Greenwood professor of hellenistic greek and indo-european philology in the victoria university of manchester tutor in new testament language and literature

Download 3.25 Mb.
Date conversion08.11.2016
Size3.25 Mb.
1   ...   18   19   20   21   22   23   24   25   ...   33
de u[pa

mou a]podo

combined where the single word would be adequate." He

quotes another papyrus, w[j o!ti xreostei?tai e]c au]tou? o[ ku


in the sense of w[j or o!ti alone: see Roberts-Gardner 179.

Winer (p. 771) cites Xenophon, Hellen. III. ii. 14, ei]pw>n

w[j o!ti o]knoi2) and Plummer

repeat the reference; but the editors have agreed to eject

o!ti from the text at that place. Its isolation in earlier

Greek seems adequate reason for flouting the MSS here.

Winer's citation from the Argument to the Busiris of Isocrates,

kathgo daimo

dispose of Blass's "unclassical" (as Plummer supposes), since

the argument is obviously late.1 We may follow Lightfoot

and Blass without much hesitation.

Nominative for In classical Greek, as any fifth-form boy

Accusative. forgets at his peril, the nominative is used

regularly instead of the accusative as subject

to the infinitive when the subject of the main verb is the

same: e@fh ou]k au]to>j a]lla> Kle

is by no means obsolete in NT Greek, as passages like 2 Co

102, Rom 93, Jn 74 (WH text), serve to show; but the ten-

dency towards uniformity has produced a number of violations

of it. Heb 724 has a superfluous au]to4:

Mt 2632 inserts me, Phil 313 e]mauto

p. 238 f., gives instances, and remarks that translations

from Latin (Viereck, SG 68) exhibit this feature.a Kalker

(p. 280) anticipates Viereck in regarding this as a case of

propter hoc as well as post hoc. But the development of
1 Dr J. E. Sandys (Aristotle's Constitution of Athens, p. xxviii) makes the

author of the u[poAreopagitieus "a Christian writer of perhaps the

sixth century." He kindly informs me that we may assume the same age for

that to the Busiris. [a See p. 249


Greek in regions untouched by Latin shows that no outside

influence was needed to account for this levelling, which

was perfectly natural.

Mixed The accus. c. inf. and the o!ti construction

Construction. have been mixed in Ac 2710, by an inadvert-

ence to which the best Attic writers were

liable. See the parallels quoted by Winer (p. 426), and add

from humbler Greek OP 237 (ii/A.D.) dhlw?n o!ti ei] ta> a]lhqh?

fanei kri pra?gma. Also see Wellh. 23.

The Articular We will proceed to speak of the most

Infinitive. characteristic feature of the Greek infinitive

in post-Homeric language. "By the sub-

stantial loss of its dative force," says Gildersleeve (AJP iii.

195), "the infinitive became verbalised; by the assumption of

the article it was substantivised again with a decided increment

of its power." Goodwin, who cites this dictum (MT 315),

develops the description of the articular infinitive, with

"its wonderful capacity for carrying dependent clauses and

adjuncts of every kind," as "a new power in the language, of

which the older simple infinitive gave hardly an intimation."

The steady growth of the articular infinitive throughout the

period of classical prose was not much reduced in the

Hellenistic vernacular. This is well seen by comparing the

NT statistics with those for classical authors cited from Gilder-

sleeve on the same page of Goodwin's MT. The highest

frequency is found in Demosthenes, who shows an average of

1 25 per Teubner page, while he and his fellow orators

developed the powers of the construction for taking dependent

clauses to an extent unknown in the earlier period. In the

NT, if my calculation is right, there is an average of 68 per

Teubner page—not much less than that which Birklein gives

for Plato. The fragmentary and miscellaneous character of

the papyri make it impossible to apply this kind of test, but

no reader can fail to observe how perpetual the construction

is. I have noted 41 exx. in vol. i of BU (361 papyri), which

will serve to illustrate the statement. An interesting line

of inquiry, which we may not at present pursue very far,

concerns the appearance of the articular infinitive in the

dialects. Since it is manifestly developed to a high degree

in the Attic orators, we should naturally attribute its fre-


quency in the Hellenistic vernacular to Attic elements in

the Koinh<; and this will be rather a strong point to make

against Kretschmer's view (p. 33), that Attic contributed

no more than other dialects to the resultant language. To

test this adequately, we ought to go through the whole

Sammlung of Greek dialect-inscriptions. I have had to

content myself with a search through Cauer's representative

Delectus, which contains 557 inscriptions of all dialects except

Attic. It will be worth while to set down the scanty

results. First comes a Laconian inscr. of ii/B.C., 32 (= Michel

182) e]pi> to> kalw?j . . . diecagnhke

"Mysteries" inscr., no. 47 (= M. 694, Syll. 653, 91 B.C.), which

has four or five instances, all with prepositions. Four Cretan

exx. follow, all from ii/B.C., and all in the same formula, peri> tw?

