Before Bomberg: The Case of The Targum of Job in the Rabbinic Bible and the Solger Codex (MS Nьrnberg) ¹

Of all the noteworthy dates in the long and colorful history of religious publishing in the West, perhaps one of the most well-known, is that of March 13, 1517. This was of course the date on which, according to the traditional account, a professor of theology by the name of Luther presented his 95 theses to the world via the Wittenburg castle church door. But this, in several respects, "unorthodox" publication was not the only significant publishing event of 1517. Later that same year, across the Alps to the south in Venice, Daniel Bomberg, a far less famous though certainly more professional publisher, published a large volume which undoubtedly for targumists constituted a more significant landmark. The Bomberg Bible of 1517 was the first edition of what was to become throughout its many subsequent printings the standard version of the Hebrew Scriptures within the Jewish tradition.

While Daniel Bomberg would come to be well-known for his great editions of Scripture and Talmud, eventually succeeding his rival Soncino as the pre-eminent publisher of Hebrew books in Italy, it is interesting to note that his first publication was in fact a Latin translation of the Psalms executed by an apostate Jew, Felix de Prato ². Apparently it was de Prato himself who had first interested Bomberg in the business of books and when the time came for work to begin on his edition of the Rabbinic Bible, it is hardly surprising that Bomberg turned over responsibility for its assembly and editing to his former translator. But in contrast to the comparative wealth of information we possess regarding the circumstances surrounding its publication, we possess very little explicit information regarding the sources which de Prato used for his edition of the targum to the Hebrew text. In the light of such a lack, the mystery of de Prato's sources can be solved only through the study of the relationship between the various extant targum witnesses and the targum text presented by the Bomberg Bible.

While recent research has shown the importance of the targum text preserved in the Solger Codex (MS Nьrnberg) in any attempt to solve the puzzle of the pre-history of the Rabbinic Bible's targum text, many pieces of this puzzle remain as yet unexamined ³. The primary focus of the present paper is an examination of one of these pieces – the rabbinic targum of Job – in the hopes of further clarifying the relationship between the targum texts of MS Nьrnberg and the 1517 Bomberg Bible.

This study is by no means the first to take up the question of the manuscript witnesses to the Rabbinic Targum of Job ⁴. In fact, over the past 20 years it is fair to say that the question has attracted a sort of international attention. While prior studies of the targum were primarily based on the text reproduced in the printed editions, it was the Israeli scholar, Raphael Weiss who was the first to make extensive use of the manuscripts in his doctoral work on the rabbinic targum of Job undertaken during the 1970's ⁵. However, while Weiss' work clearly showed the need for a critical edition of the targum, the initiative for such a project shifted to Europe where the Spanish researcher F.J. Fernбndez Vallina published his edition in 1980 ⁶. The latter's El Targum de Job was indeed a substantial undertaking and included many important witnesses, however it was not comprehensive in its coverage of the known manuscripts and thus toward the end of the 1980's the centre of study shifted yet again – this time to Britain where in 1989 David Stec completed his introduction and critical edition of the Targum of Job ⁷.

The current stemmatological position with regard to the targum of Job is, in essence, a refinement of the general outline worked out first by Weiss. He arranged the manuscripts at his disposal into general divisions which roughly correspond to the four groupings arrived at later by Fernandez Vallina. Stec included an additional six witnesses in comparison to the nine which had been used by Vallina, but his conclusions confirmed Vallina"s work (See chart 1) ⁸.

Before going on to define the more particular relationship between Nьrnberg and the text preserved in the Bomberg Bible, it seems wise to first locate it within the above stemmatological framework ⁹.

    1. MS Nьrnberg in relation to Group 3

What light does the collation of Nьrnberg ¹⁰ shed on Group 3 texts? Stec's findings suggest that while there is no physical break of any sort in the MS 116, the targum text it presents appears to be a type of composite text ¹¹. While the first thirteen chapters of its text bear a strong affinity to that of the Bomberg Bible and thus belong to Group 3, this is not the case with the latter two thirds (chapters 14-42) which appear to belong to Group 2. In addition to this lack of affinity, Stec presents his evidence for the grouping of the first section of the MS 116 and Bomberg under three headings ¹².

Unique verse patterns ¹³

Stec notes that in chapter 13,7 Bomberg and MS 116 share an otherwise unique targum pattern (allowing for minor variations). The targum pattern found in Nьrnberg for 13,7 is the same as the one found only in the Group 3 texts and is in fact identical with that found in Bomberg. The version found in MS 116 diverges slightly from the other two. Our first clue that Nьrnberg presents a unified Group 3 text as opposed to a composite text like MS 116 lies in the fact that the remainder of the targum patterns unique to Group 3, beginning with the first in chapter 24, are shared by Bomberg and Nьrnberg, but not by MS 116 ¹⁴.

