Wednesday, October 14, 2009

Exploring our Matrix Explores the Case Against Q

James McGrath and I have often sparred on the Synoptic Problem on our blogs, and it is nice to see a post today on Exploring our Matrix about my book The Case Against Q, headed Beyond Literary Dependence and Independence. It's nice to see James saying some nice things about the book and even nicer to have some critical engagement. I would like to make a few comments on James's comments.

One of the thrusts of James's post is to question the idea of Luke's direct, literary dependence on Matthew,
. . . even after reading the book I still find it unlikely that Luke used Matthew's Gospel as a source in the sense of having it open in front of him as he composed his Gospel (Italics original).
I think the topic of the mode of one author's use of another text is a useful one. I have not claimed myself that Luke had a copy of Matthew "open in front of him", though I think that that is likely when it comes to passages of very close verbatim agreement, which are common in the double tradition. But even there, I would be wary of the idea of the author-as-scribe. I am more inclined to imagine Luke having a copy of his source text in his hands and dictating to his scribe. Nor would I rule out the use of memory. I think Luke knew his sources very well, but he knows Mark much better than Matthew, which is why he consults Matthew directly more often. In his composition, I would imagine that Luke's use of his sources varied, sometimes close, sometimes free, sometimes in between. In the case of the Birth Narratives, my guess would be that Luke did not have a copy of Matthew's Gospel in his hands, and that the agreements are the result of his memory of Matthew's text.

The differences between Matthew 1-2 and Luke 1-2 are clearly important to James and I doubt that there is anything extra that I can say here that would persuade him of my view. I would just underline that history and tradition have generally preferred Luke's version and so his revision of Matthew's idea, if that's what it was, was a success. I would again mention that contradicting source material does not necessarily indicate non-use of the source and, moreover, it is worth asking how it was that Luke came up with the idea of appending a Birth Narrative to the Marcan structure? It is particularly odd to find Luke making this move when he clearly sees the "beginning" of the key events as marked by the baptism of John (Luke 3.1-2, Acts 1.22, Acts 10.37).

I am pleased to see that James welcomes my "recognition that the dichotomy between literary dependence and independence of Gospels is a false one". I tend to think that the Q hypothesis is actually stuck in a literary mind-set because of its inclination to project everything onto a hypothetical document, so that tradition-history often gets confused with literary priority.

But James asks about the possibility that oral tradition might have conveyed Minor Agreements between Matthew and Luke like that against Mark 14.65 ("Who is it that smote you?"). Of course this is possible, but I think it is problematic in this context. The argument from the Minor Agreements is a specific response to the claim that Luke never agrees with Matthew against Mark. It is hardly an answer to that data-claim to say that the agreement could have been mediated through oral tradition. The admission nullifies the claim that Matthew and Luke do not agree together against Mark. The fact is that they often do. And regularly, it is agreement of the kind that we would normally ascribe to simple redactional reworking, and not to variant tradition. To put it another way, if we had only Matthew and Mark here, would we hesitate to say that Matthew has simply redacted Mark? It is only the presence of agreement with Luke in a theory that postulates independence that is leading to the oral tradition claim.

As it happens, I think that verbatim agreement in Greek does need to be taken seriously. A string of five words including one that is rare is striking. Similar strings of verbatim agreement also often indicate direct links between texts. James's example of Mark 2.9 // John 5.8 is a case in point. It is close and may be a good case of John's familiarity with the Synoptics, all the more so as the NA27 text of Mark is much closer than in the TR that James quotes:

Mark 2.9: Ἔγειρε καὶ ἆρον τὸν κράβαττόν σου καὶ περιπάτει

John 5.8: Ἔγειρε ἆρον τὸν κράβαττόν σου καὶ περιπάτει.

This is a six word verbatim agreement, seven if not broken by καί, including an unusual word, κράβαττόν. It's the kind of material that may show familiarity (not direct "literary dependence") between John and the Synoptics (cf. my earlier post about the direct links between the Synoptics and John 12.1-8).

