1050

CHAPTER XX

VERBAL NOUNS ( vONOMATA TOU `RHMATOS)

I. Kinship. The finite verb, verbum finitum (das bestimmte Verb), has now been discussed as adequately as the space in this grammar allows. Originally there was no difference between verb and noun (see Conjugation of the Verb). But gradually there was developed a difference. It was done largely by the help of the pronouns which were added to the verb-stems. Nouns also had their own inflection. But a considerable body of words partook of the nature of both verb and noun and yet did not cut loose from either. In a sense therefore the finite verb is a combination of verb and pronoun while the non-finite verb combines verb and noun. These verbal nouns are the non-finite verb, verbum infinitum (das unbestimmte Verb).1 They failed to add the personal pronominal endings of the finite verb and so did not become limited to a subject (finite). And yet they developed tense and voice and were used with the same cases as the finite verb. In so far they are true verbs. On the other hand they are themselves always in a case like other nouns. The verbal substantive comes to drop its inflection (fixed case-form) while the verbal adjective is regularly inflected in the singular and plural of all three genders just like any other adjective. These verbal nouns may be regarded either as hybrids or as cases of arrested development, more properly deflected development, for they continued to develop in a very wonderful way. The Greek of the Attic period would be barren indeed if robbed of the infinitives and the participles. The names are not distinctive, since both are participles2 (partake of the nature of both verb and noun) and both are non-finite or infinitives (are not limited to a subject by personal endings). The root-difference between these lies not

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1051

in the verbal idea, but in the noun. It is the difference between substantive and adjective. Both are verbals both are nouns, but one is a substantive and the other is an adjective. These general remarks may help one to understand the history and usage of both infinitive and participle.

II. The Infinitive ( h` avpare,mfatoj e;gklisij or avpare,mfaton r`h/ma)

1. ORIGIN. There is no real ground for difference of opinion on this subject, however much scholars may argue as to the significance of the infinitive.3 In the Sanskrit the infinitive did not have tense or voice. The root used was that of a substantive closely connected with a verb.2 But it is verbal in Sanskrit also in the notion of action, nomina actionis. In the Veda and Brahmana the number of these verbal nouns is very large. They are used with cases, the cases corresponding to the verb, but that phenomenon appears in Latin and Greek. In Plautus "we even find the abstract noun tactio in the nominative governing its case just as if it were tangere. Classical Greek has a few wellknown examples of a noun or adjective governing the case appropriate to the verb with which it is closely connected."4 The same thing occurs in the N. T. also. Cf. koinwni,a fwti, (2 Cor. 6:14). See chapter on Cases. These substantives have enough "verbal consciousness " to "govern" cases.5 In the old Sanskrit these verbal substantives occur in any case (except the vocative, which is not a real case). The later Sanskrit has only one such case-ending so used, the accusative in -tum or -itum (cf. the Latin supine).6 But for the developments in other languages, especially in the Greek and Latin, these Sanskrit verbal substantives would not have been called infinitives. But they show beyond controversy the true origin of the infinitive before tense and voice were added. They were originally substantives in any case, which were used as fixed case-forms (cf. adverbs) which had a verbal idea (action), and which were made on verbal roots. The Latin shows three cases used in this way: the locative as in regere, the dative as in regi and the accusative as in the supine rectum.7 The Greek infinitive shows only two caseendings, the dative - ai as in lu,sai (cf. also doF e,nai, dou/nai, with Sanskrit davane; Homeric F i,dmenai with Sanskrit vidmane) or the

1052 A GRAMMAR OF THE GREEK NEW TESTAMENT

locative in lu,ein.8 Thus in the Greek and Latin it is only oblique cases that were used to form the infinitives.9 It is then as a substantive that the infinitive makes its start. We see this in the Sanskrit davane vasunam= dou/nai tw/n avgaqw/n.10 This substantive aspect is clearly seen in the use of panto,j with tou/ zh/n in Heb. 2:15. The first11 step towards the verbal idea was in the construction dou/nai ta. avgaqa,. Moulton12 illustrates the border-land of the English inf. by the sentence: "He went out to work again." If we read "hard work" we have a substantive; but if we read "work hard," we have a verbal notion. Strictly speaking, dou/nai ta. avgaqa,= 'for giving the good things,' while ivdei/n ta. avgaqa,Ê'in seeing the good things.' This was the original etymological sense as the Sanskrit makes clear. See further chapter on Conjugation of Verb.

2. DEVELOPMENT. In the Sanskrit we see the primitive infinitive without tense or voice. In the modern Greek the infinitive, outside of the Pontic dialect, has disappeared save with auxiliary verbs, and even so it is in a mutilated state, as with qe,lei lu,eià h;qelea deqei/à e;cw de,sei, remnants of the ancient infinitives lu,einà deqh/naià de,sai (Thumb, Handb., pp. 162, 167). Between these two extremes comes the history of the rise and fall of the Greek infinitive. We may sketch that history in five periods.13

(a) The Prehistoric Period. The infinitive is simply a substantive with the strict sense of the dative or locative case. Cf. the Sanskrit. We may infer also that there was no tense or voice. This original epexegetical use of the inf. as the dative of limitation has survived with verbs, substantives and adjectives. So o` cro,noj tou/ tekei/n (Lu. 1:57). Cf. our "a wonder to behold." See du,nati douleu,ein (Mt. 6:24), o`rmh. u`bri,sai (Ac. 14:5), i`kano,j lu/sai (Mk. 1:7). See also Jas. 1:19, tacu.j eivj to. avkou/sai, where eivj to. reproduces the dative idea.

(b) The Earliest Historic Period. The case-form (dative or locative) begins to lose its significance. In Homer the dative idea is still the usual one for the infinitive, in harmony with the form.14 With verbs of wishing, commanding, expecting, beginning, being able, etc., the dative idea is probably the original explanation of

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1053

Addenda 3rd ed.

the idiom. Cf. oi;date dido,nai (Mt. 7:11), 'knows how to give' (for 'giving'). Homer has bh/ d v ive,nai= 'stepped' for 'going.' But already in Homer there are signs that the case-form is getting obscured or stereotyped. It occurs as apparent subject with impersonal verbs and as the logical object of verbs of saying in indirect discourse.15 The use of pri,n with the inf. is common also in Homer. Pri,n would naturally be used with the ablative, like pura and the infinitive in Sanskrit,16 and so the Greek idiom must have arisen after the dative or locative idea of the inf. in Greek was beginning to fade.17 In Homer the inf. is already a fixed case-form. The disappearance of - ai as a distinct case-ending in Greek may have made men forget that the usual inf. was dative. This dative inf. was probably a survival of the old and once common dative of purpose. Gradually the inf. passed from being merely a word of limitation (epexegetic) to being subject or object. We see the beginning of this process in Homer, though there is only18 one instance of the article with the inf., and that is in the Odyssey (20. 52), to. fula,ssein. But even here to, may be demonstrative.19 But in Homer the inf. has tense and voice, a tremendous advance over the Sanskrit inf. This advance marks a distinct access of the verbal aspect of the inf. But there was no notion of time in the tense of the inf. except in indir. discourse where analogy plays a part and the inf. represents a finite mode.20 This use of the inf., afterwards so common in Latin, seems to have been developed first in the Greek.21 But it was the loss of the dative force as an essential factor that allowed the inf. to become distinctly verbalized.22 As it came to be, it was an imperfect instrument of language. As a verb it lacked person, number and time except in indirect discourse. As a substantive it lacked inflection (without case or number) after it came to be limited to two cases. Even after the case-idea vanished and it was used in various cases it was still indeclinable.23

1054 A GRAMMAR OF THE GREEK NEW TESTAMENT

The addition of tense and voice to the fixed case-form of the substantive with verbal root was possible just because of the obscuration of the case-idea.

(c) The Classic Period from Pindar on. The articular infinitive is often used and there is renewed accent on its substantival aspects. The inf. is freely used with or without the article in any case (except vocative) without any regard to the dative or locative ending. Pindar first uses the neuter article to, with the inf. as the subject.24 "By the assumption of the article it was substantivized again with a decided increment of its power."25 It is to be remembered, however, that the article itself is a development from the demonstrative and was very rare in Homer with anything. Hence too much must not be made of the later use of the article with the inf. Hesiod shows two examples of the article with the inf. Pindar has nine and one in the accusative.26 The absence or ambiguous character of the article in early Greek makes it necessary to be slow in denying the substantival aspect or character of the inf. in the Homeric period.27 Hence it is best to think of the article as being used more freely with the inf. as with other nouns as the article made its onward way. The greatly increased use of the article with the inf. did serve to restore the balance between the substantival and verbal aspects of the inf. now that tense and voice had come in. The enlarged verb-force was retained along with the fresh access of substantival force. "The Greek infinitive has a life of its own, and a richer and more subtle development than can be found in any of the cognate languages."28 The infinitive, thus enriched on both sides, has a great career in the classic period of the language, especially in Thucydides, the Orators, Xenophon and Plato. It has a great variety of uses. In general, however, it may be said that the inf. was not as popular in the vernacular as in the literary style for the very reason that it was synthetic rather than analytic, that it lacked clearness and emphasis.29 But it was not till the koinh, period that the inf. began to disappear.30

(d) The Koinh, Period. The inf. begins to disappear before i[na

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1055

ion the one hand and o[ti on the other. Jannaris31 outlines the two chief functions of the inf. in its developed state to be prospective (purpose like i[na) and declarative (subject or object like o[ti and i[na ultimately also). The fondness for analysis rather than synthesis, particularly in the vernacular, gradually pushed the inf. to the wall. The process was slow, but sure. There is indeed a counter tendency in the enlarged use of tou/ and the inf. in the koinh,, particularly in the LXX under the influence of the Hebrew infinitive construct, and so to some extent in the N. T. So from Polybius on there is seen an increase of tou/ and the inf. side by side with the enlarged use of i[na and o[ti. The two contradictory tendencies work at the same time.32 On the whole in the koinh, the inf. has all the main idioms of the classic age (with the marked absence of evf v w|-te) and the new turn given to tou/ and evn tw|/. The Hebrew did not use the inf. as much as the Greek and never with the article. Certainly the inf. is far less frequent in the LXX than in the comparatively free Greek of the N. T., about half as often (2.5 to the page in the LXX, 4.2 in the N. T.).33 But the Hebrew has not, even in the LXX, introduced any new uses of the inf. in the Greek. The Hebrew inf. construct had no article and was thus unlike tou/ and the inf. The total number of infinitives in the N. T., according to Votaw,34 is 2,276. The number of anarthrous infs. is 1,957, of articular 319. The inroad of i[na and o[ti is thus manifest as compared with the Attic writers. The writings of Luke show the largest and most varied use of the inf., while the Johannine writings have the fewest.35 Paul's use is very uneven. Votaw36 finds the same inequality in the case of the apocryphal books. The papyri show a similar situation. Different writers vary greatly, but on the whole the inf. is dying save in the use with auxiliary verbs, and it is going even there as is seen from the use of i[na with qe,lw in the N. T. Cf. Mk. 9:30. In the koinh, we find i[na with bou,lomai and du,namai in Polybius, the LXX and later koinh, writers.37 As the inf. disappears in the later Greek strange combinations appear, as in Malalas and Theophanes we

1056 A GRAMMAR OF THE GREEK NEW TESTAMENT

meet pro. tou/ with the subjunctive ( pro. tou/ evpirri,ywsinà pro. tou/ e`nwqw/sin).38 The inf. never had a monopoly of any construction save as the complement of certain verbs like bou,lomaià qe,lw, etc. This was probably the origin use of the inf. with verbs and it was true to the dative case-idea.39 It was here alone that the inf. was able to make a partial stand to avoid complete obliteration.

(e) The Later Period. Outside of the Pontic dialect the inf. is dead, both anarthrous and articular, save with the auxiliary verbs.40 The use of qe,lw as a mere auxiliary is common enough in Herodotus and probably was frequent in the vernacular then as it was later.41 "The fortunes of the infinitive were determined by its nature."42 The increased use of abstract nouns made it less needed for that purpose, as the fondness for i[na and o[ti made it less necessary as a verb. The N. T. is mid-stream in this current and also midway between the rise and the end of this river. The writers will use the inf. and i[na side by side or the inf. and o[ti parallel. Even in the classical Attic we find o[pwj after peiÄ ra,omai (Xenophon).43 As o[pwj disappeared i[na stepped into its place. In Latin ut was likewise often used when the inf. could have occurred. The blending of i[na and o[ti in the koinh, helped on the process.

In the N. T. the exclusive province of the inf. is a rather narrow44 one. It still occurs alone with du,namai and me,llw. It has a wide extension of territory with tou/. But on the whole it has made distinct retreat since the Attic period. The story is one of the most interesting in the history of language.

3. SIGNIFICANCE. Originally as we have seen, the infinitive was a substantive, but a verbal substantive. This set case of an abstract substantive has related itself closely to the verb.45 The Stoic grammarians46 called it a verb, avpare,mfaton r`h/maà avpare,mfaÄ toj e;gklisij. Apollonius Dyskolos47 called it a "fifth mode" and the later grammarians followed his error. Some of the Roman grammarians actually took infnitivus in the sense perfectus,

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1057

just as they mistranslated genikh, by genitivus.48 Bopp49 rightly perceived that the inf. has a nominal origin and was later adjusted to the verb in Greek. It is not a real verb in the very height of its glory.50 And yet the consciousness of the nominal origin was partially obscured even in the time of Homer. The original case-form is so far forgotten that this dative may appear in the nominative and the accusative. The tenses and voices have developed. But Brugmann51 seems to go too far in saying that already the inf. was "only" a verb in the popular feeling. Moulton,52 indeed, harks back to Apollonius Dyskolos: "The mention of 'The Verb' has been omitted in the heading of this chapter, in deference to the susceptibilities of grammarians who wax warm when lu,ein or lu,saj is attached to the verb instead of the noun. But having thus done homage to orthodoxy, we proceed to treat these two categories almost exclusively as if they were mere verbal moods, as for most practical purposes they are." He states, it is true, that every schoolboy knows that in origin and part of the use the inf. is a substantive, but "nearly all that is distinctive is verbal."53 I venture to say that this is overstating the case. It is not a mere question of the notion of the user of the infinitive in this passage or that. The history is as it is. In the full development of the inf. we see the blending of both substantive and verb. In this or that example the substantival or the verbal aspect of the hybrid form may be dominant, but the inf. in the historical period is always both substantive and verb. It is not just a substantive, nor just a verb, but both at the same time. The form itself shows this. The usage conforms to the facts of etymology. It is not true that the article makes the inf. a substantive as Winer54 has it. As a matter of fact, therefore, the inf. is to be classed neither with the noun nor with the verb, but with the participle, and both stand apart as verbal nouns. The article did enlarge55 the scope of the inf. just as the use of tense did. The Germans can say das Trinken and French le savoir like the Greek to. gnw/nai. There is no infinitive in Arabic. As a matter of fact, the inf. because of its lack of endings (here the participle is better off with the adjective endings) is the least capable of all parts of speech of fulfilling its functions.56

1058 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

In its very nature it is supplementary. It is either declarative or prospective,57 but always a verbal substantive. There is a difference between to. pra,ssein and h` pra/xij. Both have, verbal stems and both are abstract. The difference58 lies in the tense and voice of pra,ssin. But pra,ssein has all that is in pra/xij plus tense and voice. I decline, therefore, to divide the infinitive into the anarthrous and articular uses so popular in the grammars. These uses do exist, but they simply represent two uses of the inf. in its substantival aspects. They do not affect the verbal side of the inf. at all. The inf. may properly be discussed under its substantival and its verbal aspects. But even so a number of uses cross over as indirect discourse, for instance, or the inf. to express purpose (with or without the article). We must look at both sides of the inf. every time to get a total idea of its value. A number of points of a special nature will require treatment.

4. SUBSTANTIVAL ASPECTS OF THE INFINITIVE.

(a) Case (Subject or Object Infinitive). Here I mean the cases of the inf. itself, not the cases used with it. The inf. is always in a case. As a substantive this is obvious. We have to dismiss, for the most part, all notion of the ending (dative or locative) and treat it as an indeclinable substantive. A whole series of common expressions has the inf. as subject besides the ordinary verbs. Thus note 1 Cor. 9:15 kalo,n moi ma/llon avpoqanei/n, (Heb. 4:6; 9:27) avpo,keitai toi/j avnqrw,poij a[pax avpqanei/n, (Mt. 18:13) eva.n ge,nhÄ tai eu`rei/n avuto,,grk grk(3:15) pre,pon evsti.n h`mi/n plhrw/sai, (Ac. 21:35) sune,bh basta,zesqai, (Lu. 6:12) evge,neto evxelqei/n auvto,n,grk grk(18:25) euvkoÄ pw,tero,n evstin eivselqei/n, (Jo. 18:14) sumfe,rei avpoqanei/n, (Mt. 22:17) e;xestin dou/nai, (Heb. 9:5) ouvk e;stin nu/n le,gein, (Ac. 27:24) dei/ paraÄ sth/nai, (Ac. 2:24) h=n dunato.n kratei/sqai, (Ph. 3:1) ta. auvta. gra,Ä fein ouvk ovknhro,n. So Ac. 20:16; 2 Pet. 2:21. All this is simple enough. The articular inf. is likewise found in the nominative as in Mk. 9:10, ti, evstin to. evk nekrw/n avnasth/nai. Here the article is not far removed from the original demonstrative. Cf. 10:40, to. kaqi,sai ouvk e;stin evmo.n dou/nai, where dou/nai is probably the original dative 'for giving.' One naturally feels that the articular inf. is more substantival than the anarthrous, as in Ro. 7:18, to. qe,Ä lein para,keitai, moi, but that is no correct. The subject-inf. occurs freely both with and without the article in the N. T. as in the koinh, generally. See Mt. 15:20 to. fagei/n, (Mk. 12:33) to.

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1059

Addenda 3rd ed.

avgapa/n, (Ro. 7:18) to. qe,lein and to. katerga,zesqai. Add 1 Cor. 7:26; 11:6; 2 Cor. 9:1; Ph. 1:21, 24, 29; Heb. 10:31; Ro. 14: 21. The origin of this nominative or subject is probably due to its use with impersonal expressions. Moulton59 illustrates it by the Latin humanum est errare, where the force of the locative form errare may be seen by translating: 'There is something human in erring.' This may have been the original idiom, but it has gone beyond that to mean: 'Erring is human.' English students often forget that 'erring' is here infinitive, not participle, both in sense and history. It is a step further in the N. T. to see tou/ and the inf. used as subject nominative. Cf. Lu. 17: 1; Ac. 10:25; 1 Cor. 16:4. In 2 Cor. 7:11 the substantival aspect of the inf. is shown by the use of the pronoun auvto. tou/to to. luphqh/nai in the nominative with kateirga,sato. Cf. the inf. in the predicate nom. with tou/to in Ro. 1:12, tou/to de, evstin sunparaÄ klhqh/nai. So in Ro. 13:11, w[ra h;dh u`ma/j evx u[pnou evgerqh/nai, where the inf. is in predicate apposition with w[ra. Originally it was doubtless 'time for arising.' In 1 Th. 4:6 we have both the anarthrous and articular inf. in apposition with tou/to. Cf. also the appositive inf. in Ac. 15:28; Jas. 1:27; 1 Th. 4:3; Ro. 4:13.

The object-infinitive in the accusative is quite common both with and, particularly, without the article. In the N. T. more than half of the instances of the inf. come in here, the object-inf. with verbs of various sorts.60 In the LXX, however, it is rare in proportion to the other uses. The accusative case is to us more manifest when the article occurs. See Ph. 2:6, ouvc a`rpagmo.n h`gh,sato to. ei=nai i;sa qew|/, where the articular inf. is the direct object of h`gh,sato. So in 2:13, with o` evnergw/n kai. to. qe,lein kai. to. evnergei/n. Cf. Ac. 25:11, ouv paraitou/mai to. avpoqanei/n. See further 1 Cor. 14:39; 2 Cor. 8:10. In Ph. 4:10, avneqa,lete to. u`pe.r evmou/ fronei/n, the acc. may be that of general reference. Certainly in 1 Th. 3:3, to. sai,nesqai, this is true. Blass61 calls it here "quite superfluous." In Ro. 14:13 to. mh. tiqe,nai. is in apposition with the accusative tou/to, as in 2 Cor.2:1. In 2 Cor. 10:2, de,omai to. mh. parw.n qarrh/sai, we should naturally look for the ablative with de,omai. The instances without the article are more numerous. A fairly complete list of the verbs in the N. T. that have the inf. in indirect discourse was given in the chapter on Modes (Indirect Discourse, pp. 1036 ff.). These infs. are in the acc.,

1060 A GRAMMAR OE THE GREEK NEW TESTAMENT

Addenda 3rd ed.

though some of them may possibly preserve the original dative or locative idea. But the acc. with the inf. is that of general reference, while the inf. itself is in the acc. case, the object of the verb of saying or thinking. Cf. Lu. 2:44, nomi,santej auvto.n ei=nai. The occasional use of the nom. predicate, as in Ph. 4:11, e;maqon auvta,rkhj ei=nai, accents the acc. character of the object-inf. This point is clear also in the case of indirect commands where the noun or pronoun is in the dative and the inf. in the acc., as in 1 Cor. 5:11, e;graya u`mi/n mh. sunanami,gnusqai. The illustrations are numerous and need not be multiplied (see list under Indirect Discourse). With bou,lomaià du,namaià qe,lw the dative makes a good idea and was probably so understood in the beginning.62 It may be questioned, however, if in actual usage this idiom is not also the acc. Cf. Mt. 1:19 evboulh,qh avpolu/sai,grk grk(1:20) mh. fobhqh|/j paraÄ labei/n,grk grk(5:34) le,gw u`mi/n mh. ovmo,sai,grk grk(16:12) ouvk ei=pen prose,cein, (Lu. 18:1) pro.j to. dei/n proseu,cesqai (both infs. in the acc., one with pro,j, the other general reference with dei/n), (Ro. 15:8) le,gw Cristo.n dia,konon gegenh/sqai (cf. Ac. 27:13), (2 Cor. 10:2) logi,zomai tolmh/sai, (1 Th. 4:11) parakalou/men perisseu,ein kai. filotimei/sqai h`suca,zein kai. pra,ssin ta. i;dia kai. evrga,zesqai (note the interrelation of these infs.). See further Mk. 9:28; 12:12; Lu. 16:3; Jo. 5: 18; Ro. 14:2; Gal. 3:2; 1 Cor. 10:13. In the acc. also are the articular infs. with prepositions like eivj (Ro. 1:11); dia, (Ac. 8: 11); meta, (Lu. 22:20); pro,j (Mt. 5:28).

But the inf. occurs in the other oblique cases also with more or less frequency. The genitive, for instance, appears with the prepositions avnti, (Jas. 4:15); dia, (Heb. 2:15, dia. panto.j tou/ zh/n%* e[neka (2 Cor. 7:12); e[wj (Ac. 8:40). The only instance of an attribute with the infinitive in the N. T. is Heb. 2:15, except in apposition with tou/to. It was rare in classic Greek and confined to pronouns. Cf. to. auvtou/ pra,ttein, Plato, Rep. 433. The genitive may be found with evpilanqa,nomai as in Mk. 8:14, evpela,qonto labei/n (cf. evpilaqe,sqai tou/ e;rgou in Heb. 6:10. But we have ta. ovpi,sw in Ph. 3:13). At any rate in Lu. 1:9, e;lace tou/ qumia/sai (cf. 1 Sam. 14:47), we have an undoubted genitive. Cf. also metemelh,qhte tou/ pisteu/sai (Mt. 21:32). The very common use of tou/ with the inf. must also be noted. Most of these are genitives, as in tou/ avpole,sai (Mt. 2:13). The free use of tou/ with the inf. where the case is not genitive will be discussed under a special section under the article with the inf. Cf., for instance, Lu. 17:1; Ac. 10:25; 20:3; 27:1. The gen. occurs

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1061

Addenda 3rd ed.

Addenda 3rd ed.

Addenda 3rd ed.

with substantives just as other substantives are used. This is a fairly common idiom. See Ac. 27:20 evlpi.j pa/sa tou/ sw,zeÄ sqai, (1 Cor. 9:10) evp v evlpi,di tou/ mete,cein, (Ro. 15:23) evpipo,qeian de. e;cwn tou/ evlqei/n, (1 Pet. 4:17) kairo.j tou/ a;rxasqai, (Heb. 5:12) crei,an tou/ dida,skein. Note, in particular, Ro. 11:8, e;dwken auvtoi/j o` qeo.j pneu/ma katanu,xewjà ovfqalmou.j tou/ mh. ble,peinà kai. w=ta tou/ mh. avkou,ein, where the infs. are parallel with katanu,xewj. Cf. Lu. 1:57, 74; 2:6; 10:19; 21:22; 22:6, etc. Note especially Ph. 3:21, kata. th.n evne,rgeian tou/ du,nasqai auvto.n kai. u`pota,xai. Let these suffice. They illustrate well how the inf. continued to be regarded as a real substantive. The genitive occurs also with adjectives as in bradei/j tou/ pisteu/sai (Lu. 24:25); e[toimoi, evsmen tou/ avnelei/n (Ac. 23:15). The genitive is found with a;xioj (the anarthrous inf.) as in Lu. 15:19, 21, a;xioj klhqh/nai (cf. Rev. 5:4, 9). In 1 Cor. 16:4 tou/ poreu,esqai may be due to a;xion, but is probably used as subj. nominative in a rather loose way. The inf. ins the genitive is specially common in Luke and also in Paul.63

The ablative illustrations are not very numerous, but they are clear. Thus we have the abl. with verbs of hindering as in Mt. 19:14, mh. kwlu,ete auvta. evlqei/n pro,j me, and Lu. 4:42, katei/con auvto.n tou/ mh. poreu,esqai. The classical Greek had also to, and the inf., as in 1 Cor. 14:39, and to. mh, after verbs of hindering, which last does not occur in the N. T., so that it is probable that an inf. without the art. as in Mt. 19:14 is in the abl., though not certain. Moulton (Prol., p. 220) illustrates Lu. 4:42 and Ac. 14: 18 by B. U. 164 (ii/iii A.D.) pei/sai auvto.n tou/ evlqei/n, J.H.S., 1902, 369 (Lycaonian inscription) tw|/ dicotomh,santi, me tou/ to. loepo.n zh/n, B. U. 36 (ii/iii A.D.) tou/ zh/n metasth/sai, N. P. 16 (iii/A.D.) kwlu,ontej tou/ mh. spei,rein. See further Lu. 24: 16 evkratou/nto tou/ mh. evpignw/nai auvto,nà Ac. 10:47 du,natai kwlu/sai, tij tou/ mh. baptisqh/nai, 14:18 kate,pausan tou/ mh. qu,ein. Cf. also Ac. 20:20, 27; Ro. 11:10; 15:22; 2 Cor. 1:8; Heb. 7:23; 1 Pet. 3:10. Cf. in the LXX, Gen. 16:2; 20:6; Ps. 38:2; 68:24 (quoted in Ro. 11:10); Is. 24:10; 1 Sam. 8:7; Jer. 7:10.64 The abl. occurs also with prepositions as evk in 2 Cor. 8:11, evk tou/ e;cein and pro,, in Mt. 6:8 pro. tou/ aivth/sai. In Ac. 15:28, tou,twn tw/n evpa,nagkejà avpe,cesqai, the inf. is in the abl., in apposition with the preceding words.

The only instance of the inf. in the instrumental in the N. T. occurs in 2 Cor. 2:13, tw|/ mh. eu`rei/n me Ti,ton. The inf. is not found with su,n in the N. T. Votaw (Inf. in Biblical Greek, p. 29) notes six examples of the instrumental tw|/ and the inf. in, the LXX text

1062 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

of B (2 Chron. 28 : 22; Eccl. 1 : 16; Is. 56 : 6; 4 Macc. 17: 20, 21). But other MSS. vary. Moulton (Prol., p. 220) cites L. Pb. (ii/B.C.), a;llwj de. tw|/ mhqe,n v e;cein.

The locative occurs with evn as in evn tw|/ euvlogei/n (Lu. 24:51). It is extremely frequent in the N. T., especially in Luke. The possible Hebraistic aspect of the idiom comes up under Prepositions with the Inf. There remains, of course, a possible locative use of a form like lu,ein. But one doubts if this original idea is preserved in the N. T.65 Cf. Mt. 16:3, ginw,skete diakri,nein, which is more naturally explained as a dative: 'ye have knowledge for discerning,' though 'in discerning' makes sense. But with the dative it is different. There is no instance of the dative inf. with a preposition, but the original dative is clear in all examples of purpose without tou/ or a preposition. Thus Mt. 5:17, ouvk h=lqon katalu/saià avlla. plhrw/sai, 'I came not for destroying, but for fulfilling.' So Lu. 12:58, do.j evrgasi,an avphlla,cqai, 'give diligence for being reconciled.' Cf. Mt. 7:11; 16:3 with oi=da and ginw,skw. See further Mt. 2:2, h;lqomen proskunh/sai, 'we came for worshipping'; Jo. 21:3, u`pa,gw a`lieu,ein, 'I go a-fishing.' So Ro. 3: 15, LXX, ovxei/j evkce,ai ai-ma, 'swift for shedding blood.' The substantive also has the dative inf. in Ro. 9:21, evxousi,an poih/sai, 'power for making.' See further 1 Pet. 4:3, kateirga,sqai, 'for having wrought'; Gal. 5:3, ovfeile,thj poih/sai, 'debtor for doing'; Heb. 11:15, kairo.n avnaka,myai, 'time for returning.' This was the original idiom and, with all the rich later development as verbal substantive, the inf. did not wholly get away from the dative idea.

