Ben C Smith

I hinted at this argument for Q in another thread, with no response, but would really like to give it another go. I am looking for responses from those who both (A) question the existence of Q and (B) think that Luke knew Matthew, and not the other way round.

J. Kloppenborg notes on pages 88-89 of Excavating Q that Matthew 10.24-39 appears to be a Matthean compilation derived from a single pass through Q. He divides the section up into a number of separate sayings, but it seems more relevant to me to divide it up into five distinct blocks of sayings, based on their relative positions in Matthew and Luke:

Block A: Matthew 10.24-25 (disciple and master).
Block B: Matthew 10.26-33 (fear not, and before my father).
Block C: Matthew 10.34-36 (no peace, and the divided family).
Block D: Matthew 10.37-38 (loving family more, and following after me).
Block E: Matthew 10.39 (finding and losing).

Why do I slice this section up into these five blocks? Because that is what Luke had to have done on the presumption that Luke knew Matthew. The Lucan parallels come out as follows:

Block A: Luke 6.40 (disciple and master).
Block B: Luke 12.2-9 (fear not, and before my father).
Block C: Luke 12.50-52 (no peace, and the divided family).
Block D: Luke 14.26-27 (loving family more, and following after me).
Block E: Luke 17.33 (finding and losing).

(A fuller chart illustrating how these blocks work can be found on my page on editorial and redactional evidence.)

Now, on the presumption that Luke postdates Matthew and Q is a phantom, it does not bother me that Luke would take a Matthean discourse section and break it up into discrete units, scattering them throughout his gospel. What does bother me, however, is that he would break it up into five discrete units, scatter them, yet endeavor to keep them all in the same relative order: Luke 6.40; 12.2-9, 50-52; 14.26-27; 17.33. Two or three scattered parallels in the same order could be coincidence, of course, but five? That pushes the raw odds to 1 in 120, by my calculation (5!, right?). But feel free to correct my math.

So either Luke scattered this section throughout his gospel, moving each block to a different context but for some reason deciding to keep them all in the Matthean order, or Matthew simply scanned once through Q (or Luke, but remember that this post is for those who think Luke copied from Matthew, not vice versa) on the hunt for material that would fit his mission context. The latter would explain the order nicely; on the former the common order of the parallels seems arbitrary and capricious.

I look forward to responses from any of the seemingly many Q-doubters on this board, whose number I have often tried to join.

Ben.

S.C.Carlson

I'm not sure I see what the problem is supposed to be with the relative order not being a coincidence. After all, on the Farrer theory, Luke already knows the order of Matthew, and the "coincidence" in order is due to, well, Luke's knowledge of Matthew's order.

Coincidence might be a bigger a problem for an oral tradition approach, but that's not being argued here (for good reason). In any event, the relevant calculation is 5 parallels in order out of how many total in all of Q? It seems like there's plenty enough room or coincidence, if coincidence is needed.

What's so arbitrary and capricious about Luke's taking a block of Matthean material that had not yet been used and scanning his draft (in order) for places to put them in? That's just one pass through his draft for this block of material.

I'm not sure whether my response helps, but I don't really see the problem with Luke's compositional procedure in this case.

Stephen

Vorkosigan

That's a fascinating argument, Ben.

Still, there are three reasons it won't work. First, any argument from order is an argument that cuts both ways. If Luke had Q in front of him, why did he preserve the order? Your odds argument just gets worse -- if both Matt and Luke preserved the same order, the odds must be 2 X 120! After all, it is generally agreed upon that Luke preserves Q's order more faithfully than Matt. So all you've done is stood the problem on its head.

Second, you are doing the statistics the wrong way. The question is not whether the order is preserved, but when Luke takes over blocks of material from Matt/Q, how often does he change the order, and how. In other words, you have to judge this against the whole spectrum of Luke's behavior. If Luke takes over 5-Block chunks and only changes the order of 1 or 2 parts, then it would lie within the spectrum of Luke's behavior as a whole to not change any at all. An outlier, to be sure, but acceptable. You need to establish what Luke's habits are, first.

