How early is belief in the virgin birth?

Introduction

It is a commonly repeated claim that belief in the virgin birth was not part of the original Christian preaching and was a later invention. The claim is based upon the idea that the virgin birth is absent from Mark’s gospel and the letters of Paul, and is not attested until the gospels of Matthew and Luke. Given these two gospels are traditionally dated later than Mark’s gospel and the letters of Paul, the appearance is given that the virgin birth appears late on the scene.

This reconstruction parallels another common proposal that the earliest Christologies did not hold Jesus to be the Son of God by birth but by adoption (or even, didn’t regard Jesus as Son of God at all). It fits this “evolutionary” paradigm to argue that belief in the virgin birth was an earlier invention. But does this chronology of doctrine have any basis in fact? How early is belief in the virgin birth?

Dating Matthew’s Gospel

Scholarly convention requires one to nod politely at the consensus that Mark’s gospel was written c.70 and the other Synoptics c.80 or perhaps a little later. Having paid our dues to convention, we may examine the evidence and comment on the oddity of taking as near axiomatic something that stands on so shaky foundations.

If it is argued that the Olivet Prophecy requires a date of composition after the destruction of Jerusalem then we are presented with an implied scenario whereby the evangelist, writing (ex hypothesi) a decade or more after that fateful event, attributed to Jesus the prediction that “immediately after the tribulation of those days” the Parousia (i.e. the Return) would occur (Matt 24:29). So, either the evangelist sees no problem with attributing an obvious falsehood to Jesus or we should dispense with such absurdity and conclude that the destruction of Jerusalem had not yet occurred. Even if we denied the possibility of predictive prophecy and  the authority of Jesus to issue such, it would nevertheless be no great stretch for any Jew in the early first century to observe the political situation and issue the warning about where events were headed (if indeed that is what Jesus was doing).

If it is argued that Matthew’s gospel is not cited in other early Christian writings then, first, we must respond that it is (Barn. 4:14) and, second, we must ask why would one think it should be. Textual citation presupposes not only a readily available copy to cite from but also knowledge of the same book by the readers of that citation. It is reasonable to suppose that there was a transitional period between the composition of any text and the time when it became useful for textual citations. Yet even if we suppose that no such transitional period was necessary and that the NT writers would have taken a freshly penned gospel and cited immediately, is the absence of citation a reliable indicator that a book was not yet composed? Take Paul’s pastoral letters, which, if authentic, probably date from the early 60s. The UBS 4th edition indexes only two quotations from this corpus (Deut 25:4 / 1 Tim 5:18; Num 16:5 / 2 Tim 2:19); can such a sample by taken as indicative of anything? Ironically, the quotation that the UBS does not index is from a synoptic gospel (Luke 10:7 / 1 Tim 5:18). So would it be reliable to conclude that Paul (or the pseudonymous writer) had no acquaintance with Matthew’s gospel because out of three quotations he did not include one from Matthew? Or should we conclude that this argument from silence is almost worthless?

Looking at the positive evidence for dating Matthew’s gospel it is worth reiterating the observation that the evangelist seems to imply that the temple is still standing, that the temple cult is still in operation and that the parting of ways with Judaism has yet to occur (Matt 5:23-24; 12:5-7; 17:24-7). J. A. T. Robinson writes, “Matthew’s gospel shows all the signs of being produced for a community (and by a community) that needed to formulate, over against the main body of Pharisaic and Sadducaic Judaism, its own line on such issues”.[1] Moreover, Robinson notes that this, the most Jewish of gospels, lacks any echo of the martyrdom of James (c.62) or any prominence given to his successor Simeon.[2] J. Wenham draws attention to Matthew’s apologetic regarding the report that the disciples stole the body, which, the evangelist says, is still commonly told by the Jews “until this day” (Matt 28:15). Whilst it is true that this report was told and mythologized by the Jews for centuries afterwards, the only reason for Matthew to defend against such a report is if it was detracting from the evidence of the empty tomb. “As serious apologetic this extra evidence would have had limited value outside Jerusalem, but in Jerusalem itself in the early days the story would have been known to almost everyone, and the true version of the story would carry weight”.[3] These indicators would place Matthew’s gospel more naturally in the 50s or early 60s.

It is worth noting, though it has not been widely accepted, the proposal of C. P. Thiede that the Magdalen Papyrus, which preserves fragments of Matthew 26, should be dated no later than c.66.[4] If his palaeographic analysis is accepted then the composition of Matthew’s gospel must be dated to the early 60s at the latest.

