Dating and Authorship of the Gospels – Mark and Luke

When discussing the use of the Bible as a historical source, or the primary source for knowledge about early Christianity, or its Apostolic origins, there is often much debate about who wrote the Gospels, and when. The Christian often assigns an early date and agrees with the traditional authors, while the sceptic often assigns a late date and says that we have no idea who wrote the texts. The sceptic will then use this fact in an argument for why we should not accept Christianity, either because we don’t really know anything about the early Christians, or because the texts lie about their origins, etc. Here, I intend to lay out systematically some arguments to assign an early date and traditional authors to the Gospels.

(I will also include Acts in Luke, and refer to Luke and Acts together as “Luke”, since Acts is written by the same person as Luke, and certainly written a short time after Luke.)

The order of the sections may seem a bit odd, but I think it better shows the flow of my argument, and the dependence that exists between the various facts here. We will only concern ourselves here with Mark and Luke, Matthew and John will come later.

Mark was not late

Much of the analysis of the Gospels depends on their inter-relationships. That is, if we know that Mark was first, and we know that Mark was after 70, then we know that the other Gospels must be some time after 70. And so the whole framework of the sceptical dating system really depends on one or two key arguments. But we can not only refute these one or two, we can also establish one or two key arguments from the other direction, and propose an earlier timing for each Gospel based on their inter-relationships.

The primary argument for a late Mark, a Mark written after 70, is the inclusion of Jesus’ prophecy of the destruction of the temple. As these scholars argue, Jesus can’t have known that the temple would soon be destroyed, so this prophecy is the work of a later author, writing after the event, inserting the prophecy into Jesus’ ministry.

In most circumstances, I would be on board with the methodological naturalism employed here. If you made the same argument regarding the text outside the context of a religious debate, I would accept it. However here, it is often circular. If we are going to make the assumption that Jesus is incapable of prophecy, conclude late Gospels, and then use that conclusion in an argument against Christianity, then that argument is a circular argument, as it includes the premise “Jesus is incapable of prophecy”.

Given our context of debating the truth or falsehood of Christianity, it seems most reasonable to not make this assumption of Jesus’ inability to prophecy. That assumption is simply the assumption that Christianity is false, and such assumptions are not useful to us when we are trying to determine whether Christianity actually is true. You can’t make an argument that Christianity is false if you assume as a premise that Christianity is false. Therefore in this context, we ought not use this argument for a late date of Mark.

And indeed, there are no other serious arguments for a late date of Mark. There are some very speculative arguments which I will discuss, but this is certainly the main one.

One of these speculative arguments is the argument that the legion of demons in Mark 5 would only be comprehensible to readers of Mark after the war in 70, because before this time the Roman military in the area was not legionary. I think this is a very poor argument for several reasons. First, it assumes that Mark was written to Judeans, not people more familiar with Roman legions. Second, it assumes that the Judeans would not be familiar with the concept of a “legion”, which I think is absurd. Roman legions had existed for centuries by this point, were the pride of the Roman nation, and had been instrumental in dozens of battles and wars. To say that the Judeans would have no idea what a “legion” is is absurd. In fact, it may make even more sense here, if they had only heard rumours of the legendary “legion”, an unparalleled military force, the demon’s use of this word would be far more intimidating.

Luke was not late

If we have no good reason to believe Mark was late, then we have one fewer reason to push the composition of Luke out to the end of the first century as many scholars do. However there are other arguments for a late Luke, and I would like to deal with those.

One such argument is the claim that the author of Luke used Josephus as a source, and that since Josephus only wrote by about 90 AD, Luke must be at least this late. There is only one real possible case where this happened: the references to Theudas and Judas in Acts 5 and Josephus’ Antiquities 20:97-99, 102. However Josephus and Luke do not agree precisely on the dates, so either they are referring to different events, or if Luke used Josephus as a source he didn’t trust Josephus’ timelime. In that case, why use him at all? And why are there not more similarities? Importantly, why did Luke not include an account of the martyrdom of James found in Josephus? James is very significant in Acts 15, if Luke knew of James’ martyrdom, it would certainly have been included. More on that later.

