We are left, then, with three readings, for each of which priority may be claimed (those we have indicated by (A) and (B), and that of the Syr.-Sin. (C)). It is highly probable that (C) is derived from (B); but it may be well to leave this an open question, so as to have all the possibilities before us.
(1) Can we, then, explain the textual facts already noticed, if we presume the originality of (A)?
It is certainly remarkable that, after using ἐγέννησεν in a legal sense throughout the earlier links of the Genealogy (Moffatt, Burkitt, Westcott, Box, Allen, Barnard, A. J. Maclean), the compiler should desert this practice, and use the verb of physical parentage (ἐγεννήθη) in the last link of the chain. The compiler, if we may say so, does not strike us as the kind of man who would have felt the need of this. It seems much more likely that, together with some qualifying clause in reference to Mary, he would have continued to employ ἐγέννησεν in the same sense to the end. This is conjecture; but (on the present theory) it is a conjecture supported by the procedure of the scribes who have produced (B). Their object (on the present supposition) will have been to remove the ambiguities of (A) in Mt. i. 16, so as to state the doctrine more clearly. We could understand, then, their objection to τὸν ἄνδρα Μαρίας, and the change to ᾧ μνηστευθεῖσα π. Μ. What is less easy to understand is the change from ἐγεννήθη to ἐγέννησεν. It is true that ἐξ ἧς ἐγεννήθη is not without ambiguity, as the comment of the Jew in the Dialogue of Timothy and Aquila shows.[102] But, if this was a ground of objection, why should the ambiguity be replaced by one that is much greater? As we have seen, the construction of (B) is singularly loose. It is this fact which has clearly invited the modifications represented in the Syr.-Cur. and the Old Lat. MSS., and [pg 109] perhaps the Syr.-Sin. itself. The reading (B) certainly does not commend itself as a doctrinal modification of (A). Further, the priority of (A) does not help us to account for (C). If, as we believe, (C) is derived from (B), it is needless to discuss the point. But apart from that theory of the origin of (C), our conclusion remains the same. We have seen how near in spirit the scribe of the Syr.-Sin. was to the First Evangelist. Can we suppose, then, that he would have demurred to the words, τὸν ἄνδρα Μαρίας? It is very difficult to think so. For these reasons, in spite of its strong attestation, we find it impossible to presume the originality of (A).
(2) We reach a similar conclusion, if we assume (B) to be the true text of Mt. i. 16. Its singular construction does not readily suggest the craftsmanship of the compiler of the Genealogy. It is true that we can give a very good account of (C) on the present assumption. We can adopt Burkitt's suggestion, and regard it as a paraphrase of (B). But can we derive (A) from (B)? It would be reasonable to explain ἐξ ἧς ἐγεννήθη as a correction of ἐγέννησεν by a believer who failed to understand the Evangelist's point of view, and who desired a clearer reference to the Virgin Birth. But can we imagine a scribe, or an editor, motived in this way, replacing “to whom was betrothed the Virgin Mary” by the words “the husband of Mary”? The question answers itself, and forbids the assumption of the priority of (B).
(3) Can we, then, accept Archdeacon Willoughby C. Allen's view, and find the true text in (C)?[103] It is quite possible, on this theory, to give a reasonable explanation of (B), but, as in the last case, the difficulty is to account for (A). We can follow the change from ἐγέννησεν to ἐξ ἧς ἐγεννήθη, but the substitution of τὸν ἄνδρα Μαρίας for the parenthesis found in (C) remains as before an insuperable objection. At the same time Archdeacon Allen has laid down a true and a valuable principle when he writes: “The earliest Greek form was gradually altered from a desire to avoid words which, though in the intention of the [pg 110] writer they expressed legal parentage, not paternity, in fact, might be misunderstood by thoughtless readers” (p. 8).
Our results thus far are negative, but they are not barren. We have frankly to admit that no extant reading, as a whole, commends itself as the original text of Mt. i. 16. On the other hand, we can form a reasonably good idea of what that text was like. If we are to make any further advance, we must have recourse to conjecture. It is not at all impossible that future discoveries may enable us to travel upon firmer ground. Such a discovery as that of the Syr.-Sin. MS. by Mrs. Lewis and Mrs. Gibson in 1902 shows that this hope is not unreasonable. But meantime, unless we are content to acquiesce in a negative conclusion, we have no choice but to resort to conjecture. This does not mean a leap in the dark. It is in every way likely that parts of the true text are embedded in the extant readings, and it is by no means impossible that, taken together, these readings may contain the whole. It may be, that is to say, that the true text of Mt. i. 16 has found its grave in the readings we possess. Whether its resurrection can be accomplished is another question. But, in view of the general character of the true text, as indicated above, the attempt need not be foreclosed. Obviously, our results will be tentative, but they should be something more than dubious and uncertain in the extreme.
IV.
In attempting to reconstruct the true text of Mt. i. 16, we may venture the following suggestions:
(1) We have very good ground for regarding τὸν. λ. Χ. as part of the true text (though whether we read the nom. or the acc. depends upon whether we prefer ἐγεννήθη or ἐγέννησεν). Not only does this expression occur both in (A) and (B), but it is also one which we should naturally expect the Genealogy to contain. A Genealogy constructed to show the Messiahship of Jesus ends fittingly with the words “who is called Christ”.
(2) It is very probable indeed that the original text included ἐγέννησεν and not ἐγεννήθη. (i) On this view, we can readily understand the misconceptions that would arise, and give a reasonable explanation of the textual variants which exist. (ii) As [pg 111] indicating legal parentage, the expression is not one from which we think the compiler would be likely to shrink. (iii) It is not easy to suppose that those who have employed ἐγέννησεν in the reading (B) would have used this form if they had not found it already in the text.