Unique alternative targumim

Stec also brings forward otherwise unattested alternative targumim as evidence of a close relationship between Bomberg and MS 116. One of the alternative targumim to Job 12,6 was previously attested to only by Bomberg and MS 116. These two witnesses have been joined by the Nьrnberg. While the three versions clearly represent the one and the same alternative targum, again MS 116 shows some spelling variation, while Nьrnberg and Bomberg are entirely identical.

As for the latter section (chapters 14–42), where Bomberg was formerly the lone witness to an an alternative targum to 38,25, Nьrnberg now provides an additional witness (they are the only two texts which preserve 3 as opposed to 2 targumim to this verse).

Thus in the case of both the above categories, the collation of Nьrnberg tends to corroborate Stec's conclusions about the lack of affinity between MS 116 14–42 and Group 3. In addition, the evidence from Nьrnberg 14–42 strongly suggests that unlike MS 116, the entire text of Nьrnberg belongs with Bomberg in Group 3.

Shared readings

The third category of evidence for a close relationship between Bomberg and MS 116 (1-13) comes in the form of otherwise unattested readings shared by the two manuscripts. It is this type of evidence which leads Stec to conclude that the break in MS 116 comes somewhere near the end of chapter 13 and the beginning of chapter 14 ¹⁵. He provides 93 individual readings which link the first 13 chapters of MS 116 with Bomberg ¹⁶. While a simple listing of readings in common between only 2 witnesses may be rather deceptive in that it tends to gloss over any lack of agreement and undoubtedly fails to do justice to the entire network of stemmatological connections, it is perhaps to be admitted as a general indicator of affinity. While the following treatment of individual readings is the result of a slightly different approach, it too may be subject to the same criticism.

In chapter one, (See Chart 2) we find eight readings otherwise unattested but shared by all three Group 3 texts ¹⁷ Nьrnberg and Bomberg share thirteen readings not found in any other MS including MS 116 ¹⁸, while Nьrnberg and MS 116 share only three unique readings ¹⁹. Thus already in the first chapter we see that in terms of rough indicators, there seems to be a general affinity (as befits texts of the same group) between all three texts but that Nьrnberg and Bomberg have considerably more in common than do Nьrnberg and MS 116.

In the rather short chapter thirteen, the supposed last chapter for which we have three Group 3 texts, roughly the same proportions are obtained. Nьrnberg, Bomberg and MS 116 share four readings unattested by any other MS ²⁰, while Bomberg and Nьrnberg in this chapter share a total of seven otherwise unattested readings ²¹. Nьrnberg and MS 116 do not in fact share any unique readings in chapter 13 ²².

As shown by Chart 2, chapter fourteen presents a rather different picture from the trend established in the first 13 chapters of the Group 3 texts. While the shared readings of Nьrnberg and Bomberg continue at a relatively high level (varying in absolute terms in proportion to the length of a given chapter), both the Group 3 readings and those in common between Nьrnberg and MS 116 drop dramatically from chapter 14 onwards. This fits well with Stec's observations regarding the lack of affinity between the latter two thirds of MS 116 and Group 3.

Thus in terms of locating Nьrnberg within the existing stemmatological framework on the criteria utilised by Stec it is clear that Nьrnberg stands beside Bomberg as the only other witness to Group 3 preserved in its entirety. With respect to its more specific affinity within Group 3, it seems safe to conclude that it shares far more with the text of Bomberg that it does with the Group 3 text preserved by the first 13 chapters of the MS 116.

The collation of Nьrnberg also supports the observation that in comparison to the other textual traditions, Group 3 texts have a tendency to avoid ty + pronominal suffix, preferring to represent the object through the use of a verbal suffix ²³. However, the newest member of Group 3 is of less use in confirming or contesting Stec's cautious suggestion that Group 3 may belong to a Sephardi tradition. As Nьrnberg is written in an Ashkenazic square script, Group 3 remains without a Sephardic witness in terms of paleography ²⁴. And while Nьrnberg does share many of the characteristics which Stec finds in common between Group 2 (a Sephardi tradition from Northern Spain and Africa) and Group 3, the additional witness, in and of itself, neither strengthens nor weakens the linguistic argument ²⁵.

2. The Case for Bomberg's Dependence on MS Nьrnberg

For the purposes of this enquiry, it is a happy coincidence that in the case of many verses, phrases and even individual words, the Targum of Job displays the phenomenon of multiple translation ³¹. While Bomberg, as a printed edition, has incorporated these alternative targumim into the body of its text, designating them as such through its use of an appropriate formula ()"t), some manuscripts (MSS Nьrnberg, Paris Bib. Nat. 17., Wroclaw Bib. Uniw. 1106) also present targumic material (under various rubrics) in the margins. In fact, the location and order of these alternative targumim constitute a further plank in the case for dependence of Bomberg on Nьrnberg.

