“Q” was a forgery

My purpose here is to argue for a specific solution to the synoptic problem. I argue that Mark was written first and had probably been around for at least 25-30 years before the next gospel was produced. In order to justify a new gospel, the author of the gospel of Matthew forged a saying source, and attributed it to the disciple Matthew. The author then used this forged “source” and the gospel of Mark to produce his gospel. Finally, Luke was taken in by the forgery, and accepted it as authentic. Luke was also aware of the gospel of Matthew, but because he regarded it as a contemporary work, declined to use it extensively. Instead Luke created his gospel mostly from the forged saying source, and the gospel of Mark.

Obviously, arguing that a document we do not have not only existed but was forged is a difficult task. Never-the-less, I think it can be argued to be the most probable solution to the synoptic problem. Here I begin to make that attempt.

The testimony of Papias

Papias is one of the only early external pieces of evidence we have regarding the synoptic problem. Almost all the evidence we have is internal to the gospels themselves. Papias tells us that Matthew wrote his LOGIA first.  Thus it has been argued that this testimony is consistent with an early saying source, by the disciple Matthew, which was used by the later gospel writers. However, a successfully forged early saying source, and an authentic early saying source, will have much in common, including Papias’s testimony about them. Thus although Papias testimony is itself questionable, my hypothesis is completely consistent with this bit of external evidence.

Relationship to the “Two source hypothesis”

Most New Testament scholars today subscribe to some version of the two source hypothesis. This states that Matthew and Luke both used the gospel of Mark, and both used a saying source “Q”. However, Matthew and Luke were independent of each other.

My hypothesis differs only slightly from this, and thus most arguments that support one idea will also support the other. I argue that Luke had access to the gospel of Matthew, unlike the two source hypothesis, but on the other hand, I think Luke made only very limited use of Matthew, and that use of Matthew and use of Q will be hard to distinguish because both had the same author. In another difference from the two source hypothesis, I claim that the saying source was created by the author of the gospel of Matthew. But, a successfully forged saying source, and an authentic saying source will have much in common, so again much of the evidence that supports one idea will support the other.

However, my hypothesis helps explain some things that are difficult to explain given the two source hypothesis. The so-called “minor agreements” between Matthew and Luke are a major issue for the two source hypothesis. However, if Luke has access to Matthew this ceases to be an issue. The minor agreements between Matthew and Luke, where they are both using Mark, can be explained by Luke approving of Matthew’s corrections to Mark’s Greek, and Luke’s memory of Matthew’s phrasing, while he is working with Mark.

My hypothesis also does a better job explaining the “Mark/Q overlaps”, where the two source hypothesis has some difficulty. These are sections where the two source hypothesis claims that both Mark and Q had the material. At the very least it seems that the two source hypothesis must invoke additional hypotheses to explain the relationship between Mark and Q. I plan to argue that some of the so-called overlaps are late corruptions to the text of Mark, where Mark has assimilated to Matthew, and Luke has relied on the text of Matthew and/or Q. In other cases I will argue Mark’s text is original, and the overlap is caused by Luke relying on the text of Matthew and/or Q. If Q had this material, I will argue it is secondary to Mark.

Finally, my hypothesis helps make sense of the results of my statistical study, found here:
My study finds a significant statistical relationship between the vocabulary of the sections of Matthew that are unique to his gospel, and the vocabulary of the material attributed to Q. The two source hypothesis could be consistent with this, if Luke had ignored large sections of Q, thus causing them to be found only in Matthew. However, the relationship does strongly suggest that there is no real vocabulary distinction between Matthew and Q to be found here. Supporter of the Farrer hypothesis, like Mark Goodacre3 for example, have made the same case, in a different way, arguing that the interests and vocabulary of Matthew and Q can not be separated.

My study also finds that the minor agreements look distinctly Matthian in their vocabulary, but they don’t show up as related to Luke. This suggests Luke is dependent on Matthew in the minor agreements. However, this does not constitute proof either. The two source hypothesis would claim that the minor agreements should look both like Matthew, and like Luke, and although the study fails to find a significant relationship with Luke, this is not the same as saying it finds significant evidence that there is no relationship.

