Textual problems in the Hebrew Bible

I wrote this back in 1997 for a UBS TTW conference. But I believe that this topic remains relevant to Bible translators today. The Hebrew font display is messed up in the online version at points (see Dan 11.6), but if you download the document, it is fine.

---

Extracted text from BASCOM1997.pdf:

Textual problems in the Hebrew Bible may or may not affect translation and Problems of variation in scripture translations may or may not be textual so Why is it important to keep it all straight? Introduction. In the course of translation, the Bible translator is often faced with a bewildering array of variation in the modern versions available to him or her. This is especially true of the translated text of the Hebrew Bible, or Old Testament. Without direct access to the original languages of the text, it can appear in many cases that either there is no consensus on how the text reads, or that there is no consensus on what the text means, or both. As a matter of fact, this is too often the case. Yet even more often, variations encountered in modern translations have a different source than those variations which interest text critics. All translations, ancient or modern, have varied at times in wording or style with little if any seeming underlying variation in interpretation (or text), and these translations have also varied significantly at times in interpretation with little if any apparent underlying disagreement as to the actual reading of the text involved. Furthermore, the reverse is also often the case. Ancient versions can seemingly vary significantly from the Hebrew text in terms of the forms they were reading and yet apparently be trying to get at the same meaning (though in such cases wording and style would of necessity also vary). While many text critics would be reluctant to see in these seeming cases of ancient translational adjustments variation which is truly text-critical (i.e. dependence on a different underlying text form), most would in fact include them in text-critical notes, since judgments as to meanings of obscure text across different languages are inherently difficult to prove. But deciding whether translations differ due to differences of wording or style, of interpretation, or of underlying text crucially affects the strategy a translator should use for translating the text. Thus if differences in wording or style are the cause of variation in translations, the translator must unpack the common underlying meaning and translate that meaning in the wording and style decided upon for the translation involved. If the difference is in interpretation, the translator must first decide in favor of a particular interpretation, then translate that meaning, ignoring all others. Finally, if the difference is textually based, the translator must try to translate the best meaning that can be had from what is determined to be the best text based on sound text-critical principles. Of course, this is far easier said than done. First there is the problem of how to determine what kind(s) of variation is (are) involved. Much of the time more than one source is motivating the differences in the translations, confusing the picture. Then there is the problem for the translator in deciding whether to treat what has traditionally been considered a text-critical problem (complete with notes citing the variant readings) simply as an ancient translational one. In other cases there may be a legitimate textcritical problem involved, but one which simply does not affect the meaning, and thus does not need to be taken into account in translation. These both happen more often than one might at first imagine. Finally, there are practical problems of translation and exegetical traditions which, while not unique to

Robert Bascom

TTW 1997 - Mrida

Page 2

potential textual problems, can and do make it more difficult to see clearly what strategy should be used in a given case. It is often easiest to eliminate the text-critical source for variation, since it is the one likeliest to be documented in textual notes, and because it can be checked, either directly in the critical editions and their apparatuses or in helps such as HOTTP or CCAT. For many translators, however, this task is daunting. A few examples will illustrate why.1

Reference: 1Sa 2.33

              ,          The man of you whom I shall not cut off from my altar shall be spared to weep out his eyes and grieve his heart; and all the increase of your house shall die by the sword of men. Texts: your eyes...your soul [M V S T] / his eyes...his soul [G] 2 as men [M V S T] / by the sword of men [Qa G] Analysis: This change from second person (M) to third person (G) is exactly what seems to be required by the context, and thus has unsurprisingly been followed by the majority of modern versions. HOTTP/CTAT note, however, that Abiathar will be suffering on behalf of Eli, who is the object in focus in the passage, and Eli is properly referred to in the second person. It may require a note, or some translational adjustment (e.g. weep out his eyes for you...), but translators can manage the M text. The problem is having to go against virtually the whole force of translational tradition to translate the best text. The second variant in this verse involves the addition of the word sword by G and supported by a Qumran manuscript. Text-critically the plus seems to be an adjustment to 1Sa 22.18, where the priests at Nob are killed by Doeg. The M men can be understood as adults, in the prime of life, as opposed to old men. Thus M is to be preferred. But it is cases such as these where text-critical subtleties are hard to appreciate as a translator. If M means to say that the men will be cut down in the prime of life, and not die of old age, are not G and Qumran merely making implicit information explicit?

