2

Transmission History

Shenkel regards the Greek text of Kings as “one of the primary sources for any adequate re-
construction of a chronology” (CRD 4). His understanding of the recensional development of the
Greek texts in parallel to that of the Hebrew text provides a new perspective that helps resolve
the chronological problems of 1–2 Kings. Review of the transmission history of the Hebrew and
Greek texts of the Books of 1–2 Kings may elicit information that will help reconstruct DK chro-
nology.

2.1. Early Compilation of the Books of Kings

Statements in the Books of Kings refer to earlier sources that could be consulted for addi-
tional information about the kings’ reigns: “The Book of the Acts of Solomon” (1 Kgs 11:41),
“The Book of the Chronicles of the Kings of Judah” (15:23), and “The Book of the Chronicles of
the Kings of Israel” (15:31). These chronicles — spanning several centuries and presumably re-
corded separately in Israel and Judah 1 — are mentioned in stylized regnal formulas at the end of
most kings’ narratives, implying that the narratives and/or the details of the regnal formulas were
derived from the same or similar sources. 2 Regnal formulas, the kings’ narratives, and other ma-
terial (e.g., prophetic narratives concerning Elijah and Elisha) were subsequently compiled into
one account to form what we now recognize as 1–2 Kings. But how and when the Books of Kings
attained their present literary form is still a matter of debate. 3

1. See S. R. Bin-Nun, “Formulas from Royal Records of Israel and of Judah,” VT 18 (1968): 414–32; critiqued by
R. D. Nelson, The Double Redaction of the Deuteronomistic History ( Journal for the Study of the Old Testament Supple-
ment 18; Sheffield: JSOT Press, 1981), 30–31; A. F. Campbell, Of Prophets and Kings: A Late Ninth-Century Document
(1 Samuel 1–2 Kings 10) (Catholic Biblical Quarterly Monograph Series 17; Washington, D.C.: Catholic Biblical Associ-
ation of America, 1986), 140 n. 1; and W. H. Barnes, Studies in the Chronology of the Divided Monarchy of Israel (Harvard
Semitic Monographs 48; Atlanta: Scholars Press, 1991), 138–42.
2. M. Haran views these chronicles as official records of an annalistic nature rather than “narrative and literary ma-
terial of the sort we find displayed in the canonical Books of Kings”; “The Books of the Chronicles ‘of the Kings of Ju-
dah’ and ‘of the Kings of Israel’: What Sort of Books Were They?” VT 49 (1999): 158.
3. The redaction history of the Books of Kings is an area of specialist study in itself. See M. Noth, The Deuterono-
mistic History ( Journal for the Study of the Old Testament Supplement 15; Sheffield: JSOT Press, 1981), translated from
Überlieferungsgeschichtliche Studien (2nd ed.; Tübingen: Max Niemeyer, 1957); F. M. Cross, Canaanite Myth and Hebrew
Epic: Essays in the History of the Religion of Israel (Cambridge: Harvard University Press, 1973), 274–89; Nelson, Double
Redaction of the Deuteronomistic History; G. H. Jones, 1 and 2 Kings (New Century Bible Commentary; Grand Rapids:
Eerdmans, 1984), 28–48; S. J. De Vries, 1 Kings (Word Biblical Commentary 12; Waco: Word, 1985), xxxviii–lxix; M. A.
O’Brien, The Deuteronomistic History Hypothesis: A Reassessment (Orbis Biblicus et Orientalis 92; Göttingen: Vandenhoeck

14
 Transmission History 15

Translating and copying this final compilation over subsequent centuries gave opportunity
not only for editorial activity, but also increased the chance of corruptions or variants entering
the text. Emanuel Tov points out that our earliest extant texts are the products of a transmission
process during which the original text suffered textual alterations:
Most of the texts — ancient and modern — which have been transmitted from one generation to the
next have been corrupted in one way or another. For modern compositions the process of textual
transmission from the writing of the autograph until its final printing is relatively short, so that the
possibilities of its corruption are limited. In ancient texts, however, such as the Hebrew Bible,
these corruptions (the technical term for various forms of “mistakes”) are found more frequently
because of the difficult physical conditions of the copying and the length of the process of trans-
mission. 4
Julio Trebolle Barrera concurs:
It is absolutely indispensable to know the history of transmission of the biblical text throughout
more than two thousand years for further work in criticism and restoration of the “original” bibli-
cal text. It is necessary to know the wealth of variants which the manuscripts in Hebrew, Greek
and other versions have transmitted to us, including rabbinic and patristic quotations. It is also
necessary to determine which of these variants are guaranteed to go back to the oldest stages of
textual transmission and which are the work of later revisions or the result of some type of textual
corruption. 5
Since the study of chronology deals with variant numbers in Greek and Hebrew texts, I will
thus follow Shenkel’s and Trebolle Barrera’s recommendations to elucidate the transmission his-
tory of the biblical text, in this case the Books of Kings, before seeking a methodology to recon-
struct the chronology.

2.2. Transmission History of the Old Testament Text

2.2.1. Hebrew Text and Qumran Manuscripts

The earliest known Old Testament Hebrew manuscripts — dated between the 3rd century
b.c.e. and 70 c.e. 6 — were found in the Judean Desert between 1947 and 1956. The MT was not
found at Qumran, even though its presumed forerunner, a proto-MT, is present. 7 After 70 c.e.,

& Ruprecht, 1989); B. Halpern and D. S. Vanderhooft, “The Editions of Kings in the 7th–6th Centuries b.c.e.,” Hebrew
Union College Annual 62 (1991): 179–244; S. L. McKenzie, The Trouble with Kings: The Composition of the Book of Kings in
the Deuteronomistic History (Vetus Testamentum Supplement 42; Leiden: Brill, 1991); idem, “The Books of Kings in the
Deuteronomistic History,” in The History of Israel’s Traditions: The Heritage of Martin Noth (ed. S. L. McKenzie and M. P.
Graham; Journal for the Study of the Old Testament Supplement 182; Sheffield: Sheffield Academic Press, 1994), 281–
307; A. G. Auld, Kings without Privilege: David and Moses in the Story of the Bible’s Kings (Edinburgh: Clark, 1994); A. Le-
maire, “Toward a Redactional History of the Books of Kings,” in Reconsidering Israel and Judah: Recent Studies on the Deu-
teronomistic History (ed. G. N. Knoppers and J. G. McConville; Sources for Biblical and Theological Study 8; Winona
Lake, Ind.: Eisenbrauns, 2000), 446–61.
4. E. Tov, Textual Criticism of the Hebrew Bible (2nd ed.; Minneapolis: Fortress, 2001), 8.
5. J. Trebolle Barrera, The Jewish Bible and the Christian Bible: An Introduction to the History of the Bible (trans. W. G. E.
Watson; Leiden: Brill/Grand Rapids: Eerdmans, 1998), 267.
6. F. M. Cross, “A New Qumran Biblical Fragment Related to the Original Hebrew Underlying the Septuagint,”
BASOR 132 (1953): 16–17; M. Greenberg, “The Stabilization of the Text of the Hebrew Bible, Reviewed in the Light of
the Biblical Materials from the Judean Desert,” Journal of the American Oriental Society 76 (1965): 163; S. H. Horn, “The
Old Testament Text in Antiquity,” Ministry (Nov. 1987): 5.
7. F. M. Cross, “The Contribution of the Qumrân Discoveries to the Study of the Biblical Text,” Israel Exploration
Journal 16 (1966): 95 = QHBT 292; idem, “The Evolution of a Theory of Local Texts,” in QHBT 314.
16 Chapter 2

