Bart Ehrman and the Matthew Conflator Hypothesis

The Matthew Conflator Hypothesis

1/18/2015

Thanks Chris Tilling for blogging about my take on Matthew's use of Luke. This type of solution is, of course, not new to the blogosphere. Mark Goodacre responded to a question about Matthew's use of Luke a few years ago. I've added some 'one line' responses to his arguments in the comments after that blog post. If you'd like more detail on any of Mark's points, or have other objections of your own, then please post them here.

16 Comments

Graham Stevenson

1/20/2015 10:44:54 pm

Thank you for this fascinating series of videos outlining the MCH. Here’s my question:

Regarding video three and especially Luke’s behaviour, could you explain a bit more about how the MCH ‘allows the possibility that Luke uses all his sources in blocks’? In video three, from around 6:30 onwards, the same non-Markan material in Luke is said to show Luke 'frequently switching between Q and other sources’ when considering the 2DH, but the possibility of ‘sources in blocks’ when considering the MCH.

Alan Garrow

1/21/2015 04:12:05 pm

Hi Graham. Thanks for your question. Under the MCH we only (for the time being) have access to one of Luke's sources; namely, Mark. This means that we don't know how Luke treated them, but it remains possible that he worked with them one-at-a-time - in the same way that he concentrates on Mark for a while before switching to another source. Under the 2DH we know of TWO sources for Luke: namely, Mark and Q. Luke's treatment of Q is, however, different from his treatment of Mark (he switches between Q and 'other sources' much more frequently). The critical thing to bear in the mind is that, under the MCH, the blue stripes are merely the product of where Matthew chooses to select material from Luke - they have nothing to do with Luke's behaviour. I hope that makes the video clearer.

Paul Davidson link

1/23/2015 08:55:21 pm

Thanks for the excellent presentation. You will have me chewing this over for the next little while. I would be very curious to see your paper.

One brilliant contribution of Goodacre's that makes Markan priority irrefutable (not that we were in any doubt) is the presence of errors caused by editorial fatigue. I wonder if a similar phenomenon can be observed where Matthew copies from Luke.

Additionally, while I understand the preference to work strictly from the texts in their extant form(s), surely the possibility that Luke originally had no nativity story (for which there are several lines of evidence) accounts for the divergence with Matthew equally well.

Alan Garrow

1/25/2015 01:23:26 am

Hi Paul, Thanks for your questions. I don't know of examples of fatigue in Matt's treatment of Luke. I'd be reluctant to rest much weight on this type of evidence (unless there was a really special case) because there is usually an alternative explanation for the data. For example, Mark Goodacre thinks that fatigue explains Luke's clumsy reworking of Matthew's parable of the Talents - but is just as likely that Matthew came across Luke's awkward version of the parable and chose to tidy it up.
With regard to the possibility that Matthew knew Luke without Luke's birth narrative. This is, of course, a genuine possibility. The number of subtle parallels between the two nativities suggests to me, however, that Matthew did know Luke (and sought to conflate Luke with another version). I'd need to read more on the case for proto-Luke to be persuaded otherwise. Any recommendations?

Paul Davidson link

1/25/2015 08:26:50 am

Interesting… I will have to go over Goodacre's examples of editorial fatigue with the MCH in mind and see how they are affected.

Regarding Luke's birth narrative, I have frequently read the claim that the first two chapters read differently than the style of the rest of the book (e.g. more Semiticisms), but I don't have any references handy. There is manuscript evidence for Luke without the genealogy, of course. I can also think of odd terminology differences, like "Nazareth" appearing only in the birth narrative (elsewhere being referred to as "Nazara").

One reading suggestion would be Jason Beduhn's new book from Wester Institute (Polebridge Press), The First Edition of the New Testament: Marcion's Scriptural Canon. He reconstructs the version of Proto-Luke used in the early Marcionite church and discusses some of its implications for the Synoptic Problem.

Alan Garrow

1/25/2015 03:55:07 pm

Thank you for those thoughts and suggestions on Proto-Luke.

