797

CHAPTER XVII

VOICE ( DIAQESIS, Genus)

I. Point of View. For a discussion of the nature of the verb see chapter VIII, Conjugation of the Verb, I and II.

(a) DISTINCTION BETWEEN VOICE AND TRANSITIVENESS. See (b), and chapter VIII, for a discussion of this point. The matter might have been well reserved for syntax, but it seemed worth while to set forth at once the fundamental facts about voice. It is here assumed, therefore, that one understands that voice per se does not deal with the question of transitive or intransitive action. That point concerns the verb itself, not the voice. Active and middle verbs may be either transitive or intransitive. Passive verbs may even be transitive, though usually intransitive, in one sense of "transitive." But Gildersleeve¹ holds that "a transitive verb is a verb that passes over to a passive rather than one that passes over to an object." That is truer of Latin than of Greek, which, "with a lordliness that reminds one of English;" makes a passive out of any kind of an active. Terminology in syntax is open to dispute at many points, but I see only hopeless confusion here unless voice is kept to its real meaning. In Kuhner-Gerth² it is held that "the active has a double meaning," either intransitive or transitive. My point is that the voice per se has nothing to do with that question. Some verbs are intransitive, some are transitive, some are used either way. This freedom in the use of verbs increased till in the later Greek verbs that were once intransitive become transitive.³ Brugmann⁴ properly separates the question of transitive and intransitive verbs from that of voice (cf. iterative, intensive, inchoative, desiderative verbs). Some of the intransitive uses of verbs were due to the absence of the reflexive pronoun, as in perih/ge (Mk. 6 : 6), avporri,yantaj (Ac. 27:43).⁵ The modern Greek preserves the same

798 A GRAMMAR OF THE GREEK NEW TESTAMENT

freedom in the use of transitive and intransitive verbs and has peculiarities of its own.⁶

(b) MEANING OF VOICE. Voice relates the action to the subject. The use of voice then is to direct attention to the subject, not to the object. That concerns transitive and intransitive verbs. Stahl⁷ puts it crisply: "The voice of the verb describes a relation of the verb-idea to the subject."

(c) NAMES OF THE VOICES. Cf. chapter VIII, VI, (b). The names come from Dionysius Thrax (about B.C. 30), but "he has no inkling of a middle sense,"⁸ showing that already the middle is disappearing before the passive. The terminology is very poor. Gildersleeve⁹ calls the fashion of the Germans "a positively indecent nomenclature," since they call the voices genera ( ge,nh), "based on a fancied resemblance to the genders." We in English follow the French voix (Latin vox), found first in this sense in the Grammatica graeca nova of J. Weller (A.D. 1635).¹⁰

(d) HISTORY OF THE VOICES. See chapter VIII, vi, (c), (d), (e). Cf. also Jannaris, Historical Gr., p. 362 f.; Moulton, Prol., p. 152. In the pro-ethnic language there were probably both active and middle. Cf. Delbruck, Vergl. Syntax, Bd. II, p. 413. There was no passive as there was none in the Sanskrit, save in the present system.¹¹ The rise of the passive meaning with the use of middle and active endings was sure to bring confusion and a tendency towards simplification. It was inevitable that the three voices should go back to two. In the actual outcome, the passive, though an interloper, ousts the middle of its forms and of most of its uses.¹² In the modern Greek vernacular, therefore, we find only two voices as to form, for the passive has taken over the meaning of the middle also (Thumb, Handb., p. 111 f.). In the beginning there were only active and middle. In the end we find only active and passive.

(e) HELP FROM THE SANSKRIT. The verb development in the Indo-Germanic languages has been more independent than that of nouns. Latin, for instance, has recast its verb-system, and it is quite difficult to compare the Greek and Latin voices. Sanskrit

VOICE ( DIAQESIS) 799

and Greek have preserved the voices best of all. Hence the Sanskrit can throw a good deal of light on the Greek voices.¹³

(f) DEFECTIVE VERBS. Not all verbs were used in all the voices. Some were used only in one, some in two, some in all three. Then again, some verbs had one voice in one tense, another voice in another tense. This is just like the Sanskrit,¹⁴ and just what one would expect from a living language in contrast with an artificial one. Brugmann,¹⁵ indeed, divides verbs, as to voices, according to this principle (those with active only, middle only, with both, etc.). In the N. T. Blass (Gr. of N. T. Gk., p. 180) finds the same general use of the voices as in the older Greek, the same difficulty in differentiating the voices, and the same "arbitrariness" in the use of individual verbs. But much of this difficulty is due to coming at the matter with preconceived rules. Blass' treatment of the voices is quite unsatisfactory. Cf. further for this matter, chapter VIII, VI, (d).

II. The Active Voice ( dia,qesij evnergetikh,). The Stoics called the active ovrqh, also.

(a) MEANING OF THE ACTIVE VOICE. In this voice the subject is merely represented as acting or existing, for state (cf. eivmi,) must be included as well as action. It is not certain whether the active or the middle is the older; but the active is far the more common.

(b) EITHER TRANSITIVE OR INTRANSITIVE. There is nothing peculiar in the N. T. about this. Each verb has its own history. One originally transitive may become intransitive and vice versa.¹⁶ Cf. a;gw which may be intransitive a;gwmen (Mt. 26:46; cf. the interjectional a;ge, Jas. 4:13) or transitive h;gagon auvto,n (Lu. 19:35). In a;rantej (Ac. 27:13, 17) the object is probably understood ( th.n nau/n). Cf. also auvxa,nw in Mt. 6:28 and 2 Cor. 9:10. Ba,llw is usually transitive, even in Jo. 13:2 (cf. Ac. 22:23), but it is intransitive in Ac. 27:14 ( e;balen, 'rushed'). Cf. blasta,nw in Jas. 5:18 (tr.) and in Mt. 13:26 (intr.). So bre,cw is transitive in Lu. 7:38, but intransitive in Mt. 5:45. vEgei,rw is usually tran- sitive (Mt. 10:8), but see Mt. 26:46. Euvaggeli,zw, is transitive in Rev. 10:7, but intransitive in 14:6. ;Ecw is transitive except when used with adverbs, when, as in ancient Greek, it may be intransitive. Cf. tou.j kakw/j e;contaj (Mt. 4:24), evsca,twj e;cei (Mk.

800 A GRAMMAR OF THE GREEK NEW TESTAMENT

5:23), h;dh e;conta (Jo. 11:17), ou[twj e;cei (Ac. 7:1), to. nu/n e;con (Ac. 24:25). Kli,nw is transitive in Mt. 8:20, but intransitive in Lu. 9:12. In Ac. 7:42 stre,fw is intransitive, though also transitive elsewhere. In the N. T. qriambeu,w is transitive and the same is true of maqhteu,w. But in Text. Rec. evmaqh,teuse is intransitive in Mt. 27:57. Cf. du,nw intransitive in Lu. 4:40 and fu,w in Heb. 12:15. Let these serve as specimens of many such verbs in the N. T. Modern Greek is specially rich in intransitive active verbs (Thumb, Handb., p. 112) and verbs that oscillate from one use to the other.

(c) EFFECT OF PREPOSITIONS IN COMPOSITION. These may make the verb transitive or the result may be just the opposite. As examples of transitive compounds from an intransitive simplex take diabai,nw (Heb. 11:29), but intransitive in Lu. 16:26. So dih,rceto th.n vIereicw, (Lu. 19:1), pare,rcesqe th.n kri,singrk grk(11:42). On the other hand, intransitive compounds abound. The compounds of a;gw (simplex either tr. or intr.) which are often intransitive are avpa,gw (Mt. 7:13), para,gw (Mt. 9:9), peria,gw (Ac. 13:11), proa,gw (Lu. 18:39), u`pa,gw (Jo. 3:8); but not avna,gw. Cf. also paradi,dwmi in Mk. 4:29. With ba,llw note evpiba,llw in Mk. 4:37 and the peculiar evpibalw,n, in 14:72. Examples of several intransitive compounds of e;cw occur in the N. T. Thus avpe,cw (Mk. 14:41), evne,cw (Mk. 6:19), evpe,cw (Lu. 14:7; Ac. 19:22), perie,cw (1 Pet. 2:6), prose,cw (Mt. 7:15), u`pere,cw (Ph. 4:7). Here the substantive has dropped out in most cases and the verb comes to stand alone (cf. prose,cw nou/n). Cf. avnaka,mptw (Mt. 2:12), evkkli,nw (Ro. 16:17) and prosko,ptw (Jo. 11:9). Katapau,w is transitive in Ac. 14:18, but intransitive in Heb. 4:4, 10. Cf. avporri,ptw in Ac. 27:43. Stre,fw shows intransitive compounds with avna- (Ac. 5:22), avpo- (Ac. 3:26), evpi- (Lu. 2:39). The modern Greek surpasses even the koinh, in its facility for making all sorts of compound verbs (tr. and intr.) and in particular verbs compounded . with nouns, like evteknotro,fhsen and evxenodo,chsen (1 Tim. 5:10). Cf. Thumb, Handb., p. 112.