(b) The Articular Infinitive. We have to cross our tracks frequently in discussing the inf. in a lucid fashion. Numerous examples of the articular inf. have already been given in treating the cases of the inf. But the matter is so important that, it calls for special investigation. If we pass by the doubtful articular inf., to. fula,ssein, in the Odyssey,66 we still find (cf. p. 1054) a few examples in the oldest Greek (two in Hesiod, nine in Pindar, nine in the Lyrics).67 The use of the article with the inf. grew with the growth of the article itself. But it is not to be overlooked that in Homer the anarthrous inf. had already developed nearly

VERBAL NOUNS ( vONOMATA TOU`RHMATOS) 1063

all the constructions of this verbal substantive.68 The addition of the article made no essential change in the inf. It was already both substantive and verb. But the use of the article greatly enlarged the range of the inf. It is extended to new uses, especially with prepositions. The article was first used with the nom., then the acc. and then the other cases. The use of tou/ and tw|/ with the inf. is wholly post-Homeric.69 In the Dramatists and Herodotus it is still chiefly in the nom. and acc., though we do find tou/ and tw|/, and we see the inf. used with prepositions also.70 In Thucydides the articular inf. suddenly jumps to great prominence, occurring 298 times,71 especially in the speeches. Of these 163 occur with prepositions.72 He even uses to, with the future inf. and with a;n and the inf. The orators likewise use the art. inf. very freely. It was especially in Demosthenes that "the power of taking dependent clauses" was fully developed.73 Only the Pontic dialects, as already noted, keep the inf. as a living form, and a few substantives preserve a mutilated form) like to. fagi, ('eating') to. fagei/nà to. fili,, ('kissing') = to. filei/n (Thumb, Handb., p. 117). In the N. T. we see all this power still retained with the further development in the use of tou/. The inf. itself, as we have seen, is retreating in the N. T., but it still possesses the full range of its varied uses. The articular inf. has all the main uses of the anarthrous inf. Votaw (The Inf. in Bibl. Gk., p. 51) finds 22 uses of the inf. (19 anarthrous, 15 articular), but some of these overlap and are artificial. Moulton (Prol., p. 214) concludes from a study of the inscriptions that the articular inf. only invaded dialects as the koinh, was starting. There is no essential differenece in idea, and the mere presence or absence of the article is not to be pressed too far. Jannaris74 admits that sometimes the verbal character is completely obscured. On that point I am move than skeptical, since the inf. continues to have the adjuncts of the verb and is used with any voice or tense. Jannaris75 thinks that in late Greek the substantival, aspect grew at the expense of the verbal and the articular inf. had an increasing popularity. I admit the popularity, but doubt the dis-

1064 A GRAMMAR OF THE GREEK NEW TESTAMENT

appearance of the verbal aspect. Jannaris makes the mistake of taking "substantival inf." as coextensive with "articular inf." Blass76 questions if the article always has its proper force with the inf. and suggests that perhaps sometimes it merely occurs to show the case of the inf. Here again I am skeptical. Why does the case of the inf. need to be shown any more than other indeclinable substantives? In Mt. 1 the article does serve to distinguish object from subject. I have never seen an articular inf. where the article did not seem in place. Moulton77 considers the use of the article "the most characteristic feature of the Greek infinitive in post-Homeric language." Blass78 seems puzzled over the frequency of the articular inf. in the N. T., since it is chiefly confined to Luke and Paul, whose writings have most affinity with the literary language. Jannaris79 notes how scarce it is in the writings of John and in unlearned papyri and inscriptions, doubtful in the mediaeval period, and absent from the modern vernacular. "The articular infinitive, therefore, could not resist any longer the tendency of the time, whether it was conceived as a noun or as a verb."80 The analytic tendency drove it out finally. Moulton81 has made some researches on the use of the articular inf. in the dialect inscriptions. He does not find a single instance in Larfield's Boeotian inscriptions. He finds one from Lesbos, one from Elis, one from Delphi, a few from Messene, etc. He notes the silence of Meisterhans on the subject. The conclusion seems to be inevitable that the articular inf. is as rare in the Attic vernacular as it was common in the Attic orators. It is "mainly a literary use, starting in Pindar, Herodotus and the tragedians, and matured by Attic rhetoric." Aristophanes uses it less than half as often as Sophocles and Aristophanes gives the Attic vernacular. And yet it is not absent from the papyri. Moulton82 counts 41 instances in vol. I of B. U. The N. T. uses it about as often to the page as Plato. He scores a point against Kretschmer's view that the Attic contributed no more to the koinh, than any one of the other dialects, since from the literary Attic "the articular inf. passed into daily speech of the least cultured people in the later Hellenist world."83 Polybius84 deserves to rank with Demosthenes in the wealth of his use of the inf. He employs the

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1065

inf. in all 11,265 times, an average of 7.95 to the page. He has the articular inf. 1,901 times, an average of 1.35 to the page. In the N. T. the inf. occurs 2,276 times, an average of 4.2 times to a page. The articular inf. is found in the N. T. 322 times, an average of .6 times to a page. The N. T. shows fewer uses, in proportion, of the articular inf. than the O. T. or the Apocrypha. Of the 303 (Moulton) instances, 120 are in Luke's writings and 106 in Paul's Epistles. But Votaw85 counts 319 in all. The MSS. vary in a number of instances and explain the difference. Moulton86 gives the figures for all the N. T. books thus: James 7, , Hebrews 23, , Gospel of Luke 71, , Paul 106, Acts 49, 1 Peter 4, Matthew 24, , Mark 13 (14), John 4, , Revelation 1, , not in Col., Philem., Past. Eps., Joh. Eps., 2 Pet., Jud. Luke has the most varied use of the articular inf., and Paul's is somewhat uneven.87 The use of the articular inf. in the various cases has already been sufficiently discussed. In general one may agree with Moulton88 that "the application of the articular infin. in. N. T. Greek does not in principle go beyond what is found in Attic writers." The special use of the articular inf. with prepositions is reserved for separate discussion. There is little doubt that the first use of to, with the inf. was demonstrative as it was with everything.89 In Mk. 9:10, ti, evstin to. evk nekrw/n avnasth/ani, the article is almost demonstrative, certainly anaphoric (cf. verse 9). The same thing is true of 10:40 where to. kaqi,sai refers to kaqi,swmen in verse 37. It is not necessary to give in detail many examples of the articular inf. in the N. T. I merely wish to repeat that, when the article does occur with the inf., it should have its real force. Often this will make extremely awkward English, as in Lu. 2:27, evn tw|/ eivsagagei/n tou.j gonei/j to. paidi,on. But the Greek has no concern about the English or German. It is simply slovenliness not to try to see the thing from the Greek standpoint. But we are not to make a slavish rendering. Translation should be idiomatic. It is hardly worth while to warn the inept that there is no connection between the article to, and the English to in a sentence like Ph. 1:21, evmoi. ga.r to. zh/n Cristo.j kai. to. avpoqanei/n ke,rdoj) Here the article to, has just the effect that the Greek article has with any abstract substantive, that of distinction or contrast. Life and death (living and dying) are set over against each other. See further Mt. 24:45; Lu. 24:29; Ac. 3:12; 10:25; 14:9; 21:

1066 A GRAMMAR OF THE GREEK NEW TESTAMENT

12; 25:11; Ro. 4:11, 13, 16, 18; 13:8; 14:21; 2 Cor. 8:10 f.; 9:1; Ph. 1:23, 29; 2:6; 4:10; 1 Th. 3:2 f.

Some special words are needed about tou/ and the inf. The question of purpose or result may be deferred for separate discussion. We have seen how the genitive inf. with tou/ occurs with verbs, substantives, adjectives and prepositions. The ablative inf. with tou/ is found with verbs and prepositions. The ablative use is not here under discussion, since it involves no special difficulties save the redundant mh,. We may note that in Critias tou/ was very common with the inf.90 We see it also in Polybius in various uses named above.91 It is an Attic idiom that became very common in the postclassical and Byzantine Greek.92 Cf. mh. avmelh,sh|j tou/ evnoclh/sai qwni,w|, O. P. 1159, 11-13 (iii/A.D.). There is no special difficulty with tou/ and the inf. with verbs as object except in a case like Mt. 21:32 where tou/ pisteu/sai "gives rather the content than the purpose of metemelh,qhte."93

The instances with substantives like Ac. 14:9, e;cei pi,stin tou/ swqh/nai, give no trouble on the score of the article. It is the case (objective genitive) that has to be noted. So with Ph. 3:21, th.n evne,rgeian tou/ du,nasqai. As to adjectives, as already noted, it is doubtful if in 1 Cor. 16:4, eva.n de. a;xion h|= tou/ kavme. poreu,esqai, the inf. is to be taken with a;xion as genitive. Moulton94 so regards it, but it may be a loose nominative, as we shall see directly. But there is a use of tou/ and the inf. that calls for comment. It is a loose construction of which the most extreme instance is seen in Rev. 12:7, evge,neto po,lemoj evn tw|/ ouvranw|/Ã o` Micah.l kai. oi` a;ggeloi auvtou/ tou/ polemh/sai meta. tou/ dra,kontoj. This inf. (note the nom. with it) is in explanatory apposition with po,lemoj. Moulton95 cleverly illustrates it with the English: "There will be a cricket match - the champions to play the rest." It is a long jump to this from a case like Ac. 21:12, parakalou/men tou/ mh. avnabai,nein auvto,n, where the simple object-inf. is natural (cf. 1 Th. 4:10 f.). Cf. also Ac. 23:20, sune,qento tou/ evrwth/sai se o[pwj kataga,gh|j. "This loose inf. of design" is found twelve times in Thucydides, six in Demosthenes and five in Xenophon.96 These writers prefer the prepositions with tou/ and the inf. Polybius in his first five books has this simple tou/ and the inf. only six times, all negative.97

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1067

Addenda 3rd ed.

The normal use of tou/ with the inf. was undoubtedly final as it was developed by Thucydides, and in the N.T. that is still its chief use.98 But many of the examples are not final or consecutive. It is only in Luke (Gospel 24, Acts 24) and Paul (13) that tou/ with the inf. (without prepositions) is common.99 They have five-sixths of the examples.100 And Luke has himself two-thirds of the total in the N. T. Matthew has seven. John avoids it. Moulton101 shows that of Paul's "thirteen" examples three (Ro. 6: 6; 7:3; Ph. 3:10) either final or consecutive, two (Ro. 15:22; 2 Cor. 1:8) are ablative, five occur with substantives (Ro. 15:23; 1 Cor. 9:10; 16:4; 2 Cor. 8:11; Ph. 3:21), four are epexegetic (Ro. 1: 24; 7:3; 8:12; 1 Cor. 10:13). In Luke calls about half are not final. It is this loose epexegetical inf. that calls for notice. We find it in the LXX (cf. Gen. 3:22; 19:19; 31:20; 47:29, etc.).102 It is possible that this very common idiom in the LXX is due to the Hebrew l.. It does not occur in Polybius.103 In the LXX also we see tou/ and the inf. used as the subject of a finite verb in complete forgetfulness of the case of tou/. Cf. 2 Chron. 6:7, evge,neto evpi. kardi,an Dauei/d tou/ patro,j mou tou/ oivkodomh/sai oi=kon) So 1 Sam. 12:23; 1 Ki. 8:18; 16:31; Ps. 91:3; Is. 49:6; Jer. 2:18; Eccl. 3:12; 1 Esd. 5:67.104 One must recall the fact that the inf. had already lost for the most part the significance of the dative ending - ai and the locative - i (- ein). Now the genitive tou/ and the dative - ai are both obscured and the combination is used as subject nominative. We have this curious construction

1068 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

in Lu. 17:1, avne,ndekto,n evstin tou/ mh. evlqei/n. See also Ac. 10: 25, evge,Ä neto tou/ eivselqei/n, and 27:1, evkri,qh tou/ avpolei/n. Cf. further 20: 3. It is naturally rarer in the N. T. than in the LXX. Moulton (Prol., p. 220) gives a papyrus example closely allied to it, 0. P. 86 (iv/A.D.) e;qoj tou/ parasceqh/nai. See Winer-Moulton, p. 411, for numerous examples in LXX. But very much like it is the use of tou/ as object-inf., with evnte,llomai in Lu. 4:10 (Ps. 90: 11); kataneu,w in 5:7; sthri,zw in 9:51; poie,w in Ac. 3:12; kako,w in 7:19; evpiste,llw in 15:20; parakale,w in 21:12; sunti,qemai in 23: 20. Cf. also e[toimoj tou/ in Ac. 23:15. This is surely "a wide departure from classical Greek."105 It is, however, after all in harmony with the genius and history of the inf., though the nominative use of tou/ comes from the LXX.

The vernacular papyri show a few examples of tou/ and the inf. It is found in the inscriptions of Pisidia and Phrygia. Cf. Compernass, p. 40. Moulton106 illustrates Lu. 1:9 with avmelei/n tou/ gra,fein, B. U. 665 (i/A.D.); Mt. 18:25 and Jo. 5:7 ( e;cw) with i[n v e;ci tou/ pwlei/n, B. U. 830 (i/A.D.); 1 Cor. 9:6 with evxousi,anÄÄÄtou/ ÄÄqe,sqai, C. P. R. 156; Lu. 22:6 with euvkairi,aj- tou/ eu`rei/n, B. U. 46 (ii/A.D.). He concludes that the usage is not common in the papyri and holds that the plentiful testimony from the LXX concurs with the N. T. usage to the effect "that it belongs to the higher stratum of education in the main." This conclusion holds as to the N. T. and the papyri, but not as to the LXX, where obviously the Hebrew inf. construct had a considerable influence. Moulton seems reluctant to admit this obvious Hebraism.

(c) Prepositions. We are not here discussing the inf. as purpose or result, as temporal or causal, but merely the fact of the prepositional usage. The idiom cannot be said to be unusual in classical Greek. Jannaris107 agrees with Birklein108 that classical writers show some 2000 instances of this prepositional construction. The writers (classic and later) who use the idiom most frequently are Thucydides, Xenophon, Polybius, Diodorus, Dionysius, Josephus, Plutarch, Dio Cassius. The most prolific user of the construction is Polybius (1053 instances) and Josephus next (651 times).109 If the prepositional adverbs be added to the strict

VERBAL NOUNS ('ONOMATA TOT TEMATOZ) 1069

Addenda 2nd ed.

Addenda 3rd ed.

Addenda 3rd ed.

Addenda 3rd ed.

list of prepositions, the number is very much enlarged, especially in Polybius, who has 90 with ca,rin, 115 with a[ma, 504 with dia,, 160 with pro,j, 74 with eivj, 24 with evn, 90 with evpi,Ã 33 with meta,, 41 with peri,, only one with para,.110 The idiom was here again later than the articular inf. itself and was also Attic in origin and literary. But it is common also in the Greek inscriptions according to Granit.111 It is rare in the papyri, according to Moulton,112 save in the recurrent formula, eivj to. evn mhdeni. memfqh/nai, and (cf. 990) in the case of pro.j to,. Cf. pro.j to. tuci/n, B. U. 226 (i/A.D.); pro.j to. mh.ÄÄevntugca,nein, 0. P. 237 (i/A.D.); pro.j to. ÄÄdehqh/nai (ib.). Votaw113 finds the prepositional inf. almost one-half of all the articular infs. in the 0. T., the Apocrypha and the N. T., the proportion being about the same in each section of the Greek Bible. Not quite, all the prepositions were used with the inf. in ancient Greek, the exception114 being avna,. vAmfi, had it only with the genitive, kata,, with the accusative, para,, with the acc., peri, with the acc. and gen., pro,j with acc. and loc., u`pe,r with the ablative, u`po, with the ablative.115 It was not therefore freely used with all the usual case with the different prepositions. As a rule the article was essential if a preposition occurred with an inf. The reason for this was due to the absence of division between words. It was otherwise almost impossible to tell this use of the inf. from that of composition of preposition with the verb if the two came in conjunction. Cf. avnti. tou/ le,gein in Jas. 4:15. A few instances are found without the article. Thus avnti. de. a;rcesqai (note presence of de, between) in Herodotus I, 210. 2. It appears thus three times in Herodotus. So also in AEschines, Eum. 737, we have plh.n ga,mou tucei/n.116 So Soph., Ph., 100. Winer117 finds two in Theodoret (cf. IV, 851, para. sugklw,qesqai). The papyri give us eivj ba,yai, 0. P. 36 (i/A.D.), and the common vernacular phrase118 eivj pei/n ('for drinking'). Cf. do,j moi pei/n in Jo. 4:10. Moulton119 cites also an example of a;cri from Plutarch, p. 256 D, and one from an inscription of iii/B.C. (0. G. I. S. 41, Michel 370) evpi. lamba,nein. The instances without the article are clearly very few. Moulton (Prol., p. 81) suggests that the significant frequency of

1070 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

Addenda 3rd ed.

eivj pei/n in the papyri is due to Ionic influence. The LXX furnishes several instances of anarthrous eivj, as eivj evkfugei/n in Judg. 6:11 (cf. 2 Esd. 22:24; Sir. 38:27; Judith 4:15). Note also e[wj evlqei/n in 1 Macc. 16:9; e[wj ou- oivkteirh/sai in Ps. 122:2 (so Ruth 3:3); me,crij ou- evggi,sai in Tob. 11:1. Cf. also plh,n with anarthrous inf. in Polybius, etc.

The tenses have their full force in this prepositional construction, as in Mk. 5:4, dia. to. ÄÄdede,sqai kai. diespa,sqai kai. ÄÄsuntetri,Ä fqai. Naturally some tenses suit certain prepositions better, as with the present tense.120 The principles of indirect discourse apply also to the inf. with prepositions. Cf. meta. to. evgerqh/nai, me proa,xw (Mk. 14:28). In the N. T. the accusative seems to occur always even when the nominative predicate would be possible,121 as in dia. to. me,nein auvto,n (Heb. 7:24). So also Lu. 11:8. But note Xen., Cyr., I, 4. 3, dia. to. filomaqh.j ei=nai.

It is not necessary for the article to come next to the inf. as in Mt. 13:25. Several words may intervene and the clause may be one of considerable extent. Cf. Mk. 5:4; Ac. 8:11; Heb. 11:3; 1 Pet. 4:2. But the N. T. does not have such extended clauses of this nature as the ancient Greek, and the adverbs usually follow the inf.122 The English "split inf." is not quite parallel.

In the O. T. there are 22 prepositions used with the inf. and the Apocrypha has 18, while the N. T. shows only 10.123 Of these only eight are the strict prepositions $avnti,à dia,à eivjà evnà evkà meta, pro,à pro,j) and two the prepositional adverbs e[neka and e[wj. It remains now to examine each in detail.

vAnti. tou/ is not rare with the inf. and is chiefly found in the Greek orators.124 But we have it in Thucydides, Xenophon and Plato. Herodotus125 has only 11 instances of the preposition with the inf., but 5 of them are with avnti,. It does not occur in Polybius. In the N. T. we have only one instance, Jas. 4:15, avnti. tou/ le,gein. Votaw gives one for the LXX, Ps. 108:4, avnti. tou/ avgapa/n.

Dia, has 33 instances in the N. T., all but one (genitive, Heb. 2:15, dia. panto.j tou/ zh/n) in the accusative. Mr. H. Scott reports the 33 exx. thus: Phil. 1:1, Jas. 1, , Heb. 4, , Mk. 5, , Mt. 3, , Lu. 9, , Ac. 9, , Jo. 1. The O. T. has it with the inf. 35 times and the

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1071

Apocrypha 26,126 all with the accusative. The idiom dia. to, is so frequent in Xenophon and Thucydides that as compared with o[ti it stands as 2 to 3.127 In later Greek ( koinh, and Byzantine) it comes to displace even i[na and o[pwj though finally shifting to dia. na,, in modern Greek (cf. English "for that").128 It is not surprising therefore to find it in the N. T. with comparative frequency. Dia. to, is frequent in Luke's writings, and once in Paul's Epistles, and rare in the other N. T. writers.129 It is always the cause that is given by dia. to,, as in Mt. 13:5 f., dia. to. mh. e;cein. It is not merely the practical equivalent o[ti and dio,ti, but is used side by side with them. Cf. Jas. 4:2f. dia. to. mh. aivtie/Ä sqai u`ma/j- dio,ti kakw/j aivtei/sqe. It may stand alone, as in Lu. 9: 7; 11:8, or with the accusative of general reference as in indirect discourse, as in Lu. 2:4; 19:11. Note two accs. in Ac. 4:2. The perfect tense occurs seven times, as in Mk. 5:4 (ter); Lu. 6:48; Ac. 8:11; 18:2; 27:9. In Mk. 5:4 it is the evidence, not the reason, that is given.130 Blass (Gr. of N. T. Gk., p. 236) unnecessarily rejects Jo. 2:24.

Eivj to. is in comon also with the inf. without much difference in sense from evpi. tw|/ and pro.j to, with the inf.131 But the N. T. does not use evpi. with the inf. There is no doubt about the final use of eivj to, whatever, is true of the consecutive idea. In the late Greek Jannaris132 notes a tendency to use eivj to, (cf. bradu.j eivj to. lalh/sai in Jas. 1:19) rather than the simple inf. Cf. 1 Th. 4:9. But this tendency finally gave way to i[na. The O. T. has eivj to, 124, the Apocrypha 28 and the N. T. 72 times.133 In the N. T. it is more common than any other preposition with the inf., evn coming next with 55 examples. Moulton134 counts only 62 instances of eivj to, in the N. T., but Votaw is right with 72. Paul has it 50 times. There are 8 in Hebrews and only one each in Luke and Acts, a rather surprising situation. The papyri135 show scattered examples of it. Cf. eivj to. evn mhdeni. memfqh/nai, P. Fi. 2 (111/A.D.) 4 times. In 1 Pet. 4:2, eivj to. ÄÄbiw/sai, note the long clause. There is no doubt that in the N. T. eivj to, has broken away to some extent from the classic notion of purpose. That idea still occurs as in Ro. 1: 11, eivj to. sthricqh/nai. This is still the usual con fiction. Cf. Ro. 3:26; 7:4; 8:29; Eph. 1:12; Ph. 1:10; 1 Th. 3:5; Jas.

1072 A GRAMMAR OF THE GREEK NEW TESTAMENT

1:18; 1 Pet. 3:7; Heb. 2:17, and other examples in Mt. and Heb., to go no further. In Paul we notice other usages. In Ph. 1:23, evpiqumi,an eivj to. avnalu/sai, we have it with a substantive and in Jas. 1:19 it occurs with the adjectives tacu,j and bradu,j) It is epexegetic also with the verbal adjective qeodi,daktoi in 1 Th. 4:9. Besides, we find it as the object of verbs of command or entreaty giving the content of the verb as in 1 The 2:12; 3:10; 2 Th. 2:2, evrwtw/men eivj to. mh. tace,wj saleuqh/nai. Cf. also 1 Cor. 8:10. So in Mt. 20:19; 26:2; 1 Cor. 11:22 there is a really dative idea in eivj to,. Just as i[na came to be nonfinal sometimes, so it was with eivj to,, which seems to express conceived or actual result (cf. tou/ also) as in Ro. 1:20; 12:3; 2 Cor. 8:6; Gal. 3:17. Cf. the double use of w[ste for 'aim' or 'result!136 The perfect tense can be used with eivj to, as in Eph. 1:18 eivj to, eivde,nai and Heb. 11:3 eivj to. gegone,nai, the only instances. But the present occurs 32 times, the aorist 38, the perfect 2=72. These developed uses of eivj to, occur to some extent in the LXX (1 Ki. 22:8; 1 Esd. 2:24; 8:84).

vEn tw|/ appears in the tragedies.137 It is found 6 times in Thucydides, 16 in Xenophon, 26 in Plato.138 But Blass139 observes that the classical writers did not use evn tw|/ in the temporal sense of 'while' or 'during.' Moulton140 sought to minimize the fact that in the O. T. evn tw|/ occurs 455 times (45 in the Apocrypha) and that it exactly translates the Hebrew B. and held that it did not in principle go beyond what we find in Attic writers. But he took that back in the second edition141 under the suggestion of Dr. E. A. Abbott that we must find Attic parallels for 'during.' So he now calls this "possible but unidiomatic Greek." In the N. T. we have. evn tw|/ and the inf. 55 times and 3/4 in Luke. In the Greek Bible as a whole it is nearly as frequent as all the other prepositions with the inf.142 The Semitic influence is undoubted in the O. T. and seems clear in Luke, due probably to his reading the LXX or to his Aramaic sources.143 Cf. Lu. 1:8; 8:5 ( evn tw|/ spei,rein); 24:51; Ac. 3:6; 4:30; 9:3, etc. Jannaris144 sees here a tendency also to displace the participle. The

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1073

idiom is not confined to Luke's writings. Cf. Mt. 13:4; 13:25; Mk. 4:4; Heb. 2:8; 3:12, etc. Ordinarily it is the present inf. as in Mt. 13:4; Lu. 8:5; Ac. 3:26, where the Attic writers would have the present participle. But in Luke we have also the aorist inf. as in 2:27 evn tw|/ eivsagagei/n,grk grk(3:21) evn tw|/ baptisqh/Ä nai, where Blass145 sees the equivalent of the aorist participle (cf. vIhsou/ baptisqe,ntoj) or a temporal conjunction with the aorist indicative. One questions, however, whether the matter is to be worked out with so much finesse as that. The aorist inf. with evn tw|/ occurs only 12 times in the N. T.146 It is more correctly just the simple action of the verb which is thus presented, leaving the precise relation to be defined by the context, like the aorist participle of simultaneous action. Cf. evn tw|/ u`pota,xai in Heb. 2: 8; Gen. 32:19, evn tw|/ eu`rei/n. This is all that evn tw|/ should be made to mean with either the present or the aorist. Cf. Mt. 13:4; 27:12; Lu. 8:40; 9:29. The idea is not always strictly temporal. In Ac. 3:26 (cf. Jer. 11:17), 4:30, it is more like means. Votaw147 sees content in Lu. 12:15; Heb. 3:12. In Heb. 8:13, evn tw|/ le,gein, the notion is rather causal. The conception is not wholly temporal in Lu. 1:21.148 No other preposition occurs in the N. T. with the inf. in the locative case. But cf. evpi. tw|/ evmai. parame,nin, 0. P. 1122, 9 f. (A.D. 407).

[Eneken tou/ appears in Xenophon, Plato and Demosthenes, usually as final, but also causal.149 Sophocles in his Lexicon quotes the construction also from Diodorus and Apophth. There is only one instance of it in the N. T., 2 Cor. 7:12, e[neken tou/ faneÄ rwqh/nai th. spoudh.n u`mw/n, where it is clearly causal as with the two preceding participles, e[neken tou/ avdikh,santojà e[neken tou/ avdikhqe,ntoj (a good passage to note the distinction between the inf. and the part.). The case is, of course, the genitive.

vEk tou, likewise, appears in the N. T. only once with the inf. (2 Cor. 8:11, evk tou/ e;cein), but the case is ablative. Its usual idea in Attic prose is that of outcome or result.150 Votaw151 gives no illustration from the 0. T., but three from the Apocrypha. Blass152 takes it in 2 Cor. 8:11, to be equivalent to kaqo. a'n e;ch|. More

1074 A GRAMMAR OF THE GREEK NEW TESTAMENT

likely it is meant to accent the ability growing "out of" the possession of property, whatever it may be. In Polybius evk tou/ with the inf. has a more varied use (departure, source of knowledge, source of advantage).153 He uses it 25 times.

;Ewj tou/, likewise, occurs but once (Ac. 8:40, e[wj tou/ evlqei/n), and with the genitive. Birklein does not find any instances of e[wj tou/ and the inf. in the classic writers, though he does note me,cri tou/ and less frequently a;cri tou/.154 Cf. me,cri tou/ plei/n, P. B. M. 854 (i/ii A.D.). But in the O. T. Votaw155 observes 52 instances of e[wj tou/ and 16 in the Apocrypha. Cf. Gen. 24:33; Judith 8: 34. We have already noted the anarthrous use of e[wj evlqei/n in 1 Macc. 16:9 A. Cf. Gen. 10:19, 30, etc. So also e[wj ou- and me,cri$j% ou- and the inf., 1 Esd. 1:49, and Tob. 11:1 B. It is rather surprising therefore that we find only one instance in the N. T. and that in the Acts. The construction is probably due to the analogy of pri,n and the inf.

Meta. to, is found only a few times in Herodotus, Plato and Demosthenes.156 It appears, however, thirty-three times in Polybius and usually with the aorisit tense.157 The idea is temporal and the aorist is a practical equivalent for the aorist participle. In the O. T. Votaw158 finds it 99 times and only 9 in the Apocrypha. There are 15 examples in the N. T. and the case is the accusative always. Meta. to, vanished with the inf. in modern Greek.159 The aorist is always used in the N. T. save one perfect (Heb. 10:15). See Mk. 1:14; 14:28, meta. to. evgeqh/nai, me. Eight of the examples occur in Luke's writings (Lu. 12:5; 22:20; Ac. 1:3; 7:4; 10: 41; 15:13; 19:21; 20:1). See also Mt. 26:32; Mk. 16:19; 1 Cor. 11:25; Heb. 10:15, 26.

Pro. tou/ in the ancient writers was used much like pri,n and in the temporal sense.160 It gradually invaded the province of pri,n, though in the N. T. we only meet it 9 times. It is not common in the papyri nor the inscriptions.161 See Delphian inscr. 220, pro. tou/ paramei/nai. Polybius has it 12 times.162 In the 0. T. we find it 46 times, but only 5 in the Apocrypha.163 The tense is always the aorist save one present (Jo. 17:5) . Cf. Gal. 3:23, pro. tou/ evlqei/n th.n pi,stin. There is no essential differ-

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1075

Addenda 3rd ed.

ence in construction and idea between rrpiv and the inf. and pro. tou/ and the inf. The use of pri,n with the inf. was common in Homer before the article was used with the inf. The usage became fixed and the article never intervened. But the inf. with both pri,n, and pro, is in the ablative case. Cf. ablative164 inf. with pura in Sanskrit. Pri,n was never used as a preposition in composition, but there is just as much reason for treating pri,n as a prepositional adverb with the ablative inf. as there is for so considering e[wj tou/, not to say e[wj alone as in e[wj evlqei/n (1 Macc. 16: 9). The use of the article is the common idiom. The fact of pri,n and the inf. held back the development of pro. tou/. In modern Greek pro. tou/ as protou/ occurs with the subj. (Thumb, Handb., p. 193). In the N. T. pri,n is still ahead with 13 examples. The instances of pro. tou/ are Mt. 6:8; Lu. 2:21; 22:15; Jo. 1:48; 13:19; 17:5; Ac. 23:15; Gal. 2:12; 3:23.

Pro.j to, is the remaining idiom for discussion. It was used by the ancients in much the same sense as eivj to, and evpi. tw|/, 'looking to,' with a view to.'165 The idiom is very common in Polybius,166 150 examples, and there are 10 of pro.j tw|/) But in the O. T. we have only 14 examples and 12 in the Apocrypha.167 The N. T. shows 12 also. Some of the LXX examples are of pro.j to, (Ex. 1:1; 2 Macc.7:14), but in the N. T. they are all pro.j to,. the papyri Moulton168 finds pro.j to, rather more common than eivj to,. In the N. T. Matthew has it five times times(5:28; 6:1; 13:30; 23:5; 26:12). These express aim unless 5:28 is explanatory of ble,pwn.169 Mark has it once, 13:22. Luke has it twice (18: 1, where pro.j to. dei/n means 'with reference to'; Ac. 3:19 only aB, while other MSS. read eivj).170 Paul's four examples (2 Cor 3:13; Eph. 6:11, DEFG eivj; 1 Th. 2:9; 2 Th. 3:8) all give the "subjective purpose."171 Both present (3 times) and aorist (9 times) tenses occur. Cf. pro.j to. qeaqh/nai in Mt. 6:1.

(d) The Infinitive with Substantives. Numerous examples of the inf. with substantives were given in the discussion of the cases of the inf. The matter calls for only a short treatment at this point. The use of the inf. with substantives was ancient172 and natural, first in the dative or locative and then in the genitive

1076 A GRAMMAR OF THE GREEK NEW TESTAMENT

with tou/. It was always common in the classic Greek.173 The usage is common in Polybius with both the anarthrous and the articular inf.174 The same thing is true of the O. T. and the Apocrypha.175 It is so frequent as not to call for illustration. The meaning is that of complement and the inf. most frequently occurs with words of time, fitness, power, authority, need, etc. It is abundantly used in the N. T. both with and without the article. Some anarthrous examples are (Mt. 3:14) crei,an baptisqh/nai, (Lu. 2:1) do,gma avpogra,fesqai, (Jo. 1:12) evxousi,an gene,sqai,grk grk(19:40) e;qoj evntafia,zein, (Ac. 24:15) evlpi,da me,llein, (Ro. 13:11) w[ra evgerÄ qh/nai, (Gal. 5:3) ovfeile,thj poih/sai, (Heb. 7:5) evntolh.n avpodekatoi/nà (Rev. 11:18) kairo.j kriqh/nai, etc. These are all real datives and the construction is common enough in the N. T., more so than in the LXX. In Ph. 1:23 note evpiqumi,an eivj to. avnalu/sai. The same substantives may have tou/ and the inf., though now, of course, the case is genitive. Cf. (Lu. 1:57) cro,noj tou/ tekei/nà (2: 21) h`me,rai tou/ peritemei/n,grk grk(10:19) evxousi,an tou/ patei/n (Ac. 14:9) pistin tou/ swqh/nai,grk grk(27:20) evlpi.j tou/ sw,zesqai, etc. It occurs ten times in Luke's writings and nine in Paul'sEpistles. It is about as common in proportion as in the LXX.176 See further Lu. 1:74; 2:6; 21:22; 22:6; Ac. 20:3; Ro. 1:24; 8:12; 11:8; 15:23; 1 Cor. 9:10; 10:13; 2 Cor. 8:11; Ph. 3:21; 1 Pet. 4:17; Heb. 5:12, etc. Since the inf. is a substantive, the genitive relation with other substantives is obvious and natural.

(e) The Infinitive with Adjectives. This idiom is likewise classical and is common from Homer on.177 As already shown, the case varies with different adjectives. This inf. is complementary as with substantives. It is natural with adjectives as any other substantive is. It held on longest with dunato,jà i`kano,j, but other adjectives in late koinh, began to give way to eivj to, (cf. Jas. 1:19, tacu.j eivj to. avkou/saià bradu.j eivj to. lalh/sai) rather than the simple inf. and finally this disappeared before i[na (cf. Mt. 8:8, i`kano.j i[na).178 In the LXX and the N. T. the inf. with adjectives is less frequent than with substantives. We have it with both the anarthrous and the articular inf. See (Mt. 3:11) i`kano.j basta,saià (Mk. 10:40) evmo.n dou/nai, (Lu. 15:19) a;xioj klhqh/nai, (Jas. 3:2) duÄ nato.j calinagwgh/sai, (1 Cor. 7:39) evleuqe,ra gamhqh/nai, (Heb. 5:11) dunsermh,neutoj le,gein, (1 Pet. 4:3) avrketo.j kateirga,sqai, etc. It is

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1077

more common with a;xiojà dunato,jà i`kano,j. The only adjective that often has tou/ and the inf. in the O. T. is e[toimoj.179 We find it also with adverbs as in Ac. 21:13, deqh/nai avpoqanei/n e`toi,mwj e;cw (so 2 Cor. 12:14). The articular examples are less frequent. But note (Lu. 24:25) bradei/j tou/ pisteu,ein, (Ac. 23:15) e[toimoi tou/ avnelei/n. Some would add 1 Cor. 16:4, a;xion tou/ poreu,esqai, but see Cases of the Inf.

(f) The Infinitive with Verbs. This usage came to be, of course, the most frequent of all. It started as a dative or locative, then a sort of accusative of reference,180 then the object of verbs with whatever case the verb used. It is both anarthrous and articular. It is not necessary to go over again (see Cases of the Inf.) the varied uses the inf. with verbs, whether the object of verbs of saying or thinking in indirect discourse, verbs of commanding or promising,direct object of verbs (auxiliary inf.), verbs of hindering,181 etc. As a matter of fact they are all object-infs. whatever the case (acc., gen., abl., dat., instr.). Votaw182 notes that in the N. T. this use of the inf. is four times as common as any other. It is usually the anarthrous inf., but not always. Even du,namai and a;rcomai (not N. T.) are used with tou/ and the inf. Jannaris183 has made a careful list of the verbs that continued for a while in late breek to use the inf. against the inroads of i[na. Radermacher (N. T. Gr., p. 150) argues that in general the N. T. use of the inf. With verbs is like that of the koinh,. The inf. lalh/sai with evparrhsiasa,meqa (1 Th. 2:2) is not a Hebraism, but a Hellenism. But surely it is not necessary to call this usage an Atticism. In the discussion of i[na (see pp. 430, 994) the displacement of the inf. by i[na even after verbs like qe,lw was sufficiently treated. Schmid184 "shows how this 'Infinitivsurrogat' made its way from Aristotle onwards."185 In the N. T. it is chiefly in the Gospel of John that we find this use of i[na) "The strong volitive flavour which clung to i[na would perhaps commend it to a writer of John's temperament."186 But after all, the inf. with verbs has not quite disappeared from John's Gospel. Jannaris187 has worked out the situation in John's Gospel as between this use of the inf. and i[na. p. 7.

