Cherry-picking in Edinburgh

            Have you ever been misled by “cherry-picking”?   I bought a new car last week! – a new Hot Wheels car.  I won half a million dollars yesterday! – in a game of Monopoly.  Details matter, and the omission of important details can result in the spread of false impressions.

            Teachers and commentators who describe evidence very selectively risk giving false impressions to their students and readers.  This is unfortunately a frequent phenomenon when it comes to the way the evidence pertaining to John 7:53-8:11 and Mark 16:9-20 is described,   The result:  students leave the classroom, or readers leave the commentary, with a thoroughly distorted picture of the evidence.  It’s not that anyone has lied to them.  They simply have not been told the whole story.  Consider an example:  the recent descriptions of evidence pertaining to those two passages made by Dr. Larry Hurtado of the University of Edinburgh.  Dr, Hurtado is a distinguished professor with impressive credentials; surely he can be trusted to describe text-critical evidence objectively and accurately and with only the mildest of bias, right?  Well let’s see: 

            In a recent blog-post titled More on Rethinking the Textual Transmission of the Gospels, Dr. Hurtado claimed that John 7:53-8:11 “first appears in the extant manuscripts in the fifth century.”  Technically, it is true that we have no manuscripts made before the 400s in which the passage appears,  just as it is technically true that I recently bought a new car and won half a million dollars.  But the impression that that statement gives – that the passage did not began to occupy that location in the Gospels until the 400s – is false. 

            The risk of conveying such a false impression could have been avoided if Dr. Hurtado had shared just one more bit of evidence:  Jerome’s testimony that he had found the story of the adulteress in many manuscripts, both Greek and Latin.  Or, if Jerome is too obscure an author to be considered worth mentioning, perhaps the testimony of Ambrose would have been sufficient.  

            A writer resorting to less cherry-picking might inform readers and students about the different types of early Latin breves, or chapter-summaries, which refer to the story about the adulteress in its usual place in the Gospel of John – including Type I (generally regarded as contemporary with Ambrose, and with Zeno of Verona) and Type Cy; the “Cy” stands for Cyprian, the prominent author and bishop in the 200s; this form of the breves has been assigned to the time of Cyprian or slightly later.  If the composition-dates that have been given to these chapter-summaries are correct, then their testimony implies that the pericope adulterae was in Latin copies of the Gospel of John in the 200s.   

            When these pieces of evidence are added to the equation, though, there is a cost:  the narrative in which John 7:53-8:11 doesn’t show up until the fifth century crumbles to pieces.  A wider, fuller view of the evidence does not support Dr. Hurtado’s contention that this passage became part of the text of the Gospel of John “not in some early “wild” period, but later, in the period of supposed textual stability.”           

            More cherry-picking is in Dr. Hurtado’s description of evidence pertaining to Mark 16:9-20.  “The first Greek manuscripts that allow us to check the matter are Codex Vaticanus and Codex Sinaiticus, which don’t have these verses.”  That is technically true, but why has Dr. Hurtado mentioned these two manscripts from the fourth century without mentioning the much earlier testimony of Tatian and Irenaeus?  Tatian incorporated the passage into his Diatessaron in the 170s, and Irenaeus specifically quoted Mark 16:19 from the Gospel of Mark, in Against Heresies Book 3, chapter 10, around the year 180.  Here we have two pieces of evidence, both well over a century older than Vaticanus and Sinaiticus.  Why are they hidden from view?  Some of Dr. Hurtado’s readers might imagine that the testimony of Tatian and Irenaeus has been avoided because if their testimony were given a spotlight, it would be extremely difficult to convince anyone that the picture that Dr. Hurtado has painted of the history of Mark 16:9-20 can be plausibly maintained. 

Codex Vaticanus, with Mark 16:9-20
added in the space that appears
in the manuscript after 16:8.

            And why did Dr. Hurtado mention Vaticanus without also mentioning its special blank space after Mark 16:8?  Why did Dr. Hurtado mention Sinaiticus without mentioning that the last part of Mark and the first part of Luke occupy a cancel-sheet – that is, four replacement-pages, on which the lettering has unusual features that indicate the copyist’s awareness of the absent verses?  Again, students and readers might be forgiven for imagining that such information has not been shared because it makes Dr. Hurtado’s theory appear contrived.

            Finally, why did Dr. Hurtado describe Mark 16:9-20 as part of “the Medieval text of Mark,” instead of “The Second-Century Text of Mark,” in light of the testimony of Tatian and Irenaeus?  (And the testimony of Apostolic Constitutions and Ambrose and Augustine and Macarius Magnes and Marcus Eremita and some others who wrote in the time of the Roman Empire).  How is that not just spin?  Spin is exactly what it is.  

