volves a transposition* of the last clause, “O land, etc.,” from the place it holds in , which is immediately before "I will bring, etc.." This is made necessary by the structure of the qina-line, which calls for the shorter part after the cæsura. The phrase "without an inhabitant" is a favourite one in Je., e. g., 4′ 7. 29 911 269 3310 3422 4619 48° 5129. That it is not to be taken too literally is shown by Je. 4422.

The fourth and last str. adds picturesqueness of detail to the announcement of Philistia's devastation.-6. And thou wilt become pastures for shepherds] RV. renders, "and the sea-coast shall be pastures, with cottages for shepherds." But this involves the difficulty of treating 'sea-coast' as a feminine noun, while it is masculine everywhere else, even in v. 7. It is better to treat it as a case of vertical dittography from v. 7. The word 'pastures' too is in an almost impossible construction in . RV.'s 'cottages' are without solid foundation; a better rendering for the word is 'cisterns' or 'wells,' or even 'caves,' as in RVm.. But a simpler way out of the difficulty is to regard the word, which occurs only here, as a corrupt dittograph of the immediately preceding word, which it so closely resembles. Another treatment of the word is suggested by 6, which interprets it as 'Crete'; this in itself is quite possible; but, if adopted, the words 'Crete' and 'pastures' must exchange places, the former becoming the subject of the verb, viz., "And Crete, the border of the sea, will become pastures."† With the omission of "border of the sea" suggested above, this latter interpretation becomes very attractive; but it is hardly convincing because it is not likely that Philistia was known as 'Crete' in Zephaniah's time and was yet so named only once in the OT.. The line is smoother with the word omitted as in . For "pastures for shepherds," cf. Am. 12 Ps. 8313.—And folds for flocks] A picture of complete depopulation, crowded towns and villages giving place to pastoral solitudes. At this point a later editor,

* So Now., Marti, van H., Fag.. Wkl. (2. c.) om. as a gloss. Du. om. "I destroy you that there will be no inhabitant."

† So Wkl. (AOF. III, 232), van H.. Wkl. sets vv. 5. against the Philistines, but the islanders of Crete itself. Dn is too slight a basis for so novel an hypothesis.

apart as a separate oracle, directed not The occurrence of the forms

and

So We., GASm., Now., Dr. (?). Cf. Or.'s rendering, "shall become pastures for shepherds, and the land of Crete sheepfolds."

6

zealous for the pre-eminence of Judah, has inserted a line safeguarding the interests of his people.-7a. And the border of the sea will belong to the remnant of the house of Judah] The late origin* of this line is shown by the way in which it breaks the close connection between v. and v. 7b; the 'they' of 7b goes back for its antecedent, not to the 'remnant' of 7a, but to the 'flocks' of v. 6 Then, too, the use of the word 'remnant' presupposes at least the first deportation as having occurred. The same hatred of the nations in general and of the Philistines in particular is manifested here as in 2o Ob. 19 f. Zc. 95 ff. Am. 912 Is. 1114 Je. 492, all of which are of exilic or postexilic origin.-7b. By the sea will they feed; in the houses of Ashkelon at evening they will lie down] The original poem is here resumed and finished. The closing scene shows the former marts of trade and busy hives of men given over to the undisturbed possession of well-fed sheep, going in and out of the vacant houses at will, "with none to make them afraid." The qina-rhythm would be restored by the transposition of the first clause 'by the sea, etc.,' to the end of the line;† but the order of thought is more natural as in . The first clause in reads, "upon them will they feed"; the antecedent of 'them' can only be the 'pastures' of v. "; but this is to make a masculine suffix refer to a feminine antecedent. Hence, in part, the general adoption of the reading "by the sea," which involves only a very slight change of . Those holding to the integrity of the verse as a whole have felt compelled to make the verbs 'feed' and 'lie down' find their subjects in the Jews themselves (cf. 313 Is. 1430 Ez. 3414 Jb. 1119), rather than in the flocks or the nomad shepherds of these flocks. But this is a forced exegesis which, with the removal of v. 7a now keeping 7b and so far apart, becomes unnecessary. As between the shepherds and the flocks, the latter furnishes the more natural subject for the verbs. The objection usually urged, viz., that the prophet would not represent flocks as occupying the vacant houses, is not well taken; in no more effective way than this could he have represented the desolate and deserted state of the once populous region. The various attempts to emend the latter

6

* So We., Wkl. (1. c.), Marti, Siev., Beer, van H., Fag., Stk., Du., Kent.
† So Now., Marti, Kent. Du. treats it as a part of the interpolated matter.

