Video Lecture 06: Some Important Manuscripts

Now at YouTube:

Lecture 06 - Some Important Manuscripts - in the series Introduction to New Testament Textual Criticism. (24 minutes, but viewers are expected to explore the links and thus take longer)

Subtitles provide a basic outline and links to supplemental materials.

https://www.youtube.com/watch?v=9SqzPnVlvWY

An excerpt:

            We are about to meet some manuscripts.  Some of these, you might encounter very often in the apparatus of the Greek New Testament.  Others are relatively small, but they are among the earliest witnesses to the readings they support.  

            In the course of this lecture, I will mention links to supplemental materials as we go. I hope that you will pause this video when a link appears, explore each resource, and then return to the lecture.   These resources include images of manuscripts that can be viewed in fine detail.

            Today I am going to refer to the Alexandrian Text, the Western Text, and the Byzantine Text.  Hopefully in a future lecture I will go into more detail about these terms.  For now, you can generally picture them as three forms of the text:  the Alexandrian Text was used in Egypt, and influenced the Sahidic version there.  The Western Text was used mainly but not exclusively in the Western part of the Roman Empire, and influenced the Old Latin text.  The Byzantine Text  was used in the vicinity of Constantinople, and is generally supported by the majority of Greek manuscripts.

 ● We begin with Papyrus 52.  This is perhaps the oldest manuscript that contains text from the New Testament.  It is small, about the size of a playing card.  It contains text from John 18:31-33 on one side, and on the other side it contains text from John 18:37-38.  Which is not a lot of text.  It was brought to light by Colin H. Roberts in 1935.  

The importance of Papyrus 52, which is at the John Rylands Library in Manchester, England, is its age:  it is probably from about the first half of the 100s.   There is a nice description of Papyrus 52 (and other papyri fragments) by Robert Waltz at the Encyclopedia of New Testament Textual Criticism. In addition, Dirk Jongkind has a brief video about P52 on YouTube. 

● Papyrus 104 is another early papyrus fragment that is a top contender for the title “earliest New Testament manuscript.”  It was excavated at Oxyrhynchus, Egypt by Grenfell and Hunt, and was brought to light in 1997 by J. D. Thomas.  If Papyrus 52 is the earliest manuscript of John, Papyrus 104 is the earliest manuscript of Matthew.  The handwriting used for P104 was executed in a fancier style than what is seen in most other manuscripts; similar handwriting appears in some non-Biblical manuscripts excavated at Oxyrhynchus, including one in which a specific date, from the year 204 or 211, has survived. 

            Papyrus 104 contains text from Matthew 21:34-37 on one side.  The text on the other side is very extremely badly damaged.  But the surviving damaged text there probably contains text from Matthew 21:43 and 45.  This would mean that Papyrus 104 is both the earliest manuscript of Matthew 21 and also the earliest witness for the non-inclusion of Matthew 21:44. 

            Greg Lanier’s detailed analysis of Papyrus 104 can be found online in Volume 21 of the TC-Journal, for 2016.

● Papyrus 23 is a fragment of the Epistle of James, probably made in the early 200s.  It contains text from part of James chapter 1. 

            You can get a very good look at Papyrus 23 by visiting the website of its present home, the Spurlock Museum in Urbana, Illinois.

 ● Papyrus 137 received some fame, before its official publication, by being heralded as if it was from the first century; it was called “First Century Mark.”  It turned out to be not from the first century.  However, this manuscript – a very small fragment containing text from Mark 1:7-9 and Mark 1:16-18 – is the oldest copy of the text it preserves.  Like several other early fragments, it has made no impact on the compilation of the text of the New Testament.

● Papyrus 45 is much more substantial – but it is still very fragmentary.  When it was made in the first half of the 200s, Papyrus 45 contained the four Gospels and Acts.  The order of books, when the manuscript was made, is unknown.  Its surviving pages at the Chester Beatty Library in Dublin contain text from Matthew 20 and 21,  Mark 4-9, Mark 11-12, Luke 6-7, Luke 9-14, John 4-5, John 10-11, and Acts 4-17.  A leaf in Vienna contains text from Matthew 25-26.  This is the earliest known manuscript that contains text from all four Gospels.

            Papyrus 45 has several readings that are especially interesting due to the impact they have on Hort’s Theory of the Lucianic Recension.  Hopefully we will take a closer look at this in a future lecture, but for now, we can just sum it up as the theory that the Byzantine Text – the text in the vast majority of Greek manuscripts – originated as the result of an editorial effort by someone in the late 200s – possibly Lucian of Antioch – who was combining readings from two earlier forms of the text:  the Alexandrian Text, and the Western Text.  Based on this theory, Hort rejected readings in the Byzantine Text that were neither Alexandrian nor Western, reckoning that they did not exist before the Byzantine Text was made.

But in Papyrus 45, which is assigned to the early 200s, there are some readings that are not supported by the flagship manuscripts of the Alexandrian Text or the Western Text.  Readings in Mark 7:35, Acts 15:40, and many other passages show that it is hazardous to assume that non-Alexandrian, non-Western readings should be rejected.

