Streeter's Four Document Hypothesis
(CORRECTED Nov 8) Students will find various online resources and many downloadable ones at BiblicalStudies.org.uk including pdfs of Streeter's famous The Four Gospels: A Study of Origins treating of The Manuscript Tradition, Sources, Authorship, & Dates.

It's funny that I never got around to reading Streeter's book (though I feel I have indirectly through others) so I haven't previously discussed Streeter's source theory here.

Streeter's influential take on a solution to the Synoptic Problem (first published 1924) is found in chapters 7–12. Much of the 20th century (in the English-speaking "West") was either based on an acceptance or rejection of Streeter's "four-document" solution. I say influential, but his four-document solution is also often misrepresented as though Luke used a Q document.

The first thing that strikes me in looking through the book is that it looks like Streeter's Luke doesn't have a copy of Q. In Streeter's chart (p.150) Luke only knows Q through his copy of "proto-Luke" (whereby Q and L are already combined).

I'm sure I have been misled by others' reading of Streeter's source theory. When Streeter talks about a "four-document" hypothesis he has two documents in mind used by Luke (proto-Luke & Mark) and three documents used by Matthew (Q, Mark and M) thus four source documents (plus other non-documented traditions) in total underlying both of these Gospels (namely: Q, Mark, M, proto-Luke) as diagrammed on his page 150.

However, I remember being taught that Streeter's "four-documents" consisted of Q, Mark, M & L even though L is not a document known or used by Luke.

According to Streeter's diagram one would not think that Streeter's Luke used Q directly (rather than only via his proto-Luke) but on page 218 he says “this same Luke some years afterwards expanded his own early work by prefixing the stories of the Infancy and by inserting extracts from Mark no doubt at the same time making certain minor alterations and additions” so it seems Streeter's own diagram misrepresents his own theory (unless we are to assume "this same Luke" lost his copy of Q? 

What was Mark for Matthew?

And in contrast to the previous two posts relaying Sim's Matthew’s Use of Mark: Did Matthew Intend to Supplement or to Replace His Primary Source? a new book on Matthew's relation to Mark by J. Andrew Doole apparently gives different answers to these questions, namely 'yes' to supplement and 'no' to replace:

J. Andrew Doole examines Matthew's sources, which the evangelist used to compile and compose his own story of Jesus. Doole suggests that Matthew was not disputing the Gospel of Mark, rather developing its tradition in a conventional manner to reinforce its authoritative position in the growing Christian movement.

I'm hoping Doole is in close dialogue with Sim's 2011 NTS article. The question remains interesting and I don't believe it is easy to answer completely one way or the other.

Matthew to Replace Mark?

An article by David C. Sim, "Matthew's Use of Mark: Did Matthew Intend to Supplement or to Replace His Primary Source?." New Testament Studies, 57, pp 176-192

Abstract:

Most scholars acknowledge Matthew's debt to Mark in the composition of his own Gospel, and they are fully aware of his extensive redaction and expansion of this major source. Yet few scholars pose what is an obvious question that arises from these points: What was Matthew's intention for Mark once he had composed and circulated his own revised and enlarged account of Jesus' mission? Did he intend to supplement Mark, in which case he wished his readers to continue to consult Mark as well as his own narrative, or was it his intention to replace the earlier Gospel? It is argued in this study that the evidence suggests that Matthew viewed Mark as seriously flawed, and that he wrote his own Gospel to replace the inadequate Marcan account.

That the author of Mt made much use of Mk indicates a degree of acceptance I presume Sim also mentions this point, I'll peruse the article next visit to college.

The Problem Never Completely Resolves
We cannot say much about what exactly the sources looked like that preceded the synoptic Gospels, only that that direct dependence does play a large part in the end result. Even if we do conclude that Mark is the earlier written Gospel it seems that Mark is still somewhat secondary to earlier sources. Mk 13 for instance looks to be older material very unlikely to have been freshly penned by the Gospel author. Also I have previously indicated my own theorizing that most of the healing stories apparently predate our written Gospels, although in these cases they would likely have existed only in oral form. So Mark is not necessarily the oldest 'source' of shared material.

So students should be aware that hypothesizing a chronology for dating Matthew, Mark and Luke still does not completely (re)solve the problem of Gospel sources. It merely gives us a simplified ‘working hypothesis’ for supposing how a particular Gospel author may have put their own stamp on the material which we suppose to have been already available to the author (and in many cases material likely already known by the audience).

