A Recent Popularisation of Professor F. M. Cross’ Theories on the Text of the Old Testament

This article was originally published in the Tyndale Bulletin 26 (1975), pp. 113–132. It is also freely available at https://tyndalebulletin.org.

Throughout the last decade one fruitful source of knowledge of the contents of the manuscript fragments from the famous Cave 4 at Qumran has been the theses of Professor F. M. Cross’ PhD students.¹ And very fine and thoroughly professional pieces of work they are. Now from the pen of the author of one of these theses,² Professor R. W. Klein, comes a work of a somewhat different kind. His Textual Criticism of the Old Testament stands fourth in the Old Testament division of a series of Guides to Biblical Scholarship, published by Fortress Press under the general editorship of Gene M. Tucker. It is written not, of course, for the expert, but for college or seminary students, some of whom may ‘not yet have mastered Greek and Hebrew’ (p. ix), for the ‘average exegete, pastor, rabbi, theological student or informed lay person’ (p. 84): and it sets out to encourage all these people to engage, according to the several abilities, in the textual criticism of the Old Testament. The book everywhere breathes enthusiasm, which in itself is no small tribute to the impact that Professor Cross makes on his pupils. It is sincerely to be hoped that this enthusiasm will inspire a new generation of students with a sense of the importance of textual criticism so that they may be willing to qualify themselves rigorously for the work. It is scarcely a field for the unqualified. A book of this kind naturally has to be not only as ‘uncomplicated’ as possible, but also selective, and doubtless no two authors would agree completely on what selection to make or how to simplify issues which in reality are very complicated. Faced with the choice Professor Klein has decided to select almost exclusively Professor Cross’s views and theories and so ‘to disseminate his insights for wider discussion’ (p. ix). To balance this onesidedness he occasionally warns his readers that some of the views he propounds are very subjectively based, and that over these views ‘leading scholars vigorously disagree with one another’ (pp. viii, 15, 69). On the other hand he rarely gives the views of the scholars who disagree with Professor Cross. Since, however, a beginner cannot be expected to know these views without being told, and without such knowledge can scarcely arrive at a balanced assessment of the present state of Old Testament textual criticism, the present article proposes to do two things: to discuss some of the views put forward by Klein, and, where it might seem helpful, to supplement the information he gives. Following a useful glossary explaining the technical terms of the discipline, Chapter I, entitled ‘From Origin to Origen’, gives a succinct account of what is known (and/or normally conjectured) about the history of the Greek Old Testament, its origins and subsequent revisions. Naturally, it devotes a good deal of space to discussing the so-called Letter of Aristeas, and on the much debated question of the meaning of its famous paragraph 30, aligns itself with S. P. Brock’s very sensible view that Aristeas is observing that some copies of the Hebrew Old Testament in third-century (BC) Alexandria had been carelessly made. The curse which Aristeas (para. 311) records was pronounced by Alexandrian Jewry on any who would seek to make changes in the translation, Klein, again following Brock, thinks was meant to forestall any revision of the Greek on the basis of some different Hebrew.

This ‘hard-line approach in Aristeas’, Klein suggests,

may have been related to the discovery at his time of discrepancies between the Greek and Hebrew Bibles. Some of these discrepancies were caused by the fact that even the original LXX was not consistently literal. Others would have arisen subsequent to the translation of the LXX as the Hebrew text underwent various changes (additions, subtractions, copyists’ errors and the like). The number of such discrepancies would be multiplied many times if the LXX, based on the Hebrew text as transmitted in Egypt, were compared with the considerably different text types at home in places like Palestine or Babylon. (p. 4)

It will be helpful at this point to explain in a little more detail than Klein does, what is meant by saying that the original LXX was ‘not consistently literal’. This is of basic importance to every aspect of the problem and might as well be dealt with fully at once. In the Pentateuch non-literalisms occur where the LXX sensibly enough translates a Hebrew idiom by a Greek idiom: e.g. at Exodus 18:7 the LXX says ‘they greeted one another’ for the Hebrew’s “they asked a man his neighbour concerning his peace”. They occur also when the translator for reasons of piety re-phrases the direct statements of the Hebrew: e.g. at Exodus 24:11, instead of saying with the Hebrew that the seventy elders ‘beheld God’, the Greek says ‘they were seen in the place of God’. Again they occur when translators instead of translating a law, give the interpretation of the law that was favoured by Alexandrian Jewry: e.g. at Exodus 21:22–23 the Hebrew has: ‘If men strive together, and hurt a woman with child, so that her fruit depart, and yet no harm follow . . . [such and such a penalty]. But if any harm follow. . . (some other penalty)’; the Greek has ‘. . . a woman with child, and there come forth her child not fully fashioned . . . But if it was fully fashioned. . .’. The difference is considerable: in the Hebrew the penalty is governed by whether the woman, after her miscarriage, suffers any bad effects or not; in the Greek, by whether the aborted fetus had, or had not, developed to the stage where it was a viable human-being. In addition to this legal type of non-literal interpretation, there is also the midrashic type of interpretation, so called because of the (to our mind) fantastic principles on which such interpretations were based. In Exodus 27:13–16, for example, the Hebrew gives the dimensions of the wall at the east end of the tabernacle court as being 50 cubits in all, made up of 15 cubits on each side of a gate of 20 cubits. The Babylonian Talmud, ʿErubin 26, interprets the figure 15 in a most bizarre manner, as referring to height; the LXX does the same.

