Some Interesting "Minor Agreements"

Usually there is no overlap between the  field of textual (lower, or post-production) criticism and higher (pre-production) criticism.  Usually.  But overlap does sometimes occur.  Today, we shall briefly investigate the Synoptic Problem, and look into a few “Minor Agreements” (i.e., readings shared by Matthew and Luke, but not by Mark) and consider the possible/probable implications of these readings for the production of the Gospel of Mark.

            To those who are entirely new to the Synoptic Problem, the first thing to know is that the term “Problem,” in this context, simply means “a puzzle to ponder,” not something that is a troublesome difficulty that threatens to undermine the Christian faith.  (Similarly, the “criticism” the terms “textual criticism” and “higher criticism” simply means “analysis;” these fields are not platforms for promoting personal critiques of the contents of the books of the New Testament, or any other book.

            The second thing to know is that the “Synoptic Problem” orbits the answer to one question:   how do the Synoptic Gospels (Matthew, Mark, and Luke) relate to one another?  Did each author write entirely independently of the other two?  Did two of them use the other?  Did one of them make use of the other two?  Or did all three use a shared source?    

            The third thing to know is that the Synoptic Problem has been solved for the most part:  although it was consistently held in early Christianity that Matthew wrote first, researchers such as William Sanday, B.H. Streeter and John Hawkins have made a very strong case that the Gospel of Mark (or something that resembled the Gospel of Mark) was used by Matthew and Luke, and that Matthew and Luke both used a second source (known as Q, which stands for Quelle, the German word for “Source”), and that Matthew and Luke each made use of source-materials that were accessed only by Matthew, or only by Luke.  

            The Four-source Hypothesis – that Matthew used Q + Mark + extra source-material, and Luke used Q + Mark + extra source-material (for a total of four sources of material) has a lot going for it, as Daniel Wallace concisely explains here and as Dennis Bratcher explains not so concisely here.  

            The ship of the simple Four-Source Hypothesis, seaworthy as it may seem, cannot survive the reefs it faces in the “Minor Agreements” – readings shared by Matthew and Luke but not shared by Mark – especially in cases where the “Minor Agreements” occur smack-dab in to middle of the triple-tradition (i.e., material shared by Matthew, Mark, and Luke).   Something more complex has happened.

           The diagram I have made here offers a simple picture of how to account for the evidence, including the “Minor Agreements.”  I propose that neither Matthew nor Luke used the Gospel of Mark in its full canonical form; rather; Luke used an early form, and Matthew used a later form.  Mark, serving as Peter’s amanuensis, did not create one definitive text of the Gospel of Mark right away:  as Peter continued to preach and teach about Jesus, his testimony – written down in the Gospel of Mark – continued to expand.   Not until Peter’s martyrdom did a definitive text of the Gospel of Mark come into being.

            In the course of creating the definitive text of the Gospel of Mark, Mark added and subtracted a variety of details that had been in earlier forms of his record of Peter’s testimony, and this resulted in “Minor Agreements.” We shall now zoom in on some of these points which are like chords in a song which is sung by all three Synoptic writers, where Matthew and Luke sing in harmony but Mark sings a different note all by himself.

            First, let’s look at the chord that occurs in Matthew 9:18, Luke 8:40, and Mark 5:21.  The scene depicted by all three Evangelists is the famous opening of the account of the raising of Jairus’ daughter.  The first interesting feature is that when we use the Western text (extant in D 05), only in the Gospel of Mark (5:22) do we find Jairus identified by name.  This detail, if it had been known to Luke, would not be something he would have omitted.  It was probably added by Mark in the course of preparing the final form of his Gospel-account.     

