What is the trouble with Q?

In a recent article in Bible and Interpretation, Dan Smith reflects on The Trouble with Q.  But the troubles he is talking about are not those niggles that make some of us question the existence of Q, the Minor Agreements between Matthew and Luke against Mark, the Mark-Q overlaps, the weakness of the arguments for Luke's independence from Matthew and so on, but the way that many scholars fail to take the implications of the Q theory seriously.  Like Kloppenborg, Dan draws attention to facile appeals to Q's hypothetical nature by those who refuse to think through what the existence of Q means for our reconstructions of Christian Origins.

But is this really the trouble with Q?  I am puzzled by Dan's focus on scholars who are uninterested in the Synoptic Problem and Q.  Those who have not invested time in studying the problem are unlikely to want to engage seriously with the implications of the hypothesis.  In my own experience, the same set of scholars, the so-called "lazy" believers in Q, are equally unwilling to invest time in engaging with Q sceptical scholarship like my own. 

It is never easy in scholarship to find yourself in a position where you are telling other people that they should be interested in what you are studying.  Engaging other scholars' interest is always tough, and it is especially tough in areas like the study of the Synoptic Problem, which requires a lot of hard work and a degree of technical expertise. 

There is always the option, though, of seeking out dialogue partners among those who are already interested in the problem.  Dan alludes to those who do not accept the existence of Q but he does not mention them by name, much less engage them directly.  Regular readers of this blog will know that one of my memes relates to the Farrer Theory getting ignored in scholarship, but I am usually in those contexts talking about introductory-level works, textbooks and the like.  It is, however, something that happens in the higher-level work too.

The problem is that there is a major alternative to the Q hypothesis and critical engagement with it can help to clarify and focus the bigger picture questions that Dan mentions here, about the degree of diversity in early Christianity, in particular the question of the role played by Jesus' death.  Dan talks about engaging with people who may help with refining the Q hypothesis, but a healthy hypothesis is also one that engages with those who are attempting to refute it.

Let me offer one example of how this works.  Dan talks about taking seriously Q's silence about the salvific death of Jesus, the resurrection and the term "Christ".  Arguments about silence are often worth hearing and in the case of a text like the Gospel of Thomas, its silence on these same features is indeed worth some serious thought.  There is a difference in the case of Q, however, that makes any study of its silence problematic.  Given that the document is reconstructed on the basis of Matthew's and Luke's double tradition, there is always the possibility that it is not Q that is silent on the matters in question but that Matthew and Luke are silent in their witness to Q's contents.  

Dan talks about what he regards as similar difficulties in reflecting on Matthew, noting that there are many things we do not know about Matthew, its author, whether he wrote other materials and so on.  But we do have textual witnesses to Matthew that are pretty clear about the scope, parameters and wording of the work, the very things that are absent in the case of Q.   Indeed, the absence of any kind of textual witness to Q is one of the things that invites us to consider the alternative, that the kind of close verbatim agreement that Dan discusses may be evidence not for a lost document but for a direct link between Matthew and Luke.  

In other words, the hypothetical nature of Q is indeed relevant in this discussion.  Dan is right that the hypothetical nature of Q should not be used as an excuse for a refusal to think.  The real issue, though, is that Q's hypothetical nature is an invitation always to think about live alternatives.  To imagine a world without Q is surely one of the best ways of testing a model where Q is central.

To put it another way, what are the chances that Dan's essay will be noticed by those who are his targets?  At least the Q sceptics are willing to have the conversation.

The New Testament in Antiquity on the Synoptic Problem: Some Further Issues

In a post last night, Yet Another NT Introduction Ignores the Farrer Theory, I talked about the treatment of the Synoptic Problem in Gary M. Burge, Lynn H. Cohick and Gene L. Green, The New Testament in Antiquity (Grand Rapids: Zondervan, 2009). I focused particularly on its ignorance of the Farrer Theory and I didn't have space to talk about some other curiosities in its presentation, and I'd like to comment on those here.

First let me pick up again on the authors' suggestion "that when Matthew and Luke make editorial changes to Mark, none of Matthew's changes show up in Luke and vice versa."  They add that "If Matthew had known Luke -- or if Luke had known Matthew -- then surely some of the changes would be apparent" (116).  It's a remarkable statement given that the triple tradition in fact features hundreds and hundreds of agreements between Matthew and Luke against Mark.  It is true that Two-Source Theorists have explanations for these, some of the major ones as "Mark-Q overlaps" and some of the minor ones as due to independent redaction, or textual corruption, but denying that they are present ought not to be a serious option in the discussion.

