My recent invocation of Matthew Larsen’s article has generated a good bit of discussion both on and off the blog. In one of these exchanges, Mike Holmes raised some good questions about the article and agreed to let me post them. I invited Matthew Larsen to respond, and I’ve posted his response below Mike’s observations. Thanks to both Mike and Matthew for sharing their thoughts.

Here is Mike’s query:

Perhaps it is the perceived (on my part) disjunction between two of [Larsen’s] closing statements that is puzzling me. Near the bottom of p. 379, he writes “So, I argue that we can no longer simply assume that a text was finished and published, especially texts that are not high literature”—a point very similar to conclusions I reach in my “From ‘Original Text’ to ‘Initial Text’” essay (e.g., “These are the sorts of things that are characteristic of the literary environment in which the early Christian documents were composed and first circulated, which means that the relationship between the earliest circulating copy or copies and the author(s) from whom the work initially issued must be investigated, not assumed (as has traditionally been done)” [p. 658]; cf. 662, top half, esp. “In effect, the traditional assumptions now function no longer as axioms that shape or channel an analysis of the evidence, but rather as hypotheses to be tested against the evidence”). But Larson also writes, just a page later: “Unless we can presuppose a finished text, the traditional philological apparatus of textual, source and redaction criticism crumbles, as each criticism bleeds into one another” (p. 380, emphasis added)—and since he has clearly demonstrated that we cannot presuppose a finished text, therefore, he  implies, textual criticism crumbles. But here, I think, he has over-generalized the implications of his essay: he himself admits that some texts were finished (p. 379 near bottom)—in which case, however infrequently, the “traditional philological apparatus” would still be applicable. In short, I think he would be truer to his own findings to say “unless we can demonstrate a finished text …”—much like he says on 379 bottom, “Unless we can determine that a text was finished, closed and published” (a conclusion that would be close to my point that each case “must be investigated, not assumed”). In this case a key question—one that Larson does not address—is this: how might we determine, on the basis of the surviving evidence, whether a work was (to use Larson’s term), “finished” or not? … What sort of evidence is necessary, or what sorts of criteria might enable us to differentiate between “finished” or “unfinished” texts?

And here is Matthew’s response:

For what it is worth, I used “presuppose” and “assume” as more or less analogous terms. That is, I argue, based on a very brief survey (in the article, longer in my forthcoming book) (1) we can no longer simply or unreflectively assume texts were finalized in antiquity, (2) the gospels strike me as a clear example of an unfinalized textual tradition, and so, (3) if we cannot assume/presuppose gospel was a finalized textual tradition, what does it mean to do textual criticism vis-à-vis redaction criticism of such a tradition? The scribe/author distinction doesn’t make sense to me in such a framework. Each person who makes and remakes the tradition has a creative role to play.

I think Holmes’s concluding questions get to the heart of the issue. The field of early Christian textual studies suffers (probably from its Protestant/religious heritage) from an undertheorized sense of finality or unfinalizability. This is why in my forthcoming book I start and end with Bakhtin, who has the best theory of unfinalizability that I’ve come across. The real answer to his question is that from one point of view, finality is in the eye of the beholder.  Thus, the beholder, of course, however, could be the author, the reader, the corrector, and so forth. That is, the question of finality/unfinishedness is perspectival. Sometimes writers worked hard to give the impression their work is finished. Sometimes readers perceive unfinishedness in a work, even if a writer or author goes out of their way to say their work is finished (or not), and the readers feel liberty to rework and improve the tradition. Almost none of the writings that have come to be regarded as part of the New Testament strike me as coming from writers who had a sense of their writings as finished pieces of literature. The Apocalypse seems the exception: it does want to be finished, though it is not exactly what I would call literature. Whether other readers regarded it as finished is another question. The case of the openness of the textual tradition seems especially true of the gospels, to me. With many other scholars, I argue they initially lacked titles, ascribed authors, and (with the exception of the Gospel according to Luke) named readers. See the forthcoming essay that Brent mentioned in a comment.

From another point of view (and this is a logical necessity of the first point of view) there can never be a truly, unequivocally “finished” text. Josephus is a case in point here. He seemed to have regarded his Antiquities as “finished” each time he rewrote it. He says he published it initially in 93/94 CE. But then Justus seems to have written a rival account and Josephus revises his Antiquities and added his Life. But even then he concludes his Life with “Here I stop my writing for the present.” That is, after he had written, revised, rewritten, and finished his Antiquities, he then revised his Antiquities again and added his Life. Yet even then he still says he is stopping for the present, leaving a door open for future possible revision. Thus, the arbiter of finality from our perspective, based on an undertheorized sense of finality, becomes “we philologists.” My point is to ask myself and other scholars to reflect on our own position in the process of closing/controlling a text. What point in the process do we privilege for such a textual tradition? Whatever we choose, we must acknowledge our own role in the shaping of tradition.

Thus, it is a remarkably tricky thing to talk about “determining” a finished text from any objective point of view what constitutes a finished text. Whose perspective are we determining? The author’s? The readers? Or perhaps even making our own judgment? We can make a determination from a certain perspective, but that determination is not the only or the final one. That said, my research in my book does lead me, as a cultural historian of ancient writing practices, to conclude that any textual tradition that lacks a title and an ascribed author is a good candidate for being a practical, “para-literary” text that remains open to alteration. Any text that exists in multiple versions of the same work is a candidate. Any work that seems to be continually reworked is a candidate. The gospel tradition fits all these criteria. For that reason, I suspect a framework something like the one Brent mentioned (Short Recension of X and the Long Recension of X) will offer an improvement to our scholarly discourse. In my forthcoming book, I suggest creating proportional venn diagrams of the overlap between each recension and then creating a 3-D “time lapse” of the growth of the textual tradition of the gospel up to Irenaeus.