            The second interesting feature in that in Matthew 9:20 and  Luke 8:44, there is an explicit reference to the hem of Jesus’ garment; meanwhile Mark 5:27 does not.  The parallel and non-parallel is easily shown in Greek: 

            Matthew:  ἥψατο τοῦ κρασπέδου τοῦ ἱματίου αὐτοῦ

            Mark:  ἥψατο τοῦ ἱματίου αὐτοῦ

            Luke:  ἥψατο τοῦ κρασπέδου τοῦ ἱματίου αὐτοῦ

It looks as if Matthew and Luke both perpetuated a Markan text which had τοῦ κρασπέδου after ἥψατο, but in the final form of the Gospel of Mark, τοῦ κρασπέδου had fallen out of the text, perhaps via simple parablepsis.  Again, this implies that a distinction must be maintained between the Markan texts (“Ur-Mark” or “Proto-Mark”) used by Matthew and by Luke, and the final definitive form of the Gospel of Mark.  Again:  whatever Markan texts were used by Matthew, and by Luke, were not the same as the final form of the Gospel of Mark.   At least two extra steps were yet to be taken before the Gospel of Mark was finished:  Jairus’ name was added, and τοῦ κρασπέδου was omitted.

             Throughout the text of Mark, isolated “Minor Agreements” occur at points where it appears that Mark added or modified small details that were not in Ur-Mark.  Some examples:  (1) the detail in Mark 1:13 that Jesus was tempted by Satan, σατανᾶ, not by “the devil” (διαβόλου),

(2) the detail in Mark 1:39 that Jesus cast out demons (τά δαιμόνια ἐκβάλλων),

(3) the detail in Mark 2:2 that Jesus preached the word,

(4) the detail in Mark 1:45 that Jesus gave strict instructions to the leper he had cleansed

(5) the detail in Mark 2:27 that Jesus said, “The Sabbath was made for man, and not man for the Sabbath”

(6) The detail in Mark 3:5 that Jesus looked around with anger

(7) the detail in Mark 3:20-32 that Jesus’ mother and brothers came to him because they thought he was out of his mind

(8)  the detail in Mk 9:32 that Jesus spoke openly the word concerning his sufferings, death, and resurrection

(9) the details in Mark 9:21-27 about Jesus’ questions to the father of the young man with an unclean spirit,

(10) the detail in Mark 10:50 that Bartimaeus cast away his garment as soon as he heard that Jesus was calling for him. 

More examples could be given, but these ten should sufficiently show that there is a difference between Ur-Mark used by Luke, Ur-Mark used by Matthew, and the final form of the Gospel of Mark.

            This has an effect on another issue: the ending of Mark.  Stephen Boyce recently chimed in about this.  Several of the unique details in the Gospel of Mark are the sort of thing an eyewitness could add as expansions of an episode he had described previously; but they are not the sort of thing a non-eyewitness would throw in arbitrarily.  (Jairus’ name, for example, might not have been known by Peter when he composed Ur-Mark, but Peter and/or Mark may have discovered it prior to the composition of the Gospel of Mark.)  While nothing about this brings anything new against the idea that Mark 16:9-20 was composed by Mark as a freestanding text, and was then attached to 16:8 by Mark’s colleagues in Rome, it must also be granted that nothing stands in the way of a somewhat simpler solution:  that Mark, on the same occasion when he tidied up Ur-Mark and thus produced the Gospel of Mark, composed verses 9-20, and the narrative disconnect in vv. 9-10 is simply an effect of leaving the narrative thread dangling, so to speak, on an earlier occasion, or else it is an effect of replacing a now-lost ending of Ur-Mark with a fuller summary of the post-resurrection appearances of the risen Lord Jesus Christ.

            All three possibilities lead to an embrace of Mark 16:9-20 as part of the canonical text of the Gospel of Mark.

 

 

Conflations (Part 1)

Codex Macedonianus -
the end of Luke

            What is a conflation?  A conflation is the effect of a textual collision.  Suppose a copyist sat down to make a copy of the Gospel of Luke, and checked his primary exemplar by consulting a secondary exemplar.  Suppose, further, that in Luke 24:53, his primary exemplar said that the disciples “were continually in the temple, praising God,” and that his secondary exemplar said that the disciples “were continually in the temple, blessing God.”  The copyist might decide to follow one exemplar and not the other one – or he might combine them, creating a new reading:  the disciples “were continually in the temple, praising and blessing God.”  Such a combination is a conflation.

            In 1881, Hort argued that the Byzantine Text as a whole is a secondary text – a combination of readings harvested from earlier Alexandrian and Western texts.  One of Hort’s key points was that the Byzantine Text contained conflations, and he listed eight of them:  in Mark 6:33, Mark 8:26, Mark 9:38, Mark 9:49, Luke 9:10, Luke 11:54, Luke 12:18, and Luke 24:53.  In each passage, the Alexandrian Text has a short reading, the Western Text has a short reading, and the Byzantine Text has a longer reading which, Hort argued, is a combination of the Alexandrian and Western readings.