1078 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

He finds i[na, about 125 times and the inf. with verbs about 129 times. Of these 57 belong to du,namai (37) and qe,lw (20). There are besides, 10 with dei/ and 12 each with zhte,w and with me,llw. The rest are scattered with di,dwmià e;cwà ovfei,lwà doke,wà avfi,hmià aivte,wà evrwta,wà a;rcomai, etc. It is clear, therefore, that the inf. with verbs is by no means dead in the N. T., though the shadow of i[na is across its path. As illustrations of the great wealth of verbs with the inf. in the N. T. note (Mt. 11:20) h;rxato ovneidi,zein,grk grk(27:58) evke,leusen avpodioqh/nai, (Mk. 12:12) evzh,toun krath/sai, (Lu. 16:3) ska,ptein ouvk ivscu,wà evpaitei/n aivscu,nomai. Almost any verb that can be used with a substantive can be used with the inf. The use of the inf. with prosti,qemai is a Hebraism. Cf. Ex. 14: 13. See Lu. 20:11 f., prose,qeto pe,myai. It means 'to go on and do' or 'do again.' It is the one Hebraism that Thumb188 finds in Josephus, who is Atticistic. The articular inf. with verbs is much less frequent. But note to. avgapa/n after ovfei,lw (Ro. 13:8); paraitou/mai to. avpoqanei/n, (Ac. 25:11); tou/ peripatei/n after poie,w (Ac. 3:12); evpistei/lai tou/ avpe,cesqaigrk grk(15:20); katei/con tou/ mh. poreu,esqai, (Lu. 4:42). In 1 Ki. 13:16 we have tou/ evpistre,yai with du,namai. These are just a few specimens. See Cases of the Inf.

(g) The Appositional Infinitive. The grammars draw a distinction here, but it is more apparent than real as Votaw189 well says. The inf. in apposition is that with nouns; the epexegetical inf. is used with verbs. But at bottom the two uses are one. They are both limitative. With nouns the appositional inf. restricts or describes it. It is a common enough idiom in classical Greek190 and is found also in the LXX. In the N. T. observe Ac. 15:28 plh.n tou,twn tw/n evpa,nagkejà avpe,cesqai, (Jas. 1:27) qrhskei,a kaqara. kai. avmi,antoj- au[th evsti,nà evpiske,ptesqai. Cf. further Ac. 26:16; 2 Cor. 10:13; Eph. 3:6, 8; 4:17; 1 Th. 4:3 f.; Heb. 9:8; 1 Pet. 2:15 ( ou[twj). The articular inf. may also be appositional as in Ro. 14:13, tou/to kri,nate ma/llonà to. mh. tinqe,nai. So also 2 Cor. 2:1; 7:11; Ro. 4:13; 1 Th. 4:6 bis. In the N. T. and the Apocrypha it is only to, (in the articular use) that is appositional, but in the O. T. 15 out of the 17 instances have tou/ without any reference to the case of the noun.191 It is worth noting that i[na is common also in appositional clauses (cf. Lu. 1:43; 1 Cor. 9:18), especially in the writings of John (Jo. 4:34; 15:8;

VERBAL NOUNS. ( vONOMATA TOU `RHMATOS) 1079

17:3; 1 Jo. 3:11, 23; 4:21; 5:3, etc.). We find o[ti also in 1 Jo. 2:3; 3:16).192

5. VERBAL ASPECTS OF THE INFINITIVE. It is worth repeating (p. 1057) that the inf. is substantive as well as verb. Each inf. does not, of course, have all the substantival and verbal uses, but each inf. has both substantival and verbal aspects. The uses vary with each example. The verbal aspects do not exclude the substantival, though some193 writers say so. Per contra, Jannaris194 holds that "the verbal nature of the substantival infinitive was sometimes completely lost sight of." This I do not concede. After tenses came to the verbal substantive its dual character was fixed. But, pp. 1050, 1056 f., the inf. did not come to the rank of a mode.

(a) Voice. The Sanskrit inf. had no voice. In Homer the inf. already has the voices, so that it is speculation as to the origin. It is possible that the original Greek inf. had no voice. This is an inference so far as the Greek is concerned, but a justifiable one. Moulton195 illustrates it well by dunato.j qauma,sai, 'capable for wondering,' and a;xioj qauma,sai, 'worthy for wondering,' when the first means 'able to wonder' and the second 'deserving, to be wondered at.' They are both active in form, but not in sense. "The middle and passive infinitives in Greek and Latin are merely adaptations of certain forms, out of a mass of units which had lost their individuality, to express a relation made prominent by the closer connection of such nouns with the verb."196 There was so much freedom in the Greek inf. that the Sanskrit -tum did not develop in the Greek as we see it in the Latin supine. Gradually by analogy the inf. forms came to be associated with the voices in the modes. Practically, therefore, the Greek inf. came to be used as if the voices had distinctive endings (cf. the history of the imper. endings).197 Thus in Lu. 12:58, do.j evrgasi,an avphlla,cqai avp v auvtou/, it is clear that the passive voice is meant whatever the origin of the form - sqai. The reduplication shows the tense also. The same remark applies to Mk. 5:4, dia. to. dede,sqai kai. diespa,sqai u`p v auvtou/ ta.j a`lu,seij. See also 5:43, ei=pen doqh/nai auvth|/ fagei/n. No special voice significance is manifest in fagei/n, which is like our

1080 A GRAMMAR OF THE GREEK NEW TESTAMENT

'eating' and is the acc. of general reference with doqh/nai which in turn is the direct object of ei=pen. But doqh/nai has the passive force beyond a doubt. Cf. further avpolelu,sqai evdu,nato in Ac. 26:32 and e[neken tou/ fanerwqh/nai in 2 Cor. 7:12. In general, therefore, after the inf. is fully developed, the voice in the inf. appears exactly as in the modes. So tou/ avpe,cesqai (Ac. 15:20); avpogra,yasqai (Lu. 2:5); evpilaqe,sqai (Heb. 6:10); gamhqh/nai (1 Cor. 7:39); klhqh/nai ui`o,j (Lu. 15:19). Cf. qea,sasqai (Lu. 7:24) and qeaqh/nai (Mt. 6:1).

(b) Tense. See chapter on Tenses for adequate discussion of this point. Some general remarks must here suffice. As the Sanskrit inf. had no voice, so it had no tense. In the original Greek there was possibly no tense in the inf., but in Homer the tense is in full force.198 There is no time-element in the inf. (cf. subj., opt. and imperative) except as the future inf. echoes the expectation of a verb like evlpi,zw (or me,llw% or as the inf. represents a fut. ind. in indirect discourse (see Indirect Discourse under Modes). It is probably true that originally there was no distinction between aorist (punctiliar) and present (linear) action in the inf. In Sanskrit and Latin the infinitives and supines have no necessary connection with the present stem (cf. supine tactum and inf. tangere).199 "The s in lu/sai has only accidental similarity to link it with that in e;lusa."200 Moulton201 tersely adds: "But when once these noun-forms had established their close contact with the verb, accidental resemblances and other more or less capricious causes encouraged an association that rapidly grew, till all the tenses, as well as the three voices, were equipped with infinitives appropriated to their exclusive service." But even so at first the tense of the inf. had only to do with the kind of action (punctiliar, linear, state of completion), not with time.

In general, as with the subj., opt. and imper., the aorist inf. came to be the natural202 one unless some reason for the present or perf. or fut. existed. Cf. katabh/nai (Lu. 9:54); paqei/n (Lu. 24: 46); katalu/sai (Mt. 5:17); proseu,xasqai (Lu. 18:10); avkou/sai (Ac. 10:33); evkce,ai (Ro. 3:15), etc. Sometimes, as in e;dei poih/sai (Mt. 23:23), the inf. was used to suggest antecedent action. But the timeless aorist may point to what is future, as in Lu. 24: 46 above. Cf. also Lu. 2:26; Ac. 3:18. Essentially, it does neither. Cf. me,llw with aor. inf. So me,llonta evnegk[ ei/] n, P. Grenf.,

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1081

ii, 77 (iii/A.D.). In indirect assertions the aorist inf. represents the aor. indicative, but the N. T. seems to show no instance like this.203 However, that is a mere accident, for note evn tw|/ eivsagagei/n tou.j gonei/j to. paidi,on tou/ poih/sai auvtou,j. (Lu. 2:27) where the same principle applies. Contrast the tense of poih/sai and pei,qeij in Ac. 26:28. In Lu. 24:46, ge,graptai paqei/n to.n Cristo,n, we have the timeless aorist in indirect discourse.

The present inf. with some verbs would accent linear action and with others the inf. would not draw the point, sharply. Some writers have a fondness for the present.204 One can see the force of linear action in h`ma/j dei/ evrga,zesqai (Jo. 9:4) and in to. avgapa/n auvto,n (Mk. 12:33). Cf. also stoicei/n, in 3:16. In 1 Jo. 3:9, ouv du,natai a`marta,nein, the linear notion is prominent (cf. ouvc a`marta,nei in verse 6). It is also quite normal with me,llw, with which it occurs 84 times in the N. T. to 6 of the aorist. See Mt. 14:22 for both aorist evmbh/nai and present proa,gein in same sentence. Cf. also Ac. 15:37 f. The usual tense-distinction may be assumed to exist, though in a case like le,gein (Heb. 5:11) the point is not to be stressed. The present inf. in indirect assertion represents the same tense of the direct, as in Mt. 22:23; Lu. 11: 18, etc. Rarely the present inf. represents an imperfect indicative as in Lu. 20:6.

The perfect inf. is common also in indirect discourse to stand for the same tense of the direct, as in Jo. 12:2. A.c. 12:14; 14: 19; 16:27. This is natural enough. But the perfect inf. is found also in the complementary inf. as Ac. 26:32, avpolelu,sqai evdu,nato. Note Lu. 12:58, do.j evrgasi,an avphlla,cqai. But we also find the perfect tense with the articular inf. (so aorist and present) as in Mk. 5:4; Lu. 6:48; Ac. 27:9. In the N. T. there are in all 47 perfect infs. and the same number in the 0. T.205 Of the N. T. examples 23 are anarthrous, 8 articular. The papyri show the articular perf. inf. Cf. evpi. tw|/ gegone,nai, P. Oxy. 294 (A.D. 22); u`pe.r tou/ avpolelu,sqai se, P. Br. M. 42 (B.C. 168).

The future inf. is increasingly rare. Thucydides even used to, with the future inf. The same construction is found in Polybius.206 But in the koinh, the future inf. is weakening rapidly. This disappearance of the fut. inf. is partly due to the retreat of the fu-

1082 A GRAMMAR OF THE GREEK NEW TESTAMENT

ture tense in general207 and partly to the apparent kinship between the future and aorist forms. In the papyri Moulton208 notes that the future inf. is sometimes used in the koinh, as equivalent to the aorist or even the present, since the sense of the future was vanishing. Cf. cwrh,sein in Jo. 21:25 ( aBC), while the other later MSS. give cwrh/sai. In the O. T. the fut. inf. (anarthrous always) occurs only 14 times and only 6 in the N. T. The Apocrypha has, however, 54, but almost all in 2 and 3 Maccabees.209 Three of the N. T. examples are with me,llw (Ac. 11:28; 24:15; 27:10). Another is in Ac. 23:30 and is dependent on a participle after a past indicative. In Ac. 26:7 the margin of W. H. (after B) has katanth,sein (text - h/sai) with evlpi,zei. In Heb. 3:18 note w;mwsen mh. eivseleu,sesqai (LXX). Another example is in Jo. 21:25, after oi=mai. Moulton (Prol., p. 219) cites crh. e`toima,sein, B. U. 830 (i/A.D.).

(c) Cases with the Infinitive. In general the inf. uses the same case that the finite verb does. So the genitive in Heb. 6:10 evpilaqe,sqai tou/ e;rgou, the dative in 1 Cor. 7:39 w|- qe,lei gamhqh/nai, the acc. in Ac. 23:15 tou/ avnelei/n, the instrum. in Mt. 15:20 to. avni,ptoij cersi.n fagei/n, the locative in Ac. 21:21 mhde. toi/j e;qesin peripatei/n, the ablative in Ac. 15:20 tou/ avpe,cesqai tw/n avlisghma,Ä twn, the predicate nominative in Ac. 17:18 kataggeleu.j ei=nai, the predicate accusative in Ro. 2:19 pe,poiqaj seauto.n o`dhgo.n ei=nai, or the acc. of general reference in ind. discourse in Mk. 12:18. But this brings us again to the acc. in indirect assertion, a matter already treated at some length. (See Accusative Case, Indirect Discourse, and the next section.) But the thing to note is the real verbal nature of the inf. in the matter of cases. Note the three accusatives with tou/ dida,skein in Heb. 5:12, two objects, one of general reference. The cognate neuter plural is seen in polla. paqei/n (Mt. 16:21).

(d) The Infinitive in Indirect Discourse. The frequent obscuration of the cases with the inf. in indirect discourse justifies some additional remarks besides those in the chapter on Modes. The inf. is not finite and, like the participle, has no subject. By courtesy the grammars often say so, but it beclouds more than it clears to do so. The case of the predicate210 with the inf. is the

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1083

place to start. Cf. Mt. 19:21, eiv qe,leij te,leioj ei=nai. See also 2 Cor. 10:2, de,omai to. mh. parw.n qarrh/sai, where the nominative occurs within the domain of the accusative articular inf. But note Mk. 14:28, meta. to. evgerqh/nai, me proa,xw. The true nature of the acc. with the inf. as being merely that of general reference comes out well in the articular inf., as in Jas. 4:2, ouvk e;cete dia. to. mh. aivtei/sqai u`ma/j. It is not necessary here to go over again the steps taken under Modes, but simply to insist on the true nature of the accusative with the inf. It stands, indeed, in the place of a finite verb of the direct statement, but does not thereby become finite with a subject. From the syntactical standpoint the construction is true to both the substantival and verbal aspects of the inf. The subject of the finite verb, when thrown into the acc., takes this turn because of the limitations of the inf. When it is retained in the nominative, it is by apposition with the subject of the principal verb or by attraction if in the predicate. Draeger sees this point clearly in his treatment of the latter in Latin where the acc. with the inf. is much more frequent than in Greek.211 "The name is confessedly a misnomer," say King and Cookson.212 Schmid213 also sees the matter clearly and makes the acc. with the inf. the acc. of general reference. The usual beaten track is taken by Jolly,214 but the truth is making its way and will win. Schmitt215 admits that the acc. is not the grammatical subject, but only the logical subject. But why call it "subject" at all? Schroeder216 properly likens it to the double accusative with dida,skw, as in dida,skw auvto.n peripatei/n. The late Sanskrit shows a few examples like English "if you wish me to live:"217 The use of the acc. with the inf. early reached a state of perfection in Greek and Latin. Schlicher218 notes 130 instances of it in Homer with fhmi, alone as against 15 with w`j o[ti. We see it in its glory in historians like Xenophon and Thucydides in Greek and Cesar in Latin. Votaw219 notes the rarity of the construction in the O. T. and Apoc. (46 verbs), while the N. T. has 27 (83 exx.) verbs which use the idiom. But even in the N. T., as compared with the ancient Greek, the construction is greatly narrowed. The particular

1084 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

verbs in the N. T. which may use the acc. and the inf. in indirect assertion were given under Modes. A general view of the matter discloses a rather wide range still. But the idiom, being largely literary, is chiefly found in Luke, Rom. and 1 Cor. The other writers prefer o[ti. Luke, in fact, is the one who makes the most constant use of the idiom, and he quickly passes over to the direct statement. There is with most of them flexibility as was shown. Blass220 has a sensible summary of the situation in the N. T. There is, in truth, no essential difference in the Greek construction, whether the inf. is without a substantive, as in Ac. 12:15 diiscuÄ ri,zeto o[twj e;cein, with the acc., Ac. 24:9 fa,skontej tau/ta ou[twj e;cein, or with the nom. Ro. 1:22 fa,skontej ei=nai sofoi,. Cf. Ac. 17:30; 1 Pet. 3:17. Words like dei/à avna,gkh may be followed by no substantive (Mt. 23:23; Ro. 13:5). Cf. Lu. 2:26. In 1 Pet. 2:11, we have only the predicate w`j paroi,kouj- avpe,cesqai. Freedom also exists. In Mk. 9:47 we have kalo,n se, evstin mono,Ä fqalmon eivselqei/n, while in Mt. 18:8 we read kalo,n soi, evstin mono,Ä falmon eivselqei/n. Even in Matthew the predicate adj. is acc., though it might have been dative, as in Ac. 16:21. Further examples of the predicate dative when an accusative is possible are seen in Lu. 1:3; 9:59; Ac. 27:3 ( aAB); 2 Pet. 2:21. But see Ac. 15:22, 25; Heb. 2:10. The case of the inf. itself is not the point here. There are besides verbs of willing, desiring, allowing, making, asking, beseeching, exhorting, some verbs of commanding, the inf. with pri,nà w[steà to,à tou/à prepositions and the articular infinitive. With all these the acc. may occur. A difficult inf. occurs in Ac. 26:28, evn ovli,gw| me pei,qeij Cristiano.n poih/sai) Is me the object of pei,qeij or of poih/sai? Can pei,qeij be 'try by persuasion'? Prof. W. Petersen suggests that this is a contamination of evn ovli,gw| me pei,qeij Cristiano.n ei=nai and evn ovli,gw| me poih,seij Cristiano,n. But verbs differ. Keleu,w, for instance, always has the acc. and the inf., while the dative comes with ta,ssw (Ac. 22:10), evpita,ssw (Mk. 6:39), and verbs like evnte,llomaià evpitre,pwà paragge,llw, and impersonal expressions like sumfe,reià e;qoj evsti,n avqe,mitonà aivscro,n, etc. As shown above, kalo,n evstin is used either with the acc. or the dative, as is true of le,gw (cf. Mt. 5:34, 39 with Ac. 21:21; 22:24). Blass221 adds also Ac. 5:9, sunefwnh,qh u`mi/n peira,sai. He notes also that prosta,ssw occurs with the acc. (Ac. 10:48) as is true of evpita,ssw (Mk. 6:27) and ta,ssw (Ac. 15:2). Even sumfe,rei appears with the acc. and inf. (Jo. 18:14) and e;xestin (Lu. 6:4, where D has the dative, as is true of Mt.

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1085

12:4). With evge,neto Blass222 observes how clumsy is evge,neto, moi- gene,sqai me (Ac. 22:17). The acc. and inf. occurs with evge,neto (Ac. 9:32) and the dative also in the sense of it 'befell' or 'happened to' one, as in Ac. 20:16. In Ac. 22:6, evge,neto, moi- periastra,yai fw/j, the two constructions are combined. Blass223 further observes the independence of the inf. in adding an ace. of general reference besides the acc. with a verb of asking, as in Ac. 13:28 h|vth,santo Peila/ton avnaireqh/nai auvto,n, (1 Th. 5:27) o`rki,zw u`ma/j avnagnwsqh/nai th.n evpistolh,n. In Ac. 21:12, parekalou/men- tou/ mh. avnabai,nein auvto.n eivj vIerousalh,m, the auvto,n is acc. of general reference with the inf., which is itself in the genitive as to form, though the real object of the verb. There is no instance in the N. T. of the inf. in a subordinate clause unless we follow Nestle in 1 Pet. 5:8, zhtw/n ti,na katapiei/n. There are sporadic examples of such a construction due to analogy of the inf. in the main clause.224 Cf. 0. P. 1125, 14 (ii/A.D.), oua}j kai. kurieu,ein tw/n karpw/n.

(e) Personal Construction with the Infinitive. Many verbs and adjectives allowed either the personal or the impersonal construction with the infinitive. The Greek developed much more freedom in the matter than the Latin, which was more limited in the use of the impersonal.225 In the N. T. the impersonal construction occurs with fixed verbs like dei/, Ac. 25:24, bow/ntej mh. dei/n auvto.n zh/n mhke,ti, where note inf. dependent on inf. as is common enough (Ac. 26:9; Lu. 5:34; Heb. 7:23; Mk. 5:43; Lu. 6:12; 8:55). So also with e;xestin, etc. The impersonal construction is seen also in Lu. 2:26; 16:22; Ph. 3:1; Heb. 9:26, etc. The inf. with impersonal verbs is somewhat more frequent in the N. T. than in the LXX. On the whole the personal construction with the inf. is rare in the N. T.226 But in the N. T. doke,w has the personal construction, as in Ac. 17:18, dokei/ kataggeleu.j ei=nai (cf. Jas. 1:26; Gal. 2:9, etc.), but we find e;doxe, moi in Lu. 1:3 (cf. Ac. 15:28, etc.) and even e;doxa evmautw|/ dei/n pra/xai (Ac. 26:9). The koinh, seems to use it less frequently than the ancient Greek. Radermacher (N. T. Gr., p. 148) quotes Vett. Valens, p. 277, 19, do,xei- u`pa,rcein auvth/n th.n ai;resin. We have dedokima,smeqa pisteuÄ qh/nai (1 Th. 2:4) and evmarturh,qh ei=nai (Heb. 11:4). One may compare the personal construction with o[ti (1 Cor. 15:12; 2 Cor.

1086 A GRAMMAR OF THE GREEK NEW TESTAMENT

3:3; 1 Jo. 2:19). The personal construction occurs with pre,pei (Heb. 7:26). The impersonal has the acc. and the inf. (1 Cor. 11:13), the dative and the inf. (Mt. 3:15), both the dative and the acc. (Heb. 2:10). Cf. W. F. Moulton in Winer-Moulton, p. 402. The love of the passive impersonal appears in Ac. 13: 28 h|vth,santo Peila/tonà avnaireqh/nai auvto,n and in 5:21, avpe,stilan avcqh/nai auvtou,j (Radermacher, N. T. Gr., p. 148). The nominative predicate with the inf. and the nom. in indirect discourse is to be noted also.

(f) Epexegetical Infinitive. As already remarked, there is no essential difference between the appositional and the epexegetical use of the infinitive. The epexegetical inf. is added to a clause more or less complete in itself, while the merely appositional is more simple.227 It is common in the dramatists. This use is probably adnominal228 in origin, but it drifts into the verbal aspect also. We see a free use of the limitative229 inf. in w`j e;poj eiv pei/n which only occurs once in the N. T. (Heb. 7:9). Brugmann does not agree with Granewald that this is the original epexegetical or limitative inf., though it is kin to it. Blass230 applies "epexegetical" merely to the appositional inf. It is in the epexegetical inf. that we see more clearly the transition from the original substantive to the verbal idea. It is hard to draw the line between do,gma avpogra,fesqai pa/san th.n oivkoume,nhn (Lu. 2:1) and pare,dwken auvtou.j eivj avdo,kimon nou/nà poiei/n ta. mh. kaqh,konta (Ro. 1:28). The first is appositional, the latter epexegetical. A good instance of the epexegetical inf. is seen in 2 Cor. 9:5, where tau,thn e`toi,mhn ei=nai w`j euvlogi,an is subsidiary to the i[na clause preceding, as is often the case. Viteau231 notes that the construction is frequent in the Epistles. Cf. Eph. 1:16-18 $ i[na- eivj to. eivde,nai%, 3:16 f. $ i[na - krataiwqh/naià katoiÄ kh/sai, Col. 1:10 $i[na - peripath/nai), 4:3 ( i[na- lalh/sai). Further examples occur in Lu. 1:54 mnhsqh/nai, 1:72 poih/sai kai. mnhÄ sqh/nai, 1:79 evpifa/nai tou/ kateuqu/nai, Ac. 17:27 zhtei/n, 2 Pet. 3: 2 mnhsqh/nai. The LXX232 shows rather frequent instances of the articular inf. in this sense (cf. Gen. 3:22; Judg. 8:33; Ps. 77:18). The N. T. shows very few. Indeed, Votaw finds only one, that in Gal. 3:10, evpikata,ratoj pa/j oa}j ouvk evmme,nei pa/sin toi/j gegramme,noij evn tw|/ bibli,w| tou/ no,mou tou/ poih/sai auvta,. But certainly

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1087

tou/ avtima,zesqai (Ro. 1:24) after pare,dwken is just as truly epexegetical as is poiei/n in verse 28 after pare.dwken. So also Ro. 7:3; 8: 12; 1 Cor. 10:13. Burton233 looks at the epexegetical inf. as "an indirect object," as in Lu. 10:40, h` avdelfh, mou mo,nhn me kate,leipen diakonei/n. There is no doubt that in such instances the inf. is in the original dative case with the dative idea. See further Mk. 4:23; 6:31; Lu. 7:40; 12:4; Ac. 4:14; 7:42; 17:21; 23:17, 18, 19; Tit. 2:8, etc.

(g) Purpose. It is but a step from the explanatory or epexegetical inf. to that of design. Indeed, the epexegetical inf. sometimes is final, a secondary purpose after i[na, as in Eph. 1:18; 3: 17; Col. 1:10, etc. The sub-final or objective use of the inf. is also a step on the way. This use was very common in the ancient Greek, but was partially taken up by i[na in N. T.234 But many verbs, as we have seen, retain the sub-final inf. in the N. T. as in the rest of the koinh,. Blass' careful lists and those of Viteau were given under Indirect Discourse. This notion of purpose is the direct meaning of the dative case which is retained. It is the usual meaning of the inf. in Homer,235 that of purpose. It goes back to the original Indo-Germanic stock.236 It as always more common in poetry than in prose. The close connection between the epexegetical inf. and that of purpose is seen in Mk. 7:4, a pare,labon kratei/n ('for keeping,' 'to keep'). So Mt. 27:34, e;dwkan auvtw|/ piei/n oi=non ('for drinking,' 'to drink'). So Mt. 25:35, evdw,Ä kate, moi fagei/n. The inf. with the notion of purpose is exceedingly frequent in the LXX, second only to that of the object-inf. with verbs.237 It was abundant in Herodotus.238 Hence Thumb239 thinks its abundant use in the koinh, is due to the influence of the Ionic dialect. Moulton240 agrees with this opinion. This is true both of the simple inf. of purpose and tou/ and the inf. The Pontic dialect still preserves the inf. of purpose after verbs like avnabai,nw, etc. It is noteworthy that this inf. was not admitted into Latin except with a verb of motion. Moulton (Prol., p. 205) cites Par. P. 49 (ii/B.C.) eva.n avnabw/ kavgw. proskunh/sai, as parallel to Lu. 18:

1088 A GRAMMAR OF TSE GREEK NEW TESTAMENT

Addenda 3rd ed.

10, avne,bhsan- porseu,xasqai. Moulton241 notes this correspondence between the ancient and the modern vernacular and agrees with Thumb's verdict again that the result is due to the two conflicting tendencies, one the universalizing of i[na, which prevailed in Western Hellenism and resulted in the disappearance of the inf. in modern Greece, while the idealizing of the inf. in Pontus serves to illustrate to-day the N. T. idiom. The N. T. use of the inf. of purpose includes the simple inf., tou/ and the inf., eivj to, and the inf., pro.j to, and the inf. w[ste and the inf. There is no example of evf v w|- te. First note the simple inf., all in the original dative case. This use had a wider range in Homer than in the Attic writers. Thus Mt. 2:2 h;lqomen proskunh/sai auvtw|/;grk grk(5:17) ouvk h=lqon katalu,saià avlla. plhrw/sai;grk grk(7:5) diable,yeij evkbalei/n to. ka,rfojgrk grk(11:7) ti, evh,lqate eivj th.n e;rhmon qea,sasqai (so verse 8, ivdei/n); 20:28; (Mk. 3:14) avposte,llh| auvtou.j khru,sseingrk grk(5:32) perieble,Ä peto ivdei/n* $Lu. 18:10) avne,bhsan proseu,xasqai; (Jo. 4:15) die,rcwmai evnqa,de avntlei/n; (Ac. 10:33) pa,resmen avkou/sai; (2 Cor. 11:2) h`rmoÄ sa,mhn u`ma/j - parasth/sai; (Rev. 5:5) evni,khsen- avnoi/xai;grk grk(16:9) ouv meteno,hsan dou/nai. These examples will suffice. It is very common in the N. T. It is not necessary to multiply illustrations of tou/ after all the previous discussion. The O. T. shows the idiom in great abundance, though the construction is classic. It was used especially by Thucydides.242 This was a normal use. We have already noticed that Paul makes little, if any, use of this idiom.243 It is possible in Ro. 6:6; Ph. 3:10. Indeed, Votaw244 notes only 33 instances of tou/ and inf. of purpose in the N. T., and these are chiefly in Matthew, Luke and Acts. Note (Mt. 2: 13) zhtei/n tou/ avpole,sai,grk grk(13:3) evxh/lqen tou/ spei,rein, (Lu. 21:22) tou/ plhsqh/nai pa,nta,grk grk(24:29) tou/ mei/nai. See further Ac. 3:2; 5: 31; 26:18; 1 Cor. 10:7; Gal. 3:10; Heb. 10:7, etc. The use of tou/ mh, is, of course, the same construction. Cf. Ro. 6:6, tou/ mhke,ti douleu,ein h`ma/j) Cf. Ac. 21:12. In Lu. 2:22 note parasth/Ä sai, and in verse 24 tou/ dou/nai. Purpose is also expressed by eivj to, as in 1 Th. 3:5, e;pemya eivj to. gnw/nai, and by pro.j to, as in. Mt. 6:1, pro.j to. qeaqh/nai. In the N. T. w[ste with the inf. of purpose is rare. Originally purpose was the idea with w[ste, or conceived result. Actual result with w[ste was expressed by the indicative.

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1089

Addenda 3rd ed.

In the LXX the notion of purpose is still common, especially in the books of Genesis, Exodus, Leviticus.245 In the N. T. there are only 8 instances, leaving out Ac. 20:24, according to W. H., and only 7 if we follow W. H. in Lu. 9:52. See Mt. 10:1, e;dwken auvtoi/j evxousi,an w[ste evkba,llein kai. qerapeu,ein. And w[ste $Êw`jà teà 'and so') is simply 'so as,' not 'so that.' See also Lu. 4:29, w[ste katakrhmni,sai. Cf. further Mt. 15:33; 27:1; Lu. 20:20. Burton246 thinks that in Mt. 27:1 w[ste gives rather content than purpose. One must not confuse with tou/ and the inf. of purpose the somewhat analogous construction of tou/ and tou/ mh, after verbs of hindering. This is in reality, as was shown, the ablative and the regular object-inf. (substantival aspect). Cf. Lu. 4:42; Ac. 20:27; Ho. 15:22. Votaw247 notes 22 verbs in the LXX and the N. T. that use this idiom. The only common one is kwlu,w. See further Final Clauses in chapter on Modes for papyri examples.

