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.
Size3.25 Mb.
1   ...   17   18   19   20   21   22   23   24   ...   33
(ib. and pp. 193 f.) in Eph 117 it will have to be a virtual wish

clause, i!na serving merely to link it to the previous verb; but

dw<^ is preferable. This banishment of the final optative only

means that the NT writers were averse to bringing in a

1 Meanwhile we may observe that Blass's dictum (p. 213) that the ei] c. opt.

form is used "if I wish to represent anything as generally possible, without

regard to the general or actual situation at the moment," suits the NT exx.

well; and it seems to fit the general facts better than Goodwin's doctrine of a

"less vivid future" condition (Goodwin, Greek Gram. 301).


construction which was artificial, though not quite obsolete.

The obsolescence of the optative had progressed since the

time of the LXX, and we will only compare the writers

and papyri of i/A.D. and ii/A.D. Diel in his program De

enuntiatis finalibus, pp. 20 f., gives Josephus (1/A.D.) 32

per cent. of optatives after i!na, o!pwj and w[j, Plutarch

Lives (i/A.D.) 49, Arrian (ii/A.D.) 82, and Appian (ii/A.D.) 87,

while Herodian (iii/A.D.) has 75. It is very clear that the

final optative was the hall-mark of a pretty Attic style. The

Atticisers were not particular however to restrict the optative

to past sequence, as any random dip into Lucian himself will

show. We may contrast the more natural Polybius (ii/B.c.),

whose percentage of optatives is only 7,1 or Diodorus (i/B.C.),

who falls to 5. The writer of 4 Mac (i/A.D.) outdoes all

his predecessors with 71, so that we can see the cacoethes

Atticissandi affecting Jew as well as Gentile. The papyri

of our period only give a single optative, so far as I have

observed: OP 237 (late ii/A.D.) i!na . . . dunhqei

little later we have LPw (ii/iii A.D.) i!n ] eu@odon a@rti moi

ei@hi, in primary sequence; and before long, in the Byzantine

age, there is a riot of optatives, after e]a

The deadness of the construction even in the Ptolemaic

period may be well shown from TP 1 (ii/B.C.) h]ci


and citations will suffice to show why the NT does not

attempt to rival the litterateurs in the use of this resuscitated


Potential We turn to the other main division of

Optative. the Optative, that of which ou] and a@n are

frequent attendants. With a@n the Potential

answers to our own I should, you or he would, generally

following a condition. It was used to express a future in

a milder form, and to express a request in deferential style.

But it is unnecessary to dwell upon this here, for the table

given above (p. 166) shows that it was no longer a really

living form in NT times. It was literary, but not artificial,

as Luke's use proves. It figures 30 times in LXX, or

19 times when 4 Mac is excluded, and its occurrences are

1 See Kalker's observations, Quaest. 288 f.


tolerably well distributed and not abnormal in form. We

should note however the omission of a@n, which was previously

cited in one phrase (p. 194 n.).1 We shall see that a@n tends

to be dropped with the indicative; the general weakening of

the particle is probably responsible for its omission with the

optative as well. Ti31 al, does not differ

from ti

conveyed by such an omission as appears in 4 Mac 513


forgive." In other ways we become aware how little differ-

ence a@n makes in this age of its senescence. Thus in Par

P 35 (ii/B.C.) e]ch2 the dropping

of a@n would affect the meaning hardly at all, the contingent

force being practically nil. So when Luke says in 162

e]ne ti< a}n qe

Ac 1017, Lk 1526 1836 (D) 946,--there is a minimum of

difference as compared with Ac 2133 e]punqa

he might be," or Lk 1836 xAB ti< ei@h tou?to. Not that a@n

c. opt. in an indirect question is always as near as in this case

to the unaccompanied optative which we treat next. Thus in

the inscr. Magn. 215 (i/A.D.) e]perwt%? . . . ti< au]t&? shmai

ti< a}n poih

tence, "If I were to do what, should I be secure?" i.e. "what

must I do that I may . . . ?" So in Lk 611 ti< a}n poih

is the hesitating substitute for the direct ti< poih24

ti< a}n gewill this come to?"

