃ, 4QDeutᵏ²·ᵐ·ᵒ which contain about 100 words respectively, can theoretically be explained without exception by the assumption that these manuscripts are descendants of β or γ (4QDeutᵒ and 4QDeutᵏ²) resp. of γ or δ (1QDeutᵃ and 4QDeutᵏ²). The remaining manuscripts cover less than twenty words and are therefore excluded from the analysis: 2QDeutᵃ·ᵇ·ᶜ, 4QDeutᵏ³, 4QDeutᵖ, 6QpapDeut?, 6QDeut?, and 11QDeut. I made an exception discussing the very small fragment 4QpaleoDeutˢ (see above) since this is perhaps the oldest biblical manuscript from Qumran. Understandably, there is no certainty about its place in the stemma. I mark for this manuscript no less than three possible places in the stemma(identity with α, β, or γ), but the manuscript might be identical with δ or η or a descendant of α, β, γ, δ or η as well. I exclude the manuscripts 4QDeutʲ·ᵏ¹·n·q and 5QDeut for another reason: they are not actual Deuteronomy manuscripts but excerpted texts.142 139

140

141 142

The 41 words of 4QDeutˡ cover respectively one family variant for each possibility (F1, F2, and F3), all minor. So there is no serious basis for judging which of them is inherited and which came to pass by accident. Deut 3:20 (after ‫)אשר יניח יהוה‬: ‫ אלוהיכמה‬4QDeutᵐ (supralinear) 𝔊 (ὁ θεὸς ὑμῶν) 〛 > 𝔐 ⅏. Deut 13:3: ‫ תשמעו‬1QDeutᵃ 𝔊 〛 ‫⅏ 𝔐 תשמע‬. Cp. Duncan, Excerpted Texts, for the 4Q manuscripts, and Lange, Handbuch, for 4QDeutj (93: “eine Textsammlung aus verschiedenen Pentateuchbelegen”), 4QDeutn·q and 5QDeut (97, 99, 102 respectively: “Exzerpthandschrift”).

182

BENJAMIN ZIEMER

4QDeutq as a manuscript of the Song of Moses is an exceptional case: The archaic and so-called “polytheistic” 𝔊 variants in Deut 32:8, 43 witnessed now by 4QDeutʲ and 4QDeutq have found such a massive amount of interest in the scientific debate on the textual history that it is necessary to discuss these readings. 4.12. The 𝔊 version of the Song of Moses (Deut 32:1–43) as a result of contamination with an older version transmitted separately, witnessed by 4QDeutj and 4QDeutq The textual situation for Deut 32 is as follows: the only two complete Hebrew witnesses, 𝔐 and ⅏, show a lot of variants but no systematic divergences. A substantial part of the differences are agreements between ⅏ and 𝔊 against 𝔐, which have to be, according to our stemma, inherited by the hyparchetype γ. According to the Tov-Polak synopsis,143 there are more than twenty such agreements, among them the only three differences between ⅏ and 𝔐 concerning the presence of absence of whole words: ‫ ישראל‬at the end of v. 9 and ‫ יאכל יעקב‬at the beginning of v. 15 are pluses shared by ⅏ and 𝔊 against 𝔐, whereas ‫ כי‬in v. 30 is absent in ⅏ and 𝔊 while present in 𝔐. 𝔊, as usual, shows far more variants. The largest differences appear at the end of the song: v. 43 contains three stichoinot present in 𝔐 and ⅏, with the invocation of the sons or angels of God, and v. 44 contains a remark on the special written documentation of this song, using the same wording as Deut 31:22. The mention of celestial powers is connected to vv. 1 and 8 in the 𝔊 version, whereas the text common to 𝔐 and ⅏ shows a shorter version which has obviously undergone an antipolytheistic revision.144 How can the more archaic version of the Song of Moses behind 𝔊 be explained? Here 4QDeutʲ, 4QDeutᵏ¹, 4QDeutq and 4QPhylN, the four Qumran manuscripts which are not actual Deut manuscripts but contained the Song of Moses, help answer, especially 4QDeutq,145 the best preserved witness for the end of the poem. The synopsis of 4QDeutq, 𝔐 and 𝔊 for v. 43 shows that 𝔊 “knew two Hebrew forms of the text.”146 This is obvious by the double rendering of the beginning of v. 43. 𝔐 (supported 143 144 145

146

Tov and Polak, Parallel Text. See DJD XIV, 137 (Lit.). Publication: Patrick W. Skehan and Eugene Ulrich, DJD XIV, pp. 137–142 and pl. XXXI. So Ulrich, DJD XIV, 141.

A STEMMA FOR DEUTERONOMY

183

by ⅏) has one stichos(‫)ה ְר ִנ֤ינוּ גוֹיִ ֙ם ַע ֔מּוֹ‬ ַ whereas 4QDeutq has two stichoi (‫)הרנינו שמים עמו והשתחוו לו כל אלהים‬. 𝔊 first translates the two stichoi attested in 4QDeutq, rendering ‫ אלהים‬by υἱοὶ θεοῦ resp. ἄγγελοι θεοῦ, adds a translation of the stichos attested by 𝔐 and ⅏, and rounds up with a varied translation of the second stichos.147 Undoubtedly, this is a conflation of two sources. Since we speak about the Greek translation of Deuteronomy, it is, strictly speaking, a contamination of the text of the book of Deuteronomy with another source. This source need not be another version of the book of Deuteronomy; presumably it has been a traditional version of the Song preserved seperately. We are not in the dark, and the (by far) easiest explanation of the conflation is obvious: for the Greek translation of Deut 32 or its Hebrew Vorlage, a version of the book of Deuteronomy similar to 𝔐 and ⅏ has been complemented and partially replaced through conflation by an obviously more original form of the poem transmitted separately, similar to 4QDeutq. Moreover, the separate transmission of this song both in writing and in oral tradition is stated explicitly by the addition of the sentence immediately after v. 43, doubling Deut 31:22: “And Moses wrote this song the same day, and taught it to the children of Israel.” This sounds like a justification of using an alternate but obviously more original version of this poem. The consequence may come as a surprise: the original form of the Song of Moses in the book of Deuteronomy as part of the Pentateuch is well represented by 𝔐 and ⅏, but the original form of thepoemitselfis better represented by 𝔊. The same seems true for v. 8. Since the stemma should be based on actual Deuteronomy manuscripts, it is inevitable to take the strong agreement between ⅏ and 𝔐 in Deut 32 as inherited by their common Vorlage: ‫בהנחל עליון גוים בהפרידו בני אדם יצב גבל)ו(ת עמים למספר בני ישראל‬

So this has to be the original reading for the book of Deuteronomy. The agreement between ⅏ and 𝔐 determines the reading of manuscripts α, β and γ. However, a small fragment of 4QDeutʲ,148 a manuscript conflating several passages known to us from Ex and Deut, is extant for this verse and agrees with 𝔊GÖ in reading ‫ בני אלהים‬instead of ‫בני ישראל‬. Of course this reading sounds more archaic, and fits perfectly with the more archaic version of v. 43 in 4QDeutq. Again we get one original version 147 148

See ibid. Publication: Julie Ann Duncan, DJD XIV, pp. 75–91 and pls. XX–XXIII.

184

BENJAMIN ZIEMER

for the book of Deuteronomy, and another version seemingly original for the poem transmitted separately and used by 𝔊. This raises at least two questions, one about the intention of the “book” version (⅏ 𝔐), and one about analogies for the restoration of the archaic features in 𝔊. First, concerning the intention of the “book” version, it has long been recognised that the omission of the “Gods” or “sons of God” in v. 43 and their replacement by the “sons of Israel” in v. 8 can be understood as a theological revision towards monotheism. This is plausible for any time in the Assyrian, Babylonian or Achaemenid period. But there was no necessity to perform such theological revision. A plurality of Gods or sons of God(s) is mentioned elsewhere in the Pentateuch (see Gen 6:2, 4) as well as in the Psalter (see Ps 97:7b, the closest parallel to Deut 32:43 𝔊). Indeed, there was an additional motivation for the changes: the composition of the Pentateuch. The Pentateuch is structured by a more or less coherent system of numbers and personal and geographic names. A “number of the sons of God” (Deut 32:8 𝔊GÖ and 4QDeutʲ) does not at all fit in this system. In contrast, the “number of the sons of Israel” is well established in the Pentateuch. Seventy names are specified in Gen 46:8ff. under the rubric “These are the names of the sons of Israel,” and the number “seventy” is mentioned explicitly Gen 46:27 (𝔐+⅏), repeated in Ex 1:5 (𝔐+⅏) and again in Deut 10:22 (𝔐+⅏+𝔊).149 This number equals the names of peoples specified in the table of the nations in Gen 10 (𝔐+⅏+𝔊).150 So the equation of the “borders of the nations” with the “number of the sons of Israel” (‫אָדם יַ ֵצּ ֙ב גְּ ֻבֹל֣ ת ַע ִ֔מּים ְל ִמ ְס ַ ֖פּר ְבּ ֵנ֥י‬ ֑ ָ ‫גּוֹים ְבּ ַה ְפ ִר ֖ידוֹ ְבּ ֵנ֣י‬ ִ֔ ‫ְבּ ַהנְ ֵ ֤חל ֶע ְל ֙יוֹן‬ ‫ יִ ְשׂ ָר ֵ ֽאל׃‬Deut 32:8 𝔐+⅏) is well founded in the Pentateuchal composition, but was less impressive for the Greek translator of Deuteronomy,151 and had no relevance at all for the independent transmission of the poem.152 Second, it sounds odd that 𝔊 or its parent text secondarily – in terms of the textual history of the book of Deuteronomy – restored a form of the song which seems to be moreoriginalin terms of religion history and literary history.

149

150 151

152

It is especially noteworthy that 𝔐, ⅏ and 𝔊 agree in the only case this number appears in Deuteronomy. For the counting cp. Jacob, Genesis, 277–296. Some of these relations did not exist for the Greek translator of Deuteronomy: in Gen 46 and in Ex 1 (in this latter case with support by some Qumran manuscripts), 𝔊 counts 75 souls in the house of Jacob, not 70. Cp. the fluctuation in the transmission of the Decalogue.

A STEMMA FOR DEUTERONOMY

185

But this is indeed a widespread phenomenon. Let us take the Temple Scroll: it assimilates the perspective of Deuteronomy, a speech of Moses, to the perspective of Exodus and Leviticus, a speech of God. This has been apparently the form of at least some of the sources of the book of Deuteronomy, namely the book of the covenant (Ex 21–23*) and the cultic Decalogue (Ex 34*). These were well known to the author of the Temple Scroll, and he conflated them with texts of Deuteronomy. In turning the perspective back into a divine speech, the author presumably intended to present a more original form of the Deuteronomic legislation. Or let us take another example, from the book of Genesis. Deuteronomy often remembers the solemn oath to the fathers, beginning with Abraham. But it is not mentioned in Deut that these fathers built several altars in the promised land. According to the book of Genesis, Abram built altars in Shekhem, in Bethel and in Elone-Mamre, and Isaac built one in BeerSheva. At least there is no mention that they offered sacrifices at these places, they simply “called the name of the Lord.” Can this be original? In terms of religion and literary history it is self-evident that it was once told that Abraham and Isaac celebrated offerings at these places. Surprisingly one can find this feature, which belongs to an older stage in religion and literary history, in the book of Jubilees which is clearly dependent on the book of Genesis, and in 1QapGen from the first c. BCE. There it is told at great length that Abraham celebrated offerings at several places. So, an older, archaic feature is restored in a late stage of textual development.153 Differentiating between the original reading of the book of Deuteronomy and the original reading of the Song of Moses, the following considerations are to be remembered: First, it is quite obvious that the Song of Moses is, presumably, much older than our archetype α, the common ancestor of all actual manuscripts of the book of Deuteronomy. Second, it is quite plausible that this song was present in the liturgy and that the scribes knew this song by heart, as well as the Decalogue. Third, it is certainly witnessed that this song was transmitted independently, in manuscripts other than Deut manuscripts. This is true for the Hebrew: so-called 4QDeutʲ contained the Decalogue, some other passages from Exodus and Deuteronomy, and the Song of Moses. So-called 4QDeutq contained, apparently, only the Song of Moses. And this is true for the 153

Raik Heckl presents another possible example for the phenomenon that an older piece of text, preserved in external sources, secondarily reentered the text of Deuteronomy in 𝔊: The introduction of Deut 6:4 in 𝔊 and the Nash Papyrus (Heckl, Anfang, 87).

186

BENJAMIN ZIEMER

Greek: the manuscripts of Deuteronomy, the Pentateuch or the Octateuch were not the only ones transmitting that song; besides there were numerous manuscripts of psalms with the attached Ωδαι. So it is very plausible, if not self-evident, that an older form of the Song of Moses, transmitted independently, has influenced the text of 𝔊 (or one of its parent texts δ, ε or ζ). This does not challenge the validity of the stemmaas a whole (with 𝔊 as a witness relatively far away from the archetype of the book of Deuteronomy) but reminds us that contamination is always a factor of relevance for explaining the variants in broadly transmitted and well known liturgical texts. 4.13. 4QpaleoDeutr (4Q45) showing relatively close proximity to ⅏ 4QpaleoDeutʳ (4Q45)154 is of interest as the largest paleo-Hebrew manuscript of Deuteronomy. Again we look at the variants listed in the DJD edition. It should be noted that singularity readings of 𝔊 are not noted in DJD IX, in contrast to the edition of the square script Qumran Deut manuscripts in DJD XIV. For the stemmait is sufficient to state that there are such readings, so excluding the possibility 4QpaleoDeutʳ is dependent on the Vorlageof 𝔊.155 Judith Sanderson and Eugene Ulrich specified 23 variants: “There are 23 variants preserved in this scroll. One (28:19) involves an entire verse, but the other 22 involve only one word or less.”156

The editors did not mention the agreements or disagreements with 𝔊 for these variants, so I had to add them since these are important for the stemma.157

154

155

156 157

Publication: Patrick W. Skehan, Eugene Ulrich and Judith E. Sanderson, DJD IX, pp. 131–152 and pls. XXXIV–XXXVI. To give an impression, I itemise just four singularity readings of 𝔊 where the extant fragments of 4QpaleoDeutʳ clearly agree with 𝔐 ⅏ out of the first verses covered by this manuscript. Instead of ‫[א ̇להי̇ ̇ך‬ ֯ ‫ יהוה‬before ‫ לכל העמים‬in Deut 7:19, 𝔊 presupposes ‫יהוה אלהינו‬. At the end of Deut 7:21, 𝔊 lacks ‫ונור[א‬. In Deut 7:22, 𝔊 has ἵνα μὴ γένηται ἡ γῆ ἔρημος καὶ πληθυνθῇ translating apparently ‫פן תהיה הארץ שממה ותרבה‬, instead of the shorter reading ‫ פן תרבה‬commonly witnessed by 4QpaleoDeutʳ 𝔐 ⅏. In Deut 7:23, 𝔊 alone apparently renders ‫בידיך‬, not ‫לפנ]יך‬. There are far more such variants. DJD IX, 134. Ulrich, Scrolls, 192–246, adds the witness of 𝔊 too. I disagree with him only in case of variant (4), Deut 7:24 (Ulrich, Scrolls, 195f.); see above the discussion on 4QDeutᶠ, showing that 𝔊 read in all likelihood ‫לפניך‬, as witnessed by 4QDeutᶠ, 4QpaleoDeutʳ and ⅏, which is thereby determined as text of γ, δ and ε, against ‫ בפניך‬in 𝔐.

A STEMMA FOR DEUTERONOMY

187

a) “Of the 23, in 3 variants Deutʳ agrees with 𝔐 against ⅏ (]‫ ח[לב‬vs. ‫חמת‬, 32:14; syndeton vs. asyndeton, 15:8; asyndeton vs. syndeton, 33:8).”158

Concerning the variants where 4QpaleoDeutʳ agrees with 𝔐 against ⅏, the reading of 4QpaleoDeutʳ is the common reading presupposed by 𝔊 in two of the three cases,159 isolating the ⅏ readings (replacing a word, and adding a copula) as singularity readings not shared by any other witness (S2), so excluding the possibility that 4QpaleoDeutʳ depends on ⅏. In the third case (presence of ‫ו‬-copulativum) 4QpaleoDeutʳ agrees with 𝔐 against the common text of ⅏ and 𝔊.160 This would be a family variant defining exclusive relationship (F1), but the presence of a ‫ו‬-copulativum, admittedly, is not strong enough to determine this. So accident is a better option to explain this agreement, supported by the fact that this manuscript shows a relatively high number of singularity readings against all other witnesses (S4, see below). b) “In 8 variants Deutʳ agrees with ⅏ against 𝔐 (presence vs. absence of ]‫ו[הטוב‬, 13:19; absence vs. presence of ‫את‬, 15:8; -‫ ל‬vs. -‫ב‬, 7:24; asyndeton vs. syndeton, 7:19; verb clearly Hiphil, 7:23 (see VAR.); singular vs. plural verb, 7:4; ‫יהם‬- vs. ‫ם‬-, 12:3; ‫ החו]צה‬vs. ‫חוץ‬, 23:14.”161

Among the eight variants where 4QpaleoDeutʳ agrees with ⅏ against 𝔐, there are six common readings shared with 𝔊, isolating the singularity reading of 𝔐 (S1).162 The Hebrew Vorlageof 𝔊 is not determined in two cases of “inner-Hebrew” variants, isolating 𝔐 by the exclusive agreement of 4QpaleoDeutʳ with ⅏ (H2).163 Hence, in virtually all cases of readings shared with ⅏ against 𝔐 where a statement on the Vorlageof 𝔊 is possible, 4QpaleoDeutʳ agrees with 𝔊 too. This means that our manuscript is clearly dependent on the common parent text of ⅏ and 𝔊, our manuscript γ in the stemma. Let us now turn to variants where 4QpaleoDeutʳ disagrees with both ⅏ and 𝔐: 158 159

160 161 162

163

DJD IX, 134. (17) Deut 32:14 ]‫ח[לב‬, = (the) fat (of the lambs, ‫)ח ֶלב ָכּ ִ ֜רים‬, ֵ֨ with 𝔐 𝔊 vs. ‫⅏ חמת‬ (āmat) = butter (defective for ‫ ;)חמאת‬and (21) Deut 33:8 asyndetic ‫תרי֯ ]בהו‬ ֯ with 𝔐 𝔊 vs. syndetic ‫⅏ ותריבהו‬. (11) 15:8 syndetic ‫והע ֯ב ֯ט‬ ֯ with 𝔐 vs. asyndetic ‫𝔊 ⅏ העבט‬. DJD IX, 134. (1) 7:4; (2) 7:19; (3) 7:23; (4) 7:24 (see the note above); (8) 13:19 (one of the few variants on word level, concerning ]‫ ו[הטוב‬present in resp. supposed by 4QpaleoDeutʳ ⅏ 𝔊 but absent from 𝔐) and (12) 23:14. (7) 12:3; (10) 15:8, absence of ‫את‬, one of the few variants on word level.

188

BENJAMIN ZIEMER

c)

“In 10 variants Deutʳ disagrees with both 𝔐⅏ (absence vs. presence of verse, 28:19; presence vs. absence of ‫לו‬, 33:7; absence vs. presence of definite article, 28:15; presence vs. absence of -‫ל‬, 33:7; syndeton vs. asyndeton, 12:1; asyndeton vs. syndeton, 33:29; different verbal form, 23:14; plural vs. singular verb, 7:25 and 28:15; ‫ון‬- vs. ‫ו‬-, 33:3).”

In seven of these ten cases there is no agreement with 𝔊 (S4): (1) 28:19 absence of verse with six words (parablepsisthrough homoioteleuton): ‫ ארור אתא בבאך וארור אתה בצאתך‬91 (‫)עשתרות צאנך‬ against 𝔐 ⅏ 𝔊, S4 (congruent) (2) 33:7 ‫ תהיה לו‬presence of ‫לו‬: against 𝔐 ⅏ 𝔊, S4 (congruent) (4) 33:7 ‫ לקול‬presence of –‫ל‬: against 𝔐 ⅏ 𝔊, S4 (congruent) (6) 33:29 ‫ יכחשו‬asyndeton: against 𝔐 ⅏ 𝔊, S4 (congruent) (7) 23:14 ‫ והשי[בות‬hifilagainst ‫ושבת‬: against 𝔐 ⅏ 𝔊, S4 (congruent) (8) 7:25 ‫ תחמדו‬plural against ‫תחמד‬: against 𝔐 ⅏ 𝔊, S4 (congruent)164 (9) 28:15 ‫ תשמעון‬plural against ‫ תשמע‬: against 𝔐 ⅏ 𝔊, S4 (congruent)

The Hebrew Vorlage of 𝔊 is not determined in two cases (“innerHebrew variants”): (3) 28:15 ‫ אלה‬without definite article: inner-Hebrew, against ⅏ 𝔐 = H4 (congruent) (10) 33:3 ‫ תכון‬with paragogic nun : inner-Hebrew, against ⅏ 𝔐 = H4 (congruent)

There is exclusive agreement with 𝔊 only in one minor case (presence of ‫ו‬-copulativum), explainable as accidental: (5) 12:1 ‫ ואלה‬syndeton: with 𝔊 against ⅏ 𝔐 = F3 (incongruent)

In two final cases all Hebrew witnesses are at variance: (1) ‫ תאוכל‬at the end of the line165: against 𝔐 and ⅏=𝔊, which disagree with each other (14:19) (2) ‫ גאיתך‬4QpaleoDeutʳ disagrees with 𝔐 ‫ 𝔊 = גַּ ֲאוָ ֶ ֑תָך‬καύχημά, and is closer to ⅏ ‫ גאתך‬ga᾿utak, which disagrees in orthography (33:29)

To sum up: 4QpaleoDeutʳ is a sister of ⅏, derived from the same parent text, γ. As a rule, 4QpaleoDeutʳ agrees with ⅏ against 𝔐 in those cases, where the reading of ⅏ is supported by 𝔊, whereas 4QpaleoDeutʳ agrees with ⅏ against 𝔊 in those cases where ⅏ is supported by 𝔐. On 164

165

There is a partial inconsistency with the stemma: the Temple scroll, which depends on a text closely related to 𝔊 but conflates it with other sources, agrees in this minor case with 4QpaleoDeutʳ against 𝔐, ⅏ and 𝔊. Therefore it is possible that 4QpaleoDeutʳ was in agreement with ⅏ and 𝔊.

189

A STEMMA FOR DEUTERONOMY

the other hand, 4QpaleoDeutʳ does not support any of the exclusive textual readings of ⅏. Since the manuscript itself shows several unique readings, among them the omission of Deut 28:19, it cannot be identical with γ or be itself an ancestor of ⅏ or 𝔊. Hence it is to be placed in the stemmaas a fourth manuscript dependent on γ, next to ⅏, δ and 4QDeutᵈ. If one looks at the inner-Hebrew evidence – script, textual variants and orthography – the picture changes slightly. The paleo-Hebrew script is, of course, a striking similarity with ⅏ against 𝔐. At the same time, 4QpaleoDeutʳ uses spaces for word division, like 𝔐 and the Qumran scrolls in the square script, whereas 4QpaleoDeutˢ and ⅏ use dots as word divisors. But both features have no relevance for the stemmasince they can easily be reversed. We know that 4QDeutᵇ·ᶜ, descendants of γ through δ, used the square script. But we cannot exclude the possibility that δ was copied from or was itself a paleo-Hebrew manuscript, or that 4QpaleoDeutʳ was copied from a parent text in the square script. Of the 4 inner-Hebrew textual variants, there are, as stated, 2 agreements of 4QpaleoDeutʳ with ⅏ against 𝔐 (H2), and none with 𝔐 against ⅏ (H1). Concerning the orthography, among the 27 deviations to 𝔐 and/or ⅏ stated in the DJD edition,166 there are 12 readings shared with ⅏, 4 shared with 𝔐, and 11 unique readings. The spelling practice is therefore closer to ⅏ than to 𝔐. The inner-Hebrew evidence, showing significantly more agreements in minor variants and in orthography with ⅏ than with 𝔐, allows hypothesizing a further common parent text for 4QpaleoDeutʳ and ⅏, namely η. In fact, 4QpaleoDeutʳ turns out to be, via dependence on η, at least as close to ⅏ as the so-called “pre-Samaritan” texts 4QpaleoExᵐ or 4QNumᵇ. However, if one looks at some specifics of the Samaritan reading tradition, the disagreements of 4QpaleoDeutʳ with some defective ⅏ spellings are remarkable too: reading of agreements / 4QpaleoDeutr disagreements (spelling) Deut 7:23

‫⅏ 𝔐 ≠ והומם‬ ‫⅏ ≠ 𝔐 = מהומה‬

Deut 14:21

166

DJD IX, 133.

𝔐

‫וְ ָה ָמ ֙ם‬

⅏ reading tradition

‫ והמם‬wāmimma

‫הוּמה‬ ֣ ָ ‫ מהמה ְמ‬māma

‫⅏ 𝔐 ≠ גדולה‬

‫גְ ד ֔ ָֹלה‬

‫ גדלה‬gādalla

‫⅏ ≠ 𝔐 = קדוש‬

‫דוֹשׁ‬ ֙ ‫ָק‬

‫ קדש‬qādesh

190

BENJAMIN ZIEMER

4QpaleoDeutʳ is relatively close to ⅏ via the dependence on γ, the common ancestor of ⅏ and 𝔊, and on η, a further parent text of ⅏. But, as determined by 4QpaleoDeutʳ, even η did not yet contain the genuine Samaritan idiosyncrasies, like the revised formula for the chosen place,167 or the specifics of the Samaritan reading tradition. 4.14. MasDeut (Mas1c) showing absolute agreement with 𝔐Aleppo The Masada scroll of Deuteronomy (Mas1c),168 a Herodian de luxe scroll,169 is one of the outstanding proto-masoretic manuscripts, although there are only 68 words or parts of words left of the last chapters of the book. There is one major difference between ⅏ on the one hand and 𝔐 and 𝔊 on the other hand in the section covered by MasDeut: the description of the promised land in Deut 34:1–3. The Masada manuscript agrees with 𝔐 and 𝔊 against ⅏. There are several minor differences between 𝔐, ⅏ and 𝔊, and the Masada manuscript agrees with 𝔐 in all these cases. It is perhaps more impressive yet that even in matters of orthography MasDeut agrees completely with the medieval Aleppo Codex. So it cannot be excluded that this manuscript has itself been a parent text for a parent text of the Aleppo Codex. That these two manuscripts are so close one to another in textual details is not self-evident. Benjamin Kennicott found more than 20 variants (including orthography) in medieval Hebrew manuscripts for the 68 words or parts of words present in MasDeut.170 In all these cases MasDeut agrees with the Aleppo Codex (both manuscripts not known to Kennicott) and the textusreceptus(t.r.) determined by Jacob ben Hayyim in his Bombergiana and presented by Benjamin Kennicott. But there is an orthographic difference between Codex L and the Aleppo Codex: ‫פוּנ֖י‬ ֵ ‫וּשׂ‬ ְ in the blessing for Zebulun (Deut 33:19) is written plene in Codex L, but defective in the Codex Aleppo (‫)וּשׂ ֻפ ֵנ֖י‬ ְ and in the textusreceptus. MasDeut (‫)ו[שפנ֯ י‬ ̇ agrees in this case with the Codex Aleppo, the t.r.and their masoraparva(‫וחס‬ ̇ ‫ל‬,̇ “unique and defective”) against Codex L.171 167

168 169 170 171

According to considerations of space it seems inevitable to follow the judgement of the editors that, in Deut 12:5, 4QpaleoDeutʳ read ‫אשר ]יב[חר‬, not ‫אשר ]ב[חר‬. Note that in the paleo-Hebrew script of 4QpaleoDeutʳ the ‫ י‬is one of the widest letters. (Without the ‫י‬, there would be additional empty space in the width of the word ‫ מן‬of Deut 12:3 in the line above, which is implausible.) Publication: Talmon, Scroll. Tov, Scribal Practices, 126. Kennicott, Vetus Testamentum, 437, 440–442. Talmon, Scroll, 156 (cited by Lange, Textbuch, 105) counts this spelling as “against the MT plene spelling ‫שפוני‬.” Talmon mentioned that “some printed editions exhibit

A STEMMA FOR DEUTERONOMY

191

5. CONCLUSION We get the following consistent stemmafor the book of Deuteronomy: 6RXUFHVRI'HXWHURQRP\DPRQJWKHPWKH6RQJRI0RVHV _ æ ĮWKHDUFKHW\SH 4SDOHR'HXWV" _ ╱_╲ _ _ȕ 4SDOHR'HXWV" _ ╱ _ ╱ 4'HXWL_╲ I V ╱ 4'HXW Ȗ 4SDOHR'HXW " _ ╱_╲ ࢫ(archaic version 0DV'HXW0XU'HXW4'HXWJ ╱ ╱4'HXWGį æ╲of the Song ╱___╲ æ╲of Moses) ╱ Ș ╱ ╱_4'HXWE4'HXWE╲ æ ╲ 4SDOHR'HXWU _4'HXWF4'HXWHİ æ 4QDeutq _ ╱╱æ╲ æ _ 4'HXWD" șTemple╲ æ _ ╱æScrollȗ+HEUHZVorlageRI 4'HXWK"  æ_ ╲ 4QTest 11QT

Truncated vertical lines ( ¦ ) indicate relations which actually do not belong to the stemmaof Deuteronomy since they concern the relationship to manuscripts of other literary works. These are introduced due to the striking proximities to proper Deuteronomy witnesses (4QTest: 4QDeutʰ, 11QT and 4QDeutq: 𝔊). Therefore the definition of ε, ϛ,172 ζ and θ is not as certain as the definition of α, β, γ, δ and η. It is, for instance, possible that the conflated version of the Song of Moses was already part of ε, or even of δ,173 since 4QDeutᵃ·ᵇ·ᶜ·ᵉ·ʰ are not extant in Deut 32; in this case ζ would be influenced by ϛ only indirectly. However, the excerpted text 4QDeutʲ is also influenced by ϛ. For convenience, I summarise data relevant for the evaluation of Judean desert Deuteronomy manuscripts included in the stemma:

172

173

the defective spelling ‫( ”שפני‬ibid., note 17), but he did not check the reading of Codex Aleppo. So he failed to realise that ‫ ושפוני‬is merely an isolated spelling of Codex L (agreeing by accident with ⅏!) against the Masoretic tradition witnessed by the best and oldest extant manuscripts (Codex Aleppo, Sassoon 507, Sassoon 1053) which spell ‫ושפני‬. According to the Masora parva of these manuscripts (Codex Aleppo, Sassoon 1053: “unique and defective”, cp. Breuer, Biblical Text, 42) the spelling of Codex L is nothing but an error (indeed a very frequent one in late medieval manuscripts as Kennicott, Vetus Testamemtum, 440, shows). I use here the stigmasign, since it replaces the old digamma(ϝ), as the letter between ε and ζ. Since 4QDeutᵃ·ᵇ·ᶜ·ᵉ·ʰ show only insignificant portions of Deut 32:1–3, we cannot know whether δ (or ε) already introduced the major deviations witnessed by 𝔊 or whether δ (and/ or ε) still retained the version of γ which is clearly defined by the agreement of ⅏ and 𝔐. However, the excerpted text 4QDeutʲ is also influenced by ϛ, either directly or indirectly.

192 Deuteronomy manuscripts included in the stemma

BENJAMIN ZIEMER

paleographic words / sum of agreements decisive for dating174 parts of textual stemmatic relationship words variants175 (* = see discussion above: partial agreement or agreement by accident)176

last defined parent text

F1 F2 F3 S4 Orthography 32

–

–

5

6

δ

175–150 BCE 97

5

1* –

2

–

ε (?)

150–100 BCE 177

17

–

6

2

δ

4QDeutᶜ (4Q30)

150–100 BCE 562

81

1* 1* 11 18

δ

4QDeutᵈ (4Q31)

125–75 BCE

253

24

–

2* 2* 3

γ

4QDeutᵉ (4Q32)

50–25 BCE

126

7

–

–

1

–

δ

4QDeutᶠ (4Q33)

75–50 BCE

404

45

5

–

1* 2

β

4QDeutᵍ (4Q34)

1–25 CE

151

18

1

–

–

–

4QDeutʰ (4Q35)

50–1 BCE

376

44

2* 1* 8

9 –

1QDeutᵇ (1Q5)

25–68 CE

4QDeutᵃ (4Q28) 4QDeutᵇ (4Q29)

i

237

–

θ (?) α

100–50 BCE

107

10

2

4QpaleoDeutʳ (4Q45)

100–25 BCE

380

23

1* –

1* 9

closer to ⅏

η

4QpaleoDeutˢ (4Q46)

ca. 400 BCE

11

1

–

–

–

–

=𝔐 =⅏

α?

MasDeut (Mas1c) 30–1 BCE

68

6

2

–

–

–

=𝔐

α

MurDeut (Mur2) 1–70 CE

32

2

–

–

–

–

=𝔐

α

175

176

–

α

4QDeut (4Q36)

174

–

=𝔐

Data according to Lange, Handbuch, 84–105, with one exception: for 4QpaleoDeutˢ Lange names the third century, but I follow the well founded new proposal of M. Langlois in this volume (p. 291: about 400 BCE). Data for 1QDeutᵇ–4QpaleoDeutʳ according to Lange. Though Lange gives sometimes lower figures for his statistics (causing discrepancies between this chart and the discussion above), he relies on the variants recorded in DJD. That means that some purely orthographic variants are counted like textual variants for 1QDeutᵇ (DJD I); singularity readings of 𝔊 (S3), a highly subjective matter, are counted for all manuscripts published in DJD I (1QDeutᵇ) and XIV (4QDeutᵃ–4QDeuti), but not at all for 4QpaleoDeutʳ published in DJD IX. Their number depends on the preferences of the particular editors; they sometimes even include inner-Greek variants. The statistics for 4QpaleoDeutˢ, MasDeut and MurDeut are my own, since Lange does not give any figures. The data represent my own evaluation of the variants recorded in DJD. The few differences to the evaluation by the DJD editors are discussed above. To be sure, I checked all family variants with the actual editions, and counted only agreements with 𝔊GÖ as F3 variants. The asterisked variants are explained above as accidental agreement; they are therefore no true family variants.

A STEMMA FOR DEUTERONOMY

193

To sum up the results for the textual history of Deuteronomy, 𝔐 turns out to be a witness faithful to content, words and letters of its parent texts. Its textual singularity readings are, as a rule, shorter and/or more difficult than the majority readings, and the orthography shows idiosyncrasies. For both text and orthography, it cannot be excluded that the text of 𝔐 is identical with that of the archetype α. The same is true for 4QDeutᵍ, MasDeut and MurDeut,177 three luxury manuscripts. So far as preserved, they do not show any textual or orthographical variant to 𝔐.178 4QDeuti is another offspring of α not specifically related to any other branch of the stemma, but by reason of one (inner-Hebrew) textual and three orthographical deviations, the manuscript cannot itself be an ancestor of 𝔐. ⅏ turns out to be an offspring of one of the first copies of Deuteronomy, the slightly revised manuscript γ. Thereby ⅏ is closely related to δ (one of the ancestors of 𝔊), 4QDeutᵈ and 4QpaleoDeutʳ. Inner-Hebrew textual and orthographical agreements define a further common ancestor η for ⅏ and 4QpaleoDeutʳ. Text and orthography are, as a rule, transmitted faithfully by ⅏; ⅏ shows less singularity readings (regarding textual details) than most Qumran manuscripts. However, there are singularity readings which affect the content of the book: the Gerizim passages (Deut 5:18 ⅏; 27:4 ⅏), the chosen place formula (Deut 12:5.11 etc. ⅏), the itinerary (Deut 10:6–7 ⅏) and the borders of the promised land (Deut 34:1–3 ⅏) are isolated by 𝔐 and 𝔊. These idiosyncrasies were not yet present in γ or η but are later developments. This isolation is confirmed by the Qumran evidence,179 so that Gesenius’ preconception that these features are secondary Samari177

178

179

Publication: J. T. Milik, DJD II/1, 78–79, and DJD II/2, pl. XXI. There are two singularity readings of 𝔊 in Deut 11:2 (‫ )ואת גדלו‬and 3 (‫ )ואת מופתיו‬isolated by MurDeut 𝔐 ⅏ (‫ את גדלו‬and ‫)ואת מעשיו‬, and in Deut 10:1–2 MurDeut shares two morphological variants (‫ ואכתב‬and ‫ )ואעש‬and one orthographical variant (‫ )הלחת‬exclusively with 𝔐 against ⅏. For the alleged exception of Deut 33:19 ‫ו[שפנ֯ י‬ ̇ in MasDeut see above: ‫ ושפני‬is the spelling of 𝔐, whereas ‫ ושפוני‬is in fact an orthographic deviation of 𝔐L against the Masoretic tradition. ‫ יבחר‬as reading of γ is already determined by the agreement of 𝔐 and 𝔊, but strengthened by the clear attestation in 4QDeutᵇ for Deut 31:11 and the reconstruction with a high level of certainty for Deut 12:5 in 4QpaleoDeutʳ. The same is true for the other named readings preserved in Dead Sea scrolls: see the testimony of all phylacteries and excerpted manuscripts against ⅏ at the end of Deut 5; the testimony of 4QDeutᶜ against ⅏ in Deut 10:6–7, and of 4QpaleoDeutʳ and MasDeut against ⅏ in Deut 34:1–3. For Deut 27:4, there is no additional evidence for the definition of γ in the Dead Sea scrolls. 4QDeutᶠ is not extent for the one differing word, whereas XQDeut?, the small fragment of unknown provenance published by Charlesworth (Charlesworth, Scroll), shows the Samaritan reading Hargerizim. However, since this fragment does not contain a single variant agreeing with 𝔐 or 𝔊 against ⅏, the dependence on ⅏ (or its immediate forerunner) cannot be excluded; it has therefore in no way an independent

194

BENJAMIN ZIEMER

tan developments cannot be refuted. The same is true for the peculiarities of Samaritan grammar, like the absence of an absolute infinitive. 𝔊 turns out to be a translation of a relatively inferior text,180 containing deviations cumulated in several stages. However, the evidence from the Judean desert enables us to show this development in more detail: A first copy, β, determined as branch-off for 4QDeutᶠ, eliminated some of the idiosyncrasies of α – mainly unusual readings and spellings. Two succeeding manuscripts are characterised by a slight tendency towards assimilating additions – γ determined as branch-off for ⅏, 4QDeutᵈ and 4QpaleoDeutʳ, and δ as branch-off for 1QDeutᵇ181 and 4QDeutᵇ·ᶜ·ᵉ. Most changes of β, γ and δ are explainable as conscious changes towards a greater conformity of the text. In contrast, the next copy, ε, shows a less clear pattern of deviations, including unconscious deviations like homoioteleuta. However, this poorly corrected copy was seemingly the first form of the text broadly known and copied outside the inner circles of Jerusalem and Samaria. So, this manuscript is important for the early Wirkungsgeschichteof Deuteronomy, as branch-off for the Temple Scroll and for ζ, the Hebrew 𝔊 Vorlage. Seemingly the 𝔊 translator was, at least partially, aware of this prehistory and tried to improve the text, thus creating further deviations. It is perhaps the translator himself who (re-)introduced the archaic elements in Deut 32:8, 43, taken from the separately transmitted Song of Moses, into the book of Deuteronomy. Beside ζ – the manuscript carried to Alexandria –, the manuscript ε and its descendants in the land of Israel served as parent texts for some of the oldest known actual Deuteronomy manuscripts (4QDeutᵃ) as well as excerpted manuscripts (5QDeut, 4QTest). In later times, manuscripts of this branch were less popular, and copies of the more faithful written and well corrected manuscripts β, γ or δ were accessible and popular. The Samaritans adopted a copy of such a manuscript (η). At least by the end of the first c. BCE, α itself or exact copies of it were used as model for luxury manuscripts of Deuteronomy like MasDeut, MurDeut and 4QDeutᵍ. The text of such manuscripts was later preserved by the Masora which thereby shows one of the oldest text types of the book Deuteronomy, if not the text of the archetype itself.

180 181

weight for the reconstruction of α or γ, irrespective the high probability that it is a forgery (cp., among others, Lange, Handbuch, 106). Cp. Tov, Samaritan Pentateuch, 410, with similar conclusions. Publication: D. Barthélemy, DJD I, pp. 57–62 and pls. X. The five F3 variants (1QDeutᵇ 𝔊 〛 𝔐 ⅏) are the following: Deut 15:15 ‫ ;מצוך 〛 מצוך לעש]ות‬Deut 24:14 ‫ ;שכיר 〛 שכר‬Deut 31:1 ‫ ;וילך 〛 ויכל‬Deut 31:1 ‫ ;וידבר 〛 לדבר‬Deut 31:1 ‫〛 כל הד]ברים‬ ‫הדברים‬.

A STEMMA FOR DEUTERONOMY

195

6. BIBLIOGRAPHY BEN-HAYYIM, Ze᾿ev. TheLiteraryandOralTraditionofHebrewandAramaic amongsttheSamaritans,Vol.4:TheWordsofthePentateuch. Jerusalem: The Academy of the Hebrew Language, 1977. BREUER, Mordechai. TheBiblicalTextintheJerusalemCrownEditionandits SourcesintheMasoraandManuscripts[Hebrew]. Jerusalem: Keren HaMasora 2003. BROOKE, George J. “Deuteronomy 5–6 in the Phylacteries from Qumran Cave 4.” Pages 58–70 in: Emanuel:StudiesinHebrewBible,SeptuagintandDead SeaScrollsinHonorofEmanuelTov. Supplements to Vetus Testamentum 94. Edited by Sh. M. Paul et al. Leiden: Brill, 2003. CHARLESWORTH, James Hamilton. “‫ – ַה ְבּ ָר ָכה ַעל ַהר־גְּ ִרזִ ים‬An Unknown Dead Sea Scroll and Speculations Focused on the Vorlage of Deuteronomy 27:4.” Pages 393–414 in: Jesus,PaulusunddieTextevonQumran.Wissenschaftliche Untersuchungen zum Neuen Testament 2, 390. Edited by J. Frey and E.E. Popkes.Tübingen: Mohr Siebeck, 2015. DJD I: BARTHÉLEMY, D., and MILIK, J.T. Qumran Cave I. Discoveries in the Judaean Desert I. Oxford: Clarendon Press, 1955. DJD II: BENOIT, P., et al. LesGrottesdeMurabba‘ât. Discoveries in the Judaean Desert II. Oxford: Clarendon Press, 1961. DJD III: BAILLET, M., et al. LesPetitesGrottesdeQumrân. Discoveries in the Judaean Desert III. Oxford: Clarendon Press, 1962. DJD IX: SKEHAN, Patrick W. et al., eds. QumranCave4.IV.Paleo-Hebrewand GreekBiblicalManuscripts. Discoveries in the Judaean Desert IX. Oxford: Clarendon Press, 1992. DJD XIV: ULRICH, Eugene et al., eds. QumranCave4.IX.Deuteronomy,Joshua, Judges, Kings. Discoveries in the Judaean Desert XIV. Oxford: Clarendon Press, 1995. DUNCAN, Julie A. “New Readings for the ›Blessing of Moses‹ from Qumran.” JournalofBiblicalLiterature114 (1995): 273–90. —. “Excerpted Texts of Deuteronomy from Qumran.” Revue de Qumran 18 (1997–98): 43–62. ESHEL, Esther, and ESHEL, Hanan. “Dating the Samaritan Pentateuch’s Compilation in Light of the Qumran Biblical Scrolls.” Pages 215–40 in: Emanuel:Studies inHebrewBible,SeptuagintandDeadSeaScrollsinHonorofEmanuelTov. Supplements to Vetus Testamentum 94. Edited by Sh. M. Paul et al. Leiden: Brill, 2003. HECKL, Raik. “Der ursprüngliche Anfang des Deuteronomiums und seine literarische Transformation.” ZeitschriftfüraltorientalischeundbiblischeRechtsgeschichte20 (2014): 71–96. JACOB, Benno. DasBuchGenesis.Berlin: Schocken Verlag, 1934. Repr., Stuttgart: Calwer Verlag, 2000. LANGE, Armin. Handbuch der Textfunde vom Toten Meer. Band I: Die HandschriftenbiblischerBüchervonQumranunddenanderenFundorten.Tübingen: Mohr Siebeck, 2009. KENNICOTT, Benjamin. Vetus Testamentum Hebraicum, cum variis lectionibus. TomusPrimus.Oxford: Clarendon, 1776.

