• Author / Uploaded
• aslancm

• Categories
• Manuscritos del Mar Muerto
• Septuaginta
• Biblia hebrea
• Talmud
• Crítica textual

Citation preview

TEXTUAL CRITICISM OF THE HEBREW BIBLE

TEXTUAL CRITICISM OF THE HEBREW BIBLE THIRD EDITION, REVISED AND EXPANDED

EMANUEL TOY

Fortress Press Minneapolis

For Lika, companion, friend, and source of inspiration

4QCanta (Cant 4:1)

TEXTUAL CRITICISM OF THE HEBREW BIBLE Third Edition, Revised and Expanded Copyright © 2012 Fortress Press. All rights reserved. Except for brief quotations in critical articles or reviews, no part of this book may be reproduced in any manner without prior written permission from the publisher. Visit www.augsburgfortress.org/ copyrights/ or write to Permissions, Augsburg Fortress, Box 1209, Minneapolis, MN 55440. Cover image: The large Isaiah Scroll (lQisa•), cols. IV-VII © Jim Hollander I epa I Corbis Cover design: Ivy Palmer Skrade Book design: Janice Karnis

Library of Congress Cataloging-in-Publication Data is available ISBN 978-0-8006-9664-1 The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences- Permanence of Paper for Printed Library Materials, ANSI 2329.48-1984. Manufactured in the U.S.A. 15 14 13 12 2 3 4 56 7 8 9 10

CONTENTS List of Tables

xiii

List of Plates

xv

Preface Abbreviations and Sources

xvii XIX

Periodicals, Reference Works, and Serials

xxv

Bibliographical Abbreviations

xxix

System of Transliteration

!vi

Brief Didactic Guide

!vii

Chapter 1: Introduction

1

The Need for Textual Criticism of Hebrew-Aramaic Scripture

2

1. Differences among the Many Textual Witnesses a. Sequence of Books b. Chapter Division c. Layout of the Text d. Verse Division e. Single Letters and Words f. Vocalization and Accentuation g. Notes of the Masorah h. Different Editions Based on the Same Manuscript j. Differences Due to Printing Errors 2. Mistakes, Corrections, and Changes in the Texts, Including m 3. m Does Not Reflect the "Original Text" of the Biblical Books 4. Differences between Inner-Biblical Parallel Texts in m B. A Modem Approach to Textual Criticism c. Beginnings of Critical Inquiry into the Scripture Text D. Text, Canon, and Sacred Status E. Subjectivity of This Book

3 4 4 5 6 6 7 7 7 8 9 11 12 17 19 20 22

A.

Chapter 2: Textual Witnesses

23

I. Hebrew Witnesses

23 24 26 26 27 36

A.

Proto-Masoretic Texts and the Masoretic Text (m-Group) 1. The Consonantal Framework: Prato-Masoretic Texts and m a. Background b. Development of the Consonantal Text c. Origin

Contents

VI

d. Evidence e. Significance of the Differences between Medieval Manuscripts 2. Vocalization a. Background b. Systems of Vocalization c. Differences between the Vocalization Systems d. Character of the Tiberian Vocalization 3. Para-Textual Elements a. Text Divisions: Sections, Verses, Chapters, and Parashot b. Pisqah beem?a · pasuq c. Inverted Nunim d. Extraordinary Points (Puncta Extraordinaria) e. Suspended Letters (Litterae Suspensae) f. Special Letters g. Ketib-Qere h. Sebirin i. Corrections of the Scribes j. Omission of the Scribes k. Layout 4. Accentuation 5. Apparatus of the Masorah a. Content b. Masoretic Handbooks c. Editions of the Masorah d. Importance of the Masorah 6. Editions of m B. Pre-Samaritan Texts and the Samaritan Pentateuch (w.-Group) 1. 2. 3. 4.

Background Date and Origin Manuscripts and Editions Nature of the Text a. Early (Pre-Samaritan) Elements in w. a. Editorial Changes Small Harmonizing Alterations y. Linguistic Corrections 8. Small Content Differences E. Linguistic Differences b. Samaritan Elements a. Ideological Changes Changes c. Orthography

37 38 39 40 42 43 46 47 48 50 51 51 52 53 54 59 59 61 61 62 65 66 68 69 70 70 74 75 76 78 79 80 80 82 83 85 86 87 87 88 89

Contents 5. Pre-Samaritan Texts

c. The Biblical Texts Found in the Judean Desert 1. 2. 3. 4. 5.

Background Evidence Dating Publication Texts Written in the Qumran Scribal Practice a. Orthographic Features b. Morphological Features c. Scribal Features d. Contextual Adaptations e. Consistency and Statistical Analysis 6. Variants in the Qumran Scrolls 7. Classification of the Scrolls According to Textual Character 8. Contribution of the Judean Desert Scrolls to Biblical Research D. Additional Witnesses 1. Silver Rolls from Ketef Hinnom 2. The Nash Papyrus 3. Tefillin and Mezuzot from the Judean Desert 4. The Severus Scroll and R. Meir's Torah 5. Non-Biblical Sources: Quotations and "Rewritten Scripture" Texts 6. Texts That Have Been Lost II. The Ancient Translations A. Use of the Ancient Translations in Textual Criticism 1. Background 2. Exegesis a. Linguistic Exegesis b. Contextual Exegesis c. Theological Exegesis d. Midrashic Tendencies 3. Representation of Hebrew Constructions in the Translation 4. Inner-Translational Phenomena 5. Reconstruction of the Hebrew Source of the Translations B.Evidence 1. a. Name and Nature b. Scope c. Sequence of the Books d. Original Form Origin, and Date e. Evidence a. Direct Witnesses

vii 90 93 94 95 98 99 100 101 102 103 103 104 105 107 110 111 111 111 112 112 114 115 115 115 115 117 118 119 120 122 122 122 122 127 127 128 129 129 131 132 132

Contents

Vlll

Indirect Witnesses: Daughter Translations f. Editions g. Auxiliary Tools and Its Text-Critical Value h. Hebrew Source 2. Revisions of the Septuagint a. Background b. Emergence of the Revisions c. Nature of the Revisions d. Pre-Hexaplaric Revisions a. Kaige-Theodotion

e. Hexapla f. Post-Hexaplaric Revisions: Lucian 3. Targumim ( p. 224; 40:7-8 ..... pp. 223--4. 14 E.g. 1 Sam 1:24 ..... p. 236; 4:21-22 ..... p. 226; 2 Sam 23:31 ..... p. 233; 2 Kgs 11:13 ..... p. 226; Jer 23:33 ..... p. 276; 29:26 ..... p. 238; 41:9-> p. 276. 15 NJPS and McCarter, I Samuel, 222 use the same technique.

Chapter 1: In traduction

11

instance respectively: "The number is lacking in the Heb text"; "Two is not the entire number; something has dropped out." As a result of these mistakes, a literal understanding or translation of m yields a very difficult meaning. We are thus left with the assumption that the received text contains a textual error and that the earlier (correct?) text probably mentioned realistic numbers for Saul's age at the beginning of his reign, such as 30 years in d3Luc (b-mg oe2) (accepted by the REB), 21 years in 5, or 50 years suggested by the NEB.

The following are two additional examples of such mistakes. Judg 16:2

m (\j

;,m ]1tDr:liD M:::l ,r:lM';i C'ml'';i To the Gazites as follows: "Samson has come here" KOL GVTJYYEAT] (MS B: GTTTJYYEAT]) nile; ra(ao'ic; AEYOVTEC: The Gazites were told as follows (= ([; .. D 5)

The only reading that is understandable is that of l\3 and the other versions, which was followed by the NRSV ("The Gazites were told") and NJPS ("The Gazites [learned] that. .. "). Both modern translations reflect the verb in l\3. Jer 27:1

m NJPS

;"l'r:l,' ';ill(

m;,

,:::11;, ;"1';"1 ;"111;"1' 1';ir:l 1;"1'iD1M' ]:::l t:lp,1i1, n:;)';ir:lr:l n'iDM,:::l ,r:lM';i ';, nMr:l (= D ([)

At the beginning of the reign of king fehoiakim son of Josiah of Judah, this word came to Jeremiah from the LORD.

This verse serves as the heading of ch. 27, which speaks of actions taking place in the time of Zedekiah. -+ vv 3, 12; 28:1. Therefore, the mentioning of Jehoiakim in the heading does not suit the contents of the chapter and it probably erroneously repeats the first verse of the previous chapter, 26. The heading of ch. 27 was probably added in the forerunner of most textual witnesses at a later stage in the development of the book, while the earlier stage, in which it was lacking, is represented by l\3 ...... p. 289. (Zedekiah) of 5 (= NRSV) should probably be understood as a contextual correction. 16

3.

mDoes Not Reflect the "Original Text" of the Biblical Books

One of the postulates of biblical research is that many details in the text preserved in the various representatives (manuscripts, editions) of what is commonly called the Masoretic Text, do not reflect the "original text" of the biblical books. 17 Even though the concept of an "original text" necessarily remains vague -+ ch. 3B, differences between m and the other textual witnesses will continue to be recognized. Scholars will constantly hesitate regarding the originality of the readings of either m or one of the other sources. However, one thing is clear, it should not be postulated that m better or more frequently reflects the original text of the biblical 16 As a conclusion to this paragraph it is appropriate to quote the words of Kennicott, Dissertation (1753), Part the First, 269: "And now, if there certainly are Errors in the printed Text of the Old Testament, may we not be permitted to discover them?" 17 This perception goes back to Cappellus* (1650) 384-5. Also Eichhorn, Einleitung (4' 11 ed., 1823) 1.278-83 described in detail why the "oldest manuscripts were not without mistakes" (title of the section).

12

Chapter 1: Introduction

books than any other text. Furthermore, even were we to surmise that m reflects the "original" form of Scripture, we would still have to decide which form of m reflects this "original text," since m itself is represented by many witnesses that differ in small details.- ch. 21A 4. Differences between Inner-Biblical Parallel Texts in

m

The textual witnesses of the biblical books often contain parallel versions of the same unit. Some of these reflect different formulations in m itself of the same psalm (Psalm 18 II 2 Samuel 22; Psalm 14 II Psalm 53), a genealogical list (Ezra 2 II Neh 7:6-72), segments of books (Jeremiah 52 !I 2 Kgs 24:18-25:30; Isa 36:1-38:8 II 2 Kgs· 18:13-20:11), and even large segments of a complete book, viz., Chronicles, large sections of which run parallel to the books of Samuel and Kings. Some of these parallel sources are based on ancient texts that already differed from one another before they were incorporated into the biblical books, and which additionally underwent changes after they were transmitted separately from one generation to the next. In some instances, textual differences between parallel sources in m can easily be located, as in the texts presented in Table 1. Table 1 Gen 10:1-29 m Compared with the Parallel Text 1 Chr 1:4-23 m m:l"

inl(

iD"iDim ;-ttD•'-,l( J1"

c;,'-, 1i'-,1•1 nD"1 en t:JiD m

nD"i1

Oi"n1

iD"l( l(:ln:::lo1

""l(

n:::lni nl(1

nl( ]:::l"1 i1iDl(

nl(1 "01:::l";"T 111(1 16 nn 111(1 1i:l:::l

C:iD

p

ni" 111(1

m;,• i:ll(1

]Oi nl(1 12 n'-,:;, nl(1 i"ll nl(1 l:l"i1.,

c•inD:l nl(1 c•ntv'-,D

nl(1 "Pill;"! nl(1 •m;, nl(1 17

nl(1 inl(1 •non;, nl(1

;,'-,1( 20 lltv'-, ill

;-t:ll(:::l ;-t!ll

im1 '-,m1 r1ll l:lil( inl(;, CiD

iD1l( nl(129l(:::liD nl(1

fil(:::l

nl(1 c•::J;,'-, nl(1

nl( i'-,•

nD• •nl( i:::lll •j::J '-,:;, ":::ll( 1(1;-t

i"" iiD:lDil(1 24

11

•;,•1 19

;-t:ll(:::l cn

p '-, ll

i

1(1;-t;-t fil(;-t

13 ;,'-,i;;, i"ll;"T 1(1;-t n'-,:;, p1

nl(1 "i1il(;-t nl(1 18

nl( i'-,• iD1:l1 8 jii1 !(:liD

m•;,'-, '-,n;, 1(1;"1

iiDl( c•n'-,o:;, nl(1 c•oinD nl(1 14

ill

c:•ni1 iD1:l en

n•tDl(i •;,m 10 ;,1;-t"

lil(1 '-,:::l:::l

cmtD'-,'-,

'-,::Jm )1"1

;,n::Jo1 ;,'-,•m1 l(::JO iD1:l

;,•;, l(1;-t 9 ril(:::l

nl( i'-,•

m'-,1n ;,'-,1(1

i'-,• ctv'-,1 21 l:;-t"1;:::l

Clil(1 i1.,1 iiD:lDil(1 i1iDl(1 c'-,•l)

i'-,• i:::lll'-,1 25 i:::lll 111( i"" n'-,iD1 n'-,tv 111(

iil(1 "]',tv nl(1 ii1o'-,l( nl( i'-,•

1·nl( c:tv1 ril(;"l

111(1 '-,:::l1!J nl(1 28 ;,',pi 111(1 '-,nl( i1l(1 l:i1i;"T i1l(1 27 ;,'-,!( '-,:; :::l:::l1" 111(1_;,'-,•m iil(1

Chapter 1: In traduction Differences between Gen 10:1-29 sections: 3

m and

13

1 Chr 1:4-23

m in

the overlapping

1 Chr 1:6 1 Chr 1:6 tv'iDim ] ;"TiD'iDim 1 Chr 1:7 C':lii1 ] c•m11 1 Chr 1:7 n5:l'i1 ] 1'15:l'i1

;,o,,m ] ;,oi,1m

4

7 8

;"TO.Di . . .

;,o.Di 1 ;,n::101 ]

...

1

1 Chr 1 :9

1 Chr 1:10 1 Chr 1:10 l:l'i1'? ] l:l"i1'? ChrKetib 1 Chr 1:11 (Qere: l:l'i1'?) ] 1 Chr 1:13 ':1::11] i1'?1 1 Chr 1:17 r1.D 1r1.D1 1 Chr 1:17 iD01 ] liD01 1 Chr 1:17 '?::11.1' ] '?::!'.!' 1 Chr 1 :22 ] 1 Chr 1:23 iiO:I] i1i0:1 i::l' ] i1::l'

13 15 23

28 29

This table presents the text of Gen 10:1-29 m in which the overlaps with the parallel text in 1 Chr 1:4-23 in m (including some differences in details) are underlined, while the non-overlapping text is not. The high level of agreement between the two texts clearly shows that the lists in Genesis and Chronicles are closely related. 18 In this case, it is not difficult to distinguish between textual and other differences.19 The list in 1 Chr 1:4-23 reproduces the genealogical data of Genesis as well as a few narrative segments. The agreement between the two lists in m involves the smallest details, implying that little textual corruption occurred in these units, neither when the Chronicler copied the list from his source, which must have been very close to the present text of m in Genesis, nor during the course of the textual transmission in the ensuing centuries. The only differences between the two texts are in spelling (11 x),20 a different linguistic form (v 4), an added waw (v 23), and a few scribal interchanges, undoubtedly representing errors in one of the two texts (vv 3, 4, 23, 28). Vv 4 and 28 are discussed on p. 16. The nature of one case remains unresolved, 21 but all other differences fall under the heading of scribal transmission such as described in ch. 4.

18 1 Chronicles 1 contains summaries of genealogies in Genesis 5, 10-11, 25, 35-36. Also within the sections from Genesis 10, the Chronicler presented a summary of his source, since he left out some segments (Gen 10:5,9-12, 18 end, 19-21). Furthermore, the reworking of the Chronicler included the removal of headings and conclusions in Genesis. 19 When the texts differ, we may correct details in one text to those in the parallel text (see the discussion of vv 4 and 28 on p. 16), but the basic dichotomy between the texts should be maintained since they were composed by different authors. 2 Chronicles always presents a fuller orthography+.-+ p. 213 21 Gen 10:22 mentions the five sons of Shem, among them Aram, while the next verse lists Aram's four sons. However, in 1 Chr 1:17, the words "the sons of Aram" are missing in m+, a waw is added before Uz, and as a result Shem is presented as having nine

°

14

Chapter 1: In traduction

The relation between the parallel Psalms 14 and 53 (Table 2) is more complicated than between the texts presented in Table 1, but they represent the same Psalm transmitted in two different versions. Table 2 Psalm 14m Compared with the Parallel Text of Psalm 53

m

i1iDll )'!'\ i1'?•'?ll 1:::l'llni1 1n'niDi1 c•i1'?1'\ )'!'\ 1:::l'?:::l '?:::lJ iOI'\ 111'? iO '?::li1 3 C'i1'?1'\ nl'\ iDi1 '?•:liDO iD'i1 n11'\i'? C11'\ 'J:::l '?ll en'? 1'?::ll'\ 'Oll •'?;:,!'\ )11'\ •'?ll5:1 '?;:, 1ll1' l'\'?i1 4 1nl'\ CJ )'!'\ i11i1' •;:, 1iD':::ln 'Jll

C'OiDO i11i1'2 i1iDll )'!'\ m'?l'\: 11n•

i11:::l c•i1'?1'\ •;:, 1n5:1 11n5:1 ctvs 11'\ip !'\'? i11i1'

'?1'\iiD' noiD' :::lPll' '?J• 10ll m:::liD i11i1' :l1iD:::l '?1'\iiD' nll1iD'

1n· •o 7 mono

Differences between Psalm 14m and Psalm 53 m: 1 111'? ] 111'? '?•:liDO n'?no '?ll Ps 53:1 1:::l'llni1 ] 1:::l'llni11 Ps 53:2 i1'?•'?ll ] '?1ll Ps 53:2 i11i1' ] C'i1'?1'\ Ps 53:3 2 '?::li1 ] 1'?;:, Ps 53:4 3 iO ] JO Ps 53:4 '?;:, ] > Ps 53:4 4 i11i1' ] c•i1'?1'\ Ps 53:5 1n5:1 ] + 1n5:1 i1'i1 !'\'? Ps 53:6 5 i11:::l ] il5:1 Ps 53:6 ] > Ps 53:6 6 'Jll ] 1m Ps 53:6 1iD':::ln ] i1niD':::li1 Ps 53:6 •;:,] > Ps 53:6 i11i1' ] C'i1'?1'\ Ps 53:6 1i10no ] COI'\O Ps 53:6 7 nll1iD' ] n1ll1iD' Ps 53:7 i11i1' ] C'i1'?1'\ Ps 53:7 This table presents the text of Psalm 14m in which the overlaps between the two texts are underlined, while the differences are not. While there are several substantial differences between the two psalms in m, they share the same ideas and are therefore analyzed in the commentaries as a single psalm. Some scholars attempt to reconstruct the text that was at the base of the two transmitted psalms.22 children. The difference between the two texts was probably caused by a textual error. Alternatively, the difference may have been intentional, involving a different view of the source of the Arameans. Japhet, Chronicles, 59 is undecided. 22 C.C. Torrey, "The Archetype of Psalms 14 and 53," /BL 46 (1927) 186-92; K. Budde, "Psalm 14 und 53," JBL 47 (1928) 160-83. In some details, these scholars consider the

Chapter 1: In traduction

15

In their common text, the two psalms speak out against the nations who deny God's existence and devour His people, and end with a prayer for salvation emanating from Zion. The differences between the two versions were created during the course of their scribal transmission, with additional differences created by separate exegesis. It is sometimes difficult to distinguish between these two levels since several of these differences can be explained in different ways. Most differences were created during the scribal transmission: - pp. 219-62 1 il'?''?l) ] '?1.!) Ps 53:2. After the word pair 1n'niVil, the general il'?•'?l) (deed) is appropriate and not ',ul, which creates a tautology. The two words are tantalizingly close, and therefore a graphical interchange is likely. 3 10 ] JO Ps 53:4. It is difficult to decide which reading is preferable, but the words clearly developed from one another. 5 p•1:;: ] li5l Ps 53:6. An interchange of and li5l is likely, resulting in words carrying opposite meanings. According to Psalm 14, "God is present in the circle of the righteous" (NJPS) including a word p•1:;: not found in Psalm 53, while according to Psalm 53 "God has scattered the bones of your besiegers." ] ilniD'::lil 1:ln Ps 53:6 and ] Ps 53:6. The two 6 1iD'::ln ':ll) psalms display diametrically opposing pictures. The most clearly recognizable difference is that God "rejected" Israel's enemies in Psalm 53, while in Psalm 14 God serves as the "refuge" of his righteous followers (second group of variants). However, graphically the Hebrew words are very similar. Likewise, the first group of variants (Psalm 14: "You would confound the plans of the poor" [NRSV] II Psalm 53: "[God has scattered] the bones of your besiegers. You have put to shame.") offers two different pictures, while the Hebrew letters are similar as indicated in the printing (in addition, the aleph and the het are phonetically close). Both formulations have their own internal logic, and because they are graphically similar, one version developed from the other one. -+ ch. 38 Other differences between the two versions are exegetical. 3 '?:liT ] 1'?;:, Ps 53:4. 4 '?;:, ] > Ps 53:4. Ps 53:6. This additional hemistich+ runs parallel to the 5 1n5l ] + 1n5l il'il previous one, and may reflect a double, alternative, reading. Note further an expanded superscription in Ps 53:1 and the replacement of (God) in vv 2, 4, 7 in Psalm 53, the Tetragrammata+ of Psalm 14 (il1il') with as usual in the Elohistic+ Psalter (Psalms 42-89 [or: 83]) ...... Tov, "Coincidental Textual Nature," 164-6

The analysis in Tables 1 and 2 showed that it is often difficult to decide whether a certain variation reflects a scribe's content exegesis or a textual development. Assuming that such a variation often reflects scribal activity, the parallel texts in m provide a major source of information about ancient scribal activity, 23 similar to the differences between ancient scrolls of the same text. same reading original, usually that of Psalm 14, while in v 5 they go their separate ways. 23 See the data on pp. 221-39 and Sperber, Grammar.

Chapter 1: In traduction

16

Even though there is no direct, archeological, evidence (that is, ancient scrolls) for the earliest stages of the transmission of most biblical books, indirect evidence does exist for this stage in these parallel texts within m itself. Such differences attest to readings developed in one of the first stages of the textual transmission, as, for example, between the two parallel versions of the "Table of the nations" in m (Genesis 10 II 1 Chronicles 1;- Table 1 above): Gen 10:4 m C'J111 C'n::l iD'iD,m i1iD'"l!t ]1' 'J:l1 ( = ([ 0 Ps-J No and o in

1 Chr 1:7

m

Gen 10:28

m

1 Chr 1:7) The descendants of Javan: Elishah and Tarshish, the Kittim and Dodanim. 24 C'J11i1 c'n::l i1iD'iD,m 11' 'J:l1 (= w. C'J,,,, and II) 'P68Lm in Gen 10:4) The descendants of Javan: Elishah and Tarshishah, the Kittim and Rodanim. nl!t1 ":ll!J nl!t1 ( = S, also in Gen 36:23)

Obal and Abimael 1 Chr 1:22

nl!t1 ":l"ll nl!t1 (= Gen 10:28

m

w. ([ o; Gen 36:23 m

II)

([ l))

Ebal and Abimael

Similar internal differences are found in the two versions of the list of David's mighty men in m: 2 Sam 23:28-29 m

i1Jll:l

p :::l?r 29

(= ([ s)

Maharai the Netophathite, 29 Heleb son of Ba'anah the N etopha thi te 1 Chr 11:30

m

in Samuel) Maharai the Netophathite, Heled son of Ba'anah the Netophathite i1Jll:lJ:l 17D

(=II) ([ D; = o

The scribal background of differences of this type was already recognized by R. David Kimhi (RaDaK) in his commentary on "and Rodanim" in 1 Chr 1:7:25 This word is written with a resh at the beginning. And in the book of Genesis it is written with two daleths: "and Dodanim." Since the daleth and resh are similar in appearance, and among the readers of the genealogies which were written in ancient times, some read a daleth and some read a resh, some names were preserved for posterity in two 24 For a similar interchange, see Ezek 27:15 m [ 0 Jii - I I ) Po&[wv. In ancient manuscripts, several modem translations, and some editions (Letteris, Sinai), the differences between these two names have been removed by changing the Chronicles text in accord with that of Genesis. Seep. 374 for similar examples. 25 RaDaK's explanation continues the view expressed in Gen. Rab. 37.1 (ed. Theodor [Berlin: Ittskovsky, 1903-1929] 344), according to which these forms are two different exegetical variations of the same entity.

Chapter 1: Introduction

17

forms with either a daleth or a resh. Thus it is written in the book of Genesis with one of the readings and in this book with the other one. This goes to show that both forms represent one name whether read with a daleth or with a resh. This applies also to "Riblatah" (2 Kgs 25:6,20; Jer 39:5; 52:9,10,26) written with a resh and "Diblatah" (Ezek 6:14) with a daleth ... Likewise, words with waw and yod are interchanged as they are similar in appearance.