            Hort then considered the writings of Irenaeus, Hippolytus, Clement of Alexandria, Origen, Tertullian, Cyprian, Methodius, and Eusebius of Caesarea, and concluded that they contain no distinctly Byzantine variants:  “Before the middle of the third century, at the very earliest, we have no historical signs of the existence of readings, conflate or other, that are marked as distinctively Syrian.”  Hort concluded from these two points – (1) Eight readings in the Syrian text appear to be combinations of readings in the Alexandrian and Western texts, and (2) There are no distinctly Syrian readings in the writings of Irenaeus, Hippolytus, Clement of Alexandria, Origen, Tertullian, Cyprian, Methodius, and Eusebius of Caesarea – that the Syrian Text should be regarded “as not only partly but wholly derived from the other known ancient texts.”  And he further extrapolated that “all readings in which the Pre-Syrian texts concur must be accepted at once as the apostolic readings, or to speak more exactly, as the most original of recorded readings.”  In other words, readings shared by the Alexandrian and Western Texts should be accepted as a matter of course no matter what the Syrian Text says.

            Hort’s proposal essentially rendered the Syrian (Byzantine) Text superfluous – and, by extension, a vast number of New Testament manuscripts which support the Syriac Text were regarded as unimportant in the enterprise of reconstructing the original text of the New Testament:  only representatives of the Alexandrian and Western Text really mattered, and in 1881 less than 50 such Greek manuscripts were known.  Hort relied on two of them above all others:  Codex Vaticanus (B) and Codex Sinaiticus (À).  Hort proposed “(1) that readings of ÀB should be accepted as the true readings until strong internal evidence is found to the contrary, and (2) that no readings of ÀB can safely be rejected absolutely.”

            Thus, for the establishment of the base-text of the Revised Version, the testimony of thousands of Greek manuscripts was set aside, mainly in favor of a very small number of manuscripts that represented the Alexandrian Text, especially the Alexandrian Text as displayed in Vaticanus and Sinaiticus.

            With all that in the background, let’s return to the subject of conflation. 

            Although Hort’s research made an enormous impact, not everyone was persuaded.  In the 1897 Oxford Debate on New Testament Textual Criticism, Edward Miller considered it absurd to conclude that that the Syrian (Byzantine) Text is altogether secondary because of merely eight readings, and he challenged his fellow debater Dr. Sanday to produce 30 conflate readings from the Syrian Text.  In reply, Dr. Sanday did not produce 30 conflate readings, and conceded the point; his reply should be famous:  “Whatever person or whatever school produced the Traditional Text, did not systematically combine the Texts.  They were combined occasionally, and that is all one can say.  I am prepared to admit for myself that the conflations are not conclusive proof of the rightness of Dr. Hort’s theory; they could only belong to the region of hypothesis.  It is all hypothesis.”
            Nevertheless, Hort’s eight conflations are still treated as if they show that the Byzantine Text is derivative of the Alexandrian and Western Texts.  For example, on page 45 of Interpreting the New Testament Text (2006, Bock and Fanning, editors), Daniel Wallace stated the following:  “Hort argued that the Byzantine text (what he called the Syrian text) was inferior.  His arguments are still essentially valid today:  (1) conflations (i.e., a new reading combined from two earlier readings) show that the Byzantine text is secondary, because the Byzantine Text is the only text-type to conflate (cf. Luke 9:10, 24:53); (2) no ante-Nince fathers seem to quote distinctive Byzantine readings, demonstrating that the Byzantine text is late; (3) internal evidence reveals that the Byzantine text is inferior.”

            Two things should be emphasized here:  First, in 2006, conflations were presented as the number one piece of evidence for the secondary nature of the Byzantine Text.  Second, in 2006, Hort’s transmission-model was not presented as some antique theory of yesteryear, but as if it is a theory which still merits adherence – and this presentation did not come from some fringe element, but from a professor at a leading American evangelical seminary.  I conclude that those who claim that the New Testament base-text of the NIV, ESV, NLT, and CSB is not (at over 95% of its points of disagreement from the Byzantine Text) Hort’s compilation are in a fantasy-land.