The synoptic question is also a factor in dating the gospels. If, as is generally supposed, Mark was written first and the other Synoptics depended on this gospel, then Matthew’s gospel must post-date Mark (generally dated c.70). Yet this proposal is not universally accepted. J. Rist, for example, has argued for the independence of Matthew and Mark; he proposes a date for both between 60 and 70. [5] Wenham has proposed that the three Synoptics draw separately from a shared tradition and has defended the priority of Matthew, as was universally affirmed before the modern era. He argues that Luke was already well-known by the mid-50s (cf. 1 Tim 5:18), that Mark was written following Peter’s first visit to Rome in the early 40s and so dates Matthew to c.40.[6] Robinson, though accepting a more conventional solution to the synoptic question, still dates the composition of Matthew broadly between c.40-60+.[7]

Therefore, whilst nodding politely at the c.80 date of composition, it seems more probable that Matthew’s gospel dates several decades earlier. Let us then, for sake of this argument, date Matthew’s gospel to c.60 (though I see no strong reason why it could not be earlier).

Dating Luke’s Gospel

When we come to date Luke’s gospel then much of the argumentation is similar. We nod politely at the scholarly consensus that would date the composition c.80 and then proceed to reject the arguments upon which this consensus is based (see above).

There are two additional arguments to consider. Firstly, Luke ends Acts with Paul dwelling in prison for “two whole years” (Acts 28:30). Two whole years till when? Not his death (c.66), Luke makes no mention of that. Not his appearance before Caesar, though that would be the natural terminus for his narrative arc.[8] So why end there? The conclusion that seems so obvious to many scholars (and yet is strangely considered insignificant to many others) is that Luke ends there because the rest hadn’t happened yet. And if that is the case then that provides a (relatively) small window in which Acts could have been composed (Robinson gives 57-62). And given that Acts is a sequel to Luke (cf. Acts 1:1) then Luke must pre-date Acts (though, perhaps, not by much).

The second consideration is the quotation of Luke’s gospel in 1 Tim 5:18. If 1 Timothy is authentic, then it dates from sometime between 55 and 66; certainly no later than Paul’s martyrdom (c.66). If 1 Timothy is pseudonymous then it dates no later than its earliest citation (c.110; Poly .Phil 4.1, 12.3).[9]

Therefore, once again we nod politely at the c.80 date of composition, but judge that c.60 is more likely and, for sake of this argument, accept that date.

Gospel Sources

So far so dull, but we must not suppose that by dating Matthew and Luke that we have thereby dated the origin of belief in the virgin birth. For there is not one virgin birth narrative but two and, laying aside the staggering coincidence that two evangelists invented the same conception, we must conclude that the idea did not originate with them.

The relationship between Matthew and Luke is disputed (the synoptic question) and I do not propose to settle the question in this paper. Rather, I will explore what each of the major alternatives would mean for the virgin birth narrative.

The Austin Farrer Theory

The theory that postulates the greatest level of dependence between Matthew and Luke is the ‘Farrer’ theory, more recently defended by M. Goulder and M. Goodacre. According to this theory, Mark is the earliest gospel; Matthew depends on Mark and Luke depends on these two.[10]  The implication for the infancy narratives is that Luke was aware of Matthew’s account when he penned his own. For instance, J. Knight claims that “Luke writes with a knowledge of Matthew’s Gospel” as the source of the virginal conception, though he makes no visible attempt to show how the one account could be derived from the other.[11] Given the differences between the two narratives, this would mean that either Luke choose to edit out Matthew’s account in favour of another source or else to deliberately alter Matthew’s account  for narrative purposes.

There is very little that is similar between Matthew and Luke’s account, which suggests it is unlikely that Luke is using Matthew’s as a base. G. Parrinder writes “Luke writes blithely in ignorance or disregard of Matthew”.[12] More likely, then, that Luke chooses to ignore Matthew entirely. Goulder proposed that both Matthew and Luke simply invented incidents, like the infancy narratives. However, Rist responds, arguing that even had Matthew invented his infancy narrative, it is unlikely that Luke, depending on Matthew, would have done so.[13] It would be odd for Luke to ignore Matthew’s infancy narrative, given ex hypothesi, Luke follows Matthew nearly word-for-word in places. It is not obvious that Luke’s theological purposes would lead him to so utterly reject Matthew’s account. The possibility that Luke is using another source for his infancy narrative looks more promising but still does not explain why Luke chose to reject all of Matthew’s account, except perhaps that he found this approach easier than trying to integrate the two accounts. Perhaps then we have reason to be sceptical about the Farrer approach.

