“Double Translations in the Greek Versions of Daniel” in Interpreting Translation. Edited by F. Garcia Matinez and Marc Vervenne. FS J. Lust. Leuven: University Press – Peeters, 2005.
In Interpreting Translation. Edited by F. Garcia Matinez and Marc Vervenne. FS J. Lust. Leuven: University Press – Peeters, 2005.
Double Translations in the Greek Versions of Daniel
R. Timothy McLay, Assoc. Professor
St. Stephen’s University, Calais, ME 04619
The book of Daniel is particularly valuable for Septuagint studies because we have two complete Greek translations of the Hebrew. The existence of the two Greek versions, known respectively as the Old Greek (OG) and Theodotion (Th), raises a variety of questions about the nature of their relationship to one another as well as the source text or texts from which they were translated. The differences between the two versions are particularly noticeable in chapters 4–6 when compared to the rest of the book where both versions more closely reflect a Semitic text that is very similar to the MT . Most scholars are agreed that the differing version of chapters 4–6 in the OG is based on a Semitic Vorlage , and Wills and Albertz have made important contributions to understanding the alternative redactions of Daniel for these chapters while arguing that the OG version witnesses to a more original text for the book .
The significance of chapters 4–6 for examining the relationship between the Hebrew and Greek texts in Daniel becomes especially important when one considers the phenomena of double translations. A comparison of the texts yields quite a number of double translations throughout Daniel, but they are most easily isolated in chapters 4–6 because the versions differ there so markedly. For the purposes of textual criticism and this paper, it seems appropriate to define double translations on the basis of the relationship of the translation to the source text. That is, with Talshir , I would agree that double translations emerge at the level of the translation, though the reasons for two renditions for the same lexeme, phrase, or clause may differ. In most instances, double translations were created as scribal glosses or were hexaplaric additions. In some cases, however, what appears to be a double translation may actually reflect an accurate rendition of the Vorlage. And, to confuse the textual issues even further, there are instances in the translation where the exact same text is erroneously repeated. Though these repetitions have arisen in the translation, they really are not two translations of the same text. Thus, it seems more appropriate to label such instances more simply as doublets (see the discussion of 6,3 below). Comparing the instances of double translations and doublets in the texts of Daniel provides evidence of the corruption of the OG text by Th readings as well as important insights for textual criticism of the MT. In the following I will focus on larger double translations and note doublets in the OG only where they occur in a text under analysis.
Double Translations in Chapters 4–6
I begin with these chapters because they are replete with additions and expansions in the OG. Montgomery had argued that the additions were midrashic in origin , while others like Jahn and Charles argued for the originality of its text . Though it may now be stated with some confidence that it has become a scholarly consensus that the OG in these chapters witnesses to an earlier Vorlage than MT/Th, there has been little analysis of the nature of the additions and glosses in the OG text . In a forthcoming article, however, I have undertaken a thorough examination of the Greek texts in these chapters, and the results of that are comparison yield significant results. For example, many have commented on the state of the OG text. It contains obvious differences in the telling of the tales in these chapters, yet also shares similar vocabulary with Th and seems to reflect the MT in some places. On closer inspection, however, one discovers that the shared vocabulary in the Greek versions of Daniel in chapters 4–6 is actually minimal. If one uses the Th version as the point of departure there are over 2500 individual words that occur in these three chapters, but the two texts only have shared readings approximately 479 times. That is, there is shared vocabulary less than 20% of the time. What is even more interesting, however, is that when they do agree, it is almost verbatim. The statistics reveal that 398/479 or 83% of the agreements between the OG and Th are word for word. These agreements cover almost every possible type of agreement that is present in the narratives including conjunctions and particles. Based on the high percentage of verbatim agreements in the texts despite the low percentage of shared readings between the texts overall, the most reasonable explanation for these agreements is that they stem from textual corruption . The importance of this conclusion for the relationship between the OG and Th is that Th is not a revision of the OG in these chapters. To state the point more clearly, the two versions were completely independent translations . Since the Th version became the predominant Greek version it stands to reason that some of its readings would be incorporated into the transmission of the OG. The result is that textual corruptions from Th have often replaced the OG, but some of these replacements can be distinguished as double translations.
