Underestimating serious problems
While researching background information on a post I’ve been picking away at for several weeks, I came across a problem that bothered me to the point where I had to pull some books out of storage.
As you no doubt recall, the consensus explanation for the Synoptic Problem posits a “Q” source that Matthew and Luke used. But they also copied Mark.
According to the theory, the authors of those two later gospels used their sources completely independently, and edited their material according to their own tendencies. So when we happen upon a passage in which Matthew and Luke redact Markan source material in exactly the same way, we take notice. We call these passages “minor agreements,” in keeping with NT scholarship’s penchant for underestimating potentially fatal flaws.
Sometimes these agreements span just two or three words, and even in this case it’s only five words, but remarkable nonetheless. As the woman with the issue of blood approaches Jesus through the crowd, she reaches out.
- Mark 5:27:
ἐλθοῦσα ἐν τῷ ὄχλῳ ὄπισθεν ἥψατο τοῦ ἱματίου αὐτοῦ
(having come in the crowd behind touched the clothing of him)
- Matthew 9:20b:
προσελθοῦσα ὄπισθεν ἥψατο τοῦ κρασπέδου τοῦ ἱματίου αὐτοῦ
(having approached behind touched the fringe of the clothing of him)
- Luke 8:44a:
προσελθοῦσα ὄπισθεν ἥψατο τοῦ κρασπέδου τοῦ ἱματίου αὐτοῦ
(having approached behind touched the fringe of the clothing of him)
Interpolations and non-interpolations
Preliminary checks online showed that the reading in the extant manuscripts of Luke can either look like Matthew or like Mark. The Markan reading — without the fringes — is much less common. However, its existence causes us to wonder which is correct, and what are the arguments for preferring one over the other.
I especially wanted to see what Burnett Hillman Streeter and Bruce Manning Metzger had to say about the matter. Streeter’s 1924 book, The Four Gospels, contains an entire chapter dealing with the minor agreements in which he explains them in light of the Two-Source Hypothesis.
Here’s what Streeter had to say on the subject:
τοῦ κρασπέδου in Luke, om. in D a ff2 rl, a “Western non-interpolation.” (Streeter 1924, p. 313)
In layman’s terms he’s saying that the Greek words meaning “the fringe” found in Luke are omitted in manuscripts designated by the following sigla: D (Codex Bezae), a (Codex Vercellensis), ff2 (Codex Corbeiensis 2), r (Codex Usserianus), and l (Codex Rehdigeranus). And by “Western non-interpolation,” he’s referring to the theory posited by Fenton John Anthony Hort (of Westcott and Hort fame) that the Western text tradition frequently contains useful examples of non-assimilation. That is to say, it is often free of cross-contamination and scribal harmonization.
For Streeter, then, the interpolation of “the fringe” occurred when scribes, presumably remembering the text of Matthew, added it while copying Luke.
Here is how Luke 8:44a appears in Codex Bezae (see the online version, courtesy of Cambridge):
- Luke 8:44a (D):
προσελθουσα ηψατο του ϊματιου αυτου
(having approached touched the clothing of him)
- Mark 5:27:
ἐλθοῦσα ἐν τῷ ὄχλῳ ὄπισθεν ἥψατο τοῦ ἱματίου αὐτοῦ
(having come in the crowd behind touched the clothing of him)
The stylistic change from elthousa to proselthousa found in both Matthew and Luke doesn’t surprise us. Neither does the loss of the mention of the crowd. Luke, after all, starts this pericope with a reference to the throng surrounding Jesus.
Curiously, D omits the word opisthen (behind), which seems important to the narrative. We assume that Jesus didn’t see who touched him, not simply because of the crowd, but also because she came up from the rear. Does Streeter think that the original text in Luke omitted Mark’s “in the crowd behind”? The alternative explanation is that a later scribe looked at Luke’s text which did include “the crowd,” “behind,” and “fringe,” and removed all of it. But why?
What the committee decided
I had wrung about as much information out of Streeter as I could, and I turned to Metzger’s A Textual Commentary on the Greek New Testament. As you may recall, this book represents the work of an esteemed committee consisting of Kurt Aland, Matthew Black, Carlo M. Martini, Bruce M. Metzger, and Allen Wikgren. The committee looked at all variant readings, debated and discussed the issues, weighed the evidence, and then voted.
You may also recall that many mainstream NT scholars, on hearing that the Jesus Seminar followed a similar protocol — discussion followed by voting — doubled over with derisive laughter.