(d) DIFFERENT TENSES VARY. Thus where both second and first aorists occur, the second is intransitive and the first transitive. Cf. e;sth (Lu. 6:8), but e;sthsen auvto, (Mk. 9:36). This distinction applies to all the compounds of i[sthmi. Acts 27:28 ( diasth,santej) is no exception, as th.n nau/n is to be supplied. Some of the "strong" or primitive perfect actives are intransitive when the present is transitive. Thus avne,w|ga (1 Con 16:9) from avnoi,gwà avpo,lwla (Mt. 10:6) from avpo,llumià e`sta,nai (Lu. 13:25) from i[sthmià pe,poiqa <LE>

VOICE ( DIAQESIS) 801

(Ro. 2:19) from pei,qwà se,shpa (Jas. 5:2) from sh,pw. Moulton¹⁷ seems to confuse "transitive" with "active," and "intransitive" with "middle" in his discussion of these perfects: "We have a number of cases in which the 'strong' perfect active attaches itself in meaning to the middle." The middle is not in itself intransitive, nor is the active in itself transitive. "The conjecture that the perfect originally had no distinction of active and middle, its person-endings being peculiar throughout, affords the most probable explanation of the facts: when the much later - ka perfect arose, the distinction had become universal." It is doubtless true that in the primitive - a perfect there was no distinctive middle form. But why seek for a middle sense in the primitive perfect active because it happens in many cases to be intransitive? It does happen that ge,gona (Jo. 1:4) is found with gi,nomai and evlh,luqa (Jo. 17:1) from e;rcomai, two intransitive middles. It is also true that future middles are the rule with a few verbs which have this primitive, but not always intransitive, perfect. So it is with avkh,koa (trans., Ac. 6:11), ei;lhfa (trans., Rev. 11:17), pe,ponqa (intr. as the verb itself is, Lu. 13:2), te,tuca (trans., Heb. 8:6). So with ke,kragen (Jo. 1:15, intr. like the verb itself), though kekra,xomai (some MSS. in Lu. 19:40) is future perfect middle. Oi=da (Jo. 10:4) is transitive, though defective, while e;oika (Jas. 1:6), like ei;wqa (Mk. 10:1), is intransitive. But ge,grafa (Jo. 19:22) is transitive.

(e) THE ACTIVE AS CAUSATIVE. But this usage is not due to the voice, and is, besides, common to all languages.¹⁸ Cf. the Hebrew Hiphil conjugation. Viteau (" Essai sur la Syntaxe des Voix dans le Grec du N. T.," Revue de Philologie, 1894, p. 2) says that the Greek voices would not be strange to a Jew who was used to the seven conjugations of the Hebrew verb. But the point is not strictly parallel. In one sense this idiom is due to the fact that what one does through another he does himself.¹⁹ Cf. to.n h[lion auvtou/ avnate,llei. (Mt. 5:45), strictly causative. But in Jo. 19:1, e;laben o` Pila,toj to.n vIhsou/n kai. evmasti,gwsen, the other kind of causative occurs. So also with perie,temen (Ac. 16:3). There was indeed a remarkable increase in the LXX in the number of verbs used in the causative sense, many of which had been usually intransitive. Cf. basileu,w, which occurs 36 times in the causative sense in the LXX (cf. Judg. 9:6).²⁰ The Hebrew Hiphil is partly

802 A GRAMMAR OF THE GREEK NEW TESTAMENT

responsible for this increase.²¹ See further verbs in - ow, like katadoulo,w (Gal. 2:4).

(f) ACTIVE WITH REFLEXIVES. Certainly there is nothing unusual in this construction. Cf. sw/son seauto,n (Mk. 15:30), e;balen e`auto,n (Jo. 21:8), prose,cete e`autoi/j (Lu. 17:3). Cf. Jo. 21:18. Blass²² indeed says that the "active for middle" occurs. One hesitates to subscribe to that dictum. It is indeed true that the use of the reflexive pronoun with the active brings out much more sharply the reflexive relation than the mere middle. It is not necessary to say that katadouloi/ (2 Cor. 11:20) is used "for" the middle. It is true that peira,zw in the koinh, supplants the Attic peira,omai, but this is not due to a confusion of voice. With poie,w the N. T. does show a number of examples of the active where the middle was more common in the Attic, though the N. T. generally has poiei/sqai avnabolh,nà lo,gonà porei,anà spoudh,n. And the MSS. vary greatly between active and middle of poie,w with words like monh,n (Jo. 14:23), kopeto,n (Ac. 8:2), sunwmosi,angrk grk grk(23:13), but not with sumbou,lion (Mk. 15:1), evkdi,khsin, (Lu.18:7 f.), sustrofh,n (Ac. 23:12), po,lemon (Rev.11:7). But this is precisely what we find in the (inscriptions and papyri). Cf. Radermacher, N. T. Gr., p. 120. So even bia,zw and evpilanqa,nw (Mayser, Gr., p. 386). The same tendency appears in modern Greek (Thumb, Handb., p. 114). Cf. die,rrhxen ta. i`ma,tia au`tou/ (Mt. 26:65). In these examples Blass has in my judgment read too much into the active voice. But it is certain that in prose,cete e`autoi/j (Lu. 12:1) there is more emphasis on the reflexive idea than in fula,ssesqegrk grk(12:15). Cf. Moulton, Prol., p. 157.

(g) IMPERSONAL ACTIVE. Some impersonal verbs occur in the active. Cf. perie,cei evn th|/ grafh|/ (1 Pet. 2:6), and e;brexen (Jas. 5:17).

(h) INFINITIVES. These do not always reflect the force of the voice, especially in the " epexegetic" use,²³ like our English "fair to see," "good to eat." Cf. kriqh/nai and labei/n, Mt. 5:40. The infinitive has no voice in Sanskrit. See further under Infinitive (ch. XX, Verbal Nouns).

(i) ACTIVE VERBS AS PASSIVES OF OTHER VERBS.²⁴ Thus avpoÄ qnh,skw is more common than the passive of avpoktei,nw ( ÄÄkte,nnw) though examples of this passive occur in the N. T. (Rev. 6:11, etc.). W. H. read kakw/j e;cei in Mt. 17:15 rather than kakw/j pa,scei (cf. poiw/ kalw/j, etc.). So evkpi,ptw (Ac. 27:17, 26, 29) occurs

VOICE ( DIAQESIS) 803

as passive of evkba,llw, but note evkba,llesqai in Mt. 8:12. Cf. Gildersleeve, Syntax, p. 75. In 1 Cor. 11:18 avkou,w has the classic turn I am told.' But in 5:11 avkou,etai the passive itself occurs in the sense 'It is reported.' But in all such cases the distinction between the voices is not really lost.

III. The Middle Voice ( dia,qesij me,sh)

(a) ORIGIN OF THE MIDDLE. See chapter VIII, vi, (c), for the uncertainty as to the priority of active and middle. That question is an open one and must be left open. Both active and middle appear in Sanskrit and in Homer. The prehistoric situation is purely speculative. Logically the active would seem to come first, though the difference in form may be due to variation in sound (ablaut).²⁵ Probably at first there was neither active nor middle, the distinction being a development. In the Sanskrit²⁶ we meet a full system of both active and middle forms for all the tenses (not all the modes), the participle, however, having only a partial system and the infinitive no voice at all. But each verb has its own development and that was by no means uniform. Some had a very limited use as to voice, tense and mode. In Homer indeed the middle is rather more common than in later Greek.²⁷ It is only in the Sanskrit, Zend (Old Persian), Greek and Gothic that the middle is kept as a distinct voice.²⁸ In the Gothic only remnants of the middle are found,²⁹ while in Latin the middle as a separate voice disappears.³⁰ It is very difficult to run a parallel between the Latin and Greek voices. But there is a considerable remnant of Latin middles like miror, sequor, utor (cf. Draeger, Hist. Syntax, pp. 145 ff.). The final disappearance of the Greek future and aorist middle before the passive is well sketched by Jannaris.³¹ But at first we are not to think of the passive at all, that interloper that finally drove the middle out of use.

(b) MEANING OF THE MIDDLE. It is urged that the term "middle" is good because the voice in meaning stands between the active and the passive.³² But, unfortunately for that idea, the middle is older than the passive. It is true that the passive arose out of the middle and that the middle marks a step towards

804 A GRAMMAR OF THE GREEK NEW TESTAMENT

the passive. The passive idea existed before there was a separate passive form, a thing never true of all tenses and all verbs. The Hebrew Hithpael conjugation is somewhat parallel,³³ but not wholly so. The only difference between the active and middle voices is that the middle calls especial attention to the subject. In the active voice the subject is merely acting; in the middle the subject is acting in relation to himself somehow. What this precise relation is the middle voice does not say. That must come out of the context or from the significance of the verb itself. Gildersleeve³⁴ is clearly right in holding that the interpretation of the difference between active and middle is in many cases more lexical than grammatical. "The middle adds a subjective element."³⁵ Sometimes the variation from the active is too minute for translation into English. This "word for one's self" is often very difficult of translation, and we must not fall into the error of explaining the force of the middle by the English translation.