(once tou?) gene

56, 54, 60). (The Gortyn Code (Michel 1333, v/B.C.) has no

ex., for all its length.) Then 148 ( = M. 1001, the Will of

Epikteta), dated cir. 200 BC., in which we find pro> tou? ta>n


end of iv/B.C., is with one exception the oldest ex. we have:

paragenon e]poih

taj tou>j poli pot ] au[tou>j politeu

No. 171, from Carpathus, Michel (436) assigns to ii/B.C.: it

has pro> tou? misqwqh

apparently iii/B.C., has [peri> t]ou? parorig xw

The Delphian inscr. no. 220 has pro> tou? paramei?nai. Elis

contributes one ex., no. 264 ( = M. 197), dated by Michel in

the middle of iv/B.C., and so the oldest quoted: peri> de> t&?

a]postala?men . . . to> . . . ya

us (no. 431 = M. 357), from ii/B.C., e]pi> tw?i pragmateuqh?nai.

I have looked through Larfeld's special collection of Boeotian

inscriptions, and find not a single example. Unless the

selections examined are curiously unrepresentative in this

one point, it would seem clear that the articular infinitive

only invaded the Greek dialects when the Koinh< was already

arising, and that its invasion was extremely limited in extent.

To judge from the silence of Meisterhans, the Attic popular

speech was little affected by it. It would seem to have been

mainly a literary use, starting in Pindar, Herodotus, and the

tragedians, and matured by Attic rhetoric. The statistics of


Birklein (in Schanz Beitr., Heft 7) show how it extends during

the lives of the great writers, though evidently a matter of

personal taste. Thus Sophocles has 94 examples per 100

lines, Aeschylus 63, and Euripides only 37. Aristophanes

has 42; but if we left out his lyrics, the frequency would be

about the same as in Euripides. This is eloquent testimony

for the narrowness of its use in colloquial speech of the Attic

golden age; and the fact is significant that it does not appear

in the early Acharnians at all, but as many as 17 times in

the Plutus, the last product of the poet's genius. Turning to

prose, we find Herodotus showing only 07 examples per Teubner

page, and only one-fifth of his occurrences have a preposition.

Thucydides extends the use greatly, his total amounting to 298,

or more than 5 a page: in the speeches he has twice as many

as this. The figures for the orators have already been alluded

to. The conclusion of the whole matter—subject to correction

from the more thorough investigation which is needed for

safety—seems to be that the articular infinitive is almost

entirely a development of Attic literature, especially oratory,

from which it passed into the daily speech of the least

cultured people in the later Hellenist world. If this is true,

it is enough by itself to show how commanding was the part

taken by Attic, and that the literary Attic, in the evolution

of the Koinh<.

The application of the articular infin. in NT Greek does

not in principle go beyond what is found in Attic writers.

We have already dealt with the imputation of Hebraism which

the frequency of e]n t&? c. inf. has raised. It is used 6 times

in Thucydides, 26 times in Plato, and 16 in Xenophon; and

the fact that it exactly translates the Hebrew infin. with b

does not make it any worse Greek, though this naturally in-

creases its frequency.a Only one classical development failed

to maintain itself, viz. the rare employment of the infin. as a

full noun, capable of a dependent genitive: thus in Demos-

thenes, to< g ] eu# fronei?n au]tw?n, "their good sense"; or in Plato,

dia> panto>j tou? ei#nai. Heb 215 dia> panto>j tou? zh?n is an exact

parallel to this last, but it stands alone in NT Greek, though

Ignatius, as Gildersleeve notes, has to> a]dia

The fact that zh?n was by this time an entirely isolated

infinitive form may account for its peculiar treatment.b A

a b See D. 249.


similar cause may possibly contribute to the common verna-

cular (not NT) phrase ei]j pei?n,1 which we compared above

(p. 81) to the Herodotean a]nti< c. anarthrous infin. The

prepositions which Birklein (p. 104) notes as never used

with the infin. retain this disqualification in the NT: they

are, as he notes, either purely poetical or used in personal

constructions. It may be worth while to give a table of

relative frequency for the occurrences of the articular infini-

tive in NT books. Jas has (7 =) 108 per WH page;

Heb (23 =) 109; Lk (71 =) nearly 99; Paul (106 = )