78

part of this line (v. i.) seem wholly unnecessary.*-For Yahweh, their God, will visit them and turn their captivity] This line belongs with v. 7a and completes the editorial addition.† It clearly refers to the remnant of Judah and presupposes the exile. The reference to Judah here introduces a foreign element into a context which is concerned entirely with the Philistines. 'Visit,' frequently used of Yahweh's punitive activity, here denotes the exercise of his forgiveness and mercy. The promise of return from exile hardly accords with the view presented by the writer of v. 3, who contemplates the possibility of Judah's pious ones escaping from the approaching calamity. For the phrase 'turn their captivity,' V. H.AH, 199 f. 292. The alternative rendering 'turn their fortune' is less definite and forceful here.‡

The opening str. of this oracle is in tetrameter; the remaining three take on the gina-rhythm. The alien elements betray their character by their failure to conform to either of these measures.

3

Vv. 2b. . . . are omitted from the reconstructed poem as later accretions. Vv. 2b. are variants of a gloss explaining the figurative language of 2. It is impossible to say which line presents the gloss in its original form. The late origin of v. is shown by its conception of religion and by the fact that it evidently addresses itself to the Israelites, whereas the context is concerned with the Philistines. The same objection applies to v. 7. . Indeed, on the strength of vv..", this whole section is denied to Zephaniah by Schw., while Sta.GVI, 645, athetizes vv. 1-3 and Bu. (SK. 1893, pp. 394 ff., and Gesch. 89), vv. 4-7 (so also Kent). The argument against vv. 4-7 is that whereas in the genuine material Israel is represented as having done wrong and is therefore threatened with punishment, here Israel has been wronged by the nations and it is they that are to be punished. This, however, is not true of vv. 4-7, for not a word occurs in them charging Philistia with having injured Judah. The same kind of argument would also eliminate Am. 13-5. 6-8. 13-15 21-3, which are quite generally accepted as genuine. Just as Amos believed that Philistia would suffer in the general destruction about to be wrought either by the people of Urartu or by the Assyrians, so Zephaniah includes her in the universal devastation he anticipates. The prophets were men of broad vision, not limited in their range of interest and observation by a provincial horizon. They saw *There is no good reason for including this line with the rest of the verse as a late addition, as is done by Wkl., Marti, van H..

† So We., Preuschen, Now., Wkl., Marti, Dr., Siev., van H., Fag., Stk., Du., Kent. On the origin and meaning of the phrase, v. Preuschen, ZAW. XV, 1–74.

the history of their own people against the background of world-history. Not one of them looked upon his nation as a thing apart from the world's life, Amos, Jeremiah, Nahum, Habakkuk, Ezekiel and others prophesied the downfall of nations other than Israel. Zephaniah cannot be denied prophecies of the same sort, unless there be other evidence against them than the simple fact that they are directed against non-Israelites. What the relations between Philistia and Judah were in the days of Josiah we do not know, nor does Zephaniah tell us anything upon that subject, unless I' be an allusion to Philistine influence. But if Zephaniah looked for disaster to overwhelm the whole of western Asia, no special cause would be needed for a threat hurled against the Philistines.

The argument for treating vv. 1-3 as the conclusion of ch. 1 is unconvincing (contra Hi., GASm., Bu., et al.). The first chapter is complete as it stands. Moreover, the of 24 needs some antecedent material as a basis and this is supplied by 211.. The chief reason for combining vv. 1-3 with the preceding rather than the following context lies in the ex-, traneous material incorporated in this section which makes close connection with vv.. difficult. The treatment of this material as late removes this difficulty.

1. wipi wwipna] G, ovváxonte kal ovvdéônte; SO S. Ge... HP. 62, 86, 95, 147, 185 have the variant, συνδεήθητε. Σ, συλλέγητε σύνετε (probably an error for σúvre). I, convenite congregamini. Many mss. have, without), which is the normal writing of this form; v. Baer. Che. (Proph. of Is. on Is. 29'; but abandoned in CB.), win winn; so Gr., Bu. (SK. LXVI, 396), BDB., Now.K, Dr., Fag., Bew. (JBL. XXVII, 165), Kent. Hal. . Siev. wwpnai wip. Another suggestion is, derivingʼn from √ vp 'be hard' and

Both of the forms in .(וְקִישׁוּ or) הִתְקַשְׁשׁוּ וְקָשׁוּ .Van H .קשה / from ק'