Luke was early

If Luke really was written by Luke, Paul’s travelling companion, then Luke and Acts must have been written during Luke’s lifetime. This makes a date as late as the end of the first century or the start of the second century unlikely. We do not know exactly when Luke died, but it is unlikely he lived that long.

However, this is my main argument: if Luke knew about the deaths of James, Peter, or Paul, he would have certainly included them. The persecution and martyrdom of Christians is a central theme of Acts, and Peter, Paul and James are central figures of Acts. And indeed, their deaths are some of the most significant events in first-century Christian history, outside the life of Jesus Himself. It is absurd to suggest that the meticulous Luke, concerned with persecution, concerned with the lives of Peter, Paul and James, and concerned with Christian history in general, would omit these events. Neither would it omit the more general persecution of Nero, or the destruction of the Temple (surely he would include that in order to validate Jesus’ prophecy).

James was martyred in 62, and the latest event recorded in Acts is in 62, so I think we can pretty accurately date acts to around 62: after the last recorded event, before the news of James had reached Luke.

Consider also the account of the events of Paul’s last missionary journey that Luke records. Unlike the earlier events from many years before this point, this journey is described in great detail. The details described are often somewhat irrelevant to the theological point that Luke is attempting to make, and seem to be included simply because they are true.

This phenomenon has been explained by critical scholars as the author of Luke, writing in the late first or early second centuries, adapting some previously written work and incorporating it into his. However, there is no reason to believe this other than their presupposition that Luke was not an eyewitness and Acts was written late. There is no change in the kind of language, no change in the themes of the work, nothing in the text that would indicate that it was another work incorporated into Luke. Instead the natural conclusion here is that Luke spends time on these events because they had just happened, and describes them in detail because he was there. Which leads us to…

Luke was written by Luke

This was the universal position of the early church, as far back as we have sources. And indeed, this is the testimony of the work of Acts itself. We know at what point in the narrative Luke joins Paul on his journey, and this is the same point at which the language switches from “they went” to “we went”. 

Of course many scholars dispute the “we” passages, since as Luke must have been written late, these cannot be genuinely written by a travelling companion of Paul. But I think their arguments are not very good. There seems to be no real textual evidence that the “we” passages are forgeries as Ehrman claims, or a stylistic choice that doesn’t indicate Lukan authorship as other scholars claim. Their claims are mere suggestions, without any textual evidence to support them. 

One such argument is given not on the basis of textual evidence, but on the basis of “irreconcilable” differences between how Luke presents Paul and how Paul presents himself in his epistles. However these differences are quite tiny, and Luke’s representation of Paul is compatible with Paul’s own description of himself in his epistles. 

There is more positive evidence of Luke being the author. The author was clearly someone very familiar with technical medical terminology, due to the technical medical terminology that appears several times throughout these works. The author was also familiar with the geography of first century Judea, therefore someone who likely had first hand knowledge of the area.

Mark was early

We have gone one direction and refuted some arguments for a late Mark. We should now go the other direction, and produce some arguments for an early Mark. The strongest argument is the date of Luke: if I am right and Luke was finished by 62 then Mark must have been completed some time earlier, since it seems like Mark drew on Luke.

Some other arguments include Mark’s use of Latinisms. Mark uses many Latin terms which he expects the reader to be familiar with. These show a certain affinity for Rome, rather than hostility, but we’d expect hostility towards Rome if the great persecution had begun, if Paul and Peter or even James had been executed etc. This along with the dating of Luke places Mark certainly before 64.

Some scholars have used the existence of these latinisms indicates that Mark is writing after the war in 70, when Roman occupation had caused the Judeans to become more familiar with these Roman concepts. However some scholars suggest an alternative: if Mark was written for a Roman audience while Peter was in Rome, as the early sources claim, then it makes perfect sense for Mark to include these Latinisms. Peter seems to have been in Rome in 42, so this suggests an early date for us.

