CHAPTER XIX
MODE ( ;EGKLISIS)
Introductory. For a brief sketch of the number of the modes and the reasons for treating the indicative as a mode see Conjugation Of the Verb, chapter VIII, v, (a). References are there given to the pertinent literature. The use of a;n is given a brief treatment below in connection with the modes. The subject of conjunctions is divided for logical consistency. The Paratactic Conjunctions belong to the same division with Paratactic Sentences, while Hypotactic Conjunctions fall under Hypotactic Sentences. The conjunctions could of course be treated in separate or as a division of the chapter on Particles (XXI). That will be there done (v, 1) for Paratactic Conjunctions. Hypotactic Conjunctions will there receive only summary treatment and can best be discussed in detail in connection with subordinate clauses. And there are advantages in the present method. It needs to be said also that the division of the treatment of modes into those of Independent and Subordinate Sentences (A and B) is purely arbitrary and for the sake of clearness. There is no real difference in the meaning of a mode in an independent and a dependent sentence. The significance of each mode will be sufficiently discussed under A (Independent Sentences). The inclusion of all the subordinate clauses under mode is likewise for the sake of perspicuity. Voice, tense, mode thus stand out sharply.1 The difficulty of making a clear distinction in the significance of the mode has already been discussed in chapter VIII, pp. 321 ff. A mood is a mode of statement, an attitude of mind in which the speaker conceives the matter stated.2 Apollonius Dyskolos first described (moods as yucikai. diaqe,seij. That is a correct description of the function of mood as distinct from voice and tense.3
912 A GRAMMAR OF THE GREEK NEW TESTAMENT
The mode is the manner of the affirmation, while voice and tense have to do with the action of the verb (voice with relation of the subject to the action of the verb, tense with the state of the action). But even so the matter is not always clear. The mode is far and away the most difficult theme in Greek syntax. Our modern grammatical nomenclature is never so clumsy as here in the effort to express "the delicate accuracy and beauty of those slight nuances of thought which the Greek reflected in the synthetic and manifold forms of his verb."4 So appeal is made to psychology to help us out. "If the moods yucikai. diaqe,seij, why is not every utterance modal? Why does not every utterance denote a state of the soul? A universal psychology would be a universal syntax."5 Every utterance does denote a state of the soul. This is one argument for treating the indicative as a mode. The verb is necessarily modal from this point of view. But the term is naturally confined to the finite verb and denied to the infinitive and participle. Dionysius Thrax does call the infinitive a mode, but he is not generally followed.6 Gildersleeve7 notes also that "moods are temporal and tenses modal." He sees that the order moods and tenses is the natural sequence in the English (cf. chapter VIII, v, p. 320), but he follows the order tenses and moods in his Syntax of Classical Greek, though it is hard to separate them in actual study. Gildersleeve8 laments also that dia,qesij came to be applied to voice and e;glisij to mode (cf. enclitic words as to accent), "but after all tone of utterance is not so bad a description of mood." It is possible that at the beginning the indicative was used to express all the various moods or tones of the speaker, as the accusative case originally included the whole field of the oblique cases. It was only gradually that the other moods were developed by the side of the indicative (thus limiting the scope of the ind.) to accent certain "moods of mind, i.e. various shades of desire,"9 more sharply. Thompson calls this development "artificial," since no other race but the Greeks have preserved these fine distinctions between indicative, subjunctive, optative, imperative, not to say injunctive
MODE ( EGKLISIS) 913
and future indicative (almost a mode to itself). But that is too severe a term, for the modes were a gradual evolution. The injunctive was the unaugmented indicative, like lu,ouà lu,esqeà lu,sasqeà lu,qhteà lu,eteà lu,sateà sce,j.10 Moulton11 says: "Syntactically it represented the bare combination of verbal idea with the ending which supplies the subject; and its prevailing use was for prohibitions, if we may judge from the Sanskrit, where it still remains to some extent alive. The fact that this primitive mood thus occupies ground appropriate to the subjunctive, while it supplies the imperative ultimately with nearly all its forms, illustrates the syntactical nearness of the moods. Since the optative also can express prohibition, even in the N. T. (Mk. 11:14), we see how much common ground is shared by all the subjective moods." Yes, and by the indicative also. The present indicative is often a practical future. Originally the subjunctive had the short vowel (cf. i;omen in Homer). The distinction between the indicative and subjunctive is not always clear.12 The subjunctive in Homer is often merely futuristic. The affinity between the subjunctive and the optative is very close. The indicative continued to be used in the volitive sense (past tenses) and of command (future tense). Thus the other modes were luxuries of the language rather than necessities, while the indicative was the original possessor of the field. As already shown (chapter VIII, v) the injunctive survived in the imperative and subjunctive. The future indicative continued to fulfil the function of all the modes (cf. the indicative before the rise of the other modes). Thus the future indicative may be merely futuristic, or volitive, or deliberative. The same thing is true of the subjunctive and the optative. Cf. Moulton, Prol., p. 184 f. Thompson (Syntax, p. 186) curiously says that "thee indicative, however, assumed some of the functions of the other moods." If he had said "retained," he would have it right. He had just said properly enough: "It would be an error, with regard both to their origin and functions, to regard the moods as separate and water-tight compartments." The early process was from simplicity to variety and then from variety to simplicity (cf. again the history of the cases). The struggle between the modes has continued until in the modern Greek we have practically only the indicative and the subjunctive, and they
914 A GRAMMAR OF THE GREEK NEW TESTAMENT
are in some instances alike in sound (Thumb, Handb., p. 115 f.). The subj. is "considerably reduced" in use in the modern Greek. The optative has disappeared entirely, and the imperative, outside the second person, and the future indicative are expressed by periphrasis. Even the infinitive and the participle in the koinh, have felt the inroads of the subjunctive.13 It is true that as a rule we see the modes to best advantage in the simple sentence,14 though essentially the meaning in the compound sentence is the same. But it is true, as Gildersleeve15 urges, that "the predominance of parataxis over hypotaxis is a matter of style as well as of period. Hypotaxis holds fast to constructions that parataxis has abandoned. The futural subjunctive abides defiantly in the dependent clause of temporal sentences and dares the future indicative to invade its domain. The modal nature of the future, obscured in the principal sentence, forces itself upon the most superficial observer in the dependent clause." In a broad sense the indicative is the mode of objective statement in contrast with the subjective modes developed from it. But the description needs modification and is only true in a general sense. The N. T. idiom as of the koinh, in general will be found to differ from the classic Greek idiom here more than is true of the construction of the tenses.16 The disappearance of the optative is responsible for part of this change. But the effort must now be made to differentiate the four modes in actual usage whatever may be true of the original idea of each. That point will need discussion also. The vernacular in all languages is fond of parataxis. See Pfister, "Die parataktische Darsteliungsform in der volkstumlichen Erzahlung" (Woch. f. klass. Phil., 1911, pp. 809-813).
A. INDEPENDENT OR PARATACTIC SENTENCES ( PARATAKTIKA vAXIWMATA)
I. The Indicative Mode ( lo,goj avpofantiko,j or h` o`ristikh. e;gklisij).
1. MEANING OF THE INDICATIVE MODE. The name is not distinctive, since all the modes "indicate." It is not true that the indicative gives "absolute reality,"17 though it
MODE ( EGKLISIS) 915
is the "modus rectus." It does express "l'affirmation pure et simple."18 The indicative does state a thing as true, but does not guarantee the reality of the thing. In the nature of the case only the statement is under discussion. A clear grip on this point will help one all along. The indicative has nothing to do with reality ("an sich").19 The speaker presents something as true.20 Actuality is implied, to be sure, but nothing more.21 Whether it is true or no is another matter. Most untruths are told in the indicative mode. The true translation into Latin of o`ristikh, would be finitus or definitus.22 Indicativus is a translation of avposfantiko,j. The indicative is the most frequent mode in all languages. It is the normal mode to use when there is no special reason for employing another mode. The assertion may be qualified or unqualified.23 This fact does not affect the function of the indicative mode to make a definite, positive assertion. Cf. Jo. 13:8, for instance. A fine study of the indicative mode is afforded in Jo. 1:1-18, where we have it 38 times, chiefly in independent sentences. The subjunctive occurs only three times times(1:7 f.). The use of
2. KINDS OF SENTENCES USING THE INDICATIVE.
(a) Either Declarative or Interrogative. The mere declaration probably (and logically) precedes in use the question.24 But there is no essential difference in the significance of the mode. This extension of the indicative from simple assertion to question is true of all Indo-Germanic tongues.25 Cf. Mt. 2:2; Mk. 4:7; Jo. 1:19. The simple assertion is easily turned to question. Cf. evpei,nasa ga.r evdw,kate, maoi fagei/nà evdi,yhsa kai. evpoti,sate, meà ktl, and po,te se ei=domen peinw/nta kai. evqre,yamenà ktl) (Mt. 25:35-39). For the change fromquestion to simple assertion see pisteu,eij tou/to; evgw/ pepi,steuka (Jo. 11:26 f.). Cf. Ac. 26:27. The formula su. le,geij is sometimes used for the answer, as in Mt. 27:11; Lu. 22:70; Jo.18:37. So also su. ei=paj in Mt. 26:25, 64. The question without interrogative words is seen in Mt. 13:28; Jo. 13:6; Ac. 21:37; Ro. 2:21-23; 7:7, etc. Sometimes it is diffi-
916 A GRAMMAR OF THE GREEK NEW TESTAMENT
cult to tell whether a sentence is declarative or interrogative, as in 1 Cor. 1:13; Ro. 8:33 f.
For this very reason the Greek used various interrogatory particles to make plain the question. Thus a=ra, ge ginw,skeij aa} avnaÄ ginw,skeij; (Ac. 8:30. Note the play on the verb). Cf. Lu. 18:8; Gal. 2:17. It is rare also in the LXX (cf. Gen. 18:9; 37:10; Jer. 4:10), but a;ra is common.26 It is a slight literary touch in Luke and Paul. The use of eiv in a question is elliptical. It is really a condition with the conclusion not expressed or it is an indirect question (cf. Mk. 15:44; Lu. 23:6; Ph. 3:12). It is used in the N. T., as in the LXX quite often (Gen. 17:17, etc.). This construction with a direct question is unclassical and may be due to the Septuagint rendering of the Hebrew h. by eiv as well as by mh,.27 Cf. Mt. 12:10, Eiv e;xestin toi/j sa,bbasin qerapeu/sai; see also Mt. 19:3; Mk. 8:23; Lu. 13:23; 22:49; Ac. 1:6; 7:1; 19: 2; 21:37; 22:25. Note frequency in Luke. In Mk. 10:2 (parallel to Mt. 19:3) the question is indirect. The idiom, though singular, has "attained to all the rights of a direct interrogative"28 by this time. The idiom may be illustrated by the Latin an which in later writers was used in direct questions. So si, used in the Vulgate to translate this eiv, became in late Latin a direct interrogative particle. A similar ellipsis appears in the use of eiv (cf. Heb. 3:11) in the negative sense of a strong oath (from the LXX also).29 The particle h= is found in the LXX Job 25:5 B, but not in the N. T.30 So far the questions are colourless.
The use of interrogative pronouns and adverbs is, of course, abundant in the N. T. Thus ti,j, either alone as in Mt. 3:7, with a;ra as in Mt. 24:45, with ga,r as in Mt. 9:5, with all as in Lu. 3:10.31 See the double interrogative ti,j ti, in Mk. 15:24. For ti, tou/to (predicative use of tou/to) see Lu. 16:2. For the ellipsis with i[na ti, (cf. dia. ti, in Mt. 9:11; eivj ti, in Mk. 14:4) see Mt. 9:4, and for ti, o[ti note Lu. 2:49 (cf. ti, ge,gonen o[ti in Jo. 14: 22). The use of ti, in Ac. 12:18 and 13:25 is interesting. Ti, is an accusative adverb in Mk. 10:18. A sort of prolepsis or double accusative occurs in oi=da se. ti,j ei= (Mk. 1:24). Other pronouns used in direct questions are poi/oj (Mk. 11:28), po,soj (Mk. 6:
MODE ( EGKLISIS) 917
38), potapo,j (Mt. 8:27). The sense of o [ in Mt. 26:50 is disputed, as of o[ti in Mk. 2:16; 9:11, 28; Jo. 8:25.32 The use of interrogative adverbs is frequent. Cf. po,te (Mt. 25:38); e[wj po,te (Mt. 1:17); pw/j (Lu. 10:26); pou/ (Lu. 8:25); posa,kij (Mt. 18:21).
Alternative questions are expressed by h; alone as in 1 Cor. 9: 8, or with ti, - h; as in Mt. 9:5. The case of h' ti,j is different (Mt. 7:9).
Exclamations are sometimes expressed by the relative forms, like w`j w`rai/oi in Ro. 10:15, but more frequently by the interrogative proouns like po,sa (Mk. 15:4); phli,koj (Gal. 6:11); ti, (Lu. 12:49); posa,kij (Mt. 23:37). Cf. po,son in Mt. 6:23.
(b) Positive and Negative. If an affirmative or negative answer is expedited, then that fact is shown by the use of ouv for the question expecting the affirmative reply and by mh, for the negative answer. As a matter of fact, any answer may be actually given. It is only the expectation that is presented by ouv or mh,. This use of ouv is like the Latin nonne. So ouv tw|/ sw|/ ovno,mati evprofhteu,samen * (Mt. 7:22). Cf. Mt. 6:25; 13:27; 13:55; Lu. 12:6; 15:8; 17:17; 1 Cor. 9:1; 14:23; Jas. 2:5; Heb. 3: 16, etc. This lis the common classic construction. The use of ouv may suggest indignation as in ouvk avpokri,nh| ouvde,n; (Mk. 14:60. Cf. ouvk avpekri,nato ouvde,n in verse 61). So with ouv pau,sh| diastre,fwn; (Ac. 13:10). Surprise is indicated by ouvk a;ra in Ac. 21:38. Ouvci, is common. Cf. Lu. 6:39. Ouvkou/n occurs once in the N. T. (Jo. 18:37). The presence of mh, shows that the answer "no" is anticipated (the only instance of mh, with the indicative in a principal sentence). Gildersleeve33 calls ouv "the masculine negative" and mh, "the feminine negative." There is certainly a feminine touch in the use of mh, by the woman at Jacob's well when she came to the village. She refused to arouse opposition by using ouv and excited their curiosity by mh,. Thus mh,ti ou-to,j evstin o` Cristo,j; (Jo. 4:9).34 The examples in the N. T. are very numerous. The shades of negative expectation and surprise vary very greatly. Each context supplies a slightly different tone. Cf. Mt. 7:9, 16; 1:23; 26:22, 25; Mk. 4:21; Lu. 6:39; Jo. 6: 67; 7:26, 35, 47, 51 f.; 21: 5; Ro. 9:14; 11:1. Both ouv and mh, may occur in contrast in the same sentence. So mh. kata. a;nqrwpon tau/ta lalw/Ã h' kai. o` no,moj tau/ta ouv le,gei; (1 Cor. 9:8). Cf. Lu. 6: 39 mh,ti du,natai tuflo.j tuflo.n o`dhgei/n * ouvci. avmfo,teroi eivj bo,qunon evmpeÄ
918 A GRAMMAR OF THE GREEK NEW TESTAMENT sou/ntai; The use of mh,ti is common (cf. ouvci).35 The combination mh. ouv will be discussed in the chapter on Particles, but it may be noted here that ouv is the negative of the verb while mh, is the interrogative particle expecting the answer "no." The English translation expects the answer "yes," because it ignores mh, and translates only ouv. Cf. 1 Cor. 9:4, 5; 11:22; Ro. 10:18, 19. The construction is in the LXX (Judg. 6:13, etc.) and in classic Greek. It is a rhetorical question, not a simple interrogative.36 The kinds of sentences overlap inevitably so that we have already transgressed into the territory of the next group. As already shown, the indicative is used indifferently with or without the negative in either declarative or interrogative sentences. The groups thus overlap. Cf., for instance, Jo. 1:2-8. The negative of a declarative independent sentence with the indicative is ouv. This outright "masculine" negative suits the indicative. With questions, however, it is different, as has already been shown. Thus it is true that mh, made a "raid" into the indicative, as ouv did in the early language into the subjunctive.37 The optative uses either ouv or mh,, but that is another story. The indicative with ouv makes a pointed denial. Note the progressive abruptness of the Baptist's three denials in Jo. 1:20 f. 3. SPECIAL USES OF THE INDICATIVE. (a) Past Tenses. ( a) For Courtesy. It is true that the indicative "is suited by its whole character only to positive and negative statements, and not to the expression of contingencies, wishes, commands or other subjective conceptions."38 That is perfectly true. The indicative is the normal mode for saying a thing. The other modes Gildersleeve39 aptly terms "side moods." I consider, as already explained, the indicative the mode par excellence, and I doubt the value of such language as "the modal uses of the indicative."40 It is not so much that the indicative "encroached upon the other moods, and in so doing assumed their functions, especially in dependent sentences,"41 as that the indicative, particularly in dependent sentences, retained to some extent all the functions of all the modes. It is true, as already said, that the indicative was
MODE ( EGKLISIS) 919 always the most virile of all the modes and has outlived them all. But, after the other modes became fully developed, these less frequent uses of the indicative seemed anomalous. The courteous or polite use of the imperfect indicative is the simplest of these special constructions. Here the indicative is used for direct assertion, but the statement is thrown into a past tense, though the present time is contemplated. We do this in English when we say: "I was just thinking," "I was on the point of saying," etc. So Ac. 25: 22, evboulo,mhn kai. auvto.j tou/ avnqrw,pou avkou/sai. Agrippa does not bluntly say bou,lomai (cf. Paul in 1 Tim. 2:8; 5:14) nor evboulo,mhn a;n, which would suggest unreality, a thing not true. He does wish. He could have said bouloi,mhn a;n, (cf. Ac. 26:29, where Paul uses the optative), but the simple evboulo,mhn is better. The optative would have been much weaker.42 In 2 Cor. 1:15 evboulo,mhn pro,Ä teron has its natural reference to past time. Cf. evboulh,qhn in 2 Jo. 1:12 and Phil. 1:13, evboulo,mhn, not 'would have liked' as Blass (Gr. of N. T. Gk., p. 207) has it. In Gal. 4:20, h;qelon de. parei/nai pro.j u`ma/j a;rti, Paul is speaking of present time (cf. o[ti avporou/mai). He puts the statement in the imperfect as a polite idiom. The use of qe,lw is seen in Ro. 16:19. The usual force of the mode and tense appears in h;qelon in Jo. 6:21. The negative brings out sharply the element of will (cf. Lu. 19:14; Mt. 22:3). In Ro. 9:3, huvco,mhn ga.r avna,qema ei=nai autvo.j evgw. avpo. tou/ Cristou/, the same courteous even passionate) idiom occurs. It is not eu;comai as in 2 Cor. 13:7 (he does not dare pray such a prayer), nor did he do it (cf. hu;conto Ac. 27:29). He was, however, on the verge of doing it, bit drew back. With this example we come close to the use of the indicative for unreality, the so-called "unreal" indicative. See also chapter on Tense. ( b) Present Necessity, Obligation, Possibility, Propriety in Tenses of he Past. This is the usual "potential" indicative. The imper ect of such verbs does not necessarily refer to the present.43 Thus in Jo. 4:4, e;dei auvto.n die,rcesqai dia. th /j Samari,aj, it is simply a necessity in past time about a past event. So dei/ in Jo. 4:20, 24 expresses a present necessity. This use of the imperfect e;dei thus differs from either the present or the ordinary imperfect. The idiom is logical enough.44 It was a necessity and the statement may be confined to that phase of the matter, though the necessity still exists. So Lu. 24:26, ouvci. tau/ta e;dei paqei/n to.n Cristo,n; Cf. also Mt. 18:33; 23:23; 25:27; Lu. 11:42; 13:16 (cf.
920 A GRAMMAR OF THE GREEK NEW TESTAMENT dei/ in verse 14); Ac. 27:21. It is an easy step from this notion to that of an obligation which comes over from the past and is not lived up to. The present non-fulfilment of the obligation is left to the inference of the reader or hearer. It is not formally stated. It happens that in the N. T. it is only in the subordinate clauses that the further development of this use of e;dei comes, when only the present time is referred to. Thus in Ac. 24:19, oua}j e;dei evpi. sou/ parei/nai. They ought to be here, but they are not. Our English "ought" is likewise a past form about the present as well as about the past.45 So 2 Cor. 2:3, avf v w-n e;dei me cai,rein. In Heb. 9:26, evpei. e;dei auvto.n polla,kij paqei/n, there is an implied condition and e;dei is practically an apodosis of the second-class condition, which see. The same process is seen in the other words. Thus in 2 Cor. 12:11, evgw. w;feilon u`f v u`mw/n suni,stasqai, we have a simple past obligation. So in Lu. 7:41; Heb. 2:17. Note common use of the present tense also, as in Ac. 17:29. Cf. oa} wvfei,lomen poih/sai pepoih,kamen (Lu. 17:10), where the obligation comes on from the past. But in 1 Cor. 5:10, evpei. wvfei,lete a;ra evk tou/ ko,smou evxelqei/n, we have merely present time under consideration and a practical apodosis of a second-class condition implied. I do not agree with Moulton46 that a;n, in such instances has been "dropped." It simply was not needed to suggest the unreality or non-realization of the obligation. The context made it clear enough. Crh, occurs only once in the N. T. (Jas. 3:10), whereas prosh,kei (Attic) is not found at all, nor e;xesti (but evxo,n) nor evxh/n.47 But evdu,nato is used of the present time. So Jo. 11:37. Cf. the apodosis in the second-class condition without to in Jo. 9:33; Ac. 26:32. The use of w`j avnh/ken (Col. 3:18) and aa} ouvk avnh/ken (Eph. 5:4) are both pertinent, though in subordinate clauses. Note in particular ouv ga.r kaqh/ken auvto.n zh/n (Ac. 22:22), 'He is not fit to live.' In Mt. 26:24, kalo.n h=n auvtw|/ eiv ouvk evgennh,qh, we have the apodosis without a;n of a condition of the second class (determined as unfulfilled). There is no condition expressed in 2 Pet. 2:21, krei/tton ga.r h=n auvtoi/j mh. evpegnwke,nai th.n o`do.n th/j dikaiosu,nhj. Moulton48 finds the origin of this idiom in the conditional sentence, but Winer49 sees in it merely the Greek way of affirming what was necessary, possible or appropriate in itself. So Gildersleeve.50 The modern Greek preserves this idiom (Thumb,
MODE ( EGKLISIS) 921 Handb., p. 128). The use of e;mellon in Rev. 3:2 approaches this potential indicative. Cf. Thompson, Syntax, p. 274. For the use of the infinitive rather than the indicative see h' - pesei/n, in Lu. 16: 17. So also i[na and subjunctive as in Jo. 6:7. Cf. Viteau, Le Verbe, p. 21. The use of ovli,gou or mikrou/ with an aorist does not occur in the N. T. Cf. Jannaris, Hist. Gk. Gr., p. 445. ( g) The Apodosis of Conditions of the Second Class. This matter has already been touched on slightly and is treated at length under Conditional Sentences. It can be merely sketched here. The condition is not always expressed and a;n usually is present. The use of a;n, however, in the apodosis is not obligatory.51 We know very little about the origin and meaning of a;n anyhow. It seems to have a demonstrative sense (definite, then, in that case) which was shifted to an indefinite use. Cf. to.n kai. to,n ta. kai. ta,.52 Gildersleeve interprets it as a particle "used to colour the moods of the Greek language." With the past tenses of the indicative in independent sentences it is a definite particle. The effort to express unreality by the indicative was a somewhat difficult process. In Homer "the unreal imperfect indicative always refers to the past."53 So in Heb. 11:15. Nothing but the context can show whether these past tenses are used in opposition to the past or the present. The koinh, received this idiom of the unreal indicative "from the earlier age as a fully grown and normal usage, which it proceeded to limit in various directions."54 In Jo. 15:22 we have a good illustration of this construction. We know that a`marti,an ouvk ei;cosan is in opposition to the present reality because it is followed by nu/n de. pro,fasin ouvk e;cousin. The same thing is seen in verse 24 when nu/n de. e`wra,kasin follows. In verse 19 a'n evfi,lei is used, the usual construction. In Lu. 17:6 evle,gete a;n and u`ph,kousen a;n are used after the protasis eiv e;cete (firstclass condition). This is a mixed condition. So also the marginal reading in W. H. in Jo. 8:39 is evpoiei/te after eiv evste, and is followed by nu/n de. zhtei/te (cf. above). The absence of to seems more noticeable in John's Gospel. Cf. Jo. 19:11, ouvk ei=cej evxousi,an kat v evmou / ouvdemi,an eiv mh. h=n dedome,non soi a;nwqen.55 Paul has the same56 idiom. Thus Gal. 4:15 eiv dunato.n tou.j ovfqalmou.j u`mw/n evxoru,xantej evdw,kate, moi and Ro. 7:7 th.n a`marti,an ouvk e;gnwn eiv mh. dia. no,mouà th,n te 922 A GRAMMAR OF THE GREEK NEW TESTAMENT ga.r evpiqumi,an ouvk h|;dein eiv mh. o` no,moj. The MSS. vary in the support of a;n as in Gal. 4:15, where EKLP (and ac Dc) have it. In Jo. 18:36, B does not have a;n, while in 8:19, D does not have it, and the other MSS. differ in the position of a;n.57 This particle comes near the beginning of the clause, though not at the beginning. It does not precede ouvk (cf. Gal. 1:10). It is sometimes repeated in successive apodoses (cf. Jo. 4:10), but not always (cf. Lu. 12:39). Cf. Kuhner-Gerth, Bd. I, p. 247. On the use of a;n in general see Thompson, Syntax, pp. 291 ff. Hoogeveer: (Doctrina Partic. Linguae Graecae, ed. sec., 1806, p. 35) makes a;n mean simply debeo, a very doubtful interpretation. "The addition of a;n to an indicative apodosis produced much the same effect as we can express in writing by italicizing 'if.'"58 This emphasis suggests that the condition was not realized. The papyri likewise occasionally show the absence of a;n.59 The condition is not always expressed. It may be definitely implied in the context or left to inference. So kavgw. evlqw.n su.n to,kw| a'n e;praxa auvto, (Lu. 19:23) and kai. evlqw.n evgw. evkomisa,mhn a'n to. evmo.n su.n to,kw| (Mt. 25:27). Here the condition is implied in the context, a construction thoroughly classical. But, in principal clauses, there is no instance of a;n with a past tense of the indicative in a frequentative sense.60 It only survives in relative, comparative or temporal clauses (cf. Mk. 6:56; Ac. 2:45; 4:35; 1 Cor. 12:2; Mk. 3:11; 11:19). So D in Mk. 15:6, oa}n a'n h|vtou/nto. Both the aorist and the imperfect tenses are used thus with all in these subordinate clauses. There was, considerable ambiguity in the use of the past tenses for this "unreal" indicative. No hard and fast rule could be laid down. A past tense of the indicative, in a condition without a;n, naturally meant a simple condition of the first class and described past time (cf. Heb. 12:25). But in certain contexts it was a condition of the second class (as in Jo. 15:22, 24). Even with a;n it is not certain61 whether past or present time is meant. The certain application to present time is probably postHomeric.62 The imperfect might denote63 a past condition, as in Mt. 23:30; 24:43 (Lu. 12:39); Jo. 4:10; 11:21, 32; 1 Jo. 2:
MODE ( EGKLISIS) 923 19; Heb. 11:15, or, as commonly, a present condition (cf. Lu. 7: 39). The aorist would naturally denote past time, as in Mt. 11: 21. The two tenses may come in the same condition and conclusion, as in Jo. 14:28. The past perfect is found in the protasis, as in Mt. 12:7; Jo. 19:11. Once the real past perfect meets us in the conclusion (1 Jo. 2:19). And note a a'n h|;deite in Jo. 14:7. ( d) Impossible Wishes. These impracticable wishes were introduced in Attic by ei;qe or eiv ga,r, which used also w;felon with the infinitive. From this form a particle was developed o;felon (augmentless) which took the place of ei;qe and eiv ga,r. The dropping of the augment is noted in Herodotus (Moulton, Prol., p. 201). As a matter of fact, this unfulfilled wish occurs only three times in the N. T.: once with the aorist about the past, o;felo,n ge evbasiÄ leu,sate (1 Cor. 4:8), and twice with the imperfect about the present (2 Cor. 11:1; Rev. 3:15). ;Ofelon occurs once also with the future (Gal. 5:12). Many of the MSS. (DcEFGKL) read w;felon in 2 Cor. 11:1, and a few do the same in 1 Cor. 4:8. The idiom occurs in the LXX and in the inscriptions. Cf. Schwyzer, Perg., p. 173. The modern Greek expresses such wishes by na, or a;j and imperf. or aorist (Thumb, p. 128). For e;dramon in Gal. 2: 2, of unrealized purpose, see Final Clauses. Radermacher (N. T. Gr., p. 127) quotes o;felon e;meinaj, Achilles Tatius, II, 24, 3, and w;felon evgw. ma/llon evpu,ressonà Epict., Diss., 22, 12. (b) The Present. In Mt. 12:38, dida,skaleà qe,lomen avpo. sou/ shÄ mei/on ivdei/n, the present seems rather abrupt.64 In Jo. 12:21, ku,rieà qe,lomen to.n vIhsou/n ivdei/n, this is felt so strongly that it is translated: 'Sir, we would see Jesus.' See also Jo. 6:67. Cf. evboulo,mhn in Ac. 25:22 and euvxai,mhn a;n, in 26:29. There does not seem to be the same abruptness in qe,lw in 1 Cor. 7:7. Cf. also fei,domai in 7:28. There were probably delicate nuances of meaning which sufficiently softened these words, shadings which now escape us. There is no difficulty about avrkei/ in 2 Cor. 12:9. In a case like u`pa,gw a`lieu,ein (cf. evrco,meqa) in Jo. 21:3, the suggestion or hint is in the fact, not in the statement. The indicative is a definite assertion. The nature of the case supplies the rest. In 1 Cor. 10:22, h' parazhlou/men to.n ku,rion; the indicative notes the fact, while the surprise and indignation come out in the interrogative form. The question in Jo. 11:47, ti, poiou/men; is very striking. It may be questioned65 if the point is the same as ti, poiw/men; (cf. Jo. 6:28), like the Latin Quid faciamus? The subjunctive of de-
924 A GRAMMAR OF THE GREEK NEW TESTAMENT liberation suggests doubt on the whole subject or expresses a wish to do something. Blass66 cites the colloquial Latin for parallels for this idiom. But we do not need such parallels here. The inquiry of Caiaphas is rather indignant protest against the inactivity of the Sanhedrin than a puzzled quandary as to what they should do. The indicative suits exactly his purpose. He charges them with doing nothing and knowing nothing and makes a definite proposal himself. Winer sees the point clearly.67 The same use of qe,lw noted above appears in questions of deliberation as in qe,leij sulle,xwmen; (Mt. 13:28). So bou,lesqe avpolu,sw; (Jo. 18:39). Cf. Lu. 18:41. Possibility or duty may be expressed in questions also, as in pw/j du,nasqe avgaqa. lalei/n ponhroi. o;ntej; (Mt. 12:34); ti, me dei/ poiei/n i[na swqw/; (Ac. 16:30). This is the analytical method rather than trusting to the mode.68 "It is found possible, and more convenient, to show the modal character of a clause by means of particles, or from the drift of the context, without a distinct verbal form."69 (c) The Future. The future indicative "was originally a subjunctive in the main"70 and it has a distinct modal development. This fact comes out in the fact that the future tense of the indicative is a rival of the subjunctive, the optative and the imperative.71 Like the subjunctive and optative the future may be merely futuristic (prospective) or deliberative or volitive. This matter has been discussed at length under Tenses, which see. As an example of the merely futuristic note Mt. 11:28, of the volitive see Lu. 13:9, of the deliberative note Jo. 6:68. II. The Subjunctive Mode ( h` u`potaktikh. e;gklisij). Some of the Greek grammarians called it h` distaktikh,, some h` sumbouleutikh,, some h` u`poqetikh,. But no one of the names is happy, for the mode is not always subordinate, since it is used freely in principal clauses, nor is it the only mode used in subordinate clauses. But the best one is h` distaktikh,. 1. RELATIONS TO OTHER MODES. The development of the modes was gradual and the differentiation was never absolutely distinct. (a) The Aorist Subjunctive and the Future Indicative. These are closely allied in form and sense. It is quite probable that the future indicative is just a variation of the aorist subjunctive. Cf. e;domaià pi,omaià fa,gomai. The subjunctive is always future, in
MODE ( EGKLISIS) 925 subordinate clauses relatively future. Hence the two forms continued side by side in the language. There is a possible distinction. "The subjunctive differs from the future indicative in stating what is thought likely to occur, not positively what will occur."72 But in the beginning (cf. Homer) it was probably not so. Brugmann (Griech. Gr., p. 499) pointedly contends that many so-called future indicatives are just "emancipated short-vowel conjunctives." Cf. Giles, Manual, pp. 446-448; Moulton, Prol., p. 149. (b) The Subjunctive and the Imperative. These are closely allied. Indeed, the first person imperative in Greek, as in Sanskrit,73 is absent in usage and the subjunctive has to be employed instead. There is a possible instance of the subjunctive as imperative in the second person in Sophocles, but the text is uncertain.74 The use of mh, and the aorist subjunctive in prohibitions of the second and third persons is also pertinent. Thus the subjunctive is in close affinity with the imperative. (c) The Subjunctive and the Optative. They are really variations of the same mode. In my Short Grammar of the Greek N. T.75 I have for the sake of clearness grouped them together. I treat them separately here, not because I have changed my view, but in order to give a more exhaustive discussion. The closeness of the connection between the subjunctive and the optative is manifest in the Sanskrit. "Subjunctive and optative run closely parallel with one another in the oldest language in their use in independent clauses, and are hardly distinguishable in dependent."76 In the Sanskrit the subjunctive disappeared before the optative save in the imperatival uses. It is well known that the "Latin subjunctive is syncretistic, and does duty for the Greek conjunctive and optative."77 Delbruck, indeed, insists that the two modes originally had the same form and the same meaning.78 Delbruck's view has carried the bulk of modern opinion. But Giles79 is justified in saying: "The original meaning of these moods and the history of their development is the most difficult of the many vexed questions of comparative syntax." It is true that
926 A GRAMMAR OF THE GREEK NEW TESTAMENT the subjunctive in Greek refers only to the future, while the optative is not bound to any sphere.80 But the optative is usually relatively81 future like our "should," "could," etc. The use of the subjunctive was greater in Homer's time than afterwards. The independent subjunctive in particular was more freely used in Epic than in Attic. In the modern Greek82 the subjunctive has not only displaced the optative, but the future indicative and the infinitive. But even so in modern Greek the subjunctive is relatively reduced and is almost confined to subordinate clauses (Thumb, Handb., pp. 115, 126). The fut. ind. in modern Greek is really qa, ( qana,) and subj. G. Hamilton83 overstates it in saying: "This monarch of the moods, which stands absolute and alone, has all the other moods dependent on it." It is possible that originally these two moods were used indifferently.84 Vandacle85 argues for a radical difference between the two moods, but he does not show what that difference is. There were distinctions developed beyond a doubt in actual use,86 but they are not of a radical nature. The Iranian, Sanskrit and the Greek are the only languages which had both the subjunctive and optative. The Sanskrit dropped the subjunctive and the Greek finally dispensed with the optative as the Latin had done long ago.87 2. ORIGINAL SIGNIFICANCE OF THE SUBJUNCTIVE. Delbruck88 is clear that "will" is the fundamental idea of the subjunctive, while "wish" came to be that of the optative. But this position is sharply challenged to-day. Goodwin89 denies that it is possible "to include under one fundamental idea all the actual uses of any mood in Greek except the imperative." He admits that the only fundamental idea always present in the subjunctive is that of futurity and claims this as the primitive meaning from the idiom of Homer. Brugmann90 denies that a single root-idea of the subjunctive can be found. He cuts the Gordian knot by three uses of the subjunctive (the volitive, the deliberative, the futur-
MODE ( EGKLISIS) 927 istic). W. G. Hale91 identifies the deliberative and futuristic uses as the same. Sonnenschein92 sees no distinction between volitive and deliberative, to which Moulton93 agrees. "The objection to the term 'deliberative,' and to the separation of the first two classes, appears to be well grounded." He adds: "A command may easily be put in the interrogative tone." That is true. It is also true "that the future indicative has carried off not only the futuristic but also the volitive and deliberative subjunctives." But for practical purposes there is wisdom in Brugmann's division. Stahl94 sees the origin of all the subjunctive uses in the notion of will. The future meaning grows out of the volitive. Mutzbauer95 finds the fundamental meaning of the subjunctive to be the attitude of expectation. This was its original idea. All else comes out of that. With this Gildersleeve96 agrees: "The subjunctive mood is the mood of anticipation," except that he draws a sharp distinction between "anticipation" and "expectation." "Anticipation treats the future as if it were present." He thinks that the futuristic subjunctive is a "deadened imperative."97 But Monro98 on the whole thinks that the futuristic meaning is older than the volitive. So the grammarians lead us a merry dance with the subjunctive. Baumlein99 denies that the subjunctive is mere possibility. It aims after actuality, "a tendency towards actuality." At any rate it is clear that we must seek the true meaning of the subjunctive in principal clauses, since subordinate clauses are a later development, though the futuristic idea best survives in the subordinate clause.100 In a sense Hermann's notion is true that three ideas come in the modes (Wirklichkeit, Moglich- keit, Notwendigkeit). The indicative is Wirklichkeit, the imperative is Notwendigkeit, while the subjunctive and the optative are Moglichkeit. I have ventured in my Short Grammar101 to call the subjunctive and optative the modes of doubtful statement, to call,
928 A GRAMMAR OF THE GREEK NEW TESTAMENT while the indicative is the mode of positive assertion and the imperative that of commanding statement. The modes, as already seen, overlap all along the line, but in a general way this outline is correct. The subjunctive in principal sentences appears in both declarative and interrogative sentences. Cf. eivrh,nhn e;cwmen por.j to.n qeo,n (Ro. 5:1), ti, ei;pw u`mi/n; (1 Cor. 11:22). It is found in both positive and negative statements. Cf. dw/men h' mh. dw/men; (Mk. 12:14), mh. sci,swmen auvto,nà avlla. la,cwmen (Jo. 19:24). It is the mood of doubt, of hesitation, of proposal, of prohibition, of anticipation, of expectation, of brooding hope, of imperious will. We shall, then, do best to follow Brugmann. 3. THREEFOLD USAGE. The three uses do exist, whatever their origin or order of development.102 (a) Futuristic. This idiom is seen in Homer with the negative ouv as in ouvde. i;dwmi, 'I never shall see.' It is an emphatic future.103 This emphatic future with the subjunctive is common in Homer with a;n or ken and once without. Gildersleeve104 calls this the "Homeric subjunctive," but it is more than doubtful if the usage was confined to Homer. Moulton (Prol., p. 239) quotes P. Giles as saying: "This like does for many dialects what the subjunctive did for Greek, putting a statement in a polite, inoffensive way, asserting only verisimilitude." Note the presence of the subjunctive in the subordinate clauses with eva,n ( eiv).105 The presence of here and there with the subjunctive testifies to a feeling for the futuristic sense. See h[tij ouv katoikisqh|/ (Jer. 6:8). In the modern Greek, Thumb (Handb., p. 195) gives a' de.n pisteu,h|j, where is for ouvde,n. The practical equivalence of the aorist subjunctive and the future indicative is evident in the subordinate clauses, particularly those with eivà i[naà o[j and o[stij. Cf. oa} prosene,gkh| (Heb. 8:3). This is manifest in the LXX, the N. T., the inscriptions and the late papyri.106 Blass107 pronounces w`j a;nqrwpoj ba,lh| (Mk. 4:26) "quite impossible" against aBDI, A. But Moulton108 quotes ouv teqh|/ from inscriptions 317, 391, 395, 399 al. in Ramsay's Cities and Bishoprics of Phrygia, ii, 392. For the papyri, Moulton (Prol., p. 240) notes B. U; 303 (vi/A.D.) para,scw= 'I will furnish,' A. P. 144 (v/A.D.) e;lqw= 'I will come.' The itacisms in - sh| and - sei prove less, as Moulton notes. The examples in the papyri of itacistic - seià ÄÄsh|, are "innumerable." In Ac. 5:15, W. H.
MODE ( EGKLISIS) 929 print i [na- evpiskia,sei (B, some cursives). Radermacher (N. T. Gr., p. 136) is quite prepared to take pw/j fu,ghte (Mt. 23:33) Êpw/j feu,xesqe. This is probably deliberative, but he makes a better case for evn tw|/ xhrw|/ ti, ge,nhtai (Lu. 23:31). Blass109 notes that "the mixture of the fut. ind. and aorist conj. has, in comparison with the classical language, made considerable progress." He refers to Sophocles, Lexicon, p. 45, where ei;pw soi is quoted as = evrw/ soi)2 In a principal clause in Clem., Hom. XI. 3, we have kai. ou[twj- dunhqh|/, and Blass has noted also in Is. 33:24 avfeqh|/ ga.r auvtoi/j h` a`marti,a. We cannot, indeed, trace the idiom all the way from Homer. "But the root-ideas of the subjunctive changed it remarkably little in the millennium or so separating Homer from the Gospels; and the mood which was more and more winning back its old domain from the future tense may well have come to be used again as a 'gnomic future' without any knowledge of the antiquity of such a usage."110 It was certainly primitive in its simplicity111 even if it was not the most primitive idiom. The use of ouv with the subj. did continue here and there after Homer's day. We find it in the LXX, as in Jer. 6:8 (above) and in the Phrygian inscription (above). In fact, in certain constructions it is common, as in mh. ouv after verbs of fearing and caution. Cf. 2 Cor. 12:20 and MSS. in Mt. 25:9 $mh, pote ouvk avrke,sh|). It is even possible that the idiom ouv mh, is to be thus explained. Gildersleeve112 remarks on this point: "It might even seem easier to make a belong to aivschnqw/, thus combining objective and subjective negatives, but it must be remembered that ouv with the subjunctive had died out (except in mh. ouv) before this construction came in." The vernacular may, however, have preserved ouv with the subj. for quite a while. Jannaris113 confidently connects ouv in this idiom with the subj. and explains mh, as an abbreviation of mh,n. If either of these explanations is true, the N. T. would then preserve in negative principal sentences the purely futuristic subjunctive. Burton114 is clear that anyhow "the aorist subjunctive is used with ouv mh, in the sense of an emphatic future indicative." The ancient Greek sometimes employed the present subjunctive in this sense, but the N. T. does not use it. But the LXX has it, as in Jer. 1:19. So in Is. 11:9 we find ouv mh. kakopoih,sousin ouvde. mh. du,nwntai. The future ind. with ouv mh, is rare in the N. T., but ouv mh, with the aorist
930 A GRAMMAR OF THE GREEK NEW TESTAMENT subj. appears in the W. H. text 100 times.115 It cannot be said that the origin of this a construction has been solved. Goodwin116 states the problem well. The two negatives ought to neutralize each other, being simplex, but they do not (cf. mh. ouv). The examples are partly futuristic and partly prohibitory. Ellipsis is not satisfactory nor complete separation (Gildersleeve) of the two negatives. Perhaps ouv expresses the emphatic denial and mh, the prohibition which come to be blended into the one construction. At any rate it is proper to cite the examples of emphatic denial as instances of the futuristic subjunctive. Thus ouv mh, se avnw /à ouvd v ouv mh, se evgkatali,pw (Heb. 13:5); ouv mh. avpole,sh| (Mk. 9:41); ouvke,ti ouv mh. pi,w (Mk. 14:25). Cf. Lu. 6:37 etc. See ouv mh, in both principal and subordinate clauses in Mk. 13:2. See also Tense. It is a rhetorical question in Lu. 18:7 (note also makroqumei/.) rather than a deliberative one. In Rev. 15:4 we have the aor. subj. and the fut. ind. side by side in a rhetorical question, ti,j ouv mh. fobhqh|/à ku,rieà kai. doxa,sei to. o;noma; See also the ti,j evx u`mw/n e[xei fi,lon kai. poreu,setai pro.j auvto,n- kai. ei;ph| auvtw|/; (Lu. 11:5). It is difficult to see here anything very "deliberative" about ei;ph| as distinct from e[xei) It may be merely the rhetorical use of the futuristic subj. in a question. Have the grammars been correct in explaining all these subjunctives in questions as "deliberative"? Certainly the future incl. is very common in rhetorical and other questions in the N. T. (b) Volitive. There is no doubt about the presence of the volitive subjunctive in the N. T. The personal equation undoubtedly cuts some figure in the shades of meaning in the moods, here as elsewhere.117 Gildersleeve118 would indeed make this "imperative sense" the only meaning of the mood in the standard language after Homer. He does this because the deliberative subjunctive expects an imperative answer. But, as already seen, that is a mooted question. Brugmann119 takes pains to remark that the element of "will" in the volitive subjunctive belongs to the speaker, not to the one addressed. It is purely a matter of the context. It occurs in both positive and negative sentences and the negative is always mh,. The usage is common in Homer.120 Monro interprets it as expressing "what the speaker resolves or in-
MODE ( EGKLISIS) 931 sists upon." In principle the hortatory subjunctive is the same as the prohibitive use with It was a necessity for the first person, since the imperative was deficient there. Moulton121 ventures to treat this hortatory use of the first person subj. under the imperative, since the Sanskrit grammars give the Vedic subjunctive of the first person as an ordinary part of the imperative. The other persons of the Sanskrit subj. are obsolete in the epic period. Thus bharama, bharata, bharantu are compared with fe,rwmen fe,reteà fero,ntwn (Attic for koinh, fere,twsan). Moulton122 appeals also to the combination of the first and second persons in constructions like evgei,resqe a;gwmen, (Mk. 14:42). This example illustrates well the volitive idea in a;gwmen.123 The first person is usually found in this construction. Cf. also a;gwmen (Jo. 11:7); fa,gwmen kai. pi,wmen (1 Cor. 15:32); e;cwmen (Ro. 5:1, correct text); fronw/men (Ph. 3:15); grhgorw/men kai. nh,fwmen (1 Th. 5:6). Cf. Lu. 9:33 in particular (infinitive and subj.). In 1 Cor. 5: 8, w[ste e`orta,zwmen, the subjunctive is hortatory and w[ste is an inferential particle. Cf. further Heb. 12:1; 1 Jo. 4:7. As examples with mh, see mh. sci,swmen (Jo. 19:24); 932 A GRAMMAR OF THE GREEK NEW TESTAMENT 4) = Lu. 6:42 and deu /ro avpostei,lw (Ac. 7:34, LXX). Moulton128 cites a;fej evgw. auvth.n qrnhn,sw from 0. P. 413 (Roman period). We do not have to suppose the ellipsis of i[naà for a;fej is here the auxiliary. In Jo. 12:7, a;fej auvth.n i[na thrh,sh|, it is hardly probable that a;fej is just auxiliary,129 though in the modern Greek, as already stated, as is used with the third person. In the second person we have only the negative construction in prohibitions with the aorist subjunctive, a very old idiom130 (see Tenses, Aorist). "The future and the imperative between them carried off the old jussive use of the subjunctive in positive commands of 2d and 3d person. The old rule which in ('Anglicistic') Latin made sileas an entirely grammatical retort discourteous to the Public Orator's sileam? - which in the dialect of Elis" (to go on with Moulton's rather long sentence) "produced such phrases as evpime,leian poih,atai Niko,dromor- 'let Nicodromus attend to it,' has no place in classical or later Greek, unless in Soph., Phil., 300 (see Jebb). Add doubtfully Ll. P. 1, vs. 8 (iii/B.c.), Tb. P. 414 131ff. (ii/A.D.)." See Moulton, Prol., p. 178. In the LXX, Jer. 18:8, note kai. evpistrafh|/, parallel with avpostraÄ fh,tw in 18:11. In the modern Greek we have wishes for the future in the subj., since the opt. is dead. So o` qeo.j fula,xh|, God forbid' (Thumb, Handb., p. 127). Radermacher (N. T. Gr., p. 135) finds the subj. for wish in late papyri and inscriptions. It is even in the LXX, Ruth 1:9 A, dw|/ ku,rioj u`mi/n kai. eu[rhte avna,pausinà but B has optative. In the Veda the prohibitive ma is found only with the conjunctive, thus seeming to show that the imperative was originally used only in positive sentences. This idiom of mh, and the aorist subj. held its own steadily in the second person. This point has been discussed at some length under Tenses. Take as illustrations the following: mh. fobhqh|/j (Mt. 1: 20); mh. nomi,shtegrk grk(5:17); mh. eivsene,gkh|jgrk grk(6:13). The use of o[ra and o`ra/te with mh, and the aorist subj. is to be noted. Some of these are examples of asyndeton just like a;fej. Thus o[ra mhdeni, mhde.n ei;ph|j (Mk. 1:44; cf. Mt. 8:4). So also o[ra mh, (Rev. 22:9) where the verb poih,sh|j is not expressed. Cf. LXX o[ra poih,seij (Heb. 8: 5) o`ra/te mhdei.j ginwske,tw (Mt. 9:30) and o`ra/te mh. qroei/sqegrk grk(24:6). With ble,pete it is not always clear whether we have asyndeton (parataxis) or a subordinate clause (hypotaxis). In Lu. 21:8, (Anderson-Cumont-Gregoire). Radermacher (N. T. Gk., p.128) cites also sunt mhqei,hsan kai. ge,nwntai, Acta Thomae, p. 129.
