Luke 1 & 2 Infancy Narratives as a Later Interpolation

Luke 1 & 2 Infancy Narratives as a Later Interpolation December 21, 2020

Luke 3:1 opens with an elaborate chronological statement: “In the fifteenth year of the reign of Tiberius, when Pontius Pilate was … the word of the Lord came to John the son of Zacharias in the wilderness”. This surely reads as if it was originally written as the opening section of a book. The impression is strengthened by the curious position of the genealogy of our Lord (3:23). If this had been inserted by the last editor of the Gospel, we should have expected to find it, like the genealogy in Matthew, somewhere in chapters 1 or 2 in connection with the account of the Birth and Infancy. If, however, it was originally inserted in a book which only began with Luke 3:1, its position is explained – for it occurs immediately after the first mention of the name of Jesus.

– Burnett Hillman Streeter, The Four Gospels: A Study of Origins, 1924, p.209

In this piece, I am going to support my claim that Luke 1 & 2 appear to be a later interpolation and weren’t a part of the original Gospel. There is also a theory that Matthew did this too, but I will concentrate on Luke today.

As the great Catholic exegete Raymond Brown (The Birth of the Messiah: A Commentary of the Infancy Narratives in the Gospels of Matthew and Luke, p. 33) states: “Indeed, the body of the Gospel shows that the people among whom Jesus had been reared knew nothing about an extraordinary infancy (Matt 13:53-58; Luke 4:31-32, 36-37).” There is something afoot here and it needs explaining.

What Are the Implications?

The question, if one accepts some kind of interpolation, is whether it was Luke himself who wrote Luke 1 and 2 after Luke-Acts for some reason or whether the interpolation came from a third party. I am ambivalent on this since I think it leaves us with the same conclusion: that the infancy was not deemed important enough to originally include, and was later included in answer to some challenge or need, probably with theology in mind. This leads to giving it some historical doubt since it appears symbolic or theological in nature.

Raymond Brown (for example) believes the two chapters are later interpolations but by Luke himself (that they are more reflective of Acts – p. 243), though he admits there are some issues with this approach as opposed to third parties (discussed on p. 241). If you don’t have access to The Birth of the Messiah (essential reading if you are interested in the infancy narratives), then you can always read Rethinking the Western Non-interpolations: A Case For Luke Re-editing His Gospel by Giuseppe Capuana, a thorough analysis that can be read here, concluding that it was Luke himself who did it. I’m not so sure, as we shall see.

What this means is that the birth narratives were added in as an afterthought or in reaction to some perceived need or desire brought about by some external conflict or issue. This leads to confirmation that the birth narratives are not historically reliable.

It is not controversial that the Gospels were written backwards. That is, to think of the most important aspect of Jesus’ life in terms of his followers and contemporaries (for those who believe he existed) is to think of his death and supposed resurrection. His followers and sources would have known him as an adult and so the recounts of his adult life would have been front and centre of the evangelists’ thoughts and writings, and the thoughts of their sources. It is why there are barely any childhood accounts (other than the noncanonical Infancy Gospel of Thomas). All of Jesus’ followers would have been obsessed with his death after his crucifixion, and the developing stories would have concerned the man they knew, not the child they didn’t.

The birth accounts were not known, I would posit, and not theologically necessary, until there became a need to retrospectively add them.

At this point, I think Matthew and Luke (or whoever added the narratives) drew on a developing idea of a virgin birth and a Bethlehem birth and wrote those two aspects in separately to their infancy narrative. This coherently explains the differences, as are the issues and differences between Luke 1 and 2 and the rest of his Gospel; and the only two commonalities – a virgin birth in Bethlehem – are seen in both Matthew and Luke because this was a developing or received idea. Nothing else is the same and, indeed, the two Gospels even contradict.

They were absent in the early Lukan Marcionite manuscripts, and a note in a medieval Armenian copy of Ephraem’s Commentary on the Diatessaron led F.C. Conybeare to believe that Ephraem had access to an old manuscript without the infancy narrative. As you will see, there are plenty of good reasons to believe Luke 1 and 2 were later additions, and you will be in good scholastic company for believing such.


Since the first few chapters are missing from the Marcionite versions of Luke, some of the earliest manuscripts we have, there has been much debate as to how this came to be (some claim the Evangelion – Marcion’s Luke – is what the canonical Luke is based on or even the first Gospel). Marcion and his followers essentially saw Luke’s Gospel as their Bible. They also didn’t see Jesus as fully man. This led early Church Fathers to assert that it wasn’t that the chapters were not originally written but that Marcionites lopped them off. Bart Ehrman sees it as more likely that (see his article for a discussion of the Marcionite docetist views)…:

Marcion didn’t know about those chapters because they were not in the version of Luke that he knew. And that’s because Luke didn’t originally have those chapters.

