J. Gresham Machen Bibliography

THE INTEGRITY OF THE LUCAN NARRATIVE OF THE ANNUNCIATION ¹

The Lucan narrative of the birth and infancy in Lk. i. 5ii. 52 is strikingly Jewish and Palestinian both in form and in content. ² That narrative contains an attestation of the virgin birth of Christ. But according to the prevailing view among those who deny the historicity of the virgin birth, the idea of the virgin birth was derived from pagan sources. If so, the question becomes acute how such a pagan idea could have found a place just in the most strikingly Jewish and Palestinian narrative in the whole New Testament. This question has been answered by many modern scholars by a theory of interpolation. It is perfectly true, they say, that Lk. i. 5-ii. 52 is of Palestinian origin; and it is perfectly true that an attestation of the virgin birth now stands in that narrative; but, they say, that attestation of the virgin birth formed no original part of the narrative, but came into it by interpolation. It would be difficult to exaggerate the importance of this question; indeed we may fairly say that if the interpolation theory is incorrect the most prominent modern reconstruction proposed in opposition to the historicity of the virgin birth falls to the ground. The view as to the origin of the idea of the virgin birth which has been most widely held by those modern historians who deny the fact of the virgin birth stands or falls with the interpolation theory.

The interpolation theory ³ has been held in various forms. A classification of these various forms is possible from two points of view. The first point of view concerns the sense in which the supposed interpolation is to be called an interpolation. A three-fold division is here possible. In the first place, the interpolation may be regarded as an interpolation into the completed Gospel—a gloss introduced into the Third Gospel at some point in the manuscript transmission. In the second place, the interpolation may be regarded as an interpolation made by the author of the Gospel himself into a Jewish Christian source which elsewhere he is following closely. In this case the words attesting the virgin birth would be an original part of the Gospel, but would not belong to the underlying Jewish Christian narrative. In the third place, the interpolation may be regarded as an interpolation made by the author himself, not into a source but into the completed Gospel—that is, the author first finished the Gospel without including the virgin birth, and then inserted the virgin birth as an afterthought. This third possibility has been suggested —for the first time so far as we know—by Vincent Taylor, the author of the latest important monograph on the subject. ⁴ The second point of view from which a classification is possible concerns the extent of the supposed interpolation. Whether the interpolation is to be regarded as an interpolation into the completed Gospel by a scribe, or into the source by the author of the Gospel, or into the completed Gospel by the author of the Gospel, how much is to be regarded as interpolated? With regard to this latter question, there have been various opinions. The earliest and probably still the commonest view is that the interpolation embraces verses 34 and 35 of the first chapter. That view received its first systematic

grounding from Hillmann in 1891. ⁵ It has since then been advocated by Usener, Harnack, Zimmermann, Schmiedel, Pfleiderer, Conybeare and others. A second view was suggested by Kattenbusch ⁶ and defended by Weinel. ⁷ It is to the effect that only the words, “seeing I know not a man” ⁸ in Lk. i. 34, 35, are to be eliminated. A third view includes verses 36 and 37 with verses 34 and 35 in the supposed interpolation. ⁹ With regard to the former classification—that is, the classification according to the sense in which the supposed interpolation is to be taken as an interpolation—it may be noticed at the start that the first view, which regards the interpolation as an interpolation made by a scribe into the completed Gospel, is opposed by the weight of manuscript attestation. There is really no external evidence worthy the name for the view that Lk. i. 34, 35 or any part of it is an interpolation. Manuscript b of the Old Latin Version, it is true, does substitute verse 38 for verse 34, and then omits verse 38 from its proper place. But that may either have been a mere blunder in transmission, especially since the two verses begin with the same words, “And Mary said” ¹⁰; or else may be due to the desire of a scribe to save Mary from the appearance of unbelief which might be produced by her question in verse 34. ¹¹ At any rate the reading of this manuscript is entirely isolated; as it stands, it produces nonsense, since it represents the angel

as continuing to speak (verses 35-37) after he has already departed; and certainly it cannot lay the slightest claim either to be itself, or to enable us to reconstruct, the true text. As for the testimony of John of Damascus in the eighth century to the omission of the phrase, “seeing I know not a man,” in some Greek codices, that is clearly too late to be of importance. ¹² Thus the unanimity of manuscript evidence for the inclusion of Lk. i. 34, 35 is practically unbroken. And it is difficult to see how such unanimity could have arisen if the verses were interpolated in the course of the transmission. In view of the many widely divergent lines of transmission in which the text of the Gospel has come down to us, it would be surprising in the extreme if the true reading should in this passage have nowhere left even the slightest trace. This argument, of course, applies only to that form of the interpolation hypothesis which regards the supposed insertion as having been made into the completed Gospel. It does not apply to the view that the author of the Gospel himself made the insertion into the narrative derived from his source or into the Gospel which he had already written but had not published. But possibly these forms of the hypothesis may be found to be faced by special difficulties of their own. At any rate, what we shall now do is to examine these three forms of the interpolation hypothesis so far as possible together—noting, of course, as we go along, the cases where any particular argument applies only to one or to two of the three forms rather than to all. In other words, we shall examine the question whether or not Lk. i. 34, 35 is an original part of its present context or else has been inserted into that context either by the author of the Gospel into a source or by the author of the Gospel into his own completed work or by some scribe. The first consideration which we may notice as having been adduced in favor of the interpolation theory is of a

general character. The rest of the narrative, it is said, outside of Lk. i. 34, 35 is perfectly compatible with a birth of Jesus simply as the son of Joseph and Mary, indeed it is even contradictory to the notion of a virgin birth; if, therefore, we accomplish the simple deletion of these two verses, all inconsistence is removed and the story becomes perfectly smooth and easy. With regard to this argument, it should be noticed, in the first place, that the simple deletion of Lk. i. 34, 35 will not remove the virgin birth from the Third Gospel in general, or from the infancy narrative in particular; for the virgin birth is clearly implied in several other places. The first of these places is found at Lk. i. 26 f., where it is said: “And in the sixth month the angel Gabriel was sent from God unto a city of Galilee whose name was Nazareth, to a virgin betrothed to a man whose name was Joseph, of the house of David, and the name of the virgin was Mary.” Here Mary is twice called a virgin, and in what follows nothing whatever is said about her marriage to Joseph. This phenomenon is perfectly natural if the virgin birth was in the mind of the narrator, but it is very unnatural if the reverse is the case. Advocates of the interpolation theory are therefore compelled to offer some explanation of the language in Lk. i. 27. Two explanations are open to them. In the first place, it may be said that verse 27 has been tampered with by the same interpolator who inserted verses 34, 35, and that originally Mary was not here called a virgin. But against this explanation may be urged the fact that the word “virgin” occurs twice in the verse, and that if that word was not originally there the whole structure of the verse must have been different. The second possible explanation is that although the form of verse 27 which we now have is the original form—that is, although Mary was really designated there as a virgin—yet the mention of her marriage to Joseph has been omitted, by the interpolator of Lk. i. 34, 35, from the subsequent narrative. But it may be doubted whether this explanation quite

accomplishes the purpose for which it is proposed. Even if the writer of Lk. i. 27 were intending to introduce later on a mention of Mary’s marriage to Joseph, his designation of her as a virgin would seem to be unnatural. In the Old Testament narratives of heavenly annunciations, the annunciations are represented as being made to married women; and if the narrator of Lk. i, ii intended the promised son to be regarded as having a human father as well as a human mother, as in those Old Testament narratives, why did he not, as is done there, represent the annunciation as being made to a married woman? Why does he insist so particularly, by a repetition of the word, that it was made to Mary when she was a “virgin”? It must be remembered that according to all or nearly all of the advocates of the interpolation theory, the narrative is quite unhistorical; so that the narrator, according to their view, was not hampered by any historical consideration from placing the annunciation either before or after the marriage, exactly as he pleased. Why then does he insist so particularly that it took place before the marriage, or while Mary was still a “virgin,” instead of representing it as taking place after the marriage? Surely this latter representation would have been far more natural, as well as more in accord with Old Testament analogy, if the narrator really intended the promised son to be regarded as being, in a physical sense, the son of Joseph. A possible answer to this argument of ours might be based upon Lk. ii. 7, where it is said that Jesus was the “firstborn son” of Mary, and upon Lk. ii. 23 where there is recorded compliance in the case of Jesus with the Old Testament provisions about the firstborn. Perhaps, the advocates of the interpolation hypothesis might say, the emphasis in Lk. i. 27 upon the virginity of Mary at the time when the annunciation was made to her, is due only to the desire of the narrator to show that she had not previously had children. But we do not think that this answer is satisfactory. Isaac was the firstborn son of his mother Sarah, in accordance with the Old Testament narrative; and yet the annunciation

of his birth is represented as having come to his mother when she was already married. Similar is the case also with the birth of Samson and of Samuel. Why could not these models have been followed by the narrator of the birth of Jesus? Surely he could have represented Jesus as the firstborn son without placing the annunciation, in so unnatural and unprecedented a way, before instead of after his mother’s marriage. At any rate, whether we are correct or not in regarding this second explanation of Lk. i. 27 as inadequate, it should be noticed that both the two explanations result in an overloading of the interpolation hypothesis. Whether it be held that Lk. i. 27 has been tampered with, or that something has been removed by the interpolator at a later point in the narrative, in either case the activities of the interpolator must be regarded as having extended farther than was at first maintained. What becomes, then, of the initial argument that a simple removal of Lk. i. 34, 35 will suffice to make the narrative all perfectly smooth and easy as a narrative representing Jesus as being in a physical sense the son of Joseph? Moreover, Lk. i. 27 is not the only verse which requires explanation if Lk. i. 34, 35 be removed. What shall be done with Lk. ii. 5, which reads: “to be enrolled with Mary who was betrothed to him being great with child.” How could Mary be said to be only betrothed to Joseph, when she was already great with child? Certainly this form of expression, coming from a narrator who of course intended to record nothing derogatory to the honor of Mary, implies the virgin birth in the clearest possible way. It is true, the matter is complicated in this case, as it was not in the case of Lk. i. 27, by variation in the extant manuscript transmission. The reading “who was betrothed to him” appears, indeed, in the best Greek uncials, including the typical representatives of the “Neutral” type of text, the Codex Vaticanus and the Codex Sinaiticus. It also appears in the Codex Bezae, which is a representative of the “Western” type of text, and in a number of the versions. But

certain manuscripts of the Old Latin Version and the “Sinaitic Syriac” manuscript of the Old Syriac Version read “his wife”; and a number of the later uncials with the mass of the cursive manuscripts, representing what Westcott and Hort called the “Syrian Revision,” read “his betrothed wife.” This last reading is generally rejected as being a “conflate reading”; evidently, it is held, some scribe combined the reading “betrothed” with the reading “wife” to make the reading “betrothed wife.” But what decision shall be reached as between the other two readings? The external evidence certainly seems to favor the reading “betrothed,” which appears in the great early uncials, representative of the “Neutral” type of text, whereas the reading “wife” appears in no Greek manuscript at all but is attested only in Latin and in Syriac. Despite all that has been said in criticism of Westcott and Hort’s high estimate of the Neutral text, recent criticism has not really succeeded in invalidating that estimate. Nevertheless, the combination of important Old Latin manuscripts with the Sinaitic Syriac in favor of the reading “wife” shows that that reading was in existence at a rather early time. It must, therefore, at least be given consideration. ¹³ At first sight, transcriptional probability might seem to be in favor of it. If Mary at this point was in the original text spoken of as Joseph’s “wife,” it is possible to conceive of some scribe, who was eager to protect the virginity of Mary from any possible misunderstanding, as being offended by the word “wife” and so as substituting the word “betrothed” for it. But it is possible also to look at the matter in a different light. If the word “betrothed” is read in this verse, then at least a verbal contradiction arises as over against the Gospel