Cf Esth 133 puqome

might be brought to pass" (RV). In direct question we

have Ac 1718 ti< a}n qe

4 in a softened assertion meets us in Ac 2629 xcAB, eu]cai

a@n "I could pray." Among all the exx. of a@n c. opt. in Luke

there is only one which has a protasis, Ac 831 pw?j ga>r a}n

dunain mh< tij o[dhgh1 Par P 63 (ii/B.C.) has a dropped de in a place where it is needed badly:

a@lla me>n ou]qe
n o!ti e!lkesqai bebouou]qe>n a}—if one may conjecture without seeing the papyrus. (So Mahaffy

now reads: he also substitutes a]lla>, and kakw?j for e!lkesqai.)

2 It is unfortunate that this crucial 43 is missing, for e]reuna?to (an unaug-

mented form) is quite possible, though less likely. The papyrus has another

optative, in indirect question, ei@hsan ei]sporeusa


condition with the less vivid form in the apodosis.1 No

more need be said of this use; nor need we add much about

the other use of the Potential, that seen in indirect questions.

The tendency of Greek has been exactly opposite to that of

Latin, which by the classical period had made the optative

("subjunctive") de rigueur in indirect questions, whatever

the tense of the main verb. Greek never admitted ti

= quis sim into primary sequence, and even after past tenses

the optative was a refinement which Hellenistic vernacular

made small effort to preserve. On Luke's occasional use of it

we need not tarry, unless it be to repeat Winer's remark

(p. 375) on Ac 2133, where the opt. is appropriate in asking

about the unknown, while the accompanying indicative, "what

he has done," suits the conviction that the prisoner had com-

mitted some crime. The tone of remoteness and uncertainty

given by the optative is well seen in such a reported question

as Lk 315 mh

j ei@h o[ Xristo23 to> ti

. . . o[ tau?ta me

observes the rule of sequence, as he does in the use of pri

(p. 169).2

"Unreal" The Indicative—apart from its Future,

Indicative. which we have seen was originally a sub-

junctive in the main is suited by its whole

character only to positive and negative statements, and not

to the expression of contingencies, wishes, commands, or other

subjective conceptions. We are not concerned here with the

forces which produced what is called the "unreal" use of the

indicative, since Hellenistic Greek received it from the earlier

age as a fully grown and normal usage, which it proceeded to

limit in sundry directions. Its most prominent use is in the

two parts of the unfulfilled conditional statement. We must

1 It is sentences of this kind to which Goodwin's "less vivid form "does

apply: his extension of this to be the rule for the whole class I should ven-

ture to dissent from—see above, p. 196 n.

2 On the general question of the obsolescence of the optative, reference may

be made to F. G. Allinson's paper in Gildersleeve Studies 353 ff., where itacism

is alleged to be a contributory cause. Cf OP 60 (iv/A.D.) i!n ] ou#n e@xoite . . . kai>

katasth dokoi? is

similarly a misspelt subj. (or indic.). When oi had become the complete

equivalent of h, ^, ei, and ai of e, the optative forms could no longer preserve

phonetic distinctness. Prof. Thumb dissents: see p. 240.


take this up among the other Conditional Sentences, in

vol. ii., only dealing here with that which affects the study of

the indicative as a modus irrealis. This includes the cases of

omitted a@n,1 and those of ou] instead of mh<. It happens that

the only NT example of the latter has the former character-

istic as well: Mk 1421 ( = Mt 2624) kalo>n au]t&? ei] ou]k


the ultimate sense which makes this "unreal" at all: as far

as form goes, the protasis is like Heb 1225 ei] e]kei?noi ou]k


was a warning to us" might have formed the apodosis, and so

that sentence and this would have been grammatically similar.

We might speak thus of some villain of tragedy, e.g. "A good

thing if (nearly = that) there never was such a man." Trans-

ferred as it is to a man who is actually present, the saying

gains in poignancy by the absence of the contingent form.