This is in the context of a textual variant of Luke 3:23:

I have been discussing the intriguing textual variant found in Luke 3:23, where Jesus is said to be baptized. When he comes out of the water the heavens open up, the Spirit descends upon him in the form of a dove, and voice then comes from heaven. But what does the voice say? In most manuscripts the voice says exactly what it does in Mark’s Gospel: “You are my beloved Son; in you I am well pleased.” But in a few ancient witnesses it says something slightly but significantly different: “You are my Son, today I have begotten you” (or: “given you birth”).

I am arguing that the latter may in fact be the original text of Luke, but that it was changed by scribes who were alert to the problems it posed. But if that’s what the voice said, then doesn’t that indicate that it was at that moment (Note: “Today”!) that Jesus became the Son of God?

There are some fascinating reasons as to why and how Luke’s final Gospel built on the Evangelion (source blockquote in bold):

Firstly, there are many passages found in Marcion’s gospel that seem to contradict his own theology, which is unexpected if Marcion was simply removing passages from Luke that he didn’t agree with. Matthias Klinghardt has argued:

The main argument against the traditional view of Luke’s priority to [Marcion] relies on the lack of consequence of his redaction: Marcion presumably had theological reasons for the alterations in “his” gospel which implies that he pursued an editorial concept. This, however, cannot be detected. On the contrary, all the major ancient sources give an account of Marcion’s text, because they specifically intend to refute him on the ground of his own gospel. Therefore, Tertullian concludes his treatment of [Marcion]: “I am sorry for you, Marcion: your labour has been in vain. Even in your gospel Christ Jesus is mine” (4.43.9).[15]:7 (emphasis in original)

Secondly, Marcion himself claimed that the gospel he used was original, whereas the canonical Luke was a falsification.[15]:8 The accusations of adulteration are therefore mutual:

Tertullian, Epiphanius, and other ancient witnesses, all of whom knew and accepted the same Gospel of Luke we know, felt not the slightest doubt that the “heretic” had shortened and “mutilated” the canonical Gospel; and on the other hand, there is every indication that the Marcionites denied this charge and accused the more conservative churches of having falsified and corrupted the true Gospel which they alone possessed in its purity. These claims are precisely what we would have expected from the two rival camps, and neither set of them deserves much consideration.[6]:78

Thirdly, John Knox and Joseph Tyson have shown that, of the material that is omitted from Marcion’s gospel but included in canonical Luke, the vast majority (79.5-87.2%) is unique to Luke, with no parallel in the earlier gospels of Mark and Matthew.[6]:109 [16]:87 They argue that this result is entirely expected if canonical Luke is the result of adding new material to Marcion’s gospel or its source, but that it is very much unexpected if Marcion removed material from Luke.

The Baptism of Jesus

Mark and John start their Gospels with the baptism of Jesus and so for Luke to have originally done this is actually expected and how it pretty much starts without 1 and 2.

Raymond Brown sees this as compelling:

Granted that the birth material [of Luke 1-2] had an origin and transmission different from the stories of Jesus’s ministry, how did the evangelist proceed in joining birth material to the story of the ministry? Did he begin writing with the birth stories, or did he begin with the account of the ministry and, as an afterthought, prefix the birth stories? [The] evidence points in [the latter] direction. Although there have been occasional attempts to join the infancy story to the next two chapters, so that a continuous narrative-unit of the Gospel would extend from 1:5 to 4:15, the solemn beginning of the ministry in 3:1-2 could well have served as the original opening of the Lukan Gospel. Support for this is found not only in the fact that Mark and John open the Gospel story with the events surrounding the baptism of Jesus, but also in the reference to this baptism by John the Baptist as a beginning in Acts 1:22 (the latter passage suggests that the infancy narrative may have been prefixed to the Gospel after the Book of Acts was completed). The placing of the genealogy in the third chapter of Luke makes more sense if that had been done before an infancy narrative had been prefixed. As was true also with Matthew’s Gospel, none of the Lukan infancy narrative has had major influence on the body of the Gospel, so that, if the first two chapters had been lost, we could never have suspected their existence.

– Raymond E. Brown, 1977, p.239-240

The moment of baptism is also where Jesus becomes the Messiah later on in Luke. However, the birth account has Jesus becoming the Messiah then, as Bart Ehrman explains in another piece:

First Messiah. In Luke 2, right after Jesus is born, the angel of the Lord appears to the Shepherds to announce to them that in Bethlehem has been born the Christ (2:11). OK, so he’s the Christ/messiah at his birth, right? But in one of the speeches in Acts, Peter indicates that God “anointed” Jesus (that is, made him the “anointed one”—the literal meaning of “messiah”; same Greek word) at his baptism (10:37-38). And in another speech he explicitly states that Jesus became the Christ at his resurrection (2:36: by raising Jesus “God made him the Christ” after his crucifixion). Huh? How can that be? Luke is simply not interested in terminological consistency. Each of these key moments is when Jesus became the Christ.