MODE ( EGKLISIS) 933 ble,pete mh, planhqh/te, we may (p. 996) have parataxis as is possible132 in Heb. 12:25, ble,pete mh. paraith,shsqe. Cf. Ac. 13:40; Gal. 5:15. These forms occur with the third person also, as ble,pete mh, tij u`ma/j plan,ash| (Mt. 24:4). But, per contra, see 1 Cor. 10:12 ( mh, e;stai in Col. 2:8). In 1 Th. 5:15, o`ra/te mh, tij kako.n avnti. kakou / tini. avpodw|/, parataxis is probable. But the third person aorist subj. occurs with mh, alone as in mh, tij ou=n auvto.n evxouqenh,sh| (1 Cor. 16:11); mh, ti,j me do,xh| a;frona ei=nai (2 Cor. 11:16); mh, tij u`ma/j evxaÄ path,sh| (2 Th. 2:3). Elsewhere mh, and the aorist imperative occur in the third person. Radermacher (N. T. Gr., p. 134) quotes mh, and 3d person aor. subj. from koinh, writers, inscr. and papyri. Careless writers even use mh. ou=n a;llwj poih|/j, B. G. U. III, 824, 17. Even Epictetus (II, 22, 24) has mh. auvto,qen avpofai,nh|. No less volitive is an example with ouv mh,, like ouv mh. eivse,lqhte (Mt. 5: 20), which is prohibitive. So ouv mh. ni,yh|j (Jo. 13:8); ouv mh. pi,h| (Lu. 1:15). There is the will of God in ivw/ta e[n ha} mi,a kerai,a ouv mh. pare,lqh| (Mt. 5:18) in the third person. In Mt. 25:9, mh, pote ouv mh. avrke,sh| h`mi/n kai. u`mi/n, the subj. is probably futuristic (or deliberative). In a late papyrus, 0. P. 1150, 6 (vi/A.D.), note dei/xon th.n du,nami,n son kai. evxe,lqh| where the 3d pers. subj. imperative like Latin. There are examples in the N. T. where i[na seems to be merely an introductory expletive with the volitive subjunctive. Thus i[na evpiqh|/j (Mk. 5:23); i[na avnable,ywgrk grk(10:51); i[na perisseu,hte (2 Cor. 8:7); i[na mnhmoneu,wmen (Gal. 2:10. Note present tense); i[na fobh/tai, (Eph. 5:33) parallel with avgapa,tw. Cf. i[na- dw,h| ( dw|/% margin of W. H., Eph. 1:17. Moulton133 finds in the papyri (B. U. 48, ii/iii A.D.) eva.n avnabh|/j th|/ e`orth|/ i[na o`mo,se genw,meqa. So also he cites ei[na auvto.n mh. duswph,sh|j, F. P. 112 (99 A.D.), and i[na mhde. tw/n to,kwn ovligwrh,sh|j (Cicero, Att. vi. 5). The modern Greek uses na, and subj. as imperative for both second and third persons (Thumb, Handb., p. 127 f.). Note also mh. i[na avnastatw,sh|j h`ma/j, B. G. U. 1079 (A.D, 41), not i[na mh,. Moulton (Prol., p. 248) quotes Epict., IV, 1, 41, i[na mh. mwro.j h|=Ã avll v i[na ma,qh|. The use of qe,lw i[na (cf. Mk. 6:25; 10:35; Jo. 17:24) preceded this idiom. Moulton134 even suggests that proseu,cesqe i[na mh. e;lqhte eivj peirasmo,n (Mk. 14:38) is as much parataxis as o`ra/te kai. fula,ssesqe (Lu. 12: 15). This "innovation" in the koinh, takes the place of o[pwj and the future ind. Moulton (Prol., p. 177 note) cites o[pwj moi mh. evrei/j, Plato, 337 B, 'don't tell me,' where o[pwj='in which case.' The use of mh, after words of caution and apprehension is probably
934 A GRAMMAR OF THE GREEK NEW TESTAMENT paratactic in origin.135 Moulton136 notes the use of the present subj. with expressions of warning as well as the aorist. Thus in Heb. 12:15, evpiskopou/ntej mh, tij r`i,za pikri,aj evnoclh|/. But this construction borders so closely on subordinate clauses, if not clear over the line, that it will be best discussed there. Subordinate clauses show many examples of the volitive subjunctive (as clauses of design, probably paratactic in origin, Moulton, Prol., p. 185). See di v h-j latreu,wmen (Heb. 12:28). See discussion of Sub. Clauses. (c) Deliberative. There is no great amount of difference between the hortatory (volitive) subjunctive and the deliberative. The volitive is connected with the deliberative in Mk. 6:24 f., ti, aivth,swmai* qe,lw i[na dw|/j. Thus poih,swmen, 'suppose we do it,' and ti, poih,swmen; 'what are we to (must we) do?' do not vary much. The interrogative137 is a quasi-imperative. Gildersleeve138 notes in Plato (rare elsewhere in Attic) a "number of hesitating halfquestions with mh, or mh. ouv and the present subjunctive." It is possible that we have this construction in Mt. 25:9, mh, pote ouv mh. (W. H. marg. just ouv) avrke,sh| h`mi/n kai. u`mi/n. It is but a step to the deliberative question.139 This is either positive or negative, as in Mk. 12:14, dw/men h' mh. dw/men; So also ouv mh, as in Jo. 18: 11, ouv mh. pi,w auvto,; Cf. also Lu. 18:7; Rev. 15:4. The aorist or the present tense occurs as in Lu. 3:10, ti, ou=n poih,swmen* and in Jo. 6:28, ti, poiw/men; so le,gw in Heb. 11:32. Cf. the indicative ti, poiou/men; in Jo. 11:47 and the future ti, ou=n evrou/men* (Ro. 9:14). The question may be rhetorical (cf. Mt. 26:54; Lu. 14:34; Jo. 6:68; Ro. 10:14) or interrogative (cf. Mt. 6: 31; 18:21; Mk. 12:14; Lu. 22:49).140 The kinship between delib. subj. and delib. fut. ind. is seen in Mk. 6:37, avgora,swmen kai. dw,somen; The first person is the one of most frequent occurrence (cf. Ro. 6:1), ti, aivth,swmai (Mk. 6:24). But examples are not wanting for the second and third persons. Thus pw/j fu,ghte avpo. th /j kri,sewj th /j gee,nnhj; (Mt. 23:33); ti, ge,nhtai; (Lu. 23:31). See further Mt. 26:14; Ro. 10:54. It is sometimes uncertain whether we have the subjunctive or the indicative, as in e[teron prosdokw/men; (Mt. 11:3) and evpaine,sw u`ma/j; (1 Con 11:22). But note ti, ei;pw u`mi/n; in the last passage. In Lu. 11:5 we have both
MODE ( EGKLISIS) 935 ti,j e[xei and ei;ph|. So ti, doi/, (Mk. 8:37, ACD dw,sei) may be compared with ti, dw,sei (Mt. 16:26).141 This ambiguity appears in ti, poih,sw; and e;gnw ti, poih,sw in Lu. 16:3 f. The deliberative subj. is retained in indirect questions. Cf. Mt. 6:31 with Mt. 6:25. The kinship between the deliberative subj. in indirect questions and the imperative and the volitive subjunctive is seen in Lu. 12:4 f., mh. fobhqh/te- u`podei,xw de. u`mi/n ti,na fobhqeh/te\ fobh,qhe ktl) The deliberative subj., like the volitive, has various introductory words which make asyndeton (parataxis). These become set phrases like a;fesjà o[ra. Thus pou / qe,leij e`toima,swmen; (Mt. 26:17), qe,leij ei;pwmen; (Lu. 9:54). In Lu. 18:41 we have ti, soi qe,leij poih,sw; and i[na avnable,yw as the reply, using i[na in the brief answer. Cf. further Mt. 13:28. In Jo. 18:39, bou,lesqe ou=n avpolu,sw, we probably have the subj. also. Some MSS. have eiv pata,xwmen; in Lu. 22:49.142 We may leave further discussion of the subj. to the subordinate clauses. We have no examples in the N. T. of a;n with the subj. in independent sentences (but see ke, and the subj. in Homer). In subordinate clauses a;n is very common, though not necessary, as will be seen.143 (Cf. discussion of eivà o[stij.) But Jannaris144 gives instances of a;n with the subj. in principal clauses (futuristic) in Polybius, Philo, Plutarch, Galen, etc. With the disappearance of the fut. ind., the opt. and the imper., the subj. has the field as the "prospective mood." It is found in the modern Greek as in ti, na. gi,nh| (Thumb, Handb., p. 126). III. The Optative Mode ( h` euvktikh. e;gklisij). It has already been shown that the optative does not differ radically from the subjunctive. Jannaris145 calls the optative the "secondary subjunctive." 1. HISTORY OF THE OPTATIVE. For the facts see chapter on Conjugation of the Verb. It is an interesting history and is well outlined by Jannaris146 in his Appendix V, "The Moods Chiefly Since A. (Ancient Greek) Times." It retreated first from dependent clauses and held on longest in the use for wish in independent sentences like ge,noito. But even here it finally went down before the fut. ind. and subj. The optative was a luxury
936 A GRAMMAR OF THE GREEK NEW TESTAMENT of the language and was probably never common in the vernacular. Certainly it is very rare in the vernacular koinh, (both inscriptions and papyri). It is a literary mood that faded before the march of the subj. In a hundred pages of the Memorabilia of Xenophon the optative occurs 350 times. He had a "hyperorthodox love of the mood."147 Plato's Phaedo shows it 250 times in a corresponding space, but Strabo has it only 76, Polybius 37, Diodorus Siculus 13 times in a hundred pages.148 The 67 examples in the N. T. are in harmony with the koinh, usage. Gildersleeve pithily says: "The optative, which starts life as a wish of the speaker, becomes a notion of the speaker, then a notion of somebody else, and finally a gnomon of obliquity" (A. J. of Phil., 1908, p. 264). In the LXX the optative is rare, but not so rare as in the N. T., though even in the LXX it is replaced by the subj. (Thackeray, Gr., p. 193) as in the late papyri and inscriptions (Radermacher, N. T. Gr., pp. 128, 135). 2. SIGNIFICANCE. There is no definite distinction between the subjunctive and the optative in the Sanskrit.149 The Latin put all the burden on the subj., as the Greek finally did. The Sanskrit finally made the optative do most of the work. In a word, the optative is a sort of weaker subjunctive.150 Some writers make the opt. timeless and used definitely of the past.151 It is rather a"softened future"152 sometimes flung back into the past for a Standpunkt. We do not153 know "whether the opt. originally expressed wish or supposition." The name does not signify anything. It "was invented by grammarians long after the usages of the language were settled."154 They just gave it the name euvtikh, because at that time the only use it had without a;n was that of wishing. The name is no proof that wishing was the primitive or the only function or the real meaning of the mode. We have precisely the same difficulty as in the subjunctive. Indeed, the
MODE ( EGKLISIS) 937 optative has three values, just like the subjunctive, viz. the futuristic (potential), the volitive (wishes) and the deliberative.155 In the first and third kinds a;n is usually present, but not always. Brugmann156 notes only two, omitting the deliberative as some scholars do for the subj. He does reckon a third use in indirect discourse, but this is merely the opt. in subordinate sentences and may be either of the three normal usages. The rare fut. opt. in indirect discourse illustrates the point (not in the N. T.). There is no doubt of the distinction between the futuristic (potential) with negative ouv (cf. futuristic subj. in Homer) and the volitive use with mh, (cf. subj. again).157 But there was also a "neutral sense" that can hardly be classed either as futuristic or volitive.158 Gildersleeve159 calls this the "optative in questions," usually with a;n. This is the deliberative use. 3. THE THREE USES. (a) Futuristic or Potential. We begin with this whether it is the first in time or not. Delbruck160 has taken several positions on this point. The use of the negative ouv here shows its kinship with the future (cf. fut. ind. and aorist subj. in Homer).161 The a;n was not always present in Homer and it is not the a;n that gives the potential idea to the mode. In poetry the use without a;n continued. "The optative is the ideal mood of the Greek language, the mood of the fancy."162 Moulton163 puts it clearly: "It was used to express a future in a milder form, and to express a request in deferential style." Radermacher cites from Epictetus, II, 23, 1, a'n h[dion avnagnw|,h - a;n tij r` /on avkou,sei, showing clearly that the opt. and the fut. ind. are somewhat parallel. Moulton (Prol., p. 194) cites Deut. 28:24 ff., where the opt. and fut. ind. alternate in translating the same Hebrew. I do not agree with Radermacher (N. T. Gr., p. 128) in seeing in h;qelon p`arei/nai (Gal. 4:20) a mere equivalent of qe,loimi a;n. See imperfect ind. The presence of a;n gives "a contingent meaning"164 to the verb and makes one think of the unexpressed protasis of the fourth-class condition. The
938 A GRAMMAR OF THE GREEK NEW TESTAMENT idiom has vanished as a living form from the vernacular koinh, in the N. T. times.165 It appears only in Luke's writings in the N. T. and is an evident literary touch. The LXX shows it only 19 times outside of 4 Maccabees and 30 with it.166 Moulton167 notes one papyrus which does not have ay (cf. Homer), though he would MODE ( EGKLISIS) 939 ment is undoubtedly present; see (c). The same thing is true of Lu. 3:15 ( mh, pote) Ac. 17:27 ( eiv%, but Ac. 25:16 ( pri.n h; in indirect discourse for subj. of the direct) is futuristic. (b) Volitive. Moulton169 calls this use the "Optative Proper," a curious concession to the mere name. It has been the most persistent construction of the optative, and (in independent clauses) thirty-eight of the sixty-seven examples of the N. T. come under this category.170 Fifteen of the thirty-eight instances belong to mh, ge,noito, once in Lu. 20:16, and the other fourteen in Paul's Epistles (10 in Romans, 1 in 1 Cor., 3 in Gal.). Thumb considers the rare use of mh. ge,noito in modern Greek (the only relic of the optative) a literary phenomenon, but Moulton171 notes that Pallis retains it in Lu. 20:16. Moulton compares the persistence of the English optative in the phrase "be it so," "so be it," "be it never so humble," etc.172 So he notes it in the papyri for oaths, prayers and wishes.173 0. P. 240 (i/A.D.) eu= ei;h, O. P. 715 (ii/A.D.) e;nocoi ei;hmen, 0. P. 526 (ii/A.D.) cai,roij, L. Pb. (ii/B.C.) oa}j didoi,h soi, B. M. 21 (ii/B.C.) soi. de. ge,noito. The N. T. examples are all in the third person except Phil. 1:20, evgw, sou ovnai,mhn. One is a curse mhke,ti mhdei.j fa,goi (Mk. 11:14) and is equivalent to the imperative. "There is a strong inclination to use the imperative instead of the optative, not only in requests, where the imperative has a legitimate place in classical Greek as well, but also in imprecations, where it takes the place of the classical optative: avna,qema e;stw, Gal. 1:8 f. Cf. 1 Cor. 16:22."174 Only in Mk. 11: 14 and Ac. 8:20, to. avrgu,rio,n sou su.n soi. ei;hà do we have the optative in imprecations in the N. T. The opt. comes very near the imper. in ancient Greek sometimes (Gildersleeve, p. 155). Cf. gi,noito, P. Par. 26 (B.c. 163). In Ac. 1:20, where the LXX (Ps. 109:8) has la,boi, Luke gives labe,tw.175 There are only 23 examples of the volitive optative in independent clauses outside of mh. ge,noito. Paul has 15 of this 23 "(Ro. 15:5, 13; Phil. 1:20; 2 Tim. 1:16, 18; 4:16, and the rest in 1 and 2 Th.), while Mark, Luke, Acts, Hebrews, 1 Peter and 2 Peter have one apiece, and Jude two."176 They are all examples of the aorist optative except the present in Ac. 8:20. The negative is mh, and a;n is not used. In
940 A GRAMMAR OF THE GREEK NEW TESTAMENT 2 Th. 3:16 dw|,h is opt., not the subj. dw,h|. In 1 Th. 3:12 the context shows that perisseu,sai is opt. (not aor. inf. nor aor. middle imperative).177 The rare use of the volitive opt. with eiv (twelve cases in the N. T., but four belong to indirect questions), will be discussed under Conditional Sentences. If i[na dw|,h is the correct text in Eph. 1:17, we probably have a volitive optative, the i[na being merely introductory (cf. examples with the subj.).178 It is hardly a case of final i[na with the optative. Blass179 reads dw|/ here subj. after B. In modern Greek Dr. Rouse finds people saying not mh. ge,noito, but o` qeo.j na. fula,xh| (Moulton, Prol., p. 249), though na, is not here necessary (Thumb, Handb., p. 127). The ancient idiom with ei;qe and eiv ga,r is not found in the N. T., as stated already several times. ;Ofelon with the future ind. occurs for a future wish (Gal. 5:12). (c) Deliberative. There is little more to add here. The LXX180 gives instances of ti,j dw|,h; (Num. 11:29; Judg. 9:29; 2 Sam. 18:33, etc.) without a;n as in Homer, where a deliberative subj. would be admissible. See also Ps. 120 (119):3, ti, doqei,h soi kai. ti, prosteqei,h soi; In Lu. 6:11 Moulton181 remarks that ti, a'n poih,Ä saien in the indirect question is "the hesitating substitute for the direct ti, poih,somen;" Why not rather suppose a "hesitating" (deliberative) direct question like ti, a'n qe,loi o` spermolo,goj ou-toj le,gein; (Ac. 17:18). As already remarked, the context shows doubt and perplexity in the indirect questions which have a'n and the opt. in the N. T. (Lu. 1:62; 6:11; 9:46; 15:26; Ac. 5:24; 10:17). The verbs ( evne,neuonà diela,lounà eivsh/lqen dialogismo,jà evpunÄ qa,netoà dihpo,roun) all show this state of mind. See indirect question eiv bou,loito in Ac. 25:20 after avporou,menoj. Cf. 27:39. The deliberative opt. undoubtedly occurs in Lu. 3:15, dialogizome,nwn mh, pote auvto.j ei;h o` Cristo,j. It is not therefore pressing the optative unduly to find remnants of the deliberative use for it (cf. subj. and fut. indicative).
MODE ( EGKLISIS) 941 IV. The Imperative ( h` prostatikh. e;gklisij%) 1. ORIGIN OF THE IMPERATIVE. See chapter on Conjugation of the Verb for discussion of the various devices used by this latest of the modes in order to get a foothold. Giles,182 after giving the history of the imperative forms (five separate strata), curtly dismisses it as not properly a mode and declines to discuss it under syntax. So Radermacher passes it by in his N. T. Gr. Moulton,183 on the other hand, takes it up "first among the moods" because "it is the simplest possible form of the verb." It is the simplest in one of its forms like the interjectional a;ge, but it is also the latest of the modes and is without a distinct set of endings. Besides, it never dislodged the aorist subj. from the second person in prohibitions and finally gave up the fight all along the line. The modes were slower than the tenses in making sharp distinctions anyhow, and in the Sanskrit "no distinction of meaning has been established between the modes of the present-system and those (in the older language) of the perfect- and aorist-systems."184 The ambiguity of the imperative persists in the second person plural present where only the context can decide the mode. Thus evrauna/te (Jo. 5:39); pisteu,etegrk grk(14:1); avgallia/sqe (1 Pet. 1:6); oivkodomei/sqegrk grk(2:5); telei/te (Ro. 13:6); kaqi,zete (1 Cor. 6:4); cf. Jo. 12:19. The perfect form i;ste (Jas. 1:19; Heb. 12:17) shows the same situation. 2. MEANING OF THE IMPERATIVE. In original significance it was demand185 or exhortation. But, as will be shown, it was not confined to this simple idea. Besides, the notion of command (or prohibition) was expressed in various ways before the imperative was developed. These uses of the other modes continued to exist side by side with the imperative till the N. T. time. Examples of this will be given directly. The imperative itself was extended to include various shades of the future ind., the subj. and the opt. There is a general sense in which the imperative is distinct, as is seen in avgapa/te tou.j evcqrou.j u`mw/n (Mt. 5:44), but this idea of command easily softens to appeal as in ku,rieà sw/sonà avpollu,meqa (Mt. 8:25). 3. DISAPPEARANCE OF THE IMPERATIVE FORMS. It was the last mode to get on its feet. It followed the optative into oblivion save in the second person (Thumb, Handb., p. 154). There the forms held on in the main, but the present subjunctive with came also into use instead of mh, and the present imper., and
942 A GRAMMAR OF THE GREEK NEW TESTAMENT finally the hortatory (positive) subj. also appeared as imper. In the third person (both positive and negative with mh,) a;j and the subj. drove out the imperative. Thus the imperative forms in modern Greek present a wreck, if indeed they were ever much else.186 The imperative, like the subjunctive, is always future in time, though it may apply to the immediate future as in "quit that." 4. ALTERNATIVES FOR THE IMPERATIVE. These, under all the circumstances, can be logically treated before the imperative itself. Indeed, they have already been discussed in the preceding remarks on tense and mode, so that little in addition is required. (a) The Future Indicative. See ch. XVIII, Tense, where it is shown that the Volitive Future is the equivalent of the imperative. The fut. ind., like the subj. and the opt., may be merely futuristic or volitive, or deliberative. The volitive future is a matter of context and tone of voice, to be sure, but that is true also of the subj. and opt., and, in truth, of the real imperative. But more of the "tone of the imperative" further on. English, as well as Greek, continues to use this volitive future. Both positive and negative ( ouv) commands are given by the fut. ind. The negative is sometimes mh, as in mh. boulh,sesqe eivde,nai (Demosthenes), mh. evxe,stai (B. U. 197, i/A.D.), mhde,na mish,sete (Clem., Hom., III, 69).187 So also ouv mh, with the fut. ind. is sometimes prohibition, as in ouv mh. e;stai soi tou/to (Mt. 16:22). Cf. also Gal. 4:30. But it is commonest in the simple future like su. o;yh| (Mt. 27:4); u`mei/j o;yesqe grk(27:24); evkko,yeij (Lu. 13:9); ouvk e;sesqe (Mt. 6:5), etc. It is true that this use of ouv proves the origin of this idiom to be "a purely futuristic form,"188 as is the case with the question ouv pau,sh| diastre,fwn; (Ac. 13:10), but the tone of this future is volitive (imperatival). The Latin use of the volitive future coincides with that of the Greek. Gildersleeve189 says: "It is not a milder or gentler imperative. A prediction may imply resistless power or cold indifference, compulsion or concession." The exact shade of idea in this volitive future must be watched as closely as the imperative itself. Cf. kale,seij (Mt. 1:21) with su. o;yh| (Mt. 27:4). Blass190 denies that this is a "classical" idiom (against
MODE ( EGKLISIS) 943 Gildersleeve) and rather minimizes its use in the N. T. Many of the examples do come from the O. T. (LXX) legal language. Certainly in the LXX the fut. ind. often replaces the imperative under the influence of the Hebrew (Thackeray, Gr., p. 194). But examples occur where the two are equivalent. Cf. avgaph,seij in Mt. 5:43, with avgapa/te in 5:44, evrei/te in Mt. 21:3, with ei;pate in Mk. 11:3. Some MSS. have e;stw rather than gurat in Mt. 20:26. (b) The Subjunctive. The volitive subjunctive is quite to the point. In the first person this use of the subj. held its own always in lieu of the imperative. It is needless to repeat the discussion of this matter (see Subjunctive in this chapter). The use of i[na with the subj. in an imperatival sense is seen Mk. 5:23 (6:25); Eph. 5:33 is there discussed also. Cf. Tit. 2:4. Let mh. sci,swmen auvto,nà avlla. la,cwmen, (Jo. 19:24) serve as an example. So in the second person the aorist subj. held its place in prohibitions past koinh, times to the practical exclusion of the aor. imper. with mh,. The two constructions existed in the koinh, side by side with the third person. Thus mh. gnw,tw (Mt. 6:3) and mh, tij evxouqenh,sh| (1 Cor. 16:11). Cf. do,j and mh. avpostrafh|/j in Mt. 5:42. The final triumph of the subj. over the imperative (save in the second person) has been shown. Cf. the fate of the opt. before the subj. (c) The Optative. There is only one example, mhke,ti mhdei.j fa,goi (Mk. 11:14) in the N. T. The distinction between a curse and a prohibition is not very great. The parallel passage in Mt. 21: 19 has191 ouv mhke,ti evk sou/ karp.j ge,nhtai (volitive subj.). (d) The Infinitive. The idiom is very frequent in Homer.192 It occurs chiefly after an imperative. The command is carried on by the infinitive. There is no need for surprise in this construction, since the probability is that imperative forms like dei/xai (like the Latin legimini, Homeric lege,Ämenai) are infinitive in origin.193 It is true that the accent of the editors for the aorist active optative is different from the aorist active inf. in forms like kateuqu,naià perisÄ seu,sai (1 Th. 3:11 f.), but the MSS. had no accent. We could properly print the infinitive if we wished.194 So as to parakale,sai (2 Th. 2:17) where the accent is the same for both infinitive and optative (the imper. form aor. mid. sec. singl. is paraka,Ä lesai). Cf. ba,ptisai and bapti,sai, one and the same form. The idiom is less frequent in the Attic195 outside of laws and maxims,
944 A GRAMMAR OF THE GREEK NEW TESTAMENT but happens to be the one infinitive construction that is alive in the Pontic dialect to-day.196 Moulton197 expresses surprise at the rarity of this use of the inf. in the N. T., since it is common in the papyri. Cf. evxei/naià misqw/sai, A. P. 86 (1/A.D.). Moulton (Prol., p. 248) notes that Burkitt (Evang. da-Mepharr. ii, 252 f.) reads tau/ta de. poih/sai kavkei/na mh. avfei/nai in Mt. 23:23. Blass198 notes also a revival of the simple inf. or the accusative and infinitive in the later language in legal phraseology. He explains the idiom as an ellipsis, but Moulton is undoubtedly correct in rejecting this theory. There is no need of a verb of command understood in view of the etymology of a form like ba,ptisai. The use of cai,rein as greeting in epistles (with the nominative) is explained in the same way. Cf. Ac. 15 23; 23:26; Jas. 1:1. It is the absolute use of the infinitive as often. It is very common in the papyri, as Polukra,thj tw/i patri. cai,rein, P. Petr. II, xi, 1 (iii/B.C.). So Moulton (Prol., p. 180) denies the necessity of the ellipsis of a verb of command. In Ro. 12:15 cai,rein and klai,ein are clearly parallel with euvlogei/te kai. mh. katara/sqe. So in Ph. 3:16 stoicei/n is to be compared with the hortatory fronw/men. Blass199 needlessly wishes to emend the text in 2 Tim. 2:14, so as not to read mh. logomacei/n. This use of the inf. occurs also in Tit. 2:9. We probably have the same construction in mh. sunanami,gnusqai (2 Th. 3:14), though it may be explained as purpose. In 1 Cor. 5:12 kri,nein is the subject inf. In Lu. 9:3 after ei=pen the quotation begins with Mhde.n ai;rete and is changed to mh,te e;cein (indirect command). In Mk. 6:8 f. both forms are indirect (one with i[na mhde.n ai;rwsin-the other with mh. evndu,sasqai. The marg. in W.H. has mh. evndu,shsqe. The MSS. often vary between the middle inf. and imper. or subj. Winer200 thinks that expositors have been unduly anxious to find this use of the infinitive in the N. T. But it is there. See further chapter XX, Verbal Nouns. (e) The Participle. Winer201 found much difficulty in the absolute use of the participle in the N. T. The so-called genitive absolute is common enough and the participle in indirect discourse representing a finite verb. It would seem but a simple step to use the participle, like the infinitive, in an independent sentence without direct dependence on a verb. Winer admits that Greek prose writers have this construction, though "seldom." He ex-
MODE ( EGKLISIS) 945 plains it on the ground of ellipsis of the copula as is so common with adjectives (cf. Mt. 5:3-11). He passes the poets by (often the truest index of the vernacular) and admits "the Byzantine use of participles simply for finite verbs." T. S. Green202 says: "The absolute use of the participle as an imperative is a marked feature of the language of the N. T." He explains it as an "Aramaism." To this W. F. Moulton203 expresses surprise and admits only "the participial anacoluthon," which, by the way, is very much the same thing. But J. H. Moulton204 has found a number of examples in the papyri where the participle is fairly common for the indicative. The instances in the papyri of the participle in the sense of the imperative are not numerous, but one of them seems very clear. Thus Tb. 59 (i/B.c.) evn oi-j eva.n prosde,hsqe, mou evpita,ssonte,j moi proqumo,teron. It is preceded by a genitive absolute. Moulton gives another equally so: G. 35 (i/B.C.) evpimelo,menoi i[n v u`giai,hte. Moulton205 cites also the Latin form sequimini (= e`po,menoi) for the second middle plural present indicative. The similar looking form sequimini imperative has an infinitive origin, as already shown. See chapter XX, Verbal Nouns, for other examples and further discussion. On the whole, therefore, we must admit that there is no reason per se why the N. T. writers should not use the participle in lieu of the imperative. It is, of course, a loose construction, as ellipsis is and anacoluthon is, but it is not the mark of an uneducated person. In the papyrus example (Tb. 59) given above Grenfell and Hunt call the writer "an official of some importance." Moulton206 also translates Thumb207 concerning the "hanging nominative" (common in classical and koinh, Greek) as saying that the usage "is the precursor of the process which ends in modern Greek with the disappearance of the old participial construction, only an absolute form in - ontaj being left." In the ellipsis of the copula it is not always clear whether the indicative or the imperative is to be supplied. Cf. euvloghto.j o` qeo,j (2 Cor. 1:3). Shall we supply evstin or h;tw ( e;stw) as we have it in 1 Cor. 16:22? In a case like 1 Pet. 3:8 f. it is plain that the unexpressed e;ste would be imperative, but Moulton notes the curious fact that e;ste (imperative) does not appear in the N. T. at all, though we have i;sqi five times, e;stw or h;tw fourteen, and e;stwsan twice.208 There are in- 946 A GRAMMAR OF THE GREEK NEW TESTAMENT stances more or less doubtful, as evpiri,yantej (1 Pet. 5:7), which is naturally taken with tapeinw,qhte as Moultonl now admits. He evidently reacted too strongly against Winer. This use of the participle should not be appealed to if the principal verb is present in the immediate context. Sometimes it is a matter of punctuation as in Lu. 24:47, where W. H. give in the margin avrxa,menoi avpo. vIerousalh.m u`mei/j ma,rturej tou,twn, instead of vIerousalh,m\ u`mei/j. The marginal punctuation takes the participle as an imperative. The MSS. sometimes vary, as when aC give evndei,xasqe in 2 Cor. 8:24, while B, etc., have evndeiknu,menoi.209 But a number of unmistakable examples appear both in Paul and Peter, though "Paul was not so fond of this construction as his brother apostle."'210 Thus e;contej (1 Pet. 2:12) must be so explained or taken as anacoluthon (cf. avpe,cesqai). So u`potasso,menoi (1 Pet. 2: 18; 3:1) reminds one of Eph. 5:22, an "echo" according to Moulton. Other examples occur in 1 Pet. 3:7, 9, possibly 16 also; 4:8 ff. Besides avneco,menoi and spouda,zontej (Eph. 4:2 f.) and u`potasso,menoigrk grk(5:2) in Paul the most outstanding example is in Ro. 12:9 f., 16 f. These participles occur in the midst of imperatives or infinitives as imperatives imperatives(12:15). The asyndeton makes it impossible to connect with any verb. In verse 6 e;contej appears as a practical indicative. Moulton211 adds to these 2 Cor. 9:11 f. and Col. 3:16. See also Heb. 13:5. But Lightfoot212 put in a word of caution when he said: "The absolute participle, being (so far as regards mood) neutral in itself, takes its colour from the general complexion of the sentence." The participle is not technically either indicative, subjunctive, optative or imperative. The context must decide. In itself the participle is non-finite (non-modal) like the infinitive, though it was sometimes drawn out into the modal sphere. 5. USES OF THE IMPERATIVE. (a) Command or Exhortation. In general the imperative keeps within the same limits observed in the classical language, but that is not a narrow groove.213 It is the mood of the assertion of one's will over another or the call of one to exert his will. Thus
MODE ( EGKLISIS) 947 avgapa/te tou.j evcqrou.j u`mw/n (Mt. 5:44); ei;selqe eivj to. tamiei/o,n sou kai. pro,seuxaigrk grk(6:6); pa,ntote cai,rete (1 Th. 5:16). Moulton214 finds the imperatives "normal in royal edicts, in letters to inferiors, and among equals when the tone is urgent, or the writer indisposed to multiply words." The imperatives in Rev. 22:11 are probably hortatory. (b) Prohibition. This is just a negative command and differs in no respect save the presence of the negative mh,. Thus mh. kri,Ä nete (Mt. 7:1), mh. fobei/sqe (Jo. 6:20). Often the presence of the imperative in the midst of indicatives is shown by mh, as in mh. plana/sqe (1 Cor. 6:9). We do, indeed, have a with the imperative in marked contrast, where the force of the negative is given to that rather than to the mode. Thus in 1 Pet. 3:3, e;stw ouvc o` ÄÄko,smojà avll v o` krupto.j th/j kardi,aj a;nqrwpoj. The same explanation applies to ouv mo,non- avlla, kai, in 1 Pet. 2:18, but mh, mo,non is regular in Jas. 1:22, etc., because of gi,nesqe understood. In cases of contrast with ouv- avlla, (with participles and imperatives) the reason for ouv is thus apparent (H. Scott). In Mt. 5:37 ou' ou; (like nai. nai,) is the predicate (like a substantive), not the negative of e;stw. In 2 Tim. 2:14 evp v ouvde.n crh,simon (a parenthetical expression of mh. logomacei/n used as an imperative), the negative goes specifically with the single word crh,simon. Cf. also 1 Cor. 5:10. The upshot is that mh. remains the negative of the imperative. Cf. mh, moi ko,pouj pa,rece (Lu. 11:7). (c) Entreaty. A command easily shades off into petition in certain circumstances. The tone of the demand is softened to pleading.215 Moulton216 notes that the imperative has a decided tone about it. "The grammarian Hermogenes asserted harshness to be a feature of the imperative; and the sophist Protagoras even blamed Homer for addressing the Muse at the beginning of the Iliad with an imperative."217 948 A GRAMMAR OF THE GREEK NEW TESTAMENT true term for instant prayer.'"219 Gildersleeve220 denies that the N. T. shows "the absolute indifference that some scholars have considered to be characteristic of Hellenistic Greek" in the use of the imperative. He credits Mr. Mozley with the observation that "the aorist imperative is regularly used in biblical Greek when the deity is addressed; and following out this generalization Herr Krieckers, a pupil of Thumb's, has made a statistical study of the occurrences of the two tenses in Homer, Hesiod, Sappho, YEschylos, Sophocles, Euripides, Aristophanes, with the result that in prayers addressed by men to men both present and aorist are often used, whereas in prayers addressed by men to gods the aorist largely predominates." Examples221 of the imperative in petitions appear in Mk. 9:22, boh,qhson h`mi/n (Lu. 17:5) pro,sqej h`mi/n pi,stin, (Jo. 17:11) th,rhson auvtou.j evn tw| / ovno,mati, sou. (d) Permission. All this is in strict line with the ancient Greek.222 A good illustration is seen in Mt. 26:45, kaqeu,dete loipo.n kai. avnaÄ pau,esqe. This is not a question nor necessarily irony. It is too late to do Christ any good by keeping awake. He withdraws his plea for watchfulness. There is irony in plhrw,sate (Mt. 23:32), though it is the permissive use of the imperative. The note of permission is struck in evlqa,tw and evpistrafh,tw (Mt. 10:13). Cf. the fut. ind. in Lu. 10:6. See further cwrize,sqw (1 Cor. 7:15); avgnoei,twgrk)grk grk(14:38, W. H. marg.). In 2 Cor. 12:16 e;stw de, is like our 'Let it be so' or 'Granted.' In Mt. 8:31 avpo,steilon is entreaty, u`pa,gete (32) is permissive. In 1 Cor. 11:6 keira,sqw is probably hortatory. (e) Concession or Condition. It is an easy step from permission to concession. This also is classical.223 Take Jo. 2:19, lu,sate to.n nao.n tou/tonà kai. evn trisi.n h`me,raij evgerw / auvto,n) This is much the same as eva.n lu,shte. It is not a strict command. We have parataxis with kai,, but it is equivalent in idea to hypotaxis with eva,n. So with avnti,sthte tw|/ diabo,lw|à kai. feu,xetai avf v u`mw/n (Jas. 4:7 f.); avna,sta evk tw/n nekrw/n (LXX), kai. evpifau,sei soi o` Cristo,j (Eph. 5:14). See also mh. kri,neteà kai. ouv mh. kriqh/te\ kai. mh. katadika,zeteà kai. ouv mh. katadikasqh/te\ avpolu,eteà kai. avpoluqh,sesqe\ di,doteà kai. doqh,setai u`mi/n (Lu. 6:37 f.). Then again makroqu,mhson evp v evmoi,à kai. pa,nta avpodw,sw
MODE ( EGKLISIS) 949 soi (Mt. 18:26). So also tou/to poi,ei kai. zh,sh| (Lu. 10:28); e;rcesqe kai. o;yesqe (Jo. 1:39). Cf. deu/te kai. poih,sw (Mt. 4:19). Sometimes two imperatives are connected by kai, when the first suggests concession. Thus Eph. 4:26, ovrgi,zesqe kai. mh. a`marta,nete. So also evrau,nhson kai. i;de (Jo. 7:52). Cf. e;rcou kai. i;de (Jo. 1:46). This seems simple enough. (f) In Asyndeton. It is a regular classic idiom224 to have a;ge, fe,re with another imperative. ;Age with klau,sate (Jas. 5:1) is an interjection like deu/ro avkolou,qei moi (Mt. 19:21) and deu/te i;dete (Mt. 28:6). See also Jo. 4:29; 21:12; Rev. 19:17. More common is u[page and u`pa,gete with another imperative. So u[page prw/ton dialla,ghqi (Mt. 5:24); u`pa,gete avpaggei,lategrk grk(28:10). See further Mt. 8:4; 18:15; 21:28; 27:65; Mk. 1:44; 6:38, etc. In Mt. 16:6 we have o`ra/te kai. prose,cete. Cf. also Lu. 12:15. But asyndeton occurs in Mt. 24:6, o`ra/te mh. qroei/sqe) So o`ra/te ble,pete (Mk. 8:15). In Mt. 9:30 the persons and numbers are different, o`ra/te mhdei.j ginwske,tw. In Rev. 19:10, o;ra mh,, the verb with mh, is not expressed. For o[ra poih,seij see also Heb. 8:5 (LXX). The simplest form of asyndeton is seen in Ph. 3:2, ble,peteà ble,peteà ble,pete. (g) In Subordinate Clauses. The reason for treating this subject here is that it is so rare that one may not catch it in the discussion of subordinate clauses. It is well established, though rare, in Demosthenes, Lysias, Plato, Thucydides and the tragic poets.225 The case of w[ste at the beginning of a clause is not pertinent, for there it is a mere inferential conjunction, as, for instance, 1 Cor. 3:21, w[ste mhdei.j kauca,sqw. Here w[ste is not a hypotactic conjunction. Neither is the recitative o[ti, in point, as in 2 Th. 3:10, tou/to parhgge,llomen u`mi/nà o[ti ei; tij ouv qe,lei evrga,zeÄ sqaià mhde. evsqie,tw. In 1 Cor. 1:31 there is probably an ellipsis of ge,nhtai after i[na, and the imperative kauca,sqw is in the direct quotation after ge,graptai. In 1 Pet. 1:6, evn w|- avgallia/sqe (probably imperative), W. H. begin a new sentence, but w|- points back directly to kairw|/ as its antecedent. The same situation occurs in 1 Pet. 3:3 with w-n e;stw. In both examples the imperative appears with the relative. Two other instances of this construction are found in 1 Peter (a peculiarity of this Epistle). They are w|- avnti,sthtegrk grk(5:9) and eivj h'n sth/tegrk grk(5:12). We see it also in Heb. 13:7, w-n- mimei/sqe, and in 2 Tim. 4:15, oa}n kai. su. fula,ssou. Cf. 0. P. 1125, 19 (ii/A.D.), w=n qe,ma kaqaro.n avpo. pa,ntwn avnado,tw.
950 A GRAMMAR OF THE GREEK NEW TESTAMENT Dio, at the beginning of the sentence was hardly felt as a relative (inferential particle), but see 1 Cor. 14:13, dio. proseuce,sqw.226 (h) The Tenses. This matter received adequate discussion under Tenses. It may simply be noted here that in positive sentences the aorist imperative is naturally common, especially frequent in the N. T. Cf. ei;selqeÄÄpro,seuxai (Mt. 6:6). The distinction between the present and the aorist is well seen in a=ron to.n kra,batto,n sou kai. peripa,tei (Jo. 5:8). See also Jo. 2:16 and Ac. 12:8. As an example of the periphrastic present note i;sqi e;cwn (Lu. 19:17). The perfect is almost non-existent, but note pefi,mwso (Mk. 4:39). The present imper. second person alone occurs in prohibitions which are forbidden as in course of action or as a present fact ('quit doing it').227 Cf. Ro. 6:13 for sharp differences in idea between mh. parista,nete (course of action) and parasth,sate (at once and for all). In the third person a prohibition may be either in the aorist imperative or the aorist subj. See the subj. mode for further remarks concerning the failure of the second person imperative aorist in prohibitions. (i) In Indirect Discourse. This subject will receive adequate treatment under this head (see below). All that is attempted here is to indicate that, when the imperative is not quoted directly (cf. 2 Th. 3:10), it may be expressed in an indirect command either by the infinitive (cf. le,gwn mh. perite,mnein mhde. peripatei/n in Ac. 21:21) or by a conjunction like i[na as in Mk. 6:8, or thrown into a deliberative question as in u`podei,xw ti,na fobhqh/te (Lu. 12 :5): B. DEPENDENT OR HYPOTACTIC SENTENCES ( UPOTAKTIKA vAXIWMATA) Introductory. (a) Use of Modes in Subordinate Sentences. There is no essential difference in the meaning of the modes in subordinate clauses from the significance in independent sentences. The division is not made on the basis of the modes at all. Leaving out the imperative because of its rarity in subordinate sentences,the other three modes occur in almost all the subordinate clauses. The same mode-ideas are to be sought here as there. The subordinate clauses make no change in the meaning of mode, voice or tense. Burton228 does say: "Others, however, give to the mood or
MODE ( EGKLISIS) 951 tense a force different from that which they usually have in principal clauses. Hence arises the necessity for special treatment of the moods and tenses in subordinate clauses." I cannot agree to this as the reason for the separate treatment. Sometimes in indirect discourse after secondary tenses there may be a sequence of modes (true also in ancient Greek with final clauses after secondary tenses), but that is so slight a matter that it bears no sort of proportion to the subordinate clauses as a whole. Gildersleeve (A. J. of Phil., XXXIII, 4, p. 489) regards the subordinate sentence as "the Ararat in the flood of change" and parataxis and hypotaxis as largely a matter of style. Some of the modal uses have survived better in the subordinate clauses, as, for instance, the futuristic aorist subj. (cf. o[stij avrnh,shtai in Mt. 10:33), but the subordinate clause did not create the idiom. Originally there were no subordinate sentences.229 "In dependent, clauses the choice of the mood is determined by the nature of each individual case"230 as is true also of independent sentences. The qualification made above about the sequence of modes was always optional and is absent from the N. T. except a few examples in Luke. The great wealth of subordinate clauses in Greek with various nuances demand separate discussion. But we approach the matter with views of the modes already attained. (b) The Use of Conjunctions in Subordinate Clauses. In chapter XXI, Particles, full space will be given to the conjunctions (co-ordinating, disjunctive, inferential, subordinating). Here it is only pertinent to note the large part played in the Greek language by the subordinating conjunctions. It must be admitted that the line of cleavage is not absolute. The paratactic conjunctions were first on the field.231 Popular speech has always had a fondness for parataxis.232 In the modern Greek vernacular "the propensity for parataxis has considerably reduced the ancient Greek wealth of dependent constructions " (Thumb, Handb., p. 185). Hence long periods are rare. So the Hebrew used w. both as paratactic and hypothetic. In the Greek kai, we see a partial parallel.233 In Mt. 26:15, ti, qe,lete, moi dou/nai kavgw. u`mi/n paradw,sw, the kai, is almost equivalent to eva,n) So often in Luke, as in 9:51, evge,neto de.ÄÄkai,, the kai, clause is (like o[ti% the logical subject of evge,neto. The common use of the recitative o[ti illustrates well the close connection besubordinate and independent sentences. The o[ti shows
952 A GRAMMAR OF THE GREEK NEW TESTAMENT that the clause is the object of the preceding verb, but the clause is preserved in the direct (co-ordinate) form. Cf. le,gete o[ti blaÄ sfhmei/j (Jo. 10:36). Thus again a subordinate clause may be so loosely connected with the principal clause as to be virtually independent.234 Thus the relative, as in Latin, often introduces a principal sentence, a paragraph, forsooth, as evn oi-j (Lu. 12:1) and avnq v w-ngrk grk(12:3). But, on the whole, we can draw a pretty clear line between the independent and the dependent clause by means of the conjunctions. The case of asyndeton, treated elsewhere (cf. The Sentence), concerns chiefly parataxis, but some examples occur in hypotaxis, as in kai. evge,neto- ei=pe,n tij (Lu. 11:1) where the ei=pe,n clause is the logical subject of evge,neto. (c) Logical Varieties of Subordinate Clauses. Each subordinate clause sustains a syntactical relation to the principal clause after the analogy of the case-relations. The normal complete sentence has subject, predicate, object. Each of these may receive further amplification (see chapter X, The Sentence). The predicate may have a substantive (as subject or object). This substantive may be described by an adjective. An adverb may be used with predicate, adjective or substantive. Thus the sentence is built up around the predicate. In the same way each subordinate sentence is either a substantive (subject or object like an o[ti clause), an adjective like o[stij or an adverb like o[pou. This is therefore a point to note about each subordinate clause in order to get its exact syntactical relation to the principal clause. It may be related to the predicate as subject or object, or to the subject or object as adjective, or to either as adverb. A relative clause may be now substantive, now adjective and now adverb. In simple truth most of the conjunctions have their origin as relative or demonstrative pronouns. In Kuhner-Gerth235 the subordinate clauses are all discussed from this standpoint alone. Thumb (Handb., pp. 186 ff.) follows this plan. One questions the wisdom of this method, though in itself scientific enough. Burton236 has carefully worked out all the subordinate clauses from this standpoint, though he does not adopt it. Then, again, one may divide these clauses according to their form or their meaning.237 Viteau238 combines both ideas and the result is rather confusion than clarification. There may be a series of subordinate clauses, one dependent on the other. So in 1 Cor.
MODE ( EGKLISIS) 953 1:14, euvcaristw/ o[ti ouvde,na u`mw/n evba,ptisa eiv mh. Kri,spon kai. Gai/on * i[na mh, tij ei;ph| o[ti eivj to. evmo.n o;noma evbapti,sqhte. See also Mk. 6:55 and section 10 in this chapter. The infinitive and the participle are used also in subordinate clauses, but they do not directly concern the problem of the modes save in indirect discourse. They are so important and partake of the functions of both noun and verb to such an extent that they demand a separate chapter -XX. 1. RELATIVE SENTENCES. (a) Relative Sentences Originally Paratactic. The relative o[j, as is well known, was first an anaphoric substantive pronoun.239 At first the relative clause was paratactic, a principal sentence like the other.240 Cf. o[j ga,r in Homer, where o[j may be taken241 as demonstrative or relative. In its simplest form the relative was unnecessary and was not even a connective. It was just a repetition of the substantive.242 "The relative force arises where o[j (and its congeners) connects and complements."243 Indeed, the relative sentence is probably the oldest form of parataxis.244 It is only by degrees that the relative clause came to be regarded as a subordinate clause.245 As a matter of fact, that was not always the case, as has been seen in such examples as evn oi-jà avnq v w-n (Lu. 12:1, 3). But it is not true that this subordination is due to the use of the subjunctive mode.246 The effect of case-assimilation (cf. gender and number) and of incorporation of the antecedent was to link the relative clause very close to the principal sentence.247 Cf. Heb. 13:11. (b) Most Subordinate Clauses Relative in Origin. This is true not merely of o[ti and o[te which are accusative forms248 of o[, but also of other adverbs, like the ablative w`jà o[pwjà e[wj. These subordinating conjunctions therefore are mostly of relative origin."
954 A GRAMMAR OF THE GREEK NEW TESTAMENT Cf. i[naà o`po,te and perhaps eiv. Pri,nà evpei,à a;crià me,cri are not relative. Thus the subordinate clauses overlap. Burton,249 indeed, includes e[wj under relative sentences. That is not necessary, since thus nearly all the subordinate clauses would properly be treated as relative sentences. See the relative origin of various conjunctions well worked out by Schmitt,250 Weber251 and Christ.252 These clauses are mainly adverbial, though objective (and subject-clause also) o[ti (indirect discourse) is substantive simply. The word w`j occurs in Homer with the three values of demonstrative, relative and conjunction (cf. English "that").253 But here we pass by these conjunctions from relative or demonstrative roots.254 The relative pronoun alone, apart from the adverbial uses, introduces the most frequent subordinate clause, probably almost equal in some authors to all the other classes put together. In 1 Peter the relative construction is very common. Cf. 1 Pet. 1:6-12; 2:21-24. At any rate it is the chief means of periodic structure.255 Take as an instance the period in Ac. 1:1-2. Note w-nà a;cri h-j h`me,rajà ou;jà oi-j, all the subordinate clauses in the sentence except infinitive and participles. See also 1 Cor. 15:1-2, where four relatives occur and ti,ni lo,gw| is almost like a relative. Cf. further Ro. 9: 4 f. The relative sentence may be repeated indefinitely with or without kai,. (c) Relative Clauses Usually Adjectival. They are so classed by Kuhner-Gerth.256 The descriptive use followed the original substantive idiom just as the relative itself was preceded by the demonstrative. Thus the use of the relative clause as subject or object like o` and the participle is perfectly consistent. So oa}j a'n evme. de,xhtai de,cetai to.n avpostei,lanta, me (Lu. 9:48). Cf. also Mk. 9:37; Ac. 16:12. The descriptive character of the relative clause is well shown in th.n ma,cairan tou/ pneu,matoj o[ evstin r`h/ma qeou/ (Eph. 6:17). Cf. o[j in 1 Tim. 3:16. The adjectival use of the relative sentence is accented by the use of the article with it in Ro. 16:17, skopei/n tou.j ta.j dicostasi,aj kai. ta. ska,ndala para. th.n didach.n ha}n u`mei/j evma,qete poiou/ntaj. Here the relative clause is adjectival, but in itself a mere incident between tou,j and poiou/ntaj.