            Dr. Hurtado stated, “To find the variant in the manuscript tradition we have to go to later, to the fifth/sixth century, in Codex Alexandrinus, Codex Bezae, and others.”  (As if the presence of Mark 16:9-20 in Codices A, D, W, the Vulgate, and the Peshitta does not imply a much earlier ancestry.)  But we have more evidence besides just manuscripts; why would anyone put on blinders by ignoring the widespread patristic evidence that demonstrates that Mark 16:9-20 was in widespread use in the early centuries of Christianity?  Why point out the testimony from Sinaiticus (c. 350) without mentioning the testimony of Aphrahat (337)?  It might seem to some readers and students that a lot of evidence that is inconvenient for their professor’s proposal has not been presented – at least, it would seem so, if somehow they were to learn about that evidence’s existence.

            A third variant was mentioned by Dr. Hurtado, and I will mention it just for the sake of thoroughness.  The scenarios involving John 7:53-8:11 and Mark 16:9-20 are nothing like the scenario involving Dr. Hurtado’s third variant, the Comma Johanneum.  Its adoption in the Textus Receptus was the result of Erasmus’ statement (after he had compiled the Greek text without the Comma) that if he had possessed a Greek manuscript with the passage, he would have included it), plus two other things:  Erasmus’ desire to make another edition, and the premiere of Codex Montfortianus.  This is no more like the situation regarding Mark 16:9-20 – a passage with second-century patristic support, and which is included in over 99% of the Greek manuscript of Mark – than Barney Fife is like the Incredible Hulk.    

            If you want to be taught about the transmission of the early New Testament text in a way that treats the evidence fairly, without having your professor mold the evidence, and pick and choose which evidence gets a megaphone, and which evidence is silenced – my impression is that you won’t find what you’re looking for at the University of Edinburgh.  Asbury Theological Seminary isn’t a good option either.

            One more thing:  Dr. Hurtado recommended Bruce Metzger’s Textual Commentary to those who want more information about the variants he mentioned.  Let it be noted that Metzger’s Textual Commentary contains misleading statements about Mark 16:9-20.  Also, dislocations of John 7:53-8:11 occurred due to the influence of lection-cycles, not (contra Metzger, Wallace, White, et al) due to the untenable idea that the pericope adulterae was a “floating” text (a theory which has been tested, and dismantled).

                     One more one more thing:  from now till Christmas, upon the request of any student at the University of Edinburgh and Asbury Theological Seminary, I will gladly send a digital copy of my research-books, Authentic:  The Case for Mark 16:9-20, and A Fresh Analysis of John 7:53-8:11, free of charge. 



Readers are invited to explore the embedded links in this post for additional resources.

The Pericope Adulterae and Some Early Manuscripts

          The Greek manuscripts which are often cited as the primary external evidence against the pericope adulterae (John 7:53-8:11 – a passage which some evangelical seminary professors and influential preachers, including John Piper, do not regard as Scripture) are Papyrus 66, Papyrus 75, À (01, Sinaiticus), B (03, Vaticanus), A (02, Alexandrinus), C (04, Ephraemi Rescriptus), L (019, Regius), N (022, Petropolitanus Purpureus), W (032, Washingtoniensis), and Δ (038, Sangallensis).  None of these manuscripts has John 7:53-8:11 between 7:52 and 8:12.  However, the testimony of some of these witnesses is significantly nuanced by additional details.

Codex Delta's blank space.
Color page-views are online.

          For example, in Codex Δ (from the 800’s), the copyist provided a clear indication of his recollection of the passage, even though it was absent from his exemplar.  After John 7:52, the copyist wrote the first seven words of 8:12, but then left the rest of the page blank, and resumed writing after leaving three additional blank lines on the following page.  Then he restarted the text of 8:12, and proceeded on from there.  Thus, while Codex Δ attests to the absence of the pericope adulterae in its exemplar, it also attests to the copyist’s memory of the presence of the passage in some other manuscript.

          Similarly, in Codex L (from the 700’s), the copyist left a long blank space between the end of John 7:52, on one page, and the beginning of 8:12, on the following page.  This blank space in Codex L includes more than an entire blank column.  In codices Δ and L, the blank space is not sufficient to include John 7:53-8:11, but the copyists’ intention to leave “memorial space,” acknowledging their awareness of the absent passage, remains obvious.  It therefore seems somewhat selective when commentators such as Metzger, Wallace, and White (among others) mention the absence of John 7:53-8:11, but fail to mention these blank spaces, of equal age, which attest to the presence of the passage in the memories of these manuscripts’ scribes.

