The "ur-source" theory of undesigned coincidences
(Originally published at What's Wrong With the World. Link to original post at 'permalink' below.)
In my book Hidden in Plain View: Undesigned Coincidences in the Gospels and Acts, I discuss possible alternative explanations for the coincidences. Alternative to what? Alternative to the idea that the authors of the books either were eyewitnesses or had access to eyewitness testimony (as in the case of Mark and Luke) and that the details of the works fit together because they (or their human sources) were witnesses to the truth.
The alternatives I discussed, where relevant, were a) an author deliberately "planted" this appearance of coincidence to make his story seem plausible or b) an author was influenced by his knowledge of the other document(s) in question, even if he did not intend to deceive, or c) bare coincidence of one kind or another with no other special explanation in reality connecting the accounts.
I've noticed a couple of people in different contexts making reference, usually without any detail, to another supposed explanation, and recently someone wrote to me with a question about it. Roughly (and it's never spelled out to any great extent) the theory here is what I'll call an "ur-source" or "ur-text" theory, and it hypothesizes that there is some now-lost source or tradition (the implication is that this was written, though it usually isn't specified whether written or oral), for which we have no other evidence (in other words, now lost), that may or may not have been true, that contained both parts of the coincidence and from which the subsequent authors merely selectively copied. This theory, by postulating another "source," allows the question of the truth of what is stated to be postponed, pushed off into the mists. Maybe this ur-source itself was true, maybe it wasn't. Since we don't have it, we don't know what it was like in other respects. We don't know its provenance. We don't know if it came from an actual eyewitness. So maybe these things happened and maybe they didn't. But anyway, the idea of actual truthful testimony from people "in the know" has been cast into doubt and replaced with a fuzzy picture of "some other source," remarkably detailed (as far as the specifics that go to make up the coincidence in question), from which the authors of the documents we actually have copied little bits, giving the illusion of an undesigned coincidence between documents derived from independent sources.
Since the argument from undesigned coincidences has been so neglected, there isn't a scholarly literature out there responding to it, and I hesitate to make any strong predictions as to whether or not some type of ur-source or ur-text theory will prove to be popular among scholars who are dubious about the argument. My bet right now is that, at most, it will be used in some isolated cases rather than being used widely. But I could be wrong, and NT scholars are very "good" at waving hands at an unspecified "source" as an explanation for everything, so maybe the temptation to do so here will prove too strong. I'm going to predict that mainstream NT scholars will not explicitly postulate a single, ginormous ur-source containing everything (all coincidences, both sides, other contents of the gospels, etc.), because that would be too obviously in conflict with popular theories of the origins of the gospels such as the two-source hypothesis (see below). If they use this type of response at all, I'm guessing they will content themselves with vaguely hypothesizing "some earlier source" for some selected coincidences.
I wrote something quite long in response to my recent correspondent, and the rest of this post will be a modified version of what I wrote to him.
1) There is no evidence for the existence of such an ur-source. An ur-text or ur-source theory as an alternative to eyewitness testimony (including, in the case of Luke and Mark, witnesses to whom they spoke or whose accounts they possessed) is thus entirely ad hoc.
In the case you discuss (of "many coming and going" mentioned in Mark explained by the nearness of the Passover mentioned in John), such a theory (if it were a true alternative to what I'm proposing) would presumably mean that neither Peter nor any other eyewitness was the source for Mark's account of the feeding of the five thousand and that the detail in Mark about "many coming and going" did not come to Mark from someone who was there, that neither the author of John nor someone he talked to was there, but that there was some earlier statement or textual source whose connection to the truth of the events is unknown (it might or might not be true) that was longer than either account and that contained both of the details, from which these authors were selectively copying.
For that alternative, there is no evidence at all. If such a thing ever existed, it disappeared without a trace.
That hypothesis, then, merely adds a further, unneeded "layer" between the documents we actually have and real events. It is therefore in blatant violation of the maxim not to multiply entities without necessity.
Such an ur-source bears a suspicious resemblance to a real event (the feeding of the five thousand and the occurrences leading up to it) that contains both facts and the causal connection between them, but for some reason we're talking about another source rather than simply saying that Mark and John knew about the real events. The much simpler hypothesis is that reality--the actual events--contained both facts (the "many coming and going" and the time of the Passover) and that different people recounted, as we know witnesses do, different parts of reality. There is no reason to hypothesize an extra, otherwise unknown source except to put off arriving at, or cast further question on, the simpler conclusion that the events actually occurred with these various aspects to them.
Since the entity postulated in such theories is entirely hypothetical and without independent support, it becomes an arbitrary matter what form it takes. Were there a bunch of different ur-sources, one per coincidence? That would be multiplying entities with a vengeance. Or was there some ginormous ur-source (not just reality, of course!), all trace of which has been (astonishingly) lost, that literally contained everything we now find in all of the gospels, including all of the information for all of the coincidences and a description of all of the causal coincidences between them (which aren't given in any of the gospels!)? Once one starts spelling this out in detail either way, the strained nature of hypothesizing such entities without independent evidence should start to become clear.
