Wednesday, December 09, 2009

The Synoptic Problem, Part II

Read Part I


As we have seen, wide consensus exists within modern scholarship that Mark was written first, with both Matthew and Luke using Mark as a primary source. The evidence to support this is profound and compelling, while the evidence against it is very weak.

That does not solve the entire Synoptic Problem, however.

At the start of this account, I noted that all three Gospels share a significant portion of their material in common. The material shared between Mark, Matthew, and Luke is commonly referred to as the “Triple Tradition.” This material appears in all three accounts, with Mark being the originator - thus "Markan priority."

However, in addition to containing a majority of Mark’s material, Matthew and Luke also share a significant portion of material that does not come from Mark. This shared material accounts for about 25% of Matthew and Luke’s content. Since it is shared exclusively between Matthew and Luke, it is frequently referred to as the “Double Tradition.”

The stories accounting for the "Triple Tradition" make up 76% of Mark, 45% of Matthew, and 41% of Luke (the percentages are different because the lengths of these three texts are different). Additionally, stories from the "Double Tradition" account for 25% of Matthew and 23% of Luke. Among the synoptics, Luke has the most "unique" material, at 35% of his Gospel's total content.

If we accept Markan priority, as virtually all scholars do, then where did this Double Tradition material come from? Since it isn’t found in Mark, the writers of Matthew and Luke must have gotten it from somewhere else.

As with the issue of Gospel priority in the Triple Tradition, many theories abound to explain the content unique to Matthew and Luke. In fact, this issue of the Double Tradition has occupied scholars far more frequently than the issue of the Triple Tradition.


Beginning in the 19th century, several German scholars developed the theory that this common Matthew-Luke material came from a source no longer in existence – a source used by both Matthew and Luke in addition to their use of Mark. Based on the fact that this common material is almost exclusively made up of sayings attributed to Jesus, these scholars proposed a “Sayings Gospel” – an early text made up of sayings of Jesus, available to Matthew and Luke, but no longer in existence. This source has come to be known as “Q” – which is simply short for the German word quelle, which means “source.”

Over the years, the Q Hypothesis has become more and more widely accepted, and today many scholars base much of their work on the idea that Q existed.

A significant number of scholars, however, have been skeptical of this hypothetical source and the numerous historical conclusions that have been drawn from it. For a period of time in the early 20th century, the theory began to lose steam as scholars tended to think it was too fantastical.

However, the discovery in the 1940’s of a complete text of the Gospel of Thomas gave new life to the Q Hypothesis. Although debate continues to this day over the appropriate dating of Thomas, its discovery proved that at least part of the Q Hypothesis was true: Sayings Gospels did exist in earliest Christianity.

The Gospel of Thomas. This text was found in a cache near Nag Hammadi, Egypt, in 1945. It is a 4th century text written in the Egyptian language Coptic. Most scholars date the original text of the Gospel of Thomas somewhere between 60 and 140 C.E., and originally written in Greek.

Though Thomas is definitely not the lost Q document, the text of Thomas is in the format of a Sayings Gospel. It provides some 100 or so sayings attributed to Jesus, without a chronological framework or any biographical narratives. It’s just a list of things Jesus said – meaning it is exactly like the kind of document that German scholars of the 19th century predicted in regards to Q.

In addition to Thomas’ proof that Sayings Gospels existed in early Christianity, there are a number of other strong arguments supporting the Q Hypothesis.

First, the study of Markan priority has already shown that Matthew and Luke were using other texts to create their accounts. They weren’t just making it up or using stories they knew from oral tradition. Furthermore, in the opening passage of Luke, the writer tells us explicitly that other writers have already written about Jesus, and that he is basing his account on the “investigation” of these other sources. Thus, it is not unreasonable to assume that Matthew and especially Luke might have used textual sources besides Mark.

Second, some of the sayings from Q are repeated virtually verbatim in both Matthew and Luke, indicating very strongly that these stories were coming from a textual source, and not from two different writers independently telling the same oral stories with exactly the same words.

Finally, when evaluating the Q content for its literary quality, certain literary themes become evident. One of the primary themes is apocalypticism. Q imagines God bursting violently into human time to end the material world and inaugurate the kingdom of heaven. Yet neither Luke nor Matthew’s account, outside of the Q material, is nearly so apocalyptic in its theology and eschatology (ideas about the end of the world). This indicates, then, that Q was a separate document, much more apocalyptic in nature than either Matthew or Luke’s account.