While most MSS order alternative targumim in generally the same fashion (ie., T2 in one MS corresponds to T2 in other MSS) Bomberg diverges from this tendency frequently, often reversing the order of the various targumim which it includes for a given verse. For example, in 14,22, where all MSS but Bomberg have a common order. Bomberg reverses the order giving T2 first, followed by T1. Stec suggests that the most reasonable explanation of the idiosyncratic position (and therefore ordering) of alternative targumim is that they were originally written in the margin of some manuscripts and only later imported into the body of a subsequent text during the copying of such manuscripts ³².

Nьrnberg seems to provide support for this suggestion. Because no alternative material in Nьrnberg is contained within the body of the text itself, alternative words, phrases or targumim are given in the margins. These alternative textual materials must then in some fashion or another, be connected with their corresponding locations in the base text through the use of a reference system (usually identical marks at the location and in the margin) The relationship between this reference system (and the alternative targumim it serves to incorporate) and the idiosyncratic ordering of Bomberg's alternative targumim are extremely suggestive of a close link between the two texts. In Nьrnberg, the targum and the Hebrew text of Job alternate verse by verse. The evidence seems to suggest that Bomberg's idiosyncratic ordering may be best explained by the fact that de Prato in his edited Bomberg text, almost always includes the alternative targumim in the margin of Nьrnberg following the targum offered in the body of Nьrnberg ³³. It was surely neither incidental nor accidental that as a rule, when Nьrnberg Mg does contain an alternative (second) Aramaic translation, the marginal hand of Nьrnberg often provides an insertion mark following the targum which is in the body of the text ³⁴. In the light of this general practice, it is not surprising that de Prato's approach produced the "correct" order of alternative targumim in many cases ³⁵. But there are many others where the following of this default procedure resulted in Bomberg having a radically different ordering of alternative targumim ³⁶.

For example, in chapter 36, while a significant proportion of the witnesses (12) preserve three targumim to verse 33, the order in which these three targum are preserved in Bomberg is unique. The targum which most witnesses cite first (T1) is in fact cited by Bomberg as the second targum, while T2 corresponds to the third targum preserved in Bomberg. To complete the confusion, Bomberg gives T3 as its first targum. Nьrnberg and de Prato's reading of it seems to provide the answer to the question of Bomberg's mysterious ordering. (See Illustr. 1) Nьrnberg preserves T3 in the body of its text. De Prato, accordingly includes it in Bomberg as his first targum. The remaining two targumim (T1 and T2) are preserved by Nьrnberg in its margin in the same order and are therefore included by de Prato in the "proper" sequence. This then would seem to explain the unique order of the alternative targumim in Bomberg's text of Job 36,33. While it is undoubtedly possible that Bomberg's tendency toward unique ordering of its targumim could have an alternative explanation, it is suggested that in terms of its ability to account for every case of Bomberg's irregular ordering, the above hypothesis is a very reasonable one.

The use of the term reasonable at this juncture is not incidental. While the combined force of the above evidence is powerfully suggestive, in terms of proving a case of dependence of one text on another, shared readings, and even the similar arrangement of multiple targumim can in fact only be admitted as circumstantial evidence. In other words, while this type of evidence establishes emphatically the link between the two texts, it provides no means for adjudicating the case of dependence. For example, we might easily imagine a case in which the MS tradition which underlies Bomberg (it seems reasonable to think it was a MS of some importance) had been in existence for centuries before de Prato encountered it. In this case, Nьrnberg could have been been corrected and supplemented by this "proto-Bomberg" MS, to which it may have had some affinity anyway. Insofar as the shared readings and even otherwise unattested ordering of alternative targumim might also fit this set of circumstances, we must look for evidence of a decisive nature before a verdict may be reached regarding the direction of dependence. This type of decisive evidence is often found in cases where the copyist has made a mistake of interpretation in which later observers may, so to speak, "catch him in the act" ³⁷.

Further evidence

The relationship between the marginal material of Nьrnberg and the text preserved in Bomberg sheds light not only on Bomberg's ordering of alternative targumim, but also provides us with the means to explain an even more distinctive feature found in Bomberg — that of dislocated targumim.

