Aramaic Studies (formerly Journal for the Aramaic Bible) http://www.continuumjournals.com/as Volume 2, Issue 2 July 2004 Contents Peter J. Gentry, Propaedeutic to a Lexicon of the Three: The Priority of a New Critical Edition of Hexaplaric Frag- ments ....................................................... 145 Gillian Greenberg, Indications of the Faith of the Translator in the Peshitta to the ‘Servant Songs’ of Deutero-Isaiah ..... 175 Dan Levene and Beno Rothenberg, Word-Smithing: Some Metallurgical Terms in Hebrew and Aramaic .......... 193 Matthew Morgenstern, Notes on a Recently Published Magic Bowl ................................................. 207 Harry Sysling, Three Harsh Prophets: A Targumic Tosefta to Parashat Korah .......................................... 223 Wido van Peursen, The Peshitta of Ben Sira: Jewish and/or Christian? .................................................. 243 Reviews ........................................................ 263 Bibliography of the Aramaic Bible .............................. 274 Index of Texts .................................................. 283 Index of Authors ............................................... 292 Index of Subjects ............................................... 297 Abstracts ...................................................... 305 [AS 2.2 (2004) 145-74] ISSN 1477-8351 Propaedeutic to a Lexicon of the Three: The Priority of a New Critical Edition of Hexaplaric Fragments Peter J. Gentry TheSouthernBaptistTheologicalSeminary,Louisville,KY Introduction OnOctober31,2003,somefiftyscholarsandstudentsgatheredinLeu- ven to congratulate J. Lust, E. Eynikel, and K. Hauspie in celebration of the publication of their careful revision of A Greek-English Lexicon of the Septuagint, which first appeared in 1992–1996.1 All are greatly indebted to them for putting into the hands of students and scholars alike such a useful and up to date lexicon of the Septuagint, a work longoverduesincethepioneeringeffortsofJ.F.Schleusnerappearedin 1820–21, almost one hundred and eighty years before. We are further indebted to the ‘Centre for Septuagint Studies and Textual Criticism’ of the Katholieke Universiteit Leuven and their sponsors, Deutsche Bibelgesellschaft, Stuttgart, for organising a symposium in which the next step in Septuagintal lexicography is to be vigorously addressed and pursued: the task of assembling a lexicon of the remains of the three Jewish revisers, Aquila, Symmachus, and Theodotion. From the beginning, my interest in the Three has been virtually as strong as my interest in the Septuagint itself. Interest in both and study of both has led me to convictions concerning interrelationship 1Symposium: Septuagint Lexicography and Beyond: Symmachus, Aquila, and Theodotion, on the occasion of the publication of the revised one volume edition of J. Lust, E. Eynikel, and K. Hauspie, Greek-English Lexicon of the Septuagint, Revised Edition (Stuttgart: Deutsche Bibelgesellschaft, 2003). I would like to ex- press my thanks to the organisers of the symposium, Prof. Johan Lust and Dr. KatrinHauspie,forinvitingmetospeak,andtotheparticipants,forconstructive commentsandquestions. (cid:13)c TheContinuumPublishingGroupLtd2004,TheTowerBuilding,11YorkRoad, LondonSE17NXand15East26thStreet,suite1703,NewYork,NY10010,USA. 146 Aramaic Studies 2.2 (2004) of the lxx and the Three. After twenty-five years of study devoted to the Septuagint, I am convinced that one cannot study the Septu- agint without studying the Three and vice-versa. During my doctoral research on Theodotion in Job and the quest for the so-called kaivge Tradition, I became aware of just how interwoven the text-history of the Septuagint is with the texts of the Three. As I attempted to lo- cate and place the Theodotionic materials in Job