676

CHAPTER XV

PRONOUNS ( vANTWNUMIAI)

For the antiquity and history of pronouns see iv in chapter VII (Declensions). We are here concerned, not with the form, but with the use of pronouns.1 As a matter of fact all pronouns fall into two classes, Deictic ( deiktikai,) and Anaphoric ( avnaforikai,). They either "point out" or they "refer to" a substantive. So we get the modern terms, demonstrative and relative (cf. Monro, Homeric Gr., p.i 168 f.). But some pronouns may be demonstrative or relative according to the context. The demonstrative or deictic was the original usage. For practical purposes we have to follow a more minute division.

I. Personal Pronouns ( prwto,tupoi h' proswpikai. avntwnumi,ai). The personal pronouns (first and second persons) are deictic (I, thou). The reason for the use of pronouns, as already explained, was to avoid the repetition of the substantive. In Jo. 11:22 note the repetition of qeo,j. Cf. also Lu. 6:45.

(a) THE NOMINATIVE. As already explained, the verb uses the personal pronoun as personal suffixes, so that as a rule no need was felt for the separate expression of the pronoun in the nominative. All verbs had the personal endings like eiv─mi,├ evs─si,├ evs─ti,. The use of the personal pronoun in addition to the personal ending of the verb Was due to desire for emphasis. Then the separate expression of the pronoun led to the gradual sloughing off of the personal ending. In modern English this process is nearly complete. In Greek this process was arrested, though in modern Greek all verbs ei=nai are -- w verbs. In most cases, therefore, in Greek the existence of the personal pronoun in the nominative implies some emphasis or contrast. But this is not quite true of all examples. "The emphasis of the first and second persons is not to be insisted on too much in poetry or in familiar prose.

PRONOUNS ( vANTWNUMIAI) 677

Notice the frequency of evgw|=da├ evgw|=mai."2 In conversation it was particularly common to have the personal pronoun in the nominative. In the later Greek generally the personal pronouns show a weakening of force,3 but never to the actual obliteration of emphasis, not even in the Modern Greek.4 Moulton5 agrees with Ebeling6 that there was "no necessary emphasis in the Platonic h=n d v evgw,├ e;fhn evgw,├ w`j su. fh,j." Clearly then the frequency of the pronoun in the N. T. is not to be attributed to the Semitic influence. Even Conybeare and Stock7 see that it is not necessary to appeal to the well-known Hebrew fondness for pronouns for this usage. But Blass8 thinks that some of the MS. variations may be due to Semitic influence. We are free therefore to approach the N. T. examples on their merits.9

1. The First Person, evgw, and h`mei/j. It is easy to find in the N.T. numerous examples where evgw, shows contrast. So evgw. crei,an e;cw u`po. sou/ baptisqh/nai (Mt. 3:14), evgw. de. le,gwgrk grk(5:22), evgw, se evdo,xasa (Jo. 17:4). Cf. evgw, and su, in Jo. 17:23. The amount of emphasis will vary very greatly according to circumstances and may sometimes vanish entirely so far as we can determine. Different shades of meaning appear also as in u`pe.r ou- evgw. ei=pon (Jo. 1:30), 'I, myself.' Cf. kavgw. ouvk h|;dein auvto,n (Jo. 1:33) and kavgw. e`w,raka kai. memartu,rhka (verse 34) and note absence with second verb. Cf. Jo. 6:48; 16:33; 1 Cor. 2:1, 3. Note absence of evgw, in Mt. 5:18, 20, le,gw u`mi/n. Cf. also ti,j avsqenei/ kai. ouvk avsqenw/; (2 Cor. 11:29) with ti,j skandali,zetai kai. ouvk evgw. purou/mai; (ib.) as proof that the point must not be pressed too far in either direction.10 Further examples of evgw, may be seen in Ro. 7:17; Jo. 5:31, 34; 10:30; Eph. 5:32; Ph. 4:11. For the plural h`mei/j see h`mei/j proskunou/men (Jo. 4:22) in opposition to u`mei/j, but then follows merely oa} oi;damen. So in Ac. 4:20 note ouv duna,meqa h`mei/j aa} ei;damen and ti, kai. h`mei/j kindu─ neu,omen; (1 Cor. 15:30). Cf. Mt. 6:12. The "editorial" 'we' has already received discussion (cf. The Sentence) and may be merely illustrated here. Blass11 considers it a "wide-spread tendency among Greek writers, when they speak of themselves to say h`mei/j instead

678 A GRAMMAR OF THE GREEK NEW TESTAMENT

of evgw,." This is not always true in Paul's Epistles (Ro. 1:5), for sometimes he associates others with him in the address at the beginning. There are undoubted examples in the N. T. like oi-oi, evsmen (2 Cor. 10:11), peiqo,meqa (Heb. 13:18), gra,fomen (1 Jo. 1:4). But sometimes the plural merely associates the readers or hearers with the writer or speaker. So evfore,samen (1 Cor. 15:49), o`moiw,sw─ men, (Mk. 4:30). Sometimes the first person singular is used in a representative manner as one of a class (cf. the representative article like o` avgaqo,j). Blass12 does not find this idiom so common in Greek as in other languages, but it occurs in Demosthenes and certainly in Paul. So ti, e;ti kavgw. w`j a`pmartwlo.j kri,nomai; (Ro. 3:7). Cf. in next verse blasfhmou,meqa. See 1 Cor. 10:30; Gal. 2:18. In Ro. 7:7-25 special difficulties occur.

2. The Second Person, su, and u`mei/j. Thus in Jo. 17:5 note the contrast in me su,. Cf. Jo. 1:42 su. ei= Si,mwn- su. klhqh,sh|├ 2:10 su. teth,rhkaj, 4:9 pw/j su. vIoudai/oj, 4:10 su. a'n h|;thsaj, Ro. 2:3 o[ti su. evkfeu,xh|, Lu. 1:76 kai. su. de, etc. Cf. also Mt. 27:11. Sometimes su, has a very emphatic position, as in su. ti,j ei= (Ro. 9:20; 14:4). In 1 Cor. 15:36, a;frwn├ su. oa} spei,reij, it is possible,13 though not necessary, to take su, with a;frwn (cf. Ac. 1:24). In kai. su. evx auvtw/n ei= (Lu. 22:58) one is reminded of the Latin Et tu, Brute. See Lu. 10:15; Ac. 23:3; h' kai. su. ti, evxouqenei/j (Ro. 14:10). As examples of the plural take e;sesqe u`mei/j (Mt. 5:48), do,te auvtoi/j u`mei/j fagei/n (Mk. 6:37). See evkei/noj and u`mei/j contrasted in Jo. 5:38; u`mei/j in verse 39 and also in 44 f. Cf. Ac. 4:7; Lu. 10:24, and in particular u`mei/j o;yesqe (Mt. 27:24). For u`mei/j and h`mei/j contrasted see Jo. 4:22. In Jo. 4:35, ouvc u`mei/j le,gete, we have the same inclusive use of the second person that we noticed in the first. In Ro. 2:3, 17, the second person singular occurs in the same representative sense that the first has also. Cf. also Ro. 9:20; 11:17, etc. In Jo. 3:10, su. ei= o` dida,skaloj, we have a case of distributed emphasis. Cf. also Mt. 16:16; Jo. 9:34; 2 Cor. 1:23, as examples of this sustained emphasis, where the emphasis of the pronoun passes on to the remainder of the sentence and contributes point and force to the whole.14 On the whole the Greek language has freedom in the construction of the pronouns.15 Moulton raises16 the question if in su. ei=paj (Mt. 26:64), su. le,geijgrk grk(27:11), u`mei/j le,gete (Lu. 22:70), we do not have the equivalent of 'That is right,'

PRONOUNS ( vANTWNUMIAI) 679

but plh,n (Thayer) is against it in Mt. 26:64. Su, occurs in John more frequently than in all the Synoptics put together (Abbott, Johannine Gr., p. 297).

3. The Third Person. It has had a more radical development or lack of development. As a matter of fact the Greek had and has no definite third personal pronoun for the nominative like evgw, and su,. No nominative was as used for ou-├ oi=, etc., and this pronoun was originally reflexive. Besides it is not used in the N. T., though literary koinh, writers like Aristides, Arrian, Lucian, Polybius use it.17 Where another pronoun was desired for the third person besides that in the personal ending, various devices were used. The Attic writers usually employed a demonstrative ( o` de,├ o` me,n├ ou-toj├ evkei/noj├ oa}j de,├ o[de, etc.). The N. T. shows examples of all these constructions which will be illustrated in the discussion of the demonstrative pronouns. But the N. T. uses also auvto,j as the subject, an idiom foreign to Attic writers, but found already in Homer18 and common in the modern Greek, where indeed it has come to be itself a demonstrazive.19 Simcox20 rightly remarks that the main point to observe is not whether it has emphasis, but its appearance at all as the mere subject. All the personal pronouns in the nominative have more or less emphasis. The use of auvto,j in contrast with other persons is natural like auvto.j kai. oi` met v auvtou/ (Mk. 2:25). We are not here considering the intensive use of auvto,j as 'self' nor the use of o` auvto,j 'the same.' There is no dispute as to use of auvto,j as emphatic 'he' in the N. T. like the Pythogorean21 (Doric) auvto.j e;fa. So Ac. 20:35 auvto.j ei=pen, as much as to say 'The Master said.' Cf. the way in which some wives refer systematically to their husbands as "He." Other undoubted examples are auvto.j ga.r sw,sei to.n lao,n (Mt. 1:21). Here the emphasis is so clear that the Revised Version renders: "For he it is that shall save." In Mt. 12:50 auvto,j mou avdelfo,j is resumptive, gathering up o[stij, and is distinctly emphatic. Cf. likewise auvto.j bapti,sei, referring to o` evrco,menoj in Mt. 3:11; ov thrw/n- kai. auvto,j, 1 Jo. 3:24; oa}n a'n filh,sw auvto,j evstin, Mk. 14:44. Strong emphasis also appears in examples like; kai. auvto.j e;stin pro. pa,ntwn (Col. 1:17). In Mt. 8:24 auvto.j de, and Mk. 4:38 kai. auvto,j Jesus is the chief person in the story and the pronoun has emphasis. Cf. likewise Lu. 1:16, 17; 24:21; Mt. 16:20. In Lu. 19:2 W. H. and Nestle

680 A GRAMMAR OF THE GREEK NEW TESTAMENT

follow B in reading kai. auvto,j twice. Some emphasis is present both times. In Ac. 7:21 (Rec.) the pronoun auvto,n appears three times. As regards kai. auvth,, the editors differ between this accent and kai. au[th in Lu. 7:12; 8:42; 1 Cor. 7:12; Ro. 7:10. In Lu. 2:37; Ro. 16:2, Nestle agrees with W. H. in kai. auvth,. But in Lu. 2:37 auvth. ch,ra may be a 'widow by herself.'22 There is no real reason for objecting to the feminine use of this idiom. The plural auvtoi,, appears in Mk. 7:36; Lu. 2:50; 9:36. The only remaining question is whether auvto,j occurs in the nominative free from any emphasis just like the personal ending in a word. It is in Luke's Gospel and the Apocalypse23 that such instances occur. It is not a question whether auvto,j is so used in ancient Greek. Winer24 denies that any decisive passages have been adduced in the N. T. of such unemphatic use. Certainly the matter is one of tone and subjective impression to a large extent. And yet some examples do occur where emphasis is not easily discernible and even where emphasis would throw the sentence out of relation with the context. What emphasis exists must be very slight. Cf. Lu. 1:22; 2:50; 6:8; 8:1, 22; 15:14; 24:14, 25, 31; Rev. 14:10; 19:15. Thus we see all grades of emphasis. Abbott25 holds that in John auvto,j never means 'he,' either emphatic or unemphatic, but always 'himself.' But in Jo. 2:12 ( auvto.j kai. h` mhth.r auvtou/) there is little difference between the emphatic 'he' and 'himself.' Cf. also 18:1. But the intensive idea is clear in Jo. 4:2, 12. In 4:53 it might be either way. In the LXX we find auvto,j sometimes unemphatic. Cf. Gen. 3: 15 f.; 1 Sam. 17:42; 18:16.

(b) THE OBLIQUE CASES OF THE PERSONAL PRONOUNS.

1. Originally Reflexive. In pre-Homeric times the pronominal stem was reflexive.26 The reflexive form, as distinct from the personal pronoun, was a later development. The personal pronouns may be reflexive in Hesiod, the Homeric Hymns, Pindar and the other Lyric poets.27 Indeed, the early Attic inscriptions28 show the same thing, not to mention the Dramatic poets and Herodotus.29 It was only gradually that the distinctively reflexive form came into common use in the Attic prose, first for the third person, and

PRONOUNS ( vANTWNUMIAI) 681

then for the first and second persons.30 The use of the personal pronoun in the reflexive sense survived longest in the vernacular. It is not "abnormal" therefore to find in the N. T. (vernacular koinh,) the personal pronouns where a reflexive form might have been used. The N. T. does not here exactly represent Attic literary prose. Cf. avra,tw to.n stauro.n auvtou/ (Lu. 9:23), meta. to. evgerqh/nai, me proa,xw (Mk. 14:28; cf. Lu. 10:35), ba,le avpo. sou/, (Mt. 5:29). See Ro. 15:16, 19. It is not necessary to split hairs here as to whether the reflexive idea is present. It is in perfect harmony with the Greek history. Indeed English does not differ here from the Greek.

2. Auvtou/. The use of autou/ rather than ou- and sfw/n is noticeable. As a matter of fact, however, auvtou/ had long been the main pronoun for the oblique cases of the third person. In archaic and poetic forms the early use of ou- and sfw/n survived.31 In the N. T. auvtou/, is the only form found, as in auvtw/├ auvtoi/j├ auvto,n (Mt. 17:22 f.), ktl.

3. Genitive for Possession. The genitive of the personal pronoun is very common as a possessive rather than the possessive pronoun or the mere article. In Jo. 2:12 auvtou/ occurs twice, but once ( oi` avdelfoi,) we do not have it. These examples are so common as to call for mere mention, as o` path,r mou (Jo. 5:17), to.n kra,batto,n sougrk grk(5:8), to.n kra,batton auvtou/grk grk(5:9). The presence of the personal pronoun in the genitive is not always emphatic. Thus no undue emphasis is to be put upon auvtou/ even in its unusual position in Jo. 9:6, nor upon sou in 9:10, nor upon mou in 9:15. See chapter on The Sentence. See also evpa,raj tou.j ovfqalmou.j auvtou/ eivj tou.j maqhta.j auvtou/ (Lu. 6:20), evn th|/ u`pomonh|/ u`mw/n kth,sesqe ta.j yuca.j u`mw/n (Lu. 21: 19). See also position of you in Mt. 8:8 and Jo. 11:32. As a matter of fact the genitive of personal pronouns, as is common in the koinh, (Moulton, Prol., p. 40 f.), has nearly driven the possessive pronoun out. The use of the article with this genitive will be discussed in that chapter (The Article). Cf. to.n pate,ra mou (Mt. 26: 53) and fi,loi mou (Jo. 15:14). Both u`mw/n in Paul (1 Cor. 9:12) and auvtou/, (Tit. 3:5) may be in the attributive position. The position of auvtou/ is emphatic in Eph. 2:10 as is that of u`mw/n, in 1 Cor. 9:11 and h`mw/n in Jo. 11:48. The attributive position of h`mw/n (2 Cor. 4:16) and auvtou/ with other attributes (Mt. 27:60) is not unusual.

4. Enclitic Forms. The first and second persons singular have enclitic and unenclitic forms which serve to mark distinctions of emphasis in a general way. We may be sure that when the long

682 A GRAMMAR OF THE GREEK NEW TESTAMENT

form evmou/ occurs some slight emphasis is meant, as in u`mw/n te kai. evmou/ (Rom. 1:12). But we cannot feel sure that all emphasis is absent when the short form is used. Thus oivkodomh,sw mou th.n evkklhsi,an (Mt. 16:18), pa,nta moi paredo,qh u`po. tou/ patro,j mougrk grk(11:27). With prepositions (the "true" ones) the long form is used as in ancient Greek except with pro,j, which uniformly has me even where emphasis is obvious.32 Thus deu/te pro,j me (Mt. 11:28), kai. su. e;rch| pro,j me grk(3:14). Some editors here and in the LXX print pro.j me,. But in Jo. 6:37 pro.j evme, is the true text. Cf. pro.j evme, also in P.Tb. 421 (iii/A.D.). With sou/ the only difference is one of accent and we have to depend on the judgment of the editor. It is difficult, if not impossible, to lay down any fundamental distinction on this point. On sou and sou/ see chapter VII, iv, 4, (a). See also evxomologou/mai, soi (Mt. 11:25) and kavgw. de, soi le,gwgrk grk(16:18). Cf. evgw, se (Jo. 17:4) and me su,grk grk(17:5). Blass (Gr. of N. T. Gk., p. 168) says that evmou/ and sou/, the emphatic forms, occur only with other genitives like auvtou/ kai. evmou/ (Ro. 16:13). Simcox (Language of the N. T., p. 55) argues that the enclitic form occurs always except when there is emphasis. But the trouble is that the enclitic form seems to occur even where there is emphasis. The genitive of the third person can be used with emphasis. Cf. auvtw/n in Lu. 24:31. See further chapter VII, v, 4.

(c) THE FREQUENCY OF THE PERSONAL PRONOUNS. It is at bottom a differentiation from the substantive, though the roots are independent of verb and substantive and antedate historical evidence.33 This pronoun came into play where the sense required it. Thus kai. evpiqe,ntej ta.j cei/raj auvtoi/j avpe,lusan (Ac. 13:3). Cf. Mk. 6:5. There is no doubt of the fact that the N. T. uses the pronoun in the oblique cases more frequently than is true of the older Greek.34 What is the explanation of this fact? The Hebrew pronominal suffixes at once occur to one as the explanation of the situation and Blass accepts it.35 The LXX shows a similar "lavish use of pronouns."36 But a glance at the modern Greek reveals the same fondness for pronouns, and the papyri abundantly prove that the usage belongs to the vernacular koinh,.37 Cf. avnu,gw tou.j ovfqalmou,j mou Par.P. 51 (ii/B.C.), La,mpwni muoqhreuth|/ e;dwka auvtw|/ O.P. 299 (i/A.D.). Thumb38 suggests that this abundance of pro-

PRONOUNS ( vANTWNUMIAI) 683

nouns is natural in the vernacular. Blass39 finds "a quite peculiar and tiresome frequency" of the pronoun in the N. T. This is only true in comparison with literary Attic. The N. T. is here a natural expression of the vernacular. Thus in Lu. 6:20 note auvtou/ twice, u`mw/n twice in Lu. 21:19, sou in Mt. 6:17 as the reflexive twice ( a;leiyai, sou th.n kefalh.n kai. to. pro,swpo,n sou ni,yai%. It is not necessary to go as far as Moulton does and deny that there is any Semitic influence in the N. T. on this point. It was here in harmony with the current Greek. Cf. Lu. 24:50 for three examples of auvtou/ (- ou,j). Cf. se- se in Jo. 1:48. For auvto,= 'it' see Ro. 7: 20. In Lu. 1:62 auvto, and autou/ both refer to paidi,on.

(d) REDUNDANT. The pronoun, was sometimes redundant. This was also a Hebrew idiom, but the vernacular koinh, shows similar examples. The two streams flow together as above. With participles note tw|/ qe,lonti- a;fej auvtw|/ (Mt. 5:40), kataba,ntoj auvtou/ - hvkolou,qhsan auvtw|/grk grk(8:1), evmba,nti auvtw|/ eivj ploi/on hvkolou,qhsan auvtw|/grk grk(8:23). There are besides the anacolutha like o` nikw/n kai. o` thrw/n- dw,sw auvtw|/ (Rev. 2:26). Cf. also to. poth,rion- ouv mh. pi,w auvto, (Jo. 18:11) which does not differ radically from the other examples.40 Cf. also the redundant personal pronoun with the relative like the Hebrew idiom with the indeclinable rv,a] ou-──auvtou/ (Mt. 3:12), h-j - auvth/j (Mk 7:25), oua}j- auvtou,j (Ac. 15:17), oi-j- auvtoi/j (Rev. 7:2). But this idiom appeared also in the older Greek and is not merely Semitic.41 It occurs in Xenophon and Sophocles. Indeed in Rev. 17:9; e`pta. o;rh o[pou h` gunh. ka,qhtai evp v auvtw/n, we have o[pou in sense of relative pronoun much like modern Greek pou/. For the redundant antecedent see further under Relative.

(e) ACCORDING TO SENSE. See also chapter X, VII, VIII, The personal pronouns are sometimes used freely according to the sense. In Ac. 26:24, ta. polla, se gra,mmata eivj mani,an peritre,pei, the position of se is probably a matter of euphony and a case in point. Sometimes there is no immediate reference in the context for the pronoun. The narrative is compressed and one must supply the meaning. So with auvtou/ (Lu. 1:17), auvtoi/j (Mt. 8:4), auvtw/n (12: 9), auvtw/n, (Mt. 11:1), auvto,n (Jo. 20:15), auvtw/n (1 Pet. 3:14). But this is no peculiarity of N. T. Greek or of the koinh,. It is common at all times. In Jo. 8:44, yeu,sthj evsti.n kai. o` path.r auvtou/├ the auvtou/ refers to yeu/doj suggested by yeu,sthj) In 2 Cor. 5:19 auvtoi/j refers to ko,smon, as in Ro. 2:26 auvtou/ has in mind avkro,bustoj

684 A GRAMMAR OF THE GREEK NEW TESTAMENT

suggested by avkrobusti,a. So in Ac. 8:5 auvtoi/j refers to po,lin. In Mk. 5:4 auvth|/ follows the natural gender of paidi,on rather than the grammatical. But in Jo. 6:39 auvto, agrees grammatically with the abstract collective pa/n o[. In Lu. 6:6 we find a usage much like the original Homeric absence of the pure relative.42 We have kai. auvtou/ used with a;nqrwpoj much as ou- was. In Mt. 28:19 auvtou,j refers to e;qnh. In Mk. 6:46 auvtoi/j points to o;clon.

(f) REPETITION OF THE SUBSTANTIVE. Sometimes indeed the substantive is merely repeated instead of using the pronoun. Thus in Jo. 11:22 we have to.n qeo,n- o` qeo,j. This is usually due to the fact that the mere pronoun would be ambiguous as in the use of vIhsou/j in Jo. 4:1. Sometimes it may be for the sake of emphasis as in o` ui`o.j tou/ avnqrw,pou (Lu. 12:8) rather than evgw,. Sometimes antithesis is better sustained by the repetition of the substantive. Thus with ko,smw|──ko,smou (Jo. 9:5), a`marti,a a`marti,aj (Ro. 5:12). But this is no peculiarity of Greek.

II. The Possessive Pronouns ( kthtikai. avntwnumi,ai).

(a) JUST THE ARTICLE. It is not merely the possessive relation that is here under discussion, but the possessive pronoun. Often the article alone is sufficient for that relation. Thus in evktei,naj th.n cei/ra (Mt. 8:3) the article alone makes the relation clear. Cf. also ta.j cei/raj (Mk. 14:46), th.n ma,cairangrk grk(14:47), to.n avdelfo,n (2 Cor. 12:18). The common use of the genitive of the personal pronoun is not under consideration nor the real reflexive pronoun like e`autou/.

(b) ONLY FOR FIRST AND SECOND PERSONS. There is in the N. T. no possessive form for the third person. The other expedients mentioned above (usually the genitive auvtou/├ auvtw/n) are used. The personal pronouns are substantival, while the possessive forms are adjectival. In modern Greek no adjectival possessive exists. Just the genitive occurs (Thumb, Handbook, p. 89). The possessive evmo,j and so,j are disappearing in the papyri (Radermacher, N. T. Gk., p. 61). Originally the accent43 of evmo,j was * e;moj. The forms h`me,─teroj and u`me,─teroj are both comparative and imply emphasis and contrast, the original meaning of the comparative.44

(c) EMPHASIS, WHEN USED. When these possessive forms occur in the N. T. there is emphasis. But it is not true, as Blass45

PRONOUNS ( vANTWNUMIAI) 685

affirms, that there is no emphasis when the genitive forms are used. See I, (b), 4. The possessives do not occur often in the N. T. For details see chapter VII, iv, 4, (d).

(d) WITH THE ARTICLE. The possessives in the N. T. usually have the article save when predicate.46 Thus h` evmh, (Jo. 5:30), th/j evmh/j (Ro. 10:1), to. evmo,n, (Mt. 18:20), tw|/ sw|/ (Mt. 7:3), etc. When the article is absent the possessive is usually predicate as in ta. evma. pa,nta sa, evstin├ kai. ta. sa. evma, (Jo. 17:10; Lu. 15:31). In mh. e;cwn evmh.n dikaiosu,nhn th.n evk no,mou (Ph. 3:9) the possessive is attributive, a righteousness of my own, though the article comes later. In Jo. 4:34 we have evmo.n brw/nma, evstin i[na where the attributive use also occurs. But see Mt. 20:23. One may note u`mw/n in predicate (1 Cor. 3:21).

(e) POSSESSIVE AND GENITIVE TOGETHER. Paul's free use of the possessive and genitive together as attributives is well illustrated by to. evmo.n pneu/ma kai. to. u`mw/n (1 Cor. 16:18). In 1 Cor. 16:17 the MSS. vary between to. u`mw/n u`ste,rhma and to. u`me,teron (BCD) u`st. So in 1 Jo. 2:2 we have both peri. tw/n a`martiw/n h`mw/n and also peri. tw/n h`mete,rwn) Indeed the genitive may be in apposition with the genitive idea in the possessive pronoun. Thus th|/ evmh|/ ceiri. Pau,lou, (1 Cor. 16:21). Cf. 2 Th. 3:17; Col. 4:18; Jo. 14:24.

(f) OBJECTIVE USE. The possessive pronoun may be objective just like the genitive. This is in full accord with the ancient idiom. So th.n evmh.n avna,mnhsin (Lu. 22:19; 1 Cor. 11:24), th.n u`me─ te,ran kau,chsingrk grk(15:31), tw|/ u`mete,rw| evle,ei (Ro. 11:31), th.n h`mete,ran didaskali,angrk grk(15:4). Cf. th/j u`mw/n paraklh,sewj (2 Cor. 1:6).

(g) INSTEAD OF REFLEXIVE. The possessive, like the personal pronoun, occurs where a reflexive might have been used. Thus tw|/ sw|/ with katanoei/j in Mt. 7:3, avkou,w ta. evma. te,kna (3 Jo. 1:4), e;graya th|/ evmh|/ ceiri, (Phil. 1:19). The pronoun i;dioj is possessive, but is best treated as a reflexive.

III. The Intensive and Identical Pronoun ( su,ntonoj avntw─ numi,a). The use of auvto,j was originally "purely anaphoric."47 As the third personal pronoun it was, of course, anaphoric. The intensive use is more emphatic.

(a) THE NOMINATIVE USE OF Auvto,j. As already remarked, it is not always clear whether we have the emphatic 'he' or the intensive 'self' with auvto,j the nominative. Cf. auvto.j kai. h` mh,thr

686 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 2nd ed.

auvtou/ (Jo. 2:12). The intensive auvto,j appears in all persons, genders and numbers. Thus auvto.j evgw, (Ro. 7:25; cf. evgw. auvto,j Ac. 10:26), auvtoi. avkhko,amen (Jo. 4:42), du,nasai- auvto,j (Lu. 6:42), auvtoi. u`mei/j (1 Th. 4:9; cf. Ac. 18:15), auvto.j o` vIwa,nhj (Mt. 3:4), auvtoi. profh/tai (Ac. 15:32), auvto. to. bibli,on (Heb. 9:19), auvta. ta. evpoura,niagrk grk(9:23), auvta. ta. e;rga (Jo. 5:36). The article is not always used. Cf. auvto.j Dauei,d (Lu. 20:42), auvth. Sa,rra (Heb. 11: 11), auvtoi. profh/tai, (Ac. 15:32). Cf. evgw, de. auvto,j P.Oxy. 294 (A.D. 22). In 2 Cor. 10:1 note auvto.j evgw. Pau/loj. There is nothing particularly essential in the order whether auvto.j evgw. or evgw. auvto,j (see above). ;Egwge is not in the N. T.

(b) VARYING DEGREES OF EMPHASIS. For a list of the various shades of meaning possible with auvto,j see Thompson, Syntax of Attic Greek, p. 59 f. In Ro. 15:14 auvto,j occurs with the first person and auvtoi, with the second in sharp contrast. In Shakespeare we have "myself" as subject: "Myself have letters" (Julius Caesar, iv. 3).48 Cf. Latin ipse. In Jo. 2:24, auvto.j de. vIhsou/j, we have Jesus himself in distinction from those who believed on him. In 1 Cor. 11:14 h` fu,sij auvth, is 'nature of itself.' Note auvtoi. oi;date (1 Th. 3:3), 'ye for yourselves.' In Ac. 18:15, o;yesqe auvtoi,, we find 'ye by yourselves.' Each instance will vary slightly owing to the context. Cf. auvtoi,, (Ac. 16:37); auvto.j mo,noj $Mk. 6:47). On auvtoi. me.n ou=n see Ac. 13:4. See avf v e`autw/n (Lu. 12:57), not auvtoi,.

(c) Auvto,j WITH Ou-toj. In Ac. 24:15, 20, the classical idiom auvtoi, ou-toi occurs. Cf. eivj auvto. tou/to (Ro. 9:17), pepoiqw.j auvto. tou/to (Ph. 1:6), auvto. tou/to (2 Pet. 1:5, accusative of gen. reference). Cf. 2 Cor. 7:11. The other order is found in e;graya tou/to auvto, (2 Cor. 2:3).

(d) Auvto,j ALMOST DEMONSTRATIVE. In Luke auvto.j o` is sometimes almost a pure demonstrative as it comes to be in later Greek. The sense of 'very' or 'self' is strengthened to 'that very.' Thus auvth|/ th|/ w[ra|, (Lu. 2:38), evn auvtw|/ tw|/ kairw|grk grk(13:1), evn auvth|/ th|/ h`me,ra| (23: 12). The modern Greek freely employs this demonstrative sense. Cf. Thumb, p. 90. Moulton (Prol., p. 91) finds this demonstrative use of auvto.j o` in the papyri. So auvto.n to.n vAnta,n, O.P. 745 (i/A.D.). Moulton thinks that auvto,j is demonstrative also in Mt. 3:4. See VI, (h), for further discussion.

(e) IN THE OBLIQUE CASES. It is not so common as the nominative. So auvtoi/j toi/j klhtoi/j (1 Cor. 1:24). Cf. kai. auvtou,j in Ac. 15:27 (cf. 15:32). But examples occur even in the first and

PRONOUNS ( vANTWNUMIAI) 687

second persons. Thus evmou/ auvtou/ (Ro. 16:2), sou/ auvth/j (Lu. 2:35), auvtou.j h`ma/j (2 Th. 1:4), evx u`mw/n auvtw/n (Ac. 20:30, probable text). Here the use is intensive, not reflexive. The same thing is possible with u`mw/n autw/n in 1 Cor. 7:35 (cf. 11:13). But I think this reflexive. This intensive use of auvto,j with evmou/ and sou/ is found in Attic. In auvtw/n h`mw/n and u`mw/n only the context can decide which is intensive and which reflexive. Cf. Thompson, A Syntax of Attic Greek, p. 64. Cf. evx autvw/n tw/n nekrota,fwn, 'from the grave-diggers themselves,' P. Grenf. ii, 73 (iii/A.D.).