89 (in Pastorals not at all); Ac (49 =) 7 (73 in cc. 1-12,

68 in cc. 13-28); 1 Pet (4 =) 59; Mt (24 =) 35; Mk

(13 =) 32; Jn (4 =) 076; Rev (1 =) 027. [Mk] 169-20

has one ex., which makes this writer's figure stand at

143: the other NT books have none. It will be found

that Mt and Mk are about level with the Rosetta Stone.2

Tou? c. inf. The general blurring of the expressions

which were once appropriated for purpose,

has infected two varieties of the articular infinitive. That

with tou? started as a pure adnominal genitive, and still

remains such in many places, as 1 Co 164, a@cion tou?


of the ordinary genitive categories in a fair proportion of

its occurrences, the correspondence seems generally to be

accidental: the extension which began in the classical period

makes in later Greek a locution retaining its genitive force

almost as little as the genitive absolute. The normal use of

tou? c. inf. is telic. With this force it was specially developed

by Thucydides, and in the NT this remains its principal

use. We will analyse the exx. given in the concordance,

omitting those in which tou?, is governed by a preposition,

and those which are due to the LXX. Mt has 6 exx.:

in one of them, 2132, tou? pisteu?sai gives rather the content

than the purpose of metemelh

of the total for the NT. In Lk we have 23 exx., of which

5 may be due to dependence on a noun, and about one-half

1 But not to ei]j bafrom Theodoret. See Kuhner3 § 479. 2. Add an ex. with a@xri from Plutarch

p. 256 D. An inscription of iii/B.C. (OGIS 41, Michel 370) has a]postalei>j . . .

e]pi> ta>j parabola>j tw?n dikw?n lamba2 See p. 241.


seem clearly final; in Ac there are 21, with 2 adnominal,

and less than half final. Paul shows 13 (only in Rom, Gal,

1 and 2 Co, Phil), but there is not one in which purpose is

unmistakable. In Heb there is one adnominal, one (115)

final or quasi-final. Jas 517 (object clause), 1 Pet 417

(adnominal), and the peculiar1 Rev 127 supply the remainder.

Before turning to grammatical detail, let us parenthetically

commend the statistics just given to the ingenious analysts

who reject the unity of the Lucan books. The uniformity

of use is very marked throughout Lk and Ac: cf Ac 271

("We"-document) with 1520 203, Lk 2122 with Ac 915, Ac 2027

("We"-document) with 1418. Note also the uniform pro-

portion of final tou?, and the equality of total occurrences.

When we observe that only Paul makes any marked use of

tou? c. inf., outside Lk and Ac (the two writers together

accounting for five-sixths of the NT total), and that his use

differs notably in the absence of the telic force, we can

hardly deny weight to the facts as a contribution to the

evidence on the Lucan question. In classifying the uses of

this tou?, we note how closely it runs parallel with i!na. Thus

Lk 171 a]ne e]lqei?n, and Ac 1025

e]ge12), where the tou? clause represents

a pure noun sentence, in which to< would have been more

correct, may be paralleled at once by Lk 143, po

tou?to i!na e@lq^; After verbs of commanding we may have

tou? or i!na. We find the simple infin. used side by side with

it in Lk 176f.. (purpose) and 179. It is not worth while to

labour any proof that purpose is not to be pressed into

any example of tou? where the context does not demand

it; but we must justify our assertion about Paul. It is

not meant that there are no possible or even plausible

cases of final tou?, but only that when Paul wishes to express

purpose he uses other means. In the majority of cases tou?

c. inf. is epexegetic (Rom 124 73 812, 1 Co 1013), adnominal

(Rom 1523, 1 Co 910 164, 2 Co 811, Phil 321) or in a regular

ablative construction (Rom 1522, 2 Co 18). The rendering

1 WH make this a quotation from Dan 1013.20: the former verse names

Michael, who in the latter says e]pistre ktl (Theodotion).

See below.


"so as to" will generally express it. The nearest to pure final

force are Rom 66 and Phil 310; but in both it would be

quite as natural to recognise result as purpose—the main

purpose is expressed by a clause with i!na in each case, and

the tou? c. infin. comes in to expound what is involved in

the purpose stated. An extreme case of explanatory infin.

is that in Rev 127, where po

polemh?sai with subject in the nominative. The construction

is loose even for the author of Rev, but the meaning is clear:

we might illustrate the apposition by Vergil's "et cer ta-

men erat, Corydon cum Thyrside, magnum;" or more closely

still—if we may pursue our former plan of selecting English

sentences of similar grammar and widely different sense—

by such a construction as "There will be a cricket match,

the champions to play the rest."