within the history of thetransmissionoftheseso-calledkaivgetextsandoftheSeptuagint,I concludedthat‘thepreciselineofdemarcationbetweenkaivgeandlxx hasyettobetakenseriouslybymodernscholarship.’2 NextIspenttwo years analysing the translation technique of the Greek Lamentations aspreparationforproducingthenetstranslation.3 WhileIdidnotdo an exhaustive analysis of the translation technique of Lamentations, as I had done for Job, the character of the translation as well as the absence of any materials attributed to Theodotion in the Second Ap- paratus of the G¨ottingen Edition strongly suggest that the Old Greek of Lamentations might possibly be the work of Theodotion. So again I was struck by the fact that the Septuagint and the Three were deeply interwoven together. In1997Iwasassignedthetask,begunbyJosephZiegler,ofproduc- ingacriticaleditionoftheearliestGreekVersionofEcclesiastesforthe G¨ottingen Septuaginta Series. The first two or three years engaged in this task were entirely focused on studying the Collation Books, trans- lation technique in Ecclesiastes, and the manuscript tradition of the Greek witnesses. Initially the critical text could be established largely from the study of the Greek witnesses apart from the witness of the daughter versions and patristic citations, although preliminary col- lation and study of these latter witnesses was already in hand. One daughter version, however, had to be analysed in detail before begin- ning the task of establishing the text, and that was the Syro-Hexapla. It is an extremely literal translation of the text in Origen’s Fifth Col- umnandoneofthemostreliablewitnessestotheOldGreekasOrigen received it. It is also a key source for readings from the Three, fre- quently not preserved elsewhere. Careful study of this version over the last three years revealed once again deep interconnections between the 2P.J.Gentry,TheAsteriskedMaterialsintheGreekJob (SBLSCS,38;Atlanta: ScholarsPress,1995),p.496. 3‘Lamentations’inTheNewEnglishTranslationoftheSeptuagint,editedbyA. PietersmaandB.G.Wright(Oxford:OxfordUniversityPress,forthcoming). (cid:13)c TheContinuumPublishingGroupLtd2004. Gentry Propaedeutic to a Lexicon of the Three 147 text history of the Septuagint and the text of the Three. In the course of my textual work, it was frequently necessary to establish the text of the Three in order to establish the original text of the Septuagint and vice-versa: the text of the Three could not be established without first establishing the text of the Septuagint. MyresearchontheSyro-HexaplaofEcclesiastesintersectsdirectly, therefore, with the interest of the sponsors of this Symposium to pro- duce a lexicon of the Three. While my area of expertise is not in lex- icography per se, this paper is entitled ‘Propaedeutic to a Lexicon of the Three’. According to the Oxford English Dictionary, the word propaedeutic means ‘serving as introduction to a higher study, pre- liminary study.’ Here, the higher study is lexicography of the Three. I would like to show from the Syro-Hexapla of Ecclesiastes study that is both necessary and preliminary to this higher study as well as integral to it: establishing a critical text of the Three.4 Severalimportantissuesareinvolvedinestablishingthetextofthe Three. To illustrate these I discuss problems in the text of Ecclesiastes arrangedaroundthreetopics:(1)determiningthetextoftheThreeby critical use of the O(ld) G(reek), (2) the relation of the Three and the OldLatininthetexthistoryoftheOGEcclesiastes,and(3)determin- ing the text of the OG by critical use of the Three. This apparently circulardiscussionisintendedtoshow,then,theinterconnectednessof the text history of OG and of the Three, methodological issues in es- tablishingthecriticaltextoftheThree,andtheneedforsuchacritical text. The format