Κυριακή, 5 Μαρτίου 2017

Richard Carrier : Mark 16:9-20 as Forgery or Fabrication (2)

The External Evidence

When we turn our attention to the external evidence, this conclusion is confirmed. External evidence consists, first, of the evidence of the actual surviving manuscripts themselves, their evident dates and relationships, and the actual text they contain, as well as other physical evidence in them, such as scribal marks and marginal notes, and, second, the evidence of outside witnesses. In this case that means the Church Fathers, who are the earliest Christian writers outside the NT, several of whom quote or cite the Gospel of Mark, or even discuss what they saw in different manuscripts of Mark. The former is called the manuscript evidence, the latter is called the Patristic evidence.
======

The Manuscripts: Textual Evidence

A common misconception is that counting manuscripts decides what reading to regard as original. But a later reading will often have been copied many more times, precisely because it was more popular (and often for the very same reasons the emendation occurred in the first place). So often the original reading is the rarest in surviving manuscripts, not the most common. But occasionally the reverse is the case. So a more judicial analysis of the evidence is necessary. Ancient translations of the Bible afford an important source of information, as they will reflect the state of the text at the time the translation was first made, no matter how late the surviving copies of that translation are. Likewise, from many surviving copies of the original text we can often reconstruct what the manuscripts they were copied from contained, and even date when those 'source manuscripts' were made or copied from, even though those manuscripts are now lost. And some manuscripts carry far more weight than others, because they are the oldest, or used very early in contexts that entail their text held wide authority, or both.

In Greek

The oldest and most authoritative manuscripts of Mark are found in the Codex Sinaiticus (Aleph) and Codex Vaticanus (B), both of which lack the LE and the SE. There are a few older papyrus fragments of Mark, but none contain any part of chapter 16 and thus are of no help in determining the state of Mark's ending.26 Both Sinaiticus and Vaticanus date to the mid-4th century and bear signs of having been treated as authoritative texts within the Church. Many of their readings agree with numerous other early mss. Both do leave a blank space at the ending of Mark, which some scholars believe may indicate awareness of a missing ending (although, of course, a lost ending may have simply been assumed). But the Vaticanus usually indicates known textual variants with a scribal mark, which is absent here, arguing against awareness of any lost ending; the space left is only large enough for the SE, which argues against awareness of the LE; and the Vaticanus leaves blank spaces after other books, demonstrating that such does not in fact indicate awareness of a lost ending.27 Likewise, the Sinaiticus also leaves a blank space after Acts, thus such does not entail awareness of a lost ending to Mark, either.28 And experts have determined the original form of Codex Sinaiticus also lacked enough room for the LE, which also argues against knowledge of the LE.29
J.K. Elliott asserts that Vaticanus and Sinaiticus were produced by the same scribe (in Black, PEM, pp. 85-86), but as he adduces no arguments or evidence in support of that claim, I'm compelled to reject it as spurious. Even if they derive from the same scriptorium (a more plausible claim, although it's widely debated), Elliott himself admits such mss. can still derive from different exemplars (ibid., p. 83 n. 4), and we know for a fact these two must have, as their texts frequently do not agree. For example, Mark 1:40, 2:22, 10:26, and 15:44, all differ between the Vaticanus and Sinaiticus, and I just flipped to four random pages of the Aland text. Such disagreements between them number in the thousands.30 Moreover, expensive projects like these would not have relied on a single exemplar but been checked against several (e.g. the Vaticanus frequently indicates the existence of variant readings, and shows influence from both major text types, the Western and Alexandrian). Apologists like to denigrate the Vaticanus and Sinaiticus as aberrant texts, 'exceptions to the rule' (combining the fallacies of special pleading and poisoning the well) when in fact all early NT mss. are at least as deviant and flawed as they are (so cannot claim any greater authority over them on grounds of 'accuracy'), and yet these two were clearly very authoritative texts, expensively produced by the church, based on multiple exemplars, and of the earliest date among all known mss. (some scholars estimate their exemplars dated as early as the late 2nd century; and no extant mss. date earlier than these mss. themselves). They are therefore far more authoritative than deniers would have it.

In Syriac

The SE and LE are also absent from the oldest Syriac manuscript (an erased palimpsest of the late 4th century), the Sinaitic Syriac. The LE finally appears in the Syriac tradition a century later, the earliest being the Curetonian Syriac (dated to the 5th century), which shows signs of revision from a Greek exemplar, unlike the Sinaitic which appears to be more original and, unlike the Curetonian, shows direct influence from (or upon) the Diatessaron, which rather supports the conclusion that the original Diatessaron also lacked the LE (see section 5.3.3).31 The fact that other translations whose early representatives lack the LE were ultimately derived from the earliest Syriac confirms that the original Syriac tradition lacked the LE (see section 5.1.6).

In Coptic

In Coptic, all but one include the LE, but all surviving mss. date centuries after the translations were originally made, and the earliest version indicates it wasn't originally there. According to P.E. Kahle, the Coptic translation in the Sahidic dialect is the oldest (originating in the late 2nd century), yet "of the Sahidic manuscripts" that contain the LE "only one...regards 16:9-20 as part of the original text," while all "the other Sahidic manuscripts...contain evidence that some (older) manuscripts ended at 16:8." And now we know one Coptic ms. indeed lacks the LE altogether (see below). Of the others, all but one include the SE and LE "but indicate by short notes that these are alternatives found [only] in some manuscripts" (many Greek mss. indicate the same, see section 5.2; as also the Ethiopic, see section 5.1.4). The same thing is observed in the only surviving Fayyumic ms. containing the ending of Mark (extant only in fragments, whose date is unknown but must be very ancient), despite having been translated from a different Western Greek text type than the Sahidic (no later than the early 4th century). Here, "in a short note after [the SE] it points out that [the LE] was not read by all the manuscripts before the translator."
Confirming these scribal notes, we have at least one Sahidic ms. (Codex P. Palau Rib. 182, from the 5th century) that clearly lacks the LE (ending with the OE), without any indication of knowing any other text, thus confirming the conclusion that the earliest Coptic translator did not know the LE. Only mss. containing the Coptic translation in the Bohairic dialect (rendered in the 3rd or 4th century) all contain the LE without comment, so either the LE was added to the Bohairic in the later 4th century or the Bohairic derives from a copy of Mark to which the LE had become appended in the 3rd or early 4th century—while the earlier Sahidic did not (it appears to have had it added later—unless it was dropped without comment by or before the Palau scribe, but even that entails the original Sahidic translator knew the LE was not in some mss., because then the original translation must have indicated this fact, as that indication is preserved in almost all subsequent copies surviving). Then the later Fayyumic was produced by a translator aware of the fact that some mss. lacked the LE (because he said so).32 All of these facts combined indicate the LE was a rare reading and not original to Mark when the earliest translations to Coptic were made, but became incorporated later.

In Ethiopic

The Ethiopic manuscripts all contain the SE and LE (or only the LE), but all date well after the 4th or 5th century when the translation was made. The earliest are the Garima Gospels, recently re-dated to the 7th century, which contain the LE alone, and beyond that the earliest surviving ms. dates no earlier than the 9th century. And as with the Coptic, evidence suggests the original Ethiopic translation lacked the LE. Of 65 Ethiopic mss. now extant, 18 contain the LE alone, while the other 47 contain the SE followed by the LE, and 13 of those indicate the LE was an addition (with symbols or terminations separating it from the SE, or actual scribal notes declaring it).33 Although a few of the oldest mss. (one dating as far back as the 7th century) have only the LE, the later mss. that indicate otherwise (i.e. that the LE was later appended and earlier mss. ended with the SE alone) likely derive from an even earlier tradition.
Since the original Ethiopic translation was made at the end of the 5th century, there had been plenty of time (around four centuries) for one tradition to append the LE and another tradition to append the SE (or the original translation may have simply begun with the SE). The second tradition then came to append the LE by influence from the first tradition. One might instead hypothesize that the original translation was derived from a Greek exemplar containing the DE and scribal indications of the LE being unknown in some mss. (which by the 5th century, when the Ethiopic translation was made, would be entirely plausible), but that would not explain the Ethiopic mss. that lack the SE. So one tradition must have contained the LE alone, and the other the SE alone, and then the SE tradition was merged with the LE tradition by adding the latter to the former. The reverse is far less likely, as it would require interpolating the SE between the OE and LE, which makes no logical sense, since the SE and LE contradict each other, and the SE adds nothing not already in the LE. And if the SE were appended as an alternative to an original tradition that ended with the LE, then the SE would more likely be placed after the LE, or in the margins. Even more likely, the SE would simply be rejected (and thus not appear at all), or else the LE would be replaced with the SE (see sections 4.3.4 and 5.1.8).
Consequently, the only plausible way so many Ethiopic mss. could have the SE followed by the LE (and for so many of those to clearly indicate that the LE was not original and for there to be so many Ethiopic mss. that contain only the LE and no hint of the SE) is if the LE was not in the original Ethiopic but came to be appended to some Ethiopic mss. sometime between the 5th and 7th centuries, while all other mss. in that period contained (or acquired) only the SE—and then these two traditions became combined in the later middle ages (exactly as would happen in the Greek, and possibly even inspired thereby). Thus if the Ethiopic translation began without the LE, all the evidence is easy to explain, but if it began with the LE, that same evidence is harder to explain. Therefore, the original Ethiopic tradition probably lacked the LE. And even if not, it must still have begun with explicit knowledge of the fact, by outright stating it, that many of the mss. it was translated from lacked the LE.

In Latin

The late-4th century Vulgate translation contains the LE, but the earliest Latin translation lacks it: the 4th century Codex Bobiensis contains only the SE (altered, as noted in section 2.3). This represents a translation dating at least as far back as the 3rd century and possibly even the late 2nd century (based on telltale evidence in the mss., according to experts who have examined it), which establishes that the absence of the LE predates the 4th century (and possibly even the 3rd). This demonstrates that the LE did not exist in the exemplar used by one of the earliest Latin translators. Codex Vercellensis dates from around the same time, containing yet another Latin translation (thus originating from a different Greek archetype), yet it, too, lacked the LE. Vercellensis actually had a page containing the LE tacked into it by a later scribe. Experts have verified that the original leaves lacked the space to have ever contained the LE before this.34 Thus the two oldest Latin mss. (which are in fact older than even most Greek mss.) directly attest the absence of the LE.
Other non-Vulgate Latin translations (collectively called Old Latin) contain the LE, but all extant mss. of these are of late date. The only early mss. in this category date from the 5th or even as late as the 6th century, exhibiting translations made in the 3rd or 4th century (though we still can't confirm the LE was in these original translations). There are only three of these: Codex Bezae, Codex Sangallensis 1394, and Codex Corbeiensis II.35 All these Old Latin mss. are thus late enough that they could have had translations of the LE added onto them well after it had already become popular in Greek mss. (just as happened in every other translation tradition). Or any of them could have been translated from a copy of Mark containing the LE circulating in the 4th century (see section 5.3.10). Otherwise, the Bobiensis and Vercellensis translations predate these, and they lacked the LE. Only a century or more later does the LE appear in any Latin translations (just as we see in the Syriac and Coptic traditions), and in every case these later translations either derive from a time after the LE was already being accepted as the ending of Mark (e.g. the Vulgate was translated by Jerome exactly when the LE was starting to become popular in the Greek: see section 5.3.12) or are suspect as later additions. As noted above, we can already see one case of the LE being surreptitiously 'inserted' into a Latin tradition. So we have good reason to suspect this is how the LE may have ended up in other Old Latin texts—because the oldest Latin mss. and translations lacked the LE.