(f) Auvto,j SIDE BY SIDE WITH THE REFLEXIVE. So auvto.j e`autw|/ (Eph. 5:27), auvtoi. evn e`autoi/j (Ro. 8:23). Cf. 2 Cor. 1:9; 10:12. The distinctively reflexive pronouns are, of course, compounded of the personal pronouns and auvto,j. They will be treated directly. The N. T. does not have auvto,tatoj (cf. Latin ipsissimus). Some N. T. compounds of auvto,j are auvta,rkhj (Ph. 4:11), auvtokata,kritoj (Tit. 3:11), auvto,matoj (Mk. 4:28), auvto,pthj (Lu. 1:2).

(g) `O Auvto,j. The use of o` auvto,j for identity ('the same,' the very') is close kin to the original 'self' idea. Cf. ipse and idem. The idiom is frequent in the N. T. Thus o` auvto.j ku,rioj (Ro. 10:12), h` auvth. sa,rx (1 Cor. 15:39), ta. auvta.j qusi,aj (Heb. 10:11), and with substantive understood to. auvto, (Mt. 5:47), tw/n auvtw/n (Heb. 2: 14), ta. auvta, (Lu. 6:23). In 1 Cor. 11:5 we have the associative instrumental case with it, to. auvto. th|/ evxurhme,nh|. But in 1 Pet. 5:9 we actually have the genitive ('the same sort of'), ta. auvta. tw/n paqhma,twn.

IV. The Reflexive Pronoun ( avntanaklastikh. avntwnumi,a%)

a) DISTINCTIVE USE. As already explained in this chapter under Personal Pronouns, the originals of the personal pronouns in oblique cases were also reflexive.49 Only gradually the distinction between personal and reflexive arose. But even so the personal pronouns continued to be used as reflexive. Hence I cannot agree with Blass50 that evmautou/├ seautou/├ e`autou/ "have in the N. T. been to some extent displaced by the simple personal pronoun." It is rather a survival of the original (particularly colloquial) usage. Thus we have in Mt. 6:19 f. qhsauri,zete u`mi/n qhsaurou,j, 5:29 f. and 18:8 f. ba,le avpo. sou/, 6:2 mh. salpi,sh|j e;mrposqe,n sou├ 11:29 a;rate to.n zugo,n mou evf v u`ma/j, 17:27 do.j avnti. evmou/ kai. sou/, 18:15 e;legxon ) ) ) metaxu. sou/ kai. auvtou/. Matthew has rather more of these survivals. But see avfi,dw ta. peri. evme, (Ph. 2:23), to. kat v evme. pro,qumoj (Ro. 1:15). For this idiom in Attic see Thompson, Syn-

688 A GRAMMAR OF THE GREEK NEW TESTAMENT

tax of Attic Greek, p. 64. This is not indeed the classic Attic idiom, but the vernacular Attic (as in the koinh,) is not so free from it. In particular the third person presents peculiar problems, since the ancient MSS. had no accents or breathings. The abbreviated reflexive au`tou/ and auvtou/ would look just alike. It is a matter with the editors. See chapter VI, iv, (f), for details. Thus W. H. give avra,tw to.n stauro.n auvtou/ (Lu. 9:23), but ouvk evpi,steuen au`to.n auvtoi/j (Jo. 2:24). In Lu. 9:24 we have th.n yuch.n auvtou/, but in 14:26 th.n yuch.n e`autou/. In the last passage e`autou/ occurs with pate,ra and yuch.n, but not with the other words. Cf. auvtw|/, Ac. 4:32. In the light of the history of the personal pronouns the point is not very material, since auvtou/, can be reflexive also. The Attic Greek used to have dokw/ moi. But Luke in Ac. 26:9 has e;doxa evmautw|/ as Paul in 1 Cor. 4:4 says evmautw|/ su,noida. Old English likewise used the personal pronouns as reflexive. Thus "I will lay me down and sleep," "He sat him down at a pillar's base," etc.51 Cf. Ac. 19:21, me twice. See also chapter VII, Iv, 4, (c).

(b) THE ABSENCE OF THE REFLEXIVE FROM THE NOMINATIVE. It is impossible to have a reflexive in the nominative. The intensive pronoun does occur as auvto.j evgw, (2 Cor. 10:1). The English likewise, as already shown, early lost the old idiom of "myself," "himself " as mere nominatives.52 Cf. avf v e`autou/, Jo. 11:51, where auvto,j could have been employed.

(c) THE INDIRECT REFLEXIVE. It is less common in the N. T. It does indeed occur, as in the ancient Greek. So qe,lw pa,ntaj avnqrw,pouj ei=nai w`j kai. evmauto,n (1 Cor. 7:7), sunei,dhsin de. le,gw ouvci, th.n e`autou/ avlla. th.n tou/ e`te,rougrk grk(10:29). But on, the other hand, note evgw. evn tw|/ evpane,rcesqai me avpodw,sw soi (Lu. 10:35), parakalw/- sunagwni,sasqai, moi (Ro. 15:30). Cf. 2 Cor. 2:13. This on the whole is far commoner and it is not surprising since the personal pronoun occurs in the direct reflexive sense. Cf. h[n hvkou,sate, mou (Ac. 1:4). In Thucydides the reflexive form is generally used for the indirect reflexive idea.53

(d) IN THE SINGULAR. Here the three persons kept their separate forms very well. Hence we find regularly evmauto,n (Jo. 14: 21), seautw|/ (Ac. 16:28), e`autw|/ (Lu. 18:4). Indeed e`autou/ never stands for evmautou/.54 For seautou/ or seauto,n some MSS. read e`autou/ in Mk. 12:31; Jo. 18:34; Gal. 5:14; Ro. 13:9. In 1 Cor. 10:29 e`autou/='one's own' (Moulton, Cl. Rev., 1901, p. 441; Prol., p. 87). There was some tendency towards this usage in the an-

PRONOUNS ( vANTWNUMIAI) 689

cient Greek,55 though the explanation is not perfectly clear.56 But the usage is clearly found in the Atticists, Dio Chrys., Lucian and Philost. II.57 In Rev. 18:24 evn auvth|/ is a sudden change from evn soi, of the preceding verses, but is hardly to be printed au`th|/, for it is not strictly reflexive. The same58 use of auvth,n rather than se, appears in Mt. 23:37 and parallel Lu. 13:34. Cf. also Lu. 1:45. But Moulton (Cl. Rev., Dec., 1901, p. 441, April, 1904, p. 154) finds in the papyri several examples of this "un-. educated use of e`autou/ for first and second persons singular, sug─ cwrw/ meta. th.n e`autou/ teleuth,n, B.U. 86 (ii/A.D.). Radermacher (N. T. Gr., p. 61) cites evpe,graya e`autw|/ (Petersen-Luschan, Reisen etc., p. 26, n. 32). Thucydides has a few possible examples and certainly the Latin is is in point (Draeger, Historische Synt. d. Lat. Spr., p. 84). In early Greek Delbruck finds the reflexive referring indifferently to either person. The recurrence is not surprising. In the modern Greek the singular e`autou/ occurs constantly for first and second persons and even tou/ e`autou/ mou├ tou/ e`autou/ sou for emphasis. Cf. "myself," "thyself," "herself" and vulgar " hisself." See Simcox, Language of the N. T., p. 63. In translation from Semitic originals we sometimes find yuch,n rather than e`auto,n as in Lu. 9:24 (cf. Mk. 8:36). Cf. Moulton, Prol., p. 87; Robinson, Study of the Gospels, p. 114. The form au`to,n (Jo. 2:24), au`tw|/ (Lu. 12:21) is preserved in some 20 passages by W. H. and Nestle.

(e) IN THE PLURAL. Here the matter is not in any doubt. It is rather too much to say with Simcox that e`autw/n is the only form for the reflexive plural. This is indeed true for the first and third persons as avneqemati,samen e`autou,j (Ac. 23:14). In 2 Th. 1:4 autou,j h`ma/j is intensive, as already shown (chapter VII). In the third person also only e`autw/n occurs as in Mt. 18:31. In the second person plural a few examples of the reflexive u`mw/n auvtw/n apparently survive, as in Ac. 20:30; 1 Cor. 5:13 and probably so in 1 Cor. 7:35; u`mi/n auvtoi/j in 1 Cor. 11:13. But the common idiom for the second person plural is undoubtedly e`autw/n, as prose,cete e`autoi/j (Lu. 17:3). Cf. Mt. 25:9; Ro. 6:13; 1 Jo. 5:21, etc. There are some seventy examples of e`autw/n for first and second persons plural in the N. T. (Moulton, Prol., p. 87), as is the custom in the papyri, chiefly in illiterate documents. Cf. i[na geinw,meqa pro.j toi/j kaq v e`autou,j, Tb.P. 6 (ii/B.C.); i[na komisw,meqa ta. e`autw/n, Tb.P. 47.

690 A GRAMMAR OF THE GREEK NEW TESTAMENT

The LXX (Conybeare and Stock, Sel., p. 30) has this use of e`autw/n for first and second persons plural. We even find reflexive and personal together like u`mi/n e`autoi/j (Ex. 20:23).

(f) ARTICLE WITH. The reflexive is used with or without the article and in any position with the article. But curiously enough seautou/ is never so found and evmautou/ only once in sharp contrast, mh. zhtw/n to. evmautou/ su,mforon avlla. to. tw/n pllw/n (1 Cor. 10:33). Instead of this reflexive genitive (possessive) we have the genitive of the personal pronoun. Cf. timw/ to.n pate,ra mou (Jo. 8:49), a;fej to. dw/ro,n sou (Mt. 5:24). The examples of e`autou/ are, of course, abundant as in th.n e`autou/ auvlh,n (Lu. 11:21), the common idiom in the older Greek. But note also the order to. e;rgon e`autou/ (Gal. 6:4), e`autou/ tou.j po,daj (Ac. 21:11), dou,louj e`autou/ (Lu. 19:13), kh/pon evautou/ (Lu. 13:19). These are all attributive, but the sense is not quite the same in the two last. The use of auvtou/ in such examples has already been noted as in Mt. 16:24. Sometimes the MSS. vary between e`autou/ and auvtou/ as in Lu. 4:24. The plural e`autw/n is likewise found thus, tou.j e`autw/n nekrou,j (Mt. 8:22), tw|/ kuri,w| e`autw/n (Mt. 18:31), e`autw/n ta. i`ma,tia (Mt. 21:8). See further chapter XVI. The Article.

(g) REFLEXIVE IN THE RECIPROCAL SENSE. This use of e`autw/n does not really differ in idea from avllh,lwn. This is in harmony with the ancient Greek idiom. The papyri show this same blending of e`autw/n with avllh,lwn.59 Cf. P.P. 8 (ii/B.C.) three times, O.P. 260 (i/A.D.), C.P.R. 11 (ii/A.D.) twice. Thus we may note o[ti kri,mata e;cete meq v e`autw/n (1 Cor. 6:7), lalou/ntej e`autoi/j (Eph. 5:19), nouqetou/n─ tej e`autou,j (Col. 3:16), etc. Sometimes it occurs side by side with avllh,lwn as if by way of variety, as in avneco,menoi avllh,lwn kai. cari─ zo,menoi e`autoi/j (Col. 3:13). Cf. also avllh,lwn and au`tou,j in Lu. 23:12. In Ph. 2:3 avllh,louj h`gou,menoi u`pere,contaj e`autw/n, each word retains its own idea.

(h) REFLEXIVE WITH MIDDLE VOICE. Sometimes indeed the reflexive occurs with the middle voice where it is really superfluous, as in diemeri,santo e`autoi/j (Jo. 19:24, LXX), where60 Mt. 27: 35 (free paraphrase of LXX) has only diemeri,santo. So also seauto.n pareco,menoj (Tit. 2:7). But usually such examples occur where the force of the middle is practically lost, as in h[ghmai evmauto,n (Ac. 26:2), avrnhsa,sqw e`auto,n (Lu. 9:23). On the use of the reflexive in Anglo-Saxon see Penny, A History of the Reflexive Pronoun in the English Language, p. 8. Cf. paralh,myomai pro.j evmauto,n (Jo. 14:3). Moulton (Prol., p. 87) admits that sometimes

PRONOUNS ( vANTWNUMIAI) 691

e`autou/ occurs without great emphasis. This use of the reflexive with the middle may be compared with the reflexive and the personal pronoun in the LXX. So lh,myomai evmautw|/ u`ma/j lao.n evmoi, (Ex. 6:7), ouv poih,sete u`mi/n e`autoi/jgrk grk(20:23). So English "me myself," "you yourselves." Cf. Thackeray, p. 191. See further chapter XVII, Voice.

(i) THE USE OF ;Idioj. This adjective is frequent in the N. T. It is usually treated as a possessive, opposed61 to koino,j or dhmo,sioj. In the N. T. we find it, especially (17 times) in kat v ivdi,an (cf. Lu. 9:10), in the sense of 'private.' So this sense occurs also in Ac. 4:32 and Heb. 7:27. Cf. ivdiw/tai in Ac. 4:13 (1 Cor. 14:16). Sometimes also the word implies what is peculiar to one, his particularity or idiosyncrasy, as 1 Cor. 3:8; 7:7 (cf. the classic idiom). Cf. our "idiot." But in general o` i;dioj or i;dioj without the article (cf. e`autou/) means simply 'one's own,' a strong possessive, a real reflexive. To all intents and purposes it is interchangeable in sense with e`autou/. The examples of this reflexive idea are many. Thus in Mt. 9:1; Lu. 6:41; 10:34; Jo. 1:41; 4:44, etc. The use of oi` i;dioi for 'one's own people' (cf. also of oivkei/oi, 1 Tim. 5:8, classic idiom) is not strange. Cf. Jo. 1:11; 13:1, etc. Moulton62 finds the singular in the papyri as a term of endearment. The use of ta. i;dia for 'one's home' (Jo. 1:11; 19:27; Ac. 21:6) is seen also in the papyri. Moulton (Cl. Rev., 1901, p. 440) cites ta. i;dia, B.U. 86 (ii/A.D.), 183 (i/A.D.), 168 (ii/iii A.D.) bis, etc. The papyri also illustrate Jo. 1:11, oi` i;diiioi, for 'one's relations.' So pro.j tou.j ivdi,ouj, B.U. 341 (ii/A.D.). Examples without the article are despo,taij ivdi,oij (Tit. 2:9), kairoi/j ivdi,oij (1 Tim. 6:15). Cf. o` i;dioj lo,goj, B.U. 16 (ii/A.D.). Moulton, Cl. Rev., 1901, p. 440. In Jo. 1:41 Moulton63 rightly agrees with Westcott in seeing in to.n i;dion an implication that some one else went after his brother also. The only other point that here calls for remark is the question whether o` i;dioj is used in an "exhausted" or unemphatic sense. Blass64 finds it so in eivj to.n i;dion avrgro,n (Mt. 22:5). Meisterhans (p. 235) finds a few examples in the Attic inscriptions and Deissmann finds the weakened use of i;dioj in the literary koinh,. Deissmann65 argues further that this exhausted sense may be assumed in the N. T. because some examples in the LXX (Job 24:12; Prov. 27:15), etc., seem to occur. Moulton66

692 A GRAMMAR OF THE GREEK NEW TESTAMENT

finds that the papyri do not support this contention. Emphasis is beyond dispute in most of the N. T. instances like Mt. 9:1; Lu. 6:41; Jo. 1:41; Ac. 1:25; Gal. 6:5, etc. Moulton (Prol., p. 89) refers with point to Ro. 14:5, evn tw|/ ivdi,w| noi>, as showing i;dioj the equivalent of e`autou/. The N. T. passages may be assumed to show emphasis in spite of the later Byzantine i;dioj mou (cf. e`autou/ you in modern Greek). Moulton67 agrees with the Revisers in using 'own' in Mt. 22:5 as a "counter-attraction." The only difficult passage is Ac. 24:24 where B may be wrong. But is it not possible that ivdi,a| may have a covert hint at the character of Drusilla? For the present she was with Felix. In Tit. 1:12 note i;dioj auvtw/n profh,thj. Moulton (Cl. Rev., 1904, p. 154) cites h`mw/n i;dion, Ch. P. 4 (ii/A.D.), i;dion auvtou/ N. P. 25 (ii/A. D.), and eivj ivdi,an mou crei,an, B.U. 363 (Byz., Moulton, Cl. Rev., 1901, p. 440). In modern Greek o` i;dioj╩ o` auvto,j (Thumb, Handb., p. 97) or 'self,' evgw. o` i;dioj, 'I myself.' Cf. th/i auvth/i in the papyrus of Eudoxus (ii/B.C.), but Moulton (Prol., p. 91) observes that it does not occur in the N. T. in this sense.

V. The Reciprocal Pronoun $h` avmoibai,a avntwnumi,a). The use of the reflexive in the reciprocal sense has just been discussed (cf. personal pronouns as reflexive). From one point of view it might seem hardly necessary to give a separate discussion of reciprocal pronouns. But, after all, the idea is not exactly that of the mere reflexive. vAllh,lwn is, of course, reduplicated from a;lloj, one of the alternative pronouns. Cf. the Latin alias and alter alters. The Latin idiom is common in the classic Greek and is found in Ac. 2:12, a;lloj pro.j a;llon le,gontej* 19:32, a;lloi a;llo ti e;krazon; 21:34, a;lloi a;llo ti evpefw,noun. Cf. in the papyri a;llo evgw,├ a;llo pa,ntej, B.U. 1079 (A.D. 41). But the true reciprocal avllh,lwn has no nominative and is necessarily plural or dual (in older Greek). It occurs 100 times in the N. T. (W. H.) and is fairly well distributed. We have examples of the genitive (Ro. 12:5 avllh,lwn me,lh), the ablative (Col. 3:13 avneco,menoi avllh,lwn), the accusative (1 Cor. 16: 20 avspa,sasqe avllh,louj, 1 Jo. 4:7 avgapw/men avllh,louj), the locative (Ro. 15:5 evn avllh,loij), the dative (Gal. 5:13 douleu,ete avllh,loij). The prepositions are used 48 times with avllh,lwn. This pronoun brings out the mutual relations involved. In 1 Th. 5:11, para─ kalei/te avllh,louj kai. oivkodomei/te ei-j to.n e[na, note the distributive explaining the reciprocal. Moulton (Prol., p. 246) compares the modern Greek o` e[naj to.n a;llon. In Ph. 2:3 note both avllh,louj and e`autw/n. In 1 Th. 5:15 we have eivj avllh,louj kai. eivj pa,ntaj.

PRONOUNS ( vANTWNUMIAI) 693

In 2 Th. 1 : 3 note e`no.j e`ka,stou and eivj avllh,louj. The N. T. does not, like the LXX (Ex. 10:23), use avdelfo,j as a reciprocal pronoun. The middle voice is also used in a reciprocal sense as in. sunebouleu,santo (Mt. 26:4). Cf. chapter XVII, Voice.

VI. Demonstrative Pronouns ( deiktikai. avntwnumi,ai).

(a) NATURE. Curiously enough the demonstrative pronoun, like all pronouns, has given the grammarians a deal of trouble to define. For a discussion of the various theories during the ages see Riemann and Goelzer.68 Originally all pronouns were "deictic," "pointing." The "anaphoric" use came gradually.69 Indeed the same pronoun often continued to be now deictic, now anaphoric, as o[j, for instance, originally demonstrative, but later usually relative. Indeed the anaphoric use blends with the relative. Monro70 marks out three uses of pronouns, not three kinds of pronouns. The "deictic" "marks an object by its position in respect to the speaker." Thus evgw,├ su,├ o[de├ ou-toj├ evkei/noj all fall under this head. The "anaphoric" pronoun "is one that denotes an object already mentioned or otherwise known." Thus the resumptive use of o[de├ ou-toj├ evkei/noj├ o[j├ o[stij. The "relative" in the modern sense would be only o[j├ o[stij├ oi-oj├ o[soj, etc. As a matter of fact, for practical purposes the two Greek terms "deictic" and "anaphoric" may be placed beside the Latin "demonstrative" and "relative." See further chapter VII, iv, 4, (e) .

(b) DIFFERENT SHADES OF MEANING. The demonstrative pronouns do not indeed always have the same shade of meaning. They may point out, as far or near $o[de├ ou-toj├ evkei/noj%, as in apposition ( evkei/noj), as well known ( evkei/noj), as already mentioned (resumptive ou-toj├ o[de%.71 These uses belong to the various demonstratives and will come out in the context. I do not care to press the parallel with the personal pronouns (first, second, third person demonstratives) as applied to o[de├ ou-toj├ evkei/noj. The pronouns had best be treated separately, not according to the special uses.

(c) `O├ h`├ to,. This was the simplest demonstrative.72 The grammarians73 call this word a;rqron protaktiko,n as distinct from o[j which is a;rqron u`potaktiko,n. As a matter of fact o`├ h`├ to, is the same word as the Sanskrit sa (sas), sa, tad.74 The Lithuanian nominative sing-

694 A GRAMMAR OF THE GREEK NEW TESTAMENT

ular was ta-s, ta, and the Greek nominative plural oi`├ ai` came "instead of toi,├ tai,," (Brugmann, Comp. Gr., vol. III, p. 327). This form, like der in German and this in English, was used either as demonstrative, article or relative. See Kuhner-Gerth, I, p. 575. One is not to trace actual historical connection between o` and der (cf. Brugmann, Griech. Gr., p. 559). Its old use was a sort of personal demonstrative (cf. su. de, in Lu. 1:76).75 Cf. also su. de. ti, and it h' kai. su. ti, (Ro. 14:10) and su. ti,jgrk grk(14:4). Cf. Brugmann, Griech. Gr., p. 428. This substantival use is the main one in Homer.76 Indeed, as a demonstrative it means rather contrast than far or near like o[de├ ou-toj├ evkei/noj, but after all o[de is nothing but o` with the ending - de. The demonstrative use of o` is seen in tou.j o[soi in Agathias77 and tw/n o[sa in Maximus of Tyre.78 This demonstrative as antecedent to the relative ( tou.j oi[) appears in Justin Martyr79 and Tatian's Oration to the Greeks.80 Plato shows a good many examples81 (like to.n o[j├ to.n o[soj). We meet in Xenophon and Demosthenes82 kai. to,n as demonstrative, especially to.n kai. to,n├ to. kai. to,├ ta. kai. ta,. The modern Greek uses tou/├ th/j├ tw/n├ etc., as short forms of auvtou/, etc., and Jebb83 pertinently asks if this is not "a return to the earliest use of o`├ h`├ to, as a pronoun." The demonstrative o` is frequent in the comic writers. Cf. Fuller, De Articuli in Antiquis Graecis Comoedus Usu, p. 9. Volker (Syntax, p. 5) gives papyri illustrations of demonstrative o` $o` de,├ tou/ de, pro.j tou/├ pro. tou/├ ta. me,n├ ta. de, etc.)." The oblique cases have only two examples in the N. T., one a quotation from Aratus, tou/ kai, (Ac. 17:28), the other tou.j me,n├ tou.j de, (Eph. 4:11), where contrast exists. It is possible indeed that to,n in Ph. 1:11 is demonstrative. Cf. also to.n avp v avrch/j in 1 Jo. 2:13 and th,n in 1 Cor. 10:29. In Mt. 14:2 (Mk. 6:14) ai` is nearly equivalent to 'these.' In Mk. 12:5 the correct text is oua}j me,n, etc. But in the nominative the examples of this demonstrative in the N. T. are quite numerous. There are three uses of the nominative in the N. T. (1) One is the demonstrative pure and simple without any expressed contrast. So oi` de. evra,pisan (Mt. 26:67), oi` de. evdi,─ stasan (Mt. 28:17). In Mt. 26:57 oi` de. krath,santej we may have

PRONOUNS ( vANTWNUMIAI) 695

Addenda 3rd ed.

this usage or merely the article. In Acts we often have oi` me.n ou=n in this sense, usually with the participle (Ac. 1:6; 8:4, 25). But even in these examples there is apparently an implied contrast. In Mt. 16:14 and Lu. 9:19 the use of oi` de, (3, below) refers to those already mentioned in an oblique case. (2) The use of o` me,n├ o` de,, etc. This is no longer very frequent in the N. T.84 So o` me.n ou[twj├ o` de. ou;twj (1 Cor. 7:7); oi` me,n├ o` de, (Heb. 7:20, 23); oi` me,n├ oi` de, (Ac. 14:4); oi` me,n├ a;lloi de,├ e[teroi de, (Mt. 16:14 f:). In Mt. 13:23 we most likely have oa} me,n├ oa} de., not o` me,n├ o` de,. Cf. oa} me,n (Lu. 8:5). In Ac. 17:18 note tinej├ oi` de, and in Ro. 14:2 oa}j me,n├ o` de,. (3) The most common use of the demonstrative is where o` de,├ h` de,├ oi` de, refer to persons already mentioned in an oblique case. Thus in Mt. 2:5 oi` de, refers to par v auvtw/n. So in of oi` de, (Lu. 23:21) the reference is to auvtoi/j, while o` de, in the next verse points to auvto,n. In Mk. 14:61 o` de, refers to vIhsou/n, as in Ac. 12:15, h` de, to auvth,n. In Lu. 22:70 o` de, has no antecedent expressed, but it is implied in the ei=pan pa,ntej before.

(d) [Oj. The grammarians call it a;rqron u`potaktiko,n or relative.85 It did come to be chiefly relative, as already the Sanskrit yas, ya, yad has lost its original demonstrative force.86 But in the Lithuanian j-i-s Brugmann (Comp. Gr., III, p. 332) finds proof that the pro-ethnic i-o was demonstrative as well as relative. Cf. also i[─na in Homer- both 'there' and 'where' and then 'that.' In Homer o[j, like w[j $w`j), is now demonstrative, now relative, and was originally demonstrative.87 This original demonstrative sense eontinues in Attic prose, as in the Platonic h= d v o[j* kai. o[j* oa}n me,n├ oa}n de, etc.88 However, it is not certain that the demonstrative use of o[j ( kai. o[j├ h= d v o[j) is the same word as the relative. Brugmann89 indeed finds it from an original root, *so-s like Sanskrit sa-s. The examples of this demonstrative in the nominative are few in the N. T. Thus note in Jo. 5:11 (correct text) oa}j de. avpekri,qh, and also oa}j de. ouvk e;laben, in Mk. 15:23. Indeed oa}j dh, in Mt. 13:23 is close to the same idea. But this verse furnishes a good example of this demonstrative in contrast, oa} me.n e`kato.n oa} de. e`xh,konta oa} de. tria,konta. This example happens to be in the accusative case (cf. Ro. 9:21), but the nominative appears also as in aa} me.n e;pesen (Mt. 13:4), oa}j me.n eivj to.n i;dion avgro,n├ oa}j de, evpi, th.n evmpori,an (Mt. 22:5), oa}j me.n pisteu,ei (Ro. 14:2), o[j me.n ga.r kri,nei- o[j de. kri,nei,grk grk(14:5). So 1 Cor. 11: 21.

696 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

Addenda 3rd ed.

Instances of other cases occur also. I see no adequate reason for refusing to consider oa}n me.n e;deiran├ oa}n de. avpe,teinan├ oa}'n de. evliqobo,lhsan (Mt. 21:35) examples of the demonstrative o[j.90 Cf. Lu. 23:33. In the accusative plural note oua}j me,n├ oua}j de,├ Mk. 12:5; Ac. 27:44; Ju. 22 f. For the dative singular, w|- me,n├ w|- de,├ note Mt. 25:15. In 1 Cor. 12:8 we have w|- me,n ├ a;llw| de,├ ktl. For the dative plural see oi-j me,n├ oi-j de,├ 2 Cor. 2:16. In 1 Cor. 12:28 we have oua}j me,n as demonstrative without any corresponding as oua}j de,) Cf. oi` me.n ou=n in Ac. 8:4, 25; 11:19; 15:3, 30, and o` me.n ou=n in Ac. 23:18 as above in (c). The relative at the beginning of sentences or paragraphs, like evn oi=j in Lu. 12:1 (cf. avnq v w-n verse 3), may indeed at bottom be a reminiscence of the old demonstrative. Cf. Latin and English usage. The demonstrative is often used to connect sentences, as in Mt. 11:25; 12:1; Mk. 8:1, etc. Cf. Blass, Gr. of N. T. Gk., p. 276. In Mt. 26:50, evf v oa} pa,rei, we may also have an instance of the demonstrative. But we do not have in the N. T. kai. o[j├ kai. to,n├ to.n kai. to,n├ pro. tou/) Radermacher (N. T. Gr., p. 62) finds demonstrative o[sde in an inscription in Heberdey-Wilhelm, Reisen. N. 170.

(e) [Ode. Brugmann91 finds the enclitic - de the same that we have in de─u/ro├ dh,├ iv─de,, (?), Latin quan-de. It corresponds to the Latin hic, German der hier, English this here. It refers to what is "immediately near" in space or time,92 and is of relatively more importance than ou-toj. As a matter of fact o[de occurs only ten times in the N. T. In the LXX " o[de is much commoner than in the N. T." (Thackeray, Gr. of the O. T. in Gk., vol. I, p. 191), especially in the more literary parts. For its rarity in papyri and inscriptions see Mayser, Gr., etc., p. 308. It is already failing in the first century B.C. (Radermacher, N. T. Gr., p. 62). For ta,de see chapter VII, iv, 4, (e). In Lu. 16:25 w-de is the correct text. In Ac. 15:23 ta,de is not well supported and in 2 Cor. 12:19 ta. de, is right. In one of the remaining examples, th|/de h=n avdelfh, (Lu. 10: 39), Blass93 bluntly calls it "not even used correctly," a rather curt judgment. But he cites the LXX (Gen. 25:24; 38:27). In Winer-Schmiedel94 this example is not considered as o[de used for ou-toj but rather; like the classic o[de evgw,├ oi[de h`mei/j (cf. Ex. 8:25; Gen. 50:18). In Jas. 4:13, poreuso,meqa eivj th,nde th.n po,lin, it is hardly necessary to take th,nde as like the classical th.n dei/na or th.n kai. th,n (cf. Plato), though that is a possible construction. Cf.

PRONOUNS ( vANTWNUMIAI) 697

poih,somen tou/to h' evkei/no in verse 15. Plutarch95 seems to use th,nde in this sense. More likely in James th,nde merely means 'this' city which the enterprising Jew exploits for a year before he passes on to the next.

(f) Ou-toj. Of doubtful etymology, possibly an original root u.96 With this combine o`├ h`├ to,╩ouv├ au`├ tou. Then add to─j├ ta$h%├ to. In reality, therefore, ou-toj is a doubled demonstrative (combination of so and to, Giles, p. 296). It is like the Latin is-te (double also). Ou-toj is more often anaphoric than deictic.97 In Homer98 it (deictic) expresses an object present to the speaker, but not near him. The word is limited in use in Homer and usually refers to what is previously mentioned (anaphoric).99 It is very common in the N. T. and on the whole the usage accords with that of the older Greek. Naturally there is much diversity in the context.

1. The Purely Deictic. This use is not wanting. Thus in Mt. 3:17, ou-to,j evstin o` ui`o,j mou, the demonstrative identifies the one present as the Son of God. For further examples of the purely deictic use see Mt. 12:23; 17:5; 21:10 f. (a particularly good illustration); 21:38; 27:37, 47, 54; Mk. 6:3; 15:39; Lu. 4:22; 8:25, etc. But a still plainer example is in Jo. 21:21, when Simon pointed to John as ou-toj de. ti,.

