of Jesus and of related events is contained in Luke 1:5–2 (with Luke
have had abundant opportunity to communicate the story to the early disciples, either directly or through the company of women described in the latter course of the Gospels. In the case of Matthew's account, Joseph seems rather to be indicated as the channel of communication—at any rate he could have been such a channel. /pg. 642/ So if the facts are real, the explanation of the rise of the narratives is, in general, if not in detail, an easy task.
Therefore, we may examine, first, the hypothesis that the narrative is to be regarded as a copy of the facts, reserving the alternative hypothesis for subsequent discussion. Is the narrative near enough to the facts to be a copy of them, and if so, can the facts themselves be reasonably explained? If the facts are extremely unlikely, then only enormous difficulty in explaining the narrative without reference to the facts could force us to this explanation of the narrative through the facts.
1. The external attestation. The New Testament account of the birth of Jesus and of related events is contained in Luke 1:5–2 (with Luke 3:23–38) and in Matt. 1, 2. This account is therefore contained in two of the New Testament books, whose attestation is so strong as to make it practically impossible that they were written after the close of the first century, and exceedingly probable that they were written very much earlier. Nor is there any external evidence really worth considering to show that these Gospels did not originally contain the accounts of the birth. These accounts appear in all the Greek manuscripts, in all the ancient versions and in the Diatessaron of Tatian (omitting the genealogies). It is true that Cerinthus and Carpocrates and a class of Jewish Christians did not believe in the virgin birth, and did not accept those portions of the Gospels which supported that doctrine; but it is pretty evident that their action was motived by dogmatic rather than historical considerations. Even if it is held that heresy in the early Church was, in most cases, a tenacious holding to the ancient simplicity in the face of the developing theology of the Church, yet this does not affect the narrower textual question now under discussion. It may be perfectly true, for example,
that a certain class of Ebionites were not mistaken in regarding the natural birth of Christ as the correct original belief; yet it is evident that their omission of the opening chapters of Matthew and Luke was not textually justified. Perhaps the Ebionites were right in refusing to assert that the virgin birth was fact; in any case, there is no good reason to suppose that they were right in omitting the account of that supposed fact from their copies of the first and third Gospels.1 Marcion's rejection of the first two /pg. 643/ chapters of Luke shares in the low estimate which is to be attached to his other numerous alterations of the text of the New Testament books.2 As Harnack says, Marcion felt himself to be a reformer, and so the principle3 that heretics become heretical only because they faithfully maintain conditions beyond which the main body of the Church has since the separation advanced, is certainly, in his case at least, not to be applied.4
One other supposed testimony to an original form of Luke's Gospel which did not contain the first two chapters must be mentioned for the sake of completeness. In 1902, Conybeare5 called attention to the fragments added to the two manuscripts (both from the year 1195) of the Armenian translation of Ephraem's Commentary on the Diatessaron. These manuscripts, which, Conybeare believes, represent widely separated texts, both add to the Commentary various fragments, which are attributed to Ephraem. One of them—a brief account of the manner of writing of the Gospels—contains a notice about Luke, which Conybeare translates as follows: Lucas autem initium fecit a baptismo Ioannis, sicut primum de carnalitate eius locutus est et de regno quod a Davide, et deinde quidem a Abrahamo, incepit. This notice, Conybeare supposes, was found by Ephraem at the end of the Diatessaron, and, since it follows the more ancient tradition in various particulars, is very old. The text and the interpretation of the latter part of the notice about Luke are exceedingly uncertain, and this might seem to suggest the notion that the text is corrupt in the first clause; but Conybeare insists that the reading "baptism" could never have arisen if the reading "birth" had been original. With regard to this point we should certainly not
be too positive, but it does not seem altogether impossible that a scribe /pg. 644/ might have been confused by the notice about Mark which immediately precedes, and thus might have been led to change the unusual phrase "birth of John" to the more usual one "baptism of John." It must be borne in mind that Ephraem's copy of the Diatessaron, without the slightest doubt, contained the first two chapters of Luke, so that if Ephraem appended the note in question to his Commentary, or left it as he found it at the close of the Diatessaron, he must have done so without observing its real meaning. It seems more probable to suppose that the corruption of the text of the notice extends further than Conybeare thinks; but if not, it is possible that the note was written by one of those heretics who, as we have already observed, did not accept the first two chapters of Luke. In any case, it cannot be said that this notice, existing only in manuscripts of the year 1195 and there attached to a work of the fourth century, carries us back to the fact of an addition to the third Gospel, which, if made at all, was made early in the second century; especially since we can point to circles where such an idea about the Gospel arose at a later time from dogmatic considerations, and whence the notice in question might have come. We conclude, then, that there is no external evidence of any account to show that the Gospel of Luke ever existed without the first two chapters.