But there's another issue:

Here's the text in question (USCCB-NAB):

• 24 No disciple is above his teacher, no slave above his master.
  25 It is enough for the disciple that he become like his teacher, for the slave that he become like his master. If they have called the master of the house Beelzebul, 12 how much more those of his household!
  26 "Therefore do not be afraid of them. Nothing is concealed that will not be revealed, nor secret that will not be known. 13
  27 What I say to you in the darkness, speak in the light; what you hear whispered, proclaim on the housetops.
  28 And do not be afraid of those who kill the body but cannot kill the soul; rather, be afraid of the one who can destroy both soul and body in Gehenna.
  29 Are not two sparrows sold for a small coin? Yet not one of them falls to the ground without your Father's knowledge.
  30 Even all the hairs of your head are counted.
  31 So do not be afraid; you are worth more than many sparrows.
  32 14 Everyone who acknowledges me before others I will acknowledge before my heavenly Father.
  33 But whoever denies me before others, I will deny before my heavenly Father.
  34 "Do not think that I have come to bring peace upon the earth. I have come to bring not peace but the sword.
  35 For I have come to set a man 'against his father, a daughter against her mother, and a daughter-in-law against her mother-in-law;
  36 and one's enemies will be those of his household.'
  37 "Whoever loves father or mother more than me is not worthy of me, and whoever loves son or daughter more than me is not worthy of me;
  38 and whoever does not take up his cross and follow after me is not worthy of me.
  39 Whoever finds his life will lose it, and whoever loses his life for my sake will find it.

Luke has Mark in front of him, and Matthew too. So he consults his Mark and lo and behold, what does he find? That the reference to Beezebul comes in Mark 3, after the healing of the withered hand and the disciple naming. So in Luke 6 we have the same sequence of healing of withered hand, naming of disciples. When Luke goes back to Matt he sees the Beelzebub reference in 10:25, which occurs only one place in Mark -- Mark 3. So he shoves the disciple and master material in there. Luke isn't following Matthean order -- he's repairing Markan order.

The next set of Matthean material, Block B, is 10:26-33. Open Mark 4 and you will find similar references to hidden being revealed.

Block C has no Markan parallel, so Luke sticks it wherever he wants.

Block D has no Markan parallel, so Luke sticks it wherever he wants.

Block E parallels Mark 8:34.

In other words, A, B, and E more or less echo Markan order. Luke preserves Markan order more faithfully than Matthew. So really, are we looking at A Case of Q, or a preference for Markan order when possible? I would argue that Luke has simply chosen to follow Markan order -- in fact, to return the Matthean material to its Markan place -- and then preserved the order of Blocks C + D.

Vorkosigan

Vorkosigan

BTW, have I told you what a welcome addition you are to the halls of IIDB, Ben?

Vorkosigan

Ben C Smith

But if Luke is willing to break apart and scatter the discourse in the first place (again, not a problem on its own), it seems odd to me that he still intends to keep the separate pieces in order.

How many total parallels in Q may be important to the 2SH, but not so much to me. I am starting from the ground up.

In fact, I admit that the Kloppenborg argument is a little inconvenient for me. I have been wishing to see the Q material as a combination of Luke knowing Matthew (the Mark-Q overlaps especially), oral tradition, liturgical tradition, and a loose body of separate sayings texts. This particular Q argument, however, does make it appear that a certain sayings text was extensive enough for Luke to follow (at least roughly) in order throughout his own gospel. I have tried to wish it away, but so far with little success.

Because the Matthean catena is all in one spot. Scanning Luke in order is natural; but that is not my problem. My problem is scanning Matthew 10.24-39 in order.

It seems coincidental that the five separate contexts that Luke lit upon to situate the five blocks happen to fall in the same order throughout his own text. I can see where if he has the entire gospel of Matthew before him, and is trying to use most of it, he might try to stay mostly in the Matthean order for the sake of convenience. However, in this case all he has to worry about is one small section of Matthew, 10.24-39. As he is scrolling or flipping through his own gospel draft, looking for good spots into which to drop the pieces, there seems to be nothing compelling Luke to use 10.24-25 before he uses 10.26-33, and nothing compelling him to use 10.26-33 before 10.34-36, and so forth.

This is what makes the argument unidirectional, IMHO. The concept of the pass is important when scanning an entire gospel for material to use; it seems quite irrelevant when scanning only 16 verses. Scanning the gospel of Luke (or a Q document) and thus keeping its material in order is not problematic. Scanning (if that is even the word for it) Matthew 10.24-39 and keeping its material in order seems coincidental or arbitrary.

Thanks for the thoughtful response.

Ben.

Ben C Smith

Thanks, Michael. High praise indeed. I am blushing.