2. The Contemptuous Use of ou-toj. It is merely one variation of the purely deictic idiom due to the relation of the persons in question. It is rather common in the N. T. So in Mt. 26:61 ou-toj e;fh we find a "fling" of reproach as the witnesses testify against Jesus. Cf. Mt. 26:71 (parallel Lu. 22:56 kai. ou-toj), the maid about Peter; Mk. 2:7, the Pharisees about Jesus; Lu. 15:2; Jo. 6:42; 9:24; 12:34; Ac. 7:40, Jews about Moses; 19:26; 28: 4, about Paul; Lu. 15:30, the elder son at the younger; 18:11, the Pharisee at the publican, etc. A striking example occurs in Ac. 5:28.

3. The Anaphoric Use. The pronoun here refers to one previously mentioned, as in Mt. 27:58 where ou-toj alludes to vIwsh,f in verse 57, where note the anacoluthon. So in Heb. 7:1 ou-toj points to the mention of Melchizedek in the preceding verse. There are many variations in the anaphoric idiom. The simplest is the one already mentioned, where the subject of discussion is merely continued by ou-toj, as in Mt. 3:3 (cf. the Baptist in verse 1). In particular observe kai. ou-toj, as in Lu. 8:41; 16:1. In Lu. 22:59

698 A GRAMMAR OF THE GREEK NEW TESTAMENT

kai. ou-toj is rather deictic. A striking example of the continuative ou-toj occurs in Ac. 7:35, 36, 37, 38, 40. Here the pronoun is repeated as often as is desired. So Jo. 6:42. Cf. the use of the pronoun because of prolepsis (Ac. 9:20). The more frequent use is the resumptive or epexegetical use which is rather more abundant in the N. T.100 Here ou-toj is really in apposition. In Ro. 7:10, h` evntolh. h` evij zwh.n au[th eivj qa,naton, we seem to have the resumptive use with a substantive. But a clear example (different in number and gender)101 occurs in Mt. 13:38, to. de. kalo.n spe,rma├ ou-toi, eivsin. One may note a similar use of evkei/noj (Jo. 12:48; 16:13) and of auvto,j (Jo. 12:49). Another plain instance is in Ac. 2:23, where tou/ton refers to vIhsou/n (verse 22). Cf. also tou/ton (2d) in Ac. 7:35. In Ac. 4:10 evn tou,tw| is resumptive referring to the preceding substantive followed by two relative clauses, while ou-toj is deictic. In verse 11 again ou-toj is continuative. In Ro. 9:6, oi` evx vIsrah,l├ ou-toi (cf. Gal. 3:7), the resumptive use is plain. The participle before ou-toj is a very common idiom, as o` de. u`pomei,naj eivj te,loj ou-toj (Mt. 10:22; 24:13); o` evmba,yaj met v evmou/ - ou-toj grk(26:23). Cf. 1 Cor. 6:4; Lu. 9:48; Jo. 7:18, etc. The participle, of course, often follows ou-toj, not resumptive, as in Jo. 11:37. The relative is followed by resumptive ou-toj as in oa}j d v a'n avpole,sh|── ou-toj (Lu. 9:24), oa} qe,lw tou/to pra,ssw (Ro. 7:15 f., 20). So Mt. 5: 19; Mk. 6:16; Ac. 3:6; Gal. 5:17; 6:7; 2 Tim. 2:2. The plural is seen in Jo. 8:26, a;──tau/ta; also in Ph. 4:9. For a[tina ──tau/ta see Ph. 3:7, and o[soi─ou-toi Ro. 8:14; Gal. 6:12; Ph. 4:8. Cf. Winer-Schmiedel, p. 218. See o[tan- to,te├ kaqw,j- tau/ta (Jo. 8:28). In Ph. 1:22 tou/to resumes to. zh/n. In 2 Th. 3:14 tou/ton is resumptive with ei; tij as in Jas. 1:23. Cf. also 1 Cor. 8:3; Ro. 8:9; Jas. 3:2.102 For eva,n tij see Jo. 9:31. Sometimes only the context can clear up the exact reference of the anaphoric ou-toj. So in Ac. 8:26 au[th points to h` o`do,j.

4. In Apposition. See also chapter X, ix. Ou-toj itself may be expanded or explained by apposition. The simplest form of this construction is where a substantive103 is in apposition as in 2 Cor. 13:9, tou/to kai. euvco,meqa├ th.n u`mw/n kata,rtisin, where agreement in gender does not occur. Cf. the nominative h` pi,stij in 1 Jo. 5:4. Cf. 1 Th. 4:3. Ou-toj is, of course, the antecedent of the relative o[j, as in Mt. 11:10; Jo. 7:25; tou/to o[ in Jo. 16:17. In

PRONOUNS ( vANTWNUMIAI) 699

Ph. 2:5 note tou/to- oa} kai,. Sometimes a clause is in apposition with ou-toj which may be either nominative or in an oblique case. Thus with o[ti we have the nominative (with feminine predicate noun), as in au[th evsti.n h` kri,sij o[ti (Jo. 3 : 19). Cf. 1 Jo. 1:5; 5:9, 11, 14. In Mk. 4:41, ti,j a;ra ou-to,j evstin o[ti, the o[ti is almost equal to w[ste. The accusative with o[ti we have in tou/to o[ti (Ro. 2:3; 6:6; Lu. 10:11; Ac. 24:14; 1 Cor. 1:12; 15:50; 2 Cor. 5: 14; 10:7, 11; 2 Th. 3:10; Ph. 1:6 ( auvto. tou/to), 25; 1 Tim. 1: 9; 2 Tim. 3:1; 2 Pet. 1:20; 3:3, 8. Cf. also dia. tou/to o[ti in Jo. 12:39.104 In Gal. 3:17, after tou/to le,gw, we have the direct discourse without recitative o[ti, but the quotation is really in the accusative in apposition with tou/to. Cf. also Lu. 12:18, tou/to poih,sw\ kaqelw/ mou ta.j avpoqh,kaj, and Jo. 4:17. The genitive with o[ti appears in peri. tou,tou o[ti (Jo. 16:19). The locative appears in evn tou,tw| o[ti, 1 Jo. 4:9, 10, 13. Cf. evn tou,tw| o[ti (Jo. 16:30; 1 Jo. 3:19, 24) in a slightly different sense where o[ti, is really the accusative. But in general these substantive clauses have the same case as tou/to.

Closely allied to this use of o[ti is that of i[na. Thus the nominative, po,qen moi tou/to i[na e;lqh|, occurs in Lu. 1:43. In Jo. 17:3, au[th de, evstin h` aivw,nioj zwh. i[na, the pronoun is feminine because of the predicate substantive. Cf. Jo. 15:12; 1 Jo. 3:11, 23; 5:3; 2 Jo. 1:6. The accusative as the direct object of the verb is seen in tou/to proseu,comai i[na in Ph. 1:9. Cf. also tau/ta- i[na, Jo. 15:11, 17; 1 Jo. 5:13. The feminine substantive occurs in the accusative also, as in tau,thn th.n evntolh.n e;comaen avp v autou/├ i[na, 1 Jo. 4:21. The accusative is found also with prepositions. So eivj tou/to├ i[na, Ac. 9:21; Ro. 14:9; 2 Cor. 2:9; 1 Pet. 3:9; 4:6; 1 Jo. 3:8. In Eph. 6:22 we have eivj auvto. tou/to i[na. Cf. Col. 4:8. Likewise note dia. tou/to├ i[na in 2 Cor. 13:10; 1 Tim. 1:16; Phil. 1:15. In 2 Cor. 2:3, e;graya tou/to auvto. i[na, we probably have the direct accusative, though tou/to auvto, could be adverbial accusative, 'for this very reason.' The locative appears in evn tou,tw| evdoxa,sqh i[na, Jo. 15:8. Cf. 1 Jo. 4:17. The ablative case appears in Jo. 15:13, mei,zona tau,thj avga,phn ouvdei.j e;cei├ i[na) In 3 Jo. 1:4 the ablative plural is found, meizote,ran tou,twn- i[na. The apposition in these various constructions varies in degree of directness. An example of o[pwj with eivj auvto. tou/to occurs in Ro. 9:17 quoted from the LXX (Ex. 9:16). Cf. also stello,menoi tou/to mh. in 2 Cor. 8:20.

In 1 Pet. 2:19 note also the use of eiv with tou/to (though ca,rij

700 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

is predicate), tou/to ga.r ca,rij eiv. Here the eiv clause is in the same case as tou/to, nominative. So in 1 Jo. 2:3 we have eva,n in apposition with evn tou,tw| (locative).

In 1 Jo. 5:2 the correct text has o[tan in similar apposition with evn tou,tw|. The infinitive also occurs in apposition with tou/to) In Heb. 9:8 the perfect infinitive in indirect discourse with the accusative is in apposition to tou/to which is itself accusative, tou/to dhlou/ntoj tou/ pneu,matoj tou/ a`gi,ou├ mh,pw pefanerw/sqai th.n ktl. In Eph. 4:17 likewise mhke,ti peripatei/n, in apposition to tou/to (after le,gw, is in indirect discourse, though here it is indirect command, not indirect assertion. But in 1 Cor. 7:37 threi/n th.n e`autou/ parqe,non is merely explanatory of tou/to kekriken. The same thing is true in 2 Cor. 2:1, where the article is added to the infinitive which is also in the accusative, e;krina evmautw|/ tou/to├ to. mh.──evlqei/n) In Ac. 26:16 the infinitive proceiri,sasqai is in the accusative like eivj tou/to. Cf. ou[twj, 1 Pet. 2:15. The nominative infinitive in Jas. 1:27 is in apposition with au[th $qrhskei,a kaqara. ──au[th├ evpi─ ske,ptesqai). So also note ou[twj evsti.n to. qe,lhma tou/ qeou/ - fimoi/n in 1 Pet. 2:15.105 Cf. Ro. 1:12 where tou/to- sunparaklhqh/nai, are merely subject and predicate. In 2 Cor. 7:11 the nominative infinitive, to. luphqh/nai, occurs with auvto. tou/to. Indeed in Mk. 12: 24 the causal participle is really explanatory of tou/to $dia. tou/to plana/sqe├ mh. eivdo,tej. It is possible to see a similar example106 in Lu. 8:21, avdelfoi, mou ou-toi, eivsin oi` ──avkou,ontej. Here in truth ou-toi seems unnecessary.

5. Use of the Article. The article commonly occurs with the noun when the noun is used with ou-toj. The noun is by no means always necessary with ou-toj. See 6. Indeed the resumptive dem. alone is often sufficient, as in Jo. 1:2, 7, etc. So auvtoi. ou-toi (Ac. 24:15, 20). In a sense a double demonstrative thus occurs, since the article was originally demonstrative. This is in exact accord with classic usage and calls for no special comment, except that it is an idiom foreign to Latin and English. The modern Greek preserves this idiom with the demonstrative. So tou,th h` gunai,ka├ auvto.j o` a;ndraj (Thumb, Handb., p. 92). It is immaterial whether ou-toj comes first, as ou-toj o` telw,nhj (Lu. 18:11), or last, as o` a;nqrw─ poj ou-toj (Lu. 23:47). Cf. Jo. 9:24. When an adjective is used with the substantive, then the article may be repeated with the adjective, as h` ch,ra au[th h` ptwch, (Mk. 12:43), or ou-toj may, like the adjective, be brought within the rule of the article. So ti,j h`

PRONOUNS ( vANTWNUMIAI) 701

Addenda 3rd ed.

kainh. au[th [ h` ] u`po. sou/ laloume,nh didach, (Ac. 17:19).107 Even if the second article be admitted here, the point made still applies. The position of ou-toj with the article, ou-toj o` rather than o` ou-toj, does not mean simply the predicate idea, though the position is predicate. But not so th.n evxousi,an tau,thn a[pasan in Lu. 4:6. Here the real predicate notion appears. In Kuhner-Gerth (I, p. 628) the explanation is given that it is either apposition ( ou-toj o` avnh,r╩ 'this, the man') or predicative sense ( o` avnh.r ou-toj= 'the man here'). Probably so, but in actual usage the connection is much closer than that. See Lu. 15:24, ou-toj o` ui`o,j mou. Cf. the French idiom La Republique Francaise. Gildersleeve (Syntax, p. 324) takes the predicate explanation. See also chapter XVI, The Article.

6. Article Absent. The article does not always occur with substantives when ou-toj is used. When ou-toj occurs with proper names in the N. T., the article is present. So Ac. 1:11 ou-toj o` vIhsou/j, 19:26 o` Pau/loj ou-toj, 7:40 o` ga.r Mwu?sh/j ou-toj, 2:32 tou/ton to.n vIhsou/n, Heb. 7:1 ou-toj ga.r o` Melcisede,k, except in Ac. 6:14 vIhsou/j o` Nazwrai/oj ou-toj, where the article is used with the adjective, not with vIhsou/j. So uniform indeed in the Greek is the presence of the article with the noun and ou-toj, that the absence of the article causes something of a jolt. In Ro. 9:8 the conjunction of the words tau/ta te,kna must not deceive us. The copula evstin must be supplied between. The American Revision indeed calls in the English relative to render the idiom ouv ta. te,kna th/j sarko.j tau/ta te,kna tou/ qeou/. Cf. the simple predicate use in 1 Cor. 6:11, kai. tau/ta, tinej h=te) In Lu. 1:36, ou-toj mh.n e[ktoj evsti,n, the substantive is predicate. The same thing is clearly true of Lu. 2:2, au[th avpografh. prw,th evge,neto. Cf. also tou/to u`mi/n shmei/on in Lu. 2:12. Some MSS. have to,, but in either case the copula is supplied. The remaining examples are not so simple, but ultimately resolve themselves into the predicate usage unless one has to except Ac. 24:21 (see below). In Lu. 7:44, tau,thn th.n gunai/ka, the article does not occur in L 47ev. Winer108 considers the reading without the article "unexceptionable," since the woman was present. In Lu. 24:21 the predicate accusative really is found, tri,thn tau,thn h`me,ran a[gei avf v ou- tau/ta evge,neto, a common Greek idiom difficult to put into English. It is not 'this third day,' but 'this a third day.' Cf. also 2 Pet. 3:1, tau,thn deute,ran gra,fw evpistolh,n. In this instance the English translation resorts to the relative 'that' to bring out the predicate relation, 'this is the second epistle that I write.' In Jo. 2:11,

702 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 2nd ed.

Addenda 3rd ed.

tau,thn evpoi,hsen avrch.n tw/n shmei,wn even the American Revision has a wrong translation, 'this beginning of miracles.' It is rather 'this Jesus did as a beginning of miracles.' But a and Chrys. here have th,n. In Jo. 4:18, tou/to avlhqe.j ei;rhkaj, the English relative is again necessary, 'this is a true thing that thou didst say' or 'thou didst speak this as a true thing.' The translation 'truly' rather obscures the idea. In Ac. 1:5, ouv meta. polla.j tau,taj h`me,raj, several difficulties appear. The litotes, ouv meta. polla.j, does not have the usual order.109 Cf. Ac. 27:14 for met v ouv polu,. There is besides a use of meta, somewhat akin to that of pro, in pro. ea}x h`merw/n tou/ pa,sca (Jo. 12:1).110 The order would more naturally be a ouv polla.j h`me,raj meta. tau,taj or ouv pollw/n h`merw/n meta. tau,taj. However, the predicate use of tau,taj without the article permits the condensation. The free translation 'not many days hence' is essentially correct. It is literally 'after not many days these' as a starting-point (from these). In Jo. 21:14, tou/to h;dh tri,ton evfanerw,qh vIshou/j, the matter is very simple, 'this already a third time,' or to use the English relative, 'this is now the third time that.' So also in 2 Cor. 12:14 and 13:1, tri,ton tou/to. The most difficult instance to understand is in Ac. 24:21, peri. mia/j tau,thj fwnh/j h-j evke,kraxa. Here 'concerning this one voice which I cried' makes perfectly obvious sense. The trouble is that it is the only N. T. example of such an attributive usage without the article. Blass111 takes it to be equivalent to h` fwnh. ha} evge,neto h=n mi,a au[th. This is, of course, the normal Greek idiom and is possibly correct. But one wonders if a lapse from the uniform idiom may not occur here. Radermacher (N. T. Gr., p. 92) cites tou,tou pra,gmatoj├ tau/ta avdikh,─ mata├ tou/to kth/ma from inscriptions in Magnesia (Petersen-Luschan, Reisen in Lykien, p. 35, n. 54) and e;sthsan to,de mnh/ma from a Bithynian inscription (Perrot, Exploration arch. de la Galatie, p. 24, N. 34). Hence one had best not be too dogmatic as to Luke's idiom in Ac. 24:21. After all, the predicate use may be the original use, as with evkei/noj. Cf. Brugmann, Griech. Gr., p. 426 f.; Thompson, Syntax of Attic Greek, p. 67. See also chapter XVI.

7. Ou-toj in Contrast with evkei/noj. The distinction between o[de for what follows and ou-toj for what precedes112 (not strictly observed in the ancient Greek) amounts to little in the N. T., since o[de is so rare. But ou-toj does, as a rule, refer to what is near or last mentioned and evkei/noj to what is remote. See au[th and ou-toj in

PRONOUNS ( vANTWNUMIAI) 703

2 Jo. 1:6 f. and tou/to in 2 Cor. 13: 9. This idiomatic use of ou-toj is plain in Ac. 7:19. In 1 Jo. 5:20 ou-toj really refers to auvtou/ ( evn tw|/ ui`w|/ auvtou/) and so no difficulty exists. In Ac. 4:11 ou-toj is resumptive and takes up the main thread of the story again (cf. ou-toj in verse 9). In Ac. 8:26 au[th may refer to Ga,zan, but more probably (see 3, end) refers to o`do,j, a more remote substantive, indeed. In Lu. 16:1 again only the sense113 makes it clear ( a;nqrw─ po,j tij h=n plou,sioj o[j ei=cen oivkono,mon├ kai. ou-toj) that ou-toj refers to oivkono,mon. In Lu. 18:14, kate,bh ou-toj dedikaiwme,noj eivj to.n oi=kon auvtou/ par v evkei/non the two pronouns occur in sharp contrast, one pointing out the publican, the other the Pharisee. In such contrasts ou-toj refers to the last mentioned. This is clearly one example (besides 2 Jo. 1:6 f.) in the N. T., which curiously enough Blass (Gr. of N. T. Gk., p. 171) does not recognise. Cf. also Jo. 13:24; evkei/noj tou,tw| in Jo. 5:38, and tau/ta evkei,noij in 1 Cor. 10:11. In Jo. 1:7 f. both ou-toj and evkei/noj are used of John and in proper idiom.114 Instead of evkei/noj we might have had ou-toj properly enough because of auvtou/, but evkei/noj calls us back pointedly to vIwa,nhj. Cf. Abbott, Johannine Grammar, p. 236. Note ou-toj o` lo,goj- o` maqhth.j evkei/noj in Jo. 21:23. In 1 Cor. 6:13, o` de. qeo.j kai. tau,thn kai. tau/ta katargh,sei, we find ou-toj used for both the near and the remote. The number and gender make it clear. In 1 Cor. 9:3 au[th points to what follows. In a case like evn tou,tw| cai,rw (Ph. 1:18), the main thought is meant by the demonstrative. So with evn tou,tw| di,dwmi\ tou/to ga.r u`mi/n sumfe,rei (2 Cor. 8:10). Cf. tou/to Ac. 24:14, etc.

8. As Antecedent of the Relative Pronoun. The absence of the demonstrative pronoun before the relative pronoun will be discussed later. This absence is in the case of a possible pronoun before the relative and after it also. The resumptive use of the demonstrative pronoun after the relative sentence has been already treated. But115 it is "the normal correlative" ou-toj- o[j. So ou-toj peri. ou- (Mt. 11:10) ou-toj o[n, (Jo. 7:25), ou-toj o[j (Ac. 7:40), tou/to- o[ (Ph. 2:5). See interrogative demonstrative and relative in ti,j evstin ou-toj o[j (Lu. 5:21; 7:49); ti, tou/to o[ (Jo. 16:17 f.). Cf. Lu. 24:17. On the whole, however, the demonstrative before the relative is not common in the N. T. In Gal. 2:10 both auvto, and tou/to are incorporated into the relative clause, oa} kai. evspou,dasa auvto. tou/to poih/sai.

704 A GRAMMAR OF THE GREEK NEW TESTAMENT

9. Gender and Number of ou-toj. See chapter X. In general, like other adjectives, ou-toj agrees with its substantive in gender and number, whether predicate or attributive. Cf. Jo. 2:11. In 1 Cor. 6:13, kai. tau,thn kai. tau/ta, note the number and gender. But sometimes the construction according to sense prevails. So the masculine, not feminine, in Ac. 8:10, ou-to,j evstin h` Du,na─ mij tou/ qeou/. So skeu/oj evklogh/j evsti,n moi ou-toj (Ac. 9:15), ou-toi and e;qnh (Ro. 2:14). Cf. also Ju. 12, <, ou-toi- nefe,lai├ de,ndra├ ku,mata├ avste,rej; 2 Pet. 2:17, ou-toi, eivsin phgai. and ou-toi - evlai/ai (Rev. 11: 4). In these examples assimilation to the gender of the predicate does not occur. Cf. tau/ta ti,, Jo. 6:9. In Mt. 21:42 (Mk. 12:11), para. kuri,ou evge,neto au[th the feminine occurs where the neuter would be natural in Greek. This is a piece of "translation" Greek (Ps. 118:23). In Hebrew the feminine is the case for abstract words, the Hebrew having no neuter gender. In Eph. 2:8, th|/ ga.r ca,riti, evste seswsme,noi dia. pi,stewj\ kai. tou/to ouvk evx u`mw/n, there is no reference to pi,stewj in tou/to, but rather to the idea of salvation in the clause before. But in 1 Pet. 2:19 f. we have two examples of the neuter ( tou/to) on purpose to present a more separate and abstract notion than au[th would have done, an ancient Greek idiom, tou/to ga,r ca,rij eiv- tou/to ca,rij para. qew|/. In 1 Cor. 10:6 the same principle applies, tau/ta de. tu,poi h`mw/n evgenh,qhsan. A striking example is found in 1 Cor. 6:11, kai. tau/ta, tinej h=te) Here tau/ta is much like toiou/toi, but more definite and emphatic. For this use of ou-toj see also Jo. 12:34. In Ph. 3:7, a[tina h=n moi ke,rdh├ tau/ta h[gnmai- zhmi,an, assimilation to the gender of the predicate is also absent.

Sometimes the plural tau/ta occurs where a single object is really in mind. The adverbial phrase meta. tau/ta (Lu. 12:4) can refer either to one or more incidents. It is not necessary to consider tau/ta as singular in idea in Jo. 19:36 and 1 Cor. 9:15. But the usage does appear in 3 Jo. 1:4, meizote,ran tou,twn ouvk e;cw ca,rin (or cara,n), and the adverbial accusative kai. tau/ta in Heb. 11:12. Some MSS. have kai. tau/ta instead of kai. tou/to in 1 Cor. 6:8.

But assimilation to the predicate both in gender and number occurs. So in Lu. 8:14 f. to. ) ) ) peso,n├ ou-toi, eivsin oi` avkou,santej) The same thing116 appears in Gal. 4:24, a[tina, evstin avllhgorou,mena\ au-tai ga,r eivsin du,o diaqh/kai. Note the assimilation of au[th in Lu. 2:2; 8:11; 22:53; Jo. 1:19; Ro. 11:27; 1 Cor. 9:3; 1 Jo. 2: 25; 5:3, 4, 9, 11, etc., and ou-toj in Mt. 7:12.

10. The Adverbial Uses of tou/to and tau/ta. See chapter XII.

PRONOUNS ( vANTWNUMIAI) 705

Here we have kai. tou/to (adverbial accusative or nominative absolute) like Latin idque (English 'and that too') in 1 Cor. 6 : 6 (CDb tau/ta), 8 (L tau/ta); Ro. 13:11; Eph. 2:8 (this last could be otherwise explained). Kai. tau/ta, the usual classical idiom,117 appears in Heb. 11:12 with a concessive participle. In tou/to me,n├ tou/to de, (Heb. 10:33) Blass118 sees a literary usage. In 2 Cor. 2:3 Paul has tou/to auvto, in the adverbial sense, while Peter (2 Pet. 1:5) turns the phrase around kai. auvto. tou/to de,) Cf. the adverbial use of kefa,laion in Heb. 8:1. The case of ou-toj in Jo. 21:21 is noteworthy.

11. The Phrase tou/t v e;stin. See also chapter X, viii, (c). It is used without any regard to the number, gender or case of the word in apposition with it, exactly like the Latin id est. There are eighteen examples of it given in Moulton and Geden's Concordance, all but three of them from the Acts, Romans, Philemon and Hebrews. It is a mark of the more formal literary style. In Mt. 27:46 the case explained is the vocative, in Mk. 7:2 the instrumental, in Ro. 7:18 the locative, in Heb. 2:14 the accusative, in Heb. 9:11 the genitive, in Heb. 7:5 the plural, in 1 Pet. 3:20 the plural. In Ro. 1:12 the uncontracted form occurs with In 1 Macc. 4:52 ou-toj o` mh.n Caseleu/ is in apposition with the genitive.119 Here ou-toj performs the function of tou/t v e;stin. Cf. the case-irregularities in the Apocalypse.

12. In Combination with Other Pronouns. Mention may be made of evn tou,tw| ou-toj (Ac. 4:10) and other instances of the double use of ou-toj. Cf. Mk. 6:2. Cf. ou-toj ou[tw in Mk. 2:7, tau/ta ou[twj (Ac. 24:9), ou[twj tou/to; (1 Cor. 5:3), and in 2 Pet. 3:11 tou,twn ou[twj pa,ntwn. Examples of auvto. tou/to are common in Paul (Ro. 9:17; 13:6; 2 Cor. 7:11; Ph. 1:6. Cf. 2 Pet. 1:5). For tou/to auvto, see 2 Cor. 2:3, auvto. tou/to Ro. 13:6. For auvtoi. oi-toi see Ac. 24:15, 20. For tou/to o[lon cf. Mt. 1:22; 26:56. There is no doubt some difference between tau/ta pa,nta (Mt. 4:9; Lu. 12:30; 16:14) and pa,nta tau/ta (Mt. 6:32). "In the first expression, pa,nta is a closer specification of tau/ta; in the second, pa,nta is pointed out demonstratively by means of tau/ta."120

13. Ellipsis of ou-toj. The demonstrative is by no means always used before the relative. Often the relative clause is simply the object of the principal verb, as in oa} le,gw u`mi/n evn th|/ skoti,a| ei;pate (Mt. 10:27). Sometimes the implied demonstrative must be expressed in the English translation. The simplest form of this

706 A GRAMMAR OF THE GREEK NEW TESTAMENT

idiom is where the case of the demonstrative would have been the same as that of the relative. Thus suggenh.j w'n ou- avpe,koyen Pe,troj to. wvti,on (Jo. 18:26). Cf. o[n in Ac. 1:24. In Ac. 8:24 w-n is for tou,twn aa} by attraction. But the ellipsis occurs also when a different case would have been found.121 So in Mt. 19:11 oi-j de,dotai would have been ou-toi oi-j de,d. In Jo. 13:29 w-n would have been preceded by tau/ta. Cf. also Ac. 8:19; 13:37, etc. In Ro. 10:14, pw/j pisteu,swsin ou- ouvk h;kousan, the antecedent of ou- would be either tou,tw| (or evpi. tou,tw|) or more probably eivj tou/ton (preposition also dropped). When a preposition is used, it may belong to the relative clause, as in pw/j evpikale,swntai eivj oa}n ouvk evpi,steusan (Ro. 10:14; cf. Jo. 19:37), or to the implied demonstrative, as in i[na pisteu,shte eivj oa}n avpe,steilen (Jo. 6:29). In Ro. 14:21 evn w|- illustrates the preposition with the relative, while in the next verse it illustrates the preposition with the antecedent. In Jo. 11:6 evn w|- to,pw| is an example where evn would have been used with both antecedent and relative. So as to avf v w-n in 2 Cor. 2:3, etc.122 The same principle of suppressed antecedent applies to relative adverbs, as in h=lqen o[pou h=n (Jo. 11:32), strictly evkei/se o[pou.

14. Shift in Reference. It is possible that in Ac. 5:20, lalei/te evn tw|/ i`erw|/ tw|/ law|/ pa,nta ta. r`h,mata th/j zwh/j tau,thj, a slight change in sense has occurred, tau,thj more naturally going with r`h,mata. Cf. evk tou/ sw,matoj tou/ qana,tou tou,tou (Ro. 7:24). But the point is not very material.

(g) vEkei/noj. Cf. Latin ille. The old form (Epic, Pindar, Tragic poets) was kei/noj or kh/noj (Doric and Lesbian).123 Brugmann124 indeed connects it with the old Indo-Germanic root ko. The locative adverb ev─kei/ (cf. kei/─qi├ kei/─qen, Doric, Lesbian) is the immediate source of the pronoun kei/─noj├ ev─kei/─noj. Cf. English hi-ther. The original usage was therefore predicate.125 Thus in Thuc. i, 52. 2, nh/ej evkei/nai evpiple,ousi ('ships yonder are sailing ahead'), we must not confuse it with ai` nh/ej evkei/nai ('those ships'). Cf. the "adverbial" use of ou-toj. By a strange coincidence, while at work on this paragraph (Nov., 1908), I received a letter from Rev. R. H. Graves, D.D., of Canton, China, concerning Chinese pronouns, suggested by the chapter on Pronouns in my Short Grammar of the Greek N. T. He says: "The ordinary pronoun for the third person is k'ei. In Canton we also use k'ni. Compare He mentions other accidental similarities, but I dare not venture into Chinese etymology.

PRONOUNS ( vANTWNUMIAI) 707

1. The Purely Deictic. We have a few examples in the N. T. So in Jo. 13:26, evkie/no,j evstin w|- evgw. ba,yw to. ywmi,on kai. dw,sw auvtw| for Judas was present at the table. In Mt. 26:23 we have ou-toj. A gesture may also have accompanied the remark of the Pharisees in Jo. 9:28, su. maqhth/j ei= evkei,nou) Cf. also Jo. 19:21. If evkei/noj in Jo. 19:35 be taken as an appeal to God as a witness to the truth of what the writer is saying (possible, though by no means certain), the usage would be deictic. Blass126 considers that "everything is doubtful" as to this verse, a doubt shared by Abbott.127 For myself I think that evkei/noj is here anaphoric and refers to auvtou/ (cf. the similar reference of ou-toj to auvtou/ in 1 Jo. 5:20; but see Remote Object). Another possible deictic example is in Jo. 7:11. Jesus was not present, but in the minds of the people a subject of discussion. Cf. also 9:12.

2. The Contemptuous Use (cf. ou-toj). It appears unmistakably (see 1) in Jo. 9:28, su. maqhth.j ei= evkei,nou. It may also exist128 in Jo. 19:21. Cf. the solemn repetition of evkei/noj with o` a;nqrwpoj in Mt. 26:24, as well as the change from ou-toj in verse 23.