Ei] ou] occurs fairly often with the indicative, but elsewhere

always in simple conditions: see above, p. 171. The dropping

of a@n in the apodosis of unfulfilled conditions was classical with

phrases like e@dei, e]xrh?n, kalo>n h#n. Such sentences as "If he

did it, it was the right thing," may be regarded as the

starting-point of the use of the indicative in unfulfilled

condition, since usage can easily supply the connotation "but

he did not do it." The addition of a@n to an indicative

apodosis produced much the same effect as we can express in

writing by italicising "if": "if he had anything, he gave

it," or "if he had anything, in that case (a@n) he gave it,"

alike suggest by their emphasis that the condition was not

realised. We further note the familiar fact that the imper-

fect in all "unreal" indicatives generally denotes present

time:2 cf the use with o@felon in Rev 315 and 2 Co 111.

(These are the sole NT examples of this kind of unreal

indicative. The sentences of unrealised wish resemble

those of unfulfilled condition further in using the aorist

(1 Co 48) in reference to past time; but this could

1 Cf OP 526 (ii/A.D.) ei] kai> mh> a]ne to>n lo

OP 5:30 (ii/A.D.) ei] plei?on de< moi pare

(ii/B.C.) ou]k a]pe h]na

2 In Lk 176 note present, in protasis. Cf Par P 47 (ii/B.C.,=Witk. p. 641

mh> mikro


hardly have been otherwise.1 The difference of time in

the real and unreal imperfect will be seen when we drop

the a@n in the stock sentence ei@ ti ei#xon, e]di

had anything (now), I should give it," which by eliminating

the a@n becomes "if (i.e. whenever) I had anything, I used to

give it." Goodwin (MT § 399, 410 ff.) shows that this use

of the imperf. for present time is post-Homeric, and that it is

not invariable in Attic—see his exx. For the NT we may

cite Mt 2330 2443 (^@dei). Lk 1239, Jn 410 1121. 32, 1 Jn 219

as places where ei] with imperf. decidedly denotes a past

condition; but since all these exx. contain either h@mhn or ^@dein,

which have no aorist, they prove nothing as to the survival

of the classical ambiguity—we have to decide by the context

here, as in all cases in the older literature, as to whether

present or past time is meant. The distribution of tenses in

the apodosis (when a@n is present) may be seen in the table on

p. 166. The solitary pluperf. is in 1 Jn 219. It need only

be added that these sentences of unfulfilled condition state

nothing necessarily unreal in their apodosis: it is of course

usually the case that the statement is untrue, but the sen-

tence itself only makes it untrue "under the circumstances"

(a@n), since the condition is unsatisfied. The time of the

apodosis generally determines itself, the imperfect regularly

denoting present action, except in Mt 2330 (h@meqa).

Unrealised purpose makes a minute addition to the tale of

unreal indicatives in the NT. The afterthought e@dramon in

Gal 22, with which stands 1 Th 35, has plenty of classical

parallels (see Goodwin MT § 333), but no further exx. are

found in NT writers, and (as we saw above, p. 193 n.) the

former ex. is far from certain. Such sentences often depend

on unfulfilled conditions with a@n, and the decadence of these

carries with it that of a still more subtle and less practical

form of language.

1 There is one ex. of o@felon c. fut., Gal 512, and there also the associations of

the particle (as it now is) help to mark an expression never meant to he taken

seriously. The dropping of augment in w@felon may be Ionic, as it is found

in Herodotus; its application to 2nd or 3rd pers. is probably due to its being

felt to mean "I would" instead of "thou shouldst," etc. Note among the

late exx. in LS (p. 1099) that with me . . . o]le

ment. Grimm-Thayer gives LXX parallels. See also Schwyzer Perg. 173.