In ch. 4c, many similar differences between parallel texts are presented.26 The differences between Psalm 18 and 2 Samuel 22 and Isa 36:1-38:8 II 2 Kgs 18:13-20:11 are listed in Sof 8.1-2. B. A Modern Approach to Textual Criticism

Since the discovery in 1947 of Hebrew and Aramaic texts in the Judean Desert dating from approximately 250 BCE until 135 CE, our knowledge about the Scripture text has increased greatly. - ch. 21A-c. It should be remembered that until the time of those discoveries no early Hebrew and Aramaic Scripture texts were known, except for the Nash papyrus of the Decalogue - p. 111,27 and as a result the manuscripts of m from the Middle Ages served as the earliest Scripture sources in the original languages. Therefore, the research before 1947 was based on HebrewAramaic texts that had been copied 1200 years or more after the composition of the biblical books. At the same time, scholars also relied on manuscripts and early papyrus fragments of the ancient translations - ch. 211, especially of the Septuagint (11)) and the Vulgate (D), which brought them much closer to the time of the composition of the biblical books. All these, however, are translations, and the reconstruction of their Hebrew-Aramaic sources will always remain uncertain. - pp. 1227. Therefore, the discovery in .the Judean Desert of many HebrewAramaic texts dating from two millennia ago has considerably advanced our knowledge of the early witnesses and the procedure of the copying and transmitting of texts. This new knowledge has necessarily changed our understanding of the Scripture text and, accordingly, our approach to writing an introduction to textual criticism. Such a new approach is not reflected in 26

27

It is exactly these parallel biblical passages that prompted the development of textual

criticism of Hebrew-Aramaic Scripture, because they necessitated the comparison of texts. See especially Kennicott, Dissertation, Part the First compares I CHRON. XI with 2 SAM. V and XXIII (1753); Owen, Critica Sacra (1774). Further studies on this topic are listed by I. Kalimi, Chronicles, The Books of Chronicles: A Classified Bibliography (Simor Bible Bibliographies; Jerusalem: Simor, 1990) 52-66. However, this liturgical papyrus does not represent a witness for the biblical text in the generally accepted sense of the word.

18

Chapter 1: Introduction

most previously written introductions. The influential introduction of Roberts, OTTV, was written in 1951, after the discovery of the first texts in the Judean Desert, but its author was not able to incorporate the new discoveries in his description. In our view, the introductions of Klein, Textual Criticism 1974, Deist, Text 1978, id., Witnesses 1988, McCarter, Textual Criticism 1986, and Wiirthwein, Text 1988 (5th edition) although written at a time when several Judean Desert texts were known, in many respects still reflect the approach of the period before the discovery of the new data. As a matter of fact, the main Judean Desert texts were published after the publication of these introductions. On the other hand, the more recent introductions by Wegner* 2006 and Fischer* 2009 pay much attention to the new discoveries. In our opinion, the new discoveries have not only added new data that are of major importance, but have also necessitated a new approach to the texts that were known before 1947.- pp. 158-60 The new insights are visible in the different amounts of attention given to the textual sources. Ever since the 17th century, introductory analyses have given equal attention to all Scriptural texts. Scholars regarded the ancient translations, especially the Greek and Latin versions, with esteem, because their manuscripts preceded those of mby many centuries, and also because Greek and Latin sources were highly valued in the Church and in the centers of learning in Europe. Therefore, much attention has been given not only to mbut also to the Greek, Latin, and Aramaic versions, including the Peshitta (5), and even to the "daughter" (or secondary) versions made from ®, such as the Latin, Armenian, Coptic, and Ethiopic translations. - p. 127. After some time, scholars realized that most of these translations were only of limited value for the textual criticism of the Hebrew-Aramaic Bible, and that their importance was confined mainly to biblical exegesis. Nevertheless, these primary and secondary translations were still given extensive treatment in textual descriptions, commentaries, and even in the BH series. 28 In our view, this approach is no longer relevant for modern textual criticism. Therefore, the different coverage of the various sources in this introduction reflects our textual outlook. This monograph devotes less attention than in the past to the description of the medieval manuscripts of m and to most of the ancient versions, whose importance for the textual criticism of Hebrew-Aramaic Scripture-as opposed to 28 Nevertheless, the BH series contains almost no notes referring solely to the Aramaic or Latin translations, or one of the "daughter" translations+ of lB. Such evidence is mainly mentioned in conjunction with that of additional sources.-+ Tov, "Aramaic"

Chapter 1: In traduction

19

exegesis-is limited. On the other hand, much attention is devoted to texts whose relevance has been proven, that is, m, the Judean Desert texts, the Samaritan Pentateuch (w.), and 113. The study of the biblical text was initiated as an auxiliary science to biblical exegesis. Therefore, the results of textual investigation have been taken into consideration in exegesis, and that practice continues to be followed today. Textual criticism has a distinctly practical aspect for biblical exegesis, but as a rule this feature has not been reflected sufficiently in the extant handbooks on textual criticism. Unlike those handbooks, chapters 6-8 of this book deal extensively with exegetical aspects. Within this framework, the relevance of textual criticism to literary analysis, a topic that is usually not treated in handbooks such as this, is covered extensively inch. 7. C. Beginnings of Critical Inquiry into the Scripture Text Barthelemy, Critique textue/le 1992, *1-*63; B. Chiesa, "Appunti di storia della critica del testo dell' Antico Testamento ebraico," Henoch 12 (1990) 3-14; L. Diestel, Geschichte des A/ten Testaments in der christlichen Kirche Qena: Mauke, 1869); Eichhorn, Einleitung; CashenGottstein, "Biblical Manuscripts"; K.F. Keil, Manual of Historico-Critical Introduction to the Canonical Scriptures of the Old Testament, vol. II (Edinburgh: T & T Clark, 1892); Konig, Einleitung; H.J. Kraus, Geschichte der historisch-kritischen Erforschung des A/ten Testaments (3'd ed.; Neukirchen: Neukirchener Verlag, 1982); F. Laplanche, L'Ecriture, le sacre et /'histoire: Erudits et politiques protestants devant Ia Bible en France au xvne siec/e (Amsterdam/ Maarssen: APA-Holland University Press, 1986); Rosenmiiller, Handbuc/z; Steuernagei, Einleitung, § 22.

Interest in the text of Scripture began in the first centuries CE when learned church fathers compared the text of Jewish Scripture with its Greek versions. In the 3rct century, Origen prepared a six-column edition (hence its name: Hexapla edition]) of Hebrew-Aramaic Scripture, which contained the Hebrew-Aramaic text, its transliteration into Greek characters, and four different Greek versions. -+ pp. 145-6. Likewise, Jerome+ included various notes comparing words in the Hebrew text and their renderings in Greek and Latin translations in his commentaries. -+ ch. 2, Table 8 (p. 46) The critical investigation of the relation between the various textual witnesses began in the 17th century with the appearance of the Polyglot text editions+, which invited their comparison because they presented the various witnesses in parallel columns. The first extensive textual treatises are those by Cappellus, Morinus, and Richard Simon, displaying an interest in either m or one or more of the other versions, usually colored

20

Chapter 1: Introduction

by Catholic-Protestant polemics. 29 The reader of Cappellus, Critica Sacra (1650) is amazed at the level of knowledge, acumen, and critical insight displayed by this scholar, foreshadowing all modern analyses. After the middle of the 17th century, a great many critical treatises appeared on the Scriptural text, though it should be recognized that philological analysis and theological discussion often went hand in hand in this and the following century. The three aforementioned studies, as well as many by Buxtorf, Glassius, Hottinger, Houbigant, Kennicott, Rosenmi.iller, and de Rossi, contributed much to the development of the critical approach to the biblical text. 30 The Einleitung of Eichhorn* (17801823) stands out as a work of immense learning and sophistication in the 18th century, which also had a major influence on subsequent generations. 31 Among the many names that may be mentioned from the 19th century, see especially de Lagarde, Perles, Cornill, and Wellhausen, who displayed a remarkable insight into textual criticism (-+the general bibliography on pp. xxix-lv). In many areas of textual criticism, it is often best to start with these older works, since an intuitive grasp of the issues underlying divergent texts is just as important in this art as information deriving from recently discovered data (e.g. the Qumran texts). Wellhausen, BUcher Samuelis and the introductions of Konig* and Steuernagel* exhibited this type of intuition in very particular ways. However, the modern description of textual criticism differs significantly from earlier discussions as a result of the contribution of the newly discovered Qumran texts to almost every area of textual criticism. For many other respects of the history of the investigation of the biblical text, see ch. 3A. D. Text, Canon, and Sacred Status

The books of Hebrew Scripture were gradually accepted as binding (authoritative) and at some point were integrated into different collections of sacred writings, among them those of m, \13, and ill. However, scribal processes started a long time before the books obtained authoritative status, also named canonization+ when referring to Scripture as a whole. For example, Jeremiah 36 describes how the 29 Morinus, Exerc.; Cappellus, Critica Sacra; Richard Simon, Histoire critique du Vieux Testament (Paris: Editions de Paris, 1680/Rotterdam: Leers, 1685; repr. Frankfurt: Minerva, 1973) =A Critical History of the Old Testament (London: Davis, 1682). 3 For bibliographical references, see the general bibliography and ch. 3Al. 31 The works of these scholars have been described in detail by Rosenmtiller*, Keil*, Barthelemy* 1992, 1*--63*, and Childs, Introduction, 89-92.

°

Chapter 1: In traduction

21

prophet dictated the contents of a second scroll to Baruch following the burning of the first one by the king. That scroll thus constituted a second stage of the scribal development and, in a way, of the growing process of the book. --+ p. 181, n. 66. The contents of the second scroll cannot be reconstructed, let alone the first one, but in other cases we know more about the development stages of the books. Thus, in ch. 7s1 we analyze the relation between them and \1J versions of Jeremiah, suggesting that \1J (together with 4QJerb,d) represents an early stage in the literary (editorial) development of that book, which preceded the edition of m that became canonical. Elsewhere in ch. 7B, we describe many similar examples of literary variants+ preserved in non-Masoretic sources. On the basis of this understanding, we therefore submit that writing processes and textual transmission can and should be discussed beyond those seen in the final literary shape of the biblical books, often included in m. This approach involves the opening up of new horizons beyond m. When opening up new sources that are relevant for textual criticism, we thus should keep an open mind with regard to different Scripture collections. --+ p. 284. When discussing the textual criticism of Hebrew Scripture, in the view of many scholars we should not consider the canonical status of m binding for the analysis 32 since the concept of sacred Scripture was more encompassing than solely that included in m. Since textual criticism deals with all forms of Hebrew Scripture, it also covers the content of other sacred collections, namely those included in \1J, ill, and some Qumran scrolls. Some books included in these collections preceded the literary crystallization of m, while others were composed afterwards. In ch. 7s21-23, the reader will therefore find an analysis of such non-Masoretic works as Psalm 151 (\1J and 11QPsa), the editorial changes inserted in ill, and the exegetical edition of the Torah in 4QRP.33 Finally, there is not necessarily a connection status of the Scripture books and the nature of the Even the most sacred Scripture book, the Torah, more carefully than the other books, neither in scrolls nor in other text traditions.--+ p. 188 32 33

between the sacred scribal transmission. was not transmitted the proto-Masoretic

On the other hand, some scholars claim that before the canonical process was completed, the biblical books as known to us did not yet exist, and therefore the scribal transmission of the biblical books had not yet taken place. At the same time, "rewritten Scripture" compositions and abbreviated Qumran scrolls are excluded from textual analysis with regard to their major deviations from m, even though some of these compositions were considered authoritative in antiquity. As remarked on p. 189, deviations from m in these compositions in small details are included in the analysis.

22

Chapter 1: In traduction

E. Subjectivity of This Book Subjectivity pervades all chapters of this book, although we try to be as objective as possible. The objective elements pertain to the description of facts and textual sources, but these descriptions also include many subjective components. For example, the recognition of the Qumran Scribal Practice (-+ pp. 100-105), the characterization of the textual character of the Qumran scrolls(-+ pp. 105-7), and the description of the text-critical value of the various translations (-+ pp. 135-40, 146-53) are all subjective. The recognition of variants+ in the ancient translations i'S based on a combination of objective criteria and intuition. In the case of the evaluation of variants+ (ch. 6), we stress that objectivity is impossible, and that scholars must develop their own subjective reasoning (-+ pp. 279-81). On the whole, something considered a solid fact by one scholar is contested by another. Thus, almost every paragraph in this book attests to subjectivity, which needs to be taken into consideration.

2 TEXTUAL WITNESSES I. HEBREW WITNESSES The Scripture text is known from many textual witnesses (sources, texts), in Hebrew and in translation. The discussion of the Hebrew sources in this chapter is central to this book as a whole since all other chapters are based upon this description and constantly refer to it. In the description that follows, the analysis of the Hebrew witnesses (part I of this chapter) is separated from that of the ancient translations (part II). Part I describes the direct evidence, while the data covered by part II are indirect because they are based on the reconstructed Hebrew source of translations. Within each group, the sources are described in chronological order, although precision is impossible. The first two sections in part I (A, B) discuss texts that are well known from medieval sources (m and ru), and they are both discussed here together with the less known early texts from which they developed (proto-Masoretic and pre-Samaritan texts from the Judean Desert). In our terminology, a distinction is made between the proto-Masoretic+ (proto-rabbinic) texts, which are the actual forerunners of the medieval m and belong to the same group, and the pre-Samaritan+ texts from Qumran on one of which ru was based. In other words, the proto-Masoretic texts were Masoretic, so to speak, while the pre-Samaritan texts were not Samaritan. One of our goals is to advance the understanding of the biblical text as having developed diachronically in different directions from the earliest periods to the Middle Ages. At the same time, we would like to advance the synchronic understanding of the biblical text during certain periods, especially that of the Second Temple, for which relatively much manuscript evidence is available. For this purpose, it would be advantageous to depict the oldest layer of m (the protoMasoretic texts) together with other early witnesses (mainly ® and the m-like Qumran texts) without involving the medieval elements of m, which are visible, for example, in vocalization and accentuation. However, for practical reasons, this separation of the layers of m is not possible because the description of the late elements of m is intertwined with that of its early layer.

24

Chapter 2: Textual Witnesses

A. Proto-Masoretic Texts and the Masoretic Text (m-Group) Barthelemy, Critique textue/le 1992, vii-cxvi; Beit-Arie, "Practices"; Breuer, Aleppo Codex (1976); Cohen, "Masoretic Text" (1986); A. Diez Macho, Manuscritos hebreos y arameos de Ia Biblia (SEAug 5; Rome: Institutum patristicum Augustinianum, 1971); A. Dotan, "Masorah," EncJud (1971) 16.1401-82; Ginsburg, Introduction; Glatzer, "Aleppo Codex"; CashenGottstein, Biblia Rabbinica, 5-16; A.M. Habermann, "Bible and Concordance," in Thesaurus of the Language of the Bible (ed. S.E. Loewenstamm & J. Blau; Jerusalem: Bible Concordance Press, 1957) l.xix-xxxviii; M.J. Mulder, "The Transmission of the Biblical Text," in id., Mikra, 87-135; Rabin, mhqrym; A. Schenker, "Der Ursprung des massoretischen Textes im Licht der literarischen Varianten im Bibeltext," Textus 23 (2007) 51-67; A. Sperber, Grammar; I. Yeivin, "mqr ·, ktby yd sl hmqr'," EncBib Oerusalem: Bialik Institute, 1968) 5.418-38; id., "mswrh," ibid., 130-59; id., mbhr ktby-yd bsytwt nyqwd .tbrny w 'r;;-ysr 'ly Oerusalem: Akademon, 1973); id., Introduction (1980); id., Masorah (2003).

toors: Accordance, BibleWorks, The Dead Sea Scrolls Electronic Library, Global Jewish Database (Responsa Project), Gramcord for Windows, Jewish Classics Library, Libroi, Logos, Mechon Mamre, SESB, Snwzit, and WordSearch include biblical modules based on either codex L+ or BHS 1 with certain editorial liberties (e.g. the determining of the vocalization of the Ketib+ words in order to enable word searches). For details, see Tov, "Electronic Resources" 2003, 2008. The Bar-Ilan database Tokhnit "HaKeter"-Ma 'agar HaTanakh, part of the Miqra 'ot Gedolot "HaKeter" project, is based on codex A, which differs only slightly from L, but for students of grammar these differences are important.

The name Masoretic Text refers to a group of manuscripts (the m-group) that are closely related to one another. Many of the elements of these manuscripts, including their final form, were determined in the early Middle Ages, but they continue a much earlier tradition. The name Masoretic Text was given to this group because of the apparatus of the Masorah attached to it. -+ § 5. This apparatus, which was added to the consonantal base, developed from earlier traditions in the 7th to the 11th centuries-the main developments occurring in the beginning of the lOth century with the activity of the Ben Asher+ family in Tiberias. As a rule, the term Masoretic Text is limited merely to a segment of the representatives of the textual tradition of m, namely, that textual tradition that was given its final form by Aaron Ben Asher of the Tiberian branch of the Masoretes+. Since all the printed editions and most manuscripts reflect this Ben Asher+ tradition, the term Masoretic Text is imprecise, for it is actually used only for part of the Masoretic tradition, viz., that of Ben Asher. In order to avoid this imprecision, Goshen-Gottstein* distinguished between MT in general and the Tiberian MT (TMT). When using the term MT, one usually refers to the Tiberian MT. The term Masoretic Text is imprecise for another reason, too, form is not attested in any one single source. Rather, m is an abstract unit 1

In principle, these two sources should be identical, since BHS is based on codex L, but in practice they are not.

IA: Prato-Masoretic Texts and the Masoretic Text (!TI-Group)

25

reflected in various sources that differ from one another in many details. 2 Another aspect pointing to the inadequacy of the term Masoretic Text is the fact that the notes of the Masorah+ do not pertain to all the Masoretic manuscripts, as Cohen* 1986 has demonstrated. Therefore, a term like mgroup or Masoretic Texts would reflect the evidence more precisely. In this book, we continue to use the conventional term Masoretic Text or m alongside that of them-group. The principal component of m is that of the letters, evidenced fragmentarily in antiquity in the Judean Desert texts, and to this text all other elements were added during the early Middle Ages. Therefore, although the medieval form of m is relatively late, its consonantal framework reflects an ancient tradition that was in existence more than a thousand years earlier in many sources, among them many Judean Desert texts from places other than Qumran, copied in the period between 50 BCE and 115 CE. --+ p. 29. Accordingly, scholars usually designate the consonantal base of m (evidenced in the Second Temple period) as proto-Masoretic although sometimes also, anachronistically, as the Masoretic Text. The terms proto-rabbinic and rabbinic texts are used less frequently, although they actually describe the nature of m and its forerunners more precisely.--+ p. 27 m contains: 1. The consonantal framework 3 attested m proto-Masoretic (proto-

rabbinic) texts from the Judean Desert; 2. Vocalization(§ 2); 3. Para-textual elements (§ 3); 4. Accentuation (§ 4); 5. The apparatus of the Masorah (§ 5). m has served as the most commonly used form of the Hebrew-Aramaic Bible ever since its acceptance as authoritative by all Jewish communities in its consonantal form probably in the 1st century CE, and several centuries later in conjunction with its vocalization, accentuation, and the apparatus of Masoretic notes. Because of this acceptance, initially of the proto-Masoretic text by rabbinic Judaism and later of m by all sections of the Jewish people, m is attested in a very large number of sources. More than six thousand manuscripts belonging to the !11-group are known. In the words of Beit-Arie*, 72, " ... of some 2700 extant dated Hebrew manuscripts prior to 1540, six dated codices from the tenth century, eight 2

3

See below, n. 7, on the possibility of a single archetype at the base of m. This term is used here for all the letters, including the vowel letters '"lint

26

Chapter 2: Textual Witnesses

from the eleventh century, and 22 from the twelfth century are known to us, most of them Oriental. In addition, there are about sixty small fragments of Oriental codices dated before 1200 among the geniza fragments." In addition, all printed editions of the Hebrew Bible(--+ ch. 9) are based on m. The Masoretic codices, consisting of quires bound like books .... plates 11 *-15*, were written by scribes in accordance with the halakhot, "religious instructions," relating to the conventions for the preparation of the leather, ruling, writing, and correcting ..... ch. 4B1-2 The various components of the text were inserted by different people. Soferim, "Scribes" wrote down the consonantal text, naqdanim, "vocalization experts'' added the vowels and accents, and Masoretes (ba 'ale ha-Masorah, "masters of the Masorah") wrote the notes of the Masorah. However, the Masoretes were often also involved with the other components of m and, therefore, the term Masoretes often refers to those involved with all aspects of the copying of the medieval manuscripts.

1. The Consonantal Framework: Proto-Masoretic Texts and m F.l. Andersen & D.N. Freedman, "Another Look at 4QSamb," RevQ 14 (1989) 7-29; Cohen, "Some Basic Features of the Consonantal Text in Medieval Manuscripts of the Hebrew Bible," Studies in Bible and Exegesis, 1.123-82; id., "Masoretic Text" (1986); Goshen-Gottstein, "Biblical Manuscripts"; Lange, "Plurality" (2010); Y. Maori, "mwb'wt mqr'ywt bsprwt hz"l," Mahanayim 70 (1962) 90-99; Penkower, "Pentateuchal MS" (1988); id., "The Development of the Masoretic Bible," in The Jewish Study Bible (ed. A. Berlin & M.Z. Brettler; Oxford: Oxford University Press, 2004) 2077--84; Tov, "The Text of the Hebrew I Aramaic and Greek Bible Used in the Ancient Synagogues," HB, GB, and Qumran (2008) 171-88; id., "Aramaic"; E. Ulrich, "Two Perspectives on Two Pentateuchal Manuscripts from Masada," in Paul, Emanuel (2003) 453--M; Young, "Stabilization." See further the literature on p. 219.

a. Background The representatives of m (the m-group) form a tight group (family) of texts that differ from the other witnesses. The m-group shares a large number of readings and characteristic features, even though no defining textual characteristics can be identified in the content of m, except for the careful copying, fine quality, and antiquity of its text in most of the biblical books. On the other hand, on a socio-religious level this text is unique since, at a certain stage, it was preferred to the other texts by the group that was later named the rabbinic movement. As a result, the mgroup is the largest among the Scripture witnesses. From m's preservation, it is clear that the text, including the para-textual+ notations --+ § 3, has been more or less frozen since at least the 3rct century BCE. When evaluating the quality of m, one should realize that the preference form within Judaism does not necessarily imply that it contains the best (earliest) evidence of the Scripture text; both the Hebrew parent text of 11)

IA: Prato-Masoretic Texts and tile Masoretic Text (In-Group)

27

and several Qumran manuscripts reflect excellent texts, often better and/ or earlier than m...... pp. 110, 135-40 When m became the central text, at first within rabbinic Judaism and later for the Jewish people as a whole, no further changes, additions or omissions were allowed ..... § c, not even in small details such as the use of matres lectionis+. Therefore, throughout the centuries, m preserved the biblical text in the exact form in which it was current at a particular time within a particular circle; it even preserved such scribal minutiae as the extraordinary points+ and other paratextual+ elements (§ 3d) most of which were not meant to be preserved for the following generations. Precisely the same text(s) of m is (are) first attested to in the scrolls from Masada and other sites in the Judean Desert (copied between 50 BCE and 115 CE), but probably they go back to an earlier period. Very similar texts (!11-like texts) were found at Qumran from the mid-third century BCE onwards. For both groups of texts, see §b. After the proto-Masoretic text had become the accepted text, especially within rabbinic Judaism, it was copied many times and, as a result of its central status, most of the ancient translations were based upon one of its slightly different representatives: the Targumim, 5, the revisions (recensions) of (among them kaige-Th+, Aquila+, Symmachus+, and the fifth column of the Hexapla+), and 0. 4 ..... part II of this chapter. Likewise, m is quoted exclusively in both early and late rabbinic literature ...... pp. 33-4

b. Development of the Consonantal Text

The evidence from the Judean Desert allows us not only to comprehend the textual situation at the time of the copying of these scrolls, but also to improve our understanding of the development of m and to postulate a stage before the earliest available manuscript evidence. The different attestations of the consonantal text of m point to three main stages that reflect a growing measure of agreement between the sources forming the m-group. A different type of manuscript evidence characterizes each stage. These three stages, recognized on the basis of evidence, were preceded by an additional, hypothetical stage.s The internal development of the m-group may be described as one of increased textual stability, not to be confused with the possibility of a standardization process for the biblical text as a whole, which is rejected in ch. 3c2. The description that follows refers only to the precursors of m (the proto-Masoretic texts) and m itself, and not to the development of the biblical text as a whole ...... ch. 3c i. The assumption of the existence of a stage of m that preceded that of the manuscript evidence from the Judean Desert is hypothetical. The origins of m are unclear, but similar texts must have existed before the stage for which we have manuscript evidence, that is, before them-like texts from Qumran (c. 250 BCE). 4 5

Accordingly, in this book, the siglum m+ refers tom q: 5o. The description of these periods differs from TCHB 2.