(c) OFTEN DIFFERENCE FROM ACTIVE ACUTE. As examples note: ai`re,w, 'I take'; ai`re,omaià 'I take to myself' ('choose'); avnamimnh,Ä skw, 'I remind'; avnamimnh,skomai, 'I remind myself' ('remember'); avpe,cw, 'I hold off'; avpe,comai, 'I hold myself off' ('abstain'); avpodi,dwmi, 'I give back'; avpodi,domai, 'I give back of my own' ('sell'); avpo,llumi, 'I destroy'; avpo,llumai, I perish'; a[ptwà 'I fasten'; a[ptomai, 'I touch'; a;rcw, 'I rule'; a;rcomai, 'I begin'; bouleu,w, 'I counsel'; bouleu,omai, 'I take counsel' ('deliberate'); game,w, 'I marry' ('bridegroom'), game,omai ('bride'); geu,w 'I give to taste'; geu,omai, 'I taste'; gra,fw, 'I enrol'; gra,fomaià 'I indict' (but 'enrol one's self' in Lu. 2:5); danei,zw, 'I lend'; danei,zomai, 'I borrow'; dida,skw, 'I teach'; dida,skomai, 'I get taught'; i[sthmi 'I place'; i[stamai, 'I stand'; lanqa,nw, 'I escape notice'; lanqa,nomai, 'I forget'; misqo,w, 'I let,' misqo,omai, 'I hire'; pau,w, 'I make to cease'; pau,omai, 'I cease'; pei,qw, 'I persuade'; pei,qomai, 'I obey'; fai,nw, 'I show'; fai,nomai, 'I appear'; fobe,w, 'I frighten'; fobe,omai, 'I fear.' These examples in the N. T. illustrate the difference between the two voices.³⁶

(d) THE USE OF THE MIDDLE NOT OBLIGATORY.³⁷ This remark may sound like a truism, but it is justified when one can read this: "As the active is used in place of the middle, so the middle

VOICE ( DIAQESIS) 805

often stands for the active which would naturally be expected."³⁸ Winer³⁹ also speaks of the two voices being used "interchangeably." But Winer loses one of his examples, for W. H. have sugkalei/ in Lu. 15:9, as in verse 6. Winer correctly says that "it depended on the writer" which he would use. Of course, but that is not to say that no distinction existed. In Jas. 4:2 f., aivtei/te kai. ouv lamÄ ba,neteà dio,ti kakw/j aivtei/sqe, the middle seems rather on purpose ('ye ask for yourselves amiss,' Farrar, Gk. Syntax, p. 118). Blass⁴⁰ calls this "an arbitrary interchange," though he admits in general the N. T. use of aivte,w for ordinary requests (as from God), but aivte,omai in business transactions (its usual use in he N. T., Mt. 27:20; Lu. 23:23). This may be the very point in Jas. 4:2 f. and 1 Jo. 5:14. Moulton⁴¹ agrees with Mayor (James in loco) on the correctness of the distinction. Mayor (in loco) says: "When aivtei/te is thus opposed to aivtei/sqe, it implies using the words, without the spirit of prayer." See the same distinction drawn in Mk. 6:22-25; 10:35, 38 (Mt. 20:20, 22); 1 Jo. 5:15. Blass (Gr. of N. T. Gk., p. 186 note) observes that Herod's offer to Salome gave her business relations to him justifying her use of the middle (Mk. 6:24 f.). When the active and the middle occur side by side the attention is drawn to the distinction. It is to be recalled again that the same verb varied in different stages of the language in the voice used. Hence it is hardly pertinent to bring an indictment against the N. T. writers, because the middle is not used with all verbs just as it was in the Attic Greek. As a matter of fact, Homer differs from the Attic. Blass (Gr. of N. T. Gk., p. 186) succinctly says that "the New Testament writers were perfectly capable of preserving the distinction between the active and the middle." So in Mk. 14:47 note spasa,menoj th.n ma,cairan, while in Mt. 26:51 we have avpe,spasen th.n ma,cairan auvtou/. In Matthew we have the pronoun auvtou/ and avpo, supplanting the middle in Mark (cf. Radermacher, N. T. Gr., p. 120 f.). Radermacher (op. cit., p. 119), however, as a result of his researches, finds in the koinh, "Unsicherheit im Gebrauch des Mediums." The point of the middle is not the same always. So in Ac. 7:24 avmu,nesqai= 'assist,' not 'ward off from one's self,' but the force of the middle is present. So in Col. 2:15, avpekdusa,menoj ta.j avrca,j, it is not 'undress,' but 'throw off from one's self.' Cf. also plhrou/sqai in

806 A GRAMMAR OF THE GREEK NEW TESTAMENT

Eph. 1:23 and plhrou/n in 4:10. Moulton⁴² shows that there is as much freedom in the papyri in the use of active and middle as in the N. T. Thus eva.n ai`rh/te and eva.n ai`rh/sqe (G. H. 36, B.C. 95) occur side by side. So gamei/sqai=nubere fell out of use. See also it, (f).

(e) EITHER TRANSITIVE OR INTRANSITIVE. Thus eva.n mh. ni,ywntai ta.j cei/raj (Mk 7:3) and h[yanto auvtou/ grk grk(6:56), but evxi,stanto (6: 52) and eivseporeu,ontogrk grk(6:56) are intransitive. The middle is not, therefore, intransitive in itself: That is a matter that belongs to the verb-stem. As to the future middles, like bh,somai, see discussion a little later. Some verbs, indeed, are transitive in the active, but intransitive in the middle ( avpo,llumià avpo,llumaià fai,nwà fai,nomai). Cf. Hatzidakis, Einl., pp. 201 ff.; Thompson, Syntax, p. 161.

(f) DIRECT MIDDLE. It is necessary to discuss the various uses of the middle, but the divisions made by the grammarians are more or less arbitrary and unsatisfactory. They are followed here merely for convenience. The middle voice is very broad in its scope and no one word, not even reflexive, covers all the ground. It is essentially the voice of personal interest somewhat like the dative case. Grosse (Beitrage zur Syntax des griechischen Mediums and Passivums, 1891, p. 4) denies that the reflexive is the original use of the middle. But Rutherford (First Gk. Syntax, 1890, p. 74), derives both passive and middle out of the reflexive use. For the various uses of the middle in Homer, who is specially fond of this voice, see Monro, Homeric Gr., p. 7. But, curiously, Monro mentions "the Intransitive use" as one of the separate idioms of the middle. Nearly every grammarian⁴³ has his own division of these "uses" of the middle, none of which the Greeks themselves had. Gildersleeve⁴⁴ is justly impatient with this overrefinement and observes that "one must needs fall back on the way of the language," which "is capricious in such matters." It is needless to take up philosophical abstractions like "subjective" and "objective." It is not possible to tell whether the direct middle (reflexive middle) was the original use of the voice or not. The direct middle is comparatively rare in Homer and in the early Greek generally.⁴⁵ It began in the koinh, to disappear, before the active and the reflexive pronoun (cf. N. T.), but the direct middle

VOICE ( DIAQESIS) 807

revived again as the indirect middle disappeared before the passive because of "its subtle meaning."⁴⁶ Hence in Neo-Hellenic "almost every transitive verb, if active, admits of a direct middle."⁴⁷ In modern Greek this direct reflexive is nearly the sole use of the middle.⁴⁸ The modern Greek has no distinction in forms between middle and passive, but the middle signification survives. Thus lou,zomai means 'I bathe myself' (Thumb, Handb., pp. 111, 114). Thumb finds the direct reflexive use common. Moulton⁴⁹ practically confines this idiom in the N. T. to avph,gxato (Mt. 27:5), 'he hanged himself,' and even here Moulton suggests 'choked' as a truer English translation. This is indeed "a survival from classical Greek," but there seem to be other N. T. examples also. The example cited by Winer⁵⁰ from Jo. 8:59 (cf. also 12:36), evkru,gbh, is passive, as Moulton⁵¹ points out. But in u-j lousame,nh (2 Pet. 2:22) the direct middle is evident, as Moulton admits in the Appendix (p. 238). Cf. lou,sasqe (Is. 1:16), 'wash you.' Note also avpelou,sasqe, 'washed yourselves' (1 Cor. 6:11, correct translation in margin of Rev. V.). A good example also is qermaino,menoj (Mk. 14:54), 'warming himself' (Rev. V.). It is rather gratuitous to doubt the direct middle parakskeua,setai, 'prepare himself' (1 Cor. 14:8). But Moulton adds mh. sku,llou (Lu. 7:6) to Winer's list and illustrates by "the illiterate contemporary papyrus O.P. 295, mh. sklu,lle e`ath,n" (active and reflexive pronoun). So also r`anti,swtai, (W. H., Mk. 7:4) and bapti,swntai (marg.) are both direct middles. Zw/sai (Ac. 12:8), 'gird yourself,' is also direct middle. [Aptomai (Col. 2:20) is probably direct middle, subject yourselves to ordinances.' And u`pota,ssesqe (Col. 3:18) may be also. [Aptomai ('fasten myself to,' 'touch') is really the direct middle (Mk. 8 : 22). vEpekteino,menoj (Ph. 3:13) is 'stretching myself forward.' Cf. also u`pesteila,mhn (Ac. 20:27), 'withdraw myself'; avntitasso,menoj (Ro. 13:2), 'line one's self up against.' In the case of periba,llomai it is probable that we have the direct middle 'clothe one's self' (Mt. 6:29). The accusative of the thing is added in Rev. 3:18. It is possible to regard avnapau,esqe (Mt. 26:45) as direct middle. vApogra,yasqai, (Lu. 2:5) may be merely the direct middle, 'enrol himself,' though the causative idea is possible. In Lu. 12:15 vAneco,menoi avllh,lwn ('guard yourselves from') follows the classic idiom. vApogra,yasqai (Eph. 4:2) is also the direct middle, 'holding yourselves back from one an-

