Dunn and Burkett

One aspect of the synoptic problem is particularly unclear, and I wish someone would at least bring some clarity in discussing it. I think it is relevant to mention Delbert Burkett's approach here because it is really the flip side of Dunn's approach (discussed here and here) where I made observations regarding our inability as yet to know whether we can properly distinguish between an oral source and a written source. And an extension of this is our inability to recognize "re-oralization" of a tradition (i.e. deliberately writing in an "oral mode").

I must mention another article of Dunn's, "Matthew's Awareness of Markan Redation," in The Four Gospels--1992: Festschrift Frans Neirynck, ed. F. van Segbroeck et al.; 3 vols., BETL 100 (Leuven University Press), 2:1349-59, in which he proposes that Matthean avoidance of so-called Markan redaction is often due to Mt not knowing it to be from his own tradition and thus avoiding it by inserting his own oral version (i.e. rather than making a literary/editorial 'change' the decision is already made for him to simply include his known version).

It seems like a pre-empted answer to Burkett's reasoning as to why Markan redaction is so mysteriously missing from both Mt and Lk if one assumes a theory of Markan priority. Is Markan redaction missing from Mt & Lk because they didn't know those parts (as per Burkett) or because Mt and Lk recognized these parts as somehow foreign sounding and so simply replaced them with their own known/home version of traditions (per Dunn)?

There is something appealing about Dunn's approach in that certain Markan redactional features missing in Mt and Lk (and highlighted by Burkett) are not so surprising when we grant the Evangelists the ability to recognize and avoid Markan redaction. And I think this would answer Burkett's objection that Mk cannot be the source of Mt and/or Lk. It also seems like an attempt to save the two-source theory!

Strangely, Dunn also wishes to see the relative lack of variation in the synoptic passion accounts as evidence that it was "relatively more fixed at a very early stage." But can we really have it both ways? A presence of variation indicates oral (i.e. against redactional changes) and a lack of variation indicates oral? Would not a lack of variation indicate more literary dependence according to Dunn's own logic?