3. The Anaphoric. This is the more frequent use of this pronoun. Thus in Jo. 1:8 evkei/noj takes up ou-toj of verse 7 ( vIwa,nhj of verse 6). In Jo. 18:5 o` de. maqhth.j evkei/noj resumes the story of a;lloj maqhth,j immediately preceding. Cf. a;lloj and evkei/noj in Jo. 5:43. In Jo. 13:25 evkei/noj refers indeed to the preceding tou,tw| (cf. evkei/noj ou[twj). In Jo. 5:19 the reference is to pate,ra just before. Cf. Jo. 4:25. vEkei/noj de, (3 (Jo. 2:21) is continuative like ou-toj. The articular participle may be followed by the resumptive evkei/noj. So o` pe,myaj me- evkei/noj Jo. 1:33). Cf. Jo. 5:11; 2 Cor. 10:18. So in Jo. 1:18 the pronoun refers to qeo,j followed by o` w;n. Cf. Mk. 7:20 evkei/no) See Jo. 14:21. For distinction between evkei/noj and auvtou/ see 2 Tim. 2:26; 3:9.

4. The Remote Object (Contrast). This is not always true, as is shown by Jo. 18:15. Cf. Tit. 3:7. It is common thus to refer to persons who are absent. So in Jo. 3:28.(cf. Jo. 7:11) John speaks of Christ in contrast to himself, avpestalme,noj eivmi. e;mrposqen eivkei,nou. So in verse 30, evkei,noij- h`mw/n) In 1 Cor. 9:25 note evkei/noi me.n- h`mei/j de,) So in 10:11 evkei,noij- h`mw/n, 15:11 ei=te evgw. ei;te evkei/noi. In Ac. 3:13 the contrast is sharp between u`mei/j- evkei,nou, and in 2 Cor. 8:14 between u`mw/n───evkei,nwn (cf. evkei,nwn- u`mw/n in same verse). Cf. u`mi/n - evkei,noij in Mt. 13:11. In Jo. 5:39 evkei/nai

708 A GRAMMAR OF THE GREEK NEW TESTAMENT

is in opposition to u`mei/j├ as evkei/noj to u`mei/j in the preceding verse. Cf. 2 Cor. 8:9. For a contrast between those present in the same narrative see ou-toj in Lu. 18:14. Cf. evkei/noj and auvto,j in 1 Jo. 2:6 and tou/to h' evke/no in Jas. 4:15. It is common in expressions of place, like dia. th/j o`dou/ evkei,nhj (Mt. 8:28), eivj o[lhn th.n gh/n evkei,nhngrk grk(9:26; cf. evn 9:31), etc. It is frequent also with general phrases of time, like evn tai/j h`me,raij evkei,naij (Mt. 3:1). Cf. Mk. 8:1; Lu. 2:1. It usually occurs at a transition in the narrative and refers to something previously mentioned. Blass129 notes that Lu. Lu.(1:39) uses also tau,taij in this phrase and that in 6:12 D has evkei,naij rather than tau,taij. In particular observe the phrase evkei,nh h` h`me,ra for the Last Day (Mt. 7:22; Mk. 14:25; Lu. 21: 34; 17:31; Jo. 16:23, etc. Cf. Jo. 6:40, etc.).

5. Emphasis. Sometimes evkei/noj is quite emphatic. Abbott130 notes that in John's Gospel, outside of dialogue, evkei/noj usually has considerable emphasis. Instance Jo. 1:8, 18, 33; 2:21; 3:30; 4:25; 5:19, 38; 6:29; 8:42; 14:26; 15:26, etc. In the First Epistle of John he observes that it occurs only seven times and all but one refer to Christ. He is the important one in John's mind. Cf. auvto,j in Ac. 20:35. But evkei/noj is not always so emphatic even in John. Cf. Jo. 9:11, 25; 10:6; 14:21; 18:17; Mk. 16:10 ff ; 2 Tim. 3:9.

6. With Apposition. It is not common with words in apposition. But note Jo. 16:13, evkei/noj├ to. pneu/ma th/j avlhqei,aj (cf. Jo. 14:26). Note also evkei/no ginw,skete├ o[ti (Mt. 24:43) after the fashion of ou-toj with o[ti. Cf. also the resumptive uses with participles (Jo. 1:18, etc.).

7. Article with Nouns except when Predicate. When the noun is used with evkei/noj in the N. T., the article always appears, except when predicate. In Jo. 10:1, evkei/noj kle,pthj evsti,n the substantive is predicate, as in 10:35, evkei,nouj ei=pen qeou,j. With adjectives we may note the repetition of the article in Jo. 20:19 and the ambiguous position of evkei,nh in Heb. 8:7 due to the absence of diaqh,kh. With o[loj we find this order, eivj o[lhn th.n gh/n evkei,nhn (Mt. 9:26, etc.) and pa/j the same, pa/san th.n ovfeilh.n evkei,nhn (Mt. 18:32, etc.).

8. As Antecedent to Relative. So evkei/no,j evstin w|- (Jo. 13:26), evkei/non u`pe.r ou- (Ro. 14:15) evkei,noij di v ou[j (Heb. 6:7). Note also evkei/no,j evstin o` avgapw/n (Jo. 14:21) where the articular participle is the practical equivalent of a relative clause.

9. Gender and Number. Little remains to be said about variations in gender and number. Two passages in John call for re-

PRONOUNS ( vANTWNUMIAI) 709

Addenda 3rd ed.

mark, inasmuch as they bear on the personality of the Holy Spirit. In 14:26, o` de. para,klhtoj├ to. pneu/ma to. a[gion oa} pe,myei o` path.r evn tw|/ ovno,mati, mou├ evkei/noj u`ma/j dida,xei, the relative o[ follows the grammatical gender of pneu/ma. vEkei/noj, however, skips over pneu/ma and reverts to the gender of para,klhtoj. In 16:13 a more striking example occurs, o[tan de. e;lqh| evkei/noj├ to. pneu/ma th/j avlhqei,aj. Here one has to go back six lines to evkei/noj again and seven to para,klhtoj. It is more evident therefore in this passage that John is insisting on the personality of the Holy Spirit, when the grammatical gender so easily called for evkei/no. Cf. o[ in Jo. 14:17, 26 and auvto, in 14:17. The feminine evkei,nhj in Lu. 19:4 evidently refers to o`dou/ unexpressed.

10. Independent Use. The frequency of evkei/noj in John's Gospel may be noticed, but the Synoptics and Acts are not far behind. More curious, however, is the fact that in the Synoptics evkei/noj is nearly always used with a substantive (adjectival) while the independent pronominal use of the singular is almost confined to the Gospel of John (and First Epistle).131 All the uses in the First Epistle and nearly all in the Gospel are independent. As exceptions note Jo. 4:39, 53; 11:51, 53; 16:23, 26, etc. On the other hand only two instances appear in the Apocalypse Apocalypse(9:6; 11:13) and both with substantives.

(h) Auvto,j. It has undoubtedly developed in the koinh, a demonstrative force as already shown on p. 686, and as is plain in the modern Greek. Moulton132 quotes plain examples from the papyri (see above). In the N. T. it is practically confined to Luke (and Mt. 3:4 perhaps), where it is fairly common, especially in the Gospel. So evn auvth|/ th|/ oivki,a| (Lu. 10:7), 'in that house.' Moulton133 notes that in Mt. 11:25 (parallel to Lu. 10:21) we have evn evkei,nw| tw|/ kairw|/ and in Mk. 13:11 evn evkei,nh| th|/ w[ra| (parallel to Lu. 12:12 evn auvth|/ th|/ w[ra|). The tendency was not foreign to the ancient Greek and it is common enough in the modern vernacular134 to find auvto.j o` ╩ 'this.'

(i) THE CORRELATIVE DEMONSTRATIVES. Only four occur in the N. T. One of them appears only once and without the article, fwnh/j evnecqei,shj auvtw|/ toia/sde (2 Pet. 1:17). It has died in the vernacular (Radermacher, N. T. Gr., p. 63) like o[de├ thliko,sde and toso,sde) Thlikou/toj appears once as predicate, thlikau/ta o;nta (Jas.

710 A GRAMMAR OF THE GREEK NEW TESTAMENT

3:4), elsewhere attributive. The article is not used. This correlative of age always refers to size in the N. T. (2 Cor. 1:10; Heb. 2:3). Once indeed it is in connection with ou[twj me,gaj (Rev. 16:18) and so redundant. The other two are toiou/toj and tosou/toj) Toiou/toj is the demonstrative of quality (Latin talis) and it is used with a good deal of freedom. It is, of course, merely toi/oj and ou-toj combined. The compound form alone occurs in the N. T. and became more frequent generally.135 Toiou/toj without a substantive is used either without the article (Lu. 9:9) or more usually with the article in the attributive position (Mt. 19: 14; Ac. 19:25; Ro. 1:32; 1 Cor. 7:28; 2 Cor. 10:11, etc.). In Jo. 4:23, toiou,touj zhtei/ tou.j proskunou/ntaj, the articular participle is in the predicate accusative. When used with substantives toiou/toj may be anarthrous, as in Mt. 9:8; 18:5; Mk. 4: 33; Heb. 7:26; 8:1; Jas. 4:16, etc., but the article occurs also (Mk. 6:2; 9:37; 2 Cor. 12:3). In Mk. 6:2 we have the order ai` duna,meij toiau/tai (cf. ou-toj├ evkei/noj). It comes before the substantive (Jo. 9:16) or after (Ac. 16:24). It is used as the antecedent of oi-oj (Mk. 13:19; 1 Cor. 15:48; 2 Cor. 10:11) following oi-oj. But note also toiou,touj o`poi/oj in Ac. 26:29, toiou/toj o[j in Heb. 7:26 f.; 8:1, and in 1 Cor. 5:1 toiau,th h[tij) We even have toiou/toj w`j in Phil. 1:9. Cf. poi/oj- toiou/toj in a Logion of Jesus, P.Oxy. IV, p. 3, 1. tosou/toj $to,soj├ ou-toj% is the pronoun of degree (Latin tantus), both size, tosau,thn pi,stin (Mt. 8:10), and quantity, a;rtoi tosou/toi (Mt. 15:33). It occurs with the article only once, o` tosou/toj plou/toj (Rev. 18:16). Sometimes it appears without a substantive, as in Ac. 5:8; Gal. 3:4; Heb. 1:4, etc. It is the correlative with o[soj in Heb. 1:4 tosou,tw| - o[sw|, 7:20-22 kaq v o[son- kata. tosou/to, and in 10:25 tosou,tw|─o[sw|. It is worth while at this point to note the correlative adverbs, ou[twj w[ste (Ac. 14:1), ou[twj w`j (1 Cor. 4:1), ou[twj- o[pwj (Mt. 5:16). Cf. w[ste- ou[twj de, (Ro. 15:20).

VII. Relative Pronouns ( vanaforikai. avntwnumi,ai).

(a) LIST IN THE N. T. The only relatives in the N. T. (not counting adverbs) are o[j├ o[stij├ oi-oj├ o`poi/oj├ o[soj├ h`li,koj, and o` in the Apocalypse. The others have fallen by the way. Some MSS. read o[nper in Mk. 15:6, while o`sdh,per in Jo. 5:4 is not in the critical text. The LXX has o[per $a[per% five times,136 but h`li,koj not at all. These relative pronouns do not occur with uniform frequency as will be seen. [Oj is the only one very common.

PRONOUNS ( vANTWNUMIAI) 711

(b) THE NAME "RELATIVE." It is not very distinctive.137 The idea of relation (anaphoric use) belongs to the demonstrative and to the personal pronouns also. The anaphoric demonstrative use is indeed the origin of the relative.138 The transition from demonstrative to relative is apparent in Homer in the case of both o` and o[j. Sometimes it is difficult in Homer to tell the demonstrative and the relative apart.139 Cf. English that, German der. Homer often used te and tij with o` and o[j to distinguish the relative from the demonstrative.140 Gradually the relative use, as distinct from the anaphoric demonstrative, won its way.

(c) A BOND BETWEEN CLAUSES. The relative becomes then the chief bond of connection between clauses. Indeed many of the conjunctions are merely relative adverbs, such as w`j├ o[te├ o[pwj, etc. The relative plays a very important part in the structure of the subordinate sentence in Greek. That matter will receive due treatment in chapter XIX, Mode. The agreement of the relative with antecedent in person, number, gender, and sometimes case, is just the natural effort to relate more exactly the two clauses with each other. These points will receive discussion under o[j which best exemplifies them. The assimilation is at bottom the same that we see in other adjectives (cf. demonstrative pronouns). The assimilation of the relative in person, gender, number, and even case of the antecedent may be compared to assimilation in the adjective and even verbs (compound verbs especially) and prepositions. Cf. Josef Liljeblad, De Assimilatione Syntactica aped Thuc. Questiones, 1900, p. 1).

(d) [Oj.

1. In Homer. See discussion of the demonstrative o[j for origin.141 But already in Homer the relative sense, a;rqron u`potaktiko,n, is the main one, and the demonstrative is on the decline.142

2. Comparison with Other Relatives. Though o[j in the N. T. far outnumbers all the other relatives, yet the distinction between

712 A GRAMMAR OF THE GREEK NEW TESTAMENT

o[j and the other relatives is breaking down. Indeed in the vernacular it may be questioned if it was ever preserved. One may compare the unchangeable Hebrew rv,a]. Moulton143 observes that in Polybius the distinction between o[j and o[stij has "worn rather thin." In the LXX o[j is frequent,144 but in the modern Greek o[j "is used rarely even in writing."145 It is wholly absent in the vernacular. The modern Greek vernacular uses pou/ or o[pou. In the oblique cases the conjunctive pronoun tou/├ th/j is added to pou/ (cf. the Hebrew idiom). See Thumb, Handb., p. 93. Jebb (Vincent and Dickson's Handb., etc., p. 303) calls it "a curious example of false analogy" and finds an instance in Aristophanes (Birds, 1300), me,lh o[pou. Here o[pou╩evn oi-j. The vernacular carried it further. He cites modern English vernacular, "The men as he met." Indeed in Rev. 2:13 o[pou really points to an unexpressed par v u`mi/n. In Col. 3:11 o[pou is almost personal. The occasional apparent confusion between o[j and interrogative pronouns will be discussed directly. On the whole, o[j in the N. T., as in the koinh, generally, is still used in accord with the classic idiom.

3. With Any Person. In itself, of course, o[j, like all relatives, has no person. So the first person in 1 Cor. 15:10, the second person in Ro. 2:23, the third person in Mt. 5:19; Lu. 6:48 f.; 1 Cor. 4:17. These examples may suffice.

4. Gender. This is not so simple. The normal thing is for the relative to agree with the antecedent in gender, as in 1 Cor. 4:17, Timo,qeon├ o[j evsti,n mou te,knon. So in Col. 1:24 u`pe.r tou/ sw,matoj auvtou/├ o[ evstin h` evkklhsi,a; Col. 2:10 evn auvtw|/├ o[j evstin h` kefalh, (cf. Eph. 4:15) Col. 2:17 sabba,twn├ a[ (some MSS. o[) evstin skia. tw/n mello,ntwn* Rev. 5:6 ovfqalmou.j e`pta,├ oi[ eivsin ta. [ e`pta.] pneu,mata. In Rev. 21:8, to. me,roj auvtw/n evn th|/ li,mnh| th|/ kaiome,nh| puri. kai. qei,w| o[ evstin o` qa,natoj o` deu,teroj, the agreement is regular, but the idea of o[ may be more inclusive than merely146 me,roj. Cf. 1 Pet. 3 : 4.

On the other hand the relative is assimilated in gender to the predicate substantive. This is also a perfectly natural agreement. Winer147 considers that this is true particularly when the predicate presents the main idea. See Mk. 15:16, th/j auvlh/j├ o[ evstin praitw,rion* Gal. 3:16, tw|/ spe,rmati, sou├ o[j evstin Cristo,j; Eph. 6:17, th.n ma,cairan

PRONOUNS ( vANTWNUMIAI) 713

tou/ pneu,matoj├ o[ evstin r`h/ma qeou/; Rev. 4:5, lampa,desj- a[ eivsin ta. e`pta. pneu,mata (but some MSS. ai[). Cf. 2 Th. 3:17. The MSS. vary in a number of instances between agreement with antecedent and predicate. So Col. 1:27, tou/ musthri,ou tou,tou- o[j (or o[) evstin Cristo,j. Cf. also 1 Tim. 3:16, where the true text o[j is changed in the Western class of documents to o[ to agree with musth,rion. See also Eph. 1:13 f., tw|/ pneu,mati - o[ (MSS. o[j) evstin avrrabw,n. So ai[ or a[ in Rev. 5:8. In Mt. 13:31 f. ko,kkw| is followed first by o[n and then by o[ (cf. sperma,twn).

In another group of passages the change is made according to the real gender rather than the grammatical. Thus in Ac. 15:17 ta. e;qnh evf v ou[j (cf. 26:17), Jo. 6:9 paida,rion o[j e;cei, Ro. 9:23 f. skeu,h evle,ouj- ou[j├ Col. 2:19 kefalh.n evx ou-, Phil. 1:10 te,knou o[n, Rev. 13:14 qhri,w| o[j. In Gal. 4:19 as is preceded by both u`ma/j and tekni,a. In 2 Jo. 1:1, evkleth|/ kuri,a| kai. toi/j te,knoij auvth/j├ ou[j the grammatical gender (feminine and neuter followed by masculine) is ignored entirely. Cf. Ph. 2:15.

In a passage like 1 Cor. 15:10, eivmi. o[ eivmi, there is no mistake. See o[j above in verse 9. It is not 'who I am,' but 'what I am,' not exactly oi-oj either, but a more abstract idea than that. Cf. o[ in Jo. 4:22, used twice for the object of worship, God. So in 1 Jo. 1:1 observe oa} h=n- oa} avkhko,amen├ oa} e`wra,kamen (cf. verse 3) for Jesus. One may recall here that the collective abstract neuter, pa/n o[ (Jo. 6:37, 39; 17:2), is used for the disciples. Cf. o[───kavkei/noi (Jo. 17:24).

Sometimes also the relative agrees neither with the antecedent nor with a predicate substantive, hut gathers the general notion of 'thing.' A good example occurs in 1 Jo. 2:8, evntolh.n kainh.n gra,fw u`mi/n├ o[ evstin avlhqe,j, 'which thing is true.'148 So Eph. 5:5, pleone,kthj├ o[ (Western and Syrian classes read o[j) evstin eivdw─ lola,trhj, 'which thing is being an idolater.' A particularly good example is Col. 3:14 where o[ comes in between a feminine and a masculine, th.n avga,phn├ o[ evstin su,ndesmoj. In Mk. 12:42 we have a similar example, lepta. du,o├ o[ evstin kodra,nthj.

Indeed o[ evstin comes to be used as a set expression, like tou/t v e;stin, without any regard to the antecedent or the predicate, as o[ evstin ui`oi. bronth/j, Mk. 3:17. Three phrases go together in this matter, o[ evstin├ o[ e`rmhneu,etai├ o[ le,getai. The two latter occur in the periphrastic form also. Indeed the examples just noted above may very well be explained from this point of view. So Mt. 1: 23, vEmmanouh.l o[ evstin meqermhneuo,menon meq v h`mw/n o` qeo,j, where ob-

714 A GRAMMAR OF THEGREEK NEW TESTAMENT

serve the neuter participle like o[. Cf. Ac. 4:36. In Mt. 27:33, Golgoqa. o[ evstin krani,ou to,poj lego,menoj, the participle is masculine like to,poj (cf. Mk. 15:22). In Jo. 1:39 oa} le,getai meqermhneuo,─ menon connects two vocatives. Cf. 20:16. In Jo. 1:41 note the accusative and nominative connected with neuter participle, Messi,an o[ evstin meqermhneuo,menon Cristo,j, occurs between verb-forms, as in Mk. 5:41; 7:34; or genitives as in Heb. 7:2; Rev. 20:12; 21:17; or whole clauses, as in Mk. 15:34. But see Jo. 9:7; Rev. 20:2. In Ac. 9:36, however, the personal construction occurs, Tabeiqa,├ ha} diermhneuome,nh le,getai Dorka,j. See also chapter X, VIII, (c).

Once more, o[ is used to refer to a verbal idea or to the whole sentence. Instance Mt. 12:4, tou.j a;rtouj th/j proqe,sewj e;fagon o[ ouvk evxo.n h=n auvtw|/ fagei/n. Here probably to. fagei/n is the idea referred to,149 though in Mk. 2:26 and Lu. 6:4 we have ou[j. The neuter gender is only natural here. In Ac. 2:32 ou- is most likely 'whereof,' though 'of whom,' referring to vIhsou/n, is possible. So as to 3:15. But there is no doubt as to Ac. 11:30, oa} kai. evpoi,hsai; 26:10, oa} kai. evpoi,hsa; Gal. 2:10, oa} kai. evspou,dasa auvto. tou/to poih/sai (note here the use of auvto. tou/to in the relative clause); Col. 1:29 eivj oa} kai. kopiw/ (cf. eivj o[ in 2 Th. 1:11; 2:14; 1 Pet. 2:8). Cf. also oa} kai. u`ma/j avnti,tupon nu/n sw,zei ba,ptisma (1 Pet. 3:21). Per contra see in the papyri o[n used like o[ after analogy of toiou/to$n%.150 Note in passing o[ o` in Lu. 2:15, like h|- h[ te in Heb. 9:2.

5. Number. Here again, as a rule, the relative concurs with the antecedent in number, as in avsth.r o[n (Mt. 2:9), qeou/ o[j (Ro. 2:6). The construction according to sense is not infrequent, as in plh/qoj oi[ (Lu. 6:17 f.), kata. po,lin pa/san evn ai-j (Ac. 15:36, note distributive idea), mwrologi,a h' euvtrapeli,a a[ (Eph. 5:4, where feminine singular could have occurred because of h;%├ genea/j- evn oi-j (Ph. 2:15), deu─ te,ran u`mi/n gra,fw evpistolh.n├ evn ai-j (2 Pet. 3:1, referring to both, probably). Cf. o[ - le,gontaj (Rev. 5:13). On the other hand note the change from the plural to the singular in h`me,rai dw,deka avf v h-j (Ac. 24:11), and evn ouvranoi/j - evx ou- (Ph. 3:20). For the neuter plural in the relative (cf. tau/ta) to cover a vague general idea see w-n, in 1 Tim. 1:6, avnq v w-n Lu. 1:20, evn oi-j Lu. 12:1 (cf. Ac. 26:12), evf v oi-j Ro. 6:21, etc. Cf. Col. 2:22.

6. Case.

(a) Absence of attraction normal. The obvious way is for the case of the relative to be due to the construction in which it is used or to follow the same law as other nouns and pronouns (so

PRONOUNS ( vANTWNUMIAI) 715

with prepositions). That is to say, assimilation of case is not a necessity. It was indeed in a sense an after-refinement. One must not get the notion that assimilation of case had to be. Thucydides,151 for instance, did not use it so extensively in his rather complicated sentences, where the relative clauses stand to themselves. Indeed the absence of it is common enough in the N. T., outside of Luke. Cf. Mt. 13:31 ko,kkw| o[n, Mk. 13:19 kti,sewj h[n Jo. 2:22 lo,gw| o[n, (cf. 4:50), Jo. 4:5 cwri,ou o[ (CD ou-), Tit. 3:5 e;rgwn a[├ Mt. 27:60 mnhmei,w| o[, Ac. 8:32 grafh/j h[n) Not to be exhaustive, one may refer to the rather long list in Winer-Schmiedel152 (Mt. 13:44, 48; 23:35; Lu. 13:19, 21; Ac. 1:4; 4:10; 1 Tim. 6:21; Heb. 6:19; 8:2; 9:7; 1 Pet. 1:8; Rev. 1:20, etc.). The absence of assimilation in case is not only common in the old Greek, but also in the LXX, the Apocrypha and the papyri. In Aristotle attraction is nearly confined to the more recondite essays (Schindler, De Attractionis Pronominum Rel. Usu Aristotelico, p. 94).

( b) Cognate accusative. The accusative in Ro. 6:10, oa} avpe,qanen├ oa} zh|/├ and Gal. 2:20, oa} zw/, may be called adverbial. In reality it reproduces the idea of the verb (cognate acc.). Cf. Mk. 10:38 f.

( g) Attraction to the case of the antecedent. This is very common in the N. T., especially in the writings of Luke. The papyri, even "the most illiterate of them,"153 show numerous examples of attraction, "a construction at least as popular in late as in classical Greek." This applies to the LXX also. The MSS. naturally vary sometimes, some having attraction, others not. Indeed Blass154 finds this "always" in the passages in W. H. without attraction save in Heb. 8:2. Cf. h[n $h-j) in Mk. 13:19, o[n $w|-% in Jo. 2:22; 4:50, etc. On the whole attraction seems the more common. But this "idiomatic attraction of the relative" "occurs only twice in Matthew Matthew(18:19; 24:50) and once in Mark Mark(7:13)," whereas it "is very common in Luke" (Plummer, Comm., p. li). The effect of "this peculiar construction" was to give "a sentence more internal unity and a certain periodic compactness."155 No instance of attraction of a nominative to an oblique case occurs in the N. T., though this idiom is found in the ancient Greek.156

716 A GRAMMAR OF THE GREEK NEW TESTAMENT

It is usually the accusative case that is assimilated into another oblique case. Thus the accusative may be attracted into the genitive, as pra,gmatoj ou- (Mt. 18:19), lo,gou ou- (Jo. 15:20), pa,ntwn w-n (Ac. 1:1; 3:21; 22:10), diaqh,khj h-j (Ac. 3:25), evpaggeli,aj h-j grk grk grk(7:17), evqnw/n w-ngrk grk(7:45), pneu,matoj a`gi,ou ou- (Tit. 3:6). Cf. also Ac. 9:36; 22:10; 1 Cor. 6:19; 2 Cor. 10:8, 13; Eph. 1: 8; Heb. 6:10; 9:20; Jas. 2:5. In several instances it is the accusative of the inner object that is attracted. Cf. Eph. 1:19 f. So paraklh,sewj h-j parakalou,meqa (2 Cor. 1:4), ca,ritoj h-j evcari,twsen (Eph. 1:6), klh,sewj h-j evklh,qhtegrk grk(4:1), fwnh/j h-j evke,kraxa (Ac. 24:21), e;rgwn avsebei,aj w-n hvse,bhsan (Ju. 15).157 There are examples also of the accusative attracted to the ablative. So evk tw/n kerati,wn w-n (Lu. 15:16), evk tou/ u[datoj ou- (Jo. 4:14), avpo. tw/n ovyari,wn w-ngrk grk(21:10), evk tou/ pneu,matoj ou- (1 Jo. 3:24). Cf. Jo. 7:31. Then again the assimilation of the accusative to the pure dative might have been expected, but curiously enough I find so far no example of it in the N. T. In 1 Cor. 7:39 there is an instance of the relative attracted from the accusative to the dative of an omitted antecedent, evleuqe,ra evsti.n w|- qe,lei gamhqh/nai, unless gamhqh/nai be repeated, when w|- is the necessary case. However, several examples occur where the accusative is attracted to the locative or the instrumental. Instances of the locative are found in evn h`me,ra| h|-──evn w[ra| h|- (Mt. 24:50. This is not an instance of one preposition for antecedent and relative), evpi. pa/sin oi-j (Lu. 2:20; 9:43; 24:25), evn tw|/ ovno,mati, sou w|- (Jo. 17:11 f.),2 evn- qli,yesin ai-j (Ac. 7:16), evn avndri. w|-grk grk(17:31), evpi. tw|/ lo,gw| w|-grk grk(20:38), evpi. th|/ avkaqarsi,a| h|- (2 Cor. 12:21), evpi. e;rgoij avgaqoi/j oi-j (Eph. 2:10),158 evn tw|/ mnh,mati w|- (2 Th. 1:4), evn tw|/ pothri,w| w|- (Rev. 18:6). This is probably true also of 1 Cor. 7:20, evn th|/ klh,sei h|- evklh,qh, where h[n would have been the cognate accusative.159 For attraction to the instrumental see para─ do,sei h|- (Mk. 7:13), do,xh| h|- (Jo. 17:5, but W. H. have h[n in margin), shmei,oij oi-j (Ac. 2:22), qusi,aij ai=j (Heb. 10:1, but W. H. as). In a few instances it is an open question whether we have attraction or not. Thus in Jo. 13:5, tw|/ lenti,w| w|- h=n diezwsme,noj├ either the instrumental w|- or the accusative o[ (cf. Jo. 21:7) is correct. In Ac. 9:17, evn th|/ o`dw|/ h|- h;rcou, the cognate accusative h[n is possible, though the locative originally is more likely. In 1 Th. 3:9, evpi. pa,sh| th|/ cara|/ h|- cai,romen, a cognate accusative was possible ( h[n) attracted

PRONOUNS ( vANTWNUMIAI) 717

to the locative or an original instrumental. In Col. 1:23, tou/ euv─ aggeli,ou ou- hvkou,sate, either the accusative or the genitive might occur with avkou,w. But in 2 Tim. 1:13, lo,gwn w[n par v evmou/ h;kousaj, the accusative was almost certainly the original form.160 Cf. Ac. 1:4 ha}n hvkou,sate, mou. Plummer (On Luke, p. li) notes that this attraction in Luke is particularly frequent after pa/j (Lu. 2:20; 3:19; 9:43, etc.). In Lu. 5:9, evpi. th|/ a;gra| tw/n ivcqu,wn w-n $h|-% sune,labon, the attraction in some MSS. is to the locative, in others to the genitive.

A few instances are found in the N. T. where the attraction is from some other case than the accusative. A clear case of a locative assimilated to a genitive appears in Ac. 1:22, e[wj th/j hpme,raj h-j avnelh,mfqh. This is in accord with the ancient Greek idiom. The very same construction appears in the LXX (Lev. 23:15. Cf. Bar. 1:19). In 1 Tim. 4:6 A reads didaskali,aj h|- parhkolou,qhkaj├ but the rest have h-j. A dative has been attracted into the genitive along with incorporation and the preposition in Ro. 4:17, kate,nanti ou- evpi,steusen qeou/╩ kate,nanti tou/ qeou/ h|- evpi,steusen. So the phrase avf v h-j (Ac. 24:11; 2 Pet. 3:4, but Lu. 7:45 w[raj) is an abbreviation of etc, avf v h`me,raj h|- (locative attracted to ablative). In Ac. 20:18 we actually have avpo. prw,thj h`me,raj avf v h-j evpe,bhn, but as a point of departure (ablative) rather than a point of location (locative). Cf. also avf v h-j h`me,raj (Col. 1:6, 9) where the incorporation resolves itself into avf v h`me,raj h|-) So likewise a;cri h-j h`me,raj (Mt. 24:38; Lu. 1:20; 17:27; Ac. 1:2) really comes from a;cri h`me,raj h|- (locative to genitive). In Heb. 3:9 a can be regarded as adverb 'where' or as relative 'wherewith' (marg. of the Ameriican Revision). If it is relative, w|- was probably the unattracted form (instrumental to genitive like peirasmou/). In Mk. 10:38 f., to. ba,ptisma oa} bapti,zomai, the relative is in the cognate accusative retained with the passive verb.161 See further chapter on Cases.

