“The Old Greek Translation of Daniel Chapters 4–6 and the Formation of the Book of Daniel” in Vetus Testamentum 55(3):304-323.
The Old Greek Translation of Daniel Chapters 4–6 and the Formation of the Book of Daniel
by Assoc. Prof. R. Timothy McLay
St. Stephen’s University, Calais, ME 04619
ABSTRACT: The content of the Old Greek translation of Daniel iv–vi is significantly different compared to the so-called Theodotion version and the Masoretic Text. In addition, the best witness to the Old Greek version (papyrus 967) has an alternative order for the chapters: chapters vii and viii intervene between iv and v. The proposals by J. Lust and O. Munnich that 967 preserves a more original version of the content and order of the chapters for the Vorlage of Daniel are critiqued. Additional linguistic evidence that supports the theory that the Old Greek translation of chapters iv–vi circulated together independently is also provided. Finally, a hypothesis for the growth and stages of the book of Daniel that includes an explanation for the origins of the Greek versions is outlined.
The Old Greek (OG) translation of Daniel chapters iv–vi presents the Septuagintalist with a host of interesting textual problems. While the Theodotion (Th ) version reads like a translation based on formal equivalence of a text that is very similar to our existing Semitic texts of Daniel, the OG exhibits little textual similarity to either Th or the MT. The vast differences between the versions of the chapters represented respectively by the OG and Th/MT are particularly striking when compared to the rest of the book where the Greek versions are both based on Vorlagen that are very similar to the MT. Scholars are agreed that the differing version of chapters iv–vi in the OG is based on a Semitic Vorlage, and Wills and Albertz have offered important contributions for understanding the differing literary redactions of Daniel for these chapters while both also argued for the priority of the OG version. Indeed, the essential link between textual and literary criticism that is increasingly acknowledged in biblical studies owes a debt of gratitude to the discovery of the ancient manuscripts in a cave in the Judean desert in 1947. For Daniel there are two Greek versions that can be compared to one another as well as to the Semitic version. Th is generally quite close to the MT throughout the book; therefore, the texts of the Greek versions for chapters iv–vi differ considerably.
In general, the basic story in Daniel iv–vi is the same in the versions transmitted respectively by the OG and MT/Th, but there are numerous differences in details. For example, chapter iv still narrates the story of Nebuchadnezzar’s madness, but his confession and the publication of his decree occurs in a much expanded form at the end of the chapter in vv. 34(37)-34c rather than the beginning, and there are other pluses to vv. 14(17), 19(22), 23-25(26-28), 28(31), 30(33). There are also significant minuses that involve vv. 20-22(23-25) when one compares the OG to MT/Th and there is no equivalent for iv 3-6(6-9). Chapter five recounts the mysterious writing on the wall, but the OG version includes an abbreviated version of the story as a preface and omits significant portions of vv. 3, 10-13 and has no equivalent for vv. 14-15, 18-22 and 24-25. Chapter six, in which Daniel is thrown into the lion’s den, is much closer in length in the different versions, but there are large pluses in the OG vv. 3(4), 5(6), 12a, 14(15), 17-18(18-19), 22(23) and minuses in 15(16), 23(24). Even where there are no pluses or minuses in these chapters the Greek versions share little relationship with regard to style, grammar, and, more importantly, vocabulary.
Though he has not given evidence to support his view, Eugene Ulrich has suggested that the OG translation for chapters i–xii “is of one piece”. This view presupposes that the alternative translation of chapters iv–vi stems from a Semitic Vorlage that was complete for chapters i–xii. Given the fact that the content of chapters iv–vi (the madness of Nebuchadnezzar, the writing on the wall, Daniel and the lion’s den) exhibits significant literary differences as a stand alone corpus compared to the rest of the book it is doubtful that one can make a decisive argument in this matter. However, a case can be made that the chapters do not originate with the same translator as the rest of the book. Albertz has argued that the differences between OG and MT/Th in iv–vi are due to the fact that the translator of chapters i–iii, vii–xii adopted the earlier “popular” translation of iv–vi into his translation because it exhibits a theological emphasis on monotheism, which is also detectable in iii 17. One of the problems with this argument is that the OG does not exhibit a concern for monotheism elsewhere, but the fact that there is a distinct emphasis on monotheism expressed in iv 34c![37], which is echoed by repeated statements that their God is “the living God” who reigns “in heaven” (see iv 23[26], 28[31], 34[37]; v 23), remains an argument that these chapters stem from a different hand. Furthermore, there are instances where the translation of vocabulary in chapters iv–vi was different than elsewhere in the book. An analysis of the translation of the vocabulary in these chapters provides additional evidence that supports a different translator (#n), ht), )(b, hlm, Nd( ). These examples are all based on different Greek translation equivalents for Semitic vocabulary occuring both inside and outside of Daniel iv–vi; therefore, they are weighty indicators of different translation hands at work. Another type of evidence would be to note the instances where certain Greek vocabulary is restricted to chapters iv–vi, regardless of the Semitic Vorlage. However, despite the fact that the evidence favors a separate translation for these three chapters, it might be imprudent to claim that the evidence alone is compelling. Though the combined evidence of the examples does exhibit distinct patterns of translation, the judgments for the first types of examples are based on the assumption that one can reconstruct the Vorlage for the vocabulary for OG based on the content and comparable material existing in the MT. Since the content of the stories is similar in OG and Th/MT, despite the clear differences, this assumption seems reasonable in most cases, whether the Vorlage for OG was Aramaic or the translation is based on cognate Hebrew words.
Though one may not claim that an overwhelming case for the distinct nature of the translation for OG chapters iv–vi has been made previously, it has not seemed to this writer that there are serious rival explanations. However, the publication of recent theories about the origins of Daniel are indicators that a more complete examination of the available evidence is in order. In particular, there have been several scholars who have argued for an earlier Vorlage for Daniel and that it had an alternative order for the chapters based on the evidence of the best witness to the OG text, papyrus 967. In 967 chapters vii and viii intervene between iv and v. The result is a somewhat smoother chronology because the result has the narrative of chapters i–iv situated in the reign of Nebuchadnezzar, vii–viii and v in the time of Belshazzar, vi and ix in the period of Darius, and x–xii are dated in the first year of Cyrus. The alternative order of the material in papyrus 967 raises additional questions about the literary development of Daniel and its relationship to what became the MT. How is the content of the OG and the different alignment of chapters iv–vi (in papyrus 967) in relationship to the MT and the TH version to be explained? A necessary consequence of arguing for the alternative order for chapters iv–vi has also meant that these scholars have challenged the view that OG chapters iv–vi represent a distinct and separate translation.
In what follows two different proposals that advocate that 967 preserves a more original version of the content and Vorlage for Daniel will be examined. More space will be devoted to the most recent proposal by Olivier Munnich, because he argues that 967 witnesses to an alternative literary edition that was complete for the book of Daniel and that one can demonstrate linear development from its Vorlage to the MT in some cases. Significant criticisms are raised against both attempts to posit a Semitic edition of Daniel with an alternative order of the chapters as in 967, and of Munnich’s argument for a direct literary development from the Vorlage of the OG to the MT. Following the examination of the proposals by Johan Lust and Munnich a hypothesis for the outlines of the growth and stages of the book of Daniel that includes the Greek versions will be outlined.
967 Witnesses to an Early Collection of Daniel Stories from (i)ii–viii
Whether chapters iv–vi were part of an integrated literary whole encompassing at least chapters i–xii (Ulrich does attempt to provide an explanation for the relationship of the deutero-canonical additions to the known Hebrew version) is fundamental for understanding the stages of growth and composition of Daniel. Typically, it is suggested that papyrus 967 has rearranged the order of the chapters in order to fix the chronology, but Johan Lust has argued for the priority of papyrus 967. One of the strengths of Lust’s argument is that it builds on the observation that originally the tales of Daniel in chapters ii–vi circulated as independent compositions. They seemed to have been part of an existing tradition of Daniel stories, which is supported by the additions that are known to us only in their Greek versions, as well as by the Daniel fragments and the Prayer of Nabonidus discovered at Qumran. One can readily agree with Lust that the OG, including 967, witnesses to an earlier Semitic collection of Danielic tales; however, his views regarding the priority of the order of the chapters in 967 are problematic.
