The External Evidence. The essential facts of the evidence are included in the following summary: For Enviorás: [? *], B, D2, E, H, L, P, almost all uncials, all cursives except one (including 13, 61, etc.), [Pst. ?], Eus. [? Chrys.]. For λnvas, A, D*, cser. [= Hort's 112], [? Chrys.]. In explanation of this summary we need to remark : — (1) Cis here defective; but in no other case in Acts does it desert the mass of documents when they read either Έλληνες or ἑλληνισταί. (2) It is exceedingly doubtful whether should be cited for ἑλληνιστάς. It actually reads εὐαγγελιστάς, which is usually assumed to presuppose λŋoτás, on account of its like termination. But since it seems certain that evayyediσrás was suggested by, and results from, the proximity of evayyedisóuevo, the inference does not seem secure. No doubt ἑλληνιστάς could be more readily than Ἕλληνας mistaken for evayyediσrás; but if any substantive were derived from εὐαγγελιζόμενοι, it could not fail to take the form εὐαγγελιστάς. It is only with grave doubt, therefore, that the weight of * can be thrown in favor of ἑλληνιστάς. (3) The force of A, as a witness for Aλnvas, is somewhat weakened by the fact that this MS. reads Aŋvas also at ix. 29, where the true reading is undoubtedly Anorrás. D is defective at ix. 29; but, as ἑλληνιστάς. Mr. Purves notes, both A and D insert kai before ¿Aývov in xvii. 4— as do also the good cursives, 13 and 61. If this be due, as he suggests, to a tendency in A and D to put forward the Gentile work of the Church, the testimony of these MSS. here to Anvas should be somewhat suspected. The existence of such a tendency in A and D needs, however, justification. (4) The versions fail to distinguish between the terms and Aŋvas, and hence are not valid witnesses in this matter. Only the Peshitto may be an exception, inasmuch as it reads, at ix. 29, "those Jews who knew Greek"; but even it reads "Greeks" at vi. 1. ed. 3, 1883 [defends "Envac]; HAMMOND, C. E., in his Outlines of Textual Crit., etc., ed. 2, 1876, p. 113 [defends "Envac]; HORT, F. J. A., in his Notes on Select Readings, Gr. Test. vol. II., p. 93, 1881 [defends 'E/2qriorác]; PURVES, G. T.," The Reading" Envaç in Acts xi. 20," in The Presbyterian Review, vol. IV., p. 835 sq., 1883 [defends "Emaç against Hort]. See also the elaborate notes in the critical editions; in the commentaries of Alford, Wordsworth, Bloomfield, Plumptre and Howson and Spence, in loc.; and in Farrar's Life of Paul, I. 285, etc. (5) Chrysostom (whose words, ἴσως διὰ τὸ μὴ εἰδέναι ἑβραϊστὶ ἔλληvas avroùs ¿kádov, both Theophylact and Ecumenius repeat) reads "Greeks" in his commentary clearly, although Anvorús stands in the text commented on. This throws his testimony somewhat in doubt. It may be that the quotation from Acts has been conformed by later copyists to the Syrian type of text (which undoubtedly read ŋνιστάς); or it may be that Chrysostom understood ἑλληνιστάς 25 equivalent to Myras, either in the general import of the word or in this context, and hence, though reading the former, could cry out, ὅρα, ἔλλησιν εὐαγγελίζονται. The weight of his evidence for Έλληνας is weakened in proportion to the probability of his being able to thus interpret ἑλληνιστάς. The evidence being thus before us, its estimation is not without its difficulties, although the issue can scarcely remain doubtful. The Genealogical Evidence.—The application of genealogical considerations leads immediately to the conclusions that both readings are pre-Syrian, and that neither is Alexandrian in its origin, as, indeed, the presence of B in the one group and of D* in the other sufficiently evinces. Beyond that, progress is more difficult. It is certainly striking that, with the exception of D*, as is not supported by any of the typical Western documents. It is not easy to suppose, on the one hand, that Mŋras arose as a Western corruption and yet failed to propagate itself in the later Western texts, or, on the other, that On norus was originally Neutral or Neutral-Alexandrian, and thence seeped, by mixture, into all late Western texts. One is almost tempted to suppose the support of Anas due to the accidental conformity of independent obvious conjectural emendation. closer consideration, however, it appears that all the documents which class here with B have Neutral or Neutral-Alexandrian elements; and thus Murrás is readily accounted for as the NeutralAlexandrian reading, and Aras as the Western. On genealogical considerations, therefore, there is a probability that Morás is the more original reading. This probability fails to be decisive only because genealogical evidence only assigns readings to their respective classes, and leaves it to internal evidence to determine the relative purity of the classes; and internal evidence of classes can only determine usual, not invariable, relations. Although, therefore, it is certain that the Neutral-Alexandrian readings are generally better than the Western, the rule is not absolutely without exceptions, and there is a possibility that the present