Friday, August 4, 2017

Nestle-Aland, Edition 28: Cracks in the Text

          The Nestle-Aland Novum Testamentum Graece is the base-text of most major modern English versions of the New Testament.  The editors of the 28th edition of NTG introduced 34 changes to the text, all in the General Epistles (James-Jude).  Since the text of the 27th edition of NTG was the same as the text in the 26th edition (published in 1979), this means that all of the discoveries and research of the past 38 years might as well not exist as far as the standard critical text of the other 20 books of the New Testament are concerned.  
          But changes are coming.  If the changes introduced in NA28 are indicators of the kinds of changes yet to be seen in the rest of the New Testament, two things are clear: 
          (1) Whereas NA27 disagreed with the Byzantine Text in James-Jude 279 times, NA28 disagrees with the Byzantine Text in those seven books 272 times.  There is thus a very slight increase in the value assigned to Byzantine readings.
          (2) One of the changes introduced in NA28 was a conjectural emendation.  In Second Peter 3:10, where previous editions read εὑρεθήσεται (“shall be found”), NA28 adds a word, with the result that the sentence conveys precisely the opposite meaning – οὐχ εὑρεθήσεται (“shall not be found”). Although Second Peter 3:10 also ends with “shall not be found” in the Sahidic version and in a manuscript of the Harklean Syriac version (or of the Philoxenian Syriac; I have seen this evidence cited in more than one way), there is no Greek support for the inclusion of the word οὐχ.  Thus there is a slight increase in the value assigned to conjectural emendations of the Greek text.  (One could say that Greek support is non-existent for 3% of the readings (1 out of 34) introduced into the text of NA28.)

Second Peter 3:9b-12a in Sahidic.
        It may seem at first that these two tendencies represent two opposite approaches, as if one group of editors valued the reading found in the vast majority of Greek manuscripts, and another group of editors was willing to adopt readings that are not found in any Greek manuscripts.  However, they are both effects of the rise of the same approach:  thoroughgoing eclecticism, in which the textual editors adopt whatever reading they consider to best account for its rivals, even when its Greek support is weak or non-existent.        
          What might future editions of the critical text look like if the editors apply thoroughgoing eclecticism to the text from Matthew to Revelation?  Here is a list of some cracks in the text – places where changes to the Nestle-Aland compilation are most likely to occur in the future, mainly due to the editors’ openness to adopt readings with minimal or non-existent Greek support, and sometimes due to a higher value assigned to Byzantine readings with widespread support.      

● Matthew 1:16 – Instead of the usual text that means that Joseph was betrothed to Mary, and that Jesus, who is called the Christ, was born of Mary, the Sinaitic Syriac manuscript says that Joseph, to whom the virgin Mary was betrothed, begat Jesus, who is called the Christ.”  This was all it took, in Hermann von Soden’s compilation of the Greek New Testament (published in a series of volumes in 1902-1910), to rationalize altering the Greek text of Matthew 1:16b so as to read Ἰωσὴφ δε. ᾧ ἐμνηστεύθη παρθένος Μαριάμ, ἐγέννησεν Ἰησοῦν τὸν λεγόμενον Χριστόν, as if the original text meant the same thing as the Syriac text in the Sinaitic Syriac manuscript. 
            When J. R .Moffatt made his 1917 version of the New Testament, he used von Soden’s compilation as his base-base text; consequently, Moffatt rendered Matthew 1:18 as “Jacob the father of Joseph, and Joseph (to whom the virgin Mary was betrothed) the father of Jesus, who is called ‘Christ.’”  (The Sinaitic Syriac also adds the clause “to you” in Matthew 1:21, and omits the phrase “and he knew her not” in Matthew 1:25, and then states that “She bore to him a son.”)  One might suppose that Christian academia would consider this scandalous or heretical, but as it turned out, Moffatt was later chosen to serve on the translation-committee for the Revised Standard Version.  Von Soden’s treatment of Matthew 1:16, though, has been abandoned . . . so far.      

● Matthew 21:44 – Although this verse is only missing in a few Greek manuscripts and assorted versional and patristic witnesses (mainly Old Latin copies and the Sinaitic Syriac manuscript), and it could be argued that simple scribal careless is the reason (inasmuch as Καὶ begins this verse and the next one), its non-inclusion is noted in footnotes in the ESV, CSB, and NRSV, and it is feasible that future editors may relegate it to a footnote.   

