Retroversion in Medieval Jewish Biblical Exegesis: A Study in the Book of Jeremiah

Ben Zion Katz is an Associate Professor of Pediatrics at Northwestern University School of Medicine and an Attending Physician at the Children’s Memorial Hospital in Chicago. He also teaches at the Florence Melton Adult Mini-School Institute (JCC Chicago) of the Melton Center for Jewish Education in the Diaspora at Hebrew University in Jerusalem.

RETROVERSION IN MEDIEVAL

JEWISH BIBLICAL EXEGESIS:

A STUDY IN THE BOOK OF JEREMIAH

Modern biblical scholarship is perceived as operating under a set of

assumptions different from those of traditional Jewish Bible study.

Traditional Jewish study of the Bible is thought to take for granted the sacred and immutable nature of the masoretic text. Critical biblical scholarship, on the other hand, in many instances will suggest corrections for what are deemed to be mistaken masoretic readings. Many of these proposed

emendations are suggested by ancient translations such as the Greek

Septuagint,1

or ancient texts such as the Dead Sea Scrolls found at Qumran.2

In the case of the Qumran scrolls, actual manuscript evidence of a different Hebrew text from the masoretic may exist. In the case of the Septuagint, one seeks to determine whether the Hebrew text from which the Greek translation was made (referred to by scholars by the German word Vorlage) may have

differed from the masoretic based on a conjectured “back-translation” or retroversion from the Greek.

Medieval Jewish commentators probably knew nothing of the Dead Sea

Scrolls, and were likely unfamiliar with the Septuagint except as it was referred to in classical sources.3

However, they were keenly aware of at least

some of the Aramaic Targumim, particularly those of Onkelos, the official Targum of the Torah, and Jonathan, the Targum on the Prophets.4

Nahum

Sarna has previously shown that many medieval Jewish Bible commentators, especially those of Sephardi heritage, have used linguistic techniques similar to those of so-called “lower” biblical criticism in their commentaries, which include the “modern” techniques of comparative linguistics.

For example, it is well known that as far back as the Talmud, biblical

Hebrew was sometimes explained using cognate Aramaic or even Arabic

roots. It is also well established that techniques such as ellipsis (that is, particles or even entire words that may be missing from the biblical text) or

substitutions (that is, what the masoretic text “really” meant) were used by Ibn Janach and others.5

In this paper I will demonstrate that retroversion was also known to and used by medieval and pre-modern biblical commentators. I will show this using examples from the commentaries of Rabbi David ben-Joseph Kimhi

(Radak; ca. 1160 - ca. 1235) and Solomon Yedidiah ben-Abraham Norzi

(1560-1626) whose commentary is called Minchat Shai. Most of the

examples cited will be from the Book of Jeremiah. In the course of the

discussion I will also compare the comments of these authorities on the verses cited to those of the standard modern commentators6

on these same verses.

AL OR V’AL

The first example is from Jeremiah 17:2, which can be translated, following the masoretic accentuation, as follows: As their children remember their altars and their sacred poles, by verdant trees, upon lofty hills. Norzi commented on the verse, emphasizing that the verse should not end with by verdant trees and upon lofty hills even though it appears to be so from the Targum, because the Targum added a “vav” to the third from last word of this verse. In other words, Norzi realized that the Targum added a particle to this verse that is not present in the Hebrew. In this case he did not postulate that the Targum was translated from a different Hebrew text (Vorlage) than the one we have before us, even though he found such a manuscript. According to the Biblia Hebraica Stuttgartensia (BHS) the latest edition of the only critical edition of the entire Bible, other ancient translations such as Theodotian and some Syriac and Arabic translations also translated as if the “vav” were present.7

(The “vav” allows one to change the structure of the last

half of the verse. Disregarding the trope and adding the “vav,” one can translate: As their children remember their altars and sacred poles, on verdant trees and upon lofty hills, allowing the sentence to flow more

smoothly.)