Hebrew texts of a provenance other than the proto-MT began to diminish in number, whether
by accident or design, and their exemplars were lost to history — apart from those deposited at
Qumran or kept by an isolated community such as the Samaritans 8 — leading to the widespread
adoption of the MT by about the 8th century. 9
Frank Cross identifies three Hebrew text-types in the Pentateuch and Former Prophets
( Joshua, Judges, Samuel, Kings), which he traces to three stages in the recensional development
of the LXX, and suggests a local-text theory of provenance, elaborating an earlier suggestion by
W. F. Albright. 10 Even though Cross’s hypothesis has not been accepted by all scholars (some pro-
pose even greater textual diversity), 11 his theory (table 2.1) is useful as a basis for discussion. Ac-
cording to Cross, the Qumran Hebrew texts developed from a 5th-century b.c.e. Palestinian
archetype, and from this text developed at least three different branches or text-types, identified
according to their assumed provenance: Egypt, Palestine, and Babylon. 12

2.2.1.1. Egyptian Text

Cross finds “evidence that the Septuagint of Samuel and Kings was translated from an Egyp-
tian Hebrew text that separated from the Old Palestinian textual tradition no later than the
fourth century b.c. This text differed sharply from the textus receptus, and while more closely
allied to Palestinian texts from Qumrân, nevertheless is distinct from them.” 13 Elsewhere he com-
ments: “The Egyptian and Palestinian families are closely related. Early exemplars of the Pal-
estinian text in the Former Prophets, and Pentateuchal texts which reflect an early stage of the
Palestinian tradition, so nearly merge with the Egyptian, that we are warranted in describing the
Egyptian text-type as a branch of the Old Palestinian family.” 14
Objecting to Cross’s theory that this Egyptian text was utilized by the LXX translators, Tov
points out that “the reconstructed Hebrew Vorlage of Ì does not reflect any proven Egyptian
characteristics; rather, it is more likely that Ì was translated from Palestinian texts, as claimed by
the Epistle of Aristeas.” 15 Whatever its origin, the first Greek translation is not extant but is be-

8. Cross, “Contribution of the Qumrân Discoveries,” 95 = QHBT 292; cf. S. Talmon, “The Old Testament Text,” in
The Cambridge History of the Bible, vol. 1: From the Beginnings to Jerome (ed. P. R. Ackroyd and C. F. Evans; Cambridge:
Cambridge University Press, 1970), 170–71, 187 = QHBT 12–13, 29; Tov, Textual Criticism of the Hebrew Bible, 34–35,
194–95, 316.
9. Tov, Textual Criticism of the Hebrew Bible, 33, 35.
10. F. M. Cross, “The History of the Biblical Text in the Light of Discoveries in the Judaean Desert,” Harvard Theo-
logical Review 57 (1964): 297–99 = QHBT 193–95; idem, “Contribution of the Qumrân Discoveries,” 84–91 = QHBT
281–88; “Evolution of a Theory,” 306–20; W. F. Albright, “New Light on Early Recensions of the Hebrew Bible,” BASOR
140 (1955): 30–33 = QHBT 143–46.
11. Talmon, “Old Testament Text,” 193–99 = QHBT 35–41; idem, “The Textual Study of the Bible — A New Out-
look,” in QHBT 324–27; E. Tov, “A Modern Textual Outlook Based on the Qumran Scrolls,” Hebrew Union College An-
nual 53 (1982): 11–27; idem, The Text-Critical Use of the Septuagint in Biblical Research (2nd ed.; Jerusalem Biblical Studies
8; Jerusalem: Simor, 1997), 183–87; idem, Textual Criticism of the Hebrew Bible, 160–63, 185–97; see also Trebolle Bar-
rera, Jewish Bible and the Christian Bible, 292–95.
12. Cross, “History of the Biblical Text,” 296–97 = QHBT 192–93.
13. Ibid., 295; cf. 297 = QHBT 191, 193.
14. Cross, “Contribution of the Qumrân Discoveries,” 86 = QHBT 283.
15. Tov, Textual Criticism of the Hebrew Bible, 186, 136–37; see also idem, Text-Critical Use of the Septuagint, 183–87;
idem, “The Septuagint,” in Mikra: Text, Translation, Reading, and Interpretation of the Hebrew Bible in Ancient Judaism and
Early Christianity (ed. M. J. Mulder; Compendia Rerum Iudaicarum ad Novum Testamentum 2/1; Philadelphia: Fortress,
1988), 164–65.
 Transmission History 17

Table 2.1. Textual History of the Old Testament according to F. M. Cross

Early Textual Sources

Palestinian

Babylonian
Egyptian

Old Greek LXX proto-Masoretic

proto-Lucianic LXX

kaige LXX

Josephus
Masoretic Text

lieved to be represented in the OG sections of Codex Vaticanus, the latter dating to the 4th cen-
tury c.e. 16

2.2.1.2. Palestinian Text

Cross identifies the dominant Hebrew text found at Qumran as having its provenance in
Palestine:
By far the majority of the Hebrew witnesses from Qumrân belong to the Palestinian family. The
evidence for the identification of the Palestinian family is most easily delineated in Samuel. The
three manuscripts of Samuel from Cave IV, while not directly filiated, contain a single textual tradi-
tion, known at Qumrân as early as the third century (4QSamb), the early first century b.c.
(4QSamc), and in the late first century b.c. (4QSama). The earliest distinctive witness to a text of
the type of these manuscripts is found in the Chronicler. As has been shown elsewhere, the Chron-
icler, shortly after 400 b.c., cited a text of Samuel which stands in close agreement with the manu-
scripts of Qumrân, but, as is well known, sharply diverges from the received text. In the second or
early first century b.c., the same Palestinian text-form was used to revise the Septuagint: the Proto-
Lucianic recension. Finally Josephus at the end of the first century a.d., made use of this Palestin-
ian Greek recension in writing his Antiquities. 17
Elsewhere, Cross writes:
In studying the text of 4QSama, I have been forced to note a series of readings in which the
Hebrew of 4QSama reflects the so-called Lucianic recension preserved in the Greek minuscules
boc2e2, and the Itala [i.e., Old Latin]. In other words, 4QSama stands with LXXL against MT and

16. For more on OG origins, see E. Ulrich, The Dead Sea Scrolls and the Origins of the Bible (Studies in the Dead Sea
Scrolls and Related Literature; Leiden: Brill / Grand Rapids: Eerdmans, 1999), 205–14.
17. Cross, “Contribution of the Qumrân Discoveries,” 88 = QHBT 285.
 18 Chapter 2

LXXB. These are proper Proto-Lucianic readings in a Hebrew text of the first century b.c., four cen-
turies before the Syrian Father to whom the recension is attributed. 18
The significance of these comments for the reconstruction of chronology should not be over-
looked. Since the L minuscules (except c2) have the same system of chronological data in 1 Kings
as the OG sections of Codex Vaticanus and since the L manuscripts of Samuel go back to a He-
brew Vorlage represented by the 1st-century b.c.e. 4QSama, we have grounds for expecting that
the L text of Kings also derives from the same Vorlage, a conclusion with which Shenkel concurs:
“Although a proto-Lucian text type that is extensively preserved has not been found among the
scanty remains of Hebrew manuscripts for the Books of Kings at Qumran, it is reasonable to sup-
pose that the history of the development of the text for Kings is substantially the same as for the
Books of Samuel, considering the intimate relationship of these four books to each other in the
Greek text” (CRD 10).