MODE ( EGKLISIS) 955 The clause is simply adjectival with pa/j o[j in Lu. 12:8. That comes to be its most usual character. So with div h=j is in Heb. 12:28. (d) Modes in Relative Sentences. There is nothing in the relative pronoun or the construction of the clause per se to have any effect on the use of the mode.257 The relative, as a matter of fact, has no construction of its own.258 In general in dependent clauses the choice of the mode is determined by the nature of the individual case.259 Outside of relative clauses the choice in the N. T. is practically confined to the indicative and the subjunctive. The optative holds on in one or two examples. With the relative some examples of the imperative occur, as has already been shown. Cf. 1 Cor. 14:13; Tit. 1:13; 2 Tim. 4:15; 1 Pet. 5:9; Heb. 13:7. Cf. o;qen katanoh,sate (Heb. 3:1). But the mode is not due at all to the relative. In a word, the relative occurs with all the constructions possible to an independent sentence.260 The indicative is, of course, the natural mood to use if one wishes to make a direct and clear-cut assertion. Thus ouvdei.j e;stin o[j avfh/ken th.n oivki,an (Mk. 10:29). Cf. Jo. 10:12. The various uses of the subjunctive occur with the relative. The deliberative subj. is seen in pou/ evsti.n to. kata,luma, mou o[pou to. pa,sca meta. tw/n maqhtw/n mou fa,gw; (Mk. 14:14; Lu. 22:11).261 Prof. Earle, in a fine paper on "The Subj. of Purpose in Relative Clauses in Greek" (Class. Papers, 1912, pp. 213 ff.) shows how Xenophon, Soph., Eurip., Plato and other Attic writers use the idiom. Cf. Xen., Anab., II, 4, 20, ouvc e[xousin evkei/noi o[poi fu,gwsin. See also Tarbell, Class. Review, July, 1892, "The Deliberative Subj. in Relative Clauses in Greek." The subj. may be volitive as in Ac. 21:16, a;gontej par v w|- xenisqw/men Mna,swni, tini, and in Heb. 8:3, o[qen avnagkai/on e;cein ti kai. tou/ton oa} prosene,gkh| (cf. oa} prosfe,rei in Heb. 9 : 7). In Heb. 12: 28, di v h-j latreu,wmen, the subj. may be conceived as either volitive (hortatory) or merely futuristic, more probably volitive like e;cwÄ men. Clearly futuristic is the subj. in Mt. 16:28, oi[tinej ouv mh. geu,swntai qana,tou. These examples appear isolated. Cf. subj. with w[ste (not relative) as in 1 Cor 5:8, w[ste e`orta,zwmen (deliberative). But the futuristic subj., so rare in the independent sentence after Homer, is very common in the relative clause with
956 A GRAMMAR OF THE GREEK NEW TESTAMENT a'n and sometimes without a;n. It is not the a;n that determines the subj., but the subj. usually has a;n. Thus oa}j ga.r eva.n qe,lh| and oa}j d v a'n avpole,sh| Rec. (Mk. 8:35). Cf. o[stij thrh,sh| (Jas. 2:10), though AKLP read thrh,sei (itacism). Cf. Mt. 10:33 and 38. In such relative sentences the future indicative is also very common, the two forms being closely allied in form and sense. Cf. oa}j a'n o`moÄ logh,sei. (Lu. 12:8). See also o[stij o`mologh,sei and o[stij avrnh,shtai (Mt. 10:32 f.). (e) Definite and Indefinite Relative Sentences. Goodwin262 has made popular the custom of calling some relative sentences " conditional relatives." He has been followed by Burton.263 Jannaris264 considers conditional relative clauses "virtually condensed clauses capable of being changed into conditional protases." Almost any sentence is capable of being changed into some other form as a practical equivalent. The relative clause may indeed have the resultant effect of cause, condition, purpose or result, but in itself it expresses none of these things. It is like the participle in this respect. One must not read into it more than is there. Cf. oa}j e;cei w=ta (Mk. 4:9) and o` e;cwn w=ta (Mt. 13:9). Cf. ei; tij in Mk. 4:23. One might as well say that o` lamba,nwn (Jo. 13:20) is the same thing as oa}j lamba,nei (cf. Mt. 10:38). There is a change from participle to relative clause in Mt. 10:37 f., 41 f. Cf. Mt. 12: 30, 32; Lu. 9:50. So then a'n tina pe,myw (Jo. 13:20) is a conditional clause.265 It is true that o[n tina does not occur in the N. T., but ei; tij and o[jtij differ in conception after all, though the point is a fine one. The MSS. sometimes vary between ei; tij and o[stij as we see in Mk. 8:34; 1 Cor. 7:13. In Jo. 14:13 f. note o[ti a'n aivth,shte and eva,n ti aivth,shte. Note the distinction between oa} keca,rismai and ei; ti keca,rismai, in 2 Cor. 2:10. In Mk. 8:34 f. note ei; tij qe,lei - oa}j eva.n qe,lh|. What is true is that the relative sentences are either definite or indefinite. It is not a question of mode nor of the use of a;n, but merely whether the relative describes a definite antecedent or is used in an indefinite sense. The definite relative is well illustrated by 2 Th. 3:3, pisto.j de, evstin o` ku,rioj o[j sthri,xei, or Mk. 1:2, to.n a;ggelo,n mou oa}j kataskeua,sei th.n o`do,n mou. So also ca,rin di v h-j latreu,wmen (Heb. 12:28). Cf. o[ prosene,gkh| (Heb. 8:3). But indefinite is oa}j e;ceià doqh,setai auvtw|/ (Mk. 4:25). In the same verse kai. oa}j ouvk e;cei is indefinite, but kai. oa} e;cei is definite. Indefinite also is o[soi h[yanto (Mt. 14:36) and
MODE ( EGKLISIS) 957 o[soi a'n h[yanto (Mk. 6:56). So also with pa/j oa}j evrei/ (Lu. 12:10) and pa/j oa}j a'n o`mologh,seigrk grk(12:8). Cf. oa}j e;staigrk grk(17:31) with oa}j eva.n zhth,sh|grk grk(17:33) and oa}j d v a'n avpole,sei. Cf. Ac. 7:3, 7; Gal. 5:17. That it is not a question of mode is thus clear. Cf. oa}j eva.n qe,lh| with oa}j a'n avpole,sei (Mk. 8:35). Thus note in Mk. 4:25 oa}j ga.r e;cei doqh,setai auvtw|/, but in Lu. 8:18 oa}j a'n ga.r e;ch| doqh,setai auvtw|/.266 So in Lu. 12:8 we have pa/j oa}j a'n o`mologh,sei evn evmoi,, but in Mt. 10:32 pa/j oa}stij o`mologh,sei evn evmoi,) The use of o[stij is pertinent. It is either indefinite, as here, from the sense of tij= 'any one' or definite from the sense of tij = 'somebody in particular,' as in Lu. 9:30, a;ndrej du,o sunela,loun auvtw|/ oi[tinej h=san Mwu?sh/j kai. vHlei,aj. Examples of the definite use of o[stij may be seen in Mt. 7:26; 16:28; 22:2; 27:55, 62, etc. The indefinite use is seen in pa/j o[stij avkou,ei (Mt. 7:24), o[stij e;cei (Mt. 13:12), o[stij u`yw,sei (Mt. 23:12), but apparently no instance of o[stij a;n and the future ind. occurs. The indefinite use of o[stij with the subj. and a;n is uniform (11 examples), as in o[stij eva.n h|= (Gal. 5:10), o[stij a;n poih,sh| (Mt. 12:50). Cf. Col. 3:17. We also find o[stij avrnh,shtai (Mt. 10: 33), o[stij thrh,sh| (Jas. 2:10), but the definite use in Mk. 9:1. In 2 Cor. 8:12, eiv h` proqumi,a pro,keitaià kaqo. eva.n e;ch|à euvpro,sdektojà ouv kaqo. ouvk e;cei, there is a pointed distinction between the subjunctive and the indicative modes.267 Thus the indicative occurs with either the definite or the indefinite and the subjunctive with the indefinite 122 times, the definite only Mk. 9 : 1= Mt. 16:28. One may make a positive statement about either a definite or an indefinite relative or a doubtful assertion about either. The lines thus cross, but the matter can be kept distinct. The distinction is clearly perceived by Dawson Walker.268 The subjunctive with the indefinite relative, like that with o[tan and eva,n, is futuristic (cf. also future indicative). Moulton (Prol., p. 186) argues that, since this subj. is futuristic and the aorist describes completed action, the aorist subj. here is really a future perfect. "Thus Mt. 5:21, oa}j a'n foneu,sh|, 'the man who has committed murder.'" But this seems rather like an effort to introduce the Latin idiom into the Greek and is very questionable. (f) The Use of a;n in Relative Clauses. This is the place for more discussion of a;n, though, sooth to say, the matter is not perfectly clear. See also Conditions. It is probably kin to the Latin an and the Gothic an, and had apparently two meanings,
958 A GRAMMAR OF THE GREEK NEW TESTAMENT 'else' and 'in that case rather.' Monro269 argues that the primary use of a;n and ke,n is with particular and definite examples. Moulton (Prol., p. 166) translates evgw. de, ken auvto.j e[lwmai by the Scotch 'I'll jist tak her mysel'.' There was thus a limitation by circumstance or condition. The use of a;n with relative, temporal and conditional clauses "ties them up to particular occurrences" (Moulton, Prol., p. 186). It is not always quite so easy as that. This use of modal a;n appears rarely in modern Greek (Thumb, Handb., p. 188). "It is a kind of leaven in a Greek sentence; itself untranslatable, it may transform the meaning of a clause in which it is inserted" (Moulton, Prol., p. 165). That is putting it a bit strong. I should rather say that it was an interpreter of the sentence, not a transformer. Moulton counts 172 instances of modal a;n, ( eva,n) in the N. T. (p. 166). Matthew leads with 55, then Mark 30, , Gospel of Luke 28 and Acts only 10, Paul's Epistles 27, the Johannine writings only 20, Hebrews 1, , James 1. Mr. H. Scott fears that these figures are not correct, but they are approximately so. The MSS. vary very much. These examples occur with incl. or subj. Moulton finds 739 cases of modal a;n in the LXX (Hatch and Redpath). Of these 40 are with opt. (26 aorist), 56 with ind. (41 aorist, 6 imp., 1 plup., 1 pres., 7 fut. ind.), the rest with subj. Radermacher (N. T. Gr., p. 165) finds modal a;n in the koinh, decreasing and unessential with ind., subj. or opt. in relative, temporal, final or conditional clauses. The use with indefinite or general statements was rare in Homer, but gradually came to be more frequent. But in the N. T. some examples of the definite use of a;n survive especially in temporal clauses. So in Rev. 8:1, o[tan h;noixen. But o[tan sth,kete (Mk. 11:5) may be general. There is doubt also about o[tan ovye. evge,netogrk grk(11:19). But in Mk. 6:56, o[soi a'n h[yanto, the construction is rendered more definite by a;n, though o[pou a'n eivseporeu,eto in the same verse is indefinite. In Mt. 14:36 we have o[soi h[yanto, which is not more definite than Mark's construction.270 In Rev. 14:4, o[pou a'n u`pa,gei, the construction is indefinite. In Ac. 2:45 and 4:35, kaqo,ti a;n tij ei=cen, we have repetition and so a general statement to that extent. In Mk. 3:11, o[tan auvto.n evqew,roun, it is general. In most instances in the N. T., therefore, the use of a;n is clearly in indefinite relative clauses whether with the indicative or subjunctive.271 It
MODE ( EGKLISIS) 959 cannot he said that a;n is necessary with the indefinite relative and the indicative. It does not occur in the N. T. with o[stij and the future incl, but we have both o[stij o`mologh,sei (Mt. 10:32) and o[j a'n o`mologh,sei, (Lu. 12:8); oa}j e;stai (Lu. 17:31) and oa}j a'n avpole,sei (Mk. 8:35). For oa}j a'n and fut. ind. see Compernass, De Sermone Pis., p. 38. Radermacher (N. T. Gr., p. 145) cites oa}j d v a'n avdikh,seià Inscr. Petersen-Luschan, Reisen, p. 174, N. 223, 21. As already seen, the relative with the subj. usually has a;n, as eivj ha}n a'n po,lin eivse,rchsqe (Lu. 10:8); o[ti a'n prosdapanh,sh|j (10: 35). Cf. w|- a'n bou,lhtaigrk grk(10:22). In a few examples the best MSS. do not have a;n, as in o[stij avrnh,shtai (Mt. 10:33); o[stij thrh,sh|ÄÄptai,sh| de, (Jas. 2:10). The use of eva,n like a;n has been shown (cf. Orthography) to be very common with relatives at this period. It is immaterial which is found. So oa}j eva.n lu,sh| and oa}j a'n poih,sh| (Mt. 5:19). The MSS. often vary between eva,n and a'n, as in Mt. 10:14; Ac. 7:7. So also o[sa eva.n qe,lhte (Mt. 7:12) and o[sa a'n aivth,shte (Mt. 21:22). But in the N. T., as in the papyri, a'n is twice as common in relative clauses. Radermacher (N. T. Gr., p. 145) quotes o[soi - evgli,pwsi, Inscr. Perg. 249, 26, and oa}j avnaspara,xh| (or a'n avsp)) I. Gr. XII, 1, 671. Moulton (Prol., p. 169) cites C.P.R. 237 (ii/A.D.), o[sa auvtw|/ proste,khtai. He (ib., p. 168) quotes o[s v a'n pa,scete F.P. 136 (iv/A.D.), o[sa eva.n pareÄ labo,mhn B.M. 331 (ii/A.D.). The a'n is not repeated with the second verb. So oa}j a'n poih,sh| kai. dida,xh| (Mt. 5:19). There is no instance of a;n in a relative clause with an optative in the N. T. But in Gen. 33:10 the LXX has w`j a;n tij i;doi pro,swpon qeou/) So oi-j eva.n tu,coi, F.P. (see Moulton, Cl. Rev., 1901, p. 32). Radermacher (N. T., Gr., p. 131) cites kaq v oa} a'n me,roj stre,goito from Philo. There is one instance of a;n with the infinitive in the N. T. (2 Cor. 10: 9), i[na mh. do,xw w`j a'n evkrofbei/n u`ma/j, but a;n is here probably the same as eva,n and w`j a;n = 'as if.' The upshot of it all is that a;n has no peculiar construction of its own. It is more frequent with the subjunctive than with the indicative in relative sentences, but is not absolutely essential with either mode.272 In the Attic the subj. is invariable with a;n, but "in the less cultured Hellenistic writers" (Moulton, Prol., p. 166) it occurs with the ind. also. Curiously in the Gospel of John ap occurs with o[stij only in the neuter (Abbott, Johannine Grammar, p. 304). Always in the N. T. o[ti eva,n= o[ti a;n unless in Mk. 6:23 the correct text is o[ti oa} eva,n as in margin of W. H. The text is probably correct (cf. Lu. 10:35; Ac. 3:23, etc.).
960 A GRAMMAR OF THE GREEK NEW TESTAMENT (g) Special Uses of Relative Clauses. As in Latin,273 the relative clause may imply cause, purpose, result, concession or condition, though the sentence itself does not say this much. This is due to the logical relation in the sentence. The sense glides from mere explanation to ground or reason, as in oa} kai. evspou,dasa auvto. tou/to poih/sai, (Gal. 2:10). In 1 Cor. 3:17, o` nao.j tou/ qeou/ a[gio,j evstin oi[tine,j evste u`mei/j, there is an argument in oi[tinej. This is clearly true274 in Ro. 6:2, oi[tinej avpeqa,nomen th|/ a`marti,a|à pw/j e;ti zh,somen evn auvth|/; Cf. also Ac. 10:41, oi[tinej sunefa,gomen kai. suneÄ pi,omen auvtw|/. See Gal. 5:4, oi[tinej evn no,mw| dikaiou/sqe. Cf. Latin qui, quippe qui. A good example is seen in Ro. 8:32, or o[j ge tou/ ivdi,ou ui`ou/ ouvk evfei,sato. Cf. also aa} e;mellon (Rev. 3:2) and the common avnq v w-n (Lu. 1:20). Cf. Ac. 10:47; Ro. 1:25, 32; Ph. 2: 20; Col. 3:5. Only the ind. mode occurs in the N. T. in this construction.275 Purpose is also found in relative clauses (cf. Latin qui= ut is). Either the future ind. or the subj. is used for this construction. When the subj. occurs it is probably volitive.276 So Burton277 would explain all the cases of subj. of purpose with relatives, but wrongly. The use in Mk. 14:14 is analogous to the retention of the subj. of deliberation in an indirect question. Cf. the subj. of purpose with relative clause in Attic Greek.278 But the subj. construction is Homeric (like Latin also). The Attic idiom is the future ind., and the future ind. also appears in the N. T. So oa}j kataskeua,sei (Mk. 1:2= Mt. 11:10 = Lu. 7:27), oa}j u`ma/j avnamnh,seià (1 Cor. 4:17) which may be contrasted with the merely explanatory relative o[j evsti,n mou te,knon in the same sentence. So oi[tinej avpodw,sousin auvtw|/ (Mt. 21:41); oia} proporeu,Ä sontai. (Ac. 7:40; Ex. 32:1); ouvk e;cw oa} paraqh,sw (Lu. 11:6) where the Attic Greek would279 have o[ti. Sometimes i[na occurs where a relative might have been used. So 2 Cor. 12:7 evdo,qh moi sko,loy ÄÄi[na me kolafi,zh|, (Jo. 5:7) ouvk e;cw a;nqrwpon i[na ba,lh| me,grk grk(9:36) i[na pisteu,sw eivj auvto,n. Cf. Gal. 4:5; Rev. 19:15. Viteau280 strikingly compares Mt. 10:26, oa} ouvk avpokalufqh,setai and oa} ouv gnwsqh,Ä setai, with Mk. 4:22, eva.n mh. i[na fanerwqh|/ and i[na e;lqh| eivj fanero,n. The variety of construction with o[j is illustrated by Mt. 24:2 (Lu. 21:6), oa}j ouv kataluqh,setai, and Mk. 13:2, oa}j ouv mh. kataluqh|/.
MODE ( EGKLISIS) 961 The classic idiom preferred the fut. ind. for purpose with the relative (Schmid, Atticismus, IV, p. 621), but Isocrates (IV, 44) has evf v oi-j filotimhqw/sin. Radermacher (N. T. Gr., p. 138) cites for the koinh, Diod. XI, 21, 3, di v ou- tro,pou- avne,lh|; XIV, 8, 3, di v w-n evxe,lwsin; Ach. Tatius, IV, 16, 13, o[son- la,bh|, etc. Purpose is often contemplated result so that the consecutive idea follows naturally that of design. Only the ind. future is used in the N. T., unless one follows Blass281 in taking oa} prosene,gkh| (Heb. 8:3) as result. A good instance of the future ind. is in Lu. 7:4, a;xio,j evstin w|- pare,xh|, which may be profitably compared282 with the non-final use of i[na in Jo. 1:27, a;xioj i[na lu,sw. Burton283 prefers to call this a "complementary limitation of the principal clause," a sort of secondary purpose. But the notion is rather that of contemplated result. The relative denotes a kind of consequence from a particular quality or state.284 See also Ph. 2:20 ouvde,na e;cw ivso,yucon o[stij- merimnh,sei, Mk. 10:29 ouvdei.j e;stin oa}j avfh/ken th.n oivki,an, Lu.7:49 ti,j ou-to,j evstin oa}j kai. a`marti,aj avfi,hsin; Cf. 2 Th. 3:3 pisto.j o[j with 1 Jo. 1:9 pisto.j i[na) An example285 of the concessive use of oi[tinej is seen in Jas. 4:14, oi[tinej ouvk evpi,stasqe th/j au;rion poi,a h` zwh. u`mw/n. The conditional use of the relative clause is only true in a modified sense, as already shown. The relative o[j and o[stij, whether with or without does not mean ei; tij or eva,n tij, though the two constructions are very much alike. There is a similarity between ei; tij qe,lei (Mk. 9:35) and oa}j a'n qe,lh|grk grk(10:43). But I do not agree to the notion of Goodwin286 and Burton287 that in the relative clauses we have a full-fledged set of conditional sentences on a par with the scheme with the conditional particles. That procedure is entirely too forced and artificial for the Greek freedom and for the facts. There is a general sort of parallel at some points, but it is confusion in syntax to try to overdo it with careful detail as Viteau288 does. ;An is not confined to the relative and conditional sentences, but occurs with e[wjà pri,nà w`j, and o[pwj (temporal and final clauses). The indefinite relative like o[j eva.n qe,lh| (Mk. 8:35) or o[stij o`mologh,sei (Mt. 10:32) is quite similar in idea to a conditional clause with eva,n tij or ei; tij. But, after all, it is not a conditional sentence any more than the so-called
962 A GRAMMAR OF THE GREEK NEW TESTAMENT causal, final, consecutive relative clauses are really so. It is only by the context that one inferentially gets any of these ideas out of the relative. All that is true about the indefinite relative clauses has already been explained under that discussion. I therefore pass by any treatment of the kinds of conditional sentences in connection with the relative clauses. (h) Negatives in Relative Clauses. When the subj. occurs the negative is mh,, as in is oa}j a'n mh. e;ch| (Lu. 8:18), but ouv mh, is found in Mk. 13:2, oa}j ouv mh. kataluqh|/. So in Mk. 9:1= Mt. 16:28 we have ouv mh,. With the indicative the negative is 2. CAUSAL SENTENCES. (a) Paratactic Causal Sentences. These do not properly belong here, but there are so many of them that they compel notice. The common inferential particle ga,r introduces an independent, not a dependent, sentence. Paul uses it usually to introduce a separate sentence as in Ro. 2:28; 1 Cor. 15:9. In 1 Cor. 10:17 both o[ti and ga,r occur. It will be treated in the chapter on Particles. Phrases like avnq v w-n (Lu. 12:3), dio, (Mt. 27:8), dio,per (1 Cor. 8:13), o[qen (Ac. 26:19), di v ha}n aivti,an (2 Tim. 1:6, 12), ou- ca,rin (Lu. 7:47) are not always regarded as formally causal. The construction is sometimes paratactic. Indeed, the subordination of the o[ti and dio,ti clauses is often rather loose.290 Thus there is very little difference between o[ti (begins the sentence with W. H.) in 1 Cor. 1:25 and ga,r in 1:26. Cf. also evpeidh, in 1:22. See further o[ti in 2 Cor. 4:6; 7:8, 14, and dio,ti in Ro. 3:20; 8:7. The causal sentence is primarily para-
MODE ( EGKLISIS) 963 tactic. See Mt. 6:5; Lu. 11:32; 1 Cor. 15:29; Heb. 10:2. The subordinate relative is a later development.291 (b) With Subordinating Conjunctions. One may say at once that in the N. T. the mode is always the indicative. There is no complication that arises save with evpei, when the apodosis of a condition of the second class is used without the protasis as in Heb. 10:2, evpei. ouvk a'n evpau,santo. Here the construction is not due at all to evpei,. In the same way we explain evpei. e;dei in Heb. 9:26 and evpei. wvfei,lete a;ra in 1 Cor. 5:10. There is ellipsis also in the rhetorical question in 1 Cor. 15:29, evpei. ti, poih,sousin; But in Ac. 5:38 f. two complete conditional sentences ( eva,n and eiv, protasis and apodosis) occur with o[ti. In a word, it may be said that the indicative is used precisely as in the paratactic sentences. Cf. Jo. 14:19, o[ti evgw. zw/ kai. u`mei/j zh,sete) The negative is usually ouv as in 1 Jo. 2:16. Once in the N. T., Jo. 3:18, o[ti mh. pepi,steuken, we have mh,, but ouv is seen in 1 Jo. 5:10, o[ti ouv pepi,steuken. "The former states the charge, quod non crediderit, the latter the simple fact, quod non credidit" (Moulton, Prol., p. 171). Cf. o[ti mh, in Epictetus IV, 4, 11; IV, 5, 8-9. Cf. Abbott, Joh. Gr., pp. 162, 535. The distinction is subtle, mh, being more subjective and ideal. In Heb. 9:17, evpei. mh. to,te (or mh, pote) ivscu,ei, we likewise meet mh,. In B. G. U. 530 (i/A.D.), evpi. mh. avnte,grayaj auvth|/- o[ti ouvk e;pemyaj pro,j se, note evpi. $eiv% mh, and o[ti ouvk with true distinction. With ouv we have the objective fact, with mh, the element of blame ( me,mfetai) appears. "The comparison of Plutarch with the N. T. shows a great advance in the use of o[ti mh," (Moulton, Prol., p. 239). Cf. also E. L. Green, Gildersleeve Studies, pp. 471 ff.; Radermacher, N. T. Gr., p. 171. He cites o[ti mh. e;ceij, Epictetus IV, 10, 34. It is making inroads on o[ti ouv) We sometimes have avnq v w-n in a truly causal sense as in Lu. 1: 20, and that is true also of o[qen in Mt. 14:7. In Heb. 2:18 evn w|- is practically causal. So also evf v w|- is causal in Ro. 5:12; 2 Cor. 5:4; Ph. 4:10. Cf. kaqa. = 'if right,' P. Oxy. 38 (A.D. 49). The classical evf vw-te does not occur in the N. T. See evf v w|- dw,sei, on condition that he give,' P. Oxy. 275 (A.D. 66). Then w`j may have almost the force of a causal particle as in Jo. 19:33; Mt. 6:12 (cf. Lu. 11:4, kai. ga,r); 2 Tim. 1:3. The same thing is true of kaqw,j in Jo. 17:2. Kaq v o[son is causal in Heb. 7:20 (9:27) and evf v o[son in Mt. 25:40, 45. So kaqo,ti in Lu. 19:9 (cf. 1:7). In Ac. 17:31 HLP. read dio,ti. None of these
964 A GRAMMAR OF THE GREEK NEW TESTAMENT particles are strictly causal, but they come to be so used in certain contexts in the later Greek. We have w`j o[ti in 2 Cor. 5:19; w`j o[ti qeo.j h=n evn Cristw|/ ko,smon katalla,sswn e`autw|/ (cf. our "since that"). Here the Vulgate has quoniam. But in 2 Cor. 11:21 the Vulgate renders w`j o[ti by quasi, as in 2 Th. 2:2, w`j o[ti evne,sthken. Cf. Blass, Gr. of N. T. Gk., p. 321 f. It is found also in Esther 4:14 and is post-classical.292 Dio,ti is found in the Lucan writings, the Pauline Epistles, Hebrews, James and 1 Peter. In the modern Greek293 it takes the form giati,. Once (Ro. 8:21) some MSS. (W. H. read o[ti) have dio,ti in the sense of objective o[ti ('that') as in later Greek (cf. late Latin quia = quod). Instances of causal dio,ti may be seen in Lu. 1:13; Ro. 1:19, etc. It is compounded of dia, and o[ti (cf. English "for that"). In Ph. 2:26 dio,ti is causal and o[ti is declarative. In modern Greek dio,ti, survives in h` kaqareu,ousa. The vernacular has avfou/à evpeidh,à giati, (Thumb, Handb., p. 194). But all other causal particles are insignificant beside o[ti which grew steadily in use.294 It was originally merely relative and paratactic.295 In 1 Jo. 4:3 note o[ - o[ti and o[ti o[ in Ro. 4:21. It is accusative neuter rel. o[ti (cf. o[ti a'n prosdapanh,sh|j, Lu. 10:35) and is more common as the objective particle in indirect discourse (subject or object clause) than as a causal conjunction. In 1 Jo. 5:9 o[ti occurs twice, once as causal and once as objective particle. In 2 Th. 3:7 f. exegesis alone can determine the nature of o[ti. In Jo. 3:19 Chrysostom takes o[ti = 'because.' Cf. also Jo. 16:8-11 (see Abbott, Johannine Gr., p. 158). The English "the reason that" (vernacular "the reason why") is similar. It is very common in 1 John in both senses. In Jo. 1:15 causal o[ti occurs three times in succession. In Lu. 9:49, evkwÄ lu,omen auvto.n o[ti ouvk avkolouqei/ meq v h`mw/n, the present is used because of a sort of implied indirect discourse. In Mk. 9:38 W. H. read o[ti ouvk hvkolou,qei. A good example of causal o[ti is seen in Ro. 5:8. The precise idea conveyed by o[ti varies greatly. In Jo. 9:17, ti, su. le,geij peri. auvtou/à o[ti hvne,w|xe,n sou tou.j ovfqalmou,j; the use of o[ti wavers between objective and causal. Cf. also Mk. 6:17. But we need not appeal to the Hebrew296 for a justification of this balancing of two ideas by o[ti. So in Jo. 2:18, ti, shmei/on deiÄ knu,eij h`mi/nà o[ti tau/ta poiei/j; Akin to this construction is that in
MODE ( EGKLISIS) 965 Jo. 14:22, ti, ge,gonen o[ti, which is shortened into ti, o[ti in Ac. 5: 4, 9. There is a correspondence sometimes between dia. tou/to and o[ti. (Jo. 10:17); dia. ti, and o[ti (Ro. 9:31 f.). Ouvc o[ti may be either objective or causal as in Ph. 4:11, 17; 2 Th. 3:9. In the ancient Greek it meant 'not only do I say that, but I also say.' But in the N. T. it either means 'I say this not because' or 'I do not mean to say that,' and usually the latter according to Abbott.297 We must have a word about evpei,Ã evpeidh,Ã evpeidh,per. As a matter of fact evpeiÄdh,Äper (note the composition) appears in the N. T. only in Lu. 1:1 (Luke's classical introduction). This is undoubtedly a literary touch.298 vEpeidh, is read by W. H. in Lu. 7:1 and Ac. 13:46, but evpei. de, is put in the margin. Eight other examples remain, all in Luke (Gospel and Acts) and Paul (1 Corinthians and Philippians). Cf. Lu. 11:6; 1 Cor. 1:21 f. vEpei,, obsolescent in the late Greek,299 is almost confined to Luke, Paul, the author of Hebrews. Elsewhere in Matthew, Mark and John. Two of these are examples of the temporal use (Mk. 15:42; Lu. 7:1 W. H. marg.). The ordinary causal sense is well illustrated in Mt. 21:46, evpei. eivj profh,thn ei=con. The classical idiom of the ellipsis with evpei, has already been mentioned and is relatively frequent in the N. T. Cf. Ro. 3:6; 11:22; 1 Cor. 14:16; 15:29; Heb. 9:26; 10:2. It occurs in the simplest form in evpei. pw/j (Ro. 3:6) and evpei. ti, (1 Cor. 15:29). In 1 Cor. 14:16, evpei. eva,n, it is equivalent to 'otherwise' and in Ro. 11:22 to 'else,' evpei, eva,n su. evkkoph,sh|. The apodosis of a condition of the second class occurs in 1 Cor. 5:10; Heb. 9:26; 10:2. Verbs of emotion in classical Greek sometimes used eiv (conceived as an hypothesis) rather than o[ti (a direct reason).300 The N. T. shows examples of qauma,zw eiv in this sense (Mk. 15:44; 1 Jo. 3:13), though qauma,zw o[ti is found also301 (Lu. 11:38; Gal. 1: 6). [Oti is the N. T. construction302 with avganakte,w (Lu. 13:14); evxomologe,omai (Mt. 11:25); euvcariste,w (Lu. 18:11); me,lei (Mk. 4: 38); cai,rw (Lu. 10:20); cola,w (Jo. 7:23). Cf. o[ti and evf v w|- in Ph. 4:10. On the possible causal use of o[te and o[tan see article by Sheppard, The Cl. Rev., Sept., 1913. (c) Relative Clauses. This matter received sufficient discussion under Relative Clauses. For examples of o[j take Ro. 8:32;
966 A GRAMMAR OF THE GREEK NEW TESTAMENT Heb. 12:6. For o[stij note Mt. 7:15; Ho. 6:2. See also ou- ca,rin (Lu. 7:47) and di v ha}n aivti,angrk grk(8:47). (d) Dia. to, and the Infinitive. The construction is common in the N. T., occurring thirty-two times according to Votaw303 as compared with thirty-five for the O. T. and twenty-six for the Apocrypha. It is particularly frequent in Luke.304 Cf. Lu. 2:4; 18:5; Ac. 4:2; 8:11, etc. It is not in John except in 2:24, dia. to. auvto.n ginw,skein. Blass305 rejects it here because the Lewis MS. and Nonnus do not have the passage. Here note that o[ti is used side by side with dia. to,. So in Jas. 4:2 f. we have dia. to. mh. eivtei/sqai u`ma/j and dio,ti kakw/j aivtei/sqe on parity. Cf. Phl. 1:7 kaqw,j and dia. to,) In Mk. 5:4, dai. to. dede,sqai kai. diespa,sqai kai. sunteteri,fqai, note the perfect tense and the repetition of the infinitive. Burton306 thinks that here dia, gives rather the evidence than the reason. Why not both? There is one example of the instrumental use of the infinitive to express cause, tw|/ mh. eu`rei/n me (2 Cor. 2:13). The text of B has six examples in the LXX307 (cf. 2 Chron. 28:22, tw|/ qlibh/nai auvto,n). No examples of evpi. tw|/ occur.308 (e) The Participle. We do not have a;teà oi-onà oi-a, as in classical Greek, to give the real reason. That is given simply by the participle as in di,kaioj w'n kai. mh. qe,lwn auvth.n deigmati,sai (Mt. 1:19). It is "exceedingly common" (Moulton, Prol., p. 230). Cf. Jas. 2: 25; Ac. 4:21. But w`j occurs with the participle to give the alleged reason, which may be the real one or mere assumption. Thus in Mt. 7:28 f., w`j evxousi,an e;cwn kai. ouvc w`j oi` grammatei/j, the first w`j gives the ostensible (and true ground) of the astonishment of the people. Cf. also Lu. 16:1;. Ac. 2:2. But in Lu. 23:14, w`j avpostre,fonta to.n lao,n, Pilate does not believe the charge against Jesus to be true. So also with w`j mello,ntwn in Ac. 27:30. 3. COMPARATIVE CLAUSES. The discussion in my Short Grammar309 forms the basis of this section. The conjunctions employed are all of relative origin, but the construction deserves separate treatment. (a) The Relative o[soj. This is a classic idiom and occurs only in Hebrews, except once in Mark. In Heb. 1:4 the correlative is expressed and the comparative form of the adjective is found
MODE ( EGKLISIS) 967 in both clauses. Both correlative and relative are here in the instrumental case, tosou,tw| krei,ttwn geno,menoj tw/n avgge,lwn o[sw| diaÄ forw,teron par v auvtou.j keklhrono,mhken o;noma. The same phenomena are present in 8:6, save that the correlative is absent. In 10:25 there is no comparative in the relative clause. The others are examples of kaq v o[son. In 3:3 there is no correlative, but the comparative appears in both clauses. In 7:20 f. the correlative is kata. tosou/to, but there is no comparative in the relative clause. This is probably causal in idea, as is true of kaq v o[son in 9:27, where there is no comparative, though we have the correlative ou[twj kai,. The example in Mk. 7:36, o[son de. auvtoi/j dieste,lleto auvtoi. ma/llon perisso,teron evkh,russon, lacks the correlative and has no comparative with the relative, but has a double comparison in the principal clause. In Jo. 6:11 and Rev. 21:16, o[son is simply relative, not a conjunction. The causal and temporal uses of o[son are discussed elsewhere. (b) Relative o[j with kata,. The singular kaqo, is found only in Ro. 8:26 kaqo. dei/, 1 Pet. 4:13 kaqo. koinwnei/te, and 2 Cor. 8:12 kaqo. eva.n e;cw| euvpro,sdektojà ouv kaqo. ouvk e;cei, where a good distinction is drawn between the subjunctive and the indicative. Cf. 0. P. 1125, 14 (ii/A.D.) kaqo. misqoi/ me,roj. The construction with eva,n is like that of the indefinite relative with eva,n ( a;n) and the subj. The plural kaqa,, is found only once in the N. T. (Mt. 27:10). Kaqa,per, however, is found seventeen times (three doubtful as compared with kaqw,j, Ro. 9:13; 10:15; 2 Cor. 3:18) and all in Paul's writings save in Heb. 4:2 (without verb). It is thoroughly Attic and a slight literary touch. Cf. 1 Cor. 10:10. The mode is always indicative, but cf. kaqa. avre,skh| in Gen. 19:8. In Ro. 12:4 the correlative is ou[twj. (c) Kaqo,ti in a Comparative Sense. It occurs only twice (Ac. 2:45; 4:35) and the same idiom precisely each time, kaqo,ti a;n tij crei,an ei=cen. Here ay seems to particularize each case from time to time (note imperfect tense), the iterative use of a;n, (Moulton, Prol., p. 167). This usage approaches the temporal in idea. The classic idiom of the aorist ind. with a;n, no longer appears with these conjunctions. (d) `Wj and its Compounds. These are the most common comparative particles. The most frequent of all is w`j itself which has various other uses as exclamatory ( w`j w`rai/oi oi` po,dej in Ro. 10: 15), declarative like o[ti (Ac. 10:28), causal (Mt. 6:12), temporal (Lu. 12:58), with the infinitive (Lu. 9:52; Heb. 7:9), as a final particle ( w`j teleiw,sw, Ac. 20 : 24, W. H. text), with superlative
968 A GRAMMAR OF THE GREEK NEW TESTAMENT adverbs ( w`j ta,cista, Ac. 17:15), with the sense of 'about,' as w`j disci,lioi (Mk. 5:13) and with participles ( w`j me,lwn, Ac. 23: 20). The richness of this particle is thus illustrated. But the comparative relative adverb is the origin of them all. In Heb. 3:11; 4:3 w`j may be consecutive 'so,' but w`j is more often comparative than anything else. Usually w`j has a correlative. Thus ou[twj- w`j (1 Cor. 4:1); w`j- ou[twj (Ac. 8:32); w`jÄÄÄou[twj kai, (2 Cor. 7:14); w`j- kai, (Gal. 1:9); i;sojÄÄw`j kai, (Ac. 11:17); kai,ÄÄ w`j kai, (Mt. 18:33). But often no correlative is expressed (cf. Mt. 8:13).310 The verb is not always expressed. Thus w`j oi` u`pokriÄ tai, (Mt. 6:5). This predicate use of w`j is very extensive. Cf. w`j kai, (1 Cor. 7:7). The mode is usually the indicative, as in Mk. 10:1, but the subj. occurs in Mk. 4:26, w`j a;nqrwpoj ba,lh| (cf. w`j ouvk oi=den). Blass311 considers this "quite impossible," but it is read by aBD. Some late MSS. add eva,n and others read o[tan, but surely eva,n ( a;n) is not "indispensable" to the subj. (cf. Mt. 10:33). In Gal. 6:10, w`j kairo.n e;cwmen, the temporal w`j is likewise minus a;n. See Relative Clauses and discussion of a;n which is by no means necessary in these subj. clauses. Cf. Radermacher, N. T. Gr., p. 164. In 1 Th. 2:7, w`j eva.n trofo.j qa,lph| ta. e`auth/j te,kna, we do have eva,n, but the construction in Mark is not lawless. Kaqw,j comes next to w`j in frequency (chiefly with Luke and Paul). It sometimes has the correlative. So ou[twj kaqw,j (Lu. 24:24); kaqw,j - ou[twj (Jo. 3:14); kaqw,j - ou[twj kai, (2 Cor. 8:6); kaqw.j kai, - ou[twj kai, (Col. 3:13); kai, ÄÄkaqw.j kai, (Ro. 1:13); kaqw,j- kai, (Jo. 15:9); o`moi,wj kaqw,j (Lu. 17:28), and note kata.ta. auvta, in verse 30. The correlative is not always expressed (Mt. 21: 6). So in Col. 1:6, kaqw.j kai,. Sometimes the principal clause is unexpressed as in 1 Tim. 1:3, or only ouv occurs, as ouv kaqw,j (1 Jo. 3: 12; Jo. 6:58). It is a late word but is abundant in the papyri. In the N. T. it occurs only with the indicative. The word, as already noted, sometimes has a causal sense (Ro. 1:28). It may have a temporal signification in Ac. 7:17. It occurs in indirect question in Ac. 15:14, and is epexegetical in 3 Jo. 1:3. Kaqw,sper is read only once in the N. T. (Heb. 5 : 4), though W. H. put it in the margin in 2 Cor. 3:18 (text kaqa,per). `Wsei, is classical, but has no verb (cf. Mt. 3:16; Mk. 9:26, etc.) in the N. T., though it occurs with the participle w`sei, pro,bata mh. e;conta poime,na (Mt. 9:36). Cf. also Ro. 6:13. It is used in the sense of 'about' as in Lu 9:14, 28, etc. It is commonest in the Gospels and Acts.
MODE ( EGKLISIS) 969 In 2 Cor. 10: 9 we have w`j a'n evkfobei/n (here alone in the N. T. with infinitive) = 'as if to frighten.' [Wsper occurs with the indicative as in Mt. 6:2. In Mt. 25:14 a parable is thus introduced, but with no correlative. But we have the correlative in Ro. 5:19 (6:4), w[sper- ou[twj kai,. So Jo. 5:21. So w[sper- w`sau,twj (Mt. 25:14-18); w[sper- ou[twjgrk grk(13:40). We find w[sper also with the participle (cf. Ac. 2:2). Often the verb is wholly wanting as in Mt. 6:7. We meet w`sperei, only once (1 Cor. 15:8) and that without a verb. 4. LOCAL CLAUSES. These are all relative adverbial sentences and are usually treated with relative sentences, but they are worthy of a separate note. The adverbs (conjunctions) used are o[qenà ou-à o[pou. With o[qen only the indicative is found as in Lu. 11:24, o[qen evxh/lqon. More common than o[qen is ou- as in Mt. 2: 9, ou- h=n to. paidi,on. Cf. past perfect in Ac. 20:8. It occurs mainly in Luke's writings and always with the indicative save once in 1 Cor. 16:6, ou- eva.n poreu,wmai. Here the indefinite relative naturally has a;n and the subjunctive. Ou- is used with verbs of motion as well as with those of rest as this passage shows. Cf. also Lu. 10:1, ou- h;mellen auvto.j e;rcesqai. But o[pou is the usual local conjunction in the N. T., particularly in Matthew, Mark and John (Gospel and Revelation). It occurs with verbs of rest as in Mk. 2:4, o[pou h=n, and of motion as in Jo. 7:34, o[pou u`pa,gw. The indicative is the usual mode. Once, Mk. 6:56, o[pou a'n eivsepoÄ reu,eto, we find a;n to emphasize the notion of repetition in the imperfect tense, but this is not necessary. Cf. o[pou h;qelej (Jo. 21: 18). Note the emphatic negative in o[pou ouv qe,leij (ib.). Cf. also o[pou a'n u`pa,gei (Rev. 14:4) where a;n occurs with the present ind. (indefinite relative). In o[pou fa,gw (Mk. 14:14; Lu. 22:11), as noted on p. 964, the subj. is probably deliberative, answering to pou/ fa,gw in the direct question. Cf. ouvk e;cei pou/ th.n kefalh.n kli,nh| (Lu. 9:58). But the subj. with eva,n in o[pou eva.n avpe,rch| (Lu. 9: 57) is the common futuristic subj. So in the parallel passage in Mt. 8:19. See further Mt. 24:28; 26:13; Mk. 6:10; 9:18; 14:9, 14. Curiously enough all the N. T. instances of o[pou with the subj. are found in the Synoptic Gospels. There is ellipsis of the copula in Rev. 2:13, as is not infrequent with relatives. [Opou is used also in metaphorical relations, as in Heb. 9:16. The correlative adverb evkei/, occasionally appears with o[pou as in Lu. 12:34; 17:37; Jo. 12:26. Kai, is a correlative in Jo. 17:24. The use of o[pou in classical Greek is confined to indefinite sentences, but the N. T. shows a frequent use (especially in John)
970 A GRAMMAR OF THE GREEK NEW TESTAMENT where there is a definite antecedent.312 Cf. Jo. 1:28; 4:46; 7: 42; 10:40; 12:1, etc. 5. TEMPORAL CLAUSES. (a) Kin to Relative Clauses in Origin and Idiom. Blass313 bluntly says that temporal clauses introduced by o[te and o[tan "are generally only a special class of relative sentence, and exhibit the same constructions." The same thing is true of local sentences. Burton314 carries this conception to such a point that he has no separate treatment of temporal sentences at all. This is surely going too far. Thompson315 sees the matter rightly when he says: "The vague original relative import becomes specialized." Hence we expect to find both definite and indefinite temporal clauses as with other relative (and local) clauses. Definite temporal clauses may be illustrated by Mt. 7:28, o[te evte,lesen o` vIhÄ sou/j tou.j lo,gouj tou,toujà evxeplh,ssonto oi` o;cloi. The indefinite is shown in Jo. 15:26, o[tan e;lqh| o` para,klhtoj. The temporal clause may be indefinite in its futurity, frequency and duration.316 Indefinite futurity is the most common, indefinite duration the least common. The modes used in temporal clauses in the N. T. are the indicative and the subjunctive. These uses conform to the historical development of the two modes. There is one example of the optative in a temporal clause (Ac. 25:16, pro.j oua}j avpekri,Ä qhn o[ti ouvk e;stin e;qoj `Rwmai,oij cari,zesqai, tina a;nqrwpon pri.n h' o` kathgorou,menoj kata. pro,swpon e;coi tou.j kathgo,rouj to,pon te avpologi,aj la,boi peri. tou/ evgklh,matoj). Here, as is evident, the optative is due to indirect discourse, not to the temporal clause. The subjunctive with a'n ( pri.n h' a'n e;ch|ÄÄla,bh|) occurs rather than the optative according to sequence of modes. This sequence was optional and a classic idiom, and so is found in the N. T. only in Luke's writings. Observe that e;stin, is retained in the indicative. This sentence is a fine illustration of the Greek subordinate clauses. In the context in Acts it is seen that four dependent clauses precede the pri.n h; clause in the long sentence. The use of a;n or eva,n in temporal clauses has very much the same history as in other relative clauses. The usage varies with different conjunctions and will be noted in each instance. The point of time in the temporal clause may be either past, present or future. It is a rather complicated matter, the Greek temporal clause, but not so much so as the Latin cum clause, "in which the Latin lan-
MODE ( EGKLISIS) 971 guage is without a parallel."317 The different constructions may be conveniently grouped for discussion. Just as the optative with temporal clauses vanished, so there came a retreat of various temporal conjunctions. As a result in the later Greek the construction is much simpler.318 (b) Conjunctions Meaning 'When.' The classic use of the optative for repetition with such clauses has been effectually sidetracked in the vernacular koinh, (Radermacher, N. T. Gr., p. 130). Only the ind. and subj. modes occur in these clauses. vEpei has vanished319 in this sense, save in Lu. 7:1 where it is a variant (margin in W. H. and Nestle) for evpeidh,, the correct text. Curiously enough this is also the only instance of the temporal use of evpeidh, in the N. T., evpeidh. evplh,rwsen. It is a definite point of time in the past and naturally the indicative occurs. There are three examples of 972 A GRAMMAR OF THE GREEK NEW TESTAMENT e[wj ei;phte in Lu. 13:35, but Nestle still reads e[wj h[xei o[te ei;phte. The text is in much confusion, but at any rate here is manuscript evidence for the subjunctive with o[te without a;n. This is in harmony with what we saw was true of o[j and o[stij. It is also a well-known Homeric idiom.321 Radermacher (N. T. Gr., p. 164) cites o[te a;rxhtai (Vettius, pp. 106, 36). [Otan, naturally occurs more frequently with the subjunctive for indefinite future time. It is usually the aorist tense, as in Mt. 24:33, o[tan i;dhte. The present subj. does occur when the notion of repetition is implied, as in Mt. 15:2, o[tan a;rton evsqi,wsin. Cf. Mt. 6:2. Once the idea of duration seems manifest (Jo. 9:5, o[tan evn tw|/ ko,smw| w=), but usually it is future uncertainty simply. It is not necessary to take the common aorist subj. here as the Latin futurism exactum.322 Cf. o[tan paradoi/ in Mk. 4:29. The a;n ( o[te a;n) is always present save in the doubtful o[te ei;phte of Lu. 13:35. [Ote with the subj. is found in poetry and in the Byzantine writers.323 So Test. XII Pat. Levi 2:10 o[te avne,lqh|j evkei/) On the other hand a number of examples occur of o[tan with the indicative (cf. eva,n and o[pou a;n) with the indicative). Homer, Iliad, 20, 335, has o[te ken xumblh,Ä seai auvtw|/. So in Rev. 4:9 we find o[tan dw,sousin. The close affinity in form and meaning of the aorist subj. with the future indicative should cause no surprise at this idiom. In Lu. 13: 28 BD read o[tan o;yesqe, though W. H. put, o;yhsqe in the text. A good many manuscripts likewise have o[tan with the future ind. in Mt. 10:19 and 1 Tim. 5:11. Cf. o[tan e;stai in Clem., Cor. 2, 12, 1. Moulton (Prol., p. 168) notes in the papyri only a small number of examples of a;n with temporal clauses and the ind. Thus o[tan e;bhmen in Par. P. 26 (ii/B.C.); evpa.n evpuqo,mhn in B. U. 424 (ii/iii A.D.); o`po,tan avnairou/ntai in B. U. 607 (ii/A.D.). It is common in the LXX, Polybius, Strabo, etc. See Jannaris, Hist. Gk. Gr., p. 463; Radermacher, N. T. Gr., p. 164. Ramsay (Cit. and B., ii, p. 477, no. 343) gives o[tan e;zwn evgw, a "curious anti-Christian inscription" (Moulton, Prol., p. 239). A few instances occur of o[tan, with the present indicative. So o[tan sth,Ä kete in Mk. 11:25. Here324 some MSS. have the subj., as in Ro. 2:14 some read o[tan poiei/. Cf. also various readings in Mk. 13:4, 7. This construction is not unknown in earlier writers, though more common in the koinh,. Cf. Ex. 1:16; Ps. 101:3;
MODE ( EGKLISIS) 973 Prov. 1:22; Josephus, Ant., xii, 2, 3; Strabo, I, 1, 7; Act. Apocr., 126. In 2 Cor. 12:10, o[tan avsqenw/, we probably have the present subj. Cf. 1 Th. 3:8, eva.n sth,kete. The examples of o[tan with the aorist or imperfect indicative are more numerous. In Thucydides o[te was always definite and o`po,te indefinite.325 [Otan, with the optative appears in Xenophon.326 The Atticists have evpeida,n and o`po,tan (sic) with the opt. (Radermacher, N. T. Gr., p. 165). In the koinh, the field of o[tan is widened, as already shown. Agathias uses o[tan with the aorist indicative.327 It is common in the Septuagint to have o[tan with past tenses (Gen. 38:11; 1 Sam. 17:34, o[tan h;rceto; Ps. 119:7, o[tan evla,loun; Num. 11:9; Ps. 118:32; Dan. 3:7).328 The usual notion is that of indefinite repetition. Thus we note it in Polybius 4, 32, 5, o[tan me.n ou-toi h=san evge,neto to. de,on. Strabo I, 1, 7 has o[tan fhsi,n. Cf. also 13, 7, 10. In Tobit 7:11 observe o`po,te eva,n) In Mk. 3:11 we have o[tan auvto.n evqew,rounà prose,pipton auvtw|/. Cf. o[pou a;n and o[soi a;n in Mk. 6: 56. But the koinh, writers used o[tan with the aorist indicative for a definite occurrence. This is common in the Byzantine329 writers. In the modern Greek o[tan is freely used with the indicative.330 See Philo II, 112, 23, o[tan eivj e;noia h=lqen. Blass331 calls this quite incorrect, though the LXX has w`j a'n evxh/lqen vIakw,b (Gen. 27:30; cf. 6:4) of "a single definite past action.332" There are two examples in the N. T., Mk. 11:19, o[tan ovye. evge,netoà evxeporeu,onto e;xw th/j po,lewj (possible to understand it as repetition), and Rev. 8: 1, o[tan h;noixen th.n sfragi,da th.n e`bdo,mhn. But, as Moulton (Prol., p. 248) observes, it is possible to regard evxeporeu,onto in Mk. 11: 19 as pictorial rather than iterative and the papyri examples of o[tan, as seen above, allow either usage. Simcox333 explains this "lapse" on the ground that Mark and the author of the Apocalypse are the least correct of the N. T. writers. But the idiom belonged to the vernacular koinh,. See Ex. 16:3, o;felon avpeqa,noÄ men- o[tan evkaqi,samen evpi. tw/n lebh,twn kai. hvsqi,omen a;rtouj) `Osa,kij is only used with the notion of indefinite repetition. It occurs
974 A GRAMMAR OF THE GREEK NEW TESTAMENT four times in the N. T. (1 Cor. 11:25 f.; Rev. 11:6), each time with eva,n and the subjunctive. These points are all obvious. `Wj is rather common in the N. T. as a temporal conjunction. It is originally a relative adverb from o[j and occurs in a variety of constructions. The temporal use is closely allied to the comparative. Cf. w`j evla,lei h`mi/n evn th|/ o`dw|/, (Lu. 24:32). So Jo. 12: 36. The temporal aspect is sharp in Mk. 9:21 where w`j means 'since.' The examples in the N. T. are usually in the aorist or imperfect indicative as in Jo. 6:12, 16; Ac. 8:36 and chiefly refer to definite incidents. In 1 Cor. 12:2, w`j a'n h;gesqe, we have the imperfect ind. with a;n for the notion of repetition (cf. o[tan). So in Aristeas 7, 34, w`j a'n hu;xanto. In modern Greek sa,n, (from w`j a;n) is used for 'when' (Thumb, Handb., p. 192). The use of w`j a;n= 'as if' is that of conditional, not modal, a;n, and is very common in the papyri (Moulton, Prol., p. 167). See Conditions., As early as i/B.C. the papyri show examples of w`j a;nÊo[tan (originally w`j a;n='as soon as'). Cf. Radermacher, N. T. Gr., p. 164;. Rhein. Mus., 1901, p. 206; Hib. P. I, 44, 45. Radermacher (N. T. Gr., p. 164) gives w`j a;n oi=mai, Dion. Hal. and Dio Chrys., w`j a'n a;meinon e;doxenà Luc. Alex. 22. But w`j is used a few times with the subjunctive, thrice with a;n (Ro. 15:24; 1 Cor. 11:34; Ph. 2: 23), once without a;n (Gal. 6:10), w`j kairo.n e;cwmen. In classical Greek this futuristic subj. would have a;n (Moulton, Prol., p. 248 f.). With the last construction compare Mk. 4:26. In the temporal use w`j a;n is not common in Attic. In Mk. 9:21 note po,soj cro,noj- w`j. In Ac. 17:15 we have w`j ta,cista, a remnant of the rather frequent use of w`j with superlative adverbs. It is possible that kaqw,j has a temporal sense in Ac. 7:17 (cf. 2 Macc. 1:31). (c) The Group Meaning 'Until' ('While'). The words in this list have a more complex history than those in the preceding one. They are a;crià me,crià e[wj and pri,n. ;Acri (twice in the N. T., a;crij, Gal. 3:19 and Heb. 3:13) is more frequently a preposition (cf. a;cri kairou/, Lu. 4:13) than a conjunction. It is rare in Greek prose and a;cri a;n only in poetry.334 But Philo (I, 166, 20) has a;crij a'n- sbe,seie. But the simple conjunction is less frequent than the compound form (preposition and relative), as a;cri ou- (Lu. 21:24) and a;cri h-j h`me,raj (Mt. 24:38). Sometimes the MSS. vary between a;crià me,cri, and e[wj, as in Mt. 13:30 (preposition). Cf. Ac. 1:22. Past tenses of the indicative are used of an actual historical event. No example of the simple a;cri ap-
MODE ( EGKLISIS) 975 pears in this construction in the N. T., but we have a;cri ou- avne,sth (Ac. 7:18) and a;cri h-j h`me,raj eivsh/lqen (Lu. 17:27). The only instance of the present ind. is in Heb. 3:13, a;crij ou- to. sh,meron kalei/tai. Here the meaning is 'so long' (linear) or 'while' (cf. e[wj). The more common use is with reference to the indefinite future. In two instances (Rev. 17:17, a;cri telesqh,sontai, and 2:25, a;cri ou- a'n h[xw) This latter could be aorist subj.) the future indicative is read. Elsewhere we meet the subjunctive, either without a'n $a;cri sfragi,swmen, in Rev. 7:3 and a;cri telesqh|/ in 20:3, 5; a;cri ou- e;lqh| in 1 Cor. 11:26; a;cri h-j h`me,raj ge,nhtai in Lu. 1:20) or with a;n ( a;crij a'n e;lqh| in Gal. 3:19, though W. H. put just a;crij ou- in the margin). Here the time is relatively future to the principal verb prosete,qh, though it is secondary. The subj. is retained instead of the optative on the principle of indirect discourse. As a matter of fact a;n occurs only twice, the other instance being Rev. 2:25 above. Cf. a;crij o[tan plhrwqh|/, 0. P. 1107, 3 (v/A.D.). Me,crij (so twice, Mk. 13:30; Gal. 4:19, and once me,cri, Eph. 4:13) occurs only three times as a conjunction. In Eph. 4:13 it is me,cri simply, in the other examples me,crij ou-. In all three instances the aorist subj. is used without a;n, for the indefinite future. The use as a preposition is more frequent. Cf. me,cri vIwa,nou (Lu. 16:16) and me,crij ai[matoj (Heb. 12:4). It means 'up to the point of.'335 The koinh, writers show a rather varied use of me,cri (cf. Diodorus, Strabo, Polybius, Josephus, Justin Martyr). They, like the papyri, have me,cri and me,crij ou- with and without a;n, (Radermacher, N. T. Gr., p. 140). [Ewj is much more frequent in the N. T. both as preposition (cf. e[wj o[tou, Mt. 11:23) and as conjunction. The prepositional use is illustrated also in e[wj tou/ evlqei/n (Ac. 8:40). The prepositional use (more frequent than the conjunctional) goes back as far as Aristotle and denotes the terminus ad quem. [Ewj is Attic for Homeric h-oj and Doric a[j.336 As with a;cri and me,cri, we find e[wj alone as a conjunction (Mt. 2:9), e[wj ou- (Mt. 14:22) and e[wj oovranou/ grk(5:25). It is used both with the indicative and the subjunctive. When an actual event is recorded in the past only the aorist indicative is used. This is the usual classic idiom.337 So e[wj h=lqen (Mt. 24:39), e[wj ou- e;teken (1: 25), e[wj o[tou evfw,nhsan, (Jo. 9:18). When the present ind. appears with e[wj the notion is 'while,' not 'until,' and it is either a contemporaneous event, as in e[wj auvto.j avpolu,ei to.n o;clon (Mk. 6:45.
