John 7:46 - Neither Shortest Nor Longest

In his obsolete Textual Commentary of the Greek New Testament, regarding the end of John 7:46 Bruce Metzger briefly stated, "The crisp brevity of the reading supported by p^{66c, 75} B L T W copt^bo al was expanded for the sake of greater explicitness in various ways, none of which, if original, would account for the rise of the others."  

Let's test that.

Following ἄνθρωπος, we see the following variety: 

ὡς οὗτος ὁ ἄνθρωπος - Byz K M N U Γ Δ Θ Λ Π Ψ  f¹ f¹³ 2 28 33 69 124 157 565 579 1071 1424 1505

ὡς οὗτος λαλει ὁ ἄνθρωπος  - P66* 01* (There is an itacism in 01 and P66*, and 01 has a singular reading at the beginning of the verse, pictured.)

ὡς οὗτος λαλει (after ἄνθρωπος ἐλάλησεν) - 05

That's not a lot of variety.   03 P66^c 019 T and 032 appear to be the only manuscripts which support the reading adopted in UBS4.

Meanwhile, support for a longer reading comes not only from all other Greek manuscripts (with GA 13 dissenting due to a scribal error, initially failing to include ἐλάλησεν earlier in the verse, and with a transposition - ἐλάλησεν οὗτως - in N Ψ 33 1071 1241) but also from the Sahidic, Armenian, Ethiopic, Georgian, and Syriac (Sinaitic, Curetonian, Peshitta, and Harklean) versions, as well as the Palestinian Aramaic and the Vulgate.  A very impressive array.

While some commentators point out that the Byzantine text displays a tendency to clarify via embellishment, one should also be aware of the opposite tendency in the Alexandrian text to economize via abbreviation - i.e., to attempt to express the same idea using fewer words.  

If one were to treat the reading supported by the vast majority of manuscripts and versions as original here, the reading of 03 and allies is readily explained as either the result of a parableptic leap from the first ἄνθρωπος to the second ἄνθρωπος., or as an intentional attempt to eliminate superfluity.  

An early scribe could conceivably consider the Alexandrian reading in need of embellishment, and add "like this" or "like this man."  On the other hand, the addition of "like this" and "like this man" adds nothing that anyone could not figure out in a moment.  If John wrote ὡς οὗτος ὁ ἄνθρωπος, his reason for doing so would be obvious:  that is what he overheard the soldiers say.  In addition, the reading in P66* and 01 is accounted for as a conflation of the Byzantine reading and the reading in 05.

Instead of defending the Alexandrian reading by assigning to scribes a desire to make a frivolous embellishment, it is better in this case to regard the reading of 03 and allies as an accidental or intentional truncation of what John wrote.

One medieval scribe - the copyist of 2483(2866) - illustrated that a scribe in the Middle Ages could commit dittography while copying John 7:46-47.  And where dittography is possible, parablepsis tends to be possible too.

For those who may be interesting in how English versions treat this variant:  KJV NKJV MEV RSV Message NASB95 NET NIV EHV EOB all support the longer reading, demolishing any  assumption that those who reject Metzger’s premise here must harbor a pro-Byzantine prejudice.    

(Thanks to Ben Crawford for sharing this photo of GA 2483 from the Benjamin Crawford Collection, Alabama.)

0    

John 6:15 - An Interesting Little Detail

           Near the end of John 6:15 – following John’s account of the feeding of the 5,000 – there is an interesting variant-unit:  did John write that Jesus withdrew again (ἀνεχώρησεν πάλιν) to the mountain by himself, or that Jesus withdrew (ἀνεχώρησεν) (without “again”) to the mountain by himself, or that Jesus “fled again” φεύγει πάλιν) to the mountain by himself?

          John does not use ἀνεχώρησεν anywhere else, so perhaps there is something to be said in favor of φεύγει:  the idea is that copyists changed the wording of Jesus’ movement to avoid giving the impression that Jesus was running away like Brave Sir Robin.  On the other hand, John does not use φεύγει anywhere else, either. 

          Constantine von Tischendorf was willing to let Codex Sinaiticus overrule all other Greek manuscripts in existence in John 6:15, in his eighth edition of the Greek New Testament.  Scholars less entranced with À have not concurred, arguing (as Metzger reports in his Textual Commentary) that φεύγει originated as a Western reading.  It had to be introduced as a sort of paraphrase fairly early, inasmuch as it is supported by Tertullian, De Rebaptismate, Ambrose, Jerome (who placed fugit in the Vulgate, agreeing with several Old Latin copies) and Augustine.  (Tertullian’s reference is not airtight, however:  in On Idolatry XVIII, he simply says that Christ “shrank back from being made a king,” without explicitly quoting John 6:15.)     

          Ἀνεχώρησεν is read in all text-types:  Alexandrian (P75, B, L), Western (D), Byzantine (A, K, Λ, Π, N) and Caesarean (f¹).   But did John write only ἀνεχώρησεν or ἀνεχώρησεν πάλιν?  Manuscripts without πάλιν here include W, Δ, Ψ, and most Byzantine copies.  The Peshitta, likewise, does not support πάλιν.  The word could perhaps have been skipped accidentally, if a copyist’s line of sight slid from the ν at the end of ἀνεχώρησεν to the ν at the end of πάλιν.  Or, having just read John 6:3’s account of Jesus going up into the mountain, early copyists couldn’t resist harmlessly adding πάλιν – but one would have to believe this happened at least twice for πάλιν to drift into A, D, Π, N, and Y.    

          The Curetonian Syriac, interestingly, supports a conflate reading:  fled and withdrew again.  

          Readers of the Nestle-Aland Novum Testamentum Graece compilation would never detect that the reading found in the Byzantine Text, W, and the Peshitta exists:  the N-A apparatus mentions the reading in À but not the reading of most manuscripts – an omission that is difficult to explain except as an effect of the editors’ pro-Alexandrian bias.  The UBS apparatus doesn’t mention the non-inclusion of πάλιν either; nor does they Tyndale House GNT’s apparatus.    

