¹ For bibliography on the discussion to date, see Cross, F. M., The Ancient Library of Qumrân, rev. ed. (New York, 1961)Google Scholar [hereafter ALQ ²], especially the footnotes to chapter IV; The History of the Biblical Text in the Light of Discoveries in the Judean Desert, HTR 57 (1964), 281–99; and a forthcoming article, The Contribution of the Discoveries at Qumrân to the Study of the Biblical Text. As examples of old problems which have been treated from the new vantage point, we may cite the monograph of Barthélemy, D., Les devanciers d'Aquila (Leiden, 1963)Google Scholar, and the unpublished Harvard dissertations of Werner E. Lemke, Synoptic Studies in the Chronicler's History (1964), J. Donald Shenkel, Chronology and Recensional Development in the Greek Text of Kings (1964), Bruce K. Waltke, Prolegomena to the Samaritan Pentateuch (1965), and Ralph W. Klein, Studies in the Greek Texts of the Chronicler (1966).

² Eissfeldt, Otto, The Old Testament: An Introduction (New York and Evanston, 1965), 349Google Scholar; Einleitung in das AT ³ (Tübingen, 1964), 470. In this third edition, Eissfeldt has added the following words: “Among the fragments of the Hebrew text of Jeremiah which have come to light near Qumrân there are texts (alongside some which represent or an earlier form of it) which attest the shorter form presupposed by .”

³ Studies in the Text of Jeremiah, unpub. diss., Harvard University, 1965.

⁴ Perhaps the prime example of this type of textual development is the Lucianic recension of the Septuagint. For convenient description of the character of this recension, see the introductions to J. Ziegler's critical editions of the Septuagint.

⁵ For a thorough discussion of the origin and formation of double readings in the Massoretic text, together with abundant examples, see Talmon, S., Double Readings in the Massoretic Text, Textus 1 (1960), 144–84Google Scholar; Synonymous Readings in the Old Testament, Scripta Hierosolymitana 8: Studies in the Bible, ed. Rabin, Chaim (Jerusalem, 1961), 335–83Google Scholar; and Aspects of the Textual Transmission of the Bible in the Light of Qumran Manuscripts, Textus 4 (1964), 95–132.

⁶ Nachrichten der Akademie der Wissenschaften in Göttingen: I. Phil.-Hist. Klasse (Göttingen, 1958), Dritter Beitrag.

⁷ Where reference is by name alone, the work referred to is the most recent edition of that author's commentary on Jeremiah, in loc.

⁸ “Leading in the way/wilderness” is a common motif in the O.T. Note the correction of L–62–613 Arm Tht: εν τω καιρω αγοντος σε εν τω ερημω!

⁹ In 23.17, θελημασιν probably renders, not חזךךש, but חזצעמ. (In all other occurrences of the cliché in Jeremiah, -/בל is rendered της καρδιας/-; it is unlikely, therefore, that τοις θελημασιν αντων translates ךבל חזךךשב. -Vorlage of 23.17 probably read םיכלח לכו םחיחזצעמב.) If such is the case, 23.17 and illustrate the rise of the variants conflated in 7.24 .

¹⁰ Cn. Volz, Paul (Studien zum Text des Jeremia: BWAT 25 [Leipzig, 1920], 60Google Scholar): “Es empfiehlt sich nicht, חזךךש zu streichen, da dieser Ausdruck dem Bearbeiter, der hier schreibt, von jeremian. Texten her geläufig war; andererseits wüsste man nicht zu sagen, wie חזצעמ hereingekomraen sein sollte.” But Volz's attempt to apply the principle lectio difficilior will not stand up. For one thing, a possible source for חזצעמב is easily found — as others have already pointed out, Ps. 81.13 provides ample occasion for the secondary intrusion of this word into Jer. 7.24. On the other hand, it is unlikely that the original text would have such a variant in an otherwise set phrase for which (in contrast to the synonymous variants referred to immediately below) there is no synonymous variant in the O.T. The secondary infection of a standard cliché which originally was fixed in its literary context may be illustrated in two interestingly related series of expressions: (a) Jer. 30.11 (םש) ךיחוצפת (ךשא םיונח לבב) vs. 46.28 (and 10 other places) ךיחחךח. The atypical verb in 30.11 occurs, significantly, in a passage which is omitted in and is a secondary doublet of 46.28. (b) Ezek. 4.13 (םש) םח;יךא (ךשא םיזגב) vs. Ezek. 20.41 (and 5 other places similarly) בחצפנ. Again, the atypical usage in Ezekiel 4.13 is about in , and is to be taken as cross-infection from Jeremiah.