of Matthew; for without doubt Matthew lays great stress upon the fact that when Jesus was born Mary was in a legal sense not merely betrothed to Joseph but actually his wife. The contradiction need not indeed be anything more than formal; for there is no reason why Luke may not be using a terminology different from that of Matthew, so that by the word “betrothed” he is designating the extraordinary relationship which according to Matthew prevailed after Joseph had obeyed the instructions of the angel—that is, the relationship in which Mary is legally the wife of Joseph but in which he “knew her not until she had borne a son.” ¹⁴ But although the contradiction may not actually be more than formal, it might well have seemed serious to a devout scribe. The change from “betrothed” to “wife” may therefore fall into the category of “harmonistic corruptions.” This hypothesis, we think, is more probable than the alternative hypothesis, that “wife” was changed to “betrothed” for doctrinal reasons. Transcriptional considerations are thus not opposed to the reading of the Neutral text, and that reading should in all probability be regarded as correct. But if the reading “betrothed” at Lk. ii. 5 is correct, then we have another overloading of the interpolation hypothesis with regard to Lk. i. 34, 35: the advocates of that hypothesis must suppose that the interpolator tampered with Lk. ii. 5 as well as with Lk. i. 27 or with a supposed subsequent insertion mentioning the marriage of Mary to Joseph. Obviously the removal of all mention of the virgin birth from Lk. i-ii is by no means so simple a matter as was at first supposed. There is of course still another place in the Third Gospel where the virgin birth is clearly alluded to—namely Lk. iii. 23. The words “as was supposed” in that verse—“being, as was supposed, the son of Joseph”—clearly imply that Jesus was only “supposed” to be the son (in the full sense) of Joseph, and that really his relationship to Joseph was of a different kind. In this case there is no manuscript evidence for the omission of the words; the words appear in all the extant witnesses to the text, the variants (of order and the like) being unimportant for the matter now under discussion. The verse, therefore, constitutes an additional weight upon at least one form of the interpolation theory regarding Lk. i. 34, 35; it constitutes a weight upon the hypothesis that those verses are an interpolation into the completed Gospel. For if Lk. i. 34, 35 is an interpolation, the words “as was supposed” in Lk. iii. 23 must also be an interpolation; and the more numerous such interpolations are thought to be, the more difficult does it become to explain the disappearance from the many lines of documentary attestation of all traces of the original, uninterpolated text. Of course, this verse, Lk. iii. 23, has no bearing against the other principal form of the interpolation hypothesis, which supposes that the interpolation of Lk. i. 34, 35 was made by the author of the Gospel himself into his source; for Lk. iii. 23 does not stand within the infancy narrative. But even that form of the hypothesis is faced, as we have seen, by the difficulties presented by Lk. i. 27 and ii. 5. Thus it is not correct to say that if the one passage Lk. i. 34, 35 were deleted, the attestation of the virgin birth would be removed from the Lucan infancy narrative. If that passage is an interpolation, then at least one and probably two other passages must also be regarded as having been tampered with. But obviously every addition of such ancillary suppositions renders the original hypothesis less plausible. Nevertheless, the advocates of the interpolation hypothesis may still insist that although one or two verses in the infancy narrative outside of Lk. i. 34, 35 do imply the virgin birth, yet the bulk of the narrative proceeds upon the opposite assumption that Jesus was the son of Joseph by ordinary generation. The arguments in favor of this contention may perhaps be classified under three heads. In the first place, it is said, the narrative traces the Davidic descent of Jesus through Joseph, not through Mary, so that it must regard Joseph as His father. In the second place, Joseph is actually spoken of in several places as the “father” of Jesus, and

Joseph and Mary are spoken of as His “parents.” In the third place, there is attributed to Mary in certain places a lack of comprehension which, it is said, would be unnatural if she knew her son to have been conceived by the Holy Ghost. The fact upon which the first of these arguments is based should probably be admitted; it is probably true that the Lucan infancy narrative traces the Davidic descent of Jesus through Joseph. Whether it does so depends to a considerable extent upon the interpretation of Lk. i. 27. Do the words “of the house of David,” in that verse refer to Joseph or to Mary? ¹⁵ It seems more natural to regard them as referring to Joseph. This is so for two reasons. In the first place, the words come immediately after the name of Joseph; and in the second place repetition of the noun, “the virgin,” would not have been necessary at the end of the verse if Mary had just been referred to in the preceding clause; if “of the house of David” referred to Mary, the wording would be simply “to a virgin betrothed to a man whose name was Joseph, of the house of David, and her name was Mary.” Some modern Roman Catholic scholars have indeed argued with considerable force against this conclusion. The repetition of the word “virgin” instead of the use of the simple pronoun “her,” they argue, is to be explained by the desire of the narrator not merely to mention, but to emphasize, the virginity of Mary; and since Mary is evidently the chief person in the narrative, it is natural, they say, to take the three phrases; (1) “betrothed to a man whose name was Joseph,” (2) “of the house of David,” and (3) “the name of the virgin was Mary,” as being all of them descriptive of Mary. These arguments are certainly worthy of consideration—more consideration than they have actually received. And yet they are hardly sufficient to overthrow the prima facie evidence. It does seem more natural, after all, to refer the words “of the house of David” to Joseph.

If so, the Davidic descent of Mary is not mentioned in the narrative. There is indeed nothing in the narrative to prevent us from holding, if we care to do so, that Mary was descended from David. Certainly her kinship with Elisabeth ¹⁶ does not preclude such an opinion; for intermarriage between the tribe of Levi, to which Elisabeth belonged, and the other tribes was perfectly permissible under the law. No positive objection, therefore, can be raised to the view, which is held even by some scholars who reject the reference of the words “of the house of David” in Lk. i. 27 to Mary, that the narrator means to imply in his account of the annunciation to the virgin that Mary as well as Joseph was descended from David. But certainly the Davidic descent of Mary, even though it be held to be implied (which we for our part think very doubtful), is at any rate not definitely stated. If so, it looks as though the Davidic descent of Jesus were traced by the narrator through Joseph. But how can that be done if the narrator regarded the line as broken by the fact that Joseph was not really the father of Jesus? In reply it may be said that some persons in the early Church certainly did regard the two things—(1) the Davidic descent of Jesus through Joseph and (2) the virgin birth of Jesus—as being compatible. Such persons, for example, were the author of the first chapter of Matthew and the man who produced the present form of the first chapter of Luke, even though this latter person be thought to have been merely an interpolator. But if these persons thought that the two things were compatible, why may not the original author of the narrative in Lk. i-ii have done so? And if the original author did so, then the fact that he traces the Davidic descent through Joseph does not prove that he did not also believe in the virgin birth; so that the tracing of the Davidic descent through Joseph ceases to afford any support to the interpolation theory. It is another question, of course, whether the virgin birth is really compatible with the Davidic descent through Joseph.

All that we need to show for the present purpose is that it may well have been thought to be compatible by the author of the infancy narrative. However, it would be a mistake to leave the question, even at the present point in our argument, in so unsatisfactory a condition. As a matter of fact, there is, we think, a real, and not merely a primitively assumed, compatibility between the Davidic descent through Joseph and the virgin birth; the author of the first chapter of Matthew and also (if we are right in rejecting the interpolation theory) the author of the first two chapters of Luke had a perfect right to regard Jesus as the heir of the promises made to the house of David even though He was not descended from David by ordinary generation. We reject, indeed, the view of Badham that, according to the New Testament birth narratives, although Mary was a virgin when Jesus was born, yet in some supernatural way, and not by the ordinary intercourse of husband and wife, Joseph became even in a physical sense the father of Jesus. ¹⁷ This suggestion fails to do justice, no doubt, to the meaning of the narratives. In the first chapter of Matthew, and also really in the first chapter of Luke, the physical paternity of Joseph is clearly excluded. Yet it ought to be observed, in the first place, that the Jews looked upon adoptive fatherhood in a much more realistic way than we look upon it. In this connection we can point, for example, to the institution of Levirate marriage. According to the Old Testament law, when a man died without issue, his brother could take the wife of the dead man and raise up an heir for his brother. Evidently the son was regarded as belonging to the dead man to a degree which is foreign to our ideas. Because of this Semitic way of thinking, very realistic terms could be used on Semitic ground to express a relationship other than that of physical paternity. Then so eminent an expert as F. C. Burkitt, who certainly cannot be accused of apologetic motives, maintains that the word “begat” in the Matthaean genealogy does not indicate

physical paternity but only the transmission of legal heirship, so that even if the genealogy had ended with the words “Joseph begat Jesus,” that would not have afforded the slightest indication that the author did not believe in the virgin birth. ¹⁸ The truth is that in the New Testament Jesus is presented in the narratives of the virgin birth as belonging to the house of David just as truly as if he were in a physical sense the son of Joseph. He was a gift of God to the Davidic house, not less truly, but on the contrary in a more wonderful way, than if he had been descended from David by ordinary generation. Who can say that this New Testament representation is invalid? The promises to David were truly fulfilled if they were fulfilled in accordance with the views of those to whom they were originally given. In the second place, the relation in which Jesus stood to Joseph, on the assumption that the story of the virgin birth is true, was much closer than is the case with ordinary adoption. By the virgin birth the whole situation was raised beyond ordinary analogies. In an ordinary instance of adoption there is another human being—the actual father—who disputes with the father by adoption the paternal relation to the child. Such was not the case with Joseph in his relationship to Jesus, according to the New Testament narratives. He alone and no other human being could assume the rights and the duties of a father with respect to this child. And the child Jesus could be regarded as Joseph’s son and heir with a completeness of propriety which no ordinary adoptive relationship would involve. Thus the fact that in the Lucan infancy narrative Jesus is presented as the descendant of David through Joseph does not at all show that the narrative in its original form contained no mention of the virgin birth. Moreover, in refuting the first supposed proof of contradiction between the verses that attest the virgin birth and the rest of the narrative, we have really already refuted the second supposed proof. The second argument, as we observed, is based upon the application, in the second chapter of Luke, of the term “father” to Joseph and of the term “parents” to Joseph and Mary. ¹⁹ Of the instances where this phenomenon occurs, Lk. ii. 48 clearly belongs in a special category; for there the term “father” is not used by the narrator in his own name but is attributed by the narrator to Mary. Evidently, whatever may be the narrator’s own view of the relationship of Joseph to Jesus, it is unnatural that even if the virgin birth was a fact, Mary should have mentioned the special nature of that relationship in the presence of her Son. Thus in attributing the term “father” to Mary, in her conversation with Jesus, the narrator, if he did know of the virgin birth, is merely keeping within the limits of historical probability in a way which would not be the case if he had endeavored to make the virgin birth explicit at this point. But even the other occurrences of the term “father” or “parents” are thoroughly natural even if the narrator knew and accepted the story of the virgin birth. For, as we have just observed in connection with the matter of the Davidic descent, such terms could well be used on Semitic ground to describe even an ordinary adoptive relationship—to say nothing of the altogether unique relationship in which, according to the story of the virgin birth, Joseph stood to the child Jesus. Thus those manuscripts of the Old Latin Version which substitute in these passages the name “Joseph” for the term “father” and the phrase “Joseph and his mother” for the term “parents” are adopting an apologetic device which is altogether unnecessary. The absence of any such meticulous safeguarding of the virgin birth in the original text of Lk. ii shows not at all that the virgin birth was unknown to the author of that chapter, but only that the chapter was composed at an early time when naïvely direct narration had not yet given place to apologetic reflection.