196

BENJAMIN ZIEMER

MAAS, Paul. Textkritik.4th ed. Leipzig: B.G.Teubner Verlagsgesellschaft, 1960. MCCARTHY, Carmel. Deuteronomy. BHQ 5. Stuttgart: Deutsche Bibelgesellschaft, 2007. Josef OFER and Alexander LUBOTZKY. “The Masora as an Error Correcting Code.” [Hebrew]. Tarbiz82 (2014) 89–113. PAKKALA, Juha. God’s Word Omitted. Omissions in the Transmission of the HebrewBible. Forschungen zur Religion und Literatur des Alten und Neuen Testaments 251. Göttingen: Vandenhoeck & Ruprecht, 2013. POLAK, Frank H. “Statistics and Textual Filiation. The Case of 4QSama/LXX (With a Note on the Text of the Pentateuch).” Pages 215–276 in: Septuagint, ScrollsandCognateWritings.PapersPresentedtotheInternationalSymposiumontheSeptuagintandItsRelations.Society of Biblical Literature Septuagint and Cognate Studies 33. Edited by G.J. Brooke and B. Lindars. Atlanta: Scholars Press,1990. SCHIFFMANN, H. Lawrence “The Septuagint and the Temple Scroll: Shared ‘Halakhic’ variants.” Pages 277–97 in: Septuagint,ScrollsandCognateWritings. PapersPresentedtotheInternationalSymposiumontheSeptuagintandIts Relations. Society of Biblical Literature Septuagint and Cognate Studies 33. Edited by G.J. Brooke and B. Lindars. Atlanta: Scholars Press,1990. TALMON, Shemaryahu. “Fragments of a Deuteronomy Scroll from Masada: Deut 33.17–34.6 (1043/A-D).” Pages 150–161 in BoundariesoftheAncientNear Easternworld.AtributetoCyrusA.Gordon.Journal for the Study of the Old Testament: Supplement series 273. Edited by M. Lubetski et al. Sheffield: Sheffield Academic Press, 1998. THIESSEN, Matthew. “The Form and Function of the Song of Moses (Deuteronomy 32:1-43).” JournalofBiblicalLiterature123 (2004): 401-424. TOV, Emanuel. “The Contribution of the Qumran Scrolls to the Understanding of the LXX.” Pages 11–47 in: Septuagint,ScrollsandCognateWritings.Papers PresentedtotheInternationalSymposiumontheSeptuagintandItsRelations. Society of Biblical Literature Septuagint and Cognate Studies 33. Edited by G.J. Brooke and B. Lindars. Atlanta: Scholars Press,1990. —. “The Samaritan Pentateuch and the Dead Sea Scrolls: The Proximity of the Pre-Samaritan Qumran Scrolls to the SP.” Pages 387–410 in: TextualCriticism of the Hebrew Bible, Qumran, Septuagint. Collected Essays, Vol. 3. Supplements to Vetus Testamentum 167. Leiden: Brill 2015. —. ScribalPracticesandApproachesReflectedintheTextsFoundintheJudean Desert.Studies on the Texts of the Desert of Judah 54. Leiden: Brill, 2004. —. “Textual Criticism of Hebrew Scripture and Scripture-Like Texts.” Pages 205– 220 in: TextualCriticismoftheHebrewBible,Qumran,Septuagint.Collected Essays,Vol.3. Supplements to Vetus Testamentum 167. Leiden: Brill 2015. —. TextualCriticismoftheHebrewBible. 3rd ed., rev. and exp. Minneapolis: Fortress Press, 2012. —. “Textual Harmonizations in the Ancient Texts of Deuteronomy.” Pages 271– 282 in: HebrewBible,GreekBible,andQumran.CollectedEssays.Texts and Studies in Ancient Judaism 121. Tübingen: Mohr Siebeck, 2008. TOV, Emanuel, and POLAK, Frank. TheRevisedCATSSHebrew/GreekParallel Text. Electronic text hypertexted and prepared by OakTree Software, Inc. © Emanuel Tov and Frank Polak, 2009.

A STEMMA FOR DEUTERONOMY

197

ULRICH, Eugene, ed. The biblical Qumran scrolls. Transcriptions and Textual Variants.VolumeI:Genesis–Kings.Leiden: Brill, 2013. WHITE, Sidnie. “Special Features of Four Biblical Manuscripts from Cave IV, Qumran: 4QDta, 4QDtc, 4QDtdand 4QDtg.” RevuedeQumran15 (1991– 92): 157–67. —. “Three Deuteronomy Manuscripts from Cave 4.” JournalofBiblicalLiterature112 (1993): 23–42. ZIEMER, Benjamin. “Prophetenrede und Gottesrede im Pentateuch und der Ausgang der Schriftprophetie.” Pages 441–466 in: DieunwiderstehlicheWahrheit.StudienzuralttestamentlichenProphetie.FestschriftfürArndtMeinhold. Arbeiten zur Bibel und ihrer Geschichte 23. Edited by R. Lux and E.-J. Waschke. Leipzig: Evangelische Verlags-Anstalt, 2006.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN: LEGAL MATERIAL OF LEVITICUS AND DEUTERONOMY Innocent HIMBAZA

1. LEGAL MATERIAL OF LEVITICUS AND DEUTERONOMY 12-26 It is often alleged that the legal sections of the Torah were not subject to large-scale textual changes, since legal material is less susceptible to change1. The book of Leviticus is usually taken as a prime example of the lack of large-scale textual intervention. In this contribution, I would like to examine in more detail the question raised in recent years of the character of the Samaritan Pentateuch (SP) compared with some harmonistic Qumran manuscripts, labeled “biblical” or not, such as 4QReworked Pentateuch (or 4QPentateuch) and the Temple Scroll. I would like particularly to study the legal material contained in Leviticus and Deuteronomy of the SP, compared with such harmonistic manuscripts. I shall sometimes refer to the Septuagint (LXX) in order to include more textual witnesses. However the conclusions will focus on the SP and Qumran manuscripts. In her book “Rewriting Scripture in Second Temple Times,” Sidnie White Crawford identified a specific exegetical tradition among some Rewritten Scripture texts labeled “the priestly-levitical/essene line of interpretation” or “the priestly-levitical/essene exegetical tradition”.2 This line of interpretation dated at least to the 3rd century BCE is characterized (1) by the idea that the Law was observed by the righteous ancestors before Moses, (2) by an emphasis on the Levites and the enhancing of their status, and (3) by the expansion of the Temple purity to the sphere of everyday life. 1

2

Michael Segal, “The Text of the Hebrew Bible in Light of the Dead Sea Scrolls,” Materaia Giudaica 12, (2007) 5-20, esp. 14; Alex P. Jassen, ScriptureandLawintheDeadSea Scrolls, (New York: Cambridge University Press, 2014), 25. Sidnie White Crawford, Rewriting Scripture in Second Temple Times, Grand Rapids, Michigan/Cambridge, U.K.: William B. Eerdmans, 2008. The topic is addressed throughout the book. For a synthesis, see 146-149.

200

INNOCENT HIMBAZA

The question raised here is to what extent legal material from Leviticus and Deuteronomy may say anything about this kind of claim. While the Temple Scroll has been studied many times3, only a few publications have compared its text with the SP. In his statistical study published more than thirty years ago (in 1982), Emanuel Tov compared the Temple Scroll (11QTa = 11Q19) on the one hand and the Massoretic Text (MT), the LXX and SP on the other hand. He observed the following textual relationship: 11QTa = LXX = SP ≠ TM 22× 11QTa = LXX ≠ SP ≠ TM 26× 11QTa = SP ≠ LXX 2× 11QTa = SP = TM 6× 11QTa ≠ SP 11×

He concluded that there was “no solid base for assuming a special relationship between the scroll and either the Masoretic Text, the Septuagint or the Samaritan Pentateuch. The Scroll contains a textual tradition which agrees now and then with one or the other of these early texts.”4 The Temple Scroll has a particular position in this discussion since it claims the status of a law code5 where God is speaking in the first person singular and parallel commandments are harmonized. The earliest representative of this text, 4QT (4Q524), is dated between 150 and 125 BCE, however the well preserved version, 11QTa (11Q19), is from the first century CE6. We now know that the Temple Scroll is not a creation of the Qumran 3

4

5

6

For instance, Lawrence Schiffman dedicated no less than thirty-three studies to it. See Lawrence H. Schiffman, TheCourtyardsoftheHouseoftheLord.StudiesontheTemple Scroll, ed. Florentino García Martínez, STDJ 75, Leiden, Boston: Brill, 2008; Emanuel Tov, «Deuteronomy 12 and 11QTemplea LII-LIII: A Contrastive Analysis », in Hebrew Bible,GreekBible,andQumran.CollectedEssays, ed. Emanuel Tov, (Tübingen: Mohr Siebeck, 2008), 17-20. Emanuel Tov, ‫“( ׳מגילת המקדש׳ וביקרת נוסח המקרא‬The ‘Temple Scroll’ and Old Testament Textual Criticism”), in Eretz-Israel 16, Archaeological, Historical and Geographical Studies,HarryM.OrlinskyVolume, ed. Baruch A. Levine, Abraham Malamat, (Jerusalem: Israel Exploration Society, Hebrew Union College, Jewish Institute of Religion, 1982), 100-111 (English summary, 255). In 11QRTa LVI, 20-21 citing Deut 17:18, the text reads: ‫על ספר מלפני הכוהנים וכתבו לו‬ ‫( התורה הזואת‬and they shall write for him this Torah in the book in the presence of the priests), instead of ‫“ את משנה התורה הזאת‬this copy of the Torah” of MT. Emile Puech, QumrânGrotte4.XVIII.TextesHébreux(4Q521-4Q528;4Q576-4Q579), (DJD 25, Oxford: Clarendon Press, 1998), 87; James H. Charlesworth (ed.), Hebrew, Aramaic,andGreekTextswithEnglishTranslations,Volume7:TempleScrolland RelatedDocuments, (The Dead Sea Scrolls, Tübingen: Mohr Siebeck; Louisville: Westminster John Knox Press, 2011), 247-249; Before the publication of 4Q524, Michael Wise dated the Temple Scroll around 150 BCE. See Michael Owen Wise, ACritical StudyoftheTempleScrollfromCave11 (Studies in the Ancient Oriental Civilization 49,

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

201

Community but among their received literature7. According to many scholars, chapters LI-LXVI of the Temple Scroll are from a Deutoronomy Paraphrase that Wise8 dates between 300 and 190 BCE and which Schiffman considers to stem from the author/redactor of the Temple Scroll himself 9. The material for the comparative study shall be taken from those chapters. The text of the SP is harmonistic, and harmonization aims at avoiding contradictions and divergences. This principle is also well known outside the pre-Samaritan text, for example in 4QDeutn, whose status is the subject of much discussion. However, Samaritans may have received books of the Pentateuch already harmonized. Thus their own harmonistic interventions should be considered as very limited10. 2. LEVITICUS 2.1. Textual Comparison References

Qumran manuscripts

Pluses to MT

Pluses contained in SP

Lev 1:17

4QLevb

Lev 2:1 Lev 2:1

‫ ו‬of ‫ולא‬

+

b

‫ ו‬of ‫קרבנו‬

–

b

‫מנחה ]היא‬

+

4QExod-Lev 4QExod-Lev

Lev 3:1

4QLev

‫ליהוה‬

–

Lev 3:6

4QLeve

‫ ה‬of ‫השלמים‬

+

Lev 4:6

4QpapLXXLevb

δακτυλωι (‫)באצבעו‬

+

‫חטא אשר‬

–

Lev 4:23

7 8 9

10

4QLev

b

c

Chicago, Illinois: The Oriental Institute of the University of Chicago, 1990), 189-194; Casey D. Elledge, TheStatutesoftheKing:TheTempleScroll’sLegislationofKingship(11Q19LVI12-LIX21) (Cahiers de la Revue biblique 56, Paris: Gabalda, 2004), 37-45. Among which Enoch, Jubilees and the Aramaic Levi Document are counted. Michael Owen Wise, ACriticalStudy, 60. Lawrence Schiffman, “The Deuteronomic Paraphrase of the Temple Scroll,” in TheCourtyardsoftheHouseoftheLord, 443-469, esp. 447, 468-469. According to Schiffman, the author used a Vorlage of Deuteronomy containing textual differences compared to MT. Sidnie White Crawford, RewritingScripture, 89, 93-102, agrees with Schiffman. Esther Eshel and Hanan Eshel, “Dating the Samaritan Pentateuch’s compilation in light of the Qumran biblical scrolls,” in Emanuel.StudiesinHebrewBibleSeptuagintandDead SeaScrollsinHonorofEmanuelTov, ed. Shalom M. Paul, Robert A. Kraft, Lawrence H. Schiffman, Weston W. Fields (VTS 94, Leiden, Boston: Brill, 2003), 215-240. The Samaritan tenth commandment found in Exod 20 and Deut 5 does not result from a harmonization since it is a late insertion for a different literary reason.

202

INNOCENT HIMBAZA

References

Qumran manuscripts

Pluses to MT

Pluses contained in SP

Lev 4:25 ?

11QpaleoLeva

‫ ישחט את העלה ך]ושפ‬/‫)≠( ו‬ ‫( ישפך דמו‬LXX ‫מזבח )כל‬ ‫העלה ואת‬

–

Lev 4:26 ?

11QpaleoLeva

‫השלמים וכפר )≠( יהוה וכפר‬

–

Lev 7:25

4QLevg

‫אשה )≠( קרבן‬

–

Lev 9:23

11QLevb

‫כו]ל‬

–

Lev 9:24

b

11QLev

‫השלמ]ם‬

–

Lev 10:1

11QLevb

‫שנ]י‬

–

Lev 10:1

b

‫( יהו[ה‬in paleo-hebrew)

–

‫ ו‬of ‫וכל‬

+

Lev 11:20

11QLev 4QRP

c

Lev 11:25

2QpaleoLev

‫ורחץ במים‬

Lev 11:27

11QpaleoLeva

‫כפיו )≠( גחו]ן‬

–

Lev 11:42

4QRPc

‫על )≠( א]ל‬

–

Lev 11:44

4QRP

c

Lev 11:45

4QRPc

Lev 13:42

11QpaleoLeva

Lev 14:17

a

11QpaleoLev

+

[‫]יהוה אלוהיהמה‬

–

‫אלהיכם‬

+

‫ ו‬of ‫בקרחתו‬

+

‫תנוך ≠ ראתנ]ו[כ‬

– –

Lev 14:36

4QLev

‫הצרעת‬

Lev 14:43

4QLev-Numa

‫ י‬of ‫אחרי‬

+

Lev 14:45

a

֯‫ ו‬of ֯‫ו[נ֯ ֯ח ֯צו‬

+

d

4QLev-Num

Lev 14:46

4QLev-Num

֯‫ ו‬of ] ֯‫הסג[י֯ ֯רו‬

–

Lev 14:49

4QLev-Numa

‫לחטא )≠( לטחר‬

–

Lev 14:50

a

suffix ‫ ו‬of ‫ושחטו‬

+

‫אל )≠( על‬

=

‫ימי ז]ב בשרו או והחתים בשרו‬ ‫מזבו הטמאתו היא [בו כל‬

+

Lev 14:51

a

4QLev-Num

a

4QLev-Num

a

Lev 15:3

11QpaleoLev

Lev 17:3

– 4QLevd

‫והגר ה[גר בישראל‬

–

Lev 17:3

– 11QpaleoLev

‫ו‬

–

Lev 17:4

– 4QLevd

‫]לעשות אתו עלה [או שלמים‬ ‫ליהוה לרצונכם ל]ריח ניחח‬ ‫וישחטהו בחוץ ואל פתח אהל‬ ‫מו[עד לא יביאנו‬

+ (‫)…הביאו‬

Lev 17:4

– 4QLevd

‫ ו‬of ‫להקריבו‬

+

‫את‬

+

Lev 17:10

d

4QLev

a

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

References

Qumran manuscripts

Pluses to MT

Lev 17:11

4QLevd

‫ה )≠( כול‬

–

Lev 17:11

4QLevd

‫ ו‬of ‫בדמו‬

–

Lev 18:27

11QpaleoLeva

‫ ̇ה‬of ‫האלה‬ ̇

+

‫ ואמר לכמ א[תמ‬24 ‫ואקצ בם‬ ‫אד]מ[תם‬ ̇ ‫תירשו] [את‬

–

‫כי‬

+

a

Lev 20:23-24 11QpaleoLev inserted in 18:27 after ‫התעבות האלה‬ ‫עשו‬ Lev 18:30

11QpaleoLeva

Pluses contained in SP

Lev 20:2

11QpaleoLev

‫מבני )≠( מבית‬

=

Lev 21:9

4QLeve

‫בית‬

–

Lev 22:5

e

‫טמא‬

+

‫ ו‬of ‫ונפ]ש‬

–

Lev 22:6

a

4QLev

a

1QpaleoLev-Num

Lev 22:18

4QLev

‫הגר‬

+

Lev 22:18

4QLevb

‫ו )≠( או‬

–

Lev 22:21

4QLevb

‫ ב‬of ‫לנדבה )≠( בנדבה‬

≠ ‫נדבה‬

Lev 22:22

b

4QLev

‫]אש[ך‬ ֯ ‫מרוח או‬

–

Lev 22:22

4QLevb

‫אלה )≠( אשה‬

–

Lev 22:24

11QpaleoLeva

Lev 23:43

b

‫אלה‬

–

c

‫בני ישראל )≠( אבותיכם‬

–

c

4QRP

Lev 24:2++

4QRP (frg 23)

Large-scale textual change, cf. § 2.5.

–

Lev 24:10

4QLevb, 11QpaleoLeva

‫ ה‬of ‫והאיש‬

–

Lev 24:12

11QpaleoLeva

‫ויניחוהו ≠ ויניחו אתו‬

–

Lev 24:14

a

‫ ו‬of ‫הציאו‬

–

‫תרועה‬

–

Lev 25:9

11QpaleoLev c

4QRP

Lev 25:31

11QpaleoLev

‫ ו‬of ‫יחשבו‬

+

Lev 26:4

4QLXXLeva

τ[ηι γηι

–

Lev 26:24

4QRPc

‫אף )≠( גם‬

=

a

Lev 26:24

11QpaleoLev

‫חמת‬

–

Lev 27:17

4QLev-Numa

‫ ו‬of ‫ואם‬

+

Lev 27:19

11QpaleoLeva

‫את‬

–

‫כול‬

–

Lev 27:34

a

c

4QRP

203

204

INNOCENT HIMBAZA

2.2. Statistics11 Qumran Manuscript a

1QpaleoLev-Num

Pluses to MT 1

Pluses contained in SP 0

2QpaleoLev

1

1

4QExod-Levb

2

1

4QLevb

8

2

c

1

0

d

7

3

e

4QLev

3

2

4QLevg

1

0

7

5

1

0

1

1

9

3

4QLev 4QLev

a

4QLev-Num

a

4QLXXLev

b

4QpapLXXLev c

4QRP

b

4

0

11QpaleoLeva

18

6

Total

64

24

11QLev

2.3. Additional text in Lev 15:3 MT ‫וזאת תהיה טמאתו בזובו רר בשרו את־זובו או־החתים בשרו מזובו טמאתו הוא׃‬ This shall be his uncleanness of his discharge: his flesh flows with his discharge or (when) his flesh is blocked from his discharge, it is his uncleanness.

SP ‫וזאת תהיה טמאתו בזובו רר בשרו את זובו או חתום בשרו מזובו טמא הוא כל‬ ‫ימי זוב בשרו או החתים בשרו מזובו טמאתו היא׃‬ This shall be his uncleanness of his discharge: his flesh flows with his discharge or his flesh is blocked from his discharge. He is unclean all

11

The new edition published by Schorch confirms the pluses of SP, except in Lev 11:20 and 14:50. See Stefan Schorch (ed.), Leviticus,TheSamaritanPentateuch.ACritical EditioMajor,Volume3, Berlin / Boston: De Gruyter, 2018.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

205

the days of discharge of his flesh or (when) his flesh is blocked from his discharge, it is his uncleanness.

11QpaleoLeva, Fragment G, line 7 ‫הוא[ בו כל ימי ז]ב‬

According to Freedman and Mathews, the fragmentary ‘plus’ of 11Qpaleo-Leva is placed at the end of the verse.12 Its reading corresponds to ἐν αὐτῷ·πᾶσαι αἱ ἡμέραι ῥύσεως of LXX. According to Ulrich and KuglerBaek13, the reconstruction of 11Qpaleo-Leva for Lev 15:3 shows the following text : ‫וזאת תה[יה טמאתו בזב]ו רר בשרו את־זבו או־החתים בשרו מזבו טמאתו היא [בו‬ ‫כל ימי ז]ב בשרו או החתים בשרו מזבו טמאתו היא׃‬ This shall] be his uncleanness of his disc[harge: his flesh flows with his discharge or his flesh is blocked from his discharge. It is his uncleanness] in him all the days of disc[harge of his flesh or (when) his flesh is blocked from his discharge, it is his uncleanness].

The comparison between SP on the one hand, and 11QpaleoLeva on the other hand, shows that they contain two redactional differences. Firstly, the ‘plus’ of SP is located in the middle of the verse, while that of 11QpaleoLeva is at the end of the verse. Secondly, the equivalent of the preposition ‫ בו‬is lacking in SP. LXX, which is not studied in this contribution, shows many connections with 11Qpaleo-Leva, though according to Wevers, LXX depends on SP14. The majority of scholars consider the additional text of SP, 11QpaleoLeva (and LXX) to be original, while MT reflects a parablepsis by homoioteleuton.15 The parablepsis in 11Qpaleo-Leva is explained as ‫טמאתו‬: 1⁀2, 12

13

14

15

David N. Freedman and Kenneth A. Mathews, The Paleo-Hebrew Leviticus Scroll (11QpaleoLev), (Winona Lake, Indiana: American School of Oriental Research, 1985) 32-33. Cf. Robert A. Kugler and Kyung S. Baek, LeviticusatQumran.TextandInterpretation, (VTS 173; The Text of the Bible at Qumran, Leiden, Boston: Brill, 2016), 24. Eugene Ulrich (ed.), TheBiblicalQumranScrolls.TranscriptionsandTextualVariants, (VTS 134; The Text of the Bible at Qumran, Leiden, Boston: Brill, 2010), 120; Robert A. Kugler and Kyung S. Baek, LeviticusatQumran, 24. See John William Wevers NotesontheGreekTextofLeviticus, (SCSS 44, Atlanta, GA: Scholars Press, 1997), 225. Wevers ad other scholars assert that from the LXX’s point, the paraplepsis is explained as ‫מזובו‬: 1⁀2. See Paul Harlé, Didier Pralon, LeLévitique, (La Bible d’Alexandrie 3, Paris: Editions du Cerf, 1988), 146-147; Jacob Milgrom, Leviticus1-16.ANewTranslationwithIntroductionandCommentary, (AB 3, New York, London, Toronto, Sydney, Auckland: Doubleday, 1991), 908-909. Jacob Milgrom, Leviticus1-16, 908-909; John E. Hartley, Leviticus, (WBC 4, Nashville: Thomas Nelson Publishers, 1992), 203; René Péter-Contesse, Lévitique1-16, (CAT IIIa, Genève : Labor et Fides, 1993), 234-235. For a different view, see David Andrew Teeter,

206

INNOCENT HIMBAZA

while from the SP’s point, it seems to be ‫או‬: 1⁀216. SP may reflect the earliest location and formulation of the additional text, while 11QpaleoLeva and LXX reflect its later insertion. Thus the common witness of SP, 11QpaleoLeva and LXX allows us to recover the missing text in MT, while its exact formulation and its transmission history remain complex. 2.4. Additional text in Lev 17:4 The reading of Lev 15:3 is important for the discussion of Lev 17:4. Indeed, while these two verses take the same general form (‫ואל פתח אהל‬ ‫מועד לא הביאו‬: 1⁀2), they are often analyzed differently. MT (= 11QpaleoLeva) ‫ואל־פתח אהל מועד לא הביאו להקריב קרבן ליהוה לפני משׁכן יהוה דם יחשׁב‬ ‫לאישׁ ההוא דם שׁפך ונכרת האישׁ ההוא מקרב עמו׃‬ … And does not bring it to the entrance of the tent of meeting, to present an offering to Yhwh before the tabernacle of Yhwh, that man shall be held guilty of bloodshed; he has shed blood, and that man shall be cut off from the people.

SP (= 4QLevd, LXX) ‫ואל פתח אהל מועד לא הביאו לעשות אתו עלה או שלמים ליהוה לרצונכם לריח‬ ‫ניחח וישחטהו בחוץ ואל פתח אהל מועד לא הביאו להקריבו קרבן ליהוה לפני‬ ‫משכן יהוה דם יחשב לאיש ההוא דם שפך ונכרת האיש ההוא מקרב עמיו׃‬

SP, 4QLevd and LXX contain some slightly orthographic and redactional differences: ‫ לא הביאו‬SP | ‫ לוא יביאנו‬4QLevd (cf. MT v. 9) ‫ עלה‬SP | εἰς ὁλοκαύτωμα LXX ‫ לרצונכם‬SP | δεκτὸν LXX. … And does not bring it to the entrance of the tent of meeting, to make it a burnt offering or a peace offering to Yhwh for acceptance on your behalf as a pleasing odor and slaughter it outside and does not bring it to the entrance of the tent of meeting to present it as an offering to Yhwh before the tabernacle of Yhwh, that man shall be held guilty of bloodshed; he has shed blood, and that man shall be cut off from the people.

16

Scribal Laws, Exegetical Variation in the Textual Transmission of Biblical Law in the LateSecondTemplePeriod, (FAT 92, Tübingen: Mohr Siebeck, 2014), 94-99. In their synoptic comparison between MT and SP, Tal-Florentin locate the ‘plus’ of SP in the middle of the verse (‫או‬: 1⁀2). See Abraham Tal & Moshe Florentin, ThePentateuch.TheSamaritanversionandtheMasoreticVersion, (Tel Aviv: The Haim Rubin Tel Aviv University Press, 2010), 363.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

207

According to many scholars the additional text as attested in SP, 4QLevd and LXX should not be understood as the earliest text, while it is missing in MT due to parablepsis occasioned by homoioteleuton, as was the case in Lev 15:3. Those scholars also observe that it does not result from a harmonization with a parallel verse.17 At least it may have been composed with some elements from vv. 5, 8 and 9 where parallel words are found (5: ‫שלמים‬, 8: ‫ עלה‬and 9 ‫)לעשות אתו ואל פתח אהל מועד לא יבאנו‬. Now, the question is: what is the difference between the character of Lev 15:3 and that of 17:4? Scholars assert that the additional text was composed in order to precise that the ‫( קרבן‬offering in its generic meaning) of this verse contains both ‫( עלה‬burnt offering) and ‫( שלמים‬peace offering). Indeed in this paragraph (v. 3-7) only ‫ שלמים‬is mentioned by MT18. If the additional text of Lev 17:4 was composed for this purpose, it is of course not a large-scale textual change, but a textual change all the same. In that case, one should conclude that some Qumran manuscripts and SP contain at least one scale textual change added to the legal material of Leviticus. The goal may have been avoiding misreading by specifying sacrificial issues. 2.5. Additional text of 4QRPc (4Q365), Frg 23 Lev23:44-24:1 One of the main questions of this fragment is whether it should be considered as part of 4Q365 or as a different composition. And if it is a different composition the question arises as to whether it can be part of 4QTemple (?). Tov and White have discussed this question and showed its complexity19. In my opinion, the deciding argument in favor 17

18

19

Juha Pakkala, God’sWordOmitted.OmissionintheTransmissionoftheHebrewBible, (FRALANT 251, Göttingen: Vandenhoeck & Ruprecht, 2013), 95-96; David Andrew Teeter, ScribalLaws, 76-94; Robert A. Kugler, Kyung S. Baek, LeviticusatQumran, 97-98. Christophe Nihan, who does not refer to the additional text of SP, 4QLevd and LXX, points out problems in understanding the MT. First, should ‫ שחט‬of Lev 17:3 be considered as “mere killing of animals” or “interpreted in a sacrificial sense”; and second, “why this sacrifice should be a well-being offering and not a burnt offering.” Christophe Nihan, FromPriestlyTorahtoPentateuch.AStudyintheCompositionoftheBookof Leviticus (FAT 2 25, Tübingen: Mohr Siebeck, 2007), 410. It seems to me that the additional text is a literary expansion which answers such questions. See David Andrew Teeter, ScribalLaws, p. 89-92. Emanuel Tov, Sidnie White, “Reworked Pentateuch,” in QumranCave4.VIII:ParabiblicalTexts,PartI, ed. Harold Attridge et. al. (DJD XIII, Oxford: Clarendon Press, 1995, 187-352, esp. 290-296.

208

INNOCENT HIMBAZA

of keeping it as part of 4Q365 is that this fragment contains a literary quotation of a narrative text. Biblicalnarrative ‫ וידבר משוה את מועדי יהוה אל בני ישראל‬23:44 3 ‫ צו את בני ישראל לאמור‬2 ‫ וידבר יהוה אל מושה לאמור‬24:1 4

Actually, the Temple Scroll functions differently. God is directly speaking to Moses, or seemingly to Moses and to the whole people, in the first person. Sometimes the people or individuals such as Aaron, the priests or the king can be indirect addressees, but Moses is never in that position. In the Temple Scroll there is no place for a narrator who would say what God commanded Moses to do or what Moses did in order to perform a commandment of God. Actually Moses in not named in the Temple Scroll in that manner20. Thus, the text of Lev 23:44 and 24,1, which is a narrative text, cannot be included in the Temple Scroll. That is why frg 23 should be considered as belonging to a different type of text, that is a “biblical text” contained in 4Q36521. Additionaltext ‫אש ̇ר‬ ̇ ‫ לאמור בבואכמה אל הארץ‬add ... ̇ ‫]א[נ֯ וכי נותן לכמה לנחלה וישבתם עליה לבטח תקריבו‬ ‫עש ̇צים לעולה ולכול‬ [‫מל ֯אכ]ת‬ ̇ [...]‫ל]י[ם‬ ֯ ‫הע ֯ג‬ ֯ ‫]ו[את‬ ̇ ‫העולה‬ ̇ ‫]הב[ית אשר תבנו לי בארץ לערוך אותם על מזבח‬ [...]‫יום‬ ̇ ‫[◦ם לפסחים ולשלמים ו̇ לתו֯ ֯דות ולנדבות ולעולות דבר‬ ̇ ] ̇ ‫]ל[תות ולכול מלאכת הבית‬ ֯ ‫ולד‬ ̇ ‫[ל◦] [ל◦] [◦מים‬...] [... ‫יקרי֯ ]בו‬ [... ‫ יקריבו את העצים שנים ]שנים‬.‫ מ[ו֯ עד היצהר‬...] [...]◦ ‫הריש]ו[ן֯ לוי‬ ̇ ‫המקריבים ביום‬ ֯ ‫[◦◦י‬...] [... ‫הרב]יעי‬ ֯ ‫ראו[בן ושמעו̇ ן] וב[י֯ ום‬ ֯ ...] [...]‫[ל‬...] ̇ 4 5 6

20

21

4 5 6 7 8 9 10 11 12

… [add:] saying, when you come to the land which I am giving to you for an inheritance, and you dwell upon it securely, you will gather wood for the burnt offering and for all the wo[r]k of [the h]ouse which you will build for me in the land, to arrange them upon the altar of burnt-offering, [and] the calv[e]s

James VanderKam, Peter Flint, TheMeaningoftheDeadSeaScrolls,TheirSignificance forUnderstandingtheBible,Judaism,Jesus,andChristianity (New York: T&T Clark International, 2002), 211; Alex P. Jassen, ScriptureandLawintheDeadSeaScrolls (New York: Cambridge University Press, 2014), 28. According to Zahn, such a text would have been inspired by Deut 26,1-11. See Molly M. Zahn, RethinkingRewrittenScripture.CompositionandExegesisinthe4QReworked PentateuchManuscripts (STDJ 95, Leiden, Boston: Brill, 2011), 103-104.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

209

7 […] for Passover sacrifices and for whole burnt-offerings and for thanksgiving offerings and for free-will offerings and for burntofferings, daily […] 8 […]and for the d[o]ors and for all the work of the house, [they] will brin[g …] 9 [… fe]stival of New Oil, they will bring wood two […] 10 […] those who bring on the fir[s]t day, Levi […] 11 [… Reu]ben and Simeon [and on the] four[th] day […]22

This additional text is fiercely debated in scholarship. It prescribes the festivals of new oil and wood (cf. Neh 10:35; 13:31: wood; 10:38: oil; Jubilees XXI,12-14: wood; 11QTa XXI,12-XXV,01: oil, wood and order of tribes beginning with Levi). The order of tribes to offer the wood and other sacrifices begins with Levi, as is the case in 11QTa. The precedence of Levi over other tribes is also shared with Jubilees XXXI. The placement of this prescription in Lev 24:2 would be explained as an attempt to specify and complete sacrificial prescriptions after Lev 23:44. Lev 24 begins with the prescriptions on the oil for the lamp. Thus the redactor may have connected the oil for the lamp and the festival of new oil described in this text. This connection corresponds to what is said in the Temple Scroll. Indeed, according to 11QTa XXI,14-XXII, the oil is used not only for sacrifices but it is also offered to be burnt in the lamps. Comparing frg 23 of 4Q365 and the Temple Scroll, I favor the opinion that frg 23 would have served as a source for the Temple Scroll instead of being its summary23. Of course, this additional text is not attested in SP. Thus one would well assert that there is no large-scale textual change in the Samaritan text of Leviticus. However, if we consider that 4Q365 with its frg. 23 is a biblical text, then we should accept that there is at least one large-scale textual change in the legal material of Leviticus (although it is attested by one witness). 2.6. Lev 24:3 (SP) ∥ 11QTa IX,13-14 It is interesting to observe that in Lev 24:3, SP and LXX contain a plus “and his sons,” a text which is dealing only with Aaron in MT.

Lev 24:3 22

23

MT

SP (=LXX)

‫יערך אתו אהרן מערב עד בקר‬

‫יערך אתו אהרן ובניו מערב עד בקר‬

Following the new translation of Sidnie White Crawford, RewritingScripture, 50. An earlier translation with slight differences was given in Emanuel Tov, Sidnie White, “Reworked Pentateuch,” 292-293. Sidnie White Crawford, RewritingScripture, 51; Molly M. Zahn, RethinkingRewritten Scripture, 107-108, n. 74.

210

INNOCENT HIMBAZA

Do SP and LXX want to enlarge the circle of those who are in charge of the lamp? In that case, all the descendants of Aaron, that is including the Levites, would be allowed not only to set up/arrange the lamp but also to burn it. Indeed the verb ‫ערך‬, which means “to arrange,” seems to mean “to burn” in Lev 24:3, since the lamp will be “arranged,” or rather “burnt” ‫( מערב עד בקר‬from evening to morning). With the verb καίω, the LXX shows that it has clearly understood the meaning “to burn.” First, it is to be observed that 11QTa IX:1-14 refers both to the making of the Menorah ‫( מנורה‬lampstand), cf. Exod 25:31-40, and to its setting up, cf. Exod 27:20-21 and Lev 24:2-3. The goal of this precision is to recall that what should be set up/arranged regularly is not any luminary (‫)מאור‬, but the Menorah. Lev 24:4 recalls also that the luminary (‫ )מאור‬is not any kind of lamp, but the Menorah prescribed in Exod 25. However, on the one hand, this verse repeats the same prescription of the preceding verse so that it may be considered a later addition. Baentsch considers it secondary, Milgrom, a redactional supplement, and Marx, redundant.24 It is even interesting to observe that Targum Pseudo-Jonathan does not contain this verse. The absence of this verse in that Targum, one could say, or the addition of this verse in MT and its followers, may reflect a different tradition in the textual transmission history. Milgrom cannot decide whether Lev 24,4 comes from P or a basic H stratum.25 On the other hand, this verse prescribes that (all) the lamps of the Menorah will be “arranged” regularly. Here the SP and the LXX read ‫“ עד בקר‬until morning” instead of ‫( תמיד‬regularly). Thus the SP may have interpreted the verb ‫ ערך‬in the context of a lamp as an action of preparation and lighting. One should understand that the lamp is arranged when it begins to burn. In that case, the word ‫תמיד‬, which can mean “without interruption” or “with regular repetition,” may have been understood in the SP and LXX (διὰ παντός) in the first meaning given here. This may explain why the word ‫ תמיד‬has been replaced by ‫“ עד בקר‬until morning” at the end of the verse: the lamps the Menorah don’t burn during the day.26 24

25 26

Bruno Baentsch, Exodus-Leviticus-Numeri, (HAT 1,2, Göttingen: Vandenhoeck und Ruprecht, 1903), 419; Jacob Milgrom, Leviticus23-27.ANewTranslationwithIntroductionandCommentary (AB 3B, New York, London, Toronto, Auckland: Doubleday, 2000), 2091; Alfred Marx, Lévitique17–27 (CAT IIIb, Genève: Labor et Fides, 2011), 162. For the opposite point, see Christophe Nihan, FromPriestlyTorahtothePentateuch, 511, n. 450. Jacob Milgrom, Leviticus23-27, 2091. For ancient discussions on the number of lamps which burn during the day, see Flavius Josephus, Antiquities 3.199, Mishna, Tamid 3,9; Babylonian Talmud, Shabbat 22b; Menaḥoth 98b.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

211

Second, the same text of 11QTa IX,13-14 specifies that the lamps will be arranged (‫ )ערך‬by the priests, sons of Aaron, while Exod 27:20-21 (MT, SP, LXX) reads Aaron and his sons, and MT of Lev 24:3 contains Aaron alone. The harmonization of Exod 27:21 and Lev 24:3 in the SP and LXX led to the reading “Aaron and his sons” in Leviticus. In that case, this harmonization introduced a new idea in Leviticus that all the priests are in charge of the Menorah. And interestingly, this idea is in agreement with the Temple Scroll (11QTa).

3. DEUTERONOMY 12–26 3.1. Textual comparison Reference Deut 12:1 Deut 13:5 Deut 13:7 Deut 13:19 Deut 15:14 Deut 15:15 Deut 16:6 Deut 16:8 Deut 16:8 Deut 18:19 Deut 19:11 Deut 20:1 Deut 20:8 Deut 20:10 Deut 20:17 Deut 24:3 Deut 26:3 Deut 26:4 Deut 26:18 Deut 26:19

Qumran manuscripts 4QpaleoDeutr 1QDeuta 4QDeutc 4QpaleoDeutr 1QDeutb 1QDeutb 1QDeuta 4QDeutc 4QDeutc 4QTest 4QDeutk2 4QDeutf 4QDeutk2 4QDeutk 4QDeutk 4QDeutk2 4QDeutk2 4QDeutk2 4QDeutf 4QDeutc

Pluses to MT ‫ ו‬of ‫ואלה‬ ‫תדבקון )≠( תשמ[ר‬ ‫בן[ אביך] או‬ ‫ו[הטוב‬ ‫יהוה )≠( א[דני‬ ‫לעש]ות‬ ‫בו‬ ‫ששת )≠( ש[בעת‬ ‫בו כל‬ ‫הנבי‬ ‫מן‬

Pluses contained in SP – – (= 4QDeutc) + + – – – (= 4QDeutc) – ‫כל‬

– – ‫ ו‬of ‫ועם‬ + ‫– השטרים )≠( השופטים‬ ‫ ו‬of ‫וכיא‬ – ‫ ו‬of ‫והחוי‬ + ‫ =( – לקחה אשר[ לו לאשה‬4QDeuta) ‫ל )≠( לפני‬ – ‫את‬ – ‫את‬ – ‫לך‬ –

212

INNOCENT HIMBAZA

3.2. Statistics Qumran manuscript 1QDeuta 1QDeutb 4QDeutc 4QDeutf 4QDeutk 4QDeutk2 4QpaleoDeutr 4QTest Total

Pluses to MT 2 2 4 2 2 5 2 1 20

Pluses contained in SP 0 0 2 1 1 0 1 0 5

3.3. Deut 13:7 ∥ 4QDeutc ∥ 11QTa LIV:19 The two interesting additional texts in Deut 12–26 are found in Deut 13:7 and 13:19.27 Other cases are especially conjunctions. It has been observed that some additional readings are from harmonization or assimilation to parallel verses. The reading ‫ והטוב‬of Deut 13:19 as attested in the SP, 4QpaleoDeutr, LXX (different order, cf. Deut 12:25.28) and 11QTa LV,14 is counted among such cases (cf. Deut 6:18; 12:28), while Deut 13:7 is of a different nature. Carmel McCarthy asserts that the additional material of Deut 13:7, ‫בן‬ ‫( אביך או‬son of your father or), which is also known by the LXX, comes from an exegesis of the scribe/redactor. It is analyzed as possibly reflecting “an incomplete understanding of the internal relationships within a polygamous marriage situation in earlier times.” This reading may thus reflect a textual development influenced by the later exegetical traditions concerning the incest law in Lev 18:9.28 According to Levinson, “the apposition that defines the brother in relation to both the father and the mother is found only in the Second Temple period literature.”29 If these analyses are correct, one should conclude that SP contains not only additional material from harmonization, but also material from later exegesis shared with Qumran manuscripts. These later exegeses, however, are not samaritanisms. 27

28 29

See Carmel McCarthy, ‫ אלה הדברים‬Deuteronomy (Biblia Hebraica Quinta 5, Stuttgart: Deutsche Bibelgesellschaft, 2007), 39-75. See Carmel McCarthy, Deuteronomy, 91*. Bernard M. Levinson, “Textual Criticism, Assyiology, and the History of Interpretation: Deuteronomy 13:7a as a Text Case in Method,” JBL 120 (2001), 211-243.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

213

4. TEMPLE SCROLL (11QTA) LI–LXVI COMPARED TO DEUT 12–26 4.1. Textual comparison 11QTa

Pluses to MT Biblical reference

Pluses contained in SP

LI,12

Deut 16:19 ‫ ו‬of ‫ולא‬ + verbs 3rd pl

+ –

LI,15

‫ובאתה‬

Deut 16:20

–

LIII,4

‫בכה‬

Deut 12:22 (∥15:22)

+ (– in 15:22)

LIII,7

‫עד עולם‬

Deut 12:25

–

LIII,7

‫והטוב‬

Deut 12:25

–

LIII,8

‫אלוהיכה‬

Deut 12:25

–

LIV,6

‫היום‬

Deut 13:1

+

LIV,10

‫ונעבודה‬

Deut 13:3

–

LIV,19

‫בן אביכה או‬

Deut 13:7

+

LV,04

‫עליו‬

Deut 13:9

–

LV,3

‫כול‬

Deut 13:14

–

LV,6

‫כול‬

Deut 13:16

–

LV,14

‫והטוב‬

Deut 13:19

+

LV,17-18

‫או‬

Deut 17:3

–

LVI,07-1

‫או‬

Deut 17:9

–

LVI,5

‫ לשכין שמי עליו‬Deut 17:10

–

LVI,16

‫( סוס‬sg)

Deut 17:16

pl

LVI,17

‫לו‬

Deut 17:16

–

LX,11

‫לפני‬

Deut 18:5

‫לפני יהוה אלוהך‬

LX,11

‫ולברך‬

Deut 18:5

+

LXI,8

‫ולפני‬

Deut 19:17

–

LXI,12

‫עליו‬

Deut 19:21

–

LXII,4

‫פן‬

Deut 20:8

–

LXII,15

‫הגרגשי‬

Deut 20:17

+

LXIII,5

‫ראוש‬

Deut 21:6

–

LXIII,8

‫הטוב‬

Deut 21:9

–

LXIII,8

‫אלוהיכה‬

Deut 21:9

nd

LXIII,12-13 2 sg masc. LXIV,13

‫או את חמורו‬

– rd

Deut 21:12-13: 3 sg fem 3rd sg fem Deut 22:1

‫או את כל בהמתו‬

214

INNOCENT HIMBAZA

11QTa

Pluses to MT Biblical reference

Pluses contained in SP

LXV,12

‫לה‬

Deut 22:17

+

LXV,14

‫ההוא‬

Deut 22:18

+

LXVI,10

‫( ונמצא‬sg)

Deut 22:28

pl

4.2. Statistics 11QTa

Pluses to TM

Plus contained in SP

Total

33

13

4.3. Deut 18:5 (MT and SP) ∥ 11QTa LX,11 I would like to focus on the text concerning the ministry of the whole tribe of Levi. This topic is found in 11QRTa LX which is parallel to Deut 18. It has already been observed that the MT of Deut 18 gives the Levites great consideration, since the whole tribe of Levi was chosen by God to minister in his name. Thus, what is interesting to observe here is how Levites might be integrated into the privileges of the priests, and what is the exact content of their common ministry. The comparison between MT on the one hand, and SP, 11QTa and LXX on the other hand, shows that the prestige that the whole tribe of Levi should enjoy has been enhanced in these last witnesses.30 While MT of Deut 18:5 describes the task of the tribe of Levi as ‫“ לעמד ולשרת בשם יהוה‬to stand and to minister in the name of the Lord,” the common witness of SP, 11QTa LX:11 and LXX contains: (SP) ‫“ לעמד לפני יהוה אלהיכה ולשרתו ולברך בשמו‬to stand before the Lord your God, to minister (for) him and to bless in his name.” According to Deut 18:5-7, and contrary to what is said in Ezek 44:10-16, even Levites will stand and minister before the Lord. The additional text of SP, 11QTa and LXX on blessing in the name of the Lord, which results from a harmonization with Deut 10:8, indicates that Levites can in fact perform the same ritual practices as priests31. 30

31

Harald Samuel, “Levi, the Levites, and the Law”, in Rewriting and Interpreting the HebrewBible.TheBiblicalPatriarchsintheLightoftheDeadSeaScrolls, ed. Devorah Dimand, Reinhard G. Kratz (BZAW 439, Berlin, Boston: De Gruyter, 2013), 215-230. The enhancing of the status of Levites is well known in the books of Chronicles and Nehemiah. See Paul D. Hanson, “1 Chronicles 15-16 and the Chronicler’s View on the Levites,” in “Sha῾areiTalmon”:StudiesintheBible,QumranandtheAncientNearEast PresentedtoShemaryahuTalmon, ed. Michael Fishbane, Emanuel Tov (Winona Lake, Indiana: Eisenbrauns, 1992), 69-77.

LOOKING AT THE SAMARITAN PENTATEUCH FROM QUMRAN

215

It seems that the text of SP has been reformulated since there are stylistic differences in addition to the plus of the verb “to bless.” SP may have avoided repeating the tetragrammaton. Reference

MT

SP

Deut 18:5b

‫לעמד לשרת בשם יהוה‬ ‫הוא ובניו כל הימים‬

‫לעמד לפני יהוה אלהיך‬ ‫ולשרתו ולברך בשמו‬ ‫הוא ובניו כל הימים‬

Thus, on that precise point, the SP (as it is the case for the LXX) enhances the status of the Levites in the same line as some harmonistic texts that Sidnie White Crawford identified as “priestly-levitical/essene line of interpretation.”32 However it is also noteworthy that the Temple Scroll (11QTa LX:7 ∥ XX:10-11) goes a step further. First, it specifies that Levites will receive the shoulder (‫ שכם‬not ‫ )זרע‬as their special sacrificial portion. In that case, the portions of the priests should not include the shoulder but only the foreleg (‫“ האזרוע עד עצם השכם‬the foreleg until the shoulder bone”: 4Q524, Frg 6-13,6; 11QTa XX:16[; LX:05]). Second, the step of the Temple Scroll is also observed elsewhere, as in 11QTa LXI:8 where Levites are explicitly enumerated among those who judge the people, while the source of this statement in Deut 17:9 and 19:17 contains only priests and judges. Contrary to the Temple Scroll, the SP and the LXX do not ascribe the Levites the role of judging the people. Yet, the Temple scroll continues to understand priests and Levites as two distinct classes since they are often enumerated separately and the nature or the quantity of their portions are not always the same (11QTa LVIII:13).