Outside the Pentateuch, as Klein points out (pp. 63–5, and especially footnote 3 on p. 63), extreme non-literalisms sometimes occur. The Greek Book of Proverbs, for example, not only contains proverbs of Greek origin (and thus manifestly not taken from some Hebrew biblical Vorlage), but not infrequently rejects the parallelism of the Hebrew poetry which it is supposedly translating, and substitutes antithetic statements, so changing the meaning completely.³ Again, both translations of Daniel put into the mouth of the three young men in the furnace a long prayer, which, whatever its source, is certainly not original: it is part of the habit that developed in later times of composing (and of sometimes inserting into the text of Scripture) prayers, letters or edicts to which Scripture itself alludes but which it does not record.⁴

Now these observations are not meant to deny, nor even to obscure, the fact that in practically all books of the Greek Old Testament there are long passages where the Greek follows its Hebrew Vorlage closely, sometimes word for word with the same word-order, whether that Vorlage was the proto-MT or some other Hebrew text-tradition. But they do mean that non-literalisms of all sorts and kinds can suddenly occur without warning even in the most sober and straightforward parts of the LXX such as the Pentateuch. And that in turn means that where the LXX disagrees with the MT and no other Hebrew text has survived, conjecturing exactly what the LXX’s Vorlage read is beset with uncertainty.

But to return to Klein’s history of the LXX and its revisions. As concrete examples of the type of early revision of the LXX against which Aristeas may have been polemicizing Klein cites the pre-Christian proto-Lucian revision, the so-called kaige revision found in the Greek Minor Prophets scroll from the Nahal Hever and elsewhere in the Greek Old Testament, the translation-revisions by Aquila, Symmachus and Theodotion, the revision by Origen, that by Lucian and the disputed revision of Hesychdus.

In this connection one should be aware firstly that the existence of a pre-Christian proto-Lucianic revision is a part of Cross’s textual theory that has not convinced all scholars outside Cross’s own school;⁵ and secondly that the Lucianic text has not survived (if ever it existed) for all books of the Greek Old Testament,⁶ This means that text-histories written on the basis of the evidence taken from books where Lucian’s recension has survived and is easily identified, are not necessarily valid for other books of the Old Testament.

One further piece of information would be useful in a history of the LXX and its revisions written for beginners, and that is that we do not know exactly by whom the various parts of the Greek Old Testament were translated nor for what primary purpose. So we are obliged to speculate; and the danger in speculating is not always in speculating too much, but sometimes in not speculating enough. We must be wary of simplistic reconstructions of the past that do not allow for the fact that the ancient world was in many respects as complicated as our own. Scholars generally reject Aristeas’ story that the translations were done at the order of Ptolemy II so that the royal library should possess a copy of the Hebrew Scriptures, and they conclude —probably correctly—that they were made for the Jewish community. But that conclusion hides uncertainties: were they produced as ad hoc translations of the selected passages read in the synagogue on the sabbath? Or were they the work of a scholar, or a group of scholars, concerned to translate their national literature into Greek quite apart from the practical needs of the synagogue? And once translated, were they subject to any revision in Alexandria? Klein suggests that the curses which Aristeas says were pronounced on any who should change the translation were meant to forestall revision such as would bring the Greek closer to the Hebrew text with which it was being compared; and he considers that Cross’s suggested proto-Lucianic revision and the kaige revision were instances of such revision. But both of these revisions are presumed to have been made in Palestine. Were there never any revisions made in Alexandria, and made, not to bring the Greek nearer to the Hebrew, but to remove difficulties in the subject matter, revisions that would take the translation further from the Hebrew? We do know that Pemetrius the Jew, who lived and worked in Alexandria around the end of the third century BC, engaged with others in intense midrashic study⁷ of the Old Testament, apparently in the LXX translation; he was interested in two things in particular: in developing a detailed chronology of biblical events, and in explaining difficulties presented by the biblical narrative. In this latter activity he resembles the Greek scholar Sosibius, who lived in Alexandria in the first half of the third century BC and was known as ὁ λυτικός because of his learned attempts to explain difficulties which other scholars saw in Homer. Sosibius was famous in particular for his use of the device known to the ancients as anastrophe, i.e. mental transposition of words which, read in their original order, would make Homer appear to be saying something unreasonable. Now it is the fact that embedded in the text of the Greek Book of 1 Kings,⁸ there is an editorial note which seems to object to the order in which the MT⁹ records the building of the various parts of the temple and the royal palace: temple, temple, royal palace, temple; and insists that the order was temple, temple, temple, royal palace—which is in fact the order which we now find in the main body of the Greek translation.¹⁰ Again, another such editorial note¹¹ seems to object to the order in which the MT records the building by Solomon of certain cities in relation to the time of the completion of the Lord’s house;¹² and the editor, whoever he was, seems to have relieved the Greek translation of the ‘fault’ by removing the offending passage from the main Greek text to stand elsewhere.¹³