The third supposed contradiction between Lk. i. 34, 35 and the rest of the narrative that has been detected by advocates of the interpolation theory, is found in those places where Mary is represented as being puzzled by evidences of the high position of her son. How could she have been surprised by such things, it is asked, if from the beginning she knew that the child had been conceived by the Holy Ghost? With regard to this argument, it may be said, in the first place, that the argument proves too much. If the wonder, or lack of comprehension, which Mary is represented as displaying at various points of the narrative shows that she could not have been regarded by the narrator as having passed through the experience predicted in Lk. i. 34, 35, it also shows that she could not have been the recipient even of the other angelic words. If Mary had had promised to her a son who was to be called a Son of the Most High ²⁰ and of whose kingdom there was to be no end, ²¹ why should she have been surprised by the prophecies of the aged Simeon or have failed to understand the emergence in the boy Jesus of a unique filial consciousness toward God? Surely the angel’s words, even without mention of the virgin birth, might have provided the key to unlock all these subsequent mysteries. Logically, therefore, the argument with which we are now dealing would require excision, not merely of Lk. i. 34, 35, but of the whole annunciation scene. But such excision is of course quite impossible, since the annunciation is plainly presupposed in the rest of the narrative and since the section Lk. i. 26-38 is composed in exactly the same style as the rest. Evidently the argument with which we are now dealing proves too much. But that argument faces an even greater objection. Indeed it betokens, on the part of those who advance it, a woeful lack of appreciation of what is one of the most beautiful literary touches in the narrative and at the same time an important indication of essential historical trustworthiness.

We refer to the delicate depiction of the character of Mary. These modern advocates of mechanical consistency seem to suppose that Mary must have been, or rather must have been regarded by the original narrator as being, a person of a coldly scientific frame of mind, who, when she had passed through the wonderful experience of the supernatural conception, proceeded to draw out the logical consequences of that experience in all their minutest ramifications, so that thereafter nothing in heaven or on earth could affect her with the slightest perplexity or surprise. How different, and how much more in accord with historical probability, is the picture of the mother of Jesus in this wonderful narrative! According to this narrative, Mary was possessed of a simple and meditative—we do not say dull or rustic—soul. She meets the strange salutation of the angel with fear and with a perplexed question; but then when mysteries beyond all human experience are promised her says simply: “Behold the handmaiden of the Lord; be it unto me according to thy word.” Then she journeys far to seek the sympathetic ear of a woman whom she can trust; and, when she is saluted in lofty words, she responds with a hymn of praise which is full of exultation but also full of reserve. Then when the child is born, and the shepherds come with their tale of the angelic host, others marvel, but Mary “kept all these things, pondering them in her heart.” But when Simeon uttered his prophecy about the light which was to shine forth to the Gentiles, Mary, with Joseph, marvelled at the things which were spoken about her child. No doubt, if she had been a modern superman, she would have been far beyond so lowly an emotion as wonder; no doubt, since her son had been born without human father, she would never have been surprised by so comparatively trifling a phenomenon as an angelic host that appeared to simple shepherds and sang to them a hymn of praise. But then it must be remembered that according to this narrative Mary was not a modern superman, but a Jewish maiden of the first century, nurtured in the promises of the God—the recipient, indeed, of a wonderful experience, but despite that experience still possessed of some capacity for wonder in her devout and meditative soul. And surely in the Palestine of the first century such a Jewish maiden is a more natural figure than the scientific monstrosity which some modern scholars seem to demand that she should be. Finally, when she saw her twelve-year old son in the Temple, in the company of the doctors of the law, she was astonished, and when her son said, “Wist ye not that I must be about my Father’s business,” she actually failed to understand. Truly that was unpardonable dullness—so we are told —on the part of one who knew that the child had been conceived by the Holy Ghost. We can only say that if it really was dullness, that dullness has been shared from that day to this by the greatest minds in Christendom. Has the utterance of the youthful Jesus ever fully been understood—understood, we mean, even by those who have been just as fully convinced of the fact of the supernatural conception as Mary was convinced if the experience actually was hers? There are depths in this utterance which have never been fathomed even by the framers of the Nicene and Chalcedonian creeds. It will be a sad day, indeed, if the Church comes to suppose that nothing in this word of the boy Jesus can be understood; but it will also be a sad day if it supposes that all can be understood. Mary can surely be pardoned for her wonder, and for her failure to understand. She had indeed passed through a unique experience; her son had been conceived in the womb without human father as none other had been conceived during all the history of the human race. But then when He had been born, with the mother’s very human pangs, He was wrapped in swaddling clothes and laid in a manger; and then He grew up like other boys, in the Nazareth home. No doubt from the point of view with which we are now dealing His lowly birth and childhood ought to have caused no questioning or wonder in Mary’s heart; no doubt she ought to have deduced from

these things, when they were taken in connection with the miracle of His conception, the full Chalcedonian doctrine of the two natures in the one person of the Lord; no doubt she ought to have been expecting the emergence, in the human consciousness of her child, of just such a sense of vocation and divine sonship as that which appeared when she found Him with the doctors in the Temple; no doubt she ought to have been far beyond all capacity for perplexity or surprise. But then we must reflect, from our modern vantage ground, that Mary was just a Jewish woman of the first century. It is perhaps too much to expect that she should be a representative of the “modern mind.” Perhaps she may even have retained the now obsolete habit of meditation and of quiet communion with her God; perhaps, despite her great experience, she may never have grasped the modern truth that God exists for the sake of man and not man for the sake of God; perhaps God’s mercies had to her not yet come to seem a common thing. Perhaps, therefore, despite the miracle of the virgin birth, she may still have retained the sense of wonder; and when angels uttered songs of praise, and aged prophets told of the light that was to lighten the Gentiles, or when her child disclosed a consciousness of vocation that suddenly seemed to place a gulf between her and Him, she may, instead of proclaiming these things to unsympathetic ears, have preferred to keep them and ponder them in her heart. So understood, the picture of Mary in these chapters is profoundly congruous with the verses that narrate the virgin birth. By the contrary argument modern scholars show merely that even for the prosecution of literary criticism something more is needed than acuteness in the analysis of word and phrase; one must also have some sympathy for the spirit of the narrative with which one deals. And if one approaches this narrative with sympathy, one sees that the supernatural conception is not only not contradictory to what is said about the thoughts of Mary’s heart but profoundly congruous with it. The words that recur like a refrain—

“Mary kept all these words and pondered them in her heart,” “Mary kept all these words in her heart”—place Mary before the readers in a way that is comprehensible only if she alone and not Joseph is the centre of interest in the narrative. And what made her the centre of interest save the stupendous wonder of the virgin birth? How delicate and how self-consistent is this picture of the mother of the Lord! Others might pass lightly over the strange events that occurred in connection with the childhood of her Son; others might forget the angels’ song; others might be satisfied with easy solutions of the problem presented by the consciousness of divine vocation which the youthful Jesus attested in the answer which He rendered in the Temple to His earthly parents. But not for Mary was such superficiality sufficient, not for the one who had been chosen of God to be the mother of the Lord. Others might be satisfied with easy answers to questions too deep for human utterance, but not so the one who had been overshadowed by the Holy Ghost. No, whatever others might do or say, Mary kept all these things and pondered them in her heart. We are, indeed, as far as possible from accepting the Roman Catholic picture of the Queen of Heaven. But we also think that Protestants, in their reaction against Mariolatrous excesses, have failed to do justice to the mother of our Lord. Few and simple, indeed, are the touches with which the Evangelist draws the picture; fleeting only are the glimpses which he allows us into the virgin’s heart. And yet how lifelike is the figure there depicted; how profound are the mysteries in that pure and meditative soul! In the narrative of the Third Gospel the virgin Mary is no lifeless automaton, but a person who lives and moves—a person who from that day to this has had power to touch all simple and childlike hearts. Whence comes such a figure into the pages of the world’s literature? Whence comes this lifelike beauty; whence comes this delicacy of reserve? Such questions will never be asked by those historians who reconstruct past ages by rule of

thumb; they will never be asked by those who know the documents without knowing the human heart. But to historians fully worthy of that name, the picture of Mary in the Third Gospel may seem to possess a self-evidencing power. Was such a picture the product of myth-making fancy, an example of the legendary elaboration which surrounds the childhood of great men? Very different, at least, were certain other products of such fancy in the early Church. Or is this picture drawn from the life; is the veil here gently pulled aside, that we may look for a moment into the depths of the virgin’s soul; is the person here depicted truly the mother of our Lord? Whatever answers may be given to these questions, whether the picture of Mary in these chapters is fiction or truth, one thing is clear—an integral part of that picture is found in the mention of the supernatural conception in the virgin’s womb. Without that supreme wonder, everything that is here said of Mary is comparatively meaningless and jejune. The bewilderment in Mary’s heart, her meditation upon the great things that happened to her son—all this, far from being contradictory to the virgin birth, really presupposes that supreme manifestation of God’s power. That supreme miracle it was which rendered worth while the glimpses which the narrator grants us into Mary’s soul. Thus general considerations will certainly not prove Lk. i. 34, 35 to be an interpolation; no contradiction, but rather the profoundest harmony, is to be found between these verses and the rest of the narrative. The Davidic descent could clearly be traced through Joseph, and was elsewhere traced through Joseph, even if Jesus was not by ordinary generation Joseph’s son; the term “father” as applied to Joseph does not necessarily imply physical paternity; the wonder in Mary’s heart at various things that happened during the childhood of her son does not exclude the greater miracle of His conception in the womb, but on the contrary contributes to the picture of which that greater miracle is an integral part. It certainly cannot be said upon general principles, therefore, that the writer of the rest of the narrative could not have written Lk. i. 34, 35. But if such general considerations—such considerations based upon the central content of the verses—will not establish the interpolation theory, what shall be said of the two verses considered in detail and in the immediate context in which they appear? Is it possible to discern elements of style in these verses which designate them as foreign to the narrative in which they now appear; or else is it possible to exhibit between them and their present context imperfect joints which would disclose an interpolator’s hand? The former of these questions must certainly be answered in the negative. Harnack, it is true, discovers in the use of two conjunctions in the verses evidences of a hand other than that of Luke. One of these conjunctions, ²² he says, occurs, indeed, a number of times in Acts, but nowhere in the rest of the Third Gospel (unless it is genuine in Lk. vii. 7 ²³); and the other ²⁴ according to the best text of Lk. vii. 1 (where it is probably not genuine) occurs nowhere else in the Lucan writings. ²⁵ But surely the facts with regard to the former of these two words are rather in favor of Lucan authorship than against it; the word, on Harnack’s own showing, does occur a number of times in Luke’s double work. And with regard to the other word, it may simply be remembered that an author’s choice of such words is seldom completely uniform. Bardenhewer ²⁶ gives a list of other particles beside this one that occur only once in the Lucan writings. In general it is significant that Zimmermann ²⁷ and, more recently, Vincent