Surely the much-respected Committee would help me understand the evidence at hand and why nearly all modern translations keep the non-Western variant of Luke with the fringes. Guys?
The words constitute one of the so-called minor agreements of Matthew and Luke against Mark. The Committee [capitalized in the original] regarded this as accidental and decided to follow the overwhelming weight of the external evidence supporting the inclusion of the words. (Metzger 1971, pp. 145-146, emphasis mine)
And that’s it. Why do Matthew and Luke agree here? It was an accident. Oops.
What exactly does Metzger mean by “external evidence”? Well, when text critics discuss the evidence they divide the criteria into the external and internal variety. (See pp. xxiv-xxvii.) I’ll briefly summarize the kinds of external criteria based on the material on pp. xxv-xxvi.
A. The date and character of the witnesses. . .
B. The geographical distribution of the witnesses that support a variant. . .
C. The genealogical relationships of the text and families of witnesses. . .
D. Witnesses are to be weighed rather than counted. . .
In a nutshell, far too many of the extant, trustworthy texts we have of Luke’s gospel include the words found in Matthew, but not Mark. And why does Luke look like that? Happenstance — and by that they mean Luke’s edits just happened to coincide with Matthew’s.
I wondered, though, if the Committee had bothered to investigate the internal criteria surrounding this text, what would they have said? If we accept the notion that by fringes, Matthew meant us to picture Jesus wearing a prayer shawl, a tallit, then Matthew’s addition makes sense. Matthew presents Jesus as a great rabbi and a new Moses. But why did Luke add it? Was it in the Rich Oral Tradition™?
And if later scribes were uncomfortable with the “too-Jewish” nature of a prayer shawl, perhaps that’s why they dropped it from Luke. But notice that these supposedly embarrassed scribes never removed it from Matthew, only Luke.
On the other hand, if you believe Q is a mirage and Luke copied from Matthew (as “proved” by these minor agreements), then why would Luke follow Mark closely for several sentences, stop abruptly to pick up a few words from Matthew, and then return to Mark? Or was Luke “remembering” something he read in Matthew, something that popped into his head while redacting Mark’s “rough Greek”?
The problem of textual assimilation
The fact that this textual anomaly represents what Metzger called a “so-called minor agreement” helps explain scholarly focus on why the fringes appear in most witnesses rather than why they’re missing in some. But any convincing explanation certainly has to account for both.
I tend to think the original Gospel of Luke, contra Metzger, did not contain the fringes, and that we can account for the witnesses that do contain the minor agreement by appealing to internal evidence. Quoting Metzger:
Since scribes would frequently bring divergent passages into harmony with one another . . . that reading which involves verbal dissidence is usually to be preferred to one which is verbally concordant. (Metzger 2007, p. xxvii)
We must admit we can’t be certain, because of one nagging problem — namely, the issue of evidence cross-contamination. Either a scribe assimilated the text of Luke to be more in line with Mark or another scribe at some other time assimilated the text to match Matthew. But in either case, we have proof that scribes changed the text, and that fact alone is enough to plant the seed of doubt in the validity of the minor agreements argument. Always in the back of our minds, we have to wonder, “Or is it because some scribe was remembering what Matthew said here?”
We should remember, too, that although Mark early on joined the other three gospels in the canon, the early church greatly preferred Matthew. Mark was short, rough, and somewhat cryptic. Matthew’s gospel filled in the gaps; it contained the nativity, the Sermon on the Mount, and post-resurrection vignettes.
Consider this: What is the likelihood that scribes would think a “correct” reading of any scripture would be found in Mark? Textual assimilation had a natural, built-in bias away from a reading that resembles Mark and toward one that resembles Matthew.
What’s the point?
My point here is not to convince you one way or the other as to whether I or Metzger or Streeter or anyone else is correct. What I want you to notice is that we’re dealing with very deep, thorny problems. However, two towering figures in NT scholarship swept them aside — Streeter with one sentence, Metzger with two.
Nor do I wish to convince you that they weren’t experts or that they lacked knowledge or intellect. On the contrary, their extraordinary competence had blinded them to other possibilities. Their decades of study gave them an understanding of the transmission of the New Testament texts few others would ever attain. Similarly, their knowledge of the original languages — Greek, Latin, Syriac, etc. — had no parallel.
