Chapter 12:

Four Theories about How the Ending was Lost

In Part One, external evidence established that Mark 16:9 to 20 was treated as part of the Gospel of Mark in the 100’s by Justin, Tatian, and Irenaeus, all of whom were personally familiar with the city of Rome, where the Gospel of Mark had been composed. Other early patristic writers attest to the use of Mark 16:9 to 20 in the second and third centuries in diverse locales. In the early 300’s, the testimony of Eusebius of Caesarea shows that the passage was in “some copies” at Caesarea but was absent from Eusebius’ cherished codices, but the use of this passage by numerous writers in the 300’s and early 400’s demonstrates that in other locations it was treated as a normal part of the Gospel of Mark.

A detailed comparison of the variants and annotations in manuscripts with the Double-Ending established that the Short Ending originated in Egypt. An examination of the historical background of Codex Aleph, and the detection of features in Aleph which are shared by B, combine with an examination of their texts to establish that these two important codices share a close historical connection, and that they both descend from exemplars taken from Egypt. Likewise, Codex Bobbiensis has special traits which indicate that it was made in Egypt. An early stratum of the Armenian version, which was used as the base-text for the Old Georgian version, was conformed to the text of codices taken from Constantinople in about 430, and because the text of early Georgian and Armenian versions is Caesarean, it appears likely that those exemplars were, or were descended from, the copies which Eusebius had produced for Constantine ca. 330, for which the exemplars had been manuscripts at Caesarea.

When the influence of the early Alexandrian text-stream is not in the equation, neither the abrupt ending nor the Short Ending arises. The provenance of the Sinaitic Syriac manuscript is not as easily discerned as that of the other witnesses to the abrupt ending, but the rare readings that it shares with Codex Bobbiensis indicate that they are related, and since k is Egyptian, this elicits the deduction that the Sinaitic Syriac manuscript either was produced there also, or was influenced by an exemplar or ancestor-manuscript from there.

In Part Two, internal evidence established that Mark 16:9 to 20 was not initially composed as a continuation of the narrative that stops at the end of 16:8. Yet internal evidence also shows that Mark 16:9 to 20 was not composed as a pastiche, or patchwork derived from verbiage in the other Gospels. In addition, neither the vocabulary nor the style of Mark 16:9 to 20 precludes the identity of Mark as its author, provided that he initially composed it as a freestanding text.

Consideration of the external evidence and the internal evidence yields a hypothesis that Mark 16:9 to 20 was originally part of the Gospel of Mark, attached to 1:1 through 16:8 while the Gospel of Mark was still in production. Either the person who attached these verses was someone other than Mark, or Mark himself attached them in an uncharacteristic hurry to finish his account, perhaps reasoning that a sketched-out ending was better than no ending at all. They may have previously been a text that was used by Peter and Mark in the church at Rome to teach about Jesus’ post-resurrection appearances.

If they were initially part of the text of the Gospel of Mark, why were these verses absent in the earliest recoverable stratum of the Egyptian text-stream? Textual critics of the past have offered different explanations, and and it may be instructive to consider their strengths and weaknesses before advocating another theory.

Theory One: Excision by a Harmonist/Apologist. Origen, Eusebius, Jerome, Augustine, and other patristic writers occasionally reveal an interest in demonstrating the harmony of the Gospels. This was, in part, in anticipation of questions which readers might be expected to ask about differences between the accounts, and it was also, in part, a response to objections from non-Christian writers such as Celsus and Porphyry. Christian leaders in the early church were highly motivated to maintain that the Gospels smoothly interlocked with one another, free of contradictions and discrepancies. Eusebius, as we have seen, went so far as to blame some apparent difficulties on scribal corruption, and proposed emending the text – that is, from his perspective, reconstructing the original, harmonious text – to maintain harmonization. In Ad Marinum, Eusebius posited scribal error to resolve the difficulty between Mark 15:25 and John 19:14, and he noted that scribal error could be the cause of an apparent discrepancy between John 20:1 and the account in Matthew 28.

Origen, likewise, was not reluctant at all to resolve text-critical issues by preferring the reading which was apologetically advantageous. Commenting on John 1:28, he acknowledged that almost all copies read “Bethany,” and that it was attested by Heracleon, an older source, but he preferred to reading “Bethabara,” and he explained why: “Bethany . . . is fifteen stadia from Jerusalem, and the Jordan river is about one hundred and eighty stadia distant from it. Nor is there any other place of the same name in the neighborhood of the Jordan, but they say that Bethabara is pointed out on the banks of the Jordan.” In other words, Origen rejected the reading “Bethany” because it seemed to pose a difficulty.

On occasion, even when his manuscripts all said one thing – something which seemed, to Origen, problematic – Origen proposed that the original text of the Gospels said something else. According to Metzger, Origen offered a conjectural emendation of the text of Matthew 5:45, and Origen also suggests that scribal errors might be to blame for the harmonization-difficulties in the text of Matthew 21:9 and Mark 14:61.

Origen may not have been the first influential person in the early church to favor variants which were more apologetically convenient than their rivals. And if a person with an interest in maintaining the harmony of the Gospels were to approach them in the Western order (Matthew, John, Luke, Mark) then several difficulties would present themselves when he came to the ending of Mark:

(1) Mark 16:9, unpunctuated, gives the time of Jesus’ rising as “early on the first day of the week,” which could be misinterpreted to imply a disagreement with Matthew 28:1.

(2) Mark 16:13 says that the disciples did not believe the report of Cleopas and his fellow-traveler, but Luke 24:33 and 34 seems to present the disciples believing already when they hear the two travelers’ report that the Lord was risen.

(3) Mark 16:14 mentions an appearance to “the eleven” when they were sitting down at a table, but John 20:19 through 24, while it describes an appearance to the disciples, says that Thomas was absent, lowering the count to ten.

(4) Mark 16:14 to 20 seems to describe the statement of the Great Commission and the Ascension on the same occasion, whereas in Matthew 28, verses 16 to 20, the Great Commission is given in Galilee and Acts 1:6 through 12 locates the Ascension in Jerusalem, on the Mount of Olives.

All of these discrepancies can be resolved when one perceives the summarized nature of Mark 16:9 to 20, but if a person did not perceive that, he might be tempted to conclude that the difficulties had arisen as a result of textual corruption, and it was his duty to repair the damage. Facing a series of difficulties in his attempt to harmonize Mark 16:9 to 20 with the already-harmonized parallels in Matthew, John, and Luke, an early harmonist may have noticed that the first difficulty appears in 16:9, noticed the discontinuity between 16:8 (where a group of women is present) and 16:9 (where Mary Magdalene is present), and proceeded to make an intuitive leap: the entire discordant passage must be spurious.