I am suddenly pressed for time, but will get back to your arguments as soon as I can.

Ben.

the_cave

Huh, this conversation could prove interesting...

S.C.Carlson

Most critics seem to have a problem with the scattering aspect of it. The ordering just means that there is a literary phenomenon, as opposed to an oral one.

Well, it's a question of cherry-picking. There are lots of five-block sections of sequence of Double Tradition either in Matthew or Luke, but this is the only one with order I can find for all five of them. It's not hard to flip 5 heads in a row if you've got hundreds of attempts to do it.

Under Luke's knowing Matthew, however, there was a sayings text that was extensive enough for Luke to follow roughly in order. That sayings text is Matthew 10.

With Matt 10:24-39 before Luke, wouldn't scanning it in order be the most natural way of going through it? How else should Luke do it? Backwards?

If there's nothing compelling Luke, wouldn't it be easy just to default to the order of his source?

Maybe we're really back to the collecting/scattering argument: scanning to collect material seems less arbitrary than scanning to scatter material? "Scatter" may be a biased term in that it implies that the dispersal is into artribrary contexts--but has that value judgment really been established?

In sum, I don't understand why Matt 10:24-39 and its Lukan parallels are problematic. The ordering may well be coincidental, and if not, it only indicates that the ordering was transmitted by a document. Even so, I can't see how, without more, that this agreement in ordering could tell us whether the documentary source of the ordering was Matthew, Q, or Luke.

Stephen

Ben C Smith

Drat. First my car refuses to start, and now this!

Preserving the order of your source material is easy. What might require an explanation, particularly if we are imagining a scroll, is scrambling the order of your source material. On my scenario for Matthew 10.24-39, at least, Luke preserved the Q (or call it what you will) order because he was scrolling through Q while composing his own gospel. This part is quite natural, and needs no explanation. Matthew, on the other hand, has mixed up the Q order in his effort to create a discourse related to missions in his tenth chapter (just one of five blocks of teaching material in the first canonical gospel). How did he do this? In this case, if there was a Q, we get a precious glimpse into the actual logistics involved. All he had to do was scan once through Q, extracting, one pericope at a time, any (at least loosely) related material that he could find. If he is scrolling or flipping through Q in order, then his new block of mission material will also come out in order. Again, this seems quite natural, needing no explanation.

My argument is that the opposite argument, that Luke was taking material from Matthew, is less natural.

This is the crux of the issue. Arguments from order cut both ways only if they are comparing texts of similar length or complexity. In this case we are comparing the entire gospel of Luke (or at least from Luke 6.40 to Luke 17.33) with one little passage in Matthew (the material in 10.24-39).

That single Matthean block winds up as 5 widely separated Lucan blocks. No problem so far. On the Mark-Matthew (Farrer) theory Luke is demonstrating his customary willingness to break up Matthean discourse segments. But now we notice that those 5 blocks happen to fall in the same order as Matthew had them. With entire documents convenience is at issue; retaining the same relative order as the source material seems a relatively easy thing. With a single limited passage, however, what is there to compel an author, one whom we have already demonstrated to be willing and able to break it apart, to keep all the pieces in the original order? Is it somehow easier to pick out Matthew 10.24-25 before 10.26-33? I think not. The entire passage, only 16 verses, is visible at a glance. Little or no scrolling, page turning, note flipping, or what have you ought to be necessary.

But let us get back to the presumption that Q did not exist. It ought to be noticed that, on the presumption that Luke knew Matthew, in breaking up Matthew 10.24-39 he is reordering Matthew in the broad sense. To place the first block at Luke 6.40, the last at Luke 17.33, and the middle blocks at intervals in between is to make a play at scrambling Matthew as a whole, since there is a lot of Matthean material in between Luke 6.40 and 17.33 that is now out of order relative to the scattered pieces of 10.24-29.

The Kloppenborg argument, at least as I am applying it, is very local and specific, not general across the entire gospel of Matthew. On the Mark-Matthew hypothesis, Luke would be preserving the Matthean order only in a very narrow and quite puzzling sense: He has removed Matthew 10.24-39 in an absolute sense from its original Matthean order and context (note that none of it survives in either of the parallel Lucan mission speeches in Luke 9 and 10!), yet for some reason has simultaneously kept the relative order of its constituent parts, despite the fact that those parts are now strewn across a vast portion of his gospel, inserted into different contexts entirely.