2.2.1.3. Babylonian Text

Cross theorizes that a third Hebrew text-type, “distant from both the Egyptian and Palestinian
families,” 19 was taken to Babylonia and isolated there during the exile, giving rise to a Babylonian
text-type. Reintroduced to Palestine in the Maccabean period, this text-type became dominant by
the end of the 1st century c.e. 20 and is now assumed to be the precursor of the MT (i.e., proto-
MT), which reached its final form in medieval times. 21
The proto-MT is reflected in the kaige recension, so-called for its characteristic translating of
Hebrew µg (‘also’) with Greek kaÇge. 22 The KR was first identified in Dominique Barthélemy’s 1963
publication of the Greek Minor Prophets scroll (8ÓevXIIgr) found at Nahal Óever in 1952. 23 This
KR manuscript is dated to the mid-1st century b.c.e. 24 and is presumed to have brought an earlier
Greek text into conformity with the proto-MT. 25 Once widespread, 26 the kaige text-type is extant
in sections of Samuel and Kings in Codex Vaticanus, Greek manuscripts of Lamentations and
Judges, the Theodotionic edition of Exodus, and the Theodotionic additions to the Greek text of
Job and elsewhere. 27 The KR was formerly attributed to Theodotion, who lived near the end of
the 2nd century c.e., but because the so-called Theodotionic text antedates the historical Theo-
dotion by two centuries, the text is also known as kaige-Theodotion or proto-Theodotion. 28
Only small fragments — none containing chronological data — of the Books of Kings have
been found at Qumran. Some are of the proto-MT text-type, while others appear to reflect the

18. Cross, “History of the Biblical Text,” 292 = QHBT 188 (emphasis original).
19. Cross, “Contribution of the Qumrân Discoveries,” 90 = QHBT 287.
20. Ibid., 91, 95 = QHBT 288, 292.
21. Tov, Textual Criticism of the Hebrew Bible, 27.
22. Ibid., 145.
23. D. Barthélemy, Les devanciers d’Aquila: Première publication intégrale du texte des fragments du Dodécaprophéton
(Vetus Testamentum Supplement 10; Leiden: Brill, 1963).
24. Tov, Textual Criticism of the Hebrew Bible, 145.
25. For discussion see D. Barthélemy, “Redécouverte d’un chaînon manquant de l’histoire de la Septante,” Revue
Biblique 60 (1953): 18–29 = QHBT 127–39; repr. in Études d’histoire du texte de l’Ancien Testament (ed. D. Barthélemy; Or-
bis Biblicus et Orientalis 21; Göttingen: Vandenhoeck & Ruprecht, 1978), 38–49; P. Katz, “Justin’s Old Testament Quo-
tations and the Greek Dodekapropheton Scroll,” Studia Patristica 1 (1957): 343–53, repr. in Studies in the Septuagint:
Origins, Recensions, and Interpretations (ed. S. Jellicoe; New York: Ktav, 1974), 530–40.
26. Tov, Textual Criticism of the Hebrew Bible, 144.
27. K. G. O’Connell, “Greek Versions (Minor),” in Interpreter’s Dictionary of the Bible: Supplementary Volume (ed.
K. Crim; Nashville: Abingdon, 1976), 379; see also Trebolle Barrera, Jewish Bible and the Christian Bible, 316.
28. Tov, Textual Criticism of the Hebrew Bible, 145; Trebolle Barrera, Jewish Bible and the Christian Bible, 314–15.

spread is 6 points long

 Transmission History 19

Vorlage now identified with the proto-L substratum of L. 29 Being based on Hebrew texts that no
longer exist, the early Greek manuscripts are vital witnesses to the Hebrew text from which they
were translated and, hence, equally vital witnesses to the chronological data contained in the
early Hebrew Books of Kings.

2.2.2. Old Greek and Kaige

In Samuel–Kings, Codex Vaticanus exhibits the two distinct text-types referred to above: OG
and KR. The identification of the OG and KR sections of 1–2 Kings according to text-type is a sig-
nificant first step in seeking to establish to which recension any given chronological datum be-
longs. The KR’s characteristic translation µg ª kaÇge was the initial criterion used to differentiate
this text-type from the OG. But according to Shenkel (CRD 12), more significant characteristics
are the KR’s greater conformity to the MT away from the OG text-type and “an extreme concern
that at least certain important Hebrew expressions receive a uniform rendition into Greek even
at the expense of sound Greek usage and idiom.” 30 For example, the OG often translates Hebrew
verbs with the historical present, whereas the KR uses the aorist. 31 Thus, when Hebrew perfect
˚lm means “he began to reign” in the regnal formulas, the OG translates it with historical present
basileuvei, while the KR translates it with aorist ejbasÇleusen (CRD 50–53). The KR also follows the
MT in its sequencing of regnal formulas (see §5.4).

2.2.2.1. Textual Divisions in Codex Vaticanus

Thackeray identifies five sections in Samuel–Kings in Codex Vaticanus according to their tex-
tual characteristics and assigns Greek sigla to them (table 2.2). 32 Shenkel (CRD 117–20) revised
Thackeray’s bb and bg sections.

Table 2.2. Text-Types in Samuel–Kings in Codex Vaticanus

Text-Type Siglum Thackeray’s Divisions Shenkel’s Divisions

OG a 1 Samuel (same)
OG bb 2 Sam 1:1–11:1 2 Sam 1:1–9:13
KR bg 2 Sam 11:2–1 Kgs 2:11 2 Sam 10:1–1 Kgs 2:11
OG gg 1 Kgs 2:12–21:43 (same)
KR gd 1 Kgs 22:1–2 Kgs 25:30 (same)

29. Cf. J. Gray, 1 and 2 Kings (Old Testament Library; London: SCM, 1970), 46–47; Shenkel, CRD 122–23 n. 14;
De Vries, 1 Kings, liv.
30. Other characteristics that differentiate the OG and KR have been identified; see Barthélemy, Les devanciers
d’Aquila, 48–80; M. Smith, “Another Criterion for the kaÇge Recension,” Biblica 48 (1967): 443–45; Shenkel, CRD 12–18,
113–16; J. A. Grindel, “Another Characteristic of the Kaige Recension: jxn/nikoÍ,” Catholic Biblical Quarterly 31 (1969):
499–513; E. Tov, “Transliterations of Hebrew Words in the Greek Versions of the Old Testament: A Further Character-
istic of the kaige-Th. Revision?” Textus 8 (1973): 78–92; S. Jellicoe, “Some Reflections on the Kaige Recension,” VT 23
(1973): 15–24; O’Connell, “Greek Versions (Minor),” 378–79; Trebolle Barrera, Jewish Bible and the Christian Bible, 315–16.
31. H. St. J. Thackeray, The Septuagint and Jewish Worship: A Study in Origins (Schweich Lectures, 1920; London: Ox-
ford University Press for the British Academy, 1921; repr. Munich: Kraus-Thomson, 1980), 20–22; cf. Barthélemy, Les
devanciers d’Aquila, 63–65.
32. H. St. J. Thackeray, “The Greek Translators of the Four Books of Kings,” Journal of Theological Studies 8 (1907):
262–78; idem, Septuagint and Jewish Worship, 16–28.
20 Chapter 2

