صور الصفحة
النشر الإلكتروني

The accusative of extent is common, and needs no illustration. But even here, particularly with the so-called adverbial accusative, eis was called in to help: e.g. εἰς πολλὰ [τῇ πό]λει χρήσιμον (R.E.G., '90, p. 75, Νο. 41), τῆς TAτρúdos Eis Távta evepyéτnv (J.H.S., '02, p. 123, No. 53). This usage has its roots in classical Greek; cf. Thuc. 4. 81, Spaσrýριov és тà Távтa. Robertson (Gram. of Gk. N.T.3, p. 594) notes eis with accusative of extent of time in N.T.; on the other hand, the adverbial accusative has no preposition in Rom. 15. 22, ένεκοπτόμην τὰ πολλά.

The accusative of the direct object occasionally takes the place of an earlier dative or genitive. The dative gives way in K(úpɩe) Bońłn (= -ei) Táτnv (J.H.S., '02, p. 368, No. 104). The distinction between the genitive of the object (partial control) and the accusative (complete control) failed to maintain itself. It was becoming blurred even in the fourth-century dialect inscriptions.1 Our inscriptions provide the following examples: ἰχθὺν . . . ὃν ἐδράξατο παρθένος ἅγνη (C. and B., No. 657), ἐπιτυχόντα τὰς . . . φιλοδοσίας (ibid., Νο. 299), ἧς ἰστόργην οὐδ ̓ ἐν νέκροις λαθοίμην (Ramsay, Ε.Κ.Ρ., p. 143, Α), τοῖς . . . ὀνηθεῖσιν τὸ τῆς ζωῆς μέρος (C. and B., No. 658). With the first example, cf. N.T., 1 Cor. 3. 19, ó Spaσσóμevos тoùs σopoús; for the second, Moulton (C.R., 'o1, p. 437) gives a papyrus example of the verb with both cases, τῆς Ῥωμαίων πολιτείας καὶ ἐπιγαμίαν ἐπιτυχόντες.

Of the cognate accusative, which does not seem to be very common in the papyri (C.R., 'o1, p. 436), we have an example from the Lycos valley, πλεύσας . . . πλοὰς ἑβδομήκοντα δύο (C. and B., No. 420), and another from Dorylaion, ζήσαντι ἀμέμπτως τὸν ἐπιμόχθητον βίον (C.I.G., 3816), with which cf. N.T., I Pet. 4. 2, τὸν ἐπίλοιπον ἐν σαρκὶ βιῶσαι χρόνον.

The double accusative is commoner in modern than in ancient Greek. Verbs which formerly took accusative and genitive or instrumental dative now generally take two accusatives.2 This preference for the double accusative is already evident in the Kový: e.g. in the LXX., Ex. 31. 3, Évéπλŋσa aỶtòv πνεῦμα θεῖον. In the inscriptions of Asia Minor the verbs τιμῶ and κοσμῶ take a double accusative : e.g. πρίν σε νυνφικὸν ἰστέφανον κοσμήσαμεν (Ramsay, E.R.P., p. 139, D). From these the construction was extended to ȧvaríonμi, ἀνίστημι, possibly helped by the Attic ἱστάναι τινὰ χαλκοῦν. Sometimes the verb is left unexpressed: e.g. Tóvde σe Muydovín (J.R.S., '12, p. 80). The frequency of expressions like ἐκόσμησεν τὸν ἔγγονον and ἀνέστησεν τὴν γυναῖκα caused the two sets of verbs to become interchangeable, so that Tiμ even takes the dative: e.g. Téкvo èтeiμnoav (Ramsay, E.R.P., p. 209, No. 6). A different double accusative, in place of accusative and true dative, occurs with ἐπιγράφω: λίθους ἐπέγραψα γράμματα (J.R.S., 12, p. 92, No. 19); we have also τὴν στήλην ἐπέγραψα, without the γράμματα (Sterrett, Epigraphical Journey to Asia Minor, No. 196). The accusative was gaining on all hands.

1 See Giles, Manual, p. 312.

2 See Thumb, Handb., p. 37 (E.T.).

The adverbial accusative begins to replace adverbs in -ws. The reason was doubtless phonetic, since with the change from pitch to stress accent, -ws and -os could not be distinguished. The accusative was the natural case to come to the rescue. Though it is the plural ending -a that has finally prevailed, examples still exist of the adverbial use of the singular: e.g. Xiyo, Myári, beside λiya.1 Our inscriptions provide us, as I think, with an example λιγάκι, λίγα.1 of either in a confession inscription of Badinlar (C. and B., Nd 48),2 dıñ0a τὴν κώμη β' ἄναγνα, and in another of the same group (No. 50), παραγ(γ)έλ(λ)ων πâσi μndis avayvov ȧvaẞîтe (-ai). There can be no doubt of the 'popular' character of their language!