Lust suggests that the pre-Maccabean collection of Daniel stories included a form of chapter vii and the tales were redacted together according to different patterns: one of which is witnessed to by 967. Chapter viii, like the rest of the Hebrew sections, was added later. There are two main issues, however, that Lust does not explain. First, there is no reason provided for the hypothetical ordering of the chapters according to OG. Lust correctly notes that the order of chapters ii–vii reflects conscious and deliberate editorial activity, but this does not in any way demonstrate dependence on the order witnessed by 967. Second, how does Lust account for chapter viii? According to Pablo David, who expanded on Lust’s views in his doctoral thesis, chapter viii was appended to chapter vii in this pre-Maccabean collection of Aramaic tales along with i–ii 4a, and this was later followed by x–xii and chapter ix. But, how likely is it that a chapter written in Hebrew was sandwiched between chapters written in Aramaic, particularly if this editor was already attaching an introduction in Hebrew? Chapter eight causes an even more significant problem for the Greek. David suggests that the original Aramaic collection of ii, iii, iv, vii, v, vi, Bel and the Dragon was translated into Greek and that the translation for these chapters was retained and used by the later translator for chapters i, viii, and ix–xii. How did the Greek translator know to put chapter viii after vii if these were separate editions? According to David, the Greek translator would have had to have knowledge of three separate collections: ii–vii plus Bel and the Dragon; i–viii plus Bel; i–xii (plus additions). This is possible, but is it likely, especially when there could not have been a great deal of time between the production of the chapters that would have been added to the original core? This scenario offers no explanation for the setting in which the OG was translated and how the OG managed to preserve an alternative Semitic order of the chapters.
967 Witnesses to an Earlier Literary Edition than the MT
Recently, a new proposal has been advanced by O. Munnich, which, similar to Lust, advocates the priority of the order of papyrus 967 as a witness to the original order of Daniel. In addition, in agreement with the views of Ulrich, Munnich believes that the OG witnesses to an alternative literary edition. Munnich is concerned to demonstrate literary development from the OG to MT, and to this end he does note several ways that chapters iv and v in the MT exhibit that it is a later redaction than the OG, particularly in its development of the role and character of Daniel. For example, Daniel is not mentioned in OG chapter iv until v. 15(18) compared to the MT in which he is described as “endowed with the a spirit of the holy gods” and “no mystery is too difficult” for him in iv 8-9(5-6). The intelligence and wisdom of Daniel are also emphasized in the MT in v 11-16. Munnich is also no doubt correct when he identifies later additions to MT iv 3-6 and v 3.
It is interesting to examine Munnich’s argument more closely, because it is an excellent example of the relationship between literary and textual criticism. In particular, most cases of different literary editions witnessed to by the MT and LXX exhibit a linear connection, i.e. one developed from the other. Munnich’s examples from chapter iv 3-6 and v 3 are part of his argument that seeks to demonstrate the redactional process from the Semitic Vorlage of the OG to the MT. Moreover, it is essential to note that Munnich is arguing for the priority of the order of the chapters according to 967 in this hypothetical original Vorlage as well. In what follows it is argued that Munnich is mistaken on both counts. Though the MT reflects later and more extensive redaction in chapters iv–vi than does the OG and the OG witnesses to an earlier Vorlage, Munnich has neither offered sufficient evidence to demonstrate a linear connection between the Vorlage of the OG and the MT nor a satisfactory argument for the priority of the order of the chapters according to 967. His views will be examined in two parts. First, the omission in OG iv 3-6, which Munnich offers as evidence for redactional development from the OG to the MT will be discussed. Second, the notion that the order of the chapters in 967 is a witness to an earlier Semitic edition of Daniel will be scrutinized in detail. Some of the same criticisms that can be brought against Munnich’s position would apply equally to Lust, especially when it comes to their common view of the originality of the order of the chapters in papyrus 967.
In his most recent study Munnich argues not only that iv 3-6 are evidence of later redaction in the MT, but that one can also reconstruct the editorial link between the MT and the OG 967. The OG and MT are presented in parallel alignment below. Lines 1-5 are v. 2 and 6-8 represent the beginning of v. 7.
OG iv 2(5), 7 (10) MT 4, 2(5).7(10)
1.e)nu/pnion ei}don tyzx Mlx
2.kai\ eu)labh/qhn ynn%lxdyw
3.kai\ fo/boj moi e)pe/pesen Nyrhrhw
4. e)pi th=j koi/thj mou
5.
ybk#$m-l(
ynn%lhby y#$)r ywzxw
6.
7.
8. e)ka/qeudon kai\ i)dou/ y#$)r ywzxw
ybk#$m-l(
wl)w tywh hzx
Munnich begins by noting that line 2 is not quite equivalent in the OG to the MT and offers a possible Vorlage presumed by lines 3-4 (ybk#$m-l( hlpn hdrxw), which are quite different. He also notes that lines 3-4 are absent from the Peshitta, so the combined evidence of the OG and Peshitta leads one to suspect the originality of these words. Next, Munnich observes that the beginning of v. 7 is almost identical to the ending of v. 2 in the MT. Hence, he surmises, correctly, that the repetition in v. 7 is a redactional element that was inserted to pick up from v. 2, because vv. 3-6 are a secondary insertion. Finally, he examines 4QDand and suggests that if one were to reconstruct the ending of v.6 to be rm[), then it is possible that v.7 began
[tywh hzx ybk#$m l(] in 4QDand, i.e. minus y#$)r ywzxw (line 6 above). Based on this analysis, the original Semitic Vorlage can be reconstructed from the OG with some support from 4QDand and without the secondary interpolation in vv. 3-6 in the MT. Munnich suggests:
tywh hzx ybk#$m l( (?yl) hlpn hdrxw ynn%lxdyw tyzx Mlx.
This reconstruction is all rather neat, but it also has two serious shortcomings and both are related to the reality that Munnich has limited textual support for his view. First, the aim of the reconstruction of 4QDand is to remove y#$)r ywzxw in order to support the omission of this syntagm from the OG Vorlage at the beginning of v. 7, but this reconstruction of 4QDand is completely hypothetical for these verses since there are only portions of a few words that are visible for all of vv. 6-7 (9-10)! Thus, Munnich would seek to offer evidence based on text that does not exist. Furthermore, this reconstruction still assumes that 4QDand contains the secondary insertion of vv. 3-6. Second, it is interesting that at no point in his discussion does Munnich note the evidence for his own reading of the OG. The fact of the matter is that e)pi th=j koi/thj mou should be seriously questioned. The phrase is contained in 967, but in 88 it is clearly indicated by an asterisk that it is an addition to the text and it agrees with Th. It is for these reasons that the phrase is relegated to the apparatus in the critical edition by Ziegler, who also had the reading of 967 for this passage. Particularly in chapters iv–v where there is almost no relationship between a presumed Vorlage for OG and the MT, which is also confirmed by the almost complete lack of agreement between the OG and Th for these chapters, the most reasonable presupposition for textual criticism is that any textual agreements between the OG and Th here are likely to be secondary. The OG reads perfectly well without e)pi th=j koi/thj mou, though the inclusion of the phrase does allow for one to posit a linear development from the Vorlage of the OG to the MT.
The case of the plus in the MT iv 3-6 is an interesting example of the way in which text-critical decisions and the understanding of the literary development of a book are intertwined when comparing the OG versions with the MT. Though Munnich is generally correct in his analysis of the insertion in MT iv 3-6(6-9) and that 967 witnesses to an older Vorlage, the divergent nature of the texts of chapters iv–vi make any attempts to draw narrow literary connections between the Vorlage of the OG and the MT extremely dubious. This point is reinforced by the rest of Munnich’s article in which he seeks to argue for the literary development of the MT from the Vorlage witnessed to by the OG, and by 967 in particular.
