I. Difficulty of the Subject. The discussion of the verb gives greater difficulty than that of the noun for two reasons especially. For one thing the declension ( kli,sij) of nouns is more stable than the conjugation ( suzugi,a) of the verb. This difficulty applies to both the forms and the syntax of the verb.1 There is besides special difficulty in the Greek verb due to the ease and number of new verbal formations.2 Sanskrit and Greek can be compared with more ease than Greek and Latin. Giles3 indeed calls the Latin verb-system "only a mutilated fragment" of the original parent stock, so that "a curious medley of forms" is the result, while in the syntax of the verb no two Indo-Germanic languages are further apart than Greek and Latin. Both noun and verb have suffered greatly in the ravages of time in inflection. It is in declension (cases) and conjugation (personal endings) that noun and verb mainly differ.4 "These suffixes [used for the present tense], however, are exactly parallel to the suffixes in the substantive, and in many instances can be identified with them."5

II. Nature of the Verb.

(a) VERB AND NOUN. In itself verbum is merely 'word,' any word, and so includes noun also. As a matter of fact that was probably true originally. In isolating languages only position and the context can determine a verb from a noun, and that is often true in English to-day. But in inflected tongues the case-endings and the personal endings mark off noun and verb. But in simple truth we do not know which is actually older, noun or verb; both probably grew up together from the same or similar roots.6 Schoemann,7 however, is much more positive that "the first word


Addenda 2nd ed.

which man spoke was essentially much more a verb than a noun." But, whether the verb is the first word or not, it is undoubtedly the main one and often in the inflected tongue forms a sentence in itself, since the stem expresses the predicate and the ending the subject.8 It is worth noting also that by the verb-root and the pronominal root (personal endings) the verb unites the two ultimate parts of speech. The verb and noun suffixes, as already said, are often identical (Giles, Manual, etc., p. 424). In all sentences the verb is the main part of speech (the word par excellence) save in the copula ( evsti,% where the predicate is completed by substantive or adjective or adverb (another link between verb and noun). "A noun is a word that designates and a verb a word that asserts" (Whitney, Am. Jour. of Philol., xiii, p. 275). A man who does not see that "has no real bottom to his grammatical science."

(b) MEANING OF THE VERB. Scholars have found much difficulty in defining the verb as distinct from the noun. Indeed there is no inherent difference between nouns and verbs as to action, since both may express that.9 The chief difference lies in the idea of affirmation. The verb affirms, a thing not done by a noun except by suggested predication. Verbs indicate affirmation by the personal endings. Affirmation includes negative assertions also.10 Farrar11 cites also the German "abstract conception of existence" (Humboldt) and action (Ttigkeitswort), but they do not fit the facts. Curiously enough many ancient grammarians found time to be the main idea in the verb.

(c) PURE AND HYBRID VERBS. The close kinship between nouns and verbs appears in the verbal nouns which partake of both. The infinitive is a verbal substantive, and the participle is a verbal adjective. There is also the verbal in - toj and - te,oj Some of the properties of both verb and noun belong to each. They are thus hybrids. They are generally called non-finite


verbs, because they do not make affirmation. They have no personal endings. They fall short of being mere verbs, but they are more than the noun. The pure verb has personal endings and is thus finite (limited). The two must be kept distinct in mind, though they run together sometimes in treatment. The finite verb has person and number expressed in the personal ending.12 The verbum finitum has modes while the verbum infinitum, (infinitive and participle) has no modes.

III. The Building of the Verb. This is not the place for a full presentation of the phenomena concerning verb-structure. The essential facts as to paradigms must be assumed. But attention can be called to the fact that the Greek verb is built up by means of suffixes and affixes around the verb-root. So it was originally, and a number of such examples survive. Afterwards analogy, of course, played the main part. The oldest verbs are those which have the simple root without a thematic vowel like fhmi, or e;bhn. This root is the ground floor, so to speak, of the Greek verb. On this root the aorist and present-tense systems were built by merely adding the personal endings. This was the simplest form of the verb. There is no essential difference in form between e;fhn and e;sthn. We call one imperfect indicative and the other second aorist indicative, but they are originally the same form.13 The term second aorist is itself a misnomer, for it is older than the so-called first aorist - sa or - a. The thematic stem (vowel added to root) is seen in verbs like - lipo/ e. On this model the rest of the verb is built. So all Greek root-verbs are either nonthematic or thematic. The denominative verbs like tima,w are all thematic. On roots or stems then all the verbs (simple or compound) are built. The modes, the voices, the tenses all contribute their special part to the whole. The personal endings have to carry a heavy burden. They express not only person and number, but also voice. There are mode-signs and tense-suffixes, but no separate voice suffixes apart from the personal endings. The personal pronouns thus used with the verb-root antedate the mode and tense suffixes. The Sanskrit preserves the person-endings more clearly than the Greek, though the Greek has a more fully developed system of modes and tenses than the later classical Sanskrit.14 It seems certain that these pro-


Addenda 2nd ed.

nominal suffixes, like -- mi, - si, -- ti are not in the nominative, but an oblique case15 connected with the stem: me se ti (cf. demonstrative to,). But the subject of personal endings is a very extensive and obscure one, for treatment of which see the comparative grammars.16 There is a constant tendency to syncretism in the use of these personal endings. Homer has fewer than the Sanskrit, but more than Plato. The dual is gone in the N. T. and other endings drop away gradually. The nominative pronoun has to be expressed more and more, like modern English.

IV. The Survival of mi Verbs.

(a) A CROSS DIVISION. Before we take up modes, voices, tenses, we are confronted with a double method of inflection that cuts across the modes, voices and tenses. One is called the - mi, inflection from the immediate attachment of the personal endings to the stem. The other is the - w inflection and has the thematic vowel added to the stem. But the difference of inflection is not general throughout any verb, only in the second aorist and the present-tense systems (and a few second perfects), and even so the -- mi conjugation is confined to four very common verbs ( i[hmi, i;sthmi, di,dwmi, ti,qhmi), except that a number have it either in the present system, like di,knumi. (with nu inserted here), or the aorist, like e;bhn.17 The dialects differed much in the use of non-thematic and thematic verbs (cf. Buck, "The Interrelations of the Greek Dialects," Classical Philology, July, 1907, p. 724).

(b) THE OLDEST VERBS. This fact is a commonplace in Greek grammar. It is probable that originally all verbs were - mi verbs. This inflection is preserved in optative forms like lu,oimi, and in Homer the subjunctive18 evqe,lwmi i;dwmi, etc. The simplest roots with the most elementary ideas have the - mi form.19 Hence the conclusion is obvious that the - mi conjugation that survives in some verbs in the second aorist and present systems (one or both) is the original. It was in the beginning le,gomi with thematic as well as fhmi, with non-thematic verbs.20

(c) GRADUAL DISAPPEARANCE. In Latin the - mi, ending is seen only in inquam and sum, though Latin has many athematic stems. In English we see it in am. Even in Homer the - mi


forms are vanishing before the - w conjugation. Jannaris (Hist. Gk. Gr., p. 234) has an excellent brief sketch of the gradual vanishing of the - mi forms which flourished chiefly in pre-Attic Greek. The LXX MSS. show the same tendency towards the disappearance of - mi forms so noticeable in the N. T., the papyri and other representatives of the koinh,. See numerous parallel illustrations in Helbing, Gr. d. Sept., pp. 104-110. In the LXX the transition to - w verbs is less advanced than in the N. T. (Thackeray, Gr., p. 244) and the middle - mi forms held on longest. In the koinh, this process kept on till in modern Greek vernacular ei=mai is the only remnant left. In the Attic dei,knumi, for instance, is side by side with deiknu,w. In the N. T. we find such forms as didw/ (Rev. 3:9), i`stw/ (Ro. 3:31, EKL), sunistw/ (2 Cor. 3:1, BD).

(d) N. T. USAGE AS TO mi VERBS. The - mi verbs in the N. T. as in the papyri are badly broken, but still in use.

1. The Second Aorists (active and middle). We take first the so-called second aorists (athematic) because they come first save where the present is practically identical. In some verbs only the second aorist is athematic, the stem of the verb having dropped the - mi inflection. A new view21 makes the second aorist sometimes "a reduced root," but this does not show that in the parent stock the old aorist was not the mere root. Analogy worked here as elsewhere. Kaegi22 properly calls the old aorists of verbs like ba,llw $e;blhto instead of the thematic and later evba,leto) "primitive aorists." In the early Epic the root-aorists and strong thematic aorists outnumber the s or weak aorists by three to one.23 The important N. T. - mi verbs will now be considered.

Bai,nw. Only in composition in N. T. ( avna prosana sun ana avpo dia evk evm kata meta para pro sum%) In the LXX it is rare in simplex. The papyri use it freely with nine prepositions.24 Note the common forms like avne,bh (Mt. 5:1) . The "contract" forms are in the imperative as in the Attic poets ( ei;sba kata,ba).25 Mayser26 gives no examples from the papyri, nor does the LXX have any (LXX only avna,bhqi kata,bhqi bhtebh,tw bh,twsan).27 So avna,ba (Rev. 4:1), avna,bate (Rev. 11:12), kata,ba (Syrian class in Mk. 15:30), kataba,tw (Mt. 24:17; 27:42. Cf.


Addenda 3rd ed.

also Mk. 13:15; 15:32; Lu. 17:31), meta,ba (Mt. 17:20). On the other hand note the usual kata,bhqi (Mt. 27:40, etc.), meta, bhqi (Jo. 7:3), prosana,bhqi (Lu. 14:10). The forms in - a,tw, -- ate, -- a,twsan are like the Doric.

Ginw,skw. This verb in the Ionic and koinh, gin. form is very common in John's Gospel and the First Epistle. It is used in composition with avna dia evpi kata pro, the papyri adding still other compounds.28 The N. T. shows the usual second aorist forms like e;gnwn (Lu. 16:4). What calls for remark is the second aorist subjunctive gnoi/ instead of gnw|/. W. F. Moulton's view29 on this point is confirmed by the papyri30 parallel in avpodoi/ and accepted by W. H. and Nestle. Analogy seems to have worked here to make gnoi/ like doi/. But Winer-Schmiedel (p. 115) cite gnoi/ from Hernias, Mand. IV, 1, 5 a. It is in accordance with the contraction of - ow verbs when we find forms like gnoi/ doi/, etc., o,h|= oi/ instead of o,h|- w|/. For gnoi/ see Mk. 5:43; 9:30; Lu. 19:15. But see also gnw|/ in Jo. 7:51; 11:57 (D has gnoi/); 14:31; Ac. 22:24 ( evpi). But the MSS. vary in each passage. In the LXX the regular gnw|/ occurs save in Judith 14:5, where B has evpignoi/)

Di,dwmi. This very common verb is frequently compounded $avna avnt avpo dia evk evpi meta para pro% as in the papyri.31 The old indicative active appears only in paredosan in the literary preface to Luke's Gospel Gospel(1:2).32 Elsewhere the first aorist forms in -- ka (like h-ka e;qhka) sweep the field for both singular and plural. These k forms for the plural appear in the Attic inscriptions in the fourth century B.C.33 and rapidly grow. In the papyri Mayser34 finds only the k aorists. The other modes go regularly do,j, dw/, etc. The indicative middle occasionally, as the imperfect, has e for o of the root. This is possibly due to proportional analogy ( evxe,deto: evxedo,mhnevlu,eto: evluo,mhn).35 These forms are avpe,deto (Heb. 12:16), evxe,deto (Mk. 12:1; Mt. 21:33; Lu. 20:9). The usual form avpe,dosqe, etc., appears in Ac. 5:8; 23) 7:9. The subjunctive active third singular shows great variation between doi/ dw|/ (cf. gnoi/ above), and dw,h| (especially in Paul's Epistles).36 The LXX MSS. occasionally give -- doi/ and


Addenda 3rd ed.

even dh|/ by assimilation (Thackeray, Gr., p. 255 f.). For papyri examples see references under ginw,skw. Mark four times (all the examples) has also doi/ according to the best MSS. MSS.(4:29; 8:37; 14:10 f.) and John one out of three three(13:2). Tisch. (not W. H.) reads avpodoi/ in 1 Th. 5:15, but all MSS. have avpodw|/ in Mt. 18:30. W. H. accept dw|/ in Jo. 15:16; Eph. 3:16; 1 Th. 5: 15 ( avpo). Most MSS. read dw,h| in Eph. 1:17 and 2 Tim. 2: 25, in both of which places W. H. put dw|,h (opt. for doi,h) in the text and dw,h| in the margin. The opt. dw|,h appears in the LXX (Jer. 9:2) in the text of Swete. Con. and Stock, Sel. from LXX, p. 45, give dw|,h twenty-nine times in LXX and doi,h three times as variant. They give an interesting list of other forms of di,dwmi and its compounds in the LXX. Hort37 is doubtful about such a subjunctive in dw,h| except in the epic poets. Blass38 is willing to take dw,h|, and Moulton39 cites Boeotian and Delphian inscriptions which preserve this Homeric form. He adds that the subjunctive seems "a syntactical necessity" in Eph. 1:17 and 2 Tim. 2:25. The opt. dw|,h= doi,h (cf. subjunctive do,h|= dw|/) is without variant in 2 Th. 3:16; 2 Tim. 1:16, 18.40 Blass41 scouts the idea of a possible first aorist active e;dwsa from i[na dw,sh| (Jo. 17:2 acAC), dw,swmen (Mk. 6:37, aD), on the ground that h| and ei o and w so often blend in sound in the koinh,. The so-called future subjunctive will be discussed later (ch. XIX).

[Ihmi) Not in simplex in N. T. (see p. 314 for details), but avfi,hmi is quite common (especially in the Gospels), and suni,hmi less so. Besides a few examples occur also of avni,hmi kaqi,hmi pari,hmi. The papyri42 use the various prepositions freely in composition with i;hmi. The common mi second aorists, like a;fej (Mt. 3:15), avfh|/ (Mk. 12:19), avne,ntej (Ac. 27:40), are found. In the indicative active, however, the form in - ka is used alone in both singular and plural, as avfh,kamen (Mt. 19:27), avfh,kate (Mt. 23: 23), avfh/kan (Mk. 11:6). This is true of all the compounds of i[hmi in the N. T. as in LXX (Thackeray, Gr., p. 252). The form avfh/kej (Rev. 2:4) is on a par with the second person singular perfect active indicative as accepted by W. H. in kekopi,akej (Rev. 2:3), pe,ptwkej (Rev. 2:5), ei;lhfej (Rev. 11:17).43 vAfh,kaemn is aorist in Mk. 10:28 as well as in its parallel Mt. 19:27


( = Lu. 18 : 28). So also as to sunh,kate in Mt. 13:51. The perfect in - ei/ka does not, however, occur in the N. T. nor in the LXX (cf. Blass, Gr. of N. T. Gk., p. 51), though the papyri have it (Mayser, Gr. d. griech. Pap., p. 331).

[Isthmi. This verb is used freely by itself, especially in the Gospels, and occurs in twenty prepositional combinations according to Thayer ( avn evpan evxan avnq avf di evn evx evp evf katef sunef kaq avntikaq avpoaq meq par peri pro sun), going quite beyond the papyri in richness of expression.44 The second aorist active indicative e;sth ( avpe,sth, etc.) is common and is intransitive as in Attic, just like evsta,qh (cf. Blass, Gr. of N. T. Gk., p. 50). The other forms are regular ( stw/ sth/qi, etc.) save that avna,sta(like avna,ba% is read in a few places (Ac. 9:11; 12:7; Eph. 5:14), but sth/qi avna,sthqi (Ac. 9:6, 34), evpi,sthqi sth/te a`nti,sthte avpo,sthte avposth,tw.45 Winer46 cites avpo,sta para,sta also from late writers and a few earlier authors for avna,sta. The LXX shows a few examples also.47

vOni,nhmi. This classic word (not given in the papyri, according to Mayser's Grammatik) is found only once in the N. T., the second aorist opt. middle ovnai,mhn (Phil. 1:20).

Ti,qhmi) The compounds of ti,qhmi in the N. T. ( avna prosana avpo dia avntidia evk evpi sunepi kata sunkata meta pa ra peri pro pros sun u`po-) vie with those of i[sthmi and equal the papyri use.48 The first aorist active in - ka alone appears (so LXX) in the indicative singular and plural as e;qhkan (Mk. 6:29), but the subjunctive in - qw/ (Mt. 22:44), imperative pro,sqej (Lu. 17:5). The middle has the regular second aorist e;qeto (Ac. 19:21 and often).

Fhmi,) If one is surprised to see this verb put under the list of second aorist, he can turn to Blass,49 who says that it is "at once doubt," some MSS. read e;dwkej (Jo. 17:7 f.) and avfh,kete (Mt. 23:23), not to say e`w,rakej (Jo. 8:57), evlh,luqej (Ac. 21:22, B also). Moulton (Prol., p. 52) considers - ej a "mark of imperfect Gk." For further exx. of this - ej ending in the LXX and koinh, see Buresch, Rhein. Mus. etc., 1891, p. 222 f. For i[hmi and its compounds in the LXX see C. and S., Sel. fr. LXX, p. 45 f., showing numerous -- w forms, afh/kan (Xen. h=kan), etc.


Addenda 3rd ed.

imperfect and aorist." It is common in the N. T. as aorist (Mt. 4:7, for instance, e;fh. It is not always possible to decide.

2. Some - mi Presents. It is difficult to group these verbs according to any rational system, though one or two small groups (like those in - numi, - hmi) appear. The presents are more common in the N. T. than the aorists. The list is based on the uncompounded forms.

Dei,knumi. Already in the Attic deiknu,w is common, but Blass50 observes that in the N. T. the middle-passive - mi forms are still rather common. It is compounded with avna avpo evn evpi u`po, No presents (or imperfects) occur with avna and uvpo. The word itself is not used very extensively. The form dei,knumi is found once (1 Cor. 12:31), - u,w not at all. So on the other hand dei,k nu,eij occurs once (Jo. 2:18), - uj not at all. Dei,knusin is read by the best MSS. (Mt. 4:8; Jo. 5:20). The middle evndei,knuntai appears in Ro. 2:15. The - mi participle active is found in Ac. 18: 28 ( evpideinu,j) and 2 Th. 2:4 ( avpodeiknu,nta). The middle - mi participle is seen in Ac. 9:39; Tit. 2:10; 3:2 (- u,menoj, etc.). In Heb. 6:11 the infinitive evndei,knusqai is read, but deiknu,ein (Mt. 16: 21 B - u,nai).51 The other N. T. verbs in - umi ( avpo,llumi zw,nnumi u`po zw,nnumi o;mnumi sbe,nnumi strw,nnumi u`postrw,nnumi ktl.) will be discussed in alphabetical order of the simplex. The inscriptions show these forms still in use (Schweizer, Perg. Inschr., p. 178). The verbs in -- numi were the first to succumb to the - w inflection. In the LXX the - mi forms are universal in the middle, but in the active the - w forms are more usual (Thack., Gr., p. 245).

Di,dwmi. See under (d), 1, for list of compounds in the N. T. Attic Greek had numerous examples from the form dido,w $di,dou, evdi,doun, - ouj, - ou). This usage is extended in the N. T. as in the papyri52 to didw/ (Rev. 3:9), though even here BP have di,dwmi. In Wisd. of Sol. 12:19 didoi/j occurs, but Lu. 22:48 has the regular paradi,dwj. Di,dwsi is common (in LXX, Ps. 37:21, didoi/ appears) and dido,asin in Rev. 17:13. The uniform imperfect evdi,dou (Mt. 15:36) is like the Attic. Hort observes that Mk. Mk.(15:23) and Ac. Ac.(4:33; 27:1) prefer evdi,doun. Jo. Jo.(19:3) has, however, evdi,do san and Acts once also also(16:4). Di,dou (Attic present imperative) is read by Syrian MSS. in Mt. 5:42 for do,j) In Rev. 22:2 the


text has participle avpodi,dou/n, for - o,n (marg. - ou,j), while paradi dw/n is read by a in Mt. 26:46 and D in Mk. 14:42, etc.53 The middle-passive forms in - eto (imperfect) from a present di,dw are like the aorist forms, which see above. So diedi,deto (Ac. 4:35) and paredi,deto (1 Cor. 11:23). So also subjunctive paradidoi/ is found only once (1 Cor. 15:24) and is probably to be rejected (BG), though the papyri amply support it.54 In the imperfect evdi,dosan holds its place in the LXX, while in the present the forms generally prevail (Thackeray, Gr., p. 250). The LXX is quite behind the N. T. in the transition from -- mi to - w forms.

Du,namai. The use of du,nh| (Mk. 9:22; Lu. 16:2; Rev. 2:2) instead of du,nasai argues for the thematic du,nomai. Elsewhere du,nasai (Lu. 6:42, etc.). This use of du,nh| is found in the poets and from Polybius on in prose (Thayer), as shown by inscriptions55 and papyri.56 Hort57 calls it a "tragic" form retained in the koinh,. It is not surprising therefore to find B reading du,nomai (also - o,meqa, o,menoj) in Mk. 10:39; Mt. 19:12; 26:53; Ac. 4:20; 27:15; Is. 28:20 (so a in Is. 59:15). The papyri58 give plenty of illustrations also. MSS. in the LXX give du,nomai and du,nh|.

Eivmi,. The compounds are with avp evn e`vx (only e;xestin evxo,n), par sun sunpar. The papyri59 show a much more extended use of prepositions. This very common verb has not undergone many changes, though a few call for notice. In the present indicative there is nothing for remark. The imperfect shows the middle h;mhn h;meqa regularly (as Mt. 25:43; 23:30), as modern Greek uniformly has the middle present ei=mai, etc., as well as imperfect middle. Cf. already in ancient Greek the future middle e;somai. The use of seen in the papyri60 and inscriptions61 also, served to mark it off from the third singular h=n. But examples of h=men still survive (Ro. 7:5, etc.). Moulton62 quotes from Ramsay63 a Phrygian inscription of ei=mai for early fourth century A.D. He cites also the Delphian middle forms h=tai e;wtai, Messenian h=ntai,


Lesbian e;sso, as early instances of this tendency, not to mention the Northwest Greek.64 The peculiar classical second person h=sqa is found in Mk. 14:67; Mt. 26:69, but elsewhere is (Jo. 11:21, 32, etc.), the common form in the koinh,)2 =Hte (Ro. 6:20, for instance) is regular. So with the imperative (Mt. 2:13, etc.). ;Htw (as 1 Cor. 16:22) is less common65 than the usual e;stw (Gal. 1:8). ;Estwsan (never o;ntwn nor e;stwn), as in Lu. 12:35, is a form found in Attic inscriptions since 200 B.C.66 Some of the papyri even have h;twsan.67 Mention has already (Orthography) been made of the irrational n with the subjunctive in the papyri,68 as in o[tan h=n- dhlw,sw. The use of e;nie;nesti (as 1 Cor. 6:5; Gal. 3:28, etc.) an old idiom. ;Enievn and in modern Greek has supplanted evsti, in the form ei=ne or ei=nai (so for eivsi,, also).69 Cf. Sir. 37:2. N. T. has no example of imperative e;ste)

Ei=mi. Only in compounds ( avp eivj evx evp sun). The papyri70 and the inscriptions71 show only the compound forms. Blass72 indeed denies that even the compound appears in the popular koinh,, but this is an overstatement. The Attic employed e;rcomai for the present indicative and kept ei=mi for the future indicative. The koinh, followed the Ionic (and Epic) in the use of e;rcomai for all the tenses to the neglect of ei=mi. In the N. T. only Luke and the writer of Hebrews (once) use these compound forms of ei=mi and that very rarely. ;Apeimi, only occurs in the imperfect indicative (Ac. 17:10, avph|,esan%. Ei;seimi appears four times, two in the present indicative (Ac. 3:3; Heb. 9:6), two in the imperfect indicative (Ac. 21:18, 26), while eivse,rcomai, appears over two hundred times. ;Exeimi also occurs four times, all in Acts Acts(13:42; 17:15; 20:7; 27:43), against a host of instances of evxe,rcomai) ;Epeimi is read five times in Acts and all of them in the participle th|/ evpiou,sh| (Ac. 7:26, etc.). Su,neimi is found only in Lu. 8:4. B reads ei;siqi in Ac. 9:6, not ei;selqe. Blass73 rather


needlessly construes evxio,ntwn (Ac. 13:42) in the aoristic sense (so as to 17:10, 15; 21:18, 26). Ei=mi is nearly gone from the LXX (Thackeray, Gr., p. 257).

vEpi,stamai. This verb occurs fifteen times in the N. T., chiefly in Acts Acts(10:28, etc.) and always in the present tense.74

Zeu,gnumi. Only in the compound suzeu,gnumi and in the aorist active alone, sune,zeuxen (Mk. 10:9 =Mt. 19:6).