When Streeter concluded with certainty that “fringes” was not in the original text of Luke, but had been added later by a scribe, he was standing on his vast expertise. When Metzger concluded with certainty that “fringes” was in the original text, and that it accidentally conformed to Matthew at that point, he was also standing on his vast expertise.
Their very expertise created a kind of searchlight effect, in which what they were looking for came under intense illumination. This bright cone of light cutting through the darkness tends to reveal confirmatory evidence in sharp detail, while hiding potentially falsifying evidence outside the circle.
They both found what they were looking for.
You should not infer from what I wrote that this is some special problem in NT scholarship. It happens to all of us. As Nassim Taleb wrote in The Black Swan:
Even in testing a hypothesis, we tend to look for instances where the hypothesis proved true. Of course we can easily find confirmation; all we have to do is look, or have a researcher do it for us. I can find confirmation for just about anything, the way a skilled London cabbie can find traffic to increase the fare, even on a holiday. (Taleb 2010, p. 92)
Experts fall into the same trap as non-experts. The difference, unfortunately, is that their confidence in confirmatory evidence rises. The more we think we know, the greater our confirmation bias. It takes a great deal of discipline to look for evidence that proves we’re wrong. It takes even greater effort to conclude that we just don’t know the answer.
The false sense of surety
The market for books that conclude Jesus may or may not have existed — that we cannot know — must be vanishingly small, or perhaps publishers simply don’t want to deal with dull ambiguity. But I also suspect that most people with the skill, training, knowledge, and aptitude for writing books on the New Testament believe they’ve learned enough to “know” the correct answer. You can tell how sure they are by how they refer to the people on the other side of the fence. Good God, don’t they get it? Can’t they read?!
Getting back to our “fringe” issue, both Metzger and Streeter believed in the priority of Mark and the existence of Q. Whether they were conscious of it or not, those beliefs affected the way they looked at the evidence and how they evaluated what is important and what is not.
Quoting Taleb again:
This confirmation problem pervades our modern life, since most conflicts have at their root the following mental bias: when Arabs and Israelis watch news reports they see different stories in the same succession of events. Likewise, Democrats and Republicans look at different parts of the same data and never converge to the same opinions. Once your mind is inhabited with a certain view of the world, you will tend to only consider instances proving you to be right. Paradoxically, the more information you have, the more justified you will feel in your views. (Taleb 2010, p. 95)
In my previous post, I brought up the issue of context, and how knowing the mechanics of languages is not the same as knowing the culture and context that produced the New Testament. Unfortunately, the process of reading and rereading the texts reinforces the illusion that the text confirms our understanding of the text. Or, more to the point, our continual study of the New Testament consists of looking for information that confirms what we already believe to be true. This self-reinforcing process increases our surety of things we do not and cannot know with certainty.
Tim Widowfield
Latest posts by Tim Widowfield (see all)
- How Did We Get Here? Part 1 - 2024-11-21 01:22:29 GMT+0000
- What Did Marx Say Was the Cause of the American Civil War? (Part 1) - 2024-05-12 19:09:26 GMT+0000
- How Did Scholars View the Gospels During the “First Quest”? (Part 1) - 2024-01-04 00:17:10 GMT+0000
If you enjoyed this post, please consider donating to Vridar. Thanks!
There’s a fairly vocal group of NT scholars who advocate the position that Luke used Matthew as a source rather than Q. I tend to come down in the middle on this one: if Luke is to be believed that he made a thorough investigation, then it’s quite likely that he had a copy of Matthew to hand. I think he simply didn’t trust Matthew, for rather obvious reasons, and also had a copy of Q available as well as other sources.
This doesn’t, of course, answer the question of why he preferred Matthew’s reading here rather than Mark’s, but that’s essentially unknowable. If I was to guess, I’d say that he saw touching a ritual garment as having more power, or at least being more meaningful to his intended audience.
There’s also some arguing that Mt used some early version of “Luke” (eg, the Marcionite Gospel). I think the evolution of the gospels was far more complex than can adequately summed by simple source-theories, with the later canonicals being end-results of periods of mutual interaction. The Lukan prologue should at least be taken at its word that many gospels already existed.
Both Epiphanius and Tertullian discuss this passage in their copies of the gospel of Luke. Neither of them mentions the fringes, nor that the woman touched from behind, lending support that these features were absent in their versions of Luke.
