My rating: 2 of 5 stars
I recently read Delbert Burkett’s “The Case for Proto-Mark.” That was after someone on the AcademicBiblcal subreddit recommended his previous book, “Rethinking the Gospel Sources: From Proto-Mark to Mark,” where he first broaches his idea of TWO Proto-Marks.
That book was back in 2004, and I thought the second book, being much newer, would have more information. Unfortunately, it talks just about the idea of “a” Proto-Mark while referring readers to his previous book for details on the idea of PM-A, used by Matthew, and PM-B, used by Luke.
I dropped a brief review on Goodreads, linking to longer at StoryGraph, but I wanted to do a set of blog posts in much more academic-like depth.
Sidebar: I had asked the Redditor if Burkett discusses dates, or loci, of composition on any of the canonical gospels (or Proto-Mark[s]) in that first book. He was going to check his notes, but never responded, probably in part because I said I had gotten this one on ILL. Sadly, Burkett has none of that in this book, which also lessens its value.
And, that value is, per the icon? Two stars.
Let’s dig in on the first post, tackling his basic thesis, from my detailed notes, partly explicating his basic thesis already.
Burkett’s thesis is basically this:
1. The two-source theory has too many shortcomings to be salvageable
1A. This includes, more in passing than in active discussion, the idea of Mark vs a deutero-Mark salvaging the two-source theory.
2. Therefore, we turn to proto-Mark, which he is at pains to stress is compatible with other theories besides an updated two-source theory … and
2A. More specifically, Burkett’s idea of both a proto-Mark-A, used by Matthew, and a proto-Mark-B, used by Luke. (That’s explicated much more, it appears in his previous monograph “Rethinking the Gospel Sources: From Proto-Mark to Mark.”)
My response?
1. While the two-source theory’s problems are greater for sure than Streeter claimed, and may be somewhat greater than more modern people like Fitzmayr say, they’re not as severe as Burkett claims and certainly not irresolvable
2. His passing by deutero-Mark largely in silence is “interesting”
3. His stipulation that two versions of proto-Mark are required is tendentious at worst, unnecessary within a proto-Mark theory in general. (One could have just one proto-Mark and a final mark, and have Mt use one and Lk another, or have both use just the one proto-Mark)
4. Burkett’s failure in this book to discuss datings of either proto-Mark, or a final Mark, or a proposed developmental, editorial and redactional history of Mark should be seen as militating against his other claims in general.
5. Burkett’s failure to discuss in depth, or even semi-depth, a defense of his particular idea of a Proto-Mark A AND a Proto-Mark-B, not even in the conclusion, yet nowhere saying he has rejected this idea of his previous book, is also problematic. Yes, he says, in essence, that’s not his focus, but, he could have incorporated bits of that in his conclusion chapter, IMO. I say that because I am not that convinced of just a single Proto-Mark, and while he tries to spin Occam in his favor, I find multiple proto-Marks even less likely.
As for the bottom line?
It was fun to dig into some serious gospel criticism. That, as well as Burkett's diligence, rescue him from a flat one star. For more of why he fell to two stars, or rather one-and-fractional stars, go link: here. And, yes, I said my review here was relatively short; that one is relatively relatively short.
My Bayesian probabilities will now go 10 percent on some version of a proto-Mark, tho 0 percent on his two-proto idea, while still saying that this is within the two-source tradition, rejecting his idea that it's outside that. Deutero-Mark goes to 30 percent, whereas before this, some version of the traditional two-source theory would be at 80 percent, and is now at 60 percent, and a deutero-Mark would be at 20 percent, not 30.
View all my reviews