● Matthew 28:19 – The reference to the baptismal formula “in the name of the Father, and of the Son, and of the Holy Spirit” is in all Greek manuscripts of the passage.  In a quotation of the passage by Eusebius of Caesarea (who lived in the late 200’s and early 300’s), the passage is cited as if Jesus said, “Go and make disciples of all nations in My name.”  In his Oration in Praise of Constantine (16:8), Eusebius likewise utilizes Matthew 28:19 as if Jesus said, “Go and make disciples of all nations in My name.”  In 1902, researcher F. C. Conybeare argued in an article in The Hibbert Journal that the triune formula was an early interpolation, and that although Eusebius also quoted the passage with the entire formula included, the shorter reading is the original one.  (Responses to Conybeare’s case were hammered out almost as soon as his article was released.) 
The textual apparatus for the final
verses of Matthew 28 in the 16th
edition of the Nestle compilation.
           The reading εν τω ονοματι μου (“in my name”) used to be in the apparatus of the Nestle compilation – see for example the 16th edition, from 1935 – but was later expunged.  It is possible that it might reappear in the apparatus in future editions, which would probably delight some Unitarians and Oneness Pentecostals.       

● Mark 1:1 – The phrase “Son of God” is missing from a smattering of manuscripts, and this has led some textual critics (including Michael Holmes, editor of the SBLGNT) to regard the words as an addition by copyists.  It is possible that future editors of the Nestle-Aland compilation may agree.

● Mark 1:41 – Although only one Greek manuscript – Codex Bezae – reads ὀργισθεὶς (the Greek equivalent of “angry”), the 2011 edition of the NIV used that as its base-text, instead of σπλαγχνισθεὶς (“filled with compassion”) which was adopted (without a footnote) in the 1984 NIV.  Σπλαγχνισθεὶς is read by the Byzantine Text, by Codex Vaticanus, by Codex Sinaiticus, and by pretty much all other Greek manuscripts of any significance.  (A few medieval copies don’t have either term, harmonizing the passage with the text of the parallel-account in Matthew 8.)  I have explained that the reading in Codex Bezae is intrinsically improbable, and is likely the result of an attempt to retro-translate the Greek text to correspond to the Latin text (Codex Bezae is a Greek-Latin manuscript, in which the text is written in Greek, and then the same, or approximately same, passage is written on the opposite page).  Nevertheless it is possible that future editors, favoring the shallow arguments that have been offered by Bart Ehrman in favor of ὀργισθεὶς, may introduce it into the text.

● Mark 5:22 – Whose daughter did Jesus raise from the dead?  Mark and Luke tell us his name.  Why did Matthew leave out this little detail?  Researcher J. K. Elliott has suggested that one possible explanation is that Matthew left it out because it was not in his copy of the Gospel of Mark – and that although the non-inclusion of the words ὀνόματι Ἰάϊρος (“named Jairus”) is attested only by Codex Bezae and a few Old Latin copies, internal evidence tips the scales toward non-inclusion of these words, the idea being that there seems to be no obvious reason why they would be removed if they were originally present. 

● Mark 15:25 – Various commentators on the Gospels have perceived an apparent discrepancy between Mark’s statement that Jesus was crucified at the third hour, and the statement in John 19:14 that Jesus was being sentenced by Pilate at the sixth hour.  Rather than imagine that different methods of hour-reckoning are involved, some individuals, ancient and modern, have proposed that either the text of Mark 15:25 or the text of John 19:14 contains an ancient error, and that the Greek numeral ϝ (the obsolete letter digamma, which stands for “6” when written as a numeral, and which was written in a few different forms, one of which looks like the English letter F) was misread as if it was Γ (the letter gamma, which stands for “3” when written as a numeral).  Some copyists apparently thought that this idea must be correct, and wrote the Greek equivalent of “sixth” in Mark 15:25; a few others (including the copyists of the important uncials L and Δ) wrote the equivalent of “third” in John 19:14.  Future editors might consider either reading original, arguing that John was aware of the Gospel of Mark and would not starkly contradict Mark.   

● Luke 1:46 – Although all Greek manuscripts of this passage state that Mary is the person who spoke the Magnificat, a few early Old Latin witnesses attribute the song to Elizabeth instead.  Jeffrey Kloha, who was recently recruited by the Museum of the Bible to be its director of Collections Operations, has offered an interesting case for this reading.