Finally Kenneth Olsen4 cites Sharon Lea Mattila5 as arguing that “the procedure Matthew must have used to rearrange and conflate Mark and Q in detail seems more complex than might be expected in an ancient author”. However, according to my hypothesis Matthew and Q were produced together as a unit. The author of Matthew wrote his gospel straight from Mark, and created a saying source to justify this new gospel at the same time.

There are some objections to the idea that Luke knew and used Matthew, but most of these objections depend on the idea that Luke used Matthew directly for all of the double tradition Material. It has been argued that Luke would be unlikely to disassemble Matthew’s sayings, and reassemble them into the large blocks where they are found in Luke, and that Luke would not have completely re-written the birth narrative. But if Luke was working primarily from a saying source he regarded as authentic, and did not like the contemporary Matthew, this argument has no force. Luke tells us, in his prolog, that he is interested in witnesses, original sources, and facts. So mostly ignoring a contemporary source would be consistent with this. In addition, if Luke agreed completely with Matthew, he would not have been motivated to write, and we can see from what he did write, areas where his view and that of Matthew diverged.

In short, I believe my hypothesis can account for all the data accounted for by the two source hypothesis, and can better account for the minor agreements, the overlap texts, and the statistical relationships discovered in my study.

Comparison to the Farrer hypothesis

My hypothesis also shares much in common with Farrer hypothesis. This hypothesis claims that Mark was first, and that Matthew used Mark only, and that Luke then used both. While not supported by the majority of scholars today, it does have a strong following among those scholars interested in the synoptic problem. And, like the two source hypothesis, it fits fairly well with my statistical results that show Mark to be first, Matthew second, and Luke third. My hypothesis will differ from this one only slightly. The Farrer hypothesis says Matthew worked from the text of Mark, and I would agree, although I would add that Matthew also claimed to be using a saying source. This could help account for the large number of doublets in Matthew. If “Matthew” wanted to have his saying source viewed as authentic, he would have to include not just new material, but also material found in the gospel of Mark, in order to lend authenticity to the saying source. Then when “Matthew” wrote his gospel from this “source”, and from the gospel of Mark, the Markian quotes would show up both in their original Markian context and mixed in with the new “Q” material.

Luke’s behavior will also be similar on both hypotheses. On both, Luke is using Mark, and one other primary source, that is written by the author of the gospel of Matthew.

Why not just say that Luke used Matthew and forget about the saying source? My primary issues here is that Luke shows “respect” for the text of Matthew in some ways, but not in others. Luke tells us in his prolog that he is interested in witnesses, and sources, and facts. So, what is Luke’s attitude towards the gospel of Matthew? Is Matthew a disciple or someone in a position to know the facts? Or is Matthew a contemporary of Luke’s with no better access to the facts than Luke himself? The fact that Luke uses many verbatim quotes from Matthew, sometimes in favor of the wording found in Mark, seems to indicate that Luke thinks Matthew is in a position to know. On the other hand, the fact that Luke shows no regard for Matthew’s order of events seems to indicate that he does not view Matthew as being in a position to know the facts. Luke easily could have left the additional material from Matthew, in roughly Matthew’s order, but he does not do that. This brings to mind a saying source which could easily have authoritative historical wording, without authoritative historical order.

In addition we can look at Matthew’s behavior. By the time this gospel is written, we know from Paul that there is already a concept of Christian scripture. In that environment, writing in the late 1st century, does Matthew have a reasonable expectation that his gospel will be well received and read if he does not claim an authentic source? I would argue that this would be unlikely. This brings us to the last competing hypothesis to be considered.

