The Hebraic style of the Apocalypse has been acknowledged in a general sense till the present generation, but scholars have hitherto done little to establish the fact by actual and detailed evidence. Now, owing on the one hand to this fact that the Hebraic character of the Apocalypse had not been established by actual proofs, and on the other to the vast mass of fresh knowledge of vernacular Greek brought to light by the researches of Grenfell, Hunt, Thumb, Moulton, milligan, and others, a new attitude has recently been adopted by certain scholars on this question, and some have gone to the extreme length of denying altogether the presence of Hebraisms in the Apocalypse except in sections that are translated from the Semitic. Thus professor Moulon (Gr. 8-9) affirms that "even the Greek of the Apocalypse itself does not seem to owe any of its blunders to 'Hebraisms.' The author's uncertain use of cases is obvious to the most casual reader. . . . We find him perpetually indifferent to concord. But the less educated papyri give us plentiful parallels from a field where Semitism cannot be suspected. . . . Apart from places where he may be definitely translating a Semitic document, there is no reason to believe his grammar would have been materially different had he been a native of Oxyrhnchus, assuming the extent of Greek education to be the same."

{cxliii} This is not only an extravagant, but, as we shall presently discover, a wrong statement of the case, and called forth a rejoinder from Professor Swete (Apoc.² p. cxxiv, note), who wrote: "It is precarious to compare a literary document with a collection of personal and business letters, accounts, and other ephemeral writings; slips in word-formation or in syntax, which are to be expected in the latter, are phenomenal in the former, and if they find a place there, can only be attributed to lifelong habits of thought. Moreover, it remains to be considered how far the quais-Semitic colloquialisms of the papyri are themselves due to the influence of the large Greek-speaking Jewish population of the Delta." My own studies, which have extended from the time of Homer down to the Middle Ages, and have concerned themselves specially with Hellenistic Greek, so far as this Greek was a vehicle of Hebrew thought, have led me to a very different conclusion on this question, and this is, that the linguistic character of the Apocalypse is absolutely unique.⁽¹⁾

Its language differs from that of the LXX and other versions of the O.T., from the Greek of the Apocrypha and Pseudepigrapha, and from that of the papyri. Of course it has points in common with all these phases of later Greek, but nevertheless it possesses a very distinct character of its own. No literary document of the Greek world exhibits such a vast multitude of solecisms. It would almost seem that the author of the Apocalypse deliberately set at defiance the grammarian and the ordinary rules of syntax. But such a description would do him the grossest injustice. He had no such intention. He is full of his subject, and like the great Hebrew prophets of old is a true artist. His object is to drive home his message with all the powers at his command, and this he does in many of the sublimest passages in all literature. Naturally with such an object in view he has no thought of consistently breaking any rule of syntax. How then are we to explain the unbridled license of his Greek construction? The reason clearly is that, while he writes in Greek, he thinks in Hebrew, and the thought has naturally affected the vehicle of expression. Moreover, he has taken over some Greek sources already translated from the Hebrew and has himself translated and adapted certain Hebrew sources. Besides he has rendered many Hebrew expressions literally and not idiomatically -- constantly in his own original work and occasionally in his translations. His translations {cxliv} in a few cases presuppose corruptions in the Hebrew sources. But this is not all. He never mastered Greek idiomatically -- even the Greek of his own period. To him very many of its particles were apparently unknown, and the multitudinous shades of meaning which they expressed in the various combinations into which they entered were never grasped at all, or only in a very inadequate degree. On the other hand, he is more accurate in the use of certain Greek idioms than the Fourth Evangelist. Notwithstanding its many unusual and unheard expressions, the Book stands in its own literature without a rival, while in the literature of all time it has won for itself a place in the van.

I will not give a list of the chief Hebraisms in the Apocalypse which are sufficient to prove that it is more Hebraic than the LXX itself.

(i.) The Greek text needs at times to be translated into Hebrew in order to discover its meaning and render it correctly in English.