28

Chapter 2: Textual Witnesses

There probably are no sound criteria for establishing the original date of !TI based on its contents. In any event, such a date would have differed from book to book. 6 In the establishing of this date or these dates, we need not take into consideration the canonization+ process of Hebrew Scripture, since copies of m of some books could have existed before the establishment of m as authoritative Scripture. We do not know whether we should postulate one or more copies for the first stage of m, possibly in the 5 1h, 4 1h, or the beginning of the 3'd century BCE. If there was a single text, it may well have been the predecessor of the Scroll of the Temple Court+ mentioned in rabbinic literature. - pp. 30-31? In any event, if there ever was an early single archetype of m, it cannot be reconstructed. At this first hypothetical stage, m already must have been a good text, although not written with the same precision a.s that displayed in later periods. The orthography+ was inconsistent - ch. 4B4, and at least Samuel contained relatively many mistakes and was marred by theological tendencies. - pp. 189, 254-6. Although there is no manuscript evidence pertaining to the internal differences within the m-group in the first period, it would appear from a comparison of parallel texts within m itself that such differences already existed between the various textual witnesses at an early stage.- pp. 12-17 ii. The second stage of the textual transmission extended over a long time-span. Its beginning and end are determined by the presence of texts in the Judean Desert dating from c. 250 BCE (the earliest Qumran evidence) to the time of the Bar Kochba revolt (132-135 CE). The texts themselves (found at Wadi Murabba'at, Wadi Sdeir, Nahal 1-:lever, and Nahal Se'elim) are assigned to the period until c. 115 CE. The earliest textual evidence of m displays a rather uniform picture, including m-like texts from Qumran and proto-Masoretic texts, which are identical to the medieval m, from the other sites. The destruction of the Second Temple in 70 CE, occurring in the middle of this period, did not trigger a stabilization process as is often believed. Since rabbinic Judaism was the only group to survive, its texts likewise were the only survivors.- pp. 178-80 The witnesses for this stage may be divided into two groups, defined in accordance with their closeness to codex L (Leningrad codex B19A- plate 13*) of 6 Only unstable evidence is available: (l) On the basis of its orthography, m has been dated by different scholars to the periods between 550 and 350 BCE, 400 and 100 BCE, and to the end of the 3'd century or the beginning of the 2"d century BCE.- p. 212. Some of these dates are early, and precede the stage of the available evidence. (2) Since three paleo-Hebrew texts found at Qumran are close tom (4QpaleoGen-Exod 1, 1QpaleoLev, 4QpaleoDeutr), they probably provide information about m from a period preceding its attestation in Masoretic manuscripts written in the square script: The texts written in the paleo-Hebrew script were probably copied from other texts written in the paleoHebrew script or in the early Hebrew script rather than from texts written in the square script, allowing us to obtain information concerning an early period in the history of the biblical text.- pp. 206-8 7 Referring to the medieval representatives of m, de Lagarde surmised that all of them attest to a single archetype in antiquity because all the copies reflect its distinctive scribal features, such as the extraordinary points+ and distinct common errors, for example in Samuel.- pp. 171, 189. This assumption is reasonable, but the date of the archetype of the medieval manuscripts (consonantal text only) is probably later than the 51h-3rdcenturies BCE mentioned above.

IA: Prato-Masoretic Texts and the Masoretic Text (In-Group)

29

m, the best complete representative of the medieval text. The evidence of the latter group (b) is earlier than that of the former (a), but typologically it reflects a later development. (a) All the 25 texts that were found in the Judean Desert at sites other than Qumran display almost complete identity with codex L. 8 These texts were found at both the earlier site of Masada (texts written between 50 BCE and 30 CE)9 and the later sites of Wadi Murabba'at, Wadi Sdeir, Nahal Bever, Nahal Arugot, and Nahal Se'elim, dating to the period of the Bar Kochba revolt in 132-135 CE (texts copied between 20 and 115 CE). This identity can be seen best in an examination of the well-preserved texts MasPs 3 (end of the 1st century BCE), MasLevb (30 BCE30 CE), 5/6HevPs (50-68 CE), and MurXII (c. 115 CE), as shown by Young* 10 and Lange* 2010. Recognizing that few differences exist between L and the other medieval sources of m, we note that these differences are of the same nature as those between L and the Judean Desert texts. The relation between L and the ancient Judean Desert texts is one of almost complete identity 11 since the consonantal framework of m changed very little over the course of one thousand years. Some translations that derive from this period reflect m: some of the Targumim and 5 (to a lesser extent), both originating during either the second or third stage of m, as well as two revisions of the oc+, kaige-Th+ probably dating to the middle of the pt century BCE and Aquila+ probably dating to 125 CE. -+ pp. 143-4, 149, 152. In fact, all translations that originated from the pt century CE onwards are bound to reflect the Hebrew text that was to become the medieval m. At that time, there simply were no other Jewish Hebrew texts to be translated. Besides, the Targumim were the in-house texts of rabbinic Judaism and therefore by definition they were based on m. Reflecting this situation, this book uses the siglum m+ referring to the combined evidence of m ([ 5 0. 12 The key to understanding the background of the different scrolls found in the Judean Desert lies in the correlation between their nature and the socio-religious background of the archeological sites. -+ Tov* 2008. What the earlier site of Masada (scrolls written between 50 BCE and 30 CE) and the Bar Kochba sites 8

The agreement between MasLevb, one of the manuscripts in this group, and codex L pertains to the intricacies of orthography, including details in which the orthography ad lac. goes against the conventions elsewhere in the book such as the defective in Lev 9:2, 3 (col. I II, 13) and the defective hiph 'i/ form 1::l1P'1 in Lev 9:9 (col. I 21). Likewise, all the Qumran manuscripts represent the anomalous m form as while only MasLevb has (Lev 10:17; 11:6). 9 Good sources for analysis are MasPs 3 and MasLevb (see below). 10 Young records the number of variants from m included in each text, and then calculates the ratio of variation for each text by dividing the number of preserved words by the variants. The lower that number, the greater is the divergence from L. In this way, he demonstrates the clear difference between the status of the Qumran scrolls as somewhat remote from m, and those from the other Judean Desert sites as identical tom. 11 There has been some discussion as to which terminology best describes the consonantal framework of the proto-Masoretic texts of the last century BCE, but their almost complete identity is stable. However, on the basis of his study of MasGen and MasLev 3 , Ulrich* 2003, 456 challenges the closeness of these scrolls tom. 12 The readings of kaige-Th, Aquila, Symmachus, and the Hexapla usually join the evidence of m ([ 5 D, but since their evidence is fragmentary, the siglum !H+ does not cover them.

30

Chapter 2: Textual Witnesses

(scrolls written between 20 BCE and 115 CE) have in common, in contradistinction with the Qumran scrolls (analyzed on p. 31), 13 is that the people who left the scrolls behind at these sites (the Masada rebels and the freedom fighters of BarKochba) closely followed the guidance of the Jerusalem spiritual center in religious matters. They exclusively used the proto-Masoretic ("proto-rabbinic") text embraced by the spiritual leadership of Jerusalem. Some scholars even stress the priestly influence on the leadership of the revolt.-+ Tov* 2008, 177 As a background explanation for the virtual lack of differences between the copies of m in early times and through the centuries, Tov* 2008 pointed to rabbinic traditions regarding precision in the copying of scrolls, 14 the existence of a master copy of the Torah in the Temple Court, 15 and the correction procedure of scrolls according to this master copy. 16 On the basis of these traditions, Tov 13 The differences between the two sites are most visible when texts from the same period are compared. All the texts copied from the beginning of the 1'' century CE until 65 CE from Qumran are of a varied nature, with only very few m-like texts, while the texts from the same period (until 115 CE) from Judean Desert sites other than Qumran only reflect m.-+ Tov, "Diffusion" 14 The meticulous care in the transmission of m is reflected in the words of R. Ishmael: "My son, be careful, because your work is the work of heaven; should you omit (even) one letter or add (even) one letter, the whole world would be destroyed" (b. Sot. 20a). This precision also pertained to matters of orthography, since various halakhot, "religious instructions," were seemingly based on the precise spelling of words. For example, the number of the walls of the sukkah (four) is determined according to the majority m spelling (b. Sukk. 6b), disregarding a spelling n1:>10 with five letters (cf. 1Qlsaa in [sa 1:8 [p. 105]) or n:>o (Neh 8:15) with three letters. 15 Rabbinic sources deriving from a period later than the Judean Desert evidence provide descriptions of earlier textual procedures, which were also their own. In these descriptions, we read of a master copy of the Torah found in the Temple Court, and of scrolls copied from or revised according to that copy. The term sefer ha- 'azarah (.,ElO i1.,Hlii, with a variant K.,llJ .,ElO, the book of Ezra) probably referred only to the Torah, but it stands to reason that (the) other Scripture books were also found in the Temple. For details, see Tov* 2008, 177-9. The depositing and preserving of holy books in the Temple, as implied by various references in rabbinic literature, resembles the modern concept of publication. -+ Lieberman, Hellenism, 85-7. This procedure can be paralleled by evidence from ancient Egypt, Greece, and Rome. Thus, already, Eichhorn, Einleitung (4'h ed.; 1823) 1.19. See further, ]. Leipoldt and S. Morenz, Heilige Schriften: Betrachtungen zur Religionsgeschichte der antiken Mittelmeerwelt (Leipzig: Harrassowitz, 1953) 89-91, 165-71; R.T. Beckwith, The Old Testament Canon of the New Testament Church and Its Background in Early Judaism (London: SPCK, 1985) 80-86; M. Haran, "Scribal Workmanship in Biblical Times: The Scrolls and the Writing Implements," Tarbiz 50 (1980-1981; Heb.) 65-87 (70-71). These parallels in external sources were stressed greatly in the argumentation of Kutscher, Language, 82 in favor of the assumed presence of master scrolls in the Temple. 16 On several occasions, rabbinic literature mentions a "corrected scroll," sefer muggah. The Temple employed professional maggihim, "correctors" whose task it was to safeguard precision in the copying of the text: i1:>W'?i1 nr:mnt:l )"1:>W J''?01l 1'i1 C'-,ElO 'i1'Jt:l, "maggihim of books in Jerusalem received their fees from the Temple funds" (b. Ketub. 106a). This description implies that the correcting procedure based on the master copy in the Temple was financed from the Temple resources that thus provided an imprimatur. This was the only way to safeguard the proper distribution of precise copies of Scripture. These scrolls must have been used everywhere in Israel, for public reading

IA: Prato-Masoretic Texts and the Masoretic Text (m-Group)

31

postulated that the Judean Desert scrolls were in fact "corrected copies" that circulated in ancient Israel. The textual unity described above has to start somewhere, and the assumption of a master copy is therefore necessary.l 7 (b) Many Qumran scrolls, copied between c. 250 BCE and 68 CE, are very similar to codex L but not almost identical to it as are the other Judean Desert texts, and they form a sizable group among the Qumran scrolls. They are named here nt-like texts.18-+ p. 108. Among them, 4QGenb (Gen 1:1-4:11) is identical to codex L, while all other texts display several differences in small details and orthography+, always remaining close to L, and differing from other textual witnesses.I9 This closeness is exemplified by a listing of the relatively few differences from L in 1Qisab, dating to the 1'1 century BCE. -+plate 6*. Table 1 shows the close relation, sometimes almost identity, between these two texts. In col. XXI (Isa 48:17-49:15, remains of 20 verses), one finds 20 differences all concerning minutiae: 9 differences in orthography and 11 minor, mainly linguistic, differences.20

Table 1 Differences between Land 1Qlsab in !sa 48:17-49:1521 lQisab 48:17 18 20 21 49:3 4

17 18 19

20

21

Codex L

t('?i i11

•n'-,:l

't:1''?.:;l

as well as for instruction, public and private, as suggested by b. Pesah. 112a, where one of the five instructions of R. Akiba to his student R. Simeon was: "and when you teach your son, teach him from a corrected scroll." Another such precise copy was the "Scroll of the King," which accompanied the king wherever he went. Y. San. 2.20c and Sifre Deuteronomy 160 (ed. Finkelstein [New York/Jerusalem: Jewish Theological Seminary, 1969] 211) tell us that this scroll was corrected to "the copy in the Temple Court in accordance with the court of seventy-one members." This suggestion was already voiced by Krauss, Talmudische Archiiologie, III.171; Lieberman, Hellenism, 22; Greenberg, "Ancient Versions," 141-2. Lange, Handbuch, 16 and passim names them-like texts from Qumran "semi-masoretic." It remains subjective to characterize a Qumran scroll as being close or identical to codex L; however, such a characterization is probably correct as long as the number of deviations from L is "small," that is, less than 2% (thus Lange, Handbuch, 16). Ideally, the text common to codex L and the Qumran text should differ from other ancient witnesses, but this condition cannot always be met due to a paucity of evidence. The details are included in OJO XXXII. For a detailed comparison of 1Qlsab with other sources and a discussion of its broto-Masoretic status, see Barthelemy* 1992, cii-cxvi; M.G. Abegg, "1Qlsaa and 1Qlsa : A Rematch," in The Bible as Book, 221-8. Against the majority view regarding the character of 1Qlsab, G. Garbini, "1Qlsab et le texte d'Isare," Hen 6 (1984) 17-21 suggested that this scroll is quite remote from m. According to OJO XXXII, col. XXI, reproduced in plate 6* below.

32

Chapter 2: Textual Witnesses 5

',pm

6

•n.o1ib

c•p;:r7

riN 7

;"11;"1"

;"11;"1"

',N,, 101p•

tzmp 8 10

tzi"'Tj?

noo[iD] "]lbO

Table 1 refers to one column in lQisab. The complete text of lQisab (segments of 46 chapters) displays the following types of differences between the scroll and codex L+, all of which concern minutiae.

Table 2 Types of Differences between 1Qlsab and Codex L22 Orthography Addition of conjunctive waw Lack of conjunctive waw Article (addition/ omission) Differences in consonants Missing letters Differences in number Differences in pronouns Different grammatical forms Different prepositions Different words Omission of words Addition of words Different sequence

107 16 13 4 10 5 14 6 24 9 11 5 6 4

To this almost complete identity of the medieval m, the proto-Masoretic and m-like texts from the Judean Desert in orthography, one should add the similarities in the details of the layout of the poetical texts and the text divisions. - pp. 61-2;202,n.22 iii. The third stage of transmission, characterized by a relatively high level of textual consistency (except for the Severus Scroll, whose text differs from m in small details- pp. 112-13), extends from the Bar Kochba revolt (132-135 CE) until the 81h century CE. Most of the witnesses for this period pertain to its latter end: In the 1890s, more than 200,000 fragments of manuscripts, among them tens 22 According to Cohen, qdwst hnwsh, 86, n. 4.

IA: Prato-Masoretic Texts and the Masoretic Text (m-Group)

33

of thousands of biblical fragments, dating from the 9th century onward, were found in the Cairo Genizah+."2 3 All these texts reflect m...... plates 14*, 15* The following translations reflect the medieval text of m: the Targumim and 5 (to a lesser extent), both originating during either the second or third stage of m, Symmachus+, the fifth column of the Hexapla+, and D ...... pp. 145-9, 152-3. (Note: the siglum m+ refers to the combined evidence of m ([ 5 D ..... notes 4, 12). The great majority of the biblical quotations in rabbinic literature24 and the piyyu_tim (liturgical hymns)2 5 reflect the text of m. The following examples of divergences from rabbinic literature point to exceptions rather than the rule. -+ n. 24

Table 3 Differences between Codex Land Biblical Quotations in Rabbinic Literature Isa 1:1

m Gen. Rab. 13.1

Isa 1:3

m Sifre Deut 309 ed. , (p. 349) 26

Isa 1:18

m

Sifre Deut 6 ed.

J',tv:;,) 1

(p. 15); 28 MS 1 (p. 45)

p1:Jm !'(', 'r:lll 'r:lll, (c:;,'l'(t:ln 1w t:Jl'() 1Qisaa)

23 In 2011, most of these fragments remain unpublished, available only on microfilm and by electronic means. For annotated catalogues, see M.C. Davis, Hebrew Bible Manuscripts in the Cambridge Genizah Collections, vols. 1-4 (Cambridge: Cambridge University Library, 1978-2003); M. Ducan, Bibliotheque de /'Alliance Israelite Universelle: fragments bibliques en hebreu provenant de guenizot (Turnhout: Brepols, 2008); I. Yeivin, Geniza Bible Fragments with Babylonian Massorah and Vocalization (Heb.; Jerusalem: Makor, 1973). For an evaluation, see Kahle, Cairo Geniza, 3-13; J. Hempel, "Der textkritische Wert des Konsonantentextes von Kairener Geniza-fragmenten in Cambridge und Oxford zum Deuteronomium nach Kollationen von H.P. Ruger untersucht," NAWG I, Phil.-hist. Kl. 1959, 10, pp. 207-37; M.H. Goshen-Gottstein, "Biblical Manuscripts in the United States," Textus 2 (1962) 28-59; M. Saeb0, "Bemerkungen zur Textgeschichte von Leviticus. Welchen Wert haben die Varianten aus der Kairoer Geniza?" in Die Hebriiische Bibel und ihre zweifache Nachgeschichte: Festschrift for Rolf Rendtorff zum 65. Geburtstag (ed. E. Blum et al.; Neukirchen-Vluyn: Neukirchener Verlag, 1990) 131-9. 24 The biblical quotations in rabbinic literature differ occasionally from m, both in direct quotations and in the exposition. For an analysis and for the history of research, see Y. Maori, "The Text of the Hebrew Bible in Rabbinic Writings in the Light of the Qumran Evidence," in The Dead Sea Scrolls: Forty Years of Research (ed. D. Dimant & U. Rappaport: Leiden/Jerusalem: Briii/Magnes and Yad Ben-Zvi, 1992) 283-9, and id., "Rabbinic Midrash as Evidence for Textual Variants in the Hebrew Bible: History and Practice," in Modern Scholarship in the Study of Torah: Contributions and Limitations (ed. S. Carmy; Northvale, NJ: J. Aronson, 1996) 101-29. The material has been collected by S. Rosenfeld, spr mspbt swprym (Vilna: Romm, 1882); V. Aptowitzer, Das Schriftwort in der rabbinischen Literatur, vols. I-IV (Vienna: Holder, 1906-1915; repr. New York: Ktav, 1970) pertaining to Joshua-Samuel. The most complete collection is found in the HUB. ..... pp. 357-9 25 Cf. M. Wallenstein, "The Piyyut, with Special Reference to the Textual Study of the Old Testament," BJRL 34 (1952) 469-76. 26 Ed. Finkelstein (New York/Jerusalem: Jewish Theological Seminary, 1969).

Chapter 2: Textual Witnesses

34

m

Jer 30:4

I

Sifre Deut 1 1, t:l,

m

Hab 1:13

i1 ,:::l, ,iZlN C',:::l1i1 i1'?N1 ,:::l1 ,iZlN C',:::l1i1 i1'?N1

'? .ll,

Pesiq. Rab Kah. 4.10; 25.1

n1N,O n1N,O (=

1QpHab)

Table 3 does not include cases of 'al tiqre (..... p. 55) or of midrashim+, as these reflect an exegetical play with readings that would have been possible in the context. 27 All textual evidence preserved from the second period reflects m because the communities that fostered other textual traditions either ceased to exist (the Qumran covenanters) or dissociated themselves from Judaism (Samaritans and Christians) ...... pp. 178-80 iv. The fourth stage of transmission, characterized by almost complete textual unity, extends from the 81h century until the end of the Middle Ages. The main sources for this period are Masoretic manuscripts containing the apparatus of the Masorah and biblical quotations in the writings of the traditional Jewish commentators. 28 The earliest dated Masoretic manuscripts are from the 91h century. 29 During this period m became almost completely standardized, due largely to the addition of the apparatuses of vocalization+, accentuation+, and Masorah+, which necessitated the fixation of the consonants that formed their base. The sources from this period are subdivided into manuscripts from the early Middle Ages (until about 1100) and later manuscripts. The early manuscripts are more reliable.

Table 4 Differences between Medieval Masoretic Manuscripts a. Variants Collated by Kennicott 1776-1780 and de Rossi 1784-1788 Gen 1:14

all

MSS

C'OiZli1 .ll'p,:::l n1NO 'i1' C'i1'?N ,ON'1

MS 776

Lev 10:1 1 Kgs 11:20a

of Kennicott adds r,Ni1 '?.ll ,'Ni1'? (= tu (this addition is influenced by v 15 .ll'P,:::l n11NO'? 1'i11 p 'i1'1 r,Ni1 '?.ll ,'Ni1'? C'OiZli1 and v 17) all MSS N1i1':::lN1 :::l1J )1i1N 'J:::l 1np'1 MSS 5,181 of Kennicott 'm 'J:::l mp'1 (= all MSS i1.1l,5:1 11n:::l OJ5:1nn 1i1'?0Jn1 MSS 23,154,182,271A,283A of Kennicott (= cf. v 20b i1.1l,5:1 'J:::ll1n:::l)

27 See the extensive discussion by D. Rosenthal, "The Sages' Methodical Approach to Textual Variants within the Hebrew Bible," in /sac L. Seeligmann Volume, Essays on the Bible and the Ancient World (ed. A. Rofe & Y. Zakovitch; Heb. with Eng. summ.; Jerusalem: Rubenstein, 1983) Yelin, hdqdwq, 183-5; Goshen-Gottstein-Talmon, HUB, Ezekiel, xxxi, n. 86. 28 See e.g. S. Esh, "Variant Readings in Mediaeval Hebrew Commentaries; R. Samuel Ben Meir (Rashbam)," Textus 5 (1966) 84-92; E.L. Greenstein, "Misquotation of Scripture in the Dead Sea Scrolls," in The Frank Talmage Memorial Volume (ed. B. Walfish; Haifa: Haifa University Press, 1993) 71-83 (71-3). 29 According to Birnbaum and others, a sheet of a Torah scroll was written somewhat earlier: S.A. Birnbaum, "A Sheet of an Eighth Century Synagogue Scroll," VT 9 (1959) 122-9.

IA: Prato-Masoretic Texts and the Masoretic Text (!TI-Group) 1 Kgs 12:12

all

MSS

MS

202 of Kennicott

cll:m, Cll:Jn,

35

(= 1!3)

b. Early Manuscripts ofm (following Breuer* 1976, index) Josh 3:3

L A,C,S 1, RB1-2 MSS L, RB1-2 MSS A,C,S 1 MS 5 1 MSS A,L,C, RB1-2 MS C MSS A,L,S 1, RB1-2 MS

MSS

Josh 3:4 mK Josh 6:6 Josh 6:9

1J,:J1 1,J:J1

';,

r1,1:::l

';, m::ntDi1 n1,::l1tDi1

The differences in group b in Table 4 characterize the minute differences between the medieval manuscripts. Group a records greater differences that are less characteristic of this period. Divergencies within the m-group are also reflected in three groups of differences that have been institutionalized in m:

Ginsburg, Introduction (1897) 197-240; id., "On the Relationship of the So-called Codex Babylonicus of A.D. 916 to the Eastern Recension of the Hebrew Text," Recueil des trauaux ... M.D. Chwolson (Berlin: Calvary & Co., 1899) 149-88; Yeivin, Introduction, 139-41.

Even though the scribes of m meticulously preserved a uniform text, breaches in this unity are occasionally visible. Among the early sources of m, consonantal differences existed between manuscripts from the West (Palestine) and manuscripts from the East (Babylon). Some 250 such differences are mentioned in the notes of the Masorah+ as 'e, "Masoretes of the East" and Ma 'arba 'e, "Masoretes of the West." For example, 2 Kgs 8:16b

C,1i1' (Jehoram)--=--

C,1' (Joram)

For "Jehoram," a variant "Joram" is known, as recorded by the Masoretes of the East.

Notes of this type referred especially to differences between Ketib+ and Qere+. For example, Job 17:10 (BH, not BHS) Ma 'arba'e: MedinM'e:

znK

("and come!") ("they will come") ("and come!")

Most of the manuscripts of m that have been preserved are Tiberian .... p. 24, that is, Western, and therefore the majority of the Masoretic notations comparing different traditions refer to readings of the when the Tiberian manuscripts differed from Eastern sources. These differences were collected in the Middle Ages in separate lists that preserved evidence of this type even if in the manuscripts themselves such evidence was lacking or was not denoted

36

Chapter 2: Textual Witnesses

consistently. The lists refer only to discrepancies in consonants, which may indicate that their origin must have been early, before vowels and accents were inserted.

Variants in Manuscripts Written in Different Vocalization Systems B. Chiesa, L'antico Testamento ebraico secondo Ia tradizione palestinese (Torino: Bottega d'Erasmo, 1978); G. Miletto, L'Antico Testamento ebraico nella tradizione babilonese i frammenti della Genizah (Quaderni di Henoch 3; Torino: Silvio Zamorani, 1987); id., "II testa consonantico della tradizione babilonese: Ms. Opp. Add. 4 154 della biblioteca bodleiana," Sefarad 54 (1994) 333--61.

The Masoretes added the vowels to the consonants after they had been fixed, and therefore one would not expect to find many consonantal differences among the manuscripts writ.ten in different vocalization systems. - pp. 39-47. Nevertheless, manuscripts vocalized in the Palestinian and Babylonian tradition often differ from Tiberian manuscripts in consonants. - Chiesa* and Miletto* 1987, 199430 y. Masoretic Notes

The Qere readings recorded in the Masorah preserve many ancient variants to the Ketib. - pp. 54-9 All these internal differences within the m-group point to a certain degree of textual variation at an early stage of the development of m, in contrast with its later unity. The above-mentioned differences were institutionalized in the notation of the Masorah+, but a still larger number of internal differences have not been recorded. c. Origin

The almost complete identity of the medieval m with the proto-rabbinic texts found in the Judean Desert and the assumption that this identity derives from the copying from a single source, probably the Torah Scroll of the Temple Court+ - pp. 30-31, leads to further thoughts about the background of m. Since m contains a carefully transmitted text, that is well documented in a large number of copies, and since it is reflected in the rabbinic literature as well as in the Targumim and several JewishGreek revisions of it may be surmised that it originated in protorabbinic circles. - pp. 29-30. Though this assumption cannot be proven, it is supported by the late tradition that the Temple employed correctors maggihim) who scrutinized scrolls on its behalf. - p. 30, n. 16 30 Differences of this type are included in Dfez Merino, Biblia babi/6nica and in the HUB+. On the other hand, the internal unity of the Hebrew tradition is emphasized by E.J. Revell, Biblical Texts with Palestinian Pointing and Their Accents (SBLMasS 4; Missoula, MT: Scholars Press, 1977).