Comparison with the three source hypothesis

The three source hypothesis says that Matthew used both Mark and an authentic saying source, and that Luke used all three. One merit a forged saying source may have over an authentic saying source here is that it offers a simple explanation for Mark’s failure to use the saying source in his gospel. But in general this hypothesis is very close to mine, and thus there is little that can serve to separate them. One approach would be to see if the material can be divided up into “Q-like” material and “Matthew-like” material.Ron Price attempts that here -
Recently we tested his proposed division with my statistical method -
Here thanks are due to Ron, for doing the needed data entry. The results showed a relationship between the vocabulary of Ron’s sQ, and material found only in the gospel of Matthew. But these results were, in the end, inconclusive. The most important issue was probably that we could not separate the effects of genera i.e. “sayings” vs. narrative, from the effects of authorship on vocabulary. This might cause sayings by Matthew and sayings by “Q” to have a similar vocabulary of say pronouns. A look at some of the details in the study, however, suggested that at least some of Ron’s “sQ” sayings still contained Matthian themes and vocabulary. But Ron would argue that according to his hypothesis Matthew had an early saying source, and followed it extensively simply because he agreed with its themes, and this would account for thematic words being similar across both. Together these possibilities could easily account for the observed relationship. I would also note that according to my hypothesis Ron could be substantially correct in his reconstruction of the saying source, which on my hypothesis would be expected to share a common vocabulary profile with Matthew. Finally, even if we were to accept the suggestion that Ron’s partition does not correctly separate Matthew from Q, it could still be the case that another partition would succeed here But again the result of the study must be viewed as inconclusive, and at most the work suggests that separating Matthew from Q is difficult to accomplish.

Overlap texts

A more promising approach is a careful study of the so-called Mark-Q overlap texts. Here I would hope to show that the text of Q is secondary to the text of Mark. In some cases I will try to show the text of Mark is more original. In other cases I will attempt to show that the text of canonical Mark is dependent on the text of Matthew and/or Q but this is a corruption of the original text of Mark. I plan to do a complete survey of these agreements, but here I’ll just give a quick overview of a couple of cases.

The first case is based on my study of the use of “salt” in Mark.
Here I argue Mark has crafted an intricate reference to Leviticus. It is implausible that Mark was using a corrupted Q saying here, and conflating it with a reference to Leviticus. I argue that Mark has the original version of this saying and that Matthew is a deliberate rewrite. But then we can look at Matthew’s use of the word “MWRANQH” (or "Moranthe"). It has been argued1 that Matthew got this from an Aramaic saying source which had a word with the root “TPL”, and meant “becomes tasteless” The theory continues that Matthew then mistranslates it to mean “becomes insipid”, and thus chooses to use the Greek word “MWRANQH”, “foolish”. However, I would argue that Matthew wrote “foolish”, because he meant “foolish”, that is he was deliberately re-writing Mark. If that was the case, then Matthew would have most likely been well aware of the similarly between “foolish” and “tasteless” in Aramaic, and probably used this fact to help justify his changing the gospel of Mark’s “salt-less” to the gospel of Matthew’s “foolish” and the re-contextualization of the salt saying.

Connected with the salt saying, Matthew has a saying about the lamp and the lamp stand (Mt. 5:14-16). And in Matthew we are told that the disciples are the light of the world. I think it can be argued, however, that in Mark we have another apparent fulfillment of Hebrew scripture by Jesus. In Mark 4:21-23 the lamp is almost personified. Rather that being brought in, the Lamp “comes”. (MHTI ERCETAI O LUCNOS INA UPO TON MODION TEQH). Then in Mark 10:35-40 the sons of Zebedee ask about places at Jesus’s right and left, but are informed that they are reserved for the ones anointed. In Mark 15:27 we then have the bandits placed “one on his right and one on his left”. From this in can be argued that Mark is evoking Zachariah 4:1-14. There we are told about a lamp on a lamp stand, and two olive trees, “one to the right and one to the left” and that these are the two anointed ones in attendance on the Lord of the whole world. Mark’s point then would be that Jesus is the Lamp on the cross(the lamp stand), with the other crosses(olive trees), one to the right and one to the left of him, and the bandits then are the anointed ones in attendance on the Lord of the whole world.

Crafting a fulfilled prophecy like this would take some creativity by itself, but it is highly improbable empirically that Mark had to work from both a saying source and Hebrew scripture and was able to produce this. Far more likely is the hypothesis that the Lamp saying originates in Mark. And this supports the idea that both “salt” and “light” are intentional re-writes in Matthew. Why might Matthew have objected to the text of Mark here? Perhaps readings of the text of Mark which were tending toward the ideas expressed in the gospel of John, or even by proto-Gnostics were too radical for the more conservative Matthew. Identification of Jesus as God and rejection of Hebrew scripture may have been developing trends when Matthew wrote, trends that he wished to counter.