I'm perpetually agnostic concerning 'Q.' How is it that the material in Matthew and Luke concerning John the Baptizer is written virtually word-for-word? This would be the result of someone copying slavishly from a written source. So if the text in Matthew is not being copied here into Luke (or theoretically from Luke into Matthew) whether by a secondary Gospel author (or inserted by a scribe within the first hundred years of copying) we must suppose that both have here accessed the same written source concerning John the Baptizer. Yet the other shared material in Matthew and Luke (and not found in Mark) is less likely to be from the same shared written source since everywhere else the doubly-shared material in Mt-Lk is phrased independently by both authors making it impossible to know much about the immediate source of such traditions or whether these traditions originated from the very same source as the John-Baptizer material. How can we conclude the same way (all stems from a singular written source?) with this differing evidence?

So the precise sources of all the triply and doubly shared traditions are still largely unknown even after having ‘solved’ the problem of the likely order the Gospels were published. There will always remain the problem of whether an earlier version of Mark, or Matthew or Luke was known or accessed by any of the other Gospel authors (aurally or in written form) on top of the problem of other pre-Gospel traditions.

A Year of Source Theory Blogging - Part Two (Dunn)

Recapping my four previous posts on James Dunn's source theory [Aug 27; Oct 14; Dec 9; Dec 28]: Dunn's theory has an unresolved tension in that both Mk and Q are regarded as simultaneously written and oral sources. Whether Dunn regards 'Q' as being an oral source (or sources) is still not completely clear--Dunn has only explicitly argued that the first layer of Q (Q1) be seen as 'oral ' [see his "Q1 as oral tradition," in Markus Bockmuehl and Donald A. Hagner (eds.) The Written Gospel (Cambridge: Cambridge University Press, 2005), 45-69]. I concluded humorously that "It seems Dunn wants to have his Q and eat it too!" Perhaps Dunn eventually intends to demonstrate that the remainder of 'Q' should also be seen as oral.

However, what is more surprising is Dunn's article "Matthew's Awareness of Markan Redation," in F. van Segbroeck et al (eds.), The Four Gospels--1992: Festschrift Frans Neirynck, (3 vols.; BETL 100; Leuven University Press), 2:1349-59, in which Dunn argues that Matthew recognised Markan redaction in Mark and so avoided it consciously. Dunn appears to suggest that this indicates that Matthew was already familiar with many of the stories written in Mark and/or that Matthew used his Markan source in an 'oral mode.'

I am not aware whether other scholars have drawn attention to this but it seems to have anticipated Delbert Burkett's assertion that Markan redaction is suspiciously absent is Matthew and Luke. Whereas Burkett sees such an absence as undermining the notion that Matthew and Luke used Mark as a source Dunn sees it as evidence that Matthew could readily recognise Mark's own redaction of oral traditions. There is something persuasive about such an argument and I wish other scholars would comment on it.

Somewhere in my posts on Dunn I mentioned that when we speak of the 'oral period' we really should recognise that we are merely referring generally to the period prior to the Gospels being written down (and published?) since we do not know whether it was really a distinct period of 'oral transmission.'

Also I mentioned that the healing traditions appear to share less verbal/phrasing aggreements than other traditions which may be a consequence of them being widely used oral traditions. I have not researched this properly, but it deserves more attention (as a good candidate for oral story-telling that Matthew and Luke need not be completely dependent on Mark for these stories).

I also mentioned two reviews of Dunn's source theory:
David Neville, "The Demise of the Two-Document Hypothesis? Dunn and Burkett on Gospel Sources," in PACIFICA 19 (Feb 2006), 78-92.
Dennis Ingolfsland, "Jesus Remembered: James Dunn and the Synoptic Problem," Trinity Journal (Fall, 2006), 187-97.

Each of these reviews sees Dunn's source theory as undermining the Mk-Q hypothesis. But it is still not clear exactly what Dunn's source theory entails for Mark and 'Q.' I am amicable to seeing most of 'Q' as potentially stemming from oral sources. I am also comfortable with seeing the authors of Matthew and Luke capable of recognising (and avoiding) Markan redaction (so as to maintain a more 'oral mode' of writing).

However, like Dunn, I am not willing to abandon literary dependence and I think Dunn has tried to incorprate oral and written together in what seems to be an impossibly complex and contradictory task. I wish I knew what to call such a theory (I wonder if Dunn has named his theory yet)?

A Year of Source Theory Blogging - Part One

In less than two weeks this blog will be a year old so for the next couple of weeks I will present a series of posts bringing together the previous year's blogging.

This blog has intended to focus on "Introducing the Synoptic Problem to students" as "A blog on the synoptic problem aimed at furthering student thought and participation." There are already websites around introducing the synoptic problem (see sidebar) and many of these are certainly more helpful and informative than most introductions available in print. Yet students need to be encouraged to see and understand the types of arguments being used. We should be seeing more students (and eventually more scholars) agnostic as to the 'best' source theory rather than settling for what one's lecturers espouse (e.g. 2ST, Farrer, Griesbach).