(h) Result. Purpose is only "intended result," as Burton248 argues. Radermacher (N. T. Gr., p. 153) says that the difference between purpose and result in the inf. is often only in the more subjective or objective colouring of the thought. It is hard to draw a line between conceived result and intended result. Blass249 explains a number of examples as result that I have put above under Purpose, as Rev. 5:5; 16:9. It is largely a matter of standpoint. The line of distinction is often very faint, if not wholly gone. Take Rev. 5:5, for instance, evni,khsen o` le,wn avnoi/xai. The lion had opened the book and so it was I actual result. So also Ac. 5:3, dia. ti, evplh,rwsen o` satana/j th.n kardi,an souà yeu,sasqai, se. Ananias had actually lied. In the ancient Greek also the distinction between purpose and result was not sharply drawn.250 The inf. may represent merely the content251 and not clearly either result or purpose, as in Eph. 3:6, ei=nai ta. e;qnh) Cf. also 4:22, avpoÄ qe,sqai. This is not a Hebraistic (Burton) idiom, but falls in naturally with the freer use of the inf. in the koinh,) See also Ac. 15:10 evpiqei/nai zugo,n, (Heb. 5:5) genhqh/nai avrciere,a. Where it is clearly result, it may be actual or hypothetical.252 The hypothetical is the natural or conceived result. The N. T. shows but 12

1090 A GRAMMAR OF THE GREEK NEW TESTAMENT

instances of the simple inf. with the notion of result, according to Votaw.253 In the O. T. it is quite common. The 12 examples in the N. T. are usually hypothetical, not actual. So Ro. 1:10 euvoÄ dwqh,somai evlqei/n pro.j u`ma/j (Eph. 3:17) krataiwqh/naià katoikh/sai, (6: 19) gnwri,sai, (Col. 4:3) lalh/sai,grk grk(4:6) eivde,nai, (Heb. 6:10) evpilaÄ qe,sqai. It is here that the kinship with purpose is so strong. Cf. Rev. 16:9. But some examples of actual result do occur, as in Lu. 10:40; Ac. 5:3; Rev. 5:5. In the 0. T.254 we have actual result with tou/ and the inf., but no examples occur in the N. T. Not more than one-half of the examples of tou/ and the inf. in Luke, who gives two-thirds of the N. T. instances, are final.255 Some of these are examples of hypothetical result. See discussion of Result in chapter on Mode for further discussion and papyri examples. It is rather common in the 0. T., though not so frequent in the N. T.256 It is possible to regard Mt. 21:32, metemelh,qhte tou/ pisteu/Ä sai, thus, though in reality it is rather the content of the verb.257 There is similar ambiguity in Ac. 7:19, evka,kwsen tou/ poiei/n. But the point seems clear in Ac. 18:10, ouvdei.j evpiqh,setai, soi tou/ kakw/sai, se, and in Ro. 7:3, tou/ mh. ei=nai auvth.n moicali,da) If tou/ can be occasionally used for result, one is prepared to surrender the point as to eivj to, if necessary. It is usually purpose, but there is ambiguity here also, as in Mt. 26:2; 1 Cor. 11:22, where the purpose shades off toward hypothetical result. In Ac. 7:19 we seem to have hypothetical result, eivj to. mh. zwogonei/sqai. So also Ro. 6: 12, eivj to. u`pakou,ein. It is true also of Heb. 11:3, eivj to. gegone,nai. See further Ro. 12:3; 2 Cor. 8:6; Gal. 3:17.258 Votaw259 argues for actual result in Ro. 1:20, eivj to. ei=nai auvtou.j avnapologh,touj. It is hard to deny it in this passage. But it is w[ste and the inf. that is the usual N. T. construction for this idea with the inf. As already shown (see Mode) nearly all of the 62 examples of w[ste and the inf. in the N. T. have the notion of result. Once Votaw260 notes an instance of hypothetical result in the N. T., 1 Cor. 13: 2, ka'n e;cw pa/san th.n pi,stin w[ste o;rh meqista,nein. Burton261 goes further and includes in this category Mt. 10:1; 2 Cor. 2:7. But these debatable examples are n harmony with the usual am-

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1091

biguity as to result and purpose. There is no doubt about the examples of actual result with w[ste. Thus Mt. 13:54 evdi,dasken auvtou.j w[ste evkplh,ssesqai kai. le,gein, (Mk. 9:26) w[ste tou.j pollou.j le,gein, (Lu. 12:1) w[ste peripatei/n avllh,louj, (Ac. 5:15) w[ste evkfe,Ä fein. See also Ac. 15:39; Ro. 7:6; 2 Cor. 7:7; Ph. 1:13, etc. There is one instance in the text of W. H. where w`j occurs with the inf., Lu. 9: 52, w`k e`toima,sai with the idea of purpose involved. Cf. w`j scei/n 0. P. 1120, 19 f. (iii/A. D.). The use of w`j e;poj eivpei/n (Heb. 7:9) is the absolute idea, as already shown. Different also is w`j a'n evkfobei/n (2 Cor. 10:9) = 'as if.' A clear case of result occurs in Epictetus, IV, 1, 50, ou[twj- mh. avpodu,Ä rasqai.

(i) Cause. There is only one example in the N. T. of the articular inf. without a preposition in this sense. That is in 2 Cor. 2:13, tw|/ mh. eu`rei/n, and it is in the instr. case as already shown. The LXX shows a half-dozen examples, but all with variant readings.262 But it is common with dia. to, to have the causal sense, some 32 times in the N. T.263 See Prepositions and Substantival Aspects of the Infinitive. Cf. Mt. 13:5 f.; Mk. 5:4; Lu. 6:48; Jas. 4:2 f. There is one instance of e[neken tou/ in 2 Cor. 7:12.

(j) Time. Temporal relations are only vaguely expressed by the inf. See Tense in this chapter for the absence of the timeelement in the tenses of the inf. except in indirect discourse. Elsewhere it is only by prepositions and pri,n, (an adverbial preposition in reality) that the temporal idea is conveyed by the inf. Antecedent time is expressed by pri,n or pro. tou/. For pro. tou/, see Mt. 6:8; Lu. 2:21, etc. Pri,n, or pro. tou/ (so in Mt. 1:18; Mk. 14:30; Ac. 7:2; W. H. have pri.n h; in the margil in Ac. 2:20) occurs with the inf. 11 times, all aorists (all in Gospels and Acts). We have it only twice with finite verb after negative sentences, once with the subj. (Lu. 2:26), once with the opt. (Ac. 25:16), both in Luke (literary style). See, for the inf.,264 Mt. 26:34 pri.n avle,ktora fwnh/sai, (Jo. 4:49) pri.n avpoqanei/n. See further Mt. 26:75; Mk. 14:72; Lu. 22:61 (five of the instances are practically identical); Jo. 8:58; 14:29; Ac. 2:20. In Herodotus, under the influence of indirect discourse, the inf. occurs with o[kwjà evpei, evpeidh,à eivà dio,ti and the relative pronouns.265 Con-

1092 A GRAMMAR OF THE GREEK NEW TESTAMENT

temporaneous action is described by evn tw|/, especially in Luke. Cf. Lu. 1:21, evn tw|/ croni,zein. See Prepositions with Infinitive for further remarks. Subsequent action is set forth by meta. to, as in Mt. 26:32; Lu. 12:5, etc. In Ac. 8:40, e[wj tou/ evlqei/n, we have the prospective future.

(k) The Absolute Infinitive. This idiom is very common in Homer, especially as an imperative and in the midst of imperatives.266 R. Wagner267 notes that in Homer this use of the inf. occurs with the nom. The papyri still show examples like o` dei/na tw|/ dei/na cai,rein.268 Gerhard269 holds that in such cases there is ellipsis of le,gei. The Attic inscriptions270 frequently have the absolute infinitive as imperative. Deissmann (Light from the Anc. East, p. 75) notes that, as in German, it is common in edicts and notices. Cf. imperatival use of infinitive in modern French. He quotes from the "Limestone Block from the Temple of Herod at Jerusalem" (early imperial period): Mhqe,na avllogenh/ eivsporeu,esqai evnto.j tou/ peri. to. i`ero.n trufa,ktou kai. peribo,lou, 'Let no foreigner enter within,' etc. See also Epictetus, IV, 10, 18, i[na de. tau/ta ge,nhtaià ouv mikra. de,xasqai ouvde. mikrw/n avpotucei/n. The imperatival use was an original Indo-Germanic idiom.271 It flourishes in the Greek prose writers.272 Burton273 and Votaw274 admit one instance of the imperatival inf. in the N. T., Ph. 3:16, tw|/ auvtw|/ stoicei/n. But Moulton275 rightly objects to this needless fear of this use of the inf. It is clearly present in Ro. 12:15, cai,reinà klai,ein. The case of Lu. 9:3 is also pertinent where mh,te e;cein comes in between two imperatives. Moulton himself objects on this point that this inf. is due to a mixture of indirect with direct discourse. That is true, but it was a very easy lapse, since the inf. itself has this imperatival use. In 1 Th. 3:11; 2 Th. 2:17; 3:5 there is the nominative case and the whole context besides the accent to prove that we have the optative, not the aorist active infinitive. See Mode for further discussion. Moulton276 quotes Burkitt as favouring the mere infinitive, not e;dei, in Mt. 23:23, tau/ta de. poih/sai kavkei/na mh. avfei/nai, after the Lewis Syriac MS., and also kauca/sqai.- in 2 Cor. 12:1 after a. The

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1093

imperatival use of the inf. was common in laws and maxims and recurs in the papyri.277 So A. P. 86 (i/A.D.) evxei/naià misqw/sai. Radermacher (N. T. Gr., p. 146) quotes Theo, Progymn., p. 128, 12, fe,re zhtei/n, where the inf. is used as a deliberative subj. would be. He gives also the Hellenistic formula, eivj du,namin ei=nai th.n evmh,n, Inscr. Pergam., 13, 31; 13, 34. Hatzidakis278 notes that in the Pontic dialect this construction still exists. The epistolary inf. has the same origin as the imperatival inf. It is the absolute inf. This is common in the papyri. See Ac. 15:23; 23:26; Jas. 1:1, cai,rein. The nom. is the nominative absolute al Cf. 2 Jo. 1:10, where cai,rein is the object of le,gete. Radermacner (N. T. Gr., p. 146) notes how in the later language the ace. comes to be used with the absolute inf., as in C. Inscr. lat. V. 8733, doune autwnÊ dou /nai auvto,n. It is just in this absolute inf. that we best see the gradual acquirement of verbal aspects by the inf. It is probably the oldest verbal use of the inf.279 The construction in Heb. 7:9, w`j e;poj eivpei/n, is but a step further on the way. There is but one example of this absolute inf. with w`j in the N. T.280 Cf. tou/ poleÄ mh/sai in Rev. 12:7, where it is an independent parenthesis.

(1) Negatives. The ancient Greek used mh, chiefly with the inf. except in indirect assertion where ouv of the directs was retained. But we see ouv with the inf. after verbs of saying as early as Homer, fh.j ouvc u`pomei/nai, Iliad, XVII, 174. Thus ouv won a place for itself with the inf., but many verbs retained mh, as verbs of swearing, hoping, promising, etc. But special phrases could have ouv anywhere and strong contrast or emphasis would justify ouv.281 Votaw282 finds 354 instances in the Greek Bible where the inf. itself is modified by the negative. Of these 330 have mh, and the rest have compounds of mh,. The anarthrous inf. with he notes 59 times in the 0. T., 32 in the Apocrypha and 47 in the N. T., 139 in all. The articular inf. with mh, he finds in the O. T. 136 times ( tou/ 99, to, 37), in the Apocrypha 21 times ( tou/10, to, 11), in the N. T. 35 times ( tou/ 15, to, 20), 192 in all ( tou/ 124; to, 68). With the anarthrous inf. the negative more frequently occurs with the principal verb as in ouv qe,lw. We do have ouv in infinitival clauses, as will be shown, but in general it is true to say that the inf. directly is always negatived by mh, in the N. T. This is true of

1094 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

Addenda 3rd ed.

all sorts of uses of the inf. So the subject-inf. uses mh,, as krei/tÄ ton h=n auvtoi/j mh. evpegnwke,nai (2 Pet. 2:21), both the anarthrous as above and the articular as in Lu. 17:1. The object-inf. likewise has mh,, as in Lu. 21:14, qe,te evn tai/j kardi,aij u`mw/n mh. promeleta/n. For the articular accusative with mh, see Ro. 14:13. We have it with indirect commands as in Mt. 5:34, le,ge u`mi/n mh. ovmo,sai, and in indirect assertion as in Ac. 23:8, le,gousin mh. ei=nai avna,stasin mh,te a;ggelon mh,te pneu/ma) We have it with tou/ mh, as in Jas. 5:17, tou/ mh. bre,xai, and with prepositions as in 2 Cor. 4:4, eivj to. mh. auvga,sai. With verbs of hindering and denying the negative mh, is not necessary, but it was often used by the ancients as a redundant negative repeating the negative notion of the verb, just as double negatives carried on the force of the first negative. It was not always used. When the verb itself was negatived, then mh. ouv could follow.283 But we do not find this idiom in the N. T. Examples of the N. T. idiom have already been given in this chapter. The variety in the N. T. may be illustrated. See Lu. 23:2 kwlu,onta fo,rouj Kai,sari dido,nai, (Ac. 4:17) avpeilhsw,meqa auvtoi/j mhke,ti lalei/n, (Gal. 5:7) ti,j u`ma/j evne,koyen avlhÄ qei,a| mh. pei,qesqai, (Ro. 15:22) evnekopto,mhn tou/ evlqei/n (Lu. 4:42) katei/con auvto.n tou/ mh. poreu,esqai, (Mt. 19:14) mh. kwlu,ete auvta. evlqei/n pro,j me, (1 Cor. 14:39) to. lalei/n mh. kwlu,ete (Ac. 14:18) mo,lij kate,pausan tou.j o;clouj tou/ mh. qu,ein auvtoi/j (Ac. 8:36) ti, kwlu,ei me baptisqh/nai,grk grk(10:47) mh,ti to. u[dwr du,nati kwlu,sai, tij tou/ mh. bapÄ tisqh/nai,grk grk(20:20) ouvde.n u`pesteila,mhn tou/ mh. avnaggei/lai. Radermacher (N. T. Gr., p. 149) illustrates "the Pauline to. mh, with the infinitive" by Sophocles' Electra, 1078, to, te mh. ble,pein e`toi/ma and the inscr. (Heberdey-Wilhelm, Reisen in Kilikien, 170, 2), to. mhde,n v a;llonÄÄevpeisenenkei/n. We may note also Ac. 4:20, ouv duna,meqa mh. lalei/n, where the negative is not redundant. Cf. also Jo. 5:19, ouv du,natai poiei/n ouvde,n, where the second negative is redundant, but it repeats the ouv. Some MSS. have a redundant negative mh, with eivde,nai in Lu. 22:34 (cf. 1 Jo. 2:22 after o[ti) and with prosteqh/Ä nai in Heb. 12:19. So AP read avntile,gontej in Lu. 20:27.

Even in indirect discourse the same negative is repeated, as in Ac. 26:26, lanqa,nein auvto.n tou,twn ouv pei,qomai ouvqe,n. Here ouvqe,n strictly goes with lanqa,nein in spite of its position after pei,qomai, but ouv is construed with pei,qomai, and so ouvqe,n, is used rather than mhqe,n or mhde,n. But in Mk. 7:24, ouvde,na h;qelen gnw/nai, it is not best to explain ouvde,na with the inf. in this fashion. This looks like the retention of the old classic use of ouv with the inf. which

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1095

the grammars are not willing to allow in the N. T.284 Epictetus uses ouv with the inf. as in IV, 10, 18, ouv mikra. de,xasqai ouvde. mikrw/n avpotucei/n. As a matter of fact we have a number of other examples of ouv with the inf., too many to rule out without ceremony. There is the case in.Heb. 7:11, ti,j e;ti crei,a kata. th.n ta,xin Melcisede.k e[teron avni,stasqai kai. ouv kata. th.n ta,xin le,gesqai; It is true that a comes just before kata. th.n ta,xin but it is rather orced to deny it any connection with le,gesqai. Sec also Ro. 8:12, ovfeile,tai ouv th|/ sarki. tou/ kata. sa,rka zh/n, where, however, a occurs outside of tou/ and is directly concerned with th|/ sarki,. Other examples of sharp contrast by means of ouv are found, as in Ac. 10:40 f., e;dwken auvto.n evmfanh/ gene,sqaià ouv panti. tw|/ law|/ avlla. ma,rtusi; Ro. 7:6, w[ste douÄ leu,ein evn kaino,thti pneu,matoj kai. ouv palaio,thti gra,mmatoj* Heb. 13:9, bebaiou/sqai ouv brw,masin (but here no contrast is expressed). In Ro. 4:12, 16, with eivj to,, we find ouv mo,non- avlla. kai,.

(m) ;An with the Infinitive. This classic idiom has vanished from the N. T. save in 2 Cor. 10:9, w`j a'n evkfobei/n. Even here it is not a clear case, since evkfobei/n depends on do,xw and w`j a;n to comes in as a parenthetical clause, 'as if' ('as it were').

The treatment of the infinitive has thus required a good many twists and turns due to its double nature.

III. The Participle ( h` metoch,).

1. THE VERBALS IN - toj AND - te,oj. These verbals are not exactly participles inasmuch as they have no tense or voice. They are formed from verb-stems, not from tense-stems, and hence are properly called verbal adjectives.285 In the broadest sense, however, these verbals are participles, since they partake of both verb and adjective. Originally the infinitive had no tense nor voice, and the same thing was true of the participle. For convenience we have limited the term participle to the verbal adjectives with voice and tense. The verbal in - toj goes back to the original Indo-Germanic time and had a sort cf perfect passive idea.286 This form is like the Latin -tus. Cf. gnwto,j, notus; a;gnw toj, ignotus. But we must not overdo this point. Strictly this pro-ethnic -tos has no voice or tense and it never came to have intimate verbal connections in the Greek as it did in Latin and English.287 Thus amatus est and avgaphto,j evstin do not correspond, nor, in truth, does 'he is loved' square with either. "Even in Latin, a word like tacitus illustrates the absence of both tense

1096 A GRAMMAR OF THE GREEK NEW TESTAMENT

and voice from the adjective in its primary use."288 Already in the Sanskrit voice and tense appear with some of the participles, but "the division-line between participial and ordinary adjectives is less strictly drawn in Sanskrit than in the other IndoEuropean languages."289 The ambiguity due to the absence of voice in the verbal in -- toj was inherited from the original IndoGermanic time.290 It becomes, therefore, a lexical, not a syntactical problem to decide in a given instance whether the verbal is "active" or "passive" in signification. In itself it is neither. A similar problem is raised in compound adjectives like qeoÄma,coi (Ac. 5:39), 'fighting God.' In modern Greek the verbal in -- toj is rare and is little more than an adjective (Thumb, Handb., p. 151), though the new formation in - a,toj has more verbal force. This ambiguity appears in Homer and all through the Greek language.291 Blass292 overstates it when he says that in the N. T. "the verbal adjective has practically disappeared, with the exception of forms like dunato,j, which have become stereotyped as adjectives." As a matter of fact the verbal in - toj is still common in the N. T. as in the koinh, in general. Take, for instance, avgaphto,jà a;gnwtojà avdu,natojà avkata,gnwtojà avnama,rthtojà avnekto,jà avo,ratojÃa;pistojà avpo,blhtojà avresto,jà avrketo,jà gennhto,jà grapto,jà didakto,jà dunato,jà euvloghto,jà zesto,jà qaumasto,jà qnhto,jà qeo,pneustojà o`rato,jà paqhto,jà parei,saktojà pisto,jà fqarto,jà crhsto,j, etc. It is true293 that the tendency is rather to accent the adjectival aspect at the expense of the verbal idea of these words. But this also was true at the start, as we have just seen in the Sanskrit. The point to note is that the verbal does not denote voice. In Ac. 14:8; Ro. 15: 1, avdu,naton is 'incapable,' whereas usually it is 'impossible,' as in Mt. 19:26 =Mk. 10:27, etc. In Ro. 8:3, therefore, it is doubtful whether to. avdu,naton tou/ no,mou is the 'impotency' or the 'impossibility' of the law.294 There is no' notion of tense or of Aktionsart in these verbals in - toj and so avgaphto,j does not distinguish295 between avgapw,menojà avgaphqei,j and hvgaphme,noj. Moulton thus properly notes the fact that in Mt. 25:41 we have kathrame,noi, 'having become the subjects of a curse,' not kata,raÄ toi, 'cursed.' It is interesting to note cara|/ avneklalh,tw| kai. dedoÄ xasme,nh| in 1 Pet. 1:8, but here avnekla,lhtoj is active in sense,

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1097

'inexpressible.' The ambiguity comes also in our English participle 'borne' used for aivro,menon in Mk. 2:3, and the punctiliar 'brought' used for evnecqei/san in 2 Pet. 1:18. With these Moulton296 contrasts hvrme,non ('taken away') in Jo. 20:1. It is worth while to study a few more examples from the lexical point of view. In general297 the passive sense is more common, as in avgaphÄ to,j (Mt. 3:17); eu;qetoj (Lu. 9:62); didakto,j (Jo. 6:45); qeo,pneuÄ stoj (2 Tim. 3:16); qeodi,daktoj (1 Th. 4:9); grapto,j and krupto,j (Ro. 2:15 f.).298 Here (Ro. 2:15 f.) ta. krupta, is used just like a substantive (neuter adjective in plural). But zesto,j (Rev. 3:15) is active in sense as avsu,netoj (Ro. 1:31), though avsu,nqetoj next to it (paronomasia) is made from the middle sunti,qemai ('covenant').299 Suneto,j, sometimes passive in sense in the old Greek, is always active in the N. T., as in Mt. 11:25, but qnhto,j (Ro. 6: 12) is 'liable to death,' not 'dying,' as paqhto,j (Ac. 26:23) is 'capable of suffering.' Cf. the Latin adjectives in -bilis.

The verbal in - te,oj is later than that in - toj and does not occur in Homer. It is probably a modification of the verbal - toj to express the idea of the predicate-infinitive, like 'this is not to eat (to be eaten).'300 It is really a gerundive and is used in the personal or impersonal construction, more commonly the latter.301 The personal is always passive in sense, while the impersonal is active and may be formed from transitive or intransitive verbs.302 It expresses the idea of necessity. It wall never as common as the verbal in - toj and is not unknown in the papyri,303 though not frequent. It is more like the verb (and participle) than the verbal in - toj in one respect, that it often uses the cases of the regular verb.304 This is seen in the one example n the N. T. (Lu. 5:38) oi=non ne,on eivj avskou.j blhte,on. It is the impersonal construction, though the agent is not here expressed. This example of - te,on in Luke is a survival of the literary style (cf Viteau, "Essai sur la Syntaxe des Voix," Revue de Philologie, p. 38). See Theo, Progymn., p. 128, 12, eiv gamhte,on.

1098 A GRAMMAR OF THE GREEK NEW TESTAMENT

2. HISTORY OF THE PARTICIPLE.

(a) The Sanskrit Participle. This was more advanced in its development than the Sanskrit infinitive, which had no voice or tense. In the Veda the aorist present, perfect and future tenses have participles.305 The distinction in the structure of the participle as compared with the other verbal adjectives lies just in this point. The mere verbal is formed on the verb-stem, while the participle is formed on the tense-stem.306 In the Sanskrit also both voices (active and middle) show these participles. Thus already in the original Indo-Germanic tongue it appears probable that the participle existed with voice, tense, Aktionsart and government of cases.307 The Greek participle is thus rooted in this pro-ethnic participle as seen by the very suffixes -nt--, -meno-, -wos- (-us).308

(b) Homer's Time. Already in Homer and Hesiod the participle occurs as a fully developed part of speech. It occurs on an average of 8 1/6 times per page of 30 lines.309 In Hesiod the participle is chiefly attributive, while the predicate participle is less common than in Homer.310 This use of the participle as the practical equivalent of the hypotactic clause is a purely Greek development (copied by the Latin to some extent) within historical times.311 The participle is a literary device, and flourished best with writers of culture who were filome,tocoi.312 Broadus used to call the Greek "a participle-loving language," and, taken as a whole, this is true. Certainly the participle had its most perfect development in the Greek. The aorist participle died in the Sanskrit and did not appear in the Latin. It is the aorist active participle which made the participle so Powerful in Greek. The English, like the Sanskrit and the Greek, is rich in participles, though the German is comparatively poor. "We gain a certain grandeur and terseness by the constructidn, a certain sweep, a certain periÄ bolh,, such as Hermogenes recognises as lying in the participle."313 This wealth of participles gives flexibility and swing td the language.

(c) The Attic Period. In Herodotus the participle jumps to

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1099

17 1/2 times per page of 30 lines.314 But Sophoeles has it only 9 times on the same scale. Williams315 runs the parallel on with 13 for Thucydides, 125 for Xenophon, 10 1/6 for Platt, 10 3/4 for Demosthenes. It is thus in the historians and orator and not the poets, that we see the participle in its glory.

(d) The Koinh,. Here we note a sharp difference in the several styles of writing. The Atticists like Josephus with 20, and 2 Maccabees with 23 1/2, lead in conscious imitation of the ancients. They go beyond them in fact. But the writers of the literary koinh, follow close behind, as Polybius with 17 4/5, Strabo with 13 1/2 and Plutarch with 14. Certainly there is no sign of decay here. But in the LXX, Exodus, Deuteronomy and Judges give only 6 1/6 while316 the papyri show 6 4/5. This confirms the judgment that the vernacular was not fond of the participle and found it clumsy. Jannaris317 quotes striking passages from Thucydides, Plato and Demosthenes which illustrate well the clumsiness and ambiguity of the participle in long, involved sentences. Even in the older Greek in unconventional or unscholarly composition the accumulation of participles is shunned. The clearer and easier analysis of co-ordinate or subordinate clauses was used instead.318 In the N. T. we see the participle used on the whole more frequently than in the LXX and the papyri. The Hebrew had a certain restraining influence on the participle in the LXX. Inthe vernacular papyri the participle was held back on the principle just stated above. It is Luke who make most frequent use of the participle with 161 in the Gospel and 17 1/6 in the Acts per page of 30 lines.319 But 1 Peter follows close behind with 15 2/3 and Hebrews with 14. In the other Gospels Matthew has it 12 1/2, Mark 11 2/3 and John 10 2/5.320 James has it 10 per page, while in the Epistles and Revelation it drops back to 8 and 9. On the whole it is much as one would expect. The more literary books lead (after Paul with only 9 per page average in Gal., 1 Cor., and Rom.).321 The historical books surpass the Epistles, while Hebrews here reveals its hortatory, sermonic character. For a succession of participles see Ac. 12:25; 23:27; Heb. 1:3 f.; Mk. 5:15. The details of the N. T. situation will come later.

(e) Modern Greek. The participle more and more came to be

1100 A GRAMMAR OF THE GREEK NEW TESTAMENT

scholastic and dropped out of the vernacular.322 In particular was this true of the circumstantial participle. The classic Greek by means of the participle developed the periodic style ( le,xij katestramme,nh) and is seen at its highest in Isocrates. See, for example, the "Ciceronian period" in Isocrates, p. 82. Jebb323 contrasts this with le,xij eivrome,nh, simply tacking clause to clause as in Mt. 7:25, 27 and colloquial repetition of finite verbs as in Jo. 1 47; 7:4. But ble,peteà ble,peteà ble,pete (Ph. 3:2) has rhetorical effect. In the vernacular modern Greek, therefore, we see a retreat of the participle all along the line. It is not dead as the infinitive, but is dying, though some vernacular writers are bringing back the use of the participle for literary purposes (Thumb, Handb., p. 168). The analytic tendency of modern language is against it. See Jebb's remarks for the various devices used instead of the participle. The only participles left in modern Greek are the indeclinable present active in - ontaj (cf. gerund in Latin), some middle (or passive) parts. in - ou,menoj or - a,menoj and perfect passives like deme,noj (no reduplication).324 A few are made from aorist stems like ivdwme,noj (Thumb, Handb., p. 150). The use of the part. in the modern Greek is very limited indeed.

3. SIGNIFICANCE OF THE PARTICIPLE.

(a) Originally an Adjective. The infinitive was originally a substantive, as we have seen. In the Sanskrit it did not acquire voice and tense, though it had the verbal idea of action. The participle, as we have seen, had made more progress in the Sanskrit, but it was also originally an adjective. It never got away from this original adjectival idea.325 But we are not left to history and logic to prove this point. It so happens that some participles in form never became participles in fact. They are merely adjectives. Homer shows a number of such words.326 Cf. a;sÄmenoj) We see remnants of this usage in the N. T. like e`kw,n (Ro. 8:20), a;kwn (1 Cor. 9:17). Other participles come in certain uses to be only substantives (adjectives, then substantives), though the true participial use occurs also. Cf. a;rcwn, 'a ruler' (Mt. 20: 5); h`gou,menoj, 'a governor' (Ac. 7:10); ta. u`pa,rconta u`mw/n, 'your belongings' (Lu. 12:33). In general "the adjective represents a quality at rest, the participle represents a quality in motion."327 But

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1101

not all verbs express motion. The mere adjectival notion is more common in the Latin, as in praeteritus, quietus, tacitus, etc. In Mt. 17:17, genea. a;pistoj kai. diestramme,nh, the verbal adjective and participle occur together.

(b) The Addition of the Verbal Functions. These functions are tense, voice and case-government. There was originally no notion of time in the tense, nor does the tense in the participle ever express time absolutely. It only gives relative time by suggestion or by the use of temporal adverbs or conjunctions.328 The verbal idea in the participle thus expands the adjectival notion of the word.329 But the addition of these verbal functions does not make the participle a real verb, since, like the infinitive, it does not have subject.330

(c) The Double Aspect of the Participle. The (very name participle (pars, capio) indicates this fact. The word is part adjective, part verb. Voss calls it mules, which is part horse and part ass.331 Dionysius Thrax says: Metoch, evsti le,xij mete,cousa th/j tw/n r`hma,twn kai. th/j tw/n ovnoma,twn ivdio,thtoj. In the true participle, therefore, we are to look for both the adjectival and the verbal aspects, as in the infinitive we have the substantival and the verbal. The emphasis will vary in certain instances. Now the adjectival will be more to the fore as in the attributive articular participle like o` kalw/n.332 Now the verbal side is stressed as in the circumstantial participle. But the adjectival notion never quite disappears in the one as the verbal always remains in the other (barring a few cases noted above). One must, therefore, explain in each instance both the adjectival and verbal functions of the participle else he has set forth only one side of the subject. It is true that the verbal functions are usually more complicated and interesting,333 but the adjectival must not be neglected.

(d) Relation between Participle and Infinitive. As already explained, they are closely allied in use, though different in origin. Both are verbal nouns; both are infinitival; both are participial. But the participle so-called is inflected always, while the infinitive so-called has lost its proper inflection. The infinitive, besides, expresses334 the action in relation to the verb, while the participle expresses the action in relation to the subject or the object of the

1102 A GRAMMAR OF THE GREEK NEW TESTAMENT

verb (or some other substantive or pronoun).335 The distinction between the participle and the infinitive thus becomes quite important. Thus in Lu. 16:3, evpaitei/n aivscu,nomai, the idea is 'I am ashamed to beg and do not do it,' while evpaitw/n aivscu,nomai would be 'I beg and am ashamed of it.'336 Cf. the analytic expression in 2 Tim. 1:12. In Xenophon, Mem., 2, 6, 39, we have aivscu,noÄ mai le,gwn. So a;rcomai in Attic Greek took the infinitive as a rule, linking the infinitive with the verb. But sometimes the participle occurred, linking the action to the subject (or object) and so contrasting the beginning with the end.337 In the N. T. all the examples have the present infinitive except Lu. 13:25 e`sta,nai. In Lu. 3:23, avrco,menoj w`sei. evtw/n tria,konta, we have neither with avrco,menoj. Cf. Lu. 14:30, h;rxato oivkodomei/n. Radermacher (N. T. Gr., p. 169) compares avrxa,menoj evxeti,qeto (Ac. 11: 4) with avrxame,nh- kate,comai (Xen. of Eph., p. 388, 31). On the other hand, in the N. T. pau,omai occurs only with the participle, as in Lu. 5:4, evpau,sato lalw/n. Cf. Ac. 5:42; 6:13; Eph. 1:16; Col. 1:9; Heb. 10:2. But in Ac. 14:18 note kate,pausan tou/ mh. qu,ein, which well illustrates the difference between the inf. and the part. The use of evte,lesen diata,sswn (Mt. 11:1) Blass338 calls unclassical. The part. alone occurs with evnkake,w (Gal. 6:9; 2 Th. 3:13). Note also evpe,menon evrwtw/ntej (spurious passage in Jo. 8:7), but a;sitoi diatelei/te (Ac. 27:33) without o;ntej. Cf. Ac. 12:16, evpe,menen krou,wn, and Lu. 7:45, ouv die,lipen katafilou/sa. Radermacher (N. T. Gr., p. 169) finds the part. with evpime,nw in "vulgar literature." He observes that many of these neater classical idioms with the part. do not appear in the N. T. Contrast with this the inf. in Ac. 20:20, 27, ouv ga.r u`pesteila,mhn tou/ mh. avnaggei/lai. There is no example of the inf. with fai,nomai in the N. T., but the part. occurs in Mt. 6:16, 18 ( nhsteu,wn). The adjective alone is seen in Mt. 23:27, 28. Cf. also Ro. 7:13. It is hardly on a par with the participle in Mt. 6:17 in spite of Blass's insistence.339 Thoroughly classical also are proe,fqasen auvto.n le,gwn (Mt. 17:25) and e;laqon xeni,santej (Heb. 13: 2), specimens of literary style. The infinitive with profqa,nw occurs in Clem., Cor., II, 8, 2. The part. with tugca,nw does not occur in the N. T. In the later koinh, the inf. takes the place of the participle with lanqa,nwà pau,omai and fqa,nw (Radermacher, N. T. Gr., p. 169). The part. is found with u`pa,rcw

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1103

(Ac. 8:16) and prou?pa,rcw (Lu. 23:12). It is doubtful if the participle belongs to the verb in 1 Tim. 5:13, avrgai. manqa,nousin perierco,menai, but, if so, it is not to be understood as like the inf.340 In Ph. 4:11; 1 Tim. 5:4, the inf. occurs with manqa,nw according to classic idiom. At any rate, if perierco,menai (1 Tim. 5:13) is a circumstantial part., something has to be supplied with avrgai,. The part. in 1 Tim. 1:12, pisto,n me h`gh,sato qe,menoj, is certainly circumstantial. The distinction between the inf. and the part. comes out sharply in indirect discourse also. The inf. is more objective. Thus note h;kousan tou/to auvto.n pepoihke,nai to. shmei/on (Jo. 12:18) and avkou,omen ga,r tinaj peripatou/ntaj (2 Th. 3:11). The participle is a descriptive adjective even though in indirect discourse (cf. Lu. 4:23; Ac. 7:12). See 1 Cor. 11:18 for the inf. again. In Mt. 7:11, oi;date do,mata avgaqa. dido,nai, the inf. with oi=da means 'know how to give.' But in Lu. 4:41, h|;deisan to.n Cristo.n auvto.n ei=nai, it is mere indirect discourse. For lithe part. see 2 Cor. 12:2, oi=da- a`rpage,nta to.n toiou/ton (cf. Mk. 6:20). In Ac. 3:9 note ei=den auvto.n peripatou/nta. Here we have the same root, though a different sense. Oi=da is common with o[ti. But ginw,skw occurs both with the inf. as in Heb. 10:34, ginw,skontej e;cein e`auÄ tou.j krei,ssona u[parxin, and the participle as in Heb. 13:23, ginw,Ä skete to.n avdelfo.n h`mw/n Timo,qeon avpolelume,non. Cf. Lu. 8:46, evgw. e;gnwn du,namin evxelhluqui/an, where the tense and participle both accent the vivid reality of the experience. But note the inf. in Mt. 16:13. The same thing is true of o`mologe,w as in Tit. 1:16, qeo.n o`mologou/sin eivde,nai, and 1 Jo. 4:2, oa} o`mologei/ vIhsou/n evn sarki. evlhluqo,ta (cf. 2 Jo. 1:7). Cf. also Ac. 24: 10 o;nta se krith.n evpista,menoj and dokima,zw in 1 Th. 2:4 and 2 Cor. 8:22. Note difference between i[na eu[rwsin kathgorei/n auvtou/ (Lu. 6:7) and eu`ri,skei auvtou.j kaqeu,dontaj (Mk. 14:37). Cf. Indirect Discourse. Further examples of the supplementary participle come later. These sufficiently illustrate the difference between the use of inf. and part.