In chapter 24, Nьrnberg, along with virtually all other witnesses, preserves three targumim to verse 19 (See Illustr. 2). While at first glance, Bomberg seems to provide only two targumim (T1 and T2) to this verse, when the following verse is examined it becomes apparent that T3 has been mistakenly attached to verse 20. One might well imagine that marginal material has played a role in this dislocation and indeed on the basis of Nьrnberg, we see that this is the case. The unusually crowded margin includes first a variant reading followed by two alternative targumim (T2 and T3). While the marginal variant to verse 19 increases the length of the verse, the version contained in the body is rather short. This combination of too much material and a short verse proved to be de Prato's downfall. While T2 begins in the margin adjacent to the end of the targum contained in the body (T1), T3 is shunted down so that it is in fact slightly below the beginning of the targum to verse 20. Without an insertion mark, it would seem harsh to fault de Prato too much for mistakenly attaching 19T3 to verse 20.

A similar case is that of 28,17 where again most witnesses preserve three targumim to the verse (See Illustr. 3). As was the case in chapter 24, Bomberg's text has 28,17 T3 erroneously positioned following the 28,18 T1 (the sole targum to this verse in all other witnesses). And here again, a look at the margins of Nьrnberg provide an explanation for the dislocation. While the reference mark indicating the insertion point for the marginal material (T2 and T3) is provided following 28,17 T1, the Hebrew text of 28,18 is again quite short and its corresponding Aramaic translation (28,18 T1) begins in the body of the text adjacent to the beginning of the second of the two alternative targumim (T3) in the margin. Again, the dislocation is explained by de Prato's confusion regarding the correspondence of marginal material and the body of the text.

Although of a slightly different sort, one final example related to multiple targumim should be brought forward in favour of Bomberg's dependence on Nьrnberg. In verse 36 of chapter 38, Stec notes that Bomberg includes only half of T2 while other witnesses preserve the entire targum. The reason for this otherwise inexplicable omission of the first half of the verse seems to lie again in the margin of Nьrnberg. Because the first half of both T1 and T2 are substantially similar, the marginal hand in Nьrnberg has here and elsewhere saved a repetition of the similar portion by providing an insertion mark referring to the marginal targum only at the point where the two targumim diverge. De Prato apparently misunderstood the intention and cited the last half of T2 contained in the margin of Nьrnberg as a complete, self-contained targum. While the insertion mark is clear, its location is in fact not far from the end of T1 (where one might normally have expected to find it) and this may help to explain de Prato's error.

Finally two smaller, yet no less significant, pieces of evidence are presented by the margins of Nьrnberg. Only Group 3 texts preserve an alternative targum to chapter 12 verse 6 (See Illustr. 4). However, this otherwise unattested alternative targum (T2) is preserved not in the margin but rather in the body of the Nьrnberg text itself. The targum common to 15 witnesses is found in the margin of Nьrnberg and true to form de Prato has included it in Bomberg as the second targum (identical in every respect to Nьrnberg) to the verse. While most of the otherwise unattested readings preserved in this targum consist of variation of matres lectionis or transposition of words, there is one exception. The targum, triggered by the MT's mention of robbers and enemies, supplements its base translation text, supplying the names of Israel's conventional enemies.

While all other witnesses to this verse provide dyb Nwrsmty, Bomberg provides a unique variant, reading instead dyb Nwrs)ty. While it is true that an ithpa'el formation of rs) is attested ³⁸, and in fact provides tolerable sense in the context, it is an unexpected formulation when compared to the comparatively common expression which employs rsm. In fact this stock expression occurs in two other locations in the targum of Job and in neither case do we find any witness presenting significant variation ³⁹. Nьrnberg seems to provide a good explanation for this unexpected variation on the part of Bomberg. The key lies in the fact that unlike the square Ashkenazic script utilized by the scribe responsible for Nьrnberg itself, the hand responsible for the marginal material favoured, as Klein puts it, "a cursive rabbinic hand" ⁴⁰. As with other examples of square script, certain letters within the body of Nьrnberg (w and y, d and r, h and x) resemble each other to the extent that they may easily be confused ⁴¹. While ) and m (non-final) are not easily confused in the square Aramean script, the form of these letters in the cursive script is often very similar and in the case of the marginal targum of 12,6 easily confused. It seems likely that de Prato, after having read the mem of yb)wm (whose cross-stroke is unusually high) and the regularly formed aleph of the same word, then read the cursive mem of Nwrsmty as an aleph by mistake. This error then, which as we have seen would be highly unlikely in the square scripts, seems to provide a further causal link between the marginal material of Nьrnberg (in this case as preserved only in Nьrnberg) and the text which de Prato used for his edition of the Rabbinic bible.