5. CONCLUSION 1. If one considers only the pluses as one of the factors to compare textual witnesses, one observes that even though legal materials are less amenable to change, Leviticus and Deuteronomy 12-26 have changed both in the SP and Qumran manuscripts. 2. 4QRP c (4Q365), Frg 23 is a biblical text. Its nature leads to the conclusion that it cannot be part of the Temple Scroll. Its content reflects a 32

Sidnie White Crawford, RewritingScripture, 146-149.

216

3.

4.

5.

6.

INNOCENT HIMBAZA

short report on the oil and wood festivals, while the Temple Scroll seems to expand it. There is at least one large-scale textual change in the legal material of Leviticus after Lev 24:2. This change is attested only by 4QRPc (4Q365), Frg 23. If the editor of the SP knew this text, he decided to avoid it. However, there is no indication that he may have known it. The plus attested by the SP and 4QLevd in Lev 17:4 is widely analyzed as a Wiederaufnahme rather than an earlier text omitted in MT by homoioteleuton. If this is really the case, this plus can be counted among the textual changes introduced in the legal material of Leviticus. In this case, it was introduced early in the textual history of the book of Leviticus since it is common to SP, LXX and 4QLevd. Harmonizations in the SP as well as in many Qumran manuscripts have introduced new textual material in the legal chapters of Leviticus and Deuteronomy, especially in the Holiness Code and the Deuteronomic Code. It has been observed in this study that through harmonization, the status of the Levites was enhanced in Deuteronomy since they are allowed to perform the same acts as priests. Like other priests, Levites can bless the people. However, the Temple Scroll goes a step further in instructing them to judge the people and allowing them the shoulder (‫ )שכם‬as a special sacrificial portion. When one looks at the SP from Qumran, one observes that when Samaritans adopted a harmonized text, at the same time they also adopted some slight literary differences compared with MT. These differences are shared with other harmonistic texts from Qumran. The clearest example is the enhancement of the status of Levites, which is not a specific ideological product of the SP.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM: NUM 20-21 AND DEUT 2-3 Jonathan BEN-DOV

Recent studies in the history of the biblical text created a new and refined understanding of the textual phenomenon commonly called ‘harmonization.’1 Essentially, ‘harmonization’ means duplication of a certain amount of text in one point based on its occurrence elsewhere. Harmonizations can vary in size and extent, however, as well as in motivation and technique. It now becomes apparent that only some text duplications may be justly called ‘harmonizations’, while many other occurrences of this phenomenon do not harmonize the two conflicting accounts but are rather powered by a different motivation. In addition, while some duplications are recorded in all known variants of the Pentateuch, others are typical of only one or more text-types and may thus be considered a later development. While the pre-Samaritan Pentateuch is often mentioned in this regard, it is by no means the only text-type which employs text duplications. This refined understanding of duplication as an empirical model for the development of biblical texts may now carry us from the confines of lower criticism to the wider field of higher criticism. Literary-historical critics should be aware of these new developments and of their potential use in explaining the formation of biblical texts. In some cases, explanations from the discipline of lower criticism may supersede more radical 1

For the classical definition of ‘harmonization’ see Emanuel Tov, “The Nature and Background of Harmonizations in Biblical Manuscripts,” JSOT31 (1985): 3-29; Ron Hendel, TheTextofGenesis1-11.TextualStudiesandCriticalEdition (New York and Oxford: Oxford University Press, 1998), 81-85. For recent advancements see Michael Segal, “The Text of the Hebrew Bible in Light of the Dead Sea Scrolls,” MateriaGiudaica 12 (2007): 5-20; Molly M. Zahn, RethinkingRewrittenScripture.CompositionandExegesisinthe 4QReworkedPentateuchManuscripts (STDJ 95; Leiden: Brill, 2011), 143-156; Jonathan Ben-Dov, “Early Texts of the Torah: Revisiting the Greek Scholarly Context,” JAJ 4 (2013): 210-234; Emanuel Tov, Textual Criticism of the Hebrew Bible3 (Minneapolis: Fortress, 2012), 80-82; Magnar Kartveit, The Origin of the Samaritans (VTSup 128; Leiden: Brill, 2009), 265-288.

218

JONATHAN BEN-DOV

historico-literary reconstructions. My motivation in this article thus in many ways overlaps that of David Carr, who recently evaluated the phenomenon of harmonization as standing between the formation and transmission stages of biblical literature. Carr indicates that this mechanism of textual expansion “would represent a fairly mechanical, non-theological form of harmonization.”2 Other recent writers would refer to this technique as ‘scribal.’3 I wish to define categories of text-duplication and trace their circulation in various text-types (part 1). Part 2 will illuminate the textual dynamics of the reports on Israel’s sojourns east of the Jordan. In part 3, I shall connect the two earlier parts, applying the finds of part 1 to the specific casestudy of part 2. In part 4, I hope to suggest an improved explanation for the formation of Num 20-21 and Deut 2-3.

PART 1: CATEGORIES OF TEXT-DUPLICATION Many small-scale harmonizations appear throughout the Old Greek text of the Pentateuch. A typical example is Gen 8:12, where ‫ וישלח‬of MT and SP becomes ‫ ויסף שלח‬as retroverted from G, matching the same phrase in 8:10 above.4 Examples of a larger scale appear in the text of Genesis 1. In this chapter, as within priestly literature in general, there is a tendency to correlate commandments with their execution. Some of these correlations are recorded in all textual versions of Genesis 1, as for example the full correspondence of the act of creation in v. 12 to the command in v. 11. In verse 9 MT, however, the long commandment of 9a is matched only by the laconic statement ‫ ויהי כן‬at the end of the verse. This statement did not satisfy a certain later reader, who inserted a full report on the execution of God’s intention at the end of v. 9. This report is preserved in the LXX:5 2

3

4 5

David Carr, The Formation of the Hebrew Bible. A New Reconstruction (Oxford and New York: Oxford University Press, 2011), 260. Notable examples are Karel van der Toorn, ScribalCultureandtheMakingoftheHebrew Bible (Cambridge MA; Harvard University Press, 2007); D. Andrew Teeter, Scribal Laws. Exegetical Variation in the Textual Transmission of Biblical Law in the Late SecondTemplePeriod (FAT 92; Tübingen Mohr Siebeck, 2014). Hendel,Genesis1-11, 91. For the harmonizations in Gen 1 see already Julius Wellhausen, DieCompositiondes HexateuchesundderHistorischenBücherdesAltenTestaments3 (Berlin: Reimer, 1899), 184; Tov, “The Nature,” 9-10; Hendel, Genesis1-11, 20-23.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

219

Καὶ συνήχθη τὸ ὕδωρ τὸ ὑποκάτω τοῦ οὐρανοῦ εἰς τὰς συναγωγὰς αὐτῶν, καὶ ὤφθη ἡ ξηρά. 6

‫ויקוו המים מתחת השמיים אל מקויהם ותראה היבשה‬

In contrast to the small harmonization of Gen 8:12 described above, the motivation for this one is better defined by a more careful reading of the text. The scrupulous scanning of the text, seeking for commands that require fulfillment, is motivated by a refined awareness to the dynamics of the divine word. As acknowledged already by Gesenius, this type of textduplications is not vulgar or popular, as is often suggested; to the contrary, it seems ‘academic’ in nature, driven by a systematic study of the text and by clear guidelines.7 These changes do not derive from a ‘free’ attitude to the transmitted text, but rather from a highly controlled and deliberate set of practices regulating the inserted changes. Tracing the different kinds of text-duplication is therefore not only an act of philology, but also, one would even say predominantly, an act of theology. If we succeed in tracing the exact modes of correction applied by authors or editors, we may also be able to unlock their theological presuppositions. This ‘academic’ type of variants is clearly apparent in one well-defined group of textual duplications, which involves the verification of speech-acts in the Pentateuchal narratives.8 For example, in the account of the spies in Numbers 13-14, the pre-Samarian and SP text add excerpts from the similar but not identical story in Deut 1:20-30. The result is an awkward story, speaking as it were in two voices. For example, did God command Moses to send the spies of was it the people’s initiative? There are many more examples for such duplications throughout the Pentateuch, concentrated in several clusters. While this group of duplications is often called ‘harmonization,’ this appellation is apparently insufficient; when the two stories are juxtaposed, the result is not harmony but rather contradiction. Another reason must have existed for these duplications other than the need to harmonize. A new term for their definition should accordingly be sought. These large-scale duplications never occur in biblical law, a classical locus for the harmonization of contradictions, but rather only in narratives. 6 7

8

Retroversion of Hebrew is based on BHS. Guilielmus Gesenius, Depentateuchisamaritaniorigineindoleetauctoritate.Commentatio philologico-critica (Halle, 1815), 61-64.In contrast, most other scholars follow Paul Kahle in appreciating the text-duplications as popular or even vulgar. See in detail Kartveit, The Origin, 261-263. For my preference of Gesenius’ model see Ben-Dov, “Early Texts”. For the definition of this group see the studies by Segal, Zahn, Ben-Dov, Tov, quoted above (n. 1). Kartveit, TheOrigin, 276-288, diagnosed the find in SP similarly, but his explanation for it differs.

‫‪220‬‬

‫‪JONATHAN BEN-DOV‬‬

‫‪In addition, they are carried out by copying the exact wording of the pro‬‬‫‪tagonist in the Torah and pasting it in places where it seems to be required.‬‬ ‫‪Only minimal changes meant to monitor the grammatical person of the‬‬ ‫‪speaker are inserted.9‬‬ ‫‪The reason for these duplications is the wish “to increase the consistency‬‬ ‫‪of speech events,” or “increasing the self-referentiality of the Torah.”10‬‬ ‫‪In other words, they were not driven by the wish to compromise contradict‬‬‫‪ing narratives, but rather by a formal interest of rendering full credibility‬‬ ‫‪to speech acts. Take for example the duplication of Deuteronomy 5:24-27‬‬ ‫‪within the Sinai account of Exodus 20:18-19:‬‬ ‫‪MT Exodus 20:18-19‬‬

‫‪MT Deuteronomy‬‬ ‫‪5:24-27‬‬

‫‪ ...‬וַ ַיּ֤ ְ רא ָה ָע ֙ם וַ יָּ ֻ֔נעוּ‬ ‫וַ ַיּ ַֽע ְמ ֖דוּ ֵ ֽמ ָר ֽחֹק׃‬ ‫רוּ ֶאל־מ ֶֹ֔שׁה‬ ‫אמ ֙‬ ‫‪ 19‬וַ ֽיּ ֹ ְ‬

‫ר־א ָ ֥תּה ִע ָ ֖מּנוּ וְ נִ ְשׁ ָ ֑מ ָעה‬ ‫ַדּ ֵבּ ַ‬ ‫ֹלהים‬ ‫וְ ַאל־יְ ַד ֵ ֥בּר ִע ָ ֛מּנוּ ֱא ִ ֖‬ ‫ֶפּן־נָ ֽמוּת׃‬

‫‪SP Exodus 20:16‬‬ ‫ויראו כל העם וינעו‬ ‫ויעמדו מרחק׃‬

‫אמ ֗רוּ‬ ‫וַ תּ ֹ ְ‬

‫‪ Exod 20:16‬ויאמרו אל‬ ‫משה‬

‫ינוּ‬ ‫ֹלה ֙‬ ‫הו֤ה ֱא ֵ֨‬ ‫ֵ ֣הן ֶה ְר ָ֜אנוּ יְ ָ‬ ‫ת־כּב ֹ֣דוֹ וְ ֶאת־גָּ ְד ֔לוֹ וְ ֶאת־ק ֹ֥לוֹ‬ ‫ֶא ְ‬ ‫ָשׁ ַ ֖מ ְענוּ ִמ ֣תּוְֹך ָה ֵ ֑אשׁ‬ ‫ַהיּ֤ וֹם ַהזֶּ ֙ה ָר ִ֔אינוּ ִ ֽכּי־יְ ַד ֵ ֧בּר‬ ‫ת־ה ָא ָ ֖דם וָ ָ ֽחי׃‬ ‫ֹלהים ֶא ָ ֽ‬ ‫ֱא ִ ֛‬ ‫וְ ַע ָתּ ֙ה ָל ָ֣מּה נָ ֔מוּת ִ ֣כּי‬ ‫אכ ֔ ֵלנוּ ָה ֵ ֥אשׁ ַהגְּ ד ָֹל֖ה‬ ‫ֽת ֹ ְ‬ ‫ַה ֑זּ ֹאת ִאם־י ְֹס ִ ֣פים ׀ ֲא ַ֗נ ְחנוּ‬ ‫֠ ִל ְשׁמ ַֹע ֶא ֨‬ ‫הו֧ה‬ ‫ת־קוֹל יְ ָ‬ ‫ֹלהינוּ ֖עוֹד וָ ָ ֽמ ְתנוּ׃‬ ‫ֱא ֵ ֛‬

‫הן הראנו יהוה אלהינו‬ ‫את כבודו ואת גדלו ואת‬ ‫קולו שמענו מתוך האש‬ ‫היום הזה ראינו כי ידבר‬ ‫אלהים את האדם וחי‬ ‫ועתה למה נמות כי‬ ‫תאכלנו האש הגדלה‬ ‫הזאת אם יספים אנחנו‬ ‫לשמע את קול יהוה‬ ‫אלהינו עוד ומתנו‬

‫שׁר‬ ‫ל־בּ ָ֡שׂר ֲא ֶ ֣‬ ‫ִ ֣כּי ִ ֣מי ָכ ָ‬ ‫ֹלהים ַח ִ֜יּים‬ ‫ָשׁ ַ ֣מע קוֹ ֩ל ֱא ִ֨‬ ‫מנוּ‬ ‫תּוְֹך־ה ֵ ֛אשׁ ָכּ ֖ ֹ‬ ‫ָ‬ ‫ְמ ַד ֵ ֧בּר ִמ‬ ‫וַ ֶיּ ִֽחי׃ ְק ַ ֤רב ַא ָתּ ֙ה וּֽ ֲשׁ ָ֔מע‬ ‫הו֣ה‬ ‫אמר יְ ָ‬ ‫ל־א ֶ ֥שׁר י ֹ ַ ֖‬ ‫ֵ ֛את ָכּ ֲ‬ ‫ֹלהינוּ וְ ַ ֣א ְתּ ׀ ְתּ ַד ֵבּ֣ר‬ ‫ֱא ֵ ֑‬ ‫ל־א ֶ֨שׁר יְ ַד ֵ֜בּר‬ ‫ֵא ֗ ֵלינוּ ֵא ֩ת ָכּ ֲ‬ ‫ֹלהינוּ ֵא ֶל֖יָך‬ ‫הו֧ה ֱא ֵ ֛‬ ‫יְ ָ‬ ‫וְ ָשׁ ַ ֥מ ְענוּ וְ ָע ִ ֽשׂינוּ׃‬

‫כי מי כל בשר אשר שמע‬ ‫קול אלהים חיים מדבר‬ ‫מתוך האש כמונו ויחי‬ ‫קרב אתה ושמע את כל‬ ‫אשר יאמר יהוה אלהינו‬ ‫ואתה תדבר אלינו את‬ ‫כל אשר ידבר יהוה‬ ‫אלהינו אליך ושמענו‬ ‫ועשינו‬ ‫ואל ידבר עמנו האלהים‬ ‫פן נמות׃‬

‫‪These characteristics are developed by Segal, “The Text of the Hebrew Bible,” 12-17.‬‬ ‫‪Zahn, RethinkingRewrittenScripture, 57, 173.‬‬

‫‪9‬‬ ‫‪10‬‬

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

221

Once Deuteronomy reports that the people said certain words at Sinai, that report must be upheld. The absence of these words from the parallel account in Exodus constitutes a flaw in Moses’ speech, and a duplication of text from Deuteronomy to Exodus is required in order to mend it.11 Alternatively, one may see these duplications with Segal as an attempt to provide a “source” for Moses’ later speech in Deuteronomy.12 In this and similar cases, duplications appear where the fulfillment of direct speech is involved. While most duplications render credibility to the direct speech of Moses – as do by definition all the cases based on Deuteronomy – some duplications verify the speech of other figures. Thus for example, Jacob’s speech in Genesis 30:36ff SP is copied (selectively) from Gen 31:11-13 in order to provide a source for the latter; the speech of the Israelites in Exodus 6:9f SP is copied from the speech reported later in Exod 14:12. The efforts to perfect the direct divine speech are particularly concentrated in central locations within the Pentateuch where a series of divine statements may be found: the Sinai theophany, the narrative of the Ten Plagues, the parallels to Deuteronomy 1–3, and the divine creation statements in Genesis 1.13 While the former three are represented in SP, the latter appears only in G. This explanation for the presence of duplicated verses has wider consequences for biblical criticism. It is now apparent that many insertions in early texts of the Pentateuch were added to the original stories not in order to design the narrative’s message, as literary critics would read them, but rather in order to fulfill the extraneous, formalistic need of perfecting 11

12 13

On the theological and canonical implications of copying a text from Deuteronomy (Moses’ direct speech) to the Exodus (a priestly-redacted composition representing the word of God), see Benjamin Ziemer, “Prophetenrede und Gottesrede im Pentateuch und der Ausgang der Schriftprophetie,” in: R. Lux and E.-J. Waschke (eds.), Die unwiderstehliche Wahrheit. Studien zur alttestamentlichen Prophetie. Festschrift fürArndtMeinhold (Leipzig: Evangelische Verlagsanstalt, 2006), 441–466. According to Ziemer, these duplications in the pre-SP are an attempt to equate the authority of Moses’ speech (Deuteronomy) with what is conceived as divine speech in other parts of the Torah. Segal, “The Text,” 17. It seems that duplications occur wherever there are typologicalnumbers of divine pronouncements, with the Ten Commandments and Ten Plagues as obvious examples. In addition, there is a sevenfold occurrence of the speech formula ‘at that time YHWH said to me’ in Deut 1-3 (1:42; 2:2, 9, 17, 31; 3:2, 26). In Genesis 1 there are of course seven days, but in addition later traditions know ‘10 statements by which the world was created’ reflecting ten divine speeches in this chapter (see m. Avot 5:1). There are also major changes in the tabernacle chapters of Exodus 26-30 (Kartveit, TheOrigin, 279), but these changes are not text duplications; rather, they involve mainly transposition and some minor changes in wording.

222

JONATHAN BEN-DOV

the speech act. These verses are only obliquely relevant for their environment. They are thus not part of an ‘editorial layer’ or a distinct literary source in the full sense of the word, but rather a technical supplement to it. While textual critics are well aware of these additions, they are often not taken into account by literary critics. I aim to show that in some cases this type of duplications offers tools for explaining literary difficulties, and may even be more useful than the usual literary-critical methods.

PART 2: THE CIRCULATION OF VARIOUS TYPES OF DUPLICATIONS The text duplications specifically concerned with the veracity of speech acts differ from other, more general, types of harmonization. ‘General’ harmonizations are typical of all text-types of the Torah, appearing in all text witnesses or in one or more of them. In the former cases, harmonizations were part of the formation of the Pentateuch. In the later cases, they were added in one or more of the versions. Such harmonizations may be found in all major versions: MT, SP, and G. In contrast, the special type of speech-act duplications is typical – almost exclusive – of pre-SP and SP. With this distinction I mean to address Zahn’s recent objection that the duplications in (pre-)SP are not uniform and should not be simplistically considered as occurring in one stage of the formation of pre-SP.14 Indeed, I agree that we need to define various kinds of duplications which all appear in SP. While the simpler sort is typical of all Pentateuchal versions at all stages of formation and transmission, the latter kind of speechact duplications is the fruit of a well-defined wave of subsequent scribal activity. Harmonizations of the first kind existed already before the parting of the stemma into various textual branches. Thus, P in general made it a case to supply a record of execution for every commandment.15 It may even be claimed that the very composition of Exodus 35–39, tediously reporting the execution of the commandments in chapters 25–31, is itself an act of duplication that took part after the composition of chapters 25–31.16 This 14

15

16

Molly Zahn, “The Samaritan Pentateuch and the Scribal Culture of Second Temple Judaism,” JSJ 46 (2015): 285-313. Sean McEvenue, “Word and Fulfillment: A Stylistic Feature of the Priestly Writer,” Semitics1 (1970): 104–10, 104-06; Itamar Kislev, “The Investiture of Joshua (Numbers 27:12–23) and the Dispute on the Form of the Leadership in Yehud” VT59 (2009): 429–45, esp. n. 13 and literature quoted there. The question of the motivation to compose chapters 35-39 is distinct from a second question, regarding the original text of these chapters, given the large changes of order

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

223

duplication is extant in all versions of the Torah, but it is not a strictly verbatim repetition as was later done in SP, since the later chapters 35–39 contain several linguistic and even literary differences vis-à-vis the older chapters 25–31.17 This act of duplication thus came relatively late in the composition of the Pentateuch, but it came relatively early in the history of its transmission. Against the common opinion, MT does include duplications, not only of the generally harmonizing kind, but also of the more specific, speech-act type of duplications.18 A good example for this second type of duplications is Exodus 32:9, recounting the narrative of the Golden Calf in parallel to a version of that story in Deut 9-10. ‫ם־ק ֵשׁה־‬ ְ ‫ת־ה ָע֣ם ַה ֶ֔זּה וְ ִה ֵנּ֥ה ַע‬ ָ ‫יתי ֶא‬ ֙ ִ ‫הו֖ה ֶאל־מ ֶ ֹ֑שׁה ָר ִ֨א‬ ָ ְ‫אמר י‬ ֶ ֹ ‫ וַ ֥יּ‬Exod 32:9 MT, SP ‫֖עֹ ֶרף ֽהוּא׃‬

Quite a few considerations render v. 9 problematic in its context. While the divine speech to Moses about the sin of the Golden Calf continues smoothly from 32:8 to 32:10, verse 9 introduces yet another opening formula which interrupts the order. In addition, verse 9 refers to the Israelites as ‘a people’ in the singular, as opposed to the plural references in vv. 8 and 10.19 Importantly for our purposes, this verse is missing from G,

17

18

19

and content between MT and LXX in them. This second question raises a whole set of text-critical issues (see Anneli Aejmelaeus, OntheTrailoftheSeptuagintTranslators. Collected Essays [CBET 50; Leuven: Peeters, 2007], 107-121) which seem to me to obfuscate the force of the first question: why were these chapters ever composed. Abraham Kuenen, An Historical-Critical Inquiry into the Origin and Composition of theHexateuch(trans. P.H. Wicksteed; London: Macmillan, 1886), 73-80: “very short original account of the execution of the commands of Ex. XXV sqq. was gradually elaborated till at last it was brought almost into the form of the instructions themselves”. Kuenen followed a preliminary idea by Julius Popper, DerBiblischeBericht über die Stiftshütte: Ein Beitrag zur Geschichte der Composition und Diaskeue des Pentateuch (Leipzig: Hunger,1862). The duplications in Exod 35-39 are not verbatim as in SP but rather incorporate differences: both linguistic (Late Biblical Hebrew reflected in Exod 36:10-22, see Kuenen, Historico-CriticalInquiry, 79) and theological-priestly (Brevard S. Childs, Exodus:ACommentary [OTL; London: SCM, 1974], 533-537). For a list of ‘general’ harmonizations in MT see Hendel, Genesis1-11, 86. For speech-act harmonizations in MT see Alexander Rofé, “Historico-Literary Aspects of the Qumran Biblical Scrolls,” in: TheDeadSeaScrollsFiftyYearsAftertheirDiscovery.ProceedingsoftheJerusalemConferenceJuly1997 (ed. L.H. Schiffman et al; Jerusalem: Israel Exploration Society and Israel Museum, 2000), 30-39. Another difficulty is that the term “stiff-necked” in Exod 32:9 denotes disobedience to authority or negative attitude, but does not fit the religious act of the golden calf as described here. See Gili Kugler, “The Threat of Annihilation of Israel in the Desert: An Independent Tradition within Two Stories,” CBQ 88 (2016): 642.

224

JONATHAN BEN-DOV

being represented only in MT and SP. The finds of literary criticism thus correspond to those of textual criticism. Finally, verse 9 duplicates the direct speech from the parallel account in Deut 9:13: ‫ם־ק ֵשׁה־ ֖עֹ ֶרף ֽהוּא׃‬ ְ ‫ת־ה ָע֣ם ַה ֶ֔זּה וְ ִה ֵנּ֥ה ַע‬ ָ ‫יתי ֶא‬ ֙ ִ ‫מר ָר ִ֨א‬ ֹ ֑ ‫הו֖ה ֵא ַל֣י ֵלא‬ ָ ְ‫אמר י‬ ֶ ֹ ‫וַ ֥יּ‬

These considerations lead to the conclusion that Exod 32:9 duplicates Deut 9:13, in the wake of the pre-SP duplications.20 It is a duplication meant to give credence to a divine speech act: while Deut 9:13 reports that God uttered this statement to Moses, the interpolator (or should we say duplicator) was concerned that it is not reported in Exodus, and thus inserted it there. Hermeneutically speaking, the presence of verse 9 in Exod 32:7-14 was not meant to alter to the message of the story, but rather to fulfill a formalistic need. Interestingly, in this case the duplication took place in both MT and SP, but does not appear in G.21 We may now draw the following scenario. Already during the composition of the Pentateuch and during all subsequent stages of its transmission, there was a rather wide trend of harmonizations, present in all or part of the text types. They seem rather frequent in the LXX of Genesis, for example.22 A more specific sort of duplications came at a rather defined 20

21

22

Curiously, only few commentators share this idea, while others do not see a problem in the smoothness of vv. 7-14. Childs, Exodus, 567, sees the entire unit 7-14 as a deuteronomistic expansion, with v. 9 naturally playing a central part in this chracterization. However, the rest of the unit is not compellingly deuteronomistic, and the presence of v. 9 can be explained differently, as suggested here. I thus see little justification for Childs’ view. William Wevers (NotesontheGreekTextofExodus [SBLSCS 30; Atlanta, Scholars Press, 1990], 30) misses the point when claiming that the Greek translator of Exodus omitted 32:9 due to theological and literary considerations; theological corrections of this scale are very rare, if at all, and can always be explained on different grounds. Christine Hayes (“The Golden Calf Stories: The Relationship of Exodus 32 to Deuteronomy 9-10,” in TheIdeaofBiblicalinterpretation.EssaysinHonorofJamesL.Kugel [JSJSup 83; ed. H. Najman and J. H. Newman; Leiden: Brill, 2004, 45-93) made notable efforts to defend the unity of vv. 7-14 (as well as of Exod 32 in general). According to her (pp. 54-56), 32:7-14 contains two divine speeches in order to accentuate the drama, a rare but not unattested phenomenon in biblical narrative. However, she does not refer to the other difficulties raised specifically in v. 9, as indicated above. I agree with Hayes that this unit is uniform in the sense that none of the elements contained in it promotes a different message. Even verse 9, which I see as interrupting the unit, was not meant to change the message of the unit but rather to fulfill a formalistic need of a later reader. Kartveit, TheOrigin, 276, notes Exod 32:9 as a harmonization without much further reasoning. See the figures for Genesis 1-11 in Hendel, Genesis1-11, 84. Hendel quotes Z. Frankel (1841) who observed that LXX of Genesis contains 270-280 harmonizations. For further analysis with ample bibliography see Tov, TextualCriticism,136 and n. 228.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

225

stage in an attempt to validate speech-acts in the Torah. This stage, active mainly – but not only – in such focuses as Creation, Sinai, and Moses’ speech, appears almost exclusively in pre-SP.

PART 3: THE DUPLICATIONS IN NUM 20:14-21, 21: 21-35 || DEUT 2:1 - 3:3 I will now examine the relations between the episodes on Israel’s encounters with the nations east of the Jordan, according to Numbers 20– 21 and Deuteronomy 2–3, in both MT and SP. Attention to the details enables an excavation through layers of duplications, while determining their relative order and pinpointing the interest of various interpolators. This issue being a classical locus for historical-literary criticism, I believe that an examination in text-critical tools may produce better results, superseding the finer historico-critical observations. In other words, many of the layers accumulated in the above-noted chapters were not meant to alter the message of the original accounts by means of accretion of new material, but rather fulfilled a technical-formalistic, scribal interest. Numbers At the core of the discussion stand the narratives in Num 20:14-21 and 21:21-24, 33-35. I shall describe the content of these units using a bird’seye view, resisting the temptation to treat the multiple side issues that were related to them in past research. After long wanderings in the desert, in 20:14 the Israelites reach the (southwestern) edge of Edom, somewhere in the south of the Arava valley.23 By the end of chapter 21 and after long additional wandering, they are ready to enter the land of Canaan from the east, at the plains of Moab, far north of Edom (22:1).24 Since there are several nations and territories east of the Jordan between Edom and Moab, the narrator should explain how the Israelites passed through or around these territories from south 23

24

The city of Kadesh, mentioned in 20:16 “a city at the edge of your territory,” is probably not Kadesh Barnea but rather a different site, located far to its south-east. For a recent re-affirmation of this opinion and a summary of literature see D. Ben-Gad Hacohen, “Kadesh: The Hidden Toponym,” Shnaton 15 (2005): 3-19 (Hebrew). This verse (22:1) belongs with the previous section rather than with the Balak narrative. See George B. Gray, ACriticalandExegeticalCommentaryonNumbers (ICC; New York: Scribner, 1903), 306-307. Gray sees this verse as continuing the itinerary in 21:11.

226

JONATHAN BEN-DOV

to north. Of these nations and territories, Numbers 20–21 recounts only the encounters with Edom and with Sihon (‘king of the Emorites’), as well as with Og of the Bashan. Curiously, this last territory lies far to the north of the Plains of Moab. The main line of narrative in Numbers traces the path of the Israelites around Edom and through Sihon’s land. Israel approaches Edom, asks for safe passage, is refused, and is forced to bypass Edom, probably by circumventing its eastern border (21:4). The narrative continues from 20:14-21 (Edom) to the Sihon pericope in 21:21-24. Approaching Sihon, an account rather similar to Edom appears in similar wording, but this time the Israelites choose to fight Sihon in response to his aggressiveness, conquering his entire land (21:25-32). The material between Edom and Sihon (20:22–21:20) is a collection of miscellaneous fragments recording episodes from the itinerary of encircling Edom and approaching Moab.25 There is no clear narrative of an encounter with the neighboring nations of Moab and Ammon, about which only a handful of episodes is shortly reported in 21:10-20. Both Edom and Sihon are approached in Num 20-21 by means of envoys requesting free passage, with or without a request to supply food and water to the wandering camp. In neither of the narratives is the envoy sent by divine decree, nor are the Israelites commanded to make war with the nations. Rather, the encounter and its results are depicted as human decisions. Every such envoy event is marked with an ellipse in the table below. The core Sihon pericope (21:21-24) is uniform from a literary-critical point of view. The Edom pericope (20:14-21) is slightly more problematic, since the Israelites address Edom and are answered twice, first in 20:1418 and again in 19-20. Although this duplicity is sometimes taken as a sign of multiple layers,26 it may also be read as a reasonable complication 25

26

For the reason for inserting those episodes see Jacob Milgrom, Numbers (JPS Torah Commentary; Philadelphia, 1990), 318, 564. The direct continuity between the episodes on Edom and Sihon is manifest by the literary design of similarity and contrast between the two. According to W.A. Sumner, “Israel’s Encounter with Edom, Moab, Sihon, and Og According to the Deuteronomist,” VT 18 (1968): 216-228, esp. 225, the direct move from Edom to Sihon contrasts Israel’s weakness in the former case with their power in the latter. Gray, Numbers, 264-267 considers this duplicity as a sign for multiple authorship, identifying the sources J and E respectively. Similarly Martin Noth, Numbers:ACommentary (trans J. D. Martin; OTL; London: SCM, 1968), 149, explained it by recourse to J and E although not distinguishing them in detail. Siegfried Mittmann, “Num 20,1421 – eine redaktionelle Kompilation,” in Wort und Geschichte. Festschrift für Karl Elliger zum 70. Geburtstag (ed. H. Gese and H.P. Rüger; Kevelaer: Butzon and Bercker, 1973), 143-149, sees vv. 18-20 as late additions.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

227

of the plot, with no clear signs for source-critical distinction.27 Nothing in these passages betrays deuteronomistic or priestly hands,despite recent attempts to depict it as post-dtr or post priestly.28 Following the Sihon narrative and more material on the conquest of his land, the account moves to describe the Israelite occupation of the land of Og, king of Bashan, in 21:33-35. This account uses entirely different vocabulary and style than those used for the previous Edom and Sihon narratives: no envoys are sent, being replaced by a direct divine instruction to make war with Og (v. 34). In fact the language almost exactly duplicates the narrative on Og in Deut 3:1-3, suggesting that it was duplicated from there, as will be argued in more detail below. The Og narrative is incommensurate with the context in Numbers 20–21 not only in terms of style but also with regard to the geographical sequence. The Israelite expedition east of the Jordan should end at the plains of Moab, mid-way on the north-south axis of Canaan, as reported immediately below in Num 22:1. They will camp in the Plains of Moab for the rest of the Pentateuch, and will depart from there to cross the Jordan. If, as 21:3335 reports, they fought their way as far as Mount Hermon, how and why did they then return southwards to the Plains of Moab?29 The Og narrative is thus foreign to the message of Num 21. It is marked by a rectangle rather than an ellipse in the table below, in contrast to the previous narratives. 27

28

29

Explanations for the duplicity without recourse to multiple authorship of the Edom pericope were given by Milgrom,Numbers, 168-169; Jacob Licht, CommentaryonNumbers (Jerusalem: Magnes, 1985-1995), II, 211 (Hebrew). Scholars viewing the entire unit as stemming from the E source are: Sumner, “Israel’s Encounters with Edom,” and with further argumentation Joel Baden, J,E,andtheRedactionofthePentateuch (FAT 68: Tübingen: Mohr Siebeck, 2009), 130-132. In contrast, Menahem Haran, The Biblical Collection (Jerusalem: Magnes and Bialik, 2003), II, 198 (Hebrew) assigned all of it to J. More recently, Ludwig Schmidt, “Sihon und Og in Num 21,2ff.* und Dtn 2,24ff.* Ein Beitrag zur Enstehung des Buches Numeri,” in TorahandtheBookofNumbers (ed. C. Frevel et al.; Tübingen: Mohr Siebeck, 2013), 314-333, considers Num 20:14-21 to be a uniform and early textual unit, whle emphasizing that it is an elaboration of the even earlier Sihon narrative in Num 21:21-24. John van Seters, “The Conquest of Sihon’s Kingdom: A Literary Examination,” JBL 91 (1972): 182-197; Reinhard Achenbach, Die Vollendung der Tora. Studien zur RedaktionsgeschichtedesNumeribuchesinKontextvonPentateuchundHexateuch (BZAR 3; Wiesbaden: Harrassowitz, 2003), 341, 358; Eckart Otto, DasDeuteronomiumimPentateuchundHexateuch (FAT 30; Tübingen: Mohr Siebeck, 2000), 132; Wolfgang Oswald, “Die Revision des Edombildes in Numeri XX 14-21,” VT 50 (2000): 218-232. Rofé, “Historico-Literary Aspects,” 32, points out another discrepancy in the Og narrative. This report contradicts the report in Num 32:39, 41, 42 according to which the Bashan was not taken in a unified national campaign but rather as a local family campaign by Yair son of Menashe.

228

JONATHAN BEN-DOV

A short note is due here about the similar report of Israelite envoys east of the Jordan in Judges 11:16-22. This short report closely resembles those of Num 20 and 21, possibly relying on a shared source.30 This article focusing on Numbers and Deuteronomy only, it will relate to Judges 11 only in the following points: 1. The report in Judges mentions envoys sent to all nations mentioned in the narrative, i.e., ellipses rather than squares on the table below. 2. In both Judg 11 and Num 20–21 Ammon was not approached by the Israelites as part of the above itinerary. 3. Judg 11:17b mentions in passing that Moab was also approached alongside Edom and Sihon, a detail that is not mentioned in Num 20-21 but may be hinted in Deut 2:29.31 Summing up, Numbers 20–21 offers two similar and distinctly styled narratives about the sending of envoys to Edom and Sihon. The account of the encounter with Og, in contrast, is highly problematic in the context of Num 20–21. Deuteronomy Israel’s encounters east of the Jordan are lavishly expanded in Deut 2–3 as part of Moses’ final speech. Chapter 2 recounts the encounters with the following nations: Edom (called here: Seir, 2:2-8a), Moab (2:8b-13), Ammon (2:17-23), Sihon (2:24-37), followed by a report on Og in 3:17. The classical theory in historical criticism holds that the accounts in Exodus–Numbers served as sources for reworked reports in Deuteronomy 1–3.32 The author of these chapters, it is claimed, knew the earlier sources but refashioned them into a new construction, which is stamped by new Dtr ideology. 30 31

32

See Licht, Numbers, II, 209-210. This detail can be explained in one of two ways. It might have been added as a gloss in Judg 11:17b in order to accommodate for Deut 2:29. Alternatively, Licht (Numbers, II, 209) speculates – attractively in my opinion – that such a narrative of approaching Moab had originally existed in numbers 20 as in Judges 11, but that it was later abolished from the Book of Numbers in order to make way for the Moab-Balak-Bilam block in chapters 22-24. The absence of envoys to Ammon from Judges 11 runs counter to the idea by Sumner, “Israel’s Encounters,” that JE had originally known of a letter to Ammon as well. Weinfeld, Deuteronomy1–11, 130; David A. Glatt-Gilad, “The Re-Interpretation of the Edomite-Israelite Encounter in Deuteronomy II,” VT 47 (1997):441-455; This classical opinion was updated and presented in a comprehensive manner by Marc Z. Brettler, TheCreationofHistoryinAncientIsrael (London and New York: Routledge, 1995), 71-76.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

229

The Deuteronomy account addresses a full list of all five nations east of the Jordan, from the very south to the very north, in contrast to the three nations in Num 20–21 (in fact only two if the report of Og in Num 21:33-35 is dismissed as secondary). Deuteronomy conceives the large section 2:2–3:22 as a full report on the conquest, apportioning and settlement of the East Jordan, with such programmatic statements as 3:8, 21-22 attesting to this aim. Although Moab – and certainly Ammon – were not part of the older account preserved in Numbers and Judges, the Deuteronomy account does relate to them, rendering them part of a comprehensive account of the East Jordan.33 The fate of each nation is divinely decreed in Deuteronomy and instructed to Moses prior to the encounter with it, hence no envoys or negotiation are required. In the table below, the motif of divine command to make or avoid war is marked by a rectangular frame which can be seen all through the Deuteronomy column, as opposed to the ellipses from the Numbers account.

PART 4: NUM 20–21 AND DEUT 2–3 BETWEEN LOWER HIGHER CRITICISM

AND

The table summarizes the find. The sign → in the table indicates that the content of the left hand column is represented also in the column to its right. For example, SP of Numbers 20–21 includes the material from MT (marked →) of those chapters, in addition to several plusses, noted in the respective column. This synoptic presentation divides rather neatly, with ellipses in MT Numbers and rectangles in MT Deuteronomy. In several points, an ellipse appears in Deuteronomy or vice versa, a rectangle in Numbers. This kind of bird’s-eye view of the whole section allows for a refined analysis of the find, improving on earlier attempts to explain the differences. While earlier attempts of explanation focused on each short pericope seeking a local explanation for it in historico-literary terms, a wider view may yield simpler results using methods which are usually associated with textcritical work. All the textual data in the table may be explained by the motivation to duplicate segments from one account to the other, yet this general motivation could be classified into more sensitive categories. The analysis below streamlines the categories as follows. 33

To be clear, Deuteronomy does not mention a negotiation with Ammon so it does not directly contradict Judg 11 and Num 20. But it does depict Ammon as part of the routine of east-Jordanian nations.

230

JONATHAN BEN-DOV

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

231

Duplication of Speech Acts SP of Numbers 20–21, as may be expected, contains both the ellipses of MT Numbers and the rectangles of Deuteronomy. Deuteronomistic commands to do or to avoid war were consistently inserted in SP Numbers for all five nations east of the Jordan. Rather than mere mechanic duplication, they were meant for smoothing speech acts as defined above. Let us examine an example: SP Numbers 21, 4QNumb34

MT Deuteronomy 2

‫ויסעו מאבות ויחנו בעיי‬ ‫העברים במדבר אשר על‬ -‫פני מואב מזרח השמש׃‬ ‫]ו[יואמ]ר יהוה אל משה‬ ‫אל תצור את מואב ואל‬ ‫תתגר בם מלחמה כיא‬ ‫[לוא אתן ֗מ]ארצו ירשה‬ ‫כי לבני לוט נתתי את ער‬ ‫ירשה‬

MT Numbers 21 ‫א ֑בֹת ַ ֽו יַּ ֲחנ֞ וּ‬ ֹ ‫ וַ יִּ ְס ֖עוּ ֵמ‬11 ‫ְבּ ִע ֵיּ ֣י ָ ֽה ֲע ָב ִ ֗רים ַבּ ִמּ ְד ָבּ ֙ר‬ ‫מוֹאב‬ ָ֔ ‫ל־פּ ֵנ֣י‬ ְ ‫ֲא ֶשׁ ֙ר ַע‬ ‫ִמ ִמּזְ ַ ֖רח ַה ָ ֽשּׁ ֶמשׁ׃‬

‫הוה ֵא ֗ ַלי ֶאל־‬ ֜ ָ ְ‫אמר י‬ ֶ ֹ ‫ וַ ֨יּ‬9 ‫ת־מוֹאב וְ ַאל־‬ ָ֔ ‫ָ֨תּ ַצ ֙ר ֶא‬ ‫ִתּ ְת ָגּ֥ר ָ ֖בּם ִמ ְל ָח ָ ֑מה ֠ ִכּי ֽל ֹא־‬ ‫ֶא ֵ֨תּן ְלָך֤ ֵ ֽמ ַא ְר ֙צוֹ יְ ֻר ָ֔שּׁה ִ ֣כּי‬ ‫ת־ער‬ ֖ ָ ‫י־לוֹט נָ ַ ֥ת ִתּי ֶא‬ ֔ ֵ‫ִל ְבנ‬ ‫יְ ֻר ָשּׁה׃‬

‫משם נסעו ויחנו[ בנחל‬ ‫ז֗ ֯ר]ד׃‬

‫ ִמ ָ ֖שּׁם נָ ָ ֑סעוּ ַ ֽו יַּ ֲחנ֖ וּ‬12 ‫ְבּ ַנ ַ֥חל ָז ֶֽ רד׃‬ ‫ ָה ֵא ִ ֥מים ְל ָפ ִנ֖ים ָי ְ֣שׁבוּ‬10 ‫ָ ֑בהּ ַע֣ם גָּ ֥דוֹל וְ ַ ֛רב וָ ָ ֖רם‬ ‫ ְר ָפ ִ ֛אים‬11 ‫ָכּ ֲענָ ִ ֽ קים׃‬ ‫ף־הם ָכּ ֲענָ ִ ֑ קים‬ ֖ ֵ ‫יֵ ָח ְשׁ ֥בוּ ַא‬ ‫מּ ָא ִ֔בים יִ ְק ְר ֥אוּ ָל ֶ ֖הם‬ ֹ ֣ ‫וְ ַה‬ ‫וּב ֵשׂ ֞ ִעיר יָ ְשׁ ֣בוּ‬ ְ 12 ‫ֵא ִ ֽמים׃‬ ‫וּב ֵנ֧י ֵע ָ ֣שׂו‬ ְ ‫ם‬ ֒ ‫ים ְל ָפנִ י‬ ֮ ‫ַהח ִֹר‬ ‫ידוּם‬ ֙ ‫יִ ָֽיר ֗שׁוּם וַ יַּ ְשׁ ִמ‬ ‫יהם וַ יֵּ ְשׁ ֖בוּ ַתּ ְח ָ ֑תּם‬ ֶ֔ ֵ‫ִמ ְפּנ‬ ‫ַכּ ֲא ֶ ֧שׁר ָע ָ ֣שׂה יִ ְשׂ ָר ֵ֗אל‬ ‫ְל ֶ֨א ֶר ֙ץ יְ ֻר ָשּׁ ֔תוֹ ֲא ֶשׁר־נָ ַ ֥תן‬ ‫הו֖ה ָל ֶ ֽהם׃‬ ָ ְ‫י‬

34

As may be expected, the reading in 4QNumb is not entirely consistent with the later SP. This is true for the entire Book of Numbers. Thus for example in 13:20 4QNumb reads ‫ו[אם רזה היאה‬ ֗ for MT and SP ‫אם רזה‬. Such is the case for other pre-Samaritan manuscripts of the Pentateuch as well; see Emanuel Tov, “The Samaritan Pentateuch and the Dead Sea Scrolls: The Proximity of the Pre-Samaritan Qumran Scrolls to the SP,” in KeterShemTov.EssaysontheDeadSeaScrollsinMemoryofAlanCrown (ed. S. Tzoref and I. Young; Piscataway NJ: Gorgias, 2013), 59-88.