1a) Since the hypothesis of an otherwise unknown ur-source just puts an unnecessary layer between the accounts we actually do have and the facts, it is not subject to principled limits. Hence, that type of method would be a history killer.
Consider an event like the Battle of Gettysburg. Historians talk about things that happened there, like Pickett's Charge on Cemetery Ridge, as real events. They learn about various aspects of the battle from primary sources like military dispatches, speeches, putative eyewitness accounts, etc.
Suppose we applied the "ur-source" method to the Battle of Gettysburg. Then we could say that there is some ur-account (which might or might not be true) that claims that the battle happened and that this and this and this happened during the battle--all of which details have become subsequently dispersed through other accounts, which aren't really eyewitness accounts either but are just somehow copying bits and pieces of this entirely hypothetical ur-source.
This, of course, could be used to cast doubt on whether the Battle of Gettysburg ever happened at all!
The more reasonable hypothesis is that there was a real Battle of Gettysburg at which various specific things, like Pickett's Charge, actually occurred, and that these have become known to us through these varied other accounts from people's knowledge of the real events.
Otherwise we never get to talking about the real world at all. Every source we do have is taken to be based on some other source or tradition that we don't have, and we never get back to knowing the facts. This is a reductio of the method in question.
Apropos of which, here is part of a wonderful quotation by a little-known author of a book called John Who Saw. The author was a lawyer named Adrian Green-Armytage, writing about New Testament criticism in his own day (in the early 1950s):
In my world, if The Times and The Telegraph both tell one story in somewhat different terms, nobody concludes that one of them must have copied the other, nor that the variations in the story have some esoteric significance. But in that world of which I am speaking this would be taken for granted. There, no story is ever derived from the facts but always from somebody else's version of the same story....In my world we say, "The first world-war took place in 1914–1918." In that world they say, "The world-war narrative took shape in the third decade of the twentieth century."
2) The undesigned coincidence pattern does not fit well with the way that people would be likely to use a single source that they are copying but does fit the way that witnesses give accounts of reality.
If you were reading a source that said that the reason that there were many coming and going was that it was Passover time, and if you were writing a later document copying from that source, the natural thing to do would be to copy both points and point out the causal relationship between them. You're just copying, after all. And the source (ex hypothesi) emphasizes the causal connection.
But when you witness a reality, you may or may not make the causal connection yourself between two facts, or you may not remember it at all or with enough vividness to want to mention it. What is vivid is the fact that stuck with you--e.g., the busyness in Capernaum (probably it was Capernaum) and Jesus' and the disciples' desire to escape it. And, on John's side, the general recollection of the time of year. Each witness mentions what he does mention because of aspects of it that stuck in his head or that stand out to him at the time of talking or writing. E.g. Peter remembers seeing all those people around in Capernaum. If neither you nor the person who is telling you the story (e.g., Peter talking and Mark listening) has such a selective memory and/or selective recounting of a real event, but you are just copying a source that analyzes and puts everything together, there is no further rationale for the selectivity and the apparent randomness of the selectivity. In fact, your narrative would make more sense to the reader if you actually explained the causal connection.
This is especially relevant where the other fact (the explanation) could not be readily inferred and/or where the effect and the explanation occur in quite different passages or contexts in the books we actually have. Mere carelessness in copying an earlier source would often be a very poor explanation.
Think of how this point applies to the "to his servants" coincidence concerning Luke and Matthew. Luke may well not even have known that Herod said that about John the Baptist risen again to his servants. He certainly doesn't mention to whom Herod was speaking in Luke 9, and his mention of Joanna the wife of Chuza is in an entirely different context (in Luke 8). It is highly implausible that Luke got this information about Joanna from an ur-text that said, "One of the people who followed Jesus out of Galilee was the wife of Herod's household manager, and that was how we found out that Herod said this about John the Baptist when he was talking to his servants." There is no hint of such a source anywhere. This isn't even like either gospel that we actually have. If Luke were copying from that kind of a source, he would have been likely to put in the bit about Joanna in Luke 9 where he mentions Herod's perplexity, and he would have been likely to draw the connection. But Luke doesn't even give any evidence of knowing the connection. He appears to have received the list of women who came out of Galilee with Jesus from someone who knew about that as a separate fact in itself, and he connects it in Luke 8 with the portion of Jesus' ministry when those women were helping him. Similarly, we find in the synoptic gospels (but not in John) that witnesses said at Jesus' trial before the Sanhedrin that he would destroy the Temple and raise it again in three days. The explanation for this rumor is found in John, in an entirely different context (namely, the early cleansing of the Temple, several years before in Jesus' ministry), where Jesus says, metaphorically, "Destroy this temple, and in three days I will raise it up," when he is asked for his authority to cleanse the Temple. Critical scholars cast doubt upon the truth of this cleansing at all, and even (too) many evangelicals think that John "moved" the cleansing. There is no trace of such a comment in the context of the cleansing in Passion Week recounted in the synoptic gospels. It is unique to John and hence is treated as a late addition by critical scholars. (See point 4.) It would be extremely implausible to hypothesize an ur-source (other, of course, than reality itself!) that contained the question to Jesus and Jesus' answer in one context and the later rumor testified to at the trial in another context.