of the discussion is as follows. First to be cited is the Masoretic Text as the putative parent text of the Greek Translation.5 Next listed is the Greek Translation of Eccl. according to Rahlfs’ Edi- tion.6 If the critical text established for the forthcoming G¨ottingen Edition differs from that of Rahlfs, the siglum Ra designates the for- mer and Ge (i.e. Gentry) the latter. Third, the text tradition is given usingthelistofText-GroupsprovidedbyZieglerinanarticlepublished 4For strictures on the evidentiary value of Septuagint usage for lexicography, seeCameronBoyd-Taylor,‘TheEvidentiaryValueofSeptuagintalUsageforGreek Lexicography:Alice’sReplytoHumptyDumpty’,BIOSCS 34(2001),pp.47-80. 5K.Elliger,W.Rudolphetal.(eds.),BibliaHebraicaStuttgartensia (Stuttgart: DeutscheBibelgesellschaft,1983). 6A. Rahlfs (ed.), Septuaginta, Id est Vetus Testamentum graece iuxta LXX in- terpretes (Stuttgart:Wu¨rttembergischeBibelanstalt,1935). (cid:13)c TheContinuumPublishingGroupLtd2004. 148 Aramaic Studies 2.2 (2004) posthumously in the Robert Hanhart Festschrift.7 Fourth, the text of Syh is cited and fifth, the marginal notation in Syh is given with an indication of exactly where the index is placed. Sixth, evidence extant fortheThree inGreek sourcesis providedalongwiththeretroversions from Syriac proposed by Field (signalled by a smaller font). This is designated App II to conform to the Second Apparatus in the G¨ottin- gen Editions. After the data for each note are set forth a discussion ensues to evaluate the evidence. Determining the Text of the Three Marginal Notes in Syh Indicating One or Two of the Three are oJmoivw" (6= oJmoivw" toi'" oV) In previous papers presented at IOSCS meetings in Basel, 2001 and in Toronto, 2002, I focused on one kind of marginal note almost entirely unique to the Syro-Hexapla among our sources of hexaplaric materials for Ecclesiastes.8 This is the notation that indicates that one, or two, or even all of the Three Revisers, are ‘oJmoivw" toi'" oV’, i.e. the word or words which are marked by an index in Syh are identical to the text of the Seventy, the Fifth Column of Origen’s Hexapla. Approximately 70 notations of this type are extant according to Field’s Edition: 68 of them derive from Syh and Jerome is the main source for 2 not extant in Syh. Fifty-five of these notes contain the words oJmoivw" toi'" oV, one employs wJ" instead of oJmoivw" (wJ" oiJ oV 3.10), and in 14 only oJmoivw" is used.9 A chart displays the distribution of the notes: 7J.Ziegler,‘DerGebrauchdesArtikelsinderSeptuagintadesEcclesiastes’,inD. Fraenkeletal.(eds.)StudienzurSeptuaginta—RobertHanhartzuEhren (Abhand- lungen der Akademie der Wissenschaften in G¨ottingen, Philologisch-Historische Klasse, 190; Mitteilungen des Septuaginta-Unternehmens, 20; G¨ottingen: Vanden- hoeck & Ruprecht, 1990), pp. 83-120. Two manuscripts in Ziegler’s list must be excludedbecausetheyhavenotextofEccl.(334,695).Threemanuscriptsbelong moreproperlyunderhissiglumOl(Olympiodorus’CommentaryonEccl.),i.e.562, 733and734.Twofurthermanuscriptsaretobeincludedastheywerenotavailable to Ziegler, but have now been collated by the Septuaginta-Unternehmen(161 and 770). Two minor changes have been made in the sigla: cII is employed instead of C3 fortheCatenatriumpatrum(C´’=C +cI +cII)and766isemployedwhen 766Iand766IIagree. 8See Peter J. Gentry, ‘Hexaplaric Materials in Ecclesiastes and the Rˆole of the Syro-Hexapla’,Aramaic Studies 1(2003),pp.5-28andidem,‘TheRelationshipof Aquila and Theodotion to the Old Greek of Ecclesiastes in the Marginal Notes of theSyro-Hexapla’,Aramaic Studies 2(2004),pp.63-84. 