808 A GRAMMAR OF THE GREEK NEW TESTAMENT

other.' The same thing is true of avpe,cesqai eivdwloqu,twn (Ac. 15:29). In 1 Pet. 5:5 tapeinofrosu,nhn evgkoubw,sasqe, 'gird yourselves with humility,' we may have the same idiom. In Ac. 18: 5, sunei,ceto tw|/ lo,gw|, we may have the direct middle, 'held himself to the word.' There are to be added, besides, some of the causative middles, like ba,ptisai. (Ac. 22:16), 'get yourself baptized' (cf. evbapti,santo, 1 Cor. 10:2). It is true that the list is not a large one, but the idiom is clearly not obsolete in the N. T. The causative middle has a wider use also, as will be shown directly.

(g) CAUSATIVE OR PERMISSIVE MIDDLE. Cf. the German sich, lassen. This occasional use of the middle does not distinguish it from the active and occurs both with the direct and the indirect use of the middle.⁵² It is just so in modern Greek (Thumb, Handb., p. 114 f.). It is, like transitive and intransitive, more the notion of the word than a phase of the middle voice.⁵³ In later Greek the causative sense occurs only with the direct middle.⁵⁴ It is not to be forgotten that originally there was no passive form at all. The verb-idea and the context then alone decided the voice as between middle and passive. Even in the aorist and future, where the passive later has a distinct form, the line was not always sharply drawn, especially in the future. More about this a little later. But in the aorist in particular one hesitates to find a passive voice in the middle form, though it sometimes happens. Some few of these causative middles could be explained as passives, but by no means all. Certainly evklexame,nouj (Ac. 15: 22) is a true middle. A considerable residuum remains. "In Tb.P. 35 (ii/B.C.) e`auto.n aivtia,setai, 'will get himself accused,' is a middle."<sup>55</sup> In Ac. 22:16, ba,ptisai kai. avpo,lousai ta.j a`marti,aj sou, we have the causative middle, one a direct, the other an indirect, middle, 'get yourself baptized and get your sins washed away.' So then evbapti,santo (W. H. text in 1 Cor. 10:2) is causative, though many MSS. read evbapti,sqhsan. Blass⁵⁶ has eccentric notions of textual criticism, for he rejects the middle here and contends for it in Lu. 11:38 on the authority of one minuscule! Blass⁵⁷ also argues that the sense of 'let' or ' allow' belongs to the passive rather than to the middle, but this is by no means certain. Thus avdikei/sqe and avposterei/sqe (1 Cor. 6:7) may be middles (cf. actives in next verse), ' let yourselves be wronged and robbed.'

VOICE ( DIAQESIS) 809

This permissive sense of the middle is closely allied to the causative and approaches the passive.⁵⁸ In Lu. 2:5 avpogra,yasqai may be (see (f) above) causative, 'have himself enrolled,' though avpoÄ gra,gesqaigrk grk(2:1) is passive. In Mt. 5:42 dani,sasqai is to have money lent' ('to borrow'). Misqw,sasqai (Mt. 20:1) is to let out for wages' ('to hire'). In 1 Cor. 11:6, keira,sqwà kei,rasqai h' xura/sqai, (or xu,rasqai), we find the permissive middle. Cf. xurh,Ä sontai th.n kefalh,n (Ac. 21:24). But avpoko,yontai (Gal. 5:12) is causative, 'have themselves castrated' (cf. Deut. 23:1). So avpeÄ lou,sasqe, according to text of Rev. V. (1 Cor. 6:11). In Rev. 3:5 peribalei/tai comes rather close to the passive sense. See (f) above. In Lu. 14:18, 19, e;ce me parh|thme,non, we have a construction more like modern English. The causative idea in avnaÄ kefalaiw,sasqai ta. pa,nta evn tw|/ Cristw|/ (Eph. 1:10) is not due to the voice, but to the verb itself (- o,w).

(h) INDIRECT MIDDLE. In the flourishing period of the language this was by far the most frequent use, but it finally faded before the active and the intensive (reflexive) pronoun or the passive.⁵⁹ In 1 Cor. 15:28, u`potagh,setai, the passive may bear the middle force (Findlay, Expos. Gr. T., in loco). But in general the indirect middle is abundant and free in the N. T. In the modern Greek Thumb gives no instances of the indirect middle. The precise shade of the resultant meaning varies very greatly. The subject is represented as doing something for, to or by himself. Often the mere pronoun is sufficient translation. Each word and its context must determine the result. Thus in Heb. 9:12, aivwni,an lu,trwsin eu`ra,menoj, Jesus is represented as having found eternal redemption by himself. He found the way. In Mt. 16:22, proslabo,menoj auvto,n, 'Peter takes Jesus to himself.' In Mk. 9:8, peribleya,menoi, 'the disciples themselves suddenly looking round.' In Lu. 8:27, ouvk evnedu,sato i`ma,tion, 'did not put a garment on himself.' In 8:52, evko,ptonto auvth,n the word has really changed meaning, 'they beat themselves for grief as to her' ('bewailed her'), actually a direct middle. "We have, in fact, to vary the exact relation of the reflexive perpetually if we are to represent the middle in the form appropriate to the particular example."<sup>60</sup> That is precisely the case. So proskalesa,menoj (Mt. 10:1) represents Jesus as calling the disciples to himself. Cf. eivskalou/mai (Ac. 10:23). So proslamba,nesqe (Ro. 15:7; cf. also prosela,beto) is 'take to yourselves.' Kai,sara evpikalou/mai (Ac. 25:11) is 'I call upon Caesar in my behalf.' Ai`rh,so-

810 A GRAMMAR OF THE GREEK NEW TESTAMENT

mai (Ph. 1:22) is 'I take for myself' ('choose'), while kth,shsqe (Mt. 10:9), though only in the middle, means 'provide for yourselves' ('procure'). In spasa,menoj th.n ma,cairan (Mk. 14:47), the possessive is probably sufficient, 'drawing his own sword' (cf. avpe,spasen auvtou/ in Mt. 26:51). vEktinaxa,menoj ta. i`ma,tia (Ac. 18:6) is rather 'shaking out his clothes from himself,' while avpeni,yato ta.j cei/raj (Mt. 27:24) is probably 'he himself washed his hands.' In avpwqei/sqe auvto,n (Ac. 13:46; cf. Ro. 11:1) the idea is 'ye push it away from yourselves' ('reject'). vApe,dosqe (Ac. 5:8) is 'ye gave away for your own interest' ('sold'). vEnosfi,sato (Ac. 5:2) means 'kept back for himself.' In evpideiknu,menai citw/naj (Ac. 9:39) the women were 'showing garments belonging to themselves.' Note the fulness of meaning in periepoih,sato (Ac. 20:28). Cf. paraÄ threi/sqe (Gal. 4:10), avpeipa,mhn, (2 Cor. 4:2), evktre,pomai (1 Tim. 6:20). In diezw,sato (Jo. 21:7) we have 'he girded round himself.' Paraith,shsqe (Heb. 12:25) is 'beg off from yourselves' ('reject'). In Col. 4:5, to.n kairo.n evxagorazo,menoi, we have 'buying the opportunity for yourselves out of the open market.' vApoqe,menoi. (Heb. 12:1) is 'laying aside from yourselves every weight.' In evxele,xato (Lu. 10:42) we have 'she selected for herself' ('chose'). vEneÄ didu,sketo (Lu. 16:19) is 'he put clothes on himself,' though this may be direct middle with accusative of thing added. KatoptriÄ zo,menoi (2 Cor. 3:18) is probably 'beholding for ourselves in a mirror.' In Ro. 3:25, oa}n proe,qeto o` qeo,j, note that it was God's own Son whom he set forth. This free indirect reflexive use came to be the typical middle in the flourishing period of the Greek language. No fixed rule can be laid down for the translation of this or any other use of the middle. Even "deponents" like cra,omai may be indirect middles. This word from crh, ('necessity') means 'I make for myself what is necessary with something' (Moulton, Prol., p. 158). An interesting group of middles occurs in Ac. 24: 22-25, avneba,letoà diagnw,somaià diataxa,menojà parageno,menojà metepe,myatoà dialegome,nouà poreu,ouà metakale,somai. These are not all "indirect" middles, as is obvious. Cf. also evkballo,menoi. (Ac. 27:38) and prosela,beto (Ro. 14:3). It is interesting to note the difference between parei/ce in Ac. 16:16 (the damsel who furnished gain for her masters) and parei,ceto in Ac. 19:24 (Demetrius who furnished gain for his craftsmen and himself). So pei,qw is 'to exercise suasion,' and pei,qomai 'to admit suasion to one's self' (Moulton, Prol., p. 158).

