Alan Garrow: Scrolls are so 'last century'

Covering all topics of history and the interpretation of texts, posts here should conform to the norms of academic discussion: respectful and with a tight focus on the subject matter.

Moderator: andrewcriddle

User avatar
Peter Kirby
Site Admin
Posts: 8483
Joined: Fri Oct 04, 2013 2:13 pm
Location: Santa Clara
Contact:

Alan Garrow: Scrolls are so 'last century'

Post by Peter Kirby »

You can view this here (video 3/5): https://www.alangarrow.com/mch.html/
Here's the associated article: https://www.alangarrow.com/uploads/4/4/ ... demia_.pdf

Alan Garrow considers that there are different limitations in the use of scrolls that don't apply with the use of codexes. Garrow considers that scroll users "paraphrase rather than copy verbatim," "rarely change source order," and "switch between sources infrequently" (e.g. Josephus). But codex users may "copy verbatim," "change source order radically," and "switch between sources frequently" (e.g. Tatian). These fall out of the different technologies. The "rarely change source order" and "switch between sources infrequently" aspects come from the scroll being unwieldy. The "paraphrase rather than copy verbatim" idea makes sense under the assumption of a scribe working alone, but of course two scribes working together as reader and writer would not have a difficulty with this, so this is one potential weakness of the model. The codex was easier to handle as it held itself open for viewing well enough, so it was easy even for a scribe working alone to copy verbatim and to switch between sources frequently. The codex also made random access relatively trivial, much unlike rolling a scroll.

I found this interesting because it is frequently remarked that we don't pay enough attention to the practicalities of how authors and scribes operate, but seldom is this remedied in practice. I appreciate Garrow's effort to do so here.

Garrow assumes somewhat reasonably that a scroll user is a scroll user, and a codex user is a codex user. That is, we should reasonably expect to find that they use their sources similarly in one case and another, with respect to the habits noted above regarding each. A hypothesis that requires them to treat two different sources differently, with respect to these habits that fall out of the technology used, is either weaker or ad hoc (like, I suppose, assuming they had both a scroll and a codex that they alternated between).

This setup allows Garrow to inquire about how Matthew and Luke, respectively, treat Mark, given that Markan priority is widely accepted. This can provide some insight into how each of them can be considered, which would provide a framework for considering their use of other sources.

Garrow finds that Matthew switches in and out of Mark frequently, while also reordering Mark and copying verbatim.

He also finds that Luke alternates much less frequently, while also reordering Mark less and copying less verbatim.

This lets Garrow conclude that he regards Matthew as a codex user and Luke as a scroll user. In general, the evidence suggests that the codex was ascendant in this time period, and that Christians were definitely among those making this transition, so there's a general plausibility in assuming that these two different technologies could be used by two different early Christian authors.

Then, Garrow looks at the double tradition in Luke, and he finds that there is frequent alternation between this double tradition and Luke's special material (not paralleled with Mark or Matthew), even within the sections where Luke has set aside Mark. This requres that Luke would have alternated frequently between looking at his source (whether a lost source or Matthew) and working in material from elsewhere. On the other hand, if Luke expanded Mark, then there's no evidence for frequent alternation between double tradition and sondergut Luke as separate sources. Only Lukan posteriority implies that Luke was switching frequently and in a manner dissimilar to his use of Mark.

Matthew is again consistent with the habits of a codex user in the double tradition, on Matthean posteriority, both rearranging material and alternating frequently. This allows Matthew to create the Sermon on the Mount and to find various, different ways to weave the double tradition into the Markan narrative.

This leads Garrow to conclude in favor of Matthean posteriority.

PS - if you're so inclined, substitute "Evangelion" for Luke above (?). It's not clear to me that all the details of the argument works exactly the same, but if you think Luke is posterior to Evangelion, then you'd seem to assume that Evangelion is evinced as a scroll user (according to Garrow's argument) in alternating between blocks of Mark and blocks of non-Markan material. Then likewise you'd conclude that Matthew used Mark and Evangelion. So perhaps the same basic principles apply.