Zw,nnumi. The compounds are with avna dia peri u`po) Curiously enough the verb does not appear in Mayser, Nachmanson nor Schweizer, though Mayser (p. 397) does mention zeu,gnumi which on the other hand the N. T. does not give save the one form above. But the uncompounded form is read in the N. T. only three times, one aorist indicative (Ac. 12:8), one future indicative (Jo. 21:18), and one imperfect (Jo. 21:18, evzw,nnuej a form in - u,w, not - umi%. There is only one instance of the compound with avna and that an aorist participle (1 Pet. 1:13). The three examples of diaz., all in Jo. Jo.(13:4, etc.), yield no presents nor imperfects. The same thing is true of the half-dozen instances of periz., as Lu. 12:35, The LXX has perizw,nnutai (Thackeray, Gr., p. 269). The one instance of u`poz. is in Ac. 27:17 and shows the form in - umi u`pozwnnu,ntej)

-Hmai. It is only in the compound form ka,qhmai that this verb is seen in the N. T. and thus very frequently, twice with sun prefixed (Mk. 14:54; Ac. 26:30). It is usually the participle kaqh,menoj that one meets in the N. T. (as Mt. 9:9). The imperfect is regularly evka,qhto, etc. (as Mt. 13:1), the future kaqh,somai, (as Mt. 19:28). No - w forms appear in the present, though ka,qh| (Ac. 23:3) is a contract form like du,nh| for ka,qhso (already in Hyperides).75 The short imperative ka,qou for ka,qhsai (as Jas. 2:3) is already in the LXX (cf. Mt. 22:44 from Ps. 110:1) and indeed in the late Attic (Blass, ib.), though chiefly postclassical.76

[Ihmi. Like ei=mi this verb only appears in the N. T. in the compounded form ( avn-, avf kaq par sun). The same thing appears to be true of the papyri as given by Mayser,77 though fifteen combinations greet us in the papyri. But the papyri and the koinh, inscriptions have not yet furnished us with the - w formation with i[hmi compounds which we find in avf and suni,hmi


in the N. T.78 and the LXX.79 But Philo80 and the N. T. Apocrypha and early Christian writers81 follow the LXX and the N. T. vAni,hmi indeed has only avnie,ntej (Eph. 6:9) in the present stem. So also kaqi,hmi shows only kaqie,menon $me,nhn% in Ac. 10:11; 11:5, while pari,hmi has no present, but only an aorist (Lu. 11: 42) and a perfect passive (Heb. 12:12). vAfi,hmi is the form of the verb that is common in the N. T. In Rev. 2:20 avfei/j is probably a present from avfe,w.82 But Blass (p. 51, of N. T. Grammar) compares the Attic avfi,eij and ti,qeij. Only avfi,hmi (Jo. 14: 27) and avfi,hsi (Mt. 3:15) occur, but in Lu. 11:4 avfi,omen, is from the Ionic avfi,w (cf. di,dw). So also in Rev. 11:9 avfi,ousin and in Jo. 20:23 marg. W. H. have avfi,ontai. Elsewhere avfi,entai (Mt. 9:2, etc.). In the imperfect h;fien from afi,w is read in Mk. 1:34; 11:16. vAfe,wntai (Lu. 5:20, 23, etc.) is a perfect passive (Doric Arcadian, Ionic).83 Cf. Ionic e;wka. Simcox (Language of the N. T., p. 38) quotes also avne,wntai from Herodotus. With suni,hmi the task is much simpler. Blass84 sums it up in a word. In Ac. 7:25 sunie,nai, gives us the only undisputed instance of a - mi form. All the others are - w forms or have - w variations. However sunie,ntoj is correct in Mt. 13:19 and sunie,nai in Lu. 24:45. There is a good deal of fluctuation in the MSS. in most cases. W. H. read suni,ousin (Mt. 13:13), suni,wsin, (Mk. 4:12), suni,wn (Ro. 3 : 11). In 2 Cor. 10:12 W. H. read sunia/sin after B. In the LXX only the compounded verb occurs, and usually the - mi forms save with suni,hmi (Thackeray, Gr., p. 250 f.).

[Isthmi. Cf. also evpi,stamai. (see above) and sth,kw (from e[s thka, imperfect e;sthke in Rev. 12:4, ste,kw in modern Greek). For the list of compounds85 see list of aorists (1). But the essential facts can be briefly set forth. The - mi form in the present stem has disappeared in the active voice save in kaqi,sthsin, (Heb. 7:28; 2 Pet. 1:8), suni,sthmi (Ro. 16:1) and suni,sthsi (2 Cor. 10:18; Ro. 3:5; 5: 8).86 The middle (passive) forms retain the - mi inflection regularly with i[sthmi and its compounds ( avn avf auvq evx evf pro sun), as kaqi,stati (Heb. 5:1), peri,istaso


(2 Tim. 2:16).87 Two -- w forms supplant the -- mi conjugation of i[sthmi and its compounds, that in - a,w and that in - a,nw, though usually the MSS. vary greatly between the two.88 In 1 Cor. 13:2 aBDEFG read meqista,nai, though W. H. follow ACKL in meqi sta,nein.89 The form in - a,w is found in various MSS. for i`sta,w (as i`stw/men Ro. 3:31), avpokaq (Mk. 9:12 Rec.), evxista,w kaqista,w meqista,w sunista,w, but is nowhere accepted in the W. H. text, though Hort90 prefers sunista|/n to sunista,nein in 2 Cor. 3:1. In 2 Cor. 4:2 a threefold division occurs in the evidence. For suni sta,nontej we have ABP (so W. H. and Nestle), for sunista,ntej aCD*FG, for sunistw/tej DcEKL.91 The form in - a,nw is uniformly given by W. H., though the form in - a,w comes from Herodotus on and is frequent in the LXX.92 But the mi forms hold their own pretty well in the LXX (Thackeray, Gr., p. 247). The form in - a,nw, may be compared with the Cretan stanu,ein and is found in the late Attic inscriptions.93 Instances of the form in - a,nw in the W. H. text are Ac. 1:6; 8:9; 17:15; 1 Cor. 13:2; 2 Cor. 3:1; 5:12; 6:4; 10:12, 18; Gal. 2:18; Ro. 3:31; 6:13, 16). In Mk. 9:12 W. H. (not so Nestle) accept the form avpokatista,nei after B, while aD read avpokatasta,nei (cf. Cretan stanu,w). D has this form also in Ac. 1:6 and 17:15.

Kei/ma. This defective verb is only used in the present and imperfect in the N. T. as in the papyri,94 and with a number of prepositions in composition like the papyri also. The prepositions are avna sunana avnti avpo evpi kata para peri pro) The regular -- mi forms are always used, and sometimes as the passive of ti,qhmi, as peri,keimai (Ac. 28:20; Heb. 5:2). For avna,keimai only the participle avnakei,menoj appears (so Mt. 9:10) save once avne,keito (Mt. 26:20) and twice with su,n (Mt. 9:10 Mk. 2:15). In Lu. 23: 53 h=n kei,menoj follows the Attic, but aB have h=n teqeime, noj in Jo. 19:41.95 So in the LXX ti,qhmi partially replaces kei/mai (Thackeray, Gr., pp. 255, 272).

Kre,mamai. This verb is used as the middle of the active krema,n numi (this form not in N. T.) and does not appear in Mayser's list


for the papyri. The form96 kre,matai is read in Mt. 22:40 and the participle krema,menoj( n) in Gal. 3:13; Ac. 28:4. In Lu. 19:48 aB (so W. H. anq Nestle) read evxekre,meto, an - w form and the only compound form of the verb in the N. T. The other forms are aorists which come from an active present krema,nnumi annu,w a,w or - a,zw) They are krema,santej (Ac. 5:30) and kremasqh|/ (Mt. 18: 6). But none of these presents occurs in the N. T. Cf. Veitch, Greek Verbs, p. 343 f., for examples of the active and the middle. So also no present of kera,nnumi, (compound sun) is found in the N. T., but only the perfect passive (Rev. 14:10) and the aorist active (Rev. 18:6).

Mi,gnumi. The only - mi form is the compound sunanami,gnusqai (1 Cor. 5:9, 11) and so 2 Th. 3:14 according to W. H., instead of sunanami,gnusqe. Elsewhere, as in the papyri,97 the N. T. has only the perfect passive (Mt. 27:34) and the aorist active (Lu. 13:1).

Oi;gnumi. This verb does not appear in the N. T. in the simple form, but always compounded with avn- or dian--. Besides it is always an - w verb as in the papyri98 and the LXX.99 It is worth mentioning here to mark the decline of the - mi forms.

;Ollumi. Only in the common avp and once with sunap (Heb. 11:31). In the active only the - w forms are found as avpollu,ei (Jo. 12:25), avpo,llue (Ro. 14:15). But in the middle (passive) only the - mi, forms100 meet us, as avpo,llutai (1 Cor. 8:11), avpw,llunto (1 Cor. 10:9). So the LXX.

;Omnumi. A half-dozen examples of the present tense of this verb occur in the N. T. All but one ( ovmnu,nai, Mk. 14:71) belong to the - w inflection, as ovmnu,ei. (Mt. 23:21 f.). The Ptolemaic papyri also have one example of o;mnumi, the rest from ovmnu,w.101 The LXX sometimes has the - mi form in the active and always in the middle (Thackeray, Gr., p. 279). Neither ph,gnumi. (aorist Heb. 8:2) nor prosph,gnumi (aorist Ac. 2:23) appears in the present in the N. T.

Pi,mplhmi. No present tense in the N. T., though a good many aorists, save the compound participle evmpiplw/n, from the - w verb -- a,w) Mayser102 gives no papyri examples. LXX has - w form usually.


Pi,mprhmi) The simple verb occurs once only, pi,mprasqai. (Ac. 28:6) according to W. H.103 This is the only instance where a present occurs at all in the N. T. The papyri give no light as yet. No simplex in the LXX, but evnepi,mprwn in 2 Macc. 8:6 (Thackeray, Gr., p. 249).

`Rh,gnumi. The compounds are with dia peri pros) No presents appear save in the simple verb and diar. With diar) only the - w forms are used as dierh,sseto (Lu. 5:6), diarh,sswn (Lu. 8:29). But we have r`h,gnuntai (Mt. 9:17) and r`h,ssei (Mk. 9:18). Mayser gives no papyri examples of the present.

`Rw,nnumi has no presents at all in the N. T., but only the, perfect passive imperative e;rrwsqe (Ac. 15:29).

Sbe,nnumi. This verb has only three presents in the N. T. and all of theform, one active sbe,nnute (1 Th. 5:19, Tisch. zbenn), two middle sbe,nnutai (Mk. 9:48) and sbe,nnuntai (Mt. 25: 8). The LXX has only -- mi forms and in the more literary books (Thackeray, Gr., p. 284).

Strw,nnumi. The compounds are with kata u`po) There are only two present stems used in the N. T., evstrw,nnuon (Mt. 21:8) and u`post) (Lu. 19:36). Thus the - mi form is wholly dropped as in the papyri104 and the LXX.105

Ti,qhmi. For the list of compounds see Aorist (1). This verb has preferred the - mi form of the present stem as a rule in the koinh,. The inscriptions106 do so uniformly and the papyri107 use the - w inflection far less than is true of di,dwmi. In the present indicative D has ti,qi ( ti,qei% for ti,qhsi108 (Lu. 8:16). In the imperfect evti,qei is read twice (Ac. 2:47; 2 Cor. 3:13) from tiqe,w, as already in the Attic. So likewise evti,qoun (as in Attic) twice (Ac. 3:2; 4:35), but the best MSS. have evti,qesan in Mk. 6:56 ( aBL D) and Ac. 8:17 ( aAC, though B has - osan and C - eisan).109 The reading of B in Ac. 8:17 ( evti,qosan) calls for a present ti,qw which the papyri supply against the idea of Winer-Schmiedel,110 as paratiqo,menoj (BM 239), para katati,qomai (B.0 326).111 Good cursives show that the late language used tiqe,w in the present (Mk. 10:16; 15:17). Cf. u`potiqou/sa in second century papyrus (B.U. 350).112 In the LXX forms prevail in the present and imperfect (Thackeray, Gr., p. 250).


Fhmi,. The only N. T. compound is with sun-, none in the papyri according to Mayser.113 In the papyri fa,skw (lengthened form) is usually employed for the participle and infinitive114 of fhmi,. The participle is so used in the N. T. (Ac. 24:9; Ro. 1: 22). Su,nfhmi appears only once (Ro. 7:16). The - mi inflection is uniform in fhmi, both in the present and the imperfect (aorist). The only forms in the N. T. are fhmi, (1 Cor. 7:29), fhsi,n (Mt. 13:29), fasi,n (Ro. 3:8), and the common e;fh (Mt. 4:7). It is regular - mi in the LXX.

Crh,) This impersonal verb had a poetic infinitive crh/nai of the -- mi inflection, but Veitch (p. 627) and L. and S. get it from cra,w. At any rate crh, is found only once in the N. T. (Jas. 3:10), dei/ having supplanted it. Mayser does not find it in the papyri nor Nachmanson. and Schweizer in the inscriptions.

3. Some - mi Perfects. There are only three verbs that show the active perfects without $k%a in the N. T. (mere root, athematic).

qnh,skw. The compounds are avpo (very common), sunapo (rare). The uncompounded verb occurs nine times and forms the perfect regularly as an - w verb ( te,qnhka), save that in Ac. 14: 19 DEHLP read teqna,nai instead of teqnhke,nai, but the - mi form is not accepted by W. H. The N. T. has always teqnhkw,j, never tenqnew,j. In the LXX these shorter second perfect forms occur a few times in the more literary books (Thackeray, Gr., pp. 253, 270). They show "a partial analogy to verbs in - mi," (Blass, Gr., p. 50).

Oi=da is a - mi perfect in a few forms ( i;smen i;ste) from root id- (cf. Latin vid-eo Greek ei=don). The word is very common in the N. T. and su,noida is found twice (Ac. 5:2; 1 Cor. 4:4). The present perfect indicative like the papyri115 usually has oi=da oi=daj oi=de oi;damen, - ate asin, which was the Ionic inflection and so naturally prevailed in the koinh,. Three times indeed the literary Attic i;ste appears (Jas. 1:19; Eph. 5:5; Heb. 12:17). The passage in James may be imperative instead of indicative. In Ac. 26:4 i;sasin (literary Attic also) is read. The imperfect also runs h|;dein h|;deij, etc. ;Hideisan (Mk. 1:34; 14:40) is like i`sth,keisan (Rev. 7:11).116 The other modes go regularly eivdw/ (Mt. 9:6), eivde,nai (1 Th. 5:12), eivdw,j (Mt. 12:25). The LXX usage is in accord with the N. T. Cf. Thackeray, Gr., p. 278.

[Isthmi. See Aorist (1) for compounds. The second perfect is in the N.T.the infinitive e`sta,nai (Lu. 13:25; Ac. 12:14;


1 Cor. 10:12) and the participle e`stw,j (Mt. 20:3, 6, etc.) though e`sthkw,j (- w form) also sometimes (Mk. 13:14; 15:35, etc.), e`stw/sa (1 Cor. 7:26; 2 Pet. 3:5), e`sto,j (Mt. 24:15; Rev. 14:1) although e`sthko,j also (Rev. 5:6 text, W. H. marg. - w,j). The same variation occurs in the papyri. Curiously enough the earlier LXX books show less of the short perfect than the later ones and the N. T. Thackeray (Gr., p. 253) suggests an "Atticistic reversion" for a while. The form e[staka (papyri also) belongs to the - w form as well as the late present sth,kw from the perfect stem. These - mi perfects of i[sthmi in are always intransitive, while e[sthka is intransitive and e[staka is transitive.117 This in brief is the story of the - mi verbs in the N. T.118 The new transitive perfect e[staka is common in the koinh, from second century B.C. onwards. Cf. Schweizer, Perg. Inschr., p. 185; Mayser, Gr., p. 371.

V. The Modes ( evgkli,seij). The meaning and use of the modes or moods belongs to syntax. We have here to deal briefly with any special items that concern the differentiation of the modes from each other by means of mode-signs. There is no clearly proper method of approaching the study of the verb. One can begin with tense, voice and then mode or vice versa. The first is probably the historical order to a certain extent, for the matter is complicated. Some tenses are later than others; the passive voice is more recent than the other two, the imperative as a complete system is a late growth. Since no purely historical treatment is possible by reason of this complicated development, a practical treatment is best. There are reasons of this nature for taking up modes first which do not apply to syntax. The two main ideas in a verb are action and affirmation. The state of the action is set forth by the tense, the relation of the action to the subject by voice, the affirmation by mode. Tense and voice thus have to do with action and mode with affirmation. Mode deals only with the manner of the affirmation. The same personal endings used for voice limit the action (hence finite verbs) in person and number.

(a) THE NUMBER OF THE MOODS OR MODES (Modi). This is not so simple a matter as it would at first appear. Modern grammarians generally agree in declining to call infinitives, participles and the verbal adjectives in - to,j and - te,oj moods. Some refuse to call the indicative a mood, reserving the term for the variations


from the indicative as the normal verb by means of mode-signs. Thus Clyde119 thinks of "only two moods, viz. the subjunctive and the optative, because, these only possess, in combination with the personal endings, a purely modal element." There is point in that, and yet the indicative and imperative can hardly be denied the use of the term. Jannaris120 admits three moods; indicative, subjunctive and imperative. He follows Donaldson121 in treating the subjunctive and optative as one mood. Others, like Monro,122 find the three in the subjunctive, optative and imperative. Once again five moods are seen in early Greek by Riemann and Goelzer123: the indicative, injunctive, subjunctive, optative, imperative. On the injunctive see Brugmann, Griechische Grammatik, p. 332, though he does not apply the term mode to the indicative. So Hirt, Handbuch, p. 421 f. Moulton124 admits this primitive division, though declining to call the indicative a mode save when it is a "modus irrealis." The injunctive is no longer regarded as a separate mood, and yet it contributed so much to the forms of the imperative that it has to be considered in an historical review. The indicative can only be ruled out when it is regarded as the standard verb and the moods as variations. Certainly it is best to let the indicative go in also. The modern Greek, having no optative, has a special conditional mode ( u`poqe tikh,). Cf. Sanskrit. Indeed, the future indicative is considered by some grammarians as a separate mode. Cf. Thompson, Syntax of Attic Greek, p. 494; Moulton, Prolegomena, p. 151. Thumb accepts the four modes in modern Greek (Handbook, p. 115).

(b) THE DISTINCTIONS BETWEEN THE MOODS. These are not absolute, as will be seen, either in form or in syntax. The indicative and the imperative blend in some forms, the subjunctive and the indicative are alike in others, the injunctive is largely merged into the imperative and subjunctive, while the subjunctive and optative are closely akin and in Latin blend into one. Greek held on to the optative with separate values to each mood.125 Moulton126 indeed despairs of our being able to give the primitive root-idea of each mood. That subject belongs to


syntax, but the history of the mode-forms is in harmony with this position. As with the cases so with the moods: each mood has fared differently in its development and long history. Not only does each mood perform more functions than one, but the same function may sometimes be expressed by several127 moods. The names themselves do not cover the whole ground of each mood. The indicative is not the only mood that indicates, though it does it more clearly than the others and it is used in questions also. The subjunctive not merely subjoins, but is used in independent sentences also. The optative is not merely a wish, but was once really a sort of past subjunctive. The imperative has the best name of any, though we have to explain some forms as "permissive" imperatives, and the indicative and subjunctive, not to say injunctive, invade the territory of the imperative." It is probable, but not demonstrable, that the indicative was the original verb-form, from which the, others were evolved by morphological changes" (Thompson, Syntax of Attic Greek, p. 494). The origin of the mode-signs cannot yet be explained.

(c) THE INDICATIVE ( o`ristikh. e;gklisij). There is indeed little to say as to the form of the indicative since it has no mode-sign. It is the mode that is used in all the Indo-Germanic languages unless there is a special reason to use one of the others. In fact it is the normal mode in speech. It is probably the earliest and the one from which the others are derived. Per contra it may be argued that emotion precedes passionless intellection. The indicative continues always to be the most frequent and persists when others, like the injunctive and optative, die. It is the only mode that uses all the tenses in Sanskrit and Greek. In the Sanskrit, for instance, the future is found only in the indicative (as in Greek save in the optative in indirect discourse to represent a future indicative of the direct) and the perfect appears only in the indicative and participle, barring many examples of the other modes in the early Sanskrit (Vedas). In the Sanskrit the modes are commonest with the aorist and the present.128 And in Greek the imperfect and past perfect never got beyond the indicative. The future barely did so, never in the subjunctive till the Byzantine period. The perfect subjunctive and optative, not to say imperative, were always a rarity outside of the periphrastic forms and


in the koinh, have practically vanished.129 Thus we can clearly see the gradual growth of the modes. In modern English we have almost dropped the subjunctive and use instead the indicative. In the modern Greek the indicative survives with as much vigor as ever. The N. T. peculiarities of the indicative can best be treated under Syntax. It may be here remarked, however, that besides the regular indicative forms a periphrastic conjugation for all the tenses of the indicative appears in the N. T. The present is thus found as evsti.n prosanaplhrou/sa (2 Cor. 9:12), the perfect as evsti.n pepragme,non (Ac. 26:26), the imperfect as h=n dida, skwn (Lu. 5:17), the past perfect as h=san proewrako,tej (Ac. 21:29), even the aorist as h=n blhqei,j (Lu. 23:19), the future as e;sesqe la lou/ntej (1 Cor. 14:9), the future perfect as e;somai pepoiqw,j (Heb. 2:13). This widening of the range of the periphrastic conjugation is seen also in the LXX. Cf. Thackeray, Gr., p. 195.

(d) THE SUBJUNCTIVE ( u`potaktikh,). The function of the subjunctive as of the other modes will be discussed under Syntax. Changes come in function as in form. Each form originally had one function which varied with the course of time. But the bond between form and function is always to be noted.130 The German grammarians (Blass, Hirt, Brugmann, etc.) call this the conjunctive mode. Neither conjunctive nor subjunctive is wholly good, for the indicative and the optative both fall often under that technical category.131 It is in the Greek that mode-building reaches its perfection as in no other tongue.132 But even in the Greek subjunctive we practically deal only with the aorist and present tenses, and in the Sanskrit the subjunctive rapidly dies out save in the first person as an imperative.133 In Homer i;men is indicative134 and i;omen is subjunctive so that non-thematic stems make the subjunctive with the thematic vowel o/ e. Thematic stems made the subjunctive with a lengthened form of it wh. Cf. in the Ionic, Lesbian, Cretan inscriptions135 forms like avmei,yetai. The same thing appears in Homer also in the transition period.136 Jannaris137 indeed calls the aorist subjunctive a future subjunctive because he


conceives of the aorist as essentially past, a mistaken idea. The subjunctive does occur more freely in Homer than in the later Greek, partly perhaps because of the fact that the line of distinction between it and the indicative (especially the aorist subjunctive and the future indicative) had not been sharply drawn.138 Add to this the fact that poih,sh| and poih,sei came to be pronounced exactly alike and one can see how the confusion would come again. Cf. i[na dw,sei ( dw,sh|) in the N. T. MSS.139 On the short vocal ending of the subjunctive and its connection with the indicative one may recall e;domai pi,omai fa,gomai, in the N. T., futures which have a strange likeness to the Homeric subjunctive i;omen. They are really subjunctives in origin. It is still a mooted question whether the future indicative is always derived from the aorist subjunctive or in part corresponds to the Sanskrit sya.140 The only aorist subjunctives that call for special comment in the N. T. are the forms gnoi/ and doi/, for which see this chapter, iv, (d), 1.141 There are parallels in the papyri as is there shown. The form o;yhsqe in Lu. 13:28 (supported by AL, etc., against o;yesqe, BD) is probably a late aorist form like e;dwsa ( dw,sh|) rather than the Byzantine future subjunctive.142 As already pointed out, the examples in N. T. MSS. of the Byzantine future subjunctive are probably due to the blending of o with w, ei with h|, e with h, etc. N. T. MSS., for instance, show examples of avrkesqhsw,meqa (1 Tim. 6:8), gnw,swntai (Ac. 21:24), genh,shsqe (Jo. 15:8), dw,swsin (Lu. 20:10; Rev. 4: 9), eu`rh,swsin (Rev. 9:6), zh,shtai (Mk. 5:23), h[xwsin (Rev. 3:9), kauqh,swmai (1 Cor. 13:3), kerdhqh,swntai (1 Pet. 3:1) , pore,swmai (Ro. 15:24), swqh,shtai (Ro. 11:26), etc. It is to be admitted, however, that the Byzantine future subjunctive was in use at the age of our oldest Greek N. T. MSS. Cf. Winer-Schmiedel, p. 107. Hort dismisses them all (Appendix, "Notes on Orthography," p. 172). The present subjunctive didoi/, is parallel to doi/. No ex-


Addenda 3rd ed.

ample of the periphrastic present subjunctive appears in the N. T. In Gal. 4 :17 ( i[na zhlou/te) the contraction of oh is like that of the indicative oe,143 unless indeed, as is more probable, we have here (cf. also 1 Cor. 4:6, fusiou/sqe) the present indicative used with is as in 1 Jo. 5:20 ( ginw,skomen). In Gal. 6:12 ACFGKLP read i[na mh. diw,kontai. Cf. Ro. 14:19. Cf. Homer. The perfect subjunctive does not exist in the N. T. save in the second perfect eivdw/ ( i[na eivdw/men, 1 Cor. 2:12) and the periphrastic form as h|= pepoihkw,j (Jas. 5:15. Cf. pepoiqo,tej w=men, 2 Cor. 1:9) and usually in the passive as h|= peplhrwme,nh (Jo. 16:24). In Lu. 19:40 Rec. with most MSS. read kekra,xontai (LXX). In the papyri h=n sometimes is subjunctive = h=i. Cf. Moulton, Cl. Rev., 1901, p. 38, 1904, p. 108; Prolegomena, pp. 49, 468. He cites o[sa eva.n h=n in Gen. 6:17E. But the modern Greek constantly uses eva,n with the indicative, and we find it in the N. T. and papyri (Deissmann, Bible Studies, pp. 203 ff.). Some of the papyri examples may be merely the indicative with eva,n, but others undoubtedly give the irrational n. In the LXX the subjunctive shows signs of shrinkage before the indicative with eva,n o[tan i[na (Thackeray, Gr., p. 194).