Yet whatever the case, on the Farrer theory, Luke, writing with knowledge of Matthew’s account, chose to change every part of the narrative except the central core. Both accounts record that Mary was a virgin,[14] that Jesus was conceived by the Holy Spirit,[15] the angelic prescription of the name ‘Jesus’,[16] his birth in Bethlehem[17], the fact that Mary was betrothed (not married)[18] and the name of her intended (‘Joseph’).[19] If Luke felt so free as to change every other part of Matthew’s account, then we must provide some reason why Luke felt compelled to preserve this core. And the most straightforward explanation is that he knew this core from some other tradition, in addition to Matthew’s account.

Drawing out the implications, if we date Luke to c.60 (see above) then, on the Farrer theory, Matthew must be dated earlier. Let us assign a notional gap of five years and date Matthew c.55. I have proposed that there is also the other tradition known to Luke, which is of unknown date and provenance. Let us assign an arbitrary date of c.55 for this tradition. Since Matthew and this other tradition have a shared core, then the belief in the virgin birth must pre-date both. Let us assign a notional gap of five years and say that this belief must be earlier than c.50.

Two-Source Theory

The majority view regarding the synoptic question is that Matthew and Luke are independent of each other but draw on two shared sources: Mark and (the hypothetical) Q. This theory derives from the observation that both Matthew and Luke replicate much that is in Mark but that they also parallel each other in passages not present in Mark. This non-Markan overlap is thought to be derived from Q. Grand claims have been made for Q; it has been called the “lost gospel”,[20] the “first gospel”[21] and the authentic teaching of Jesus.[22] For some Q, or its earliest redaction, is their core for reconstructing the historical Jesus.[23] Yet it must be recognised that Q is a hypothetical source and anything said about Q is conjecture based purely upon textual overlap between two gospels. As C. Evans writes, “scholars say far too much about the hypothetical source”.[24]

Mark does not include an infancy narrative, but might Q have included one? The simple answer is we cannot know. We are not able to determine the true extent of Q. As Evans has argued, if we attempted to reconstruct Mark based upon overlap between Matthew and Luke we would recover only around 60% of the gospel, even though 80% is attested, and important sections would be lost.[25] If Matthew (say) depended on Q for his infancy narrative and Luke did not (for whatever reason) then there would be no overlap and hence no evidence for the presence of this narrative in Q. It is sometimes argued that Q was a ‘sayings’ gospel, based upon comparison with the Gospel of Thomas and the a priori suspicion that the earliest records of Jesus would be a collection of sayings. However, the overlap between Matthew and Luke does not consist purely of sayings (e.g. Luke 4:13) and so we cannot rule out the possibility that it included narrative passages. (It is not clear why one would think that Thomas is representative of the earliest gospel form when it evidently was not earlier than canonical gospels). So we cannot rule out the possibility that Q (if such a source existed) contained an infancy narrative. And if (for some reason unfathomable to me) we judge that Q must be earlier than Mark then the possibility that it included an infancy narrative is highly significant. But, as I have said, anything we say about Q is conjecture (if not rampant speculation).

Let us suppose that neither Mark nor Q contained an infancy narrative; whence came Matthew and Luke’s infancy narratives? It might be proposed that neither evangelist had a source for what they wrote, despite the fact that according to the Two-Source theory they both depended heavily on sources for the majority of their respective gospels. Yet this proposal is immediately rendered untenable by the shared core between the two narratives. However creative the two evangelists might have been, it is necessary to suppose that they shared a source according to which Jesus was conceived by the Holy Spirit in the womb of the Virgin Mary, who was betrothed to Joseph, and that his birth took place in Bethlehem and that his name was given by an angel. And if that source is not Mark (evidently) or Q then it is some third source/tradition.[26]

However, if this third source was a written source upon which both Matthew and Luke depended then we once again encounter the problem that their two accounts are so different. We might speculate that Matthew (say) followed the source closely and that Luke only followed the source loosely, but then once again it is left inexplicable why Luke should depend upon a source and yet deviate so widely from it. The more promising alternative is that there was more than one source.