The use of the accusative in apposition to a whole sentence is well established in classical Greek. In an inscription of Eumenia (C. and B., No. 368) we have, τὸ δὲ πάντων μεῖζον, ἔσται αὐτῷ πρὸς τὸν θεόν. χάριν, an accusative of this kind, occurs occasionally without a dependent genitive, thus showing that its substantival force lived alongside of the more frequent adnominal use: e.g. Παρ]θενὶς Ηφαιστίωνι χάριν (C.I.G., 3848), Τειμαῖος . . . μητρὶ . Xápu (Ath. Mitt., 'oo, p. 425, No. 40). Sometimes it seems to assume a concrete meaning (= μvnμeîov), but the development was only made possible by its frequent use in this construction.3 Similar accusatives are euxýv (very common) and xaipeσrnpíov (C. and B., No. 289). Arising possibly from the same construction is an idiomatic use of Taûτа, apparently favoured by the Hellenistic vernacular. In C. and B., No. 232, a certain Gaius finishes an exposition of his principles with the words ταῦτα, φίλοι. Another (ibid., No. 635) seems to be combating Christian views, and says: ovк nμŋv. ἐγενόμην· οὐκ ἔσομαι· οὐ μέλι μοι ὁ βίος· ταῦτα. It gives the nature of an ultimatum to the schoolboy's threats in Pap. Oxy. 119. Here again the ubiquitous preposition intrudes, e.g. eis μvýμns xápiv (C. and B., No. 261). Günther (I.F. XX., p. 162) notes eis... Xápw in a dialect inscription of Phokis; even Sophocles (Antig. 30) has πpòs xáρiv Bopâs, though perhaps that means χάριν βορᾶς. more than xápu Bopas. Possibly the preposition sometimes stresses the idea of purpose: e.g. Ῥοῦφος . . . τὸ σύνκρουστον καὶ τὸν . . γράδον . . . κατεσκεύασεν εἰς ἡρῶον (C. and B., No. 212), and in Acta S. Marinae, the Greek of which represents the vernacular of the district of Pisidian Antioch, κρατήσω αὐτὴν ἐμαυτῷ εἰς γυναῖκα (Usener, p. 17, 1. 12). With this cf. LXX., Gen. 4I. 45, ἔδωκεν αὐτῷ τὴν ̓Ασεννέθ . . . εἰς γυναῖκα. The modern preposition here is γιά, as in τὴν ἐζήτησε γιὰ γυναῖκα, τὸν κλαίγω yià Tεlaμévo. We have, in fact, to note three constructions, not differing sensibly in meaning: (1) simple apposition, (2) os with accusative, (3) eis with accusative. Blass (Gram. of N.T. Gk., p. 93) says of the second use that it may be a Hebraism.' But it is found in the modern dialect of Silli: e.g. va πáρηs čŋv gópŋ 5 évaîxa (Dawkins, Mn. Gk. in Asia Minor, p. 288, l. 31).

1 See Dieterich, Untersuchungen, p. 182.

The reading, which differs from Ramsay's,

was suggested by Mr. M. N. Tod.

3 For other examples, see Nachmanson, in Eranos, '09, p. 44 sqq.

See Thumb, Handb., § 50 (a).

Still less can eis with accusative be regarded as a Hebraism. Apart from the examples cited above, we have a papyrus example quoted by Moulton (Proleg., p. 72). Its use, in fact, whether instead of a simple accusative, or with elvai instead of a predicate nominative, only represents in Greek a language tendency which was developed earlier and to a far greater extent in Latin in the case of the predicate dative: 'ludibrium aliquem habere' and 'ludibrio aliquem habere' will both stand, and to give Latin also a third construction we can add 'pro ludibrio habere.'

The inscriptions illustrate the growing preference of the prepositions for the accusative. For éri we have éπì ő (C. and B., No. 45), 'in view of which,' though a levelling of quantity is possible here. For ὑπό, ὁ ὑπὸ αὐτὴν τόπος (J.P., '91, p. 91, No. 12 A), and vπò émiμéλeiav (C. and B., No. 299), with which one might contrast Antiphon 123, 20, vπò éπiμeλeías Oεoû. This agrees with the general Kowý tendency to confine vπó to the accusative, except where agency is denoted. Brugmann (Gr. Gr., p. 517 n.) notes that vπép ('for the sake of ') with accusative is found in the N.W. dialects in the third century B.C., and states that the construction is not a characteristic of the Kown. In an inscription of Akmonia (C. and B., No. 561) we have vπèp exń(v): the inscription is mutilated, but apparently we have the usual appositional accusative strengthened this time by vπép, not eis; that this is so seems certain from a number of examples of the same use of eveкev and xápiv: e.g. μνήμην ἕνεκεν (J.H.S., '99, p. 113, Νο. ΙΟΙ), μνήμην [χά]ριν (ibid., P. 120, No. 121). Finally, we note, beside the simple accusative with ỏdúpoμai, e.g. τὰ τέκνα τὸν ἐμὸν πότμον ὠδύροντο (Ramsay, E.R.P., p. 144, Β), διά with accusative in the Pelagian Legends, wdúpero di' avτýv (Usener, p. 13, l. 14).


