Rethinking the Gospel Sources with Burkett


Image result for rethinking the gospel sources delbertThe Synoptic Problem isn’t my main squeeze but, since the Synoptic tradition is, I find myself dipping my toes in the source-critical debate from time to time. Usually the debate follows familiar divisions, with almost everyone agreeing that Mark wrote first and served as a main source for Matthew and Luke. The crowd splits over whether Luke knew and used Matthew, or whether Luke and Matthew independently made use of a now-lost source, “Q,” in addition to Mark. Breaking from the traditional framework, Delbert Burkett’s 2004 monograph, Rethinking the Gospel Sources: From Proto-Mark to Mark, picks a fight with everyone by arguing trenchantly against Markan Priority: “[n]o one Synoptic served as a source for either of the other two” and that, instead, all three of the Synoptics as we have them made use of a now-lost source—not Q, but a “Proto-Mark!”[1]

Two chapters into the book and the argumentation is clear and, frankly, brilliant if also problematic. I am inclined to agree with the editorial blurbs, hailing B.’s book as an “important challenge” that must be refuted “as quickly as possible.” It’s an important book with a serious argument; it’s also fun to read even when I disagree. B.’s argument is cumulative, but we can get a sense of the logic of the argument by working through one of its planks.

Distinctive Markan Use of πολύς

Using the story of Jairus’s daughter (Mark 5:21-43 and parallels), B. presents a curious phenomenon across the three Synoptics: πολύς, a “distinctive” yet “benign” feature of Mark’s text, drops out of both Matthew and Luke in the parallel stories. Why should that be so? It’s not a word to which Matthew or Luke were likely theologically opposed. Its absence in Matthew and Luke is curious. Here I recreate B.’s Table 2.4A (see p. 15):

—                     parallel material “but not the relevant Markan feature” (in this case, πολύς)
[blank space] “no material parallel to Mark”

Mark Matthew  Luke
5:21 πολύς 9:18 ⁠— 8:40⁠ ⁠—
5:23 πολλά 9:18 ⁠— 8:41 ⁠—
5:24 πολύς 9:19 ⁠— 8:42
5:26 πολλά 8:43 ⁠—
5:26 πολλων 8:43 ⁠—
5:38 πολλά 9:23 ⁠— 8:52 ⁠—
5:43 πολλά 8:56 ⁠—

On B.’s reading, Matthew and Luke parallel Mark (Luke more closely than Matthew), but never retain this distinctive feature. For B., this pattern is difficult to align with Markan Priority. Lacking an ideological reason for Matthew or Luke to suppress Mark’s πολύς (and I’m hard-pressed to come up with a good one), B. suggests that Markan Priorists

would have us believe either that Matthew and Luke shared an aversion to these common expressions of size and degree or that the editorial process resulted in the coincidental elimination of this word to a highly improbably degree. […] Is there some other explanation? The most obvious explanation is that the instances of the expression that occur uniquely in Mark did not occur in the material that Matthew and Luke shared with Mark. (17, emph. added).

If Matthew and Luke have indeed omitted Mark’s πολύς in the Jairus story, then one needs to explain that omission. B. offers three options: (1) Matthew and Luke have some objection to the word (but we don’t know what); (2) its omission is coincidental; or (3) the Mark we know isn’t the Mark known to Matthew or Luke. The “Mark” known to Matthew and Luke was a “Proto-Mark.” B. will spend the rest of the book advancing and defending (3) as an answer not just to the question of Mark’s πολύς in the Jairus narrative, but to the Synoptic Problem itself.

Problems in Parallel

Farrer-Goulder-Goodacre theorists (that is, those who think Luke used Matthew) might be tempted to assume that Matthew omitted Mark’s πολὺς across this story and Luke, having Matthew at-hand, simply followed Matthew’s tendency. But strikingly Luke does not align closely with Matthew here, but rather Mark.

A problem does emerge, however, on B.’s reading of the Synoptic parallels in the Jairus story, but it has nothing to do with whether Luke used Matthew (and less still to do with Q). Whereas B. counts four examples of Matthew having “material parallel to Mark but not the relevant Markan feature” (Mark 5:21/Matt 9:18—; 5:23/9:18—; 5:24/9:19—; and 5:38/9:23—), I count none. This renders the πολὺς problem less problematic, with Matthew and Luke reproducing the Markan tradition in their own, highly distinctive ways (thus: no need for Proto-Mark).

We can speak of Jairus’s daughter as a parallel story in all three Synoptics, but on the level of text, Matthew does not offer any true parallels at B.’s key citations (9:18; 9:19; 9:23). Luke, on the other hand, does tend to follow Mark more closely; B.’s em dashes in Luke’s column seem well-placed to me. Matthew’s column, however, is stripped of its putative parallels. The table looks quite different after I have counted the beans (my proposed challenges in red):

Mark Matthew (orig) Matt (challenges) Luke
5:21 πολὺς 9:18 ⁠— [blank]
Matt has no crowd //
8:40⁠ ⁠—
5:23 πολλά 9:18 ⁠— [blank]
Matt has no “besought” //
8:41 ⁠—
5:24 πολύς 9:19 ⁠— [blank]
Matt has no crowd //
8:42 —
5:26 πολλά 8:43 ⁠—
5:26 πολλων 8:43 ⁠—
5:38 πολλά 9:23 ⁠— [blank]
Matt has no “weeping” //
8:52 ⁠—
5:43 πολλά 8:56 ⁠—

Each Gospel has a parallel version of this story, but on the micro-level of text, Matthew does not parallel Mark as closely as does Luke. Matthew has a different editorial/storytelling program in mind.[2] There are simple explanations for Matthew’s omissions:

Matthew doesn’t have the Markan πολὺς in Matt 9:18, 19 because Matthew also doesn’t have the Markan crowd (ὄχλος πολὺς; 5:23, 24). Matthew doesn’t have the Markan πολλά emphasizing Jairus’s plea in 9:18 because, well, Jairus isn’t pleading in Matthew like he is in Mark and Luke (cf. their use of παρακαλεω). Matthew doesn’t feature any people weeping or wailing in 9:23, so there is no reason to reproduce Mark’s πολλά from 5:38. Matthew is doing something different with the story than did Mark; Luke does something different with it, too, but it bears more family resemblances to Mark, which I still take to be its progenitor.

So, on my reading of these texts, Matthew and Luke are doing nothing more or less than recasting Mark’s story in their own narrative contexts (or, in terms borrowed from John Miles Foley, Matthew and Luke perform Mark’s tradition). Both follow Mark’s broad contours, but Matthew zigs when Mark and Luke tend to zag. If Matthew doesn’t parallel Mark on the textual level (no crowd, no pleading, no weeping or wailing), then we are left only with the problem of explaining Luke’s omission of πολὺς in his much closer textual parallels, but this would not require an appeal to “Proto-Mark,” at least as far as I understand the issues and evidence in the story of Jairus’s daughter.

I’m back, baby!

[1] Delbert Burkett, Rethinking the Gospel Sources: From Proto-Mark to Mark (London: T&T Clark, 2004), 5.

[2] “Matthew’s abbreviation of this story is drastic. The first three instances of πολύς in Mark 5,21-24 are absent because in the Matthean context there is no crowd present.” David J. Neville, “The Phantom Returns: Delbert Burkett’s Rehabilitation of Proto-Mark.” Ephemerides Theologicae Lovainienses, 84 (2008), 148. Neville retains Matthew 9:23 as “one remaining potential parallel,” but the textual parallel seems to run between Markan and Matthean use of “tumult.” Matthew lacks the specific context in which Mark’s πολλά might reside, so I do not count this as a true parallel.