(i) RECIPROCAL MIDDLE. Since e`autw/n was used in the reciprocal sense, it was natural for the middle to fall in with this idiom.

VOICE ( DIAQESIS) 811

Addenda 2nd ed.

Thus sunebouleu,santo (Mt. 26:4), 'they counselled with one another,' does not differ radically from evxele,gonto (Lu. 14:7), 'they selected the first seats for themselves.'⁶¹ So also evboule,santo (Jo. 12:10), sunete,qeintogrk grk(9:22), sunanami,gnusqai (1 Cor. 5:9), kri, nesqaigrk grk(6:1), evma,conto (Jo. 6:52), dialego,menoj (Ac. 19:8. In Mk. 9:34, pro.j avllh,louj diele,cqhsan, we have passive deponent with reciprocal pronoun).⁶² The reciprocal middle survives in modern Greek (Thumb, Handb., p. 114). For classic examples see Gildersleeve, Syntax, p. 66.

(j) REDUNDANT MIDDLE. Here the pronoun and the middle both occur. This idiom is found as early as Homer and indicates a dimness in the force of the middle on the part of the speaker. "The effect is artificial" according to Thompson.⁶³ Gildersleeve (Syntax, p. 68) sees in this idiom the effort to bring out more clearly the reflexive force of the middle. Moulton (Prol., p. 162) cites from the papyri e`atuto.n aivtia,setai. Tb.P. 35 (ii/B.C.). This redundance probably began very naturally. Thus in Ac. 7:58, avpe,qento ta. i`ma,tia auvtw/n, the personal pronoun is added, not the reflexive. So in u`po,dhsai ta. sanda,lia, sou and peribalou/ to. i`ma,tio,n sougrk grk(12:8) and a;leiyai, sou th.n kefalh,n (Mt. 6:17). Cf. ni,ptontai ta.j cei/raj (Mt. 15:2) without the pronoun. So in Lu. 14:1, kai. auvtoi. h=san parathrou,menoi, the auvtoi. wavers between mere personal and intensive. Cf. the active in Eph. 5:27, parasth,sh| auvto.j e`autw|/. But in Jo. 19:24 the LXX quotation is given as diemeri,santo- e`autoi/j, while in Mt. 27:35 it is merely diemeri,santo. Note also seauto.n pareco,menoj (Tit. 2:7) and poiou/mai - e`mautw|/ (Ac. 20:24). See also avneqre,yato auvto.n e`auth|/ eivj ui`o,n, (Ac. 7:21) and 1 Tim. 3:13 e`autoi/j peripoiou/ntai. Most of the examples, however, in the N. T. occur with verbs which are not found in the active. Cf. Lu. 9: 23 avrnhsa,sqw e`auto,n, Ac. 24:10 ta. peri. evmautou/ avpologou/mai, 26: 2 h[gnmai evmauto,n, Ph. 3:12 evmauto.n ou;pw logi,zomai)

$k) DYNAMIC (DEPONENT) MIDDLE. "I would fain call the drip-pan middle, the pande,kthj middle, the middle that is put at the bottom to catch the drippings of the other uses."⁶⁴ And this is the most difficult use of the middle to explain. Some writers distinguish between the dynamic and the deponent. Others, like Thompson,⁶⁵ make the dynamic include the deponent. The name "deponent" is very unsatisfactory. It is used to mean the laying

812 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 2nd ed.

aside of the active form in the case of verbs that have no active voice. But these verbs in most cases never had an active voice. Moulton⁶⁶ is clearly right in his contention that the term in reality applies as well to active verbs that have no middle as to middle verbs that have no active. The term is usually applied to both middles and passives that have no active (Clyde, Gk. Syntax, p. 61). Others⁶⁷ use the term for middle verbs that have no longer a reflexive idea. But "deponent" is a very poor definition. Nor is the word "dynamic" much better. Winer's remark⁶⁸ is not very lucid: "From Middle verbs are to be carefully distinguished Deponents." They are indeed either transitive or intransitive, but some are in the middle voice, others passive. But the point about all the "dynamic" middles is that it is hard to see the distinctive force of the voice. The question is raised whether these verbs have lost the middle idea or never had it. "Like the rest of us, Stahl has to go into bankruptcy," Gildersleeve⁶⁹ remarks on Stahl's attempt to explain this use of the middle. Moulton (Prol., p. 158) thinks that in these verbs "it is useless to exercise our ingenuity on interpreting the middle, for the development never progressed beyond the rudimentary stage." But these verbs persist in the modern Greek (Thumb, Handb., p. 113). It is possible that the Greeks were more sensitive to the exact force of this middle than we are, just as they used the intensive particles so freely. Where guessing is all that we can do, is it not clear that these "dynamic" middles represent the original verb before the distinction was drawn between active and middle? The French says je m'apercois, 'I perceive.' The intensive force of this middle is partially seen in verbs of mental action which are so common in Greek, like aivsqa,nomai (Lu. 9:45), avrne,omai (Lu. 12:9), proaitia,omai (Ro. 3:9), avspa,zomai (Ac. 25:13), diabebaiou/mai (Tit. 3:8), kataÄ lamba,nomai (Ac. 4:13, but note katalamba,nw in the same sense in Ph. 3:12), evnte,llomai (Heb. 11:22), evpilanqa,nomai (Mt. 16:5), eu;comai (Ro. 9:3), h`ge,omai (Ph. 3:8), logi,zomai (Ph. 4:8), mai,nomai (Ac. 26:25), me,mfomai (Ro. 9:19), fei,domai (Ro. 8:32). I imagine that the personal interest of the subject is not so difficult to recognise in such verbs, especially since in a word like katalamba,nomai it is not "deponent," but occurs also in the active. The papyri vary,⁷⁰ as does the N. T. in the use of poiou/mai and poiw/ with nouns. Thus we have sumbou,lion poih,santej (Mk. 15:1), but mnei,an poiou,-

VOICE ( DIAQESIS) 813

menoj (Eph. 1:16). There is the utmost freedom in the matter in the N. T. Not all the "deponents" of mental action are middles in the aorist. Cf. bou,lomaià evnqume,omaià evpimele,omaià euvlabe,omai. These are commonly called passive deponents in the present as well as in the aorist and future, but the matter is not clear by any means. At any rate there are middle verbs which are very hard to explain, like gi,nomai. (Mt. 8:26), a;llomai (Jo. 4:14), avfikne,omai (Ro. 16:19), diamartu,romai (Ac. 2:40), e;rcomai, (Jo. 1:39), evrga,zomai (Mt. 25: 16), kaqe,zomai (Mt. 26:55), ka,qhmai (Mt. 13:1), sune,pomai. (Ac. 20: 4; cf. sequor). Kei/mai is probably passive. It is not hard to see the reflexive idea in de,comai. (Mt. 10:14). Perible,pomai is always middle in the N. T. (cf. Mk. 3:5), accenting the movement of the eyes or concern expressed in the look. There are also passive deponents that correspond to this list that really do not seem to be passive in idea, like bou,lomaià du,namaià fobe,omai. Some of these verbs have both middle and passive forms, like gi,nomai $evge,netoà evgenh,qhn%, de,comai $evde,xatoà evde,cqhn%. Not all of these middle "deponents" have middle forms in all tenses. Cf. ge,gonaà h=lqon, evlh,luqaà e;laqon. Then, again, some verbs have the deponent or dynamic middle only in the future, like o;yomai, though Homer is fond of the middle forms of this verb.⁷¹ But the aorist and future middle call for special treatment.

(l) MIDDLE FUTURE, THOUGH ACTIVE PRESENT. Some verbs, active in the other tenses, have the future only in the middle. No real explanation of this phenomenon is known. For a list see chapter VIII, VI, (d). Some of them are really separate verbroots, as o`ra,wà o;yomai* evsqi,wà fa,gomai. Others represent a special variation of the future form, like avpoqanou/maià pesou/maià pi,omai, but both komi,somai and komiou/mai. Others are regular enough, like avkou,somai, - bh,somaià gnw,somaià e;somaià qauma,somaià te,xomaià feu,xomai) In other instances the old classic middle has vanished in the N. T. before the active future, as in a`marth,swà avpanth,swà a`pra,swà gela,swà klau,swà kra,xwà pai,xwà r`eu,sw, etc. Some verbs, like avkou,w, za,w, use either voice in the future. Some of these middle futures create no difficulty. Thompson⁷² calls them all "strict middles," but most of them are as "deponent" as the verbs in the previous section. Clyde⁷³ quotes Curtius' explanation that an act in the future lies mainly in the mind of the speaker. But on the whole the matter remains unexplained, though the number has greatly decreased in the N. T. as in the koinh, generally.⁷⁴ See also Dieterich,

814 A GRAMMAR OF THE GREEK NEW TESTAMENT

Untersuch., p. 205; Radermacher, N. T. Gr., p. 120. Moulton⁷⁵ justly takes "the existence of this large class of futures as additional evidence of a close connection between the middle flexion and the stressing of the agent's interest in the action of the verb." The use of the middle future (and occasionally aorist) as passive comes under the passive voice, for it is really passive. See under IV.