The Q Hypothesis is one that is accepted by many scholars. As we saw above, however, there are a significant number of scholars and theologians who doubt Q’s existence. Many of these scholars, instead, accept the theory put forth by British scholar Austin Farrer. The Farrer Hypothesis solves the problem of the Double Tradition by simply suggesting that Luke, writing after Matthew, used both Mark and Matthew as sources. Thus, the Double Tradition material – the material common between only Matthew and Luke – is explained by Luke’s reliance on Matthew’s account. Luke was copying Matthew.

On the surface, this seems like a reasonable argument. It follows in line with the scientific principle of Occam’s Razor. That principle states that the simplest answer is usually the right answer. In this case, it is certainly simpler to just assume Luke used Matthew, rather than to argue a hypothetical source used independently by both Luke and Matthew and no longer in existence to modern scholarship.

Some of the same arguments put forth in support of Q are also put forth in support of Lukan reliance on Matthew.

First, since some of the stories are written identically, word-for-word, in both texts, this could indicate that Luke was simply copying Matthew.

Second, when Luke tells us in his opening passage that he is using other textual sources, it is not unreasonable to assume that those other sources were Mark and Matthew (as opposed to Mark and Q).

Third, there is no obvious reference to anything like a Q document in the writings of the earliest Church fathers. If such a document existed, it was already long lost by the 2nd century.

Finally, there is the aforementioned issue of “minor agreements” between Luke and Matthew. We saw this issue earlier in support of the idea of Matthean priority in the Triple Tradition (the material common between Matthew, Mark, and Luke). It is also used to support the idea that Luke was using Matthew as a source.

Recall that the “minor agreements” argument centers on stories that are shared between all three synoptic Gospels (again, the Triple Tradition). In those shared stories, there are a number of places where Matthew and Luke use the same word against Mark. Thus, the example of Jesus and his captors: Matthew and Luke say he was “hit,” while Mark says he was “struck.” One scholar, Franz Neirynck, has identified as many as 347 instances of these “minor agreements,” including at least sixteen spots where Luke and Matthew agree on five or more words against Mark. For those who support the Farrer Hypothesis, this suggests strongly that Luke was copying Matthew - how else could one explain these frequent word agreements? Extrapolating from there, this would mean that the Double Tradition material (the stories common between Luke and Matthew but not found in Mark) didn’t come from a second text (Q), but simply from Luke copying it out of Matthew. This also would mean that Markan priority still holds, because the “minor agreements” come from Luke copying Matthew, not from Mark editing the other two.


Both the Q Hypothesis and the Farrer Hypothesis have strong evidence to support them. So which one is right?

Ultimately, unless the Q document is found, we probably won’t ever know for sure. But it does seem clear that one or the other is the correct answer, and my personal feeling is that the Q Hypothesis probably best explains all the available evidence.

Consider a few things:

First, supporters of the Farrer Hypothesis have a fairly strong argument with the “minor agreements” of Matthew and Luke. Mark wrote first, Matthew copied, and then Luke copied both Mark and Matthew, frequently using Matthew’s choice of words when telling a Markan story (thus, the "minor agreements").

However, these word agreements could easily be the result of later scribes attempting to harmonize the two accounts. We don’t, of course, have original copies of any of the texts of the New Testament. Our earliest copies are hundreds of years removed from the originals, and while numerous copies of the Gospels exist from the early Medieval period, they have a notorious lack of unanimity. Scribes were hand-copying these texts. Mistakes were to be expected and were very common. In addition, these scribes frequently edited the texts to fit their own particular theology or literary purposes. It is not outside the realm of possibility that many of the “minor agreements” of Luke and Matthew could be the result of scribal editing. Additionally, as we saw above, coincidence could account for many of the “minor agreements,” particularly considering that Mark was writing in “Pidgin-Greek” and Luke and Matthew both write in a higher form of Greek.

Second, there is the issue of no references to a Q-like document by any early Church fathers. While this is an interesting point, a lack of references by folks like Origen and Iranaeus does not preclude the existence of such a text. Early Church fathers were not attempting to provide exhaustive lists of all the Christian texts they knew about. Mostly, their discussion of non-canonical Christian texts centered on those texts they believed to be heretical. There is no reason to suppose that any early Church father would have found Q to be heretical. Furthermore, we know that Sayings Gospels, like the Gospel of Thomas, existed among ancient Christian communities, so a lack of commentary by the Church fathers may simply indicate that Q was already long lost by the time Christianity was becoming institutionalized.