PPS - If you think Evangelion preceded Mark, I award you no points, and may God have mercy on your soul. /s
User avatar
Ken Olson
Posts: 1341
Joined: Fri May 09, 2014 9:26 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by Ken Olson »

Peter Kirby wrote: Fri Jun 23, 2023 9:09 pm PPS - If you think Evangelion preceded Mark, I award you no points, and may God have mercy on your soul. /s
:D

IMNSHO, Garrow is one of many scholars of the synoptic problem who sees the data through the lens of his own hypothesis (and that of the 2DH). I have criticized his lack of even-handedness in using the term 'unpicking' (in the "Streeter's Other Synoptic Solution' paper to which you linked) in a blog post here:

https://kenolsonsblog.wordpress.com/202 ... %9d%90%a6/

I will try to comment anon on some recent discussions involving Alan Garrow, Mark Goodacre and me, on the specific issue of conflations and how it supports or does not support the Matthean Posteriority Hypothesis (of which Garrow's Matthew Conflator Hypothesis is a form) or Farrer more.

Best,

Ken
gryan
Posts: 1120
Joined: Fri Jun 01, 2018 4:11 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by gryan »

Ken Olson wrote: Sat Jun 24, 2023 7:24 am
I will try to comment anon on some recent discussions involving Alan Garrow, Mark Goodacre and me, on the specific issue of conflations and how it supports or does not support the Matthean Posteriority Hypothesis (of which Garrow's Matthew Conflator Hypothesis is a form) or Farrer more.
You have my attention!

In the video, Scrolls are so 'last century', the argument that Matt used a codex and Lk used a scroll was, imho, very persuasive.
https://www.alangarrow.com/mch.html/
User avatar
Ken Olson
Posts: 1341
Joined: Fri May 09, 2014 9:26 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by Ken Olson »

I'm first going to address the unsubstantiated 'unpicking' charge in the first video (video 1/5) and look at two of the synopses Garrow uses in the video to attempt to substantiate it:

https://www.alangarrow.com/mch.html/

Garrow: If Luke is using Matthew, there are occasions where he has to remove Matthew's additions to Mark as if he has a surgeon's instinct for this green material, leaving behind the orange material that come from Mark. Here's another example in the Beelzebul Controversy and again in the Parable of the Mustard Seed. This is a very difficult operation to perform practically, and it's also difficult to see why Luke might have been motivated to do such a complicated thing. So this is a good argument against Luke's use of Matthew.

Is what Garrow claims in the narration actually what we see in the synopsis? That is, do we actually see that, if Luke used Matthew, he must have intentionally, and complicatedly, gone through Matthew and, with a surgeon's instinct, carefully removed Matthew's additions to Mark (in green) to use in his own gospel while leaving the material Matthew took over from Mark (in orange) behind.

So, first, here's Garrow's Synopsis of the Beelzebul Pericope

Garrow's Beelzebul Pericope Synopsis.png
Garrow's Beelzebul Pericope Synopsis.png (1.48 MiB) Viewed 1295 times
When we look at the synopsis, the first four words Garrow colors in green (Matthew's additions to Mark used by Luke) Βεελζεβοὺλ ... ἄρχοντι τῶν δαιμονίων are also found in Mark, so if Luke's intention was to carefully remove all of Matthew's Markan material and use on Matthew's addition to Mark, his surgeon's instinct seems to have failed him in this case. More likely, Garrow is simply wrong and Luke had no such intention and carried out no such complicated practice. Garrow has mis-categorized the four words as one of Matthew's additions to Mark, when in fact it is Matthew's use of Mark, and he has also failed to realize he needs more categories. He has color codes two types of double agreement - Mark-Matthew (orange) and Matthew-Luke (green), but he does not have a category for triple agreements (Mark-Matthew-Luke) and this category counts against his claim that Luke has used only material Matthew added to Mark.

If one were to protest that the word Beelzebul is separated from the other three words in the sequence, and therefore they are not in continuous verbatim agreement, we should note that this is the case in Luke as well. It isn't really clear what rule Garrow is following in his color coding, or if he's following a rule at all. Sometimes word roots, minus prefixes and endings count as agreements, and sometimes they do not. Matthew has the word μερισθεῖσα twice. Garrow colors one of these uses green and the other orange. The same thing happens with Matthew's two uses of σταθήσεται - one green, one orange. ὁ Σατανᾶς is colored orange (a Mark-Matthew agreement) even though it also occurs in Luke. The word βασιλεία occurs twice in each of the gospels, but is marked as a Matthew-Luke agreement in green.