IA: Prato-Masoretic Texts and the Masoretic Text (m-Group)

37

However, there is a snag in this description. A strong desire for textual unity existed within them-group, but this tendency could not erase the differences that already existed among copies belonging to the same group. In the last centuries before the Common Era and the first century of the Common Era, almost identical copies, such as those found in the Judean Desert at sites other than Qumran ..... p. 29, coexisted with less precise copies of the same group, the m-like texts ..... p. 31. Phrased differently, the wish to preserve a unified textual tradition remained an abstract ideal that could not be accomplished for the country as a whole ...... Cohen, qdwst hnwsl1. The various texts of them-group from the Second Temple period thus differed from one another, but over the course of the centuries the number of these differences decreased due to the addition of vocalization+, accentuation+, and the apparatus of the Masorah+. These notations added an element of precision and prevented changes in consonants.

d. Evidence

The differences among the large numbers of medieval manuscripts have been recorded in several editions and collections of variants. The first five collections mentioned below pertain mainly to manuscripts written after 1100, while the last two, more recent, collections also include variants in early manuscripts. 31 All collections except for that of Kennicott include variants in both consonants and vowels. • Min/:lat Shay (1626) by Yedidya Shelomo Norzi was printed in various Bible editions starting with Mantua 1742-1744, among them the Rabbinic Bibles+, and subsequently also in a separate edition (Vienna, 1813-1815). 32 Z. Betzer, Minhat Shay on the Torah: Critical Edition and Notes (Jerusalem: World Union of Jewish Studies, 2005) provides a new edition of Min/:lat Shay to the Torah. • B. Kennicott, Vetus Testamentum hebraicum cum variis lectionibus, vols. 1-11 (Oxford: Clarendon, 1776-1780) ...... plate 18*. This influential edition lists medieval variants from more than 600 different European manuscripts. The variants to mare grouped around the text of van der Hooght's edition (Amsterdam/Utrecht, 1705) and those to w. are grouped around the text of w. from Walton's London Polyglot (1653-1657). ..... p. 71. The details of the Hebrew manuscripts are described in Kennicott's Dissertatio generalis and id., Dissertation. • J.B. de Rossi, Variae lectiones Veteris Testamenti, vols. I-IV (Parma: Regio, 1784-1788; repr. Amsterdam: Philo, 1969). The variants collected in this edition from several hundred manuscripts were meant to supplement the Kennicott 1776-1780 edition. • J.C. Doderlein & J.H. Meisner, Biblia Hebraica . . . ad fidem recension is masoreticae cum variis lectionibus ... Benj. Kennicotto et f. B. de-Rossi (Halle /Berlin: 31

Apart from the editions mentioned in this section, the first printed editions contained readings not known from other sources: ..... p. 70. It appears that the editors of these editions had access to manuscripts that were subsequently lost. 32 See Y. Ofer, "Methods and Sources of Yedidya Shelomo Norzi in His Treatise Minhat Shay," Textus 24 (2009) 287-312.

38

Chapter 2: Textual Witnesses

Libraria Orphanotrophei, 1818). This useful edition records the most important variants from the editions of Kennicott 1776-1780 and de Rossi 1784-1788. • C.D. Ginsburg. - p. 73 • HUB+ (Isaiah, Jeremiah, Ezekiel). See the third and fourth apparatuses in plate 30*. - ch. 9B2 • BH+ and BHS+ (plates 27*-28*) quote from the collections of Kennicott 17761780 and de Rossi 1784-1788 without detailed information on the manuscripts themselves. -Table 5

e. Significance of the Differences between Medieval Manuscripts Scholarly opinions concerning the value of the differences among the medieval manuscripts are divided. Many scholars, among them the editors of BH and BHS, attach considerable significance to the readings attested to in the above-mentioned collections, while others are more reserved. Scholars who value the readings contained in medieval manuscripts are influenced by procedures developed in biblical criticism in the previous centuries rather than by content considerations relating to the readings themselves. This is due to the fact that manuscripts from the Middle Ages formed the almost exclusive source of information for the study of Hebrew Scripture when critical biblical scholarship began to develop, and every detail in those manuscripts received attention that exceeded its real significance. Thus, in BH-BHS, the number of manuscripts (according to Kennicott 1776-1780) evidencing a certain reading is mentioned specifically.- p. 355; plates 27* and 28*; and Table 5:

Table 5 Quotations in BHfrom Medieval MSS (according to Kennicott 1776-1780) Isa 1:3 Jer 14:14 Ezek 7:5

ca 30 MSS II} 50 Bet nonn MSS ca 30 MSS Edd were created," Gen 2:4) explained as two words, :1:!, "with the letter he," and "He created them."

g. Ketib-Qere J. Barr, "A New Look at Kethibh-Qere," OTS 21 (1981) 19-37; M. Breuer, "'mwnh wmd' bnwsb hmqr '," Death 47 (1978) 102-13; P. Cassuto, "Qere-Ketiv et Massora Magna dans Ie manuscrit B 19a," Textus 15 (1990) 84-119; Maimon Cohen, The KethiQ.and Qeri System in the Biblical Text: A Linguistic Analysis of the Various Trad.itions Based on the Manuscript 'Keter Aram Tsova' (Heb.; Jerusalem: Magnes, 2007); Gordis, Biblical Text; Y.M. Grintz, mbw ·y mqr · (Tel Aviv: Yavneh, 1972) 60-82; S. Levin, "The 'ip as the Primary Text of the 1"Jn," Hagut Ivrit be'Amerika I (Heb.; Tel Aviv: Yavneh, 1972) 61-86; Moshkowitz-f:lamiel, Introduction, 1.7286; Y. Ofer, "Ketiv and Qere: The Phenomenon, Its Notation, and Its Reflection in Early Rabbinic Literature," Leshonenu 70 (2008) 55-73 (Heb.); H.M. Orlinsky, "The Origin of the Kethib-Qere System: A New Approach," VTSup 7 (1960) 184-92; Morinus, Exerc., 603-34; Sperber, Grammar, 493-510; J. Simonis, Analysis et explicatio lectionum masorethicarum, Kethiban et Krijan vulgo dictarum, Ea forma, qua il/ae in textu 5. exstant, Ordine alphabetico digesta (Amsterdam: Wetstein, 1753); E. Tov, "The Ketiv-Qere Variations in Light of the Manuscript Finds in the Judean Desert," in id., HB, GB, and Qumran (2008) 199-205; G.E. Wei!, "Qere-Kethib," IDBSup, 716-23; Yeivin, Introduction (1980) 52--62.

toors: Ketib-Qere variations are searchable in Accordance, BibleWorks, Logos, SESB.

Qere notations are found only in medieval manuscripts of m and printed editions, and not in ancient scrolls. In a large number of instancesranging from 848 to 1566 in the different sources-the Mp notes that one should disregard the written form of the text (in the Aramaic language of the Masorah: ::::l't:1f, k'tib, "what is written") 69 and read instead a different word or words (in Aramaic: 'Jj?, qere, or '!)?, qeri, "what is read").7° Different systems are used in manuscripts and editions for recording the K-Q since the vocalization of the Ketib has not been transmitted in the manuscripts. The vowels of the Ketib forms are hypothetically provided by Simonis* 1753, Ginsburg (- p. xxi), and computer modules such as Accordance, BibleWorks, Logos, and SESB. In the manuscripts and most editions (e.g. Ginsburg, the BH series, HUB, Cohen, Miqra 'ot Gedolot "Haketer"), the vowels of the Qere are superimposed on the consonants of 69 A few of the Qere words have been indicated in some Masorah manuscripts as yatir, "superfluous" (usually: yatir yod, or yatir waw), i.e., when reading, one must disregard the yod or waw. For an example, see Josh 10:24 (p. 216, Table 6). 70 In early manuscripts, the Qere was sometimes denoted by a circullus or, more frequently, a vertical sign similar to a final nun or possibly zayin (Yeivin* 1980, 52). This system is also followed in BHQ.

IA: Prato-Masoretic Texts and the Masoretic Text (m-Group)

55

the Ketib, while the unvocalized Qere is presented elsewhere, mainly in the margin as part of the Mp+? 1 Josh 6:13

mK mO

2 Sam 22:51

mK mO

'1'?1;:r l1'?i1 (= li'?;:r) (consonants= Ps 18:51 (=

The evidence for the Ketib and Qere in the manuscripts of m is relatively late, but the practice itself was already mentioned in the rabbinic literature (the opinions of the medieval commentators are quoted in Sperber* and MoshkowitzI-:Iamiel*). For example, b. 'Erub. 26a notes that in 2 Kgs 20:4 "It is written 'the city,' but we read 'court'." Manuscripts and editions likewise indicate here: Ketib ,,lm, "the city," Qere "court." 73 Rabbinic literature also mentions 'al tiqre formulas phrased as "do not read ( 'al tiqre) X, but Y," but their nature differs from the Qere system. These formulas do not necessarily reflect readings that would have been known to the rabbis. Rather, they point to an exegetical play on words, especially words with an added or omitted mater lectionis+ that would have been possible in the context.74

Occurrences of the "constant Qere" (Qere perpetuum) are not indicated with a Masoretic note, but in such cases the Ketib is vocalized with the on the basis vowels of the Qere. Thus mK ii1i1', YHWH, is vocalized as of its Qere 'adonay (or, when appearing next to as representing 'elohim).75 71 In some modern editions (e.g. Koren, Adi, Dotan 2001), the unvocalized but accented Ketib forms are included in the text, while the vocalized and accented Qere is recorded in the margin. In the NJPS 2nd ed., 1999, both forms are juxtaposed in this way in the text itself. In Ginsburg's edition, the vocalized Qere is contained in the apparatus under the text. In other editions (Breuer, Horev, and Jerusalem Crown), both the Ketib (in the text) and the Qere (in the margin) are vocalized with the vowels of the Qere. European translations usually follow the Qere (-+Daley, Textual Basis, 171-202); when following the Ketib, the translations rarely doq.tment their choice, although NJPS 2nd ed. does remark on such choices. -+ pp. xx-xxii for bibliographical references to the editions 72 Accordingly, the Ketib and Qere probably represent two different readings. 73 For further examples, see b. Yoma 21b (on Hag 1:8); b. Ned. 37b; Gen. Rab. 34.8; Sof 7. See also Midrash Qere we-la Ketib included in the collection of A. Jellinek, Bet ha-Midrasch 5 (Vienna: Herzfeld & Bauer, 1873; repr. Jerusalem: Sifre Vahrman, 1967) 27-30. 74 The 'a/ tiqre formulae have been collected by N.H. Torczyner, "1 tqr '," Eshko/, 'n?yqlwpdyh ys'r'lyt (Berlin: Sholem, 1932) ll.376-386 (Heb.) and classified by A. Rosenzweig, "Die Al-tikri-Deutungen," in Festschrift zu Israel Lewy's siebzigstem Geburtstag (ed. M. Brann & J. Elbogen; Breslau, 1911; repr. Jerusalem: Makor, 1972) 20453. For example, b. Ber. 64a: "R. Eleazar said in the name of R. Hanina: The disciples of the wise increase peace in the world, as it says, 'And all thy children shall be taught of the LORD, and great shall be the peace of thy children.' Read not ('a/ tiqre) banayik, 'thy children,' but bonayik, 'thy builders,' or· 'those of you who understand'" (lsa 54:13). It appears that this formulation, as with several others, is based on a variant reading known from the supralinear addition of a waw in 1Qisaa.-+ Talmon, Qumran, 71-116 75 Special scribal practices for writing the Tetragrammaton+ are also known from a Qumran scroll. The dicolon ( : ) before the occurrences of i11i1' in 4QRPb (40364)+

Chapter 2: Textual Witnesses

56

In some instances, the Mp+ directs the reader to read a word that is not included in the text. Qere wela · ketib indicates a word that is "read but not written." In such cases, only the vocalization is included in the text.76 2 Sam 8:3

mK

,,,

•,

to restore his power at the river (P)e(r)a(t) mQ

Judg 20:13

m:l ... Perat

(=

[ 5 D and 1 Chr 18:3)

!11 K

and the (s)o(ns) of Benjamin would not listen sons of

(=

5 [D)

In other instances, the Mp instructs the reader to disregard a word included in the text. Ketib wela' qere indicates a word that is "written but not read." In these cases, the word is not vocalized.7 7 2 Sam 13:33

Mp:

i"':;!"(

':;1

is written and not read"

Background. The Ketib-Qere instances have been classified into different categories in Okhlah we-Okhlah+, Massoreth ha-Massoreth, and studies by Cordis*, Cassuto*, and Mairnon Cohen* 2007.78 While different explanations for the background of this notation have been given, the first one is preferable. a. The Qere Represents a Reading Tradition Undoubtedly, the Qere represented a reading tradition (Levin*; Breuer* 1978; Barr*; Maimon Cohen*, 311; Tov* 2008) that accompanied the written text, rather than a living written recording of differences. The strongest argument in favor of this view is the terminology used referring to a spoken word (Qere) as opposed to a written one (Ketib). This assumption may be supported by the fact that the K-Q is the only para-textual feature of !11 that is not paralleled by the Judean Desert scrolls (-+ Ofer* 2008, 59), and therefore probably developed after the 1"' century CE.

The fact that the manuscripts never contain more than one Qere word also points to a reading tradition, which is naturally limited to one word ...... Barr* 1981. According to this assumption, the Qere words derived from an obligatory text such as an exemplary manuscript. Such a source could also have contained

76 77 78

probably resembles the Qere, indicating that the word should be read differently or not at all. In many Qumran manuscripts, the Tetragrammaton+ is written in paleo-Hebrew+ characters or indicated with four dots (Tetrapuncta+).-+ Tov, Scr. Prac., 218 Several examples are mentioned in b. Ned. 37b-38a and Sof 6.8. Okhlah we-Okhlah+, list 97 provides the complete evidence. The complete evidence is found in Sof 6.9 and Okhlah we-Okhlah+, list 98. In addition to the examples of Ketib-Qere given in this many instances are mentioned elsewhere, especially inch. 4C (see Index 3).

IA: Prato-Masoretic Texts and the Masoretic Text (In-Group)

57

inferior variants, and therefore not all the corrections of this type were necessarily preferable.

The Written Qere Corrects the Ketib as Indicated by the Masorah According to this assumption, the Qere words were originally added to the written text as corrections replacing the Ketib text. Maimon Cohen*, 311 points to the early linguistic background of many Ketib forms replaced by later Qere forms. Further, the perpetual Qere of YHWH as 'adonay (p. 55) is a correction, as are the replacements of possibly offensive words with euphemistic expressions. 79 The main examples are: Deut 28:27 80

mK mO

c•'?D.tJ:n c•int:l:n

and with hemorrhoids (?) and with tumors (?)

Deut 28:30 81

mK mo

i1J:::l::ltD'

he shall enjoy (?) her he shall lie with her

However, some aspects of the assumption of a correction are problematical. • Exactly the same words-with identical meaning-sometimes form the Qere word in one verse, and the Ketib word in another one. For example, Gen 39:20 Judg 16:21, 25

mK mO mK mO

(= w.)

There are also many similar Ketib-Qere interchanges in both directions for the pairs C',J.tl n,:::ltD • While several instances of specific Ketib words were "corrected" by a Qere, identical words have not been corrected elsewhere. For example, Gen 24:33

mK mO

1'JD'?) C'i;;tO:::l J1ilot:::l

Gen 50:26

was placed CtD,'1 (= w.) was placed

(w. CtD, '1)

• The Qere words include several forms that are less plausible than the Ketib with regard to either context or grammar. For example, Gen 8:17 2 Sam 3:25

mK mO mK mo

• It is unlikely that the presumed correctional activity would have been limited to similar letters (the Qere words are almost always very similar to the Ketib word).

79 See b. Meg. 25b: "Our rabbis taught: wherever an indelicate expression is written in the Torah, we substitute a more polite one in reading. m'?Jtv', 'he shall enjoy (?) her,' ;"ll::l::ltD', 'he shall lie with her'." See list 2 in Cordis* and Schorch, Euphemismen. For further instances, see Sof. 9.8. For other euphemisms used in biblical manuscripts, see pp. 250-52. 80 The same Ketib-Qere is found in 1 Sam 5:6, 9, 12; 6:4, 5. 81 The same Ketib-Qere is found in lsa 13:16; Jer 3:2; Zech 14:2.

Chapter 2: Textual Witnesses

58

y. The Qere Word Served as a Written Variant to the Ketib The Qere could have been a variant found either in the same manuscript or in a different source. 82 Indeed, most of the K-Q variants pertain to small differences, especially interchanges+ of similar letters, which are also known as variations between manuscripts. For example,83 Josh 3:16 Josh 4:18 Josh 15:47 2 Kgs 16:6

Prov 20:21

mK mO mK mO mK

at Adam from Adam(= (11 cases). Verses in which the second word is i10iD, e.g. Gen 29:3 (10 verses). Verses containing i1Ji1 and afterwards i1Ji11, e.g. Gen 31:51 (5 verses). and afterwards e.g. Verses containing Judg 3:24 (8 verses). Verses in which I:Jj occurs three times, e.g. Gen 24:25 (12 verses).

Additional Masoretic works, early and late (-+ Yeivin, Introduction, 128-55): Diqduqqe ha- ye 'amim by Aaron Ben Asher+ (first half of lOth century).

Masoret Seyag La-Torah by Meir ben Todros ha-Levi Abulafia (11801244). 'Eyn ha-Qore' by Yequti'el ben Yehuda ha-Naqdan (probably from the second half of the 12th century). Shay (1626) by Yedidyah Shelomo from Norzi.-+ p. 37 c. Editions of the Masorah The Masorah of RB2+ was published with a translation and notes by Ginsburg 1880-1905 (p. xxxvii), tog.ether with various Masoretic treatises. S. Frensdorff, Die Massora Magna (Hannover: Cohen & Risch, 1876; repr. New York: Ktav, 1968) contains an index of the Masorah of RB2. Since the Masorah of RB2, culled from different manuscripts, is imprecise, scholars often prefer to consult the Masorah of a specific manuscript, especially those contained in the following four editions: • Wei!, Massorah Gedolah ... Leningrad (1971).-+ p.liv • D.S. Loewinger, Massorah Magna of the Aleppo Codex Oerusalem: Shrine of the Book, 1977). • M. Breuer, The Masorah Magna of the Pentateuch by Shemuel ben Ya 'aqov (Ms. r.'?), vols. 1-2 (New York: Lehman Foundation, 1992). • E. Fernandez Tejera, Las Masoras del libra de Genesis: C6dice Ml de Ia Universidad Complutense de Madrid (TECC 73; Madrid: CSIC, 2004); see also additional volumes in this series.

70

Chapter 2: Textual Witnesses

d. Importance of the Masorah

The main importance of the apparatus of the Masorah is for the internal study of m, and not the comparison of m with the other textual witnesses. 6. Editions ofm L. Blau, "Dr Ginsburg's Edition of the Hebrew Bible," JQR 12 (1900) 217-54; Cohen, "Consonantal Character" (1981 ); id., Miqra 'ot Gedolot "Haketer"; id., "Introduction"; Ginsburg, Introduction (1897) 779-976; id., Jacob Ben Chajim Ibn Adonijah's Introduction to the Rabbinic Bible (London: Longmans, Green, Reader, and Dyer, 1867; repr. New York: Ktav, 1968); L. Goldschmidt, The Earliest Editions of the Hebrew Bible, With a Treatise on the Oldest Manuscripts of the Bible by Paul Kahle (New York: Aldus, 1950); Goshen-Gottstein, Biblia Rabbinica (1972) 5-16; id., "Editions" (1992); Habermann, Ketav; M.J. Mulder, "The Transmission of the Biblical Text," in id., Mikra, 87-135 (116-21, 133-4); H.M. Orlinsky, "Prolegomenon" to Ginsburg, Introduction (1966) x-xviii; Penkower, Jacob Ben-Hayyim (1982); id., "Bamberg's First Bible Edition and the Beginning of His Printing Press," Kiryat Sefer 58 (1983) 586-604 (Heb.); id., "New Evidence" (1992); id., "Ben Asher" (1999); B. Pick, "History of the Printed Editions of the Old Testament, Together with a Description of the Rabbinic and Polyglot Bibles," Hebraica 9 (1892) 47-116; C. Rabin, "mqr', dpwsy hmqr'," EncBib Oerusalem: Bialik Institute, 1968) 5.368-86; Roberts, "Hebrew Bible"; Walton, Polyglotta.

toofs: Tokhnit "HaKeter"-Ma 'agar HaTanakh, Bar-Ilan University, Ramat Gan; part of the Miqra 'of Gedolot "HaKeter" Project, based on codex A.

The text of m has been printed many times from various sources, 101 usually without critical principles.102 Since the Tiberian branch of the Ben Asher+ system of m became the determinative text in Jewish tradition, it was followed in all editions of m and, to all intents and purposes, that text served as the printed text of Hebrew-Aramaic Scripture as a whole. RB2+ was very influential among these editions, and served almost as the "received text" of the Bible. Goshen-Gottstein* 1992. As a consequence, most of the subsequent editions are based on this edition. However, in the second half of the 20th century several editions appeared that are based on a single manuscript (see below). The history of the printing of m is described by Ginsburg* 1897, Rabin*, and Mulder* and only the major facts are mentioned here (the central editions are listed on pp. xx-xxii). For a list of all the editions of Hebrew-Aramaic Scripture, see ch. 9 and Table 1 there (p. 344). 101 The printed editions of rn are often erroneously regarded as representing "the Bible," even though they represent only one textual tradition, albeit a central one. - ch. 964 102 see, for example, the introductory words in the Appendix to the Koren 1962 edition explaining its textual basis: " ... on the basis of the opinions of the Masoretes, the grammarians, and the interpreters and according to what was found in the majority of the manuscripts and printed editions accepted as authoritative, and not as a slavish copy of a specific edition or manuscript" (my italics, E.T.).

IA: Prato-Masoretic Texts and the Masoretic Text (m-Group)

71

Several unvocalized editions of m preceded the first printed editions that included vowels and accents. 103 The first printed edition of the Hagiographa with vowels and accents appeared in Naples in 1487, while the first complete biblical text was printed in 1488 in Soncino, a small town in the vicinity of Milan. These editions were followed by the appearance of several complete or partial editions of Hebrew Scripture. All these editions include many mistakes and idiosyncrasies, described in detail by Ginsburg* 1897. Sometimes words or entire verses were left out or duplicated. The publishers did not indicate which manuscript(s) they used for the edition, but Cohen* 1981 maintains that the consonantal base of the early editions reflects specific manuscripts and not a combination of such sources. The first editions actually included more than may be expected from a text edition, since they included a Targum and some medieval commentaries, mainly by R. David Kimhi (RaDaK) and Rashi. At that time, the earlier manuscripts that were to dominate 20th century editions (codices Land A among others) were not known to the editors or recognized as important sources. Particularly important for the progress of biblical research were the so-called Polyglots, or multilingual editions. With the development of biblical criticism, scholars increasingly based their work on these editions because of their rich content. The Polyglot editions present the biblical text in parallel columns in Hebrew (m and w.), Greek, Aramaic, Syriac, Latin, and Arabic, accompanied by Latin versions of these texts and by grammars and lexicons of these languages, while the earlier editions presented a smaller range of texts. The first Polyglot is the Complutensian (-+ p. xxxiii) prepared by Cardinal Ximenes in Alcala (in Latin: Complutum), near Madrid (1514-1517).1° 4 The second Polyglot was prepared in Antwerp (1569-1572), 105 the third in Paris (1629-1645),106 and the fourth, the most extensive of all, was published by B. Walton* and E. Castellus in London (1653-1657). The textual basis of the last three editions is unclear. The first Polyglot edition was followed by the Rabbinic Bibles (later to be named Miqra 'ot cedolot, "folio edition"). The name of these editions derived from the inclusion of several medieval commentaries on m and one or more Targumim ...... plate 26*. The first two Rabbinic Bibles (RB) were printed at the press of Daniel Bamberg in Venice; the earlier one, RB1 (1516-1517), was edited by Felix Pratensis and the later, RB2 (1524-1525), by Jacob Ben-Hayyim ben Adoniyahu. RB2 differed from RB1, among other things, due to the addition of 103 Psalms (Bologna (?], 1477), Torah (Bologna, 1482), Prophets (Soncino, 1485-1486), Hagiographa (Naples, 1486-1487). 104 According to Ginsburg, Introduction, 917-25, m in this edition was based on at least three sources, MS Madrid University Library codex 1 (MS 59 of Ginsburg*, 771), the Naples Bible (1491-1493), and the Lisbon Pentateuch (1491). According to P. Kahle, "The Hebrew Text of the Complutensian Polyglot," in Homenaje a Millds-Vallicrosa (Barcelona: CSIC, 1954) 1.741-51, some of the manuscripts used for this edition were close to the Ben Asher+ tradition. On \13 in this edition, see S. O'Connell, From Most

Ancient Sources: The Nature and Text-Critical Use of the Greek Old Testament Text of the Complutensian Polyglot Bible (OBO 215; Fribourg/ Gottingen: Academic Press/Vandenhoeck & Ruprecht, 2006). 105 Biblia Sacra Hebraice, Chaldaice, Graece et La tine etc. (Antwerp: Christophorus Plantinus, 1569-1572). 106 Biblia Hebraica, Samaritana, Chaldaica, etc. (Paris: Antonius Vitre, 1629-1645).