In Georgic and Armenian

The oldest Georgic manuscript (dating to the 9th century) lacks the LE. The LE starts to appear in the Georgic tradition a century later. The Georgic translation is believed to have been made in the late 5th century, and not from the Greek but from the Armenian translation, which was made in the early 5th century by Mesrop Mashtots, itself originally from a Syriac translation, later corrected against the Greek. Although extant Armenian manuscripts are much later, most of them (nearly a hundred) lack the LE, including the earliest. Based on the trend already exhibited by the Latin, Syriac, and Georgic (and the trend evident in Coptic and Ethiopic), this suggests the LE was not known to Mesrop and only added later. This agrees with the fact that most Armenian mss. lack the LE (the LE being added so late, it had less time to propagate) and the fact that the earliest Georgic mss. lack the LE (having derived from the original Armenian, which thus must have lacked the LE), which in turn confirms the Syriac began without the LE (as the Armenian translation was originally based on it), which further argues the Diatessaron lacked the LE (see sections 5.1.2 and 5.3.3).36 These translation traditions are very early and wildly diverse geographically and culturally, and in every case the absence of the LE is earlier. Though the Armenian and Georgic ultimately derive from the earliest Syriac translation of the late 2nd century, the Latin and Coptic and Ethiopic are all independent of that, and yet all of these attest the LE was not commonly known until the 4th century. That this is directly confirmed by two expert witnesses (Eusebius and Jerome, per sections 5.3.10 and 5.3.12) settles the fact. This supports the conclusion that the LE was a late addition to the text of Mark.
Corroborating this conclusion is the fact that an Armenian author, Eznik of Kolb, quotes the LE in the middle of the 5th century, a decade or two after the Armenian Bible was translated, yet he does not quote any known translation of the Bible, but composes his own, possibly from a Greek original, which verifies the Armenian translation originated without the LE. And since we already know there were Greek mss. of Mark containing the LE at that time, Eznik's awareness of it affords no proof of its originality.37 Likewise, an Armenian translation of the Syriac of Aphraates a few decades after Eznik also attests the LE, but that also doesn't derive from the Armenian Bible, but a late Syriac copy of the Diatessaron (see sections 5.3.3 and 5.3.11).

In Gothic

The only early translation that likely began with the LE is the Gothic: a 6th century Gothic ms. (the Speyer fragment of Codex Argenteus) attests the LE in a translation probably made by Ulfilas shortly after 348 A.D. in what is now Bulgaria (just north of Greece). But as we know there were mss. of Mark containing the LE by then (see section 5.3.10), this only confirms the rarity of source mss. containing the LE, as apparently only one early translation tradition began with one (apart from perhaps one or two Latin translations of the 3rd or 4th century: see section 5.1.5).38

The SE-LE Sequence and the Robinson Thesis

The existence of the SE in numerous mss. (in several languages, including the original Greek) entails there were many root mss. that lacked the LE. The invention of the SE itself entails the LE was absent very early in the history of the text, necessitating the creation of the SE in order to address growing dissatisfaction with the OE. An even more essential clue is that all the manuscripts that include both the SE and LE always place the SE before the LE, whether in Greek or any other language (see section 5.1.4). Since the SE was most likely created by an author unaware of the LE (thus all these mss. still attest that the LE did not exist in earlier copies of Mark), any manuscript that places the SE before the LE has clearly added the LE, i.e. their ultimate 'source manuscript' (or archetype) must have contained only the SE, to which the LE was appended later. This is decisively confirmed in the physical evidence of the mss. (see section 5.2). It's also inherently obvious. No one would interpolate the SE before the LE anyway (see sections 4.3.4 and 5.1.4, and following). A large number of manuscripts containing the LE thus attest to the previous absence of the LE in the very act of including it. This happens to include numerous Greek and Latin manuscripts, and most Ethiopic manuscripts, and the earliest Coptic manuscripts that even contain the LE at all. When all those examples are thus rightly excluded, the evidence from all the earliest mss. (and translations) strongly favors the LE being a late addition to Mark.
This evidence is fairly damning. Which is why Maurice Robinson desperately advances the claim that this universal sequence (SE followed by LE) is explained by a lectionary use of the SE as a forged "optional ending" (in Black, PEM, pp. 58-59). Thus he can maintain the LE was the original ending. But his theory is too absurd to credit. Indeed, it's incredible five times over:
(1) It's implausible to presume all extant mss. (even in the various translation traditions) derive from a lectionary (which at any rate would be special pleading, and that against all probability).

(2) There is no evidence of such a practice (of providing an optional shorter ending to a whole story, much less interposed before the longer genuine one) in any lectionary. So his theory is not only wholly without precedent, it stands against all extant precedent; indeed, his own evidence of editing in lectionaries contains no instance comparable to what he is proposing: the insertion of an entire elaborate verse from whole cloth (cf. Black, PEM, p. 59 n. 74).

(3) It's self-defeating. Such a practice would entail Christians so little valued the canonical text of their scriptures that they felt free to substantially alter it just to suit lectionary convenience, and then let this error infect all other Bibles in the whole of the world, and that without any marginal note explaining the fact, but instead passing off the alteration as "according to Mark," which fact if accepted undermines rather than supports the authenticity of the LE, as it ensures Christians would have no compulsion against inventing the LE for the very same reason Robinson alleges would motivate them to invent the SE. Moreover, Robinson's theory guarantees pervasive biblical errancy. It thus kills the doctrine of inerrancy in the very effort to save it.

(4) It's directly refuted by the physical evidence in the manuscripts themselves, which uniformly declare a divergence of mss. and not a reliance on lectionary practice (see section 5.2), and by the testimony of Eusebius and Jerome (see sections 5.3.10 and 5.3.12), who would certainly not be so uninformed as Robinson's theory requires them to have been—for if his theory were correct, we would have heard it from them. To the contrary, Eusebius and Jerome don't even know about the SE, and know only mss. with or without the LE. If the SE originated in texts with the LE, their testimony would be impossible. As their testimony exists, it's Robinson's theory that's impossible.

(5) It suffers the final defect that the problem this egregious and implausible doctoring of the text is supposed to have solved (not wanting to end a daily reading at such a defeatist place as verse 8) would have been far more easily and plausibly solved by simply ending the lection at verse 10 (or even verse 6 or 7), a solution so vastly more probable that Robinson's theory fails even on the mere consideration of its prior probability. Indeed, as Darrell Bock notes, "The liturgical unit of Mark 15:43-16:8 is not long (13 verses). So why cut it off at v. 8?" (Black, PEM, p. 133). Indeed. Why not just continue all the way to verse 20? Clearly Robinson so badly wants his theory to be true that he can't even see how ridiculous it is.
So Robinson's theory is to be rejected. We must conclude that the universal presence of the SE before the LE (where they appear together) argues against the authenticity of the LE.

Assessment of Textual Evidence

Combine the above fact with all the more direct evidence that the earliest mss. and traditions lacked the LE, and we have a strong external case against the authenticity of the LE. There are only three theories that can explain all this evidence: (1) neither the LE nor the SE were in the original text of Mark (and are therefore forgeries, either of composition or insertion); (2) either the LE or SE was original to Mark but then lost by accident, and very early (and whichever was original, whether LE or SE, the other is not original and therefore a forgery); (3) either the LE or SE was original to Mark but then deliberately removed, and very early (and whichever was deleted, whether LE or SE, the other is not original and therefore a forgery). Thus, no matter which theory you adopt, you cannot escape the conclusion that Mark contains a forgery. Inerrancy is thus defeated.
And only the first theory is credible. Not only does all the other internal and external evidence confirm this, but the other two theories are deficient. The SE is not likely to have been accidentally lost, as it is much too short. Even the LE is too short. The loss of a codex page could destroy up to four whole columns of text, but the LE consumes not even two; the SE, a mere fraction of one. And early loss from a scroll is prohibitively improbable (the ending would be on the inside of the roll, attached to the cog, the least likely section to lose). The SE is also unlikely to have been deliberately removed, because it cannot possibly have contained anything anyone would want to remove. Even the LE is unlikely to have been deliberately removed, for though it contains some content that might have been undesired by some (though its prevalence in the record suggests hardly anyone disliked its content, rendering that theory implausible from the start), most other instances of motivated deletion in the manuscript tradition involve excising only the offensive material, leaving the rest—or simply altering the material to be agreeable. This is particularly evident in how material in Mark was redacted by Luke and Matthew, and how passages in Mark were emended by later scribes.39 Only occasionally did anyone delete whole sections of Mark, and not (so far as we can tell) because they were doctrinally offensive. Thus, for example, if the remark about handling snakes was offensive, we would more likely find manuscripts in which simply that one phrase or verse was removed (as indeed it was in one 15th century lectionary), or if Jesus upbraiding the Apostles was offensive, we would find altered manuscripts in which Jesus simply didn't upbraid them. If the transition was recognized as awkward, we would find manuscripts in which this was repaired by emendation. And so on. In other words, deliberate deletion cannot explain the loss of the whole LE. Hence the second and third theories are improbable, while the first theory is very probable. That it is fully corroborated in the remaining evidence (internal and external) only confirms this. Therefore, Mark did not write the SE or the LE.

The Manuscripts: Physical Evidence

Apart from the textual evidence of the manuscripts, the surviving manuscripts also contain physical clues to the late origin of the SE and LE. Annotations to this effect are actually found in numerous mss. Some mss. indicate the end of the Gospel after 16:8 by subscribing the title of the book there or placing some other symbol there (the same ways the ends of other Gospels were indicated), and then follow that with the LE (or SE and LE), demonstrating the scribe was aware of the fact that the LE (or even SE) was not originally the ending of Mark.40 This practice is evident even in other languages, including the Coptic and Ethiopic (see sections 5.1.3 and 5.1.4) and the Armenian.41 The most likely explanation of this strange juxtaposition is that their ultimate 'source manuscript' lacked the LE originally (and thus had the concluding subscription after 16:8), and then the LE (or SE & LE) was added by a second hand (i.e. a later scribe than the one who originally transcribed the ms.), and when this whole collage was copied out again it was simply copied verbatim in exactly that order (by a third scribe, transcribing either the ms. we have now or the archetype from which ours ultimately derives). That entails each 'source manuscript' lacked the LE, and the LE was snuck in later on. We actually have examples of this process in the making: actual mss. in which the LE was clearly added later in a second hand.42 We even have a medieval scribe confessing to doing this (see section 5.3.10).
Not only is the LE (or SE & LE) "often separated from 16:8 by scribal signs" like these but in some mss. there are actual "notations that state or suggest that what follows is not found in some witnesses," e.g. minuscule 199 (from the 12th century) says "in some of the copies this [the LE] is not found; rather, it stops here."43 Some of these notes derive from common ancestors, but even counting archetypes there are numerous independent notations like this, and (as just noted above) many more indicators in other mss. besides these explicit scribal notations. This confirms that numerous root mss. lacked the LE. And though in some medieval mss. there are scribal notes claiming the LE is the older reading, by then it may have appeared to be—especially to medieval scribes, who only had a few mss. to compare and no knowledge of the modern science of textual criticism. In addition to scribal markings in many mss. and scribal notes in many other mss., some mss. (like minuscule 274, and several Syriac and Coptic mss., and in a similar way even Codex Regius, commonly known as manuscript L) add the SE in the margins as an alternate ending.44 This also suggests knowledge of other now-lost mss. in which Mark ended only with the SE. The scribe of L is the most explicit, concluding Mark at 16:8 with a dotted line in one column, and then using the other column for endnotes stating that "some" mss. "also" had the SE (by itself) and that others had only the LE (and in each note providing the text of the respective ending), which could even mean L's exemplar had neither, but at the very least it means some mss. had the SE by itself. Similarly, the 7th century manuscript 083 ends with the SE and then adds a note "there is also this, appearing after 'and they were afraid'" and appends the LE. The SE and LE are even found attached in some mss. to the ending of Gospels other than Mark (usually Luke or John).45 This is most peculiar, and a fact that may be a clue to the origin of the LE.