(m) THE MIDDLE RETREATING IN THE N. T. This is happening because of the active (cf. a`marth,sw above) as well as the passive. This is true of the koinh, in general.<sup>76</sup> There was a considerable amount of variation and even of confusion among writers in the later period.<sup>77</sup> Different words had different histories in the matter. But we have just seen from the list of "dynamic-deponent" middles plenty of evidence that from the day of Homer on the function of the middle voice was indistinct in many verbs.<sup>78</sup> "The accuracy with which the middle was used would naturally vary with the writer's Greek culture."<sup>79</sup> And, it may be added, with the author's feelings at the moment. The judgment of Simcox<sup>80</sup> is right, that the middle "is one of the refinements in Greek idiom which is perhaps beginning to be blurred in some of the N. T. writers, but is preserved to a greater or less extent in most." But it is no more "blurred" than in other writers of the koinh,. It is simply that all the distinctions of earlier times did not survive with all the verbs. On the whole, in the N. T., aivtw/ is used colloquially and aivtou/mai for the more elevated style, but usage varies with different writers as in the LXX. Cf. Abbott, Johannine Gr., p. 389. So u`stere,w in Heb. 4:1, but u`sterou/mai in Ro. 3:23. But the change in the N. T. is mainly in the disuse of the middle, not in a new use of it. From Homer to modern Greek plenty of middles are hard to define, and the N. T. is no more erratic than the rest of Greek, not to say of the koinh, (Moulton, Prol., p. 159). But the delicate distinctions between the active and the dynamic middle are lost in modern Greek (Thumb, Handb., p. 112), if indeed they ever really existed.

IV. The Passive Voice ( dia,qesij paqhtikh,).

(a) ORIGIN OF THE PASSIVE. See chapter VIII, VI, (e), for a discussion of the rise of the passive voice.⁸¹ In Sanskrit the middle

VOICE ( DIAQESIS) 815

was liable to be used in the passive sense.⁸² As is well known in Homer, the future passive forms do not occur except two, migh,sesqai and dah,seai (Stahl, Syntax, p. 66), and the distinction between aorist middle and aorist passive is indistinct. Indeed, strictly speaking, there was no passive voice as to form in Greek, as there was none in the original Indo-Germanic speech.⁸³ The passive sense was developed in various languages in different" ways. This sense may be due to verbs of state, but Greek fell upon various devices like the active of some verbs ( kakw/j e;cwà pa,scw), the mere use of the middle, the development of two special tenses by the use of active endings (aorist) and middle (future) with a special suffix. In Homer⁸⁴ evblh,mhnà evkta,mhnà evsco,mhn occur as passives just like evsce,qhnà e;comai) "Even in Attic evsco,mhn appears as a passive, evsce,qhn being late."⁸⁵ In Homer also the distinctive aorist passive form sometimes has practically the active or middle signification.⁸⁶ This much of repetition is necessary to get the position of the passive clearly before us. It is really no voice at all in form as compared with the active and middle. Cf. French je me trouve and the use of reflexive pronouns in English.

(b) SIGNIFICANCE OF THE PASSIVE. The subject is represented as the recipient of the action. He is acted upon. The name "passive" comes from patior (cf. pa,scw u`po, in Mt. 17:12). vApoktanqh/nai (Mk. 9:31) occurs as well as avpoqnh,skein. The use of peri,keimai as the transitive passive (Ac. 28:20) of periti,qhmi is somewhat different. The idea of having an experience is very vague and allows wide liberty. The point to note is that at first this idea had no distinctive form for its expression. Only the context and the force of the verb itself could make it clear. The future passive, being built upon the earlier aorist passive, reflects the Aktionsart of the aorist.⁸⁷

(C) WITH INTRANSITIVE OR TRANSITIVE VERBS. "Theoretically the passive ought to be formed from transitive verbs only with an accusative object."⁸⁸ But Greek follows no such narrow rule. That is an artificial rule of the Latin which Greek knows nothing about.⁸⁹ Cf. kathgorei/tai u`po. tw/n vIoudai,wn (Ac. 22:30). Other N. T. examples are diakonhqh/nai (Mk. 10:45), evgkalei/sqai (Ac.

816 A GRAMMAR OF THE GREEK NEW TESTAMENT

19:40), euvarestei/sqai, (Heb. 13:16), kategnwsme,noj (Gal. 2:11), marturei/sqai (Ac. 6:3), crhmati,zesqai (Mt. 2:12). Blass (Gr. of N. T. Gk., p. 185) notes that "only in Lu. 2:26 do we have h=n auvtw|/ kecrhmatisme,non." The passive is used with both active and middle verbs. Thus we have from logi,zomai both evlogisa,mnh and evlogi,sqhn. Cf. egveno,mhn and evgenh,qnh from gi,nomai.

(d) THE PASSIVE USUALLY INTRANSITIVE. But it is not necessarily so. Dida,skw, for instance, is transitive in the passive, as evdida,cqhte (2 Th. 2:15), and note kathchme,noj thn o`do,n (Ac. 18:25). See also 1 Cor. 9:17; Lu. 7:25; 9:25; Gal. 2:7. Transitive passives are usually verbs that in the active have two accusatives or an accusative of the thing with the person in the dative or ablative. This accusative of the thing is retained in the passive. Cf. evpisteu,qhsan ta. lo,gia tou/ qeou/ (Ro. 3:2), peribeblhme,nouj stola.j leuka,j (Rev. 7:9). For full list see "Accusative" in chapter XI, Cases. Cf. also th.n a[lusin tau,thn peri,keimai (Ac. 28:20). The transitive passive " deponents," like mh. fobhqh/te auvtou,j (Mt. 10: 26), call for special discussion a little later. Certainly there is no "passive" sense in poreuqh/nai. The vernacular<sup>90</sup> in later times preferred the active to passive. Cf. aivtou/sin (Lu. 12:20) as a N. T. illustration. In a`gni,sqhti, (Ac. 21:24) the passive apparently has the force of 'let' or 'get' (cf. the causative middle). Cf. also perite,mnhsqe (Gal. 5:2).⁹¹ It is possible so to regard avdikei/sqe and avposterei/sqe (1 Cor. 6:6 f.). Sometimes, indeed, it is difficult to tell whether a verb is middle or passive. Cf. ptwcoi. euvaggeÄ li,zontai (Mt. 11:5), proeco,meqa (Ro. 3:9), evndunamou/sqe (Eph. 6: 10). Indeed, as already said, in all the Greek tenses save the aorist and the future it is always an open question whether we have middle or passive. "The dividing-line is a fine one at best" (Moulton, Prol., p. 162). Only the context and the verb-idea can decide. So with evgei,romai (Mt. 27:63), periespa/to (Lu. 10:40) and qoruba,zh|grk grk(10:41), bia,zetai (Mt. 11:12). Cf. perfects in Ac. 13:2; 25:12; Ro. 4:21; 1 Pet. 4:1; Jo. 9:22.

(e) AORIST PASSIVE. This tense calls for special comment. As already stated, in Homer the aorist middle form, like the other middle forms, was sometimes used as passive.⁹² In itself there is no reason why this should not be so. The distinctive passive aorist (second and first) grew up side by side with this use of the aorist middle. vEfa,nhn and e;bhn are really the same form at

VOICE ( DIAQESIS) 817

bottom.⁹³ Out of this intransitive aorist active (cf. avpo,lwla) grew the so-called second aorist passive forms (- hn) with active endings. We have evku,bhn (Jo. 8:59) from the transitive kru,ptw) (cf. evsta,lhn from ste,llw, etc.) and evca,rhn (Jo. 14:28) from the intransitive cai,rw. It is probable that hvge,rqh sometimes (as in Mk. 16:6) is merely intransitive, not passive, in idea. Moulton (Prol., p. 163) says "often." In 1 Cor. 15:15 f., etc., the true passive "emphasizes the action of God." But u`peta,ghsan (Ro. 10:3) is more likely passive in sense, like evkoimh,qhn (1 Th. 4:14), 'was put to sleep' (Moulton, Prol., p. 162). Moulton quotes from the papyri "a purely middle use of koimhqh/nai, 'fell asleep'," h`ni,ka h;mellon koimhqh/nai e;graya, Ch.P. 3 (iii/B.C.). He finds a "clear passive" in i[na ta. pro,bata evkei/ koimhqh/i, F.P. 110 (i/A.D.), but evkollh,qh (Lu. 15:15) can be explained as passive or middle in sense. In a few verbs ( e;sthn evsta,qhn) a distinction was developed.⁹⁴ W. F. Moulton thinks (Winer-M., p. 315, n. 5) that "a faint passive force" may be observed in staqh/nai in the N. T., but hardly in Mk. 3:24. Cf. also intransitive staqh,somai in Mt. 12:25, 26. vEsta,qhka in modern Greek is aorist passive for ste,kw, 'stand,' and evsth,qhka for sth,nw, 'place' (Thumb, Handb., p. 145). The correct text (W. H.) in Ac. 21:3 is avnafa,nantej th.n Ku,pron (active), not avnafane,ntej (passive). But still some MSS. do have this transitive second aorist passive participle. If one keeps in mind the origin of this aorist passive form (from the active), he may be the less surprised to find it also transitive like the active. Already in Homer this was true.