            Now let’s zoom in on the two passages mentioned by Dr. Wallace:  Luke 9:10 and 24:53.  I will list the major variants in each passage, followed by some analysis of each variant-unit.  

Luke 9:10

            P75 B L 33:  εις πόλιν καλουμένην Βηθσαϊδα (to a city called Bethsaida) [P75 reads Βηδθσαιδα; L reads Βιθσαϊδαν]

            À 157:  εις τόπον ερημον (to a remote place)

            D:  εις κωμην καλουμένην Βηδθσαϊδα (to a village called Bedthsaida)

            Θ:  εις κωμην καλουμένην Βηδθσαϊδα εις τόπον ερημον (to a village called Bethsaida, to a remote place)

            A:  εις ερημον τόπον πόλεως λεγομένην Βηδσαϊδα (to a remote place of the city called Bethsaida)

            f¹:  εις τόπον πόλεως καλουμένης Βηθσαϊδα (to a place of the city called Bethsaida)

            Byz W:  εις τόπον ερημον πόλεως λεγομένην Βηδσαϊδα (to a remote place of the city called Bethsaida)

            C E F G M Π 565 f¹³ 1424:  εις τόπον ερημον πόλεως καλουμένης Βηθσαϊδα (to a remote place of the city called Bethsaida)

            K N:  εις τόπον ερημον πόλεως καλουμένην Βιθσαϊδαν (to a remote place of the city called Bithsaida)

Those who attempt to produce the reading found in most manuscripts (εις τόπον ερημον πόλεως λεγομένην Βηδσαϊδα) from the Alexandrian reading εις πόλιν καλουμένην Βηθσαϊδα and the Western reading εις κωμην καλουμένην Βηδθσαϊδα will soon find themselves frustrated, for neither one says anything about a deserted place.   But Hort’s proposal did not involve such a conflation.  Instead, Hort saw Sinaiticus’ reading as a truncated form of a Western reading (attested in Old Latin copies):  εις τόπον ερημον Βηδσαϊδα (to a deserted place, Bethsaida) or εις τόπον ερημον καλουμένον Βηδθσαϊδα (to a deserted place called Bethsaida).

However, there are simpler explanations for the Byzantine reading.  For example, a copyist wishing to harmonize the text of Luke here to the text of Matthew 14:13 (where, immediately before the feeding of the 5,000, Jesus departs εις ερημον τόπον) or Mark 6:32 (where, immediately before the feeding of the 5,000, Jesus instructs His disciples to go with Him εις ερημον τόπον) would not need a secondary exemplar to introduce εις ερημον τόπον into the text of Luke 9:10.  He would only need the parallel-passages in Matthew and Mark.

Another possibility is that the original text is preserved in C E F G M N Π 565 f¹³ 1424, and that this reading explains each of its rivals, along the following lines:

B’s reading is a simplification, elicited by a scribe’s sense that a single place cannot be both remote  (or deserted, or wilderness) and belong to a city. 

D’s reading is the same simplification, with Bethsaida downsized to a village.

            À’s reading is a harmonization, replacing Luke’s verbiage with verbiage from the parallel-passage in Matthew 14:13 or Mark 6:32.

            Θ’s reading is D’s reading with εις τόπον ερημον inserted from Mt. 14:13 or Mk. 6:32.

            A’s reading is the same as the usual Byzantine reading, with a minor transposition.

            f¹’s reading is the reading of CEFG etc., except ερημον is absent, either due to parableptic error or due to a scribe’s sense that a remote/deserted place cannot be said to belong to a city.

            Byz’s reading is the reading of CEFG etc., with the word λεγομένην taking the place of its synonym (used more frequently by Luke, and supported across multiple transmission-lines) καλουμένης.

            K and N’s reading is the reading of CEFG etc., slightly tweaked.

           

Luke 24:53    

            P75 À  B C* L and the Sinaitic Syriac, Palestinian Aramaic, Coptic, and Georgian support ευλογουντες (blessing)

            D and several Old Latin copies support αινουντες (praising).