A variant on the ‘Two Source’ solution to the synoptic question is the ‘Four Source’ solution that posits that there were two other sources, in addition to Mark and Q that the evangelists utilised: Matthew depended on M for the material unique to his gospel and Luke depended on L for the material unique to his gospel. It is generally supposed that M, if it existed, contained a number of parables and that L, if it existed, contained narrative sections and some sayings. Yet, in fact, we can only guess at the full extent of these hypothetical sources. If they were, in any sense, cohesive gospels (that is, something analogous to the Synoptics), then it is entirely plausible that they might have included an infancy narrative. Or, if it is judged their style and form make it unlikely that they contained infancy narratives, (and I don’t know how one reliably makes those judgments), then we are left to speculate about some other sources or traditions that formed the basis of these narratives.

Again, drawing out the implications, if we date both Luke and Matthew to c.60 (see above) then the source(s) on which they depend for their infancy narratives must be earlier. Following the procedure above, let us set a notional gap of five years and date Q, M, L or any other written source for the infancy narratives to c.55. As we have argued above, it is unlikely that Matthew and Luke can have used the same source for their narratives (given their clear differences), yet given the shared core between these narratives the belief in the virgin birth must pre-date both sources (whatever they were). Again, assigning a notional gap of five years, let us say that the belief must be earlier than c.50.

Oral Tradition

A third form of solution to the synoptic question is the proposal that, rather than depend one upon another, the evangelists wrote independently based upon an oral tradition. It should not be assumed that oral means vague. Indeed, for this proposal to work, it requires that the tradition is stable and consistent; how else would it explain the close similarity between the Synoptics? There are various approaches to this theory. For example, Rist proposes that Matthew and Mark are independent, working from the same oral tradition(s) but that Luke depended on these prior gospels, whereas Wenham proposes that all three evangelists worked independently from the same tradition. Yet the point of significance is that according to these theories there was a body of oral tradition predating the Synoptic gospels upon which these gospels drew and that oral tradition would include the infancy narratives.

Attempting to date something amorphous as a hypothetical oral tradition is impossible. Suffice to say that it must pre-date Matthew (say, pre-60) and at least some of that tradition must pre-date Mark. Whether the infancy narrative was an extant oral tradition prior to Mark’s composition is impossible to determine but it would be arbitrary to rule it out.

Matthean Priority

Finally we can briefly mention those theories that posit Matthew as the earliest gospel on which the other two depend (Augustinian theory; Griesbach theory). Because Matthew contains an infancy narrative, on these theories the virgin birth forms part of the earliest known gospel tradition.

Dating Mark

It is time to draw together some implications. I have sought to explain the shared core between the Lukan and Matthean infancy narratives based upon the different theories that respond to the synoptic question. Historians would not rule out the possibility that Matthew simply made up his infancy narrative and Luke took only the core of Matthew’s infancy narrative when he made up his own, but this does not seem particularly likely. The more likely alternative is that both Matthew and Luke drew on other written sources (whether gospels or separate narratives). Yet if Matthew and Luke had separate sources, we still need to explain the shared core between these sources, and that (on most theories) requires a two stage process.

I have, for sake of argument, dated Luke c.60. I have then assigned a notional gap of five years between sources and, on that basis, argued that belief in the virgin birth must pre-date 50. These figures are all notional but take on special significance for the question of the absence of the infancy narrative in Mark. If Luke (and/or Matthew) depends on Mark then, given the goose/gander sauce equivalency, we should assign the same notional five year gap and date Mark c.55, that is, we should date the composition of Mark after the origin of the belief in the virgin birth. It may be felt that the five year gap is too meagre and that Mark should be dated earlier but, given this criteria alone, we should also date Matthew and Luke’s sources earlier. Only based on indications independent of our present considerations would it be justified to date Mark earlier; indeed, some scholars think there are such indicators and have dated Mark as early as c.45.[27]  Yet, if we date Mark earlier than we have, there is no strong reason to date Matthew as late as c.60, so the conclusion persists. Similarly, if we reject the early date for Matthew and Luke proposed above and choose to follow conventional dates (i.e. c.80), then we do not destroy the conclusion because, given a two stage source process, belief in the virgin birth can be dated as early as Mark’s gospel (i.e. c.70).

The only way[28] to escape the conclusion that belief in the virgin birth pre-dates the composition of Mark’s gospel is to suppose that Matthew and Luke, or their immediate sources, fabricated their infancy narratives without prior traditions. Historians do not rule out the possibility that Matthew fabricated his infancy narrative and then Luke changed all but the bare bones for his own (on the Farrer theory). Nor do they rule out the possibility that one author invented the virgin birth, that it was then incorporated into two separate narratives and that these narratives were copied by Matthew and Luke (on the Two/Four-Source theory). But the fact that we cannot rule these ideas out does not make them likely because either proposal imagines at least two stages of fabrication.