1

Unless indicated otherwise, text in Hebrew will be from the BHS, in Greek from the Rahlfs edition of the Septuagint, and English translation will be from the RSV. 2 Following CTAT convention, M = MT, Q = Qumran, V = Vulgate, T = Targum, G = LXX, Th = Theodotion, S = Syriac. Also, for the limited purposes of this paper, minor witnesses will be ignored, witnesses aligning with or differing from M in different ways will sometimes be placed together (usually in parentheses in such cases), and in general the cases will be treated as simply as possible, rather than in full detail.

2

Robert Bascom Reference: 1Ki 10.8 (// 2Ch 9.7)

TTW 1997 - Mrida

Page 3

          '        Happy are your wives! Happy are these your servants, who continually stand before you and hear your wisdom! Text: men [M V T] / wives [G S] Analysis: The M has men, which can be understood either as officials, soldiers or people in general. But HOTTP/CTAT prefer Gs wives in 1Ki 10.8 and see men as a scribal correction there. On the other hand, men is considered by HOTTP/CTAT to be the original text of the exact parallel in 2Ch 9.7. Both the scribal correction in 1Ki 10.8 and the original Hebrew text of 2Ch 9.7 are thus seen as late attempts to avoid a positive reference to Solomons wives, who by that time had bad reputations in the biblical tradition (cf. 1Ki 11.1-4). This kind of argumentation, which comes so naturally to scholars used to historical investigation and relative chronological placement of biblical writings, is quite puzzling to many if not most translators in the field. In such a case, however, perhaps all that can be done is to educate translators as to the relative dates of Kings and Chronicles in relation to the canonical process, and defend what on the surface seems to many as an overly sophisticated way of reasoning.

Reference: Isa 43:14

                     Thus says the LORD, your Redeemer, the Holy One of Israel: For your sake I will send to Babylon and break down all the bars, and the shouting of the Chaldeans will be turned to lamentations. Texts: fugitives [M 1Qa G S] / bars [Th V T] ships [all] / lamentations [] Analysis: Both of these cases involve a variant vocalization of M. In the first case, several ancient witnesses have followed the alternate vocalization, but in the second case, all ancient witnesses agree with M, and it is only modern interpreters who have attempted a different reading. The alternations in vocalization give either fugitives or bars in the first case, and ships or lamentations in the second. Both alternate readings are appealing in context, and the second even more than the first. There is a wide and confusing range of choices which have been made by modern translations, many of which 3

Robert Bascom

TTW 1997 - Mrida

Page 4

will choose to repoint one, but not both of the words in question (cf. REB). Translators are thus faced with what looks like a difficult text-critical problem with little agreement, when actually the issue lies with most modern translations not having figured out how to make sense of a difficult, but textcritically sound Hebrew text.3

Reference: Dan 11.6

                                 '      After some years they shall make an alliance, and the daughter of the king of the south shall come to the king of the north to make peace; but she shall not retain the strength of her arm, and he and his offspring shall not endure; but she shall be given up, and her attendants, her child, and he who got possession of her. Texts: arm [M G] / seed [Th V (S) 4] he who has begotten her (i.e. her father) [M] / her child [Th V S (G)] Analysis: Both of these cases are rated C in HOTTP/CTAT, but for opposite reasons. In the first case, arm (in the sense of power) is well attested, but both the grammar and the context are confusing enough to render the revocalized meaning seed (offspring in RSV) a reasonable alternative, and one followed by some ancient versions as well. In the second case M stands alone, but on careful examination, the father in the latter part of the verse would be the same as the king of the south in the first part, and thus makes good sense in the context. Nevertheless, this is a fairly complicated passage in a book written in an intentionally obscure manner about historical details with which the writer expected his original audience to be intimately familiar, and which now have to be reconstructed. And when text critics can only give a C rating (or worse at times) to the text they are supporting as the best text to translate, translators often feel (perhaps somewhat rightly) they are not as obliged to follow the text-critical advice as in cases where the case has been made on a firmer basis.5

