The Greek text of Luke 1.34 reads, εἶπεν δὲ Μαριὰμ πρὸς τὸν ἄγγελον· πῶς ἔσται τοῦτο, ἐπεὶ ἄνδρα οὐ γινώσκω;
In English: “And Mary said to the angel, ‘How shall this be since I do not know a man?’” This is in response to the angel’s statement in v 31: “You will become pregnant and give birth to a son, and you will name him Jesus.” A fifth-century Latin manuscript, Italab, however, has a unique reading in v 34. Instead of Mary denying that she has had sexual relations with any man, she instead says, “Behold, I am the Lord’s servant; let this happen to me according to your word” (ecce ancilla domini; contingat mihi secundum verbuntuum).
In his book, The Incredible Shrinking Son of Man, Robert M. Price argues that “without this verse there is nothing in Luke that even implies a supernatural conception or birth.”1 He goes on to argue that this manuscript alone follows the original text of Luke 1.34.
Recently, in Reinventing Jesus: How Contemporary Skeptics Miss the Real Jesus and Mislead Popular Culture,2 a book I co-authored with J. Ed Komoszewski and M. James Sawyer, I discussed Price’s argument that Italab by itself reflected the original wording in Luke 1.34. After considering the internal evidence (what the author would likely have written, judging by coherence in the context; what the scribes would likely have written, based on their known alterations due to intention or accident) and showing how it does not support Price’s conclusions (Reinventing Jesus, 98-99), I then examined the external evidence. I introduced that section with these words:
Perhaps the external evidence is stronger for the omission. If so, it would have to be nearly unanimous in order to overcome such strong opposition from the internal evidence. But Price acknowledges that external evidence is terribly weak—one Latin manuscript! Basically, there is no external evidence to support his claim. One fifth-century Latin manuscript involves no geographical distribution, no genealogical solidarity, and only minimal date and character credentials. The rest of the Latin manuscripts have this verse, as well as all the Greek manuscripts. And patristic writers have commented on this verse from early times.
Price is entirely too generous in his assessment of the omission when he gives it equal billing with the inclusion. His conclusion that “the evidence is too meager for us ever to be able to settle the question” sounds as if we need to suspend judgment because the evidence is so evenly balanced. Rather, the evidence for the omission is too meager to take Price’s suggestion seriously.
… The evidence for the lack of Luke 1:34 is so palpably weak that it is not even entertained by any serious New Testament scholar. How is it possible for one lone Latin manuscript to have gotten the wording right when all the other thousands of manuscripts—many of which are significantly earlier and with far better credentials than this one manuscript—let it slip through their nets? Price offers no plausible way in which the transmission of the text could have occurred so that the true text somehow was missed through more than four hundred years of copying but was caught by this one scribe. A good historian must at least offer some plausible explanation for such a unique anomaly. And he should also give evidence that, elsewhere in the text, a versional manuscript—or even a group of versional manuscripts—can contain the original wording when all the others produce an error.3
All of this is a matter of record. The evidence is very clear that Italab is in error here. I will discuss this more in a moment. For now, however, I need to note an error that was made not by this scribe but by me. After this point in Reinventing Jesus, I charged Price with inadvertently mixing up the data. The full statement is below:
Furthermore, Price got his facts mixed up. It is not Old Latin manuscript b that lacks this verse, but Old Latin manuscript β. That is a later manuscript than b by two centuries (seventh century vs. fifth century). It lacks Luke 1:34 because the manuscript is a fragmentary manuscript that contains only Luke 1:64–2:51. But Price does not tell us that the manuscript is fragmentary; he gives the clear impression that the scribe was unaware of this one verse and thus did not include it in his copy of Luke. If we were to apply that kind of logic to other manuscripts, we would have to say that the scribe of Codex Vaticanus thought that Hebrews ended in the middle of 9:14 because there is no more text after that—in fact, the page breaks off in the middle of a word.
The reality is that manuscripts suffer the ravages of time. Hundreds of them are missing a leaf or two or are mere fragments of a larger manuscript that is no longer extant. Price bases his argument on absolutely no evidence at all—no external evidence, no internal evidence. Rather, his philosophical presuppositions are driving his decisions. How then can he say, “The evidence is too meager for us ever to be able to settle the question”? The evidence, on the contrary, is absolutely solid that the Gospel of Luke never lacked 1:34. As William Lane was fond of saying, “An ounce of evidence is worth a pound of presumption.”
In this case, we have a pound of evidence versus an ounce of presumption. Even scholars who deny the virgin birth know that the texts that speak of it are not in question. It is a mere grasping at straws to even entertain the possibility that this is not the case, and it unmasks a wholesale agenda of destroying the faith of Christians by playing fast and loose with historical data. This is not the way any bona fide textual critic applies his trade. The most charitable thing we can say is that Price was sloppy and irresponsible in handling the data. And again, scholarly judgments can never properly be a matter of the will to disbelieve.4
Recently, I rechecked the data and discovered that, though it is true that Itala β does indeed lack Luke 1.1-63, Price was correct about Italab lacking Mary’s self-description as a virgin (the reading of this MS is noted at the beginning of this paper). In this case, I was in error about the data and I owe Robert Price an apology for charging him with being sloppy and irresponsible. In this instance, I was the one who was sloppy and irresponsible! I have alerted the publisher to fix the comments for the next printing. I have written to Dr. Price directly. And I have posted this apology on three different websites. Although I may strongly disagree with Dr. Price’s views, I was wrong in my assessment of the actual MS that he had in mind.