            Byz and A C^c K M U W Δ Θ Λ Π Ψ f¹ f¹³ 33 157 565 579 1424 etc. and the Vulgate, the Peshitta, and the Armenian version support αινουντες και ευλογουντες (praising and blessing). 

            Hort claimed that this case “needs no explanation.”  Here, it is claimed, we face a simple and uncomplicated blend:  using αινουντες τον Θν in a Western exemplar and using ευλογουντες τον Θν in an Alexandrian exemplar, an editor created the reading αινουντες και ευλογουντες τον Θν.  However, what if the Alexandrian reading exists because an Alexandrian scribe considered it superfluous to say that the disciples both blessed and praised God?  And what if the Western reading exists because a Western scribe lost his place in the text and his line of sight drifted from the final letters of αινουντες to the final letters of ευλογουντες, accidentally skipping all the letters in between?  One might say that a conflation is simpler – if one finds it simple to posit an editor with two exemplars who would combine their readings here, after refraining from making any other conflations involving the many distinct Western readings elsewhere in Luke 24 – but neither theory is unfeasible.   

            Losses of text (whether deliberate or accidental) in both the Alexandrian and Western transmission-lines can give the impression that the Byzantine reading is a combination of the other two, but this impression is not necessarily the last word.  Consider Mark 6:33 – where one finds the proposed conflation upon which Hort focused the most in his 1881 Introduction.

Mark 6:33b

            À B 0187 892 and the Vulgate and some lectionaries:  και προηλθον αυτους (and they arrived first)

            L 579 1241 and the Armenian version and some lectionaries:  και προσηλθον αυτους (and they arrived first)

            Δ Θ:  και προσηλθον αυτοις (and arrived ahead of them)

            f¹:  ηλθον εκει (they came there) [without the εκει after συνέδραμον]

            f¹³:  και προηλθεν αυτους (and they arrived first)

            D:  και συνηλθον αυτου (and came together to Him)

            33:  προς αυτους.  και συνηλθον προς αυτον (before them, and gathered together to Him)

            700:  και συνηλθον αυτω (and came together to Him)

            565:  και ηλθον αυτου (and came to Him)

            Byz P84^vid E F G H K M U V Γ Π 157:  και προηλθον αυτους και συνηλθον προς αυτον (and they arrived first, and came together to Him)

            W:  verse 33 ends at συνέδραμον εκει          

            A:  και προηλθον αυτους και συνέδραμον προς αυτον (and they arrived first, and ran together to Him) [repeating a verb that occurs earlier in the verse]

            The next word in the text (in all transmission-lines) is και.

            As Willker noted, the scholar Bousset wrote about this textual contest in 1894:  “If we accept the long reading as original, then the short readings are quite easy to understand:  συνηλθον αυτου was omitted [accidentally] and D et al omitted [deliberately] the difficult προηλθον αυτους.” Is this feasible?  Let’s see: 

The loss of προηλθον αυτους is fully capable of being lost accidentally, via an accidental skip from και to και :

1.  Text before parablepsis:  και προηλθον αυτους και συνηλθον προς αυτον και

            2.  Parablepsis, και¹ to και²

            3.  Text after parablepsis:  και συνηλθον προς αυτον

            Thus the reading in D is mostly accounted for by the Byzantine reading; all that is needed at this point is for a Western scribe to replace προς αυτον with αυτου.  Meanwhile, how does one get to και συνηλθον αυτου if all one has to start with is και προηλθον αυτους?

           

            The loss of συνηλθον προς αυτον is accounted for as follows:

            1.  Text before parablepsis:  και προηλθον αυτους και συνηλθον προς αυτον και

            2.  Parablepsis, και² to και³

            3.  Text after parablepsis:  και προηλθον αυτους

            Thus the reading in B et al is accounted for by the Byzantine reading.

            So far, three “conflations” have been examined, and none of the three prove the posteriority of the Byzantine reading, let alone the posteriority of the entire Byzantine Text.  But even if all eight of Hort’s proposed conflations were airtight, this would not demonstrate the posteriority of the Byzantine Text, for a very simple reason.  God willing, I will explore that reason in Part 2.

Readers are invited to double-check the data in this post.