Reference: Mic 2.8

3 4

Even the NIV, which translates both fugitives and ships, changes shouting to pride (or interprets it as such?). In a comment only a translator can truly appreciate, Barthlemy notes in CTAT: In the lengthy historicising paraphrase that S offers here, it is impossible to find one word which corresponds to 6A. 5 A good example of HOTTP/CTAT themselves sending mixed signals is Zep 3.8, where Ms prey is revocalized as witness by two of the five committee members, and both readings are given a C rating.

4

Robert Bascom

TTW 1997 - Mrida

Page 5

                 But you rise against my people as an enemy; you strip the robe from the peaceful, from those who pass by trustingly with no thought of war. Texts: yesterday [all] / against [] cloak [M V] / peace [G S] honor [M V] / cloak [G S] pass by [M V T] / impose [G S] turn away from [M] / disaster [G] / return from [] Analysis: There is no phrase in M of this verse which has not been treated as a text-critical problem and emended in some translation, ancient or modern. In several cases the ancient witnesses have paraphrased to the point that they are not even usefully listed as witnesses in the case. The result has been a chaotic scattering of readings in modern translations. On the other hand, HOTTP/CTAT give the M readings an A and four B ratings, in order. The trouble comes for translators (and one suspects this has always been the case) in making sense of M. The suggested translation in HOTTP/CTAT is somewhat less than fully intelligible: yesterday still my people rose up as a enemy (of God and their poor brothers): <being> in front of a cloak, they strip off the honor of those who pass by in confidence, being averse to war/strife. Given enough time and effort, a readable translation could probably be made from this suggestion, but in its present form it is far from helpful to translators. It is not just that M is difficult (what does it mean to be in front of a cloak?), but the wording is such that at every turn one is tempted to emend the Hebrew to what text critics have come to call the easier reading. A translators commitment to text criticism is truly tested by a verses like this one, especially when the translation suggestions are little more than literal back-translations of an admittedly obscure Hebrew source text. Making the translators task more difficult in cases such as this one and others like it (see the next case as well) is that fact that the UBS handbook for this book assumes the text followed RSV/TEV for its comments and ignores HOTTP/CTAT completely.

Reference: Mic 6.9/10

            .             The voice of the LORD cries to the cityand it is sound wisdom to fear thy name: Hear, O tribe and assembly of the city! Texts: he who appoints it. [M] / assembly of [] Still [M V S] / the city [G] 5