Munnich argues that a Maccabean reviser is responsible for the redaction of MT, and this is evident in the changes introduced in the Hebrew. For this view, Munnich also refers to an article in which he made the case that the interpolation in ii 13-23 was written in Hebrew. According to him, the aim of the editor was to centralize the collection around the figure of Daniel. Thus, this individual wrote the preface for the collection (i–ii 4a) and is responsible for the overall redaction of the book. With respect to the alignment of the original order of the chapters, witnessed to by 967, he states, “La difference des sequences s’explique plutot par le fait qu’on a affaire à l’adjonction—sans doute assez tardive dans le texte premasoretique—d’un episode assez vaguement lié à Daniel et situé sous Baltazar” (emphasis his). The loose connection of chapter v and vi to the rest of the tales is also noticeable because of the lack of the dating at the beginning of these chapters, which contrasts with the more precise dating in vii 1 and viii 1. Munnich compares these chapters to the other additions, and even refers to them as “additions de l’intérieur”. The rationale for this explanation is puzzling. Why would the later redactor of the MT have moved the chapters if the original version was successful? What is accomplished by interrupting the better chronology that is reflected in the arrangement in 967?
Munnich does not clearly state, but he seems to imply that since MT chapters vii and viii were associated with Belshazzar and were dated to the first (ch. vii) and third (ch. viii) years of his reign that they were originally put prior to Belshazzar’s death in chapter v. But, why was it necessary for this Maccabean editor to date these chapters to Belshazzar’s reign? Clearly, and Munnich would agree, at least chapter viii and the final form of chapter vii is a Maccabean composition that focuses on Antiochus. To this writer’s knowledge, there are no current scholars, apart from some who are more conservative, who would date the composition of chapter viii as early as the tales in chapters ii–vi(vii). Therefore, chapters vii and viii could have been dated to any reign. Under what conditions and for what reason would this later redactor order the material as Munnich supposes? The answer that Munnich appears to offer to this question is the linguistic connections that he draws between the OG in chapters iv and vii–viii that seem to focus on the person of Antiochus. There are several difficulties with this proposal. First, since chapter iv is concerned with the fall of a king(dom) would one not expect to discover some linguistic similarities between chapters vii and viii, whether in Hebrew or Greek? Second, the linguistic connections are based on the OG translation. If there were a pre-existing Greek version of chapter iv, editorial revisions due to the Maccabean crisis of that chapter would be expected and would presumably result in linguistic connections with chapters vii-viii. Therefore, even if the linguistic connections were compelling, why would that require that the chapters run consecutively? Third, the specific terminological links that Munnich offers to anchor chapter iv to vii and viii are not very strong and certainly do not sufficiently outweigh linguistic links that exist between other chapters, particularly if one was intentionally making a connection. The specific linguistic connections he makes are: iv 19 u(yw/qh sou h( kardi/a // viii 25 h( kardi/a au)tou= u(ywqh/setai; iv 19 e)chrh/mwsaj to\n oi!kon tou= qeou= tou= zw=ntoj // viii 11 kai\ to\ a3gion e)rhmwqh/setai; iv 21 kai\ o( u3yistoj kai\ oi( a1ggeloi au)tou= e)pi se\ katatre/xousin // vii 25 kai\ r(h/mata ei)j to\n u#yiston lalh/sei kai\ tou\j a(gi/ouj tou= u(yi/stou katatri/yei; iv 34 au)to\j . . . a)lloioi= kairou\j kai\ xro/nouj // vii 25 kai\ prosde/cetai a)lloiw=sai kairou\j kai\ no/mon. These examples are not particularly striking. If there were that close a connection between these chapters, would one not expect to find at least one of these phrases to be identical? For example, it would seem that v 2 a)nuyw/qh h( kardi/a au)tou= is just as close a parallel to iv 19(22) u(yw/qh sou h( kardi/a (see also the compound verb earlier in the verse) as the example in viii 25 and why does iv 21(24) not have oi( a#gioi instead of oi( a1ggeloi like vii 25? Given the composition of Daniel’s visions in chapter viii (and the final form of chapter vii) during the Maccabean period it is hard to conceive of a scenario in which a scribe would place those visions between the tales of iv and v, especially when chapter viii was written in Hebrew. There remains the possibility that chapters vii and viii could have been composed with the link to chapter iv in mind (which would be in keeping with Munnich’s proposal), but even this theory fails. It would require that one ignore the obvious formal links that vii and viii have with chapters ix-xii and the evidence for those specific links is wanting. Without any other recourse to explain the order of 967, it seems that the only reason to accept the alternative order is the dating in vii 1 and viii 1. Of course, if the only reason for the order is chronological, then an alternative Semitic Vorlage as suggested by Munnich and Lust is unnecessary.
Even if it is supposed that the order of 967 reflected a Semitic Vorlage, Munnich’s argument would fail because it cannot offer an explanation for the existing OG translation. Though he has not outlined a detailed description of his understanding of the growth of the book and how this related to the OG translation, Munnich has argued that 967 witnesses to an alternative Semitic literary edition that was basically complete for chapters i–xii plus the additions. As previously noted, he connects chapter iv with a Maccabean reviser of vii–viii as well as with chapters xi–xii and 1 Maccabees. He also connects the interpolation of ii 13-24 with the initial stages of the linking of the tales in ii-vi to the visions in vii–xii, the writing of an introductory chapter i–ii 4a, and the overall aim of this reviser, which was to centralize the collection around the figure of Daniel. Munnich is not clear on this point, but when would it have been possible for the Greek translation of his hypothetical Semitic Vorlage to have been accomplished? According to Munnich this hypothetical collection had the chapters arranged and presumably the different content for iv–vi (or iv, vii, viii) according to 967 and this original Semitic edition underwent further editing, which produced proto-MT. But, since the reordering of the chapters and other editorial revisions of the chapters in question must have occurred around the same time, how would the Greek translator have had access to the earlier arrangement and non-edited version? Finally, an earlier observation bears repeating. If this original edition had been compiled as is supposed, why would chapters vii and viii have been moved to a different place? The answer to this question would have to be that it was due to the fact that chapters vii–viii were placed with the rest of the visions because that is where they would have been seen to belong as part of the newer material. But, would they not have belonged with that material in the first place?
The alternative order of 967 presents several variables for any theory that seeks to advocate that it witnesses to an alternative Semitic Vorlage for the book of Daniel. One must give an account for both the alternative order of the chapters as well as the different content in OG iv–vi, while explaining how the visions in vii–viii, one of which was written in a different language!, could have been part of an alternative literary edition; and that this collection managed to be translated into Greek. Under what conceivable circumstances could these events have happened in light of the evidence? The final form of Daniel had to be around 164 BCE, the OG version is generally dated around the beginning of the first century BCE in Alexandria, and the Qumran manuscripts, which date from the same period, are proto-MT. The only real evidence to support the alternative order presented by 967 is 967 itself. Though there are some links between chapter iv and vii–viii, they are not particularly strong given the content of the chapters. Most importantly, the identification of the visions in vii 1 and viii 1 with Belshazzar has to become a reason for their location in the Vorlage to which 967 supposedly witnesses. But, if chronology is in any way an issue, then this becomes another blow to any theory advocating the originality of the order in 967. It is at least as likely that 967 changed the order to fix the chronology as it is that an editor has attributed these visions to the time of Belshazzar and created a collection such as we have in 967, particularly when 967 also has chapter 37 after 39 in Ezekiel. Would it not be much simpler to argue that the specific dating in vii 1 and viii 1 was inserted by the Maccabean editor in order to date these later visions prior to the events of chapter 5? Both the explanations by Lust and Munnich are based solely on the alternative order of one manuscript. They are unable to account for the significant evidence that unifies the translation of OG chapters iv–vi, the Qumran evidence that witnesses to the MT, and they have not offered a plausible scenario to support their hypotheses.
The OG Witnesses to an Earlier Collection of the Tales
The explanatory value of a hypothesis is determined by how well it explains the known data. Based on the extant evidence, it is unlikely that there is any hypothesis that attempts to posit an alternative Semitic edition based on the order of the chapters in 967 that can offer an adequate account for the rest of the data. In what follows a hypothesis is offered that outlines the growth and stages of the book of Daniel that includes the Greek versions. It would be inappropriate to identify it as a new proposal because it relies in most instances on what is current scholarly consensus. For the sake of simplicity it is presented in a series of steps.