Ariston the Presbyter

In a 10th century Armenian ms. the LE is uniquely separated from the rest of the Gospel with a note saying 'of Ariston the Presbyter'. This note appears to have been added to that ms. by a later scholar in the 13th or 14th century, and thus could be a mere conjecture.46 But it would be a strange thing to conjecture—in fact, the only plausible motive for anyone to scribble this in the margin would be their discovery that it was true. Although Metzger concludes "the probability that an Armenian" scribe of such late date "would have access to historically valuable tradition on this point is almost nil" (TNT, p. 325), that's not a sound argument, because it's even less probable that an Armenian scribe of any date would write such a note unless he did have a 'historically valuable tradition' confirming the very point being noted.
The name most likely refers to Aristion, an early 2nd century Christian elder who may have written lost commentaries on the Gospels.47 Some scholars conjecture instead that it refers to an 'Ariston' believed to be an actual disciple of Jesus, and thus (the note would be claiming) the LE was written by an eyewitness. Although passing his testimony off as Mark's would still be an act of forgery, it would also be foolish, since to pass off eyewitness testimony as instead the testimony of another author (Mark), whom everyone believed wasn't an eyewitness, would actually diminish that testimony's authority. There is thus no reason for any disciple to have done this, nor does the LE read at all like an eyewitness report (quite the contrary, as shown in section 4.3). This conjecture is thereby implausible. There is no evidence to support it anyway.
A 2nd century author is far more likely. There were two men of similar name around the same time (early-to-mid 2nd century): a certain Aristion the Elder, who (as noted above) may have written a commentary on the Gospels, and an Ariston of Pella, who composed a now-lost Dialogue of Jason and Papiscus the Jew, which was known to Origen and Jerome and which many scholars suspect was employed by Justin Martyr. Either would explain any use Justin may have made of the LE, i.e. if the LE originally appeared in either of those works (the Commentaries of Aristion or the Dialogue of Ariston), Justin could have employed it without having any idea of it being passed off later as the ending of Mark (see section 5.3.2).

Accidental or Deliberate Transfer

There is an actual commentary on the Gospels that does survive (from another author), on which Maurice Robinson observes, "the primary matter [in ms. 304] is the commentary. The gospel text is merely interspersed between the blocks of commentary material, and should not be considered the same as a 'normal' continuous-text MS. Also, it is often very difficult to discern the text in contrast to the comments" and "following gar at the close of [16:8], the MS has a mark like a filled-in 'o', followed by many pages of commentary, all of which summarize[s] the endings of the other gospels and even quote[s] portions of them" before continuing on (emphasis mine).48 Note the eerie relevance of his remarks: it was often difficult to tell where the Gospel text ended and the commentary began, and commentaries on the Gospel of Mark naturally inspired commentators into summarizing the endings of the other gospels, a perfect description of the LE. Could someone have deliberately (or even accidentally) copied out a paragraph from such a commentary and inserted it into an actual copy of the Gospel? Like, say, a commentary by an Aristion whom at least one medieval scholar had reason to believe originally wrote it?
Even Bruce Metzger has suspected something like this, concluding that "in view of the inconcinnities between verses 1-8 and 9-20, it is unlikely that the long ending was composed ad hoc to fill up an obvious gap; it is more likely that the section was excerpted from another document, dating perhaps from the first half of the second century" (TCG, p. 125). A brief summary and harmonization of all the actual appearance narratives (from the other three Gospels and Acts), is exactly the sort of paragraph we might expect to find in a Commentary on the Stories of the Lord (such as Aristion may have written), or even in Ariston's Dialogue of Jason and Papiscus, such as a summary of appearances from the extant Gospels placed in the mouth of the dialogue's Christian advocate Jason (which might even have been later mistaken as a quotation of the LE, once it had crept into some mss. of Mark). Its extraction and transfer to Mark would not be unheard of in ancient practice, particularly for someone keen on borrowing a more satisfying ending. This is all the more credible when we observe that the Western text, in which the LE first appears, typically placed Mark at the end of the Gospels, thus inviting the need to summarize (and harmonize) the appearances of the other Gospels that would all have just been read.49
This theory would explain every single oddity in the evidence: (1) it would explain the origin of the Armenian scholar's marginal note (only if he found the LE in its original context—whether an actual work by Ariston, or by some later author who clearly indicated deriving it from Ariston—would he be likely to have made a note attributing it to such an obscure author, especially an author who wrote early enough to actually be the LE's author, which is otherwise a remarkable coincidence); (2) it would explain the fact that the LE mysteriously became appended to other Gospels, not just Mark (as if originally it was not associated with Mark alone but all the Gospels, as a commentary would be—in fact, the passage may have originally been appended to the Gospels as a whole and thus became attached to whatever Gospel ended each individual collection, which in the majority Western text was the Gospel of Mark); (3) it would explain the fact that the LE shows no awareness of having just followed verses 16:1-8 (as noted in section 4.1); (4) it would explain why the author of the LE made no notable effort to emulate Markan style (and yet exhibits influence from the style of all the texts of the NT, including Mark); (5) it would explain the LE's brief summarizing character and its evident harmonizing intent (the LE reads just like a paragraph taken out of context from a commentary on the Gospels, or even a dialogue in which their content was summarized); (6) it would explain the LE author's knowledge of the whole NT (and why he limits his summary of accounts to stories appearing only in the Canonical NT); (7) it would explain the LE author's manifest assumption that his readers must know or have access to the NT (especially if the LE appeared in a commentary on the Gospels or, as in the Western text, Mark was positioned at the end of them, for then those other stories would already be in the reader's hands—being, in fact, in the very same book); (8) it would explain why the OE-to-LE transition is both illogical and ungrammatical (which makes no sense for a deliberate forger of the LE but makes perfect sense if the LE was simply cut and pasted from another book); (9) it would explain why all the physical evidence in the mss. suggests the LE began as an appendix to Mark and not an actual continuation of Mark's narrative; (10) and, of course, it would explain why all the indications are that the manuscript tradition for Mark originally and widely lacked the LE.
The LE therefore almost certainly derives from another work (whether of Aristion, Ariston, or someone else) and was transferred to the end of Mark and thus mistaken (or passed off) as Markan material.

The Patristic Evidence

That leaves only one more category of evidence: the Patristic. A major problem with relying on Patristic authority is that the manuscripts of the Church Fathers have themselves been doctored to reflect later canonical readings of the Bible. This is particularly a problem for the mss. of Irenaeus, which is thus a problem for the ending of Mark because Irenaeus is the only 2nd century author who clearly attests the existence of the LE.
The MS traditions of virtually all the church fathers show that later copyists tended to "correct" quotations of the Bible to the form of text prevalent in their own day. Consequently, Patristic writings that survive only in Medieval MSS or that are available only in uncritical editions, such as Migne's Patrologia Graeca, are of practically no value for establishing the original wording of the NT.50

Before patristic evidence can be used with confidence, however, one must determine whether the true text of the ecclesiastical writer has been transmitted. As in the case of the New Testament manuscripts, so also the treatises of the fathers have been modified in the course of copying. The scribe was always tempted to assimilate scriptural quotations in the fathers to the form of the text that was current in the later manuscripts of the New Testament.51
Quotations in the Church Fathers also commonly contradict each other and are in other ways notoriously unreliable. We even have some confirmed instances in which later Christian redactors added entire sentences or paragraphs to an earlier Patristic text.52 While the manuscripts we have now exhibit several very different textual traditions of equal antiquity, only the Western Text (whose best extant representative is Codex Bezae, although it still deviates from the Western text-type in numerous ways) is most commonly used by early Patristic authors (especially Justin, Irenaeus, and Tertullian), "all of which are characterized by longer or shorter additions and by certain striking omissions," while other text-types, such as the Alexandrian, may be closer to the originals.53 Patristic authors after the 4th century are also of no use in the present case, since we know the SE and LE were circulating as endings of Mark by then, and manuscripts containing them were growing more numerous thereafter, eventually eclipsing altogether the original text of Mark. Keeping all these cautions in mind, only the following authors are of use in evaluating how Mark originally ended.

Papias

Papias reported the miracle of Justus Barsabbas drinking poison and coming to no harm "by the grace of the Lord," which is sometimes cited as evidence Papias knew the LE (see section 4.3.1). But in fact this entails Papias didn't know the LE, which further argues the LE did not exist in the Gospel of Mark at that time (early 2nd century). For Papias definitely knew the Gospel of Mark.54 Yet he credits all this information to oral tradition, not the Gospel of Mark, and shows no knowledge of Jesus having predicted it (which surely he would mention) or that this was in any way a common sign among apostles. He instead appears to have reacted as though the effect were a surprise (uniquely "by the grace of God" and experienced by Barsabbas alone). Though we do not have a full direct quote from Papias to confirm these conclusions, they seem undeniably apparent from Eusebius' account of them. This then stands as evidence against the authenticity of the LE, not in favor of it. To the contrary, the LE's inclusion of immunity to poison may have been inspired by stories like this, not the other way around (see section 5.3.6).

Justin

The earliest author usually cited is Justin Martyr (c. 160 AD), but he provides no real evidence of the presence of the LE in Mark. In only one passage (Apology 1.45.5) he uses together the same three words appearing in Mark 16:20, but does not indicate he is quoting any Gospel there, much less Mark. Where Justin mentions the OT had predicted "the powerful word that His Apostles preached everywhere after having left Jerusalem," the "preached everywhere after having left" is all that echoes the LE, just three words in Greek and not even in the same order (the LE word logos is also used by Justin elsewhere in the same sentence but is common and expected here and thus not telltale). In contrast, the LE does not have the words 'Apostles' or 'Jerusalem', nor does Justin mention anything else that would suggest knowledge of the LE (such as the specific signs declared there, or even its appearances of Jesus). The similarity thus appears to be coincidental, or at most evidence of an idiom in wide use that separately influenced both Justin and the author of the LE (the predicted sentiment is already inherent in Luke 24:47-52 and Acts 1:8, as well as Matthew 28:19). Moreover, even if we could consider this as evidence of the LE's influence on Justin, as noted in section 5.2, Justin may have only known the LE in a text other than Mark. For Justin doesn't in fact say he is citing a Gospel, and everyone agrees he is not quoting one. Therefore, this passage cannot demonstrate the LE was in Mark at that time. Of course, even if it was, it could have been appended to copies of Mark in the early 2nd century, so even a direct quotation from Justin would be insufficient to establish the LE was originally in Mark. But we don't have any such quotation anyway.55

Tatian

It's possible that Justin's pupil Tatian incorporated the LE in his Diatessaron (or 'Harmony of the Four Gospels') after 175 A.D. But we cannot confirm that this was originally the case, as we do not have Tatian's version of the Diatessaron.56 We know the Diatessaron had additions and changes made to it over the centuries, few versions agree, and the texts we have now date centuries after Tatian. For example, a famous interpolation in John (on the adulteress, John 7:53-8:11) was evidently not originally in the Diatessaron, yet found its way in centuries later.57 The LE may have done the same. In fact, different textual traditions have the LE incorporated into the Diatessaron in different ways. Other sections of the Diatessaron also differ among the various textual traditions. So it does look like the LE was added later by different editors in different ways. At best the earliest references to the LE being in the Diatessaron appear in the works of Aphraates and Ephrem in the mid-4th century (though these are to some extent questionable: see the footnote in section 5.3.11 below). But they also attest to many other interpolations in their copies of the Diatessaron (i.e. many passages that do not now exist—and certainly did not originate—in any of the four Canonical Gospels). Thus the Diatessaron had already become corrupt by then. It is therefore of little use in determining the origin of the LE. Moreover, even if Tatian incorporated the LE, that would only confirm that it had entered some mss. of Mark by mid-2nd century, which still would not establish that it was originally a part of Mark almost a hundred years earlier.