72

Chapter 2: Textual Witnesses

the Masorah, which held a special place in that edition. Further differences between the two editions include a number of readings, mainly in orthography, vocalization, and accentuation (1% of all the details according to Penkower* 1982, vii). For a modern version of the Miqra 'ot cedolot, see Cohen, Miqra 'ot Gedolot "Haketer" (1992-2007). 107 This edition follows the layout of RB2, but its base text is the Aleppo codex ...... ch. 9, Table 1 (p. 344) The Rabbinic Bibles were based on several unnamed late 108 manuscripts, to which the editors applied their editorial principles. The editor of RB1 based his biblical text on several Ashkenazi manuscripts and the first printed editions (Soncino 1488; Brescia 1494), while RB2 was based upon several "accurate Spanish manuscripts" that were close to the text of the "accurate Tiberian manuscripts" such as L and A.1° 9 Sometimes, the editors changed details of these manuscripts in their editions according to own grammatical insights; this applies especially to the system of the ga 'yot+ of RB2. RB1 recorded some variants from other manuscripts in the margins. Because of the inclusion of the Masorah, Targumim, and traditional Jewish commentaries in RB2, that edition was hailed as the Jewish edition of the Hebrew Bible. Consequently, all subsequent editions, with the exception of a few critical editions issued in the second half of the 20 1h century, reflect this edition. 11 0 The new editions deviated from RB2 only when changing or adding details on the basis of other manuscripts or when removing or adding printing errors ...... pp. 89

RB2 became the leading edition because of its status within Judaism and the scholarly world. However, the uncertainty regarding the textual base of this edition is problematic for precise scholarship, and therefore several new editions have tried to improve upon RB2 in various ways. 111 Sometimes, readings were changed based on specific Masoretic manuscripts (e.g. J.H. Michaelis 1720 and N.H. Snaith 1958, the latter following B.M. Or 2375, 2626, 2628 112 At the same 107The editorial principles are described in Joshua-Judges (1992) and 1-2 Kings (1995). l08 According to humanist principles, editions should be based on late manuscripts that are readily available. --+ Timpanaro, Genesis, 45 109renkower* 1982, xxviii-xxix, xliv-xlv. Thus also id., "Rabbinic Bible," in Dictionary of Biblical Interpretation (ed. J .H. Hayes; Nashville: Abingdon Press, 1999) 2.361-4 (363); Cohen, "Introduction," 11*-15*. In the words of Cashen-Gottstein* 1992, 224: "(w]ith a view to the fact that this is the first eclectic text arranged in the early sixteenth century, it seems amazing that, until the twentieth century, this early humanistic edition served as the basis for all later texts." The differences between RB2 and codex A are recorded in the HUB.--+ Goshen-Gottstein-Talmon, HUB, Ezekiel, xli llOin the 17'h-20'h centuries, a few hundred good, yet non-critical editions have appeared, the most important of which are those of J. Buxtorf (1618), J. Athias(1661), J. Leusden (2"d ed., 1667), D.E. Jablonski (1699), E. van der Hooght (1705), J.H. Michaelis (1720), A. Hahn (1831), E.F.C. Rosenmiiller (1834), M.H. Letteris (1852), the first two editions of BH (Leipzig 1905, 1909-1913), and M. Koren (1962). The dates mentioned refer to the first editions followed by revised editions and subsequent printings. Especially valuable is the collection of variants of Kennicott* 1776-1780 based on the van der Hooght edition (1705).--+ pp. 34, 37 111 See Cashen-Gottstein* 1992, 221-6. 112The Snaith edition did not follow the B.M. manuscripts exactly, as pointed out in detail by Cohen-Freedman, "Snaith."

IA: Prato-Masoretic Texts and the Masoretic Text (!TI.-Group)

73

time, since all these editions reflect the Ben Asher text, the centrally accepted text in Judaism, the recognition developed that any new edition should involve an exact representation of that tradition. - p. 346. Thus S. Baer and F. Delitzsch attempted to reconstruct the Ben Asher text on the basis, among other things, of Ben Asher's grammatical treatise Diqduqqe ha-'f''amim.1 13 - p. 69. The edition of Baer-Delitzsch was based on RB2, corrected according to the editors' principles, particularly with regard to the system of the ga 'yot+. C.D. Ginsburg, while reproducing RB2, tried to adapt that edition to the presumed original form of the Ben Asher text on the basis of his thorough knowledge of the Masorah. Ginsburg's edition included a critical apparatus+ containing variants from manuscripts, printed editions, and the ancient versions. 114 Cassuto's "Jerusalem Bible" intended to reach the same goal by changing details in the edition of Ginsburg on the basis of some readings in codex A that he had consulted in Aleppo. 11 5 Only in later years did the search for the most precise Bible text lead scholars to systematically use a manuscript presumably vocalized by Aaron Ben Asher himself (codex A), or one corrected according to that manuscript (codex L), or codex C. 116 Several editions, some of them critical, are based on single sources or combinations of such sources. • Codex L from 1009 (-plate 13*) formed the base for the third edition of BH (1929-1937, 1951), as well as BHS (1967.:...1977 - plates 27* and 28*) and BHQ (2004-;- plate 29*), in contrast to the first two editions of BH (Leipzig: Hinrichs, 1905, 1909-1913), which were still based on RB2. The two editions by A. Dotan, Adi 1973 and Dotan 2001, are likewise based upon this codex. The great majority of the efectronic e8itions of Hebrew-Aramaic Scripture are also based on this manuscript.- p. 24, bibliography • Codex A is used as the base text for the HUB+ and for Cohen, Miqra 'ot Gedolot 'Haketer.' The editions of Breuer (1977-1982 and 1997 [Breuer, Horev]) 117 113 s. Baer & F. Delitzsch, Textum masoreticum accuratissime expressit, e fontibus Masorae varie illustravit, notis criticis confirmavit (Leipzig: Tauchnitz, 1869-1894). 114 Because of the inclusion of all these sets of data, this edition may be considered a critical edition; however, its text base is not a manuscript, but the RB2 edition that lacked scholarly precision. - p. 349. Ginsburg will be better remembered for his monumental Introduction (1897) that served as a preparation for his editorial activity on HebrewAramaic Scripture.- Blau* 1900 115 This edition was criticized much by P. Kahle, "The New Hebrew Bible," VT 3 (1953) 416-20; M. Medan, "The Jerusalem Bible: Comments on the Text, Vocalization, and Accentuation," Leshonenu 18 (1952-1953) 181-92 (Heb.); Roberts, "Hebrew Bible," 254-55; M.H. Goshen-Gottstein, "The Authenticity of the Aleppo Codex," Textus 1 (1960) 1758 ( 30, n. 35) asserts that Cassuto's edition does not contain readings from codex A; Ofer, "Preparation," 202-3. L. Simon, "The Jerusalem Bible," VT 4 (1954) 109-10 wrote in defense of this edition that it was prepared after Cassuto's death in 1951 according to his instructions. . 116 Penkower* 1999 summarizes these tendencies among editors. 117The Jerusalem Crown (2000) described on the title page as The Bible of the Hebrew University of Jerusalem, follows "the methods of Rabbi Mordechai Breuer" (thus the subtitle of the Companion Volume [ed. M. Glatzer; Jerusalem: Ben Zvi, 2000]), especially his Horev edition. The same subtitle records the basis of this edition as following "the

74

Chapter 2: Textual Witnesses

reflect this codex in orthography, vowels, and accents, although in practice they were based on a number of Tiberian sources (codex L, B.M. Or. 4445, RB2 among others). 118 • The edition of Snaith is based on a combination of the Sephardic manuscripts B.M. Or. 2375,2626,2628 and the "Shem Tov" Bible.119

B. Pre-Samaritan Texts and the Samaritan Pentateuch (w.-Group) A.D. Crown & R. Pummer, A Bibliography of the Samaritans (3'd ed.; ATLA Bibliography Series 51; Lanham, MD: Scarecrow, 2005). M. Baillet, "Les divers etats du Pentateuque Samaritain," RevQ 13 (1988) 531-45; id., "Samaritains," DBSup (Paris: Letouzey & Ane, 1990) Xl.773-1047; Z. Ben-Hayyim, "The Samaritan Vowel-System and Its Graphic Representation," ArOr 22 (1954) 515-30; id., LOT (1957-1977); id., Grammar of Samaritan Hebrew (2000); I. Ben-Zvi, The Book of the Samaritans (Heb.; Jerusalem: Yad Ben-Zvi Press, 1976); Cohen, "Orthography" (1976); A.D. Crown, "Studies in Samaritan Scribal Practices and Manuscript History: III. Columnar Writing and the Samaritan Massorah," BJRL 67 (1984) 349-81; id., The Samaritans (1989); id., "Samaritan Literature and Its Manuscripts," BJRL 76 (1994) 21-49; id., Samaritan Scribes (2001); id., "Samaritan Scribal Habits with Reference to the Masorah and the Dead Sea Scrolls," in Paul, Emanuel, 159-77; F. Dexinger, "Das Garizimgebot im Dekalog der Samaritaner," in Studien zum Pentateuch Walter Kornfeld zum 60 GeburtstaK (ed. G. Braulik; Vienna/Freiburg/ Basel: Herder, 1977) 111-33; E. Eshel, "4QDeutn" (1991); Gesenius, Pent. Sam.; Kartveit, Samaritans; Macuch, Grammatik; J. Margain, "Samaritain (Pentateuque)," DBSup (Paris: Letouzey & Ane, 1990) Xl.762-73; R. Pummer, "The Samaritans and Their Pentateuch," in Pentateuch as Torah, 237-69; J.D. Purvis, The Samaritan Pentateuch and the Origin of the Samaritan Sect (HSM 2; Cambridge, MA: Harvard University Press, 1968); J.-P. Rothschild, "Samaritan Manuscripts," in Crown, The Samaritans, 771-94; Sanderson, Exodus Scroll; Sperber, Grammar, 234-97; Schorch, "Korrekturen" (1994); id., "The Significance of the Samaritan Oral Tradition for the Textual History of the Pentateuch," in Samaritan Researches V (Studies in Judaica 10; Sydney: Mandelbaum Publishing: 2000) 1.03-1.17; id., Vokale (2004); id., "The Latent Masorah of the Samaritans," in Samaritans: Past and Present: Current Studies (SJ 53; ed. M. Mor & F.V. Reiterer; Berlin/ New York: De Gruyter, 2010) 123-32; M. Segal, "The Text of the Hebrew Bible in Light of the Dead Sea Scrolls," Materia giudaica XII

text and Masorah of the Aleppo codex and related manuscripts." The principles of the Jerusalem Crown were explained by Glatzer, Companion Volume, 51-9; Ofer, "Preparation." l18According to the description in vol. II of the edition, Breuer produced an eclectic+ text based on the majority of the sources examined; he notes that his preferred readings always agreed with codex A. According to Breuer, this procedure ensured that he was able to recover the text of that codex, including its lost pages, notably in the Torah, by following the Tiberian manuscripts. Breuer's editions do not contain an apparatus documenting the sources selected for the printed text. However, his later study, The Biblical Text in the Jerusalem Crown Edition and Its Sources in the Masora and Manuscripts (Heb.; Jerusalem: Keren Ha-Masora, 2003) may serve as such an apparatus. Cohen, "Introduction," 61 *-9* criticized the eclectic choices of the Breuer system regarding the ga ·yot. 119 see N.H. Snaith, "The Ben Asher Text," Textus 2 (1962) 8-13. For criticisms, see CohenFreedman, "Snaith."

lB: Pre-Samaritan Texts and the Samaritan Pentateuch (w.-Group)

75

(2007) 5-20; A. Tal, "Samaritan Literature," in Crown, The Samaritans (1989) 413-67; id., "Divergent Traditions of the Samaritan Pentateuch as Reflected by Its Aramaic Targum," journal for the Aramaic Bible 1 (1999) 297-314; J.H. Tigay in id., Models, 53-96; E. Tov, "Rewritten Bible Compositions and Biblical Manuscripts, with Special Attention to the Samaritan Pentateuch," in id., HB, GB, and Qumran (2008) 57-70; id., Consistency (2010); R. Weiss, bmqr' Qerusalem: Rubinstein, 1976) 317-37; M.M. Zahn, Rethinking Rewritten Scripture: Composition and Exegesis in the 4QReworked Pentateuch Manuscripts (STDJ 95; Leiden: Brill, 2011); J. Zsengeller, "Origin or Originality of the Torah? The Historical and Textcritical Value of the Samaritan Pentateuch," in From Qumran to Aleppo, 189-202.

-€fectronic toofs: A module of w. is included in Accordance (representing Tal, Shekhem) and Bible Works (edition of von Gall). For both see§ 3. In Accordance, the differences between lll and m are automatically indicated. Resident morphological analyses+ in Accordance allow for a wide range of comparative searches of w. and m, including vocalization. The module takes into consideration the Samaritan reading tradition as recorded by Ben-Hayyim, LOT (e.g. a noun at the base of '1:::l,:::l in w. Exod 19:9 against the verbal form '""9':T"/l in !H).

Summary. The Samaritan Pentateuch (lll) contains the text of the Torah

written in a special version of the early Hebrew script+ (-+plate 17*), preserved for centuries by the Samaritan community ....... § 1. ll.l contains a few ideological elements that form a thin layer added to an otherwise non-sectarian early text -+ § 4, very similar to so-called pre-Samaritan texts found at Qumran. -+ § 5. Scholars are divided in their opinion on the date of the creation of the Samaritan text. -+ § 2. Often the preSamaritan texts and ll.l are together named the ll.l-group. The preSamaritan texts are typologically older than ll.l, but as they have been preserved fragmentarily, ll.l is discussed first. Since the differences between ll.l and the pre-Samaritan texts are minor, the characterization of the former essentially pertains also to the latter. 1. Background

Beyond the medieval copies of ll.l, scholars are interested in the early (non-sectarian) Qumran manuscripts on which this text is based. Since these texts are known only fragmentarily, our main source of information is the medieval text of ll.l, which is sectarian. ll.l is the Samaritan text of the Torah, 120 written in a special version of the early Hebrew script+ and preserved by the Samaritan community in many copies. This text is purely consonantal, and the reading tradition that developed alongside the text remained oral for a long period. In due course, the Samaritans 120 The Samaritans also possess a version of Joshua that, among other things, contains several readings agreeing with against !)l. This text was published by Gaster, "Buch Josua." See also Ben-Zvi*, 292-322. The status of this version is contested; seep. 315, n. 51. The Samaritans likewise possess a historical work containing material parallel to the biblical books of Joshua-Kings and 2 Chronicles ("Chronicle II") ...... P. Stenhouse, "Samaritan Chronicles," in Crown, The Samaritans, 222-3. For further details, see Baillet* 1990.

76

Chapter 2: Textual Witnesst•s

developed vowel signs, but only rarely were some of these insertedinconsistently-into late manuscripts. It was only in the 20 1h century that parts of the Torah were written with partial (Ben-Hayyim* 2000, o; Schorch* 2004, 27) or full vocalization. - r. Sadaqa (p. 78). The reading tradition has been recorded in scholarly transliteration, especially by Ben-IJayyim* 1957-1977, 2000*. This tradition is also reflected in translations of w. into Aramaic and Arabic. 12 1 The importance of this reading tradition for textual criticism was investigated especially by Schorch* 2000 who concluded (1.16) that, as a rule, w. agrees with m. The differences between the two reading traditions are recorded in TaiFlorentin, Samaritan Version (2010) and Schorch, Vokale (2004). The scribal tradition of w. reflects several features that are similar to those of the Qumran scrolls(- Scr. Prac., Index, 392) and the Masorah of m. The "Samaritan Masorah" pertains to sense divisions similar to the open and closed sections' of m, the fixed written form of certain sections (- pp. 198-200; Crown* 1984), and also, at an earlier stage, to musical directions similar to the Masoretic accentuation'.- Schorch* 20Hl Because the consonantal framework of w. was largely based on preSamaritan texts such as those found at Qumran - § 5, the descriptive name "Samaritan" of w. is somewhat misleading at a textual level. The later reading tradition of w. indeed reflects the Samaritan tradition, but the main content of the text, including its typological characteristics often differing much from !U+ and were already found in the earlier nonsectarian texts. Thus, whereas the proto-Masoretic texts derive from the same circles as the medieval m - pp. 29-31; the pre-Samaritan texts, while agreeing much with w., lack the ideological Samaritan features. 2. Date and Origin R.j. Coggins, Samaritans and Jews: Tilt• Origins of S11maritanism Reconsidt·n·d (Atl

'i1:::l::l

141 Cohen* 1976 showed that in certain words and grammatical categories m and J.U display similar spellings but that the texts behaved differently in the five books of the Torah. 142 4QpaleoExodm and 4QExod-Levf are somewhat fuller than the medieval text of J.U (and m), while 4QNumb, reflecting the spelling of the QSP+, is much fuller. For details, see the text editions mentioned on p. 90. The full spelling of the pre-Samaritan texts does not follow any specific pattern.

90

Chapter 2: Textual Witnesses 1(!:1,•

10 11 13 14 15 17 21 24 26

1(1::!• ii"tD 1tD1::1" m•J1l( ,,on iin1Jo j1£l£ltD iin1"tv "ll1,T .,ii

n1"Jl( ,bn

iin'(tv "ll,T .,,ii

Table 18 shows that w is written with a fuller orthography than m in twelve words, and that in two instances the situation· is reversed. In v 17 m I w ]15l5ltv the two texts have different plene spellings. Only rarely does the medieval text of lU contain very full spellings of the type occurring in the QSP+ and in some pre-Samaritan scrolls.143 Gen 24:41b

m lU

Deut 10:1,3

m lU

•pJ l(•pJ cf. v 41 a m iipJn, w l('pJn C"JtDl(,:l C"J1tD'l(,:l

=

4QRPb (4Q364) C"J1)tD"l(,:l

5. Pre-Samaritan Texts Z. Ben-ljayyim, "Comments on the Use of the Term 'Proto-Samaritan'," in Language Studies V-VI (Heb.; Jerusalem: Hebrew University, 1992) 13-23; Dexinger• 1994; Eshel-Eshel* (2003); Kartveit*, 259-312; Lemmelijn, Plagues Narrative; Tov• 2008; Zahn* 2011. Editions: DJD V (4QRPa [4Q158), 4QTest [4Q175)), IX (4QpaleoExodm), XII (4QExod-Levf; 4QLevd), XIII (4QRPb [4Q364)), XIV (4QNumb). toofs: The pre-Samaritan texts are searchable in the DSS module in Accordance, Bib/eWorks, Logos, and SESB. In Accordance, the differences between these texts and m are automatically indicated. Resident morphological analyses+ allow for a wide range of comparative searches and concordances.

Before the discoveries made at Qumran, scholars conceived of the medieval manuscripts of w. as reflections of an ancient text, whose nature could not be easily determined. However, this situation has changed since the discovery at Qumran of texts that are very close to w. in significant details. These texts probably preceded the creation of w., and they are now called pre-Samaritan on the assumption that one of them was adapted to suit the views of the Samaritans and subsequently served 143See Cohen• 1976. In some details, the morphology reflected in the QSP bears a similarity to w, for which see Kutscher, Language, 566-7; Z. Ben-Hayyim, "mswrt wzyqth lmswrt h/sw11 Sl mgylwt ym wgwn bz"l," Leshonenu 22 (1958) 223--45; M. Mansoor, "Some Linguistic Aspects of the Qumran Texts," JSS 3 (1958) 46-9.

IB: Pre-Samaritan Texts and the Samaritan Pentateuch (w.-Group)

91

as their Scripture. 144 The use of the term pre-Samaritan is thus based on the assumption that while the connections between w. and the preSamaritan texts are exclusive, they reflect different realities. The preSamaritan texts are not Samaritan documents, 145 as they lack the specifically Samaritan readings, 146 but they share with w. its major features. Most pre-Samaritan texts are written in the square script+, while 4QpaleoExodm is written in the paleo-Hebrew script+. The w.-group reflects a popular textual tradition of the Torah that circulated in ancient Israel in the last centuries BCE in addition to the mgroup and many additional texts. It is represented by 4QpaleoExodm, 4QExod-Levf, 4QNumb (secondarily also close to and possibly also 4 4QLevd_1 7 It is also reflected in the biblical manuscripts 4QRPa (4Q158) and 4QRPb (4Q364)+, and is used by 4QTest (4Q175) in the quotation from Exod 20:17 w. = 20:21 m (containing m Deut 5:28-29 [25-26], 18:18-19) and by 4QComm Gen A (4Q252) and the author of Jubilees.l 48 The earliest pre-Samaritan text is 4QExod-Levf, dated to the mid-third century BCE. Some scholars name these texts "Palestinian" 149 since they are not evidenced outside Palestine, but this term implies that no other groups of texts were extant in Palestine. The best-preserved pre-Samaritan text is 4QpaleoExodm (large sections of 44 columns from Exodus 6-37).1 50 -+ plate 2*. Most of the evidence pertaining to editorial additions described in § 4 pertains to this scroll and 4QNumb. Likewise, a verse that is based on the parallel description in Deut 9:20 is added in 4QpaleoExodm and w. after Exod 32:10: [ )1ii1

1.l)::l i1iDO '?'?Eln(' ]1 .11'0iDii'?

i11 [ i1'

'?m

[1n

... of [you] a great nation. [And the LO]RD[ was angry with Aaron,] so much (that He was ready) to destroy him; and Moses interce[de]d for A[aron]-thus also ®MS 58

144 ch. 7 of Kartveit* is therefore appropriately named "The Pentateuch That the Samaritans Chose" (pp. 259-312). Ben-Hayyim* 1992 does not accept the logic behind the term "pre-Samaritan." 145 on the other hand, M. Baillet claimed that several Qumran texts actually witness lll. itself: "Le texte samaritain de I'Exode dans les manuscrits de Qumran," in Hommages a Andre Oupont-Sommer (ed. A. Caquot & M. Philonenko; Paris: Adrien-Maisonneuve, 1971) 363--81. 146 The tenth commandment of lll. (-+ p. 88) is found neither in 4QpaleoExodm (-+ Sanderson, Exodus Scroll, 13, 235; 0]0 IX, 101-2) nor in 4QRPa (4Q158). 147 4QDeutn is not a pre-Samaritan text; see E. Owen, "4QDeutn: A Pre-Samaritan Text?" oso 4 (1997) 162-78. 148 According to J.C. VanderKam, Textual and Historical Studies ill the Book of jubilees (HSM 14; Missoula, MT: Scholars Press, 1977) 137, Jubilees especially reflects readings of lll. and 113, texts that were "at home in Palestine." 149 Cross, OJO XII, 136. 150 P.W. Skehan, E. Ulrich, & J.E. Sanderson, 0]0 IX; J.E. Sanderson, "The Contribution of 4QpaleoExodm to Textual Criticism," RevQ 13 (1988) 547-60.

Chapter 2: Textual Witnesses

92

Furthermore, the following additions in 4QpaleoExodm record the explicit fulfillment of the divine command to Moses and Aaron to warn Pharaoh before each plague.-§ 4a After

7:18 7:29 8:19 9:5 9:19 10:2

(based (based (based (based (based (based

upon upon upon upon upon upon

7:16-18 ) 7:26-29 freconstructed D 8:16b-19 [reconstructed]) 9:3-5 rreconstructed)) 9:16-19) 10:5-6)

(col. II 6-11) (col. III2-4) (col. IV 4-9) (col. V 1-3) (col. V 28-31) (col. VI 27-29) 4QpaleoExodm and w also display a major rearrangement vis-a-vis m.- p. 82 The close relationship between the pre-Samaritan texts and can be viewed in Table 19.

U1

in contrast tom

Table 19

A Pre-Samaritan Text Compared with !lt+ and w.

mNum 20:12, Deut 3:24-27])

4QNumb Xl24-30 (Num 20:12-13b w [=

The underlining in § 1 indicates the extensive differences between 4QNumb and !H+, while§ 2 shows the proximity of 4QNumb to U1 (the only differences pertain to orthography). 151 (1) Compared with

m

[lines 1-23 missing] ,tDK l',Kil l:lh[-m3:24i

.

nK 13u

I

25

iltDl1' ,tDK r,KJ::l1 ,tDK ili?Tnil il:l1'1 . m 25[ ,tDK nK1Til r,Kil nK iTK,K1 I

27

nK

. ,ilil P,'ill

24

. .

tv1p'1 il1il' nK

nK1

li:')01Jti

iTTil

l

nK n,K,iTJ'?

il1il'

..

,tDK

,,, ..

I..

ilnK il1il'

l-11121>1 I..

tDK1J,l ?K j17l1

vucut

1-m 271

iTTiT

iTT iT

I

I..

2il 29 30

hoi/om mar}{ in

(2) Compared with •nm ,tDK l',Kil '

))h r-m 3:24[

.

nN1 iT:l';l

iTTil

tv1p'1 il1il' nK nN

nN n1N,il 1'?