(e) Method of Treating the Participle. The hybrid character of the participle has led to a great deal of diversity in its treatment in the grammars. Prof. Williams341 gives an interesting summary in his monograph. None of them are satisfactory because they do not follow a consistent plan. Part of the, divisions are from the adjectival, part from the verbal point of view. They are not parallel. Thus we have Kuhner's complementary, attributive, adverbial participles; Goodwin's attributive, circumstantial, supplementary; Burton's adjectival, adverbial, substantival;

1104 A GRAMMAR OF THE GREEK NEW TESTAMENT

Jannaris' adjectival and adverbial; Blass' attributive and in additional clause; Hadley and Allen's attributive and predicate; Delbruck-Brugmann's external, objective, adverbial. Then Williams342 adds another that no better, ascriptive, adverbial, complementary. Thompson343 gives the attributive and the supplementary participle after saying that the nominal and the verbal classification is more elastic. The only way to get symmetry in the treatment of the participle is to follow the line of its double nature (adjectival and verbal) and discuss the adjectival functions and the verbal functions separately. See the discussion of the infinitive. That is to say, each participle must be considered as both adjectival and verbal. Not all the adjectival aspects will be true of any one participle nor all of the verbal, but each one will have some adjectival and some verbal functions. Thus alone can one get a clear statement of the many participial combinations and permutations. As an Adjective the participle is attributive (anarthrous or articular) or predicate. It may even be substantival, especially with o`. It is always declinable. As a verb there is always voice and tense and there may be cases. But any given anarthrous predicate participle may be either supplementary (complementary) or circumstantial (additional) or wholly independent (as indicative or imperative). The articular participle is ruled out of this three-fold alternative, though it still has voice, tense and governs cases. The articular participle is always attributive (or substantival). The lines thus cross and recross in the nature of the case. But a clear statement of all the essential facts can be made by taking the adjectival and the verbal aspects separately. In any given instance there is thus a double problem. Both sides of the given participle must be noted.

4. ADJECTIVAL ASPECTS OF THE PARTICIPLE.

(a) Declension. The free declension of the participle in number and gender and case (cf. per contra the infinitive) makes the task of noting the adjectival aspects comparatively simple. There are anomalies of agreement in these three points as with other adjectives. Thus in Rev. 3:12 h` katabai,nousa in apposition with th/j kainh/j vIer. does not conform in case. There is a difficulty of both case and gender in pepurwme,nhj in Rev. 1:15. See also plh/qoj kra,zontej (Ac. 21:36) where the number and gender both vary. In Mk. 4:31 note o[j oa}n pa,ntwn tw/n sperma,twn where o;n takes the gender of spe,rma. Cf also h=n kaqh,menai (Mt. 27:61).

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1105

But these matters are discussed adequately in chapter on The Sentence.

(b) Attributive Participle.

(a) Anarthrous. The article is not of course necessary with the attributive participle any more than with any other attributive adjective. Thus we have u[dwr zw/n (Jo. 4:10), 'living water,' which is just as really attributive as to. u[dwr to. zw/n (Jo. 4:11). When the article is used there is no doubt abort the participle being attributive. When it is absent, it is an open question to be examined in the light of the context. Note also 1 Cor. 13:1, calko.j hvcw/n h' ku,mbalon avlala,zon. This construction (the anarthrous attributive) is not so common as the other uses of the participle,344 and yet it is not wholly absent from the N. T. See h=coj w[sper ferome,nhj pnoh/j biai,aj (Ac. 2:2) and qu,ra hvnew|gme,nh (Rev. 4:1). It is not always easy to draw the line between the anarthrous attributive participle and the predicate participle of additional statement. Cf. avnh.r gegennhme,noj evn Tarsw|/ avnateqramÄ me,noj de. evn tw|/ po,lei tau,th| (Ac. 22:3). If o` occurred before these participles, we should have the articular-attributive participle which is equivalent to a relative.345 So in Ac. 10:18, we have o` evpikalou,Ä menoj Pe,troj, but in 10:32, oa}j evpikalei/tai Pe,troj. Cf. Lu. 6:48, o[moio,j evstin avnqrw,pw| oivkodomou/nti oivki,an, with Mt. 7:24, avndri. ou[stij w|vkodo,mhsen auvtou/ th.n oivki,an. See also Lu. 6:49. Cf. Ro. 8:24, evlpi.j blepome,nh ouvk e;stin evlpi,j. Cf. Mt. 27:33. The problem is particularly real in Mk. 5:25, 27. W. H. indicate by the comma after evlqou/sa that they regard the participles with gunh, ( ou=saà paÄ qou/saà dapanh,sasaà wvfelhqei/saà evlqou/sa) up to that point as attributive. They describe the woman who comes. Then the sentence proceeds with the predicate-circumstantial participles ( avkou,sasa evlqou/sa% before h[yato. Luke Luke(8:43) makes the matter plainer by putting a relative clause after the first participle. The anarthrous attributive participle is closely bound to the substantive or pronoun even when it is an additional statement. See Mt. 12:25, pa/sa basilei,a merisqei/sa kaq v e`auth/j evrhmou/tai. See also Lu. 6:0; 2 Th. 2:4; Rev. 2:15. In Mt. 13:19, panto.j avkou,onÄ toj, we probably have the genitive absolute and so predicate circumstantial, but even here auvtou/ occurs, though remote. Cf. pa/j o` avkou,wn (Mt. 7:26) and pa/j o[stij avkou,eigrk grk(7:24), where we see how nearly these constructions approach each other.346 But the anar-

1106 A GRAMMAR OF THEGREEK NEW TESTAMENT

Addenda 3rd ed.

Addenda 3rd ed.

throus indefinite participle is clearly found in Jas. 4:17, eivdo,ti ou;n kalo.n poiei/n kai. mh. poiou/ntià a`marti,a auvtw|/ evstin) This passage may throw some light on Mt. 12:25. In Mt. 13:35, dia. tou/ profh,tou le,gontoj, we probably have the articular attributive participle, since the Greeks did not always place the attributive participle between the article and the substantive.347 The use of e;cwn is interesting in Rev.15:1, ei=don avgge,louj e`pta. e;contaj plhga,j. The anarthrous indefinite participle is seen also in a few constructions like proseti,qento pisteu,ontej tw|/ kuri,w| (Ac. 5:14), where the participle means 'believing men' and has plh,qh in apposition with it. See also fwnh. bow/ntoj (Mk. 1:3, LXX), evxeleu,setai h`gou,Ä menoj (Mt. 2:6, LXX), ouvk e;stin suni,wn and ouvk e;stin evkzhtw/n (Ro. 3:11, LXX) where o` is morel common, e;ceij evkei/ kratou/ntaj (Rev. 2:14). It is worth noting in this connection also the fact that occasionally a preposition occurs with an anarthrous participle (cf. infinitive). So cwri.j khru,ssontoj (Ro. 10:14). Here the idea is not 'without preaching,' but 'without one preaching,' 'without a preacher.' For 'without preaching' we must have cwri.j tou/ khru,ssein. See once (more cai,rein meta. cairo,ntwnà klai,ein meta. klaio,ntwngrk grk(12:15) and evpi. poiou/ntaj (1 Pet. 3:12). In 1 Cor. 15:27, evkto.j tou/ u`pota,xantoj, we have the usual articular construction.

( b) Articular. The articular participle occurs a few times in Homer.348 In general the Book of Acts has the articular participle in about the same proportion as the great Attic writers.349 All articular participles are, of course, attributive. But the matter has some points of interest and cannot be dismissed with this general statement. The examples are very numerous. The substantives may be expressed as in th.n h`toimasme,nhn u`mi/n basilei,an (Mt. 25:34); oi` grammatei/j oi` avpo. vIerosolu,mwn kataba,ntej (Mk. 3: 22). Like other articular adjectives, the participle may come between the article and the substantive, as in th|/ u`giainou,sh| didaskaÄ li,a| (1 Tim. 1:10); tou/ fainome,nou avste,roj (Mt. 2:7); th/j prokeime,nhj auvtw|/ cara/j (Heb. 12:2). Cf. Jude 1:3. The substantive may precede and the article may be repeated, as to. u[dwr to. zw/n (Jo. 4:11); to. sw/ma to. genhso,menon (1 Cor. 15:37); tw|/ qew|/ tw|/ dido,nti (1 Cor. 15: 67). Cf. Mt. 26:28; 27:44; Jam. 5:1; Ro. 2:10. In Mk. 12:38 the article is repeated as in 1:40 (apposition) when the nominative reminds us of the common anacoluthon in Revelation.

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1107

With proper names note vIhsou/j o` lego,menoj Cristo,j (Mt. 1:16); o` evpikalou,menoj Pe,troj (Ac. 10:18). Cf. 1 Th. 1:10; 2 Tim. 1: 8 f. For a long passage see o` - dida,skwn (Ac. 21:28). The order of the words is not insisted on and in long passages the participle may follow without the repetition of the article, as in Mt. 6:30, to.n co,rton tou/ avgrou/ sh,meron o;nta kai. au;rion eivj kli,banon ballo,menon) See also Ac. 12:10; 13:32; 26:4, 6; Heb. 2:2; Heb. 12:3, where in the long clause the participle with toiau,thn, comes in between to.n and u`pomemenhko,ta and a good distance from avntilogi,an. Sometimes the article is, used with the participles but not with the substantive, as in paidi,oij toi/j evn avgora|/ kaqhme,noij (Lu. 7: 32); crusi,ou tou/ avpollume,nou (1 Pet. 1:7); o;noma to. dedome,non (Ac. 4:12); polu.j avriqmo.j o` pisteu,saj (Ac. 11:21); polloi. pla,noi oi` mh. o`mologou/ntej (2 Jo. 1:7); a;nqrwpoi oi`ÄÄavrnou,menoi, (Jude 1:4, where note the series of participles and one adjective avsebei/j parallel with the participles). Cf. also 1 Cor. 2:7. The articular participle also occurs with pronouns,350 as in su. o` evrco,menoj (Mt. 11: 3); tina.j tou.j pepoiqo,taj (Lu. 18:9); tij o` sulagwgw/n (Col. 2:8); auvtoi/j toi/j pisteu,ousin (Jo. 1:12); su. o` kri,nwn (Jas. 4:12); tine.j oi` tara,ssontej (Gal. 1:7); polloi. oi` fronou/ntej (Ph. 3:18. f.). Particularly in address do we find the articular participle, as in Mt. 7:23; 27:40; Lu. 6:25 (but note dative in 6:24); Ac. 2: 14; 13:16. The use of the articular participle with pa/j is common, as pa/j o` ovrgizo,menoj (Mt. 5:22); pa/j o` avkou,wn (Mt. 7:26), pa/j o` le,gwngrk grk(7:21). This is equal to the relative clause pa/j o[stij (Mt. 7:24). In Ro. 2:1 pa/j o` kri,nwn is used with a;nqrwpe. Cf. pa,ntej oi` avkou,ontej in Ac. 9:21. Here also o` porqh,saj, is continued by kai. evlhlu,qei as if it were a relative clause. The articular participle sometimes occurs where it is followed by an infinitive. Here it is still further complicated, but it is clear. See th.n me,llousan do,xan avpokalufqh/nai (Ro. 8:18); ta. dokou/nta me,lh- u`pa,rcein (1 Cor. 12:22). Cf. also 2 Pet. 3:2. The use of o` w;n in Acts calls for special remark. In Ac. 13:1, kata. th.n ou=san evkklhsi,anà we see this idiom, which Moulton351 translates 'the local church.' Note 14:13 D, tou/ o;ntoj Dio.j Propo,lewj (or pro. po,lewj). Cf. Ramsay's remark (Ch. in Rom. Emp., p. 52, quoting J. A. Robinson), that in Acts o` w;n, "introduces some technical phrase, or some term which it. marks out as having a technical sense (cf. 5:17; 13:1 28:17), and is almost equivalent to tou/ ovnomazonme,nou." An ingenious person might apply this in Eph. 1:1 to the text with evn vEfe,sw| absent; but the usual view needs no defence against such an alternative.

1108 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

With ai` ou=sai in Ro. 13:1 we may compare Par. P. 5 (ii/B.C.), evf v i`ere,wn kai. i`ereiw/n tw/n o;ntwn kai. ouvswn. So N. P. 49 (iii/A.D.), tou/ o;ntoj mhno,j 'the current month.' The passage in Ac. 5:17 reads h` ou=sa ai;resij, and 28:17 has tou.j o;ntaj tw/n vIoudai,wn prw,touj. Moulton agrees, we may note, with Sanday and Headlam (in loco) in taking o` w'n evpi. pa,ntwn (Ro. 9:5) as referring to Jesus. As is well known, the difficulty here is a matter of exegesis and the punctuation of the editor will be made according to his theology. But it may be said in brief that the natural way to take o` w;n and qeo,j is in apposition to o` Cristo,j. It is a very common thing in the N. T., as already noted, to have o` and the participle where a relative clause is possible. But this idiom is common in the older Greek. See Ac. 10:18, 32, and chapter on Article. It remains then to speak of the frequent use of thearticular participle without a substantive or pronoun. This idiom is too common for exhaustive treatment, but some examples are given. Cf. Mt. 10: 40, o` deco,menoj u`ma/j evme. de,cetaià kai. o` evme. deco,menoj de,cetai to.n avpostei,lanta, me. Note also o` deco,menoj and the next verse and oa}j a'n poti,sh| in verse 42. See further Mt. 10:37; Ac. 10:35; Rev. 1: 3. The question of the tense is interesting in some of these examples, as in o` eu`rw.n th.n yuch.n auvtou/ avpole,sei auvth,n in Mt. 10:39, but that will be discussed a bit later. Like a relative clause, the articular participle may suggest352 the notion of cause, condition, purpose, etc., as in Mt. 10:37, 39, 40, 41; Lu. 14:11; Ro. 3:5. But this notion is very indefinite.

(c) Predicate Participle. From the adjectival standpoint all participles that are not attributive are predicate. This aspect of the participle must be elucidated further. The verbal aspect comes into special prominence with all the predicate participles. They will be touched very lightly here and receive full discussion under Verbal Aspects. It may be said at once that all the supplementary and circumstantial participles are predicate. One must not confuse the articular participle in the predicate like su. ei= o` evrco,menoj (Lu. 7:19) with the real predicate participle. Cf. Lu. 16:15; 22:28.353 The predicate participle is simply the adjective in the predicate position. That is, it is not attributive. There are obviously many varieties of the predicate participle. But the predicate adjective has had adequate treatment. Cf. e;ce me parh|Ä thme,non (Lu. 14:18). Cf. also Heb. 5:14; Ac. 9:21.

(d) The Participle as a Substantive. The adjective, though a variation from the substantive, is sometimes used as a substantive

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1109

as in to. avgaqo,n. It is not strange, therefore hat the participle also shows substantival uses. These are sometimes anarthrous, as in a;rcwn, (Mt. 9:18), h`gou,menoj (Mt. 2:6). But, as a rule, the participle as a substantive is articular. Cf. Lu. 12:33, ta. u`pa,rconta u`mw/n, where the genitive shows the substantival character of this participle. Cf. further 2:27 to. eivqisme,non tou/ no,mou, (1 Cor. 7:35) pro.j to. u`mw/n auvtw/n sumfe,ron, (Ph. 3:8) dia. to. u`pere,con th/j gnw,sewj, (Mt. 14:20) to. perisseu/on tw/n klasma,twn (Ro. 7:23) tw|/ o;nti (Heb. 12:11) pro.j to. paro,n, etc. There are also the many examples where o` and the part. is used without a subst. or pron., as in Mt. 10:39, o` eu`rw,n and o` avpole,saj (cf. o` avgaqo,jà o` kaÄ ko,j). The substantive use of the participle is a classic idiom.354 The use of the neuter participle as an abstract substantive is not so common in the N. T. as in the ancient Greek.355 But see further to. gegono,j (Lu. 8:56), ta. gino,mena grk(9:7), to. avpolwlo,jgrk grk(19:10), ta. evrco,mena (Jo. 16:13), to. nu/n e;con (Ac. 24:25), ta. mh. o;ntaà ta. o;nta (1 Cor. 1:28), to. auvlou,menongrk grk(14:7), to. dedoxasme,non (2 Cor. 3:10 f.), to. dokou/n (Heb. 12:10), etc. In Lu. 22:49 note to. evso,menon. One is not to confuse with this idiom the so-called "substantive participle" of some grammars, which is a term used for the substantivizing of the verbal force of the participle, not the adjectival. Thus Burton356 calls the supplementary participle like that in Ac. 5:42, ouvk evpau,onto dida,skontej, and in Lu. 8:46, e;gnwn du,namin evxelhluqui /an avp v evmou/, the "substantive participle." I confess that I see nothing to be gained by applying "substantive" to the purely verbal aspects of the participle. Confusion of thought is the inevitable result. See 5, (d), ( d).

(e) The Participle as an Adverb. The formation of adverbs from participles is due to its adjectival function. Cf. o;ntwj (Mk. 11:32), o`mologoume,nwj (1 Tim. 3:16), u`perballo,ntwj (2 Cor. 11: 23). Besides, the participle itself (cf. neuter adjective polu,, etc.) sometimes has an adverbial force. In particular note tuco,n (1 Cor. 16:6). See also evpibalw.n e;klaien (Mk. 14:72). This obscure participle expresses coincident action (cf. Moulton, Prol., p. 131). Cf. h=lqan speu,santej (Lu. 2:16), speu,saj kata,bhqi and speu,saj kate,bhgrk grk(19:5 f.). We cannot always draw a distinction between this use and the circumstantial participle of manner. The verbal and the adjectival standpoints come together. A number of the grammars apply the term "adverbial" to all the circumstantial participles.357 But it is more than doubtful if

1110 A GRAMMAR OF THE GREEK NEW TESTAMENT

one gains as much as he loses thereby. It is true that logically a sort of adverbial relation may be worked out, an adverbial addition to the sentence.358 But it does not help much from the syntactical point of view to insist on this fact in the exposition of the circumstantial participle. As to form the circumstantial participle is still adjectival. The adverbial notion is inferential and purely logical. There is something, however, to be said for the adverbial aspect of the redundant participle in ble,pontej ble,pete (Mt. 13:14, LXX), which is on a par with avkoh|/ avkou,stete. are attempts to translate the Hebrew inf. absolute. Moulton359 has found the idiom in AEschyluls and Herodotus, but the N. T. usage is clearly due to the LXX, where it is very common. Cf. also ivdw.n ei=don (Ac. 7:34), euvlogw/n euvlogh,sw (Heb. 6:14), from the LXX again. Blass (Gr. of the N. T. Gk., p. 251) calls this construction "thoroughly un-Greek." There are other pleonastic participles like the common avpokriqei.j ei=pen (Mt. 3:15) which is somewhat like the vernacular "He ups and says" (Moulton Prol., p. 15 f.). Cf. also tou/to eivpw.n le,gei (Jo. 21:19), avpelqw.n pe,praken (Mt. 13:46), 'he has gone and sold.' So also avnasta.j h=lqen (Lu. 15:20), 'he arose and came.' Once again note labou/sa evne,kruyen (Mt. 13:33), 'she took and hid.' This idiom is more Aramaic than Hebraic and is at any rate picturesque vernacular. But it is also Greek. Pleonasm belongs to all tongues. Radermacher (N. T. Gr., p. 179) quotes Herod. VI, 67, 10, ei=pe fa,j; VI, 68, 5, e;fhÄÄle,gwn. Mr. Dan Crawford finds in the Bantu language "dying he died" for the, irrevocableness of death. We now turn to the verbal aspects of the participle, which are more complex.

5. VERBAL ASPECTS OF THE PARTICIPLE.

(a) Voice. There is nothing of al distinctive nature to say about the voice of the participle in addtion to what has already been said (see ch. on Voice). The voices run in the participles precisely as in the verb itself. We find the voice in the earliest Greek as in the Sanskrit. All the nuances of the voices appear in the participle. Cf. the active in dida,skwn (Lu. 13:10), zw/n, (Jo. 4:10); the middle in prosdecome,noij (Lu. 2:36), evpikalesa,menoj (Ac. 22: 16), spasa,menoj (Mk. 14:47); the passive in lupou,menoj (Mt. 19: 22), th.n avpokekrumme,nhn (1 Cor. 2:7), avpolelume,non (Heb. 13:23), evpistrafei,j (Mk. 5:30), kwluqe,ntej (Ac. 16:6). We may note in particular e;ce me parh|thme,non (Lu. 14:18 f.), e;sesqe misou,menoi (Mt. 10:22) and e;sesqe lalou/ntej (1 Cor. 14:9). In Mk. 5:26,

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1111

paqou/sa u`po. pollw/n ivatrw/n, the active participle has the construction of the passive, but this is due to the verb pa,scw, not to the voice. Cf. also Gal. 4:9, gno,ntej qeo.n ma/llon de. gnwsqe,ntej u`po. qeou/)

(b) Tense.

(a) Timelessness of the Participle. It may be said at once that the participle has tense in the same sense that "the subjunctive, optative and imperative have, giving the state of the action as punctiliar, linear, completed. In the beginning360 this was all that tense meant in the participle. The participle was timeless. Indeed the participle in itself continued timeless, as is well shown by the articular participle.361 Thus in Mk. 6:14, vIwa,nhj o` bapti,zwn, it is not present time that is here given by this tense, but the general description of John as the Baptizer without "regard to time. It is actually used of him after his death. Cf. oi` zhtou/ntej (Mt. 2:20). In Mt. 10:39, o` eu`rw.n avpole,sei, the principal verb is future while the participle is aorist, but the aorist tense does not mean past or future time. So in Mt. 25:20 and 24 o` labw,n and o` eivlhÄ fw,j have no notion of time but only the state of the action. But the tenses of the participle may be used for relative time. In relation to the principal verb there may be suggested time. Thus o` eu`rw.n avpole,sei above implies that eu`rw,n is antecedent to avpole,sei which is future. In Ac. 24:11, avne,bhn proskunh,swn the principal verb is past, but the participle is relatively future, though absolutely past. The relative time of the participle approximates the indicative mode and is able to suggest antecedent (aorist, present, perfect tenses), simultaneous (aorist, present tenses) and subsequent (present, future tenses) action. The tenses of the participle must be studied with this distinction in mind. But this notion of relative time "is deeply imbedded in the nature of the participle and the use is universal."362 Certainly this notion of relative time is more obvious in the Greek participle than in the Latin or in the modern languages.363 In the chapter on Tense the participial tenses were treated with reasonable clompleteness,but some further remarks are necessary at this point. A word needs to be said about the idiom ou-toj h=n o` eivpw,n $Jo. 1:15), ou-toj h=n o` ÄÄ kaqh,menoj (Ac. 3:10), where the principal verb is thrown into the past.

1112 A GRAMMAR OF THE GREEK NEW TESTAMENT

( b) The Aorist. The Aktionsart of the aorist participle is sufficiently illustrated in the discussion of the aorist tense. There is, of course, no reason for not having the constative, ingressive or effective aorist in the participle.364 Schaefer365 argues that in most cases the participle uses the effective aorist. That may be true, though there is nothing in the nature of the participle itself to cause it. Blass366 thinks that the aorist participle contains the idea of completion, but even so that motion may be merely constative or ingressive. Goodwin367 holds that the aorist participle generally represents the action as antecedent to the principal verb. Burton368 has it more nearly correct when he insists that the aorist participle conceives of the event indefinitely or simply. So Blass369 denies that the aorist tense implies antecedent action. It is usually assumed that the proper use of the aorist participle is antecedent action and that only certain verbs (as exceptions) may occasionally express simultaneous action. But this is a misapprehension of the real situation. It is doubtless true, as Burton370 notes, that the antecedent use furnishes the largest number of instances, but that fact does not prove priority or originality of conception. "The aorist participle of antecedent action does not denote antecedence; it is used of antecedent action, where antecedence is implied, not by the aorist tense as a tense, but in some other way."371 Moulton372 is equally explicit: "The connotation of past time was largely fastened on this participle, through the idiomatic use in which it stands before an aorist indicative to qualify its action. As point action is always completed action, except in the ingressive, the participle naturally came to involve past time relative to that of the main verb." It is probable that the original use of the aorist participle was that of simultaneous action. From this was developed quite naturally, by the nature of the various cases, the antecedent notion. Cf. nhsteu,saj evpei,nasen (Mt. 4:2) where the fasting expressed by the participle is given as the reason for the hungering expressed by the principal verb. For further examples of anteceden action see Mt. 2:14; 2:16; 27:3; 2 Cor. 2:13. For the articular aorist see Mt. 10:39; Lu. 12:47; Jo. 5:15. While this came to be the more common idiom

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1113

from the nature of the case, the original use of the aorist participle for simultaneous action continued. One has no ground for assuming that antecedent action is a necessary or an actual fact with the aorist participle.373 The aorist participle of simultaneous, action is in perfect accord with the genius and history of the Greek participle. For numerous examples of both uses see the chapter on Tense. A good instance is seen in Mt. 27:4, h[marÄ ton paradou.j ai-ma avqw|/on. So also u`polabw.n ei=pen (Lu. 10:30). See Ac. 2:23, tou/ton prosph,xantej avnei,late, where the slaying was manifestly done by the impaling on the cross. The two actions are identical per se. Moulton (Prol., p. 131) observes that when the verb precedes the aorist participle it is nearly always the participle of coincident action. He (Prol., p. 132) cites 0. P. 530 (ii/A.D.), evx w=n dw,seij- lutrw,sasa, mou ta. i`ma,tia. It so happens that the N. T. shows a great number of such examples. See Mk. 15:30 sw/son kataba,j, (Lu. 2:16) h=lqan spei,santej, (Ac. 10:33) kalw/j evpoi,hsaj parageno,menoj. Cf. Mt. 26:75. In Ac. 10:29, h=lqon metapemfqei,j, the participle is antecedent in idea. Acts, however, is particularly rich in examples of the coincident aorist participle which follows the verb. See 10:39; 11:30; 13:33; 15:8, 9; 19:2; 23:22, 25, 30; 25:13; 26:10. It is in point of fact a characteristic of Luke's style to use frequently the coincident participle (both aorist and present) placed rater the principal verb. This fact completely takes away the point of Sir W. M. Ramsay's argument374 for the aorist of subsequent action in Ac. 16:6, where, however, it is more probably antecedent action, as is possible in Ac. 23:22. The argument made against it under Tense need not be repeated here.375 Burton assents376 to the notion of the aorist of "subsequent" action in the participle, but no real parallels are given. I have examined in detail the N. T. examples adduced and shown the lack of conclusiveness about them all. See chapter on Tense. It is even claimed that subsequent action is shown by the participles (present as well as aorist) in Ac. 5:36; 6:11; 8:10, 18; 14:22; 17:26; 18:23; 28:14, but with no more evidence of reality. Actual examination of each passage shows the action to be either simultaneous or antecedent. See also Lu. 1:9, e;lace tou/ qumia/sai eivselqw.n eivj to.n nao,n, where it is obviously coincident. The same thing is true of Heb. 11:27, kate,lipen Ai;guptonà mh. fobhqei,j. Cf. also Ac. 7:35 oa}n hvrnh,santo

1114 A GRAMMAR OF THE GREEK NEW TESTAMENT

eivpo,ntej,grk grk(13:22) ei=pen marturh,saj. A case like 1 Pet. 1:20 f. is not, of course, pertinent. However, the common use of the aorist participle in indirect discourse (as with all the supplementary participles) without any notion of time is to the point. So Ac. 9:12, ei=den a;ndra eivselqo,nta kai. evpiqe,nta. So evqew,roun to.n Satana/n peso,nta (Lu. 10:18). The action is purely punctiliar with no notion of time at all. It is true that the articular participle is occasionally used (see chapter on Tense) for time past to the time of the writer, but future to the time of the principal verb. As a matter of fact this aorist participle is timeless, as is shown by the use of o` parabou,j in Mt. 10:4 and o` paradidou,j in 26:25. So o` eipw,n in Jo. 5:12; o` poih,saj 5:15; h` avlei,yasa 11:2. It is the action alone that is under consideration, not the time of its performance. See, per contra, o` gnou,j- kai. mh. e`toima,saj h' pooih,Ä saj darh,setai (Lu. 12:47) where the aorist participle gives the simple action with a future verb. Cf. Lu. 6:49 for the articular aorist part. with the present indicative. Burton377 feels the weakness of his contention for "subsequent" action in the aorist participle when he explains that it is "perhaps due to Aramaic influence." There is no need for an appeal to that explanation, since the fact does not exist. It is only in the circumstantial participle that any contention is made for this notion. It is certainly gratuitous to find subsequent action in Ro. 4:19, mh. avsqenh,saj th|/ pi,stei kateno,hsen, not to mention 4:21; Ph. 2:7; Heb. 9:12. Burton reluctantly admits that, though in 1 Pet. 3:18 zwopoihÄ qei,j is "clearly subsequent to avpe,qanen," yet it "is probably to be taken together with qanatwqei,j as defining the whole of the preceding clause." This latter view is, of course, true, since the order of the participles is qanatwqei,j zwopoihqei,j. The timelessness of the aorist participle is well shown in Jo. 16:2, o` avpoketi,naj [ u`ma/j] do,xh| latrei,an prosfe,rein tw|/ qew|/) Cf. also avgago,nta- teleiw/sai. (Heb. 2:10). This coincident use of the aorist participle is by no means so rare in the ancient Greek as is sometimes alleged.378 The action was specially likely to be coincident if the principal verb was also aorist.379 Like the other articular participles, the aorist participle may be the practical equivalent of the relative. So in Lu. 12:8 f. oa}j a'n o`mologh,sei and o` avrnhsa,menoj are used side by side.