(a) The resolution of the participle in one of the oblique cases (gen. dat. or acc.), or of an infinitive, into a finite verb in the following clause, which finite verb should have been rendered idiomatically in Greek by a participle or by an infinitive respectively. We have here a frequent Hebrew idiom whicih cannot be explained from vernacular Greek and which, not having been recognized, has led to mistranslations of the text in every version of the Apocalypse down to the present day.⁽²⁾
     "It is," writes Driver (Hebrew Tenses, 163), "a common custom with Hebrew writers, after employing a participle or infinitive, to change the construction, and if they wish to subjoin other verbs, which logically should be in the participle or infinitive as well, to pass to the use of the finite verb." Here we have the explanation of a dozen of passages in our author, which have been generally mistranslated in all the versions. In a few cases they are rightly translated, and then only through deliberate emendation of the text.⁽³⁾
     The idiom of a participle continued by a finite verb is rendered literally into Greek in the LXX in Gen 27³³, Is 14¹⁷, and idiiomatically in Is 5^8.23, Ezek 22³. But it is rendered literally comparatively seldom in the LXX, whereas our text it occurs ten times and most probably eleven originally, as we shall see presently. In a few cases the Syriac, Latin, Bohairic, and A.V. are right, but probably unconsciously. This idiom emerges in the first chapter in ^5-6 and recurs in ¹⁸ 2^2.9.20.23 3⁹ 7¹⁴ 14^2-3 15³.

(a) In 1^5-6 we have tw/| avgapw/nti h`ma/j kai. lu,santi h`ma/j ) ) ) kai. evpoi,hsen h`ma/j basilei,an, which should therefore by rendered, "Unto Him that loveth us . . . and He  made us."

(b) The failure to recognize this idiom in 1¹⁸ has led most scholars to mispunctuate the text, and the rest, like Wellhausen and Haussleiter, to excise o` zw/n. The translation of o` zw/n kai. evgeno,mhn nekro,j should be 1^17c "Fear not: I am the first and the last, 1¹⁸ And He that liveth and was dead." Thus we recover the right sense.

(g) Again we have in 2²³ evgw, eivmi o` evraunw/n ) ) ) kai. dw,sw another example of this idiom = "I am He that trieth . . . and giveth." HEre the Hebrew in our author's mind would be yttnw !TeBoh; or even !Tea,w>: cf. Dan 12¹², and see vol. ii. 392 n. For a further treatment of this idiom the reader can consult the note in vol. i. 14 sq.

(d) Next, attention should be drawn in 20⁴, where originally I feel assured there was another instance of this idiom; for the oi[tinej in tw/n pepelekisme,nwn ) ) ) kai. oi[tinej ouv proseku,nhsan is obviously an insertion made by John's literary executor, who edited 20⁴-22 after John's death. {cxlvi} See vol. ii. 182, 183. The insertion of oi[tinej is against our author's usage. In practically every instance the failure to recognize this idiom has led both to a mistranslation of the text and a misrepresentation of the meaning. Since the various instances of this idiom are dealt with as they arise, alike in the Commentary and Translation, I will bring forward only two more here to show how important it is that it should be accurately rendered.

(e) In 14^2.3h` fwnh. h] h;kousa w`j kiqarw|dw/n kiqarizo,ntwn evn tai/j kiqa,raij auvtw/n⁽⁴⁾kai. a;|dousin w`j wv|dh.n kainh,n = "The voice which I heard was as the voice of harpers, harping with their harps and singing as it were a new song":

(z) 2²⁰h` le,gousa e`auth.n profh/tin kai. dida,skei = "who calleth herself a prophetess and teacheth" (not "and she teacheth," R.V.).

(b) In 13¹⁵ we have a resolution of the infinitive into a finite verb in the following clause as in Hebrew (see quotation above from Driver's Hebrew Tenses). Thus kai. evdo,qh // auvth/| // dou/nai ) ) ) kai. poih,sh| = ttl hl !tnyw = "And it was given  unto him to give . . . and to cause." See vol. ii. 420, footnote.