Let me know if you find one of Mark Goodacre's instances of fatigue (with Luke using Matthew) particularly convincing. The same applies, of course, to any other readers of this thread.

Paul Davidson link

1/25/2015 05:50:38 pm

One of my favourite examples of fatigue is the Parable of the Sower, which clearly illustrates Luke’s use of Mark. You can read the specifics in Goodacre’s “The Synoptic Problem”, p. 74ff. Essentially, Luke makes two telling mistakes: (1) He omits Mark’s mention that the seed on the rocky soil “sprang up quickly”, yet he gives Mark’s interpretation that “those who hear with joy receive the word”. (2) Luke changes the parable so that the seed withers for lack of moisture; but in the explanation, he forgets this and gives Mark’s explanation about shallow soil, which makes no sense. This makes Luke’s use of Mark a near-certainty.

Anyway, for fun I compared the three versions (Mk 4, Lk 3, Mt 13) side by side to see if anything stood out. After a *very* brief look, a few things stand out.

1. Matthew greatly expands the middle of the parable, not only keeping Mark’s Isaiah quote but fixing it and the quoting it again in longer form. It seems unlikely that Luke used Matthew, because none of this expanded material appears in Luke’s version.

2 There are some very short “minor agreements” between Luke and Matthew.

Mk 4:11 / Lk 8:10 / Mt 13:11: Luke and Matthew agree on “ὁ δὲ εἶπεν” vs. Mark’s “καὶ ἔλεγεν”, and on “μυστήρια” vs. “μυστήριον”.

Mk 4:14 / Lk 8:11 / Mt 13:19: This is kind of interesting. Mark has “word”, Luke amends it to “word of God”, and Matthew has “word of the kingdom”. Given his penchant for replacing “God” with circumlocutions, you could argue that Matthew has Luke’s wording in mind. (On the other hand, it is less likely Luke would turn “word of the kingdom” to “word of God”.)

Mk 4:15 / Lk 8:12 / Mt 13:19: Both Luke and Matthew use the word “heart”.

If none of that is useful, I apologize in advance.

Alan Garrow

1/25/2015 09:00:55 pm

Thanks Paul. A lot of data from a quick look.

The big question is ... are there ever occasions where Matthew's fatigue in using Luke is the only possible explanation for their differences (I don't think there are - but I'd be happy to be proven wrong). Equally, are there ever occasions where Luke's fatigue in using Matthew is the only possible explanation for their differences (I don't think there are - but I'm happy to discuss any examples that others would like to put forward).

Anyway thanks for illustrating what happens when you look at a passage through the lens of Matthew using Luke.

Scott de B.

2/23/2015 07:10:28 am

I found your videos very interesting. Although not a NT scholar (rather a Classicist by training), I enjoy these literary puzzles quite a bit.

I had a few comments. One in particular involves dating, which you don't really address. You argue that Luke is operating from a 'scroll' mentality, while Matthew is operating from a 'codex' mentality. You suggest that Luke is working from (and also writing?) scrolls, while Matthew has access to codices. Matthew is not just later, he is sufficiently later that Mark, Luke and 'Q' have all been converted to codex form and in the same place.

I think that is rather unlikely given the usual chronological range. Yes, there is evidence for codices or codex-like works as early as the late 1st century, but they don't become widespread until much later. However, given the usual dating ranges, the possibility for Matthew is quite narrow. It is frequently held that he is quoted by Irenaeus c. 120. Meanwhile, some scholars (though certainly not all) think Acts shows knowledge of Josephus' Antiquities, so after 97. That isn't really a lot of time for Luke to write, get converted to a codex, then come to Matthew with other works also in codex form. So I'd like to get a sense of what date parameters you'd assign to Luke and Matthew. I take it you are not suggesting a considerably later date for Matthew, such as c. 150?

Alan Garrow

2/25/2015 06:17:32 pm

Hi Scott. Thanks for your question. The most important point, for me, is that Luke and Matthew appear to use different compositional techniques. Less important is why they differ - scroll use vs. codex use is just my best guess. Your point still stands, however. The relative dates of Luke and Matthew is relevant to the discussion.