This poses the question: Is this resultant common order a coincidence (that is, Luke did not try to maintain the same order, but it fell out that way anyhow), or is it by design (that is, Luke wanted to retain the same order)?

If it is coincidence, then it is a bit larger coincidence than I, for one, am able to swallow, pending further evidence. If it is by design, then Luke, though quite willing to break up the discourse, is yet for some reason bent on keeping the material in the same order, despite the fact that it no longer seems to make one whit of difference, as each block is now in a different context anyway. What would it matter if block E now preceded block A in the gospel of Luke? What is at stake? What reason did Luke have for retaining the order of this Matthean material while abandoning its cohesion?

As to the question of Lucan coincidence, you state:

I am not certain what you are up to here. On the one hand, the statistics are valid only if Luke did not intend to retain the relative Matthean order, only if the results as we have them now were coincidental; but in that case Lucan habits elsewhere do not matter, since the results were accidental, not by Lucan design. On the other hand, only if Luke wanted to retain the relative order do his habits elsewhere matter, but then the statistics have no meaning, since they are meant only to show the odds in the case of an utter coincidence! Looking to his authorial habits in order to explain the statistical probability of a chance arrangement is a non sequitur.

The path of argumentation bifurcates at coincidence versus design. You have to either explain why Luke changed the absolute order, but kept the relative order, of Matthew 10.24-39 on purpose or reckon with the odds of it coming out that way by chance. Not neither, not both. Just one at a time.

As to the question of Lucan design, you state:

An ingenious argument in the abstract. In your view, then, the blocks of Matthew 10.24-39 end up in the same relative order, not because they appeared in that order in Matthew, but rather because they appeared in that order in Mark, and Luke likes to preserve the Marcan order of things. Let us look more closely at this possibility, block by block.

First, block A:

I have three major problems with this scenario:

1. You are correct that Mark 3 contains the healing of the withered hand, the commission of the twelve, and the Beelzebub controversy. You are also correct to note that the Lucan parallel to Mark 3 is basically Luke 6, in which we find the healing of the withered hand and the commission of the twelve... but no Beelzebub controversy. For whatever reason, Luke has moved that pericope to Luke 11.14-23. So your hypothesis requires Luke, influenced by the Beelzebub reference in Matthew 10.24-25, to have moved that saying to where the Marcan Beelzebub pericope used to be. What was it supposed to do there, serve as a placeholder?

2. In point of fact, Luke does not move Matthew 10.24-25 into any Marcan context at all; he moves it right into the middle of the sermon on the plain, an utterly Matthean context if there is no Q. He has moved the saying, in other words, not from the Matthean order into the Marcan order, but rather from one Matthean context to another.

3. On your hypothesis it was the reference to Beelzebub in the saying about disciples and masters that influenced Luke to move it in parallel with Mark 3, where Mark (but not Luke!) has the Beelzebub controversy, yet Luke did not retain the reference to Beelzebub! Luke 6.40 does not mention Beelzebub; it has only the disciple and the master. If the Beelzebub reference was important enough to affect his decision on placement, why was it not important enough to even retain in the saying?

Now block B:

Block B, which is Matthew 10.26-33, does indeed contain the saying about the hidden being revealed. And Luke has moved block B to his central section, 12.2-9. But how does that relate to the Marcan state of affairs? First, the Lucan central section is an addition to the Marcan framework; it is not a Marcan context. Second, the Lucan parallel to Mark 4, such as it is, is not anywhere near 12.2-9; rather, it is in Luke 8.4-18:

Mark 4.2-25.

4.2-9: Parable of the sower.
4.10-12: The mystery of the kingdom.
4.13-20: Explanation of the parable of the sower.
4.21-25: The nature of parables (the hidden revealed).

Luke 8.4-18.

8.4-8: Parable of the sower.
8.9-10: The mysteries of the kingdom.
8.11-15: Explanation of the parable of the sower.
8.16-18: The nature of parables (the hidden revealed).

Your hypothesis of Luke following the Marcan order explains nicely why Luke has placed his parallel to Mark 4.22 at Luke 8.17. But how does it even touch why Luke has placed Matthew 10.26-33 (which includes much more than just the hidden revealed) at his own 12.2-9?

On to block C:

Not exactly true. Block C has Matthew 10.35-36 = Luke 12.51-52, the divided family, which closely resembles Mark 13.12 = Luke 21.16. Why not stick this pericope where Mark has it, or something like it, on Olivet?