Now we know that later Palestinian rabbis freely used the methods of Alexandrian scholarship in their expositions of Scripture,¹⁴ though, of course, they did not alter the Hebrew text on those principles. What we have to ask ourselves is this: if Palestinian rabbis used Alexandrian methods in the interpretation of Scripture, is it likely that the earlier Alexandrian Jewish scholars did not? And secondly, did the Jews of Alexandria use all of the Greek Old Testament (if, indeed, they ever used any part of it) as a substitute for the Hebrew for the official reading of Scripture in the synagogue, or did they regard at least some parts of it as midrashic exegesis in which the devices of Alexandrian scholarship could be used, where necessary, to produce better sense than the text, as it stood before them, seemed to yield?¹⁵

Admittedly much of this is speculative; but when we are all at the stage, as Professor Cross and his school are, of constructing theoretical text-histories of the LXX, we should not content ourselves with simplistic reconstructions of the past, nor generalise too freely from one part of the LXX to another.

And with this in mind we turn now to Klein’s Chapter II.

II

Chapter II, entitled ‘The LXX and the Scrolls’, sets out to illustrate how the LXX can be used for getting behind the Masoretic stage of the Hebrew text-tradition and reconstructing the Hebrew text of pre-Christian times. The first major step in the argument is that the Scrolls have shown that when the LXX diverges from the MT, its divergences arise much more often than some scholars used to think from the variant type of Hebrew Vorlage which it was following than from any freedom of translation. Klein chooses to illustrate it by quoting the opening six verses of Exodus as they are found in the MT, the LXX and a Hebrew manuscript from Qumran, 4QEx^a. His analysis reveals first that on four of the five occasions on which the LXX disagrees with the MT, its divergence is not of its own making, since it shares its reading with another Hebrew manuscript. Secondly, it shows that nevertheless none of the three texts is in complete agreement with either of the others. Thirdly, by a qualitative analysis of the variants Klein seeks to show that none of the text types is necessarily superior. And then Klein adds: ‘Finally, it is a good example of the subjective element in the decision-making process of textual criticism. Not everyone would accept our interpretation of numbers 1, 5 and 6’.

Now where the LXX is supported in its divergences from the MT by some Hebrew manuscript from Qumran, all scholars of course, agree that here the LXX reading was not ‘invented’ by the translators but was based on a Hebrew Vorlage. Disagreement arises when the LXX diverges from the MT in places (and they are the vast majority) where no other form of Hebrew text is extant. Here a certain subjective element inevitably enters into each scholar’s decision, as Professor Goshen-Gottstein¹⁶ has helpfully pointed out, however hard the scholar may try to be objective. It is very fairminded and brave, therefore, of Klein to call attention to the subjective element in his own analysis of these verses from Exodus. On the other hand, one cannot help wondering if the amount of subjectivity which Klein shows is altogether inevitable and irreducible. Let us look closely at his analysis and grouping of the six variants (p. 14):

A Unique Reading in Qumran

1. Joseph is included among those who went to Egypt with Jacob.

Qumran and LXX Agree against MT

2. Qumran and LXX add ‘their father’ after Jacob; MT does not.

3. Qumran and LXX read ‘And all the persons of Jacob’; MT reads ‘And all the persons coming out of the loins of Jacob’.

4. Qumran and LXX read ‘seventy-five’; MT reads ‘seventy’.

5. Qumran and LXX read ‘persons. And Joseph died’. MT reads “persons. And Joseph was in Egypt. And Joseph died’.

Qumran and MT Agree against LXX

6. Qumran and MT read ‘Asher. And all the persons’; LXX reads ‘Asher. And Joseph was in Egypt. And all the persons’.

Since one of the main purposes of the list is to show that the LXX’s divergences from the MT are often based on some Hebrew Vorlage different from the MT, it is understandable that the list should show that in four of the LXX’s five disagreements with the MT it has the support of Qumran, whereas Qumran supports the MT only once.

But other analyses are certainly possible, if not preferable. The first heading ‘A Unique Reading in Qumran’ could, of course, be rephrased: ‘MT and LXX agree against Qumran’. One wonders why it is not, when variant No. 6 is not headed ‘A Unique Reading in LXX’, but ‘Qumran and MT agree against LXX’. Actually, however, it is doubtful whether variant No. 1 should be listed as a separate variant at all; it is part of a single complex variant which for some reason Klein has broken into three pieces, variants 1, 5 and 6. MT and LXX both have the explanatory phrase: ‘And Joseph was in Egypt’, but they have it in different positions. Qumran does not have the phrase: instead, contrary to MT and LXX, it includes Joseph among those who went to Egypt with Jacob. This is really a case of one variant over which all three texts disagree, or at the most two variants: one in which MT and LXX agree against Qumran, and one in which all three disagree.

More disturbing, and yet instructive because it illustrates the pitfalls that beset conjectural retroversion of the Greek, is Klein’s statement under variant 3 that Qumran and LXX read ‘And all the persons of Jacob’. The LXX does not in fact so read: it has ‘And all the persons out of (ἐξ) Jacob’. What exactly 4QEx^a read here we shall never know: the text at this point has perished.¹⁷ This uncertainty, then, over 4QEx^a’s exact reading ought strictly to eliminate its evidence; but even if the reading which Klein conjectures for it is allowed, the variant ought still to be listed under the heading ‘All three texts disagree’, and not ‘Qumran and LXX agree against MT’.