● Luke 4:44 – Two of the most important textual critics of the 1800’s – Samuel Tregelles and Constantine von Tischendorf – adopted the Byzantine reading Γαλιλαίας (of Galilee) rather than the Alexandrian reading Ἴουδαίας (of Judea).  The reason is clear:  before verse 44, Luke describes Jesus’ ministry in Galilee, and after verse 44, Luke is still describing Jesus’ ministry in Galilee.  There seems to be no impetus for the sudden mention of Judea.  The NASB resorts to guiding its readers to a resolution of this problem, not by adopting or even mentioning the reading found in the vast majority of manuscripts, but by explaining in a footnote that Luke intended to refer to “the country of the Jews (including Galilee).”  This seems rather facile – but how else can one accept the text of Luke 4:44 in the Nestle-Aland compilation without conceding that Luke inexplicably confused Judea and Galilee
            Of course the scholars in charge of the Nestle-Aland compilation harbor no reluctance to attribute errors to Luke; their greater difficulty here is to explain how Luke – whose grasp of the geography in his narratives seems well-grounded – managed to make such an incredible mistake.  Future editors may conclude that such a mistake is less likely to have originated with Luke and more likely originated with a very early copyist (earlier than Papyrus 75, which supports Ἴουδαίας), not unlike the reading in Codex Sinaiticus in Luke 1:26, which says that Nazareth was a city in Judea
● Luke 6:1 – It is possible that instead of σαββάτω, future editors might prefer the very widely attested, but difficult to understand, Byzantine reading, σαββάτω δευτερωπρώτω.  All manner of hypotheses have been offered to explain the Byzantine reading and its tenacious ability to survive being copied by scribes who are supposed to have been allergic to difficulties.  The theory proposed by the late Bruce Metzger (in his Textual Commentary on the Greek New Testament) seems rather dubious.  Future editors might surrender to the force of the internal evidence, and admit that in this case it was Alexandrian copyists, rather than Byzantine ones, who excised a difficult feature in their exemplars.

● John 1:13 – All Greek manuscripts of this passage display a text which refers collectively to those “who were born, not of natural descent, or of the will of the flesh, or of the will of man, but of God,” as the CSB renders the passage.  However, some early writers – including Irenaeus and Tertullian – cite the verse as if it refers specifically to Christ, via a singular reference, rather than a plural one.  Tertullian (who wrote in Latin in North Africa) shows that he is also aware of the plural reading, but he insists that the singular reading is correct, and in chapter 19 of his composition On the Flesh of Christ, Tertullian boldly proposes that the plural reading is an effect of tampering.  This might be enough to provoke a footnote here in some English translations in the future.     
● John 1:18 – Although “only-begotten God” (misrendered “the only God” in the ESV) has achieved wide acceptance, especially after it was found in Papyrus 66, there may be something to the suspicion that the adoption of this reading was a sort of theological trade-off in the shift from the Textus Receptus to the text of Westcott & Hort, and then to the Nestle-Aland compilation(s):  theologians who were reluctant to say good-bye to proof-texts such as the Comma Johanneum and First Timothy 3:16 could say hello to new affirmations of Christ’s deity via the acceptance of the Granville Sharp rule in Second Peter 1:1 (as long as the text of Codex Sinaiticus was avoided) and Titus 2:13 – and via the acceptaince of the reading Θεός in John 1:18. 
            However, not only is μονογενὴς Θεός an entirely non-Johannine term (unless one accepts it as genuine, of course), but some translators of major English versions that are based on the Nestle-Aland compilation seem reluctant to translate it accurately.  The CSB, for example, renders John 1:18 as follows:  “No one has ever seen God.  The one and only Son, who is himself God and is at the Father’s side – he has revealed him.”  Just what is the CSB’s textual justification for including both the word “Son” and the phrase “who is himself God” in this verse??? 
            The NIV is equally bad:  “No one has ever seen God, but the one and only Son, who is himself God and is in closest relationship with the Father, has made him known.”  The reading ὁ μονογενὴς Θεός, which is supported by Papyrus 75,  demands a rendering like what is in the NASB:  “the only begotten God.”  But what if the article (ὁ) disappears, as it does in P66, Sinaiticus, Vaticanus, and the Nestle-Aland compilation?  The annotator of the NET has argued for the article-free reading, on the grounds that “θεός without the article is a much harder reading.”  The NET also states that “Although υἱός fits the immediate context more readily, θεός is much more difficult.”  That is true, no doubt, but is it likely that John would produce such a difficult term, and never mention it again?  The NET resorts to an inventive rendering to circumvent what would otherwise seem to be a reference to a god:  “The only one, himself God.”
            A purely theological case against the reading μονογενὴς Θεός might be proposed in light of the various ways in which several of the translations have mangled its meaning and thus confused their readers, on the grounds that God is not the author of confusion.  A more scientific case against μονογενὴς Θεός might be put together via (a) careful consideration of the immense span of the patristic support for ὁ μονογενὴς υἱός, and (b) the observation that the case for Θεός, with or without ὁ, involves the admission that Byzantine scribes declined to adopt a reading which, according to some apologists for the Alexandrian Text, would have been theologically enormously helpful.
            It is not impossible that a third reading, ὁ μονογενὴς, despite being poorly attested – it is supported by Ephrem Syrus (citing the Diatessaron), by the Palestinian Syriac version, by two Vulgate manuscripts, and by Pseudo-Vigilius – might be considered by some future editors to be the variant which best explains its rivals.  Charles Burney also offered a conjecture:  μονογενὴς θεοῦ (“only-begotten of God”).
            In any event, it would not be surprising if future editors and translators alter their treatment of this passage, perhaps by returning to the very widespread reading ὁ μονογενὴς υἱός (the only-begotten Son) in John 1:18 on the basis of the theory that the introduction of Θεός – rather than being the effect of Valentinian tampering, as some have suspected – was due to the early accidental confusion of one sacred-name contraction for another, and that the loss of the article was just one of many such losses typical of the Alexandrian Text.

