The Birth Narratives

Mark begins with a grown man walking out of Galilee to the Jordan. John begins before creation. Matthew and Luke begin with a birth, and they do it so differently that the difference is itself the first thing to teach.

Matthew opens with a list: 'This is the record of the genealogy of Jesus Christ, the son of David, the son of Abraham:' Those two names are the argument. Son of David means the promise of 2 Samuel 7 has an heir. Son of Abraham means the promise of Genesis 12 has a destination. Matthew then tells the story almost entirely from Joseph's side. Joseph is the one who receives all four of the dreams given to him, the one who decides, the one who moves the family. Mary does not speak a word in Matthew's account.

Luke opens with a formal preface addressed to a patron, then drops into a world that sounds like the Septuagint: an elderly childless couple, a priest at the altar, an angel, a barren woman who conceives. Every note is borrowed from Genesis and 1 Samuel. Luke tells the story from the women's side. Mary speaks, Elizabeth speaks, and Mary's song in 1:46-55 is a political poem about thrones brought down and the hungry filled. Luke's Jesus is announced to shepherds, who in the social ranking of the day were nobody in particular.

So the two writers are not producing two attempts at the same paragraph. They are making two arguments, each with a different set of readers in view, and each choosing the material that carries its argument. That is how ancient biography worked, and noticing it early saves a student a great deal of unnecessary anxiety later.

One more thing to notice before the difficulties. Both accounts have the conception attributed to the Holy Spirit, both have Bethlehem, both have Nazareth, both have Herod the Great on the throne, both have the name given before the birth, and both trace the royal line through Joseph while denying that Joseph is the biological father. Two independent-looking traditions agreeing on that cluster is not nothing, and it is worth saying so before the columns of difference are drawn up.

Here is the difficulty, stated without cushioning. Matthew 1:6-16 runs from David through Solomon and the kings of Judah down to 'Jacob the father of Joseph, the husband of Mary'. Luke 3:23-31 runs backwards from 'the son of Joseph, the son of Heli' up through a line that leaves David not by Solomon but by Nathan. The lists share Shealtiel and Zerubbabel in the exilic stretch, and otherwise, between David and Joseph, they are largely different names. Joseph has two fathers. This is not a modern discovery. Julius Africanus was writing about it in the early third century, and Eusebius preserved his answer in Ecclesiastical History 1.7. Four proposals are worth a student's time. The first is levirate marriage, which is Africanus' own solution: Jacob and Heli were half-brothers with the same mother, one died childless, the other raised up a son for him, so that Joseph had a legal father and a biological father with different names. This has the merit of being the oldest answer on record and of using a legal institution that actually existed. Its weakness is that it needs a specific chain of family circumstances for which there is no evidence beyond the problem it was invented to solve.

The second is that Luke gives Mary's line, with Heli as Mary's father and Joseph entered in the register as his son-in-law. This is defended vigorously in a good deal of modern evangelical commentary and it has an obvious attraction: it gives Jesus a biological descent from David through his mother. Its weakness is that Luke's text says Joseph, not Mary, and the qualification Luke does insert, that Jesus 'was regarded as the son of Joseph', is attached to Jesus' relation to Joseph rather than to Joseph's relation to Heli. You have to supply the move the text does not make.

The third is that Matthew gives the legal line of royal succession, the throne of David passing from one heir to the next, while Luke gives actual physical descent. Adoption, which was ordinary in the ancient world, would then explain how the two lines converge on one man. This handles the Solomon-Nathan divergence elegantly, because a throne line and a family line need not be the same list. Notice that it also concedes something: on this reading, at least one of the two lists is not a biological record, which is a step most readers have not consciously taken.