(e) THE OPTATIVE ( euvktikh,). Like the subjunctive the optative is poorly named, as it is much more than the wishing mood. As Giles144 remarks, difference of formation is more easily discerned in these two moods than difference of meaning. In the Sanskrit the subjunctive (save in first person) gave way before the optative, as in Latin the optative largely (sim originally optative) disappeared before the subjunctive.145 The Greek, as already stated, is the only language that preserved both the subjunctive and the optative,146 and finally in the modern Greek the optative has vanished, mh. ge,noito being merely "the coffin of the dead optative."147 It is doubtful if the optative was ever used much in conversation even in Athens (Farrar, Greek Syntax, p. 142), and the unlearned scribes of the late Greek blun-


Addenda 2nd ed.

dered greatly when they did use it (Jannaris, Hist. Gk. Gr., p. 204). Moulton (Prol., p. 240) agrees with Thumb that the optative was doomed from the very birth of the koinh, and its disappearance was not due to itacism between oi, and h|, which was late. Clyde,148 however, suggests that the blending of sound between oi and h| had much to do with the disappearance of the optative. But apart from this fact the distinction was never absolutely rigid, for in Homer both moods are used in much the same way.149 And even in the N. T., as in Homer and occasionally later, we find an instance of the optative after a present indicative, ouv pau,o mai euvcaristw/n i[na dw|,h (Eph. 1:17, text of W. H., subj. dw,h| or dw|/ in marg., question of editing). Jannaris150 calls the Greek optative the subjunctive of the past or the secondary subjunctive (cf. Latin). Like the indicative (and originally the subjunctive) the non-thematic and thematic stems have a different history. The non-thematic stems use ih ( ie) and the thematic oi (composed of o and i). The s aorist has a+ i besides the form in - eia. This two-fold affix for the optative goes back to the earlier Indo-Germanic tongues151 (Sanskrit ya and i). The optative was never common in the language of the people, as is shown by its rarity in the Attic inscriptions.152 The Boeotian dialect inscriptions show no optative in simple sentences, and Dr. Edith Claflin reports only two examples in subordinate clauses.153 The optative is rare also in the inscriptions of Pergamum.154 The same thing is true of the papyri.155 In the N. T. the future optative no longer appears, nor does the perfect. The classic idiom usually had the perfect subjunctive and optative in the periphrastic forms.156 Examples of the periphrastic perfect optative survive in the papyri,157 but not in the N. T. There are only sixty-seven examples of the optative in the N. T. Luke has twenty-eight and Paul thirty-one (not including Eph. 1:17), whereas John, Matthew and James do not use it at all. Mark and Hebrews show it only once each, Jude twice and Peter four times. The non-thematic aorist appears in the N. T. sometimes, as dw|,h (perhaps by analogy). So W. H. read without reservation in 2 Th. 3:16; Ro. 15:5; 2 Tim. 1:16, 18. This is the


preferred text in Eph. 1:17; 2 Tim. 2:25, but in Jo. 15:16; Eph. 3:16, W. H. read dw|/ (subjunctive). In Eph. 1:17 the margin has dw,h| (subjunctive) also.158 The inscriptions159 and the papyri160 show the same form (-- w|,hn instead of - oi,hn). In Eph. 1:17 Moulton161 considers dw,h| (subjunctive) absolutely necessary in spite of the evidence, for dw|h (optative). But see above. The aorist optative in - ai is the usual form, as kateuqu,nai (1 Th. 3:11), pleona,sai kai. perisseu,sai (1 Th. 3:12), katarti,sai (Heb. 13:21), etc., not the AEolic-Attic - eie. So also poih,saien (Lu. 6:11), but yhlafh,seian (Ac. 17:27) according to the best MSS. (B, etc.).162 Blass163 comments on the fact that only one example of the present optative appears in the simple sentence, viz. ei;h (Ac. 8:20), but more occur in dependent clauses, as pa,scoite (1 Pet. 3:14). The optative is rare in the LXX save for wishes. Thackeray, Gr., p. 193.

(f) THE IMPERATIVE ( prostaktikh,%. The imperative is a later development, in language and is in a sense a makeshift like the passive voice. It has no mode-sign (cf. indicative) and uses only personal suffixes.164 These suffixes have a varied and interesting history.

1. The Non-Thematic Stem. An early imperative was just the non-thematic present stem.165 In the imperative the aorist is a later growth, as will be shown directly. Forms like i[sth dei,knu are pertinent.

2. The Thematic Stem. Cf. a;ge le,ge. This is merely an interjection (cf. vocative lo,ge).166 This is the root pure and simple with the thematic vowel which is here regarded as part of the stem as in the vocative lo,ge. The accent eivpe, evlqe, eu`re, ivde, labe, was probably the accent of all such primitive imperatives at the beginning of a sentence.167 We use exclamations as verbs or nouns.168


In Jas. 4:13 we have a;ge nu/n oi` le,gontej, an example that will illustrate the origin of a;ge. Note the common interjectional use of i;de (so N. T.). Cf. also accent of la,be. The adverb deu/ro (Jo. 11:43, La,zare deu/ro e;xw) has a plural like the imperative in - te (Mt. 11:28, deu/te pro,j me pa,ntej oi` kopiw/ntej).

3. The Suffix - qi. The non-thematic stems also used the suffix - qi (cf. Sanskrit dhi, possibly an adverb; cf. "you there!"). So gnw/qi for second aorist active, i;sqi for present active, fa,nhqi lu, qhti for second and first aorist passive.169 In the N. T. sometimes this - qi is dropped and the mere root used as in avna,ba (Rev. 4: 1), meta,ba (Mt. 17:20), avna.sta (Eph. 5:14; Ac. 12:7) according to the best MSS.170 The plural avna,bate (Rev. 11:12) instead of avna,bhte is to be noted also. The LXX MSS. exhibit these short forms ( avna,sta avpo,sta, but not avna,ba) also. Cf. Helbing, Gr. d. Sept., p. 70; Con. and Stock, Sel. from LXX, p. 46. See e;mba, kata,ba, etc., in Attic drama. But avna,sthqi (Ac. 8:26), evpi,sthqi 2 Tim. 4:2), meta,bhqi (Jo. 7:3), kata,bhqi (Lu. 19:5), prosana,bhqi (Lu. 14:10) occur as usual. In the papyri - qi has practically disappeared save in i;sqi.171

4. The Suffix - tw. It is probably the ablative of the demonstrative pronoun (Sanskrit tad). It is used with non-thematic ( e;stw) and thematic stems ( lege,tw). The Latin172 uses this form for the second person also (agito). In the case of e;stw (Jas. 1:19) the N. T. has also h;tw (Jas. 5:12).173 The form kataba,tw (Mt. 24: 17) has the unlengthened stem, but evlqa,tw is like the first aorist evpistreya,tw. The N. T. like the koinh, generally174 has the plural only in twsan which is made by the addition of san to tw. Cf. e;stwsan (Lu. 12:35). The middle sqw (of uncertain origin)175 likewise has the plural in the N. T. in sqwsan. So proseuxa,sqwsan (Jas. 5:14). This is true of the plural of both present and aorist as in papyri and inscriptions. So the LXX cf. Helbing, Gr. d. Sept., p. 69 f.

5. The Old Injunctive Mood. It is responsible for more of the imperative forms than any other single source. "The injunctive


was simply an imperfect or aorist indicative without the augment."176 So labou/ corresponds to evla,beso la,besqe was evla,besqe lh,fqhte was evlh,fqhte la,bete was evla,bete.177 So sce,j ( e;scej) may be compared with e;luej ( qi,gej with e;qigej), but do,j e[j qe,j Brugmann considers of uncertain origin, possibly subjunctive.178 Forms like lue,te may be injunctive ( evlu,ete)179 or merely the indicative.180 Note the difficulty of deciding on imperative and indicative in forms like evrauna/te (Jo. 5:39), pisteu,ete (Jo. 14:1) i;ste (Jas. 1: 19). But in these cases, except Jo. 5:39, we probably have the imperative. In the case of i;ste the N. T. indicative would be oi;date.181 In the N. T. ka,qou (Jas. 2:3) is the shorter form of ka,qhso, though not by phonetic processes. The injunctive survives to some extent in the Sanskrit and borders on the subjunctive and the imperative and was specially common in prohibitions.182 It consists of the bare stem with the personal endings.

6. Forms in - sai) These, like ba,ptisai (Ac. 22:16), are probably just the infinitive sigmatic aorist.183 Cf. dei/xai. Cf. also Latin legimini with the Homeric infinitive lege,menai.184 The infinitive is common in the Greek inscriptions in the sense of an imperative.185 In the N. T. as in the papyri this use is not infrequent. So cai,rein (Jas. 1:1), stoicei/n (Ph. 3:16), mh. sunanami,gnusqai (2 Th. 3:14). In modern Greek instead of the imperative in - sai the form lu,sou occurs with the sense of lu,qhti.186

7. The Form in - son $lu/son). It is difficult of explanation. It may be injunctive or a verbal substantive.187 The N. T. has eivpon (Mt. 4:3) rather than eivpe, (Mt. 8:8) in about half the instances in W. H.188 This is merely in keeping with the common koinh, custom of using first aorist endings with second aorist stems. The form eivpo,n is traced to the Syracusan dialect.189

8. First Person. The Sanskrit used the first person subjunctive as imperative of the first person. Cf. English "charge we the foe." The Greek continued this idiom. But already in the N. T. the use of the imperative a;fej (Cf. modern Greek as and third person subjunctive) is creeping in as a sort of particle with the subjunctive. So a;fej evkba,lw (Mt. 7:4). Cf. English "let" with infini-


tive. Cf. An deu/te avpoktei,nwmen in Mt. 21:38. Besides a;ge deu/te we may have o[ra with the subjunctive (Mt. 8:4), ble,pete with future indicative (Heb. 3:12).

9. Prohibitions. Here the aorist subjunctive with mh, held its own against the aorist imperative quite successfully. In the Sanskrit Veda the negative ma is never found with the imperative, but only with the subjunctive.190 Later the Sanskrit uses the present imperative with ma, but not the aorist. This piece of history in the Greek191 is interesting as showing how the imperative is later than the other modes and how the aorist imperative never won its full way into prohibitions. However, in the N. T. as in the inscriptions and papyri, we occasionally find the aorist imperative with mh, in 3d person. So mh. kataba,tw (Mt. 24:17).

10. Perfect Imperative. In the Sanskrit the imperative is nearly confined to the present tense. The perfect imperative is very rare in the N. T. (only the two verbs cited) as in all Greek. We find e;rrwsqe (Ac. 15:29; in 23:30 W. H. reject e;rrwso) and peri,mwso (Mk. 4:39). The perfect imperative also occurs in the periphrastic form as e;stwsan periezwme,nai, (Lu. 12:35).

11. Periphrastic Presents. Other periphrastic forms of the imperative are i;sqi euvnow/n (Mt. 5:25), i;sqi e;cwn (Lu. 19:17), mh. gi,nesqe e`terozugou/ntej (2 Cor. 6:14) and even i;ste ginw,skontej (Eph. 5:5).

12. Circumlocutions. But even so other devices (see Syntax) are used instead of the imperative, as the future indicative ( avga ph,seij, Mt. 5:43); i[na and the subjunctive (Eph. 5:33); a question of impatience like ouv pau,sh| diastre,fwn (Ac. 13:10), etc.

VI. The Voices ( diaqe,seij).

(a) TRANSITIVE AND INTRANSITIVE. The point is that "transitive" is not synonymous with "active." Transitive verbs may belong to any voice, and intransitive verbs to any voice. Take evdi,daxa evdidaxa,mhn evdida,cqhn, which may be transitive in each voice. On the other hand eivmi, gi,nomai evlu,qhn, are intransitive. The same verb may be transitive or intransitive in the same voice, as a;gw. A verb may be transitive in Greek while intransitive in English, as with katagela,w and vice versa. This matter properly belongs to syntax, but it seems necessary to clear it up at once before we proceed to discuss voice. Per se the question of transitiveness belongs to the idea of the verb itself, not to that of voice. We


actually find Green192 making four voices, putting a neuter ( ouvde, teron) voice (using active and middle endings) on a par with the others! The Stoic. grammarians193 did speak of a neuter voice as neither active ( kathgo,rhma ovrqo,n% nor passive ( u[ption), meaning the middle ( me,sh). Jannaris194 confounds transitiveness with voice, though he properly says (p. 356) that "the active voice is usually transitive," i.e. verbs in the active voice, not the voice itself. Even Whitney195 speaks of the antithesis between transitive and reflexive action being effaced in Sanskrit. Was that antithesis ever present? Farrar196 speaks of verbs with an "active meaning, but only a passive or middle form," where by "active" he means transitive. Even the active uses verbs which are either transigive ( avllopaqh,j) or intransitive ( auvtopaqh,j). So may the other voices. If we clearly grasp this point, we shall have less difficulty with voice which does not deal primarily with the transitive idea. That belongs rather to the verb itself apart from voice.197 On transitive and intransitive verbs in modern Greek see Thumb, Handb., p. 112.

(b) THE NAMES OF THE VOICES. They are by no means good. The active ( evnergetikh,) is not distinctive, since the other voices express action also. This voice represents the subject as merely acting. The Hindu grammarians called the active parasmai padam ('a word for another,') and the middle ( me,sh) atmane padam ('a word for one's self').198 There is very little point in the term middle since it does not come in between the active and the passive. Indeed reflexive is a better designation of the middle voice if direct reflexive is not meant. That is rare. The middle voice stresses the interest of the agent. Cf. Moulton, Prolegomena, p. 155 f. In truth we have no good name for this voice. Passive ( paqhtikh,) is the best term of all, for here the subject does experience the action even when the passive verb is transitive, as in evdida,cqhn. But this point encroaches upon syntax.


(c) THE RELATIVE AGE OF THE VOICES. It is a matter of doubt as between the active and middle. The passive is known to be a later development. The Sanskrit passive is the ya class.199 In Homer the passive has not reached its full development. The passive future occurs there only twice. The aorist middle is often used in passive sense ( blh/to, for instance).200 That is to say, in Homer the passive uses all the tenses of the middle with no distinct forms save sometimes in the aorist. In later Greek the future middle (as timh,somai) continued to be used occasionally in the passive sense. The aorist passive in fact used the active endings and the future passive the middle, the passive contributing a special addition in each case ( h qh hs qhs). Some languages never developed a passive (Coptic and Lithuanian, for instance), and in modern English we can only form the passive by means of auxiliary verbs. Each language makes the passive in its own way. In Latin no distinction in form exists between the middle and the passive, though the middle exists as in potior, utor, plangor, etc. Giles201 thinks that the causative middle (like dida,skomai, 'get taught') is the explanation of the origin of the Greek passive. Cf. ba,ptisai (Ac. 22:16). It is all speculation as between the active and middle. An old theory makes the middle a mere doubling of the active (as mamimai).202 Another view is that the middle is the original and the active a shortening due to less stress in accent, or rather (as in ti,qemai and ti,qhmi) the middle puts the stress on the reflexive ending while the active puts it on the stem.203 But Brugmann204 considers the whole question about the relation between the personal suffixes uncertain. Of one thing we may be sure, and that is that both the active and the middle are very old and long antedate the passive.

(d) THE SO-CALLED "DEPONENT " VERBS. These call for a word (cf. ch. XVII, (k)) at the risk of trespassing on syntax. Moulton205 is certainly right in saying that the term should be applied to all three voices if to any. The truth is that it should not be used at all. As in the Sanskrit206 so in the Greek some verbs were used in both active and middle in all tenses (like lu,w); some verbs in some tenses in one and some in the other (like bai,nw,


bh,somai); some on one voice only (like kei/mai). As concerns voice these verbs were defective rather than deponent.207 Note also the common use of the second perfect active with middle verbs ( i,nomai ge,gona).208 A number of verbs sometimes have the future in the active in the N. T. which usually had it in the middle in the older Greek. These are: avkou,sw (Jo. 5:25, 28, etc., but avkou,somai, Ac. 17:32), a`marth,sw (Mt. 18:21), avpanth,sw (Mk. 14:13), a`rpa,sw (Jo. 10:28), ble,yw (Ac. 28:26), gela,sw (Lu. 6:21), diw,xw (Mt. 23:34), zh,sw (Jo. 5:25), evpiorkh,sw (Mt. 5:33, LXX), klau,sw (Lu. 6:25), kra,xw (Lu. 19:40), pai,xw (Mk. 10:34), r`eu,sw (Jo. 7:38), siwph,sw (Lu. 19:40), spouda,sw (2 Pet. 1:15), sunanth,sw (Lu. 22: 10). But still note avpoqanou/mai e;somai zh,somai qauma,somai lh,myo mai o;yomai pesou/mai pi,omai te,xomai fa,gomai feu,xomai etc. Cf. Blass, Gr. of N. T. Gk., p. 42 f.; Winer-Schmiedel, p. 107; Moulton, Prol., p. 155. See Helbing, Gr. d. Sept., p. 89 f.; Thackeray, pp. 231 ff., for illustrations in the LXX. The term "deponent" arose from the idea that these verbs had dropped the active voice. Verbs do vary in the use of the voices in different stages of the language.

(e) THE PASSIVE SUPPLANTING THE MIDDLE. In Latin the middle and passive have completely blended and the grammars speak no more of the Latin middle. Greek indeed is the only European speech which retains the original middle form and usage.209 In fact, when we consider other tongues, it is not strange that the passive made inroads on the middle, but rather that there was any distinction preserved at all.210 In most modern languages the middle is represented only by the use of the reflexive pronoun. The Greek itself constantly uses the active with reflexive pronoun and even the middle. Jannaris211 has an interesting sketch of the history of the aorist and future middle and passive forms, the only forms where the two voices differ. As already remarked, the old Greek as in Homer212 did not distinguish sharply between these forms. In Homer the middle is much more common than in later Greek,213 for the passive has no distinct form in the future and not always in the aorist. In the modern Greek the middle has no distinctive form save lu,sou (cf. lu/sai)


Addenda 3rd ed.

and this is used as passive imperative second singular.214 Elsewhere in the aorist and future the passive forms have driven out the middle. These passive forms are, however, used sometimes in the middle sense, as was true of avpekri,qh, for instance, in the N. T. The passive forms maintain the field in modern Greek and appropriate the meaning of the middle. We see this tendency at work in the N. T. and the koinh, generally. Since the passive used the middle forms in all the other tenses, it was natural that in these two there should come uniformity also.215 The result of this struggle between the middle and passive in the aorist and future was an increasing number of passive forms without the distinctive passive idea.216 So in Mt. 10:26 ( mh. fobhqh/te auvtou,j) the passive is used substantially as a middle. Cf. the continued use of timh,somai as future passive in the earlier Greek as a tendency the other way. The history of this matter thus makes intelligible what would be otherwise a veritable puzzle in language. Here is a list of the chief passive aorists in the N. T. without the passive idea, the so-called "deponent" passives: avpekri,qhn (Mt. 25:9 and often, as John, Luke chiefly having Attic avpekri,nato also, Ac. 3:12), diekri,qhn (Ro. 4:20), sunupekri,qhn (Gal. 2:13), avpelogh,qhn (Lu. 21:14, but see 12:11), hvgallia,qhn (Jo. 5:35), evgenh,qhn (Mt. 6:10, but also evgeno,mhn often, as Ac. 20:18); cf. ge,gona and gege,nhmai, evdeh,qhn (Lu. 5:12); hvge,rqhn (Lu. 24:34), hvduna,sqhn (Mk. 7:24, as New Ionic and LXX) and hvdunh,qhn (Mt. 17:16), diele,cqhn (Mk. 9:34), evqauma,sqhn (Rev. 13:3, but passive sense in 2 Th. 1:10), evqambh,qhn (Mk. 1:27), evnqumhqei,j (Mt. 1:20), metemelh,qhn (Mt. 21: 32), evfobh,qhn (Mt. 21:46), euvlabhqei,j (Heb. 11:7), etc. For the LXX usage see Thackeray, p. 238. The future passives without certain passive sense are illustrated by the following: avnakliqh,so mai (Mt. 8:11), avpokriqh,somai (Mt. 25:37), evpanapah,setai (Lu. 10:6), qaumasqh,somai (Rev. 17:8), koimhqh,somai (1 Cor. 15:51), evntraph,sontai (Mk. 12:6), metamelhqh,somai (Heb. 7:21), fanh,somai (Mt. 24:30), fobhqh,somai, (Heb. 13:6). But we have genh,somai, dunh,somai evpimelh,somai poreu,somai. Cf. Blass, Gr. of N. T. Gr., p. 44 f.; Winer-Schmiedel, p. 108. For the rapid development of this tendency in later Greek see Hatzidakis, Einl., p. 192 f. See Helbing, Gr. d. Sept., pp. 97-100, and Thackeray, p. 240 f., for similar phenomena in the LXX. These so-called deponents appear in modern Greek (Thumb, Handb., p. 113). Cf. ch. XVII, iv, (e) .


Addenda 3rd ed.

(f) THE PERSONAL ENDINGS. They are probably pronominal,217 though Brugmann218 does not consider the matter as clear in all respects. One point to note is the heavy burden that is placed upon these endings. They have to express voice, person and number, everything in truth that has to do with the subject. Mode and tense are indicated otherwise. There was a constant tendency to slough off these personal endings and get back to the mode and tense-stems. Hence di,dwmi becomes di,dw (papyri) in late Greek. Le,gw was originally le,gomi.219

(g) CROSS-DIVISIONS. These personal endings have two cross-divisions. The active and middle have a separate list, the passive having none of its own. Then there is another cleavage on the line of primary and secondary tenses in the indicative, i.e. the unaugmented and the augmented tenses. The subjunctive mode falls in with the primary endings and the optative uses the secondary endings. But the first person active singular of the optative has one primary ending (as lu,oimi).220 But may it not be a reminiscence of the time when there was no distinction between subjunctive and optative? The imperative has no regular set of endings, as has already been shown, and does not fall in with this development, but pursues a line of its own. As a matter of fact the imperative always refers to the future.

(h) THE ACTIVE ENDINGS. They have received some modification in the N. T. Greek. The imperative can be passed by as already sufficiently discussed. The disappearance of the -- mi forms in favour of the - w inflection has been carefully treated also, as avfi,omen (Lu. 11:4). The subjunctive doi/ and optative dw|,h have likewise received discussion as well as the optative - ai and - eie. But some interesting points remain.

The use of - osan instead of - on is very common in the LXX (as Jer. 5:23, 26) and was once thought to be purely an Alexandrian peculiarity (Simcox, Language of the N. T., p. 37). For the LXX phenomena see Helbing, Gr. d. Sept., pp. 65-67; Con. and Stock, Sel. from the LXX, p. 32 f. The LXX is the principal witness to the - osan forms (Thackeray, Gr., p. 195), where they


are exceedingly frequent (ib., pp. 212 ff.). It is not so abundant outside of the LXX, but the Boeotians used it for the imperfect and optative.221 Mayser222 has found more examples of it in the Tebtunis Papyri, both aorist and imperfect, than Moulton223 had discovered. The inscriptions also show it.224 In the N. T. the contract verb evdoliou/san (Ro. 3:13) is a quotation from the LXX. In Jo. 15:22, 24, the imperfect ei;cosan has to be admitted. In 2 Th. 3:6 parela,bosan is read by aAD and W. H. put it in the margin. The text parela,bete is supported by BFG. This use of the - mi inflection may be compared with the use of twsan in the imperative. In the modern Greek it is common with contract verbs (cf. LXX) like evdoliou/san above. The modern Greek evrwtou/sa is a new formation (Thumb, Handb., p. 171) modelled after it.

Blass225 needlessly hesitates to accept - an in the present perfect instead of the usual - asi, and even Moulton226 is reluctant to admit it for Paul and Luke, preferring to regard it "a vulgarism due to the occasional lapse of an early scribe." It is certainly not a mere Alexandrianism as Buresch227 supposed. The ending - anti in the Doric usually dropped n and became - asi in Attic, but the later Cretan inscriptions show - an after the analogy of the aorist.228 The Alexandrian koinh, followed the Cretan. The papyri examples are very numerous229 and it is in the inscriptions of Pergamum230 also. Hort (Notes on Orthography, p. 166) considers it "curious," but has to admit it in various cases, though there is always some MS. evidence for - asi. Thackeray (Gr., pp. 195, 212) thinks that in some instances an with the perfect is genuine in the LXX. The earliest examples are from Lydia, parei, lafan (246 B.C.) and avpe,stalkan (193 B.C.). Cf. Dieterich, Unters., p. 235 f. The N. T. examples are avpe,stalkan (Ac. 16:36), ge,go


Addenda 3rd ed.

nan (Ro. 16:7; Rev. 21:6), e;gnwkan (Jo. 17:7), ei;rhkan (Rev. 19: 3), eivselh,luqan (Jas. 5:4), e`w,rakan (Lu. 9:36; Col. 2:1), pe,ptwkan (Rev. 18:3), teth,rhkan (Jo. 17:6). On the other hand the Western class of documents ( aADN Syr. Sin.) read h[kasin in Mk. 8:3 instead of eivsi,n. But it is in the LXX (Jer. 4:16), and Moulton231 finds h[kamen in the papyri. The form of h[kw is present, but the sense is perfect and the k lends itself to the perfect ending by analogy.

Another ending that calls for explanation is the use of - ej instead of - aj in the present perfect and the first aorist (in - ka especially). Hort considers the MS. evidence "scanty" save in Revelation. The papyri give some confirmation. Moulton232 cites avfh/kej e;grayej, etc., from "uneducated scribes" and thinks that in Revelation it is a mark of "imperfect Greek." Deissmann233 finds the phenomenon common in a "badly written private letter" from Fayum. Mayser234 confirms the rarity of its occurrence in the papyri. In the inscriptions Dieterich235 finds it rather more frequent and in widely separated sections. In Mt. 23:23 B has avfh,kete; in Jo. 8:57 B has e`w,rakej; in Jo. 17:7 and in 17:8 B has e;dwkej; once more in Ac. 21:22 B gives evlh,luqej.236 It will hardly be possible to call B illiterate, nor Luke, whatever one may think of John. D has avpeka,luyej in Mt. 11:25.237 W. H. accept it in Rev. 2:3 ( kekopi,akej), 2:4 ( avfh/kej), 2:5 ( pe,ptwkejgrk) 11:17 ( ei;lhfej), all perfects save avfh/kej. It is rare in the LXX (Thackeray, Gr., p. 215); found in A (Ex. 5:22, avpe,stalkej) and in e;dwkej (Ezek. 16:21; Neh. 9:10). The modern Greek has it as in e;desa ej (Thumb, Handb., p. 152).

We have both h=sqa (Mt. 26:69) and is (Mt. 25:21). The form in - qa is vanishing (Schweizer, Perg. Inschr., p. 166). Cf. also Mayser, Gr. d. griech. Pap., p. 321. The papyri have (acts, as N. T., and e;fhj. But see - mi Verbs.