          The non-original reading φεύγει is more interesting than it looks – not only because it was temporarily favored by a leading textual critic in the 1800s, but because it reveals that retro-translation sporadically affected the Greek text in the exemplar of the opening chapters of John in Sinaiticus – the same sort of Western phenomenon that has unfortunately affected the NIV 2011’s base-text of Mark 1:41.

 

Glossary of Textual Criticism: A-C

“Understandest thou what thou readest?”

That was Philip’s question to the Ethiopian eunuch in Acts 8:30, as rendered in the KJV.  Every field of scientific study involves some specialized terms, or jargon, which might initially be difficult to understand, and New Testament textual criticism is no exception.  It is easier when you know the jargon.  Gary Dykes has made a very thorough dictionary of terms that pertain to textual criticism and manuscript-production that can be read at http://www.biblical-data.org/TC_stuf/definitions.html .  (You might have to cut-and-past the address in your browser.)  The website of the Hill Museum & Manuscript Library offers a helpful multi-part review of terminology relevant to the study of Latin manuscripts.  The British Library’s online glossary of terms used in its descriptions of illuminated manuscripts is also informative.  Robert Waltz’s Encyclopedia of New Testament Textual Criticism includes a very thorough review of the terminology used in this field.  Here is the first of several parts of a concise introductory list of some of the technical terms used in New Testament textual criticism, with their definitions.   

Alexandrian Text:  The form of New Testament text which was dominant in Egypt in the early church, displayed most accurately by Codex Vaticanus and the early Sahidic version.  Since papyrus tends to naturally rot away except in low-humidity climates such as the climate of Egypt, almost all surviving papyrus manuscripts – especially the ones found as the result of excavations in or near Oxyrhynchus, Egypt – support the Alexandrian Text.  Where Alexandrian readings deviate from the Byzantine Text, the Alexandrian reading frequently has internal characteristics that commend it as original.  In some cases, however, Alexandrian variants can be plausibly attrributed to scribal carelessness and conscious editing.  The Nestle-Aland compilation of the Greek New Testament, the primary basis for most modern English versions (the ESV, CSB, NIV, NLT, etc.) is mainly based on the Alexandrian Text.    

 

Yellow arcs (added) accompany
Section-numbers and
 Canon-numbers  in the margin
of this Gospels-manuscript.

Ammonian Sections:  The segments into which the text of the Gospels was divided for identification in the cross-reference system developed by Eusebius of Caesarea.  There are 355 sections in Matthew, 234 in Mark, 343 in Luke, and 232 in John – at least, these are typical.  This system of text-segmentation is named after Ammonius of Alexandria, who, according to Eusebius in his letter Ad Carpianus (which often precedes the Canon-Tables), developed a cross-referencing method in which the text of Matthew was supplemented by the parallel-passages, or the numbers of parallel-passages, in the other Gospels.  It was Eusebius, however, who developed the Sections as we know them, for they cover passages in Mark, Luke, and John that are not paralleled in Matthew.

            In very many Gospels-manuscripts, the Section-numbers appear in the margin alongside the text, accompanied by the canon-number (written below it, separated by a horizontal line).  The numerals are typically written in red.  It is not unusual to see that in the text itself, the first letter on the first line after the beginning of a Section is given special treatment – either by being written larger, or by being written in different ink (often red) slightly to the left of the left margin, or both.    

Bifolio:  A sheet of writing-material (whether parchment, or papyrus, or paper), vertically folded in the middle so as to form four pages upon which text could be written.  Typically, groups of four bifolium were combined – picture a stack of four flat sheets; then picture them vertically folded, all at once, so as to form a small blank 16-page book.  Such a 16-page book is called a quire, or quaternion.  (Quires could take other forms – consisting of different numbers of sheets – and could be supplemented and repaired in a variety of ways.) 

Another way to picture a quire is as a booklet consisting of eight leaves, or folios, each leaf consisting of the front (recto) and back (verso) of half of a bifolio.  To prepare books large enough to contain all four Gospels, or large enough to contain the book of Acts and the Epistles, or even the entire New Testament, quires were sewn together to make a multi-quire codex.  Not all quires consisted of only four sheets – for example, Papyrus 45 is a single-quire codex; all its sheets were laid flat in a single stack before being sewn together.      

Breves:  chapter-summaries, especially those that appear in Latin manuscripts.  Some forms of breves appear to have originated very early in the Old Latin transmission-line, including one form – developed in the mid-200s or slightly thereafter – that includes a reference to the pericope adulterae (John 7:53-8:11, which is absent from most early Greek manuscripts of the Gospel of John) in its usual location in the text of John.   

Byzantine Text:  The Greek text of the New Testament that is supported by a strong majority of manuscripts, as represented in the Byzantine Textform compiled by Robinson and Pierpont.   This form of the text was dominant in Constantinople and its environs (i.e., Byzantium) from the 400s onward.  Many Byzantine readings are supported by patristic testimony from the 300s and earlier; the Gothic version and the Peshitta version also provide strong (but not uniform) support for the Byzantine Text. Compared to the Alexandrian Text, the Byzantine Text tends to be longer and easier to understand.  This is, however, a general description; there are variant-units in which the Alexandrian reading is longer.

            When Westcott and Hort issued the 1881 Revised Text, Hort maintained that all distinctly Byzantine readings (which he described as “Syrian,” reckoning that the core of the Byzantine Text had previously been developed at Antioch, in Syria) should be rejected, on the grounds that the Byzantine Text as a whole was the product of a recension, that is, a carefully edited form of the text made by someone – perhaps Lucian of Antioch – whose editorial work consisted of selecting variants from exemplars drawn from Alexandrian and Western transmission-lines.  Readings that deviated from the Alexandrian and Western variants, Hort theorized, must have originated in the mind of the editor who produced the Antiochan text.  Since Hort proceeded to reject the Western Text as having been thoroughly contaminated by expansions, the 1881 Revised Text was almost 100% Alexandrian at points where these three major forms of the text disagree – and distinctly Byzantine readings, despite being supported by almost all surviving Greek manuscripts, were very few and far between.