Students should be encouraged to look at the assumptions for how a theory resolves a problem (and the type of problem being solved). My own experience with the Gospel pericopes does not yet indicate the nature of the non-Markan sources appearing in Luke (granting that Luke depended on something almost identical to Mark). I can now see this agnosticism present throughout my past year's blogging. OK, now for a year's recapping and distillation of thoughts part one...

This blog began when I posted my lecture notes for BN101 held at Whitley College August 2006. Here I used the terminology look-alike gospels as a student-friendly description for the 'Synoptic Gospels.' The lecture emphasised the three kinds of similarities:
similar CONTENT (displayed in a Venn diagram),
simliar PHRASING (displayed in a Synopsis), and
similar ORDER of passages/pericopes (as displayed in Allan Barr's diagram/chart).
It is these three features (and the lack of these three features held in common with the fourth Gospel) which suggests evidence of a literary relationship between the Synoptic Gospels. The 'problem' (or type) of this relationship is more so a modern problem arising as a historical enterprise to explain the use of 'sources' though pre-modern responses demonstrate some awareness of a concern with the problem of Gospel differences (thought to potentially undermine the historicity or veracity of the Gospel accounts). Ironically, it is not now the differences but the similarities in CONTENT, PHRASING and ORDER which present reason to suspect direct copying and often thought to be potentially more devastating (in terms of historicity).
My initial lecture notes helpfully grouped arguments for Markan priority into three categories:
Flawed Arguments (order of pericopes; length of pericopes; lower/higher Christology);
Inconclusive Arguments (difficult passages-Mark has more of them; vocab statistics); and
Substantial Arguments (editorial fatigue; tight plotting of Mark).

What I could have made clearer is that those who oppose Markan priority are basically correct that Markan priority is more of a belief (or 'working hypothesis') than a scientific fact given the current tests and results are rather tentative and often reversable. Assuming Mark to be a source of both Matthew and Luke makes reasonable sense of all three Synoptic Gospels (but so do other source theories).

My initial interest in the synoptic problem is mentioned briefly in my third post and in the latter part of my sixth post. Basically as an undergraduate student I found the introductions to the synoptic problem available in print very disapointing and unhelpful so I wrote an essay demonstrating this(part A evaluating six standard intros; part B looking at arguments for and against Matthean posteriority). The essay is made available in the side-bar (Solving the Synoptic Problem for Students?).

Thanks to Stephen Carlson in January I had to correct my earlier assertion that Henry Owen (1764) was the first to suggest direct literary dependence (interrelationship) since it appears that the notion might be dated even earlier (e.g. Hugo Grotius, John Mill, J.J. Wettstein). However I concluded that we still need to distinguish between:

(a) those who believe that the later Gospel compositions were written with some knowledge of the predecessors work (Augustine’s position); and
(b) those who believe that an author fully ‘depends’ on an earlier work for the material (without which they would probably lack the material).

The latter category has produced various models all claiming some notion of 'dependence' but due to the belief that certain material would have (or could have) been available in more than one document or in the form of oral traditions (rendered even more complex by secondary and tertiary orality--i.e. as a written work develops an oral life of its own) it is impossible to know whether Luke's Gospel depends on the text/document of Matthew's Gospel for all the overlapping material (and/)or whether Matthew and Luke shared a common source or whether Matthew's Gospel (or portions of it) had entered into a second level of orality and could be recalled/remembered by the author of Luke.

To update on an older exchange between Carlson and myself (see link in sidebar) which debated whether Luke's dependence on Matthew should be seen as any more plausible than the two's dependence on shared source/s (i.e. Mk-Mt-Lk vs. Mk-Q-Lk). I think our definitions of plausibility were not identical. What is it that would make us perceive that the Mt-Lk double tradition (1) originated with Matthew's text and/or (2) was only available to Luke via Matthew? So neither the Mk-Q or Mk-Mt theory should be presupposed as more logically plausible since even though the Mk-Q theory demands a common hypothetical source (Q) the Mk-Mt theory demands that Luke owes all of its 'Q' material solely to the text of Matthew making the Farrer theory less agnostic about the material, which could actually represent multiple traditions/sources. Only at a basic mathematical/logical level does Farrer hold the advantage by simplifying the Mt-Lk overlap. Yet only the John Baptist traditions share a close enough verbal similarity (phrasing) to demand literary dependence for this material (which would either indicate very little direct borrowing from Matthew or give us an extremely small Q!)

My next post will review my posts on Dunn's source theory and attempt to bring some clarity to my earlier disorganized posts.