The final piece of evidence in favour of Bomberg's direct dependence on (at least) Nьrnberg's text of the targum of Job is this time to be found not in the margins of Nьrnberg, but rather within the body of the text itself. At several junctures in Nьrnberg, where the body of the text preserves the suffixed preposition ywmwq a (later?) corrector has crossed out the first w and placed a d above the crossed out letter, producing the more common prepositional phrase ywmdq ⁴². While, six of the seven occurrences are located in 5 chapters (21–25) toward the middle of the book, the first such correction is to be found much earlier in chapter 13 (See Illustr. 5). When we turn to Bomberg, we see that Bomberg agrees with the corrected form of Nьrnberg — ywmdq is read in place of ywmwq. But in the isolated first occurrence, Bomberg alone amongst the witnesses, preserves the unexpected ywmmq (?). The explanation for this form lies in a combination of sloppy penmanship on the part of the corrector and the fact that this correction occurs both first and in relative isolation from the other examples in the middle of the book. While the superscript d provided by a corrector in 21,08; 21,27 and 21,31 is clear and well-constructed, that which occurs in 13,15 is anything but. (see Illustr. 5). It is scarcely anything more than an arch and with its proximity to the m which follows it immediately in the same word, we can see how de Prato might have taken it as a mem.

    * * *

The phenomenon of correction raises one final issue which remains to be addressed. Is it possible that de Prato has relied on the codex by which N has been corrected? Willem Smelik, on the basis of his investigation of the relationship between Bomberg and Nьrnberg in Targum Jonathan to Judges 5, observes that: a) de Prato follows a text very similar to the corrected text of Nьrnberg, rather than the original, but that b) he did not always feel obliged to include a marginal variant reading under the heading )"s ⁴³. Primarily on the basis of this latter observation he quite rightly suggests that the lack of correspondence between Bomberg and the marginal material of Nьrnberg allows for the possibility that the witness which was used in the correction of Nьrnberg might have been the MS used by de Prato, as opposed to Nьrnberg itself.

While both of these observations generally hold for the case of the targum of Job, it seems clear that on the basis of both the idiosyncratic ordering and dislocation of targumim in Bomberg and de Prato's confusion of aleph and mem in the cursive script, there is a demonstrated causal link between the material presented in the margins of Nьrnberg and the text included by de Prato in his edition. While much, in fact most, of the marginal material included under the rubric )"s is not included in the text of Bomberg, there are in fact occasions when de Prato has included even entire alternative targumim (28,7T2; 28,8T2) which has been so labelled. And as it is clear that the same hand is responsible for virtually all the marginal material, whether included under the formula )"s or the more usual )"t, it seems more than reasonable to conclude that in the case of the targum of Job at least, there is an inescapable link between the marginal material of Nьrnberg and the text of Bomberg and that therefore Nьrnberg, and not the codex used in its correction, was the text used by de Prato.

While it is hoped that the evidence above has served to present the case for de Prato's dependence on the text of the targum of Job preserved in MS Nьrnberg with sufficient clarity, perhaps it is best to conclude with a brief survey of what I have not shown. Such is our lack of knowledge regarding the sources at de Prato's disposal, that we must allow, at least the possibility (and perhaps probability) that different MSS may have been employed in different ways by de Prato in his production of the various targum texts which are included in the Bomberg Bible. In any case, this study has in no way disproved that other texts were not used by de Prato in his editing process. Second, the question of the relationship between the targum texts of the two witnesses as a whole (Bomberg tg) (Nьrnberg tg) can only be meaningfully addressed after each of the constituent targum texts (ie., Tg. Ruth, Tg. Judges, Frg. Tg. and now Tg. Job etc.) has been examined in its own right. In this sense, it is suggested that the present study, consitutes not the closing but rather a continuance of the case of de Prato's sources, and it is hoped that when all the evidence is finally in we will be somewhat closer to knowing which text it was that lay before not Bomberg, as my title suggests, but before de Prato, the true inspiration behind the Rabbinic Bible.

Summary

It is a well-known fact that even in its earliest edition, an Aramaic translation or targum was amongst the vast and varied material assembled for inclusion in the Rabbinic Bible. But in contrast to the comparative wealth of information we possess regarding the circumstances surrounding its publication, we possess little knowledge with regard to the sources used by Felix de Prato when he took up the task of editing the 1517 Rabbinic Bible for the Venetian publisher Daniel Bomberg. While prior research has shown the importance of the targum text preserved in the Solger Codex (Stadtbibliothek Nurnberg) in any attempt to solve the puzzle of the pre-history of the Rabbinic Bible's targum text, many pieces of this puzzle remain as yet unexamined. The present study locates the targum text preserved in MS Nurnberg (Solger Codex) within the stemmatological framework proposed by D. Stec in the introduction to his critical edition of the Targum of Job. More importantly, the present paper presents decisive evidence (through the detection of editorial errors) that the editor of the first Rabbinic Bible (Felix de Prato) copied his targum text of Job directly from Codex Solger preserved in the Stadtbibliothek Nurnberg.