So too with the term “Lord.” The angel tells the shepherds that Jesus was “born” as the Lord (2:11). And in Luke 10:1 Jesus is designated as the Lord during his public ministry. But in Acts 2:36 we are told that Jesus became the Lord when he was raised from the dead.

So too with “Savior.” Jesus is said to have been born as the Savior in Luke 2:11; and during his lifetime he is said to have been the Savior in Acts 13:23-24. But according to the speech in Acts 5:31, he was made the Savior at the resurrection.

This is a real problem and one that is perfectly solved by seeing Luke 1 and 2 as later interpolations. I would also argue, contra-Brown, that this better points toward a third party interpolation rather than this being Luke.

Theological Differences

Many differences, as we can see from the previous sections, are theological (and much has been covered there). The great Lukan analyst H. Conzelmann “virtually ignores the infancy narrative, for he has found it different from and even contrary to the main thought of Luke/Acts” (Brown, p. 241). The schema that he proposes that Luke used to structure his Gospel (the period of Israel, the Satan-free period of Jesus, the period of the Church and the Spirit) illustrates how problematic Luke’s infancy narratives are in terms of theological coherence; as Brown states, (p.242): “One can understand how poorly the infancy narrative fits into this schema…”

The Virgin Birth; No Internal or External Corroboration

Luke’s infancy narratives fit in very thematically to the non-canonical gospels that are seen to be even greater embellishments of the Jesus story. The Gospel of James was structurally and content-wise very like Luke, but then including the magi and the slaughter of the innocents at the end. It was a pretty popular (non-canonical) Gospel. The Infancy Gospel of Thomas goes on to include stories of Jesus’ childhood that we see nowhere in the canon.

The virgin birth seems to be a later development, as we can see with these, and is conspicuously absent from Paul and Mark, as well as John. As far as Matthew is concerned, there are Church Fathers who tell us that the Gospel of Matthew’s infancy is absent from early manuscripts, too. Given the silence from all other earlier documents, and from later canonical ones, in terms of a virgin birth, we can see that the interpolation thesis fits in perfectly. There simply was no known virgin birth story in other sources (Matthew excepted, but they are not coherent).

Not only is there no sign of this birth narrative in Paul, Mark or John, and Matthew fundamentally disagrees on what happened (other than virgin birth and Bethlehem) but, worse still, Luke doesn’t mention it again, either. As Ehrman states in a third article:

Moreover, it is worth pointing out that nowhere else in Luke’s Gospel does he refer back to the miraculous birth of Jesus; never later does he say anything about his mother being a virgin or allude to the spectacular events surrounding his birth.

Nor is there any reference in the book of Acts to any of the stories in these opening two chapters, even though Luke would have had ample opportunity to refer to them, especially in the speeches of the apostles in Acts, when they talk about what an amazing savior Jesus was.

Linguistic Analysis

I refer you back to the detailed analysis by Capuana mentioned earlier.

From a linguistic point of view, Luke 1:5-2:52 stands in great difference to the rest of Luke(-Acts). Even Craig Blomberg is aware of this:

Abruptly, with 1:5, Luke adopts a very Semitic form of Greek writing. From chapter 3 on, he uses standard koiné, though with a bit more literary artistry than the other evangelists, but not as elegantly as the preface or as Hebraic in style as the rest of [chapters 1-2]

– Craig Blomberg, Jesus and the Gospels: Introduction and Survey, p.202

W. G. Kümmel, states:

…the few linguistic deviations between Lk 1–2 and Lk 3–24 can perhaps be traced back…to the fact that the prehistory [that is, the Infancy Narrative] was written later than the rest of the Gospel.

–  Introduction to the New Testament, trans. Howard Clark Kee, 1975, p. 1370)

And Ehrman agrees:

That he may have done so is suggested by the fact that the style and emphases of chs. 1-2 are different from the rest of the Gospel. In particular, these opening chapters are written in a style very much reminiscent of what you find in the Greek Old Testament (the “Septuagint” as it is called). The rest of Luke’s book is not. It does make sense that chs. 1-2 would be. For the author of these chapters, Jesus’ birth is in close continuity with the earlier acts of God as found in Scripture, so much so that one can easily draw parallels between Jesus’ miraculous birth and the miraculous births of famous OT figures (especially, for example, the prophet Samuel).