             The text of Papyrus 45 does not agree particularly strongly with Codex Vaticanus, and it does not agree particularly strongly with Codex Bezae either.  In the parts of Mark where Papyrus 45 is extant, its closest textual relative is Codex W – but Codex W’s text in those parts of Mark is not particularly Alexandrian or Western either.

            When Papyrus 45 was first studied, after it was brought to light in the 1930s, there was a tendency to call its text Caesarean, like the text of family-1.  But the late Larry Hurtado showed that whatever Papyrus 45’s text is, it is not closely related to the Caesarean Text.  And while it repeatedly agrees with the Byzantine Text, it is not consistently Byzantine either.

● Papyrus 46 is the earliest substantial copy of most of the Epistles of Paul, basically arranged in order according to their length, with Hebrews between Romans and First Corinthians.  There is some uncertainty about how many epistles the copyist intended to include in the codex.  Part of this manuscript is at the Chester Beatty Library in Dublin, and part of it is at the University of Michigan.  Its most likely production-date is around 200, give or take 50 years.  The text of Papyrus 46 tends to agree with Codex Vaticanus, but not as strictly as one might expect. For example, in Ephesians 5:9, Papyrus 46 agrees with the Byzantine Text, reading “the fruit of the Spirit” instead of “the fruit of the light.”

● Papyrus 66 contains most of the Gospel of John, with some gaps due to incidental damage.  It was found in Egypt in the early 1950s, and was published in 1956.  Its production-date was initially assigned to around 200, but a wider range is possible.  The copyist who wrote the text in Papyrus 66 made over 400 corrections of what he had initially written. 

● Papyrus 75 is also assigned to around 200.  It is a damaged but substantial codex that contains text from Luke and John.  Its surviving text of Luke begins in chapter 3; its surviving text of John ends in chapter 15.  The text of Papyrus 75 is close to the text found in Codex Vaticanus, but the two manuscripts are not related in a grandfather-and-grandson kind of relationship.  Page-views of Papyrus 75 can be found online at the website of the Vatican Library.

Each of the next three manuscripts was designed as a pandect, that is, a large one-volume collection of the entire Bible.  We tend to assume that it is not unusual to have a single volume that contains all of the books of both the Old Testament and the New Testament.  But that is because we are part of a post-printing-press generation.   In the world of manuscripts, Greek pandects of the Bible are rare.

● Codex Vaticanus is a very important manuscript of the Bible, housed, along with many other manuscripts, at the Vatican Library in Rome.  Its New Testament portion was not the subject of scholarly study until the early 1800s, and since then its reputation has grown.  Today it is generally regarded as the most important manuscript of the New Testament.

 Textually, Codex Vaticanus is the paramount representative of the Alexandrian Text.

Vaticanus was produced in the early 300s.  Its text, in the New Testament, is formatted in three columns per page.  This is usually its format in the Old Testament books too, although in the books of poetry the format is two columns per page.  Codex Vaticanus does not contain the entire New Testament; it has no text from First Timothy, Second Timothy, Titus, Philemon, or the book of Revelation; a text of Revelation is in the codex, written in minuscule lettering, but it is not really the same codex. 

Vaticanus also does not contain the text of the book of Hebrews after Hebrews 9:14.     

            The lettering in Codex Vaticanus has been extensively reinforced; that is, someone, long after the codex was made, traced over the lettering, except where, rightly or wrongly, he thought that the text was inaccurate.  The exact date when this was done is a matter of debate.  I suspect that Codex Vaticanus, before it ended up at the Vatican Library, was previously in the hands of an important character in the 1400s named Bessarion, and scribes working for Bessarion may have been responsible for sprucing it up a bit.  This did not materially affect its text.

The entire manuscript can be viewed page by page at the website of the Vatican Library.

● Codex Sinaiticus is the wingman of Codex Vaticanus.  Its text is not as good, but it is more complete.  The New Testament text of Codex Sinaiticus has survived in more or less the same form in which it left its scriptorium in the mid-300s.  “More or less,” that is, because a few centuries after its production, someone attempted to adjust many of its readings, but those attempts can be detected.  In addition to containing the text of every book of the New Testament, Codex Sinaiticus also contains the Epistle of Barnabas and part of the Shepherd of Hermas.

Most of the text of Codex Sinaiticus is Alexandrian.  However, in the first eight chapters of John, more or less, its text tends to be more like the Western Text.  It is as if the copyists were working from an exemplar of the Gospels that was Alexandrian, but in these opening chapters of John, their main exemplar was damaged, and so they used a drastically different exemplar as their back-up.

That would be consistent with a historical scenario that is mentioned by Jerome, who states that Acacius and Euzoius, at Caesarea in the mid-300s, labored to replace texts written on decaying papyrus in the library there with more durable parchment copies.  Whereas Codex Vaticanus does not have the Eusebian Section-numbers in its margins in the Gospels, Codex Sinaiticus does – but in a somewhat mangled form. 

            This indicates that Eusebius of Caesarea was not involved in the production of Codex Sinaiticus, because it is extremely unlikely that he would have allowed his own cross-reference system to be presented so carelessly.  At the same time, as the place where Eusebius was bishop until his death, Caesarea was one of the first places where the Eusebian Canons were used.

