Just over 10 years ago, the historian Robin Lane Fox wrote of the problem with the date of the birth of Jesus as presented in Luke’s Gospel in the starkest terms: “The Gospel…assumes that Quirinius and King Herod were contemporaries, when they were separated by ten years or more” (The Unauthorized Version, Viking/Penguin, 1991, p.28). Yet fundamentalists of various description assert that the historical evidence does not in any way undermine the credibility of the Gospel account. Who is right? Having found many web pages that deal with some of the evidence, but also having found it rare that all the relevant evidence is presented in a clear manner in one place, I have decided to attempt to do just that: to work through the evidence in a manner that does justice to the problem, and to the arguments that are used by both sides, so that those interested may navigate their way through this difficult problem. Otherwise, one will be tempted simply to trust the word of ‘experts’ who support one’s own view, whether that is in favor of the reliability of the Gospels or calling them into question.
Let us begin with the text of Luke 2:2 so as to be as clear as possible regarding what it means from the outset. The verse in question reads:
Which in English is normally translated as something like
‘This was the first census, which took place while Quirinius was in charge of Syria’
If this were all the information we had on the subject, then everything would appear reasonably straightforward in terms of both what the sentence means grammatically and what it implies historically. However, we do have additional information, and it does not agree with this statement. The problem is how Luke 2:2 relates to two other texts, one within the Christian canon and one from outside it. On the one hand, there is the information Matthew gives that Jesus was born roughly two years before the death of Herod the Great. That Jesus was born while Herod the great was alive is immediately clear from the Gospel of Matthew. That Jesus was thought to have been around 2 years old when Herod died is implied by the fact that, in Matthew’s narrative, Herod asks the Magi when the star they saw appeared, and then he proceeds to give an order that all male children in Bethlehem below the age of two be killed. At any rate, Herod died in 4 B.C.E., and thus if Matthew’s information is even close to being accurate, then Jesus was born sometime between 6-4 B.C.E. In fact, this would agree with what Luke’s Gospel says in 1:5, placing the birth of John the Baptist in the time of King Herod, and the birth of Jesus less than a year after that of John.
Our other source of information is the Jewish historian Josephus. Josephus tells how Quirinius became legate of Syria after Archelaus, the son of Herod the Great, was deposed from power. This happened in 6 C.E. At this point, the difficulty becomes clear. How could Jesus be born in Bethlehem in or after 6 C.E., when Quirinius became legate of Syria, and yet before the death of Herod the great, which occurred in 4 B.C.E. This was the problem that Robin Lane Fox was referring to in the quotation we gave earlier: it is not that there was no overlap between the lifetimes of Herod and Quirinius (and so in that sense they were ‘contemporaries’), but that Herod died before Quirinius came into office. That this was the case is glaringly obvious, since Quirinius came as legate precisely as a result of the end of the Herodian dynasty.
Now, any historian will admit that there is always a great deal of uncertainty involved in fixing dates in the ancient world and reconstructing events. And so there is nothing illegitimate in the attempts of scholars to seek a way to harmonize the relevant sources: Matthew, Luke, and Josephus. The only question is whether this can be done in a way that does justice to the available evidence, and to all the relevant evidence, without engaging in special pleading or in forcing the evidence to fit in order to defend the Bible’s historical accuracy at any cost. The cost of being dishonest about the evidence is far greater than the cost of acknowledging that the Bible does not always give accurate historical data. And so let us now look at some of the attempts that have been made to demonstrate the accuracy of Luke’s information.
One approach has been to argue that the sentence in Luke 2:2 should be translated in a different way:
‘This census happened before Quirinius was in charge of Syria’
The two questions we must ask about this suggestion are (1) Is this translation legitimate? and (2) Would it solve the problem? We shall take each question in turn.
First, can one make a case for this translation? The word translated ‘first’ can at times also mean ‘before’, and the phrase ‘Quirinius being in charge of Syria’ is in the genitive case, as it would need to be after protos when used in its comparative sense of ‘before’. However, these considerations do not by any means settle the matter. As Daniel B. Wallace points out in his web article [http://www.bible.org/page.php?page_id=1146], the lack of a definite article and the position of the words all make it almost impossible that the alternative translation using ‘before’ could be correct. For one thing, when ‘protos’ is used to mean ‘before’, it is followed immediately by the word it is modifying. Here, it is before the verb, and so to take it as ‘before’ would be the equivalent of saying in English:
‘This census before happened Quirinius being in charge of Syria’
If the phrase in the genitive is taken as a genitive absolute construction, on the other hand, and the phrase is translated according to the norms of usual Greek usage, then the phrase must mean something like:
‘This first census took place, while Quirinius was in charge of Syria’
or
‘This census happened for the first time while Quirinius was in charge of Syria’
The census of Quirinius in 6 C.E. is the first known Roman census to have taken place in Judea, and so it fits the bill. To suggest that this refers to a census that took place ‘before’ the first one known from historical sources, and which provoked quite a reaction for this very reason (see Josephus’ account of what happened), is stretching the limits of credibility.