            Hort’s theory, however, was greatly weakened by the discovery – in papyrus manuscripts which had been excavated in Egypt, and which appeared to have been produced before or during the lifetime of Lucian of Antioch – of readings which did not agree with the flagship manuscripts of the Alexandrian Text, nor with the Western Text.  This implied that whatever the origins of every distinctive Byzantine reading might be, they could not all have originated during the undertaking of a recension made in the late 200s or early 300s, because at least some distinctive Byzantine readings already existed at that time.  If the Lucianic recension ever happened, it had to involve the consultation of not only Alexandrian and Western exemplars, but also exemplars containing at least some Byzantine readings – in which case, Hort’s basis for rejecting all distinctive Byzantine readings falls to the ground. 

            Nevertheless, even after the discovery of distinctive Byzantine readings in Egyptian papyri, the heavily Alexandrian Revised Text continued to be promoted, especially in Nestle’s Novum Testamentum Graece, and in the United Bible Societies’ Greek New Testament, which are the primary base-texts currently used by most translators.  In A Textual Commentary on the Greek New Testament, author Bruce Metzger – a member of the UBS compilation-committee – defended over 1,000 rejections of Byzantine readings that have an impact on translation.

Caesarean Text:  The form of the text of the Gospels displayed in manuscripts 1582, 1, and some Armenian and Georgian manuscripts.  The testimony of manuscripts 1 and 1582 is augmented by support from an assortment of other manuscripts including 118, 131, and 209).  Researcher Kirsopp Lake established that the distinct readings shared by 1, 118, 131, and 209 descend from a shared ancestor in 1901 in the volume Codex 1 of the Gospels and Its Allies.  (The recognition of 1582 as a member of the same family – and as its best Greek representative – came later).  This cluster of Greek manuscripts is called family 1,  and is generally (but not always) characterized by its members’ unusual treatment of the pericope adulterae:  the passage is put after the end of John 21, having been uprooted and transplanted as the note in 1 and 1582 explains: 

“The chapter about the adulteress:  in the Gospel according to John, this does not appear in the majority of copies; nor is it commented upon by the divine fathers whose interpretations have been preserved – specifically, by John Chrysostom and Cyril of Alexandria; nor is it taken up by Theodore of Mopsuestia and the others.  For this reason, it was not kept in the place where it is found in a few copies, at the beginning of the 86th chapter [that is, the 86th Eusebian Section], following, ‘Search and see that a prophet does not arise out of Galilee.’”

            The Caesarean Text is also notable for referring to “Jesus Barabbas” in Matthew 27:16-17.   Advocates of the genuineness of this reading argue that early Christians suppressed it, considering it to be embarrassing that a criminal such as Barabbas had the same name as the Messiah.  Others have noted that appearance of the name “Jesus in this passage may have originated when an early scribe accidentally repeated the letters ΙΝ at the end of the word ϒΜΙΝ in verse 17, and this was misunderstood as the contraction for the word Ιησους (that is, “Jesus”). 

            It is evident that a Caesarean Text exists for all four Gospels.  It is less evident that there is a Caesarean Text of Acts and the Epistles; however, minuscule 1739 represents a distinct transmission-line, and it was copied by the same copyist who made minuscule 1582, so this should not be ruled out.  

Cancel-sheet:  a parchment sheet, folded in the middle and written on both sides, so as to constitute four pages of a manuscript, made to replace the work of the main copyist.  The most well-known examples of cancel-sheets are in Codex Sinaiticus, including the bifolium that contains Mark 14:54-Luke 1:76 (without Mark 16:9-20).       

Catena:  A commentary consisting of a series of comments by patristic authors who accompanies the Biblical text.  Unlike commentaries written by a single author, a catena combines extracts from the writings of several authors, forming a chain (Latin:  catena) of comments.  The identity of the writer being quoted is sometimes, but not always, written in the vicinity of his comments.  The earliest known Greek catena is in Codex Zacynthius (040, Ξ), an incomplete copy of the Gospel of Luke.

           

Codex (plural:  Codices):  A handmade book. 

Colophon:  a note added to the text of a manuscript.  The contents of such notes can vary; the most useful colophons are those which mention the year and location where the manuscript was produced.  They may also convey the name of the scribe, the name of the patron who sponsored the manuscript’s production, and even declare a curse against whoever might think about taking the manuscript away from the library to which it was entrusted. 

Commentary manuscripts:  A manuscript in which the text of a commentary by one individual accompanies the Biblical text.  Such material is similar to a catena, especially since although a commentary may be written by a single individual, that individual may make free and generous use of the works of other commentators, sometimes acknowledging his source and sometimes not.  As Robert Waltz mentions in his article on Commentaries in the online Encyclopedia of New Testament Textual Criticism, manuscripts with commentaries tended to have one of two forms:  one in which the commentary frames the text, and one in which segments of the text and segments of the commentary alternate.  Frame-commentaries were capable of accompanying texts unrelated to the commentary itself; alternating-commentaries, meanwhile, were always copied at the same time as the Scripture-text they accompany.  For this reason, whenever the same alternating-commentaries accompany the same text, their testimony should be “boiled down,” so to speak, to the testimony of their shared ancestor.

            Some commentaries were more popular than others.  For the Gospel of Matthew, John Chrysostom’s commentary was most popular; for Mark, the Catena-Commentary of Victor of Antioch (a.k.a. the Catena in Marcum) was widely disseminated (and sometimes wrongly attributed to other authors/compilers such as Cyril of Alexandria or Peter of Laodicea).  The commentary of Titus of Bostra was the dominant commentary on Luke.  And for the Gospel of John, copies of both the commentary by John Chrysostom and the commentary by Theophylact are abundant; the latter appears mainly in the alternating format.  Among the other commentators whose work accompanies the New Testament text in some manuscripts are Andreas of Caesarea (in specially formatted copies of Revelation), Andreas the Presbyter (in some copies of Acts and the Epistles),  Oecumenius, and Euthymius Zigabenus.        