Taylor ²⁸ can point to the Lucan character of the diction in these verses positively in support of their view that Luke himself, and not some scribe, was the interpolator. The truth is that the arguments of Zimmermann and Vincent Taylor, on the one hand, and of Harnack on the other, at this point simply cancel each other: the language of the two verses displays exactly the same combination of Jewish character with Lucan diction which appears everywhere else in the narrative. It is quite impossible to prove by stylistic considerations either that the verses are a Lucan interpolation into the source (or as Vincent Taylor would say into the original form of the Gospel) or a non-Lucan interpolation by a scribe. Nothing could be smoother, from a stylistic point of view, than the way in which these verses harmonize with the rest of the infancy narrative. If then no support for the interpolation theory can be obtained from stylistic considerations, what shall be said of the way in which the thought of the two verses fits into the immediate context? May any loose joints be detected by which the verses have been inserted, or does the whole section appear to be of a piece? In this connection, some of the arguments which have been advanced by advocates of the interpolation theory are certainly very weak. Thus when Harnack says ²⁹ that the question and answer in Lk. i. 34, 35 unduly separate the words, “Behold thou shalt conceive,” in verse 31, from the corresponding words, “Behold Elisabeth thy kinswoman has conceived, she also,” in verse 36, surely he is demanding a perfect regularity or obviousness of structure which is not at all required in prose style. Even if verses 34, 35 are removed, still the two phrases that Harnack places in parallel are separated by the important words of verses 32 f. As a matter of fact, it is by no means clear that the parallelism is conscious at all. But what is truly surprising is that Harnack can regard the content of this reference to Elisabeth as an

argument in favor of the interpolation theory instead of regarding it as an argument against it. The words in verses 36 f., Harnack argues, obtain a good sense only if no mention of Mary’s conception by the Holy Spirit has gone before; for if the most wonderful thing of all has already been promised, then it is weak and unconvincing, he thinks, to point, in support of this wonder, to the lesser wonder of Elisabeth’s conception in her old age. Surely this argument should be exactly reversed. The fact that in verses 36 f. the angel points, not to the career of Elisabeth’s son as the forerunner of Mary’s greater Son, but to something extraordinary in the manner of his birth, shows plainly that this example is adduced in illustration of something lying in the same sphere—namely in illustration of the greater miracle involved in the conception of Jesus entirely without human father in the virgin’s womb. If all that had been mentioned before was the greatness of a son whom Mary was to bear simply as the fruit of her coming marriage with Joseph, then nothing could be more pointless than a reference to the manner in which John was born. As a matter of fact, the plain intention is to illustrate the greater miracle (birth without human father) by a reference to the lesser miracle (birth from aged parents). It is perfectly true, of course, that there could be in the nature of the case no full parallel for the unique miracle of the virgin birth. But what the angel could do was to point to a happening that was at least sufficient to illustrate the general principle “with God nothing shall be impossible.” ³⁰ It is not surprising, therefore, that Hilgenfeld ³¹ apparently makes the reference to Elisabeth an argument, not against, but in favor of, the integrity of the passage, and that Spitta ³²

and others make it an argument for including verses 36 f. in the supposed interpolation. Against this latter hypothesis there are, indeed, the gravest possible objections. Against the view that the whole passage, embracing verses 34-37, constitutes an interpolation, the argument from the stylistic congruity of the supposed interpolation with the remainder of the narrative tells with crushing force. That argument was strong even if only verses 34 f. were regarded as interpolated. But in that case it might conceivably (though even then not plausibly) be said that the interpolation is too brief to disclose the stylistic variations from the rest of the narrative which in a longer interpolation might be expected to reveal the interpolator’s hand. But if the interpolator inserted so long a passage as verses 34-37, then it is truly a most extraordinary thing that he should have been able to catch the spirit of the infancy narrative so perfectly that nowhere in the whole course of his long insertion has he struck a single discordant note. Interpolators are not apt to be possessed of such wonderfully delicate skill. Moreover, it may turn out that there are still other special difficulties in the way of this modified form of the interpolation hypothesis. But unlikely though this modification of the interpolation hypothesis is, it does at least show a salutary feeling for the weakness of the more usual view. Certainly verses 36 f. are connected with 34 f. in the most indissoluble way; it is inconceivable that the reference to Elisabeth’s conception in her old age should be separated from the reference to Mary’s conception by the Holy Ghost. What we have here is a rather clear instance of the fate that frequently besets interpolation theories. The critic starts hopefully to remove something from a literary production. At first he thinks it is an easy matter. But then he discovers, to his consternation, that great shreds of the rest of the book are coming up along with the thing that he is trying to remove; the book proves to be not an agglomeration but an organism. So it is with Lk. i. 34, 35. At first it seems to be an easy matter just to remove these verses and so get rid of the disconcerting attestation of the virgin birth in a Palestinian narrative. But the thing proves to be not so easy as it seemed. For one thing, as we observed above, something has to be done with Lk. i. 27 and probably with Lk. ii. 5 and iii. 23. And then here in the immediate context it is quite evident that if Lk. i. 34 f. is to go, verses 36 f. must go too. We may, before we have finished, discover connections with still other parts of the context. At any rate, it should certainly be disconcerting to the advocates of the interpolation theory that what Harnack regards as a loose joint showing verses 34 f. to be no original part of their present context, is regarded by equally acute observers as being so very close a connection that if what appears in one side of the connection is interpolated what appears on the other side must also go. If the interpolation theory were correct, we might naturally expect some sort of agreement among the advocates of it as to the place where the joints between the interpolation and the rest of the narrative are to be put. Not much stronger, perhaps, though no doubt more widely advocated, than the arguments mentioned so far is the argument to the effect that verses 34 f. constitute a “doublet” with verses 31-33, and so could not originally have stood side by side with those former verses. In verses 31-33, it is said, Jesus is called Son of David and Son of the Most High; in verse 35 he is called Son of God because of the manner of his birth. If—so the argument runs—the writer had had in his mind the “Son of God” of verse 35, he would not have written the “Son of the Most High” and the “David His father” of verses 31-33.

With respect to this argument, it should be remarked in the first place that there is clearly no contradiction between the representation in verses 31-33 and that in verse 34 f. Offense has indeed been taken at the grounding of divine sonship in verse 35 upon the physical fact of divine paternity— “therefore also that holy thing which is begotten shall be called the Son of God.” How different, it is said in effect, is the Messianic conception of divine sonship in verses 31-33! But the question may well be asked whether the divine sonship of the child in verse 35 is grounded so clearly upon a physical fact of divine paternity as the objection seems to suppose. It is perfectly possible to take the word “holy” in that verse not as the subject but as part of the predicate. In that case, the words should be translated: “therefore also that which is begotten shall be called holy, Son of God.” On this interpretation it is not particularly the divine sonship but the holiness of the child which is established by the physical fact of the supernatural conception, and the divine sonship becomes merely epexegetical of the holiness. The decision between the two ways of construing the word “holy” is difficult. But even if the word is regarded not as predicate but as subject, still we do not think that there is the slightest antinomy as over against verses 31-33. Even if the meaning is: “therefore also that holy thing that is begotten shall be called Son of God,” we still do not see how such a grounding of the fact of divine sonship is contradictory to that which appears in the preceding verses. Certainly this verse does not intend to present the only way in which the divine sonship of the child is manifested. The verse says (in the construction that we are now discussing) that because of the supernatural conception the child shall be called Son of God; but it does not say that because of the supernatural conception the child shall be Son of God. We do not indeed lay particular stress upon this distinction. No doubt the distinction between “to be” and “to be called” is often not to be pressed; no doubt the passive of the verb “to call” in the New Testament sometimes implies not merely that a thing is designated as this or

that, but that it is rightly so designated. So here, “shall be called Son of God” may be taken as meaning by implication, “shall be rightly called Son of God”, and the emphasis may be upon the fact that justifies the calling rather than the calling itself. But whatever stress may be laid or may not be laid upon the distinction between “to be called” and “to be,” it is certainly absurd to take this sentence in an exclusive sense, as though it meant that the fact of the supernatural conception is the only reason why the child should “be called” or should “be” the Son of God. All that is meant is that the activity of the Holy Spirit at the conception of Jesus is intimately connected with that aspect of His being which causes Him to be called Son of God. One who was conceived in the womb by such a miracle must necessarily be the Son of God; a child who was conceived by the Holy Ghost could not be just an ordinary man. But clearly the verse does not mean that the supernatural conception was an isolated fact, and that it was the only thing that grounds the divine sonship of Jesus. Certainly the modern, exclusive way of interpreting such an utterance is quite foreign to the Semitic mind, which could place side by side various aspects of the Messiah’s person even before they were united in a systematic scheme. And at this point we are bound to think that the Semitic mind is preferable to the “modern mind.” Nothing could be more consistent than the passage, verses 31-35, as it stands. First the greatness of the promised child is celebrated in general terms; then, in response to Mary’s question, the particular manner of His birth is mentioned, and mentioned in a way thoroughly congruous with the generally supernatural character which has been attributed to Him before. How the divine sonship which appears in verses 31-33, can be regarded as incongruous with the virgin birth, or as rendering superfluous the mention of it, is more than we can understand. Verses 34 f. are not a disturbing or unnecessary doublet as over against verses 31-33; but render more specific one point which is included in this more general assertion.

At any rate, it is quite incorrect to regard verse 35 as connecting the divine sonship of Jesus with the supernatural conception in any anthropomorphic way. It is the creative activity of the Holy Spirit, and not any assumption of human functions of fatherhood, which is in view. The chaste language of verse 35 is profoundly congruous with verses 31-33, and in general with the lofty monotheism of the Old Testament; and it is profoundly incongruous with the crassly anthropomorphic interpretation which has sometimes been forced upon it by modern scholars. The arguments for the interpolation theory that have been mentioned so far are, we think, very easily refuted. Much more worthy of consideration is the argument with which we now come to deal. It is not indeed cogent as a support of the interpolation hypothesis; but at least it does call attention to a genuine exegetical difficulty which must be examined with some care. We refer to the argument based upon Mary’s question in verse 34: “How shall this be, seeing I know not a man?” This question has been regarded as being inconsistent with the context for two reasons. In the first place, why did not Mary simply assume that the child who has just been promised was to be the fruit of her coming marriage with Joseph? Since she was betrothed to Joseph, the fact that she was not yet living with him constituted no objection to the promise that she should have a child. In the second place, why is it that Mary should be commended, in the sequel, for her faith, if she had uttered this doubting question, which is very similar to the question for which Zacharias was so severely punished? Of these two objections it is the former which most deserves attention. The latter objection, despite the great stress that has been laid upon it by many advocates of the interpolation hypothesis, can surely be dismissed rather easily. It is true, indeed, that in the narrative Zacharias is represented as punished for his question, ³³ whereas Mary, despite her