Tertullian

Supposed evidences of Tertullian's knowledge of the LE (c. 190 A.D.) are invalid because they can more easily derive from the other Gospel texts and Christian teachings that the LE itself drew upon. Passages from Tertullian exhibit no features distinctive of the LE, nor give any indication Tertullian is quoting anything, much less the Gospel of Mark. Tertullian, On the Soul 25.8 (seven demons expelled from Mary) derives from Luke; On the Cure for Heretics 30.16 (Apostles given powers) derives from Acts and the Epistles; On the Resurrection 51.1 and Against Praxeas 2.1 and 30.5 (Jesus rising to sit at the right hand of God) obviously derive from Acts 1:11 (and Mark 12:36 and 14:62), not the LE; likewise, On Fleeing Persecution 10.2 (believers given power over demons) can just as easily derive from Mark 6:7 and elsewhere. See table in section 4.3.1 for obvious alternative sources. Other references are even less relevant, e.g. On Baptism 10.7 explicitly interprets a saying of John the Baptist, not Jesus, and conspicuously lacks reference to the Gospel of Mark; and Tertullian's remarks here otherwise doesn't resemble the LE at all, but merely echo standard Christian belief of the time. In other words, we can't argue from any of this evidence that Tertullian knew the LE. To the contrary, the absence of any direct reference in any of these passages to what Jesus says in the LE argues Tertullian didn't know the LE. For a declaration of Jesus on these facts would have clinched Tertullian's point in almost every case, which makes the absence of the LE in these passages far more telling. Nor can we argue from any of these passages that Tertullian knew the LE was in Mark, for like Justin, even if we could prove Tertullian knew the LE (and we can't), that would not prove he knew it as the ending of Mark, rather than as a text in some other work (see section 5.2).

Irenaeus

The only other relevant author from the 2nd century is Irenaeus (c. 185 AD). He appears to provide the only reliable evidence that the LE was in any copies of Mark in the 2nd century. But the mss. of Irenaeus are notoriously corrupt and problematic. He only mentions the LE once, and that in a passage that only survives in Latin translation, yet the Latin texts of Irenaeus are among those most tampered with. The claim has been made that Theodoret of Cyrrhus (c. 450 A.D.) quotes this passage in the original Greek, confirming that if it had been interpolated, it happened in the Greek before the Latin translation was made (which would certainly be possible). But this is not in fact true. Theodoret's quotation is from a previous section of Irenaeus, not this one.58 It has also been claimed this passage is quoted in Greek in a marginal note added next to the LE in a medieval Bible, but that's also not true.59 The scholium in question only says "Irenaeus, who was near to the apostles, in the third book against heresies quotes this saying as found in Mark." It does not quote the text of Irenaeus. As that ms. dates to the mid-10th century (and the author of it's marginalia dates no earlier than the 5th century), this testimony confirms nothing, for the referenced passage could be an interpolation made anytime in the two hundred years or more after Irenaeus wrote—even in the Greek, yet for all we know this scholar could be referring to a Latin text of Irenaeus.60 So we have no Greek text of this passage. It exists only in the medieval Latin.
Certainly, on its face we would still accept this passage as confirmation that Ireneaus' copy of Mark by the late 2nd century contained the LE. But there is a persuasive argument to be made that this passage was not written by Irenaeus but interpolated (at least within two or three centuries, or even later), quite possibly by accident. The passage looks like a marginal note added by a scribe intending to add to Irenaeus' arguments in that chapter. As there was no standard notation for distinguishing marginal notes from accidentally omitted text, we have countless examples of such notes being accidentally interpolated into the text of other manuscripts. This could be one such case. According to manuscript specialist F.W. Hall, "the casual jottings of readers and correctors are often imported into the text," hence in his manual on textual criticism he dedicates an entire section to "Insertion of interlinear or marginal glosses or notes" as a common cause of erroneous interpolation in manuscripts. Robert Renehan agrees, "marginal confusions...occur frequently in mss.," giving several examples (e.g. § 35 shows several "marginal scholia which have been incorporated into the text" of the letters of Epicurus, in some cases entire sentences). In his own brief survey, Miroslav Marcovich documents at least 33 examples of this kind of mistake in the works of the early Church Fathers.61
To understand why the passage attesting the LE in Irenaeus may be an interpolation, the entire section must be quoted to reveal the flow of Irenaeus' argument, and why the LE does not appear to fit. Before this Irenaeus has spent an entire chapter arguing that Jesus is God and there is only one God in Jesus, extensively quoting the NT and OT, every instance confirming his thesis that he can find. He then concludes (emphasis added):
Wherefore also Mark, the interpreter and follower of Peter, does thus commence his Gospel narrative: "The beginning of the Gospel of Jesus Christ, the Son of God; as it is written in the prophets, Behold, I send My messenger before Thy face, which shall prepare Thy way. The voice of one crying in the wilderness, Prepare ye the way of the Lord, make the paths straight before our God." Plainly does the commencement of the Gospel quote the words of the holy prophets, and point out Him at once, whom they confessed as God and Lord; Him, the Father of our Lord Jesus Christ, who had also made promise to Him, that He would send His messenger before His face, who was John, crying in the wilderness, in "the spirit and power of Elijah," "Prepare ye the way of the Lord, make straight paths before our God." For the prophets did not announce one and another God, but one and the same, under various aspects, however, and many titles. For varied and rich in attributes is the Father, as I have already shown in the book preceding this, and as I shall show from the prophets themselves in the further course of this work. Also, towards the conclusion of his Gospel, Mark says: "So then, after the Lord Jesus had spoken to them, He was received up into heaven, and sat on the right hand of God," confirming what had been spoken by the prophet: "The LORD said to my Lord, Sit Thou on My right hand, until I make Thy foes Thy footstool." Thus God and the Father are truly one and the same: He who was announced by the prophets, and handed down by the true Gospel, whom we Christians worship and love with the whole heart, as the Maker of heaven and earth, and of all things therein. (Irenaeus, Against All Heresies 3.10.5)
Note that before the sentence in bold, Irenaeus appears already to have concluded his argument. Yet then, out of the blue, he adds, as if an afterthought, "Also..." and quotes Mark 16:19 as verifying Psalms 110:1 (which had already been verified, and by Jesus himself, in Mark 12:35-37). Even more strangely, in none of this additional sentence does the word 'Father' appear, yet this passage is supposed to support Irenaeus's argument that 'thus God and the Father' are one and the same, because this is the argument of the preceding sentence, and the conclusion declared in the following sentence. If the material in bold is removed, we have a consistent argument from premise to conclusion. But reinsert the material in bold and there is an illogical disconnect between the argument Irenaeus is supposed to be making, and the passage being quoted—because that passage does not support this argument.
So why is it here? It would make sense as an addition to the whole theme of chapters 9 through 12, but it makes no sense appearing exactly here. And even though his section 3.10.5 is where we would expect all his quotations from Mark to appear, this particular quotation still does not fit the specific argument Irenaeus is making. It would support only a different argument, albeit one that would reinforce his overall thesis, and thus should appear as a separate argument either before or after the present one, not inexplicably inserted in the middle of it. But a scholar who wished to add reinforcing evidence from Mark to Irenaeus' overall theme would certainly place it in the margins of section 3.10.5, if he would place it anywhere. Which would explain how it later came to be so arbitrarily inserted into the text.
In further support of this conclusion, Irenaeus knows that his argument from Mark 1:1-3 requires considerable elucidation (consisting of several sentences), but the comparably required elucidation of his supposed argument from Mark 16:19 is missing. As written, the text in bold actually refutes rather than supports Irenaeus, for it plainly says Jesus was a different entity from God (sitting next to him, not in his place, and addressing each other in the third person), and it is not explained how the Psalm quoted makes any different conclusion out of this. Irenaeus would have needed to explain the connections here: how the Psalm supports reinterpreting Mark 16:19 as a confirmation rather than refutation of the thesis that Jesus and God are one and the same. He would certainly have called into service Mark 12:35-37, and explicitly identified the links we are supposed to make between the different Lords named and God and the Messiah and why we are to presume David is speaking of the latter in Psalm 110:1. Yet none of this is present. A marginal note would easily consist of a single sentence, leaving the connecting arguments implied, but Irenaeus himself would not likely deliver such a presumptuous and unfinished argument, especially one so manifestly supporting his opponents (the heretics he is here engaged in refuting). The fact that it doesn't even support the argument it is attached to only confirms the conclusion that Irenaeus didn't write this. I conclude this testimony is probably spurious.
Another passage in Irenaeus is sometimes adduced as evidence he knew the LE, but the passage in question actually argues against such knowledge.62 For it neither quotes the LE, nor uses the same vocabulary as the LE, nor even implies he is drawing any information from the Gospels at all—for he is providing his own description of current activity in the Church, which he lists not as exorcism, speaking in tongues, immunity to poison, and healing, but exorcism, prophecy, healing, and resurrecting the dead (and each described elaborately), thus showing no congruity with the LE. His list simply reflects common Christian practice and belief at the time.63 And since his point is that these powers prove the Christian gospel true, the fact that Jesus himself had said so (16:17-18: "these signs shall accompany them that believe," thereby confirming the truth of the gospel) would so soundly secure his argument that for him to neglect citing it here is patently strange. This all but proves he did not know the LE. Similarly, Irenaeus mentions "speaking with all kinds of tongues" as a power Christians displayed, but only far away from this list, in a completely different book, showing no awareness that this was ever predicted by Jesus, much less in the same place as healing and exorcism. The phenomenon is already ubiquitously discussed in the Epistles and obviously still going on, so this passage does not attest knowledge of the LE. Indeed, again, this argues against such knowledge, since here as elsewhere he fails to associate the powers listed in the LE, and fails to mention that these powers were predicted by Christ himself. That he never shows any knowledge of 'immunity to snakes and poison' being a power any Christians should or did have only confirms the point. So from these passages as well it seems much more likely that Irenaeus did not know of the LE.

Hippolytus

Hippolytus (c. 210 A.D.) refers to eaters of the Eucharist becoming immune to poison, which is said to demonstrate knowledge of the LE, but it cannot be anything of the kind.64 It neither quotes the LE, nor mentions snakes, nor even attributes the claim to Jesus, despite the supreme authority this would establish. And unlike the LE, Hippolytus associates the power with the eucharist, not baptism. Since other tales of immunity to poison were already circulating (see sections 4.3 and 5.3.1), the LE is not the only possible source of Hippolytus' claim. In fact, given the incongruities, it's the least likely source for it. Instead, unless the author of the LE was making that claim up out of whole cloth (and thus the claim is completely false, which then refutes inerrancy), the author of the LE must have been drawing on independent traditions regarding immunity to poison, which traditions could just as well be what informed Hippolytus. And even if there was no such tradition, an inference to this same conclusion from Luke 10:19 would be as obvious to Hippolytus as to the author of the LE (see section 4.3.1), so knowledge of the LE would not be required (and by his associating the power with the eucharist rather than baptism, his knowledge of the LE should even be rejected). Other alleged references to the LE in Hippolytus are nothing of the kind, e.g. that Jesus sat on the right hand of God (Treatise on Christ and Antichrist 46) derives from Luke-Acts (and elsewhere). Thus, there is nothing in Hippolytus that confirms the LE even existed, much less was known as the ending of Mark.