976 A GRAMMAR OF THE GREEK NEW TESTAMENT Note dependence on hvna,gkasen, like indirect discourse), or a lively proleptic future in terms of the present, as in e[wj e;rcomai pro,sece th|/ avnagnw,sei (1 Tim. 4:13) and in Jo. 21:22 f. It is possible to take Mk. 6:45 as this proleptic future.338 Indeed some MSS. here give also avpolu,sh| and - ei. In Mt. 14:22 the reading (in the parallel passage) is e[wj ou- avpolu,sh|. Cf. the construction with the Latin dum. In Lu. 19:13 W. H. read evn w|- e;rcomai instead of e[wj e;rcomai. Instead of e[wj h`me,ra evsti,n (Jo. 9:4) W. H. have w`j in the margin, though keeping e[wj in text (as does Nestle). If e[wj is genuine, it is clearly 'while,' not 'until.' In Jo. 12:35 f. W. H. read in the text an, not gun. We have, besides, e[wj o[tou ei= in Mt. 5:25. Most of the examples of e[wj deal with the future and have only the subj. after the classic idiom.339 The future, being identical in form with the aorist subj., is possible in the cases of e[wj ou- avnape,myw (Ac. 25:21) and e[wj o[tou ska,yw (Lu. 13:8), but the regular subj. is the probable idiom. In Lu. 13:35 some MSS. have e[wj h[xei (see (b)), but W. H. reject h[xei o[te) Both e[wj ou- and e[wj o[tou are used, but always without a;n. So e[wj ou- avne,lwsin (Ac. 23:21) and e[wj o[tou plhrwqh|/ (Lu. 22:16). With simple gun it is more common to have a;n. So e[wj a'n a`podw|/j (Mt. 5:26), but note e[wj e;lqh|grk grk(10:23). ;An is not essential in this construction. Cf. Lu. 12:59; 15:4; 22:34. In Mk. 14:32, e[wj proseu,xwmai, the notion is rather 'while' than 'until.' Cf. Mt. 14:22; 26:36; Lu. 17:8. But the note of expectancy suits the subjunctive. In Mt. 18:30, e;balen auvto.n eivj fulakh.n e[wj avpodw|/ to. ovfeilo,menon, the subj. is retained after secondary tense of the indicative as in indirect discourse. [Ewj occurs after negative verbs also (cf. pri,n), as in Lu. 22:34. Moulton (Prol., p. 169) quotes Tb. 6 (ii/B.C.) e[wj me,nwsin, G. H. 38 (i/B.C.) e[wj katabh|/j. In the papyri a;n, as in the N. T., is often absent from these conjunctions meaning 'until.' Radermacher (N. T. Gr., p. 140) finds e[wj and the subj. common in the papyri, the inscrs. and the koinh, writers. Blass340 thinks he sees a certain affinity with final sentences in the subj. with these conjunctions for the future indefinite. At any rate it is good Attic and should cause no trouble. The koinh, fully agrees with the ancient idiom. It is, of course, a matter of taste with the writer whether he will regard a future event as a present reality or a future uncertainty to be hoped for and attained. Pri,n is a comparative form (cf. superlative prw/Ätoj) like the Latin
MODE ( EGKLISIS) 977 prius.341 It is the neuter accusative singular. It is really the same in idea as pro,teron, 'before,' 'formerly.' Pindar uses it as a preposition with the ablative pri.n w[rajÊpro. w[raj. The original construction with pri,n, was the infinitive, though the subj. and the optative occur with it in Homer.342 Homer has it 81 times with the infinitive, 6 with the subj., once with the opt. and not at all with the indicative.343 The word developed so much importance in the later Greek that Goodwin in his Moods and Tenses gives it a separate extensive discussion (pp. 240-254). In the N. T. there are only fourteen examples of it and all of them in the Gospels and Acts. Eleven of the fourteen are with the infinitive (cf. Homer). Cf. pri.n avpoqanei/n (Jo. 4:49), pri.n vAbraa.m gene,sqaigrk grk(8:58). Six times we have pri.n h;, as in Mt. 1:18. Luke alone uses the classic idiom of pri,n with the subj. or opt. after negative sentences. In both instances it is only relative future after secondary tenses, but in Lu. 2:26, mh. ivdei/n qa,naton pri.n [ h'] a'n i;dh| to.n Cristo.n kuri,ou, the subj. is retained according to the usual rule in indirect discourse in the koinh, (so often in the Attic). In Ac. 25:16, as already explained (p. 970), pri.n h' e;coiÄÄla,boi after avpekri,qhn o[ti ouvk e;stin, is changed from the subj. to the opt. as is possible in indirect discourse, a neat classic idiom found in Luke alone in the N. T. Some of the MSS. do not have ay in Lu. 2:26 and reads e[wj a'n here. A few MSS. have pri,n h; in Lu. 22:34.344 The papyri writers do not show the same consistency as Luke in the use of pri,n.345 But note mh,te dido,tw- pri.n auvtw|/ evpiste,llhtai., 0. P. 34 (ii/A.D.). For 'until' e[wj kept the field. Indeed in Lu. 22:34, ouv fwnh,sei sh,meron avle,ktwr e[wj tri.j avparnh,sh|, we see e[wj where pri,n would usually come (Radermacher, N. T. Gr., p. 164). Very early pro. tou/ and inf. also began to displace pri,n (see Verbal Nouns). In the modern Greek pri,n holds its place (also pri. na,à o[soà protou/) with ind. and subj. (Thumb, Handb., p. 193). The N. T. does not have e;ste, but the papyri show it. Cf. e;st v a;n, Amh. P. II, 81, 11 (iii/A.D.). See also Job 13:22 a. (d) Some Nominal and Prepositional Phrases. We have already seen in the case of a;crià me,ri and e[wj how they occur with relative pronouns as conjunctional phrases. The same thing occurs with a number of temporal phrases. Thus avf v ou-. In Lu. 13:7 avf v ou- is preceded by tri,a e;th as the terminus a quo. It
978 A GRAMMAR OF THE GREEK NEW TESTAMENT means 'since.' Cf. tri,thn tau,thn h`me,ran a;gei avf v ou- in Lu. 24:21. In Rev. 16:18 it is the simple equivalent of avpo. tou,tou o[te as in the Attic Greek and Herodotus. In these examples the indicative occurs, but in Lu. 13:25, avf v ou- a'n evgerqh|/, the construction of e[wj is used for the uncertain future, the subj. with a;n. The conception of avpo. tou,tou o[te has to be appealed to, 'from that moment when,' 'when once' the master arises. In like manner we see avf v h-j used for 'since' in Lu. 7:45; Ac. 24:11; 2 Pet. 3:4. In Col. 1:6, 9 we have the form avf v h-j h`me,raj. vEn w|- is not always temporal. It may be merely local (Ro. 2:1), instrumental (Ro. 14:21) or causal (Ro. 8:3). The temporal use is much like e[wj in the sense of 'while,' as in Mk. 2:19 (Lu. 5:34) evn w|- o` numfi,oj met v autvw/n evsti,n. Cf. Jo. 5:7, evn w|- e;rcomai with e[wj e;rcomai in Jo. 21:22. In Lu. 19:13 the Text. Rec. has e[wj e;rcomai, but evn w|- is the true reading. In 1 Pet. 1:6 evn w|- has its antecedent expressed in the preceding sentence and means 'wherein.' In Mk. 2:19 we see o[son cro,non for duration of time. In Mt. 9: 15 the shorter evf v o[son occurs, while in Heb. 10:37 note o[son o[son (a Hebraism from the LXX, though paralleled in the papyri). In Ro. 7:1 we read evf v o[son cro,non the fullest form of all. Moulton (Prol., p. 169) cites C.P.R. 24, 25 (ii/A.D.) evf v oa}n h|= cro,non (note absence of a;n). (e) The Temporal Use of the Infinitive. There are nine examples of pro. tou/ and the infinitive. In the LXX there are 35 examples (Votaw, The Infinitive in Bibl. Gk., p. 20). These examples all have the accusative with the infinitive, as in pro. tou/ u`ma/j aivth/sai auvto,n (Mt. 6:8. Cf. Lu. 2:21; 22:15; Jo. 1:48 f.; 17:5; Ac. 23: 15; Gal. 2:12; 3:23), except Jo. 13:19, pro. tou/ gene,sqai, but even here it is implied. The tense is aorist except a present in Jo. 17:5. The sense is quite like pri,n (see before). The inscriptions (Moulton, Prol., p. 214) show scattered examples of pro. tou/ and inf. The use of evn tw|/ as 'when' or 'while' is much more common. It occurs only 6 times in Thucydides, Plato 26 times, Xenophon 16 times.346 But it is very common in the Septuagint as a translation of the Hebrew B. and the infinitive construct. Moulton347 admits a Hebraism here in the sense of 'during,' a meaning not found in the vernacular koinh, so far. The construc-_ tion is, however, very common in Luke, the most literary of the N. T. writers, and in all parts of his Gospel. It is found both in the sense of 'while' and 'when.' Usually it is the present tense that has the notion of 'while' and the aorist that of 'when.' So
MODE ( EGKLISIS) 979 in Lu. 1:8 note evn tw|/ i`erateu,ein auvto,n,grk grk(2:27) evn tw|/ eivsagagei/n tou.j gonei/j to. paidi,on vIhsou/n) The examples are numerous (55 in the N. T.), but the LXX shows 500 instances,348 undoubted proof of the influence of the Hebrew there, where it is nearly as common as all other prepositions with the infinitive. This use of evn tw|/ and the infinitive is not always temporal. In Lu. 12:15 it is rather the content than the time that is meant, In Lu. 1:21 it may be causal. Meta. to, and the infinitive we find fifteen times in the N. T. In the LXX the construction appears 108 times according to Votaw.349 It has the resultant meaning of 'after' and always has the aorist infinitive except the perfect in Heb. 10:15. It is found in Luke, Paul, Matthew, Mark, Hebrews, and chiefly in Luke. A good example is found in meta. to. avpoÄ ktei/nai (Lu. 12:5). See also Ac. 7:4; 10:41. Mention should also be made of e[wj tou/ evlqei/n in Ac. 8:40, as in the LXX (Judith 1:10; 11:19). It occurs 52 times in the O. T. and 16 in the Apocrypha. But note me,cri tou/ plei/n , P. B. M. 854 (i/A.D.). On prepositions and inf. see Verbal Nouns. (f) Temporal Use of the Participle. This subject will demand more extended treatment under the head of the Participle (Verbal Nouns). Here it may be noted that the participle does not of itself express time. We may in translation render the participle by a temporal clause with 'as,' 'while,' 'since,' 'when,' 'after,' etc., like the Latin cum.350 As a rule the unadorned participle in English is enough to bring out the idea. The participle may be co-ordinated in translation with the principal verb by the use of 'and.' The present participle is merely descriptive and contemporaneous, as avpoqnh,skwn (Heb. 11:21). The aorist participle has either simultaneous action, as avspasa,menoi (Ac. 25:13), or antecedent, as evm ba,nta (Mt. 13:2). The wealth of participles gave the Greek a great advantage over the Latin in this matter. In the flourishing period of the language the temporal participle vied with the conjunctions in the expression of temporal relations. In the koinh, this use of the participle is still quite live, as almost any page of the N. T. shows, though it has manifestly in places shrunk before the analytic tendency to use conjunctions and finite verbs. This tendency to use conjunctions is still more noticeable in modern Greek.351
980 A GRAMMAR OF THE GREEK NEW TESTAMENT 6. FINAL AND CONSECUTIVE CLAUSES. (a) Kinship. It is a difficult matter to correlate properly these subordinate clauses. They nearly all have relative adverbs as conjunctions. Often the same conjunction is used indifferently in a number of different kinds of clauses. So w`j in comparative, declarative, causal, temporal, final, consecutive, indirect interrogative, exclamatory. In like manner o[pwj has a varied use. Cf. the Latin ut, which is comparative, final, apprehensive, consecutive. The English that and German dass have a like history. Goodwin,352 therefore, treats "final and object-clauses" together as pure final clauses, object-clauses with verbs of care and effort, clauses with verbs of fearing. He gives a separate discussion of consecutive clauses.353 Burton354 practically follows Goodwin. Viteau355 blends them all into one. Winer practically ignores consecutive clauses. Jannaris356 pointedly says that the popular speech "avoids the consecutive construction" and uses w[ste and the infinitive for either final or consecutive (cf. Latin ut and English that) "thus confounding consecutive with final clauses." It was not quite that. As a matter of fact the various points of view shade off into one another very easily and sometimes quite imperceptibly. It is not always easy to distinguish purpose and result in the mind of the writer or speaker. The very word finis may be the end aimed at (purpose) or attained (result). My colleague, Prof. W. 0. Carver, D.D., has suggested grouping these ideas all under result, either contemplated, feared or attained. Some such idea is near the true analysis and synthesis. The later Greek showed a tendency to gather most of these ideas under i[na.357 (b) Origin in Parataxis. It seems clear that these final clauses had their origin in parataxis, not hypotaxis. The conjunctions, when used, were an after-development. The step from parataxis to hypotaxis has already been taken when we meet the Greek of Homer,358 though the paratactic construction continued side by side in isolated instances. Examples like a;fej evkba,lw (Lu. 6:42), bou,lesqe avpolu,sw (Jo. 18:39), qe,leij e`toima,swmen (Mk. 14:12) are probably instances of this original idiom rather than of a mere ellipsis of i[na.359 Cf. also the possible origin of ouv mh, as ou;\ mh,) This
MODE ( EGKLISIS) 981 disconnected idiom was felt to be especially bare in the positive form, but the negative paratactic construction with mh, with verbs of fearing is present in Homer.360 Gildersleeve361 quaintly says: "Parataxis, which used to be thrust into the background, has come forward and claimed its rights." This grammatical sage, barring the infinitive and participle, adds: "Nihil est in hypotaxi quod non prius fuerit in parataxi." The subjunctive, therefore, in final clauses is merely the volitive subj. of parataxis.362 It was natural that the parataxis should be plainer in negative sentences, for alongside of mh, (originally the mere negative in parataxis and the negative conjunction in hypotaxis) there came i[na mh,Ã o[pwj mh,.363 The whole matter is carefully worked out by Weber364 with careful discussion of each construction in the various writers during the long course of Greek linguistic history from Homer through the Attic writers. (c) Pure Final Clauses. Here conscious purpose is expressed. This class constitutes the bulk of the examples and they are the easiest to understand. The Greek is rich in variety of construction for this idea. We can deal only with the idioms in the N. T. ;Ofra is not in the N. T. or LXX, nor is the idiom of o[pwj with the future indicative after verbs of striving. (a) [Ina. The etymology of i[na is not certain. A fragment365 of Hesiod has i[n auvtw|/. Perhaps i[nÄa is derived from this form. But at any rate in Homer i[naÊ evkei/ in Iliad, 10, 127. After Homer, especially in the poets, it has the meaning 'where,' 'in what place,' 'whither.'366 The exact connection between this local demonstrative and relative sense and the final 'that' (ut) is not clear.367 But we have a similar transition in the Latin ut, English that, German dass. Sophocles in his Lexicon of the Roman and Byzantine Periods gives nineteen uses of i[na for the Greek of that era. They may all be whittled down to three, viz. the pure final, the object-clauses or sub-final, the consecutive. There is no doubt that i[na came to be used in all these ways in the Byzantine period. In the koinh, of the N. T. time the first two are abundantly shown. The ecbatic or consecutive use is debatable in the N. T. But each in its order. Curiously enough the Attic inscriptions make a very sparing use
982 A GRAMMAR OF THE GREEK NEW TESTAMENT of i[na, much preferring o[pwj and o[pwj a;n.368 So in epic and lyric poetry i[na is overshadowed by o;fra and in tragedy by w`j, though Aristophanes uses it in three-fourths of his final sentences and Plato and the Attic orators use it almost exclusively (Goodwin, Moods and Tenses, p. 109). The original use of i[na, after the demonstrative and the relative, stage, was the pure final. It is so in Homer, though Monro admits one instance of the object-clause.369 Only the subj. occurs with it in Homer in this construction. This is the natural mode for the expectant note in clauses of purpose.370 But it must not be overlooked that i[na in no way controls the mode, for the idiom is at bottom paratactic in origin.371 But the indicative had a use also as well as the optative, as will presently be shown. A word further is needed concerning the tremendous development in the use of i[na. Thucydides used o[pwj three times as often as i[na, and w`j as a final particle only twice. Xenophon in the first three books of the Anabasis has o[pwj one and a half times as often as i[na, and w`j nearly as often as i[na. But Polybius (books I-V) uses i[na exclusively, and the N. T. has i[na about twelve times as often as o[pwj and w`j perhaps once. It is thus not simply that i[na displaced o[pwj and w`j, but it gradually usurped the final use of the infinitive also. It comes to be almost the exclusive means of expressing purpose, and in the modern Greek vernacular every phase of the subj. and the old future ind. can be expressed by na, ( i[na) and the subj.372 Na, is used also with the ind. The intention in modern Greek is brought out a bit more sharply by gia. na, (Thumb, Handb., p. 197). But the distinction is sometimes faint. All in all it is one of the most remarkable developments in the Greek tongue. The eight and a half pages of examples in Moulton and Geden's Concordance bear eloquent testimony to the triumph of i[na in the N. T. Nearly a page and a half of these examples are in the Gospel of John. But we are now specifically concerned with the pure final use of i[na. Here i[na is in the accusative case of general reference. Thus in evlh,luqa i[na ma,qw (cf. veni ut discam, 'I am come that I may learn') i[na is really a demonstrative. 'I am come as to this,' viz. 'I may learn.' The conjunction is supplied to avoid the asyndeton and is in apposition with ma,qw. As already explained, the subj. is the predominant mode, as in tou/to de. o[lon ge,gonen i[na plhrwqh|/ (Mt. 1:
MODE ( EGKLISIS) 983 22). Cf. Ph. 3:8. The negative with i[na is mh,, as in i[na mh. kriqh/te (Mt. 7:1). The aorist subj. is the normal tense, of course, as in i[na metadw/ (Ro. 1:11), though the present occurs to denote a continuous action, as in i[na pisteu,hte (Jo. 13:19). Cf. i[na gnw/te kai. ginw,skhte (Jo. 10:38). The perfect subj. occurs in eivdw/Ã as i[na eivdh|/j (1 Tim. 3:15); i[na eivdw/men (1 Cor. 2:12); i[na eivdh/te (1 Jo. 5:13). Cf. also Jo. 17:19, 23; 1 Cor. 1:10; ,2 Cor. 1:9 ( i[na mh. pepoiqo,tej w=men); i[na pareskeuasme,noi h=te (2 Cor. 9:3). The subj. is regularly retained after a secondary tense of the indicative as in avne,bh i[na i;dh| (Lu. 19:4); evpeti,mhsen i[na mhdeni. ei;pwsin (Mt. 16:20). Cf. Mk. 8:6. There is no instance in the N. T. of the optative used with i[na after a secondary tense of the indicative. It is true that W. H. read i[na dw|,h in the text of Eph. 1:17 ( i[na dw,h| or dw|/ in the margin), but this is after a primary tense, ouv pau,omai. It is the volitive use of the optative and is not due to i[na. It is like the optative in a future wish.373 This use of the opt. with i[na after a wish is not unknown to classic Greek.374 It is the subj., not the opt., that is seen in i[na plhroi/j (Col. 4:17), i[na paradoi/ (Mk. 14:10) and in the sub-final i[na ovnoi/ (Mk. 9:30).375 In Homer and the early writers generally the rule was to use the opt. with the final clauses after secondary tenses, but in the Attic orators the two modes (subj. and opt.) are on a par in such a construction, while Thucydides prefers the subj., though Xenophon is just the reverse.376 In the N. T. the optative in final clauses after secondary tenses is non-existent. In 2 Tim. 2:25 mh, pote dw|,h is after a primary tense as in Eph. 1:17, and here again the text is uncertain (cf. dw,h| in margin and avnah,ywsin in text.) The Atticists (Arrian, Appian, Herodian, 4th Macc., Plutarch) made a point of the opt. with i[na as "the hall-mark of a pretty Attic style" (Moulton, Prol., p. 197). The N. T. writers, more like Diodorus and Polybius, fail "to rival the litterateurs in the use of this resuscitated elegance." Moulton speaks also of "the
984 A GRAMMAR OF THE GREEK NEW TESTAMENT riot of optatives" in the artificial Byzantine writers. On the whole subject of final clauses see Gildersleeve on "The Final Sentence in Greek," 1883, p. 419, A. J. of Philol., IV, pp. 416 ff., VI, pp. 53 There is no trouble to find in the papyri, inscr. and koinh, writers generally abundant examples of i[na and the subj. in pure design (Radermacher, N.T.Gr., p.138). But while the subj. is the normal construction, the indicative is also present. In classical Greek i[na was not used with the future ind.377 It was not common even with o[pwjà w`j and mh,. The similarity in form and sense (not to mention itacism of - h| and - ei) made the change very easy and, indeed, the text is not always certain as between the aorist subj. and the future ind. Thus in 1 Cor. 13:3 i[na kauch,Ä swmai is supported by aAB, i[na kauqh,swmai by CK and i[na kauqh,Ä somai by late documents.378 In Gal. 2:4 the best documents have i[na katadoulw,sousin instead of - swsin. In Jo. 17:2 the MSS. vary between i[na dw,sei and dw,sh|. So in Jo. 15:8 note i[na fe,rhte kai. ge,nhsqe ( genh,sesqe in margin of W. H.); Eph. 6:3, i[na ge,nhtai kai. e;sh|. But the idiom is well established in the N. T., especially in the Apocalypse. Thus i[na qewrh,sousin (Jo. 7:3); i[na xurh,sontai (Ac. 21:24); i[na evrei/ (Lu. 14:10); i[na qh,sw (1 Cor. 9:18); i[na dw,sousin (Lu. 20:10); i[na kenw,sei (1 Cor. 9:15); i[na kerdhqh,sonÄ tai (1 Pet. 3:1); i[na sfa,xousin (Rev. 6:4); i[na dw,seigrk grk(8:3); i[na h[xousin- gnw/singrk grk(3:9); i[na e;stai kai. eivse,lqwsingrk grk(22:14), etc. This last example may be non-final. In some of these examples the subj. and ind. future occur side by side. In Mk. 6:56 and Ac. 5:15 note i[na ka;n (only instances of a;n with i[na in the N. T.). This is not modal a;n, but ka;n as 'even' = kai, (Jannaris, Hist. Gk. Gr., p. 165; Moulton, Prol., p. 167). In Rev. 13:15 the MSS. vary between i[na poih,sh| and - ei, and in 16 between i[na dw/sin and dw,sei ( poiei/ i[na sub-final). The usage is thus on a firm foundation in the N. T. It is in the LXX also. See i[na e;stai in Lev. 10:6 and in other writers of the koinh, (Iren., 584 A, i[na e;sh|).379 But i[na occurs also with the present ind. This is a rare construction in the N. T. and is not a classic idiom. It occurs only three times in the N. T. Thayer calls it "a solecism frequent in the eccl. and Byzantine writers." It is so common in late writers as not to surprise us in the N. T.380 Thus 1 Cor. 4:6 i[na mh. fusiou/sqe, Gal, 4:17 i[na zhlou/te and 1 Jo. 5:20 i[na ginw,skomen. The first two are possible subjunctives. W. H. read i[na mh,tij du,natai in the margin of Rev. 13:17, and various
MODE ( EGKLISIS) 985 MSS. support the present ind. with i[na in Jo. 4:15; 5:20; 17: 3; Gal. 6:12; 1 Th. 4:13; Tit. 2:4; 2 Pet. 1:10; Rev. 12:6.381 In the earlier Greek writers we do find i[na used with past tenses of the indicative.382 The idea was to show that the purpose was dependent on an unfulfilled wish or unattained action. But this refinement does not appear in the N. T. except in two examples with mh, pwj. With all the wide extension of i[na in Western Hellenistic,383 at the heart of it there is the pure telic idiom. [Ina with the imperative in 1 Cor. 1:31 is due, of course, to the quotation. [Ina is repeated three times in 2 Cor. 12:7. In Jo. 11:37, poih/sai i[na kai. ou-toj mh. avpoqa,nh|, one is reminded of the Latin facere ut (sub-final). Westcott (Hebrews, p. 342 f.) gives a list of all the examples of i[na in the Epistle (20). Only two of o[pwj. ( b) [Opwj. It is compounded of the neuter accusative relative o[ and the indefinite adverb pw,j.384 It occurs in indirect questions as in Lu. 24:20 in the sense of 'how,' the usual interrogative sense, and note article also as in to. pw/j (Lu. 22:2). [Opwj in a sense is the connecting link between the various kinds of final sentences.385 Thucydides and Xenophon preferred o[pwj to i[na, and Aristotle has i[na only a few times (W. Schmid, Atticismus, III, p. 87). Polybius does not use o[pwj at all in books I-V. The N. T. has i[na 493 times, o[pwj 52 (Jannaris, p. 417) as far as Colossians. Scott counts i[na 746 times in text of W. H. (not including 6 of i[na ti,-) and 58 of o[pwj. Thumb does not give o[pwj as a final particle in modern Greek (Handb., p. 197). Even in later Greek o[pwj was a sign of literary affectation.386 As already noted, in the fourth and fifth centuries B.C. o[pwj was quite the rule in the Attic inscriptions.387 It is rare in Homer and never has ke, or a;n in pure final clauses in the Homeric language.388 This idiom with a;n first appeat's in AEschylus. In the great Attic writers and the Attic inscriptions the subjunctive, the future indicative and the optative after secondary tenses, all are found. The future indicative occurred chiefly with verbs of striving, though sometimes in pure final clauses.389 The negative with this future indicative was mh, ( o[pwj mh,%Ã though no example
986 A GRAMMAR OF THE GREEK NEW TESTAMENT occurs in the N. T. Moulton (Prol., p. 177 note) finds in the papyri a few survivals of o[twj mh, and the fut. ind., though mostly ousted by i[na mh,. Cf. Hb. P. 45, 60, 168 (iii/B.C.), Tb. P. 414 (ii/A.D.). Stahl (Syntax, p. 360) calls o[pwj and fut. ind. Attic. In the N. T. the optative does not occur in this construction. In the Atticists it is revived as with i[na.390 The fut. ind. with o[pwj in pure final clauses has practically vanished from the N. T. The one example in Ro. 3:4, o[pwj a;n dikaiwqh|/j kai. nikh,seij is a quotation from the LXX (Ps. 51:6), but changed from subj. there. But o[pwj qanatw,sousin is a variant reading in Mt. 26:59, and the future ind. is possible in Mt. 2:8, o[pwj proskunh,sw, though it is probably the aorist subj. Other variant readings where the future ind. is supported with o[pwj are 1 Cor. 1:29, kauch,setai, and Mk. 5:23 o[pwj zh,setai, (here W. H. read i[na zh,sh|). But at any rate the use of the future ind. with o[pwj in pure final clauses is not quite dead in the N. T. period, though surely dying. Elsewhere the aorist subj. alone occurs save in Lu. 16:26 (bis), 28 and Mt. 6:4. [Owpj no longer391 has a;n in final clauses save in the quotation from Ps. 51:6 (Ro. 3:4) and three passages in Luke's writings (Lu. 2:35 o[pwj a'n avpokalufqw/sin Ac. 3:19 f. o[pwj a'n e;lqwsinÄÄkai. avpostei,lh|, 15:17 o[pwj a'n evkzhth,swsin from Amos (so A, but B without a;ngrk) 9:12). ;An is a variant reading in Mt. 6:5 and is found very often in the LXX. Radermacher (N. T. Gr., p. 158) finds o[pwj a;n in Diodorus XIV, 80, 8, Aristeas, § 239, inscr. of Halicarnassus (iii/B.C.), Jahrb. d. Ost. Inst. XI, 56. But it is rare and o[pwj steps into the background before i[na. The revival of o[pwj in the third and fourth cent. A.D. was Atticistic and did not affect, the vernacular. The inscriptions and the papyri for the first century A.D. show the prevalence of i[na over o[pwj (Radermacher, N. T. Gr., p. 157 note). The negative is, of course, always mh,, as in Ac. 20:16, o[pwj mh. ge,nhtai. The subj. is used indifferently after primary tenses (Mt. 6:2, poiou/sin o[pwj doxasqw/sin) and secondary tenses (Ac. 9:24, parethrou/nto o[pwj auvto.n avne,lwsin). Cf. Ro. 9:17. It is interesting to note that in the N. T. o[pwj is almost confined to Matthew and Luke's writings. The literary flavour of Luke explains his use of the idiom, but we do not look for literary ear-marks in Matthew. The one example in John John(11:57) occurs side by side with i[na ( i[na mhnu,sh|Ã o[pwj pia,swsin) and may be used for the sake of variety as in i[na ge,nhtai o[pwj ge,nhtai, (2 Cor. 8:14). Cf. also Lu. 16:28;
MODE ( EGKLISIS) 987 1 Cor. 1:29; 2 Th. 1:12, though i[na- i[na appear in 1 Cor. 4:6; Gal. 4:5.392 In 1 Cor. 1:17 note i[na mh, and o[pwj mh, 1: 29. But i[na has "invaded the territory of o[pwj, as with fronti,zein and spouda,zein" (Moulton, Prol., p. 206). In modern Greek o[pwj has lost all telic force (Thumb, Handb., p. 198). Sometimes o[pwj represents the main purpose and the infinitive the subordinate purpose, a construction amply illustrated in the papyri.393 So then, though o[pwj as a pure final conjunction is disappearing in the N. T., it yet occurs with the same concept on the whole. ( g) `Wj. It was not a favourite final particle with Thucydides (only twice), though Xenophon used it nearly as much as i[na. It is not surprising to find only one instance of it in the N. T. and that one not certain. aB read w`j teleiw,sw in Ac. 20:24 instead of w`j teleiw/sai (cf. Lu. 9:52). W. H. and Neste read teleiw,sw, but Souter (Rev. V.) gives teleiw/sai. It is the last leaf on the tree and a fluttering one at that. The form could be the future incl. or aorist subj. Radermacher (N. T. Gr., p. 158) finds final w`j merely a reminiscence in the koinh,, but it is needless to cite Mk. 4:26 f., w`j a;nqrwpoj ba,lh|, since this is not final at all, but comparison. On w`j a;n in final sentences see Schmidt, Joseph. Bloc., p. 409, for statistics. Radermacher quotes F. P. 118 (110 A.D.), poreu,ouÄÄe[wj to.n evkei/ evlaiw/na poti,sh|j, where e[wj is used as final w`j. Per contra in modern Greek, Moulton (Prol., p. 249) notes that w`j takes the meaning of e[wj as well as its own. ( d) Mh,à mh, poteà mh, pwj. Negative purpose is expressed by i[na mh,à o[pwj mh, also, but originally it was done merely by mh, in a paratactic sentence.394 In Homer and the early writers mh, is far in excess of i[na mh,à o[pwj mh,, but in Aristophancs and Herodotus the reverse is true, while in Plato and Xenophon mh, as a final conjunction has about gone. It is rare in the Attic historians and orators generally.395 Originally a negative adverb (subjective negative) it came to be used also as a conjunction. Cf. Latin ne. The idioms mh. ouv appears in Homer in a few final clauses, and after Homer mh. ouv is used with verbs of fearing.396 In the N. T. i[na mh, (1 Cor. 1:17) and o[pwj mh, grk(1:29) have the run over the conjunction mh,) Only the subj. is used, though in Ac. 27:42 mh, tij diafu,goi is a variant reading, but diafu,gh| is correct after the secondary tense of the incl. In Mk. 13:36, mh. eu[rh| a primary tense occurs in the principal verb. In Col. 2:4 W. H. read i[na mhdei.j 988 A GRAMMAR OF THE GREEK NEW TESTAMENT paralogi,zhtai instead of mh, tij (the variant reading). See also mh, tij logi,shtai (2 Cor. 12:6). Both mh, and mh, pwj are preserved as final conjunctions in the modern Greek (Thumb, Handb., p. 198). The use of mh. pote and mh, pwj is practically the same. Mh, pwj appears with the subj. (Paul) after (secondaryand primary tenses. So e;pemya mh, pwj kataiscunqw/men (2 Cor. 9:3 f. Note also i[na mh, in 9:3, 4) and mh, pwj ge,nwmai (1 COr. 9:27). In Gal. 2:2 $mh, pwj e;dramon) and 1 Th. 3:5 ( mh, pwj evpei,rasen) we have a difficult construction. One view is to take it as an indirect question. This is possible in Gal. 2:2, but not in 1 Th. 3:5. Even in Gal. 2:2 there would be an ellipsis of a participle like zhtw/n maqei/n) Moulton (Prol., p. 201) suggests that e;dramon as an "after-thought" in Gal. 2:2 has plenty of classical Parallels. Cf. Goodwin, Moods and Tenses, § 333. In 1 Th. 3:51 we have mh, pwj evpei,rasen kai. ge,nhtai side by side. It is better therefore to take tre,cw in Gal. 2:2 as subj. also. Thus in both examples we have the subj. and the aorist ind. This is in accord with the ancient idiom where in pure final sentences a past tense of the incl. was used if it is distinctly implied that the purpose was not attained.397 That is precisely the case here. Paul did not run in vain. The tempter did not succeed with the Thessalonians. It is thus unfulfilled purpose that Paul neatly expresses in accord with the Attic diction. Mh, pote loses the notion of time in pote and has rather the idea of contingency, 'but perchance' rather than 'lest at any time.' Radermacher (N. T. Gr., p. 158) thinks that pote, and pw,j often distinguish deliberative (dubitative) from final mh,. As a strictly final particle it occurs either with the subj. or the future ind., though the subj. is more common.398 For the fut. ind. note Mt. 7:6 mh, pote katapath,sousin (correct text, though the aorist subj. has support), Mk. 14:2 mh, pote e;stai. In Lu. 12:58 note mh, pote katasu,rh| kai. avpodw,sei. Both subj. and fut. ind. likewise occur in Mt. 13:15 (Ac. 28:27) mh, pote i;dwsinÄÄkai. iva,somai (LXX, Is. 6:10). So also in Lu. 14:8 f., mh, pote h|= keklhme,noj (note perfect subj.) kai. evrei/ (cf. i[na evrei/ in verse 10). The normal subj. is seen in Lu. 14:12, mh, pote avntikale,swsin. The opt. in the N. T. is wanting in final sentences as in .Lses of repetition (Radermacher, N. T. Gr., p. 131). W. H. read mh, pote dw|,h (opt.) in 2 Tim. 2:25. But even so, if true, it not a pure final clause but a kind of indirect question as in Lu. 3:15, only in 2 Tim. 2:25 the opt. occurs after a primary tense. It is hardly just to say
MODE ( EGKLISIS) 989 with Moulton399 that here Paul "misused an obsolete idiom," since the opt. after primary tenses occurs occasionally with i[na in the papyri.400 Cf. mh, pote auvtw/n crei,a ge,noitoà euvqe,wj auvtou.j evxe,laÄ son, P. Oxy. I, 118, 38. But it is more than likely, as Moulton argues, that in 2 Tim. 2:25 we should read subj. dw,h|, since avnah,Ä ywsin undoubtedly is subj. The epic dw,h| is supported by eva.n gnw,h|, Clem., Paed., III, 1. (Moulton, Prol., p. 193.) ( e) Relative Clauses. This construction in the earlier Greek, like the Latin, had either the subj. or the opt. The Attic added the future ind. which largely displaced the subj. and the opt.401 The N. T. follows the Attic use of the fut. ind. Cf. oi[tinej avpoÄ dw,sousin (Mt. 21:41); oua}j katasth,somen (Ac. 6:3). See 1 Cor. 4: 17, o[j avnamhn,sei. Blass402 explains the occasional return to the subj. as due to i[na. See o[pou fa,gw (Mk. 14:14); par v w|- xenisqw/men (Ac. 21:16); o[ prosene,gkh| (Heb. 8:3); di v h-j latreu,wmengrk grk(12:28). Radermacher (N. T. Gr., p. 138) quotes B. U. III, 822 (ii/A.D.) eu-ron georgo,n ti,j (= oa}j) auvta. e`lu,sh|, Diodorus, XIV, 8, 3, di v w-n evxe,lwsi ta. tei,ch. The N. T. hardly uses the relative clause of purpose as freely as the Attic Greek. ( z) The Infinitive. A brief statement is alone necessary here, since the infinitive receives full discussion in the next chapter. Suffice it to say that the infinitive is exceedingly common in the N. T. for the notion of pure purpose. Votaw403 counts some 1,285 such instances of the simple infinitive of purpose in "biblical Greek." He gives the figures for the N. T. alone as 211. He notes that "this use of the infinitive is second only to that of general object in order of relative frequency of occurrence." Moulton (Prol., p. 205) notes that the inf. of purpose is more common in the N. T. than in Attic, and he agrees with Thumb (Theol. Lit., 1903, p. 421) in the theory that this frequency of the inf. of purpose in the koinh, is due to the Ionic dialect. It has survived in the Pontic dialect of modern Greek, though elsewhere displaced by na, and the subj. Cf. e`toima,swmen fagei/n (Mt. 26:17) and e`toima,swmen i[na fa,gh|j (Mk. 14:12). The telic inf. is common in the koinh, writers generally (Radermacher, N. T. Gr., p. 152). Cf. Xenophon of Eph., 393, 28, evlhlu,qei proseu,xasqai. It is commonest with verbs of movement (Moulton, Prol., p. 205), as in eva.n avnabw/ kavgw. proskunh/sai, Par. P. 49 (ii/B.C.). This infinitive may be resolved easily into the original dative (or locative), as in Jo.