None of the modern commentators mentions the Targum or any of the other variants in BHS. Even Carroll, who adds the particle “and” to his translation, omits this discussion.8

RETROVERSION IN MEDIEVAL JEWISH BIBLICAL EXEGESIS: JEREMIAH

BETZUROT - NETZUROT

In the next examples, however, instead of warning the reader not to follow the Targum, Kimhi and Norzi assume the Targum had or may have had a

different Hebrew text (Vorlage) from which it was translated. For Jeremiah 33:3, Norzi utilized the Targum to postulate a different reading than is found in the masoretic text. This verse can be translated literally as: Call to Me, and I will answer you, And I will tell you wondrous and fortified [betzurot] things you have not known. The obvious difficulty is with the word

“betzurot,” which I have translated using its typical meaning of “fortified.” Rashi translated the word as “guarded” or “secret” [shemurot belibi la-asot], which makes the verse read smoothly, but is not what betzurot typically means.

Norzi commented that in the book of the Targum (literally, in the Hebrew book of Jonathan ben Uzziel, which we would call the Vorlage of Targum

Jonathan) it was written “netzurot” [guarded, treasured, watched over], as in the almost identical phrase found in Isaiah 48:6, because Targum Jonathan translated betzurot as “nateran,” from the root “natar” [a guard]. Norzi continued: “And so it was written [also] in the book of Rashi because he translated [the word as] guarded [shmurot belibi la-asot].” In other words, Norzi added that Rashi’s Vorlage also read netzurot, thereby retroverting from the Aramaic and postulating an original Hebrew reading of netzurot, using both the Targum and Rashi’s commentary as proof. He does this despite the fact that Rashi may not be an independent witness, since Rashi had the Targum in front of him and often translated according to the Targum

(although he does not use the term “ketargumo,” denoting the Aramaic

Targum, on this verse). It is not important whether or not one can adduce from the evidence here that the Vorlage of Rashi or Targum Jonathan actually read netzurot, only that Norzi felt that this was the case. Also, the lecto difficilior argument, that the more difficult text is probably original, is irrelevant to the current discussion; what is important again is that Norzi believed this was the case based on the evidence available to him.

Incidentally, such comments of his are not limited to Nakh. To cite one example from the Torah: Rashi, Ibn Ezra and Hizkuni commented on the

phrase ve-et kol asher atzaveh [and all that I will command you] in Exodus 25:22, especially its initial “vav.” In the masoretic text, however, the phrase

does not start with a “vav.” Norzi there commented that the “book of Rashi” [as well as that of Ibn Ezra and Hizkuni] had an extra “vav” [shekakh hayah katuv besifrayhem].

Oddly, the alternate reading of netzurot in Jeremiah 33:3 is not accepted as superior by all moderns.9

Bright and Carroll mention that netzurot is found in

a few manuscripts (Oriental manuscripts according to BHS10

) and note the

similarity to the verse in Isaiah 48:6. Holladay most emphatically stated that netzurot was “doubtless a secondary correction.” McKane, however, did call it “probably the right nuance.”

BAGDAROT - BEGDUDOT

The next example is from Jeremiah 49:3, whose beginning can be literally translated as: Howl, O Heshbon, for Ai is ravaged! Cry out, O daughters of Rabbah! Gird on sackcloth, lament, And run to-and-fro in the [sheep] pens [bagderot]. The difficulty here is with the word “bagderot” which translated literally means “sheep pens” but which the Targum rendered as “in

companies“ [besayan], which smoothes out the reading but is not what the Hebrew word bagderot usually means. Kimhi recognized the implications of the Targum’s translation and stated: “And I am surprised at how the Targum . . . translated . . . in groups . . . [for] it appears that he read begdudot . . ., as is written in Jeremiah 48:37. There are no other places in Tanakh where the Targum translates bagderot as besayan.“11

(Kimhi did not state that Rashi

also seemed to have understood the Targum in this manner.) Kimhi therefore used similar reasoning to that of Norzi in the previous example, but did not conclude (as Norzi did previously) that the Targum had a different Vorlage. This is consistent with his comments in Isaiah 38:19 and elsewhere, as noted previously by Uriel Simon and Frank Talmage.12

There, Kimhi also thought

that Targum pseudo-Jonathan and Ibn Ezra commented on the text as if it were vocalized differently, and was again surprised because he could find no evidence of such a vocalization.13

Again, as pointed out by Simon and

Talmage, Kimhi will often use the lack of a manuscript as evidence against the apparant Vorlage of the Vulgate, where certain Christologic references may be found. Perhaps it is for that reason that Kimhi was reluctant to retrovert from the targumim without a corroborating text.