Conflation:  a reading which is a combination of two earlier readings.  The presence of conflations implies that the text containing them emerged later than the text that contains its component-parts.  Eight apparent conflations in the Byzantine Text of the Gospels, comprised of component-parts that appear to be combinations of component-parts consisting of Alexandrian and Western readings, were a major part of Hort’s case against the Byzantine Text. 

            However, conflations appear in major representatives of all text-types, not just in the Byzantine Text.  In Codex Sinaiticus, in John 13:24, where the Alexandrian Text reads και λεγει αυτω ειπε τις εστιν and the Byzantine Text reads πυθεσθαι τις αν ειη, Sinaiticus’ text appears to combine those two phrases, reading πυθεσθαι τις αν ειη περι ου ελεγεν, και λεγει αυτω ειπε τις εστιν.  A conflation also appears in Codex Vaticanus at Colossians 1:12:  the Western Text reads καλεσαντι, the Byzantine Text reads ικανωσαντι, and Vaticanus reads καλεσαντι και ικανωσαντι, a combination of the Western and Byzantine readings.  And in Codex D, a conflation appears in John 5:37:  the Alexandrian Text (supported by Papyrus 75) reads has εκεινος μεμαρτυρηκεν, and the the Byzantine Text (supported by Papyrus 66) reads αυτος μεμαρτυρηκεν; the reading in Codex Bezae is εκεινος αυτος μεμαρτυρηκεν and this is precisely what would be produced by a copyist wishing to preserve two different readings in two different exemplars.

            Researcher Wilbur Pickering, in Appendix D of his book The Identity of the New Testament Text, investigates several more cases of apparent conflation in non-Byzantine manuscripts; while some of his examples are capable of more than one explanation, it seems sufficiently clear that the appearance of conflations in a manuscript or text-type cannot validly condemn the entire text-type as late or as posterior to other text-types.

Conjectural emendation:  A reading which is proposed as original but is not supported in any extant Greek manuscript.  The apparatus of the 27^th edition of the Nestle-Aland Novum Testamentum Graece mentioned many of these from a wide variety of sources, but only one was adopted in the text (in Acts 16:12).  In the 28^th edition, all mentions of conjectural emendations were removed from the apparatus – and one conjectural emendation was adopted into the text of Second Peter 3:10, altering the meaning of the sentence.   

[Continued]

A Conflation in John 16:4?

          In John 16:4a, where Jesus warns His disciples about difficult times ahead, there is an interesting textual variant.  In the Byzantine Text, the text means, strictly translated into English:  “But these things I have told you, that when comes their hour, you may remember those things of which I told you.”  

          The Alexandrian Text means something very slightly different:  “But these things I have told you, that when comes the hour, you may remember those things of which I told you.”  The Western Text, as attested by Codex D, has its own variation:  “These things I have told you, that when comes the hour, you shall remember that I told you.”  The Caesarean Text has its own reading too:  “But these things I have told you, that when comes the hour, you may remember that I told you.” 

These four renderings represent the following texts: 

Western:  Ταυτα λελάληκα υμιν ινα οταν ελθη η ωρα μνημονεύτε οτι εγω ειπον υμιν  

Alexandrian:  Αλλα ταυτα λελάληκα υμιν ινα οταν ελθη η ωρα μνημονεύητε αυτων οτι εγω ειπον υμιν

Caesarean (f¹³):  Αλλα ταυτα λελάληκα υμιν ινα οταν ελθη η ωρα αυτων μνημονεύητε οτι εγω ειπον υμιν

Byzantine:  Αλλα ταυτα λελάληκα υμιν ινα οταν ελθη η ωρα αυτων μνημονεύητε αυτων οτι εγω ειπον υμιν

          I want to focus here on the variant in the middle of the sentence, shown in bold print.  The longer reading, αυτων μνημονεύητε αυτων, looks like a combination of the two shorter readings that are listed before it; that is, this variant appears to be a conflation.  

          It may be helpful to view the data about this variant-unit that is found in the fourth edition of the United Bible Society’s Greek New Testament (page 384).  Setting aside minor variants, the contest is between the following rivals:

Western (D):  μνημονεύτε (altered to μνημονευσητε via an addition above the line; see image, from the online presentation of Codex Bezae at the Cambridge Digital Library.)

Caesarean (f¹³):  μνημονεύητε αυτων
Alexandrian (B):  αυτων μνημονεύητε
Byzantine:  αυτων μνημονεύητε αυτων

The citation of D* in UBS4 is incorrect.  D* reads μνημονευτε, as shown here

            Although Hort did not list this passage as a Western Non-Interpolation, he could have.  The Western reading is shorter here, which might tempt some textual critics to propose that an early copyist, seeking to augment the clarity of the sentence, added αυτων in the margin, and some subsequent copyists placed it before μνημονεύητε, and some other copyists placed it after μνημονεύητε, yielding the Alexandrian and Caesarean readings. 

             In addition, Hort did not list this passage as an example of Byzantine conflation, although he could have:  take the Alexandrian reading (with αυτων preceding μνημονεύητε) and the Caesarean reading (with αυτων following μνημονεύητε) and blend them together, and the result is the Byzantine reading:  αυτων μνημονεύητε αυτων. 

             According to advocates of “reasonable eclecticism,” this sort of thing illustrates the secondary and derivative nature of the Byzantine Text.  When a Byzantine reading looks like a combination of the Western and Alexandrian readings, this (they say) is because the Byzantine reading is a combination of the Western and Alexandrian readings, revealing that the Byzantine Text is essentially an amalgamation of the Alexandrian Text and the Western Text.  