¹¹ It is odd that the Hexaplaric “correction” is not really a correction to (pace Ziegler's apparatus), but a conflation of εξανηλωσαν and its later replacement, συνετελεσαν. The absence of a Hexaplaric correction to the full text of further suggests the latter's late, conflated character.

¹² ע and ש are easily confused in some periods; see, for example, Cross, F. M., The Development of the Jewish Scripts, The Bible and the Ancient Near East, ed. Wright, G. E. (New York, 1961), 137, fig. 1, lines 3, 5Google Scholar.

¹³ Interestingly, the verb of -Vorlage is the verb of (B)-Ketib! That is, following זעמל, the form would have to be pointed as an infinitive. The Ketib suggests a Hebrew text tradition in which (B) contained the negative אל זעמל, allowing a perfect form to follow.

¹⁴ In 7.6, םכל עךל is translated εις κακον υμιν. -Vorlage may have read םבל עךחל here (F. M. Cross has informed me that the development עךל from עךחל — omitting ח which was not pronounced — is frequent at Qumrân). But in Jeremiah α (chapters 1–29) the hiphil otherwise is rendered by κακοποιεω or πονηρευω, or is mistaken for the noun.

¹⁵ Syriac has (A) and (B), but in reverse order. For another example of the variants (A) and (B), compare 1 Kgs. 10.15 בךעח יבלמ לל// 2 Chr. 9.14 יכלמ לככוע.

¹⁶ חךמח ילככ םחלםנז does not make sense. κριοι suits the context, and is to be preferred. While κριοι translates םיךיךא in verses 34, 35, 36, it is the standard translation of ליא in the Septuagint. Probably read יליא; and possibly the variants arose by attraction of חךמח יליאכ to the phrase חךםח ילכ (5 times elsewhere in O.T.). The form םכיחזעזםחז is obscure, but probably reflects some form of the verb עםנ “to shatter” (cp. Syriac, and other instances of עפנ used to describe breaking of vessels). In our projected variants we follow Bright, 159.

¹⁷ So, e.g., Rudolph in BH³ and Bright.

¹⁸ With , Targum, and 46.16, 50.16; is by anticipation of the following phrase.

¹⁹ Following S. Talmon, Textus 1 (1960), 180. Cn. Rudolph, 146: “dl das überflüssige ‘Männer nach Ägypten’; lässt statt dessen 22b weg … aber eine solche Nachricht ist unerfindlich. wollte wohl nicht Wort haben, dass ein Jude einen solchen Auftrag ausführte … und ging dabei in den Bahnen von weiter, wo der Zusatz in 22a ermöglicht, 22b bei der Verlesung in der Synagogue wegzulassen.” If readings (A) and (B) were indeed in -Vorlage, one could explain the absence of the latter quite sufficiently by haplography, without recourse to the strained explanation given by Rudolph. It may be noted that such ad hoc explanation of the translator's Tendenz is characteristic of comments by many scholars on minuses in Jeremiah. For criticism of this kind of procedure, see Lemke, Werner E., The Synoptic Problem in the Chronicler's History, HTR 58 (1965), 361fGoogle Scholar. With respect to our present reading, we may ask what is offensive about the task of the “posse,” that anyone should want to suppress reference to its Jewish make-up.

²⁰ Synonymous, that is, in the context of this book, in which Nebuchadrezzar is represented as king of all the earth. On the synonymous variants חכלממ/חלשממ see further, S. Talmon, Scripta Hierosolymitana 8, 380, and Isa. 39.2 vs. 7QIsa^a.

²¹ See below, note 44.

²² The development of this conflate reading is mechanically similar to that in 25.6–7 (see above, reading no. 11). This development is obscured by the apparatus to BH³, which has omitting the second variant. Further, BH³ “dl c S” is misleading; Syriac does not support , but omits from ךלמח חזנב חא through to חפצמכ, and hence is defective by haplography, חפצמכ 1° to 2°, thereby indirectly attesting the conflate tradition of .