(c) Just as in (a,b), the constructions under this head are quite impossible and unintelligible as Greek, but are full of meaning as literal reproductions of a Hebrew idiom. 

(a) The first is 12⁷o` Micah.l kai. oi` a;ggeloi auvtou/ tou/ (> a 046) polemh/sai. We have here a classical Hebrew idiom: see vol. i. p. 322. The words righly understood are most vivid: "Michael and his angels had to fight with the dragon." It is remarkable that the MSS allowed this astonishing Greek to survive in any form.

(b) The same idiom recurs in 1310 where only A has preserved it in a slightly corrupt form: ei; tij ) ) ) avpoktanqh/nai( // auvto.n // evn macai,rh| avpoktanqh/nai (= greh'le awh ) ) )⁵greh'le rva) = "if any man is to be slain with the sword, with the sword must he be slain." In vol. i. 356, I have shown that the Greek translators found great difficulty in rendering this idiom, and resorted to at least half a dozen different ways. The same idiom is to be found in Ethiopic. In kau,swn e;stai (Luke 12⁵⁴) the e;stai is rendered by the Eth. lamedh before the infinitive. Thus our author introduces a new use of the inf. into Greek which none of the grammarians has recognized. 

(d) Again an expression may be possible in Greek as regards form but wrong in regard to sense. Thus in 2²² ba,llw eivj kli,nhn as a piece of Greek is meaningless in its context but full of significance if retranslated into Hebrew. See vol. i. 71.

{cxlvii} (e) The finite verb in Hebrew is translated literally, when idiomatically it should be rendered by a participle. Cf. 1¹⁶ h` o;yij auvtou/ w`j o` h[lioj fai,nei (= ryay wmwk) = "his face was as the sun shining" (not "shineth"). See vol. i. 31.

(f) The Greek phrase ku,rioj o` qeo.j o` pantokra,twr requires to be retranslated in order to punctuate and translate it rightly. It should not be punctuated as in WH with a comma after ku,rioj and another after qeo,j) In fact no commas should intervene at all. The entire phrase is found in 2 Sam 5¹⁰, 1 Kings 19^10.14, Hos 12⁵⁽⁶⁾, Amos 3¹³ 4¹³ 5¹⁴ etc. (= twabch yhla hwhy), and often ku,rioj pantokra,twr, Hab 2¹³, Hag 1^2.5, Zech 1³. Next it is to be observed that o` pantokra,twr in all three cases is a rendering of twabc (with or without the art.) following the construct case. Hence o` pantokra,twr is the equivalent of a gen. in Greek dependent on the  noun that precedes it. Thus nothing -- not even a comma (as in WH) should intervene between o` qeo,j and o` pantokra,twr: cf 4⁸ 11¹⁷ 15³ 16^7.14 19^6.15 21²².⁽⁶⁾ Thus we see that on textual grounds 1⁸ (ku,rioj o` qeo,j( o` w]n kai. o` h=n kai. o` evrco,menoj( o` pantokra,twr) is the interpolation of an ignorant scribe, who was unacquainted with the origin of this divine title. The context also is against it. See vol. ii. 38, n. 4. Furthermore, it follows that it is not to be rendered "the Lord God, the Almighty," as in R.V., but as "the Lord God Almighty."

(g) When Hebrew and Greek words agree as to their primary meanings, the secondary meanings of the Hebrew words are in a few cases assigned to the Greek. Here retranslation is necessary.

(a) In 10¹ we have the extraordinary phrase oi` po,dej auvtou/ w`j stu,loi puro,j) Here, as I have shown in vol. i. 259 sq., po,dej is to be rendered as "legs."

(b) Again poimai,nein is to be rendered as "to break" in 2²⁷ 12⁵ 19¹⁵ for the same reason: see vol. i. 75 sq.