Chart 1: Stemmatological Divisions

Group 1

MS Camb.Univ.Lib. Ee.5.9
MS Paris, Bib.Nat. Heb.17
MS Vat., Bib.Apostol. Urbinas I
MS Rome, Bib. Angelica N.72
MS Florence, Bib. Medic.Laur. Plut. III.I
MS Milan, Bib. Ambros. B.35 inf.
MS Parma, Biblioteca Palatina 3189
MS Parma, Biblioteca Palatina 3232
MS Wroclaw, Bib. Uniw. M.1106

Group 2

The Antwerp Polyglot, 1570
MS London, Jews' College H.116 (14-42)
MS Paris, Bib. Nationale Heb.110
MS Salamanca, Bib. Univ. M-2
MS Madrid, Bib. de la Universidad Complutense 116-Z-40

Group 3

Bomberg Bible, Venice 1517
MS London, Jews' Col. H.116 (1-13)
(MS Nьrnberg, Stadtbibliothek Nьrnberg)

Group 4

MS Paris, Bib. Nationale Heb.17 (marginal alternative targumim)
MS Paris, Bib. Nationale Heb.17 (corrections)
MS Parma, Biblioteca Palatina 3231

Chart 2

Illustration 1: MS Nьrnberg RtgJob 36,33

Translation:

Nьrnberg body [Bomberg T1] (T3): 
   "…go up from him"

Nьrnberg Mg¹ [Bomberg T2] (T1):
   They tell about the law each one with another; they are
   agitated in anger and rise to him on high.

Nьrnberg Mg² [Bomberg T3] (T2):
   Righteousness declares concerning him (his companion);
   jealousy and anger go up from upon him.

Illustration 2: MS Nьrnberg RtgJob 24,19

Translation:

Nьrnberg Mg¹ [Bomberg add.to 24,19 T1] (T1a):
   and they are not there, thus to She"ol will be carried away
   those who have been condemned to She"ol.

Nьrnberg Mg² [Bomberg 24,19 T2] (T2):
   Like dryness and heat which snatch away the waters
   of the snow and they are not there, thus to She"ol will be
   carried those who have been comdemned to She"ol.

Nьrnberg Mg³ [Bomberg 24, 20 T2] (24, 19 T3):
   Because they suspended the gift of the tithes that were
   commanded in the time of the heat of the Summer, the waters of
   the rain in the time of the Winter and the snow were restrained
   from them; they were condemned to the burial place.

Illustration 3: MS Nьrnberg RtgJob 28,17

Translation:

Nьrnberg body [Bomberg 28,17 T1] (28,17 T1):
   …pure gold.

Nьrnberg Mg¹ [Bomberg 28,17 T2] (28,17 T2):
   A man rich in gold and glass does not put it in order, nor will he
   give in exchange vessels of fine gold.

Nьrnberg Mg² [Bomberg 28,18 T2] (28,17 T3):
   It is not possible that it should be found with ease; but rather in
   something prcious like gold, and whoever does not guard it will
   pass away and be broken like glass…

Illustration 4: MS Nьrnberg RtgJob 12,6

Translation:

Nьrnberg Mg [Bomberg T2] (T1):
And the sons of Ishmael who are at ease in their tents will
  be delivered (Nwrsmty) to robbers, and the Ammonites and
  Moabites who are dwelling securely will be delivered into
  the hand(s) (dyb Nwrsmty) of the wicked and those who
  anger God, like Pharaoh, upon whom God brought the
  plague with this hand.

Illustration 5: MS Nьrnberg RtgJob 13,15

NOTES:

¹ A draft of this paper was read at the congress of the International Organisation for Targum Studies, held in Oslo, Norway (July, 1998). The research for this publication included a five month stay with the team preparing the Bilingual Concordance to the Targum of the Prophets at the Kampen Theological University in 1997-98. This was made possible by a research grant of the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO). Particular thanks are due first to Willem Smelik for directing my attention toward Nьrnberg and the Solger's text of the targum of Job and no less to Johannes DeMoor for allowing me the time to examine the MS in person. Finally, I"d like to thank Willem Smelik, David Stec and my supervisor, Peter Hayman for their constructive comments on a draft of this paper.

For a complete list of variants to the targum of Job see, D. STEC, The Text of the Targum of Job (Leiden 1994). With the exception of this verse, the English translations which appear in this study are those of C. MANGAN, The Targums of Job, Proverbs and Qohelet (AramB 15; Edinburgh 1987). Following the convention observed by the Aramaic Bible series, italics indicate divergences from MT (in this case supplementary material but elsewhere also substitutions) while the base translation is represented by the non-italicized text.