Other statements that I would like to question relate to their brief arguments for Marcan Priority, e.g. the statement that "While Mark's gospel is shorter, each of Mark's narratives is longer, and often details are removed in Matthew (but still assumed)"  (115).  I like the drift of the latter element here, which coheres with my argument from Fatigue, but the statement that "each of Mark's narratives is longer" is not true at all.  Many of Matthew's parallels to Marcan narrative pericopes are longer.  The impression of consistent Matthean brevity is derived largely from the triple tradition material in Matt. 8-9, where Matthew is shorter in his parallels with Mark.  Elsewhere, this is less often the norm.

A third puzzling claim is the following:

In Mark 6.5 we learn that Jesus "could not do any miracles [in Nazareth]." Matthew appears to supplement this potentially embarrassing admission by saying Jesus did no mighty work there "because of their lack of faith" (Matt. 13.58). (115).

The line quoted from Matthew is identical in Mark 6.6; Matthew does not so much supplement Mark here as contract it.  I think the authors are aiming to make a contrast between Mark's "could do no . . ." and Matthew's "did not . . .", which is the way that this argument is usually set up, but the re-statement of it loses the necessary clarity to make that point (and, incidentally, it's worth taking a look at Peter Head's Christology and the Synoptic Problem for a nuanced discussion of this example).

A fourth issue is a diagram that appears on the bottom of p. 115.  The diagram has arrows illustrating Marcan Priority plus Luke's use of Matthew.  In other words, it appears to be an inadvertent and unlabelled diagram of the Farrer theory.  But it appears to be used, in context, as a diagram that is supposed (just) to illustrate Marcan Priority, which is what is being discussed either side of it.   It may be that the arrow pointing from Matthew to Luke is deliberate, and that a subtle allusion to the Farrer theory is intended, but I doubt it.  The text goes on to talk about "whether Matthew used Luke or Luke used Matthew", which would suggest alternative arrows, one going in each direction between Matthew and Luke.  I think this lay-out could prove quite confusing to students.

Yet another NT Introduction ignores the Farrer Theory

It's déjà vu!

After noticing a recent case of Another Introduction to the Bible, Another Chance to Ignore the Farrer Theory, curiosity compelled me to check out another recent New Testament Introduction, Gary M. Burge, Lynn H. Cohick and Gene L. Green, The New Testament in Antiquity (Grand Rapids: Zondervan, 2009).  I had previously noticed just how gorgeous looking the book was -- a lavishly illustrated historical introduction to the New Testament with a stress on its cultural contexts.

What I had not previously noticed was that it featured a several page discussion on the Synoptic Problem (112-7) with nice diagrams, a Synopsis of a passage (Matt. 8.16-17 // Mark 1.32-4 // Luke 4.40-41) and pictures of Augustine, J. J. Griesbach, B. H. Streeter and William Farmer.  And this itself is something of a first -- I don't think I've ever seen a picture of Streeter in a New Testament introduction before, let alone Farmer.

Sadly, my initial enthusiasm soon gave way to the now all-too-familiar experience of seeing the Farrer Theory ignored.  And this example is a particularly striking one in the genre.  As far as this textbook is concerned, there are only two solutions to the Synoptic Problem actively discussed in New Testament scholarship today, the Two-Source Theory and the Griesbach Theory.  The Two-Source Theory is represented as "most popular today", indeed "so well received that some scholars refer to this result as an 'assured finding'" (116).  The latter is a quotation from Willi Marxsen in 1968.

The authors go on to explain, however, that there is "another generation of scholars", led by William Farmer, who adhere to the Griesbach Theory.  And they add that "Farmer has a growing following" (117).

The discussion of the Synoptic Problem itself proceeds along familiar lines.  Augustine gives way to Griesbach, but Griesbach does not work because Marcan Priority is more plausible (115). But once Marcan Priority is accepted, "the next question is whether Matthew used Luke or Luke used Matthew" (115).  At this point, there is reason for the Q sceptic to feel encouraged -- the right questions are being asked!

Such hopes are soon disappointed.  The authors suggest "that when Matthew and Luke make editorial changes to Mark, none of Matthew's changes show up in Luke and vice versa" (116).  "If Matthew had known Luke -- or if Luke had known Matthew -- then surely some of the changes would be apparent" (116).  And so, when the double tradition is introduced, it becomes inevitable that Q is the answer.

Technically, I suppose, the Farrer theory is represented in one half of that couple of theoretical sentences even though it is not introduced by name or properly explained.  But what I find particularly remarkable about this example is that the Q theory is introduced on the basis of an erroneous claim, the statement that Luke does not feature any of Matthew's editorial changes to Mark.  Indeed, for many, this is the appeal of the Farrer Theory, that it is able to account for the many cases of Matthew's and Luke's agreements against Mark in triple tradition.

