(3) We may explain (C), with Burkitt, as derived from (B). The Syriac translator was not satisfied with the loose construction of (B). Taking ᾧ to refer to ἐγέννησεν as well as to μνηστευθεῖσα, he made the connexion clearer by inserting a second Ἰωσήφ as the subject of the verb. In taking this last step, he either returned unconsciously to part at least of the true reading, or had access to good Greek MSS. which we no longer possess.
It is of interest to compare the reading we have suggested as the original text of Mt. i. 16 with others which have been put forward. In discussing one of these possibilities, Sanday writes (Outlines, p. 200): “If we may suppose that the original text ran Ἰωσὴφ τὸν ἄνδρα Μαρίας ἣ ἐγέννησεν Ἰησοῦν τὸν λεγόμενον Χριστόν, that would perhaps account for the two divergent lines of variants better than any other”. In spite of its advantages, this text suffers under two disadvantages from which the one we have preferred is free. (i) Not only is γεννάω used in a different sense from that which it has in the rest of the Genealogy, but it is the very same form of the verb which is employed differently. (ii) The reading is too smooth and clear. Apart from the phrase τὸν ἄνδρα Μ. no loophole is left for misunderstanding, and so no sufficient starting-point is provided for the subsequent textual variants.
Burkitt has instanced the reading we have preferred. In rejecting the view that the Syr.-Sin. represents the true text, he writes (p. 264): “Had we such a text as Ἰακ. δὲ ἐγένν. τὸν Ἰωσὴφ τὸν ἄνδρα Μαρίας· Ἰωσὴφ δὲ ἐγέννησεν κτλ. the case would have [pg 113] been different”. In reference to this suggestion, however, Burkitt says, (i) the evidence does not point that way, (ii) in that case the Syr.-Sin. would be further from the original than that of א B and Tertullian, (iii) Syr.-Sin. and k would “agree in a common corruption”, and we should have to speak of the “Western” text in the singular number.
The last point raises a large question which it is impossible to consider here. As regards the second objection, while in some respects (C) would be further from the original than (A), in other and more important respects it would be appreciably nearer. In its use of ἐγέννησεν it would be nearer to the original than any reading we possess. As regards the first objection, we have frankly to agree that the textual evidence does not point that way. We cannot point to a shred of MS. evidence to support the conjectured reading. A generation ago this would have been considered a fatal objection. But, in view of the freedom with which the text of the Gospels was handled during the first half of the second century, and which the textual variants illustrate, this objection can no longer be regarded as insuperable. So long as we restrict ourselves to the attested readings, the problem remains insoluble. If, then, we can reach a reasonable conclusion on other lines, we are free to do so. Doubtless, in default of attestation, we can describe our results as no more than tentative. But we have no desire to claim more. As the problem stands at present, the test to be applied is, What reading, conjectured or attested, furnishes the best explanation of the facts at our disposal?, it being remembered that these facts include, not only the textual variants, but also the unique character of the Genealogy itself. It may be, as we have suggested, that new discoveries await us. But, unless we have entirely misread the evidence we already possess, no discovery is to be expected which will completely transform the textual problem.
In conclusion, we may state certain propositions (apart from the question of the exact wording of the true text of Mt. i. 16) which have in their favour a high degree of probability.
(1) The readings which we have called (A) and (B) are independent attempts to alter the original text in the interests of the Virgin Birth; that is, they are “doctrinal modifications”.
(2) The reading of the Sinaitic Syriac is not unfavourable to the [pg 114] doctrine. It should no longer be spoken of as “the eccentric reading”, nor should we describe the translator as influenced by “heretical tendencies”.
(3) The original text of Mt. i. 16 implied the Virgin Birth, but it was stated from the unique point of view reflected in the Genealogy itself.
(4) The text was liable to misunderstanding, and the history of the textual variants is the history of that misunderstanding.