This leads us on to the question of whether translating the phrase so that it refers to an enrollment or census that took place before Quirinius was legate of Syria would solve the problem. The answer to this question appears to be ‘no’, but let us not prejudge the matter too hastily. The main reason why the attempt to make this refer to an earlier census does not help is the lack of any evidence for such a census whatsoever. Reference is occasionally made to a census around the year 3 B.C.E. The fifth century historian Orosius mentions it, and during this period there is also evidence of the swearing of oaths of allegiance to Caesar Augustus and images of him being set up – not surprising, since it was at this time that he came to power. But the census taken was of Roman citizens. The people of vassal kingdoms may have been requested to swear allegiance to Caesar (See Josephus, Antiquities 17.2.4 #42), but even this would have raised objections among the Jews. This enrollment or registration of Roman citizens would not have applied to Joseph of Nazareth, who was not a Roman citizen. Nor was Jesus, since had he inherited this legal status from Joseph, he would not have been crucified, since crucifixion was not a punishment that would be carried out against a Roman citizen [See the discussion in chapter 6 of Martin Hengel’s book Crucifixion, SCM, 1977]. So reference to this empire-wide registration of Roman citizens has no relevance to Jesus, and no link whatsoever to Luke 2:1, which refers to a decree regarding taxation. [The article by Glenn Miller at http://www.christian-thinktank.com/quirinius.html clouds the issue precisely by focusing attention on this registration of Roman citizens]. Had the reference been to such an enrollment of Roman citizens, then there would be no obvious reason for mentioning Quirinius’ famous tax census in this context! [as Raymond Brown points out in Birth of the Messiah, Doubleday, 1993, p.552]
Another reason for thinking that the attempt to make Luke 2:2 refer to an enrollment prior to the census of Quirinius is unable to bring Luke into agreement with Josephus is that while Herod was king, a Roman census in his territory is highly unlikely to have taken place, and certainly such a controversial event could not have taken place without leaving some mark in the literature of that time. It is precisely because it is the first occurrence of such an event that the census for purposes of taxation which took place under Quirinius with the introduction of direct Roman rule caused such a commotion, left its mark in Josephus’ writing, and was so famous that Luke could simply refer to ‘the registration that took place when Quirinius was in charge of Syria’ and everyone would know precisely what he meant. And so, for both grammatical and historical reasons, the attempt to make this a census that took place before Quirinius was legate of Syria is unconvincing.
Another approach has been to accept that this is the ‘first’ census that occurred while Quirinius was legate, but that Quirinius had been legate (or in some other way ‘in charge of Syria’) prior to his appointment in connection with the famous census in 6 C.E. In connection with this suggestion, it is common to mention a fragmentary inscription known as the ‘Lapis Tiburtinus’. It reads:
…(R)EGUM QUA REDACTA IN POT … AUGUSTI POPULIQUE ROMANI SENATU(S) … SUPPLICATIONES BINAS OB RES PROSP …
… IPSI ORNAMENTA TRIUMPH … PRO CONSUL ASIAM PROVINCIAM OP … DIVI AUGUSTI (I)TERUM SYRIAM ET PH….
The gist of it is that it refers to someone who was ‘legate a second time in Syria’. Scholars and historians trying to solve the dilemma of Luke 2:2 found this helpful to their quest to resolve the problem. However, it does not in fact state clearly that the person in question was ‘legate of Syria a second time’, but only that the person was ‘legate a second time, in Syria’, which could easily mean that the person was previously legate, but not of Syria. Be that as it may, the only two periods prior to 6 C.E. when we do not know who was legate of Syria, and thus Quirinius could theoretically have served in this position, are prior to 10 B.C.E. and between 4-1 B.C.E. A second (or I should say first) stint by Quirinius as legate of Syria during either of these periods could potentially make Luke internally consistent on this point, but neither would fit the date of around 6 B.C.E. for the birth of Jesus suggested by Matthew’s Gospel, unless one were to posit a 4-year delay between the instituting of the census by Quirinius before 10 B.C.E., and its completion by S. Sentius Saturninus in between 9-6 B.C.E. while he was legate. In fact, Tertullian would later state that Joseph and Mary went to Bethlehem as a result of a census held while Saturninus was in charge. But we may presume that he was simply aware of the same difficulty concerning dates in Luke’s Gospel that we are, and was trying to ‘fix’ the problem.