1. MT 4,29 Nwm(+y Kl Nyrwtk )b#(
Th 4,29 xo/rton w(j bou=n ywmiou=si/ se
OG 4,29 xo/rton w(j bou=n se ywmiou=si/ kai\ a)po\ th=j xlo/hj th=j gh=j e!stai h( nomh/ sou
The Th text agrees with the MT. Though there are significant differences in the OG version it is quite evident that the OG has a double translation for this text and that the first portion agrees verbatim with Th. The same clause xo/rton w(j bou=n ywmiou=si/n me has been added as a doublet in the following verse as well.
2. MT 5,23 Nwhb Nyt# )rmh . . . Kynbrbrw htn)w Kymdq wytyt htyb-yd )yn)mlw
Th 5,23 kai\ ta\ skeu/h tou= oi1kou au)tou= h!negkan e)nw/pio/n sou, kai\ su\ kai\ oi( megista=ne/j sou . . . oi]non e)pinete e)n au)toi=j
OG 5,23 su\ e)poih/sw e)stiatori/an toi=j fi/loij sou kai\ e!pinej oi!non, kai\ ta\ skeu/h tou= oi1kou tou= qeou= tou= zw=ntoj h)ne/xqh soi, kai\ e)pinete e)n au)toi=j su\ kai\ oi( megista=ne/j sou
The OG already refers to the king drinking wine with his friends: su\ e)poih/sw e)stiatori/an toi=j fi/loij sou kai\ e!pinej oi!non, while e)pinete e)n au)toi=j su\ kai\ oi( megista=ne/j sou is an addition that agrees with Th. It is best explained as a corruption from Th that could easily be removed as secondary to the OG’s narrative. It should also be noted that some of ta\ skeu/h tou= oi1kou tou= qeou= tou= zw=ntoj h)ne/xqh soi exhibits word for word agreement with Th/MT and this material is not required by the narrative in the OG either.
3. MT 6,4 yd lbq-lk )ynpdd#x)w )ykrs-l( xcntm )wh hnd l)ynd Nyd)
)twklm-lk-l( htwmqhl ty#( )klmw hb )ryty xwr
Th 6,3 kai\ h}n Danihl u(pe\r au)tou/j o#ti pneu=ma perisso\n e)n au)tw=,| kai\ o( basileu\j kate/sthsen au)to\n e)f ) o3lhj th=j basilei/aj au)tou=
OG 6,3-4a u(pe\r pa/ntaj e!xwn e)cousi/an e)n th=| basilei/a,| kai\ Danihl h}n e)ndedume/noj porfu/ran kai\ me/gaj kai\ e!ndocoj e!nanti Darei/ou tou= basile/wj, kaqo/ti h}n e)pisth/mwn kai\ suneto/j, kai\ pneu=ma a3gion e)n au)tw=|, kai\ eu)odou/menoj e)n tai=j pragmatei/aij tou= basile/wj, ai}j e!prasse, [to/te o( basileu\j e)bouleu/sato katasth=sai to\n Danihl e)pi\ pa/shj th=j basilei/aj au)tou= kai\ tou\j du/o a(/ndraj, ou$j kate/sthse met\ au)tou=, kai\ satrapaj e(kato\n ei1kosi e(pta] o3te de\ e)bouleusato o( basileu\j katasth=sai to\n Danihl e)pi\ pa/shj th=j basilei/aj au)tou=
Though Th exhibits fairly close formal agreement with the MT, the choice of u(pe\r au)tou/j for )ynpdd#x)w )ykrs-l( xcntm is one of several places (cf. vv. 2.4.5) in the opening verses of chapter six where Th has minuses. Comparing the texts that remain one is left with one of the more interesting examples of the way in which the texts have been corrupted in the course of their transmission.