Codex Regius' blank space.

          Before we turn to some other interesting features in these manuscripts, it should be pointed out that of the 1,476 manuscripts that contain the pericope adulterae, about 60 manuscripts have it in a location other than between John 7:52 and 8:12.  One particular group of manuscripts, which includes the important minuscules 1 and 1582, has the passage after the end of the Gospel of John, preceded by a note stating that because most manuscripts did not contain the passage, and because it was not commented upon by venerable patristic writers (such as John Chrysostom), it was moved to the end of the book, having been previously found after John 7:52 (the end of which the annotator quotes). 

          Although the minuscules that contain John 7:53-8:11 after John are not particularly early, their agreements are considered to echo an ancestor-manuscript which was produced in the 400’s.  In addition, their distinct readings tend to have an affinity with readings used by Origen, a patristic author who died in 254. 

          The transfer of John 7:53-8:11 from the usual place in chapters 7 and 8 to the end of the Gospel of John thus did not begin when these manuscripts were produced, but centuries earlier, when their shared ancestor was made.  This raises a question:  could some of the manuscripts which have been cited against the pericope adulterae, and which do not have it in chapters 7 and 8, have had it at the end of the Gospel of John? 
          In the case of Codex N (from the 500’s), there is no way to verify if it contained the pericope adulterae after the end of John or not, because the manuscript is damaged; the last extant bit of John 21 is in verse 20, and so there is no way to know if 21:25 was followed by the pericope adulterae when Codex N was in pristine condition or not.

          In Codex W (from about the year 400), the Gospels are arranged in the order Matthew-John-Luke-Mark.  Commentator Wieland Willker has noticed that between the end of John and the beginning of Luke, there is a blank page – blank on both sides.  No such similar feature exists in Codex W between Matthew and John, or between Luke and Mark.  This might be an attempt, by a copyist aware of the existence of the pericope adulterae, to provide space where it could be added.    

          In Codex A (from the early 400’s), the pages containing the text from John 6:50 to 8:52a have been lost.  Thus, we cannot see directly that in Codex A, John 7:52 was followed by 8:12; we have to rely on space-calculations.  Here, again, Willker’s commentary is very helpful:  he notes that the copyist accidentally omitted John 8:52, when his line of sight wandered too far down the page.  When this is accounted for, a reconstruction of the missing text, without John 7:53-8:11, fits the space that would have been on the absent pages, whereas if the pericope adulterae had been present, the space would not be remotely sufficient. 

Codex Alexandrinus -
The end of John,
and a blank column.

          At the end of the Gospel of John, the copyist of Codex A put the closing-title of the book at the end of the first column on the page.  The second column is completely blank.  One might argue that this is to be expected at the end of the Gospels – yet, at the end of Acts, there is no similar blank column; the column in which the book of Acts ends is followed immediately by a column in which the Epistle of James begins.  On the other hand, between the end of Philemon and the beginning of Revelation, there are two blank columns, that is, one side of the page is blank.  There is little way to discern, from this evidence alone, if the blank column at the end of John in Codex A is filler-space, or memorial-space.

          In Codex Sinaiticus (from the mid-300’s), after the Gospel of John concludes in the fourth column of a page, the next four columns are completely blank.  Once again, while it is probable that this is simply filler-space, it is not impossible that this feature represents copyists’ recollection of the presence of the pericope adulterae at the end of the Gospel of John. 

          In Codex Vaticanus (from the early 300’s), John 7:53-8:11 does not appear after 7:52, and there is no unusual blank space after the end of John (on page 1382 of the codex) – just the usual leftover space below the end of the book.  However, in the outer margin alongside that blank space after John 21:25, there is an interesting feature:  an umlaut, also known as a distigme.    Very many of these symbols appear in the margins of the New Testament books in Codex Vaticanus; researcher Philip Payne brought them to the attention of his fellow-researchers in the late 1990’s and early 2000’s, and established that they were added to denote the locations of textual variants. 

          It has not been determined beyond reasonable doubt that these symbols were added in ancient times.  Payne has contended that most of the distigmai are contemporary with the production of the manuscript; Peter Head, however, has challenged this position.  My own suspicion is that these marks are all late.  However, because some scholars, including Daniel Wallace, have treated them as if they are ancient, let’s consider their possible significance in the case at hand:  the only known textual variant that would elicit the addition of a distigme in the blank space after the end of the Gospel of John is the presence of the pericope adulterae.   

