MetaCatholic

Delighted to learn that Mark Goodacre, of all people, has a friend called Q. Even more delightful, he’s been celebrating Q’s marriage to M. Any chance they might have twins called Matthew and Luke?

Michael Halcomb has responded in some detail to my post yesterday (which was in turn a comment on this post of his), arguing for the compatibility of the Matthean and Lukan birth narratives. (He is not alone, as this comment by Peter Kirk reveals.) I would again like to repay the compliment. I have a number of points I want to make, not necessarily in the same order as Michael’s argument.

First, I note that he refers to my approach as “minimalistic”. I confess that outside of a fairly narrow (and I think mistaken) approach to the OT, I’m not entirely sure what this means, and even there I think it’s become unhelpful. But generally, I would say that in terms of relating the gospel traditions to Jesus-questing (a fascinating pursuit of limited value) I’m generally much nearer Wright’s “maximalist” approach than the Jesus’ seminar’s “minimalist” one. My stance on the historicity of the birth narratives is, I think, a consequence of my reading of them side by side. My reading is not a consequence of my stance, and is somewhat different in its judgement on historicity from my reading of the main narratives.

The first major point Michael makes is this one:

Doug, in my opinion, is wrong that in order to maintain the integrity of the narratives, one must choose which narrative they will use. Clearly, Doug fears the word harmonization. Now, I’m not calling for another Diatessaron but it is not unthinkable that the Gospel writers’ stories at some points, should not harmonize.

I agree, in principle, that the stories at some points may be harmonised, but I would also assert than any attempt at doing so has to begin with the integrity of each as a literary and theological work. Only when we have reached an understanding of what any particular author is doing can we compare and contrast their work fully, though in the case of the Synoptics this is to some extent a circular process, since the comparison aids the understanding of authorial intent. My view is that the birth narratives present such significantly different narratives, tailored to the role of prologue in terms of their respective gospel’s themes, that harmonisation as a means of historical or literary study violates their integrity. I have previously noted some of the differences in this post. It is not unthinkable that at some points the stories should harmonise, it is, in my view inappropriate in this particular case.

Michael goes on to say:

Doug seems to forget that they used one another’s works (in what order, as we know, is debatable).

I’m not sure how he makes this observation on the basis of my post. I would note, however, that different solutions to the Synoptic problem actually suggest different routes through this material.

• If one adopts the two document hypothesis, then it is possible to notice some common elements to the two narratives (perhaps supported in part by an independent cave tradition shown in the Protoevangelium of James and Justin) which might suggest an earlier tradition on which both Matthew and Luke build. This hypothetical tradition knows of an announcement of Jesus birth by an angel (but perhaps not to whom); the idea that Jesus name is given by the angel; the conception of Jesus by the Virgin through the Holy Spirit; the birth at Bethlehem and the final settlement of the Holy Family at Nazareth.
• This first route through the material permits harmonising, though it also suggests that both Matthew and Luke may be embellishing an early source for narrative and theological purposes. It doesn’t particularly encourage us to suppose that the one selected accidentally exactly the bits of the story that the other omitted and vice versa. They treat no other source in this way, whether Mark or the contested Q.
• If one follows the Farrer hypothesis, then we have to accept that Luke almost completely rewrites Matthew, either because he doesn’t like Matthew’s material, believes Matthew has got it wrong and that he (Luke) has better sources / better ideas / better stories, or some combination of the lot. In that case there is less evidence (if any) for an earlier tradition.
• This second route through the material makes it fairly obvious, I would have thought, that Luke would not approve of harmonising his story with Matthew. After all, he chose to leave out the greater part of its detail, and tell it very differently. He is hardly going to thank his later readers for putting it all back in. On the Farrer hypothesis, at least, there are clear grounds for saying that harmonising violates Luke’s integrity.