1. In the beginning. As Lust and David argued, and in agreement with virtually all critical scholarship, the court tales in Daniel ii–vi owe their origins to a number of independent compositions that circulated in the Mesopotamian region prior to and during the Hellenistic period. The existence of the so-called additions to Daniel (The Prayer of Azariah and the Hymn of the Three Young Men, Susanna, and Bel and the Snake) and the manuscript fragments of other Daniel stories at Qumran argues that this is the case.
2. The First Collection. There was an original collection of the tales involving chapters (iii 31) iv–vi. Albertz and Wills conducted independent form-, source-, and redaction-critical investigations of chapters iv–vi in the MT and the OG and concluded that the OG reflects an older, Aramaic Vorlage. Wills convincingly demonstrates that the shared redactional characteristics in these chapters show these tales circulated as an independent collection . Albertz provided linguistic links between chapters iv–vi in the OG as evidence that these chapters reflect a different translator from the rest of the book, which I have confirmed in previous publications as well as in the present article. Thus, the best witness for the Vorlage of the original collection of tales is the OG.
3. The Addition of Chapters ii–iii. In all likelihood the next stage of the collection was the addition of a form of chapters ii–iii. Chapter ii did not include ii 13-23(24), 29-30, (40)41-43, 47, 49 at this stage and chapter iii did not have the additions.
4. A Maccabean Collection of Chapters(i)ii–vi(vii). The most disagreement about the early formation of Daniel would revolve around this stage. C. C. Torrey was one of the first to conceive and argue for an Aramaic collection of ii–vi and propose that chapter vii was intentionally written in Aramaic in order to connect the visions with the tales. Thus, a couple of stages in the chronological development may be combined here, but most scholars who have analyzed these chapters would agree that at least the core of chapter vii (without the references to the horns and their interpretations) was added to ii–vi before any of the other visions. An early version of chapter i may have been created in this period as an introduction to the collection, but the final version derives from the final redaction of the book. Whether it was just before or soon after the persecution by Antiochus, the addition of the core of chapter vii with the vision of four kingdoms was accompanied by the redaction of chapter ii to include the reference to the fourth kingdom, but without any references to “the toes” (vv. 41, 42).
5. An Egyptian Collection of Chapters(i)ii–vi(vii). Given the fact it seems most likely that chapters viii–xii and the final introduction were added to the core collection within a fairly short period of perhaps three years, the best way to account for the separate literary editions of Daniel is to assume a fairly major intervention that caused a separate development. The Maccabean crisis provides such an intervention and it fits the broad scholarly consensus about the development of the book. Prior to or during the Maccabean crisis a version of Daniel (i)ii–vi(vii) ended up in Alexandria. (If any part of chapter vii was not part of this original collection it would not significantly affect the overall argument.) Since the collection of chapters iv–vi had circulated independently for many decades it had already been translated into Greek. There may have been a preliminary translation of ii and iii as well, but given the process of revision and the effects of textual corruption it is difficult to determine an answer to this question. Though I discuss the nature of the OG translation and its relationship to the Theodotion version in chapters iv–vi and the rest of the book in other articles, for the time being it only needs to be noted that a translation of the rest of the core collection was made in Egypt sometime after 167 BCE so that there was a Greek version of chapters (i)ii–vi(vii).
6. The Final Redaction of proto-MT. The original Semitic version underwent extensive redaction after the desecration of the temple by Antiochus. Following the addition of chapter viii to the core collection, chapters x–xii were added along with interpolations in chapter ii (“the toes”) and vii (the “horns”) with their accompanying interpretations. Chapter ix was probably the last of the visions to be written. An introduction was either written at this point or the initial one was translated into Hebrew and revised to emphasize the role of the maskilim who were the final redactors (compare i 4, 17 with xi 33; xii 3). The maskilim are also responsible for additions to chapter ii 13-23, 27-29, 47, 49 and throughout chapters iv–vi where Daniel’s role as an inspired interpreter are pronounced.
7. Redaction of the Egyptian Collection of Chapters(i)ii–vi(vii). Separated from the extensive redaction going on to the proto-MT, the core collection underwent some redaction in Egypt. The main initiative was that The Prayer of Azariah and the Hymn of the Three Young Men was inserted into chapter iii. There would have been some other development within the version as well, but it is difficult to isolate. A corollary to the insertion of the addition to chapter iii may have been that Nebuchadnezzar’s lengthy confession and epistle were probably moved to the conclusion of chapter iv in order to separate it from the hymns. It was most likely in this setting that Susanna, and Bel and the Snake became attached more or less permanently to the Greek version of the book, but there must have been a long association with the core collection because the versions of these Daniel stories are quite different in Theodotion.
8. The Egyptian (OG) Version Gets Reacquainted with the proto-MT. At some later point the OG version added a translation of viii–xii to the core translation of (i)ii–vi(vii) and engaged in some revision of chapters i, ii, and vii, which had undergone changes in the Semitic version. For example, 967 has no references to the “toes” in ii 40-42. The OG of the core collection had already established itself as an independent version, which explains why chapters iv–vi remained substantially different. The so-called additions remained as an integral part of this tradition. Thus, Ulrich’s view that the OG and MT are two literary editions of Daniel that have undergone separate redaction from a common Vorlage is sound, but not that the OG was a uniform translation. It should also be pointed out that the earlier argument against Munnich’s reconstruction of OG iv 2-7 does not have to have any bearing on the reconstruction offered here. Obviously, the Vorlage of what became the OG could have been as Munnich argues where the Maccabean editor redacted the proto-MT. However, Munnich’s argument for a direct development literary from the Vorlage of the OG to the MT in this passage cannot be grounded in the texts.
9. The So-called Theodotion Version. Given the fact that the proto-MT had undergone extensive and separate redaction in Palestine apart from the Vorlage preserved in the OG, eventually a separate translation enterprise was undertaken in the same region. As Di Lella notes, this translation must have been prior to the common era. However, the forces that eventually created the MT could not have been too much at work, at least in the case of Daniel, because the Greek additions were retained in this new version. Susanna and Bel and the Snake must have had strong roots with the Daniel collection because they were part of the translation enterprise. The Prayer of Azariah and the Hymn of the Three Young Men is almost verbatim when one compares the two Greek translations, so it would seem that the new translator adopted that version from the OG.
10. The Transmission Process. During the course of the transmission process there was some scribal editing of all the versions and textual corruption of the Greek versions because they were now both accessible. The OG suffered the most from this process due to the dominance of the MT and the fact that the Theodotion version was based on a similar Vorlage. In the end, 967 is the sole surviving pre-hexaplaric witness to the OG, but along the way chapters vii–viii were inserted after chapter iv by a well-meaning scribe to fix the chronology. The issue of the relationship between the two Greek versions can only be mentioned here, but I have made the case that the Theodotion version is basically an independent translation. The issue is complicated because OG and Th do not reflect the same relationship consistently throughout the book. But, as a summary of what is available in other publications the relationship between the OG and Th is described below.
The Greek versions of chapters iv–vi are quite distinct compared to the remainder of the book and it is unlikely that any scholar would contest that view. Furthermore, a comparison of chapters i–iii in the Greek versions reveals that the vast majority of these chapters exhibit relatively minor points of agreement. It is only in chapters vii–xii that one generally finds the common agreements between the two Greek versions to run around 50% and portions, particularly in viii 5–x 21 and other isolated verses, where the agreement is considerably stronger. Some of the places where there is agreement are due to textual corruption, which can be demonstrated. As I have previously noted, “where OG exhibits a marked agreement with TH and formal equivalence to MT . . . we have every reason to suspect that TH readings have corrupted the OG”. These findings are consistent with the hypothesis that has been given in this paper to explain the origins of Daniel and the OG version. Generally, there is very little relationship between the two Greek versions in chapters i–vi(vii), because their origins are chronologically and geographically distinct and their core translations were based on different Vorlagen. In chapters iv–vi, in particular, I have clearly demonstrated that the reason for common readings between the OG and Th is the pervasive corruption of the OG with Th readings. Chapters (vii)viii–xii are based on very similar Vorlagen and were translated much closer in time. Thus, for portions of chapters vii–xii the common agreements between the two Greek versions are what would be expected from two independent translators working on the same text when one allows for the textual corruption that occurred during the transmission process. However, there are also portions of the Greek versions in these chapters where the verbal agreements between the texts are so strong that based on the available textual evidence it would appear that one is a revision of the other.