Nominal Verbs THE mention of "The Verb" has been omitted

and Verbal in the heading of this chapter, in deference to

Nouns. the susceptibilities of grammarians who wax

warm when lu

Verb instead of the Noun. But having thus done homage

to orthodoxy, we proceed to treat these two categories almost

exclusively as if they were mere verbal moods, as for most

practical purposes they are. Every schoolboy knows that

in origin and in part of their use they belong to the

noun; but on this side they have been sufficiently treated

in chapters iv. and v., and nearly all that is distinctive is


The Infinitive:— The Greek Infinitive is historically either

Its Origin. a locative (as luei@nai, etc.) from a noun base closely connected

with a verb.1 We can see this fact best from a glance at

Latin, where regere is obviously the locative of a noun like

genus, reigi, the dative of a noun much like rex except in

quantity, and rectum, -tut, -tu the accusative, dative, and loca-

tive, respectively, of an action-noun of the 4th declension. In

Plautus we even find the abstract noun tactio in the nomi-

native governing its case just as if it were tangere. Classical

Greek has a few well-known exx. of a noun or adjective

governing the case appropriate to the verb with which it is

closely connected. Thus Plato Apol. 18B ta> mete

sthAnt. 789 se> fu1 On the morphology of the Infinitive see Giles Manual2= 468 ff. It should be

noted that no syntactical difference survives in Greek between forms originally

dative and those which started in the locative.


Sanskrit would show us yet more clearly that the so-called

infinitive is nothing but a case—any case—of a noun which

had enough verbal consciousness in it to "govern" an object.

The isolation and stereotyping of a few of these forms produces

the infinitive of Greek, Latin, or English. It will be easily

seen in our own language that what we call the infinitive is

only the dative of a noun: Middle English had a locative with

at. In such a sentence as "He went out to work again," how

shall we parse work? Make it "hard work," and the Noun claims

it: substitute "work hard," and the Verb comes to its own.

One clear inference from all this is that there was originally

No voice no voice for the infinitive. Dunato>j qauma<-

distinction. sai, "capable for wondering," and a@cioj,


verbal noun in the same way; but one means "able to

wonder," and the other "deserving to be wondered at." The

middle and passive infinitives in Greek and Latin are merely

adaptations of certain forms, out of a mass of units which

had lost their individuality, to express a relation made

prominent by the closer connexion of such nouns with

the verb.

Survivals of There are comparatively few uses of the

Case force. Greek Infinitive in which we cannot still

trace the construction by restoring the dative

or locative case from whence it started. Indeed the very

fact that when the form had become petrified the genius of the

language took it up afresh and declined it by prefixing the

article, shows us how persistent was the noun idea. The

imperative use, the survival of which we have noticed above

(pp. 179 f.), is instructive if we are right in interpreting it in

close connexion with the origins of the infinitive. A dative

of purpose used as an exclamation conveys at once the

imperatival idea. The frequent identity of noun and verb

forms in English enables us to cite in illustration two lines of

a popular hymn :—

“So now to watch, to work, to war,

And then to rest for ever!”

A schoolmaster entering his classroom might say either "Now

then, to work!" or "at work!"—dative or locative, express-


ing imperative 2nd person, as the hymn lines express 1st

person. Among the NT exx., Phil 316 has the 1st,1 and the

rest the 2nd person. The noun-case is equally traceable in

many other uses of the infinitive. Thus the infinitive of

purpose, as in Jn 213 a[lieu2 proskunh?sai

for worshipping, —of consequence, as Heb 610 e]pilaqeto

the extent of forgetting,—and other "complementary" infini-

tives, as Heb 1115 kairo>n a]nakaopportunity for returning,

2 Tim 112 dunato>j fulacompetent for guarding. The force

of such infinitives is always best reached by thus going back

to the original dative or locative noun.

