Here is a question that is sometimes raised for dating Luke’s Gospel or Luke's competence as a historian.
The NIV translation of the Bible, among some others, takes the liberty of translating Luke 2:1 to the effect that a “census” was conducted across the Roman Empire around the time of Jesus' birth. But did that really happen? And are there any implications if it never happened?If we think that Luke 2:1 is claiming a "census" took place, then there could be implications for dating Luke’s Gospel. Here’s why. While there is a historical record of an Empire-wide census - counting people and property - in 74AD, there isn’t one for the time Luke is talking about, unless you are basically talking about Quirinius' 6AD undertaking instead. (In the era of Luke 2:1, the Emperor Augustus was counting everything and anything he could get away with, but not necessarily an empire-wide census of people and property covering Judea.) Where could Luke have got the idea of such a "census" in Herod's reign from (if he meant that)? Would it mean (the critics' view) that the 74AD Roman census gave Luke the idea of imagining a massive census in the year of Jesus’ birth? Well, if so, that means that Luke wrote his Gospel sometime post-74AD. That is a reason why some scholars use Luke 2:1 to date the gospel after 70AD. But there is already a heap of ‘ifs’ here. Let’s step back and unpick the overlapping issues.
The devil is in the detail, and especially the assumptions of translators. In Luke 2:1, the specific Greek words matter. Firstly, whichever way you choose to read it, Luke says there was more than one registration. There are two main options about that. Option A: Luke could be saying that 6AD was the "first" of more than one registration under Quirinius - which is odd because neither Luke nor Josephus ever claims that Quirinius had a second one after 6AD. Or option B: Luke is saying that the one around Jesus' birth was the one "before" the 6AD Qurinius one, but fewer scholars like the translation "before". Either way, we ought to have some sense of why Luke undoubtedly thought there was more than one registration, but I can't resolve that in this post. What would be interesting to develop further is that when Luke talks about the registration in the nativity story, it is here that we find the word translated "first"; whereas when Gamaliel refers to a registration, it is not claimed to be the first. This word "first" is plausibly how Luke distinguishes between the two.
Secondly, my focus here: these words report a decree to REGISTER i.e. “apographesthai” (not literally the word “census”, and there are alternative meanings - see below), to register ALL THE WORLD i.e. “pasan ten oikoumenen” (not literally “the Roman Empire”). That’s all we have to go on within the verse: "to register all the world". Everything else is interpretation. Let’s break this down, because there are reasons to think "census" is a mistranslation, and it wouldn't be the first time that an exegetical problem has been thrown up by a mistranslation or a misunderstanding.
• “REGISTER”
Simply following the traditions of some translators, and their assumptions about what “register” means, is problematic. Clearly, we can’t absolutely assume without evidence that “register” here means “census” (NIV) - certainly not in the modern sense of a head count or in the ancient sense of mustering an army; nor can we absolutely assume it means “pay taxes” (KJV); or “register loyalty to Augustus” (see this suggestion which I think needs more attention). Census, taxation and loyalty oath are all within the range of activities covered by "register." Perhaps the safest translation is to leave it ambiguous as “register”, admitting that we can’t confirm what was meant by Luke to have been registered around the time of Jesus' birth.
For the purpose of this post, it ultimately has little bearing as it is an inconclusive translation. Even if we were to suppose that Luke made an error in Luke 2:1, writing some decades later, nothing about dating Luke-Acts pre- or post-70AD is determined by it. That is, an error in dating the registration could be made by a writer before or after 70AD.
But, as said, a date after 70AD will be often be suggested by those who hold that "register" means "census," and this is because the best external explanation for anyone writing "census" is found post-74AD.
So, the issue of dating can be influenced by assumptions of what this "registration" actually was.
There is another reason why some assume that Luke means "census". Luke uses the same word "register" when talking about the 6AD one (Acts 5:37), but this does not mean that Luke is using such a generic word to mean the exact same kind of thing with laser-sharp certainty. Here I need to expand on the suggestion of an error by Luke himself, as to his placing of Quirinius. In Acts 5:37, Luke a second time refers to a "registration" but which he now associates with Judas the Galilean's troubles (one which historian Josephus happens to uniquely connect with a tax-related registration in 6AD, when Quirinius held an office). Whereas in Luke 2:1, a "registration" is associated with an earlier time - Herod's era (which Luke uniquely connects with Quirinius having some kind of governing influence). Why be certain as to why he gives the same meaning to the generic word "registration" each time?