Much more common is the use of the first aorist endings - a, - aj, etc., with the second aorist stem and even with the imperfect. This change occurs in the indicative middle as well as active. This matter more technically belongs to the treatment of the


aorist tense, as the - a is part of the tense-stem, but it is also conveniently discussed here. The Attic already had ei=pa e;pesa h;negka. The Attic inscriptions indeed show e;sca eu`ra,mhn and even the imperfects h;lpiza e;fera.238 This tendency towards uniformity spread in the koinh, somewhat extensively.239 Moulton240 finds the strong aorists with -- a chiefly in "uneducated writing" in the papyri, but common in general. This process of assimilation of the strong with the weak aorist was not yet complete.241 Blass242 thinks it an "intermediate" form already in the ancient Greek which spread in the koinh,. Cf. the liquid form h;ggeila. But both the strong and the weak aorists appear in the N. T. Thackeray (Cr., p. 195; cf. also pp. 210 ff.) notes that the -- an termination was finally extended to all past tenses, though in the LXX the imperfect forms are due to later copyists. In the modern Greek we note it regularly with kate,laba h;qela ei=ca, etc. (Thumb, Handb., pp. 152, 160, etc.). Hort243 has a detailed discussion of the matter in the N. T. This mixture of usage is shown in ei=pa and ei=pon. The - a form is uniform with endings in - t $ei;pate eivpa,tw eivpa,twsan). Both eivpo,n and eivpe, occur. We have avpeipa,meqa (2 Cor. 4:2) and proei,pamen (1 Th. 4:6). The participle is usually - w,n, but sometimes ei;paj. Both ei=paj and ei=pej ei=pon and ei;pan meet us. We always have the h;negka inflection save in the infinitive and the imperative. And even here we once have avne ne,gkai (1 Pet. 2:5) and once also prose,negkon (Mt. 8:4 BC). So also with e;pesa we have the weak or first aorist inflection in the indicative and imperative plural pe,sate (Lu. 23:30; Rev. 6:16). But in these two examples Hort244 (against W. 1-1.) favours pe,sete on MS. grounds ( aABD, aBC). In Lu. 14:10; 17:7 avna,pese is correct. The other forms that are accepted by W. H. are e;balan


once (Ac. 16:37); evpe,balan twice (Mk. 14:46; Ac. 21:27); ei=dan, ei;damen in a few places (Mt. 13:17; Lu. 10:24; Mt. 25:37, etc.); the indicatives avnei/lan (Ac. 10:39), avnei,late (Ac. 2:23), avnei,lato (Ac. 7:21), ei[lato (2 Th. 2:13), e`xeila,mhn (Ac. 23:27), evxei,lato (Ac. 7:10; 12:11); eu-ran once (Lu. 8:35, or avneu/ran eu[ramen once (Lu. 23:2), and eu`ra,menoj once (Heb. 9:12); the imperatives e;l qate evlqa,tw uniformly; both h=lqan and h=lqon, once avph/lqa (Rev. 10:9), regularly h;lqamen (Ac. 21:8). There are many other examples in various MSS. which W. H. are not willing to accept, but which illustrate this general movement, such as avpe,qanan (Mt. 8:32, etc.), e;laban (Jo. 1:12), evla,bamen (Lu. 5:5), evla,bate (1 Jo. 2:27), evxe,balan (Mk. 12:8), e;pian (1 Cor. 10:4 D), e;fugan (Lu. 8: 34 D), kate,fagan (Mk. 4:4 D), sune,scan (Ac. 7:57 D), gena,menoj (Lu. 22:41 a), etc. But let these suffice. Moulton245 is doubtful about allowing this - a in the imperfect. But the papyri support it as Deissmann246 shows, and the modern Greek247 reinforces it also as we have just seen. W. H. receive ei=can in Mk. 8:7; Ac. 28:2 ( parei/can); Rev. 9:8; ei;camen, in 2 Jo. 1:5. But D has ei=can in Jo. 15:22, 24; a has e;legan in Jo. 9:10; 11:36, etc. There is a distinct increase in the use of the sigmatic aorist as in h`ma,rthsa (Mt. 18:15), o;yhsqe (Lu. 13:28). It appears already in the LXX (Thackeray, Gr., p. 235). But see further under vii, (d).

The past perfect has the - ein forms exclusively as uniformly in the koi,nh.248 So ei`sth,keisan (Rev. 7:11), h|;deisan (Mk. 14:40), pe poih,keisan (Mk. 15:7). So the LXX. Cf. Helbing, Gr. d. Sept., p. 68. But the imperfect evxh|,esan (Ac. 17:15) is to be observed.

(i) THE MIDDLE ENDINGS. These call for less remark. bou, lei (Lu. 22:42) is the only second singular middle form in - ei, for o;yh| (Mt. 27:4) displaces o;yei. The inscriptions249 sometimes show bou,lh|. Blass250 regards bou,lei a remnant of literary style in Luke,


but the papyri also have bou,lei. The occasional use of du,nh| (Mk. 9:22 f.) has been discussed under - mi) Verbs. It appears only once in the LXX, but the "poetic and apparently Ionic" evpi,sth| is more frequent (Thackeray, Gr., p. 217). Cf. also ka,qou (Jas. 2:3) as LXX and ka,qh| (Ac. 23:3). On the other hand we have fa,gesai and pi,esai (Lu. 17:8). This revival of the use of - sai parallel with - mai, - tai in the perfect of vowel verbs in the vernacular amounts to a "new formation" in the view of Blass.251 So Moulton, Prol., p. 54 f. To call this revival a "survival" is "antediluvian philology." In the LXX pi,esai is universal and fa,gesai outside of the Pentateuch where fa,gh| holds on (Thackeray, p. 218). The - sai form is universal in modern Greek. The love of uniformity made it triumph. But see Contract Verbs for further discussion. The middle form h;mhn (Mt. 25:35) and h;meqa (Mt. 23:30) is like the koinh, generally and the modern Greek ei=mai. Cf. also e;somai. For evxe,deto (Mt. 21:33) with loss of root o and w inflection (thematic e) see - mi Verbs. Cf. also evxekre,meto (Lu. 19:48). The LXX has -- ento for - onto (Thackeray, p. 216).

(j) PASSIVE ENDINGS. As already observed, the passive voice has no distinctive endings of its own. The second aorist passive, like evfa,nhn, is really an active form like e;bhn ( evfa,nhn, is the proper division).252 Cf. Latin tace-re. So evca,rhn from caire,w. The first aorist in - qhn seems to have developed by analogy out of the old secondary middle ending in - qhj $evdo,qhj) parallel with so (Sanskrit thas).253 The future passive is a late development and merely adds the usual soe and uses the middle endings. The ending in - qhn is sometimes transitive in Archilochus,254 as the middle often is, and perhaps helps to understand how in the koinh, these forms (first aorist passive) are so often transitive ("deponents") as in avpekri,qhn evfobh,qhn, etc. The second aorist passive as noticed above is really an active form. So the passive forms have a decidedly mixed origin and history. There is nothing special to note about these passive endings in the N. T. save the increased use of them when even the passive idea does not exist. In some verbs s is inserted contrary to Attic practice. So ke,k leistai (Lu. 11:7), le,lousmai (Heb. 10:22). It is a common usage in the LXX (Thackeray, Gr., pp. 219 ff.). See also VII,


(g), 9. In Rev. 8:12; 18:23, W. H. print fa,nh| (first aorist active, cf. evpifa/nai in Lu. 1:79) rather than the passive fanh|/. Note evkfu,h| (Mt. 24:32, but Rec. evkfuh|/, though evkfu,h| in Mk. 13:28), sunfuei/sai (Lu. 8:7) and pareisedu,hsan (Ju. 4) for e;dun (Rec. Mk. 1:32) which the LXX retains (Thackeray, Gr., p. 235). In the LXX, when a verb had both first and second aorist passive forms, the first disappeared (ib., p. 237). But see (d), for further discussion.

(k) CONTRACT VERBS. The use of - sai was mentioned above. It appears255 in kauca/sai (1 Cor. 4:7; Ro. 2:17, etc.) and ovduna/sai (Lu. 16:25) where ae regularly contracts into a. See cari,esai (=- ei/sai) P. Oxy. 292 (A.D. 25).

Verbs in - aw. The confusion with verbs in - ew is already seen in the Ionic (Herodotus). The LXX in general preserves the distinction between - aw and - ew verbs, but aAB occasionally have the confusion (Thackeray, Gr., p. 241). In the modern Greek the blending is complete. One conjugation is made up, some forms from - aw, some from - ew (Thumb, Handb., p. 169 f.). The N. T. MSS. vary. W. H. receive hvrw,toun in Mt. 15:23 ( aBCD), but hvrw,twn in Mk. 4:10 though - oun, is here supported by aC and by single MSS. elsewhere. Hatzidakis (Einl. in d. Neug., p. 128 f.) considers hvrw,toun due to Ionic influence. In Mt. 6:28 we have kopiou/sin, LP in B 33, but W. H. reject256 it, as they do nikou/nti in Rev. 2:7, 17; 15:2, and katege,loun (Lu. 8:53). In Mk. 14:5 W. H. read evnebrimw/nto ( aC - ou/nto) and in Jo. 11:38 evmbrimw,menoj ( aA - ou,menoj). So there is a variation as to h`ttw/ntai (2 Pet. 2:20) from h`tta,omai and h`ssw,qhte (2 Cor. 12:13) from e`sso,w after the analogy of evlasso,w.257 W. H. print zh|/n (Ro. 8:12). This is a matter of much dispute with the editors, but it is more than doubtful if W. H. are correct. On the other side see WinerSchmiedel258 and Moulton.259 But both - za,w (Ro. 8:12) and cra,o mai (1 Tim. 1:8) have the 17 contraction rather than a (-- hw verbs, Moulton, Prol., p. 54). In Ro. 7:9 B even has e;zhn for e;zwn. But the koinh, uses cra/sqai, though not in the N. T.260 Paul


has crh/tai (pres. subj.) in 1 Tim. 1:8. Elsewhere also the a forms prevail in the koinh, as in diya/n and peina/n. So peina|/ (1 Cor. 11:21), peina/n (Ph. 4:12), diya|/, (Ro. 12:20) as subjunctive (so peina|/ same verse). The LXX keeps Attic zh/n and crh/sqai, but diya/n and peina/n (Thackeray, Gr., p. 242).261

Verbs in -- ew sometimes show forms in - aw. So evllo,ga in Phil. 1:18, evlloga/tai in Ro. 5:13, evlea/te in Ju. 22, 23, , and evlew/ntoj in Ro. 9:16, but evleei/ in Ro. 9:18. LXX has both forms. The koinh, usually has the - ein/ forms.262 For further examples of this confusion between - aw and - ew in LXX and isolated N. T. MSS. see Winer-Schmiedel.263 In 1 Cor. 11:6 all editors print xura/sqai (cf. kei,rasqai just before), though in 1 Cor. 11:5 evxurhme,nh and xurh, sontai (Ac. 21:24) probably come from xure,w.264 Cf. eva,w eva,sw.265

Contraction does not always take place with ee in verbs in - ew. In Lu. 8:38 W. H. follow BL in giving evdei/to but Hort266 admits that it is not free from doubt. Blass267 and Moulton268 consider evde,eto correct and the contraction a mere correction, and it is supported by the LXX and papyri. AP even have evdei/to. In Rev. 16:1 evkce,ete is undoubtedly right and evxe,ceen, in 16:2, but note evkcei/tai (Mt. 9:17). In Mk. 14:3 kate,ceen is to be noticed also (cf. Attic aorist). On the other hand in Jo. 3:8 note pnei/ evxe,plei (Ac. 18:18), plei/n avpoplei/n (Ac. 27:1 f.). In the LXX these words appear now one way, now the other.269 De,w ('to bind'), r`e,w have no ee forms in the N. T. W. H. accept in text only evxouqene,w in all the dozen examples in the N. T. (as Lu. 18:9, evxouqenou/ntaj), but in Mk. 9:12 they have d instead of q.270 Observe also avfe,wn tai (Lu. 5:20, etc.) instead of avfw/ntai or the regular avfei/ntai. In the N. T., W. H. give evrre,qh (Gal. 3:16; Mt. 5:21, etc.), but Hort271 thinks the Attic evrrh,qh should appear always in Matthew.

Verbs in - ow have two knotty problems. In Gal. 4:17 zhlou/te and 1 Cor. 4:6 fusiou/sqe are regular if indicative. But if they are subjunctive, the contraction oh is like the indicative oe (cf. indica-


tive and subjunctive of - ow verbs). So Blass272 and Moulton.273 Hort274 doubts the indicative here. If euvdw/tai (1 Cor. 16:2) be regarded as a present subjunctive no problem in contraction is raised.275 But in Col. 4:17 we have the subjunctive in i[na plh roi/j as in Attic for both indicative and subjunctive. In Ro. 3:13 evdoliou/san is the common LXX form in - osan. The other point is the infinitive in - ou/n, or - oi/n. W. H. give - oi/n for this infinitive everywhere except plhrou/n in Lu. 9:31.276 Cf. and - h|/n in W. H. Blass277 considers the - oi/n termination "hardly established for the N. T." since even in the N. T. the evidence is "small," though "of good quality " Hort contends.278 In Mt. 13: 32 kataskhnoi/n is supported by BD (in Mk. 4:32 by B), in 1 Pet. 2:15 fimoi/n has a, and in Heb. 7:5 avpodekatoi/n has BD. Moulton279 finds no support earlier in date than B save one inscription cited in Hatzitiakis (Einl., p. 193) and one papyrus of second century A.D. Mayser280 likewise finds no infinitive in - oi/n till after first century and gives some references for this late infinitive form. It looks as if the case will go against W. H. on this point. The form is probably due to some late grammarian's refinement and is linguistically unintelligible.

Piei/n, is often contracted (sounded finally ii, then i% into pei/n (so W. H., Jo. 4:7, 9, etc.) and in some MSS. ( a 8/9 times) into pi/n. But piei/n is the Syrian reading (Mt. 20:22, etc.).281 Contraction in - aw ew ow verbs, of course, takes place only in the present, imperfect and present participle.

VII. The Tenses ( cro,noi).

(a) THE TERM TENSE. It is from the French word temps, 'time,' and is a misnomer and a hindrance to the understanding of this aspect of the verb-form. Time does come finally to enter relatively into the indicative and in a limited way affects the optative, infinitive and participle. But it is not the original nor the general idea of what we call tense.282 Indeed it cannot be shown of


any verb-form that it had originally any reference to time. We must therefore dismiss time from our minds in the study of the forms of the tenses as well as in the matter of syntax. It is too late to get a new name, however.

(b) CONFUSION IN NAMES. The greatest confusion prevails in the names given to the various tenses. The time idea appears in the names present, past perfect and future. The state of the action rules in the names aorist, imperfect and perfect. Thus it is clear that the time idea did not prevail with all the names that the grammarians used. In the indicative, indeed, in the past three tenses appear, in the present two, in the future one (sometimes two). In the other modes as a rule only three tenses are found; in truth, in the subjunctive, optative and imperative practically only two are in common usage, the aorist and the present.

As a matter of fact there are nine possible tenses for each voice in the indicative: the aorist present, the imperfect present, the perfect present, the aorist past, the imperfect past, the perfect past; the aorist future, the imperfect future, the perfect future. These ideas do occur. In the past the distinction is clear cut. In the present no sharp line is drawn between the aorist and durative (unfinished or imperfect) save when the periphrastic conjugation is used or when Aktionsart comes in to help out the word itself. In the future, as a rule, no distinction at all is made between the three ideas. But here again the periphrastic conjugation can be employed. As a rule the future is aoristic anyhow. For further discussion see Jannaris, Hist. Gk. Gr., p. 180; Farrar, Greek Syntax, p. 120, and the references there to Harris' Hermes, Harper's Powers of the Greek Tenses, and H. Schmidt's Doctrina Temporum Verbi Graeci et Latini. The modern Greek preserves as distinct forms the aorist, present, imperfect; the future, the perfect and past perfect using periphrastic forms. Mr. Dan Crawford reports 32 tenses for Bantu.

(c) THE VERB-ROOT. There were originally two types of verbroots, the punctiliar and the durative. The tense called aorist ( avo,ristoj, 'undefined action') is due to the use of the punctiliar verbs (the idea of a point on a line). The present tense comes out of the durative verb-root. But it is worth repeating that tenses are a later development in the use of the verb.283

Hence it was natural that some verbs never developed a present tense, like ei=don, and some made no aorist, like o`ra,w. The defective verbs thus throw much light on the history of the tenses.


Out of these two ideas grew all the tenses. Each language had its own development. Some aorists in Sanskrit had no presents, like the Greek ei=pon. Each tense in the Greek pursued its own way. It is a complex development as will be seen. The idea of comparing the aorist to a point and the present to a line is due to Curtius, but it has since been worked out at length.284 Instead of saying "irregular" verbs, Delbruck (Vergl. Syntax, Tl. II, p. 256) speaks of "several roots united to one verb."

This Aktionsart or kind of action belongs more specifically to syntax.285 But it is not possible to make a modern study of the tense formations without having clearly in mind this important matter. It will come out at every turn. Along with the various tense-suffixes which came to be used to express the tense-distinctions as they were developed there remains also the meaning of the verb-root itself. This is never to be left out of sight. Prepositions also enter into the problem and give a touch much like a suffix (perfective). So qnh,skein is 'to be dying' while avpoqanei/n is 'to die' and avpoteqnhke,nai is 'to be dead.' Cf. e;cei, and avpe,cei e;fagon and kate,fagon. But more of this in Syntax. The point here is simply to get the matter in mind.

(d) THE AORIST TENSE ( avo,ristoj cro,noj). It is not true that this tense was always the oldest or the original form of the verb. As seen above, sometimes a durative root never made an aorist or punctiliar stem. But the punctiliar idea is the simplest idea of the verb-root, with many verbs was the original form, and logically precedes the others. Hence it can best be treated first. This is clearer if we dismiss for the moment the so-called first aorists and think only of the second aorists of the -- mi form, the oldest aorists. It is here that we see the rise of the aorist. Henry286 has put this matter tersely: "The ordinary grammars have been very unfortunate in their nomenclature; the so-called second perfects are much more simple and primitive than those called first perfects; the same is the case with the second aorists passive as contrasted with the first aorists," etc. The same remark applies to second aorists active and middle. The non-thematic second aorists represent, of course,


the most primitive form. The survivals of these forms in the N. T. have been discussed under - mi Verbs. The difference between the strong aorist (both thematic and non-thematic) and similar presents is syntactical and not formal.287 The point is that the strong aorists and the corresponding presents represent the simple stem of the verb. Brugmann288 indeed treats them together. It is not possible to make an etymological distinction between the imperfects e;fhn e;grafon and the aorists e;sthn e;fugon. The imperfect, of course, differs from the present only in the augment and secondary endings.289 The kinship between the aorist and present stems is further shown in reduplication. Reduplication in the aorist, as h;gagon, is supposed to be originally causative.290 Cf. the use of it with inceptive presents like gi$g%nw,skw. The aorist was quite common in the older Sanskrit, but is rare in the later language.291 Cf. the blending of the aorist and the present perfect forms in Latin. The strong aorist (both non-thematic and thematic) is far more common in Homer than in the later Greek.292 Indeed in the modern Greek the strong aorist has wellnigh vanished before the weak aorist.293

As often, the grammars have it backwards. The so-called second is the old aorist, and the so-called first is the late form of the verb. This weak form of the aorist has a distinct tense-sign, s, the sigmatic aorist. The s (- sa) was not always used, as with liquid verbs,294 like e;steila. This sigmatic aorist appears also in the Sanskrit.295 The distinction was not always observed between the two forms, and mixed aorists of both kinds occur in Homer, like h;xonto h;neika.296 No wonder therefore that uniformity gradually prevailed at the expense of the strong aorist in two ways, the disuse of the strong aorist (so h=xa) and the putting of first aorist endings to the second aorist stems, as ei=pa e;sca.

The k aorists in the indicative ( e;dwka e;qhka h-ka) continued to hold their own and to be used usually in the plural also. An ex-


tension of this usage (after the analogy of the perfect) is seen in the Byzantine and modern Greek297 form evlu,qhka for evlu,qhn.

There is one more aorist form, the aorist passive. As already shown, the so-called second aorist passive (-- hn), like evfa,nhn evca,rhn, is merely the second aorist active.298 The so-called first aorist passive in -- qhn is a Greek creation after the analogy of the old Indo-Germanic.299 Homer makes little use of either of these passive aorists, but the second is the more frequent with him and the form in - qhn is very rare.300

If this emphasis upon the aorist forms seem unusual to modern students, they may be reminded that in English we have only two tenses (apart from the periphrastic conjugation) and that they are usually punctiliar, as "I sing," "I sang." One is a present aorist, the other a past aorist.301 We do not here enter into the Aktionsart of the aorist (whether ingressive, constative or effective).302 That belongs to syntax.

The inscriptions agree with the development shown above in the aorist and support the N. T. phenomena.303 Mayser304 gives a careful discussion of the papyri development. In brief it is in harmony with what has already been observed. The non-thematic strong aorist is confined to a few verbs like bh/nai gnw/nai dou/nai du/nai qei/nai pri,asqai sth/nai. The k aorists are used exclusively in both singular and plural. The thematic strong aorist is disappearing before the weak sigmatic aorist.

In the N. T. the k aorists e;dwka e;qhka avfh/ka occur always except that Luke Luke(1:2 in the literary introduction) has pare,dosan. Elsewhere evdwkate (Mt. 25:35), e;qhkan (Mk. 6:29), avfh,kate (Mt. 23:23), etc., and quite frequently.305 The LXX also nearly always has k with these aorists in the plural.306

The non-thematic aorists in the N. T. are not numerous. The list is found in the discussion of - mi verbs and includes avne,bhn e;gnwn e;sthn e;fhn wvna,mhn, and all the forms of dou/nai ei-nai and qei/nai save the indicative active.


Addenda 3rd ed.

Addenda 3rd ed.

The thematic strong aorist in the N. T. shows the two developments noted above. The use of - a instead of - on with the strong aorist-stem is very common. See this chapter, vi, (h), for N. T. list like e;balan, etc. The MSS. vary much in the matter.307 The other change is the increased use of the sigmatic aorist. Here again Blass308 has a careful presentation of the facts. vEbi,wsa (1 Pet. 4:2) is a case in point instead of the old Attic evbi,wn. So is evbla, sthsa (Mt. 13:26; Heb. 9:4; Jas. 5:18) rather than e;blaston. Both evga,mhsa (Mt. 5:32) and e;gnma (Mt. 22:25) occur. Cf. Helbing, Gr. d. Sept., p. 93 f., and Thackeray, Gr., pp. 233 ff., for LXX illustrations.

=Hxa occurs a few times instead of the common h;gagon, as evpa,xaj (2 Pet. 2:5), evpisuna,xai (Lu. 13:34). Blass justifies it as appearing at least in dialects, LXX and late writers.309 It is part of the tendency towards the sigmatic aorist. Likewise a`marth,sw is slipping in beside a`ma,rtw (Mt. 18:15; Ro. 5:14, 16, cf. verse 12). Blass finds it in Emped., LXX, Lob., Phryn., 732. W. H. accept e;dusen (Mk. 1:32 on the authority of BD ( aA, etc., e;du). Luke in Ac. 24:21 has the reduplicated aorist evke,kraxa like the LXX, but usually the N. T. has the late form e;kraxa as in Mt. 8:29 ( e;kraxan), though once the Attic avne,kragon appears (Lu. 23:18). Once Luke (Ac. 6:2) has katalei,yantaj, a form that Blass310 finds in Herm., Vis. VIII, 3. 5, and Mayser311 observes avnteilh/yai in the papyri.

;Oyhsqe (Lu. 13:28) finds a parallel in an old Homeric aorist wvya,mhn (Winer-Schmiedel, p. 109). In Rev. 18:14 the Text. Rec. (without any known authority) has an aorist form eu[rhsa. So in Jas. 4:13 some MSS. have evmporeusw,meqa. Indeed some verbs have dropped the strong aorist form entirely like bio,w blasta,nw evgei,ro mai ktei,nw. See careful discussion of Winer-Schmieclel, p. 109 f. MSS. frequently read dw,sh| dw,swmen, etc., as if from an aorist e;dwsa as Jo. 17:2; Rev. 4:9. Cf. Winer-Schmiedel, p. 120. Cf. Helbing, Gr. d. Sept., p. 90 f., for LXX examples that further parallel these illustrations.

Conversely is to be noted a new strong aorist avne,qalon (Ph. 4: 10) which Blass312 takes in a causative sense ( avneqa,lete to. u`pe.r evmou/ fronei/n).

Verbs in -- zw make the aorist both in s and x) Most of these


verbs have dental stems in Attic, but some have guttural. Hence the s forms prevail till to-day. The LXX agrees with the N. T. (Thackeray, Gr., p. 222 f.). So evnu,staxan (Mt. 25:5), evmpai/xai, (Mt. 20:19), evpesth,rixan (Ac. 15:32); but on the other hand evsth,risen (Lu. 9:51), h[rpasen (Ac. 8:39), h`rmosa,mhn, (2 Cor. 11:2), slapi,sh|j (Mt. 6:2).313 The tendency in the papyri and the inscriptions on the whole is towards the use of s and not x with the verbs in zw.314 Cf. bapti,zw logi,zomai nomi,zw, etc.

Like kale,w and tele,w315 we have e in evfore,samen (1 Cor. 15:49) and evrre,qh (Mt. 5:21), but euvfo,rhsa (Lu. 12:16), r`hqe,n (Mt. 1:22) and evpepo,qhsa (1 Pet. 2:2). Cf. also h|;nesa h;rkese evme,sai. Cf. evpei,nasa (Mt. 4:2), but diyh,sw, though D has - a- in Jo. 6:35 and a in Rev.

The liquid verbs in - ai,nw and - ai,rw generally retain a even when not preceded by e or i as in Attic. So evba,skana (Gal. 3:1); once ker danw/ (1 Cor. 9:21), elsewhere - hsa* evxeka,qara (1 Cor. 5:7); evleu,kanan (Rev. 7:14); evsh,mana (Rev. 1:1); evpifa/nai (Lu. 1:79). In Rev. 8:12 and 18:23 note fa,nh|, not fanh|/. The koinh, begins to use - ana and -- ara with all verbs, and it is well-nigh universal in modern Greek. The LXX agrees with the N. T. (Thackeray, Gr., p. 223). A few -- nha forms survive in modern Greek (Thumb, Handb., p. 140 f.).