There are obvious doublets within the OG text that Munnich has correctly placed in brackets . The inclusion of kai\ tou\j du/o a(/ndraj, ou$j kate/sthse met\ au)tou= and kai\ satrapaj e(kato\n ei1kosi e(pta is based on the larger addition of to/te o( basileu\j e)bouleusato katasth=sai to\n Danihl e)pi\ pa/shj th=j basilei/aj au)tou=. The number of satraps is repeated from 6,1 and, though the reference to the two men strictly speaking is not a doublet from anywhere else in the texts, it is included as part of this larger addition for the sake of narrative coherence. The clause to/te o( basileu\j e)bouleu/sato katasth=sai to\n Danihl e)pi\ pa/shj th=j basilei/aj au)tou= is a doublet for 6,4a . At the same time, these texts hearken back to kate/sthse . . . e)pi\ pa/shj th=j basilei/aj au)tou= in OG 6,1, and the similar readings in Th 6,3 kai\ o( basileu\j kate/sthse au)to\n e)f ) o3lhj th=j basilei/aj au)tou= and Th 6,1 kai\ kate/sthse e)pi\ th=j basilei/aj au)tou=0. . . e0n o3lh| th=| basilei/a| au)tou=. Elsewhere, I have suggested that these references to Daniel’s appointment along with the references to the satraps (who are not mentioned in OG after v. 3) are probably all secondary intrusions from Th . For example, this view is supported by the fact that Daniel is described as u(pe\r pa/ntaj e!xwn e)cousi/an e)n th=| basilei/a| and e)ndedume/noj porfu/ran kai\ me/gaj kai\ e!ndocoj e!nanti Darei/ou tou= basile/wj in v. 3, which is equivalent in sense (in fact it offers more information) to the secondary double translation to/te o( basileu\j e)bouleu/sato katasth=sai to\n Danihl e)pi\ pa/shj th=j basilei/aj au)tou=.
4. MT 6,21 l)yndl rm)w )klm hn( q(z byc( lqb
Th 6,20 e)bo/hse fwnh=| i)sxura=| Danihl
OG 6,20 e)ka/lese to\n Danihl fwnh=| mega/lh| meta\ klauqmou= le/gwn
The double translation fwnh=| mega/lh for meta\ klauqmou= is best explained as a corruption from Th.
Double Translations in Chapters 1-3
Comparing the Greek translations for double readings in chapters 1–3 differs from analyzing chapters 4–6 because even though the frequency of shared vocabulary readings is still relatively low, it is higher than chapters 4–6. In general, the discrepancy is explained by the significant difference in the Vorlage(n) for the OG in chapters 4–6 compared to the MT. Since the Vorlagen for the OG and Th appear to be much more similar in these chapters, the likelihood of shared readings that are based on rendering the same Vorlage increases. A comparison of these chapters reveals that the frequency of verbatim agreements exhibits a slight increase from chapter1 (30%) to 2 (36%), and an even higher increase in chapter 3 (44%). I have discussed the relevance of the verbatim agreements for understanding the relationship between the OG and Th versions elsewhere . For the purposes of this paper I will focus on examining some of the double translations that are evident in the OG when one compares it to Th and the MT.
1. MT 2,11 yhwty) )l )r#b-M( Nwhrdm yd Nyhl) Nhl
Th 2,11 a)ll ) h@ qeoi/, w{n ou)k e1stin h( katoiki/a meta\ pa/shj sarko/j.
OG 2,11 ei) mh/ tij a1ggeloj, ou{ ou)k e1sti katoikhth/rion meta\ pa/shj sarko/j: o#qen ou)k e)nde/xetai gene/sqai kaqa/per oi1ei.
The plus o#qen ou)k e)nde/xetai gene/sqai kaqa/per oi1ei is evidence that ou{ ou)k e1sti katoikhth/rion meta\ pa/shj sarko/j is rooted in corruption from Th’s text. Technically, the plus does not create a double translation of the MT because the OG plus would not be based on the MT, but it is this type of reading that underscores the fact that the Vorlage for the OG was different from the MT. Apart from the variant katoikhth/rion in the OG for Th’s katoiki/a, the reading exhibits a lengthy agreement in a section where there is hardly any shared vocabulary between the Greek versions.