But our proof of the early date of the accounts of the birth is not indirect and negative merely. We are not forced to rely solely on the argument that the chapters under discussion are firmly fixed in the first and third Gospels, that these Gospels have early attestation, and that therefore the chapters are early. On the contrary, there is the strongest kind of evidence for the early use, not only of the first and third Gospels in general, but of those very parts of the Gospels which contain an account of the birth.
For the virgin birth—the most remarkable thing narrated in the chapters under discussion—was part of the firmly fixed Christian belief at a very early time. In the first place, it formed part of the
original "Apostles' Creed" (though expressed in slightly different words from those we use to-day), which arose, according to Harnack, about 150 A.D., according to Zahn, certainly not later than 120. And even aside from the question as to the origin of the Creed as a whole, more or less fixed and creed-like statements of the virgin birth— statements pointing to what Harnack calls "an Eastern Christological :V20:""—can be detected in early writers.6 /pg. 645/
It is beyond dispute that Irenæus gave to the virgin birth a place in the rule of faith, at least in so far as he had any definite rule of faith at all. As to Justin Martyr, Hillmann7 has raised objections, not, indeed, to the fact of Justin's testimony, but to the manner of it. He says that Justin is evidently a pioneer in the support of the virgin birth, because he regards as Christians (•BÎ J@L$ º:,JXD@L (X<@LH) those who deny the doctrine (Dial., 48). But how else would you expect him to speak of those who accepted Christ as the Messiah, though holding a peculiar view of the manifestation of His Sonship? In other words, how else could he express the idea of "heretic" as opposed to "unbeliever"? And to hold that Justin regarded the virgin birth as something uncertain or unimportant is to run counter to the large number of passages (both in the Dialogue and in the Apology) where it is mentioned as one of the fundamental facts about Christ.
That Aristides believed in the virgin birth is attested by the Syriac and Armenian versions as well as by what remains of the original Greek,8 and it is probable that the phrase "born of the Virgin Mary" found a place in his creed.9
In regard to Ignatius, it would seem that the two passages, Eph. xix. 1, 6"Â §8"2,< JÎ< –DP@<J" J@L$ "ÆT$<@H J@L$J@L º B"D2, <\" 9"D\"H 6"Â Ò J@P,JÎH "ÛJ0$H Ò:@\TH 6"Â Ò 2V<"J@H J@L$ 6LD\@LA JD\" :LFJZD4" 6D"L(0$H, "J4<" ¦< ºFLP\" 2,@L$ ¦BDVP20, and Smyrn. i. 1, •802T$H Ð<J" ¦6 (X<@LH )"$Â* 6"J( FVD6", LÊÎ< 2,@L$ 6"J( 2X80:" 6"Â *b<":4< 2,@L$ (,(,<0:X<@< •802T$H ¦6 B"D2X<@L, were sufficient. Hillmann, however, by a
process of reasoning, arrives at the conclusion that the author did not know Luke 1:34, 35, 3:23. The author, he says, in Eph. xviii. 2, xx. 2, and Smyrn. i. 1, regards Jesus as begotten (1) ¦6 FBXD:"J@H :¥< )"$\*, (2) B<,b:"J@H *¥ ,(\@L. This can be explained only on the Adoptionist view, for the generation from the seed of David cannot be regarded as coming through Mary, since in the first passage it is parallel with B<,b:"J@H *¥ ,(\@L, and since in Trall. ix. 1 ¦6 (X<@LH )"$Â* is regarded as distinct from ¦6 9"D\"H. Now, even if we admit that Ignatius regarded Mary as not of the tribe of Judah (which does not seem to me to be clearly proved by the passages cited above), it does not follow that because he then derived Jesus' Davidic descent through Joseph, he did not know Luke 1:34, 35, 3:23. For if those passages stood where they stand to-day, the very same supposed contradiction was present in the first part of the third Gospel as is present in /pg. 646/ Ignatius. Ignatius simply