This would require a rather sophisticated and bold approach to the text; however, such a thing was not unheard of in the early church. Dionysius of Alexandria, in the mid-200’s, made incisive comments about the authorial style of Revelation as he attempted to justify his view that it was not written by the same person who wrote the Gospel of John. Other writers appealed to stylistic features in the book of Hebrews to build a case that Paul was its author. In addition, an immediate removal of the passage need not have occurred: an early harmonist could add an obelus at, or a series of obeli alongside, Mark 16:9 to 20, to indicate that the remainder of the text was suspect, and a subsequent copyist, also aware of harmonization-difficulties, could proceed to interpret the obelus to indicate outright spuriousness.

Additionally, an apologetic difficulty may have been felt by early Christians who faced misinterpreters of Mark 16:17 and 18. This difficulty is still felt by apologists to this day; for example, James White mentions one theological “problem” after another involving these verses.

Within the church, misinterpreters may have used the passage to teach that miraculous signs, such as tongues-speaking and healings, were divinely sanctioned as normative Christian practices. And outside the church, anti-Christian writers derisively challenged Christians to drink poison, citing Mark 16:18. Such considerations could provoke an apologist to conclude that the entire passage is so problematic that it cannot be authentic, and on that basis he would reject it.

Dealing with such objections from anti-Christian writers was no idle exercise for Christians in the 200’s and 300’s. In about 250, Origen wrote an extensive reply (consisting of eight books) to the objections and jibes of the anti-Christian writer Celsus, who had written in about 180. Later, sometime before 300, Eusebius of Caesarea wrote a response to the anti-Christian writer Porphyry – a verbose response consisting of 25 books, none of which are extant.

To these apologetically driven authors, a text without Mark 16:9 to 20 almost certainly seemed more advantageous than a text which included it. To them, the resultant puzzling stop at the end of Mark 16:8 was something they could afford, in exchange for the maintenance of the consistency of the Gospels with one another. It probably was not difficult for these apologists to convince themselves of the correctness of a deduction that a problematic passage must not be authentic.

Jerome shares an interesting example of the extent to which some apologists would go to maintain the consistency and perfection of the Scriptures: in the Prologue to his Commentary on Daniel, he describes how an objection from Porphyry was answered. Porphyry had showed that in the story of Susanna, in the passage where Daniel cross-examines the elders, there are two expressions which form puns in Greek: “To split from the mastic tree (απο του σκηινου σκηισαι [apo tou sheinou skeisai])” and “to saw from the evergreen oak (και απο του πρινου πρισαι [kai apo tou prinou prisai]),” thus convicting the book of being a Greek forgery. Although, in his Introduction to Daniel, Jerome had acknowledged that the portions about Susanna and Bel and the Dragon were “spread throughout the world,” he responded to Porphyry’s objection in the following way:

“Both Eusebius and Apollinarius have answered him in the same manner, explaining that the stories about Susanna and of Bel and the Dragon are not contained in the Hebrew text; instead they constitute part of the prophecy of Habbakkuk ben-Joshua of the tribe of Levi. Likewise we find in the title of that same story of Bel, in the Septuagint, ‘There was a certain priest named Daniel the son of Abda, a close advisor of the king of Babylon,’ but Holy Scripture testifies that Daniel and the three Hebrew children were from the tribe of Judah.

“For this very reason, when I was translating Daniel many years ago, I marked these phenomena with an obelus, showing that they were not in the Hebrew text. And in this connection I am surprised to be told that certain fault-finders complain that I have taken the liberty of shortening the book. After all, Origen, Eusebius and Apollinarius, and other outstanding churchmen and teachers who were conversant in Greek acknowledge that, as I have said, because these phenomena are not found amongst the Hebrews, they are not obliged to explain to Porphyry difficulties which he found in these portions which exhibit no authority as Holy Scripture.”

These comments instructively display two things. First, they show that some readers could misunderstand the meaning of an obelus. Although, in Jerome’s translation, the story of Susanna was part of the book of Daniel, he reports that “certain fault-finders” had complained that he had shortened the book, simply because he had obelized the story of Susanna. Second, they show that apologists such as Eusebius, and Jerome himself, were willing to reject well-known passages in order to answer an objection against the veracity of the authors of Scripture. In Origen’s Hexapla, as Origen conveniently explains in a letter to Africanus, Origen attached an obelus to passages in the Hebrew text to which there was no corresponding text in the Greek translations, and in the Greek texts that were included in the Hexapla, he attached an asterisk to passages which had no corresponding text in his Hebrew copies.

Initially these obeli and asterisks were intended strictly as reports; Origen made it perfectly clear to Africanus that he preferred to use the Septuagint, and that to rely on the Hebrew text exclusively would cause all sorts of problems; he had marked the differences, he explained, so that he would know what passages were recognized by the Jews, so that in his discussions with them he would not be mocked for citing a text they did not acknowledge as Scripture. Origen treated the story of Susanna as Scripture in Book Ten of his Stromata and elsewhere in his writings.

Jerome attests, though – with sympathetic approval – that Eusebius and another writer had gone farther, and had removed Porphyry’s objection by declaring that it was an objection against an inauthentic passage. Considering how well-known the story of Susanna was, this was a remarkable and drastic step, and it shows that these apologists were willing to use a sort of theologically driven textual criticism as an apologetic weapon.

F. H. A. Scrivener, a prominent textual critic of the 1800’s, proposed that this sort of step had been taken to deal with objections about Mark 16:9 to 20: “In fact, after having been cited as genuine by the Fathers of the second and third centuries, from Irenaeus downwards, the difficulty of harmonizing their narrative with the other Gospels (a circumstance which ought to plead in their favour) brought suspicion upon these verses, and caused their omission in some copies seen by Eusebius (Questiones ad Marinum), whose influence over the Scripture codices of his age we

have seen to be very considerable.” William Farmer, in his detailed book about this passage, similarly proposed that an early Alexandrian editor had excised the passage because this was the simplest way to resolve objections to its several difficulties, real or imagined. Such an editor may have realized that the several small adjustments, concentrated in one passage, that would be needed to resolve the difficulties would be easily noticed and exposed as such when compared to unadjusted copies, whereas a simple excision of the entire problematic passage would leave no fingerprints, so to speak.

However, other researchers, including Burgon and Warfield, have expressed skepticism about Scrivener’s theory. Their reasoning is simple: a copyist who sensed harmonistic difficulties in some passages within Mark 16:9 to 20 would have altered the phrases which he perceived as non-harmonious; he would not delete the entire passage.

Theory Two: Accidental Loss Involving a Liturgical Note. In many copies of the Gospels, the words αρχη (arche, “beginning”) and τελος (telos, “end”) appear at the beginnings and endings of lections, the segments of text assigned to be read on certain days of the church-calendar. Usually the arche and telos symbols consistently appear throughout the Gospels, but in some copies, they only appear at a few selected lections. Sometimes one mark or the other will appear more frequently than its expected counterpart.

If an early copy of Mark contained only a few τελος (telos) symbols, and if, in this copy, a lection ended at the end of 16:8, and if, in this copy, 16:8 concluded at the end of a page (as it does in Codices 1, 15, and some other copies), then when used as an exemplar by a copyist who was equally unfamiliar with the text of Mark and the meaning of the telos-marks, such a copy could be misinterpreted to say that the Gospel of Mark ended there at the end of 16:8, and that the remaining text represented some other composition.