Third, recall the discussion above about literary themes in the Q document. Q is notoriously apocalyptic. Almost every story from Q has an apocalyptic twist. Yet within the Gospels of Matthew and Luke, virtually no apocalypticism exists except in that Double Tradition (Q) material. If you remove the Double Tradition material, neither Matthew nor Luke could be described as texts that are apocalyptic in nature. This lends profound support to the idea that this Double Tradition material is coming from a second source that was theologically slanted toward apocalypticism, rather than from Matthew inserting these apocalyptic stories into his otherwise non-apocalyptic account and then Luke copying him. I can simply think of no good way to explain this without the Q Hypothesis.

Finally, and most importantly, there is the issue of the birth and resurrection accounts in Matthew and Luke. This final point is the real “deal breaker” for me in terms of supporting the Q theory over the Farrer Hypothesis.

Anyone who reads the birth stories of Matthew and Luke will notice immediately that they differ in dramatic and profound ways. There is practically nothing similar between Luke and Matthew’s accounts of Jesus’ birth. Even the genealogy of Jesus between the two accounts is totally different. If Luke was using Matthew as a source, why does his virgin birth story vary so dramatically from Matthew’s?

The same is true of the resurrection accounts in Matthew and Luke. There are hardly any similarities at all. They are essentially two completely different versions of the same story. Again, if Luke was using Matthew, shouldn’t he have followed Matthew in his resurrection accounts?

Recall that Mark’s Gospel contains no birth narrative and no resurrection narrative. Mark begins with Jesus getting baptized, and ends with the women finding the empty tomb and fleeing in terror. Luke’s birth and resurrection accounts, then, could not have come from Mark. If he also was using Matthew, one would expect his birth and resurrection stories to be at least somewhat similar, if not very similar, to Matthew’s. Yet they could not possibly be more different and still be describing the same thing. Why would Luke give precedence to Matthew on sayings of Jesus (the Double Tradition material), but totally ignore Matthew on the birth and resurrection of Jesus?

To accept the Farrer Hypothesis – that Luke used Matthew as a second source – one would have to assume that Luke found Matthew’s depiction of Jesus’ sayings to be compelling and historically accurate, but found his depictions of Jesus birth and resurrection to be totally off-base. Furthermore, since Luke clearly didn’t get his birth and resurrection stories from Matthew, then where, exactly did he get them from? Not Mark. Mark doesn’t contain any such stories. One is left, again, with hypothesizing documents that no longer exist, which Luke had access to, and which he found more compelling than Matthew’s accounts.

As I said, this last point is really the deal breaker for me. Matthew and Luke have such profoundly varying accounts of birth and resurrection that the Farrer Hypothesis of Lukan reliance on Matthew begins to break down in irreversible ways. Since the only other logical conclusion is that Matthew and Luke wrote independently but used the same secondary source for their Double Tradition material, I accept that the Q Hypothesis is the most appropriate answer to this aspect of the Synoptic Problem.


Here is how I believe the formation of the New Testament Gospels most likely took place.

The Gospel of Mark began as a collection of stories about Jesus based on the teachings of Peter. They were perhaps first written down by Peter’s secretary – a man tradition tells us was named Mark. This “Proto-Mark” was perhaps written in the early 60’s C.E.

Around 70-72 C.E., after Peter had died and after Jerusalem had been sacked by the Romans, Proto-Mark was expanded into roughly the form we have today. Its stories were categorized to fit with a portion of the Jewish liturgical calendar, thus making it a Jewish-Christian liturgical text to be used in the synagogue. It fell into circulation among Christian communities over the next ten to fifteen years.

Around 85 C.E., the writer of Matthew took the Gospel of Mark and expanded it, because Mark had only covered about six months of the Jewish calendar. Matthew may have originally been written in Aramaic. Matthew completed what Mark had started, covering all twelve months of the liturgical calendar. As part of that expansion, he included a number of sayings of Jesus that he drew from the now-lost source we call the Q document. He may also have used a third source – called “M” in scholarly circles – from which he drew his unique resurrection and birth accounts. Like Mark’s Gospel, Matthew may also have started out as a “proto” Gospel and was later expanded into the version we know today.

A few years later, around 90 C.E., the writer of Luke took up pen and parchment to write his own account of Jesus’ life. He was not familiar with the Gospel of Matthew or Matthew’s source M, but he did have the Gospel of Mark and the Q document. He also had at least one other source called “L” – representing the material that is unique to his Gospel. Like Matthew with the M material, Luke’s L material informed his unique birth and resurrection accounts.