Mare ȧn'. For similar combinations of Qal and Hithpo., v. Is. 29' Hb. 1'. The derivation of the vb. remains doubtful. Some would make it a denominative from wp, 'stubble,' meaning 'to gather stubble, sticks, etc.'; but when so used the obj. p,tan or 'sy always accompanies it, —a fact which seems to point to the vb. itself as having only the simple meaning 'collect,' 'gather.' In any case, the vb. cannot be here used denominatively. Van H.'s reading connects it with √ wp or wp, corresponding to the Arabic = 'mensuravit'; but the resulting sense is hardly satisfactory enough to warrant the necessary change in pointing involved. Mau. attaches it to wip (= Arabic 3) and renders 'bend yourselves'; but no such vb. occurs in Heb. and the Arabic vb., as Dav. points out, is a denominative, meaning not 'bend' but 'be bowshaped' or 'be curved in the back.' Stei. suggests √p, connected with up, 'be hard.' Ew. proposes the Aram. √ vưp · 'be old,' with a supposed primary meaning 'be withered,' and renders 'turn pale.' But

=

meaning.

none of these is more than a barren conjecture, providing no suitable nh] G тò åwaldevtov; SO S. I non amabilis. Van H., Schw.

.

Bew. (1. c.), 733 NS (√ D). The

foll. 6, p. most plausible explanation of is suggested by GASm., viz. Arabic ksf in classical speech 'cut a thread' or 'eclipse the sun'; but in the colloquial, 'to rebuff,' 'disappoint,' 'put to shame'; in forms IV and VIII it means 'be disappointed,' 'shy,' or 'timid' (v. Spiro's ArabicEnglish Vocabulary). This meaning as possible for the Heb. is supported by the Aram. √ which means 'lose colour,' 'be ashamed' (v. Jerusalem Targum on Nu. 1214 Ps. 35 697). Barth, Etymologische Studien, 61, derives it less easily from an Arabic ksf· 'be oppressed,' 'afflicted.'-2. pn n¬b] Rd. pr? vạn, tr. the letters nn (with a slight change in the second) to precede. For similar transpositions, cf.

=

עלוה, 10 ;ימשחו for ישמחו ן ;ואין for ואני,5 .Ho ;לנו for בול,312 .Am

for hy; 1310. 14, 178 for mix. This reading accounts for all the elements in , does away with the rare usage of an inf. cstr. with a (found besides only in Hg. 21), finds an exact parallel in Is. 295 (cf. 4013), is possibly supported in part by 6 (v. i.), and yields a line of the right length and structure. « Toû yevéobaɩ úμâs; similarly son þ Gr. (Monatsschrift, 1887, p. 506), . Schw. . We. S

n; so GASm., OortEm., Or., CB., Now., Marti, Hal., Dr., Siev. (adding ), Fag. (om. x), Roth., Du., Kent. But (1) this involves the introduction into classical Heb. of the usage, not found otherwise except in the late gloss upon this passage, which immediately follows; (2) it does not satisfactorily account for either the 7 or the of M; (3) it yields a line shorter than the measure set by the context; and (4) it is by no means certain that it represents the text that lay before 6, for the inf. construction of G suggests in its present form in that yevéolai might easily be the rendering of n in such a difficult context. The corruption may easily antedate 6. Bew. (1. c.), pn nibạ '(before) the appointed time is at an end.' Bu. (SK., 1893, p. 396), oppŋ nxhp (using first letters of p).—' pido] G ws aveos = 12 or EP. Bew. (1. c.), on. The only possible rendering of in this context is, "like chaff a day has passed away"; but this is altogether pointless.-] 6 = ¬ay; so S I SH; also Gr., We., GASm., OortEm., Or., CB., Now., Marti, Dr., Siev., Fag., Roth., Du.. Bu.. Van H 7. Hal. (using foll. Dr) Day.—0»] S L GB.N.A.Q HP. 48, 233 om.; so Schw., Gr., We., OortEm., CB., Now., Marti, Dr., Siev., van H., Du.. SH has it under asterisk. is supported by Y HP. 22, 36, 40, 42, 51, 62, 68, 86, 87, 91, 95, 97, 114, 147, 153, 185, 228, 238, 240.— ora] Explicable only as a strengthened negative, Ges. 152y; nowhere else in the list of fifty-one occurrences of ' is a second negative employed with it. The accumulation of particles is characteristic of late Heb..—197