This leaves us with two variants, Nos. 2 and 4, as firm instances of agreement between Qumran and LXX against MT. The analysis, therefore, would now appear as follows:

All Three Texts Disagree

    Two variants: Klein’s Nos. 3, and 5-and-6

LXX Agrees with Qumran against MT

    Two variants: Klein’s Nos. 2 and 4

LXX Agrees with MT against Qumran

    One variant: Klein’s No. 1.

This does not, of course, do away with Klein’s point that some of the LXX’s divergences from the MT are based on other Hebrew texts; but in this instance it does considerably alter the proportions of agreement and disagreement between the LXX and the sundry Hebrew texts, and so illustrates what Klein means when he warns us that his own analysis is subjective.

Klein is not merely concerned to demonstrate that the LXX’s divergences from the MT are often based on some different Hebrew Vorlage. The question is, he says, ‘whether the divergences are merely a series of individual variants or whether different Hebrew text types existed before the MT emerged’ (p. 13). Following Cross, he believes that the divergences reveal the existence of definite text-types, and he devotes the rest of this chapter not to proving this view, but to illustrating his argument for it.

Taking as an extreme example the Samaritan text of the Pentateuch, he shows what is meant by the term ‘expansionist’ as applied to a text-type. He then suggests that in the Pentateuch the Samaritan text, some of the Scrolls, and the LXX are in varying degrees expansionist, the MT not so; that for Isaiah all texts are in different places and in varying degrees expansionist; that in Jeremiah the MT is very expansionist, while one Scroll, 4QJer^b, and the LXX are not; and that in Samuel-Kings three, if not four, types of Hebrew text can be discerned: a Hebrew text developed in Egypt witnessed to by the parts of the LXX Samuel-Kings that still contain the un-revised Old Greek translation; a Hebrew text developed in Palestine witnessed to by certain Scrolls and by the proto-Lucianic recension of the LXX; a proto-MT Hebrew text developed in Babylon and thence introduced into Palestine, witnessed to by certain Scrolls and by the kaige recension of the LXX; and the MT witnessed to by later revisions of the LXX. Both MT and proto-MT texts are defective through much accidental omission.

All these theories Klein illustrates with copious material; but to come to a balanced assessment of them, the beginner ought to be told in addition that other scholars are not equally convinced about all the elements in them, and why they are not convinced. Professor S. Talmon, for instance, rejects the theory of three local texts completely.¹⁸ Père D. Barthélemy, who identified the kaige recension, thinks that F. M. Cross should give up his ‘Egyptian Hebrew text’ of Samuel-Kings.¹⁹ And Ε. Τον questions whether there ever was a proto-Lucianic recension.²⁰

III

Professor Klein entitles his third chapter ‘The LXX—It Does Make a Difference’. The purpose of the chapter seems to be to convince the beginner, who might think that after all the study of the LXX is not likely to affect one’s understanding of the Old Testament very much, that it does and will make an enormous difference. To do this the chapter begins by assuming as fact the theories which the previous chapter was content to illustrate rather than prove (p. 13), and then selects a number of ‘dramatic examples’ to show that in certain large and important variants the LXX should be followed against the MT, even where no Hebrew evidence has survived which supports it in its disagreement with the MT. Moreover it suggests that these ‘dramatic examples’ are ‘in some respects only the top of the iceberg’ (p. 50), and so seems to imply that larger and more numerous examples of the dramatic difference made by the LXX could be found throughout the whole of the LXX. This is all exciting enough; but one should here remember Klein’s own introductory statement:

Textual criticism has often been considered a ‘safe’ discipline, in contrast to the methods of ‘higher criticism’. The reader of this book will discover that in textual criticism, too, bold subjective decisions must sometimes be made . . . (p. viii).

The first thing that strikes one about the examples offered is the simplistic rigour with which general concepts are applied to complicated particular decisions without careful weighing of all the possibilities. It is, for example, the belief of Cross and his school that the MT of Samuel is marked by extensive haplography while in Jeremiah it is expansionist. Doubtless there is truth in these evaluations. But because they may be true general descriptions, it does not mean that they are automatically true in every possible instance. Take the following two examples cited by Klein:

1 Samuel 3:15 (p. 28): And Samuel slept until the morning, and he got up in the morning, and he opened the doors . . .

Jeremiah 26:22 (p. 31): Then King Jehoiakim sent (a) men to Egypt (b) Elnathan son of Achbor and men with him to Egypt.

In the first example the phrase in italics is present in the LXX but absent in the MT; in the second, absent in the LXX but present in the MT. In the first example Klein simply asserts that the MT has accidentally lost the phrase through haplography (morning . . . morning). It could nevertheless be argued that on grounds of literary style the fuller reading of the LXX is a wooden addition concerned lest a reader of the MT’s faster moving narrative should imagine wrongly that Samuel opened the doors in his sleep without getting up. But Klein does not as much as mention any possible explanation other than haplography.