On the whole the genitive of the object governed by verbs maintains its ground: e.g. του βωμού κήδεται (C. and B., No. 7), προνοησάντων τῆς ȧvaoτáσe[ws (ibid., No. 26). Other verbs which take the genitive are iepaтeúw, ἀναστάσε[ως ἱερῶμαι, στρατηγώ, τυγχάνω, and (generally) καταφρονεῖν. This genitive still survives in some of the dialects.1

For the adnominal genitive note åμaρтwλòs deŵv in confession inscriptions, the construction arising out of the equivalence of the notions of 'debt' and 'sin'; ȧμаρτwλós follows opeiλérŋs (Ramsay in Exp. Times, X., p. 55). Beside the genitive we have εἰς with accusative: ἁμαρτωλὸς ἔστω εἰς τὴν Λητώ (C.I.G. 4303). Tỷ TEKOVσŋ AỦToû (C. and B., No. 376), where the participle has lost much of its verbal force, is classical.

The genitive of 'reference' is quite frequent in the earlier Greek, and Nachmanson (in Eranos, '09, p. 30 sqq.) has shown that it is not uncommon in the Kown. Perhaps it is most at home with verbs of judicial action (accusing, condemning, etc.), but it is by no means confined to these. Its origin seems Brugmann, Gr. Gr., p. 446.

1 See Thumb, Handb., p. 35.


clear; it is related to what appears to be the pre-ethnic meaning of the adverbial genitive-the sphere within which the verbal action takes place. The relation is seen, e.g. in Thuc. I. 93. 4, τῆς . . . θαλάσσης . . . ἐτόλμησεν eiπeîv és ávlekтéa éσTí, where the genitive denotes in the maritime sphere.' But whenever the local force recedes, the genitive assumes a vaguer meaning, that of reference. For Latin cf. Plaut. Bacch. 3. 4. 5, 'nam mihi diuini nunquam quisquam creduat.' The following are the clearest examples from the inscriptions of Phrygia: Π. τέκνων σωτηρίας Μηνὶ εὐχήν (Gött. Gelehrte Anz., '97, p. 405), Ε. και Π. . . . ἐποιήσαμεν μνήμης (J.H.S., '98, p. 98, Νο. 39), ανεστήσαμεν . μνήμης (C.I.G. 3989b), κατέδωκεν . . . στεφα[νω]τικοῦ δην. διακόσια (C. and B., No. 4II), ἀποτείσι προστείμου τῷ ταμείῳ Env. p' (ibid., No. 28). With this last cf. Leg. XII. Tab. VIII. 2, ‘si iniuriam faxsit XXV. poenae sunto.' But the vagueness of the idea of reference led to the desire for closer definition, and this was attained by means of prepositions. Thus, besides the very common μνήμης ἕνεκα, etc., we have δώρου χάριν (C. and B., No. 389), 'by way of a gift,' and also πроσтeíμov óvóμati (J.H.S., '97, p. 411, No. 14), σтeдаνwτιкоû óvóμaтi (C.I.G. 3919), where óvóμari might be rendered 'under the heading of.' Cf. Thuc. 4. 60, óvóμati évvóμg Evμμaxías, 'under the honourable name of an alliance.' One might quote from N.T., Mark 9. 41, év óvóμaтı öтi Xpiσтoû éσté, where the defining genitive is replaced by a or clause. Finally in an inscription of Hierapolis (Judeich, No. 327) we have eis πроσтíμον Xóyov, a good example of the analytic tendency of Hellenistic Greek.

