Why Are The Infancy Narratives So Different?

Robert Stein lists eleven points of agreement between the infancy narratives of Matthew and Luke (The New American Commentary: Luke [Nashville, Tennessee: Broadman Press, 2003], n. 9 on pp. 70-71). Darrell Bock lists nineteen (Luke, Volume 1, 1:1-9:50 [Grand Rapids, Michigan: Baker Books, 1994], pp. 71-72). We can arrive at a lower or higher number, depending on how many details we include. Critics make much of the differences between the two gospels, but there's a lot of overlap, and the differences can be harmonized:

"But the stories [in the infancy narratives] have long been shown to be compatible (Machen), even mutually complementary. Moreover literary independence of Matthew and Luke at this point does not demand the conclusion that the two evangelists were ignorant of the other's content. Yet if they were, their differences suggest to some the strength of mutual compatibility without collusion. Matthew focuses largely on Joseph, Luke on Mary. R.E. Brown (Birth of Messiah, p. 35) does not accept this because he finds it inconceivable that Joseph could have told his story without mentioning the Annunciation or that Mary could have passed on her story without mentioning the flight to Egypt. True enough, though it does not follow that the evangelists were bound to include all they knew. It is hard to imagine how the Annunciation would have fit in very well with Matthew's themes. Moreover we have already observed that Matthew was prepared to omit things he knew in order to present his chosen themes coherently and concisely." (D.A. Carson, The Expositor's Bible Commentary: Matthew, Chapters 1 Through 12 [Grand Rapids, Michigan: Zondervan, 1995], p. 71)

Mark would have known of some of Jesus' resurrection appearances (Mark 16:7, 1 Corinthians 15:5-8), yet he chose not to include them in his gospel. John's gospel closes with a reference to other material that could have been included, but wasn't (John 21:25). Matthew and Luke, covering months or years of events surrounding Jesus' birth, and doing so with different sources and different objectives, can plausibly differ to a significant degree. One of Luke's objectives in writing his gospel was to supplement what was already available (Luke 1:1-4).

As I've mentioned in previous articles (here and here), the gospels of Matthew and Luke were widely accepted and interpreted as giving harmonious historical accounts early on. A small minority of sources, such as Marcion and his followers, would reject one or both of the two gospels on other grounds (not because of inconsistencies in the infancy accounts), but the large majority of professing Christians accepted both. Not only were Matthew and Luke accepted across the Christian mainstream, but Irenaeus comments that they were widely accepted among heretical groups as well (Against Heresies, 3:11:7, 3:12:12).

It can't be argued that the early Christians accepted both gospels and interpreted them as harmonious historical accounts because they were unaware of the differences. Gospel harmonies such as Tatian's wouldn't have existed if there was no awareness of the differences among the gospels. Eusebius refers to how "every believer" offered an explanation for the differences between the genealogies of Jesus in Matthew and Luke (Church History, 1:7:1). The early Christians were willing to dispute the canonicity of books like 2 Peter and Revelation, so it doesn't seem that they accepted the canonical books, like the gospels, uncritically. If Matthew and Luke were giving radically contradictory accounts of the events surrounding Jesus' infancy, as modern critics often allege, the earliest Christians, including contemporaries of the apostles and people who lived just after their time, don't seem to have been aware of it. There doesn't seem to have been any separation between a Matthean community and a Lukan community that's left any traces in the historical record. Rather, the two gospels were accepted together as harmonious from as far back as we can trace the issue. Given Luke's apostolic ties and his positive portrayal of the apostles and other early church leaders in Luke and Acts, it makes sense that he and Matthew would hold similar views.

To keep the differences between the infancy narratives in perspective, let's consider what sort of contradictions could have existed. If the gospel authors were as unconcerned with evidence and as isolated from other Christians as modern critics often suggest, then why don't we see something like one gospel describing a virgin birth while the other describes conception by sexual intercourse? Why doesn't one gospel have Jesus come to earth as a grown man around 20 B.C. while the other has Jesus born into a family around 10 A.D.? Why do they even mention Nazareth, if the gospel authors were as unconcerned with historical accuracy as some critics suggest? Why not just have Jesus in Bethlehem throughout His pre-ministry life? Etc. More reasonable critics will acknowledge that there's a significant core of agreement between Matthew and Luke and that there was some significant concern about historical accuracy by both authors.

Harmonization is commonplace in historical research. Different sources will often focus on different issues. A biographer of John F. Kennedy who wants to focus on Kennedy's foreign policy will differ significantly from another biographer who's focusing on Kennedy's influence on the civil rights movement. When discussing the role of religion in the founding of America, an atheist web site might focus on the anti-religious and unorthodox sentiments of men like George Washington and Thomas Jefferson, while a Christian web site might focus on their pro-religious and orthodox beliefs. Most of the time, both representations of the men are accurate, but in different contexts. The same man who speaks negatively of religion in the context of discussing state churches might speak positively of the influence of religious principles on society. Quotes that may seem significantly different on the surface can be harmonized upon closer analysis.