It is an interesting state of affairs to have the Q theory expounded, for introductory students, on the basis of the denial of a large body of data that is actually foundational for those who are sceptical of the existence of Q.

Another Introduction to the Bible, Another Chance to Ignore the Farrer Theory

Regular readers will be familiar with the Farrer theory ignored trope.

Here we go again.

Robert Kugler and Patrick Hartin, An Introduction to the Bible (Grand Rapids: Eerdmans, 2009) is the latest in a long line of Bible Introductions to ignore the Farrer Theory, to talk as if the Two-Source Theory is gospel, to suggest that Marcan Priority is the same thing as the Two-Source Theory and to mention only Griesbach offered as an alternative (with a brief mention also of Augustine).

Kugler and Hartin's discussion of the Synoptic Problem is found on pp. 351-5 of their introduction. They have a brief discussion of some data (352) and then they proceed to "Solutions Proposed".   Three are offered, first, "St. Augustine's Solution", then "The Griesbach Hypothesis", which is "still supported by a small number of scholars today" and then "The Two-Document Hypothesis", which is given as "the most widely accepted view" and which "helps to explain most of the difficulties in the relationship among the Gospels" (352).

The authors then provide a series of "five stages": (1) oral traditions, (2) written compilations (including Q), (3) Mark, (4) Matthew, (5) Luke-Acts.  They conclude with a section headed "The Two-Document Hypothesis Illustrated".

No arguments for the existence of Q are offered but several reasons are given for Marcan Priority (353-4), with the implication that this entails acceptance of Q.  Surprisingly, the passages offered as illustrations of the Two-Document Hypothesis include some famous examples of difficulties for the Two-Source Hypothesis, including Matthew's and Luke's major agreements against Mark 1.2-13, with special attention given to the Mal. 3.1 quotation in Mark 1.2.

One of the difficulties they face in using the Mark 1 material to illustrate the Two-Source Theory is self-contradiction.  Thus they begin with the statement, "A passage occurring in all three Synoptic Gospels is referred to as the triple tradition" (354).  They continue with the Mark 1 / Matthew 3-4 / Luke 3-4 material, noting parallels between Matthew and Luke that are not in Mark and adding, "This is known as the double tradition".

Oversimplification leads to difficulties at other points, for example:

 The Q document contains practically no narrative. Instead it presents a series of Jesus' sayings (very similar to the Gospel of Thomas which was discovered in 1945).  These sayings are recorded chiefly in the sermons in Matthew and the journey of Jesus to Jerusalem in Luke (353).

Familiarity with the reconstructed text of Q actually shows that it contains a good amount of narrative and that the contrast with the Gospel of Thomas is striking.  It is true that a lot of the double tradition is found in Matthew's sermons, but an awful lot of it occurs outside of them (Matt. 3-4, 8, 11, 23, etc.).  Quite a lot occurs outside of Luke's journey section too (Luke 3-4, 6-7).

I do understand, of course, the constraints of the New Testament introduction, but I can't help feeling a disappointment not only that the Farrer Theory is once again ignored but also that data is is not described with the kind of precision that would help students to begin to see the Synoptic Problem accurately.  The authors do not provide additional reading or links to works that would help them to fill out their knowledge in this area and to assess competing claims.

Sometimes I think that I am a bit too noisy on this subject, writing books, articles, websites, podcasting, blogging.  "There's Goodacre going on about Q again!"  But clearly, as far as these authors are concerned, I am as quiet as a mouse.  In a funny sort of way, it's quite reassuring -- I'm clearly not as irritating as I sometimes worry I might be!

Farrer Theory ignored: yet another new example

Regular readers of the NT Blog will be familiar with my occasional notices on how the Farrer Theory often gets ignored in scholarship.  The attempt to make the Griesbach Hypothesis the sole alternative to the Two-Source Theory gives scholars an easier ride, especially in introductory works, in which the establishment of Marcan Priority becomes an argument for the existence of Q.  The latest example of the phenomenon is found in Grant R. Osborne, Matthew (Zondervan Exegetical Commentary on the New Testament; Grand Rapids: Zondervan, 2010): 35-8, a preview of which can now be read online.