But even if one were to posit an earlier term served by Quirinius as legate, or in some other official capacity, in Syria, this would still leave other objections, such as that regarding the plausibility of a Roman census taking place in Herod’s kingdom. Again, while not impossible, such an event is unlikely to have taken place without it leaving any mark, and would make the fuss regarding the census of Quirinius in 6 C.E. hard to understand. And while some appeal to the evidence in Josephus’ Antiquities, 18.1.1, that relations between Herod and Caesar Augustus were at times strained, there is no evidence that he did demote Herod from the status of ‘friend’ and thus impose Roman taxation in the form more appropriate for a Roman province.
It is thus difficult if not impossible to read the two birth narratives (found in Matthew’s and Luke’s Gospels) in conjunction with one another and with other historical sources of information, and to work out a logical chronology. In fact, this problem is not only one relating to information found in Josephus. The impression one gets from reading Matthew’s Gospel is that Bethlehem is the home of Joseph, Mary, and Jesus. It is the place where they are found, it is the place the Magi find them roughly two years after Jesus is born, it is the place they would return to had not fear of Herod’s son Archelaus forced them to seek refuge in Galilee instead (Matthew 2:22). In Luke’s Gospel, on the other hand, Joseph and Mary are from Nazareth, and it is only the census that forces them to Bethlehem (Luke 2:1-5). Luke 2:39 says that ‘once they had done everything the Law required, they went back to Galilee, to their own town of Nazareth’. Not only did the Law of Moses not require a two-year stay for them to do everything it required, but it is unlikely that Mary would have had to accompany Joseph to register. Indeed, Joseph himself would not have been required to register in Judea, since Galilee was not affected by the census in 6 C.E. Of course, it is possible to maintain that Joseph had property in Bethlehem, and thus was only temporarily absent from there while living in Nazareth. But this attempted ‘solution’ simply raises another problem, namely why someone with property in Bethlehem is forced to rely on the hospitality of others when he goes there for the census. [That Joseph and Mary depended on the hospitality of others in Luke’s version of the story is clear, even though in Matthew’s version they are living in a house that is presumably their own. For the cultural background to the Lucan story follow THIS LINK]. And so Luke’s version, which has Joseph and Mary not only rely on the hospitality of others, but also offer the sacrifice the Law specified for those who were not wealthy (Luke 2:24; cf. Leviticus 5:7), excludes the one possible reason for the journey to have taken place: namely, the possibility that Joseph owned property in that area. Luke also tells us that Mary had relatives not far away, and so there is simply no reason why she should not have stayed with them, since women did not normally own property and were not required to register in the case of a census of this source.
See also Richard Carrier's web page, "The Date of the Nativity in Luke".
And so we can see how, from a historian’s perspective, the birth/infancy narratives are extremely problematic. Even the genealogies do not agree. The accounts of Jesus' birth and infancy in the Gospels are best regarded as legendary and theological attempts to say something about Jesus by filling in the gaps in an unknown period in his life. As we shall see slightly later, it was quite normal procedure to attribute a miraculous birth to a person of importance in the ancient world. If that is what Matthew and Luke do in their Gospels, then perhaps we should read them precisely as just that: as attempts not to give us historical data about when and where Jesus was born, but to tell stories that will help the reader of their Gospel to understand who Jesus is and what his significance is.
Please proceed now to the next document in this series, about the cultural background to Luke’s infancy narrative, if you have not done so already by following the link above. The next two documents will give you a chance to look at and think about other ways of reading and approaching these stories in the Gospels. To proceed, just CLICK HERE.
Other relevant links:
N. F. Gier, "Serious Problems With Luke's Census" http://www.class.uidaho.edu/ngier/census.htm
Felix Just, "The Birth of Jesus" http://catholic-resources.org/Bible/Jesus-Birth.htm
Jesus' Genealogies http://www.usbible.com/Gospel/jesus_genealogy.htm
Ronald Troxel,
"Luke's Infancy Narrative"
http://imp.lss.wisc.edu/~rltroxel/gospels/Lect14.pdf