990 A GRAMMAR OF THE GREEK NEW TESTAMENT 21:3, u`pa,gw a`lieu,ein, 'I go a-fishing'; Mt. 2:2, h;lqomen proskunh/sai, we went up for worshipping.'404 It is easy to see the purpose in the dative form of proskunh/sai, but less clear in the locative a`lieu,ein, (probably due to syncretism). Moulton405 suggests that the locative was originally a sort of designed result and gradually the line of cleavage vanished between the two forms as was true of i[na (and ut). "The burden of making purpose clear is in all these cases thrown on the context; and it cannot be said that any difficulty results, except in a minimum of places." This idiom has a much wider range in Homer than in Attic writers and is again more prevalent in the N. T. than in the Attic.406 A few examples must suffice: ouvk h=lqon katalu/saià avlla. plhrw/sai (Mt. 5:17); o` vIhsou/j avnh,cqh- peirasqh/nai u`po. tou / diabo,lou grk(4:1); ouvk h-lqon kaÄ le,sai dikai,uj (Mk. 2:17); pa,resmen avkou/sai (Ac. 10:33). Cf. Lu. 18:10; Ac. 11:25; 12:13; 13:44, etc. Less frequent is the inf. with tou/ for the idea of purpose. Votaw407 notes but 34 such examples of direct purpose in the N. T., though the O. T. shows 734. These 34 are almost confined to Matthew, Luke and Acts. Cf. tou / avpole,sai (Mt. 2:13); to/ spei/rai (Lu. 8:5); tou/ aivtei/n (Ac. 3:2). See both together in Lu. 1:76 f., 79; 2:22, 24, parasth/sai- kai. tou/ dou/nai. For a full discussion see "Articular Infinitive" (Verbal Nouns). Paul seems to avoid it as a rule. But see Ro. 6:6; Ph. 3:10. The use of w[ste and the inf. for pure purpose is rare in the N. T., some half-dozen instances.408 Only probable examples should be claimed (p. 1089). Thus w[ste evkba,llein (Mt. 10:1). Cf. Mt. 15:33; 24:24; 27:1; Lu. 4:29; 20:20. Radermacher (N. T. Gr., p. 160) cites P. Oxy. I, 52, 7 (325 A.D.), evpoistale,ntoj w[ste th.n dia,qesin e;ggrafon prosfwnh/sai. For further examples of telic w[ste in the inscriptions and writers of the koinh, see Koch, Observationes grammaticae, p. 20. It is more frequent in the LXX. Radermacher even cites a case of final w[ste with the subj. in a late papyrus; B. G. U. III, 874, gegra,fhka u`mi/n w[ste pe,myhte. There are two examples of w`j in W. H., w`j e`toima,sai (Lu. 9:52, other editors w[ste) and w`j e;poj eivpei/n (Heb. 7:9). In Ac. 20:24 most editors have w`j teleiw/sai, but not W. H. The articular infinitive with prepositions is very common in the N. T. as in the LXX, about one-half of all the examples of the articular infinitives.409 For a discussion of prepositions with the inf. see Verbal Nouns. Both eivj to, and pro.j to, occur with the inf. in the papyri, the latter
MODE ( EGKLISIS) 991 more frequently. They both seem "to carry the thought of a remoter purpose." (Moulton, Prol., p. 220.) Moulton cites B. U. 226 (i/A.D.) o[pwj eivdh|/ pare,sestai (= qai) - pro.j to. tuci/n, 0. P. 237 (ii/A.D.) o[pwj fronti,sh|j- pro.j to. mh.ÄÄevntugca,nein. The papyri have eivj to. evn mhdeni. memfqh/nai as a "recurrent formula." Cf. P. Fi. 2 (iii/A.D.) 4 times. Moulton gives numerous papyri references for telic eivj to,. The examples with eivj to, are the most common of all in the N. T. (72 instances). As a rule these indicate purpose more or less strong, though not always. It is particularly common in Paul (50 exx., H. Scott). So eivj to. sthÄ ricqh/nai (Ro. 1:11), eivj to. ei=naigrk grk(8:29). Cf. 4 Th. 3:5; Eph. 1: 12; Ph. 1:10). The instances of pro.j to, are ew (12) and chiefly in Matt. and Paul. Cf. pro.j to. qeaqh/nai (Mt. 6:1); pro.j to. dun,nasqai (Eph. 6:11).410 ( h) The Participle. The future participle, so common in this construction in the Attic Greek, has nearly vanished from the N. T. as from the rest of the koinh,. A few remnants survive like e;rcetai vHlei,aj sw,swn (Mt. 27:49), avne,bhn proskunh,swn and poih,swn, (Ac. 24:11, 17). Cf. Ac. 8:27. So also the present participle occasionally occurs where purpose is implied. Thus avpesta,lkaÄ men avpagge,llontaj (Ac. 15:27). Cf. e;pemyan avgge,llontaj (Thuc. VII, 26, 9).411 Cf. also Mk. 3:31. A good example is Ac. 3: 26, avpe,steilen auvto.n auvtolou/nta. See Participle (Verbal Nouns) and Tense for further remarks. (d) Sub-Final Clauses (really object or subject clauses like o[ti clauses ). There are a considerable number of clauses which are not pure purpose and yet are not result. They are the bridge, in a sense, between the two extremes. They are found with verbs of striving, beseeching, commanding, fearing. In some instances the clause is hardly more than an object-clause. The same conjunctions are here used in general, and this shows that no hard and fast line was drawn in the matter. Various divisions are made of these verbs.412 Burton413 calls them object-clauses of exhorting, of striving, of fearing, of subject and predicate, of complementary and epexegetic clauses, of conceived result. But even so they overlap and run into one another. (a) [Ina. Here again the main conjunction is i[na. All these varieties noted by Burton are seen with i[na save with verbs of
992 A GRAMMAR OF THE GREEK NEW TESTAMENT fearing. As we have seen,414 there were two tendencies in the koinh,. One was the spread of the Ionic use of the inf. of purpose, the other was the wide extension of i[na in Western Hellenistic. So the i[na in the non-final or sub-final sense, once rare,415 now comes to be exceedingly common. The development came on soon after the close of the classical age.416 But Thackeray (Gr ., pp. 24, 194) finds it rare in the LXX. It came to be used in almost any sense that the infinitive bore and finally displaced it. This weakened use of ba is one of the characteristics of the koinh, and is richly illustrated in the N. T., particularly in the writings of John. Thus in Mt. 5:29, sumfe,rei i[na avpo,lhtai, the i[na clause is the subject of sumfe,rei and is a subject-clause in the nominative case. There is a great variety of phrases417 which thus use i[na. So avrketo.n i[na ge,nhtai (Mt. 10:25; 18:6). Cf. 1 Pet. 4:3 (inf.). See also i`kano.j i[na (Mt. 8:8), though elsewhere inf.; a;xioj i[na (Jo. 1:27), but inf. in 1 Cor. 16:4, as often; sunh,qeia u`mi/n i[na (Jo. 18:39); evlh,luqen w[ra i[na (Jo. 12:23); evmoi. eivj evla,cisto,n evstin i[na (1 Cor. 4:3); evmo.n brw/ma, evstin i[na (Jo. 4:34); lusitelei/ÄÄÄi[na (Lu. 17:2); tou/toà i[na e;lqh| (Lu. 1:43); zhtei/tai i[na (1 Cor. 4:2); cara.n i[na (Ph. 2:2). Thus the i[na clause is seen to be either nom. or acc., simply, or in apposition with a substantive. In John418 the appositional use is very frequent. So au[th i[na (Jo. 17:3); mei,zona tau,Ä thjà i[nagrk grk(15:13, ablative); evn tou,tw| i[nagrk grk(15:8, locative); ca,rinà i[na (3 John 4, , accusative). Cf. Jo. 6:39; 1 Jo. 3:1, 11, 23; 4:21; 2 Jo. 1:6; 1 Cor. 9:18; Rev. 2:21. In Jo. 15:12 i[na avgapa/te (subj.) is in apposition with evntolh,. Some of these are complementary or epexegetic clauses. In the subject and object (or appositive) clauses the subjunctive is usually found, though occasionally the fut. ind., as in evrre,qh i[na avdikh,sousin (Rev. 9:4). See further examples of the fut. ind. in Rev. 3:9; 6:11; 13:12; 14:13 (especially common in the Apocalypse). In Rev. 9:5 we have evdo,qh i[na mh. avpoktei,nwsin auvtou,jà avll v i[na basanisqh,sontai. In Jo. 17:3 some MSS. read i[na ginw,skousin (read by Treg. and Tisch.). Object-clauses with i[na after verbs of striving, beseeching, etc., largely displace o[pwj. Many of these verbs use also the infinitive and a few retain o[pwj.419 Blass420 gives a careful list of the construction in the N,. T. with each of these verbs. See also
MODE ( EGKLISIS) 993 Thayer under i[na (2). Cf. Acta Pauli et Theclae, 29, pro,seuxai u`pe.r tou/ te,knou mouà i[na zh,setai. With these verbs i[na gives the purport or object rather than the purpose. This use of i[na is very rare421 in classic Greek, though in itself not out of harmony with the Greek genius. The parallel between i[na in this sense and o[ti is seen in Jo. 11:50; 1 Jo. 5:3, 9, 11. Per contra see 1 Jo. 5: 13 for distinction. Cf. also o[ti in. Mt. 13:13 with i[na in Lu. 8:10. It is worth repeating that in the modern Greek (except in the Pontic dialect) it is universal ( na,) to the exclusion of the inf. and o[pwj. It is common after verbs of saying (Thumb, Handb., p. 189). The examples in the N. T. are too numerous to give a complete list. But note i[na after avggareu,w (Mt. 27: 32); avgallia,omai, (Jo. 8:56); avgwni,zomai (Jo. 18:36); aivte,omai (Col. 1:9); avpagge,llw (Mt. 28:10. So paragge,llw, Mk. 6:8); avpoÄ ste,llw (Ac. 16:36); avfi,hmi, (Mk. 11:16); bouleu,omai (Jo. 12:10); and sumb) (Mt. 26:4); ble,pw (1 Cor. 16:10); gra,fw (Mk. 9:12); diaste,llomai (many MSS. in Mt. 16:20); de,omai (Lu. 9:40); di,dwmi (Mk. 10:37); evntolh.n di,dwmi ( lamba,nw), as in Jo. 11:57 (13:34; 15:12); evnte,llomai (Mk. 13:34); evpitima,w (Mt. 12:16; 16:20, W. H.); evxorki,zw (Mt. 26:63); evrwta,w (Mk. 7:26); ei=pon (Mt. 4: 3); and le,gw (Ac. 19:4); qe,lw (Mk. 6:25); e;stin qe,lhma (Mt. 18: 14); zhlo,w (1 Cor. 14:1); zhte,w (1 Cor. 4:2); khru,ssw (Mk. 6:12); merimna,w (1 Cor. 7:34); parakale,w (Mt. 14:36); pei,qw (Mt. 27: 20); poie,w (Jo. 11:37); proseu,comai (Mk. 14:35); sunti,qemai (Jo. 9:22 and inf.); ti,qhmi (Jo. 15:16); fula,ssomai (2 Pet. 3:17). This is a most interesting list. Kalker (Questiones de elocutione Polybiana, 1880. Cf. Moulton, Prol., p. 20) has shown how Polybius favours i[na with verbs of commanding like aivte,omai, paÄ ragge,llw), etc. No real distinction in sense can here be drawn between the inf. and i[na. The later koinh, (and so the N. T.) carried this use of i[na much further than did Polybius, who had more affinity with the old literary Greek. There is no need to appeal to Latin influence for this sub-final use of i[na, as Moulton (p. 208) abundantly shows from the papyri. So 0. P. 744 (i/B.c.) evrwtw/ se i[na mh. avgwnia,sh|j, N. P. 7 (i/A.D.) e;graya i[na soi fulacqw/si, B. U. 531 (ii/A.D.) parakalw/ se i[na kata,sch|j, 0. P. 121 (iii/A.D.) ei=pa, soi ei[na dw,swsin. Moulton (Prol., pp. 177, 208) recalls the old jussive subj. as sufficient explanation of this use of i[na. Radermacher (Rh. M., LVI, 203) and Thumb (Hellen., p. 159) support Moulton against the Latin influence theory. Per contra see Goetzeler, De Polybii El., pp. 17 ff.; Kalker, Quest.; Viereck, Sermo Grae-
994 A GRAMMAR OF THE GREEK NEW TESTAMENT cus, p. 67. Moulton scores his point and observes also that the inf. was not driven out by i[na in the papyri, see ( e). Cf. A. P. 135 (ii/A.D.), evrwtw/ se mh. avmelei/n mou. Radermacher (N. T. Gr., p. 155 f.) gives numerous other examples of non-final i[na in papyri and inscriptions. The subj. is the usual mode employed even after secondary tenses. Thus evbouleu,santo i[na avpoktei,nwsin (Jo. 12:10). In Mk. 9:30, ouvk h;qelen i[na tij gnoi/, we have still the subj., not the opt. As already noted, i[na dw|,h in Eph. 1:17 is an optative of wish after a primary tense. It is here also the subfinal i[na. Cf. Phil. 1:14; Col. 4:12. Moulton422 points out how closely akin are proseu,cesqe i[na mh. e;lqhte (Mk. 14:38) and o`ra/te kai. fula,ssesqe (Lu. 12:15). The paratactic origin of the i[na construction is thus well illustrated. "An innovation in Hellenistic is i[na c. subj. in commands, which takes the place of the classic o[pwj c. fut. indic."423 Moulton cites a moderate number of examples of this abrupt use of i[na in the papyri. So F. P. 112 (99 A.D.) evpe,con (= wn) Zwli,lwi kai. ei[na auvto.n mh. duswph,sh|j, letter of Cicero (Att. 6:5) tau/ta ou=n prw/ton me,nà i[na pa,nta sw|,zhtai\ deu,teron de,à i[na mhde. tw/n to,kwn ovligwrh,sh|j, B. U. 48 (ii/iii A.D.). i[na o`mo,se genw,meqa. There is a doubtful ex. of this sense of i[na in Soph., Oed. C. 155, though o[pwj was so used.424 It appears in Arrian and Epictetus. In the modern Greek the na, clause sometimes "approaches the nature of a principal sentence" (Thumb, Handb., p. 198). But this elliptical imperative is undoubted in the N. T. Cf. Mk. 5:23, i[na evlqw.n evpiqh|/j. So also Mt. 20:32; 1 Cor. 7: 29; 2 Cor. 8:7; Eph. 4:29; 5:33. With this construction compare the asyndeton without i[na in Mk. 10:36, ti, qe,lete poih,sw u`mi/n. As already explained (p. 430), this may be parataxis (two questions). Cf. i[na in Mk. 10:35 and Gal. 5:17.425 ( b) [Opwj. It is much rarer in the N. T. in these constructions. It no longer occurs with the future ind. after verbs of striving. The papyri show o[pwj occasionally in this sense also. Moulton (Prol., p. 208) cites B. M. 21 (ii/n.c.) hvxi,wsa, se o[pwj avpodoqh|/, while " avxiw/ c. infin. occurs in the same, papyrus." Radermacher (N. T. Gr., p. 141 f.) quotes Theoph. ad Autolycum, 2, 34 e;stw soi evreuÄ na/n ta. tou/ qeou/ o[pwj dunh,sei, inscr. from Magn., 90, 12 (ii/B.C.) evfro,ntisen o[pwj- avpokatastw/sin. The few examples in the N. T. are all in the subj. Burton notes only three (Mt. 12 14; 22:15; Mk. 3:6), and all three after sumbou,lion e;labon ( evdi,doun). The clause thus thus partakes of the nature of an indirect deliberative MODE ( EGKLISIS) 995 question (cf. Mk. 11:18, pw/j). They are all after secondary tenses. There are some instances in the N. T. of o[pwj after verbs of beseeching, though many verbs that in Attic had this idiom no longer have it. Thus o[pwj and the subj. occur with de,omai. (Mt. 9:38), aivte,omai (Ac. 25:3), evrwta,w (Lu. 7:3), parakale,w (Mt. 8: 34), proseu,comai (Ac. 8:15). ( g) Mh,à mh, pwjà mh, pote. The usual construction in the negative sub-final clauses is i[na mh,à but a small list of verbs commonly have mh,, as the conjunction. This is true of verbs meaning 'to take heed,' to care for, 'fear.'426 It is a much narrower range than the sub-final use of i[na. In the N. T. the subj. always occurs with mh, except in Col. 2:8 ble,pete mh, tij e;stai. Thus ble,pete mh, tij u`ma/j planh,sh| (Mt. 24:4). Treg. and Tisch. read the fut. ind. in 2 Cor. 12:21, but W. H. and Nestle rightly have tapeinw,sh| (cf. verse 20). The pres. subj. occurs in Heb. 12:15 evpiskopou/ntej mh. evnoclh|/. Elsewhere we have only the aor. subj. Thus after ble,pw (Mk. 13:5); o`ra,w (Mt. 18:10); skope,w (Gal. 6:1); fobe,oÄ mai, (Ac. 27:17). In Ac. 23:10 some MSS. have euvlabe,omai, but fobe,omai is correct. This construction with fobe,omai is rare in the N. T. (Luke, Paul and Hebrews) and is apparently a literary touch. Cf. Ac. 27:29. In Ac. 5:26, evfobou/nto ga.r to.n lao.n mh. liqasqw/sin (note subj. after secondary tense), there is a prolepsis of to.n lao,n)427 Mh, pwj is found after ble,pw with the aor. subj. (1 Cor. 8:9) and fobe,omai (2 Cor. 11:3; 12:20). Cf. Gal. 2:2 in 6, (c), ( d) Pure Final Clauses. If the fear is about an object in the present or past, the ind. is used. Cf. p. 1045. Thus in Lu. 11:35, sko,pei mh. ÄÄevsti,nà and in Gal. 4:11, fobou/mai u`ma/j mh, pwj eivkh|/ kekopi,aka eivj u`ma/j. This is in strict accord with Attic idiom.428 The papyri show it also (Moulton, Prol., p. 193). So Par. P. 49 (ii/B.c.) avgwniw/ mh, pote avrrwstei/, N. P. 17 (iii/A.D.) u`fwrou/me mh. a;ra evnqrw,skwn e;laqen u[dati. Radermacher (N. T. Gr., p. 141) adds examples of fut. ind., as Enoch 6:3, fobou/mai mh. ouv qelh,sete; Dio Chrys., xxxiv, 44, ouv ga.r e;sti ki,ndunojà mh. Mallwtw/n evsome,nwn avsqene,Ä steroi do,xete. The negative in such a clause is ouv. Thus fobou/mai mh, pwj ouvc oi[ouj qe,lw eu[rw (2 Cor. 12:20). This is to show contrast to Cf. Col. 2:8, mh, tij e;staiÄÄkai. ouv. Sometimes a verb of fearing is implied, though not expressed (cf. elliptical use of i[na and i[na mh,). Thus Ac. 5:39, mh, pote eu`reqh/te. This is a possible explanation of mh, pote ouv mh. avrke,sh| (or mh, pote ouvk) in Mt. 25:9
996 A GRAMMAR OF THE GREEK NEW TESTAMENT (note negatives) and mh, pote dw|,h (2 Tim. 2:25). Mh, pote is used with the aorist subj. after prose,cw (Lu. 21:34; Heb. 2:1), with a present subj. after fobe,omai (Heb. 4:1), with a pres. opt. after dialogi,zomai (Lu. 3:15, ind. question), with a fut. ind. after ble,pw (Heb. 3:12). These clauses are of paratactic origin.429 This paratactic construction survives in the use of o[ra with the imperative (Mt. 9:30; 24:6), but even so the clause may be dependent in actual use as in Mt. 18:10; 1 Th. 5:15. Some doubt430 arises concerning the clauses with ble,pw which have a paratactic origin, but are practically dependent. Those in the third person are clearly so (Mk. 13:5; Ac. 13:40, etc.). This argues for a like usage in Lu. 21:8; Gal. 5:15; Heb. 12:25. ( d) The Relative Clause. It is a classic idiom for complementary relative clauses to be used in a sub-final sense.431 As examples of this idiom in the N. T. note a;xio,j evstin w|- pare,xh| (Lu. 7:4); ouvk e;cw oa} paraqh,swgrk grk(11:6); ouvde,na e;cw o[stij merimnh,sei (Ph. 2:20). Cf. scw/ ti, gra,yw (Ac. 25:26) and ti. gra,yai ouvk e;cw (ib.). Radermacher (N. T. Gr., p. 138) quotes from Achilles Tatius, IV, 16, 3, avpogeu,somai tosou/ton o[son kavkei,nh la,bh|) ( e) The Infinitive. With verbs of exhorting, beseeching, etc., the infinitive was the normal idiom in the ancient Greek. In the N. T. it still occurs twice as often as i[na and o[pwj together.432 Some of these verbs have only the inf. in the N. T., as aivscu,nomaià avxio,wà avske,wà bou,lomaià doke,wà eva,wà evpiqume,wà evpipoqe,wà evpitre,pwà evpiceire,wà keleu,wà ovkne,wà paraine,wà peira,wà spouda,zwà ta,ssw and compounds, fronti,zwà fobe,omai in the sense of 'to be afraid to do' (Mt. 2:22). Many of the verbs that use sub-final i[na may have the inf. also. Thus poih,sw u`ma/j gene,sqai (Mk. 1:17). So also bouleu,omaià aivte,omai, proseu,comaià le,gw etc. Cf. a;xioj lu/sai (Ac. 13:25) and a;xioj i[na lu,sw (Jo. 1:27). In 2 Cor. 9:5 the inf. is used after the i[na clause to express an epexegetic or complementary purpose ( tau,thn e`toi,mhn ei=nai), a rather common usage. Cf. in 1 Cor. 9:15 both i[na and the inf. in a broken sentence. Moulton433 argues that in Paul the majority of cases of tou/ with the inf. are epexegetic (Ro. 1:24; 7:3; 8:12; 1 Cor. 10:13) or adnominal (Ro. 15: 23; 1 Cor. 9:10; 16:4; 2 Cor. 8:11; Ph. 3:21) or the ablative construction (Ro. 15:22; 2 Cor. 1:8). Certainly tou/ mh. evlqei/n in Lu. 17:1 is not purpose, nor tou/ eivselqei/n in Ac. 10:25. Cf. also Mt. 21:32, tou/ pisteu/sai. Luke uses tou/ and the inf. more than
MODE ( EGKLISIS) 997 any other N. T. writer. The papyri show this non-final use of tou/ and the inf. (Moulton, Prol., p. 219 f.). So B. U. 1031 (ii/A.D.) fro,nhson tou/ poih/sai, B. U. 164 (ii/iii A.D.) pei/sai auvto.n tou/ evlqei/n, B. M. 23 (ii/B.C.) prosdeome,nou mou tou/ peripoih/sai. In Lu. 18:1, pro.j to. dei/n, is not final. Eivj to, and the inf a we find chiefly in Paul (44 examples, Moulton, Prol., p. 218. Mr. H. Scott makes 50 by counting the verbs instead of the preposition). The construction is always final in the other N. T. writers. But Paul has non-final uses, as in 1 Th. 2:12; 4:9. ( z) Eiv and o[ti. In Lu. 17:2 we have lusitelei/ eiv e;rriptai h' i[na skandali,sh|, where eiv and i[na introduce subject-clauses. Cf. also eiv= o[ti in Mk. 9:42. In Lu. 19:21, evfobou,mhn se o[ti a;nqrwpoj auvsthro.j ei=, the rare use of o[ti with fobe,omai may be causal. It is made easier by the proleptic use of se. The usual object-clause with o[ti belongs to indirect discourse. (e) Consecutive Clauses. (a) [Ina. It is debatable whether i[na has ecbatic use in the N. T. There is in itself no reason why it should not have it, since undoubtedly it was so used in the later Greek.434 It occurs also in modern Greek, as ei=nai na. ca,sh| kanei.j to. mualo, tou, 'that is for one to lose his reason' (Thumb, Handb., p. 197). The parallel of the Latin ut may have had some influence on this late Greek. The development, however, was in the vernacular, and out of the subfinal use of i[na, and the Latin influence was not needed. There is not space to follow the long debate in the grammars and commentaries on this subject. Kuhner435 held that i[na had the ecbatic sense, but Thayer436 boldly accepts the verdict of Fritzsche and Winer who "have clearly shown that in all the passages adduced from the N. T. to prove the usage the telic (or final) force prevails." W. F. Moulton437 agreed with Winer as against Fritzsche in the admission of the sub-final use of i[na, but he balked at the consecutive idea. "But it does not follow that the weakened i[na is generally equivalent to w[ste: this use of i[na is rather, as we can still perceive in most cases, an extension of eo eonsilio ut." Yes, in most cases, beyond a doubt. I once had just this feeling and stood against438 the admission of the consecutive force of i[na. J. H. Moulton439 confesses to a similar development of opinion on this subject. He had once440 committed himself against the ec-
998 A GRAMMAR OF THE GREEK NEW TESTAMENT batic i[na, but now he confesses himself "troubled with unsettling doubts." He boldly advocates441 the freedom of commentators to interpret i[na as the context demands (final, sub-final, consecutive). Ellicott442 had defended just this principle, and he is the most severely grammatical of commentators. The commentator must have grammar, but he needs the grammar of the author on whose work he is making comments. So, also Sanday and Headlam on Ro. 11:11 ( mh. e;ptaisan i[na pe,swsin) pointedly interpret it thus: i[na expresses the contemplated result." They appeal to Ellicott, Lightfoot and Evans in support of this laxer use of i[na as against Winer and the Germans. They also (p. 143) quote Chrysostom's exposition of i[na in Ro. 5:20: to. de. i[na evntau/qa ouvk aivtioÄ lagi,aj pa,lin avll v evkba,sew,j evstin. Lightfoot admits the consecutive force of i[na in Gal. 5:17; 1 Th. 15:4. He is correct in both instances. See also Lu. 1:43. In Jo. 16:2, e;rcetai w[ra i[na do,xh| it is almost temporal. It is argued that, where i[na seems to be used in a consecutive clause, it is the divine purpose that is to be considered. But certainly no such explanation is possible in Ro. 11:11. There is such a thing as the divine purpose and it is seen443 in Lu. 9:45, h=n parakekalumme,non avpo v auvtw/n i[na mh. ai;sqwntai auvto,. Cf. also Mt. 1:22, i[na plhrwqh|/. But surely no such purpose444 appears in Jo. 6:7 ouvk avrkou/sin auvtoi/j i[na e[kastoj bracu. la,bh|) Here we have contemplated result, it is true, but it is result just the same. It is probably just out of this idiom (conceived result) that the use of i[na for actual result came. Burton445 admits this conceived result as in Heb. 10:36, and seeks to explain Jo. 9:2, ti,j h[marten- i[na tuflo.j gennhqh|/; But the effort is not successful. He denies that there is a certain, "scarcely a probable, instance in the N. T. of a clause denoting actual result conceived as such."446 He considers447 Rev. 13:13, poiei/ shmei/a mega,laà i[na kai. pu/r poih|/ evkÄ tou/ ouvranou/ katabai,nein, as the most probable instance of i[na denoting actual result. But there are others just as plain, if not clearer. Thus 1 Jo. 1:9, pisto,j evstin kai. di,kaiojà i[na avfh|/ ta. a`marÄ ti,aj. Blass448 places this beside a;dikoj evpilaqe,sqai (Heb. 6:10) and thinks that the consecutive use of i[na grew out of the infinitive in that sense. With this Moulton449 agrees. Cf. also Rev. 9:20, ouv meteno,hsanà i[na mh. proskunh,sousin, with ouv meteno,hsan dou/nai auvtw|/ do,xan MODE ( EGKLISIS) 999 in 16:9. Note in particular 1 Jo. 3:1, where the clause kai, evsmen accents the ecbatic force of i[na. This use is possible also in Jo. 9:36; Mk. 11:28. In Mk. 4:22, eva.n mh. i[na fanerwqh|/, we have i[na (cf. avll v i[na) used like w[ste and the inf. (cf. Blass, Gr. of N. T. Gk., p. 218). In Mk. 2:10 i[na we have real purpose. The consecutive i[na appears outside of the N. T. as in Arrian (Diss. Epict., II, 2, 16) ou[tw mwro.j h=nà i[na mh. i;dh|) Sophocles in his Lexicon gives a quite extensive list of passages in the koinh, writers where i[na has the consecutive sense. He has probably claimed too many, but some of them are real instance. Even Josephus has i[na in the sense of conceived result.450 Radermacher (N. T. Gr., p. 156) cites Epictetus, IV, 3, 9, evleu,qeroj ga,r eivmi kai. fi,loj tou/ qeou/ i[n v e`kw.n pei,qwmai auvtw|/. Several other examples occur in Epictetus. So, then, we conclude that i[na has in the N. T. all three uses (final, sub-final, consecutive), and thus runs a close parallel with the infinitive which it finally displaced.451 Sophocles cites several examples of consecutive i[na from the LXX. One of these is certainly pertinent, Wisdom of Sol. 13:9, for i[na du,nwntai follows tosou/ton and i[na has the force of w[ste. ( b) [Wste. This conjunction is merely w`j and te,= 'and so.' In Homer w`j is both a demonstrative and a relative. Either idea may appear in w[ste. It is really a comparative particle.452 In the early writers the inf. was more common than the ind. with w[ste. Thus in Euripides the inf. occurs 130 times to 20 indicatives. In Thucydides it is 144 to 82, but in Plato it is 253 to 240. The consecutive sentence began with the inf. and was extended to the finite verb.453 In late Greek it returned to the inf. construction. Cf. Green, Diodorus and the Peloponnesian War, 1899, p. 21. Of the 95 instances454 of w[ste in the N. T. probably 30 do not come up for discussion under either final or consecutive clauses. The word in these examples is merely an introductory inferential particle like all. The structure is wholly paratactic. In this sense of 'therefore' the particle occurs with the ind. nineteen times. Cf. Mt. 12:12, w[ste e;xestin. Once the subj. appears, 1 Cor. 5: 8, w[ste e`orta,zwmen. Radermacher (N. T. Gr., p. 161) even quotes P. Oxy. IV, 743, 27 (ii/B.C.) w[st v a'n tou/to se qe,lw ginw,skeinà and there are other instances like it. The other eleven instances have the
1000 A GRAMMAR OF THE GREEK NEW TESTAMENT imper. (pres.). Cf. w[ste blepe,tw (1 Cor. 10:12). See 1 Cor. 3:21; 11:33, etc. Of the hypotactic examples 62 have the infinitive and only two the indicative. In the Attic Greek actual result was expressed by w[ste and the indicative, while w[ste and the inf. ('so as to') denoted a result naturally or necessarily following the preceding cause.455 In the N. T. there are only two instances of the ind. with w[ste (as a hypotactic conjunction). They are Jo. 3:16, o[twj ga.r hvga,phsen o` qeo.j to.n ko,smon w[ste to.n uio.n to.n monogenh/ e;dwken, and Gal. 2:13, kai. sunupekri,qhsan auvtw|/ oi` loipoi. vIoudai/oi w[ste kai. Barna,baj sunaph,cqh auvtw/n th|/ u`pokri,sei. Here the actual result is distinctly accented. Blass456 on the flimsiest grounds seeks to oust w[ste in Jo. 3:16 by o[ti and to put the inf. in Gal. 2:13, so as to get rid of this construction entirely in the N. T. Moulton457 rightly shows small patience with such "summary" methods in textual criticism. The construction with the ind. is not quite obsolete in the vernacular koinh,, but in the LXX it is almost absent. This classic idiom stands, therefore, in the N. T., but only to make the contrast sharper. Of the 62 instances of w[ste with the inf. in the N. T. they are nearly all consecutive, not final nor even sub-final. Even in the classical Greek the inf. with w[ste in the sense of actual result was displacing458 the ind. and in the vernacular it grew rapidly. Cf. w[ste- avpolelu,sqai, B. G. U. 27 (ii/A.D.). This is a distinct encroachment on the old idiom and has a wider range than in Attic.459 In Ac. 14:1 note ou[twj w[ste. See Mt. 13:32 w[ste evlqei/n ta. peteina. tou/ ouvranou/ kai. kataskhnoi/n evn toi/j kla,doij auvtou/, (Mk. 4:37) w[ste h;dh geni,zesqai to. ploi/on, (Ac. 15:39) w[ste avpocwrisqh/nai auvtou.j avp v avllh,lwn. Tatian took w[ste consecutive in Lu. 4:29 (Moulton, Prol., p. 249). Consecutive w[ste and inf. is too common in the inscriptions and papyri for Radermacher to mention (N. T. Gr., p. 160). We do not have w[ste after a comparative ( h' w[ste) in the N. T. There is no example of w[ste nor of evf v w|-te in the sense of 'on condition that.' In Gal. 2:9 i[na has practically that idea. ( g) `Wj. Thayer considers that in Heb. 3:11 and 4:3 we have the consecutive use of w`j. It is a quotation from the LXX (Ps. 94:11) and is possible, though the simple 'as' is sufficient.460 But
MODE ( EGKLISIS) 1001 (in has kept its place as a consecutive particle in the koinh, (Radermacher, N. T. Gr., p. 160). ( d) [Oti. There is no doubt about the consecutive use of OT in the later Greek.461 We find it in the LXX, as in Ex. 3:11, ti,j eivmi evgw. o[ti poreu,somai pro.j Faraw,; Cf. also 2 Ki. 8:13. The instances in the N. T. are not numerous, but they are very clear. Thus Mk. 4:1, ti,j a;ra ou-to,j evstin o[ti kai. o` a;nemoj kai. h` qa,lassa u`poakou,ei auvtw|/; In Mt. 8:27 note potapo.j o[ti (cf. ou[twj w[ste). See also Heb. 2:6 (Ps. 8:5); Lu. 4:36. Radermacher (N. T. Gr., p. 160) quotes Acta Christophori, 68, 18, toiou/toi ga,r eivsin oi` qeoi. u`mw/n o[ti u`po. gunaiko,j evkinh,qhsan. Moulton (Prol., p. 249) gives ti, didoi/j toi/j avmnoi/j souà o[ti zwh.n aivw,nion e;cousin; Pelagia, 20. It occurs in Theocritus 25 me,gaj- tosou/ton o[ti- die,koya, x, 14 evj tosou/Ä ton o[ti. C. Abbott (Joh. Gr., p. 534) takes o[ti as consecutive in Jo. 14:22, ti, ge,gonen o[ti h`mi/n me,lleij evmfani,zein; Abbott finds no instance of consecutive o[ti in the Egyptian papyri. The idiom is common in the late Greek. Akin to it is the modern Greek use of pou/ as consecutive (Thumb, Handb., p. 197). The same idea is found in Jo. 7:35. ( e) The Relative. This is a common classic idiom. The mode is the ind. and the negative ouv.462 In Latin the subj. is the mode with qui. The tense is usually the fut. ind., though the construction is rare463 in the koinh,. But one may note in the N. T., Mt. 10:26 and in particular 24:2, ouv mh. avfeqh|/ w-de li,qoj evpi. li,qon o[j ouv kataluqh,setai. See also Lu. 8:17; 1 Cor. 6:5; Ro. 8:32. In Jo. 5:7, a;nqrwpon ouvk e;cw i[na ba,lh|, we see i[na usurping this province of the relative. Cf. Rev. 19:15. See "Relative" under Sub-final. ( z% The Infinitive. The inf. with w[ste has been discussed, but we have left the simple inf., the articular ( tou/) inf., eivj to, and the inf. There are apparently examples of each construction in the N. T. Thus the simple inf. of result is seen in Lu. 1:54, avntela,Ä beto vIsrah.l paido.j auvtou/ mnhsqh/nai evle,ouj; at any rate it is used here very freely. Blass464 considers the infinitives in Lu. 1:72 used "quite incoherently." But in Ac. 5:3 yeu,sasqai has a consecutive idea, as has evpilaqe,sqai in Heb. 6:10. See also avnoi/xai in Rev. 5:5 and dou/nai in 16:9. Cf. Lu. 1:76, 78 f. It is probable that originally the Dative - ai in the inf., do,menai as opposed to do,men,
1002 A GRAMMAR OF THE GREEK NEW TESTAMENT expressed "designed result" (Moulton, Prol., pp. 204, 207), but this idea shrank into the background. This idiom is found in the papyri,465 as in 0. P. 526 (ii/A.D.), ouvk h;mhn avpaqh.j avlo,gwj se avpolei,Ä pein. Meyer on Ro. 7:3, tou/ mh. ei=nai, argues that tou/ and the inf. never expresses result, a position which I once held.466 But the evidence is too strong to resist. See Infinitive for distinction between actual and hypothetical result. Radermacher (N. T. Gr., p. 154) quotes Acta Barnabae, 10, mh. bia,sh| Barna,ban tou/ mh. poÄ reu,esqai, as consecutive. The idiom is not common in the papyri as is true of tou/ and inf. (Moulton, Prol., p. 220). It belongs chiefly to the LXX and Byzantine writers, and Moulton puts it in "the higher stratum of education in the main." The epexegetic use occurs, as in C. P. R. 156 evxousi,an- tou/ÄÄqe,sqai, 0. P. 275 tou/ avpospasqh/nai evpi,teimon. This construction ( tou/ and the inf.) had a very wide development in the N. T. in opposition to the encroachments of i[na. See Lu. 17:1 and Ac. 10:25, where tou/ and the inf. is practically the subject of the verb (cf. original dative and locative cases). Luke has two-thirds of the examples of tou/ and the inf. in the N. T. Only half of these (in Gospel and Acts) seem clearly final according to Moulton.467 He holds that of the 13 examples in Paul none are unmistakably final, though Ro. 6:6 and Ph. 3:10 are probably so. In both instances tou/ and the inf. is epexegetic of a i[na clause (Moulton, Prol., p. 218). In Paul 'so as to' will usually express his idea with tou/ and the inf. A clear instance in Luke is seen in Ac. 7: 19, evka,kwsen tou.j pate,raj tou/ poiei/n='so as to make.' Blass468 cites a parallel from the LXX (1 Ki. 17:20), su. evka,kwsaj tou/ qanatw/sai to.n ui`o.n auvth/j. Other LXX instances are Gen. 3:22; 19:21; Is. 5:14. Cf. Ro. 7:3 (epex., consec., p. 1067), tou/ mh. ei=nai. It is probable in Lu. 9:51; Ac. 18:10; 20:3; 27:1; Ro. 1:24. Cf. tou/ evrwth/sai and o[pwj kataÄ ga,gh|j in Ac. 23:20. So with eivj to, and the inf. Its most natural signification is aim or purpose, but, just as with i[na, so here result is sometimes the idea. Meyer in his note on Ro. 1:20, eivj to. ei=nai auvtou.j avnapologh,touj, insists that the meaning of eivj to. is always purpose. In this particular instance divine purpose may be the idea, though result is the probable conception. See Sanday and Headlam in loco. Ellicott on 1 Th. 2:12, eivj to. peripatei/n (after parakalou/ntej ktl)), admits the sub-final use of eivj to, (cf. i[na) after verbs of exhorting (cf. 1 Th. 3:10), though denying the ecbatic use. But it is only a step to go on and that
MODE ( EGKLISIS) 1003 the N. T. writers took. See the epexegetic use of eivj to, in 1 Th. 4 9. Winer469 admitted the consecutive use of eivj to, and the inf. as in 2 Cor. 8:6, as eivj to. parakale,sai h`ma/j Ti,ton, 'so that we besought Titus.' This idiom is not present in the Johannine writings, though it is very frequent in Paul's writings especially Ro. and 1 Th.) and Hebrews. Notice tacu.j eivj to. avkou/saià bradu.j eivj to. lalh/sai (Jas. 1:19). In Heb. 11:3, eivj to. gegone,nai, we have a clear example of result. Note the perfect tense with notion of permanence.470 See also fronei/n eivj to. swfronei/n (Ro. 12:3), where purpose is impossible. Cf. Gal. 3:17. As to pro.j to, and the inf. the point is not clear. Purpose is undoubtedly present as in Mt. 6:1; Eph. 6:11, and there is total absence of purpose in Lu. 18:1, pro.j to. dei/n. It is not certain, in spite of Blass' comment,471 that in the N. T. pro.j to, expresses result. In Mt. 5:28, pro.j to. evpiqumh/sai, either purpose or result is possible. W. F. Moulton472 denies that the idiom ever conveys mere result, but admits that it may have subjective purpose as in 1 Th. 2:9. J. H. Moulton473 holds that this is the idea in all the four examples in Paul's writings. See further 2 Th. 3:8; 2 Cor. 3:13. 7. WISHES. The use of the optative for a future wish like a`gia,sai (1 Th. 5:23), mh. ge,noito (Gal. 6:14), is not a hypotactic construction. This is pure parataxis and has already been discussed under the Optative.474 See Optative Mode. The only hypotactic sentence for the expression of a wish in the N. T. is that with ovfei,lw, which comes in the late Greek to be used as a particle. Even here it is possible to regard the construction as paratactic, but note eiv ga,r and ei;qe. It is the second aorist ind. of ovfei,lw without the augment. ;Ofelon, with the inf. occurs in Herodotus, and the form is thus probably Ionic.475 For koinh, parallels see "Impossible Wishes" under Indicative Mode. Cf. w;feiÄ lon suni,stasqai in 2 Cor. 12:11. It is found in the LXX476 as a conjunction, as in Ex. 16:3, o;felon avpeqa,nomen. Cf. Num. 14:2; 20:3. Moulton477 suggests that its application to the second and
1004 A GRAMMAR OF THE GREEK NEW TESTAMENT third persons is due to the meaning 'I would' rather than 'thou shouldst.' As a matter of fact its use in the N. T. is very limited, though ei;qe and eiv ga,r are wanting as particles of wishing. For a wish about the past we have the aorist ind. So o;felo,n ge evbasiÄ leu,sate (1 Cor. 4:8). Cf. Ps. 118 (119):5. For a wish about the present we have the imperfect ind. So 2 Cor. 11:1, o;felon avnei,cesqe, and Rev. 3:15, o;felon h=j. The Text. Rec. here has o;felon ei;hj, but it is baseless. However, we do find the fut. ind. for a future wish. So Gal. 5:12, o;felon avpoko,yontai. Wishes as a separate idiom are vanishing in the N. T. But o;felon appears in Lucian, Athenagoras, Greg. Naz., Socrates. Cf. Sophocles' Lexicon. To compensate for this loss we have the strong asseverations with ouv mh, (Mt. 13:14), the use of eiv like the Hebrew ~ai (Mk. 8:12; Heb. 4:3), eiv mh,n (Heb. 6:14), the use of the participle like the Hebrew inf. absolute (Mt. 13:14). The distinction between wish and supposition with eiv was sometimes hard to make in Homer.478 The relation between wishes and conditions is not clear. 8. CONDITIONAL SENTENCES. (a) Two Types. No hypotactic clause is more important than this. For some reason the Greek conditional sentence has been very difficult for students to understand. In truth the doctors have disagreed themselves and the rest have not known how to go. The theory of Hermann, followed by most Germans (Winer,479 Blass480), is the one that I learned from Broadus and have expounded in my Short Grammar.481 It is also that of Gildersleeve.482 This theory in brief is that there are four classes of conditions which fall into two groups or types. The two types are the determined and the undetermined. The point in "determined" is that the premise or condition is assumed to be true (or untrue). A positive statement is made in either case and the conclusion follows logically from this premise. The indicative is the one used for this type (the first and second class conditions, real and unreal, or fulfilled and unfulfilled). The other type is the undetermined condition. Naturally the indicative is not allowed here. The element of uncertainty calls for the subj. or the optative. The difference therefore between the third and fourth class conditions is just that between the subj. and the opt. They are both modes of doubtful, hesitating affirmation, but the optative
MODE ( EGKLISIS) 1005 is more remote than the subj. In this type the premise is not assumed to be either true or untrue. The point is in the air and the cloud gathers round it. But there is less mist over the subj. than the opt. In broad outline this is the classification of the conditional sentences which I hold to be true. Thompson483 is surely right in saying that no division can claim any higher right than that of convenience and intelligibility, except that I should like to add that the exposition should be in harmony with the facts of the historical development of the Greek language. There is no nobler achievement in syntax than the Greek conditional sentence before it broke down from the loss of the optative and the future indicative. In the modern Greek it is therefore a wreck, and there is corresponding obscurity between the various classes of conditions, as in English, in spite of special developments to make atonement for the loss.484 In broad outline these four classes of conditions may be termed Reality, Unreality, Probability, Possibility. The word Probability is, however, too strong a term for the third-class condition ( eva,n and the subj.). La Roche485 prefers "objektive Moglichkeit" for the third class and "subjektive Moglichkeit" for the fourth class ( eiv and the opt.). This is also the language of Winer,486 "objective possibility" and "subjective possibility." Farrar487 prefers the words Possibility, Impossibility, Slight Probability, Uncertainty. Radermacher (N. T. Gr., p. 142) calls eiv with ind. "objektiv," eva,n with subj. "an sich objektiv," eiv with opt. "subjektiv," eiv with past tenses of ind. "Irrealitat." So it goes. Radermacher thinks also that, to understand the Greek conditions, we must distinguish sharply between the vernacular and the koinh, ("so mussen wir Scharf scheiden zwischen Volkssprache and der Koine"), a mistaken view in my judgment. It is best to use koinh, for both the vernacular and literary language. This brings us face to face with the other theory, the one adopted by Farrar. It was expounded by Goodwin488 and has had quite a vogue in America and England.489 This theory calls for "particular" and "general" suppositions as a fundamental element. This is a false step in itself. As
1006 A GRAMMAR OF THE GREEK NEW TESTAMENT Gildersleeve490 shows, each of the four classes of conditions may be particular or general. That point has no bearing on the quality of the condition. Goodwin's past general supposition, where alone a show of distinct structure is made, is a mixed condition (see later under fourth class condition): But the point on which I wish to attack Goodwin's scheme is chiefly in his definition of the first and second class conditions. That involves the third also, as will be seen. Goodwin confuses the "fact" with the "statement" of the fact. He describes the first condition thus: "When the protasis simply states a present or past particular supposition, implying nothing as to the fulfilment of the condition, it takes a present or past tense of the indicative with eiv." The words to which I object, besides "particular," are "implying nothing as to the fulfilment of the condition." This condition pointedly implies the fulfilment of the condition. It is the condition of actuality, reality, Wirklichkeit, and not mere "possibility" as Farrar has it (see above) a la Goodwin. This is the crux of the whole matter. Once see that the first class condition with the ind. implies the reality of the premise, all else follows naturally. In the discussion of the second class condition Goodwin491 properly says: "When the protasis states a present or past supposition, implying that the condition is not or was not fulfilled, etc." This is the condition of unreality as the other is that of reality and the indicative is, of course, used with both. Hence the subj. and the opt. conditions fall apart to themselves as undetermined. The point about all the four classes to note is that the form of the condition has to do only with the statement, not with the absolute truth or certainty of the matter. Examples will be given directly to show that the second class condition is sometimes used where the fact is just the opposite. The same thing is true of the first class condition. We must distinguish always therefore between the fact and the statement of the fact. The conditional sentence deals only with the statement. This point is clearly seen in KuhnerGerth, II, p. 465, except that the third class is lost sight of and merged with the first. Burton492 follows Goodwin through all his
MODE ( EGKLISIS) 1007 ramifications. A word further is demanded by way of warning. One must not try to explain the Greek condition by the English or German translation. The English is often hopelessly ambiguous, while the Greek is perspicuous if one will only give it a chance to speak for itself. The true explanation is only possible by the approach from the Greek standpoint. And that is by the mode, not by eiv or eva,n. vEa,n is nothing but eiv a;n. The a;n is not essential to either protasis or apodosis. Homer493 used eiv with the subj. with or without ke, or a;n. The Attic Greek494 sometimes has eiv a;n with the opt. and Demosthenes used eiv a;n with the past ind. Radermacher (N. T. Gr., p. 127) quotes Joh. Philop. De aeternitate 430, 28 (iii/A.D.) eiv- hvdu,nato a;n. He gives also (p. 163) ka'n - bohqoi,h, Diod. XI, 37, 3; eva.n mh.- r`u,saito, Diod. I, 77, 3. The modern Greek uses a;n (for eva,n) with any tense of the ind. (Thumb, Handb., p. 194). There is no principle involved in a;n, simply custom. In modern Greek the subj. is used, of course, more freely since the fut. ind. and the opt. have vanished.495 Jolly holds that the ind. was a later development with conditional sentences in Greek and that the first attempt was made with the subj. and the opt. He thinks that the use of the ind. was the result of a clearer conception of the logical possibilities of the conditional clause. The subj. was more common in the Zend and the Sanskrit (and Latin) than in the Greek.496 Here as always a;n is difficult to explain. "Now it has a definite reference, now it is indefinite. Sometimes the reference is supplied by the context, sometimes by the opposite."497 See The Use of a;n in Relative Sentences in this chapter. We shall first examine the standard forms of the conditional sentence and then note the variations and modifications. (b) Four Classes. (a) Determined as Fulfilled. This class of condition assumes the condition to be a reality and the conclusion follows logically and naturally from that assumption. Gildersleeve (Am. Jour. of Philol., 1882, p. 435) observes that this is the favourite condition: "It is the favourite condition when one wishes to be or seem fair, the favourite condition when one is sure of the premiss." The construction is eiv (sometimes eva,n)498 and any tense of the in-
1008 A GRAMMAR OF THE GREEK NEW TESTAMENT dicative in the protasis. The apodosis varies very greatly. It all depends on what one is after, whether mere statement, prediction, command, prohibition, suggestion, question. Hence the apodosis may be in the indicative (any tense) or the subjunctive or the imperative. There is no necessary correspondence in tense between protasis and apodosis. The variation in the mode of the apodosis has no essential bearing on the force of the condition. This condition, therefore, taken at its face value, assumes the condition to be true. The context or other light must determine the actual situation. The apodosis is the principal clause, but since the protasis is the premise, the protasis usually precedes the apodosis. The apodosis may be declarative or interrogatory, positive or negative. This condition is so frequent in the N. T. that no exhaustive list can be given, but representative examples must suffice. Thus in Mt. 12:27, eiv evgw. evn Beezebou.l evkba,llw ta. daimo,niaà oi` ui`oi. u`mw/n evn ti,ni evkba,llousin; This is a good example (cf. also Gal. 5:11) to begin with, since the assumption is untrue in fact, though assumed to be true by Jesus for the sake of argument. The question is a reductio ad absurdum. In verse 26, eiv o` Satana/j to.n Satana/n evkba,lleià evf v e`auto.n evmeri,sqh, there is the additional point of change of tense in the apodosis. He was already divided against himself, in that case, before he casts himself out. But the tense may be merely due to a quick change of view-point as accomplished (timeless aorist in reality). This point comes out well in verse 28, eiv de. evn pneu,mati qeou/ evgw. evkba,llw ta. daimo,niaà a;ra e;fqasen evf v u`ma/j h` basilei,a. Note a;ra with the aorist. For the past ind. in both clauses see Ac. 11:17 ( eiv e;dwkenà ti,j h;mhn); 1 Cor. 15:2; Rev. 20:15 ( ei; tij ouvc eu`re,qhà evblh,qh). For the present ind. in both clauses note Mt. 19:10 ( eiv ou[twj evsti,n ouv sumfe,rei); Ro. 8:9; Jo. 15:18; 1 Cor. 15:12. The presence of the perfect in protasis protasis(15:14, 17, 19) or apodosis apodosis(15:13, 16) does not vary the point. In 2 Cor. 2:5, the perfect is followed by the perfect. The fut. ind. may, though rarely in the N. T., occur in both clauses, as in Mt. 26:33 ( eiv skandalisqh,sontaià skandalisqh,Ä somai%. Cf. Mk. 14:29; Lu. 19:40; 1 Cor. 3:15; 2 Tim. 2:12; 1 Cor. 3:14 f. But such little niceties cut no figure in this construction. There is perfect liberty to mix the tenses ad libitum. So past and present (Lu. 19:8 f.; 11:18; 2 Cor. 7:8, 14; Ro. Lange (Der horn. Gebr. der Partikel Eiv) makes it exclamatory. But Hale (The Orig. of Subj. and Opt. Cond. in Gk., Harv. Stu. in Class. Philol., 1901) treats it as a demonstrative in the locative case, meaning 'in that case.' This is more probable.