The Old Greek Translation of Daniel Chapters 4–6 and the Formation of the Book of Daniel
by Assoc. Prof. R. Timothy McLay
St. Stephen’s University, Calais, ME 04619
ABSTRACT: The content of the Old Greek translation of Daniel iv–vi is significantly different compared to the so-called Theodotion version and the Masoretic Text. In addition, the best witness to the Old Greek version (papyrus 967) has an alternative order for the chapters: chapters vii and viii intervene between iv and v. The proposals by J. Lust and O. Munnich that 967 preserves a more original version of the content and order of the chapters for the Vorlage of Daniel are critiqued. Additional linguistic evidence that supports the theory that the Old Greek translation of chapters iv–vi circulated together independently is also provided. Finally, a hypothesis for the growth and stages of the book of Daniel that includes an explanation for the origins of the Greek versions is outlined.
The Old Greek (OG) translation of Daniel chapters iv–vi presents the Septuagintalist with a host of interesting textual problems. While the Theodotion (Th ) version reads like a translation based on formal equivalence of a text that is very similar to our existing Semitic texts of Daniel, the OG exhibits little textual similarity to either Th or the MT. The vast differences between the versions of the chapters represented respectively by the OG and Th/MT are particularly striking when compared to the rest of the book where the Greek versions are both based on Vorlagen that are very similar to the MT. Scholars are agreed that the differing version of chapters iv–vi in the OG is based on a Semitic Vorlage, and Wills and Albertz have offered important contributions for understanding the differing literary redactions of Daniel for these chapters while both also argued for the priority of the OG version. Indeed, the essential link between textual and literary criticism that is increasingly acknowledged in biblical studies owes a debt of gratitude to the discovery of the ancient manuscripts in a cave in the Judean desert in 1947. For Daniel there are two Greek versions that can be compared to one another as well as to the Semitic version. Th is generally quite close to the MT throughout the book; therefore, the texts of the Greek versions for chapters iv–vi differ considerably.
In general, the basic story in Daniel iv–vi is the same in the versions transmitted respectively by the OG and MT/Th, but there are numerous differences in details. For example, chapter iv still narrates the story of Nebuchadnezzar’s madness, but his confession and the publication of his decree occurs in a much expanded form at the end of the chapter in vv. 34(37)-34c rather than the beginning, and there are other pluses to vv. 14(17), 19(22), 23-25(26-28), 28(31), 30(33). There are also significant minuses that involve vv. 20-22(23-25) when one compares the OG to MT/Th and there is no equivalent for iv 3-6(6-9). Chapter five recounts the mysterious writing on the wall, but the OG version includes an abbreviated version of the story as a preface and omits significant portions of vv. 3, 10-13 and has no equivalent for vv. 14-15, 18-22 and 24-25. Chapter six, in which Daniel is thrown into the lion’s den, is much closer in length in the different versions, but there are large pluses in the OG vv. 3(4), 5(6), 12a, 14(15), 17-18(18-19), 22(23) and minuses in 15(16), 23(24). Even where there are no pluses or minuses in these chapters the Greek versions share little relationship with regard to style, grammar, and, more importantly, vocabulary.
Though he has not given evidence to support his view, Eugene Ulrich has suggested that the OG translation for chapters i–xii “is of one piece”. This view presupposes that the alternative translation of chapters iv–vi stems from a Semitic Vorlage that was complete for chapters i–xii. Given the fact that the content of chapters iv–vi (the madness of Nebuchadnezzar, the writing on the wall, Daniel and the lion’s den) exhibits significant literary differences as a stand alone corpus compared to the rest of the book it is doubtful that one can make a decisive argument in this matter. However, a case can be made that the chapters do not originate with the same translator as the rest of the book. Albertz has argued that the differences between OG and MT/Th in iv–vi are due to the fact that the translator of chapters i–iii, vii–xii adopted the earlier “popular” translation of iv–vi into his translation because it exhibits a theological emphasis on monotheism, which is also detectable in iii 17. One of the problems with this argument is that the OG does not exhibit a concern for monotheism elsewhere, but the fact that there is a distinct emphasis on monotheism expressed in iv 34c![37], which is echoed by repeated statements that their God is “the living God” who reigns “in heaven” (see iv 23[26], 28[31], 34[37]; v 23), remains an argument that these chapters stem from a different hand. Furthermore, there are instances where the translation of vocabulary in chapters iv–vi was different than elsewhere in the book. An analysis of the translation of the vocabulary in these chapters provides additional evidence that supports a different translator (#n), ht), )(b, hlm, Nd( ). These examples are all based on different Greek translation equivalents for Semitic vocabulary occuring both inside and outside of Daniel iv–vi; therefore, they are weighty indicators of different translation hands at work. Another type of evidence would be to note the instances where certain Greek vocabulary is restricted to chapters iv–vi, regardless of the Semitic Vorlage. However, despite the fact that the evidence favors a separate translation for these three chapters, it might be imprudent to claim that the evidence alone is compelling. Though the combined evidence of the examples does exhibit distinct patterns of translation, the judgments for the first types of examples are based on the assumption that one can reconstruct the Vorlage for the vocabulary for OG based on the content and comparable material existing in the MT. Since the content of the stories is similar in OG and Th/MT, despite the clear differences, this assumption seems reasonable in most cases, whether the Vorlage for OG was Aramaic or the translation is based on cognate Hebrew words.
Though one may not claim that an overwhelming case for the distinct nature of the translation for OG chapters iv–vi has been made previously, it has not seemed to this writer that there are serious rival explanations. However, the publication of recent theories about the origins of Daniel are indicators that a more complete examination of the available evidence is in order. In particular, there have been several scholars who have argued for an earlier Vorlage for Daniel and that it had an alternative order for the chapters based on the evidence of the best witness to the OG text, papyrus 967. In 967 chapters vii and viii intervene between iv and v. The result is a somewhat smoother chronology because the result has the narrative of chapters i–iv situated in the reign of Nebuchadnezzar, vii–viii and v in the time of Belshazzar, vi and ix in the period of Darius, and x–xii are dated in the first year of Cyrus. The alternative order of the material in papyrus 967 raises additional questions about the literary development of Daniel and its relationship to what became the MT. How is the content of the OG and the different alignment of chapters iv–vi (in papyrus 967) in relationship to the MT and the TH version to be explained? A necessary consequence of arguing for the alternative order for chapters iv–vi has also meant that these scholars have challenged the view that OG chapters iv–vi represent a distinct and separate translation.
In what follows two different proposals that advocate that 967 preserves a more original version of the content and Vorlage for Daniel will be examined. More space will be devoted to the most recent proposal by Olivier Munnich, because he argues that 967 witnesses to an alternative literary edition that was complete for the book of Daniel and that one can demonstrate linear development from its Vorlage to the MT in some cases. Significant criticisms are raised against both attempts to posit a Semitic edition of Daniel with an alternative order of the chapters as in 967, and of Munnich’s argument for a direct literary development from the Vorlage of the OG to the MT. Following the examination of the proposals by Johan Lust and Munnich a hypothesis for the outlines of the growth and stages of the book of Daniel that includes the Greek versions will be outlined.
967 Witnesses to an Early Collection of Daniel Stories from (i)ii–viii
Whether chapters iv–vi were part of an integrated literary whole encompassing at least chapters i–xii (Ulrich does attempt to provide an explanation for the relationship of the deutero-canonical additions to the known Hebrew version) is fundamental for understanding the stages of growth and composition of Daniel. Typically, it is suggested that papyrus 967 has rearranged the order of the chapters in order to fix the chronology, but Johan Lust has argued for the priority of papyrus 967. One of the strengths of Lust’s argument is that it builds on the observation that originally the tales of Daniel in chapters ii–vi circulated as independent compositions. They seemed to have been part of an existing tradition of Daniel stories, which is supported by the additions that are known to us only in their Greek versions, as well as by the Daniel fragments and the Prayer of Nabonidus discovered at Qumran. One can readily agree with Lust that the OG, including 967, witnesses to an earlier Semitic collection of Danielic tales; however, his views regarding the priority of the order of the chapters in 967 are problematic.