Tenses. From the account just given of the

genesis of the infinitive it follows that it

was originally destitute of tense as much as of voice. In

classical Sanskrit the infinitive is formed without reference

to the conjugation or conjugations in which a verb forms its

present stem: thus  cru (klucrotum, pres. crnomi--

yuj (iungo), yoktum, yunajmi— bhu (fufui, be), bhavi-

tum, bhavami. We can see this almost as clearly in Latin,

where action-nouns like sonitum, positum, tactum and tactio,

etc., have no formal connexion with the present stem seen

in sonat, penit, tangit. The s in lu?sai has only accidental

similarity to link it with that in –e@lusa. But when once

these noun forms had established their close contact with the

verb, accidental resemblances and other more or less capricious

causes encouraged an association that rapidly grew, till all

the tenses, as well as the three voices, were equipped with

infinitives appropriated to their exclusive service. Greek had

been supplied with the complete system from early times,

and we need say nothing further on the subject here, since

the infinitive presents no features which are not shared with

other moods belonging to the several tenses.2

1 Brugmann, Gram.3 517 n., regards w[j e@poj ei]pei?n as being for ei@pwmen, and

coming therefore under this head. It is a literary phrase, found only in Heb

79: cf the would-be literary papyrus, OP 67 (iv/A.D.). On this and other exx.

of the "limitative infin." see Grunewald in Schanz Beitrage II. iii. 22 ff.,

where it is shown to be generally used to qualify pa?j or ou]dei
2 The Hellenistic weakening of the Future infinitive, which in the papyri

is very frequently used for aorist or even present, would claim attention here

if we were dealing with the Koinh< as a whole. See Kalker 281, Hatzidakis

190 f., 142 f. The NT hardly shows this form: apart from e@sesqai, I

Infinitive of Some important questions arise from the

Purpose, etc. free use in NT of the infinitive which is

equivalent to i!na c. subj. In ThLZ, 1903,

p. 421, Prof. Thumb has some suggestive remarks on this

subject. He shows that this infinitive is decidedly more

prominent in the Koinh< than in Attic, and is perhaps an

Ionic element, as also may be the infin. with tou?, of which the

same is true. In the Pontic dialect of MGr—as mentioned

above, pp. 40 f.—the old infin. survives, while it vanished

in favour of na< c. subj. in European MGr, where the infin.

was less prominent in ancient times.a Now the use of the

infin. in Pontic is restricted to certain syntactical sequences.

To these belong verbs of movement, like come, go up (cf Lk

1810, Par P 49—ii/B.C., = Witk. 29—e]a>n a]nabw? ka]gw> pros-

kunh?sai), turn, go over, run, rise up, incline, etc. The NT (and

LXX) use generally agrees with this; and we find a similar

correspondence with Politic in the NT use of the infinitive

after such verbs as bou

e]pixeirw?, ai]sxu

e]w?, e]pitre

. With other verbs, as

parakalw?, the i!na construction prevails. This correspondence

between ancient and modern vernacular in Asia Minor, Thumb

suggests, is best explained by assuming two tendencies within

the Koinh<, one towards the universalising of i!na, the other

towards the establishment of the old infinitive in a definite

province: the former prevailed throughout the larger, western

portion of Hellenism, and issued in the language of modern

Hellas, where the infinitive is obsolete; while the latter held

sway in the eastern territory, exemplifying itself as we should

expect in the NT, and showing its characteristic in the dialect

spoken to-day in the same country. Prof. Thumb does not

pretend to urge more than the provisional acceptance of this

theory, which indeed can only be decisively accepted or rejected

when we have ransacked all the available inscriptions of Asia

Minor for their evidence on the use of the infinitive. But it

can only cite He 318, Ac 267 (WH mg). Jn 212 has xwrhby xwrh?sai in the later MSS; but the future is wanted here. The aorist may

be due to the loss of future meaning in xwrh

scribes wrote. The obsoleteness of fut. infin. with me

been remarked already (p. 114 n.). [a See p. 249.


is certainly very plausible, and opens out hints of exceedingly

fruitful research on lines as yet unworked.

Ecbatic" i!na The long debated question of " i!na e]k-


new light which has come in since H. A. W. Meyer waged heroic

warfare against the idea that i!na could ever denote anything

but purpose. All motive for straining the obvious meaning

of words is taken away when we see that in the latest stage

of Greek language-history the infinitive has yielded all its

functions to the locution thus jealously kept apart from it.