MODE ( EGKLISIS) 1009 4:2; 15:27; 1 Jo. 4:11), past and future (Jo. 3:12; 15:20; Lu. 16:11), present and future (Mt. 17:4; Jo. 5:47; 11:12; Ac. 5:39; 19:39; Ro. 8:11). In 1 Cor. 9:11 eiv evspei,ramen and eiv qeri,somen occur side by side. Examples of the imperative in the apodosis occur as in Mk. 4:23 ei; tij e;cei w=ta avkou,einà avkoue,tw. Cf. Mt. 5:29; 8:31; Lu. 4:3; Ac. 16:15; Jo. 7:4; 18:23. In Lu. 4:3, eiv ui`o.j ei= tou/ qeou/à eivpe,, we have a good example of the first class condition. The devil would not, of course, use the second class (assumed to be untrue), for that would be an affront to Christ. The third and fourth classes would throw doubt on the point. The temptation, to have force, must be assumed as true. The devil knew it to be true. He accepts that fact as a working hypothesis in the temptation. He is anxious to get Jesus to prove it, as if it needed proof for Christ's own satisfaction and for his reception. If the devil used Aramaic, then we have Christ's own translation of it or that of the Evangelist. In Jo. 18:23 ( eiv kakw/j evla,lhsaà martu,rhson peri. tou/ kakou/), however, the assumption is not a fact, though Christ treats it as such for argument's sake. Cf. Lu. 23:35, 37. In Jo. 20:15 note the aorist ind. ( eiv evba,stasj) and the imper. ( eivpe,). Blass (Gr. of N. T. Gk., p. 215) takes eiv qe,leij in the late koinh, to be like the French s'il vous plait. Cf. Mt. 17:4. For the subj. in the apodosis note Gal. 5:25, eiv zw/men pneu,matià pneu,mati kai. stoicw/men. The use of eva,n with the ind. is rather more frequent in the late koih,. Finally eiv came to be "a mere literary alternative."499 In the koinh, in Pisidia and Phrygia eva,n occurs with the aorist ind., the pres. ind. and the future ind. as well as with the subj.500 The papyri examples are unmistakable, as eva.n dei/ in Tb. P. 58 (ii/B.C.), eva.n oi=den B. U. 546 (Byz.), eva.n fai,netai, A. P. 93 (ii/A.D.), eva.n d v eivsi,n 0. P. (ii/A.D.), eva.n keleu,eij 0. P. 1150, 2 f. (vi/A.D.), eva.n macou/sin Par. P. 18, eva,nper evkplhrw,sousin Par. P. 62 (ii/B.c.).501 Radermacher (N. T. Gr., pp. 83, 163) cites others from the papyri and inscriptions. So Heberdey-Wilhelm, Reisen, p. 137, eva.n de, tij qh,sei* Eum. Hippiatr., p. 244, 30, eva,nper evno,rchj evsti,n. Perhaps examples like eva.n h=n are not to be counted as instances, since h=n for h|= is sometimes subj.502 In general, the difference between eiv and eva,n is considerably lessened in the koinh,, though it must be
1010 A GRAMMAR OF THE GREEK NEW TESTAMENT remembered that eva,n was never confined to the subj. nor eiv to the ind. and opt. vEa.n h=sqa occurs in Job 22:3, and Moulton503 quotes it from Hb. P. 78 (iii/B.c.) as "certainly subj." Cf. also eva.n h=san Tb. P. 333 (iii/A.D.), and a number of undoubted examples of eva,n with past, present and future tenses of the ind. from koinh, writers are given in Sophocles' Lexicon under eva,n) Thayer calls it "a somewhat negligent use, met with from the time of Aristotle on." It was just a normal development in the koinh, till in the modern Greek a;n, is used indifferently with either ind. or subj. So a'n to, vkanej, 'if you did so,' a'n diya,sh|j, 'if you thirst' (Thumb, Handb., p. 194 f.). Theophylact in his Proem to Luke has eva.n mh. evqa,rrei, In the N. T. we note eva.n oi;damen (1 Jo. 5:15); eva.n sth,kete (1 Th. 3:8), where the distinction is clear between the two modes (ind. and subj.). In 1 Th. 3:8 aD have sth,khte, but in Lu. 6:34 there is considerable support for eva.n danei,zete, as there is for eva,n te avpoqnh,skomen in Ro. 14: 8. In Gal. 1:8 a few MSS. read eva.n euvagÄ geli,zetai. It is possible to treat eva.n marturw/ as pres. ind., Jo. 5: 31; 8:14. There is undue scepticism on Blass' part504 concerning eva,n and the fut. ind. It is true that the MSS. are generally divided, but there is no real room for doubt about following aBCE in Ac. 8:31, eva.n o`dhgh,sei, except for possible itacism with - h|. That is possible also in Rev. 2:5 where W. H. read eva.n metanoh,sh|j. But there is no room for itacism in Mt. 18:19 eva.n sumfwnh,sousin, supported by aBDEL D 33, although rejected by W. H. and Nestle (FGKM have - wsin), nor in Lu. 19:40 eva.n siwph,sousin, nor in Rev. 2:22 eva.n mh. metanoh,sousin. In Mt. 18:19 the editors seem unwilling to follow the MS. evidence for the fut. ind. It is mere tradition to feel that eva,n has to have the subj. Besides, we have eva.n e;sh| and eva.n mhke,ti prosqh,sw Hermas, Mand. V, 1. 2 and Mand. IV, 3. 7. In Lev. 22:9 we find eva.n bebhlw,sousin. There is at any rate no great difference in the resultant sense between the fut. ind. and the aor. subj. and it was a very natural development. Cf. Homer's use of ke, with both. But, when all is said, as a matter of fact, in the N. T. as in the koinh, generally, the rule is for eiv to appear with the incl. and eva,n with the subj. In 1 Cor. 7:5 we have eiv mh,ti a;n (bracketed by W. H.) without a verb. It is matched by the papyri.505 Thus B. U. 326 ei; ti eva.n - katali,pw, 0. P. 105 (ii/A.D.) ei; ti a;llo aiva.n ( e;) cw, B. M. 233 (iv/A.D.) ei; ti a'n- avnalw,sh|j, Tb. P. 28 (ii/B.C.) eiv ka'n du,natai. In these the modal a;n ( eva,n) is separated from eiv and used as if with o[jà o[pou. Rader-
MODE ( EGKLISIS) 1011 macher (N. T. Gr., p. 162) cites also Joh. Philop., De aetern., p. 85, 19, eiv ouvk a'n - u`pa,rch|. Deissman506 sees no analysis of eva.n mh, ti in this, though Moulton contends for this explanation. The use of eiv peri,keitai in Mk. 9:42 in the sense of o[ti Blass (Gr. of N. T. Gk., p. 215) calls "quite incorrect." He means it is not "classic." Note the irony in 1 Cor. 14:38, ei; tij avgnoei/, avgnnoei/tai. The negative of the protasis in the first class condition is practically always ouv in the N. T. We have eiv ouv as a rule, not eiv mh,. In the classic Greek the rule was to use eiv mh,, and eiv ouv appeared only where the ouv coalesced with a single word (the verb generally) or for sharp antithesis or emphasis.507 But in the N. T., as in the koinh, generally and occasionally in the Attic,508 we meet eiv ouv in the condition of the first class. Jannaris509 notes 34 examples of eiv ouv in the N. T., but Moulton510 finds only 31 of this class of condition. There are only two in the second, so that there is a slight discrepancy. In truth eiv mh, occurs only five times with the simple logical condition, and the examples are not quite normal except the one in Mk. 6:5, ouvk evdu,nato eiv mh. evqara,peusen (a simple past condition), and in 1 Tim. 6:3, ei; tij- mh. prose,rcetai. (Blass calls this an "abnormal" instance from the literary style and unlike the N. T. idiom). But see 1 Cor. 15:2 evkto.j eiv mh. eivkh|/ evpisteu,sate, 2 Cor. 13:5 eiv mh,ti avdo,kimoi, evsteà Gal. 1:7 eiv mh, tine,j eivsin. Elsewhere the negative is ouv. This is in harmony with the meaning of ouv and the ind. mode. The definite negative goes with the definite mode. This is the condition of supposed reality and eiv ouv is the natural combination. In general Blass511 is correct in saying that ouv is the negative of the ind. and of the other modes including the inf. and part. This, of course, was not the Attic standard, but that was hopelessly gone even for the Atticists.512 In the modern Greek de,n, (from ouvde,n) supplants ouv with the ind. and mh,( n) goes with the subj. That is the goal, as Moulton observes,513 which is not yet reached in the N. T., for mh, occurs in questions of doubt with the ind. and eiv mh, still holds on. Even in the modern Greek, Thumb (Handb., p. 195) gives de,n, with subj. or ind. in conditions as a' de.n pisteu,h|j and a' de.n ph,gaina. Rader-
1012 A GRAMMAR OF THE GREEK NEW TESTAMENT macher (N. T. Gr., p. 172) cites Pap. Wess. xxvi, eiv ouv di,dotai,. But the point to get clear is that in the first class condition the normal negative in the koinh, is eiv ouv. Moulton counts the idiom 6 times in Luke, 3 in John, 16 in Paul, 2 in James, and one each in Matthew, Hebrews, 2 Peter and Revelation. As examples take Lu. 18:4 eiv kai. to.n qeo.n ouv fobou/mai ouvde. a;nqrwpon evntre,pomai and Jo. 1:25 eiv su. ouvk ei= o` Cristo,j. In the latter case the negative is very emphatic. So in Jo. 5:47 eiv ouv pisteu,ete. Cf. further Lu. 12:26; 16:11, 31; Jo. 3:12; Ro. 11:21; 1 Cor. 15:13, 1517; 2 Th. 3:10. Sometimes ouv practically coalesces with the verb, as in Lu. 14:26; 1 Cor. 7:9; 11:6; 16:22; 1 Tim. 5:8; Rev. 20:15. The notion of contrast is seen in Jo. 10:37 eiv ouv poiw/à eiv de. poiw/) Note also ka'n mh. pisteu,hte. So in 5:46 eiv piÄ steu,eteà eiv de.ÄÄouv pisteu,ete. See further Lu. 11:8; Jas. 2:11; 2 Pet. 2:4. In Mt. 26:42 note eiv ouv du,natai tou/to parelqei/n eva.n mh. pi,w. In Ro. 11:21, eiv ouvk evfei,satoà ouvde. sou/ fei,setai, it is hardly possible to translate eiv ouv by 'unless.' The same thing is true in 1 Cor. 9:2 and 15:29. Cf. eva.n mh, in 9:16. ( b) Determined as Unfulfilled. In this somewhat difficult condition only past tenses of the ind. occur. The premise is assumed to be contrary to fact. The thing in itself may be true, but it is treated as untrue. Here again the condition has only to do with the statement, not with the actual fact. A good illustration is found in Lu. 7:39 ou-toj eiv h=n o` profh,thjà evgi,nwsken a;n) The Pharisee here assumes that Jesus is not a prophet because he allowed the sinful woman to wash his feet. Jesus is therefore bound to be ignorant of her true character. The form of the condition reveals the state of mind of the Pharisee, not the truth about Jesus' nature and powers. As a matter of fact it is the Pharisee who is ignorant. For this reason I cannot agree with Moulton's statement514 that the ind. is not suited to the expression of contingencies, wishes, commands or other subjective conceptions. On p. 201 Moulton recovers himself by saying that "these sentences of unfulfilled condition state nothing necessarily unreal in their apodosis," and "the sentence itself only makes it untrue under the circumstances." I should add " as conceived by the speaker or writer." Surely the ind. is the mode for positive and negative statements, for directness of statement and clarity of expression. But one must emphasize the words "statement" and "expression." The ind. does not go behind the face value of the record. Most untruths are told in the ind. mode. The
MODE ( EGKLISIS) 1013 statement of unreality here from the standpoint of the speaker or writer, is as clear cut and positive as that of reality in the first class condition. The term "unreal" as applied to this use of the ind. properly belongs only to the standpoint of the user. To him the case is impossible and he makes a positive statement to that effect with the ind. By the ind. mode the condition is determined. Whether it is fulfilled or unfulfilled is a more difficult matter. This idea has to be conveyed by suggestion. It is not a question of positive or negative, but of definite assumption of unreality. The "unreality" does not come from the ind. That in its origin is a matter wholly of the context. Take Mk. 6:5, for instance, ouvk evdu,nato eiv mh. evqera,peusen. In the abstract it is not possible to tell which class of condition we have here. Its is either first or second, we know. If the writer is talking about the present time in terms of past time, then it is a second class condition determined as unfulfilled. The Greek fell upon the use of the past tenses of the ind. as a device to help in this matter. An unfulfilled condition about present time was expressed in terms of the imperfect ind. An unfulfilled condition about past time was expressed in terms of the aorist or the past perfect ind. There is the analogy of wishes to justify it, if, indeed, wishes did not come out of this construction ( ei;qeà eiv ga,r). The origin of this precise point is obscure.515 In the context one must seek for light and help. In Mk. 6:5 ( ouvk evdu,nato evkei/ poih/sai ouvdemi,an du,naminà eiv mh. ovli,goij avrrw,stoij evpiqei.j ta.j cei/raj evqera,peusen) it is clear that a definite past event is chronicled. So it is a condition of the first class, determined as fulfilled. But in Jo. 15:22 (and 24) eiv mh. h;lqon kai. evla,lhsa auvtoi/jà a`marti,an ouvk ei;cosan, how is it? Is it a simple historical narrative about a past situation? Is it a hypothesis about the present time in terms of past time to suggest its unreality? Fortunately here the context shows. The very next words are nu/n de. pro,fasin ouvk e;cousin peri. th/j a`marti,aj auvtw/n (Cf. also nu/n de, in verse 24). The contrast with the present and actual situation is made in plain terms. In Jo. 9:41 we have nu/n de, even after a;n. This is not always clone in the context and one is either left to his wits or a;n is added to the apodosis. In verse 19 of John 15 we have eiv evk tou/ ko,smou h=teà o` ko,smoj a'n to. i;dion evfi,lei. "The addition of a;n to an indicative hypothesis produced much the same effect as we can express in writing by italicising 'if' "516 or by add-
1014 A GRAMMAR OF TH GREEK NEW TESTAMENT ing to the apodosis 'in that case.' This is the definite use of a;n. But it is a mistake to say, as some writers517 do, that a;n in the apodosis is essential to the second class condition. Even Moulton518 says: "The dropping of a;n in the apodosis of unfulfilled conditions was classical with phrases like e;deià evcrh/nà kalo.n h=n." The absence was so undoubtedly, but was a;n ever really necessary with these verbs? When a;n was used with them, there was a slight change of meaning. The N. T. is in perfect accord with ancient idiom when it has kalo.n h=n eiv ouvk evgennh,qh (Mt. 26:24); evdu,nato eiv mh. evpeÄ ke,klhto (Ac. 26:32); eiv mh. h=nà ouvk evdu,nato (Jo. 9:33), not to mention the apodosis alone in Mt. 25:27; Lu. 19:23; Ac. 22:22; 27: 21; 2 Cor. 2:2 ; 12:11; 2 Pet. 2:21. In Ac. 24:19, as oua}j e;dei evpi. sou/ parei/nai kai. kathgorei/n ei; ti e;coien pro.j evme, it is a mixed cond. (protasis in fourth class) and the apodosis is itself a relative clause. But the idiom goes further than these verbs of propriety and possibility and obligation, as is seen in Gal. 4:15, eiv dunato,nà evdw,Ä kate, moi; Jo. 15:22, 24; 19:11, ouvk ei=ceijà eiv mh. h=n soi dedome,non Ro. 7:7, ouvk e;gnwn eiv mh. dia. no,mou and ouvk h|;dein eiv mh. e;legen. In 1 Cor. 5:10, evpei. wvfei,lete, we have the apodosis of this condition. Moulton (Prol., p. 200 note) cites 0. P. 526 (ii/A.D.) eiv kai. mh. avne,bene evgw. ouv pare,benon; 0. P The simplest form of this condition is when the imperfect occurs in both clauses or the aorist in both. In the former case present time is generally meant, as in Lu. 7:39 eiv h=nà evgi,nwsken a;n , Jo. 5:46 eiv episteu,eteà evpisteu,ete a;n. So also Jo. 8:42; 9:41; 15:19;
MODE ( EGKLISIS) 1015 18:36; 1 Cor. 11:31; Gal. 1:10; Heb. 8:4, 7.519 In Jo. 8:19, eiv h|;deiteÄÄa'n h|;deite, we have the same construction, for this past perfect has the sense of the imperfect. In Heb. 11:15, eiv evmnhÄ mo,neuon- ei=con a;n, however, the reference is to past time as the context makes clear. It is descriptive of an unreal hypothesis in the.past of a continuous nature. 'If they had kept on remembering, 'they would have kept on having.' This is a classical idiom, though uncommon. Another example is seen in Mt. 23:30, eiv h;meqa evn tai/j h`me,raij tw/n pate,rwn h`mw/nà ouvk a'n h;meqa. Only the context can help one tell the kind of condition in 1 Cor. 12:19 and Heb. 7:11, for the apodosis appears in the form of a question without a;n and the verb. The other normal condition of this class is where the aorist ind. occurs in both clauses, as in Mt. 11: 21 eiv evge,nontoà pa,lai a;n meteno,hsan, Mk. 13:20 eiv mh. evkolo,bwsen ouvk a'n evsw,qh. This refers to past time. Cf. Mt. 25:27; 1 Cor. 2: 8; Jo. 14:2; Heb. 10:2 (only apodosis). Sometimes one tense occurs in one clause, another in the other. The standpoint is shifted. Thus in Jo. 14:28 eiv hvgapa/teà evca,rhte a;n, Gal. 3:21 eiv evdo,qhà a'n h=n, Heb. 4:8 eiv kate,pausenà ouvk a'n evla,lei. Cf. also Jo. 15:22, 24. It is not always certain that the present reference of h=n can be insisted on, since there was no separate aorist form of eivmi,. Sometimes h=n is aorist. So as to Jo. 11:21, 32, eiv h=jà ouvk a'n avpe,qanen. But the point of difference is certainly made in Jo. 18: 30, eiv mh. h=n poiw/nà ouvk a'n paredw,kamen. Cf. Ac. 18:14; Mt. 26:24. In Jo. 4:10, eiv h|;deijà su. a'n h|;thsaj, we have the same thing. Cf. also Mt. 24:43. In Ac. 18:14 note in the next verse eiv de, evstinà o;yesqe (first class). In 1 Jo. 2:19 we have the past perfect in the apodosis eiv h=sanà memenh,keisan a;n the solitary example.520 But the past perfect occurs in the protasis as in Ac. 26:32, avpoleÄ lu,sqai evdu,natoà o` a;nqrwpoj ou-toj eiv mh. evpeke,klhto Kai,sara. Cf. also eiv evgnw,keiteà ouvk a'n katedika,sate (Mt. 12:7), though Westcott521 takes this as a "real imperfect" like h|;dein above. The periphrastic past perfect we find in Jo. 19:11 ouvk ei=cejà eiv mh. h=n dedome,non. Moulton522 has given a list of the times that a;n appears in the apodosis in the N. T. with the ind. imperf. (17 times), the ind. aor. (24) and the past perfect (1). In Lu. 17: 6 we have the pres. ind. and the imperf. combined, eiv e;ceteà evle,gete a;n. This is really a mixed condition (first and second classes). Cf. Jo. 8:39, eiv
1016 A GRAMMAR OF THE GREEK NEW TESTAMENT evste,à evpoiei/te (the margin of W. H.). Radermacher (N. T. Gr., p. 163) quotes P. Oxy. IV, 729 (137 A.D.) eva.n de. mh. evkw,lusen Zeu,j e;zhsen a;n, where note eva,n with aorist ind. like the modern Greek a'n to. h;xeura (Thumb, Handb., p. 195). The negative of the second class condition is in the N. T. always mh, except twice, Mt. 26:24 (Mk. 14:21) kalo.n h=n auvtw|/ eiv ouvk evgennh,qh. Here the ouv is very emphatic. Elsewhere we have eiv mh, as in Mt. 24:22 (note mh, in protasis, ouv in apodosis); Jo. 9:33; 15:22, 24; 18:30; 19:11; Ac. 26:32; Ro. 7:7. In itself eiv mh, is three times as common in the N. T. as eiv ouvà but outside of the five examples of eiv mh, in the first class conditions above and one in the third class (Lu. 9:13) eiv mh, is confined to the second class condition and to the elliptical use like plh,n in the sense of 'except' or the phrase eiv de. mh, meaning 'otherwise' without a verb (cf. eiv mh, thus in Mt. 12:4; Lu. 4:26; eiv de. mh, in Jo. 14:11).523 See a bit later on this point. As already noted, modern Greek uses a'n de,n in this condition (Thumb, Handb., p. 195). ( g) Undetermined, but with Prospect of Determination. This class uses in the condition clause the mode of expectation (Erwartung), the subj. It is not determined as is true of the first and second class conditions. But the subj. mode brings the expectation within the horizon of a lively hope in spite of the cloud of hovering doubt. W. G. Hale524 considers that the subj. in this condition is due "to a fusion of volitive subj. and the anticipatory subj." Monro525 thinks it is the quasi-imperative sense (volitive subj.). He argues that the use of mh, with the subj. (cf. prohibitions) proves this. But Moulton526 replies that "the negative mh,, originally excluded from this division of the subjunctive, has trespassed here from the earliest times." So he urges that the subj. with eva,n (as with o[tan) is the futuristic, not the volitive, use. The futuristic subj. in Homer may have ouv, but usually mh, with the subj. in conditions, and yet some cases of eiv ouv with the subj. occur in Homer when a coalesces with the verb as eiv ouvk evqe,lwsin, Iliad 3. 289, eiv ouvk eivw/sin, 20. 139. In Jer. 6:8 we still have h[tij ouv katoikisqh|/ in B. The truth probably is that in some instances this subj. is futuristic, in others volitive or deliberative. The point is a fine one as one can readily see. Gildersleeve527 finds the
MODE ( EGKLISIS) 1017 prevalence of the subj. in conditional (as in temporal) clauses due to the-greater exactness of the subj. here. It enables one, since it has a "tendency to realization" (Tendenz zur Wirklichkeit),528 to make a difference between the indicative and the optative conditions, though it has more affinity with the optative, except in the case of some future indicative conditions which come very close to the subj. idea. The kinship in origin and sense529 of the aorist subj. and fut. ind. makes the line a rather fine one between eiv and the fut. ind. and eva,n and the subj. indeed, as we sometimes have evan and the fut. ind. in the first class condition, so we occasionally meet eiv and the subj. in the third class condition. Radermacher (N. T. Gr., p. 162) notes eiv and subj. at first as a "vulgarism," but surely the classic usage answers that. The inscriptions have usually only eva,n and aorist subj. he finds. But he finds also abundant instances of eiv and subj. in koinh, and late writers. So Epictetus, II, 18, 11 eiv mh, tij evxalei,yh|, Vettins, 274, 11 evi de, tij logi,shtai, Hippiatr., 177, 2 eiv prossch|/j, Demetrius, De eloc. 21, 11 eiv ge,nhtai, Pausanias, II, 35, 3 eiv u`dreu,wntai. So in Lu. 9:13 eiv mh,ti avgora,swmen, 1 Cor. 14:5 evkto.j eiv mh, deirmhneu,h| Ph. 3:12 eiv katala,bw (possibly also ei; pwj katanÄ th,sw in verse 11), Rev. 11:5 ei; tij qelh,sh| (text of W. H., but margin qe,lei or qelh,sei). In Ro. 11:14, ei; pwj parazhlw,sw kai. sw,sw, we may also have the aorist subj. In 1 Th. 5:10 we have ei;te grhgorw/men ei;te kaqeu,dwmen. It is in the midst of a final sentence with i[na. In 1 Cor. 9:11 some MSS. read eiv qeri,swmen. This construction occurs occasionally in classical Greek. It was frequent in Homer and in the Attic poets, but is rare in our normalized texts of Attic prose, though a few examples occur in Time., Plato, Xenophon.530 This "laxity" increased till finally eiv, like o[te, vanishes before eva,n ( a;n) which is used indiscriminately with ind. or subj., while eiv is a mere "literary alternative." In modern Greek a;n has driven eiv out of the vernacular. In Deut. 8:5 AF have ei; tij paideu,sh|. Cf. Judg. 11:9. Moulton531 finds the same construction in the papyri as does Deissmann,532
1018 A GRAMMAR OF THE GREEK NEW TESTAMENT though it is rare in the early papyri.533 Moulton (Prol., p. 187) cites 0. P. 496 (ii/A.D.) eiv de. h=n (= h|=), though he534 seems curiously unwilling to admit the examples in the N. T. As Just as the second class condition was debarred from the future, so the third class condition is confined to the future (from the standpoint of the speaker or writer). The first class condition covers past, present and future. In I Cor. 10:27 note ei; tij kalei/ and eva,n tij ei;ph. In Ac. 5:38, eva.n h|= and eiv- evsti,n, a real distinction is preserved. Gamaliel gives the benefit of the doubt to Christianity. He assumes that Christianity is of God and puts the alternative that it is often in the third class. This does not, of course, show that Gamaliel was a Christian or an inquirer. He was merely willing to score a point against the situation, but eva,n and the subj. ontemplate the future result Sadducees. Here, indeed, the su position is about a present (turn out to be). So eva,n e;chte in 1 Cor. 4:15; eva.n h|= in Mt. 6:22. vEan qe,lh|j in Mt. 8:2 is future in conception. In Jo. 5:31, eva.n marturw/ (possibly pres. ind.), the idea would be 'if perchance I bear witness.' Cf. also 8:14. I such instances the matter may be looked at as a present realty (so eiv skandali,zei. Mt. 5:29)
MODE ( EGKLISIS) 1019 or a future possibility (so eva.n skandali,sh|, Mk. 9:43). Cf. also eva.n avgaph,shte in Mt. 5:46 with eiv a`gapa/te in Lu. 6:32 (in verse 33, eva.n avgaqorpoih/te).536 In Jo. 13:17 note eiv tau/ta oi;dateà maka,rioi, evste eva.n poih/te auvta,) Here we have the first and third class conditions happily combined with clear distinction. Jesus assumes the knowledge as a fact, but the performance is doubtful. The tense is usually the aorist, though sometimes the pres. subj. occurs. Thus eva.n avkou,sh| (Mt. 18:15); 1020 A GRAMMAR OF THE GREEK NEW TESTAMENT Tenses where papyri parallels are given). That may be the explanation here. It is possible also to explain it as a change of standpoint. The protasis looks to the future, while the apodosis turns back to the past. Such vivid changes in language are due to the swift revolution in thought. See Mt. 18:15, eva.n avkou,sh|, evke,rdhsaj; Jo. 15:6, eva.n mh, tij me,nh| evn evmoi,à evblh,qh e;xw kai. evxhra,nqh (cf. evdoxa,sqh i[na fe,rhte also of the future); 1 Cor. 7:28, eva.n kai. gamh,sh|jà ouvc h[martej\ kai. eva.n gn,mh| h` parqe,nojà ouvc h[marten. For a similar idiom see Ignatius, Ep. to Romans 8:3; to Polycarp 5:2. Moulton (Prol., p. 247) cites Epict., a'n me.n strateu,swmaià avphlla,ghn. See also Abbott, Joh. Gr., p. 586. In Mk. 10:30, eva.n mh. la,bh|, we have eva.n mh, almost in the sense of oa}j mh,) Cf. also eva.n mh. i[na in Mk. 4:22. The use of eiv ouv and eva.n mh, side by side is seen in Mt. 26: 42, eiv ouv du,natai tou/to parelqei/n eva.n mh. auvto. pi,w. Cf. also Jo. 10: 37, eiv ouv poiw/ and ka'n mh. pisteu,hte. ( d) Remote Prospect of Determination. Hale539 attributes "the Greek optative assumption to a fusion of the true opt. and the potential opt." The use of the opt. in the protasis of this condition is probably volitive, since the negative540 is mh,) That is certainly true of the optative in wishes with eiv or eiv ga,r $ei;qe%.541 But the deliberative use occurs a few times with eiv in indirect questions. The potential opt. in the apodosis with a;n is more difficult to explain. It is certainly not volitive any more, not more than mere fancy (Vorstellung), the optative of opinion,542 and apparently futuristic. This fourth class condition is undetermined with less likelihood of determination than is true of the third class with the subj. The difference between the third and fourth classes is well illustrated in 1 Pet. 3:13 f. So Jesus draws a distinction in Lu. 22:67. The use of the opt. in both apodosis and protasis accents the remoteness of the hypothesis. And yet it is not in the category of unreality as in the second class. It floats in a mirage, but does not slip quite away. It is thus suitable not merely for real doubt, but it also fits well the polite temper of courteous address. It is evident that this condition will be comparatively infrequent. It is an ornament of the cultured class and was little used by the masses save in a few set phrases (or wishes). It is not strange, therefore, that no complete example of this fourth class condition appears in the LXX, the N. T. or the papyri so far as examined.543 Radermacher (N. T. Gr., pp.
MODE ( EGKLISIS% 1021 133, 143) with all his diligence produces no example of the opt. in both condition and conclusion in the current koinh,. In the modern Greek it has disappeared completely. In the N. T., as in the LXX, the instances of the protasis are very few. Moulton544 notes only 13 in the LXX apart from the Atticistic 4 Maccabees. Of these he observes that 2 are wishes, 5 are cases of w[s$per) ei; tij and 2 are indirect questions. There are in the N. T. only 11 examples. Some of these are indirect questions. Thus in e;legon eiv bou,loito poreu,esqai (Ac. 25:20) we have the opt. of ind. discourse. The direct was eiv bou,lh|. The same thing is true of 27:39, evbouleu,onto eiv du,naito evksw/sai to. ploi/on. There is implied indirect discourse or purpose (cf. the classic use of eiv for purpose).545 So we see aim in Ac. 17:27, zhtei/n eiv a;ra ge yhlafh,seian auvto.n kai. eu[roien, and 20:16, e;speuden eiv dunato.n ei;h. In 27:12, pw/j du,nainto, we have both purpose and implied indirect discourse. In 24:19, ei; ti e;coien, the protasis is more nearly that of the proper fourth class condition, but even so it is a mixed condition, since the apodosis e;dei belongs to the second class. Blass546 ventures to suggest ei; ti e;cousin as more correct. But it is needless to change the text. These examples are all in Acts, one of the more literary books of the N. T. Paul has only the stereotyped phrase eiv tu,coi. (1 Cor. 14:10; 15:37), which is a, true example of this protasis, "if it should happen." The two other examples are in 1 Pet. 3:14 eiv kai. pa,scoite dia. dikaiosu,nhnà maka,rioi, and 3:17 krei/tton avgaqopoiou/ntajà eiv qe,loi to. qe,lhma tou/ qeou/à pa,scein. The idiom is a mere torso, as is evident. In 0. P. 1106, 7 (vi/A.D.), eiv ga.r evpime,noienà plh/qoj evpisth,setai stratiwtiko,n, we have a mixed condition. The apodosis with a;n (the less definite a;n) is more frequent and occurs both in direct and indirect discourse. Since the potential opt. in the N. T. never occurs in connection with the protasis, the matter was discussed sufficiently under The Optative Mode in Independent Sentences (see this chapter, III, 3, (b) ). This potential opt. is practically the apodosis of an unexpressed protasis. But the exx. occur in questions save one (Ac. 26:29). Twice the questions are direct (Ac. 8:31; 17:18). The rest are indirect (opt. preserved as in the direct). Cf. Lu. 1:62 ti, a'n qe,loi, Ac. 5:24 ti, a'n ge,noito. So Lu. 6:11. The deliberative element in some of these questions is well illustrated in Lu. 9:46; Ac. 10: 17. The MSS. vary in some cases about the presence of a;n as
1022 A GRAMMAR OF THE GREEK NEW TESTAMENT in Lu. 18:36. The examples are all in Luke's writings. In Ac. 8:31 we do indeed have a protasis, but not of the fourth class. It is a mixed condition. The disappearance of this opt. condition led to the enlarged use of the first and third classes. In Ro. 3:6 and 1 Cor. 15:35 the fut. ind. is used where the potential opt. would have suited the Attic idiom.547 (c) Special Points. (a) Mixed Conditions. The human mind does not always work in stereotyped forms, however excellent they are. Grammatical construction is merely the expression of the mental conception. Freedom must be acknowledged without any apology. I say these somewhat commonplace things because of the bill of "exceptions" which meet us in so many grammars at this point. It would have been a miracle if the four classes of conditions were never "mixed," that is, if the protasis did not belong to one class, while the apodosis fell in another. In P. Goodsp. 4 (ii/B.C.), eiv e;rrwsaià ei;h a;n, we have the protasis of the first class and the apodosis of the fourth. Radermacher (N. T. Gr., p. 132) quotes Pastor Hermae, Sim. IX, 12, 4 ouvdei.j eivseleu,setai eiv mh. la,boi, Theoph. Ad Autolycum eiv ma.r ba,boi- evkkau,sei. Thus in Lu. 17:6, eiv e;ceteà evle,gete a;n, we have a protasis of the first class (determined as fulfilled) and the apodosis of the second (determined as unfulfilled). The same thing is true of the marginal reading in the text of W. H. in Jo. 8:39, eiv evste,à evpoiei/te. In Ac. 24:19, oua}j e;dei evpi. sou/ parei/nai kai. kathgorei/n ei; ti e;coien pro.j evme,, we find a protasis of the fourth class with an apodosis of the second class. Then again in Ac. 8:31, pw/j ga/r a'n dunai,mhn eva.n mh, tij o`dhgh,sei me* we have a protasis of the first class (barring. itacism) and an apodosis of the fourth. The examples like 1 Cor. 7:28 do not amount to mixed condition, since it is merely a question of the standpoint in time of the apodosis, though this apodosis does more naturally go with the first class condition. There may be two protases, as in 1 Cor. 9:11, and both of the same class, or the two may belong to different classes, as in Jo. 13:17. ( b) Implied Conditions. Sometimes the apodosis is expressed, while the protasis is merely implied by a participle, an imperative or a question. In such examples one must not think that the participle, for instance, means 'if.' Thus in Ro. 2:27 teÄ lou/sa with krinei/ suggests a condition of either the first or the third class according as one conceives it. The condition is hinted at, not stated. The same thing is true of lambano,menon in 1 Tim.
MODE ( EGKLISIS) 1023 4:4 and metatiqeme,nhj in Heb. 7:12. Cf. also Heb. 2:3; 1 Cor. 11:29; Gal. 6:9. This use of the participle is still very frequent548 in the N. T. In Mt. 16:26 we have eva.n kerdh,sh|, while in Lu. 9:25 note kerdh,saj. In Lu. 19:23, kavgw. evlqw.n su.n to,kw| a'n auvto. e;praxa, the apodosis calls for a condition of the second class (context). The imperative is used where a protasis might have been employed. Thus in Mk. 1:17, deu/te ovpi,sw mouà kai. poih,sw. The adverb deu/te has the force of an imperative. There is an implied condition here. So also 11:24 pisteu,ete kai. e;stai. Cf. Mt. 7: 7; 11:28; 19:21; Lu. 7:7; Jo. 2:19; 14:16; Jas. 4:7. The imp. may be (Jas. 1:5) the apodosis of an expressed condition and the implied protasis of another conclusion.549 In Eph. 4:26; ovrgi,zeÄ sqe kai. mh. a`marta,nete, two imperatives together practically answer as protasis and apodosis. In Mt. 7:10, h' kai. ivcqu/n aivth,sei- mh, o;fin evpidw,sei auvtw|/; the two questions do the same thing in a rough sort of way (anacoluthon), not technically so. In Mt. 26:15, ti, qe,lete, moi dou/nai kavgw. u`mi/n paradw,sw auvto,n, the question takes the place of the protasis. Here kai, joins the two parts of the sentence, but in Jas. 5:13 we have question and imperative in separate sentences. Cf. also 1 Cor. 7:21. These devices are all found in the classic idiom.550 ( g) Elliptical Conditions. An incomplete condition is really a species of ellipsis or aposiopesis and is common to all languages.551 Ellipsis of the copula in the apodosis (1 Cor. 12:19) or the protasis (Ro. 8:17) is not the point. That is, of course, common. So Ro. 4:14; 8:17; 11:16; 1 Cor. 7:5; 1 Pet. 3:14; 2 Cor. 11:16. There may be the absence of either protasis or apodosis. The apodosis is wanting in some instances. The suppression of the apodosis in Lu. 13:9, ka'n me.n poih,sh| karpo.n eivj to. me,llonÄÄ amounts to aposiopesis.552 See also 19:42, eiv e;gnwj kai. su,) Cf. further Mk. 7:11; Jo. 6:62; Ac. 23:9. In Lu. 22:42 the aposiopesis disappears from the text of W. H. ( pare,negke, not parenegkein). In 2 Th. 2:3, eva.n mh. e;cqh|, we have a mere anacoluthon as in Ph. 1:22. These protases belong to either the first, second or third classes. The lonely protases of the fourth class discussed above (cf. 1 Pet. 3:14, 17) come in here also. We have a species of anacoluthon. The structure of the sentence is changed so that the corresponding apodosis does not follow. In the same
1024 A GRAMMAR OF THE GREEK' NEW TESTAMENT way (suppression of apodosis) is to be explained the use of eiv like ~ai in the sense of 'not,' in solemn oaths or questions. The apodosis is wanting. So eiv doqh,setai th|/ geneh|/ tau,th| shmei/on (Mk. 8: 12). So Heb. 3:11 (4:3, 5) eiv evleu,sontai (Ps. 94-95:11). This is aposiopesis. The full expression is seen in Gen. 14:23; Num. 14:30; 1 Sam. 14:45. It is an apparent imitation of the Hebrew idiom, though not un-Greek in itself. Radermacher (N. T. Gr., p. 184) treats this idiom in Mk. 8:12 as due to translation from the Hebrew (Aramaic). Analogous to this is eiv mh,n, in Heb. 6:14, if eiv is not really h= changed by itacism (cf. Ezek. 33:27; 34:8). Hort553 holds to the difference between eiv mh,n and h= mh,n and would take eiv in Heb. 6:14 as the true eiv. But Moulton554 makes out a good case from the papyri and the inscriptions for taking it as merely a variation of h= mh,n. He finds eleven papyri examples of ei= mh,n from ii/B.C. to i/A.D. Particularly clear is the Messenian Mysteries inscr., Michel 694, ei= ma.n e[xein. If so, it does not come in here. But the use of eiv in questions is pertinent. Thus eiv ovli,goi oi` swzo,menoi; (Lu. 13:23). Cf. Mt. 12:10; Lu. 12:26; 22:49; Ac. 17:27; 19:2. Radermacher (N. T. Gr., p. 136) takes eiv in questions= h= as in Lu. 22:49. This is possible on grounds of itacism, but it does not entitle Radermacher to say "werden muss." The use of the condition in the sense of 'to see if' borders on this elliptical construction. Something has to be supplied before the protasis in order to make the idea clear. The apodosis is virtually contained in the protasis. It is a classic555 idiom and reappears in the papyri.556 So 0. P. 743, o[loj diaponou/mai eiv [E) calÄ kou/j avpo,lesen. The protasis here may conform to the first class condition as in eiv e;cei (Lu. 14:28); ei; pwj h;dh pote. euvodwqh,somai (Ro. 1:10). So Mk. 11:13; Ac. 8:22. In Ph. 3:12, eiv kai. kataÄ la,bw, we have the third class and possibly also in Ro. 11:14. But in Ac. 27:12 it is the fourth class, ei; pwj du,nainto. The use of eiv in the indirect question, as in Mk. 3:2, eiv qerapeu,sei, corresponds closely with the preceding. Cf. also 11:13. The same thing is true of eiv in the sense of o[ti, as in Ac. 26:23. This is also true of eiv with verbs of wonder, as in Mk. 15:44; Ac. 26:8. The protasis itself is sometimes abbreviated almost to the vanishing point, as in eiv mh, without a verb, in the sense of 'except' (Mt. 5:13). Here eiv and mh, seem to coalesce into one word like plh,n. Cf. 11:27, ouvdei.j evpiginw,skei to.n ui`o.n eiv mh. o` path,r. This is very common as in classic Greek. Sometimes we have eiv mh. mo,non as in
MODE ( EGKLISIS) 1025 Mt. 21 : 19. The origin of this use of eiv mh, was the fact that the verb was identical with the preceding one in the apodosis and so was not repeated. From this ellipsis the usage spread to mere exceptions to the previous statement, a limitation simply. Eiv mh, may make exception to a preceding negative as in Gal. 1:19, e[teron de. tw/n avposto,lwn ouvk ei=don eiv mh. vIa,kwbon to.n avdelfo,n. The effect here is to make eiv mh, seem adversative instead of exceptive. Cf. Mt. 12:4. For eva.n mh, in this construction see Gal. 2:16. In 1 Cor. 7:17 eiv mh, has the sense of 'only' and is not to be construed with peripatei,tw. The use of eiv mh, occurs in questions expetting a negative answer, as in Mk. 2:7, ti,j du,natai avfie,nai a`marÄ ti,aj eiv mh, ei-j o` qeo,j; In 1 Cor. 7:5, eiv mh,ti [ a;n], we have ti (cf. ei; ti in Mt. 18:28) added and possibly also a;n. B here omits a;n, possibly to "case a difficulty" as Moulton557 suggests. If genuine, it would be a sort of analysis of eva,n into eiv a;n, that occurs in the illiterate papyri. For examples see under 8, (b), (a). For eiv mh,ti with the ind. pres. see 2 Cor. 13:5 and the subj. aorist. See Lu. 9:13. The use of evkto.j eiv mh, probably comes by analogy from evkto.j eiv (cf. Latin nisi), but it occurs in the N. T. without verbs only in 1 Tim. 5:19. Elliptical also are eiv mh. i[na (Jo. 10:10); eiv mh. o[ti (2 Cor. 12:13); eiv mh. o[tan, (Mk. 9:9). In Jo. 14:11 note eiv de. mh, in the sense of 'but if not,' otherwise.' Cf. Mk. 2: 21; Rev. 2:5, 16. For eiv de. mh,ge see Lu. 5:36. Other forms of eiv used elliptically are ei; per (Ro. 3:30); w`sei, (Mt. 3:16); w`speÄ rei, (1 Cor. 15:8). Eiv de. mh, and eiv de. mh, ge became such fixed phrases558 that they occur even when the preceding sentence is negative (Mt. 9:17) or where eva.n mh, would be more natural (Lu. 10:6, where the phrase answers to eva.n h|=. Cf. Lu. 13:9. In Jo. 14:2, eiv de. mh, ei=pon a;n, the conclusion is expressed. In 2 Cor. 10:9 we have w`j a;n without a verb= 'as if.' It is common to have ei;teÄÄei;te (1 Cor. 8:5) without the verb. The use of ka;n without the verb is also found in the sense of 'if only,' 'at least.' So in Mk. 5:28; 6:56. In 2 Cor. 11:16 we have both eiv de, mh, ge and ka;n ( de,xhsqe to be supplied). In Lu. 12:38 note ka;n - ka;n) The suppression of the protasis occurs in all the examples of the potential opt. already discussed, as in Ac. 26:29. Even in the deliberative questions of the opt. with Ccv the same thing is true. Cf. Ac. 17:18 (direct); Lu. 1:62 (indirect). The protasis is also suppressed sometimes with evpei,. Cf. 1 Cor. 15 29, evpei. ti, poih,sousin; Here a protasis of the first or (more probably) of the third class must be supplied. So in Ro. 3:6; 11:6,
1026 A GRAMMAR OF THE GREEK NEW TESTAMENT 22. In 1 Cor. 14:16, evpei. eva.n euvlogh|/j pw/j evrei/, the ellipsis still occurs in spite of eva,n. In Heb. 9:26, evpei. e;dei, and 10:2, evpei. ouvk a'n evpau,santo, the protasis would belong to the second class, as is true also of evpei. wvfei,lete a;ra in 1 Cor. 5:10. In 7:14, evpei. a;ra evsti,n, the protasis would be of the first class. ( d) Concessive Clauses. These are really just conditional559 clauses with the addition of kai,. In kai. eiv and kai. eva,n ( ka;n) the sense is 'even if' and is climacteric. Burton560 seeks to draw quite a distinction between concessive and conditional clauses. He cites Mt. 26:33, eiv pa,ntej skandalisqh,sontai evn soi,à evgw. ouvde,pote skandalisqh,somai, as an instance of the concessive idea without kai,. It is possible that we may read the idea into this passage because in the parallel passage in Mk. 14:29 we read eiv kai, avll v evgw,) Cf. also ka'n de,h| in Mt. 26:35 with eva.n de,h| in Mk. 14: 31. The use of eiv ( eva,n) in the sense of 'though' shows that there is at bottom no essential difference. The structure is precisely the same as the conditional sentence. They are, to repeat, nothing but conditional sentences of a special tone or emphasis. The use of kai, was to sharpen this emphasis either up or down. With kai. eiv the supposition is considered improbable.561 With kai. eiv the truth of the principal sentence is stoutly affirmed in the face of this one objection. It is rhetorically an extreme case. In 1 Cor. 8:5, kai. ga.r ei;per eivsi.n - [ avll] h`mi/n ei-j qeo,j, we have an instance. In Mk. 14:29 the true text is eiv kai,, not kai. eiv. In 1 Pet. 3:1 W. H. read simply eiv. In late Greek kai. eiv vanishes before kai. a;n ( eva,n).562 So in the N. T. we have kai. eva.n kri,nw (Jo. 8:16). So also Gal. 1:8. For ka;n see Jo. 8:14, ka'n marturw/) So Mt. 21:21; 26:35. See Jo. 10:38, eiv de. poiw/à ka'n evmoi. pisteu,hte. The clauses with eva,n and the subj. are, of course, third class conditions. Sometimes563 kai. eiv and ka;n can hardly564 be considered as strong as 'even if.' They may be resolved into 'and if.' So Mt. 11:14; Lu. 6:32; Mk. 16:18; Jo. 8:55; Rev. 11:5. Much more common is eiv kai,. This phrase means 'if also.' Here the protasis is treated as a matter of indifference. If there is a conflict, it makes no real difficulty. There is sometimes a tone of contempt in eiv kai,. The matter is belittled. There is often some particle in the conclusion in this construction as in Lu. 18:4, eiv kai. to.n qeo.n ouv fobou/mai ouvde. a;nqrwpon evntre,pomaià dia, ge to. pare,ceinà ktl) Note ge as in 11:8. Cf. Col. 2:5, eiv kai, - avlla,.
MODE ( EGKLISIS) 1027 There is considerable variety with eiv kai,. Thus in 2 Cor. 7:8 we have a condition of the first class (so Lu. 11:8; 18:4, etc.), while in 1 Pet. 3:14, eiv kai. pa,scoite, we have one instance of the fourth class. With eva.n kai, and the subj. we find, of course, the third class. So Gal. 6:1, eva.n kai. prolhmfqh|/. Cf. 2 Tim. 2:5. In 1 Cor. 7:28, eva.n kai. gamh,sh|j, the notion is 'if even' rather than 'also' (cf. kai. eva.n gh,mh|). In Mt. 18:17 note eva.n parakou,sh| auvtw/n, and eva.n de. kai. th/j evkklhsi,aj parakou,sh|. There is nothing peculiar about Ro. 14:8, eva,n te zw/men- eva,n te avpoqnh,skwmen (Cf. Ex. 19:13.) Cf. ei;teÄÄei;te with the ind. (1 Cor. 3:22) or the subj. (1 Th. 5:10). The use of the participle for concession (see kai,per w;n, Heb. 5:8) will be treated under the Participle. For the use of ka;n even after eva,n see Mk. 5:28. ( e) Other Particles with eiv and eva,n. These have no effect on the condition as a distinct class, though they modify the precise idea in various ways. This point will be treated more exactly under Particles. But note eiv a;ra (Mk. 11:13; Ac. 8:22); ei; ge (Eph. 4:21); eiv a;ra ge (Ac. 17:27 opt.); ei; ge kai, (2 Cor. 5:3); eiv de. mh,ge (Lu. 5:36); eiv ou=n (Mt. 6:23; Heb. 7:11); ei;per (Ro. 3: 30); eva,nper (Heb. 3:14; 6:3); ei; pwj (Ro. 1:10, the fut. ind.; Ac. 27:12, the opt.). In Mk. 8:23 ei; ti is in direct question. 9. INDIRECT DISCOURSE (Oratio Obliqua). (a) Recitative [Ote in Oratio Recta. Direct quotation is more frequent in primitive language, in the vernacular, and in all vivid picturesque narrative. It is the dramatic method of reporting speech. It is natural in Homer, in the Old Testament and in the Gospels, in Aristophanes and in Shakespeare, and in Uncle Remus. The prolonged indirect discourse in Thucydides and in Livy, in Xenophon and Caesar, is more or less artificial. In the LXX little use is made of indirect discourse. The direct quotation may not be as verbally exact as the indirect,565 but it is more lively and interesting. As a rule the direct discourse is simply introduced with a word of saying or thinking. The ancients had no quotation-marks nor our modern colon. But sometimes tin was used before the direct quotation merely to indicate that the words are quoted. We find this idiom occasionally with o[ti, more seldom with w`j, in the Attic writers.566 It is very rare567 in the LXX, since the Hebrew so frequently has a special participle like 'saying.' But see Gen. 28: 16. In the N. T. Jannaris568 counts 120 instances of recitative o[ti.