One may note as supporting evidence for the corruption of the OG with Th readings in v. 11 the text baru/j e)sti kai\ e)pi/docoj earlier in the verse. This is the only occurrence of baru/j in the Greek texts of Daniel and it is a hapax legommena for ryqy in the LXX. Therefore, the shared reading of baru/j is a distinctive agreement in the Greek texts. However, the OG includes a double translation for ryqy in kai\ e)pi/docoj. For this reason, the double translation in the OG, and the fact that both terms are rare equivalences in the LXX, is evidence that the agreement of baru/j is a secondary addition from Th.
2. MT 3,2 )ytpt )yrbtd )yrbdg )yrzgrd) )twxpw )yngs )ynprd#x)l
)tnydm yn+l# lkw
Th 3,2 tou\j u(pa/touj kai\ tou\j strathgou\j kai\ tou\j topa/rxaj, h(goume/nouj kai\ tura/nnouj kai\ tou\j e)p ) e)cousiw=n kai\ pa/ntaj tou\j a!rxontaj tw=n xwrw=n
OG 3,2 pa/nta ta\ e1qnh kai\ fula\j kai\ glw/ssaj, satra/paj kai\ strathgou/j, topa/rxaj kai\ u(pa/touj, dioikhta\j kai\ tou\j e)p ) e)cousiw=n kata\ xw/ran kai\ pa/ntaj tou\j kata\ th\n oi)koume/nhn
It is almost impossible to sort out the corruption in the texts as they have been transmitted in this verse. Immediately, one is struck by the repetitious nature of the OG’s references to various groups of people:
i. pa/nta ta\ e1qnh kai\ fula\j kai\ glw/ssaj
ii. satra/paj kai\ strathgou/j, topa/rxaj kai\ u(pa/touj, dioikhta\j
iii. kai\ tou\j e)p ) e)cousiw=n kata\ xw/ran
iv. kai\ pa/ntaj tou\j kata\ th\n oi)koume/nhn
It may be observed that i. and iv. generally refer to the whole population while ii. and iii. refer to those who are authorities. This repetition does not require that two of the four groups are secondary, but a comparison to Th/MT suggests that some of these additions are doublets that are rooted in Th or harmonization to the MT. For example, it is a characteristic of the OG, which is particularly evident in this chapter, for it to abbreviate any lists (e.g. vv. 3, 7, 10, 15). Therefore, it is unlikely that OG has both the pluses pa/nta ta\ e1qnh kai\ fula\j kai\ glw/ssaj as well as a long list of officials. At the very least, it is probable that the verbatim agreements strathgou/j, topa/rxaj and u(pa/touj are secondary additions to the OG based on Th. This would allow satra/paj kai\ dioikhta\j in the OG to serve as general terms that would include all officials. Though i. and iv. appear to be double translations, there is no Vorlage for either of these pluses in the MT. There is a partial agreement of kai\ pa/ntaj tou\j with Th, so it is more likely that iv. is the doublet.
The shared reading kai\ tou\j e)p ) e)cousiw=n is a rare example of an OG reading that can be proven to have infiltrated Th. The reason for this is that Th has two references to all those in authority: kai\ tou\j e)p ) e)cousiw=n kai\ pa/ntaj tou\j a!rxontaj tw=n xwrw=n. Th renders the identical Aramaic with kai\ pa/ntaj tou\j a!rxontaj tw=n xwrw=n in the very next verse; therefore, kai\ tou\j e)p ) e)cousiw=n in Th is a secondary double translation from the OG.