             And now, the plot twist.  Contrary to the impression I have given here up to this point, this variant-unit in John 16:4 illustrates an inconsistency on the part of practitioners of “reasoned eclecticism.”  In the preceding paragraphs, I mixed up the Alexandrian and Byzantine readings.  Reverse the Byzantine and Alexandrian readings as I gave them previously, and you will see the evidence as it actually exists: 

The Alexandrian variant is longer than the Byzantine variant: 

            Codex B (supported, according to Swanson, by 118 124 157 and 1071 and, according to UBS4, by P66^vid À² 0233 205 Peshitta, Harklean Syriac, and the lemma of Cyril of Alexandria) reads αυτων μνημονεύητε αυτων. 

            Codex A agrees with B but its orthography is not so good:  αυτων μνημονεύηται αυτων. 

            Codex Θ concurs, but again with a spelling-variation:  αυτων μνημονεύειτε αυτων. 

            Codex Π is a little different, with μνημονεύσητε in the middle, but it still agrees with B by having αυτων both before and after the word. 

   

The Byzantine variant is shorter than the Alexandrian variant:

            Byz (supported by K U Ψ 1582 2^c 700 Lect^pt tr) reads μνημονεύητε αυτων.

            Ε Γ Δ Λ 565 and 1424 concur, albeit with an itacism:  μνημονεύετε αυτων.  This is listed in UBS4 as a separate variant, with additional support from 1006 1241 1243 1342 Lect^pt (l1016 with αυτου).

            Codex 1 stands alone, with a different itacism:  μνημονεύειτε αυτων.

            Codex ﬡ* likewise stands alone with an itacism (also seen in A):  μνημονεύηται αυτων.

            Additional support for the Byzantine reading, according to UBS4:  0141  180 597 (892^supp l866 with αυτου) 1010 1292 1505 ff² Palestinian Syriac Boh^pt (arm) (eth) geo slav (Chrysostom).

             

Before we proceed further, we ought to take a closer look at the testimony of Papyrus 66.  As you can see from this image, there is not a lot left of the lower half of the page of P66 that contains John 16:4.  Nevertheless, a calculated guess can be made about whether or not it read αυτων μνημονεύητε αυτων, αυτων μνημονεύητε, αυτων μνημονεύητε, or just μνημονεύητε when it was in pristine condition.

            Here is the Alexandrian Text of John 16:2b-4, formatted to correspond to the array of letters in P66, with sacred names contracted:

λατρείανπροσφέρειντωθωκαιταυ

ταποιήσουσινοτιουκεγνωσαντον

πραουδεεμεαλλαταυταλελάλη

καυμινιναοτανελθηηωρααυτων

μνημονεύητεαυτωνοτιεγωειπον

υμινταυταδευμινεξαρχηςουκ

ειπονοτιμεθ’υμωνημην.

Here is the Alexandrian Text of John 16:2b-4 again – this time with a slightly different array, and without the contraction of patēra: 

δόξηλατρείανπροσφέρειντωθωΚαι

ταυταποιήσουσινοτιουκεγνωσαντον

πατέραουδεεμεΑλλαταυταλελ

άληκαυμινιναοτανελθηηωρααυ

τωνμνημονεύητεαυτωνοτιεγωει

πονυμινταυταδευμινεξαρχηςουκ

ειπονοτιμεθ’υμωνημην.

Here is the Byzantine Text of John 16:2b-4, formatted to correspond to the array of letters in P66, with all sacred names contracted:

δόξηλατρείανπροσφέρειντωθωκαι

ταυταποιήσουσινοτιουκεγνωσαν

τονπραουδεεμεαλλαταυταλελ

άληκαυμινιναοτανελθηηωρα

μνημονεύητεαυτωνοτιεγωειπον

υμινταυταδευμινεξαρχηςουκ

ειπονοτιμεθ’υμωνημην.

Here is the Byzantine Text of John 16:2b-4 again – this time with a slightly different array, and without the contraction of patēra: 

ξηλατρείανπροσφέρειντωθωκαιταυ

ταποιήσουσινοτιουκεγνωσαντονπα

τέραουδεεμεΑλλαταυταλελάλ

ηκαυμινιναοτανελθηηωρα

μνημονεύσητεαυτωνοτιεγωειπ

ονυμινταυταδευμινεξαρχηςουκ

ειπονοτιμεθ’υμωνημην.

Here is another configuration of the Byzantine Text, with patēra contracted:

λατρείανπροσφέρειντωθωκαιταυ

ταποιήσουσινοτιουκεγνωσαντον

πραουδεεμεαλλαταυταλελάλη

καυμινιναοτανελθηηωρα

μνημονεύητεαυτωνοτιεγωειπ

ονυμινταυταδευμινεξαρχηςουκ

ειπονοτιμεθ’υμωνημην.

          In this array, the Byzantine Text neatly corresponds to the remains of P66 – but αυτων would fit neatly in the space after ωρα at the end of the fourth line.

         All things considered, it seems probable that P66 contained αυτων both before and after μνημονεύητε when it was in pristine condition.

          It might also be a good idea to take a closer look at Codex Sinaiticus.  No mystery here:  it initially read μνημονευηται αυτων after ωρα, thus disagreeing with Codex Vaticanus and supporting the Byzantine reading.  A later “corrector” has added αυτων in the right margin, to be read immediately after ωρα, thus creating an agreement with B.   

          Now let’s get back to that four-way contest.  Setting aside the lightweights, here again are the heavyweight contenders:

            Western (D):  μνημονεύτε

            Caesarean (f¹³):  αυτων μνημονεύητε
            Byzantine :  μνημονεύητε αυτων
            Alexandrian:  αυτων μνημονεύητε αυτων

            Nothing stands in the way of conclusion that the Caesarean reading + the Byzantine reading = the Alexandrian reading.  Nevertheless, instead of concluding that a conflation has been made in the Alexandrian Text, the advocates of “reasoned eclecticism” resort to other hypotheses.  