Nor does this hypothesized fabrication have significant explanatory power when trying to explain the origin of belief in the virgin birth. As M. Davies argues, it is difficult to explain the origin of this belief if it was not common amongst the early Christians, since it does not derive from either pagan or Jewish sources. Matthew’s text does not draw on Hellenistic biographical form of the semi-divine Greek heroes as is sometimes naively proposed.[29] Nor does Matthew base the virgin conception on the Hebrew Scriptures; “the Isaiah prophecies themselves, read in their own context, could hardly have given rise to an expectation of a miraculous conception”.[30]

Though Matthew certainly uses the Old Testament to justify his story to a Jewish audience,[31] this is certainly not its derivation. France goes further stating that the Old Testament passages are so far from prompting the Matthean narrative that “it is hard to see why they should ever have been introduced into a Christian account of Jesus’ origins”[32] unless the story was already circulating that Jesus was born of a virgin. Therefore we can provide no explanation as to why Matthew would choose to fabricate the virgin birth nor why Luke should choose to fabricate a similar story (either in ignorance of Matthew’s account or in disregard of it).

If we conclude that it looks just a little bit too unlikely that both Matthew and Luke composed their infancy narratives without prior sources then we are forced to conclude that belief in the virgin birth pre-dates Mark’s gospel and it follows that the absence of an infancy narrative in Mark’s gospel cannot be taken as evidence that believe in the virgin birth is a late invention.

Absence in Mark’s Gospel

Why then is there no infancy narrative in Mark’s gospel? One possibility we must consider is that belief in the virgin birth was not universal. After all, we have insufficient evidence from this period to guarantee the universality of all Christian doctrine in this period. Perhaps some Christians believed in the virgin birth, whilst others did not; perhaps Mark was one of the ones who didn’t. Yet he gives no hint that he is arguing against this belief, nor does he present any opposing infancy or genealogical information. He does not, as later the Ebionites would do, claim that Jesus was the son of Joseph, as we might expect were he seeking to contradict the virgin birth.

Another possibility might be that though belief in the virgin birth was extant, Mark was simply unaware of it. Perhaps he had received no tradition, either way, as to Jesus’ origins and so did not comment on this issue. Yet there is evidence within Mark’s gospel that he presupposed the virgin conception when he wrote. Had Mark received no tradition about Jesus’ origins, he would have assumed that Jesus would have had a human father (like everyone else). But Mark never mentions Joseph or any other earthly father; instead God is presented as Jesus’ father (Mark 1:11, 9:7). Further Mark records that Jesus was called ‘the son of Mary’ (Mark 6:3), something he is unlikely to have done had he assumed that Jesus had a human father. “Apart from the birth narratives of Matthew and Luke … there would seem to be no reason to reverse the normal Semitic usage and refer to Jesus as his mother’s son instead of his father’s”.[33] Evans goes further asserting that “among Jews a man was not denoted the son of his mother unless illegitimate”.[34] Though Parrinder objects that there are instances of the mother being named, he cannot provide any examples earlier than Mohammed.[35] The other explanation, that Joseph was not referred to as he was dead by this time,[36] is nullified by the numerous examples of individuals being denoted by the name of their dead father or ancestor (e.g. 1 Sam 23:6).

One further piece of evidence is Jesus’ use of Psalm 110 to confound his critics, recorded in Mark 12:35-37. He asks if David calls the Messiah ‘Lord’, how he can be David’s son. Now Mark believed that Jesus was the ‘son of David’ (Mark 10:47-48) and it would be a great departure from Jewish ideas if he did not understand that genealogically. Yet the implication of Jesus’ recorded teaching is that the Messiah is more than just the son of David.[37] This implies that Mark believed that there was something transcendent about Jesus. An appeal to the virgin conception is probably the easiest explanation.[38]

If then Mark presupposed belief in the virgin birth when he wrote his gospel, why does he not include an infancy narrative?