The so-called passive "deponents," verbs which had no active, formed the aorist with the passive form. But they were not always intransitive. Some of them were so, like poreu,omai (Mt. 8:9), metame,lomai, (Mt. 27:3), du,namai (Mt. 17:16), but most of them are really transitive. They probably represent a survival of the old active origin of the aorist passive forms.⁹⁵ As examples of the transitive passive deponents note evboulh,qh (Mt. 1:19), evdeh,qh (Lu. 5:12), evnqumhqe,ntoj (Mt. 1:20), evpemelh,qh (Lu. 10:34), evfobh,qh (Mt. 14:5). These passive aorists have precisely the construction that the middle or active would have so far as case is concerned. The distinctive passive sense is absent. Some of the "deponents" have both a middle and a passive aorist with a distinct passive sense. Thus note the middle and passive voices side

818 A GRAMMAR OF THE GREEK NEW TESTAMENT

by side in avrnhsa,menoj and avparnhqh,setai (Lu. 12:9). It so happens that this context is full of passive forms. Some of them in the strict passive sense, like evpisunacqeisw/n grk(12:1), sugkekalumme,non evsti.n oa} ouvk avpokalufqh,setaigrk grk(12:2), gnwqh,setaigrk grk(12:2), avkousqh,setai and khrucqh,setaigrk grk(12:3), pwlou/ntai and ouvk e;stin evpilelhsme,nongrk grk(12:6), hvri,qmhntaigrk grk(12:7), avfeqh,setaigrk grk(12:10). But note also the passive deponents fobhqh/te grk(12:4 f.), fobh,qhtegrk grk(12:5), fobei/sqegrk grk(12:7). Cf. also avpode,xasqai (Ac. 18:27) and parede,cqhsangrk grk(15:4), where the voices are distinguished, qea,sasqai tou.j avnakeime,nouj (Mt. 22:11) and pro.j to. qeaqh/nai auvtoi/j (Mt. 6:1), logisa,menoj (Heb. 11:19) and evlogi,sqh (Lu. 22:37), iva,sato (Lu. 9:42) and iva,qh (Mt. 8 : 13), evru,sato (Col. 1:13) and evru,sqhn (2 Tim. 4:17), evcari,sato (Lu. 7:21) and carisqh/nai (Ac. 3:14). One may note also parh|th,santo (Heb. 12:19) and e;ce me parh|thme,non (Lu. 14:19, perfect passive); evxele,xato (Mk. 13:20), but o` evklelegme,noj (Lu. 9:35); koresqe,ntej trofh/j (Ac. 27:38) and h;dh kekoresme,noi evste, (1 Cor. 4:8). It is possible to see a difference also between evge,neto (Jo. 1:14) and genhqh,tw (Mt. 6:10). vApekri,qhn (Mt. 25:9) steadily drove out avpekri,nato (Ac. 3:12), though both are used transitively with no difference in sense. The papyri more frequently<sup>96</sup> have avpekrina,mhn, though both forms continue in the koinh,) Cf. also avpologhqh/nai (Lu. 21:14), diele,cqhsan (Mk. 9:34), evqauma,sqh (Rev. 13:3), though with passive sense in 2 Th. 1:10. As a result of this inroad of the comparatively new passive forms the aorist middle forms vanished. In modern Greek the passive aorist form is almost invariably used for both the middle and the passive ideas. This tendency seen in the N. T. (and the rest of the koinh,) has triumphed over the aorist middle.<sup>97</sup> In Ro. 10:3, th|/ dikaiosu,nh| tou/ qeou/ ouvc u`peta,ghsan, the Rev. V. translates `they did not subject themselves to the righteousness of God.'

(f) FUTURE PASSIVE. As has been mentioned several times already, Homer has only two future passive forms (second futures). The passive voice indeed occurs but rarely in the Boeotian dialect.⁹⁸ The future in - qh,somai is comparatively late. At first, certainly, the distinction between passive and middle (and active also, -- hn, - qhn) was "a distinction of function, not of form."⁹⁹ It is not surprising to find the middle future form in Homer used with the passive sense (cf. all the other tenses save aorist), where the forms

VOICE ( DIAQESIS) 819

for the two voices are identical. In later prose the future middle form continued to be used in the passive sense even in the great prose writers (Herodotus, Thucydides, Xenophon, Plato, Demosthenes).¹⁰⁰ In the LXX Conybeare and Stock (Selections, p. 75 f.) find the same idiom. Cf. Ex. 12:10, ouvk avpolei,yetai avp v auvtou/ e[wj prwi,Ã kai. ovstou/n ouv suntri,yetai avp v auvtou/) It is quite within bounds, therefore, to speak of "medio-passives" in the future as in the aorist.¹⁰¹ The idiom appears in the papyri.¹⁰² So narrow is the dividing-line between middle and passive. Is peribalei/tai (Rev. 3:5) middle or passive in sense? The same ambiguity exists as to avpoko,yontai (Gal. 5:12). Considering the rather large list of verbs¹⁰³ that once used the middle future as passive in sense the idiom is rare in the N. T. In general, therefore, the future passive form has made its place secure by the time of the koinh,. Even verbs that have no active form have the future passive as well as the future middle. Thus avparnh,somai. (Mk. 14:31), but avparnhqh,somai. (Lu. 12:9); iva,somai (Ac. 28:27), but ivaqh,setai (Mt. 8:8); and in Ro. 2:26 logisqh,setai is passive in sense. But the future passive form was destined, like the other futures, to disappear as a distinct form. Only the compound tense occurs in the modern Greek.¹⁰⁴ But, meanwhile, the future passive form took over the uses of the vanishing future middle forms.¹⁰⁵ It is possible to find a passive sense in evpanapah,setai. (Lu. 10:6), metamelhqh,setai (Heb. 7:21), avnakliqh,sontai (Mt. 8:11), koimhqhso,meqa (1 Cor. 15:51), kollhqh,Ä setai. (Mt. 19:5). Cf. also qaumasqh,sontai (Rev. 17:8), peisqh,sontai (Lu. 16:31), fanh,setai (Mt. 24:30), u`potagh,setai (1 Cor. 15:28).<sup>106</sup> In 1 Cor. 15:28 note also u`potagh|/, which reinforces the argument for the true passive. But the future passive may also be devoid of the passive idea and even transitive just like the aorist passive. Cf. avpokriqh,somai (Mt. 25:37), evntraph,sontai to.n ui`o,n (Mt. 21:37), fobhqh,somai (Heb. 13:6). The passive avfaireqh,setai (Lu. 10:42) has the usual sense, but one wonders if in w-n te ovfqh,somai, soi (Ac. 26:16) the passive voice is transitive and even causative (cf. Is. 1:12). Cf. the examples of reflexive passives in the LXX (Conybeare and Stock, Sel., p. 76), like o;fqhti= 'show thyself' (1

820 A GRAMMAR OF THE GREEK NEW TESTAMENT

Ki. 18:1). It is possible, of course, for w- to be attracted to the case of tou,twn from oi-j ('in which,' 'wherein'). Then ovfqh,somai, soi would be 'I will appear to thee.' Note the new present ovpta,Ä nomai (Ac. 1:3). But the future middle persisted in genh,somai, dunh,somaià evpimelh,somaià poreu,somai.

(g) THE AGENT WITH THE PASSIVE VOICE. As already noted, the Greek has no difficulty in using a verb in the passive which was not used with the accusative in the active. Thus note evgkalei/Ä sqai (Ac. 19:40), kathgorei/tai u`po. tw/n vIoudai,wn (Ac. 22:30), pepi,Ä steumai to. euvagge,lion (Gal. 2:7).<sup>107</sup> A few verbs idiomatically use the dative with the passive. Thus evgnw,sqh tw|/ Sau,lw| (Ac. 9:24), eu`re,qhn (Ro. 10:20), evfa,nh (Mt. 1:20), w;fqh (1 Cor. 15:7 f.), qeaqh/nai (Mt. 6:1).¹⁰⁸ The direct agent is most commonly expressed by u`po, (Mt. 4:1), the intermediate by dia, (Mt. 1:22). The agent (see chapter on Prepositions) is also expressed by avpo, (2 Cor. 3:18), e`k (Gal. 4: 4), para, (Jo. 17:7). See also discussion under Instrumental Case (chapter XI, Cases) for discussion of auvtw|/ with evsti.n pepragme,non (Lu. 23:15), whether dative or instrumental. In the N. T., as in ancient Greek (Gildersleeve, Syntax, p. 72), the instrument is sometimes personified and treated as an agent. Cf. ka,lamon u`po. avne,mou saleuo,mon; (Mt. 11:7).