Osborne says that "now there are two major camps"..  The first is the "Griesbach Hypothesis" and he goes on, "there are stronger arguments for Markan Priority, especially the form called the "two- or four- source" hypothesis formulated by B. H. Streeter, which says that Matthew and Luke used Mark and Q as well as their own special material (M and L)" (36).   Osborne then provides four arguments for Marcan Priority (36-7) and concludes "All of this makes it likely that Mark was the first Gospel and that Matthew and Luke used Mark as well as Q" (37).  None of the arguments is in fact an argument for Matthew's and Luke's independent use of Mark, and so the conclusion that Marcan Priority necessarily leads to acceptance of the Q hypothesis is illegitimate.

Osborne briefly mentions the Gospel of Thomas as providing an analogue for Q, without exploring the important question of the apparent generic contrasts between the two and he adds that "recent work has made the likelihood of Q more viable" (37), footnoting Carlston and Norlin's 1999 article on statistics and Q. That article certainly points in the direction of a Q document on the assumption of the independence of Matthew and Luke, but it does not speak to the issue of Luke's familiarity with Matthew.

I am now so used to this kind of by-passing of the Farrer Theory that fresh examples come as no surprise.  I have in fact come to expect it whenever I open a new book.  I must confess to a touch of additional disappointment in this case, however, given my careful review of a book to which Osborne contributed, and concerning which we had corresponded, in which I spoke of the problems that arise from ignoring the Farrer theory and discussing the Synoptic Problem solely as a question of the Griesbach Theory vs. the Two-Source Theory (Mark Goodacre, Review of David Alan Black and David R. Beck (eds.), Rethinking the Synoptic Problem, NovT 49/2 (2007): 197-9).

John Poirier on New Testament Introduction and the Synoptic Problem

Over on Verily Verily, Rafael discusses Poirier on NT Introductions and the Synoptic Problem, following on from a fine article in the new JSNT, John Poirier, "The Synoptic Problem and the Field of New Testament Introduction", JSNT 32/2, 179–190. The abstract is as follows:

The synoptic problem is an important and visible subfield within NT studies, yet, for some reason, almost every NT introduction written in the past forty or so years has passed on a defective understanding of that subfield. Two problems in particular plague these NT introductions: (1) their discussions of the synoptic problem tend to rely on a logical argument disproven almost sixty years ago, and (2) they tend to misrepresent the current state of the question by marginalizing the Farrer hypothesis, which today is the Two-Source Theory’s leading competitor.

It's an excellent article, based on a paper Poirier gave in the Synoptics section at the SBL Annual Meeting in 2008. As Rafael points out, the article's claims are further born out by the recent publication of Mark Allan Powell, Introducing the New Testament, which I recently discussed in a post on Mark Allan Powell on the Synoptic Problem.

Mark Allan Powell on the Synoptic Problem

In my previous post, I offered some enthusiasm for the splendid looking new textbook for introductory New Testament classes, Introducing the New Testament by Mark Allan Powell, published by Baker. Regular readers will not be surprised to see me revisiting something of a bête noire in complaining about the way that the Synoptic Problem is treated in introductory texts, and there is no exception here. I must admit that the personal disappointment on this occasion was a little more pronounced than usual. It is several years now since I complained about the way that the Synoptic Problem gets treated in introductory books (The Case Against Q, Chapter 1), and I have done what I can, through books, articles, websites, blogs, to try to generate some awareness of a major alternative to the consensus view. Sometimes, this has met with success. But all too often, I have felt like Kierkegaard's clown, and no more so than here, apparently having made next to no impact on the way that the debate is framed, let alone the solutions that are offered. Let me try to explain why, and please forgive me for focusing in a major way on something that is clearly peripheral to Mark Allan Powell's interests, as it is to many other scholars. The difficulty for people like me is that the introductory textbook can do more than anything else in embedding ideas in students' minds, and it is a shame if they are not even given the framework within which to explore the problem in a balanced way.

Powell's discussion of the Synoptic Problem (here called "the Synoptic Puzzle", 92-100), presents the Two-Source Theory as the solution to the Synoptic Problem and gives the Griesbach Theory as the main alternative (diagrammed on 97, sidebar on 99). I am grateful to Mark Allan Powell for adding a sentence about the Farrer theory (99) after I drew attention to the lack of mention in the manuscript, but all the books under "For Further Reading" are by defenders of the Two-Source Theory, and there are no arguments provided for the existence of Q. Q is largely taken for granted, and then explained. New students, therefore, do not have any framework within which they are able to question Q at the same time as affirming Marcan Priority.

