The First Christmas Liturgy (Part II)

What was the Church’s first Christmas liturgy like?

In my first post on this subject, I suggested that there is very good evidence that parts of the infancy narrative of Luke (specifically the canticles and hymns) were part of the earliest Christmas liturgy in the Jewish-Christian Church.

If you are interested in the subject, the questions you will be asking will include: what was the very early Jewish-Christian liturgy like? When in the Jewish year would the birth of Christ have been celebrated and what would the liturgical context of the celebration have been? When in the year, in fact, was Jesus actually born?

There are actually answers to all these questions, but some of them are easier to describe and summarize than others, and for a couple of them, I have yet to completely finish my research. But I can begin quite usefully with the question: when in the year was Jesus born? The traditional date for the Western Church of course, is December 25; there is no complete answer as to how this date was arrived at. The best answer is that the date was selected based on the idea of the “integral” age of the prophets, whose day of death must have been the same as the day of their conception or birth. They came up with March 25 for the day of Jesus’ death as well as his conception, and added nine months to get December 25. The dating at earliest surfaced in the 3rd century. The Eastern Church doesn’t celebrated Christmas but rather Epiphany, which I will discuss later. (1) Most people realize that December 25 is probably not the actual date of Christ’s birth. But is it possible even to find out when he was born?

There is a intriguing and very full attempt at an answer provided by Ernest L. Martin in his book The Star of Bethlehem: The Star that Astonished the World. He tackles the historical indications of Christ’s birth from several angles and in each case, comes up with a remarkably consistent answer, both as to the year and the date. His book is conveniently available to read online. (2) Since some of the evidence he discusses about the time of the Jewish calendar and celebrations depends on the actual year of Jesus’ birth, I’ll tackle that part first.

The Year of Herod’s Death

One of the primary chronological indicators for the birth of Jesus is the death of Herod the Great, because both Matthew and Luke say that he was born in during Herod’s reign (Mt. 2:1, Lk. 1:5). For over a hundred years, following the work of Emil Schürer, (3) historians have almost unanimously assigned the death of King Herod the Great to 4 B.C. This was based on several indications. First is the length of his reign given by the Jewish historian Josephus, who said that he reigned 34 years from his conquest of Jerusalem, traditionally thought to have taken place in 37 B.C. Then there is the fact that his sons, Archelaus and Antipas, claimed to have begun their reigns in 4 B.C. Lastly, Josephus says that Herod died shortly after a lunar eclipse – and there was a partial lunar eclipse in March of 4 B.C.

Herod the Great

Herod the Great

Unfortunately, this dating is severely at odds with the traditional dating of Jesus’ birth in the Gospels and the early Church Fathers. If Herod’s death is as far back as 4 B.C., then Jesus could have been born at earliest, in 5-6 B.C. But the chronological data given by Luke (3:1, 23) that Jesus was about 30 years old in the “fifteenth year of the reign of Tiberius Caesar” would yield a birth date of around 3 B.C. Something is wrong with the chronological data.

Martin manages to straighten everything out. Following the lead of W. E. Filmer (4), he establishes that the exact date of the fall of Jerusalem to Herod was not 37 B.C., but the end of 36 B.C., and that the death of Antigonus, which Josephus actually uses to date Herod’s reign, actually would have occurred in early 35 A.D. Subtract 34 years and we get 1 B.C. for Herod’s death.

As for the lunar eclipse, the one in 4 B.C would not have allowed sufficient time for all the events Joseph lists as occurring between then and the Passover after Herod’s death – only 29 days. But there was also a full lunar eclipse on January 10, 1 B.C., which would have allowed a more reasonable time of over 3 months for these events.

Herod’s sons did date their reigns from 4 B.C., but this is most likely because they were attempting to legitimize their reigns by claiming the co-rulership that Herod had granted to his oldest son Antipater, whom he made co-ruler in 4 B.C. Herod subsequently had Antipater executed for attempting to assassinate him, and then named Archelaus his successor. As client kings, the rule of Archelaus and Antipas would have had to be ratified by Emperor Augustus. But Augustus never formally proclaimed them kings, though he did permit them to rule, so shoring up the legitimacy of their royal succession this way would have been a natural reaction. Antedating reigns was actually common in the ancient world.

In addition to all this, Martin shows that events in Roman history that otherwise make no sense if Herod died in 4 B.C. become completely comprehensible. This is perhaps the most impressive evidence of all. For instance, it makes sense of the war that Rome fought to quell the uprising that broke out in Judea shortly after Herod’s death, known as “the war of Varus.” Historians have long sought for confirmation of this war in Roman sources, but could find nothing. But because they thought Herod died in 4 B.C., they were looking in the wrong year.

On the other hand, we have very clear evidence that Augustus was demobilizing his troops and granting pensions to his officers during the whole period from 5 B.C. to 2 B.C., something he would certainly not have done if he were fighting a major war in the East that began in 4 B.C. The new date also makes sense of the imperial victory that Augustus claimed in 1 A.D. Historians have long been puzzled by this, since according to the traditional dating, Augustus fought no known wars during this period. But if we accept that the war of Varus began in the summer of 1 B.C. and ended in early 1 A.D., the victory celebration would fit perfectly. Martin has much more evidence as well, which I don’t have room for.

Martin also analyzes the statement in Luke 3, and determines exactly how the beginning of the fifteenth year of Tiberius’ reign would have been understood, and concludes that Jesus would have been born in the fall of 3 B.C.