²³ (ο) ανηρ του πολεμου occurs in Ex. 15.3 (F^b), Nun, 31.28 (p^b-), Josh. 5.4 (M^mgv^mgz^mg), 5.6 (εβρ᾽ i^mg), 6.3 (O×), 1 Sam. 18.5 (^l καιγε), 2 Kgs. 25.4 ( καιγε). Significantly, in 1 Sam. 18.5 the old Greek text is absent. Since in this verse l is otherwise infected with καιγε readings (εν οφθαλμοις twice [on this as a καιγε replacement of earlier ενωπιον in Samuel-Kings, see the dissertation of J. D. Shenkel cited in note 1 above, pp. 24–33]; καιγε once), its agreement with καιγε here likewise seems to be due to late infection. On the other hand, in 2 Kgs. 25.4, where καιγε reads οι ανδρες του πολεμου, earlier reads ανδρες οι πολεμισται. The recensional character of the former term here, and generally, is clear, especially when taken with the evidence from Jeremiah (see immediately below).

²⁴ των ανθρωπων των πολεμουντων] τ. a. τ. πολεμιστων A–410 C'–613 Bo. The participle in Ziegler may reflect attraction to 32.24, 29, 37.10 (םימחלנח), either in (with A-etc. original?) or in -Vorlage (with A-etc. harmonizing to in 49.26, etc.)

²⁵ 2 Kgs. 25.25 omits לכ, which probably is secondary in Jeremiah.

²⁶ The two passages would have stood at roughly the same place in adjoining columns, and a scribe who did not understand the purpose of the gloss would have attached the phrase to םיךבנ. Compare 42.17 (םישנאח לב) + και παντες οι αλλογενεις = םיךח לכז) and 43.2 (םיךזח םישנאח לכ] omits םיךזח); the respective plusses are alternate absorptions of the same gloss םיךזח, which in -Vorlage was corrupted to, or mistaken for, םיךזח and prefixed with לכז. Other examples of two-way absorption of glosses can be adduced.

²⁷ If ανδρας εν πολεμω in fact does represent old Greek, the related development of the two readings is not affected; then becomes a witness to text tradition (B). But the existence in Ziegler of other recensional corrections (e.g., twice in 27.6, on which see Studies in the Text of Jeremiah, 113–16), and the existence of a number of inner-Greek doublets composed of old Greek and recensional Greek readings (see section C below), gives plausibility to our analysis.

²⁸ Both verbs are used elsewhere with this cliché: 11.22, 21.9, 27.13, 38.2, 42.17, 22; and 44.18, 27.

²⁸ The pejorative connotations of חילזלנ are clear from, e.g., Dt. 29.16, 2 Kgs. 23.24. חיצע may have been inspired by Isa. 46.1. For another example of conflation of the divine name and its pejorative substitute, see above, reading no. 5.

³⁰ An examination of parallel passages in Jer. 39.9–10 and 2 Kgs. 25.11–12, and the related passage in 2 Kgs. 24.14, in and versions, reveals frequent fluctuation between the synonymous variants םעח חזלך/עךאח חזלך.

³¹ Cf. Talmon, S., Textus 1 (1960), 165Google Scholar. For other examples of these synonymous variants, see Jud. 13.7, 1 Sam. 1.11, and the comments of Cross, F. M., A New Qumrân Biblical Fragment Related to the Original Hebrew Underlying the Septuagint, BASOR 132 (1953), 18Google Scholar (on 4QSam^a, v. 22).

³² For example, 30.10–11//46.27–28, 48.40b,41b///49.22 (italicized passages are absent from ; note that in each example the order of the contexts is reversed in ).

³³ Bright, cxxiii. In fairness, one should note Bright's admission (cxxiv) that he has not made a systematic investigation of the textual problems of the book of Jeremiah. In this statement he follows earlier studies, which could be cited to the same effect.

³¹ For example, 5.9//5.29//9.8; 6.22–24//50.41–43; 10.12–16//51.15–19; 11.20//20.12; 23.19–20//30.24–25; 21.9//38.2; 49.18//50.40; 49.19–21//50.44–46. For a full discussion of variants to large doublets, see Studies in the Text of Jeremiah, 225–34.