Robert Bascom

TTW 1997 - Mrida

Page 6

Analysis: It is helpful for several reasons to quote the UBS handbook at length at this point: The Hebrew text is practically unintelligible, but it is followed literally by KJV and even by the recent NIV. NIV has "Heed the rod and the One who appointed it." It is hard to see what this is supposed to mean. The large majority of modern translations (RSV, JB, NAB, Mft, TEV) and commentaries differ from this interpretation in three ways. First, they take the word translated "rod" in NIV to mean "tribe" instead. In other passages, this Hebrew word sometimes means "rod" and sometimes means "tribe." Second, they divide the Hebrew text so as to include the first word of verse 10 with verse 9. Third, they change three consonants in the Hebrew which a scribe could easily have copied wrongly. The meaning of their revised text is "tribe and assembly of the city" (RSV). The term "assembly" may refer to a formal gathering, but this is not certain. The two nouns "tribe" and "assembly" together stand for the people, and TEV expresses this in more natural English as you people who assemble in the city. The word assemble may suggest that not all the people addressed actually lived in the city. This may in fact be one reason why the word "tribe" is used, to refer to the people of the tribe of Judah, whether they were citizens of Jerusalem itself or not. Even if they lived in smaller towns or villages outside of Jerusalem, they would gather or assemble in Jerusalem for important occasions. The HOTTP/CTAT rating C for the text of M indicates some uncertainty on the part of text critics for this reading. On the other hand, G is alone in support of the second alternate reading and arguments about misreading of similar letters cut both ways. Worse, while G, V, and S do not fully align with M (they support differing vocalizations), neither do they support the conjectural reconstruction of assembly from who appoints it. Once again M has the harder reading. The handbook has laid out most of the essential points of case simply and clearly, and in non-technical language. It has noted the major difficulties with the Hebrew text as it stands and has given the major alternative. Unfortunately, the handbook has not only followed a reading which is a mixture of a conjecture and a variant reading on insufficient evidence, and then has spent the last half of its paragraph on the subject exegeting the dubious result. If HOTTP and CTAT need to give translators more help in understanding and using the text they prefer, the handbook authors should be careful not to spend quite so much time explicating what in the end is probably not the text at all.

Reference: Hab 2.5

                      '           

6

Robert Bascom

TTW 1997 - Mrida

Page 7

Moreover, wine is treacherous; the arrogant man shall not abide. His greed is as wide as Sheol; like death he has never enough. He gathers for himself all nations, and collects as his own all peoples. Text: boaster [G 1QpHab (lemma)] / wine [M V T] / wealth [1QpHab (pesher)] Analysis: On the basis of the consonantal text of 1QpHab (interpreted as wealth by the pesher) and G (itself corrected here by Rahlfs and Ziegler) HOTTP/CTAT propose a kind of conjecture--a form not found elsewhere in biblical Hebrew but which can be posited on the basis of cross-linguistic evidence from Arabic. This conjectural form (found after all in the consonants of 1QpHab and indirectly attested in the uncorrected G) would be a rare word for boaster (which fits well in context), presumably substituted here by the more common words for wine (M, V, and T) and wealth (1QpHab-pesher). It is highly unusual to find HOTTP/CTAT proposing what looks like a conjectural emendation, even one which has as much basis as this one does. The text-critical arguments are sound, however, and translators are normally encouraged to follow the best text available. One factor making this problematic is that HOTTP does not give the evidence, and its presentation of the case makes the emendation look entirely conjectural, even labeling it as such. CTAT does give the evidence in full detail, but the languages and work required to use this tool make it impractical for many translators.6 Another factor inhibiting the translator from following the reading suggested by HOTTP/CTAT is the narrow basis the reading has in modern translations. The only major modern translation tradition to follow this variant reading has been NEB/REB,7 precisely on the basis of the consonants of 1QpHab. This relative lack of translational support alone should not keep translators from using the best text, but it often plays a greater role than many would like to admit. Most distressing in this case and others like it is the existence of a UBS handbook recommending not the reading preferred by HOTTP/CTAT, but the variant wealth of the pesher of 1QpHab.

Reference: Zep 1.3

                    I will sweep away man and beast; I will sweep away the birds of the air and the fish of the sea. I will overthrow the wicked; I will cut off mankind from the face of the earth, says the LORD. Text: what makes the wicked stumble [M Sym] / I will make the wicked stumble [Th V S T (G)]

6

A more accessible form of the information in CTAT would be of great value to many translators, and is a project already being undertaken by the UBS. 7 It should be noted that both the Spanish and French common language versions seem to follow this reading, however.