          If the distigmai are as ancient as the manuscript itself, Codex Vaticanus testifies to a fourth-century copyist’s awareness of the pericope adulterae’s presence at that location in at least one manuscript older than Codex Vaticanus itself.  This would imply that the transfer of the passage to the end of John 21 was not initially due to its lack of use by Chrysostom and other patristic writers, but was caused by some other factor.

          Papyrus 75 (usually assigned a production-date in the early 200’s) is only extant in John up to 15:10, so there is no way to tell whether or not the pericope adulterae was present after the end of chapter 21.          

Papyrus 66 -
not much remains
of John 21:17ff. except
the page-number.

          Similarly, Papyrus 66 (which Robert Waltz describes as “a notably inaccurate copy”), also from the 200’s, is very fragmentary in John 21, and no text can be confidently reconstructed beyond 21:17.  Thus we cannot tell with certainty that Papyrus 66 did not contain the pericope adulterae after John 21. 

          A few things should be clear from this review of the major early witnesses for the non-inclusion of the pericope adulterae.

● First:  the evidence strongly supports the view that the text of John used in Egypt in the 200’s did not contain the passage after John 7:52.  

● Second:  codices L and Δ should be considered witnesses for non-inclusion and for inclusion.  

● Third:  the testimony of most of the major Greek manuscripts that support the non-inclusion of the pericope adulterae in chapters 7 and 8 is not nearly as clear or one-sided when they are asked to testify about the passage’s presence or absence following John 21; on this question, most of the early Greek manuscript-evidence is open to interpretation.   

Codex Vaticanus and the Ending of Mark

            “Some of the earliest manuscripts do not include 16:9-20.”   So says a bracketed heading-note in the English Standard Version.  The number of Greek manuscripts in which the text stops at the end of 16:8 is three.

            Minuscule 304 is one of those three.  It is a medieval commentary-manuscript in which the text of Matthew and Mark is interspersed with commentary-material.  Its text in Mark is essentially Byzantine and the manuscript appears to have undergone some damage near the end.  There’s nothing about 304 that would suggest that it has more weight than any other medieval manuscript. 

            The other two manuscripts in which Mark’s text stops at 16:8 are another story:  Codex Vaticanus (produced c. 325) and Codex Sinaiticus (produced c. 350) are the oldest and second-oldest Greek manuscripts of Mark 16.  (These two manuscripts are not the earliest evidence pertaining to the ending of Mark, just the earliest manuscripts.  Patristic writers in the 100’s, 200’s, and early 300’s utilized the contents of Mark 16:9-20, but the manuscripts used by those writers are not extant.)  I have previously described the unusual features in Codex Sinaiticus involving the ending of Mark.  Today, let’s examine the last page of Mark in Codex Vaticanus. 

The last page of Mark in Codex Vaticanus.

          The text on this page begins in 15:43, and ends at the end of 16:8, on the 31^st line of the second column.  The closing book-title appears a little further down the second column.  The third column is completely blank.  It was normal for copyists to begin books at the tops of columns, and thus some space was typically left below the end of each book before the next book began at the top of the next column (except in those cases where the book happened to conclude right at the end of a column).  It was not normal, however, for the copyist of Vaticanus to leave an entire column blank.  This is the only blank column in the New Testament portion of Codex Vaticanus.

           Of course this raises a question:  why did the copyist leave this column blank?  The obvious answer is that the copyist was aware of copies that contained verses 9-20, and although his exemplar lacked these verses, he left space to give the eventual owner of the manuscript the option of including them in the event that another exemplar was available.

            The blank space is not quite adequate to include verses 9-20.  If one were to erase the closing-title and write the contents of verses 9-20, beginning at the end of v. 8, using the copyist’s normal handwriting, there would still be four lines of text yet to be written when one reached the end of the last line of the third column.  It is perhaps for this reason that Daniel Wallace, referring to this blank space in his chapter of the 2008 book, Perspectives on the Ending of Mark, has said, “The gap is clearly too small to allow for the LE.”  In the same book, Maurice Robinson affirmed, “The space is insufficient to contain the entire LE.”  Their co-author J. K. Elliott stated less definitively, “Vaticanus actually contains a blank column after 16:8 that could possibly contain verses 9-20, suggesting that its scribe was aware of the existence of the longer reading.”

The last page of Mark in Codex Vaticanus with verses 9-20
in the blank space after v. 8, using cut-and-pasted characters
from Mark 15:43-16:8 on the same page.  

           If a copyist were to resort to compacted lettering – the script that the copyist of Sinaiticus used in the first six columns of the text of Luke – then the blank space is practically an exact fit.  In a reconstruction of Mark 16:9-20 (shown here) in the blank space after 16:8, using characters that were written elsewhere on the page in Mark 15:43-16:8, verse 20 concludes on the next-to-last line of the third column.     