Just how early was it? I think a compelling argument can be made that Mark was written in the late 30s or early 40s. I will not go into all of these arguments, as I think it is sufficient to push Mark earlier by pushing Luke earlier. But for anyone who wants to examine more arguments, these are good. Another similar argument is given for a date of 45, which is also worth investigating. Once we dispense with the prophecy argument, there are good reasons to put it as early as 40, and that is where I will place it also.

This early date does raise a question: why did Paul not make any mention of Mark in his epistles? I think the answer is simply: he didn’t need to. Paul rested on his authority as an Apostle in his writings, and had no need to cite a work written by a non-Apostle to justify himself, even if Mark contains Jesus’ words.


We have very briefly examined some of the reasons why Luke and Mark are dated to be quite late, and shown these reasons to not be good ones. We have also very briefly given some positive reasons for early dates of Mark and Luke, and traditional authorship of Luke. These arguments have, to the best of my ability, been based on objective reasoning which assumes neither the truth nor falsehood of Christianity, and simply engages with the historical evidence.

Further Reading:

Saturday Links 25/8/18


Sorry posts have been a bit sparse lately, I have just started a full time (secular) job. I still need to work out how I am going to manage my time between all of my projects.

John 1:3, NA28, and the Deity of Christ

During a recent discussion with a Jehovah’s Witness, I deployed the standard argument for the deity of Christ from John 1:3. The argument is one that I use in my previous summary of scriptural support for the Trinity. Here is the passage in question, from the NASB;

All things came into being through Him, and apart from Him nothing came into being that has come into being. In Him was life, and the life was the Light of men.

The argument briefly is this: Here the Logos (Christ, clearly, from v14) is the mechanism by which all things came into being. All things which came into being did so through Him. We create two categories of things: Things that didn’t come into being, and things that came into being through Christ. The non-Trinitarians here believe that Christ is something that came into being, so they can’t put Christ in either box. And so Christ cannot have come into being.

However in the most recent edition of the critical text, the NA28, we see that the Greek is rendered slightly differently. You can see the new Greek here, but since this isn’t a textual criticism blog I won’t go into many details. It concerns the placement of some grammatical marks that didn’t exist at all in the original Greek. So the underlying Greek text (no spaces between words, no grammar) remains unchanged, but the scholars now believe that the sentence division should be two words earlier. I am also no Greek scholar, but I will offer a tentative amateur translation of the new Greek here:

All things came into being through Him, and apart from Him, not even one did. What emerged in Him was life, and the life was the Light of men.

You can see here the difference right at the boundary between the two sentences. And my interlocutor claimed that this difference invalidated my argument. They claimed that since the ὃ γέγονεν is connected to the second sentence, we immediately are given an exception to the “all things”: the life and light. And since that exception exists, then we have grounds for thinking that Christ Himself is another exception.

I do not think this is a good counter-argument. I think even the new sentence 1 is very clear: It not only says “All things”, but it still contains the phrase “not even one” (οὐδὲ ἕν). John is doubly emphatic here: all things emerged through Him, and not even one emerged apart from Him. It doesn’t seem like John is leaving a lot of room for exceptions here.

So can the life and the light be an exception? First, notice that the previous sections discuss what emerged through the Logos, but here we are talking about what emerges in the Logos. For us to have an exception, we’d have to have that the light and life emerged in the Logos but not through the Logos. Is that reasonable?

Certainly not, especially when we consider that the life and the life are the Logos Himself. He Himself is the Light, He Himself is the Light. But John describes Him this way because they emphasize two aspects of the Logos: the Light because He opposes the darkness and shines on men, so that we may see God. And the Life, because not only is He the origin and source and guiding principle of all life, but also because by Him we may “have life, and have it to the full”, that we may have true life in this life and eternal life in the next.

The life and light emerged in Him, they became apparent, they were unfolded, because He came down to us. The Logos being the Life and Light are related to the Incarnation. This doesn’t get the Unitarian out of the “not even one”. Through Christ, everything was made. Without Christ, not even one of them was. The clear, natural, obvious reading is that Christ is unmade. Otherwise, there would be no “not even one”.