232

JONATHAN BEN-DOV

Moab is mentioned in passing in 21:11. This verse does not depict an encounter with Moab, but rather mentions the name of a place which stands east of it. At this point the (pre-)SP reviser decided to insert the report from Deuteronomy 2:9 of the divine command to avoid war with Moab. The words duplicated from Deuteronomy are only those which constitute a speech act, while the rest of the Moab section in vv. 10-12 is not duplicated, being a narrative rather than a report of direct speech.35 The interpolator’s only concern was the integrity of speech acts.36 The encounter with Sihon in SP Numbers constitutes another significant example. MT reports a simple plot: envoys being sent to Sihon and rejected, he marches to war only to be countered by the Israelites and defeated. SP adds elements from the parallel account in Deut 2:24-33. It introduces divine speeches to Moses, commanding him to make war with Sihon not once but twice (Deut 2:24 → SP Num 21:20a; Deut 2:31 → SP Num 2:23a).37 This addition corresponds to the speech-act duplications, featuring direct speech by God. The Sihon section in SP Numbers contains additional verses from Deuteronomy, all constituting direct speech that needs to be reaffirmed. This time it is not God’s speech but rather Moses’ speech: parts of the envoys’ message pertaining to specific requests from the host nation and mentioning previous deals with other nations (Deut 2:27-29). 35

36

37

One could argue that the “ethnological” reports in Deut 2:10-12 (and similarly 2:20-23, 3:9) are later additions into Deuteronomy, which did not exist in the copy of Deuteronomy used by pre-SP. Otto (DasDeuteronomium, 133-134, n. 105, with additional bibliography) claimed for the lateness of these “insertions” without invoking the argument from SP, but rather only on historico-critical considerations. I do not wish to argue with Otto about his diachronic literary analysis, as there are ample other scholars who disagree with him (e.g. Norbert Lohfink, “Die Stimmen in Deuteronomium 2,” BZ n.f. 37 [1993]: 209-235). However, the evidence of SP cannot be used to claim for the lateness of the ethnological reports, there being little evidence for extensive literary accretions in Deuteronomy as late as the second century BCE. Another sort of non-speech-act harmonization may be encountered in the Moab narrative in Num 21. Angela Roskop Erisman, “Transjordan in Deuteronomy: The Promised Land and the Formation of the Pentateuch”, JBL 132 (2013): 771, claims that Num 21:11bb and 13ab are glosses inserted on the basis of the parallel account in Deuteronomy. In that account, the Israelites did not cross directly through Moab, but rather bypassed it, arriving at Sihon’s land from the east. Roskop Erisman suggests further far-reaching ideas about the literary formation of Numbers 21. She treats the relationship between the Numbers and Deuteronomy accounts on the literary-editorial level rather than on the scribal level, the opposite attitude to that embraced here. I claim in this article that the point of view of textual criticism is as effective and useful as that of literary-historical criticism, and that it is more empirical than the latter, based on numerous examples from the preSamaritan Pentateuch. The reason for the existence of two such speeches in Deuteronomy will be examined below.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

233

An Early Non-Speech-Act Duplication in MT The only rectangle in the column of MT Numbers is 21:33-35, the section on war with Og, otherwise containing ellipses only. The reason why this paragraph must be seen as duplicating Deut 3:1-3 was explained above.38 This must be a very early duplication, being attested in all major text-types. Note that it is not a speech-act duplication of the pre-SP special type. While typical duplications in pre-SP operate only on speech acts, Num 21:33-35 duplicates a third person narrative. This particular case may be properly called ‘harmonization’ inasmuch as it involves mere general interest in harmonizing the stories. MT Deuteronomy 3:1-3

MT Numbers 21:33-35

‫עוֹג‬ ֩ ‫ וַ ֵנּ ֶ֣פן וַ ַ֔נּ ַעל ֶ ֖דּ ֶרְך ַה ָבּ ָ ֑שׁן וַ יֵּ ֵצ֣א‬1 ‫ל־ע ֛מּוֹ‬ ַ ‫אתנוּ ֧הוּא וְ ָכ‬ ֵ֜ ‫ְך־ה ָבּ ָ֨שׁן ִל ְק ָר‬ ַ ‫ֶ ֽמ ֶל‬ ‫ַל ִמּ ְל ָח ָ ֖מה ֶא ְד ֶ ֽר ִעי׃‬ ָ ְ‫אמר י‬ ֶ ֹ ‫ וַ ֨יּ‬2 ‫ירא א ֹ֔תוֹ‬ ֣ ָ ‫ל־תּ‬ ִ ‫הו֤ה ֵא ַ ֙לי ַא‬ ‫ת־א ְר ֑צוֹ‬ ַ ‫ל־ע ֖מּוֹ וְ ֶא‬ ַ ‫ת־כּ‬ ָ ‫ִ ֣כּי ְביָ ְד ָ֞ך נָ ַ ֧ת ִתּי א ֹ֛תוֹ וְ ֶא‬ ָ ‫וְ ָע ִ ֣שׂ‬ ‫חֹן ֶ ֣מ ֶלְך‬ ֙ ‫ית ְל ִסי‬ ָ ‫שׁר ָע ִ֗שׂ‬ ֣ ֶ ‫ית ֔לּוֹ ַכּ ֲא‬ ‫יוֹשׁב ְבּ ֶח ְשׁ ֽבּוֹן׃‬ ֖ ֵ ‫ָ ֽה ֱאמ ִ ֹ֔רי ֲא ֶ ֥שׁר‬ ‫ת־עוֹג‬ ֥ ‫ֹלהינוּ ְבּיָ ֵ ֗דנוּ ַגּ֛ם ֶא‬ ֵ֜ ‫הוה ֱא‬ ֨ ָ ְ‫ וַ יִּ ֵתּן֩ י‬3 ַ ‫ת־כּ‬ ָ ‫ְך־ה ָבּ ָ ֖שׁן וְ ֶא‬ ַ ‫ֶ ֽמ ֶל‬ ‫ד־בּ ְל ִ ֥תּי‬ ִ ‫ל־ע ֑מּוֹ וַ נַּ ֵ֕כּהוּ ַע‬ ‫יר־לוֹ ָשׂ ִ ֽריד׃‬ ֖ ‫ִה ְשׁ ִ ֽא‬

‫עוֹג‬ ֩ ‫ וַ יִּ ְפ ֙נוּ ַ ֽו יַּ ֲע ֔לוּ ֶ ֖דּ ֶרְך ַה ָבּ ָ ֑שׁן וַ יֵּ ֵצ֣א‬33 ‫ל־ע ֛מּוֹ‬ ַ ‫אתם ֧הוּא וְ ָכ‬ ָ֜ ‫ְך־ה ָבּ ָ֨שׁן ִל ְק ָר‬ ַ ‫ֶ ֽמ ֶל‬ ‫ַל ִמּ ְל ָח ָ ֖מה ֶא ְד ֶ ֽר ִעי׃‬ ‫ירא א ֹ֔תוֹ‬ ֣ ָ ‫ל־תּ‬ ִ ‫הו֤ה ֶאל־מ ֶֹשׁ ֙ה ַא‬ ָ ְ‫אמר י‬ ֶ ֹ ‫ וַ ֨יּ‬34 ‫ת־א ְר ֑צוֹ‬ ַ ‫ל־ע ֖מּוֹ וְ ֶא‬ ַ ‫ת־כּ‬ ָ ‫ִ ֣כּי ְביָ ְד ָ֞ך נָ ַ ֧ת ִתּי א ֹ֛תוֹ וְ ֶא‬ ‫חֹן ֶ ֣מ ֶלְך‬ ֙ ‫ית ְל ִסי‬ ָ ‫שׁר ָע ִ֗שׂ‬ ֣ ֶ ‫ית ֔לּוֹ ַכּ ֲא‬ ָ ‫וְ ָע ִ ֣שׂ‬ ‫יוֹשׁב ְבּ ֶח ְשׁ ֽבּוֹן׃‬ ֖ ֵ ‫ָ ֽה ֱאמ ִ ֹ֔רי ֲא ֶ ֥שׁר‬ ָ ‫ וַ יַּ ֨כּוּ א ֹ֤תוֹ וְ ֶא‬35 ‫ל־ע ֔מּוֹ ַעד־‬ ַ ‫ת־כּ‬ ָ ‫ת־בּנָ ֙יו וְ ֶא‬ ‫יר־לוֹ ָשׂ ִ ֑ריד וַ ִיּ ְֽיר ֖שׁוּ ֶאת־‬ ֖ ‫ִבּ ְל ִ ֥תּי ִה ְשׁ ִ ֽא‬ ‫ַא ְר ֽצוֹ׃‬

Duplication and the Literary Composition of Deuteronomy A scrutiny of the table (above) calls for a reevaluation of the common assumption that Deuteronomy used the (JE) narratives of Num 20–21 as a source. The table, rather surprisingly, does not support this assumption. Curiously, the column of MT Deuteronomy contains many rectangles but only one ellipse. Had there existed direct borrowing, one would expect to see more traces for the motif of envoys. In the Edom narrative (2:2-7) for instance, neither envoys nor Edom’s answers are reported in Deut 2:2-7. 38

This conclusion is shared by nearly all scholars, for example: Wellhausen, Composition, 109; Gray, Numbers, 306; Licht, Numbers, II 212; Rofé, “Historico-Literary Aspects” (with earlier bibliography); Noth, Numbers, 162, 166. Opposition to it comes from Haran (TheBiblicalCollection, II, 198), Brettler (TheCreationofHistory, 75) and Baden (J, E, and the Redaction, 140-141), who claim that Num 21:33-35 is an earlier source which in turn served as a source for the Og report in Deut 3:1-3. The early texts regularly quoted in the historical speeches of Deuteronomy, however, never correspond to the speech word by word as they do here.

234

JONATHAN BEN-DOV

This absence and others like it raise the possibility that Deut 2–3 did not rely on Num 20–21 as a source, but rather developed an independent narrative line using distinct vocabulary and style. Cases of verbal similarity between Deut 2 and Num 20–21 may be explained as later scribal accretions based on duplication. The main example is the single ellipse in Deut 2, placed in the Sihon narrative (2:24-32). ‫ְך־ח ְשׁ ֧בּוֹן ָ ֽה ֱאמ ִ ֹ֛רי‬ ֶ ‫ת־סי ֨חֹן ֶ ֽמ ֶל‬ ִ ‫אַר ֹנן֒ ְר ֵ ֣אה נָ ַ ֣ת ִתּי ֠ ְביָ ְדָך ֶא‬ ְ ‫ת־נ ַ֣חל‬ ַ ‫רוּ ֶא‬ ֮ ‫ ֣קוּמוּ ְסּ ֗עוּ וְ ִע ְב‬24 ‫אָחל֙ ֵ ֤תּת ַפּ ְח ְדּ ָ֙ך וְ יִ ְראָ ְ֣ת ָ֔ך‬ ֵ ‫ ַהיּ֣ וֹם ַה ֶ֗זּה‬25 ‫ת־אַר ֖צוֹ ָה ֵ ֣חל ָ ֑רשׁ וְ ִה ְת ָגּ֥ר ֖בּוֹ ִמ ְל ָח ָ ֽמה׃‬ ְ ‫וְ ֶא‬ ֙ ‫ל־ה ָשּׁ ָ ֑מיִם ֲא ֶ ֤שׁר יִ ְשׁ ְמ‬ ַ ‫ל־פּנֵ ֙י ָ ֽה ַע ִ֔מּים ַ ֖תּ ַחת ָכּ‬ ְ ‫ַע‬ ‫ וָ ֶא ְשׁ ַל֤ח‬26 ‫עוּן ִשׁ ְמ ֲע ָ֔ך וְ ָרגְ ז֥ וּ וְ ָחל֖ וּ ִמ ָפּ ֶנֽיָך׃‬ ֹ ֽ ‫יחוֹן ֶ ֣מ ֶלְך ֶח ְשׁ ֑בּוֹן ִדּ ְב ֵ ֥רי ָשׁל֖ וֹם ֵלא‬ ֖ ‫ל־ס‬ ִ ‫אָכ ֙ים ִמ ִמּ ְד ַ ֣בּר ְק ֵד ֔מוֹת ֶא‬ ִ ‫ַמ ְל‬ ‫ ֶא ְע ְבּ ָ ֣רה ְבאַ‬27 ‫מר׃‬ ֹ ֣ 28 ‫וּשׂ ֽמ ֹאול׃‬ ְ ‫יָמין‬ ֥ ִ ‫אָסוּר‬ ֖ ‫ְר ֶ֔צָך ַבּ ֶ ֥דּ ֶרְך ַבּ ֶ ֖דּ ֶרְך ֵא ֵלְ֑ך ֥ל ֹא‬ ‫אָכ ְל ִתּי וּ‬ ַ֔ ְ‫א ֶכל ַבּ ֶכּ ֶ֤סף ַתּ ְשׁ ִבּ ֵ ֙רנִ ֙י ו‬ ִ ‫ן־לי וְ ָשׁ ִ ֑ת‬ ֖ ִ ‫ַ ֛מיִם ַבּ ֶכּ ֶ֥סף ִתּ ֶתּ‬ ‫שׂוּ־לי ְבּ ֵנ֣י ֵע ָ֗שׂו ַה ֹֽיּ ְשׁ ִב ֙ים‬ ִ ֜ ‫ ַכּ ֲא ֶ֨שׁר ָ ֽע‬29 ‫יתי ַ ֖רק ֶא ְע ְבּ ָ ֥רה ְב ַרגְ ָ ֽלי׃‬ ַ ‫ר־א ֱעב ֹ֙ר ֶא‬ ֽ ֶ ‫וֹאָבים ַהיּ ְֹשׁ ִ ֖בים ְבּ ָ ֑ער ַ ֤עד ֲא ֶ ֽשׁ‬ ִ֔ ‫ְבּ ֵשׂ ֔ ִעיר וְ ַה ֣מּ‬ ‫הו֥ה‬ ָ ְ‫ל־האָ ֶ֕רץ ֲא ֶשׁר־י‬ ָ ‫ת־היַּ ְר ֵ ֔דּן ֶא‬ ֙ ‫אָבה ִסי‬ ָ֗ ‫ וְ ֣ל ֹא‬30 ‫ֹלהינוּ נ ֵ ֹ֥תן ָ ֽלנוּ׃‬ ֖ ֵ ‫ֱא‬ ‫ֹלהיָך‬ ֶ֜ ‫הוה ֱא‬ ֨ ָ ְ‫י־ה ְק ָשׁ ֩ה י‬ ִ ‫חֹן ֶ ֣מ ֶלְך ֶח ְשׁ ֔בּוֹן ַה ֲע ִב ֵ ֖רנוּ ֑בּוֹ ִ ֽכּ‬ ְ ‫ת־רוּחוֹ וְ ִא ֵמּ ֙ץ ֶא‬ ֗ ‫ֶא‬ ‫אמר יְ הוָ ֙ה ֵא ֔ ַלי ְר ֵ֗אה‬ ֶ ֹ ‫ וַ ֤יּ‬31 ‫ת־ל ָב ֔בוֹ ְל ַ ֛מ ַען ִתּ ֥תּוֹ ְביָ ְדָך֖ ַכּיּ֥ וֹם ַה ֶזּֽה׃ ס‬ ִ ‫תי ֵ ֣תּת ְל ָפ ֶ֔ניָך ֶא‬ ֙ ִ ‫ַ ֽה ִח ֹּ֙ל‬ ‫ וַ יֵּ ֵצ ֩א ִסי ֨חֹן‬32 ‫ת־אַר ֽצוֹ׃‬ ְ ‫ת־אַר ֑צוֹ ָה ֵ ֣חל ָ ֔רשׁ ָל ֶ ֖ר ֶשׁת ֶא‬ ְ ‫ת־סי ֖חֹן וְ ֶא‬ ‫ל־ע ֛מּוֹ ַל ִמּ ְל ָח ָ ֖מה ָי ְֽה ָצה׃‬ ַ ‫אתנוּ ֧הוּא וְ ָכ‬ ֵ֜ ‫ִל ְק ָר‬

The “ellipse’ (i.e., the motif of envoys) in vv. 26-29 stands out of its environment. This section creates complications both in the general setting of chapter 2 and in the unit 2:24-32 in particular. Verses 26-29 are enveloped by decisive commands to make war with Sihon (vv. 24-25, 31), declaring that war as a pivotal act in the Israelite campaign. As Shimon Gesundheit has clarified, it is inconceivable that Moses in v. 26 sends envoys to Edom, and explicitly defies the divine word!39 The envoy is called here ‫דברי‬ ‫‘ שלום‬words of peace’ (2:26), in glaring contrast to the divine command to start war, ‫( מלחמה‬2:24).40 The two divine commands (vv. 24-25, 31) form a Wiederaufnahme, marking the envoys account contained within the frame (vv. 26-29) as a later insertion into the original narrative.41 39

40

41

Shimon Gesundheit, “Die Midrasch-Exegese im Dienst der Literarkritik. Zum Beispiel, Krieg und Frieden in Dtn 2,24-32,” in Congress Volume Munich 2013 (VTSup 163; ed. C. Maier; Leiden: Brill, 2014), 111-124. Beginning the section in v. 26 rather than v. 24, Brettler (TheCreationofHistory, 71-72) does not underscore this contradiction. Weinfeld, Deuteronomy 1-11, 171, tried to harmonize this difficulty seeing Moses’ message only as a “pretext” for starting war. Gesundheit (“Midrasch-Exegese,” 113115) quotes him and other similar commentators and convincingly rejects them. The rabbis noted the contradiction and condemned Moses for his peace-seeking frailty (Deuteronomy Rabbah, Shoftim 5). I.e., as Gesundheit clarifies, it is not the case that the entire unit 2:24-31 is a late insertion (as claimed by Carl Steuernagel, DasDeuteronomium[HKAT; Göttingen: Vandenhoeck & Ruprecht, 1923], 58), but rather that vv. 26-31 were inserted, while vv. 24-25 constitute the original text of the Sihon account.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

235

According to Gesundheit, vv. 26-29 were inserted for a theological purpose, the intention being to balance the message of vv. 24-25, which was no longer valid in the eyes of the reviser. While according to the initial command the war with Sihon already constituted the beginning of the holy war on the promised land whose border was the Arnon River, vv. 26-29 clarified in contrast that the Land only begins when “I cross the Jordan to the land which Yhwh our God gives us” (v. 29).42 The inserted material, thus Gesundheit, is a piece of “inner-biblical Midrash”, assembling bits and pieces from various sources in order to design a new theological message for the unit. Two different motivations should be distinguished in this pericope, however. Gesundheit is certainly right about the desire to moderate the bellicose tone of the original divine command (v. 24) with a more peaceful approach in v. 26.43 But this motivation could have been accomplished without importing to Deut 2 the element of envoys, i.e. the ellipse, so alien to this chapter. The detailed presence of this motif exceeds the motivation indicated by Gesundheit, and must have had a different reason. His solution thus only solves part of the problem – i.e., the short phrase ‫דברי‬ ‫( שלום‬2:26b) – but does not account for the wider questions arising about the dominant motifs in this chapter (i.e, rectangles vis-à-vis ellipses). In order to explain Deut 2:26-29 one needs to consider the entire array of East Jordanian encounters in both Numbers and Deuteronomy. More recently after Gesundheit, Ludwig Schmidt noted the indebtedness of Deut 2:26-29 to Num 21.44 Constructing an elaborate narrative of Fortschreibung for the entire array of accounts east of the Jordan, his study is indeed aware of the need of a wider explanations for the duplications. He acknowledges that Deut 2:26-32 (Sihon) depends on Num 21:21-23, whose core is indeed pre-Dtr and pre-P. According to him, Num 20:7-14 (Edom) is a later post-exilic passage, which reworks the older Sihon passage. In some aspects Schmidt’s argument is inferior to Gesundheit’s, having missed the sting of the contradiction between the command to make war in 2:24-25 and the sending of peace envoys in the immediately subsequent verse. Instead, Schmidt accommodates this 42 43

44

Gesundheit, “Midrasch-Exegese,” 115. This nuance is not motivated by different conceptions of the Promised Land (pace Gesundheit) but rather by the Deuteronomic law of war in Deut 20:10-18. See Brettler, The Creation of History, 73-74; Eckart Otto, Deuteronomium 1–11. Erster Teilband: 1,1–4,43 (HTKAT; Freiburg: Herder, 2012), 457. Brettler, ibid, 67-69 similarly explains how Deut 16:18-20 influenced the design of the narrative in Deut 1:9-18. Schmidt, “Sihon und Og”.

236

JONATHAN BEN-DOV

contradiction in the ongoing cycle of Fortschreibung. However, both Gesundheit and Schmidt agree, each in his own way, that the development of the Sihon narrative seeks a refinement of the theological or literary message of the stories. In contrast, I think that the solution cannot be sought only on the theological level, but rather also as fulfilling scribal needs.45 One may even think of two stages of textual growth, with a scribal-formalistic move adding first vv. 26-29, and a later ‘pacifistic’ interpolator adding the words ‫ דברי שלום‬in v. 26 at a second stage. The interpolators of the bulk of Deut 2:26-29 did not act with the intention of placing two contrastive messages next to each other. The effect created by this positioning is similar to the effect of placing two contradicting spy narratives next to each other in SP of Numbers 13.46 In the same way that the latter was not meant to create a unified theological message but rather only fulfil a formalistic need, this must be the case in the MT of the Sihon narrative as well. It is a scribal duplication, seeking to smooth out a speech-act or some other formalistic purpose. Other duplications known to us operate in the same way. The scribal mechanism of duplication was thus a dominant principle in the composition of the accounts of east-Jordanian encounters. The mechanism operated in both ways, copying material from Numbers to Deuteronomy and vice versa in the various versions. As encountered above, some duplicators came relatively early in the text’s formation, while other operated at a later stage. Taking place at an early stage, the changes made by the interpolator of Deut 2:26-31 are reflected in all text-types and in all versions. It is an early intervention, similar to the Og pericope in Num 21:33-35. Curiously, it functioned in the opposite direction: while the Og narrative copies text from Deuteronomy to Numbers, the Sihon narrative copies text from Numbers to Deuteronomy. Being an early case of duplication, which does not correspond to the later characteristics of pre-SP duplications, the Sihon account 45

46

Gesundheit (in his useful chart on p. 124) is right that Deut 2:26-29 does not strictly copy material from Num 21 as is usually the case, and cannot thus be considered a straightforward duplication like those of SP. Rather, the passage is a pastiche of phrases from Numbers 20, 21 and even from elsewhere in Deuteronomy 2. Yet the Sihon narrative in Num 21:21-23 functions as the skeleton of the entire passage Deut 2:26-29, with a variety of elements – some typically deuteronomistic, others characteristic of the similar Edom accounts – accumulating on it. Being an early scribal duplication its use of the sources must have been more free than the later SPtype duplications. See Segal, “The Text of the Hebrew Bible,” 11-12.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

237

in Deuteronomy uses different techniques: not strictly copy-paste but rather a collection of phrases from several related places.47 Post-MT Developments in Deuteronomy A long process of development by duplication was depicted thus far: two early duplications represented in all versions, and a systematic expansion represented in pre-SP. The vitality of the idea of duplication did not stop there, however, and continued through even later textual manifestations of the account. The first case to be examined is the SP of the Edom pericope in Deut 2:1-7.48 In general, SP of Deut 2 reflects a text similar to MT with only a handful of minor changes.49 It represents all the ‘rectangles’ of MT and in addition the ‘ellipse’ already included in the Sihon pericope MT Deut 2:2629. In addition however, while MT Deuteronomy does not know a request and refuse (‘ellipse’) component in Edom, the SP hastens to complete the picture, adding it to the chapter. SP Deut 2:2-7b

MT Deut 2: 2-7

MT Num 20:14-18

‫ ויאמר יהוה אלי לאמר׃‬2 ‫מר׃‬ ֹ ֽ ‫הו֖ה ֵא ַ ֥לי ֵלא‬ ָ ְ‫אמר י‬ ֶ ֹ ‫ וַ ֥יּ‬2 ‫ רב לכם סוב את ההר‬3 ‫ת־ה ָ ֣הר‬ ָ ‫ב־ל ֶ֕כם ֖סֹב ֶא‬ ָ ‫ ַר‬3 ‫הזה פנו לכם צפונה׃‬ ‫ַה ֶזּ֑ה ְפּנ֥ וּ ָל ֶכ֖ם ָצ ֽ ֹפנָ ה׃‬ ֣ ֶ ֒‫מר‬ ֹ ‫ת־ה ָע ֮ם ַצ֣ו ֵלא‬ ָ ‫ וְ ֶא‬4 ‫ ואת העם צוי לאמר אתם‬4 ‫אַתּם‬ ‫ֽעֹ ְב ִ ֗רים ִבּגְ בוּל֙ ֲא ֵח ֶיכ֣ם‬ ‫עברים בגבול אחיכם בני‬ ‫י־ע ָ֔שׂו ַהיּ ְֹשׁ ִ ֖בים ְבּ ֵשׂ ִ ֑עיר‬ ֵ ֵ‫ְבּנ‬ ‫עשו היושבים בשעיר‬ ‫וְ יִ ְֽיר ֣אוּ ִמ ֶ֔כּם וְ נִ ְשׁ ַמ ְר ֶ ֖תּם‬ ‫וייראו מכם ונשמרתם‬ ִ 5 ‫אד׃‬ ֹ ֽ ‫ְמ‬ ‫ אל תתגרו בם כי‬5 ‫אַל־תּ ְתגָּ ֣רוּ ָ֔בם ֠ ִכּי מאד׃‬ ‫אַר ָ֔צם ַ ֖עד‬ ְ ‫א־א ֵ ֤תּן ָל ֶכ ֙ם ֵ ֽמ‬ ֶ ֹ ‫ֽל‬ ‫לא אתן לכם מארצם‬ ֑ ָ ‫ִמ ְד ַ ֣רְך ַכּ‬ ‫ף־רגֶ ל ִ ֽכּי־יְ ֻר ָ ֣שּׁה‬ ‫ירשה עד מדרך כף רגל‬ ‫ת־הר‬ ֥ ַ ‫ְל ֵע ָ֔שׂו נָ ַ ֖ת ִתּי ֶא‬ ‫כי ירשה לעשו נתתי את‬ ‫ אכל תשבירו‬6 ‫הר שעיר׃‬ ‫א ֶכל ִתּ ְשׁ ְבּ ֧רוּ‬ ֹ ֣ 6 ‫ֵשׂ ִ ֽעיר׃‬ ‫מאתם בכסף ואכלתם‬ ‫ֵ ֽמ ִא ָ ֛תּם ַבּ ֶכּ ֶ֖סף וַ ֲא ַכ ְל ֶ ֑תּם‬ ַ֜ ַ‫וְ ג‬ ‫וגם מים תכירו‬ ‫ם־מיִם ִתּ ְכ ֧רוּ‬ 47

48 49

A concise and useful description of the technique used in 2:26-29, and of Deut 2 in general was supplied by Brettler, theCreationofHistory, 73-75. Brettler rightly points out that harmonization with the Numbers account was not the only operative principle but also assimilation of adjacent narratives within Deut 2. This version is not represented in any copy of Deuteronomy from Qumran. Some examples are: in 2:5 the phrase ‫( לא אתן לכם מארצם‬MT) is longer in SP: ‫לא אתן‬ ‫לך מארצם ירושה‬, following MT Deut 2:9; in 2:31 the name Sihon is supplemented by the title ‫סיחון מלך חשבון האמורי‬.

‫‪238‬‬

‫‪JONATHAN BEN-DOV‬‬

‫‪MT Num 20:14-18‬‬

‫‪MT Deut 2: 2-7‬‬

‫‪SP Deut 2:2-7b‬‬

‫יתם׃‬ ‫וּשׁ ִת ֶ ֽ‬ ‫ֵמ ִא ָ ֛תּם ַבּ ֶכּ ֶ֖סף ְ‬ ‫ֹלהיָך ֵ ֽבּ ַר ְכ ָ֗ך‬ ‫הוה ֱא ֶ֜‬ ‫‪ִ 7‬כּי֩ יְ ָ ֨‬ ‫שׂה יָ ֶ ֔דָך יָ ַ ֣דע‬ ‫ְבּכֹל֙ ַמ ֲע ֵ ֣‬ ‫ת־ה ִמּ ְד ָ ֥בּר ַהגָּ ֖ד ֹל‬ ‫ֶל ְכ ְתּ ָ֔ך ֶא ַ‬ ‫אַר ָבּ ִ ֣עים ָשׁ ָ֗נה‬ ‫ַה ֶזּ֑ה ֶז֣ה ׀ ְ‬ ‫ֹלה ֙יָך ִע ָ֔מְּך ֥ל ֹא‬ ‫הו֤ה ֱא ֶ֙‬ ‫יְ ָ‬ ‫ָח ַ ֖ס ְר ָתּ ָדּ ָ ֽבר׃‬

‫מאתם בכסף ושתיתם׃‬ ‫‪ 7‬כי יהוה אלהיך ברכך‬ ‫בכל מעשה ידיך ידע‬ ‫לכתך את המדבר הגדול‬ ‫הזה זה ארבעים שנה‬ ‫יהוה אלהיך עמך לא‬ ‫חסרת דבר׃‪--‬‬

‫אָכים‬ ‫‪ 14‬וַ יִּ ְשׁ ֨ ַלח מ ֶ ֹ֧שׁה ַמ ְל ִ ֛‬ ‫ִמ ָקּ ֵ ֖דשׁ ֶא ֶ ֣‬ ‫ל־מ ֶלְך ֱא ֑דוֹם ֤כֹּה‬ ‫אַתּה‬ ‫אָחיָך יִ ְשׂ ָר ֵ֔אל ָ ֣‬ ‫אָמ ֙ר ִ ֣‬ ‫ַ‬ ‫ל־ה ְתּ ָלאָ֖ה‬ ‫יָ ַ ֔ד ְע ָתּ ֵ ֥את ָכּ ַ‬ ‫ֲא ֶ ֥שׁר ְמ ָצאָ ְֽתנוּ׃ ‪ 15‬וַ יֵּ ְר ֤דוּ‬ ‫ינוּ ִמ ְצ ַ ֔ר ָ‬ ‫ֲאב ֵֹ֙ת ֙‬ ‫יְמה וַ ֵנּ ֶ֥שׁב‬ ‫יָמים ַר ִ ֑בּים וַ יָּ ֵ ֥רעוּ‬ ‫ְבּ ִמ ְצ ַ ֖ריִם ִ ֣‬ ‫ָל֛נוּ ִמ ְצ ַ ֖ריִם וְ ַל ֲאב ֵ ֹֽתינוּ׃‬ ‫‪ 16‬וַ נִּ ְצ ַ ֤עק ֶאל־יְ הוָ ֙ה וַ יִּ ְשׁ ַ ֣מע‬ ‫ק ֔ ֵֹלנוּ וַ יִּ ְשׁ ַל֣ח ַמ ְלאְָ֔ך וַ יּ ִֹצ ֵ ֖אנוּ‬ ‫ִמ ִמּ ְצ ָ ֑ריִם וְ ִהנֵּ ֙ה ֲא ַנ ְ֣חנוּ‬ ‫בוּלָך׃‬ ‫ְב ָק ֵ ֔דשׁ ִ ֖עיר ְק ֵ ֥צה גְ ֶ ֽ‬ ‫‪17‬‬

‫‪ 7a-b‬ואשלחה מלאכים‬ ‫אל מלך אדום לאמר‬

‫אַר ֶ֗צָך ֤ל ֹא‬ ‫ה־נּ֣א ְב ְ‬ ‫נַ ְע ְבּ ָר ָ‬ ‫וּב ֶ֔כ ֶרם וְ ֥ל ֹא‬ ‫נַ ֲעב ֹ֙ר ְבּ ָשׂ ֶ ֣דה ְ‬ ‫נִ ְשׁ ֶ ֖תּה ֵ ֣מי ְב ֵ ֑אר ֶ ֧דּ ֶרְך‬ ‫יָמין‬ ‫ַה ֶ ֣מּ ֶלְך נֵ ֗ ֵלְך ֤ל ֹא נִ ֶטּ ֙ה ִ ֣‬ ‫ְ‬ ‫וּשׂ ֔מ ֹאול ַ ֥עד ֲא ֶ ֽשׁר־נַ ֲע ֖בֹר‬ ‫אמר ֵא ָל ֙יו‬ ‫בוּלָך׃ ‪ 18‬וַ ֤יּ ֹ ֶ‬ ‫גְּ ֶ ֽ‬ ‫ֱא ֔דוֹם ֥ל ֹא ַת ֲע ֖בֹר ִ ֑בּי ֶפּן־‬ ‫אתָך׃‬ ‫ַבּ ֶ ֖ח ֶרב ֵא ֵ ֥צא ִל ְק ָר ֶ ֽ‬

‫אעברה בארצך לא אטה‬ ‫בשדה ובכרם ולא נשתה‬ ‫מי בור דרך המלך נלך‬ ‫לא נסור ימין ושמאל עד‬ ‫אשר נעבר גבולך ויאמר‬ ‫לא תעבר בי פן בחרב‬ ‫אצא לקראתך‬

‫‪This duplication copies the parallel section from Num 20 only partly.‬‬ ‫‪While Num 20 knows two separate stories of envoys to Edom, one (vv. 14‬‬‫‪17) by Moses and the other (v. 19) by the Israelites, the SP copies only the‬‬ ‫‪first of the two. Even that section is not fully copied, as the historical‬‬ ‫‪introduction of that envoy (vv. 14b-16) is omitted.50 SP thus copies only‬‬ ‫‪the essential request of v. 17 and the response of v. 18. Note that this‬‬ ‫‪duplication does not specifically address the fulfillment of a speech act:‬‬ ‫‪For this introduction see Wolfgang Oswald, “Die Revision des Edombildes in Numeri‬‬ ‫‪XX 14-21,” VT 50 (2000): 219-221; earlier Calum Carmichael, “A New View of the‬‬ ‫‪Origin of the Deuteronomic Credo,” VT 19 (1969): 273-289.‬‬

‫‪50‬‬

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

239

it copies narrative next to it, as well as the speech of the king of Edom, hardly a sacred kind of speech that requires particular protection. The SP does not even copy the direct speech in its entirety. This duplication is thus part of a different wave from those encountered in the SP of Numbers. Like the duplication in the MT of the Sihon narrative explained in detail above, this one was moved from Numbers to Deuteronomy, while sharing the essential feature of not copying the exact direct speech, and only it. An even later stage of the same process can be seen in a small fragment of the Reworked Pentateuchc (4Q365 frg. 37).51 ]◦ [ ]◦ ‫[ד‬ ]‫[וכול העם המלחמה‬ ]‫[מנחל ארנון ויחנו‬ ̊ ‫בא ̊ר]נון‬ ̊ ‫[י̊ ו ויחנו‬ ]‫בית‬ ֗ ‫[מים עד‬ ֗

1 2 3 4 5

This fragment mentions the Arnon River (line 3) and the ‘people of war’ (line 2, a phrase typical of Josh 8-11). Tov and White suggested that the fragment refers to Deut 2:24 or 2:36, where the Arnon is mentioned, but in principle it can also refer to Num 21:13 ‫ויחנו מעבר ארנון‬. Zahn later suggested that it better fits in Numbers because it employs the third person plural, rather than the first person of Deut 2.52 Be as it may, the fragment hints to both Deuteronomy and Numbers, as well as language taken from elsewhere. It seems to contain new material which emulates all of the above sources into a new formulation. It constitutes a further rewriting of the parallel passages about Moab, exemplifying further elaboration of the principle of expansion by duplication.53

PART 4: CONCLUSIONS (1) Numbers 20–21 and Deuteronomy 2–3 were originally two independent lines of story. The story in Numbers recounted the episodes of Edom and Sihon, and possibly also Moab, describing envoys sent to the respective kings and answered accordingly. The story in Deuteronomy 2–3, 51

52 53

Quoted after Emanuel Tov and Sidney White, “365. 4QReworked Pentateuchc,” in QumranCave4VIII.ParabiblicalTexts,Part1 (DJD 13; Oxford: Clarendon, 1994), 311. Zahn, RethinkingRewrittenScripture, 113-115. For a similar argument applied to other Reworked Pentateuch scrolls see e.g. Christoph Berner, “The Redaction of the Sinai Pericope (Exod 19-14) and Its Continuation in 4Q158,” DSD 20 (2013): 378-409.

240

JONATHAN BEN-DOV

in contrast, knew five nations east of the Jordan River, all controlled and administered by the divine command with Israel instructed by God accordingly. (2) As these two cycles gained authority, the impetus arose to compare them and create an even, harmonious picture. This impetus is driven by the idea “… that Scripture is perfect and perfectly harmonious… to discover the basic harmony underlying apparently discordant words, since all of Scripture, in their view, must speak with one voice”.54 Since the text is inspired, corrections of this sort tended to use the text already attested in the Pentateuch rather than to create entirely new text units. (3) Such changes did not derive from a wish to update the theological or literary message of the pericope at hand. The correction is in fact extraneous to the content of the textual unit. In the words of David Carr, “It would represent a fairly mechanical, non-theological form of harmonization”.55 (4) A first, general wave of corrections was motivated by the simple urge to harmonize different narratives. This early wave is reflected in all textual versions known to us. It was not a comprehensive effort, and did not follow strict rules in choosing the elements of speech to duplicate. This wave was represented by three elements in the above discussion, the first two of them represented in all textual witnesses: – (a) copying the Og narrative from Deut 3:1-3 to Num 21:33-35. – (b) creating an envoy report (Deut 2:26-29) within the Sihon narrative of Deuteronomy and concluding it with Wiederaufnahme, causing significant interruption in the resulting composition. This change does not involve pure duplication of Num 21 into Deuteronomy, but rather a more elaborate work of pastiche, probably involving also an insertion of the phrase ‫ דברי שלום‬in 2:26b. – (c) In the SP of Deuteronomy only, an envoy report was also copied to the story on Edom, from Numbers to Deuteronomy, as if to render the diagram of ellipses and rectangles even. (5) A more well-defined wave of duplications arose later, probably sometime in the second century BCE.56 The motivation for this wave was not simple textual harmony, but rather a unique interest in the perfection of speech acts. It was active in several focuses within the Pentateuch 54

55 56

James Kugel, TraditionsoftheBible (Cambridge MA; Harvard University Press, 1998), 17-19. Carr, TheFormation, 260. For the dating see Ben-Dov, “Early Texts of the Torah,” 221.

TEXT DUPLICATIONS BETWEEN HIGHER AND LOWER CRITICISM

241

where a significant concentration of speech acts could be found: the Plagues narratives, the Sinai covenant, Genesis 1, and Deuteronomy 1–3. This wave is highly characteristic of the pre-SP, but traces of it can also be seen within MT (for example Exod 32:9). The people who carried out this wave were highly informed and consistent, running what may be defined as an academic task rather than a vulgar or popular one. In Numbers 20–21 SP they took pains to duplicate all the direct speeches from Deuteronomy to Numbers, making sure to copy only the direct discourse, without the accompanying narrative. (6) Students of the formation of biblical literature should be aware of these scribal insertions alongside the theological or literary expansions better known in literary-critical circles. In this sense, the finds of lower criticism may sometimes prove to be more useful than those of higher criticism.

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ FOLLOW THE SAME MODEL? Abraham TAL

Apparently, an oxymoronic pairing. The first member is none other than the Samaritan Pentateuch, which, by Moses’ command, is the only book to be related to what we are used to call “the Bible”. This is clearly stated in Deut 4:2, ‫לא תוסיפו על הדבר אשר אנכי מצוה אתכם היום ולא תגרעו ממנו‬ ‫לשמר את מצות יהוה אלהיכם אשר אנכי מצוה אתכם‬, “You must neither add anything to what I command you nor take away anything from it, but keep the commandments of the LORD your God with which I am charging you,” and further stressed in Deut 13:1, ‫את כל הדבר אשר אנכי מצוך היום‬ ‫אֹתו תשמרו לעשות לא תוסיפו עליו ולא תגרעו ממנו‬, “You must diligently observe everything that I command you; do not add to it or deduct anything from it.” The latter is the introductory verse to a series of interdictions regarding ‫“ נביא או חולם חלום‬prophets or dream-diviners.” The “finale” of the entire Tora designates Moses as “the last prophet,” as the community understands Deut 34:10, ‫ולא קם נביא עוד בישראל כמשה‬, “never again did arise in Israel a prophet like Moses.” This solemn statement is developed in the Samaritan medieval hermeneutic composition, Tibåt Mårqe, which specifies that the word ‫ עוד‬of Deuteronomy regards the future. Book 4, paragraph 33 states: ‫איך משה לא קעם ולא יקום וכן אלה כתב‬ ‫בדילה ולא קם עוד נביא בישראל כמשה‬, “like Moses never arose and never will arise (‫)ולא יקום‬, and thus God wrote about him ‫ולא קם עוד‬.” Similar statements are repeated several times in the book, such as in paragraphs 6, 7, 89, etc.,1 and tens of times in their late medieval liturgy, probably as an understated response to the muslim claim of Muhammad being God’s prophet.2 Under such circumstances, obviously, the biblical prophets were utterly rejected, Isaiah making no exception.

1 2

Ben-Hayyim 1988, 309f. E.g.: ‫לא נשמע לאכתב אלא לה לית אלה רב הך יהובה ולא נשמע ממלל מן נביא אלא מן נביה‬ ‫דקבלה‬, “we shall not obey other book, but this (i.e., the Tora); there is no great God but He, and we shall not listen to other prophet but to the prophet that wrote it” (Cowley 1909, 60).

244

ABRAHAM TAL

Nevertheless, both the phonologic/morphologic shape and word arrangement of the Qumranic copy of Isaiah, known as “the great Isaiah scroll”, make it a reasonable match to the linguistic model reflected in the SP, namely, the language and style of the Second Temple period, more precise, the turn of the last millennium B.C.

LINGUISTIC FEATURES Right from the beginning: the scribe omitted the ‫ ע‬in the prophet’s name, which he, or a corrector tried to restore. The recession of the guttural consonants at that time is common knowledge. Eduard Kutcher’s brilliant treatment of the subject in his book about the scroll is famous.3 This is nicely paralleled by the SP interpretation of Lev 17:7 ‫ל ְשּׂ ִע ִירים‬,ַ ‘to the goatdemons,’ as ‫לשערים‬, pronounced as laššrəm, ‘to the gates.’ This attests the loss of the guttural – another piece of common knowledge, amply treated by Ben-Hayyim.4 A similar case of lack of acquaintance with the gutturals is ‫ כמאפכת‬in v. 7, where the scribe omitted the ‫ ה‬of ‫כמהפכת‬, and then the corrector tried to restore it, making a wrong decision by choosing an aleph. Wrong decision, indeed? I rather believe the scribe/corrector applied here the spelling with ‫ א‬in line with the spelling in ch. 13:19: ‫כמאפכת‬ ‫אלהים את סודם ואת עומרה‬. The Muraba῾at fragment of Isaiah displays, however, the ordinary spelling ‫כמהפכת זר]ים‬, with a ‫ה‬. There is reason to believe that the ‫ א‬spelling attests an Aramaic influence, as the latter uses constantly the verb ‫ אפך‬with ‫ א‬as first consonant, when “turning away” in battle is involved. Let me repeat: the original misspelling ‫ כמפכת‬reflects faithfully the actual speech pattern of the word, in a historical period when the glottal consonants were no longer articulated. A conspicuous Aramaism occurs in verse 1. The scribe copied correctly ‫בימי‬, ‘in the days of,’ but the corrector, devoted to his Aramaic education, produced a hybrid creature, namely ‫ביומי‬, just as the Aramaic Targum says at this very spot: ‫ביומי עזיה יותם אחז חזקיה מלכיא דבית יהודה‬. Aramaic is also the factor that implanted the 3rd pers. fem. sg. ending of the perfect ‫ת‬- in ‫ ונתרת‬in verse 8, instead of the ‫ה‬- ending in the masoretic ‫ונותרה‬. And so is the Aramaizing form ‫עצרתה‬, for the masoretic ‫עצרה‬, assembly in v. 13.5 The scribe took the word as the popular name of Pentecost, on which Josephus’ testifies: τῇ πεντηκοστῇ, ἣν Ἑβραῖοι ἀσαρθὰ καλοῦσι 3 4 5

Kutscher 1974, passim; especially 56-58. Ben-Hayyim 2000, 38-43. Kutscher 1974, 24.

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ

245

(Antiquities 3:252). Remarkably, this is the name of the festival in the Mishna.6 In the next chapter, verse 2 displays the pure Aramaic 3rd person masc. sing. pronoun ‫ עלוהי‬instead of the Hebrew ‫עליו‬. Chapter 2:24 has twice the Aramaic preposition ‫תחות‬, ‘instead’, for the Hebrew ‫ ַתּ ַחת‬of the MT, which also occurs twice in the scroll in the same verse. I believe I don’t need to stress the position of Aramaic in the area at the turn of the last century before Christ. In order to appreciate its deep penetration into the Hebrew literature of the epoch, suffice to mention the grammatical structure of Mishnaic Hebrew, which was re-shaped at that time in accordance with the Aramaic system. Among the Aramaisms implanted in the scroll by the careless scribe is the 2nd pers. fem. sg. suffixed pronoun ‫כי‬-. It occurs in our scroll in few cases such as ‫שריכי‬, ‘your princes’ (1:23), ‫עלי לכי‬, ‘get you up’ (“dativus ethicus”) followed by the habitual pronoun ‫ קולך‬a few words further (40:9); ‫גואלכי‬, ‘your savior,’ soon after the regular pronoun ‫ מושיעך‬in 49:26; ‫לנפשכי‬, ‘to yourself,’ a few words before ‫‘ גֵ וֵ ְך‬your back’ in 51:23. Such forms occur in the Elephantine documents.7 Aramaisms are rather frequent in SP. For example the masoretic Exod 15:1 has ‫גָ אֹה גָ ָאה‬, a classical sequence of an absolute infinitive preceding a finite verb, a predicate whose subject is God: ‘He (the Lord) has triumphed’. SP has no absolute infinitive, much like its contemporary Mishnaic Hebrew.8 It features a different structure: ‫גוי גאה‬, pronounced guwwi g, with ‫גוי‬, ‘a people,’ as subject, and ‫ גאה‬as its predicate: ‘a people has struggled.’ The latter, ‫גאה‬, seems to be no other than the perfect ‫ גח‬of the Aramaic verb ‫( גוח‬omitting the guttural according to SH phonetic laws). This calls to mind the frequent ‫אגיח‬, Af῾el of ‫גוח‬, by which the Targumim render as a rule the Hebrew ‫נלחם‬, e.g., Exod 1:10 where Pharaoh’s words ‫ ונלחם בנו‬is rendered in Onqelos as ‫ויגיחון בנא‬, ‘they will combat us.’ 1QIsaa shows a pronounced preference for the “lengthened” 3rd pers. masc. pl. personal pronoun ‫( המה‬v. 2; 38:1; 41:21; 56:11; 61:9 and many others), which implies a vocalic ending. The ending ‫מה‬- is shared by the masoretic text as far as the 3rd pers. masc. pl. is concerned: ‫ה ָמּה‬, ֵ outnumbering the short form ‫ ֵהם‬by about 4o percent.9 The SP is unequivocal in 6 7

8 9

‫ע ֶצ ֶרת‬, ֲ e.g., tract. Shevi᾿it 1:1. Leander 1928, 27-29. The same pronoun occurs occasionally – and rarely – in Late Biblical Hebrew, apparently as a literary ornament: ‫שובי נפשי למנוחיכי כי יהוה גמל עליכי‬, ‘Return, O my soul to your rest, for the Lord has dealt bountifully with you’ (Ps 116:7). Segal 1908, 684. Joüon-Muraoka 2006, §39.