The website provides some supplementary materials, but here things are no more encouraging. There are Chapter 4 flashcards that give "Two Source Theory", "Q", "M", "L", "Four Source Theory" and "Griesbach Theory". Note that Marcan Priority is treated only in partnership with the Q hypothesis and not separately. Under "Extra Content", 4.5 Proposed Solutions to the Synoptic Problem (also available as a PDF), four solutions are listed, "Augustine's Solution", "The Two-Gospel Hypothesis", "The Two Source Hypothesis" and "The Four-Source Hypothesis". In a summary on the "Status of the Synoptic Puzzle in the Twenty-First Century", the Farrer theory and its proponents are not even mentioned.

Then under 4.6, there is Evidence to Support the Two-Source Hypothesis (also available as a PDF). Given the absence of any arguments for the existence of Q in the book itself, I was interested to see what the grounds would be here. Four arguments are given as "Evidence That Matthew and Luke Were Produced Independently of Each Other". All will be familiar to those who have spent any time studying the Synoptic Problem, alas:

(1) With regard to sequence of events, Matthew and Luke frequently agree with one another and with Mark, but they never agree with one another against Mark. This suggests that Mark served as a basic outline used independently by both Matthew and Luke, who sometimes followed him and sometimes did not. If (as an alternative proposal suggests) Mark had copies of both Matthew and Luke and produced an abbreviation of their works, we would expect instances in which Mark departed from a sequence of events followed by both Matthew and Luke, but that never happens.

This argument is problematic because it is expressed in terms of opposition to the Griesbach Theory. The fact that Luke usually follows Mark's order is not a problem for adherents of the Farrer theory, for whom Luke is prioritizing Mark over Matthew and, like Matthew, using it as "a basic outline". The "never" is also incorrect since there are minor Matthew-Luke agreements against Mark in order.

(2) Neither Matthew nor Luke ever includes the other’s additions to the Markan text.

This is simply an error. The reason that the Two-Source Theory invokes categories like "Mark-Q overlaps" and has to spend time discussing "Minor Agreements" is that there are many triple tradition passages in which there are substantial agreements between Matthew and Luke against Mark.

(3) The likelihood that either Matthew or Luke used the other as a source is reduced by what would then be inexplicable omissions of material.

The argument from omission is always a problematic one, not least given the fact that every theory has to cope with omission at some point. It often simply amounts to a statement of what we would have done if we were one of the evangelists. The only example Powell gives of an implausible omission by Luke is Matt. 25.31-46 (Sheep and the Goats). If I were Luke, I don't think I would include it, with its characteristically Matthean judgement day scenario of divisions into two groups, sheep and goats, wheat and tares, wise and foolish, good and bad, and eternal hell-fire for the second group. But that may not be right; it is always a guess to say why we imagine that Luke omitted any material, likewise with the huge chunks omitted from Mark, including material we might think to have been congenial.

(4) The material that Matthew and Luke have in common but that is not found in Mark is never found at the same place in their Gospels.

This is another error. One of the difficulties for the Q theory is that Matthew and Luke sometimes integrate double tradition material with triple tradition material, especially in Matt. 3-4 and Luke 3-4, but also elsewhere. The reason that there is a category called "Mark Q overlap" is that "the material that Matthew and Luke have in common but that is not found in Mark is" sometimes "found at the same place in their Gospels."

So, once we get to arguments for the existence of Q, they are not especially impressive, which is a shame given that the actual text seems to express some confidence in the existence of Q. All in all, I remain optimistic, though. Where Q is sustained either by ignoring the arguments against it, or by reproducing weak arguments in its favour, it is surely only a matter of time before an Introduction to the New Testament is produced that at least takes seriously the Q sceptical view.

Farrer fails to get a look in

Today's post on Jesus Creed, Knowing the Currents 2 was one of those useful reality checks for Q sceptics like me. In summarising the "basic streams that flowed into this current", i.e. "Gospel Criticism", Scot McKnight summarizes source criticism and writes that "The basic theories are what I call the Oxford hypothesis [= the Two-Source Theory] and then also the Griesbach hypothesis." So the Farrer Theory fails to get a look in. Perhaps Scot is right that the Two-Source Theory and Griesbach deserve to be listed side by side as the basics, but I would want to add several brief comments: (1) Griesbach has had no real currency in the UK since Farmer revived it in the US; (2) An adherent of the Griesbach theory recently told me that he thought Farrer had now edged ahead of Griesbach in the US, and it clearly gave him no pleasure to say this; (3) Curiously, the Farrer Theory is itself an "Oxford hypothesis" since it originated there and was dominant for a long time (though not since Christopher Tuckett arrived in Oxford in the 90s); (4) One of the three books Scot recommends for exploring the currents, Sanders and Davies's Studying the Synoptic Gospels, in fact advocates the Farrer Theory.