This leads us to consider the use of Toû with the infinitive, which our Hellenistic grammarians are either content to call 'epexegetic,'1 a term which leaves the case still unexplained, or to regard as a construction which could apparently be used capriciously, as subject, as object, to denote remote purpose,' and what not.2 Now if we accept Brugmann's view (Gr. Gr., p. 448) that this is also a genitive of reference, it is possible to obtain a consistent view of its use in Hellenistic Greek. Its frequency in LXX. is doubtless due to the influence of the Hebrew infinitive-construct. We note at the outset that in N.T. the construction is for the most part confined to Luke and Paul, whose writings belong most closely to the literary language. As Moulton observes, 'in the main it belongs to the higher stratum of education,' and that is where violent grammatical innovations are least expected. We, of course, exclude those cases in which the genitive is clearly ablatival or adnominal. Moulton (Proleg., p. 216) begins by stating that the infinitive with Toû 'started as a pure adnominal genitive, and still remains such in many places, as I Cor. 16. 4, ǎžioν тоû πореúεσlat.' But if this is the starting-point we have to explain somehow the process by which such a genitive as that with ăţios came to be the purely adverbial genitive of Thucydides, the Attic orators, and Hellenistic Greek, and we shall find ourselves back among the 'elliptical

1 Moulton, Proleg., p. 217.

? Robertson, p. 1066 sqq.

[ocr errors]

grammarians. What needs to be emphasized in each case, it seems to me, is the distinction between the resultant meaning, which depends upon the context, and the inherent meaning, which is always there and bears no relation to the context. In Thucydides, who first made extensive use of the construction, the resultant meaning is generally that of purpose, though that is by no means always prominent. Where it is not, the grammarians call it a 'loose infinitive of design.' The same applies to Demosthenes and Xenophon (cf. Jannaris, Hist. Gr. Gr., p. 578). To take some N.T. examples-Acts 23. 20, oi 'Iovdaîo συνέθεντο τοῦ ἐρωτῆσαί σε, is the Jews have made a compact in regard to asking thee.' This is called 'the loose infinitive of design' (Robertson, p. 1066). Substitute Táρeiσi for ovvélevтo, and purpose becomes quite clear; but that has to do with the context and the total resultant meaning. The adverbial genitive of the gerundive in Latin is on precisely the same footing. In Germanicus Aegyptum proficiscitur cognoscendae antiquitatis purpose is evident; but in concordiam ordinum, quam dissoluendae tribuniciae potestatis rentur esse (Livy V. 3. 5) the idea is certainly not one of purpose. In one group of examples Robertson calls this curious construction' the subject infinitive. He quotes especially LXX., 2 Chron. 6. 7, ἐγένετο ἐπὶ καρδίαν Δαυεὶδ . . . τοῦ οἰκοδομῆσαι οἶκον. Now τοῦ οἰκοδομῆσαι οἶκον may be the logical subject, but surely grammatically the verb is impersonal, and the sentence is no stranger than the English, 'it occurred to me about going.' Luke 17. I is the same, ἀνενδεκτόν ἐστιν τοῦ τὰ σκάνδαλα μὴ ἐλθεῖν, which is, 6 with regard to pitfalls not coming, it is impossible.' Others Robertson calls the object infinitive; here again the principle is the same. Matt. 21. 32 well illustrates the position, οὐδὲ μετεμελήθητε ὕστερον τοῦ πιστεῦσαι αὐτῷ. Here Moulton denies purpose to the genitive: it 'gives rather the content than the purpose of μereμeλńOnte.' μετεμελήθητε. Robertson calls it the object' of the verb. Translate, 'you did not change your minds even afterwards with reference to (i.e. in the sphere of) belief in him.' The genitive in 2 Cor. 1. 8, σте ¿§аπорη¤ñνaι ημâs кaì toû Çîv, is called by Robertson ablatival; but that case is not natural to aπopeîv (è§intensive). Cf. Plato, Phaedo 84 C, ei dé tɩ πeρì тоÚтшν àπореîтov, with πeρì reinforcing the genitive. It is again 'reference': 'even with regard to life.' In one N.T. example, at any rate, we have a like use of eveκev for closer definition : 2 Cor. 7. 12, ἕνεκεν τοῦ φανερωθῆναι. Compare μνήμης and μνήμης ἕνεκα. Rom. 8. 12, ὀφειλέται ἐσμέν τοῦ κατὰ σάρκα ζῆν shows a use of the genitive which is not very far from that with verbs of judicial action. We suggest then that, while in each separate case the meaning of the genitive is conditioned by its surroundings, its inherent force is simply that of reference. The context applies the appropriate colour, and it is not always a purposecolour. Indeed, Moulton states of Paul that when he wishes to express purpose he uses other means.' And naturally, since Toû with infinitive did not in itself express purpose. We cite from the inscriptions-τοῦ καὶ τοὺς ἄλλους . . . πειρᾶσ[θαί . ἀ]γαθοῦ τινος παραιτίους ἔσεσθαι (C. and B., No. 497), Α. κατε[σκεύασεν τὸ μνημεῖον] τοῦτο . . . τοῦ ἔχειν ἡμᾶς μόν[ους ἐξουσίαν

[ocr errors]
[ocr errors]
« السابقةمتابعة »