³⁵ Compare, for example, the orthography in 30.10//46.27, 10.13//51.16.

³⁶ Note especially the following details: (i) Outside of Jeremiah, זלכ occurs 15 times, חלכ 10 times; in Jer., זלכ occurs only in 6.13 bis, while חלכ (8.10 bis) occurs also in 2.21, 8.6, 15.10. 20.7, 48.31, 48.38. (ii) The hiphil םילכח is inappropriate to the context. What we have in 6.15 is niphal inf. cstr., with full orthography (cf. , Eb 10 םילכח, and 8.12 ; also 3.3). (iii) The phrase -/יחךקפ חע (כ) occurs 3 times in Jer. (6.15, 49.8, 50.31) ; twice reads the noun-/חךקפ. On the other hand, the phrase םחךקפ חנש/חע (כ) occurs 8 times in Jer. (8.12, 10.15//51.18, 11.23, 23.12, 31.44, 46.21, 50.27 [see also Isa. 10.3, Hos. 9.7, Mic. 74]). The latter form probably is original, while the former arose when the phrase was read as a defectively written םיחדקפ. Again, therefore, 6.13–15 represents an orthographically more advanced text than 2.10b–12.

³⁷ Aided, conceivably, by one (or both) of the following factors: (i) the passage would take up exactly 3 lines in a column of 50–52 letters; (ii) a manuscript with paragraph breaks after the prophetic clichés would be particularly vulnerable to haplography.

³⁸ In the absence of versification, the only way to find one's place in an ancient manuscript would be by context. The relative proximity of the two contexts in a manuscript several yards long and containing dozens of columns would aid in the mistaken identification of context.

³⁹ 6.13–15 presupposes a defectively-written Vorlage: συνετελεσαντο for עצזכ; την ατιμιαν αυτων for םילכח; εν τη πτωσει αυτων for םילפנכ; επισκοπης (αυτων) םיחךקפ. For a fine illustration of the orthographic contrast between two parts of a conflate reading, and orthographic differences generally between text and marginal readings in iQIsa^a, see S. Talmon, Textus 4 (1964), 109f.

⁴⁰ These readings may be found in Ziegler's work referred to above, page 3. It should be noted that the following observations are my own, following close study of the textual data, and that Ziegler is not to be held accountable for the conclusions which I draw from his analysis.

⁴¹ In addition, the following (with the Greek citation first, to facilitate reference to Ziegler's study): 8.16, 10.20, 17.11, 19.15 26(46).21, 30(49)4, 30.9(49.31), 31(48).13, 34.12(27.15), 39(32).17, 39(32).35, 49(42).16, 50(43).6.

⁴² In addition, the following: 2.2–3, 3.8, 9.16–17(15–16), 9.22(21), 11.16, 18.20,22, 21.12, 23.17, 23.29, 27(50),2, 29.12(49.11), 31(48).2, 37(30).6, 31(48).18, 50(43).9. These readings are discussed fully in Studies in the Text of Jeremiah, 50–55.

⁴³ The likelihood is strong in 18.20,22 and 23.29; possible instances are 3.8, 21.12, 50(43).9.

⁴⁴ See Studies in the Text of Jeremiah, chaps. III, IV, VI. The manner in which fills out proper names is illustrated above, in readings no. 14, 15, 19, 20. F. M. Cross generously has allowed me to examine photographs of the fragments of 4QJer^b, of which a partial text is published in ALQ ², 187 and n. 38. This fragmentary ms. agrees strikingly with in reading short names versus the long names , and refutes once and for all the notion that 's short-name readings are the result of a desire to condense the text in translation.

⁴⁵ Assuming that the prophets were translated after the Pentateuch, but significantly earlier than the time of the καιγε recension of the old Greek toward a proto-Massoretic Hebrew text type (this recension dates probably from the second half of the first century of the Christian era; cf. F. M. Cross, HTR 57 [1964], 282).

⁴⁶ See F. M. Cross, ALQ ², 187, and The Development of the Jewish Scripts, The Bible and the Ancient Near East, 137, Fig. 1, lines 3, 5, and commentary. Further, D. N. Freedman, The Massoretic Text and the Qumran Scrolls: A Study in Orthography, Textus 2 (1962), 87–102.