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1115

<tab>$g% The Present. As the aorist participle timeless and punctiliar, so the present participle is timeless and durative. The participle is thus, like the infinitive, ahead of the present indicative, which does not distinguish between punctiliar and durative action. A careful treatment of the force of the present participle has been given under Tense. The real timelessness of this participle is shown in the fact that it is used indiscriminately with past, present or future tenses of the indicative. So pwlou/ntej e;feron (Ac. 4:34); avpoqnh,skwn euvlo,ghsen (Heb. 11:21); kai,per w'n ui`o.j e;maqen (Heb. 5:8); merimnw/n du,natai (Mt. 6:27); e;sesqe lalou/nÄ tej (1 Cor. 14:9). The articular present especially shows the absence of time. So oi` dokou/ntej ouvde.n prosane,qento (Gal. 2:6); proseti,qei tou.j swzome,nouj (Ac. 2:47); o` deco,menoj u`ma/j evme. de,cetai (Mt. 10:40); evsqi,ete ta. paratiqe,mena (Lu. 10 :8) ; o` ble,pwn evn tw|/ krufai,w| avpodw,sei (Mt. 6:18). There will be Aktionsart in this participle also. Some of these words are really punctiliar ( de,coÄ mai, for instance). But, in general, the present participle gives linear action. The present participle may have relative time. This relative time is usually simultaneous or coincident. This is only natural. Sometimes, however, this relative time may be antecedent action, a classic idiom.380 Example of this idiom were given under Tense, but add Jo. 9:8, oi` qewrou/ntej to. pro,teron where the adverb of time helps to throw the participle back of e;legonà as a;rti with ble,pw makes the verb later than tuflo.j w;n in 9: 25. Cf. also Gal. l:23, o` diw,kwn h`ma/j pote. nu/n euvaggeli,zetai, where both participle and verb have adverbs of time by way of contrast. For other instances like these see Mt. 9:20 Mk. 5:25 Lu. 8: 43; Jo. 5:5; Ac. 24:10; Eph. 2:13; Col. 1:21; 1 Tim. 1:13, etc. There are also undoubted instances of the present participle to express the notion of purpose, futuristic in conception, though present in form. Add to the instances already given the following: Mk. 3:31, e;xw sth,kointej avpe,steilan kalou/ntej. Here the first participle is only noticeable as the usual linear action (with aorist indicative). The second participle, however, is practically purpose. 'They sent to him calling him.' 'They sent to call him.' So also Lu. 13:6 h=lqen zhtw/n,grk grk(13:7) e;rcomai zhtw/n. It is not strictly true that here the present participle means future or subsequent time. It is only that the purpose goes on coincident with the verb and beyond. This prospective present part. (cf. present ind.) appears in Ac. 21:3, h=n avpofortizo,menon to.n go,mon. 'The ship was appointed to unload her cargo.' Cf. Mt. 6 : 30;

1116 A GRAMMAR OF THE CREEK NEW TESTAMENT

11:3; 26:28; Lu. 7:19; 1 Chr. 15:57; Jas. 5:1; Ac. 3:26. The future is "simulated"381 also by the present participle when it is used for conative action. It is, of course, not the participle that brings out this notion. See (Mt. 23:14) ouvde. tou.j eivsercome,Ä nouj avfi,ete eivselqei/ngrk grk(27:40) o` katalu,wn to.n nao,n (Ac. 28:23) pei,Ä qwn auvtou,j. The notion of repetition (iterative present) occurs also as in Ac. 2:47, proseti,qei tou.j swzome,nouj, 'kept adding those saved from time to time.' So pwlou/ntej e;feron kai. evti,qooun (Ac. 4: 34). 'They would from time to time sell and bring and place at the feet of the apostles.' There is thus a sharp contrast from the specific instance of Barnabas, of whom it is said: pwlh,saj h;negken grk(4:37). It is not clear, (however, why the present participle occurs in 3:8, evxallo,menoj e;sth kai. periepa,tei, unless it is to note that he kept on leaping and walking (alternately). Cf. this notion in verse 8, peripatw/n kai. avllo,menoj. Cf. also in 5:5, avkou,wn pesw.n evxe,yuxen, where pesw,n is antecedent to the verb, but avkou,wn is descriptive (linear). The notion of distribution is perhaps present in Heb. 10:14, tou.j a`giazome,nouj, 'the objects of sanctification.'382 Certainly o` kle,ptwn is iterative in Eph. 4:28. Cf. Ac. 1:20; Col. 2:8. It is interesting to note the difference between the present and the aorist participle in Mt. 16:28, e[wj a'n i;dwsin to.n ui`o.n tou/ avnqrw,pou evrco,menon, and in Ac. 9:12, ei=den a;ndra eivselqo,nta. The perfect participle of the same verb and in the same construction occurs in Mk. 9:1, e[wj a'n i;dwsin th.n basilei,an tou/ qeou/ evlhluqui/an evn duna,mei. The three tenses of the participle of pi,ptw may also be illustrated by the punctiliar notion of the aorist in peso,nta in Lu. 10:18, the durative notion pipto,ntwn in Mt. 15:27 and of pi,ptontej in Mk. 13:25, the perfect notion of peptwko,ta in Rev. 9:1.

(5) The Perfect. This tense brings little that is distinctive in the participle. Cf. teteleiwme,noi $Jo. 17:23), pepoihko,tej (18: 18), prosfa,twj evlhluqo,ta (Ac. 18:2), kekopiakw,j (Jo. 4:6), peptwko,ta (Rev. 9:1), evlhluqo,ta (1 Jo. 4:2), o` eivlhfw,j (Mt. 25:24). The distinction between intensive and extensive was drawn under Tense. Some of the intensive uses have lost the notion of completion (punctiliar) and hold on to the linear alone in the present sense. Cf. e`stw,j eivmi (Ac. 25:10), eivdw,j (Mt. 12:25) with which contrast oi` evgnwko,tej (2 Jo. 1:1), suneidui,hj (Ac. 5:2), teqnhkw,j (Lu. 7:12), paresthkw,j (Jo. 18:22). The periphrastic use of the perfect participle in past, present and future time has been sufficiently illustrated already. So has the rare com-

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1117

bination of perfect and present participle in Eph. 4 : 18; Col. 1:21. The perfect participle also is either articular or anar-throus, attributive or predicate. For the predicate use see in particular Lu. 13:6 sukh/n ei=ce,n tij pefuteume,nhn, (Heb. 5:14) ta. aivsqhth,ria gegumnasme,na evco,ntwn. It needs to be noted again that the perfect participle has no time in itself. In th nature of the case the act will be antecedent except where the tense has lost its true force as in evstw,jà teqnhkw,jà eivdw,j. But it is only relative time, not absolute, and the leading verb may itself be punctiliar, linear or perfect, in the past, present or future.383 Just as the present participle may suggest antecedent action and so be a sort of "imperfect" participle (past time), so the perfect participle is sometimes384 used where a sort of past perfect sense results. The action was finished and is now no longer the fact, though the state represented by the perfect once existed. So evpi. tw|/ sumbebhÄ ko,ti auvtw|/ in Ac. 3:10. Cf. Mk. 5:15, qewrou/sin to.n daimonizo,menon kaqh,menon i`matisme,non ka. swfronou/ntaà to.n evschko,ta to.n legiw/naà kai. evfobh,qhsan. This is a most instructive passage. The historical present and the aorist indicative here occur side by side. The attributive and the predicate participles appear side by side. The present and the perfect participles come together. Of the two perfect participles, one, i`matisme,non, is still true (punctiliar plus linear) and describes the man's present state; the other, to.n evschko,ta, is no longer true and describes the state of the man before Jesus cast out the demon, which casting-out is itself in the past. This participle is therefore a sort of past perfect. Cf. also Jo. 8:31. Another striking example is Jo. 11:44, evxh/lqen o` teqnhkw.j dedeÄ me,noj. Here dedeme,noj is still true, though teqnhkw,j is not. Lazarus had been dead, but is not now. We see the same situation in 1 Cor. 2:7, th.n avpokekrumme,nhn. The wisdom of God is no longer hidden. The point is still clearer in Ro. 16:25 musthri,ou cro,Ä noij aivwni,oij sesighme,nou fanerwqe,ntoj de. nu/n, where the long silence is now expressly said to be broken. Note the sharp contrast in the aorist participle with nu/n. This distinction between the perfect and aorist participle is often clearly drawn. See 2 Cor. 12:21 tw/n prohmarthko,twn kai. mh. metanohsa,ntwn, (1 Pet. 2:10); oi` ouvk hvlehÄ me,noi nu/n de. evlehqe,ntej. The same act may be looked at from either standpoint. One may not always care to add the linear aspect to the punctiliar. Cf. o` gegenhme,noj and o` gennhqei,j in 1 Jo. 5:18, to.n evschko,ta to.n legiw/na in Mk. 5:15 and o` daimonisqei, in 5:18,

1118 A GRAMMAR OF THE Gitii,EK NEW TESTAMENT

o` labw,n in Mt. 25:18 and o` eivlhfw,j in 25:24. Cf. e;gnwn du,namin evxelhluqui/an avp v evmou/ (Lu. 8:46) aud evpignou.j th.n evx auvtou/ du,namin evxelqou/san (Mk. 5:30). Adverbs lof time may occur with the perfect as with other tenses of the participle. Cf. Jo. 19:33, h;dh teqnhko,ta. There is a sort of harmony in o` e`wrakw.j memartu,rhkengrk grk(19:35). The difference between the perfect and present tenses after ei=don is strikingly shown in Revelation. Cf. ei=don ta.j yuca.j tw/n evsfagme,nwngrk grk(6:9), a;llon a;ggelon avnabai,nontagrk grk(7:2), avste,ra evk tou/ ouvranou/ peptwko,ta grk(9:1). Cf. also Mk. 5:33, fobhqei/sa kai. tre,mousaà eivdui/a. One must not confuse the perf. part. in Gal. 2:11 and Rev. 21:8 with a present like yhlafwme,nw| in Heb. 12:18 ('touchable').

( e) The Future. The future participle, like the future tense in general, was later in its development than the other tenses. It is usually punctiliar also and has something of a modal value (volitive, futuristic) like the subjunctive (aorist).385 See discussion under Tense. The future participle is always subsequent in time to the principal verb (cf. the present participle by suggestion), not coincident and, of course, never antecedent. Hence the future participle comes nearer having a temporal notion than any of the tenses. But even so it is relative time, not absolute, and the future participle may occur with a principal verb in the past, present or future. This idiom grew out of the context and the voluntative notion of the future tense.386 This point is well illustrated by the parallel use of me,llwn to express intention. Cf. o` paradw,swn auvto,n (Jo. 6:64) and o` me,llwn auvto.n paradido,naigrk grk(12:4). As already shown, the future participle is much less frequent in the N. T. (as in LXX) than in the koinh, generally (as in the papyri). Another rival to the future participle is evrco,menoj (Jo. 1:9), o` evrco,menoj (Lu. 7:19). Both me,llw and evrcomai (cf. ei=mi% are anticipatory presents.387 Cf. evnestw/ta and me,llonta in Ro. 8:38. Nearly all the N. T. examples of the future participle (see chapter on Tense for discussion) are in Luke and Paul and Hebrews (the three best specimens of literary style in the N. T.). But see Mt. 27:49, sw,swn; Jo. 6:64, o` paradw,swn; 1 Pet. 3: 13, o` kakw,swn. For the Gospel of Luke see 22:49, to. evso,menon. The rest of his examples are in the Acts, as 8:27, proskuÄ nh,swn,grk grk(20:22) ta. sunanth,sontagrk grk(22:5) a;xwn,grk grk(24:11) proskunh,Ä swn,grk grk(24:17) poih,swn. For Paul see Ro. 8:34, o` katakrinw/n (a

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1119

question of editing, but cf. o` avpoqanw,n, 1 Cor. 15:37, to. genhso,menon. For Heb. see 3:5, tw/n lalhqhsome,nwn,grk grk(13:17) avpodw,sontej. We find w`j in Heb. 13:17. In conclusion one must note that the future part. disappeared wholly from the later Greek. The modern Greek does not know it at all. Instead it uses na, and the subjunctive.388 But in general in the N. T. the participle is still used in thorough accord with the ancient idiom so far as the tenses are concerned.389 In the papyri I note it more frequently than in the N. T. Cf. koinologhso,menon, P. Goodsp. 4 (ii/B.C.); ta. - [ s] taqhso,mena, P. Tb. 33 (B.C. 112).

(c) Cases. There is no need to tarry here tp prove the verbal force of the participle as to cases. Precisely the same cases occur with the participle as with the finite modes of the verb. Cf. evkbalw.n pa,ntaj (Mk. 5:40) and krath,saj th/j ceiro.j tou/ paidi,ou (5: 41). These illustrations illustrate the point and that is enough.

(d) The Supplementary Participle. The term supplementary or complementary is used to describe the participle that forms so close a connection with the principal verb that the idea of the speaker is incomplete without it. The participle does not differ in reality from the adjective in this respect, and it is still an adjective like pisto.j me,nei (2 Tim. 2:13). But it is the verbal aspect of the participle that is here accented. The participle fills out the verbal notion.

(a) The Periphrastic Construction. The general aspects of this idiom were treated in chapter on Tense (cf. also Conjugation of Verbs). It is only necessary here to stress the close connection between this participle and the principal verb as in h=n evkba,llwn daimo,nion kwfo,n (Lu. 11:14). In Ac. 19:36, de,on evsti.n u`ma/j katestalÄ me,nouj u`pa,rcein, we have two examples of this idiom. Cf. Lu. 13:11. Sometimes we find the periphrastic participle alone without the copula as in evxo,n (Ac. 2:29), eiv de,on (1 Pet. 1:6). But note evxo.n h=n (Mt. 12:4) and de,on evsti,n (Ac. 19:36). So pre,pon evsti,n (Mt. 3:15). Particularly interesting is eivsin gegono,tej (Heb. 7:23). The periphrastic participle, as already noted, was far more common in the N. T. and the LXX than in the older Greek. But the reverse is true of certain verbs frequently so used in the Attic. Radermacher390 thinks that the commonness of the periphrastic participle in the N. T. is due to the rhetorical tendency.

1120 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

Addenda 3rd ed.

This might apply to Hebrews, but surely not to the Synoptic Gospels and Acts. Moulton (Prol., p. 226) admits that the Semitic sources of part of the Gospels and Acts account for the frequency of the periphrastic imperf. (cf. Aramaic). Certainly the LXX is far ahead of the classic Greek and of the koinh, in general. The papyri (Moulton, Prol., p. 226) show it often in fut. perfects and in past perfects. Schmid (Attic., p. 113 f.) finds it rare in literary koinh, save in fut. perfects. Moulton finds periphr. imperf. in Matthew 3 times, Mark 16, , Luke 30, , John 10, , Acts (1-12) 17, Acts (13-28) 7, Paul 3. And even sb some of these examples are more adjectival than periphrastic. Cf. Ph. 2:26. See p. 888.

( b) A Diminution of the Complementary Participle. This decrease is due partly to the infinitive as with a;rcomaià doke,w. See discussion in this chapter on Relation between the Inf. and the Participle. But it is due also to the disappearance of the personal construction and the growth of the impersonal with o[ti or i[na. In Mk. 2:1, eivselqw.n pa,lin eivj Kafarnaou.m di v h`merw/n hvkou,sqh o[ti evn oi;kw| evsti,nà the personal construction is retained even with the circumstantial participle. Cf. also 2 Cor. 3:3, fanerou,menoi o[ti evste. evpistolh. Cristou/. But it is vanishing with the verbs where it was once so common. See under Infinitive, 5, (e), for further remarks. Jannarisl has made a careful study of the facts in the later Greek. It may be noted that oi;comai does not occur at all in the N. T., though the LXX (and Apocrypha) has it 24 times, twice with the inf. it disappeared from the vernacular. As to tugca,nw it occurred only once with the participle (2 Macc. 3:9). It has the inf. as well as i[na ( na) in the later Greek, though it is very abundant with the participle in the papyri.391 Cf. t[ ug] ca,nei Nei/loj r`e,wn, P. B. M. 84 (ii/A.D.). But tugca,nw fi,loj without w;n occurs also in the koinh, (Radermacher, N. T. Gr., p. 169). Curiously enough lanqa,nw appears once with the participle in the LXX (Tob. 12:13) as in the N. T. (Heb. 13:2). In the koinh, the inf. supplants the part. as it had already gained a foothold in the old Greek.392 Note also the adverb as in la,qra| evkba,llousin (Ac. 16:37). Fqa,nw continued in use through the koinh, but with the sense of 'arrive,' 'reach,' not the idiomatic one 'arrive before.' This latter notion appears in profqa,nw (cf. prolamba,nw), which has it once only in the N. T. (Mt. 17:25), while the inf. is seen in proe,laben muri,sai (Mk. 14:8). As early as Thucydides the inf. is found with fqa,nw, and see also 1 Ki. 12:18. It is, common in the koinh,.393 The

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1121

tendency to reverse the construction by using one of these verbs in the participial form is seen in tuco,n (participial adverb) in 1 Cor. 16:6. It is possible that fai,nomai still shows the participial construction in Mt. 6:16, 18, but not in Ro. 7:13, where the participle is circumstantial, not complementary. The impersonal construction gains394 on the personal in the koinh,. In the N. T. we no longer have dh/loj eivmi, nor fanero,j eivmi,. But we do have eu`re,qh e;cousa in Mt. 1:18. ;Arcomai has lost the part. in the N. T., but u`pa,rcw holds on to it, but not in the sense of 'begin,' rather of 'existing.' Cf. both adjective and part. in Jas. 2:15 and 1 Tim. 4:3. It tends to sink into the level of eivmi, as an auxiliary verb with the periphrastic participle, as in Ac. 8:16; 19:36. The same thing is true of prou?pa,rcw in Lu. 23:12, but not in Ac. 8:9 where mageu,wn is circumstantial. We have seen that pau,omai is true to the part. (cf. Lu. 5:4; Ac. 5:42, etc.) and that the part. occurs also with evpime,nw (Jo. 8 :7), tele,w (Mt. 11:1), and that diatele,w has the adj. without w;n (Ac. 27:33). Cf. also dialei,pw in Lu. 7 :45. See also the part. with evgkake,w in Gal. 6:9; 2 Th. 3:13. The part. with kartere,w in Heb. 11:27 is circumstantial, as is that with avne,comai in 1 Cor. 4:12 and with ka,mnw in Heb. 12:3. The doubtful participle with manqa,nw in 1 Tim. 5:13 has already been discussed (Relation between Inf. and Part., 3, (d)). Moulton395 is positive that the absolute construction advocated by Weiss is intolerable and that we must either admit the supplementary participle here or boldly insert ei=nai with Blass. Moulton396 is probably right in opposing the incorrectness of the part. with eu= pra,ssw in Ac. 15:29, evx w-n diathrou/ntej e`autou.j eu= pra,xete. At bottom this is the same idiom as we have in 10:33, kalw/j evpoi,hÄ saj parageno,menoj. Cf. also Ph. 4:14; 2 Pet. 1:19; 3 Jo. 1:6. Blass397 is right in including here ti, poiei/te lu,ontej (Mk. 11:5), ti, poiei/te klai,ontej (Ac. 21:13), h[marton paradou,j (Mt. 27 :4).

( g) Verbs of Emotion. As a matter of fact it is not beyond controversy that the part. with these verbs of emotion is the supplementary and not the circumstantial participle. At any rate the idiom comes to the border-line between the two constructions. I do not wish to labour the point and so treat the construction as complementary. The connection is not, however, so close with these verbs as is true of those in the two preceding lists. Indeed, the connection varies with different verbs and with the same verb in different contexts. It seems clear enough in

1122 A GRAMMAR OF THE GREEK NEW TESTAMENT

Ac. 16:34, hvgallia,sato pepisteukw,j, and in 2 Pet. 2:10, ouv tre,mouÄ sin blasfhmou/ntej. The examples with avganakte,w (Mt. 21:15, etc.) and cai,rw (Mt. 2:10, etc.) all seem to be circumstantial.398 The same thing is true of lupe,w. The participle does not occur in the N. T. with aivscu,nomai. The step over to the circumstantial participle of manner or cause is not very far to take.399

( d) Indirect Discourse. This participle is clearly supplementary and in the N. T. is usually connected with the object of the principal verb. The nom.400 of the part. e;cousa appears with the passive eu`re,qh in Mt. 1:18 as noted above. The active in the N. T. would have had o[ti and the ind., if the reference was to Mary. The classic Greek could have said eu-ren e;cousaà but the N. T. Greek, eu-ren o[ti e;cei. Cf. also eu`reqei.j w`j a;nqrwpoj in Ph. 2:8. But 1 Tim. 5:13 has to be noted. This subject was treated in detail under Indirect Discourse (see Modes). See that discussion for details about the different verbs, some of which, besides the participial construction, may instead use the inf. or o[ti and the indicative. Here it is sufficient to give enough illustrations of this participle in indirect discourse with verbs of mental action to show the real complementary nature of the participle. The tense, of course, represents the tense of the direct. With most of these verbs (especially401 oi=daà manqa,nwà o`mologe,w) the participle is giving way to the inf. or o[ti, but still the idiom is common enough to attract notice in all parts of the N. T. Cf. gei,nwske sauto.n e[xonta, P. B. M. 356 (i/A.D.). It is common to explain this participle as the object of the principal verb after the analogy of the inf. in indirect discourse. So Jannaris402 calls it "the objective participle" and Burton403 "the substantive participle as object." Blass404 more correctly perceives that it is the substantive or pronoun that is the object while the participle is a predicate adjective agreeing with this object. It is easy to see this point where no indirect discourse occurs, as in Heb. 7:24, avpara,baton e;cei th.n i`erwsu,nhn, where e;cw does not mean to 'opine' and where the verbal adj. occurs. But see the participle in 5:14, tw/n ta. aivsqhÄ th,ria gegumnasme,na evco,ntwn, or, still better, Lu. 14:18, e;ce me parh|Ä thme,nonà where e;cw means 'consider' and we have the participle.

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1123

Addenda 3rd ed.

Cf. Mk. 3:1; Ac. 9:21, i[na dedeme,nouj auvtou.j avga,gh|. See also 24:27. Then note Ph. 2:3, avllh,louj h`gou,menoi u`pere,contaj.405 The addition of w`j does not change the real construction as in tou.j logizome,nouj h`ma/j w`j kata. sa,rka peripatou/ntaj, 2 Cor. 10:2; w`j evcqro.n h`gei/sqe 2 Th. 3:15. In principle it is the double accusative, too common with some verbs, only the second ace. is a predicate adj., not a substantive. Cf. Ro. 10:9 (margin of W. H.), eva.n o`mologh,sh|j ku,rion vIhsou/n, and 2 Jo. 1:7, o`mologou/ntej vIhsou/n Cristo.n evrco,menon evn sarki,. The presence or absence of the copula does not materially change the construction when an adj. or substantive is the second ace. Thus note 2 Cor. 8:22, oa}n evdokima,samen spoudai/on o;nta, and Mk. 6:20, eivdw.j auvto.n a;ndra di,kaion. So we have no part. after ei=don in Jo. 1:50; Mt. 25:37, 38, 39; Ac. 8:23; 17:16. Blass406 calls this an "ellipse" of the participle, an idiom common in classical Greek. It is hardly necessary to appeal to the "ellipse" to explain it. The predicate force of o;nta, comes out well in Ac. 8:23. If no substantive or adjective is used, the participle is itself the full predicate and represents the predicate of the direct discourse. Cf. Mk. 12:28 avkou,saj auvtw/n sunzhtou,ntwn (Lu. 8:46) e;gnwn du,namin evxelhluqi/an avp v evmou/) The point to note is that even here in indirect discourse, where the participle represents the verb of the direct, the participle is still an adjective though the verbal force has become prominent. The examples are too numerous to discuss in detail or even to quote in full. As representative examples see Mt. 16:28 after ei=don $evrco,menon, but Mk. 9: 1 has evlhluqui/an), Mk. 5:30 after evpi,stamai, 7:30. after eu`ri,skw (cf. also Lu. 23:2), Lu. 10:18 after qewre,w (cf. in particular Ac. 7:56), Jo. 1:38 after evpi,stamai, 7:32 after avkou,w, Ac. 19:35 after ginw,skw, 24:10 after evpi,stamai, Heb. 2:9 after ble,pw, Heb. 13:23 after ginw,skw, 2 Cor. 8:22 after dokima,zw, Ph. 2:3 after h`ge,omai, 2 Jo. 1:7 after o`mologe,w. The punctiliar idea is present as in peÄ so,nta in Lu. 10:18, or the linear as in evggi,zousan (Heb. 10:25), or the perfected state as in peptwko,ta (Rev. 9:1). Cf. also Ac. 2:11; 24:18; Mk. 9:38; 1 Jo. 4:2. Burton407 explains as "the substantive participle" (see 4, (d)) also Jo. 4:39, th|/j gunaiko.j marÄ turou,shj and Heb. 8:9, evn h`me,ra| evpilabome,nou mou. The first example is really the attributive participle liken tou/ profh,tou le,gontoj (Mt. 21:4). The second example is more difficult, but it is a quotation from the LXX (Jer. 31:32) and is not therefore a model of Greek. The mou has to be taken with h`me,ra| and the

1124 A GRAMMAR OF THE GREEK NEW TESTAMENT

participle would be a circumstantial temporal use. It is probably suggested by the original Hebrew, as Moulton (Prol., p. 47) admits. Cf. Barn. 2:28, evn h`me,ra| evnteilame,nou sou auvtw|/. Cf. evpi. parou/sin u`mei/n, B. G. U. 287 (A.D. 250). The reference of Burton to Josephus, Ant. 10, 4. 2, does not justify the interpretation which he gives.

(e) The Circumstantial Participle or Participial Clauses.

(a) The General Theory. There is but one difference between the supplementary and the circumstantial participle. It lies in the fact that the circumstantial participle is an additional statement and does not form an essential part of the verbal notion of the principal verb. The circumstantial participle may be removed and the sentence will not bleed. It is still a true participle, predicate adjective as well as circumstantial addition to the verb. In point of agreement the circumstantial may be related to the subject of the principal verb or the object, or indeed any other substantive or pronoun in the sentence. It may have also an independent construction with a substantive or pronoun of its own (genitive or accusative absolute) or have no substantive or pronoun at all. Once again the participle may be so independent as to form a sentence of its own and not merely be a subordinate clause. See the section on The Independent Participle as a Sentence. Here we are dealing with the independent participle in a subordinate clause with various stages of independency from mere addition and agreement with a substantive or pronoun to complete isolation though still subordinate. Some of the grammars, Burton408 for instance, call this the "adverbial" participle. There is a slight element of truth here, but only so far as there is a sort of parallel with the subordinate conjunctional clauses which are adverbial (cf. o[teà i[naà w`j, etc.). But it is distinctly misleading to treat this participle as adverbial. In fact, there is a constant tendency to read into this circumstantial participle more than is there. In itself, it must be distinctly noted, the participle does not express time, manner, cause, purpose, condition or concession. These ideas are not409 in the participle, but are merely suggested by the context, if at all, or occasionally by a particle like a[maà euvqu,jà kai,perà pote,à nu/nà w`j. There is no necessity for one to use the circumstantial participle. If he wishes a more precise note of time, cause, condition, purpose, etc., the various subordinate clauses (and the infinitive) are at his command, besides the co-ordinate clauses. The vernacular increasingly

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1125

preferred the co-ordinate or the subordinate clause with conjunctions to the rather loose circumstantial participle.410 We see the triumph of this analytic tendency in the modern Greek.411 But it remains true that the participial clause was one of the great resources of the Greek language and in contrast the Latin seems very poor.412 The English comes next to the Greek in its rich use of the circumstantial participle. Moulton413 notes the failure of the English, even with the help of auxiliary verbs, to express the precise difference between lu,saj and lelukw,j ( o` labw,n and o` eivlhfw,j, for instance, in Mt. 25:18, 24). He rightly also calls attention to the weakness of the Greek because of its wealth of participles, since so much ambiguity is possible. Does a given circumstantial participle bear the notion of 'because' or 'although'? Only the context can tell, and men do not always interpret the context correctly. One more remark is necessary. By means of the circumstantial participle the sentence may be lengthened indefinitely. Good illustrations of this freedom may be seen in the periodic structure in Thucydides, Isocrates, Lysias and Demosthenes. But the N. T. itself has examples of it as is seen in 2 Pet. 2:12-15, blasfhmou/ntejà avdikou,menoià h`gou,menoià evntruÄ fw/ntej.

( b) Varieties of the Circumstantial Participle. Here are treated only those examples which have syntactical agreement in case with some substantive or pronoun in the sentence. It may be repeated that this participle does not express the ideas called by the usual classification into participles of time, manner (means), cause, purpose, condition, concession. Hence it is proper to group the examples together. The classification is only justified by the context and occasional use of a particle.414 The same classification is possible also for the absolute use of the participial clause. The examples are too numerous for exhaustive treatment. A few must suffice.

Time. It is not the tense that is here under discussion, though naturally the different tenses will vary in the way that time is treated (antecedent, simultaneous, future), as already shown. The point more exactly is whether a given circumstantial parti-

1126 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

ciple occurs in a context where the temporal relation is the main one rather than that of cause, condition, purpose, etc. It is usually a mistake to try to reproduce such participles by the English 'when,' 'after,' etc., with the indicative. To do this exaggerates the nuance of time as Moulton415 observes. It is generally sufficient to preserve the English participle or to co-ordinate the clauses with 'and.' The slightness of the temporal idea is well seen in the pleonastic participles avnasta,j (Mt. 26:62), avpokriqei,j (Mt. 3: 15, very common in the Synoptic Gospels. John usually has avpekri,qh kai. ei=pen as in 1:49), avpelqw,n (Mt. 13:46), labw,n (13: 31, cf. verse 33), poreuqe,ntejgrk grk(21:6). Here the notion is temporal, but very slightly so. Cf. also prosqei.j ei=pen in Lu. 19:11. The use of avrxa,menoj as a note of time is seen in Mt. 20:8 f.; Lu. 23: 5; 24:47; Ac. 1:22. In Ac. 11:4, avrxa,menoj Pe,troj evxeti,qeto auvtoi/j kaqexh/j, the part. is slightly pleonastic,416 but note contrast with kaqexh/j as with e[wj tw/n prw,twn in Mt. 20: 8. Cf. evrco,menoj[ j] e;rcouà P. Tb. 421 (iii/A.D.). Sometimes the temporal idea is much more prominent, as in diodeu,santej (Ac. 17:1), evlqw.n evkei/noj evle,gxei to.n ko,smon (Jo. 16:8). So also Mt. 6:17, su. de. nhsteu,wn a;leiyai. Here the descriptive force of the participle is distinctly temporal. In examples like Mk. 1:7 ku,yaj lu/sai to.n i`ma,nta, Ac. 21:32 paraÄ labw.n sratiw,taj kate,dramen evp v auvtou,j, there is precedence in order of time, but it is mere priority with no special accent on the temporal relation.417 Cf. Mt. 2:16; 13:2. In Ac. 24:25 f. we have some interesting examples of the participle. In dialegome,nou auvtou/ we see the temporal notion of 'while' with the genitive absolute. In tou/ me,llontoj the temporal notion in this attributive part. is due to me,llw. In geno,menoj it is mere antecedence with avpekri,qh (almost simultaneous, in fact). In to. nu/n e;con the attributive participle again has the temporal idea due to the words themselves. In metalabw,n we have antecedence emphasized by kairo,n. In a[ma kai. evlpi,zwn we have the linear notion stressed by a[ma. In pukno,teron auvto.n metapempo,menoj w`mi,lei auvtw|/ the note of repetition in pukno,teron reappears in participle and verb. An interesting example is also seen in Heb. 11:32, evpilei,yei me dihÄ gou,menon o` cro,noj, where in a poetic way time is described as going off and leaving the writer discoursing about Gideon and the rest. In 1 Pet. 5:10, ovli,gon paqo,ntaj, the adverb of time makes it clear. The note of time may appear in any tense of the participle and with any tense in the principal verb. It is not always easy to

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1127

discriminate between the temporal participle and that of attendant circumstance or manner. Moulton418 and Blass419 make no distinction. These two uses are the most frequent of all. A good example of this ambiguity occurs in Ac. 21:32, where paralabw,n (cf. labw,n in ancient Greek) may be regarded as merely the attendant circumstance. So also the notion of occasion wavers between time and cause. Cf. avkou,ontej (Lu. 4:28). For w`j with this participle see 1 Cor. 7:29

Manner. The ancient use of e;cwn in the sense of 'with' occurs in Mt. 15:30 e;contej meq v e`autw/n cwlou,j, Mk. 14:3 e;cousa avla,baÄ stron mu,rou, Ac. 21:23 euvch.n e;contej avf v e`autw/n. Cf. also fe,rwn in Jo. 19:39. In Jo. 18:3 we have labw,n used in practically the same sense as meta, in Mt. 26:47. Cf. also labw,n in Mt. 25:1. In Lu. 1:64, evla,lei euvlogw/n, the part. is one of manner, as in Mt. 19:22 avph/lqen lupou,menoj, (Mk. 1:22) w`j evxousi,an e;cwn, where w`j makes the point plainer, plainer,(1:4) khru,sswn, where the participle is not the periphrastic construction with evge,neto,grk grk(1:5) evxomologou,meÄ noi, (Ac. 3:5) evpei/cen auvtoi/j prosdokw/n ti, (a picturesque bit of description), (2 Th. 3:11) mhde.n evrgazome,nouj avlla. periergazome,nouj (a real pun). It is hard to tell how to classify a participle like that in Gal. 6:3, mhde.n w;n. It makes sense as temporal, causal or modal. But there is no doubt in a case like Lu. 19:48 . evxekre,meto auvtou/ avkou,wn or Ac. 2:13 diacleua,zontej e;legon or w`j ouvk ave,ra de,rwn (1 Cor. 9: 26). This notion of manner appears in the participles that have an adverbial notion like speu,saj (Lu. 19:5 f.),. evpibalw,n (Mk. 14:72), tuco,n (1 Cor. 16:6), ble,pontej (Mt. 13:14); prosqei.j ei=pen (Lu. 19:11). Cf. also avnable,yaj ei=pen in verse 5. So also the pleonastic participles like avpokriqei,j (see above) may be looked at either as temporal or modal or even adverbial. See further kriÄ ma,santej (Ac. 5:30), sumbiba,zwngrk grk(9:22) as good examples of the modal participle. Burton420 makes a separate division for the participle "of attendant circumstance," but this is not necessary and leads to overrefinement. These examples are either temporal as in evxelqo,ntej (Mk. 16:20), evklexame,nouj (Ac. 15: 22) or modal as doxazo,menoj (Lu. 4:15), avnalabw,n (2 Tim. 4:11) or pleonastic as avpekri,qhsan le,gousai (Mt. 25:9). Blass' term "conjunctive" (Gr. of N. T. Gk., p. 249) throws no particular light on the point. In 1 Tim. 1:13 avgnow/n is manner. In Ac.

1128 A GRAMMAR OF THE GREEK NEW TESTAMENT

18:18, keira,menoj, we have in truth both the temporal and the modal. But it is easy to split hairs over the various circumstantial participles and to read into them much more than is there. Cf. 2 Cor. 4:1 f. See bapti,zontej and dida,skontej in Mt. 28:19 f. as modal participles. So avgnow/n in 1 Tim. 1:13. Cf. kata. a;gnoian in Ac. 3:17.