Or, if an early copy had incurred damage, and accidentally lost its final page – a page containing 16:9-20 – then such a copyist, unfamiliar with the text and with the use of lections, would naturally understand the telos after 16:8 to mean that he had reached the end of the Gospel of Mark. This theory was proposed by Burgon in 1871:

“Of course it will have sometimes happened that St. Mark 16: 8 came to be written at the bottom of the left hand page of a manuscript. And we have but to suppose that in the case of one such Codex the next leaf, which would have been the last, was missing, - (the very thing which has happened in respect of one of the Codices at Moscow) - and what else could result when a copyist reached the words,

ΕΦΟΒΟΥΝΤΟ ΓΑΡ ΤΟ ΤΕΛΟΣ

[Efobounto gar to telos]

but the very phenomenon which has exercised critics so sorely . . . . The copyist will have brought St. Mark’s Gospel to an end there, of course. What else could he possibly do?”n

The lectionary-related notes in manuscript 274 illustrate another aspect to this theory. In 274, between the end of 16:8 (on the left part of a line) and the beginning of 16:9 (on the right part of the same line), an abbreviated lectionary note says, “τελ. Β′ εωθ.” – “End of the second Heothina,” that is, the second lection in the eleven-part cycle for matins.n This was intended to notify the reader that Mark 16:1 through 8, the second Heothina-reading, ended there. Another abbreviated note, added in the right margin, states, “Third Gospel of the Resurrection; this is also the Gospel-reading at Matins on Ascension Day.”

A copyist familiar with the Gospel of Mark would not interpret a note signifying the end of the second Heothina-lection, “τελος ευαγγελιον Β′” (“B′” being the Greek representation of the numeral 2) as if it signified the end of the second Gospel itself. However, if the Heothina-readings were established in the second century (becoming only much later a component of the Byzantine lectionary) a copyist in Egypt whose usual duty was the production of copies of the Septuagint, and who was not familiar with the text of Mark, could understandably misinterpret such a note. In the Septuagint, numerals typically accompany the names of the Minor Prophets in the titles and subscriptions of each book. A copyist of the Septuagint in the second century would thus be accustomed to writing “‘Ωσηε Α′, Αμως Β′,” (Hosea – 1, Amos – 2) and so forth, up to Μαλαχιας ΙΒ′ (Malachi – 12), at the beginning and end of each book of the Minor Prophets.

If some early Christians were to hire such a copyist, and if they instructed him to prepare the first volume of a two-volume set of the Gospels in which Matthew and Mark were contained in the first volume, and Luke and John were in the second volume (fitting the format that seems to be used for Papyrus 75), then what might happen if, as his sole exemplar, he used a lector’s copy in which the ends of the Heothina-readings were noted?

Upon reaching the end of Matthew 28, verse 20, the copyist would encounter the note “end of Gospel #1” (τελος ευαγγελιον Α′.” He would understandably conclude that the note meant that the first Gospel ended there. After all, the text of Matthew does indeed end at that point. With that assumption in place, the copyist would proceed to copy the text of Mark, and when he reached the end of 16:8, he would find the note, “end of Gospel #2” (τελος ευαγγελιον Β′”). Because he was used to enumerating the Minor Prophets, and because he had understood the similar note at the end of Matthew to signify the end of Gospel #1, he would assume that this note signified the end of Gospel #2 – not the end of a Heothina-reading, but the end of the second Gospel-account. He therefore would regard Mark 16:8 as the end of the Gospel of Mark, and place the subscription after εφοβουντο γαρ (efobounto gar), and include the remaining 12 verses after that. This, then, would raise questions about the legitimacy of the separated verses, resulting in their eventual excision.

The disadvantage of this form of the theory is that it requires several fine-tuned conditions, foremost of which is that the Heothina-series would have to be in use in the 100’s or early 200’s. Although it is a very early arrangement of lections, there is no direct evidence that it is that early. Also, a lector would have had to mark the ends of lections but not the beginnings. The lector’s copy would have to be subsequently used by a copyist who was unfamiliar with the Gospel of Mark. And the copy produced by this copyist would have to be accepted by other copyists; otherwise its shortened text would be immediately repaired instead of being reproduced – thus this theory requires not just one, but several copyists to be unfamiliar with the text of Mark. Such a combination of events is unlikely, although it might be possible in a setting in which the leaders of a Christian community had been martyred or forced to evacuate, leaving novices in charge of manuscript-production.

Hort rejected Burgon’s theory, arguing as follows: “The last leaf of a Manuscript of Century 2 might easily be filled with verses 9 through 20, and might easily be lost; and thus the manuscript would naturally become the parent of transcripts having a mutilated text. It is not so easy to understand how a defect of this magnitude in so conspicuous a part of the Gospels could be widely propagated and adopted, notwithstanding the supposed existence of a fuller text in the copies current all around. Nevertheless the loss of a leaf in Century 2 does afford a tenable mode of explaining omission, and would deserve attention were the Documentary and the Intrinsic evidence ambiguous.”

In other words, setting aside Burgon’s hypothesis about lection-related marks, Hort granted that there is nothing impossible about the ideas that the last page of a codex of Mark

produced in the 100’s could have contained 16:9 to 20, and the idea that the last page of such a codex could be lost, and the idea that this codex, thus damaged, could become the parent of subsequent copies which displayed a text ending at the end of 16:8. The difficulty, Hort insisted, is that undamaged copies would become available, and copyists would quickly repair the damaged text.

Would they? If a damaged codex of Mark were used for several years in an isolated locale, without competition, its text would soon be regarded as the standard text in that area. When a new text was introduced from some other place, the locals might test it before accepting it, not unlike the way in which new Bible translations are evaluated before being adopted for use in church services. And to doctrinally vigilant, apologetically motivated copyists, the new passage might have seemed objectionable: the statement that Jesus appeared in another form, and the harmonization difficulties, and the statement about serpent-handling and poison-drinking, may have been too much for the local church-leaders to immediately accept. The evidence from Egypt shows that 16:9 to 20 was eventually accepted, but its acceptance was not immediate and unanimous.

If evidence were to be discovered that the Heothina-series existed in the 100’s or early 200’s – or that any manuscripts at all from that time contained lectionary-related marks and/or marginalia – then the plausibility of this theory would be greatly enhanced. Since this cannot be verified, however, a key component of this theory must be considered highly tentative.

Another theory may deserve to be mentioned, even though it has not been strongly endorsed by anyone, because it sheds an interesting light on the evidence that involves the Short Ending.

Theory Three: Accidental Loss Involving the Short Ending. Mark 16:9 to 20 forms an idea lection for Eastertime and Ascension-day, and it is thus not surprising that it was selected to be read on both those occasions in the lectionary. As a result, the preceding passage ends on a rather downbeat note: the women, mute with fear, flee from the tomb. This may have disturbed an early bishop or lector, who remedied the situation by creating a liturgical flourish, summarizing 16:9 to 20 so as to conclude on a positive note. After creating this liturgical flourish (the text we call the Short Ending), the bold embellisher may have written it in the margin of his copy of Mark, with instructions for the lector to read up to the end of 16:8, and proceed to continue reading to the end of the material in the margin, and stop there.