These “M” and “L” sources, by the way, may have been textual or oral, or may have simply been the writers of Luke and Matthew developing their own Jesus stories based on their reading of the Jewish scriptures.

Finally, around 100 C.E., the Gospel of John was produced, with its writer being at least familiar with Matthew, Mark, and Luke. This familiarity was most likely indirect, meaning that the writer was familiar with the content of the three synoptic Gospels, but did not necessarily have copies of these Gospels on the desk in front of him as he wrote. He also likely used a number of other sources, both textual and oral, with a heavy reliance on the traditions passed down by the apostle John and his followers.


The Synoptic Problem is one that has kept scholars busy for many centuries. Two hundred years of academic research into the Triple Tradition has overturned the long standing conclusion that Matthew wrote his Gospel first. At this point, Markan priority is well enough established that without new and profound evidence, the conclusion is not likely to change.

The theories about secondary sources, however, including the Q Hypothesis and the Farrer Hypothesis (not to mention half a dozen others that I have not illustrated here), practically beg for more research.

If the Q document is ever found, the question will, of course, be convincingly answered. But barring such a monumental archaeological discovery, the question of the Double Tradition in Matthew and Luke will continue to remain a topic of disagreement and debate for years to come in the world of New Testament scholarship.


adiaphthoria said...

Nice assessment. I doubt that the Q document is out there or else it would've been found. I always lean heavily though to the oral tradition possibilities for the Q source. People back then remembered oral traditions more clearly and accurately than we do today, because they didnt have the luxury of jotting down a note or too to help us remember. Thus, Matthew and Luke could still have strong consistencies IF Q is only an oral tradition. As far as the M & L theories go, I favor the view that Matthew and Luke had their own independent stories to share, rather than just see them as copycats who plagarized Mark, the Q tradition, and then someone else as well. Luke especially is a great writer, and great writers dont just "borrow" from everyone else, but regardless of what their work is get their own unique voice into whatever they are writing. At least, that's what I try to do.

Scott said...

Honestly, after doing this in-depth look at the various theories, I am beginning to lean away from Q ever so slightly. In the past, I felt like Q was a pretty solid explanation, but as I've delved deeper into it, it seems that neither Q nor Lukan reliance on Matthew really seems to adequately explain all the evidence.

It's possible that Q was an oral tradition, but I think J.D. Crossan has done a fantastic job of showing how that is unlikely, in his book The Birth of Christianity. He devotes a chapter or two just to the anthropological studies of illiterate story tellers of Albania in the 1930's (no, I'm not kidding). I won't go into all the detail, but he shows how oral traditions depend on keywords and key phrases, more so than entire passages of word-for-word memorization. It seems unlikely that the sometimes very long, virtually identical "Q" passages of Matthew and Luke would come strictly from oral tradition.

I agree that the "M" and "L" sources may very likely simply be original creations of Matthew and Luke. I totally agree that there is no reason to assume that they copycatted every single story they told. In fact, I would say that's highly unlikely. Perhaps some of the "M" and "L" material came from other sources, but I'd guess that much of it was their own midrashic creations based on their own study of Jewish scripture.

John C. Poirier said...

I know this comment is really late, compared to the date of your post, but anyways:

Luke's non-use of Matthew's birth narrative is easy to explain on the terms of the Farrer hypothesis: Luke, like the vast majority of Christians in his day, would have been horrified at Matthew's implying that Jesus' birth was astrologically determined. Jesus' ascendency over the sidereal powers, in fact, was a very big part of many early Christians' understanding of the Christ event. Luke probably would have reacted very strongly to Matthew's birth narrative -- as indeed many later Christians found that they had to explain away those same astrological references -- and he simply replaced the whole thing.

Scott said...

Thanks for your comment John.

Yes, going with the Farrer Hypothesis, we could speculate that Luke disagreed with Matthew's depiction, and so changed the story.

One of the whole problems with the entire discussion, of course, is that it sort of presupposes that the Gospels as we have them now were in essentially the same form in the 1st century. For instance, Luke might have used Matthew (as the Farrer Hypothesis says), but the "Matthew" he used didn't include the birth story, so Luke inserted his own.

As I have argued in other essays, I think Matthew, in particular, is probably heavily edited and redacted from some original "proto-Matthew" document. I've even gone so far as to suggest that "proto-Matthew," might, in fact, be the Q document.

Serene Musings Books of the Year, 2005-2015