Matthew was a Jew writing largely for a Jewish audience, whereas Luke was a Gentile writing largely for a Gentile audience. Matthew begins his genealogy with Abraham, the father of the Jewish people, whereas Luke traces Jesus' genealogy back to Adam. That's a difference, but not a contradiction, and it's a difference on an issue about which all would agree that both authors knew the relevant facts. Nobody would argue that Matthew was unaware that the human race can be traced back further than Abraham. Similarly, Matthew and Luke would have known that Joseph and Mary both needed an explanation for Mary's pregnancy. Matthew chooses to focus on how Joseph came to the conclusion that God was involved, and Luke focuses on Mary. But it would be ridiculous to suggest that Matthew thought that Mary didn't need any explanation and that she never discussed the issue with Joseph or anybody else. Matthew had to have known about and thought about more than what he writes about. Matthew would have known that Jesus was circumcised and that He and His parents went through other traditional Jewish practices, even though he doesn't discuss such issues as Luke does. Similarly, Luke decides to discuss John the Baptist's background and to parallel it with the background of Jesus, whereas Matthew doesn't. Yet, as the remainder of Matthew's gospel shows, Matthew was aware that Jesus and John were around the same age and had close ties (Matthew 3:14). There are many events and themes surrounding the infancy of Jesus, and it makes sense for two authors to make some different choices about what to address and in what manner. If they could differ when discussing issues where both are known to have been aware of the relevant facts, such as in the examples discussed above, then we should be cautious in reaching the conclusion that they contradicted each other in areas where we have less information.

To make an accurate assessment of the consistency of the birth narratives, we have to know what to compare. Instead of comparing Matthew 1-2 to Luke 1-2 and the genealogy in Luke 3, it makes more sense to compare Matthew 1 to Luke's material. In other words, the events of Matthew 2 seem to postdate the events Luke addresses. Matthew 2:16 suggests that a significant amount of time may have passed ("two years old and under"). Luke is concerned with addressing the family's submission to the Jewish law, and so he concludes his account of Jesus' infancy there and mentions that the family settled in Nazareth afterward (Luke 2:39). If the proper comparison is primarily between Matthew 1 and Luke's material (Matthew 2 has some relevance, but is from a later timeframe), then there isn't much to explain. The material in Matthew 1 is of a general nature and is consistent with what Luke records. The most significant difference would be between the genealogies, and it was common practice in ancient times to arrange genealogies according to numerical patterns, to follow legal rather than biological relationships, to leave out some generations, etc.:

"In general, people map out genealogies to accomplish political, social, or religious purposes such as to explain how peoples are related to one another or to legitimate a political union of groups and peoples or to authenticate leadership status for certain families. They often begin with a common ancestor, finish with the immediate ancestors of the person making the genealogy, and are abbreviated or have variations in the middle." (Anthony J. Saldarini, in James D.G. Dunn and John W. Rogerson, ed., Eerdmans Commentary On The Bible [Grand Rapids, Michigan: Eerdmans, 2003], p. 1006)

There are multiple ways to harmonize the genealogies of Matthew and Luke (see, for example, here and Steve Hays' comments on pages 155-158 of his e-book This Joyful Eastertide). The primary issue here, when considering historicity, ought to be the upfront plausibility of the gospel writers’ composing reliable genealogies. We know that at least some sources kept genealogical records during the timeframe in question, and we know that Matthew and Luke were early sources living at a time when many relatives of Jesus were still alive and in prominent places in the church. The availability of such sources is more significant than the differences between the genealogies, since it was commonplace for genealogies to be rendered in different ways according to different principles.

What about the geographical setting of Matthew 1? Critics make much of the reference to Bethlehem in Matthew 2:1 by suggesting that it justifies an assumption that everything that occurred in chapter 1 occurred in Bethlehem, which would contradict Luke's account. But Matthew 2:1 only places the birth itself there, which agrees with Luke. Rather than creating a problem, Matthew 2:1 is further evidence of agreement between the two accounts, particularly given the fact that Matthew refrains from mentioning a location earlier:

"In the larger shape of Matthew's narrative it is a little surprising to have the location and timing of Jesus' birth introduced for the first time here [in Matthew 2:1] and not at the point where Mary, Joseph, and Jesus first enter the narrative...Matthew offers no comment on how it is that the birth happened to take place in Bethlehem." (John Nolland, The Gospel Of Matthew [Grand Rapids, Michigan: Eerdmans, 2005], p. 107)

Contrast Matthew's initial silence on the location of the events with his emphasis on locations in chapter 2 (verses 1, 3, 5-6, 8, 13-16, 20-23). Matthew may have had more information on locations for the later events, so that he refrained from commenting on locations earlier, associating Bethlehem with the birth itself, not the earlier events.

Issues like these have been addressed in many places by conservative scholarship, such as in Craig Keener's commentary on Matthew (Grand Rapids, Michigan: Eerdmans, 1999) and Darrell Bock's commentary on Luke (Grand Rapids, Michigan: Baker Books, 1994). Even if we were to conclude that there are some inconsistencies between the accounts in Matthew and Luke, we'd still have to address the historicity of the material. The accounts could have some errors and inconsistencies, yet still be largely historical. Many people, such as Raymond Brown, have accepted some significant portions of the infancy narratives as historically plausible or probable while rejecting inerrancy. Some people accept the historicity of something like the virgin birth or Jesus' resurrection without believing in the inerrancy of scripture. As I've said before, though I don't consider such a position the best option, it is more reasonable than the radical skepticism of naturalists and some other critics of Christianity.

However, we don't have any good reason to conclude that the infancy accounts are inconsistent. As the large number of events discussed in both gospels illustrates, and as the timeframe referred to in Matthew 2:16 suggests, Matthew and Luke would have had many issues, themes, and angles to choose from in compiling an account. They agree in the general outlines of their narratives, and they differ in much of what they include and don't include. The differences aren't sufficient to justify the conclusion that either account is erroneous, much less that both accounts are as unhistorical as some critics claim.