It is possible that the data do somehow support Garrow's claim (I doubt it, but it's possible) but his synopsis fails to do so and just seems arbitrary.

Garrow's Synopsis of the Parable of the Mustard Seed
Garrow's Mustard Seed Synopsis.png
Garrow's Mustard Seed Synopsis.png (865.06 KiB) Viewed 1295 times
The Parable of the Mustard seed is shorter, but we see the same phenomena in it as we did in Beelzebul. The words κόκκῳ σινάπεως are colored green (Matthew-Luke agreement) even though they are also found in Mark. And he final five words in Mark τὰ πετεινὰ τοῦ οὐρανοῦ
κατασκηνοῦν are marked in orange (Mark-Matthew) even though they are a three way agreement, with Luke having a different ending for the final word and both Matthew and Luke interrupting the sequence by insertion of a single word.

The color coding of Garrow's synopsis seems arbitrary. The data in the synopses do not support the claim that Luke has complicatedly distinguished between Matthew's Markan words and his non-Markan words and omitted the former and included only the latter. To be sure, Luke has omitted some words that Matthew took over from Mark, but he has included others. To use and abstract model, if three writers A, B, and C tell a story at roughly similar length, and B uses A, using 50% of A's words and 50% of his own, and then C uses B's story and keeps 50% of B's words and 50% of his own, then we would expect to see B agree about 50% with A and C, but A an C will agree only 25% with each other. There is no need to hypothesize a complicated intentional process to explain this.

Best,

Ken
User avatar
Ken Olson
Posts: 1341
Joined: Fri May 09, 2014 9:26 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by Ken Olson »

I'm reposing this that I originally posted to Facebook's Synoptic Problem Study Group (of which Alan Garrow is a member) back on Feb. 27.

This is part 2 of a critique of Alan Garrow’s ‘Gnats, Camels, and Matthew’s use of Luke'. JSNT online 2023, PDF in this forum’s Files section. I posted Part 1 back on January 27.

Alan Garrow: Before leaving this passage, it is worth pausing to notice that, when ‘followed by the crowds’ is viewed in context, a more obvious and reliable method for determining the direction of dependence becomes available (see Synopsis 1).

Synopsis 1 shows Matthew = Mark agreements in italic underline, and Matthew = Luke agreements in bold. This arrangement highlights a phenomenon that occurs elsewhere, most famously in the Beelzebul Controversy. In these passages, Mark and Luke have somewhat different accounts of the incident in question whereas Matthew agrees closely with portions of both. The simplest explanation for this is that Matthew conflates the two. By contrast, FH Luke is required to have a surgeon’s instinct for elements in Matthew that Matthew has not taken from Mark and a distinct distaste for elements of Matthew that Matthew has taken from Mark. [‘Gnats, Camels’, 9-10]

Leaving aside for the moment whether the phenomenon Garrow describes occurs elsewhere, I want to examine Garrow’s claim that it occurs in the case of Synopsis 1, which he provided. Garrow’s method (insofar as he describes it here) may be obvious, at least to him, but it is far from reliable, as analysis of the synopsis Garrow has provided will show.

Synopsis 1 - The Feeding of the Five Thousand.jpeg
Synopsis 1 - The Feeding of the Five Thousand.jpeg (79.64 KiB) Viewed 1270 times

By my count:

Mark has 37 words

Matt also has 37 words, 20 of them (a bit over half) from Mark

Luke has 32 words, of which 6 whole words and 2 partial words agree with Matthew. Of the 6 whole words, 2 agree with Mark and are in sequence and the other four are from the characteristically Matthean phrase under discussion.

How does this data justify Garrow’s claim that on the Farrer Hypothesis ‘Luke is required to have a surgeon’s instinct for elements in Matthew that Matthew has not taken from Mark and a distinct distaste for elements of Matthew that Matthew has taken from Mark’?

Of the 17 words in Matthew that were not in Mark, Luke has only 4 whole words (in unbroken sequence in Matthew, but interrupted sequence in Luke) and 2 partial words, and he also has takes over 2 of the words from Matthew that were in Mark (in sequence). Garrow’s description of the data does not correspond to the data what we actually see in the synopsis.