² D. AMRAM, The Makers of Hebrew Books in Italy (London 1963) 150. Psalterium ex Hebraeo ad verbum fere tralatum was apparently printed by Bomberg not at his own press but at that of Herman Liechtenstein.

³ W.F. SMELIK, The Targum of Judges (Leiden 1995) 150-153. D.R.G. BEATTIE, "The Textual Tradition of Targum Ruth", in: The Aramaic Bible: Targums in their Historical Context (D.R.G BEATTIE, M.J. MCNAMARA eds.) (JSOTSS 166; Sheffield 1994) 341-347. B. GROSSFELD, The Two Targums of Esther (AramB 18; Edinburgh 1991) 6-7; M.L. KLEIN, "The Extant Sources of the Fragmentary Targum to the Pentateuch", HUCA 46 (1975) 115-137; M.L. KLEIN, The Fragment Targums of the Pentateuch According to their Extant Sources (AnBib 76; Rome 1980) 1, 26.

⁴ For a fuller account of the history of codicological work on the Targum of Job see the introduction in STEC, The Text of the Targum of Job, 1-3.

⁵ R. WEISS, The Aramaic Targum of Job (Heb.) (Tel-Aviv 1979)

⁶ F.J. FERNБNDEZ VALLINA, El Targum de Job (ediciуn crнtica) (Madrid 1980).

⁷ STEC, The Text of the Targum of Job.

⁸ STEC, The Text of the Targum of Job, 68.

⁹ A computer-assisted stemmatological investigation of the targum of Isaiah is currently being undertaken by A. Houtman in association with the BCTP team in Kampen. Some time soon, this technology may be of use in further refining and confirming the existing stemma for the targum of Job. In its absence however, I would suggest that my attempt here consists of locating Nьrnberg generally within the existing framework rather than determining the entire network of relationships between Nьrnberg and the other textual witnesses.

¹⁰ MS Nьrnberg is a 13^th c. codex preserved in 7 volumes at the Stadtbibliothek Nьrnberg. See KLEIN, The Fragment Targums of the Pentateuch, 31 for a full description of the MS. TgJob and the Hebrew text alternate verse by verse in two columns of 33 lines per folio.

¹¹ Computer-based stemmatologists are developing tools (i.e., cladistics, graph theory) to detect the boundaries of sections in such cases of "consecutive contamination". See E. WATTEL – M.VAN MULKEN, "Shock Waves in Text Traditions", Studies in Stemmatology (P.VAN REENEN – M. VAN MULKEN eds.) (Amsterdam – Philadelphia 1996) 105-122.

¹² STEC, The Text of the Targum of Job, 77-79.

¹³ STEC, The Text of the Targum of Job, 112. Stec makes the following distinction between alternative targumim and alternative verse patterns: "If a verse has only one targum in all the texts used in this edition the alternative patterns of the verse are indicated by (a) (b) etc. If a verse is given more than one targum in at least one text, the labelling system T1, T2, etc is used".

  ¹⁴ 24,24; 28,7T2 (b); 31,34c; 34,24 (amongst the wide variety of different targum patterns to this verse, Nьrnberg agrees with Bomberg with the exception of one vowel letter; MS 116's pattern is substantially different) 38,13T1 and T2; 41,17 Nьrnberg agrees with Bomberg with the exception of one marginal reading.

¹⁵ STEC, The Text of the Targum of Job, 78. Readings linking MS 116 to Group 2 texts begin in chapter 14 despite the fact that the entire manuscript appears to have been written in one hand and there is no physical break evident.

¹⁶ STEC, The Text of the Targum of Job, 77-78. Both identical readings and those exhibiting matres lectionis variation are included.

¹⁷ ynwb 1,4; dk 1,5; ytyk 1,6; Mdq 1,7; ydb(b 1,8;
)h 1,12; )yyrwt 1,14; Nmt < 1,21.

¹⁸ )ylmgd 1,3; #ymx 1,3; )(b# 1,4; )mwyb 1,6;
yt)w 1,6; Nwh(ycmb 1,6; )n#) 1,7; r#p)h 1,8;
lyxd bwy) 1,9; ywtygw 1,10; )mwy ylwxt) 1,13;
rdsm 1,22; y+xym 1,22; (Readings unique to only Bomberg and Antwerp are not included in this category despite the strong evidence that Montano made use of Bomberg in his production of Antwerp. Cf., STEC, The Text of the Targum, 11.)

¹⁹ )klhmw 1,7; Ngmld 1,9; rwsm 1,12;

²⁰ Pw) 13,2; ywwl 13,5; ywp)h 13,8; hytytrw 13,11T1.