question, is praised. ³⁴ But are the two questions the same? In form, it must be admitted, there is a certain similarity. Both Zacharias and Mary, instead of accepting the lofty promises of the angel without remark, ask a question betokening at least bewilderment; and both of them ground their bewilderment in an explanatory clause. But there the similarity ceases. Zacharias’ question reads: “According to what shall I know this?” That question can be interpreted as nothing else than a definite request for a sign; the wonder that is promised must be able to exhibit an analogy with something else before Zacharias will consent to “know” it. Mary on the other hand says simply, “How shall this be?”. She does not express any doubt but that it shall be, but merely inquires as to the manner in which it is to be brought to pass. Certainly she does not ask for a sign in order that she may “know” what the angel has told her will be a fact. To the modern reader, indeed, Mary’s question may seem to indicate doubt. In our modern parlance, the words: “I do not see how that can be,” or the like, may often mean that we do not think that it will be. Politeness, at the present time, is often a very irritating thing. But we have no right to attribute such politeness to Mary or to the writer who reports her words. And her question, as it stands, attests not a refusal to believe without further proof, but only perplexity as to what is involved in the angel’s words. Even in its wording, then, Mary’s question is different from that of Zacharias. But still greater is the difference in the situation which the two questions respectively have in view. Zacharias had been promised a son whom he had long desired, a son whose birth would bring him not misunderstanding and slander (as Mary’s son might bring to her) but rather a removal of the reproach to which, by his childlessness, he had been subjected. Moreover the birth of such a son, even in the old age of his parents, would be in accordance with the Old Testament analogies which Zacharias

knew very well. What except sinful unbelief could lead, under such circumstances, to the request for a sign? Mary, on the other hand, when the angel, prior to her marriage, spoke of a son, was promised something which seemed at first sight to run counter to her maidenly consciousness. Old Testament analogies, moreover, quite contrary to what was the case with Zacharias, could give her no help. Where in the Old Testament was it recorded that a son had been promised to a maid? Surely it is small cause for wonder that in such bewilderment she should have asked the angel for light? Even, therefore, if the wording of the two questions were more similar than it actually is, the underlying mind of the two speakers may still have been quite different. Zacharias was promised that which was quite in accord with Old Testament analogies and would mean the fulfilment of hopes that he had cherished for many a year; Mary was promised a strange, unheard of, thing, which might subject her to all manner of reproach. And yet finally (despite the strange explanation from the angel, which rendered the danger of that reproach only the more imminent) she said, in simple submission to the will of God: “Behold the handmaiden of the Lord, be it unto me according to thy word.” It is surely no wonder that Zacharias was punished and Mary praised. Much more worthy of consideration, we think, is the other one of the two objections to which Mary’s question has given rise. Indeed, the former objection, as has just become evident in the last paragraph, receives what weight it may have only from this objection with which we shall now have to deal. We have argued that if the angel’s promise to Mary seemed inconsistent with her maidenly consciousness, her question, unlike that of Zacharias, was devoid of blame. But, it will be objected, why should the promise have been interpreted by her in any such way; why should it have seemed inconsistent with her maidenly consciousness at all? The angel in the preceding verses has said nothing about anything peculiar in the birth of her son; why then did she not understand the promise as referring simply to her approaching marriage?

If she was going to ask any question, surely it ought to have been—thus the objection runs—a question about the greatness of her son rather than about the manner of his birth; the thing which ought to have caused surprise in view of the preceding words is not the mere fact that she was to have a son (for in view of her approaching marriage that was to be expected) but that she was to have such a son—that the son of a humble maiden at Nazareth was to assume the throne of David, that He was to be called the son of the Most High and that of His Kingdom there was to be no end. Her question in other words ought in view of the context to have been: “How shall this be, seeing I am a humble woman?”, instead of: “How shall this be, seeing I know not a man?” As it is, verse 34, we are told, reveals clearly an interpolator’s hand; it is entirely unnatural in view of the context, and merely constitutes a clumsy device for the introduction of an idea (the virgin birth) that was quite foreign to the original story. To this argument, Roman Catholic scholars have a ready answer. The question of Mary in verse 34, they say, is to be explained by the fact that she had already either made a vow, or at least formed a fixed resolve, never to have intercourse with a man; the present tense, “I know,” in the clause “seeing I know not a man,” is to be taken in a future sense, or rather as designating what was already a permanent principle of Mary’s life. Thus the meaning of the verse is: “How shall this be, since as a matter of principle I have determined not to know a man?” This solution certainly removes in the fullest possible way the difficulty with which we now have to do. And no objection to it can be raised from a linguistic point of view; there seems to be no reason why the present indicative, “I know,” could not be taken as designating a fixed principle of Mary’s life that would apply to the future as well as to the present. But the question is whether in avoiding one difficulty this Roman Catholic solution does not become involved in other difficulties that are greater still. In the first place, this solution runs counter to the prima facie evidence regarding the brothers and sisters of Jesus, who are mentioned in a number of places in the New Testament. Despite the alternative views—that these “brethren of the Lord” were children of Joseph by a former marriage or that they were merely cousins of Jesus, the word “brother” being used in a loose sense—it still seems most probable that they were simply children of Joseph and Mary. This conclusion is in accord with Lk. ii. 7, where Mary is said to have “brought forth her firstborn son”; for the word “firstborn” may naturally be held to imply that afterwards she had other children. The implication here is, indeed, by no means certain; for under the Jewish law the word “firstborn” was a technical term, which could be applied even to an only child, and in the sequel of this narrative stress is actually laid upon the fact that the legal provisions regarding the “firstborn” were fulfilled in the case of Jesus. Still, despite such considerations, the phrase does seem slightly more natural if Mary was regarded by the narrator as having other children. Such an interpretation would agree, moreover, with Mt. i. 25, where it is said that “Joseph knew her not until she had borne a son.” Here again the natural implication of the words can conceivably be avoided; it may be insisted that the author does not say that Joseph knew her after she had borne a son, but only that he did not know her before she had borne a son. And yet it does seem strange that if the narrator supposed that Joseph never lived with Mary as with a wife he should not have said that in simple words. In rejecting the Roman Catholic solution of our difficulty, we are not merely influenced by the positive historical evidence for the existence of other sons of Mary. Equally cogent is the negative consideration that if the narrator in the first chapter of Luke had meant that Mary had formed a resolve of perpetual virginity, he would naturally have indicated the fact in a very much clearer way. Such a resolve in a Jewish maiden of the first century would have been an unheard of thing. Asceticism, with the later prejudice against

marriage and the begetting of children, was quite foreign to the Jewish circles that are depicted in Lk. i-ii in such a vivid manner. If, therefore, the narrator were intending to attribute so extraordinary a resolve to Mary, he would naturally have taken pains to make his meaning perfectly clear; he might, for example, have been expected to tell of the special divine guidance which alone could have led a Jewish maiden to depart in such an unheard of way from all the customs and all the ingrained sentiments of her people. As a matter of fact, the narrator has done nothing of the kind. On the contrary, he has simply told us that Mary was betrothed to Joseph; and he has not hinted in any way whatsoever that the approaching marriage was to be a marriage in name only. Such a marriage is indeed set forth with great clearness in the apocryphal Protevangelium of James; but there is not the slightest hint of any such thing in our Third Gospel. If then the Roman Catholic solution is to be rejected, what shall be put in its place? If when Mary said: “How shall this be, seeing I know not a man?” she was not giving expression to a resolve of perpetual virginity with which a child in her approaching marriage with Joseph would seem inconsistent, how shall her question be understood? Why did she not simply assume that the son whom the angel had promised would be the fruit of her approaching union with her betrothed? Some modern scholars find an answer in the hypothesis of a mistranslation, in our Greek Gospel, of a Hebrew or Aramaic original of the angel’s words. If the future “thou shalt conceive” in verse 31, it is said, only were a present instead of a future, all would be plain; in that case the conception in Mary’s womb would be represented by the angel as taking place at once, so that Mary could not understand it as referring it to a marriage which still lay in the future, and so her bewildered question would be easily explained. Now although in our Greek text, it is said, the word translated, “thou shalt conceive,” is unequivocally future, the original of it in Hebrew or Aramaic would be a participle; and the participle might be meant to refer to the present as well as to the future—the decision in every individual case

being determined only by the context. In the present passage, it is said, the participle was intended, in the Semitic source, to refer to the present; and the whole difficulty has come from the fact that the Greek translator, who gave us our present form of Lk. i-ii, wrongly took it as referring to the future. If, then, the Semitic original is here restored, Mary’s question—since she could not explain a present conception in her womb by her future union with Joseph—becomes thoroughly suited to the context, so that there is no longer any indication of an interpolator’s clumsy hand. This solution, of course, assumes the existence of a Semitic original for the first chapter of Luke. That assumption is by no means improbable. But the question might arise how the Greek translator came to make the mistake. Would a translator be likely—for no particular reason, since the participle in the source might be translated by a present, even though it might also be translated by a future—would a translator be likely to introduce such serious confusion into the narrative in its Greek form? Obviously it would be more satisfactory, if possible, to find an interpretation which would suit the Greek narrative as it stands. Such an interpretation, we believe, is actually forthcoming, though it appears in a number of slightly different forms, between which we may not be able to decide. This true interpretation of the Greek text is not without affinity for the hypothesis of mistranslation which has just been discussed; indeed what it actually proposes is to find in the Greek words a meaning rather similar to that which the advocates of the theory of mistranslation have found in the Hebrew or Aramaic original. The Greek word, “thou shalt conceive,” is indeed future; but would it necessarily be referred by Mary to the time of her marriage with Joseph; might it not rather be referred by her to an immediate future? The latter alternative, we think, is correct. Annunciations, as they were known to Mary from the Old Testament, were made to married women; and when such an annunciation

came to her, an unmarried maiden, it is not unnatural that she should have been surprised. No doubt the influence upon her of the Old Testament narratives was not conscious; in the bewilderment caused by the angel’s greeting it is not likely that she reviewed consciously in her mind the stories of Hannah or of the wife of Manoah. But the unconscious effect of these stories may have been very great; they may well have served to create in her subconscious mind a close connection between angelic annunciations and the condition of a married woman as distinguished from that of a maid. Hence to her maidenly consciousness the promise of a son may well have occasioned her the utmost surprise. If, indeed, she had looked at the matter from the point of view of cold logic, her surprise might possibly have been overcome. She could have reflected that after all she was betrothed, and that the annunciation could in her case, as was not so in the Old Testament examples, be taken as referring to a married state that was still to come. But would such reflection have been natural; is it not psychologically more probable that she should have given expression, in such words as those in Lk. i. 34, to her first instinctive surprise? We have, then, in the current objection to Mary’s question another instance of that failure to understand the character of Mary, of that attempt to attribute to her, as she is depicted in this narrative, the coldly scientific quality of the “modern mind,” which has already been noticed in another connection. Suppose it be granted that in her question to the angel Mary was not strictly logical; is that any objection either to the ultimate authenticity of the question as a question of Mary, or to its presence in the narrative in Lk. i-ii? We might almost be tempted to say that a certain lack of logic in Mary’s words is a positive indication of their authenticity and of their original presence in this narrative. This absence of an easy, reasoned solution of all difficulties, this instinctive expression of a pure, maidenly consciousness, is profoundly in accord with the delicate delineation, all through this narrative, of the mother of the Lord.