Such an adaptation would work as long as the lector understood the annotation in his copy of Mark. But if such a copy from outside Egypt was placed into the hands of professional Egyptian copyists who were completely unfamiliar its contents, such an annotation could trigger a terrible mistake. Instead of interpreting the marginalia as instructions for a lector, an Egyptian copyist could have interpreted them as instructions for a copyist. He thus could interpret those instructions to mean that the Gospel-account should continue to the end of the material in the margin, and stop there. Even with 16:9 to 20 present on the page in front of him, a copyist who misinterpreted such a margin-note in such a way would assume that its author had supervised, and corrected, the manuscript’s scribe, and thus the note, rather than the subsequent text, was to be followed. With that misunderstanding in place, an Egyptian copyist could begin a week of busy copying with an exemplar that contained 16:9 to 20 in the text and the Short Ending in the margin, and end the week with a new copy in which 16:9 to 20 was not included, and in which the Short Ending appeared immediately after 16:8 in the text.

Were Alexandrian copyists capable of this kind of error? A variant in the Alexandrian text of Matthew 27:49 may be instructive. At Matthew 27, verse 49, the Alexandrian text reads, “αλλος δε λαβων λογχην ενυξεν αυτου την πλευραν, και εξηλθεν υδωρ και αιμα” (“And another took a spear and pierced his side, and out came water and blood”). This feature is explained as being originally a note in the margin (drawn from John 19:33 and 34) which was inserted into the text. As Metzger stated, the chance that this variant is original is low for two reasons: first, if copyist working with other text-types had found it in their exemplars, it is more likely that they would have either relocated it, rather than uniformly removed it altogether, and, second, simply because Matthew was written first in Gospels-books, the presence of this verse in Matthew would have tended to elicit the removal or relocation of the verse in John, which has not occurred.

According to Robert Waltz, this reading is supported by “Aleph B C L U Γ 1010 1293 dubl eptmarg kenan lich mac-regol mull mae slav” but “is omitted by all other texts, including A D E F G H K M S W Δ Θ Σ Byz it am cav ful hub tol cur pesh hark sa bo arm geo.” The agreement of Aleph, B, and L is particularly significant. (The recently-published Middle Egyptian (abbreviated as “mae”) manuscript of Matthew, from ca. 400, displays “and the name of the servant was Malchus” at Matthew 26:51; the phrase was borrowed from John 18:10. In the same manuscript at Matthew 27, verse 49, the phrase from John 19:33 and 34 appears, with “blood” and “water” in the same order in which they appear in John.)

Considering Metzger’s case against this variant, and considering that the addition of this variant results in a text of Matthew 27, verse 49 that collides with the chronology of events in John (where the body of Christ is speared after His death), we may deduce that the copyist who initiated this Alexandrian reading was not very familiar with the text of the Gospels. A scenario in which early Egyptian copyists were capable of badly misinterpreting marginalia such as one might find in an exemplar that had been a lector’s copy is not merely conceivable, but observeable.

Hort claimed that “The documentary evidence for the Shorter Conclusion resolves itself into additional evidence (indirect, it is true, in form, but specially certified by the nature of the indirectness) for the omission of verses 9 through 20.” and Metzger, adopting some of Hort’s verbiage, also claimed that “The external evidence for the shorter ending (2) resolves itself into additional testimony supporting the omission of verses 9-20.”

However, such as assumption is not automatic. If the adoption of the Short Ending was simultaneous with the excision of 16:9 to 20 due to a copyist’s misunderstanding of a lector’s embellishment, then the situation is just the reverse: the Short Ending and 16:9 to 20 would both exist prior to the abrupt ending, which would come into existence as the result of reluctance to adopt either earlier ending.

Hort apparently never considered this possibility, regarding the Short Ending as “4 colourless lines” which would never be preferred over “12 verses rich in interesting matter.” Likewise Metzger, again adopting Hort’s verbiage, wrote, “No one who had available as the conclusion of the Second Gospel the twelve verses 9-20, so rich in interesting material, would have deliberately replaced them with four lines of a colorless and generalized summary.” However, if a second-century copyist who was unfamiliar with the text of Mark really thought that his exemplar was telling him to include the Short Ending, and stop there, that is probably what he would have done.

Such an error could only be perpetuated when and where Christians would not immediately perceive and correct it – either at a very early stage of transmission, when the text of

the Gospel of Mark was first introduced, or at a point immediately following a disruptive persecution. Such conditions may have existed in Alexandria in the very early 200’s, following severe persecutions.

The possibility of such a sequence of events justifies a closer consideration of the history and contents of the Short Ending. Codex Bobbiensis, from ca. 430, is its earlier witness, but by attesting to adparuit, Codex Bobbiensis supports a Greek form of the Short Ending that includes the word εφανη (efane). This is almost certainly a later form of the Short Ending than the form attested by Codex L, in which εφανη (efane) does not appear. Thus, although the parchment of Codex L is younger than that of Codex Bobbiensis, the text of the Short Ending in Codex L is older than k, and probably older than k’s exemplar. This implies that the Short Ending, without εφανη (efane) or εφανη αυτοις (efane autois), was extant at a time prior to the production-dates of Aleph and B.

That implication raises an interesting possibility regarding the testimony of Codex B. It is possible that the copyist of B possessed an exemplar with the Short Ending, and an exemplar with 16:9-20. If he was unsure which exemplar should be followed, he may have considered that his least worst option was to leave the manuscript somewhat unfinished, and formatted so that either ending could be added after 16:8: either the Short Ending, via letter-stretching, could be extended into the third column, or verses 9 to 20, via letter-squashing (or via the addition of an extra line in columns two and three), could be crowded into the third column. Codex B would thus be an indirect witness to both the Short Ending and to verses 9 to 20, and rather than testify to the existence of an exemplar which displayed the abrupt ending, it would be among the first generation of manuscripts to do so.

In addition, the view of Hort and Metzger that the Short Ending must have been created by a copyist whose exemplar contained the abrupt ending, and who had no awareness of 16:9 to 20, should not be accepted casually. If a copyist possessed a damaged exemplar, and knew that its final portion, containing 16:9 to 20, had been lost, and if he only vaguely recollected the missing passage, he might well produce the Short Ending as a summary of its contents: the Short Ending states that the women reported (εξήγγειλαν [exengeilan]) to Peter and those with him (i.e., with Peter); 16:10 states that Mary Magdalene reported (απήγγειλεν [apengeilen]) to those who had been with him (i.e., with Jesus). The Short Ending uses the phrase Μετα δε ταυτα (Meta de tauta); 16:12 begins with exactly the same three words. The Short Ending states that Jesus sent them out from east to west; Mark 16:15 states that Jesus commanded them to go into all the world. The Short Ending describes the good news (κήρυγμα (kerugma)) of eternal salvation (σωτηρίας (soterias)); in Mark 16:15-16 Jesus commands the disciples to preach (κηρύξατε (keruxate)), and that the believer will be saved (σωθήσεται (sothesetai)). The secondary form of the Short Ending has the word εφανη (efane); the same word appears in Mark 16:9. Are these all mere coincidences? Is it not possible that these elements in the Short Ending are faint recollections of 16:9 to 20? If this is the case, then the Short Ending may have been composed not as a liturgical flourish, but as a place-holder for 16:9 to 20, and, after being perpetuated in an isolated locale, it became entrenched, against the expectations and intentions of its creator.