Pro>j to< and Two other modes of expressing purpose

ei]j to< c. infin. have been, to a more limited extent, infected

by the same general tendency. Pro>j to<

c. infin. occurs 5 times in Mt and once in Mk, with clearly

final force, except perhaps in Mt 528, where it might rather

seem to explain ble


and Ac 319 stand alone in Luke, and the former is hardly

final: we go back to a more neutral force of pro

reference to the duty" (Winer). Paul has it 4 times,

and always to express the "subjective purpose" in the

agent's mind, as W. F. Moulton observes (WM 414 n., after

Meyer and Alford). This then is a locution in which the

final sense has been very little invaded. Ei]j to< c. infin.

is almost exclusively Pauline. It occurs thrice in Mt, in

very similar phrases, all final; Mk, Lk and Ac have it once

each, with final force fairly certain. Jas and 1 Pet have

two exx. each, also final; and the same may probably be

said of the 8 exx. in Heb. The remaining 44 exx. are evenly

distributed in Paul, esp. Rom, Th, and Co--none in Col,

Philem and the Pastorals. Westcott on Heb 51 distinguishes

between i!na and ei]j to<, which he notes as occurring in

close connexion in a considerable number of passages: " i!na

appears to mark in each case the direct and immediate

end, while ei]j to< indicates the more remote result aimed

at or reached." This seems to be true of both tou? and

ei] to>. Since we have seen that i!na itself has largely lost

its appropriation to telic force, it would naturally follow

that ei]j to< would lose it more easily: on the whole,

however, this is hardly the case. On Heb 113, Moulton

and Westcott, independently, insist on the perseverance of

the final meaning, in view of the writer's usage elsewhere.

The ei]j to> gegone

depend on kathrti

of the fiat in Gen 1. Paul's usage is not so uniform. It is

difficult to dispute Burton's assertion (MT § 411) that in

Rom 123, 2 Co 86, Gal 317 (not, I think,1 in 1 Th 216) ei]j to<

"expresses tendency, measure of effect, or result, conceived

or actual." Add (with WM 414 n.) exx. of ei]j to< expressing

the content of a command or entreaty (as 1 Th 212), or

acting for the epexegetic inf. (1 Th 49). Purpose is so

remote here as to be practically evanescent. We must

however agree with SH in rejecting Burton's reasoning as

to Rom 120; for this belongs to the category of passages

dealing with Divine action, in which contemplated and actual

results, final and consecutive clauses, necessarily lose their

differentia. It has been often asserted--cf especially a

paper by Mr A. Carr on "The Exclusion of Chance from the

Bible," in Expos. v. viii. 181 ff.--that Hebrew teleology is

responsible for the blurring of the distinction between pur-

pose and consequence: it is a "subtle influence of Hebrew

thought on the grammar of Hellenistic Greek." This might

be allowed—as a Hebraism of thought, not language--in

passages like that last mentioned, where the action of God

is described. But the idea that "Hebrew teleology" can

have much to do with these phenomena as a whole is put

out of court by the appearance of the same things in lan-

guage which Semitic influences could not have touched. We

Evidence of the have already shown this for i!na. A few exx.

Papyri, etc. may be cited for 70 from vernacular

witnesses:—BU 665 (1/A.D.) a]melei?n tou?

gra ou#n e[toima proairei?n,

i!n ] e@xi tou? pwlei?n: cf Mt 1825, Jn 57, for parallel construe-

1 See Findlay CGT in loc., where strong reasons are given for accepting

Ellicott's interpretation, seeing here the purpose of God.

Lions with e@xw. BU 1031 (ii/A.D.) fro

JHS, 1902, 369 (Lycaonian inscr., iii/A.D. or earlier) t&?

dixotomh loepo>n zh?n ei$j (cause). NP 16

(iii/A.D.) kwlu spei42, Ac 1418, etc.

BU 36 (ii/iii A.D.) tou? zh?n metasth?sai: cf 2 Co 18. BU

164 (ii/iii A.D.) parakalw? se . . . pei?sai au]to>n tou? e]lqei?n.

BM 23 (ii/B.C.) prosdeome

(i/A.D.) tou? se> mh>i eu[reqh?nai, apparently meaning "because

of your not being found," as if t&?:1 the document is illiterate

and naturally ejects the dative. OP 86 (iv/A.D.) e@qoj e]sti>n

tou? parasxeqh?nai. OP 2'75 (i/A.D.) tou? a]pospaqh?nai


1 Co 96. BU 46 (ii/A.D.) eu]kairi

Lk 226. BU 625 (ii/iii A.D.) pa?n poi

1   ...   18   19   20   21   22   23   24   25   ...   33

The database is protected by copyright © 2017
send message

    Main page