He follows up this point (which I hope I have shown does not lead to his conclusion) by saying:

If we follow the minimialistic approach that Doug uses, to its logical conclusions, we end up with a number of different Jesuses, a number of different Christ’s crucified and resurrected, etc. We no longer just have four perspectives on the same event but four different events in and of themselves.

I think this is simply mistaken, although it is a mistake that much Jesus research seems to tempt one to. From an historian’s point of view, one interrogates the four witnesses, as one might any evidence before a jury, and from listening to their different accounts and standpoints, seeks to determine what actually happened. (It is, of course, more complex than that analogy suggests, but widely different witness statements are not unknown before a jury.) I think that such an exercise has a limited apologetic value, and is intellectually fascinating to all of us who love detective stories.

From the viewpoint of faith, however, we are listening to people tell stories about someone both we and they know and experience in word, sacrament, prayer and action. Just as many of our family anecdotes capture an aspect of a person without necessarily being factually accurate, so do theirs. We are not primarily doing history in reading the gospels, we are relating to Jesus. I think a firm, and overall, historically trustworthy core is both a necessary and real component of the gospels taken together, but they are not there to satisfy the curiosities of a twenty-first century historian, although they give (I think) a remarkably good historical picture when compared to any other bioi from the period.

But they also throw up many inconsistencies and anomalies, not just as all testimony does in the judicial world of weighing evidence, but as narratives told to shape faith in line with the particular responses and beliefs of their authors, for whom, like us, a common faith has very diverse emphases. Harmonisation is simply mistaken as an overall approach: how many times did Jesus cleanse the temple? If we harmonise, we must say three, in which case you might have expected the money changers to say “Hey, quick lads, pack up your bags, it’s that Jesus bloke coming in again.” It doesn’t work. Michael is simply wrong: It is if we harmonise that we have several different versions of the same event, rather than three perspectives on one event.

Finally, I can’t help noting that Michael shows a general tendency to be more certain than the evidence allows, which extends also to his linguistic analysis.

A little knowledge of Luke and Matthew’s Greek (which I’m sure Doug is capable of understanding or may already be familiar with) shows clearly that the Jesus family stayed in the front room of a multi-roomed house.

I am not sure what Michael is getting at here. As far as Matthew goes, the only possible relevant word is οἰκία (house). That contains no connotations about any number of rooms. In Luke, the only relevant word is κατάλυμα (room). Exactly what this means is not clear, but depends on its context, as is the case with words generally, which are not depositories of fixed meanings. It often means a large room, and appears in some places to mean guest-room. But it does not necessarily imply “the front room of a multi-roomed house.” I call to my aid the LXX of Jeremiah 32:38 ἐγκατέλιπεν ὥσπερ λέων κατάλυμα αὐτου — “Like a lion he has left the front room of his multi-roomed house” Hmm, perhaps not.

At every level, I can’t help but feel that Michael is pushing the evidence into an already existing shape. It is the particular peril of harmonisers that they are not sufficiently patient with the details, especially when they reveal ambiguities, aporiai, contradictions and questions.

In a comment on my quick Q addendum, Steve Allison says something I felt needed more interaction than a further response in the comments.

When I first read they were attempting to stratify Q I thought they were crazy. But, something I read in Burton Mack seemed to make sense that it has some validity. The Gospel of Thomas has only sayings from the first two hypothetical levels of Q and not the third. Could that be chance or was it cunningly devised? Other options?

I’d really like someone who knows what they’re talking about, like Mark Goodacre, to give a more detailed answer to this. I’m going to confine myself to some general observations.