Origen, Clement, and Other 3rd Century Authors

There are a large number of Christian authors from 100-300 A.D. who never mention the LE, which taken together is significant but not compelling (since many NT verses are likewise unattested but still certainly authentic). But most telling is the silence of Origen and Clement (c. 200-230 A.D.), who each left us a huge corpus erudite with discussions and quotations of the Gospels. Similarly other copious authors, like Tertullian and Cyprian, erroneously believed to have attested the LE, in fact very curiously did not. Likewise Lactantius, despite his having written extensive treatises on Christian abilities and beliefs. Though it is always possible they just never happened to strike upon an occasion to reference the LE, given the vast extent of their respective writings this at least approaches the improbable, the more so when combined with the silence of all other authors before the 4th century (apart from, at most, Irenaeus, per section 5.3.5).
Clement actually had credible occasions to quote the LE yet didn't (e.g. Stromata 4.6; On the Rich Man 34; Comments on the Epistle of Jude; etc.). So his silence is notable, even if still not conclusive. Tertullian might likewise be expected to cite the LE in several passages yet doesn't (e.g. Against Marcion 5.8; Exhortation to Chastity 4; and the passages noted in section 5.3.4). Cyprian, too (see section 5.3.8). Origen also had occasion to quote or address the LE in his extensive treatise Against Celsus (e.g. 1.6, 1.67, 2.48, 2.56-70, etc.), but most especially where he had to rebut Celsus' claim that Mary was insane. Some now claim Celsus was there referring to Mary having once been possessed by demons (and hence he must be referring to Mark 16:9), but the context disproves this. In Against Celsus 2.55 Origen tells us Celsus said only two people saw the wounds Jesus had suffered, one woman who was paroistros ("driven frantic; beside herself" or "half-mad; practically insane") and one other man ("from among those engaged in the same charlatanry"). This is clearly the scene in John (the one man being Thomas), not the LE, which contains no reference to seeing wounds, nor any appearance to a single man. Origen assumes nothing else in his rebuttal (in 2.59-62). In fact, Origen's most direct rebuttal (in 2.60) is that Celsus' claim that Mary was insane is "a statement which is not made by the history recording the fact" but a calumny entirely made up by Celsus. That suggests neither Celsus nor Origen knew of Mark 16:9, which would be a historical record of the fact (directly declaring her an ex-demoniac). Moreover, Celsus would surely have lambasted the Gospel of Mark for including other material in the LE (such as its claim of immunity to poisons and snakes), compelling Origen to make a rebuttal. Yet instead all Celsus attacks is the account in John. So the silence here argues the LE was not known to Celsus, and it supports (though does not prove) the LE was not known to Origen.65

Vincentius (via Cyprian)

Cyprian reports that in 256 A.D. bishop Vincentius of Thibaris had said at a council that the Lord "commanded his apostles, saying, 'Go ye, lay on hands in my name, expel demons'. And in another place: 'Go ye and teach the nations, baptizing them in the name of the Father, of the Son, and of the Holy Ghost'." But the latter is an exact quotation of Matthew 28:19, while the former is not an exact quotation of Mark 16:17: the LE does not have Jesus giving this as a command (but rather as promising it as a sign, and descriptively in the third person, not in the imperative; the imperative is only used for his command to go and preach the gospel two verses earlier), and in the LE Jesus does not link laying on hands and the expelling of demons, but connects laying on hands and healing (a whole verse later). Hence Vincentius must be quoting some other lost Gospel or agrapha, or some tertiary source that conflated the contents of the LE (which, as already noted in previous sections, could have originated in some source other than Mark, and probably in the early-to-mid 2nd century), or he is conflating several passages from the NT (e.g. Mark 6:7-13 and 9:38-40 and Acts 14:3, cf. also Luke 10:17 and 9:49-50). Since Vincentius does not say this passage is in Mark (or even a Gospel) and the quoted words do not match those of the LE, we can derive no conclusion from this that the LE was then in Mark. To the contrary, that none of the many dozens of bishops quoted on the role and importance of baptism in this text ever quote the LE is rather an argument against anyone knowing it as scripture even by the middle of the 3rd century.66

Other Dubious Witnesses Before the 4th Century

Other documents from the 2nd century have been proposed as witnesses to the LE, but none are credible. The Epistle of Barnabas 15:9, which quotes no Gospel, merely says Jesus ascended the day he rose, which claim derives from Luke. Similarly Hermas 102 (Parable 9.25.2) contains common phrases shared with Mark 16:15, but used in a different way, in no similar order or even together, and without any indication of deriving any of this from any source, much less a Gospel (to the contrary, it is there portrayed as a direct communication from an angel). Like the reference in Justin (see section 5.3.2), this only looks like a common set of idioms and phrases in Christian preaching (derived from Gospel passages other than the LE: see table in section 4.3.1), which independently influenced Hermas, Justin, and the author of the LE. Similarly, the last verses of the extant fragment of the Gospel of Peter (vv. 58-60) do not attest the LE but in fact contradict it, saying there were twelve, not eleven disciples, no appearance to Mary, and Jesus doesn't appear to the disciples indoors at dinner, but outdoors in Galilee. All the contents of this passage are more clearly adapted from John 21 (or a common source shared by John 21) and possibly other passages tabulated in section 4.3.1.
The 3rd century Didascalia has one passage that comes close to the LE, reading (in Syriac), "But when we had divided the whole world into twelve parts, and were gone forth among the Gentiles into all the world to preach the word, then Satan set about and stirred up the People to send after us false apostles for the undoing of the word." But this is clearly not a quotation of the LE (Jesus is not even the one speaking), and it conspicuously does not conform to Mark 16:15 or 16:20. It appears to merely embellish Matthew 28:19, or simply derives from Justin (see section 5.3.2).67 Several passages in the anonymous Epistula Apostolorum (originally composed mid-2nd century) likewise bear no demonstrable connection to the LE (deriving instead from the other Gospels), and we have no reliable text of the latter anyway, only distant translations of it.68 And the extant portion of the Acts of Pilate that clearly employs the LE is unmistakably late. Some form of the Acts of Pilate may derive from the early 2nd century, but such cannot have been the text that cites the LE. Justin refers to the Acts of Pilate (in Apology 1.35.9 and 1.48.3), but the only part of the extant Acts of Pilate that could be of such early date is the appendix called The Report of Pilate to the Emperor Claudius which lacks any reference to the LE, yet contains all the material Justin claims to have found there. All other content of the extant Acts of Pilate dates from the late 4th century and later. So it is also of no use in answering the present question.69

Eusebius

So that leaves us with the 4th century and later. The next relevant author is therefore Eusebius (c. 320 A.D.). In his Letter to Marinus (a.k.a. Ad Marinum) Eusebius specifically addresses the authenticity of the LE. He says "it is not current in all the copies," and in fact not only do "the accurate copies" end at verse 16:8, but "nearly all the copies" do, the LE only "being rarely found in some copies." Eusebius was a renowned publisher of Bibles, supervised a scriptorium, and had charge of the most extensive Christian library then in the world, whose members had actively sought the gathering of countless manuscripts of the Bible on an ongoing basis for over a century (from Origen to Pamphilus to Eusebius himself), and Eusebius' authority on the Biblical text was universally accepted by his peers and successors. So the fact that he observed the LE to be rare, and not present at all in the most trusted manuscripts, proves that the later mss. tradition, in which most copies contain the LE, is a later medieval development.
This testimony supports the conclusion that the LE is not original to Mark, but was interpolated in only a few mss. sometime before the 4th century. The Eusebian Canons also exclude the LE, so Eusebius himself considered it non-canonical.70 And there is no valid basis for rejecting his testimony. He had seen vastly more manuscripts of the first three centuries than any modern scholar could ever hope to, and thus we are in no position to gainsay him. No early witness contradicts his testimony. And even if he could be exaggerating, he can't be lying, since Jerome corroborates him (see section 5.3.12), and Jerome would know, having extensive experience with even more manuscripts than Eusebius. In fact, were the evidence any different Eusebius would have defended the LE's authenticity, not doubted it, much less have supported that doubt with a lie. We must conclude Eusebius has given us a sufficiently accurate report on the state of the LE text. Eusebius shows no knowledge at all of the SE.
Eusebius' testimony alone is clear and authoritative, at least establishing the existence of the LE as of c. 300 A.D. Had it originated any later, Eusebius would have been aware of its recent appearance, but he shows no certainty as to its origin, so it can't have been composed and inserted later than the 3rd century. Accordingly, I find none of the later patristic attestations of any relevance. They merely repeat what we already know from Eusebius. Some even appear to have been using Eusebius as their source on the matter.71 Kelhoffer even shows how a remark attributed (possibly pseudonymously) to the 6th century author Victor of Antioch deviously rewrites the same argument from Ad Marinum into an argument for exactly the opposite conclusion, thus betraying knowledge of the Ad Marinum in the very effort to gainsay it. This very same passage from Pseudo(?)-Victor then confesses to having added the LE to manuscripts that lacked it! We can thus see how the LE came to proliferate in copies of Mark and the OE eclipsed.72

Aphraates, Ephrem, Ambrose

A certain Aphraates composed a collection of Demonstrations in Syriac, the relevant portion of which in 337 A.D. We don't have the originals, only much later copies, so we can't be sure he actually quoted the LE. Aphraates was employing the Diatessaron in some form. Ephrem the Syrian then composed a Syriac commentary on the Diatessaron about forty years later (c. 375 A.D.). Again we lack the originals and have only much later copies. Both seem to quote the LE and attest its presence in their copies of the Diatessaron (though one might still have doubts: see the following note). But we already know the LE had crept into copies of Mark by then (see section 5.3.10). So it could just as easily have also crept into the Diatessaron (see section 5.3.3 above). Neither author, therefore, affords any useful evidence regarding the origin of the LE.73
Ambrose (c. 375 A.D.) also unmistakably quotes the LE. But this again post-dates Eusebius, and Eusebius already attests the existence of the LE in some mss. of Mark.74 Although Ambrose never specifies what document he knew the LE from (and though he quotes it many times, he only ever quotes exactly the same section: vv. 15-18), from the evidence of Eusebius we can assume Ambrose found it in a copy of Mark and regarded it as of that Gospel. This and all later examples in Patristic sources afford no further evidence, as they merely corroborate what has already been proved: that after the 3rd century copies of Mark were circulating that ended with the LE.

Jerome and Later

The next relevant author is Jerome (late 4th century). In his Letter to Hedybia (Epistles 120.3) Jerome explicitly says essentially the same thing Eusebius did: the LE is not in most mss. and in none of the best (and he also shows no knowledge of the SE at all). He is almost certainly relying on Eusebius for this. But he would have known if Eusebius' observation was at all dubious, since Jerome's own acquaintance with the mss. was unrivaled in his day. So by approving what Eusebius said Jerome in fact corroborates him. Elsewhere Jerome says the VLE appears in some mss., especially in Greek mss., thus attesting the VLE was forged sometime in the 4th century (and thus barely one or two centuries after the LE itself was interpolated).75
Everything else after Eusebius is useless, only verifying what we already know: that by the 4th century the LE was circulating in some copies of Mark. The Apostolic Constitutions (AC) are of the late 4th century and thus of no use. Indeed, their primary source document, the 3rd century Didascalia, lacks any quotation of the LE (see section 5.3.9). Hence those elements were added to the AC later, exactly when the LE was circulating in copies of Mark.76 Later in the 5th century, Hesychius appears to reject the LE without argument (though only in a vague context), while in the 6th century Severus (in a work once attributed to Hesychius) essentially just repeats what Eusebius said about it (see section section 5.3.10 above). Since unlike Jerome their vast knowledge of the mss. is not established, their testimony doesn't independently corroborate Eusebius. The same goes for all subsequent Patristic sources.77 All other texts attesting to the LE (e.g. the late appendix added to the Pistis Sophia, which itself is in no way earlier than the late 4th century) are of such late date as to have no use in deciding the question. We know some copies of Mark ended with the LE in the 4th century, and this version began thereafter to gain in popularity. Hence further evidence only attests to what we already know. There is no other relevant evidence.

