Orthography, Textual Criticism, and the Poetry of Job

In an article published in Eretz Israel in 1969, David Noel Freedman called attention to the great number of forms in the book of Job that are spelled without internal vowel markers (matres lectionis) where one might expect them.¹ This is particularly remarkable in cases where diphthongs are contracted and unmarked by matres. Thus, we have instead of (18:12; 40:16), instead of (9:34; 13:21; 20:25; 33:7), instead of (12:19; 33:19), instead of (13:23; in 32:7), instead of (10:18; 15:13; in 12:22; in 28:11), instead of (34:32; in 12:7, 8; in 30:19), instead of (30:13; 35:3), instead of (9:9; in 22:1), and so forth. By Freedman's count, there are forty-two such forms with contracted and unmarked diphthongs in the book, though one might also add conservative spellings like (2:2), (2:9), (11:3), (11:15), (11:17), (18:16), (37:2; in 28:26), (29:16; in 13:6; in 31:13), and (29:24).² Moreover, there are a number of Ketiv readings that suggest a conservative orthography: for (5:18), for (9:13), for (15:15), for (20:11), for (21:20), for (24:1), for (27:15), for (31:20), for (37:12), for (38:41), for (39:26), for (39:30), and for (40:17).

Drawing on the results of his joint dissertation with Frank Moore Cross, which argued in part that such spellings typified epigraphic Hebrew of the preexilic period and generally of the north (Israel as opposed to Judah),³ a thesis that despite a few rare exceptions has held true,⁴ Freedman concluded that the conservative orthography indicates a provenance in the north during the preexilic period. The study has prompted at least one scholar to judge that it is now difficult to maintain a date later than the seventh century B.C.E.⁵ Not all are so sanguine about Freedman's approach, to be sure. Yet even James Barr, Freedman's most severe critic, has conceded that the book of Job manifests an unusually high concentration of such forms, indeed, "the highest anywhere in the Bible."⁶ The implications of this judgment for the exegesis of Job, however, have not been explored. Such an exploration is the purpose of this essay. Before doing so, though, I would like to expand Freedman's database.

I. Beyond MT^L

The publication of 4QpalaeoJob^c from Qumran has lent credibility to Freedman's article, since internal matres in that manuscript are, with only two exceptions, entirely absent:⁷

	MT	4QpalaeoJobc
13:24		(so, too, MSKenn 170)
13:36		(also 3 MSS)
13:26		(cf. in several MSS at 31:18)
13:27		(cf. in 33:11)
14:16		(so, too, many MSS)
14:17		(similarly 2 MSS)

While it may not be possible to prove that the orthographic archaism in this manuscript reflects the autograph, 4QpalaeoJob^c does demonstrate that the MT cannot be decisive on the question of orthography. If the orthography suggested by 4QpalaeoJob^c is original, the introduction of internal matres in the other Qumran manuscripts of Job (2QJob; 4QJob^a; 4QJob^b) as well as in the MT is part of the book's history of interpretation.⁸ Moreover, while Freedman based his analysis on the Leningrad Codex (MT^L) by way of BHK³, the Aleppo Codex (MT^A), the Bomberg Bible (MT^B), and other manuscripts often show additional forms without internal matres. The following is only a small sample of literally hundreds of Hebrew variants without internal vowel markers:

MTL	Other Manuscripts
(1:3)	(3 MSS)
(1:4)	(1 MS)
(1:5)	(8 MSS)
(1:14)	(5 MSS)
(3:22)	(1 MS)
(5:17)	(1 MS)
(5:17)	(1 MS)
(6:16b)	(1 MS)
(6:24)	(2 MSS)
(8:10)	(1 MS; 7 MSS )
(10:3)	(1 MS)
(10:22)	(MTAB, many MSS)
(12:5)	(4 MSS)
(13:1)	(1 MS)
(13:6)	(1 MS)
(13:11)	(1 MS)
(14:3)	(2...