In addition, there are several clues embedded in the text of Codex Sinaiticus that suggest that it was made at Caesarea, during the time when Acacius, an Arian, was bishop there.  I think it is very probable that this is when and where it was made.

Details about the origin of Codex Sinaiticus, and the quality of its text, have tended to be overshadowed by stories about its discovery in the 1800s by Constantine Tischendorf at Saint Catherine’s Monastery on Mount Sinai.  In this lecture I will not go into detail about all that, except to say, first, that the most generous interpretation of Tischendorf’s account of his first encounter with pages from Codex Sinaiticus is that he did not understand what he was being shown and what he was being told, and, second, all of the pages that Tischendorf took should be returned to the monastery from which they came.   

Codex Sinaiticus has a secondary set of section-numbers in its margin in Acts that is, for the most part, shared by Codex Vaticanus.  This indicates that when these numbers were added, probably in the 600s, these two manuscripts were at the same place.

Codex Sinaiticus has its own website, CodexSinaiticus.org , and there one can find not only good photographs of the manuscript but also some interesting information about its background and how it was made.

● Next is Codex Alexandrinus.  This codex, from the early 400s, has undergone significant damage:  it is missing the first 24 chapters of Matthew.  The surviving Gospels-text of Alexandrinus is particularly important because it tends to support the Byzantine Text, unlike Vaticanus and Sinaiticus.  In Acts and the Epistles, its text agrees much more often with the two flagship Alexandrian codices, but this is a tendency, definitely not a two-peas-in-a-pod level of agreement.  For Revelation, Codex A is the best manuscript we have. The entire New Testament portion of Codex Alexandrinus can be viewed page by page at the British Library’s website.     

● The worst Greek manuscript we have is Codex Bezae, a diglot manuscript, with alternating pages in Latin, which was produced in the 400s.  It has undergone some damage, but it still contains most of the four Gospels, in the order Matthew – John – Luke – Mark, part of Third John in Latin, and most of the book of Acts.

More important than its production-date is the date of the readings that it supports:  many of them are supported by Old Latin witnesses, and by early patristic writers who used what is called the Western Text. 

 The high level of textual corruption in Codex Bezae makes the text found in relatively young manuscripts look excellent in comparison.  Codex D’s text demonstrates that what really matters is not the age of a manuscript, as much as how well the copyists in the transmission-line of a manuscript did their job. 

Once one comes to terms with the awful quality of Codex Bezae’s text, though, many of its readings are awfully interesting.  It echoes a time in the text’s history when copyists prioritized conveying the meaning of the text – or what they thought was its meaning – above the form of the text found in their exemplars.

 Codex Bezae can be viewed online page by page at the University of Cambridge’s Digital Library.

● Also from the 400s, and probably earlier than Codex Bezae, is Codex Washingtonianus.  Codex W was acquired by the American businessman Charles Freer in 1906.  It is the most important Greek Gospels-manuscript in the United States.  Part of what makes Codex W important is not only its age, but its attestation to different forms of the text collected in a single volume:  its text in Matthew is strongly Byzantine.  Its text in Mark 1:1 to Mark 1-5 is similar to Western Text.  Its text in the rest of Mark tends to agree with the surviving text of Papyrus 45, at least in the parts where P45 is extant.   In Luke, up to chapter 8, its text is Alexandrian, but the rest of Luke tends to agree with the Byzantine Text.  In the first four chapters of John, Codex W has supplemental pages, copied from a different exemplar than the rest.  In the rest of John, it tends to agree with the Alexandrian Text.

This has led some researchers to suspect that although most of Codex W appears to have been made in the 400s, it may be a copy of an earlier codex that was based on exemplars that had been partly destroyed in the Diocletian persecution, in the very early 300s, just before Codex Vaticanus was made.  Page-views of Codex W can be accessed at the website of the Center for the Study of New Testament Manuscripts.

● Codex Ephraemi Syri Rescriptus, also known as Codex C, is a palimpsest.  Its surviving pages contain text from almost every book of the New Testament, as well as pages from some of the books of Poetry in the Old Testament, and two apocryphal books.  It was made some time in the 400s.  Its text is somewhat Alexandrian, with significant Byzantine mixture.  It is one of the few Greek manuscripts that support the reading “six hundred and sixteen” as the number of the beast in Revelation 13:18.

The parchment of Codex C was recycled to provide material on which some of the works of Ephraem the Syrian were written; this accounts for the name of the manuscript.  Its Biblical text was established in the 1840s, after much effort, by Constantine Tischendorf, the same individual who brought Codex Sinaiticus to the attention of European scholars.   The text has undergone extensive correction.

● 0176 is a fragment, probably produced in the 400s, that contains text from Galatians 3:16-24.  This manuscript was excavated from Oxyrhynchus, Egypt, which is somewhat intriguing, because the text of this fragment is thoroughly Byzantine, not Alexandrian.