Assessment of External Evidence

From the Patristic evidence we can say with certainty that if the LE existed in the 2nd century, it was extremely rare and hardly anyone knew of it. And there is good reason to believe it was not then known as a part of Mark at all. The only evidence of such is a single passage in Irenaeus (which we have seen is of questionable authenticity) and its use in the Diatessaron (of which we have no 2nd century copies but only later corrupted ones). Only by the 4th century can we be certain it was clearly known, and known as an ending to Mark, yet even then it was explicitly known to be rare. Eusebius' account, paraphrased or gainsaid by many authors thereafter, is fairly damning: he certainly had access to numerous mss., and it was his observation that most mss. lacked the LE and that none of the mss. he trusted as the most reliable contained it. He does't even know about the SE.
Such was the state of the matter in the early 4th century. More than half a century later Jerome approvingly echoes Eusebius on this point. But we know Jerome also had access to numerous mss. and was a famous philologer and linguist (himself producing the Vulgate translation of the Bible still used by the Catholic Church and discussing in his letters many manuscripts and variants), so he would not have echoed Eusebius if he did not observe the same still to be true. Only in later centuries did the LE become increasingly more common, eventually eclipsing nearly all mss. that lacked it, even gradually leaking back into all foreign translations in the middle ages. The evidence of the manuscripts (particularly in the myriad early traditions of translation) corroborates this sequence (see sections 5.1 and 5.2). Hence "what became the majority reading in the Middle Ages started out as a minority reading," which indicates the LE was not in the original.78
Before the 4th century, none of the evidence that has been touted actually attests to the LE at all, much less as the ending of Mark, except one single reference in Irenaeus. Which by that very fact comes under suspicion. How could no one else ever show any awareness of Mark ending with the LE for nearly two whole centuries, except Irenaeus alone? Indeed, even more inexplicably, he is even represented as taking it for granted, as an undisputed fact, as if he knew of no mss. that lacked the LE and it therefore was in everyone's copy of the Bible—everyone who themselves failed to notice it. And that despite the fact that Eusebius informs us it was a rarely found reading a century later. It can easily be just luck that no one else who knew of it found occasion to reference it. But this and all other evidence still weighs against this reference in Irenaeus being authentic, and when we combine that fact with the actual evidence of that reference being a later interpolation (and thus not by the hand of Irenaeus after all), it carries force (as shown in section 5.3.5). And even if we accept that passage's authenticity, it can only establish that the LE had been appended to some copies of Mark by 185 A.D., not that it originated with Mark over a century earlier, much less was widely known as the ending of Mark. To the contrary, the most likely way no one else could know of it and the majority of mss. a century later lack it is if it did not originate with Mark.
The correct theory must explain why so many diverse mss. lack the LE—which means not just the extant ones that do (section 5.1), but all the ones we know must have (5.2, corroborated by Eusebius and Jerome: sections 5.3.10 and 5.3.12). This includes the evidence of the large number of extant mss. that append the LE to the SE, or indicate it as an uncommon reading; the evidence of the many early translation traditions, which entail that several root mss. (the mss. on which the original translations were made) lacked the LE; the evidence of the earliest extant mss. (Vaticanus and Sinaiticus in the Greek, Bobiensis and Vercellensis in the Latin), which all lack the LE; the evidence of the earliest extant Gospel texts (Matthew and Luke both follow Mark closely up to verse 6:8 but then deviate wildly, confirming that even they didn't know the LE); and the evidence of Eusebius and Jerome who both attest directly to the fact that most mss. lacked the LE (again, sections 5.3.10 and 5.3.12). Though all of this is possible if the LE were original and lost very early, all of it would be far more likely if the LE was added later. When this conclusion is combined with the internal evidence, the case against the LE is decisive.
Either way, the LE has no sound Patristic support as being original to Mark. Meanwhile the SE has no support from the Church Fathers at all (Marcus, MNT, p. 1089).

Conclusion

Kelhoffer argues (in MAM) that the LE was composed between 120-150 A.D. and possibly originated in a text other than Mark and was transferred. Other scholars have concluded the same. And I have presented considerable evidence supporting this conclusion. However, none of the evidence, even that Kelhoffer presents, establishes the conclusion that the LE had already been appended to Mark by the end of the 2nd century. As I have argued, even the testimony of Irenaeus and the Diatessaron are doubtful. However, it's certainly possible. The LE must have become appended to a copy of Mark at least by the end of the 3rd century, and there is no reason to suppose this can't have happened in the 2nd century. And whenever it occurred, all the same evidence confirms that this is indeed what happened: the LE was appended to Mark, a century or more after Mark was originally written. The style, logic, and content of the LE all demonstrate against Markan authorship, indeed decisively even by themselves, the more so together (section 4). The manuscript evidence and even the Patristic evidence strongly confirm this conclusion in every respect (section 5). And all the leading experts agree (section 3). There is therefore no rational basis for believing the LE was written by Mark. Yet it is presented as such and appears as such in the canonical Bible. The authenticity of the SE is even more indefensible. We have seen ample evidence to confirm it is a forgery, and all experts are now unanimous that it is. Thus, whether the LE or SE or both, canonical Mark contains a forgery. This conclusively proves the Bible is not inerrant but contains at least one egregious interpolation, falsely represented as original text, which can be neither true nor inspired.