That i!na normally meant "in order that" is beyond ques-

tion. It is perpetually used in the full final sense in the

papyri, having gained greatly on the Attic o!pwj. But it

has come to be the ordinary construction in many phrases

where a simple infinitive was used in earlier Greek, just as

in Latin ut clauses, or in English those with that, usurp the

prerogative of the verbal noun. "And this is life eternal,

that they should know thee" (Jn 173), in English as in

the Greek, exhibits a form which under other circum-

stances would make a final clause. Are we to insist on

recognising the ghost of a purpose clause here?a Westcott

says that i!na here "expresses an aim, an end, and not only

a fact." The i!na clause then, as compared with (to>) ginw<-

skein, adds the idea of effort or aim at acquiring knowledge of

God. I will not deny it, having indeed committed myself

to the assumption as sufficiently established to be set down

in an elementary grammar.1 But I have to confess myself

troubled with unsettling doubts; and I should be sorry now

to commend that i!na as strong enough to carry one of the

heads of an expository sermon!

Let us examine the grounds of this scepticism a little

more closely. In Kalker's often quoted monograph on the

language of Polybius, pp. 290 ff., we have a careful presenta-

tion of i!na as it appears in the earliest of the Koinh< writers,

who came much nearer to the dialect of common life than

the Atticists who followed him. We see at once that i!na

has made great strides since the Attic golden age. It has

invaded the territory of o!pwj, as with fronti
1 Introd.2 217. [a See p. 249,


daformer occurs only in Tit 38; the latter eleven times. And

instead of Attic o!pwj, or Polybian i!na, behold the infinitive

in every occurrence of the two! Under Kalker's next head

Polybius is brought into an equally significant agreement

with the NT. He shows how the historian favours i!na after

words of commanding, etc., such as diasafei?n, ai]te?sqai,


quoted: suneta

penth pezou>j pentakosi pro>j Messhni

i!na tou>j i@souj tou pezou>j e]capostei

The equivalence of infin. and i!na c. subj. here is very plain.

In the later Koinh< of the NT, which is less affected by

literary standards than Polybius is, we are not surprised to

find i!na used more freely still; and the resultant idiom in

MGr takes away the last excuse for doubting our natural

conclusions. There is an eminently sensible note in SH on

Rom 1111, in which the laxer use of i!na is defended by the

demands of exegesis, without reference to the linguistic

evidence. The editors also (p. 143) cite Chrysostom on

520: to> de> i!na e]ntau?qa ou]k ai]tiologi

e]stin. It will be seen that what is said of the weakening

of final force in i!na applies also to other final constructions,

such as tou? c. infin. And on the other side we note that

w!ste in passages like Mt 271 has lost its consecutive force

and expresses a purpose.a It is indeed a repetition after

many centuries of a development which took place in the

simple infinitive before our contemporary records begin. In

the time when the dative do

were still distinct living cases of a verbal noun, we may

assume that the former was much in use to express designed

result: the disappearance of distinction between the two

cases, and the extension of the new "infinitive mood" over

many various uses, involved a process essentially like the

vanishing of the exclusively final force in the normally final

constructions of Greek, Latin, and English. The burden of

making purpose clear is in all these cases thrown on the

context; and it cannot be said that any difficulty results,

except in a minimum of places. And even in these the diffi-

culty is probably due only to the fact that we necessarily

a See p. 249.


read an ancient language as foreigners: no difficulty ever

arises in analogous phrases in our own tongue.

Latinism? The suggestion of Latin influence in this

development has not unnaturally been made

by some very good authorities;1 but the usage was deeply

rooted in the vernacular, in fields which Latin cannot have

touched to the extent which so far-reaching a change

involves. A few exx. from papyri may be cited :—OP 744

(i/B.C.) e]rwtw? se i!na mh> a]gwnia

i!na soi fulaxqw?si (cf BU 19 (ii/A.D.)). BU 531 (ii/A.D.)

parakalw? se i!na kata


BM 21 (ii/B.C.) h]ci

occurs in the same papyrus. Par P 51 (ii/B.C.) le

i!na proskunh

immediately of Mt 43 1620, Mk 510 39 etc., the naturalness

of the development is obvious from the simple fact that the

purpose clause with i!na is merely a use of the jussive sub-

junctive (above, pp. 177 f.), which makes its appearance after

a verb of commanding or wishing entirely reasonable. The

infinitive construction was not superseded: cf AP 135 (ii/A.D.)