case may be an exception. Internal Evidence of Groups. We appeal, consequently, to In 1 ternal Evidence of Groups for additional evidence and greater surety. Here we find ourselves embarrassed at once by the doubt resting on the testimony of *. If its witness were clearly for Anvarrás, the known high character of the combination B, here increased greatly by the adjunction of many other important witnesses, would throw the weight of the external evidence overwhelmingly for that reading. Just in the degree that we judge it probable that the present reading of * is only a stupid blunder for Anorrás, must the testimony for that reading appear to us to approach the overwhelming point. Even when we lay aside the testimony of **, however, the internal evidence of groups appears still to support ¿Aŋviσtás,— B being rarely wrong when in conjunction with such a train as here sides with it. Still another mode of procedure is open to us, by which we may reach an independent result, and thus test the probabilities already raised. We may try, by internal evidence of groups, the special value of the group which here appears as the evidence for Anvas. We have noted something over a hundred cases in which the group, A, D* occurs in the Book of Acts. In the great majority of these, however, it has either actually or practically the support of all other MSS. except *; in other words, the rival reading is a mere individualism or slip of the careless scribe of *, which has been corrected into conformity with the universally supported reading by the scholarly hand whom we know as . These cases are only valuable in helping us estimate the value of, to whom hardly due credit is usually attached. The remaining instances may be conveniently classified as follows: (1) Instances in which, A, D* have the support of two or more of the primary documents :1 Tischendorf's viii. ed.; T.vii. Tischendorf's vii. ed.; H. = Westcott and Hort; and A. Alford. (*) BCEL P al. 13, L. T. Tr. H. A. right.] * BE al.10+ vg. Cop. L. T. Tr. H. A. right. (2) Instances in which they are supported by B and secondary au (3) Instances in which they are supported by C and secondary (4) Instances in which they are supported by * and secondary authorities only: 21 [xxi. 22 AVT. SEL. Ovvε2.] * C2 EHI L P al.plu. | L. T. A. [xxi. 22 wrong.] (5) Instances in which they are supported by secondary authorities only: vuov after 0805 | 61, al. vg. Ir. Or. Chrys.| L. T.vii. Tr. A. wrong. This last list, of course, furnishes the truest parallels to our present passage, and it must be confessed that the most of them are clearly wrong, while none of them are clearly right, and (besides xi. 20) only two seem capable of being plausibly defended. The case is little better with the other instances which lack the support of B; out of nine cases, only three apparently can be plausibly defended, and these are all of such character that internal evidence is of somewhat doubtful value in regard to them. The result of this investigation also, thus, is to discredit Anvas. Three, or perhaps four, independent methods of examining the evidence thus elicits from the external testimony a consentient witness for the probable originality of EλAŋviorás. The exact force of this cumulative probability is not easy to estimate. It is certainly strong enough to give us full confidence in the correctness of λAŋviorrás, in the absence of strong rebutting considerations drawn from internal evidence. And in the presence of such rebutting considerations, it is strong enough to demand from us very anxious questionings and very strenuous efforts after harmony before we set it aside. The Internal Evidence. Transcriptional Probability. Ἕλληνας, That the transcriptional probability goes with the external in favor of ¿Mŋuorás is scarcely open to doubt. Any ordinary reader would naturally expect λλŋyas here; and, therefore, a scribe, finding it here, would be very unlikely to alter it into the difficult reading and rare word, ¿AAŋviorás. This is not to assume in scribes a nice appreciation of the true course of the history, but only a slight attention to the immediate context in its most obvious appearances. The contrast with 'Iovdaíovs that would inevitably suggest itself to the mind of any copyist would be the standing one, which he would almost venture to write without reference to his copy; only if he had just written Ἑβραίους, would he think of ἑλληνιστάς a5 its contrast. The strengthening kai before the pós would render it all the more inevitable that he should expect to find, and hence should write, Anvas. The general progress of the narrative from v. 19 points in the same direction. All combined renders ¿quorás so difficult a reading as to forbid our supposing that any scribe would (consciously or unconsciously) write it here for Anvas, - points out gras as so obvious a correction as to make it very probable that scribes might even independently (consciously or unconsciously) write it here for ἑλληνιστάς. On the assumption that Anvas is the original reading, explanations |