But was maidenly instinct here really at fault; was Mary wrong in not simply referring the angel’s promise to her approaching marriage? Was she wrong in thinking that an immediate conception in her womb was naturally implied in the angel’s words? We are by no means certain that this is the case. On the contrary the very appearance of the angel and his momentous greeting would seem clearly to indicate some far more immediate significance in that moment than could be found merely in a promise concerning the indefinite future. After all, it was really strange in itself, as well as an offence to the consciousness of the virgin, if a child to be born in the approaching union with Joseph should be promised before instead of after the marriage. The future tense, “thou shalt conceive,” therefore, though not actually equivalent to a present, does refer most naturally to an immediate future. Thus the interpretation of the angel’s previous words which is implied in verse 34 is a very natural interpretation, and cannot possibly stamp verses 34 f. as an interpolation. This view avoids one difficulty that faces that theory of mistranslation which we have rejected. If the Hebrew or Aramaic participle of which the Greek, “thou shalt conceive,” is a translation were intended in a strictly present sense, there would seem to be a contradiction with Lk. ii. 21, where the name Jesus is said to have been given by the angel before the child had been conceived in the womb. If the conception were represented as taking place at the very moment when the word translated “thou shalt conceive” was uttered, then the name was given not before, but at the very moment of, the conception. On our view, on the other hand, it is possible to take Lk. ii. 21 in the strictest way, and yet find no contradiction with Lk. i. 31. The conception was represented by the angel as taking place in the immediate future, but not at the very moment when the word, “thou shalt conceive,” was spoken. It is impossible to say just when the conception is to be put. Many have thought of the moment when Mary said, “Be it unto me in accordance with thy word, " ³⁵

and this view has sometimes been connected with speculations about the necessity, for the accomplishment of the incarnation, of Mary’s act of submission. The salvation of the world, it has sometimes been held, depended upon Mary’s decision to submit herself to God’s plan; here as elsewhere, it has been held, God had respect to human free will. Such a way of thinking is contrary to ours. Of course our rejection of it does not by any means involve rejection of the view that puts the moment of the conception at the time when Mary uttered her final words. Yet on the whole we think it better to treat the question as it is treated by the narrator—with a cautious reserve. All that is involved in our view is that the “thou shalt conceive” in verse 31 refers to the near future, and would not naturally be taken by Mary as referring to her approaching marriage. It is quite possible that at this point we have claimed too much; it is quite possible that Mary’s question in verse 34 is not strictly logical; it is quite possible that she might well have taken the angel’s promise as referring to her approaching marriage. But that admission would not at all seriously affect our argument. Even if Mary’s question was not strictly logical, it was at least very natural; it was natural as expressing her bewilderment; like Peter at the Transfiguration, she knew not what she said. She was terrified at the angel’s greeting, and as a pure maiden she had not expected then the promise of a son. What wonder is it that her maidenly consciousness found expression in words that calm reflection might have changed? We are almost tempted to say that the less expressive of calm reasoning are Mary’s words in verse 34, so much the less likely are they to be due to an interpolator’s calculating mind, and so much the more likely are they to be due to Mary herself or to have been an original part of a narrative which everywhere depicts her character in such a delicate way. So far, we have been considering the arguments that have been advanced in favor of the interpolation theory. It is now time to consider a little more specifically the positive arguments that may be advanced against it. What positive indications, as distinguished from the mere burden of proof against the interpolation theory, may be advanced in favor of the view that Lk. i. 34 f. was an original part of the narrative in which it now stands? The strongest indication of all, perhaps, is found in the total impression that the narrative makes. We have been accustomed to read Lk. i-ii with appreciation of its unity and of its beauty only because the virgin birth is in our mind. But if we could divest ourselves of that thought, if we could imagine ourselves as reading this narrative for the first time and reading it without Lk. i. 34 f., it would seem disorganized and overwrought almost from beginning to end. The truth is that the child whose birth was prophesied by an angel and was greeted, when it came, by a choir of the heavenly host, is inconceivable as a mere child of earthly parents. No, what we really have here in this Christmas narrative is the miraculous appearance upon the earth of a heavenly Being—a human child, indeed, but a child like none other that ever was born. Not merely this detail or that, but the entire inner spirit of the narrative involves the virgin birth. Only partially can this total impression be analyzed. Yet such analysis is not without its value. It may serve to remove doubts, and so may allow free scope at the last for a new and more sympathetic reading of the narrative as a whole. Some of the details in Lk. i-ii which presuppose the virgin birth are of a subsidiary kind. But their cumulative effect is very great. Thus it has been well observed that Mary’s words of submission in Lk. i. 38 are without point if there has been no prophecy of the virgin birth in what precedes. If all that the angel has said is a prophecy that in her coming marriage Mary is to be the mother of the Messiah, why should there be this parade of submission on her part? These words are natural only if what has been promised involves possible shame as well as honor; then only do they acquire the pathos which has been found in them by Christian feeling throughout all the centuries and which the narrator evidently intended them to have.

It is such considerations, perhaps, which have led a few advocates of the interpolation theory to suggest that verse 38, as well as verses 36 f., may be regarded as part of the interpolation. But this suggestion only heaps difficulty upon difficulty. Without Mary’s final words of submission, the whole annunciation scene is left hanging in the air. Let the reader just imagine that verse 39 originally followed upon verse 33, and then let him see what effect is made by such an account of the scene. It will be evident enough that an artistic whole has been subjected to mutilation. What point is there, moreover, in the praise of Mary’s faith in verse 45—“Blessed is she who has believed; because there shall be fulfilment of the things that have been spoken to her from the Lord”—if Mary has not in what precedes given any expression to her faith? Evidently verse 45 refers to verse 38 in the clearest possible way. But verse 45 presupposes far more than verse 38; it also presupposes the stupendous miracle the promise of which Mary had believed. How comparatively insignificant would Mary’s faith have been if all that had been promised her was that her son in her coming marriage was to be the Messiah! Is it not perfectly evident that the faith for which Mary is praised is something far more than that; is the reference not plainly to her acceptance of an experience that involved possible shame for her among men and that was quite unique in the history of the human race. We have here a phenomenon that appears in the narrative from beginning to end. The truth is that this account of the birth and infancy of Jesus is all pitched in too high a key to suit a child born by ordinary generation from earthly parents. The exuberant praise of Mary’s faith, like many other features of the narrative, and indeed like the spirit of this narrative from beginning to end, seems empty and jejune unless the reader has in his mind the miracle which really forms the centre of the whole. But this is not the only point at which the account of Mary’s visit to Elisabeth presupposes the virgin birth. Certainly the account of the visit constitutes a clear refutation at

least of that form of the interpolation theory which includes in the interpolation verses 36 and 37. When the angel is represented in those verses as pointing to the example of Elisabeth, evidently the motive is being given for the journey that Mary immediately undertakes. “And Mary arose in those days and went with haste into the hill country into a city of Judah.” Why did she go at all, and especially why did she go in haste? Is it not perfectly clear that it was because of the angel’s words? Without verses 36 f. the whole account of the visit to Elisabeth is left hanging in the air. Verses 36 f., therefore, were clearly in the original narrative. But, as we have already pointed out, verses 36 f. presuppose verses 34 f. in the clearest possible way. As it stands, the narrative hangs together; but when the supposed interpolation is removed all is thrown into confusion. Hilgenfeld ³⁶ has pointed out still another way in which the account of Mary’s visit to Elisabeth presupposes Lk. i. 34, 35. Evidently at the time of the visit the conception is regarded as already having taken place. When Elisabeth says to Mary: “Blessed art thou among women, and blessed is the fruit of thy womb. And whence is this to me, that the mother of my Lord should come to me, ³⁷ her words seem overwrought if the conception is still to come. But if the conception has already taken place at the time of Mary’s journey, how is the journey to be explained? Surely it cannot be explained if Mary is regarded as already married to Joseph. In that case, as Hilgenfeld has well intimated, what would have been in place for Mary, if there was to be any journey at all, would have been a bridal tour with her husband, not a hasty journey far away from her husband to the home of a kinswoman. Is it not perfectly clear that the whole account of Mary’s visit to Elisabeth presupposes the supernatural conception? If Mary has passed through the wonderful experience promised in Lk. i. 34, 35, then everything falls into its proper place; then it is the most natural thing in the world for the angel to

suggest, and for Mary to carry out, a journey to visit her kinswoman, who also has passed through a wonderful, though of course far inferior, experience of God’s grace. But if Lk. i. 34 f. is omitted, everything is at loose ends. Even at the very end of the infancy narrative, the virgin birth seems to be presupposed. When it is said in Lk. ii. 51 that Jesus “went down with them, and came to Nazareth, and was subject unto them,” the sentence seems without point if Jesus was born of Joseph and Mary by ordinary generation. Why should it be thought a thing so remarkable that a child of earthly parents, even if the child was the Messiah, should be subject to its parents? The very way in which the submission of the boy Jesus to His earthly parents is introduced in the narrative suggests that His relationship to them was such as to make the submission an extraordinary and noteworthy thing. We should not, indeed, be inclined to lay particular stress upon this point if it were taken by itself. Perhaps one might say that if there was in the boy Jesus so extraordinary a consciousness of sonship toward God as is attested by His answer in the Temple, it was remarkable that He should subject Himself to earthly parents even if He were descended from them by ordinary generation. But that only pushes the difficulty in the way of an acceptance of the interpolation theory a step farther back. Is it likely that a son born of earthly parents by ordinary generation should have had such a stupendous consciousness of unique sonship toward God at all? We are really led back again and again, wherever we start, to one central observation. That central observation is that only a superficial reading of Lk. i-ii can find in this narrative an account of a merely human child; when the reader puts himself really into touch with the inner spirit of the narrative he sees that everywhere a supernatural child is in view. There is therefore a certain element of truth in the view advanced by the school of comparative religion to the effect that the child depicted in this narrative is a Gotteskind. That view is certainly wrong in detecting a polytheistic and mythological background for the stories of Lk. i-ii; but at least it is quite correct in observating that what the narrator has in view is no ordinary, merely human child. The whole atmosphere that here surrounds the child Jesus is an atmosphere proper only to one who has been conceived by the Holy Ghost. ³⁸ But it is time to turn from such general considerations to an argument of a much more specific kind. The argument to which we refer is found in the remarkable parallelism that prevails between the account of the annunciation to Mary and that of the annunciation to Zacharias. ³⁹ This parallelism shows in the clearest possible way that the verses Lk. i. 34, 35 belong to the very innermost structure of the narrative. In both accounts we find (1) An appearance of the angel Gabriel, (2) fear on the part of the person to whom the annunciation is to be made, (3) reassurance by the angel and pronouncement of a promise, (4) a perplexed question by the recipient of the promise, (5) a grounding of the question in

a causal clause, (6) reiteration of the promise with reference to something which in both cases is in the nature of a sign. The facts may best be indicated if we place the two sections in parallel columns: ⁴⁰

Lk. i. 11-20 1 Verse 11 And there appeared unto him an angel of the Lord standing on the right side of the altar of incense.

2 Verse 12 And when Zacharias saw him, he was troubled, and fear fell upon him.

3 But the angel said unto him, Fear not, Zacharias: for thy prayer is heard; and thy wife Elisabeth shall bear thee a son, and thou shalt call his name John. And thou shalt have joy and gladness; and many shall rejoice at his birth. For he shall be great in the sight of the Lord, and shall drink neither wine nor strong drink: and he shall be filled with the Holy Ghost, even from his mother’s womb. And many of the children of Israel shall he turn to the Lord their God. And he shall go before him in the spirit and power of Elias, to turn the hearts of the fathers to the children, and the disobedient to the wisdom of the just; to make ready a people prepared for the Lord.

4 Verse 18a And Zacharias said unto the angel, Whereby shall I know this?

Lk. i. 28-38 1 Verse 28 And the angel came in unto her, and said, Hail, thou that art highly favoured, the Lord is with thee.