3. MT 3,17 Nyxlp )nxn)-yd )nhl) yty) Nh
Th 3,17 e!sti ga\r qeo/j, w|{ h(meij latreu/omen
OG 3,17 e!sti ga\r o( qeo/j o( e)n ou)ranoi=j ei{j ku/rioj h(mw=n, o{n fobou/meqa
This is an excellent example of how a secondary Th reading has been adopted into the OG. The equation of qeo\j for hl) (Nyhl)) is unlikely in the OG. In fact, at least up until 3,20, the OG seems to prefer forms of ku/rioj for references to the divinity. qeo\j does not appear in chap. 1 in the OG, according to Munnich’s text, and only three times in chap. 2 (vv. 44.45.47). The rendering in v. 44 is dubious because in all three prior instances (vv. 19.28.37) where )ym# is collocated with hl) the OG renders hl) with ku/rioj. The lengthy plus in v. 47 e)sti\n o( qeo\j u(mw=n qeo\j tw=n qew=n kai\ ku/rioj tw=n kuri/wn kai\ ku/rioj tw=n basile/wn where Th has o( qeo\j u(mw=n au)to/j e)sti qeo\j qew=n kai\ ku/rioj tw=n basile/wn is likewise due to the insertion o( qeo\j u(mw=n qeo\j tw=n qew=n from Th. Without the secondary addition from Th, the OG would read something like e)sti\n ku/rioj tw=n kuri/wn kai\ ku/rioj tw=n basile/wn. That leaves v. 45 as the only instance in the first two chapters where qeo\j renders hl), when everywhere else ku/rioj is the designated term for the deity. The lone instance seems improbable, but we do not have the textual evidence for what was probably ku/rioj. Prior to 3,20 there are only three other occurrences of qeo\j in the translated portions of the OG according to Munnich in chapter 3: vv. 14.15.17. It is possible that the translator employed qeo\j in vv. 14.15 because these verses are direct speech from Nebuchadnezzar. The translator may have wanted to indicate that the king believed that the image was a god.
Thus, qeo\j is a word for the deity that one finds only rarely, if at all, in the first three chapters of the OG , and, in v. 17, it is also part of a double translation that agrees with Th. Finally, it can be observed that in a very similar text in 2,28 the OG has ku/rioj for hl). The similarity to 2,28 may also explain the addition of o( e)n ou)ranoi=j. The core reading of the OG appears to be ei{j ku/rioj h(mw=n, o{n fobou/meqa, which includes ei{j as a plus emphasizing monotheism, while something similar to o( e)n ou)ranoi=j may have been part of the original as well. There is no way to determine whether o( e)n ou)ranoi=j was part of the original text and/or whether it was based on a Vorlage similar to 2,28, or whether it was adopted into the text to smooth out the reading after the Th reading was added.
4. MT 3,24 hlhbthb Mqw hwt )klm rcndkwbn Nyd)
Th 3,91 Kai\ Nabouxodonosor h!kousen u(mnou/ntwn au)tw=n kai\ e)qau/mase kai\ e)cane/sth e)n spoudh|=
OG 3,91 Kai\ e)ge/neto e)n tw=| a)kou=sai to\n basile/a u(mnou/ntwn au)tw=n kai\ e(stw\j e)qew/rei au)tou\j zw=ntaj, to/te Nabouxodonosor o( basileu\j e)qau/mase kai\ a)ne/sth speu/saj
The agreement of e)qau/mase, which in Th translates hwt, would have to be considered a distinctive agreement between the two Greek texts. Though e)qau/mase is a good rendering, it is also a unique equivalent. This shared reading, which agrees with the MT (apart from the plus u(mnou/ntwn au)tw=n) and the plus in the OG requires closer examination. For example, e)qau/mase kai\ a)ne/sth speu/saj is very similar to kai\ e)qau/mase kai\ e)cane/sth e)n spoudh|= in Th and reads as a double translation with kai\ e(stw\j e)qew/rei au)tou\j zw=ntaj. Furthermore, it should be noted that u(mnou/ntwn au)tw=n is not required by the syntax of the OG, while it is in Th where the finite verb is employed. Thus, the OG has been expanded by secondary material from Th.
Double Translations in Chapters 7–12
In their present form the vision chapters generally exhibit a much closer correspondence between the Greek texts. This correspondence is reflected in a higher percentage of verbatim agreements in the vocabulary and a much closer formal equivalence of the OG to the MT. In all likelihood, at least some of the connections between the OG and Th/MT may be explained by the hypothesis that the Vorlagen for the Greek versions were more similar for the latter chapters than they were in chapters 1–6; however, the witnesses still preserve instances where the OG diverges quite significantly from Th/MT. That there are occasionally larger pluses and minuses in the OG is an indication of the divergent character of the OG text. For example, 9,24-27 is obviously corrupt, though Munnich’s text only has brackets in v. 27 . The following will examine other instances of larger double translations and doublets in the OG.