            Metzger asserted (in A Textual Commentary on the New Testament, page 247) that the Alexandrian reading should be accepted “because of the strength of the external evidence (p^66vid A B Θ Π* 33) and because αυτων after ωρα was more likely to be removed as superfluous than added by copyists.”  But obviously not all copyists thought along those lines:  if the Caesarean reading was descended from the Alexandrian reading, then some copyists must have chosen to retain the αυτων after ωρα and remove the other one. 

            Before forming a deduction about why copyists did what they did, we should consider the matter of authorial style.  In the Gospel of John, ωρα is used 19 times, but never accompanied by αυτων except here in John 16:4 in the Alexandrian Text.  This consideration thus weighs in against the Alexandrian reading.  Why did this go unmentioned by Metzger? 

           

             John 16:4 is not the only passage where we encounter something that appears to be an Alexandrian conflation.  Let’s take a look at Matthew 3:12.  This variant-unit is not featured in the apparatus of UBS4, although it was included in the apparatus in UBS2 and commented upon by Metzger in A Textual Commentary.  Apparently when a different reading was adopted for UBS3 than what had been in UBS2 (in which the second αυτου is bracketed), someone made a decision to remove the variant-unit from the apparatus.  

            In Matthew 3:12 the rival variants witnesses line up as follows:

            Groucho:  εις την αποθηκην                             ([the wheat] into the barn)

            Chico:  εις την αποθηκην αυτου                      ([the wheat] into his barn)

            Zeppo:  αυτου εις την αποθηκην                     (his [wheat] into the barn)

            Harpo:  αυτου εις την αποθηκην αυτου           (his [wheat] into his barn)

 

            Can you guess which variant is attested by which text-type?  According to the apparatus in UBS2 and Swanson’s volume on Matthew, here are the witnesses’ testimonies (versional witnesses are in blue):

            εις την αποθηκην:  f¹³ it^a it^q geo^{1, A} Justin Clement Irenaeus

            εις την αποθηκην αυτου:  E L U 157 892 1195 1253 1424 1646 it^b it^ff1 it^g1 Curetonian Syriac, Sinaitic Syriac, Peshitta, Harklean, Armenian, Irenaeus Ambrose Cyril

            αυτου εις την αποθηκην:  ﬡ C D^supp K M S Δ Ω  f¹ 2 28 33 565 700 1009 1010 1230 1241 1365 2148 2174 Byz it^aur it^c it^d it^f it^l Vulgate, Sahidic, Bohairic, Hilary Augustine

            αυτου εις την αποθηκην αυτου:  B W 1071 1216 eth^{ro, pp} (eth^ms? εις την αποθηκην) geo^B 

            Although this is the same kind of scenario that is confronted when Byzantine conflations are posited, I submit that in Matthew 3:12 and John 16:4, the Alexandrian Text does not display a true conflation.  (Similarly I do not grant that every Byzantine conflation identified by Hort is actually a conflation, but that’s another subject.)  Instead, it displays the effects of a scribal phenomenon in which a copyist anticipated an approaching word and wrote it twice instead of once – once where it did not belong, and once where it appeared in his exemplar.  In other words, the double-appearance of αυτων in the Alexandrian Text of John 16:4, and the double-appearance of αυτου in the Alexandrian Text of Matthew 3:12, did not originate thoughtfully, but are essentially cases of dittography.  The Alexandrian reading in John 16:4 is not necessarily a conflation, but it is a corruption.

_______________

Postscript: 

Why Is Today’s “Reasoned Eclectic” Base-text

for Bible Versions Becoming Less Eclectic?

            The first English version to reflect the Alexandrian Text of John 16:4a was, as far as I can tell, Granville Penn’s The Book of the New Covenant, published in 1836:  “but these things I have told you, that when their time cometh, ye may remember that I told you of them.”  

            Penn did not have the resources that textual critics have today; he did not even have the resources that Westcott and Hort had when they compiled their revision in 1881.  Instead, as Penn explained in a supplemental volume, Annotations to the Book of the New Covenant, he used the text of Codex Vaticanus as the basis for his revision-work:  “I have taken the continued and entire text of the most ancient surviving manuscript, the “Codex Vaticanus” or Vatican MS., noted 1209 in the Vatican catalogue, and marked B by Wetstein; making it the basis and substance of the revision.”  

            As far as the text of the Gospels is concerned, Penn’s 1836 translation thus echoed Vaticanus (or, the imperfect editions of its text that were available in 1836), except where its readings seemed beyond hope of salvaging.  The compilation of Penn’s base-text was thus determined by the most non-eclectic principle imaginable; he gave Codex B a virtual monopoly of influence.       

            And yet, in 2015, when we examine the “reasoned eclectic” text as it currently exists, and compare it to earlier compilations used in the 1900’s as the base-texts for English translations, its degree of agreement with Penn’s base-text frequently seems to increase rather than decrease.  Consider, for example the text of John 16:4a in the NIV as it existed in 1984 (and as it was still being printed in 2002):

            “I have told you this, so that when the time comes you will remember that I warned you.” (Scripture taken from the Holy Bible, New International Version ®,  Copyright © 1973, 1978, 1984 by International Bible Society.  Used by permission of Zondervan.  All rights reserved.)

            The compilers of the NIV’s base-text apparently adopted the Western reading of John 16:4a. 

            The Contemporary English Version, likewise, seems to follow the Western reading:

            “I am saying this to you now, so that when the time comes, you will remember what I have said.” (Contemporary English Version (CEV) Copyright © 1995 by American Bible Society.)

            The New Century Version, likewise, conforms to the Western reading of John 16:4a: 

            “I have told you these things now so that when the time comes you will remember that I warned you.” (The Holy Bible, New Century Version ®, Copyright © 2005 by Thomas Nelson, Inc.)

            However, in the TNIV and in the new editions of the NIV, the text of John 16:4a is different:

            “I have told you this, so that when their time comes you will remember that I warned you about them.”  [Bold print added for emphasis.]  (Scripture taken from the Holy Bible, New International Version ®, NIV® Copyright © 1973, 1978, 1984, 2011 by Biblica, Inc. ®.  Used by permission.  All rights reserved worldwide.)