(5) Inverse attraction. What is called inverse attraction is due to the same tendency to identify antecedent and relative, only the assimilation is that of the antecedent to the relative. In itself this phenomenon is no more peculiar than the other. Plato" who uses the ordinary attraction very often, seldom has inverse attraction (Cleef, De Attractionis in Enuntionibus Rel. Vsv Platonico, pp. 44-46). No inverse attraction is found in Pisidian Greek (Compernass, De Serm. Gr., p. 13). The examples are not very numerous in the N. T., but the ancient Greek amply supports the

718 A GRAMMAR OF THE GREEK NEW TESTAMENT

idiom.162 One example, li,qon oa}n avpedoki,masan, occurs in Mt. 21:42; Mk. 12:10= Lu. 20:17. It is from the LXX (Ps. 118:22). In 1 Pet. 2:7 W. H. read Mos. Cf. also Lu. 1:73, o[rkon oa}n w;msen├ which might have been o[rkou ou- after mnhsqh/nai.163 See also 1 Cor. 10:16, to.n a=rton oa}n klw/men. Hence also to. poth,rion oa} euvlogou/men, of verse 16. If o[n is a part of the text (not W. H.) in Ac. 10:36, we have to.n lo,gon o[n.164 Sometimes anacoluthon occurs also as in pa/n r`h/ma avrgo.n o[- peri. auvtou/, Mt. 12:36; pa/j oa}j evrei/- avfeqh,setai auvtw|/├ Lu. 12:10; panti. w|- evdo,qh- zhthqh,setai par v auvtou/, 12:48; pa/n oa} de,dwken- evx auvtou/, Jo. 6:39; pa.n oa} de,dwkaj auvtw|/ dw,sei auvtoi/j, 17:2. In 2 Cor. 12:17, mh, tina w-n - di v auvtou/, we have anacoluthon, but not attraction. In Mt. 25:24, suna,geij o[qen ouv diesko,rpisaj, we have evkei/qen o[pou shortened to o[qen. There is not inverse attraction in ouvdei.j o[j (1 Cor. 6:5) since e;ni precedes ouvdei,j.

( e) Incorporation. But the most striking instance of this close unity between antecedent and relative is the incorporation of the antecedent into the relative clause with identity of case. I count 54 such examples in Moulton and Geden.165 They are fairly well distributed through the different portions of the New Testament. 1) The simplest form of such incorporation is where no change of case is required. Thus Lu. 24:1, fe,rousai aa} h`toi,masan avrw,mata; Jo. 6:14, ivdo,ntej aa} evpoi,hsen shmei/a (W. H.); Mt. 7:2, evn w|- ga.r kri,mati kri,nete kriqh,sesqe├ kai. evn w|- me,trw| metrei/te metrhqh,seati u`mi/n= Mk. 4: 24 = Lu. 6:38; Mt. 24:44, h|- ouv dokei/te w[ra|=Lu. 12:40 (not Mt. 24:50). For further examples of this simple incorporation see Mt. 23:37 = Lu. 13:34 (the set phrase, adverbial accusative, oa}n tro,pon), so also Ac. 1:11; 7:28; 15:11; 27:25; Mk. 2:19 ( o[son cro,non; but not Lu. 12:46= Mt. 24:50); Lu. 17:29 f.; Jo. 9:14; 11:6; 17:3; Ac. 7:20; 25:18; probably 26:7; Ro. 2:16; 7: 19; 9:24 ( ou[j- h`ma/j note); 16:2; Ph. 3:18 (but probably only predicate accusative like Mk. 15:12); 2 Tim. 1:6 ( di v h[n). In 1 Jo. 2:25 there is not exactly incorporation, but apposition to the relative. In Lu. 8:47; Ac. 22:24 and Heb. 2:11 the case is the same also, but the preposition would have been needed only with the relative. Cf. Phil. 1:10; 2 Tim. 1:12; Heb. 13:11. See w-n- ponhrw/n, Ac. 25:18, where there is incorporation and attraction to the case of the antecedent. The same thing is true

PRONOUNS ( vANTWNUMIAI) 719

of Rev. 17:8, where blepo,ntwn agrees with w-n. In Heb. 13:11, w-n zw,wn- tou,twn, the substantive is incorporated, but the demonstrative is repeated afterwards. Cf. also oa}──auvto. tou/to (Gal. 2:10). It is possible that Ro. 4:17 belongs here, the preposition kate,nanti, being understood twice. The same thing may be true of Lu. 1:4, peri. w-n kathch,qhj lo,gwn th.n avsfa,leian (either lo,gwn [or peri. lo,gwn] peri. w-n or peri. lo,gwn ou[j).

2) But sometimes besides incorporation there has resulted a change of case also. The antecedent may be drawn into the case of the relative (cf. inverse attraction) as in Mk. 6:16, o[n evgw. avpe─ kefa,lisa vIwa,nhn ou-toj hvge,rqh. Here the demonstrative pronoun is resumptive. The change is made from nominative to accusative. The same thing is true of the spurious passage in Jo. 5:4, 4.) dh,pote katei,ceto nosh,mati (change from genitive to instrumental). This is probably true of Ac. 21:16, a;gontej par v w|- xenisqw/men Mna,─ swni, tini Kupri,w|. The resolution of this passage is not certain, but it may be a;gontej Mna,swna par v w|- (change from accusative to locative).166 But pro.j Mna,swna may be correct.

In Ro. 6:17, u`phkou,sate eivj oa}n paredo,qhte tu,pon didach/j, the resolved form would probably be tu,pw| didach/j eivj oa}n paredo,qhte. In Heb. 7:14, eivj ha}n fulh,n, the substantive would have been in apposition with evx vIou,da (the ablative). In Heb. 10:10 evn w|- qelh,mati the accusative to. qe,lhma is present in the preceding sentence. The same thing is true of 1 Pet. 1:10, peri. h-j swthri,aj ( swthri,an just before). In 2 Cor. 10:13 we have in the same sentence the substantive repeated (once incorporated and attracted to the case of the relative, but the relative itself attracted to the case of kano,noj), kata. to. me,tron tou/ kano,nouj ou- evme,risen h`mi/n o` qeo.j me,trou.

3) In a few instances the attraction has been that of the relative to the case of the antecedent, transferred to the relative clause. See Ac. 25:18, w-n evgw. u`peno,oun ponhrw/n. For examples with prepositions (see chapter on Prepositions) note: peri. pa,ntwn w-n evpoi,shen ponhrw/n (Lu. 3:19), peri. paw/n w-n ei=don duna,mewn grk(19:37), where the incorporation is only partial. It is clear therefore that in the great majority of instances there is no change of case required. Very many also are set phrases like oa}n tro,pon├ h|- w[ra|├ h|- h`me,ra|├ di v ha}n aivti,an, etc. For presence of the antecedent see Jo. 16:17 f.

7. Absence of Antecedent. It so often happens that the relative has no antecedent that it calls for special consideration.

720 A GRAMMAR OF THE GREEK NEW TESTAMENT

The clause indeed often becomes a substantive rather than an adjective clause. [Oj thus occurs in general statements as in Mt. 10:14; 23:16, 18 (cf. also pa/j o[j, Lu. 12:48; 14:33; Ac. 2:21; Gal. 3:10). Blass (Gr. of N. T. Gk., p. 173) gives a large number of such instances of the general or indefinite use of o[j. So as o[j e;cei w=ta avkou,ein avkoue,tw (Mk. 4:9), where the relative clause is the subject of avkoue,tw. This is the indefinite relative. Cf. Mk. 4:25. Here the relative and the antecedent (if expressed) are in the same case (nominative). Cf. 1 Cor. 15:10, eivmi, o[ eivmi; Lu. 9:50, etc. Both may be in the accusative as in oa} de. u`mi/n le,gw pa/sin le,gw (Mk. 13:37), mh. eivdw.j oa} le,gei (Lu. 9:33). Cf. Mk. 15:12; Lu. 11:6; Jo. 1:45; 6:29; 19:37, etc. But the relative may be in the accusative when the antecedent would have been in the nominative. So oa} lalei/ gi,netai (Mk. 11:23). Cf. Jo. 1:26; 4:18, etc.

So both may be examples of the genitive, as suggenh.j w'n ou- avpe,─ koyen Pe,troj to. wvti,on (Jo. 18:26) where ou-= tou,tou ou-) So in 1 Cor. 7:1 peri. w-n╩peri. tou,twn (or pragma,twn) peri. w-n. But in a;cri ou- (Rev. 2:25) we really have a;cri kairou/ w|- (or evn w|-). In Lu. 23:41, a;xia w-n evpra,xamen, the resolution is tou,twn a[ (gen. and acc.). So in Jo. 17:9 peri. w-n de,dwkaj╩peri. tou,twn ou[j. In Ac. 21:24 chntai peri. sou/ ouvde,n╩tou,twn a[, etc. Exactly so w-n in Lu. 9:36; 23: 14; Ac. 8:24; 22:15; 25:11; Ro. 15:18; 2 Cor. 12:17. In Ac. 26:16, ma,rtura w-n te ei=de,j me w-n te ovfqh,somai, soi, it is the second w-n that gives trouble. The antecedent would be tou,twn and the relative before attraction either a (ace. of general reference) or oi-j (locative or instrumental). In Ro. 4:7 w-n has as its unexpressed antecedent ou-toi. CF. also Ac. 13:25. In Mt. 6:8 (so Jo. 13:29), w-n crei,an, the antecedent would be in the accusative. So also peri. w-n. Ac. 24:13. In Lu. 17:1 di v ou- is resolved into tou,tw| di v ou- (dative). In Ro. 10:14, pw/j pisteu,swsin ou- ouvk h;kousan, we probably have ou-╩eivj tou/ton (or tou,tw|) ou-.

The examples of the ablative are not many. See Jo. 7:31 where w-n after plei,ona shmei/a is to he resolved into tou,twn a[ (abl. and acc.). So in Ac. 26:22 evkto.j w-n╩ evkto.j tou,twn a[. In Heb. 5:8 avf v w-n╩ avpo. tou,twn a[, while in 2 Cor. 2:3 avf v w-n ╩ avpo. tou,twn avf v w-n) Cf. Lu. 6:34, par v w-n; 1 Cor. 10:30. In Ac. 13:39, avpo. pa,ntwn w-n, the one preposition covers both ablatives.

For the dative I note oi-j de,dotai (Mt. 19:11), where the antecedent like pa,ntej would have been in the nominative. Cf. Lu. 7:43, 47 w|-; Ro. 15:21 oi-j and 2 Pet. 1:9 w|-. In 1 Cor. 7:39, w|- qe,lei gamhqh/nai, the antecedent would have been in the dative also. So also 2 Cor. 2 : 10 w|-* Ro. 6:16 w|- twice. In 2 Tim. 1.12, oi=da w|-

PRONOUNS ( vANTWNUMIAI) 721

pepi,steuka├ it is the accusative rather followed by dative, auvton w|-. In Mt. 20:23 (Mk. 10:40) the antecedent of oi-j is probably tou,twn. In Ro. 10:14 the antecedent of ou- would be tou,tw|.

Some few examples of the locative appear also. Cf. evf v oi-j, Ro. 6:21, where the antecedent would have been evpi. tou,toij. So Ro. 2:1 and 14:22 evn w|- implies evn tou,tw| (cf. also 1 Pet. 2:12; 3:16), but not so verse 21 where evn w|- refers to an involved ti or mhde,n. In Ro. 7:6 evn w|- may involve tou,tw| evn w|-. In Heb. 2:18 evn w|- (= evn tou,tw| evn w|-) really has a causal force. In Ph. 4:11 evn oi-j╩evn tou, toij evn oi-j, but in 2 Tim. 3:14 evn oi-j╩evn tou,toij aa}. Cf. 2 Pet. 2:12 (but tau/ta evn oi-j may be correct).

I have noticed no examples of the instrumental. But great freedom and variety are manifest.

8. Prepositions with the Antecedent and Relative. The preposition may be used twice167 "in the case of a, sharper division of the relative clause." So eivj th.n gh/n tau,thn├ eivj h[n, Ac. 7:4; avpo. prw,thj h`me,raj avf v h-j 20:18. Then again the preposition may occur with the antecedent, but not with the relative, though implied, as in evn panti. cro,nw| w|- eivsh/lqen Ac. 1:21. So the margin in Ro. 2:16 evn hpme,ra| h|-. Cf. Lu. 1:25. It is possible also so to understand evn th|/ o`dw|/ h|- h;rcou Ac. 9:17. But it is clearly true of avpo. pa,n─ twn w-n, Ac. 13:39.

On the other hand the preposition may occur with the relative, but not with the antecedent. Thus evkei,nh| th|/ w[ra| evn h|-├ Jo. 4:53. When the antecedent is absent, the preposition may be the one common to both, as in avf v w-n (2 Cor. 2:3), or which belongs to only one. Cf. par v w-n (Lu. 6:34), evf v oi-j (Ro. 6:21), evn oi-j (Ph. 4:11), u`pe.r ou- (1 Cor. 10:30), evn w|- (Ro. 14:22), as eivj o[n (Ro. 10:14), peri. w-n (1 Cor. 7:1), etc. This "one" may be the antecedent, as in the following examples, eivj o[n (Jo. 6:29) = eivj tou/ton o[n├ peri. w-n (Jo. 17:9) = peri, tou,twn ou[j├ u`pe.r a[ (1 Cor. 4:6) = u`pe.r tau/ta a[├ avf v w-n, (Heb. 5:8)= avpo. tou,twn a[├ eivj o[n (Jo. 19:37)= eivj tou/ton o[n├ etc. Or the "one" may be the relative, as di v ou- (Lu. 17:1)= tou,tw| di v ou-├ evf v o[n (Heb. 7:13)= ou-toj evf v o[n, etc. The use of prepositions is common in the same way with the relative and its incorporated antecedent. See evn w|- kri,mati (Mt. 7:2), a;cri h-j h`me,raj (Lu. 1:20), di v ha}n aivti,an (Lu. 8:47), par v w|-──Mna,swni (Ac. 21:16), eivj oa}n- tu,pon (Ro. 6:17), avf v h-j h`me,raj (Col. 1:9), peri. h-j swthri,aj (1 Pet. 1:10), etc. Cf. Ro. 16:2.

9. Relative Phrases. Some of the abbreviated prepositional clauses come to be used at the beginning of principal sentences

722 A GRAMMAR OF THE GREEK NEW TESTAMENT

like the free use of conjunctions and relatives. Cf. Latin use of qui. Cf. Draeger, Hist. Syntax, Bd. II, p. 512. So avnq v w-n (Lu. 12:3), evn oi-jgrk grk(12:1), dio, (Heb. 3:7), peri. w-n (1 Cor. 7:1), ou- ca,rin (Lu. 7:47), di v ha}n aivti,an (2 Tim. 1:6). Cf. o[qen (Heb. 3:1). Indeed (Winer-Schmiedel, p. 228) evn w|- may be here equal to tou,tw| o[ti├ avnq v w-n╩ avnti. tou,twn o[ti├ evf v w|-╩evpi. tou,tw| o[ti (2 Cor. 5:4), dio,ti (1 Th. 2:8)= dia. tou/to o[ti├ evf v oi-j (Ro. 6:21), etc. The temporal and causal use of the relative phrases is common. Cf. evn w|- (Heb. 2:18). Indeed kaqo, (Ro. 8:26) is kaq v o[[j kaqo,ti (Ac. 2:45) is kaq v o[ti├ kaqa,per (Ro. 4:6) is kaq v a[per. Cf. evf v o[son (Mt. 9:15), kaq v o[son (Heb. 3:3).

Adverbs show the same phenomena as other relative forms. Thus in Ro. 5:20 ou- has no antecedent. In 1 Cor. 16:6 ou-╩ evkei/se ou-. So o[pou in Jo. 11:32 ╩evkei/se o[pou and in Jo. 20:19 = evntau/qa o[pou. In 2 Sam. 14:15 o[= conjunction.

10. Pleonastic Antecedent. The redundant antecedent incorporated into the relative clause has attracted considerable attention. In Herodotus 4, 44 oa}j- ou-toj occurs,168 and Blass169 cites Hyper. Eux. ž 3, w-n- tou,twn. But in ancient Greek it was a very rare usage. In Winer-Schmiedel170 examples of pleonastic ou-toj are cited from Xenophon, Diodorus Siculus, Pausanias, Sophocles. Pleonastic auvto,j appears in Aristophanes, Birds, 1237, oi-j qute,on auvtoi/j) Reference also is made to Sophocles and Lucian. In the LXX the idiom is extremely common, manifestly under the influence of the Hebrew Al rv,a] (cf. Aramaic d.). It "is found in all parts of the LXX and undoubtedly owes its frequency to the Hebrew original. But the fact that it is found in an original Greek work, such as 2 Macc. (xii, 27 evn h|- ) ) ) evn autth|/) and a paraphrase such as 1 Esdras (iii, 5, 9; iv, 54, 63; vi, 32), is sufficient to warrant its presence in the koinh,."171 For numerous examples of the idiom in the LXX see Winer-Schmiedel, p. 200, and Winer-Moulton, p. 185. Cf. also Conybeare and Stock, Selections, pp. 65 ff. As a matter of fact the examples are not very numerous in the N. T. It occurs several times in Rev. Rev.(3:8 ha}n- auvth,n, 7:2 oi-j evdo,qh auvtoi/j, 7:9 oa}n- auvto,n, 13:8 ou-──auvtou/, 20:8 w-n- auvtw/n%) Outside of the Apocalypse, which so strongly bears the influence of the LXX, the usage is infrequent. See Mt. 3:12, ou- to. ptu,on evn th|/ ceiri. auvtou/, an example hardly parallel as a matter of fact. But a clearer instance is Mk. 1:7 ( = Lu. 3:16 f.), ou-──auvtou/, and still more so 7:25, h-j ei=ce to. quga,trion auvth/j. Cf. also oi[a- toiau,th (Mk. 13:19), oi-oj-

PRONOUNS ( vANTWNUMIAI) 723

thlikou/toj (Rev. 16:18), oi[a- ou[twj (Mk. 9:3), o[pou- evkei/ (Rev. 12:6, 14), o[pou- evp v auvtw/n, (Rev. 17:9).172 In Arc. 15:17, evf v oua}j- evp v auvtou,j, we have a quotation from the LXX (Amos 9:12). "The N. T. examples are all from places where Aramaic sources are certain or suspected" (Moulton, Prol., p. 95). One almost wonders, after this admission, why Moulton, p. 94, seems so anxious to prove that the idiom in the N. T. is not a Hebraism. By his own admission it seems a practical Hebraism there, though the idiom had an independent development in the Greek. The early sporadic examples in the ancient Greek173 blossom out in the later Greek again and in the modern Greek become very common. Psichari174 considers it rather far-fetched in Moulton to appeal to the modern Greek vernacular, o` giatro.j pou/ to.n e;steila, 'the doctor whom I sent for,' since the modern Greek vernacular just as readily uses pou/ without auvto,n. Psichari complains that Thumb175 also has not explained clearly this idiom. But Psichari, believes that the idiom existed in the vernacular koinh, (and so fell in readily with the Hebrew usage) and has persisted to the present day. He considers176 the example from a papyrus of the third century A.D. (P.Oxy. I, 117,15) decisive, evx w-n - evx auvtw/n. See also P. Amh. II, 11, 26, o[per fanero.n tou/to evge,neto. Moulton177 has given abundant examples from Old English. So in Chaucer (Knightes Tale, 1851 f.):

" Namely oon,

That with a spore was thirled his brest-boon." He compares also the German der du bist. Simcox178 cites vernacular English "a thing which I don't like it." Evidently therefore the idiom has had independent development in various languages in the vernacular. According to Jannaris (Hist. Gk. Gr., p. 353) the relative is in such cases regarded as "a mere, connective."

In Gal. 3:1, oi-j - evn u`mi/n, W. H. reject evn u`mi/n) In Gal. 2:10, oa} auvto. tou/to├ we have the intensive use of auvto,, but tou/to is pleonastic. In 1 Pet. 2:24, oa}j - auvto,j, we have again intensive auvto,j.

11. The Repetition of o[j. Winer179 rightly remarks that it is a misapprehension of the Greek genius to expect the relative rather than auvto,j or ou-toj in a case like Jo. 1:7; Lu. 2:36; 19:2; Ac.

724 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 2nd ed.

10:36. The old Greek could, and commonly did,180 use ou-toj or more usually auvto,j with kai, to continue the narrative. Blass181 rather curiously calls it "negligent usage." Cf. Lu. 13:4, evf v oua}j e;pesen o` pu,rgoj kai. avpe,kteinen auvtou,j; 1 Cor. 8:6, evx ou-──kai. eivj auto,n and di v ou- - kai. di v auvtou/ (cf. Heb. 11:4); 2 Pet. 2:3, oi-j - kai. auvtw/n; Rev. 17:2, meq v h-j - kai. auvth/j. In Lu. 17:31 kai. o` occurs rather than kai. auvto,j. Cf. Jo. 13:24. In Jo. 1:33, evf v oa}n - kai. evp v auvto,n, the repetition of the relative would have been impracticable. But in 1 Cor. 7:13 Paul might very well have written h[tij- kai. o[j rather than kai. ou-toj (a sort of parenthesis). It is common,182 also, to have neither the relative repeated nor the demonstrative. So o[j ge tou/ ivdi,ou ui`ou/ ouvk evfei,sato├ avlla. u`pe.r h`mw/n pa,ntwn pare,dwken auvto,n, (Ro. 8:32). Cf. Ph. 4:9.

But the relative may be repeated. A good many such examples occur in the N. T. Kai, may be used, as w-n kai. w-n (Ro. 4:7). Cf. also ou-- w|- kai, (Ac. 27:23) and w-n te - w-n te - (Ac. 26:16). Cf. 1 Cor. 15:1 f., oa}──oa} kai. ── evn w|- kai. ── di v ou- kai.. See Jo. 21:20. But examples occur also of the repetition of the relative without any conjunction, as in oa}j - oa}n- par v ou- (Ac. 24:6). See 1 Cor. 4:17. Cf. o[sa - o[sa, etc. (Ph. 4:8). This repetition of o[j is specially frequent in Paul. Cf. Col. 1:24, 28 f.; Eph. 3:11 f.; 1 Cor. 2:7 f., though it is not exactly "peculiar" to him (WinerMoulton, p. 209). In 1 Jo. 1:1 oa} is repeated without conjunction three times, while in verse 3 oa} is not repeated with the second verb. In 1 Pet. 1: 6-12 four sentences begin with a relative. In Ro. 9:4 f. we have oi[tinej - w-n- w-n - kai. evx w-n.

The use of avnq v w-n o[sa together (Lu. 12:3) finds abundant parallel in the LXX, easily falling in with the Hebrew construction183 with rv,a]. Thus a double relative occurs.

In Ro. 4:21 the conjunction of o[ti o[ is merely accidental; but that is not true of o[ - o[ti in 1 Jo. 4:3. Cf. also oi-on o[ti in Ro. 9:6.

12. A Consecutive Idea. This may be implied in o[j. Thus in Lu. 7:4, a;xioj evstin w|- pare,xh| tou/to. One is reminded of qui in Latin.184 Cf. also ti,j evstin ou-toj o[j kai. a`marti,aj avfi,hsin; (Lu. 7:49). A particularly good example is 1 Cor. 2:16, ti,j ga.r e;gnw nou/n kuri,ou├ oa}j sunbiba,sei auvto,n; See chapter XIX, Mode.

13. Causal. [Oj may also introduce a causal sentence. So o[j

PRONOUNS ( vANTWNUMIAI) 725

ge in Ro. 8:32. Cf. Latin quippe qui. This is perfectly regular in ancient Attic. Cf. Thompson, Syntax of Attic Greek, p. 374. See also chapter XIX, Mode.

14. In Direct Questions. The passage in Mt. 26:50, e`tai/re├ evf v oa} pa,rei, is the only one in the N. T. where such a construction is possible. There is no doubt as to the occasional use of o[stij (see (e), 9), o`po,soj├ o`po,teroj├ o[pwj in direct questions in the ancient Greek. For examples see Jannaris, Hist. Gk. Gr., p. 473 f. See further chapter XIX, Mode. This double use of relative pronouns is on a par with the double use of interrogative stems (cf. indefinite) so common in the Indo-Germanic tongues.185 The Latin qui and quis are kin in root and usage. Moulton186 rightly considers it "superfluous to say that this usage cannot possibly be extended to direct question." Winer187 explained the "misuse" as belonging to late Greek. A few examples188 of o[j in a direct question do occur. So in Euseb., P. E. vi, 7: 257 d, Gaisford edition, w-n e[neka; Just., Cohort. 5 (p. 253 A), di v ha}n aivti,an - prose,ceij `Omh,rw|; Apophth., 105 C, vArse,nie├ di v oa} evxh/lqej. Certainly the idiom was chiefly in the vernacular and rare even there. Blass189 conjectures a slip in the text, ai=re having been changed to e`tai/re, and Chrysostom had an imperative in his text. We may suppose "a rather harsh ellipsis" of the principal verb and treat it as an ordinary relative.190 [Oj may indeed here be demonstrative as suggested by Noah K. Davis.191 There was undoubtedly in the later Greek considerable confusion in the use of the relatives and the interrogatives. It is not impossible for o[j here to be interrogative. That is as much as one can at present say. Blass thought it "quite incredible."

15. In Indirect Questions. Here the matter is much clearer. Even Blass192 admits that "relatives and interrogatives become confused in Greek as in other languages." In the classical language o[j (still more o[stij) is "frequently" so employed. This use comes from Homer on down and occurs in Aristophanes, Sophocles, Herodotus, Xenophon, Plato, Lysias. Thucydides193 uses it side by side with o[stij. The papyri have it as Moulton has shown.194

726 A GRAMMAR OF THE GREEK NEW TESTAMENT

Cf. fra,zontej evn h-i kw,mhi oivkou/sin├ R. L. 29 (iii/B.C.); fronti,saj di v w-n dei/ tau/ta evrgasqh/nai, P.P. ii. 37 (ii/B.C.). It is a little surprising, however, to find Blass195 saying that this usage "is wanting in the N. T." W. F. Moulton196 in his footnote gives undoubted examples of o[j in indirect questions after verbs of knowing, declaring, etc. So oi=den- w-n crei,an e;cete, Mt. 6:81; avpaggei,late aa} avkou,ete, 11:4; eivdui/a oa} ge,gonen├ Mk. 5:33; avne,gnwte oa} evpoi,hsen├ Lu. 6:3 (cf. Mt. 12:3 ti,%* mh. eivdw.j oa} le,gei,, 9:33; di v ha}n aivti,an h[yato auvtou/ avph,ggeilen, 8:47 (cf. Ac. 22:24); dida,xei u`ma/j a[ dei/ eivpei/n, 12:12. But not 2 Tim. 1:12. And then in 1 Tim. 1:7 we find aa} le,gousin and peri. ti,nwn diabe─ baiou/ntai used side by side after mh. noou/ntej. Cf. also Jo. 18:21. One may compare197 also Lu. 11:6, ouvk e;cw oa} paraqh,sw auvtw|/, with Mk. 8:2 (Mt. 15:32), ouvk e;cousin ti, fa,gwsin. See also w`j iva,qh in Lu. 8 : 47, and note w`j in Lu. 23:55; 24:35, not to mention o[soj├ o`poi/oj.

16. The Idiom ouvdei,j evstin o[j. It occurs in the N. T., as Mk. 9:39; 10:29; Lu. 1:61; 18:29; 1 Cor. 6:5. For ouvdei,j evstin oa}j ouv see Mt. 10:26 (cf. Lu. 8:17). Here one is reminded of the old idiom ouvdei.j o[stij. Mayser (Grammatik, p. 310) calls attention to the papyri use of o[n╩o[ after analogy of tosou/to$n%. Cf. ti,j- oa}j ouv in Ac. 19:35. The N. T. does not use198 e;stin o[j├ eivsi.n oi[╩ti.j├ tine,j.

(e) [Ostij.

1. Varied Uses. The form is, of course, merely o[j and tij. But we have seen a variety of uses of o[j, and tij likewise is not entirely uniform. Hence the combination cannot be expected to be so.

2. The Distinction between o[j and o[stij. It was not ironclad in the ancient language, as may be seen by reference to the Epic, Ionic, Attic poets, and to Herodotus (once Thucydides).199 Blass200 finds that the distinction between them is no longer regularly preserved in the N. T., least, of all in Luke, best of all in Paul. Moulton201 finds some examples in the papyri of o[stij in the sense of o[j, but doubts if the two relatives are ever absolutely convertible and thinks that on the whole the classical distinction remains undisturbed, though sometimes during the koinh, period it had worn rather thin.202 But Jannaris203 holds that o[stij, having a wider scope

PRONOUNS ( vANTWNUMIAI) 727

than o[j, in postclassical times was used indiscriminately for o[j. He is supported by Kaelker about Polybius.204 But in the vernacular modern Greek o[ti is alone common, other forms of o[stij being rare, though o[tinoj and o[tinwn are found (Thumb Handb., p. 93 f.). Kruger205 calls o[j "objective" and o[stij "qualitative and generic." W. F. Moulton206 defines o[stij as properly indicating the class or kind to which an object belongs. But no exact parallel can be drawn nor uniform distinction preserved. Each has its own history. Jebb207 takes o[stij to refer to class in ancient Greek and hence is either indefinite or causal. In the modern Greek it is still indefinite, but has also in the vernacular displaced o[j in the masculine and feminine nominative. In the LXX o;stij is less frequent than o[j and is almost confined to the nominative and accusative.208 In the papyri209 it is less frequent than o[j and is usually in the nominative as in the N. T. (Moulton, Cl. Rev., 1904, p. 154).

3. The Indefinite Use. This is, as a matter of fact, still the least frequent in the N. T. There are about 27 of the indefinite and 120 of the definite use (Scott). Cf. o[stij se. r`api,zei eivj th.n dexia.n siago,na (Mt. 5:39), o[stij avrnh,shtai, megrk grk(10:33), o[ti a'n aivth,shte (Jo. 14:13), o[stij eva.n h|- (Gal. 5:10). Thus it is used with indicative or subjunctive, with or without a;n ( eva,n). Cf. Mt. 13:12. In Mk. 8:34 ei; tij does not differ very greatly from o[stij. Cf. also eva,n mh,, Mk. 10:30. Pa/j o[stij is, of course, indefinite also. Thus Mt. 7:24; pa/n o[ ti eva.n poih/te (Col. 3:17), etc. For pa/sa yuch. h[tij a;n see Ac. 3:23 (LXX). In P. Par. 574 (iii/A.D.) note o[stij pot v ou=n ei=.

4. The Definite Examples. These are partly causal clauses. Some indeed seem merely descriptive. Thus Mt. 7:15, tw/n yeudoprofhtw/n o[tinej e;rcontai. Cf. also Mt. 7:26; 13:52; 21:33, etc. The value of the pronoun sometimes does not differ greatly from oi-oj and expresses quality. Thus euvnou/coi oi[tinej, Mt. 19:12; a;lloij gewrgoi/j oi[tinej, 21:41; parqe,noij ai[tinej, 25:1, etc. Once indeed we actually have toiau,th h[tij (1 Cor. 5:1). Cf. also potaph. h` gunh. h[tij (Lu. 7:39). See also Gal. 4:24, 26. Then again it may be merely explanatory as in gunai/kej pollai.──ai[tinej hvkolou,qhsan tw|/ vIhsou/ (Mt. 27:55). Cf. Mk. 15:7; Lu. 12:1; Col. 3:5; Rev. 11:8, etc. This use of o[stij is particularly frequent with proper names.

5 Thack., Gr., p. 192.

728 A GRAMMAR OF THE GREEK NEW TESTAMENT

So Lu. 2:4, eivj po,lin Dauei.d h[tij kalei/tai Bhqlee,m. Cf. also Lu. 8:26; Ac. 16:12, etc. Abbott, Joh. Gr., p. 303, takes the explanatory or illustrative examples= 'now he,' one that.' Moulton210 points out that o[stij at the beginning of a parable (cf. Mt. 20:1) is really a type and so appropriate. In an example like Lu. 1:20, toi/j lo,goij mou oi[tinej plhrwqh,sontai, Moulton takes it to be 'which for all that' (almost adversative), while in Lu. 10:42 h[tij ouvk avfaireqh,setai auvth/j= 'and it shall not be taken away from her.' There is no doubt about the causal use of o[stij (cf. qui and quippe qui). See Jo. 8:53, vAbraa.m o[stij avpe,qanen ('seeing that he died'); Ac. 10:47, oi[tinej to. pneu/ma to. a[gion e;labon ('since they received the Holy Spirit'). Cf. also Ac. 7:53; Ro. 2:15; 6:2; Heb. 8:6; 10:35; Eph. 3:13; Ph. 4:3; Col. 3:5; Jas. 4:14; 1 Pet. 2:11, etc.