Conversely in the second passage the possibility that the LXX might have omitted the italicised phrase by haplography (to Egypt . . . to Egypt) is not even discussed. The MT in Jeremiah, it has been decided, is in general expansionist: it is therefore expansionist in this particular passage too, and a reason is sought to explain the presumed process of expansion.²¹

Take another longer example. On p. 16 Klein explains in detail what is meant by ‘expansions’ and the kind of motivation that lies behind them: they involve, inter alia, clarifications, and interpolations both large and small from parallel passages.

So when Yahweh gives Moses a message for Pharaoh in the plague account, the Samaritan Pentateuch frequently adds a passage which reports Moses’ actual deliverance of the message.

On pp. 29–31 Klein records another variant. In 2 Samuel 11, Joab sends a messenger to King David and instructs him that if, when he has related the war news, the king says X, he is to say Y. In the MT the messenger relates the war news and says Y; and we are then left to presume that the king said X. The LXX however, relates that the king actually said X word for word as Joab had envisaged. One cannot help thinking that had this occurred in the Samaritan Pentateuch Klein would have straightway dubbed it an expansionist addition. But it does not occur in the Samaritan Pentateuch but in Samuel, and therefore the possibility that the LXX text is expansionist is not even mentioned: homoeoteleuton and homoeoarchton are said to have caused the omission in the MT.

It is of course sound procedure, When all other considerations have been weighted and found indecisive, to let the known general character of the texts at variance decide the issue. But to resort at once without discussion of other considerations to the general character of the texts and to decide on this ground alone is to make the same mistake as those New Testament textual critics made who, deciding that Vaticanus B was in general a pure text, tended to follow it automatically everywhere. And it is liable to give the beginner the impression that provided one knows the basic quality of text in each book (‘expansionist’, ‘haplographic’ or whatever), decision making over any one variant is simple; whereas in fact it is much more often a complicated process.

Another doubtful piece of methodology is the way in which Klein suggests that the LXX confirms certain theories of the literary composition of Jeremiah. He starts by observing (p. 34) that certain long sections of Jeremiah are written in what is often called Deuteronomistic prose. He next observes that some scholars argue that many of those prose sections are original, while others hold that many of them are secondary. He then claims that ‘it can now be shown on the basis of the LXX that some of these passages came into the text during the process of textual transmission’. As illustration of this he cites three passages from Jeremiah where the LXX has a shorter (in two of them much shorter) text than the MT; and on the grounds that the LXX lacks the verses in question, he suggests that the verses came into the Hebrew text during the process of transmission.

But their absence from the LXX does not automatically prove that. The case first has to be argued that the absence of these particular passages from the LXX (and its Hebrew Vorlage) is original and not a subsequent omission. If that can first be proved, then it would follow that the entry of the material into the MT was late, and this in turn would be an additional reason for thinking that the material was Deuteronomistic. First, then, what is the proof that the absence from the LXX is original? Klein does not explicitly say, but his comment on the first omission is:

These verses are absent from the LXX in their entirety, and they bristle with Deuteronomistic terminology. Apparently a scribe steeped in this tradition composed these verses for insertion [i.e. in the MT] at this place.

In other words he now seems to start from ‘the acknowledged fact’ that the verses contain Deuteronomistic phraseology, and on that ground to argue that their omission from the LXX is original. But what if it were not first acknowledged that the verses are Deuteronomistic in their phraseology? Why, then Klein loses the only evidence he cites for deciding that the LXX omissions are original and the MT secondary. It is interesting to see, therefore, that in a recent article W.L. Holladay²² claims that H. Weippert’s new study of the prose passages in Jeremiah ‘Source C’ demonstrates that they have no connection at all with the literary circles associated with Deuteronomy or the Deuteronomistic historical work, but rather with Jeremiah’s own poetic diction and theological work. Now this is no place to enter into this large controversy. The point is simply that if the claim of Deuteronomistic origin is disputed, it would seem unwise to cite it as front-line evidence of the LXX’s originality; and if the LXX cannot be proved original on other grounds, one cannot cite the LXX’s ‘originality’ to bolster the claim that the verses are Deuteronomistic additions; for that way lies circular reasoning.

Klein then (p. 35) goes further and insists that the LXX must be used for investigating the broader questions of the book’s composition, and cites as an example the messianic passage 33:14–26. He says that literary critical judgement shows 33:14–26

to be a late insertion, updating 23:5–6. While the latter speaks of God raising a righteous branch for David, 33:14–26 calls Jerusalem a righteous branch and promises many successors for the Davidic kingdom and the Levitic priests. Literary criticism indicates that 33 is from a much later writer who is trying to apply Jeremiah’s words to the situation of the post exilic Jerusalem. In this case, the literary critical judgement is confirmed by the text critical data since 33:14–26 is absent from the LXX.

But something has surely gone wrong with the literary judgement here, or at least Klein’s version of it. Jeremiah 33:14–26 nowhere calls Jerusalem a righteous branch. What perhaps Klein meant to say is that while 23:5–6 calls David’s descendant both ‘a righteous branch’ and ‘The Lord our Righteousness’, 33:14–16 calls David’s descendant ‘a righteous branch’ and Jerusalem ‘The Lord our Righteousness’. But the mistake²³is worth citing as an example of the ease with which subjective literary judgements can err through inexact observation.