● The Purple Triplets is my pet name for three uncial manuscripts from the mid-500s:  Codex N, Codex O, and Codex Σ.      Codex N is also known as 022, Codex Petropolitanus Purpureus.  Codex O is also known as 023, Codex Sinopensis.  It contains text from the Gospel of Matthew.  And Codex Σ is also known as 042, Codex Purpureus Rossanensis, or the Rossano Gospels.  It contains text from Matthew and Mark.

            These are not the only Greek uncial manuscripts written on purple parchment.  What is especially interesting about these three is that they are related to each other like siblings, copies of the same master-copy.  Codex Σ is known not only for its mainly Byzantine text, but also for its illustrations, which can be viewed at http://www.codexrossanensis.it .

● Codex Regius, also known as Codex L, contains most of the text of the four Gospels.  It was probably made in the 700s, probably by an Egyptian copyist.  Codex L is one of six Greek manuscripts that attest to both the Shorter Ending and the Longer Ending of Mark.  Codex L also has a large distinct blank space in the Gospel of John where most manuscripts have John 7:53-8:11, the story of the adulteress.    

● Codex Pi, also known as Codex Petropolitanus, is a Gospels-manuscript assigned to the 800s.  Its text is a very early form of the Byzantine Text.

● Codex K, also known as Codex Cyprius, is another Gospels-manuscript that was also probably produced in the 800s.  In the first 20 verses of the Sermon on the Mount in Matthew 5, compared to the text in Codex Sinaiticus, the text of Codex Cyprius is much closer to the original text.

● Minuscule 2474, the Elfleda Bond Goodspeed Gospels, from the 900s, contains an example of the text of the Gospels that dominated Greek manuscript-production in the Byzantine Empire.  This manuscript can be viewed page by page at the website of the Goodspeed Manuscript Library of the University of Chicago.

            There are also several clusters, or groups, of manuscripts, that share readings that indicate that they share the same general line of descent: 

            In the Gospels, the text of some members of a group of manuscripts that display a note called the Jerusalem Colophon is above average importance.

            Readings shared by the main members of Family 1 in the Gospels, best represented by minuscule manuscripts 1, 1582, and 2193, probably echo an ancestor-manuscript from the 400s.

            Members of Family 13 in the Gospels tend to echo an ancestor-manuscript with many reading that diverge from the Byzantine standard.

            Also, in the General Epistles, members of the Harklean Group echo a form of the text that has some unusual readings that are earlier than Codex Sinaiticus.

            Some other minuscules, such as minuscules 6, 157, 700, 892, and 1739, are as important as some of the uncials.  Their existence should remind us that when we ask how much weight ought to be given to a particular manuscript, the primary consideration should not be “How old is it”, but “How well did the copyists in its transmission-stream do their job?”. 

            No manuscript sprang into being out of nothing, and any manuscript, early or late, if it is independent from another known manuscript, has the potential to contribute something to a reconstruction of the text of the New Testament.

             To get some idea of the appearance of New Testament manuscripts, I encourage you to explore the online presentations of manuscripts at the following institutions:

            Saint Catherine’s Monastery at Mount Sinai, the Vatican Library, the British Library, the National Library of France, the Walters Art Museum, the Goodspeed Manuscript Collection at the University of Chicago, the Kenneth W. Clark Collection at Duke University, and the  Center for the Study of New Testament Manuscripts.

    

John 6:11 - An Instructive Variant-unit

           In John 6:11, as John presents his account of the Feeding of the Five Thousand, an instructive textual variant appears:  where the KJV and NKJV say that Jesus, after giving thanks, gave the pieces of fish to the disciples, and the disciples gave to those who were seated, the EHV and ESV, NIV, CSB, etc., do not specifically mention the disciples in this verse about the distribution of the pieces of fish. 

            Textually, the difference between the Greek text of the KJV and NKJV, versus the Greek base-text of the other versions, is uncomplicated:  the Textus Receptus, supported by a large majority of manuscripts, has, after the word διέδωκεν (“gave”), the words τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ (“to the disciples, and the disciples”). 

            Theoretically, these words could be lost by accident:  they begin with exactly the same word that follows them (τοῖς), and a copyist who skipped from one τοῖς to the next τοῖς might not realize his mistake.  On the other hand, these words could be deliberately inserted in order to bring John’s account into closer harmony with the other Gospel-writers’ descriptions of the same event: 

            ● Matthew 14:19 says that after Jesus gave thanks for the bread, He broke it and gave to the disciples the bread (τοῖς μαθηταῖς τοὺς ἄρτους), and the disciples (οἱ δὲ μαθηταὶ) to the crowd (τοῖς ὄχλοις).

            ● Mark 6:41 says that after Jesus broke the bread, He gave (ἐδίδου) to the disciples (τοῖς μαθηταῖς)

            ● Luke 9:16 says that Jesus gave (ἐδίδου) the broken bread and fish to the disciples (τοῖς μαθηταῖς)

            The flagship manuscripts of the Alexandrian text, beginning with Papyrus 28, Papyrus 66, Papyrus 75, À (Codex Sinaiticus), and B (Codex Vaticanus) weigh in for non-inclusion of τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ.  Their allies in this case include Codex A (fol. 46r), L, N, W, Π, and the Curetonian Syriac, the Peshitta, the Gothic version, and the Armenian version.   This is an interesting array of versional support, because the Peshitta and the Gothic version tend to favor the Byzantine Text in the Gospels.