Theory Four: Simple Accidental Loss. Those who might consider the preceding theory too complex might prefer this theory. It resembles Burgon’s theory, without the involvement of lection-related notes, and is simple: the first copy of the Gospel of Mark to reach an isolated locale in Egypt was a copy of Mark which, en route, was damaged, losing its final

portion, on which 16:9 to 20 had been written. The history of the text in Egypt would thus match the sequence in which the abrupt ending, the Short Ending, and 16:9 to 20 are found in Egyptian witnesses. The text of the damaged exemplar (with the abrupt ending) was transferred to numerous copies, and for a while the abrupt ending was the only known ending of Mark in that locale. Then someone, dissatisfied with the abruptness of the final scene, wrote and attached the Short Ending in his copy, and it, too, was used as an exemplar, resulting in a growing family of copies with the Short Ending. Then, when an undamaged copy of Mark was introduced from a nearby Egyptian locale, copyists resorted to making a non-decision by formatting the Double-Ending with notes about their puzzling situation, in which some copies ended at 16:8, some copies ended with the Short Ending, and some copies had 16:9 to 20 after 16:8. Next, when the abrupt ending had ceased to be perpetuated, and copyists faced one group of exemplars with the Short Ending and another group of exemplars with 16:9 to 20, they again resorted to making a non-decision by formatting the Double-Ending with notes about their puzzling situation, this time with no mention of the abrupt ending.

The external evidence fits neatly into such a reconstructed history: the exemplars of Codices Aleph and B and Codex P. Palau-Ribes Inv. Nr. 182 (and the abruptly-ending copies to which Eusebius referred in Ad Marinum) thus descend from the earliest stage of the Egyptian text of Mark. Codex 083 descends from a stage when all three endings were extant. Codices L and Ψ (Psi) descend from a stage when copies with the Short Ending, and copies with 16:9 to 20, were known. Later still, the Short Ending was blended directly into the text between 16:8 and 16:9, in an Ethiopic version. The only special mechanism that is required to trigger that series of events is that the accidental loss would need to occur at precisely the point where the stark transition between 16:8 and 16:9 occurs.

Each of these four theories has some shortcomings, but none of them render the theories altogether implausible:

● A harmonist who excised Mark 16:9 to 20 would have to be extremely confident and clever. On the other hand, early copyists did not always act with restraint, and early apologists, like other people, were capable of ruthless cleverness.

● Evidence that the Gospel of Mark was divided into lections in the 100’s and 200’s remains elusive. On the other hand, our knowledge of second-century manuscripts is not so complete that we can exclude the possibility that in some very early manuscripts the text was divided into lections. The Diatessaron’s 55-chapter arrangement would be convenient for weekly reading, as would the scheme of chapter-divisions attested by Codices B and 579.

● If the Short Ending was added in the margin of a copy of Mark 1:1 through 16:20 as a flourish for a lection that would otherwise end at 16:8, it would take a rather dense copyist, isolated from any advisors familiar with the text, to misinterpret this so as to excise verses 9 to 20 and retain the Short Ending as part of the text. On the other hand, some copyists were unfamiliar with the texts that they were copying, and some mistakes by early copyists show that they sometimes misunderstood margin-notes in their exemplars.

● An accidental loss of the final page or portion, occurring exactly where a remarkable transition of scene and style occurs, seems rather lucky. On the other hand, nothing precludes the idea that 16:9 began the last (or second-to-last) page of an early codex of Mark. The text is so formatted in several extant copies.

Any future argument against the passage must answer all these theories. Any argument in favor of the passage, if it does not adopt one of these theories, must be more persuasive than all of them. In the following chapter, such an argument is proposed.

●●●●●●●

Chapter 13:

Why Mark 16:9 to 20 Was Excised in Egypt

At this point, some readers may already be satisfied that five considerations render the case against the authenticity of Mark fatally weak:

(1) Internally, the style of the abrupt ending is radically unMarcan, and externally, it is attested in a narrow transmission-stream.

(2) The longer ending is attested by very early and very widespread witnesses.

(3) Several plausible mechanisms have been identified which could cause 16:9 to 20 to be lost after its initial inclusion in the production-stage of the Gospel of Mark. Even though it is impossible to empirically demonstrate that one of these mechanisms is the cause of the loss, it may seem reasonable to diagnose that the early Alexandrian Text is not healthy at this point, even if the exact germ that caused this cannot be identified.

(4) Attempts to dismiss Mark 16:9 to 20 as a Byzantine reading which invaded the other text-types are futile: if the text of this passage in leading members of each text-type is Byzantine, that would show that the passage has been grafted on, while if its text in those witnesses contains unique variants, this would show that those variants and the text in which they are embedded were not acquired by invasion, but are indigenous. Consider the test-results:

● Five readings appear in Caesarean witnesses but tend to be absent in the definitive Byzantine, Western, and Alexandrian witnesses:

(a) f13 omits δε (de) and inserts the contracted name “Jesus” after Αναστας (Anastas) in 16:9.

(b) Codex Θ (038) has μαθηταις (mathetais) in 16:10 instead of μετ’ (met’).

(c) Codex Θ (038) has εφανη (efane) instead of εφανερωθη (efanerothe) in 16:12.

(d) Codex Θ (038) has πορευθεντες (poreuthentes) instead of απελθοντες (apelthontes) in 16:13.

(e) f1, f13, 28, and 565 (and A, Δ, and C) add εκ νεκρων (ek nekron) after εγηγερμενον (egegermenon) in 16:14.

● As was mentioned in chapter four, Codex D, representing the Western text-type, has seven non-Byzantine readings in Mark 16:9 to 15:

(a) Codex D has εφανερωσεν πρωτοις (efanerosen protois) instead of εφανη πρωτον (efane proton) in 16:9,

(b) Codex D has αυτοις (autois) after απηγγειλεν (apengeilen) in 16:10,

(c) Codex D has και ουκ επιστευσαν αυτω (kai ouk episteusan auto) instead of ηπιστησαν (epistesan) in 16:11,

(d) Codex D has και (kai) at the beginning of 16:12,

(e) Codex D has προς αυτους (pros autous) instead of αυτοις (autois) in 16:15,

(f) Codex D omits απαντα (apanta) in 16:15, and

(g) Codex D inserts και (kai) before κηρυξατε (keruxate) in 16:15.