The second aorist passive has a few late developments of its own. This substitution of the second aorist passive for the first is a favorite idiom in the N. T.316 The koinh, shows likewise fondness for the - hn formations.317 This is true of the inscriptions' and the papyri.318 This development is directly the opposite of that in the case of the second and first aorist active and middle. It has already been observed that in Homer the passive aorist is very rare. Perhaps the increase in the use of -- hn forms is partly due to the general encroachment of aorist passive forms on the middle, and this is the simplest one. The Attic, of course, had many such forms also. Here are the chief N. T. examples: hvgge,lhn ( avp avn di kat Lu. 8:20, etc.) is in the LXX and the papyri; hvnoi,ghn (Mk. 7:35, etc.), but hvnoi,cqhsan also (Rev. 20:12); h`r pa,ghn (2 Cor. 12:2, 4), but the Attic h`rpa,sqh (Rev. 12:5); dio rugh/nai is read by some MSS. in Mt. 24:43; dieta,ghn (Gal. 3:19), u`peta,ghn (Ro. 8:20, etc.), but the Attic diatacqe,nta (Lu. 17:9 f.);


kateka,hn (Rev. 8:7; 1 Cor. 3:15), but Attic evxekau,qhsan (Ro. 1: 27); katenu,ghn (Ac. 2:37); evkru,bhn (Jo. 8:59). So also evfu,hn instead of e;fun follows the analogy of evrru,hn (Heb. 2:1) and evca,rhn (Lu. 22:5). Thus we have evkfuh|/ (Mk. 13:28)319 and sumfuei/sai (Lu. 8:6-8). Forms like evplh,ghn (Rev. 8:12) and evfa,nhn (Mt. 1:20) are Attic. On the other hand the poetical evkli,qhn (Mt. 14:19 avnakliqh/nai) has displaced the Attic evkli,nhn. vApekta,nqhn occasionally appears (as in Mk. 8:31 and Rev, six times) where the Attic would have avpe,qanon and evte,cqhn (Lu. 2:11) when the Attic would usually have evgeno,mhn. Both evgenh,qhn (Mt. 6:10 and often in 1 Th.) and evgeno,mhn (Mt. 7:28) are common, as hvdunh,qhn (Mt. 17:16) and hvduna,sqhn (Mk. 7:24). The many aorist passives in the deponent sense have already been noticed under VI, (e).

(e) THE PRESENT TENSE ( o` evnestw.j cro,noj). The present indicative, from the nature of the case, is the most frequent in actual use and hence shows the greatest diversity of development. Brugmann320 finds thirty-two distinct ways of forming the present tense in the Indo-Germanic tongues and thirty of them in the Greek. But some of these represent very few verbs and for practical purposes a much simpler classification is sufficient.321 Unfortunately the grammars by no means agree on the simplification. As samples see Giles, Man. of Comp. Philol., p. 425 f.; Hadley and Allen, p. 122 f.; Monro, Homeric Grammar, p. 9; Riemann and Goelzer, Phonetique, pp. 394 ff.; Kuhner-Blass, II, pp. 88 ff. In simple truth the facts are so varied that they lend themselves to many combinations more or less artificial. One of the most satisfactory is that of Monro, who has the historical instinct at least in his arrangement.

1. The Root Class. This is the simple non-thematic present like fhmi,. This is the logical one to put first, as with the aorist like e;bhn. This class is disappearing in the N. T. though duna mai eivmi, ei=mi in composition ( eis evx% ka,qhmai kei/mai kre,mamai appear.

2. The Non-Thematic Reduplicated Present. So di,dwmi i[h mi i[sthmi ki,crhmi ovni,nhmi pi,mplhmi ti,qhmi. It was never a very large class, but holds on in the N. T. And - w forms are common with these verbs.


3. The Non-Thematic Present with - na- and - nu-. So in the N. T. avmfie,nnumi avpo,llumi dei,knumi zeu,gnumi zw,nnumi kat a,gnumi kera,nnumi kore,0nnumi krema,nnumi mi,gnumi o;mnumi ph,g numi r`h,gnumi sbe,nnumi strw,nnumi, but these all have more commonly the - w forms.322

4. The Simple Thematic Present. So le,gw, lu,w. This was a constantly increasing class at the expense of the - mi verbs. It had several branches also including root-verbs like a;gw gra,fw, a strengthened vowel like pei,qw $piq% lei,pw $lip% feu,gw $fug% sh,pw th,kw trw,gw qli,bw pni,gw, etc., Hadley and Allen's "strong vowel class,"323 and the many contract denominative verbs like tima,w file,w avxio,w. But see the i. Class for these contract verbs. New verbs were added to this list from nouns and some also from verb-stems, grhgore,w from the old perfect evgrh,gora (this tense never in the N. T.),324 sth,kw (Mk. 11: 25) from e[sthka (modern Greek ste,kw).325 In Lu. 1:24 perie,kruben is probably imperfect, not aorist, from kru,bw $kru,ptw%. Cf. evkru,bhn.326 The LXX shows these new presents from perfect stems (Thackeray, Gr., p. 224 f.).

5. The Reduplicated Thematic Present. So gi,nomai $gi,gnomai * gige,nomai% pi,ptw (* pipe,tw), ti,ktw (* tite,kw), gn pt kt being weak forms of - gen pet tek) The N. T. has also ivscu,w from i;scw (* sise,cw).

6. The Thematic Present with a Suffix. There are five (- i, - n, -- sk - t q). Each of these divisions furnishes a number of verbs.

(a) The i class. It is very large. This suffix is used to make verbs from roots and substantives. It is probable327 that originally the suffix was - gi. It is thought that contract verbs in - aw ew, -- ow, etc., originally had this i as j or y which was dropped.328 It is thus the chief way of forming denominative verbs and is preeminently a secondary suffix.329 Some of these verbs are causative, some intensive, some desiderative.330 The special Greek desiderative in -- sei,w does not appear in the N. T., but forms like kopia,w are found. In particular, forms in - izw become so common that they no longer have an intensive, iterative or causative force,331


but are used side by side with the older form, as ba,ptw bapti,zw* r`ai,nw r`anti,zw, etc. In all the -- zw forms the i has united with a palatal (guttural) or lingual (dental), a matter determined by the aorist or future. So fula,ssw is from fula,kj w, fra,zw from fra,d j w. Other familiar combinations are i and l, as ba,lj w= ba,llw i with n by transposition, as fa,n-j wfai,nw, i with r likewise, as a;rj wai;rw. In kai,w and klai,w the u has dropped between a and i. In the N. T. verbs in - ai,nw, - ai,rw have - ana ara in the first aorist active as already shown under the aorist tense (d). vAmfia,zw (Lu. 12:28) is an example of a new present for avfie,nnumi. Cf. also avpoktennontwn (Mt. 10:28) in some MSS. for the older avpoktei,nw, - nnw, - nj w). See Blass332 for the variations in the MSS. at many places in the N. T. with this word. So evkcu,nnw (Mt. 26:28, etc.) in the best MSS. for evkce,w. Only in Mt. 9:17 we have evkcei/tai from evkce,w and in Rev. 16:1 evkce,ate333 in some MSS.

( b) The n class is also well represented in the N. T. with thematic stems. It takes various forms. There is the n alone, as ka,mnw an as a`marta,nw ne as avfikne,omai. Sometimes the n is repeated in the root, as lamba,nw $lab% manqa,nw $maq% tugca,nw $tuc%) In the koinh, (so LXX. and N. T.) this inserted n ( m) is retained in the aorist and future of lamba,nw $evlh,mfqhn lh,myomai) contrary to literary Attic. So the papyri.

( g) The sk class. It is commonly called Inceptive,334 but Delbruck335 considers these verbs originally terminative in idea, while Monro336 calls attention to the iterative idea common in Homer with the suffix - ske sko) The verbs with sk may be either without reduplication, as bo,skw qhn,skw i`la,skomai fa,skw, or with reduplication as gi$g%nw,skw dida,skw (for dida,cskw% mimnh,skw, pa,scw (for pa,qskw). Cf. avre,skw gami,skw ghra,skw eu`ri,skw mequ,skw. Reduplication is thus a feature with root-verbs (nonthematic) like di,dwmi and thematic like gi,$g%nomai as well as the sk class. For reduplication in the aorist and the perfect see (h). The iterative idea of some of these sk verbs suits well the reduplication.

(6) The t class. It is not a very numerous one (about 18 verbs), though some of the verbs are common. The verb has


always a labial stem like a[ptw ba,ptw tu,ptw. The root may end in b as in kalu,ptw, p as in tu,ptw, or f as in ba,ptw. It is even possible that pt may represent an original pj (cf. iota class).

( e) The q class. Cf. avlh,qw e;sqw knh,qw nh,qw in the present. The modern Greek has developed many new presents on the basis of the aorist or the perfect (Thumb, Handb., p. 143).

(f) THE FUTURE TENSE ( o` me,llwn cro,noj). The origin of this tense has given rise to much discussion and some confusion. Vincent and Dickson337 even say that the first aorist is derived from the s future! Like the other tenses there has been a development along several lines. No general remark can be made that will cover all the facts. As already remarked, the future tense is fundamentally aoristic or punctiliar in idea and not durative or linear. The linear idea can be accented by the periphrastic form, as e;sesqe lalou/ntej (1 Cor. 14: 9). Cf. also Mt. 24:9; Lu. 1:20; 5:10; Mk. 13:25. But as a rule no such distinction is drawn. The truth is that the future tense is a late development in language. In the Sanskrit it is practically confined to the indicative and the participle, as in the Greek to the indicative, infinitive and participle (optative only in indirect discourse, and rarely then, not at all in N. T.). And in the Rigveda the sya form occurs only some seventeen times.338 The Teutonic tongues have no future form at all apart from the periphrastic, which existed in the Sanskrit also.339 In the modern Greek again the future as a distinct form has practically vanished and instead there occurs qa, and the subjunctive or qe,lw and the remnant of the infinitive, like our English "shall" or "will."340 Giles341 thinks it uncertain how far the old Indo-Germanic peoples had developed a future.

Probably the earliest use of the future was one that still survives in most languages. It is just the present in a vivid, lively sense projected into the future. So we say "I go a-fishing" as Simon Peter did, u`pa,gw a`lieu,ein (Jo. 21:3). The other disciples respond evrco,meqa kai. h`mei/j su.n soi,. This usage belongs to the realm of syntax and yet it throws light on the origin of the future tense. So Jesus used (Jo. 14:3) the present and future side by side ( e;rco


mai kai. paralh,myomai). We have seen already that a number of aorists and presents like fhmi, had identically the same root and with no original distinction. That is, the durative idea was not distinguished from the aoristic or punctiliar. It is not strange, therefore, to see a number of these roots with primary endings (cf. subj. and opt. aorists) used as futures without any tense-suffix at all. Some were originally either present or future in sense (cf. e;rcomai above), others came to be used only as future. These verbs appear in Homer naturally, as bi,omai e;domai ei=mi pi,omai, etc.342 Cf. N. T. fa,gomai. It is possible that those with variable vowel like e;domai may really be the same form as the Homeric subjunctive (like i;omen as opposed to i;men%.343 Pi,omai is common in Attic. (N. T.) and is from aorist root ( e;pion). The form fa,gomai (LXX and N. T.) is analogous (aorist, e;fagon). The Attic used ce,w as future also, but LXX and N. T. have cew/ (Blass, Gr. of N. T. Gk., p. 42). Cf. Helbing, Gr. d. Sept., p. 88, for LXX illustrations to the same effect. The LXX has the classic e;domai; not in the N. T. (Thackeray, p. 231).

It used to be said that the a future was merely a variation of the Sanskrit sya, the y or j sound disappearing in the Greek. This gave a simple explanation of the s futures. But a rival theory has been advanced which derives the s future from the a aorist.344 The frequency of the aorist subjunctive in Homer with ke, $a;n) in principal clauses much like the future indicative in Attic, and the absence of a future passive, not to say future optative, in Homer give some colour to this contention.345 Thus dei,xw and the Latin dixo would be identical in form and meaning.346 But Brugmann347 has perhaps solved the problem by the suggestion that both explanations are true. Thus gra,yw he derives from the aorist subjunctive gra,yw, a mixed tense with a double origin. The use of - sioe in the Doric lends weight to the derivation of these verbs at least from the sya (Sanskrit) type.348 Hirt349 regards seoe (Doric) as a combination of the s future and the e future (liquid verbs, for instance) and considers it a new Greek formation. This Doric future therefore may be as old as any,


if not the oldest suffix, in fact the really distinctively future suffix. In the N. T. this Doric form survives in pesou/mai1 (Mt. 10:29). `Re,w has r`eu,sw (Jo. 7:38), klai,w has klau,sw (Lu. 6:25), while feu,gw has feu,xomai (Jo. 10:5). The other forms common in Attic have no future in the N. T. This mixed350 origin of the future (partly aorist subj., partly Indo-Germ. sio) shows itself in the Aktionsart of the tense. So Moulton notes proa,xw (Mk. 14:28) as durative, but a;xei (1 Th. 4:14) as aoristic. Cf. Thumb, Handb., p. 123.

Thus we may gain further light351 on the Ionic-Attic future of verbs in - izw) It is like the Doric - seoe. So we have - ise,w, dropping s we get - ie,wiw/) These verbs in are very common in the later Greek. In the N. T. the usage varies between this form of the future and the aoristic form in - soe. The LXX, like the Ptolemaic papyri (Thackeray, p. 228), has usually - iw/ in first singular and so metoikiw/ (Ac. 7:43) and parorgiw/ (Ro. 10:19), both quotations. Elsewhere W. H.352 prefer the forms in - i,sw, and Blass353 thinks that in the original passages of the N. T. the - i,sw forms are genuine. So the forms in - i,sei (like bapti,sei) are uniform in the N. T. (Lu. 3:16) save kaqariei/ (Heb. 9:14) and diakaqariei/, (Mt. 3:12).354 MSS. vary between avforiei/ and - i,sei fwtiei/ and - i,sei croniei/, and - i,sei. Cf. Blass.355 So in Eph. 6:8; Col. 3:25, the MSS. vary between komiei/tai and komi,setai. Some MSS. read komiou,menoi in 2 Pet. 2:13.356 All editors357 accept komiei/sqe in 1 Pet. 5:4. The active plural W. H.358 print as - iou/si always (as maka riou/sin, Lu. 1:48) save in gnwri,sousin (Col. 4:9).

The syncopated futures359 from the dropping of s do not survive in the N. T. in kale,sw tele,sw which always retain the s.360 So even avpole,sw (Mt. 21:41), though avpolw/ is common in the LXX and


is quoted once in the N. T. (1 Cor. 1:19). However, the middle avpolou/mai is the N. T. form (Lu. 5:37) like avpoqanou/mai. vElau,nw has no future in the N. T. The N. T., like the LXX, has a future form avfe,lw/ (Rev. 22:19) from the aorist ei=lon of aivre,w.

The liquid verbs in l, n r present few problems. They belong to the aorist subjunctive type of formation.361 Here again we have syncopation of the s. Verbs like ba,llw $balw/% me,nw $menw/% ai;rw $avrw/) form the future with the variable vowel oe added to the stem without a in the N. T. as in the earlier Greek.

Blass362 has shown that in the N. T. the future active has largely displaced the future middle with verbs that were defective in the active voice. These futures are as follows: a`marth,sw (Mt. 18:21), avpanth,sw (Mk. 14:13), a`rpa,sw (Jo. 10:28), ble,yw (Ac. 28:26), gela,sw (Lu. 6:21), diw,xw (Mt. 23:34), klau,sw (Lu. 6:25), kra,zw (Lu. 19:40 aBL), pai,xw (Mk. 10:34), r`eu,sw (Jo. 7:38), spouda,sw (2 Pet. 1:15), sunanth,sw (Lu. 22:10). We see this tendency already in the LXX (Thackeray, Gr., p. 231 f.). On the other hand the future middle alone occurs with avpoqanou/mai (Jo. 8:24), gnw,somai (1 Cor. 4: 19), lh,myomai (Mt. 10:41), o;yomai (Mt. 24:30), pesou/mai (Doric, Mt. 10:29), pi,omai (Mk. 10:39), fa,gomai (Lu. 14: 15), feu,xomai (Jo. 10:5). Carh,somai (Lu. 1:14) Blass363 regards as Attic future from the aorist ( evca,rhn) as compared with the future cairh,sw from the present. Both avkou,sw (Jo. 5:25) and avkou,somai, (Ac. 21:22, chiefly in the Acts) are found, and zh,sw (Jo. 5:25) and zh,somai (Jo. 11:25).

The so-called second future passive as seen in the case of carh,so mai above is really just the middle ending with s put to the aorist active stem. There is no difference in form or sense between bh,somai and stalh,somai save the - h which was really a part of the active stem of these verbs.364 The point is that fundamentally these so-called second future passives are really future middles corresponding to active aorists like the future middles and presents above ( lh,myomai, for instance). This point is made clearer by the fact that the Doric365 used only active endings like avnagra fhsei/ (not - etai). Homer, besides, only has one second future passive ( migh,somai, really middle) and none in - qhs--.366 Instead he uses the middle future as later Greek continued to do with verbs like timh,somai. Cf. genh,somai from evgeno,mhn. Some verbs indeed used both this second future passive like fanh,somai (Mt. 24:30) which


is punctiliar and fanou/mai (1 Pet. 4:18) which may be durative like the Attic as Moulton367 argues. So pau,sontai (1 Cor. 13:8) and evpanapah,setai (Lu. 10:6). Cf. also avnoigh,somai (Mt. 7:7), a`rpagh,somai (1 Th. 4:17), fanh,somai (Mt. 24:30), u`potagh,somai (1 Cor. 15:28), yugh,somai (Mt. 24:12), carh,somai (Lu. 1:14, see above).

The first future passive so-called is built upon the distinctively368 Greek aorist in - qh-. It is unknown to Homer, as stated above, and, like the second aorist passive, is aorist in origin and idea. Here again the Doric used the active endings369 like sunacqhsou/nti. This later form in - qhs-- grew continually in usage over the merely middle form like timh,somai. But the passive future did not always have the passive sense, as has been shown in the case of avnakiqh, somai (Mt. 8:11), avpokriqh,somai (Mt. 25:37), etc.370 vAnoicqh,somai also appears in Lu. 11:9 f. in some MSS. As an example of the usual forms in the N. T. take gnwsqh,somai (1 Cor. 14:7). Only mnhsqh,somai (not memnh,somai) and staqh,somai (not e`sth,xw) appear in the N. T.371

For a periphrastic future passive expressing continuance see e;sesqe misou,menoi (Mt. 10:22).372 This is naturally not a very common idiom for this tense, though the active periphrastic future is less infrequent as already shown.

(g) THE PERFECT TENSES ( te,leioi cro,noi).

1. The Name. It does fairly well if we do not think of time in connection with the tense, a mistake that Clyde makes.373 The completed state does not of itself have reference to present time. That comes later and by usage in the indicative alone in contrast to past and future. Originally the perfect was merely an intensive or iterative tense like the repetition of the aoristic present.374

2. The Original Perfect. The Greek perfect is an inheritance from the Indo-Germanic original and in its oldest form had no reduplication, but merely a vowel-change in the singular.375 Indeed 191:6a (Sanskrit veda, Latin vidi, English wot) has never had reduplication.376 It illustrates also the ablaut from id to oid in the singular, seen in Sanskrit and Gothic also.377 Cf. Latin capio, cepi (a to e). Note also kei/mai in the sense of te,qeimai.


But the vowel-change characteristic of the original perfects is seen in other verbs which did use reduplication. Reduplication will receive separate treatment a little later, as it pertains to the present and aorist tenses also. It may be here remarked that the reduplicated form of some iterative presents doubtless had some influence in fastening reduplication upon the perfect tense. Note the English "mur-mur " (Greek goggu,zw avrari,skw), where the syllable is doubled in the repetition. It was a natural process. A number of these reduplicated forms with the mere change in the vowel appear in the N. T. This so-called second perfect, like the second aorist, is a misnomer and is the oldest form.378 In Homer indeed it is the usual form of the perfect.379 These old root-perfects, old inherited perfect forms according to Brugmann,380 persist in the koinh, and are reasonably common in the papyri,381 the inscriptions382 and the N. T. They are of two classes: (1) real mi perfects without any perfect suffix, like e`sta,nai (Ac. 12:14); (2) second perfects in - a, like ge,gona le,loipa. As N. T. examples may be mentioned avkh,koa (Ac. 6:11), ge,gona (1 Cor. 13:1)), ei;wqa (Lu. 4:16), ge,graqa (Jo. 19:22), oi=da (Jo. 10:4), o;lwla ( avp, Mt. 10:6), etc. These forms are found in the LXX. Cf. Helbing, Gr. d. Sept., p. 103; Thackeray, Gr., p. 252 f. But the koinh, gave up the shorter (without - a) forms of the plural indicative active perfect of i[sthmi ( e[stamen e[state e`sta/sin). See this chapter, iv, (d), 3, for details.

3. The k Perfect. This is a new type created by the Greek language of which no adequate explanation has yet been offered. The Attic inscriptions already had the k form (Meisterhans, p. 189 f.). It is apparently at first in the singular, as in e[sthka (pl. e[stamen), etc.383 One might think that just as h[kw has a perfect sense like kei/mai and finally had a few perfect forms384 (like h[kasin), so by analogy some k verbs became the type and analogy did the rest. But Giles385 observes that the stems of the twelve or fourteen k perfects in Homer all end in a vowel, a liquid or a nasal, not one in k. And then the


three k aorists ( e;dwka e;qhka h-ka) call for explanation. But per contra there are some perfects in Homer which have k stems like de, dorka e;oika te,thka, etc. So that after all analogy may be the true explanation of the k perfects which came, after Homer's time, to be the dominant type in Greek. But the - ka perfects are rare in Homer. The examples are so common ( de,dwka, etc.), in the koinh, as in the classic Greek, as to need no list. Note e[sthka intransitive and e[staka transitive.

4. The Aspirated Perfects. They are made from labials and palatals ( f, c) and are absent from Homer. Even in the early classical period they are confined to pe,pomfa and te,trofa.386 Homer did use this aspirate in the peculiar middle form like tetra fatai.387 He has indeed te,trofa from tre,fw388 and probably just here, we may see the explanation by analogy of te,trofa from tre,pw and so of all the aspirated forms.389 An important factor was the fact that k g c were not distinguished in the middle perfect forms. As a N. T. example of this later aspirated perfect take prosenh,noca (Heb. 11:17). Cf. also ei;lhfa pe,praca te,taca.

5. Middle and Passive Forms. It is only in the active that the perfect used the k or the aspirated form ( f c). We have seen already that in the koinh, some active perfect forms drop the distinctive endings and we find forms like e`w,rakan and e`w,rakej. Helbing (Gr. d. Sept., pp. 101-103) gives LXX examples of rootperfects like e;rrwga k perfects like te,qeika e[sthka and transitive e[staka, aspirated perfects like e;rrhca. The middle and passive perfects did use the reduplication, but the endings were added directly to this reduplicated stem as in le,lumai. On the history of the ending - ka see Pfordten, Zur Geschichte des griechischen Perfectums, 1882, p. 29.

6. The Decay of the Perfect Forms. In the Sanskrit the perfect appears in half the roots of the language, but in the later Sanskrit it tends more and more to be confused with the mere past tenses of the indicative (aorist and imperf.) and grows less common also.390 In the Latin, as is well known, the perfect and the aorist tenses blended. In vidi and dedi we see preserved391 the old perfect and in dixi we see the old aorist. The Greek of the Byzantine period shows a great confusion between the perfect and the aorist, partly due to the Latin influence.392 Finally


Addenda 3rd ed.

Addenda 3rd ed.

Addenda 3rd ed.

in the modern Greek vernacular the perfect form is lost save in the perfect passive participle like keklhme,noj. The perfect active is now made with e;cw and the passive participle ( e;cw deme,no) or with e;cw and a root similar to the third singular aorist subjunctive ( e;cw de,sei or de,sh|). Cf. Thumb, Handb., p. 161. The only k perfect in modern Greek is eu[rhka, "the only certain remnant of the ancient perfect" (ib., p. 148). Cf. e;ce me parh|thme,non (Lu. 14:18). Cf. also pepwrwme,nhn e;cete th.n kardi,an u`mw/n (Mk. 8 : 17). This is much like the English perfect in reality, not like the Greek e;cw and aorist participle (like e;cw avkou,saj). Cf. Sonnenschein, Greek Grammar, Syntax, 1894, p. 284. The perfect passive in modern Greek vernacular is formed like e;cw luqh/ (- ei) or lelume,noj ei=mai.393 But we are in no position to throw stones at the Greeks, for we in English have never had a perfect save the periphrastic form. How far the perfect and the aorist may have become confused in the N. T. in sense is a matter of syntax to be discussed later.394

7. The Perfect in the Subjunctive, Optative, Imperative. Hence the perfect is practically395 confined to the indicative. No example of the perfect optative occurs even in the periphrastic form. The subjunctive perfect, except the form eivdw/ $eivdh/te 1 Jo. 5 : 13), appears only in the periphrastic conjugation, of which a few examples remain. So the active, as h|= pepoihkw,j (Jas. 5:15), pepoiqo,tej w=men (2 Cor. 1:9), and the passive, as w=sin teteleiwme,noi (Jo. 17:23), h|= keklhme,noj (Lu. 14:8), h|= peplhrwme,nh (Jo. 16:24). So also Jo. 17: 19, 1 Cor. 1:10, etc. The imperative makes a little worse showing. We still have i;ste (Jas. 1:19; Eph. 5:5; Heb. 12:17 all possible indicatives), pefi,mwso (Mk. 4:39) and e;rrwsqe (Ac. 15:29). The periphrastic imperative perfect is also found as e[stwsan periezwsme,nai (Lu. 12:35). In simple truth, as previously remarked (see proof in Prof. Harry's articles), the perfect subjunctive, optative and imperative never had any considerable vogue in Greek, not as much as in Sanskrit. In Homer the perfect subjunctive active is more common than in later Greek, but it is rare in Homer.396

8. The Perfect Indicative. It is to the indicative that we turn


for the real development of the perfect. Here the perfect was for long very frequent indeed, and the time element comes in also. The ancients did not agree in the names for the three tenses of perfect action in the indicative. The Stoics397 called the present perfect sunteliko.j (or te,leioj) cro,noj evnestw,j, the past perfect sunteliko.j $te, leioj% cro,noj parw|chme,noj, the future perfect sunteliko.j $te,leioj% cro,noj me,llwn. Sometimes the present perfect was called merely o` para kei,menoj cro,noj, the past perfect o` u`persunteliko.j cro,noj and the future perfect o` met v ovli,gon me,llwn cro,noj (futurum exactum). The name plu-perfect is not a good one. The tense occurs in the N. T. with 22 verbs and 15 have the augment (H. Scott). Thus teqeme li,wto (Mt. 7:25) and evlhlu,qei (Jo. 6:17), but evbe,blhto (Lu. 16:20) and periede,deto (Jo. 11:44). Cf. ei=con avpokeime,nhn (Lu. 19:20) in the light of modern Greek. In the N. T. the past perfect is not very frequent, nor was it ever as abundant as in the Latin.398 It goes down as a distinct form with the present perfect in modern Greek. Hirt399 calls attention to the fact that Homer knows the past perfect only in the dual and the plural, not the singular, and that the singular ending is a new formation, a contraction of -- ea into - h. In the N. T., however, only - ein is used. It is not certain whether the past perfect is an original Indo-Germanic form. The future perfect was always a very rare tense with only two active forms of any frequency, e`sth,xw and teqnh,xw. The middle and passive could make a better showing. In Heb. 8:11 eivdh,sousin, is probably future active (from LXX),400 and in Lu. 19:40 some MSS., but not aBL (rejected by W. H.), give kekra,xontai (cf. LXX). In Heb. 2:13 (another quotation from the LXX) we have the periphrastic form e;somai pepoiqw,j. The future perfect passive occurs in the N. T. only in the periphrastic form in such examples as e;stai dedeme,non (Mt. 16:19), e;stai lelume,na (Mt. 18:18), e;sotnai diamemerisme,noi (Lu. 12 : 52). Cf. e;sh| kat[ a] teqeim[ e,] no$s% B.G.U. 596 (A.D. 84). In the nature of the case the future perfect would not often be needed. This periphrastic future perfect is found as early as Homer.401 The papyri likewise show some examples.402


Addenda 3rd ed.