2 Verse 29 And she was troubled at the saying, and cast in her mind what manner of salutation this might be.

3 And the angel said unto her, Fear not, Mary: for thou hast found favour with God. And behold, thou shalt conceive in thy womb, and bring forth a son, and shalt call his name Jesus. He shall be great and shall be called the Son of the Highest: and the Lord God shall give unto him the throne of his father David: and he shall reign over the house of Jacob for ever; and of his kingdom there shall be no end.

4 Verse 34a Then said Mary unto the angel, How shall this be,

5 Verse 18b for I am an old man, and my wife well stricken in years.

6 Verses 19-20 And the angel answering said unto him, I am Gabriel that stand in the presence of God; and am sent to speak unto thee, and to shew thee these glad tidings. And behold, thou shalt be dumb, and not able to speak, until the day that these things shall be performed, because thou believedst not my words, which shall be fulfilled in their season.

5 Verse 34b seeing I know not a man?

6 Verses 35-37 And the angel answered and said unto her, The Holy Ghost shall come unto thee, and the power of the Highest shall overshadow thee; therefore also that holy thing which is begotten shall be called the Son of God. And behold, thy cousin Elisabeth, she also hath conceived a son in her old age: and this is the sixth month with her, who was called barren. For with God nothing shall be impossible. It may be remarked in passing that even this exhibition does not fully set forth the connection between the two accounts. It does not show, for example, that in both cases the name of the angel is Gabriel, that the description of Mary in verse 27 is very similar in form to that of the parents of John in verse 5, that the Holy Spirit is mentioned in connection with the beginning of the earthly life both of John and of Jesus, and that the two accounts are specifically linked together by the words “in the sixth month” in Lk. i. 39. But even in itself the parallelism, when the two accounts are set forth as above in parallel columns, is so striking as to render almost inconceivable the hypothesis that it came by chance. No one who really attends to the structure of both sections should doubt but that they came from the same hand. In both cases the narrative is cast in the same mould. But if verses 34 and 35 were removed, this parallelism would be marred at the most important point. What, then does the interpolation hypothesis involve? It involves something that is certainly unlikely in the extreme—namely the supposition that an interpolator, desiring to insert an idea utterly foreign to the original narrative, has succeeded in inserting that idea in such a way as not only to refrain from marring the existent parallelism—even that would have been difficult enough—but actually to fill up in the most beautiful fashion a parallelism which otherwise would have been incomplete! We should have to suppose that the original narrator, though he did not include the virgin birth, left a gap exactly suited to its inclusion. And then we should have to suppose the appearance of an interpolator gifted with such marvellous literary skill as to be able, in the first place, to construct an interpolation that in spirit and style should conform perfectly to the body of the narrative, and then, in the second place, to insert that interpolation in just the place necessary to complete a parallelism which, when it was thus completed, makes upon every attentive reader the impression of being an essential element in the original framework of the narrative. Surely this entire complex of suppositions is improbable in the extreme. How, then, can we possibly avoid the simple conclusion that the parallelism between the two accounts, including the part of it which appears in Lk. i. 34 f., was due to the original narrator? At this point, however, there may be an objection. May it not be said that the very perfection of the parallelism that appears if verses 34, 35 are included constitutes an argument not for but against the originality of those verses? Have we not, in other words, in the inclusion of verses 34 f., something in the nature of a “harmonistic corruption”? May not an interpolator, observing the large measure of parallelism between the accounts of the annunciations, have decided to make that parallelism a little more complete than it actually was? A little reflection, we think, will show that these questions must be answered with an emphatic negative. The analogy with what is called a “harmonistic corruption” in textual criticism would not hold in this case at all. To show that it would not hold, we need only to glance at the harmonistic corruptions that actually appear in the text of the Synoptic

Gospels. What is the nature of these corruptions? An example will make the matter plain. The verse Mt. xvii. 21, “Howbeit this kind goeth not out but by prayer and fasting,” in the account of the healing of the demoniac boy after the descent from the mount of the transfiguration, is omitted by the so-called “Neutral” type of text as attested by the Codex Vaticanus and the Codex Sinaiticus. It is universally recognized as a gloss. But if it were genuine it would not add anything to our knowledge of the incident; for in Mk. ix. 29 very similar words are certainly genuine. It is perfectly evident that the text of Matthew has been made to conform to that of Mark. We have here, therefore, a typical example of a “harmonistic corruption.” But how totally different is this case from the case of Lk. i. 34 f., if these latter verses are really an addition to the original narrative! In the case of Mt. xvii. 21, a sentence is taken over in a mechanical way from a parallel account; in the case of Lk. i. 34 f., all that would be derived from the parallel account would be the sequence of question, grounding of the question, and answer: and the content of the interpolation would be of a highly original kind. Such originality would be quite unheard of among “harmonistic corruptions.” What we should have here would be no mere obvious filling out of a narrative by the mechanical importation of details from a parallel account, but the addition of a highly original idea—by hypothesis foreign to the original narrative—and the expression of that idea in a way profoundly congruous, indeed, with the inner spirit of the narrative, but at the same time quite free from any merely literary dependence upon what has gone before or upon what follows. It is doubtful whether any parallel could be cited for such a phenomenon in the entire history of textual corruptions. It appears, therefore—if we may use for the moment the language of textual criticism—that “intrinsic probability” and “transcriptional probability” are here in admirable agreement. On the one hand, the verses Lk. i. 34 and 35 are really in the closest harmony with the rest of the narrative; but on

the other hand that harmony is not of the obvious, superficial kind that would appeal to an interpolator. Indeed the very difficulty that we found in the interpretation of Mary’s question in verse 34 may be turned into an argument not for, but against, the interpolation theory. The difficulty is of a superficial kind that would probably have been avoided by an interpolator; the underlying harmony is of a kind worthy only of such a writer as the original composer of Lk. i-ii. Shall we attribute to an interpolator the delicate touch that is really to be found in Mary’s question? Is not the question rather—we mean not the invention of the question but the preservation of it—to be attributed to the writer who has given us the rest of this matchless narrative? In what has just been said, we have been using the language of textual criticism; we have been speaking of “intrinsic probability” and of “transcriptional probability” as though this were an ordinary question of the text. Such language would, of course, apply in fullest measure to that form of the interpolation hypothesis which finds in Lk. i. 34 f. an interpolation into the completed Gospel; for in that case we should actually be dealing with scribal transmission in the strictest sense. But the language could really apply in some measure also to the other forms in which the interpolation hypothesis has been held. In any case, we have in Lk. i. 34 f. an element that on one hand is in underlying harmony with the rest of the infancy narrative and yet, on the other hand, cannot be understood as being due to the effort of a later writer—whether the author of Luke-Acts or someone else—to produce that harmony by an insertion into this Palestinian narrative. Real harmony with the rest of the narrative, and superficial difficulty—these are the recognized marks of genuineness in any passage of an ancient work. And both these characteristics appear in Lk. i. 34 and 35. At any rate, whatever may be thought of our use of the terminology of textual criticism, the parallelism with the account of the annunciation to Zacharias stamps Lk. i. 34 f. unmistakably as being an original part of the account

of the annunciation to Mary. The argument comes as near to being actual demonstration as any argument that could possibly appear in the field of literary criticism. It is very clear that the two verses in question were part of the original structure of the narrative. But before this phase of the subject is finally left, it will be necessary to consider the alternative view as to the extent of the interpolation, which was suggested by Kattenbusch and has been advocated by Weinel and others. ⁴¹ According to these scholars, not the whole of Lk. i. 34 f. constitutes the addition to the narrative, but only the four words translated “seeing I know not a man” ⁴² in verse 34. If these four words are removed, it may be argued, there is in Mary’s question no reference to the manner in which her child is to be born; she is puzzled merely by the greatness of her promised son, and asks therefore, “How shall this be?”, without at all thinking of anything other than the son that she was to have in her approaching marriage with Joseph. In reply—so the hypothesis may be held to run—the angel in verse 35 points to an activity of the Holy Spirit securing the greatness and holiness of the son, without at all excluding the human agency in His conception in the womb; the child will be in a physical sense the son of Joseph and Mary; but just as the son of Zacharias was to be filled with the Holy Spirit at the very beginning of

his life, ⁴³ so the son of Joseph will be fitted by the same Spirit for a far higher function. In comment upon this hypothesis, it may be said, in the first place, that the hypothesis hardly accomplishes what it undertakes to accomplish; it hardly succeeds in removing the supernatural conception from Lk. i. 34, 35. Surely the minimizing interpretation which Weinel advocates for verse 35 is unnatural in the extreme. When Mary is told by the angel, “The Holy Ghost shall come upon thee, and the power of the Highest shall overshadow thee: therefore also that holy thing which is begotten shall be called the Son of God,” it seems very improbable that no more is meant than a sanctifying action of the Spirit upon a child conceived by another agency in the womb. Why should it be said, “The Holy Ghost shall come upon thee,” if the activity of the Spirit terminates upon the child in the womb rather than upon Mary? Why should not some expression like that in Lk. i. 15—“He shall be filled with the Holy Ghost”—be used if the work of the Spirit in both cases is essentially the same? Perhaps, indeed, the advocates of the hypothesis will maintain that on their view the work of the Spirit is not the same in both cases; perhaps they will say that in the case of John merely a sanctifying influence is meant, whereas in the case of Jesus the Spirit, though working indeed with the human factor, becomes constitutive of the very being of the child. But when that is said we are getting back very close indeed to the view that the Spirit’s action excludes the human father altogether. The truth is that in verse 35 the human father is quite out of sight; only two factors are in view—the mother Mary and the Spirit of God. “Conceived by the Holy Ghost and born of the virgin Mary” is really a correct summary of that verse. Even without the disputed words in verse 34, therefore, the following verse, verse 35, still presupposes the virgin birth. But if so, all ground for suspecting the words “seeing I know not a man” disappears. A second objection to Weinel’s hypothesis is found in the

parallelism with the annunciation to Zacharias to which attention has already been called. Weinel himself performed a very useful service by urging that parallelism as an objection to the ordinary form of the interpolation theory, which would remove all of verses 34 and 35. But he did not seem to observe that it tells also against his own view. If the words, “seeing I know not a man,” are removed from verse 34, then there is nothing to correspond to the grounding of Zacharias' question in verse 18. Let it not be said that we are expecting too perfect a similarity between the two parallel accounts. On the contrary, we recognize to the full the freshness and originality of verses 28-38 as over against verses 11-20; there are many details in one account that are not also in the other; the parallelism is by no means mechanical. But the point is that if Mary’s grounding of her question be removed from verse 34, it is not merely one detail that is subtracted but an essential element in the structural symmetry of the passage. It is really essential to the author’s manner of narrating the annunciation to Zacharias that Zacharias’ question should not merely indicate bewilderment in general, but should point the way for the explanation that was to follow. It seems evident that a similar plan is being followed in the case of the annunciation to Mary. But that plan is broken up if the words, “seeing I know not a man,” are not original in verse 34. Weinel’s hypothesis would force us to suppose that the original narrator left a gap in the structure of one of his parallel accounts, and a gap so exceedingly convenient that when by the insertion of four words an interpolator introduced into the narrative a momentous new idea, the most beautiful symmetry of form was the result. Surely such a supposition is unlikely in the extreme. It is perfectly evident, on the contrary, that the symmetry that results when Mary’s grounding of her question is retained is due not to mere chance or to what would be a truly extraordinary coincidence between a defect in the fundamental structure and an interpolator’s desires, but to the original intention of the author.