²¹ tylw 13,2; qyt#md 13,5; ytwwpy# 13,6; )m+ql 13,12;
ywmdq 13,16; )t#h 13,19; ylgyr 13,27.

²² It is significant to note that in many of the cases where Nьrnberg and MS 116 agree against all other witnesses, the corrections found both in Nьrnberg (body) and Nьrnberg (Mg) often agree with Bomberg.

²³ STEC, The Text of the Targum, 78. Nьrnberg agrees with all the examples adduced by Stec in chapters 1-13. Others include 15,24; 31,15; 35,10.

²⁴ This of course does not preclude a Sephardic origin for as recognised by Stec, paleography alone cannot be decisive in the determination of a ms tradition.

²⁵ STEC, The Text of the Targum, 78. The features brought forward in favour of this affinity are a) the use of the shorter form of the 3^rd masc pronom suffix with plural nouns, i.e., yw versus yhw; a preference for –m as opposed to the separate form Nm of the same preposition; the use of )"t instead of )"l when referring to alternative targumim; a paucity of alternative targumim to individual words and phrases (although in many cases they may be found in Nьrnberg (Mg).

²⁶ See note 3 above.

²⁷ GROSSFELD, The Two Targums of Esther, 6-7.

²⁸ By which Beattie intends the "Rabbinic Bibles of the sixteenth century and their lineal descendants, and the editions of Lagarde and Sperber" (Beattie, "The Textual Tradition of Targum Ruth", 342, n.10).

²⁹ BEATTIE, "The Textual Tradition of Targum Ruth", 343.

³⁰ STEC, The Text of the Targum, 86, prefers this term to that suggested by Weiss ("duplicate targumim") on the grounds that "duplicate" is suggestive of "two" and does not reflect the cases where three or even four targumim are attested. In addition, it seems probable that the use of "duplicate" might lead to the incorrect assumption that the various targumim are merely copied versions of a single targum.

³¹ See STEC, The Text of the Targum, 85-88 for a synopsis and evaluation of explanations of this phenomenon beginning with that of Bacher in the late 19^th c. This distinctive feature of the Targum of Job is also encountered occasionally in TgPsalms. Interestingly, an avoidance of explicit multiple translation is seen by Alexander Samely as a basic (essential?) formal feature of the targumim to the Pentateuch. Cf., A. SAMELY, The Interpretation of Speech in the Pentateuch Targums (Tьbingen 1992) 179.

³² STEC, The Text of the Targum, 89.

³³ The only exceptions to this rule come toward the beginning of the targum. At 4,11 and 7,12, two locations where the insertion marks are clearly indicated before the targum given in the body, de Prato has correctly given the marginal targum priority over that contained in the body of the text. This preserves the order intended by the marginal hand and indeed the order of Bomberg accords with other manuscripts for this verse.

³⁴ Eg., 6,6.7; 11,12; 14,18; 15,10.18.20.32; 18,12.13.16; 29,15; 30,2.24; 34,26; 36,19; 37,1.24; 38,12.15.

³⁵ Eg. 6,7; 11,12; 14,18; 15,10.18.20.32; 18,4.12.13.16; 29,15; 30,2; 30,4.19.24; 34,26; 36,19; 37,1.12.21.22.24; 38,12.14.15.

³⁶ Eg., 5,7; 6,6; 12,5.6; 14,22; 28,7*.8*; 34,9*; 36,20*.33; 37,23*; 38,25.37; (*indicates insertion mark is missing or location is unclear)

³⁷ While KLEIN, "The Extant Sources of the Fragmentary Targum", 127 clearly improves upon Ginzburger (Das Fragmententhargum Berlin 1899, VIII-IX) in the evaluation of the errors in the Fragmentary Targum, Ginzburger's approach is essentially sound and Klein recognizes this decisive nature of errors in his own argument for Bomberg's dependence on Nurn.

³⁸ See M. JASTROW, Dictionary of the Targumim, The Talmud Babli and Yerushalmi and the Midrashic Literature, (Jerusalem 1903) who glosses the resulting form as "to be imprisoned" (98).

³⁹ The expression also occurs in 1,12 and again in 9,24.

⁴⁰ KLEIN, The Fragment Targums of the Pentateuch, vol.1, 31.

⁴¹ A good example of the confusion of r and d may be seen in 37,22 and again at 40,10 where in both cases Bomberg reads )dwhby# where the other witnesses preserve )rwhby#. The letter preserved in Nьrnberg at these two points may be plausibly read as either r or d.

⁴² 21,08, 21,27, 21,31, 22,37, 23,04, 25,05. JASTROW, A Dictionary of Targumim, 1332, sees ywmwq as a contraction of ywmdq.

⁴³ SMELIK, The Targum of Judges, 153.