            Wieland Willker’s Textual Commentary on the Greek Gospels gives an overview of the relevant external evidence.  Greek support for τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ includes D (Codex Bezae), S, Δ, Θ, Ψ, and versional allies include the Sinaitic Syriac, Lycopolitan John (the Qau Codex), and some Old Latin manuscripts.

            Here is an additional sample of the large majority of manuscripts that support the Byzantine reading:  Codex G (011), Codex Y (034), 72, 115, 116, 157 (with οχλοις (“the crowd”) instead of τοῖς ἀνακειμένοις (“those who were seated”)), 178 , 484, 490, 505, 537, 715, 699, 714, 796, 892 (with αυτου (“His”) describing the disciples), 1186, 1187, 1200, 1209, 1318, 1478, 2612, 2099, and 2290.   

            In addition, two witnesses have notable features:

            ● In GA 1346 the main text does not include τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ but in the outer margin, partly trimmed away, the words have been written, and in the margin and in the text where the correction was intended to appear, the ⁜ symbol appears.

           ● In GA 775 the main text did not include τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ but a corrector has erased part of the shorter ending and replaced it with the longer reading.

            If representatives of the Alexandrian text stood alone, it would be very tempting to reckon that an early Alexandrian copyist accidentally lost τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ via a parableptic error when his line of sight drifted from τοῖς to τοῖς.  However, representatives of the early core of the Byzantine Text – representatives of Family Π – are allied with Codex A, Codex N (a purple uncial), and distinctly non-Alexandrian versional evidence in support of the non-inclusion of τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ:  114, 265, 389, 1079, 1188 , 1219, 2615, 2291(with συν before ἀνακειμένοις), and 2757 support the non-inclusion of τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ. 

            It may be plausible to posit one parableptic error, but not the same parableptic error in multiple transmission-lines – which is what is required to explain the lack of support for τοῖς μαθηταῖς, οἱ δὲ μαθηταὶ in (1) the Peshitta, (2) in the Vulgate, (3) in Gothic, (4) in the P75-B-L-Sah cluster, and (5) in members of Family Π, if these words are original in John 6:11.

            What has happened – and it must have happened fairly early to be echoed in witnesses as diverse as D, the Sinaitic Syriac, and the Qau Codex – is a very common kind of textual variant in the Gospels:  a harmonization.  This introduction of words from Matthew 14:19 into John 6:11 was probably intended to guide readers to an accurate understanding of the scene, so that they would not misunderstand John’s words to mean that Jesus gave the bread Himself without the agency of the disciples.  It may be considered a benign harmonization – even a pastorally motivated one – but it would have been better to let John’s own words stand as they were.

            This variant-unit shows that a simple appeal to the majority of manuscripts is not a safe path to the original form of the text.  It also shows that the early production-date of a witness is no guarantee that its text is better than that of a much later manuscript (such as 114). 

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

             

Mary, Martha, and John 11

            Back in 2016, an interesting text-critical thesis was proposed in Harvard Theological Review:   unusual readings in Papyrus 66, considered alongside textual variants in many other manuscripts, indicate that the character of Martha did not originally appear in the Gospel of John; she was inserted by a later writer who understood Mary of Bethany and Mary Magdalene to be the same person, and who wished to diminish the role of Mary Magdalene.

            Lately this theory has been getting some attention;  in 2018, Candida Moss (of Notre Dame University) concluded an article about it by stating, “for the first time there is a plausible scholarly argument for the idea that Mary Magdalene was written out of the Bible and the history books.”  And in July of 2019, Elizabeth Schrader, the thesis-writer, made an appearance at the Religion for Breakfast video-show, promoting the theory.

            Is Schrader’s main idea plausible, or has she misread the evidence?  She has misread the evidence, mainly by consistently misinterpreting scribal errors as if they have implications that they simply do not have.  This may be concisely demonstrated beyond a reasonable doubt by undertaking the challenge that Schrader issued near the end of her Religion for Breakfast interview:  to demonstrate that Martha is not an addition to the Gospel of John, one needs to do the following:
            ● explain the unusual readings in P66.

            ● explain why the names are always changing in John 11:5. 

            ● explain why there’s only one sister in so much early artwork.

            ● explain why there is not similar confusion involving the names of Mary and Martha in Luke 10:38-42.

            ● explain the reading of Codex Alexandrinus in John 11:1, where the name Mary is changed to Martha, and the verse ends by referring to his sister rather than her sister.

Let’s begin.

● THE UNUSUAL READINGS IN JOHN 11:1-3 IN PAPYRUS 66

            The copyist who transcribed the text of P66 was not particularly competent.  Occasionally, he got ahead of himself and over-anticipated the text he was copying, somewhat it the same way that a typist, upon encountering the phrase “The quick brown fox jumped” at the end of a page, might continue to type “over the lazy dogs,” without bothering to turn the page – only to find a different phrase after the page is turned. 