Means. It is usual421 to distinguish means from manner in the participle. There is a real point, but it is not always clear where manner shades off into means. But some instances are clear. Cf. Mt. 6:27, ti, merimnw/n du,natai prosqei/nai; So also manteuome,nh in Ac. 16:16. Thus the maid furnished the revenue for her masters. In Heb. 2:10 avgago,nta and 2:18 peirasqei,j we may also have instances of this notion, but the first may be temporal and the second causal. Jannaris422 blends the treatment of manner and means and notes how this participle disappears in the later Greek.

Cause. The ground of action in the principal verb may be suggested by the participle. Cf. di,kaioj w'n kai. mh. qe,lwn auvth.n deigmati,Ä sai evboulh,qh, Mt. 1:19; h[marton paradou.j ai-ma, 27:4; evca,rhsan ivdo,ntejà Jo. 20: 20. As a matter of fact this idiom is very frequent. Cf. further Mt. 2:3, 10; Jo. 4:45; 21:12; Ac. 4:21; 9:26; 24:22, eivdw,j- ei;paj, Ro. 6:6, ginw,skontej, and eivdo,tej* 2 Pet. 3:9; Col. 1:3 f.; 1 Tim. 4:8; Jas. 2:25. For w`j with this participle see 1 Cor. 7:25, w`j hvlehme,noj. In Ac. 24:22 eivdw,j may be taken as 'wishing to know,' though Felix may also have actually had some knowledge of Christianity (cf. Paul's appearance before Gallio). So also eivdw,jgrk grk(24:22) may mean 'wishing to know.' The N. T. no longer has a;teà oi-onà oi-a with the part. as classic Greek did.423 In Jo. 5:44 a 'causal participle lamba,nontej is coordinate with zhtei/te)

Purpose. The use of the participle to express aim or design has already been discussed several times from different points of view (Tense, Final Clauses, Tense of the Participle). This fine classic idiom is nearly gone in the N. T. Purpose is expressed chiefly by i[na or the inf. For the future part. of purpose see Mt. 27:49; Ac. 8:27; 22:5; 24:11, 17. In Heb. 13:17, w`j avpodw,sontej, there is as much cause as purpose. Blass424 wrongly accepts avspaso,menoi in Ac. 25:13. The present part. is also used in the sense of purpose where the context makes it clear. So Ac. 3:26, avpe,steilen auvto.n euvlogou/nta. Cf. Lu. 13:6 f.; Ac. 15:

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1129

27; RO. 15:25. But it is not absent from the papyri. Cf. P. Goodsp. 4 (ii/B.C.) avpesta,lkamen- koinologhso,meno,n soi. So also the present part., P. Oxy. 275 (A.D. 66), diakonou/[ n] ta kai. poio[ u/] nta.

Condition. The use of the conditional disappeared more rapidly than the temporal and causal in the later Greek.425 It is only the protasis, of course, which is here considered. It is still a common idiom in the N. T. In Mt. 16:26 we have eva.n to.n ko,smon o[lon kerdh,sh|, while in Lu. 9:25, we find kerdh,saj to.n ko,smon o[lon. Here it is the condition of the third class plainly enough. See poih,saj e;sh| ktl)Ã in B. G. U. 596 (A.D. 84). In 1 Cor. 11:29, mh. diakri,nwn, it may be the first class condition with eiv that is the equivalent, but one cannot always be certain on this point. Cf. Ro. 2:27, telou/sa; Gal. 6:9, mh. evkluo,menoi; 1 Tim. 4:4, lambano,Ä menon; Heb. 2:3, avmelh,santej; 7:12, metati,qeme,nhj) Moulton426 denies that the participle stands in the N. T. for a condition of the second class (unreal condition). In Lu. 19:23, kavgw. evlqw.n su.n to,kw| a'n auvto. e;praxa, the participle is part of the apodosis, while the condition is implied in the 'preceding question. Moulton427 rightly notes that one can no longer decide by the presence of mh, with the participle that it is conditional or concessive, since mh, has come in the koinh, to be the usual negative of participles. There is no instance of ay with the participle in the N. T., though Moulton (Prol., p. 167) quotes one in a koinh, inscr., I. M. A. iii, 174, diakio,teron a'n swqe,nta (in a despatch of Augustus). For w`j a;n see Particles with Participles.

Concession. This is also a frequent construction. Cf. Mt. 14:9, luphqei,j. The context calls for the adversative idea in 7:11, ponhroi. o;ntej. See further Mt. 26:60; 14:5; Mk. 4:31; Jo. 12:37; 21:11; Jas. 3:4; Ac. 13:28; Ro. 1:21, 32; 9:22; 1 Cor. 9:19; Jude 1:5. To avoid ambiguity the Greek often used particles to make the concessive idea plain, and this idiom survives in the N. T. Cf. kai, ge- u`pa,rconta (Ac. 17:27), kai, toi genhqe,ntwn, (Heb. 4:3), kai,per more frequently as in Ph. 3:4; Heb. 5:8; 7:5; 12:17; 2 Pet. 1:12. In Heb. 11:12 we also have kai. tau/ta nenekrwme,nou) Kai,toige occurs only with the finite verb as in Jo. 4:2.428 So koi,toi in Ac. 14:17. It is worth while to note the survival of ouv with kai, ge in Ac. 17:27.429 Moulton (Prol., p. 231), admits Wellhausen's (Einl., p. 22) claim that lalei/ blasfhmei/. (Mk. 2:7) is an Aramaism for two Aramaic participles,

1130 A GRAMMAR OF THE GREEK NEW TESTAMENT

"the second of which should appear as a participle" as in: Lu. 22:65, blasfhmou/ntej e;legon. But W. H. punctuate lalei/; blasfhmei/.

( g) The Absolute Participle in Subordinate Clauses. It is not strange that the participle should have been used in clauses that stand apart from the rest of the sentence. There it has its adjectival agreement. It is but a step further than the ordinary circumstantial participle which makes an additional statement. All the varieties of the circumstantial participle can appear in the absolute participle.

Nominative Absolute. It is possible thus to explain some examples of anacolutha in ancient Greek430 and the N. T., though Blass431 demurs. Cf. o` pisteu,wn eivj evme,ÄÄpotmaoi. evk th/j koili,aj auvtou/ r`eu,sousin (Jo. 7:38) ; evpigno,ntejà de. ÄÄfwnh. evge,neto mi,a evk pa,ntwn (Ac. 19:34); o` nikw/n dw,sw auvtw|/ (Rev. 3:21). Cf. also tw/n qelo,ntwn and oi` kate,sqontej (Mk. 12:40). So Mk. 7:19; Rev. 2:26. At any rate it is the nominativus pendens, and there is not any special difference. In the modern Greek (Thumb, Handb., p. 169) the nominative absolute with the participle occurs, though rare, and usually a conjunctional clause has supplanted the genitive absolute.

Accusative Absolute . This construction was used with impersonal verbs or phrases like de,onà evxo,nà paro,n. It was probably an appositional addition to the sentence.432 It has nearly, if not quite, disappeared from the N. T. The adverb tuco,n (1 Cor. 16:6) is really an instance of it, but not so evxo,n in Ac. 2:29, where evsti,n is probably to be supplied. Cf. evxo.n h=n (Mt. 12:4) and de,on evsti,n (Ac. 19:36). Cf. also ouv sumfe,ron me,n in 2 Cor. 12:1. But a possible accusative absolute is gnw,sthn o;nta (Ac. 26:3), though it is very rare to see the accusative absolute with a substantive of its own.433 In such instances it was usual to have also w`j or w[sper.434 The accusative is an old idiom, appearing in the oldest Greek title known to us.435 But it came to be rather common in Thucydides.436 It was rare in the Attic orators. Luke avoids the accusative absolute in Ac. 23:30, by an awkward437

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1131

use of the genitive absolute, mhnuqei,shj de, moi evpiboulh/j eivj to.n a;nÄ dra e;sesqai. The papyri use evxo,ntoj rather than evxo,n.438 We do not have the acc. absolute in Ph. 1:7, since u`ma/j o;ntaj is a resumption (apposition) of u`ma/j before. Genitive Absolute. It is by no means certain that the case is always genitive. Indeed, it is pretty clear that some of these examples are ablative. Probably some are real genitives of time.439 The Sanskrit uses chiefly the locative in these absolute constructions. It is possible that the Latin ablative absolute may sometimes be locative or instrumental.440 The use of the true genitive in the Greek idiom is probably to be attributed to expressions of time in the genitive case with which participles were used. Then the temporal circumstantial participle was right at hand. It is in Attic prose, particularly the orators, that we see the highest, development of the idiom.441 The accusative absolute was just as idiomatic as this genitive-ablative construction, but it did not get the same hold on the language.442 See Cases for further remarks. The koinh, shows a rapid extension of the genitive absolute. "In the papyri it may often be seen forming a string of statements, without a finite verb for several lines."443 In the N. T. different writers vary greatly, John's Gospel, for instance, having it only one-fourth as often as the Acts.444 The most frequent use of the idiom is when the substantive (or pronoun) and the participle stand apart with no syntactical connection with any part of the sentence. Cf. Mk. 4:17, ei=ta genoÄ me,nhj qli,yewj h' diwgmou/ dia. to.n lo,gon euvqu.j skandali,zontai; Ac. 12: 18, genome,nhj de. h`me,raj h=n ta,racoj ouvk ovli,goj; 18:20; 7:5; Eph. 2:20; Mk. 8:1; 2 Pet. 3:11; Heb. 9:6-8, 15, 19. These are perfectly regular and normal examples. But sometimes the genitive absolute occurs where there is already a genitive in the sentence. So Mt. 6:3, sou/ de. poiou/ntoj- h` avristera, sou; Jo. 4:51; Ac. 17:16. In Mk. 14:3 we find a double gen. absolute o;ntoj auvtou/- katakeime,nou auvtou/) Even in the classical Greek the genitive absolute is found when the participle could have agreed with some substantive or pronoun in the sentence.445 It was done apparently to make the

1132 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

participial clause more prominent. The papyri show illustrations. of the same thing,446 as in B. U. 1040 (ii/A.D.) cai,rw o[ti moi tau/ta evpoi,hsajà evmou/ metamelome,nou peri. mhdeno,j. It is fairly common in the N. T. We have it even when the part. refers to the subject of the verb, as in Mt. 1:18, mnhsteuqei,shj th/j mhtro.j auvtou/ Mari,aj- eu`re,qh e;cousa. In Ro. 9:1 the construction is regular, though moi and mou occur. In Mt. 8:1 we find kataba,ntoj auvtou/- hvkolou,qhÄ san auvtw|/. Cf. 5:1; 9:18; 17:22; 2 Cor. 4:18, etc. Likewise the genitive and the accusative come together as in Jo. 8:30, auvtou/ lalou/ntoj- evpi,steusan eivj auvto,n. Cf. also Mt. 18:25; Ac. 28:17. Quite unusual is Ac. 22:17 where we have moi u`postre,Ä yantià proseucome,nou mou and gene,sqai me. The N. T. occasionally uses the participle alone in the genitive absolute according to the occasional classic usage.447 In the papyri it is more frequent than in the N. T.448 In particular note the common evxo,ntoj, P. Oxy. 275 (A.D. 66). Cf. also dhlwqe,ntoj, B. U. 970 (ii/A.D.). See Mt. 17:14, evlqo,ntwn; 17:26, eivpo,toj; Ac. 21:31, zhtou,ntwn. In Lu. 12:36, evlqo,ntoj kai. krou,santoj euvqe,wj avnoi,xwsin auvtw|/, we have the genitive participle although auvtw|/ is present. Cf. B. G. U. 423 (ii/A.D.) o[ti mou kinduneu,santoj eivj qa,lassan e;swse, where me the object of e;swse is not expressed.

(f) The Independent Participle in a Sentence. There is no doubt that the use of the absolute participle (nominative, accusative, genitive-ablative) is a sort of "implied predication."449 It remains to be considered whether the participle ever forms an independent sentence. We have seen that the inf. is occasionally so used. It is but a step from the independent clause to the independent sentence. Did the participle take it? The nominative absolute as a sort of anacoluthon appears in the ancient Greek. Cf. Plato, Apol. 21 C, kai. dialego,menoj auvtw|/ e;doxe, moi o` avnh.r ei=nai sofo,j. As the genitive-absolute, like other circumstantial participles, retreated before the conjunctional clauses, there was an increasing tendency to blur or neglect the grammatical case agreements in the use of the participles. The N. T., like the koinh, in general, shows more examples of the anacoluthic nominative participle than the older Greek.450 The mental strain of so many participles in rapid conversation or writing made anacolutha

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1133

easy.451 "Hence even writers of systematic training could not but occasionally blunder in the use of the circumstantial participle." Jannaris had thus concluded that the late Greek showed an independent use of the participle as anacoluthon.452 Blass453 would go no further than this. Viteau454 found abundant illustration of the independent use of the anacoluthic participle in the LXX. Viteau explains it as a Hebraism. But Moulton455 claims that the subject is removed from the realm of controversy by the proof from the papyri. Thumb456 finds the idiom in classical Greek and in the koinh, (in the LXX, N. T., papyri, inscriptions, etc.). It is easy to be extreme on this point of dispute. In the chapter on Mode (the Imperative) adequate discussion appears concerning the participle as imperative. That discussion need not be repeated. It may be insisted, however, again that the participle in itself is never imperative nor indicative, though there seem to be examples in the N. T., as in the papyri, where, because of ellipsis or anacoluthon, the participle carries on the work of either the indicative or the imperative. In examples like 2 Cor. 1:3, euvloghto.j o` qeo,j, either evsti,n or e;stw may be supplied with the verbal adjective. It must not be forgotten that this is the work of the interpreter to a large extent rather than of the grammarian. The manuscripts often vary in such examples and the editors differ in the punctuation. But the grammarian must admit the facts of usage. The papyri and the N. T. show that sometimes the participle was loosely used to carry on the verbal function in independent sentences.457 Cf. avpostuÄ gou/ntej to. ponhro,nà kollw,menoi tw|/ avgaqw|/ (Ro. 12:9), for instance, where we have a complete sentence without connection with anything else. The preceding sentence is h` avga,ph avnupo,kritoj (an independent sentence itself) and it is followed by a series of independent participles (verses 10-13). In verse 14 we have abruptly euvlogei/te- kai. mh. katara/sqe (imperatives) and then the absolute infinitive cai,rein (imperatival also). The point seems to be incontrovertible. Cf. also Col. 3:16. It is only necessary to add a word about the independent participle in the midst of indicatives, since this use is far more frequent than the imperative idiom just noted. In general it may be said that no participle

1134 A GRAMMAR OF THE GREEK NEW TESTAMENT

should be explained in this way that can properly be connected with a finite verb. In Ro. 12:6, e;contej de, it is clear that we cannot carry on the participle as subordinate to e;comen or evsmen in the preceding verses. W. H. boldly start a new sentence. In either case, whether we have comma or period before, we must take e;contej as imperatival or indicative, on the one hand, or, on the other hand, supply evsme,n or w=men as poiei/te is supplied in Ro. 13: 11 with kai. eivdo,tej to.n kairo,n.458 But other examples leave no such alternative. We may first summarize Moulton's satisfactory exposition of the matter. There is a striking similarity between the third person plural indicative and the participle in the IndoGermanic tongues (*bheronti, ferunt, fe,rousi, bairand, etc.). The frequent ellipsis of est in the Latin perfect and passive is to be noted also. The probability that the Latin second plural middle indicative is really a participle which has been incorporated into the verb inflection (cf. sequimini and e`po,menoi) is also suggestive. This fact may point to the prehistoric time when the Latin used the participle as indicative. The papyri re-enforce the argument strongly. We quote a bit from Moulton459: "Tb. P. 14 (ii/B.C.), tw/i ou=n shmainome,nwi `Hra/ti parhggelko,tej evnw,pionà 'I gave notice in person' (no verb follows). Tb. P. 42 (ib.), hvdikhme,noj (no verb follows). A. P. 78 (ii A.D.) bi,an pa,scwn e`ka,stote, etc. (no verb)." This may serve as a sample of many more like it. Moulton (Prol., p. 223) adds that use of the part. as ind. or imper. in the papyri is "not at all a mark of inferior education." See 1 Pet. 2: 12 where e;contej does not agree with the paroi,kouj. We may now approach the passages in dispute between Winer460 and Moulton.461 Moulton passes by Winer's suggestion that in 2 Cor. 4:13 e;contej is to be taken with pisteu,omen. This is probable, though awkward. So in 2 Pet. 2:1 the participles can be joined with pareisa,xousin. But in Ro. 5:11 it is, Moulton argues, somewhat forced to take ouv mo,non de,à avlla. kai. kaucw,menoi otherwise than as independent. If we once admit the fact of this idiom, as we have done, this is certainly the most natural way to take it here. Moulton is silent as to stello,menoi in 2 Cor. 8:20. Winer connects it with sunepe,myamen in verse 18 and he is supported by the punctuation of verse 19 as a parenthesis by W. H. But even so in verse 19 we have ouv mo,non de. avlla. kai. ceirotonhqei,j (cf. Ro. 5:11) stranded with no verb. Moulton also passes by Heb. 6:8 and 2 Pet. 3:5. In Heb. 7:1 Moulton follows W. H. in reading o` (not

VERBAL NOUNS ( vONOMAT TOU `RHMATOS) 1135

oa}j% sunanth,saj on the authority of C*LP against aABC462DEK 17. So he sees no necessity for taking e`rmhneuo,menoj as an indicative. In Heb. 8:10; 10:16, Moulton takes didou,j as parallel with evpiÄ gra,yw, whereas Winer would resolve evpigra,yw into a participle. Here Moulton is clearly right. In Ac. 24:5, eu`ro,ntej ga,r, we have anacoluthon as both Winer and Moulton agree. Moulton adds: "Luke cruelly reports the orator verbatim." Moulton omits to comment on Winer's explanation of the parenthetical anacoluthon in 2 Pet. 1:17, labw.n ga,r. It is a violent anacoluthon and Winer does not mend it. Note 2 Cor. 5:6, qarrou/ntej, where after a parenthesis we have qarrou/men de, (resumptive). But Moulton takes 2 Cor. 7:5 qlibo,menoi as an example of the "indicative" participle. So does he explain Ro. 12:6 e;contej, and e;cwn in Rev. 10:2. In Ac. 26:20 the MSS. vary between avpagge,llwn and avph,ggellon. In Heb. 10:1 e;cwn will also be independent if du,natai be read. In Ph. 1:30 e;contej has u`mi/n, above and halts in the case agreement. On the whole, therefore, we may conclude that, while every instance is to be examined on its merits, a number of real examples of the idiom may be admitted in the N. T. Viteaul has entirely too large a list of such instances. Many of them admit a much simpler explanation as in Ph. 1:30 above. In Revelation, it is true, there is more than usual laxity in the agreement of the participle, especially when it is in apposition. There is also a change from nominative to accusative between ivdou, and ei=don as in Rev. 4:1-5; 7:9; 14:1-3; 14:14, etc. But there are real examples in Rev., as kai. e;cwngrk grk(1:16), le,gwn,grk grk(11:1). With all this development along a special line we must not forget that the participle is both adjective and verb. Blass463 has a careful discussion of "the free use of the participle." In Col. 1:26 he notes that the participle avpokekrumme,non is continued by the indicative evfanerw,qh. Cf. Jo. 5:44.

(g) Co-ordination between Participles. Blass464 uses the term "conjunctive" participle instead of a special use of the "circumstantial" participle. It is not a particularly happy phrase. But it does accent the notion that this participle, though an addition to the principal verb, is still joined to it in grammatical agreement. Blass465 shows clearly how identity of action may be expressed by two finite verbs, as well as by the pleonastic participle of identical action. Cf. Jo. 1:25 kai. hvrw,thsan auvto.n kai. ei=pan (Mt. 15:23 hvrw,toun le,gontej), 12:44 e;kraxen kai. ei=pen

1136 A GRAMMAR OF THE GREEK NEW TESTAMENT

(Mt. 8:29 e;kraxan le,gontej), 13:21 evmartu,rhsen kai. ei=pe (Ac. 13: 22 ei=pen marturh,saj), 18:25 hvrnh,sato kai. ei=pen, (Mt. 26:70 hvrnh,Ä sato le,gwn), where John prefers the particularity of the finite verb. But see also Lu. 6:48, e;skayen kai. evba,qunen, 'he dug and deepened'= 'he dug deep.' Cf. Jo. 8:59. There remains the relation of participles to each other when a series of them comes together. There is no rule on this subject beyond what applies to other words. Two or more participles may be connected by kai, as in Ac. 3:8, peripatw/n kai. a`llo,menoj kai. aivnw/n to.n qeo,n. But we have asyndeton466 in Ac. 18:23, dierco,menoj th.n Galatikh.n cw,ranà sthri,zwn tou.j maqhta,j. Cf. Lu. 6:38, me,tron kalo.n pepiesme,non sesaÄ leume,non u`perekcunno,menon dw,sousin. Sometimes kai, occurs only once as in Mk. 5:15, kaqh,menon i`matisme,non kai. swfronou/nta. may be a subtle reason for such a procedure as in Ac. 18:22, katelqw.n eivj Kaisari,anà avnaba.j kai. avspasa,menoj, where the first participle stands apart in sense from the other two. Cf. also Mk. 5: 32. In a list of participles one may be subordinate to the other as in Mk. 5:30, evpignou.j evn e`autw|/ th.n evx auvtou/ du,namin evxelqou/san evpistrafei,j) This accumulation of participles is only occasional in the Synoptic Gospels (cf. Mt. 14:19; 27:48; and, in particular, Mk. 5:25-27), but very common in Acts and the Pauline Epistles. Blass467 concedes to Luke in Acts "a certain amount of stylistic refinement" in his use of a series of participles, while with Paul it is rather "a mere stringing together of words," an overstatement as to Paul. Luke was not an artificial rhetorician nor was Paul a mere bungler. When Paul's heart was all ablaze with passion, as in 2 Corinthians, he did pile up participles like boulders on the mountain-side, a sort of volcanic eruption. Cf. 2 Cor. 3:8-10; 6:9 f.; 9:11ff. But there is always a path through these participles. Paul would not let himself be caught in a net of mere grammatical niceties. If necessary, he broke the rule and went on (2 Cor. 8: 20). But Moulton468 is right in saying that all this is "more a matter of style than of grammar." It is rhetoric.

(h) Ouv and mh, with the Participle. It is worth noting that in Homer469 ouv is the normal negative of the participle, mh, occurring only once, Od. 4. 684, and in an optative sentence of wish. It cannot be claimed that in Homer has won its place with the participle. In modern Greek mh, alone occurs with the present participle (Thumb, Handb., p. 200). It is generally said that

VERBAL NOUNS ( vONOMATATOU `RHMATOS) 1137

Addenda 2nd ed.

in classical Attic ouv is always the negative of the participle unless condition or concession is implied when the negative is mh,. But if one looks at all the facts up to 400 B.C. he will go slow before he asserts that mh, is proof that the participle shows a conditional or concessive force.470 Jannaris471 claims the rule only for Attic, "though even here ouv is not rarely replaced by mh,," that is to say, the rule does not apply even in Attic. The use of "replaced" is wholly gratuitous when it is admitted that the rule does not apply outside of Attic. It is so hard to be historical always even in an historical grammar. If one takes the long view, from Homer with its one use of to the modern Greek with nothing but ouv he sees a steady progress in the use of mh, which gradually ousted ouv altogether. The Attic marks one stage, the koinh, another. It is true that in the Attic there is a sort of correspondence between ouv and the participle and the indicative with ouv on the one hand, while, on the other, mh, and the participle correspond to the subjunctive or the optative with mh,. But ouv occurred in Homer with the subj. and persisted with the indicative. The lines crossed and the development was not even, but on the whole mh, gradually pushed ouv aside from the participle. In the N. T., as in the koinh, generally, the development has gone quite beyond the Attic. In the Attic the use of ouv was the more general, while in the koinh, the use of mh, is normal. In the N. T. there is no need to explain mh, with the participle. That is what you expect. Cf. Lu. 12:33 mh. palaiou,mena, Jo. 5:23 o` mh. timw/n Ac. 17:6 mh. eu`ro,ntej, Heb. 11:13 mh. komisa,menoi. In the N. T. it is ouv that calls for explanation, not mh,. But it may be said at once that the N. T. is in thorough accord with the koinh, on this point. Even in a writer of the literary koinh, like Plutarch472 one notes the inroads of mh,. The papyri go further than Plutarch, but still have examples of ouv, like ouv kekomisme,nai P. Par. (B.C. 163), to.n ouvk evn leukai/j evsqh/sin evn qea,trw| kaqi,santa 0. P. 471 (ii/A.D.), ovde,pw peplhÄ rwko,twn 0. P. 491 (ii/A.D.), ouv duna,menoj A. P. 78 (ii/A.D.).473 Moulton474 thinks that in many of these papyri examples there is "the lingering consciousness that the proper negative of a downright fact is ouv." In general it may be said of the koinh, that the presence of a with the participle means that the negative is clear-cut

1138 A GRAMMAR OF THE GREEK NEW TESTAMENT

and decisive. Cf. Mt. 22:11 ouvk evndedume,non e;nduma ga,mou, (Lu. 6: 42) ouv ble,pwnà (Jo. 10:12) o` misqwto.j kai. ouvk w'n poimh,n, (Ac. 7:5) ouvk o;ntoj auvtw|/ te,knougrk grk(17:27) kai, ge ouv makra.n - u`pa,rconta, (26: 22) ouvde.n evkto.j le,gwn,grk grk(28:17) ouvde.n poih,saj, (1 Cor. 4:14) ouvk evnÄ tre,pwn,grk grk(9:26) w`j ouvk ave,ra de,rwn, (2 Cor. 4:8) avll v ouv stenocwrou,Ä menoi, (Ph. 3:3) kai. ouvk evn sarki. pepoiqo,tej, (Col. 2:19) kai. ouv kratw/n, (Heb. 11:1) pragma,twn ouv blepoume,nwn,grk grk(11:35) ouv prosdexa,meÄ noi, (1 Pet. 1:8) ouvk ivdo,ntej grk(2:10) oi` ouvk hvlehme,noi. In all these we have no special departure from the Attic custom, save that in Ac. 17:27 the participle is concessive. But we have just seen that the Attic was not rigid about ouv and mh, with the participle. In two of the examples above ouv and mh, come close together and the contrast seems intentional. Thus in Mt. 22:11 we have ouvk evndeduÄ me,non e;nduma ga,mou, while in verse 12 we read mh. e;cwn e;nduma ga,mou. The first instance lays emphasis on the actual situation in the description (the plain fact) while the second instance is the hypothetical argument about it. In 1 Pet. 1:8 we read oa}n ouvk ivdo,ntej avgapa/teà eivj oa}n a;rti mh. o`rw/ntej pisteu,ontej de. avgallia/te. Here ouv harmonizes with the tense of ivdo,ntej as an actual experience, while mh, with o`rw/ntej is in accord with the concessive idea in contrast with pisteu,ontej) Cf. Hort in loco who holds that the change of particles here is not capricious. "Though Blass thinks it artificial to distinguish, it is hard to believe that any but a slovenly writer would have brought in so rapid a change without a reason."475 It may be admitted further that "in Luke, Paul and Hebrews we have also to reckon with the literary consciousness of an educated man, which left some of the old idioms even where mh,) had generally swept them away."476 See also ta. mh. kaqh,Ä konta (Ro. 1:28) and Text. Rec. ta. ouv avnh,konta (Eph. 5:4). Cf. mh, and ouv in Ac. 9:9. Blass477 notes that the Hebrew alo is regularly translated in the LXX by ouv without any regard to the Greek refinement of meaning between ouv and mh, with the participle. Hence in the N. T. quotations from the LXX this peculiarity is to be noted. Moulton478 observes also that, while this is true, the passages thus quoted happen to be instances where a single word is negatived by ouv. Cf. Ro. 9:25 th.n ouvk hvgaphme,nhn, (Gal. 4:27) h` ouvk ti,ktousaà h` ouvk wvdi,nousa. A case like Ac. 19:11, ouv ta.j tucou,saj, is, of course, not pertinent. It is a "common vernacular phrase,"479 besides the fact that ouv is not the

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1139

negative of. the participle480 any more than it is in Ac. 19:11; 28: 21. Moulton481 also rules out ouvk evxo,n (2 Cor. 12:4) on the ground that it is the equivalent of the indicative. The copula is not expressed. But note ouvk evxo,ntoj, Oxy. 275 (A.D. 66). On this count the showing for ouv with the participle is not very large in the N. T. Luke has ouv five times with the participle (Lu. 6:42; Ac. 7:5; 17:27; 26:22; 28:17). Paul leads with a dozen or so (Ro. 9:25; Gal. 4:27 twice; 1 Cor. 4:14; 9:26; 2 Cor. 4:8, 9; Ph. 3:3; Col. 2:19; 1 Th. 2:4). Hebrews has two (11: 1, 35) and Peter three (1 Pet. 1:8; 2:10; 2 Pet. 1:16, ouvÄÄ avlla,). Matthew has only one one(22:11), and note mh. e;cwn in the next verse. The MSS. vary also between the negatives as in Mt. 22:11, where C482D have mh, which Blass483 adopts with his whimsical notions of textual criticism. At any rate Matthew, Luke (Gospel) and John use mh, almost exclusively with the participle, while Mark, James, the Johannine Epistles and Revelation do not have ouv at all with the participle. In Ro. 8:20, ouvc e`kou/sa, the old participle is merely an adjective as in Heb. 9:11. In Ro. 9:25, to.n ouv lao,n, the negative occurs with a substantive (quotation from LXX). The ancient Greek would usually have added o;nta.

(i) Other Particles with the Participle. The ancient Greek484 had quite a list of adverbs (particles) that were used with the circumstantial participle on occasion to make clearer the precise relation of the participle to the principal verb or substantive. Some of these (like a[teà oi-onà oi-a% no longer occur with the part. in the N. T. But some remain in use. These particles, it should be noted, do not change the real force of the participle. They merely sharpen the outline. The simplest form of this usage is seen in the adverbs of time like to. pro,teron (Jo. 9:8); pote, (Gal. 1:23. Cf. Eph. 2:13; Lu. 22:32); puno,teron (Ac. 24: 26). In Mk. 9:20; Jo. 5:6 note other expressions of time. More idiomatic is the use of euvqu,j as in eivselqou/sa euvqu,j (Mk. 6:25). Cf. also h;dh ovyi,aj genome,nhj (Mk. 15:42), e;ti w;n (2 Th. 2:5) and a;rti evlqo,ntoj Timoqe,ou (1 Th. 3:6). Blass485 denies that a[ma with the participle in the N. T. suggests simultaneousness or immediate sequence. He sees in a[ma kai. evlpi,zwn (Ac. 24:26) only 'withal in the expectation,' not 'at the same time hoping.' I question

1140 A GRAMMAR OF THE GREEK NEW TESTAMENT

the correctness of Blass' interpretation on this point. Cf. also a[ma avne,ntejgrk grk(27:40); proseuco,menoi a[ma kai. peri. h`mw/n (Col. 4:3), where it requires some overrefinement to refuse the classic idiom to Luke. Under the concessive participle we saw examples of kai, ge (Ac. 17:27), kai,toi (Heb. 4:3), kai,per (Heb. 5:8, etc.). There is also the use of o[mwj in the principal sentence to call attention to the concessive force of the participle (1 Cor. 14:7). So ou[twj points back to a participle of time or manner (Ac. 20: 11). Worth noting, besides, is kai. tou/to as in Ro. 13:11, though here a finite verb may be implied. So also kai. tau/ta nenekrwme,nou (Heb. 11:12). There remain w`jà w`si,à w[sper. The use of w`sei, (Ro. 6:13) and of w[sper (Ac. 2:2) is limited to condition or comparison. It is only with w`j that there is any freedom or abundance. Blass486 notes the absence of the accusative absolute with w`j in the N. T. and its absence from the future participle save in Heb. 13:17, where it is not strictly design. There is nothing specially significant in the phrase ouvc w`j, 'not as if,' in Ac. 28:19; 2 Jo. 1:5. The N. T., like the classical Greek, uses w`j without the participle in abbreviated expressions like w`j tw|/ kuri,w| (Col. 3:23); w`j evn h`me,ra| (Ro. 13:13); w`j di v h`mw/n (2 Th. 2:2), etc., where the participle is easily supplied from the context.487 In some instances one must note whether the particle does not belong with the principal verb. But, common as w`j is with the participle, it does not change the nature of the participle with which it occurs.488 The participle with w`j may be causal, temporal, conditional, manner, etc. Then again w`j may be used to express the notion of the speaker or writer as well as that of one who is reported. In truth, cos implies nothing in itself on that point. The context alone must determine it.489 The various uses of w`j itself should be recalled. There may be nothing but comparison, as in w`j evxousi,an e;cwn (Mk. 1:22); w`j ouvk ave,ra de,rwn (1 Cor. 9:26). So also Mk. 6:34; 2 Cor. 6: 9 f.; 1 Pet. 2:13, 16. In Lu. 22:26 f. observe w`j o` diakonw/n. The causal idea is prominent in w`j hvlehme,noj (1 Cor. 7:25). Cf. Heb. 12:27 and D in Ac. 20:13, w`j me,llwn. The concessive or conditional notion is dominant in 1 Cor. 7:29 f.; 2 Cor. 5:20, w`j tou/ qeou/ parakaÄ lou/ntoj di v h`mw/n. So also in Ac. 3:12; 28:19; 2 Jo. 1:5. In Lu. 16:1, w`j diaskorpi,zwn, the charge is given by Jesus as that of the

VERBAL NOUNS ( vONOMATA TOU `RHMATOS) 1141

slanderer ( dieblh,qh) and the context implies that it is untrue (only alleged).490 Pilate makes a similar use of w`j avpostre,fonta to.n lao,n in Lu. 23:14. He declines by the use of w`j to accept the correctness of the charge of the Sanhedrin against Jesus. For a similar use see w`j me,llontaj (Ac. 23:15); w`j me,llwngrk grk(23:20); proÄ fa,sei w`j mello,ntwn (genitive absolute 27:30). But in 2 Cor. 5: 20 (see above) Paul endorses the notion that he is an ambassador of God and w`j is not to be interpreted as mere pretence. God is speaking through Paul. There is no instance of a;n with the participle in the N. T. as appears in classic Greek. Winer491 notes two instances of w`j a;n with the participle in the LXX (2 Macc. 1:11; 3 Macc. 4:1). To these Moulton492 adds another (2 Macc. 12:4) and a genitive absolute example in the papyri, Par. P. 26 (11/B.C.), w`j a'n euvtakthqhsome,nwn. Cf. also ib., w`j a'n u`po. th/j limh/j dialuo,menoi. The inscrs. show it also, 0. G. I. S. 90, 23 (ii/B.C.), w`j a'n- sunesthkui,aj. Blass493 finds a genitive absolute with w`j a;n Barnabas 6:11. All this is interesting as foreshadowing the modern Greek use of sa,n as a conjunction.494

1 K.-B1., Bd. II, p. 4.

2 In K.-G. (Bd. II, p. 1) the ch. begins thus: "Lehre von den Partizipialen; dean Infinitiv and dem Partizipe." Both are "participles" and both are "infinitives."