e]rwtw? se mh> a]melei?n mou. We need add nothing to Winer's

remarks (WM 422 f.) on qe5

is a particularly good ex. under this head, in that qe

has both constructions: we may trace a greater urgency

in that with i!na, as the meaning demands. From such

sentences, in which the object clause, from the nature of

the governing verb, had a jussive sense in it which made

the subjunctive natural, there was an easy transition to

object clauses in which the jussive idea was absent. The

careful study of typical sentences like Mt 1025 88 (contrast

311) 186, Jn. 127 (contr. Lk 1519) 434 158. 13, Lk 143 (for which

Winer quotes a close parallel from Epictetus), will show

anyone who is free from predisposition that i!na can lose the

last shred of purposive meaning.2 If the recognition of a

purpose conception will suit the context better than the denial

1 So Gotzeler De Polybi elocutione 17 ff. for prosemh<: also Kalker op. cit., and Viereck SG 67. Against these see Radermacher

RhM lvi. 203 and Thumb Hellen. 159. 2 See further pp. 240 f.


of it, we remain entirely free to assume it; but the day is

past for such strictness as great commentators like Meyer

and Westcott were driven to by the supposed demands of

grammar. The grammarian is left to investigate the extent

to which the i!na construction ousted the infinitive after

particular expressions, to observe the relative frequency of

these usages in different authors, and to test the reality of

Thumb's proposed test (above, p. 205) for the geographical

distribution of what may be to some extent a dialectic


Consequence. The consecutive infin. with w!ste has

been already alluded to as admitting some-

thing very much like a purely final meaning. The total

occurrences of w!ste in the NT amount to 83, in 51 of which

it takes the infin. A considerable number of the rest,

however, are not by any means exx. of what we should call

w!ste consecutive with the indicative: the conjunction be-

comes (as in classical Greek) little more than "and so" or

"therefore," and is accordingly found with subj. or imper.

several times. Of the strict consecutive w!ste c. indic. there

are very few exx. Gal 213 and Jn 316 are about the clearest,

but the line is not easy to draw. The indicative puts the

result merely as a new fact, co-ordinate with that of the

main verb; the infinitive subordinates the result clause so

much as to lay all the stress on the dependence of the result

upon its cause. Blass's summary treatment of this construc-

tion (p. 224) is characteristic of a method of textual criticism

which too often robs us of any confidence in our documents

and any certain basis for our grammar. "In Gal 213 there is at

any rate a v.l. with the infin."—we find in Ti "ascr sunupaxqh-

nai"--,"while in Jn 316 the correct reading in place of w!ste

is o!ti which is doubly attested by Chrys. (in many passages)

and Nonnus."a Those of us who are not impressed by such

evidence might plead that the text as it stands in both places

entirely fits the classical usage. It is just "the importance

attaching to the result"—to quote one of Blass's criteria

which he says would have demanded the indic. in Ac 1539 in

a classical writer—which accounts for the use of the indica-

tive: in Jn 316, "had the other construction—w!ste dou?nai,

so much as to give—been used, some stress would have been

a See p. 249.


taken off the fact of the gift and laid on the connexion

between the love and the gift."1 Even if the indicative

construction was obsolete in the vernacular—which the

evidence hardly suffices to prove—, it was easy to bring in the

indicative for a special purpose, as it differed so little from

the independent w!ste = and so. The infinitives without

w!ste in consecutive sense were explained above (p. 204),

upon Heb 610. So in OP 526 (ii/A.D.), ou]k h@mhn a]paqh>j


Sometimes we meet with rather strained examples, as those in

the Lucan hymns, 154.72 especially. The substitution of i!na

c. subj. for the infin. occasionally makes i!na consecutive, just

as we saw that w!ste could be final: so 1 Jn 19, Rev 920,

Jn 92—where Blass's "better reading" o!ti has no authority

earlier than his own, unless Ti needs to be supplemented.

