I’m currently reading James D G Dunn’s Neither Jew nor Greek. Like the first two volumes of his Christianity in the Making project, it’s a careful and broadly conservative account of early Christian history, with this volume taking us into the second century, with Irenaeus as a kind of horizon, but the biblical books as a primary focus. (By conservative I mean both in many of his conclusions (e.g. the broad trustworthiness of Acts) but also his attachment to a traditional behind-the-text historical approach to the texts as sources.
But this morning I imagined Mark Goodacre spluttering over his cornflakes as I read this sentence:
“[T]here are no persuasive indications either that Matthew knew Luke’s Gospel or that Luke knew Matthew’s Gospel.” (p.246)
There is an unstated protasis which this sentence desperately needs: “If you are convinced of the existence of Q, then …”
Broadly speaking, there are exactly the same indications, exactly the same evidence that a) Matthew and Luke used Q, b) Luke used Matthew and c) Matthew used Luke. The difference is not in the evidence itself, but in the framework used to interpret the evidence. Only a Q-shaped framework allows Dunn to say what he says.
Now it so happens I largely agree with the heart of Dunn’s position: there is a literary source Matthew and Luke share, but there are also other oral sources, some of which they share and some of which they don’t. It may also be that there are occasions when either one of Matthew or Luke make use of their shared literary source while the other doesn’t. If there is (as I think) some sort of Q, we can have no certain knowledge of its upper or lower limits.
However, I hold this theory not because of any persuasive character of any individual piece of evidence, but because it seems to me that an untidy theory such as this makes better sense of the variegated nature of the evidence as a whole. The theories which try to produce a single explanatory model, whether an over-confident Q, the Farrer-Goulder-Goodacre theory, or a version of Matthean posteriority, all seem to me too tidy for the nature of the evidence, even if I can sometimes feel the force (almost thou dost persuade me) of say, Goodacre’s Case Against Q, or Alan Garrow’s brilliant series of video presentations of the Matthew Conflator Hypothesis.
It just seems to me more than pedantry to keep a sense of what belongs to the theory that makes sense of the evidence, and what is a property of the evidence itself. The evidence is only evidence of some kind of interrelationship. But which interrelationship it is belongs to the theory. And that in the end is a judgement about what makes greatest amount of sense of the greatest amount of the evidence.
It’s like all the best detective stories, really. And it wasn’t Luke in the library with the scissors and paste.