Lust suggests that the pre-Maccabean collection of Daniel stories included a form of chapter vii and the tales were redacted together according to different patterns: one of which is witnessed to by 967. Chapter viii, like the rest of the Hebrew sections, was added later. There are two main issues, however, that Lust does not explain. First, there is no reason provided for the hypothetical ordering of the chapters according to OG. Lust correctly notes that the order of chapters ii–vii reflects conscious and deliberate editorial activity, but this does not in any way demonstrate dependence on the order witnessed by 967. Second, how does Lust account for chapter viii? According to Pablo David, who expanded on Lust’s views in his doctoral thesis, chapter viii was appended to chapter vii in this pre-Maccabean collection of Aramaic tales along with i–ii 4a, and this was later followed by x–xii and chapter ix. But, how likely is it that a chapter written in Hebrew was sandwiched between chapters written in Aramaic, particularly if this editor was already attaching an introduction in Hebrew? Chapter eight causes an even more significant problem for the Greek. David suggests that the original Aramaic collection of ii, iii, iv, vii, v, vi, Bel and the Dragon was translated into Greek and that the translation for these chapters was retained and used by the later translator for chapters i, viii, and ix–xii. How did the Greek translator know to put chapter viii after vii if these were separate editions? According to David, the Greek translator would have had to have knowledge of three separate collections: ii–vii plus Bel and the Dragon; i–viii plus Bel; i–xii (plus additions). This is possible, but is it likely, especially when there could not have been a great deal of time between the production of the chapters that would have been added to the original core? This scenario offers no explanation for the setting in which the OG was translated and how the OG managed to preserve an alternative Semitic order of the chapters.
967 Witnesses to an Earlier Literary Edition than the MT
Recently, a new proposal has been advanced by O. Munnich, which, similar to Lust, advocates the priority of the order of papyrus 967 as a witness to the original order of Daniel. In addition, in agreement with the views of Ulrich, Munnich believes that the OG witnesses to an alternative literary edition. Munnich is concerned to demonstrate literary development from the OG to MT, and to this end he does note several ways that chapters iv and v in the MT exhibit that it is a later redaction than the OG, particularly in its development of the role and character of Daniel. For example, Daniel is not mentioned in OG chapter iv until v. 15(18) compared to the MT in which he is described as “endowed with the a spirit of the holy gods” and “no mystery is too difficult” for him in iv 8-9(5-6). The intelligence and wisdom of Daniel are also emphasized in the MT in v 11-16. Munnich is also no doubt correct when he identifies later additions to MT iv 3-6 and v 3.
It is interesting to examine Munnich’s argument more closely, because it is an excellent example of the relationship between literary and textual criticism. In particular, most cases of different literary editions witnessed to by the MT and LXX exhibit a linear connection, i.e. one developed from the other. Munnich’s examples from chapter iv 3-6 and v 3 are part of his argument that seeks to demonstrate the redactional process from the Semitic Vorlage of the OG to the MT. Moreover, it is essential to note that Munnich is arguing for the priority of the order of the chapters according to 967 in this hypothetical original Vorlage as well. In what follows it is argued that Munnich is mistaken on both counts. Though the MT reflects later and more extensive redaction in chapters iv–vi than does the OG and the OG witnesses to an earlier Vorlage, Munnich has neither offered sufficient evidence to demonstrate a linear connection between the Vorlage of the OG and the MT nor a satisfactory argument for the priority of the order of the chapters according to 967. His views will be examined in two parts. First, the omission in OG iv 3-6, which Munnich offers as evidence for redactional development from the OG to the MT will be discussed. Second, the notion that the order of the chapters in 967 is a witness to an earlier Semitic edition of Daniel will be scrutinized in detail. Some of the same criticisms that can be brought against Munnich’s position would apply equally to Lust, especially when it comes to their common view of the originality of the order of the chapters in papyrus 967.
In his most recent study Munnich argues not only that iv 3-6 are evidence of later redaction in the MT, but that one can also reconstruct the editorial link between the MT and the OG 967. The OG and MT are presented in parallel alignment below. Lines 1-5 are v. 2 and 6-8 represent the beginning of v. 7.
OG iv 2(5), 7 (10) MT 4, 2(5).7(10)
1.e)nu/pnion ei}don tyzx Mlx
2.kai\ eu)labh/qhn ynn%lxdyw
3.kai\ fo/boj moi e)pe/pesen Nyrhrhw
4. e)pi th=j koi/thj mou
5.
ybk#$m-l(
ynn%lhby y#$)r ywzxw
6.
7.
8. e)ka/qeudon kai\ i)dou/ y#$)r ywzxw
ybk#$m-l(
wl)w tywh hzx
Munnich begins by noting that line 2 is not quite equivalent in the OG to the MT and offers a possible Vorlage presumed by lines 3-4 (ybk#$m-l( hlpn hdrxw), which are quite different. He also notes that lines 3-4 are absent from the Peshitta, so the combined evidence of the OG and Peshitta leads one to suspect the originality of these words. Next, Munnich observes that the beginning of v. 7 is almost identical to the ending of v. 2 in the MT. Hence, he surmises, correctly, that the repetition in v. 7 is a redactional element that was inserted to pick up from v. 2, because vv. 3-6 are a secondary insertion. Finally, he examines 4QDand and suggests that if one were to reconstruct the ending of v.6 to be rm[), then it is possible that v.7 began
[tywh hzx ybk#$m l(] in 4QDand, i.e. minus y#$)r ywzxw (line 6 above). Based on this analysis, the original Semitic Vorlage can be reconstructed from the OG with some support from 4QDand and without the secondary interpolation in vv. 3-6 in the MT. Munnich suggests:
tywh hzx ybk#$m l( (?yl) hlpn hdrxw ynn%lxdyw tyzx Mlx.
This reconstruction is all rather neat, but it also has two serious shortcomings and both are related to the reality that Munnich has limited textual support for his view. First, the aim of the reconstruction of 4QDand is to remove y#$)r ywzxw in order to support the omission of this syntagm from the OG Vorlage at the beginning of v. 7, but this reconstruction of 4QDand is completely hypothetical for these verses since there are only portions of a few words that are visible for all of vv. 6-7 (9-10)! Thus, Munnich would seek to offer evidence based on text that does not exist. Furthermore, this reconstruction still assumes that 4QDand contains the secondary insertion of vv. 3-6. Second, it is interesting that at no point in his discussion does Munnich note the evidence for his own reading of the OG. The fact of the matter is that e)pi th=j koi/thj mou should be seriously questioned. The phrase is contained in 967, but in 88 it is clearly indicated by an asterisk that it is an addition to the text and it agrees with Th. It is for these reasons that the phrase is relegated to the apparatus in the critical edition by Ziegler, who also had the reading of 967 for this passage. Particularly in chapters iv–v where there is almost no relationship between a presumed Vorlage for OG and the MT, which is also confirmed by the almost complete lack of agreement between the OG and Th for these chapters, the most reasonable presupposition for textual criticism is that any textual agreements between the OG and Th here are likely to be secondary. The OG reads perfectly well without e)pi th=j koi/thj mou, though the inclusion of the phrase does allow for one to posit a linear development from the Vorlage of the OG to the MT.
The case of the plus in the MT iv 3-6 is an interesting example of the way in which text-critical decisions and the understanding of the literary development of a book are intertwined when comparing the OG versions with the MT. Though Munnich is generally correct in his analysis of the insertion in MT iv 3-6(6-9) and that 967 witnesses to an older Vorlage, the divergent nature of the texts of chapters iv–vi make any attempts to draw narrow literary connections between the Vorlage of the OG and the MT extremely dubious. This point is reinforced by the rest of Munnich’s article in which he seeks to argue for the literary development of the MT from the Vorlage witnessed to by the OG, and by 967 in particular.