U1

nK " '

24

,tDN ili1N il1il'

!)a

I

25

I

26

151 A similnrly close relation tow is shown in 4QNumb Xlll13-17 (Num 21:1lb 12a-b w ]= m Dcut 2:9, Num 21:12, Dcut 2:1H-19]), and XIII 26-29 (Num 21 :20a-b w (=!It Num 21:20, Dcut 2:24-25]).

IB: Pre-Samaritan Texts and the Samaritan Pentateuch (w.-Group) ;"T::>i1i1:lj::>1 ;"[2'iV170::> ;"liV17' iiVM fiM]:J1 C'OiV:J[ '?M '0 iiVM •

;,pm;,

0

j;"l;"l pi';"T[ i:l17:l iiVM nM,m ];"l:l1c:l;"lf riM;"l nM ;"TMiM, MJ ;"li:l17M '?M ;,:;>'-, :l"i[ ;"liV10 '?M ;"11;"1' ]"iOM[1'1 0

0

1..

(-m 26]

1..

]:iO' i1::>'J['17 MiV1 ;"l.i05:li1 iVM1]i[ ;M i1 )17

vacat

[=!11 27]

93

;"1::>,']

(-m 25]

1

]1J:l'?;"l1 ;"Tii1 1..

i1ii1 i:l,:l ,,17 '7M i:l,]

27 2s

29 30

Even though the pre-Samaritan texts and w. share distinctive typological traits and many small details, they also diverge occasionally. For example, the number of editorial additions (pp. 80-82) differs somewhat in the various sources: 4QpaleoExodm reflects fewer editorial interventions than w., while 4QNumb has more. In addition, individual texts of the w.-group also display unique readings. However, in spite of these differences, the editorial and harmonizing readings common to the above-mentioned texts are exclusive and hence form a distinct group. w. is somewhat removed from the pre-Samaritan texts on account of its ideological corrections, phonological changes, and slightly more defective orthography+. 152 ..... p. 87-90 Little can be said with certainty on the internal relation between the preSamaritan texts. Their agreement in idiosyncratic features would indicate a common text that was subsequently developed in various ways in the different manuscripts. 153 It is difficult to know why the Samaritan community chose a so-called preSamaritan text as the basis for its Scripture in all five Torah books. In all probability, there was no special reason for this choice other than the fact that popular texts such as these must have been current throughout Israel. ..... p. 187. Neither the proto-Masoretic text, usually associated with the Temple circles, nor the text underlying (\j was chosen for this purpose.

The editorial developments visible in the w.-group make it an important source for our understanding of the growth of the books of Hebrew Scripture, in addition to what is known from m and (\j and some Qumran texts ...... ch. 7s21 C. The Biblical Texts Found in the Judean Desert Fitzmyer, Dead Sea Scrolls; F. Garda Martinez & D.W. Parry, A Bibliography of the Fi11ds i11 the Desert of Judah 1970-95 (STDJ 19; Leiden: Brill, 1996); Lange, Ha11dbuch (2009); http://orion.huji .ac.il I resources I bib I bibliosearch.shtml Barthelemy, Critique textuel/e 1992, xcviii--cxvi; Cross, ALQ 3; Cross-Talmon, QHBT; Debe(, "Variant Literary Editions"; T. Elgvin, "Sixty Years of Qumran Research: Implications for 152 Aithough the pre-Samaritan texts are more than one thousand years older than the medieval manuscripts of ill, the somewhat more defective spelling of the medieval copies of ill reflects a typologically earlier stage in the development of the ill-group. Apparently, the medieval copies of ill derive from a copy that is older or less doctored than the copies found at Qumran. 153 Alternatively, scribes independently produced copies of the biblical text reflecting the same editorial-scribal tendencies. However, the high degree of agreement between the pre-Samaritan texts does not support such an assumption.

94

Chapter 2: Textual Witnesses

Biblical Studies," SEA 73 (2008) 7-28; Lange, Handbuch; A. Paul, La Bible avant Ia Bible, Ia grande revelation des manuscripts de Ia mer Marte (Paris: Cerf, 2005); M. Segal, "The Text of the Hebrew Bible in Light of the Dead Sea Scrolls," Materia giudaica XII (2007) 5-20; P.W. Skehan, "Qumran, Litterature de Qumran, A. Textes bibliques," DBSup (Paris: Letouzey et Ane, 1979) IX.805--22; Y. Sussmann, "The History of Ha/akha and the Dead Sea Scrolls: Preliminary Observations on 4QMiqsat Ma'ase haTorah (4QMMT)," Tarbiz 59 (1989-1990) 11-76; Talmon, Qumran; Tov, Samuel; id., "Excerpted and Abbreviated Biblical Texts from Qumran," in Tov, HB, GB, and Qumran (2008a) 27-41; id., "The Biblical Texts from the Judaean Desert: An Overview and Analysis," in Tov, ibid. (2008b) 128-54; id., Scr. Prac.; id., Revised Lists of the Texts from the Judaean Desert (Lei den: Brill, 2010); E. Ulrich, "The Dead Sea Scrolls and the Biblical Text," in Flint-VanderKam, DSS 1.79-100 (1998); id., DSS (1999) 121-47; id., "The Qumran Scrolls and the Biblical Text," in Schiffman, Dead Sea Scrolls (2000) 51-9; Washburn, Catalog.

toofs: The biblical scrolls are searchable tn the DSS module (M.G. Abegg) in Accordance. The differences between the scrolls and m are automatically indicated and the program enables the creation of concordances.

1. Background The main find site of biblical manuscripts in the Judean Desert area is Qumran, some

15 km south of Jericho near the Dead Sea, where

remnants of 930 biblical and non-biblical scrolls, once complete, were found in eleven caves.1 54 The texts deposited in the caves probably had been collected by the Qumranites, 155 but we possess no information regarding the role of these texts, or their possible use in the daily life of the community over a period of more than two hundred years. The term

library is often used for this collection, mainly for the texts found in cave 4, but it would probably be more appropriate to consider the caves as text depositories. It stands to reason that at an earlier stage most of the cave 4 scrolls were housed in a central location in one of the community buildings. The nature of the groups of texts found in individual caves has not been clarified.156 154 some of the Judean Desert texts have already been discussed in sections A and B, dealing with the proto-Masoretic and pre-Samaritan texts. Those sections treated important textual witnesses (m, ut) attested not only in medieval sources but also in the Judean Desert texts. In this section, the complete evidence of the discoveries in the Judean Desert is presented, making a slight overlap with those sections. 155 The Qumran scribes were also involved in the copying of scrolls. Possibly, the room in which this activity took place, the so-called scriptorium, can be identified. -+ Tov, Scr. Prac., 15, n. 33 156The only characteristics of the individual caves seem to be: (1) Cave 7 contains only Greek papyrus fragments (19 items), probably mainly Scripture texts. (2) Most of the texts from cave 6 are Hebrew papyri, including a few that are biblical. These two collections must have derived from a special source, different from that of the main depository of texts in cave 4. (3) The collection of texts from cave 4, by far the largest among the different caves, probably constituted the main collection of the Qumran community, as no compositions (as opposed to copies) were found in caves 1-3 and 5--

IC: The Biblical Texts Found in the Judean Desert

95

Another group of 25 texts was found in various additional places in the Judean Desert. ..... § 2b below While the identification of the community living at Qumran is relevant for an understanding of the non-biblical texts found there, it is of limited importance for the discussion of their Scripture texts since many of the biblical scrolls found at Qumran were apparently taken there from other places. It appears that Qumran was inhabited by Essenes (possibly identical to the Boethusians mentioned in rabbinic literature}, whose halakhic practices may have derived from those of the Sadducees, as suggested by an analysis of 4QMMT.-+ Sussmann*

2. Evidence

a. Qumran. Between 1947 and 1956, fragments of 210-212 biblical scrolls from Qumran (representing 224-226 copies of biblical books) were found in the eleven Qumran caves. 157 - Table 20. Most of the fragments are small, containing no more than one-tenth of a biblical book, while 1Qisa3 contains the complete text of Isaiah. Within the Qumran corpus of some 930 texts, the biblical texts constitute 22% (not counting the tefillin and mezuzot). 158 The number of copies of the individual books shows the differing levels of interest in them: Note the exceptionally numerous copies of Genesis (23-24}, Deuteronomy (32}, Isaiah (21}, and Psalms (36}, reflecting the interest of the Qumran covenanters ...... Table 20

The main depository of texts is cave 4, which contains copies of all the books of the Hebrew Bible with the exception of Esther 159 and possibly 11 that were not matched by cave 4 copies. (4) A large percentage of the identifiable texts from cave 11 reflect the QSP+ or are sectarian.-+ Tov, HB, GB, and Qumran, 421-37 157 Many doubts remain in matters of detail rendering the total numbers approximate only. (1) Do certain groups of fragments now regarded as single scrolls possibly represent more manuscripts (e.g. Mur 1)? Conversely, do e.g. 4QJerb,d,e indeed form three manuscripts (thus DJD XV), and not a single one? (2) Our analysis refers only to the books contained in the traditional canon of the Hebrew-Aramaic Bible since we are uncertain regarding the canonical conceptions of the persons who left these texts behind.-+ E. Ulrich, "Qumran and the Canon of the Old Testament," The Biblical Canons, 57-80; A. Lange, "The Status of the Biblical Texts in the Qumran Corpus and the Canonical Process," in The Bible as Book, 21-30. (3) The figures are based on the views of the scholars who published the texts. If texts that are considered liturgical by some or most scholars have been published as canonical texts, such as 4QPsf, 11QPs 3 (-+ pl. 8*) and llQPsb -+ pp. 320-21, they are included in the figures. (4) Tefillin and mezuzot, although containing segments of Hebrew Scripture, are excluded from the statistics. 158 In the Masada corpus, the biblical texts constitute a larger percentage, 46.6 or 43.75% depending on a calculation of either 15 or 16 literary texts. Within the biblical corpus, a special interest in the Torah is visible at all the Judean Desert sites: 89 texts or 42.5% of the Qumran biblical corpus, and 15 texts or 60.0% at the other sites. 159 By coincidence, so it seems, no fragments of this relatively small book have been preserved. On the other hand, S. Talmon, "Was the Book of Esther Known at Qumran?" DSD 2 (1995) 249--67 (with earlier literature) and J. Ben-Dov, "A Presumed Citation of Esther 3:7 in 4QDb," DSD 6 (1999) 282-4 claimed that Esther was known to the

96

Chapter 2: Textual Witnesses

Chronicles, 160 as well as of some of the so-called Apocrypha and Pseudepigrapha. 161 Although most of the scrolls contain only one biblical book, ten Torah scrolls contain two consecutive books. - Table 20. Likewise, the books of the Minor Prophets were treated as one book and appeared together in one scroll. At the same time, some scrolls contained mere selections: (Deuteronomy 32), 4QExodd, 4QCanfl,b; 4QPs8, 4QPsh, SQPs (all: Psalm 119).- ch. 7s18 Table 20 Biblical Scrolls Found at Qumran (2009)162 Book

Square Script

PaleoHebrew Script

Notes Referring to Additional Books Included

Genesis

20-21

3

Exodus

15

1

Leviticus

10

4

4QGen-Exoda includes Exodus; 4QpaleoGenExod1; 4QRPa: Gen, Exod; 4QRPb: Gen, Exod, Num, Deut; 4QRrc: Gen-Deut 4QExodb includes Gen; 4QExod-Levf; 4QRpd: Exod, Num, Deut 4QLev-Numa includes Numbers

Numbers Deuteronomy Joshua Judges Samuel Kings Isaiah Jeremiah Ezekiel

5 30 2 3

1 2

4QDeuti includes Exod

4

3 21 6 6

covenanters. Talmon suggested that the Qumranites did not accept this book as canonical. 1604QChron (4Ql18) is extremely fragmentary, and G.J. Brooke, "The Books of Chronicles and the Scrolls from Qumran," in Reflection and Refraction: Studies in Biblical Historiography in Honour of A. Graeme Auld (ed. R. Rezetko eta!.; VTSup 113; Leiden: Brill, 2007) 35-48 doubts that this fragment contains a segment of the book of Chronicles. 161 Fragments of Tobit, Jubilees, Levi ar, Tfud ar, TNaph. In the wake of the Qumran finds, the terms "Apocrypha" (non-canonical compositions contained in Jewish-Greek Scripture -+ p. 129) and "Pseudepigrapha" Oewish religious works written between c.200 BCE and c.200 CE not included among the Apocrypha) are now anachronistic. All these compositions are actually exponents of the Second Temple Jewish literature. 162The data are based on Tov* 2010. Ten unpublished fragments are not included in the calculation. Fragments covering more than one biblical book are counted only once.

IC: The Biblical Texts Found in the ]udean Desert Minor Prophets

8-9

Psalms Proverbs Job Canticles Ruth Lamentations Qoheleth Daniel Ezra-Nehemiah Chronicles

36

97

4QXIIb,g contain more than one book

4

3

1

4 4 4 2

8 2 1

Fragments of twelve biblical scrolls written in the paleo-Hebrew script+ have been found at Qumran. 163 -+ Table 20. The writing in this script must have been preserved for the most ancient biblical books, the Torah and Job-note that the latter is traditionally ascribed to Moses. 164 The longest preserved texts written in the paleo-Hebrew script are llQpaleoLeva (-+Freedman-Mathews, Leviticus) and the pre-Samaritan 4QpaleoExodm.-+ pp. 90-93. These texts reflect textual variety -+Ulrich* 1999, 121--47, and accordingly their textual character cannot serve as a key for solving the riddle of the writing in this script. The texts written in the paleo-Hebrew script differ from those written in the square script+ in the scribal characteristics inherent with the writing in that script, the almost complete lack of scribal intervention, and in additional scribal features. It is not impossible that these texts derived from Sadducean circles -+ Scr. Prac., 246-8; in any event, paleographical analysis suggests that these texts do not belong to the earliest group of Qumran scrolls. -+ n. 123 b. Other Sites in the fudean Desert. The medieval consonantal text of m more closely resembles the 25 texts found outside Qumran than the mlike Qumran texts. -+ p. 29

163 1QpaleoLev-Numa; lQpaleoLevb?; 2QpaleoLev; 4QpaleoGen-Exod1, 4QpaleoGenm, 4QpaleoExodm, 4QpaleoDeutr,s, 4QpaleoJobc; 6QpaleoGen, 6QpaleoLev; llQpaleoLeva . ..... Tov, Scr. Prac., 54-6; E. Ulrich, "A Revised Edition of the 1QpaleoLev-Numa and lQpaleoLevb? Fragments," RevQ 22 (2006) 341-7. The identification of the unprovenanced paleo-Hebrew fragment XpaleoDaniel is very unlikely. See E. Puech, "Note additionelle sur le fragment en paleo-Hebreu," RevQ 19 (1999) 449-51. 164 0. b. B. Bat. 14b-15a and also ancient (Jacobite) manuscripts of the Peshitta in which Job follows the Torah.

98

Chapter 2: Textual Witnesses

Table 20a

Biblical Scrolls Found at fudean Desert Sites Other Than Qumran Book

Genesis Exodus Leviticus Numbers Deuteronomy Joshua Judges Isaiah Ezekiel Minor Prophets Psalms Daniel

Square Script

PaleoHebrew Script

4 1 4 4

0 0 0

3 1 3 1 1 1 3

0

0

(2009)165

0

1? (-+ n. 163)

3. Dating G. Doudna, "Dating the Scrolls on the Basis of Radiocarbon Analysis," in FlintVanderKam, Dead Sea Scrolls (1998) 1.430-71; A.].T. Jull et al., "Radiocarbon Dating of Scrolls and Linen Fragments from the Judean Desert," Radiocarbon 37 (1995) 11-19 = "Atiqot 28 (1996) 85-91; E. Tov, "The Sciences and the Reconstruction of the Ancient Scrolls: Possibilities and Impossibilities," in The Dead Sea Scrolls in Context: Integrating the Dead Sea Scrolls in the Study of Ancient Texts, l.Jlnguages, and Cultures (ed. A. Lange et al.; VTSup 140 I 1; Leiden/Boston: Brill, 2011) 3-25.

The first system used for dating scrolls was that of paleography (dating on the basis of an analysis of the handwriting), and this is still our major resource for dating. 166 At the same time, at an early stage in the study of the scrolls, C-14 examinations 16 7 of the leather and papyrus fragments 165The data are based on Tov* 2010, 126-9. Sixteen unpublished fragments are not included in the calculation. Fragments covering different books are counted only once. This table covers the following sites: Wadi Murabba'at, Wadi Sdeir, Nahal Arugot, Nabal Hever, Nabal Hever/Seiyal, Nahal Se'elim, Masada, and unknown sites. 166 For a summary of the paleographical dates given to the scrolls, see B. Webster, DJD XXXIX (2002) 351-446. 167 C-14 analysis is based on a comparison of the animal hides that contained carbon-14 atoms when the animal was alive and tree rings also containing C-14 atoms. The logic behind this comparison is that the number of carbon-14 atoms in the hides decreased at a measurable rate after the death of the animal, when they became C-12 atoms, all to be compared with the tree rings.

/C: Tile Hi/!liml Tt•xls Fo1111d i11 lilt' /lldl'tlll l>t•st•rl

99

bL•canw instrunwntal in determining their dates .... Doudnn*, usually supporting pall•ographical dating.lhH Tlw paleographical datl'S applied to tlw documents rangL' from 2!i0 IICF to hH n: for tlw Qumrnn texts, 1h 9 from 50 lll'F to 30 n: for the Masada texts, nnd from 20 lin to 115 l'F for the texts from Wndi Murnbba'at, Wadi Sdeir, Nahallkwr, and Nahnl SL•'elim. 17° Tlw earliest Qumran biblical fragments(-+ n. 170) postdate the date of writing of thl' latest biblical books by several centuries. However, 4QDan'····, containing portions of the second part of the book, were probably copil•d between 125 and 100 BCI;, not more than sixty years after the completion of the finalL•diting stage of that book. 4. Publication

Most of the biblical texts nrc included in the 0/0 series (for which vol. XXXIX and Tov, Lists servl' as an index): D/ D I (Qutnran cave I); II (Wndi Murabba'at); Ill (minor Qumrnn cnvcs); IV (liQPs·•), IX, XII, XIV-XVII (nil: Qumran caw 4); XXIII (Qumrnn cave 11 ); XXXII (rel'dition of 1 An abbn•viated version of nil these L•ditions is included in Ulrich, RQS (2010). Very few editions were published beyond D/D. .... Freedman-Mathews, Ln,iticus. The ft:{illill and mczuzot were published in vnrious ndditionaleditions ..... n. 195 Nahal !·.lever, and Tlw texts found at Wadi Sdeir, Nahal Murnbba'at Wl'rl' published in D/D II nnd XXXVIII. For the Masadn texts, Sl'l' S. Talmon in Masada VI, 1-149. Sel' also E. Tov, "Recording the Dead Sea Scrolls in the Text Editions of Hebrl'W Scripture," in fiR, CR, 1111d Qumm11 (200H) 221-7 ..... ch. 911 lhHFm l'X,lmpll•, with tlw aid of .1 l'-14 ll'::;l, 1Qisa'1 was d.11l'd to lwlWl'l'n .md IO:IJICI; (pilll'ogr..Jphic,ll dall': BCJ;) -o Jull*, whill' C. Bonani l'l .11., "Radiocarbon Dating oltlw lh•,1d S1•a Scrolls," '1\liqol 20 (1991) 27-:12 . 14 (1992) H4:1-49 sugg1•st 202107 ll( '1·: For comparalivl' lahl1•s n•cmding tlw p.lll'ographic.ll .md l'-14 data, Sl'l' Wl'hsl••r ( n. I hh). 169 Sonw Qumr.m srroll!> carry d,lll'S th.ll .ll'l' l'arlil'r th,m thosl' ol tlw aclLt.ll sl'llll•mt•nl at Qumran, which is Ill'!: until 6H n; according to tlw n•vis1•d chronology of J. M.1gnl•ss, '1'111· 1\rdwo/ogy o((,JIIII/1'1111 111111 1/11• I >t•ttd St'il Sao/Is (( ;r,md Rapids: 2002) Wlwn moving out to tlw dl'Sl'rl, tlw QumJ\1n inh.1bitants look with tlwm scrolls that h.1d hl'l'n wrill1•n .11 .m l'Mlil•r 1wriod. Tlw pl'•lf.. 1wriod for srroll production mincidt•d with tlw JWriod of sl'llll'llll'nl ,ll Qumr,m. -o Tov, "I >iflusion" 170 According lo this .N. l:n•••drn.m, 'Tiw M.1sorl'lic Tl•xland tlw ()llllll\111 Snolls: A Study in Orthogr.1phy," Tn/11.' 2 ( 1%2) H7 102; A. Y.ll'lil•ni, 'Tiw l'al.wogr,,phy of 4(Jjtor'1 : A Comp.lr,llivl' Study," TniiiS ( llJ9()) 211 6H.

100

Chapter 2: Textual Witnesses

5. Texts Written in the Qumran Scribal Practice Cross, ALQ3, 174-7; Freedman-Mathews, Leviticus, 51-82; Kutscher, Language; Martin, Scribal Character, 1.393--402, 11.71{}--11; P. Muchowsk.i, Hebrajski Qumrdnski jako jezyk m6wiony (Poznan: Wydawnistwo Naukowe, 2001); Qimron, DSS (1986) 17-24, 57-64; E.J.C. Tigchelaar, "In Search of the Scribe of 1QS," in Paul, Ema11ue/, 439-52; Tov, Scr. Prac. (2004) 261-73; id., "Scribal Features of Two Qumran Scrolls," forthcoming (2012).

The 210-212 Hebrew-Ararnaic biblical texts found at Qumran do not share any major textual, linguistic, or scribal characteristics. Since they were written in different periods and at different places, they reflect textual and linguistic variety as described .in § 6. However, one group of texts is idiosyncratic and is closely related to the scribes who copied the Qumran sectarian scrolls: Within the Qumran corpus, a group of 167 non-biblical and biblical texts has been isolated as reflecting an idiosyncratic practice, the characteristics of which are visible in in orthography+, morphology+, and scribal features. -+ Tov* 2004. 1 1 These texts do not share a common textual typology. -+ p. 109. This group of texts is closely connected with the Qumran community, since it includes virtually all commonly agreed upon sectarian writings. 172 The texts written in the Qumran scribal practice (QSP) could have been penned anywhere in Palestine, but they were probably written mainly at Qumran. 173 Early scrolls, such as 4QQoha (175-150 BCE), must have been copied elsewhere by similarly oriented scribes as this and a few other texts predate the settlement at Qumran. 174 171 Two similar texts were found at Masada (MasShirShabb [Mas 1k] and MasQumranType Fragment [Mas 1n]). These texts, as well as others, were probably taken to Masada by one of the fleeing Qumran covenanters. 172seven or eight sectarian writings do not display these characteristics.-+ Tov* 2004, 262 173It must be conceded that the term Qumran scribal practice used here may be misleading since the described practice was not invented by the Qumran scribes and may well have been used elsewhere in ancient Israel, but no better term suggests itself. In many aspects, this is a Palestinian scribal system, but it would be misleading to call these texts Palestinian, since the use of such terminology would imply that there are no other Palestinian texts. The name QSP merely indicates that as a scribal system it is known mainly from a large number of Qumran scrolls, without implying that this practice was used only at Qumran. 174The statistical analysis is based on the Qumran corpus containing fragments of 930 texts, from which 150 Aramaic (including 17 Nabatean-Aramaic texts) and 27 Greek texts are excluded, since they display no features comparable to the orthographic and morphological peculiarities recognized for the Hebrew texts. By the same token, at least another 150 items should be excluded due to their fragmentary state. From among the 600 remaining texts, 400--500 are large enough for analysis. Within this group, Tov* 2004, 277-85 lists 167 texts (including biblical texts and 8 tefillin) that presumably reflect the orthographic and morphological features of the QSP (of these 167 texts, some 130 are good candidates, while the remainder are probable candidates). The 28 biblical texts, not all equally convincing, are listed on p. 104.

IC: The Biblical Texts Found in the Judean Desert

101

The QSP is characterized by orthographic+, morphological+, and scribal 175 features.176

a. Orthographic Features Many Qumran texts are characterized by a distinctive orthography+ that has no equal among the known documents from other places, while a few features are reflected in the letters from the period of the Second Jewish Revolt, in Mishna manuscripts-+ Kutscher*, 20, and in the orally transmitted reading tradition of tu, 177 but the evidence known to date does not provide any close parallels to the combined features of the QSP. In addition to the very full orthography, it has some special orthographic features, which occur in conjunction with a series of morphological and scribal features (see below). Cross* describes the orthography of these texts as a "baroque style" and includes the morphological features described in§ b under the heading of orthography.