The present perfect and the past perfect also have the periphrastic conjugation. So we find with comparative indifference403 e;stin gegramme,na (Jo. 20:30) and in the next verse ge,graptai. So also h=n gegramme,non (Jo. 19:19) and evpege,grapto (Ac. 17:23). Cf. also Lu. 2:26. The active has some examples also, though not so many, as e`stw,j eivmi (Ac. 25:10), and h=san proewrako,tej (Ac. 21:29).

9. S in Perfect Middle and Passive and Aorist Passive. It may be due to a variety of causes. Some of these verbs had an original s in the present stem, like tele,s%w avkou,$s%w. Hence tete,lesmai, h;kousmai ( hvkou,sqhn) etc.404 Others are dental stems like pei,qw pe, peismai. Others again are n stems which in Attic (apparently analogical) changed to s, as fai,nw pe,fasmai, but in the N. T. this n assimilates to the m as in evxhramme,noj (Mk. 11:20) from xhrai,nw, memiamme,noj (Tit. 1:15) from miai,nw. Then again some verbs take the s by analogy merely, as in the case of e;gnwsmai evgnw,sqhn (1 Cor. 13:12), ke,kleismai (Lu. 11:7), le,lousmai (Heb. 10:22).

(h) REDUPLICATION ( diplasiasmo,j or avnadi,plwsij).

1. Primitive. Now this primitive repetition of the root belongs to many languages and has a much wider range than merely the perfect tense. Hence it calls for separate treatment. It is older, this repetition or intensifying of a word, than either the inflection of nouns or the conjugation of verbs.405 Root reduplication existed in the parent language.406

2. Both Nouns and Verbs. Among nouns note avgwgo,j ba,r baroj be,bhloj, etc. But it was among verbs that reduplication found its chief development.407

3. In Three Tenses in Verbs. It is in the aorist, the present and the perfect. This is precisely the case with the Sanskrit, where very many aorists, some presents and nearly all perfects have reduplication.408 In Homer409 the reduplication of the second


Addenda 3rd ed.

aorist is much more frequent than in later Greek, but forms like h;gagon h;negkon ei=pon, persist in N. T. Greek and the koinh, generally. Cf. evke,kraxa in Ac. 24 : 21. The Greek present shows reduplication in three classes of presents, viz. the root class (like di,dwmi i[hmi i[sthmi, etc.), the thematic presents (like gi,gnomai pi,ptw, etc.), inceptive verbs (like gignw,skw, etc.). The most common reduplication in Greek is, of course, that in the perfect tense, where it is not like augment, mode-sign or personal endings. It is an integral part of the tense in all modes, voices and persons, until we see its disappearance (p. 365) in the later Greek. In the vernacular the extinction is nearly complete.410 Even presents411 like gnw,skw occur in modern Greek. Dieterich412 gives numerous examples of dropped reduplicatiion in inscriptions and papyri. It is absent in the modern Greek vernacular, even in the participle.413

4. Three Methods in Reduplication. Perhaps the oldest is the doubling of the whole syllable, chiefly in presents and aorists, like goggu,zw avrari,skw h;gagon, etc. This is the oldest form of reduplication414 and is more common in Greek than in Latin.415 The later grammarians called it Attic reduplication because it was less common in their day,416 though, as a matter of fact, Homer used it much more than did the Attic writers.417 But perfects have this form also, as avkh,koa evlh,luqa, etc. But the reduplication by i is confined to presents like di,dwmi gi,gnomai gignw,skw, etc. And most perfects form the reduplication with e and the repetition of the first letter of the verb as le,kuka. But Homer had pe,piqon and other such aorists. Ei=pon is really an example of such an aorist.

5. Reduplication in the Perfect. The history is probably as follows in the main. Originally there were some perfects without reduplication,418 a remnant of which we see in oi=da. The doubling of the whole syllable was the next step like avkh,koa evgrh,gora evlh,luqa avpo,lwla, etc., like the present and aorist usage.419 Then comes the e with repetition of the initial letter of a consonant-


stem like leloipa. But here some further modifications crept in. The aspirates did not repeat, but we have te,qeika. Those with a did not repeat it, but instead used the rough breathing as e[sthka or the smooth like e;schka. This was all for euphony. But forms like e;schka e;spasmai fall under another line also, for, if the verb begins with a double consonant, the consonant need not be used. So e;gnwka, but be,blhka ge,grafa. The Cretan dialect has indeed e;grattai= ge,graptai.420 So far the N. T. phenomena are in harmony with the general Greek history, as indeed is the case with the papyri421 and the inscriptions.422 In Lu. 1:27 and 2:5, we have evmnhsteume,nh not memn. (cf. me,mnhmai). Just as s verbs did not repeat, so with r` verbs sometimes. So evrimme,noi (Mt. 9:36), e;rrwsqe (Ac. 15:29), etc. But in Rev. 19:13 W. H. read rverantisme,non, though Hort423 advocates rvaramme,non. D has rverimme,noi in Mt. 9:36 above. This reduplication of initial r` is contrary to Attic rule. For the LXX see Thackeray, Gr., p. 204 f. This use of e begins to spread in the koinh, and is seen in LXX MSS., as in A evpe,grapto (Deut. 9:10). For similar forms in Ionic and late writers see WinerSchmiede1.424 Once more several verbs that begin with a liquid have ei as the reduplication in the Attic and Ionic, though not in all dialects. Perhaps euphony and analogy entered to some extent in the case of ei;lhfa $lamba,nw% ei;rhka (cf. evrrh,qhn). Note also ei;lhca and ei;loca. With verbs beginning with a vowel there was sometimes the doubling of the syllable as avkh,koa or the mere lengthening of the vowel as h;kousmai, or the addition of e alone with contraction as eivqisme,noj, or uncontracted as e;oika $from ei;kw). Cf. ei;wqa. In Jo. 3:21 (so 1 Pet. 4:3) we have ei;rgasmai as in Attic and eivlkwme,noj in Lu. 16:20. In o`raw we have e`oraka in Paul's Epistles (1 Cor. 9:1) and sometimes a sort of double reduplication (like ei;wqa) as e`w,raka (Jo. 1:18). So Attic. See Additional Note. In Col. 2:1 the form e`o,rakan calls for notice both for its reduplication and its ending (cf. e`w,rakan Lu. 9: 36). So also avne,w|gen (1 Cor. 16:9; a hvnew|gw,j, Jo. 1 : 51) and avnew|gme,nh (2 Cor. 2:12). Indeed in this last verb the preposition may readditional reduplication (treble therefore), as in hvnew|gme,nh (Rev. five times). See also hvmfiesme,non (Mt. 11:8 = Lu. 7:25) from avmfie,nnumi. But as a rule with compound verbs in the N. T. re-


duplication comes only between the prepositions and the verb. Sometimes the reduplication is not used, as in euvaresthke,nai (Heb., 11:5), but aDEP have euvhr) We have w|vkoko,mhto (Lu. 4:29), but oivkodomh/sqai (Lu. 6:48).425 Cf. oivkodomh,qh (Jo. 2:20) for absence of augment. Reduplication in the perfect has disappeared from the modern Greek (Thumb, Handb., p. 119) and is showing signs of decay in the koinh,. For suppression of reduplication in papyri see Mayser, p. 341.

(i) AUGMENT ( au;xhsij).

1. The Origin of Augment. It has never been explained. It is generally conceded to be an independent word, an adverb, added to the verb, which is an enclitic after the augment like e;lipe.426 We have mere conjectures for the origin of the adverb, possibly a locative of the pronoun-stem. In Sanskrit it is a.

2. Where Found. It is found in Sanskrit, Iranian, Armenian and Greek, and only in the past tenses of the indicative. But in Mt. 12:20 we actually have katea,xei. (fut. ind. of kata,gnumi, and in Jo. 19:31 katearw/sin) (aor. pass. subj.), probably to distinguish these forms from kata,gw). So Winer-Schmiedel, p. 98. This "false augment" is very common in later Greek (Hatzidakis, Einl., p. 64). Augment persists in modern Greek (Thumb, p. 117).

3. The Purpose of Augment. It denotes past time. The secondary endings do that also and with sufficient clearness at first. More than half of the past tenses of the Sanskrit do not have the augment.427 In Homer some verbs like o`ra,w never had augment, and often for metrical reasons the augment is not found in Homer. He used much freedom in the matter.428 Jannaris429 is probably right in the opinion that this freedom is due to the original fulness of the verb-endings. Augment won a firm foothold in prose before it did in poetry,430 but never was everywhere essential. It varied greatly in its history as will be shown.

4. The Syllabic Augment ( au;xhsij sullabikh,). Its use with the past tenses of the indicative was not exactly uniform, being less constant with the past perfect than with the aorist and imperfect. The syllabic augment occurs also with some initial vowel verbs due to original digamma F, s in the anlaut. So ei;asen (Ac. 28:4),


ei;domen (Mt. 2:2), ei=pen (Mt. 2:8), ei[lato (2 Th. 2:13), etc. Cf. Thackeray, Gr., p. 200 f. In the N. T. it is absent from the past perfect more frequently than it is present, as is true of the papyri431 and late Greek generally.432 So, for instance, teqemeli,wto (Mt. 7:25), pepoih,keisan (Mk. 15:7), paradedw,kwisan (Mk. 15:10), evlhlu,qei. (Jo. 6:17), etc. On the other hand the augment does appear in such examples as evpepoi,qei (Lu. 11:22), evbe,blhto (Lu. 16:20), evgego,nei (Jo. 6:17), sunete,qeinto $Jo. 9:22), periede,deto (Jo. 11:44), etc. It was only in the past perfect that both augment and reduplication appeared. The koinh, strove to destroy the distinction between reduplication and augment so that ultimately reduplication vanished (Thumb, Hellenismus, p. 170). But first the augment vanished in the past perfect. The Attic sometimes had e`sth,kein (Winer-Schmiedel, p. 100). Hort (Notes on Orthography, p. 162) contends for i`sth,kein uniformly in the N. T. as more than mere itacism for ei`sth,kein, for even B has i five times in spite of its fondness for ei. So W. H. uniformly, as Rev. 7:11 and even in Jo. 1:35 and Lu. 23:49. Cf. similar itacism between ei=don and i;don in the MSS. (Hort, Notes on Orthography, p. 162). On augment in the LXX see Conybeare and Stock, Sel. from LXX, pp. 36 ff.; Swete, Intr. to 0. T., p. 305; Thackeray, Gr., pp. 195 ff. Syllabic augment was much more tenacious with the aorist and imperfect than the temporal.

5. The Temporal Augment ( au;xhsij cronkh,). The simplicity of the syllabic and the resulting confusion of the temporal had undoubtedly something to do with the non-use of the temporal augment in many cases.433 The koinh, shows this tendency.434 Even the Attic was not uniform in the use of the temporal augment. At bottom there is no real distinction between the temporal and syllabic augment. Both express time and both make use of the syllabic e. The difference is more one of the eye and ear than of fact. What we call the temporal augment is the result of the contraction of this e with the initial vowel of the verb.435 As remarked above, this very confusion of result, difficult to keep clear as the vowel-sounds tended to blend more and more, led to the disuse of this e and contraction with initial vowel verbs, especially with diphthongs.436 Hence in the N. T. we meet such examples as the


following: of ai evpaiscu,nqh (2 Tim. 1:16); of eu euvlo,ghsen (Mt. 14:19), euvdo,khsa (Mt. 17:5), euvnou,cisan (Mt. 19:12), euvkai,roun (Mk. 6:31), euvfrai,nonto (Ac. 7:41), euvporei/to (Ac. 11:29), euvqu dromh,samen, (Ac. 16:11), euvcari,sthsen (Ac. 27:35).437 But on the other hand we have hu[riskon (Mk. 14:55), proshu,xanto (Ac. 8:15), huvco,mhn (Ro. 9:3), huvdo,khsan (Ro. 15:26); of oi oivkodomh,qh (Jo. 2:20), etc., but w|vkodo,mhsen (Lu. 7:5), etc.; of ei ei;xamen (Gal. 2: 5) just like Attic; of e diermh,nusen (Lu. 24:27), diegei,reto (Jo. 6: 18), avne,qh (Ac. 16:26), avfe,qhsan (Ro. 4:7, Ps. 32:1); of o pro orw,mhn (Ac. 2:25; Ps. 16:8), and some MSS. in Lu. 13:13 ( avnor qw,qh) and Ro. 9:29 ( o`moiw,qhmen); of i i;scusen (Lu. 8:43), i`ka,nwsen, (2 Cor. 3:6) and iva/to (Lu. 9:11); of w wvne,omai has no augment, wvnh,sato (Ac. 7:16), and the same, thing is true of wvqe,w as avpw, sato (Ac. 7:27), evxw,sen (Ac. 7:45). vErga,zomai has h, not ei, as its augment according to W. H. So hvrga,zonto (Ac. 18:3), but always ei;con.

6. Compound Verbs ( parasu,nqeta). The language varied in the way it regarded compound verbs, though usually a verb derived from a compound is treated as a unit. So evqhrioma,chsa evliqo bo,lhsan evmoscopoi,hsan (Ac. 7:41), evnaua,ghsa evprofh,teusen (Mk. 7: 6), evparrhsia,sato (Ac. 9:27), evsukofa,nthsa, but euvhggeli,sato (Ac. 8:35) in late Greek and proeuggeli,sato (Gal. 3:8). If the compound embraces a preposition, the augment as in Attic usually follows the preposition like avph,nthsan (Lu. 17:12). Some verbs derived from nouns already compounded are augmented like verbs compounded with a preposition, as dihko,nei (Mt. 8 : 15) unlike Attic. As further examples note avpedh,mhsen (Mt. 21:33), evpequ,mhsan (Mt. 13:17), kathgro,roun (Mk. 15:3), evpecei,rhsan (Lu. 1:1), avpe logei/to (Ac. 26:1), sunh,rgei (Jas. 2:22). Cf. Winer-Schmiedel, p. 102. But in Mt. 7:22 and 11:13 the Syrian class of MSS. have proefhteu,samen and - san. Sometimes the preposition itself is treated as a part of the verb when put directly to the verb, as h;fien (Mk. 1:34), h;noixen (Rev. 6:1), dih,noigen (Lu. 24:32), evka, qeudon (Mt. 25:5), evka,qhto (Mt. 13:1), evka,qisen (Jo. 19:13), evka qezeto (Jo. 4:6). In Mt. 13:15 evka,mmusan (from Is. 6:10) is assimilation of katamu,w. Verbs beginning with euv-- vary in augmented tenses between euv-- and huv, but when followed by a vowel, the verb is treated as a compound like euvhggeli,sato above.

7. Double Augment. It is fairly common in the N. T. In the


case of h;gagon and ei=pon the augment is added to the aoristic reduplication. But in e`w,rwn (Jo. 6:2 in Tischendorf's text, W. H. evqew,roun) there is a clear case of double augment like the double reduplication in e`w,raka. So also the N. T. regularly hvdunh,qhn (Mt. 17:16) and even hvduna,sqh (Mk. 7:24). Both evdu,nato (Mk. 6:5) and hvdu,nato (Mk. 14:5) appear and the MSS. vary much. This h (analogy to h;qelon) first arises in the Attic in 300 B.C.438 With me,llw h;mellon is the usual form (Jo. 4:47), though e;mellon occurs also (Jo. 7:39). Bou,lomai in the N. T. never has h, though the Text. Rec. has it in 2 Jo. 1:12. On the other hand qe,lw always has h (Gal. 4:20, h;qelon) even after the initial e was dropped. vApoka qi,sthmi has always a double augment, one with each preposition. So avpekate,sth (Mk. 8:25) and avpekatesta,qh (Mk. 3:5).439 So LXX and later Greek.440 But in Heb. 12:4 avntikate,sthte is the true text.441 vAnoi,gw has a peculiar history. It now has single augment on the preposition, as h;noixen (Rev. 6:3), now double augment of the verb, as avne,w|xen (Jo. 9:14), now a triple augment on verb and preposition, as hvnew|,cqhsan (Mt. 9:30). vAnecomai, on the other hand, has only one augment, as avnesco,mhn (Ac. 18:14) and avnei,cesqe (2 Cor. 11:1). For double augment in the LXX see Thackeray, Gr., pp. 202 ff.

VIII. The Infinitive ( h` avpare,mfatoj e;gklisij%) The most striking development of the infinitive in the koinh, belongs to syntax, and not accidence.442 Hence a brief discussion will here suffice. Blass, for instance, in his Grammar of N. T. Greek, has no discussion of the infinitive under "Accidence," nor has Moulton in his Prolegomena. But the infinitive has a very interesting history on its morphological side.

1. No Terminology at First. Originally it was a mere noun of action (nomen actionis). Not all nouns of action developed into infinitives. Brugmann443 quotes from Plato th.n tou/ qeou/ do,sin u`mi/n where a noun of action ( do,sij) is used with the dative. This is, of course, not an infinitive. The older Sanskrit shows quite a variety of nouns of action used in a "quasi-infinitive sense,"444 governing cases like the verb, but having no tense nor voice.

2. Fixed Case-Forms. The first stage in the development was reached when these nouns of action were regarded as fixed ease-


forms. That stage was obtained in the Sanskrit. At first the dative was the most common case so used along with the accusative, genitive, ablative and sometimes the locative. In the later Sanskrit the accusative supplanted the rest (tum or itum). Cf. the Latin supine.445 But the Sanskrit infinitive, while governing cases, never developed tense nor voice, and so remained essentially a substantive.

3. With Voice and Tense. But the second stage appears in the Greek and Latin where it had its most characteristic development.446 The infinitive becomes a real verbal substantive. Here voice and tense are firmly established. But while, by analogy, the Greek infinitive comes to be formed on the various tense and voice stems, that is an after-thought and not an inherent part of the infinitive. There was originally no voice, so that it is even a debatable question if timh/sai, for instance, and haberi are not formed exactly alike.447 The active and the passive ideas are both capable of development from dunato.j qauma,sai, 'capable for wondering.'448 The passive infinitive had only sporadic development in single languages.449 The middle is explained in the same way as active and passive. The tense-development is more complete in Greek than in Latin, the future infinitive being peculiar to Greek. The Latin missed also the distinctive aorist infinitive. But here also analogy has played a large part and we are not to think of lu/sai, for instance, as having at bottom more kinship with e;lusa than with lu,sij.450 Indeed the perfect and future infinitives are both very rare in the N. T. as in the koinh, generally.451 This weakening of the future infinitive is general452 in the koinh,, even with me,llw as well as in indirect discourse. In Jo. 21:25 late MSS. have cwrh/sai instead of cwrh,sein. Indeed the papyri in the later koinh, show a hybrid infinitive form, a sort of mixture of aorist and


future like evpeleu,sasqai (even in early papyri).453 In the LXX we find teu,xasqai (2 Macc. 15:7) and evkfeu,xasqai in 2 Macc. 9:22. In other cases the two are used side by side. It is only in the state of the action that the infinitive has any true tense-action developed save in indirect discourse where the infinitive tense represents the time of the direct discourse. The infinitive thus is like a verb in that it expresses action, governs cases, has voice and tense.454

4. No Personal Endings. The infinitive never developed personal endings and remained undefined, unlimited. The infinitive and the participle are thus both infinitives in this sense, that they are the unlimited verb so far as personal endings are concerned. They are both participles in that they participate in both noun and verb. The terms have no inherent distinction, but serve merely as a convenience.455 In the nature of the case neither can have a subject in any literal sense. But it is to be admitted even here that the line between the finite and the infinite verb is not absolute.456 Cf. the forms fe,re and fe,rein, for instance. But the cases used with the infinitive will be discussed in Syntax.

5. Dative and Locative in Form. The infinitive continued a substantive after the voice and tense-development. At first the case-idea of the form was observed, but gradually that disappeared, though the form remained. The Greek infinitives are always either datives or locatives, "dead datives or locatives" usually.457 All infinitives in - ai are datives. Thus all those in - nai, - sai e,nai menai (Homer), - sqai, (- qai). Those in -- sqai alone give any trouble. It is probably a compound ( s qai), but its precise origin is not clear.458 The locative is seen in - ein, and Homeric - men, but the origin of - ein is again doubtful.459 But no distinction remains between the two cases in actual usage.460 In Homer461 the dative sense as well as form remain extremely common, as indeed is true of all Greek where the infinitive remains. The very common infinitive of purpose, like h=lqon avgora,sai, is a true dative. (Cf. Mt. 2:2.) But the very essence of the infinitive as a complete development is that this dative or locative form could be


used in any case like any other substantive without inflection, an indeclinable substantive in a fixed case-form.

6. The Presence of the Article. After Homer's day it was common and chiefly in the Attic,462 but this is a matter to be treated further in Syntax. The point to observe here is that the article did not make a substantive of the infinitive. It was that before voice and tense were used with it. But it is true that even in Homer the verbal aspect is more prominent than the substantival. In the vernacular the article was never much used with the infinitive; perhaps for convenience it was not so employed.

7. The Disappearance of the Infinitive. The old forms in - ein and - nai remain longest (Thackey, Gr., pp. 210, 257). The causes for the, disappearance of the infinitive in later Greek till in the modern Greek vernacular it is (outside of the Pontic dialect) dead and gone, lie largely in the region of syntax. The infinitive as a whole disappears before o[ti and i[na (modern Greek na,). Farrar463 calls attention to the absence of the infinitive in Arabic. It was always a matter of discretion with a Greek writer whether in certain clauses he would use the infinitive or an object-clause ( o[ti o[pwj i[na).464 Cf. Latin. The English infinitive has an interesting history also as the mutilated form of the dative of a gerund.465

8. Some N. T. Forms. Not many N. T. forms call for special remark and those have been explained already, such as - oi/n (Mt. 13:32; Heb. 7:5), pei/n and even pi/n for piei/n (Jo. 4:9). In Lu. 1:79 evpifa/nai instead of the Attic evpifh/nai is noticeable. In Ph. 4:12 we have peina/n, not - h/n) The Coptic has the infinitive pa stiggoi/n (cf. W. H. kataskhnoi/n, Mt. 13:32 = Mk. 4:32, and avpode katoi/n in Heb. 7:5). In 1 Cor. 11:6 we find both kei,rasqai and xura/sqai. In Mk. 14:71 ovmnu,nai is the regular - mi form. In Heb. 11:5 euvaresthke,nai is without reduplication in AKL. In Lu. 9: 18 18(11:1) a periphrastic infinitive appears, evn tw|/ ei=nai auvto.n pro seuco,menon. The augment occurs with avnew|cqh/nai. in Lu. 3:21. Cf. e;somai dido,nai in Tob. 5:15 B.

IX. The Participle ( h` metoch,%.

1. The Name. This does not really distinguish this verbal adjective from the verbal substantive, the infinitive. Both are par-


ticiples and both are infinitives. Voss466 calls the participles "mules" because they partake of both noun and verb, but the infinitives are hybrid in exactly the same sense. Like the infinitive, the Greek participle has voice, tense, and governs cases, and may use the article. Unlike the infinitive the participle has regular inflection like other adjectives. Clyde467 would include participles in the infinitive. So Kuhner-Blass.468 Dionysius Thrax469 puts the participle right: Metoch, evsti le,xij mete,cousa th/j tw/n r`h ma,twn kai. th/j tw/n ovnoma,twn ivdi,othtoj.