The fourth proposal is the one that asks a different question. It holds that the two lists are doing different things, and that reconciling them name by name is a demand neither writer invites. Matthew's structure is openly artificial: he claims three sets of fourteen in 1:17, and a straight count of the last set comes out at thirteen on the usual reckoning, and he has skipped Ahaziah, Joash and Amaziah between Joram and Uzziah, and Jehoiakim later on. He has trimmed a known list to a shape. The shape is likely deliberate, and the most common explanation is that the numerical value of the consonants of David's name in Hebrew is fourteen. Luke's structure is equally purposeful: he runs past Abraham to Adam and ends 'the son of God', and he places the whole list immediately after the voice at the baptism says 'You are My beloved Son'. Ancient genealogies routinely telescoped, omitted and selected in order to make a claim about who someone was. On this view Matthew is filing a royal document and Luke is filing a human one, and the question 'which is the real list' is a question from a different century.

What should a student conclude? That the difficulty is genuine and old, that the four proposals are all defensible and none is demonstrable, and that the two writers agree on the point both of them are making, which is that this child stands in David's line. Anyone who tells you the problem is solved, in either direction, is telling you more than the evidence carries.

Both writers put the birth in Bethlehem, and Matthew quotes Micah 5:2 to say why it matters. Luke gets the family there by a different route: a registration ordered by Caesar Augustus, 'This was the first census to take place while Quirinius was governor of Syria' (Luke 2:2, BSB).

That verse carries a real historical problem and you should not hide it from students, because they will find it in an afternoon. Josephus dates the census conducted under Quirinius in Judea to AD 6, when Judea became a Roman province after Archelaus was deposed, and he treats it as a new and resented thing that provoked the revolt of Judas the Galilean (Antiquities 18.1-2). Herod the Great died in 4 BC. Both Matthew and Luke place the birth in Herod's lifetime. On the plain reading of the dates, Luke 2:2 is about a decade adrift.

The main responses are these. Some argue Quirinius held an earlier commission in the region, perhaps a military one, and that an earlier registration took place under his oversight; the evidence for this is thin and depends on a fragmentary inscription whose subject is not named. Some translate the Greek of 2:2 as 'this registration was before the one when Quirinius was governor of Syria', which the grammar can bear and which some good scholars defend, though most translators have not adopted it. The BSB prints the other reading in its text, 'the first census to take place while Quirinius was governor', and puts this one in a footnote ('Or This was the census before'), so both are on the page for a student to weigh. Some argue for a registration begun earlier and completed under Quirinius. And some, including many who hold a high view of scripture, conclude that Luke, writing decades later and at a distance, has attached the birth to the most famous census in living memory and got the dating wrong, and that this bears on Luke's precision rather than on the substance of what he reports. State the options, say which you find least strained, and say that the question is open.

The manger deserves a separate note, because the picture almost everyone carries is built partly on a translation. Luke says Mary laid the child in a manger because there was no room for them in what the BSB calls the inn, and the Greek word there is kataluma. Kataluma is the word Luke uses in 22:11 for the upper guest room where the last supper is eaten. When Luke wants a commercial inn he writes pandocheion, as he does in 10:34. Many houses of the period had an animal area on a lower level of the same structure, and a manger indoors is entirely ordinary. The likely scene is a crowded family house with no room in the guest space, not a hard-hearted innkeeper turning a pregnant woman into the cold. The innkeeper is not in the text at all.

This matters less for the sentiment than for the method. A student who learns to check one Greek word against the same writer's usage elsewhere has learned something transferable. The point is not to strip the story of warmth. It is to let the story be the one Luke wrote.

Matthew's infancy account is built on five fulfilment quotations and on dreams, four of them Joseph's and one the magi's in 2:12, and the pattern it draws is Israel's own story replayed in one child. A ruler orders the killing of infant boys. A child escapes. The child is carried into Egypt and then called out of it, and Matthew quotes Hosea 11:1, 'Out of Egypt I called My Son' (Matthew 2:15, BSB), a verse whose original subject is plainly the nation. One small detail of typography deserves a sentence: the BSB capitalises Son in Matthew 2:15 but prints 'out of Egypt I called My son' in Hosea 11:1 itself. Greek and Hebrew have no capitals, so that capital is the translators' reading of Matthew, and it is exactly the move the lesson is describing. Matthew knows that. He is not misreading Hosea; he is claiming that what was true of Israel is now true of this one Israelite, which is the move his whole Gospel makes. The magi are Gentiles who read the heavens and come to worship, while the king in Jerusalem and his scribes, who know exactly where the Messiah is to be born, do not travel the six miles to look.

