I'm indebted to Alan for a number of reasons, so my comments here are not meant to be disparaging in any way. In fact, what I've read of his essay so far helpfully and honestly takes up Birger Gerhardsson's work on memory and tradition and sets it on a firmer footing. As one example, Kirk has an enviable grasp of the dynamics of fluidity and fixity at work among the evidence for the early Jesus tradition:
The formal features of such genres ensure stability and thus continuity across many oral enactments. However, through their equally core property of variability, or better, multiformity, the tradition is brought to expression in ways responsive to the different social and historical contexts in which it is enacted. (160; my emphasis)
I love that formulation: "their equally core property [relative to their stability] of . . . multiformity." Too often we think of the tradition as some "thing" that exists on its own and that suffers corruption if/as it changes. Research on memory, oral tradition, and oral performance, however, have encouraged us to reconfigure our understanding of the Jesus tradition to account for the ways the tradition could be multiply expressed in different forms, for different purposes, on different occasions, etc. The tradition contains within itself the capability of multiform expression and variation, so that change ≠ corruption.
Even so, Kirk shares an assumption (if I may call it an assumption) with the vast majority of biblical and related scholarship that the codification of tradition in written texts—even handwritten texts—stabilizes that tradition. Immediately following the excerpt quoted above, Kirk adds,
Oral genres, in other words, though stable are not fixed in the sense that the written medium fixes a text. To fix them would be to impair their capacity for oral (as opposed to written) transmission, for loss of adaptability to different social and historical settings entails erosion of relevance and hence survivability. (160–61)
I understand the basic logic undergirding this assumption. If I give a speech this afternoon, I can't "re-hear" that speech a week or even ten minutes later without the aid of some recording technology (which obviously did not exist in antiquity). But if I write a text this afternoon, I can return to that text ten minutes, a week, even ten years later and the wording will be the same then as it is today. Once inscribed, a text can be corrected, erased, commented upon, whatever. But unless I'm stuck in some J. K. Rowling story, the words set in ink won't change and, within limits, aren't even subject to change.
But Kirk's essay isn't focused on written texts; it's focused, instead, on the conjunction of memory and tradition. And I cannot see how an inscribed text fixes tradition and places limits on its variability (or, to use Kirk's preferred term, multiformity). I'm not suggesting that written traditions can't be fixed, relatively or absolutely. But the fixing (= stabilizing) of tradition requires certain social forces that transcend the presence or the absence of written texts. One need look no further than the synoptic gospels to see that, if Matthew and Luke are dependent on Mark and Q (a proposition Kirk accepts), nothing about the entextualization of the Markan or Q tradition rendered that tradition fixed. Bart Ehrman's work on "the orthodox corruption of scripture" [my comments are available here] highlights the way [hand]written traditions were still responsive to (even subject to) the social contexts in which they were employed.
Within the Hellenistic milieux of early Christianity, second Temple Judaism, Roman Egypt, etc., tradition in both written and oral expressions experienced the "core properties" of stability and multiformity. That may strike us as odd, given the ease with which we naturally expect multiple copies of [printed] texts to be identical, whether we bought them in east Tennessee or South Yorkshire. But we simply cannot even recognize the data of the ancient world—let alone account for it—as long as we think that "the written medium fixes a text."