RETROVERSION IN MEDIEVAL JEWISH BIBLICAL EXEGESIS: JEREMIAH

Modern commentators differed as to whether to adopt the reading of

begdudot. Bright and Carroll cited it as a possible reading. Holladay prefered the masoretic text. McKane, in a thorough discussion of the issue, concluded that “[i]t would be remarkable if Targum had a Hebrew text with” begdudot because there is no other evidence of its existence and because the correct plural of “gdud” is “gedudim,” not “gdudot.“14

(This could be another

unstated reason why Kimhi did not want to accept the possibility of begdudot as the Vorlage of the Targum.)

AHARAYHEM - AHARETAM

The final example is found in Jeremiah 50:21, where the text reads literally as: Advance against her – the land of Meratayim – And against the

inhabitants of Pekod; utterly destroy them after them [aharayhem] – says the Lord. Here the obvious difficulty is that the last words of the text are awkward. The Targum rendered the word “aharayhem” as “she-aret-hone

[their remnant],“ making for a smoother reading. Radak commented that

Targum Jonathan “rendered [the word] aharayhem as if aharetam [their

remnant] [was written]“ as in Amos 9:1, where the word “aharetam” is used in this sense. Here Radak seemed to be working with the notion that the Targum may have translated from a different Vorlage, and he did not express as much surprise as he did in the previous example.

Again, the modern commentators varied in their enthusiasm for this

possible emendation. Bright mentioned the reading aharetam in a footnote. Holladay omitted the word aharayhem from his translation altogether, as did the Septuagint. Carroll did not mention the variant, and, as in the previous example, translated according to the masoretic text. McKane, in another thorough discussion, noted the Targum’s translation, but concluded, like Holladay, that the word aharayhem should be deleted.15

In this survey, essentially from a single biblical book, several examples were cited of retroversion, in which Norzi and perhaps Kimhi postulated a different Hebrew Vorlage for the Targum in several verses where the

masoretic text was problematic. It is important to note that not all p’shatoriented medieval commentators were open to this idea. As shown by Uriel Simon, as careful a reader of the biblical text as Ibn Ezra, was generally not predisposed to this type of analysis.16

Without an ancient translation or

manuscript as a basis, or evidence from multiple earlier authorities, Norzi would likely reject an alternative text, as he rejected the reading of “yagodu” for “yagoru” in Psalms 56:7, despite the fact that Ibn Ezra may have had such a text.17

Nevertheless, we can discern from a careful reading of classical

Jewish biblical commentators, both “standard” Jewish interpretations of Scripture and also certain types of modern “critical” interpretations.

NOTES

1. Despite the use of the phrase “the Septuagint,” there is not one single Septuagint translation. See the articles by L.J. Greenspoon. “Mission to Alexandria: The truth and legend about the creation of the Septuagint, the first Bible translation,” Bible Review, 5:34-41 (August 1989), and M.K.H. Peters, “Why study the Septuagint?” Biblical Archeologist, 36:174-181 (September 1986). See also the brief discussion of the Septuagint by Louis Jacobs in Principles of Jewish Faith, An Analytical Study, (London: Valentine Mitchell, 1964) pp. 254-255 and the discussion by Emmanuel Tov in his magisterial Textual Criticism of the Hebrew Bible (Minneapolis: Augsburg Fortress, 1992) esp. pp. 134-48.