It should be mentioned, however, that this point came at the end of the discussion. The conclusions about Luke 1.34 were already solidly made. The evidence is so overwhelmingly against the reading of b that it doesn’t even register a blip in the apparatus of Nestle-Aland27 (the standard Greek New Testament used today, which records over 10,000 textual problems).
How is the reading of b in v 34 to be explained? The first thing that a textual critic must do when such an anomaly occurs in the text is to ask if there is any way to explain it by way of unintentional alteration on the part of the scribe. If it can be explained that way—especially if the reading finds no support in any other witnesses—then that is the most probable solution. In this case, an unintentional error seems evident.
Both v 34 and v 38 begin with exactly the same introduction: εἶπεν δὲ Μαριάμ (‘and Mary said’; or, in Latin, dixit autem Maria). The scribe’s eye could easily have skipped down to v 38 and he could have written Mary’s response in that verse and placed it in v 34. The editors of The Gospel according to St. Luke, vol 1, IGNTP, note at 1.38, “Lvt (b) places these words after verse 34”—indicating that b does not place them instead of v 34. This comment, however, may be overgenerous, for the evidence is that b simply replaces Mary’s response in v 34, not the whole of v 34.
It seems that the reading of b is to be explained as a case of haplography (writing once what should have been written twice) in which the scribe’s eye skipped to v 38 when reading dixit autem Maria. After he wrote the response from v 38, he continued reading from v 35. (It would require the scribe to look ahead at one line of text in his exemplar [i.e., the ‘master’ copy he was copying from], write that down in his copy, then return to the exemplar to the verse that followed the one he skipped originally. Anyone who has copied texts knows how easy it is make this sort of mistake.) When he got to v 38, he recognized that he had written this text already and so he omitted the entire response in this location. The reading of b in v 38 is simply et disces sit ab illa angelus (‘and the angel departed from her’). Thus, Mary’s response in v 38 has been removed and placed in v 34. That there are no other MSS—Greek, Latin, Coptic, Syriac, etc.—that do this is instructive.
Further, there is internal evidence that this is an error of sight on part of b’s scribe rather than a reflection on a primitive tradition that denied Mary’s virginity at the time of the conception: in v 27, b agrees with the rest of the manuscript tradition in calling Mary a virgin (virginem; Greek παρθένον) at the time of the angel’s visit. This seems to contradict what Price says: “without [verse 34] there is nothing in Luke that even implies a supernatural conception or birth….”5 But v 27 indicates that Mary was a virgin at the time of the angel’s visit, and in v 35 the angel answers ( ἀποκριθείς) that Mary would become pregnant by divine intervention. (Further, the fact that the angel answers Mary suggests that Mary asked a question. In Italab the question is removed, leaving the structure of the dialogue somewhat unnatural. But if the question that is found in all other witnesses in v 34 is retained, then the response of the angel is perfectly fitting.)
It also seems confirmed by those Latin MSS that are closest to the text of b (such as ff2 and q, from the fifth and seventh6 centuries, respectively). Old Latin ff2 and b “in St. Luke have certain common renderings—some of them blunders—which cannot have arisen independently.”7 And q had “a far stronger resemblance to b than to any other Old-Latin MS” according to its editor.8 Yet neither of these manuscripts—nor any others—lack Mary’s response in v 34.
All in all, the evidence is rather overwhelming that Luke 1.34 did indeed contain Mary’s question to the angel: “How shall this be since I do not know a man?” I know of no other textual variant that is found in a single versional witness that ought to be regarded as authentic. And the fact that this particular variant can be explained as an accidental alteration, and is not corroborated in b’s closest allies—not to mention any other witnesses—is convincing. And this probability is hardly lessened by my misclassification of Itala β as the manuscript that Price was writing about. Whether there are no Latin manuscripts or only one, the external and internal evidence against such a reading is securely grounded. That even the Jesus Seminar recorded Mary’s question in Luke 1.34 as the original text—as much as they collectively would like to expunge it from the biblical record—speaks volumes.9
1 Robert M. Price, The Incredible Shrinking Son of Man: How Reliable Is the Gospel Tradition? (Amherst, NY: Prometheus, 2003) 70.
2 Grand Rapids: Kregel, 2006.
3 Reinventing Jesus, 99-100.
4 Ibid., 100-101.
5 Price, Incredible Shrinking Son of Man, 70.
6 However, Bonifatius Fischer, Die lateinischen Evangelien bis zum 10. Jahrhundert, vol 3: Varianten zu Lukas (Freiburg: Herder, 1990), 13*, dates Italaq as sixth-seventh century. He also dates as late fifth century (ibid.).
7 E. S. Buchanan, editor, The Four Gospels from the Codex Veronensis (Oxford: Clarendon, 1911) xx-xxi.
8 Henry J. White, editor, The Four Gospels from the Munich MS. (q) Now Numbered Lat. 6224 in the Royal Library at Munich (Oxford: Clarendon, 1888) li.
9 Their fresh translation, based on examination of the original languages, renders Luke 1.34 as “And Mary said to the messenger, ‘How can this be, since I’ve not had sex with any man[?]” Although textual variants are regularly listed in the footnotes, there is none here for Luke 1.34. See The Complete Gospels, revised and expanded version, ed. Robert J. Miller (Sonoma, CA: Polebridge, 1994) 119.