5. Value of o[j? It is a matter of dispute whether in the N. T., as usually in modern Greek, o[stij has come already to have merely the force of o[j. There are undoubted examples where it is equal to o[sper ('which very'). So Ac. 11:28, h[tij evge,neto├ evpi. Klaudi,ou. Cf. also Ac. 13:31; 16:16; 1 Cor. 3:17, etc. Blass211 goes further and finds o[stij in Luke purely in the sense of o[j. He is supported by Jebb212 who says that "no natural interpretation can make it more in Lu. 2:4." In Acts at any rate a fairly good case can be made out for this weakened sense of o[stij. Cf. 8:14 f. Pe,tron kai. vIwa,nhn oi[tinej, 12:10 th.n pu,lhn h[tij, 17:10. See also Rev. 12:13. Moulton213 gives an exact parallel from the papyri for Mt. 27:62, th|/ evpau,rion h[tij evsti.n meta. th.n paraskeuh,n ( au;rion h[tij evsti.n ie). He quotes Hort also (Comm., 1 Pet. 2:11) in favour of the position that in some places in the N. T. no distinction can be drawn between o[j and o[stij) Blass214 denies that Paul uses o[stij as the equivalent of o[j. I confess that I fail to see a great deal of difference between o[itinej and oi-j in Ro. 16:4, oi[tinej and oi[ in 16:7. Cf. also o[j and h[tij in verses 5 f.

6. Case. There is little here that calls for comment. We do not have attraction or incorporation. As a matter of fact only three cases occur (nom., gen., ace.).215 The stereotyped phrase

PRONOUNS ( vANTWNUMIAI) 729

with e[wj and the genitive, e[wj o[tou, occurs five times. Cf. Mt. 5:25; Lu. 12:50 (Luke three times, Matthew and John once each). This is the only form of the shortened inflection. The LXX once216 (2 Macc. 5:10) has h[stinoj, elsewhere o[tou . The accusative is found in the N. T. only in the neuter singular o[ti (absent from modern Greek). But see (note 6, p. 728) occasional o[ntina and h[ntina in the papyri. So Lu. 10:35, o[ti avn prosdapanh,sh|j. Cf. o[ti a;n, Jo. 2:5; 14:13; 15:16; o[ti eva,n, Mk. 6:23; 1 Cor. 16: 2 f.; Col. 3:17; o[ti alone, Jo. 8:25; Ac. 9:6. The other examples are all in the nominative. In Ac. 9:6 the clause is nominative.

7. Number. In general the number of o[stij agrees with that of the antecedent. But in a few instances o[stij agrees with the predicate. So with 1 Cor. 3:17, nao.j oi[tinej- u`mei/j, Eph. 3:13, qli,yesin h[tij- do,xa. Cf. Ac. 16:12.

8. Gender. Likewise o[stij in general agrees with the antecedent in gender. So Eph. 1:22 f. evkklhsi,a h[tij- to. sw/ma, Gal. 4:24 mi,a h[tij── [Agar. Cf. Rev. 11:8. But the gender of the predicate may be followed as in Ac. 16:12, Fili,ppouj (fem., H. Scott says, but Thayer has oi`) h[tij - po,lij; 1 Tim. 3:15, oi;kw| qeou/ - h[tij - evkklhsi,a. In Ph. 1:28, h[tij - e;ndeixij, the antecedent is the general idea of the preceding clause. One example of o[ti is neuter singular (2 Cor. 3:14, o[ti evn Cristw|/ katargei/tai), and several times the neuter plural (Jo. 21:25, a[tina eva.n gra,fh─ tai). So Gal. 4:24; 5:19. Cf. the absence of the neuter in the modern Greek. The masculine and feminine, both singular and plural, are very frequent. Cf. Mt. 2:6; 7:15; Lu. 2:4; 23: 55. See further for number, gender and case, chapter X, VII, VIII, IX.

9. Direct Questions. Examples of o[stij in direct questions are found in Aristophanes and Plato as quoted by Jannaris.217 An example of it occurs also in 1 Chron. 17:6, o[ti ouvk wvkodomh,sate, moi oi=kon ke,drinon; Here the Hebrew has hm'l'. Cf. also 2 Ki. 8:14 in AB, o[ti where other MSS. have ti,. In Barn. Ep. c. 10 we have o[ti de. Mwu?sh/j ei;rhken; Vulgate has quare.218 Jannaris219 gives a number of instances for the later Greek. And yet Blass220 calls it "quite incredible," a remark impossible to justify in the light of the facts. It is, indeed, unusual, but there is no a priori reason

730 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

why the N. T. writers could not occasionally use o[stij as a direct interrogative. One may note also the use of el in a direct question.221 The N. T. examples are all confined to o[ ti. In Mt. 7:14 o[ti is certainly merely causal, not exclamatory nor interrogative. In Mk. 2:16 o[ti (sec.) read by BL 33, is accepted by W. H. and Nestle as interrogative. AC al. read ti, o[ti, while aD have dia. ti,. It is possible, to be sure, that o[ti may be an "abbreviation"222 or "ellipsis"223 for ti, o[ti. But it is more probable that it is here regarded as tantamount to an interrogative ( ti, o[ti or dia. ti,). Moulton (Cl. Rev., 1904, p. 154) quotes o[ti ti, in B.U. 607 (ii/A.D.) gra,yon moi o[ti ti, e;praxaj. But in Mk. 9:11 the Greek uncials all give the first o[ti. This is all the more remarkable since the second o[ti is clearly a conjunction. The Latin MSS. give variously quare, quia, quid, etc., and some Greek cursives pw/j ou=n. 'Why' is the natural and obvious idea.224 So in Mk. 9:28 o[ti is read by the great mass of MSS. (including aBCL), though AD and a number of others have dia. ti,, some even have o[ti dia. ti, (conflate reading), a few ti, o[ti. In John 8:25 both W. H. and Nestle print as a question, Th.n avrch.n o[ti kai. lalw/ u`mi/n* The Latin versions have quod or quia. It is a very difficult passage at best. Th.n avrch.n o[ti may be taken to mean 'Why do I speak to you at all?' ( th.n avrch,n ╩ o[lwj). But there may be ellipsis,225 'Why do you reproach me that ( o[ti% I speak to you at all?' If necessary to the sense, o[ti may be taken here as interrogative.226 Moulton227 admits the N. T. use of o[stij in a direct question. Recitative o[ti is even suggested in Winer-Schmiede1,228 but the occasional interrogative use of o[ti is sufficient explanation. But the passage in Jo. 8:25 is more than doubtful. Chrysostom takes o[ti there as relative, Cyril as causal.229

10. Indirect Questions. In ancient Greek o[stij is exceedingly common in indirect questions, sharing the honours with ti,j.230 The astonishing thing about this use of o[stij is its almost entire absence from the N. T. (cf. modern) Greek, where it is not used in this sense). No example has yet been shown from the papyri. Indeed the relative forms, the se-called indirect interrogatives, are not common in the N. T. in that sense. The direct interroga-

PRONOUNS ( vANTWNUMIAI% 731

tives are the rule in the N. T. in indirect questions.231 Only one instance of o[ti in an indirect question is found in the N. T., Ac. 9:6, lalhqh,setai, soi o[ti se dei/ poiei/n. Even this reading, though supported by aABC, Blass232 rejects "in view of the general practice elsewhere," a needless conclusion. Why not call it a "literary" mark in Luke? [Opwj is so used once (Lu. 24:20), o[pou not at all (not even Jo. 14:4), oi-oj in 1 Th. 1:5, and o`poi/oj only in 1 Cor. 3:13; Gal. 2:6; 1 Th. 1:9; Jas. 1:24. See further chapter XIX.

(f) Oi-oj.

1. Relation to o[j. This correlative form is related to o[j as qualis is to qui. The antecedent toiou/toj is not, of course, always expressed. But it is qualitative, and not a mere relative like o[j or even o[stij. In the modern Greek the word has disappeared except the form o[gioj ( o` oi-oj%233 in the dialects and is rare (14 times) in the N. T. Mayser234 merely mentions it in his Grammatik d. griech. Papyri. It is in the N. T. usually without toiou/toj, as in Mt. 24:21, but it is several times followed by toiou/toj, as in 1 Cor. 15:48; 2 Cor. 10:11. A rather unusual instance is oi-oj── thlikou/toj seismo.j ou[pw me,gaj (Rev. 16:18). In 2 Cor. 12:20 oi-on is, of course, first person. So oi-oi 1 Th. 1:5.

2. Incorporation. No instance of attraction occurs, but an example of incorporation is found in 2 Tim. 3:11, oi[ouj diwgmou.j u`ph,negka. In Rev. 16:18 the addition of thlikou/toj ou[tw me,gaj after oi-oj is by way of explanatory apposition. But in Mk. 13:19, oi[a ouv ge,gonen toiau,th, the incorporation is redundant after the fashion of oa}n - auvto,n.

3. Indirect Question.235 Like o[j we have oi-oj so used. Cf. 1 Th. 1:5, oi;date oi-oi evgenh,qhmen. In 2 Tim. 3:11 we may have an indirect question also. The Textus Receptus for Lu. 9:55 (D has poi,ou) has another instance of the use of oi-oj in an indirect question, ouvk oi;date oi[ou pneu,mato,j evste u`mei/j.

4. Number. Oi-oj may agree in number with the predicate rather than the antecedent. So 1 Cor. 15:48, oi-oj - toiou/toi. Note the difference in the position of the negative in of ouvc oi[ouj and oi-on ouv, 2 Cor. 12:20. Blass, Gr. of N. T. Gk., p. 179, calls to.n auvto.n- oi-on (Ph. 1:30) peculiar.

732 A GRAMMAR OF THE GREEK NEW TESTAMENT

5. Oi-o,n te, evstin. The only example236 in the N. T. is in Ro. 9: 6, ouvc oi-on de. o[ti, where note the absence of te. It does not occur in exclamations.

(g) `Opoi/oj.

I. Qualitative. It corresponds td the interrogative poi/oj. It is very rare in the N. T. (see Declensions), but occurs in modern Greek vernacular for 'whoever' (Thumb, p. 93). In the literary modern Greek o` o`poi/oj, Jannaris237 thinks that the use of the article was due to the Italian il quale and the French lequel (cf. Old English the which), since educated scribes objected to the vernacular o[pou and pou/.238

2. Double Office. Like oi-oj├ o[soj and h`li,koj it has the double office of relative and indirect interrogative.239 Four of the N. T. instances are indirect questions (1 Cor. 3:13; Gal. 2:6; 1 Th. 1:9; Jas. 1:24). In Gal. 2:6, o`poi/oi, pote, we have the indefinite form ('whatever kind').240 Note here the use of ti and o`poi/oi. In 1 Cor. 3:13 the antecedent is expressed and repeated by redundant auvto,.

3. Correlative. Only one instance is correlative, Ac. 26:29, toiou,touj o`poi/oj. Cf. qualiscumque. Note here the difference in number.

(h) [Osoj.

1. Quantitative. It is found in the LXX like oi-oj and o`poi/oj6 and survives in the modern Greek.241 There are a hundred and eight instances in the N. T. (W. H. text) which display great variety of usage. Radermacher (N. T. Gr., p. 63) notes that in Philo o[soj is often equal to oi[.

2. Antecedent. The presence of the antecedent is not common outside of pa,ntej o[soi (Ac. 5:36, 37), pa,nta o[sa (very common, as Mt. 7:12; 13:46; 18:25; Mk. 11:24, etc.), o[soi- ou-toi (also frequent, as Ro. 8:14; Gal. 6:12, etc.). Cf. o[soi- auvtoi/j in Jo. 1:12. But in Mk. 3:28 o[sa has a`marth,mata and blasfhmi,ai as antecedents and naturally is neuter. Cf. Ac. 3:24; 9:39; Rev. 21:16. It is common without antecedent both in the masculine ( o[soi Mt. 14:36) and the neuter ( o[sa Mk. 9:13).

3. Attraction. This was possible in Jo. 6:11, evk tw/n ovyari,wn

PRONOUNS ( vANTWNUMIAI) 733

o[son h;qelon, but it does not occur. In Lu. 11:8, dw,sei auvtw|/ o[swn crh|,zei, the regular construction occurs. In Winer-Schmiedel242 it is stated that attraction is found in the N. T. with o[soj. I find no real examples outside of the few cases of incorporation now to be mentioned.243

4. Incorporation. In Ac. 9:13 o[sa kaka, is an instance. Mk. 2:19 has o[son cro,non. The other examples (Ro. 7:1; 1 Cor. 7:39; Gal. 4:1) are all instances of evf v o[son cro,non.

5. Repetition. In Mk. 6:30 we have in W. H. o[sa kai. o[sa (not Tisch.). But in Ph. 4:8 o[sa is repeated six times without kai,. In Heb. 10:37 o[son o[son (LXX) is in imitation of the Hebrew in Hab. 2:3. Cf. also Is. 26:20 and D on Lu. 5:3 where o[son o[son= ovli,gon of the other MSS.244 But that this is not an essential Hebraism, but a vernacular idiom in harmony with the Hebrew, is now clear.245

6. With a;n. Note the use as an indefinite relative (Mk. 6:56; Lu. 9:5; Jo. 11:22; Ac. 2:39; 3:22, etc.) and with eva,n (Mt. 7:12; 18:18; 23:3; Mk. 3:28, etc.).

7. Indirect Questions. The instances are fairly numerous. So avkou,ontej o[sa poiei/ (Mk. 3:8); avpa,ggeilon o[sa- pepoi,hkengrk grk(5:19). Cf. 5:20; Lu. 8:39; 9:10; Ac. 4:23; 2 Tim. 1:18, etc.

8. In Comparison. [Oson $o[sw|) is used in comparative sentences usually with tosou/to ( tosou,tw|). Cf. Mk. 7:36; Heb. 1:4; 8:6; 10:25.

9. Adverbial. vEf v o[son (Mt. 9:15; 25:40;. Ro. 7:1, etc.) and kaq v o[son (Heb. 3:3; 7:20; 9:27) partake of the nature of conjunctions.

(i) `Hli,koj. This form was used to express both age and size. Hence the corresponding ambiguity of h`liki,a. Cf. for age Jo. 9:21, for stature Mt. 6:27. The pronoun is absent from the LXX, never very common, but survives in the literary modern Greek.246 It appears also in the papyri.247 Like the other relatives it might have had a double use in the N. T. (relative and indirect interrogative). But the few examples are all indirect interrogatives: Col. 2:1 eivde,nai h`li,kon avgw/na e;cw, Jas. 3:5 ivdou. h`li,kon pu/r h`li,khn

734 A GRAMMAR OF THE GREEK NEW TESTAMENT

u[lhn avna,ptei. The examples in James may be regarded as exclamatory. Note also that h`li,kon refers to smallness and h`li,khn to greatness of the size. In Gal. 6:11 W. and Nestle read phli,koij in the text and h`li,koij in the margin. This again is indirect question after i;dete.

(j) `O AS RELATIVE. The use of the t forms of o`├ h`├ to, as relative is very old in Greek. It appears in Homer248 and is common in Herodotus. In Arkadian o` appears as demonstrative, as article and as relative (Meister, Die griech. Dialekten, Bd. II, p. 116). Cf. also South Ach. (Hoffmann, Griech. Dial., pp. 257, 292-300). Jannaris249 gives examples of it from Ionic (where very common), Doric and Attic (inscriptions), and sporadically in the later Greek. In modern Greek it survives only in sententious sayings with and in Crete and Southeast Greek (Thumb, p. 94). Mayser250 finds a few doubtful instances in the papyri. Wilcken (Archiv, I) gives some examples from B. M. as to, moi de,dwkej (p. 292), th.n avga,phn th.n poiei/j (p. 301), and Moulton (Cl. Rev., 1904, p. 155) quotes pro.j to. du,nome from B.U. 948 (iv/v A.D.) "very illiterate." Mayser (op. cit.) gives numerous examples of o` kai, which "first in Roman time" appears in the nominative. He compares this with the relative use o[j kai, and is inclined to regard o` kai, as relative. The analogy of the Latin qui et favours the relative idea, but the article alone is sufficient in Greek. I would not insist on the relative for Sau/loj o` kai. Pau/loj (Ac. 13:9), though admitting the possibility of it. It means (Deissmann), not 'Saul who is henceforth Paul,' but 'also Paul.' Cf. also Hatch, Jour. of Bibl. Lit., Pt. II, p. 141 f., 1908. In truth this use of o` kai, with double names was very common in N. T. times.251 Dieterich252 sees no instance of o` as relative in the N. T. But in Rev. 1:4, 8; 11:17, we have o` h=n. One either has to say that here o` is used as a relative or that it is a relative. It all comes to the same in the end. It may be a bit artificial, o` w'n kai. o` h=n kai. o` evrco,menoj, but the antique and vernacular relative o` came in as a resource when John did not wish to use geno,menoj of God, and since there is no aorist particinle for eivmi,. Psychologically

PRONOUNS ( vANTWNUMIAI) 735

the article is called for here between two articles, but grammar can do nothing with it. If h=n is treated as a substantive, that would call for to, as in to. de, avne,bh (Eph. 4:9). Moulton253 finds several examples in late papyri of o` as relative (for o` as demonstrative see pp. 693 ff.), like th.n ci/ra th.n de,dwken (p. 304). The only real difficulty in Rev. 1:4, 8, etc., is the nominative use, and that was not insuperable when the exigencies of the sentence demanded it. It is possible that this phrase had come to be a set phrase among the Christians for the eternity and unchangeableness of God. For the possible use of ti,j as relative see under VIII.

VIII. Interrogative Pronouns ( avntwnumi,ai evrwthtikai,)

(a) Ti,j. The root of the interrogative ti,j (Thess. ki,j. Cf. Ionic kw/j├ ko,teroj), indefinite tij (cf. te%, is at bottom the same as the Indo-Germanic root quis and Latin quis (aliquis, que).254 Curiously enough some of the grammars, Monro's Homeric Grammar, for example, give no separate or adequate discussion of the interrogative pronouns.

1. Substantival or Adjectival. Ti,j is either adjectival as ti,na misqo.n e;cete; (Mt. 5:46), or, as more commonly, substantival like ti,j u`pe,deixen; (Mt. 3:7).

2. The Absence of Gender. That it appears only in the nominative and accusative is noteworthy. This fact probably had something to do with the gradual retreat of ti,j before poi/oj.255 The neuter in the N. T. occurs with adjectives only, as ti, avgaqo,n in Mt. 19:16.

3. Ti,j╩poi/oj. An opposite tendency is seen in the use of ti,j= poi/oj.256 Hatzidakis257 has shown examples of this idiom as early as Euripides. As New Testament illustrations one may note ti,j ou-to,j evstin o[j (Lu. 7:49), ti,nej oi` lo,goi ou-toi oua}j avntiba,llete (Lu. 24:17; cf. poi/a 24:19), ti,j evstin ou-toj o` ui`o.j tou/ avnqrw,pou (Jo. 12:34). Cf. Lu. 4:36. Only once258 is poi/oj used with the article (Jas. 4:14, and here B omits h`), while we find ti,j h` sofi,a (Mk. 6:2), ti,j h` aivti,a (Ac. 10:21); etc. Sometimes ti,j and poi/on are used together. It might seem at first as if the distinction were here insisted on, as in eivj ti,na h' poi/on kairo,n (1 Pet. 1:11) and poi/on oi=kon - h' ti,j to,poj (Ac. 7:49). But, tautology seems plain in the last example and may be true of 1 Pet. 1:11, but not certainly

736 A GRAMMAR OF THE GREEK NEW TESTAMENT

so.259 In Mk. 4:30 W. H. read evn ti,ni, but some MSS. have evn poi,a|. Cf. also ti,j kai. potapo,j in Lu. 7:39, which is not tautological.

4. Indeclinable ti,. In Jo. 18:38, ti, evstin avlh,qeia, the neuter in the predicate calls for no special remark. So Gal. 3:19. Cf. Latin quid and English what in such a sentence. This idiom belongs to the ancient Greek and distinguishes between the essence of a thing ( ti,) and the classification of a thing ( ti,j), as Gildersleeve puts it (Syntax of Cl. Gk., p. 59). Cf. u`mei/j ti,nej evste,* (Ac. 19:15) and ti, evstin a;nqrwpoj (Heb. 2:6). But this explanation will not hold for 1 Jo. 3:2, ti, evs,meqa, nor Ac. 13:25, ti, evme. u`ponoei/te. The text in Acts is not certain. The koinh, shows this development outside of the N. T.260 In the modern Greek "the neuter ti,, is used with all genders and cases both in the singular and plural" (Vincent and Dickson, Handb., p. 55). Cf. ti, w[ra ei=nai 'what o'clock is it?' Ti, gunai/ka; 'which woman?' Thumb, Handb., p. 94. It is not unusual in classical Greek261 to have ti, as predicate to tau/ta, as in Lu. 15:26 ti, a'n ei;h tau/ta, Jo. 6:9 tau/ta ti, evstin. So probably ti, tau/ta poiei/te; (Ac. 14:15), though ti, here may be 'why' and not predicative. The usual construction appears in Ac. 17:20 ti,na qe,lei tau/ta ei=nai (cf. Jo. 10:6), 11:17 evgw. ti,j h;mhn; cf. Lu. 8:9. In Ac. 21:33 ti,j and ti, are sharply distinguished. The use of ti, with gi,nomai is hardly in point here (Ac. 5:24; 12:18) as it is found in the Attic262 ti, ge,nwmai. In Jo. 21:21 ou-toj de. ti,; we must supply genh,setai.

5. Predicate Use of ti, with tou/to. In Ac. 23:19, ti, evstin oa} e;ceij, we find the full expression. In Lu. 16:2, ti, tou/to avkou,w peri. sou/, we meet the abbreviated idiom. Cf. Ac. 14:15 ti, tau/ta (see also 9). Cf. Lu. 1:66; Ac. 5:24. The phrase ti, pro.j h`ma/j (Mt. 27: 4), ti, pro.j se, (Jo. 21:22) is matched by the Attic ti, tau/ta evmoi, (Kuhner-Gerth, II, 417; Blass, Gr. of. N. T. Gk., p. 177). Cf. ou-toj ti,, (Jo. 21:21). Blass (ib.) also cllnpares ti, ga,r moi tou.j e;xw kri,nein, (1 Cor. 5:12) with the infinitive in Arrian, Diss. Epict., ii, 17. 14. Ti, evmoi. kai. soi, (Jo.12:4, etc.) is in the LXX (2 Ki. 3:13), but it is also a Greek idiom (ellipsis, KuhnerGerth, ib.).

6. In Alternative Questions. Quality in general is nearly gone from the koinh,. Ti,j when po,teroj might have been used is not unknown in ancient Greek.263 Indeed even in Latin quis occurs sometimes instead of the more usual uter.264 In the LXX po,teroj

PRONOUNS ( vANTWNUMIAI) 737

is supplanted by ti,j and the particle po,teron occurs only once, and that in Job (literary).265 Moulton266 finds only one example of po,te roj in the papyri, and that unintelligible. So in the N. T. po,teroj does not occur as an adjective. So in Mt. 9:5 ti, ga,r evstin euvko─ pw,teron eivpei/n- h' eivpei/n, 21:31 ti,j evk tw/n du,o evpoi,hsen, 27:21 ti,na qe,lete avpo. tw/n du,o. Cf. also 23:17, 19; 27:17; Mk. 2:9; Lu. 7:42; 22:27; 1 Cor. 4:21; Ph. 1:22. Moulton267 notes that "whether, adjectivally, is as archaic as po,teroj," and predicts that "the best of the two" will be the English of the future.

7. The Double Interrogative. Cf. ti,j po,qen in Soph., Tr. 421. It is common in other Indo-Germanic languages.268 Cf. ti,j ti,noj evsti.n evrga,thj, Hom. Clem. 2, 33. So ti,j ti, a;rh| in Mk. 15:24. Some MSS. have ti,j ti, also in Lu. 19:15, but not aBDL (W. H. and Nestle read ti,). Cf. h`li,kon───h`li,khn in Jas. 3:5.

8. As Relative. Just as o[j and o[stij came to be used as interrogatives, so ti,j drifted occasionally to a mere relative. We have seen (1 Tim. 1: 7) how the relative and the interrogative come to be used side by side. "In English, the originally interrogative pronouns 'who' and 'which' have encroached largely on the use of the primitive relative 'that.'"269 Moulton's sketch of the facts270 makes it clear that in the N. T. ti,j may be relative if the exigencies call for it. Moulton finds it only in the illiterate papyri, but the usage is supported by inscriptions271 and by the Pontic dialect today.272 Moulton273 gives from the papyri, eu-ron georgo.n ti,j auvta. e`lku,sh|├ B.U. 822 (iii/A.D.); ti,noj eva.n cri,an e;ch|j, B. M. 239 (iv/A.D.). From the inscriptions see ti,j a'n kakw/j poih,sei, J.H.S., XIX, 299. Moulton274 also quotes Jebb on Soph., O. T. 1141: " Ti,j in classical Greek can replace o[stij only where there is an indirect question." The plainest New Testament example of ti,j as o[j appears to be Mk. 14:36 ouv ti, evgw. qe,lw avlla. ti, su,. Cf. Mt. 26:39 ouvc w`j evgw. qe,lw├ avll v w`j su,) But it is not much more so than Mt. 15:32 ouvk e;cousin ti, fa,gwsin (cf. Mk. 8:1 f.) and Mk. 6:36 i[na- avgo─ ra,swsin e`autoi/j ti, fa,gwsin. Cf. ouvk e;cei pou/ - kli,nh| (Mt. 8:20), but o[pou- fa,gw (Mk. 14:14). See in the papyri, ouvde.n e;cw ti, poi─ h,sw soi, B.U. 948 (iv/v A.D.), as quoted by Moulton (Cl. Rev., 1904, p.155). But even so Xenophon has this idiom, and Sophocles, Oed.

p. 94).

738 A GRAMMAR OF THE GREEK NEW TESTAMENT

Addenda 3rd ed.

Col. 317, has ouvk e;cw ti, fw/, which looks like an indirect question. Cf. Winer-Moulton, p. 211; Winer-Schmiedel, p. 240. It is not necessary to bring275 under this construction ouv ga.r h|;dei ti, avpokriqh|/ (Mk. 9:6) nor Mk. 13:11. Here the idiom is really that of indirect question (deliberative question). Cf. the direct question in Mt. 6:31 with the indirect in 6:25. So in Mt. 10:19 (first example) and see 9. But the second example in Mt. 10:19 ( doqh,se tai- ti, lalh,shte) may be the relative use. Cf. also Lu. 17:8. In Ac. 13:25 the punctuation can (so Nestle, but not W. H.) be made so that ti, is relative, ti, evme. u`ponoei/te ei=nai├ ouvk eivmi. evgw,) It is possible also thus to construe Lu. 19:3, ivdei/n vIhsou/n ti,j evstin, instead of taking ti,s evstin as an accusative of general reference. Cf. Mk. 1:24, oi=da, se ti,j ei= (Lu. 4:34 also). Cf. the prolepsis su. ti,j ei= in Jo. 8:25. So Ro. 14:4, 10. The rhetorical questions in Lu. 11:5; 15:4, 8; Jas. 3:13 are not, of course, instances of this usage.276 Perhaps the anacoluthon in Lu. 11:11 ( ti,na de. evx u`mw/n to.n pate,ra aivth,sei- evpidw,sei) may have arisen because of this idiom. The distinction between ti,j and o[j is, of course, usually maintained (Jo. 16:18; Ac. 23:19; Heb. 12:7). It is at least noteworthy that in 1 Cor. 15:2 Paul changes from o[j (used four times) to ti,ni lo,gw|. An indirect question comes with a jolt and makes one wonder if here also the relative use of ti,j does not occur. In Mt. 26:62 ( ouvde.n avpokri,nh| ti, ou-toi, sou katamarturou/sin) we may have an indirect question (cf. Mk. 14:60), though pro,j would be usual (cf. Mt. 27:14). It is better to follow W. H. with two separate questions277 and even so ti,= ti, evstin o[. The use of ti,j as relative Blass278 calls "Alexandrian and dialectical." The LXX (Lev. 21: 17 a;nqrwpoj ti,ni eva.n h|-├ Deut. 29:18 avnh.r- ti,noj, Ps. 40:6 ouvk e;stin ti,j) does show examples of it, but it is not confined to Egypt, as has been already shown.279 Brugmann (Griech. Gr., p. 561) finds ti,j as relative in Boeotian and even rarely in the older Attic.

9. Adverbial Use. The neuter accusative ti, is frequently used in the sense of 'why' in the N. T. This is classical and common and calls for little comment. It still appears in modern Greek (Thumb, p. 94). See Mt. 7:3 ( ti, ble,peij to, ka,rfojgrk) 8:26 ( ti, deiloi, evste*grk) 19:17; 20:6, etc. In Ac. 14:15 ti, tau/ta poiei/te we probably have ti,= 'why.' Cf. Mk. 11:3. In Mk. 2:24 ti, poiou/sin toi/j sa,bbasin oa} ouvk e;xestin; note 'why,' though ti, is followed by o[. It

PRONOUNS ( vANTWNUMIAI) 739

Addenda 3rd ed.

is interesting to note pw/j h' ti,, Mt. 10:19 Lu. 12:11. In Jo. 14:22 ti, ge,gonen o[ti we see the full form of the idiom ti, o[ti (Lu. 2:49; Ac. 5:4, 9). Here ti, still = 'why.' But in i[na ti, (1 Cor. 10:29 and Mt. 9:4; 27:46; Lu. 13:7; Ac. 4:25; 7:26) ti, is really the subject of ge,nhtai (ellipsis). It is not unknown in Attic Greek.280 W. H. never print i`nati, (cf. Mt. 9:4; Lu. 13:7). It is common in LXX.

10. With Prepositions. There is very little difference between ti,= 'why' and dia. ti,= 'because of what' (Mt. 15:2, 3; 17:19; Lu. 24:38, etc.). Kata . ti, ('according to what') is practically 'how.' Cf. Lu. 1:18. For evn ti,ni see Mt. 5:13. But pro.j ti, (Jo. 13:28)= 'for what purpose.' In Jo. 13:22 peri. ti,noj le,gei there is no such idea. But purpose again is expressed by eivj ti, (Mt. 14:31; 26:8; Mk. 14:4; Ac. 19:3).

11. With Particles. Paul in particular is fond of the rhetorical use of ti, ga,r (Ro. 3:3; 4:2, etc.), ti, ou=ngrk grk(3:1, 9, etc.), ti, e;tigrk grk(3:7; 9:19), avlla. ti,grk grk(11:4), h; ti, grk(11:2). Cf. ti,j a;ra in Lu. 22:23 and ti, a;ra 1:66; Ac. 12:18.