Best,

Ken

PS - Yes, I realize I have not yet addressed the scrolls vs codices issue and its relationship to the issues of Matthew and Luke's order of the double tradition and relative levels of verbatim agreement. Still hope to get there.
User avatar
Ken Olson
Posts: 1341
Joined: Fri May 09, 2014 9:26 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by Ken Olson »

This was a follow up post to the previous one, examining Garrow's claim that he has 'a more obvious and reliable method for determining the direction of dependence' and that 'Mark and Luke have somewhat different accounts of the incident in question whereas Matthew agrees closely with portions of both. The simplest explanation for this is that Matthew conflates the two' (see previous post).

Let’s look at another synopsis from later in the same peircope: Matt 14.9a / Mark 6.39-40 / Luke 9.14-15.

The three evangelists are telling the same story, but here there are few agreements (even fewer in the Greek than in the English translation). Matthew and Mark agree on the sitting down on the grass and Mark and Luke agree on the sitting down (though they use slightly different words) and that the people sitting down are in fifties. So we see in the synopsis that Mark has some Matthean material and some Lukan material, so obviously Mark has conflated Matthew and Luke, right?

Synopsis of the Feeding of the 5000.jpeg
Synopsis of the Feeding of the 5000.jpeg (108.96 KiB) Viewed 1264 times

Well, no. Conflation is one of the possibilities, and possibly it’s even the most obvious. But an explanation that might seem to be the most obvious is very often not the most logical one, nor the correct one.

Griesbachians look at the triple tradition passages where Mark is the middle term between Matthew and Luke (i.e., besides the material all three have, Mark also has some material in common with Matthew and some in common with Luke, but Matthew and Luke have very little in common with each other) and think it’s obvious that Mark has conflated Matthew and Luke (after all, he has Matthean material and Lukan material, doesn’t he?). Some Griesbachians go so far as to claim that Markan priority (which entails the rejection of the theory of Markan conflation of Matthew and Luke) is an untenable theory because it would require that Matthew and Luke intentionally collaborated to produce the phenomenon we see in the texts.

They are wrong about that. In the abstract, if we presume three writers A, B, and C, and B and C independently rewrite A using 50% of A’s words and 50% of their own, it will (almost) always produce the phenomenon that the Griesbachians see as conflation. There will be 25% three way agreement among all three, A and B will agree an additional 25% of the time and A and C will also agree an additional 25% of the time. Agreements of B and C against A will be rare, occurring only when B and C coincidentally agree in the 50% of the words they are providing themselves and not taking from A.

As a general rule, people argue for conflation as an explanation for the middle term only when and where it supports their favored source hypothesis and are skeptical about it the rest of the time. Griesbachians argue for Markan conflation as explanation for those triple tradition passages where mark is the middle term, but Markan priorists rightly reject that inference. It is a possible explanation, but not a required one, and in many cases there are strong arguments against it.

As a Markan priorist, Garrow does not want (at least I don’t think he does) to argue for Markan conflation of Matthew and Luke as an explanation for those triple tradition passages in which Mark is the middle term. He also should not assume that Matthean conflation of Mark and Luke is the best explanation for passages in which Matthew is the middle term. It may be simple to do so, and it may even appear obvious to him, but it’s not a reliable method.

Best,

Ken
User avatar
Peter Kirby
Site Admin
Posts: 8483
Joined: Fri Oct 04, 2013 2:13 pm
Location: Santa Clara
Contact:

Re: Alan Garrow: Scrolls are so 'last century'

Post by Peter Kirby »

I agree with your criticism of the synopsis and "unpicking" argument from Garrow.
User avatar
Ken Olson
Posts: 1341
Joined: Fri May 09, 2014 9:26 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by Ken Olson »

I started a long post responding to Garrow’s third video, ‘Scrolls Are So Last Century’, but I realize that will take a long time to write and I may never finish, so I will post shorter individual criticisms and may convert them into a blog post later.

So, first criticism: Garrow oversimplifies the issue by suggesting there are two possible models and then giving overgeneralized descriptions of those models.

I allow that the invention or adoption of the codex (c. 100 CE?) did make it easier to find any particular location of a text, but it is false to suggest that ‘scroll users’ paraphrased rather than copying verbatim. A writer’s decision whether to copy his source closely or not depends on a number of things, but whether his source is a scroll or a codex is not one of them. I’m not aware that anyone before Garrow ever claimed this (and if anyone after Garrow has, I suspect they are proponents of the MPH dependent on him).