Munnich argues that a Maccabean reviser is responsible for the redaction of MT, and this is evident in the changes introduced in the Hebrew. For this view, Munnich also refers to an article in which he made the case that the interpolation in ii 13-23 was written in Hebrew. According to him, the aim of the editor was to centralize the collection around the figure of Daniel. Thus, this individual wrote the preface for the collection (i–ii 4a) and is responsible for the overall redaction of the book. With respect to the alignment of the original order of the chapters, witnessed to by 967, he states, “La difference des sequences s’explique plutot par le fait qu’on a affaire à l’adjonction—sans doute assez tardive dans le texte premasoretique—d’un episode assez vaguement lié à Daniel et situé sous Baltazar” (emphasis his). The loose connection of chapter v and vi to the rest of the tales is also noticeable because of the lack of the dating at the beginning of these chapters, which contrasts with the more precise dating in vii 1 and viii 1. Munnich compares these chapters to the other additions, and even refers to them as “additions de l’intérieur”. The rationale for this explanation is puzzling. Why would the later redactor of the MT have moved the chapters if the original version was successful? What is accomplished by interrupting the better chronology that is reflected in the arrangement in 967?
Munnich does not clearly state, but he seems to imply that since MT chapters vii and viii were associated with Belshazzar and were dated to the first (ch. vii) and third (ch. viii) years of his reign that they were originally put prior to Belshazzar’s death in chapter v. But, why was it necessary for this Maccabean editor to date these chapters to Belshazzar’s reign? Clearly, and Munnich would agree, at least chapter viii and the final form of chapter vii is a Maccabean composition that focuses on Antiochus. To this writer’s knowledge, there are no current scholars, apart from some who are more conservative, who would date the composition of chapter viii as early as the tales in chapters ii–vi(vii). Therefore, chapters vii and viii could have been dated to any reign. Under what conditions and for what reason would this later redactor order the material as Munnich supposes? The answer that Munnich appears to offer to this question is the linguistic connections that he draws between the OG in chapters iv and vii–viii that seem to focus on the person of Antiochus. There are several difficulties with this proposal. First, since chapter iv is concerned with the fall of a king(dom) would one not expect to discover some linguistic similarities between chapters vii and viii, whether in Hebrew or Greek? Second, the linguistic connections are based on the OG translation. If there were a pre-existing Greek version of chapter iv, editorial revisions due to the Maccabean crisis of that chapter would be expected and would presumably result in linguistic connections with chapters vii-viii. Therefore, even if the linguistic connections were compelling, why would that require that the chapters run consecutively? Third, the specific terminological links that Munnich offers to anchor chapter iv to vii and viii are not very strong and certainly do not sufficiently outweigh linguistic links that exist between other chapters, particularly if one was intentionally making a connection. The specific linguistic connections he makes are: iv 19 u(yw/qh sou h( kardi/a // viii 25 h( kardi/a au)tou= u(ywqh/setai; iv 19 e)chrh/mwsaj to\n oi!kon tou= qeou= tou= zw=ntoj // viii 11 kai\ to\ a3gion e)rhmwqh/setai; iv 21 kai\ o( u3yistoj kai\ oi( a1ggeloi au)tou= e)pi se\ katatre/xousin // vii 25 kai\ r(h/mata ei)j to\n u#yiston lalh/sei kai\ tou\j a(gi/ouj tou= u(yi/stou katatri/yei; iv 34 au)to\j . . . a)lloioi= kairou\j kai\ xro/nouj // vii 25 kai\ prosde/cetai a)lloiw=sai kairou\j kai\ no/mon. These examples are not particularly striking. If there were that close a connection between these chapters, would one not expect to find at least one of these phrases to be identical? For example, it would seem that v 2 a)nuyw/qh h( kardi/a au)tou= is just as close a parallel to iv 19(22) u(yw/qh sou h( kardi/a (see also the compound verb earlier in the verse) as the example in viii 25 and why does iv 21(24) not have oi( a#gioi instead of oi( a1ggeloi like vii 25? Given the composition of Daniel’s visions in chapter viii (and the final form of chapter vii) during the Maccabean period it is hard to conceive of a scenario in which a scribe would place those visions between the tales of iv and v, especially when chapter viii was written in Hebrew. There remains the possibility that chapters vii and viii could have been composed with the link to chapter iv in mind (which would be in keeping with Munnich’s proposal), but even this theory fails. It would require that one ignore the obvious formal links that vii and viii have with chapters ix-xii and the evidence for those specific links is wanting. Without any other recourse to explain the order of 967, it seems that the only reason to accept the alternative order is the dating in vii 1 and viii 1. Of course, if the only reason for the order is chronological, then an alternative Semitic Vorlage as suggested by Munnich and Lust is unnecessary.
Even if it is supposed that the order of 967 reflected a Semitic Vorlage, Munnich’s argument would fail because it cannot offer an explanation for the existing OG translation. Though he has not outlined a detailed description of his understanding of the growth of the book and how this related to the OG translation, Munnich has argued that 967 witnesses to an alternative Semitic literary edition that was basically complete for chapters i–xii plus the additions. As previously noted, he connects chapter iv with a Maccabean reviser of vii–viii as well as with chapters xi–xii and 1 Maccabees. He also connects the interpolation of ii 13-24 with the initial stages of the linking of the tales in ii-vi to the visions in vii–xii, the writing of an introductory chapter i–ii 4a, and the overall aim of this reviser, which was to centralize the collection around the figure of Daniel. Munnich is not clear on this point, but when would it have been possible for the Greek translation of his hypothetical Semitic Vorlage to have been accomplished? According to Munnich this hypothetical collection had the chapters arranged and presumably the different content for iv–vi (or iv, vii, viii) according to 967 and this original Semitic edition underwent further editing, which produced proto-MT. But, since the reordering of the chapters and other editorial revisions of the chapters in question must have occurred around the same time, how would the Greek translator have had access to the earlier arrangement and non-edited version? Finally, an earlier observation bears repeating. If this original edition had been compiled as is supposed, why would chapters vii and viii have been moved to a different place? The answer to this question would have to be that it was due to the fact that chapters vii–viii were placed with the rest of the visions because that is where they would have been seen to belong as part of the newer material. But, would they not have belonged with that material in the first place?
The alternative order of 967 presents several variables for any theory that seeks to advocate that it witnesses to an alternative Semitic Vorlage for the book of Daniel. One must give an account for both the alternative order of the chapters as well as the different content in OG iv–vi, while explaining how the visions in vii–viii, one of which was written in a different language!, could have been part of an alternative literary edition; and that this collection managed to be translated into Greek. Under what conceivable circumstances could these events have happened in light of the evidence? The final form of Daniel had to be around 164 BCE, the OG version is generally dated around the beginning of the first century BCE in Alexandria, and the Qumran manuscripts, which date from the same period, are proto-MT. The only real evidence to support the alternative order presented by 967 is 967 itself. Though there are some links between chapter iv and vii–viii, they are not particularly strong given the content of the chapters. Most importantly, the identification of the visions in vii 1 and viii 1 with Belshazzar has to become a reason for their location in the Vorlage to which 967 supposedly witnesses. But, if chronology is in any way an issue, then this becomes another blow to any theory advocating the originality of the order in 967. It is at least as likely that 967 changed the order to fix the chronology as it is that an editor has attributed these visions to the time of Belshazzar and created a collection such as we have in 967, particularly when 967 also has chapter 37 after 39 in Ezekiel. Would it not be much simpler to argue that the specific dating in vii 1 and viii 1 was inserted by the Maccabean editor in order to date these later visions prior to the events of chapter 5? Both the explanations by Lust and Munnich are based solely on the alternative order of one manuscript. They are unable to account for the significant evidence that unifies the translation of OG chapters iv–vi, the Qumran evidence that witnesses to the MT, and they have not offered a plausible scenario to support their hypotheses.
The OG Witnesses to an Earlier Collection of the Tales
The explanatory value of a hypothesis is determined by how well it explains the known data. Based on the extant evidence, it is unlikely that there is any hypothesis that attempts to posit an alternative Semitic edition based on the order of the chapters in 967 that can offer an adequate account for the rest of the data. In what follows a hypothesis is offered that outlines the growth and stages of the book of Daniel that includes the Greek versions. It would be inappropriate to identify it as a new proposal because it relies in most instances on what is current scholarly consensus. For the sake of simplicity it is presented in a series of steps.
1. In the beginning. As Lust and David argued, and in agreement with virtually all critical scholarship, the court tales in Daniel ii–vi owe their origins to a number of independent compositions that circulated in the Mesopotamian region prior to and during the Hellenistic period. The existence of the so-called additions to Daniel (The Prayer of Azariah and the Hymn of the Three Young Men, Susanna, and Bel and the Snake) and the manuscript fragments of other Daniel stories at Qumran argues that this is the case.