Thackeray sees the end of the first KR section at 1 Kgs 2:11 with the death of David as evi-
dence of a division between Samuel and Kings earlier than what now appears in the MT. 33 If the
change in text-type at 1 Kgs 22:1 to the second KR section indicates where one scroll ended and
another began, then 1 Kings 22 was part of the same scroll as 2 Kings, making it, in effect, the
first chapter of 2 Kings.
In addition, 1 Kings 20 MT and 1 Kings 21 MT are transposed in the Greek, with 1 Kings 21
coming before 1 Kings 20 (i.e., the last verse of the OG text-type occurs at 20:43 MT = 21:43
Greek). The Greek account of war between Syria and Israel in 1 Kings 21 is then followed by
22:1: “For 3 years Syria and Israel continued without war.” The MT placement of Ahab’s murder
of Naboth in 1 Kings 21 between Ahab’s battle with Syria (1 Kings 20) and another battle with
Syria (1 Kings 22) exhibits no textual continuity between the last verse of 1 Kings 21 (“I will bring
the evil upon his house”) and 1 Kgs 22:1. The order of the Greek text is more logical, and the di-
vision of the Greek text shown by the change in text-type appears to support an earlier division
than is now exhibited by the MT.
While the Hebrew Books of Samuel filled only one scroll, the Greek version of 1–2 Samuel
took twice as much space, because the Greek script includes vowels. 34 This raises the question,
however, why 2 Samuel exhibits two text-types (both OG and KR) while 1 Samuel, 1 Kings, and
2 Kings each exhibit only a single text-type, roughly corresponding to scroll divisions. The two
text-types in 2 Samuel may indicate where one scroll ended and another began. Alternatively, the
change may indicate where one copyist stopped and another started, the first copying from an
OG scroll, and the second transcribing from a KR scroll.
A similar circumstance could be theorized for the change from OG in section gg (roughly
1 Kings) to KR in section gd (roughly 2 Kings). Scrolls of different origins and exhibiting different
text-types were subsequently copied and combined to create the great codexes, including Codex
Vaticanus, all the while preserving the recensional characteristics of each section. 35 On the other
hand, the mixing of two text-types or scrolls may have come about through purely mechanical
causes. 36 Not knowing why two recensions were utilized in the compilation of Samuel–Kings
leaves unanswered the question whether sections bg and gd are the only KR sections ever produced
or whether KR versions once existed for sections a, bb, and gg — that is, for all of Samuel–Kings.

2.2.2.2. Chronological Data in Codex Vaticanus

Whatever the reason for the two recensions in Samuel–Kings, Codex Vaticanus’s exhibiting
two different text-types is important for understanding the chronology of the Books of Kings. In
fact, the most important feature with regard to chronology in the Books of Kings in Codex Vati-

33. Thackeray, “Greek Translators,” 262–65.

34. This doubled length likely caused the present division of these books into 1 Samuel and 2 Samuel; see Trebolle
Barrera, Jewish Bible and the Christian Bible, 97; and Thackeray, “Greek Translators,” 265.
35. Trebolle Barrera, Jewish Bible and the Christian Bible, 318.
36. E. Tov, “The State of the Question: Problems and Proposed Solutions,” in 1972 Proceedings IOSCS and Pseudepig-
rapha (ed. R. A. Kraft; Society of Biblical Literature Septuagint and Cognate Studies 2; Missoula, Mont.: Scholars Press,
1972), 5. T. Muraoka proposes that two translators worked side by side on the scrolls and that at a later time bg and gd
were revised by the KR redactor; “The Greek Texts of Samuel–Kings: Incomplete Translations or Recensional Activity?”
in 1972 Proceedings IOSCS and Pseudepigrapha (ed. R. A. Kraft; Society of Biblical Literature Septuagint and Cognate
Studies 2; Missoula, Mont.: Scholars Press, 1972), 90–107, revised in Abr-Nahrain 21 (1982–83): 28–49. Barthélemy
regards this as improbable; “Prise de position sur les autres communications du colloque de Los Angeles,” in Études
d’histoire du Texte de l’Ancien Testament (ed. D. Barthélemy; Orbis Biblicus et Orientalis 21; Göttingen: Vandenhoeck &
Ruprecht, 1978), 257–58.
 Transmission History 21

canus is that 1 Kgs 15:1–21:43 OG contains chronological data that for the most part are at wide
variance with the MT data but concur with the L data in boe 2 (but not c2). On the other hand, the
KR data in section gd concurs in the majority of instances with the MT data. The OG text is not
extant for section gd but is presumed, consistent with section gg, to be represented by the data of
boe2 where these manuscripts differ from the MT/KR data. 37 And the KR, not being extant in
section gg, is considered to be represented in this section by the MT data, both being derived
from a Hebrew proto-MT precursor.
The conflicting chronological data found in the two different recensions of the Books of
Kings in Codex Vaticanus may be illustrated by the regnal formula at 1 Kgs 22:41–42. KR and
MT both have, “Jehoshaphat son of Asa began to reign over Judah in the 4th year of King Ahab
of Israel, . . . and he reigned 25 years.” This verse is not extant in OG or L at this position (L does
not have 1 Kgs 22:41–51), but appears instead at 16:28a: “In the 11th year of Omri, Jehoshaphat
son of Asa began to reign, . . . and he reigned 25 years.” These statements are in conflict, and
without the expediency of some unattested dating system to explain the difference in accession
years, Jehoshaphat’s reign cannot have commenced in both Ahab’s 4th year and Omri’s 11th year.
Jehoshaphat’s regnal formula is the only such formula to appear in both the OG and KR text-
types in Codex Vaticanus. If the L chronological data in boe2 is representative of the no longer
extant OG, then a second example of conflicting regnal data appears in the two text-types at
22:52: KR/MT have, “Ahaziah-I son of Ahab reigned over Israel in Samaria in the 17th year of Je-
hoshaphat king of Judah and he reigned 2 years”; and L has, “In the 24th year of Jehoshaphat
king of Judah, Ahaziah-I son of Ahab began to reign and he reigned 2 years.”
A more complex example is the accession synchronism of Joram (2 Kgs 1:17 MT = 1:18a
L/KR). The MT and L synchronisms state that Joram became king in Jehoram’s 2nd year. The KR
says that Joram began to reign in Jehoshaphat’s 18th year and repeats this synchronism at 3:1,
where it appears also in MT. The MT, therefore, has two synchronisms for the accession of Joram:
Jehoram’s 2nd year and Jehoshaphat’s 18th year.
These three examples — Jehoshaphat, Ahaziah-I, and Joram — are sufficient to show that con-
flicting data are represented in the OG and KR sections of Codex Vaticanus. Further analysis of
the data is required to determine why the OG and KR exhibit variant data, why the respective
data in the texts are not internally consistent, and whether any numbers have been altered acci-
dentally or intentionally.

2.2.3. Lucianic Manuscripts

2.2.3.1. Proto-Lucian and Lucian
L is named after its putative author, Lucian of Antioch (ca. 240–311/312 c.e.). 38 Lucian, how-
ever, cannot have been the originator of this recension, because it appears as early as the end of