1. MT 8,11 )bcw w#dqm Nwkm Kl#hw dymth Myrh wnmmw
Th 8,11 kai\ di' au)to\n qusi/a e)rra/xqh, kai\ e)genh/qh kai\ kateuodw/qh au)tw=|,
kai\ to\ a#gion e)rhmwqh/setai
OG 8,11 kai\ di' au)to\n ta\ o!rh ta\ a)p' ai)w=noj e)rra/xqh, kai\ e)ch/rqh o( to/poj
au)tw=n kai\ qusi/a, kai\ e!qhken au)th\n e3wj xamai\ [e)pi\ th\n gh=n] kai\ eu)odw/qh kai\ e)genh/qh, kai\ to\ a#gion e)rhmwqh/setai:
8,11–12 is a notorious text. In the Greek versions it is part of a larger portion (8,11–17a) where there is extensive verbatim agreement between the OG and Th. Other scholars have already argued for a variety of doublets in the OG, but, in keeping with what one finds in the earlier chapters, my contention is that the nature of the verbatim agreements argues for Th corruption of the OG.
Montgomery has noted, ta\ o!rh ta\ a)p' ai)w=noj is the OG translation for dymth Myrh (therefore qusi/a is a double translation from Th) and o( to/poj + au)tw=n renders Nwkm . It also seems probable that kai\ e!qhken au)th\n e3wj xamai\ (glossed by e)pi\ th\n gh=n) is a gloss that originated from v. 12 . Against Montgomery and Jeansonne, however, it seems more probable that e)rra/xqh originated as a doublet from Th for e)ch/rqh. Both Montgomery and Jeansonne presuppose that Th is revising the OG; therefore, Th readings are too easily accepted as reliable guides for reconstructing the OG. Besides the additions of qusi/a and e)rra/xqh to the OG from Th, the readings kai\ di' au)to\n and kai\ eu)odw/qh kai\ e)genh/qh, kai\ to\ a#gion e)rhmwqh/setai exhibit substantial agreement between the OG and Th. Montgomery suggests that kai\ to\ a#gion e)rhmwqh/setai is the equivalent for )bcw w#dqm, while kai\ eu)odw/qh kai\ e)genh/qh are secondary glosses . Again, Montgomery’s reconstruction assumes that Th is a reliable guide to OG readings when it seems more likely that the lengthy plus in the OG actually stems from Th influence. There are no reliable means to determine whether the OG even had an equivalent for )bcw w#dqm or what might have been in its Vorlage.
2. MT 8,16 h)rmh-t) zlhl Nbh l)yrbg rm)yw )rqyw
Th 8,16 kai\ e)ka/lese kai\ ei]pe Gabrihl, sune/tison e)kei=non th\n o#rasin
OG 8,16 [kai\ e)ka/lese kai\ ei}pe Gabrihl, sune/tison e)kei=non th\n o#rasin]kai\ a)naboh/saj o( a!nqrwpoj ei}pen )Epi\ to\ pro/stagma e)kei=no h( o#rasij
Munnich correctly has the double translation from Th in brackets. Jeansonne argues that the double translation was preserved in the OG and that Th chose the reading that most closely corresponded to the MT. Besides the obvious insistence that Th is a revision of the OG, this view must presuppose that the Th version is quite late . Furthermore, Jeansonne’s view fails to account for the fact that similar double Th corruptions can be isolated in the OG.
3. MT 10,13 Myn#)rh Myr#h dx) l)kym hnhw
Th 10,13 kai\ i0dou\ Mixahl ei[j tw=n a)rxo/ntwn tw=n prw/twn
OG 10,13 i0dou\ [ei[j tw=n a)rxo/ntwn tw=n prw/twn], ei[j tw=n a(gi/wn a)gge/lwn
Though Munnich correctly demonstrates that ei[j tw=n a(gi/wn a)gge/lwn derives from the OG translator and identifies ei[j tw=n a)rxo/ntwn tw=n prw/twn as secondary, he remains cautious in his assessment of the bracketed text as corrupt . The agreement with Th is clearly the result of secondary influence from Th.