12. As Exclamation. In Mt. 7:14 W. H. read o[ti (causal), not ti, stenh. h` pu,lh. But in Lu. 12:49 kai. ti, qe,lw eiv h;dh avnh,fqh there is no doubt of the text. W. H. punctuate as a question, but Nestle as an exclamation. Examples of exclamatory ti,-'how' are found in 2 Sam. 6:20; Song of Sol. 7:6 and in the modern Greek, ti, kalo.j a;nqrwpoj! Cf. Mullach, Vulg., pp. 210, 321; Winer-Moulton, p. 562. Blass281 compares the Hebrew hm'. On the whole it is best to take ti, in Lu. 12:49= 'how.'

13. Indirect Questions. It is, of course, the ancient idiom282 to have ti,j in an indirect question. But in the N. T. the indirect interrogative o[stij has disappeared in this idiom save in Ac. 9:6 (MSS. divided here). A good example of ti,j occurs in Ac. 10: 29 punqa,nomai ti,ni lo,gw| metepe,myasqe, me. In Luke we meet the neuter article rather frequently before the indirect question. So to. ti, a'n qe,loi grk(1:62), to. ti,j a'n ei;hgrk grk(9:46). Cf. 22:23, 24, etc. Blass (Gr. of N. T. Gk., p. 158) sees no special point in the article (cf. English " the which"). Paul sometimes uses it also (Ro. 8:26; 1 Th. 4:1 to. pw/j). The question is brought out rather more sharply by the article. The Attic use of to. ti,├ to. poi/on) (Thompson, Synt., p. 74) in reference to something previously mentioned is like our "The what?" Cf. Herm., Sim., VIII, i, 4, Clem., Hom., i, 6. 14. Ti,j or ti,j. Sometimes it is difficult to decide whether ti,j

740 A GRAMMAR OF THE GREEK NEW TESTAMENT

or ti.j is right. So 1 Pet. 5:8 W. H. have zhtw/n katapiei/n with tina. in the margin. But Nestle actually prints zhtw/n ti,na katapiei/n. In Heb. 5:12 W. H. read tina. and Nestle tina (both indefinite). In Jas. 5:13 the reading is, of course, tij, not ti,j. So 1 Cor. 7:18.

(b) Poi/oj.

1. Qualitative. It occurs sixteen times in direct questions. It is still used in its original qualitative sense. Clearly this is true in Jo. 12:33, shmai,nwn poiw| qana,tw| h;mellen avpoqnh,skein (cf. 18:32), Ro. 3:27 ( dia. poi,ou no,mou* tw/n e;rgwn*). The same thing is true of 1 Cor. 15:35 ( poi,w| sw,mati e;rcontai), cf. also 1 Pet. 2:20. In 1 Pet. 1:11 we find both ti,na and poi/on in apparent contrast. Other possible instances are Jo. 10:32; Ac. 7:49 (LXX); Jas. 4: 14. The common evn poi,a| evxousi,a|. (Mt. 21:23; Mk. 11:28; Ac. 4:7, LXX, etc.) seems also to retain the qualitative force. Cf. also Lu. 24:19. The qualitative sense is clear in poi,ou pneu,mato,j evste (Lu. 9:55), a spurious passage, however.

2. Non-qualitative. But some examples clearly have lost the qualitative sense. In the modern Greek poio,j is used regularly283= ti,j, and is the usual interrogative. Note the accent poio,j. Indeed examples of this weakened sense of poi/oj Jannaris284 finds as early as AEschylus and Euripides. See (a), 3. In Mt. 24:42 ouvk oi;date poi,a| h`me,ra| o` ku,rioj u`mw/n e;rcetai there seems to be merely the force of ti,j, not quality. Cf. also 24:43 poi,a| fulakh|/, Lu. 12:39 poi,a| w[ra|, Ac. 23:34 poi,aj evparcei,aj, Rev. 3:3 poi,an w[ran. This is probably true also of Mt. 22:36 poi,a evntolh, (Mk. 12:28). In Lu. 5:19 poi,aj and 6:32 f. poi,a ca,rij either point of view will answer.

3. In Indirect Questions. It occurs sixteen times (not counting Lu. 9:55) in this construction against four for o`poi/oj. Cf. indicative in Mt. 21:24; 24:42; Jo 12:33; 21:19, and the subjunctive in Lu. 5:19 mh. poi,aj eivsene,gkwsin. Poi/oj is found in the LXX and in the papyri.

(c) Po,soj.

1. Less Frequent than poi/oj. It occurs chiefly in the Synoptic Gospels (twenty-seven times in W. H. text).

2. Meaning. It is used in the sense of 'how much' $po,sw| Mt. 12:12), 'how great' ( po,son Mt. 6:23), and of 'how many' ( po,souj a;rtouj e;cete; Mt. 15:34). Eleven examples of po,sw| occur almost like an adverb (Mt. 7:11; 10:25, etc.). The use of po,soj cro,noj - w`j (Mk. 9:21) is noteworthy.

PRONOUNS ( vANTWNUMIAI) 741

3. In Indirect Questions. See ouvk avkou,eij po,sa sou katamarturou/sin; (Mt. 27:13). Cf. Ac. 21:20, etc.

4. The Exclamatory Use. This is found in Lu. 15:17 po,soi mi,sqioi tou/ patro,j mou, and in 2 Cor. 7:11 po,shn kateirga,sato u`mi/n spoudh,n. The exclamatory use of tw/j may be mentioned (Mk. 10:23 f.; Jo. 11:36). Cf. w`j in Ro. 10:15 and 11:33. Cf. po,soj- w`j in Mk. 9:21.

(d) Phli,koj)

1. Rare. It is found only twice in the N. T. (Gal. 6:11; Heb. 7:4) and W. H. put h`li,koij in the margin of Gal. 6:11. It is rare also285 in the LXX (cf. Zech. 2:2), and has disappeared from the modern Greek vernacular.

2. Indirect Questions. Both of the N. T. examples are indirect questions. The example in Heb. 7:4 describes greatness of Melchisedek (how great), the one in Gal. 6:11 presents the size of the letters (how large).

(e) Potapo,j.

It is the late form for podapo,j. It no longer in the N. T. means 'from what country,' but merely 'of what sort'= poi/oj. It is found only once in LXX (Susanna 0 54, "where it keeps something of its original local meaning").286 It exists in the late Greek vernacular.287 It occurs once in a direct question (Mt. 8:27) and once probably in an exclamation (2 Pet. 3:11). Four times we find it in indirect questions (Mk. 13:1; Lu. 1:29; 7:39; 1 Jo. 3:1). In Lu. 7:39 it is contrasted with ti,j.

(f) Po,teroj.

As a pronoun it has vanished from the LXX (Thackeray, Gr., p. 192) and from the papyri (Moulton, Prol., p. 77). The only example in the N. T. (cf. LXX, Thackeray, p. 192) is in an alternative indirect question as the conjunction po,teron (Jo. 7:17). Cf. Latin utrum-an. Blass (Gr. of N. T. Gk., p. 176) cites Herm., Sim., ix, 28. 4.

IX. Indefinite Pronouns ( avntwnumi,ai avo,ristoi).

(a) Ti.j.

1. The Accent. Jannaris288 calls it "irrational" to accent the nominative ti.j rather than ti,j. But then the nominative singular never has an accent unless at the beginning of a sentence or in philosophical writings (Thompson, Syntax, p. 76) and cannot otherwise be distinguished in looks from ti,j the interrogative.

2. Relation to ti,j. The same connection is seen in the Latin

742 A GRAMMAR OF THE GREEK NEW TESTAMENT

quis, ali-quis and quis-quis (cf. ti,stij in Argive dialect).289 Brugmann290 considers - ki- in ouv─ki,├ polla,ki─j the same word as ti and cites ki.j in the Thessalian dialect. Just as in modern Greek ti,j disappears before poio,j, so tij vanishes before kanei,j (Thumb, Handb., p. 95). But in the N. T. tij is still very common, especially in Luke and Acts. In general the usage is in harmony with that of ancient Greek. We do not have e;nioi in the N. T. In Ac. 25:26 note ti gra,yai and ti, gra,yw. Cf. Lu. 7:40. See tij ti,├ Ro. 8:24, in margin of W. H.

3. Tij as Substantive. As a substantive tij may be equal to 'any one,' 'anybody' or 'anything,' as in ouvde. to.n pate,ra tij evpi─ ginw,skei, Mt. 11:27; pw/j du,natai, tij, 12:29; ei; tij qe,lei, 16:24; eva,n tij u`mi/n ei;ph| ti (note both examples like tino,j ti Lu. 19:8; cf. Mk. 11:25; Col. 3:13), Mt. 21:3. For several instances of ti╩ 'anything' see Ac. 25:5, 8, 11. But the substantive use of tij may be = 'somebody' or 'something,' as e;rcetai, tij Lu. 8:49, dramw.n de, tij Mk. 15:36, u`po, tinoj Heb. 3:4. Cf. Lu. 8:46. Often the partitive genitive (or ablative) occurs with tij as substantive. So tine.j tw/n grammate,wn Mt. 12:38, tij tw/n maqhtw/n Lu. 11:1, tij evk tou/ o;clou 12:13. The plural is usually = 'some,' as Mk. 9:1; 1 Cor. 9:22. In Homer tij was sometimes "public opinion, the man in the street" (Gladstone, quoted in Thompson's Syntax, p. 75). This idiom is very nearly represented by ei=pen de, tij evk tou/ o;clou, Lu. 12:13 (cf. 11:1; 7:36). In Heb. 2:6, diemartu,rato pou, tij├ the tij is really quite definite in the writer's mind, though he writes thus.

4. With Numerals = 'About.' With numerals tij sometimes in classical Greek gives an approximate idea rather than exact reckoning, like our "about." No certain instances of this idiom appear in the N. T. Certainly not Ac. 19:14, where tinoj, not tinej, is the correct text. In Lu. 7:19, proskalesa,menoj du,o tina.j tw/n maqhtw/n, the meaning may be 'about two,' but it could mean 'certain two' just as well. The same thing is true of Ac. 23:23, proskalesa,meno,j tinaj du,o, where it is even less likely that the idea is 'about two.' Classical also is ei-j tij (Lu. 22:50; Jo. 11:49, and probably Mk. 14:47). The adjectival uses of tij are quite varied.

5. With Substantives. Here tij may = a kind of,' as avparch,n tina, Jas. 1:18. Cf. Ac. 17:20, though this is not true of Col.

PRONOUNS ( vANTWNUMIAI) 743

2:23 because of the negative.291 But the commonest use of tij with substantives is= 'certain' (really rather uncertain!). Thus i`ereu,j tij, Lu. 1:5; a;nqrwpo,j tij, Lu. 14:2, 16; 15:11, etc. Cf. ti u[dwr, Ac. 8:36. Sometimes it is difficult to give more force to tij than the English indefinite article. Cf. nomiko,j tij, Lu. 10:25; krith,j tij h=n e;n tini po,lei, Lu. 18:2. Indeed it is nearly always true that our "certain" is too emphatic.

6. With Adjectives. The effect is rhetorical.292 There is "a double adjectival sense."293 Thus Ac. 8:9, tina me,gan,='a very great man' ('some great man'), in his own estimation. Blass294 needlessly considers this passage an interpolation. Cf. also Heb. 10:27, fobera, tij evkdoch,, where tij rather intensifies fobera,. The tone may tend to soften the matter as in Heb. 2:7, 9, bracu, ti. But in Lu. 24:41 ti brw,simon, Jo. 1:46 ti avgaqo,n, Ac. 25:26 avsfale,j ti, Ro. 14:14 ti koino,n, 2 Cor. 11:16 mikro,n ti, we have rather the substantive use of ti. But in tuflo,j ti, Lu. 18:35, both are adjectives. Cf. a;lloj tij (Lu. 22:59) and e[tero,j tij (Ac. 27:1).

7. As Predicate. Here tij may be emphatic = 'somebody in particular,' as Ac. 5:36, le,gwn ei=nai, tina e`auto,n (cf. 8:9). See also Gal. 2:6, avpo. tw/n dokou,ntwn ei=nai, ti, where note difference between ti and tinej. In Gal. 6:3 note in eiv dokei/ tij ei=nai, ti mhde.n w;n, both senses of tij. But the predicate may have the other meaning of ti ('anyone,' 'anything'). So 1 Cor. 3:7; 10:19; Gal. 6:15. In Gal. 2:6 compare ti, and o`poi/oi.

8. The Position of tij. It is not material. It naturally follows the substantive or adjective as in eivj kw,mhn tina,, Lu. 10:38, but we often have the other order as in tina ch,ran Lu. 21:2. Tine,j may indeed begin a sentence (Ph. 1:15; 1 Cor. 8:7).

9. As Antecedent. In Mt. 16:28 tinej is the antecedent of oi[tinej, but here oi[tinej is more definite than of would have been. Cf. Lu. 9:27. In 2 Cor. 10:2 note tinaj tou.j l)

10. Alternative. It is used to express alternative ideas, as tine.j me,n- tine.j de, in Ph. 1:15. Cf. u`po. tinw/n - u`po. tinw/n- a;llwn de, in Lu. 9:7 f. and tij - e;teroj in 1 Cor. 3:4.

11. The Negative Forms ou; tij├ mh, tij. These are not printed as single words by W. H., except mh,ti as an interrogative particle expecting the answer No, as in Mt. 26:22, mh,ti evgw, eivmi├ ku,rie* cf. Jo. 4:33. It is all a matter with the editor whether in i[na mh, tij

744 A GRAMMAR OF THE GREEK NEW TESTAMENT

ei;ph|, 1 Cor. 1:15 (cf. Eph. 2:9), we may not really have mh,tij. The separation in Heb. 3:13; 4:11 is against it. Cf., for instance, mh, tina (2 Cor. 12:17) and mh,ti in the next verse. The anacoluthou with tina here is noticeable.

12. Indeclinable ti. The use of tij with spla,gcna kai. oivktirmoi├ (Ph. 2:1) may be compared with indeclinable ti. Indeclinable ti itself survives in modern Greek ka;ti (Moulton, Prol., p. 244),

(b) Ei-j = Tij.

This is merely one usage of ei-j, the cardinal numeral. The idiom is common after Plutarch, but traces of it occur earlier.295 Moulton296 sees no difference between ei-j and tij in Aristophanes, Av., 1292. The papyri furnish similar examples. "The fact that ei-j progressively ousted tij in popular speech, and that even in classical Greek there was a use which only needed a little diluting to make it essentially the same, is surely enough to prove that the development lay entirely within the Greek language, and only by accident agrees with Semitic."297 This use of ei=j alone, with genitives, with substantives, was treated at the close of the chapter on Adjectives. For ei-j tij see tij) For ei-j - ei-j as alternative pronoun see later, and for ei-j- ouv and ouvdei,j $mhdei,j) see Negative Pronouns under xi.

(c) Pa/j ='any one' no matter who, 'anything' no matter what. Cf. quidvis.298 We see this construction in Ac. 2:21 (LXX), pa/j oa}j eva.n evpikale,shtai. So Gal. 3:10 (LXX); Lu. 14:33. Pa/j with a participle may have the same force, like panto.j avkou,ontoj to.n lo,gon, Mt. 13:19 (cf. Lu. 11:4), and pa/j o` ovrgizo,menoj, Mt. 5:22, etc. For pa/j- ouv 'no one' see negative pronouns. For the adjectival uses of pa/j, see chapter on Adjectives and chapter on Article.

(d) `O Dei/na. This rare pronoun was current chiefly in colloquial speech (Jannaris, Hist. Gk. Gr., p. 166). It survives in-the modern Greek (Thumb, p. 98). It means "Mr. So-and-So." It occurs only once in the N. T., pro.j to.n dei/na, Mt. 26:18.

X. Alternative or Distributive Pronouns ( avntwnumi,ai dath─ ri,ai).

I apply a term from AEschylus in lieu of a better one. The reciprocal pronoun avllh,lwn has been already treated.

(a) vAmfo,teroi. ;Amfw has vanished299 from the koinh,. vAmfoteroi has taken its place. It continues in the later Greek,300 but Thumb

PRONOUNS ( vANTWNUMIAI) 745

does not give it for modern Greek. It is frequent in the LXX,301 but is found only fourteen times in the N. T. It occurs without the article in all but five instances. So Mt. 9:17. Once the article is used with the substantive, avmfotera ta. ploi/a, Lu. 5:7. The other four examples have the article before the pronoun, like oi` avmfo,teroi, Eph. 2:18. It is possible, even probable, that in two instances duality has disappeared from the word. It seems certain that three items are referred to in Ac. 23:8 and in Ac. 19: 16 the seven sons of Sceva are alluded to. A corruption of the text is possible (cf. the Bezan text for 19:16), but it is hardly necessary to postulate that in view of "the undeniable Byzantine use"302 of avmfo,teroi for more than two (cf. "both" in old English). The papyri show undoubted examples also and "the Sahidic and some later versions took avmfote,rwn, as all.'"303 But Moulton304 hesitates to admit in Luke "a colloquialism of which early examples are so rare," a rather surprising objection from Dr. Moulton. On the whole one is safe in the two passages in Acts here quoted to admit the free use of avmfo,teroi. The papyri examples bearing on this usage include N.P. 67, 69 (iv/A.D.) "where it is used of four men" (Moulton, Cl. Rev., 1904, p. 154), probably also B.M. 336 (ii/A.D.). See Bury, Cl. Rev., XI, p. 393, for the opposite view. Nestle (Berl. Phil. Woch., 1900, N. 47) shows that German also uses "beide" for three and more persons.

(b) [Ekastoj. In the LXX e`ka,teroj is still used to a limited extent (Gen. 40:5) and occasionally= e[kastoj, without dual idea (cf. avmfo,teroi), as often in the papyri.305 In O.P. 256 (i/A.D.) and B.M. 333 (ii/A.D.) e`ka,teroj is used of three and of four in G. H. 23a (ii/B.C.). See Moulton, Cl. Rev., 1901, p. 440, and proper use of e`ka,teroj in P.Oxy. 905 (A.D. 170), pro.j to. e`ka,teron me,roj. But in the N. T. e`ka,teroj does not appear. [Ekastoj is common in the N. T., but comes to be replaced in modern Greek by ka,qe├ kaqei,j and kaqe,naj (cf. kaq v ei-j in the N. T.).306

1. Without Substantive. This is indeed the usual idiom, as in Mt. 16:27; Jo. 6:7.

2. With Substantive. Never with the article. So Eph. 4:16; Heb. 3:13; Rev. 22:2. Thus very rare.

746 A GRAMMAR OF THE GREEK NEW TESTAMENT

3. With ei-j. This is very frequent. So ei-j e[kastoj Mt. 26:22, etc. We even have avna. ei-j e[kastoj, Rev. 21:21. But in Ac. 21:19, evxhgei/to kaq v ea}n e[kaston w-n evpoi,hsen we must not307 connect e[kaston with e[n.

4. With Genitive. It is common also with the genitive, as in Lu. 13:15; Eph. 4:7.

5. Partitive Apposition. This is frequent also. Thus avfh/te e[kastoj Mt. 18:35, evporeu,onto pa,ntej───e[kastoj Lu. 2:3, etc. The same thing is true in Eph. 5:33 u`mei/j kaq v e[na e[kastoj. This is a classical construction.308

6. Rare in Plural. So e[kastoi. Ph. 2:4, but even here W. H. have e[kastoj in the margin.

7. Repetition. Note the repetition of e[kastoj in Heb. 8:11 (from Jer. 31:34). This translation of vyai, by e[kastoj rather than avnh,r is an instance of independence of Hebrew literalism. Cf. Mt. 18:35 with Gen. 13:11; Ro. 15:2 and Eph. 4:25 with Is. 3:5 (Winer-Schmiedel, p. 246). For avnh,r╩e[kastoj in the LXX (literal books) see Thackeray, Gr., p. 192.

(c) ;Alloj. Cf. Latin alius, English else.

1. Used absolutely = 'An-other,' 'One Other.' This is the commonest use of the pronoun. Cf. 1 Cor. 12:8-10 where a;llw| occurs six times. So Mt. 13:5-8 where a;lla appears three times. But it is found alone also, as a;llouj, Mt. 27:42. For a;lloj tij see Lu. 22:59. Cf. ouvde.n a;llo (Gal. 5:10) 'nothing else.' It occurs in modern Greek vernacular.

2. For Two. But a;lloj occurs where the idea of two is present (pair). Here e[teroj might have been used, but even in Euripides, I. T. 962 f., Blass309 finds qa,teron- to. d v a;llo, though he considers it a "most striking encroachment" for a;lloj to supplant e[teroj in this fashion. Moulton (Cl. Rev., 1901, p. 440) cites th/j me.n mia/j- th/j d v a;llhj G. H. 23a (ii/B.c.); du,o├ to.n me.n e[na- kai. to.n a;llon B.U. 456 (iv/A.D.). Moulton310 explains the existence of kai. th.n a;llhn $siago,na% in Lu. 6:29 as a failure on Luke's part to correct his source, a like failure appearing in Mt. 5:39, unless that was his source. But the matter goes much further than that. In Mt. 12:13 h` a;llh refers to the other hand ( cei,r). In Jo. 19:32 note tou/ prw,tou- kai. tou/ a;llou.311 Cf. also Jo. 18:16; 20:3 f. In Jo. 5:32 evgw, and a;lloj are contrasted. So Mt. 25:16, ta. pe,nte ta,lanta- a;lla pe,nte, for which Blass6 finds "complete illustration in classi-

PRONOUNS (vANTWNUMIAI) 747

cal authors." There are other N. T. examples such as a;llhn in Mt. 19:9, ta. du,o- a;lla du,o Mt. 25:17, a;llhn Mk. 10:11, a;llon 10:12, a;llon para,klhton Jo. 14:16.

3. As Adjective. Common. Cf. Mt. 2:12; 4:21; and in particular Rev. 14:6, 8, 15, 17, 18 and 1 Cor. 15:39, 41.

4. With the Article. It is not frequent. The article sharply refers to a preceding example. Cf. Mt. 5:39; Mt. 27:61. John alludes to himself in his Gospel as o` a;lloj maqhth,jgrk grk(18:16; 20:2, 3, 4). The article may be repeated, as in Jo. 18:16; 19:32.

5. The Use of a;lloj a;llo = 'One One Thing, One Another.' This is classical and is illustrated in Ac. 19:32; 21:34. In Ac. 2:12, a;lloj pro. a;llon, the idiom is almost reciprocal like avllh,lwn.

6. In Contrast for 'Some - Others.' We have a;llh me,n- a;llh de,, 1 Cor. 15:39 and 41; aa} me,n - a;lla de,├ Mt. 13:4 f. (cf. kai. a;llo, Mk. 4:5); oi` me,n - a;lloi de,──e[teroi de,, Mt. 16:14; kai, a;lloi──a;lloi de,├ Mk. 8:28; u`po. tinw/n- a;llwn, Lu. 9:8; o` ei-j o` a;lloj, Rev. 17:10.

7. Ellipsis of a;lloj is possible in Ac. 5:29, Pe,troj kai. oi` (sc. a;lloi) avpo,stoloi. Blass312 cites also Ac. 2:14, Pe,troj su.n toi/j (sc. loipoi/j) e[ndeka. But psychologically this explanation is open to doubt.

8. The Use of a;lloj and e[teroj Together. Blass313 finds this "probably only for the sake of variety." Certainly in 1 Cor. 12:9 f. no real distinction can be found between a;lloj and e[teroj, which are here freely intermingled. But I am bound to insist on a real difference in Gal. 1:6 f. The change is made from e[teron to a;llo for the very reason that Paul is not willing to admit that it is a gospel on the same plane ( a;llo) as that preached by him. He admits e[teron, but refuses a;llo. The use of eiv mh, by Paul does not disturb this interpretation. The same thing would seem to be true of 2 Cor. 11:4, a;llon vIhsou/n - pneu/ma e[teron - euvagge,lion e[teron. It may be that variety (as in 1 Cor. 12:9 f.) is all that induces the change here. But it is also possible that Paul stigmatizes the gospel of the Judaizers as gmpov (cf. Gal. 1:6) and the Spirit preached by them, while he is unwilling to admit another ( a;llon) Jesus even of the same type as the one preached by him.

9. ='Different.' Besides, it is not to be forgotten that in ancient Greek a;lloj itself was used for 'different kind.' Thompson (Syntax, p.76) cites a;lla tw/n dikai,wn from Xen., Mem., IV, 4. 25. Cf. also avlla, in the sense of 'but.' Cf. avlla. a;llh in 1 Cor. 15:39.

748 A GRAMMAR OF THE GREEK NEW TESTAMENT

Indeed in 1 Cor. 15:39, 41, a;llh me,n- a;llh de, it is expressly stated that the glory is not h` auvth,. In verse 40 e`te,ra occurs. Here a;lloj seems to be used in the sense of 'dfferent,' like e[teroj. In Latin alius was often used where earlier Latin would have used alter. Cf. Draeger, Hist. Synt., p. 105.

10. vAllo,trioj. This variation of a;lloj has the same relation to it that alienus has to alius. It means 'belonging to another,' and occurs fourteen times in the N. T. Cf. Ro. 15:20. The contrast with auvtw/n is seen in Mt. 17:25. In Heb. 11:34 it has the notion of alienus.

(d) [Eteroj.

1. Absolutely. So often as in Lu. 14:19 f., but it is also used more frequently with substantives than is a;lloj. Cf. Lu. 4:43; Ac. 7:18 (LXX), etc. For e[tero,j tij see Ac. 8:34; Ro. 13:9. For the genitive with e[teroj cf. Mt. 8:21; Gal. 1:19.

2. With Article. The article is also more common with e[teroj than with a;lloj. Cf. Mt. 10:23; 11:16, etc.

3. Second of Pair. A common, probably the original, use of gmpos is for the second of a pair. Cf. Latin alter. It is the only surviving dual pronominal word in the N. T. (except avmfo,─ teroi), and is common in the LX.X314 and the papyri.315 For su.n e`te,ra| mia|/, see P.Tb. 421 (iii/A.D.). The examples are rather abundant in the N. T. of this dual (comparative) sense ( e[─teroj). So to.n e[na- to.n e[teron, Mt. 6:24; su,──h' e[teron, 11:3; evn tw|/ e`te,rw| ploi,w|, Lu. 5:7. Cf. also Lu. 7:19 f.; 14:31; 16:13; 17:34 f.; 18:10; 20:11.316 Not radically different from this conception is the use of it for 'next,' as in Lu. 6:6, evn e`te,rw| sabba,tw|, 9:56 eivj e`te,ran kw,mhn Ac. 20:15 th|/ e`te,ra|. Cf. also Mt. 10:23. See also, to.n e[teron in Ro. 2:1; 13:8 = 'neighbour.'

4. ='Different.' The sense of 'different' grows naturally out of the notion of duality. The two things happen just to be different. Cf. Latin alius and alienus. The word itself does not mean 'different,' but merely 'one other,' a second of two. It does not necessarily involve "the secondary idea of difference of kind" (Thayer). That is only true where the context demands it. But note how Latin alter lends itself to the notion of change. Thompson317 suggests that this sense may be "an euphemism for kako,j." The N. T. examples are rather numerous. So evge,neto- to. ei=doj tou/ prosw,pou auvtou/ e[teron, Lu. 9:29. Cf. also Ac. 2:4; Ro. 7:23; 1 Cor. 14 : 21; 2 Cor. 11 : 4; Gal. 1 : 6; Heb. 7 : 11, 13, 15; Ju. 7.

PRONOUNS ( vANTWNUMIAI) 749

Cf. also e`te,rwj in Ph. 3:15 and evn e`te,ra| morfh| Mk. 16:12 (disputed part of Mark.)318 Cf. Ac. 17:21. We have already seen that a;lloj may be equal to 'different' (1 Cor. 15:39). [Eteroj occurs in verse 40 in the sense of 'different.' Ramsay (on Gal. 1:6) argues that, when e[teroj occurs in contrast with a;lloj, it means not 'different' (as Lightfoot in loco), but 'another of the same kind.' Moulton (Prol., p. 246) stands by Lightfoot in spite of Ramsay's examples.

5. ='Another' of Three or More. But e[teroj comes also to be employed merely for 'another' with more than two and with no idea of difference. This usage probably grew out of the use with two groups. So Lu. 10:1, avne,deixen e`te,rouj e`bdomh,konta du,o. In Mt. 12:45, e`pta. e[tera pneu,mata ponhro,tera e`autou/, the notion of difference is present. This difference may also be implied by Luke in 23:32, kai. e[teroi kakou/rgoi du,o. Cf. Lu. 8:3. But this is hardly true of Ac. 2:13. In Ac. 4:12 the point of e[teron is rather that no other name at all than that of Jesus, not that of difference in kind. In Lu. 19:16-20 we have this order, o` prw/toj├ o` deu,te─ roj├ o` e[teroj. So in 1 Cor. 4:6, ei-j u`pe.r tou/ e`no.j fusiou/sqe kata. tou/ e`te,rou, the third is again presented by e[teroj. Then, again, e[teroi occupies third place in Mt. 16:14 and Heb. 11:36. In Mt. 15: 30 it comes in the fifth place. Blass319 admits that this use of grepoc "at the close of enumerations may be paralleled from Attic writers." See further Lu. 3:18; Ro. 8:39; 1 Tim. 1:10. But in 1 Cor. 12:8-10 e[te,rw| occurs in the third and the eighth places. We are not surprised then to learn that the papyri furnish plenty of examples where e[teroj refers to more than two.320 Blass indeed considers this extension not correct, and Moulton seems surprised that Luke should change the correct a;lloj (Mk. 4:5-8 Mt. 13:5-8) to e[teron in Lu. 8:6-8. But Luke is reinforced by Paul in this laxity as to e[teroj. Cf. plla. kai. e[tera in Lu. 3:18. Moulton (Cl. Rev., 1904, p. 154) calls this "incorrect e[teroj" and finds it in the papyri, as in O.P. 494 (ii/A.D.). But we do not need to hold grepos in leading strings. The "subtlety" (Cl. Rev., 1901, p. 440) is only called for in that case.

6. In Contrast. [Eteroj may also be used in contrast for 'the one,' 'the other.' So 1 Cor. 15:40, e`te,ra me,n- e`te,ra de,. It is common in contrasts with other pronouns. Thus with ei-j in Mt. 6:24; o` ei-j in Lu. 7:41; Lu. 17:34 ff.; with tij, Lu. 11:15 f.; with oa} me,n├ Lu. 8:5 f.; with oi` me,n and a;lloi, Mt. 16:14. But

750 A GRAMMAR OF THE GREEK NEW TESTAMENT

neither ouvde,teroj $mhd──% nor ouvqe,teroj ( mhq──) occurs in the N. T., though mhqe,teroj is read in Prov. 24:21. In Clem. Hom. XIX, 12 we have ouvqe,teroj.

(e) OTHER ANTITHETIC PRONOUNS. For ei-j- ei-j (Mk. 10:37), qei-j - o` de, (Gal. 4:24 f.), o` ei-j- o` a;lloj (Rev. 17:10) see ei-j under Numeral Adjectives. So likewise tij may be contrasted with tij (Ph. 1:15), with a;lloj (Lu. 9:7 f.), with e[teroj (1 Cor. 3:4). For the very common o` me,n- o` de.├ oa}j me,n- oa}j de, see Demonstrative Pronouns. The repetition of the substantive is to be noted also. So oi=koj evpi. oi-kon pi,ptei, Lu. 11:17; o` satana/j to.n satana/n evkba,llei, Mt. 12:26 ( = Lu. 11:18). This notion of repetition is seen in h`me,ra| kai. h`me,ra| (2 Cor. 4:16; cf. Heb. mAyd' mAy). Cf. also ei-j kai. ei-j (Mt. 20:21; 24:40 f.; 27:38, etc.); o` ei-j - o` e[teroj, Lu. 7:41. For ei-j- kai. ei-j- kai. ei-j see Mk. 9:5 = Mt. 17:4 = Lu. 9:33. This threefold repetition of ei-j is rhetorical.321 The distributive use of ei-j with kata, and avna, $ea}n kaq v ea}n├ ei-j kaq v ei-j├ avna. ei-j) was treated under Numeral Adjectives.