Matthew is commonly dated prior to Ignatius - but on rather thin grounds. When Ignatius mentions events that also appear in Matthew there is no way of telling whether he got these from Matthew direct or from Matthew's source. Matthew could, therefore, have been composed (I would guess) at any time within 80-135 CE.

Luke's relationship to Josephus is, of course, critical to its date. If the Synoptic Problem has taught us anything, however, it is that confirming the direct dependence of one text on another is exceptionally difficult - even when they share very extensive material. Personally, I favour an early date for Luke - but the evidence is not clear cut.

And this is really the point. Compared to the standard methods for dating Luke and Matthew the evidence of their differing compositional techniques is relatively concrete: Matthew appears to have used a more advanced technique than Luke ... and this suggests (but does not prove) that Matthew wrote some time after Luke.

Paul D. link

2/26/2015 08:53:42 am

@Scott: Irenaeus is c. 180 CE, so that would be the terminus ante quem for Matthew. I am skeptical that (pseudo-) Ignatius shows any awareness of canonical Matthew.

@Alan Garrow: I wonder if you've read John Sturdy's book (published posthumously) on the revised dating of the NT texts. He puts Luke at 110 and Matthew at 130, with well-reasoned arguments.

Gavin

3/23/2015 09:15:15 am

It's so refreshing to hear an argument on the subject laid out with such unobfuscating rigor... At least the best case since Klinghard's. I wonder how well your hypotheses might be harmonized; that is, what if we consider Marcion's gospel in place of GLuke?

Alan Garrow

3/23/2015 06:14:45 pm

Gavin, Thanks for this.

The first question to determine, I think, is whether it is probable that Matthew used a version of Luke. If this does seem likely then there is a follow-up question: which version of Luke did Matthew use? If Matthew used Marcion's Gospel (or some other form of Proto-Luke) then this could explain some of Matthew's omissions from Luke. Having Luke composed in two stages - both before and after Matthew - might also explain some instances of Alternating Primitivity. So, incorporating Marcion into the MCH could offer some useful refinements. The downside of taking this route is that it brings in fresh elements of speculation and complexity. My preference at this stage, therefore, is to work with the broad notion that Matthew used something similar to canonical Luke - while accepting that there is always scope for greater complexity.

David Inglis link

8/13/2015 09:06:58 am

Alan, apologies for the late reply, but I have only recently become aware of the MCH. It is very refreshing to come across someone not rejecting out of hand the possibility that Marcion's Gospel could have been a source for Luke. This is the stance I take in my MwEL (Mark with Early Luke) hypothesis, in which an early version of Luke (with Marcion's gospel being a possibility) is a source for Matthew. However, I then allow for both the early version of Luke and Matthew being sources for the Luke that we know, with the result that Q is superfluous. However, as I am only now watching your videos, I may well have to revise my hypothesis accordingly.

Alan Garrow link

8/13/2015 06:04:05 pm

David, it is equally refreshing to come across someone who uses the phrase 'I may well have to revise my hypothesis'. I am conscious that, ultimately, Marcion needs to be placed somewhere in the picture. I hope you will keep me in touch with your thinking on this.

David Inglis link

8/14/2015 03:03:25 am

Alan, thank you for your reply. Having now been able to watch your Didache videos, it does seem that you have presented a convincing case for at least the 'Two Ways' section being the source of the parallel material in Mt and Lk (another 'middle term'?). However, I am concerned that you may be relying on a text that has been assimilated to Mt, so do you have evidence that this is not the case? Also, as this material is only a tiny faction of what the IQP define as Q, in your opinion what DT material (if any) cannot be explained on the MCH, either by Mt using Lk, or via the Didfache?

Your comment will be posted after it is approved.

the problem page

The Matthew Conflator Hypothesis

Leave a Reply.

Author

Archives

Categories

the problem page

The M﻿atthew Conflator Hypothesis

Leave a Reply.

Author

Archives

Categories

The Matthew Conflator Hypothesis