1028 A GRAMMAR OF THE GREEK NEW TESTAMENT The idiom appears chiefly in the historical books. See Mt. 7:23, o`mologh,sw o[ti ouvde,pote e;gnwn u`ma/j. This particular instance can be looked upon as indirect discourse, since the person is the same in both clauses and the tense and mode are unaffected. It is probable that indirect declarative clauses grew out of constructions of this nature.569 But in Mt. 27:43, ei=pen o[ti qeou/ eivmi. ui`o,j, there is no doubt at all. See 26:74, ovmnu,ein o[ti ouvk oi=da to.n a;nqrwpon, and 26:75, eivrhko,toj o[ti pri.n avle,ktora fwnh/sai tri.j avparnh,sh| me. So Mk. 1:37; 2:12, 16; 4:21; 8:28; Jo. 10:36; Ac. 25:8; Ro. 4:17. In Mt. 16:7 we have (W. H., but R. V. marg. has causal) recitative o[ti ( o[ti a;rtouj ouvk evla,bomen); while in verse 8 the indirect (probably causal) use, o[ti a;rtouj ouvk e;cete* In Mk. 6:23 (W. H. marg.) we have a direct quotation with o[ti, in Mt. 14:7 the same thing appears as indirect discourse without o[ti. In Jo. 10:34, avpekri,qh- ouvk e;stin gegramme,non o[ti evgw. ei=pa qeoi, evste, note a treble direct quotation, once with o[ti and twice without. In Jo. 1:50 the first o[ti is causal, the second is indirect discourse. The o[ti in the beginning of Jo. 20:29 is causal. In Jo. 20:18 o[ti is recitative, causal in 3:18, declar. in 3:19. It is doubtful whether first o[ti is recitative or causal in Jo. 21:17. In Ro. 3:8, o[ti poih,Ä swmen (hortatory subj.), o[ti is also recitative. So in 2 Th. 3:10 o[ti is merely recitative. The instances of direct quotation without o[ti are very numerous. Cf. Mt. 8:3; 26:25. Sometimes the same thing is reported with o[ti (Mt. 19:9) or without o[ti (Mk. 10:11). For single words quoted without agreement with the word with which they are in apposition note o` dida,skaloj and o` ku,rioj in Jo. 13:13. W. H. seek to indicate the presence of recitative o[ti by beginning the quotation with a capital letter as in all their quotations. Cf. Jo. 9:9. This redundant o[ti may occur before direct questions as in Mk. 4:21; 8:4. It continues common in the koinh, and the modern Greek uses pw/j in this idiom.570 (b) Change of Person in Indirect Discourse. Sometimes this was not necessary, as in Jo. 18:8. So in Mt. 16:18, kavgw. de, soi le,gw o[ti su. ei= Pe,troj, there is no change in the second person. Cf. also Jo. 11:27; Gal. 2:14. But in Mt. 20:10, eno,misan o[ti plei/on lh,myontai, the direct discourse would have lhmyo,meqa. So Lu. 24:23. Compare evla,bomen in Mt. 16:7 with e;cete in v. 8. Note ti, fa,gwmen (direct) in Mt. 6:31, but ti, fa,ghte (indirect) in 6:25. In Mk. 9:6, ouv ga.r h|;dei ti, avpokriqh|/, the direct would be ti, avpokriqw/;
MODE ( EGKLISIS) 1029 The person may be both ways in the same sentence, as in Ac. 1: 4, parh,ggeilen- perime,nein th.n evpaggeli,an tou/ patro.j ha}n hvkou,sate, mou. See further under Mixture. (c) Change of Tense in Indirect Discourse. Mr. H. Scott objects to the wide scope here given to the term "inqirect discourse" to cover "object clauses" after o`ra,wà ktl., but I conceive the principle to be the same. After primary tenses there is, of course, no change in mode or tense. Note Mt. 16:18 above. See also Mk. 11:24, pisteu,ete o[ti evla,bete kai. e;stai u`mi/n) It is only after secondary tenses that any change occurs. Usually even then there is no change of tense in Greek. Thus o[pou h;kouon o[ti e;stin (Mk. 6: 55). So with avkou,saj o[ti basileu,ei- evfobh,qh (Mt. 2:22). So hvlpi,Ä zomen o[ti auvto,j evstin (Lu. 24:21). See also Mt. 21:45; Mk. 6:49, Lu. 1:22; Jo. 2:17; 6:24. Cf. Gal. 2:14, ei=don o[ti ouvk ovrqopodou/Ä sin. So Jo. 11:13. In Jo. 21:19 the future ind. is retained after ei=pen shmai,nwn. Cf. Mt. 20:10. So in Lu. 5:19 the aorist subj. occurs. In Mk. 2:16 we have o[ti evsÄi,ei twice, the first in ind. discourse and the second with interrogative o[ti. But sometimes the ancient Greek, even the Attic,571 used a past tense of the indicative in ind. discourse where the direct had the tenses of present time. The N. T. shows occasionally the same construction. In a case like Jo. 1:50, ei=po,n soi o[ti ei=do,n se, the aorist tense belonged to the direct. Cf. 9:30, 32, 35. So as to the imperfect h=n and aorist avne,bleyen in Jo. 9:18. Cf. also Lu. 13:2. In Mt. 27:18, h|;dei o[ti dia. fqo,non pare,dwkan auvto,n, the aorist is used for antecedent action. Cf. paradedw,keisan in Mk. 15:10. See also Mt. 16:12, o[ti ouvk ei=pen. But in Jo. 2:25, auvto.j ga.r evgi,nwsken ti, h=n evn tw|/ avnÄ qrw,pw|, the direct form572 would have evstin, not h=n. So with h|;dei ti, e;mellen poiei/ngrk grk(6:6); ouvk e;gnwsan o[ti to.n pate,ra auvtoi/j e;legengrk grk(8:27). Cf. also 11:51; 12:16, 33; 18:32. In Ac. 19:32, ouvk h;deisan ti,noj e[neka sunelhlu,qeisan, the past perfect stands when the direct would have the present perfect. In Ac. 16:3, h|;deisan o[ti [Ellhn o` path/r auvtou/ u`ph/rcen, the imperfect may indicate that Timotheus' father was no longer living, though it is not the necessary meaning, as we have just seen. Cf. Mk. 11:32; Jo. 6:22-24; 16:19; Ac. 22: 2; 1 Pet. 1:12. In Ac. 22:29, evfobh,qh evpignou.j o[ti `Rwmai/o,j evstin kai. o[ti auvto.n h=n dedekw,j, we see both constructions combined. In
1030 A GRAMMAR OF THE GREEK NEW TESTAMENT Jo. 11:40, ouvk ei=po,n soi o[ti eva.n pisteu,sh|j o;yh| the subj. and the fut. ind. are retained after secondary tense, unless o[ti is recitative. This preservation of the original tense appears in clauses not strictly in indirect discourse. In Lu. 9:33, ei=pen- mh. eivdw.j oa} le,gei, the present tense is retained in the relative clause oa} le,gei, as it is in the causal clause in 9:49, evkwlu,omen auvto.n o[ti ouvk avkolouqei/ meq v h`mw/n. In Jo. 21:25, cwrh,sein, the future inf. stands for the future ind. in the direct, as teqnhke,nai does in Ac. 14:19 for the perfect ind. In Lu. 20:6 ei=nai really represents the imperfect indicative of the direct. (d) Change of Mode in Indirect Discourse. The rule with the Greek was not to change the tense. The mode after past tenses, with more freedom, was either retained573 or changed to the corresponding tense of the optative mode. The optative, as the most remote in standpoint of the modes, suited this idiom very well. The imperfect and past perfect indicative were, however, retained, though even here the optative sometimes appeared.574 When the aorist optative represented an aorist indicative of the direct discourse the opt. represented past time.575 Usually the optative and subjunctive are future as to time. We have the optative in the N. T. in indirect discourse only in Luke. It was in the koinh, a mark of literary care, almost Atticism, quite beyond the usual vernacular. And with Luke the idiom is almost confined to indirect questions. Luke never has the opt. after o[ti or w`j. Once (Ac. 25:16) in a subordinate temporal clause the optative occurs where the subj. with (cf. Lu. 2:26) or without all would be in the direct, pri.n h' e;coiÄÄte la,boi. And even here ouvk e;stin, after o[ti comes just before. This change in the subordinate clause was also optional in the ancient idiom.576 If a;n was used with the subj. in the direct it was, of course, dropped with the change to the optative in the indirect. Similar to this is the use of eiv and the optative with dependent single clause either as protasis with implied apodosis or purpose like eiv yhlafh,seian (Ac..17: 27); eiv dunato.n ei;hgrk grk(20:16); ei; pwj du,naintogrk grk(27:12). Here after primary tenses we should have eva,n and the subj. or eiv and the future ind. Cf. Ph. 3:12; no. 1:10. Cf. ti, gra,yw in Ac. 25:26. As already explained also, the indirect questions with eiv and the
MODE ( EGKLISIS) 1031 optative (Ac. 25:20; 27:39) are instances where the indicative would be used in the direct. Even in indirect questions Luke usually keeps the mode of the direct. So the indicative as in to. ti,j- dokei/ (Lu. 22:24), the subjunctive as in to. pw/j- avpodw|/ grk grk grk(22:4) or the optative as in to. ti, a'n qe,logrk grk(1:62). The indicative is never changed to a subjunctive as in Latin. When the subj. in Greek occurs in an indirect question it does so because it was the subj. in the direct. Thus ouv ga.r h;dei ti, avpokriqh|/ (Mk. 9:6). Cf. Mt. 6:25, 31, ti, fa,ghteà ti, fa,gwmen. So Lu. 22:2, 4; Ac. 4:21. Cf. subj. with i[na after secondary tenses (Ro. 1:13; 1 Pet. 4:6). The use of the optative (as distinct from subj.) in indir. discourse was a Greek development. We see the beginning of it in Homer. The optative, however, does occur in Lu. Lu.(18:36, W. H. text, margin a;n) in an indirect question where the direct had the indicative. Cf. potapo.j ei;h in 1:29. So 8:9, evphrw,twn ti,j ei;h. In Ac. 21.33, evpunqa,neto ti,j ei;h kai. ti, evstin pepoihkw,j, both constructions occur side by side. The variation here in the mode (retention of the ind.) gives a certain vividness to this part of the question. See Optative in Paratactic Sentences where the koinh, parallels are given. In gi,noito kratei/n pa,shj h-j a'n ai`rh/sqe cw,raj, P. Par. 26 (B.C. 163), there is no sequence of mode. The subj. is with the indefinite relative and the opt. is a wish. It has been already (under Optative) shown that a;n, and the opt. in an indirect question is there because it was in the direct (cf. Ac. 17: 18, ti, a'n qe,loi; with Lu. 1:62, to. ti, a'n qe,loi). Sometimes, one must admit, the difference between the two is reduced to a minimum, as in the papyri occasionally.577 So in Lu. 9:46, to. ti,j a'n ei;h (cf. to. ti,j ei;h) in Lu. 22:23). See also Lu. 15:26; Ac. 10:17. But there is always a shade of difference. The manuscripts reflect this haziness in the variations between ind. and opt. as in Lu. 22:23; Ac. 2:12, et cet. In Lu. 3:15, mh, pote ei;hà we also have the opt. in an indir. question. Radermacher (N. T. Gr., p. 165) quotes Diod. I, 75, 5, evpeida.n- pro,sqoito. The Atticists used it often. (e) The Limits of Indirect Discourse. It is not always easy to draw the line between indirect discourse and other constructions. Thus Jannaris578 uses it only for declarative clauses with o[ti or w`j) Burton579 confines it to indirect assertions and indirect questions, but admits that it also covers indirect commands and promises. Take Mt. 14:7, w`molo,ghsen auvth|/ dou/nai oa} eva.n aivth,shtai. The in-
1032 A GRAMMAR OF THE GREEK NEW TESTAMENT finitive dou/nai is the direct object of the verb and does not seem to be in indir. discourse, for in Mk. 6:23 the direct form has dw,sw. But, after all, it is practical indir. discourse, though the analogy of tense construction breaks down in this instance. But note fut. infinitive with w;mosen in Heb. 3:18, according to the principle of indirect discourse. On the whole it is best to consider three classes or kinds of indirect discourse: declarative clauses, indirect questions, indirect commands. (f) Declarative Clauses (Indirect Assertions). (a) [Oti and the Indicative. There is no clear instance of w`j in this sense in the N. T. It was common in the ancient Greek.580 Just as final o[pwj retreated before i[na, so declarative w`j did before o[ti.581 In late Greek i[na monopolized the field as a final particle and divided it with o[ti as a declarative conjunction. We do have w`j in indirect questions a few times as will be shown. This is more likely the meaning even in Ac. 10:28, evpi,stasqe w`j avqe,miton. Reeb582 points out that Demosthenes uses w`j for what is false and o[ti for what is true. The German wie is used like w`j with verbs of reading, narrating, testifying. With these verbs w`j is more than just on ('that'). [Oti expresses the thing itself and w`j the mode or quality of the thing (Thayer). With this explanation it is possible to consider it as declarative, though really meaning 'how.' Cf. Lu. 24:6, mhn,sqhte w`j evla,lhsen. So in Lu. 8:47 with avpagge,llw, 23:55 after qea,omai, Ac. 10:38 after oi=da, Ac. 20:20 with evpi,stamai, Ro. 1:9 with ma,rtuj (so Ph. 1:8; 1 Th. 2:10). The manuscripts vary in some passages between w`j and o[ti and pw/j. W. H. bracket do in Lu. 6:4 and read pw/j in Mk. 12:26 and o[ti in Jude 1:5, though w`j is retained in 7.583 In all these passages it is possible to regard w`j as the 'how' of indirect question rather than declarative. The encroachment of pw/j on o[ti is to be noticed also. Cf. Mt. 12:4 after avnaginw,skw (and Mk. 12: 26), Mk. 12:41 after qewre,w, Mk. 5:16 after dihge,omai, Lu. 14: 7 after evpe,cwn, Ac. 11:13 after avpagge,llw (so 1 Th. 1:9). In the later Greek pw/j comes gradually to be equivalent to o[ti.584 Gradually pw/j gained the ascendency over o[ti till in the modern Greek it became the regular declarative particle. See Thumb, Handb., p. 190. In Ro. 10:15; 11:33, w`j is exclamatory. The koinh, writers and the papyri show this same retreat of w`j before
MODE ( EGKLISIS) 1033 o[ti and the inroad of pw/j on o[ti (Radermacher, N. T. Gr., p. 159). Cf. B. U., I, 37 (51 A.D.), oi=daj pw/j- crh,zwi Epictetus often after o`ra,w. There is, however, no doubt of the use of w`j o[ti in the declarative sense= 'that.' It is an unclassical combination, but it appears in the LXX (Esther 4:14) and in the koinh, writers.585 It is like the Latin quasi in the Vulgate. The late papyri (fourth cent. A.D.) show that w`j o[ti came in the vernacular to mean simply 'that.'586 Moulton cites also two Attic inscriptions from the first century B.C. which have w`j o[ti in the sense of w`j or o[ti alone. The editors have removed o[ti from w`j o[ti in Xenophon's Hellen. III, ii, 14, eivpw.n w`j o[ti ovknoi,h. Moulton agrees to Blass' stigma of "unclassical" on w`j o[ti, but Paul has koinh, support for his use of it in 2 Cor. 5:19; 11:21; 2 Th. 2:2. But o[ti has won its place in the N. T. not only over w`j, but also over the infinitive. The use of the inf. in indir. discourse587 takes quite a subordinate place in the N. T. Luke alone uses it to any extent. The periphrasis with o[ti has superseded it in nearly all the N. T. writers.588 The use of o[ti is the common way of making a declaration in indirect discourse in the N. T. There arose also dio,ti in the declarative sense589 (cf. late Latin quia=quod), but no example occurs in N. T. The classic causal sense of dio,ti prevailed. It is sometimes doubtful whether o[ti is causal or declarative as in Ac. 22:29. The context must decide. Finally, as noted, pw/j came to be the normal declarative conjunction in the vernacular over the inf. as over w`j and o[ti) as the infinitive disappeared from indir. discourse.590 The only mode used with o[ti in the N. T. is the ind. In Ro. 3:8 (subj.) o[ti is recitative. At bottom o[ti is just o[ ti, and Homer sometimes used o[ te in the declarative sense (and 6). Cf. o[ti o[te together in 1 Cor. 12 The verbs after which o[ti is used in the N. T. cover a wide range. Indeed, o[ti comes also after substantives like avggeli,a (1 Jo. 1:5); kri,sij (Jo. 3:19); lo,goj (Jo. 15:25); marturi,a (1 Jo. 5: 11); ma,rtuj (2 Cor. 1:23); parrhsi,a (1 Jo. 5:14), causal in Ac. 22: 14; fa,sij (Ac. 21:31). It is in apposition also with evn ovno,mati (Mk. 9:41). We see also evn tou,tw| o[ti (1 Jo. 3:16). Some- 1034 A GRAMMAR OF THE GREEK NEW TESTAMENT times o[ti itself seems to imply evn tou,tw| (Ro. 5:8) or peri. tou,tou (Mk. 1:34) or eivj evkei/no (Jo. 2:18). Cf. tou/to o[ti (Rev. 2:6). Another irregularity of construction is the prolepsis of the substantive before o[ti (and change of case) as in 1 Cor. 16:15. This idiom is sometimes called the epexegetic use of o[ti. Cf. further Ac. 9:20. It is a rather common idiom. Cf. Mt. 25:24. See especially Jo. 8:54. In Ro. 9:6 note ouvc oi-on de. o[ti) In 1 Cor. 15:27 dh/lon o[ti is almost adverbial, but that is not true of pro,Ä dhlon o[ti in Heb. 7:14. The elliptical ti, o[ti (Lu. 2:49) may be compared with ti, ge,gonen o[ti, in Jo. 14:22. The elliptical ouvc o[ti (cf. Jo. 6:46) is like the corresponding English "not that." The o[ti clause may be in the nominative (subject clause) as in Mk: 4:38, ouv me,lei soi o[ti avpollu,meqa; More usually it is, of course, in the accusative (object clause) as in Jo. 11:27, pepi,Ä steuka o[ti. The o[ti clause may also be in apposition with the locative as in Mk. 9:41. In Gal. 1:20, ivdou. evnw,pion qeou/ o[ti, we have a solemn oath as in avlh,qeia o[ti (2 Cor. 11:10); pisto.j o[tigrk grk(1:18); pa,rtuj o[ti (2 Cor. 1:23); ovmhu,w o[ti (Rev. 10:6); zw/ evgw,Ã o[ti (Ro. 14:11, LXX). Rarely the personal construction occurs with o[ti as in 1 Cor. 15:12, Cristo.j khr,ssetai o[ti. In Jas. 1:13 we either have recitative o[ti or oratio variata. In Jo. 4:1 we have one o[ti clause dependent on another. [Oti may be repeated in parallel clauses as in Jo. 6:22; Ac. 17:3; 22:29; 1 Cor. 15:3 ff. In 1 Jo. 5:9 we have two examples of o[ti, but one is causal. In Jo. 1: 15 ff. the three are all causal. In Jo. 11:50 we have o[ti and i[na in much the same sense. Not so 1 Jo. 5:13. Cf. i[na in 1 Jo. 5: 3 with o[ti in 5:11. The verbs that use declarative o[ti in the N. T. are very numerous. A few have only o[ti. Thus Mk. 11:32, a[pantej ei=con to.n vIwa,nhn o[ti profh,thj h=n (note h=n). Blass591 calls this use of e;cw a Latinism like habeo. Cf. also u`polamba,nw o[ti (Lu. 7:43), a classical construction. So also lale,w (Heb. 11:18); sumbiba,zw (Ac. 16:10); sfragi,zw (Jo. 3:33); gnwri,zw (1 Cor. 12:3); evmfaÄ ni,zw (Heb. 11:14); evxomologe,w (Ph. 2:11); kathce,w (Ac. 21: 21); khru,ssw (1 Cor. 15:12); avpodei,knumi (2 Th. 2:4); mhnu,w (Lu. 20:37); u`podei,knumi (Ac. 20:35); fanero,omai (2 Cor. 3:3); avpokalu,ptw (1 Pet. 1:12); paradi,dwmi (1 Cor. 15:3); parati,Ä qhmi (Ac. 17:3); profhteu,w (Jo. 11:51). The great mass of the verbs of perceiving, showing (contrary to Attic), knowing, believing, hoping, thinking, saying, declaring, replying, testifying, etc., use either the declarative o[ti or the infinitive. In Lu.
MODE ( EGKLISIS) 1035 9:18 f. with le,gw we have the inf. and o[ti side by side. So also in Ac. 14:22 with parakale,w. Outside of the verbs le,gw, evpimarture,w, katakri,nw and parakale,w the infinitive in indir. discourse in the N. T. is confined to the writings of Luke and Paul and Hebrews according to Viteau,592 "comme vestige de la langue littóraire." But even with Luke and Paul the rule is to use o[ti. Blass593 has a careful list of the uses of these verbs. In margin of W. H. in Jo. 5:15 we have avnagge,llw with o[ti, but the text has ei=pon. But see o[ti also in Ro. 2:4 ( avgnoe,w), Mt. 12:5 ( avnaginw,skw), Lu. 18:37 ( avpagge,llw), Ac. 25:16 ( avpokri,nomai), 1 Jo. 2:22 ( avrne,oÄ mai), Ac. 17:6 ( boa,w), 1 Pet. 2:3 ( geu,omai), Ro. 10:5 ( gra,fw), Mt. 16:21 ( deiknu,w), 1 Cor. 1:11 ( dhlo,w), Ac. 10:42 ( diamartu,romai), Ac. 17:3 ( dianoi,gw), Mk. 8:31 ( dida,skw), Mt. 6:7 ( doke,w), Ac. 9:27 ( dihge,omai), Lu. 24:21 ( evlpi,zw), Mt. 6:26 ( evmble,pw), 1 Cor. 11:2 ( dihge,omai), Ac. 13:32 ( euvaggeli,zomai), Lu. 18:11 ( euvcariste,w), Rev. 2:4 ( e;cw kata, tinoj), Lu. 11:38 ( qauma,zw), Jo. 6:5 ( qea,omai), Ac. 4: 13 ( katalamba,nomai), Lu. 12:24 ( katano,ew), 2 Cor. 5:14 ( kri,nw), 2 Pet. 3:5 ( lanqa,nw), Mt. 3:9 ( le,gw), Ac. 23:27 ( manqa,nw), 2 Cor. 1: 23 ( ma,rtura to.n qeo.n evpikalou/mai), Heb. 7:8 ( marture,w), Ac. 20:26 ( martu,romai), Mt. 27:63 ( mimnh,skw), Mt. 5:17 ( nomi,zw), Mt. 15:17 ( noe,w), Mt. 26:74 ( ovmnu,w), Jas. 1:7 ( oi;omai), Ro. 9:1 ( ouv yeu,domai), 1 Cor. 15:3 ( paradi,dwmi), Heb. 13:18 ( pei,qomai), Jo. 6:69 ( piÄ steu,w), Ro. 4:21 ( plhrofore,w), 2 Cor. 13:2 ( proei,rhka kai. prole,gw, cf. Gal. 5:21), Ac. 23:34 ( punqa,nomai), Lu. 15:6, 9 ( sugcai,rw), Jo. 18:14 ( sumbouleu,w), Ro. 8:16 ( summarture,w), Mt. 16:12 ( suni,hmi), Ju. 5 ( u`pomimnh,skw) 1 Cor. 10:19 ( fhmi,%, Lu. 10:20 ( cai,rw), 1 Tim. 1:12 ( ca,rin e;cw tini,). I cannot claim that this is a complete list, but it is the best I can do with the help of H. Scott, Blass, Thayer, Moulton and Geden, and Viteau's list. At any rate it gives one a fairly clear idea of the advances made by o[ti on the classic infinitive idiom. Some verbs still share the participle with o[ti, but not verbs of showing. These no longer appear in the N. T. with the participle.594 So with o[ti note ble,pw (Heb. 3:19); qewre,w (Mk. 16:4). Cf. Ac. 19:26, qewre,w and avkou,w. So also evpiginw,skw (Lu. 7:37); evpi,stamai (Ac. 15:7); eu`ri,skw (Ro. 7:21); mnhmoneu,w (Ac. 20: 31); o`ra,w (Mk. 2:16). Besides some verbs appear with either o[ti, the infinitive or the participle. Thus avkou,w (Mt. 5:21; Jo. 12:18; Lu. 4:23); ginw,skw (Mt. 21:45; Heb. 10:34; Lu. 8:46); logi,zomai (Ro. 8 : 18; 2 Cor. 10 : 2 both inf. and part.); oi=da (Ac. 16 : 3; Lu. 4:41; 2 Cor. 12:2); o`mologe,w (Mt. 7:23 unless recitative o[ti;
1036 A GRAMMAR OF THE GREEK NEW TESTAMENT Tit. 1:16; 2 Jo. 1:7). In Ac. 27:10 we find used with the infinitive "quite irregularly" Blass595 calls it. But it is just the classic mingling of two constructions seen in the more usual form in Ac. 14:22, where a change is made from the inf. to o[ti and dei/. Different verbs had varying histories in the matter of o[ti. It was not a mere alternative with many. With avkou,w, for instance, o[ti is the usual idiom. The same thing is true with ginw,skwà oi=daà le,gwà nomi,zwà pisteu,w. But with fhmi,, in classical Greek almost always with the infinitive (Ro. 3:8), we twice have o[ti (1 Cor. 10:19; 15:50). For o[ti and then the inf. see Mk. 8:28 f. The substantive nature of the o[ti clause is well shown in 1 Th. 3:6. Radermacher (N. T. Gr., p. 159) cites o[tiÄu`pa,rcein from Proklus' In rem publ., II, 225, 22. The o[ti clause is often called an object clause and may be in the nominative or in the accusative. ( b) The Infinitive. With some verbs we have only single instances of the infinitive of indir. discourse in the N. T. So with boa,w (Ac. 25:24); ginw,skw (Heb. 10:34); katalamba,nomai (Ac. 25: 25); h`ge,omai (Ph. 3:8); noe,w (Heb. 11:3). vApokri,nomai has it only thrice (Lu. 20:7; Ac. 25:4). See also avpagge,llw (Ac. 12: 14); avparne,omai (Lu. 22:34); diiscuri,zomai (Ac. 12:15); dhlo,w (Heb. 9:8); evpagge,llomai (Mk. 14:11; Ac. 7:5); evpimartu,romai (1 Pet. 5:12); katakri,nw (Mk. 14:64); marture,w (Ac. 10:43); proaitia,omai (Ro. 3:9); prokatagge,llw (Ac. 3:18); shmai,nw (Ac. 11:28); crhmati,zw (Lu. 2:26). Some of these are words that are not used with any construction very often, some occur only with the infinitive, like evpideiknu,w (Ac. 18:28); prosdoka,w (Ac. 3:5; 28 6); u`pokri,nomai (Lu. 20:20); u`ponoe,w (Ac. 13:25; 27:27). There is, besides, the inf. with bou,lomaià qe,lwà keleu,w, etc., more exactly the simple object inf. Other verbs that have occasionally the inf. are in the list given under (a), those with either o[ti or the inf. like avrne,omai, (Heb. 11:24); gra,fw (Ac. 18:27); deiknu,w (Ac. 10:28); dida,skw (Lu. 11:1); diamartu,romai (Ac. 18:5); dianoi,gw (Ac. 16:14. Cf. tou/ in Lu. 24:45); euvaggeli,zomai (Ac. 14:15), sumbouleu,w (Rev. 3:18). In Luke and Paul the inf. of indir. discourse is fairly common with le,gw (Lu. 9:18, 20, etc. Cf. Mt. 12:24; Mk. 3: 28) and with nomi,zw (Lu. 2:44; Ac. 7:25, etc.). In the old Greek the inf. was the favourite construction in indirect discourse.596 The Latin had it in all its glory, but the gradual disappearance of the inf. from late Greek made it wither away. Indeed, it was a comparatively late development in Greek
MODE ( EGKLISIS) 1037 anyhow and is rare in Homer.597 It is not easy to draw the line between bou,lomai and keleu,w with the inf. on the one hand and le,gw and nomi,zw with the inf. on the other.598 At bottom the construction is the same. The question of the case of the substantive or adjective used with this inf. is not vital to the idiom. It is really a misnomer to call it "the accusative and infinitive." That is, in fact, more frequently the case found with this inf., but it is so, not because the idiom calls for it per se, but simply because the infinitive can have no subject, not being a finite verb (cf. the participle). Hence when a noun (not the object) occurs with the inf. in indir. discourse it is put in the accusative of general reference, if there is no word in the sentence in another case for it naturally to agree with by apposition. This matter was discussed under Cases, but will bear some repetition at this point since it is so often misunderstood. Clyde599 correctly sees that, since the inf. itself is in a case and is non-finite, it cannot have a subject. Monro600 thinks that the accusative was a late development to assist the "virtual" predication of the later inf. Sometimes this acc. itself is the direct object of the principal verb (so verbs of asking, ad.). Gildersleeve has a pertinent word: "I look with amazement at the retention [by Cauer in his Grammatica Militans] of Curtius' utterly unsatisfactory, utterly inorganic explanation of the acc. c. inf. in oratio obliqua, against which I protested years ago (A. J. P., XVII, 1890, 517): h;ggeilan o[ti o` Ku/roj evni,khse becomes e;ggeilan to.n Ku/ron o[ti evni,khsen, but o[ti evni,Ä khsenÊnikh/sai" (A. J. P., XXXIII, 4, p. 489). To go no further, Gildersleeve shows that the o[ti construction is later than the acc. c. inf. But the grammarians went astray and called this accusative the "subject" of the inf., and, when some other case appears with the inf., it is an "exception" to the rules of the grammarians, though in perfect harmony with the genius of the Greek inf. Even Moulton601 says: "In classical Greek, as any fifth-form boy forgets at his peril, the nominative is used regularly instead of the accusative as subject to the infinitive when the subject of the main verb is the same." Now, there is no doubt about the presence of the nominative in such an instance. But why say "instead of the accusative"? The nominative is normal and natural in such a construction. This construction probably, almost certainly, antedated the accusative with the inf.602 We still
1038 A GRAMMAR OF THE GREEK NEW TESTAMENT meet it in the N. T. The oldest idiom was to have no noun with the inf., as in Lu. 24:23, h=lqan le,gousai kai. ovptasi,an avgge,lwn e`wraÄ ke,nai. The context makes it perfectly clear that the word ovptasi,an is the object of e`wrake,nai and the rest is matter of easy inference. Cf. Ac. 26:9 (with dei/n); Jas. 2:14; 1 Jo. 2:6, 9; Tit. 1:16. In the majority of cases in the N. T. the noun is not repeated or referred to in the predicate. So in Lu. 20:7 we have avpekri,Ä qhsan mh. eivde,nai, but in Ac. 25:4 Fh/stoj avpekri,qh threi/sqai to.n Pau/Ä lon eivj Kaisari,anà e`auto.n de. me,llein. It is easy to see why Pau/lon has to be in the acc. if expressed at all. We could have had auvto,j rather than e`auto,n which probably is just co-ordinated with Pau/lon. Cf. krith.j ei=nai in Ac. 18:15; Mt. 19:21 te,leioj ei=naià Ph. 4:11 e;maqon auvta,rkhj ei=nai, where the principle is the same, though not technically indirect discourse; it is the predicate nominative. So with bou,lomaià qe,lwà zhte,w, etc. The personal construction is a good illustration of the nominative. Cf. Heb. 11:4, evmarturh,qh ei=nai di,kaioj. The nominative occurs also in Ro. 1:22, fa,skontej ei=nai sofoi,. See further Ro. 9:3; 1 Cor. 3: 18; 8:2; 14:37; 2 Cor. 10:2; Heb. 5:12; Jas. 1:26; Jo. 7:4 (W. H. text). In a case like Lu. 20:20 dikai,ouj ei=nai is inevitable because of u`pokrinome,nouj. But there are a good many examples in the N. T. where the nominative could have been properly retained and where the accusative has crept in, perhaps owing to a tendency towards uniformity rather than to any special Latin influence as Blass supposed.603 Moulton604 notes the same tendency in the koinh, outside of Latin influence. Moulton (Prol., p. 249) refers to AEschylus, P. V. 268 f., with the note of Sykes and WynneWilson, and to Adam's note on Plato, Apol., 36 B., for classical examples of acc. with inf. where nom. could have occurred. Cf. Ro. 6:11, u`mei/j logi,zesqe e`autou.j ei=nai nekrou,j. It is rare in the classical Greek for the accusative to occur in such sentences.605 The N. T. undoubtedly shows an increase of the ace. where the nominative was the rule for the older Greek. So Ro. 2:19, pe,poiqaj seauto.n o`dhgo.n ei=nai tuflw/n, where auvto,j (cf. Ro. 9:3) would have been sufficient. Cf. also Ac. 5:36 (cf. 8:9) le,gwn ei=nai, tina e`auto,n, (Ph. 3:13) evgw. evmauto.n ou;pw logi,zomai kateilhfe,nai, (Heb. 10:34) ginw,Ä skontej e;cein e`autou.j krei,ssona u[parxin, (Eph. 4:22) avpoqe,sqai u`ma/j (some distance from the verb evdida,cqhte). See also Ac. 21:1; Ro. 1:20 f. Blass, p. 238, thinks that in 2 Cor. 7:11 the class. Greek would have had o;ntaj, not ei=nai. Even so, but the N. T. has
MODE ( EGKLISIS) 1039 ei=nai. An example like Lu. 20:20 (see above) is hardly pertinent, since the participle on which the inf. depends is itself in the accusative. Cf. 6:4.606 In Ac. 25:21, tou/ Pau,lou evpikalesame,nou threi/sqai auvto,n the pronoun could have been assimilated to the case of Pau,lou ( auvtou/). So also in Lev. 2:9; 3:9, tw/n lego,ntwn vIoudai,ouj ei=nai e`autou,j (different order in 3:9). We find the same lack of assimilation in Ac. 22:17, moi- mouÄÄme, and in 25:27 moi- pe,mponta and in Heb. 2:10 auvtw|/- avgago,nta. In 2 Pet. 3:3, ginw,skontej is clue to anacoluthon (cf. 1:20) as with avpe,cesqai- e;contej (1 Pet. 2:11 f.) and with stello,menoi (2 Cor. 8:20). So Lu. 1:74 h`mi/n r`usqe,ntaj, 5:7 mete,coij evlqo,ntaj. The Greek of the N. T. did sometimes have assimilation of case as in Ac. 16:21, aa} ouvk e;xestin h`mi/n marade,cesqai ouvde. poiei/n `Rwmai,oij ou=sin. So also 15:25, e;doxen h`mi/n genome,noij o`moqumado.n evklexame,noij (- ouj margin of W. H.) pe,myai (cf. accusative retained in verse 22, evklexame,nouj). Cf. also Lu. 1:3; 9:59; 2 Pet. 2:21. Contrast e;doxe, moi of Lu. 1:3 with e;xoxa evmautw|/ of Ac. 26:9. The same situation applies to the cases with the articular infinitive. Cf. Mt. 26:32, meta. to. evgerqh/nai, me proa,xw. Here the me is not necessary and auvto,j could have been used. So with Lu. 2:4, dia. to. ei=nai auvto,n. The auvto,n is superfluous, as in Heb. 7:24.607 Cf. Lu. 10:35, evgw. evn tw|/ evpane,rceÄ sqai, me avpodw,sw soi. See further Lu. 1:57; 2:21; 24:30; Ac. 18:3. It is easy to show from this use of the articular inf. that the inf. has no proper "subject." The accusative is due to other reasons. Take Lu. 2:27, evn tw|/ eivsagagei/n tou.j gonei/j to. paidi,on vIhsou/n, where the context makes plain that paidi,on is the object of eivsagagei/n and gonei/j the acc. of general reference. The article tw|/ must be considered in explaining this instance. Cf. Lu. 18: 5; Ac. 1:3; 27:4; Heb. 5:12 (three accusatives in W. H.'s text). The acc. with the inf. was normal when the substantive with the inf. was different from the subject of the principal verb. Cf. Ro. 3:8, fasi,n tinej h`ma/j le,gein o[ti (note inf. after fhmi,, and o[ti after le,gw, but it is recitative o[ti. In Lu. 24:23, le,gousin auvto.n zh/n we see le,gw with the acc. and inf. Typical examples are seen in Mt. 17:4, kalo,n evstin h`ma/j w-de ei=nai Ac. 12:14; 14:19; 16:13; 24:15; 1 Pet. 3:17; 5:12; 1 Cor. 14:5; Heb. 9:8. See further Verbal Aspects of Inf., (d), in next chapter. The tense of the original is preserved in the inf. as a rule. A case like Mt. 14:7, w`molo,ghsen auvth|/ dou/nai oa} eva.n aivth,shtai, may
1040 A GRAMMAR OF THE GREEK NEW TESTAMENT seem a bit disconcerting since in the direct discourse in Mk. 6:23 we find dw,sw. But the future is aoristic anyhow. The line between indir. discourse and the simple object inf. is not sharply drawn. Cf. Ac. 23:12. In Lu. 20:6, pepeisme,noj ga,r evstin vIwa,Ä nhn profh,thn ei=nai, the inf. represents h=n of the direct. There was no help for this, since there is no imperfect inf. The future inf. in indir. discourse is rare, but see Jo. 21:25; Ac. 23:30 (see Tenses). Examples of the perfect inf. in this idiom occur in Ac. 12:14; 14:19; 16:27; 25:25; Heb. 9:8. Cf, o`mologei/ eivlhfe,naià P. Oxy. 37 (A.D. 49). There is little more to say. The use of tou/ and the inf. as subject has been noted (pp. 996, 1002). See tou/ evlqei/n, Lu. 17:1, where ta. ska,ndala is the acc. of general reference while this genitive inf. is itself in the nominative case. See also Ac. 10:25. We do not have a;n with the inf. in indir. discourse. In 2 Cor. 10: 9, i[na mh. do,xw w`j a'n evkfobei/n, we have w`j a;nÊas if.' It is not the a;n in apodosis. Nestle in his N. T. gives at 1 Pet. 5:8 zhtw/n ti,na katapiei/n, but surely tina. is the correct accent. W. H. places even this in the margin. Souter prints tina., departing from R. V. which has tina. But Radermacher (N. T. Gr., p. 147) cites Callinicus in Vita Hypatii, 57, 12, pou/ eu`rei/n, and 113, 11, ti, poih/sai (cf. German Was tun?). It may be worth while to add that frequently we meet an inf. dependent on an inf. (cf. inf. on part. in Lu. 20:20). I have noticed the idiom in Luke, Paul, Mk., Heb. Cf. Lu. 6:12, evxelqei/n auvto.n eivj to. o[roj proseu,xasqai, where the first is in indirect discourse, and Ac. 18:2, dia. to. diatetace,nai Klau,Ä dion cwri,zesqai pa,ntaj tou.j vIoudai,ouj, where the second is indirect discourse (indir. command). Cf. Ro. 15:8. ( g) The Participle. Middleton608 suggests that the use of the participle in indir. discourse is older than the inf. This may be true, since in the Sanskrit it developed much more rapidly than the inf. But there were cross-currents at work in indirect discourse. Just as the inf. was circumscribed by the declarative o[ti, so the participle was limited by o[ti or the infinitive. Thus verbs of showing ( dei,knumià dhlo,w) and of manifesting ( fanero,w) no longer occur with the participle in the N. T. However, we have the participle with fai,nomai ('appear'), as in Mt. 6:16. Besides, the participle has disappeared from use with aivsqa,nomaià manqa,nwà me,Ä mhnmaià suni,hmi. The participles with manqa,nw in 1 Tim. 5:13 are additional statements, as the Revised Version correctly translates. With the inf. manqa,nw means 'to learn how,' not 'to learn that.'
MODE ( EGKLISIS) 1041 Cf. Ph. 4:11; Tit. 3:14. But some verbs in the N. T. still have the participle in indir. discourse. They are verbs of perception by the senses (hearing, seeing, knowing). In the ancient Greek the nominative was used when the participle referred to the subject of the verb. Thus o`rw/ h`marthkw,j meant 'I see that I have sinned.' In the N. T., however, we have declarative o[ti in such clauses (Mk. 5:29; 1 Jo. 3:14).609 Viteau610 rightly insists on a real difference between the participial conception and the declarative o[ti or the inf. If the idea is one of intellectual apprehension merely, an opinion or judgment, we have o`rw/ o[ti (Jas. 2:24). If it is a real experience, the participle occurs as in Mk. 8:24, w`j de,ndra o`rw/ peripateou/ntaj. So in Ac. 8:23, his eivj su,ndesmon o`rw/ se o;nta. There is something in this distinction. Cf. ble,pw o[ti (Jas. 2:22), but the participle in Heb. 2:9, vIhsou/n evstefanwÄ me,non. In Mk. 8:24 we have o[ti with ble,pw and the part. with o`rw/. The realistic quality of the part. is finely brought out in Mk. 9:1, e[wj a'n i;dwsin th.n basilei,an tou/ qeou/ evlhluqui/an evn duna,mei) Note the tense as in Lu. 10:18, evqew,roun to.n Satana/n- peso,nta. Cf. 9:49; 21:20; Ac. 11:13; 17:16. See Jo. 19:33, w`j ei=don h;dh auvto.n teqnhko,ta) The tense of the direct is preserved. See for qewre,w, Mk. 16:4 and Lu. 24:39, kaqw.j evme. qewrei/te e;conta. For evpi,stamai take Ac. 15:7 and 24:10. Cf. also mnhmoneu,w with o[ti. (Ac. 20:31) and the part. (2 Tim. 2:8). It is very clear in eu`ri,skw (see o[ti in Ro. 7:21) which, as in classic Greek, is commonly used with the participle. See Mt. 1:18; 12:44; Lu. 23:2; Ac. 9:2. In Mt. 1:18 we have the passive construction eu`re,qh e;cousa. In Lu. 23:2 we find three participles. Dokima,zw in the N. T. has only the inf. (Ro. 1:28) and the participle (2 Cor. 8:22). So with h`ge,omai (Ph. 2:6; 3:7). Cf. also e;ce me parh|thÄ me,non (Lu. 14:18). In 2 Jo. 1:7 note the part. with o`mologe,w. In verse 4, peripatou/ntaj with eu`ri,skw, the case agrees only in sense with evk tw/n te,knwn. The difference between o[ti with oi=da (Ac. 23: 5) and the part. is clear (2 Cor. 12:2), though this is the only instance of the part. with this verb. It prefers o[ti, but may have the inf. (Lu. 4:41). The difference is even clearer in ginw,skw. See o[ti in Mt. 21:45, the inf. in Heb. 10:34. The usual idiom is o[ti, but note Lu. 8:46, e;gnwn du,namin evxelhluqui/an avp v evmou/, where Christ thus graphically describes the terrible nervous loss from his healing work. He felt the power "gone" out of him. In our vernacular we speak of a sense of "goneness." See also Ac. 19:35; Heb. 13:23. But see Mk. 5:29, e;gnw tw|/ sw,mati o[ti i;atai.
1042 A GRAMMAR OF THE GREEK NEW TESTAMENT Mk. 5:30 evpiginw,skw has the attributive participle after it. vAkou,w also occurs with declarative o[ti (Mt. 5:21; 32 times), the inf. (Jo. 12:18; 1 Cor. 11:18) or the part. (Ac. 7:12; 14: 9; 3 Jo. 1:4; 2 Th. 3:11, etc.). These examples have the accusative when the thing is understood. Blass611 curiously calls the acc. incorrect in Ac. 9:4; 26:14. The genitive with fwnh, does occur in 11:7; 22:7. Blass has an overrefinement on this point. As with the acc. construction of the part. with avkou,w, so most of the genitive examples are found in the Acts. So 2:6; 6:11; 14:9, etc. But see also Mk. 12:28, avkou,saj auvtw/n suzhÄ tou,ntwn. So 14:58; Lu. 18:36; Jo. 1:37. The perfect part. in this construction is seen in Lu. 8:46; Jo. 19:33, etc. For the aorist see Lu. 10:18. In Mk. 6:8 we have oratio variata. The sentence starts with i[na and concludes with the inf. Hence the part. u`podedeme,nouj is construed with the inf. See the acc. part. in Rev. 4:4 as explained by ei=don in verse 1, though ivdou, and the nominative have come between. ( d) Kai. evge,neto. One hardly knows whether to treat this construction as indirect discourse or not. It is a clear imitation of the Hebrew yhiy.w; and is common in the LXX with two constructions. It is either kai. evge,neto kai, with finite verb (or evge,neto de,) as in Gen. 24:30; 29:13; Josh. 5:1, etc.), or we have asyndeton, kai. evge,neto plus finite verb (Gen. 22:1; 24:45, etc.). For evge,neto we often find evgenh,qh (1 Sam. 4:1; 11:1, etc.). This asyndeton is also common in the future as kai. e;stai with finite verb (Is. 9:16; 10:20, 27, etc.). This kai. e;stai construction is quoted a few times in the N. T. (Ac. 2:17, 21; Ro. 9:26) from the LXX. For kai. e;stai kai, see Ex. 13:11 f. W. F. Moulton612 has pointed out that the idiom occurs when the principal sentence has some note of time. J. H. Moulton613 quotes Driver (Tenses, § 78) as describing the yhiy.w; construction in a similar fashion, "a clause specifying the circumstances under which an action takes place." All the examples of these two constructions in Luke fit this description. Luke has in the Gospel eleven of the kai. evge,neto kai, examples and twenty-two of the kai, evge,neto type. For kai. evge,neto kai, see Lu. 17:11; without the second kai, 17:14. See in particular Lu. 8 and 9. It is frequently the case that Luke has evn tw|/ and the inf. with the idiom. So 9:51, evge,neto de. evn tw|/ sumplhrou/sqai- kai. auvto.j evsth,risen. Here kai. is almost equivalent to o[ti. So kai. evge,Ä neto evn tw|/ ei=nai- ei=pe,n tij grk(11:1). We have kai. evge,neto kai, also in
MODE ( EGKLISIS) 1043 Mt. 9:10. The form kai. evge,neto Moulton614 counts outside of Luke only twice in Mark and five times in Matthew with the phrase evge,neto o[te evte,lesen. Cf. Mt. 7:28. Moulton is concerned to show against Dalman that the idiom is not Semitic. He admits the Hebraism in kai. evge,neto kai,, but doubts as to kai. evge,neto (asyndeton). But surely the LXX has left its mark in this point also. The LXX does not have evge,neto (or gi,netai) and the infinitive (but cf. 2 Macc. 3:16 h=n - titrw,skesqai). In the N. T. we find it in Mt. 18:13; Mk. 2:15; five times in Luke and seventeen times in Acts. Cf. u`mi/n gi,noito kratei/n, P. Par. 26 (B.C. 163-2). The other two constructions are absent from the Acts, showing that in the Gospel Luke was more directly using Semitic sources or imitating the LXX on the point. But even inf. with evge,neto is not ancient Greek, which used sune,bh. We have sune,bh and the inf. in Ac. 21:35. The modern Athenian vernacular has sune,bh o[ti on while the country districts615 use e;tuce na,. Moulton finds the inf. with gi,netai in the papyri and rightly in the vernacular koinh, the origin of this idiom. There is no essential difference between the inf. with gi,netai and evge,neto. Cf. Ac. 9:32; 16:16; 9:32, 37, 43; 11:26, etc. Outside of Luke (Gospel and Acts) the inf. with evge,neto is confined to Mk. 2:23, which Moulton calls "a primitive assimilation of Lu. 6:1." See Ac. 10:25, evge,Ä neto tou/ eivselqei/n. This is Moulton's presentation, which is certainly more just than the mere description of "Hebraism" for all these constructions.616 We do not have the o[ti clause with gi,netai or evge,neto in the N. T. (g) Indirect Questions. (a) Tense. See (c) under Indirect Discourse. It may here be simply stated that when the principal verb is primary no change in tense occurs. When it is secondary, still no change appears as a rule, though occasionally one does see it, as in, Jo. 2:25; 6:6; 18:32. But note evpunqa,neto pou/ genna/tai (Mt. 2:4); evqew,roun pou/ te,qeitai (Mk. 15:47). Cf. Ac. 10:18. Note difference between present perfect in Mk. 15:44 and the aorist in the same verse. For the future ind. see Jo. 21:19; Mk. 11:13. ( b) Mode. It is only necessary to say that as a rule the same mode is retained in the indirect question that was in the direct. Thus see Mk. 5:14; 15:47; Lu. 8:36; 23:55; Ac. 10:29, where the indicative occurs. We have the ind. after secondary as well as primary tenses. This is the common idiom in the N. T. as in
1044 A GRAMMAR OF THE GREEK NEW TESTAMENT the koinh,) In all instances where a subj. appears in this construction it is due to the fact that the subj. would have been present in the direct (deliberative subj.). Note ti, fa,gwmen; in Mt. 6:31 and ti, fa,ghtegrk grk(6:25). See also pou/ me,neij; of Jo. 1:38 and ei=dan pou/ me,nei of verse 39 for the retention of the indicative. The Latin changed the ind. to the subj. in indirect questions, but the Greek did not. This deliberative subj. occurs after primary tenses as in Lu. 9:58, ouvk e;cei pou/ th.n kefalh.n kli,nh|, and after secondary tenses also as in Mk. 9:16, ouv ga.r h|;dei ti, avpokriqh|/. Cf. also Mk. 6:36; Lu. 5:19; 12:36. So also the optative occurs a few times where it was in the direct. This is the construction with a;n which has already been discussed twice. See Ac. 17:18, ti, a'n qe,loi, for the direct form, and Lu. 1:62, ti, a'n qe,loi, for the indirect. Cf. Lu. 9:46; Ac. 5:24. In 2 Tim. 2:25, mh, pote dw|,h (W. H. have dw,h| in margin), we have the optative without a'n after a primary tense if dw|,h be correct. Moulton617 considers the subj. here a "syntactical necessity." We heed not moralize, therefore, on this instance of the optative even if it is genuine. Radermacher (Neut. Gr., p. 132) shows that the Atticists frequently used the opt. after a primary tense, as copyists often fail to catch the spirit of a thing. The papyri (ib.) have some illustrations of the same idiom. The other examples of the opt. in indirect questions are all after secondary tenses and the change is made from an indicative or a subj. to the optative. These examples all occur in Luke. As instances of the opt. where the direct had the incl. see Lu. 1: 29; 3:15; 18:36. See Ac. 21:33 for both modes. In Ac. 17: 27, eiv a;rage yhlafh,seian, the opt. represents a subj. with eva,n after a primary tense. So in Ac. 27:12. In no instance where the opt. without a;n occurs in the indirect discourse is it necessary. In all these examples the indicative or the subj. could have been retained. The infinitive with ti,na in 1 Pet. 5:8 is read by Nestle, but not by W. H. or Souter. See under (f),( b). ( g) Interrogative Pronouns and Conjunctions Used. One notes at once the absence of o[stij in this construction, the common classic idiom. We do have o[ti once in Ac. 9:6, lalhqh,setai, soi o[ti se dei/ poiei/n. Elsewhere the most usual pronoun is ti,j and ti, as in Ac. 10:29; 21:33. We even have ti,j ti, a;rh| in Mk. 15:24 (double interrogative). Tischendorf reads ti,j ti, in Lu. 19:15, but W. H. have only ti,. Thin double use appears rarely in the older Greek.618 As a rule the distinction between ti,j and o[j is pre-
MODE ( EGKLISIS) 1045 served in indirect questions, as in Jo. 13:24 (cf. 13:12). The occasional confusion between ti,j and o[j was discussed under Pronouns. See 1 Tim. 1:7 and Jas. 3:13. Now and then the simple relative pronoun or adverb is used in an indirect question, as was true of classical Greek also. So Mk. 5:19 f. o[sa, Lu. 8:47 di v ha}n aivti,an, Ac. 15:14 kaqw,j, 1 Th. 1:5 oi-oià and the various examples of w`j discussed in connection with Indirect Assertions (Lu. 8:47; Ac. 10:28, 38, etc.) which are more likely to be understood in the sense of 'how,' and so indirect questions. Cf. Lu. 6:3 f. ( o[ and w`j), Mt. 10:19 ( doqh,setai pw/j h' ti, lalh,shte) Lu. 17:8 ( ti,%. Other interrogative words used are pou/ (Mt. 2:4), po,qen (Jo. 8:14), poi/oj (Rev. 3:3), po,te (Lu. 12:36), pw/j (Lu. 8:36), phli,koj (Gal. 6:11), po,soj (Mt. 16:9), potapo,j (Lu. 1:29). The correlative words, besides the lone instance of o[ti in Ac. 9:6, are o[pwj (Lu. 24:20), o`poi/oj (1 Th. 1:9). In Mk. 14:14 (Lu. 22:11) pou/- o[pou fa,gw; most likely the o[pou clause is an indirect question with the deliberative subj., but it may be the volitive subj. simply. There are plenty of instances of eiv in indirect questions (see Conditional Sentences) as in Mk. 15:44 after qauma,zw and evperwta,w; Lu, 14:28 after yhÄ fi,zw; 14:31 after bouleu,omai; Mt. 26:63 after ei=pon; 27:49 after o`ra,w; Mk. 3:2 after parathre,w; Jo. 9:25 after oi=da; Ac. 4:19 after kri,nw; 10:18 after punqa,nomai; 19:2 after avkou,w; 2 Cor. 2:9 after ginw,skw; 13:5 after peira,zw. There are, besides, those passages619 where a word is suppressed, like Mk. 11:13; Eph. 3:2; Ph. 3:12; 2 Th. 2:15. See also the optative with eiv in Ac. 17:27; 25:20; 27:12. This is all quite classical and gives no trouble. We find mh, also used like an indirect question after skoÄ pe,w (cf. p. 995) with the ind. (Lu. 11:35) and mh, pote after dialogi,Ä zomai with the opt. (Lu. 3:15). In Jo. 7:17 an alternative indirect question occurs with po,teron- h;) The only other alternative construction in an indirect question is in 2 Cor. 12 : 2 f. after oi=da, and is ei;teÄÄei;te. In all these points the N. T. is in harmony with the koinh,. The use of ti, with the subj. (Mk. 6:36) or the future ind. (Ac. 25:26 possibly subj. aor.) may be compared with pou/ after e;cw in Lu. 9:58. In Col. 4:6 pw/j after eivde,nai is to be distinguished from the use of the inf. after oi=da ('know how to do.' Cf. Lu. 11:13). In Mk. 2:24, i;de ti, poiou/sin; the i;de is probably just the interjection as in Mt. 25:25. For the acc. and the ind. question side by side see Mt. 16:9. ( d) The Article with Indirect Questions. This classical idiom
1046 A GRAMMAR OF THE GREEK NEW TESTAMENT appears in Luke and Paul. See to. ti, (Lu. 1:62), to. ti,jgrk grk(9:46), to. pw/jgrk grk(22:4). So Paul has to. pw/j in 1 Th. 4:1 and to. ti, in Ro. 8:26 (cf. ti, to, in 8:27). See also 22:23 f.; Ac. 4:21; 22:30. The substantive nature of the indirect question is well shown also in Jo. 4:10. Cf. Lu. 24:19 f. (h) Indirect Command. As already explained, this construction is somewhat vague and the line is hard to draw between this and other idioms. (a) Deliberative Question. A direct command may be turned into a deliberative question in the indirect with the subjunctive. The volitive idea of the imperative thus glides into the deliberative. In Lu. 12:5, u`podei,xw de. u`mi/n ti,na fobhqh/te\ fobh,qhte to.nà ktl), we have the point illustrated both in the direct (imperative) and the indirect (deliberative subj.). Here the only difference between the two forms is the accent. Cf. mh. fobhqh/te in verse 4. In Mt. 10:28 we have fobei/sqe. Obviously this is a natural, though not very frequent, turn for the Command to take. ( b) The Conjunctions i[na and o[pwj. These may be used after verbs of commanding and beseeching. This idiom does not differ clearly from the sub-final construction. It is a species of purpose (or sub-final. See Final Clauses). The examples there given might suffice, but note the following: Mk. 6:8 parh,ggeilen auvtoi/j i[na mhde.n ai;rwsin, Mt. 16:20 evpeti,mhsen toi/j maqhtai/j i[na mhdeni. ei;pwÄsin, 2 Th. 3:12 paragge,llomen kai. parakalou/men evn kuri,w| vIhsou/ Cristw|/ i[na- evsqi,wsin, Ac. 25:3 aivtou,menoi o[pwj metape,myhtai. See further Mt. 8:34; Lu. 16:27; 1 Cor. 1:10. In Lu. 16:27 f. we have the purely final idea in both o[pwj and i[na which are subordinate to the first i[na after evrwtw/. But we cannot follow this use of i[na after qe,lw and such verbs where it is more or less purely objective. The recitative o[ti with the imperative in 2 Th. 3:10 is not an instance of indirect command, but simply the direct command preserved. ( g) The Infinitive. It seems more obvious and is still common in the koinh,, though retreating before i[na. The negative is, of course, mh,. This use of the infinitive must not be confounded with the idiom for indirect assertion (declarative) as in Mk. 12: 18, oi[tinej le,gousin avna,stasin mh. ei=nai. Note Ac. 21:21, le,gwn mh. perite,mnein auvtou.j ta. te,kna mhde. toi/j e;qesin peripatei/n, where we have prohibition, not assertion (note incidentally the two Accusatives) with le,gwn (same verb as above). So also 23:12, le,gontej mh,te fagei/n mh,te pei/n. Cf. 21:4. Simple enough is the construction after ei=pa in Lu. 9:54, ei;pwmen pu/r katabh/nai; See also Mk. 8:
MODE ( EGKLISIS) 1047 7. In Mt. 16:12, sunh/kan o[ti ouvk ei=pen prose,cein (cf. prose,cete in verses 6 and 11), we have the declarative o[ti and the indicative followed by the inf. in indirect command. In im. 2:26, h=n auvtw|/ kecrhmatisme,non mh. ivdei/n qa,naton, the construction is like that of indirect command, but the sense comes nearer to the mere object infinitive. See the direct dw,sw in Mk. 6:23 reproduced in the indirect by dou/nai (Mt. 14:7). There is a certain amount of freedom taken in such transference to the indirect. In Ac. 18:2, dia. to. diatetace,nai Klau,dion cwri,zesqai pa,ntaj, the inf. is dependent on an inf. Other instances of the inf. in indir. command are seen in Ac. 25:24, bow/ntej mh. dei/n auvto.n zh/nà 26:20, avph,ggellon metanoei/n. In 2 Th. 3:6 we have paragge,llomen ste,llesqai, while in verse 12 we have i[na. In verse 10 the direct quotation follows this same verb. In Mk. 6:8 f. we have both i[na mh. ai;rwsin and mh. evndu,sasqai (marg. of W. H., Mh. evndu,shsqe) after parh,ggeilen. Luke Luke Luke(9:3-5) gives it all in the direct form. In 2 Th. 3:14, tou/ton shmeiou/sqeà mh. sunanami,gnusqai auvtw|/, the inf. is not in indirect command, but rather the inf. used in the direct as the equivalent of the imperative. But in 1 Cor. 5:11, e;graya u`mi/n mh. sunanami,gnuÄ sqai (so also verse 9), we de have indirect command. (i) Mixture. Strictly this point belongs to the chapter on Figures of Speech (cf. also, Oratio Variata, The Sentence), but a word is called for here. We have mixture of several sorts as in the classic Greek. In Ac. 19:1 f., Pau/lon evlqei/n kai. eu`rei/n, ei=pe,n te, we have the infinitive (object-clause subject of evge,neto) and the finite clause ei=pe,n te side by side. Cf. Ac. 4:5 f. for inf. followed by kai, and the indicative. So in Lu. 9:19 we have the infinitive construction and the o[ti construction side by side after avpokriqe,ntej ei=pan. In Ac. 14:22, parakalou/ntej evmme,nein th|/ pi,stei kai. o[ti- dei/, the construction glides from the inf. into o[ti. In Ro. 3:8 the recitative o[ti is dependent on the inf. le,gein after fasi,n. In Ac. 9:27, dihgh,santo pw/j evn th|/ o`dw|/ ei=den to.n ku,rion kai. o[ti evla,lhsen auvtw|/à kai. pw/j ktl., we have a change from ind. question to indirect assertion and then back again to indirect question. The change may be from the indirect to the direct as in Ac. 1:4, perime,nein th.n evpaggeli,an tou/ patro.j ha}n hvkou,sate, mou. Cf. also 23:22. See also Jo. 12:29. This change appears in Mk. 6:8 f., if the true text is evndu,shsqe. But the change may be just the reverse, from the direct to the indirec is in Ac. 23: 23, ei=pen `Etoima,sate- kth,nh te parasth/sai. In 27:10 o[ti occurs with the inf., a mixture of the o[ti and the infinitive constructions in indirect assertions. This use of o[ti with the inf. appears in.