• The date, provenance and theological tendencies of Thomas are sufficiently disputed to make it a bit of a wax nose, capable of being pulled and shaped in a number of different directions. Thomas is an important piece of evidence, but whether it provides independence evidence of the Jesus tradition, or of a particular tendency or grouping within the late-first / early second century church, depends entirely on one’s overall reconstruction and hypothesis.
• The earliest datable fragments of Jesus tradition we have come in 1 Corinthians, and include both sayings used in ethical discussion, and a narrative controlling the performance of the Eucharist. Even here there is no “sayings only” tradition, which might make us wary of reconstructions which postulate “sayings-only” as more primitive.
• Q, (granting its existence for a moment) understood as that material common to Matthew and Luke, but not (largely) in Mark, is not a “sayings-only” source, but contains narrative and order, evidenced most strongly in the material around John the Baptist.
• Q, understood as above, is a source of which we cannot know the limits. It is at least possible that some of the material designated Q in fact comes from a second shared source. It is more likely that not everything in Q was used both by Matthew and Luke, as not everything in Mark is used by both Matthew and Luke. Any reconstruction of Q which assumes we have all of it is historically and evidentially improbable. When one begins to attempt the stratification of a reconstruction of part of a hypothetical source into earlier layers, one is counting angels on the head of a pin.
• The argument that Thomas is very primitive depends in significant part on the existence of an early Q as a pure sayings source. The argument that Q was originally a pure sayings source depends in large part on the existence of Thomas as a primitive document. This is the apotheosis of a circular argument, incapable in the light of the evidence as we currently have it of verification, and there are, as noted above, good reasons for doubting the methodology of this deconstruction of Q.
• Finally, one may simply doubt, on the basis of cross-cultural comparisons, whether any oral tradition or history anywhere is entirely without narrative. Whatever the problems of the different views put forward by, say, Dunn in Jesus Remembered, or Bauckham in Jesus and the Eyewitnesses, the fundamental point that both before and after the crucifixion, people told stories about Jesus, seems well grounded both in the sources and in normal human experience. It is inherently implausible that Jesus did nothing worth talking about except teach, and that what he did wouldn’t have had an impact worth remembering and recounting. To put it bluntly, and adapting a point that has been made in other contexts, no one could have been bothered to crucify Thomas’ oracular sage.

The Quixotic Infidel contributes to the ongoing Q discussion. He offers a footnote on this post of mine, saying:

I would object to coming to a negative conclusion regarding Q simply based on how “irritating” the deconstruction and parsing of a merely theoretical source by its proponents can be.

Of course, I quite agree, although I would also note how much emotional judgement is (however disguised) very often caught up in intellectual judgement. But let me put it another way. When something gets magnified from a working hypothesis to a full-blown reconstruction, its faults become ever more obvious. The sheer preposterousness of claims about the early tradition history and stratification of a non-existent text is itself a reason to go back and question whether something that leads to such strange conclusions is at all well founded. Or again, the taller the house of cards is built, the stronger the temptation to pull one out and see what happens.

James McGrath follows up his posts on Q with a suggested Synoptic meme. He’s partially right that we’re talking about paradigms (a much overused word that I’m not entirely sure isn’t a bit too overblown for discussing models) for viewing the gospels here, where any individual piece of evidence won’t work. The crucial piece of evidence is a bit like that moment on a flight. Your hearing’s all weird as you take off, and then at a certain point, your ears pop, and hearing is restored to normal. What James is after is those moments in Synoptic study when our ears pop. What passage was the proximate cause?

It’s a great idea, and I’m afraid I can’t oblige him, which is probably just another illustration of the ways in which we’re all different, but let me say why.

I have, since I first began to engage in critical study, always believed in Markan priority. I guess that says something about my teachers, including the parish priest of my parents’ parish, who didn’t believe in hiding these things from the simple and unlearned. I’ve never come across an argument or exegesis that seriously made me question that.

When it comes to the second half of the problem, the relationship of Matthew and Luke, I haven’t really got an ear-popping moment either. I am, and I’m sure this is a disgraceful thing to say, a “don’t know”. There are some passages which make me think a direct literary dependence is a good explanation, as in my previous post on the infancy narratives. There are other places where Q appears to be the most economical explanation: “the finger of God” (Luke 11:20) is one such. Luke’s version seems to fit his theology less well than Matthew’s version would have.