The sources for the ‘fringes’ that Matthew (and some vesrions of Luke) mention possible originate in Mark 6:56 (NIV): “And wherever he went—into villages, towns or countryside—they placed the sick in the marketplaces. They begged him to let them touch even the edge of his cloak, and all who touched it were healed.” One could even argue that a scribe added the edge/fringes to his copy of Luke based on Mark 6:56, not Matthew 9:20b.
I appreciate the interest in what is, for the purposes of the post, a MacGuffin. Any thoughts about confirmation bias, the problem of experts, etc.?
I considered saying something, but originally decided not to. However…
As far as I’m concerned, there are too many scholars, all of whom are overanalyzing the same data, and who have a tacit gentleman’s agreement to not step on each other’s toes. In the example you gave, there is another possibility: the tassels may have originally been in Mark and got removed. Or Bob de Jong’s point – they may have been brought in from Mark 6:56.
The result is always going to be the scholar’s best guess, and that’s all it is: a guess even though it’s informed by a set of text critical rules.
By the time you add young turks who need something to write about to make a reputation, you’ve got a scholarly stew that will never yield a clear result. I’m not sure why anyone would expect it to.
I’ve never wanted to like a post here more than this.
I really find the whole field fascinating, but the academia on it is pure gaudy chrome which adds nothing of value. I prefer to look at it is a fictional genre where we are trying to imagine a story which would produce the results we see. Anyone who thinks they know what happened is stuffed.
The extent of Paul’s sources when he states “according to the scriptures” is indeterminate, nor is it definitively known how Paul may of been interpreting said sources per Pesher, Midrash, Second Temple period texts, Jewish mysticism, the Jewish angelic panoply, etc.
If a scholar “knows” that “according to the scriptures” = the Septuagint, they may look no further.
Trusting experts is something that we shouldn’t do blindly. They’re just as susceptible to bias as anyone else. There are a few studies that show this. What’s worse is that experts are obviously intelligent, so their bias comes off as more sophisticated (and thus seems more correct) than a layman’s.
On the other hand, immediately distrusting experts can lead to similar or even worse problems.
I think that not all experts are created equally. It depends on the field in question. I’m a lot more trusting of experts in STEM fields than I am in humanities. This is because in STEM, a scholar’s personal politics or religious beliefs can rarely be used to predict the outcome of their research, while the opposite is generally true for the humanities.
In my albeit limited experience, in the STEM world you often have the joy of instant feedback when you’re wrong. It’s humbling, even though you realize you’re making progress. We learn by false steps.
Ehrman accepting one of Price’s key points and then shrugging it off.
Bart Ehrman (20 June 2017). “Was Jesus Made Up? A Blast from the Past.”. The Bart Ehrman Blog. Retrieved 13 October 2017. “[Per Ehrman] In browsing through some old posts, I came across this one from five years ago…”:
Ha, I learned another word, a “MacGuffin”! I understand your concern that we may miss the point of your post; I didn’t, but just agree with it. People make – wrong – decisions based on confirmation bias all the time, not just in Bible studies.
As Thucydides (5th C BC) said: “for it is a habit of mankind to entrust to careless hope what they long for, and to use sovereign reason to thrust aside what they do not fancy”
I would recommend the following book:
Burkett, Delbert. Rethinking the Gospel Sources: From Proto-Mark to Mark. T & T Clark International, 2004.
It casts significant doubt on both the two-source hypothesis that Streeter and Metzger supported and the Farrer–Goulder–Goodacre hypothesis.
There are so many elements that argue against the Q hypothesis it’s hard to know where to start.
Just for starters I’d recommend Perrin and Goodacre’s “Questioning Q” with particular note of the chapters by Perrin [in which he shows the historical development of the hypothesis in the 1800s and how politically convenient it was – a theme touched on in the last chapter by Goodacre] and Olson [who demonstrates how mechanically improbable the hypothesis is by deconstructing Streeter and Downing in contrast to Farrer]. Olson’s article is available here:https://drum.lib.umd.edu/handle/1903/1378
I like Goulder’s use of doubled animal imagery to shows/suggests the dependency between “Luke” and “Matthew” which is part of his article “Is Q a juggernaut”.
But, for now, I’ll just chuck in this thought.
We don’t know who “Luke” was, or where and when he wrote.
Ditto for “Matthew”.
And ditto for “Mark,” which is relevant if we go with the Q hypothesis sub set that postulates that “Mark” also had access to Q.