37. This applies, however, only to the end of the EDK, which ends with the deaths of Ahaziah-J and Joram at the
hands of Jehu (2 Kgs 9:24–28). In the LDK, which commences with the accessions of Athaliah in Judah and Jehu in Is-
rael and ends with the fall of Samaria, there is no difference between the boe2 data and the MT/KR data except for the
regnal years for Pekahiah (MT/KR has 2 years; boe2 has 10 years). It is, therefore, in the EDK that the greater number
of variants occur between OG/L and MT/KR in Kings.
38. For discussion of the L manuscripts, see the anonymous article “Lucian’s Recension of the Septuagint,” Church
Quarterly Review 51 (1900–1901): 379–98; B. M. Metzger, “Lucian and the Lucianic Recension of the Greek Bible,” New
Testament Studies 8 (1961–62), 189–203 = “The Lucianic Recension of the Greek Bible,” in Chapters in the History of New
Testament Textual Criticism (Grand Rapids: Eerdmans, 1963), 1–41; S. P. Brock, “Lucian Redivivus: Some Reflections on
Barthélemy’s Les devanciers d’Aquila,” Studia Evangelica 5 (1968): 176–81; S. Jellicoe, The Septuagint and Modern Study
22 Chapter 2

the 1st century c.e. in the writings of Josephus, who used proto-L to construct his Samuel narra-
tive. 39 Moreover, proto-L Samuel appears to derive from a text like the Palestinian Hebrew text
(4QSama) found at Qumran dating to the 1st century b.c.e. Proto-L is also found in the sixth col-
umn of Origen’s Hexapla at 2 Sam 11:2–1 Kgs 2:11 (Thackeray’s section bg), 40 in the Old Latin
for the Books of Samuel dating from the 2nd century c.e., 41 in some readings in the Old Testa-
ment Peshitta in 1 Samuel, and in other manuscripts of the Old and New Testaments and various
church fathers. 42
Many scholars believe that the substratum used by Lucian was itself a revision of the OG (i.e.,
the first Greek translation) or of what is now known as proto-L. 43 Other scholars claim that proto-
L is either the OG 44 or any single OG translation. 45 Depending on the view held, L is seen as a
composite text having two or three layers — a composite text that Trebolle Barrera describes as
follows:
It is not easy to determine precisely what was the work Lucian carried out, both in respect of the
pre-Hexaplar tradition and in respect of Origen’s work. However, the characteristics of the Lucianic
text are quite obvious: frequent additions, inserted into the text to adapt it to the rabbinic Hebrew
text; many duplicate readings in which the old Septuagint is juxtaposed to the Hexaplar reading,
which in turn is closer to the rabbinic text; grammatical corrections and stylistic improvement of
the text; the insertion of explanatory elements, such as proper names, pronouns, articles, etc.;
replacement of Hellenistic forms by the Attic equivalents, etc. 46

2.2.3.2. Lucianic Variants

The L variants are published in the second apparatus of the Larger Cambridge Septuagint
beneath the text of Codex Vaticanus. 47 Manuscripts boc2e2, dating from the 10th to 14th

(Oxford: Clarendon, 1968), 157–290; Shenkel, CRD 8–11; E. Tov, “Lucian and Proto-Lucian: Toward a New Solution of
the Problem,” Revue Biblique 79 (1972): 101–13 = QHBT 293–305; N. Fernández Marcos, “The Lucianic Text in the
Books of Kingdoms: From Lagarde to the Textual Pluralism,” in De Septuaginta: Studies in Honour of John William Wevers
on His Sixty-Fifth Birthday (ed. A. Pietersma and C. Cox; Mississauga, Ont.: Benben, 1984), 161–74; Trebolle Barrera, Jew-
ish Bible and the Christian Bible, 310–11.
39. H. St. J. Thackeray, Josephus the Man and the Historian (New York: Jewish Institute of Religion Press, 1929), 82–
89; Metzger, “Lucian and the Lucianic Recension,” 199 = Chapters in the History, 34; E. C. Ulrich, The Qumran Text of
Samuel and Josephus (Harvard Semitic Monographs 19; Missoula, Mont.: Scholars Press, 1978), 22, 27–28, 257–59.
40. As Cross notes (“History of the Biblical Text,” 295 = QHBT 191), the sixth column of the Hexapla was normally
that of Theodotion, but Adam Mez recognizes that in this section it was proto-L.
41. Trebolle Barrera, Jewish Bible and the Christian Bible, 352–53.
42. Metzger, “Lucian and the Lucianic Recension,” 196–200 = Chapters in the History, 31–35; Tov, “Lucian and
Proto-Lucian,” 103–5 = QHBT 295–97.
43. Cross, “History of the Biblical Text,” 295–96 = QHBT 191–92; idem, “Evolution of a Theory,” 312–15; Ulrich,
Qumran Text of Samuel and Josephus, 257–59; Fernández Marcos, “Lucianic Text in the Books of Kingdoms,” 169–73.
C. D. Stanley appears to support the position of a proto-L text: “The Significance of Romans 11:3–4 for the Text His-
tory of the LXX Book of Kingdoms,” JBL 112 (1993): 43–54, esp. 54.
44. Barthélemy, Les devanciers d’Aquila, 126–27; cf. idem, “Prise de position,” 271–73; D. G. Deboys, The Greek Text
of 2 Kings (M.Litt. thesis; Oxford University, 1981), 187–89; A. Pietersma, “Proto-Lucian and the Greek Psalter,” VT 28
(1978): 67.
45. Tov suggests that “the substratum of boc2e2 contains either the Old Greek or any single Old Greek translation”;
“Lucian and Proto-Lucian,” 110 = QHBT 302 (emphasis original). See Trebolle Barrera, Jewish Bible and the Christian
Bible, 310–11.
46. Trebolle Barrera, Jewish Bible and the Christian Bible, 310 (emphasis original); see Metzger, “Lucian and the Lu-
cianic Recension,” 194–95 = Chapters in the History, 25–27.
47. Manuscript r also shows proto-L readings in parts of 2 Kings; see CRD 127 n. 39.
 Transmission History 23

Table 2.3. Lucianic Manuscripts

Larger Cambridge Göttingen

Septuagint Siglum Number Date
19 (bu) 11th–12th century
b
{ 108 (b) 13th–14th century
o 82 12th–13th century
c2 127 10th century
e2 93 13th–14th century

centuries, 48 are denoted in the Göttingen Septuagint by numbers (table 2.3). 49

The agreement of the L chronological data in manuscripts boe 2 with the OG data in section
gg of 1 Kings attests the early pre-Hexaplaric nature of the substratum containing the regnal for-
mulas. The main difference between the chronological data exhibited by L and OG in Vaticanus
is at 1 Kgs 16:15. Manuscripts be2 have a synchronism giving Zimri’s accession in Asa’s 22nd year
and giving to Zimri a reign of 7 days. Codex Vaticanus does not have the synchronism and at-
tributes to Zimri a 7-year reign. The MT has an alternative synchronism giving Zimri’s accession
in Asa’s 27th year, but agreeing with L in giving Zimri a reign of 7 days. Bruce Metzger com-
ments, “Despite the numerous secondary features which Lucian introduced into his recension of
the Old Testament, one may expect to find here and there in it certain readings, not extant in the
other forms of the Septuagint, which will be useful in ascertaining the most ancient form of the
Hebrew text.” 50 The unique variant in be2 giving Zimri’s accession in Asa’s 22nd year may thus
preserve an important witness to a very ancient text not found elsewhere.
On the other hand, specific readings — especially when the Hexaplaric signs are present —
show signs of having been conformed to the Greek text in Origen’s Hexapla, which was itself
conformed to the MT. In the absence of Hexaplaric signs, Shenkel comments, “the coincidence
of a reading in the Lucianic text with the reading of the normal hexaplaric witnesses, when such
a reading is not found in the Old Greek or the KR, justifies the presumption that the reading in
question is an assimilation from the hexaplaric text introduced into the proto-Lucianic text by a
later redactor” (CRD 19).
For example, at 2 Kgs 1:17 the Larger Cambridge Septuagint provides in its apparatus addi-
tional text appearing in boc2(*)e2: kaµ ejbasÇleusen ∆Iwra;m oJ ajdelfo;Í ∆OcozÇou ajnt∆ aujtou` o§ti oujk h®n
aujtå` u¥ovÍ (and Joram the brother of Ahaziah-I reigned after him because he did not have a son). 51
This verse in boc2e2 does not contain a synchronism, but Hexaplaric minuscules x and y include
the synchronism that Joram began to reign in Jehoram’s 2nd year. Since the same verse appears
in 1:17 MT but not in KR, the synchronism was, according to Shenkel, “inserted into the MT
after the KR revision was completed” (CRD 74). The absence of the synchronism at 1:17 KR and