● John 7:52 – It is sometimes said that the excellence of the preservation of the New Testament text in the extant manuscripts is demonstrated by the observation that despite all of the discoveries of papyrus manuscripts in the late 1800’s and 1900’s, no reading that is supported exclusively by the papyri has been adopted in place of readings that were already extant.  However, a reading of Papyrus 66 might come closer to doing so than any other variant in the papyri. 
John 7:52 in Papyrus 66.
          William Bowyer
’s 1782 book Critical Conjectures and Observations on the New Testament included a theory expressed by Dr. Henry Owen about John 7:52:   “The Greek text, I apprehend, is not perfectly right:  and our English Version has carried it still farther from the true meaning.  Is it possible the Jews could say, “that out of Galilee hath arisen no prophet;” when several (no less perhaps than six) of their own prophets were natives of that country?  When they tell Nicodemus to search the Scriptures (see Cambr. MS. and Vulgate Version), they plainly meant, for the birth-place of the prophet that was to come, i.e., the Messiah; which he would find to be, not any town of Galilee, but Bethlehem in the land of Judea.  Hence then I conclude, that what they really said, and what the reading ought to be, was – ὅτι Ὁ ΠΡΟΦΗΤΗΣ ἐκ τῆς Γαλιλαίας οὐκ ΕΓΕΙΡΕΤΑΙ:  That the prophet is not to arise out of Galilee:  from whence they supposed Jesus to have sprung.”   
            The second part of Owen’s proposal was adopted by textual critics in the 1800’s, for ἐγείρεται is supported by a variety of important manuscripts (including א B D N W Δ).  The first part, though, was not vindicated until the discovery of Papyrus 66, which has the Greek equivalent of “the” before the word “prophet” – exactly what Owen, in the 1780’s, thought was the original reading. 
            (It is sometimes claimed that Papyrus 75 also supports this reading; however, the digital online image of the passage in Papyrus 75 shows that Papyrus 75 has no testimony about this one way or the other, due to thorough damage to the papyrus after the Γαλιλα of the second occurrence of Γαλιλαίας.) 
            It is possible that future editors, impressed by the cogency of Owen’s argument as well as by the early date of the manuscript with which it interlocks, may move the article into the text.        

            The Gospels are not the only part of the New Testament in which non-extant readings, and variants with very slight external support, may be embraced in the future.  God willing, in my next post, I will examine ten other passages, outside the Gospels, which have invited conjectural emendation in the past, and which may do so in the future.


Quotations from the ESV have been taken from The Holy Bible, English Standard Version. ESV® Permanent Text Edition® (2016). Copyright © 2001 by Crossway Bibles, a publishing ministry of Good News Publishers.

Quotations from the NIV have been taken from the Holy Bible, New International Version®, NIV® Copyright ©1973, 1978, 1984, 2011 by Biblica, Inc.®  Used by permission. All rights reserved worldwide.

Scripture quotations marked CSB have been taken from the Christian Standard Bible®, Copyright © 2017 by Holman Bible Publishers.  Used by permission.  Christian Standard Bible® and CSB® are federally registered trademarks of Holman Bible Publishers.

Scripture quotations marked NASB taken from the New American Standard Bible®, Copyright © 1960, 1962, 1963, 1968, 1971, 1972, 1973, 1975, 1977, 1995 by The Lockman Foundation.  Used by permission.

1 comment:

maurice a. robinson said...

I for one would be reluctant to speculate overmuch as to what the editors of future Editio Critica Maior or Nestle-Aland volumes might or might not do since we have no way of knowing either what those future CBGM results might claim in regard to textual "flow" nor as to how the editors' application of eclectic principles might affect those results. Unanticipated surprises for everyone will probably be the most that can be said.

Based on the ECM results in regard to NA28 in the General Epistle and at present what is occurring in ECM Acts, I at least remain impressed with whatever readings move the NA text more toward the Byzantine (which Westcott and Hort likely would not approve).