2. For examples, see Tov, esp. pp. 100-17; F.M. Cross and S. Talmon, Qumran and the History of the Biblical Text (Cambridge, MA: Harvard University Press, 1975).

3. E.g., Soferim 1:8, B. Meg. 9a-b.

4. Encyclopedia Judaica, Vol. 4, columns 841-51 (Jerusalem: Keter, corrected edition [no date]).

5. Nahum Sarna, “Hebrew and Biblical Studies in Medieval Spain,” in The Sephardi Heritage: Essays on the history and cultural contribution of the Jews of Spain and Portugal, ed. by R.D. Barnett, Vol. I: The Jews in Spain and Portugal before and after the expulsion of 1492 (NY: Ktav, 1971), reprinted in JPS Scholar of Distinction Series, N.H. Sarna, Studies in Biblical Interpretation, (Philadelphia: Jewish Publication Society, 2000) pp. 81-125. 6. W. McKane, A Critical and Exegetical Commentary on Jeremiah, (The International Critical Commentary) (Edinburgh: T & T Clark, Ltd, 1986); R.P. Carroll, Jeremiah: a commentary, (The Old Testament Library) (Philadelphia: Westminister Press, 1986); W.L. Holladay, A Commentary on the Book of the Prophet Jeremiah (Hermeneia), (Philadelphia: Fortress Press, 1986) Vol. 1, chapters 1-25, Vol. 2, chapters 26-52, 1989; J. Bright, The Anchor Bible Jeremiah (NY: Doubleday, 1965).

6. R. Kittel, Biblia Hebraica Stuttgartensia (Deutsche Biblegesellschaft, 1990) p. 815. 8. Holladay, Vol. 1, p. 483; McKane, p. 384; Carroll., p. 348; Bright, p. 114. 9. Holladay, Vol. 2, pp. 221, 224; McKane, p. 855; Carroll, pp. 632-3; Bright, pp. 292, 296. 10. Kittel, p. 850.

7. There are, however, at least two references in the Tanakh where the Targum translates the root “gadod” with the root “sayan” (Mic. 4:14, Jer. 5:7).

8. F.E.Talmage, David Kimhi: the man and his commentaries (Cambridge and London: Harvard University Press, 1975) pp. 54-185; U. Simon, “Rabbi Abraham Ibn Ezra and Rabbi David Kimhi - two approaches to the question of the reliability of the traditional text“ (Hebrew), Shnaton Bar Ilan, Vol. 6, 1968, pp. 191-237.

RETROVERSION IN MEDIEVAL JEWISH BIBLICAL EXEGESIS: JEREMIAH

13. Kimhi made similar comments referring to II Samuel 19:15, Isaiah 43:14 and 58:3 and Hosea 12:1, and to a lesser extent in Jeremiah 11:2.

14. Holladay, Vol. 2, pp. 366, 369; McKane, pp. 1207-8; Carroll, pp. 798-99; Bright, p. 324. 15. Holladay, Vol. 2, p. 393; McKane, pp. 1271, 1274-5; Carroll, pp. 825, 827; Bright, p. 342. 16. op. cit.

15. But see S.Z. Leiman, “Masorah and Halakha: a study in conflict,” in Tehillah Le-Moshe (Greenberg), ed. Cogan, 1997, for a fascinating study on medieval opinions regarding altering the masoretic text when it was obvious the Talmud had a different reading on halakhically relevant verses.

QQUUEESSTTIIOONNSS

FFRROOMM RRAABBBBII HHAAYYYYIIMM HHAALLPPEERRNNSS BBOOOOKK

TTOORRAAHH DDIIAALLOOGGUUEESS

1. In Genesis 12:1, God instructs Abraham to leave his

land, his kindred and his father’s house. Which is the

most difficult of the three?

2. In Genesis 21:9-21, Ishmael is referred to in various

ways. (The English translations do not convey all the

variations.) Find the significance of each appellation.

3. What previous occurrence might Laban be alluding

to when he says: In our place this is not done, to put

the younger before the elder (Gen. 29:26)?

RESPONSES ON PAGE