Notes & References

1 Yes, some Christians actually defend the forgery as inspired (mainly Pentecostals who desperately need the snake handling pronouncement to be true), because they confess the evidence that it is a forgery is simply beyond any reasonable challenge (a noteworthy confession indeed): e.g. "'And the Signs Are Following': Mark 16.9-20—A Journey into Pentecostal Hermeneutics," John Christopher Thomas and Kimberly Ervin Alexander, Journal of Pentecostal Theology 11.2 (2003): pp. 147-170.
2 For other common examples of forgeries and interpolations, see the relevant sections of Bart Ehrman, Jesus, Interrupted (2009) and Misquoting Jesus (2005), and Paul Tobin, The Rejection of Pascal's Wager: A Skeptic's Guide to the Bible and the Historical Jesus (2006).
3 Helmut Koester, "The Text of the Synoptic Gospels in the Second Century," in William Petersen, ed., Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission (1989), pp. 19-27.
4 Quoted by Daniel Wallace, "Inspiration, Preservation, and New Testament Textual Criticism," Grace Theological Journal 12.1 (1992): p. 44 [pp. 21-50], available as of August 2009 at Ted Hildebrandt's site at Gordon College.
5 Many manuscripts omit "in their hands" in verse 18, so I have placed those words in brackets, as having a 50/50 shot at being a later addition or original to the LE. One very late manuscript even omits the entire reference to serpents, but that was later doctrinal meddling; the original LE certainly included it. Some rare manuscripts containing the SE say "Jesus himself appeared [and] sent out" the word or "Jesus himself appeared to them [and] sent out" the word, but all scholars agree these are later interpolations (the added words are missing from almost all mss., especially the oldest and the best). There are a few even rarer manuscript deviations of little importance (in both the SE and LE), but discussing them further is unnecessary for the present thesis. Note also that some scholars confusingly call the OE the SE. Here the terms will be consistently employed as I have defined them. But that doesn't mean I'm certain the OE was in fact the way the Gospel originally ended, only that among biblical experts it is most commonly (but not universally) thought to be.
6 Only one extant ms. clearly contains the SE alone (the Latin Codex Bobiensis, see section 2.2). Others that may have are ambiguous as to their original condition.
7 The ms. literally reads "truth power" (both words framing "to comprehend God's"), which is ungrammatical and thus corrupt. Some scholars suggest "and" has been dropped ("to comprehend God's truth [and] power") but the word order makes this less likely than a corruption of alêthinên into alêtheian (hence "true power").
     The full Greek of the VLE reads: kakeinoi apelogounto legontes hoti ho aiôn houtos tês anomias kai tês apistias hupo ton Satanan estin, ho mê eôn ta hupo tôn pneumatôn akatharta tên alêtheian tou theou katalabesthai dunamin, dia touto apokalupson sou tên dikaiosunên êdê. ekeinoi elegon tô Christô, kai ho Christos ekeinois proselegen hoti peplêrôtai ho horos tôn etôn tês exousias tou Satana, alla eggizei alla deina kai huper hôn egô hamartêsantôn paredothên eis thanaton hina hupostrepsôsin eis tên alêtheian kai mêketi hamartêsôsin, hina tên en tô ouranô pneumatikên kai aphtharton tês dikaiosunês doxan klêronomêsôsin. alla [poreuthentes eis ton kosmon hapanta kêruxate to euaggelion pasê tê ktisei]. The last bracketed clause is identical to the LE text.
8 Unless otherwise noted I rely on the critical edition of the Greek text provided by Barbara Aland, Bruce Metzger, et al., The Greek New Testament, 4th Revised ed. (1983). See pp. 189-92 for the endings of Mark (and p. 191, n. 6 for the VLE, on which also see C.R. Gregory, Das Freer-Logion (1908)). This includes an apparatus distinguishing which endings appear in which manuscripts (hereafter mss.), although in other cases I have found the apparatus of this edition frequently omits variants that I have personally seen even in the mss. they attest to using, which means such omissions may also exist here (so we should not assume their apparatus is complete).
9 A fact well summarized by Daniel Wallace in Black, PEM (see section 3), pp. 33-38; cf. also Darrell Bock, ibid., pp. 134-37.
10 See the Wikipedia entry on "Codex Bobiensis," with Metzger, TNT, p. 73. My translation is adapted from William Lane, The Gospel according to Mark (1974), p. 582, n. 3, and the original Latin.
      The Latin text of the BE reads: Subito autem ad horam tertiam tenebrae diei factae sunt per totum orbem terrae, et descenderunt de caelis angeli et surgent in claritate vivi Dei simul ascenderunt cum eo, et continuo lux facta est. Tunc illae accesserunt ad monimentum.
     Literally this says the angels "descended from heaven and rose up in the splendor of the living God ascended with him" but that is oddly stated and grammatically incorrect. Almost certainly the middle verb has been corrupted in transmission, from a singular to a plural. The original must have had something like surgente eo rather than sergent, and my translation reflects this. The phrase surgente eo is the ablative absolute for "as he rose up" and such a construction and sense is entirely expected here, or possibly it was sergente iu, as the ablative Iesu was often thus abbreviated among the nomina sacra even in early Latin mss., while the Greek (if this interpolation derives from a Greek source) would have used the genitive absolute (with the very same abbreviation iu), in which case this verse said "as Jesus rose up."
11 Wieland Willker, "A Textual Commentary on the Greek Gospels, Vol. 2b: The Various Endings of Mark," 6th ed. (2009), part of Wilker's extensive "Online Textual Commentary on the Greek Gospels."
12 As of October 2009 the best discussion at Wikipedia is "Mark 16:9–20 in the manuscript tradition."
13 For example, the most extensive attempt to argue the LE was and is the original ending of Mark (and thus not a forgery) is still that of William Farmer, The Last Twelve Verses of Mark (1974), but his arguments have been refuted many times over by the scholars just listed. Some of his more egregious errors had already been noted in J.N. Birdsall's review of Farmer's book in The Journal of Theological Studies (n.s.) 26 (1975): pp. 151-60. Similarly, a recent debate pitting scholars pro and con reads decisively in favor of the negative (Black, PEM; see Bock's assessment therein, pp. 124-41). And other surveys come to essentially the same conclusion, e.g. Steven Cox, A History and Critique of Scholarship Concerning the Markan Endings (1993).
14 Bruce Terry, "The Style of the Long Ending of Mark," originally published as "Another Look at the Ending of Mark," Firm Foundation 93 (14 Sept. 1976).
15 For a general discussion of the principles of stylistic forgery detection see: Donald Foster, Author Unknown: Tales of a Literary Detective (2000) and Robert Eagleson, "Forensic Analysis of Personal Written Texts: A Case Study," in Language and the Law, ed. by John Gibbons (1994): pp. 362-73.
16 See, for example, the studies of Randel Helms, Gospel Fictions (1988) and Dennis MacDonald, The Homeric Epics and the Gospel of Mark (2000).
17 Ezra Gould, Critical and Exegetical Commentary on the Gospel according to St. Mark (1896), p. 303.
18 Ezra Gould, op. cit., pp. 301-03. Analysis of several of these stylistic incongruities has been repeated by several scholars since, generating an overall consensus: e.g. Paul Danove, "Determination of the Extent of the Text of the Gospel of Mark," The End of Mark's Story: A Methodological Study (1993): pp. 119-31; James Keith Elliott, "The Text and Language of the Endings to Mark's Gospel," Theologische Zeitschrift 27.4 (July-August 1971): pp. 255-62.
19 The kakeinon used twice in Mark 12:4-5 is still a demonstrative, i.e. it references preceding nouns in each case: "he sent another slave, and that one they bashed in the head...he sent another [slave], and that one they killed" (contrast Mark 14:2-3, where "he sent a servant...and him they beat up," using auton instead of ekeinon). The author of the LE uses ekeinos (by itself) as a synonym of autos. Mark never does.
20 Black, PEM, p. 133 (Bock), pp. 30-31 (Wallace), p. 89 (Elliott, emphasizing the fact that the peculiar features occur repeatedly in the LE, but comparable deviations do not occur 'repeatedly' in any other extended section of Mark).
21 Marcus, MNT, p. 1089.
22 Bruce Terry, "Style."
23 Reported by Papias, according to Eusebius, History of the Church 3.39.9. This tale may have independently influenced the LE, but it does not reflect awareness of the LE—to the contrary, it entails ignorance of it (see section 5.3.1).
     Against Maurice Robinson's false claim that there are other elements of the LE not found in the Gospels (besides this one) see Bock's concise refutation in Black, PEM, p. 134 (esp. n. 8), although even Bock erroneously claims the 'weeping and mourning' of 16:10 is novel (it is not: see table), and misses the fact that (a) Jesus' rebuke (and proffering of evidence) in Luke 24:35-46 entails the two from Emmaus were disbelieved (as even Bock seems aware in n. 8) and (b) four of the five signs do derive from the Gospels and Acts (see table), leaving only one novel fact: the immunity to poison drink (which, as will be argued, is a logical inference from Luke 10:19, and thus not really so very novel).
24 On the early (mid-2nd century) assembly of the NT: David Trobisch, The First Edition of the New Testament (2000) and "Who Published the Christian Bible?" CSER Review 2.1 (2007), pp. 29-32. On there being over forty Gospels to choose from (over forty are known): Christopher Tuckett, "Forty Other Gospels," The Written Gospel, Markus Bockmuehl and Donald Hagner, eds. (2005): pp. 238-53. On how ancient schools taught students to summarize, paraphrase, and rewrite passages in their own voice: David Gowler, "The Chreia," The Historical Jesus in Context, Amy-Jill Levine, Dale Allison, Jr., and John Dominic Crossan, eds. (2006): pp. 132-48; Raffaella Cribiore, Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt (2001); and see Dennis MacDonald, The Homeric Epics and the Gospel of Mark (2000), pp. 4-6, for a summary account.
25 On this ironic inclusio in Mark see Richard Carrier, "The Spiritual Body of Christ and the Legend of the Empty Tomb," in Robert M. Price & Jeffery Jay Lowder, eds., The Empty Tomb: Jesus Beyond the Grave (2005: pp. 105-232), pp. 163-64.
     Mark's frequent use of irony is documented by Paul Danove, The End of Mark's Story: A Methodological Study (1993) and Jerry Camery-Hoggatt, Irony in Mark's Gospel (1992). For the OE as such: Adela Collins, "The Empty Tomb in the Gospel According to Mark," in Eleonore Stump & Thomas Flint, eds., Hermes and Athena: Biblical Exegesis and Philosophical Theology (1993), pp. 107-40.
26 See the Wikipedia entries on these mss. for more infomation: "Codex Sinaiticus" (designation: Aleph) and "Codex Vaticanus" (designation: B). Both also have project websites devoted to them: see The Sinaiticus Project and The Vaticanus Project.
27 All demonstrated by Daniel Wallace in Black, PEP, pp. 17-18. Maurice Robinson claims the Vaticanus scribe must have miscounted the number of words in the LE (in Black, PEP, p. 52 n. 44), but that's just special pleading.
28 See Wallace's discussion in Black, PEP, p. 18 n. 42.
29 Demonstrated in H.J. Milne and T.C. Skeat, Scribes and Correctors of the Codex Sinaiticus (1938), pp. 9-11; admitted reluctantly by Maurice Robinson in Black, PEP, pp. 51-52 n. 43.
30 See Wikipedia entries (cited above) for evidence and bibliography. On the debate whether they derive from the same scriptorium or even the same half of the century see Dirk Jongkind, Scribal Habits of Codex Sinaiticus (2007): pp. 18-21. Jongkind also demonstrates throughout his text that Vaticanus and Sinaticus used different exemplars. This is now generally beyond dispute. The evidence that they nevertheless derive from the same scriptorium is much weaker. They do show many second-hand corrections aligning each other, and bear other similarities, but many of these corrections were made centuries later, some as late as the 12th century (and thus do not indicate origin in the same scriptorium), and their other similarities no more indicate a common scriptorium than a common fashion among all scriptoria of the period. Even shared decorative devices at best may indicate scribes trained in the same school, but such is not entailed, as such elements were commonplace. Moreover, Vaticanus and Sinaiticus bear significant differences (e.g. they do not contain all the same books), which argues against a common origin.
31 F. Crawford Burkitt, Evangelion da-Mepharreshe: The Curetonian Version of the Four Gospels, with the Readings of the Sinai Palimpsest and the Early Syriac Patristic Evidence (1904), pp. 215-17; and the Wikipedia entries for "Syriac versions of the Bible" and "Syriac Sinaiticus.
     The conclusions in this and following paragraphs of the main text are based (only in part) on the critical apparatus provided in Aland & Metzger (cited earlier).
32 P.E. Kahle, "The End of St. Mark's Gospel: The Witness of the Coptic Versions," Journal of Theological Studies 2 (1951): pp. 49-57; Gerald Browne, "The Gospel of Mark in Fayumic Coptic," The Bulletin of the American Society of Papyrologists 13.2 (1976): pp. 41-43; and see the Wikipedia entry for "Coptic Versions of the Bible." On the recent discovery of Codex P. Palau Rib. 182 lacking the LE see Kurt Aland & Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism (2nd rev. ed., 1995), p. 202, and Hans Quecke, Das Markusevangelium saïdisch: Text der Handschrift PPalau Rib. Inv. Nr. 182 mit den Varianten der Handschrift M 569 (1972).
     Lectionary 1602 (8th century) has the Greek on one side, Sahidic on the other, and the Greek includes the LE, while the Sahidic ends with verse 16:6 (according to Aland & Aland, ibid., p. 203), but as I do not read Coptic I could not verify whether this was where the text ended or only where the damaged mss. ends. It appears to be the latter, so I consider its testimony inconclusive. Several Greek manuscripts likewise 'lack the LE' only because of lost pages and thus are of no use as evidence (minuscules 2386, 1420, 16; and even 304, the commentary text discussed in section 5.2.2, though not exhibiting actual ms. damage, nevertheless appears to be missing numerous concluding pages, exactly where a reference to the LE might appear).
33 Bruce Metzger, "The Ending of the Gospel according to Mark in Ethiopic Manuscripts," in John Reumann, ed., Understanding the Sacred Text (1972), pp. 167-80. See also: Rochus Zuurmond, Novum Testamentum Aethiopice, vol. 1.2 (1989): 44-52, and "The Ethiopic Version of the New Testament," in Bart Ehrman & Michael Holmes, eds., The Text of the New Testament in Contemporary Research (1994): 142-56; and Martin Bailey, "Discovery of Earliest Illuminated Manuscript: Revised Dating Places Garima Gospels before 650" The Art Newspaper 214 (June 2010), which transmits the findings of J. Mercier, "La peinture éthiopienne à lépoque axoumite et au XVIII siècle," Comptes-rendus des séances de l'Académie des inscriptions et belles-lettres (2000): 35-71. Notably Mercier only tested two of the illuminated pages, not the leafs with the Gospel text, and the Gospels were rebound centuries after being compiled (with pages out of order) so it is still uncertain if the Garima texts of Mark are actually as old as the pictures inserted among them.
34 Confirming this conclusion: C.H. Turner, "Did Codex Vercellensis (a) Contain the Last Twelve Verses of St. Mark?" Journal of Theological Studies 29 (1927-28): pp. 16-18; with more supporting evidence in Kurt Aland, "Bemerkungen zum Schluss des Markusevangeliums," Neotestamentica et Semitica (1969: pp. 157-80): pp. 169-78. See the remarks of Daniel Wallace in Black, PEM, pp. 24-25 n. 6, and Bruce Metzger, The Early Versions of the New Testament (1977), pp. 312-13. See also the Wikipedia entry for "Codex Vercellensis."
35 See Wikipedia entries on "Codex Bezae," "Codex Corbeiensis II," "[ http://en.wikipedia.org/wiki/Old_Latin_Bible Vetus Latina]," and "List of New Testament Latin Manuscripts." On Sangallensis see John Wordsworth, Portions of the Gospels according to St. Mark and St. Matthew (1886), pp. xxix-xxx.
     Codex Bezae actually lost the page that would have contained the Latin text of LE, and pages were added to the Codex centuries later replacing that loss with a borrowed translation from the standard Vulgate. My examination of the evidence in Frederick Scrivener, Bezae Codex Cantabrigiensis (1864) leads me to conclude that the Latin of Codex Bezae probably did contain the LE, but that this was derived largely from the Greek opposite, with knowledge of the Latin Vulgate, and thus is not an early Latin translation, but a late translation made from a late 4th century (or even later) Greek ms. containing the LE (such as we already know existed). See also Kurt Aland & Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism (2nd rev. ed., 1995), p. 189, and Bruce Metzger, The Early Versions of the New Testament (1977), pp. 317-18.
36 Metzger, TCG, pp. 122-26; J. Neville Birdsall, "The Georgian Version of the New Testament," in Bart Ehrman & Michael Holmes, eds., The Text of the New Testament in Contemporary Research (1994: pp. 173-87), pp. 178 and 180; Joseph Alexanian, "The Armenian Version of the New Testament," ibid. (pp. 157-72), p. 157.
     See also "www.armenianbible.org" and Ernest Cadman Colwell, "Mark 16:9-20 in the Armenian Version," Journal of Biblical Literature 56.4 (December 1937): pp. 369-86. Colwell provides eight converging lines of evidence establishing that the LE did not exist in Mesrop's original Armenian translation, producing a fairly decisive case. This is further supported by the evidence in Albert Edmunds, "The Six Endings of Mark in Later Manuscripts and Catholic and Protestant Imprints of the Old Armenian Version," The Monist 29 (1919): pp. 520-25.
37 On Eznik's knowledge of the LE (from a source other than the Armenian Bible), see Colwell, "Mark," p. 384. The reference appears in Eznik, On God or Sects 112, quoting Jesus (first from Luke 10:19 and then from Mark 16:17-18), implying the Gospels were his source (though he doesn't specifically say so).
38 On the Gothic translation see summary and bibliography in the entries at Wikipedia for "Ulfilas" and "Codex Argenteus."
39 See Bart Ehrman, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament (1993).
40 Metzger & Ehrman, TNT, pp. 40-41; Collins, MAC, pp. 804-06.
41 That all the same phenomena are observed in the Armenian manuscripts: Colwell, "Mark," pp. 375-78.
42 See Willker, op. cit., pp. 6-7 (note 8 above) and Kelhoffer, "The Witness," pp. 104-09 (cited below) and Edmunds, "Six Endings," p. 524.
43 Marcus, MNT, p. 1089; Collins, MAC, p. 805; cf. Kurt Aland, "Der wiedergefundene Markusschluss? Eine methodologische Bemerkung zur Textkritischen Arbeit," Zeitschrift für Theologie und Kirche 67 (1970): pp. 3-13.
44 Metzger & Ehrman, TNT, p. 324. For the evidence of L see the facsimile in John Burgon, The Last Twelve Verses of the Gospel according to S. Mark (1871), p. 126 (with the scribal notes translated on p. 123). For the SE in the marginalia of several Syriac and Coptic (Bohairic) mss. see: Clarence Russell Williams, The Appendices to the Gospel according to Mark: A Study in Textual Transmission (1915), pp. 367, 372-73, 392-95, 441 (and for ms. 274, cf. p. 418)."
45 See Colwell, "Mark," pp. 378-81.
46 On this Armenian marginal note see Colwell, "Mark," pp. 383-84.
47 Eusebius, History of the Church 3.39.14 says Papias "in his own book passes on other commentaries on the stories of the Lord from the aforementioned Aristion, as well as traditions from John the Elder," where the key phrase (tôn tou kuriou logôn diêgêseis) could actually be the title of a book (Commentaries on the Sayings [or Stories] of the Lord), or referring to such a book. Although Eusebius earlier quotes a passage (3.39.7-8) in which Papias implies he did not read the works of Aristion but asked other people about the things Aristion was saying (Aristion was evidently a contemporary), Papias only says he preferred the living word, not that he consulted it exclusively (i.e. that he preferred asking Aristion's disciples about Aristion's teachings does not mean he did not already know Aristion's teachings in writing, like the proposed Commentaries, just as Papias knew of some of the Gospels).
     It has been suggested that this Armenian scribal note refers to Aristion attesting the Barsabbas story of surviving poison (see sections 4.3.1 and 5.3.1) but the note neither contains such a remark nor is placed anywhere near verse 16:18 (where such a remark would belong). The note precedes the whole LE, was added by a later scholar (not the copyist who produced the ms.), and is so brief, there is no plausible case to be made that some prior note about Barsabbas had become corrupted into this state.
48 Maurice Robinson as quoted in The Encyclopedia of New Testament Textual Criticism, entry for "Manuscript 304" available online.
49 Noted by Elliott in Black, PEM, p. 92.
50 Bart Ehrman, Didymus the Blind and the Text of the Gospels (1986), p. 6 (cf. pp. 6-7 for discussion and references).
51 Metzger & Ehrman, TNT, p. 12 (cf. pp. 126-34 for discussion and references).
52 For a general survey of why Patristic evidence "involves the greatest difficulties and the most problems" see Bruce Metzger, "Patristic Evidence and the Textual Criticism of the New Testament" in NTS, pp. 167-88 (quoting p. 167), supported by Gordon Fee, "The Use of the Greek Fathers for New Testament Textual Criticism," The Text of the New Testament in Contemporary Research, Bart D. Ehrman and Michael Holmes, eds. (1994): pp. 191-207.
53 Metzger & Ehrman, TNT, pp. 308-09. For background see the Wikipedia entries on the "Western text-type" and the "Alexandrian text-type."
54 Eusebius, History of the Church 3.39 contains both Papias' story about Barsabbas and Papias' declaring familiarity with the Gospel of Mark, as well as all the other details mentioned.
55 James Kelhoffer makes the best case for this passage being evidence Justin knew the LE (MAM, pp. 170-75), but even his argument doesn't overcome the reasons just noted. Nevertheless, his argument is equally compatible with the conclusion that Justin knew this material from another source, not Mark's Gospel.