This latter could, of course, be explained by Luke knowing this particular tradition well already from an oral source, and falling back on the more familiar wording. But once you do that, you’re admitting that neither the Farrer theory, nor the Two Document theory, are entirely adequate explanations. In fact, I think that’s right, they can’t be, without reducing what must have ben a more complex process to a literary theory. Either theory needs some chastening supplementation from oral tradition. But that doesn’t solve the problem of the underlying model one works with.

What has made me receptive to the Farrer theory (apart from sitting in the pub with Mark) is something different. It’s the dogmatism of those who talk about “The extant text of Q.” It’s the faux-archaeological literary approach of those who detect layers within a non-existent text. It’s the assumption that between them Matthew and Luke left hardly anything out of Q. Things like that. In short, it is less a case of evidence than irritation. The evidence is constructed by the theory which explains it. The attraction of the Farrer theory is that (even chastened by the acceptance of an ongoing interplay with oral tradition) it doesn’t mean constructing a literary source, but working with the ones we’ve got.

I am aware that falls well short of an argument, but it increasingly seems to me that the most sound approach would be first to explore the literary relationship between the texts we have, and only when it has been demonstrated beyond reasonable doubt that the Synoptics cannot be accounted for in those terms, should we turn to postulating hypothetical documents. The problem with the 2DH is that it took the field in advance of that exploration, before, as it were, it earned the right to do so. I am becoming ever more persuaded by Mark and others that Q was premature, and that a good case can be made for the Farrer theory.

I’m not quite there yet. But the Q hypothesis doesn’t need attacking: the Farrer theory needs exploring. If Farrer can be shown to offer a reasonable and reasonably full (not complete) account of the evidence, then Q is simply redundant.

James McGrath has dared to take Mark Goodacre on over Q. First, he posted about why Q should be deduced from the overlap between Matthew and Mark. Mark Goodacre responded pointing out that one shouldn’t make too much of these differences.

We have an idea of the degree of verbatim agreement that they [Matt and Luke] exhibit when they share a written source, in the triple tradition material, and it is not as high as the degree of agreement we see between them in double tradition material. The link between Matthew and Luke is a direct one.

One of the interesting features of this ongoing debate between Mark Goodacre and a good many Q-theorists is that Mark continues to refine and strengthen his arguments. I look forward to the future post he mentions. Undaunted, James comes back with a post specifically on the infancy narratives. He argues that they are so different that one cannot imagine either Matthew or Luke having had the other’s gospel available to them while they write.

Although I look forward to Mark’s rebuttal of this, I do want to offer a few observations, since it seems to me that this doesn’t make a particularly strong argument.

Firstly, the infancy narratives are themselves somewhat different material to the bulk of the gospel. Not least, the creative writing of infancy narratives as a prelude seeking to characterise the subject of the life of the hero seems to be part of the literary genre. It is unclear that either Matthew or Luke are seeking to convey history in these introductions, but rather introduce their respective and different stories.

Then it is worth noting the similarities in the narratives. Jesus is born in Bethlehem. His birth is announced by an angel. Mary is a virgin, Jesus is her first born, and his conception occurs without male intervention. It is spoken of in both as the work of the Holy Spirit. Complete outsiders come to witness his birth, which is conveyed to them by miraculous portents. I could add, but leave out of this list, things that might have been deduced, such as Jesus’ birth taking place in the reign of Herod, or things known elsewhere in the gospel tradition, such as Jesus coming from Nazareth, or his parents’ names being Joseph and Mary.

How is this commonality to be accounted for? There is more here than complete independence from one another can easily account for. But if we postulate a common oral (or written) tradition that both know, it must have contained rather more than these bare facts which scarcely make up a sufficient story to have been narrated. In that case this tradition must have looked either more like Matthew or more like Luke, but we can say no more, except that for some reason, one or the other, or both, developed this tradition in various quite incompatible ways. Or we can say that Luke, with Matthew’s exemplar before him, felt he could write a much better story that served his purposes in introducing the hero of his Life. History would seem to agree: apart from Herod, the magi, and the massacre, it is Luke’s story that dominates the Christmas narrative, despite the fact that Matthew was always the most popular gospel.