48. Ulrich, Qumran Text of Samuel and Josephus, 21.

49. See Jellicoe, Septuagint and Modern Study, 163–64 n. 5, 167; Ulrich, Qumran Text of Samuel and Josephus, 19–21;
P. K. McCarter, I Samuel (Anchor Bible 8; New York: Doubleday, 1980), 9 n. 19.
50. Metzger, “Lucian and the Lucianic Recension,” 200 = Chapters in the History, 35.
51. Shenkel notes (CRD 73–74): “The asterisk after the Lucianic manuscript c2 shows that this passage was not in
the proto-Lucianic stage of the text, but was inserted at a later date to bring the Lucianic text into partial conformity
with the MT and the hexaplaric recension, here represented by the text of the two minuscules, x and y.”
24 Chapter 2

the asterisk after c2 showing a Hexaplaric addition at 1:17 indicates that the text and synchronism
were not in the Vorlage of 1:17 L/KR.
If the position of the datum at 1:17 MT giving Joram’s accession in Jehoram’s 2nd year is a
Hexaplaric addition, what then shall we make of the MT’s second synchronism at 3:1 for Joram’s
accession in Jehoshaphat’s 18th year (a synchronism also present in 1:18a KR and 3:1 KR)? Is this
an earlier synchronism in an earlier position than that at 1:17? If so, why do L and KR each have
Joram’s synchronism at 1:18a but with conflicting data? Bearing this in mind we consider the next
example of L variation.
L agrees with MT and KR in synchronizing Jehoram’s accession in Joram’s 5th year (8:16). But
this is in conflict with its previous synchronism at 1:18a, in which Joram’s accession is synchro-
nized with Jehoram’s 2nd year, inferring that Jehoram became king in Ahaziah-I’s 2nd year (syn-
chronism not extant). Why does L have a conflict in its data? It could be that the L data in 1:18a
and 8:16 belong to different strata of L, one datum from the proto-L stratum and one from the
later Hexaplaric stratum in which proto-L has been conformed to the MT. Recognition of which
L stratum each datum belongs to may help resolve the enigma of conflicting variants in L.

2.2.4. Aquila, Symmachus, and Theodotion

Three recensions based on the KR originated in the 2nd century: Aquila, Symmachus, and
Theodotion occupied the third, fourth, and sixth columns of Origen’s Hexapla and together were
known as “the Three” (o¥ gu). 52 Of these, only Theodotion is relevant to my study. The recension
formerly attributed to Theodotion of the 2nd century c.e. is related closely to the 1st-century
b.c.e. KR. 53 Shenkel notes, however, that “Theodotion’s revision can be distinguished from the
KR, or proto-Theodotion, by its greater fidelity to the MT” (CRD 20; see also 11–12). In other
words, the traditional 2nd-century Theodotion cannot be “completely erased from history. 54

2.2.5. Origen’s Hexapla

By the 4th century, three recensions of the LXX were known: Origen, Hesychius, and Lu-
cian. 55 Little is known about Hesychius’s edition; Lucian’s recension was discussed above. Ori-
gen (ca. 185–254) was a scholar of Alexandria who probably composed his Hexapla in Caesarea,
completing it in 245. 56 The Hexapla was an enormous six-column volume (hence its name) that
contained various Hebrew and Greek texts in its six columns (table 2.4). Unfortunately for text-
critical purposes, it is not easy, according to Trebolle Barrera, to determine whether the fifth
column
was the LXX as known by Origen or a text he had already revised which he supplied with the
appropriate hexaplaric signs and additions. . . . Origen certainly completed a later, truly hexaplar
edition. . . . This later edition contained only the text of the LXX; the lacunae in the LXX in respect
of the MT (shorter) are completed by the text of Theodotion. As a result, this edition was supplied

52. For further information see P. W. Skehan, “Texts and Versions,” in The Jerome Biblical Commentary (ed. R. E.
Brown, J. A. Fitzmyer, and R. E. Murphy; Englewood Cliffs, N.J.: Prentice-Hall, 1968), 571–72; O’Connell, “Greek Ver-
sions (Minor),” 378–79; Tov, “Septuagint,” 182–84; idem, “State of the Question,” 5–7; idem, Textual Criticism of the He-
brew Bible, 144–47; Trebolle Barrera, Jewish Bible and the Christian Bible, 312–17.
53. Tov, Textual Criticism of the Hebrew Bible, 145; Trebolle Barrera, Jewish Bible and the Christian Bible, 314–17.
54. Trebolle Barrera, Jewish Bible and the Christian Bible, 316. For discussion of the identity of Theodotion, see Bar-
thélemy, Les devanciers d’Aquila, 144–57.
55. Trebolle Barrera, Jewish Bible and the Christian Bible, 309.
56. Ibid., 309, 311, 528–29.
 Transmission History 25

Table 2.4. Origen’s Hexapla

Column Contents
1 Hebrew text
2 Greek transliteration
3 Aquila
4 Symmachus
5 LXX
6 Theodotion

with the requisite diacritical signs. The hypothesis of this hexaplar edition replaces the other
hypothesis, according to which the fifth column of the Hexaplar contained the text revised by Ori-
gen, marked by asterisks and obeli, the same text later made into the edition of the LXX alone. 57
“Origen’s intention,” according to Shenkel, “was to make the Greek text chosen as the basis of
his revision conform with the definitively established MT” (CRD 18). Accordingly, material added
to the LXX from another source was marked with an asterisk (ì) at the beginning and a metobe-
lus (Û or Ù or ù) at the end. Material in the LXX not found in the MT was marked with an obelus
(– or ˆ or ÷) at the beginning and again a metobelus at the end. In the course of transmission
many of the symbols were deleted or incorrectly transcribed, leading to confusion. 58
The entire Hexapla was never copied, and the original was lost sometime in the 7th century
(except for the Psalms fragments rediscovered by Mercati). The fifth column, however, was
widely circulated in Palestine as early as the 4th century. 59 And this text was largely preserved in
the Syriac translation of Paul of Tella done in the early 7th century, known as the Syro-Hexaplar,
which shows many of the Hexaplaric signs. 60 That the letter lomadh (l) is found in the margin of
the Syro-Hexaplar beside readings also found in L cursives boe 2 indicates that lomadh refers to L
readings. 61 Tov proposes that Lucian’s 4th-century revision of an earlier Greek text is the most
important post-Hexaplaric text. 62