In the third place, Mary’s question in verse 34, in the shortened form to which Weinel’s hypothesis reduces it, seems unnatural and abrupt even apart from any comparison with the parallel account. According to Weinel, Mary said merely, in reply to the angel’s promise: “How shall this be?” In that form the question seems to have no point; it is a meaningless interruption of the angel’s speech. And it does not seem to prepare in any intelligible way for what follows in verse 35. No doubt there are narrators to whom such clumsiness could be attributed; but certainly the author of Lk. i-ii was not one of them. In this narrative, such banality would be singularly out of place. It is perfectly evident that in verse 34 the author is preparing for verse 35 in some far more definite and intelligible way than by the meaningless words, “How shall this be?”; Mary’s question is plainly intended to point the way to the special explanation that is given in the following verse. Thus on Weinel’s hypothesis the original narrator would at this point have suddenly descended to banality; and the beautiful naturalness and symmetry which now appears in the passage would be due not to the author but to an interpolator. Who can believe that such a supposition is correct? Such objections would be decisive in themselves. But there is another objection that is perhaps even more serious still. It is found in the extraordinary restraint which Weinel’s hypothesis is obliged to attribute to the supposed interpolator. An interpolator, we are asked to believe, desired to introduce into a Jewish Christian narrative of the birth of Jesus a momentous idea—the idea of the virgin birth—which by hypothesis was foreign to that narrative. How does he go to work? Does he insert any express narration of the event that he regarded as so important? Does he even mention it plainly? Not at all. What he does is simply to insert four words, which will cause the context into which they are inserted to appear in a new light, so that now that context will be taken as implying the virgin birth. Where was there ever found such extraordinary restraint,

either in an ordinary interpolator who tampered with the manuscripts of a completed book, or in an author like the author of Luke-Acts who desired to introduce a new idea into one of his sources? Is it not abundantly plain that if an interpolator desired to introduce the virgin birth into the narrative of Lk. i-ii he would have done so in far less restrained and far more obvious manner than Weinel’s hypothesis requires us to suppose. On the ordinary form of the interpolation hypothesis, which includes in the supposed insertion all of verses 34 and 35, we were called upon to admire the extraordinary literary skill of the interpolator, which enabled him to construct a rather extensive addition that should be highly original in content and yet conform so perfectly to the innermost spirit of the rest of the narrative. On Weinel’s hypothesis, on the other hand, it is the extraordinary restraint of the interpolator which affords ground for wonder. The surprising thing is that if the interpolator was going to insert anything—in the interests of the virgin birth—he did not insert far more. We have enumerated four special objections to the hypothesis of Weinel. With the exception of the one based on the parallelism with Lk. i. 11-20, they apply only to this hypothesis and not also to the more usual view as to the extent of the interpolation. That more usual view is in turn faced by some special objections that the view of Weinel avoids. But it must be remembered that some of the weightiest objections apply to both hypotheses alike. All that we have said regarding the plain implication of the virgin birth in Lk. i. 27 and ii. 5, and regarding the subtler implication of it at other points in the narrative, tells against any effort to find in the original form of Lk. i-ii a narrative that presented Jesus as being by ordinary generation the son of Joseph and Mary. What needs finally to be emphasized is that in holding the virgin birth of Christ to be an integral part of the representation in Lk. i-ii we are not dependent merely upon details. At least equally convincing is a consideration of the narrative as a whole. With regard to the results of such a general consideration, it may be well now to say a final word. In what precedes, we have laid special stress upon the parallelism between the account of the annunciation to Mary and that of the annunciation to Zacharias. That parallelism, we observed, establishes Lk. i. 34, 35 in the clearest possible way as belonging to the basic structure of the narrative; the (evidently intentional) symmetry of form between the two accounts is hopelessly marred if these verses, either as a whole or in part, are removed. But what now needs to be observed is that the difference between the two accounts is at least as significant, in establishing the original place of the virgin birth in Lk. i-ii, as is the similarity. In fact the very similarity finds its true meaning in the emphasis which it places upon the difference. One obvious difference, of course, is that the annunciation of the birth of John comes to the father of the child, while the annunciation of the birth of Jesus comes to the mother. What is the reason for this difference? Is the difference due merely to chance? Is it due merely to the way in which the tradition in the two cases happened to be handed down— merely to the fact that, as Harnack thinks, ⁴⁴ the stories regarding Jesus were preserved by a circle that held Mary in special veneration and had been affected in some way by the impression that she had made? If this latter suggestion is adopted, we have a significant concession to the traditional opinion, which has always been inclined to attribute the Lucan infancy narrative, mediately or immediately, to the mother of the Lord. Such an admission will probably not be made by many of those who reject, as Harnack does, the historicity of the narrative. And for those who will not make the admission, who will not admit any special connection of the narrative with Mary or with her circle, the central place of Mary instead of Joseph in the annunciation scene remains a serious problem. But even if we accept

the Marianic origin of the narrative—and do so even in a way far more definite than is favored by Harnack—still the unique place of Mary in the narrative requires an explanation. The point is not merely that Mary receives special attention—that her inmost thoughts are mentioned and the like—but that she is given an actual prominence that would seem unnatural if the child belonged equally to Joseph and to her. The fact is that we find ourselves here impaled upon the horns of a dilemma. If, on the one hand, the narrative is quite unhistorical, and not based upon any tradition connected with the actual Mary, then we do not see how the narrative or the legend lying back of it, ever came—since in this case it had full freedom of invention—to attribute such importance to the mother unless she was regarded as a parent of the child in some sense that did not apply to Joseph. Certainly the narrative displays no general predilection in favor of women as over against men; for in the case of John the Baptist the annunciation is regarded as being made to Zacharias, not to Elisabeth. If, therefore, it regards the relation of Joseph to Jesus as being similar to that of Zacharias to John, why does it not make him, like Zacharias, the recipient of the angelic promise? So much may be said for one horn of the dilemma. But if the other horn be chosen—if the narrator be regarded as being bound by historical tradition actually coming from Mary—still the prominence of Mary in the narrative remains significant. Are we to suppose that Mary attributed that prominence to herself without special reason? This supposition, in view of Mary’s character, as it appears in the narrative itself, is unlikely in the extreme. Thus, whatever view we take of the ultimate origin of the narrative, the prominence in it of Mary as compared with Joseph, which is so strikingly contrasted with the prominence of Zacharias as compared with Elisabeth, clearly points to something specially significant in her relation to the promised child, something which Joseph did not share. In other words it points to the supernatural conception, which is so plainly

attested in Lk. i. 34, 35. The removal of these verses by the advocates of the interpolation theory has really deprived us of the key that unlocks the meaning of the narrative from beginning to end. There is, moreover, another way also in which the relation between the two accounts of annunciations presupposes the virgin birth. What sympathetic reader can fail to see that the relation between the two accounts is a relation of climax? It is clearly the intention of the narrator to exhibit the greatness of Jesus in comparison with His forerunner, John. But in the annunciation of the birth of John the manner of the birth is given special prominence. The child, it is said, is to be born of aged parents; and around this feature a large part of the narrative revolves. The unbelief of Zacharias and the punishment of that unbelief are occasioned not by the prediction of later events in the life of the promised child, but by the prediction of the wonderful manner of his birth. Are we to suppose that in the parallel account there was nothing to correspond to this central feature of the annunciation to Zacharias? Are we to suppose that after laying such special stress upon the unusual manner of the promised birth of John the narrator proceeded to narrate a promise of a perfectly ordinary birth of Jesus; are we to suppose that it is the intention of the narrator that while John was born of aged parents by a special dispensation of divine grace, Jesus was simply the child of Joseph and Mary? No supposition, we think, would more completely miss the point of the narrative. Verses 36 and 37 surely provide the true key to the relation between the two accounts; the angel there points to the coming birth of John the Baptist from an aged mother as an example of that omnipotence of God which is to be manifested in yet plainer fashion in the birth of Jesus. In the light of this utterance, the whole meaning of the parallelism between the two accounts of annunciations becomes plain. The very similarities between the two cases are intended to set off in all the greater plainness the stupendous difference; and the difference concerns not merely the relative greatness of

the two children that are to be born but also the manner of their conception in the womb. A wonderful, if not plainly supernatural, conception in the case of John followed by a merely natural conception in the case of Jesus, which the interpolation hypothesis requires us to find, would have seemed to the composer of the narrative to involve a lamentable anticlimax. The entire structure of the narrative protests eloquently against any such thing. At this point, however, an objection may possibly be raised. It is not an objection against our argument in itself, but an argumentum ad hominem against our use of it. We have insisted that there is a conscious parallelism between the account of the annunciation to Zacharias and that of the annunciation to Mary, and that the author evidently intends to exhibit the superiority, even in the manner of birth, that Jesus possesses over against John. But—so the objection might run—does not such a view of the author’s intentions involve denial of the historicity of the narrative? If the author was ordering his material with such freedom as to exhibit the parallelism that we have discovered, and if he was deliberately setting about to show the superiority of Jesus over John, must he not, in order to pursue these ends, have been quite free from the restraint which would have been imposed upon him by information concerning what actually happened to Zacharias and to Mary? In other words, does not the artistic symmetry which we have discovered in the narrative militate against any acceptance of its historical trustworthiness? And since we are intending to defend its historical trustworthiness, have we, as distinguished from those who deny its trustworthiness, any right to that particular argument against the interpolation theory which we have just used. In reply, it may be said simply that our argument has not depended upon any particular view as to the way in which the symmetry, upon which we have been insisting, came into being. It would hold just as well if the author merely reproduced a symmetry which was inherent in the divine ordering

of the facts, as it would if he himself constructed the symmetry by free invention. In either case, the symmetry would be intentional in his narrative. Moreover, even in a thoroughly accurate narrative there is some possibility of such a selection and ordering of the material as shall bring certain features especially into view. A portrait, with its selection of details, is sometimes not less truthful but more truthful than a photograph. So in this case, the author, we think, was not doing violence to the facts when he presented the annunciation to Mary as in parallel with the annunciation to Zacharias. That parallelism, we think, was inherent in the facts; and the writer showed himself to be not merely an artist but a true historian when he refrained from marring it. But the point is that although the argument for the integrity of the passage which we have based upon the parallelism holds on the view that the narrative is historical, it holds equally well on the hypothesis that it is the product of free invention. In either case—however the parallelism came to be there—it certainly as a matter of fact is there; and an interpolation theory which holds that it was originally defective at the decisive point is faced by the strongest kind of objections that literary criticism can ever afford. Our conclusion then is that the entire narrative in Lk. i-ii finds both its climax and its centre in the virgin birth of Christ. A superficial reading may lead to a contrary conclusion; but when one enters sympathetically into the inner spirit of the narrative one sees that the virgin birth is everywhere presupposed. The account of the lesser wonder in the case of the forerunner, the delicate and yet significant way in which Mary is put forward instead of Joseph, the lofty key in which the whole narrative is pitched—all this is incomprehensible without the supreme miracle of the supernatural conception in the virgin’s womb. The interpolation hypothesis, therefore, not merely fails of proof, but (so fully as can reasonably be expected in literary criticism) is positively disproved. Princeton. J. GRESHAM MACHEN.