● Three readings – the second and third of which are particularly distinctive – occur almost exclusively in representatives of the Alexandrian Text:

(a) C*, L, 33, 579, and 892 (and D and W) have παρ’ (par’) instead of αφ (af’) after Μαρια τη Μαγδαληνη in 16:9.

(b) C*, L, Δ (Delta), and Ψ (Psi) omit καιναις (kainais) at the end of 16:17. 099 also omits γλωσσαις λαλησουσιν (glossais lalesousin), which is probably the result of accidental line-skipping. This implies that 099’s exemplar read:

δαιμονια εκβαλουσιν

γλωσσαις λαλησουσιν

και εν ταις χερσιν etc.

(c) C, L, Δ (Delta), Ψ (Psi), 099, 579, and 892 have και εν ταις χερσιν (“And in their hands”) at the beginning of 16:18.

The non-Byzantine features in Caesarean, Western, and Alexandrian copies cannot reasonably be supposed to have been derived from an invading Byzantine Text. This shows that Mark 16:9 to 20 was not grafted onto the text of Mark in these text-types; it is indigenous to them.

(5) Patristic evidence generally does not support the abrupt ending or the Short Ending; not only is the patristic evidence in favor of Mark 16:9 to 20 ancient and widespread, but no patristic authors prior to Eusebius say anything about the abrupt ending of Mark, and after Eusebius, as we have seen, none of the patristic writers who mention the abrupt ending after Eusebius write independently of his comments. They used his comments and altered them in the course of advocating the inclusion of 16:9 to 20.

Although these five points form a sufficient case for the authenticity of Mark 16:9 to 20, I believe that there is an additional solution which is even more attractive than the ones previously described. Before presenting this solution, it may be helpful to briefly review some traditions about the circumstances under which the Gospel of Mark was written, and about the ministry and martyrdom of Mark himself.

Eusebius of Caesarea, in Church History Book Three, chapter 39, preserves Papias’ statement that “The Elder” reported the following: “Mark, who had been Peter’s interpreter, wrote down accurately, though not in order, whatever he remembered of the things said or done by Christ. For he neither heard the Lord nor followed him, but afterward, as I said, he followed Peter, who adapted his teaching to the needs of those who listened to him, but with no intent to give a sequential account of the Lord’s discourses. So that Mark committed no error while he thus wrote some things as he remembered them. For he was careful of one thing: not to omit any of the things which he had heard, and not to state any of them falsely.”

In Church History Book Five, chapter 8:1 through 3, Eusebius quotes from the beginning of the third book of Irenaeus’ Against Heresies (where Irenaeus seems to rely on Papias’ writings): “Matthew published his Gospel among the Hebrews in their own language, while Peter and Paul were preaching and founding the church in Rome. After their departure (έξοδον, exodon), Mark the disciple and interpreter of Peter also transmitted to us in writing those things which Peter had preached.”

In addition, in Church History Book Six, 14:5 through 7, Eusebius presents a statement that he attributes to Clement of Alexandria:

“Clement gives the tradition of the earliest presbyters, as to the order of the Gospels, in the following manner: the Gospels containing the genealogies, he says, were written first. The Gospel according to Mark had this occasion: as Peter had preached the Word publicly at Rome, and declared the gospel by the Spirit, many who were present requested that Mark, who had followed him for a long time and remembered his sayings, should write them out. And having composed the Gospel he gave it to those who had requested it. When Peter learned of this, he neither directly forbade nor encouraged it.”

The accounts of Irenaeus and Clement conflict: Irenaeus states that Mark wrote after the departure of Peter and Paul, but Clement states that Mark was distributing the Gospel while Peter was still alive. This should be compared to what Jerome, relying on his recollection of earlier compositions, wrote in the eighth chapter of De Viris Illustribus:

“Mark, the disciple and interpreter of Peter, wrote a short gospel at the request of the brethren at Rome, embodying what he had heard Peter tell. When Peter heard this, he approved it and published it to the churches to be read by his authority, as Clement in Book 6 of his Hypotyposes, and Papias, bishop of Hierapolis, record. Peter also mentions this Mark in his first epistle, figuratively indicating Rome under the name of Babylon: “She who is in Babylon elect together with you salutes you, and so does Mark my son.”

“So, taking the gospel which he himself composed, he [Mark] went to Egypt. And first preaching Christ at Alexandria, he formed a church so admirable in doctrine and continence of living that he constrained all followers of Christ to his example. Philo ~ most learned of the Jews ~ seeing the first church at Alexandria still Jewish in a degree, wrote a book on their manner of life as something creditable to his nation, telling how, as Luke says, the believers had all things in common at Jerusalem, so he recorded what he saw was done at Alexandria under the learned Mark. He died in the eighth year of Nero and was buried at Alexandria, Annianus succeeding him.”

Jerome was clearly relying on earlier accounts, including Eusebius’ Church History; the statement about the year of Mark’s death seems to be drawn directly from Eusebius’ Church History, Book Two, chapter 24: “When Nero was in the eighth year of his reign, Annianus succeeded Mark the evangelist in the administration of the parish of Alexandria.”

Eusebius provides a second affirmation of the year of the beginning of the bishopric of Annianus in Church History, Book Three, chapter 14: “In the fourth year of Domitian, Annianus, the first bishop of the parish of Alexandria, died after holding office twenty-two years, and was succeeded by Abilius, the second bishop.” Figuring that Domitian’s reign began in Septemner of 81, adding four years brings us to September of 85. By subtracting 22 from 85, we arrive at the year 63. If Annianus served as bishop for a bit more than 22 years but less than 23 full years, Eusebius’ two statements agree.

On the question of whether Mark wrote his Gospel before Peter’s death, or afterward, the accounts are divided. Their discord may decrease a little if Jerome’s statement is understood as an incorrect deduction based on Eusebius’ statement that Annianus succeeded Mark in the eighth year of Nero’s reign. If Eusebius’ statement means that Mark, instead of dying in that year, departed from Alexandria to go to Rome, then if Nero’s eighth year is calculated to be 62 (since his reign began on October 13, in the year 54), the emerging picture is that Mark established a Christian community in Alexandria, and then went to Rome, possibly at the urging of Timothy (see Second Timothy 4:11). According to this hypothesis, Peter and Mark were both ministering in Rome in the year 62.

In the mid-60’s, severe persecution against Christians arose in the city of Rome, and both Paul and Peter were martyred. What then happened to Mark? He apparently did not remain in Rome; as Peter’s assistant he would have been a natural choice to lead the congregation there; yet a man named Linus is reported by Eusebius (in Church History Book Three, 3:2) to have been the first bishop of Rome after the martyrdoms of Paul and Peter. A detailed tradition is found in the medieval composition History of the Patriarchs of the Coptic Church of Alexandria by Severus of Al-Ushmunain (in the mid-900’s), who stated that he accessed source-materials from the monastery of St. Macarius and other monasteries in Egypt, and from Alexandria. Severus of Al-Ushmunain states that Mark was martyred in Alexandria.