Consider how this point applies to cases where the one event is really quite puzzling without the other and not easily supplied. For example, it is quite odd that, in Luke's account of Pilate and Jesus, Pilate asks if Jesus is the King of the Jews, receives an ambiguous (at best) reply, or an outright affirmation (depending on how you read Jesus' answer), and goes back out and tells the people definitely that he finds no fault in Jesus. Why would Pilate do that? If Luke were copying from a source that also contained the fuller dialogue given in John, in which Jesus affirms his kingship but says that his kingdom is not of this world, it would make much more sense for Luke to copy the further dialogue as well (or at least summarize it), rather than leaving Pilate's, "I find no fault in him" hanging out in the wind, inexplicable and surprising. Reality, as recorded accurately in John, supplies the other side of the puzzle.
In short, an ur-text theory does not explain well the casualness of the bits that contribute to the coincidences and the way we have to glean the connections that turn out (surprisingly) to be there, whereas selective testimony, relating real events, does.
3) Such an ur-source theory is an extremely poor explanation of the coincidences in the gospels (as I've been explaining), but if possible it's even more bizarre if one attempts to use such a method for the coincidences between Acts and the Pauline epistles. (My correspondent didn't suggest this as an explanation of the Acts coincidences, but I bring it up for completeness' sake and also because I think it's illuminating concerning the underlying epistemic issues.)
Indeed, it's difficult to imagine what it would look like for the Acts coincidences. One would have to get really creative! Presumably an ur-source theory would be that the epistles weren't really written by Paul (even the universally acknowledged ones) and that Acts wasn't really written by a companion of Paul, but that both the epistles and Acts were copied from bits and pieces of some otherwise unknown and now unavailable ur-source that told the stories in question, included the bits (in all the right places) that we find out now from the epistles, and made the causal connections between them! Then the authors of Acts and the (forging) author(s) of the epistles just inexplicably left out any mention of those causal connections when copying, and just happened to copy bits that fit together causally as aspects of reality do.
This, on top of all its other faults, would cast doubt upon the existence of the Apostle Paul itself, which would be historically completely irrational.
It's a good exercise to see how crazy this would get as applied to the epistles and Acts, because it illustrates the point in 1a, above: Namely, once we start making up invisible ur-sources that contain everything that later got dispersed among other documents that (darn it!) look just like historical accounts of real events, there's no principled place to stop.
4) This ur-text theory cannot be used consistently by any "critical" New Testament scholar (that is to say, by any very mainstream or skeptical or liberal NT scholar), so it is neither fish nor fowl.
Such an ur-text theory would fly in the face of the usual liberal scholarly approach, which is modern source and redaction criticism. An ur-source theory, applied to most of the undesigned coincidences in the gospels, would make all sorts of things that are supposed to be later additions to the story, later developments, into (instead) aspects of a hypothetical document that was even earlier than Mark. A skeptical theorist cannot have it both ways. He cannot have it that, say, the resurrection appearances of Jesus in the gospel of John are late "accretions" to the narrative and at the same time hypothesize that the conversation between Jesus and Peter in John 21 after Jesus' resurrection was part of a very early, now lost, ur-text that predated Mark. It can't be both of those.
To put it rather briefly, the usual source critical idea is that earlier sources were shorter and more minimal and that unique, additional material in later gospels is what was added as time went on. This is part of the reason that critics assume that Mark, the shortest gospel, predated Matthew, for example. Even hypothetical "sources" (like Q, for example, or the so-called "Pre-Markan Passion narrative") are supposed to have been individually shorter than the gospels that later cobbled them together.
Now, of course, I hold no brief for the "late development" theory. But that's because I think we have good evidence that these events really happened and that the gospels really were recounted by those who knew what they were talking about. If I were trying to make up an alternative theory that would cast doubt upon the eyewitness nature of the accounts we have in the gospels, I certainly wouldn't hypothesize that they were all widely believed so early in the Christian community that they were contained in a (now lost) document or documents even older than the gospel of Mark, from which all later authors copied.
Liberal and other mainstream New Testament scholars are sometimes inconsistent, but that makes their scholarship even worse. I (unfortunately) wouldn't put it past someone who was used to hypothesizing "sources, sources everywhere" to imply, on one day of the week, that unique material in John is (for that reason alone) a later addition or change and, on another day of the week, respond to an undesigned coincidence by suggesting that there must have been an ur-source, prior to Mark, that contained material we now find unique to John! But this would just be further evidence that it is a case of "any stick with which to beat a dog"--in other words, anything to avoid saying that these authors were close to the facts and were telling us what really happened. Even skeptical theorists ought to try harder than that to be internally consistent.
I really hope that the ur-source theory doesn't become popular as an attempted response to undesigned coincidences. But to the extent that it might concern someone who is interested in the argument, I think it is good to make a response available.
Posts
No comments:
Post a Comment