3 Goodwin, M. and T., p. 297.

4 Moulton, Prol., p. 202.

5 Ib., p. 203.

6 Whitney, Sans. Gr., pp. 347 ff.

7 Cf. Moulton, Prol., p. 202; Giles, Man. of Comp. Philol., p. 469; Vogrinz, Gr. d. hom. Dial., 1889, p. 139.

8 Cf. Giles (Man., p. 470) for lu,Äein, and its relation to the Sans. -san-i.

9 Brug., Griech. Gr., p. 515.

10 Ib.

11 Ib.

12 Prol., p. 203.

13 Burton, N. T. M. and T., p. 143, has four. But see Robertson, Short Gr. of the Gk. N. T., p. 188.

14 Monro, Hom. Gr., p. 154.

15 Ib., pp. 157, 159.

16 Whitney, Sans. Gr., § 983.

17 Monro, Hom. Gr., p. 158. It seems a bit odd to find Radermacher (N. T. Gr., p. 145) saying of the inf.: "in seiner urspriinglichen Bedeutung als Modus." The inf. is not a mode and the original use was substantival, not verbal.

18 Monro ib., p. 179.

19 Birkletn, Entwickelungsgesch. des substantivierten Infin., 1888, p. 2 f.

20 Monro, Hom. Gk., pp. 158 ff. Cf. Brug., Griech. Gr., p. 515.

21 Goodwin, Moods and Tenses, p. 299.

22 Gildersl. Am. Jour. of Philol., 1882, p. 195.

23 Jann., Hist. Gk. Gr., p. 568.

24 Burton, N. T. M. and T., p. 143.

25 Gildersl., Am. Jour. of Philol., 1882, p. 195.

26 Birklein, Entw. d. subst. Infinitivs, p. 4 f.

27 Jann., Dist. Gk. Gr., p. 576. Hesseling (Essai hist. sur l'infinitif grec, 1892, p. 5) puts the matter too strongly.

28 Gildersl., Am. Jour. of Philol., 1882, p. 195.

29 Jann., Hist. Gk. Gr., p. 569.

30 Ib., p. 480.

31 Hist. Gk. Gr., p. 568.

32 Kalker, Questiones de Elocutione Polyb., 1880, p. 302.

33 Votaw, The Use of the Inf. in Bibl. Gk., 1896, p. 55

34 lb., p. 50.

35 Ib., p. 52.

36 Ib.

37 Thompson, Synt. of Attic Gk., p. 248. Cf. Jann., Hist. Gk. Gr., p. 574, for list of verbs with ba in late Gk.

38 Rueger, Beitr. zur hist. Synt. d. griech. Sprache, 1895, p. 11.

39 Monro, Hom. Gr., p. 154.

40 Jebb in V. and D.'s Handb., p. 324.

41 Ib., p. 326. G. Meyer (Essays and Studien, 1885, p. 101) says that the Albanians are the only Slavic folk "dem ein Infinitiv abgeht." It .is due to the mod. Gk.

42 Thompson, Synt. of th,,, Attic Gk., p. 247.

43 Blass, Gr. of N. T. Gk., p. 221.

44 Ib., p. 222.

45 Curtius, Erlaut., p. 296.

46 Jolly, Gesch. des Inf. im Indoger., 1873, p. 16.

47 Ib., p. 22.

48 Ib., pp. 31 ff.

49 Vergl. Gr., p. 3.

50 Cf. Schroeder, Vber die formelle Untersch. der Redet. im Griechischen and Lateinischen, p. 10.

51 Griech. Gr., p. 515.

52 Prol., p. 202.

53 Ib.

54 W.-M., p. 406.

55 Goodwin, M. and T., p. 298.

56 W.-M., p. 399,

57 Jann., Hist. Gk. Gr., p. 568, f. Cf. Henry, Revue de Linguistique de la Philologie. Comparee, vol. XX, ii.

58 Monro, Hom. Gr., p. 153.

59 Prol., p. 210.

60 Votaw, Inf. in Bibl.Gk., p. 57.

61 Gr. of N. T. Gk., p. 234. Cf. 2 Esd. 6:8 to. mh. katarghqh/nai.

62 Monro, Hom. Gr., p. 154.

63 Blass, Cr. of N. T. Gk., p. 234.

64 Cf. Viteau, Le Verbe., p. 172.

65 Moulton, Prol D. 210.

66 Cf. Monro, Hom. Gr., p. 179. Gildersl. (Am. Jour. of Philol., 1912, p. 488) gave this name ("articular infinitive") to the idiom. "I watch the fate of my little things with a benevolent detachment."

67 Birklein, Entwickelungsgeschichte, p. 91.

68 Goodwin, M. and T., p. 315.

69 Monro, Hom. Gr., p. 164.

70 Goodwin, M. and T., p. 315.

71 Birklein, Entwickelungsgeschichte, p. 91.

72 Gildersl., Contrib. to the Hist. of the Inf., Transac. of the Am. Philol. Asso., 1878, pp. 5-19.

73 Goodwin, M. and T., p. 315. Hypereides, he adds, even exceeds Demosthenes.

74 Hist. Gk. Gr., p. 576.

75 Ib., p. 577.

76 Gr. of N. T. Uk., p. 233.

77 Prol., p. 213.

78 Gr. of N. T. Gk., p. 233.

79 Hist. Gk. Gr., p. 579.

80 Ib.

81 Prol., pp. 213 ff.

82 Ib., p. 213.

83 Ib., p. 215.

84 Allen, The Inf. in Polyb. Compared with the Inf. in Bibl. Gk., p. 47.

85 Inf. in Bibl. Gk., pp. 50 ff.

86 Prol., p. 216.

87 Votaw, Inf. in Bibl. Gk., p. 52.

88 Prol., p. 215.

89 Monro, Hom. Gr., p. 164.

90 Birklein, Entwick., p. 9.

91 Allen, The Inf. in Pdyb., pp. 29 ff.

92 Jann., Hist. Gk. Gr., p. 578.

93 Moulton, Prol., p. 216.

94 Ib.

95 Ib., p. 218.

96 Jann., Hist. Gk. Gr., p. 578. Cf. Birklein, Entwick., p. 101.

97 Jann., ib.

98 Moulton, Prol., p. 216.

99 Ib., p. 217.

100 Mr. H. Scott gives the following list for tou/ and the inf.:
 Pres.Aor.
Paul134
Synoptics922
Acts1113
Heb.13
Rev.-1
Jas.-1
1 Pet.-1
 3445
  
 79
 

(less 9 fr. LXX, 4 Paul, 5 Act.=70)

101 Prol., p. 217. Cf. also Gal. 3:10.

102 Cf. W.-M., p. 410 f.

103 Allen, Inf. in Polyb., p. 53. Cf. Gildersl., Am. J. of Phil., vol. XXVII, p. 105 f.

104 Votaw, The Inf. in Bibl. Gk., p. 28.

105 Burton, N. T. M. and T., p. 159. In late Gk. this use of tou/ and the inf. came to displace the circumstantial participle and even finite clauses, only to die itself in time. Cf. Jann., WA. Gk. Gr., p. 483.

106 Prol., p. 219 f.

107 Entwickelungsgesch., p. 103.

108 Hist. Gk. Gr., p. 576.

109 Krapp, Der substantivierte Inf., 1892, p. 1.

110 Allen, The Inf. in Polyb., p. 33. vEk 25, pro, 12=1179 for all.

111 De Inf. ef Part. in Inscr. Dialect. Graec. Questiones Synt., 1892, p. 73.

112 Prol., p. 220.

113 Inf. in Bibl. Gk., p. 19.

114 Goodwin, M. and T., p. 320.

115 Cf. Birklein, Entwickelungsgesch., p. 104. These preps. "retain this disqualification in the N. T." (Moulton, Prol., p. 216).

116 Thompson, Synt. of Attic Gk., p. 246.

117 Moulton, Prol., p. 216.

118 W.-M., p. 413.

119 Ib.

120 Burton, N. T. M. and T., p. 50.

121 W.-M., p. 415.

122 Ib., p. 413.

123 Votaw, Inf. in Bibl. Ck., p. 20.

124 Birklein, Entwiek., p. 104.

125 Helbing, Die Prapositionen bei Herod., p. 148.

126 Votaw, Inf. in Bibl. Gk., p. 20.

127 Birklein, Entwick., p. 107.

128 Birklein, Entwick., p. 107.

129 Hist. Gk. Gr., p. 487.

130 Jann., Hist. Gk. Gr., p. 373 f.

131 Votaw, Inf. in Bibl. Gk., p. 20.

132 Viteau, Le Verbe, p. 165.

133 Prol., p. 218.

134 Burton, N. T. M. and T., p. 161.

135 Ib., p. 220.

136 Cf. Blass, Gr. of N. T. Gk., p. 236; Moulton, Prol., p. 219; Burton, N. T. M. and T., p. 161.

137 Birklein, Entwick., p. 108.

138 Moulton, Prol., p. 215.

139 Prol., p. 215.

140 Gr. of N. T. Gk., p. 237.

141 P. 249.

142 Votaw, Inf. in Bibl. Gk., p. 20.

143 But Dalman, Worte Jesu, p. 26 f., denies that it is an Aramaean constr.

144 Hist. Gk. Gr., p. 379.

145 Gr. of N. T. Gk., p. 237.

146 Inf. in Bibl. Gk., p. 20.

147 Burton, N. T. M. and T., p. 50.

148 Blass, Gr. of N. T. Gk., p. 237.

149 Birklein, Entwick., p. 106. It is found in Polyb. also. Cf. Kalker, Questiones, p. 302; Allen, Inf. in Polyb., p. 35. Lutz (Die Casus-Adverbien bei Att. Redn., 1891, p. 18) finds it "zuerst bei Antiphon."

150 Birklein, Entwiek. p. 105.

151 Inf. in Bibl. Gk., p. 20.

152 Gr. of N. T. Gk., p. 237.

153 Allen, Inf. in Polyb., p. 34 f.

154 Jann., Hist. Gk. Gr., p. 386.

155 Entwick., p. 105.

156 Entwick., p. 105.

157 Inf. in Bibl. Gk., p. 20.

158 Moulton, Prol., p. 214.

159 Birklein, Entwick., p. 108.

160 Allen, Inf. in Polyb., p. 33.

161 Allen, Inf. in Polyb., p. 41.

162 Votaw, Inf. in Bibl. Gk., p. 20.

163 Inf. in Bibl. Gk., p. 20.

164 Whitney, Sans. Gr., § 983; Monro, Hom. Gr., p. 15S. Homer used pri,n with the inf. after both positive and negative clauses.

165 Birklein, Entwiek., p. 107.

166 Ib., p. 218.

167 Allen, Inf. in Polvb., p. 33.

168 Cf. Blass, Gr. of N. T. Gk., p. 236.

169 Votaw, Inf. in Bibl. Gk., p. 20.

170 W.-M., p. 414 note.

171 Prol., p. 220.

172 Monro, Hom. Gr., p. 154.

173 Goodwin, M. and T., p. 301.

174 Votaw, Inf. in Bibl. Gk., pp. 15, 25.

175 Allen, Inf. in Polyb., pp. 23, 32.

176 Ib., p. 27.

177 Monro, Hom. Gr., p.155 f. For Polyb. see Allen, Inf. in Polyb., pp. 23, 32.

178 Jann., Hist. Gk. Cr., p. 487.

179 Votaw, Inf.in Bibl. Gk., p. 27.

180 Monro, Hom. Gr., p. 154.

181 Inf. in Bibl.

182 See Jann., Hist. Gk. Gr., p. 4S7.

183 Hist. Gk. Gr., p. 574 f.

184 Atticismus, Bd. IV, p. Sl. Cf. also Hatz., Einl., p. 215.

185 Moulton, Prof, p. 211.

186 Ib.

187 Hist. Gk. Gr., p. 572 f. For an extended list of the verbs in the N. T. used with the complementary inf. see Viteau, Le Verbe, pp. 157

188 Hellen., p. 125. Cf. Moulton, Prol., p. 233.

189 Inf. in Bibl. Gk., p. 17.

190 Cf. Hadley and Allen, § 950; Goodwin, § 1517.

191 Votaw, Inf. in Bibl. Gk., p. 29.

192 See Blass, Gr. of N. T. Gk., p. 229.

193 As, for instance, Szczurat, De Inf. Horn. Usti, 1902, p. 17. He claims that the Horn. inf. came to serve almost all the ideas of the finite verb.

194 Hist. Gk. Gr p. 576.

195 Prol., p. 203.

196 Ib.

197 In Ac. 26:28, pei,qeij Cristiano.n poih/sai, one notes a possible absence of the strict voice in poih/sai. But it is a hard passage.

198 Monro, Hom. Gr., p. 160.

199 Ib.

200 Moulton, Prol., p. 204

201 Ib.

202 Votaw, Inf. in Bibl. Gk., p. 59, notes 5,484 aorists and 3,327 presents in the Gk. Bible. In the N. T. the ratio is 4:3, in the 0. T. 2:1.

203 Burton, N. T. M. and T., p. 53.

204 Gilders1., Am. Jour. of Philol., 1882, p. 193. Madvig, Bemerkungen uber einige Punkte des Griech., 1848, p. 321, shows how the inf. has only the time of the principal verb.

205 Votaw, Inf. in Bibl. Gk., p. 50.

206 Allen, Inf. in Polyb., p. 48.

207 Jann., Hist. Gk. Gr., pp. 486, 552 ff.

208 Prol., p. 204 f. Cf. Hatz., Einl., pp. 142, 190; Kalker, Quest., p. 281.

209 Votaw, Inf. in Bibl. Gk., p. 59.

210 Cf. Delbruck, Vergl. Synt., Tl. II, p. 460. Brug. (Griech. Gr., p. 518) takes the acc. as originally the obj. of the verb. That was not always true, as we have seen in Indirect Discourse (pp. 1037 ff.).

211 Hist. Synt., Bd. II, pp. 380, 446.

212 Uber den Infinitiv, p. 40.

213 Introd. to Comp. Gr., 1890, p. 214.

214 Gesch. de Inf., p. 247.

215 Uber den Urspr. des Substantivsatzes, p. 5.

216 Uber die formelle Untersch. der Redet., p. 28.

217 Wilhelmius, De Inf. linguarum Sanscritae, Beoticae, Persicae, Graecae, Oscae, Vmbricae, Latinae, Goticae Forma et Vsv, 1873, p. 5.

218 Moods of Indirect Quotation, Am. Jour. of Theol., Ja ., 1905.

219 Inf. in Bibl. Gk., p. 9.

220 Gr. of N. T. Gk., pp. 239-241.

221 Ib., p. 240.

222 Gr. of N. T. Gk., p. 241.

223 Ib.

224 Cf. Middleton, Analogy in Synt., p. 9. Maximus of Tyre has it in a rel. clause. Durr, Sprachl. Unters., p. 43.

225 Thompson, Synt. of Attic Gk., p. 239.

226 Blass, Gr. of N. T. Gk., p. 239.

227 Thomspon, Synt. of Attic Gk., p. 239.

228 Brug., Griech. Gr., p. 517.

229 Granewald, Der freie formelhafte Inf. der Limit. im Griech., p. 21 f.

230 Gr. of N. T. Gk., p. 229.

231 Le Verbe, p. 161.

232 Votaw, Inf. in Bibl. Gk., p. 26.

233 N. T. M. and T., p. 147.

234 Cf. Blass, Gr. of N. T. Gk., p. 255 f.; Humphreys, Th Problems of Greek, Congress of Arts and Sciences, 1904, vol. III, pp. 171 ff.

235 Monro, Hom. Gr., p. 154.

236 Brug., Griech. Gr., p. 516; Delbruck, Grundr., IV, pp. 463 ff.

237 Votaw, Inf. in Bibl. Gk., p. 10.

238 Thompson, Synt. of Attic Gk., p. 240.

239 Theol. Lit., 1903, p. 421.

240 Prol., p. 205.

241 Prol., p. 205. Allen gives no ex. of the simple inf. of purpose in Polyb., only tou/à w[steà evf v w|-te. Cf. Inf. in Polyb., p. 22.

242 Moulton, Prol., p. 216. Thuci was the first to use tou/ and the inf. for purpose (Berklein, Entwickelungsgesch., p. 58).

243 Ib., p. 217 f.

244 Inf. in Bibl. Gk., p. 21.

245 Votaw, Inf. in Bibl. Gk., p. 10.

246 N. T. M. and T., p. 150.

247 Inf. in Bibl. Gk., p. 24. Cf. W.-M., p. 409.

248 N. T. M. and T., p. 148.

249 Gr. of N. T. Gk., p. 224.

250 Baumlein, Modi, p. 339.

251 W.-M., p. 400. See Burton, N. T. M, and T., p. 150 f.

252 Allen, Inf. in Polyb., p. 21,

253 Inf. in Bibl. Gk., p. 13.

254 Votaw, Inf. in Bibl. Gk., p. 25. Cf. Ruth 2:10, ti, o[ti eu-ron ca,rin evn ovfqalÄ moi/j sou tou/ evpignw/nai, me; See also 2 Chron. 33:9; 1 Macc. 14:36.

255 Moulton, Prol., p. 217.

256 Votaw, Inf. in Bibl. Gk., p. 25.

257 Moulton, Prol., p. 216.

258 Cf. Burton, N. T. M. and T., p. 161; Moulton, Prol., p. 219.

259 Inf. in Bibl. Gk., p. 21.

260 Ib., p. 14.

261 N. T. M. and T., p. 149.

262 Votaw, Inf. in Bibl. Gk., p. 29.

263 Burton, N. T. M. and T., p. 161, mentions only 23.

264 The inf. with pri,n is common in Hom. See Monro, p.58.

265 Benard, Formes verbales en Grec d'apres le Texte d' Orodote, 1890, p. 196. See also Sturm, Die Entwick. der Konstrukt. mit pri,n, 1883, p. 3.

266 Monro, Hom. Gr., p. 162.

267 Der Gebr. des imper. Inf. ion Griech., 1891, p. 12.

268 Reinach, Pap. grecs et demotiques, 1905.

269 Unters. zur Gesch. des griech. Briefes, Phil. Zeitschr., 1905, p. 56.

270 Meisterh., p. 244.

271 Brug., Griech. Gr., p. 516.

272 Inf. in Bibl. Gk., p. 18.

273 W.-M., p. 397.

274 Prol., p. 179.

275 N. T. M. and T., p. 146.

276 Ib., p. 248.

277 Ib., p. 179.

278 Einl., p. 192.

279 Moulton Prol., p. 203.

280 For the variety of uses of the absolute inf. in ancient R. see Goodwin, M. and T., pp. 310 ff.

281 Thompson, Synt. of Attic Gk., p. 414.

282 Inf. in Bibl. Gk., p. 58.

283 See Thompson, Synt., pp. 425 ff.

284 Cf. Blass, Cr. of N. T. Gk., p. 255.

285 Thompson, Synt. of Attic Gk., p. 262.

286 Brug., Griech. Gr., p. 200.

287 Moulton, Prol., p. 221.

288 Moulton, Prol., p. 221.

289 Stahl, Krit.-hist. Synt., p. 761.

290 Whitney, Sans. Gr., p. 347.

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

292 Brug., Griech. Gr., p. 200.

293 Cf. Viteau, Essai sur la Synt. des Voix, Revue de Philol., p. 41.

294 Moulton, Prol., p. 221.

295 Ib.

296 Ib., p. 222.

297 Riem. and Goelzer, Synt., p. 707.

298 In Sans. the verbal adjs. in -ta are sometimes called passive participles (Whitney, Sans. Gr., p. 340). This form does not belong to the tense system.

299 Moulton, Prol., p. 222.

300 Brug., Griech. Gr., pp. 184, 525.

301 Goodwin, M. and T., p. 368 f.

302 Riein. and Goelzer, Synt" p. 707.

303 Moulton, Prol., p. 222.

304 But even with - toj this sometimes appears as in didaktoi. qeou/ (Jo. 6:45) where we have the ablative. Cf. Brug., Griech. Gr., p. 5221

305 Whitney, Sans. Gr., p. 202.

306 Thompson, Synt. of Attic Gk., p. 262.

307 Brug., Griech. Gr., p. 521 f.

308 Brug., Indoger. Forsch., V, pp. 80 ff.; Giles, Man., p. 473; Moulton, Prol., p. 221.

309 Williams, The Part. in the Book of Acts, 1909, p. 7.

310 Bolling, The Part. in Hesiod, Cath. Univ. Bull., 1S97, III, p. 423.

311 Ib.

312 Jann., Hist. Gk. Or., p. 505.

313 Gildersl., Stylistic Effect of the Gk. Part., Am. Jour, of Philol., 1888, p. 142.

314 Williams, The Part. in Acts, p. 7.

315 Ib., p. 10.

316 Ib., p. 505.

317 Ib.

318 Williams, Part. in Acts, p. 23.

319 Hist. Gk. Cr., p. 504.

320 Ib.

321 Ib., p. 22. Williams did not count 2 Cor. and the other Pauline Epistles.

322 Jann., Hist. Gk. Gr., p. 505.

323 V. and D., Handb., p. 333.

324 Thumb, Handb., p. 167. pf. also Blass, Gr. of N. T. Gk., p. 242.

325 Brug., Griech. Gr., p. 522.

326 Monro, Hom. Gr., p. 54. Cf. Stahl, Krit.-hist. Synt., p. 681.

327 Bolling The Part. in Hesiod, Cath. Univ. Bull., 1897, III, p. 422.

328 Brug., Griech. Gr., p. 522.

329 Farrar, Gk. Syn p. 169.

330 Ib.

331 Brug., Griech. Gr., p: 522.

332 Monro, Hom. Gr., p. 53.

333 Burton, N. T. M. and T., p. 163.

334 Blass, Gr. of N. T. Gk., p. 242. In general, on this Point, see Goodwin, M. and T., p. 357.

335 Cf. Schoemann, Die Lehre von den Redet. nach den Alten, 1862, p. 34.

336 Robertson, Short Gr., p. 194.

337 Ib.

338 Blass, Gr. of N. T. Gk., p. 245.

339 Ib.

340 W.-M., p. 436.

341 The Part. in Acts, pp. 1 ff.

342 The Part. in Acts, p. 5.

343 Synt. of Attic Gk., p. 249.

344 Goodwin, M. and T., p. 330.

345 Blass, Gr. of Nf T. Gk., p. 242.

346 This use of pa/j without art. occurs occasionally in class. Gk. See K.-G., II, p. 608 f.

347 Cf. Goodwin, M. and T., p. 330.

348 Vogrinz, Gr. des hom. Dialektes, 1889, p. 184.

349 Williams, The Part. in the Book of Acts, p. 46.

350 Cf. Blass, Gr. of N. T. Gk., p. 243.

351 Prol., p. 228.

352 Burton, N. T. M. and T., p. 167.

353 Ib., p. 169.

354 Goodwin, M. and T., p. 331.

355 N. T. M. and T., p. 175 f.

356 Blass, Gr. of N. T. Gk., p. 244.

357 So Burton, N. T. and T., p. 169 f.

358 Brug., Griech. Gr., p. 523.

359 Prol., pp. 14,

360 Brug., Griech. Gr., p. 522.

361 Moulton, Prol., p. 126. He notes Heb. 10:14, tou.j a`giazome,noujà as a good ex. of the timelessness of the part.

362 Gildersl., Synt. of Class. Gk., Pt. I, p. 139.

363 W.-M., p. 427.

364 Schaefer, Das Partizip des Aoristes bei den Tragikern, 1894, p. 5.

365 Ib.

366 Gr. of N. T. Gk., p. 197.

367 M. and T., p. 48. So Monro, Hom. Gr., p. 48.

368 N. T. M. and T., p. 59.

369 Gr. of N. T. Gk., p. 197.

370 Ib.

371 N. T. M. and T., p. 61.

372 Prol., p. 130.

373 Moulton, Prol., p. 131.

374 St. Paul the Traveller, p. 212.

375 See Ballentine, Biblibtheca Sacra, 1884, p. 7S7, for discussion of N. T, exx.

376 N. T. M. and T., p. 65.

377 N. T. M. and T., p. 66.

378 See Leo Meyer, Griech. Aor., p. 125.

379 Gilders1., Synt., Pt. I, p. 140. See Seymour, The Use of the Gk. Aorist Part., Trans. Am. Philol. Assoc., XII, p. 88 f.

380 Goodwin, M. and T., p. 47; Gildersl., Synt., Part 1, p. 139.

381 Gildersl., Synt., Pt. I, p. 140.

382 Moulton, Prol., p. 127.

383 Cf. Gildersl., Synt., Pt. I, p. 142.

384 Cf. Burton, N. T. M. and T., p. 72.

385 Cf. Delbruck, Synt. Forsch., IV, p. 97.

386 Brug., Griech. Gr., p. 523.

387 There is an expectant note in to. evkcunno,menon (Mt. 26:23).

388 Cf. Jebb in V. and D., p. 335.

389 The fut. part. is rare in the inscr. Cf. Granit, De Inf. et Partic. in Inscr. Dial. Grace. Questiones Synt., 1892, p. 122.

390 N. T. Gk., p. 166.

391 Hist. Gk. Gr., p. 493.

392 Jain., Hist. Gk. Gr., p. 493.

393 Moulton, Prol., p. 228.

394 Ib., p. 494.

395 Ib.

396 Ib., p. 228 f.

397 Prol., p. 229,

398 Gr. of N. T. Gk., p. 245.

399 Blass, Gr. of N. T. Gk., p. 245.

400 Jann., Hist. Gk. Gr., p. 495.

401 Blass, ib., p. 247.

402 The pap. show the same tendency. Cf. Moulton, Prol., p. 229. See Radermacher, N. T. Gr., p. 169.

403 Hist. Gk. Gr., p. 497.

404 N. T. M. and T., p. 176.

405 Gr. of N. T. Gk., p. 246.

406 Cf. Goodwin, M. and T., pp. 359

407 Gr. of N. T. Gk., p. 246.

408 N. T. M. and T., p. 176.

409 N. T. M. and T., pp. 166

410 Blass, Gr. of N. T. Gk., p. 247.

411 Jann., Hist. Gk. Gr., p. 499.

412 Jebb, in V. and D., p. 333.

413 Moulton, Prol., p. 229.

414 Ib. Cf. Alexander, Partic. Periphrases in Attic Orators (Am. Jour. of Philol., IV, p. 291 f.).

415 Certainly we cannot admit the idea that the part. itself has different meanings. Cf. Paul, Prin. of the Hist. of Lang., p. 158.

416 Prol., p. 230.

417 Blass, Gr. of N. T. Gk., p. 249.

418 Ib., p. 248.

419 Prol., p. 230.

420 Gr. of N. T. Gk., p. 248.

421 N. T. M. and T., p. 173. Cucuel and Riemann (Rêgles Fondamentales de la Synt. Grecque, 1888, p. 110) consider this notion an "exception," but it is not necessary to do that.

422 Goodwin, M. and T., p. 333.

423 Cf. Goodwin, M. and T., p. 335.

424 Hist. Gk. Gr., p. 504.

425 Gr. of N. T. Gk., p. 248.

426 Jann., Hist. Gk. Gr., p. 502.

427 Blass, Gr. of N. T. Gk., p. 248.

428 Prol., p. 230.

429 Moulton, Prol., p. 230,

430 Ib., p. 229.

431 Thompson, Synt. of Attic Gk., p. 259.

432 Gr. of N. T. Gk., p. 251. He calls it "antiquated." It was never very common.

433 Brug., Griech. Gr., p. 524.

434 Goodwin, M. and T., p. 339.

435 Thompson, Synt., p. 261.

436 Deiss., Exp. Times, 1906, Dec., p. 105.

437 Lell, Der Absolut-Akkusativ im Griech. bis zu Arist., 1892, p. 17.

438 Blass, Gr. of N. T. Gk., p. 252.

439 Ouvk evxo,ntoj, P. Oxy. 275 (A.D. 66).

440 Brug., Griech. Gr., p. 524.

441 Monro, Hom. Gr., p. 167 f.

442 Cf. Spieker, The Genitive Abs. in the Attic Orators, Am. Jour. of Philol., VI, pp. 310-343.

443 Blass, Gr. of N. T. Gk., p. 251.

444 Moulton, Prol., p. 74.

445 Gildersl., Styl. Effect of the Gk. Part., Am. Jour. of Philol. 1888, p. 153.

446 Goodwin, M. and T., p. 338.

447 Cf. Moulton, Prol., pp. 74, 236; Cl. Rev., XV, p. 437.

448 Goodwin, M. and T., p. 338.

449 Moulton, Prol., p. 74. This idiom is common in Xen. Roche, Beitr., p. 128.

450 Monro, Hom. Gr., p. 167.

451 Thompson, Synt. of Attic Gk., p. 259.

452 Jann., Hist. Gk. Gr., p 505.

453 Le Verbe, pp. 200

454 Ib., pp. 500, 505.

455 Prol., pp. 180 ff., 222 ff.

456 Gr. of Gk. N. T., p. 283.

457 Hellen., p. 131.

458 Moulton, Prol., p. 180, cites Meisterh., pp. 244-246, for the use of the imp. part. in decrees. It is the nominativus pendens applied to the part.

459 Moulton, Prol., pp. 180, 183 f.

460 W.-Th., p. 351 f.

461 Ib., pp. 223 f.

462 Prol., p. 224 f.

463 Le Verbe, pp. 201

464 Gr. of N. T. Gk., p. 247.

465 Gr. of N. T. Gk., p. 284 f.

466 Ib., p. 250.

467 Gr. of N. T. Gk., p. 250.

468 Prot., p. 231.

469 Ib., p. 251.

470 Monro, Hom. Gr., p. 262 f.

471 Howes, The Use of mh, with the Part., Harv. Stu. in Class. Philol.,1901, pp. 277-285.

472 Hist. Gk. Cr., p. 430.

473 Blass, Cr. of N. T. Gk., p. 255.

474 See further exx. in Moulton, Prol., p. 231.

475 Prol., p. 232.

476 Moulton, Prol., p. 232.

477 Prol., p. 232.

478 Ib.

479 Ib., p. 231.

480 Gr. of N. T. Gk., p. 255.

481 Blass, Gr. of N. T. Gk., p. 255 f.

482 Prol., p. 231.

483 Gr. of N. T. Gk., p. 255. Cf. Gildersleeve, Encroachments of mh, on ouv in later Gk., Am. Jour. of Philol., I, p. 45 f.

484 Cf. Goodwin, M. and T., pp. 340 ff.

485 Gr. of N. T. Gk., p. 252.

486 Gr. of N. T. Gk., p. 253.

487 Ib.

488 Fuhrer, De Particulae w`j cum Participiis et. Praepositionibus punctae Usu Thucydideo, 1889, p. 7.

489 Goodwin, M. and T., p. 343.

490 Cf. Blass, Gr. of N. T. Gk., p. 253.

491 W.-M., p. 378.

492 Gr. of N. T. Gk., p. 253.

493 Prol., p. 167.

494 Cf. Moulton, Prol., p. 167; Hatz., Einl., p. 217.