246

ABRAHAM TAL

this respect: imma is the universal pronunciation of the pronoun in the reading of the Tora, even when the spelling has no final he: ‫הם‬, whether independent or suffixed: ‫לאבותיכם‬, pronounced lbūtīkimma(Deut 1:8).10 An interesting case is 44:9: ‫ועדיהמה המה בל יראון‬, ‘and their witnesses do not see,’ with ‫ המה‬as an interlinear addition (of the corrector?). The MT has approximately the same sequence with ‫ המה‬bearing punctaocultantes, deleting dots, on each letter (a masoretic note says: “four words have puncta in the prophets”). This demonstrates that the presence of the pronoun was unstable at this spot in both MT and the scroll, suspected as redundant. The same rule applies to the suffixed pronoun, such as ‫ לכמה‬as against MT ‫( ָל ֶכם‬2:22). The same “long” ending occurs when the pronoun is suffixed to a noun, e.g., ‫ ידיכמה‬in 1:15 – 2nd pers. masc. pl.11 The masoretic Isaiah 1:12 displays an asyndetic connection: ‫י־ב ֵ ֥קּשׁ ֛ז ֹאת‬ ִ ‫ִמ‬ ‫מס ֲח ֵצ ָ ֽרי‬ ֹ ֥ ‫מיֶּ ְד ֶכ֖ם ְר‬, ִ ‘who asked this from you trample my courts?’; the verb ‫ בקש‬governs the infinitive ‫ רמס‬as its object. This is an impossible construction in Mishnaic Hebrew, which uses the infinitive only when preceded by a Lamed, e.g., ‫מבקש ליטע‬, ‘he wishes to plant’ (Kil᾿ayim 3:6); ‫בקשו לשקול דינרים‬, ‘they wanted to pay denarii’ (Sheqalim 2:4).12 In conformity with this Mishnaic rule, the scribe of our scroll added the Lamed: ‫לרמוס חצרי‬. However, the scribe of 4Q60 followed the masoretic type ‫רמס‬. Similarly, our scribe added a Lamed to the infinitive in the following verse 13: ‫לוא תוסיפו להביא ̇מנחת שוא‬, ‘no longer shall you bring vain offerings,’ achieving a Mishnaic-like sequence, e.g., ‫אם הוסיפו ְלנַ ֵטּף‬. ‘If (the rains) continue dripping’ (Machshirin 2:4), etc. The Muraba῾at fragment of Isaiah keeps faithfully the masoretic string ‫לא תוסיפו הביא מנחת שוא‬. One may detect a certain hesitation in the scribe’s approach to the “lengthened imperfect.” On the one hand he reproduces it faithfully: ‫לכו‬ ‫נא ונוכחה‬, ‘let us reason together’ (1:18); ‫ואשיבה שופטיך‬, ‘I shall restore your judges’ (1:26). On the other hand he transforms the “lengthened imperfect” into a regular one: ‫ ואסיר כל בדיליך‬... ‫והשיב ידי עליך‬, ‘I shall turn my hand against you … and I shall remove all your slag’ (1:25) for the masoretic ‫יליִ ְך‬ ֽ ָ ‫ל־בּ ִד‬ ְ ‫אָס ָירה ָכּ‬ ֖ ִ ְ‫אָשׁ ָיבה יָ ִ ֙די ָע ֔ ַליִ ְך וְ ֶא ְצ ֥ר ֹף ַכּ ֖בֹּר ִס ָיג֑ יִ ְך ו‬ ֤ ִ ְ‫ו‬. This indecision reflects the status of the “lengthened imperfect” in the post exilic period. 10

11 12

The isolated form of the independent 2nd pers. fem. pl. ‫ אתן‬has no final vowel (Gen 31:6): wattən. This “long” ending occurs in the fragment 4Q63 as well. Segal 1927, 165-167. Kutscher 1974, 346.

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ

247

Though it is totally absent from Mishnaic Hebrew,13 it has not disappeared from Late Biblical Hebrew. Only this time it is devoid of its original denotation of self entreat (cohortative), such as e.g., ‫אמלטה‬, ‘let me escape’ (Genesis 19:20), but, in Late Biblical Hebrew, as a literary device in narrative context, with no cohortative connotation, e.g., ‫ ואֹמרה‬... ‫ואתפללה‬, ‘and I prayed … and I said’ (Dan 8:15); ‫ואתאבלה‬, ‘and I mourned’ (Nehemia 1:4). In such utterances the form expresses a past action. This seems to be the background of the vacillation between one form to another in the scroll.14 And this is what the SP displays as well. It has ‫ ואברך‬for the masoretic ‫וַ ֲא ָ ֽב ֲר ָכ ֙ה‬, ‘and I shall bless’ (Gen 12:3); ‫ אמלט‬for the masoretic ‫א ָמּ ְל ָ֨טה‬, ִ ‘let me escape’ (Gen 19:20); ‫ ונשאל‬for the MT ‫וַ נִּ שאלה‬, ‘let us ask her’ (Gen. 24:57), all of them in the cohortative, with no formal indication. On the other hand, the “lengthened imperfect” still exists in SP, but with no particular function, merely denoting the past tense. Such is ‫ונלכדה את‬ ‫‘ כל עריו‬and we captured all his towns,’ as against the masoretic ‫וַ נִּ ְלכֹּד‬ (Deut 2:34); ‫ ואשליכה את עפרו‬... ‫ואשרפה אתו באש‬, ‘and I burned it with fire … and I threw its ashes,’ for the masoretic ‫יכה‬ ָ ‫וָ ֶא ְשׂר ֹף אתו באש … וָ ַא ְשׁ ִל‬ ‫( את עפרו‬Deut 9:21).15 The Masoretic verbal system distinguishes carefully between pausal and internal forms of the tenses. At the end of a clause, when disjunctive accents are associated with a verb, the stress is retreated from the ultima syllable to the penultima, and, as a result, the formerly short (or reduced to shewa) vowel becomes long. The pattern ‫ ְיִכ ְתּבוּ‬of the imperfect becomes ‫יכתובו‬, and the imperative ‫ ִכּ ְתבוּ‬changes to ‫כתובו‬.16 Apparently, our scroll represents a linguistic phase in which no such distinction exists any longer, since the penultima stress prevailed, as it did in Aramaic too, and full vowels appear: ‫דרושו משפט‬, ‘seek justice,’ for the masoretic ‫( ִדּ ְרשׁוּ‬Isa 1:17); ‫ותחפורו מהגנות‬, ‘and you shall blush for the gardens’ (1:29), where MT vocalizes ‫ ;וְ ַת ְח ְפּרו‬and so is ‫ותעלולים ימשולו בם‬, ‘and babes shall rule over them’ (3:4), facing the masoretic ‫;יִמ ְשׁלוּ‬ ְ and the imperative fem. sg. ‫אמורי‬ in Isa 40:9 with the vocalic waw added above the line, as against the masoretic ‫א ְמ ִרי‬, ִ ‘say.’ Amazingly, for the masoretic ‫יאדימו כתולע‬, ‘they are red like crimson,’ the scroll has ‫ידומו כתולעת‬, where ‫ ידומו‬is the pausal form of ‫ידמו‬, ‘will be 13 14 15

16

Segal 1927, 72. Qimron 1987, 149-161. This case may also be attributed to the distinct inclination of SP towards harmonization. The regular imperfect ‫ וָ ֶא ְשׂר ֹף‬has been “lengthened” by the force of the following ‫יכה‬ ָ ‫וָ ַא ְשׁ ִל‬. Harmonization is treated below. Joüon-Muraoka 2006, 96-97.

248

ABRAHAM TAL

resemblant to crimson’ (1:18). By this phenomenon, the scroll represents the treatment of the imperfect in all the DSS documents, as for example the second column of the Temple Scroll says: ‫לא תחמודו כסף וזהב‬, ‘you shall not covet neither silver nor gold,’ paraphrasing Deut 7:25 (which has the singular: ‫)לא תחמד‬. Obviously, these forms can be identified in the scrolls only when a materlectionis is used. The indifferent use of so-called pausal forms in the Second Temple period is recorded in Mishnaic Hebrew.17 The problem is that in the process of transmission of the Mishna, the image became blurred. The copyists of Mishnaic manuscripts, well educated in the spirit of Biblical Hebrew, often applied Biblical rules that defaced the grammatical particularities of the manuscripts. In the present case, they omitted the waw that could have revealed the phonology involved. As a result, the famous Kaufmann codex of the Mishna (Kaufmann A50 of the Hungarian Academy) is very poor in such “full” forms when non pausal. And then, when the Naqdan made his work of vocalization, also in the spirit of Biblical Hebrew, these forms nearly disappeared.18 Nearly, because in several places the scribe maintained the waw in nonpausal forms. Such is ‫יַ ֲחלוקו יורשי הבעל עם יורשי האב‬, ‘The heirs of the [levirate] husband and the heirs of [the woman’s] father shall divide it’ (Yevamot 4:3). The Naqdan deleted the waw by puncta ocultantes following biblical grammar, but left untouched the ḥaṭef, contradicting the rule: ‫יַ ֲח ְלקוּ‬. The same is repeated in Ketubot 8:6 with the same procedure, and so in Bava Mesia 8:8 with the same verb: ‫יחלוקו את חודש העיבור‬, ‘let them divide the (money for) the intercalated month.’ Samaritan Hebrew is a penultima-stressed language. Ben-Hayyim has demonstrated many years ago that SH underwent a secondary recession of the stress after a period in which it did not differ from the Tiberian vowel system.19 Accordingly, its imperfect has always a full vowel, where the regular masoretic vocalization has a shewa (a reduced vowel) being ultima stressed. This is barely visible when the consonantal skeleton of the SP is consulted, because the regular vowel of the imperfect of the simple and the doubled stem (qal/piel) is a, for which no mater lectionis exists. It is only the ritual pronunciation that reveals it, e.g., ‫ולא תשקרו איש בעמתו‬ wlatešåqqåruīšbammitu, ‘you shall not lie one to another’ – non-pausal (Lev 19:11), as against the masoretic ‫ולא ְת ַשׁ ְקּרוּ‬. No distinction is made 17 18

19

Segal 1927, 71-71. A good example of a “Tiberianized” manuscript of the Mishna is codex Parma De Rossi 497. See Bar-Asher 2009, 28-69. Ben-Hayyim 2000, 68.

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ

249

between non-pausal and pausal forms, e.g., ‫‘ את חקתי תשמרו‬you shall keep my statutes,’ is pronounced itaqqutitišmru, just as the MT vocalizes the pausal form with its o vowel: ‫תּ ְשׁמֹרוּ‬. ִ

REPLACEMENT OF RARE WORDS AND FORMS So far with grammatical matters which exemplify the tendencies of the Second Temple period. In the following, several instances are presented in which rare words and forms are replaced by more familiar ones. The masoretic collocation ‫וְ ֵאין ֵק ֶצה‬, ‘there is no limit,’ occurs twice in chapter 2:7. It is restricted to this passage and to Nahum 2:10; 2:3 and 3:9. The competitive ‫ אין קץ‬is present in five different places, among which Isaiah 9:6; Job 22:5; Qohelet 4:8, 16; 12:12. ‫ קץ‬alone occurs 62 times in the Bible. It is therefore little wonder that our scribe adopted the more common form, present in Mishnaic Hebrew too: ‫הקּץ‬, ַ ‘it is the end’ (Mishna Eduyot 2:9); ‫הגיע הקץ‬, ‘the the end has come’ (Sifra, Behuqqotay 1:4). This is also the form preferred by 4Q56, though 4Q55 remained dedicated to the masoretic ‫קצה‬. In chapter 1:18 ‫כשני‬, ‘like scarlet’ replaced the scarce masoretic ‫כּ ָשּׁנִ ים‬. ָ The latter occurs just one more time in Proverbs 31:21: ‫יתהּ ָל ֻבשׁ ָשׁנִ ים‬ ָ ‫כּל ֵבּ‬, ָ ‘all her household wears scarlet (garments)’, while ‫ שני‬occurs 42 times in the Bible. The following ‫ כתולעת‬replaces the masoretic ‫תּוֹלע‬ ָ ‫כ‬, ַ ‘as purple’ (1:18). The latter occurs one more time in Lamentations 4:5: ‫תוֹלע ִח ְבּ ֖קוּ‬ ָ ֔ ‫ָה ֱא ֻמנִ ֙ים ֲע ֵל֣י‬ ‫אַשׁ ַפּ ֽתּוֹת‬, ְ ‘those who were brought up in purple, cling to ash heaps,’ while the feminine ‫תולעת‬, in the same sense, in direct connection with ‫ ָשׁנִ י‬occurs 33 times in the Bible. This feminine form occurs in the Mishna and the Tosephta as well, mostly in quotations from the Tora. Being aware of the shorter masculine form, the scribe/corrector utterly deleted the final ‫ת‬, in contrast with 4Q60, which maintains the popular feminine ‫כתולעת‬. This is conspicuous in the SP as well. Loyal to the language of its time, this document gives preference to frequent lexemes, where the masoretic text has infrequent ones. In Genesis 24:21 there is the story about Abraham’s servant gazing at Rachel, as she waters the flock: ‫והאיש משתאה לה‬. The hapax ‫ משתאה‬is written in SP ‫ משתי‬and pronounced mašti, ‘drinks’ as if from the frequent ‫שתה‬, ‘to drink,’ referring to his request: ‫הגמיאיני‬ ‫נא מעט מים‬, ‘give me some water to drink’ (v. 18). The Gen 49:21 hapax ‫שׁ ֶפר‬, ָ believed to denote ‘beautiful’, ‘goodly’, within the collocation ‫אמרי‬ ‫שפר‬, ‘goodly words,’ is replaced with ‫שׁוֹפר‬, ָ ‘trumpet’ which occurs

250

ABRAHAM TAL

5 times in the Pentateuch and 23 times in the Mishna. This reflects the Samaritan understanding of the passage, which incidentally is also a homiletic Jewish interpretation of the word that appears in the Genesis Rabba Midrash: ‫אמרי שופר שהן משפרים אמרים שניתנו בשופר ובתרועה‬, ‘trumpeted words that improve words, stated with trumpet and clarion’ (section 98, §21). In chapter 2:20 MT has a strange sequel: ‫אָדם ֵא ֚ת‬ ֔ ָ ‫הוּא יַ ְשׁ ִ ֣ליְך ָה‬ ֙ ‫ַבּיּ֤ וֹם ַה‬ ‫שׂוּ־לוֹ ְל ִ ֽה ְשׁ ַתּ ֲחוֹ֔ת ַל ְח ֹ֥פּר ֵפּ ֖רוֹת וְ ָל ֲע ַט ֵלּ ִ ֽפים‬ ֙ ‫א ִל ֵיל֣י ַכ ְס ֔פּוֹ וְ ֵ ֖את ֱא ִל ֵיל֣י זְ ָה ֑בוֹ ֲא ֶ ֤שׁר ָ ֽע‬, ֱ “on that day people will throw they idols of silver and their idols of gold which they made for themselves to ‫ ַל ְח ֹ֥פּר ֵפּ ֖רוֹת‬and to the bats”. For ‫עשו‬ ‫לו‬, the scroll has a better syntax: ‫עשו אצבעותיו‬, “his fingers made,” and for the cryptic ‫ל ְח ֹ֥פּר ֵפּ ֖רוֹת‬,ַ the scroll has ‫לחפרפרות‬, in one word, assuming that it is a kind of animal parallel to the following ‫ולעטלפים‬. This is how Jerome rendered the word too: talpas = moles. (LXX: τοῖς ματαίοις = Peshitta: ‫סריקותא‬, both meaning “vanity”). In parenthesis I would like to mention that the “corrections” in the spirit of the MT abound in the scroll, as e.g., ‫( והוכיח בין עמים רבים‬chapter 2:4), where the corrector deleted ‫ בין‬and added a supralinear ‫ ל‬to ‫עמים‬. The original reading of the scribe was attracted by the preceding ‫ושפט בין גויים‬.

HARMONIZATION Harmonization is one of the favorite occupations of the scribe. The somewhat syntactically isolated ‫וּשׁ ָמ ָ ֖מה ְכּ ַמ ְה ֵפּ ַ ֥כת זָ ִ ֽרים‬, ְ ‘it is desolate as the overthrow of strangers,’ which strangely repeats the beginning of the verse: ‫אַר ְצ ֶכ֣ם ְשׁ ָמ ָ֔מה‬, ְ ‘your land is desolate,’ is transformed in the scroll into ‫ושממו עליה‬, ‘they shall be appalled at it’ (1:7). The change is inspired by similar expressions of consternation, e.g., ‫‘ ַכּ ֲא ֶ֨שׁר ָשׁ ְמ ֤מוּ ָע ֶ ֙ל ֙יָך ַר ִ֔בּים‬many were astonished at you’ (Isa 52:14); ‫יכם‬ ֶ֔ ‫א ֵיְב‬ ֹ ֽ ‫יה‬ ָ֙ ‫וְ ָ ֽשׁ ְמ ֤מוּ ָע ֶ ֙ל‬, ‘your enemies shall be appalled at it’ (Lev 26:32). See also Ezek 26:16; 27:35; 28:19. A more lenient passage is achieved by turning the noun ‫ שממה‬into a verb, creating a verbal clause in line with the precedent verbal clause: ‫זרים‬ ‫אוכלים אותה‬, ‘aliens devour your land’. The Isaiah fragment of Muraba῾at follows the masoretic ‫ושממה‬, ‘and devastation.’ In 1:15 the scribe added ‫אצבעותיכם בעאון‬, ‘your fingers (are full) with iniquity’ to the end of the verse, which looks abrupt: ‫ידיכם דמים מלאו‬, ‘your hands are full of blood,’ achieving a perfect parallelismus membrorum. The addition was taken from chapter 59:3, triggered by the presence of the word ‫ כפיכם‬in both verses, and by their striking resemblance:

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ

251

1:15: ‫י־ת ְר ֥בּוּ ְת ִפ ָלּ֖ה ֵא ֶינ֣נִּ י שׁ ֵ ֹ֑מ ַע יְ ֵד ֶיכ֖ם‬ ַ ‫אַע ִ ֤לים ֵעינַ ֙י ִמ ֶ֔כּם ַגּ֛ם ִ ֽכּ‬ ְ ‫יכם‬ ֶ֗ ‫וּב ָפ ִר ְשׂ ֶכ֣ם ַכּ ֵפּ‬ ְ ‫דּ ִ ֥מים ָמ ֵ ֽלאוּ׃‬,ָ ‘When you spread forth your hands, I will hide my eyes from you; even though you make many prayers, I will not listen; your hands are full of blood.’ 59:3: ‫רוּ־שׁ ֶקר ְלשׁוֹנְ ֶכ֖ם‬ ֶ֔ ‫יכ ֙ם ִדּ ְבּ‬ ֶ ‫וֹת‬ ֵ ‫עוֹת ֶיכ֖ם ֶ ֽבּ ָעוֹ֑ ן ִשׂ ְפ ֽת‬ ֵ ‫יכ ֙ם נְ ג ֲֹאל֣ וּ ַב ָ ֔דּם וְ ֶא ְצ ְבּ‬ ֶ ‫ִ ֤כּי ַכ ֵפּ‬ ַ ‘For your hands are defiled with blood and your fingers with ‫עוְ ָ ֥לה ֶת ְה ֶגּֽה‬, iniquity; your lips have spoken lies, your tongue mutters wickedness.’

However, 4Q60 follows the masoretic text ending in ‫מ ֵ ֽלאוּ‬. ָ We may encounter a similar textual intervention in 3:24, where the MT looks abrupt: ‫גוֹרה נִ ְק ָפּ ֙ה וְ ַ֨ת ַחת ַמ ֲע ֶ ֤שׂה ִמ ְק ֶשׁ ֙ה ָק ְר ָ֔חה וְ ַ ֥ת ַחת‬ ֤ ָ ‫וְ ָהיָ ֩ה ַ֨ת ַחת ֜בֹּ ֶשׂם ַ ֣מק יִ ְֽה ֶ֗יה וְ ַ֨ת ַחת ֲח‬ ‫י־ת ַחת ֹֽי ִפי‬ ֖ ַ ‫ְפּ ִת ִיג֖יל ַמ ֲח ֹ֣ג ֶרת ָ ֑שׂק ִכּ‬

The more or less accepted understanding of this verse, rich in hapaxlegomena, is: ‘Instead of perfume, there shall be rot; And instead of an apron, a rope; Instead of a diadem of beaten-work, A shorn head; Instead of a rich robe, A girding of sackcloth; A burn instead of beauty’ (JPS).

Translators and commentators had much trouble with the end ‫י־ת ַחת‬ ֖ ַ ‫ִכּ‬ ‫י ִֹֽפי‬, and, in order to maintain the principle of parallelismus membrorum dictated by the four precedent phrases, they where compelled to treat the otherwise conjunction ‫ ִכּי‬as if it was a noun, albeit mutilated: ‫כּוִ י‬, ְ derived of ‫כוה‬, ‘to burn.’ Remarkably, the Septuagint ignored the phrase altogether. Our scroll perceives ‫ ִכּי‬as the ordinary conjunction having a conclusive denotation, and, harmonizing the phrase with the preceding ones, adds: ‫כי‬ ‫תחת יפי בשת‬, ‘for instead of beauty – shame.’ The casus adverbialis in 1:20 ‫ח ֶרב ְתּ ֻא ְכּ ֔לוּ‬, ֣ ֶ ‘you will be devoured by sword,’ is rendered as ‫בחרב תאכלו‬, which, by the addition of betinstrumentalis, equalizes the form with 3:25: ‫בחרב יפולו‬, ‘they will fall by sword’ (MT: ‫)בּ ֶ ֣ח ֶרב יִ ֹ֑פּלוּ‬. ַ However, the scribe was not alert enough in verse 7, where he left the casusadverbialis unchanged ‫שרופות אש‬, ‘burn by fire.’ Like our scroll, The SP is very rich in harmonizations. The masoretic Gen. 3:16 has two different nominal patterns for labour pangs: ‫אַר ֶבּ ֙ה‬ ְ ‫ַה ְר ָ ֤בּה‬ ‫בוֹנ ְך וְ ֵ ֽהר ֵֹ֔נְך ְבּ ֶ ֖ע ֶצב ֵ ֽתּ ְל ִ ֣די ָב ִנ֑ים‬ ֣ ֵ ‫ע ְצּ‬, ִ ‘I will greatly increase your pain and your childbearing; in pain you shall bring forth children,’ which SP equalized promptly: ‫בעצבון תלדי בנים‬. In Gen. 31:33 Zilpah and Bilhah, Jacob’s two maids, are called in MT ‫ ֲא ָמהוֹת‬while in all other places they are named ‫שׁ ָפחוֹת‬. ְ SP removes the difference by using ‫ ְשׁ ָפחוֹת‬in 31:33 as well. The only exception is 30:3, where Bilhah is characterized as ‫אמתי‬, ‘my maid,’

252

ABRAHAM TAL

but this instance has escaped the harmonizer’s eye. In Gen. 7:2 the masoretic text uses an uncommon expression where males and females regard animals: ‫ן־ה ְבּ ֵה ָ֡מה‬ ַ ‫וּמ‬ ִ ‫ח־לָך֛ ִשׁ ְב ָ ֥עה ִשׁ ְב ָ ֖עה ִ ֣אישׁ וְ ִא ְשׁ ֑תּוֹ‬ ְ ‫הוֹרה ִ ֽתּ ַקּ‬ ֗ ָ ‫ִמ ֣כֹּל ַה ְבּ ֵה ָ ֣מה ַה ְטּ‬ ‫א ֶשׁר ֣ל ֹא ְטה ָ ֹ֥רה ִ ֛הוא ְשׁ ַנ֖ יִ ם ִ ֥אישׁ וְ ִא ְשׁ ֽתּוֹ׃‬. ֲ ֠ SP replaces ‫איש ואשתו‬, ‘man and his wife’ (twice) with the more adequate ‫זכר ונקבה‬, ‘male and female,’ in accordance with the following verse: ‫ַגּ֣ם ֵמ ֧עוֹף ַה ָשּׁ ַ ֛מיִם ִשׁ ְב ָ ֥עה ִשׁ ְב ָ ֖עה זָ ָכ֣ר וּנְ ֵק ָ ֑בה‬ (see also 6:19). And, of course, in accordance with lexical logic. Speaking about logic, one may observe the formal incongruence in the masoretic Isaiah 1:23: ‫סוֹר ִ ֗רים וְ ַח ְב ֵ ֙רי גַּ נָּ ִ֔בים ֻכּ ֙לּוֹ א ֵ ֹ֣הב ֔שׁ ֹ ַחד וְ ר ֵ ֹ֖דף ַשׁ ְלמ ִֹנ֑ים‬ ְ ‫ָשׂ ַ ֣ריִ ְך‬ ‫יהם‬ ֽ ֶ ‫א־יָבוֹא ֲא ֵל‬ ֥ ֹ ‫אַל ָמ ָנ֖ה ֽל‬ ְ ‫יָתוֹם ֣ל ֹא יִ ְשׁ ֔ ֹפּטוּ וְ ִ ֥ריב‬, ֙ ‘Your princes are rebels and companions of thieves. Everyone loves bribe and runs after gifts. They do not defend the orphan, and the widow’s cause does not come before them.’ The plural subject, “your princes” has indeed plural predicates: ‫סוֹר ִ ֗רים‬ ְ ‫וְ ַח ְב ֵר֙י גַּ נָּ ִ֔בים‬, “rebels and companions of thieves,” and ‫יָתוֹם ֣ל ֹא יִ ְשׁ ֔ ֹפּטוּ‬, ֙ “do not defend the orphan,” but between these plurals, distant from each other, the appositional clause is in the singular: ‫כּ ֙לּוֹ א ֵ ֹ֣הב ֔שׁ ֹ ַחד וְ ר ֵ ֹ֖דף ַשׁ ְלמ ִֹנ֑ים‬. ֻ The Scroll mends this incongruence by pluralizing it: ‫כולם אוהבי שוחד רודפי‬ ‫שלמונים‬. And so it does in 2:18: ‫והאלילים כליל יחלופו‬, ‘and the idols shall utterly pass away,’ to resolve the incongruence of MT: ‫ילים ָכּ ִ ֥ליל יַ ֲח ֹֽלף‬ ֖ ִ ‫וְ ָה ֱא ִל‬. SP is very much engaged in plural vs. singular matters. Thus in the masoretic Gen. 38:25, Tamar, accused of harlotry, produces the proof of Yehudah’s fatherhood: ‫ילים וְ ַה ַמּ ֶ ֖טּה‬ ֛ ִ ‫ל ִ֞מי ַהח ֶ ֹ֧ת ֶמת וְ ַה ְפּ ִת‬,ְ ‘to whom belong the signet, the cords and the staff?’. SP remembers that in verse 18 she asked for one cord as pledge: ֖‫וּמ ְטָּך‬ ַ ‫וּפ ִת ֔ ֶילָך‬ ְ ‫ח ָ ֹֽת ְמ ָ֙ך‬, ‘your signet, and your cord and your staff.’ Accordingly, SP puts the cord in the singular: ‫והפתיל‬. Moreover, MT has the masculine ‫ חותם‬in v. 18 and its synonym ‫ החותמת‬in v. 25. SP does not allow such mixture, featuring ‫ חתים‬in both verses.20 In Gen 41:8 Pharaoh relates his dream to the magicians: ‫וַ יְ ַס ֨ ֵפּר ַפּ ְר ֤עֹה ָל ֶה ֙ם‬ ‫אוֹתם ְל ַפ ְר ֽעֹה‬ ֖ ָ ‫ין־פּוֹתר‬ ֵ֥ ‫ֹלמוֹ וְ ֵא‬ ֔ ‫ת־ח‬ ֲ ‫א‬, ֶ ‘Pharaoh told them his dream, but there was no one who could interpret them to Pharaoh,’ with ‫אוֹתם‬ ֖ ָ (plural pronoun) acting as an anaphoric reference to ‫ֹלמוֹ‬ ֔ ‫( ֲח‬singular), despite the fact that verses 1-7 relate about two consecutive dreams. SP solves the contradiction by putting the dream in v. 8 in the plural: ‫חלמיו‬, ‘his dreams.’ In the opposite direction, MT features an unjustified plural related to Jacob in Gen 46:7, where the displacement of the whole family is described: ‫יְמה‬ ָ ‫וּבנ֥ וֹת ָבּ ָנ֖יו וְ ָכל־זַ ְר ֑עוֹ ֵה ִ ֥ביא ִא ֖תּוֹ ִמ ְצ ָ ֽר‬ ְ ‫וּב ֵנ֤י ָבנָ ֙יו ִא ֔תּוֹ ְבּנ ָ ֹ֛תיו‬ ְ ‫בּ ָ֞ניו‬, ָ ‘his sons, and his sons’ sons with him, hisdaughters, and his sons’ daughters; 20

SP differentiates between our signet, ‫חתים‬, and the ornamental carving ‫ חותם‬in Exod 28:11, 21, 36, etc. MT has ‫ ח ָֹתם‬for both.

DO THE SAMARITAN PENTATEUCH AND 1QISAᵃ

253

all his offspring he brought with him into Egypt.’ The plural ‫ ְבּנ ָ ֹ֛תיו‬contradicts the fact, related in v. 15, that Jacob had only one daughter: ‫ֵ ֣א ֶלּה ְבּ ֵנ֣י‬ ‫לאָ֗ה ֲא ֶ֨שׁר ָי ְֽל ָ ֤דה ְליַ ֲעק ֹ֙ב ְבּ ַפ ַ ֣דּן ֲא ָ ֔רם וְ ֵ ֖את ִדּ ָינ֣ה ִב ֑תּוֹ‬.ֵ SP promptly replaced the plural with the singular: ‫בנתו‬, distinctly pronounced as such: bintu.

CONCLUSION Within the short span of time allowed to a conference talk, I tried to pick some examples that show how both documents follow the same model of – so to say – “improving” an ancient text no longer adequate to the linguistic, stylistic, and also textual logic of the contemporary environment. What was “en vogue” in the eighth century BC could not satisfy the demands of a generation living many centuries later, in a different atmosphere. They made efforts to avoid inappropriate constructions – with limited success. Both transmitters of SP and of the Isaiah scroll were only partly alert, skipping over many such cases, apparently by lack of attention and careless transmission. Or, perhaps, because both transmitters considered only a part of the problematic spots as neuralgic points worthy of redactional intervention. Is this a proof that, in essence, ideological matters excluded, they had a masoretic-like source as prototype?

BIBLIOGRAPHY BAR-ASHER, M. 2009. StudiesInMishnaicHebrew, I-II, Jerusalem. BEN-HAYYIM, Z. 1988. Tibåt Mårqe, a Collection of Samaritan Midrashim, Edited,TranslatedandAnnotated, Jerusalem (in Hebrew). —. 2000. AGrammarofSamaritanHebrew, Winona Lake, Indiana. COWLEY, A. E. 1909. TheSamaritanLiturgy, I-II, Oxford. JOÜON, P., MURAOKA, T. 2006. AGrammarofBiblicalHebrew, Roma. KUTSCHER, E. Y. 1974. TheLanguageandLinguisticBackgroundoftheIsaiah Scroll, Leiden. LEANDER, P. 1928. Laut-undFormenlehredesÄgyptisch-Aramäischen, Göteborg. QIMRON, E. 1987. “Consecutive and Conjunctive Imperfect. The Form of the Imperfect with waw in Biblical Hebrew”, JQR 77 (1987), 149-161. SEGAL, M. H. 1908 “Mishnaic Hebrew and its Relation to Biblical Hebrew”, JQR 20 (1908), 647-737. —. 1927. AGrammarofMishnaicHebrew, Oxford.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH Michael LANGLOIS

Seventy years after their discovery, the Dead Sea Scrolls continue to shed light on the Samaritan Pentateuch. In this contribution, I would like to focus on palaeography and explore two ways in which the study of the scribal hands that copied the scrolls may be relevant to the history of the Samaritan Pentateuch (⅏). In the first part, we will examine the scripts of biblical Dead Sea Scrolls allegedly close to ⅏ and date their copy on the basis of their palaeographical features. These dates will be used as boundaries for various stages of the redaction history of ⅏ or its ancestors. In the second part, we will examine the so-called “Palaeo-Hebrew” script used by a few Dead Sea Scrolls. Taking into account other attestations of this script, we will outline its development and conclude as to the history of the Samaritan script.

1. THE SCRIPTS

DEAD SEA SCROLLS RELATED SAMARITAN PENTATEUCH

OF BIBLICAL THE

TO

In their recent synthesis, Anderson and Giles1 mention the following Dead Sea Scrolls as preserving a possible pre-Samaritan text: 4Q17 (4QExod-Levᶠ); 4Q22 (4QpaleoExodᵐ); 4Q27 (4QNumᵇ); 4Q158 and 4Q364–367 (4QRP); and perhaps 4Q26 (4QLevᵈ). They also refer to 4Q37 (4QDeutʲ), 4Q38 (4QDeutᵏ¹), and 4Q41 (4QDeutn), whose relation to ⅏ is debated. I will refrain, at this stage, from discussing the alleged relationship of these manuscripts to ⅏, but rather focus on their palaeographical features. I will also add to this list 4Q45 (4QpaleoDeutʳ), which is one of the closest scrolls to ⅏ for the book of Deuteronomy, according 1

Robert T. Anderson and Terry Giles, TheSamaritanPentateuch:AnIntroductiontoIts Origin,History,andSignificanceforBiblicalStudies, Resources for Biblical Study 72 (Atlanta: Society of Biblical Literature, 2012), 46–47.

256

MICHAEL LANGLOIS

to Ziemer.2 Unless stated otherwise, the analyses presented below are based on the examination of high-resolution photographs provided by the Israel Antiquities, either on their dedicated website3 or directly to me upon special request.4 I refrained from consulting previous palaeographical studies of these scrolls in order to avoid any bias. My conclusions were then (and only then) compared to previous studies, and their differences (if any) are briefly discussed below. The thirteen manuscripts will now be presented according to their inventory numbers. 4Q17 (4QExod-Levᶠ)5 The script of this manuscript is rather naïve. The calamus is not beveled, and the letters are quite big, with an average caliber of 2.5±0.5 mm. Several features indicate a pre-Hasmonaean period: the absence of medial forms for ‫ כ‬and ‫נ‬, the narrow ‫ל‬, etc. Compared to an ostracon from Maresha dated to 136 of the Seleucid era (i.e. 176 BCE),6 some letters seem to be drawn with an older ductus, such as the ‫ ת‬with a longer left leg and the absence of final ‫ ם‬and medial ‫נ‬. I conclude that this manuscript was probably copied around the second half of the third century BCE. Cross dates it “to the mid-third century CBE”,7 which is possible, but I would not go much earlier. 4Q22 (4QpaleoExodᵐ) This manuscript is one of the few Dead Sea Scrolls that have been copied using the so-called “Palaeo-Hebrew” script; it will therefore be dealt with in the second part of this study. As we will see, it may have been copied sometime in the second half of the second century BCE or during the early first century BCE.

2 3 4

5

6

7

See his contribution to this volume pp. 127ff. The Leon Levy Dead Sea Scrolls Digital Library, http://www.deadseascrolls.org.il/. I want to thank Pnina Shor, Oren Ableman, Orit Rosengarten-Kuslansky, Shai Halevi and Yael Barschak for their kindness and support. Frank Moore Cross, “17. 4QExod-Levᶠ,” in QumranCave4.VII.GenesistoNumbers, Discoveries in the Judaean Desert XII (Oxford: Clarendon Press, 1994), 133– 44. Esther Eshel and Amos Kloner, “An Aramaic Ostracon of an Edomite Marriage Contract from Maresha, Dated 176 B.C.E.,” Israel Exploration Journal 46.1/2 (1996): 1–22. Cross, “17. 4QExod-Levᶠ,” 134.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

257

4Q26 (4QLevᵈ)8 4Q26 is one of the scrolls whose relation to ⅏ is debated.9 It is preserved by a few damaged fragments, which complicates the palaeographical analysis. The hand is confident; the script is formal, spacious, regular. ‫ה‬ features a thick protruding traverse; the top of ‫ ל‬is thickened, and some ‫ ל‬have an angular hook; ‫ ג‬is symmetrical; ‫ ב‬is drawn without lifting the pen; ‫ ת‬features an angular foot. Overall, no clear Herodian features can be detected. I conclude that this scroll was copied sometime in the middle of the first century BCE. The DJD editor does not discuss the palaeographical features, whereas Eshel qualifies the script as “early Herodian” to be dated “between 30 BCE and 20 CE,” pointing similarities with 4QNumᵇ.10 But as we will see below, 4QNumᵇ is more at home in the mid-first century BCE or shortly after; I believe it may also be the case with 4QLevᵈ, though its fragmentary condition prevents a more accurate dating. 4Q27 (4QNumᵇ)11 The hand of the scribe who copied this scroll is skilled, confident, and consistent. ‫ י‬and ‫ ו‬are penned in a similar way and can easily be confused, which is typical of first century BCE scripts. The head of ‫ ל‬is thickened. ‫ ב‬is drawn without lifting the pen, as opposed to the two-step ductus (with a left-to-right base protruding to the right) that develops in the Herodian period. Final ‫ ם‬features a triangular ornament at the left, which suggests further development from Hasmonaean hands. ‫ א‬sometimes has a tick at the bottom of the left leg as well as that of the diagonal; this is also indicative of a late Hasmonaean to early Herodian period. ‫ג‬, on the contrary, features symmetrical legs, unlike the later Hasmonaean shape. I conclude that this scroll was copied around the mid-first century BCE, preferably in the third quarter, but not much later. Jastram mentions Cross’ dating between 30 BCE to 20 CE, but adds that in a personal communication Cross expressed a “preference for the earlier portion of that range,” a 8

9 10

11

Emanuel Tov, “26. 4QLevd,” in QumranCave4.VII:GenesistoNumbers, Discoveries in the Judaean Desert XII (Oxford: Clarendon Press, 1994), 193–95. See, in this volume, the contribution by Michaël van der Meer pp. 41ff. Esther Eshel, “4QLevd: A Possible Source for the Temple Scroll and MiqṣatMa῾aśe Ha-Torah,” DSD 2.1 (1995): 1. Nathan Jastram, “27. 4QNumb,” in QumranCave4.VII.GenesistoNumbers, Discoveries in the Judaean Desert XII (Oxford: Clarendon Press, 1994), 205–67.

258

MICHAEL LANGLOIS

conclusion that Jastram adopts.12 Not only was Cross right in preferring the earlier portion of his range, but I believe that the third quarter would even be better given the relatively few Herodian features. 4Q37 (4QDeutj)13 This is one of the scrolls whose relation to ⅏ is debated and complicated by its composite nature.14 It was copied by a trained scribe in a formal script; several Herodian features can be observed: the upper arm of ‫ א‬is sometimes ornamented; ‫ ב‬is penned in two steps, with the base drawn from left to right and protruding to the right; ‫ ג‬is made up of a slanted vertical and an oblique left leg. ‫ ד‬features two horns: the left horn is a slightly slanted tick, while the right horn is sometimes drawn as a triangular loop joining the traverse and the leg. ‫ ל‬has a fully developed hook, but little or no upper tick. ‫ ק‬features a triangular loop joining the vertical and the traverse. All these features are consistent with a typical developed Herodian formal script. Few late Herodian features are detected, which leads me to conclude that this scroll was probably copied in the first half of the first century CE, preferably around the second quarter. Cross dates it to ca. 50 CE, which is possible, but I want to emphasize that the script is not as typologically “late” as other late Herodian scrolls.15 4Q38 (4QDeutᵏ¹)16 4Q38 is one of the scrolls whose relation to ⅏ is debated and complicated by its composite nature.17 The DJD edition consists of one large fragment and four small fragments. My analysis is based on frag. 2, which was copied by a somewhat inconsistent hand; the strokes are not 12 13

14

15

16

17

Jastram, “27. 4QNumb,” 211. Julie Ann Duncan, “37. 4QDeutj,” in QumranCave4.IX.Deuteronomy,Joshua,Judges, Kings, ed. Eugene C. Ulrich et al., Discoveries in the Judaean Desert XIV (Oxford: Clarendon Press, 1995), 75–91. See the contributions by Benjamin Ziemer and Emanuel Tov in the present volume, pp. 127ff and 19ff respectively. Frank Moore Cross, “The Development of the Jewish Scripts,” in TheBibleandthe AncientNearEast.EssaysinHonorofWilliamFoxwellAlbright, ed. G. Ernest Wright (London: Routledge & Kegan Paul, 1961), 139. Julie Ann Duncan, “38. 4QDeutk1,” in Qumran Cave 4. IX. Deuteronomy, Joshua, Judges, Kings, ed. Eugene C. Ulrich et al., Discoveries in the Judaean Desert XIV (Oxford: Clarendon Press, 1995), 93–98. See the contributions by Benjamin Ziemer and Emanuel Tov in the present volume, pp. 127ff and 19ff respectively.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

259

always very precise, and the ductus can vary. The script is semiformal and exhibits cursive features as well as a tendency to add curved ornamentations (see for instance ‫ע‬, ‫ש‬, sometimes ‫א‬, and so on). ‫ ל‬features a rounded hook of varying size, sometimes almost closed. ‫ ה‬features a double traverse, while ‫ ש‬has a descending left stroke. These features can already be observed in semicursive scripts in the Hasmonaean period, and it is difficult to know when they impregnated formal hands. But since no typical Herodian development can be observed, I conclude that this fragment was likely copied in the second half of the first century BCE. The DJD editor dates it to ca. 30–1 BCE,18 which is possible, though I would not exclude a slightly earlier date due to the influence of semicursive ductus. 4Q41 (4QDeutn)19 The relation of this beautiful scroll to ⅏ is likewise complicated by its composite nature.20 It contains two sheets; the first one contains one column, while the second contains five columns. They have been copied by the same scribe. The script is quite small, partially regular; the size of final ‫ ך‬and ‫ ם‬are reminiscent of a second-century BCE ductus, but most features are otherwise in line with first-century Hasmonaean scripts. ‫ ע‬is quite developed, as is ‫ י‬which can easily be confused with ‫ל‬. ‫ ו‬features thickening at its top by means of a forward loop. The left leg of ‫ א‬sometimes joins the diagonal at its top, but is not drawn together with it. ‫ ב‬is drawn without lifting the pen. ‫ ת‬features a sometimes longer right leg, especially (but not exclusively) in final position. All these characteristics lead me to conclude that this manuscript was copied around the first half of the first century BCE, preferably around the second quarter. The DJD editor dates it to ca. 30–1 BCE,21 but she is mislead in her palaeographical analysis: ‫ א‬does not use “the ‘inverted-v’ form”; when the left stroke joins the diagonal at its top, it is nonetheless 18 19

20

21

Duncan, “38. 4QDeutk1,” 94. Sidnie White Crawford, “41. 4QDeutn,” in Qumran Cave 4. IX. Deuteronomy, Joshua, Judges,Kings, ed. Eugene C. Ulrich et al., Discoveries in the Judaean Desert XIV (Oxford: Clarendon Press, 1995), 93–98. Elizabeth Owen, “4QDeutn: A Pre-Samaritan Text?,” DeadSeaDiscoveries 4.2 (1997): 162–78; Sidnie White Crawford, “A Response to Elizabeth Owen’s ‘4QDeutn: A PreSamaritan Text?,’” DeadSeaDiscoveries 5.1 (1998): 92–94. See also the contributions by Benjamin Ziemer and Emanuel Tov in the present volume, pp. 127ff and 19ff respectively. White Crawford, “41. 4QDeutn,” 117–18.