Time for block D:

Again not exactly true. The saying about taking up the cross is found at Matthew 16.24 = Mark 8.34 = Luke 9.23.

Finally, block E:

I presume you mean Mark 8.35, on finding and losing. But again, Luke has a parallel for Mark 8.35, and so does Matthew; the parallels are Matthew 16.25 = Mark 8.35 = Luke 9.24. The parallel at Mark 8.35 explains nicely why Luke has his own 9.24 there, but how does it explain why he moves Matthew 10.39 to Luke 17.33, again part of his own central section, with no Marcan context?

Not at all. Blocks B and E, if Luke were thinking of Mark at all, would fall at Luke 8.17 and 9.24, respectively, instead of at Luke 12.2-9 and 17.33. And block A, as discussed above, has the most problems of all if Mark had anything to do with the Lucan decision.

Ben.

Ben C Smith

If Matthew knew Luke, then Matthew scattered a lot of Lucan material. If Luke knew Matthew, then Luke scattered a lot of Matthean material. If both Matthew and Luke knew Q, then either or both of them scattered a lot of Q material.

The scattering itself, therefore, does not bother me, at least not very much. Of course, there may be some cases in which the scattering itself seems weird, but I have not yet argued that here.

Oh, not nearly hundreds. We are not even talking about pericopes here. We are talking about contiguous (or nearly contiguous) blocks, whether each block contains one pericope or two or many.

I intentionally turned what Kloppenborg had as 10 (IIRC) individual pericopes in order into my own arrangement of 5 individual blocks in order, precisely for this reason. Dividing the passage up into smaller pieces may be quite reasonable from a literary standpoint, but does not change the fact that, for example, the close connection of Luke 12.2-7 with 12.8-9 probably has something to do with the close connection of Matthew 10.26-31 with 10.32-33, whether Q mediated or not. So we have to consider Matthew 10.26-33 = Luke 12.2-9 as a single block.

If you wish to use pericopes instead of blocks to try to push the overall number into the hundreds, then my 5 blocks quickly become at least 8 pericopes, and the odds (since we are on the coincidence branch of the argument at this point) suddenly skyrocket.

Of course, this kind of blocking is not without its own issues. For example, shall we split the sermon on the mount up into several blocks because Luke omits several sections and removes others? Or shall we retain it all as a single block because it is all a single sermon, and the position of each individual pericope or block within the whole is thus explained even if others have been removed? I tend to the latter. On the other hand, surely each removed pericope (like Matthew 5.25-26 = Luke 12.57-59) would be a single block, unless (like Matthew 6.19-21; 24-34 = Luke 12.32-34; 12.22-31) they end up in the same context yet again.

On this score I even considered regarding Luke 12.2-9 and 12.50-52 as a single block, but the intervening matter (Luke 12.10-49) switches gears several times, thus rendering a common context meaningless.

All this just to say that the number of coin flips is not in the hundreds. Perhaps I shall take a look into roughly how many it would be. But it could not possibly be more than 70, since AFAICT even the most generous estimates of the extent of Q tend to divide it into 50-70 pericopes, and we would be talking about blocks, so there should be quite a bit fewer yet.

I am not certain whether you mean Matthew 10 as a whole or just 10.24-39 as in my OP. If the former, please elaborate. If the latter, then how are sixteen verses extensive enough to compel Luke to reproduce them in relative, though not absolute, order? Surely he would not have to flip notes, turn pages, or unroll a scroll very much to have 16 verses before him virtually all at once.

No, I would think that he would default to the order in which acceptable contexts appeared in his gospel. This has been my contention all along, as when I wrote, for instance:

And you implied that the Lucan contexts are the issue when you asked in post 2 on this thread:

If indeed the issue is where in the gospel of Luke this material will fit, as we seem to agree, then the order of the 16 verses in Matthew 10.24-39 should have little or nothing to do with where the 5 blocks wind up. Unless just by chance the suitable contexts strewn across Luke happen to match the five Matthean blocks in order. And with that we are back to coincidence.

No, not in the abstract. But this is not an abstract argument. To scatter from a very compact text into a very expanded text, yet retain the same order, seems rather less natural than to scan the expanded text in order so as to compile the compact text.

I agree that scatter may have that connotation, and in fact I am against that connotation in this case, since what I am seeing is a suspicious pattern of dispersal, not a random distribution. Perhaps disperse is a better term.

Ben.