Raymond Brown discusses the linguistic differences on p. 244-245 in some depth.

Chapter 3 Looks Like the Start

If you take out the opening two chapters, Luke 3 actually looks like the start of the Gospel. This is pretty compelling:

Now in the fifteenth year of the reign of Tiberius Caesar, when Pontius Pilate was governor of Judea, and Herod was tetrarch of Galilee and his brother Philip was tetrarch of the region of Ituraea and Trachonitis, and Lysanias was tetrarch of Abilene, in the high priesthood of Annas and Caiaphas, the word of God came to John, the son of Zechariah, in the wilderness. And he came into all the region around the Jordan, preaching a baptism of repentance [a]for the forgiveness of sins; as it is written in the book of the words of Isaiah the prophet:

It really does read like the start of the Gospel as should be:

The beginning of Chapter 3 reads like the beginning of a bios-history / gospel: “In the fifteenth year of the reign of Emperor Tiberius, when Pontius Pilate was governor of Judea… ” etc., and is shortly followed by Jesus’s baptism and genealogy, etc. From the canonical point of view, it seems an odd place to insert Jesus’s genealogy – at the point of his adult baptism and after he has already “grown in wisdom and divine favor” (2:52) instead of at his conception or birth where genealogies would normally be placed. Indeed, one has to question the relevance and coherency of a genealogy that traces Jesus’s lineage through his father’s side if Jesus did not have a human father as is stated in chapters 1-2. Of course none of this would be odd if the gospel that underpinned Luke originally began at Chapter 3 – right at the point where the gospel sinks back into standard koiné style, and where we know other early versions started, and where the silence of further allusion to the virgin birth begins. Is this just all coincidence?

Again, Ehrman agrees with this element, too:

Even more striking, 3:1 reads very much like the *beginning* of a narrative, not the continuation of a narrative. The author dates the events he is about to narrate, which would normally happen when he begins his account.

What is yet more striking, the *whole* of chapter 3 makes better sense as the first chapter of a book, not a third. That is because right after Jesus’ baptism (an account very similar to Mark’s opening chapter), we find a genealogy of Jesus (an account not found in Mark). Now why would you give a person’s genealogy NOT at the beginning of your biography, at the time of his birth (which is where one could expect to find a genealogy; see, e.g., Matthew 1), but in a later chapter, at the time of his baptism???

If Luke 3 *began* this Gospel though (that is, if it originally did not have chs. 1-2), then the placement of the genealogy makes perfect sense – especially if the words spoken by God from heaven are a quotation of Psalm 2:7, as I’m proposing “You are my son; today I have begotten you.” In that case, the Gospel begins by providing a dating for the events it is about to narrate, it opens with an account of Jesus’ baptism by John (as in Mark’s Gospel), that account ends with God indicating that this is the day on which he has “given birth to” (or “begotten”) Jesus, and then the author launches directly into the genealogy of Jesus’ birth.

The Nazareth Synagogue

One other conundrum solved by the chapters being interpolations is as follows:

If this is how an earlier form of the gospel once stood, this would actually prevent a further little oddity found within canonical Luke. In the canonical version, Jesus mocks the Nazareth synagogue attendees by suggesting they will want to see a miracle sign like the ones they have heard he performed in Capernaum (4:23). But at this point in canonical Luke, Jesus hasn’t performed any miracles in Capernaum yet. The miracles he performs in Capernaum occur AFTER these verses (4:31-41). If the Nazareth episode (4:16-30) wasn’t part of the gospel (as the Church Fathers attest was the case for the Marcionite version) then there is no oddity. It seems to me that the oddity was created unintentionally by a careless splicing together of sources. The redactor wants to have Jesus’s ministry expand from Nazareth (Jesus’s home town) to Capernaum, to Jerusalem, and ultimately to the rest of the world. However in constructing it this way, by splicing in a miracle narrative that references Jesus’s earlier Capernaum deeds before he even goes to Capernaum, he makes a little error that gives us a hint of the underlying redactional processes.


I think that the evidence presented here, and in-depth elsewhere, provides a really good, coherent case as to why there are very many differences between the infancy narratives and the rest of Luke, differences that simply can’t be explained by “Luke wrote the whole thing as was”. It also better explains the differences between Luke and Matthew.

As such, I would argue, this also lends to the case that the infancy narratives are not historical and were later-developed embellishments.

Please grab a copy of my book The Nativity: A Critical Examination [UK].

Stay in touch! Like A Tippling Philosopher on Facebook:

A Tippling Philosopher

You can also buy me a cuppa. Or buy some of my awesome ATP merchandise! Please… It justifies me continuing to do this!

Browse Our Archives