92.19, 4.10, 17; 5.1, 19; 6.2, 7.27, 8.17, 9.1, 10.11, 20; 11.4, 9; 12.7. Nine of the fourteenwerediscussedearlier,seeGentry,‘HexaplaricMaterialsinEcclesiastes’. (cid:13)c TheContinuumPublishingGroupLtd2004. Gentry Propaedeutic to a Lexicon of the Three 149 Type of Note Instances Totals oV = aV sV qV 6= NIL 11 × oV = aV 6= sV 3 × 4 oV = aV 6= qV 1 × oV = aV 6= NIL 2 × oV = sV 6= aV qV 1 × 1 oV = sV 6= NIL 1 × oV = qV 6= aV 7 × 18 oV = qV 6= sV 9 × oV = qV 6= aV / sV 2 × oV = qV 6= NIL 7 × oV = aV qV 6= sV 12 × 12 oV = aV qV 6= NIL 8 × oV = aV sV 6= NIL 3 × oV = sV qV 6= aV 1 × 1 oV = sV qV 6= NIL 2 × The first line of the table is oV = aV sV qV 6= NIL. This means there are eleven instances where a marginal note indicates that aV sV qV are all identical to oV and NIL, i.e. no contrasting reading is provided by Syh. The totals include only the instances where a contrasting reading is provided. This demonstrates how to read the table. In the paper presented in Basel, all instances where all Three agree with the Old Greek and no contrasting reading is supplied were care- fully investigated. The main result of this study was that while the text of the Syro-Hexapla is for the most part a faithful representa- tion of the hexaplaric text, the marginal notes were probably not de- rivedfromthesamesource,butrather,weredrawnfromthemarginof a different Greek manuscript whose biblical text was non-hexaplaric. The Catena manuscripts seem to be the best candidate for the puta- tive source. The focus and purpose of the scholiast who excerpted the hexaplaric materials also differed from other traditions and this may (cid:13)c TheContinuumPublishingGroupLtd2004. 150 Aramaic Studies 2.2 (2004) explain why many of the readings of the Three are uniquely preserved by the Syro-Hexapla. Furthermore,inpreparingtheBaselpaperforpublication,aninter- esting discovery was made. If the lemma of the manuscript from which themarginalnotesinSyhwerecopiedandtranslateddifferedfromthe oV text, the marginal note would read only oJmoivw" (indicating that aV sV qV agreed with the lemma, but not with the oV text). If, however, the lemma and the oV text were the same, the note would read oJmoivw" toi'" oVtoshowthataV sV qVagreedwiththeoVtextaswellthelemma, or in some instances with the oV text against the lemma. Nine of the fourteen notes in which only oJmoivw" is used were dis- cussed previously.10 The remaining five notes in which oJmoivw" is used but only one or two of the Three Revisers are indicated are examined here. This analysis will demonstrate not only that the earlier proposal is correct, but also show how this proposal is critical for establishing the correct text of the Jewish Reviser in question. 2.19 (cid:141)m(cid:141)h tjt ytmkj(cid:141)w ytlm[(cid:141) w|/ ejmovcqhsa kai; w|/ ejsofisavmhn uJpo; to;n h{lion w|/ 2◦ Bc Sc A C O−637 797 254-342 68 149 161 248 252 260 296 311 339 443∗ 534´ 543 547c 548c 549 602 698 705 706 795 Fa Sa Syh = Compl Sixt Gra Ra å] absc 998; o 637 357-754; in quo Hi; w" 125II 542 Dam; > B∗-S∗ 411 C´’−797 a b 155 336´ 338 443c 547∗ 548∗ 602 645 705 766II GregAg 825 GregNy 362,12 368,8 Ol = Ald >acmç tyjt tmkjtaød YhO w toyaøld YhO .atwmUdbO hOb >t>a> YhO w Syhmg: Index super Field: aV qV oJmoivw" ThetextasestablishedbyRahlfsissound.Herightlychoseasorig- inal text the reading of A, C, the Hexaplaric text and the majority of minusculesagainstB-S,theCatenaGroupandanumberofminuscules influenced by them, since the omission of the second instance of the relative pronoun w|/ was not likely due to freedom on the part of the translator but rather is quite probable as an inner Greek improvement of the text.11 Field’s oV text is identical to the lemma of Rahlfs and therefore the note that Aquila and Theodotion are oJmoivw" can only mean that their 10SeeGentry,‘HexaplaricMaterialsinEcclesiastes’. 