When this is compared to the report from Irenaeus that Mark composed his Gospel-account after the departure – that is, the martyrdoms – of Peter and Paul, the situation becomes more clear: after assisting Paul on his first missionary journey, and after assisting Barnabas in Cyprus (as related in Acts 13:1 to 13), and then with Barnabas (as related in Acts 15:36 to 39), Mark established churches in Egypt in the 50’s, and traveled from there to Rome in 62, leaving behind Annianus. Immediately after the deaths of Paul and Peter, Mark left Rome and returned to Egypt.

The martyrdoms of Paul and Peter are generally assigned to the year 67. Eusebius of Caesarean, in Book Two, chapter 25 of Church History, states that Paul was beheaded in Rome, and that Peter was crucified in the reign of Nero. He also reports that they were both martyred at the same time, and cites as his source for this information a man named Dionysius of Corinth. Dionysius of Corinth is a fairly early source; Eusebius reports that he served the church in the early 170’s. Jerome, in the first and fifth chapters of De Viris Illustribus, echoes Eusebius’ information, stating that Peter and Paul were both martyred “in the fourteen year of the reign of Nero, which is the 37th year after the Lord’s Sufferings.”

The account preserved by Severus of Al-Ushmunain specifically states that Mark was seized by unbelievers in Alexandria on Easter, when one of their religious festivals, dedicated to the deity Serapis, occurred, on the 29th day of the month called Barmudah (the eighth month of the Egyptian calendar), and that he died the next day. Although this is a late document, its author states that he relied upon earlier sources. One such earlier text, although it does not say anything about the specific date of Mark's martyrdom, agrees regarding the location: the author of The Martyrdom of Peter of Alexandria (this bishop was martyred in 311) states, “They took him up and brought him to the place called Bucolia, where the holy St. Mark underwent martyrdom for Christ.” The same author states that Peter of Alexandria entreated his persecutors “to allow him to go to the tomb of St. Mark.”

Only in certain years would Easter coincide on the calendar with the festival of Serapis, and the year 68 is one of those years. Thus, it appears Mark was martyred in 68, in Alexandria, less than a year after Paul and Peter were martyred in 67 in Rome. If the gist of the tradition preserved by Irenaeus is followed, then Mark must have had only a small window of opportunity after the martyrdoms of Paul and Peter to write his Gospel-account.

This does not mean that the tradition reported by Clement of Alexandria is entirely untrue. After Mark had been in Rome long enough to be recognized as Peter’s assistant and interpreter, he may have responded to requests for copies of collections of Peter’s sayings. These collections, though, may have been shorter than the final form of the Gospel of Mark; a definitive collection of all of Peter’s remembrances would not be feasible until after Peter stopped recollecting.

The tradition preserved by Irenaeus is not likely to be a later invention, because creative tradition-inventors would tend to emphasize the apostolic authority of the text. Clement’s tradition, by stating that Peter neither approved nor disapproved Mark’s undertaking, certainly does not seem to have been designed to ensure that readers would regard the Gospel of Mark as apostolically approved, but Irenaeus’ tradition, by stating that Mark wrote the Gospel of Mark after Peter had departed (that is, died), is even less positive, inasmuch as the martyred apostle Peter cannot even acquiesce to the text’s contents.

If we thus accept Irenaeus’ basic version of events, and assign a date in 67 for the martyrdom of Peter in Rome, and a date in 68 for the martyrdom of Mark in Alexandria, then the date for the composition of the Gospel of Mark must be somewhere in between.

All this provides the background for the following hypothesis:

In the second half of the year 67, following the martyrdoms of Peter and Paul, as Mark was almost finished writing his Gospel-account, he was in imminent danger and had to suddenly stop writing his nearly-complete text, leaving it, and whatever else he had written, in the hands of his colleagues. Thus, when Mark left Rome, his definitive collection of Peter's remembrances was unfinished and unpublished.

Mark’s Roman colleagues were thus entrusted with an incomplete and unfinished text. They had no desire to insert material of their own invention into Mark’s text, but they also had no desire to publish a composition which they all knew was not only unfinished, but which would be recognized as unfinished by everyone who was familiar with Peter’s preaching, indeed, by everyone acquainted at all with the message about Jesus. Therefore, rather than publish the Gospel of Mark without an ending (that is, with the abrupt ending), they completed it by supplementing it with a short text which Mark, at an earlier time, had composed about Jesus’ post-resurrection appearances. Only after this supplement was added did the Roman church begin to copy the Gospel of Mark.

That hypothesis explains internal features in 16:9 to 20, such as the awkward transition between 16:8 and 16:9, and the concise writing-style, but it does not explain why 16:9 to 20 is not found in the earliest stratum of the Alexandrian Text. To answer that question, we must consider the nature of Mark 16:9 to 20, and the nature of Alexandrian copyists.

B. H. Streeter, in his influential book The Four Gospels, made an insightful surmise about Mark 16:9 to 20: “The hypothesis that Mark 16:9-20 was originally a separate document has the additional advantage of making it somewhat easier to account for the supplement in the text of W (cf. p. 337 f.) known as the “Freer logion.” A catechetical summary is a document which lends itself to expansion; the fact that a copy of it had been added to Mark would not at once put out of existence all other copies or prevent them suffering expansion. No doubt as soon as the addition became thoroughly established in the Roman text of Mark, it would cease to be copied as a separate document. But supposing that a hundred years later an old copy of it in the expanded version turned up. It would then be mistaken for a fragment of a very ancient manuscript of Mark, and the fortunate discoverer would hasten to add to his copy of Mark – which, of course, he would suppose to be defective – the addition preserved in this ancient witness.”

That is a very plausible explanation of the origin of the Freer Logion; slightly adapted, Streeter’s theory implies that the Freer Logion did not originate as an expansion in the Gospel of Mark, but as an expansion of the freestanding Marcan summary of Jesus’ post-resurrection-appearances which Mark’s colleagues incorporated into the text of the Gospel of Mark. But what was such a text doing in Egypt? If Mark was the author of this summary, then it is possible that he composed it not in the 60’s at Rome, but earlier, during the period in the 50’s-62 when he was in Egypt – the only locale in which the Freer Logion is known to have existed.

If Mark’s summary of Jesus’ post-resurrection appearances was already known to some of the Christians in Egypt, who used it as a freestanding composition, then when the Gospel of Mark arrived from Rome in the late 60’s or early 70’s, sometime after Mark’s colleagues there had begun its dissemination, it would not be difficult for them to compare it to their copies of the Marcan summary-text and immediately see that the final portion of the text from Rome was not, and could not be, part of the Petrine Memoirs.

At least some of the first individuals in Alexandria to read the Gospel of Mark would thus be inclined to regard 16:9 to 20 as a distinct Marcan composition which, though valuable as a

Marcan text, simply did not belong in the memoirs of the apostle Peter. As a result, they declined to perpetuate it in their copies of the Gospel of Mark. This explains why the early Alexandrian Text is divided: the Egyptian Christians who accepted the Marcan ending perpetuated it in their copies, while those who did not accept it did not reproduce it as part of the text of Peter’s remembrances (which is what the Gospel of Mark was understood to be).