2. Verbal Adjectives. As a matter of fact no absolutely clear line can be drawn between verbal adjectives and other adjectives.470 An adjective may not only be used with a case like keno,j with the ablative, but may even take on a verbal, nature in certain connections.471 Some, like kluto,j were always purely adjectival.472 Most of the forms in - toj in Greek are adjectival, but many of them have a verbal idea developed also, either that of completion, as avgaphto,j ('beloved,' Mt. 3:17) , or of possibility or capability, as paqhto,j ('liable to suffering,' Ac. 26:23) . In Greek these verbals in - toj never became a part of the verb as in Latin perfect passive participle.473 Moulton474 shows how amatus est and "he is loved" represent different tenses, but scriptum est and "it is written" agree. But there was no reason why the - toj should not have had a further verbal development in Greek. For the structure of this verbal adjective see the chapter on Formation of Words, where a list of the chief examples is given. Moulton475 points out the wavering between the active and passive idea when the true verbal exists in the N. T., by the example of avdu, natoj, in Ro. 8:3. Is it 'incapable' as in Ro. 15:1 or 'impossible' as is usual? Blass476 indeed denies the verbal character of the - toj form in the N. T. to any examples except paqhto,j (Ac. 26:23). But this is too extreme, as Moulton477 clearly proves. vAsu,netoj is active in Ro. 1:31 while avsu,nqetoj is middle ( sunti,qemai). With the forms in - toj therefore two points have to be watched: first, if they are verbal at all, and then, if they are active, middle or passive. There is no doubt as to the verbal character of the form in - te,oj, which expresses the idea of necessity. This is in fact a ge-


rundive and is closely allied to the - toj form.478 It has both a personal construction and the impersonal, and governs cases like the verb. It is not in Homer479 (though - toj is common), and the first example in Greek is in Hesiod.480 The N. T. shows only one example, blhte,on (Lu. 5:38), impersonal and governing the accusative. It appears in a few MSS. in the parallel passage in Mk. 2: 22. One further remark is to be made about the verbals, which is that some participles lose their verbal force and drop back to the purely adjectival function. So e`kw,n me,llwn in the sense of 'future.' Cf. eloquens and sapiens in Latin.481 In general, just as the infinitive and the gerind were surrounded by many other verbal substantives, so the participle and the gerundive come out of many other verbal adjectives. In the Sanskrit, as one would expect, the division-line between the participle and ordinary adjectives is less sharply drawn.482

3. True Participles. These have tense and also voice. Brugmann483 indeed shows that the Greek participle endings go back to the proethnic participle. Already in the Sanskrit the present, perfect and future tenses (and in the Veda the aorist) have participles in two voices (active and middle),484 thus showing an earlier development than the infinitive. The endings of the Greek participles are practically the same as those of the Sanskrit. The Latin, unlike the Sanskrit and the Greek, had no aorist and no perfect active participle, and the future participle like acturus may have come from the infinitive.485 The Greek has, however, two endings for the active, - nt for all tenses save the perfect, just like the Sanskrit. The perfect ending (wes, -wos, -us, Greek - wj, - ot, - w) is difficult of explanation, but is likewise parallel with the Sanskrit.486 The perfect participle is more common in Homer than any other form of the perfect (Sterrett, Homer's Iliad, N. 44). The middle ending - meno is uniform and is like the Sanskrit. The Greek aorist passive participle ending (-- qent) is peculiar to the Greek and is made by analogy from the old active form like fane,ntj $fanei,j),


like Latin, manens.487 The participles survive in modern Greek, though the active, like the third declension, takes on the form gra,fontaj ( gra,fwn).488

The modern Greek uses chiefly the present active, the past passive participle (Dieterich, Unters., p. 206), and some middle or passive participles in - ou,menoj or - a,menoj (Thumb, Handb., p. 167). The use of the aorist and perfect active participles gave Greek a great superiority over the Latin, which had such a usage only in deponent verbs like sequor, secutus. But Greek used the other participles far more than the Latin. English alone is a rival for the Greek in the use of the participle. One of the grammarians calls the Greeks filome,tocoi because they were a participle-loving people.489 The use of the tenses of the participle belongs to syntax. One may merely remark here that the future participle is very rare in the N. T. as in the papyri and koinh, generally (cf. Infinitive). The LXX has it seldom (Thackeray, Gr., p. 194). It is found chiefly in Luke in the N. T., as Lu. 22:49; Ac. 8:27; 20:22; 22:5; 24:11, 17.490 The N. T. itself presents no special peculiarities as to the forms of the participle. In Rev. 19:13 r`eramme,non has been cited under the question of reduplication. `Estw,j is more frequent than e`sthkw,j. Other perfects like avpolwlw,j call for no comment.

4. In Periphrastic Use. The participle is common in the N. T. in the periphrastic tenses. These have been given in detail under the various tenses, but a summary at this point is desirable. This use of the participle with various forms of the verb "to be" is so common in all languages, ancient and modern, as hardly to require justification. Modern English uses it largely in its verbinflection, as does modern Greek. The use of the participle as the predicate is found all through the Indo-Germanic languages.491 It is very frequent in the Sanskrit, especially in the later language.492 Its oldest usage seems to be in the perfect tense, which exists as far back as we can go.493 In the N. T. the perfect optative does


Addenda 3rd ed.

Addenda 3rd ed.

not appear, though once a good chance for the periphrastic perfect optative arises as in Ac. 21:33, evpunqa,neto ti,j ei;h kai. ti, evstin pepoih kw,j. The perfect subj. save eivdw/ is seen in the N. T. only in the periphrastic form both in the active, as h|= pepoihkw,j (Jas. 5:15), and the passive, as h|= peplhrwme,nh (Jo. 16:24).494 So 2 Cor. 9:3. The periphrastic perfect imperative is illustrated by e;stwsan perie zwsme,nai (Lu. 12:35). No example of the periphrastic perfect infinitive appears in the N. T., so far as I have noticed, except katestalme,nouj u`pa,rcein (Ac. 19:36). A periphrastic perfect participle also is observed in o;ntaj avphllotriwme,nouj (Col. 1:21). Colloquial Attic has it (Arist. Ran. 721) and the inscriptions (Syll. 928495 ii/B.C.) avpokekrime,nhj ou;shj (Moulton, Prol., p. 227). In the indicative the periphrastic form is the common one for the future perfect, both active, as e;somai pepoqw,j (Heb. 2:13), and passive, as e;stai lelume,na (Mt. 18:18). Cf. Lu. 12:52. Moulton (Prol., p. 227) finds three papyri with aorist participles in future perfect sense. With gi,nomai note gego,nate e;contej (Heb. 5:12). Cf. Rev. 16:10, evge,neto evskotisme,nh. Cf. 2 Cor. 6:14; Col. 1:18; Rev. 3:2. The past perfect is very common in the passive, as h=n gegramme,non (Jo. 19:19), but less frequent in the active, as h=san proewrako,tej (Ac. 21:29). In Ac. 8:16 we not only have h=n evpi peptwkw,j, but even bebaptisme,noi u`ph/rcon (cf. also 19:36). Cf. also h=n kei,menoj as equal to h=n teqeime,noj (Lu. 23:53); h=n e`stw,j (Lu. 5:1); ei=con avpokeime,nhn (Lu. 19:20), like e;ce parh|thme,non (Lu. 14:18), since kei/mai is perfect in sense. The present perfect is more common in the periphrastic form than in the active, as e`stw,j eivmi (Ac. 25:10), and especially in the passive, as gegramme,non evsti,n (Jo. 6:31).

The periphrastic aorist appears only in h=n blhqei,j (Lu. 23:19) and only in the indicative.496 But note evge,neto sti,lbonta (Mk. 9:3).

The periphrastic future indicative is found several times in the active, as e;sontai pi,ptontej (Mk. 13:25), and the passive, as e;sesqe misou,menoi (Lu. 21 : 17).

The present tense is written periphrastically in the imperative, as i;sqi euvnow/n (Mt. 5:25; cf. Lu. 19:17), and even with gi,nomai, as mh. gi,nesqe e`terozugou/ntej (2 Cor. 6:14). Cf. Rev. 3:2. In Col. 1:18 we find an aorist subjunctive with a present participle, i[na ge,nhtai prwteu,wn. The present infinitive occurs in evn tw|/ ei=nai auvto.n proseuco,menon (Lu. 9:18; 11:1). As an example of the present indicative active take a[ evstin e;conta (Col. 2:23), and of


the passive take o[ evstin meqermhneuo,menon, (Jo. 1:42), though this last is not strictly an instance in point. Cf. also evsti.n prosana plhrou/sa (2 Cor. 9:12).

The periphrastic imperfect is the most common of all. It is not unknown to the old Greek, and is abundant in the papyri and the koinh, generally, but it is even more frequent in the LXX (Thackeray, Gr., p. 195) and in the Aramaic. As Blass497 shows, not all the examples in the N. T. are strictly periphrastic, like h=san . . . avgraulou/ntej (Lu. 2:8). But they are abundant enough, as one can see on almost any page of the Gospels. Take h=san avna bai,nontej kai. h=n proa,gwn (Mk. 10:32). So Ac. 2:2, h=san kaqh,menoi and Gal. 1:22, h;mhn avgnoou,menoj.

For list of important verbs in the N. T. see Additional Notes and my Short Grammar of the Greek N. T. (third ed.), pp. 48-56, 241-244. For such verbs in the LXX see Thackeray, Gr., pp. 258-920 (Table of Verbs); Helbing, Gr. d. LXX, pp. 128-135. For list in the papyri see Mayser, Gr., pp. 387-415.

1 Giles, Man., p. 403 f.

2 Hirt, Handb., p. 332.

3 Man., p. 404.

4 Steinthal, Zeitschr. fur Volkerpsych. etc., p. 351. Cf. Schleicher, Unterscheidung von Nomen und Verbum etc., 4. Bd. der Abh. d. phil. etc., 1865, p. 509.

5 Giles, Man., p. 424.

6 Schroeder, Uber die form. Untersch. d. Redet. im Griech. und Lat., 1874, pp. 10 ff.

7 Die Lehre von den Redet. etc., 1864, p. 31.

8 Monro, Hom. Gr., p. 1. In the Sans. it is to be noted that the noun had an earlier and a more rapid development than the verb. The case-endings appear first in the Sans., the verb-conjugation in the Gk., though the personal endings are more distinct in the Sans.

9 Cf. Garnett, Philol. Ess.

10 Cf. Gr. Gen. of Port Royal; Farrar, Gk. Synt., p. 38.

11 Ib. He considers the verb later than the noun because of its complex idea. Cf. Schramm, Uber die Bedeutung der Formen des Verbums (1884); Curtius, Die Bildung der Tempora und Modi im Griech. und Lat. (1846); Junius, Evolution of the Greek Verb from Primary Elements (1843); Lautensach, Verbalflexion der att. Inschr. (1887); Hogue, Irregular Verbs of Attic Prose (1889).

12 Cf. Brug., Grundr., Bd. II, pp. 2, 837. On difference between finite and non-finite verbs see Curtius, Das Verbum d. griech. Spr., p. 1 f.

13 Hirt, Handb., p. 363 f. Cf. also Giles, Man., pp. 425

14 Donaldson, New Crat., pp. 570

15 Donaldson, New Crat., pp. 570 ff. Cf. Farrar, Gk. Synt., p. 39.

16 Cf. Hirt, Handb., pp. 355 ff.; Giles, Man., pp. 413 ff.

17 Jann., Hist. Gk. Gr., p. 232 f.

18> Monro, Hom. Gr., p. 51.

19 Farrar, Gk. Synt., p. 46.

20 Monro, Hom. Gr., p. 2. Cf. Clyde, Gk. Synt., 5th ed., 1876, p. 54; Riem. and Goelzer, Phonet., pp. 347

21 Cf. King and Cookson, Prin. of Sound and Inflexion, 1888, pp. 225

22 Gk. Gr., 1893, p. 245.

23 Thompson, Hom. Gr., 1890, p. 127.

24 Mayser, Gr. d. griech. Pap., p. 389.

25 Blass, Gr. of N. T. Gk., p. 50.

26 Gr. d. griech. Pap., p. 364 f.

27 W.-Sch., p. 115. Cf. Veitch, Gk. Verb, p. 110.

28 Mayser, Or. d. griech. Pap., p. 391. 2 W.-M., p. 360 note.

29 Moulton, Prol., p. 55. Cf. Mayser, Gr. d. griech. Pap., pp. 137, 325, for o[pwj doi/.. Cf. Moulton, Cl. Rev., 1901, pp. 37, 436.

30 Mayser, Gr. d. griech. Pap., p. 392.

31 Blass, Gr. of N. T.. Gk., p. 49.

32 Meisterh., Att. Inschr., p. 188 f.

33 Gr. d. griech. Pap., p. 367 f.

34 So W.-H., Notes on Orth., p. 167 f. Cf. W.-Seh., p. 121. For pap. exx. see Moulton, Cl. Rev., 1901, p. 37.

35 Cf. Blass, Gr. of N. T. Gk., p. 49.

36 Notes on Orth., p. 168. Cf. also W.-Sch., p. 121.

37 Gr. of N. T. Gk., p. 48 f.

38 Prol., p. 55. Cf. Dittenb., Syll., 462. 17, etc.

39 Hort, Notes on Orth., p. 168.

40 Gr. of N. T. Gk., pp. 49, 212.

41 Mayser, Gr. d. griech. Pap., p. 398.

42 Cf. Hort, Notes on Orth., p. 166. The evidence is "nowhere free from

43 Mayser, Gr. d. griech. Pap., p. 398.is :

44 Hort, Notes on Orth., p. 168.a W.-M., p. 94.

45 Thack., Gr., p. 254. Cf. W.-Sch., p. 122 f. On i`sta,nai and its compounds in the LXX see interesting list in C. and S., Sel. fr. LXX, p. 43 f., giving - w forms, transitive e[staka, etc.

46 Mayser, Gr. d. griech. Pap., p. 411.

47 Gr. of N. T. Gk., p. 50. The verb is mentioned here to impress the fact that it is aorist as well as imperfect.

48 Ib., p. 48.

49 In the pap. both - umi and - u,w but only - umai. Mayser, Gr. d. griech. Pap., p. 392.

50 Moulton, Cl. Rev., 1901, p. 37. Cf. Deiss., B. S., p. 192. Mod. Gk. has di,dw.

51 Hort, Notes on Orth., p. 167. Cf. also W.-Sch., p. 121.

52 Moulton, Cl. Rev., 1901, p. 37.

53 Schweizer, Perg. Inschr., p. 177.

54 Mayser, Gr. d. griech. Pap., p. 355; Moulton, Cl. Rev., 1901, p. 36. Cf. also Dieterich, Untersuch., p. 222; Schmid, Atticismus, IV, p. 597; Deiss., B. S., p. 193.

55 Notes on Orth., p. 168. Cf. Lobeck, Phryn., p. 359 f.

56 Mayser, Gr. d. griech. Pap., p. 355; Moulton, Cl. Rev., 1901, p. 36.

57 Mayser, ib., p. 394.

58 Ib., p. 356.

59 Schweizer, Perg. Inschr., p. 178.

60 Prol., p. 56. D (M. shows) alone has h=n in Ac. 20:18.

61 Cities and Bish. of Phrygia, II, 565.

62 Prol., p. 37.

63 W.-Sch., p.,117:

64 Cf. Moulton, Prol., p. 56. Both forms in pap. and inscr. On h;mhn h=j h;meqa h;tw e;stwsan in the LXX see C. and S., Sel. fr. LXX, p. 31 f. Thack., Gr., P. 256 f. Beyond this the LXX goes very little.

65 Meisterh., Att. Inschr., p. 191.

66 Moulton, Rev., 1901, p. 436.

67 Ib., p. 38. Cf. Gen. 6:17 E, according to Moulton, Prol., p. 49.

68 Cf. Blass, Gr. of N. T. Gk., p. 51 f.; Thack., p. 257.

69 Mayser, GrF d. griech. Pap., p. 355.

70 Nachm., Magn. Inschr., p. 157.

71 Gr. of N. T. Gk., pp. 52, 54.

72 Ib., p. 52.

73 Just so the pap., Mayser, Gr. d. griech. Pap., p. 395.

74 Blass, Gr. of N. T. Gk., p. 52. Cf. also for pap., Moulton, Cl. Rev., 1901, p. 38. For LXX see Thackeray, p. 272.

75 W.-Sch., p. 118; Schweizer, Perg. Inschr., p. 177; Reinhold, De Graec., p. 89.

76 Gr. d. griech. Pap., p. 398.

77 Mayser, ib., p. 354; Hort, Notes on Orth., p. 167.

78 W.-Sch p. 123. Herod. is cited for the use of evxi,ei and meti,ei as - w presents.

79 Ib.

80 Reinhold, De Graec., p. 94.

81 So Hort, Notes on Orth., p. 167; W.-Sch., p. 123; Hatz., Einl., pp. 309, 334.

82 Moulton, Prol., p. 38 f.

83 Gr. of N. T. Gk., p. 51. He gives the MS. variations and parallels in Hermas and Barn. See further A. Buttmann, Gr., p. 48.

84 Mayser, Gr. d. griech. Pap., p. 398.

85 Hort, Notes on Orth., p. 168; Blass, Gr. of N. T., p. 48.

86 Blass, Gr. of N. T., p. 49.

87 Hort, Notes on Orth., p. 168; Blass, Gr. of N. T. Gk., p. 49.

88 Here Hort (Notes, etc., p. 168) differs from Westcott and prefers - a,nai.

89 Ib.

90 Blass, Gr. of N. T. Gk., p. 48.

91 Ib. W.-Sch., p. 122.

92 Meisterh., Att. Inschr., p. 177. For many -- nw verbs in mod. Gk. see Thumb, Handb., p. 133 f.

93 Mayser, Or. d. griech. Pap., pp. 354, 399. For the Byz. and mod. Gk. usage see Dieterich, Unters., p. 223.

94 Blass, Gr. of N. T. Gk., p. 51.

95 In the LXX the active goes over to the - w class. Thack., Gr., p. 273.

96 Mayser, Gr. d. griech. Pap., p. 403.

97 Ib., p. 404. And indeed the old Attic avnoi,gw, Meisterh., p. 191.

98 Thack., Gr., p. 277.

99 So the pap. Mayser, Gr., p. 352; Thackeray, p. 246.

100 Mayser, ib., pp. 351 f., 404.

101 lb., p. 406.

102 Tisch. reads evmpipra/sqai from pipra,w. Nestle agrees with W. H.

103 Mayser, Gr., p. 352.

104 Thack., Gr., p. 286.

105 Nachm., Magn. Inschr., p. 156; Schweizer, Perg. Inschr., p. 176.

106 Mayser, Gr. d. griech. Pap., p. 352 f.

107 Blass, Gr. of N. T. Gk., p. 49.

108 Hort, Notes on Orth., p. 167.

109 P. 121.

110 Deiss., B. S., p. 192 f.; Moulton, Cl. Rev., 1901, p. 37.

111 Ib. Mod. Gk. has qe,tw.

112 Gr. d. griech. Pap., p. 355.

113 Ib. So inscr., Nachm., p. 157

114 Mayser, Gr. d. griech. Pap., p. 372.

115 Cf. W.-Sch., p. 114 f. Neither oi=sqa nor h|;deisqa appears in the N. T.

116 Mayser, Gr. d. griech. Pap., p. 370 f.

117 Ib. Cf. W.-Sch., p. 119.

118 See Hoffmann, Die griech. Dial., Bd. II, pp. 572 ff., for - mi verbs in North Achaia. For the "strong" perfects, like ge,gona, see VII, (g), 2.

119 Gk. Synt., p. 62. Cf. Hirt, Handb. etc., p. 417.

120 Hist. Gk. Gr., p. 179.

121 New Crat., p. 617 f.

122 Hom. Gr., p. 49.

123 Phonet., p. 455.

124 Prol., p. 164 f. Farrar (Gk. Synt., p. 45) refers to Protagoras as the one who first distinguished the moods.

125 Giles, Man., p. 459.

126 Prol., p. 164,

127 Clyde, Gk. Synt., p. 62. Cf. Kohlmann, Uber die Modi des griech. and des lat. Verbums (1883).

128 Whitney, Sans. Gr., p. 201.

129 See discussion bet. Profs. Harry and Sonnenschein in Cl. Rev., 1905-6. Cf. also La Roche, Beitr. zur griech. Gr., 1893; Jann., Hist. Gk. Gr., p. 197.

130 For contrary view see Burton, N. T. Moods and Tenses, p. I.

131 Farrar, Gk. Synt., p. 45 f.

132 Bd. II, p. 40.

133 Giles, Man., p. 458 f.

134 Ib., p. 459. In the Boeotian dial. the subj. does not appear in simple sentences (Claflin, Synt. of Bmotian, etc., p. 73)

135 Riem. and Goelzer, Phonet., p. 456 f.

136> Monro, Hom. Gr., p. 49.

137 Hist. Gk. Gr., p. 179.

138 Sterrett, Hom. II., Dial. of Homer, p. 27 (1907). Cf. Moulton, The Suffix of the Subj. (Am. Jour. of Philol., 10, 185 f.); La Roche, Die conj. and opt. Formen des Perfects (Beitr. I, pp. 161 ff.).

139 Cf. already in the Attic inscr. the spelling of the subj. in - ei. Meisterh., Att. Inscr., p. 166. For this phenomenon in the pap. see Mayser, Gr. d. griech. Pap., p. 324.

140 Cf. Henry, Comp. Gr. of Gk. and Lat., Elliott's transi., 1890, p. 115 f. and note; Giles, Comp. Philol., p. 459.

141 Cf. Moulton, Cl. Rev., 1901, p. 37, and 1904, p. 111, for subjs. avpodoi/ evpignoi/ in the pap.

142 Cf. avrxhsqe in Lu. 13:25, but a;rxesqe (BEG, etc.) and a;rxhsqe ( aAD, etc.) in verse 26.

143 Blass, Gr. of N. T. Gk., p. 48. But in 1 Cor. 16 : 2 we have regularly euvo- dw/tai (marg. euvodwqh|/). Hort (Notes on Orth., pp. 167, 172) is uncertain whether euvodw/tai is perf. ind. or subj. (pres. or perf.). He cites parazhlou/men (1 Cor. 10:22) and diabebaiou/ntai (1 Tim. 1:7) as possible pres. subjs.

144 Man. of Comp. Philol., p. 458. Cf. Brug., Griech. Gr., p. 337, for list of works on optative.

145 Whitney, Sans. Gr., p. 202. Giles, Comp. Philol., p. 503 f.

146 Giles, ib., p. 459. On the blending of subj. and opt. in Ital., Germ. and Balto-Slav. tongues see Brug., Kurze vergl. Gr., 2. Tl., p. 585. Cf. the Byz. Gk. mingling of subj. and ind. in Hatz., Einl., p. 216 f.

147 Clyde, Gk. Synt., p. 84.

148 Gr. S., p. 85.

149 Monro, Hom. Gr., p. 219.

150 Hist. Gk. Gr., p. 179.

151 Riem. and Goelzer, Phonet.,p. 461. Cf. K.-BI., Bd. II, p.40 f.; Brug., Gk. Gr., pp. 337 ff.

152 Meisterh., Att. Inschr., p. 166.

153 Synt. of Bceot. Dial. Inscr., pp. 77, 81.

154 Schweizer, Perg. Inschr., p. 191.

155 Mayser, Gr. d. griech. Pap., p. 326.

156 K.-B1., Bd. II, p. 99.

157 Mayser, Gr. d. griech. Pap., p. 327.

158 Hort, Intr. to N. T. Gk., p. 168. Cf. LXX.

159 Schweizer, Perg. Inschr., p. 191.

160 Mayser, Gr. d. griech. Pap., p. 326 f.; Cronert, Mem. Gr. Hercul., p. 215 f.; Moulton, Cl. Rev., 1904, p. 111 f. Doi/ also appears in pap. as opt. as well as subj.

161 Prol., p. 55. Cf. Blass' hesitation, Gr. of N. T. Gk., p. 49 f.

162 Cf. W.-Sch., p. 114. In the LXX the form in - eie is very rare. Cf. Helbing, Gr. d. Sept., p. 68 f. The LXX has also - oisan aisan 3d plu. Cf. Thack., Gr., p. 215. Opt. is common in 4 Macc.

163 Gr. of N. T. Gk., p 220.

164 K.-B1., Bd. II, p. 41.

165 Giles, Comp. Philol., p. 464.

166 Ib., p. 269.

167 Ib., p. 464. Cf. Brug., Grundr., II, 958; Riem. and Goelzer, Phonet., p. 359. It is coming more and more to be the custom to regard the thematic vowel as part of the root. Giles, Comp. Philol., p. 415.

168> Moulton, Prol., p. 171 f.

169 Cf. Brug., Griech. Gr., p. 341.

170 Hort, Notes on Orth., p. 168.

171 Mayser, Gr. d. griech. Pap., p. 327.

172 Giles, Comp. Philol., p. 466. Cf. Brug., Gk. Gr., p. 341.

173 So pap. and late inscr., Moulton, Prol., p. 56.

174 Cf. for pap. Mayser, Gr. d. griech. Pap., p. 327. Cf. Brug., Griech. Gr., p. 343. It is after iii/B.C. that - twsan completely supplants - ntwn. Cf. Meisterh., Att. Inschr., p. 167. Nachm., Magn. Inschr., p. 149. Schweizer, Perg. Inschr., p. 167.

175 Brug., Griech. Gr., p. 343 (he cfs. e`pe,sqw with e`pe,sqai); Hirt, Handb. etc., p. 430. Giles (Comp. Philol., p. 467 f.) gets it from tw by analogy of te and sqe.

176 Moulton, Prol., p. 165.

177 Brug., Griech. Gr., p. 332.

178 Ib.

179 Ib.

180 Hirt, Handb., p. 429 f.

181 W.-Sch.,p. 119.

182 Moulton, Prol., p.165.

183 Riem. and Goelzer, Phonet., p. 372. Cf. Brug., Griech. Gr., p. 345.

184> Giles, Comp. Philol., p. 468; Hirt, Handb., p. 430; Wright, Comp. Gk. Gr., p. 334.

185 Moulton, Prol., p. 179 f.

186 V. and D., Handb., p. 81. Cf. Dieterich, Unters., p. 205.

187 Brug., Griech. Gr., p. 345; Hirt, Handb., p. 427.

188 Hort, Notes on Orth., p. 164.

189> K.-B1., Bd. II, p. 45.

190 Monro, Hom. Cr., p. 240.

191 Ib.; cf. also Delbruck, Synt. Forsch., IV, p. 120. Hence Delbruck argues that the aorist imper. did not come into use until after the pres. imper. The imper. was originally only positive, not negative.

192 Handb. to the Gk. of N. T., p. 55.

193 Cf. Dion. Thr., p. 886. Cf. Farrar, Gk. Synt., p. 40.

194 Hist. Gk. Gr., p. 179.

195 Sans. Gr., p. 200.

196 Gk. Synt., p. 41. Cf. Brug., Griech. Gr., p. 467 f.

197 Giles, Comp. Philol., p. 476: "The distinction between the transitive and intransitive meanings of the active voice depends upon the nature of the root in each case."

198 Whitney, Sans. Gr., p. 200. Cf. also Brug., Kurze vergl. Gr., II, p. 492. See also Clark, Comp. Gr., p. 182, for the meaningless term "middle." It is as active as the "active" voice. Farrar, Gk. Synt., p. 119.

199 Whitney, Sans. Gr., p. 275; Thumb, Handbuch d. Skt., pp. 394ff.

200 Sterrett, Hom. Il., Dial. of Hom., p. 27.

201> Comp. Philol., p. 477.

202 Clyde, Gk. Syn., p. 55.

203 Moulton, Prol., p. 152.

204 Griech. Gr., p. 346. Cf. Kurze vergl. Gr., II, p. 599. Cf. Giles, Comp. Philol., p. 419.

205 Prol., p. 153,

206 Whitney, Sans. Gr., p. 200.

207 Brug., Furze vergl. Gr., p. 598; Moulton, Prol., p. 153.

208 Hirt, Handb., p. 334; Moulton, Prol., p. 154.

209 Delbrtick, Synt. Forsch., Bd. IV, p. 69.

210 Clyde, Gk. Synt., p. 55.

211 Hist. Gk. Gr., pp. 362 ff.

212 Sterrett, Horn. II., Hom. Dial., p. 27.

213 Monro, Hom. Gr., p. 7.

214 Thumb, Handb., p. 111. So mod. Gk. has only two voices; V. and D., Handb., to Mod. Gk., p. 81.

215 Jann., Hist. Gk. Gr., p. 362.

216 Ib. koinh, exx. are numerous, like h|`de,sqhn evnequmh,qhn evporeu,qhn evfobh,qhn, etc.

217 Clyde, Gk. Synt., p. 53.

218 Gk. Gr., p. 346.

219> Cf. Clyde, Gk. Synt., p. 54. The same thing has happened in Eng. where the loss is nearly complete save 2d and 3d pers. sing.