(g) Again in 1⁵ the primary sense of prwto,tokoj( "firstborn," is eclipsed by the secondary denoting "chief" or "sovereign" -- which secondary sense it derives originally from {cxlviii} the Hebrew rwkb. Cf. Job 18¹³ where twm rwkb = "the most deadly disease," and Is 14³⁰ ~yld yrwkb = "the poorest." See note on 1⁵ in the Commentary. 

(d) Possibly in 1⁷ko,ptesqai evpi, we have an instance in which a secondary meaning of l[ is assigned by our author to evpi,)

(h) Other Hebrew idioms literally reproduced in the Greek need to be retranslated in order to appreciate their exact meaning.

(a) 2²³dw,sw = "to requite," as !tn in Jer 17¹⁰ on which 2²³ is based.

(b) 3⁸ de,dwka evnw,pio,n sou qu,ran = "I have set," etc. See vol. i. 41.

(g) 3⁹ ivdou. didw/ = "behold I will make": vol. i. 41.

(d) 5⁶ evn me,sw| ) ) ) evn me,sw| = "between . . . and": see vol. i. 140.

(e) 6¹ le,gontoj w`j fwnh, (AC 046 and most curss.) bronth/j. Here w`j fwnh, = lwqk, which our author may have in his mind, and which = w`j fwnh, or w`j fwnh/|) By a slip our author wrote the former. The same misreading is found in Is 5¹⁷ etc.: see vol. i. 161.

(z) 12¹¹evni,khsan dia. to. ai-ma tou/ avrni,ou ) ) ) kai. ouvk hvga,phsan( ktl)) where the kai, is to be rendered by "seeing," as vav in Hebrew. The kai, (= vav) introduces a statement of the condition under which the action denoted by evni,khsan took place. See footnote 7, vol. ii. 417. The same Hebraism recurs in 18³ 19³. 

(h) 12¹⁴avpi. prosw,puou tou/ o;fewj = Xxnh ynpm -- "because of the serpent": see vol. i. 330.

(q) 21⁸toi/j de. deiloi/j ) ) ) to. me,roj auvtw/n = ~qlh ) ) ) bl ykrl. The dative is to be explained as a reproduction of the Hebrew idiom where l introduces a new subject: see vol. ii. 216, footnote.

(i) 22⁵o` qeo.j fwti,sei evpV auvtou,j)Our author uses fwti,zein as a transitive verb in 18¹ 21²³, and naturally we expect it to be used as such here. Moreover, the context itself is against using it here intransitively; for "God will shine upon them" is not a likely expression. If, however, we understand "His face" as in the Hebrew, Ps 118²⁷, we can render fwti,zein transitively as in 18¹ 21²³ and give a most excellent meaning to the passage: "we will cause his face to shine upon them": see vol. ii. 210 sq.

(i) The future is to be rendered by the pres. in 4^9-10; for here the future represents the Hebrew imperfect in a frequentative sense. Thus o[tan dw,sousin ) ) ) do,xan ) ) ) pesou/ntai, "where they give . . . glory . . . they fall down." This mistranslation {cxlix} of the Hebrew imperf. is often met with in Greek translations. It occurrence in our author, who thinks in Hebrew, is therefore very natural. See vol. ii. 399, footnote. The future in 13⁸ proskunh,sousin should be rendered as proseku,noun (= Hebrew imperf.).

(j) The present in 9⁶ is to be rendered as a future, where feu,gei represents the Hebrew imperf. in our author's mind: as a past imperf. in 7¹⁰ kra,zousin, 12⁴ su,rei, 16²¹ katafeu,gei. 