XI. Negative Pronouns ( avntwnumi,ai avrnhtikai,,).

(a) Ouvdei,j.

1. History. Note this accent rather than ouvdei/j. Ouvdei,j is supplanted in modern Greek vernacular by kanei,j, but ouvde,n survives as negative particle in form de,n. Cf. Jannaris, Hist. Gk. Gr., p. 171.

2. Ouvqei,j. This is made from are ou;te ei-j. (sometimes also from ouvde. ei-j, 'not even,' Brugmann, Griech. Gr., p. 146) and occurs sometimes in the best N. T. MSS. Cf. W. H.'s text for Lu. 22: 35; 23:14; Ac. 15:9; 19:27; 26:26; 1 Cor. 13:2; 2 Cor. 11:9. Jannaris322 finds it a peculiarity of the Alexandrian school. Meisterhans323 has shown from the inscriptions how ouvqei,j and mhqei,j came to be practically universal during the third century and the first half of the second century B.C. Thackeray324 has reinforced this position from the uncials for the LXX. The papyri are in full accord.325 In the fourth and fifth centuries A.D., the date of the great uncials, ouvqei,j and mhqei,j had disappeared from current speech, and yet a number of instances survive in the MSS. of the O. T. and the N. T., though others were probably replaced by ouvdei,j and mhdei,j.326 In-

PRONOUNS ( vANTWNUMIAI) 751

deed ouvqei,j was a sort of fashion (Moulton, Cl. Rev., Mar., 1910, p. 53) that came in iv/B.C. and vanished ii/A.D. It was nearly extinct in N. T. times. See further chapters VI, III (g), and VII, III 2.

3. Gender. The feminine form is less frequent in the N. T. than the masculine and neuter. The word occurs with substantives (Mk. 6:5), with other pronouns ( a;lloj, Ac. 4:12; e[teroj, 17:21), but usually alone, as in Mt. 5:13; 6:24. It is common with the genitive (Lu. 18:34). The adverbial use of ouvde,n is seen in Gal. 4:1 ouvde.n diafe,rei dou,lou, but the cognate accusative is a possible explanation (Gal. 2:6). Cf. ouvden in 1 Cor. 7:19. In Rev. 3: 17, ouvde.n crei,an e;cw, the neuter is not to be construed with crei,an.

4. Ouvde. ei-j. This is, of course, more emphatic than ouvdei,j. The usage appears often in Xenophon, Demosthenes and other classic writers, the LXX and the Atticists.327 For examples in the N. T. see Mt. 27:14; Jo. 1:3; Ac. 4:32; Ro. 3:10. The same principle appears in ouvk e;stin e[wj e`no,j├ Ro. 3:12 (Ps. 14:1, 3). Cf. also the separation of ouv- pote, in 2 Pet. 1:21.328

5. Ei-j - ouv. It is after the analogy of pa/j- ouv and distinctly emphatic, and is found in Demosthenes.329 Cf. Lu. 12:6, ea}n evx auvtw/n ouvk e;stin. So likewise Mt. 10:29, ea}n evx auvtw/n ouv pesei/tai. In Mt. 5:18 we have e[n- ouv mh,. For ouvdei.j o[stij see o[stij.

(b) Mhdei,j. In general the history of mhdei,j is parallel to that of ouvdei,j. It is naturally much less frequent and its use instead of ouvdei,j belongs to the discussion of Modes and Negative Particles. It follows in that matter the fate of mh,. Mhqei,j appears only once in the text of the N. T., Ac. 27:33. The use of mhde.n w;n, Gal. 6:3, may be compared with ouvqe,n eivmi, 1 Cor. 13:2. In 1 Th. 4:12 note mhdeno.j crei,an e;chte.

(c) Ou;tij AND Mh,tij. These were treated under tij. Following the editors in the separation of these forms, it is to be observed that mh,ti as mere particle occurs not merely in questions like mh,ti ou-to,j evstin o` Cristo,j; Jo. 4:29, but also with eiv) So eiv mh,ti in 1 Cor. 7:5; 2 Cor. 13:5. But in Lu. 9:13, eiv mh,ti poreuqe,ntej h`mei/j avgora,swmen, it is possible to take mh,ti as the object of avgora,swmen) Cf. Jo. 6:12, i[na mh, ti avpo,lhtai. But note mh,tige 1 Cor. 6:3. The use of tij with the conjunction mh, is not infrequent (Mk. 13:5) and with the negative adverb mh, also (Jo. 3:3, 5, etc.). So we have, contrary to the usual classic idiom, ouv- tij├ mh,──tij)330 The

752 A GRAMMAR OF THE GREEK NEW TESTAMENT

undoubted separation of ouv and mh, from tij in such examples as Mt. 11:27; 12:19; Lu. 8:51; 12:4; Jo. 7:4; 10:28; Ac. 28:21; of 1 Cor. 4: 5, etc., argues for the same thing where mh, tij and mh, ti happen to come together. The koinh, (Moulton, Prol., p. 246) supports the use of tij with the negative: Tb.P. 1 (ii/B.C.) mhdemia/j krath,sewj mhde. kuriei,aj tino.j evggai,ou periginome,nhj.

(d) WITH Pa/j.

1. Ouv pa/j. Used together the words call for little in the way of explanation. Ouv merely negatives pa/j as in classic Greek and= 'not every one.' Thus in Mt. 7:21, ouv pa/j o` le,gwn- eivseleu,setai, Jesus did not mean to say that 'no one' who thus addressed him could enter the kingdom of heaven. He merely said that 'not every one' would. Cf. also ouv pa/sa sa,rx 1 Cor. 15:39. The same principle applies to the plural ouv pa,ntej cwrou/si to.n lo,gon, Mt. 19: 11. Cf. Ac. 10:41; Ro. 9:6; 10:16. But my friend, Mr. H. Scott, notes that in Ro. 10:16 and 1 Cor. 15:39 ouv pa/j can well mean 'no,' and that in Mt. 7:21 and the other clauses where avlla,, occurs the avlla, negatives the whole of the preceding clause. This is certainly worth considering. Cf. Mt. 7:21 ouv pa/j o` le,gwn, with pa/j o` avkou,wn, in 7:26.

2. Ouv- pa/j. Here we have a different situation. The negative goes with the verb. A negative statement is made as to pa/j. The result is the same as if ouvdei,j had been used with an affirmative verb. So Mt. 24:22 (Mk. 13:20) ouvk aa}n evsw,qh pa/sa sa,rx, the idea is 'no flesh,' not 'not all flesh,' i.e. 'some flesh,' would have been lost. Cf. Lu. 1:37 ouvk avdunath,sei - pa/n r`h/ma, Ro. 3:20 (Gal. 2:16) ouv dikaiwqh,setai pa/sa sa,rx) See also Ac. 10:14 ouvde,pote- pa/n. Cf. ouvde. pa/n Rev. 7:16; 9:4. It is true that this idiom is very common in the LXX331 as a translation of lKo- alo. Cf. Ex. 12:16, 43; 20:10, etc. But it is not without analogy also 'in the papyri use of pa/j "with prepositions and adjectives of negative meaning. Thus a;neu or cwri.j pa,shj u`perqe,sewj, a recurrent formula, avnupeu,qenoi panto.j evpi,mou├ Tb.P. 105 (ii/B.C.); di,ca pa,shj evxousi,aj, Plutarch, Cons. ad Uxor., 1 (cf. Heb. 7:7)."332 Clearly the construction was in harmony with the koinh,.

3. Mh, - pa/j. The same principle applies. Cf. 1 Cor. 1:29, o[pwj mh. kauch,shtai pa/sa sa,rx) Here it is 'no flesh' as above with ouv - pa/j. See also Rev. 7:1. On the other hand mh. pa/j (1 Jo. 4:1)= 'not every' like ouv pa/j.

PRONOUNS ( vANTWNUMIAI) 753

Addenda 2nd ed.

4. Ouv mh,──pa/n in Rev. 21:27 does not differ at all from the ouv- pa/j and mh,──pa/j is in construction.

5. Pa/j- ouv). Here the ancient Greek idiom to a certain extent comes to one's relief.333 But the alo - lKo lies behind the LXX translation. It is less harsh than ouv - pa/j. Cf. Blass, Gr. of N. T. Gk., p. 178. The denial about pa/j is complete as with ouv- pa/j. See 1 Jo. 2:21, pa/n yeu/doj evk th/j avlhqei,aj ouvk e;stin. Cf. 1 Jo. 3:15; Eph. 5:5; Rev. 22:3.

6. Pa/j - mh, falls into the same category. Cf. Jo. 3:16; 6:39; 12:46; Eph. 4:29; 5:3. Here also the denial is universal. But most probably mhdei,j would have pleased an older Greek more.

7. Pa/j- ouv mh,) In Rev. 18:22 the same explanation holds.

8. Ouv- pa,ntej. With the plural ouvk eivsi.n pa,ntej evx h`mw/n, 1 Jo. 2:19, the matter is not so clear. Two translations are possible, as is seen in the American Revision. The text there is: "they all are not of us." The margin has: "not all are of us." The analogy of ouv- pa/j in the singular favours the first.

9. Pa,ntej ouv. With pa,ntej ouv koimhqhso,meqa, 1 Cor. 15:51, the ouv goes with the verb. The effect is the same as pa/j- ouv, above. 'We all shall not sleep' means that 'none' of us shall sleep. 'We shall all be changed.' Per contra, see ouv pa,ntej Ro. 10:16= 'not all.'

1 Cf. Schoemann, Die Lehre von den Redet. nach den Alten, p. 95: "Die Nomina benenrien die Dinge nach ihren Qualitaten, die Pronomina bezeichnen sie nach ihren Verhaltnissen."

2 Gildersleeve, Synt. of Cl. Gk., part i, p. 35.

3 Jann., Hist. Gk. Gr., p. 348.

4 Thumb, Handb., etc., p. 59 f.

5 Prol., p. 85.

6 Gildersleeve Studies, p. 240.

7 Sel. from the LXX, p. 65.

8 Gr. of N. T. Gk., p. 316.

9 In general the N. T. follows the classic idiom. W.-Sch., p. 194.

10 Cf. W.-Sch., p. 194.

11 Gr. of N. T. Gk., p. 166. Cf. Moulton, Prol., p. 86 f., who leaves the matter to the exegete.

12 Gr. of N. T. Gk., p. 316 f.

13 W.-Sch., p. 195.

14 Renaud, The Distributed Emphasis of the Pers. Pron., 1884.

15 Bernhardy, Wissensch. Synt. der griech. Spr., 1829, p. 45.

16 Prol., p. 86.

17 W.-Sch., p. 191.

18 Blass, Gr. of N. T. Gk., p. 164.

19 Thumb, Handb., p. 90.

20 Lang. of the N. T., p. 60. Cf. C. and S., Sel. from LXX, p. 29.

21 Prol., p. 86.

22 W.-Sch., p. 195; Blass, Gr. of N. T. Gk., p. 164.

23 Simcox, Lang. of the N. T., p. 61.

24 W.-M., p. 187.

25 Joh. Gr., p. 279.

26 Dyroff, Gesch. des Pron. Reflex., 1. Abt., p. 16.

27 Ib., pp. 68, 75, 80 f.

28 Ib., 2. Abt., p. 1 f.

29 Ib., 1. Abt., pp. 90 f., 126 f.

30 Ib., 2. Abt., pp. 69, 89.

31 Jann., Hist. Gk. Gr., p. 152.

32 Cf. Blass, Gr. of N. T. Gk., p. 165.

33 Wundt, Volkerpsych., 1. Bd., 2. Ti., 1904, p. 47.

34 Cf. W.-Th., p. 143; Blass, Gr. of N. T. Gk., p. 164.

35 Cf. also Simcox, Lang., etc., p. 53.

36 C. and S., Sel., etc., p. 65.

37 Moulton, Prol., p. 84 f.

38 Hellen., p. 108 f.

39 Gr. of N. T. Gk., p. 165.

40 Abbott, Joh. Gr., p. 281.

41 W.-Th., p. 148. Cf. C. and S., p. 65 f.

42 Farrar, Gk. Synt., p. 35.

43 Brug., Griech. Gr., p. 250.

44 Seymour, The Hom. Dial., p. 60.

45 Gr. of N. T. Gk., p. 168. Brugmann (Vergl. Gr., ii. 283) derives the poss. from the gen., while Delbruck (V, i. 213) obtains the gen. from the poss. Who can tell?

46 Simcox, Lang., etc., p. 54.

47 Monro, Hom. Gr., p. 170. Jann., Hist. Gk. Gr., p. 351, calls this the "determinative" pronoun. On the whole subject of aims see K.-G., I, pp. 651 ff.

48 Farrar, Gk. Synt., p. 35.

49 Cf. Dyroff, d. Pron. Reflex., 1. Abt., p. 16.

50 Blass, Gr. of N. T. Gk., p. 166 f.

51 Farrar, Gk. Synt., p. 33.

52 Ib.

53 Dyroff, Gesch. etc., Bd. I, 1892, p. 19.

54 W.-Sch., p. 205.

55 Blass, Gr. of N. T. Gk., p. 167.

56 Brug., Griech. Gr., p. 421. Cf. Meisterh., Att. Inschr., p. 194.

57 W.-Sch., p..205.

58 Ib.

59 Moulton, Prol., p. 87.

60 W.-Th., p. 257.

61 Blass, .Gr. of N. T. Gk., p. 169.

62 Prol., p. 90.

63 Ib.

64 Gr. of N. T. Gk., p. 169.

65 B. S., p. 123 f.

66 Cl. Rev., Dec., 1901, p. 440 f.; Prol., p. 90.

67 Prol., p. 90. Cf. Jann., Hist. Gk. Gr., p. 613.

68 Synt., p. 763 f.

69 Monro, Hom. Gr., p. 168 f.

70 Ib.

71 Riem. and Goelzer, Synt., p. 779.

72 K.-BL, I, p. 603.

73 Blass, Gr. of N. T. Gk., p. 145. See Gildersleeve, Synt., pp. ii, 216-226.

74 Whitney, Sans. Gr., p. 189.

75 Thompson, Synt. of Att. Gk., p. 67.

76 Monro, Hom. Gr., p. 176.

77 Reffel, Uber den Sprachgebr. des Agathias, 1894, p. 5.

78 Darr, Sprachl. 1899, p. 27.

79 Cf. Gildersleeve's ed. of First Apol., ch. 5 and note to p. 116.

80 Otto's ed., pp. 24, 90.

81 Cf. Gildersleeve, Justin Martyr, p. 116, for others.

82 Hadley and Allen, Gk. Gr., p. 216.

83 V. and D.'s Handb., etc., p. 297.

84 Cf. Moulton, Prol., p. 81.

85 Blass, .Gr. of N. T. Gk., p. 145.

86 K.-131., I, p. 608.

87 Whitney, Sans. Gr., p. 195.

88> Monro, Hom. Gr., p. 185.

89 Thompson, Synt. of Att. Gk., p. 68.

90 Cf. Griech. Gr. p. 241; Comp. Gr., III, p. 335.

91 So Thompson, Synt. of Att. Gk., p. 68, n. 3.

92 Griech. Or., p. 242.

93 Thompson, Synt. of Att. Gk., p. 65.

94 Or. of N. T. Gk., p. 170.

95 P. 216.

96 Quest. conviv. 1. 6. 1, th,nde th.n h`me,ran.

97 Brug., Griech. Gr., pp. 242, 428.

98 Thompson, Synt. of Att. Gk., p. 65.

99 Monro, Hom. Gr., p. 170.

100 Ib.

101 Simcox, Lang. of the N. T., p. 65 f.

102 Ib.

103 Ib., p. 66.

104 Cf. Blass, Gr. of N. T. Gk., p. 171.

105 Cf. W.-Sch., p. 217, where it is observed that elsewhere often dia. tou/to points to what goes before.

106 For exx. in earlier Gk. and literary koinh,, see W.-Sch., p. 217.

107 W.-Sch., p. 218.

108 See Gildersleeve, Synt., p. 331, for this "pseudo-attributive position."

109 W.-Th., p. 110.

110 W.-Sch., p. 221.

111 Cf. Blass, Gr. of N. T. Gk., pp. 126, 133.

112 Ib., p. 172.

113 Thompson, Synt. of Att. Gk., p. 66.

114 Blass, Gr. of N. T. Gk., p. 171.

115 Blass, ib., p. 172, explains evkei/noj as showing that the discourse passes from John to Jesus. But evkei/noj refers to John.

116 Thomp., Synt. of Att. Gk., p. 66.

117 W.-Sch., p. 219.

118 Blass, Gr. of N. T. Gk., p. 171.

119 Ib.

120 W.-Sch., p. 219.

121 W.-Th., p. 548.

122> W.-Th., p. 158.

123 Cf. ib., p. 159.

124 Brug., Griech. Gr., p. 242 f.

125 Ib.

126 Ib., p. 426 f.

127 Gr. of N. T. Gk., p. 172.

128 Joh. Gr., pp. 285, 567.

129 Abbott, ib., p. 568. He cites Mt. 27: 19, 63 as exx. of the good and the bad sense of evkei/noj. Cf. Lat. ille.

130 Gr. of N. T. Gk., p. 171.

131 Joh. Gr., p. 283.

132 Abbott, ib. For the Joh. use of evkei/noj see Steitz and A. Buttmann, Stud. in Krit. (1859, p. 497; 1860, p. 505; 1861, p. 267). Cf. Blass, Gr. of N.T. Gk., p. 172.

133 Prol., p. 91.

134 Ib.

135 Cf. Jann., Hist. Gk. Gr., pp. 320, 351.

136 Jann., Hist. Gk. Gr., p. 162.

137 Thack., Gr. of 0. T. in Gk., vol. I, p. 192.

138 Robertson, Short Gr. of the Gk. N. T., p. 81.

139 Brug., Griech. Gr., p. 556; Baron, Le Pron. Rel. et la Conj., 1891, p. 25. He notes that o[j went from dem. to rel. before O did.

140 Monro, Hom. Gr., pp. 186 ff.

141 Farrar, Gk. Synt., p. 35. [Oste survives in Pindar, Bacch., Ion. and Trag. choruses. Thompson, Synt. of Att. Gk., p. 68 f.

142 Whitney, Sans. Gr., p. 195. Baron, Le Pron. Rel. et la Conj. en Grec, p. 35. Cf. Delbruck, Vergl. Synt., III, p. 295 f.; Brug., Griech. Gr., p. 243.

143 Monro, Hom. Gr., p. 186. So 65 yap is ambiguous. On the anaphoric demonstr. o[j cf. Delbruck, Vergl. Synt., III, p. 310; Brug., Griech. Gr., p. 241.

144 Prol., p. 92.

145 Thack., Gr., vol. I, p. 192.

146 V. and D., Handb., etc., p. 56. "The disuse of o[j in common speech is characteristic; so simple a form ceased to satisfy the desire of emphasis." Jebb in V. and D., p. 302.

147 Cf. W.-Sch., p. 231 f.

148 W.-M., p. 207.

149 Cf. Abbott, Joh. Gr., p. 302.

150 W.-Sch., p. 233.

151 Mayser, Gr., p. 310.

152 Blass, Gr. of N. T. Gk., p. 173.

153> P. 226. Moulton, Prol., p. 93. Attraction of the relative to the case of the antecedent is not unknown in Lat. Cf. Draeger, Hist. Synt., Bd. II, p. 507. Hom. shows only one instance. Middleton (Analogies in Synt., p: 19) considers analogy the explanation of the origin of attraction.

154 Gr. of N. T. Gk., p. 173.

155 W.-Th., p. 163.

156 Thompson, Synt. of Att. Gk., p. 71; W.-Sch., p. 227,

157 Cf. Blass, Gr. of N. T. Gk., p. 174; Moulton, Prol., p. 93.

158 But in W.-Sch. (p. 225) ois is held to be essential to the structure. For attraction in John see Abbott, Joh. Gr., p. 298,

159 But see per contra W.-Sch., p. 223.

160 W.-Sch., p. 225. Hort in note to text says: " w-n probably a primitive error for oa}n."

161> Cf. W.-Sch., p. 226 f.

162 Cf. Thompson, Synt. of Att. Gk., p. 71.

163 Blass, Gr. of N. T. Gk., p. 175.

164 Cf. Blass, ib., and Comm. on Acts in loco.

165 This is more than "occasional," as Blass says (Gr. of N. T. Gk., p. 174). He rightly notes the absence of the article.

166 Thompson (Synt. of Att. Gk., p. 71 f.) finds this change only in the acc. But this is not Attic.

167 Blass, Gr. of N. T. Gk., p. 174.

168 K.-G., II, p. 433.

169 Gr. of N. T. Gk., p. 175.

170 P. ,201. Cf. also W.-M., p. 185.

171 Mack., Gr. of 0. T. in Gk., p. 46.

172 Cf. Blass, Gr. of N. T. Gk., p. 175; Simcox, Lang. of the N. T., p. 59.

173 Blass, Gr. of N. T. Gk., p. 175, cites ou- h` pnoh. auvtou/, from Clem. Cor. 21. 9.

174 Essai sur le grec de la Sept., p. 182.

175 Hellen., p. 128.

176 Cf. also Jann., Hist. Gk. Gr., p. 353.

177 Prol., p. 94.

178 Lang. of the N. T., p. 59. Cf. Farrar, Gk. Synt., p.1113.

179 W.-M., p. 186.

180 Bernhardy, p. 304; Jann., Hist. Gk. Gr., p. 354; Jelf, 833.2; K.-G., II, p. 432.

181 Gr. of N. T. Gk., p. 175,

182 "Normal" indeed. Thompson, Synt., p. 70.

183 Thack., Gr. of 0. T. in Gk., p. 25.

184 Cf. Thompson, Synt. of Att. Gk., p. 369.

185 Thompson, Synt. of Att. Gr., p. 74.

186 Prol., p. 93.

187 W.-M., p. 208.

188 Blass, Gr. of N. T. Gk., p. 331; Jann., Hist. Gk. Gr., p. 474.

189 Gr. of N. T. Gk., p. 176.

190 Simcox, Lang. of the N. T., p. 68.

191 Robertson, Short Gr. of the Gk. N. T., p. 178.

192 Gr. of N. T. Gk., p. 175.

193 Thompson, Synt., p. 74. Cf. also Jann., Hist. Gk. Gr., p. 473; Moulton, Prol., p. 93.

194 Prol., p. 93; Cl. Rev., Dec., 1901, p. 441.

195 Gr. of N. T. Gk., p.1175.

196 W.-M., p. 207 f.

197 W.-Sch., p. 237.

198 Ib., p. 236.

199 Cf. Thompson, Synt. of Att. Gk., p. 69, for the exx.

200 Gr. of N. T. Gk., p. 172 f.

201 Prol., p. 91.

202 Ib.; Cl. Rev., Dec., 1901, p. 441 f.

203 Hist. Gk. Gr., p. 352.

204 Quest., p. 245 f.

205 Gr., p. 139. For the confusion between o[j and o[stij see also Brug., Griech. Gr., p. 558 f.

206 W.-M., p 209, n. 3, where a very helpful discussion occurs.

207 V. and D., Handb. to Mod. Gk., p. 302.

208 Mayser, Gr., p. 310.

209 Prol., p. 92. [Ostij as 'who indeed' is common in Pisidia. Cf. Compernass, De Serm. Grace., p. 13.

210 Gr. of N. T. Gk., p. 173.

211 V. and D., Handb., p. 302.

212 Prol., p. 91.

213 Gr. of N. T. Gk., p. 173.

214 The pap. show the same situation. Moulton, Cl. Rev., April, 1904, p. 154. Thus h[ntina BM 77 (viii/A.D.), o[ntina inscr. J.H.S., 1902, p. 349, evx o[tou BM 190 (iii/?), e[wj o[tou NP. 56 (iii/A.D.).

215 Thack., Gr., p. 192.

216 Hist. Gk. Gr., p. 473.

217 Cf. W.-M., p. 208.

218 Hist. Gk. Gr., p. 473. It is more usual in the second of two questions. Cf. Riem. and Goelzer, Synt., p. 398.

219 Gr. of N. T. Gk., p. 176.

220 Lachmann, Praef., p. 43.

221 Blass, Gr. of N. T., p. 176.

222 W.-M., p. 208.

223 Simcox, Lang. of the N. T., p. 68.

224 Blass, Gr. of N. T. Gk., p. 176.

225 Simcox, Lang. of N. T., p. 68.

226 Prol., p. 94.

227 P. 238. The use of o[ti ti, lends colour to the notion of recitative o[ti.

228 Abbott, Job. Gr., p. 143.

229> Cf. Jann., Hist. Gk. Gr., p. 473.

230 Cf. Blass, Gr. of N. T. Gk., p. 175; W.-Sch., p. 236 f.; Viteau, Prop., pp. 62 ff.

231 Gr. of N. T. Gk., p. 175.

232 Jann., Hist. Gk. Gr., pp. 87, 168; Thumb, Handb., p. 94.

233 P. 311.

234 Cf. K.-G., II, p. 439, for exx. in the older Gk.

235 For a different explanation = ouv dh, pou evkpept. see Blass, Gr. of N. T. Gk., p. 179.

236 Gk. Gr., p. 167.

237 V. and D., Handb., p. 303.

238 Moulton, Prol., p. 93.

239 Blass, Gr. of N. T. Gk., p. 318. It is rare in anc. Gk. in this sense. K.-G., II, p. 439. Cf. o[pwj Lu. 24:20.

240 Thack., Gr., p. 192.

241 Jann., Hist. Gk. Or., p. 168.

242 P. 224.

243 But in the pap. Moulton finds avrourw/n - o[swn (Prol., p. 93). As a matter of fact in the N. T. o[soj nowhere occurs outside of the nom. and acc. except in Lu. 11:8 and Heb. 1:4; 8:6; 10:25.

244 Blass, Gr. of N. T. Gk., p. 179. Blass also cites Aristoph., Vesp., 213.

245 Moulton, Prol., p. 97; Blass, Gr. of N. T. Gk.,330.

246 Jann., Hist. Gk. Gr., p. 168.

247 Mayser, Gr., p. 311.

248 Monro, Hom. Gr., pp. 182 ff. For hist. of the matter see K.-B1., I, pp. 608

249 Hist. Gk. Gr., p. 353. Cf. also Brug., Griech. Gr., p. 560; Meisterh., Gr., p. 156; Dieterich, Byz. Arch., pp. 1, 198 f.

250 Gr., pp. 310 IL

251 See Schmid, Der Atticismus, p. 338; Volker, Synt. d. griech. Pap., p. 6; Ramsay, Cities and Bish. of Phrygia, XIX,. 429; Deiss., B. S., pp. 313 ff.; Moulton, Prol., p. 83.

252 Unters., p. 199. Winer (W.-Th., p. 107) rejects o` kai, as relative.

253 Cl. Rev., April, 1904, p. 155.

254 Cf. Whitney, Sans. Gr., p. 194; Brug., Griech. Gr.,.pp. 117, 244.

255 Jann., Hist. Gk. Gr., p. 163.

256 lb., p. 164.

257 Einl., p. 207 f.

258 Blass, Gr. of N. T. Gk., p. 176.

259 Blass, Gr. of N, T., p. 176.

260 Jann., Hist. Gk. Gr., p. 164.

261 Blass, Gr. of N. T. Gk., p. 177.

262 Ib.

263 Jelf, 874, obs. 4.

264 Draeger, Hist. Synt., p. 103.

265 Thack., Gr., p. 192.

266 Prol., p. 77.

267> Ib.

268> Thompson, Synt., p. 74.

269 Simcox, Lang. of the N. T., p. 67.

270 Prol., p. 93 f.; Cl. Rev., Apr., 1904, p. 154 f.

271 Dieterich, Unters., p. 200.

272 Thumb, Theol. Literzaturzeit., xxviii, p. 423, (quoted in Moulton, Prol.,

273 Prol., p. 93.

274 As Simcox does, Lang. of the N. T., p. 69 f.

275> Cf. W.-Sch., p. 241; Moulton, Prol., p. 93.

276 W.-Sch., p. 241; Blass, Gr. of N. T. Gk., p. 331.

277 Gr. of N. T. Gk., p. 175.

278 Cf. W.-Sch., p. 241.

279 W.-Sch., p. 240.

280 Gr. of N. T. Gk., p. 177.

281 Thompson, Synt., p. 74. Cf. Brug., Gliech. Gr., p. 561.

282 Thumb, Handb., p. 94.

283 Hist. Gk. Gr., p. 163. Cf. Dieterich, Unters., p. 202.

284 Thackeray, Gr., p. 192.

285 Ib.

286 Moulton, Prol., p. 95.

287 Hist. Gk. Gr., p. 163.

288 Brug., Griech. Gr., p. 244.

289 Ib. Interrogative and indefinite is at bottom the same word. Cf. Hartung, Uber die Casus in der griech. and Sprache, p. 279.

290 W.-Sch., p. 242.

291 W.-M., p. 212 f.; Blass, Gr. of N. T. Gk., p. 178.

292 Moulton in W.-M., p. 213.

293 Gr. of N. T. Gk., p. 178.

294 Hatz., Einl., p. 207; W.-Sch., p. 243.

295 Prol., p. 97.

296 Ib.

297 Thompson, Synt., p. 77.

298 Moulton, Prol., p. 57.

299 Jann., Hist. Gk. Gr., p. 320.

300 Thack., Gr., p. 192.

301 Moulton, Prol., p. 80.

302 Ib.

303 Ib.

304 Ib., p. 79. Cf. Thack., Gr., p. 192.

305 Cf. Thumb, Handb., p. 96; Jann., Hist. Gk. Gr., p. 178. On the whole subject of distrib. pron. see Brug., Die distrib. tend die kollekt. Num. der indoger. Spr., 1907.

306 W.-Sch., p. 246 f.

307 Blass, Gr. of N. T. Gk., p. 179.

308 Ib., p. 180.

309 Prol., p. 79.

310 W.-Sch., p. 245.

311 Gr. of N. T. Gk., p. 180,

312 lb.

313 Ib., p. 318.

314 Thack., Gr., p. 192.

315 Mayser, Gr., p. 312.

316 Cf. W.-Sch., p. 244.

317 Synt., p. 77.

318 Cf. W.-Sch., p. 245.

319 Gr. of N. T. Gk., p. 179.

320 Moulton, Prol., p. 79.

321 W.-Sch., p. 246.

322 Hist. Gk. Gr., p. 170. But see Schwyzer, Perg. Inschr., p. 114, for idea that the change is due to t and d being pronounced alike.

323 Att. Inschr., p. 259.

324 Gr., pp. 551f.

325 Thack., Gr., p. 60,

326 Thumb, Hellen., p. 14; Mayser, Gr., p. 150 f.

327 W.-Sch., p. 248; Schmid, Atticismus, II, p. 137 f.

328 Cf. W.-Sch., p. 249.

329 Ib., p. 178.

330 Cf. Blass, Gr. of N. T. Gk., p. 256.

331 W.-M., p. 215.

332 Moulton, Prol., p. 246. Cf. Cl. Rev., Dec., 1901, p. 442; Apr., 1904, p. 155.

333 W.-M., p. 215.