The next example to be cited as proving that the LXX does make a difference is the radical disagreement in chronology between the MT and the LXX in Kings (pp. 36ff). J. D. Shenkel has shown that of the two divergent chronologies to be found in the Greek, the one that agrees with the MT is the result of a later revision of the Greek (the kaige revision, in fact) and has concluded therefore that the other chronology is the Old Greek one. Klein follows Shenkel in thinking that the Old Greek represents a Hebrew chronology that may be older than the MT, though in fairness he warns his readers that ‘a completely accurate absolute chronology cannot yet be constructed’ (p. 36). In a way this last clause is an understatement. The fact is that the Old Greek chronology, in favour of which both Shenkel and Klein would discard the MT’s scheme, is self-contradictory, incomplete and incompletable. At the crucial point that has determined their peculiar order of text both the B-text and the Lucianic are disastrously confused. The Lucianic chronology is an unintelligent hotch-potch of the MT and the B-text. The B-text, though slightly more intelligent, is shown to be secondary by a major discrepancy and a consequent embarrassed omission of two normal sychronisms.²⁴ (See further pp. 15–16 below.)

Believing that the Greek chronology and the order of the narrative which it demands are original, Klein accepts Shenkel’s speculations on why the MT has changed both the chronology and the narrative-order. He states those speculations in suitably tentative language, though in his final argument (p. 39) he again shows his tendency not to cite all the evidence:

Finally, we learn from 1 Kings 22:48f and 2 Kings 8:20 that there was no king in Edom during the reign of Jehoshaphat. Since the king of Edom is explicitly mentioned in 2 Kings 3, MT’s identification of the originally anonymous “king of Judah” as Jehoshaphat is again shown to be errant.

But the crucial verse on which Klein bases his objection to the MT’s account, namely 1 Kings 22:48f, does not say simply that in Jehoshaphat’s day ‘there was no king in Edom’, but qualifies that remark with ‘a deputy was king’. Klein neither mentions this, nor explains why the MT of 2 Kings 3 could not be referring to this deputy.

As another example of the difference that the LXX could make if it were proved original Klein offers (p. 40) the LXX’s first account of Jeroboam’s rise to power (1 Kgs 12) which in an important respect differs from the MT. Klein thinks the LXX is original and the MT secondary, though again in all fairness he warns his readers of the tentative nature of his reconstruction.

To attain any cogency Klein’s reconstruction would of course have had to take into account that lying beside this version of the rise of Jeroboam, which Klein regards as Old Greek, there is another and very different version, which nevertheless bears in its language the marks of being Old Greek older Greek perhaps, and certainly more stylish than the Greek of the first version. Moreover it seems possible that the second version has influenced the first version precisely in those features of the first version which Klein considers original.²⁵

The remainder of Chapter III is devoted to showing how our understanding of the Chronicler’s work has been considerably helped by the discovery at Qumran of Hebrew texts of Samuel-Kings that differ from the MT and clearly resemble the Chronicler’s Vorlage. They show that many differences between the MT of Samuel-Kings and Chronicles were not due to the Chronicler’s ‘bias’ as earlier scholars thought, but to his Vorlage. Klein also cites examples to show that the Chronicler sometimes agrees with the LXX against the MT of Samuel-Kings in places where no Hebrew text-tradition other than the MT has survived. Klein would take this agreement to prove that the LXX is here based on a Palestinian Hebrew text, whereas the MT is a text developed in Babylon. One major point calls for elaboration. In the course of refuting the idea that the Chronicler omitted the information given in 1 Kings 9:16 because he considered it derogatory to the reputation of Solomon, Klein, following W. E. Lemke, observes (p.46)

that the LXX of Kings omits all of verses (9):15–25 at this point although some manuscripts do include some of these verses after 10:22. Even there, however, verse 16 is only included in Hexaplaric manuscripts where it is identified as an addition from the Hebrew by an asterisk.

Now it is true that the presence of verse 16 in the Hexaplaric manuscripts alone betrays it as an Origenic addition at this point (though actually the addition stands in chapter 9 and not in chapter 10 as Klein and Lemke seem to imply); but it is not true to say as Lemke does²⁶ that the verse ‘is not to be found in our best manuscripts of Bas (= Kings)’. It is to be found stationed at 5:15b (Rahlfs’ enumeration) as part of the LXX’s radical rearrangement of the material of 1 Kings.

5:14a (= MT 3:1b)	Solomon puts Pharaoh’s daughter in David’s City until completion of the Lord’s house, his own house, and the wall of Jerusalem.
5:14b (= MT 9:16)	Pharaoh’s parting-gift of Gezer to his daughter.
5:15–9:9	Building of the Lord’s house and Solomon’s palaces. Dedication of house. When all the building is finished, God appears to Solomon
9:9a (= MT 1 Kings 9:24 = LXX 1 Kings 2:35f	Solomon brings up Pharaoh’s daughter out of David’s City into his own house.

The rationale of this arrangement of the text is easy to be seen. 5:14a says that Solomon put Pharaoh’s daughter in David’s city until certain building should be complete. The very moment it is complete, 9:9a is there to say that Solomon brought her up from David’s City. In other words, the peculiar arrangement of the LXX’s text has for one, at least, of its purposes a desire for exact timetabling. We must ask therefore if this peculiar arrangement of the LXX contents is in fact the order given to the material by the original translators; is it what Cross and his school would call Old Greek? And if it is, does it go back to a Hebrew Vorlage? And if it does, has that Vorlage any claim to represent the original Hebrew?