Firstly, we should question the expectation that Mark would include an infancy narrative. Though we might expect modern biographers to give a full life story from childhood to death, there doesn’t seem strong reason to expect Mark to follow this form. In fact, it is evident that Mark didn’t feel compelled to include details about Jesus’ infancy (miraculous or otherwise) or indeed any details about Jesus’ life prior to his baptism. From the gospel itself it is clear that Mark’s purpose is to give an account of Jesus’ ministry. So we have no reason to expect an infancy narrative in Mark’s gospel. Rather we should be more intrigued by the fact that both Matthew and Luke choose to include an infancy narrative and that, presumably, tells us something interesting about their purposes.

Secondly, if we take it as given that Mark presupposed belief in the virgin birth when he wrote his gospel, then presumably he presupposed that belief was held among his audience. After all, Mark is not writing apologetic literature, seeking to defend Christianity from criticism, nor is he writing evangelical literature, seeking to convince others of the truth of Christianity. Mark’s primary audience for his gospel is fellow Christians and so he has no need to persuade them that Jesus was born of virgin, as though that was something they did not already know.

Thirdly, we should free ourselves from the false assumption that Mark had no constraints about what he could write about. Mark did not have the freedom of a novelist because Mark was writing about (what he believed to be) real events. He could not just make stuff up; he would depend upon (direct or indirect) testimony. Rist writes:

The disciples were not eyewitnesses of the events of Jesus’ childhood … As has often been observed, this fact may be one of the reasons governing Mark’s choice of a starting point: he wanted a point where his eyewitnesses had first-hand evidence.[39]

Here it is important to note the work of R. Bauckham, who proposes that the naming of individuals within Mark’s gospel, such as Bartimaeus and Simon of Cyrene, was an intentional device to link testimonies to living witnesses.[40] More significant is his analysis of the inclusio device with biographies such as Lucian’s Alexander and Porphyry’s Life of Plotinus, and the argument for similar devices in Mark (relating to Peter) and Luke (relating to the women).[41] Using both internal evidence and the testimony of Papias, [42] Bauckham can thus defend the position that the testimony of Peter stands behind the main part of Mark’s gospel.

We do not need to retrace all his arguments here. We may simply note that (1) there is nothing implausible about Papias’ testimony, and (2) Peter stands large within the Marcan narrative. Bauckham’s thesis has a priori plausibility since the early Christian message was centred on the testimony of the apostles (cf. 1 Cor 15:1-6; Gal 1:18-19). If Mark’s gospel was largely based on the testimony of Peter then it is understandable that it doesn’t include an infancy narrative because Peter wasn’t there. This doesn’t mean Mark (or Peter for that matter) didn’t believe in the virgin birth; it only means that the story of Jesus’ birth wasn’t part of the testimony to which Mark had access.

Absence in Paul

The other consideration that leads some scholars to date belief in the virgin birth late is that, as well as not appearing explicitly in Mark’s gospel, the virgin birth is not explicitly mentioned in other early Christian writings, such as the letters of Paul. Yet the relevance of this absence is much the same as the relevance of the absence in Mark’s gospel (i.e. not great). The majority of Paul’s letter date from 55 or later, by which time belief in the virgin birth was already extant. Even the earliest letters, 1 & 2 Thessalonians, do not date before 50, which is within the window identified above. So the absence of explicit reference to the virgin birth in the letters of Paul cannot be taken as evidence that this belief was not extant. Nevertheless there is still an important question as to why Paul does not mention it. The situation with Paul is slightly different than with Mark. With Mark, who is writing narrative, the question was why Mark’s narrative does not include Jesus’ infancy. With Paul, who is not primarily writing narrative, and refers to only a few events in Jesus’ life, there is no expectation that he would include an infancy narrative. Instead, the question is why Paul, in some of his Christological passages, does not choose to mention and draw significance from Jesus’ virgin birth.

H. P. Thompson urges that the absence of explicit mention “must be carefully interpreted”. He writes “for example, the silence of the speeches of Acts points only to the conclusion that the early preaching of the Gospel concentrated on the end of Jesus’ life rather than the beginning.”[43] This is significant because Luke did believe in the virgin birth but resists the temptation to write that belief into those early speeches. Thompson takes this as evidence that Luke was a faithful historian. More to the point, it demonstrates that those writing Christology need not have felt compelled to discuss the virgin birth. Indeed, it is arguable that there is no passage in the Pauline literature that would naturally feature a mention of the virgin birth. Since Paul believes that Jesus is the Son of God and since, as argued above, it is likely that the idea of virgin birth was already current when Paul wrote, it may be that Paul simply presupposes that his audience would have the same understanding of the sonship of Jesus as his own. However, though this is possible, probable even, it is not, in itself, positive evidence that Paul did believe in the virgin birth.