(h) IMPERSONAL CONSTRUCTION. This is the usual idiom in the Coptic in lieu of the absence of the passive. But it is often rather rhetorical than syntactical as Moulton shows.¹⁰⁹ He compares also the French on, the German man, the English one. Wellhausen¹¹⁰ shows how in the Aramaic this impersonal plural was common. One notes aivtou/sin (Lu. 12:20), where a passive would be possible. Cf. suna,gousin kai. ba,llousin kai. kai,etai (Jo. 15: 6) where the passive occurs in kai,etai. Note in particular evxhra,nqh kai. suna,gousin auvta, (Jo. 15:6). Cf. also tre,fwsin auvth,n (Rev. 12:6). The use of the impersonal passive like pisteu,etai and o`mologei/tai (Ro. 10:10) is another matter and calls for no comment. It is rare in Greek as compared with Latin (Gildersleeve, Syntax, p. 77). Cf. the plural in 10:14 f. See also the personal construction in 1 Cor. 15:12 eiv de. Cristo.j khru,ssetai o[ti.

¹ Am. Jour. of Philol., 1908, p. 279.

² Bd. I, p. 89.

³ Jann., Hist. Gk. Gr., p. 357.

⁴ Griech. Gr., p. 467.

⁵ Jebb.,V. and D.'s Handb., p. 318.

⁶ Thumb., Handb., p. 112 f.

⁷ Krit.-hist. Synt. d. griech. Verbums, p. 42.

⁸ Thompson, Synt., p. 158.

⁹ Notes on Stahl's Synt. of the Gk. Verb in Am. Jour. of Philol., 1908, p. 275.

¹⁰ Riem. and Goelzer, Synt., p. 233.

¹¹ Whitney, Sans. Gr., p. 201.

¹² Jann., Hist. Gk. Gr., p. 362.

¹³ Giles, Comp. Philol., p. 404 f.

¹⁴ Whitney, Sans. Gr., p. 200.

¹⁵ Griech. Gr., pp. 459ff. Cf. Thompson, Synt., p. 159.

¹⁶ Cf. Jann., Hist. Gk. Gr., p. 357.

¹⁷ Prol., p. 154.

¹⁸ Cf. Jann., list. Gk. Gr., p. 359.

¹⁹ Gildersleeve, Synt. of Cl. Gk., p. 63.

²⁰ C. and S., Sel., p. 76.

²¹ Thack., Gr. of the 0. T. in Gk., p. 24.

²² Gr. of N. T. Gk., p. 183.

²³ Cf. Gildersleeve, Synt., p. 63.

²⁴ Thompson, Synt., p. 172.

²⁵ Moulton, Prol., p. 152.

²⁶ Whitney, Sans. Gr., p. 200.

²⁷ Monro, Hom. Gr., p. 7.

²⁸ Cf. 0. Hoffmann, Das Prasens der indoger. Grundspr., 1889, p. 25. In the Bantu language Mr. Dan Crawford finds 16 voices (reflexive, reciprocal, intensive, etc., all having special forms).

²⁹ Giles, Comp. Philol., p. 406.

³⁰ Ib p. 405.

³¹ Hist. Gk. Gr., p. 362 f.

³² Clyde, Gk. Synt., p. 57.

³³ Ewald, Heb. Gr., § 243.

³⁴ Am. Jour. of Philol., 1908, p. 277.

³⁵ Viteau, Essai sur la Synt. des Voix, p. 17. Cf. Moulton, Prol., p. 153.

³⁶ Cf. Jann., Hist. Gk. Gr., p. 360; Clyde Gk. Synt., p. 58 f.; Farrar, Gr. Synt., p. 117 f.; Thompson, Synt., pp. 168 ff.

³⁷ Gildersleeve, Synt. of Class. Gk., p. 66.

³⁸ Blass, Gr. of N. T. Gk., p. 185.

³⁹ W.-Th., p. 256.

⁴⁰ Gr. of N. T. Gk., p. 186.

⁴¹ Prol., p. 160.

⁴² Prol., p. 158 f. He cites also suna/rai le,gon, B.U. 775 (ii/A.D.). But the pap. use the middle also.

⁴³ Cf. Farrar, Gk. Synt., p. 117; Brug., Griech. Gr., pp. 459 ff.; K.-G., Bd. I, pp. 100 ff.; Stahl, Krit.-hist. Synt., pp. 49 ff.

⁴⁴ Am. Jour. of Philol., 1908, p. 278.

⁴⁵ Monro, Hom. Gr., p. 7.

⁴⁶ Jann., Hist. Gk. Gr., p. 360..

⁴⁷ Ib.

⁴⁸ Moulton, Prol., p. 156.

⁴⁹ Ib.

⁵⁰ W.-Th., p. 253.,

⁵¹ Prol., p. 156.

⁵² Gildersleeve, Synt. of Class. Gr., p. 67.

⁵³ Thompson, Synt. of Att. Gk., p. 162.

⁵⁴ Jann., Hist. Gk. Gr., p. 361.

⁵⁵ Moulton, Prol., p. 162.

⁵⁶ Gr. of N. T. Gk., p. 187.

⁵⁷ Ib., p. 185.

⁵⁸ Thompson, Synt., p. 162.

⁵⁹ Jann., Hist. Gk. Gr., pp. 360, 362.

⁶⁰ Moulton, Prol., p. 157.

⁶¹ Moulton, Prol., p. 157.

⁶² Cf. Jann., Hist. Gk. Gr., p. 361.

⁶³ Synt. of Att. Gk., p. 166.

⁶⁴ Gildersleeve, Am. Jour. of Philol., 1908, p. 277.

⁶⁵ Synt., p. 161.

⁶⁶ Prol., p. 153.

⁶⁷ Thompson, Synt., p. 161.

⁶⁸ W.-Th., p. 258.

⁶⁹ Am. Jour. of Philol., 1908, p. 278.

⁷⁰ Moulton, Prol., p. 159,

⁷¹ Monro, Hom. Gr., p. 7. So the other poets. Thompson, Synt., p. 165.

⁷² Synt., p. 165.

⁷³ Gk. Synt., p. 60.

⁷⁴ Moulton, Prol., p. 154.

⁷⁵ Moulton, Prol., p. 155. Cf. Monro, Hom. Gr., p. 42.

⁷⁶ Jann., Hist. Gk. Gr., p. 363 f.

⁷⁷ Hatz., Einl., pp. 194 ff. Cf. Thumb, Hellen., p. 127.

⁷⁸ Moulton, Prol., p. 158 f.

⁷⁹ Ib., p. 159.

⁸⁰ Lang. of the N. T., p. 95.

⁸¹ Cf. K.-G., Bd. I, pp. 121 ff.

⁸² Whitney, Sans. Gr., pp. 201, 275.

⁸³ Thompson, Synt., p. 162.

⁸⁴ Cf. Brug., Griech. Gr., p. 464.

⁸⁵ Gildersleeve, Am. Jour. of Philol., 1908, p. 278.

⁸⁶ Sterrett, The Dial. of Hom., N. 27.

⁸⁷ Brug., Griech. Gr., p. 464.

⁸⁸ Gildersleeve, Am. Jour, of Philol., 1908, p. 279.

⁸⁹ Cf. Jann., Hist. Gk. Gr., p. 359.

⁹⁰ Jann., Hist. Gk. Gr., p. 359.

⁹¹ Cf. Blass, Gr. of N. T. Gk., p. 185.

⁹² Seymour, The Hom. Dial., p. 74. Cf. Brug., Griech. Gr., p. 464.

⁹³ Giles, Comp. Philol., p: 410; Brug., Griech. Gr., p. 465.

⁹⁴ Blass, Gr. of N. T. Gk., p. 181.

⁹⁵ See ch. VIII, vt, (e), for list of these N. T. passive aorists.

⁹⁶ Moulton, Prol., p. 161.

⁹⁷ Cf. Jannaris, Hist. Gk. Gr., p. 362; Hatz., Einl., pp. 196 ff.; Jebb in V. and D.'s Handb., p. 315.

⁹⁸ Claflin, Synt. of the Boeot. p. 67.

⁹⁹ Gildersleeve, Synt. of Class. Gk., p. 61.

¹⁰⁰ Gildersleeve, ib., p. 73 f. Cf. Hartel, Abriss der Gr. d. hom. and herod. Dial., 1888, p. 40.

¹⁰¹ Brug., Griech. Gr., p. 463 f.

¹⁰² Moulton, Prol., p. 162.

¹⁰³ Clyde, Gk. Synt., p. 61; Thompson, Synt. of Att. Gk., p. 171.

¹⁰⁴ Cf. Thumb, Handb., pp. 115, 125.

¹⁰⁵ Jann., Hist. Gk. Gr., p. 363.

¹⁰⁶ Moulton, Prol., p. 163. Cf., for the LXX, Helbing, Gr., p. 98.

¹⁰⁷ Cf. Gildersleeve, Synt., etc., p. 77.

¹⁰⁸ Cf. Blass, Gr. of N. T. Gk., p. 185.

¹⁰⁹ Prol., p. 58 f.

¹¹⁰ Einl., p. 25 f.