Here we can call on Cross’s own insight, which Klein has put forward in this chapter, that where the MT of Chronicles disagrees with the MT of Kings, that is because the MT of Chronicles is a Palestinian text, the MT of Kings a Babylonian; and further that, in these disagreements the reason why the MT of Chronicles is often supported by the Old Greek of Kings is because the LXX was translated from a text that was basically Palestinian (but, as Cross would have it, developed somewhat in Egypt). Consider then the table below. The Greek of 2:35f^βγ agrees with the MT of 1 Kings 9:24 exactly except for its first word. The Greek of 1 Kings 9:9a is nearer to the Hebrew of 2 Chronicles 8:11 than to that of 1 Kings 9:24 in the variant θυγάτηρ Φ. ἀνέβαινεν/ἀνήγαγεν Σ. τὴν θυγατέρα Φ. This nearness we could perhaps attribute to the Palestinian basis of LXX’s Vorlage. But the Greek of 1 Kings 9:9a disagrees with the Hebrew of Chronicles very markedly, and particularly in its timetable notes at the beginning and at the end, which as we have seen are the key elements in the Greek which relate this verse to the Greek’s highly peculiar ordering of the contents of 1 Kings. It would appear then that the Greek’s special interest in timetabling and the special ordering of the text that goes with it are not to be traced back either to a Palestinian²⁷text or to a Babylonian; and they have little claim to represent the original Hebrew. But a thorough-going interest in a timetable that is prepared to re-order the contents of the book, is not far removed from an interest in chronology. The Greek’s special chronology likewise involves a special ordering of some of the contents of Kings. Shenkel and Cross think that the chronology represents the original Hebrew. But if the timetabling does not, is it likely that the chronology does?

MT	LXX	LXX	MT	LXX
1 Kgs 9:24	1 Kgs 2:35fβγ	1 Kgs 9:9a	2 Chr 8:11	2 Chr 8:11
אך	οὕτως	τότε	יאת־	καὶ τὴν
בת־	θυγάτηρ	ἀνήγαγεν	בת־פר׳	θυγατέρα Φ.
פרעה	Φαραω	Σαλωμων	העלה	Σαλωμων
עלתה	ἀνέβαινεν	τὴν θυγατέρα Φ.	שלמה	ἀνήγαγεν
מעיר	ἐκ τῆς πόλεως	ἐκ πόλεως	מעיר	ἐκ πόλεως
דוד	Δαυιδ	Δαυιδ	דוד	Δαυιδ
אל־	εἰς τὸν	εἰς		εἰς τὸν
ביתה	οἶκον αὐτῆς	οἶκον αὐτοῦ	לביתַ	οἶκον
אשר	ὅν	ὅν	אשר	ὅν
בנה־	ᾠκοδόμησεν	ᾠκοδόμησεν	בנה־	ᾠκοδόμησεν
לה	αὐτῇ	αὐτῷ	לה	αὐτῇ
אז	τότε	ἐν
בנה	ᾠκοδόμησεν	ταῖς
את־	τὴν	ἡμέραις
המלוא׃	ἄκραν	ἐκείναις

IV

Chapter IV, entitled ‘The Greek and Hebrew Evidence’ is a helpful introduction for the beginner to the basic tools he will need to know, and know how to use, in the textual criticism of the Old Testament.

V

The final chapter, ‘Doing Textual Criticism’ shows how all the facts and theory enunciated so far should be applied to the practical decisions that have to be taken over individual variants. It therefore repeats a lot of what has been already said, particularly Cross’s theory of local texts and the arguments for the existence of a proto-Lucianic recension. But in the end, after all the dramatic examples calculated to rouse the interest and enthusiasm of the beginner, when it comes to practical advice Klein very wisely and sensibly urges a cautious attitude. ‘The MT itself’, he says (p. 74),

deserves very high respect and should be changed only with great caution. In cases where the criteria listed below under internal considerations are indecisive, the student is well-advised to prefer the MT.

No better advice could be given.²⁸

Footnotes

¹ Notably J.D. Shenkel, Chronology and recensional development in the Greek Text of Kings, Harvard University Press, 1968, p. 123; J. G. Janzen, Studies in the text of Jeremiah, Harvard University Press (1973) 173–18.

² Studies in the Greek texts of the Chronicler, Harvard Divinity School, 1966, unpublished; summarised in HTR 59 (1966) 449. See also HTR 62 (1969) 99–107.

³ See G. Gerleman, Studies in the Septuagint III, Proverbs, Lund, 1956, especially 17ff.

⁴ The Greek Esther is a notable example.

⁵ See E. Tov, ‘Lucian and Proto-Lucian’, RB 79 (1972) 101–113, and Cross’s reply, 1972 Proceedings, Septuagint and Cognate Studies 2, ed. R.A. Kraft, 118f.

⁶ Footnote text here:See especially J. W. Wevers, Text History of the Greek Genesis, Göttingen (1974) 175.