7

Robert Bascom

TTW 1997 - Mrida

Page 8

Analysis: As the UBS handbook notes, the translation I will overthrow is based on a slight change in the traditional Hebrew text. This change is followed by nearly all the ancient witnesses (the G tradition being divided and lacking the whole phrase in many mss) and most modern translations. It fits better both in the context and forms nicely parallel lines of poetry. For these reasons it is not hard to understand why once again the UBS handbook advises against the HOTTP/CTAT suggestion and for the change.8 The problem (again) is what is good for translators (texts which make good sense) are often bad for text critics (= the easier reading). CTAT has an explanation for reading the Hebrew as it stands which respects the grammar and the accentuation preserved by M. That which makes the wicked stumble can be read as (a feminine hifil participle = relative clause) referring to the beasts, birds, and fish, which are in this reading considered to be objects of idolatry, i.e. stumbling blocks. To some this will seem like over-sophisticated reasoning bent on justifying M at all costs,9 but the text-critical arguments are well-founded, and the rating given M is a B.

Reference: Zep 3.17e-18

       .        '   he will exult over you with loud singing 18 as on a day of festival. I will remove disaster from you, so that you will not bear reproach for it. [KJV-alternate reading] he will joy over thee with singing. 18 I will gather them that are sorrowful for the solemn assembly, who are of thee, to whom the reproach of it was a burden. Texts: afflicted by being deprived of the feasts [M V] as on the day of the festival [G (T)] they were [M V S T] / disaster [G] the burden [M] / who will lay shame upon her [M G V] / upon you [S T]

8

The reactions of handbook authors faced with difficult translation suggestions from HOTTP/CTAT can be seen clearly in this quote from the handbook at Zep 2.14, where Ms phrase beasts of the nations is dealt with in the following manner in the handbook: Even if the Hebrew goy is correct and carries the meaning "nations," what does the expression "all the beasts of the nations" mean? HOTTP suggests "all the beasts living in flocks (or, herds)," but this meaning is not found in any major translation. To be fair, the handbook authors go on to suggest following Ms reading, but with the interpretation of nations found in many commentaries as (all) species, and find translational support for this in several major translations as well. 9 One of the most obvious examples of this apparent tendency is at 1Sa 13.1, where HOTTP/CTAT defend the number two (in M) with reference to the length of Sauls reign, saying that was the time Saul reigned before Samuels anointing of David.

8

Robert Bascom

TTW 1997 - Mrida

Page 9

Analysis: CTAT has treated the text of this verse in excruciating detail, in some instances breaking up and in others going beyond the cases treated in HOTTP. CTAT gives six pages to the verse (two more than it gives to the book of Jonah), and in the end two out of the three cases in this verse which were dealt with in the HOTTP committee are rated C. Once again the UBS handbook is helpful in laying out the situation: Most modern translations follow the Septuagint and take the first phrase of verse 18 with verse 17 as the concluding words of the message begun in verse 16 (RSV, BJ, JB, NAB, NEB, TEV, FrCL, GeCL). Except for NEB, they also follow the text of the Greek and Syriac versions and translate as on a day of festival (RSV, JB) or something similar... versions [] which followed the Septuagint by taking the first two words of verse 18 with verse 17, also follow it in reading disaster or something similar in meaning instead of the verb "they were." They then take this word with the verb "gather" (understood in the sense of "gather for removal") and translate as I will remove disaster from you (RSV, compare NAB) or "I have taken away your misfortune" (JB)... A number of modern versions take the word translated "burden" in KJV, RV and NIV as a verb rather than a noun, having the meaning of "bear." Thus RSV renders so that you will not bear reproach for it, and JB "no longer need you bear the disgrace of it"... Some modern scholars wish to follow the ancient Syriac translation and read "upon you" for the "upon her" of RV. The handbook authors are as fully aware of the difficulties in this verse as are HOTTP and CTAT and clearly state this as well: Such are the difficulties in this verse that the TEV footnote "Verse 18 in Hebrew is unclear" can only be described as a masterpiece of understatement... In the end, the handbook authors decide for the force of translation tradition and against HOTTP/CTAT: An interpretation of verse 18 different from that of RSV or TEV may be found in HOTTP. However, this interpretation is not represented in any modern English translation and does not fit the context very readily, so it must be viewed with some reserve. The handbook is not entirely unfair in its assessment, but as in the other cases we have seen, the problem is that while HOTTP/CTAT do not seem to go out of their way to make intelligible translation suggestions, handbook authors often do not seem go out of their way to understand and try to translate the text favored by HOTT/CTAT. As we have seen, this is the case even when the textual cases are straightforward and well-grounded. In complicated cases or cases with more doubt (such as those in this verse), the HOTTP/CTAT text-critical judgments are very likely to be ignored or dismissed by UBS handbooks (and thus also by UBS translators).10 10