1048 A GRAMMAR OF THE GREEK NEW TESTAMENT classic Attic (cf. Xen., Cyr., 1, 6, 18, etc.). See Jannaris, Hist. Gk. Gr., p. 570. Moulton (Prol., p. 213) gives a papyrus example, 0. P. 237 (ii/A.D.), dhlw/n o[ti eiv ta. avlhqh/ fanei,h mhde. kri,sewj dei/sqai to. pra/gma. See further Winer-Moulton, p. 426. (j) The Subordinate Clause. A complex sentence may be quoted in indirect discourse as readily as the simple sentence. This principal clause follows the usual laws already discussed. Secondary tenses of the indicative in the subordinate clause suffer no change at all in mood or tense.620 This is obviously true after primary tenses, as in Gal. 4:15, marturw/ u`mi/n o[ti eiv dunato,n ÄÄevdw,kate, moi. Here the copula h;n is suppressed. In Lu. 19:15 note ei=pen fwnhqh/nai- oi-j dedw,kei. So after primary tenses the primary tense follows, as in Mk. 11:23, le,gw o[ti oa}j a'n ei;ph|ÄÄe;stai auvtw|/. Cf. Ac. 25:14 f. But even after secondary tenses the rule is to retain the tense and mode of the direct much more than in the Attic where the mode was quite optional.621 See Lu. 9:33, ei=pen mh. eivdw.j oa} le,gei. Another example of the relative clause appears in Mt. 18:25, evke,leusen- praqh/nai- kai. o[sa e;cei. Even after a condition of the second class the primary tense may be retained, as in Lu. 7 39, evgi,nwsken a'n ti,j kai. potaph. h` gunh. h[tij a[ptetai auvtou/ o[ti a`martwlo,j evstin. For a causal sentence see evkwlu,omen auvto.n o[ti ouvk avkolouqei/ meq v h`mw/n (Lu. 9:49). A temporal clause with the subjunctive appears in Mt. 14:22, hvna,gkasen- proa,gein- e[wj ou- avpolu,sh|. See also Ac. 23:12 avneqema,tisan- e[wj ou- avpoktei,nwsin In 25:16, however, we have the optative in the subordinate clause of time with pri.n h' $e;coià la,boi), after avpekri,qhnà the sole example. It is in Luke, as one would expect. The change here is from the subj. to the opt. In Lu. 7:43, o[ti w|-, only the subordinate relative clause is given. 10. SERIES OF SUBORDINATE CLAUSES. It is interesting to observe how rich the Greek language is in subordinate clauses and how they dovetail into each other. It is almost like an endless chain. The series may run on infinitum and yet all be in perfect conformity to the genius of the language. I have collected quite a number of examples to illustrate this complexity of structure, some of which are here given. A typical one is Mk. 11:23. After le,gw o[ti we have o[j a'n ei;ph| which has oratio recta, but the relative clause proceeds with kai. mh. diakriqh|/ avlla. pisteu,h| o[ti oa} lalei/ gi,netai. The relative oa} lalei/, is the fourth involution of subordinate clauses after le,gw. Cf also Jo. 17:24. A similar multiplicity of subordinate clauses is found in Ac. 25:14-16.
MODE ( EGKLISIS) 1049 After avne,qeto le,gwn we have oratio recta. The first step is the relative clause peri. ou-ÄÄevnefa,nisan, on which hangs pro.j oua}j avpekri,qhn, which in turn is followed by o[ti ouvk e;stin and that by cari,zesqai, and this again by pri.n h' e;coiÄÄla,boi. The pri.n h; clause is the fifth involution in the oratio recta. Cf. also Ac. 3:19 ff. ( pro.j to. evxalifqh/naià o[pwj a'nà oa}n dei/ de,xasqaià w-n). In Ac. 11:13 there are five involutions. The complications axe not, of course, always so many. In Lu. 7:39 the oratio recta has a series of three ( ti,j h[tij - o[ti). See the threefold series in Ro. 3:8, kaqw,j fasi,n tinej h`ma/j le,gein o[tià ktl) So also Mk. 6:55, perife,rein o[pou h;kouon o[ti e;stin (infinitive, relative, declarative). So again 1 Cor. 11:23 f. ( o[tià h|-à ei=pen and oratio recta). Here also the o[ clause is in apposition with the o[ti clause. Cf. Lu. 19:15 (inf., i[naà ti,%. In Ac. 7:25, evno,mizen sunie,nai tou.j avdelfou.j o[tià ktl., we have two forms of indirect assertion (the inf., then o[ti), one dependent on the other. So also o[ti follows dia. to. le,gesqai in Lu. 9:7 f. In 4:10 we have the o[ti clause and then the articular inf. In Jo. 6:24 the o[ti clause is subordinate to the o[te clause. In 1 Jo. 5:9 we have a o[ti clause dependent on a o[ti clause. In Jo. 4:1 we have w`j o[ti- o[ti. In Mt. 16:20 the sequence is i[na- o[ti) So Jo. 16: 4; 17:23. In Mk. 14:14 we have two cases of oratio recta, one dependent on the other. In Lu. 24:7 it is w`j- o[ti. Cf. i[na i[na in Gal 3:14. In Col. 1:9 the i[na clause and the infinitive peripath/sai are parallel. The instances are numerous where one infinitive is dependent on another infinitive. Thus evxelqei/n proseu,xasqai (Lu. 6:12); doqh/nai fagei/ngrk grk(8:55); pro.j to. dei/n proseu,Ä cesqai grk(18:1); dia. to. tetace,nai Klau,dion cwri,zesqai, after evlhluqo,ta (Ac. 18:2); dei/n pra/xaigrk grk(26:9); gegenh/sqai eivj to. bebaiw/sai (Ro. 15: 8); kathrti,sqai eivj to. gegone,nai (Heb. 11:3). In Ac. 23:30, mhnuÄ qei,shj moi evpiboulh/j eivj to.n a;ndra e;sesqai, the future inf. in indirect discourse is dependent on the participle in the genitive absolute. In Heb. 9:8, tou/to dhlou/ntoj tou/ pneu,matoj tou/ a`gi,ou pefanerw/sqai, the perfect inf. follows the genitive absolute. There are various other combinations. These are given as illustrations. No rules are called for about the using of a series of subordinate clauses. The presence of so many of them in Luke, Paul and Hebrews shows the literary quality of a more periodic structure. 1 Jann., Hist. Gk. Gr., pp. 445 ff., has this plan. I had already made my outline before reading his treatment of the subject.
2 Thompson, Synt. of Att. Gk., p. 185.
3 Cf. Bru., Griech. Gr., p. 498; K.-G., I, p. 200; Stahl, Krit.-hist. Synt., p. 220. See Sandys, Hist. of Class. Scholarship, III, p. 458.
4 Farrar, Gk. Synt., p. 136.
5 Gildersl., "A Syntactician among the Psychologists," Am. Jour. of Jan., 1910, p. 74.
6 Cf. Steinthal, Gesch. d. Sprachw., pp. 309, 628.
7 Am. Jour. of Philol., XXIII, p. 127; XXX, p. 1.
8 Ib., XXX, p. 1; Synt. of Classic. Gk., p. 79.
9 Thompson, Synt., p. 510.
10 Moulton, Probl, p..165.
11 Ib. Cf. also Thompson, Synt. of Attic Gk., p. 510. The injunctive had "a meaning hovering between the imperative, conjunctive and optative."
12 Giles, Man., 459.
13 Thompson, Synt., p. 494. In the Sans. it was the subjunctive that went down in the fight. Cf. Whitney, Sans. Gr., p. 201 f.
14 Ib., p. 495.
15 Am. Jour. of Philol., Jan., 1909, p. 2.
16 Blass, Gr. of N. T. Gk., p. 205.
17 Bernhardy, Wiss. Synt. der griech. Sprache, p. 384.
18 Vandacle, L'Otatif Grec, 1S97, p. 111.
19 K.-G., Bd. I, p. 201.
20 Ib. Der Redende stellt etwas als wirklich.
21 Jann., Hist. Gk. Gr., p. 445.
22 Riem. and Goelzer, Synt., p. 297 f.
23 Burton, and T., p. 73.
24 Jann., Hist. Gk. Gr., p. 445.
25 Blass, Gr. of N. T. Gk., p. 205.
26 Viteau, Etude sur le Grec du N. T. Le Verbe, p. 22. Some editors read a;ra in Gal. 2:17, but see Blass, Gr. of N. T. Gk., p. 259. See a;ra in Mt. 18:1.
27 Cf. Blass, Gr. of N. T. Gk., p. 260.
28 W.-Th., p. 509.
29 Robertson, Short Gr. of the Gk. N. T., p. 179.
30 Viteau, Le Verbe, p. 22.
31 Cf. Robertson, Short Gr. of the Gk. N. T., p. 178.
32 See ch. XV, Pronouns.
33 Am. Jour. of Philol., Jan., 1910, p. 78.
35 Blass, Gr. of N. T. Gk., p. 254.
36 Burton, N. T. M. and T., p. 179.
37 Cf. Gildersl., Am. Jour. of Philol., Jan., 1910, p. 78.
38 Moulton, Prol., p. 199.
39 Synt. of Classic Gk., Pt. I, § 365.
40 Monro, Hom. Gr., p. 235.
41 Thompson, Synt. of Attic Gk., p. 186.
42 W.-Th., p. 283.
43 K.-G, Bd. I, p. 204 f.
44 Blass, Gr. of N. T. Gk., p. 206.
45 Our transl. therefore often fails to distinguish the two senses of e;dei in Gk. Gildersl., Synt., Pt. I, p. 144 f. Cf. chapter on Tense.
46 Prol., p. 200.
47 Blass, Gr. of N. T. Glc., p. 206.
48 Prol., p. 200.
49 W.-Th., p. 282.
50 Synt., Pt. I, p. 144.
51 Blass, Gr. of N. T. Gk., p. 205.
52 Gildersl., Synt., Pt. I, p. 168 f.
53 Gildersl., Am. Jour. of Philol., Jan., 1909, p. 16. Cf. Stahl, Krit.-hist. Synt., p. 251 f.
54 Moulton, Prol., p. 199.
55 Here aA read e;ceij.
56 But not in Acts. Cf. Blass, Gr. of N. T. Gk., p. 206.
57 Blass, Gr. of N. T. Gk., p. 206.
58 Moulton, Prol., p. 200.
59 Ib. Cf. Moulton, Class. Quart., Apr., 1908, p. 140. Moulton (Prol., p. 200) cites without a;n O.P. 526 (ii/A.D.) ouv pare,benon, O.P. 530 (ii/A.D.) pa,lin soi avpesta,lkein (ii/B.c.) ouvk avpe,sthi, all apodoses of 2d class conditions. The mod. Gk. here uses the conditional qa, (Thumb, Handb., p. 195).
60 Blass, Gr. of N. T. Gk., p. 207. Cf. Gildersl., Synt., Pt. I, p. 170 f.
61 Cf. Goodwin, M. and T., § 399.
62 Monro, Hom. Gr., pp. 236 f.
63 Moulton, Prol., p. 201.
64 Cf. Viteau, Le Verbe, p. 21.
65 Against Blass, Gr. of N. T. Gk., p. 210.
66 Ib. Cf. Thompson, Synt., p. 187.
67 W.-Th., p. 284.
68 Blass, Gr. of N. T. Gk., p. 210.
69 Monro, Hom. Gr., p. 235.
70 Moulton, Prol., p. 199.
71 Thompson, Synt., p. 218.
72 Thompson, Gk. Synt., 1883, p. 133.
73 Whitney, Sans. Gr., p. 216.
74 Cf. Gildersl., Synt., Pt. I, p. 149.
75 Pp. 129-131.
76 Whitney, Sans. Gr., p. 216,
77 Thompson, Synt. of Attic Gk., 1907, p. 191.
78 Die Crundl. d. griech. Synt., p. 115 f.
79 Comp. Plilol., p. 502.
80 Gildersl., Am. Jour. of Philol., Jan., 1909, p. 11.
81 Cf. Buumlein, Unters. uber griech. Modi (1846, p. 25 f.).
82 Cf. V. and D., Handb., p. 321 f.
83 Latin of the Latins and Greek of the Greeks, p. 23.
84 Bergaigne, De conjunctivi et optativi in indoeurop. linguis.
85 L'optatif grec, p. xxiii.
86 Ib., p. iii.
87 Jolly, Ein Kapitel d. vergl. Synt., Der Konjunktiv und Optativ, p. 119.
88 Die Grundl., p. 116 f. Cf. Synt., II, pp. 349 ff.
89 M. and T., App., Relation of the Optative to the Subjunctive and other Moods, p. 371.
90 Griech. Gr., p. 499.
91 The Anticipatory Subjunctive in Gk. and Lat., Stud. Class. Phil. (Chicago), I, p. 6. See discussion of these three uses of fut. ind. under Tense.
92 Cl. Rev., XVI, p. 166.
93 Prol., p. 184.
94 Synt., p. 235 f.
95 Konjunktiv und Optativ, p. 8 f.
96 Synt., Pt. I, p. 147.
97 Ib., p. 148.
98 Hom. Gr., p. 231.
99 Unters. uber die griech. Modi, p. 35. Cf. Wetzel, De Conjunctivi et Optativi apud Graecos Usu, p. 7.
100 Hammerschmidt, Uber die Grundb. von Konjunktiv und Optativ, p. 4.
101 Pp. 129-131. As a matter of fact both Delbruck and Goodwin fail to establish a sharp distinction between the subjunctive and the optative. Cf. Giles, Man., p. 504.
102 Cf. Giles, Man., p. 505.
103 Monro, Hom. Gr., p. 198.
104 Synt., Pt. I, p. 153.
105 Brug., Greich Gr., p. 503.
106 Moulton, Prol., p. 240.
107 Gr. of N. T. Gk., p. 321.
108 Prol., p. 240.
109 Gr. of N. T. Gk., p. 208.
110 See also Hatz., Einl., p. 218.
111 Moulton, Prol., p. 186.
112 Goodwin, M. and T., pp. 2, 372.
113 Justin Martyr, p. 169.
114 Hist. Gk. Gr., p. 449.
115 N. T. M. and T., p. 78.
116 Moulton, Prol., 3d ed., p. 190. But in the Germ. ed., p. 300, Moulton names 74. He had given 78 in the first Engl. ed.
117 M. and T., pp. 389 ff. See also pp. 101-105.
118 Giles, Man., p. 505.
119 Synt., Pt. I, p. 148.
120 Griech. Gr., p. 500.
121 Monro, Hom. Gr., p. 197.
122 Prol., p. 175.
123 Ib.
124 See 1 Cor. 10 : 7-9 for the change from first to second persons.
125 Jann., Hist. Gk. Gr., p. 447.
126 Blass, Gr. of N. T. Gk., p. 208. But see a;fete i;dwmen, (Mk. 15:36), though aD here read a;fej.
127 Moulton, Prol., p. 176. Jannaris (Hist. Gk. Gr., p. 448) derives a;j from e[ase ( e;ason), a;se)
128 It was rare in classic Gk. not to have a;ge or fe,re or some such word. Cf. Goodwin, M. and T., p. 88; Gildersl., Synt., Pt. I, p. 148 f. The volitive subj. is common in mod. Gk. (Thumb, Handb., p. 126) both for exhortations, commands, prohibitions and wishes. It occurs in the late pap. for wish, as kataxiw,sh|, P. Oxy. I, 128, 9. So in the inscr. toiau/ta pa,qh|, Pontica III, 62, 8
129 Prol., p. 175.
130 Ib.
131 Delbruck, Synt., p. 120; Monro, Hom. Gr., p. 240.
132 But Blass (Gr. of N. T. Gk., p. 278) holds the opposite view.
133 Prol., p. 179.
134 Ib., p. 178.
135 Cf. Blass, Gr. of N. T. Gk., p. 212 f.
136 Prol., p. 178.
137 Monro, Hom. Gr., pp. 199, 229.
138 Synt., Pt. I, p. 152. Cf. Goodwin, M. and T., p. 92.
139 Blass, Gr. of N. T. Gk., p. 211.
140 Burton, N. T. Moods and Tenses, p. 77.
141 Blass, Gr. of N. T. Gk., p. 210. Cf. K.-G., Tl. I, p. 221.
142 Ib.
143 Cf. Paley, The Gk. Particles, p. 5. See Koppin, Beitr. zu Entwick. and Ward. der Ideen uber die Grundb. d. griech. Modi (1880).
144 Hist. Gk. Gr., p. 564. On the subj. see further Earle, Cl. Papers, p. 221.
145 Ib., p. 450.
146 Ib., pp. 560-567.
147 Gildersl., Am. Jour. of Phil., Jan., 1909, p. 19. According to Vandacle (L'Optatif Grec, p. 251) Plato et Xen. "ont donne a l'optatif la plus grande extension possible; Xenophon marque l'apogee." The optative he also describes as "un instrument d'une delicatesse infinie." See further Kupff, Der Gebr. d. Opt. bei Diod. Sic. (1903); Reik, Der Opt. bei Polyb. und Philo (1907).
148 Schmid, Der Gebr. des Optativs bei Diod. Sic., 1903, p. 2.
149 Whitney, Sans. Gr., p. 218. In the original speech there was no clear distinction between the subj. and the opt. (Curtius, Temp. und Modi, 1846, p. 266).
150 Goodwin, M. and T., p. 384.
151 Baumlein, Griech. Modi, p. 177.
152 Monro, Hom. Gr., p. 229.
153 Ib., p. 231.
154 Goodwin, M. and T., p. 375.
155 Giles, Man., p. 510.
156 Griech. Gr., pp. 504 ff.
157 Goodwin, M. and T., p. 375.
158 Ib., p. 4.
159 Synt., Pt. I, p. 154. Stahl (Krit.-hist. Synt., p. 236 f.) notes a "concessive opt.," which is an overrefinement. It is merely a weakened form of wish (K.-G., Bd. I, p. 228) or of the potential use.
160 Cf. his Konjunktiv and Optativ, Syntaktische Forschungen, Att.-indische Synt. In the last of these he suggests that the potential and wishing functions are distinct in origin.
161 Monro, Hom. Gr., p. 219.
162 Gildersl., Synt., Pt. I, p. 153.
163 Prol., p. 197.
164 Ib., p. 166.
165 Moulton, Prol., p. 197 f.; Blass, Gr. of N.T. Gk., p. 220.
166 Prol., p. 197.
167 Ib., p. 198. He notes also 4 Macc. 5:13, suggnw,seien without a;n. In the Pap. a;n is usually present with the potential opt. (Radermacher, N. T. Gk., P. 129). Sometimes 'laws occurs with the opt., i;swjÄÄavporh,seien in Joh. Philop.
168 Burton, M. and T., p. 80; Blass, Gr. of N. T. Gk., p. 220:
169 Prol., p. 194.
170 Burton, N. T. M. and T., p. 79; Moulton, Prol., p. 194.
171 Ib., p. 240.
172 Cf. Sweet, New Eng. Gr.: Synt., pp. 107 ff.
173 Moulton, Prol., p. 195 f.
174 Blass, Gr. of N. T. Gk., p. 220.
175 Ib.
176 Moulton, Prol., p. 195.
177 They are all exx. of the third person save Phil. 1:20. Here is the list (with Burton's errors corrected by H. Scott): Mk. 11:14; Lu. 1:38; 20: 16; Ac. 8:20; Ro. 3:4, 6, 31; 6:2, 15; 7:7, 13; 9:14; 11:1, 11; 15: 5, 13; 1 Cor. 6:15; Gal. 2:17; 3:21; 6:14; 1 Th. 3:11, 12 bis; 5:23 bis; 2 Th. 2:17 bis; 3:5, 16; 2 Tim. 1:16, 18; 4:16; Phil. 1:20; Heb. 13:21; 1 Pet. 1:2; 2 Pet. 1:2; Ju. 2, 9.
178 Moulton, Prol., p. 196.
179 Gr. of N. T. Gk., p. 211.
180 Moulton, Prol., p. 194.
181 Ib., p. 198. On the "development principle" of the opt. see Mutzbauer, Konj. and Opt., p. 155.
182 Man., pp. 464-473, 502.
183 Prol., p. 171.
184 Whitney, Sans. Gr., p. 220.
185 Delbruck, Die Grundl., p. 120.
186 Jann., Hist. Gk. Gr., pp. 449, 451, 555 ff.; V. and D., Handb. (Jebb), p. 322 f.; Thumb, Handb., p. 127.
187 Moulton, Prol., p. 177. Cf. Gildersl., Synt., p. 117.
188 Moulton, Prol., p. 177.
189 Synt., Pt. I, p. 116. Cf. W.-Th., p. 316.
190 Gr. of N. T. Gk., p. 209.
191 Moulton, Prol., p. 179.
192 Monro, Hom. Gr., p. 162.
193 Giles, Man., p. 468.
194 Moulton, Prol., p. 179.
195 Blass, Gr. of N. T. Gk., p. 222.
196 Hatz., Einl., p. 192. Cf. Thumb, Hellen., p. 130 f.
197 Prol., p. 179 f.
198 Gr. of N. T. Gk., p. 222.
199 Ib.
200 W.-Th., p. 316.
201 Ib., pp. 350 ff.
202 Gr., p. 180.
203 W.-Moulton, p. 732, n. 5
204 Prol., p. 223.
205 Ib.
206 Ib., p. 225.
207 Hellen., p. 131.
208 Mr. H. Scott notes the absence of e;ste in the H. R. Conc. of the LXX, in Veitch, in Kuhner-Bl., Mayser, Helbing, Thackeray. In Goodspeed's Index Pat. he finds it only in 1 Clem. 45:1, and the accent is doubtful here. He finds it also in Test. XII Pat. Reub. 6:1. It could have been used in Napht. 3:2 and in Ign. Eph. 10:2.
209 Prol., p. 181, against his former view in Expositor, VI, x. 450.
210 Ib.
211 Ib.
212 Ib.
214 Blass, Gr. of N. T. Gk., p. 221.
215 Prol., p. 173.
216 Gilders1., Synt., Pt. I, p. 158.
217 Prol., p. 172.
218 Ib.
220 Moulton, Prol., p. 173.
221 Am. Jour. of Philol., Apr., 1909, p. 235.
222 Cf. Burton, N. T. M. and T., p. 80.
223 Cf. Gildersl., Synt., Pt. I, p. 158; Miller, The Limitation of the Imperative in the Attic Orators, Am. Jour. of Philol., 1892, pp. 399-436.
224 Cf. K.-G., Bd. I, p. 236.
225 Gildersl., Synt., Pt. I, p. 162.
226 lb., p. 167.
227 Cf. Brug., Griech. Gr., p. 511.
228 Gildersl., Synt., Pt. I, p. 164. See also Thompson, Synt., p. 190 f.
229 N. T. M. and T., p. 81.
230 Brug., Griech. Gr., p. 552.
231 Jann., Hist. Gk. Gr., p. 452.
232 Brug., Griech. Gr., p. 552.
233 Jann., Hist. Gk. Gr., p. 451.
234 Cf. Monro, Hom. Gr., p. 194.
235 Monro, Hom. Gr., p. 194.
236 Ti. II, 2. Bd., pp. 354-459.
237 N. T. M. and T., p. 82.
238 Monro, Hom. Gr., p. 194 f.
239 Le Verbe: Syntaxe des Propositions, pp. 41-144.
240 Brug., Griech. Gr., p. 556.
241 Ib., p. 559.
242 Monro, Hom. Gr., p. 186. Stahl, Hist.-krit. Synt., p. 523, points out that the relative sentence is either "synthetic or parathetic."
243 Schmitt, Uber den Ursprung des Substantivsatzes mit Relativpartik. im Griech., 1889, p. 12.
244 Thompson, Synt. of Attic Gk., p. 383.
245 Frenzel, Die Entwick. des relat. Satzb. im Griech., 1889, p. 4.
246 Thompson, Synt., p. 383.
247 Baron, Le Pronom Relat. et la Conj. en Grec, 1892, p. 61.
248 Brug., Griech. Gr., p. 557. It was not always done (attraction) either in Herod. or Thuc. Cf. Reisert, Zur Attraktion der Relativsatze in der griech. Prosa, p. 30 f.
249 Brug., Griech. Gr., p. 561.
250 Thompson, Synt., p. 384.
251 N. T. M. and T., pp. 126
252 Ober den Ursprung des Substantivsatzes mit Relativpartik. im Griech.
253 Entwickelungsgesch. der Absichtsatze.
254 Der Substantivs. und das Rel. w`j.
255 Baron, Le Pronom Rel. et la Conjonction en Grec, p. 130.
256 Frenzel, Die Entw. des rel. Satzb. im Griech., p. 4.
257 J. Classen, Beob. uber den homerischen Sprachgeb., 1867, p. 6.
258 Bd. II, pp. 420 ff.
259 See, per contra, Baron, Le Pronom Rel. et la Conjonction en Grec, pp. 61
260 Monro, Hom. Gr., p. 189.
261 Jann., Hist. Gk. Gr., p. 452.
262 Thompson, Synt., p. 383.
263 Blass, Gr. of N. T. Gk., p. 217, explains this subj. as due to a "final meaning." D in Mk. reads fa,gomai.
264 Moods and Tenses, p. 197.
265 N. T. M. and T., p. 119.
266 Hist. Gk. Gr., p. 470.
267 Cf. Robertson, Short Gr. of the Gk. N. T., p. 169.
268 Viteau, Le Verbe, p. 139.
269 Cf. W.-Th., p. 307.
270 Elem. Gk. Synt., 1897, p. 7. Cf. Baumlein, Unters. etc., p. 315.
271 Hom. Gr., p. 263 f.
272 Per contra see W.-Th., p. 306.
273 Blass (Gr. of N. T. Gk., p. 217) quotes a'j a'n suntele,sousin, from an inscr. in Viereck's Sermo Graecus, p. 38.
274 Cf. K.-G., Bd. II, pp. 421, 424.
275 Draeger, Hist. Synt., Bd. II, p. 527.
276 Cf. Burton, N. T. M. and T., p. 118.
277 Cf. K.-G., Bd. II, p. 421.
278 Moulton, Prol., p. 185.
279 N. T. M. and T., p. 126.
280 Goodwin, M. and T., p. 217.
281 Cf. Blass, Gr. of N. T. Gk., p. 218.
282 See Viteau, Le Verbe, p. 135.
283 Gr. of N. T. Gk., p. 218.
284 Blass, ib., cites also i`kano.j lu/sai in Mk. 1:7.
285 N. T. M. and T., p. 126.
286 Cf. K.-G., Bd. II, p. 422.
287 Burton, N. T. M. and T., p. 118.
288 M. and T., pp. 195 ff.
289 N. T. M. and. T., pp. 119 ff.
290 Le Verbe, pp. 136 ff.
291 N. T. M. and T., p. 180.
292 Blass, Gr. of N. T. Gk., p. 274. Cf. also Burton, N. T. M. and T., p. 98.
293 Cf. Nilsson, Die Kausalsatze itn Griech. his Arist. I, Die Poesie.
294 Viteau, Le Verbe, p. 98..
295 Jann., Hist. Gk. Gr., p. 454.
296 Ib.
297 Monro, Hom. Gr., p. 189.
298 As Viteau does in Le Verbe, p. 100. The LXX does show the idiom, as in 1 Ki. 1:8, ti, e;sti soi o[ti klai,eij*
299 Joh. Gr. p. 162.
300 Viteau, Le Verbe, p. 101.
301 Jann., Hist. Gk. Gr., p. 454.
302 Cf. ib.
303 Viteau, Le Verbe, p. 101.
304 The Use of the Inf. in Bibl. Gk., p. 20. Mr. II. Scott notes pres. 24, aor. 1 (Mt. 24 : 12), perf. 7 times.
305 Blass, Gr. of N. T. Gk., p. 236.
306 N. T. M. and T., p. 161.
307 Votaw, The Use of the Inf. in Bibl. Gk., p. 29.
308 Viteau, Le Verbe, p. 101.
309 Chapter XXVIII..
310 In general correlatives are rare in the LXX. Viteau, Le Verbe, p. 142.
311 Gr. of N. T. Gk., p. 321.
312 Cf. Abbott, Joh. Gr., p. 152 f.
313 Gr. of N. T. Gk., p. 218.
314 N. T. M. and T., pp. 118, 126 ff.
315 Synt. of Attic Gk., p. 329.
316 Ib., p. 328.
317 W. G. Hale, Stud. in Class. Philol., The Cum Constructions, 1887, p. 259.
318 Jann., Hist. Gk. Gr., p. 466.
319 vEpei, was rare in Homer. Monro, Hom. Gr., p. 226.
320 Cf. Monro, Hom. Gr., pp. 189 ff.; Brug., Griech. Gr., p. 561; Riem. and Goelzer, Synt., p. 444 f.
321 Cf. Mutzbauer, Konjunktiv and Optativ, p. 97.
322 W.-M., p. 387.
323 Viteau, Le Verbe, p. 125. Cf. Jannaris, Hist. Gk. Gr., p. 463.
324 Cf. W.-M., p. 388,
325 Winifred Warren, A Study of Conjunctional Temp. Clauses in Thucydides, 1897, p. 73. [Ote is found twice in 1 Thuc. with the optative, but Miss Warren reads o`po,te.
326 Baumlein, Unters. uber die griech. Modi und die Partik. ke,n und a;n, 1846, p. 322.
327 Reffel, Uber den Sprachgebr. des Agathias, p. 24.
328 Viteau, Le Verbe, p. 123; W.-M., p. 388 f.
329 W.-M., p. 389.
330 Ib.; Mullach, Vulg., p. 368.
331 Gr. of N. T. Gk., p. 218.
332 W.-M., p. 389.
333 Lang. of the N. T., p. 111.
334 Meisterh.-Schwyzer, Gr. d. attisch. Inschr., p. 251.
335 Brug., Griech. Gr., p. 563.
336 Ib., p. 200.
337 Goodwin, M. and T., p. 235.
338 Burton, N. T. M. and T., p. 128. But the proper sense of the indic. is better as an expression of the fact. Radermacher, N. T. Gr., p. 140.
339 Goodwin, M. and T., p. 235.
340 Gr. of N. T. Gk., p. 219.
341 Cf. Sturm, Geschichtl. Entw. der Konstr. mit pri,n, 1882, p. 4; Frenzel, Die Entw. der Satze mit pri,n, 1896, p. 12.
342 Sturm, ib., p. 145.
343 Ib., p. 6.
344 Cf. Blass, Gr. of N. T. Gk., p. 219.
345 Moulton, Prol., p. 169 note.
346 Moulton, Prol., p. 215.
347 Ib., p. 249.
348 Votaw, The Inf. in Bibl. Gk., p. 20.
349 Ib.
350 Moulton, Prol., p. 230. "We should not usually put a temporal clause to represent these, as it would overdo the emphasis."
351 Jebb in V. and D.'s Handb., p. 333.
352 M. and T., pp. 105-137.
353 Ib., pp. 217-233.
354 N. T. M. and T., pp. 83-100.
355 Le Verbe, pp. 71-95.
356 Hist. Gk. Gr., p. 455.
357 Ib., p. 458. Thus o[pwj and w`j gradually disappear.
358 Brug., Griech. Gr., p. 555,
359 Goodwin, M. and T., p. 109.
360 Ib., p. 108.
361 Am. Jour. of Philol., 1883, p. 419.
362 Moulton, Prol., p. 185.
363 Goodwin, M. and T., p. 107.
364 Entwickelungsgeschichte der Absichtsatze (1884, 1885).
365 Dyroff, Gesch. des Pronomen reflexivum, 1892, p. 71.
366 Cf. Brug., Griech. Gr., p. 566.
367 Ib.
368 Meisterh.-Schw., p. 253 f.
369 Hom. Gr., p. 207.
370 Stahl, Krit.-hist. Synt., p. 479; Mutzbauer, Konj. and Opt., p. 76.
371 Goodwin, M. and. T., p. 107; Blass, Gr. of N. T. Gk., p. 211.
372 Jann., Hist. Gk. Gr., p. 416 f.; Jebb in V. and D., pp. 319-323.
373 Cf. W.-H., vol. II, App., p. 168.
374 W.-M., p. 363.
375 On the sparing use of the opt. with final sentences in late Gk. see the tables in Diel, De enuntiatis finalibus apud Graecarum rerum scriptores posterioris aetatis, 1894, pp. 20 See also Radermacher, N. T. Gr., p. 132. Moulton (Prol., p. 197) notes how the Atticists revelled in the opt. with i[naà o[pwjà w`j, Josephus has 32 per cent. opts., Plut. 49 (Lives), Arrian 82, Appian 87! Polyb. has only 7, Diodorus 5. These are true koinh, literati. Moulton finds only one pap. of this period with opt. with i[na, O.P. 237 (late ii/A.D.), i[na - dunhqei,hn. In iii/A.D. he notes L.Pw., i[n v - ei;hi in primary sequence. Tb. 1 (ii/B.C.) actually has hvxi,wsa crhmatisqh,soito.
376 Weber, Entwickelungsgeschichte der Absichtsatze, p. 243.
377 Goodwin, M. and T., p. 115.
378 Approved by Blass, Gr. of N. T. Gk., p. 212.
379 Moulton, Prol., p. 35.
380 W.-M., p. 362.
381 Cf. W.-H., App., pp. 167, 169, 171. See further Meyer on 1 Cor. 4 : 6.
382 Cf. Goodwin, M. and T., p. 120. The Mod. Gk, has na, with past tenses of the ind. (Thumb, Handb., p. 198).
383 Moulton, Pro1., pp. 41, 205, 211.
384 Brug., Griech. Or., p. 565; Delbruck, Konj. and Opt., p. 61.
385 Thompson, Synt. of Attic Gk., p. 348.
386 Jann., Hist. Gk. Gr., p. 417.
387 Meisterh.-Schw., p. 253 f.
388 Goodwin M. and T., p.
389 Ib., p. 113 f.
390 Moulton, Prol., p. 197; Jana., Hist. Gk. Gr., p. 417.
391 Blass, Gr. of N. T. Gk., p. 211.
392 Ib.
393 Moulton, Prol., p. 220.
394 Goodwin, M. and T., pp. 107, 112.
395 Ib., p. 112.
396 Ib., p. 107.
397 Goodwin, M. and T. p. 120 f.
398 Burton, N. T. M. and T., p. 86.
399 Prol., p. 194.
400 Ib., p. 197.
401 Goodwin, M. and T., pp. 216
402 Gr. of N. T. Gk., p. 217.
403 The Inf. in Bibl. Gk., p. 10.
404 Moulton, Prol., p. 204.
405 lb., p. 207.
406 Blass, Gr. of N. T. Gk., p. 223.
407 Inf. in Bibl. Gk., p. 21.
408 Ib., p. 10.
409 Ib., p. 19.
410 Cf. Burton, N. T. M. and T., p. 161 f.
411 Blass, Gr. of N. T. Gk., p. 198.
412 Cf. Goodwin, M. and T., pp. 122 ff.
413 N. T. M. and T., p. 83.
414 Moulton, Prol., p. 205.
415 It is seen as early as Demosthenes (IV, 28).
416 Jebb in V. and D.'s Handb., p. 320.
417 Cf. Blass, Gr. of N. T. Gk., p. 228.
418 W.-Th., p. 338 f.
419 Blass, Gr. of N. T. Gk., p. 225 f.
420 Tb.
421 It is found in Hom. Cf. Goodwin, M. and T., p. 128.
422 Prol., p. 178.
423 Ib.
424 W.-M., p. 396.
425 See art. by Jann., Expositor, ser. V, vol. IX, p. 296.
426 Burton, N. T. M. and T., pp. 88, 951f.
427 Cf. Burton, N. T. M. and T., p. 95.
428 Goodwin, M. and T., p. 133.
429 Moulton, Prol., pp. 185, 248.
430 Burton, N. T. M. and T., p. 89.
431 Cf. Goodwin, M. and T., p. 217.
432 Burton, N. T. M. and T., p. 87.
433 Prol., p. 218 f.
434 Jann., Hist. Gk. Gr., p. 455.
435 Gr., § 555, 2, Anm. 3.
436 Lexicon, p. 304. Abbott, Joh. Gr., p. 114, holds to the strict use of i[na.
437 W.-M., p. 421.
438 Short Gr. of the Gk. N. T., pp. 153, 155.
439 Prol., p. 206.
440 Intr. to N. T. Gr., p. 217.
441 Prol., p. 209.
442 On Epli. 1 : 17.
443 Moulton, Prol., p. 210.
444 Blass, Gr. of N. T. Gk., p. 228.
445 N. T. M. and T., p. 92 f.
446 Ib., p. 04.
447 Ib.
448 Gr. of N. T. Gk., p. 224.
449 Prol., p. 210.
450 Blass, Gr. of N. T. Gk., p. 224.
451 Moulton, Prol., p. 210.
452 Cf. Gildersl., The Consec. Sent. in Gk., Am. Jour. of Philol., 1886, p. 167.
453 Cf. Berdolt, Der Konsekutivsatz in der illtern griech. Litteratur, 1896, pp. 21-27.
454 Mr. H. Scott makes 95 times by counting the verbs, Geden 83.
455 Goodwin, M. and T., pp. 223 ff.
456 Gr. of N. T. Gk., p. 224.
457 Prol., p. 209.
458 Burton, N. T. M. and T., p. 99.
459 Blass, Gr. of N. T. Gk., p. 224.
460 In Xen. Ws rather than w[ste occurs both with the inf. and the modes. Cf. Wehmann, De w[ste particulae usu Heroditeo Thucydideo Xenophonteo, 1891, p. 40.
461 Jann., Hist. Gk. Gr., p. 455; Moulton, Prol., p. 249. Cf. Compernass, § 38. See Sophocles' Lexicon.
462 Goodwin, M. and T., p. 218 f.
463 Jann., Hist. Gk. Gr., p. 468.
464 Gr. of N. T. Gk., p. 224.
465 Moulton, Prol., p. 210.
466 Short Gr. of the Gk. N. T., p. 156.
467 Prol., p. 217.
468 Gr. of N. T. Gk., p. 236.
469 W.-M., p. 413 f.
470 Moulton, Prol., p. 219.
471 Gr. of N. T. Gk., p. 236.
472 W.-M., p. 414 note.
473 Prol., p. 218. See further Ogden, De infinitivi finalis vel consecutivi constructione apud priscos poetas Graecos, 1913.
474 See ch. on "Wishes" in my Short Gr. of the Gk. N. T., p. 157.
475 Moulton, Prol., p. 201.
476 In W.-Sch., p. 29, reference is made to eiv o;felon evfu,laxaj in Job 14: 13 and eiv ga.r o;felon dunai,mhn in Job 30:24. Evidently o;felon was not felt to be sufficient alone.
477 Prol., p. 201.
478 Monro, Horn. Gr., p. 227. Cf. Gildersl., Am. Jour. of Philol., 1909, p. 14.
479 W.-M., pp. 363 ff.
480 Gr. of N. T. Gk., p. 213 f.
481 Pp. 161
482 Am. Jour. of Philol., 1882, pp. 435 ff.
483 Synt. of Attic Gk., p. 296.
484 Jebb, V. and D.'s Handb., pp. 330 ff.; Thumb, Handb., p. 194 f.
485 Beitr. zur griech. Gr., 1893, pp. 14, 18. He uses "Wirklichkeit" and "Irrealitat" (pp. 8, 28) for the others.
486 W.-M., p. 364.
487 Gk. Synt., p. 156 f.
488 See Proc. of the Am. Acad., vol. VI; Jour. of Philol., V, pp. 186-205, VIII, pp. 13-38; M. and T., pp. 145 ff.
489 Adopted by Thompson, Synt. of Attic Gk., p. 296.
490 Am. Jour. of Philol., 1882, pp. 435 ff. Gildersl. still objects to the distinction of "particular" and "general" suppositions which Goodwin brought into fashion. That merely depends on the character of the apodosis. Cf. Am. Jour. of Philol., 1909, p. 10.
491 M. and T., p. 147.
492 N. T. M. and T., pp. 100 ff. Farnell (Gk. Conditional and Rel. Sent., 1892) also follows Goodwin, as does R. H. Smith (The Theory of Cond. Sent. in Gk. and Lat., 1894).
493 Monro, Hom. Gr., p. 210 f.
494 Baumlein, Unters., pp. 352ff.
495 Cf. Jann., Hist. Gk. Gr., p. 463; Thumb, Handb., p. 194 f.
496 Cf. Jolly, Ein Kapitel vergl. Synt., 1872, p. 122 f.
497 Gildersl., Am. Jour. of Philol., 1882, p. 449.
498 The origin of eiv is uncertain. Eiv is the same as al in Homer (and Doric).
499 Jann., Hist. Gk. Gr., p. 420.
500 Compernass, De Sermone, p. 35 f.
501 Moulton, Prol., p. 168.
502 Ib., pp. 49, 168, 187; Cl. Rev., XVIII, p. 108. For the usage of the LXX see Sterenberg, The Use of Cond. Sent. in the Alex. Version of the Pentateuch, 1908.
503 Prol., p. 168.
504 Gr. of N. T. Gk., p. 215.
505 Moulton, Prol., p. 169.
506 B. S., p. 204.
507 W.-Th., p. 477.
508 Jann., Hist. Gk. Gr., p. 429.
509 Ib.
510 Prol., T.
511 Gr. of N. T. Gk., p. 253.
512 Moulton, Prol., p. 170. Cf. Gildersl., Am. Jour. of Philol., 1880, first copy.
513 Prol., p. 170. Cf. P. Thouvenin, Les Negations dans le Nouveau Testament, Revue de Philol., 1894, p. 229,
514 Prol., p. 199. Goodwin, M. and T. (p. 147), sees clearly on this point.
515 Cf. Wilhelmus, De Modo Irreali qui Vocatur, 1881, p. 3. Mod. Gk. no longer has this idiom. It uses a;n with the past ind. and qa, in the apodosis for a;n.
516 Moulton, Prol., p. 200.
517 Bamberg, Hauptregeln der griech. Synt., 1890, p. 45.; Conditional Clauses in Gk., p. 2, Anonymous Pamphlet in Bodleian Library.
518 Prol., p. 200.
519 Cf. Westcott on Heb., pp. 111 ff., for an excellent summary of the second class conditions.
520 Moulton, Prol., p. 201.
521 On Heb., p. 113.
522 Prol., p. 166.
523 Cf. Blass, Gr. of N. T. Gk., p. 254; Moulton, Prol., p. 171.
524 The Origin of Subj. and Opt. Conditions in Gk. and Lat., Harv. Stu. in Class. Philol., 1901, p. 115.
525 Hom. Gr., p. 230. Stahl, Ciriech. histor. Synt., p. 390, makes it futuristic.
526 Prol., p. 185.
527 AM. Jour. of Philol., 1909, p. 11.
528 Griech. Modi, p. 177.
529 Gildersl. (Am. Jour. of XXXIII, 4, p. 490) complains that in Germany no standing is given to his distinction between the "minatory and monitary" use of eiv with the. future indicative. He first promulgated it in 1876.
530 Jann., Mist. Gk. Gr., pp. 420, 464.
531 Prol., p. 187. Cf. Goodwin, M. and T., p. 167.
532 B. S., p. 118.
533 The Phrygian inser. show similar exx. Cf. Ramsay, Cities and Bish. of Phrygia, II, 292. Burton (N. T. M. and I ., p. 105) admits that it is an overrefinement to rule out eiv and the subj. C . Moulton, Prol., p. 240.
534 Prol., p. 187.
535 Cf. Moulton, Prol., p. 43; Meisterh.-Sp. 225 f. In Jo. 5:19 we have both uses of a;n (conditional and modal). In Mk. 5:28 note eva.n a;ywmai ka'n tw/n i`mati,wn, not a repetition of modal a;n, but a particle ka;nÊ even.'
536 Cf. Blass, Gr. of N. T. Gk., p. 215.
537 Prol., p. 186.
538 Cf. Abbott, Joh. G p. 371.
539 Origin of Subj. and Opt. Cond., Harv. Stu. in Class. Philol., 1901, p. 115.
540 Moulton, Prol., p. 196.
541 Cf. Monro, Horn. Gr., p. 227.
542 Gildersl., Am. J. of Philol., 1909, p. 7.
543 Moulton, Prol., p. 196.
544 Ib.
545 Monro, Hom. Gr., p. 228 f.
546 Gr. of N. T. Gk., p. 221.
547 Blass, Gr. of N. T. Gk., p. 220.
548 Moulton, Prol., p. 230.
549 Burton, N. T. M. and T., p. 110.
550 Jann., Hist. Gk. Gr., p. 461.
551 Robertson, Short Gr. of the Gk. N. T., p. 166.
552 W.-Th., p. 600.
553 App., p. 151.
554 Prol., p. 46.
555 Goodwin, M. and T., pp. 180
556 Moulton, Prol., p. 194.
557 Prol., p. 169.
558 Burton, N. T. M. and T., p. 111.
559 Blass, Gr. of N. T. Gk., p. 215.
560 N. T. M. and T., p. 112.
561 Paley, Gk. Part., p. 31.
562 Jann., Hist. Gk. Gr., p. 465.
563 Thayer's Lexicon.
564 Cf. Burton, N. T. M. and T., p. 114.
565 Burton, N. T. M. and T., p. 130.
566 Goodwin, M. and T., p. 285.
567 Viteau, Le Verbe, p. 50; but see on the other hand Con. and Stock, Sel., p. 114.
568 Hist. Gk. Gr., p. 472.
569 Schmitt, Uber den Urs pr. des Substantivsatzes, 1889, p. 66.
570 Thumb, Handb., p. 192. Jann., Hist. Gk. Gr., p. 472. Kai. to,tej ei=pe pw/j De, sou to, vlega evgw,; then he said, Didn't I tell you so?'
571 Cf. Goodwin, M. and T., p. 263.
572 Cf. Robertson, Short Gr., p. 1S1. As a matter of fact, the primitive method in oratio obliqua was probably this very change of tense as in Eng. We have it more frequently in Hom. than the change of mode or the graphic retention of tense. Cf. Thompson, Synt. of Attic Gk., p,. 402.
573 In archaic Lat. the incl. was used in indirect discourse as in Gk. Cf. Draeger, Hist. Synt., Bd. II, p. 460.
574 Goodwin, M. and T., p. 263.
575 Madvig, Bemerk. Uber einige Punkte der griech. Worthig. 1848, p. 23.
576 Goodwin, M. and T., p. 273.
577 Moulton, Prol., p. 198.
578 Hist. Gk. Gr., p. 471 f.
579 N. T. M. and T., p. 131. So most of the grammars.
580 Goodwin, M. and T., p. 258.
581 Jann., Hist. Gk. Gr., p. 571.
582 De Particulorum o[ti et w`j apud Demosthenum Usu, 1890, p. 38.
583 Cf. Blass, Gr. of N. T. Gk., p. 230 f.
584 Hatz., Einl., p. 19.
585 See Sophocles' Lexicon under w`j. Cf. Jann., Hist. Gk. Gr., p. 413. Moulton (Prol., p. 212) gives C.P.R. 19 (iv/A.D.) prw,hn bibli,a evpide,dwka th|/ sh|/ evpimelei,a| w`j o[ti evboulh,qhn
586 Moulton, Prol., p. 212.
587 Blass, Gr. of N. T. Gk., p. 231.
588 Moulton, Prol., p. 211.
589 Jann., Hist. Gk. Gr., p. 413.
590 Mitsotakis, Praktische Gr. der neugriechischen Schrift- and Umgangssprache, 1891, p. 235.
591 Gr. of N. T. Gk., p. 231,
592 Le Verbe, p. 51.
593 Gr. of N. T. Gk., p. 231 f.
594 Ib., p. 233.
595 Gr. of N. T. Gk., p. 233.
596 Cf. Goodwin, M, and T., p. 267.
597 Monro, Hom. Gr., p. 162.
598 Goodwin, M. and T., p. 269.
599 Gk. Synt., p. 139.
600 Hom. Gr., p. 162.
601 Prol., p. 212.
602 Monro, Hom. Gr., p. 162.
603 Gr. of the Gk. N. T., p. 238 f.
604 Prol., p. 212 f.
605 Blass, Gr. of N. T. Gk., p. 237.
606 See also Lu. 23:2, le,gonta au`to.n ei=nai)
607 Moulton, Prol., p. 212. Cf. Zeitlin, The Accusative with Inf. and some Kindred Constr. in Eng. (1908).
608 Analogy in Synt., p. 64.
609 Blass, Gr. of N. T. Gk., p. 246.
610 Le Verbe, p. 531f.
611 Gr. of N. T. Gk., p. 246.
612 W.-M., p. 760, n. 2.
613 Prol., p. 16.
614 Ib.
615 Ib., p. 17.
616 As in Burton, N. T. M. and T., p. 142
617 Prol., pp. 55, 193. Cf. Burton, N. T. M. and T., p. 134.
618 Viteau, Le Verbe, p. 68.
619 Cf. Viteau, Le Verbe, p. 62.
620 Goodwin, M. and T., p. 273.
621 Ib., p. 272.