⁷ See F. Jacoby, Die Fragmente der griechischen Historiker, Dritter Teil, Leiden (1969) 666–71, and especially the use of ἐπιζητεῖν in fragment 5. S. Lieberman, Hellenism in Jewish Palestine, New York (1962) 48 has taught us to see in ζητέω and its cognates and compounds the Greek equivalent of מדרש.

⁸ At 1 Kings 2:35c.

⁹ At 1 Kings 6–7.

¹⁰At 1 Kings 6–7.

¹¹ At 1 Kings 2:35k.

¹² First Kings 9:17–19 records the building of the cities, 9:25 the completion of the Lord’s house.

¹³ Only part of the MT paragraph 1 Kings 9:15–25 is to be found at LXX 1 Kings 10:23–25 (Brooke-McLean’s enumeration). For the remainder of the material see LXX 1 Kings 2:35^c, ^f, ^g, ^h, ^I; 46^c, ^d.

¹⁴ See S. Liebermann, op. cit. 47ff.

¹⁵We must distinguish between the superstitious attitude to the LXX shown in later times by such people as Philo (On Moses, 2.25–44) and the much more sober attitude evidenced in Aristeas (Ad Philocratem 302).

¹⁶ In ‘Theory and practice of textual criticism: the text critical use of the Septuagint’, Textus, III, 130ff, especially 133.

¹⁷ Perhaps what Klein meant to say was what Cross observes (Ancient Library of Qumran 1st edition, London (1958) 138 n.31): ‘In reconstructing 4Q, it becomes clear that there is insufficient room for the full reading of MT. 4Q like LXX appears to be shorter. As for the original reading, we cannot speak with any certainty’. Klein’s mistake is in going further than this and stating that Qumran and LXX both read ‘And all the persons of Jacob’. He points out (p. 15) that ‘both methods of indicating descent’ [i.e. the longer ‘all the persons coming out of the loins of Jacob’, and the shorter ‘and all the persons of Jacob’] are contained in the MT of Gen 46:26’. But the shorter one reads כל־הנפש הבאה ליעקב מצכיםח, and the Greek, which does not understand the phrase to indicate physical descent from Jacob, but geographical descent into Egypt with Jacob, renders πᾶσαι δὲ ψυχαὶ αἱ εἰσελθοῦσαι μετὰ Ιακωβ εἰς Αἴγυπτον. In the next verse (Gen 46:27) the plain genitive of the Greek (πᾶσαι ψυχαὶ) οἴκου Ιακωβ corresponds to the MT’s לבית־יעקב. When therefore, the Greek of Exodus 1:5 uses neither μετὰ Ιακοωβ, nor the plain genitive, but ἐξ Ιακωβ, can anyone tell us for certain that it represents, say, a plain construct, and not + ל nor + מ nor anything else? And can he guarantee that whatever it represented was the same as what 4QEx^a had, whatever that was?

¹⁸ The Cambridge history of the Bible, I (1970) 193–199.

¹⁹ 1972 Proceedings, Septuagint and Cognate Studies 2, ed. R.A. Kraft, 61, and Cross’s reply, 121.

²⁰ See footnote 5 above.

²¹ Klein considers that the MT has conflated two synonymous variants. This is a basic idea which he acknowledges comes from S. Talmon, ‘Double readings in the Masoretic Text’, Textus 1 (1960) 144–184; 4 (1964) 95–132.

²² ‘A fresh look at “Source B” and “Source C” in Jeremiah’ VT 25 (1975) 395. 402–412.

²³ The book shows signs of having been hastily compiled. My colleague, Mr. D. F. Payne, points out that the sub-title on the cover reads ‘From the Septuagint to Qumran’, but on the title page ‘The Septuagint after Qumran’. Actually the sub-title ‘The Septuagint after Qumran’ would have been a fairer indication of the scope of the book than ‘Textual Criticism of the Old Testament’. Several matters which a beginner needs to know in order to understand, say, the footnotes in RSV, receive very short shrift. Professor K.G. O’Connell (CBQ 37 (1975) 266–267) complains that on pp. 6 and 24 his views have been wrongly reported.

²⁴ See JTS 21 (1970) 125.

²⁵See now R. P. Gordon, The second Septuagint account of Jeroboam: history or midrash?’, VT 25 (1975) 368–393 (published since Klein wrote).

²⁶ ‘The synoptic problem in the Chronicler’s history’, HTR 58 (1965) 359 n. 33.

²⁷ The LXX’s special timetable disagrees with Josephus in another respect. Josephus, Ant., VIII 99ff. has it that the grand dedication service for the temple was held immediately the seven years spent in the building of the temple were finished. The LXX goes out of its way to insist (1 Kgs 8:1) that it was not held until twenty years (seven for the building of the temple and thirteen for the building of the royal palaces) had gone by.

²⁸ The diagram of the ‘Transmission of Samuel and Kings’ (p. 71) could unintentionally give rise to misunderstandings. Its two separate lines are not meant to give the impression that the Hebrew text did not originate in Palestine before copies were taken to Babylon. Nor does it wish to imply that the kaige recension was done in Babylon (most think it was done in Palestine). The diagram in spite of its title was not in fact meant to represent the full story of the transmission of the text, but simply where the different text-types were principally developed. For a better diagram see S. Talmon, Cambridge History of the Bible, I, 195.