57. Ibid., 311 (emphasis original). P. E. Kahle believes that the LXX column did not contain diacritical signs; “The
Greek Bible Manuscripts Used by Origen,” JBL 79 (1960): 116. S. P. Brock argues that the symbols were already in the
text appropriated by Origen; The Recensions of the Septuagint Version of I Samuel (D.Phil. thesis; Oxford University, 1966),
37–42, cited in his “Origen’s Aims as a Textual Critic of the Old Testament,” Studia Patristica 10 (1970): 215 n. 3, repr.
in Studies in the Septuagint: Origins, Recensions, and Interpretations (ed. S. Jellicoe; New York: Ktav, 1974), 343 n. 3. Tov
notes (“Septuagint,” 185–86): “Recently discovered fragments of the complete Hexapla, especially the Psalm fragments
published by Cardinal Mercati [in Psalterii Hexapli Reliquiae], show no Hexaplaric signs in the fifth column. Hence
scholars have hypothesized that the original Hexapla did not contain these signs, and that these were inserted subse-
quently in a separate edition of the fifth column of the Hexapla.”
58. See H. B. Swete, Introduction to the Old Testament in Greek (2nd ed.; Cambridge: Cambridge University Press,
1914), 59–78; Klein, Textual Criticism of the Old Testament, 7–9; Trebolle Barrera, Jewish Bible and the Christian Bible,
311–12.
59. Swete, Introduction to the Old Testament in Greek, 76.
60. Ibid., 73–78; J. W. Wevers, “Septuagint,” in Interpreter’s Dictionary of the Bible (ed. G. A. Buttrick; New York:
Abingdon, 1962), 4:275; Klein, Textual Criticism of the Old Testament, 9.
61. Swete, Introduction to the Old Testament in Greek, 82–83 (citing 2 Kgs 9:9, 28; 10:24, 25; 11:1; 23:33, 35); cf.
Metzger, “Lucian and the Lucianic Recension,” 194.
62. Tov, “Septuagint,” 186.
 26 Chapter 2

Recognizing that the Hexapla was used to bring Greek texts into conformity with the MT has
implications for the chronology of the Books of Kings. If Greek chronological data at variance
with MT data were replaced by MT data to bring them into conformity, some of the data now
found in the Greek texts may actually belong to the MT system. Another result of assimilating
OG or proto-L data to the MT data may be that some original data are lost because we have no
other extant witnesses of the early Hebrew (Palestinian or Egyptian text-type). If such assimilation
took place, not all data having the same accession synchronism or length of reign necessarily re-
flect the original text. Analyzing the text in conjunction with extrabiblical synchronistic records
may determine which data are original, which intermediary, and which Hexaplaric.

2.2.6. Insights from Recensional Development

The above survey reveals some important insights. Understanding the recensional develop-
ment of the Greek texts is indispensable in evaluating the data for reconstructing the chronology
of Israel and Judah. The early Greek manuscripts are important witnesses to Hebrew texts no
longer extant. Codex Vaticanus, the earliest witness to Samuel–Kings, contains two text-types:
OG and KR. Since the OG chronological data concur with the proto-L data and the KR with the
MT data, this divergence causes Vaticanus to exhibit two chronological systems.
Cross’s analysis of the Qumran manuscripts and Shenkel’s study of the recensional develop-
ment of the Greek texts of Samuel — which Shenkel presumed also applied to the Books of Kings
(CRD 3, 11) — leads Shenkel to assert: “Viewed from the perspective of the historical develop-
ment of the Greek text, it is now evident that the Old Greek chronology, far from being the
artificial contrivance of late scribal activity, was the earliest chronology in the Greek textual tra-
dition and was already present in the Hebrew Vorlage of the earliest translation of the Books of
Kings” (CRD 110).
This may be true, but we must determine exactly which OG chronological data is original to
the OG text. It is not to be presumed that all data within the OG text necessarily reflect the Greek
chronological system. Allowance has to be made for contamination or cross-fertilization of texts
over centuries of transmission. 63 Tov views the Qumran texts as showing a multiplicity of texts
rather than Cross’s tripartite division and notes that the Qumran texts “are primarily a collection
of individual texts whose nature is that of all early texts and which relate to each other in an in-
tricate web of agreements and differences. In each text one also notices unique readings, that is,
readings found only in one source.” 64 This web might have involved transference of chronologi-
cal data if variants had arisen for some reason in one or another text. Textual corruption,
whether accidental or deliberate, can thus be attributed to many causes. 65
In addition, early Greek texts were revised to bring them into conformity with the proto-MT.
The KR data, for example, do not follow the OG/L system but the proto-MT data, now repre-
sented by the MT. Later revisions were made to conform the Greek text to the definitive MT —
which was one of Origen’s aims when he constructed his Hexapla. Subsequent revisions — for ex-
ample, proto-L — were influenced by Origen’s work. The extent to which OG, L, or KR were
made to conform to the MT or the extent of cross-fertilization between text-types in the centu-
ries before the stabilization of the MT probably cannot be determined, but analysis of the chron-

63. Cf. Cross, “New Qumran Biblical Fragment,” 23–24.

64. Tov, Textual Criticism of the Hebrew Bible, 160.
65. Ibid., 8–11, 233–85; Trebolle Barrera, Jewish Bible and the Christian Bible, 370–76.

Text is 12 points short

 Transmission History 27

ological data may indicate if, or where, data from one text may have been replaced by data from
another.

2.3. Thiele’s Approach to the Text of Kings

Thiele’s first writing on chronology was a 1944 journal article entitled “The Chronology of
the Kings of Judah and Israel.” 66 The Qumran scrolls had not yet been discovered. Nevertheless,
the Vaticanus text of 1–2 Kings had been published in 1930 in the Larger Cambridge Septuagint,
with the L variants in its apparatus. Thiele made no reference to the Greek texts, constructing his
chronology solely from the MT data and using the same principles that he applied in his later
writings (a critique of Thiele’s methodology appears in §6.3 and elsewhere). Thiele revised and
expanded his journal article into a monograph entitled The Mysterious Numbers of the Hebrew
Kings, published in three editions and interspersed with a number of journal articles. 67
In 1951, when Thiele’s first edition of MNHK was published, only initial discoveries at Qum-
ran had taken place, so the significance of the Dead Sea Scrolls was not known. Thiele does, how-
ever, discuss the Greek texts in this edition. After reciting the regnal data and the problems
associated with the MT, Thiele notes that
Already in the third century before the beginning of the Christian era there are indications that the
existence of these problems was recognized and that attempts were being made to deal with the
problem. The Septuagint, at that time translated from the Hebrew, contains a number of striking
variations in the chronological data from those contained in the Massoretic text, variations which it
seems were introduced in order to make possible a more harmonious pattern as regards the
lengths of reign of the Hebrew kings and the synchronisms. (MNHK 1 9 = MNHK 2 11; similarly
MNHK 3 39)
Right at the outset, Thiele acknowledges problems in the chronological data of the Hebrew and
Greek texts as early as the 3rd century b.c.e., but he suggests that it is the Greek text into which
variants were introduced to correct what were seen as problematic data in the Hebrew text. He
did not take into account that it might be the MT, a much later Hebrew text than that from which
the Greek was translated, that might be responsible for the variants seen in the extant Hebrew
and Greek texts.
Thiele constructed his chronology using the MT data and only then did he turn his attention
to the Greek texts. By virtually dismissing the Greek data Thiele not only excluded half of the in-
formation available to him in the Books of Kings, but he made a judgment in favor of the MT
data without first examining the evidence. The flaw in his method is obvious in light of the com-
plex transmission history reviewed earlier in this chapter.
For his Greek source, Thiele (MNHK 1 168 n. 2) relied on Burney, 68 giving the MT, LXX, and
L data in three columns and explaining, “By ‘LXX’ we mean the pattern of variations of the early
Greek texts as contrasted with the pattern now found in the Massoretic text” (MNHK 1 169). He
is apparently unaware of Thackeray’s 1907 observation that Codex Vaticanus exhibits two text-