The orthography of the QSP has been described in various studies, especially in the detailed description of 1Qisaa by Kutscher*, in an analysis of all the Qumran texts by Qimron*, and in Tov* 2004. It is characterized by the addition of many matres lectionis+ whose purpose it is to facilitate the reading ...... pp. 208-18. Below are several examples which should be viewed in conjunction with the texts reproduced in plates 3*-5* and in Table 22 on pp. 105-6 (for statistical details, see Tov* 2004, 337-43): In the orthography of the QSP, I o I and I u I are almost always represented by a waw, including (e.g. liD1n, i11::l, i1iD1o), qame? ('-,1::>, i10::>1n, and qame? (e.g. Because of scribal inconsistency, many words appear in the same text with different spellings, e.g. and in 1Qisaa and in several other texts. Yod represents not only IiI (usually: not short i), but also ?ere: (lQisaa 61:2), (38:1). Unique for certain lexemes is the 175 In the past, criticisms were limited to the arguments based on orthography: Cross, ALQ3 , 174-7; J. Lubbe, "Certain Implications of the Scribal Process of 4QSamc," RevQ 14 (1989-1990) 255--65; J. Cook, "Orthographical Peculiarities in the Dead Sea Biblical Scrolls," ibid., 293-305; Ulrich, DSS, 111; J. Campbell, "Hebrew and Its Study at Qumran," in Hebrew Study from Ezra to Ben-Yehuda (ed. W. Horbury; Edinburgh, T & T Clark, 1999) 38-52 (41); A. Lange, "Kriterien essenischer Texte," Qumran kontrovers: Beitriige zu den Textfunden vom Toten Meer (ed. J. Frey & H. Stegemann; EinblickeErgebnisse-Berichte-Reflexionen aus Tagungen der Katholischen Akademie Schwerte 6; Paderborn: Bonifatius, 2003) 59-69. For a comprehensive review, see E.J.C. Tigchelaar, "Assessing Emanuel Tov's 'Qumran Scribal Practice'," Transmission (2010) 173-207. 176 The idiosyncratic content of Qumran tefillin written in QSP orthography and morphology is distinct from the pericopes prescribed by rabbinic literature and contained in the tefillin written in the m system. This fact provides an external control that supports the assumed existence of tlie QSP, unrelated to the question of whether or not these teftl/in were sectarian.-+ Cohn, Teftllin, 73-5 177 For example, Mur pap Letter from Beit-Mashiko to Yeshua b. Galgula (Mur 42) 2 iD1i ii'Jna;;; Masada 1, inscription 449 nm iii1iDii. See also b. Meg. 11a where iD1i1iDnN (" Ahasuerus") is explained from iD1i ("head").

102

Chapter 2: Textual Witnesses

representation of I il especially in final position mainly in sometimes Less frequent examples include 59:7; 40:5), apparently by also in analogy to etc. in which the 'aleph belongs to the root. He as a mater lectionis for I a I is very frequent at the end of words, such as in qtlth (e.g. i1motv) and the second person masc. sing. suffix, e.g. i1:l::l',o, mlkkh, etc. He in final in 1Qisaa 1:4 (m and i1,1pi1 in position for I e I occurs unexpectedly in 6:4 (m 'Aleph denotes I a I in final position: 34:11 (m i1'',.1.l), 66:8 (m i1'J::l}, and even in medial position: 1:17 (m r:nn•), 30:31 (m i1::l').

The orthography of a complete section is exemplified in Table 22 (p. 105). b. Morphological Features M.G. Abegg, DJD XXXII (2010) 2.25-41; S. Fassberg, "The Preference for Lengthened Forms in Qumran Hebrew," Meghillot 1 (Heb.; Haifa/Jerusalem: Haifa University Press/Bialik Institute: 2003) 227-40; M.H. Goshen-Gottstein, Text and umguage in Bible and Qumran Qerusalem/Tel Aviv: Orient, 1960); S. Morag, "Qumran Hebrew: Some Typological Observations," VT 38 (1988) 148--64; E. Qimron, "The Nature of DSS Hebrew and Its Relation to BH and MH," in Diggers at the Well, 232-44; W.M. Schniedewind, "Linguistic Ideology in Qumran Hebrew," ibid., 235-52; See also the bibliography on p. 100.

The following six features characterize the QSP morphology+, which has a tendency towards lengthened pronominal, verbal, and in one case adverbial forms(-+ Fassberg*; Tov* 2004, App. 9 for statistical details): • Lengthened independent pronouns: hu 'ah, hi 'ah, 'atemah, 'atenah, hemah (the latter form is also found in lll and the later books of m); • Lengthened pronominal suffixes for the second and third persons plural on nouns and prepositions, e.g. bmh, bhmh, mlkmh; • Forms of the Qal imperfect (w)tqtwlw and (w)yq.twlw that serve in m as pausal forms, but occur in these texts as free forms; • Forms of the Qal imperfect o with pronominal suffixes construed as yequ.tlenu (et sim.) instead of the Tiberian forms yiqtelenu (et sim.); 178 • The form qetaltemah for the second person plural in all conjugations; • Lengthened forms of viz., i111o.l 79

Some of these features may have been created by analogy with existing forms, while others may be dialectical. Certain forms are described as archaic by Kutscher*, 52, 434-40; Qimron* 1986, 57; Cross*. Fassberg* stressed the artificial nature of the lengthened forms. 178see I. Yeivin, "The Verbal Forms 1J'?t:11p', in DSS in Comparison to the Babylonian Vocalization," in Bible and Jewish History (ed. B. Uffenheimer; Heb. with Eng. summ.; Tel . Aviv: Tel Aviv University, 1971) 256-76. 179o. P. Wernberg-Moller, "Two Biblical Hebrew Adverbs in the Dialect of the Dead Sea Scrolls," in A Tribute to Geza Vermes, Essays on Jewish and Christian Literature and History (ed. P.R. Davies & R.T. White; JSOTSup 100; Sheffield: JSOT, 1990) 21-35.

IC: The Biblical Texts Found in the fudean Desert

103

c. Scribal Features

The orthographic and morphological features, however inconsistent, allow for a distinction between a group of texts displaying an idiosyncratic practice and texts that do not display these features. These features are accompanied by objective scribal features. - Tov, Scr. Prac., 264-6 • Writing of the divine names, especially the Tetragrammaton+, in paleoHebrew+ characters or Tetrapuncta+ in texts written in the square script. Within the Qumran corpus, this practice is documented mainly in texts written in the Qumran orthography and morphology. Since this practice is based on a certain conception of the sanctity of the divine names, and since the approach of the Qumran community to this issue is known also from other indicators, this practice provides an independent control supporting the hypothesis of a QSP. • Scribal markings. The majority (84) of the 131 Hebrew Qumran texts containing scribal markings of some kind, as listed in Tov* 2004, 279-85 (e.g. the paragraphos+ sign), also reflect the orthographic and morphological features of the QSP. In some groups, this percentage is very high, e.g. for cancellation dots+, parenthesis signs+, the X-sign, and guide dots for drawing horizontal lines. • Frequent use of non-final letters in final position and final letters in nonfinal position. -+ p. 197 • Crossing out of letters and words with a line.-+ p. 204

d. Contextual Adaptations A. Rubinstein, "Notes on the Use of the Tenses in the Variant Readings of the Isaiah Scroll," VT 3 (1953) 92-5; id., "Formal Agreement of Parallel Clauses in the Isaiah Scroll," VT 4 (1954) 316-21.

The scribes of the texts written in the QSP often adapted seemingly irregular forms to the context. This was illustrated by Rubinstein* 1954 with regard to the adaptation of grammatical elements in 1Qisaa to the parallel hemistich+ and the simplification of the tense system (Rubinstein* 1953). These changes reflect a free approach to the biblical text, as exemplified in Table 21 with additional examples (a similar phenomenon in ll.l is exemplified in Table 13 [p. 85]).

Table 21 Contextual Changes in 1Qlsaa Isa 1:23

m 1Qlsaa

Isa 14:30

. . r'J,,,

1ntv :JiTN , .,:;1

... '£l11, 1n1tD ':Ji11N r::::l.,1:;1 ( cf.

,,:Jm

c,,,,o T,tD)

0: 5 D)

(ln',NtD1 ltD,tD :Jl',:J

(I will kill your stock by famine) and it shall slay (the

very last of you). J1,i1N I shall slay (cf. D)

Chapter 2: Textual Witnesses

104

Isa 46:11

m

i1JtDVN CO]N

(i1JN':::lN CO]N '1'1i:::li)

(I have spoken, so I will bring it to pass;) I have designed , so I will complete it. i1JtDVN CO]N

I have designed it, so I will complete it. Isa 51:19

!11

lOnJN '0

1Qlsaa

lOnJ' '0 (cf. 1\J)

e. Consistency and Statistical Analysis The scribes writing in this scribal practice followed certain principles but, at the same time, each scribe also maintained a certain level of independence. This is clear from a comparison of overlapping texts written in this scribal practice, such as 4Qisac//1Qisaa. These divergences are also evident when one compares the two segments written by scribes A (cols. I-XXVII, Isa 1:1-33:24) and B (cols. XXVIII-LIV, Isa 34:1-66:24) of 1Qisaa. 180 The two segments of that scroll must therefore be taken as two separate units. The idiosyncratic spellings and forms recorded above do not all appear in every text. The combined group of features is probably most clearly visible in the following 12 biblical texts and tefillin: 4QNumb, 1QDeuta, 4QDeutk2, 4QDeutm, 4QSamc, 1Qlsaa (especially scribe B), 2QJer, 4QXUC, 4QPhyl A, B, J-K, L-N. At the same time, certain features are absent from some texts that otherwise display most of the idiosyncrasies of the QSP. Thus N':J occurs in most texts belonging to this group, but not in 1Qlsaa scribe AJ81 The following 28 biblical texts may be considered to have been written in the QSP: lQDeuta, lQisaa, 2QExodb (?), 2QNumb (?), 2QDeutc (?), 2QJer, 4QExodb,j(?) 4QNumb, 4QDeuti,kl,k2,m, 4QRPa,b,c (4Q158, 364, 365), 4QSamc, 4Qisac, 4QXUC•e,g, 4QPs 0 (?), 4QLam, 4QQoha, 11QPsa,b,c(?),d(?), and 4QPhyl A, B, G-1, J-K, L-N, 0, P, Q. Virtually all the sectarian compositions written by the Qumran covenanters (such as 1QHa, lQM, lQS, and the pesharim, but see n. 172) also belong to this group. Although there is no characteristic representative of this group, 1Qisaa, 182 18°one notes the preponderance of the short form of the second person sing. masc. suffix in nouns, prepositions, and verbs (l-) in the section written by scribe A, in contrast to the longer form (;'!:l-) in the section written by scribe B. Similar differences are visible in the spelling of (N)':l and ;'!(1):l.- Tov* 2012 18 1 m, in sharp contrast to the aforementioned Qumran texts, does not reflect the features described here as characteristic of the QSP. Most of the full spellings do not occur in m, not even (with the exception of Jer 33:8), while occurs only rarely. However, 11 sporadic "typical Qumran" forms do occur in m: ;"TlnN (Gen 31:6; Ezek 13:11, 20; 34:17), (Ezek 40:16), (Ezek 23:48, 49), (Isa 34:16), (Amos 4:3), ;"T:li' (Exod 13:16), Oer 29:25). Whatever explanation is given for these sporadic exceptions, m does not reflect most of the special forms of the QSP. 182 Aramaic influence and weakening of the gutturals is more recognizable in this scroll than in other texts.- Table 22, "Language"; Kutscher, Lanxuaxe, 91-5, 505-11

IC: The Biblical Texts Found in the Judean Desert

105

which contains the longest Qumran text of a biblical book, is a good example of the QSP.

6. Variants in the Qumran Scrolls

The Qumran texts, as well as differing from one another, relate to m, \1J, m., and the other texts in a ramified system of agreements and disagreements. The more significant deviations from m in the Qumran texts are described inch. 7s1, 4, 9-13, 18, 22, 23, as well as below. Several groups of texts are recognized on the basis of variants in the Qumran texts ...... § 7. Among these groups, them-like and pre-Samaritan texts have been described in detail in sections Al and B4. The tables adduced in this section exemplify some of the more characteristic types of variants found in the Qumran texts, without exhausting the evidence. The tables exemplify, among other things, the readings found in texts written in the QSP+, represented in Table 22 by lQisaa. 4QSama, described in Table 23, presents features of groups 3 and 4. Some texts exemplified by lQisaa in Table 22 display a great number of differences in orthography+ and morphology+, whereas the relation is reversed in the texts exemplified by 4QSama in Table 23: differences in morphology and orthography are few, in contrast to the large number of other types of differences in both major and minor details. Most of the variants listed for 4QSama in Table 23 are substantial.

Table 22

Classified Differences between m and 1Qlsaa in Isa 1:1-8 1. Orthography+ m 2 3

1Qlsaa

':l !'\'?

1i1.J1p (2x) 1'\1'?

4

111ll '?1:l

5

·'?I:J'? '?:l1 6

!'\'?

(2x) 1'\'?1 7

8

•'?1n'? '?1:li iDI'\1i 1'\1'? 1iD:::l1n (2x) 1'\1'?1

\

c•'?jl'\ i1ril'\ i1:lQ:l

I(

i1n11'\ i1:10:l

106

Chapter 2: Textual Witnesses 2. Language

1

m

lQistf

1i1' lliV'

1i1'lltv'

'l:l':::l

'l:l,,:::l

1i1'lll

i1'!1ll

1i1'pm•

2 7

n::>::li11:l::l

i1'pm i11:li11 n::>IC::ll:l::l

8

i1inm

nin:l1

Ci11

Omission of 'ayin indicates the weakening of laryngeals and pharyngeals. Supralinear+ letter added. Supralinear letter; addition of waw is probably due to Aramaic influence. Short theophoric names are more frequent in the Second Temple period. Supralinear letter; see the previous item. See§ 5b above. Supralinear letter; the addition of an 'aleph is probably influenced by the Aramaic root l::lN: (= l::li1 Heb.) The variant probably reflects an Aramaic verbal form for the third person fern. sing.

Notes 1. Supralinear+ letters concern elements that were not included in the first writing ...... p. 204 2. The linguistic variants described above are typical of the scrolls written in the osr+, while the variants included in the next category are not. 3. Other Differences 2

fiN:ii

3

'l:lll

5

·n

7

'l:lll1

different word patterns ,;•'?.v 11:ll:ltv1 cf. Lev 26:32 ii11

i1J1,1:l::l

8

Table 23

Differences between m and 4QSama in 1 Sam 1:22-28 There are only three differences in orthography (vv 23 m '-,:;,).All other differences relate to matters of content.

m 22

4QSama iiVN: 1ll

1ll [CiV

CiV c'-,,v

+

1i:::l1 23 24 i11:lll m'-,vm

c'-,1v 1ll

? i11i1']

i1''iV tv'-,(DI:) ip:::l[ j:::l i::l:::l] en'-,, i11i1'

cf. \B 5 0:: D cf. \B 5 0:: D

1i1'n[ n:11] [1"n] 'l:l' '-,,:;, T::ll:l 1mM '-,vn,

i1w;i'-,w;i

Notes

1i11N:':::l'1 Cl:lll] i!JJ,i7

itDN:[::>] rbtii[ 1i1':::lN: i11i1''-, i11:l'l:l' C'l:l'l:l i1tDll']

= \B; ..... p. 256 cf. \B; ..... p. 277 =IB cf. 11>; ..... p. 236 =IB =IB

24 m 1'-,tv, 28 m

IC: The Biblical Texts Found in the Judean Desert

107

[ 1.!);,; nt\

25 28

11::lniD'1 CiD 1nniD', ii1ii'"

t:lniD['1]

[mii'" 1]nniDm ciD 1ii[:nllm]

=

®

cf. ® 2:11 - p. 255

7. Classification of the Scrolls According to Textual Character Debe!, "Variant Literary Editions"; R.S. Hendel, "Assessing the Text-Critical Theories of the Hebrew Bible after Qumran," The Oxford Handbook of the Dead Sea Scrolls (ed. T.H. Lim & J.J. Collins; Oxford: Oxford University Press, 2010) 281-302; Kreuzer, "Text," 132-8; Lange, "They Confirmed"; id., Handbuch, 1-32, 143-57; id., "Plurality" (2010) 47-58; E. Ulrich, "Methodological Reflections on Determining Scriptural Status in First Century Judaism," in Grossman, Rediscovering (2010) 145-61 (15fr60).

Due to the absence of objective criteria for classifying the Qumran scrolls, 183 they are classified here according to their textual character. Only 121 of the 210-212 biblical scrolls can be classified in this way. 184 In this classification, an attempt is made to characterize and analyze three groups and a cluster of texts, three of which were unknown before the Qumran discoveries (2, 3, 4).1 85 The QSP+ group is listed as an appendix to the classification, since these texts have different textual backgrounds. 183 I. Young, "The Biblical Scrolls from Qumran and the Masoretic Text," in Feasts and Fasts, A Festschrift in Honour of Alan David Crown (ed. M. Dacy et a!.; Mandelbaum Studies in Judaica 11; Sydney: University of Sydney, 2005) 81-139 attempted to record the variants by objective criteria. Young calculated the number of variants (deviations from m) relative to the number of words in the scrolls excluding orthographic variants, while not differentiating between small details and major content variations. 184 The remaining texts are too fragmentary for textual analysis. In this calculation, the following principles are employed, which show that the results are approximate rather than precise: (1) A few questionable attributions to textual groups are included in the statistics. (2) In accord with statistical probability, texts that are equally close to m and tu in the Torah and tom and 1\3 in the other books are counted as m, the most common text group. (3) Texts written according to the QSP are not included in the statistics; they are counted in other groups in accord with their textual affiliation. (4) Since the texts of the tu-group are not evidenced for books other than the Torah, statistics for the Torah are separated from those of the other books. All statistics, based on the data in Scr. Prac., 279-85, 332-5 with slight changes, differ slightly from the previous formulation in Tov, HB, GB, and Qumran (2008) 145. Lange* 2010 presents a different statistical analysis, in which the m-group features less prominently in the Torah, and the non-aligned texts more so. 185 At a second stage, we describe the distribution of these groups in the individual Bible books. - Tov* 2008, 151-3. Even with the limitation of the haphazard preservation of the evidence, it is relevant to examine, for example, the preservation of them-like scrolls in each of the Scripture books. For some criticisms of the attempt to recognize four, at an earlier stage five, groups of scrolls, see Davila (p. 158, n. 7); Segal* 2007; Ulrich* (2010); Hendel* 2010, 291-5; S.W. Crawford, "Understanding the Textual History of the

108

Chapter 2: Textual Witnesses

In the 46 Torah texts that are sufficiently extensive for analysis, 22 (48%) are m-like (or, in a few cases, are equally close to m and ill), 5 exclusively reflect ill (11 %), one reflects® (2%), and 18 are non-aligned+ (39%). In the remainder of Hebrew-Aramaic Scripture, in the 75 texts that are sufficiently extensive for analysis, 33 texts (44%) are m-like (or, in a few cases, are equally close tom and®), 5 reflect® (7%), and 37 form a cluster of non-aligned texts (49%). Thus, the overall preponderance of mlike and non-aligned texts in the Qumran corpus is evident; in the Torah, more texts are m-like, while in the other books non-aligned+ texts prevail, with them-like group coming a close second. -+ p. 188. These percentages reflect the reality of the texts assembled and copied by the Qumran community, and differ much from the reality of the other Judean Desert sites. At the latter sites, all the texts reflect only the medieval m, both those from the time of the settlement at Qumran (found at Masada) and those written during the Bar-Kochba revolt (the other sites).-+ p. 29 (1) m-like Texts 186

The m-like texts constitute the largest group within the Qumran biblical corpus.-+ p. 31; Tov* 2004, 332-5. These texts are exemplified by 1Qisab in Tables 1 and 2. The scribes who copied these texts from protoMasoretic texts such as those found at other Judean Desert sites took certain liberties and were less precise than the scrolls from which they copied. Yet, the scrolls are very close to m and during the first two generations of Qumran research they, too, were named proto-Masoretic. (2) Pre-Samaritan Texts

The pre-Samaritan Qumran texts 4QpaleoExodm, 4QExod-Levf, 4QNumb, 4QRPa (4Q158), and 4QRPb (4Q364) reflect the characteristic features of ill, with the exception of its ideological readings, but they occasionally deviate from it. -+ pp. 90-93. The ill-group was probably popular in Palestine. (3) Texts Close to the Presumed Hebrew Source of® Lange, Handbuch (2009) 122-38; E. Tov, "The Qumran Hebrew Texts and the Septuagint: An Overview," in Die Septuaginta: Entstehung, Sprache, Geschichte (ed. S. Kreuzer et al.; WUNT; Mohr Siebeck; Ttibingen, 2011), forthcoming. Hebrew Bible: A New Proposal," in The Hebrew Bible in Light of the Dead Sea Scrolls (ed. N. David et al.; FRLANT; Gottingen: Vandenhoeck & Ruprecht, forthcoming). l86Lange, Handbuch, 16 and passim uses the term "semimasoretic texts."

IC: The Biblical Texts Found in the fudean Desert

109

Although no identical or almost identical text to the presumed Hebrew source of ® has been found at Qumran, a few texts are very close to that translation:18? 4QJerb,d bear a strong resemblance to ® in characteristic details. -+ ch. 7Bl. Similarly close to®, though not to the same degree, are 4QDeutq (-+ ch. 4, Table 8 [pp. 249-50]), 4QSam 3 (close to ® and ®Luc; -+group 4), 4QSamb, and 11QPs3 Psalm 151.188 (4) A Cluster of Non-Aligned Texts Many Qumran texts are not exclusively close to m, ®, or ll.l and are therefore considered non-aligned. The employment of this term implies that the texts under consideration follow an inconsistent pattern of agreements and disagreements with m, ®, and ll.l and that they also contain readings not known from other sources. 189 This cluster of texts does not form a group like the others since the latter share common features, while the non-aligned texts differ from one another. The texts that are most manifestly non-aligned are those that contain (groups of) readings that diverge significantly from the other texts, such as 4QRpc--e (4Q365-367) and 4QJosh 3 • -+ ch. 7s13, 23.190 A special sub-group of nonaligned texts are scrolls written for a specific purpose, viz., "excerpted" Scripture texts, such as 4QExodd, 4QDeutkl,q, and 4QCant3,b and liturgical texts, including most of the Psalms texts from caves 4 and 11. -+ ch. 7s18.1 91 The non-aligned character of some texts may be misleading when the few preserved readings display a number of insignificant agreements and disagreements with the other sources, as in 4QDeutc,h. The number of these texts is probably inflated due to our limited knowledge of the ancient scrolls. As an appendix to the classification, we mention the scrolls written in the QSP that reflect different textual backgrounds, and not one common typology. -+ § 5. Thus, 1Qisa 3 could have been copied from them-like lQisab or a similar text and 4QNumb could have been copied from a text like tu, but these assumptions

187 There is insufficient evidence for speculating on the internal relationship between the texts that are close to ®. 188 on all these texts, see Tov* 2011. 4QNumb is not included in this group since it is, first and foremost, close to l.U. Likewise, occasional agreements with \IJ and mere statistical agreement in insignificant details are not included in the statistics. 189 For the data on the texts, see Tov* 2004, 332-5. 190 4QSam 3 holds a special position in this regard, since it is closely related to the Vorlage of ®, while also reflecting independent features. 191 In our view, these scrolls are not biblical scrolls in the usual sense of the word, and therefore the number of Scripture scrolls should be decreased by about forty. However, our analysis follows the classification of the scrolls determined by the scholars publishing them. When abandoning that classification, the number of non-aligned texts is much smaller. Furthermore, the large number of statistically independent texts results from the nature of the transmission, causing every manuscript to differ from the others.

110

Chapter 2: Textual Witnesses

cannot be verified. The 28 texts written in the QSP (not all equally convincing), often described as typical Qumran texts, comprise a sizable group among the 121 reasonably large biblical texts (23%), but this percentage is not included in the overall statistical analysis, since these texts have a different textual background. If, indeed, a large segment of the Qumran scrolls were penned by Qumran scribes, it is remarkable that they contain no sectarian readings. 192 If all the aforementioned groups and texts were copied at Qumran, or if, as we believe, only some were copied there while others were brought from elsewhere, the coexistence of the different categories of texts in the Qumran caves is noteworthy. The fact that these different texts were found in the same caves reflects a textual plurality at Qumran and in the country as a whole between the 3'd century BCE and the 1'1 century CE. -+ pp. 186-7. While no solid conclusions can be drawn about the approach of the Qumranites to the biblical text, it is safe to say that they paid no special attention to textual differences such as those described here. For one thing, no specific text was preferred in their sectarian writings. -+Lange, Handbuch, 158-65. At the same time, the great number of mlike texts found at Qumran probably reflects the dominance of m in several sectors in Israel. These were the texts that the covenanters took with them when they moved from their settlements to the Judean Desert.

8. Contribution of the Judean Desert Scrolls to Biblical Research

The Judean Desert texts contribute much to our knowledge of the biblical text at the time of the Second Temple, a period for which there was hardly any Hebrew evidence before 1947. Until that time, scholars based their analyses mainly on manuscripts from the Middle Ages. The Qumran evidence enriches our knowledge in the following areas. • Readings not known previously help us to better understand many details in the biblical text, often pertaining to matters of substance. -+ chapters 4, 6, 7 • The textual variety reflected in the four groups of texts described in § 7 provides a good overview of the condition of the biblical text in the Second Temple period ...... ch. 3c3 • The scrolls provide much background information on the technical aspects of the copying of biblical texts and their transmission in the Second Temple period. -+ ch. 4A • The reliability of the reconstruction of the Vorlage+ of the ancient translations, especially l\J, is supported much by the Qumran texts. 193 -+ pp. 122-7 192 Thus G.J. Brooke, "E Pluribus Unum"; id., "Deuteronomy 5-6 in the Phylacteries from Qumran Cave 4," in Paul, Emanuel, 57-70; E. Ulrich, "The Absence of 'Sectarian Variants' in the Jewish Scriptural Scrolls Found at Qumran," in The Bible as Book, 179-95. On the other hand, the following scholars believe that the scrolls do include sectarian readings: van der Kooij, Textzeugen, 95-6; P. Pulikottil, Transmission of Biblical Texts in Qumran: The Case of the Large Isaiah Scroll 1Qlsaa OSOTSup 34; Sheffield: Sheffield Academic Press, 2001). 193This claim was already made by G.R. Driver, "Hebrew Scrolls," JTS n.s. 2 (1951) 17-30 (25-7).