So can we envisage why Luke might have altered Matthew’s story. Let’s see.

• It begins with a genealogy, which needs to go elsewhere if the educated reader is to be enticed into the narrative.
• It needs a literary tie in with the style and content of the Bible, provided by Luke’s Septuagintal language and the poetry given to Zechariah and Mary.
• It needs to emphasise the faithfulness of the Holy Family. They do all that the Law requires. They are also obedient citizens of empire, and the census is no occasion of rebellion.
• It needs to get rid of those Magi. Luke, as the story of Simon Magus reveals, does not think highly of them.
• While it is shadowed with the hint of what is to come towards its close in the Presentation, it is essentially marked with joy, from the jubilation of Mary and Elizabeth together, through the wonder of the shepherds, to the rejoicing of Simeon and Anna.
• It needs to emphasise the place of the poor and unfavoured: childless Elizabeth and lowly Mary are in a complementary situation, and an age-based antithesis. The shepherds are among the (semi-)outcast poor.
• It needs to prepare the way for the significant place of women as disciples in the rest of the gospel.
• It needs to put Jesus’ birth on the world stage through a global census (accompanied by an ironic use of Augustan roles of Saviour and the bringer of peace) and not confine him to an outpost of civilisation under Herod.

There may be more, but this short list shows both literary and theological reasons why Luke might have wished to completely rework Matthew’s narrative, while retaining that essential core of ideas. In short, the differences between the infancy narratives in no way provide an argument for the independence of Matthew and Luke. In fact, under the vast differences, there is a surprising commonality of some key points which have to point either to a heavily redacted common tradition, or to literary dependence.

Mark Goodacre has a post today in which he succinctly sets out one of the key planks of his argument against Q. Fans of ST:TNG will know that Q is very good at distorting reality, or making new realities. Data is no more immune to his power than the human (and other species) crew of the Enterprise.

Whatever solution to the synoptic problem one prefers, Mark does illustrate a perennial problem. The theories we choose to examine the data, and turn it into information, themselves tend to determine which data we look at, and how we classify it. His example of the way in which certain material is classified as Mark-Q overlaps, while other material is classified as “minor agreements”, is a very clear example of how the data is massaged by the theory in order to prove the theory.

The data do in fact reveal the kind of continuum he outlines. That could still be a Q-continuum. But perhaps we should remember that (as Wikipedia says) “Q is a mischievous, seemingly omnipotent being.” For his tricks, Q depends on that “seemingly” and, as Mark shows, Q-theorists too are not above their own smoke and mirrors. The accurate description of data should not simply be determined by theory which purports to explain it. Unfortunately, all too often it is.

Reading David Parker’s short article Textual Criticism and Theology (Expository Times (118.12 [Sept 2007], pp. 583-89) in the course of a different conversation I almost missed the import of one of his arguments:

It is a reasonable argument that Matthew, having produced a thoroughly revised and expanded version of Mark, would not have been surprised to know that Luke took both Mark and Matthew and made a very different version.

And without a single footnote, the Farrer theory sailed on. Now I’m so used to seeing Q-theorists do this without mentioning that it’s a hypothesis not a text, that I can’t begin to express my delight at David doing the opposite. Clearly sharing teaching loads over a number of years with Mark Goodacre is healthy for his argument. (Mind you, this paper was first delivered in Birmingham, where gospels have dwelt long in Goulder’s and Goodacre’s shadow)

Of course, both should note the hypothetical nature of the underlying argument, but on this occasion it’s so much fun to see the boot on the other foot, that I really can’t carp.