2. The First Collection. There was an original collection of the tales involving chapters (iii 31) iv–vi. Albertz and Wills conducted independent form-, source-, and redaction-critical investigations of chapters iv–vi in the MT and the OG and concluded that the OG reflects an older, Aramaic Vorlage. Wills convincingly demonstrates that the shared redactional characteristics in these chapters show these tales circulated as an independent collection . Albertz provided linguistic links between chapters iv–vi in the OG as evidence that these chapters reflect a different translator from the rest of the book, which I have confirmed in previous publications as well as in the present article. Thus, the best witness for the Vorlage of the original collection of tales is the OG.
3. The Addition of Chapters ii–iii. In all likelihood the next stage of the collection was the addition of a form of chapters ii–iii. Chapter ii did not include ii 13-23(24), 29-30, (40)41-43, 47, 49 at this stage and chapter iii did not have the additions.
4. A Maccabean Collection of Chapters(i)ii–vi(vii). The most disagreement about the early formation of Daniel would revolve around this stage. C. C. Torrey was one of the first to conceive and argue for an Aramaic collection of ii–vi and propose that chapter vii was intentionally written in Aramaic in order to connect the visions with the tales. Thus, a couple of stages in the chronological development may be combined here, but most scholars who have analyzed these chapters would agree that at least the core of chapter vii (without the references to the horns and their interpretations) was added to ii–vi before any of the other visions. An early version of chapter i may have been created in this period as an introduction to the collection, but the final version derives from the final redaction of the book. Whether it was just before or soon after the persecution by Antiochus, the addition of the core of chapter vii with the vision of four kingdoms was accompanied by the redaction of chapter ii to include the reference to the fourth kingdom, but without any references to “the toes” (vv. 41, 42).
5. An Egyptian Collection of Chapters(i)ii–vi(vii). Given the fact it seems most likely that chapters viii–xii and the final introduction were added to the core collection within a fairly short period of perhaps three years, the best way to account for the separate literary editions of Daniel is to assume a fairly major intervention that caused a separate development. The Maccabean crisis provides such an intervention and it fits the broad scholarly consensus about the development of the book. Prior to or during the Maccabean crisis a version of Daniel (i)ii–vi(vii) ended up in Alexandria. (If any part of chapter vii was not part of this original collection it would not significantly affect the overall argument.) Since the collection of chapters iv–vi had circulated independently for many decades it had already been translated into Greek. There may have been a preliminary translation of ii and iii as well, but given the process of revision and the effects of textual corruption it is difficult to determine an answer to this question. Though I discuss the nature of the OG translation and its relationship to the Theodotion version in chapters iv–vi and the rest of the book in other articles, for the time being it only needs to be noted that a translation of the rest of the core collection was made in Egypt sometime after 167 BCE so that there was a Greek version of chapters (i)ii–vi(vii).
6. The Final Redaction of proto-MT. The original Semitic version underwent extensive redaction after the desecration of the temple by Antiochus. Following the addition of chapter viii to the core collection, chapters x–xii were added along with interpolations in chapter ii (“the toes”) and vii (the “horns”) with their accompanying interpretations. Chapter ix was probably the last of the visions to be written. An introduction was either written at this point or the initial one was translated into Hebrew and revised to emphasize the role of the maskilim who were the final redactors (compare i 4, 17 with xi 33; xii 3). The maskilim are also responsible for additions to chapter ii 13-23, 27-29, 47, 49 and throughout chapters iv–vi where Daniel’s role as an inspired interpreter are pronounced.
7. Redaction of the Egyptian Collection of Chapters(i)ii–vi(vii). Separated from the extensive redaction going on to the proto-MT, the core collection underwent some redaction in Egypt. The main initiative was that The Prayer of Azariah and the Hymn of the Three Young Men was inserted into chapter iii. There would have been some other development within the version as well, but it is difficult to isolate. A corollary to the insertion of the addition to chapter iii may have been that Nebuchadnezzar’s lengthy confession and epistle were probably moved to the conclusion of chapter iv in order to separate it from the hymns. It was most likely in this setting that Susanna, and Bel and the Snake became attached more or less permanently to the Greek version of the book, but there must have been a long association with the core collection because the versions of these Daniel stories are quite different in Theodotion.
8. The Egyptian (OG) Version Gets Reacquainted with the proto-MT. At some later point the OG version added a translation of viii–xii to the core translation of (i)ii–vi(vii) and engaged in some revision of chapters i, ii, and vii, which had undergone changes in the Semitic version. For example, 967 has no references to the “toes” in ii 40-42. The OG of the core collection had already established itself as an independent version, which explains why chapters iv–vi remained substantially different. The so-called additions remained as an integral part of this tradition. Thus, Ulrich’s view that the OG and MT are two literary editions of Daniel that have undergone separate redaction from a common Vorlage is sound, but not that the OG was a uniform translation. It should also be pointed out that the earlier argument against Munnich’s reconstruction of OG iv 2-7 does not have to have any bearing on the reconstruction offered here. Obviously, the Vorlage of what became the OG could have been as Munnich argues where the Maccabean editor redacted the proto-MT. However, Munnich’s argument for a direct development literary from the Vorlage of the OG to the MT in this passage cannot be grounded in the texts.
9. The So-called Theodotion Version. Given the fact that the proto-MT had undergone extensive and separate redaction in Palestine apart from the Vorlage preserved in the OG, eventually a separate translation enterprise was undertaken in the same region. As Di Lella notes, this translation must have been prior to the common era. However, the forces that eventually created the MT could not have been too much at work, at least in the case of Daniel, because the Greek additions were retained in this new version. Susanna and Bel and the Snake must have had strong roots with the Daniel collection because they were part of the translation enterprise. The Prayer of Azariah and the Hymn of the Three Young Men is almost verbatim when one compares the two Greek translations, so it would seem that the new translator adopted that version from the OG.
10. The Transmission Process. During the course of the transmission process there was some scribal editing of all the versions and textual corruption of the Greek versions because they were now both accessible. The OG suffered the most from this process due to the dominance of the MT and the fact that the Theodotion version was based on a similar Vorlage. In the end, 967 is the sole surviving pre-hexaplaric witness to the OG, but along the way chapters vii–viii were inserted after chapter iv by a well-meaning scribe to fix the chronology. The issue of the relationship between the two Greek versions can only be mentioned here, but I have made the case that the Theodotion version is basically an independent translation. The issue is complicated because OG and Th do not reflect the same relationship consistently throughout the book. But, as a summary of what is available in other publications the relationship between the OG and Th is described below.
The Greek versions of chapters iv–vi are quite distinct compared to the remainder of the book and it is unlikely that any scholar would contest that view. Furthermore, a comparison of chapters i–iii in the Greek versions reveals that the vast majority of these chapters exhibit relatively minor points of agreement. It is only in chapters vii–xii that one generally finds the common agreements between the two Greek versions to run around 50% and portions, particularly in viii 5–x 21 and other isolated verses, where the agreement is considerably stronger. Some of the places where there is agreement are due to textual corruption, which can be demonstrated. As I have previously noted, “where OG exhibits a marked agreement with TH and formal equivalence to MT . . . we have every reason to suspect that TH readings have corrupted the OG”. These findings are consistent with the hypothesis that has been given in this paper to explain the origins of Daniel and the OG version. Generally, there is very little relationship between the two Greek versions in chapters i–vi(vii), because their origins are chronologically and geographically distinct and their core translations were based on different Vorlagen. In chapters iv–vi, in particular, I have clearly demonstrated that the reason for common readings between the OG and Th is the pervasive corruption of the OG with Th readings. Chapters (vii)viii–xii are based on very similar Vorlagen and were translated much closer in time. Thus, for portions of chapters vii–xii the common agreements between the two Greek versions are what would be expected from two independent translators working on the same text when one allows for the textual corruption that occurred during the transmission process. However, there are also portions of the Greek versions in these chapters where the verbal agreements between the texts are so strong that based on the available textual evidence it would appear that one is a revision of the other.