The abrupt ending is thus explained as the effect of Egyptian copyists’ recognition of Mark 16:9 to 20 as a composition unrelated to the remembrances of Peter. These verses were rejected by some Egyptian copyists, probably as early as the early 100’s, because the copyists thought that the Gospel of Mark derived its authority from the apostle Peter, rather than from Mark, and because although the passage was useful, they knew that it was not Petrine.

This rare tendency to apply a sort of higher criticism to justify the excision of material that did not appear to have come from the primary author was apparently shared by one of the copyists of one of the two Greek manuscripts which displays the abrupt ending. In Codex Sinaiticus at the end of John, Scribe A finished the text at the end of 21:24, and followed this with the decorative coronis and the subscription. Then he had second thoughts, erased the coronis and subscription, and added 21:25, followed by a new coronis and a new subscription. Tischendorf had detected this in the 1800’s, but it was not until the page was exposed to ultraviolet light in research overseen by Milne and Skeat that the evidence came to light – literally – that the copyist who initially concluded the text of John at the end of 21:24 was the same one who erased the coronis and subscription, and added 21:25 with a new coronis and subscription.

The initial excision of John 21:25 in Aleph was probably not an altogether isolated decision; Theodore of Mopsuestia (350 to 428), in a statement preserved in Ishodad of Merv’s Commentary on the Gospels, claimed that the extra material in the Septuagint version of Job, and the sentence about the angel moving the waters in John 5:4, and this verse, John 21:25, are “Not the text of Scripture, but were put above in the margin, in the place of some exposition; and afterwards, he says, they were introduced into the text by some lovers of knowledge.” Theodoret himself may have been repeating a theory of an earlier writer whose claims were also known to Scribe A of Aleph.

A textual anomaly of a different sort may indicate that outside of Egypt, Mark 16:9 to 20 was accepted even though it was known to be in some sense secondary to the rest of the Gospel of Mark. The evidence for this is indirect, and its connection to the ending of Mark can only be posited by a series of hypotheses. That is why I have mentioned this piece of evidence last, almost as a tangent, before presenting my concluding thoughts on the main subject. Nevertheless this last piece of evidence, though its relevance is admittedly speculative, is very interesting.

Why does the Gospel of John end twice? At the end of John 20:31, the narrative is brought to an appropriate conclusion; nevertheless an additional chapter follows: a chapter which features an appearance of Jesus to a group of disciples, including Peter, in Galilee.

The concluding verses of John 21 form such an appropriate conclusion that Origen, in Book Ten of his Commentary on John, stated that the words, “Blessed are they that have not seen and yet have believed” are found “at the end of the Gospel of John.” Possibly he was merely making a generalized location, for elsewhere Origen seems to show that he knew the contents of John 21. That explanation, however, cannot account for a statement of Tertullian: in chapter 25 of his composition Against Praxeas, he stated that the phrase, “that you might believe that Jesus Christ is the Son of God” are written at the “very termination” of the Gospel of John. Yet Tertullian also shows, in chapter 50 of De Anima, that he was aware that there had been “an

ungrounded expectation that he [John] would remain alive until the coming of the Lord,” which is alluded to in John 21:23.

These statements by Origen and Tertullian suggest that in North Africa and Egypt in the early 200’s, John 21 was considered to be an appendix to the main narrative. This interpretation persisted even in the days of Augustine, who stated in Tractate 122 on John (covering John 20:30-21:11), “This paragraph [20:30 and 31] signifies, as it were, the end of the book. But there is afterwards related how the Lord manifested Himself at the sea of Tiberias, and in the catch of fishes made special reference to the mystery of the church, regarding its future character in the final resurrection of the dead. I think, therefore, that in order to give special prominence to this section, that it has been arranged here, where there is a sort of conclusion to the book, and then a sort of preface [21:1] to the narrative that was to follow, to give it a position of greater eminence.”

Perhaps, as copies of the Gospel of Mark were being circulated and recopied in the early 70’s, someone from Rome took a copy to Ephesus, to the apostle John, along with information about the background of Mark 16:9 to 20 and with a question about its appropriateness. And maybe John, in response, wrote a short text telling about how Peter and some of the other disciples, including himself, had gone to Galilee, and had encountered Jesus (as John 21:2 through 14 relates), and that Peter had been called at that time to shepherd the sheep (as John 21:15 through 19 relates), and that Jesus had prophesied at that time that Peter would be martyred (as John 21:18 to 19 shows). This text, for a short time, may have circulated with the understanding that it was an apostolic continuation of Peter’s Memoirs – and if it had been aggressively promoted, it might have displaced Mark 16:9 to 20, but instead, perhaps because copies from Rome which included 16:9 to 20 continued to circulate and multiply, it was instead attached, with some editorial adjustments, to the end of the Gospel of John.

This would explain why the Gospel of John ends twice.

It might also account for an interesting feature in the pseudepigraphical Gospel of Peter, which was described earlier in comments about Codex Bobbiensis. In this text, the narrative flows smoothly from the events related in Mark 16:8 to events related in John 21. In chapters 13 through 15 of this text, the angel at the tomb tells the women that the crucified one they seek has risen: “He is risen and departed; but if you do not believe it, look in and see the place where he lay, that he is not here. For he is risen and is departed whither he was sent.” Next: “Then the women were frightened and fled. Now it was the last day of unleavened bread, and many were coming forth from the city and returning to their own homes, for the feast was ending. But we, the twelve disciples of the Lord, were weeping and in sorrow, and each one, being grieved because of what had happened, departed to his own house. But I, Simon Peter, and Andrew my brother, took our nets and went to the sea. And there was with us Levi the son of Alphaeus, whom the Lord” - (here the text in the manuscript stops).

There are several interesting features in this part of the Gospel of Peter (not the least of which is the remotely possible adaptation of Mark 16:10’s reference to mourning and weeping; Mark 16:10 has πενθουσι και κλαίουσιν (penthousi kai klaiousin); Gospel of Peter has εκλαιομεν και ελυπουμεθα, eklaiomen kai elupoumetha). The author has designed his narrative to proceed from the scene at the beginning of Mark 16:8 (or Matthew 28:8) to the scene at the beginning of John 21. Conspicuously absent are the angel’s instructions to inform the disciples about Christ’s resurrection and summon them to Galilee. The post-resurrection appearances related by Matthew and Luke are not used. It is as if the author of Gospel of Peter considered John 21 to be, in some way, the proper continuation of the Gospel of Mark. This may not have been his own idea.

Possibly he inherited a short-lived tradition in which John 21 was interpreted as the apostolic conclusion to Peter’s Memoirs, and in which Mark 16:9 to 20 was considered to have been rendered superfluous by the Johannine ending. Possibly this tradition was briefly shared in Alexandria. If copyists there interpreted John 21 as the apostolic sequel to the Petrine Memoirs collected by Mark, they would not find the abrupt ending particularly disturbing, as long as that interpretation endured.