(a) Nominativus pendens. -- This construction is found in 2²⁶ 3^12.21 o` nikw/n dw,sw auvtw/|, 6<sup>8</sup> o` kaqh,menoj evpa,nw auvtou/ o;noma auvtw/|)⁽⁷⁾ In other passages, however, our author has assimilated the construction more to the Greek construction by changing the nom. into the dat., 2^7.17 (21⁶) tw/| nikw/nti dw,sw auvtw/|, 6⁴ tw/| kaqhme,nw| evpV // auvto.n // evdo,qh auvtw/|: cf. Matt 5⁴⁰. This construction is very frequent in the LXX owing to its frequency in the Hebrew.

(b) The oblique forms of the personal pronoun are added to relatives. 3⁸ h]n ouvdei.j du,natai klei/sai auvth,n, 7² oi-j evdo,qh auvtoi/j, 7⁹ o]n ) ) ) auvto,n, 13^8.12: cf. also 12^6.14 (o[pou ) ) ) evkei/) 17⁹ (o[pou ) ) ) evpV auvtw/n). The pronoun is, of course, pleonastic in the Greek but not in the Hebrew, where, since the relative is uninflected, it supplies the inflection needed. This pleonastic use of the pronoun is found also in Mark 1⁷ (= Luke 3¹⁶), 7²⁵ 9³ (oi-a ) ) ) ou[twj), 13⁹, J 1²⁷, Acts 15¹⁷. Examples of this idiom occur exceptionally in the koinh,. It is found also in Early English. But in our text its frequency is due to Semitic influences.

(c) (a) A noun or participial phrase, which is dependent on or in apposition to a preceding gen. dat. or acc., may stand in the nom., if it is preceded by the art., though Greek syntax would require it to agree with the oblique case that goes before it. This peculiar idiom is derived from the Hebrew, according to which the noun or phrase which stands in apposition to a noun in an oblique case remains unchanged. Instances of this usage occur in the LXX; but what is a rare phenomenon in the Greek version of the O.T. (cf. Ezek 23^7.12)⁽⁸⁾ is a well-established idiom in the Greek text of the Apocalypse.⁽⁹⁾ Our {cl} author has, in fact, adopted a Hebraism into his Greek, and naturalized it there. Thus it has become a marked characteristic of his style: cf. 1⁵ 2^13.20 3¹² [8⁹] 9¹⁴ 14¹² 20². In these passages observe that the noun is always preceded by the art. 1⁵ avpo. VIhsou/ Cristou/ o` ma,rtuj o` pisto,j( 2²⁰ th.n gunai/ka VIeza,bel( h` le,gousa e`auth.n profh/tin( 3¹² th/j kainh/j VIerousalh,m( h` katabai,nousa, [8⁹tw/n ktisma,twn ) ) ) ta. e;conta yuca,j]. How readily a Jew could adopt or fall into such a solecism when using an inflected language, is illustrated by Nestle (Textual Criticism of the Greek Testament, p. 330), who notes the following gem from Salomon Bar  in his translation of the Massoretic note at the end of the Books of Samuel (Leipzig, 1892, p. 158, "ad mortem Davidis rex Israelis." 

(b) If the art. is omitted, then the word or phrase is put in the same case as the noun that precedes it. Contrast 9¹⁴ tw/| avgge,lw|( o` e;cwn t) sa,lpigga( and 7² 9¹⁷ 13¹ 14⁶ 15² 18¹ 20¹a;ggelon ) ) ) e;conta th.n klei/n) 

(g) But this rule does not apply to le,gwn. Thus in 14⁶ we have ei=don a;llon( a;ggelon peto,menon ) ) ) e;conta euvagge,lion ) ) ) le,gwn. But le,gwn (or le,gontej) stands by itself: it appears most indeclinable. This may be due to the fact that it may reproduce rmoale in our author's mind. Cf. 4¹ h` fwnh. ) ) ) le,gwn: 5<sup>11</sup> o` avriqmo.j auvtw/n ) ) ) le,gontej, 11¹ evdo,qh moi ka,lamoj ) ) ) le,gwn( 11¹⁵ fwnai. ) ) ) le,gontej) This solecism is, of course, found in the LXX: cf. Gen 15¹ 22²⁰ 38¹³ 45¹⁶ 48²⁰ etc.