In this regard, Gal 4:4 has been the key text in contention. Broadly speaking, Catholic scholars have seen here an allusion to the virgin birth whilst Protestant scholars have not. A. D. Norris argues that for Paul to conjoin the concepts of Jesus’ filial relationship with God and Jesus’ being born of a woman is a clear statement from Paul that Jesus had a human mother and a divine father (and there aren’t so many options to make that statement explicable).

“There is a breath-taking completeness about this evidence which makes one wonder, once composure has been regained, how its teaching could ever have been overlooked”.[44]

Yet the scholarly consensus seems to be that “born of a woman” does not allude specifically to Mary but is “a typical Jewish circumlocution for the human person”.[45] As evidence for this idiom J. Bligh cites Job 14:1-2, 15:14-15, 25:4-5; Ben Sira 10:18; Matt 11:11 (par. Luke 7:28);[46] J. D. G. Dunn adds 1QS 11.21; 1QH 13.14, 18.12, 13, 16; SB 3:570; [47] S. K. Williams also cites Josephus, Antiquities 7:21; 16:382,[48] though I think he is mistaken about the appearance of the idiom in these passages.

It is noticeable that in all the Greek sources cited the word for “born” is a cognate of γεννάω, the common word for birth (Job 14:1-2, 15:14-15, 25:4-5 LXX γεννητὸς γυναικὸς; Ben Sira 10:18 γεννήμασιν γυναικῶν; Matt 11:11, Luke 7:28 γεννητοῖς γυναικῶν). Paul, however, does not. In Gal 4:4 he says γενόμενον ἐκ γυναικός, using a cognate of the root word γίνομαι (“to become”). Literally, Paul says Jesus “came in existence out of a woman” (cf. “come of a woman” Darby; “made of a woman” KJV; “come of a woman” YLT). The issue is not whether γίνομαι and its cognates can refer to human birth, they can; the issue is whether Paul’s choice of words specifically deviates from the common idiom because he wants to say something different.

Usually idioms use the same words whenever they occur; it is the consistency of language that makes them idiomatic. “Born of woman” is a Jewish idiom so, I suppose, it is possible that some may have translated it into Greek in different ways. The evidence from the examples cited, though, is that cognates of γεννάω are consistently used. It seems reasonable to suppose Paul would have used those words were he using them idiomatically.

Also, though it would be acceptable for a Greek writer to use a cognate of γίνομαι to refer to a birth, it is not usual. Norris writes,

“since [Paul] makes no categorical statement of the circumstance of Jesus’s conception, he is scrupulous that no misconception shall arise through a misconstruction of words carelessly used”.[49]

He contrasts Paul’s use of a cognate of γεννάω when referring to the births of Isaac and Ishmael in Gal 4:23.[50] Paul’s deviation both from the idiom and from his own practice suggests that he is using his words purposefully. And if Paul means to distinguish Jesus’ origins from usual human birth then this may indicate that he knows something special about Jesus’ birth.

(It is worth mentioning, in passing, that γίνομαι is far too common a word to read much into its usage here. For example, it can be used both of an origin and of a change of state. We cannot from Gal 4:4 alone determine that Paul believed that Jesus was created at his birth, any more than we can rule out that Paul believed that Jesus was incarnated at his birth. This word will not help us with regards this question).

Elsewhere in Paul’s letters there are possible allusions to the virgin birth. Bligh suggests that 1 Cor 7:36-39 contains an allusion to the situation of Mary and Joseph.[51] Norris also argues that Rom 1:3-4 hints at the virgin birth with its presentation of the dual descent of Jesus from David and from the Spirit. [52]

Wide Acceptance of the Virgin Conception

One strong argument in favour of the idea that the belief in the virgin conception was primitive is the relative ease with which the virgin conception was accepted so widely. We have seen how the virgin conception narrative existed before the gospels of Matthew and Luke, and it is obvious that this belief found acceptance amongst the circles that produced these gospels. But we know that Matthew and Luke’s gospels were accepted widely throughout the first century churches and there is no evidence of them being rejected or treated as suspect (until we come to the Gnostics and the Ebionites of the second century). The community that produced the Didache accepted Matthew’s gospel, so though it does not mention the virgin conception it is likely that they accepted it. The Johannine phrase “only-begotten” also implies a belief in the virgin conception,[53] especially since John connects it with the phrase “made flesh”.[54] The derisive tale that Jesus was an illegitimate child that arose in amongst Jews in the latter half of the first century[55] is strong evidence that the virgin conception was being preached by Christians at this time. The (almost creedal) statements of Ignatius[56] and Aristides[57] show that by the early second century the belief in the virgin conception was widely accepted amongst Christians. The elaborations of the apocryphal infancy gospels of the second century (e.g. Protevangelim of James, Infancy Gospel of Thomas) also demonstrate that this belief was common.[58]