            Thus the base-text of John 16:4 in the newly revised NIV has quietly become less eclectic than the NIV of 1984.  The New Revised Standard Version (NRSV), the English Standard Version (ESV), and the Common English Bible (CEB) reflect the same Alexandrian base-text of John 16:4.  It would seem, from this lone example, that although more text-critical materials are available now than ever before (in the form of newly discovered manuscripts, new editions of patristic writings, etc.) “reasoned eclecticism” is producing a text that very closely resembles what Penn produced in 1836 by ignoring almost all evidence except Codex Vaticanus.  In terms of results, it is almost anti-eclectic.

The Text of Reasoned Eclecticism: Is It Reasonable and Eclectic? Part Four of a Four-Part Response to Dan Wallace

Testing Reasoned Eclecticism

            Wallace attempts to separate his “reasoned eclectic” position from the pro-Alexandrian view of Westcott and Hort.  He stated that the Westcott-Hort theory has “many flaws.”  However, as Eldon Epp has observed, the Nestle-Aland text “almost always departs from the B-text only when an À versus B attestation is in question.”²⁹  If the methodologies of Nestle-Aland and Westcott-Hort are so different, why are their results so similar?  How is it that the Nestle-Aland text differs at only 558 points from a text that was based on a theory with “many flaws”?³⁰  

            Wallace describes the Nestle-Aland text as much more eclectic than the Westcott-Hort text, noting that “In scores of places the editors of the modern critical texts have adopted a Byzantine reading against an Aleph-B alignment (contra Hort).”  But what do these changes really constitute? 

            There is a way to find out:  sift through Wieland Willker’s online presentation of the Westcott-Hort text, which parenthetically displays the readings of NA27 where they diverge from WH-1881.  What differences have emerged in, say, the first ten chapters of Matthew?  A total of 55 differences, which are listed in the appendix of this article (in the post that follows this one).

            Seventeen of these changes, however, are not textual, but only consist of the addition or removal of brackets.  Of the remaining 38 changes, most of them do not make the text any more eclectic than Hort’s text, because they only consist of exchanging a reading in ﬡ for a reading in B, or a reading in B for a reading in À.  Only 17 changes in Matthew 1-10 actually make the text of NA26 more eclectic than Hort’s text, and of these 17 (indicated by red dots in the list in the appendix), eight are bracketed in NA26.  The only newly adopted readings not supported by either ﬡ or B, and also not displayed in brackets, are the following:     

           

● 3:7 – αυτου, after βαπτισμα, has been adopted, although it is absent in B and À.

● 3:14 – Ιωαννης has been added, although it is absent in B and À.  Inclusion is supported by P⁹⁶.

● 8:8 – NA reads και αποκριθεις instead of αποκριθεις δε, which is supported by B and À.

● 8:9 – NA does not include τασσομενος, which Hort included in brackets.  Inclusion is supported by B and À.

● 9:18 – NA reads ελθων instead of Hort’s προσελθων, which is supported by B and À.

● 9:32 – NA reads ανθρωπον, although non-inclusion is supported by B and À.

● 10:13 – NA, after the second occurrence of ειρηνη υμων, reads προς instead of εφ.  B and À support εφ.  

● 10:23 – NA includes αν, which B and À do not include.

            The confidently presented (i.e., not bracketed) “reasoned eclectic” text of Matthew 1-10 is more eclectic than Hort’s 1881 text – by a margin of four additions, one subtraction, and three substitutions.  In four of these eight cases (specifically, in 3:7, 9:32, 10:13, and 10:23) the adopted reading is supported by Codex D, and thus does not represent the adoption of a reading distinct to the Byzantine Text.  Thus the distinctly Byzantine contribution to the non-bracketed Nestle-Aland text of Matthew 1-10 amounts to one addition, two substitutions, and one subtraction (and the subtracted word was already bracketed by Hort).
            That’s next to nothing.  The results of “reasoned eclecticism,” as practiced by the compilers of Nestle-Aland, and by Wallace, are virtually the same text that Hort produced.  Why do they so persistently reject the Byzantine Text?  The answer is obvious:  against all evidence to the contrary, they are still entrenched in the belief that the Byzantine Text is a blend of Alexandrian and Western readings that were combined by an editor around 300.  In other words, they still adhere to Hort’s theory that the Byzantine Text originated with the Lucianic Recension, as Wallace shows when he states that the Westcott-Hort theory “was apparently still right on its basic tenet:  the Byzantine texttype—or majority text—did not exist in the first three centuries.”  

            Is that a reasonable assumption? 

            By the early 400’s, the Byzantine text of the Gospels, or a substantial strata of it – a text that resembled the Byzantine Text far more than any other text-type – had been used by Wulfilas, Basil, Chrysostom, and the unknown translator(s) of the Peshitta.  It was also being disseminated in copies such as Codex A and Codex W.  In the 500’s, a similar text was used for the Purple Uncials (N-O-Φ-Σ). 

            What was happening in the areas where these writers lived?  Although we do not have copies of the Greek New Testament from Syria, Asia, and Greece before 300 – because they were made of papyrus, and those that were not victims of Roman persecutors became victims of high humidity – it should go without saying that the Greek-speaking Christians in these areas possessed copies of New Testament books, and routinely used them in their church-services.  Did the people in those areas simply set aside their old Greek texts – Alexandrian, or Western – and pick up a new Greek text – the Byzantine Text – that was brimming with new readings? 

            The idea is inherently improbable.  Such an act would not be analogous to what happened around 425 in Syria, when Theodoret replaced 200 copies of the Diatessaron with copies of the separate Gospels.  Theodoret was replacing the work of Tatian, who – rightly or wrongly – had acquired a reputation as a heretic, with the work of apostles and their associates.  The churches of Syria had a strong theological motive to accept Theodoret’s gift as an improvement, and to set the Diatessaron aside, regarding it as a dubious proxy.  But on what basis would a newly introduced Greek text of the Gospels (or of the entire New Testament), never before seen, be favored – and rapidly favored – over the Greek text of the Gospels that had been used in these areas for generations, and which those preceding generations had risked their lived to protect and preserve?