11Onlyoncein68instancesdoestheparticlev,havenocorrespondingequivalent inGreekandthisisduetorenderingafiniteverbbyaparticiple(1.11). (cid:13)c TheContinuumPublishingGroupLtd2004. Gentry Propaedeutic to a Lexicon of the Three 151 textisthesameastheOG.Indeedtheevidencemaybeconstruedthis way,butthequestionarises,whyisthemarginalnoteonlyoJmoivw"and not oJmoivw" toi'" oV when in fact the reading of Aquila and Theodotion is equivalent to that of the oV text as Field proposes? Alternatively, we may suggest that the note originated in the mar- gin of a manuscript such as we find in the Catena Group where the second instance of w|/ is omitted. In this case the note shows that the scholiast wanted to indicate that the text of Aquila and Theodotion alsoomittedtherelativepronounlikethelemmaofhismanuscript,but thatthiswasnottheoVtext.Thisgivesamoreconsistentandeconom- ical interpretation of the note oJmoivw" and fits the pattern established forthesenotes.ItisextremelyproblematicthatnoothersourcesforaV and qV exist apart from the marginal note in Syh. It is also difficult to find a similar syntactic situation, with two clauses introduced by v, or rv,a} in close proximity and joined by waw, where text is extant for aV andqVtodeterminewhethertheymightormightnothaveomittedthe secondrelativeintheirtranslation.Icanonlynotethatinmyprevious study approximately six similar instances were considered where the text proposed for the Three was the opposite of that claimed by Field and no other witnesses exist to adjudicate the matter.12 4.17 µyhlah tyb la ûlt r(cid:141)ak ûylgr rm(cid:141) [Qr ûlgr] fuvlaxon povda sou, ejn w|/ eja;n poreuvhÊ eij" oi\kon tou` qeou` w|/ eja;n poreuvhÊ B-S∗ A C 998 C´’ (w|/[... 371) a b−(125) d 68 125II 149 155 161 248 252 260 296´ 311 336´ 338 339 443 534´ 542 543 547 549 602 645 698 706 766 795 Dam GregAg 929 Ol PsChr] tw poreuesqai Sc O-411; cum vadis Hi Isid Jud 30,6 aUhlad atybl tna Lzøad Ytma Klyd algrl rf .atwmUdbO hOb >t> Lzøad Syhmg: Index super Field: qV oJmoivw" (ejn tw`/ poreuvesqai) Fromtheperspectiveofbothtextualwitnessesandtranslationtech- nique, no doubt can be cast on the text reconstructed by Rahlfs. One cannot pit the text of the Hexaplaric Group against 998 B and S∗ without powerful evidence from other factors—evidence which does not exist here. Indeed, the reading of the O Group is supported only by Sc. The OG rendering ejn w|/ eja;n poreuvhÊ is apparently based upon reading ûlt r(cid:141)ab rather than ûlt r(cid:141)ak as parent text; confusion of b 126.2; 7.26; 9.1; 10.11; 11.4; 12.7. See Gentry, ‘Hexaplaric Materials in Ecclesi- astes’,p.28. (cid:13)c TheContinuumPublishingGroupLtd2004. 152 Aramaic Studies 2.2 (2004) and k is common in the transmission of the Hebrew text.13 This Greek construction is found elsewhere in Eccl. only in 3.22 where hyhy(cid:141) hmb is rendered by ejn w|/ eja;n gevnhtai. The Hebrew construction b plus Infinitive Construct occurs in Qohelet only in 12.4 and is not rendered by ejn plus articular infinitive in Greek; indeed ejn plus articular infini- tive is a construction not found in Eccl. although it was increasingly popularinHellenisticGreek.14 Whateverthesource,thereadinginthe Hexaplaric Group seems certain to be a smoothing of the construction found in the entire manuscript tradition in Greek. The O Group thus has the lectio facilior. It may not be possible to be certain as to which parent text in Greek is the source for the rendering in Syh, but we should probably classifyitwiththeO Group.Fieldhasassumedthatthemarginalnote qV oJmoivw" in Syh means that the text of qV read ejn tw`/ poreuvesqai. If he is right, one wonders why, however, the note did not read oJmoivw" toi'" oV. On the other side, if the note came from a manuscript such as we find in the Catena Group, it would mean that qV had the same text as the OG reconstructed by Rahlfs, but not the oV text known to the scholiast. This would explain why the note is only oJmoivw" and