took over the two sides of the account in Luke without reflection. That this view of the matter is correct is made perfectly evident by the fact that Ignatius in the two passages quoted above distinctly states the virginity of Mary—a fact which nullifies the inferences of Hillmann. Without sufficient reason, Hillmann regards the phrase (,(,<0:X<@< ¦6 B"D2X<@L (Smyrn. i. 1) as an interpolation; Eph. xix. 1 (which Swete calls the classical passage) he does not mention at all. Swete calls attention to the fact that the testimony of Ignatius is made more valuable by the nature of his argument. He is arguing with Docetics, and is urging against them the reality of the birth of Jesus. It would, therefore have suited his purpose to point to the natural birth; but instead of this he says in effect that, though of course supernatural, the birth was yet real. So there is nothing against the statement of Harnack that "Ignatius has freely reproduced a 'kerugma' of Christ which seems, in essentials, to be of a fairly definite historical character, and which contained, inter alia, the Virgin Birth, Pontius Pilate, and the •BX2"<,<."10
We have thus traced a firm and well-formulated belief in the virgin birth back to the beginning of the second century. The question at once arises whether the accounts of Matthew and Luke were the
sources of that belief. Some kind of an argument might be derived from the manner of statement of the doctrine in the early patristic writers, but this would not be absolutely convincing, for example, in the case of Ignatius. However, the decision is made very probable by the following considerations. It is just this virgin birth which is most urged as necessitating a late date for Luke 1, 2; Matt. 1, 2, or certain portions of those chapters—indeed, if it were not for the virgin birth, probably those chapters would, in view of the great weight of manuscript attestation, have passed unchallenged as original parts of the Gospels. But it is just this virgin birth which we have shown to have been accepted as a fundamental fact early as the days of the Apostolic Fathers. At the beginning of the second century, then, the first and third Gospels were used, and the virgin birth was accepted. According to a great weight of manuscript evidence, the virgin birth found a place in those Gospels. The conclusion is at least a natural one that the Christians of that time derived their belief in the virgin birth from the account of that birth which is so firmly fixed in the Gospels, or at any rate that they derived the belief partly from those Gospels. If, as seems to be /pg. 647/ possible, for example in Justin, an extra-canonical source was also used, any argument for the trustworthiness of our canonical accounts is rather increased than otherwise, since another testimony is added to the two that we already possess. If the extra-canonical source was itself the source of our two accounts, then by it we are carried still further back. Our accounts are demonstrably old; if a still older account containing the virgin birth was used along with them at the beginning of the second century, then we have worked back very near to the time of the supposed facts. If the early writers enumerated above used only some account different from our account, then it is still significant that just that element in our accounts which has met with most objections was a firmly fixed part of the Christian belief at the beginning of the second century. But there is practically conclusive evidence that these early writers did know our accounts, and this fact, coupled with the evidence of the manuscripts and versions, leads to the conclusion that Matt. 1, 2, and Luke 1, 2, were parts of the original Gospels, and were therefore written in all probability