IC: The Biblical Texts Found in the fudean Desert

111

The texts from the other sites in the Judean Desert demonstrate that m was the sole text used at those locations. This conclusion is of major importance for understanding the realm of influence of that text. -+ p. 29 D. Additional Witnesses Several texts were in existence in ancient Israel beyond those known today. These sources include texts that are not biblical in the usual sense of the word (thus§§ 1-3, 5). 1. Silver Rolls from Ketef Hinnom G. Barkay, "The Priestly Benediction on the Ketef Hinnom Plaques," Cathedra 52 (1989) 37-

76 (Heb.).

Two minute silver rolls (amulets?), dating to the 71h or 6th century BCE, contain the priestly blessing (Num 6:24-26) in a formulation that differs in some details from m+, plate 1*. Roll II lacks the words 1m'1, "He will deal graciously with you" (v 25) and ';, "The LORD will bestow his favor upon you" (v 26). Since these documents are not biblical texts, their contribution to textual criticism is limited. 2. The Nash Papyrus W.F. Albright, "A Biblical Fragment from the Maccabaean Age: The Nash Papyrus," JBL 56 (1937) 145-76; S.A. Cook, "A Pre-Massoretic Biblical Papyrus," Proceedings of the Society of Biblical Archaeology 25 (1903) 34--56; E. Eshel, "4QDeutn" (1991); N. Peters, Die iilteste Abschrift der zehn Gebote, der Papyrus Nash (Freiburg im Breisgau: Herder, 1905).

The so-called Nash papyrus, dating to the 1st or 2nd century BCE, and discovered in Egypt in 1902, contains the Decalogue (Exod 20:2-17 = Deut 5:6-21) followed by the introductory formula Deut 4:45 and the shema' pericope (Deut 6:4-5 are preserved). The Scripture verses of Deuteronomy are thus presented in a different order, while the Decalogue itself contains a mixed formulation of the text of Deuteronomy 5 and Exodus 20. In this composite text, the argument for the Sabbath commandment contains Exod 20:11 rather than Deut 5:14-15. However, details in the text of that commandment are close to Deuteronomy texts such as 4QDeutn, 4QMez A, 4QPhyl G, and 8QPhyl -+ Eshel* 1991, in all of which the Exodus pericope replaces the text of Deuteronomy or is added to it. The orthography of the Nash papyrus is fuller than that of m. scrolls were discovered in 1979 in the excavations at Ketef Hinnom in Jerusalem. Their spelling is like that of m with the exception of the pronominal suffix ending with a he in nouns: miskaboh, i1'?Nj goaloh (?).

112

Chapter 2: Textual Witnesses

Apparently, this composite text reflects a liturgical text. Its content may be compared with several tefillin and mezuzot from Qumran. - § 3 3. Tefillin and Mezuzot from the Judean Desert M. Baillet, DJD III; D. Barthelemy, DJD I; G.J. Brooke, "Deuteronomy 5-6 in the Phylacteries from Qumran Cave 4," in Paul, "Emanuel," 57-70; Cohn, Tefil/in; K.G. Kuhn, "Phylakterien aus Hohle 4 von Qumran," Abhandlungen der Heidelberger Akademie der Wissenschaften, Phil.Hist. Kl. 1957, 5-31; J.T. Milik, DJD II, III, VI; M. Morgenstern & M. Segal, DJD XXXVIII, 183-91; D. Nakman, "The Contents and Order of the Biblical Sections in the Tefillin from Qumran and Rabbinic Halakhah: Similarity, Difference, and Some Historical Conclusions," Cathedra 112 (2004) 19-44 (Heb.); J.H. Tigay, "tpylyn," EncBib Oerusalem: Bialik Institute, 1982) 8.883-95; Y. Yadin, Tefillin (Phylacteries) from Qumran (XQ Phyl 1-4) Oerusalem: IES I Shrine of the Book, 1969).

Many fragments of biblical texts contained in mezuzot, head-tefillin, and arm-tefillin include parts of Exodus 12-13 and Deuteronomy 5-6, 10-11, 32 (-plate 10*), partly reflecting the instructions of the Rabbis and partly containing additional segments. 195- n. 176; pp. 218-9; Cohn*. Details in the biblical texts contained in these tefillin and mezuzot often differ from m, 196 while sometimes agreeing with other ancient witnesses, among them several Qumran scrolls.- ch. 7s12 (4QPhyl A, B, J) 4. The Severus Scroll and R. Meir's Torah Habermann, Ketav, 166-75; D.S. Loewinger, "The Torah Scroll That Was Hidden in the Synagogue of Severus in Rome and Its Relation to the Isaiah Scrolls from the Judean Desert and 'R. Meir's Torah'," Beth Mikra 15 (1970) 237-63 (Heb.); J.P. Siegel, The Severus Scroll and 1Qlsa (SBLMasS 2; Missoula, MT: Scholars Press, 1975).

Rabbinic literature preserves various pieces of information on biblical scrolls whose text differed from m. - § 6. The largest number of such testimonies refers to a Torah scroll that Titus brought to Rome as booty after the destruction of the Temple 197 or, according to a different tradition (Gen. Rabbnti) was brought to Rome by the exiles themselves. At a later period, this scroll was given by Alexander Severus (reigned 222235 CE) to a synagogue that was being built with his permission. In rabbinic literature several individual words are quoted from this Torah tefillin and mezuzot from the 2nd and 1" centuries BCE until the 1'' and 2nd centuries were discovered in the Judean Desert, mainly at Qumran and also in Wadi Murabba'at and Nahal Se'elim (publication: Baillet*, Barthelemy*, Kuhn*, Milik*, Morgenstern* & Segal*, Yadin*). 19 6some texts were possibly written from memory, as permitted by b. Meg. 18b: )'':l!:ln :m::J:-T 1a l(';li.J mma1, "Tefillin and mezuzot may be written out without a written source ." 197According to Josephus, Bell. Jud. VII 150, 162, "a copy of the Law" was removed from the Temple by Titus and taken to Rome. CE

ID: Additional Witnesses

113

scroll, while other quotations, probably from the same source, are attributed to "R. Meir's Torah," since the Torah scroll from the synagogue of Severus was apparently known to R. Meir, a scribe (2nd century cE).l98 Although the original quotations from the Severus Scroll have often been corrupted in the rabbinic sources, they can usually be reconstructed with some degree of probability, as was done by Siegel*. From the scant information known about the contents of the Severus Scroll, it appears that its characteristic features are the weakening of the gutturals(-+ Table 17 re w. [p. 89] and Table 22 re some Qumran texts [pp. 105-6]), the writing of non-final letters in final position -+ p. 197, and the interchange+ of similar letters, as exemplified in Table 24. Thirty-three readings from the scroll are known, but from the evidence preserved in the quotations, it is sometimes difficult to determine the precise difference between this Torah text and the other texts. Loewinger* and Siegel* emphasize the typological resemblance between the readings of this scroll and 1Qisa 3 , both of which are characterized by an imprecise textual transmission displaying a free approach. -+ p. 184. In view of these characteristics, it is very unlikely that this scroll, if indeed it is the one that was brought by Titus from the Temple, was the Temple scroll.

Table 24 Select Differences between mand the Severus Scroll Gen 1:31

Gen 3:21 Sev. Gen 25:33

::J1t:l

m1:l::l

Sev. Gen 27:2

!lH

Gen 27:27

m+

Sev. Sev. Gen 36:10

very good (sources: MS Paris, Farhi Bible) as it were: death is good 11.V mm:l ( = all other ancient texts) garments of skins mm:l (Gen. Rabbati) as it were: garments of light :nt:l

Sev.

m Sev.

he sold his birthright (Gen. Rabbati, MS Paris, Farhi Bible) he sold his sword (?) l:l1" the day of my death the-day-of-my-death (MS Paris, Farhi Bible) i11tl:' field i110 (MS Paris, Farhi Bible)= 4QPhyl G Deut 5:21 i11l1 )::J the son of Adah the-son-of-Adah (MS Paris, Farhi Bible)

198 The main sources quoting from the Severus Scroll are Gen. Rab., Gen. Rabbati of Moses ha-Darshan (a collection of midrashim from the 11th century), the Far}:li Bible (14'h century), and the MS Hebr. 31, Fol. 399, Bib!. Nat., Paris, all of which are described by Siegel* and Loewinger*.

114

Chapter 2: Textual Witnesses

5. Non-Biblical Sources: Quotations and "Rewritten Scripture" Texts G.J. Brooke, "New Perspectives on the Bible and Its Interpretation in the Dead Sea Scrolls," in The Dynamics of Language and Exegesis at Qumran (ed. D. Dimant & R. Kratz; Tiibingen: Mohr Siebeck, 2009) 19-37; A. Lange, "The Status of the Biblical Texts"(-+ p. 95, n. 157); id., Handbuch (2009) 158-68; Tov, HB, GB, and Qumran (2008) 27-42, 136-9.

The Bible text as quoted, excerpted, and rewritten in antiquity evidences a textual reality that should be taken into consideration together with direct sources such as described elsewhere in this chapter. Quotations in rabbinic literature are identical to the proto-Masoretic text-+ p. 33, n. 24, while the Qumran non-biblical scrolls reflect many early variants. Several non-biblical Qumran compositions; both sectarian and non-sectarian, freely quote from and allude to passages in the Bible. For example, the sectarian Qumran Hodayot and Rules, as well as non-Qumranic compositions such as 4QNon-Canonical Psalms A-B (4Q380-381) are replete with biblical quotations and allusions. Most of these quotations are free, involving changes to the biblical text, which therefore must be used carefully in a text-critical analysis. In some cases, variants are reflected in these texts, such as 4QComm Gen A (4Q252) I 2 ,,,, ("shall dwell") form )11' ("shall abide [?]") in Gen 6:3 and ibid. V 3 c•'-,nii = lU 1•'-,n for m in Gen 49:10. -+ DJD XXII, 197, 205. So-called "rewritten Scripture" compositions such as 4-llQTemple and Jubilees also reflect many variants.-+ pp. 189-90. The textual background of several compositions has been studied, but no specific biblical text or text group is reflected in these texts. -+Lange* 2002,27 and id., 2009. For a discussion and bibliography, see Tov* 2008. Different views have been voiced regarding the text-critical value of the biblical text quoted by the pesharim. A positive position was taken by the editors of textual editions that incorporated readings from these pesharim in their textual apparatuses: BHS+ for 1QpHab, HUB+ for the pesharim on Isaiah, and the Biblia Qumranica for the Minor Prophets. On the other hand, some scholars cautioned that many so-called deviations from m in the pesharim and commentaries were due to contextual exegesis.1 99 However, although such is found in the pesharim, including very few cases of sectarian exegesis, 20 many, if not most, deviations in the lemmas in these pesharim probably reflect variants found in the biblical manuscripts used by the commentator.

Excerpted and abbreviated biblical texts contribute to textual criticism, but not to the literary analysis of the biblical books.-+ pp. 320-21 199E.g., G. Molin, "Der Habakkukkomentar von 'En Fesha in der alttestamentlichen Wissenschaft," TZ 8 (1952) 340-57; G.]. Brooke, "The Biblical Texts in the Qumran Commentaries: Scribal Errors or Exegetical Variants?" in Early jewish and Christian Exegesis: Studies in Memory of William Hugh Brownlee (ed. C.A. Evans & W.F. Stinespring; Atlanta: Scholars Press, 1987) 85-100 with references to earlier studies; id., "Some Remarks on 4Q252 and the Text of Genesis," Textus 19 (1998) 1-25. most clear-cut examples are 1QpHab VIII 3 )1;"1 (m Hab 2:5]";"1); 1QpHab XI 3 C;"!'ill10 (m Hab 2:15 C;"!'i1llr.l). -+ W.H. Brownlee, The Text of Habakkuk in the Ancient Commentary from Qumran OBL Monograph Series XI; Philadelphia, 1959) 113-8

10: Additional Witnesses

115

6. Texts That Have Been Lost Ginsburg, Introduction, 430-37; H.L. Strack, Prolegomena critica in Vetus Testamentum hebraicum (Leipzig: Hinrichs, 1873) 14-29.

The texts known to us comprise only a certain percentage of the Hebrew Scripture scrolls that circulated in antiquity. In the 71h or 5th century BCE, the number of such texts may have been minimal, but in the last few centuries BCE many texts must have circulated in ancient Israel beyond those found in the Judean Desert. -+ Tov, "Diffusion." Some such lost texts, about which a few details are known from medieval works, have been reviewed by Ginsburg*: Sefer ("codex") Hilleli, Sefer Zanbuqi, Sefer Yerushalmi, Sefer Sefer Sinai, and Sefer Babli. II. THE ANCIENT TRANSLATIONS

A. Use of the Ancient Translations in Textual Criticism Aejmelaeus, Trail; Barr, Comparative Philology, 238-72; id., "The Typology of Literalism in Ancient Biblical Translations," MSU 15 (NAWG I, Phil.-Hist. Kl. 1979) 279-325; S.P. Brock, "Bibeli.ibersetzungen, I," TRE (Berlin/New York: de Gruyter, 1980) Vl.161 ff.; id., "Translating the Old Testament," in It Is Written: Scripture Citing Scripture: Essays in Honour of B. Lindars, SFF (ed. D.A. Carson & H.G.M. Williamson; Cambridge: Cambridge University Press, 1988) 87-98; S.R. Driver, Samuel, xxxiii-xxxix; N. Fernandez Marcos, "The Use of the Septuagint in the Criticism of the Hebrew Bible," Sefarad 47 (1987) 59-72; M.H. Goshen-Gottstein, "Theory and Practice of Textual Criticism: The Text-Critical Use of the Septuagint," Textus 3 (1963) 130-58; M.L. Margolis, "Complete Induction for the Identification of the Vocabulary in the Greek Versions of the Old Testament with Its Semitic Equivalents: Its Necessity and the Means of Obtaining It," JAOS 30 (1910) 301-12; Mulder, Mikra; Hebrew Bible/Old Testament: The History of Its Interpretation, I (ed. M. Saeb0; Gi:ittingen: Vandenhoeck & Ruprecht, 1996); I. Soisalon-Soininen, Studien zur Septuaginta-Syntax (Suomalaisen Tiedeakatemian Toimituksia, AASF, SARJA-Ser. B, 237; Helsinki: Suomalainen Tiedeakatemia, 1987); Tov, TCU (1997); id., "Biliteral Exegesis of Hebrew Roots in the Septuagint?" HB, GB, and Qumran (2008) 378-97; Talmon, "Ancient Versions"; J. Ziegler, Untersuchungen zur Septuaginta des Buches Isaias (AT Abh XII.3, 1934).

1. Background

In ancient times and in the Middle Ages, the Bible was translated into different languages, the most important of which are Greek, Aramaic, Syriac, Latin, and Arabic. These translations are very significant for the textual criticism of the Hebrew Bible, since this discipline collects all the relevant material that is available from antiquity and the Middle Ages, including material derived from translated works. It goes without saying that these texts cannot be used in their own languages, since the textual discussion can only take into consideration Hebrew data. Therefore,

116

Chapter 2: Textual Witnesses

elements of the Hebrew texts underlying the various ancient translations need to be reconstructed. This reconstructed text from which a translation was made is called the parent text or Vorlage of a translation, that is, the text that lay before the translator. The importance of the ancient translations for the textual criticism of the Bible was emphasized greatly before the discovery of the Qumran scrolls, since before that time manuscripts of the ancient translations were the earliest sources for our knowledge of the biblical text. In the absence of ancient Hebrew material, scholars attached much importance to the ancient translations, since their early attestations (in the case papyrus fragments from the 2"d and 1st centuries BCE onwards and manuscripts from the 4th century CE onwards) preceded the medieval manuscripts of m by many centuries. The Qumran discoveries could have decreased the value of the ancient translations, since reliance on Hebrew texts is preferable to the use of ancient translations whose Hebrew source is not known. However, the Qumran scrolls are very fragmentary, and even if they were complete, would remain highly significant since it reflects important textual traditions differing from m, l.ll, and the Qumran texts. Several important readings are also reflected in the other translations, even though ([ s D are closely affiliated with m (together: m+ ). The Qumran discoveries enhanced the importance of also in other ways, since the agreements between the two improved our confidence in the reconstruction procedure.- § 5 The views of scholars are divided concerning the feasibility of reconstructing the Hebrew text used by the ancient translators. Some stress the ability of scholars to reconstruct words or sentences, while others emphasize the difficulties involved. Some general rules for reconstruction have been formulated - § 5, but they are of limited value. For even if scholars were to agree concerning abstract rules, the very use of one particular rule or another is based on subjective opinion. Most of the rules formulated for the reconstruction of the Hebrew source of the ancient translations were made in reference to since that translation yields more information relevant to the study of the Bible than all the other translations together. Most of the rules for the reconstruction from also apply to the other translations. In reconstructing the Hebrew source of ancient translations, one can take several points of departure. Every reconstruction is made with m in mind due to the large degree of congruence between m and the presumed Vorlage+ of the ancient translations, and because of the centrality of m in the textual procedure. The centrality of m in the reconstruction of the Vorlage of the ancient versions does not involve a value judgment on the central position of that text; it merely reflects a convention for which no better alternative has been suggested. A first convention in our approach to the ancient translations is that when the content of an ancient translation is identical tom, in all probability its Hebrew Vorlage+ was also identical to m. At the same time, this information does not simplify the task of reconstruction; since identity in content is not easily definable, all the words in the translation must be analyzed in detail. In the case of the Targumim, D, and usually also s,· their Hebrew source was almost completely identical to the consonantal framework of m - p. 29, and

IIA: Use of Ancient Translations in Textual Criticism

117

therefore reconstruction is limited to a small number of deviations from m. This identity is less pronounced in the case of l\3, and in some of its chapters there are very few overlaps with m. These data should be kept in mind when referring to the differences between m and the ancient translations. Since differences between m and the Hebrew sources of the ancient translations are often very significant for exegesis and our understanding of the textual transmission, the analysis of the ancient translations is a necessary part of textual criticism. Textual critics are especially interested in those cases in which a critical analysis of the versions yields a difference in meaning presumably resulting from deviations from m in the Hebrew Vorlage of one of the ancient translations. However, here lies the difficulty: how can we know in which cases this Vorlage indeed differed from m? Although there are thousands of differences between m and the translations, only a fraction of them were created because of a divergence between m and the Vorlage+ of the translation. Most of the differences were created by unrelated factors. These are inner-translational factors, especially in the area of exegesis ..... § 2, which created many renderings that are now described as differences between the translation and m. From a text-critical point of view, such differences are not significant since they were created by the translator and do not indicate a Hebrew source that deviated from m. 201 Another category of inner-translational factors( ..... § 4) includes corruptions in the textual transmission of the translation that caused apparent differences between it and m. The implication of this procedure is that before one makes use of a translation within the textual praxis, one has to be aware of all the intricacies of the exegetical system and translation technique of the translator of a given unit (book). Information of this type does not relate directly to the Hebrew source of the translation, but one needs to have a thorough knowledge of the character of each translation unit in order to reconstruct elements of its source ...... § 3 Tools have been developed for reconstructing the Hebrew source of a translation that has been made faithfully (literally), since such a translation usually employed the same equivalent for a particular Hebrew word or grammatical structure in most of its occurrences. On the other hand, if the translation was made freely or even paraphrastically, it is difficult and often impossible to reconstruct the Hebre"Y Vorlage+. Hence, an overall knowledge of the exegesis (--+ § 2) and representation of Hebrew constructions in specific translation units (..... § 3) is essential in order to be able to embark on the reconstruction activity ...... § 5

2. Exegesis

Within the present framework, there is only room for the most essential information about the translators' exegesis. This topic encompasses many secondary areas, and the reader is advised to peruse the bibliography relevant to the translations to be analyzed below. 201 At another level, that of the exegesis of the biblical text, these instances are significant since they provide early evidence of Bible exegesis.

118

Chapter 2: Textual Witnesses

a. Linguistic Exegesis Every translation reflects linguistic exegesis that is essential to this undertaking. This exegesis consists of three levels. • Linguistic identifications that identify all forms in the source language and the connection between the words. Without this identification, the words of the source text cannot be translated. Among other things, an analysis of most (but not all) morphological+ constituents of the nouns and verbs is essential for the translation procedure. For example, the translator needed to analyze the morphological nature of the word being translated in order to determine, for example, whether it was a noun or a verb. If it was a verb, the translator m?y have taken further steps in his analysis. In accordance with the grammatical concepts that developed from medieval times ·onwards, translators may have had an understanding of the root of the verb, as well as its conjugation (binyan), aspect, and tense. After all, translators distinguished between such homographic consonantal forms as the pi 'el wayedabber ("he spoke") and the hiph 'il wayadber ("he subdued"; correctly rendered by lfi in Ps 18:48 and 47:4 with forms of imoTCiaaw). 202 By the same token, a homograph such as ,N,' necessitated that the translator decide whether it is derived from a root r 'h, "to see," or from yr', "to fear," that is, in the Tiberian vocalization either "they will see" (passim in the Bible), or "fear!" (plural), e.g. Ps 34:10. The same decision had to be made regarding N,,,, which may be derived from either r 'h (NTJ, "and he saw") or yr' (NTJ = NT'J, "and he feared"). • Semantic exegesis of all the words in the source language. Before turning to equivalents, the translator has to determine the meaning of each Hebrew word. For example, any form of the verb NtDJ can be taken in at least four entirely different ways. Brock* 1988, 87 showed how the different translations of Gen 4:7 nNtD CN N,',ii reflect four different ways of understanding that verb. Most identifications are seemingly based on triliteral Hebrew verbs, but it does not necessarily follow that the translators followed a system of triliteral roots. The evidence merely shows that the translators were able to draw on various sources, enabling them to obtain the necessary semantic information. In most cases (e.g. , .. 1":J!J), all three letters were necessary for the identification, while in some cases two letters sufficed. Thus, in the weak verbs (patterns N"!J, '"!J, J"!J, N"!J, !J"!J, ,, ..!!, N"',, •"",), often only two radicals were used for semantic identification, and the translators often erred in this regard ...... Tov* 2008 • Determining the equivalents of words of the source language in the target language on the basis of the translator's knowledge and sensitivity in that language. 202 According to another scenario, translators had only a vague understanding of such abstractions as conjugations. It sufficed for them to distinguish between a form reflecting "something like the pi 'e/" and a form incorporating "something like the hiph'il." After all, it sufficed to distinguish between wydbr 1 (= wayedabber) carrying meaning 1 and wydbr 2 (wayadber) carrying meaning 2. The actual reading or pronunciation ("vocalization" in later times) and parsing are not a necessary part of the translation process.

IIA: Use of Ancient Translations in Textual Criticism

JJ9

All translations reflect at least these three levels of linguistic exegesis. Nevertheless, only a few translation units (such as Aquila+) are confined to such exegesis. As a rule, translations also reflect the first two types of exegesis that are described below. This feature may then be used to characterize the translations: the greater the number of exegetical elements found in a translation unit, the freer it is considered. Conversely, the greater the number of fixed equivalents in a translation unit, the more it is considered literal, and the smaller the number of such equivalents, the freer it is considered. Among the exegetical elements contained in the translations, it is possible to distinguish between linguistic exegesis that follows the text closely, and other forms of exegesis that move the translation away from a literal understanding. Some exegetical elements form a necessary part of the translation process, while others infuse the text with elements of the taste, understanding, and personality of the translator, sometimes to such an extent that the plain meaning of the text is completely obscured. For example, some translators allowed themselves the freedom to include allusions to other Bible verses or to insert their own reflections into the translation. However, even if according to our understanding the translators seem to have strayed far from the simple meaning of the verses, they were, nevertheless, reflecting what they considered to be their message. The three types of non-linguistic exegesis that are found in most of the translations are exemplified below.

b. Contextual Exegesis

A translator sometimes explains a detail based on another detail in the context or he may add or omit a detail. For example, Num 20:19

Exod 32:26

m

()'N p1)

0::: 0

(It is but a small) matter, (on my feet I would pass through.) a bad thing (i.e., a matter of offense)= 0::: Ps-J N

m

•'?N ';,'?

Who is on the LORD's side? To me! (= o:::o Ps-J N s Who is on the LORD's side? Come to me!

TLS rrpos KupLOv tTw rrp6s

NRSV

D)

Stylistic shortening is exemplified in the next example. Josh 4:14

m+

nN

mN ,N.,,,

and they revered him as they had revered Moses. KaL EopouvTo a{nov waTTE p Mwuaf)V and they revered him like Moses.

The linguistic exegesis mentioned in § a describes the determining of equivalents on the basis of linguistic-semantic identification alone. As with linguistic exegesis, contextual exegesis also has linguistic aspects, but often the overall meaning of the context is more influential for determining equivalents. For example, Exod 6:12

m

C'n::ltv '?1ll of uncircumcised lips aA.oyos

lacking verbal fluency

120

Exod 6:30 Exod 18:7

Chapter 2: Textual Witnesses

m

c'n:ltv ',ill

(\j

taxvowvo