4. MT 11,8 Mhyksn-M( Mhyhl) Mnw
Th 11,8 kai\ ge tou\j qeou\j au)tw=n meta\ tw=n xwneutw=n au)tw=n
OG 11,8 kai\ tou\j qeou\j au)tw=n katastre/yei meta\ tw=n xwneutw=n au)tw=n kai\ tou\j o!xlouj au)tw=n
This is a puzzling text. There do not seem to be any means to suggest how a text that was similar to Mhyksn-M( was rendered by the OG with (kai\) tou\j o!xlouj au)tw=n, though the presence of this text in all the OG witnesses assures that it is genuine. This is also the only time in Daniel that the verb katastre/fw appears. However, the very fact that Kysn is infrequent in the MT and that meta\ tw=n xwneutw=n au)tw=n agrees with Th indicates that the shared reading is a secondary translation. It is hard to know what the OG was reading in its Vorlage.
5. MT 12,8 hl) tyrx) hm
Th 12,8 ti/ ta\ e!sxata tou/twn
OG 12,8 ti/j h( lu/sij tou= lo/gou tou/tou [kai\ ti/ ai( parabolai\ au{tai;]
This text represents an example of a double translation where one of the readings does not stem from Th. It is included here because it is an example of the degree to which the OG can differ from Th. There are instances in chapters 7-12 where the OG diverges quite significantly from Th, even where it may be argued that they are based on a similar Vorlage. Thus, are the differences where they agree verbatim due to Th revising the OG or are they best explained as scribal corrections of the OG with Th readings? The presence of these double translations and the cases I have demonstrated in other publications argue that the latter is the case. In this example, Munnich marks the second translation as the gloss and accepts Montgomery’s suggestion that parabolai\ is based on reading twdx) in the Vorlage . Since tou/tou = hl) in the first translation and parabolai = twdx) in the second, Montgomery’s suggestion demonstrates that both readings could be characterized as free renditions of a Vorlage similar to the MT. Therefore, there is not enough evidence to determine which reading should be considered original.
6. MT 12,13 Nymyh Cql Klrgl dm(tw xwntw Cql Kl ht)w
Th 12,13 kai\ su\ deu=ro kai\ a)napau/ou: kai\ a)nasth/sh| ei)j to\n klh=ro/n sou ei)j sunte/leian h(merw=n
OG 12,13 kai\ su\ ba/dison, a)pw/qou: e!ti ga/r ei)sin h(me/rai kai\ w{rai ei)j a)naplh/rwsin suntelei/aj, kai\ a)napau/sh| kai\ a)nasth/sh| e)pi\ th\n do/can sou ei)j sunte/leian h(merw=n
Based on the number of double translations and the corruption of the OG text it is quite fitting that the book should end with one. Once again, there is a double translation that agrees almost verbatim with Th and corresponds to the MT. Though the first translation is not a formal equivalent for the present MT, it is similar in its eschatological focus. The main difference is that Th/MT refer specifically to the rising/resurrection of Daniel. This promise builds on vv. 1-3 and the Sitz im Leben for these verses are associated with the maskilim, who are the redactors of the book in its final form . It should also be noted that scholars are agreed that the book as a whole is the product of a complex series of redactions and that chapter 12 exhibits signs of more than one redaction . The secondary double translation in v. 13 is evidence of that redaction process. Whether e)pi\ th\n do/can sou should be regarded as a corruption of ei)j to\n klh=ro/n sou (by a scribe reading Klrgl for Klrgl) or as original is uncertain.
Conclusions
The evidence of secondary scribal corrections of the OG with Th readings is overwhelming. On the basis of the full access to the text of papyrus 967 Munnich has already removed some additions to the OG text compared to the earlier edition by Ziegler. The examples we have examined offer additional instances where the text of the OG has been corrupted. The primary evidence for this corruption is that the OG contains double translations and, in most of the cases analyzed, it was apparent that a majority of one of the translations for each reading in the OG exhibited verbatim agreement with Th. These examples can be isolated throughout the book of Daniel and are important evidence of the degree to which the OG text was the victim of scribal correction with Th readings in the course of its transmission. Given that these double translations can be determined to be secondary scribal corrections, and that I have demonstrated that Th is not a revision in chapters 1–6, there is a greater probability that the increased frequency of verbatim agreement in chapters 7-12 is due to secondary corruption of the OG with Th readings.