Luke's account is built on songs and on the temple. Mary's Magnificat, Zechariah's Benedictus, the angels' Gloria, Simeon's Nunc Dimittis: four poems in two chapters, each of them saturated with Old Testament language, and the first of them announcing that God 'has brought down rulers from their thrones, but has exalted the humble'. The birth is announced to shepherds on a night shift. The child is presented in the temple, and the two people who recognise him are an old man and a widow of eighty-four (the BSB footnote allows that Luke means she had been a widow for eighty-four years). Luke's Gospel will keep doing this for twenty-four chapters.

There is one more piece of honest work to do here, on the virginal conception. Matthew cites Isaiah 7:14 in the Greek form, which uses parthenos, virgin. The Hebrew of Isaiah 7:14 has almah, which means a young woman of marriageable age and does not by itself specify virginity. The word often offered as the precise alternative, betulah, is not as clean as it is usually made to sound either: Genesis 24:16 adds a clause saying she had not had relations with any man, which would be redundant if the noun settled the matter, and Joel 1:8 has a betulah lamenting the husband of her youth. The honest statement is that Hebrew had no single term that unambiguously meant virgin and nothing else. So the Hebrew word alone cannot carry the argument that Isaiah 7:14 predicts a virgin birth, and Isaiah's own oracle has a near-term reference in the reign of Ahaz. The BSB prints 'the virgin' in the text of Isaiah 7:14 and adds the footnote 'Or young woman', so its own page records both options; the text carries the translators' choice and the footnote carries the Hebrew's openness. The other side has a real case, and it deserves to be heard at its strength: almah is never clearly used in the Hebrew Bible of a married woman, the Jewish translators of the Septuagint chose parthenos here long before there were any Christians, and many interpreters read the oracle as part of Isaiah 7 to 11, where the child of 9:6 bears divine titles, and so as reaching beyond Ahaz's reign. What the philology shows is that the Hebrew word does not settle the question by itself, and a teacher who claims it does, in either direction, will be caught out.

What that observation does not do is settle the theological question, and here the honest reading runs the other way from what a student might expect. Matthew's claim does not depend on Isaiah 7:14 alone. Matthew reports a virginal conception as something he has received, gives Joseph's resistance to it as evidence that it was not obvious, and Luke reports the same thing independently, with Mary's question in 1:34 making the point in her own voice. Two accounts that share very little share this. Whether one believes it is a separate matter from whether Isaiah predicted it, and the two questions should be kept apart.

Teach the two accounts separately before you ever teach them together. Read Matthew 1-2 on its own and ask what argument it makes. Read Luke 1-2 on its own and ask the same. Only then lay them side by side. Students who meet the fused nativity first will spend years unable to say which Gospel the shepherds are in.

When the differences come up, and they will, resist two bad instincts. The first is to harmonise everything on the spot, producing a composite narrative that no evangelist wrote and that quietly teaches students that the text needs rescuing. The second is to treat every difference as a defeat, which teaches them that ancient writers were trying and failing to do modern journalism. Both instincts share a wrong assumption about what kind of writing this is.

Say what can be said with confidence: the two accounts are substantially independent, they agree on a striking cluster of particulars, and they diverge in material and emphasis in ways that fit two writers with different sources and different arguments. Say what cannot: that the genealogies have been reconciled to general satisfaction, or that Luke's census dating has been vindicated. A student who can hold both of those sentences at once has learned the habit this whole course is trying to build.

And then teach what the accounts are actually about, because it is easy to spend the whole hour on the difficulties and never reach it. Matthew's angel gives a reason for the name: 'you are to give Him the name Jesus, because He will save His people from their sins.' Yeshua means the LORD saves. The name is the job description, and both Gospels put it in the mouth of the messenger before the child is born, which is to say that the point of the story is announced before the story begins.