Thus for both the salt and the light saying which are connected in Matthew, we have a good argument that the text of the “saying source” is secondary to the text of Mark

Another example is the Beelzebub controversy in Mark 3:22-30. First of all there are signs within Mark, that this has been inserted. It separates the arrival of his family from the statement about family. Also the Holy Spirit is probably a later development. In Mark the only place this term appears is in Mark/Q overlap texts, and it is not in Luke’s parallels. The one exception is Mark 1:11 where the Spirit descends on him like a dove, but this is a rather special use of the term.

It also seems that Luke probably did not have this text in his copy of Mark. There are no Luke/Mark agreements against Matthew in this section, although they are typical elsewhere. Also Luke places this text outside of its context in Mark, and groups it with material he got from Q/Matthew. I think it is sometimes missed that there is no reason, a priori, that these actions have to coincide. That is, Luke could follow the order of Mark and use the wording of Matthew/Q, or use the wording of Mark and the order of Matthew/Q. The combination of these actions suggests that Luke did not know this text as part of Mark. That combined with the fact that it lines up exactly with what looks to be an interpolation makes it highly probable that this is a late addition to Mark, which was not in Luke’s copy of Mark.

It has been noted that Mark’s style seems to be one of “pseudo-interpolation”. But here I think we need to ask if this is really the style of Mark, or if this just appears to be the style because there have, in fact, been a number of interpolations. In any case, even if we accept pseudo-interpolation as Mark’s style, it is still rather coincidental that one of these lines up exactly with text that Luke seems to be ignorant of.

And finally here we have Fleddermann’s arguments2, which indicate that the text of Mark here is later than Q (that is later than the text reconstructed from Matthew and Luke), which also helps support my argument that this is an addition to the text of Mark. Fleddermann argues (p.59) that “’your sons’ concedes too much because it implies that Jesus operates on the same level as his opponents. The developing churches developing Christology could not admit such equality”. Thus Mark’s text is more likely secondary here.

It might be argued that given that we have no surviving texts of Mark with this reading, that interpolation here is improbable. But, I think we could easily demonstrate, given the changes in surviving copies, that we are unlikely to have complete originals of any of our texts. Some changes would have gotten into all surviving copies of all of our documents. Also, we can expect the rate of change to have been quicker, early on, before things became well established as scripture. And while there were fewer copies in existence an alteration could achieve full replacement in all existing copies with greater ease. In addition, changes would be more likely in a document that was less well suited to developing Christianity, which means the impetus to make changes in the first gospel would be the greatest. And finally, Mark has the most variation in surviving texts of any of our NT documents, also indicating it has experienced the greatest amount of change.

In most cases these originals are completely lost. However, in the case of the gospel of Mark, we do have witnesses to the text of Mark outside of the surviving copies of the gospel of Mark. We have the texts of Matthew and Luke. We can view them as non-conservative early witnesses to the text of Mark, and view our surviving copies of Mark as conservative late witnesses. In this case, our non-conservative early witness, Luke, gives us two independent indications that he does not know Mark’s text here. Then our late conservative texts, our copies of Mark, show internal signs of alteration. We can then argue for assimilation to the more popular text of Matthew.

The final example, for now, comes from Mark 1:2-3. The quote is erroneously attributed to Isaiah here. However, if the seemingly free-floating bit of overlap text (Mark 1:2) is removed then Mark’s text is correct. Mark reads better with the overlap text removed. Fleddermann2 also points out that the use of the word SOU here is not quite suited to its context, and probably reflects its original “Q” context. Given that the original author of Mark seems to have extensive knowledge of Hebrew scripture, it seems more likely that a late editor would make this addition, rather than the author of Mark, himself. Together, this argues that the original text of Mark has been expanded here, and that assimilation to Matthew/Q and/or Luke has taken place.


Thus case-by-case, I hope to show that “Q” is a later document than Mark, and that this combined with its Matthian themes and vocabulary suggests that it was a forgery in the name of the disciple Matthew, in order to justify the production of the gospel of Matthew.

1. Black, Matthew (1967), “An Aramaic Approach to the Gospels and Acts”, Oxford University Press

2. Fleddermann, H.T. (1995), “Mark and Q, a study of the overlap texts”, Leuven University Press

3. Goodacre, Mark (2002), “The case against Q, studies in Markian priority and the synoptic problem”, Trinity Press International


5. Sharon Lea Mattila, “A Question Too Often Neglected,” NTS 41 (1995)
199-217, p. 206.

Back to synoptic main page