One more point. James mentions that he sticks with Q as a "working hypothesis". My difficulty with this kind of argument for Q (cf. Case Against Q, 75-77) is that a hypothesis is not tested by its usefulness. All sorts of erroneous hypotheses are useful and, on a certain level, they work. Seven-day Biblical creation was a useful working hypothesis for millennia. The reason that erroneous hypotheses remain persuasive is often because they work; they are attractive; they appeal to culturally determined elements in our thinking.

Thanks again, though, James for some enjoyable reflections and for taking the time to think seriously about my stuff -- a real compliment.


Scott F said...

"A string of five words including one that is rare is striking"

Might not the presence of a rare word aid in the verbatim recall of an entire phrase. The unusual wording would make it stick in your mind much like bits of Shakespeare of the KJV do.

James F. McGrath said...

Thanks for replying, Mark! And well done spotting the evidence for authorial fatigue in my blog post - I went with a version of the Greek text readily available online rather than be bothered to type out the NA text. :)

I certainly agree with you that it is, at least in theory, possible to explain difference and disagreement in a way that also envisages use of the very literary source that is being disagreed with (if that makes sense). At the moment, I'm not sure that I am confident in our ability to tell with confidence which of the many scenarios I mentioned in my post is in fact the case. And so I do hope to do some more work in the near future on that methodological issue.

I wonder whether there would be interest in something on this specific topic at SBL, either under the auspices of an already-existing section, or as something new...

James F. McGrath said...

Oops, I neglected to subscribe. I'm now doing so - feel free to delete this!

Mark Goodacre said...

Thanks, James. There are some nice NA26 texts available on the web -- I share the fatigue!

In discussions with you, we tend to focus on the big differences between Matthew and Luke and that is fine because it is clearly a sticking point for you, and one that I need to address in my own thinking. But in terms of general approach, I would never begin from the areas of major difference and reason from there. In discussions about links between texts, you have to begin from the areas of high verbatim agreement and see where they lead. When I have had cases of plagiarism by students, I establish the case that they have copied by means of looking for the highest verbatim agreement sections. What I find interesting about the Synoptic Problem is that range of agreement, from 100% all the way down to a few percentage points. How we explain the sliding scale is, of course, the heart of the problem. But our best bet for establishing direct links lies in the material at the high verbatim level.

Mark Goodacre said...

PS Definitely topics worth addressing. I think that there is a lot of unclear thinking around on the question of orality and textuality, memory, text and oral tradition. Not sure I'd be too keen on a whole new SBL section, but perhaps worth exploring.

Mark Goodacre said...

Scott: good point. I think this supports the case since recall of "bits of Shakespeare or the KJV" are of course recall of texts.

James F. McGrath said...

Mark, you are of course right that if we are to prove use of a written source, verbatim agreement is the really decisive evidence(I should note that, in a strange quasi-Synoptic way, I began replying to you in the comment section on my blog before realizing that your comment was posted here).

Part of the problem is that it has been hard to quantify exactly what sorts of and extent of agreement proves a literary dependence, while taking seriously the fact that divergence in no way proves independence. And, in relation to another comment on my blog, often the argument comes down to what we can or can't imagine an author doing, which isn't a particularly strong argument - and, as you Mark show in your book, sometimes the problem is a mistaken impression of what would have been involved, rather than any genuine unlikelihood that an author would use a source in a particular way.

I think there is evidence that ancient authors, even when they used a source, rarely copied it word for word and line for line - although sometimes they did so. And so one subject I want to investigate further in the near future is whether we have any hope of telling the difference between various scenarios, such as (1) a situation in which Luke used Matthew but read a whole pericope, then turned away from the text of Matthew to write his own version without looking back at it or only doing so occasionally, and (2) Matthew and Luke using a common no-longer-extant source independently of one another. Since there is a range of agreement from slight to substantial, the question is what proves literary dependence, especially since such dependence could involve a range of forms from copying word for word to reading from Matthew and writing one's own version a few days later.

We could have a whole section as SBL dedicated to actually reading things to people in one session and then testing their memory of what they heard in a subsequent session - now that could be fun! :)