260

MICHAEL LANGLOIS

drawn separately. Likewise, the base of ‫ ב‬is not “penned from left to right.” I agree with her that ‫ ו‬and ‫ י‬can sometimes be confused, which is usually indicative of the second half of the first century BCE, but I would not consider it “a sure sign of an early Herodian hand.” My preference for the second quarter rather than the first is in part due to this confusion of ‫ ו‬and ‫י‬, but the last third of the century is typologically too late. 4Q45 (4QpaleoDeutr) This manuscript is one of the few Dead Sea Scrolls that have been copied using the so-called “Palaeo-Hebrew” script; it will therefore be dealt with in the second part of this study. As we will see, it may have been copied sometime in the second half of the second century BCE or during the early first century BCE. 4Q158 (4QRPᵃ)22 The hand of the scribe who copied 4Q158 is skilled and confident. Few Herodian features can be observed, such as the forward triangular thickening on top of ‫ ל‬that develops from the late Hasmonaean thickening and anticipates the late Herodian tick. The ductus of ‫ מ‬with a left oblique tick rather than a loop is also indicative. ‫ ג‬features an asymmetrical shape, but other shapes are more conservative: ‫א‬, ‫ע‬, final ‫ך‬, etc. I conclude that this scroll was probably copied around the last third of the first century BCE. Allegro does not discuss the palaeography of this scroll; in his review article, Strugnell qualifies the hand as “formelle, hérodienne ou légèrement préhérodienne.”23 The conservative shapes of some letters, which I mentioned above, probably led him to consider a possible pre-Herodian dating; but the presence of more developed shapes favors the early Herodian period. 4Q364 (4QRPᵇ)24 As I worked on the upcoming volume GleaningsfromtheCaves, I studied a small fragment labeled MS 5439/1.25 Very few letters were attested, 22

23

24

25

John Marco Allegro, Qumrân Cave 4. I (4Q158–4Q186), Discoveries in the Judaean Desert of Jordan V (Oxford: Clarendon Press, 1968), 1–6. John Strugnell, “Notes en marge du volume V des « Discoveries in the Judaean Desert of Jordan »,” RevuedeQumrân 7.2 (26) (1970): 168. Emanuel Tov and Sidnie White, “364. Reworked Pentateuchb,” in Qumran Cave 4. VIII.ParabiblicalTexts,Part1, Discoveries in the Judaean Desert XIII (Oxford: Clarendon Press, 1994), 197–254. Michael Langlois, “Palaeographical Analysis of the Dead Sea Scrolls in The Schøyen Collection,” in Gleanings from the Caves. Dead Sea Scrolls and Artefacts from The

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

261

but they all were consistent with a second-century BCE palaeographical dating. Later, this fragment was identified as belonging to 4Q364. I was thus able to study the shape of letters that were not attested on the fragment; some of them were slightly more developed, which led me to conclude that this scroll was copied in the second half of the second century BCE. The date suggested by the editors of 4Q364, “late Hasmonean or transitional,”26 is too late. Puech wrongly attributed this fragment to 4Q1, but his dating “de la deuxième moitié du IIe siècle avant J.-C.”27 is correct. 4Q365 (4QRPᶜ)28 The script of this manuscript is naïve and semi-formal. The pen is not beveled; the scribe likes refined, recurved strokes. ‫ ש‬has an angled right stroke; ‫ ע‬features an elbow and horizontal base; ‫ ג‬is symmetrical (unlike the later Herodian ductus); ‫ פ‬has a large curled head; ‫ ל‬has a thickened top. ‫ ב‬is sometimes penned without lifting the pen, with a concave rightto-left base, and sometimes in the new two-step ductus; ‫ ו‬and ‫ י‬are both short and can easily be confused. I conclude that the scroll was copied around the second half of the first century BCE. I agree with the DJD editors who qualify the script as “transitional between the late Hasmonaean and early Herodian periods.”29 This scroll may have been tested for radiocarbon dating:30 according to Doudna, the “Pentateuchal paraphrase” tested in Zurich31 is an “Additional

26 27

28

29 30

31

SchøyenCollection, ed. Torleif Elgvin, Kipp Davis, and Michael Langlois, Library of Second Temple Studies 71 (London: Bloomsbury T&T Clark, 2016), 81–82. Tov and White, “364. Reworked Pentateuchb,” 201. Émile Puech, “Un nouveau fragment 7a de 4QGn-Exa = 4QGn-Ex 1 et quelques nouvelles lectures et identifications du manuscrit 4Q1,” RevuedeQumrân 25.1 (2011): 105. Emanuel Tov and Sidnie White, “365. Reworked Pentateuchᶜ,” in Qumran Cave 4. VIII.ParabiblicalTexts,Part1, Discoveries in the Judaean Desert XIII (Oxford: Clarendon Press, 1994), 255–318. Tov and White, “365. Reworked Pentateuchᶜ,” 260. Two sets of radiocarbon tests were performed on Dead Sea Scrolls: Georges Bonani et al., “Radiocarbon Dating of Fourteen Dead Sea Scrolls,” Radiocarbon 34.3 (1992): 843–49; Timothy A. J. Jull et al., “Radiocarbon Dating of Scrolls and Linen Fragments from the Judean Desert,” Radiocarbon 37.1 (1995): 11–19. See recently Johannes van der Plicht and Kaare L. Rasmussen, “Radiocarbon Dating and Qumran,” in Holistic Qumran:Trans-DisciplinaryResearchofQumranandtheDeadSeaScrolls.Proceedings of the NIAS-Lorentz Center Qumran Workshop, 21-25 April 2008, ed. Jan Gunneweg, Annemie Adriaens, and Joris Dik, Studies on the Texts of the Desert of Judah 87 (Leiden: Brill, 2010), 99–121. Bonani et al., “Radiocarbon Dating of Fourteen Dead Sea Scrolls,” 845.

262

MICHAEL LANGLOIS

Frg. 3, assigned to 4Q365 (?)”;32 it has been dated 2139±32 BP. I used the calibration data set IntCal 13.14c to calibrate this result: intcal13.14c 2500 Sample ID 2044+/-65 1 and 2 sigma

2400

2300

2200

Radiocarbon Age

2100

2000

1900

1800

1700

1600 400

300

200

100

0

100

200

300

cal BC/AD

The 1σ (68.3 % confidence) ranges are 344-324 BCE, 205-148 BCE, or 141-112 BCE. Even the latest range (141-112 BCE) seems too early. The 2σ (95.4 % confidence) ranges are 353-294 BCE, 229-219 BCE, 213-84 BCE, or 80-55 BCE. The latest range is possible, in which case some of the palaeographical features previously thought to appear in the second half of the first century BCE should now considered to have appeared earlier. But other explanations are possible: (1) the sample may have been contaminated; (2) the calibration curve should be corrected; (3) the fragment that was tested does not belong to 4Q365,33 in which case my new dating for 4Q364(4QRPᵇ) would perfectly fit the third 1σ range. 32

33

Greg Doudna, “Dating the Scrolls on the Basis of Radiocarbon Analysis,” in TheDead Sea Scrolls after Fifty Years: A Comprehensive Assessment, Volume One, ed. Peter W. Flint and James C. VanderKam (Leiden / Boston / Köln: Brill, 1998), 468. See Emanuel Tov, ed., TheTextsfromtheJudaeanDesert.IndicesandanIntroductionto theDiscoveriesintheJudaeanDesertSeries, Discoveries in the Judaean Desert XXXIX (Oxford: Clarendon Press, 2002), 366 n. 48.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

263

4Q366 (4QRPᵈ)34 The hand of the scribe who copied this manuscript is confident; the script is simple and semi-formal. The narrow ‫ ל‬with no wide hook and thickening at the top is at home in the Hasmonaean period. The partially developed ‫ ע‬is also at home in the Hasmonaean period, and tends to become more angular, yet not as much as later Herodian hands. ‫ ג‬is almost symmetrical; ‫ א‬is drawn in three strokes; ‫ ב‬is drawn without lifting the pen, unlike the later two-step ductus; ‫ י‬has a nice triangular shape; final ‫ך‬ is quite simple, semi-cursive. I conclude that 4Q366 was copied around the second half of the first century BCE. I agree with the DJD editors who qualify the script as “a late Hasmonaean formal hand.”35 4Q367 (4QRPᵉ)36 The hand of the scribe who copied this manuscript is confident and consistent. The calamus is not beveled; the script is semiformal. ‫ א‬is drawn in three steps, and not according to the new two-step ductus. ‫ ב‬is drawn in one step, and not according to the new two-step ductus. ‫ ע‬is quite simple and unsophisticated, unlike later ductus. Overall, this is a typical developed Hasmonaean script without any of the new features that are commonly associated with the Herodian period. I conclude that this scroll was copied around the first half of the first century BCE. The DJD editors date the script to the “mid- or lateHasmonaean”37 period, that is, “125-50 BCE” according to the DJD index volume.38 My conclusion is in line with their dating. Synthesis The dates derived from the palaeographical analysis of the thirteen manuscripts under consideration can be summed up as follows: 34

35 36

37 38

Emanuel Tov and Sidnie White, “366. 4QReworked Pentateuchd,” in QumranCave4. VIII. Parabiblical Texts, Part 1, Discoveries in the Judaean Desert XIII (Oxford: Clarendon Press, 1994), 335–43. Tov and White, “366. 4QReworked Pentateuchd,” 336. Emanuel Tov and Sidnie White, “367. 4QReworked Pentateuche,” in QumranCave4. VIII. Parabiblical Texts, Part 1, Discoveries in the Judaean Desert XIII (Oxford: Clarendon Press, 1994), 367. Tov and White, “367. 4QReworked Pentateuche,” 346. Tov, TheTextsfromtheJudaeanDesert.IndicesandanIntroductiontotheDiscoveries intheJudaeanDesertSeries, 358.

264

MICHAEL LANGLOIS

2nd half of 3rd c. BCE 2nd half of 2nd c. BCE 2nd half of 2nd c. BCE or early 1st c. BCE 1st half of 1st c. BCE 1st half of 1st c. BCE (pref. 2nd quarter) mid-1st c. BCE mid-1st c. BCE (pref. 3rd quarter) 2nd half of 1st c. BCE

end of 1st c. BCE 1st half of 1st c. CE (pref. 2nd quarter)

4Q17 (4QExod-Levᶠ) 4Q364 (4QRPᵇ) 4Q22 (4QpaleoExodᵐ) 4Q45 (4QpaleoDeutʳ) 4Q367 (4QRPᵉ) 4Q41 (4QDeutn)? 4Q26 (4QLevᵈ)? 4Q27 (4QNumᵇ) 4Q365 (4QRPᶜ) 4Q366 (4QRPᵈ) 4Q38 (4QDeutᵏ¹)? 4Q158 (4QRPᵃ) 4Q37 (4QDeutʲ)?

These dates call for a few observations: (1) The so-called “Proto-Samaritan”39 or “pre-Samaritan” text-type is already attested in the third century BCE with 4Q17 (4QExod-Levᶠ). Cross even states that it is, “along with 4QSamᵇ, the earliest of the manuscripts found in the caves of Qumran.”40 Of course this does not mean that the pre-Samaritan text-type is the earliest, as there may have been manuscripts of other text-types that are now lost; as a matter of fact, the second part of the present essay will suggest that there are, indeed, older biblical manuscripts among the Dead Sea Scrolls. But this date confirms that the textual character of the pre-Samaritan type already existed in the third century BCE and cannot be attributed to the political and religious history of Judea in the Hasmonaean period. (2) The broad chronological distribution of these manuscripts, ranging from the late third century BCE to the late first century BCE, calls for a detailed comparative study of their textual features in order to document the development of the pre-Samaritan text-type over the centuries. Is it, for instance, a coincidence that one of the latest scrolls under consideration, 4Q158 (4QRPᵃ), appears to have a stronger exegetical character than even those other scrolls that have been called “reworked Pentateuchs”?41 39 40 41

Cross, “17. 4QExod-Levᶠ,” 136. Cross, “17. 4QExod-Levᶠ,” 134. Michael Segal, “4QReworked Pentateuch or 4QPentateuch?,” in TheDeadSeaScrolls FiftyYearsAfterTheirDiscovery.ProceedingsoftheJerusalemCongress,July20-25,

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

265

(3) The dates that were previously ascribed to those scrolls are sometimes too late. This is, for instance, the case for 4Q364 (4QRPᵇ), which turns out to be a century older than previously thought, or for 4Q41 (4QDeutn), which is half a century older. Several of the scripts had been qualified as Herodian but, as I explained, no developed or late Herodian features were found, except in 4Q37 (4QDeutʲ). Indeed, these scrolls exhibit very few features that require a date far into the Herodian period. Radiocarbon dating (cf. 4Q365 [4QRPᶜ]) might even lead us to ascribe earlier dates to some features traditionally thought to have appeared in the Herodian period. In fact, the last manuscripts in our list – 4Q365 (4QRPᶜ), 4Q366 (4QRPᵈ), 4Q38 (4QDeutᵏ¹), 4Q158 (4QRPᵃ) and 4Q37 (4QDeutʲ) – are excerpted scrolls or exhibit further textual development than the pre-Samaritan texttype,42 so that the latest scroll of (or close to43) the pre-Samaritan text-type (4Q27 [4QNumᵇ]) was copied at the end of the Hasmonaean period, and not in the Herodian period as Cross initially thought. In conclusion, the new palaeographical dates proposed here and the resulting chronological distribution of these manuscripts show that, by the end of the Hasmonaean period, the pre-Samaritan text-type had been abandoned by the scribes of the Dead Sea Scrolls, except for such works as excerpted manuscripts or so-called “reworked pentateuchs.” By the Herodian period, this text-type was no longer in use for the copy of biblical scrolls.

2. THE SCRIPTS OF “PALAEO-HEBREW” DEAD SEA SCROLLS AND THAT OF THE SAMARITAN PENTATEUCH While most Dead Sea Scrolls use the so-called “Jewish”44 script, which is a local evolution of the Aramaic script, a few manuscripts were copied using the so-called “Palaeo-Hebrew” script, that is, the script

42 43

44

1997, ed. Lawrence H. Schiffman et al. (Jerusalem: Israel Exploration Society / The Shrine of the Book, Israel Museum, 2000), 398. See also the contribution to this volume by Emanuel Tov, pp. 19ff. See the contribution by Emanuel Tov to this volume, pp. 19ff. According to Emanuel Tov, the text-type of 4Q27 (4QNumᵇ) is transitional between the common source of 𝔊⅏ and pre-Samaritan scrolls; see his contribution to this volume, pp. 19ff. See e.g. Cross, “The Development of the Jewish Scripts.” I would argue that, at least until the end of the second century BCE, there is no typical “Jewish” or “Judaean” script, since the same script is used in Samaria and Idumaea, for instance.

266

MICHAEL LANGLOIS

used by Hebrew inscriptions during the first half of the first millennium BCE. Most of them are biblical scrolls:45 1Q3 (1QpaleoLev and 1QpaleoNum), 2Q5 (2QpaleoLev), 4Q11 (4QpaleoGen-Exodˡ), 4Q12 (4QpaleoGenᵐ), 4Q22 (4QpaleoExodᵐ), 4Q45 (4QpaleoDeutʳ), 4Q46 (4QpaleoDeutˢ), 4Q101 (4QpaleoJobᶜ), 6Q1 (6QpaleoGen), 6Q2 (6QpaleoLev), and 11Q1 (11QpaleoLevᵃ). A few non-biblical scrolls have also been found: 4Q123 (4QpaleoParaJosh), 4Q124 (4QpaleoUnid1), 4Q363a (4QcryptC, which uses the Palaeo-Hebrew script with additional “cryptic” signs), 11Q22 (11QpaleoUnid), Mas1o, and Mur 17 (a palimpsest, hence the two inventory numbers Mur 17A [Mur papLetter] and Mur 17B [Mur papList of Personal Names]). I leave aside the new “Jerusalem” papyrus, whose authenticity is debated and which will be dealt with in another study.46 In what follows, we will survey the palaeographical features of most of these scrolls and compare them to other uses of the Palaeo-Hebrew script in order to outline the development of this script and conclude as to its relation with the Samaritan script. a. A Typology of the Palaeo-Hebrew Script in the Dead Sea Scrolls Few scholars have attempted to establish a typology of the PalaeoHebrew script attested in the Dead Sea Scrolls. To my knowledge, the best study so far is an unpublished dissertation by McLean.47 Yet, there are several important features that McLean overlooked and which lead me to outline a new typology. Since a full study is beyond the scope of the present volume, I will limit myself to some of the main typological developments that I have observed on the photographs available to me:48 45

46

47

48

Emanuel Tov, ScribalPracticesandApproachesReflectedintheTextsFoundinthe JudeanDesert, Studies on the Texts of the Desert of Judah 54 (Leiden: Brill, 2004), 246. See my preliminary analysis in Michael Langlois, “How a 2,700-Year-Old Piece of Papyrus Super-Charged the Debate over UNESCO and Jerusalem,” TheHuffingtonPost, 15 November 2016, http://www.huffingtonpost.com/the-conversation-global/how-a-2700year-old-piece_b_12982154.html. I have since examined Mur 17 and the new “Jerusalem” papyrus together in Jerusalem and will publish my assessment in a forthcoming study. Mark David McLean, “The Use and Development of Palaeo-Hebrew in the Hellenistic and Roman Periods” (Harvard University, 1982). Other studies are much more limited in scope and/or depend upon McLean; see e.g. Hanson’s palaeographical analysis in David Noel Freedman, K. A. Mathews, and Richard S. Hanson, eds., ThePaleo-Hebrew LeviticusScroll(11QpaleoLev) (Philadelphia: American Schools of Oriental Research, 1985). Thanks are, once again, due to Pnina Shor and her team, cf. n. 4 above.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

267

(1) In earlier Palaeo-Hebrew inscriptions, letters rest on a virtual base line and words are separated by dots drawn on the baseline. This can still be seen on a few Palaeo-Hebrew Dead Sea Scrolls, but most of them hang letters on a ceiling line, which probably betrays an influence from the Aramaic script. In at least two scrolls (4Q45 and 4Q12449), words are no longer separated by dots. McLean did not take into account these important aspects. (2) Speaking of vertical alignment, Z (= ‫ )ס‬initially features large horizontal strokes that are drawn high above the line, like W (= ‫ ;)ל‬with time, Z (= ‫ )ס‬shifts down so as to align with other letters. McLean did not pay attention to this evolution, perhaps because he mainly deals with coins, where constraints of space do not allow for an ideal vertical placement of letters. (3) ] (= ‫ )צ‬has a short (sometimes very short) left stroke slanted in opposition to the right strokes, usually counterclockwise. With time it tends to become taller and more vertical. The letter as a whole also tends to become taller and/or narrower. (4) ^ (= ‫ )ק‬features a right hook that tends to grow with time; in cursive scripts it can be drawn together with the left hook (in which case it is closed), or separately and become wide open. It can also join the shaft at a lower point, sometimes even at the bottom of the shaft. Those two types seem to develop in parallel, since one is attested in 4Q22 and the other in its repair sheet. (5) T (= ‫ )ט‬is composed of a circle surrounding an inner cross. The circle tends to open with time, which can lead to a new shape with a V-shaped head (see for instance 4Q22 and 4Q45). (6) Q (= ‫ )ו‬features a head composed of a left half-circle followed by a horizontal stroke which rests on a shaded vertical shaft. The halfcircle can become more angular, made up of two strokes forming an acute angle to the left. A more cursive ductus also develops, in which the half-circle and shaft are drawn together, while the horizontal stroke is penned last. The earlier ductus does not disappear, as can be seen in scrolls that use the two types. 49

Tov mentions only 4Q45, cf. Tov, ScribalPractices, 133. Indeed, the editor of 4Q124 in DJD has doubts as to the use of word-separating dots, see Patrick W. Skehan, Eugene C. Ulrich, and Judith E. Sanderson, “124. 4QpaleoUnidentified(1),” in QumranCave 4. IV.Palaeo-HebrewandGreekBiblicalManuscripts, Discoveries in the Judaean Desert IX (Oxford: Clarendon Press, 1992), 206. He mentions a few possible occurrences, but this does not change the fact that the general practice of this scroll is to separate words by spaces.

268

MICHAEL LANGLOIS

(7) V (= ‫ )כ‬has a horizontal traverse whose left end is pointing either downwards or upwards. In the first case, the left stroke rests on it, and in the latter case it hangs from it. A more cursive ductus develops, in which the traverse has a chevron shape; the left stroke may even disappear. The two types can coexist in the same scroll. b. A Palaeographical Dating of the Palaeo-Hebrew Dead Sea Scrolls Dating the Palaeo-Hebrew Dead Sea Scrolls is complicated by several factors. First, their number is limited, and so is their typological development compared to that of the Jewish script. Second, outside these scrolls, there are very few inscriptions that use the Palaeo-Hebrew script in the Persian, Greek and Roman periods: most inscriptions are in Aramaic and thus use the Aramaic script, with very few exceptions. Hebrew inscriptions are rare, but they are usually written in the Palaeo-Hebrew script.50 Coins are by far the largest corpus; there are also numerous tags from Masada (on pottery sherds or column drums), but they are of little help. The use of coins for dating is problematic for several reasons. First, there is a gap of more than a century between Hasmonaean coins and those produced during the Judaean War. Second, many letters of the alphabet are not attested on Hasmonaean coins. Third, the miniature size of letters on coins may have led engravers to forego features attested in contemporary scripts. Fourth, coins do not necessarily represent all the possible contemporary script types: they represent positive evidence confirming the use of a given ductus but cannot be used as negative evidence proving that a given ductus was not in use at the time; a good example is the use of a type of Q (= ‫ )ו‬in Mount Gerizim inscriptions51 which McLean believed to have appeared in the first century CE, more than a century after the destruction of the Mount Gerizim complex. Fifth, the absence of comparative material before Hasmonaean coins makes it difficult to establish a terminuspostquem; there are, however, a few exceptions, notably the Mount Gerizim inscriptions mentioned above. 50

51

For a study of the Palaeo-Hebrew script in the Persian period, see e.g. Martin Peilstöcker and Benjamin Sass, “A Hebrew Seal from Jaffa and the Hebrew Script in the Post-First Temple Period,” ‘Atiqot XLII (2001): 199–210; Gordon J. Hamilton, “Paleo-Hebrew Texts and Scripts of the Persian Period,” in “AnEyeforForm”:EpigraphicEssays inHonorofFrankMooreCross, ed. Jo Ann Hackett and Walter Emanuel Aufrecht (Winona Lake: Eisenbrauns, 2014), 253–90. Yitzhak Magen, Haggai Misgav, and Levana Tsfania, Mount Gerizim Excavations VolumeI:TheAramaic,HebrewandSamaritanInscriptions, Judea and Samaria Publications 2 (Jerusalem: Staff Officer of Archaeology, Civil Administration for Judea and Samaria, 2004).

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

269

Bearing these caveats in mind, let us establish chronological anchors for some of the typological developments of the Palaeo-Hebrew script based on dated inscriptions: (1) Mildenberg coin 15, bearing the inscription OPU (= ‫ )יהד‬and dated 340–331 BCE,52 features a U (= ‫ )י‬with a lower horizontal stroke that protrudes to the left and which is similar to some early PalaeoHebrew Dead Sea Scrolls (4Q12, 6Q2, 2Q5, 4Q11, 1Q3). This ductus later disappears in coins, as well as in later Palaeo-Hebrew Dead Sea Scrolls. (2) Mildenberg coin 16, also dated 340–331 BCE, features a P (= ‫)ה‬ in which the second and third horizontal strokes join to the left, forming a triangle. This triangular ductus also appears on a coin mentioning PS\P PU^RSU (= ‫“ )יחזקיה הפחה‬Hezekiah the governor”53 dated to the end of the Persian period, in the second half of the fourth century BCE. In fact, it already appears on a few jar handles from Gibeon (nos. 1, 5, 14, 20, 42, 52) that could be dated to the early Persian period, in the sixth or fifth century BCE.54 This ductus becomes popular on Hasmonaean coins in the early first century BCE. (3) Coins from the Judaean War use two forms of Q (= ‫)ו‬: an angular form in which the half-circle and diagonal have become a series of three strokes, and a cursive form in which the half-circle is drawn with the shaft. Coins from the Bar-Kokhba period only use the second type and exhibit further development.56 The second type is already attested in Hasmonaean coins57 and in at least one Hebrew inscription from 52 53

54

55 56 57

McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 36–37 and pl. 2.6. Ya῾akov Meshorer, ATreasuryofJewishCoins:FromthePersianPeriodtoBarKokhba (Jerusalem: Yad Ben-Zvi, 2001), 199 no. 22. Frank Moore Cross, “Epigraphical Notes on Hebrew Documents of the Eighth-Sixth Centuries B. C.: III. The Inscribed Jar Handles from Gibeon,” BASOR.168 (1962): 21, 23. For a summary of the discussions surrounding the dating of these finds, see Charles E. Carter, TheEmergenceofYehudinthePersianPeriod:ASocialandDemographic Study, Journal for the Study of the Old Testament Supplement Series 294 (Sheffield: Sheffield Academic Press, 1999), 119–22. These discussions are not over; I am not convinced by van der Veen’s recent argument and conclusion that these handles “may now be safely attributed to the reigns of Hezekiah and Manasseh in historical terms (i.e. 726-697/6 and 697/6-642 BC respectively)”; see Peter van der Veen, “An Inscribed Jar Handle from Ras El-῾Amud. A New Reading and an Absolute Date,” KUSATU 11 (2010): 109–21. Meshorer, TreasuryofJewishCoins, 132. Meshorer, TreasuryofJewishCoins, 163. See e.g. John Hyrcanus I (134–104 BCE) coins in group Ab in Meshorer, Treasuryof JewishCoins, 201.

270

MICHAEL LANGLOIS

Mount Gerizim (no. 385),58 to be dated before 111–110 BCE when the city and its sanctuary were destroyed by John Hyrcanus.59 (4) Mildenberg coin 19 uses a ^ (= ‫ )ק‬in which the right hook has grown and the left hook has almost totally disappeared. Coins from the Judaean War keep the left hook, changing it into a slanted upper tick on top of a protruding left traverse. A similar evolution can be observed in some of the latest Palaeo-Hebrew Dead Sea Scrolls (4Q22, 4Q45, 11Q1). (5) All Hebrew inscriptions discovered on Mount Gerizim (nos. 382– 389) are written with letters hanging from a virtual or real ceiling line, including Z (= ‫( )ס‬no. 384 and 389). On the basis of this comparative material, we can now sketch a chronological framework for some of the Palaeo-Hebrew Dead Sea Scrolls. 4Q46(4QpaleoDeuts) The ductus of 4Q46 (4QpaleoDeutˢ) has slowly evolved from Iron Age scripts and shares some (though not all) of the features attested on Hebrew seals from the later half of the Achaemenid period. 4Q46 would thus be at home in the fifth or fourth centuries BCE; an earlier date is not impossible but lacks clear parallels, whereas a date in the third century is possible but unnecessary. Several letters of the alphabet are not attested, however; this absence prevents a more specific dating. I should emphasize that typological development is not linear60 and indeed appears to be very slow in the Palaeo-Hebrew script of the Persian and early Hellenistic periods.61 McLean dates this scroll to the “second half of the third century” BCE;62 although this date is possible, no comparative material allows for such a precision, and the manuscript may well have been copied earlier. If I may move away from palaeography for a minute, I am pleased to note that this scroll happens to witness one of the oldest text-types of Deuteronomy according to Ziemer.63 My analysis was based on 58 59

60

61

62 63

Magen, Misgav, and Tsfania, MountGerizimExcavationsVol.I, 256. Yitzhak Magen, Mount Gerizim Excavations Volume II: A Temple City, Judea and Samaria Publications 8 (Jerusalem: Staff Officer of Archaeology, Civil Administration for Judea and Samaria, 2008), 171. Contrary to the underlying principle used for instance by Solomon A. Birnbaum, The HebrewScripts.PartOne:TheText (Leiden: E. J. Brill, 1971). See e.g. Peilstöcker and Sass, “A Hebrew Seal from Jaffa,” 202–6; Hamilton, “PaleoHebrew Texts and Scripts of the Persian Period,” 259, 269. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 57. See his contribution to the present volume, pp. 127ff.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

271

palaeographical features only, without prior knowledge of the textual character of this scroll, but the fact that its early date fits Ziemer’s stemma is telling. 4Q12(4QpaleoGenᵐ) 4Q12 (4QpaleoGenᵐ) shows little development from 4Q46 – see Q (= ‫)ו‬, U (= ‫ )י‬or V (= ‫ )כ‬for instance – and would also be at home in the fifth or fourth centuries BCE, perhaps in the third century should the development of the script be slow. McLean dates 4Q12 to the “middle of the second century” BCE;64 such a late date is unnecessary. 2Q5(2QpaleoLev) 2Q5 (2QpaleoLev) features more curved letters, as can be seen in some Q (= ‫)ו‬, the V (= ‫)כ‬, and the base of X (= ‫ )מ‬for instance. A similar tendency might be reflected in the descender of O (= ‫)ד‬, which tends to lengthen. The script also uses the new ductus for U (= ‫ )י‬with three parallel strokes, which was only nascent in 4Q12. It is difficult to pinpoint the birth of this ductus, but this manuscript could be at home in the fourth or third centuries. McLean dates it to ca. “150 to 75 BCE”65 which seems unnecessarily late. 6Q2(6QpaleoLev) 6Q2 (6QpaleoLev) is close to 2Q5 and exhibits the same ductus for U (= ‫ )י‬with further development. The single occurrence of N (= ‫)ג‬ exhibits an unusual shape, with thickening at the left and at the bottom. Overall, 6Q2 may also have been copied around the fourth or third centuries BCE. McLean acknowledges the affinities between 6Q2 and 2Q5 and ascribes them both the same unnecessarily late date between 150 and 75 BCE. 4Q11(4QpaleoGen-Exodl) 4Q11 (4QpaleoGen-Exodˡ) exhibits further development; the letters are clearly hanging from a ceiling line – a phenomenon also attested on 64 65

McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 60. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 100.

272

MICHAEL LANGLOIS

Mount Gerizim inscriptions – and ^ (= ‫ )ק‬features a large right hook joining the shaft at its base. The descender of X (= ‫ )מ‬has moved to the left. These developments suggest that 4Q11 may have been copied around the third century, though earlier and later dates are possible. McLean dates it to “the first half or first three-quarters of the first century BCE,”66 but what he views as late features, such as the non-protruding stroke of P (= ‫ )ה‬or the width of N (= ‫)ג‬, cannot be confined to a specific period, as can be seen by the absence of those features in manuscripts he ascribed to that period. 1Q3frags.1–15(1QpaleoLev;1QpaleoNum) 1Q3 (frags. 1–15, 1QpaleoLev) has affinities with 6Q2 and 2Q5 but is typologically later: the letters are clearly hanging from a ceiling line and Z (= ‫ )ס‬is drawn lower. ] (= ‫ )צ‬features a taller left stroke, while the right hook of ^ (= ‫ )ק‬sometimes joins the base of the shaft. Although a date in the fourth century is possible, 1Q3 is probably more at home in the third century, like 4Q11. McLean’s dating between “150 to 75 BCE”67 is, once again, probably late, while Birnbaum’s dating “ca. 440 B.C.E.”68 is too early, flawed by his methodology which supposes a linear evolution of the script with 12 new features corresponding to 150 years. 6Q1(6QpaleoGen),4Q101(4QpaleoJobᶜ)and4Q123(4QpaleoParaJosh) Another group of manuscripts, made up of 6Q1 (6QpaleoGen), 4Q101 (4QpaleoJobᶜ) and 4Q123 (4QpaleoParaJosh), is also typologically slightly later than 2Q5 and 6Q2 but does not feature the three-stroke U (= ‫)י‬ attested by 4Q12, 6Q2, 2Q5, 4Q11 and 1Q3. These three manuscripts (6Q1, 4Q101 and 4Q123) may have been copied around the third century BCE. McLean dates 4Q101 “between 225 and 150 BCE,”69 and 6Q1 and 4Q123 to the “last half of the second century” BCE70; these ranges are possible but too narrow and a bit late.

66 67 68 69 70

McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 66. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 100. Birnbaum, TheHebrewScripts.PartOne:TheText, col. 69. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 52. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 65–66. The paragraph that starts at the last line of p. 65 is obviously misplaced; it likely belongs before the discussion of 4QpaleoSn44a (= 4Q123) at the bottom of p. 63.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

273

11Q1(11QpaleoLevᵃ),4Q22(4QpaleoExodᵐ)and4Q45(4QpaleoDeutr) The scripts of 11Q1 (11QpaleoLevᵃ), 4Q22 (4QpaleoExodᵐ) and 4Q45 (4QpaleoDeutʳ) are much more developed. Letters are not shaded and shafts tend to be vertical rather than slanted. ] (= ‫ )צ‬is more compact. ^ (= ‫ )ק‬is wide open with a left horn. The hand of 11Q1 (11QpaleoLevᵃ) is more hesitant and has affinities with some of the Mount Gerizim Hebrew inscriptions: the shaft of L (= ‫)א‬ does not protrude to the top (compare Mount Gerizim inscription no. 387); Q (= ‫ )ו‬has a curved head drawn together with the shaft while the diagonal is drawn after (compare inscription no. 385); and T (= ‫ )ט‬has a “+” sign at its center instead of the usual “×” (compare inscription no. 385). The date of 11Q1 thus depends on that of the Mount Gerizim inscriptions; the latter may either belong to the first precinct, which was in use from the Persian period until the late third century BCE, or to the new precinct, which was built in the early second century BCE.71 In any case, 11Q1 is at home in the second century BCE; McLean dates it “between 1 and 50 CE,”72 pointing for instance to some types of Q (= ‫ )ו‬attested on coins from the Judaean War, but this ductus is now attested on Mount Gerizim inscriptions. Hanson dates it “around 100 B.C.E.,”73 which is possible but a bit late given the new evidence from Mount Gerizim. 4Q22 (4QpaleoExodᵐ) and 4Q45 (4QpaleoDeutʳ) have affinities with 11Q1; moreover, they feature a new ductus for T (= ‫)ט‬. They may be slightly later, which means that 4Q22 would be more at home in the latter half of the second century BCE or the early first century BCE – depending, again, on the date of 11Q1 and of the Mount Gerizim inscriptions. McLean dates 4Q22 to “the first half to three quarters of the first century BCE”74; this range is indeed possible but too narrow and perhaps late. Hanson’s dating “around 100 B.C.E.”75 is correct. Now, 4Q22 has been radiocarbon-dated to 2044±65 BP.76 I used intCal 13.14c to produce calibrated dates:

71 72 73

74 75

76

Magen, MountGerizimExcavationsVolumeII:ATempleCity, 143. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 87. Freedman, Mathews, and Hanson, ThePaleo-HebrewLeviticusScroll(11QpaleoLev), 23. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 78. Freedman, Mathews, and Hanson, ThePaleo-HebrewLeviticusScroll(11QpaleoLev), 23. Jull et al., “Radiocarbon Dating of Scrolls and Linen Fragments from the Judean Desert,” 14.

274

MICHAEL LANGLOIS intcal13.14c 2350 Sample ID 2139+/-32 1 and 2 sigma

2300

2250

2200

Radiocarbon Age

2150

2100

2050

2000

1950

1900 350

300

250

200

150

100

50

0

cal BC/AD

The 1σ-calibrated ranges are 159-132 BCE and 117 BCE-21 CE, while the 2σ-calibrated ranges are 346-321 BCE or 206 BCE-84 CE. There is no need to resort to 2σ-calibrated dates, since 1σ-calibrated dates ascribe the copy of 4Q22 between the mid-second century BCE to the turn of the era, which is consistent with our date, albeit much broader. 4Q45 (4QpaleoDeutʳ) is very close to 4Q22 but does not use dots to separate words, a practice also attested in Mount Gerizim inscriptions. 4Q45 would thus be at home in the later half of the second century BCE or the early first century BCE. Before moving on to the next scrolls, and although I grouped these three scrolls – 11Q1 (11QpaleoLevᵃ), 4Q22 (4QpaleoExodᵐ) and 4Q45 (4QpaleoDeutʳ) – on the sole basis of their scripts, it is tempting to look at their textual character since they exhibit more palaeographical affinities with the Mount Gerizim inscriptions than other scrolls. Indeed, 4Q22 (4QpaleoExodᵐ) is known to belong the pre-Samaritan text-type. What about the other two? It turns out that the stemma of Deuteronomy prepared by Benjamin Ziemer for the present volume identifies 4Q45 (4QpaleoDeutʳ) as a direct “sister” of ⅏.77 And as for 11Q1 (11QpaleoLevᵃ), its editors 77

See his contribution pp. 127ff.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

275

disagree: Freedman advocates a “proto-Samaritan”78 text-type, while Mathews concludes that 11Q1 has “no clear textual affiliation.”79 He agrees in this respect with Tov,80 while Puech insists that the agreements with 𝔊 are more numerous than acknowledged.81 Although the paucity of palaeographical evidence at our disposal prevents us from qualifying the script of the Mount Gerizim inscriptions – and even more so that of these three scrolls – as Samaritan, I would not be surprised if a stemma of Leviticus revealed that 11Q1 is closer to ⅏ than previously thought, as was the case for 4Q45. Speaking of 4Q45, its palaeographical dating correlated to its direct relation to ⅏ in Ziemer’s stemma indicates that their parent cannot be dated later than the second century BCE. 4Q124(4QpaleoUnid1) The absence of separating dots is also noticeable in 4Q124 (4QpaleoUnid1), which may also have been copied in the later second century BCE or in the early first century BCE. Also noticeable is the presence of a leftward curve on top of W (= ‫)ל‬, which develops the concave ductus of earlier scrolls (e.g. 6Q1 and 4Q101, see also some Hasmonaeans coins;82 compare the leftward tick on some coins from the Judaean War83). McLean dates 4Q124 to “the first half or first three-quarters of the first century BCE,”84 which is possible but a bit narrow. 4Q22(4QpaleoExodᵐ)repairsheet The curve observed on top of W (= ‫ )ל‬in 4Q124 is even more developed on the repair sheet of 4Q22. Other letters there enjoy a curved ductus, such as P (= ‫)ה‬, Q (= ‫)ו‬, \ (= ‫ )פ‬and _ (= ‫)ר‬. This repair sheet is therefore probably at home in the first century BCE. It has been 78

79

80 81

82

83 84

David Noel Freedman, “Variant Readings in the Leviticus Scroll from Qumran Cave 11,” TheCatholicBiblicalQuarterly 36.4 (1974): 533. K. A. Mathews, “The Leviticus Scroll (11QpaleoLev) and the Text of the Hebrew Bible,” TheCatholicBiblicalQuarterly 48.2 (1986): 198. He agrees in this respect with Emanuel Tov, “‫ בקומראן‬11 ‫( אופייה הטקסטואלי של מגילת ויקרא ממערה‬The Textual Character of the Leviticus Scroll from Qumran Cave 11),” Shnaton 3 (1978): 238–44. Tov, “The Textual Character of the Leviticus Scroll from Qumran Cave 11.” Émile Puech, “Notes en marges de 11QpaléoLévitique. Le fragment L, des fragments inédits et une jarre de la grotte 11,” RevueBiblique 96.2 (1989): 181. See for instance ANS 107 (pl. 15.4) and H 136 (pl. 15.13) in McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods.” Meshorer, TreasuryofJewishCoins, 132. McLean, “Palaeo-Hebrew in the Hellenistic and Roman Periods,” 66.

276

MICHAEL LANGLOIS

radiocarbon-dated to 2024±39 BP;85 I used intCal 13.14c to produce calibrated dates: intcal13.14c 2300 Sample ID 2024+/-39 1 and 2 sigma

2250

2200

2150

2100

Radiocarbon Age

2050

2000

1950

1900

1850

1800

1750 300

200

100

0

100

200

cal BC/AD

The 1σ-calibrated ranges are 88-75 BCE and 57 BCE-26 CE, while the 2σ-calibrated ranges are 160-132 BCE or 117 BCE-62 CE. Once again there is no need to resort to 2σ-calibrated dates, since 1σ-calibrated dates ascribe the copy of this repair sheet to the first century BCE or the early first century CE. This dating is in line with my assessment. c. Synthesis The dates derived from the palaeographical analysis of Palaeo-Hebrew Dead Sea Scrolls can be summed up as follows: ca. 5th or 4th c. BCE

4Q46 (4QpaleoDeutˢ) 4Q12 (4QpaleoGenᵐ)

85

Jull et al., “Radiocarbon Dating of Scrolls and Linen Fragments from the Judean Desert,” 14.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

ca. 4th or 3rd c. BCE

277

2Q5 (2QpaleoLev) 6Q2 (6QpaleoLev)

rd

ca. 3 c. BCE

4Q11 (4QpaleoGen-Exodˡ) 1Q3 frags. 1–15 (1QpaleoLev; 1QpaleoNum) 6Q1 (6QpaleoGen) 4Q101 (4QpaleoJobᶜ) 4Q123 (4QpaleoParaJosh)

ca. 2nd c. BCE

11Q1 (11QpaleoLevᵃ)

ca. 2nd half of 2nd c. or early 1st c. BCE 4Q22 (4QpaleoExodᵐ) 4Q45 (4QpaleoDeutʳ) 4Q124 (4QpaleoUnid1) ca. 1st c. BCE

4Q22 (4QpaleoExodᵐ) repair sheet

I emphasize that these dates are approximate, due (1) to the slow evolution of the Palaeo-Hebrew script – not only at that time, but from the eighth century BCE onward – and (2) to the paucity of witnesses, both among and outside the Dead Sea Scrolls. However approximate they may be, these new dates are nonetheless telling: (1) Many Palaeo-Hebrew scrolls seem to have been copied before scribes began to use the Aramaic script to write Hebrew texts, that is around the second half of the third century BCE, as seen for instance with 4Q17 (4QExod-Levᶠ), studied above in the first part of the present essay. The use of the Aramaic script to write a Hebrew text is indeed quite uncommon; for instance, only one Hebrew inscription from Mount Gerizim uses the Aramaic script (no. 150),86 which points to the turn of the second century BCE for the adoption of the Aramaic script to write Hebrew texts. A few other Mount Gerizim inscriptions use the Aramaic script with occasional Palaeo-Hebrew letters, which may indicate that this adoption was nascent or gradual. (2) Likewise, the strong development exhibited by 11Q1 and later PalaeoHebrew Dead Sea Scrolls might be explained by the fact that these manuscripts were copied when the Aramaic script came to be used for Hebrew texts. This would also explain the absence of separating dots in the latest Palaeo-Hebrew Dead Sea Scrolls. 86

Magen, Misgav, and Tsfania, MountGerizimExcavationsVol.I, 141–42.

278

MICHAEL LANGLOIS

(3) After this transitional period, the Palaeo-Hebrew script was apparently abandoned by the scribes of the Dead Sea Scrolls around the first century BCE. An exception is the repair sheet of 4Q22, but one would not expect a scribe to use the Jewish script to produce a repair sheet for a Palaeo-Hebrew scroll. (4) As a corollary, it is precisely at that time that scribes began to use Palaeo-Hebrew letters to write the tetragrammaton and other divine names or titles in manuscripts that were otherwise copied using the Jewish script. Indeed, this practice is only attested in manuscripts of the first centuries BCE and CE.87 d. The Samaritan script The new chronology that I suggest here has repercussions for Samaritan studies. Since the topic of this paper is palaeography, let us focus on the Samaritan script, whose origins are debated. According to Samaritan tradition, it is the very same script that was used by ancient Israelites. The Abisha scroll is even believed to have been penned by Abishua, the great-grandson of Aaron. At the other extreme, Barag recently concluded that the Samaritan script is a “conscious creation of a related but different alphabet” in the fourth century CE.88 His conclusion is based on the absence of any inscription using the Samaritan script earlier than the fourth century BCE. He is both right and wrong. On the one hand, I agree with him that the Samaritan script, as we know it around the fourth century CE, is not attested earlier. But its origins can now be traced thanks to the Palaeo-Hebrew Dead Sea Scrolls and to some palaeographical features mentioned above. Indeed, some of the Palaeo-Hebrew Dead Sea Scrolls witness the development of a new ductus for U (= ‫)י‬, in which the base protrudes to the left of the shaft, eventually becoming an independent third left stroke. 87

88

There are 29 such manuscripts according to Tov, ScribalPractices, 242–43. They can all be dated to the first centuries BCE and CE: 1Q11 (1QPsᵇ), 1Q14 (1QpMic), 1QpHab, 1Q15 (1QpZeph), 1Q27 (1QMyst), 1QHᵃ, 1Q35 (1QHᵇ), 2Q3 (2QExodᵇ), 3Q3 (3QLam), 3Q14 (3QUnclassified Fragments), 4Q20 (4QExodʲ), 4Q38a (4QDeutᵏ²), 4Q57 (4QIsaᶜ), 4Q161 (4QpIsaᵃ), 4Q165 (4QpIsaᵉ), 4Q171 (4QpPsᵃ), 4Q173 (4QpPsᵇ), 4Q180 (4QAgesCreat A), 4Q183 (4QHistorical Work), 4Q258 (4QSᵈ), 4Q267 (4QDᵇ), 4Q268 (4QDᶜ), 4Q406 (4QShirShabbᵍ), 4Q413 (4QComposition concerning Divine Providence), 4Q26b (4QLevᵍ), 6Q15 (6QD), 6Q18 (6QpapHymn), 11Q2 (11QLevᵇ), and 11Q5 (11QPsᵃ). Dan Barag, “Samaritan Writing and Writings,” in FromHellenismtoIslam:Cultural andLinguisticChangeintheRomanNearEast, ed. Hannah M. Cotton et al. (Cambridge: Cambridge University Press, 2009), 319.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

279

This phenomenon is nascent in 4Q12 (4QpaleoGenᵐ) and developed in 2Q5 (2QpaleoLev), 4Q11 (4QpaleoGen-Exodˡ), 1Q3 (1QpaleoLev; 1QpaleoNum) and even more in 6Q2 (6QpaleoLev).89 These manuscripts were copied around the fourth and third centuries BCE. Other manuscripts of the same period – 6Q1 (6QpaleoGen), 4Q101 (4QpaleoJobᶜ) and 4Q123 (4QpaleoParaJosh) – exhibit another development and do not feature this kind of U (= ‫)י‬. Later, none of the Palaeo-Hebrew Dead Sea Scrolls from the second century or early first century BCE uses this kind of U (= ‫)י‬. The same is true of scrolls in Jewish or Greek scripts from the first centuries BCE and CE that use Palaeo-Hebrew letters to write divine names and titles. Likewise, coins from the Hasmonaean period or the Judaean War, or even from the Bar-Kokhba period, do not use this U (= ‫ )י‬either, nor do the few inscriptions found in Jerusalem or the Masada tags. It appears only in a small scrap of papyrus from Masada (Mas 1039-320 = Mas1o),90 but guess what: this papyrus mentions Mount Gerizim and may thus be of Samaritan origin! The later use of this ductus for U (= ‫ )י‬in the Samaritan script suggests that it was alive and well throughout those centuries, though not in Judaean inscriptions. The alternative explanation, namely that the same ductus was invented again in the fourth century CE by chance, is very unlikely, especially given the fact that its only attestation at the turn of the Christian era is in a text related to Mount Gerizim (Mas1o). Even if we were to accept Barag’s hypothesis of a conscious creation in the fourth century CE, the Palaeo-Hebrew script on which the Samaritan script was based must have featured this type of U (= ‫)י‬. The likeliest explanation is that this PalaeoHebrew script derived from the one attested in such manuscripts as 6Q2 (6QpaleoLev) and developed on its own from the second century BCE, while another branch gave rise to the late Palaeo-Hebrew script attested in such manuscripts as 11Q1 (11QpaleoLevᵃ). Interestingly enough, this first branch of the Palaeo-Hebrew script – i.e. the ancestor of the Samaritan script, so to say – is not attested in the rare 89

90

For a previous observation that the script of 1Q3 (1QpaleoLev) might be at the origin of the Samaritan script, see Esther Eshel and Hanan Eshel, “Dating the Samaritan Pentateuch’s Compilation in Light of the Qumran Biblical Scrolls,” in Emanuel:Studies intheHebrewBible,theSeptuagint,andtheDeadSeaScrollsinHonorofEmanuel Tov, ed. Shalom M. Paul et al. (Leiden / Boston: Brill, 2003), 224. Yigael Yadin, “The Excavation of Masada – 1963/64: Preliminary Report,” IsraelExplorationJournal 15.1/2 (1965): 109; Shemaryahu Talmon and Yigael Yadin, Masada VI, YigaelYadinExcavations1963–1965FinalReports:HebrewFragmentsfromMasada; TheBenSiraScrollfromMasada (Jerusalem: Israel Exploration Society / The Hebrew University of Jerusalem, 1999), 138–49.