So we are asked to believe that a written document that was so important that it formed the basis for a chunk of each of the synoptic writers over whatever the geographic region involved was and the span of time, decades, involved, left otherwise no trace whatsoever.
Hmmm.
No trace whatever? I presume that you think the Gospel of Thomas is irrelevant.
Yes.
GoT is relevant to the idea that *something like* a Q source (a collection of sayings attributed to Jesus) *might have* existed prior to narrative gospels. But its not identified *as* Q by any scholar I know of.
Apart from 3 fragments in Greek c.150AD from Oxyrhynchos, GoT is a fourth century Coptic document. We can see comparing the Greek fragments with the same in the Coptic that between the two there has been considerable change to the text. Posit a similar rate of change and similar contraction and expansion occurring with a c.50AD date for the ur-text and I think the difficulties should be obvious; never mind basing argument on such a small sample of text.
“Basing an argument on such a small sample of text” is ill-advised, especially since we know there was so much more that was lost.
The gospels were all written in a time in shadow. Each of the gospels contains lots of material that was made up out of thin air, the original authors knew they were doing that. I imagine they were even commissioned to rather than some bishops or clerics doing it themselves and mucking it up much worse, and certainly not some Baldrick forgers.
Many people would have been hungry for stories about Jesus given that Christianity must have had very thin substance before the gospels, but what we see today suggests that each new gospel would have been treated with substantial skepticism and circulation would be completely closed off from many people. This would have been true of epistles as well, especially the known forgeries. If a book came out today stated to have been written by Charles Dickens a lot of people would be rightly skeptical, meanwhile works of Mark Twain and Tolkien are still coming out. Faith and the gullibility associated with it are almost always shielded by fundamentalists clinging to strict interpretation. In my experience, every translation of Scripture is beaten to death by fundamentalists whose only standard is their own interpretation. It isn’t reasonable to think a new letter or book coming out is going to be accepted as gospel from the get go no matter who backs it up.
Successive gospels were most certainly written in response to the one(s) that came before them, but copyists may not have even had access to the other gospels because of the restrictions against heretical documents. By the earliest time we are aware someone had already decided that all the gospels were gospel and that’s the point when scribes could make changes based on alternate gospels. What we want to know is technically what happened in the shadow time. This is layers of intrigue we will never know beyond the mere technical details left behind.
This also strikes me the way Trekkies are whining about Star Trek Discovery. If you can’t change Star Trek, what makes anyone think they could change Christianity.
Just turned a fellow Trekkie onto that yesterday, and he still hates the newer movies; everyone has their own “canon”..
“But in either case, we have proof that scribes changed the text, and that fact alone is enough to plant the seed ”
This is why I’m also skeptical of passages that are taken so seriously like “James the brother of Jesus “. We don’t have a fricken clue what it originally said, we know the texts were monkied with, and here we are arguing over the nuances of some vague passage when the actual physical evidence on the ground is zip. Frustrating.
James the brother of the LORD. If it said ‘Jesus’, thre would be no argument. We might not know what the original text aid but they are likeley to have been more contradictory of one another than less. We can see this with the texts we actually have tending to harmonise over time.
As you note, two versions in Luke mean some scribe messed with it at some point. This seems purely a transcription issue.
Each evangelist obviously had his own reasons for retaining or omitting words & phrases. But the error committed is bringing in one’s a priori assumptions about the those motives (attempting to mind-read), then accepting or dismissing evidence to conform to one’s assumptions.
Stepping back from all that for a moment, for an antiseptic look at the additions & deletions themselves, might prove illuminating:
Mark: [ἐλθοῦσα] [ἐν τῷ ὄχλῳ] [ὄπισθεν] [ἥψατο] [τοῦ ἱματίου αὐτοῦ]
A1 B C D E
Matt: [προσελθοῦσα] [ὄπισθεν] [ἥψατο] [τοῦ κρασπέδου] [τοῦ ἱματίου αὐτοῦ]
A2 C D X E
Luke: [προσελθοῦσα] [ὄπισθεν] [ἥψατο] [τοῦ κρασπέδου] [τοῦ ἱματίου αὐτοῦ]
A2 C D X E
Luke(D):[προσελθουσα] [ηψατο] [του ϊματιου αυτου]
A2 D E
===
(1) A1 B C D _ E
(2) A2 _ C D X E
(3a) A2 _ C D X E
(3b) A2 _ _ D _ E
You can then play around with these, rearrange the order, to see what makes most sense.