Though there is a great deal of difference in the way different handbook authors have treated text-critical issues, the tendency seen throughout the above examples to ignore text-critical decision when faced with difficult-to-translate texts is widespread (cf. for example Hab 3.16, which involves a small change but fixes an apparent grammatical/context problem).

9

Robert Bascom Conclusion:

TTW 1997 - Mrida

Page 10

To come back to the question posed in the title to this paper, it would seem obvious that it is important to determine the best text to translate before trying to actually translate that text. Otherwise the translator will apply too soon translational principles of intelligibility and naturalness which will nearly always align with ancient or modern attempts to facilitate, to make hard texts easier in one way or another. It is only after the best text has been determined that translational principle should be applied, and then to full advantage. Often it takes the best efforts of a translator to convey the meaning of a difficult (but text-critically sound) text. Clearly, the problems of the translator facing text-critical problems which were laid out in the introduction should be addressed by the proper use of HOTTP/CTAT on the one hand, and by UBS handbooks (where they exist) on the other. But as we have seen, at least two problems still exist for the translator who wishes to use the best possible text when translating Scripture. One is the difficulty in accessing some of the major tools of text-critical research, tools ironically made available at least in part through UBS efforts. HOTTP is far too cryptic for general use in many instances, and often requires a full research of the case in order for the investigator to understand what the HOTTP opinion really means. CTAT is far from cryptic, but its encyclopedic level of detail is often intimidating for the field translator, and it takes some training to be able to get to the parts which are relevant for translation.11 The UBS handbooks, on the other hand, in large measure have paid more attention to translation tradition than to text-critical suggestions made by HOTTP/CTAT. The policy not to innovate in translation has in general been a good one. There have been fads in text-criticism, and if a previous generations tendency to correct M with G or conjectures had been followed by many translations, the situation facing translators would clearly be worse.12 Yet reluctance to follow sound text-critical judgments must also be overcome. Translational traditions have too often cemented into texts wording and interpretations which are at best awkward and at worst simply wrong.13 The practical problem remains that for the majority of translators working in the field, HOTTP/CTAT suggestions are often apparently not (and often in actuality not) represented in any translation they can expect their audience to use for comparison. Handbook authors have doubtless been aware of this fact, and felt it an unfair burden on translators to expect them to translate a text they can find in scholarly volumes alone. It may be another generation or more before translations exist in every major language which will back up the best in text-critical scholarship. But first this scholarship must be made accessible and intelligible to more than those directly involved in text criticism itself. Interestingly, both assignments clearly belong to the UBS.

11 12

It is also in French, which will hinder many field translators in parts of the world which are not French-speaking. An unfortunate example of a modern translation following the tendency to correct M with the ancient versions or conjectures is the VP. 13 For example, Psa 23 does not end with forever, but with for length of days. The theological significance is such that it has been nearly impossible to change the major translation tradition (but see TEV and FC).

10

Robert Bascom

TTW 1997 - Mrida Abbreviations:

Page 11

M = Masoretic Text G = Septuagint Th = Theodotion Sym = Symachus V = Vulgate S = Syriac T = Targum HOTTP = Hebrew Old Testament Text Project CTAT = Critique Textuelle de LAncien Testament RSV = Revised Standard Version KJV = King James Version RV = Revised Version NIV = New International Version TEV = Todays English Version FC = Francais Courant VP = Versin Popular

11