Although overall, I’m more with Mark Goodacre on the Wikipedia debate than with Ben Witherington or Jim West, I found it mildly ironic that the Wikipedia article on the synoptic problem points to the need for a very careful sifting of Wikipedia’s information. To say the least, it could certainly do with Mark overhauling it and rendering it into a coherent argument and readable English.

I often use Wikipedia as a jumping off point for things, never as a final court. And any similar resource will always need to be used with caution, and the more so while people who could contribute knowledgeably to it, prefer to turn up their educated noses, and stick out their academic tongues. One of the more bizarre references in the synoptic problem page is not placed with the main solutions (let the reader understand) but has the appearance of a later more fundamentalist edit.

A handful of researchers, such as Eta Linnemann, argue that each of the evangelists are independent of one another and that the apparent literary similarities are merely coincidental. This theory agrees with the Christian belief that the whole Bible, including the Gospels, was inspired directly by the Holy Spirit and that therefore no intermediate or source documents between books are required because “all scripture is God-breathed”

I would have remained cheerfully ignorant of Linneman’s existence or views but, intrigued by this drivel, dug around. His Her demolition of Q says

The gospels report the words and deeds of our Lord Jesus. They do this partly through direct eyewitnesses (Matthew, John) and partly by those who were informed by eyewitnesses (Mark, Luke).30 In that case the similarities as well as the differences are just what one expects from eyewitness reminiscence.

I’d almost rather see Q reign unchallenged than see this put forward as a serious argument. But it gets worse:

I was a theologian for decades but did not know about the inspiration of the Holy Scripture. I had to be born again to find this out … As a theologian, I was steeped in historical-criticism. If the Lord had not taken me out of it, I would still be in it. … For many years I had taught my students the historical-critical theory that there is a synoptic problem, whose only solution is the two-source theory. I taught that Matthew and Luke copied Mark, and then added their own information from another source. Now I found this had no basis.

Obviously this highlights the need for those who combine personal faith with academic rigour to be showing a better way generally. But I will risk putting Jim West’s and the Wiki sceptics’ noses out of joint (and Jim should be more put out that this man woman was taught by Bultmann, was an associate Professor at Marburg, and still says this stuff). It seems to me that if people like Jim West and Ben Witherington and many others don’t engage in the public information forum that is Wikipedia, then the field will be left to those for whom rigorous academic engagement and true faith have at best nothing to do with each other, and at worst are opposed. And it’s too important for that.

One of the blogs I enjoy is Separated by a Common Language, exploring the differences between American and British English usages, and along with that some inevitable cultural differences. By coincidence, immediately after reading it today, I read Mark Goodacre’s post on the invisibility of the Farrer theory in the US. While the two-source hypothesis still reigns here as in the US, my feeling is that if anyone here was asked what alternatives there were, Farrer would be the one that gets mentioned.

This serendipitous sequence of reading prompted two quite different questions for me. First, given that we broadly read the same books in the same language(s), what brings about this cultural difference in academic preference? And are there any other theories that seem more dominant in one English-speaking culture than the other? I’ve heard it sometimes suggested that different solutions to the pistis Christou debate are culturally preferred, but I can’t say I’ve enough first-hand knowledge to have observed this myself.

The second area is in our different usages of the same language. One regularly reads statements such as: “The Koine dialect is a direct descendant from the earlier Attic one.” Or, “The Greek of the New Testament contains many Hebraisms” but I’ve never come across (either as a book or a reference) any real exploration of language differences within Hellenistic Greek, of which “Hebraisms” may only be one instance of local usage / dialect form. There are probably several, in which case I’d be glad to be pointed to one, although I suspect the surviving evidence may be too scanty for such a work.

But, given our own daily experience of English, in an age when communication has never been so easy, and cultures less isolated from one another’s linguistic development, wouldn’t it be reasonable to assume that there were at least as many differences between the Greek spoken in, say Ephesus, and the Greek spoken in Egypt, as there are between the English spoken in London and New York? Was “Hellenism” quite as mono-cultural and mono-linguistic as many discussions seem to assume?