That's two registrations connected separately by the two authors with Quirinius, unless Luke has awkwardly messed about with a single one (but why then does he clearly say there were two?). As hinted above, anyone who tries to say that both Josephus and Luke are correct usually has to either translate the Luke 2:1 registration as the one "before" Quirinius' registration; either that or calculate that Luke's two registrations where in two periods with Quirinius having some kind of official role in Judea in both periods. If we don't want to leap to the conclusion that either Josephus or Luke is wrong. That is one of the reasons for not leaping to the conclusion that they are all one and the same 'registration', especially as Luke unquestionably indicates there was more than one registration, and he pins one of them to the era of Herod who died in 4BC, and the other to Judas the Galilean's antics which Josephus effectively dates to 6AD. You can see why some people think one of those authors made a mistake (usually pointing the finger at Luke, not Josephus, because they can see that Luke knows about the 6AD census and could muddle things). And this is why some factor this into their thinking about dating Luke's Gospel late (after 74AD). Bear with me here.
For my part, I don't think Luke is muddled, in light of two loose ends that leave some awkwardness for the argument that Luke is conflating together two different moments in time. Firstly, that he clearly says there were two registrations. Secondly, the loose end that his colourfully detailed story of the birth of Christ (Luke 2) does not have any trace of Judas the Galilean's revolt in the background, which it surely should do if Luke has the 6AD census revolt in his mind. It would add colour to the story of Joseph and Mary's journey to Bethlehem. For that reason, I don't see that Luke thought that these two events - the birth and the 6AD census revolt - coincide in time. In that light, he cannot be placing them under the same "registration," in his mind. If Luke had included any trace of Judas the Galilean's revolt in the nativity story, then yes, we could then say that we know he made the error. But he doesn't. Luke's terse mention of Judas the Galilean (Acts 5:37) omitted mention of Quirinius and a registration, but he must have known that Judas' revolt was about that registration. So that does suggest that Luke is conscious of these being his two registrations: in separate moments in time, the story of the birth and the story of the revolt. He was probably conscious that he was referring, or alluding, to a registration in each of those separate stories. Thus, in his mind, two registrations are clearly on the table, so perhaps the first being a loyalty oath, and the second the 6AD tax-related census. Of course, even with a two-registration solution, there still has to be a satisfactory explanation for what connects Quirinius' name in each case. The explanation for 6AD being under Quirinius is supplied directly by Josephus. The explanation for one under Herod is dependent on Luke and needs a bit of reasoning that is somewhat difficult as explained here.
Let's get back to the crucial issue of translation, and how that at least helps to see what the evidence is. We've seen a range of possible meanings for "registration." Now let's look at...
• “ALL THE WORLD”
The Greek word for ‘world’ here is oikoumenen. We can’t absolutely assume it means “the whole geographic world”, and similarly we can’t absolutely assume the NIV’s interpretation that it means all of “the Roman Empire”, or that it just means “all Judea”. The question is, whose “oikoumenen” is it in Luke’s mind, given that the only illustrative example he gives merely takes a single couple to Bethlehem? The safest translation is perhaps just “world”, admitting that we can’t nail down whose “world” is intended. But clearly, at least, Luke would not intend the literal whole physical world because he will have known that that went beyond Caesar’s jurisdiction.
CONCLUSION: All we are strictly left with as an accurate translation is “register all the world”. That’s how I would translate it, with a footnote to say that we don’t know what that means; and mentioning the notable suggestions of translators and exegetes. The ESV translates it more or less the same way: "In those days a decree went out from Caesar Augustus that all the world should be registered."
If there are three options for what "registration" means, and "census” is probably an anachronistic one, we need a good rationale for why we would select that in this specific case (NIV translators ought to take note!).
In short, if we were to start from the presumption that Luke has made a mistake, we would have to choose which mistake, and there are different ways of cutting the cookie: e.g. a pre- or post-70AD mistake due to Luke misapplying an awareness of Imperial oaths or tax collection; or a post-74AD mistake of Luke misapplying an awareness of an Empire-wide census.
My tentative conclusion would be that this registration is to do with making a loyalty oath to Augustus, as the best fit with the text and the historical period (with people being registered for this purpose in Judea and probably wider afield). If it is about an oath to Augustus, then Luke may be correct about it having happened. We know that this was a practice in the era of Jesus' birth. On that basis, I lean towards that, but I wouldn't press that too firmly.
A critic might still think, even if it is about an oath, that it is a mistake made by Luke pre- or post-70AD, but there would be no certain way of knowing which. So that does not help to date this gospel anyway.
I don’t think we can date Luke’s gospel from Luke 2:1 – we might try to, but we shouldn’t press our conclusions about it too firmly into service, when no-one can be sure of the correct translation and meaning of it.