Among Greco-Roman writers, it was rare for an author to copy his or her source closely when writing in the genres of history or biography, because in those genres the author was expected to demonstrate his own literary style and not to borrow that of another (borrowing facts was fine, just not wording). (Caveat: These norms apply to Greco-Roman writings; in Semitic languages, as for instance the Chronicler’s use of Kings, borrowing the wording of a source seems to have been acceptable).In genres where the author’s own artistry was less of an issue, such as school or technical (including medical or military) treatises, using the source’s language was more acceptable. If these genre norms changed at the time the codex was invented, I’m not aware of it.

Garrow generalizes about ‘scroll users’ and ‘codex users’ giving remarkably few examples (Josephus and Tatian, and then Luke and Matthew). According to Garrow, Luke trends to paraphrase like Josephus (the 'scroll user’), while Matthew is more likely to copy verbatim (as a 'codex user').

In Robert MacEwen’s study Matthean Posteriority (2015), he compared the number of Significant Verbal Agreements (verbatim agreements of four words or more in sequence) found in Matthew’s use of Mark with that found in Luke’s use of Mark. According to MacEwen, 23.5% of Matthew’s wording when following Mark is found in SVA’s, compared to 14.5% of Luke’s.

While MacEwen’s numbers, 14.5% for Luke, and 23.5% for Matthew, do show that Matthew follows Mark’s wording more closely than Luke does, they are still relatively close to each other. They do not justify the claims that Garrow makes for them, that Luke is a typical scroll user paraphrasing his sources while Matthew is a typical codex user copying verbatim. As far as I’m aware, no one has ever done a study like MacEwen’s on other Greco-Roman authors and their sources to determine how much of their wording is found in SVA’s. My impression in looking at Josephus’ Antiquities and his ostensible sources in the LXX and Maccabees is that he has nowhere near 14.5% SVA’s, and Luke is more like Matthew than he is like Josephus. I do not think that the difference between 23.5% SVA’s in Matthew’s use of Mark and 14.5% SVA’s in Luke’s use of Mark can be used justify the claim that the two evangelists are using different technologies. They are just not that different.

Best,

Ken

I have a blog post discussing Rob MacEwen’s work here:

https://kenolsonsblog.wordpress.com/202 ... t-macewen/
User avatar
Peter Kirby
Site Admin
Posts: 8483
Joined: Fri Oct 04, 2013 2:13 pm
Location: Santa Clara
Contact:

Re: Alan Garrow: Scrolls are so 'last century'

Post by Peter Kirby »

Ken Olson wrote: Sun Jun 25, 2023 8:50 am I started a long post responding to Garrow’s third video, ‘Scrolls Are So Last Century’, but I realize that will take a long time to write and I may never finish, so I will post shorter individual criticisms and may convert them into a blog post later.
Sounds good. (I see shades of John 21:25 here.)
Ken Olson wrote: Sun Jun 25, 2023 8:50 am I allow that the invention or adoption of the codex (c. 100 CE?) did make it easier to find any particular location of a text, but it is false to suggest that ‘scroll users’ paraphrased rather than copying verbatim. A writer’s decision whether to copy his source closely or not depends on a number of things, but whether his source is a scroll or a codex is not one of them. I’m not aware that anyone before Garrow ever claimed this (and if anyone after Garrow has, I suspect they are proponents of the MPH dependent on him).
Good point. I also made this kind of criticism in the OP.
gryan
Posts: 1120
Joined: Fri Jun 01, 2018 4:11 am

Re: Alan Garrow: Scrolls are so 'last century'

Post by gryan »

Re: "Garrow generalizes about ‘scroll users’ and ‘codex users’ giving remarkably few examples (Josephus and Tatian, and then Luke and Matthew). According to Garrow, Luke trends to paraphrase like Josephus (the 'scroll user’), while Matthew is more likely to copy verbatim (as a 'codex user')."

Thanks for your comments Ken.

Do you know if the author has written this argument in a paper?

The color coding in the video gives the impression that Matt does a lot of switching back and forth whereas Lk goes from one source for a while back to another for a while (if memory serves). It made the argument "look" very impressive.
Post Reply