            Although the implication that the copyist of Codex Vaticanus clearly recollected 9-20 when he wrote the text of Mark 16:1-8 from an exemplar that did not have verses 9-20, Daniel Wallace has proposed a different explanation, namely, that the copyist was using an exemplar in which the Gospels, though containing the Alexandrian text, were arranged in the Western order (Mt-Jn-Mk-Lk), and although the copyist rearranged the Gospels into the order Mt-Mk-Lk-Jn, he added a blank space to represent the blank space at the end of his exemplar.  This theory seems like the result of a determined effort to dismiss the obvious implication of the blank column.  Where is the evidence that the Alexandrian form of the Gospels-text was ever anything but in the order Mt-Mk-Lk-Jn?  (In Papyrus 75, John follows Luke.)  And why would any copyist regard the blank space at the end of an exemplar as a feature worth replicating?  Any manuscript, unless its text happened to end at the end of its last column, would contain some blank space at the end.   And why would a copyist replicate such a blank space, but not the order of the books?  And why would a copyist consider such a blank column worth replicating, but not add a blank column between John and Acts, or between Acts and James, or between Jude and Romans? 

            In addition to the contrived idea of an Alexandrian Gospels-exemplar with the Gospels arranged in the Western order with blank space at the end which the copyist wished to replicate, Wallace has questioned the significance of the blank space by pointing out that there are three large blank spaces in the Old Testament portion of Codex Vaticanus.  However, all three of those blank spaces are accounted for by special factors:

● One of these blank spaces is the space between the end of the Old Testament and the beginning of the New Testament; the last page of the Old Testament portion concludes with the apocryphal text of Bel and the Dragon, incorporated into the Septuagint’s text of Daniel.  To expect the Gospel of Matthew to begin in the next column would be a preposterous expectation. 

● One of these blank spaces occurs at the end of Second Esdras, before the beginning of the book of Psalms.  Only two lines of text are placed in the first column of the last page of Second Esdras, and after the closing-title (and what appears to be the signature of someone named Klement, possibly a former owner of the codex), the rest of the page is blank.  But the reason for this is obvious:  the book of Psalms begins on the very next page, and the text of Psalms is formatted in two columns, rather than three.  It was absolutely necessary to begin Psalms on a new page, due to the difference in the number of columns on the page.

● One of these blank spaces occurs between the end of the book of Tobit and the beginning of the book of Hosea.  The text of Tobit concludes in approximately the middle of the second column of a page, and the third column is blank.  Wallace claimed that “The gap at the end of Tobit lacks sufficient explanation.”  However, the explanation becomes obvious upon close examination.

            One copyist’s work ended at the end of Tobit, and another copyist’s work begins with the Prophetic Books, which begin with the Minor Prophets, which begin with Hosea.  At this point where one copyist’s work was connected to another copyist’s work, what we have after the end of Tobit is simply leftover space.  This should become very obvious when we notice that the leftover space after the end of Tobit did not initially consist of just the remainder of the page.   As Dirk Jongkind mentioned on page 31 of Scribal Habits of Codex Sinaiticus, besides the one and a half columns on the remainder of the page on which Tobit concludes, there was an entire unused page (front and back) after that – the last leaf of quire 49 – that was cut out when the manuscript was sewn together. 

            To restate:  what we have in Codex Vaticanus between Tobit and Hosea is nothing but a “seam,” so to speak, that resulted from the production-process, where one copyist’s work was attached to the pages produced by another copyist.  The situation is entirely different in Mark, where Mark 16:8, and the blank space, are on one side of a page, and the beginning of Luke is on the opposite side, and the text on both sides is, of course, written by the same copyist. 

            Wallace’s claim that “All in all, the reasons for the gaps are anything but clear” is not true.  Every blank space between books in Codex Vaticanus is fully capable of obvious explanation: 

(1)  The blank space before Psalms was required by the shift from a three-column format to a two-column format. 

(2)  The blank space before Hosea is a production-seam, where one copyist’s work was attached to another copyist’s work. 
(3)  The blank space between the Septuagint’s text of Daniel (concluding with the story of Bel and the Dragon) is the end of the Old Testament portion.   
(4)  The blank space after Mark 16:8 was elicited by the copyist’s recollection of verses 9-20.

            So:  there is more to the picture than the simple statement that “Some early manuscripts do not include verses 9-20.”  As far as early Greek manuscripts are concerned, “Some” = two.  “Early” = over 100 years later than clear patristic use of the contents of verses 9-20.  And “Do not include” = do not include, but show their copyists’ awareness of, verses 9-20.