66. E. R. Thiele, “The Chronology of the Kings of Judah and Israel,” Journal of Near Eastern Studies 3 (1944): 137–86.
67. Other articles by Thiele not previously mentioned include “New Evidence on the Chronology of the Last Kings
of Judah,” BASOR 143 (1956): 22–27; “The Synchronisms of the Hebrew Kings — A Re-evaluation: I,” Andrews University
Seminary Studies 1 (1963): 121–38; “The Synchronisms of the Hebrew Kings — A Re-evaluation: II,” Andrews University
Seminary Studies 2 (1964): 120–37; “Pekah to Hezekiah,” VT 16 (1966): 83–107; “An Additional Chronological Note on
‘Yaw, Son of ‘Omri,’ ” BASOR 222 (1976): 19–23.
68. C. F. Burney, Notes on the Hebrew Text of the Books of Kings (Oxford: Clarendon, 1903), xli–xliv.
28 Chapter 2

types in 1–2 Kings. 69 Thiele does observe that Ahab’s chronological data is the last instance of a
regnal formula in which the data in the LXX and L concur and that for the next five kings (Jeho-
shaphat, Ahaziah-I, Joram, Jehoram, and Ahaziah-J), the L data differ from LXX (MNHK 1 170,
176). But he does not discern that the 1 Kings data in Codex Vaticanus and L belong to the
Greek system and that the 2 Kings data belong to the Hebrew system. Failing to distinguish these
two text-types, he views the Greek sources of 1 Kings and 2 Kings as one text.
Thiele’s analysis of the Greek data, when compared with his dating systems for the Hebrew
data, leads him to conclude that the Greek texts used a different dating system, which he calls “in-
consequent accession-year reckoning” (MNHK 1 172). Thiele notes that the Greek data, when
compared to the MT, were “models of simplicity” (MNHK 1 180). Having already constructed his
chronology using the MT data, Thiele then has to explain why he has not used the simpler Greek
data, and his answer lay in arguing that the Hebrew data were earlier than the Greek:
If, however, the Hebrew figures were first, we may well understand how scholars at some early time
might have failed to see the basic harmony behind the apparently contradictory figures, and how
such a misconception as to the basic nature of the numbers involved could have given rise to an
attempt to produce a more harmonious pattern than was considered possible from the Hebrew
data. (MNHK 1 182)
Thiele’s apparent application of the text-critical principle that a difficult reading is preferred
over an easy one fails to take into consideration that the principle is invalidated when the difficult
reading is the result of scribal error. 70 Applying the principle to numbers gives even more reason
for caution, as numbers are highly prone to corruption, 71 and how numbers were written in non-
extant Hebrew manuscripts is subject to debate (see §7.3.6). Thiele does not take into account
the late nature of the MT, assumes that the MT data represent the original Hebrew text, and at-
tempts, for example, to illustrate the secondary nature of the LXX over the MT in the EDK:
In the recorded data in the Massoretic text for this period, the total years for the kings of Judah
from Rehoboam to the death of Ahaziah[-J] is ninety-five, whereas the total for Israel from Jero-
boam [I] to the death of Joram is ninety-eight. Why should this figure for Judah be three years less
than it is for Israel? . . . The increase of the years of Abijam from three to six looks strangely like an
attempt at adjustment of a length of reign to make possible the same total of years for Judah in this
period as for Israel. (MNHK 1 184)
Thiele proposes that the Greek texts are secondary based on their agreement for the regnal to-
tals for Israel and Judah. But, if the Greek texts giving 6 years for the reign of Abijam and a total
of 98 years for Israel and Judah are, in fact, correct, no amount of “chronological procedure” will
make correct the 3 years given to Abijam in the MT. Thiele was forced to make this proposal to
defend his use of the MT data.
In his analysis of the MT data and dating systems, compared with the LXX and L data and
dating system, Thiele notes that the simplicity of the Greek numbers is combined with an irreg-
ularly applied inconsequent accession-year system, whereas the MT data have a consistent dating
system. Thiele concludes, “It seems clear then, that of these three patterns of Hebrew chronol-
ogy, that of MT. is the earliest and best, that of LXX comes next in points of time and accuracy,
and that of Luc. is the latest and the most inaccurate” (MNHK 1 203). Thiele’s conclusion is con-

69. Thackeray, “Greek Translators,” 263. Thiele does not refer to Thackeray’s article.
70. Cf. Tov, Textual Criticism of the Hebrew Bible, 302–5.
71. J. W. Wenham, “Large Numbers in the Old Testament,” Tyndale Bulletin 18 (1967): 20–24; J. B. Payne, “The Va-
lidity of the Numbers in Chronicles,” Bibliotheca Sacra 136 (1979): 110.
 Transmission History 29

trary to that proposed later by Shenkel on the basis of his analysis of the Greek texts, which indi-
cated an earlier origin for OG and L than for the MT (CRD 18–21).
Thiele’s second edition of MNHK, published in 1965, removes the chapter on the Greek texts
and replaces it with a two-page discussion at the end of the book. After reproducing variants
again taken from Burney but leaving out the important datum that Abijam reigned 6 years in the
Greek texts and 3 years in the MT (MNHK 2 198; also MNHK 3 209), 72 he writes, “A careful study
of these variations reveals the fact that they are not the result of scribal errors but constitute ed-
itorial changes made with the object of correcting what were regarded as errors in the early He-
brew text.” Thiele makes no mention of the Qumran scrolls, even though initial findings had
been published 73 and constituted a threat to Thiele’s chronology based on the assumption that
the MT exhibited the earliest data.
In 1974 Thiele entered into discussion with Shenkel’s Chronology and Recensional Development
in the Greek Text of Kings, published in 1968. 74 His main point of contention involved the chronol-
ogy for the Omride period and the identification of the king of Judah in the Moabite campaign
(2 Kings 3), an identification that turns on the interpretation given to the variant regnal data for
the accessions of Joram and Jehoram. Shenkel’s argument is that two variant chronological sys-
tems are involved, one exhibiting Greek data and the other Hebrew. Thiele, on the other hand,
asserts that the variants show a coregency between Jehoshaphat and Jehoram, and he rejects the
explanation that two chronological systems might have been involved. 75
In the third edition of MNHK, published in 1983 well after the discovery of the Qumran
scrolls, Thiele continues to utilize the principles he had put into practice in his earlier works. He
includes a few pages concerning the Greek data of the Omride period (MNHK 3 89–94), but his
discussion about Omri’s reign claims that the variant data for Jehoshaphat’s accession, given
twice in Codex Vaticanus (Omri’s 11th year in 1 Kgs 16:28a OG/L and Ahab’s 4th year in 22:41
KR), reveals that the Greek data are late and artificial (MNHK 3 93). Thiele continues to not ap-
preciate what Shenkel had earlier discussed — that the variant data follow two different chrono-
logical systems (CRD 43–60). Apart from this discussion, Thiele repeats in his third edition the
material at the end of the second edition (MNHK 2 197–99 = MNHK 3 209–10).
Thiele’s assumption that the MT displays data earlier than the Greek texts gains no support
from the transmission history of the Hebrew and Greek texts. The MT text-type is not even found
at Qumran. 76 Unfortunately, Thiele never avails himself of the new insight that came to light with
the examination and publication of the Qumran Scrolls.

72. This table is reproduced with the same omission by Jones, 1 and 2 Kings, 20.
73. E.g., Albright, “Early Recensions,” 27–33 = QHBT 140–46; F. M. Cross, The Ancient Library of Qumran and Mod-
ern Biblical Studies (1st ed.; Garden City, N.Y.: Doubleday, 1958; rev. ed.; Garden City, N.Y.: Anchor Books, 1961); idem,
“History of the Biblical Text,” 281–99 = QHBT 177–95.
74. E. R. Thiele, “Coregencies and Overlapping Reigns among the Hebrew Kings,” JBL 93 (1974): 174–200.
75. Ibid., 175–86.
76. Cross, “Evolution of a Theory,” 314; idem, “Contribution of the Qumrân Discoveries,” 95 = QHBT 292.