Blass quotes a good ex. from Arrian, ou!tw mwro

i@d^. We should not however follow him in making i!na con-

secutive in Lk 945, for the thought of a purpose of Providence

seems demanded by parakekalumme4 we can

concede, but 2 Co 117 is better treated as final: Paul is

disclaiming the mundane virtue of unsettled convictions,

which aims at saying yes and no in one breath. See p. 249.

The infinitive when used as subject or

Infinitive as object of a verb has travelled somewhat

subject or further away from its original syntax. We

object. may see the original idea if we resolve

humanum est errare into "there is something human in

erring." But the locative had ceased to be felt when the

construction acquired its commanding prevalence, and the

indeclinable verbal noun could become nom. or acc. without

difficulty. The i!na alternative appears here as it does in the

purpose and consequence clauses, and (though this perhaps

was mere coincidence) in the imperative use (pp. 176 and

178 f.). Thus we have Mt 529 al sumfe25 a]rketo

Jn 1839 sunh3 ei]j e]la34

e]mo>n brw?ma< e]stin, all with iva in a subject clause. See Blass's

full list, p. 228, and note his citation from "Barnabas" 513,

e@dei i!na pa
1 I quote from my Introduction 218, written before Blass's book.


Lk 143, 1 Jn. 53, Jn 1513, etc. The prevalence of the i!na in

Jn has its bearing on Prof. Thumb's criteria described above

(pp. 40 f. and 205); for if the fondness of Jn for e]mo

characteristic of Asia Minor, that for i!na goes the other way.

It would be worth while for some patient scholar to take up

this point exhaustively, examining the vernacular documents

among the papyri and inscriptions and in the NT, with care-

ful discrimination of date and locality where ascertainable.

Even the Atticists will yield unwilling testimony here; for a

"wrong" use of i!na, if normal in the writer's daily speech,

could hardly be kept out of his literary style there was a

very manifest dearth of trained composition lecturers to correct

the prose of these painful litterateurs of the olden time!

Schmid, Atticismus iv. 81, shows how this "Infinitivsurrogat"

made its way from Aristotle onwards. Only by such an inquiry

could we make sure that the dialectic distribution of these

alternative constructions was a real fact in the age of the

NT. Tentatively I should suggest--for time for such an

investigation lies wholly below my own horizon--that the

preference was not yet decisively fixed on geographical lines,

so that individuals had still their choice open. The strong

volitive flavour which clung to i!na would perhaps commend

it as a mannerism to a writer of John's temperament; but one

would be sorry to indulge in exegetical subtleties when he

substitutes it for the infinitive which other writers prefer.

The Accusative We might dwell on the relation of

and Infinitive the accus. c. infin. (after verbs of saying,

and substitutes. believing, and the like) to the periphrasis

with o!ti which has superseded it in nearly

all the NT writers. But no real question as to difference

of meaning arises here; and it will suffice to cite Blass's

summary (pp. 230 ff.) and refer to him for details. He

shows that "the use of the infinitive with words of believing

is, with some doubtful exceptions, limited to Luke and Paul

(Hebrews), being a 'remnant of the literary language'

(Viteau [i.] 52)." So with other verbs akin to these: Luke

is indeed "the only writer who uses [the acC. and infinitive]

at any length, and even he very quickly passes over into the

direct form." The use of w[j instead of o!ti is limited, and

tends to be encroached upon by pw?j: of Hatzidakis 19, who


(might not however to have cited Ac 421 in this connexion

The combination w[j o!ti in 2 Co 519 1121, 2 Th 22, is taken

by Blass (Gr.2 321 f.) as equivalent to Attic w[j c. gen. abs.,

the Vulgate quasi representing it correctly. It must be

noted that in the vernacular at a rather later stage it meant

merely "that": thus CPR 19 (iv/A.D.) prw

Share with your friends:
1   ...   17   18   19   20   21   22   23   24   ...   33

The database is protected by copyright © 2019
send message

    Main page