280

MICHAEL LANGLOIS

Hebrew inscriptions from Mount Gerizim. It is therefore not specifically Samaritan in origin, at least in the third and early second centuries BCE. From the second century BCE onwards, it developed on its own, but perhaps not in Judaea, given the absence of evidence for this script. The one exception, Mas1o, is perhaps not an exception at all, since it mentions Mount Gerizim and could actually be Samaritan. e. Conclusions Until new inscriptions are found, and taking into account the latest epigraphical evidence, the following conclusions can be drawn: (1) Contrary to rabbinical tradition, there is no evidence that the Hebrew script was replaced by the Aramaic script to write the Torah (or more generally Hebrew texts) at the time of Ezra, that is, in the early Persian period.91 And contrary to modern academic usage, there is no need to call this script “Phoenician”92 or “Palaeo-Hebrew”93 or “Neo-Hebrew”94 or “Neo-Palaeo-Hebrew”;95 it is just the Hebrew script. (2) The earliest use of the Aramaic script for Hebrew texts is attested in the late third century BCE, both at Qumran and on Mount Gerizim, but remains exceptional at the time. The use of the Hebrew script in early Dead Sea Scrolls is therefore not, as Cross and many others thought, “an archaistic survival from the book hand of Israelite times.”96 It

91

92

93 94 95 96

The tradition, preserved in m. Yadayim 4:5, that scrolls using the Hebrew script do not make hands impure, is likewise later and perhaps anti-Samaritan since, as the evidence suggests, a branch of the Hebrew script remained in continuous use among the Samaritans. The rabbinic tradition was also known to Christians; Epiphanius of Salamis thus states that the change of script was made at the time of Ezra, who “desired to make Israel different from the other peoples” and abandoned the Hebrew script “because that form was already in the possession of the Samaritans”; DeXIIgemmisrationalissacerdotisHebraeorumliber (Migne, Patrologia, vol. 43, p. 356), quoted by Birnbaum, The Hebrew Scripts. Part One: The Text, cols. 73–74. e.g. Dominique Barthélemy, “3. Lévitique et autres fragments en écriture ‘phénicienne,’” in QumranCaveI, ed. Dominique Barthélemy and Józef Tadeusz Milik, Discoveries in the Judaean Desert I (Oxford: Clarendon Press, 1955), 51–54. e.g. Cross, “The Development of the Jewish Scripts.” e.g. Magen, Misgav, and Tsfania, MountGerizimExcavationsVol.I. e.g. Birnbaum, TheHebrewScripts.PartOne:TheText. Cross, “The Development of the Jewish Scripts,” 189 n. 4.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

281

was, on the contrary, perfectly normal that a Hebrew text should use the Hebrew script.97 (3) Around the second century BCE, the use of the Aramaic script for Hebrew texts becomes the norm at Qumran. The Hebrew script is progressively abandoned, and from the first century BCE on it is only used sporadically, e.g. for divine names or cryptic scripts. (4) This desertion is not global. From the second century BCE onward, a branch of the Hebrew script is used in Samaritan circles, while another branch is used elsewhere in Judea in the first centuries BCE and CE, as attested on coins, tags,98 and a few inscriptions (including an Aramaic one99). (5) Contrary to the minimalist theory of Barag, the branch of the Hebrew script used in Samaritan communities evolves naturally throughout the Roman period. It eventually becomes the so-called “Samaritan” script in the Byzantine period. These conclusions, based solely on palaeographical evidence and taking into account the latest epigraphical discoveries, stand in sharp contrast with both the maximalist theory of the Samaritan tradition and the minimalist theory of Barag. It also has ramifications for the history of Judaism and Samaritanism in the Second Temple period: what are the historical reasons behind the first use of the Aramaic script for Hebrew texts sometime around the third century, both at Qumran and on Mount Gerizim? Why was the Hebrew script abandoned around the second century BCE by the scribes of the Dead Sea Scrolls? Was it a natural evolution100 or a reaction against other movements, such as the Sadducees101 or the 97

98

99 100

101

Contra, i.a., William M. Schniedewind, “Aramaic, the Death of Written Hebrew, and Language Shift in the Persian Period,” in MarginsofWriting,OriginsofCultures, ed. Seth L. Sanders, Second Printing with Postscripts and Minor Corrections, Oriental Institute Seminars 2 (Chicago: The Oriental Institute of the University of Chicago, 2007), 141, 143. Yigael Yadin, Joseph Naveh, and Yaacov Meshorer, MasadaI,TheYigaelYadinExcavations1963–1965FinalReports:TheAramaicandHebrewOstracaandJarInscriptions;TheCoinsofMasada (Jerusalem: Israel exploration society / The Hebrew University of Jerusalem, 1989). Those tags are the product of workers, not learned scribes; the Hebrew script is therefore not confined to academic circles at the time. Joseph Naveh, “An Aramaic Tomb Inscription Written in Paleo-Hebrew Script,” IsraelExplorationJournal 23.2 (1973): 82–91. This is the opinion of Naveh, who connects this shift to the development of a “Jewish” flavor of the Aramaic script in the late third century BCE; cf. Naveh, “An Aramaic Tomb Inscription Written in Paleo-Hebrew Script,” 90. But the so-called “Jewish” script is now attested outside of Judea, especially in Samaria and Idumaea. See e.g. Diringer, who believed that the Sadducees used the Hebrew script; David Diringer, “Early Hebrew Script Versus Square Hebrew Script,” in EssaysandStudies PresentedtoStanleyArthurCook, ed. David Winton Thomas (London: Taylor’s Foreign Press, 1950), 35–49.

282

MICHAEL LANGLOIS

Samaritans? If the Hebrew script used by the Samaritans evolved on its own from that period on, can the same be said about their textual tradition of the Scriptures? These are some of the questions raised by mere palaeographical analysis. BIBLIOGRAPHY ALLEGRO, John Marco. Qumrân Cave 4. I (4Q158–4Q186). Discoveries in the Judaean Desert of Jordan V. Oxford: Clarendon Press, 1968. ANDERSON, Robert T., and Terry GILES. TheSamaritanPentateuch:AnIntroductiontoItsOrigin,History,andSignificanceforBiblicalStudies. Resources for Biblical Study 72. Atlanta: Society of Biblical Literature, 2012. BARAG, Dan. “Samaritan Writing and Writings.” Pages 303–23 in FromHellenism to Islam: Cultural and Linguistic Change in the Roman Near East. Edited by Hannah M. Cotton, Robert G. Hoyland, Jonathan J. Price, and David J. Wasserstein. Cambridge: Cambridge University Press, 2009. BARTHÉLEMY, Dominique. “3. Lévitique et autres fragments en écriture ‘phénicienne.’” Pages 51–54 in QumranCaveI. Edited by Dominique Barthélemy and Józef Tadeusz Milik. Discoveries in the Judaean Desert I. Oxford: Clarendon Press, 1955. BIRNBAUM, Solomon A. The Hebrew Scripts. Part One: The Text. Leiden: E. J. Brill, 1971. BONANI, Georges, Susan IVY, Willy WOLFLI, Magen BROSHI, Israel CARMI, and John STRUGNELL. “Radiocarbon Dating of Fourteen Dead Sea Scrolls.” Radiocarbon 34.3 (1992): 843–49. CARTER, Charles E. TheEmergenceofYehudinthePersianPeriod:ASocialand DemographicStudy. Journal for the Study of the Old Testament Supplement Series 294. Sheffield: Sheffield Academic Press, 1999. CROSS, Frank Moore. “17. 4QExod-Levᶠ.” Pages 133–44 in QumranCave4.VII. GenesistoNumbers. Discoveries in the Judaean Desert XII. Oxford: Clarendon Press, 1994. —. “Epigraphical Notes on Hebrew Documents of the Eighth-Sixth Centuries B. C.: III. The Inscribed Jar Handles from Gibeon.” BASOR 168 (1962): 18–23. —. “The Development of the Jewish Scripts.” Pages 133–202 in TheBibleand theAncientNearEast.EssaysinHonorofWilliamFoxwellAlbright. Edited by G. Ernest Wright. London: Routledge & Kegan Paul, 1961. DIRINGER, David. “Early Hebrew Script Versus Square Hebrew Script.” Pages 35–49 in EssaysandStudiesPresentedtoStanleyArthurCook. Edited by David Winton Thomas. London: Taylor’s Foreign Press, 1950. DOUDNA, Greg. “Dating the Scrolls on the Basis of Radiocarbon Analysis.” Pages 430–71 in TheDeadSeaScrollsafterFiftyYears:AComprehensive Assessment,VolumeOne. Edited by Peter W. Flint and James C. VanderKam. Leiden / Boston / Köln: Brill, 1998. DUNCAN, Julie Ann. “37. 4QDeutj.” Pages 75–91 in QumranCave4.IX.Deuteronomy, Joshua, Judges, Kings. Edited by Eugene C. Ulrich, Frank Moore Cross, Sidnie White Crawford, Julie Ann Duncan, Patrick W. Skehan,

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

283

Emanuel Tov, and Julio Trebolle Barrera. Discoveries in the Judaean Desert XIV. Oxford: Clarendon Press, 1995. —. “38. 4QDeutk1.” Pages 93–98 in QumranCave4.IX.Deuteronomy,Joshua, Judges,Kings. Edited by Eugene C. Ulrich, Frank Moore Cross, Sidnie White Crawford, Julie Ann Duncan, Patrick W. Skehan, Emanuel Tov, and Julio Trebolle Barrera. Discoveries in the Judaean Desert XIV. Oxford: Clarendon Press, 1995. ESHEL, Esther. “4QLevd: A Possible Source for the Temple Scroll and Miqṣat Ma῾aśeHa-Torah.” DSD 2.1 (1995): 1–13. ESHEL, Esther, and Hanan ESHEL. “Dating the Samaritan Pentateuch’s Compilation in Light of the Qumran Biblical Scrolls.” Pages 215–40 in Emanuel:Studies in the Hebrew Bible, the Septuagint, and the Dead Sea Scrolls in Honor of Emanuel Tov. Edited by Shalom M. Paul, Robert A. Kraft, Lawrence H. Schiffman, and Weston W. Fields. Leiden / Boston: Brill, 2003. ESHEL, Esther, and Amos KLONER. “An Aramaic Ostracon of an Edomite Marriage Contract from Maresha, Dated 176 B.C.E.” IsraelExplorationJournal 46.1/2 (1996): 1–22. FREEDMAN, David Noel. “Variant Readings in the Leviticus Scroll from Qumran Cave 11.” TheCatholicBiblicalQuarterly 36.4 (1974): 525–34. FREEDMAN, David Noel, K. A. Mathews, and Richard S. Hanson, eds. ThePaleoHebrew Leviticus Scroll (11QpaleoLev). Philadelphia: American Schools of Oriental Research, 1985. HAMILTON, Gordon J. “Paleo-Hebrew Texts and Scripts of the Persian Period.” Pages 253–90 in “AnEyeforForm”:EpigraphicEssaysinHonorofFrank Moore Cross. Edited by Jo Ann Hackett and Walter Emanuel Aufrecht. Winona Lake: Eisenbrauns, 2014. JASTRAM, Nathan. “27. 4QNumb.” Pages 205–67 in QumranCave4.VII.Genesis toNumbers. Discoveries in the Judaean Desert XII. Oxford: Clarendon Press, 1994. JULL, Timothy A. J., Douglas J. DONAHUE, Magen BROSHI, and Emanuel TOY. “Radiocarbon Dating of Scrolls and Linen Fragments from the Judean Desert.” Radiocarbon 37.1 (1995): 11–19. LANGLOIS, Michael. “How a 2,700-Year-Old Piece of Papyrus Super-Charged the Debate over UNESCO and Jerusalem.” TheHuffingtonPost, 15 November 2016. http://www.huffingtonpost.com/the-conversation-global/how-a-2700year-old-piece_b_12982154.html. —. “Palaeographical Analysis of the Dead Sea Scrolls in The Schøyen Collection.” Pages 79–128 in Gleanings from the Caves. Dead Sea Scrolls and ArtefactsfromTheSchøyenCollection. Edited by Torleif Elgvin, Kipp Davis, and Michael Langlois. Library of Second Temple Studies 71. London: Bloomsbury T&T Clark, 2016. MAGEN, Yitzhak. MountGerizimExcavationsVolumeII:ATempleCity. Judea and Samaria Publications 8. Jerusalem: Staff Officer of Archaeology, Civil Administration for Judea and Samaria, 2008. MAGEN, Yitzhak, Haggai MISGAV, and Levana TSFANIA. MountGerizimExcavations Volume I: The Aramaic, Hebrew and Samaritan Inscriptions. Judea and Samaria Publications 2. Jerusalem: Staff Officer of Archaeology, Civil Administration for Judea and Samaria, 2004.

284

MICHAEL LANGLOIS

MATHEWS, K. A. “The Leviticus Scroll (11QpaleoLev) and the Text of the Hebrew Bible.” TheCatholicBiblicalQuarterly 48.2 (1986): 171–207. MCLEAN, Mark David. “The Use and Development of Palaeo-Hebrew in the Hellenistic and Roman Periods.” Harvard University, 1982. MESHORER, Ya῾akov. ATreasuryofJewishCoins:FromthePersianPeriodto BarKokhba. Jerusalem: Yad Ben-Zvi, 2001. NAVEH, Joseph. “An Aramaic Tomb Inscription Written in Paleo-Hebrew Script.” IsraelExplorationJournal 23.2 (1973): 82–91. OWEN, Elizabeth. “4QDeutn: A Pre-Samaritan Text?” DeadSeaDiscoveries 4.2 (1997): 162–78. PEILSTÖCKER, Martin, and Benjamin SASS. “A Hebrew Seal from Jaffa and the Hebrew Script in the Post-First Temple Period.” ‘Atiqot XLII (2001): 199– 210. VAN DER PLICHT, Johannes, and Kaare L. RASMUSSEN. “Radiocarbon Dating and Qumran.” Pages 99–121 in HolisticQumran:Trans-DisciplinaryResearch ofQumranandtheDeadSeaScrolls.ProceedingsoftheNIAS-LorentzCenterQumranWorkshop,21-25April2008. Edited by Jan Gunneweg, Annemie Adriaens, and Joris Dik. Studies on the Texts of the Desert of Judah 87. Leiden: Brill, 2010. PUECH, Émile. “Notes en marges de 11QpaléoLévitique. Le fragment L, des fragments inédits et une jarre de la grotte 11.” RevueBiblique 96.2 (1989): 161–83. —. “Un nouveau fragment 7a de 4QGn-Exa = 4QGn-Ex 1 et quelques nouvelles lectures et identifications du manuscrit 4Q1.” RevuedeQumrân 25.1 (2011): 103–11. SCHNIEDEWIND, William M. “Aramaic, the Death of Written Hebrew, and Language Shift in the Persian Period.” Pages 141–51 in Margins of Writing, OriginsofCultures. Edited by Seth L. Sanders. Second Printing with Postscripts and Minor Corrections. Oriental Institute Seminars 2. Chicago: The Oriental Institute of the University of Chicago, 2007. SEGAL, Michael. “4QReworked Pentateuch or 4QPentateuch?” Pages 391–99 in TheDeadSeaScrollsFiftyYearsAfterTheirDiscovery.Proceedingsofthe JerusalemCongress,July20-25,1997. Edited by Lawrence H. Schiffman, Emanuel Tov, James C. VanderKam, and Galen Marquis. Jerusalem: Israel Exploration Society / The Shrine of the Book, Israel Museum, 2000. SKEHAN, Patrick W., Eugene C. ULRICH, and Judith E. SANDERSON. “124. 4QpaleoUnidentified(1).” Pages 205–14 in QumranCave4.IV.PalaeoHebrew and Greek Biblical Manuscripts. Discoveries in the Judaean Desert IX. Oxford: Clarendon Press, 1992. STRUGNELL, John. “Notes en marge du volume V des « Discoveries in the Judaean Desert of Jordan ».” RevuedeQumrân 7.2 (26) (1970): 163–276. TALMON, Shemaryahu, and Yigael YADIN. MasadaVI,YigaelYadinExcavations 1963–1965FinalReports:HebrewFragmentsfromMasada;TheBenSira Scroll from Masada. Jerusalem: Israel Exploration Society / The Hebrew University of Jerusalem, 1999. TOV, Emanuel. “26. 4QLevd.” Pages 193–95 in QumranCave4.VII:Genesis to Numbers. Discoveries in the Judaean Desert XII. Oxford: Clarendon Press, 1994.

DEAD SEA SCROLLS PALAEOGRAPHY AND THE SAMARITAN PENTATEUCH

285

—. ScribalPracticesandApproachesReflectedintheTextsFoundintheJudean Desert. Studies on the Texts of the Desert of Judah 54. Leiden: Brill, 2004. —, ed. TheTextsfromtheJudaeanDesert.IndicesandanIntroductiontothe DiscoveriesintheJudaeanDesertSeries. Discoveries in the Judaean Desert XXXIX. Oxford: Clarendon Press, 2002. —. “‫ בקומראן‬11 ‫( אופייה הטקסטואלי של מגילת ויקרא ממערה‬The Textual Character of the Leviticus Scroll from Qumran Cave 11).” Shnaton 3 (1978): 238–44. TOV, Emanuel, and Sidnie WHITE. “364. Reworked Pentateuchb.” Pages 197–254 in Qumran Cave 4. VIII. Parabiblical Texts, Part 1. Discoveries in the Judaean Desert XIII. Oxford: Clarendon Press, 1994. —. “365. Reworked Pentateuchᶜ.” Pages 255–318 in Qumran Cave 4. VIII. ParabiblicalTexts,Part1. Discoveries in the Judaean Desert XIII. Oxford: Clarendon Press, 1994. —. “366. 4QReworked Pentateuchd.” Pages 335–43 in Qumran Cave 4. VIII. ParabiblicalTexts,Part1. Discoveries in the Judaean Desert XIII. Oxford: Clarendon Press, 1994. —. “367. 4QReworked Pentateuche.” Pages 345–51 in Qumran Cave 4. VIII. ParabiblicalTexts,Part1. Discoveries in the Judaean Desert XIII. Oxford: Clarendon Press, 1994. VAN DER VEEN, Peter. “An Inscribed Jar Handle from Ras El-῾Amud. A New Reading and an Absolute Date.” KUSATU 11 (2010): 109–21. WHITE CRAWFORD, Sidnie. “41. 4QDeutn.” Pages 93–98 in QumranCave4.IX. Deuteronomy, Joshua, Judges, Kings. Edited by Eugene C. Ulrich, Frank Moore Cross, Sidnie White Crawford, Julie Ann Duncan, Patrick W. Skehan, Emanuel Tov, and Julio Trebolle Barrera. Discoveries in the Judaean Desert XIV. Oxford: Clarendon Press, 1995. —. “A Response to Elizabeth Owen’s ‘4QDeutn: A Pre-Samaritan Text?’” Dead SeaDiscoveries 5.1 (1998): 92–94. YADIN, Yigael. “The Excavation of Masada – 1963/64: Preliminary Report.” IsraelExplorationJournal 15.1/2 (1965): 1–120. YADIN, Yigael, Joseph NAVEH, and Yaacov MESHORER. Masada I, The Yigael YadinExcavations1963–1965FinalReports:TheAramaicandHebrew Ostraca and Jar Inscriptions; The Coins of Masada. Jerusalem: Israel exploration society / The Hebrew University of Jerusalem, 1989.

VARIEGATION IN SECOND TEMPLE PERIOD HEBREW: PASSIVE T-STEMS, THE ‫ הלז‬DEMONSTRATIVE SERIES, AND ‫ אפוא‬IN SAMARITAN HEBREW AND IN THE DEAD SEA SCROLLS* Christian STADEL

The discovery of the Dead Sea Scrolls changed the study of Second Temple period Judaism drastically, in many respects. Leaving the plethora of hitherto unknown Hebrew and Aramaic texts aside, it also led to a reassessment of previously known sources, be they apocryphal and pseudepigraphic books or the various ancient versions of the Bible. Indeed, our understanding of the development and canonization of the biblical text in Second Temple times could be refined considerably as the Qumran material was successively and successfully integrated into to the broader picture. It is in this area of Pentateuchal textual criticism and canonization processes that research on the Samaritan Pentateuch and the Dead Sea Scrolls evinced the most obvious cross-fertilization. There are, however, additional fields that benefit from a comparison between the Samaritan and Qumran evidence. An important one is Hebrew linguistics. In what follows, I shall look at the impetus that Samaritan Hebrew and Qumran Hebrew evidence gave to the field, and at its implications for our view of the nature and development of Hebrew in the Second Temple period. The general overview will be followed by three case studies on grammatical features, in which the Samaritan and Qumran Hebrew varieties in particular shed light on each other. It was noted early on that the Dead Sea Scrolls are a treasure trove for Hebrew historical linguistics, and it took only about a dozen years until Edward Kutscherʼs publication of the first seminal monograph on Qumran Hebrew in 1959.1 Kutscherʼs work set the stage for Elisha Qimronʼs * Thanks to are due to Uri Mor for helpfully answering my bibliographical queries pertaining to Rabbinic Hebrew, and to Mor Shemesh for his assistance in gathering the raw data on the Qumran Hebrew t-stem. 1 Kutscher 1959; English translation Kutscher 1974.

288

CHRISTIAN STADEL

comprehensive treatment of the orthography, phonology, and morphology of the Hebrew of the Dead Sea Scrolls,2 and its general framework was adopted in virtually all subsequent publications. The original Hebrew version of Qimronʼs dissertation appeared in 1976. These were times of seismic shifts in Hebrew historical linguistics, for in the following year, Ze᾿ev Ben-Ḥayyim published the standard treatment of the Samaritan pronunciation tradition.3 While it had been known since Geseniusʼ DePentateuchi samaritaniorigine,indoleetauctoritate,commentatiophilologico-critica from 1815 that the Hebrew of the Samaritan consonantal text differed slightly from Tiberian Biblical Hebrew, Ben-Ḥayyim succeeded in forty years of painstaking research to uncover, document and analyze the oral tradition of the Samaritans, which proved to be completely unlike the other, Jewish oral traditions. This pronunciation tradition in particular is now generally referred to by the name “Samaritan Hebrew,” and wherever we adduce phonological, morphological, and in part also syntactical features of Samaritan Hebrew, they derive from this pronunciation tradition recorded by Ben-Ḥayyim. Actually, the two-layered structure of Samaritan Hebrew, with the consonantal text that reveals only very few differences4 and the oral traditions that evinces a totally new variety of Hebrew, should be supplemented by a third layer, that of the Samaritan Targum from Byzantine times. For only in this third source, the earliest Samaritan translation of their sacred scripture, can we get a glance at Samaritan Hebrew semantics, which occasionally differ as well from what we know from the Jewish traditions. In our test cases, we shall make recourse to all three layers of Samaritan Hebrew. Even though research on Qumran Hebrew and on Samaritan Hebrew developed roughly simultaneously, and comprehensive works appeared only in the 1970s, various scholars have availed themselves of data from the two Hebrew varieties for comparative studies, and with good results.5 The long forms with final /-ā/ of the independent pronouns and pronominal suffixes (e.g., Qumran Hebrew ‫ אתמה‬and Samaritan attimma) are probably the most famous case.6 It also demonstrates that the new data only complicated our prospect of Hebrew in the Second Temple period. Until the 2 3 4

5

6

Qimron 1976; abbreviated English edition Qimron 1986; Qimron 2018. Ben-Ḥayyim 1977; enhanced English edition Ben-Ḥayyim 2000. But these tend to reflect the same language system as the pronunciation tradition (where this is possible to establish), see, e.g., Schorch 2003. E.g., Ben-Ḥayyim 1958–1962; Kutscher 1974, p. 566; Ben-Ḥayyim 2000, passim; Schorch 2008. Morgenstern 2007; Qimron 2013.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

289

middle of the 20th century, when evidence for Hebrew varieties was essentially restricted to Biblical Hebrew (including Greek transcriptions), Late Biblical Hebrew, and Tannaitic Hebrew, these were either thought to represent chronologically successive stages of one language, or Biblical Hebrew was seen as a literary strand with some internal variation, that was rather loosely connected to the patois which was later put to writing as Tannaitic Hebrew. In other words, the differences between the varieties of Hebrew were explained as resulting from language change over time or were thought to reflect a dichotomy of spoken vs. written language, and/or two distinct dialects. The new Samaritan and Qumran Hebrew data fit neither of these explanations, as the example of the pronominal forms demonstrates: the 2mpl. forms ‫ אתמה‬and attimma represent a shared innovation (or a parallel development). For the 2msg. suffix, on the other hand, Qumran Hebrew retained the earlier form ‫כה‬-, with final vowel, while the Samaritans have the typologically later -åk. But for the 2fsg. pronoun, Samaritan Hebrew consistently preserves the more original form atti. Early as well as late elements appear in all varieties, without clear-cut distribution patterns, and this situation only aggravates when taking additional minor corpora into account: The proto-Mishnaic Hebrew of the Copper Scroll and MMT, epigraphic Judean Hebrew from the two revolts,7 the Hebrew reflected in Septuagint translations and transcriptions,8 and the language of the biblical vocalization (as against the consonantal text).9 It is well neigh impossible to force the variegated data from all these different kinds of Second Temple period Hebrew into the straitjacket of either time, geography, or register and genre, i.e. to explain them as reflecting changes in one parameter only. Rather, it takes a multi-dimensional approach to account for the rich and changing dialectal landscape of Second Temple Hebrew, as it has emerged from the discoveries in the 20th century. Aaron Koller has spelled out the relevant parameters for Qumran Aramaic, but mutatis mutandis his presentation holds true for Qumran Hebrew and Samaritan Hebrew as well.10 The vexing question of variegation in the Hebrew of the Second Temple period is best addressed for each language feature in its own right, and with the following defining factors in mind. (1) Chronology: In some instances, one variety of Hebrew changed over time, and the result of this change is attested in different corpora. The increase in the use of the long 1csg. Imperfect form ‫ וָ ֶא ְבנֶ ה‬with past 7 8 9 10

Mor 2015 has a comprehensive treatment. Cf., e.g., Joosten 2002. Cf., e.g., Joosten 2015; Khan 2013. Koller 2011.

290

CHRISTIAN STADEL

tense meaning in the different strata of Biblical Hebrew, in Samaritan Hebrew, and in Qumran Hebrew is a case in point.11 However, while both Samaritan Hebrew and Qumran Hebrew are typologically later than Tiberian Hebrew with respect to the form ‫וָ ֶא ְבנֶ ה‬, they preserve typologically early features in other grammatical categories. For instance, Samaritan Hebrew did not degeminate original /rr/ and retained the maqtal pattern without attenuation;12 and in Qumran Hebrew, forms with an original final vowel preserved the penultimate stress and the original full vowel of this stressed syllable (e.g., 1QIsaa ‫ ותחפורו‬vs. MT Isa 1:29 ‫)וְ ַ֨ת ְח ְפּ ֔רוּ‬.13 Language change over time alone cannot explain all the data. (2) Dialect geography: Few would disagree with the general statement that the spoken Hebrew of the Second Temple period was not uniform and that different local dialects existed, supposedly first as a dialect continuum and later on as pockets in an increasingly Aramaic-speaking environment. In theory, the relevant geographic framework would include Samaria vs. Judea, and in the case of the latter a special place would be reserved for Jerusalem. However, actually, it is notoriously difficult to tie linguistic features attested in the written material to a specific local dialect of Hebrew.14 This is true for epigraphic texts, whose find spots do not necessarily correspond to the place of origin of the scribe,15 but also for literary texts. While, in principal, Biblical Hebrew is convincingly associated with Jerusalem and the Temple milieu, the very same texts appear in a different linguistic garb in Samaritan and Qumran circles, and it is impossible to isolate dialect geography as the defining factor. (3) Genre and register: The preserved linguistic material attests to different textual genres, and supposedly reflects, at least in part, different registers of speech. While the influence of genre on linguistic diversity can be excluded in some cases (most prominent are the Masoretic, Samaritan, and Qumran versions of the language of the Pentateuch), it is probably a relevant factor in others, e.g., in the legal epigraphic material, which might well have been written in a special legalese. The difference in style between biblical and Tannaitic literary texts is usually attributed at least partly to register, and the same probably holds true for the language of the epigraphic letters as well, which were subject to only minimal editing and stylization (and thus represent adhoc language usage), unlike the literary texts. 11 12 13 14 15

Qimron 2008. Stadel 2017; Ben-Ḥayyim 2000, pp. 276–279. Qimron 1992, p. 88. But see Talshir 2003 for an idiosyncratic attempt. Koller 2011, pp. 201–202.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

291

(4) Prestige: Social factors connected to prestige and ideology certainly determine language use, and it is very likely that they did so in the case of Second Temple Hebrew as well. Thus, the choice of Hebrew over Aramaic and Greek for drawing up legal documents during the revolts could arguably have had an ideological background.16 More pertinent are studies that seek to explain the preference of one Hebrew variety over others as a result of deliberate, ideologically motivated choices, as William Schniedewind has proposed for Qumran Hebrew (though his arguments are highly hypothetical and not necessarily convincing).17 In general, it seems that linguistic prestige is a more useful and less elusive concept than supposed language ideologies. And even though we cannot ascertain the different levels of prestige assigned to different varieties of Hebrew, it is safe to assume that Biblical Hebrew enjoyed such prestige in the realm of religious literature. Essentially, it affected the Second Temple Hebrew varieties in two ways: On the one hand, it was a negative, artificially unifying factor that possibly masked additional linguistic differences. On the other hand, whenever the attempt to imitate the prestigious Biblical language was only partially successful, this resulted in new forms and syntagmas – pseudo-classicisms – , which increased the variegation.18 In sum, when looking for explanations for the variegation in Second Temple period Hebrew one should have recourse to all four factors enumerated above. And, more importantly, one should judge each case individually. It is thus unlikely that we will be able to uncover bundles of isoglosses, or of diachronic developments, that would allow us to offer an overarching explanation of all the linguistic facts. Rather, we should expect Second Temple Hebrew to be a kaleidoscope, whose changing forms do not lend themselves easily to one-dimensional explanations. And whatʼs worse, this kaleidoscope is only partially known to us, and the picture we arrive at is thus necessarily fragmentary. In the following, we shall focus on Samaritan Hebrew and Qumran Hebrew and their particular contribution to the general picture of Hebrew in the Second Temple period. In due course, we shall take a look at three subjects, in which differing defining factors were at play: the passive meaning of the tD-stem, the demonstrative series ‫הלז‬, and the reanalysis of the biblical particle ‫אפוא‬.

16 17 18

Eshel 2006. Schniedewind 2000. Joosten 1999; Joosten 2016.

292

CHRISTIAN STADEL

1. TD-STEM WITH

PASSIVE MEANING

One morphological feature of Samaritan Hebrew that has often been adduced in comparative studies is the unique niphˁal paradigm with geminated first radical in all verbal forms, i.e. also in the Perfect and Participle, where the /n/ of the stem would not normally assimilate to the following radical. Samaritan forms such as ‫ ונשמרת‬wniššmårtå (Deut 23:10) and ‫ ונברכו‬wnibbårrku (Gen 12:3) have been compared to the Tannaitic Hebrew nippəˁaland nitpaˁˁel, e.g., ‫אתה‬ ָ ‫יטּ ֵמ‬ ְ ִ‫( נ‬m. Ṭehar. 5:9, MS Parma B) and ‫ית ַא ְכּלוּ‬ ְ ִ‫( נ‬m. Tamid 2:5),19 which are also supposed to have influenced the extraordinary Tiberian vocalization of biblical forms such as ‫וְ נִ וַּ ְסּרוּ‬ (Ezek 23:48) and ‫( וְ נִ ַכּ ֵפּר‬Deut 21:8).20 As perceptively observed by various scholars, the forms in the different dialects all result from the same process of contamination of the original Biblical Hebrew niphˁal and hitpaˁˁel forms.21 It is impossible to know whether Qumran Hebrew participated in this morphological process since the scrolls do not attest to an unequivocal nitpaˁˁel (and the common spelling ‫ נפעל‬could or could not represent a nippəˁal). Therefore, and because the formal side of the process is almost universally agreed upon, we shall leave the matter aside and rather focus on a change in the semantics of the Hebrew t-stem that facilitated this contamination. In a recent overview of the linguistic underpinnings of morphological change, David Fertig states that, cross-linguistically, “[t]he input words to a (non-blend) contamination are always very closely related semantically, typically either as near-synonyms or antonyms.”22 Is this the case with the Biblical Hebrew niphˁal (N) and hitpaˁˁel (tD) stems? This is a vexed question, since the semantics of these stems (and the stem system in general) are debated. In reference grammars, both stems are often assigned the basic meaning of a “reflexive”23 and would thus evince the semantic overlap that facilitates contamination. However, more sophisticated recent studies that look at the attested meanings in detail usually interpret the stems as expressing various middle situation types.24 The diversity of semantic 19

20

21 22 23

24

Florentin 1993, p. 205; Bar-Asher 1999, pp. 48–52, 55–59, 64–65; Ben-Ḥayyim 2000, p. 223. If not indicated otherwise, Mishnah quotations follow the text of MS Kaufmann A 50. Joüon and Muraoka 1996, pp. 169–170 (§59f). The total assimilation of the /t/ to the first radical is also attested in other forms, Joüon and Muraoka 1996, p. 158 (§53e). Florentin 1993, pp. 204–205; Ben-Ḥayyim 2000, p. 117–118; Schorch 2008, p. 187. Fertig 2013, p. 63. E.g., Bauer and Leander 1922, pp. 289–291 (§38w′, f″–h″), and, more nuanced, Joüon and Muraoka 1996, pp. 150–151 (§51c), 159 (§53i). The most pertinent studies are Arnold 2005 for the tD stem and Jenni 2012 for the N stem, which give further references.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

293

sub-categories in both stems probably resulted from a reordering of the stem system, following the loss of the Gt- and Ct-stems (the reflexive counterparts of the qal and hiphˁil) at the proto-Hebrew stage.25 In Biblical Hebrew, the stems share a common semantic denominator in that both demote or fade out the agent. This common denominator is rather vague, though, and it contrasts with notable differences that pertain to passive and reflexive meanings.26 For in the broad realm of middle meanings, N-stem forms can convey passive semantics in certain contexts, while this meaning is very rare in the tD-stem.27 Conversely, the tD-stem accommodates reflexive meanings, which are few and far between in the N-stem.28 It is the former difference that is of relevance to our present investigation. According to the studies on the Hebrew stem system adduced above, Classical Biblical Hebrew employed two kinds of stems to express passive situations: the N-stem with its mediopassive meaning and the set of internal passive stems derived through Ablaut, puˁˁal (Dp), hophˁal (Cp), and – in vestiges – the internal passive of the qal (Gp). But while in the Ablaut stems the passive meaning is inherent to the form, i.e. ‫( ְתּ ֻח ָלּ֑ק‬Amos 7:17; Dp), ‫( וַ יֻּ גַּ ד‬Gen 22:20; Cp), ‫( ֻל ַקּח‬Gen 3:23; Gp) and similar forms can only be understood as passive, in the N-stem the meaning is dependent on the context. If an external agent is implied, but not mentioned, the form expresses a passive meaning, as in ‫יוֹאשׁ ְבּשׁ ְֹמרוֹן‬ ָ ‫‘ וַ יִּ ָקּ ֵבר‬and Joash was buried in Samaria’ (2 Kgs 13:13). In subsequent, post-biblical stages of Hebrew, the Ablaut stems were no longer productive,29 and their passive meaning had to be covered by other stems. In light of the Classical Biblical Hebrew system, the N-stem would appear to be the natural candidate to fulfill this function, since it was the only stem that had previously been used to express passive meanings. And indeed, in Late Biblical Hebrew, Qumran Hebrew, Samaritan Hebrew, and Tannaitic Hebrew, the N-stem is continually employed to express passive meanings. Additionally, however, the tD-stem also acquired a passive meaning to varying degrees. This is rather surprising from an inner-Hebrew perspective, but it has convincingly been explained as an areal feature and Aramaic influence. For the corresponding Ablaut passive stems were lost in contemporaneous Aramaic 25 26 27

28 29

Gzella 2009, esp. p. 318. For a different view see Benton 2009, Benton 2012. Jenni 2012, pp. 149–150, 173; Arnold 2005, p. 78. Essentially, this is also the view of Siebesma 1991, pp. 167–168. Arnold 2005, p. 181; Gzella 2009, pp. 305–306. For Samaritan Hebrew: Florentin 1993, pp. 201–203, Florentin 2012, pp. 350–351; For Qumran Hebrew: Qimron 2018, pp. 182–184; For Tannaitic Hebrew: Bar-Asher 1999, pp. 65–69, where the Cp stem is still attested in verbal forms. Mor 2015, pp. 180–181 has an overview.

294

CHRISTIAN STADEL

as well, where the t-stems naturally extended their functional domain to the passive (Aramaic having no N-stem).30 If the Hebrew tD-stem did indeed take over passive meanings in postbiblical Hebrew, this would have led to a more substantial overlap between the meanings of the niphˁal and hitpaˁˁel, which, in turn, could have provided the basis for the contamination of the two stems that supposedly produced the Rabbinical and Samaritan nippəˁal and nitpaˁˁel. Unfortunately, however, the extent to which the tD-stem takes passive meanings in the stages following Classical Biblical Hebrew has never been investigated comprehensively. We shall present the status questionis for Late Biblical Hebrew and Tannaitic Hebrew, and shall then take a closer look at the Samaritan Hebrew and Dead Sea Scrolls evidence. In Late Biblical Hebrew, passive tD-forms are few and far between, and possible examples are contested.31 The most commonly cited cases are ‫יִת ַענּוּ‬ ְ ‫יהם‬ ֶ ‫וּמ ֲעוֺנ ֵֺת‬ ֵ ‘and because of their sins they are afflicted’ (Ps 107:17), ‫‘ ִהיא ִת ְת ַה ָ ֽלּל‬she shall be praised’ (Prov 31:30), and ‫‘ וַ יִ ְשׁ ַתּ ְכּחוּ ָב ִעיר‬and they were forgotten in the city’ (Qoh 8:10). If at all, Late Biblical Hebrew only evinces the beginning of the phenomenon. Presumably, it was not pervasive enough at this stage to allow for the contamination. Following the demise of post-biblical Hebrew, on the other hand, in the language of the Tannaim, passive tD-stem forms are far more common, though just how much has not been studied endétail.32 Examples from the Mishnah include ‫ית ַא ְכּלוּ‬ ְ ִ‫‘ ַה ְפּ ָד ִרים ֶשׁלֺּא נ‬the pieces of fat that had not been eaten’ (m. Tamid 2:5) and ‫ילּוּעים‬ ִ ‫יתח ְתּכוּ ַה ִדּ‬ ַ ִ‫יתפּ ְתּחוּ ַה ֲח ִביּוֹת נ‬ ַ ִ‫‘ נ‬the casks were opened, the gourds were cut’ (m. ῾Or. 3:8), but these can easily be multiplied. It is beyond the scope of the present article to investigate the exact extent of the phenomenon in Tannaitic Hebrew, but in what follows we shall determine whether Samaritan Hebrew and Qumran Hebrew are more like Late Biblical Hebrew or Mishnaic Hebrew with respect to passive uses of the tD-stem. The situation in Samaritan Hebrew is somewhat difficult to assess.33 For while it is true that the consonantal skeleton of the Samaritan Pentateuch 30 31

32

33

This is a well known fact that was noted by many, e.g., Florentin 1993, pp. 210–211. See, e.g., Waltke and O’Connor 1990, pp. 431–432 (§26.3a), Benton 2009, pp. 81–82, Jenni 2012, pp. 149–150. The usage has been duly noted in various publications, e.g., Segal 1908, pp. 677–678, Bar-Asher 1999, p. 64, Mor 2015, p. 181, but I know of no systematic study of the subject. As stated above, the development of a passive hitpaˁˁel is assumed to have been a prerequisite for the contamination of niphˁal and hitpaˁˁel in Samaritan Hebrew. Hence, passive nippəˁal forms are not considered here.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

295

was not yet completely fixed in late Second Temple times (a pertinent example is Exod 22:6, where the Samaritans read ‫ ונגנב‬for MT ‫)וְ גֻ נַּ ב‬, there were probably restrains that prevented Samaritan scribes from changing the text freely and systematically updating it linguistically. Given that this proposition is correct, linguistic changes that affected Samaritan Hebrew in the last two centuries BCE or so, were – by and large – only reflected in the pronunciation tradition if they did not necessitate a change of the consonantal text. As a consequence, relatively few attested passive hitpaˁˁel forms in the Samaritan Pentateuch do not necessarily mean that such forms were as uncommon in spoken Samaritan Hebrew. The Samaritan Hebrew pronunciation tradition regularly employs passive hitpaˁˁel forms in two morphological categories: Both primaey and mediaew roots read as Ablaut passives of the C-stem in the Tiberian tradition are consistently interpreted as hitpaˁˁel forms (with full anticipatory assimilation of the /t/),34 as in the following instances: Exod 29:27 ‫ שוק התרומה אשר הונף ואשר הורם‬šoq attērūma ēšår uwwnəfwēšåruwwråm (MT: ‫ הוּנַ ף‬and ‫)הוּרם‬ ָ֑ Gen 39:1 ‫ ויוסף הורד מצרימה‬wyūsəfuwwrədmiṣrīma (MT: ‫)הוּרד‬ ַ Gen 24:33 ‫ וישם לפניו לאכל‬wyuwwšåmalfnolkål (MT: ‫יּוּשׂם‬ ַ ַ‫)ו‬

The initial of the Perfects prohibits an interpretation as N-stem forms. In two extraordinary cases, such hitpaˁˁel Perfects are paralleled in the Tiberian tradition by a hiphˁil Perfect with active meaning and by a niphˁal Infinitive, respectively. The Samaritan forms convey a passive meaning: Lev 25:45 ‫ אשר הולדו בארצכם‬ēšåruwwlēdubreṣkimma (MT: ‫)הוֹלידוּ‬ ִ Num 12:15 ‫ עד האספה מרים‬ˁadiyysēfamaryåm (MT: ‫)ה ָא ֵסף‬ ֵ

Additionally, there are two instances of Samaritan hitpaˁˁel forms that involve a change in the consonantal text.35 The Samaritan forms come in lieu of Tiberian hitpaˁˁel forms with assimilated /t/: Num 21:27 ‫( תבנה ותתכונן עיד סיחון‬MT: ‫)וְ ִתכּוֹנֵ ן‬ Num 24:7 ‫( ותתנשא מלכותו‬MT: ‫)וְ ִתנַּ ֵשּׂא‬

In the first case a passive meaning is likely,36 in the second it is at least possible. In sum, Samaritan Hebrew systematically employs passive hitpaˁˁel forms in primaey and mediaew verbs, and three additional forms also 34 35 36

Ben-Ḥayyim 2000, p. 178; Florentin 1993, pp. 201–202. The forms were pointed out by Schorch 2008, p. 188 n. 45. Cf. Jenni 2012, p. 272.

296

CHRISTIAN STADEL

attest to such a usage outside these narrow categories. As mentioned, scribal restraints could have masked the real extent of the phenomenon. But even based on the present data, it is safe to state that the passive tD-stem is more common in Samaritan Hebrew than in Late Biblical Hebrew. To what extent tD-stem forms convey passive meanings in Qumran Hebrew has never been systematically investigated.37 In the biblical scrolls, two cases attest to the replacement of the puˁˁal by the passive hitpaˁˁel.38 These instances resemble the Samaritan Hebrew evidence in that they necessitated the change of a given text. They come from two adjacent verses in the Great Isaiah scroll: 1QIsaa LXIII, 28 ‫( ועל בורכים תשתעשעו‬MT Isa 66:12 ‫)תּ ָשׁעֳ ָ ֽשׁעוּ‬ ְ 1QIsaa LXIII, 29 ‫( וביר]ו[שלים תתנחמו‬MT Isa 66:13 ‫)תּנֻ ָ ֽחמוּ‬ ְ

In the non-biblical texts, more than 500 instances of tD-stem forms are attested (including parallel passages in different scrolls).39 It is impossible to determine for each and every instance whether it conveys a passive meaning. Approximately 110 forms lack sufficient context to establish their precise meanings, others are morphologically ambivalent (such as 1QS III, 4 ‫ ולוא יטהר במי נדה‬or 1QS X, 15 ‫להדשן בעדני תנובת תבל‬, which could be either N- or tD-stem forms), and with a few roots a passive meaning is impossible for semantic reasons (e.g., ‫)התהלך‬.40 Excluding all these, there remain roughly 250 instances of certain tD-stem forms that could potentially be passive. In one case, a tD-stem form is used parallel to an Ablaut passive: 4Q387 3 6–7 [+ 4Q385a 5 6] ‫]והורד [בימיהם גאון מרשיעי ברית ועבדי נאכר‬ [‫ ויתקרע ישראל בדור הה]וא‬ʻ[and] in their days the pride of those who condemn the covenant and of slaves of the foreign [will be brought down], and Israel will be torn in th[at] generationʼ41

In other instances, the tD-stem parallels an N-stem, but in itself this is not indicative of a passive meaning. In the following examples, an 37

38 39 40

41

Qimron 1986, pp. 48–49 mentions the phenomenon and provides a number of examples. The list in Qimron 2018, pp. 237–238 contains more examples than ours, but the book was published too late to be incorporated systematically. However, some examples listed there are ambiguous: 4Q174 iv 4, for example, need not be interpreted as passive. Kutscher 1974, p. 362. We follow the readings of Qimron 2010–2014. In some verbs which express a religious practice, it is almost impossible to decide whether contemporaneous speakers of Hebrew understood these to imply an external agent. For the most part, ‫התברך‬, ‫התקדש‬, and ‫ הצטדק‬have therefore not been considered. A comparable instance can be found in 1QS III, 8, but there the form ‫ יטהר‬is ambivalent and could also be a N-stem.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

297

external agent is clearly implied, and middle or reflexive meanings are thus excluded: 11Q5 XXII, 10–11 (‖ 4Q88 8 5–7) ‫נבחן אדם כדרכו איש כמעשיו ישתלם‬ ‫‘ סביב נכרתו צריך ציון ויתפזרו כול משנאיך‬Man is tested to his way, according to his deeds he will be rewarded. All around, oh Zion, your adversaries are cut off, and all those who hate you are scattered.’ 4Q385 2 3 (‖ 4Q386 1 I, 2; 4Q388 7 5) ‫וא[לה מתי יהיו ואיככה ישתלמו‬ ‫‘ חסדם‬and the]se, when will they come to be, and how will they be recompensed for their piety’ 4Q448 II, 8–9 ‫‘ ועל ממלכתכ יתברכ שמכ‬and may your name be blessed upon your kingdom’ 11QTa XLIX, 16 ‫‘ ובגדים ושקים ועורות יתכבסו‬and cloths and sacks and skins shall be washed’42 4Q266 11 7–8 ‫‘ במרד מלפני הרבים ישתלח‬in rebellion, he shall be send away from before the Many’43

In addition, numerous instances of the root √ḥšb in the tD-stem denote passive situations: 1QS III, 1 (‖ 4QSc 3 2) ‫‘ לוא חזק למשוב חיו ועם ישרים לוא יתחשב‬he is not strong enough to repent in his life and among the righteous he will not be reckoned’ 1QS III, 3–4 (‖ 4QSc 3 5) ‫‘ בעין תמימים לא יתחשב‬with the core of the faultless he will not be reckoned’ 4Q181 1 3–4 ‫כבודו הגיש מבני תבל ]לשרתו [להתחשב עמו בס]וד א[לים‬ ‘his glory has brought some of the children of the world [to serve him,] to be reckoned with him in the se[cret of the g]ods’ 4Q400 2 5–6 ‫‘ ו]אנו בשר בכבודנו[ מה נתחשב ]ב[ם‬and [we, flesh in our glory,] how shall we be reckoned [among] them’ 4Q491 11 I, 14 (‖ 4QHe 1 1–2) ‫אני עם אלים אתחשב ומכוני בעדת קודש‬ ‘I will be reckoned among the gods, and my abode is in the holy congregation’

While this list of certain passive tD-forms is relatively short, it proves that the stem could express a passive meaning in Qumran Hebrew. If this was the case, some other, semantically ambivalent, tD-stem forms could also have been intended as passives. This is especially true of forms that 42

43

Cp. Lev 13:55–56, where similar materials are washed and the tD-form ‫ ֻה ַכּ ֵבּס‬is employed. Presumably, the Participle in l. 14 ‫ ויצא המשתלח‬refers back to the sentence in question (which would explain the awkward tD-stem Participle, instead of the Ablaut passive D-stem) and has a passive meaning, too.