Once I had read Martin, I was astonished at the weakness of the 4. B.C. dating. A number of other recent scholars have come to similar conclusions, and his conclusions were accepted by Jack Finegan, the “dean” of scholars of Biblical chronology. (5)

The Census in Luke

The mentioned of a census for collecting of taxes in Luke has long given rise to objections that the Romans did not collect taxes in Judea, as that was a right granted to “client kings” like Herod. In fact, Luke doesn’t specifically say that there was a census, nor does he say anything about taxes — these are all interpretations put on his words by translators. He actually spoke about an apographe, or “registration,” which might have been done for a number of purposes.

Using this knowledge, and relying on the traditional date for Jesus birth in 3-2 B.C., Martin dated the birth of Jesus to coincide with the empire-wide registration mentioned in a number of sources, requiring a loyalty oath to Augustus, that was part of the planned celebrations for the Emperor being named Pater Patriae or “Father of his Country,” which took place in February 2 B.C. The decree for the registration would have gone out the previous spring/summer of 3 B.C. and the registration would have been carried out in the fall. Once again, all the information fits: Mary and Joseph went to Bethlehem for this census and Jesus was born in the late summer-early fall of 3 B.C.

The Priestly Courses in the Temple

Martin goes even further than this. Another traditional method for dating the birth of Jesus has been to use the priestly courses in the temple, a method used as early as the 4th century by St. John Chrysostom. The method is based on the dating of the temple service of Zecharias, at which he was visited by an angel announcing that he would have a son (Lk 1:5-25). Zechariah’s priestly class was that of Abijah. He would have served twice a year. If we know when his course was, we can approximately date the conception of John (presumably right after he returned home), and the conception of Jesus six months later. Chrysostom counted back from the last known course at the Temple in 70 A.D. and hit the date of December 25. (6)

Martin conducted a fresh evaluation of the courses and when they were held each year, depending on the calendar and date of Passover. He wrote quite extensively about it, and I will leave you to look at the details for yourself. The result is that he was able to establish that one of Zechariah’s two courses for the time in question would lead to a date of early September 3 B.C. for the birth of Jesus.

The Star

Martin’s last method is astronomical and involves the identity of the Star of Bethlehem. This is one of the most amazing aspects of his work. Many previous scholars had argued in favor of various supernovas and star conjunctions that took place in 7 or 6 B.C. in accordance with Schürer’s dating. But Martin, taking a fresh look, discovered a most unusual pattern of movement between Jupiter, Venus and the star Regulus (which means literally “little king”) that took place between May 3 B.C. and December 25, 2 B.C. I will leave you to study Martin to get the whole picture, but it involves, among other things, the “little king” being crowned in the sky by Jupiter.

Martin believes that the December 25 end to the phenomenon marked not the day of Christ’s birth, but the day the star “stood still” for the Magi to find the Holy Family in Bethlehem, where they had been staying for some time. He explains that Jupiter was in Virgo at the time and in retrogression, which would have made it appear stationary in the sky. He sets the birth of Christ about 15 months earlier, when Mary and Joseph first arrived, on a date marked by an astonishing conjunction of stars and planets that he believes has a Biblical confirmation.

The Woman Clothed with the Sun

On September 11, 3 B.C., the sun and moon appeared together in the constellation Virgo. The sun was mid-body to the woman, the new moon or crescent moon beneath her feet. She was at the “head” of the twelve signs of the zodiac.

All of this points clearly, Martin says, to what John says in Revelation:

A great sign appeared in the sky, a woman clothed with the sun, with the moon under her feet, and on her head a crown of twelve stars. She was with child and wailed aloud in pain as she labored to give birth. Then another sign appeared in the sky; it was a huge red dragon, with seven heads and ten horns, and on its heads were seven diadems. Its tail swept away a third of the stars in the sky and hurled them down to the earth. Then the dragon stood before the woman about to give birth, to devour her child when she gave birth. She gave birth to a son, a male child, destined to rule all the nations with an iron rod. (Rev. 12:1-5).

Martin believes that John knew of this astrological wonder at Christ’s birth and presented it in his vision. In fact, many other details in it are symbolic. The ancient Israelites thought that the constellation Virgo represented Ruth, who was an ancestor of King David, — and hence also of Christ. Later, they identified her with the prophecy in Isaiah 7:15 – “Behold, a virgin shall conceive and bear a son.”

Such was the night sky on September 11, 3 B.C.

That day was the Jewish feast of Rosh Hashanah — the Jewish New Year and the beginning of the High Holy Days. The significance of the date and its relation to the life of Jesus and his identity are extraordinary — but I will leave that until next time.



(1) William Tighe, “Calculating Christmas,” Touchstone (December 2003).

(2) Ernest L. Martin, The Star of Bethlehem: the Star that Astonished the World (expanded ed. 1993). View the online version here. The book apparently incorporates a great deal of his earlier book The Birth of Christ recalculated (Pasadena: FBR Publications, 1978, 1980 second edition). See also his “The Nativity and Herod’s Death,” in Chronos, Kairos, Christos: Nativity and Chronological Studies Presented to Jack Finegan, ed. Jerry Vardaman and Edwin M. Yamauchi (Winona Lake, IN: Eisenbrauns, 1989) 85–92.

(3) Emil Schürer, A History of the Jewish People in the Time of Jesus Christ, 5 vols (New York, Scribner’s, 1896)

(4) W.E. Filmer, “Chronology of the Reign of Herod the Great,” Journal of Theological Studies ns 17 (1966) 283–298;

(5) Ormond Edwards, “Herodian Chronology,” Palestine Exploration Quarterly 114 (1982) 29–42; Paul Keresztes, Imperial Rome and the Christians: From Herod the Great to About 200 A.D. (Lanham, MD: University Press of America, 1989) 1–43; Jack Finegan, Handbook of Biblical Chronology, rev. ed. (Princeton U. Press, 2001).

(6) Richard Ostling, “Why is Dec. 25 the date to celebrate Christmas? Two explanations compete.” Associated Press article, December 22, 2004.