            In John 11:1, the copyist of P66 initially wrote the Greek equivalent of “Now a certain man was sick, Lazarus of Bethany, the village of Mary and Mary her sister.”  Then, realizing that he had written “Mary” twice,” he went back and corrected the text by erasing the letter iota in the second Μαριας and replacing it with the letter theta, so as to write Μαρθας.  This kind of mistake is not particularly unusual for this copyist; he made at least 15 other mistakes of dittography (writing twice what should be written once) in the text of John.  

            Apart from this careless one-letter mistake, the copyist of P66 initially wrote a normal text of John 11:1, identical to what is found in the Textus Receptus.  In verse 3, we find a reading in which, according to Schrader, “one named woman has been split into two unnamed women.”  After writing the Greek equivalent of “Sent, therefore,” (απεστιλεν ουν) the copyist initially wrote a name – either Μαριας  or Μαρθας – and continued on a little further, to the end of the line he was writing:  προς αυτον λεγουσα, that is, “to Him, saying.”  (Probably he also started the verse with και (and) and then declined to keep the word, but this does not figure into the subject at hand.)  At this point, the copyist of P66 realized that he had over-anticipated the text in his exemplar (perhaps when he finished writing λεγουσα, consulted his exemplar, and saw that it read λεγουσαι), went back, adjusted the endings of the verbs so to as to turn them into plurals (απεστιλαν and λεγουσαι), erased the name (which is why we’re not sure whether it was Μαριας  or Μαρθας , but I suspect it was Μαριας), and in the space where the name had been, wrote αι αδελφαι, that is, “the sisters.” 

            It could be said that one woman has been replaced with a reference to two woman – but to what extent is this saying anything more than that the copyist of P66 began verse 3 by assuming that it was about one woman, and then corrected his mistake?  If the presence of αι αδελφαι was the special property of an interpolated manuscript in the hands of the copyist of P66, then it certainly was well-travelled:  αι αδελφαι is the reading here in John 11:3 in Codex Vaticanus, and in Codex Sinaiticus; αι αδελφαι is the reading in Papyrus 45, and in Papyrus 75.  Likewise Origen, in his Commentary on John, VI:40, in the course of discussing a textual variant in John 1:28, mentions that John says that Bethany was the town of Lazarus, and of Martha and Mary.  If one consults Schrader’s data-tables in which the contents of many manuscripts are compared, it appears that αι αδελφαι is supported in every extant Greek manuscript in the list in which verse 3 appears – except P66, in which the copyist almost immediately fixed his mistake. 

            Schrader seems to consider problematic the inclusion of αυτης (her sister) at the end of John 11:1, arguing that the original text was αυτου.  However, by asserting that αυτου is the original reading, Schrader is arguing for a reading that originated as an expression of a tendency among some copyists (especially in Old Latin texts) to adjust the text in favor of the dominance of men – that is, in Codex A (from the 400s), 841 (from the 1400s), 1009 (from the 1200s), 1071 (from the 1100s), and in two medieval lectionaries, we see the effect of a scribal preference to refer to “his sister” instead of “her sister.”  In such a smattering of witnesses, the reading αυτου simply pops up; meanwhile in P66, P75, B, ℵ, K, L, M, S, W, Y, Δ, Θ, Π, 047, and so on, αυτης has ancient, abundant, and coherent support.

            To put it another way:  there is no genealogical connection between Codex A and the medieval minuscules 423, 841, 1009, 1071, and two lectionaries; the reading αυτου at the end of John 11:1 appears in these manuscripts not as something with ancient roots, but as something more like a weed that has sprouted from the minds of what a few copyists thought the text should say.              

           

            Before moving on to the next point, I should address a reading in the important medieval minuscule 157:  In John 11:1, the words και Μαρθας are absent.  Is this evidence that minuscule 157 echoes some ancient exemplar in which Martha did not appear in the narrative?  No; what has happened is that the preceding word Μαριας appears at the end of a line; the copyist lost his line of sight as he began the next line, shifting forward to the letters at the end of και Μαρθας.  Thus he accidentally skipped those two words – but their presence in his exemplar is obvious from the words that he wrote next:  της αδελφης αυτης (that is, her sister).  Schrader observes that 157 thus “nonsensically” applies a feminine pronoun to Lazarus, but it seems not to have registered that the obvious explanation of this nonsense-reading is that a simple scribal mistake has been made, rather than that a lost Martha-less form of John 11 is being attested.

● THE DIFFERENT ORDERS OF NAMES IN JOHN 11:5

            The text of John 11:5 in most Greek manuscripts says, “Now Jesus loved Martha and her sister and Lazarus.”  Whether one consults the UBS compilation or the Robinson-Pierpont Byzantine Textform, or Michael Holmes’ SBL-GNT, or even the Textus Receptus, they all agree:  ἠγάπα δὲ ὁ Ἰησοῦς τὴν Μάρθαν καὶ τὴν ἀδελφὴν αὐτῆς καὶ τὸν Λάζαρον.  The text of P66 is identical with the exception that Jesus’ name is contracted (as is typical in Greek manuscripts) and the word αὐτῆς is not in the text; however it is supplied in the margin.   