298

CHRISTIAN STADEL

would otherwise best be understood as middles, for example 1QS IX, 8 (‖ 4QSd VII, 8) ‫‘ אל יתערב הונם עם הון אנשי הרמיה‬their wealth shall not mix/shall not be admixed with the wealth of deceitful men.’ Of those roots that certainly express passive notions, most are unattested (√qrˁ, √šlm, √pzr, √šlḥ) or very rare (√ḥšb, √šˁšˁ, √kbs) in the tD-stem in Biblical Hebrew, while only two are common (√nḥm, √brk). Thus, the Qumran Hebrew passive tD-stem forms are not solely the result of an extended semantic range of existing forms; some of them constitute new forms with a new meaning. A close reading of possible examples thus confirms certain instances of a passive tD-stem meaning in both Samaritan Hebrew and Qumran Hebrew. The same phenomenon is also attested marginally in Late Biblical Hebrew and more extensively in Rabbinic Hebrew. Since the latter is not a direct descendant of Biblical Hebrew, the phenomenon is best interpreted as a typological similarity. Supposedly, it was triggered by two tendencies that affected the various stages of post-biblical Hebrew: The morphological loss of the Ablaut passive, which left a semantic category unrepresented, and the subsequent extension of the semantic range of the tD-stem, to cover the said category. The latter development was probably facilitated by the similar Aramaic stem system and increased the typological similarity between Hebrew and Aramaic. It also facilitated the contamination of the N- and tD-stems in Tannaitic Hebrew and Samaritan Hebrew, for it widened the semantic overlap between the two stems. In a multi-dimensional model of post-biblical Hebrew, the phenomenon is best explained by the wave model and/or seen as a Sprachbund feature. 2. THE DEMONSTRATIVE ‫הלז‬ The next case is much more complex. Apart from the usual ‫זה‬, ‫זו‬/‫זאת‬, and ‫אלה‬, Biblical Hebrew attests to another rare series of demonstratives (11 occurrences in total) that do not designate far deixis.44 The forms ‫ַה ָלּז‬ (csg), ‫( ַה ָלּזֶ ה‬msg), and ‫( ַה ֵלּזוּ‬fsg) are usually adnominal, but a pronominal usage is attested as well: Gen 24:65 ‫אתנוּ‬ ֵ֔ ‫י־ה ִ ֤אישׁ ַה ָלּזֶ ֙ה ַהה ֵֹלְ֤ך ַבּ ָשּׂ ֶד ֙ה ִל ְק ָר‬ ָ ‫ל־ה ֗ ֶע ֶבד ִ ֽמ‬ ָ ‫אמר ֶא‬ ֶ ֹ ‫וַ ֣תּ‬ Dan 8:16 ‫ת־ה ַמּ ְר ֶ ֽאה‬ ַ ‫יאל ָה ֵ ֥בן ְל ַה ָ ֖לּז ֶא‬ ֵ֕ ‫אמר גַּ ְב ִר‬ ַ֔ ֹ ‫וַ יּ‬

44

Garr 2008 offers a comprehensive overview and has references to previous studies.

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

299

The forms differ from the regular demonstratives in aspects of their syntax. They can function as attribute of an anarthrous noun and they immediately follow the noun, preceding other attributes: 1 Sam 14:1 ‫ל־הנַּ֙ ַע ֙ר נ ֵ ֹ֣שׂא ֵכ ֔ ָליו ְל ָ֗כה וְ נַ ְע ְבּ ָר ֙ה‬ ַ ‫ן־שׁאוּל֙ ֶא‬ ָ ‫אמר יוֹנָ ָ ֤תן ֶבּ‬ ֶ ֹ ‫וַ יְ ִ ֣הי ַהיּ֗ וֹם וַ ֨יּ‬ ַ ‫ֶא‬ ‫ל־מ ַצּ֣ב ְפּ ִל ְשׁ ִ֔תּים ֲא ֶ ֖שׁר ֵמ ֵע ֶ֣בר ַה ָלּ֑ז‬ Ezek 36:35 ‫ן־ע ֶ֑דן‬ ֵ ַ‫יְתה ְכּג‬ ֖ ָ ‫וְ ָא ְמ ֗רוּ ָה ָ ֤א ֶרץ ַה ֵ ֙לּ ֙זוּ ַהנְּ ַשׁ ָ֔מּה ָה‬

The use of the forms is restricted to quoted direct discourse, it is never anaphoric, and Randall Garr has argued quite convincingly that they function as medial deictics, i.e. they designate palpable objects at a moderate distance.45 These semantics are exemplified by the Goliath story, where the Philistine is first introduced to the pericope by the regular demonstrative as ‫ ָה ִ ֤אישׁ ָ ֽהע ֶֹל ֙ה ַה ֶ֔זּה‬and is then designated as ‫ה ְפּ ִל ְשׁ ִ ֣תּי ַה ֔ ָלּז‬, ַ since he was standing at some distance between the lines.46 1 Sam 17:25–26 ‫ית ֙ם ָה ִ ֤אישׁ ָ ֽהע ֶֹל ֙ה ַה ֶ֔זּה ִ ֛כּי ְל ָח ֵ ֥רף‬ ֶ ‫אמר ׀ ִ ֣אישׁ יִ ְשׂ ָר ֵ֗אל ַה ְרּ ִא‬ ֶ ֹ ‫וַ ֣יּ‬ ֶ֜ ‫ֶאת־יִ ְשׂ ָר ֵ ֖אל ע ֶֹל֑ה ְ ֽו ָ֠היָ ה ָה ִ֨אישׁ ֲא ֶשׁ‬ ‫ן־לוֹ‬ ֔ ‫יִתּ‬ ֶ ‫תּוֹ‬ ֙ ‫ת־בּ‬ ִ ‫ר־יַכּנּוּ יַ ְע ְשׁ ֶ ֥רנּוּ ַה ֶ ֣מּ ֶלְך ׀ ֣עֹ ֶשׁר גָּ ֗דוֹל וְ ֶא‬ ָ ‫אמר ָדּ ִ ֗וד ֶ ֽא‬ ֶ ֹ ‫וְ ֵא ֙ת ֵבּ֣ית ָא ִ֔ביו יַ ֲע ֶ ֥שׂה ָח ְפ ִ ֖שׁי ְבּיִ ְשׂ ָר ֵ ֽאל׃ וַ ֣יּ‬ ֒‫מר‬ ֹ ‫מּוֹ ֵלא‬ ֮ ‫ל־ה ֲאנָ ִ֞שׁים ָהע ְֹמ ִ ֣דים ִע‬ ַ ‫ישׁ ֲא ֶ ֤שׁר יַ ֶכּ ֙ה ֶא‬ ֙ ‫ַמה־יֵּ ָע ֶ֗שׂה ָל ִא‬ ‫ת־ה ְפּ ִל ְשׁ ִ ֣תּי ַה ֔ ָלּז וְ ֵה ִ ֥סיר ֶח ְר ָ ֖פּה ֵמ ַ ֣על יִ ְשׂ ָר ֵ ֑אל ִ ֣כּי ִ֗מי‬ ‫ֹלהים ַח ִיּֽים׃‬ ֥ ִ ‫ַה ְפּ ִל ְשׁ ִ ֤תּי ֶ ֽה ָע ֵרל֙ ַה ֶ֔זּה ִ ֣כּי ֵח ֵ ֔רף ַמ ַע ְר ֖כוֹת ֱא‬

The story of the Shunnamite woman (2 Kings 4) offers another good example for the medial deictic meaning.47 The distribution pattern of the forms in Biblical Hebrew does not lend itself to a diachronic explanation, since the series is attested in the classical stage of the language as well as in the transition period (Ezekiel) and in Late Biblical Hebrew (Daniel). And if we add the evidence from other post-biblical Hebrew varieties, the distribution pattern even becomes awkward. As in the previous section, we shall concentrate on Samaritan Hebrew and the Dead Sea Scrolls. Two aspects of the use of the form in the Samaritan Pentateuch are noteworthy. Biblical Hebrew attests only twice to the long (markedly masculine?) form ‫ה ָלּזֶ ה‬, ַ in Gen 24:65 and Gen 37:19, while the shorter ‫ ַה ָלּז‬is regularly found in the Prophets and Writings. In the Samaritan Pentateuch, however, the short ‫ הלז‬allåz is used in the two verses from Genesis. In other words, the Samaritans employed the more common form of the demonstrative, even though they obviously did not know this short form from the Prophets or Writings. Could it be that this usage reflected the 45 46 47

Garr 2008, pp. 387–389. Note that the verses are lacking in the Old Greek of 1 Reigns. Garr 2008, p. 388.

300

CHRISTIAN STADEL

living Samaritan Hebrew of the Second Temple period? While this might be the case, evidence from Byzantine times suggests otherwise. For in both verses, the Samaritan Targum manuscripts almost universally gloss ‫ הלז‬as ‫זהיה‬/‫זעיה‬, an adjective from the Aramaic root √zhy ‘to be joyous, happy.’48 According to the Samaritans, the verses speak of ‘the joyous man’ (‫האיש‬ ‫ )הלז‬that is Isaac and of ‘the joyous dreamer’ (‫ )בעל החלמות הלז‬that is Joseph. By the time of the Targum, the original meaning of ‫ הלז‬had been forgotten in Samaritan circles.49 But whether this situation is also representative of the Samaritan Hebrew of the Second Temple period is difficult to establish. The Qumran scrolls, on the other hand, present a straight forward picture: The non-biblical scrolls do not attest to the ‫ הלז‬series, but one of the biblical occurrences is fragmentarily preserved in 4QDanb 18 II, 6 (Dan 8:16). In the language of the literary scrolls from Qumran, demonstratives of the ‫ הלז‬series had apparently not been in use. Surprisingly, however, the documentary materials from the Judean Desert preserve an altogether different picture. For the demonstrative ‫הלז‬ as well as the feminine form ‫ הלזו‬are attested a total of 12 times in six Hebrew legal papyri from the Bar-Kokhba period, which belong to three different caches.50 The papyri 5/6Ḥev 44–46 from Nahal Hever record the details of three sequential business arrangements pertaining to two estates in Ein Gedi; they were drafted in quick succession by one and the same scribe and contain most of the examples (9 tokens).51 However, three additional attestations in deeds found in Wadi Murabbaˁat (Mur 24 deed 5 l. 8 and deed 6 l. 9) and Ketef Jericho (Jer 11 l. 2), respectively, proof that this was not an idiosyncrasy.52 As in Biblical Hebrew, the demonstratives occur in adnominal as well as pronominal use: 5/6Ḥev 44 25 ‫‘ על חשבון המחלקת הלזו‬on the account of that division’ 5/6Ḥev 45 28–29 ‫‘ עד סוף הזמן הלז‬until the end of this season’ 5/6Ḥev 45 13–14 ‫‘ הלז אחכרתי לך‬This I have leased to you.’

When used adnominally, the forms only occur with noun phrases formally marked as definite (allowing for some cases where ambiguous 48

49 50 51 52

The Samaritan Targum manuscript A to Gen 39:19 is exceptional in that it translates ‫‘ אזביסה‬the disgraceful,’ also a midrashic rendering, cf. Tal 2000, p. 15. Ben-Ḥayyim 2000, p. 237. Mor 2015, pp. 154–157 offers a comprehensive overview. Yadin et al. 2002, pp. 39–70. Yardeni 2000, pp. 107–112; 215–216 (erroneously labeled Jer 12).

VARIEGATION IN SECOND TEMPLE PERIOD HEBRE

301

orthography might mask another construction, e.g., 5/6Ḥev 45 18 ‫שבמקום‬ ‫)הלז‬, in contradistinction to the biblical example in 1 Sam 14:1.53 In another respect, the biblical and epigraphic materials differ diametrically: whereas in Biblical Hebrew, the demonstratives are only employed as deictics, never as anaphora, the opposite is true for the documents from the Judean Desert (though not all examples are unambiguous due to the fragmentary context),54 e.g.: Mur 24 5:6–8 ‫העפר שה]י[ה שלי בח>כ cII-422 551 730 18) καὶ (>M′) ἀνακλάσει M′ (s nom) C′cat 730cat 18cat 646cat These readings, whose attribution to the Samareitikon is hardly in doubt, will not be included presently. Also to be excluded are cases where an anonymous reading in M′ can be identified as belonging to the Samareitikon because it is identified as such in another passage: Lev 8:13 κιδάρεις] πιλία M′ Although the reading is anonymous in Lev 8:13, the noun πιλία “felt hat” is attributed to the Samareitikonin scholia on Exod 28:4 and Lev 8:9. The word is unexpected and rare in Greek, making it very likely that the textual source from which it is drawn is the same in all three passages. Both phenomena illustrate the well-known fact that the attribution of marginal readings was not handled systematically in the ancient sources. It is therefore reasonable to suppose that M′ transmits Samareitikonreadings sinenominein other passages as well. F. Field, OrigenisHexaplorumquaesupersunt,TomusI&II(Oxford: Clarendon, 1875), lxxxii–lxxxiv. A. Geiger, Nachgelassene Schriften, 4. Band (ed. Ludwig Geiger; Berlin: Gerschel, 1876), 121–126; S. Kohn, “Samareitikon und Septuaginta,” MGWJ 38 (1894): 1–7, 49– 67. See John W. Wevers, Levitikus(Septuagint II,2; Göttingen: Vandenhoeck & Ruprecht, 1986), 32, note 1. See Pummer, “The Greek Bible and the Samaritans,” 289-295.

318

JAN JOOSTEN

The readings to be discussed in this paper are of the same type as those collected by Fraenkel, but were not included in his list. Some may have been omitted due to oversight, others because Fraenkel did not yet have access to A. Tal’s critical edition of the Samaritan Targum.

ANONYMOUS MARGINAL READINGS IN CODEX M

AND ITS GROUP

In what follows, marginal readings in M′ that conform more or less markedly to the Samaritan Targum or reading tradition, yet are not marked out as such in the Göttingen apparatus, will be listed with a few remarks. Some of these readings had been signaled by Geiger, but a few others are presently connected to the Samareitikon for the first time. Lev 6:3(10)19 ‫ל־בּ ָשׂרוֹ‬ ְ ‫י־בד יִ ְל ַבּשׁ ַע‬ ַ ‫וּמ ְכנְ ֵס‬ ִ ‫וְ ָל ַבשׁ ַהכּ ֵֹהן ִמדּוֹ ַבד‬ LXX καὶ ἐνδύσεται ὁ ἱερεὺς χιτῶνα λινοῦν καὶ περισκελὲς λινοῦν ἐνδύσεται περὶ τὸ σῶμα αὐτοῦ The priest shall put on his linen vestments after putting on his linen undergarments next to his body χιτῶνα λινοῦν] ἐπενδύτην λευκόν M′ περισκελὲς λινοῦν] σκέπας λευκάς M′ SamTg20 ‫חוור‬/‫אבר‬/‫“ עבר‬white”

Instead of “linen vestments” and “linen undergarments,” the marginal reading has “a whitegarment”21 and “white coverings” respectively. The readings agree with the Samaritan Targum against all other available evidence. The reading is noted by Geiger.22 The interpretation finds an interesting parallel in the War Scroll, where priests are required to dress in white linen specifically: 1QM 7:9-11: When the battle lines are arranged against the enemy, battle line against battle line, there shall go forth from the middle open19

20

21

22

The MT will be quoted from BHS, the Septuagint text and its apparatus from the Göttingen edition. English translations follow the NRSV. The Samaritan Targum is quoted from Tal (see above, note 5). The textual tradition of the Samaritan Targum is very unstable. The word ἐπενδύτης is not part of the standard Septuagint vocabulary, but it is found in the kaigesection 2 Sam 13:18 and in a reading attributed to Theodotion in Exod 29:5. Geiger, NachgelasseneSchriften, 125.

BIBLICAL INTERPRETATION IN THE SAMAREITIKON AS EXEMPLIFIED

319

ing into the space between the battle lines seven priests of the sons of Aaron, dressed in fine white linen garments: a linen tunic and linen breeches (‫)בגדי שש לבן כתונת בד ומכנסי בד‬, and girded with a linen sash of twined fine linen, violet, purple and crimson, and a varicolored design, the work of a skillful workman, and decorated caps on their heads; the garments for battle, and they shall not take them into the sanctuary.

On the same principle, three more marginal readings can be attributed to the Samareitikon: Lev 16:4 λινοῦν 1o] … λευκόν M′ | λινῇ] … λευκῇ M′ 108 (= SamTg) 16:23 τὴν στολὴν τὴν λινῆν] τὰ φάρεα τὰ λευκά M′ (= SamTg)23

All these passages involve the translation of the Hebrew word ‫ בד‬as “white”. Lev 14:37 ‫ן־ה ִקּיר‬ ַ ‫יהן ָשׁ ָפל ִמ‬ ֶ ‫וּמ ְר ֵא‬ ַ καὶ ἡ ὄψις αὐτῶν ταπεινοτέρα τῶν τοίχων and if its appearance is deeper than the walls δυσειδής M′ SamTg ‫“ שני‬deteriorated”

One of the symptoms of a “leprous house” is the presence of reddish or greenish spots “deeper” than the wall according to MT and LXX. The marginal reading cannot with certainty be related to the word ταπεινοτέρα in the Septuagint, but doing so is the only way to make sense of the evidence. The reading implies that, instead of being spatially lower, the diseased spot in the house is “ugly, unsightly.” This interpretation corresponds to the Samaritan Targum but is found in no other witness to Lev 14:37. It seems to be based on a meaning of ‫ שפל‬in Aramaic and late Hebrew: “mean, lowly.”24

23

24

The rare word τὰ φάρεα (from φάρος “mantle”) for Hebrew ‫ בגדים‬is found also in marginal readings in M′ on Lev 6:4(11) and 8:2: 6:4(11) τὴν στολήν] τὰ φάρεα M′| στολὴν ἄλλην] φάρεα ἕτερα M′108 8:2 στολάς] φάρεα M′ The use of the same vocabulary makes it likely that these anonymous readings too go back to the Samareitikon. See e.g. Dan 4:14.

320

JAN JOOSTEN

Lev 16:26 ‫ת־ה ָשּׂ ִעיר ַל ֲעזָ אזֵ ל‬ ַ ‫וְ ַה ְמ ַשׁ ֵלּ ַח ֶא‬ ὁ ἐξαποστέλλων τὸν χίμαρον τὸν διεσταλμένον εἰς ἄφεσιν25 and the one who sent the goat away to Azazel ὁ ἐξαποστελλόμενος ἅμα τῷ χιμάρῳ εἰς ἀζαζήλ M′ SamTgNMB ‫ודמשתלח עם צפירה לעזאזל‬

Instead of referring to the man who took out the goat for Azazel to the desert as “the one who sent away the goat,” the marginal reading refers to him as “the one who was sent away with the goat,” reflecting perhaps a Pual participle instead of the MT’s Piel, and taking the particle as the preposition meaning “with” and not as the notaaccusativi. In both these points, M’s reading finds precise parallels in variant readings occurring in a few manuscripts of the Samaritan Targum. The interpretation seems to be found nowhere else. Note should be taken that the reading tradition of the Samaritans interprets the biblical text in the same way as the MT, with an active verb and a marker of the accusative.26 Lev 19:18 ‫א־תטֹּר‬ ִ ֹ ‫א־תקֹּם וְ ל‬ ִ ֹ‫ל‬ καὶ οὐκ ἐκδικᾶταί σου ἡ χείρ, καὶ οὐ μηνιεῖς you shall not take vengeance or bear a grudge init – μηνιεῖς] οὐ στήσῃ οὐδὲ παρατηρήσῃ M′ SamTg ‫לא תקום ולא תטור‬ Oral reading lå̄ tiqqom wlå̄ tiṭṭor “you will not stand up and bear a grudge”27

The marginal reading’s interpretation of the first verb as a form of the root ‫“ קום‬to stand up” instead of ‫“ נקם‬to take vengeance” is surprising in view of the context. It is paralleled in the reading tradition according to the analysis of Ben-Hayyim.28 Ben-Hayyim writes: 25 26

27

28

Note that ‫ לעזאזל‬is translated twice in the Septuagint version. In Samaritan Hebrew, the two particles are vocalized differently even when they are not followed by suffixes: ᾿at=“with,” but ᾿it= marker of the accusative. The second verb, tiṭṭor, is interpreted in various ways in the Samaritan tradition: “to keep a grudge,” “to do battle with” (apparently on the basis of the noun ‫“ טור‬row”), “to ambush.” See A. Tal, ADictionaryofSamaritanAramaic(Handbuch der Orientalistik 50.1; Leiden: Brill, 2000). Ben-Ḥayyim, LOT 245;Grammar, 152. The qalof peh-nun verbs is vocalized differently, Grammar, 145 (note yiqqåm “he will avenge” Deut 32:43).

BIBLICAL INTERPRETATION IN THE SAMAREITIKON AS EXEMPLIFIED

321

While the Samaritan translations of this verse do not differ from the Jewish exegetical tradition, there is nothing to be learned from them about the traditions of pronunciation.

The anonymous Greek reading in M′ provides exactly what Ben-Hayyim couldn’t find: a Samaritan translation that echoes the reading tradition. Incidentally, the same interpretation may underlie the Samaritan Targum’s majority reading ‫תקום‬. Since Samaritan Aramaic mostly uses the root ‫גבי‬ for “to avenge,” the form ‫ תקום‬may derive from ‫קום‬.29 The reading οὐ στήσῃ was attributed to the Samareitikon by Geiger, but without concrete evidence.30 Lev 26:9 ‫יכם‬ ֶ ‫יתי ֲא ֵל‬ ִ ִ‫וּפנ‬ ָ καὶ ἐπιβλέψω ἐφ᾿ ὑμᾶς I will turn to you ἐπιβλέψω ἐφ᾿ ὑμᾶς] εὐιλατεύσω ὑμῖν M′ SamTg ‫אתרחי לידכון‬

MT’s “I will turn to you” is interpreted in the marginal reading as “I will be merciful to you,” in accord with the Samaritan Targum. The interpretation seems to reflect a concern not to create too anthropomorphic a picture of God. Lev 26:13 ‫מטֹת ֻע ְלּ ֶכם‬ ֹ ‫וָ ֶא ְשׁבֹּר‬ καὶ συνέτριψα τὸν δεσμὸν τοῦ ζυγοῦ ὑμῶν I have broken the bars (LXX ‘the bond’) of your yoke δεσμόν] ῥάβδους M31 SamTg ‫“ אטרי‬sticks” Oral reading måṭṭot

Instead of MT’s “the bars of your yoke” and the Septuagint’s “the bond of your yoke,” the marginal reading has “the rods of your yoke.” The interpretation rests on a different vocalization of the Hebrew text 29 30 31

This was pointed out by Christian Stadel at the conference. Geiger, NachgelasseneSchriften, 125. In the last part of Leviticus, the M group is reduced to codex M only, the other manuscripts being lacunary.

322

JAN JOOSTEN

attested also in the Samaritan reading tradition and Samaritan Targum. The reading was signaled by Geiger.32 Lev 26:16 ‫ף־אנִ י ֶא ֱע ֶשׂה־זֹּאת ָל ֶכם‬ ֲ ‫ַא‬ καὶ ἐγὼ ποιήσω οὕτως ὑμῖν I too will do thus to you καί 1o] ὀργιζόμενος M SamTgM1 ‫רגז‬, SamTgN ‫ראגז‬

The marginal reading interprets ‫ אף‬as a verbal form from a root meaning “to be angry.” Exactly the same interpretation is found in a few manuscripts of the Samaritan Targum. Lev 26:19 ‫וְ ָשׁ ַב ְר ִתּי ֶאת־גְּ אוֹן ֻעזְּ ֶכם‬ καὶ συντρίψω τὴν ὕβριν τῆς ὑπερηφανίας ὑμῶν and I will smash the insolence of your arrogance τὴν ὕβριν τῆς ὑπερηφανίας] τοὺς ὑπερφερεῖς τῆς ἰσχύος M SamTgAVNMECB ‫“ חיולי תקיפיהון‬the mighty ones of their strong ones”

Instead of MT’s “the insolence of your arrogance”, the marginal reading has a plural (perhaps reflecting ‫ )גאוני‬which is interpreted in reference to a group: “the insolent ones of (your) strength.” The plural is found in none of the other witness to this verse, but it does find an echo in some manuscripts of the Samaritan Targum.

ANALYSIS The profile of the readings in the sample is globally similar to that of marked Samareitikon readings. 1. Like the marked readings, these anonymous variants agree with the Samaritan tradition of the Pentateuch in its various components. There are no cases in our sample of agreement with the consonantal text of the Samaritan Pentateuch against the MT, but this is just a coincidence. The 32

Geiger, NachgelasseneSchriften, 125.

BIBLICAL INTERPRETATION IN THE SAMAREITIKON AS EXEMPLIFIED

323

anonymous reading in Lev 19:25 – excluded from our sample because it was already identified by Detlef Fraenkel – agrees with SP’s ‫“ להאסיף‬to gather” against MT’s ‫הוֹסיף‬ ִ ‫“ ְל‬to increase”: πρόσθεμα - αὐτοῦ] καὶ συνάξετε ὑμῖν τὴν πρόσοδον αὐτοῦ M′: cf SamAram

Several of our readings agree with the Samaritan reading tradition against the Tiberian vocalization: in Lev 26:13 the reading reflects måṭṭot “sticks” with the Samaritan reading against the MT’s mōṭōt“beams,” and in Lev 19:18 the reading reflects a form of the root ‫ קום‬with the reading tradition against the MT which derives the form from the root ‫נקם‬. In 19:18 the agreement of the anonymous reading with the Samaritan Targum cannot be established, but in all the other cases it can. As was realized already by the very first students of the scholia reflecting the Samareitikon,the Samaritan Targum is the touchstone and the constant companion of the Greek readings. It is important to realize, however, that the Samaritan tradition is not entirely unified. In some instances the reading tradition and the consonantal text of the Samaritan Pentateuch do not agree. In other cases the Targums diverge from the oral reading. Finally, the manuscripts of the Samaritan Targum differ very often, and sometimes dramatically, among themselves. Interesting cases in this regard are the reading in Lev 16:26, which agrees pointedly with certain manuscripts of the Samaritan Targum while other manuscripts agree with the MT and all other witnesses; and the reading in Lev 26:19 where the reading and some Targum manuscripts appear to agree against the entire textual tradition. 2. Again like the marked Samareitikon readings, the readings in our sample clearly relate to the Septuagint. The Samareitikon was not a fresh translation, but a recension of the Old Greek version. The relationship is clearly visible in the use of the verbs ἐξαποστέλλω in Lev 16:26 and εὐιλατεύω in Lev 26:9, two items of typical Septuagint vocabulary that are rare in other texts. At the same time, the readings contain several lexemes that are unattested in the Septuagint, and rare in Greek generally: φάρος “mantle” in 6:4(11), 8:2, and 16:23 joins such curious items typifying the marked readings as πιλίον “felt hat” in Exod 28:4, and μαδίζω “to pluck or singe bare” in Lev 1:17. The readings tend to be more word-for-word than the Septuagint, as can be seen in the sample in Lev 19:18. This tendency characterized marked Samareitikon readings, as well as the fragmentary manuscripts enumerated above. Nevertheless, Samareitikon readings may also diverge from the

324

JAN JOOSTEN

Hebrew and the Old Greek for theological or halakhic reasons. A good example of the former in our sample is Lev 26:9, where the received Greek rendering may have been too literal to the taste of Samaritan tradents. An example of the latter may perhaps be found in the rendering of bad as λευκός “white,” which corresponds to Samaritan practice until today. Both types of non-literal renderings are also found among marked Samareitikon readings. 3. In addition to theology and religious law, an important source of variation in the Samaritan tradition is what may be called “creative philology”: polysemic lexemes are reinterpreted against their contextual meaning, new meanings are derived on the basis of bilateral roots, Aramaic or late Hebrew meanings are attributed to the words of the Torah. One wonders sometimes what the impetus for the interpretations might be, other than the sheer pleasure of twisting the biblical text into a new direction. In our sample, this creative approach is particularly manifest in the rendering of ‫“ אף‬also” as “being angry” in 26:16, but it also transpires in Lev 19:18 and 16:26. In Lev 14:37 the rendering of ‫“ שפל‬low” as δυσειδής “ugly” may be due to interference from Aramaic or post-Biblical Hebrew. CONCLUSION The principal value of the present investigation is that of completing the dossier of the Samareitikon.The Greek version of the Samaritans has been neglected in biblical research and deserves to be studied much more intensively. But it can only be studied after it has been delineated as fully and accurately as possible. Beyond this primary objective, this paper illustrates the coherence of the Samaritan tradition of the Pentateuch in its various branches. The consonantal text, the reading tradition, the Samaritan Targum, and the remains of the Samareitikon form a tightly-knit family, each of whose members helps to interpret the others. Only the Arabic translation stands somewhat apart in this respect. The Samaritan Pentateuch is to be studied holistically, taking into account all textual material including the oral reading traditions and the ancient versions – Aramaic, Greek, and Arabic. The Samaritan Pentateuch is to be explored from within the Samaritan tradition. Reducing it to a consonantal text, as biblical scholars tend to do, is unhelpful. While the consonantal text of the Samaritan Pentateuch can be traced to the first century BCE, and the reading tradition has been cogently argued to stretch back to the Second Temple period, the Samaritan Targum and

BIBLICAL INTERPRETATION IN THE SAMAREITIKON AS EXEMPLIFIED

325

the Samareitikon are hard to date. In an earlier publication I have argued for an early second century CE date for the latter and a somewhat earlier date for the former. The strong coherence among the different components is consonant with this time frame. The roots of the Samaritan tradition as represented by the reading tradition, the Targum, and the Greek version are to be sought for in the Roman or Late Hellenistic period. Although no obvious connection to any of the Qumran compositions has come to light, the fact that the consonantal text of the Samaritan Pentateuch finds clear fore-runners among the DSS makes it likely that the interpretive dimensions of this tradition too may receive illumination from the study of the Scrolls, and shed some light on them in return. The agreement in regard to the color of priestly vestments is suggestive in this respect although in isolation it does not mean much. More research is needed to explore other possible similarities between the Qumran scrolls and the Samaritan tradition in this comprehensive approach.

Contributions to Biblical Exegesis and Theology 1. J.A. Loader, A Tale of Two Cities, Sodom and Gomorrah in the Old Testament, earlyJewishandearlyChristianTraditions, Kampen, 1990 2. P.W. Van der Horst, Ancient Jewish Epitaphs. An Introductory Survey of a MillenniumofJewishFuneraryEpigraphy(300BCB-700CE), Kampen, 1991 3. E. Talstra, Solomon’s Prayer. Synchrony and Diachrony in the Composition of 1 Kings8,14-61, Kampen, 1993 4. R. Stahl, Von Weltengagement zu Weltüberwindung: Theologische Positionen im Danielbuch, Kampen, 1994 5. J.N. Bremmer, SacredHistoryandSacredTextsinearlyJudaism.ASymposiumin HonourofA.S.vanderWoude, Kampen, 1992 6. K. Larkin, The Eschatology of Second Zechariah: A Study of the Formation of a MantologicalWisdomAnthology, Kampen, 1994 7. B. Aland, NewTestamentTextualCriticism,ExegesisandChurchHistory:ADiscussionofMethods, Kampen, 1994 8. P.W. Van der Horst, Hellenism-Judaism-Christianity: Essays on their Interaction, Kampen, Second Enlarged Edition, 1998 9. C. Houtman, Der Pentateuch: die Geschichte seiner Erforschung neben einer Auswertung, Kampen, 1994 10. J. Van Seters, The Life of Moses. The Yahwist as Historian in Exodus-Numbers, Kampen, 1994 11. Tj. Baarda, EssaysontheDiatessaron, Kampen, 1994 12. Gert J. Steyn, Septuagint Quotations in the Context of the Petrine and Pauline SpeechesoftheActaApostolorum, Kampen, 1995 13. D.V. Edelman, TheTriumphofElohim,FromYahwismstoJudaisms, Kampen, 1995 14. J.E. Revell, TheDesignationoftheIndividual.ExpressiveUsageinBiblicalNarrative, Kampen, 1996 15. M. Menken, OldTestamentQuotationsintheFourthGospel, Kampen, 1996 16. V. Koperski, The Knowledge of Christ Jesus my Lord. The High Christology of Philippians3:7-11, Kampen, 1996 17. M.C. De Boer, JohanninePerspectivesontheDeathofJesus, Kampen, 1996 18. R.D. Anderson, AncientRhetoricalTheoryandPaul, Revised edition, Leuven, 1998 19. L.C. Jonker, Exclusivity and Variety, Perspectives on Multi-dimensional Exegesis, Kampen, 1996 20. L.V. Rutgers, TheHiddenHeritageofDiasporaJudaism, Leuven, 1998 21. K. van der Toorn (ed.), TheImageandtheBook, Leuven, 1998 22. L.V. Rutgers, P.W. van der Horst (eds.), The Use of Sacred Books in the Ancient World, Leuven, 1998 23. E.R. Ekblad Jr., Isaiah’sServantPoemsAccordingtotheSeptuagint.AnExegetical andTheologicalStudy, Leuven, 1999 24. R.D. Anderson Jr., GlossaryofGreekRhetoricalTerms, Leuven, 2000 25. T. Stordalen, EchoesofEden, Leuven, 2000 26. H. Lalleman-de Winkel, JeremiahinPropheticTradition, Leuven, 2000 27. J.F.M. Smit, About the Idol Offerings. Rhetoric, Social Context and Theology of Paul’sDiscourseinFirstCorinthians8:1-11:1, Leuven, 2000 28. T.J. Horner, ListeningtoTrypho.JustinMartyr’sDialogueReconsidered, Leuven, 2001 29. D.G. Powers, Salvation through Participation. An Examination of the Notion of the Believers’CorporateUnitywithChristinEarlyChristianSoteriology, Leuven, 2001 30. J.S. Kloppenborg, P. Hoffmann, J.M. Robinson, M.C. Moreland (eds.), TheSayings GospelQinGreekandEnglishwithParallelsfromtheGospelsofMarkandThomas, Leuven, 2001 31. M.K. Birge, TheLanguageofBelonging.ARhetoricalAnalysisofKinshipLanguage inFirstCorinthians, Leuven, 2004

32. P.W. van der Horst, JaphethintheTentsofShem.StudiesonJewishHellenisminAntiquity, Leuven, 2002 33. P.W. van der Horst, M.J.J. Menken, J.F.M. Smit, G. van Oyen (eds.), Persuasionand DissuasioninEarlyChristianity,AncientJudaism,andHellenism, Leuven, 2003 34. L.J. Lietaert Peerbolte, PaultheMissionary, Leuven, 2003 35. L.M. Teugels, Bibleandmidrash.TheStoryof‘TheWooingofRebekah’ (Gen. 24), Leuven, 2004 36. H.W. Shin, TextualCriticismandtheSynopticProbleminHistoricalJesusResearch. TheSearchforValidCriteria, Leuven, 2004 37. A. Volgers, C. Zamagni (eds.), Erotapokriseis. Early Christian Question-andAnswerLiteratureinContext, Leuven, 2004 38. L.E. Galloway, FreedomintheGospel.Paul’sExemplumin1Cor9inConversation withtheDiscoursesofEpictetusandPhilo, Leuven, 2004 39. C. Houtman, K. Spronk, EinHelddesGlaubens?RezeptionsgeschichtlicheStudien zudenSimson-Erzählungen, Leuven, 2004 40. H. Kahana, Esther. Juxtaposition of the Septuagint Translation with the Hebrew Text, Leuven, 2005 41. V.A. Pizzuto, A Cosmic Leap of Faith. An Authorial, Structural, and Theological InvestigationoftheCosmicChristologyinCol1:15-20, Leuven, 2005 42. B.J. Koet, DreamsandScriptureinLuke-Acts.CollectedEssays, Leuven, 2006 43. P.C Beentjes. “HappytheOneWhoMeditatesonWisdom”(SIR.14,20).Collected EssaysontheBookofBenSira, Leuven, 2006 44. R. Roukema, L.J. Lietaert Peerbolte, K. Spronk, J.W. Wesselius (eds.), TheInterpretationofExodus.StudiesinHonourofCornelisHoutman, Leuven, 2006 45. G. van Oyen, T. Shepherd (eds.), TheTrialandDeathofJesus.EssaysonthePassion NarrativeinMark, Leuven, 2006 46. B. Thettayil, InSpiritandTruth.AnExegeticalStudyofJohn4:19-26andaTheological Investigation of the Replacement Theme in the Fourth Gospel, Leuven, 2007 47. T.A.W. van der Louw, TransformationsintheSeptuagint.TowardsanInteractionof SeptuagintStudiesandTranslationStudies, Leuven, 2007 48. W. Hilbrands, Heilige oder Hure? Die Rezeptionsgeschichte von Juda und Tamar (Genesis38)vonderAntikebiszurReformationszeit, Leuven, 2007 49. J. Joosten, P.J. Tomson (eds.), VocesBiblicae.SeptuagintGreekanditsSignificance fortheNewTestament, Leuven, 2007 50. A. Aejmelaeus, OntheTrailoftheSeptuagintTranslators.CollectedEssays, Leuven, 2007 51. S. Janse, “You are My Son”. The Reception History of Psalm 2 in Early Judaism andtheEarlyChurch, Leuven, 2009 52. K. De Troyer, A. Lange, L.L. Schulte (eds.), ProphecyaftertheProphets?TheContributionoftheDeadSeaScrollstotheUnderstandingofBiblicalandExtra-Biblical Prophecy, Leuven, 2009 53. C.M. Tuckett (ed.), FeastsandFestivals, Leuven, 2009 54. M. Labahn, O. Lehtipuu (eds.), AnthropologyintheNewTestamentanditsAncient Context, Leuven, 2010 55. A. van der Kooij, M. van der Meer (eds.), The Old Greek of Isaiah: Issues and Perspectives, Leuven, 2010 56. J. Smith, TranslatedHallelujehs.ALinguisticandExegeticalCommentaryonSelect SeptuagintPsalms, Leuven, 2011 57. N. Dávid, A. Lange (eds.), QumranandtheBible.StudyingtheJewishandChristian ScripturesinLightoftheDeadSeaScrolls, Leuven, 2010 58. J. Chanikuzhy, Jesus,theEschatologicalTemple.AnExegeticalStudyofJn2,13-22in theLightofthePre70C.E.EschatologicalTempleHopesandtheSynopticTemple Action, Leuven, 2011

59. H. Wenzel, ReadingZechariahwithZechariah1:1–6astheIntroductiontotheEntire Book, Leuven, 2011 60. M. Labahn, O. Lehtipuu (eds.), ImageryintheBookyofRevelation, Leuven, 2011 61. K. De Troyer, A. Lange, J.S. Adcock (eds.), TheQumranLegalTextsbetweenthe HebrewBibleandItsInterpretation, Leuven, 2011 62. B. Lang, Buch der Kriege – Buch des Himmels. Kleine Schriften zur Exegese und Theologie, Leuven, 2011 63. H.-J. Inkelaar, Conflict over Wisdom. The Theme of 1 Corinthians 1-4 Rooted in Scripture, Leuven, 2011 64. K.-J. Lee, TheAuthorityandAuthorizationofTorahinthePersionPeriod, Leuven, 2011 65. K.M. Rochester, PropheticMinistryinJeremiahandEzekiel, Leuven, 2012 66. T. Law, A. Salvesen (eds.), GreekScriptureandtheRabbis, Leuven, 2012 67. K. Finsterbusch, A. Lange (eds.), WhatisBible?, Leuven, 2012 68. J. Cook, A. van der Kooij, Law,Prophets,andWisdom.OntheProvenanceofTranslatorsandtheirBooksintheSeptuagintVersion, Leuven, 2012 69. P.N. De Andrado, The Akedah Servant Complex. The Soteriological Linkage of Genesis22andIsaiah53inAncientJewishandEarlyChristianWritings, Leuven, 2013 70. F. Shaw, TheEarliestNon-MysticalJewishUseofΙαω, Leuven, 2014 71. E. Blachman, The Transformation of Tamar (Genesis 38) in the History of Jewish Interpretation, Leuven, 2013 72. K. De Troyer, T. Law, M. Liljeström (eds.), IntheFootstepsofSherlockHolmes.Studies intheBiblicalTextinHonourofAnneliAejmelaeus, Leuven, 2014 73. T. Do, Re-thinkingtheDeathofJesus.AnExegeticalandTheologicalStudyofHilasmos andAgapein1John2:1-2and4:7-10, Leuven, 2014 74. T. Miller, ThreeVersionsofEsther.TheirRelationshiptoAnti-SemiticandFeminist CritiqueoftheStory, Leuven, 2014 75. E.B. Tracy, SeeMe!HearMe!Divine/HumanRelationalDialogueinGenesis, Leuven, 2014 76. J.D. Findlay, FromProphettoPriest.TheCharacterizationofAaroninthePentateuch, Leuven, forthcoming 77. M.J.J. Menken, StudiesinJohn’sGospelandEpistles.CollectedEssays, Leuven, 2015 78. L.L. Schulte, MyShepherd,thoughYouDonotKnowMe.ThePersianRoyalPropagandaModelintheNehemiahMemoir, Leuven, 2016 79. S.E. Humble, ADivineRoundTrip.TheLiteraryandChristologicalFunctionofthe Descent/AscentLeitmotifintheGospelofJohn, Leuven, 2016 80. R.D. Miller, BetweenIsraeliteReligionandOldTestamentTheology.EssaysonArchaeology,History,andHermeneutics, Leuven, 2016 81. L. Dequeker, StudiaHierosolymitana, Leuven, 2016 82. K. Finsterbusch, A. Lange (eds.), Texts and Contexts of Jeremiah. The Exegesis of Jeremiah1and10inLightofTextandReceptionHistory, Leuven, 2016 83. J.S. Adcock, “OhGodofBattles!StealMySoldiers’Hearts!”AStudyoftheHebrew andGreekTextFormsofJeremiah10:1-18, Leuven, 2017 84. R. Müller, J. Pakkala (eds.), InsightsintoEditingintheHebrewBibleandtheAncient Near East. What Does Documented Evidence Tell Us about the Transmission of AuthoritativeTexts?, Leuven, 2017 85. R. Burnet, D. Luciani, G. van Oyen (eds.), TheEpistletotheHebrews.Writingatthe Borders, Leuven, 2016 86. M.K. Korada, TheRationaleforAniconismintheOldTestament.AStudyofSelect Texts, Leuven, 2017 87. P.C. Beentjes, “WithAllYourSoulFeartheLord”(Sir.7:27).CollectedEssayson theBookofBenSiraII, Leuven, 2017 88. B.J. Koet, A.L.H.M. van Wieringen (eds.), Multiple Teachers in Biblical Texts, Leuven, 2017

89. T. Elgvin, The Literary Growth of the Song of Songs during the Hasmonean and Early-HerodianPeriods, Leuven, 2018 90. D.C. Smith, The Role of Mothers in he Genealogical Lists of Jacob’s Sons, Leuven, 2018 91. V.P. Chiraparamban, TheManifestationofGod’sMercifulJustice.ATheocentricReading ofRomans3-21-26, Leuven, 2018 92. P. Paul, BeyondtheBreach.AnExegeticalStudyofJohn4:1-42asaTextofJewishSamaritanReconciliation, forthcoming 93. I. Fröhlich, DavidinCulturalMemory, Leuven, 2019

PRINTED ON PERMANENT PAPER

• IMPRIME

SUR PAPIER PERMANENT

N.V. PEETERS S.A., WAROTSTRAAT

• GEDRUKT

OP DUURZAAM PAPIER

50, B-3020 HERENT

- ISO 9706