220 It is not worth while here to take time to make a careful discussion of each of these endings. For the hist. treatment of them see Brug., Griech. Gr., pp. 345 ff.; Giles, Comp. Philol., pp. 413 ff.; Riem. and Goelzer, Phonet., pp. 348 ff.

221 Moulton, Prol., p. 33. Cf. Dieterich, Unters., p. 242.

222 Gr. d. griech. Pap., p. 323.

223 Prol., p. 52; Cl. Rev., 1901, p. 36, 1904, p. 110.

224 Nachm., Magn. Inschr., p. 148; Schweizer, Perg. Inschr., p.166. See further Dieterich, Unters., p. 242 f. Cf. Deiss., B. S., p. 191; W.-Sch., p. 112 f.

225 Gr. of N. T. Gk., p. 46.

226 Prol., p. 52.

227 Ge,gonan and anderes Vulgargriechisch, Rhein. Mus., 1891, pp. 193 ff. Cf. Moulton, Cl. Rev., 1901, p. 36.

228 K.-B1., Bd. II, p. 48 f.

229 Mayser, Gr. d. griech. Pap., p. 323 f. "A fair show in the papyri," Moulton, Prol., p. 52.

230 Schweizer, Perg. Inschr., p. 167. Thumb (Hellen., p. 170) rightly denies that it is merely Alexandrian. For LXX exx. ( e`w,rakan pe,prakan, etc.) see Helbing, Gr. d. Sept., p. 67.

231 Prol. p. 53. Cf. Hort, Notes on Orth., p. 169. The N. T. does not follow illiterate pap. in putting - asi to aorist stems (Moulton, Cl. Rev., 1901, p. 36).

232 Ib.; Prol., p. 52.

233 B. S., p. 192.

234 Gr. d. griech. Pap., p. 321.

235 Unters. etc., p. 239. Blass, Gr. of N. T. Gk., p. 46, cites Apoll., Synt., I, 10, p 37, as saying that ei;rhkej e;grayej graye,tw, etc., gave the grammarians trouble.

236 Blass, Gr. of N. T. Gk., p. 46.

237 Cf. W.-Sch., p. 113.

238 Meisterh., Att. Inschr., p. 183 f.

239 Dieterich, Unters., p. 237 f. For the inscr. see Schweizer, Perg. Inschr., p. 181 f.; Nachm., Magn. Inschr., p. 166 f.

240 Cl. Rev., 1901, p. 36. Cf. Mayser, Gr. d. griech. Pap., p. 368 f.

241 Ib. Cf. Deiss., B. S:, p. 190 f.2

242 Gr. of N. T. Gk., p. 45. The LXX is in harmony with this tendency also. Is it Cilician according to Heraclides? W.-Sch., p. 111 note. Cf. in Hom. forms like h;xonto evbh,seto where the sec. aorist endings go with the first aorist stern (Sterrett, Hom. 11., N. 42).

243 Notes on Orth., p. 164 f. See also Blass, Gr. of N. T. Gk., p. 45; W.-Sch., p. 111 f. The LXX MSS. tally with the N. T. in the use of - a. Cf. Helbing, Gr. d. Sept., pp. 62-65; C. and S., Sel. fr. LXX, p. 35 f.

244 Notes on Orth., p. 164. Moulton (Prol., p. 51) speaks of "the functionally useless difference of ending between the strong and the weak aorist."

245 Prol., p. 52. So Buresch, Rhein. Mus., 46, 224. Hort (Notes on Orth., p. 165) needlessly considers evkce,ete (Rev. 16:1) a second aorist upper. instead of the present. Cf. evxe,cean (usual form in Rev. 16:6). Cf. WT.-Sch., p. 111. But kate,ceen (Mk. 14:3) is the usual Attic aorist. Cf. Moulton, Prol., p. 55.

246 B. S., p. 191, e;legaj, etc.

247 Cf. Simcox, Lang. of the N. T., p. 36; Geldart's Guide to Mod. Gk., p. 272 note.

248 With rare variations in the inscr. and pap. Moulton, Prol., p. 53. Cf. Mayser, Gr. d. griech. Pap., pp. 320

249 Schweizer, Perg. Inschr., p. 168. Cf. also Mayser, Gr. d. griech. Pap., p. 328. The pap. do not show oi;ei and o;yei, but only bou,lei.

250 Gr. of N. T. Gk., p. 47. For oi;h| o;yh|, and bou,lh| in LXX MSS. see Helbing, Gr. d. Sept., p. 60 f.; C. and S., Sel. fr. LXX, p. 33 f. B in the LXX shows a fondness for - ei forms (itacism). Cf. Thack., Gr., p. 217.

251 Gr. of N. T. Gk., p. 47. Cf. Mayser, Gr. d. griech. Pap., p. 328.

252 Giles, Man. of Comp. Philol., pp. 410, 427.

253 Ib., pp. 411, 422. On "Passive Formations" see Hadley, Ess. Phil. and Crit., p. 199. On the strong passive forms in LXX see C. and S., Sel. fr. LXX, p.41.

254 Giles, Man. of Comp. Philol., p. 411.

255 Cf. Mayser, Gr. d. griech. Pap., p. 328, for cariei/sai. The LXX (1 Ki. 14 : 6 A) shows avpexenou/sai. The only certain instance in the LXX is kta/sai. (Sir. 6:7). See Thack., p. 218. Cf. further Hatz., Einl., p. 188.

256 Hort, Notes on Orth., p. 166.

257 Ib. Moulton (Cl. Rev., 1901, p. 36) cites evni,kei and timou/ntej from pap.

258 Pp. 42, 116 note.

259 Prol., p. 54. Cf. Brug., Griech. Gr., p. 61. The pap. support zh/n, not zh|/n) Mayser, Gr. d. griech. Pap., p. 347. So in general the pap. are in harmony with N. T. usage here, Mayser, pp. 346 ff.

260 Moulton, Prol., p. 54.

261 W.-Sch., p. 116 note. Cf. kathrame,noj (Mt. 25:41).

262 Hatz., Einl., p. 128 f. Moulton (Cl. Rev., 1904, p. 110) cites fronw/ntej and per contra avgapou/ntej from pap.

263 P. 117 note.

264 Hort (Notes on Orth., p. 166) prefers xu,rasqai after Plut. and Lucian.

265 Cf. W.-Sch., p. 116 f. See further on this mixing of contract verbs, Mayser, Gr. d. griech. Pap., p. 349. The LXX MSS. show much the same situation as to contract verbs that we find in the N. T. and the pap. Helbing (Gr. d. Sept., pp. 110-112) gives the facts in detail.

266 Notes on Orth., p. 166.

267 Gr. of N. T. Gk., p. 47.

268 Prol., p. 54.

269 Cf. Thack., Gr., pp. 242ff.; W.-Sch., p.115 note.

270 Hort, Notes on Orth., p. 166.

271 Ib. BD always have it.

272 Gr. of N. T. Gk., p. 48. Cf. K.-B1., Bd. II, p. 587.

273 Prol., p. 54.

274 Notes on Orth., p. 171 f.

275 Moulton, Prol., p. 54. Cf. Hort, Notes on Orth., p. 167.

276 Hort, ib., p. 166.

277 Gr. of N. T. Gk., p. 48.

278 Notes on Orth., p.166.

279 Prol., p. 53. Cf. Nestle (Am. Jour. of Theol., July, 1909, p. 448) for mastiggoi/n, in Coptic.

280 Cr. d. griech. Pap., p. 349; Raderm., p. 74.

281 Hort, Notes on Orth., p. 170.

282 Cf. Delbruck, Grundl. d. griech. Synt., Bd. IV, p. 80; Brug., Griech. Gr., p. 469 f.; Giles, Man. of Comp. Philol., p. 481 f. See Swete, 0. T. in Gk., p. 305, for remarks about tenses in the LXX.

283 Giles, Man. of Comp. Philol., p. 482 f.

284 Cf. Mutzbauer, Grundl. der Tempuslehre (1893); Delbruck, Grundl. d. grieeh. Synt., II, pp. 13 ff.; Brug., Griech. Gr., pp. 470 ff.; Giles, Man. of Comp. Philol., p. 480 f.; Moulton, Prol., pp. 108 ff.

285 Thumb (Handb., p. 123) likewise feels the necessity of a word about Aktionsart under Morphology.

286 Comp. Gr. of the Gk. and Lat., Elliott's transl., 1890, p. 105 f. note. Cf. Leo Meyer, Griech. Aoriste, 1879, p. 5 f.

287 Brug., Griech. Gr., p. 268.

288 Ib. Cf. also Riem. and Goelzer, Phonet., pp. 396, 410, 414. So K.-B1., II, p. 92 f.

289 Cf. Giles, Man. of Comp. Philol., p. 453 f.

290 So Whitney, Sans. Gr., p. 308. Cf. Hirt, Handb. etc., p. 371. Cf. K.-B1., II, p. 30 f., for list.

291 Whitney, Sans. Gr., p. 298.

292 See interesting lists in Sterrett's II., N. 38 ff.

293 V. and D., Handb. etc., p. 79 f.

294 K.-B1., II, p. 102 f. Cf. Brug., Griech. Gr., p. 313; Delbruck, Grundl., etc., IV, pp. 75 ff. Hartmann (De aoristo secundo, 1881, p. 21) makes too much distinction between the second and first aorists.

295 Whitney, Sans. Gr., p. 313.

296 Sterrett, Hom. Il., N. 42.

297 V. and D. Handb., etc., p. 81, but in particular Thumb, Handb., p. 144.

298 Cf. K.-B1., II, p. 93 f.

299 Hirt, Handb. etc., p. 399 f.

300 Sterrett, Hom. IL, N. 42 f.

301 Farrar, Gk. Synt., p. 126. Cf. Monro, Hom. Gr., p. 45.

302 Munro, ib., p. 47.

303 Cf. Schweizer, Perg. Inschr., pp. 180 ff.; Nachm., Magn. Inschr., pp. 162 ff.; Meisterh,, Att. Inschr., pp. 181, 185, 187.

304 Gr. d. griech. Pap., pp. 358-370.

305 Cf. W.-Sch., p. 119.

306 See Helbing, Gr. d. Sept., p.94 f., for similar exx. in the LXX, and Thack., Gr., p. 255.

307 Cf. Blass, Gr. of N. T. Gk., p. 45 f.

308 Ib., p. 43.

309 Ib. Mayser (Gr. d. griech. Pap., p. 369) finds it in the pap. as well as avgagh/sai.

310 Gr. of N. T. Gk., p. 43. Cf. katalei,yh| Mk. 12:19 a.

311 Gr. d. griech. Pap., p. 370.

312 Gr. of N. T. Gk., p. 43.

313 Cf. W.-Sch., p. 105.

314 Cf. Mayser, Gr. d. griech. Pap., pp. 360 ff., for careful discussion and references for further research.

315 So pone,w and fore,w$e% in the LXX. Cf. W.-Sch., p. 105.

316 Blass, Gr. of N. T. Gk., p. 43.

317 Cf. Schmid, Atticismus, IV, p. 594 f.

318 Nachm., Magn. Inschr., p. 171; Schweizer, Perg. Inschr., p. 190 f.

319 Mayser, Gr. d. griech. Pap., p. 381 f. Cf. Reinhold, De Graac., p. 76 f.

320 Cf. W.-Sch., p. 110, for exx. in Jos. and LXX. Cf. also Helbing, Gr. d. Sept., p. 95 f. MSS. simply read - fuh)

321 Grundr., II, pp. 836-1330. In Hom. the same root will form a present in several ways, as e;cw i;sw ivscw ivsca,nw. Cf. Monro, Hom. Gr., p. 40.

322 Giles, Man. of Comp. Philol., p. 423.

323 Blass, Gr. of N. T. Gk., p. 48. 2 Gr., p. 122.

324 Blass, Gr. of N. T. Gk., p. 40.

325 Ib., p. 41. The LXX MSS. show both grhgore,w and sth,kw. Cf. Helbing, Gr. d. Sept., p. 82.

326 Cf. Blass, Gr. of N. T. Gk., p. 41.

327 Cf. Monro, Hom. Gr., p. 34; Hirt, Handb. etc., p. 380.

328 Hirt, ib., p. 383 f.

329 Giles, Man. of Comp. Philol., p. 440.

330 Ib., pp. 445 ff. On the whole subject of contract verbs see Jann., Hist. Gk. Gr., pp. 207 ff.

331 Jann., ib., p. 222.

332 Gr. of N. T. Gk., p. 41. The LXX has these new presents. Thack., p.225.

333 Blass, ib. The LXX MSS. illustrate most of these peculiarities of verbs in the present tense. Cf. Helbing, Gr. d. Sept., pp. 82-84;

334 Giles, Man. of Comp. Philol., p. 436.

335 Grundr., IV, p. 59. Cf. Brug., Grundr., II, 669.

336 Hom. Gr., p. 34.

337 Handb. of Mod. Gk., p. 82.

338 Hirt, Handb. etc., p. 401.

339 Giles, Comp. Philol., p. 446; Whitney, Sans. Gr., p. 333 f.

340 Thumb, Handb., pp. 161 f., 173.

341 Man. of Comp. Philol., p. 446. On the whole subject of "Indo-European Futures" see Hadley, Ess. Phil. and Crit., pp. 184 ff.

342 Sterrett, Hom. IL, N. 38.

343 Giles, Man., p. 447. Cf. Moulton, Prol., p. 184; Riern. and Goelzer, Phonet., p. 438.

344 Ib., p. 446. Cf. also Hirt, Handb. etc., p. 401 f.

345 Sterrett, Hom. Il., N. 27.

346 Giles, Man., p. 446.

347 Griech. Gr., p. 320. This position is accepted by K.-Bl., II, p. 105.

348 Ib., p. 105 f.

349 Handb. etc., p. 403 f.

350 And this pesou/mai is possibly not from petsou/mai, but a change of t to s. Cf. K.-B1., II, p. 107; Brug., Griech. Gr., p. 322; Hirt. Handb., p. 404. Henry (Comp. Gr. of Gk. and Lat., p. 116) considers the Doric future to be the affix of the future twice over, as seso seo.

351 Moulton, Prol., p. 149.

352 Cf. K.-B1., II, p. 106 f.

353 Notes on Orth., p. 163. Cf. Mayser, Gr., p. 356.

354 Gr. of N. T. Gk., p. 42.

355 Ib. But Blass (ib.) prefers evggiei/ (Jas. 4:8) .

356 Ib. See Helbing, Gr. d. Sept., pp. 84 f., 87 f., for the LXX exx. of verbs in - zw)

357 Ib.

358 Notes on Orth., p. 163.

359 Ib.

360 Giles, Man., p. 446 f.

361 Blass, Gr. of N. T. Gk., p. 41 f. Brug. (Griech. Gr., p. 321) considers this a new formation after the aor. subj. suffix. The LXX keeps s. Cf. Helbing, Gr. d. Sept., p. 86; Thack., Gr., p. 230.

362 Brug., Griech. Gr., p. 321.

363 Gr. of N. T. Gk., p. 42.

364 Ib., p. 43.

365 Giles, Man., pp. 410, 427.

366 Ib., p. 447.

367 K.-B1., II, p. 111.

368 Prol., p. 150.

369 Giles, Man., pp. 420, 447.

370 Ib., p. 447.

371 See VI, (e), in this chapter.

372 Blass, Gr. of N. T. Gk., p. 36.

373 Ib p. 204.

374 Gk. Synt., p. 71.

375 Giles, Man., p. 449.

376 Hirt, Handb. etc., pp. 406, 410.

377> Giles, Man., p. 449.

378 Hirt, Handb. etc., p. 410.

379 Riem. and Goelzer, Phonet., p. 445.

380 Sterrett, Hom. II., N. 43. So ge,gona ei;wqa le,loipa pe,poiqa, etc.

381 Gk. Gr., p. 323.

382 Mayser, Gr. d. griech. Pap., pp. 372 ff.

383 Nachm., Magn. Inschr., p. 159 f.

384 Hirt, Handb. etc., p. 412 f.

385 In the LXX h[kamen h[kate h[kasin occur. The pap. add kaqhkui,aj h`ko,twn h`ke,nai. Wackern., Theol. Literaturzeit., 1908, p. 38. Cf. Helbing, Gr. d. Sept., p. 103 f.; Thack., Gr., p. 269. The pap. show the perfect forms in the plural. Mayser, p. 372.

386 Man., p. 450.

387 Giles, Man., p. 451.

388 Brug., Griech. Gr., p. 325.

389 Sterrett, Hom. IL, N. 43.

390 Giles, Man., p. 451.

391 Whitney, Sans. Gr., pp. 279, 295 f.

392 Giles, Man. of Comp. Philol., p. 451.

393 Moulton, Prol., p. 142.

394 Thumb., Handb., p. 165. Certainly the aorists in - ka are very common in the mod. Gk. (Thumb, Handb., pp. 140, 146 ff.).

395 Cf. Moulton, Prol., p. 143 f.

396 Blass, Gr. of N. T. Gk., p. 200 f. Cf. discussion between Prof. Harry and Prof. Sonnenschein in Cl. Rev., 1906, and La Roche, Beitr. z. griech. Gr., 1893.

397 Sterrett, Hom. Il., N. 43.

398 K.-B1., II, p. 2 f.

399 Blass, Gr. of N. T. Gk., p. 201. Brug. calls the past perf. a "neue Bildung."

400 Handb. etc., p. 415 f.

401 So Hirt follows Wackern. in seeing a new stem here eivdh. Cf. ib., p. 416. B in Deut. 8:3 has ei;dhsan like the aorist ei;dhsa from Arist. onwards. Cf. Mayser, Gr., p. 370; Thack., Gr., p. 278.

402 Sterrett, Hom. Il., N. 27.

403 Mayser, Gr. d. griech. Pap., p. 377. In the Boeotian inscr. the past perf. and the fut. perf. are both absent.

404 Blass, Gr. of N. T. Gk., p. 202 f.; Brug. (Griech. Gr., p. 330 f.) points out how in prehistoric times the periphrastic form alone existed in the subj. and opt. middle and passive, as indeed was practically true always for all the voices.

405 Ib., p. 326. Cf. Helbing, Gr. d. Sept., p. 100 f.; Thack., pp. 219 ff., for LXX illustr. of both s and n ( m).

406 Brug., Comp. Gr. (transl.), vol. IV, p. 10. See note there for books on Reduplication. Add Lautensach, Gr. Stud. (1899).

407 Ib., p. 11. Cf. K.-B1., II, p. 8.

408 Brug., Griech. Gr., p. 176. Fritzsche (Ques. de redupl. graeca; Curtius, Stud. zu griech. and lat. Gr., pp. 279 ff.) considers the doubling of the syllable (iteration) the origin of all reduplication like avrari,skw biba,zw)

409 Whitney, Sans. Gr., p. 222.

410 Sterrett, Hom. Il., N. 32.

411 See Jann., Hist. Gr., p. 190 f., for exx. like e;takto even in Polyb., and later gramme,noj, etc.

412 Ib. Cf. Thumb, Handb., p. 148 f.

413 Unters. etc., p. 215.

414 Thumb, Handb., p. 148 f.

415 Hirt, Handb. etc., p. 369.

416 Giles, Comp. Philol., p. 409.

417 Jann., Hist. Gk. Gr., p. 190.

418 Sterrett, Hom. Il., N. 32.

419 Cf. Brug., Comp. Gr. (transl.), IV, p. 384. Cf. also Hirt, Handb. etc. p. 407; Brug., Griech. Gr., p. 259.

420 Ib., Helbing, Gr. d. Sept., pp. 70-82, treats together augment and reduplication, not a very satisfactory method.

421 Hirt, Handb. etc., p. 408.

422 Mayser, Gr. d. griech. Pap., pp. 338

423 Nachm., p. 150 f.; Schweizer, Perg. Inschr., p. 171.

424 Notes on Orth., p. 170.

425 F. 103. Cf. also K.-B1., II, p. 23, and Blass, Gr. of N. T. Gk., p. 38.

426 Moulton (Cl. Rev., Feb., 1901, p. 36) cites avpaith/sqai e`toima,kamen from the pap.

427 Brug., Comp. Gr. (transl.), IV, p. 25. Jann. (Hist. Gk. Gr., p. 185) thinks it is an archaic form of the imperf. of eivmi, ( e en).

428 Whitney, Sans. Gr., p. 221.

429 Sterrett, Hom. Il., N. 30 f.

430 Hist. Gk. Gr., p. 185.

431 Brug., Comp. Gr. (transl.), IV, p. 32.

432 Mayser, Gr. d. griech. Pap., p. 333.

433 W.-Sch., p. 99.

434 See good discussion in Jann., Hist. Gk. Gr., p. 186.

435 Mayser, Gr. d. griech. Pap., p. 336.

436 Jann., Hist. Gk. Gr., p. 185.

437 Ib., p. 186. Hence in mod. Gk. temporal augment is nearly gone. Already in the LXX the movement toward the loss of the temporal augment is seen (Thack., Gr., pp. 196, 199 f.). The pap. often have - eire,qhn for -- h|re,qhn (Mayser, pp. 127, 335).

438 See W.-Sch., p. 100 f. Cf. Hort, Notes on Orth., p. 162 f.

439 Meist erh., Att. Inschr., p. 169.

440 So inscr. Letronne, Rec. II, p. 463.

441 W.-Sch., p. 103.

442 Hort, Notes on Orth., p. 163.

443 Dieterich, Unters., p. 209.

444 Comp. Gr. (transl.), II, p. 471.

445 Whitney, Sans. Gr., p. 203. On these infs. in posse see Brug., Comp. Gr., IV, p. 599.

446 Whitney, ib., p. 347. Cf. ger. of Lat. For special treatises on the inf. see Brug., Comp. Gr. (transl.), IV, pp. 595 ff.; Griech. Gr., p. 359. Cf. also Grunewald, Der freie formelhafte Inf. der Limitation im Griech. (1888); Birklein, Entwickelungsgesch. des substant. Inf. (1888); Votaw, The Use of the Inf. in Bibl. Gk. (1896); Allen, The Inf. in Polyb. compared with Bibl. Gk. (1907). Jann. (Hist. Gk. Gr., pp. 480 ff., 568 ff.) has a very good sketch of the history of the inf. in Gk. On p. 572 f. he discusses John's use of the inf. with verbs (129 exx.). Cf. Jolly, Gesch. des Inf. im Indog. (1873); Gildersleeve, Contrib. to the Hist. of the Articular Inf. (Transl. Am. Phil. Ass., 1878, A. J. P., vol. III, pp. 193 ff.; vol. VIII, pp. 329 ff.; vol. XXVII, p. 105 f.).

447 Brug., Comp. Gr. (transl.), II, p. 471.

448 Hirt, Handb. etc., p. 433.

449 Moulton, Prol., p. 203.

450 Hirt, Handb., p. 431.

451 Moulton, Prol., p. 204.

452 Votaw, Use of the Inf. in Bibl. Gk., p. 59.

453 Moulton, Prol., p. 204.

454 Mayser, Gr. d. griech. Pap., p. 385. Cf. Moulton, Cl. Rev., Feb., 1901, p. 36 f. Cf. Hatz., Einl., p. 190.

455 Brug., Comp. Gr. (transl.), IV, p. 7.

456 K.-B1., II, p. 4.

457 Brug., Comp. Gr. (transl.), p.7.

458 Cf. Giles, Man. of Comp. Philol., p. 469 f.; Brug., Grundr., II, 1093. 8.

459 Clyde, Gk. Synt., p. 90.

460> Hirt, Handb., p. 432; Giles, Man., p. 470.

461 Moulton, Prol., p. 202.

462 Monro, Horn. Gr., p. 154.

463 Moulton, Prol., p. 213 f.

464 Gk. Synt., p. 164.

465 Blass, Gr. of N. T. Gk., p. 221. Thumb (Handb. of Mod. Gk.) has no discussion of the infinitive.

466 Farrar, Gk. Synt., p. 169. Cf. Donaldson, New Crat., p. 603.

467 Farrar, Gk. Synt., p. 169.

468 Gk. Synt., p. 94.

469> II, p. 4.

470 19.

471 Brug., Comp. Gr., IV, p. 605.

472> Ib., II, p. 456,

473 Giles, Comp. Philol., p. 474.

474 Ib.

475 Prol., 221.

476 Ib.

477 Gr. of N. T. Gk., p. 37.

478 Prol., p. 222.

479 Brug., Comp. Gr., IV, p. 605.

480 Sterrett, Hom. Il., N. 28.

481 Hirt, Handb., p. 438. Moulton (Cl. Rev., Mar., 1904, p. 112) finds one ex. of - te,oj in the pap. and "the - toj participle is common in neg. forms." Note that he calls it a participle.

482 Brug., Comp. Gr., II, p. 457.

483 Whitney, Sans. Gr., p. 347.

484 Indog. Forsch., V, pp. 89 ff. Cf. Moulton, Prol., p. 221.

485 Whitney, Sans. Gr., p. 202.

486 Giles, Comp. Philol., p. 474.

487 Hirt, Handb., p. 436 f.

488 Giles, Comp. Philol., p. 473. Cf. the Sans. passive part. in -ta or -na, Whitney, Sans. Gr., p. 340.

489 Jann., Hist. Gk. Gr., p. 206. Cf. Hatz., Einl., p. 1432%

490 Farrar, Gk. Synt., p. 169.

491 Blass, Gr. of N. T. Gk., p. 37. He cites elsewhere Mt. 27:49, sw,swn, Jo. 6:64, 1 Cor. 15:37; Heb. 3:5; 13:17; 1 Pet. 3:13. Then there are the doubtful forms kausou,mena (2 Pet. 3:10, 12) and komiou,menoi (2 Pet. 2:13).

492 Brug., Comp. Gr., IV, p. 444.

493 Whitney, Sans. Gr., p. 394.

494 Brug., Comp. Gr., IV, p. 446.

495 Brug., Griech. Or., p. 331. Kektw/mai and kekth|,mhn had no following in Gk.

496 Blass, Or. of N. T. Gk., p. 204. I am chiefly indebted to Blass for the facts in this summary.

497 Gr. of N. T. Gk., p. 203.