Given wide acceptance of this belief in the latter first and early second centuries, it is reasonable to suppose that this belief was primitive – or, at very least, any prior belief about Jesus’ origins was ambiguous, under-stated or not positively opposed to the virgin conception. The fact that both Mark and Paul seem to presuppose this belief is a strong indication of its primacy.

How Early Is Belief in the Virgin Birth?

So far in this essay I have been arguing conservatively from the broadly accepted evidence (i.e. the gospels of Matthew and Luke) against a broadly accepted conclusion (i.e. that belief in the virgin birth is late). On this basis I have, rather coyly, concluded that belief in the virgin birth must pre-date 50. Yet I don’t for a moment suppose that belief in the virgin birth originated in the late 40s. Despite the relative silence of Mark and of Paul, I think we can be confident (as argued above) that belief in the virgin birth pre-dated both these writers and so we have no particular reason to doubt that it dates back to the earliest days of Christianity. I argued above that the birth narratives of both Matthew and Luke are best explained by their dependence on sources. Elsewhere, I have argued that these sources were the testimonies of Mary and Joseph, perhaps mediated through others to the evangelists.[59] If this is the case then belief in the virgin birth predates the virgin birth.

Jesus seems to have understood his own intimacy with God as transcending even that of the radical intimacy he teaches to his followers. This is most clearly displayed in the abba prayers of Jesus, as indicated by Jeremias. Though the Aramaic word abba is only found once in the gospels (cf. Mark 14:36), it probably underlies the vocative pater in the other prayers of Jesus. There is some evidence that God was previously addressed as ‘abi but abba seems unique to Jesus and the early Christians (cf. Rom 8:14; Gal 4:6).[60] The fact that followers of Jesus did use abba to address God in prayer demonstrates of itself abba has “no exclusively Christological weight”.[61] It does, however, indicate a special intimate relationship between Jesus and God, which led him to address only one person as father/abba (Matt 23:9; cf. Mark 3:31-35). Also, as many commentators have noted, Jesus says “my father” and “your father” but never “our father”.^[62]

The filial consciousness of Jesus is demonstrated elsewhere in the gospel tradition. The parable of vinedressers portrays Jesus as the son and heir of the vineyard (Israel). This parable is multiply attested (Mark 12:1-9; Matt 21:33-41; Luke 20:9-16; GTh 65). Even the Jesus Seminar proposes that some form of the parable goes back to Jesus and Patterson’s reconstruction of the parable includes contrast between the servants and the son. [63] Perhaps more significant is the saying of Jesus recorded in Matt 11:27 / Luke 10:22. The authenticity of this Q-saying has been queried because of the radical claims it makes. Some have objected that the self-ascription of “Son” by Jesus is uncharacteristic, though for our purposes this argument is circular. The major objection is that this saying “has a distinctively Johannine ring”.[64]

Witherington has argued that since this passage is unlikely to derive from Johannine material, parallel passages in John can be taken as independent witness of this tradition (cf. John 5:19-20, 7:27-29).[65] The idea that this verse is an early Christian expansion or commentary on the preceding verse,[66] do not account for these Johannine parallels and it is not clear in what textual cues could have led to such an expansion. In favour of the authenticity of this saying is the probability that it derives from an Aramaic original and that it contains the potentially embarrassing implication that the Son is unknowable. Accepting it as authentic, Jesus is claiming not only to be heir of “all things” but also to stand in a unique relationship with God whereby he alone can truly reveal God to mankind. “In short, Jesus saw himself as the unique mediator of the final revelation of God, and thus God’s unique Son”.[67]

All this falls short of an explicit statement from Jesus that he believed that he was born of a virgin and was thus literally the Son of God. Nevertheless Jesus does claim to stand in a unique relationship with God, from which he derives significant knowledge and authority. One may reasonably question whether such a self-understanding could arise purely from messianic expectations and if not then we must allow the possibility that Jesus himself believed in some more special form of sonship.