Some have also suggested pantachou ("everywhere") is so rare Justin can only have adapted it from the LE, but it appears so casually in Mark 1:28 and Luke 9:6 that it was clearly both a common and scripturally established idiom, and a search of the Thesaurus Linguae Graecae confirms it was a very common word in Greek literature of the early Empire. Justin himself uses it ten times in his own writings (this instance making eleven).
56 Kelhoffer is more confident than the evidence warrants (MAM, pp. 170-75). Note that earlier dates are often given for the Diatessaron, but it was most likely composed in the East, and by all accounts Tatian did not go east after his conversion until the 170's. At any rate, an earlier date of composition cannot be proved.
57 See the Wikipedia entry on "Diatessaron."
58 The error originates from mistaking Rousseau's modern 'back translation' of the Latin into Greek for an actual Greek text—and then mistaking that as deriving from Theodoret. Kelhoffer, MAM, p. 170, even presents the Greek text of 'Theodoret' as if it came from him and not Rousseau, and posits theories from the text type! Probably one of the most embarrassing errors of his career. Alas, the Greek Kelhoffer quotes is Rousseau's. I verified this myself, consulting first-hand a copy of Adelin Rousseau & Louis Doutreleau, Iréné de Lyon: Contre les hérésies livre III (1974), vol. 1 (see: pp. 64-67, 79-82, pp. 144-48) and 2 (see: pp. 128, 137-39).
59 In Minuscule 1582, according to, e.g., Maurice Robinson (in Black, PEM, p. 47 n. 26).
60 Burnett Streeter, The Four Gospels: A Study of Origins (1953), p. 124. On the origin of this marginal note in the early 5th century or after (anytime from the late 5th to 9th century is possible) see K. W. Kim, "Codices 1582, 1739, and Origen," Journal of Biblical Literature 69.2 (June 1950): 167-75. The evidence is simply that of all the sources named by the annotator (in ms. 1582 where this citation of Irenaeus appears), the latest of them date to the early 5th century (which establishes the original author added these notes to the textual tradition behind 1582 either in the late 5th century or later), which could simply reflect the annotator's limited library or preference for venerable sources (so he could still be writing even as late as the 9th century). Moreover, the annotator who compiled the bulk of these notes is not necessarily the same one who added the note referencing Irenaeus. That could have been added by anyone at any time in the intervening centuries. An identical note appears in a different location in the 11th century manuscript 72, but we cannot deduce anything useful from this (72 might be lifting that note from any manuscript related to 1582 of any possible date).
61 F. W. Hall, A Companion to Classical Texts (1913), p. 194 (pp. 193-97); Robert Renehan, Greek Textual Criticism: A Reader (1969), p. 36 (§ 32); Miroslav Marcovich, Patristic Textual Criticism (1994), s.v. "Interpolations," Index. See also: Martin West, Textual Criticism and Editorial Technique Applicable to Greek and Latin Texts (1973), p. 28; and Paul Maas (tr. by Barbara Flower), Textual Criticism (1958), pp. 34-35 (§ 33) and p. 14 (§ 16). I have personally verified numerous egregious examples, amounting to entire paragraphs, accidentally interpolated into the Weights and Measures of Epiphanius (which I presented at a conference at UC Berkeley in 2005).
62 Irenaeus, Against Heresies 2.32.4.
63 Heb. 6:2; 1 Cor. 12:8-11, 12:28-30 (cf. Mark 5:23, Luke 4:40); and Justin Martyr, Dialogue with Trypho 39. Notably, Irenaeus says Christians exhibit "in the name" of Christ the powers of God "in proportion to the gift each has received" and then lists four gifts; Justin says Christians prove the power of Jesus by "receiving gifts, each as he is worthy, illumined through the name of Christ" and then lists seven gifts; Tertullian says something similar (Against Marcion 5.8, conspicuously quoting only Paul as evidence); evidently this was a common mode of Christian preaching (most likely based on 1 Corinthians 12, cf. Romans 12:4-9, 1Corinthians 7:8, 1Corinthians 14, and Hebrews 2:4 and 6:4-6, etc.).
64 Hippolytus, Apostolic Tradition 36.1 ("The faithful shall be careful to partake of the eucharist before eating anything else. For if they eat with faith, even though some deadly poison is given to them, after this it will not be able to harm them.").
65 Kelhoffer, MAM, p. 171, n. 48 also refutes the specious suggestion that Celsus knew the LE.
66 Cyprian, The Opinions of 87 Clerics at the Seventh Council of Carthage Concerning the Baptism of Heretics 37:
Vincentius of Thibaris said: We know that heretics are worse than Gentiles. If, therefore, being converted, they should wish to come to the Lord, we have assuredly the rule of truth which the Lord by His divine precept commanded to His apostles, saying, "Go ye, lay on hands in my name, expel demons." And in another place: "Go ye and teach the nations, baptizing them in the name of the Father, of the Son, and of the Holy Ghost." Therefore first of all by imposition of hands in exorcism, secondly by the regeneration of baptism, they may then come to the promise of Christ. Otherwise I think it ought not to be done." Several other clerics at the same council likewise said heretics can only be accepted back into their church if they are exorcized by laying on hands and baptized.
The notion that heretics must be exorcised by laying on hands and baptized before being accepted back into the fold is echoed by several other clerics at the same council, but conspicuously, none cite the Lord in support of their opinion.
67 Didascalia 23.(6.8), or p. 101 of the Connolly translation. See the Wikipedia entry for "Didascalia Apostolorum."
68 See Kelhoffer, MAM, p. 171, n. 49.
69 See Kelhoffer, MAM, pp. 176-77; also "The Report of Pilate to the Emperor Claudius" at [EarlyChristianWritings.com EarlyChristianWritings.com], as well as the entire "Acts of Pilate" resource page there, and the Wikipedia entry for "Acts of Pilate."
70 James Kelhoffer, "The Witness of Eusebius' Ad Marinum and Other Christian Writings to Text-Critical Debates concerning the Original Conclusion to Mark's Gospel," Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der älteren Kirche 92 (2001): pp. 78-112 [exclusion from Canons: p. 108]. This article also shows how what Eusebius reports was a rare reading in the 4th century became the most common reading in later medieval manuscripts; and it provides an English translation of the entire Letter to Marinus with accompanying Greek text. That Eusebius was well aware of Western readings (and thus Western manuscripts) and used them on occasion (while only tending to prefer the Alexandrian text) is shown by, among others, D.S. Wallace-Hadrill, "Eusebius and the Gospel Text of Caesarea," The Harvard Theological Review 49.2 (April 1956): 105-14, so it cannot be claimed his remarks apply only to manuscripts in the Alexandrian tradition.
71 Kelhoffer, "The Witness," pp. 99-109.
72 Kelhoffer, "The Witness," pp. 104-05 (cf. also Collins, MAC, p. 805).
73 See introductions on Aphraates and Ephrem in the Nicene and Post-Nicene Fathers vol. 13. Aphraates quotes part of the LE (only a truncated version of vv. 16-17, the verses later found in the Diatessaron) in Demonstrations 1.17 (although a quotation of 16:15 is curiously absent from Demonstrations 1.8, where we would also expect it). For Ephrem see Carmel McCarthy, Saint Ephrem's Commentary on Tatian's Diatessaron: An English Translation of Chester Beatty Syriac MS 709 (Oxford University Press, 1993): p. 289 (section 19.15), where a compression of Mark 16:15 and Matthew 28:19 is quoted in a fashion resembling what we know was in some copies of the Diatessaron of that period (although again nowhere else is any material from the LE quoted or mentioned in the whole of Ephrem's commentary, and even here the only words that would derive from the LE consist of the brief and ambiguous expression "into the whole world" which is already implied by the "all nations" of Matt. 28:19). Unfortunately we know Ephrem's text has been compromised by later editors (McCarthy, pp. 31-34) and that Ephrem would have been well aware of other versions of the New Testament besides the Diatessaron (McCarthy, p. 15), so this attestation may be much less secure than is commonly supposed.
74 Ambrose of Milan, On the Holy Spirit 2.13.(151), On Repentance 1.8.(35), and other works.
75 Jerome, Against Pelagius 2.15.
76 Section 5.3.(14) of the Apostolic Constitutions clearly just summarizes Matthew, Luke and John, and in a manner conspicuously not conforming to the LE; but section 6.3.(15) directly quotes Mark 16:16 and section 8.1.(1) directly quotes Mark 16:17-18. None of these elements appears in the Didascalia. See also the Wikipedia entry for "Apostolic Constitutions."
77 Hesychius of Jerusalem, Collection of Difficulties and Solutions PG 93.1440 (5th century A.D.) simply assumes Mark ends at 16:8. Severus of Antioch, Homily 77 (5th/6th century A.D.) repeats Eusebius. See Kelhoffer, "The Witness," pp. 101-102.
78 Quoting Daniel Wallace, in Black, PEM, pp. 24.

Δεν υπάρχουν σχόλια:

Δημοσίευση σχολίου