            Instead of assuming that the local text of Antioch, Asia, and Greece that was being used in the 200’s was suddenly abandoned in the 300’s, it is more reasonable to deduce that the Byzantine Text – or at least a very large part of it – was the local text of Antioch, Asia, and Greece in the 200’s, and that is why it continued to be the local text of those areas in the 300’s and 400’s.  As other text-types infiltrated the area, conflations occurred, but only rarely.  The amount of conflation (or apparent conflation) that Hort proposed was on display in the Byzantine Text is not materially greater than what one observes in early Alexandrian witnesses.³¹        

            A handful of conflations (in which a proto-Byzantine reading was combined with either an Alexandrian or Western reading) sprinkled over a strata of earlier text should not be considered a sufficient reason to regard the entire text-type as late. 

            Even early manuscripts such as P⁵³ and P⁶⁶ contain apparent conflations – and, occasionally, the early papyri contain distinctly Byzantine readings, that is, readings that are found in the Byzantine Text which are not found in the flagship manuscripts of the Western or Alexandrian texts.  A sample of these, from data provided by Harry A. Sturz,³² may be provided:  P⁴⁵ agrees with the Byzantine Text in the Gospel of Mark, disagreeing with ÀB and D, in 5:22, 5:42, 6:2, 6:16, 6:45, 6:48, 7:12, 7:30 (twice), 7:31, 7:32, 7:35 (twice), 7:36, 9:6, 9:20 (twice), 12:6, and 12:16.  Sturz proceeds to list 26 agreements in Luke between early papyri and the Byzantine Text that disagree with ﬡB and D, and 44 agreements in John between early papyri and Byz that disagree with ﬡB and D. 

            Wallace seems to believe that every one of these Byzantine readings entered the early papyri out of nowhere.  Why not regard them, or at least some of them, as results of mixture with a local text that was infiltrating Egypt in the early 200’s?³³  

Conclusion

            One does not have to adhere to the notion that the Byzantine Text was the majority text in the 200’s to grant that a substantial strata of the Byzantine Text was the local text of Antioch and Asia in the 200’s.  All one has to grant is the inherently probable premises that (a) the Greek text in Antioch and Asia did not develop along exactly the same lines in which it developed in the regions where the Western and Alexandrian texts dominated, and (b) when copyists made the shift from papyrus to parchment in Antioch and Asia, they did not set aside their most ancient, most respected local exemplars, but continued to use the same form of the New Testament text that previous generations had been using in their area, with only a minimal amount of accretions.     

            If these premises were accepted, instead of the premise that the Byzantine Text did not exist at all until the 300’s, then text-compilers might conceivably give Byzantine readings the hearing that they deserve, instead of reaching for whatever hypothesis (often the assertion of intentional corruption) is required to dismiss the Byzantine reading and maintain the Alexandrian reading, even when it is easily attributable to scribal carelessness.  Then, and only then, will it be possible to compile a Greek New Testament based on truly reasonable premises, with truly eclectic results. 

●●●●●●●

_______________

FOOTNOTES

29 – See page 136 of Eldon Epp’s article, A Continuing Interlude in New Testament Textual Criticism? in Harvard Theological Review, Vol. 73, 1980.)  Epp raises an important question:  today’s textual critics (“today” being 1980, when he wrote – his point is even more cogent now) have “more than 80 papyri, more than 200 additional uncials, more than 2600 additional minuscules, and perhaps 2000 additional lectionaries that were unavailable to or were not utilized by Westcott-Hort.”  Why, so far, has it made so little difference?

30 – According to the statistics presented by Michael Marlowe at http://www.bible-researcher.com/stats.html , there are 551 differences between the 1979 Nestle-Aland text and the 1881 text of Westcott & Hort.  Slight fuzziness in the statistic may be due to different treatments of bracketed words in both texts.  As a peripheral point, it should be noted that some of those 558 differences exist because the compilers of the Nestle-Aland text – or at least a majority of the compilers – rejected Hort’s occasional preference for Western readings (the most famous examples being the “Western Non-Interpolations” in Luke 24).  Each time the compilers rejected a Western reading that Hort had favored, the resultant compilation became less eclectic and more Alexandrian.

31 – If one carefully sifts through Pickering’s Appendix D – Conflation or Confusion? in The Identity of the New Testament Text, fourth edition, one may find raw data indicative of conflations in either Vaticanus or Sinaiticus in Matthew 3:12, 24:38, Mark 1:4, 1:28, 1:40, 4:5, John 7:39, 13:24, 16:4, Acts 24:14, First Corinthians 7:34, Ephesians 2:5, Philippians 1:18, Colossians 1:12, 3:17, and Hebrews 9:10.  I would also draw the reader’s attention to the conflation in À in the third verse of Jude, in which the Alexandrian Text is combined with a reading of the Harklean Group.  This data tends to shatter Wallace’s claim that “Nowhere could it be shown that the Alexandrian combined Western and Byzantine readings.”

32 – See List 1 on pages 145-153 of The Byzantine Text-type & New Testament Textual Criticism, and selections from List IV, on pages 191-194.    

33 – This would account for the Byzantine readings that are found in compositions by authors such as Origen and Didymus, who were outside the locales where the Byzantine Text was popular, but were capable of encountering stray Byzantine copies and occasionally favoring their readings, such as “Bethabara” in John 1:28, which was mentioned and defended by Origen.  The early, rather than later, influence of the Byzantine Text in Caesarea and its environs may also efficiently account for the strong Byzantine element that exists in Caesarean witnesses such as 1582; regarding this see Robert Waltz’s tentatively framed research at http://www.skypoint.com/members/waltzmn/ByzPrior.html#Testing .