            An assortment of other manuscripts disagree, primarily because of two scribal tendencies:  (1)  the tendency to supply names, so as to make the text more explicit, and
(2)  the tendency to put Mary’s name first, so as to correspond to the order of names given when the characters are introduced in John 11:1.

            Under the influence of those two natural tendencies, some copyists rewrote the verse to say, “Now Jesus loved Mary and her sister Martha, and Lazarus.”  This may be considered the Caesarean form of the verse, attested in a special cluster of manuscripts (consisting mainly of Θ, f¹, f¹³, 543, 565, 828, and others), the members of which share other textual features, such as unusual placements of the pericope adulterae (John 7:53-8:11).  

            A few other manuscripts list Martha first, but add Mary’s name, so as to say, “Now Jesus loved Martha and her sister Mary, and Lazarus.”  Schrader lists two medieval manuscripts – 2561 and 2680 – which support the form, “Now Jesus loved Mary and her sister and Lazarus,” thus putting Mary’s name in the place formerly occupied by Martha’s name.  

            What is not seen in any of these Greek manuscripts is a form of the text in which Martha is entirely absent.  Even in the few relatively late manuscripts in which her name does not appear in 11:5, she is referred to as Mary’s sister.  When the rival readings are analyzed, from the more explicit to the less specific, and from those harmonized to 11:1 to those less harmonized, the anomalies are easily sorted out and the usual, ordinary text is confirmed, and the flow from more specific to less specific, and from more harmonized to less harmonized, is generally matched by the flow from the  younger to the older witnesses.  In other words, the consistent picture shown by Greek manuscripts in John 11:5 is that the insertion of Mary’s name, and the transposition of Mary’s name to the front of the list, and the loss of Martha’s name, are late scribal adjustments, not echoes of an ancient exemplar.

            Furthermore, it is not accurate to say that the names in John 11:5 are “always changing.”  The verse is altered in the Caesarean Text, i.e., in select members of f¹ and  f¹³.  But in most manuscripts (including P45, P75, ℵ, B, A, K, L, W) it is stable.  In all Greek manuscripts of John 11:5, the verse conveys that Jesus loved Martha and Mary and Lazarus, whether the names of all three individuals are supplied in this verse or not.

● THE RESURRECTION OF LAZARUS IN EARLY CHRISTIAN ARTWORK

            In Schrader’s thesis, there is very little emphasis on artwork; her appeal to artwork in the Religion for Breakfast interview may be something that was just thrown in.  Nevertheless, it may be briefly considered:  artwork is art, and the degree of detail provided in a work of art is subject to the whims, abilities, and resources of the artist.  Artists have creative freedom which copyists do not.  A depiction of the resurrection of Lazarus in the Catacomb of the Giordani shows only Jesus and Lazarus.  Similarly in a mosaic on the wall of the Basilica of Sant’Apollinare Nuovo, made around 530, a scene depicts the raising of Lazarus without any females present.  Likewise the Murano Diptych, from the 400s-500s, depicts Jesus and Lazarus, but no women.  And at the Museo Pio-Cristiano at the Vatican, a scene on a sarcophagus-lid from the cemetery of Saint Agnes depicts the resurrection of Lazarus, but without anyone except Jesus and Lazarus. 

               Should we therefore assume that the artists of these four early works of art knew a form of John 11 in which Mary and Martha (and the crowd of mourners accompanying them) do not appear?  (Meanwhile The Jonah Sarcophagus depicts two women present at the raising of Lazarus.)  I think the point is already clear:  it would be absurd to treat ancient artwork as a means to answer the question at hand.

● THE STABILITY OF NAMES IN LUKE 10:38-42

            Why, we are invited to ask, is there instability involving the names of Mary and Martha in John 11:1-12:2, but not in Luke 10:38-42, where the same two characters are depicted?  There are two very simple reasons why this is the case.  First, Luke 10:38-42 constitutes only five verses, in which Mary’s name appears twice and Martha’s name appears four times, and the two names never appear side-by-side; in contrast, John 11:1-12:2 constitutes 59 verses – or 46, if we exclude John 11:47-57, which is really a different scene – in which Mary’s name appears eight times and Martha’s name appears eight times, and both names appear in the same sentence twice (in v. 1 and v. 19).  The passage in Luke is one-eighth the length of the passage in John, and it provides very little opportunity to get the sisters’ names mixed up.

            The second reason is that while in John, each sister is described as a sister of Lazarus, and both sisters undertake similar actions (both say the same thing to Jesus, in John 11:21 and 11:32), in Luke their actions and attitudes form a stark contrast; Martha is busy, while Mary is sitting at Jesus’ feet.  It is impossible not to see this contrast in the episode in Luke; it forms the foundation of the lesson that is intended to be conveyed.  Meanwhile, in John, the two sisters are described similarly, and say similar things.  There is a stark contrast between them in Luke which precludes confusion of the two individuals, whereas in John there is not.  

A hypothetical reconstruction
of the uncorrected text of John 11:1
in Codex A.  The manuscript is online.

● THE TEXT OF JOHN 11:1 IN CODEX ALEXANDRINUS

            In Codex Alexandrinus, the text that stands in the manuscript now says, “Now a certain man was sick, Lazarus of Bethany, the village of Mary and Martha his sister,” differing from the usual text only by the presence of “his” (αυτου) instead of “her” (αυτης), a difference addressed already.  When the copyist initially wrote out this verse, however, he made another, more significant mistake (which was detected by the researcher Cowper in 1840). 