not oJmoivw" toi'" oV.Itisplausiblethatthetextof998BandSwasthelxx known to qV and he did not revise it. We do not know here the text of either aV or sV. The Bible text of Jerome’s Commentary (cum vadis) follows our MT; doubtless this is a correction based directly upon the Hebrewtext.15 PerhapstherenderingsofaVandsV—thelattersooften influenced Jerome—were the same; one cannot say. I am proposing that qV had ejn w|/ eja;n poreuvhÊ and not the articular infinitive. It seems impossibletoadjudicatebetweenmyproposalandthatofFieldexcept that the explanation proposed here makes better sense of oJmoivw" and also can fit well the proposal that the notes came from a manuscript whoselemmawasdifferentfromthatoftheoVtext.Nonethelessthecase does show that determining the critical text of OG and establishing a critical text of the Three are inextricably intertwined. 13r(cid:141)abmaybeconsideredbysomeasanerrorforr(cid:141)ak,butitiswellattestedin MT.Thereare3instancesinEccl.renderedejn oi|"(3.9),kaqovti(7.2)andkaqwv" (8.4). The renderings in 7.2 and 8.4 probably show that the Greek Translator construedhisparenttexttohavethemorecommonr(cid:141)ak.Theinstancein3.9isa clearcaseandrenderedsimilarly4.17. 14SeeG.Horrocks,Greek: A History of the Language and its Speakers (London: Longman,1997),pp.46-47. 15Furtherexamplesarediscussedinfra. (cid:13)c TheContinuumPublishingGroupLtd2004. Gentry Propaedeutic to a Lexicon of the Three 153 5.19 wyyj ymy ta rkzy hbrh al yk o{ti ouj polla; mnhsqhvsetai ta;" hJmevra" th`" zwh`" aujtou` o{ti ou]j opou V | ouj (> C´−157 299 563 571 609 797) pollav Bc O C´−157 299 563 571 609 357 68 Didlemetcomm SaI Syh = Ald Compl Sixt Gra] ouk alla B∗ 998; ou kala" 534; ou polla" S A C 411 157-299-563-571-609-cII 46s-337 b d−357 125II 149 161 248 252 260 296´ 311 336´ 338 339 443 543 547 549 602 613 645 698 706 766 795 GregAg 961 Ol; ou polloi 631; ou pollwn 542; non Fa; non multum Hi = å >hUlyd ayj|d aUtmw|y rkdton YgUsO wld Lfm .atwmUdbO hOb >t>a> Lfm Syhmg: Index super Field: aV qV oJmoivw" Two problems in the text are listed for 5.19. Uncial V reads opou for oti ou—a common palaeographic error in the uncial square series wherePandTIareeasilyconfused.Secondly,ouj pollavisclearlyorig- inal as a literal rendering of the Hebrew text, but this reading suffered corruption within the Greek textual transmission. The readings of B∗ 998 and 534 are due to errors of sight and incorrect word division; ouj pollav" is an attempt to smooth the syntax and so is the lectio facilior along with the readings of 631 and 542. The Fayyumic omits the trou- blesomepollavaltogetherandJeromecorrectstheOldLatinaccording to the Hebrew. Syhreflectsouj pollavasparenttextthusagreeingwiththeoVtext. For this problem, however, the Catena Group is split. If the lemma from which the marginal note is derived had ouj pollav and aV and qV had the same, as assumed in the edition of Field, why not employ oJmoivw" toi'" oV as the notation since this is also the oV text? Just the opposite may be proposed as a solution which better accounts for the data. If the lemma of the scholiast was ouj pollav" and he wanted to indicate that this was also the reading of aV and qV but not that of the oV text, this could be clearly signalled by employing oJmoivw" rather than oJmoivw" toi'" oV. This proposal could also explain the origin of the reading which spread in part of the Catena tradi- tion. The data can be fitted nicely into the theory that the notes came from a different manuscript than the source of the text of Syh. This manuscript had as lemma a text like that of part of the Catena tradi- tion. Consistent distinction between oJmoivw" rather than oJmoivw" toi'" oV demands that we understand ouj pollav" as the text of Aquila and Theodotion. (cid:13)c TheContinuumPublishingGroupLtd2004.