(d) e;cwn follows an acc. when not preceded by the art. in 5⁶ avrni,on e`sthko.j ) ) ) e;cwn( 14<sup>14</sup> o[moion ui`o.n avnqrw,pon( e;cwn( 17³ qhri,on ) ) ) e;cwn. But in 5⁶ 17³ it seems corrupt for e;con. In 14¹⁴e;cwn is right and kaqh,menon o[moion, which precedes, is a slip for nom.

(b) In 13¹¹ we have the extraordinary statement that the second Beast had two horns like a lamb and spake like a dragon! The first idea may be suggested by Matt 7¹⁵ "Beware of false prophets, which come to you in sheep's clothing, but inwardly are ravening wolves." See, however, vol. ii. 451 sq. But what is the explanation of the second idea -- "he spake like a dragon"? A dragon does not speak. If the text had read "like the dragon," it  might have recalled the temptation of Eve in Eden. But the lack of the article can be explained by the translator's reading !yntk as !yNIt;K. instead of !yNIT;K;; and, since kai. evla,lei = rbdtw, the latter is most probably corrupt for dbatw, as in 2 Chron 22¹⁰ (cf. 2 Kings 11¹).Thus 13^11c should be read: "but he was a destroyer like the dragon." Thus brings our text into line with Matt 7¹⁵ (quoted above) and prepares us for the statement in 13¹⁵ that this second Beast put all to death that did not worship the first Beast.

(g) Again in 15^5.6 there are two expressions, hvnoi,gh // o` nao.j th/j skhnh/j tou/ marturi,ou // evn tw/| ouvranw/|, and evndedume,noi // li,qon // kaqaro.n lampro,n, which are clearly corrupt. Inferior MSS (025. 046) have corrected the second into li,non) A new vision begins with these verses. It is clear that no Jew writing originally in Greek could have used either of the obelized phrases. But, as I have shown in vol. ii. 37 sq., what is most probably the true text can be discovered by retranslation into Hebrew. In the first passage, 15<sup>5</sup> o` nao.j th/j skhnh/j tou/ marturi,ou evn tw/| ouvranw|/ = ~ymXb d[wm lha lkyh, which was corrupt for ~ymXbX ~yhla lkyh = o` nao.j tou/ qeou/ o` evn tw/| ouvranw/|( a phrase which we find exactly in 11¹⁹ accompanied by the same verb hvnoi,gh and the repeated article. In 15⁶ // li,qon // is to be explained by a mistranslation of vve, which can be rendered either by li,qoj( ma,rmaroj( or by bu,ssinoj. Here the latter, of course, is the right rendering.

(b) These two passages naturally lead to the inquiry: Did John translate the Hebrew source himself, or did he adopt an independent Greek version of it? The fact that every phrase and construction in 15^5-8 are distinctly our author's, furnishes such strong evidence for the former hypothesis that it seems necessary to accept it. If this is right, then we must conclude {clii} that our author inserted here a translation which, while reproducing exactly the corrupt Hebrew before him in 15⁵ and a wrong rendering of a Hebrew word in 15⁶, would have been corrected later, if he had had the opportunity of revision. Repeatedly we find traces of unfinished work in our author, which a revision would have removed. Thus 12^14-16 18⁴ (see vol. i. 330-332, ii. 96 ad fin.) are meaningless survivals of earlier expectations. Unhappily the work was revised by one of his disciples who was quite unequal to the task, and to whom we owe some of the worst confusions in the Book. See,  however, p. lxiv ad fin.

(c) For other passages which need to be retranslated in order to discover their meaning, see 18²² (mousikw/n), 18¹⁹ evk th/j timio,thtoj auvth/j ) ) ) hvrhmw,qh) 
 

Scanned and edited by Brad Johnson