            Normally, the text of John 11:1 goes, Ἦν δέ τις ἀσθενῶν Λάζαρος ἀπὸ Βηθανίας, ἐκ τῆς κώμης Μαρίας καὶ Μάρθας τῆς ἀδελφῆς αὐτῆς – “Now a certain man was sick, Lazarus of Bethany, the village of Mary and her sister Martha.”  But besides shifting from αυτης to αυτου (and thus causing the text to refer to “his sister Martha”), the copyist of Codex A made a parableptic error, skipping from the letters –as at the end of Μαρίας to the same letters at the end of καὶ Μάρθας, thus skipping the two words καὶ Μάρθας.  A clever correction was made:  the word κώμης was erased, and then written in small letters at the end of the previous line, and the newly blank space was filled with the words Μαρίας καὶ Μάρθας. 

            That’s all there is to that scribal mistake and correction.  The other mistake in John 11:1 in Codex A – αυτου instead of αὐτῆς – was addressed in the first point.

            Although Schrader’s five-part challenge has been answered, there are two additional components of her thesis that I will address here. 

● IS “ΑΥΤΗ” IN JOHN 11:4 A DATIVE FEMININE SINGULAR?

            This point is slightly technical:  should the word αυτη in 11:4 be understood as if it was meant to signify the person Jesus was addressing – causing the sentence to begin, “Upon hearing that, Jesus said to her” – or should it be understood (as most English translations render it) instead as a nominative term, causing the sentence to begin, “Upon hearing that, Jesus said, “This,” and so forth.   The copyist of P66 put a comma-like mark before αυτη, as if he perceived that the text could initially seem ambiguous without it, and wished to ensure that readers would understand the αυτη to mean “This” instead of “to her.”

            And the copyist of P66 wasn’t the only scribe to do so.  Codex Sinaiticus has a separating mark between ειπεν and αυτη.  In Codex Vaticanus, ειπεν ends a line, and some empty space is leftover, before αυτη begins the next line.   Jumping ahead several hundred years, the Elfleda Bond Goodspeed Gospels displays a separating dot between ειπεν and αυτη.  I am confident that many other copies share this feature, so as to elicit the understanding that “This sickness” was the intended meaning.  In some other manuscripts, such as 138 and 1321, the risk of ambiguity has been eliminated by moving αυτη to the other side of ἡ ἀσθένεια. (Schrader lists a total of 12 Greek manuscripts with this reading, and seems to consider each one as somehow problematic, but this is simply a clarifying transposition.)   Following this clever adjustment, some copyists conflated both placements; as a result, seven manuscripts Schrader has examined have αυτη both before, and after, ἡ ἀσθένεια.  (These, too, are counted as problematic by Schrader.)

            The translator of the Latin text in the Old Latin Codex Carotensis (VL 33) seems to have been at a disadvantage; his Greek exemplar(s) apparently did not have distinction-making marks or separation-spaces in this verse, and due to this ambiguity, this manuscript has the phrase “dixit ei,” that is, “said to her,” in John 11:4.  This is a symptom of a Latin translator’s confusion, however; it does not indicate that αυτη was meant to be understood this way. 

● WHO SERVES SUPPER IN JOHN 12:2?

            In P66 – after Martha’s name has appeared in – John 12:2 begins not with the usual ἐποίησαν (“they made), but with the singular ἐποίησεν.  This is a very slight variation, probably elicited by a scribe’s desire to relieve readers of the burden of asking who “they” were; the resultant sense, with ἐποίησεν, is that Lazarus made a supper for Jesus.  Minuscules 295 and 841, Schrader has observed, share this reading.

            A little further along in the verse, P66 says that Martha served.  This is the reading of almost all manuscripts, whether early or late – but – but Schrader has observed that minuscules 27, 63, and 1194 have Mary’s name here, instead of Martha’s.  I leave it to readers to mull over the probabilities:  is this a simple effect of scribal inattentiveness, sparked by anticipation of Mary’s actions in the following verse, or do three Byzantine minuscules preserve the original reading, against all other Greek manuscripts?

IN CONCLUSION . . .

            There is more material in Schrader’s thesis that I have not considered in this brief essay.  However, the major points have been covered, and her five-point challenge has been answered.  Although Schrader has collected many variant-readings in John 11 (which must have taken considerable work), a very large majority of the readings in question, and especially the variants at the core of her arguments, are the effects of scribal carelessness, or the effects of scribes’ desire to augment the clarity the text.   

            This tends to hollow out her claim that one in five of the Greek manuscripts she has examined displays some problem involving the character of Martha in John 11:1-12:2; the evidence points toward a different and unremarkable direction:  copyists were sometimes careless, and sometimes desired to augment the clarity of the text.  None of these textual variants suggests anything remotely resembling the massive interpolation that Schrader has proposed.



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

[In the first posting of this post, I had the words απεστιλεν and απεστιλαν mixed up.  My bad.]