Although in this conversation I didn’t get a chance, in previous similar conversations I asked my colleague what John could have possibly written here to indicate that Christ really was uncreated and created all things. And their answer was that if the sentence division really was in the traditional place, rather than in the place indicated by the NA28, then that would do it.

Consider what that means. John didn’t write a full stop. John didn’t write using any grammatical marks. John only wrote the Greek letters. My Unitarian friend says that if John had written “…οὐδὲ ἕν ὃ γέγονεν. ἐν αὐτῷ ζωὴ ἦν…” instead of “…οὐδὲ ἕν. ὃ γέγονεν ἐν αὐτῷ ζωὴ ἦν…” then that would be sufficient proof that Christ was uncreated. But think about that, that’s just untrue. Because if John *did* write the first case, then they’d still read it as the second, because John didn’t write with any punctuation!. I ask them what would possibly convince them that they’re wrong, and they answer by telling me that what already exists would be enough. But the fact is that even if John did write precisely what the Unitarian says would convince him, the Unitarian would just move the sentence division around until it didn’t say that. Because that’s precisely what they have done.

And notice that for this entire argument, I have simply accepted the NA28 reading, because I don’t think it helps the Unitarian case at all. In the discussions where this has come up, the Unitarian has, even after I’ve granted the NA28 for the sake of discussion, spent much time trying to justify it. But that’s not the point. I am no Greek scholar, I’ll just accept whatever the experts say. Which at the moment is probably the NA28. But even granting that, we still deduce that Christ is the Uncreated Creator. 

Textual Variants in the Quran

When I debate with Muslims, they often claim that Islam is superior to Christianity because the Quran that they read is identical to the Quran that Muhammad received, while the bible is corrupted and changed and we are uncertain of the original. This discussion often comes down to a discussion of textual variants: differences between biblical manuscripts. And certainly there are textual variants in the Bible, we make no secret of that. And my Muslim friends will claim that since there are textual variants, the bible cannot be trusted.

However my Muslim friends often overlook the fact that there are textual variants in the history of the Quran as well. And if we apply the same standards to the Quran that we do to the Bible, then we must conclude that the Quran is also unreliable. I don’t make any claims here about whether textual variants make a text unreliable, I only want to point out the double standards.

I won’t examine many textual variants today, perhaps I will in future posts. But for now, here is an interesting example: the Sana’a manuscript. This is a Quranic manuscript, but really it is two Quranic manuscripts. It is an 80 folio collection of Quran manuscripts from 578 CE to 669 CE (radiocarbon dating), which was then erased and rewritten in the late 7th or early 8th century. The lower text has been partially recovered, and while it is a Quranic text it contains significant textual variants. A list of some of these can be found here. Not only does it add or remove or change words in some cases, there is at least one case in which an entire verse is missing. There are textual variants here, no-one would deny that.

So if the Muslim tries to undermine Christianity through use of textual variants in scripture, they undermine themselves. Now it is true that the Bible has more variants than the Quran does. But that is the result of two differences in the transmission of the texts. First, the bible was never controlled by a central authority who enforced a particular textual tradition, as the Quran was. Uthman, as is well known, attempted to destroy any copies of the Quran that differed from his own. And after him, an Islamic authority was always policing these other traditions. That doesn’t mean the tradition that they used as a standard was the accurate one, as we’ve seen there existed other traditions pre-Uthman. But they did attempt to standardize it.

Furthermore, there were simply fewer copies made of the Quran. There exist orders of magnitude more New Testament manuscripts than Quranic manuscripts, and so we would expect there to be orders of magnitude more textual variants within the New Testament textual tradition.

But as James White regularly points out, these facts actually make the Biblical text more reliable, not less. Because no central authority controlled the transmission of the Biblical text, there was no opportunity for anyone to deliberately change it. And because there were many copies made, we can compare a greater number of manuscripts in order to reconstruct the originals. The factors that lead to the Bible having more textual variants actually make the Bible more reliable in general.